diff --git "a/CompeteSMoE/competesmoe_versions/competesmoev9/trainer_state.json" "b/CompeteSMoE/competesmoe_versions/competesmoev9/trainer_state.json" new file mode 100644--- /dev/null +++ "b/CompeteSMoE/competesmoe_versions/competesmoev9/trainer_state.json" @@ -0,0 +1,124783 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 0.9999398785546805, + "eval_steps": 500, + "global_step": 8316, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "auxiliary_loss_clip": 0.04223623, + "auxiliary_loss_mlp": 0.02763417, + "balance_loss_clip": 2.25183964, + "balance_loss_mlp": 2.12044978, + "epoch": 0.00012024289063909097, + "flos": 24932483919360.0, + "grad_norm": 39.917004623791534, + "language_loss": 2.58671618, + "learning_rate": 0.0, + "loss": 1.90504694, + "num_input_tokens_seen": 20375, + "step": 1, + "time_per_iteration": 16.231332540512085 + }, + { + "auxiliary_loss_clip": 0.02802639, + "auxiliary_loss_mlp": 0.0194568, + "balance_loss_clip": 1.50709009, + "balance_loss_mlp": 1.50184035, + "epoch": 0.00024048578127818193, + "flos": 30664624377600.0, + "grad_norm": 55.15687571858483, + "language_loss": 1.88863051, + "learning_rate": 5.021476677069823e-07, + "loss": 1.9361136, + "num_input_tokens_seen": 39035, + "step": 2, + "time_per_iteration": 2.617116689682007 + }, + { + "auxiliary_loss_clip": 0.02801753, + "auxiliary_loss_mlp": 0.0195365, + "balance_loss_clip": 1.49962592, + "balance_loss_mlp": 1.5080936, + "epoch": 0.0003607286719172729, + "flos": 19026227969280.0, + "grad_norm": 40.06664818853955, + "language_loss": 1.61463404, + "learning_rate": 7.958852231401551e-07, + "loss": 1.66218805, + "num_input_tokens_seen": 57600, + "step": 3, + "time_per_iteration": 2.590442180633545 + }, + { + "auxiliary_loss_clip": 0.02773906, + "auxiliary_loss_mlp": 0.01747803, + "balance_loss_clip": 1.49286056, + "balance_loss_mlp": 1.34916675, + "epoch": 0.00048097156255636386, + "flos": 19316314206720.0, + "grad_norm": 37.024637275513705, + "language_loss": 1.64716244, + "learning_rate": 1.0042953354139647e-06, + "loss": 1.69237959, + "num_input_tokens_seen": 76465, + "step": 4, + "time_per_iteration": 2.5792222023010254 + }, + { + "auxiliary_loss_clip": 0.02804283, + "auxiliary_loss_mlp": 0.01844088, + "balance_loss_clip": 1.50558293, + "balance_loss_mlp": 1.43000305, + "epoch": 0.0006012144531954548, + "flos": 13991264893440.0, + "grad_norm": 54.987841433366945, + "language_loss": 1.93642855, + "learning_rate": 1.1659507774310057e-06, + "loss": 1.9829123, + "num_input_tokens_seen": 94350, + "step": 5, + "time_per_iteration": 2.8258097171783447 + }, + { + "auxiliary_loss_clip": 0.02772993, + "auxiliary_loss_mlp": 0.01980645, + "balance_loss_clip": 1.49133027, + "balance_loss_mlp": 1.53737712, + "epoch": 0.0007214573438345458, + "flos": 23148988225920.0, + "grad_norm": 46.050348324284656, + "language_loss": 1.60758328, + "learning_rate": 1.2980328908471373e-06, + "loss": 1.65511966, + "num_input_tokens_seen": 114595, + "step": 6, + "time_per_iteration": 2.838355302810669 + }, + { + "auxiliary_loss_clip": 0.02791279, + "auxiliary_loss_mlp": 0.01523342, + "balance_loss_clip": 1.63052297, + "balance_loss_mlp": 1.26394308, + "epoch": 0.0008417002344736367, + "flos": 67663246170240.0, + "grad_norm": 4.591884372945748, + "language_loss": 0.81458426, + "learning_rate": 1.4097067265369432e-06, + "loss": 0.85773045, + "num_input_tokens_seen": 179590, + "step": 7, + "time_per_iteration": 3.2428369522094727 + }, + { + "auxiliary_loss_clip": 0.02794086, + "auxiliary_loss_mlp": 0.01818376, + "balance_loss_clip": 1.49668646, + "balance_loss_mlp": 1.41478157, + "epoch": 0.0009619431251127277, + "flos": 21281381504640.0, + "grad_norm": 41.166582078856756, + "language_loss": 1.58437657, + "learning_rate": 1.506443003120947e-06, + "loss": 1.63050127, + "num_input_tokens_seen": 195090, + "step": 8, + "time_per_iteration": 2.81170654296875 + }, + { + "auxiliary_loss_clip": 0.02789494, + "auxiliary_loss_mlp": 0.01840737, + "balance_loss_clip": 1.49885249, + "balance_loss_mlp": 1.4207387, + "epoch": 0.0010821860157518186, + "flos": 23331342597120.0, + "grad_norm": 17.41433871112451, + "language_loss": 1.47545803, + "learning_rate": 1.5917704462803102e-06, + "loss": 1.52176023, + "num_input_tokens_seen": 211635, + "step": 9, + "time_per_iteration": 2.9129936695098877 + }, + { + "auxiliary_loss_clip": 0.02774752, + "auxiliary_loss_mlp": 0.01880867, + "balance_loss_clip": 1.49587822, + "balance_loss_mlp": 1.46258581, + "epoch": 0.0012024289063909096, + "flos": 17010166337280.0, + "grad_norm": 13.381670519085908, + "language_loss": 1.5311898, + "learning_rate": 1.6680984451379884e-06, + "loss": 1.57774615, + "num_input_tokens_seen": 224705, + "step": 10, + "time_per_iteration": 2.867178201675415 + }, + { + "auxiliary_loss_clip": 0.02784617, + "auxiliary_loss_mlp": 0.01894784, + "balance_loss_clip": 1.49399507, + "balance_loss_mlp": 1.45952666, + "epoch": 0.0013226717970300007, + "flos": 21288133261440.0, + "grad_norm": 13.923071485465337, + "language_loss": 1.3247838, + "learning_rate": 1.7371455188905097e-06, + "loss": 1.37157786, + "num_input_tokens_seen": 244635, + "step": 11, + "time_per_iteration": 2.8717381954193115 + }, + { + "auxiliary_loss_clip": 0.02777189, + "auxiliary_loss_mlp": 0.01963131, + "balance_loss_clip": 1.49344265, + "balance_loss_mlp": 1.5202446, + "epoch": 0.0014429146876690916, + "flos": 27237884935680.0, + "grad_norm": 10.60491618688204, + "language_loss": 1.25566864, + "learning_rate": 1.8001805585541196e-06, + "loss": 1.30307198, + "num_input_tokens_seen": 265765, + "step": 12, + "time_per_iteration": 2.903015613555908 + }, + { + "auxiliary_loss_clip": 0.0279024, + "auxiliary_loss_mlp": 0.01894817, + "balance_loss_clip": 1.50051546, + "balance_loss_mlp": 1.46871507, + "epoch": 0.0015631575783081825, + "flos": 19062174504960.0, + "grad_norm": 7.018694326815068, + "language_loss": 1.28936529, + "learning_rate": 1.8581671739548328e-06, + "loss": 1.33621585, + "num_input_tokens_seen": 283500, + "step": 13, + "time_per_iteration": 2.7735040187835693 + }, + { + "auxiliary_loss_clip": 0.02751891, + "auxiliary_loss_mlp": 0.01842519, + "balance_loss_clip": 1.4872216, + "balance_loss_mlp": 1.41870642, + "epoch": 0.0016834004689472734, + "flos": 48139473985920.0, + "grad_norm": 6.492068258444603, + "language_loss": 1.13368285, + "learning_rate": 1.9118543942439254e-06, + "loss": 1.17962694, + "num_input_tokens_seen": 305685, + "step": 14, + "time_per_iteration": 3.8906829357147217 + }, + { + "auxiliary_loss_clip": 0.02788704, + "auxiliary_loss_mlp": 0.01882412, + "balance_loss_clip": 1.49956298, + "balance_loss_mlp": 1.4591713, + "epoch": 0.0018036433595863645, + "flos": 34970026314240.0, + "grad_norm": 5.780825417571243, + "language_loss": 1.12738645, + "learning_rate": 1.961836000571161e-06, + "loss": 1.17409778, + "num_input_tokens_seen": 327340, + "step": 15, + "time_per_iteration": 3.942448377609253 + }, + { + "auxiliary_loss_clip": 0.02716476, + "auxiliary_loss_mlp": 0.01446762, + "balance_loss_clip": 1.60817969, + "balance_loss_mlp": 1.19956958, + "epoch": 0.0019238862502254555, + "flos": 59768284440960.0, + "grad_norm": 3.7873066574088994, + "language_loss": 0.647062, + "learning_rate": 2.0085906708279293e-06, + "loss": 0.68869436, + "num_input_tokens_seen": 382710, + "step": 16, + "time_per_iteration": 3.2174065113067627 + }, + { + "auxiliary_loss_clip": 0.02749933, + "auxiliary_loss_mlp": 0.01834439, + "balance_loss_clip": 1.48931098, + "balance_loss_mlp": 1.41711104, + "epoch": 0.0020441291408645466, + "flos": 20814543417600.0, + "grad_norm": 5.006286020415962, + "language_loss": 1.1594584, + "learning_rate": 2.0525099325728135e-06, + "loss": 1.205302, + "num_input_tokens_seen": 400890, + "step": 17, + "time_per_iteration": 2.8124237060546875 + }, + { + "auxiliary_loss_clip": 0.02690781, + "auxiliary_loss_mlp": 0.0140892, + "balance_loss_clip": 1.59938169, + "balance_loss_mlp": 1.16630507, + "epoch": 0.0021643720315036373, + "flos": 63857001582720.0, + "grad_norm": 3.534343575862486, + "language_loss": 0.72240222, + "learning_rate": 2.0939181139872922e-06, + "loss": 0.76339924, + "num_input_tokens_seen": 462605, + "step": 18, + "time_per_iteration": 3.2501676082611084 + }, + { + "auxiliary_loss_clip": 0.02730715, + "auxiliary_loss_mlp": 0.01774999, + "balance_loss_clip": 1.47984695, + "balance_loss_mlp": 1.37312078, + "epoch": 0.0022846149221427284, + "flos": 31284981192960.0, + "grad_norm": 6.8862703816956135, + "language_loss": 1.01652777, + "learning_rate": 2.1330868934640175e-06, + "loss": 1.06158495, + "num_input_tokens_seen": 483280, + "step": 19, + "time_per_iteration": 2.9732537269592285 + }, + { + "auxiliary_loss_clip": 0.02661002, + "auxiliary_loss_mlp": 0.01385613, + "balance_loss_clip": 1.59148812, + "balance_loss_mlp": 1.14757562, + "epoch": 0.002404857812781819, + "flos": 51083648161920.0, + "grad_norm": 3.5612502474647525, + "language_loss": 0.76441884, + "learning_rate": 2.170246112844971e-06, + "loss": 0.80488491, + "num_input_tokens_seen": 537620, + "step": 20, + "time_per_iteration": 3.0499982833862305 + }, + { + "auxiliary_loss_clip": 0.02699366, + "auxiliary_loss_mlp": 0.01752635, + "balance_loss_clip": 1.47069609, + "balance_loss_mlp": 1.35590661, + "epoch": 0.0025251007034209102, + "flos": 15815347309440.0, + "grad_norm": 3.9139481918980246, + "language_loss": 1.01619577, + "learning_rate": 2.2055919496770983e-06, + "loss": 1.06071579, + "num_input_tokens_seen": 555760, + "step": 21, + "time_per_iteration": 2.766576051712036 + }, + { + "auxiliary_loss_clip": 0.02687368, + "auxiliary_loss_mlp": 0.01765504, + "balance_loss_clip": 1.47233748, + "balance_loss_mlp": 1.36591482, + "epoch": 0.0026453435940600014, + "flos": 37851857458560.0, + "grad_norm": 3.5641627931915214, + "language_loss": 0.89678985, + "learning_rate": 2.2392931865974923e-06, + "loss": 0.94131863, + "num_input_tokens_seen": 578450, + "step": 22, + "time_per_iteration": 2.915733575820923 + }, + { + "auxiliary_loss_clip": 0.02650502, + "auxiliary_loss_mlp": 0.01669318, + "balance_loss_clip": 1.4653846, + "balance_loss_mlp": 1.287467, + "epoch": 0.002765586484699092, + "flos": 21141976821120.0, + "grad_norm": 4.108129821109611, + "language_loss": 1.01743567, + "learning_rate": 2.271496085962064e-06, + "loss": 1.0606339, + "num_input_tokens_seen": 596145, + "step": 23, + "time_per_iteration": 2.7851972579956055 + }, + { + "auxiliary_loss_clip": 0.02652598, + "auxiliary_loss_mlp": 0.01651745, + "balance_loss_clip": 1.46578538, + "balance_loss_mlp": 1.26893997, + "epoch": 0.002885829375338183, + "flos": 20667381396480.0, + "grad_norm": 6.4582959560058555, + "language_loss": 1.02577996, + "learning_rate": 2.3023282262611022e-06, + "loss": 1.06882346, + "num_input_tokens_seen": 614920, + "step": 24, + "time_per_iteration": 2.8213083744049072 + }, + { + "auxiliary_loss_clip": 0.0263075, + "auxiliary_loss_mlp": 0.01642362, + "balance_loss_clip": 1.46240592, + "balance_loss_mlp": 1.27672386, + "epoch": 0.003006072265977274, + "flos": 34823869873920.0, + "grad_norm": 3.4401190363791034, + "language_loss": 0.92404193, + "learning_rate": 2.3319015548620114e-06, + "loss": 0.96677303, + "num_input_tokens_seen": 636060, + "step": 25, + "time_per_iteration": 2.93013072013855 + }, + { + "auxiliary_loss_clip": 0.02629066, + "auxiliary_loss_mlp": 0.01632633, + "balance_loss_clip": 1.45917821, + "balance_loss_mlp": 1.26852, + "epoch": 0.003126315156616365, + "flos": 24422021118720.0, + "grad_norm": 3.1258550560001734, + "language_loss": 0.92938811, + "learning_rate": 2.3603148416618152e-06, + "loss": 0.97200513, + "num_input_tokens_seen": 655575, + "step": 26, + "time_per_iteration": 2.8071112632751465 + }, + { + "auxiliary_loss_clip": 0.02620075, + "auxiliary_loss_mlp": 0.01615799, + "balance_loss_clip": 1.4575659, + "balance_loss_mlp": 1.25340271, + "epoch": 0.003246558047255456, + "flos": 23622326674560.0, + "grad_norm": 2.417929924402407, + "language_loss": 1.01013279, + "learning_rate": 2.3876556694204647e-06, + "loss": 1.05249143, + "num_input_tokens_seen": 675730, + "step": 27, + "time_per_iteration": 2.8364737033843994 + }, + { + "auxiliary_loss_clip": 0.02607196, + "auxiliary_loss_mlp": 0.01619711, + "balance_loss_clip": 1.45517325, + "balance_loss_mlp": 1.25273681, + "epoch": 0.003366800937894547, + "flos": 17820275725440.0, + "grad_norm": 2.598426996230197, + "language_loss": 0.90710926, + "learning_rate": 2.414002061950908e-06, + "loss": 0.94937837, + "num_input_tokens_seen": 694605, + "step": 28, + "time_per_iteration": 2.758441925048828 + }, + { + "auxiliary_loss_clip": 0.02585771, + "auxiliary_loss_mlp": 0.01573786, + "balance_loss_clip": 1.443712, + "balance_loss_mlp": 1.22168982, + "epoch": 0.003487043828533638, + "flos": 24426115269120.0, + "grad_norm": 2.3143717970213125, + "language_loss": 0.99875557, + "learning_rate": 2.4394238264681557e-06, + "loss": 1.04035115, + "num_input_tokens_seen": 714340, + "step": 29, + "time_per_iteration": 2.841968297958374 + }, + { + "auxiliary_loss_clip": 0.02559813, + "auxiliary_loss_mlp": 0.01555236, + "balance_loss_clip": 1.44324303, + "balance_loss_mlp": 1.20485604, + "epoch": 0.003607286719172729, + "flos": 26140311002880.0, + "grad_norm": 2.0688736615282015, + "language_loss": 0.99650174, + "learning_rate": 2.4639836682781433e-06, + "loss": 1.03765225, + "num_input_tokens_seen": 734470, + "step": 30, + "time_per_iteration": 2.8280248641967773 + }, + { + "auxiliary_loss_clip": 0.02550068, + "auxiliary_loss_mlp": 0.0157824, + "balance_loss_clip": 1.44718742, + "balance_loss_mlp": 1.22385466, + "epoch": 0.00372752960981182, + "flos": 20593082113920.0, + "grad_norm": 2.5330450913470517, + "language_loss": 1.00228429, + "learning_rate": 2.487738122623307e-06, + "loss": 1.04356742, + "num_input_tokens_seen": 753380, + "step": 31, + "time_per_iteration": 2.8217709064483643 + }, + { + "auxiliary_loss_clip": 0.02490904, + "auxiliary_loss_mlp": 0.0151654, + "balance_loss_clip": 1.42332411, + "balance_loss_mlp": 1.18122792, + "epoch": 0.003847772500450911, + "flos": 22674608282880.0, + "grad_norm": 3.542931332731801, + "language_loss": 0.98856741, + "learning_rate": 2.510738338534912e-06, + "loss": 1.0286417, + "num_input_tokens_seen": 772105, + "step": 32, + "time_per_iteration": 2.7993133068084717 + }, + { + "auxiliary_loss_clip": 0.02431742, + "auxiliary_loss_mlp": 0.01512599, + "balance_loss_clip": 1.40411544, + "balance_loss_mlp": 1.17576146, + "epoch": 0.003968015391090002, + "flos": 17967796882560.0, + "grad_norm": 2.253383541762095, + "language_loss": 1.02578127, + "learning_rate": 2.5330307420306648e-06, + "loss": 1.06522465, + "num_input_tokens_seen": 788955, + "step": 33, + "time_per_iteration": 2.7649455070495605 + }, + { + "auxiliary_loss_clip": 0.02417696, + "auxiliary_loss_mlp": 0.01510276, + "balance_loss_clip": 1.40360785, + "balance_loss_mlp": 1.177634, + "epoch": 0.004088258281729093, + "flos": 27304103658240.0, + "grad_norm": 2.495465278045269, + "language_loss": 0.88010252, + "learning_rate": 2.554657600279796e-06, + "loss": 0.91938221, + "num_input_tokens_seen": 810230, + "step": 34, + "time_per_iteration": 2.9169676303863525 + }, + { + "auxiliary_loss_clip": 0.02391495, + "auxiliary_loss_mlp": 0.01516396, + "balance_loss_clip": 1.39910889, + "balance_loss_mlp": 1.18241882, + "epoch": 0.004208501172368184, + "flos": 23258587599360.0, + "grad_norm": 4.467245365527221, + "language_loss": 1.03427327, + "learning_rate": 2.5756575039679493e-06, + "loss": 1.0733521, + "num_input_tokens_seen": 829780, + "step": 35, + "time_per_iteration": 2.8402509689331055 + }, + { + "auxiliary_loss_clip": 0.02363994, + "auxiliary_loss_mlp": 0.0146534, + "balance_loss_clip": 1.39240789, + "balance_loss_mlp": 1.15139008, + "epoch": 0.0043287440630072746, + "flos": 17312104062720.0, + "grad_norm": 2.02821911294031, + "language_loss": 0.95176375, + "learning_rate": 2.5960657816942747e-06, + "loss": 0.99005711, + "num_input_tokens_seen": 848695, + "step": 36, + "time_per_iteration": 2.8113553524017334 + }, + { + "auxiliary_loss_clip": 0.02185789, + "auxiliary_loss_mlp": 0.01261796, + "balance_loss_clip": 1.46228266, + "balance_loss_mlp": 1.08403134, + "epoch": 0.004448986953646365, + "flos": 53092491160320.0, + "grad_norm": 1.3769573577586274, + "language_loss": 0.60954642, + "learning_rate": 2.6159148575788668e-06, + "loss": 0.64402229, + "num_input_tokens_seen": 906730, + "step": 37, + "time_per_iteration": 3.2650437355041504 + }, + { + "auxiliary_loss_clip": 0.0233069, + "auxiliary_loss_mlp": 0.01479655, + "balance_loss_clip": 1.38548899, + "balance_loss_mlp": 1.16913843, + "epoch": 0.004569229844285457, + "flos": 13444165866240.0, + "grad_norm": 7.688957957075786, + "language_loss": 0.98710346, + "learning_rate": 2.635234561171e-06, + "loss": 1.0252068, + "num_input_tokens_seen": 925125, + "step": 38, + "time_per_iteration": 2.8179726600646973 + }, + { + "auxiliary_loss_clip": 0.02308251, + "auxiliary_loss_mlp": 0.01391066, + "balance_loss_clip": 1.38080645, + "balance_loss_mlp": 1.0992415, + "epoch": 0.0046894727349245475, + "flos": 16209609966720.0, + "grad_norm": 2.3217026336583046, + "language_loss": 0.94282508, + "learning_rate": 2.6540523970949877e-06, + "loss": 0.97981822, + "num_input_tokens_seen": 939970, + "step": 39, + "time_per_iteration": 2.8205316066741943 + }, + { + "auxiliary_loss_clip": 0.02292014, + "auxiliary_loss_mlp": 0.01453504, + "balance_loss_clip": 1.37522054, + "balance_loss_mlp": 1.15004444, + "epoch": 0.004809715625563638, + "flos": 23914244505600.0, + "grad_norm": 6.586090051991614, + "language_loss": 0.92575103, + "learning_rate": 2.6723937805519533e-06, + "loss": 0.96320617, + "num_input_tokens_seen": 957470, + "step": 40, + "time_per_iteration": 3.745347023010254 + }, + { + "auxiliary_loss_clip": 0.02262528, + "auxiliary_loss_mlp": 0.01355701, + "balance_loss_clip": 1.36071086, + "balance_loss_mlp": 1.07932615, + "epoch": 0.00492995851620273, + "flos": 20773030273920.0, + "grad_norm": 2.258692863091866, + "language_loss": 0.93113679, + "learning_rate": 2.690282243737839e-06, + "loss": 0.96731913, + "num_input_tokens_seen": 976405, + "step": 41, + "time_per_iteration": 4.689307928085327 + }, + { + "auxiliary_loss_clip": 0.02245086, + "auxiliary_loss_mlp": 0.01397255, + "balance_loss_clip": 1.36502683, + "balance_loss_mlp": 1.1077199, + "epoch": 0.0050502014068418205, + "flos": 20338655103360.0, + "grad_norm": 3.0668836195433546, + "language_loss": 0.99495584, + "learning_rate": 2.7077396173840807e-06, + "loss": 1.03137922, + "num_input_tokens_seen": 994690, + "step": 42, + "time_per_iteration": 2.851630449295044 + }, + { + "auxiliary_loss_clip": 0.02226928, + "auxiliary_loss_mlp": 0.01408361, + "balance_loss_clip": 1.35960042, + "balance_loss_mlp": 1.11615515, + "epoch": 0.005170444297480911, + "flos": 25994872834560.0, + "grad_norm": 2.3039653245756866, + "language_loss": 0.92912245, + "learning_rate": 2.7247861909342594e-06, + "loss": 0.96547532, + "num_input_tokens_seen": 1015615, + "step": 43, + "time_per_iteration": 2.854069709777832 + }, + { + "auxiliary_loss_clip": 0.02218582, + "auxiliary_loss_mlp": 0.01399186, + "balance_loss_clip": 1.35394084, + "balance_loss_mlp": 1.11365628, + "epoch": 0.005290687188120003, + "flos": 20954055841920.0, + "grad_norm": 2.48211104897485, + "language_loss": 0.83154172, + "learning_rate": 2.7414408543044743e-06, + "loss": 0.86771935, + "num_input_tokens_seen": 1031255, + "step": 44, + "time_per_iteration": 2.860588788986206 + }, + { + "auxiliary_loss_clip": 0.02196664, + "auxiliary_loss_mlp": 0.01355682, + "balance_loss_clip": 1.34909332, + "balance_loss_mlp": 1.0806427, + "epoch": 0.005410930078759093, + "flos": 15851401585920.0, + "grad_norm": 5.791189946118453, + "language_loss": 0.7953738, + "learning_rate": 2.7577212237113157e-06, + "loss": 0.83089721, + "num_input_tokens_seen": 1048295, + "step": 45, + "time_per_iteration": 2.784562110900879 + }, + { + "auxiliary_loss_clip": 0.02177126, + "auxiliary_loss_mlp": 0.01362294, + "balance_loss_clip": 1.34673345, + "balance_loss_mlp": 1.10098732, + "epoch": 0.005531172969398184, + "flos": 21104988791040.0, + "grad_norm": 1.9008910048469272, + "language_loss": 1.04576826, + "learning_rate": 2.7736437536690466e-06, + "loss": 1.08116245, + "num_input_tokens_seen": 1067925, + "step": 46, + "time_per_iteration": 2.821108818054199 + }, + { + "auxiliary_loss_clip": 0.02158084, + "auxiliary_loss_mlp": 0.01331259, + "balance_loss_clip": 1.33289957, + "balance_loss_mlp": 1.07128739, + "epoch": 0.005651415860037276, + "flos": 20844887431680.0, + "grad_norm": 2.7626408599015058, + "language_loss": 1.08044457, + "learning_rate": 2.789223836941131e-06, + "loss": 1.11533809, + "num_input_tokens_seen": 1088060, + "step": 47, + "time_per_iteration": 2.8513307571411133 + }, + { + "auxiliary_loss_clip": 0.02130846, + "auxiliary_loss_mlp": 0.01355103, + "balance_loss_clip": 1.32841074, + "balance_loss_mlp": 1.09932756, + "epoch": 0.005771658750676366, + "flos": 13260195383040.0, + "grad_norm": 3.706042419012275, + "language_loss": 1.08738327, + "learning_rate": 2.8044758939680847e-06, + "loss": 1.12224293, + "num_input_tokens_seen": 1104130, + "step": 48, + "time_per_iteration": 2.7962374687194824 + }, + { + "auxiliary_loss_clip": 0.02131902, + "auxiliary_loss_mlp": 0.0135731, + "balance_loss_clip": 1.33328438, + "balance_loss_mlp": 1.1001997, + "epoch": 0.005891901641315457, + "flos": 24425396997120.0, + "grad_norm": 2.7857523772031847, + "language_loss": 1.02248216, + "learning_rate": 2.8194134530738863e-06, + "loss": 1.05737436, + "num_input_tokens_seen": 1122900, + "step": 49, + "time_per_iteration": 2.9043588638305664 + }, + { + "auxiliary_loss_clip": 0.021177, + "auxiliary_loss_mlp": 0.01342929, + "balance_loss_clip": 1.33270574, + "balance_loss_mlp": 1.10431981, + "epoch": 0.006012144531954548, + "flos": 23076197314560.0, + "grad_norm": 4.596196608255372, + "language_loss": 0.90385902, + "learning_rate": 2.834049222568994e-06, + "loss": 0.93846524, + "num_input_tokens_seen": 1140250, + "step": 50, + "time_per_iteration": 2.8706612586975098 + }, + { + "auxiliary_loss_clip": 0.02098522, + "auxiliary_loss_mlp": 0.01327911, + "balance_loss_clip": 1.32483566, + "balance_loss_mlp": 1.08453321, + "epoch": 0.006132387422593639, + "flos": 22528775064960.0, + "grad_norm": 1.8672425470623464, + "language_loss": 0.92674339, + "learning_rate": 2.848395155712969e-06, + "loss": 0.96100771, + "num_input_tokens_seen": 1160470, + "step": 51, + "time_per_iteration": 2.8483762741088867 + }, + { + "auxiliary_loss_clip": 0.02092292, + "auxiliary_loss_mlp": 0.01313485, + "balance_loss_clip": 1.32747495, + "balance_loss_mlp": 1.08307743, + "epoch": 0.00625263031323273, + "flos": 27628340751360.0, + "grad_norm": 3.1869505787408086, + "language_loss": 0.98127097, + "learning_rate": 2.8624625093687977e-06, + "loss": 1.01532876, + "num_input_tokens_seen": 1177605, + "step": 52, + "time_per_iteration": 2.8741133213043213 + }, + { + "auxiliary_loss_clip": 0.02071254, + "auxiliary_loss_mlp": 0.01313026, + "balance_loss_clip": 1.31911874, + "balance_loss_mlp": 1.0849067, + "epoch": 0.006372873203871821, + "flos": 23110671392640.0, + "grad_norm": 4.869864160071273, + "language_loss": 0.89042723, + "learning_rate": 2.876261897070029e-06, + "loss": 0.92427003, + "num_input_tokens_seen": 1197735, + "step": 53, + "time_per_iteration": 2.793262481689453 + }, + { + "auxiliary_loss_clip": 0.02083824, + "auxiliary_loss_mlp": 0.01337346, + "balance_loss_clip": 1.31994808, + "balance_loss_mlp": 1.09978616, + "epoch": 0.006493116094510912, + "flos": 22856028900480.0, + "grad_norm": 2.324910407338429, + "language_loss": 0.9245429, + "learning_rate": 2.889803337127447e-06, + "loss": 0.95875454, + "num_input_tokens_seen": 1216335, + "step": 54, + "time_per_iteration": 2.8237786293029785 + }, + { + "auxiliary_loss_clip": 0.02062581, + "auxiliary_loss_mlp": 0.0134203, + "balance_loss_clip": 1.31862533, + "balance_loss_mlp": 1.10094082, + "epoch": 0.006613358985150003, + "flos": 23071708114560.0, + "grad_norm": 2.7231838909469284, + "language_loss": 0.8472234, + "learning_rate": 2.903096296321516e-06, + "loss": 0.88126945, + "num_input_tokens_seen": 1234480, + "step": 55, + "time_per_iteration": 2.781245708465576 + }, + { + "auxiliary_loss_clip": 0.02047308, + "auxiliary_loss_mlp": 0.0130435, + "balance_loss_clip": 1.31794286, + "balance_loss_mlp": 1.08567297, + "epoch": 0.006733601875789094, + "flos": 26537662229760.0, + "grad_norm": 2.107508991079847, + "language_loss": 0.9191826, + "learning_rate": 2.9161497296578907e-06, + "loss": 0.95269918, + "num_input_tokens_seen": 1253870, + "step": 56, + "time_per_iteration": 2.8688066005706787 + }, + { + "auxiliary_loss_clip": 0.02026103, + "auxiliary_loss_mlp": 0.01294823, + "balance_loss_clip": 1.3041327, + "balance_loss_mlp": 1.07080531, + "epoch": 0.006853844766428185, + "flos": 15523178083200.0, + "grad_norm": 2.586444242037837, + "language_loss": 0.86095452, + "learning_rate": 2.928972116604173e-06, + "loss": 0.89416379, + "num_input_tokens_seen": 1270145, + "step": 57, + "time_per_iteration": 2.7514448165893555 + }, + { + "auxiliary_loss_clip": 0.02033069, + "auxiliary_loss_mlp": 0.01295784, + "balance_loss_clip": 1.30801511, + "balance_loss_mlp": 1.08635771, + "epoch": 0.006974087657067276, + "flos": 24243760897920.0, + "grad_norm": 2.2606741688673853, + "language_loss": 1.02155423, + "learning_rate": 2.9415714941751377e-06, + "loss": 1.05484271, + "num_input_tokens_seen": 1291365, + "step": 58, + "time_per_iteration": 2.8948347568511963 + }, + { + "auxiliary_loss_clip": 0.02018465, + "auxiliary_loss_mlp": 0.01340475, + "balance_loss_clip": 1.30330086, + "balance_loss_mlp": 1.11998522, + "epoch": 0.007094330547706367, + "flos": 25772513690880.0, + "grad_norm": 1.896149776039743, + "language_loss": 0.935812, + "learning_rate": 2.9539554871897396e-06, + "loss": 0.96940136, + "num_input_tokens_seen": 1311535, + "step": 59, + "time_per_iteration": 2.8322672843933105 + }, + { + "auxiliary_loss_clip": 0.01999226, + "auxiliary_loss_mlp": 0.01288016, + "balance_loss_clip": 1.30240142, + "balance_loss_mlp": 1.0919404, + "epoch": 0.007214573438345458, + "flos": 21319015979520.0, + "grad_norm": 2.2565043574207913, + "language_loss": 0.97484267, + "learning_rate": 2.9661313359851253e-06, + "loss": 1.00771511, + "num_input_tokens_seen": 1329420, + "step": 60, + "time_per_iteration": 2.822392702102661 + }, + { + "auxiliary_loss_clip": 0.01992725, + "auxiliary_loss_mlp": 0.01303008, + "balance_loss_clip": 1.29858661, + "balance_loss_mlp": 1.10044789, + "epoch": 0.007334816328984549, + "flos": 24937088192640.0, + "grad_norm": 2.034692532465447, + "language_loss": 0.93994868, + "learning_rate": 2.978105921839922e-06, + "loss": 0.97290605, + "num_input_tokens_seen": 1349965, + "step": 61, + "time_per_iteration": 2.8371427059173584 + }, + { + "auxiliary_loss_clip": 0.01971965, + "auxiliary_loss_mlp": 0.01270543, + "balance_loss_clip": 1.29293084, + "balance_loss_mlp": 1.07179701, + "epoch": 0.00745505921962364, + "flos": 18510586277760.0, + "grad_norm": 2.806797486479977, + "language_loss": 0.72355962, + "learning_rate": 2.9898857903302893e-06, + "loss": 0.75598472, + "num_input_tokens_seen": 1368915, + "step": 62, + "time_per_iteration": 2.816227912902832 + }, + { + "auxiliary_loss_clip": 0.01970909, + "auxiliary_loss_mlp": 0.01288995, + "balance_loss_clip": 1.29536808, + "balance_loss_mlp": 1.09740138, + "epoch": 0.007575302110262731, + "flos": 18477656484480.0, + "grad_norm": 2.7177862435754414, + "language_loss": 0.8834089, + "learning_rate": 3.001477172817253e-06, + "loss": 0.91600794, + "num_input_tokens_seen": 1386805, + "step": 63, + "time_per_iteration": 2.8601605892181396 + }, + { + "auxiliary_loss_clip": 0.01963218, + "auxiliary_loss_mlp": 0.01282016, + "balance_loss_clip": 1.29464066, + "balance_loss_mlp": 1.09643137, + "epoch": 0.007695545000901822, + "flos": 24973178382720.0, + "grad_norm": 2.9282034885741366, + "language_loss": 0.96078098, + "learning_rate": 3.012886006241894e-06, + "loss": 0.99323332, + "num_input_tokens_seen": 1406190, + "step": 64, + "time_per_iteration": 2.966465950012207 + }, + { + "auxiliary_loss_clip": 0.01939898, + "auxiliary_loss_mlp": 0.01272957, + "balance_loss_clip": 1.28693104, + "balance_loss_mlp": 1.0954783, + "epoch": 0.007815787891540913, + "flos": 21324223451520.0, + "grad_norm": 2.189189386868844, + "language_loss": 0.88356936, + "learning_rate": 3.0241179513858383e-06, + "loss": 0.91569793, + "num_input_tokens_seen": 1425500, + "step": 65, + "time_per_iteration": 2.861149311065674 + }, + { + "auxiliary_loss_clip": 0.01920437, + "auxiliary_loss_mlp": 0.01266561, + "balance_loss_clip": 1.28107762, + "balance_loss_mlp": 1.07611203, + "epoch": 0.007936030782180003, + "flos": 21575777374080.0, + "grad_norm": 2.1058098708451243, + "language_loss": 0.87943953, + "learning_rate": 3.035178409737647e-06, + "loss": 0.91130948, + "num_input_tokens_seen": 1442950, + "step": 66, + "time_per_iteration": 2.903911828994751 + }, + { + "auxiliary_loss_clip": 0.01909571, + "auxiliary_loss_mlp": 0.01244222, + "balance_loss_clip": 1.27440107, + "balance_loss_mlp": 1.07713795, + "epoch": 0.008056273672819095, + "flos": 20120785159680.0, + "grad_norm": 2.730689325212029, + "language_loss": 0.8872366, + "learning_rate": 3.046072539090907e-06, + "loss": 0.9187746, + "num_input_tokens_seen": 1460915, + "step": 67, + "time_per_iteration": 5.52929949760437 + }, + { + "auxiliary_loss_clip": 0.01891067, + "auxiliary_loss_mlp": 0.01263781, + "balance_loss_clip": 1.2763617, + "balance_loss_mlp": 1.09612513, + "epoch": 0.008176516563458186, + "flos": 18333116156160.0, + "grad_norm": 2.423526666845292, + "language_loss": 1.04947567, + "learning_rate": 3.056805267986779e-06, + "loss": 1.08102417, + "num_input_tokens_seen": 1478385, + "step": 68, + "time_per_iteration": 3.8001132011413574 + }, + { + "auxiliary_loss_clip": 0.01878643, + "auxiliary_loss_mlp": 0.01252979, + "balance_loss_clip": 1.26958108, + "balance_loss_mlp": 1.09381104, + "epoch": 0.008296759454097276, + "flos": 21872076664320.0, + "grad_norm": 2.2501614936712975, + "language_loss": 0.95346534, + "learning_rate": 3.0673813091022194e-06, + "loss": 0.98478162, + "num_input_tokens_seen": 1497605, + "step": 69, + "time_per_iteration": 2.811068058013916 + }, + { + "auxiliary_loss_clip": 0.01729984, + "auxiliary_loss_mlp": 0.01241845, + "balance_loss_clip": 1.3376143, + "balance_loss_mlp": 1.16669583, + "epoch": 0.008417002344736368, + "flos": 63408228036480.0, + "grad_norm": 1.2816250767166242, + "language_loss": 0.61995679, + "learning_rate": 3.0778051716749317e-06, + "loss": 0.64967507, + "num_input_tokens_seen": 1561150, + "step": 70, + "time_per_iteration": 3.4099793434143066 + }, + { + "auxiliary_loss_clip": 0.01834004, + "auxiliary_loss_mlp": 0.01238683, + "balance_loss_clip": 1.2567215, + "balance_loss_mlp": 1.08418751, + "epoch": 0.008537245235375458, + "flos": 22966454286720.0, + "grad_norm": 2.1354380980174175, + "language_loss": 0.90407121, + "learning_rate": 3.0880811730470094e-06, + "loss": 0.93479806, + "num_input_tokens_seen": 1580605, + "step": 71, + "time_per_iteration": 2.845784902572632 + }, + { + "auxiliary_loss_clip": 0.01696986, + "auxiliary_loss_mlp": 0.01185729, + "balance_loss_clip": 1.3269105, + "balance_loss_mlp": 1.1178273, + "epoch": 0.008657488126014549, + "flos": 61984046712960.0, + "grad_norm": 1.1405797277901175, + "language_loss": 0.58579576, + "learning_rate": 3.098213449401257e-06, + "loss": 0.61462289, + "num_input_tokens_seen": 1647535, + "step": 72, + "time_per_iteration": 3.3663692474365234 + }, + { + "auxiliary_loss_clip": 0.01820124, + "auxiliary_loss_mlp": 0.01217134, + "balance_loss_clip": 1.24857807, + "balance_loss_mlp": 1.07322478, + "epoch": 0.00877773101665364, + "flos": 30296791152000.0, + "grad_norm": 2.274691798135859, + "language_loss": 0.98814833, + "learning_rate": 3.1082059657570015e-06, + "loss": 1.01852095, + "num_input_tokens_seen": 1666770, + "step": 73, + "time_per_iteration": 2.8274106979370117 + }, + { + "auxiliary_loss_clip": 0.01805149, + "auxiliary_loss_mlp": 0.01217176, + "balance_loss_clip": 1.24877906, + "balance_loss_mlp": 1.07364845, + "epoch": 0.00889797390729273, + "flos": 23514056104320.0, + "grad_norm": 2.1129405894693063, + "language_loss": 0.96707904, + "learning_rate": 3.1180625252858496e-06, + "loss": 0.99730229, + "num_input_tokens_seen": 1685200, + "step": 74, + "time_per_iteration": 2.875420093536377 + }, + { + "auxiliary_loss_clip": 0.01784901, + "auxiliary_loss_mlp": 0.01216409, + "balance_loss_clip": 1.24410236, + "balance_loss_mlp": 1.07421625, + "epoch": 0.009018216797931822, + "flos": 23075838178560.0, + "grad_norm": 2.882351569919797, + "language_loss": 0.80470383, + "learning_rate": 3.1277867780021663e-06, + "loss": 0.83471692, + "num_input_tokens_seen": 1701835, + "step": 75, + "time_per_iteration": 2.8622748851776123 + }, + { + "auxiliary_loss_clip": 0.01766712, + "auxiliary_loss_mlp": 0.01204499, + "balance_loss_clip": 1.2335304, + "balance_loss_mlp": 1.07851851, + "epoch": 0.009138459688570914, + "flos": 15918877284480.0, + "grad_norm": 1.8948177559630652, + "language_loss": 0.95655322, + "learning_rate": 3.1373822288779824e-06, + "loss": 0.9862653, + "num_input_tokens_seen": 1718415, + "step": 76, + "time_per_iteration": 2.778977632522583 + }, + { + "auxiliary_loss_clip": 0.01771605, + "auxiliary_loss_mlp": 0.01271837, + "balance_loss_clip": 1.24195695, + "balance_loss_mlp": 1.12201464, + "epoch": 0.009258702579210003, + "flos": 27016531372800.0, + "grad_norm": 1.9910490191890653, + "language_loss": 0.79453474, + "learning_rate": 3.1468522454274533e-06, + "loss": 0.82496917, + "num_input_tokens_seen": 1738770, + "step": 77, + "time_per_iteration": 2.8348982334136963 + }, + { + "auxiliary_loss_clip": 0.0175374, + "auxiliary_loss_mlp": 0.01201536, + "balance_loss_clip": 1.23369575, + "balance_loss_mlp": 1.0680213, + "epoch": 0.009378945469849095, + "flos": 26903196984960.0, + "grad_norm": 1.988703161748972, + "language_loss": 0.91728735, + "learning_rate": 3.15620006480197e-06, + "loss": 0.94684005, + "num_input_tokens_seen": 1758040, + "step": 78, + "time_per_iteration": 2.7954719066619873 + }, + { + "auxiliary_loss_clip": 0.01739723, + "auxiliary_loss_mlp": 0.0123933, + "balance_loss_clip": 1.23104227, + "balance_loss_mlp": 1.10304999, + "epoch": 0.009499188360488187, + "flos": 35694236327040.0, + "grad_norm": 2.5968048839078777, + "language_loss": 0.74764526, + "learning_rate": 3.1654288004333087e-06, + "loss": 0.77743578, + "num_input_tokens_seen": 1776705, + "step": 79, + "time_per_iteration": 2.90067458152771 + }, + { + "auxiliary_loss_clip": 0.0173181, + "auxiliary_loss_mlp": 0.01194285, + "balance_loss_clip": 1.228459, + "balance_loss_mlp": 1.07459903, + "epoch": 0.009619431251127276, + "flos": 21503201944320.0, + "grad_norm": 2.7388527820802815, + "language_loss": 0.75929314, + "learning_rate": 3.1745414482589353e-06, + "loss": 0.78855407, + "num_input_tokens_seen": 1795915, + "step": 80, + "time_per_iteration": 2.845170021057129 + }, + { + "auxiliary_loss_clip": 0.0171174, + "auxiliary_loss_mlp": 0.01179781, + "balance_loss_clip": 1.2247057, + "balance_loss_mlp": 1.06066716, + "epoch": 0.009739674141766368, + "flos": 17421056991360.0, + "grad_norm": 3.6996124442663723, + "language_loss": 0.87095404, + "learning_rate": 3.1835408925606204e-06, + "loss": 0.89986926, + "num_input_tokens_seen": 1814055, + "step": 81, + "time_per_iteration": 2.8215763568878174 + }, + { + "auxiliary_loss_clip": 0.01710046, + "auxiliary_loss_mlp": 0.01186398, + "balance_loss_clip": 1.22452927, + "balance_loss_mlp": 1.0736258, + "epoch": 0.00985991703240546, + "flos": 27527109246720.0, + "grad_norm": 2.3191111395349435, + "language_loss": 0.89220113, + "learning_rate": 3.1924299114448214e-06, + "loss": 0.92116559, + "num_input_tokens_seen": 1834535, + "step": 82, + "time_per_iteration": 2.8716423511505127 + }, + { + "auxiliary_loss_clip": 0.01703122, + "auxiliary_loss_mlp": 0.01197441, + "balance_loss_clip": 1.22007322, + "balance_loss_mlp": 1.07513213, + "epoch": 0.00998015992304455, + "flos": 13808084509440.0, + "grad_norm": 5.414886245056712, + "language_loss": 0.83443362, + "learning_rate": 3.2012111819909055e-06, + "loss": 0.86343932, + "num_input_tokens_seen": 1851865, + "step": 83, + "time_per_iteration": 2.8668723106384277 + }, + { + "auxiliary_loss_clip": 0.01701263, + "auxiliary_loss_mlp": 0.01196403, + "balance_loss_clip": 1.2201215, + "balance_loss_mlp": 1.0771935, + "epoch": 0.010100402813683641, + "flos": 20191385341440.0, + "grad_norm": 5.01376589826675, + "language_loss": 0.94967431, + "learning_rate": 3.2098872850910627e-06, + "loss": 0.97865099, + "num_input_tokens_seen": 1868540, + "step": 84, + "time_per_iteration": 2.800844669342041 + }, + { + "auxiliary_loss_clip": 0.01694012, + "auxiliary_loss_mlp": 0.01191295, + "balance_loss_clip": 1.21967745, + "balance_loss_mlp": 1.08019185, + "epoch": 0.010220645704322733, + "flos": 17201642762880.0, + "grad_norm": 2.148799265386921, + "language_loss": 0.89274269, + "learning_rate": 3.2184607100038194e-06, + "loss": 0.92159581, + "num_input_tokens_seen": 1887180, + "step": 85, + "time_per_iteration": 2.799372673034668 + }, + { + "auxiliary_loss_clip": 0.01680011, + "auxiliary_loss_mlp": 0.01180667, + "balance_loss_clip": 1.21503663, + "balance_loss_mlp": 1.06956446, + "epoch": 0.010340888594961822, + "flos": 21470415805440.0, + "grad_norm": 3.970811256761606, + "language_loss": 0.9311657, + "learning_rate": 3.2269338586412414e-06, + "loss": 0.95977247, + "num_input_tokens_seen": 1904765, + "step": 86, + "time_per_iteration": 2.8377647399902344 + }, + { + "auxiliary_loss_clip": 0.01677953, + "auxiliary_loss_mlp": 0.01189278, + "balance_loss_clip": 1.21640086, + "balance_loss_mlp": 1.08518481, + "epoch": 0.010461131485600914, + "flos": 23002831785600.0, + "grad_norm": 2.5516737988344658, + "language_loss": 0.96542799, + "learning_rate": 3.2353090496083106e-06, + "loss": 0.99410027, + "num_input_tokens_seen": 1922600, + "step": 87, + "time_per_iteration": 2.932950019836426 + }, + { + "auxiliary_loss_clip": 0.01662784, + "auxiliary_loss_mlp": 0.01193088, + "balance_loss_clip": 1.20793819, + "balance_loss_mlp": 1.09175992, + "epoch": 0.010581374376240005, + "flos": 33546850571520.0, + "grad_norm": 2.452959729269066, + "language_loss": 0.81239879, + "learning_rate": 3.2435885220114572e-06, + "loss": 0.84095752, + "num_input_tokens_seen": 1943950, + "step": 88, + "time_per_iteration": 2.898566246032715 + }, + { + "auxiliary_loss_clip": 0.01656566, + "auxiliary_loss_mlp": 0.01179903, + "balance_loss_clip": 1.2110858, + "balance_loss_mlp": 1.08300948, + "epoch": 0.010701617266879095, + "flos": 21763087822080.0, + "grad_norm": 4.845094437822522, + "language_loss": 0.93798709, + "learning_rate": 3.2517744390519113e-06, + "loss": 0.96635175, + "num_input_tokens_seen": 1962815, + "step": 89, + "time_per_iteration": 2.782777786254883 + }, + { + "auxiliary_loss_clip": 0.01655023, + "auxiliary_loss_mlp": 0.0115731, + "balance_loss_clip": 1.20590913, + "balance_loss_mlp": 1.06389737, + "epoch": 0.010821860157518187, + "flos": 19060199256960.0, + "grad_norm": 2.1065955076799217, + "language_loss": 0.75225914, + "learning_rate": 3.259868891418298e-06, + "loss": 0.78038245, + "num_input_tokens_seen": 1980580, + "step": 90, + "time_per_iteration": 2.78857159614563 + }, + { + "auxiliary_loss_clip": 0.01653484, + "auxiliary_loss_mlp": 0.01175123, + "balance_loss_clip": 1.21142256, + "balance_loss_mlp": 1.07913554, + "epoch": 0.010942103048157278, + "flos": 25447378757760.0, + "grad_norm": 2.1149529860759384, + "language_loss": 0.84781408, + "learning_rate": 3.2678739004917757e-06, + "loss": 0.87610018, + "num_input_tokens_seen": 2000315, + "step": 91, + "time_per_iteration": 2.8099427223205566 + }, + { + "auxiliary_loss_clip": 0.01654825, + "auxiliary_loss_mlp": 0.01186354, + "balance_loss_clip": 1.21032786, + "balance_loss_mlp": 1.09461045, + "epoch": 0.011062345938796368, + "flos": 27493928058240.0, + "grad_norm": 1.6213201484364208, + "language_loss": 0.92173457, + "learning_rate": 3.275791421376029e-06, + "loss": 0.95014632, + "num_input_tokens_seen": 2023760, + "step": 92, + "time_per_iteration": 2.9584014415740967 + }, + { + "auxiliary_loss_clip": 0.01637272, + "auxiliary_loss_mlp": 0.01161917, + "balance_loss_clip": 1.20041323, + "balance_loss_mlp": 1.07670593, + "epoch": 0.01118258882943546, + "flos": 16071210864000.0, + "grad_norm": 2.107458332359186, + "language_loss": 0.96272171, + "learning_rate": 3.2836233457634622e-06, + "loss": 0.9907136, + "num_input_tokens_seen": 2041895, + "step": 93, + "time_per_iteration": 5.544355869293213 + }, + { + "auxiliary_loss_clip": 0.01631321, + "auxiliary_loss_mlp": 0.01186187, + "balance_loss_clip": 1.198982, + "balance_loss_mlp": 1.08771968, + "epoch": 0.011302831720074551, + "flos": 20668602458880.0, + "grad_norm": 2.6459537151840835, + "language_loss": 0.85516262, + "learning_rate": 3.2913715046481135e-06, + "loss": 0.88333774, + "num_input_tokens_seen": 2061640, + "step": 94, + "time_per_iteration": 3.8424623012542725 + }, + { + "auxiliary_loss_clip": 0.01626913, + "auxiliary_loss_mlp": 0.01157404, + "balance_loss_clip": 1.19884074, + "balance_loss_mlp": 1.07085836, + "epoch": 0.011423074610713641, + "flos": 13072238490240.0, + "grad_norm": 2.3200604150777995, + "language_loss": 0.88868833, + "learning_rate": 3.299037670895023e-06, + "loss": 0.91653156, + "num_input_tokens_seen": 2078255, + "step": 95, + "time_per_iteration": 2.769068956375122 + }, + { + "auxiliary_loss_clip": 0.01618645, + "auxiliary_loss_mlp": 0.01179936, + "balance_loss_clip": 1.1966815, + "balance_loss_mlp": 1.09482062, + "epoch": 0.011543317501352733, + "flos": 30335646689280.0, + "grad_norm": 3.0149522268810247, + "language_loss": 0.80281776, + "learning_rate": 3.3066235616750667e-06, + "loss": 0.83080363, + "num_input_tokens_seen": 2099490, + "step": 96, + "time_per_iteration": 2.8867015838623047 + }, + { + "auxiliary_loss_clip": 0.01613856, + "auxiliary_loss_mlp": 0.01157627, + "balance_loss_clip": 1.19869709, + "balance_loss_mlp": 1.07665944, + "epoch": 0.011663560391991824, + "flos": 15522962601600.0, + "grad_norm": 2.074021362824894, + "language_loss": 0.92351449, + "learning_rate": 3.3141308407736276e-06, + "loss": 0.95122933, + "num_input_tokens_seen": 2116125, + "step": 97, + "time_per_iteration": 2.8195738792419434 + }, + { + "auxiliary_loss_clip": 0.01613271, + "auxiliary_loss_mlp": 0.01166407, + "balance_loss_clip": 1.19244289, + "balance_loss_mlp": 1.08515429, + "epoch": 0.011783803282630914, + "flos": 19902125116800.0, + "grad_norm": 4.089269842296994, + "language_loss": 0.86541653, + "learning_rate": 3.321561120780869e-06, + "loss": 0.89321327, + "num_input_tokens_seen": 2134835, + "step": 98, + "time_per_iteration": 2.753443956375122 + }, + { + "auxiliary_loss_clip": 0.01607064, + "auxiliary_loss_mlp": 0.01160548, + "balance_loss_clip": 1.19481564, + "balance_loss_mlp": 1.08110642, + "epoch": 0.011904046173270006, + "flos": 22340674517760.0, + "grad_norm": 2.161503606093782, + "language_loss": 1.01641488, + "learning_rate": 3.3289159651708192e-06, + "loss": 1.04409099, + "num_input_tokens_seen": 2152410, + "step": 99, + "time_per_iteration": 2.758782148361206 + }, + { + "auxiliary_loss_clip": 0.01598959, + "auxiliary_loss_mlp": 0.01161793, + "balance_loss_clip": 1.19258797, + "balance_loss_mlp": 1.07901382, + "epoch": 0.012024289063909096, + "flos": 19100060375040.0, + "grad_norm": 2.0693558384665174, + "language_loss": 0.97673929, + "learning_rate": 3.3361968902759768e-06, + "loss": 1.00434685, + "num_input_tokens_seen": 2172090, + "step": 100, + "time_per_iteration": 2.7455148696899414 + }, + { + "auxiliary_loss_clip": 0.0158847, + "auxiliary_loss_mlp": 0.01166103, + "balance_loss_clip": 1.19160056, + "balance_loss_mlp": 1.09257507, + "epoch": 0.012144531954548187, + "flos": 15012205159680.0, + "grad_norm": 2.1355528154439476, + "language_loss": 0.94163895, + "learning_rate": 3.343405367163663e-06, + "loss": 0.96918476, + "num_input_tokens_seen": 2189020, + "step": 101, + "time_per_iteration": 2.7556285858154297 + }, + { + "auxiliary_loss_clip": 0.01597146, + "auxiliary_loss_mlp": 0.01149575, + "balance_loss_clip": 1.19571829, + "balance_loss_mlp": 1.08024263, + "epoch": 0.012264774845187279, + "flos": 15122020014720.0, + "grad_norm": 2.4016181965206567, + "language_loss": 0.81187534, + "learning_rate": 3.350542823419951e-06, + "loss": 0.83934259, + "num_input_tokens_seen": 2205620, + "step": 102, + "time_per_iteration": 2.7973690032958984 + }, + { + "auxiliary_loss_clip": 0.01586607, + "auxiliary_loss_mlp": 0.0115365, + "balance_loss_clip": 1.18967676, + "balance_loss_mlp": 1.07745147, + "epoch": 0.012385017735826368, + "flos": 13949248959360.0, + "grad_norm": 3.4548642284407443, + "language_loss": 0.87474859, + "learning_rate": 3.3576106448465615e-06, + "loss": 0.90215123, + "num_input_tokens_seen": 2219000, + "step": 103, + "time_per_iteration": 2.6694674491882324 + }, + { + "auxiliary_loss_clip": 0.01578514, + "auxiliary_loss_mlp": 0.01145719, + "balance_loss_clip": 1.18450177, + "balance_loss_mlp": 1.07557595, + "epoch": 0.01250526062646546, + "flos": 23623260428160.0, + "grad_norm": 3.8065248728612855, + "language_loss": 0.88140714, + "learning_rate": 3.3646101770757797e-06, + "loss": 0.90864944, + "num_input_tokens_seen": 2237790, + "step": 104, + "time_per_iteration": 2.8626224994659424 + }, + { + "auxiliary_loss_clip": 0.01577478, + "auxiliary_loss_mlp": 0.01140983, + "balance_loss_clip": 1.19057953, + "balance_loss_mlp": 1.07031584, + "epoch": 0.012625503517104552, + "flos": 34640078958720.0, + "grad_norm": 1.7734514189182784, + "language_loss": 0.85608166, + "learning_rate": 3.371542727108104e-06, + "loss": 0.88326627, + "num_input_tokens_seen": 2259965, + "step": 105, + "time_per_iteration": 2.936739683151245 + }, + { + "auxiliary_loss_clip": 0.01575753, + "auxiliary_loss_mlp": 0.01177423, + "balance_loss_clip": 1.18816781, + "balance_loss_mlp": 1.1025362, + "epoch": 0.012745746407743641, + "flos": 17821891837440.0, + "grad_norm": 2.836313427959545, + "language_loss": 0.89733952, + "learning_rate": 3.3784095647770114e-06, + "loss": 0.92487133, + "num_input_tokens_seen": 2278610, + "step": 106, + "time_per_iteration": 2.7735798358917236 + }, + { + "auxiliary_loss_clip": 0.01568697, + "auxiliary_loss_mlp": 0.0114649, + "balance_loss_clip": 1.18299913, + "balance_loss_mlp": 1.07215166, + "epoch": 0.012865989298382733, + "flos": 20595057361920.0, + "grad_norm": 2.178548831392148, + "language_loss": 0.88690662, + "learning_rate": 3.3852119241449547e-06, + "loss": 0.91405851, + "num_input_tokens_seen": 2297730, + "step": 107, + "time_per_iteration": 2.7719123363494873 + }, + { + "auxiliary_loss_clip": 0.01557718, + "auxiliary_loss_mlp": 0.01131331, + "balance_loss_clip": 1.18069398, + "balance_loss_mlp": 1.06362033, + "epoch": 0.012986232189021825, + "flos": 23948969978880.0, + "grad_norm": 4.071273657365171, + "language_loss": 0.96176612, + "learning_rate": 3.3919510048344295e-06, + "loss": 0.98865664, + "num_input_tokens_seen": 2315740, + "step": 108, + "time_per_iteration": 2.790341377258301 + }, + { + "auxiliary_loss_clip": 0.01545289, + "auxiliary_loss_mlp": 0.01145032, + "balance_loss_clip": 1.18111253, + "balance_loss_mlp": 1.08166051, + "epoch": 0.013106475079660914, + "flos": 23725425686400.0, + "grad_norm": 3.1410598254970172, + "language_loss": 0.86641157, + "learning_rate": 3.3986279732976907e-06, + "loss": 0.89331472, + "num_input_tokens_seen": 2334215, + "step": 109, + "time_per_iteration": 2.772515296936035 + }, + { + "auxiliary_loss_clip": 0.01552758, + "auxiliary_loss_mlp": 0.0112058, + "balance_loss_clip": 1.18266654, + "balance_loss_mlp": 1.06068969, + "epoch": 0.013226717970300006, + "flos": 21102438925440.0, + "grad_norm": 1.8598325427833489, + "language_loss": 0.95382446, + "learning_rate": 3.4052439640284983e-06, + "loss": 0.9805578, + "num_input_tokens_seen": 2353130, + "step": 110, + "time_per_iteration": 2.778524875640869 + }, + { + "auxiliary_loss_clip": 0.01538419, + "auxiliary_loss_mlp": 0.01131731, + "balance_loss_clip": 1.17735231, + "balance_loss_mlp": 1.06821632, + "epoch": 0.013346960860939098, + "flos": 24863902231680.0, + "grad_norm": 1.8678144110578228, + "language_loss": 0.81010193, + "learning_rate": 3.4118000807190217e-06, + "loss": 0.83680344, + "num_input_tokens_seen": 2374010, + "step": 111, + "time_per_iteration": 2.8559200763702393 + }, + { + "auxiliary_loss_clip": 0.01543133, + "auxiliary_loss_mlp": 0.01148435, + "balance_loss_clip": 1.17688894, + "balance_loss_mlp": 1.08248806, + "epoch": 0.013467203751578187, + "flos": 28181940140160.0, + "grad_norm": 1.7254685936159215, + "language_loss": 0.76038301, + "learning_rate": 3.4182973973648723e-06, + "loss": 0.78729868, + "num_input_tokens_seen": 2395220, + "step": 112, + "time_per_iteration": 2.804168939590454 + }, + { + "auxiliary_loss_clip": 0.01544715, + "auxiliary_loss_mlp": 0.01140652, + "balance_loss_clip": 1.17843163, + "balance_loss_mlp": 1.08200073, + "epoch": 0.013587446642217279, + "flos": 18916233546240.0, + "grad_norm": 4.265404946586954, + "language_loss": 0.95232821, + "learning_rate": 3.424736959321014e-06, + "loss": 0.97918177, + "num_input_tokens_seen": 2413025, + "step": 113, + "time_per_iteration": 2.7865657806396484 + }, + { + "auxiliary_loss_clip": 0.01535532, + "auxiliary_loss_mlp": 0.01138595, + "balance_loss_clip": 1.17732477, + "balance_loss_mlp": 1.08044434, + "epoch": 0.01370768953285637, + "flos": 23988615615360.0, + "grad_norm": 2.053168107885883, + "language_loss": 0.88659537, + "learning_rate": 3.431119784311155e-06, + "loss": 0.91333669, + "num_input_tokens_seen": 2432700, + "step": 114, + "time_per_iteration": 2.7541580200195312 + }, + { + "auxiliary_loss_clip": 0.01531548, + "auxiliary_loss_mlp": 0.01147593, + "balance_loss_clip": 1.17720675, + "balance_loss_mlp": 1.08851314, + "epoch": 0.01382793242349546, + "flos": 39202565512320.0, + "grad_norm": 2.260123974265431, + "language_loss": 0.77690381, + "learning_rate": 3.43744686339307e-06, + "loss": 0.8036952, + "num_input_tokens_seen": 2455020, + "step": 115, + "time_per_iteration": 2.917344331741333 + }, + { + "auxiliary_loss_clip": 0.01517219, + "auxiliary_loss_mlp": 0.01113644, + "balance_loss_clip": 1.16803658, + "balance_loss_mlp": 1.06066716, + "epoch": 0.013948175314134552, + "flos": 41353506714240.0, + "grad_norm": 2.1291389505922758, + "language_loss": 0.90782523, + "learning_rate": 3.44371916188212e-06, + "loss": 0.93413383, + "num_input_tokens_seen": 2475775, + "step": 116, + "time_per_iteration": 2.895068883895874 + }, + { + "auxiliary_loss_clip": 0.01519704, + "auxiliary_loss_mlp": 0.01099355, + "balance_loss_clip": 1.17054796, + "balance_loss_mlp": 1.04525757, + "epoch": 0.014068418204773643, + "flos": 22453542028800.0, + "grad_norm": 2.370814491669494, + "language_loss": 0.86160076, + "learning_rate": 3.449937620235143e-06, + "loss": 0.88779134, + "num_input_tokens_seen": 2496370, + "step": 117, + "time_per_iteration": 2.7861404418945312 + }, + { + "auxiliary_loss_clip": 0.01514706, + "auxiliary_loss_mlp": 0.0114594, + "balance_loss_clip": 1.17153287, + "balance_loss_mlp": 1.09539533, + "epoch": 0.014188661095412733, + "flos": 23805147922560.0, + "grad_norm": 1.7349599076193287, + "language_loss": 0.89385432, + "learning_rate": 3.456103154896722e-06, + "loss": 0.92046082, + "num_input_tokens_seen": 2517645, + "step": 118, + "time_per_iteration": 2.805415630340576 + }, + { + "auxiliary_loss_clip": 0.01504955, + "auxiliary_loss_mlp": 0.01124168, + "balance_loss_clip": 1.16691148, + "balance_loss_mlp": 1.07226419, + "epoch": 0.014308903986051825, + "flos": 23660248458240.0, + "grad_norm": 2.551736670002943, + "language_loss": 0.92354167, + "learning_rate": 3.462216659109757e-06, + "loss": 0.94983298, + "num_input_tokens_seen": 2537825, + "step": 119, + "time_per_iteration": 4.585986852645874 + }, + { + "auxiliary_loss_clip": 0.01516922, + "auxiliary_loss_mlp": 0.01137113, + "balance_loss_clip": 1.17389441, + "balance_loss_mlp": 1.08847547, + "epoch": 0.014429146876690916, + "flos": 20667991927680.0, + "grad_norm": 5.073514688903926, + "language_loss": 0.85219741, + "learning_rate": 3.4682790036921077e-06, + "loss": 0.87873775, + "num_input_tokens_seen": 2556485, + "step": 120, + "time_per_iteration": 3.653724431991577 + }, + { + "auxiliary_loss_clip": 0.01502217, + "auxiliary_loss_mlp": 0.01133611, + "balance_loss_clip": 1.16665804, + "balance_loss_mlp": 1.09305573, + "epoch": 0.014549389767330006, + "flos": 20229199384320.0, + "grad_norm": 1.9579223256953364, + "language_loss": 0.83074772, + "learning_rate": 3.4742910377810193e-06, + "loss": 0.85710597, + "num_input_tokens_seen": 2573945, + "step": 121, + "time_per_iteration": 3.6606853008270264 + }, + { + "auxiliary_loss_clip": 0.01496992, + "auxiliary_loss_mlp": 0.01115527, + "balance_loss_clip": 1.16380072, + "balance_loss_mlp": 1.07423258, + "epoch": 0.014669632657969098, + "flos": 18004174381440.0, + "grad_norm": 2.3075692733606648, + "language_loss": 0.88880432, + "learning_rate": 3.4802535895469042e-06, + "loss": 0.91492951, + "num_input_tokens_seen": 2592695, + "step": 122, + "time_per_iteration": 2.7728865146636963 + }, + { + "auxiliary_loss_clip": 0.01494781, + "auxiliary_loss_mlp": 0.01104305, + "balance_loss_clip": 1.16510689, + "balance_loss_mlp": 1.06103158, + "epoch": 0.01478987554860819, + "flos": 22741796672640.0, + "grad_norm": 2.1091646170297973, + "language_loss": 0.89493132, + "learning_rate": 3.4861674668779934e-06, + "loss": 0.92092216, + "num_input_tokens_seen": 2610925, + "step": 123, + "time_per_iteration": 2.765625 + }, + { + "auxiliary_loss_clip": 0.01492856, + "auxiliary_loss_mlp": 0.01118058, + "balance_loss_clip": 1.16337228, + "balance_loss_mlp": 1.07364094, + "epoch": 0.01491011843924728, + "flos": 17198590106880.0, + "grad_norm": 2.2348950325872954, + "language_loss": 0.84019744, + "learning_rate": 3.492033458037272e-06, + "loss": 0.86630654, + "num_input_tokens_seen": 2629495, + "step": 124, + "time_per_iteration": 2.8145558834075928 + }, + { + "auxiliary_loss_clip": 0.01486125, + "auxiliary_loss_mlp": 0.01124125, + "balance_loss_clip": 1.15971231, + "balance_loss_mlp": 1.07977903, + "epoch": 0.01503036132988637, + "flos": 17673867889920.0, + "grad_norm": 2.3165211570022333, + "language_loss": 0.87116587, + "learning_rate": 3.497852332293018e-06, + "loss": 0.89726841, + "num_input_tokens_seen": 2645070, + "step": 125, + "time_per_iteration": 2.6852188110351562 + }, + { + "auxiliary_loss_clip": 0.01494117, + "auxiliary_loss_mlp": 0.01112571, + "balance_loss_clip": 1.16479874, + "balance_loss_mlp": 1.0723021, + "epoch": 0.015150604220525462, + "flos": 18878239935360.0, + "grad_norm": 2.0444393366622418, + "language_loss": 0.96582282, + "learning_rate": 3.5036248405242356e-06, + "loss": 0.99188977, + "num_input_tokens_seen": 2663825, + "step": 126, + "time_per_iteration": 2.8324151039123535 + }, + { + "auxiliary_loss_clip": 0.01486222, + "auxiliary_loss_mlp": 0.01097395, + "balance_loss_clip": 1.15866423, + "balance_loss_mlp": 1.05338264, + "epoch": 0.015270847111164552, + "flos": 39420184060800.0, + "grad_norm": 1.8397850823158048, + "language_loss": 0.82639199, + "learning_rate": 3.509351715802146e-06, + "loss": 0.85222816, + "num_input_tokens_seen": 2684710, + "step": 127, + "time_per_iteration": 2.8748676776885986 + }, + { + "auxiliary_loss_clip": 0.01482258, + "auxiliary_loss_mlp": 0.01111754, + "balance_loss_clip": 1.15881419, + "balance_loss_mlp": 1.06716943, + "epoch": 0.015391090001803644, + "flos": 43762466286720.0, + "grad_norm": 1.9334824146799185, + "language_loss": 0.78499335, + "learning_rate": 3.5150336739488763e-06, + "loss": 0.81093347, + "num_input_tokens_seen": 2706995, + "step": 128, + "time_per_iteration": 3.051741123199463 + }, + { + "auxiliary_loss_clip": 0.01478153, + "auxiliary_loss_mlp": 0.01102907, + "balance_loss_clip": 1.1580255, + "balance_loss_mlp": 1.06718016, + "epoch": 0.015511332892442733, + "flos": 18916341287040.0, + "grad_norm": 1.6742282833146394, + "language_loss": 0.84040034, + "learning_rate": 3.5206714140744143e-06, + "loss": 0.866211, + "num_input_tokens_seen": 2727050, + "step": 129, + "time_per_iteration": 2.721099376678467 + }, + { + "auxiliary_loss_clip": 0.01483947, + "auxiliary_loss_mlp": 0.01114069, + "balance_loss_clip": 1.16582847, + "balance_loss_mlp": 1.07327616, + "epoch": 0.015631575783081827, + "flos": 24535283679360.0, + "grad_norm": 2.914011526040731, + "language_loss": 0.87533748, + "learning_rate": 3.5262656190928208e-06, + "loss": 0.90131766, + "num_input_tokens_seen": 2745350, + "step": 130, + "time_per_iteration": 2.815931558609009 + }, + { + "auxiliary_loss_clip": 0.01454196, + "auxiliary_loss_mlp": 0.01063554, + "balance_loss_clip": 1.23188353, + "balance_loss_mlp": 1.05354011, + "epoch": 0.015751818673720917, + "flos": 62328536098560.0, + "grad_norm": 1.0591200821583466, + "language_loss": 0.71535105, + "learning_rate": 3.5318169562186737e-06, + "loss": 0.74052858, + "num_input_tokens_seen": 2814195, + "step": 131, + "time_per_iteration": 3.369765520095825 + }, + { + "auxiliary_loss_clip": 0.01470228, + "auxiliary_loss_mlp": 0.01130723, + "balance_loss_clip": 1.15582299, + "balance_loss_mlp": 1.09255266, + "epoch": 0.015872061564360006, + "flos": 23878549365120.0, + "grad_norm": 1.8661025984978468, + "language_loss": 0.8227002, + "learning_rate": 3.5373260774446292e-06, + "loss": 0.8487097, + "num_input_tokens_seen": 2834645, + "step": 132, + "time_per_iteration": 2.7818856239318848 + }, + { + "auxiliary_loss_clip": 0.01464048, + "auxiliary_loss_mlp": 0.01107492, + "balance_loss_clip": 1.15208602, + "balance_loss_mlp": 1.06829643, + "epoch": 0.0159923044549991, + "flos": 23367899664000.0, + "grad_norm": 1.9324939483551153, + "language_loss": 0.90434003, + "learning_rate": 3.542793620000961e-06, + "loss": 0.93005538, + "num_input_tokens_seen": 2854120, + "step": 133, + "time_per_iteration": 2.8751535415649414 + }, + { + "auxiliary_loss_clip": 0.01466426, + "auxiliary_loss_mlp": 0.01134994, + "balance_loss_clip": 1.15409279, + "balance_loss_mlp": 1.09293675, + "epoch": 0.01611254734563819, + "flos": 17858305249920.0, + "grad_norm": 2.806781509052541, + "language_loss": 0.86767638, + "learning_rate": 3.5482202067978894e-06, + "loss": 0.89369059, + "num_input_tokens_seen": 2871330, + "step": 134, + "time_per_iteration": 2.766183376312256 + }, + { + "auxiliary_loss_clip": 0.01463971, + "auxiliary_loss_mlp": 0.01101908, + "balance_loss_clip": 1.15117621, + "balance_loss_mlp": 1.06521559, + "epoch": 0.01623279023627728, + "flos": 20954774113920.0, + "grad_norm": 2.052114824704706, + "language_loss": 0.75918067, + "learning_rate": 3.553606446851471e-06, + "loss": 0.78483939, + "num_input_tokens_seen": 2888070, + "step": 135, + "time_per_iteration": 2.7530648708343506 + }, + { + "auxiliary_loss_clip": 0.01460508, + "auxiliary_loss_mlp": 0.01096067, + "balance_loss_clip": 1.15136385, + "balance_loss_mlp": 1.06025672, + "epoch": 0.016353033126916373, + "flos": 15742412743680.0, + "grad_norm": 1.8898835298238248, + "language_loss": 0.83253968, + "learning_rate": 3.5589529356937613e-06, + "loss": 0.85810542, + "num_input_tokens_seen": 2906465, + "step": 136, + "time_per_iteration": 2.793030261993408 + }, + { + "auxiliary_loss_clip": 0.01453815, + "auxiliary_loss_mlp": 0.01099087, + "balance_loss_clip": 1.14660501, + "balance_loss_mlp": 1.06301475, + "epoch": 0.016473276017555463, + "flos": 18807280617600.0, + "grad_norm": 2.08295298689624, + "language_loss": 0.7694357, + "learning_rate": 3.5642602557679627e-06, + "loss": 0.79496473, + "num_input_tokens_seen": 2924915, + "step": 137, + "time_per_iteration": 2.831183671951294 + }, + { + "auxiliary_loss_clip": 0.01461017, + "auxiliary_loss_mlp": 0.01108018, + "balance_loss_clip": 1.1558789, + "balance_loss_mlp": 1.07442498, + "epoch": 0.016593518908194552, + "flos": 24352641999360.0, + "grad_norm": 5.223144324131421, + "language_loss": 0.83905441, + "learning_rate": 3.569528976809202e-06, + "loss": 0.86474466, + "num_input_tokens_seen": 2942130, + "step": 138, + "time_per_iteration": 2.8287088871002197 + }, + { + "auxiliary_loss_clip": 0.01449887, + "auxiliary_loss_mlp": 0.01114561, + "balance_loss_clip": 1.145818, + "balance_loss_mlp": 1.07776082, + "epoch": 0.016713761798833646, + "flos": 22346133384960.0, + "grad_norm": 1.79993498563005, + "language_loss": 0.90125805, + "learning_rate": 3.5747596562115522e-06, + "loss": 0.92690253, + "num_input_tokens_seen": 2962745, + "step": 139, + "time_per_iteration": 2.7096917629241943 + }, + { + "auxiliary_loss_clip": 0.01453449, + "auxiliary_loss_mlp": 0.01104341, + "balance_loss_clip": 1.14915371, + "balance_loss_mlp": 1.06583667, + "epoch": 0.016834004689472735, + "flos": 17821820010240.0, + "grad_norm": 2.237958535049375, + "language_loss": 0.91076106, + "learning_rate": 3.5799528393819138e-06, + "loss": 0.93633896, + "num_input_tokens_seen": 2981825, + "step": 140, + "time_per_iteration": 2.766188144683838 + }, + { + "auxiliary_loss_clip": 0.0145002, + "auxiliary_loss_mlp": 0.01105635, + "balance_loss_clip": 1.14554024, + "balance_loss_mlp": 1.07013416, + "epoch": 0.016954247580111825, + "flos": 20519501103360.0, + "grad_norm": 2.085146248802362, + "language_loss": 0.88118923, + "learning_rate": 3.585109060081286e-06, + "loss": 0.90674579, + "num_input_tokens_seen": 3001625, + "step": 141, + "time_per_iteration": 2.7242393493652344 + }, + { + "auxiliary_loss_clip": 0.01443622, + "auxiliary_loss_mlp": 0.01120435, + "balance_loss_clip": 1.14479756, + "balance_loss_mlp": 1.08526778, + "epoch": 0.017074490470750915, + "flos": 22088869200000.0, + "grad_norm": 2.479170303370634, + "language_loss": 0.78679341, + "learning_rate": 3.590228840753992e-06, + "loss": 0.81243396, + "num_input_tokens_seen": 3022055, + "step": 142, + "time_per_iteration": 2.78906512260437 + }, + { + "auxiliary_loss_clip": 0.01443154, + "auxiliary_loss_mlp": 0.01084711, + "balance_loss_clip": 1.1455791, + "balance_loss_mlp": 1.05309677, + "epoch": 0.01719473336139001, + "flos": 15997270717440.0, + "grad_norm": 2.234217031241929, + "language_loss": 0.87130845, + "learning_rate": 3.5953126928453423e-06, + "loss": 0.89658713, + "num_input_tokens_seen": 3039605, + "step": 143, + "time_per_iteration": 2.6699001789093018 + }, + { + "auxiliary_loss_clip": 0.01438591, + "auxiliary_loss_mlp": 0.01107881, + "balance_loss_clip": 1.14294839, + "balance_loss_mlp": 1.07500315, + "epoch": 0.017314976252029098, + "flos": 22492038430080.0, + "grad_norm": 2.1759740611917304, + "language_loss": 0.80694985, + "learning_rate": 3.600361117108239e-06, + "loss": 0.83241463, + "num_input_tokens_seen": 3059405, + "step": 144, + "time_per_iteration": 2.8228015899658203 + }, + { + "auxiliary_loss_clip": 0.01435109, + "auxiliary_loss_mlp": 0.011232, + "balance_loss_clip": 1.14215767, + "balance_loss_mlp": 1.09094226, + "epoch": 0.017435219142668188, + "flos": 22018053536640.0, + "grad_norm": 2.8865147524983037, + "language_loss": 0.97170866, + "learning_rate": 3.6053746038991616e-06, + "loss": 0.99729174, + "num_input_tokens_seen": 3078490, + "step": 145, + "time_per_iteration": 3.7005903720855713 + }, + { + "auxiliary_loss_clip": 0.01409797, + "auxiliary_loss_mlp": 0.01020509, + "balance_loss_clip": 1.19746685, + "balance_loss_mlp": 1.00820625, + "epoch": 0.01755546203330728, + "flos": 72240526149120.0, + "grad_norm": 1.0533939155440573, + "language_loss": 0.58424008, + "learning_rate": 3.6103536334639843e-06, + "loss": 0.60854316, + "num_input_tokens_seen": 3131755, + "step": 146, + "time_per_iteration": 5.053954601287842 + }, + { + "auxiliary_loss_clip": 0.01424828, + "auxiliary_loss_mlp": 0.01080412, + "balance_loss_clip": 1.13736534, + "balance_loss_mlp": 1.04793954, + "epoch": 0.01767570492394637, + "flos": 25337061112320.0, + "grad_norm": 2.0429523370926264, + "language_loss": 0.85490745, + "learning_rate": 3.615298676214041e-06, + "loss": 0.87995982, + "num_input_tokens_seen": 3152035, + "step": 147, + "time_per_iteration": 3.7344775199890137 + }, + { + "auxiliary_loss_clip": 0.01426392, + "auxiliary_loss_mlp": 0.01107236, + "balance_loss_clip": 1.13846445, + "balance_loss_mlp": 1.07428694, + "epoch": 0.01779594781458546, + "flos": 20449188230400.0, + "grad_norm": 2.2071411615065566, + "language_loss": 0.8883214, + "learning_rate": 3.6202101929928317e-06, + "loss": 0.91365767, + "num_input_tokens_seen": 3170625, + "step": 148, + "time_per_iteration": 2.8476035594940186 + }, + { + "auxiliary_loss_clip": 0.01422454, + "auxiliary_loss_mlp": 0.01099514, + "balance_loss_clip": 1.13643456, + "balance_loss_mlp": 1.07039094, + "epoch": 0.017916190705224554, + "flos": 16253601148800.0, + "grad_norm": 10.278207107650232, + "language_loss": 0.88482928, + "learning_rate": 3.6250886353337413e-06, + "loss": 0.91004896, + "num_input_tokens_seen": 3188155, + "step": 149, + "time_per_iteration": 2.807244300842285 + }, + { + "auxiliary_loss_clip": 0.01426472, + "auxiliary_loss_mlp": 0.01092173, + "balance_loss_clip": 1.13771033, + "balance_loss_mlp": 1.06045115, + "epoch": 0.018036433595863644, + "flos": 23330588411520.0, + "grad_norm": 2.0639935558631533, + "language_loss": 0.86386538, + "learning_rate": 3.6299344457091488e-06, + "loss": 0.8890518, + "num_input_tokens_seen": 3209015, + "step": 150, + "time_per_iteration": 2.7743980884552 + }, + { + "auxiliary_loss_clip": 0.01418854, + "auxiliary_loss_mlp": 0.01099032, + "balance_loss_clip": 1.13293123, + "balance_loss_mlp": 1.06776381, + "epoch": 0.018156676486502734, + "flos": 18588010043520.0, + "grad_norm": 2.2359040195906448, + "language_loss": 0.93747246, + "learning_rate": 3.634748057771256e-06, + "loss": 0.96265137, + "num_input_tokens_seen": 3224955, + "step": 151, + "time_per_iteration": 2.8680765628814697 + }, + { + "auxiliary_loss_clip": 0.01416504, + "auxiliary_loss_mlp": 0.01090921, + "balance_loss_clip": 1.13394749, + "balance_loss_mlp": 1.06305027, + "epoch": 0.018276919377141827, + "flos": 25448707560960.0, + "grad_norm": 1.9445584956482809, + "language_loss": 0.85740185, + "learning_rate": 3.639529896584965e-06, + "loss": 0.88247615, + "num_input_tokens_seen": 3246330, + "step": 152, + "time_per_iteration": 2.7896509170532227 + }, + { + "auxiliary_loss_clip": 0.01417164, + "auxiliary_loss_mlp": 0.01087643, + "balance_loss_clip": 1.13384056, + "balance_loss_mlp": 1.05676746, + "epoch": 0.018397162267780917, + "flos": 20047311889920.0, + "grad_norm": 3.22470799375155, + "language_loss": 0.88731223, + "learning_rate": 3.6442803788531233e-06, + "loss": 0.91236037, + "num_input_tokens_seen": 3264290, + "step": 153, + "time_per_iteration": 2.7935988903045654 + }, + { + "auxiliary_loss_clip": 0.01419562, + "auxiliary_loss_mlp": 0.01092565, + "balance_loss_clip": 1.13364184, + "balance_loss_mlp": 1.06378818, + "epoch": 0.018517405158420007, + "flos": 27565282425600.0, + "grad_norm": 2.25044072143907, + "language_loss": 0.96037996, + "learning_rate": 3.6489999131344357e-06, + "loss": 0.98550129, + "num_input_tokens_seen": 3287065, + "step": 154, + "time_per_iteration": 2.787590265274048 + }, + { + "auxiliary_loss_clip": 0.01406475, + "auxiliary_loss_mlp": 0.01076074, + "balance_loss_clip": 1.13039267, + "balance_loss_mlp": 1.05280447, + "epoch": 0.0186376480490591, + "flos": 19354056422400.0, + "grad_norm": 2.162549226971017, + "language_loss": 0.90558833, + "learning_rate": 3.653688900054313e-06, + "loss": 0.93041384, + "num_input_tokens_seen": 3305595, + "step": 155, + "time_per_iteration": 2.7302520275115967 + }, + { + "auxiliary_loss_clip": 0.01408764, + "auxiliary_loss_mlp": 0.01081644, + "balance_loss_clip": 1.12727046, + "balance_loss_mlp": 1.0514605, + "epoch": 0.01875789093969819, + "flos": 26687840993280.0, + "grad_norm": 2.142591482364704, + "language_loss": 0.76045519, + "learning_rate": 3.6583477325089526e-06, + "loss": 0.78535926, + "num_input_tokens_seen": 3326135, + "step": 156, + "time_per_iteration": 2.7781074047088623 + }, + { + "auxiliary_loss_clip": 0.01411498, + "auxiliary_loss_mlp": 0.01105502, + "balance_loss_clip": 1.13251829, + "balance_loss_mlp": 1.07617688, + "epoch": 0.01887813383033728, + "flos": 24353001135360.0, + "grad_norm": 2.539180490239585, + "language_loss": 1.04545808, + "learning_rate": 3.6629767958628916e-06, + "loss": 1.07062817, + "num_input_tokens_seen": 3343510, + "step": 157, + "time_per_iteration": 2.8653438091278076 + }, + { + "auxiliary_loss_clip": 0.01408755, + "auxiliary_loss_mlp": 0.01087106, + "balance_loss_clip": 1.13115191, + "balance_loss_mlp": 1.05809021, + "epoch": 0.018998376720976373, + "flos": 14647532330880.0, + "grad_norm": 2.14216433312879, + "language_loss": 0.85661274, + "learning_rate": 3.667576468140291e-06, + "loss": 0.88157129, + "num_input_tokens_seen": 3361325, + "step": 158, + "time_per_iteration": 2.798368453979492 + }, + { + "auxiliary_loss_clip": 0.01402854, + "auxiliary_loss_mlp": 0.01084689, + "balance_loss_clip": 1.12731051, + "balance_loss_mlp": 1.05717516, + "epoch": 0.019118619611615463, + "flos": 29305261146240.0, + "grad_norm": 2.546589081712018, + "language_loss": 0.890957, + "learning_rate": 3.672147120210184e-06, + "loss": 0.91583246, + "num_input_tokens_seen": 3377925, + "step": 159, + "time_per_iteration": 2.755950689315796 + }, + { + "auxiliary_loss_clip": 0.01403761, + "auxiliary_loss_mlp": 0.01085936, + "balance_loss_clip": 1.1295768, + "balance_loss_mlp": 1.06022298, + "epoch": 0.019238862502254553, + "flos": 20886723797760.0, + "grad_norm": 4.081286719035537, + "language_loss": 0.86315447, + "learning_rate": 3.6766891159659177e-06, + "loss": 0.88805139, + "num_input_tokens_seen": 3396335, + "step": 160, + "time_per_iteration": 2.7785587310791016 + }, + { + "auxiliary_loss_clip": 0.01400556, + "auxiliary_loss_mlp": 0.01081827, + "balance_loss_clip": 1.12615776, + "balance_loss_mlp": 1.0529424, + "epoch": 0.019359105392893646, + "flos": 21360672777600.0, + "grad_norm": 2.7304615567453103, + "language_loss": 0.88033539, + "learning_rate": 3.6812028124990075e-06, + "loss": 0.90515924, + "num_input_tokens_seen": 3413605, + "step": 161, + "time_per_iteration": 2.6959619522094727 + }, + { + "auxiliary_loss_clip": 0.01399381, + "auxiliary_loss_mlp": 0.01087746, + "balance_loss_clip": 1.12666059, + "balance_loss_mlp": 1.0620209, + "epoch": 0.019479348283532736, + "flos": 16283729681280.0, + "grad_norm": 2.684042112917351, + "language_loss": 0.81725001, + "learning_rate": 3.6856885602676016e-06, + "loss": 0.84212124, + "num_input_tokens_seen": 3429640, + "step": 162, + "time_per_iteration": 2.7083513736724854 + }, + { + "auxiliary_loss_clip": 0.01401641, + "auxiliary_loss_mlp": 0.01079469, + "balance_loss_clip": 1.12905788, + "balance_loss_mlp": 1.05410111, + "epoch": 0.019599591174171826, + "flos": 22091239497600.0, + "grad_norm": 4.4402885671150205, + "language_loss": 0.94082904, + "learning_rate": 3.6901467032597733e-06, + "loss": 0.96564019, + "num_input_tokens_seen": 3448125, + "step": 163, + "time_per_iteration": 2.724883556365967 + }, + { + "auxiliary_loss_clip": 0.01397281, + "auxiliary_loss_mlp": 0.01083608, + "balance_loss_clip": 1.1224575, + "balance_loss_mlp": 1.05791807, + "epoch": 0.01971983406481092, + "flos": 19609668581760.0, + "grad_norm": 2.547091538023541, + "language_loss": 0.87459964, + "learning_rate": 3.694577579151804e-06, + "loss": 0.89940852, + "num_input_tokens_seen": 3466535, + "step": 164, + "time_per_iteration": 2.718506097793579 + }, + { + "auxiliary_loss_clip": 0.01397586, + "auxiliary_loss_mlp": 0.01081965, + "balance_loss_clip": 1.12650073, + "balance_loss_mlp": 1.05484533, + "epoch": 0.01984007695545001, + "flos": 19099342103040.0, + "grad_norm": 2.146838082907181, + "language_loss": 0.73995268, + "learning_rate": 3.6989815194616703e-06, + "loss": 0.76474822, + "num_input_tokens_seen": 3483730, + "step": 165, + "time_per_iteration": 2.686552047729492 + }, + { + "auxiliary_loss_clip": 0.01397814, + "auxiliary_loss_mlp": 0.01102848, + "balance_loss_clip": 1.12402213, + "balance_loss_mlp": 1.07426119, + "epoch": 0.0199603198460891, + "flos": 20848406964480.0, + "grad_norm": 2.4306366865543554, + "language_loss": 0.80061477, + "learning_rate": 3.703358849697888e-06, + "loss": 0.82562137, + "num_input_tokens_seen": 3503640, + "step": 166, + "time_per_iteration": 2.7668328285217285 + }, + { + "auxiliary_loss_clip": 0.01395626, + "auxiliary_loss_mlp": 0.01086932, + "balance_loss_clip": 1.12515783, + "balance_loss_mlp": 1.06342387, + "epoch": 0.020080562736728192, + "flos": 21870747861120.0, + "grad_norm": 5.404802572152084, + "language_loss": 0.8238886, + "learning_rate": 3.7077098895038803e-06, + "loss": 0.84871411, + "num_input_tokens_seen": 3523010, + "step": 167, + "time_per_iteration": 2.7065987586975098 + }, + { + "auxiliary_loss_clip": 0.01393684, + "auxiliary_loss_mlp": 0.01067739, + "balance_loss_clip": 1.12444639, + "balance_loss_mlp": 1.04121506, + "epoch": 0.020200805627367282, + "flos": 21688788539520.0, + "grad_norm": 2.400361683364502, + "language_loss": 0.97092652, + "learning_rate": 3.712034952798045e-06, + "loss": 0.99554062, + "num_input_tokens_seen": 3541125, + "step": 168, + "time_per_iteration": 2.702672004699707 + }, + { + "auxiliary_loss_clip": 0.01392897, + "auxiliary_loss_mlp": 0.01084641, + "balance_loss_clip": 1.1229105, + "balance_loss_mlp": 1.0545404, + "epoch": 0.02032104851800637, + "flos": 33543043729920.0, + "grad_norm": 3.3572202488735776, + "language_loss": 0.84514499, + "learning_rate": 3.7163343479096656e-06, + "loss": 0.86992031, + "num_input_tokens_seen": 3562700, + "step": 169, + "time_per_iteration": 2.8590409755706787 + }, + { + "auxiliary_loss_clip": 0.01388855, + "auxiliary_loss_mlp": 0.010738, + "balance_loss_clip": 1.12308812, + "balance_loss_mlp": 1.04974353, + "epoch": 0.020441291408645465, + "flos": 31686965274240.0, + "grad_norm": 3.199448842820734, + "language_loss": 0.82903975, + "learning_rate": 3.720608377710802e-06, + "loss": 0.85366631, + "num_input_tokens_seen": 3582790, + "step": 170, + "time_per_iteration": 2.8558037281036377 + }, + { + "auxiliary_loss_clip": 0.01390502, + "auxiliary_loss_mlp": 0.01090944, + "balance_loss_clip": 1.12147546, + "balance_loss_mlp": 1.06220269, + "epoch": 0.020561534299284555, + "flos": 20886687884160.0, + "grad_norm": 2.22275582290134, + "language_loss": 0.86509115, + "learning_rate": 3.7248573397443277e-06, + "loss": 0.88990563, + "num_input_tokens_seen": 3601715, + "step": 171, + "time_per_iteration": 3.6584243774414062 + }, + { + "auxiliary_loss_clip": 0.01388703, + "auxiliary_loss_mlp": 0.01091828, + "balance_loss_clip": 1.12520528, + "balance_loss_mlp": 1.06449389, + "epoch": 0.020681777189923645, + "flos": 20996610480000.0, + "grad_norm": 2.4098357548001164, + "language_loss": 0.97487402, + "learning_rate": 3.729081526348224e-06, + "loss": 0.99967921, + "num_input_tokens_seen": 3620245, + "step": 172, + "time_per_iteration": 4.567703723907471 + }, + { + "auxiliary_loss_clip": 0.01384299, + "auxiliary_loss_mlp": 0.01078703, + "balance_loss_clip": 1.11945868, + "balance_loss_mlp": 1.05358577, + "epoch": 0.020802020080562738, + "flos": 28257532312320.0, + "grad_norm": 5.073930852535722, + "language_loss": 0.85162079, + "learning_rate": 3.7332812247762777e-06, + "loss": 0.87625074, + "num_input_tokens_seen": 3641545, + "step": 173, + "time_per_iteration": 2.762882947921753 + }, + { + "auxiliary_loss_clip": 0.01388823, + "auxiliary_loss_mlp": 0.01092816, + "balance_loss_clip": 1.12241948, + "balance_loss_mlp": 1.06755567, + "epoch": 0.020922262971201828, + "flos": 19681274344320.0, + "grad_norm": 2.5411036845636517, + "language_loss": 0.9558264, + "learning_rate": 3.737456717315293e-06, + "loss": 0.9806428, + "num_input_tokens_seen": 3660510, + "step": 174, + "time_per_iteration": 3.7231085300445557 + }, + { + "auxiliary_loss_clip": 0.01384809, + "auxiliary_loss_mlp": 0.01071287, + "balance_loss_clip": 1.12315774, + "balance_loss_mlp": 1.04878044, + "epoch": 0.021042505861840918, + "flos": 15666353694720.0, + "grad_norm": 1.9944877336291662, + "language_loss": 0.90653491, + "learning_rate": 3.7416082813989552e-06, + "loss": 0.9310959, + "num_input_tokens_seen": 3677505, + "step": 175, + "time_per_iteration": 2.693608522415161 + }, + { + "auxiliary_loss_clip": 0.01389925, + "auxiliary_loss_mlp": 0.01086637, + "balance_loss_clip": 1.12285399, + "balance_loss_mlp": 1.06188917, + "epoch": 0.02116274875248001, + "flos": 21142012734720.0, + "grad_norm": 2.237991678580765, + "language_loss": 0.89342606, + "learning_rate": 3.745736189718439e-06, + "loss": 0.91819167, + "num_input_tokens_seen": 3696760, + "step": 176, + "time_per_iteration": 2.7722327709198 + }, + { + "auxiliary_loss_clip": 0.01377674, + "auxiliary_loss_mlp": 0.01069006, + "balance_loss_clip": 1.11743963, + "balance_loss_mlp": 1.04576051, + "epoch": 0.0212829916431191, + "flos": 24715770543360.0, + "grad_norm": 3.91160134935488, + "language_loss": 0.72407073, + "learning_rate": 3.749840710329894e-06, + "loss": 0.74853754, + "num_input_tokens_seen": 3717465, + "step": 177, + "time_per_iteration": 2.8535451889038086 + }, + { + "auxiliary_loss_clip": 0.01384544, + "auxiliary_loss_mlp": 0.01097328, + "balance_loss_clip": 1.11975455, + "balance_loss_mlp": 1.06700158, + "epoch": 0.02140323453375819, + "flos": 16645493508480.0, + "grad_norm": 2.817434588701684, + "language_loss": 0.97754568, + "learning_rate": 3.7539221067588938e-06, + "loss": 1.0023644, + "num_input_tokens_seen": 3731440, + "step": 178, + "time_per_iteration": 2.846749782562256 + }, + { + "auxiliary_loss_clip": 0.01385853, + "auxiliary_loss_mlp": 0.01095152, + "balance_loss_clip": 1.12032533, + "balance_loss_mlp": 1.06739974, + "epoch": 0.021523477424397284, + "flos": 20299332689280.0, + "grad_norm": 3.4593247700036356, + "language_loss": 0.93512386, + "learning_rate": 3.757980638101964e-06, + "loss": 0.95993394, + "num_input_tokens_seen": 3744935, + "step": 179, + "time_per_iteration": 2.727851152420044 + }, + { + "auxiliary_loss_clip": 0.01384398, + "auxiliary_loss_mlp": 0.01086947, + "balance_loss_clip": 1.12069654, + "balance_loss_mlp": 1.05944562, + "epoch": 0.021643720315036374, + "flos": 26104005331200.0, + "grad_norm": 2.3288528896307206, + "language_loss": 0.89799821, + "learning_rate": 3.7620165591252806e-06, + "loss": 0.92271173, + "num_input_tokens_seen": 3763035, + "step": 180, + "time_per_iteration": 2.80619740486145 + }, + { + "auxiliary_loss_clip": 0.0138104, + "auxiliary_loss_mlp": 0.01073399, + "balance_loss_clip": 1.12005806, + "balance_loss_mlp": 1.04989088, + "epoch": 0.021763963205675464, + "flos": 24787663614720.0, + "grad_norm": 1.9407580753021745, + "language_loss": 0.94266218, + "learning_rate": 3.766030120360636e-06, + "loss": 0.9672066, + "num_input_tokens_seen": 3782665, + "step": 181, + "time_per_iteration": 2.789510488510132 + }, + { + "auxiliary_loss_clip": 0.0137737, + "auxiliary_loss_mlp": 0.01074655, + "balance_loss_clip": 1.11601186, + "balance_loss_mlp": 1.05154061, + "epoch": 0.021884206096314557, + "flos": 25813559957760.0, + "grad_norm": 2.10750822851974, + "language_loss": 0.90181005, + "learning_rate": 3.7700215681987578e-06, + "loss": 0.92633033, + "num_input_tokens_seen": 3802435, + "step": 182, + "time_per_iteration": 2.7821271419525146 + }, + { + "auxiliary_loss_clip": 0.01373865, + "auxiliary_loss_mlp": 0.01075058, + "balance_loss_clip": 1.11574745, + "balance_loss_mlp": 1.05038214, + "epoch": 0.022004448986953647, + "flos": 20082719721600.0, + "grad_norm": 3.0037735034678126, + "language_loss": 0.82240355, + "learning_rate": 3.7739911449800767e-06, + "loss": 0.84689277, + "num_input_tokens_seen": 3822490, + "step": 183, + "time_per_iteration": 2.7556846141815186 + }, + { + "auxiliary_loss_clip": 0.01373903, + "auxiliary_loss_mlp": 0.01075268, + "balance_loss_clip": 1.1148411, + "balance_loss_mlp": 1.05302393, + "epoch": 0.022124691877592736, + "flos": 20480609652480.0, + "grad_norm": 3.9223147564152976, + "language_loss": 0.80783874, + "learning_rate": 3.7779390890830114e-06, + "loss": 0.83233047, + "num_input_tokens_seen": 3841140, + "step": 184, + "time_per_iteration": 2.7697062492370605 + }, + { + "auxiliary_loss_clip": 0.01377382, + "auxiliary_loss_mlp": 0.01075726, + "balance_loss_clip": 1.11480951, + "balance_loss_mlp": 1.05070448, + "epoch": 0.02224493476823183, + "flos": 23586847015680.0, + "grad_norm": 2.090610297050497, + "language_loss": 0.85908961, + "learning_rate": 3.7818656350098723e-06, + "loss": 0.88362074, + "num_input_tokens_seen": 3862090, + "step": 185, + "time_per_iteration": 2.7179205417633057 + }, + { + "auxiliary_loss_clip": 0.01377008, + "auxiliary_loss_mlp": 0.01079651, + "balance_loss_clip": 1.1161294, + "balance_loss_mlp": 1.05568981, + "epoch": 0.02236517765887092, + "flos": 16909940413440.0, + "grad_norm": 2.6047351847284195, + "language_loss": 0.7719596, + "learning_rate": 3.7857710134704447e-06, + "loss": 0.79652619, + "num_input_tokens_seen": 3881025, + "step": 186, + "time_per_iteration": 2.746959924697876 + }, + { + "auxiliary_loss_clip": 0.01368702, + "auxiliary_loss_mlp": 0.01068517, + "balance_loss_clip": 1.11388445, + "balance_loss_mlp": 1.0468924, + "epoch": 0.02248542054951001, + "flos": 43508182930560.0, + "grad_norm": 2.287647698313398, + "language_loss": 0.79382437, + "learning_rate": 3.7896554514633234e-06, + "loss": 0.81819654, + "num_input_tokens_seen": 3905310, + "step": 187, + "time_per_iteration": 2.876145362854004 + }, + { + "auxiliary_loss_clip": 0.01367505, + "auxiliary_loss_mlp": 0.01066558, + "balance_loss_clip": 1.1108427, + "balance_loss_mlp": 1.04406333, + "epoch": 0.022605663440149103, + "flos": 23367648268800.0, + "grad_norm": 3.0594895969765066, + "language_loss": 0.83987892, + "learning_rate": 3.7935191723550955e-06, + "loss": 0.86421955, + "num_input_tokens_seen": 3924265, + "step": 188, + "time_per_iteration": 2.752495050430298 + }, + { + "auxiliary_loss_clip": 0.01367498, + "auxiliary_loss_mlp": 0.01066687, + "balance_loss_clip": 1.11202693, + "balance_loss_mlp": 1.04404902, + "epoch": 0.022725906330788193, + "flos": 29019915504000.0, + "grad_norm": 2.07137550980926, + "language_loss": 0.88616908, + "learning_rate": 3.797362395957408e-06, + "loss": 0.9105109, + "num_input_tokens_seen": 3944830, + "step": 189, + "time_per_iteration": 2.8082499504089355 + }, + { + "auxiliary_loss_clip": 0.01373307, + "auxiliary_loss_mlp": 0.01092374, + "balance_loss_clip": 1.11564422, + "balance_loss_mlp": 1.0693903, + "epoch": 0.022846149221427282, + "flos": 24496176746880.0, + "grad_norm": 2.0854258419587595, + "language_loss": 0.78107786, + "learning_rate": 3.8011853386020055e-06, + "loss": 0.80573469, + "num_input_tokens_seen": 3965735, + "step": 190, + "time_per_iteration": 2.7727606296539307 + }, + { + "auxiliary_loss_clip": 0.01367777, + "auxiliary_loss_mlp": 0.01081153, + "balance_loss_clip": 1.1125412, + "balance_loss_mlp": 1.05616641, + "epoch": 0.022966392112066376, + "flos": 15523537219200.0, + "grad_norm": 3.512390672166243, + "language_loss": 0.90043294, + "learning_rate": 3.804988213213804e-06, + "loss": 0.92492223, + "num_input_tokens_seen": 3983975, + "step": 191, + "time_per_iteration": 2.7706360816955566 + }, + { + "auxiliary_loss_clip": 0.01351124, + "auxiliary_loss_mlp": 0.01016706, + "balance_loss_clip": 1.16259527, + "balance_loss_mlp": 1.00859976, + "epoch": 0.023086635002705466, + "flos": 55650408433920.0, + "grad_norm": 1.0159613915908805, + "language_loss": 0.63148129, + "learning_rate": 3.808771229382049e-06, + "loss": 0.65515959, + "num_input_tokens_seen": 4043440, + "step": 192, + "time_per_iteration": 3.1709282398223877 + }, + { + "auxiliary_loss_clip": 0.01367259, + "auxiliary_loss_mlp": 0.01075471, + "balance_loss_clip": 1.1115638, + "balance_loss_mlp": 1.05505037, + "epoch": 0.023206877893344555, + "flos": 19313441118720.0, + "grad_norm": 2.024299483989744, + "language_loss": 0.84329081, + "learning_rate": 3.8125345934296324e-06, + "loss": 0.8677181, + "num_input_tokens_seen": 4061750, + "step": 193, + "time_per_iteration": 2.750279664993286 + }, + { + "auxiliary_loss_clip": 0.01363283, + "auxiliary_loss_mlp": 0.01103162, + "balance_loss_clip": 1.10761952, + "balance_loss_mlp": 1.0773052, + "epoch": 0.02332712078398365, + "flos": 23072965090560.0, + "grad_norm": 2.231556080008809, + "language_loss": 0.8765223, + "learning_rate": 3.81627850848061e-06, + "loss": 0.90118676, + "num_input_tokens_seen": 4082345, + "step": 194, + "time_per_iteration": 2.735203504562378 + }, + { + "auxiliary_loss_clip": 0.01360757, + "auxiliary_loss_mlp": 0.01086705, + "balance_loss_clip": 1.10815334, + "balance_loss_mlp": 1.06570053, + "epoch": 0.02344736367462274, + "flos": 24425971614720.0, + "grad_norm": 2.865457545223642, + "language_loss": 0.86196351, + "learning_rate": 3.820003174525994e-06, + "loss": 0.88643813, + "num_input_tokens_seen": 4101770, + "step": 195, + "time_per_iteration": 2.8065381050109863 + }, + { + "auxiliary_loss_clip": 0.01363658, + "auxiliary_loss_mlp": 0.01079315, + "balance_loss_clip": 1.11050081, + "balance_loss_mlp": 1.05726099, + "epoch": 0.02356760656526183, + "flos": 21579799697280.0, + "grad_norm": 2.656544916505495, + "language_loss": 0.82840598, + "learning_rate": 3.823708788487851e-06, + "loss": 0.85283566, + "num_input_tokens_seen": 4118770, + "step": 196, + "time_per_iteration": 2.747311592102051 + }, + { + "auxiliary_loss_clip": 0.01363462, + "auxiliary_loss_mlp": 0.01076717, + "balance_loss_clip": 1.11176109, + "balance_loss_mlp": 1.05654657, + "epoch": 0.02368784945590092, + "flos": 25193598192000.0, + "grad_norm": 5.694323178843094, + "language_loss": 0.84602571, + "learning_rate": 3.827395544281781e-06, + "loss": 0.87042749, + "num_input_tokens_seen": 4141110, + "step": 197, + "time_per_iteration": 3.6780877113342285 + }, + { + "auxiliary_loss_clip": 0.01364344, + "auxiliary_loss_mlp": 0.01071275, + "balance_loss_clip": 1.11140263, + "balance_loss_mlp": 1.04863763, + "epoch": 0.02380809234654001, + "flos": 27562481164800.0, + "grad_norm": 2.2973550910900387, + "language_loss": 0.79228443, + "learning_rate": 3.831063632877802e-06, + "loss": 0.81664062, + "num_input_tokens_seen": 4161430, + "step": 198, + "time_per_iteration": 3.682586669921875 + }, + { + "auxiliary_loss_clip": 0.01363098, + "auxiliary_loss_mlp": 0.01091158, + "balance_loss_clip": 1.11449647, + "balance_loss_mlp": 1.06976008, + "epoch": 0.0239283352371791, + "flos": 18259786540800.0, + "grad_norm": 2.52098730189932, + "language_loss": 0.75736284, + "learning_rate": 3.834713242359712e-06, + "loss": 0.78190541, + "num_input_tokens_seen": 4179260, + "step": 199, + "time_per_iteration": 3.7489328384399414 + }, + { + "auxiliary_loss_clip": 0.01360941, + "auxiliary_loss_mlp": 0.01085668, + "balance_loss_clip": 1.10733271, + "balance_loss_mlp": 1.06164718, + "epoch": 0.02404857812781819, + "flos": 21395110942080.0, + "grad_norm": 2.080983707036262, + "language_loss": 0.87226045, + "learning_rate": 3.838344557982959e-06, + "loss": 0.89672649, + "num_input_tokens_seen": 4200640, + "step": 200, + "time_per_iteration": 2.900596857070923 + }, + { + "auxiliary_loss_clip": 0.0135909, + "auxiliary_loss_mlp": 0.01074025, + "balance_loss_clip": 1.10715318, + "balance_loss_mlp": 1.05340159, + "epoch": 0.024168821018457284, + "flos": 16654256426880.0, + "grad_norm": 3.0803565871809457, + "language_loss": 0.8497833, + "learning_rate": 3.841957762231063e-06, + "loss": 0.87411439, + "num_input_tokens_seen": 4218170, + "step": 201, + "time_per_iteration": 3.694302797317505 + }, + { + "auxiliary_loss_clip": 0.01356894, + "auxiliary_loss_mlp": 0.01069438, + "balance_loss_clip": 1.10513818, + "balance_loss_mlp": 1.04678845, + "epoch": 0.024289063909096374, + "flos": 22820872464000.0, + "grad_norm": 2.0767712842709036, + "language_loss": 0.87909186, + "learning_rate": 3.8455530348706454e-06, + "loss": 0.90335524, + "num_input_tokens_seen": 4237770, + "step": 202, + "time_per_iteration": 2.760685443878174 + }, + { + "auxiliary_loss_clip": 0.01358527, + "auxiliary_loss_mlp": 0.01069758, + "balance_loss_clip": 1.10941875, + "balance_loss_mlp": 1.04841983, + "epoch": 0.024409306799735464, + "flos": 17748598135680.0, + "grad_norm": 2.0390525321275255, + "language_loss": 0.7744897, + "learning_rate": 3.849130553005099e-06, + "loss": 0.79877257, + "num_input_tokens_seen": 4255985, + "step": 203, + "time_per_iteration": 2.720613718032837 + }, + { + "auxiliary_loss_clip": 0.01356882, + "auxiliary_loss_mlp": 0.01062777, + "balance_loss_clip": 1.10582709, + "balance_loss_mlp": 1.04073536, + "epoch": 0.024529549690374557, + "flos": 21616213109760.0, + "grad_norm": 1.8275557051075897, + "language_loss": 0.83713204, + "learning_rate": 3.852690491126933e-06, + "loss": 0.8613286, + "num_input_tokens_seen": 4276035, + "step": 204, + "time_per_iteration": 2.7462048530578613 + }, + { + "auxiliary_loss_clip": 0.01357049, + "auxiliary_loss_mlp": 0.01075299, + "balance_loss_clip": 1.10744572, + "balance_loss_mlp": 1.05194569, + "epoch": 0.024649792581013647, + "flos": 25551662918400.0, + "grad_norm": 2.40480115463454, + "language_loss": 0.91220093, + "learning_rate": 3.856233021168845e-06, + "loss": 0.93652439, + "num_input_tokens_seen": 4295730, + "step": 205, + "time_per_iteration": 2.797361373901367 + }, + { + "auxiliary_loss_clip": 0.01350075, + "auxiliary_loss_mlp": 0.01061978, + "balance_loss_clip": 1.10179269, + "balance_loss_mlp": 1.04207051, + "epoch": 0.024770035471652737, + "flos": 34495574544000.0, + "grad_norm": 2.2041589487175957, + "language_loss": 0.91419685, + "learning_rate": 3.859758312553544e-06, + "loss": 0.93831742, + "num_input_tokens_seen": 4317950, + "step": 206, + "time_per_iteration": 2.9867398738861084 + }, + { + "auxiliary_loss_clip": 0.01357149, + "auxiliary_loss_mlp": 0.01076495, + "balance_loss_clip": 1.10781121, + "balance_loss_mlp": 1.05386901, + "epoch": 0.02489027836229183, + "flos": 21505428587520.0, + "grad_norm": 2.4426927587254226, + "language_loss": 0.91997921, + "learning_rate": 3.8632665322423735e-06, + "loss": 0.94431567, + "num_input_tokens_seen": 4337605, + "step": 207, + "time_per_iteration": 2.7333106994628906 + }, + { + "auxiliary_loss_clip": 0.01351153, + "auxiliary_loss_mlp": 0.01072847, + "balance_loss_clip": 1.10444236, + "balance_loss_mlp": 1.05397606, + "epoch": 0.02501052125293092, + "flos": 23219013790080.0, + "grad_norm": 1.8106499896474375, + "language_loss": 0.85895085, + "learning_rate": 3.866757844782762e-06, + "loss": 0.88319081, + "num_input_tokens_seen": 4358110, + "step": 208, + "time_per_iteration": 2.801748275756836 + }, + { + "auxiliary_loss_clip": 0.01350916, + "auxiliary_loss_mlp": 0.01072601, + "balance_loss_clip": 1.10488701, + "balance_loss_mlp": 1.05048752, + "epoch": 0.02513076414357001, + "flos": 26388920010240.0, + "grad_norm": 4.427335301200986, + "language_loss": 0.9141677, + "learning_rate": 3.870232412354527e-06, + "loss": 0.93840295, + "num_input_tokens_seen": 4374955, + "step": 209, + "time_per_iteration": 2.74802303314209 + }, + { + "auxiliary_loss_clip": 0.01348167, + "auxiliary_loss_mlp": 0.01072978, + "balance_loss_clip": 1.10163045, + "balance_loss_mlp": 1.05023265, + "epoch": 0.025251007034209103, + "flos": 13590430047360.0, + "grad_norm": 2.0703077059675485, + "language_loss": 0.92425954, + "learning_rate": 3.873690394815086e-06, + "loss": 0.94847095, + "num_input_tokens_seen": 4391535, + "step": 210, + "time_per_iteration": 2.7552807331085205 + }, + { + "auxiliary_loss_clip": 0.01354747, + "auxiliary_loss_mlp": 0.01077276, + "balance_loss_clip": 1.10483384, + "balance_loss_mlp": 1.05676019, + "epoch": 0.025371249924848193, + "flos": 15049229103360.0, + "grad_norm": 4.946842153183125, + "language_loss": 0.91376889, + "learning_rate": 3.877131949743587e-06, + "loss": 0.93808925, + "num_input_tokens_seen": 4408400, + "step": 211, + "time_per_iteration": 2.7201688289642334 + }, + { + "auxiliary_loss_clip": 0.01350728, + "auxiliary_loss_mlp": 0.01079846, + "balance_loss_clip": 1.10402966, + "balance_loss_mlp": 1.05755401, + "epoch": 0.025491492815487283, + "flos": 25553853648000.0, + "grad_norm": 4.015490911377151, + "language_loss": 0.7815659, + "learning_rate": 3.880557232483993e-06, + "loss": 0.80587167, + "num_input_tokens_seen": 4427840, + "step": 212, + "time_per_iteration": 2.815849542617798 + }, + { + "auxiliary_loss_clip": 0.01346691, + "auxiliary_loss_mlp": 0.01071097, + "balance_loss_clip": 1.0998385, + "balance_loss_mlp": 1.05040216, + "epoch": 0.025611735706126376, + "flos": 20630752502400.0, + "grad_norm": 2.038314591512367, + "language_loss": 0.86884731, + "learning_rate": 3.883966396187164e-06, + "loss": 0.89302516, + "num_input_tokens_seen": 4447110, + "step": 213, + "time_per_iteration": 2.7520687580108643 + }, + { + "auxiliary_loss_clip": 0.01348198, + "auxiliary_loss_mlp": 0.01070995, + "balance_loss_clip": 1.10450757, + "balance_loss_mlp": 1.04984736, + "epoch": 0.025731978596765466, + "flos": 19062282245760.0, + "grad_norm": 2.3143654300448917, + "language_loss": 0.90112615, + "learning_rate": 3.887359591851937e-06, + "loss": 0.925318, + "num_input_tokens_seen": 4464715, + "step": 214, + "time_per_iteration": 2.709774971008301 + }, + { + "auxiliary_loss_clip": 0.01343533, + "auxiliary_loss_mlp": 0.01069333, + "balance_loss_clip": 1.10042, + "balance_loss_mlp": 1.04987836, + "epoch": 0.025852221487404556, + "flos": 22163814927360.0, + "grad_norm": 2.1451496956369143, + "language_loss": 0.92294508, + "learning_rate": 3.890736968365265e-06, + "loss": 0.9470737, + "num_input_tokens_seen": 4485030, + "step": 215, + "time_per_iteration": 2.769404888153076 + }, + { + "auxiliary_loss_clip": 0.01342385, + "auxiliary_loss_mlp": 0.01079251, + "balance_loss_clip": 1.0996834, + "balance_loss_mlp": 1.05780482, + "epoch": 0.02597246437804365, + "flos": 26541971861760.0, + "grad_norm": 2.987923513296819, + "language_loss": 0.85164893, + "learning_rate": 3.894098672541412e-06, + "loss": 0.87586528, + "num_input_tokens_seen": 4505935, + "step": 216, + "time_per_iteration": 2.7365243434906006 + }, + { + "auxiliary_loss_clip": 0.01344342, + "auxiliary_loss_mlp": 0.01076395, + "balance_loss_clip": 1.10174823, + "balance_loss_mlp": 1.05438852, + "epoch": 0.02609270726868274, + "flos": 32671671696000.0, + "grad_norm": 2.482191111741443, + "language_loss": 0.75451708, + "learning_rate": 3.89744484916025e-06, + "loss": 0.77872443, + "num_input_tokens_seen": 4527045, + "step": 217, + "time_per_iteration": 2.800260305404663 + }, + { + "auxiliary_loss_clip": 0.01345583, + "auxiliary_loss_mlp": 0.01078876, + "balance_loss_clip": 1.10173464, + "balance_loss_mlp": 1.05558276, + "epoch": 0.02621295015932183, + "flos": 26243553669120.0, + "grad_norm": 1.851296659977376, + "language_loss": 0.8741731, + "learning_rate": 3.900775641004673e-06, + "loss": 0.89841771, + "num_input_tokens_seen": 4546360, + "step": 218, + "time_per_iteration": 2.7428555488586426 + }, + { + "auxiliary_loss_clip": 0.0135124, + "auxiliary_loss_mlp": 0.01085141, + "balance_loss_clip": 1.1030122, + "balance_loss_mlp": 1.06238425, + "epoch": 0.026333193049960922, + "flos": 42921402353280.0, + "grad_norm": 3.6000250241910963, + "language_loss": 0.73643303, + "learning_rate": 3.904091188897156e-06, + "loss": 0.76079684, + "num_input_tokens_seen": 4565495, + "step": 219, + "time_per_iteration": 2.8790838718414307 + }, + { + "auxiliary_loss_clip": 0.0134569, + "auxiliary_loss_mlp": 0.01073991, + "balance_loss_clip": 1.10276496, + "balance_loss_mlp": 1.05143654, + "epoch": 0.026453435940600012, + "flos": 17963846386560.0, + "grad_norm": 2.4523582509046307, + "language_loss": 0.81966025, + "learning_rate": 3.90739163173548e-06, + "loss": 0.84385705, + "num_input_tokens_seen": 4583330, + "step": 220, + "time_per_iteration": 2.797483444213867 + }, + { + "auxiliary_loss_clip": 0.01344149, + "auxiliary_loss_mlp": 0.01071641, + "balance_loss_clip": 1.10012674, + "balance_loss_mlp": 1.05038631, + "epoch": 0.026573678831239102, + "flos": 18984319776000.0, + "grad_norm": 2.4753213471419517, + "language_loss": 0.88518447, + "learning_rate": 3.910677106527646e-06, + "loss": 0.90934241, + "num_input_tokens_seen": 4600520, + "step": 221, + "time_per_iteration": 2.678847312927246 + }, + { + "auxiliary_loss_clip": 0.01339469, + "auxiliary_loss_mlp": 0.01073727, + "balance_loss_clip": 1.09949505, + "balance_loss_mlp": 1.05148292, + "epoch": 0.026693921721878195, + "flos": 29241448634880.0, + "grad_norm": 2.867785772014399, + "language_loss": 0.84243143, + "learning_rate": 3.913947748426004e-06, + "loss": 0.86656344, + "num_input_tokens_seen": 4617340, + "step": 222, + "time_per_iteration": 2.7915587425231934 + }, + { + "auxiliary_loss_clip": 0.01345864, + "auxiliary_loss_mlp": 0.0106612, + "balance_loss_clip": 1.10322285, + "balance_loss_mlp": 1.04454303, + "epoch": 0.026814164612517285, + "flos": 14128083797760.0, + "grad_norm": 3.0682302077015673, + "language_loss": 0.75991642, + "learning_rate": 3.9172036907606136e-06, + "loss": 0.78403628, + "num_input_tokens_seen": 4630820, + "step": 223, + "time_per_iteration": 2.6953227519989014 + }, + { + "auxiliary_loss_clip": 0.01342103, + "auxiliary_loss_mlp": 0.01075554, + "balance_loss_clip": 1.09859109, + "balance_loss_mlp": 1.05487144, + "epoch": 0.026934407503156375, + "flos": 23511973115520.0, + "grad_norm": 1.899332505300615, + "language_loss": 0.94918978, + "learning_rate": 3.920445065071855e-06, + "loss": 0.97336638, + "num_input_tokens_seen": 4651985, + "step": 224, + "time_per_iteration": 3.6281745433807373 + }, + { + "auxiliary_loss_clip": 0.01342109, + "auxiliary_loss_mlp": 0.01068479, + "balance_loss_clip": 1.10118461, + "balance_loss_mlp": 1.04783177, + "epoch": 0.027054650393795468, + "flos": 28950356816640.0, + "grad_norm": 3.194891665179141, + "language_loss": 0.80132067, + "learning_rate": 3.923672001142322e-06, + "loss": 0.82542658, + "num_input_tokens_seen": 4672295, + "step": 225, + "time_per_iteration": 3.6979987621307373 + }, + { + "auxiliary_loss_clip": 0.01336616, + "auxiliary_loss_mlp": 0.01063661, + "balance_loss_clip": 1.09664798, + "balance_loss_mlp": 1.04326391, + "epoch": 0.027174893284434558, + "flos": 31431568596480.0, + "grad_norm": 1.9939057834596292, + "language_loss": 0.84559351, + "learning_rate": 3.926884627027996e-06, + "loss": 0.8695963, + "num_input_tokens_seen": 4696065, + "step": 226, + "time_per_iteration": 3.73319411277771 + }, + { + "auxiliary_loss_clip": 0.0133403, + "auxiliary_loss_mlp": 0.01063367, + "balance_loss_clip": 1.09461176, + "balance_loss_mlp": 1.04229033, + "epoch": 0.027295136175073648, + "flos": 22054466949120.0, + "grad_norm": 1.8940788041008758, + "language_loss": 0.77404135, + "learning_rate": 3.930083069088744e-06, + "loss": 0.7980153, + "num_input_tokens_seen": 4716065, + "step": 227, + "time_per_iteration": 3.6557583808898926 + }, + { + "auxiliary_loss_clip": 0.01318362, + "auxiliary_loss_mlp": 0.01014418, + "balance_loss_clip": 1.14055872, + "balance_loss_mlp": 1.00631225, + "epoch": 0.02741537906571274, + "flos": 60800752972800.0, + "grad_norm": 0.9813324084807885, + "language_loss": 0.59289551, + "learning_rate": 3.933267452018137e-06, + "loss": 0.61622334, + "num_input_tokens_seen": 4775860, + "step": 228, + "time_per_iteration": 3.291841983795166 + }, + { + "auxiliary_loss_clip": 0.01337437, + "auxiliary_loss_mlp": 0.01060798, + "balance_loss_clip": 1.09713054, + "balance_loss_mlp": 1.04006791, + "epoch": 0.02753562195635183, + "flos": 24606278910720.0, + "grad_norm": 2.1613488373255363, + "language_loss": 0.84583634, + "learning_rate": 3.936437898872622e-06, + "loss": 0.86981875, + "num_input_tokens_seen": 4795835, + "step": 229, + "time_per_iteration": 2.9418070316314697 + }, + { + "auxiliary_loss_clip": 0.0133553, + "auxiliary_loss_mlp": 0.01070951, + "balance_loss_clip": 1.09561193, + "balance_loss_mlp": 1.05018508, + "epoch": 0.02765586484699092, + "flos": 34094236907520.0, + "grad_norm": 6.450047311462945, + "language_loss": 0.79906422, + "learning_rate": 3.9395945311000525e-06, + "loss": 0.82312906, + "num_input_tokens_seen": 4817460, + "step": 230, + "time_per_iteration": 2.7940518856048584 + }, + { + "auxiliary_loss_clip": 0.01339409, + "auxiliary_loss_mlp": 0.01070729, + "balance_loss_clip": 1.09850538, + "balance_loss_mlp": 1.05070186, + "epoch": 0.027776107737630014, + "flos": 14829922615680.0, + "grad_norm": 5.746856527680883, + "language_loss": 0.90964258, + "learning_rate": 3.942737468567608e-06, + "loss": 0.93374395, + "num_input_tokens_seen": 4835475, + "step": 231, + "time_per_iteration": 2.6473302841186523 + }, + { + "auxiliary_loss_clip": 0.01335956, + "auxiliary_loss_mlp": 0.01074832, + "balance_loss_clip": 1.09641957, + "balance_loss_mlp": 1.05253983, + "epoch": 0.027896350628269104, + "flos": 47920347066240.0, + "grad_norm": 2.34132846338048, + "language_loss": 0.85979587, + "learning_rate": 3.9458668295891026e-06, + "loss": 0.88390374, + "num_input_tokens_seen": 4857760, + "step": 232, + "time_per_iteration": 2.956207513809204 + }, + { + "auxiliary_loss_clip": 0.01332246, + "auxiliary_loss_mlp": 0.0107054, + "balance_loss_clip": 1.09287739, + "balance_loss_mlp": 1.05022645, + "epoch": 0.028016593518908194, + "flos": 21684550734720.0, + "grad_norm": 2.8638214368045363, + "language_loss": 0.86582541, + "learning_rate": 3.948982730951712e-06, + "loss": 0.8898533, + "num_input_tokens_seen": 4875855, + "step": 233, + "time_per_iteration": 2.8191561698913574 + }, + { + "auxiliary_loss_clip": 0.01335909, + "auxiliary_loss_mlp": 0.01066012, + "balance_loss_clip": 1.09471583, + "balance_loss_mlp": 1.04400611, + "epoch": 0.028136836409547287, + "flos": 18439483305600.0, + "grad_norm": 2.412694054171373, + "language_loss": 0.82095981, + "learning_rate": 3.9520852879421254e-06, + "loss": 0.84497893, + "num_input_tokens_seen": 4893200, + "step": 234, + "time_per_iteration": 2.753476142883301 + }, + { + "auxiliary_loss_clip": 0.01333628, + "auxiliary_loss_mlp": 0.0106609, + "balance_loss_clip": 1.0972482, + "balance_loss_mlp": 1.04680204, + "epoch": 0.028257079300186377, + "flos": 31576934937600.0, + "grad_norm": 3.9164402495304813, + "language_loss": 0.82114929, + "learning_rate": 3.955174614372137e-06, + "loss": 0.84514648, + "num_input_tokens_seen": 4912965, + "step": 235, + "time_per_iteration": 2.812668561935425 + }, + { + "auxiliary_loss_clip": 0.01332813, + "auxiliary_loss_mlp": 0.01083324, + "balance_loss_clip": 1.09542465, + "balance_loss_mlp": 1.06162739, + "epoch": 0.028377322190825467, + "flos": 23513337832320.0, + "grad_norm": 2.2174701943368134, + "language_loss": 0.84409469, + "learning_rate": 3.9582508226037045e-06, + "loss": 0.86825609, + "num_input_tokens_seen": 4933105, + "step": 236, + "time_per_iteration": 2.7460832595825195 + }, + { + "auxiliary_loss_clip": 0.01333804, + "auxiliary_loss_mlp": 0.01062909, + "balance_loss_clip": 1.09540987, + "balance_loss_mlp": 1.04023528, + "epoch": 0.02849756508146456, + "flos": 20479604071680.0, + "grad_norm": 5.838133792720258, + "language_loss": 0.93992662, + "learning_rate": 3.9613140235734636e-06, + "loss": 0.96389371, + "num_input_tokens_seen": 4950085, + "step": 237, + "time_per_iteration": 2.7716891765594482 + }, + { + "auxiliary_loss_clip": 0.01333274, + "auxiliary_loss_mlp": 0.010615, + "balance_loss_clip": 1.09605527, + "balance_loss_mlp": 1.04138947, + "epoch": 0.02861780797210365, + "flos": 14283362292480.0, + "grad_norm": 2.3324856018112707, + "language_loss": 0.8093788, + "learning_rate": 3.96436432681674e-06, + "loss": 0.83332658, + "num_input_tokens_seen": 4968075, + "step": 238, + "time_per_iteration": 2.6833083629608154 + }, + { + "auxiliary_loss_clip": 0.01333329, + "auxiliary_loss_mlp": 0.01076063, + "balance_loss_clip": 1.0948422, + "balance_loss_mlp": 1.05460536, + "epoch": 0.02873805086274274, + "flos": 25808532053760.0, + "grad_norm": 2.131936980905413, + "language_loss": 0.89105535, + "learning_rate": 3.967401840491044e-06, + "loss": 0.91514933, + "num_input_tokens_seen": 4987355, + "step": 239, + "time_per_iteration": 2.7790517807006836 + }, + { + "auxiliary_loss_clip": 0.01331971, + "auxiliary_loss_mlp": 0.01068397, + "balance_loss_clip": 1.09646177, + "balance_loss_mlp": 1.04889441, + "epoch": 0.028858293753381833, + "flos": 17304238984320.0, + "grad_norm": 2.9025693043307443, + "language_loss": 0.87766856, + "learning_rate": 3.97042667139909e-06, + "loss": 0.90167224, + "num_input_tokens_seen": 5004680, + "step": 240, + "time_per_iteration": 2.683927297592163 + }, + { + "auxiliary_loss_clip": 0.01334207, + "auxiliary_loss_mlp": 0.01072853, + "balance_loss_clip": 1.09630227, + "balance_loss_mlp": 1.05054891, + "epoch": 0.028978536644020923, + "flos": 23038347358080.0, + "grad_norm": 2.3045496097664557, + "language_loss": 0.87484908, + "learning_rate": 3.973438925011327e-06, + "loss": 0.8989197, + "num_input_tokens_seen": 5022965, + "step": 241, + "time_per_iteration": 2.7558445930480957 + }, + { + "auxiliary_loss_clip": 0.01327476, + "auxiliary_loss_mlp": 0.01060203, + "balance_loss_clip": 1.09019136, + "balance_loss_mlp": 1.03943706, + "epoch": 0.029098779534660012, + "flos": 28329712692480.0, + "grad_norm": 2.375358787678291, + "language_loss": 0.91698551, + "learning_rate": 3.976438705488002e-06, + "loss": 0.9408623, + "num_input_tokens_seen": 5042625, + "step": 242, + "time_per_iteration": 2.761855125427246 + }, + { + "auxiliary_loss_clip": 0.0133139, + "auxiliary_loss_mlp": 0.01069338, + "balance_loss_clip": 1.09556055, + "balance_loss_mlp": 1.04920292, + "epoch": 0.029219022425299106, + "flos": 13881665520000.0, + "grad_norm": 2.7299719870584647, + "language_loss": 0.9334926, + "learning_rate": 3.9794261157007744e-06, + "loss": 0.95749992, + "num_input_tokens_seen": 5060380, + "step": 243, + "time_per_iteration": 2.68790340423584 + }, + { + "auxiliary_loss_clip": 0.01333929, + "auxiliary_loss_mlp": 0.01075823, + "balance_loss_clip": 1.09648633, + "balance_loss_mlp": 1.05461526, + "epoch": 0.029339265315938196, + "flos": 19422501788160.0, + "grad_norm": 3.111544953420139, + "language_loss": 0.85150111, + "learning_rate": 3.982401257253887e-06, + "loss": 0.87559861, + "num_input_tokens_seen": 5078720, + "step": 244, + "time_per_iteration": 2.709484815597534 + }, + { + "auxiliary_loss_clip": 0.0133028, + "auxiliary_loss_mlp": 0.0106285, + "balance_loss_clip": 1.09415364, + "balance_loss_mlp": 1.04319239, + "epoch": 0.029459508206577285, + "flos": 15669550005120.0, + "grad_norm": 2.742263971805704, + "language_loss": 0.89876354, + "learning_rate": 3.985364230504893e-06, + "loss": 0.92269486, + "num_input_tokens_seen": 5096605, + "step": 245, + "time_per_iteration": 2.7257466316223145 + }, + { + "auxiliary_loss_clip": 0.01329784, + "auxiliary_loss_mlp": 0.01068844, + "balance_loss_clip": 1.09452939, + "balance_loss_mlp": 1.04801798, + "epoch": 0.02957975109721638, + "flos": 28220975245440.0, + "grad_norm": 2.139577655032125, + "language_loss": 0.84600508, + "learning_rate": 3.988315134584976e-06, + "loss": 0.8699913, + "num_input_tokens_seen": 5116285, + "step": 246, + "time_per_iteration": 2.8090007305145264 + }, + { + "auxiliary_loss_clip": 0.01330579, + "auxiliary_loss_mlp": 0.01070972, + "balance_loss_clip": 1.09280384, + "balance_loss_mlp": 1.0488106, + "epoch": 0.02969999398785547, + "flos": 24315869450880.0, + "grad_norm": 2.20477236869787, + "language_loss": 0.80438733, + "learning_rate": 3.991254067418851e-06, + "loss": 0.82840288, + "num_input_tokens_seen": 5136825, + "step": 247, + "time_per_iteration": 2.670165538787842 + }, + { + "auxiliary_loss_clip": 0.01330762, + "auxiliary_loss_mlp": 0.01060573, + "balance_loss_clip": 1.09562516, + "balance_loss_mlp": 1.0426563, + "epoch": 0.02982023687849456, + "flos": 35078584193280.0, + "grad_norm": 2.334103548911415, + "language_loss": 0.82863581, + "learning_rate": 3.994181125744254e-06, + "loss": 0.8525492, + "num_input_tokens_seen": 5158630, + "step": 248, + "time_per_iteration": 2.8579413890838623 + }, + { + "auxiliary_loss_clip": 0.01331947, + "auxiliary_loss_mlp": 0.01066673, + "balance_loss_clip": 1.09356868, + "balance_loss_mlp": 1.04641914, + "epoch": 0.02994047976913365, + "flos": 26177155378560.0, + "grad_norm": 5.050729564139736, + "language_loss": 0.74044514, + "learning_rate": 3.99709640513106e-06, + "loss": 0.76443136, + "num_input_tokens_seen": 5179510, + "step": 249, + "time_per_iteration": 2.732138156890869 + }, + { + "auxiliary_loss_clip": 0.01330238, + "auxiliary_loss_mlp": 0.01065643, + "balance_loss_clip": 1.09079337, + "balance_loss_mlp": 1.04386353, + "epoch": 0.03006072265977274, + "flos": 25625028447360.0, + "grad_norm": 2.0923263551039715, + "language_loss": 0.85462916, + "learning_rate": 4e-06, + "loss": 0.87858796, + "num_input_tokens_seen": 5199345, + "step": 250, + "time_per_iteration": 3.665703535079956 + }, + { + "auxiliary_loss_clip": 0.013319, + "auxiliary_loss_mlp": 0.0106752, + "balance_loss_clip": 1.09628868, + "balance_loss_mlp": 1.04866087, + "epoch": 0.03018096555041183, + "flos": 22127078292480.0, + "grad_norm": 2.701712016095574, + "language_loss": 0.88487113, + "learning_rate": 3.999999848300794e-06, + "loss": 0.90886527, + "num_input_tokens_seen": 5218330, + "step": 251, + "time_per_iteration": 3.6519813537597656 + }, + { + "auxiliary_loss_clip": 0.01325648, + "auxiliary_loss_mlp": 0.01063482, + "balance_loss_clip": 1.09181416, + "balance_loss_mlp": 1.04536235, + "epoch": 0.030301208441050925, + "flos": 30188197359360.0, + "grad_norm": 1.7581880027944172, + "language_loss": 0.89358282, + "learning_rate": 3.999999393203203e-06, + "loss": 0.91747415, + "num_input_tokens_seen": 5240740, + "step": 252, + "time_per_iteration": 3.6727476119995117 + }, + { + "auxiliary_loss_clip": 0.01324053, + "auxiliary_loss_mlp": 0.01057698, + "balance_loss_clip": 1.08796632, + "balance_loss_mlp": 1.03827882, + "epoch": 0.030421451331690014, + "flos": 23621392920960.0, + "grad_norm": 1.810393913219936, + "language_loss": 0.85327578, + "learning_rate": 3.999998634707293e-06, + "loss": 0.87709326, + "num_input_tokens_seen": 5260290, + "step": 253, + "time_per_iteration": 3.6450111865997314 + }, + { + "auxiliary_loss_clip": 0.01332316, + "auxiliary_loss_mlp": 0.01071135, + "balance_loss_clip": 1.09673178, + "balance_loss_mlp": 1.05104804, + "epoch": 0.030541694222329104, + "flos": 27928446883200.0, + "grad_norm": 2.274118123263435, + "language_loss": 0.96315563, + "learning_rate": 3.999997572813182e-06, + "loss": 0.98719013, + "num_input_tokens_seen": 5278100, + "step": 254, + "time_per_iteration": 2.7060842514038086 + }, + { + "auxiliary_loss_clip": 0.01321632, + "auxiliary_loss_mlp": 0.0105823, + "balance_loss_clip": 1.08817124, + "balance_loss_mlp": 1.03937101, + "epoch": 0.030661937112968194, + "flos": 18588441006720.0, + "grad_norm": 2.4471916782342698, + "language_loss": 0.87563723, + "learning_rate": 3.999996207521028e-06, + "loss": 0.89943588, + "num_input_tokens_seen": 5296810, + "step": 255, + "time_per_iteration": 2.733682870864868 + }, + { + "auxiliary_loss_clip": 0.01327613, + "auxiliary_loss_mlp": 0.0106244, + "balance_loss_clip": 1.09223247, + "balance_loss_mlp": 1.04232991, + "epoch": 0.030782180003607287, + "flos": 12969139478400.0, + "grad_norm": 2.524024614753512, + "language_loss": 0.82007045, + "learning_rate": 3.999994538831039e-06, + "loss": 0.84397101, + "num_input_tokens_seen": 5313395, + "step": 256, + "time_per_iteration": 2.7085471153259277 + }, + { + "auxiliary_loss_clip": 0.01325823, + "auxiliary_loss_mlp": 0.01060949, + "balance_loss_clip": 1.08917308, + "balance_loss_mlp": 1.03990865, + "epoch": 0.030902422894246377, + "flos": 23335364920320.0, + "grad_norm": 7.478504934985832, + "language_loss": 0.8562929, + "learning_rate": 3.99999256674347e-06, + "loss": 0.88016063, + "num_input_tokens_seen": 5333545, + "step": 257, + "time_per_iteration": 2.678931474685669 + }, + { + "auxiliary_loss_clip": 0.01291241, + "auxiliary_loss_mlp": 0.01027957, + "balance_loss_clip": 1.11998963, + "balance_loss_mlp": 1.02042341, + "epoch": 0.031022665784885467, + "flos": 55094151438720.0, + "grad_norm": 1.0176369516387302, + "language_loss": 0.5346595, + "learning_rate": 3.999990291258618e-06, + "loss": 0.55785143, + "num_input_tokens_seen": 5392235, + "step": 258, + "time_per_iteration": 3.239187240600586 + }, + { + "auxiliary_loss_clip": 0.01326963, + "auxiliary_loss_mlp": 0.01077625, + "balance_loss_clip": 1.09219849, + "balance_loss_mlp": 1.05609572, + "epoch": 0.03114290867552456, + "flos": 19317786664320.0, + "grad_norm": 2.2312271967541006, + "language_loss": 0.86930174, + "learning_rate": 3.999987712376829e-06, + "loss": 0.89334762, + "num_input_tokens_seen": 5410555, + "step": 259, + "time_per_iteration": 2.6921589374542236 + }, + { + "auxiliary_loss_clip": 0.01325252, + "auxiliary_loss_mlp": 0.01067923, + "balance_loss_clip": 1.09208441, + "balance_loss_mlp": 1.04691815, + "epoch": 0.031263151566163654, + "flos": 20959442881920.0, + "grad_norm": 2.71250932141547, + "language_loss": 0.82581884, + "learning_rate": 3.999984830098494e-06, + "loss": 0.84975058, + "num_input_tokens_seen": 5430135, + "step": 260, + "time_per_iteration": 2.6835200786590576 + }, + { + "auxiliary_loss_clip": 0.01325551, + "auxiliary_loss_mlp": 0.01070687, + "balance_loss_clip": 1.09085751, + "balance_loss_mlp": 1.04832351, + "epoch": 0.03138339445680274, + "flos": 14793006412800.0, + "grad_norm": 2.648487563053308, + "language_loss": 0.97936118, + "learning_rate": 3.999981644424051e-06, + "loss": 1.00332355, + "num_input_tokens_seen": 5444935, + "step": 261, + "time_per_iteration": 2.7563650608062744 + }, + { + "auxiliary_loss_clip": 0.01324125, + "auxiliary_loss_mlp": 0.01062813, + "balance_loss_clip": 1.08810306, + "balance_loss_mlp": 1.04201102, + "epoch": 0.03150363734744183, + "flos": 11655599022720.0, + "grad_norm": 2.8187545692333216, + "language_loss": 0.86042732, + "learning_rate": 3.999978155353982e-06, + "loss": 0.88429666, + "num_input_tokens_seen": 5462080, + "step": 262, + "time_per_iteration": 2.7677834033966064 + }, + { + "auxiliary_loss_clip": 0.01320485, + "auxiliary_loss_mlp": 0.01058096, + "balance_loss_clip": 1.08684635, + "balance_loss_mlp": 1.03690064, + "epoch": 0.03162388023808092, + "flos": 33727732485120.0, + "grad_norm": 2.5427218082598824, + "language_loss": 0.80451584, + "learning_rate": 3.9999743628888186e-06, + "loss": 0.82830167, + "num_input_tokens_seen": 5483870, + "step": 263, + "time_per_iteration": 2.8175301551818848 + }, + { + "auxiliary_loss_clip": 0.01319409, + "auxiliary_loss_mlp": 0.01063792, + "balance_loss_clip": 1.0854466, + "balance_loss_mlp": 1.04434943, + "epoch": 0.03174412312872001, + "flos": 20810952057600.0, + "grad_norm": 2.416474620010925, + "language_loss": 0.89548063, + "learning_rate": 3.999970267029133e-06, + "loss": 0.9193126, + "num_input_tokens_seen": 5502830, + "step": 264, + "time_per_iteration": 2.757908821105957 + }, + { + "auxiliary_loss_clip": 0.0131674, + "auxiliary_loss_mlp": 0.01063799, + "balance_loss_clip": 1.08823204, + "balance_loss_mlp": 1.04403377, + "epoch": 0.0318643660193591, + "flos": 23727939638400.0, + "grad_norm": 1.879746932705357, + "language_loss": 0.80074388, + "learning_rate": 3.999965867775548e-06, + "loss": 0.82454926, + "num_input_tokens_seen": 5523225, + "step": 265, + "time_per_iteration": 2.687288284301758 + }, + { + "auxiliary_loss_clip": 0.01317779, + "auxiliary_loss_mlp": 0.01059813, + "balance_loss_clip": 1.08646905, + "balance_loss_mlp": 1.04132366, + "epoch": 0.0319846089099982, + "flos": 13917863450880.0, + "grad_norm": 4.64237222666478, + "language_loss": 0.86900175, + "learning_rate": 3.9999611651287315e-06, + "loss": 0.89277768, + "num_input_tokens_seen": 5541380, + "step": 266, + "time_per_iteration": 2.653932571411133 + }, + { + "auxiliary_loss_clip": 0.01318942, + "auxiliary_loss_mlp": 0.01061476, + "balance_loss_clip": 1.08731794, + "balance_loss_mlp": 1.04156804, + "epoch": 0.03210485180063729, + "flos": 14753253035520.0, + "grad_norm": 2.9213114800962003, + "language_loss": 0.79033923, + "learning_rate": 3.999956159089396e-06, + "loss": 0.81414342, + "num_input_tokens_seen": 5558830, + "step": 267, + "time_per_iteration": 2.6636571884155273 + }, + { + "auxiliary_loss_clip": 0.0131843, + "auxiliary_loss_mlp": 0.01062511, + "balance_loss_clip": 1.08781934, + "balance_loss_mlp": 1.04403305, + "epoch": 0.03222509469127638, + "flos": 28913153304960.0, + "grad_norm": 2.0589075115570727, + "language_loss": 0.79885185, + "learning_rate": 3.999950849658302e-06, + "loss": 0.82266128, + "num_input_tokens_seen": 5577750, + "step": 268, + "time_per_iteration": 2.691850423812866 + }, + { + "auxiliary_loss_clip": 0.01321222, + "auxiliary_loss_mlp": 0.0106287, + "balance_loss_clip": 1.0885241, + "balance_loss_mlp": 1.04429734, + "epoch": 0.03234533758191547, + "flos": 16946389739520.0, + "grad_norm": 2.157258030556821, + "language_loss": 0.84268278, + "learning_rate": 3.999945236836254e-06, + "loss": 0.86652368, + "num_input_tokens_seen": 5596715, + "step": 269, + "time_per_iteration": 2.544304370880127 + }, + { + "auxiliary_loss_clip": 0.01326039, + "auxiliary_loss_mlp": 0.01058719, + "balance_loss_clip": 1.09146142, + "balance_loss_mlp": 1.03906143, + "epoch": 0.03246558047255456, + "flos": 18989096284800.0, + "grad_norm": 3.0546804813048105, + "language_loss": 0.94499832, + "learning_rate": 3.999939320624103e-06, + "loss": 0.9688459, + "num_input_tokens_seen": 5611865, + "step": 270, + "time_per_iteration": 2.568237781524658 + }, + { + "auxiliary_loss_clip": 0.0131682, + "auxiliary_loss_mlp": 0.01062202, + "balance_loss_clip": 1.08530116, + "balance_loss_mlp": 1.04216254, + "epoch": 0.03258582336319365, + "flos": 23728334688000.0, + "grad_norm": 1.9065118601945872, + "language_loss": 0.89821553, + "learning_rate": 3.999933101022749e-06, + "loss": 0.92200565, + "num_input_tokens_seen": 5632270, + "step": 271, + "time_per_iteration": 2.5680480003356934 + }, + { + "auxiliary_loss_clip": 0.01320853, + "auxiliary_loss_mlp": 0.01064042, + "balance_loss_clip": 1.08761334, + "balance_loss_mlp": 1.04419327, + "epoch": 0.032706066253832745, + "flos": 27670823562240.0, + "grad_norm": 1.8453232763762861, + "language_loss": 0.86981672, + "learning_rate": 3.999926578033132e-06, + "loss": 0.89366567, + "num_input_tokens_seen": 5652085, + "step": 272, + "time_per_iteration": 2.6887123584747314 + }, + { + "auxiliary_loss_clip": 0.01320071, + "auxiliary_loss_mlp": 0.01054513, + "balance_loss_clip": 1.08647799, + "balance_loss_mlp": 1.03470039, + "epoch": 0.032826309144471835, + "flos": 45624685968000.0, + "grad_norm": 3.314275561627356, + "language_loss": 0.62849534, + "learning_rate": 3.999919751656244e-06, + "loss": 0.65224111, + "num_input_tokens_seen": 5678985, + "step": 273, + "time_per_iteration": 2.8393986225128174 + }, + { + "auxiliary_loss_clip": 0.01319689, + "auxiliary_loss_mlp": 0.01064148, + "balance_loss_clip": 1.0857383, + "balance_loss_mlp": 1.04580188, + "epoch": 0.032946552035110925, + "flos": 25812374808960.0, + "grad_norm": 2.7339047958606373, + "language_loss": 0.75980151, + "learning_rate": 3.9999126218931195e-06, + "loss": 0.78363991, + "num_input_tokens_seen": 5697020, + "step": 274, + "time_per_iteration": 2.7169058322906494 + }, + { + "auxiliary_loss_clip": 0.01325569, + "auxiliary_loss_mlp": 0.01074045, + "balance_loss_clip": 1.09353876, + "balance_loss_mlp": 1.05622351, + "epoch": 0.033066794925750015, + "flos": 15121984101120.0, + "grad_norm": 4.3102259833700565, + "language_loss": 0.89613712, + "learning_rate": 3.99990518874484e-06, + "loss": 0.92013323, + "num_input_tokens_seen": 5713460, + "step": 275, + "time_per_iteration": 2.7765140533447266 + }, + { + "auxiliary_loss_clip": 0.01321281, + "auxiliary_loss_mlp": 0.01062733, + "balance_loss_clip": 1.09099734, + "balance_loss_mlp": 1.04294407, + "epoch": 0.033187037816389105, + "flos": 22776593973120.0, + "grad_norm": 2.286162428781538, + "language_loss": 0.92337549, + "learning_rate": 3.999897452212534e-06, + "loss": 0.94721556, + "num_input_tokens_seen": 5730790, + "step": 276, + "time_per_iteration": 2.784194231033325 + }, + { + "auxiliary_loss_clip": 0.01313799, + "auxiliary_loss_mlp": 0.01064495, + "balance_loss_clip": 1.08472705, + "balance_loss_mlp": 1.04313231, + "epoch": 0.033307280707028195, + "flos": 23331414424320.0, + "grad_norm": 8.497900163708918, + "language_loss": 1.00254107, + "learning_rate": 3.999889412297374e-06, + "loss": 1.02632403, + "num_input_tokens_seen": 5750215, + "step": 277, + "time_per_iteration": 3.616765260696411 + }, + { + "auxiliary_loss_clip": 0.01314652, + "auxiliary_loss_mlp": 0.01068726, + "balance_loss_clip": 1.083915, + "balance_loss_mlp": 1.04927146, + "epoch": 0.03342752359766729, + "flos": 28840290566400.0, + "grad_norm": 2.037054111428489, + "language_loss": 0.79026896, + "learning_rate": 3.999881069000581e-06, + "loss": 0.81410277, + "num_input_tokens_seen": 5769945, + "step": 278, + "time_per_iteration": 3.6090965270996094 + }, + { + "auxiliary_loss_clip": 0.01312392, + "auxiliary_loss_mlp": 0.01058167, + "balance_loss_clip": 1.08466995, + "balance_loss_mlp": 1.03988016, + "epoch": 0.03354776648830638, + "flos": 19384544090880.0, + "grad_norm": 4.945616068283533, + "language_loss": 0.86639559, + "learning_rate": 3.99987242232342e-06, + "loss": 0.89010119, + "num_input_tokens_seen": 5784950, + "step": 279, + "time_per_iteration": 3.512218952178955 + }, + { + "auxiliary_loss_clip": 0.0132229, + "auxiliary_loss_mlp": 0.01064749, + "balance_loss_clip": 1.09070015, + "balance_loss_mlp": 1.04591346, + "epoch": 0.03366800937894547, + "flos": 17858628472320.0, + "grad_norm": 2.437302638835073, + "language_loss": 0.79707766, + "learning_rate": 3.9998634722672026e-06, + "loss": 0.82094806, + "num_input_tokens_seen": 5805005, + "step": 280, + "time_per_iteration": 3.551593780517578 + }, + { + "auxiliary_loss_clip": 0.01316057, + "auxiliary_loss_mlp": 0.01073354, + "balance_loss_clip": 1.08786809, + "balance_loss_mlp": 1.05392313, + "epoch": 0.03378825226958456, + "flos": 35951033635200.0, + "grad_norm": 2.4504346188176807, + "language_loss": 0.78687513, + "learning_rate": 3.999854218833286e-06, + "loss": 0.8107692, + "num_input_tokens_seen": 5825825, + "step": 281, + "time_per_iteration": 2.7615606784820557 + }, + { + "auxiliary_loss_clip": 0.01316116, + "auxiliary_loss_mlp": 0.01061381, + "balance_loss_clip": 1.0876298, + "balance_loss_mlp": 1.0426892, + "epoch": 0.03390849516022365, + "flos": 25702488126720.0, + "grad_norm": 2.1167398400726425, + "language_loss": 0.82297277, + "learning_rate": 3.999844662023075e-06, + "loss": 0.84674776, + "num_input_tokens_seen": 5845700, + "step": 282, + "time_per_iteration": 2.7520153522491455 + }, + { + "auxiliary_loss_clip": 0.01309391, + "auxiliary_loss_mlp": 0.01059037, + "balance_loss_clip": 1.08418608, + "balance_loss_mlp": 1.0406425, + "epoch": 0.03402873805086274, + "flos": 21284505987840.0, + "grad_norm": 1.8396917086096978, + "language_loss": 0.9218924, + "learning_rate": 3.999834801838018e-06, + "loss": 0.94557667, + "num_input_tokens_seen": 5864680, + "step": 283, + "time_per_iteration": 2.6449475288391113 + }, + { + "auxiliary_loss_clip": 0.01315293, + "auxiliary_loss_mlp": 0.01055082, + "balance_loss_clip": 1.08820319, + "balance_loss_mlp": 1.03805852, + "epoch": 0.03414898094150183, + "flos": 22710913954560.0, + "grad_norm": 2.149616163567461, + "language_loss": 0.74183631, + "learning_rate": 3.9998246382796115e-06, + "loss": 0.76554012, + "num_input_tokens_seen": 5884260, + "step": 284, + "time_per_iteration": 2.713822841644287 + }, + { + "auxiliary_loss_clip": 0.013149, + "auxiliary_loss_mlp": 0.01064946, + "balance_loss_clip": 1.08348274, + "balance_loss_mlp": 1.04601598, + "epoch": 0.03426922383214093, + "flos": 18879927874560.0, + "grad_norm": 2.678020183692347, + "language_loss": 0.90853894, + "learning_rate": 3.999814171349399e-06, + "loss": 0.9323374, + "num_input_tokens_seen": 5902120, + "step": 285, + "time_per_iteration": 2.685696601867676 + }, + { + "auxiliary_loss_clip": 0.0130895, + "auxiliary_loss_mlp": 0.01067756, + "balance_loss_clip": 1.08398104, + "balance_loss_mlp": 1.04912388, + "epoch": 0.03438946672278002, + "flos": 34752012716160.0, + "grad_norm": 2.043309688901355, + "language_loss": 0.73876262, + "learning_rate": 3.9998034010489655e-06, + "loss": 0.76252973, + "num_input_tokens_seen": 5925810, + "step": 286, + "time_per_iteration": 2.8782293796539307 + }, + { + "auxiliary_loss_clip": 0.01316024, + "auxiliary_loss_mlp": 0.01061241, + "balance_loss_clip": 1.08879411, + "balance_loss_mlp": 1.04285908, + "epoch": 0.03450970961341911, + "flos": 22164102236160.0, + "grad_norm": 2.8168624646292724, + "language_loss": 0.76099014, + "learning_rate": 3.999792327379946e-06, + "loss": 0.7847628, + "num_input_tokens_seen": 5945185, + "step": 287, + "time_per_iteration": 2.697841167449951 + }, + { + "auxiliary_loss_clip": 0.01316228, + "auxiliary_loss_mlp": 0.0108172, + "balance_loss_clip": 1.08923674, + "balance_loss_mlp": 1.06243217, + "epoch": 0.034629952504058197, + "flos": 21725740656000.0, + "grad_norm": 2.193213600347209, + "language_loss": 0.96203339, + "learning_rate": 3.999780950344021e-06, + "loss": 0.98601294, + "num_input_tokens_seen": 5963375, + "step": 288, + "time_per_iteration": 2.6686084270477295 + }, + { + "auxiliary_loss_clip": 0.01318322, + "auxiliary_loss_mlp": 0.01059736, + "balance_loss_clip": 1.08838439, + "balance_loss_mlp": 1.03869605, + "epoch": 0.034750195394697286, + "flos": 20047994248320.0, + "grad_norm": 2.036711624077596, + "language_loss": 0.82453156, + "learning_rate": 3.999769269942916e-06, + "loss": 0.84831214, + "num_input_tokens_seen": 5983415, + "step": 289, + "time_per_iteration": 2.7607176303863525 + }, + { + "auxiliary_loss_clip": 0.0131339, + "auxiliary_loss_mlp": 0.01059222, + "balance_loss_clip": 1.08652782, + "balance_loss_mlp": 1.04160273, + "epoch": 0.034870438285336376, + "flos": 27965865876480.0, + "grad_norm": 2.434659833245424, + "language_loss": 0.80984998, + "learning_rate": 3.999757286178402e-06, + "loss": 0.83357608, + "num_input_tokens_seen": 6005850, + "step": 290, + "time_per_iteration": 2.6931276321411133 + }, + { + "auxiliary_loss_clip": 0.01314257, + "auxiliary_loss_mlp": 0.01072442, + "balance_loss_clip": 1.08628845, + "balance_loss_mlp": 1.05304658, + "epoch": 0.03499068117597547, + "flos": 22017514832640.0, + "grad_norm": 1.9095341065316793, + "language_loss": 0.90741825, + "learning_rate": 3.999744999052299e-06, + "loss": 0.93128526, + "num_input_tokens_seen": 6027240, + "step": 291, + "time_per_iteration": 2.7498619556427 + }, + { + "auxiliary_loss_clip": 0.01280759, + "auxiliary_loss_mlp": 0.0101085, + "balance_loss_clip": 1.11171186, + "balance_loss_mlp": 1.00360191, + "epoch": 0.03511092406661456, + "flos": 57242147725440.0, + "grad_norm": 0.9640892811499282, + "language_loss": 0.61102343, + "learning_rate": 3.9997324085664675e-06, + "loss": 0.6339395, + "num_input_tokens_seen": 6087470, + "step": 292, + "time_per_iteration": 3.2199602127075195 + }, + { + "auxiliary_loss_clip": 0.01311686, + "auxiliary_loss_mlp": 0.01055724, + "balance_loss_clip": 1.0844841, + "balance_loss_mlp": 1.03668618, + "epoch": 0.03523116695725365, + "flos": 22928065626240.0, + "grad_norm": 2.4161540106561716, + "language_loss": 0.92568666, + "learning_rate": 3.999719514722821e-06, + "loss": 0.94936061, + "num_input_tokens_seen": 6107600, + "step": 293, + "time_per_iteration": 2.7318496704101562 + }, + { + "auxiliary_loss_clip": 0.01314025, + "auxiliary_loss_mlp": 0.01056293, + "balance_loss_clip": 1.08810222, + "balance_loss_mlp": 1.03798223, + "epoch": 0.03535140984789274, + "flos": 36903241226880.0, + "grad_norm": 2.5901987417458985, + "language_loss": 0.74777448, + "learning_rate": 3.999706317523314e-06, + "loss": 0.7714777, + "num_input_tokens_seen": 6126160, + "step": 294, + "time_per_iteration": 2.800865411758423 + }, + { + "auxiliary_loss_clip": 0.01312159, + "auxiliary_loss_mlp": 0.01058614, + "balance_loss_clip": 1.0874877, + "balance_loss_mlp": 1.04062557, + "epoch": 0.03547165273853183, + "flos": 20449152316800.0, + "grad_norm": 2.2230020224941263, + "language_loss": 0.86440974, + "learning_rate": 3.999692816969948e-06, + "loss": 0.88811749, + "num_input_tokens_seen": 6145695, + "step": 295, + "time_per_iteration": 2.6914007663726807 + }, + { + "auxiliary_loss_clip": 0.01273317, + "auxiliary_loss_mlp": 0.01006527, + "balance_loss_clip": 1.10712838, + "balance_loss_mlp": 0.99956483, + "epoch": 0.03559189562917092, + "flos": 69850564871040.0, + "grad_norm": 0.9974928287700074, + "language_loss": 0.69315231, + "learning_rate": 3.999679013064772e-06, + "loss": 0.71595073, + "num_input_tokens_seen": 6212440, + "step": 296, + "time_per_iteration": 3.299705743789673 + }, + { + "auxiliary_loss_clip": 0.01310209, + "auxiliary_loss_mlp": 0.01061058, + "balance_loss_clip": 1.08549571, + "balance_loss_mlp": 1.042521, + "epoch": 0.03571213851981002, + "flos": 21651944163840.0, + "grad_norm": 5.251050857877003, + "language_loss": 0.85857868, + "learning_rate": 3.99966490580988e-06, + "loss": 0.88229132, + "num_input_tokens_seen": 6229800, + "step": 297, + "time_per_iteration": 2.6755144596099854 + }, + { + "auxiliary_loss_clip": 0.01311212, + "auxiliary_loss_mlp": 0.01065621, + "balance_loss_clip": 1.08523965, + "balance_loss_mlp": 1.04787064, + "epoch": 0.03583238141044911, + "flos": 43945610757120.0, + "grad_norm": 2.686390487644893, + "language_loss": 0.65705228, + "learning_rate": 3.999650495207411e-06, + "loss": 0.68082058, + "num_input_tokens_seen": 6255825, + "step": 298, + "time_per_iteration": 2.9005537033081055 + }, + { + "auxiliary_loss_clip": 0.01309862, + "auxiliary_loss_mlp": 0.01074269, + "balance_loss_clip": 1.08516288, + "balance_loss_mlp": 1.05548191, + "epoch": 0.0359526243010882, + "flos": 18910810592640.0, + "grad_norm": 3.441714111588123, + "language_loss": 0.90510696, + "learning_rate": 3.999635781259553e-06, + "loss": 0.92894822, + "num_input_tokens_seen": 6271090, + "step": 299, + "time_per_iteration": 2.599515914916992 + }, + { + "auxiliary_loss_clip": 0.0126324, + "auxiliary_loss_mlp": 0.01008483, + "balance_loss_clip": 1.10026276, + "balance_loss_mlp": 1.00118768, + "epoch": 0.03607286719172729, + "flos": 61668892782720.0, + "grad_norm": 0.9210320685029529, + "language_loss": 0.52269959, + "learning_rate": 3.999620763968535e-06, + "loss": 0.54541683, + "num_input_tokens_seen": 6329965, + "step": 300, + "time_per_iteration": 3.1225714683532715 + }, + { + "auxiliary_loss_clip": 0.01312513, + "auxiliary_loss_mlp": 0.0106618, + "balance_loss_clip": 1.08903956, + "balance_loss_mlp": 1.04752338, + "epoch": 0.03619311008236638, + "flos": 27819062991360.0, + "grad_norm": 1.631217051356969, + "language_loss": 0.86564791, + "learning_rate": 3.999605443336638e-06, + "loss": 0.88943487, + "num_input_tokens_seen": 6352095, + "step": 301, + "time_per_iteration": 2.6892127990722656 + }, + { + "auxiliary_loss_clip": 0.0131273, + "auxiliary_loss_mlp": 0.01063145, + "balance_loss_clip": 1.08676839, + "balance_loss_mlp": 1.04347515, + "epoch": 0.03631335297300547, + "flos": 13621133197440.0, + "grad_norm": 2.528008340256342, + "language_loss": 0.89689517, + "learning_rate": 3.999589819366185e-06, + "loss": 0.92065394, + "num_input_tokens_seen": 6365885, + "step": 302, + "time_per_iteration": 2.652567148208618 + }, + { + "auxiliary_loss_clip": 0.01309338, + "auxiliary_loss_mlp": 0.01075028, + "balance_loss_clip": 1.08559692, + "balance_loss_mlp": 1.05599022, + "epoch": 0.036433595863644565, + "flos": 27631788456960.0, + "grad_norm": 2.7281254903007066, + "language_loss": 0.84651351, + "learning_rate": 3.999573892059547e-06, + "loss": 0.87035716, + "num_input_tokens_seen": 6385015, + "step": 303, + "time_per_iteration": 2.7777817249298096 + }, + { + "auxiliary_loss_clip": 0.0131388, + "auxiliary_loss_mlp": 0.01066498, + "balance_loss_clip": 1.08757448, + "balance_loss_mlp": 1.04426575, + "epoch": 0.036553838754283655, + "flos": 24572020314240.0, + "grad_norm": 4.691866488515485, + "language_loss": 0.81208193, + "learning_rate": 3.999557661419138e-06, + "loss": 0.83588576, + "num_input_tokens_seen": 6405165, + "step": 304, + "time_per_iteration": 4.457868814468384 + }, + { + "auxiliary_loss_clip": 0.01311942, + "auxiliary_loss_mlp": 0.01072627, + "balance_loss_clip": 1.08910048, + "balance_loss_mlp": 1.05383945, + "epoch": 0.036674081644922744, + "flos": 23404313076480.0, + "grad_norm": 1.9124383411968353, + "language_loss": 0.81547761, + "learning_rate": 3.9995411274474225e-06, + "loss": 0.83932328, + "num_input_tokens_seen": 6424445, + "step": 305, + "time_per_iteration": 3.5503957271575928 + }, + { + "auxiliary_loss_clip": 0.01311175, + "auxiliary_loss_mlp": 0.01057857, + "balance_loss_clip": 1.08669984, + "balance_loss_mlp": 1.03806865, + "epoch": 0.036794324535561834, + "flos": 27489690253440.0, + "grad_norm": 1.9138483548310556, + "language_loss": 0.81735533, + "learning_rate": 3.999524290146908e-06, + "loss": 0.84104568, + "num_input_tokens_seen": 6444650, + "step": 306, + "time_per_iteration": 3.6221189498901367 + }, + { + "auxiliary_loss_clip": 0.01312243, + "auxiliary_loss_mlp": 0.01059596, + "balance_loss_clip": 1.0893451, + "balance_loss_mlp": 1.04053426, + "epoch": 0.036914567426200924, + "flos": 19463476227840.0, + "grad_norm": 2.4689666889775834, + "language_loss": 0.92404878, + "learning_rate": 3.9995071495201485e-06, + "loss": 0.94776714, + "num_input_tokens_seen": 6461755, + "step": 307, + "time_per_iteration": 2.6171133518218994 + }, + { + "auxiliary_loss_clip": 0.01310549, + "auxiliary_loss_mlp": 0.01061025, + "balance_loss_clip": 1.08777905, + "balance_loss_mlp": 1.04319167, + "epoch": 0.037034810316840014, + "flos": 22309324922880.0, + "grad_norm": 3.0168399595653934, + "language_loss": 0.97953033, + "learning_rate": 3.999489705569744e-06, + "loss": 1.00324607, + "num_input_tokens_seen": 6479455, + "step": 308, + "time_per_iteration": 2.6903789043426514 + }, + { + "auxiliary_loss_clip": 0.01309096, + "auxiliary_loss_mlp": 0.01050729, + "balance_loss_clip": 1.08484423, + "balance_loss_mlp": 1.03147697, + "epoch": 0.03715505320747911, + "flos": 18588333265920.0, + "grad_norm": 2.1908141917976796, + "language_loss": 0.8633033, + "learning_rate": 3.999471958298341e-06, + "loss": 0.88690162, + "num_input_tokens_seen": 6498365, + "step": 309, + "time_per_iteration": 2.638787269592285 + }, + { + "auxiliary_loss_clip": 0.01312204, + "auxiliary_loss_mlp": 0.0106483, + "balance_loss_clip": 1.08673429, + "balance_loss_mlp": 1.04350376, + "epoch": 0.0372752960981182, + "flos": 35955343267200.0, + "grad_norm": 1.963800068116651, + "language_loss": 0.76103604, + "learning_rate": 3.999453907708631e-06, + "loss": 0.78480637, + "num_input_tokens_seen": 6520770, + "step": 310, + "time_per_iteration": 2.800400495529175 + }, + { + "auxiliary_loss_clip": 0.0130688, + "auxiliary_loss_mlp": 0.01061857, + "balance_loss_clip": 1.08526289, + "balance_loss_mlp": 1.04401135, + "epoch": 0.03739553898875729, + "flos": 20814040627200.0, + "grad_norm": 2.2759080817276636, + "language_loss": 0.81504565, + "learning_rate": 3.999435553803353e-06, + "loss": 0.83873296, + "num_input_tokens_seen": 6540170, + "step": 311, + "time_per_iteration": 2.6443121433258057 + }, + { + "auxiliary_loss_clip": 0.01307682, + "auxiliary_loss_mlp": 0.01056277, + "balance_loss_clip": 1.08405471, + "balance_loss_mlp": 1.03805041, + "epoch": 0.03751578187939638, + "flos": 20264140339200.0, + "grad_norm": 3.1879417146529057, + "language_loss": 0.83361971, + "learning_rate": 3.999416896585292e-06, + "loss": 0.85725927, + "num_input_tokens_seen": 6557200, + "step": 312, + "time_per_iteration": 2.6550607681274414 + }, + { + "auxiliary_loss_clip": 0.01305286, + "auxiliary_loss_mlp": 0.01055305, + "balance_loss_clip": 1.08328557, + "balance_loss_mlp": 1.03767395, + "epoch": 0.03763602477003547, + "flos": 20668063754880.0, + "grad_norm": 2.9791069085915995, + "language_loss": 0.85875285, + "learning_rate": 3.9993979360572775e-06, + "loss": 0.88235879, + "num_input_tokens_seen": 6577340, + "step": 313, + "time_per_iteration": 2.708104372024536 + }, + { + "auxiliary_loss_clip": 0.01313278, + "auxiliary_loss_mlp": 0.01065805, + "balance_loss_clip": 1.08806562, + "balance_loss_mlp": 1.04753065, + "epoch": 0.03775626766067456, + "flos": 16691352197760.0, + "grad_norm": 4.62994817671807, + "language_loss": 0.82885134, + "learning_rate": 3.999378672222185e-06, + "loss": 0.85264212, + "num_input_tokens_seen": 6595125, + "step": 314, + "time_per_iteration": 2.6288952827453613 + }, + { + "auxiliary_loss_clip": 0.01310008, + "auxiliary_loss_mlp": 0.01052543, + "balance_loss_clip": 1.08594847, + "balance_loss_mlp": 1.03392208, + "epoch": 0.03787651055131366, + "flos": 21141797253120.0, + "grad_norm": 1.8907274625811366, + "language_loss": 0.82946378, + "learning_rate": 3.9993591050829385e-06, + "loss": 0.85308927, + "num_input_tokens_seen": 6612990, + "step": 315, + "time_per_iteration": 2.7481839656829834 + }, + { + "auxiliary_loss_clip": 0.01308528, + "auxiliary_loss_mlp": 0.01060598, + "balance_loss_clip": 1.0856837, + "balance_loss_mlp": 1.04162002, + "epoch": 0.037996753441952746, + "flos": 22018089450240.0, + "grad_norm": 1.9399290684122878, + "language_loss": 0.79563284, + "learning_rate": 3.999339234642506e-06, + "loss": 0.81932414, + "num_input_tokens_seen": 6632740, + "step": 316, + "time_per_iteration": 2.6846375465393066 + }, + { + "auxiliary_loss_clip": 0.01314519, + "auxiliary_loss_mlp": 0.01060904, + "balance_loss_clip": 1.09035742, + "balance_loss_mlp": 1.04090071, + "epoch": 0.038116996332591836, + "flos": 27709391790720.0, + "grad_norm": 3.770632929256164, + "language_loss": 0.83889383, + "learning_rate": 3.9993190609038994e-06, + "loss": 0.86264801, + "num_input_tokens_seen": 6651505, + "step": 317, + "time_per_iteration": 2.9041075706481934 + }, + { + "auxiliary_loss_clip": 0.01308581, + "auxiliary_loss_mlp": 0.01060729, + "balance_loss_clip": 1.08628953, + "balance_loss_mlp": 1.04207289, + "epoch": 0.038237239223230926, + "flos": 21178067011200.0, + "grad_norm": 1.9387497686249755, + "language_loss": 0.83215111, + "learning_rate": 3.999298583870182e-06, + "loss": 0.85584426, + "num_input_tokens_seen": 6671090, + "step": 318, + "time_per_iteration": 2.781205654144287 + }, + { + "auxiliary_loss_clip": 0.01310357, + "auxiliary_loss_mlp": 0.01065772, + "balance_loss_clip": 1.08727336, + "balance_loss_mlp": 1.04616177, + "epoch": 0.038357482113870016, + "flos": 25556618995200.0, + "grad_norm": 2.249238203406123, + "language_loss": 0.77474403, + "learning_rate": 3.999277803544458e-06, + "loss": 0.79850531, + "num_input_tokens_seen": 6691245, + "step": 319, + "time_per_iteration": 2.771157741546631 + }, + { + "auxiliary_loss_clip": 0.01250785, + "auxiliary_loss_mlp": 0.01021068, + "balance_loss_clip": 1.09020567, + "balance_loss_mlp": 1.01434469, + "epoch": 0.038477725004509106, + "flos": 59227578034560.0, + "grad_norm": 0.9578199881059267, + "language_loss": 0.62333548, + "learning_rate": 3.999256719929882e-06, + "loss": 0.64605403, + "num_input_tokens_seen": 6752520, + "step": 320, + "time_per_iteration": 3.262634038925171 + }, + { + "auxiliary_loss_clip": 0.01251256, + "auxiliary_loss_mlp": 0.0101534, + "balance_loss_clip": 1.08957171, + "balance_loss_mlp": 1.00852144, + "epoch": 0.0385979678951482, + "flos": 67317676398720.0, + "grad_norm": 1.2277723156698621, + "language_loss": 0.67142892, + "learning_rate": 3.999235333029651e-06, + "loss": 0.6940949, + "num_input_tokens_seen": 6806460, + "step": 321, + "time_per_iteration": 3.212071180343628 + }, + { + "auxiliary_loss_clip": 0.0131022, + "auxiliary_loss_mlp": 0.0106542, + "balance_loss_clip": 1.08926165, + "balance_loss_mlp": 1.04741931, + "epoch": 0.03871821078578729, + "flos": 22746752749440.0, + "grad_norm": 1.8693118905280577, + "language_loss": 0.82093418, + "learning_rate": 3.999213642847009e-06, + "loss": 0.84469062, + "num_input_tokens_seen": 6827045, + "step": 322, + "time_per_iteration": 2.742936849594116 + }, + { + "auxiliary_loss_clip": 0.01309805, + "auxiliary_loss_mlp": 0.01068349, + "balance_loss_clip": 1.08676589, + "balance_loss_mlp": 1.0480361, + "epoch": 0.03883845367642638, + "flos": 26280613526400.0, + "grad_norm": 1.7404276705173536, + "language_loss": 0.91155726, + "learning_rate": 3.999191649385247e-06, + "loss": 0.93533885, + "num_input_tokens_seen": 6848220, + "step": 323, + "time_per_iteration": 2.7841379642486572 + }, + { + "auxiliary_loss_clip": 0.01248981, + "auxiliary_loss_mlp": 0.01007949, + "balance_loss_clip": 1.08797812, + "balance_loss_mlp": 1.0007962, + "epoch": 0.03895869656706547, + "flos": 56962835568000.0, + "grad_norm": 0.9026697557896585, + "language_loss": 0.59781647, + "learning_rate": 3.999169352647702e-06, + "loss": 0.62038577, + "num_input_tokens_seen": 6909400, + "step": 324, + "time_per_iteration": 3.1913564205169678 + }, + { + "auxiliary_loss_clip": 0.01310781, + "auxiliary_loss_mlp": 0.01062479, + "balance_loss_clip": 1.08708763, + "balance_loss_mlp": 1.04238021, + "epoch": 0.03907893945770456, + "flos": 24863363527680.0, + "grad_norm": 1.8991009282882105, + "language_loss": 0.83041108, + "learning_rate": 3.999146752637755e-06, + "loss": 0.85414362, + "num_input_tokens_seen": 6930445, + "step": 325, + "time_per_iteration": 2.741779327392578 + }, + { + "auxiliary_loss_clip": 0.01307, + "auxiliary_loss_mlp": 0.01071232, + "balance_loss_clip": 1.0857929, + "balance_loss_mlp": 1.0500958, + "epoch": 0.03919918234834365, + "flos": 18368595815040.0, + "grad_norm": 2.502669447791281, + "language_loss": 0.89737463, + "learning_rate": 3.999123849358836e-06, + "loss": 0.921157, + "num_input_tokens_seen": 6948110, + "step": 326, + "time_per_iteration": 2.731886625289917 + }, + { + "auxiliary_loss_clip": 0.01304357, + "auxiliary_loss_mlp": 0.0106647, + "balance_loss_clip": 1.08233476, + "balance_loss_mlp": 1.04696751, + "epoch": 0.03931942523898275, + "flos": 25225414663680.0, + "grad_norm": 2.078040827725712, + "language_loss": 0.74878275, + "learning_rate": 3.999100642814418e-06, + "loss": 0.77249098, + "num_input_tokens_seen": 6968550, + "step": 327, + "time_per_iteration": 2.77333927154541 + }, + { + "auxiliary_loss_clip": 0.01310271, + "auxiliary_loss_mlp": 0.01066696, + "balance_loss_clip": 1.090873, + "balance_loss_mlp": 1.04755127, + "epoch": 0.03943966812962184, + "flos": 23257905240960.0, + "grad_norm": 2.3132994147358277, + "language_loss": 0.88430142, + "learning_rate": 3.999077133008022e-06, + "loss": 0.90807104, + "num_input_tokens_seen": 6987135, + "step": 328, + "time_per_iteration": 2.692312479019165 + }, + { + "auxiliary_loss_clip": 0.01309212, + "auxiliary_loss_mlp": 0.01067217, + "balance_loss_clip": 1.08736908, + "balance_loss_mlp": 1.04583097, + "epoch": 0.03955991102026093, + "flos": 29168837291520.0, + "grad_norm": 2.1032940211547637, + "language_loss": 0.90630025, + "learning_rate": 3.9990533199432145e-06, + "loss": 0.93006456, + "num_input_tokens_seen": 7008630, + "step": 329, + "time_per_iteration": 2.852144718170166 + }, + { + "auxiliary_loss_clip": 0.01311196, + "auxiliary_loss_mlp": 0.01059199, + "balance_loss_clip": 1.08858252, + "balance_loss_mlp": 1.04057825, + "epoch": 0.03968015391090002, + "flos": 17602441695360.0, + "grad_norm": 2.184630567283427, + "language_loss": 0.75585043, + "learning_rate": 3.999029203623608e-06, + "loss": 0.77955437, + "num_input_tokens_seen": 7026350, + "step": 330, + "time_per_iteration": 4.697614908218384 + }, + { + "auxiliary_loss_clip": 0.0130597, + "auxiliary_loss_mlp": 0.01061445, + "balance_loss_clip": 1.08637238, + "balance_loss_mlp": 1.04286051, + "epoch": 0.03980039680153911, + "flos": 21799285752960.0, + "grad_norm": 2.196421493314402, + "language_loss": 0.87135684, + "learning_rate": 3.99900478405286e-06, + "loss": 0.89503098, + "num_input_tokens_seen": 7045660, + "step": 331, + "time_per_iteration": 3.8041598796844482 + }, + { + "auxiliary_loss_clip": 0.01309462, + "auxiliary_loss_mlp": 0.01073797, + "balance_loss_clip": 1.08934021, + "balance_loss_mlp": 1.05522466, + "epoch": 0.0399206396921782, + "flos": 15195134148480.0, + "grad_norm": 3.41882829581653, + "language_loss": 0.82377237, + "learning_rate": 3.998980061234676e-06, + "loss": 0.84760493, + "num_input_tokens_seen": 7063575, + "step": 332, + "time_per_iteration": 2.649833917617798 + }, + { + "auxiliary_loss_clip": 0.01307491, + "auxiliary_loss_mlp": 0.01059581, + "balance_loss_clip": 1.08676195, + "balance_loss_mlp": 1.04026949, + "epoch": 0.040040882582817294, + "flos": 14422910630400.0, + "grad_norm": 2.768215121417092, + "language_loss": 0.76141548, + "learning_rate": 3.9989550351728055e-06, + "loss": 0.78508621, + "num_input_tokens_seen": 7080505, + "step": 333, + "time_per_iteration": 3.6096596717834473 + }, + { + "auxiliary_loss_clip": 0.0130327, + "auxiliary_loss_mlp": 0.01064969, + "balance_loss_clip": 1.08547282, + "balance_loss_mlp": 1.0474931, + "epoch": 0.040161125473456384, + "flos": 19280906375040.0, + "grad_norm": 3.346705753740481, + "language_loss": 0.84561354, + "learning_rate": 3.998929705871046e-06, + "loss": 0.8692959, + "num_input_tokens_seen": 7097860, + "step": 334, + "time_per_iteration": 2.6262149810791016 + }, + { + "auxiliary_loss_clip": 0.01308676, + "auxiliary_loss_mlp": 0.01047127, + "balance_loss_clip": 1.08794105, + "balance_loss_mlp": 1.02948451, + "epoch": 0.040281368364095474, + "flos": 17821101738240.0, + "grad_norm": 2.531292061670313, + "language_loss": 0.88957793, + "learning_rate": 3.99890407333324e-06, + "loss": 0.91313595, + "num_input_tokens_seen": 7116390, + "step": 335, + "time_per_iteration": 2.6366333961486816 + }, + { + "auxiliary_loss_clip": 0.01300539, + "auxiliary_loss_mlp": 0.01059671, + "balance_loss_clip": 1.08176029, + "balance_loss_mlp": 1.03984618, + "epoch": 0.040401611254734564, + "flos": 19573757959680.0, + "grad_norm": 1.6923333841474824, + "language_loss": 0.87050045, + "learning_rate": 3.998878137563275e-06, + "loss": 0.89410257, + "num_input_tokens_seen": 7135940, + "step": 336, + "time_per_iteration": 2.6309449672698975 + }, + { + "auxiliary_loss_clip": 0.01300574, + "auxiliary_loss_mlp": 0.01064537, + "balance_loss_clip": 1.08268571, + "balance_loss_mlp": 1.0466435, + "epoch": 0.040521854145373654, + "flos": 22054466949120.0, + "grad_norm": 2.237980166949956, + "language_loss": 0.85269392, + "learning_rate": 3.998851898565085e-06, + "loss": 0.87634504, + "num_input_tokens_seen": 7155745, + "step": 337, + "time_per_iteration": 2.777923345565796 + }, + { + "auxiliary_loss_clip": 0.01299927, + "auxiliary_loss_mlp": 0.01051493, + "balance_loss_clip": 1.08469248, + "balance_loss_mlp": 1.03424346, + "epoch": 0.04064209703601274, + "flos": 22674644196480.0, + "grad_norm": 1.8901208953963478, + "language_loss": 0.82992899, + "learning_rate": 3.998825356342653e-06, + "loss": 0.85344326, + "num_input_tokens_seen": 7175920, + "step": 338, + "time_per_iteration": 2.763899564743042 + }, + { + "auxiliary_loss_clip": 0.01303293, + "auxiliary_loss_mlp": 0.01057037, + "balance_loss_clip": 1.08555138, + "balance_loss_mlp": 1.03970361, + "epoch": 0.04076233992665183, + "flos": 38582172783360.0, + "grad_norm": 2.9641121505023946, + "language_loss": 0.72965097, + "learning_rate": 3.998798510900003e-06, + "loss": 0.75325429, + "num_input_tokens_seen": 7198720, + "step": 339, + "time_per_iteration": 2.9327797889709473 + }, + { + "auxiliary_loss_clip": 0.01304655, + "auxiliary_loss_mlp": 0.01070214, + "balance_loss_clip": 1.08436728, + "balance_loss_mlp": 1.05183172, + "epoch": 0.04088258281729093, + "flos": 25885309374720.0, + "grad_norm": 2.1307115802824885, + "language_loss": 0.83865154, + "learning_rate": 3.998771362241207e-06, + "loss": 0.86240017, + "num_input_tokens_seen": 7219125, + "step": 340, + "time_per_iteration": 2.640860080718994 + }, + { + "auxiliary_loss_clip": 0.01301856, + "auxiliary_loss_mlp": 0.01059774, + "balance_loss_clip": 1.08472157, + "balance_loss_mlp": 1.04247642, + "epoch": 0.04100282570793002, + "flos": 19789832223360.0, + "grad_norm": 3.3581494056337955, + "language_loss": 0.87877917, + "learning_rate": 3.998743910370385e-06, + "loss": 0.90239549, + "num_input_tokens_seen": 7237985, + "step": 341, + "time_per_iteration": 2.6399502754211426 + }, + { + "auxiliary_loss_clip": 0.0130749, + "auxiliary_loss_mlp": 0.01059827, + "balance_loss_clip": 1.09093904, + "balance_loss_mlp": 1.04311407, + "epoch": 0.04112306859856911, + "flos": 22565152563840.0, + "grad_norm": 1.8894306131017078, + "language_loss": 0.73361856, + "learning_rate": 3.998716155291702e-06, + "loss": 0.75729179, + "num_input_tokens_seen": 7255825, + "step": 342, + "time_per_iteration": 2.736938238143921 + }, + { + "auxiliary_loss_clip": 0.01304095, + "auxiliary_loss_mlp": 0.0106673, + "balance_loss_clip": 1.0877223, + "balance_loss_mlp": 1.04906321, + "epoch": 0.0412433114892082, + "flos": 25040654081280.0, + "grad_norm": 1.797522723232085, + "language_loss": 0.9056462, + "learning_rate": 3.998688097009366e-06, + "loss": 0.92935443, + "num_input_tokens_seen": 7276590, + "step": 343, + "time_per_iteration": 2.702192783355713 + }, + { + "auxiliary_loss_clip": 0.01303243, + "auxiliary_loss_mlp": 0.01057368, + "balance_loss_clip": 1.08619642, + "balance_loss_mlp": 1.04141808, + "epoch": 0.04136355437984729, + "flos": 25191371548800.0, + "grad_norm": 2.735366860140538, + "language_loss": 0.80442101, + "learning_rate": 3.998659735527636e-06, + "loss": 0.82802713, + "num_input_tokens_seen": 7295680, + "step": 344, + "time_per_iteration": 2.7618319988250732 + }, + { + "auxiliary_loss_clip": 0.01304646, + "auxiliary_loss_mlp": 0.01055923, + "balance_loss_clip": 1.08792853, + "balance_loss_mlp": 1.03854263, + "epoch": 0.04148379727048638, + "flos": 22966777509120.0, + "grad_norm": 2.1848385474228356, + "language_loss": 0.77902269, + "learning_rate": 3.998631070850813e-06, + "loss": 0.8026284, + "num_input_tokens_seen": 7316300, + "step": 345, + "time_per_iteration": 2.7131597995758057 + }, + { + "auxiliary_loss_clip": 0.01302268, + "auxiliary_loss_mlp": 0.01070247, + "balance_loss_clip": 1.08587408, + "balance_loss_mlp": 1.05294955, + "epoch": 0.041604040161125476, + "flos": 14063481187200.0, + "grad_norm": 2.363711052740664, + "language_loss": 0.83491921, + "learning_rate": 3.9986021029832455e-06, + "loss": 0.85864437, + "num_input_tokens_seen": 7333615, + "step": 346, + "time_per_iteration": 2.617974281311035 + }, + { + "auxiliary_loss_clip": 0.01304119, + "auxiliary_loss_mlp": 0.01053729, + "balance_loss_clip": 1.08544004, + "balance_loss_mlp": 1.03547788, + "epoch": 0.041724283051764566, + "flos": 12091877614080.0, + "grad_norm": 2.7288894756297797, + "language_loss": 0.91364717, + "learning_rate": 3.9985728319293285e-06, + "loss": 0.9372257, + "num_input_tokens_seen": 7347590, + "step": 347, + "time_per_iteration": 2.635206699371338 + }, + { + "auxiliary_loss_clip": 0.0130384, + "auxiliary_loss_mlp": 0.01066345, + "balance_loss_clip": 1.08271658, + "balance_loss_mlp": 1.0459125, + "epoch": 0.041844525942403656, + "flos": 12385303816320.0, + "grad_norm": 2.451707286411748, + "language_loss": 0.8538276, + "learning_rate": 3.998543257693501e-06, + "loss": 0.87752944, + "num_input_tokens_seen": 7364345, + "step": 348, + "time_per_iteration": 2.7004234790802 + }, + { + "auxiliary_loss_clip": 0.01304041, + "auxiliary_loss_mlp": 0.01063576, + "balance_loss_clip": 1.08703971, + "balance_loss_mlp": 1.04682755, + "epoch": 0.041964768833042745, + "flos": 23769345041280.0, + "grad_norm": 1.8199392264352041, + "language_loss": 0.87988174, + "learning_rate": 3.998513380280251e-06, + "loss": 0.9035579, + "num_input_tokens_seen": 7384625, + "step": 349, + "time_per_iteration": 2.6273834705352783 + }, + { + "auxiliary_loss_clip": 0.01307379, + "auxiliary_loss_mlp": 0.0106142, + "balance_loss_clip": 1.08781004, + "balance_loss_mlp": 1.04132152, + "epoch": 0.042085011723681835, + "flos": 11875336473600.0, + "grad_norm": 2.3765476529453315, + "language_loss": 0.95185882, + "learning_rate": 3.99848319969411e-06, + "loss": 0.97554684, + "num_input_tokens_seen": 7402225, + "step": 350, + "time_per_iteration": 2.678209066390991 + }, + { + "auxiliary_loss_clip": 0.01309193, + "auxiliary_loss_mlp": 0.01058734, + "balance_loss_clip": 1.09136868, + "balance_loss_mlp": 1.04143655, + "epoch": 0.042205254614320925, + "flos": 16873957964160.0, + "grad_norm": 1.9320858956590479, + "language_loss": 0.79324883, + "learning_rate": 3.9984527159396564e-06, + "loss": 0.81692815, + "num_input_tokens_seen": 7420865, + "step": 351, + "time_per_iteration": 2.628967046737671 + }, + { + "auxiliary_loss_clip": 0.0129731, + "auxiliary_loss_mlp": 0.01044377, + "balance_loss_clip": 1.08134007, + "balance_loss_mlp": 1.02624512, + "epoch": 0.04232549750496002, + "flos": 25118508810240.0, + "grad_norm": 2.086206541559121, + "language_loss": 0.8447597, + "learning_rate": 3.9984219290215154e-06, + "loss": 0.86817652, + "num_input_tokens_seen": 7441040, + "step": 352, + "time_per_iteration": 2.6945953369140625 + }, + { + "auxiliary_loss_clip": 0.0130226, + "auxiliary_loss_mlp": 0.01056421, + "balance_loss_clip": 1.08790803, + "balance_loss_mlp": 1.03980291, + "epoch": 0.04244574039559911, + "flos": 26724541714560.0, + "grad_norm": 1.6677832655324478, + "language_loss": 0.89162099, + "learning_rate": 3.998390838944356e-06, + "loss": 0.9152078, + "num_input_tokens_seen": 7462545, + "step": 353, + "time_per_iteration": 2.711015224456787 + }, + { + "auxiliary_loss_clip": 0.01304658, + "auxiliary_loss_mlp": 0.0105908, + "balance_loss_clip": 1.08717513, + "balance_loss_mlp": 1.04067457, + "epoch": 0.0425659832862382, + "flos": 20923244951040.0, + "grad_norm": 2.658823982527189, + "language_loss": 0.90407044, + "learning_rate": 3.998359445712895e-06, + "loss": 0.92770779, + "num_input_tokens_seen": 7481650, + "step": 354, + "time_per_iteration": 2.715985059738159 + }, + { + "auxiliary_loss_clip": 0.0130268, + "auxiliary_loss_mlp": 0.01063915, + "balance_loss_clip": 1.08288181, + "balance_loss_mlp": 1.04543793, + "epoch": 0.04268622617687729, + "flos": 23331127115520.0, + "grad_norm": 3.0365884393366107, + "language_loss": 0.81602764, + "learning_rate": 3.9983277493318955e-06, + "loss": 0.83969361, + "num_input_tokens_seen": 7500945, + "step": 355, + "time_per_iteration": 2.6353185176849365 + }, + { + "auxiliary_loss_clip": 0.01302898, + "auxiliary_loss_mlp": 0.01054845, + "balance_loss_clip": 1.08613539, + "balance_loss_mlp": 1.03775012, + "epoch": 0.04280646906751638, + "flos": 25994010908160.0, + "grad_norm": 3.333544601240233, + "language_loss": 0.8112613, + "learning_rate": 3.998295749806165e-06, + "loss": 0.83483875, + "num_input_tokens_seen": 7522170, + "step": 356, + "time_per_iteration": 2.732208728790283 + }, + { + "auxiliary_loss_clip": 0.01304357, + "auxiliary_loss_mlp": 0.01059841, + "balance_loss_clip": 1.09020615, + "balance_loss_mlp": 1.04234076, + "epoch": 0.04292671195815547, + "flos": 26906824258560.0, + "grad_norm": 1.9405859346698124, + "language_loss": 0.83137798, + "learning_rate": 3.998263447140558e-06, + "loss": 0.85501999, + "num_input_tokens_seen": 7542370, + "step": 357, + "time_per_iteration": 3.6667659282684326 + }, + { + "auxiliary_loss_clip": 0.01304233, + "auxiliary_loss_mlp": 0.0105439, + "balance_loss_clip": 1.08613634, + "balance_loss_mlp": 1.03631783, + "epoch": 0.04304695484879457, + "flos": 39457315745280.0, + "grad_norm": 2.6294518765752732, + "language_loss": 0.8164624, + "learning_rate": 3.998230841339976e-06, + "loss": 0.84004867, + "num_input_tokens_seen": 7564380, + "step": 358, + "time_per_iteration": 3.757214307785034 + }, + { + "auxiliary_loss_clip": 0.01301163, + "auxiliary_loss_mlp": 0.01057738, + "balance_loss_clip": 1.08601761, + "balance_loss_mlp": 1.04054844, + "epoch": 0.04316719773943366, + "flos": 19646297475840.0, + "grad_norm": 2.4668187139419104, + "language_loss": 0.85094875, + "learning_rate": 3.998197932409363e-06, + "loss": 0.87453771, + "num_input_tokens_seen": 7582390, + "step": 359, + "time_per_iteration": 2.602274179458618 + }, + { + "auxiliary_loss_clip": 0.0129758, + "auxiliary_loss_mlp": 0.01049091, + "balance_loss_clip": 1.08358955, + "balance_loss_mlp": 1.03271174, + "epoch": 0.04328744063007275, + "flos": 22452320966400.0, + "grad_norm": 2.333282202542991, + "language_loss": 0.8613742, + "learning_rate": 3.9981647203537125e-06, + "loss": 0.88484097, + "num_input_tokens_seen": 7599890, + "step": 360, + "time_per_iteration": 3.6478271484375 + }, + { + "auxiliary_loss_clip": 0.01304655, + "auxiliary_loss_mlp": 0.01062256, + "balance_loss_clip": 1.08757436, + "balance_loss_mlp": 1.04493451, + "epoch": 0.04340768352071184, + "flos": 21283033530240.0, + "grad_norm": 1.954509137440414, + "language_loss": 0.96028066, + "learning_rate": 3.998131205178063e-06, + "loss": 0.98394978, + "num_input_tokens_seen": 7618360, + "step": 361, + "time_per_iteration": 2.6298768520355225 + }, + { + "auxiliary_loss_clip": 0.01301045, + "auxiliary_loss_mlp": 0.01069018, + "balance_loss_clip": 1.0841831, + "balance_loss_mlp": 1.05145848, + "epoch": 0.04352792641135093, + "flos": 11583705951360.0, + "grad_norm": 3.2733583126171486, + "language_loss": 0.76599813, + "learning_rate": 3.998097386887498e-06, + "loss": 0.78969878, + "num_input_tokens_seen": 7635435, + "step": 362, + "time_per_iteration": 2.652233123779297 + }, + { + "auxiliary_loss_clip": 0.01297204, + "auxiliary_loss_mlp": 0.0105877, + "balance_loss_clip": 1.08467054, + "balance_loss_mlp": 1.0425874, + "epoch": 0.04364816930199002, + "flos": 23623547736960.0, + "grad_norm": 2.09351881633569, + "language_loss": 0.85107529, + "learning_rate": 3.998063265487148e-06, + "loss": 0.8746351, + "num_input_tokens_seen": 7656485, + "step": 363, + "time_per_iteration": 2.662837266921997 + }, + { + "auxiliary_loss_clip": 0.01300076, + "auxiliary_loss_mlp": 0.01055472, + "balance_loss_clip": 1.08727837, + "balance_loss_mlp": 1.03868759, + "epoch": 0.043768412192629114, + "flos": 14429734214400.0, + "grad_norm": 2.07503713703765, + "language_loss": 0.8109926, + "learning_rate": 3.99802884098219e-06, + "loss": 0.834548, + "num_input_tokens_seen": 7674595, + "step": 364, + "time_per_iteration": 2.722846031188965 + }, + { + "auxiliary_loss_clip": 0.01302402, + "auxiliary_loss_mlp": 0.01068982, + "balance_loss_clip": 1.08486605, + "balance_loss_mlp": 1.05168426, + "epoch": 0.043888655083268203, + "flos": 26468893641600.0, + "grad_norm": 2.122212853807967, + "language_loss": 0.82375866, + "learning_rate": 3.997994113377845e-06, + "loss": 0.84747249, + "num_input_tokens_seen": 7693495, + "step": 365, + "time_per_iteration": 2.676603317260742 + }, + { + "auxiliary_loss_clip": 0.01300226, + "auxiliary_loss_mlp": 0.01058641, + "balance_loss_clip": 1.08538532, + "balance_loss_mlp": 1.04220223, + "epoch": 0.04400889797390729, + "flos": 27235263242880.0, + "grad_norm": 1.9179267618714768, + "language_loss": 0.83364636, + "learning_rate": 3.9979590826793815e-06, + "loss": 0.85723507, + "num_input_tokens_seen": 7714685, + "step": 366, + "time_per_iteration": 2.86320161819458 + }, + { + "auxiliary_loss_clip": 0.01302372, + "auxiliary_loss_mlp": 0.01068193, + "balance_loss_clip": 1.08659708, + "balance_loss_mlp": 1.04990578, + "epoch": 0.04412914086454638, + "flos": 20119528183680.0, + "grad_norm": 2.150210639267906, + "language_loss": 0.80973923, + "learning_rate": 3.997923748892113e-06, + "loss": 0.83344489, + "num_input_tokens_seen": 7734005, + "step": 367, + "time_per_iteration": 2.7747983932495117 + }, + { + "auxiliary_loss_clip": 0.01299112, + "auxiliary_loss_mlp": 0.01051597, + "balance_loss_clip": 1.08425474, + "balance_loss_mlp": 1.03471744, + "epoch": 0.04424938375518547, + "flos": 22604618632320.0, + "grad_norm": 1.7263760626282272, + "language_loss": 0.88743734, + "learning_rate": 3.9978881120214015e-06, + "loss": 0.91094446, + "num_input_tokens_seen": 7755525, + "step": 368, + "time_per_iteration": 2.7346785068511963 + }, + { + "auxiliary_loss_clip": 0.01300921, + "auxiliary_loss_mlp": 0.0105553, + "balance_loss_clip": 1.08411789, + "balance_loss_mlp": 1.03558671, + "epoch": 0.04436962664582456, + "flos": 24132365844480.0, + "grad_norm": 2.7369735311886685, + "language_loss": 0.79304326, + "learning_rate": 3.997852172072652e-06, + "loss": 0.81660783, + "num_input_tokens_seen": 7776740, + "step": 369, + "time_per_iteration": 2.707648515701294 + }, + { + "auxiliary_loss_clip": 0.01301717, + "auxiliary_loss_mlp": 0.01059963, + "balance_loss_clip": 1.08508146, + "balance_loss_mlp": 1.04168844, + "epoch": 0.04448986953646366, + "flos": 18222906251520.0, + "grad_norm": 2.384412344213197, + "language_loss": 0.89485866, + "learning_rate": 3.9978159290513155e-06, + "loss": 0.91847551, + "num_input_tokens_seen": 7794820, + "step": 370, + "time_per_iteration": 2.5944697856903076 + }, + { + "auxiliary_loss_clip": 0.0130544, + "auxiliary_loss_mlp": 0.01063154, + "balance_loss_clip": 1.08622813, + "balance_loss_mlp": 1.04471219, + "epoch": 0.04461011242710275, + "flos": 30117920400000.0, + "grad_norm": 2.3435469941182303, + "language_loss": 0.80067098, + "learning_rate": 3.997779382962892e-06, + "loss": 0.82435691, + "num_input_tokens_seen": 7817705, + "step": 371, + "time_per_iteration": 2.763150691986084 + }, + { + "auxiliary_loss_clip": 0.01298882, + "auxiliary_loss_mlp": 0.01060578, + "balance_loss_clip": 1.08457696, + "balance_loss_mlp": 1.0427916, + "epoch": 0.04473035531774184, + "flos": 29752529299200.0, + "grad_norm": 1.7869607413181539, + "language_loss": 0.73561138, + "learning_rate": 3.997742533812924e-06, + "loss": 0.75920594, + "num_input_tokens_seen": 7840970, + "step": 372, + "time_per_iteration": 2.738323926925659 + }, + { + "auxiliary_loss_clip": 0.01302962, + "auxiliary_loss_mlp": 0.0107711, + "balance_loss_clip": 1.08803165, + "balance_loss_mlp": 1.05782151, + "epoch": 0.04485059820838093, + "flos": 13151565676800.0, + "grad_norm": 2.4484866609214646, + "language_loss": 0.9268626, + "learning_rate": 3.997705381607001e-06, + "loss": 0.95066333, + "num_input_tokens_seen": 7857785, + "step": 373, + "time_per_iteration": 2.7856104373931885 + }, + { + "auxiliary_loss_clip": 0.01246637, + "auxiliary_loss_mlp": 0.01024127, + "balance_loss_clip": 1.08962893, + "balance_loss_mlp": 1.01721251, + "epoch": 0.04497084109902002, + "flos": 68094209548800.0, + "grad_norm": 0.9652279889143864, + "language_loss": 0.60235959, + "learning_rate": 3.997667926350761e-06, + "loss": 0.62506717, + "num_input_tokens_seen": 7916115, + "step": 374, + "time_per_iteration": 3.156999111175537 + }, + { + "auxiliary_loss_clip": 0.01246413, + "auxiliary_loss_mlp": 0.01019916, + "balance_loss_clip": 1.09013414, + "balance_loss_mlp": 1.01300192, + "epoch": 0.04509108398965911, + "flos": 64342263346560.0, + "grad_norm": 0.8972069951575466, + "language_loss": 0.57755405, + "learning_rate": 3.997630168049886e-06, + "loss": 0.60021734, + "num_input_tokens_seen": 7974480, + "step": 375, + "time_per_iteration": 3.2336859703063965 + }, + { + "auxiliary_loss_clip": 0.0130473, + "auxiliary_loss_mlp": 0.01057201, + "balance_loss_clip": 1.08831644, + "balance_loss_mlp": 1.03822303, + "epoch": 0.045211326880298205, + "flos": 22271115830400.0, + "grad_norm": 2.0046040582993654, + "language_loss": 0.77609384, + "learning_rate": 3.997592106710101e-06, + "loss": 0.79971313, + "num_input_tokens_seen": 7993940, + "step": 376, + "time_per_iteration": 2.840975284576416 + }, + { + "auxiliary_loss_clip": 0.01295759, + "auxiliary_loss_mlp": 0.01051557, + "balance_loss_clip": 1.08272266, + "balance_loss_mlp": 1.03468847, + "epoch": 0.045331569770937295, + "flos": 32159441796480.0, + "grad_norm": 2.5454633787289147, + "language_loss": 0.65894043, + "learning_rate": 3.997553742337182e-06, + "loss": 0.68241358, + "num_input_tokens_seen": 8013365, + "step": 377, + "time_per_iteration": 2.794199228286743 + }, + { + "auxiliary_loss_clip": 0.0130127, + "auxiliary_loss_mlp": 0.01062592, + "balance_loss_clip": 1.08716381, + "balance_loss_mlp": 1.04448414, + "epoch": 0.045451812661576385, + "flos": 22163455791360.0, + "grad_norm": 1.7162929102516964, + "language_loss": 0.91541553, + "learning_rate": 3.997515074936949e-06, + "loss": 0.93905407, + "num_input_tokens_seen": 8034240, + "step": 378, + "time_per_iteration": 2.7712244987487793 + }, + { + "auxiliary_loss_clip": 0.01302176, + "auxiliary_loss_mlp": 0.01069683, + "balance_loss_clip": 1.08764935, + "balance_loss_mlp": 1.05021596, + "epoch": 0.045572055552215475, + "flos": 16581968305920.0, + "grad_norm": 2.4358622743185796, + "language_loss": 0.87152553, + "learning_rate": 3.997476104515268e-06, + "loss": 0.89524412, + "num_input_tokens_seen": 8052430, + "step": 379, + "time_per_iteration": 2.6609015464782715 + }, + { + "auxiliary_loss_clip": 0.01298573, + "auxiliary_loss_mlp": 0.01058329, + "balance_loss_clip": 1.08736563, + "balance_loss_mlp": 1.04137754, + "epoch": 0.045692298442854565, + "flos": 17603375448960.0, + "grad_norm": 1.8836738930054755, + "language_loss": 0.77591306, + "learning_rate": 3.9974368310780485e-06, + "loss": 0.79948205, + "num_input_tokens_seen": 8069605, + "step": 380, + "time_per_iteration": 2.6831490993499756 + }, + { + "auxiliary_loss_clip": 0.0130636, + "auxiliary_loss_mlp": 0.01063373, + "balance_loss_clip": 1.0879879, + "balance_loss_mlp": 1.04158211, + "epoch": 0.045812541333493655, + "flos": 26761098781440.0, + "grad_norm": 2.329868587810593, + "language_loss": 0.74842703, + "learning_rate": 3.997397254631251e-06, + "loss": 0.77212441, + "num_input_tokens_seen": 8090225, + "step": 381, + "time_per_iteration": 2.6914665699005127 + }, + { + "auxiliary_loss_clip": 0.01225887, + "auxiliary_loss_mlp": 0.01036004, + "balance_loss_clip": 1.07467175, + "balance_loss_mlp": 1.02913749, + "epoch": 0.04593278422413275, + "flos": 60250349894400.0, + "grad_norm": 0.8275555940855851, + "language_loss": 0.60080028, + "learning_rate": 3.997357375180878e-06, + "loss": 0.62341923, + "num_input_tokens_seen": 8154505, + "step": 382, + "time_per_iteration": 3.326723575592041 + }, + { + "auxiliary_loss_clip": 0.01299494, + "auxiliary_loss_mlp": 0.01063195, + "balance_loss_clip": 1.0864501, + "balance_loss_mlp": 1.04552829, + "epoch": 0.04605302711477184, + "flos": 21799249839360.0, + "grad_norm": 1.9934162149043493, + "language_loss": 0.75597775, + "learning_rate": 3.997317192732979e-06, + "loss": 0.77960467, + "num_input_tokens_seen": 8173285, + "step": 383, + "time_per_iteration": 2.7412519454956055 + }, + { + "auxiliary_loss_clip": 0.01301418, + "auxiliary_loss_mlp": 0.01056637, + "balance_loss_clip": 1.08634102, + "balance_loss_mlp": 1.03933358, + "epoch": 0.04617327000541093, + "flos": 19459705299840.0, + "grad_norm": 2.3300512704865977, + "language_loss": 0.82569182, + "learning_rate": 3.99727670729365e-06, + "loss": 0.84927237, + "num_input_tokens_seen": 8191845, + "step": 384, + "time_per_iteration": 5.526530981063843 + }, + { + "auxiliary_loss_clip": 0.01298442, + "auxiliary_loss_mlp": 0.0106523, + "balance_loss_clip": 1.08725858, + "balance_loss_mlp": 1.04823112, + "epoch": 0.04629351289605002, + "flos": 25411468135680.0, + "grad_norm": 2.497349836533345, + "language_loss": 0.78169024, + "learning_rate": 3.997235918869033e-06, + "loss": 0.80532694, + "num_input_tokens_seen": 8212880, + "step": 385, + "time_per_iteration": 2.775137186050415 + }, + { + "auxiliary_loss_clip": 0.01301196, + "auxiliary_loss_mlp": 0.01058969, + "balance_loss_clip": 1.08688092, + "balance_loss_mlp": 1.04183841, + "epoch": 0.04641375578668911, + "flos": 20558284813440.0, + "grad_norm": 1.8567365853234077, + "language_loss": 0.82800508, + "learning_rate": 3.997194827465315e-06, + "loss": 0.85160679, + "num_input_tokens_seen": 8231475, + "step": 386, + "time_per_iteration": 3.697254180908203 + }, + { + "auxiliary_loss_clip": 0.0130184, + "auxiliary_loss_mlp": 0.01060056, + "balance_loss_clip": 1.08579016, + "balance_loss_mlp": 1.04237747, + "epoch": 0.0465339986773282, + "flos": 13188661447680.0, + "grad_norm": 2.5887011665464748, + "language_loss": 0.91875792, + "learning_rate": 3.997153433088728e-06, + "loss": 0.94237697, + "num_input_tokens_seen": 8248600, + "step": 387, + "time_per_iteration": 2.743476390838623 + }, + { + "auxiliary_loss_clip": 0.01304097, + "auxiliary_loss_mlp": 0.01060889, + "balance_loss_clip": 1.08830523, + "balance_loss_mlp": 1.04342484, + "epoch": 0.0466542415679673, + "flos": 25556547168000.0, + "grad_norm": 1.7662210394906215, + "language_loss": 0.81343561, + "learning_rate": 3.997111735745554e-06, + "loss": 0.83708549, + "num_input_tokens_seen": 8271570, + "step": 388, + "time_per_iteration": 2.7210116386413574 + }, + { + "auxiliary_loss_clip": 0.01299166, + "auxiliary_loss_mlp": 0.01053262, + "balance_loss_clip": 1.08729064, + "balance_loss_mlp": 1.03567863, + "epoch": 0.04677448445860639, + "flos": 22236749493120.0, + "grad_norm": 1.8574008018604051, + "language_loss": 0.82720667, + "learning_rate": 3.997069735442118e-06, + "loss": 0.85073096, + "num_input_tokens_seen": 8291265, + "step": 389, + "time_per_iteration": 2.6964118480682373 + }, + { + "auxiliary_loss_clip": 0.01297578, + "auxiliary_loss_mlp": 0.01060541, + "balance_loss_clip": 1.08483815, + "balance_loss_mlp": 1.04385185, + "epoch": 0.04689472734924548, + "flos": 28147825198080.0, + "grad_norm": 1.5838851978086042, + "language_loss": 0.80423683, + "learning_rate": 3.997027432184792e-06, + "loss": 0.82781804, + "num_input_tokens_seen": 8315925, + "step": 390, + "time_per_iteration": 2.790759563446045 + }, + { + "auxiliary_loss_clip": 0.01297234, + "auxiliary_loss_mlp": 0.0106151, + "balance_loss_clip": 1.0846895, + "balance_loss_mlp": 1.04439211, + "epoch": 0.04701497023988457, + "flos": 23148952312320.0, + "grad_norm": 1.8267016208872877, + "language_loss": 0.89536202, + "learning_rate": 3.99698482597999e-06, + "loss": 0.91894948, + "num_input_tokens_seen": 8333605, + "step": 391, + "time_per_iteration": 2.741189479827881 + }, + { + "auxiliary_loss_clip": 0.01208038, + "auxiliary_loss_mlp": 0.01010523, + "balance_loss_clip": 1.0605979, + "balance_loss_mlp": 1.00379968, + "epoch": 0.04713521313052366, + "flos": 64827668764800.0, + "grad_norm": 0.8665410221420657, + "language_loss": 0.63861519, + "learning_rate": 3.99694191683418e-06, + "loss": 0.66080081, + "num_input_tokens_seen": 8394405, + "step": 392, + "time_per_iteration": 3.262887477874756 + }, + { + "auxiliary_loss_clip": 0.01299612, + "auxiliary_loss_mlp": 0.01065419, + "balance_loss_clip": 1.08790934, + "balance_loss_mlp": 1.04799056, + "epoch": 0.047255456021162746, + "flos": 18771585477120.0, + "grad_norm": 2.004545562592604, + "language_loss": 0.81625342, + "learning_rate": 3.996898704753867e-06, + "loss": 0.83990383, + "num_input_tokens_seen": 8412355, + "step": 393, + "time_per_iteration": 2.6839187145233154 + }, + { + "auxiliary_loss_clip": 0.01300617, + "auxiliary_loss_mlp": 0.01063913, + "balance_loss_clip": 1.08488536, + "balance_loss_mlp": 1.04544735, + "epoch": 0.04737569891180184, + "flos": 22053820504320.0, + "grad_norm": 2.03595554743196, + "language_loss": 0.87817633, + "learning_rate": 3.996855189745609e-06, + "loss": 0.90182161, + "num_input_tokens_seen": 8431620, + "step": 394, + "time_per_iteration": 2.6717026233673096 + }, + { + "auxiliary_loss_clip": 0.01294279, + "auxiliary_loss_mlp": 0.01052138, + "balance_loss_clip": 1.08209133, + "balance_loss_mlp": 1.0355916, + "epoch": 0.04749594180244093, + "flos": 29057370410880.0, + "grad_norm": 1.8910046734477384, + "language_loss": 0.92673659, + "learning_rate": 3.996811371816007e-06, + "loss": 0.95020074, + "num_input_tokens_seen": 8454045, + "step": 395, + "time_per_iteration": 2.7716076374053955 + }, + { + "auxiliary_loss_clip": 0.01297144, + "auxiliary_loss_mlp": 0.01064651, + "balance_loss_clip": 1.08569217, + "balance_loss_mlp": 1.04750848, + "epoch": 0.04761618469308002, + "flos": 35112268172160.0, + "grad_norm": 1.873319330612119, + "language_loss": 0.77994633, + "learning_rate": 3.996767250971707e-06, + "loss": 0.80356425, + "num_input_tokens_seen": 8476785, + "step": 396, + "time_per_iteration": 2.847656011581421 + }, + { + "auxiliary_loss_clip": 0.01300094, + "auxiliary_loss_mlp": 0.01052964, + "balance_loss_clip": 1.08678186, + "balance_loss_mlp": 1.03588128, + "epoch": 0.04773642758371911, + "flos": 25630702796160.0, + "grad_norm": 1.8920816026479423, + "language_loss": 0.87171018, + "learning_rate": 3.996722827219403e-06, + "loss": 0.89524072, + "num_input_tokens_seen": 8498400, + "step": 397, + "time_per_iteration": 2.7165911197662354 + }, + { + "auxiliary_loss_clip": 0.01297626, + "auxiliary_loss_mlp": 0.01065045, + "balance_loss_clip": 1.08650625, + "balance_loss_mlp": 1.04717541, + "epoch": 0.0478566704743582, + "flos": 20631506688000.0, + "grad_norm": 2.320986008943373, + "language_loss": 0.82864261, + "learning_rate": 3.996678100565833e-06, + "loss": 0.85226935, + "num_input_tokens_seen": 8517455, + "step": 398, + "time_per_iteration": 2.7452070713043213 + }, + { + "auxiliary_loss_clip": 0.01295466, + "auxiliary_loss_mlp": 0.01060621, + "balance_loss_clip": 1.08387446, + "balance_loss_mlp": 1.04424167, + "epoch": 0.04797691336499729, + "flos": 18835721210880.0, + "grad_norm": 6.836300935939038, + "language_loss": 0.88731349, + "learning_rate": 3.996633071017783e-06, + "loss": 0.91087437, + "num_input_tokens_seen": 8534085, + "step": 399, + "time_per_iteration": 2.638578414916992 + }, + { + "auxiliary_loss_clip": 0.01292637, + "auxiliary_loss_mlp": 0.01052581, + "balance_loss_clip": 1.08448911, + "balance_loss_mlp": 1.03611779, + "epoch": 0.04809715625563638, + "flos": 21099673578240.0, + "grad_norm": 2.069198433603079, + "language_loss": 0.81867951, + "learning_rate": 3.996587738582084e-06, + "loss": 0.84213167, + "num_input_tokens_seen": 8550885, + "step": 400, + "time_per_iteration": 2.6795759201049805 + }, + { + "auxiliary_loss_clip": 0.01292074, + "auxiliary_loss_mlp": 0.01058405, + "balance_loss_clip": 1.08128858, + "balance_loss_mlp": 1.04209709, + "epoch": 0.04821739914627548, + "flos": 23805650712960.0, + "grad_norm": 5.552876182070026, + "language_loss": 0.86176729, + "learning_rate": 3.9965421032656115e-06, + "loss": 0.88527215, + "num_input_tokens_seen": 8570815, + "step": 401, + "time_per_iteration": 2.716945171356201 + }, + { + "auxiliary_loss_clip": 0.01296362, + "auxiliary_loss_mlp": 0.01057213, + "balance_loss_clip": 1.08377624, + "balance_loss_mlp": 1.04035664, + "epoch": 0.04833764203691457, + "flos": 22200587475840.0, + "grad_norm": 2.2956763451408775, + "language_loss": 0.94551259, + "learning_rate": 3.99649616507529e-06, + "loss": 0.96904838, + "num_input_tokens_seen": 8589910, + "step": 402, + "time_per_iteration": 2.701979637145996 + }, + { + "auxiliary_loss_clip": 0.01202417, + "auxiliary_loss_mlp": 0.01006875, + "balance_loss_clip": 1.05866611, + "balance_loss_mlp": 1.00067639, + "epoch": 0.04845788492755366, + "flos": 65904376896000.0, + "grad_norm": 1.2080761345371382, + "language_loss": 0.63133281, + "learning_rate": 3.996449924018088e-06, + "loss": 0.65342575, + "num_input_tokens_seen": 8650370, + "step": 403, + "time_per_iteration": 3.1739790439605713 + }, + { + "auxiliary_loss_clip": 0.01289756, + "auxiliary_loss_mlp": 0.01064658, + "balance_loss_clip": 1.0823946, + "balance_loss_mlp": 1.04904151, + "epoch": 0.04857812781819275, + "flos": 19281301424640.0, + "grad_norm": 1.853913624734821, + "language_loss": 0.79533744, + "learning_rate": 3.99640338010102e-06, + "loss": 0.81888163, + "num_input_tokens_seen": 8669475, + "step": 404, + "time_per_iteration": 2.6897614002227783 + }, + { + "auxiliary_loss_clip": 0.01291952, + "auxiliary_loss_mlp": 0.0104485, + "balance_loss_clip": 1.08169556, + "balance_loss_mlp": 1.02945423, + "epoch": 0.04869837070883184, + "flos": 24062376193920.0, + "grad_norm": 4.465531405081163, + "language_loss": 0.78638864, + "learning_rate": 3.996356533331146e-06, + "loss": 0.8097567, + "num_input_tokens_seen": 8691345, + "step": 405, + "time_per_iteration": 2.7250633239746094 + }, + { + "auxiliary_loss_clip": 0.01294707, + "auxiliary_loss_mlp": 0.0106295, + "balance_loss_clip": 1.08075023, + "balance_loss_mlp": 1.04524755, + "epoch": 0.04881861359947093, + "flos": 25187169657600.0, + "grad_norm": 2.033451856409103, + "language_loss": 0.61806697, + "learning_rate": 3.996309383715573e-06, + "loss": 0.64164352, + "num_input_tokens_seen": 8710125, + "step": 406, + "time_per_iteration": 2.679795265197754 + }, + { + "auxiliary_loss_clip": 0.01294861, + "auxiliary_loss_mlp": 0.01063137, + "balance_loss_clip": 1.08226836, + "balance_loss_mlp": 1.0451479, + "epoch": 0.048938856490110025, + "flos": 16362913213440.0, + "grad_norm": 2.0196750100375214, + "language_loss": 0.73881757, + "learning_rate": 3.996261931261454e-06, + "loss": 0.76239753, + "num_input_tokens_seen": 8728705, + "step": 407, + "time_per_iteration": 2.714292526245117 + }, + { + "auxiliary_loss_clip": 0.01294972, + "auxiliary_loss_mlp": 0.01052998, + "balance_loss_clip": 1.08611977, + "balance_loss_mlp": 1.03642833, + "epoch": 0.049059099380749115, + "flos": 29895094379520.0, + "grad_norm": 1.6615366532900357, + "language_loss": 0.86450374, + "learning_rate": 3.996214175975987e-06, + "loss": 0.88798344, + "num_input_tokens_seen": 8749225, + "step": 408, + "time_per_iteration": 2.905061960220337 + }, + { + "auxiliary_loss_clip": 0.01297225, + "auxiliary_loss_mlp": 0.01060699, + "balance_loss_clip": 1.08568263, + "balance_loss_mlp": 1.04408097, + "epoch": 0.049179342271388204, + "flos": 35918858027520.0, + "grad_norm": 2.1922484795854174, + "language_loss": 0.78958458, + "learning_rate": 3.996166117866417e-06, + "loss": 0.81316382, + "num_input_tokens_seen": 8771160, + "step": 409, + "time_per_iteration": 2.9323787689208984 + }, + { + "auxiliary_loss_clip": 0.01287442, + "auxiliary_loss_mlp": 0.01057586, + "balance_loss_clip": 1.08024549, + "balance_loss_mlp": 1.04204059, + "epoch": 0.049299585162027294, + "flos": 14611226659200.0, + "grad_norm": 1.9386533942498452, + "language_loss": 0.870929, + "learning_rate": 3.996117756940035e-06, + "loss": 0.89437926, + "num_input_tokens_seen": 8787845, + "step": 410, + "time_per_iteration": 4.510929346084595 + }, + { + "auxiliary_loss_clip": 0.01292008, + "auxiliary_loss_mlp": 0.01048948, + "balance_loss_clip": 1.08246076, + "balance_loss_mlp": 1.0328548, + "epoch": 0.049419828052666384, + "flos": 19567939956480.0, + "grad_norm": 1.9885065336093548, + "language_loss": 0.97913486, + "learning_rate": 3.996069093204175e-06, + "loss": 1.0025444, + "num_input_tokens_seen": 8803805, + "step": 411, + "time_per_iteration": 2.730725049972534 + }, + { + "auxiliary_loss_clip": 0.01295016, + "auxiliary_loss_mlp": 0.01056346, + "balance_loss_clip": 1.08421421, + "balance_loss_mlp": 1.03975165, + "epoch": 0.049540070943305474, + "flos": 13659916907520.0, + "grad_norm": 2.2202028382605428, + "language_loss": 0.88102245, + "learning_rate": 3.996020126666221e-06, + "loss": 0.90453607, + "num_input_tokens_seen": 8820785, + "step": 412, + "time_per_iteration": 3.5646090507507324 + }, + { + "auxiliary_loss_clip": 0.01290158, + "auxiliary_loss_mlp": 0.01047334, + "balance_loss_clip": 1.08128393, + "balance_loss_mlp": 1.03168178, + "epoch": 0.04966031383394457, + "flos": 21832035978240.0, + "grad_norm": 2.353796120071092, + "language_loss": 0.81933606, + "learning_rate": 3.995970857333601e-06, + "loss": 0.84271097, + "num_input_tokens_seen": 8841195, + "step": 413, + "time_per_iteration": 2.8529696464538574 + }, + { + "auxiliary_loss_clip": 0.01293106, + "auxiliary_loss_mlp": 0.01066346, + "balance_loss_clip": 1.08131599, + "balance_loss_mlp": 1.04928672, + "epoch": 0.04978055672458366, + "flos": 28618793349120.0, + "grad_norm": 1.804488691368967, + "language_loss": 0.80048776, + "learning_rate": 3.995921285213789e-06, + "loss": 0.8240822, + "num_input_tokens_seen": 8861455, + "step": 414, + "time_per_iteration": 2.7929463386535645 + }, + { + "auxiliary_loss_clip": 0.01290154, + "auxiliary_loss_mlp": 0.01048359, + "balance_loss_clip": 1.08183849, + "balance_loss_mlp": 1.03187215, + "epoch": 0.04990079961522275, + "flos": 19828220883840.0, + "grad_norm": 2.1124455974236755, + "language_loss": 0.80938685, + "learning_rate": 3.995871410314305e-06, + "loss": 0.83277196, + "num_input_tokens_seen": 8880015, + "step": 415, + "time_per_iteration": 2.7432289123535156 + }, + { + "auxiliary_loss_clip": 0.01174037, + "auxiliary_loss_mlp": 0.01006319, + "balance_loss_clip": 1.0198102, + "balance_loss_mlp": 1.00028658, + "epoch": 0.05002104250586184, + "flos": 62735045293440.0, + "grad_norm": 0.9085135625214482, + "language_loss": 0.59618723, + "learning_rate": 3.995821232642714e-06, + "loss": 0.61799085, + "num_input_tokens_seen": 8938420, + "step": 416, + "time_per_iteration": 3.3494718074798584 + }, + { + "auxiliary_loss_clip": 0.0126495, + "auxiliary_loss_mlp": 0.01059542, + "balance_loss_clip": 1.04560971, + "balance_loss_mlp": 1.04378271, + "epoch": 0.05014128539650093, + "flos": 27928518710400.0, + "grad_norm": 2.1877103765409087, + "language_loss": 0.82271123, + "learning_rate": 3.995770752206629e-06, + "loss": 0.84595615, + "num_input_tokens_seen": 8959495, + "step": 417, + "time_per_iteration": 2.806138038635254 + }, + { + "auxiliary_loss_clip": 0.0129251, + "auxiliary_loss_mlp": 0.01059682, + "balance_loss_clip": 1.08249712, + "balance_loss_mlp": 1.04356503, + "epoch": 0.05026152828714002, + "flos": 17705576620800.0, + "grad_norm": 1.9398379161820145, + "language_loss": 0.97309178, + "learning_rate": 3.995719969013709e-06, + "loss": 0.99661374, + "num_input_tokens_seen": 8976675, + "step": 418, + "time_per_iteration": 2.639237403869629 + }, + { + "auxiliary_loss_clip": 0.01239565, + "auxiliary_loss_mlp": 0.01063897, + "balance_loss_clip": 1.00501752, + "balance_loss_mlp": 1.04696989, + "epoch": 0.05038177117777912, + "flos": 19133277477120.0, + "grad_norm": 4.371101631237686, + "language_loss": 0.85753793, + "learning_rate": 3.995668883071655e-06, + "loss": 0.88057256, + "num_input_tokens_seen": 8992900, + "step": 419, + "time_per_iteration": 2.836042642593384 + }, + { + "auxiliary_loss_clip": 0.01292131, + "auxiliary_loss_mlp": 0.01061153, + "balance_loss_clip": 1.08359957, + "balance_loss_mlp": 1.04517937, + "epoch": 0.050502014068418206, + "flos": 20667704618880.0, + "grad_norm": 2.2299258934220787, + "language_loss": 0.91049886, + "learning_rate": 3.995617494388219e-06, + "loss": 0.93403167, + "num_input_tokens_seen": 9011020, + "step": 420, + "time_per_iteration": 2.6927318572998047 + }, + { + "auxiliary_loss_clip": 0.01230135, + "auxiliary_loss_mlp": 0.01048505, + "balance_loss_clip": 0.99933314, + "balance_loss_mlp": 1.03080273, + "epoch": 0.050622256959057296, + "flos": 21361103740800.0, + "grad_norm": 2.0463487211447045, + "language_loss": 0.80626595, + "learning_rate": 3.995565802971196e-06, + "loss": 0.82905233, + "num_input_tokens_seen": 9030995, + "step": 421, + "time_per_iteration": 2.8453941345214844 + }, + { + "auxiliary_loss_clip": 0.01234735, + "auxiliary_loss_mlp": 0.01049805, + "balance_loss_clip": 1.00248075, + "balance_loss_mlp": 1.03402126, + "epoch": 0.050742499849696386, + "flos": 27673588909440.0, + "grad_norm": 1.7810212683675193, + "language_loss": 0.67619252, + "learning_rate": 3.995513808828427e-06, + "loss": 0.69903791, + "num_input_tokens_seen": 9053790, + "step": 422, + "time_per_iteration": 2.9919791221618652 + }, + { + "auxiliary_loss_clip": 0.01239088, + "auxiliary_loss_mlp": 0.01048503, + "balance_loss_clip": 1.00213993, + "balance_loss_mlp": 1.03367317, + "epoch": 0.050862742740335476, + "flos": 19865999013120.0, + "grad_norm": 1.8501717397427626, + "language_loss": 0.76827443, + "learning_rate": 3.9954615119678e-06, + "loss": 0.79115033, + "num_input_tokens_seen": 9072345, + "step": 423, + "time_per_iteration": 2.681816577911377 + }, + { + "auxiliary_loss_clip": 0.01250641, + "auxiliary_loss_mlp": 0.01056912, + "balance_loss_clip": 1.03947592, + "balance_loss_mlp": 1.04133153, + "epoch": 0.050982985630974566, + "flos": 22085098272000.0, + "grad_norm": 1.9818728418169338, + "language_loss": 0.80523258, + "learning_rate": 3.995408912397248e-06, + "loss": 0.82830811, + "num_input_tokens_seen": 9090240, + "step": 424, + "time_per_iteration": 2.8425939083099365 + }, + { + "auxiliary_loss_clip": 0.01238442, + "auxiliary_loss_mlp": 0.01058988, + "balance_loss_clip": 1.00468028, + "balance_loss_mlp": 1.0428952, + "epoch": 0.05110322852161366, + "flos": 20740962407040.0, + "grad_norm": 2.512777577334359, + "language_loss": 0.93550527, + "learning_rate": 3.99535601012475e-06, + "loss": 0.95847964, + "num_input_tokens_seen": 9105570, + "step": 425, + "time_per_iteration": 2.7259600162506104 + }, + { + "auxiliary_loss_clip": 0.01205259, + "auxiliary_loss_mlp": 0.00515495, + "balance_loss_clip": 0.96135706, + "balance_loss_mlp": 0.0, + "epoch": 0.05122347141225275, + "flos": 28547295327360.0, + "grad_norm": 1.5470187017123254, + "language_loss": 0.75635719, + "learning_rate": 3.995302805158333e-06, + "loss": 0.77356476, + "num_input_tokens_seen": 9128225, + "step": 426, + "time_per_iteration": 2.8370702266693115 + }, + { + "auxiliary_loss_clip": 0.01226568, + "auxiliary_loss_mlp": 0.01057596, + "balance_loss_clip": 0.99945867, + "balance_loss_mlp": 1.04087079, + "epoch": 0.05134371430289184, + "flos": 19722679747200.0, + "grad_norm": 1.9472657562662157, + "language_loss": 0.83757299, + "learning_rate": 3.9952492975060665e-06, + "loss": 0.86041468, + "num_input_tokens_seen": 9148295, + "step": 427, + "time_per_iteration": 2.7488653659820557 + }, + { + "auxiliary_loss_clip": 0.01258718, + "auxiliary_loss_mlp": 0.01052331, + "balance_loss_clip": 1.03958607, + "balance_loss_mlp": 1.03566551, + "epoch": 0.05146395719353093, + "flos": 34458945649920.0, + "grad_norm": 3.645575052231166, + "language_loss": 0.85194767, + "learning_rate": 3.995195487176067e-06, + "loss": 0.87505811, + "num_input_tokens_seen": 9168525, + "step": 428, + "time_per_iteration": 2.774198055267334 + }, + { + "auxiliary_loss_clip": 0.01287743, + "auxiliary_loss_mlp": 0.01054707, + "balance_loss_clip": 1.08167601, + "balance_loss_mlp": 1.03963327, + "epoch": 0.05158420008417002, + "flos": 21760286561280.0, + "grad_norm": 1.817046605066844, + "language_loss": 0.85528922, + "learning_rate": 3.995141374176499e-06, + "loss": 0.87871379, + "num_input_tokens_seen": 9186920, + "step": 429, + "time_per_iteration": 2.7040634155273438 + }, + { + "auxiliary_loss_clip": 0.0112894, + "auxiliary_loss_mlp": 0.00507454, + "balance_loss_clip": 0.94495553, + "balance_loss_mlp": 0.0, + "epoch": 0.05170444297480911, + "flos": 72553956226560.0, + "grad_norm": 0.8753026399816203, + "language_loss": 0.63144189, + "learning_rate": 3.995086958515572e-06, + "loss": 0.64780587, + "num_input_tokens_seen": 9244940, + "step": 430, + "time_per_iteration": 3.3600199222564697 + }, + { + "auxiliary_loss_clip": 0.01194286, + "auxiliary_loss_mlp": 0.00507452, + "balance_loss_clip": 1.05788732, + "balance_loss_mlp": 0.0, + "epoch": 0.05182468586544821, + "flos": 62416159326720.0, + "grad_norm": 0.9623008280062817, + "language_loss": 0.59934908, + "learning_rate": 3.995032240201538e-06, + "loss": 0.61636645, + "num_input_tokens_seen": 9307335, + "step": 431, + "time_per_iteration": 3.2102572917938232 + }, + { + "auxiliary_loss_clip": 0.01142645, + "auxiliary_loss_mlp": 0.01005786, + "balance_loss_clip": 0.97366929, + "balance_loss_mlp": 0.99982595, + "epoch": 0.0519449287560873, + "flos": 41225989432320.0, + "grad_norm": 0.9328756029600126, + "language_loss": 0.63146824, + "learning_rate": 3.9949772192427e-06, + "loss": 0.65295255, + "num_input_tokens_seen": 9353960, + "step": 432, + "time_per_iteration": 2.9310860633850098 + }, + { + "auxiliary_loss_clip": 0.01232779, + "auxiliary_loss_mlp": 0.01056062, + "balance_loss_clip": 0.99769944, + "balance_loss_mlp": 1.04125595, + "epoch": 0.05206517164672639, + "flos": 17494530261120.0, + "grad_norm": 1.7902601082825507, + "language_loss": 0.7984271, + "learning_rate": 3.994921895647405e-06, + "loss": 0.82131553, + "num_input_tokens_seen": 9372130, + "step": 433, + "time_per_iteration": 2.7579405307769775 + }, + { + "auxiliary_loss_clip": 0.01187883, + "auxiliary_loss_mlp": 0.01006337, + "balance_loss_clip": 1.0539763, + "balance_loss_mlp": 1.00071013, + "epoch": 0.05218541453736548, + "flos": 64002762973440.0, + "grad_norm": 0.8391807804169586, + "language_loss": 0.5527578, + "learning_rate": 3.994866269424043e-06, + "loss": 0.5747, + "num_input_tokens_seen": 9428500, + "step": 434, + "time_per_iteration": 3.1399502754211426 + }, + { + "auxiliary_loss_clip": 0.01145772, + "auxiliary_loss_mlp": 0.01053429, + "balance_loss_clip": 0.8729583, + "balance_loss_mlp": 1.03715694, + "epoch": 0.05230565742800457, + "flos": 19317319787520.0, + "grad_norm": 2.279685967385225, + "language_loss": 0.78639686, + "learning_rate": 3.9948103405810545e-06, + "loss": 0.80838883, + "num_input_tokens_seen": 9447450, + "step": 435, + "time_per_iteration": 2.9196832180023193 + }, + { + "auxiliary_loss_clip": 0.01190422, + "auxiliary_loss_mlp": 0.0105517, + "balance_loss_clip": 0.95581955, + "balance_loss_mlp": 1.03930271, + "epoch": 0.05242590031864366, + "flos": 25298636538240.0, + "grad_norm": 1.8460545421765886, + "language_loss": 0.86060911, + "learning_rate": 3.994754109126923e-06, + "loss": 0.88306504, + "num_input_tokens_seen": 9468945, + "step": 436, + "time_per_iteration": 4.777956247329712 + }, + { + "auxiliary_loss_clip": 0.01140439, + "auxiliary_loss_mlp": 0.0105388, + "balance_loss_clip": 0.88008356, + "balance_loss_mlp": 1.03698826, + "epoch": 0.052546143209282754, + "flos": 26211629456640.0, + "grad_norm": 1.8012689033172269, + "language_loss": 0.93562555, + "learning_rate": 3.994697575070181e-06, + "loss": 0.95756876, + "num_input_tokens_seen": 9488405, + "step": 437, + "time_per_iteration": 3.7453064918518066 + }, + { + "auxiliary_loss_clip": 0.01236164, + "auxiliary_loss_mlp": 0.01057706, + "balance_loss_clip": 1.00505638, + "balance_loss_mlp": 1.04297221, + "epoch": 0.052666386099921844, + "flos": 22158140578560.0, + "grad_norm": 4.46876615096418, + "language_loss": 0.91342115, + "learning_rate": 3.994640738419402e-06, + "loss": 0.93635994, + "num_input_tokens_seen": 9507780, + "step": 438, + "time_per_iteration": 3.6954023838043213 + }, + { + "auxiliary_loss_clip": 0.01256842, + "auxiliary_loss_mlp": 0.0105425, + "balance_loss_clip": 1.03978109, + "balance_loss_mlp": 1.03818107, + "epoch": 0.052786628990560934, + "flos": 23881817502720.0, + "grad_norm": 1.7861111646238026, + "language_loss": 0.80875403, + "learning_rate": 3.9945835991832075e-06, + "loss": 0.83186495, + "num_input_tokens_seen": 9529665, + "step": 439, + "time_per_iteration": 2.800873041152954 + }, + { + "auxiliary_loss_clip": 0.01287116, + "auxiliary_loss_mlp": 0.01052134, + "balance_loss_clip": 1.08354044, + "balance_loss_mlp": 1.03674388, + "epoch": 0.052906871881200024, + "flos": 24605021934720.0, + "grad_norm": 1.9860012597603227, + "language_loss": 0.92946678, + "learning_rate": 3.994526157370268e-06, + "loss": 0.95285928, + "num_input_tokens_seen": 9548280, + "step": 440, + "time_per_iteration": 2.8215529918670654 + }, + { + "auxiliary_loss_clip": 0.01132328, + "auxiliary_loss_mlp": 0.01006506, + "balance_loss_clip": 0.96516603, + "balance_loss_mlp": 1.00097513, + "epoch": 0.053027114771839114, + "flos": 56461631143680.0, + "grad_norm": 0.8988608815543666, + "language_loss": 0.59265351, + "learning_rate": 3.994468412989296e-06, + "loss": 0.61404181, + "num_input_tokens_seen": 9609690, + "step": 441, + "time_per_iteration": 3.37654185295105 + }, + { + "auxiliary_loss_clip": 0.01213676, + "auxiliary_loss_mlp": 0.01056995, + "balance_loss_clip": 0.99691355, + "balance_loss_mlp": 1.04113984, + "epoch": 0.053147357662478203, + "flos": 17311098481920.0, + "grad_norm": 2.0697456328957364, + "language_loss": 0.92747396, + "learning_rate": 3.994410366049052e-06, + "loss": 0.95018071, + "num_input_tokens_seen": 9627550, + "step": 442, + "time_per_iteration": 2.7230796813964844 + }, + { + "auxiliary_loss_clip": 0.01258907, + "auxiliary_loss_mlp": 0.01057148, + "balance_loss_clip": 1.04126048, + "balance_loss_mlp": 1.04116166, + "epoch": 0.0532676005531173, + "flos": 17164977955200.0, + "grad_norm": 2.023915516853362, + "language_loss": 0.8273446, + "learning_rate": 3.994352016558341e-06, + "loss": 0.85050511, + "num_input_tokens_seen": 9644855, + "step": 443, + "time_per_iteration": 2.6917569637298584 + }, + { + "auxiliary_loss_clip": 0.01256866, + "auxiliary_loss_mlp": 0.01045926, + "balance_loss_clip": 1.04260218, + "balance_loss_mlp": 1.03001118, + "epoch": 0.05338784344375639, + "flos": 27819960831360.0, + "grad_norm": 2.02772849863712, + "language_loss": 0.73882961, + "learning_rate": 3.994293364526014e-06, + "loss": 0.76185757, + "num_input_tokens_seen": 9665740, + "step": 444, + "time_per_iteration": 2.7271811962127686 + }, + { + "auxiliary_loss_clip": 0.01224127, + "auxiliary_loss_mlp": 0.0105304, + "balance_loss_clip": 1.00383091, + "balance_loss_mlp": 1.03614843, + "epoch": 0.05350808633439548, + "flos": 21507691144320.0, + "grad_norm": 1.9376179696698559, + "language_loss": 0.84877461, + "learning_rate": 3.99423440996097e-06, + "loss": 0.87154627, + "num_input_tokens_seen": 9685280, + "step": 445, + "time_per_iteration": 2.7683839797973633 + }, + { + "auxiliary_loss_clip": 0.01240202, + "auxiliary_loss_mlp": 0.01044301, + "balance_loss_clip": 1.00510263, + "balance_loss_mlp": 1.02867222, + "epoch": 0.05362832922503457, + "flos": 20084299920000.0, + "grad_norm": 2.0862945153671557, + "language_loss": 0.81773436, + "learning_rate": 3.994175152872152e-06, + "loss": 0.84057939, + "num_input_tokens_seen": 9704365, + "step": 446, + "time_per_iteration": 2.7183094024658203 + }, + { + "auxiliary_loss_clip": 0.01264543, + "auxiliary_loss_mlp": 0.01051841, + "balance_loss_clip": 1.04066718, + "balance_loss_mlp": 1.03691602, + "epoch": 0.05374857211567366, + "flos": 26137222433280.0, + "grad_norm": 2.111609887090029, + "language_loss": 0.78786194, + "learning_rate": 3.994115593268548e-06, + "loss": 0.81102586, + "num_input_tokens_seen": 9724145, + "step": 447, + "time_per_iteration": 2.759817600250244 + }, + { + "auxiliary_loss_clip": 0.01284744, + "auxiliary_loss_mlp": 0.01054472, + "balance_loss_clip": 1.08091724, + "balance_loss_mlp": 1.03875971, + "epoch": 0.05386881500631275, + "flos": 27486817165440.0, + "grad_norm": 2.1352955481481506, + "language_loss": 0.82135719, + "learning_rate": 3.994055731159195e-06, + "loss": 0.84474933, + "num_input_tokens_seen": 9741615, + "step": 448, + "time_per_iteration": 2.6753897666931152 + }, + { + "auxiliary_loss_clip": 0.01263934, + "auxiliary_loss_mlp": 0.01049951, + "balance_loss_clip": 1.04662192, + "balance_loss_mlp": 1.03618884, + "epoch": 0.053989057896951846, + "flos": 23585087249280.0, + "grad_norm": 1.8927161435662656, + "language_loss": 0.87387037, + "learning_rate": 3.993995566553172e-06, + "loss": 0.89700925, + "num_input_tokens_seen": 9760580, + "step": 449, + "time_per_iteration": 2.7213544845581055 + }, + { + "auxiliary_loss_clip": 0.0122135, + "auxiliary_loss_mlp": 0.0105977, + "balance_loss_clip": 0.99717754, + "balance_loss_mlp": 1.0440228, + "epoch": 0.054109300787590936, + "flos": 25228862369280.0, + "grad_norm": 1.6513930466532951, + "language_loss": 0.77212262, + "learning_rate": 3.993935099459607e-06, + "loss": 0.79493386, + "num_input_tokens_seen": 9782195, + "step": 450, + "time_per_iteration": 2.736074209213257 + }, + { + "auxiliary_loss_clip": 0.01280479, + "auxiliary_loss_mlp": 0.01055276, + "balance_loss_clip": 1.07890415, + "balance_loss_mlp": 1.0404582, + "epoch": 0.054229543678230026, + "flos": 23841525421440.0, + "grad_norm": 1.9831705643400326, + "language_loss": 0.74373329, + "learning_rate": 3.993874329887673e-06, + "loss": 0.76709086, + "num_input_tokens_seen": 9800850, + "step": 451, + "time_per_iteration": 2.6992084980010986 + }, + { + "auxiliary_loss_clip": 0.01257892, + "auxiliary_loss_mlp": 0.01053947, + "balance_loss_clip": 1.04197335, + "balance_loss_mlp": 1.03862238, + "epoch": 0.054349786568869116, + "flos": 16320933192960.0, + "grad_norm": 2.3293809414664532, + "language_loss": 0.86372465, + "learning_rate": 3.993813257846589e-06, + "loss": 0.88684303, + "num_input_tokens_seen": 9817605, + "step": 452, + "time_per_iteration": 2.680513381958008 + }, + { + "auxiliary_loss_clip": 0.01259992, + "auxiliary_loss_mlp": 0.01050161, + "balance_loss_clip": 1.04182184, + "balance_loss_mlp": 1.0334003, + "epoch": 0.054470029459508205, + "flos": 18660729127680.0, + "grad_norm": 2.221575105695271, + "language_loss": 0.93172628, + "learning_rate": 3.993751883345619e-06, + "loss": 0.95482773, + "num_input_tokens_seen": 9835965, + "step": 453, + "time_per_iteration": 2.7179136276245117 + }, + { + "auxiliary_loss_clip": 0.01227602, + "auxiliary_loss_mlp": 0.01055836, + "balance_loss_clip": 1.00445747, + "balance_loss_mlp": 1.04074359, + "epoch": 0.054590272350147295, + "flos": 17785298856960.0, + "grad_norm": 2.646076823410791, + "language_loss": 0.87439424, + "learning_rate": 3.993690206394073e-06, + "loss": 0.8972286, + "num_input_tokens_seen": 9852265, + "step": 454, + "time_per_iteration": 2.7003259658813477 + }, + { + "auxiliary_loss_clip": 0.01240605, + "auxiliary_loss_mlp": 0.01067081, + "balance_loss_clip": 1.00337505, + "balance_loss_mlp": 1.05104184, + "epoch": 0.054710515240786385, + "flos": 17785945301760.0, + "grad_norm": 3.6476264250874437, + "language_loss": 0.87820685, + "learning_rate": 3.993628227001307e-06, + "loss": 0.90128374, + "num_input_tokens_seen": 9870465, + "step": 455, + "time_per_iteration": 2.7349660396575928 + }, + { + "auxiliary_loss_clip": 0.01230297, + "auxiliary_loss_mlp": 0.0105433, + "balance_loss_clip": 0.99956113, + "balance_loss_mlp": 1.03969133, + "epoch": 0.05483075813142548, + "flos": 48210900180480.0, + "grad_norm": 1.7844054837973493, + "language_loss": 0.71407294, + "learning_rate": 3.993565945176726e-06, + "loss": 0.73691916, + "num_input_tokens_seen": 9891490, + "step": 456, + "time_per_iteration": 2.9705584049224854 + }, + { + "auxiliary_loss_clip": 0.01233618, + "auxiliary_loss_mlp": 0.01050536, + "balance_loss_clip": 1.0049119, + "balance_loss_mlp": 1.03612339, + "epoch": 0.05495100102206457, + "flos": 19682244011520.0, + "grad_norm": 1.8485485557024541, + "language_loss": 0.8421222, + "learning_rate": 3.993503360929776e-06, + "loss": 0.86496377, + "num_input_tokens_seen": 9910375, + "step": 457, + "time_per_iteration": 2.6853785514831543 + }, + { + "auxiliary_loss_clip": 0.01126283, + "auxiliary_loss_mlp": 0.01055954, + "balance_loss_clip": 0.84320545, + "balance_loss_mlp": 1.03875184, + "epoch": 0.05507124391270366, + "flos": 26360048453760.0, + "grad_norm": 2.6209520250540232, + "language_loss": 0.81300551, + "learning_rate": 3.99344047426995e-06, + "loss": 0.8348279, + "num_input_tokens_seen": 9931635, + "step": 458, + "time_per_iteration": 3.320645570755005 + }, + { + "auxiliary_loss_clip": 0.01185234, + "auxiliary_loss_mlp": 0.01051587, + "balance_loss_clip": 0.92278999, + "balance_loss_mlp": 1.03371751, + "epoch": 0.05519148680334275, + "flos": 22601314581120.0, + "grad_norm": 2.5185074368027, + "language_loss": 0.93173069, + "learning_rate": 3.993377285206789e-06, + "loss": 0.95409894, + "num_input_tokens_seen": 9951420, + "step": 459, + "time_per_iteration": 3.144214391708374 + }, + { + "auxiliary_loss_clip": 0.0118576, + "auxiliary_loss_mlp": 0.01052, + "balance_loss_clip": 0.95762932, + "balance_loss_mlp": 1.03591847, + "epoch": 0.05531172969398184, + "flos": 40552519380480.0, + "grad_norm": 1.6820788267254256, + "language_loss": 0.86246049, + "learning_rate": 3.99331379374988e-06, + "loss": 0.8848381, + "num_input_tokens_seen": 9975025, + "step": 460, + "time_per_iteration": 2.910369873046875 + }, + { + "auxiliary_loss_clip": 0.01231786, + "auxiliary_loss_mlp": 0.01049222, + "balance_loss_clip": 0.99505436, + "balance_loss_mlp": 1.0341717, + "epoch": 0.05543197258462093, + "flos": 23477894087040.0, + "grad_norm": 2.024367375453346, + "language_loss": 0.79868549, + "learning_rate": 3.993249999908852e-06, + "loss": 0.82149553, + "num_input_tokens_seen": 9995175, + "step": 461, + "time_per_iteration": 2.8193397521972656 + }, + { + "auxiliary_loss_clip": 0.01283989, + "auxiliary_loss_mlp": 0.01062477, + "balance_loss_clip": 1.07799053, + "balance_loss_mlp": 1.04619312, + "epoch": 0.05555221547526003, + "flos": 18624603024000.0, + "grad_norm": 2.469837028132748, + "language_loss": 0.87411571, + "learning_rate": 3.993185903693384e-06, + "loss": 0.89758039, + "num_input_tokens_seen": 10011975, + "step": 462, + "time_per_iteration": 3.5265212059020996 + }, + { + "auxiliary_loss_clip": 0.01230559, + "auxiliary_loss_mlp": 0.0104991, + "balance_loss_clip": 0.99978989, + "balance_loss_mlp": 1.03557527, + "epoch": 0.05567245836589912, + "flos": 23587098410880.0, + "grad_norm": 2.4530097275511693, + "language_loss": 0.82825983, + "learning_rate": 3.9931215051131995e-06, + "loss": 0.8510645, + "num_input_tokens_seen": 10032620, + "step": 463, + "time_per_iteration": 3.745349407196045 + }, + { + "auxiliary_loss_clip": 0.01239443, + "auxiliary_loss_mlp": 0.01053195, + "balance_loss_clip": 1.0004288, + "balance_loss_mlp": 1.03745961, + "epoch": 0.05579270125653821, + "flos": 27746667129600.0, + "grad_norm": 1.5384739923430089, + "language_loss": 0.80132294, + "learning_rate": 3.993056804178068e-06, + "loss": 0.82424933, + "num_input_tokens_seen": 10054165, + "step": 464, + "time_per_iteration": 3.6201422214508057 + }, + { + "auxiliary_loss_clip": 0.01171132, + "auxiliary_loss_mlp": 0.01057079, + "balance_loss_clip": 0.92054969, + "balance_loss_mlp": 1.04135489, + "epoch": 0.0559129441471773, + "flos": 27014161075200.0, + "grad_norm": 1.9879290889310754, + "language_loss": 0.84370112, + "learning_rate": 3.992991800897803e-06, + "loss": 0.86598325, + "num_input_tokens_seen": 10073970, + "step": 465, + "time_per_iteration": 2.833949565887451 + }, + { + "auxiliary_loss_clip": 0.01277951, + "auxiliary_loss_mlp": 0.01048739, + "balance_loss_clip": 1.07576537, + "balance_loss_mlp": 1.03276527, + "epoch": 0.05603318703781639, + "flos": 15229787794560.0, + "grad_norm": 2.343212354208923, + "language_loss": 0.89863884, + "learning_rate": 3.9929264952822665e-06, + "loss": 0.9219057, + "num_input_tokens_seen": 10091505, + "step": 466, + "time_per_iteration": 2.7588226795196533 + }, + { + "auxiliary_loss_clip": 0.01257596, + "auxiliary_loss_mlp": 0.0105533, + "balance_loss_clip": 1.03884423, + "balance_loss_mlp": 1.04007161, + "epoch": 0.05615342992845548, + "flos": 22266482976000.0, + "grad_norm": 1.8896494589070987, + "language_loss": 0.88373744, + "learning_rate": 3.992860887341366e-06, + "loss": 0.90686667, + "num_input_tokens_seen": 10109675, + "step": 467, + "time_per_iteration": 2.761749505996704 + }, + { + "auxiliary_loss_clip": 0.01192389, + "auxiliary_loss_mlp": 0.0105195, + "balance_loss_clip": 0.95801842, + "balance_loss_mlp": 1.03770447, + "epoch": 0.056273672819094574, + "flos": 23584979508480.0, + "grad_norm": 1.9264898641235544, + "language_loss": 0.81124604, + "learning_rate": 3.992794977085052e-06, + "loss": 0.83368945, + "num_input_tokens_seen": 10127675, + "step": 468, + "time_per_iteration": 2.8701303005218506 + }, + { + "auxiliary_loss_clip": 0.01207157, + "auxiliary_loss_mlp": 0.01052977, + "balance_loss_clip": 0.96220773, + "balance_loss_mlp": 1.03682446, + "epoch": 0.056393915709733664, + "flos": 19858708552320.0, + "grad_norm": 1.9845994330754715, + "language_loss": 0.84502804, + "learning_rate": 3.992728764523326e-06, + "loss": 0.86762947, + "num_input_tokens_seen": 10146620, + "step": 469, + "time_per_iteration": 2.847489595413208 + }, + { + "auxiliary_loss_clip": 0.01227696, + "auxiliary_loss_mlp": 0.0105359, + "balance_loss_clip": 1.00141478, + "balance_loss_mlp": 1.03851032, + "epoch": 0.05651415860037275, + "flos": 22163779013760.0, + "grad_norm": 1.7746422452498898, + "language_loss": 0.80732679, + "learning_rate": 3.99266224966623e-06, + "loss": 0.83013976, + "num_input_tokens_seen": 10167535, + "step": 470, + "time_per_iteration": 2.768880844116211 + }, + { + "auxiliary_loss_clip": 0.01214818, + "auxiliary_loss_mlp": 0.01048549, + "balance_loss_clip": 0.99653322, + "balance_loss_mlp": 1.0341959, + "epoch": 0.05663440149101184, + "flos": 19463548055040.0, + "grad_norm": 1.9119365854336674, + "language_loss": 0.88102579, + "learning_rate": 3.992595432523855e-06, + "loss": 0.90365946, + "num_input_tokens_seen": 10184825, + "step": 471, + "time_per_iteration": 2.7849392890930176 + }, + { + "auxiliary_loss_clip": 0.01187555, + "auxiliary_loss_mlp": 0.01050986, + "balance_loss_clip": 0.95916957, + "balance_loss_mlp": 1.03580475, + "epoch": 0.05675464438165093, + "flos": 22670226823680.0, + "grad_norm": 3.4048866625026823, + "language_loss": 0.86244774, + "learning_rate": 3.992528313106338e-06, + "loss": 0.88483316, + "num_input_tokens_seen": 10203025, + "step": 472, + "time_per_iteration": 2.7848799228668213 + }, + { + "auxiliary_loss_clip": 0.01279879, + "auxiliary_loss_mlp": 0.00515471, + "balance_loss_clip": 1.08002245, + "balance_loss_mlp": 0.0, + "epoch": 0.05687488727229002, + "flos": 16901177495040.0, + "grad_norm": 2.54016353581739, + "language_loss": 0.82291019, + "learning_rate": 3.9924608914238595e-06, + "loss": 0.8408637, + "num_input_tokens_seen": 10218020, + "step": 473, + "time_per_iteration": 2.766227960586548 + }, + { + "auxiliary_loss_clip": 0.01255457, + "auxiliary_loss_mlp": 0.01047499, + "balance_loss_clip": 1.04029584, + "balance_loss_mlp": 1.03363526, + "epoch": 0.05699513016292912, + "flos": 29168980945920.0, + "grad_norm": 2.126530453633368, + "language_loss": 0.84192181, + "learning_rate": 3.992393167486648e-06, + "loss": 0.86495137, + "num_input_tokens_seen": 10237170, + "step": 474, + "time_per_iteration": 2.883812427520752 + }, + { + "auxiliary_loss_clip": 0.01280689, + "auxiliary_loss_mlp": 0.01059651, + "balance_loss_clip": 1.07863247, + "balance_loss_mlp": 1.04358196, + "epoch": 0.05711537305356821, + "flos": 18916197632640.0, + "grad_norm": 2.4342034232870455, + "language_loss": 0.80456769, + "learning_rate": 3.992325141304977e-06, + "loss": 0.8279711, + "num_input_tokens_seen": 10255125, + "step": 475, + "time_per_iteration": 2.7287158966064453 + }, + { + "auxiliary_loss_clip": 0.01183806, + "auxiliary_loss_mlp": 0.01045551, + "balance_loss_clip": 0.95617759, + "balance_loss_mlp": 1.03220522, + "epoch": 0.0572356159442073, + "flos": 26758979879040.0, + "grad_norm": 2.1991127680203193, + "language_loss": 0.86625969, + "learning_rate": 3.992256812889166e-06, + "loss": 0.88855326, + "num_input_tokens_seen": 10271230, + "step": 476, + "time_per_iteration": 2.9119653701782227 + }, + { + "auxiliary_loss_clip": 0.01281354, + "auxiliary_loss_mlp": 0.01047913, + "balance_loss_clip": 1.08029628, + "balance_loss_mlp": 1.03245175, + "epoch": 0.05735585883484639, + "flos": 35116146840960.0, + "grad_norm": 59.385630888979094, + "language_loss": 0.76668811, + "learning_rate": 3.992188182249582e-06, + "loss": 0.78998083, + "num_input_tokens_seen": 10293125, + "step": 477, + "time_per_iteration": 2.819373846054077 + }, + { + "auxiliary_loss_clip": 0.01226237, + "auxiliary_loss_mlp": 0.01055081, + "balance_loss_clip": 1.00102043, + "balance_loss_mlp": 1.03920281, + "epoch": 0.05747610172548548, + "flos": 18734381965440.0, + "grad_norm": 2.0521380331936743, + "language_loss": 0.90778756, + "learning_rate": 3.992119249396633e-06, + "loss": 0.9306007, + "num_input_tokens_seen": 10311810, + "step": 478, + "time_per_iteration": 2.7535667419433594 + }, + { + "auxiliary_loss_clip": 0.0121673, + "auxiliary_loss_mlp": 0.00515286, + "balance_loss_clip": 0.9967798, + "balance_loss_mlp": 0.0, + "epoch": 0.05759634461612457, + "flos": 27964752554880.0, + "grad_norm": 2.1860972088892208, + "language_loss": 0.82009673, + "learning_rate": 3.992050014340778e-06, + "loss": 0.83741695, + "num_input_tokens_seen": 10332165, + "step": 479, + "time_per_iteration": 2.822688102722168 + }, + { + "auxiliary_loss_clip": 0.01157121, + "auxiliary_loss_mlp": 0.01008606, + "balance_loss_clip": 1.01645422, + "balance_loss_mlp": 1.00333691, + "epoch": 0.057716587506763666, + "flos": 69292009405440.0, + "grad_norm": 0.825321858424645, + "language_loss": 0.54964906, + "learning_rate": 3.99198047709252e-06, + "loss": 0.57130635, + "num_input_tokens_seen": 10393685, + "step": 480, + "time_per_iteration": 3.3563618659973145 + }, + { + "auxiliary_loss_clip": 0.01200653, + "auxiliary_loss_mlp": 0.01055799, + "balance_loss_clip": 0.95704174, + "balance_loss_mlp": 1.04086185, + "epoch": 0.057836830397402755, + "flos": 25009196745600.0, + "grad_norm": 1.7625180307481516, + "language_loss": 0.78690004, + "learning_rate": 3.991910637662408e-06, + "loss": 0.80946457, + "num_input_tokens_seen": 10413975, + "step": 481, + "time_per_iteration": 2.756929397583008 + }, + { + "auxiliary_loss_clip": 0.01275958, + "auxiliary_loss_mlp": 0.01045574, + "balance_loss_clip": 1.07791889, + "balance_loss_mlp": 1.031317, + "epoch": 0.057957073288041845, + "flos": 25593894334080.0, + "grad_norm": 1.8471455916729438, + "language_loss": 0.80777174, + "learning_rate": 3.9918404960610355e-06, + "loss": 0.8309871, + "num_input_tokens_seen": 10433005, + "step": 482, + "time_per_iteration": 2.701951026916504 + }, + { + "auxiliary_loss_clip": 0.01255099, + "auxiliary_loss_mlp": 0.01065866, + "balance_loss_clip": 1.03766465, + "balance_loss_mlp": 1.05088162, + "epoch": 0.058077316178680935, + "flos": 20777411733120.0, + "grad_norm": 2.1753983187819297, + "language_loss": 0.77444053, + "learning_rate": 3.991770052299043e-06, + "loss": 0.79765016, + "num_input_tokens_seen": 10451235, + "step": 483, + "time_per_iteration": 2.687490224838257 + }, + { + "auxiliary_loss_clip": 0.012222, + "auxiliary_loss_mlp": 0.01046148, + "balance_loss_clip": 0.99675119, + "balance_loss_mlp": 1.03234971, + "epoch": 0.058197559069320025, + "flos": 18916484941440.0, + "grad_norm": 2.3480716699775703, + "language_loss": 0.8760702, + "learning_rate": 3.991699306387118e-06, + "loss": 0.89875364, + "num_input_tokens_seen": 10469705, + "step": 484, + "time_per_iteration": 2.7855422496795654 + }, + { + "auxiliary_loss_clip": 0.01256868, + "auxiliary_loss_mlp": 0.01056917, + "balance_loss_clip": 1.03954613, + "balance_loss_mlp": 1.04292154, + "epoch": 0.058317801959959115, + "flos": 24863327614080.0, + "grad_norm": 2.2043697243116775, + "language_loss": 0.78135127, + "learning_rate": 3.991628258335991e-06, + "loss": 0.80448908, + "num_input_tokens_seen": 10491910, + "step": 485, + "time_per_iteration": 2.779918909072876 + }, + { + "auxiliary_loss_clip": 0.01187905, + "auxiliary_loss_mlp": 0.01056963, + "balance_loss_clip": 0.95570618, + "balance_loss_mlp": 1.04265761, + "epoch": 0.05843804485059821, + "flos": 23257977068160.0, + "grad_norm": 3.185629665391335, + "language_loss": 0.87601805, + "learning_rate": 3.991556908156442e-06, + "loss": 0.89846671, + "num_input_tokens_seen": 10508435, + "step": 486, + "time_per_iteration": 2.7137932777404785 + }, + { + "auxiliary_loss_clip": 0.01226466, + "auxiliary_loss_mlp": 0.01055598, + "balance_loss_clip": 0.99668074, + "balance_loss_mlp": 1.04099488, + "epoch": 0.0585582877412373, + "flos": 23150532510720.0, + "grad_norm": 1.8584879753961594, + "language_loss": 0.87517202, + "learning_rate": 3.9914852558592914e-06, + "loss": 0.89799261, + "num_input_tokens_seen": 10529485, + "step": 487, + "time_per_iteration": 2.7718820571899414 + }, + { + "auxiliary_loss_clip": 0.01251717, + "auxiliary_loss_mlp": 0.01044622, + "balance_loss_clip": 1.03851545, + "balance_loss_mlp": 1.0296371, + "epoch": 0.05867853063187639, + "flos": 23506406507520.0, + "grad_norm": 3.3524124136179587, + "language_loss": 0.80442142, + "learning_rate": 3.991413301455413e-06, + "loss": 0.82738483, + "num_input_tokens_seen": 10545935, + "step": 488, + "time_per_iteration": 4.574707508087158 + }, + { + "auxiliary_loss_clip": 0.01213118, + "auxiliary_loss_mlp": 0.01047772, + "balance_loss_clip": 0.99608815, + "balance_loss_mlp": 1.03405106, + "epoch": 0.05879877352251548, + "flos": 29495803818240.0, + "grad_norm": 2.206263796094367, + "language_loss": 0.77964354, + "learning_rate": 3.991341044955719e-06, + "loss": 0.80225241, + "num_input_tokens_seen": 10565690, + "step": 489, + "time_per_iteration": 3.6783530712127686 + }, + { + "auxiliary_loss_clip": 0.01247647, + "auxiliary_loss_mlp": 0.00516131, + "balance_loss_clip": 1.03473592, + "balance_loss_mlp": 0.0, + "epoch": 0.05891901641315457, + "flos": 20157485880960.0, + "grad_norm": 2.1290706641759067, + "language_loss": 0.81785572, + "learning_rate": 3.991268486371172e-06, + "loss": 0.83549356, + "num_input_tokens_seen": 10584245, + "step": 490, + "time_per_iteration": 3.5180275440216064 + }, + { + "auxiliary_loss_clip": 0.01225499, + "auxiliary_loss_mlp": 0.0103936, + "balance_loss_clip": 0.99661374, + "balance_loss_mlp": 1.02517366, + "epoch": 0.05903925930379366, + "flos": 24644200694400.0, + "grad_norm": 2.5273528944414294, + "language_loss": 0.87939316, + "learning_rate": 3.991195625712779e-06, + "loss": 0.90204167, + "num_input_tokens_seen": 10601210, + "step": 491, + "time_per_iteration": 2.7909281253814697 + }, + { + "auxiliary_loss_clip": 0.01271869, + "auxiliary_loss_mlp": 0.01044399, + "balance_loss_clip": 1.07622874, + "balance_loss_mlp": 1.02997518, + "epoch": 0.05915950219443276, + "flos": 21250391045760.0, + "grad_norm": 1.917361384053404, + "language_loss": 0.81583214, + "learning_rate": 3.991122462991592e-06, + "loss": 0.8389948, + "num_input_tokens_seen": 10620730, + "step": 492, + "time_per_iteration": 2.60205340385437 + }, + { + "auxiliary_loss_clip": 0.01279079, + "auxiliary_loss_mlp": 0.01046242, + "balance_loss_clip": 1.07770181, + "balance_loss_mlp": 1.03023267, + "epoch": 0.05927974508507185, + "flos": 9902727319680.0, + "grad_norm": 3.305271231801498, + "language_loss": 0.80746591, + "learning_rate": 3.991048998218712e-06, + "loss": 0.83071911, + "num_input_tokens_seen": 10634035, + "step": 493, + "time_per_iteration": 2.6347432136535645 + }, + { + "auxiliary_loss_clip": 0.01251672, + "auxiliary_loss_mlp": 0.01060448, + "balance_loss_clip": 1.03751314, + "balance_loss_mlp": 1.04573762, + "epoch": 0.05939998797571094, + "flos": 18259499232000.0, + "grad_norm": 2.690095747519396, + "language_loss": 0.76630431, + "learning_rate": 3.990975231405281e-06, + "loss": 0.78942555, + "num_input_tokens_seen": 10652485, + "step": 494, + "time_per_iteration": 2.6686460971832275 + }, + { + "auxiliary_loss_clip": 0.01248974, + "auxiliary_loss_mlp": 0.01061049, + "balance_loss_clip": 1.03753018, + "balance_loss_mlp": 1.04642248, + "epoch": 0.05952023086635003, + "flos": 28256598558720.0, + "grad_norm": 1.8064610958178353, + "language_loss": 0.78641045, + "learning_rate": 3.990901162562491e-06, + "loss": 0.80951071, + "num_input_tokens_seen": 10673175, + "step": 495, + "time_per_iteration": 2.9256503582000732 + }, + { + "auxiliary_loss_clip": 0.01190587, + "auxiliary_loss_mlp": 0.00516011, + "balance_loss_clip": 0.95059347, + "balance_loss_mlp": 0.0, + "epoch": 0.05964047375698912, + "flos": 14902498045440.0, + "grad_norm": 1.9092494001723244, + "language_loss": 0.90719241, + "learning_rate": 3.9908267917015765e-06, + "loss": 0.92425841, + "num_input_tokens_seen": 10691235, + "step": 496, + "time_per_iteration": 2.746382713317871 + }, + { + "auxiliary_loss_clip": 0.01241912, + "auxiliary_loss_mlp": 0.01056282, + "balance_loss_clip": 1.03502572, + "balance_loss_mlp": 1.04203653, + "epoch": 0.059760716647628206, + "flos": 23185581206400.0, + "grad_norm": 2.095602733224378, + "language_loss": 0.92801881, + "learning_rate": 3.990752118833821e-06, + "loss": 0.95100069, + "num_input_tokens_seen": 10708675, + "step": 497, + "time_per_iteration": 2.681750774383545 + }, + { + "auxiliary_loss_clip": 0.01277685, + "auxiliary_loss_mlp": 0.01047876, + "balance_loss_clip": 1.07905543, + "balance_loss_mlp": 1.03242707, + "epoch": 0.0598809595382673, + "flos": 22746968231040.0, + "grad_norm": 1.753404844066256, + "language_loss": 0.78019476, + "learning_rate": 3.990677143970553e-06, + "loss": 0.80345041, + "num_input_tokens_seen": 10729485, + "step": 498, + "time_per_iteration": 2.6716909408569336 + }, + { + "auxiliary_loss_clip": 0.01198192, + "auxiliary_loss_mlp": 0.01046588, + "balance_loss_clip": 0.96273017, + "balance_loss_mlp": 1.03117394, + "epoch": 0.06000120242890639, + "flos": 22127221946880.0, + "grad_norm": 2.2043504020600433, + "language_loss": 0.81495798, + "learning_rate": 3.990601867123144e-06, + "loss": 0.83740574, + "num_input_tokens_seen": 10749210, + "step": 499, + "time_per_iteration": 2.748152732849121 + }, + { + "auxiliary_loss_clip": 0.01174674, + "auxiliary_loss_mlp": 0.01060074, + "balance_loss_clip": 0.9217422, + "balance_loss_mlp": 1.04541755, + "epoch": 0.06012144531954548, + "flos": 19171773878400.0, + "grad_norm": 2.3489875861152427, + "language_loss": 0.84771812, + "learning_rate": 3.990526288303014e-06, + "loss": 0.87006557, + "num_input_tokens_seen": 10768000, + "step": 500, + "time_per_iteration": 2.758864164352417 + }, + { + "auxiliary_loss_clip": 0.01222672, + "auxiliary_loss_mlp": 0.00514681, + "balance_loss_clip": 0.99962914, + "balance_loss_mlp": 0.0, + "epoch": 0.06024168821018457, + "flos": 22783345729920.0, + "grad_norm": 1.7416226100272012, + "language_loss": 0.90938979, + "learning_rate": 3.9904504075216295e-06, + "loss": 0.9267633, + "num_input_tokens_seen": 10788760, + "step": 501, + "time_per_iteration": 2.7079949378967285 + }, + { + "auxiliary_loss_clip": 0.01203989, + "auxiliary_loss_mlp": 0.0104996, + "balance_loss_clip": 0.95733213, + "balance_loss_mlp": 1.03462994, + "epoch": 0.06036193110082366, + "flos": 18770687637120.0, + "grad_norm": 2.202177829274332, + "language_loss": 0.93935829, + "learning_rate": 3.990374224790501e-06, + "loss": 0.96189773, + "num_input_tokens_seen": 10806965, + "step": 502, + "time_per_iteration": 2.782529592514038 + }, + { + "auxiliary_loss_clip": 0.01223084, + "auxiliary_loss_mlp": 0.01051161, + "balance_loss_clip": 0.99982357, + "balance_loss_mlp": 1.03593802, + "epoch": 0.06048217399146275, + "flos": 17201570935680.0, + "grad_norm": 2.739322091897315, + "language_loss": 0.71027666, + "learning_rate": 3.990297740121185e-06, + "loss": 0.73301911, + "num_input_tokens_seen": 10824900, + "step": 503, + "time_per_iteration": 2.648216962814331 + }, + { + "auxiliary_loss_clip": 0.01249273, + "auxiliary_loss_mlp": 0.00515894, + "balance_loss_clip": 1.03580117, + "balance_loss_mlp": 0.0, + "epoch": 0.06060241688210185, + "flos": 24024131187840.0, + "grad_norm": 1.668164408397752, + "language_loss": 0.77802849, + "learning_rate": 3.990220953525284e-06, + "loss": 0.79568011, + "num_input_tokens_seen": 10842010, + "step": 504, + "time_per_iteration": 2.708341121673584 + }, + { + "auxiliary_loss_clip": 0.01210684, + "auxiliary_loss_mlp": 0.01048472, + "balance_loss_clip": 0.99182713, + "balance_loss_mlp": 1.03322482, + "epoch": 0.06072265977274094, + "flos": 14611190745600.0, + "grad_norm": 2.4786663008710454, + "language_loss": 0.7410956, + "learning_rate": 3.9901438650144465e-06, + "loss": 0.76368707, + "num_input_tokens_seen": 10858260, + "step": 505, + "time_per_iteration": 2.658539295196533 + }, + { + "auxiliary_loss_clip": 0.01239292, + "auxiliary_loss_mlp": 0.01052044, + "balance_loss_clip": 1.03408241, + "balance_loss_mlp": 1.03709471, + "epoch": 0.06084290266338003, + "flos": 20558284813440.0, + "grad_norm": 2.8293363738436503, + "language_loss": 0.9182508, + "learning_rate": 3.990066474600367e-06, + "loss": 0.94116414, + "num_input_tokens_seen": 10876230, + "step": 506, + "time_per_iteration": 2.7502763271331787 + }, + { + "auxiliary_loss_clip": 0.01233986, + "auxiliary_loss_mlp": 0.01043081, + "balance_loss_clip": 1.0311619, + "balance_loss_mlp": 1.02931845, + "epoch": 0.06096314555401912, + "flos": 22309217182080.0, + "grad_norm": 1.872505376671831, + "language_loss": 0.68209338, + "learning_rate": 3.989988782294786e-06, + "loss": 0.70486408, + "num_input_tokens_seen": 10896320, + "step": 507, + "time_per_iteration": 2.730829954147339 + }, + { + "auxiliary_loss_clip": 0.01178777, + "auxiliary_loss_mlp": 0.01055248, + "balance_loss_clip": 0.95569837, + "balance_loss_mlp": 1.04061556, + "epoch": 0.06108338844465821, + "flos": 19131374056320.0, + "grad_norm": 1.7959234385699863, + "language_loss": 0.94999582, + "learning_rate": 3.989910788109489e-06, + "loss": 0.97233611, + "num_input_tokens_seen": 10912970, + "step": 508, + "time_per_iteration": 2.7686831951141357 + }, + { + "auxiliary_loss_clip": 0.01202827, + "auxiliary_loss_mlp": 0.01049982, + "balance_loss_clip": 0.96012795, + "balance_loss_mlp": 1.0349257, + "epoch": 0.0612036313352973, + "flos": 33584018169600.0, + "grad_norm": 2.6008634756221807, + "language_loss": 0.74825686, + "learning_rate": 3.989832492056307e-06, + "loss": 0.77078497, + "num_input_tokens_seen": 10933995, + "step": 509, + "time_per_iteration": 3.008579969406128 + }, + { + "auxiliary_loss_clip": 0.0125567, + "auxiliary_loss_mlp": 0.01049809, + "balance_loss_clip": 1.04061747, + "balance_loss_mlp": 1.0333941, + "epoch": 0.06132387422593639, + "flos": 27490552179840.0, + "grad_norm": 2.1972842081120203, + "language_loss": 0.81083846, + "learning_rate": 3.989753894147119e-06, + "loss": 0.83389324, + "num_input_tokens_seen": 10954120, + "step": 510, + "time_per_iteration": 2.7990944385528564 + }, + { + "auxiliary_loss_clip": 0.01248184, + "auxiliary_loss_mlp": 0.01053954, + "balance_loss_clip": 1.04050553, + "balance_loss_mlp": 1.0400126, + "epoch": 0.061444117116575485, + "flos": 25885057979520.0, + "grad_norm": 1.7152841688772162, + "language_loss": 0.80079412, + "learning_rate": 3.989674994393846e-06, + "loss": 0.82381558, + "num_input_tokens_seen": 10973595, + "step": 511, + "time_per_iteration": 2.731245994567871 + }, + { + "auxiliary_loss_clip": 0.01250168, + "auxiliary_loss_mlp": 0.01052902, + "balance_loss_clip": 1.03920615, + "balance_loss_mlp": 1.03734565, + "epoch": 0.061564360007214575, + "flos": 28512031150080.0, + "grad_norm": 2.136353911188, + "language_loss": 0.9396472, + "learning_rate": 3.98959579280846e-06, + "loss": 0.9626779, + "num_input_tokens_seen": 10991995, + "step": 512, + "time_per_iteration": 2.7929434776306152 + }, + { + "auxiliary_loss_clip": 0.01160506, + "auxiliary_loss_mlp": 0.01048611, + "balance_loss_clip": 0.92106169, + "balance_loss_mlp": 1.03399587, + "epoch": 0.061684602897853665, + "flos": 12094355652480.0, + "grad_norm": 2.049180209802674, + "language_loss": 0.82972437, + "learning_rate": 3.989516289402973e-06, + "loss": 0.85181552, + "num_input_tokens_seen": 11007625, + "step": 513, + "time_per_iteration": 3.6544265747070312 + }, + { + "auxiliary_loss_clip": 0.01126395, + "auxiliary_loss_mlp": 0.01053177, + "balance_loss_clip": 0.87196457, + "balance_loss_mlp": 1.03912187, + "epoch": 0.061804845788492754, + "flos": 19532639865600.0, + "grad_norm": 15.374545272238752, + "language_loss": 0.80435055, + "learning_rate": 3.989436484189447e-06, + "loss": 0.82614625, + "num_input_tokens_seen": 11025570, + "step": 514, + "time_per_iteration": 3.700928211212158 + }, + { + "auxiliary_loss_clip": 0.01250606, + "auxiliary_loss_mlp": 0.01049796, + "balance_loss_clip": 1.03464174, + "balance_loss_mlp": 1.03414357, + "epoch": 0.061925088679131844, + "flos": 15341111020800.0, + "grad_norm": 3.1332957338420404, + "language_loss": 0.8066898, + "learning_rate": 3.9893563771799885e-06, + "loss": 0.82969385, + "num_input_tokens_seen": 11042045, + "step": 515, + "time_per_iteration": 3.6049580574035645 + }, + { + "auxiliary_loss_clip": 0.01277287, + "auxiliary_loss_mlp": 0.01051943, + "balance_loss_clip": 1.07688892, + "balance_loss_mlp": 1.03666008, + "epoch": 0.062045331569770934, + "flos": 25919927107200.0, + "grad_norm": 2.222502550772337, + "language_loss": 0.8616122, + "learning_rate": 3.989275968386749e-06, + "loss": 0.8849045, + "num_input_tokens_seen": 11059955, + "step": 516, + "time_per_iteration": 3.6255600452423096 + }, + { + "auxiliary_loss_clip": 0.01215131, + "auxiliary_loss_mlp": 0.01044434, + "balance_loss_clip": 0.99408221, + "balance_loss_mlp": 1.02998543, + "epoch": 0.06216557446041003, + "flos": 28110621686400.0, + "grad_norm": 1.9741509965393775, + "language_loss": 0.77066344, + "learning_rate": 3.989195257821926e-06, + "loss": 0.79325908, + "num_input_tokens_seen": 11078440, + "step": 517, + "time_per_iteration": 2.8348445892333984 + }, + { + "auxiliary_loss_clip": 0.01223461, + "auxiliary_loss_mlp": 0.01053659, + "balance_loss_clip": 1.00056732, + "balance_loss_mlp": 1.03857899, + "epoch": 0.06228581735104912, + "flos": 23478181395840.0, + "grad_norm": 2.5399235211165947, + "language_loss": 0.84077299, + "learning_rate": 3.989114245497765e-06, + "loss": 0.86354417, + "num_input_tokens_seen": 11098240, + "step": 518, + "time_per_iteration": 2.746019124984741 + }, + { + "auxiliary_loss_clip": 0.01248187, + "auxiliary_loss_mlp": 0.01041293, + "balance_loss_clip": 1.03208506, + "balance_loss_mlp": 1.0273757, + "epoch": 0.06240606024168821, + "flos": 15195205975680.0, + "grad_norm": 2.4096859756741535, + "language_loss": 0.94933963, + "learning_rate": 3.989032931426554e-06, + "loss": 0.97223449, + "num_input_tokens_seen": 11115395, + "step": 519, + "time_per_iteration": 2.709099292755127 + }, + { + "auxiliary_loss_clip": 0.01222355, + "auxiliary_loss_mlp": 0.01046903, + "balance_loss_clip": 0.99792027, + "balance_loss_mlp": 1.03353953, + "epoch": 0.06252630313232731, + "flos": 20631829910400.0, + "grad_norm": 1.9935162425621653, + "language_loss": 0.86523807, + "learning_rate": 3.9889513156206295e-06, + "loss": 0.88793063, + "num_input_tokens_seen": 11134835, + "step": 520, + "time_per_iteration": 2.7988462448120117 + }, + { + "auxiliary_loss_clip": 0.01196585, + "auxiliary_loss_mlp": 0.01055042, + "balance_loss_clip": 0.95699602, + "balance_loss_mlp": 1.04019499, + "epoch": 0.06264654602296639, + "flos": 20778058177920.0, + "grad_norm": 3.1189883479259928, + "language_loss": 0.73954415, + "learning_rate": 3.988869398092371e-06, + "loss": 0.7620604, + "num_input_tokens_seen": 11154745, + "step": 521, + "time_per_iteration": 2.8331234455108643 + }, + { + "auxiliary_loss_clip": 0.01221284, + "auxiliary_loss_mlp": 0.01054733, + "balance_loss_clip": 0.99732745, + "balance_loss_mlp": 1.03893805, + "epoch": 0.06276678891360549, + "flos": 29605798241280.0, + "grad_norm": 2.3260479503390634, + "language_loss": 0.79062808, + "learning_rate": 3.988787178854206e-06, + "loss": 0.81338823, + "num_input_tokens_seen": 11174280, + "step": 522, + "time_per_iteration": 2.7708311080932617 + }, + { + "auxiliary_loss_clip": 0.01274788, + "auxiliary_loss_mlp": 0.01053282, + "balance_loss_clip": 1.07642293, + "balance_loss_mlp": 1.03855968, + "epoch": 0.06288703180424457, + "flos": 22126288193280.0, + "grad_norm": 1.9847834183089696, + "language_loss": 0.87185705, + "learning_rate": 3.988704657918608e-06, + "loss": 0.89513767, + "num_input_tokens_seen": 11193340, + "step": 523, + "time_per_iteration": 2.6853275299072266 + }, + { + "auxiliary_loss_clip": 0.01250111, + "auxiliary_loss_mlp": 0.01046502, + "balance_loss_clip": 1.03876674, + "balance_loss_mlp": 1.03243494, + "epoch": 0.06300727469488367, + "flos": 14976689587200.0, + "grad_norm": 2.3154784138676727, + "language_loss": 0.79627287, + "learning_rate": 3.988621835298094e-06, + "loss": 0.81923902, + "num_input_tokens_seen": 11210555, + "step": 524, + "time_per_iteration": 2.650651693344116 + }, + { + "auxiliary_loss_clip": 0.01267878, + "auxiliary_loss_mlp": 0.01046445, + "balance_loss_clip": 1.07566237, + "balance_loss_mlp": 1.03267658, + "epoch": 0.06312751758552275, + "flos": 24535391420160.0, + "grad_norm": 2.0628481749739738, + "language_loss": 0.91792536, + "learning_rate": 3.988538711005229e-06, + "loss": 0.94106853, + "num_input_tokens_seen": 11230010, + "step": 525, + "time_per_iteration": 2.679626703262329 + }, + { + "auxiliary_loss_clip": 0.01247038, + "auxiliary_loss_mlp": 0.01045941, + "balance_loss_clip": 1.03979063, + "balance_loss_mlp": 1.03213608, + "epoch": 0.06324776047616185, + "flos": 21507008785920.0, + "grad_norm": 2.3244887736962796, + "language_loss": 0.88278973, + "learning_rate": 3.988455285052622e-06, + "loss": 0.90571952, + "num_input_tokens_seen": 11246190, + "step": 526, + "time_per_iteration": 2.701570510864258 + }, + { + "auxiliary_loss_clip": 0.01237797, + "auxiliary_loss_mlp": 0.0104689, + "balance_loss_clip": 1.03331709, + "balance_loss_mlp": 1.03121424, + "epoch": 0.06336800336680094, + "flos": 21688034353920.0, + "grad_norm": 2.0220157077252634, + "language_loss": 0.83830464, + "learning_rate": 3.98837155745293e-06, + "loss": 0.86115146, + "num_input_tokens_seen": 11264230, + "step": 527, + "time_per_iteration": 2.662527084350586 + }, + { + "auxiliary_loss_clip": 0.01246941, + "auxiliary_loss_mlp": 0.01047013, + "balance_loss_clip": 1.03870964, + "balance_loss_mlp": 1.03241026, + "epoch": 0.06348824625744003, + "flos": 19500895221120.0, + "grad_norm": 2.9057337060165875, + "language_loss": 0.76321846, + "learning_rate": 3.988287528218854e-06, + "loss": 0.78615797, + "num_input_tokens_seen": 11283015, + "step": 528, + "time_per_iteration": 2.711379051208496 + }, + { + "auxiliary_loss_clip": 0.01245988, + "auxiliary_loss_mlp": 0.01042669, + "balance_loss_clip": 1.03865194, + "balance_loss_mlp": 1.02938926, + "epoch": 0.06360848914807912, + "flos": 15481233976320.0, + "grad_norm": 1.9657681239339908, + "language_loss": 0.90421748, + "learning_rate": 3.98820319736314e-06, + "loss": 0.92710412, + "num_input_tokens_seen": 11299630, + "step": 529, + "time_per_iteration": 2.6128149032592773 + }, + { + "auxiliary_loss_clip": 0.01196588, + "auxiliary_loss_mlp": 0.01044519, + "balance_loss_clip": 0.95670795, + "balance_loss_mlp": 1.03150153, + "epoch": 0.0637287320387182, + "flos": 20593369422720.0, + "grad_norm": 1.733519366676995, + "language_loss": 0.85553098, + "learning_rate": 3.988118564898582e-06, + "loss": 0.87794197, + "num_input_tokens_seen": 11319170, + "step": 530, + "time_per_iteration": 2.7571449279785156 + }, + { + "auxiliary_loss_clip": 0.01184288, + "auxiliary_loss_mlp": 0.00516137, + "balance_loss_clip": 0.95639461, + "balance_loss_mlp": 0.0, + "epoch": 0.0638489749293573, + "flos": 17412222245760.0, + "grad_norm": 2.5415317940530366, + "language_loss": 0.89431167, + "learning_rate": 3.988033630838019e-06, + "loss": 0.91131592, + "num_input_tokens_seen": 11333210, + "step": 531, + "time_per_iteration": 2.674285650253296 + }, + { + "auxiliary_loss_clip": 0.01247175, + "auxiliary_loss_mlp": 0.0105487, + "balance_loss_clip": 1.03741026, + "balance_loss_mlp": 1.04073191, + "epoch": 0.0639692178199964, + "flos": 23807661874560.0, + "grad_norm": 1.7521241906014269, + "language_loss": 0.88186169, + "learning_rate": 3.987948395194334e-06, + "loss": 0.90488213, + "num_input_tokens_seen": 11355590, + "step": 532, + "time_per_iteration": 2.7608957290649414 + }, + { + "auxiliary_loss_clip": 0.01234844, + "auxiliary_loss_mlp": 0.01053521, + "balance_loss_clip": 1.03280294, + "balance_loss_mlp": 1.03945446, + "epoch": 0.06408946071063548, + "flos": 18477225521280.0, + "grad_norm": 2.014529115387943, + "language_loss": 0.77083921, + "learning_rate": 3.987862857980458e-06, + "loss": 0.79372287, + "num_input_tokens_seen": 11371535, + "step": 533, + "time_per_iteration": 2.6492903232574463 + }, + { + "auxiliary_loss_clip": 0.01197171, + "auxiliary_loss_mlp": 0.01049723, + "balance_loss_clip": 0.95963818, + "balance_loss_mlp": 1.03357053, + "epoch": 0.06420970360127458, + "flos": 27162220936320.0, + "grad_norm": 2.0883863350996426, + "language_loss": 0.76925659, + "learning_rate": 3.987777019209368e-06, + "loss": 0.79172552, + "num_input_tokens_seen": 11392050, + "step": 534, + "time_per_iteration": 2.768406629562378 + }, + { + "auxiliary_loss_clip": 0.01268933, + "auxiliary_loss_mlp": 0.01052165, + "balance_loss_clip": 1.07586884, + "balance_loss_mlp": 1.03817022, + "epoch": 0.06432994649191366, + "flos": 23659673840640.0, + "grad_norm": 1.7935024746497688, + "language_loss": 0.8104021, + "learning_rate": 3.987690878894084e-06, + "loss": 0.83361304, + "num_input_tokens_seen": 11411765, + "step": 535, + "time_per_iteration": 2.656097650527954 + }, + { + "auxiliary_loss_clip": 0.01221309, + "auxiliary_loss_mlp": 0.01048214, + "balance_loss_clip": 0.99741703, + "balance_loss_mlp": 1.03330088, + "epoch": 0.06445018938255276, + "flos": 23403953940480.0, + "grad_norm": 2.301297834218594, + "language_loss": 0.85310954, + "learning_rate": 3.987604437047673e-06, + "loss": 0.87580478, + "num_input_tokens_seen": 11431565, + "step": 536, + "time_per_iteration": 2.70731520652771 + }, + { + "auxiliary_loss_clip": 0.01241144, + "auxiliary_loss_mlp": 0.01047675, + "balance_loss_clip": 1.03451622, + "balance_loss_mlp": 1.03394246, + "epoch": 0.06457043227319184, + "flos": 19646692525440.0, + "grad_norm": 2.701974128746862, + "language_loss": 0.77833575, + "learning_rate": 3.987517693683251e-06, + "loss": 0.80122393, + "num_input_tokens_seen": 11450140, + "step": 537, + "time_per_iteration": 2.640470027923584 + }, + { + "auxiliary_loss_clip": 0.01213826, + "auxiliary_loss_mlp": 0.01050863, + "balance_loss_clip": 0.99865997, + "balance_loss_mlp": 1.0363549, + "epoch": 0.06469067516383094, + "flos": 16978744915200.0, + "grad_norm": 2.6314307229880396, + "language_loss": 0.96374297, + "learning_rate": 3.9874306488139745e-06, + "loss": 0.98638988, + "num_input_tokens_seen": 11465400, + "step": 538, + "time_per_iteration": 2.6816415786743164 + }, + { + "auxiliary_loss_clip": 0.01194253, + "auxiliary_loss_mlp": 0.01054138, + "balance_loss_clip": 0.95938671, + "balance_loss_mlp": 1.03986263, + "epoch": 0.06481091805447003, + "flos": 23296401642240.0, + "grad_norm": 1.8303263931656244, + "language_loss": 0.88135231, + "learning_rate": 3.987343302453049e-06, + "loss": 0.90383625, + "num_input_tokens_seen": 11486675, + "step": 539, + "time_per_iteration": 3.7681097984313965 + }, + { + "auxiliary_loss_clip": 0.01219496, + "auxiliary_loss_mlp": 0.01059544, + "balance_loss_clip": 0.99853408, + "balance_loss_mlp": 1.0455606, + "epoch": 0.06493116094510912, + "flos": 29172356824320.0, + "grad_norm": 1.8523723991191685, + "language_loss": 0.82414186, + "learning_rate": 3.987255654613724e-06, + "loss": 0.84693223, + "num_input_tokens_seen": 11510440, + "step": 540, + "time_per_iteration": 3.660400390625 + }, + { + "auxiliary_loss_clip": 0.0118853, + "auxiliary_loss_mlp": 0.01051693, + "balance_loss_clip": 0.95666158, + "balance_loss_mlp": 1.03807914, + "epoch": 0.06505140383574821, + "flos": 19865065259520.0, + "grad_norm": 2.2417632328822648, + "language_loss": 0.70532161, + "learning_rate": 3.987167705309296e-06, + "loss": 0.72772384, + "num_input_tokens_seen": 11529715, + "step": 541, + "time_per_iteration": 3.6365060806274414 + }, + { + "auxiliary_loss_clip": 0.01245589, + "auxiliary_loss_mlp": 0.00514675, + "balance_loss_clip": 1.03609347, + "balance_loss_mlp": 0.0, + "epoch": 0.0651716467263873, + "flos": 17924703540480.0, + "grad_norm": 1.9477505105690778, + "language_loss": 0.95004338, + "learning_rate": 3.987079454553108e-06, + "loss": 0.967646, + "num_input_tokens_seen": 11547665, + "step": 542, + "time_per_iteration": 3.595898151397705 + }, + { + "auxiliary_loss_clip": 0.01191891, + "auxiliary_loss_mlp": 0.01055367, + "balance_loss_clip": 0.95975363, + "balance_loss_mlp": 1.04195595, + "epoch": 0.0652918896170264, + "flos": 20842840356480.0, + "grad_norm": 1.7542349343574914, + "language_loss": 0.91231132, + "learning_rate": 3.986990902358546e-06, + "loss": 0.93478388, + "num_input_tokens_seen": 11564605, + "step": 543, + "time_per_iteration": 2.677884817123413 + }, + { + "auxiliary_loss_clip": 0.01241443, + "auxiliary_loss_mlp": 0.01045925, + "balance_loss_clip": 1.03377008, + "balance_loss_mlp": 1.03085732, + "epoch": 0.06541213250766549, + "flos": 21872507627520.0, + "grad_norm": 2.2660200819539784, + "language_loss": 0.93148816, + "learning_rate": 3.986902048739045e-06, + "loss": 0.95436192, + "num_input_tokens_seen": 11584550, + "step": 544, + "time_per_iteration": 2.723410129547119 + }, + { + "auxiliary_loss_clip": 0.01222351, + "auxiliary_loss_mlp": 0.01052798, + "balance_loss_clip": 0.99972624, + "balance_loss_mlp": 1.03820658, + "epoch": 0.06553237539830457, + "flos": 23110743219840.0, + "grad_norm": 2.720983978431475, + "language_loss": 0.80415297, + "learning_rate": 3.986812893708082e-06, + "loss": 0.82690442, + "num_input_tokens_seen": 11600740, + "step": 545, + "time_per_iteration": 2.7148396968841553 + }, + { + "auxiliary_loss_clip": 0.01218381, + "auxiliary_loss_mlp": 0.0104755, + "balance_loss_clip": 0.99372184, + "balance_loss_mlp": 1.03262472, + "epoch": 0.06565261828894367, + "flos": 17923769786880.0, + "grad_norm": 2.0946211851246725, + "language_loss": 0.81288493, + "learning_rate": 3.9867234372791826e-06, + "loss": 0.83554423, + "num_input_tokens_seen": 11618695, + "step": 546, + "time_per_iteration": 2.61844539642334 + }, + { + "auxiliary_loss_clip": 0.01238022, + "auxiliary_loss_mlp": 0.01052794, + "balance_loss_clip": 1.03368998, + "balance_loss_mlp": 1.03895402, + "epoch": 0.06577286117958275, + "flos": 22783058421120.0, + "grad_norm": 1.5340173854030446, + "language_loss": 0.8725282, + "learning_rate": 3.986633679465918e-06, + "loss": 0.89543629, + "num_input_tokens_seen": 11638850, + "step": 547, + "time_per_iteration": 2.699645757675171 + }, + { + "auxiliary_loss_clip": 0.01171352, + "auxiliary_loss_mlp": 0.01053356, + "balance_loss_clip": 0.92119771, + "balance_loss_mlp": 1.03909242, + "epoch": 0.06589310407022185, + "flos": 23696194993920.0, + "grad_norm": 2.179143750284412, + "language_loss": 0.80494481, + "learning_rate": 3.986543620281904e-06, + "loss": 0.82719195, + "num_input_tokens_seen": 11658500, + "step": 548, + "time_per_iteration": 2.7684285640716553 + }, + { + "auxiliary_loss_clip": 0.01201097, + "auxiliary_loss_mlp": 0.01047559, + "balance_loss_clip": 0.99273217, + "balance_loss_mlp": 1.03416014, + "epoch": 0.06601334696086093, + "flos": 26864772410880.0, + "grad_norm": 1.7946885352019188, + "language_loss": 0.91089016, + "learning_rate": 3.986453259740802e-06, + "loss": 0.93337667, + "num_input_tokens_seen": 11676670, + "step": 549, + "time_per_iteration": 2.736440658569336 + }, + { + "auxiliary_loss_clip": 0.01219303, + "auxiliary_loss_mlp": 0.01059053, + "balance_loss_clip": 1.00108755, + "balance_loss_mlp": 1.04522467, + "epoch": 0.06613358985150003, + "flos": 12567694101120.0, + "grad_norm": 3.443778737613339, + "language_loss": 0.79032183, + "learning_rate": 3.986362597856319e-06, + "loss": 0.81310534, + "num_input_tokens_seen": 11693170, + "step": 550, + "time_per_iteration": 2.7004454135894775 + }, + { + "auxiliary_loss_clip": 0.01208512, + "auxiliary_loss_mlp": 0.00515746, + "balance_loss_clip": 0.9935776, + "balance_loss_mlp": 0.0, + "epoch": 0.06625383274213913, + "flos": 18332505624960.0, + "grad_norm": 2.530594431434774, + "language_loss": 0.81877804, + "learning_rate": 3.986271634642211e-06, + "loss": 0.83602071, + "num_input_tokens_seen": 11710150, + "step": 551, + "time_per_iteration": 2.7108216285705566 + }, + { + "auxiliary_loss_clip": 0.01269129, + "auxiliary_loss_mlp": 0.01056172, + "balance_loss_clip": 1.07781911, + "balance_loss_mlp": 1.04313612, + "epoch": 0.06637407563277821, + "flos": 15375585098880.0, + "grad_norm": 2.414473700419527, + "language_loss": 0.81569469, + "learning_rate": 3.986180370112274e-06, + "loss": 0.83894765, + "num_input_tokens_seen": 11726670, + "step": 552, + "time_per_iteration": 2.5649874210357666 + }, + { + "auxiliary_loss_clip": 0.01240386, + "auxiliary_loss_mlp": 0.00515297, + "balance_loss_clip": 1.03526878, + "balance_loss_mlp": 0.0, + "epoch": 0.0664943185234173, + "flos": 24025244509440.0, + "grad_norm": 1.7390823423499497, + "language_loss": 0.74760556, + "learning_rate": 3.986088804280354e-06, + "loss": 0.76516235, + "num_input_tokens_seen": 11746400, + "step": 553, + "time_per_iteration": 2.7497506141662598 + }, + { + "auxiliary_loss_clip": 0.01219475, + "auxiliary_loss_mlp": 0.01041614, + "balance_loss_clip": 0.99627793, + "balance_loss_mlp": 1.02775049, + "epoch": 0.06661456141405639, + "flos": 20957503547520.0, + "grad_norm": 2.892589139769797, + "language_loss": 0.93785787, + "learning_rate": 3.985996937160342e-06, + "loss": 0.96046877, + "num_input_tokens_seen": 11765590, + "step": 554, + "time_per_iteration": 2.7071919441223145 + }, + { + "auxiliary_loss_clip": 0.01241968, + "auxiliary_loss_mlp": 0.01056754, + "balance_loss_clip": 1.04062033, + "balance_loss_mlp": 1.04262733, + "epoch": 0.06673480430469549, + "flos": 52223953322880.0, + "grad_norm": 2.5374945442178083, + "language_loss": 0.69034517, + "learning_rate": 3.985904768766173e-06, + "loss": 0.7133323, + "num_input_tokens_seen": 11788365, + "step": 555, + "time_per_iteration": 2.9270286560058594 + }, + { + "auxiliary_loss_clip": 0.01196686, + "auxiliary_loss_mlp": 0.01049468, + "balance_loss_clip": 0.96041262, + "balance_loss_mlp": 1.03497219, + "epoch": 0.06685504719533458, + "flos": 16217079995520.0, + "grad_norm": 2.346471112821191, + "language_loss": 0.76251292, + "learning_rate": 3.98581229911183e-06, + "loss": 0.78497446, + "num_input_tokens_seen": 11807285, + "step": 556, + "time_per_iteration": 2.8132011890411377 + }, + { + "auxiliary_loss_clip": 0.01242615, + "auxiliary_loss_mlp": 0.01042388, + "balance_loss_clip": 1.03213871, + "balance_loss_mlp": 1.02839899, + "epoch": 0.06697529008597367, + "flos": 22491535639680.0, + "grad_norm": 1.8167301145918788, + "language_loss": 0.92275143, + "learning_rate": 3.985719528211341e-06, + "loss": 0.94560146, + "num_input_tokens_seen": 11826655, + "step": 557, + "time_per_iteration": 2.7215182781219482 + }, + { + "auxiliary_loss_clip": 0.01145414, + "auxiliary_loss_mlp": 0.01019937, + "balance_loss_clip": 0.98475248, + "balance_loss_mlp": 1.01483476, + "epoch": 0.06709553297661276, + "flos": 62688216936960.0, + "grad_norm": 0.8440444331186372, + "language_loss": 0.62998986, + "learning_rate": 3.985626456078777e-06, + "loss": 0.6516434, + "num_input_tokens_seen": 11891310, + "step": 558, + "time_per_iteration": 3.386716604232788 + }, + { + "auxiliary_loss_clip": 0.01189009, + "auxiliary_loss_mlp": 0.01058032, + "balance_loss_clip": 0.95612127, + "balance_loss_mlp": 1.04440045, + "epoch": 0.06721577586725185, + "flos": 11216590997760.0, + "grad_norm": 2.2827708911274245, + "language_loss": 0.86081374, + "learning_rate": 3.985533082728259e-06, + "loss": 0.88328409, + "num_input_tokens_seen": 11906965, + "step": 559, + "time_per_iteration": 2.7106690406799316 + }, + { + "auxiliary_loss_clip": 0.01269043, + "auxiliary_loss_mlp": 0.01053223, + "balance_loss_clip": 1.07470703, + "balance_loss_mlp": 1.03936458, + "epoch": 0.06733601875789094, + "flos": 25922189664000.0, + "grad_norm": 1.8919209846231169, + "language_loss": 0.74574423, + "learning_rate": 3.985439408173951e-06, + "loss": 0.76896691, + "num_input_tokens_seen": 11927190, + "step": 560, + "time_per_iteration": 2.677762508392334 + }, + { + "auxiliary_loss_clip": 0.01269721, + "auxiliary_loss_mlp": 0.01039691, + "balance_loss_clip": 1.07678401, + "balance_loss_mlp": 1.02611876, + "epoch": 0.06745626164853002, + "flos": 20813645577600.0, + "grad_norm": 2.0443941602990203, + "language_loss": 0.70584768, + "learning_rate": 3.9853454324300634e-06, + "loss": 0.72894174, + "num_input_tokens_seen": 11946400, + "step": 561, + "time_per_iteration": 2.642897605895996 + }, + { + "auxiliary_loss_clip": 0.011483, + "auxiliary_loss_mlp": 0.01048806, + "balance_loss_clip": 0.88085961, + "balance_loss_mlp": 1.034477, + "epoch": 0.06757650453916912, + "flos": 19829262378240.0, + "grad_norm": 2.234352888743335, + "language_loss": 0.77886057, + "learning_rate": 3.985251155510852e-06, + "loss": 0.80083168, + "num_input_tokens_seen": 11965430, + "step": 562, + "time_per_iteration": 2.8143928050994873 + }, + { + "auxiliary_loss_clip": 0.01172029, + "auxiliary_loss_mlp": 0.01047919, + "balance_loss_clip": 0.92584926, + "balance_loss_mlp": 1.03410888, + "epoch": 0.06769674742980822, + "flos": 25739224761600.0, + "grad_norm": 1.781325185141186, + "language_loss": 0.80370474, + "learning_rate": 3.98515657743062e-06, + "loss": 0.82590425, + "num_input_tokens_seen": 11984895, + "step": 563, + "time_per_iteration": 2.80446457862854 + }, + { + "auxiliary_loss_clip": 0.01216422, + "auxiliary_loss_mlp": 0.01049032, + "balance_loss_clip": 0.99493194, + "balance_loss_mlp": 1.03545964, + "epoch": 0.0678169903204473, + "flos": 13074788355840.0, + "grad_norm": 1.848382773525276, + "language_loss": 0.77629882, + "learning_rate": 3.985061698203711e-06, + "loss": 0.79895335, + "num_input_tokens_seen": 12002010, + "step": 564, + "time_per_iteration": 2.705705404281616 + }, + { + "auxiliary_loss_clip": 0.01180938, + "auxiliary_loss_mlp": 0.01004259, + "balance_loss_clip": 1.05843544, + "balance_loss_mlp": 0.99915695, + "epoch": 0.0679372332110864, + "flos": 70865830788480.0, + "grad_norm": 0.881793969419887, + "language_loss": 0.63816959, + "learning_rate": 3.984966517844523e-06, + "loss": 0.6600216, + "num_input_tokens_seen": 12057255, + "step": 565, + "time_per_iteration": 3.206646680831909 + }, + { + "auxiliary_loss_clip": 0.01264856, + "auxiliary_loss_mlp": 0.01049081, + "balance_loss_clip": 1.07181025, + "balance_loss_mlp": 1.03461552, + "epoch": 0.06805747610172548, + "flos": 28256418990720.0, + "grad_norm": 2.447245859282468, + "language_loss": 0.80598092, + "learning_rate": 3.984871036367492e-06, + "loss": 0.82912034, + "num_input_tokens_seen": 12077280, + "step": 566, + "time_per_iteration": 4.529407978057861 + }, + { + "auxiliary_loss_clip": 0.01239086, + "auxiliary_loss_mlp": 0.00514785, + "balance_loss_clip": 1.03725255, + "balance_loss_mlp": 0.0, + "epoch": 0.06817771899236458, + "flos": 20120533764480.0, + "grad_norm": 1.774838283790469, + "language_loss": 0.82845539, + "learning_rate": 3.984775253787102e-06, + "loss": 0.84599411, + "num_input_tokens_seen": 12095570, + "step": 567, + "time_per_iteration": 3.7178361415863037 + }, + { + "auxiliary_loss_clip": 0.01244116, + "auxiliary_loss_mlp": 0.01045024, + "balance_loss_clip": 1.03348303, + "balance_loss_mlp": 1.0309689, + "epoch": 0.06829796188300366, + "flos": 17930629284480.0, + "grad_norm": 2.5026739458213254, + "language_loss": 0.87509936, + "learning_rate": 3.984679170117885e-06, + "loss": 0.89799076, + "num_input_tokens_seen": 12111775, + "step": 568, + "time_per_iteration": 3.7924821376800537 + }, + { + "auxiliary_loss_clip": 0.01235853, + "auxiliary_loss_mlp": 0.01043238, + "balance_loss_clip": 1.03257632, + "balance_loss_mlp": 1.02977931, + "epoch": 0.06841820477364276, + "flos": 14501627285760.0, + "grad_norm": 2.8919042350674515, + "language_loss": 0.78569496, + "learning_rate": 3.984582785374415e-06, + "loss": 0.80848593, + "num_input_tokens_seen": 12129215, + "step": 569, + "time_per_iteration": 2.824272394180298 + }, + { + "auxiliary_loss_clip": 0.01212387, + "auxiliary_loss_mlp": 0.00514673, + "balance_loss_clip": 0.99514389, + "balance_loss_mlp": 0.0, + "epoch": 0.06853844766428185, + "flos": 21938474954880.0, + "grad_norm": 1.9125114680100497, + "language_loss": 0.80797911, + "learning_rate": 3.9844860995713155e-06, + "loss": 0.82524967, + "num_input_tokens_seen": 12148755, + "step": 570, + "time_per_iteration": 2.728111982345581 + }, + { + "auxiliary_loss_clip": 0.0124133, + "auxiliary_loss_mlp": 0.0104853, + "balance_loss_clip": 1.04031646, + "balance_loss_mlp": 1.03479123, + "epoch": 0.06865869055492094, + "flos": 16800628348800.0, + "grad_norm": 2.2114357008522334, + "language_loss": 0.828655, + "learning_rate": 3.9843891127232524e-06, + "loss": 0.85155368, + "num_input_tokens_seen": 12166290, + "step": 571, + "time_per_iteration": 2.6985180377960205 + }, + { + "auxiliary_loss_clip": 0.01153437, + "auxiliary_loss_mlp": 0.01044244, + "balance_loss_clip": 0.91606331, + "balance_loss_mlp": 1.03159571, + "epoch": 0.06877893344556003, + "flos": 19937281553280.0, + "grad_norm": 5.428615300815814, + "language_loss": 0.66942, + "learning_rate": 3.984291824844938e-06, + "loss": 0.69139677, + "num_input_tokens_seen": 12181385, + "step": 572, + "time_per_iteration": 2.7443392276763916 + }, + { + "auxiliary_loss_clip": 0.01264187, + "auxiliary_loss_mlp": 0.01053583, + "balance_loss_clip": 1.07291901, + "balance_loss_mlp": 1.03887308, + "epoch": 0.06889917633619912, + "flos": 23039388852480.0, + "grad_norm": 2.246835070317811, + "language_loss": 0.84742486, + "learning_rate": 3.984194235951132e-06, + "loss": 0.87060249, + "num_input_tokens_seen": 12197530, + "step": 573, + "time_per_iteration": 2.6859920024871826 + }, + { + "auxiliary_loss_clip": 0.01269963, + "auxiliary_loss_mlp": 0.01056098, + "balance_loss_clip": 1.07900167, + "balance_loss_mlp": 1.04284179, + "epoch": 0.06901941922683821, + "flos": 20960556203520.0, + "grad_norm": 3.2897404645785104, + "language_loss": 0.84836006, + "learning_rate": 3.9840963460566375e-06, + "loss": 0.87162066, + "num_input_tokens_seen": 12216310, + "step": 574, + "time_per_iteration": 2.5930049419403076 + }, + { + "auxiliary_loss_clip": 0.01127146, + "auxiliary_loss_mlp": 0.01048608, + "balance_loss_clip": 0.87931967, + "balance_loss_mlp": 1.03429055, + "epoch": 0.06913966211747731, + "flos": 24821850384000.0, + "grad_norm": 1.6428096649894182, + "language_loss": 0.89213312, + "learning_rate": 3.983998155176305e-06, + "loss": 0.9138906, + "num_input_tokens_seen": 12236670, + "step": 575, + "time_per_iteration": 2.927769899368286 + }, + { + "auxiliary_loss_clip": 0.01172974, + "auxiliary_loss_mlp": 0.01019557, + "balance_loss_clip": 1.05335855, + "balance_loss_mlp": 1.01462173, + "epoch": 0.06925990500811639, + "flos": 58367446957440.0, + "grad_norm": 0.8210838595862684, + "language_loss": 0.57076097, + "learning_rate": 3.9838996633250305e-06, + "loss": 0.5926863, + "num_input_tokens_seen": 12297185, + "step": 576, + "time_per_iteration": 3.2793447971343994 + }, + { + "auxiliary_loss_clip": 0.01237695, + "auxiliary_loss_mlp": 0.01045557, + "balance_loss_clip": 1.03190756, + "balance_loss_mlp": 1.03318906, + "epoch": 0.06938014789875549, + "flos": 12749940731520.0, + "grad_norm": 2.0607877401976435, + "language_loss": 0.87937701, + "learning_rate": 3.983800870517753e-06, + "loss": 0.90220952, + "num_input_tokens_seen": 12313975, + "step": 577, + "time_per_iteration": 2.6529266834259033 + }, + { + "auxiliary_loss_clip": 0.01236652, + "auxiliary_loss_mlp": 0.01044872, + "balance_loss_clip": 1.03790236, + "balance_loss_mlp": 1.03230774, + "epoch": 0.06950039078939457, + "flos": 22820226019200.0, + "grad_norm": 4.093201908581122, + "language_loss": 0.77646887, + "learning_rate": 3.983701776769463e-06, + "loss": 0.7992841, + "num_input_tokens_seen": 12331385, + "step": 578, + "time_per_iteration": 2.7113680839538574 + }, + { + "auxiliary_loss_clip": 0.01229947, + "auxiliary_loss_mlp": 0.01054756, + "balance_loss_clip": 1.03395331, + "balance_loss_mlp": 1.04114223, + "epoch": 0.06962063368003367, + "flos": 21941348042880.0, + "grad_norm": 2.5933981380533133, + "language_loss": 0.85481238, + "learning_rate": 3.9836023820951885e-06, + "loss": 0.87765944, + "num_input_tokens_seen": 12350600, + "step": 579, + "time_per_iteration": 2.6409528255462646 + }, + { + "auxiliary_loss_clip": 0.01189064, + "auxiliary_loss_mlp": 0.01044779, + "balance_loss_clip": 0.95745021, + "balance_loss_mlp": 1.03176761, + "epoch": 0.06974087657067275, + "flos": 20706021452160.0, + "grad_norm": 2.4125952001945805, + "language_loss": 0.6874097, + "learning_rate": 3.983502686510011e-06, + "loss": 0.70974815, + "num_input_tokens_seen": 12371430, + "step": 580, + "time_per_iteration": 2.699791669845581 + }, + { + "auxiliary_loss_clip": 0.01238258, + "auxiliary_loss_mlp": 0.00515273, + "balance_loss_clip": 1.03151369, + "balance_loss_mlp": 0.0, + "epoch": 0.06986111946131185, + "flos": 22638230784000.0, + "grad_norm": 1.9437234108624657, + "language_loss": 0.73689806, + "learning_rate": 3.9834026900290525e-06, + "loss": 0.75443339, + "num_input_tokens_seen": 12390825, + "step": 581, + "time_per_iteration": 2.6461966037750244 + }, + { + "auxiliary_loss_clip": 0.01264324, + "auxiliary_loss_mlp": 0.0104161, + "balance_loss_clip": 1.07480955, + "balance_loss_mlp": 1.02825236, + "epoch": 0.06998136235195095, + "flos": 26943453152640.0, + "grad_norm": 2.056335487292161, + "language_loss": 1.0012939, + "learning_rate": 3.983302392667482e-06, + "loss": 1.02435327, + "num_input_tokens_seen": 12411670, + "step": 582, + "time_per_iteration": 2.7346925735473633 + }, + { + "auxiliary_loss_clip": 0.01240992, + "auxiliary_loss_mlp": 0.01046313, + "balance_loss_clip": 1.04023206, + "balance_loss_mlp": 1.03349221, + "epoch": 0.07010160524259003, + "flos": 22492505306880.0, + "grad_norm": 2.134754333253635, + "language_loss": 0.93723786, + "learning_rate": 3.983201794440517e-06, + "loss": 0.9601109, + "num_input_tokens_seen": 12431245, + "step": 583, + "time_per_iteration": 2.681957483291626 + }, + { + "auxiliary_loss_clip": 0.01201631, + "auxiliary_loss_mlp": 0.01056497, + "balance_loss_clip": 0.9922722, + "balance_loss_mlp": 1.04394412, + "epoch": 0.07022184813322913, + "flos": 18332541538560.0, + "grad_norm": 1.7651883881292536, + "language_loss": 0.67617977, + "learning_rate": 3.9831008953634165e-06, + "loss": 0.69876111, + "num_input_tokens_seen": 12450535, + "step": 584, + "time_per_iteration": 2.707390069961548 + }, + { + "auxiliary_loss_clip": 0.01144726, + "auxiliary_loss_mlp": 0.01047162, + "balance_loss_clip": 0.911807, + "balance_loss_mlp": 1.03361988, + "epoch": 0.07034209102386821, + "flos": 24675550289280.0, + "grad_norm": 4.710753679642965, + "language_loss": 0.81385469, + "learning_rate": 3.9829996954514864e-06, + "loss": 0.83577359, + "num_input_tokens_seen": 12469675, + "step": 585, + "time_per_iteration": 2.788525104522705 + }, + { + "auxiliary_loss_clip": 0.01224477, + "auxiliary_loss_mlp": 0.01052424, + "balance_loss_clip": 1.03193414, + "balance_loss_mlp": 1.0383451, + "epoch": 0.0704623339145073, + "flos": 25995878415360.0, + "grad_norm": 1.9688639019257457, + "language_loss": 0.84362465, + "learning_rate": 3.982898194720079e-06, + "loss": 0.86639363, + "num_input_tokens_seen": 12490405, + "step": 586, + "time_per_iteration": 2.825174570083618 + }, + { + "auxiliary_loss_clip": 0.01206008, + "auxiliary_loss_mlp": 0.00514631, + "balance_loss_clip": 0.99810153, + "balance_loss_mlp": 0.0, + "epoch": 0.0705825768051464, + "flos": 25338318088320.0, + "grad_norm": 2.5129303330867305, + "language_loss": 0.82440662, + "learning_rate": 3.982796393184592e-06, + "loss": 0.84161305, + "num_input_tokens_seen": 12509485, + "step": 587, + "time_per_iteration": 2.903078317642212 + }, + { + "auxiliary_loss_clip": 0.01139513, + "auxiliary_loss_mlp": 0.01006245, + "balance_loss_clip": 1.00632298, + "balance_loss_mlp": 1.00102413, + "epoch": 0.07070281969578548, + "flos": 66047552507520.0, + "grad_norm": 0.8191396693362611, + "language_loss": 0.6265707, + "learning_rate": 3.98269429086047e-06, + "loss": 0.64802837, + "num_input_tokens_seen": 12567325, + "step": 588, + "time_per_iteration": 3.233042001724243 + }, + { + "auxiliary_loss_clip": 0.0120061, + "auxiliary_loss_mlp": 0.01044294, + "balance_loss_clip": 0.99223244, + "balance_loss_mlp": 1.03053701, + "epoch": 0.07082306258642458, + "flos": 23653568528640.0, + "grad_norm": 2.8687968970242212, + "language_loss": 0.8662045, + "learning_rate": 3.982591887763199e-06, + "loss": 0.88865352, + "num_input_tokens_seen": 12584785, + "step": 589, + "time_per_iteration": 2.717587471008301 + }, + { + "auxiliary_loss_clip": 0.01163436, + "auxiliary_loss_mlp": 0.01047704, + "balance_loss_clip": 0.94534039, + "balance_loss_mlp": 1.0338583, + "epoch": 0.07094330547706366, + "flos": 13880049408000.0, + "grad_norm": 2.264217547294648, + "language_loss": 0.81810057, + "learning_rate": 3.982489183908316e-06, + "loss": 0.84021193, + "num_input_tokens_seen": 12601205, + "step": 590, + "time_per_iteration": 2.6448564529418945 + }, + { + "auxiliary_loss_clip": 0.01107606, + "auxiliary_loss_mlp": 0.0104386, + "balance_loss_clip": 0.86535347, + "balance_loss_mlp": 1.03159893, + "epoch": 0.07106354836770276, + "flos": 24645098534400.0, + "grad_norm": 1.8155660260025734, + "language_loss": 0.84673238, + "learning_rate": 3.982386179311399e-06, + "loss": 0.86824703, + "num_input_tokens_seen": 12621725, + "step": 591, + "time_per_iteration": 2.790025234222412 + }, + { + "auxiliary_loss_clip": 0.01239361, + "auxiliary_loss_mlp": 0.01044519, + "balance_loss_clip": 1.03494489, + "balance_loss_mlp": 1.02987957, + "epoch": 0.07118379125834184, + "flos": 16217223649920.0, + "grad_norm": 2.1159037820733766, + "language_loss": 0.8763842, + "learning_rate": 3.982282873988075e-06, + "loss": 0.89922297, + "num_input_tokens_seen": 12639600, + "step": 592, + "time_per_iteration": 3.597487449645996 + }, + { + "auxiliary_loss_clip": 0.01213158, + "auxiliary_loss_mlp": 0.01048063, + "balance_loss_clip": 0.99599183, + "balance_loss_mlp": 1.03564191, + "epoch": 0.07130403414898094, + "flos": 19719986227200.0, + "grad_norm": 1.647509365374299, + "language_loss": 0.86890996, + "learning_rate": 3.982179267954016e-06, + "loss": 0.89152217, + "num_input_tokens_seen": 12660030, + "step": 593, + "time_per_iteration": 3.667961835861206 + }, + { + "auxiliary_loss_clip": 0.01258913, + "auxiliary_loss_mlp": 0.01049341, + "balance_loss_clip": 1.07053876, + "balance_loss_mlp": 1.0357933, + "epoch": 0.07142427703962004, + "flos": 21871933009920.0, + "grad_norm": 3.369866194680373, + "language_loss": 0.96197259, + "learning_rate": 3.982075361224937e-06, + "loss": 0.98505509, + "num_input_tokens_seen": 12678395, + "step": 594, + "time_per_iteration": 3.5186452865600586 + }, + { + "auxiliary_loss_clip": 0.01236111, + "auxiliary_loss_mlp": 0.00514733, + "balance_loss_clip": 1.0373733, + "balance_loss_mlp": 0.0, + "epoch": 0.07154451993025912, + "flos": 18296595002880.0, + "grad_norm": 1.9334867564088754, + "language_loss": 0.87875658, + "learning_rate": 3.981971153816602e-06, + "loss": 0.89626503, + "num_input_tokens_seen": 12696000, + "step": 595, + "time_per_iteration": 2.6642980575561523 + }, + { + "auxiliary_loss_clip": 0.01263839, + "auxiliary_loss_mlp": 0.01047504, + "balance_loss_clip": 1.07851672, + "balance_loss_mlp": 1.03502238, + "epoch": 0.07166476282089822, + "flos": 22160690444160.0, + "grad_norm": 4.416426081030657, + "language_loss": 0.96266985, + "learning_rate": 3.981866645744819e-06, + "loss": 0.98578328, + "num_input_tokens_seen": 12716715, + "step": 596, + "time_per_iteration": 2.6639339923858643 + }, + { + "auxiliary_loss_clip": 0.01261338, + "auxiliary_loss_mlp": 0.00515064, + "balance_loss_clip": 1.07156372, + "balance_loss_mlp": 0.0, + "epoch": 0.0717850057115373, + "flos": 14136343925760.0, + "grad_norm": 2.2549472623576245, + "language_loss": 0.81509882, + "learning_rate": 3.9817618370254416e-06, + "loss": 0.83286285, + "num_input_tokens_seen": 12733370, + "step": 597, + "time_per_iteration": 2.6569302082061768 + }, + { + "auxiliary_loss_clip": 0.01263771, + "auxiliary_loss_mlp": 0.01053828, + "balance_loss_clip": 1.07564521, + "balance_loss_mlp": 1.04111457, + "epoch": 0.0719052486021764, + "flos": 30917794412160.0, + "grad_norm": 2.035194186982866, + "language_loss": 0.87189209, + "learning_rate": 3.9816567276743684e-06, + "loss": 0.89506805, + "num_input_tokens_seen": 12753235, + "step": 598, + "time_per_iteration": 2.6562464237213135 + }, + { + "auxiliary_loss_clip": 0.01210031, + "auxiliary_loss_mlp": 0.01046938, + "balance_loss_clip": 0.99483258, + "balance_loss_mlp": 1.03372324, + "epoch": 0.0720254914928155, + "flos": 21287019939840.0, + "grad_norm": 1.8541560838221938, + "language_loss": 0.7756722, + "learning_rate": 3.9815513177075466e-06, + "loss": 0.79824185, + "num_input_tokens_seen": 12772020, + "step": 599, + "time_per_iteration": 2.7162206172943115 + }, + { + "auxiliary_loss_clip": 0.01227595, + "auxiliary_loss_mlp": 0.01049024, + "balance_loss_clip": 1.03509843, + "balance_loss_mlp": 1.03589272, + "epoch": 0.07214573438345458, + "flos": 27819170732160.0, + "grad_norm": 1.5763813459504132, + "language_loss": 0.70063984, + "learning_rate": 3.9814456071409646e-06, + "loss": 0.72340596, + "num_input_tokens_seen": 12792555, + "step": 600, + "time_per_iteration": 2.6768133640289307 + }, + { + "auxiliary_loss_clip": 0.01167894, + "auxiliary_loss_mlp": 0.01043919, + "balance_loss_clip": 0.91766334, + "balance_loss_mlp": 1.02961349, + "epoch": 0.07226597727409367, + "flos": 25483576688640.0, + "grad_norm": 3.1522785298123646, + "language_loss": 0.85372651, + "learning_rate": 3.981339595990659e-06, + "loss": 0.87584472, + "num_input_tokens_seen": 12811085, + "step": 601, + "time_per_iteration": 2.8598594665527344 + }, + { + "auxiliary_loss_clip": 0.01234946, + "auxiliary_loss_mlp": 0.01049941, + "balance_loss_clip": 1.03411591, + "balance_loss_mlp": 1.03590989, + "epoch": 0.07238622016473276, + "flos": 23513840622720.0, + "grad_norm": 2.335849623544678, + "language_loss": 0.81131393, + "learning_rate": 3.981233284272713e-06, + "loss": 0.83416283, + "num_input_tokens_seen": 12830830, + "step": 602, + "time_per_iteration": 2.6972291469573975 + }, + { + "auxiliary_loss_clip": 0.01188862, + "auxiliary_loss_mlp": 0.01051675, + "balance_loss_clip": 0.95764595, + "balance_loss_mlp": 1.03788197, + "epoch": 0.07250646305537185, + "flos": 25453519983360.0, + "grad_norm": 1.627371898462367, + "language_loss": 0.90163672, + "learning_rate": 3.981126672003253e-06, + "loss": 0.92404211, + "num_input_tokens_seen": 12853505, + "step": 603, + "time_per_iteration": 2.837456703186035 + }, + { + "auxiliary_loss_clip": 0.01211704, + "auxiliary_loss_mlp": 0.01047647, + "balance_loss_clip": 0.98975432, + "balance_loss_mlp": 1.03498697, + "epoch": 0.07262670594601094, + "flos": 27155038216320.0, + "grad_norm": 2.115812172808568, + "language_loss": 0.7785815, + "learning_rate": 3.981019759198451e-06, + "loss": 0.801175, + "num_input_tokens_seen": 12872455, + "step": 604, + "time_per_iteration": 2.8041574954986572 + }, + { + "auxiliary_loss_clip": 0.01213395, + "auxiliary_loss_mlp": 0.01038814, + "balance_loss_clip": 0.9939034, + "balance_loss_mlp": 1.02589214, + "epoch": 0.07274694883665003, + "flos": 26651607148800.0, + "grad_norm": 2.6439858428680765, + "language_loss": 0.84243333, + "learning_rate": 3.980912545874528e-06, + "loss": 0.86495543, + "num_input_tokens_seen": 12892620, + "step": 605, + "time_per_iteration": 2.754000663757324 + }, + { + "auxiliary_loss_clip": 0.01224578, + "auxiliary_loss_mlp": 0.0051533, + "balance_loss_clip": 1.02928233, + "balance_loss_mlp": 0.0, + "epoch": 0.07286719172728913, + "flos": 29862344154240.0, + "grad_norm": 1.939315722936131, + "language_loss": 0.85344458, + "learning_rate": 3.980805032047746e-06, + "loss": 0.87084365, + "num_input_tokens_seen": 12914090, + "step": 606, + "time_per_iteration": 2.868051767349243 + }, + { + "auxiliary_loss_clip": 0.01205615, + "auxiliary_loss_mlp": 0.01049924, + "balance_loss_clip": 0.99340183, + "balance_loss_mlp": 1.03559542, + "epoch": 0.07298743461792821, + "flos": 17382057799680.0, + "grad_norm": 2.3290457381366454, + "language_loss": 0.81075954, + "learning_rate": 3.980697217734415e-06, + "loss": 0.8333149, + "num_input_tokens_seen": 12931830, + "step": 607, + "time_per_iteration": 2.6880452632904053 + }, + { + "auxiliary_loss_clip": 0.01164741, + "auxiliary_loss_mlp": 0.00515071, + "balance_loss_clip": 0.9157989, + "balance_loss_mlp": 0.0, + "epoch": 0.07310767750856731, + "flos": 19498201701120.0, + "grad_norm": 2.199457123976701, + "language_loss": 0.91564775, + "learning_rate": 3.980589102950891e-06, + "loss": 0.93244588, + "num_input_tokens_seen": 12949995, + "step": 608, + "time_per_iteration": 2.7612147331237793 + }, + { + "auxiliary_loss_clip": 0.01208666, + "auxiliary_loss_mlp": 0.01043061, + "balance_loss_clip": 0.99788916, + "balance_loss_mlp": 1.03000796, + "epoch": 0.07322792039920639, + "flos": 29168693637120.0, + "grad_norm": 2.352456475187531, + "language_loss": 0.75965512, + "learning_rate": 3.9804806877135755e-06, + "loss": 0.78217238, + "num_input_tokens_seen": 12968040, + "step": 609, + "time_per_iteration": 2.712066411972046 + }, + { + "auxiliary_loss_clip": 0.01234764, + "auxiliary_loss_mlp": 0.00515187, + "balance_loss_clip": 1.02986836, + "balance_loss_mlp": 0.0, + "epoch": 0.07334816328984549, + "flos": 23477822259840.0, + "grad_norm": 2.3556460129981622, + "language_loss": 0.85879683, + "learning_rate": 3.980371972038915e-06, + "loss": 0.8762964, + "num_input_tokens_seen": 12988530, + "step": 610, + "time_per_iteration": 2.81306529045105 + }, + { + "auxiliary_loss_clip": 0.01261293, + "auxiliary_loss_mlp": 0.01054548, + "balance_loss_clip": 1.07595015, + "balance_loss_mlp": 1.04116726, + "epoch": 0.07346840618048459, + "flos": 22962467877120.0, + "grad_norm": 1.9434481477334187, + "language_loss": 0.84292722, + "learning_rate": 3.980262955943399e-06, + "loss": 0.86608559, + "num_input_tokens_seen": 13008195, + "step": 611, + "time_per_iteration": 2.671457290649414 + }, + { + "auxiliary_loss_clip": 0.01207365, + "auxiliary_loss_mlp": 0.01052171, + "balance_loss_clip": 0.99949253, + "balance_loss_mlp": 1.03891516, + "epoch": 0.07358864907112367, + "flos": 17673903803520.0, + "grad_norm": 2.6774473228516222, + "language_loss": 0.87073058, + "learning_rate": 3.980153639443569e-06, + "loss": 0.89332592, + "num_input_tokens_seen": 13024180, + "step": 612, + "time_per_iteration": 2.6630656719207764 + }, + { + "auxiliary_loss_clip": 0.01217511, + "auxiliary_loss_mlp": 0.01045217, + "balance_loss_clip": 0.99387562, + "balance_loss_mlp": 1.03194332, + "epoch": 0.07370889196176277, + "flos": 24097029840000.0, + "grad_norm": 2.0590822887466946, + "language_loss": 0.79965079, + "learning_rate": 3.980044022556005e-06, + "loss": 0.82227802, + "num_input_tokens_seen": 13043865, + "step": 613, + "time_per_iteration": 2.6807620525360107 + }, + { + "auxiliary_loss_clip": 0.01236934, + "auxiliary_loss_mlp": 0.01050087, + "balance_loss_clip": 1.035532, + "balance_loss_mlp": 1.03727162, + "epoch": 0.07382913485240185, + "flos": 25885919905920.0, + "grad_norm": 2.754137069069843, + "language_loss": 0.72870016, + "learning_rate": 3.9799341052973375e-06, + "loss": 0.7515704, + "num_input_tokens_seen": 13063700, + "step": 614, + "time_per_iteration": 2.8224382400512695 + }, + { + "auxiliary_loss_clip": 0.01206973, + "auxiliary_loss_mlp": 0.01049641, + "balance_loss_clip": 0.99565685, + "balance_loss_mlp": 1.03620005, + "epoch": 0.07394937774304094, + "flos": 16873850223360.0, + "grad_norm": 2.2043117175291034, + "language_loss": 0.75525129, + "learning_rate": 3.979823887684241e-06, + "loss": 0.77781743, + "num_input_tokens_seen": 13082640, + "step": 615, + "time_per_iteration": 2.7544593811035156 + }, + { + "auxiliary_loss_clip": 0.01258988, + "auxiliary_loss_mlp": 0.01048334, + "balance_loss_clip": 1.0725317, + "balance_loss_mlp": 1.0351969, + "epoch": 0.07406962063368003, + "flos": 20703471586560.0, + "grad_norm": 2.8614053092449665, + "language_loss": 0.84572279, + "learning_rate": 3.979713369733434e-06, + "loss": 0.86879599, + "num_input_tokens_seen": 13100505, + "step": 616, + "time_per_iteration": 2.6491007804870605 + }, + { + "auxiliary_loss_clip": 0.01226372, + "auxiliary_loss_mlp": 0.01057111, + "balance_loss_clip": 1.03181481, + "balance_loss_mlp": 1.04281175, + "epoch": 0.07418986352431912, + "flos": 21430985650560.0, + "grad_norm": 2.087853612504908, + "language_loss": 0.8490119, + "learning_rate": 3.979602551461683e-06, + "loss": 0.87184668, + "num_input_tokens_seen": 13121285, + "step": 617, + "time_per_iteration": 2.7008414268493652 + }, + { + "auxiliary_loss_clip": 0.01207358, + "auxiliary_loss_mlp": 0.01046445, + "balance_loss_clip": 0.99592519, + "balance_loss_mlp": 1.03217566, + "epoch": 0.07431010641495822, + "flos": 12021133777920.0, + "grad_norm": 2.23457101403944, + "language_loss": 0.91543508, + "learning_rate": 3.979491432885799e-06, + "loss": 0.93797308, + "num_input_tokens_seen": 13137550, + "step": 618, + "time_per_iteration": 4.412399530410767 + }, + { + "auxiliary_loss_clip": 0.011646, + "auxiliary_loss_mlp": 0.00514821, + "balance_loss_clip": 0.94918895, + "balance_loss_mlp": 0.0, + "epoch": 0.0744303493055973, + "flos": 20957575374720.0, + "grad_norm": 1.9017342382679816, + "language_loss": 0.82836086, + "learning_rate": 3.97938001402264e-06, + "loss": 0.84515512, + "num_input_tokens_seen": 13156675, + "step": 619, + "time_per_iteration": 2.717764377593994 + }, + { + "auxiliary_loss_clip": 0.01182443, + "auxiliary_loss_mlp": 0.01044443, + "balance_loss_clip": 0.9582386, + "balance_loss_mlp": 1.03078198, + "epoch": 0.0745505921962364, + "flos": 16253134272000.0, + "grad_norm": 3.923454223913741, + "language_loss": 0.80131119, + "learning_rate": 3.979268294889105e-06, + "loss": 0.82358003, + "num_input_tokens_seen": 13172225, + "step": 620, + "time_per_iteration": 4.464945077896118 + }, + { + "auxiliary_loss_clip": 0.01258847, + "auxiliary_loss_mlp": 0.01049902, + "balance_loss_clip": 1.07355142, + "balance_loss_mlp": 1.03735518, + "epoch": 0.07467083508687548, + "flos": 50944635550080.0, + "grad_norm": 1.7872356986067168, + "language_loss": 0.7407608, + "learning_rate": 3.979156275502143e-06, + "loss": 0.7638483, + "num_input_tokens_seen": 13195885, + "step": 621, + "time_per_iteration": 2.8603675365448 + }, + { + "auxiliary_loss_clip": 0.01182193, + "auxiliary_loss_mlp": 0.01043501, + "balance_loss_clip": 0.95499265, + "balance_loss_mlp": 1.03042376, + "epoch": 0.07479107797751458, + "flos": 17529686697600.0, + "grad_norm": 2.7231443685444163, + "language_loss": 0.91441309, + "learning_rate": 3.979043955878749e-06, + "loss": 0.93667001, + "num_input_tokens_seen": 13213730, + "step": 622, + "time_per_iteration": 2.730130672454834 + }, + { + "auxiliary_loss_clip": 0.0120419, + "auxiliary_loss_mlp": 0.01050234, + "balance_loss_clip": 0.99541634, + "balance_loss_mlp": 1.03679919, + "epoch": 0.07491132086815366, + "flos": 23473943591040.0, + "grad_norm": 2.153911714910758, + "language_loss": 0.82873809, + "learning_rate": 3.978931336035959e-06, + "loss": 0.85128236, + "num_input_tokens_seen": 13232540, + "step": 623, + "time_per_iteration": 2.736889600753784 + }, + { + "auxiliary_loss_clip": 0.01233739, + "auxiliary_loss_mlp": 0.01043255, + "balance_loss_clip": 1.03579021, + "balance_loss_mlp": 1.03054142, + "epoch": 0.07503156375879276, + "flos": 20157557708160.0, + "grad_norm": 2.401840387380393, + "language_loss": 0.81980681, + "learning_rate": 3.9788184159908595e-06, + "loss": 0.8425768, + "num_input_tokens_seen": 13249670, + "step": 624, + "time_per_iteration": 2.6394879817962646 + }, + { + "auxiliary_loss_clip": 0.01195325, + "auxiliary_loss_mlp": 0.01044544, + "balance_loss_clip": 0.98974121, + "balance_loss_mlp": 1.03230095, + "epoch": 0.07515180664943186, + "flos": 15115519653120.0, + "grad_norm": 4.719569117845207, + "language_loss": 0.82342881, + "learning_rate": 3.97870519576058e-06, + "loss": 0.84582746, + "num_input_tokens_seen": 13266095, + "step": 625, + "time_per_iteration": 2.7298803329467773 + }, + { + "auxiliary_loss_clip": 0.01182023, + "auxiliary_loss_mlp": 0.00514405, + "balance_loss_clip": 0.95547831, + "balance_loss_mlp": 0.0, + "epoch": 0.07527204954007094, + "flos": 21287702298240.0, + "grad_norm": 3.4496785140884527, + "language_loss": 0.809883, + "learning_rate": 3.978591675362295e-06, + "loss": 0.82684731, + "num_input_tokens_seen": 13284810, + "step": 626, + "time_per_iteration": 2.7754740715026855 + }, + { + "auxiliary_loss_clip": 0.01152074, + "auxiliary_loss_mlp": 0.01053692, + "balance_loss_clip": 0.91792947, + "balance_loss_mlp": 1.04076374, + "epoch": 0.07539229243071004, + "flos": 21324187537920.0, + "grad_norm": 1.8388680898179066, + "language_loss": 0.87545681, + "learning_rate": 3.978477854813226e-06, + "loss": 0.89751446, + "num_input_tokens_seen": 13304150, + "step": 627, + "time_per_iteration": 2.868643045425415 + }, + { + "auxiliary_loss_clip": 0.01233264, + "auxiliary_loss_mlp": 0.01047509, + "balance_loss_clip": 1.0325613, + "balance_loss_mlp": 1.03487909, + "epoch": 0.07551253532134912, + "flos": 13042540920960.0, + "grad_norm": 2.2521792321264753, + "language_loss": 0.82531095, + "learning_rate": 3.97836373413064e-06, + "loss": 0.84811866, + "num_input_tokens_seen": 13322205, + "step": 628, + "time_per_iteration": 2.715444564819336 + }, + { + "auxiliary_loss_clip": 0.01254763, + "auxiliary_loss_mlp": 0.01047987, + "balance_loss_clip": 1.07025397, + "balance_loss_mlp": 1.03445089, + "epoch": 0.07563277821198822, + "flos": 19208761908480.0, + "grad_norm": 2.026127290839993, + "language_loss": 0.74547136, + "learning_rate": 3.978249313331848e-06, + "loss": 0.76849884, + "num_input_tokens_seen": 13340435, + "step": 629, + "time_per_iteration": 2.6554019451141357 + }, + { + "auxiliary_loss_clip": 0.01236961, + "auxiliary_loss_mlp": 0.00514379, + "balance_loss_clip": 1.03292704, + "balance_loss_mlp": 0.0, + "epoch": 0.07575302110262731, + "flos": 19537200892800.0, + "grad_norm": 2.724771896250839, + "language_loss": 0.61985087, + "learning_rate": 3.978134592434208e-06, + "loss": 0.63736427, + "num_input_tokens_seen": 13358185, + "step": 630, + "time_per_iteration": 2.6428680419921875 + }, + { + "auxiliary_loss_clip": 0.01068735, + "auxiliary_loss_mlp": 0.01013265, + "balance_loss_clip": 0.89297318, + "balance_loss_mlp": 1.00873494, + "epoch": 0.0758732639932664, + "flos": 67961808017280.0, + "grad_norm": 1.0116161152071552, + "language_loss": 0.59394336, + "learning_rate": 3.978019571455123e-06, + "loss": 0.61476338, + "num_input_tokens_seen": 13410130, + "step": 631, + "time_per_iteration": 3.329061269760132 + }, + { + "auxiliary_loss_clip": 0.01256356, + "auxiliary_loss_mlp": 0.01043954, + "balance_loss_clip": 1.07325101, + "balance_loss_mlp": 1.03227174, + "epoch": 0.07599350688390549, + "flos": 18989204025600.0, + "grad_norm": 2.4544710659539746, + "language_loss": 0.83696949, + "learning_rate": 3.977904250412042e-06, + "loss": 0.85997266, + "num_input_tokens_seen": 13429085, + "step": 632, + "time_per_iteration": 2.672748565673828 + }, + { + "auxiliary_loss_clip": 0.01211269, + "auxiliary_loss_mlp": 0.01048104, + "balance_loss_clip": 0.99373841, + "balance_loss_mlp": 1.03603363, + "epoch": 0.07611374977454458, + "flos": 21069006341760.0, + "grad_norm": 2.108833388042563, + "language_loss": 0.85854524, + "learning_rate": 3.97778862932246e-06, + "loss": 0.88113904, + "num_input_tokens_seen": 13446250, + "step": 633, + "time_per_iteration": 2.714778423309326 + }, + { + "auxiliary_loss_clip": 0.01065453, + "auxiliary_loss_mlp": 0.01044772, + "balance_loss_clip": 0.79000175, + "balance_loss_mlp": 1.03228438, + "epoch": 0.07623399266518367, + "flos": 18514536773760.0, + "grad_norm": 2.87731974663882, + "language_loss": 0.93994629, + "learning_rate": 3.9776727082039144e-06, + "loss": 0.96104848, + "num_input_tokens_seen": 13463220, + "step": 634, + "time_per_iteration": 3.186225414276123 + }, + { + "auxiliary_loss_clip": 0.01164258, + "auxiliary_loss_mlp": 0.01004362, + "balance_loss_clip": 1.04855275, + "balance_loss_mlp": 0.99980783, + "epoch": 0.07635423555582276, + "flos": 44663036077440.0, + "grad_norm": 0.8564132342471571, + "language_loss": 0.55507267, + "learning_rate": 3.977556487073991e-06, + "loss": 0.57675886, + "num_input_tokens_seen": 13517775, + "step": 635, + "time_per_iteration": 3.261197090148926 + }, + { + "auxiliary_loss_clip": 0.01195935, + "auxiliary_loss_mlp": 0.01043195, + "balance_loss_clip": 0.98833883, + "balance_loss_mlp": 1.03061259, + "epoch": 0.07647447844646185, + "flos": 21761148487680.0, + "grad_norm": 1.8086203374663248, + "language_loss": 0.81578112, + "learning_rate": 3.97743996595032e-06, + "loss": 0.83817244, + "num_input_tokens_seen": 13537815, + "step": 636, + "time_per_iteration": 2.7419040203094482 + }, + { + "auxiliary_loss_clip": 0.0125856, + "auxiliary_loss_mlp": 0.01048182, + "balance_loss_clip": 1.07328296, + "balance_loss_mlp": 1.0347116, + "epoch": 0.07659472133710095, + "flos": 23806799948160.0, + "grad_norm": 1.7184587985113393, + "language_loss": 0.81797397, + "learning_rate": 3.9773231448505804e-06, + "loss": 0.84104139, + "num_input_tokens_seen": 13559605, + "step": 637, + "time_per_iteration": 2.6581926345825195 + }, + { + "auxiliary_loss_clip": 0.012062, + "auxiliary_loss_mlp": 0.00514154, + "balance_loss_clip": 0.99454629, + "balance_loss_mlp": 0.0, + "epoch": 0.07671496422774003, + "flos": 21469984842240.0, + "grad_norm": 2.5101994880551013, + "language_loss": 0.78190053, + "learning_rate": 3.977206023792491e-06, + "loss": 0.79910409, + "num_input_tokens_seen": 13579495, + "step": 638, + "time_per_iteration": 2.7522034645080566 + }, + { + "auxiliary_loss_clip": 0.0123004, + "auxiliary_loss_mlp": 0.01047463, + "balance_loss_clip": 1.03570342, + "balance_loss_mlp": 1.03476691, + "epoch": 0.07683520711837913, + "flos": 16980971558400.0, + "grad_norm": 2.268756383909576, + "language_loss": 0.80898708, + "learning_rate": 3.97708860279382e-06, + "loss": 0.83176208, + "num_input_tokens_seen": 13597605, + "step": 639, + "time_per_iteration": 2.664716958999634 + }, + { + "auxiliary_loss_clip": 0.01173671, + "auxiliary_loss_mlp": 0.01045605, + "balance_loss_clip": 0.95367682, + "balance_loss_mlp": 1.03293908, + "epoch": 0.07695545000901821, + "flos": 23476744851840.0, + "grad_norm": 1.7158052655532137, + "language_loss": 0.77977061, + "learning_rate": 3.97697088187238e-06, + "loss": 0.80196333, + "num_input_tokens_seen": 13618120, + "step": 640, + "time_per_iteration": 2.758458375930786 + }, + { + "auxiliary_loss_clip": 0.01205675, + "auxiliary_loss_mlp": 0.01040501, + "balance_loss_clip": 0.99703968, + "balance_loss_mlp": 1.02815068, + "epoch": 0.07707569289965731, + "flos": 17634258167040.0, + "grad_norm": 4.755753145760288, + "language_loss": 0.9180305, + "learning_rate": 3.976852861046029e-06, + "loss": 0.94049227, + "num_input_tokens_seen": 13634735, + "step": 641, + "time_per_iteration": 2.7515127658843994 + }, + { + "auxiliary_loss_clip": 0.01155889, + "auxiliary_loss_mlp": 0.01044014, + "balance_loss_clip": 0.91805995, + "balance_loss_mlp": 1.03208709, + "epoch": 0.0771959357902964, + "flos": 25775674087680.0, + "grad_norm": 1.5890148038430834, + "language_loss": 0.80411041, + "learning_rate": 3.97673454033267e-06, + "loss": 0.82610941, + "num_input_tokens_seen": 13656835, + "step": 642, + "time_per_iteration": 2.9098541736602783 + }, + { + "auxiliary_loss_clip": 0.01208329, + "auxiliary_loss_mlp": 0.01050376, + "balance_loss_clip": 0.99266839, + "balance_loss_mlp": 1.03866386, + "epoch": 0.07731617868093549, + "flos": 19828651847040.0, + "grad_norm": 2.0011224588339283, + "language_loss": 0.82644159, + "learning_rate": 3.976615919750254e-06, + "loss": 0.84902859, + "num_input_tokens_seen": 13674535, + "step": 643, + "time_per_iteration": 3.758094310760498 + }, + { + "auxiliary_loss_clip": 0.01225189, + "auxiliary_loss_mlp": 0.01044686, + "balance_loss_clip": 1.03109026, + "balance_loss_mlp": 1.03155506, + "epoch": 0.07743642157157458, + "flos": 21324654414720.0, + "grad_norm": 2.0647552742308233, + "language_loss": 0.86745965, + "learning_rate": 3.976496999316775e-06, + "loss": 0.89015841, + "num_input_tokens_seen": 13693290, + "step": 644, + "time_per_iteration": 3.5650384426116943 + }, + { + "auxiliary_loss_clip": 0.01211, + "auxiliary_loss_mlp": 0.01051408, + "balance_loss_clip": 1.00104904, + "balance_loss_mlp": 1.0373652, + "epoch": 0.07755666446221367, + "flos": 19969133938560.0, + "grad_norm": 2.0883205634499755, + "language_loss": 0.84281772, + "learning_rate": 3.976377779050271e-06, + "loss": 0.8654418, + "num_input_tokens_seen": 13711420, + "step": 645, + "time_per_iteration": 3.6074013710021973 + }, + { + "auxiliary_loss_clip": 0.01220017, + "auxiliary_loss_mlp": 0.01055812, + "balance_loss_clip": 1.02806401, + "balance_loss_mlp": 1.04260349, + "epoch": 0.07767690735285276, + "flos": 23623224514560.0, + "grad_norm": 1.9822081963420077, + "language_loss": 0.84174353, + "learning_rate": 3.976258258968831e-06, + "loss": 0.86450183, + "num_input_tokens_seen": 13729965, + "step": 646, + "time_per_iteration": 3.6465742588043213 + }, + { + "auxiliary_loss_clip": 0.0118402, + "auxiliary_loss_mlp": 0.01050038, + "balance_loss_clip": 0.95750636, + "balance_loss_mlp": 1.03744388, + "epoch": 0.07779715024349185, + "flos": 22236246702720.0, + "grad_norm": 2.485155942619648, + "language_loss": 0.74523318, + "learning_rate": 3.976138439090583e-06, + "loss": 0.76757371, + "num_input_tokens_seen": 13748045, + "step": 647, + "time_per_iteration": 2.9065616130828857 + }, + { + "auxiliary_loss_clip": 0.0117956, + "auxiliary_loss_mlp": 0.01040888, + "balance_loss_clip": 0.95656633, + "balance_loss_mlp": 1.02745938, + "epoch": 0.07791739313413094, + "flos": 20955097336320.0, + "grad_norm": 2.208815096460399, + "language_loss": 0.85346639, + "learning_rate": 3.976018319433706e-06, + "loss": 0.87567091, + "num_input_tokens_seen": 13765590, + "step": 648, + "time_per_iteration": 2.7921335697174072 + }, + { + "auxiliary_loss_clip": 0.01231395, + "auxiliary_loss_mlp": 0.01040679, + "balance_loss_clip": 1.03394866, + "balance_loss_mlp": 1.02751231, + "epoch": 0.07803763602477004, + "flos": 19312327797120.0, + "grad_norm": 2.4061805208917146, + "language_loss": 0.91164297, + "learning_rate": 3.9758979000164205e-06, + "loss": 0.93436366, + "num_input_tokens_seen": 13782410, + "step": 649, + "time_per_iteration": 2.727485179901123 + }, + { + "auxiliary_loss_clip": 0.01185261, + "auxiliary_loss_mlp": 0.01047487, + "balance_loss_clip": 0.9558388, + "balance_loss_mlp": 1.0344398, + "epoch": 0.07815787891540912, + "flos": 22710806213760.0, + "grad_norm": 1.8376142454272781, + "language_loss": 0.71843964, + "learning_rate": 3.975777180856995e-06, + "loss": 0.74076706, + "num_input_tokens_seen": 13801530, + "step": 650, + "time_per_iteration": 2.7862043380737305 + }, + { + "auxiliary_loss_clip": 0.01258956, + "auxiliary_loss_mlp": 0.0105597, + "balance_loss_clip": 1.07189107, + "balance_loss_mlp": 1.04303598, + "epoch": 0.07827812180604822, + "flos": 22711129436160.0, + "grad_norm": 4.876654190512948, + "language_loss": 0.86156499, + "learning_rate": 3.975656161973742e-06, + "loss": 0.88471431, + "num_input_tokens_seen": 13820615, + "step": 651, + "time_per_iteration": 2.7131006717681885 + }, + { + "auxiliary_loss_clip": 0.01258427, + "auxiliary_loss_mlp": 0.01046443, + "balance_loss_clip": 1.07076323, + "balance_loss_mlp": 1.03213811, + "epoch": 0.0783983646966873, + "flos": 21725597001600.0, + "grad_norm": 2.578308950598618, + "language_loss": 0.88832283, + "learning_rate": 3.9755348433850194e-06, + "loss": 0.91137159, + "num_input_tokens_seen": 13835955, + "step": 652, + "time_per_iteration": 2.6611576080322266 + }, + { + "auxiliary_loss_clip": 0.01110798, + "auxiliary_loss_mlp": 0.01009508, + "balance_loss_clip": 0.96839416, + "balance_loss_mlp": 1.00502586, + "epoch": 0.0785186075873264, + "flos": 60640877537280.0, + "grad_norm": 0.9606973017422978, + "language_loss": 0.63622189, + "learning_rate": 3.975413225109232e-06, + "loss": 0.65742493, + "num_input_tokens_seen": 13896505, + "step": 653, + "time_per_iteration": 3.292229175567627 + }, + { + "auxiliary_loss_clip": 0.01230343, + "auxiliary_loss_mlp": 0.01042137, + "balance_loss_clip": 1.03299212, + "balance_loss_mlp": 1.02841592, + "epoch": 0.0786388504779655, + "flos": 23877902920320.0, + "grad_norm": 18.043956583703228, + "language_loss": 0.93506688, + "learning_rate": 3.975291307164829e-06, + "loss": 0.95779169, + "num_input_tokens_seen": 13915150, + "step": 654, + "time_per_iteration": 2.6815192699432373 + }, + { + "auxiliary_loss_clip": 0.0116683, + "auxiliary_loss_mlp": 0.01043886, + "balance_loss_clip": 0.95062828, + "balance_loss_mlp": 1.03232288, + "epoch": 0.07875909336860458, + "flos": 15158684822400.0, + "grad_norm": 1.8635196853698823, + "language_loss": 0.84949291, + "learning_rate": 3.975169089570306e-06, + "loss": 0.87160003, + "num_input_tokens_seen": 13933525, + "step": 655, + "time_per_iteration": 2.6587367057800293 + }, + { + "auxiliary_loss_clip": 0.01218288, + "auxiliary_loss_mlp": 0.01050759, + "balance_loss_clip": 1.02909636, + "balance_loss_mlp": 1.03777075, + "epoch": 0.07887933625924368, + "flos": 22236857233920.0, + "grad_norm": 1.9267803286750655, + "language_loss": 0.91507131, + "learning_rate": 3.975046572344202e-06, + "loss": 0.93776178, + "num_input_tokens_seen": 13949985, + "step": 656, + "time_per_iteration": 2.67146897315979 + }, + { + "auxiliary_loss_clip": 0.01176678, + "auxiliary_loss_mlp": 0.01042678, + "balance_loss_clip": 0.95192361, + "balance_loss_mlp": 1.03016138, + "epoch": 0.07899957914988276, + "flos": 20777734955520.0, + "grad_norm": 2.3976462975632953, + "language_loss": 0.71143717, + "learning_rate": 3.974923755505103e-06, + "loss": 0.73363072, + "num_input_tokens_seen": 13969215, + "step": 657, + "time_per_iteration": 2.717615842819214 + }, + { + "auxiliary_loss_clip": 0.01170141, + "auxiliary_loss_mlp": 0.01048825, + "balance_loss_clip": 0.95208335, + "balance_loss_mlp": 1.03553963, + "epoch": 0.07911982204052186, + "flos": 23003047267200.0, + "grad_norm": 1.7455608498063284, + "language_loss": 0.91259909, + "learning_rate": 3.974800639071641e-06, + "loss": 0.93478876, + "num_input_tokens_seen": 13989935, + "step": 658, + "time_per_iteration": 2.7600748538970947 + }, + { + "auxiliary_loss_clip": 0.01113541, + "auxiliary_loss_mlp": 0.00514281, + "balance_loss_clip": 0.87236476, + "balance_loss_mlp": 0.0, + "epoch": 0.07924006493116094, + "flos": 23111389664640.0, + "grad_norm": 1.9877273210477535, + "language_loss": 1.005018, + "learning_rate": 3.974677223062492e-06, + "loss": 1.02129614, + "num_input_tokens_seen": 14007150, + "step": 659, + "time_per_iteration": 2.832569122314453 + }, + { + "auxiliary_loss_clip": 0.01199909, + "auxiliary_loss_mlp": 0.01050269, + "balance_loss_clip": 0.99221122, + "balance_loss_mlp": 1.03778791, + "epoch": 0.07936030782180004, + "flos": 16472153450880.0, + "grad_norm": 2.155240666400648, + "language_loss": 0.74452299, + "learning_rate": 3.974553507496378e-06, + "loss": 0.76702476, + "num_input_tokens_seen": 14025725, + "step": 660, + "time_per_iteration": 2.9101479053497314 + }, + { + "auxiliary_loss_clip": 0.01197551, + "auxiliary_loss_mlp": 0.01053179, + "balance_loss_clip": 0.99300724, + "balance_loss_mlp": 1.03851032, + "epoch": 0.07948055071243913, + "flos": 23733290764800.0, + "grad_norm": 2.3438812343694337, + "language_loss": 0.89561713, + "learning_rate": 3.974429492392068e-06, + "loss": 0.91812444, + "num_input_tokens_seen": 14045750, + "step": 661, + "time_per_iteration": 2.727384567260742 + }, + { + "auxiliary_loss_clip": 0.01255179, + "auxiliary_loss_mlp": 0.00514223, + "balance_loss_clip": 1.0715425, + "balance_loss_mlp": 0.0, + "epoch": 0.07960079360307822, + "flos": 19573326996480.0, + "grad_norm": 5.038011206643484, + "language_loss": 0.91403091, + "learning_rate": 3.974305177768373e-06, + "loss": 0.93172497, + "num_input_tokens_seen": 14063960, + "step": 662, + "time_per_iteration": 2.6151492595672607 + }, + { + "auxiliary_loss_clip": 0.01170364, + "auxiliary_loss_mlp": 0.0105293, + "balance_loss_clip": 0.95435536, + "balance_loss_mlp": 1.03947687, + "epoch": 0.07972103649371731, + "flos": 23513409659520.0, + "grad_norm": 2.035399946071657, + "language_loss": 0.86339104, + "learning_rate": 3.974180563644152e-06, + "loss": 0.88562405, + "num_input_tokens_seen": 14082525, + "step": 663, + "time_per_iteration": 2.7675249576568604 + }, + { + "auxiliary_loss_clip": 0.01208849, + "auxiliary_loss_mlp": 0.01053348, + "balance_loss_clip": 0.99386495, + "balance_loss_mlp": 1.04033029, + "epoch": 0.0798412793843564, + "flos": 16726867770240.0, + "grad_norm": 2.2724560113086323, + "language_loss": 0.8903054, + "learning_rate": 3.97405565003831e-06, + "loss": 0.91292739, + "num_input_tokens_seen": 14098610, + "step": 664, + "time_per_iteration": 2.632566213607788 + }, + { + "auxiliary_loss_clip": 0.01184845, + "auxiliary_loss_mlp": 0.01051196, + "balance_loss_clip": 0.95509267, + "balance_loss_mlp": 1.03869104, + "epoch": 0.07996152227499549, + "flos": 18223337214720.0, + "grad_norm": 2.3663260606414243, + "language_loss": 0.77859366, + "learning_rate": 3.973930436969794e-06, + "loss": 0.8009541, + "num_input_tokens_seen": 14117065, + "step": 665, + "time_per_iteration": 2.7074267864227295 + }, + { + "auxiliary_loss_clip": 0.01198232, + "auxiliary_loss_mlp": 0.01052808, + "balance_loss_clip": 0.99095154, + "balance_loss_mlp": 1.0391047, + "epoch": 0.08008176516563459, + "flos": 20594877793920.0, + "grad_norm": 1.9738435577182019, + "language_loss": 0.85856569, + "learning_rate": 3.973804924457602e-06, + "loss": 0.8810761, + "num_input_tokens_seen": 14135145, + "step": 666, + "time_per_iteration": 2.670151710510254 + }, + { + "auxiliary_loss_clip": 0.01199434, + "auxiliary_loss_mlp": 0.01059115, + "balance_loss_clip": 0.99196041, + "balance_loss_mlp": 1.04588866, + "epoch": 0.08020200805627367, + "flos": 31834306863360.0, + "grad_norm": 1.7951296918932733, + "language_loss": 0.85671961, + "learning_rate": 3.973679112520771e-06, + "loss": 0.87930518, + "num_input_tokens_seen": 14156860, + "step": 667, + "time_per_iteration": 2.7715609073638916 + }, + { + "auxiliary_loss_clip": 0.01167487, + "auxiliary_loss_mlp": 0.01044668, + "balance_loss_clip": 0.94973731, + "balance_loss_mlp": 1.03200758, + "epoch": 0.08032225094691277, + "flos": 17783503176960.0, + "grad_norm": 1.7612746659716645, + "language_loss": 0.98701382, + "learning_rate": 3.973553001178389e-06, + "loss": 1.00913537, + "num_input_tokens_seen": 14174365, + "step": 668, + "time_per_iteration": 2.675635814666748 + }, + { + "auxiliary_loss_clip": 0.01176029, + "auxiliary_loss_mlp": 0.01045278, + "balance_loss_clip": 0.95480847, + "balance_loss_mlp": 1.03288651, + "epoch": 0.08044249383755185, + "flos": 24061693835520.0, + "grad_norm": 1.9488184248519163, + "language_loss": 0.76015127, + "learning_rate": 3.973426590449585e-06, + "loss": 0.78236431, + "num_input_tokens_seen": 14192320, + "step": 669, + "time_per_iteration": 3.6531670093536377 + }, + { + "auxiliary_loss_clip": 0.01155759, + "auxiliary_loss_mlp": 0.01062727, + "balance_loss_clip": 0.918697, + "balance_loss_mlp": 1.04965544, + "epoch": 0.08056273672819095, + "flos": 18223624523520.0, + "grad_norm": 2.261004302543909, + "language_loss": 0.75293756, + "learning_rate": 3.9732998803535364e-06, + "loss": 0.7751224, + "num_input_tokens_seen": 14210380, + "step": 670, + "time_per_iteration": 3.632297992706299 + }, + { + "auxiliary_loss_clip": 0.01259048, + "auxiliary_loss_mlp": 0.01046635, + "balance_loss_clip": 1.07397723, + "balance_loss_mlp": 1.03376675, + "epoch": 0.08068297961883003, + "flos": 19676856971520.0, + "grad_norm": 2.266009591620453, + "language_loss": 0.85112715, + "learning_rate": 3.973172870909465e-06, + "loss": 0.87418395, + "num_input_tokens_seen": 14225145, + "step": 671, + "time_per_iteration": 2.669497489929199 + }, + { + "auxiliary_loss_clip": 0.01212563, + "auxiliary_loss_mlp": 0.01046256, + "balance_loss_clip": 0.99455768, + "balance_loss_mlp": 1.03199267, + "epoch": 0.08080322250946913, + "flos": 23148736830720.0, + "grad_norm": 2.6209316546798798, + "language_loss": 0.80962002, + "learning_rate": 3.973045562136638e-06, + "loss": 0.83220822, + "num_input_tokens_seen": 14241960, + "step": 672, + "time_per_iteration": 3.5850160121917725 + }, + { + "auxiliary_loss_clip": 0.01235537, + "auxiliary_loss_mlp": 0.01053598, + "balance_loss_clip": 1.0357182, + "balance_loss_mlp": 1.0412004, + "epoch": 0.08092346540010822, + "flos": 21763626526080.0, + "grad_norm": 2.060511039604737, + "language_loss": 0.9173345, + "learning_rate": 3.972917954054368e-06, + "loss": 0.94022584, + "num_input_tokens_seen": 14260515, + "step": 673, + "time_per_iteration": 2.696058511734009 + }, + { + "auxiliary_loss_clip": 0.01205568, + "auxiliary_loss_mlp": 0.01049177, + "balance_loss_clip": 0.99912745, + "balance_loss_mlp": 1.03533113, + "epoch": 0.08104370829074731, + "flos": 21032485188480.0, + "grad_norm": 3.958993063266246, + "language_loss": 0.81947428, + "learning_rate": 3.972790046682013e-06, + "loss": 0.8420217, + "num_input_tokens_seen": 14279190, + "step": 674, + "time_per_iteration": 2.646238327026367 + }, + { + "auxiliary_loss_clip": 0.01182397, + "auxiliary_loss_mlp": 0.01043664, + "balance_loss_clip": 0.95251048, + "balance_loss_mlp": 1.03054523, + "epoch": 0.0811639511813864, + "flos": 20083186598400.0, + "grad_norm": 1.8025393403338765, + "language_loss": 0.79082239, + "learning_rate": 3.972661840038977e-06, + "loss": 0.81308299, + "num_input_tokens_seen": 14299480, + "step": 675, + "time_per_iteration": 2.8990747928619385 + }, + { + "auxiliary_loss_clip": 0.01230442, + "auxiliary_loss_mlp": 0.01048647, + "balance_loss_clip": 1.03343177, + "balance_loss_mlp": 1.03588557, + "epoch": 0.08128419407202549, + "flos": 16836718538880.0, + "grad_norm": 2.3246313468455755, + "language_loss": 0.83634204, + "learning_rate": 3.972533334144707e-06, + "loss": 0.85913295, + "num_input_tokens_seen": 14316405, + "step": 676, + "time_per_iteration": 2.612626075744629 + }, + { + "auxiliary_loss_clip": 0.01235225, + "auxiliary_loss_mlp": 0.01044597, + "balance_loss_clip": 1.03310144, + "balance_loss_mlp": 1.03101957, + "epoch": 0.08140443696266458, + "flos": 23769273214080.0, + "grad_norm": 2.408108633342443, + "language_loss": 0.78597116, + "learning_rate": 3.972404529018699e-06, + "loss": 0.80876935, + "num_input_tokens_seen": 14336265, + "step": 677, + "time_per_iteration": 2.7189738750457764 + }, + { + "auxiliary_loss_clip": 0.01199243, + "auxiliary_loss_mlp": 0.0104797, + "balance_loss_clip": 0.99073899, + "balance_loss_mlp": 1.03480315, + "epoch": 0.08152467985330367, + "flos": 24390132819840.0, + "grad_norm": 1.6741762719321012, + "language_loss": 0.85546219, + "learning_rate": 3.972275424680493e-06, + "loss": 0.87793428, + "num_input_tokens_seen": 14356375, + "step": 678, + "time_per_iteration": 2.714111328125 + }, + { + "auxiliary_loss_clip": 0.01255112, + "auxiliary_loss_mlp": 0.01040913, + "balance_loss_clip": 1.07218933, + "balance_loss_mlp": 1.02811038, + "epoch": 0.08164492274394276, + "flos": 19317750750720.0, + "grad_norm": 2.1329946900320125, + "language_loss": 0.92048311, + "learning_rate": 3.972146021149673e-06, + "loss": 0.94344342, + "num_input_tokens_seen": 14374650, + "step": 679, + "time_per_iteration": 2.6556236743927 + }, + { + "auxiliary_loss_clip": 0.01188676, + "auxiliary_loss_mlp": 0.01048285, + "balance_loss_clip": 0.96051514, + "balance_loss_mlp": 1.0361495, + "epoch": 0.08176516563458186, + "flos": 14830461319680.0, + "grad_norm": 2.3462590037967224, + "language_loss": 0.78515267, + "learning_rate": 3.972016318445868e-06, + "loss": 0.8075223, + "num_input_tokens_seen": 14392650, + "step": 680, + "time_per_iteration": 2.6911113262176514 + }, + { + "auxiliary_loss_clip": 0.01231949, + "auxiliary_loss_mlp": 0.0103667, + "balance_loss_clip": 1.03513646, + "balance_loss_mlp": 1.02380776, + "epoch": 0.08188540852522094, + "flos": 22602320161920.0, + "grad_norm": 2.1533171936215263, + "language_loss": 0.92088556, + "learning_rate": 3.971886316588757e-06, + "loss": 0.94357181, + "num_input_tokens_seen": 14413155, + "step": 681, + "time_per_iteration": 2.6988322734832764 + }, + { + "auxiliary_loss_clip": 0.01172695, + "auxiliary_loss_mlp": 0.01037642, + "balance_loss_clip": 0.95836973, + "balance_loss_mlp": 1.02392709, + "epoch": 0.08200565141586004, + "flos": 19463727623040.0, + "grad_norm": 2.546615685092045, + "language_loss": 0.73687679, + "learning_rate": 3.9717560155980595e-06, + "loss": 0.75898015, + "num_input_tokens_seen": 14428805, + "step": 682, + "time_per_iteration": 2.7234699726104736 + }, + { + "auxiliary_loss_clip": 0.01231058, + "auxiliary_loss_mlp": 0.01043193, + "balance_loss_clip": 1.03320229, + "balance_loss_mlp": 1.03057468, + "epoch": 0.08212589430649912, + "flos": 20594662312320.0, + "grad_norm": 1.9999759108963784, + "language_loss": 0.92329913, + "learning_rate": 3.971625415493542e-06, + "loss": 0.94604164, + "num_input_tokens_seen": 14447125, + "step": 683, + "time_per_iteration": 2.6271286010742188 + }, + { + "auxiliary_loss_clip": 0.01176834, + "auxiliary_loss_mlp": 0.01045583, + "balance_loss_clip": 0.95240033, + "balance_loss_mlp": 1.03295302, + "epoch": 0.08224613719713822, + "flos": 25953611086080.0, + "grad_norm": 1.9607408658861631, + "language_loss": 0.8774786, + "learning_rate": 3.971494516295017e-06, + "loss": 0.89970279, + "num_input_tokens_seen": 14466575, + "step": 684, + "time_per_iteration": 2.744389772415161 + }, + { + "auxiliary_loss_clip": 0.01187042, + "auxiliary_loss_mlp": 0.01041418, + "balance_loss_clip": 0.95648348, + "balance_loss_mlp": 1.02891338, + "epoch": 0.08236638008777732, + "flos": 23768734510080.0, + "grad_norm": 1.9424752513850674, + "language_loss": 0.85408568, + "learning_rate": 3.971363318022341e-06, + "loss": 0.87637031, + "num_input_tokens_seen": 14487915, + "step": 685, + "time_per_iteration": 2.762336492538452 + }, + { + "auxiliary_loss_clip": 0.01207667, + "auxiliary_loss_mlp": 0.01041108, + "balance_loss_clip": 0.99156415, + "balance_loss_mlp": 1.02813768, + "epoch": 0.0824866229784164, + "flos": 38799144887040.0, + "grad_norm": 1.7141895113700045, + "language_loss": 0.68402159, + "learning_rate": 3.971231820695417e-06, + "loss": 0.70650929, + "num_input_tokens_seen": 14511530, + "step": 686, + "time_per_iteration": 2.8466951847076416 + }, + { + "auxiliary_loss_clip": 0.01211533, + "auxiliary_loss_mlp": 0.01048724, + "balance_loss_clip": 0.99657154, + "balance_loss_mlp": 1.03618312, + "epoch": 0.0826068658690555, + "flos": 23107762391040.0, + "grad_norm": 2.0522853225588245, + "language_loss": 0.8139624, + "learning_rate": 3.971100024334193e-06, + "loss": 0.83656496, + "num_input_tokens_seen": 14529050, + "step": 687, + "time_per_iteration": 2.6930978298187256 + }, + { + "auxiliary_loss_clip": 0.01166882, + "auxiliary_loss_mlp": 0.0104188, + "balance_loss_clip": 0.95100957, + "balance_loss_mlp": 1.02960134, + "epoch": 0.08272710875969458, + "flos": 21136374299520.0, + "grad_norm": 2.0504797463959594, + "language_loss": 0.8660242, + "learning_rate": 3.970967928958663e-06, + "loss": 0.88811183, + "num_input_tokens_seen": 14546165, + "step": 688, + "time_per_iteration": 2.771681547164917 + }, + { + "auxiliary_loss_clip": 0.01160081, + "auxiliary_loss_mlp": 0.01049173, + "balance_loss_clip": 0.91809952, + "balance_loss_mlp": 1.03694856, + "epoch": 0.08284735165033368, + "flos": 19063000517760.0, + "grad_norm": 1.7571685660818757, + "language_loss": 0.83571208, + "learning_rate": 3.970835534588865e-06, + "loss": 0.85780466, + "num_input_tokens_seen": 14563660, + "step": 689, + "time_per_iteration": 2.7341864109039307 + }, + { + "auxiliary_loss_clip": 0.01212768, + "auxiliary_loss_mlp": 0.01043409, + "balance_loss_clip": 0.99992329, + "balance_loss_mlp": 1.03135109, + "epoch": 0.08296759454097276, + "flos": 16727442387840.0, + "grad_norm": 2.2487656840836237, + "language_loss": 0.8561843, + "learning_rate": 3.970702841244883e-06, + "loss": 0.87874603, + "num_input_tokens_seen": 14581980, + "step": 690, + "time_per_iteration": 2.747305393218994 + }, + { + "auxiliary_loss_clip": 0.01235198, + "auxiliary_loss_mlp": 0.01044968, + "balance_loss_clip": 1.03700483, + "balance_loss_mlp": 1.03040051, + "epoch": 0.08308783743161186, + "flos": 18004928567040.0, + "grad_norm": 1.8329131718833411, + "language_loss": 0.82326049, + "learning_rate": 3.970569848946847e-06, + "loss": 0.84606218, + "num_input_tokens_seen": 14601795, + "step": 691, + "time_per_iteration": 2.6269097328186035 + }, + { + "auxiliary_loss_clip": 0.01221095, + "auxiliary_loss_mlp": 0.01044744, + "balance_loss_clip": 1.03339362, + "balance_loss_mlp": 1.03173804, + "epoch": 0.08320808032225095, + "flos": 15079788599040.0, + "grad_norm": 2.309447419043911, + "language_loss": 0.82511818, + "learning_rate": 3.970436557714932e-06, + "loss": 0.84777665, + "num_input_tokens_seen": 14618315, + "step": 692, + "time_per_iteration": 2.694298028945923 + }, + { + "auxiliary_loss_clip": 0.01203789, + "auxiliary_loss_mlp": 0.0104687, + "balance_loss_clip": 0.9959079, + "balance_loss_mlp": 1.03391826, + "epoch": 0.08332832321289003, + "flos": 22383085501440.0, + "grad_norm": 2.1627871673965977, + "language_loss": 0.86369145, + "learning_rate": 3.970302967569358e-06, + "loss": 0.88619804, + "num_input_tokens_seen": 14636905, + "step": 693, + "time_per_iteration": 2.7711212635040283 + }, + { + "auxiliary_loss_clip": 0.01236237, + "auxiliary_loss_mlp": 0.01042766, + "balance_loss_clip": 1.03929257, + "balance_loss_mlp": 1.02977824, + "epoch": 0.08344856610352913, + "flos": 24717386655360.0, + "grad_norm": 3.1762896150510316, + "language_loss": 0.68418694, + "learning_rate": 3.9701690785303896e-06, + "loss": 0.70697689, + "num_input_tokens_seen": 14656100, + "step": 694, + "time_per_iteration": 2.8330771923065186 + }, + { + "auxiliary_loss_clip": 0.01234305, + "auxiliary_loss_mlp": 0.01045733, + "balance_loss_clip": 1.03359652, + "balance_loss_mlp": 1.03334093, + "epoch": 0.08356880899416821, + "flos": 25370206387200.0, + "grad_norm": 2.1272107371914815, + "language_loss": 0.88454592, + "learning_rate": 3.970034890618339e-06, + "loss": 0.90734631, + "num_input_tokens_seen": 14675790, + "step": 695, + "time_per_iteration": 3.6266355514526367 + }, + { + "auxiliary_loss_clip": 0.01217976, + "auxiliary_loss_mlp": 0.01042184, + "balance_loss_clip": 1.03088629, + "balance_loss_mlp": 1.03034616, + "epoch": 0.08368905188480731, + "flos": 24353072962560.0, + "grad_norm": 1.996202367693509, + "language_loss": 0.88148141, + "learning_rate": 3.969900403853562e-06, + "loss": 0.90408301, + "num_input_tokens_seen": 14694830, + "step": 696, + "time_per_iteration": 3.5841119289398193 + }, + { + "auxiliary_loss_clip": 0.01257845, + "auxiliary_loss_mlp": 0.01052224, + "balance_loss_clip": 1.07504058, + "balance_loss_mlp": 1.04010034, + "epoch": 0.08380929477544641, + "flos": 18037319656320.0, + "grad_norm": 1.6780076077397588, + "language_loss": 0.78061694, + "learning_rate": 3.96976561825646e-06, + "loss": 0.80371761, + "num_input_tokens_seen": 14711920, + "step": 697, + "time_per_iteration": 2.638977289199829 + }, + { + "auxiliary_loss_clip": 0.01161375, + "auxiliary_loss_mlp": 0.01048216, + "balance_loss_clip": 0.91668695, + "balance_loss_mlp": 1.03483498, + "epoch": 0.08392953766608549, + "flos": 26286287875200.0, + "grad_norm": 2.131966790365833, + "language_loss": 0.87188721, + "learning_rate": 3.969630533847479e-06, + "loss": 0.89398313, + "num_input_tokens_seen": 14730880, + "step": 698, + "time_per_iteration": 4.590519905090332 + }, + { + "auxiliary_loss_clip": 0.01231405, + "auxiliary_loss_mlp": 0.01044671, + "balance_loss_clip": 1.03267717, + "balance_loss_mlp": 1.03323281, + "epoch": 0.08404978055672459, + "flos": 22492146170880.0, + "grad_norm": 1.9607349315097309, + "language_loss": 0.84107095, + "learning_rate": 3.969495150647113e-06, + "loss": 0.8638317, + "num_input_tokens_seen": 14749050, + "step": 699, + "time_per_iteration": 2.6465015411376953 + }, + { + "auxiliary_loss_clip": 0.01185032, + "auxiliary_loss_mlp": 0.01044709, + "balance_loss_clip": 0.96196914, + "balance_loss_mlp": 1.03215051, + "epoch": 0.08417002344736367, + "flos": 24826878288000.0, + "grad_norm": 1.7885926968459738, + "language_loss": 0.76435202, + "learning_rate": 3.969359468675899e-06, + "loss": 0.78664947, + "num_input_tokens_seen": 14769180, + "step": 700, + "time_per_iteration": 2.753899335861206 + }, + { + "auxiliary_loss_clip": 0.01231212, + "auxiliary_loss_mlp": 0.010442, + "balance_loss_clip": 1.03512359, + "balance_loss_mlp": 1.03176045, + "epoch": 0.08429026633800277, + "flos": 16945922862720.0, + "grad_norm": 2.1608620361668938, + "language_loss": 0.89527875, + "learning_rate": 3.969223487954418e-06, + "loss": 0.91803288, + "num_input_tokens_seen": 14786640, + "step": 701, + "time_per_iteration": 2.641726493835449 + }, + { + "auxiliary_loss_clip": 0.01153172, + "auxiliary_loss_mlp": 0.01044622, + "balance_loss_clip": 0.91532904, + "balance_loss_mlp": 1.03247488, + "epoch": 0.08441050922864185, + "flos": 23841920471040.0, + "grad_norm": 7.599767559993759, + "language_loss": 0.82841712, + "learning_rate": 3.969087208503301e-06, + "loss": 0.85039508, + "num_input_tokens_seen": 14806720, + "step": 702, + "time_per_iteration": 2.771265983581543 + }, + { + "auxiliary_loss_clip": 0.01168145, + "auxiliary_loss_mlp": 0.0104242, + "balance_loss_clip": 0.95304662, + "balance_loss_mlp": 1.02884197, + "epoch": 0.08453075211928095, + "flos": 25520205582720.0, + "grad_norm": 2.4777996769241395, + "language_loss": 0.8414098, + "learning_rate": 3.968950630343219e-06, + "loss": 0.86351538, + "num_input_tokens_seen": 14823705, + "step": 703, + "time_per_iteration": 2.7549023628234863 + }, + { + "auxiliary_loss_clip": 0.01196797, + "auxiliary_loss_mlp": 0.01044501, + "balance_loss_clip": 0.98962224, + "balance_loss_mlp": 1.03141153, + "epoch": 0.08465099500992004, + "flos": 19532496211200.0, + "grad_norm": 2.0894968444600455, + "language_loss": 0.93246877, + "learning_rate": 3.968813753494892e-06, + "loss": 0.95488179, + "num_input_tokens_seen": 14841865, + "step": 704, + "time_per_iteration": 2.711225986480713 + }, + { + "auxiliary_loss_clip": 0.01167281, + "auxiliary_loss_mlp": 0.00514732, + "balance_loss_clip": 0.9509176, + "balance_loss_mlp": 0.0, + "epoch": 0.08477123790055913, + "flos": 29351299403520.0, + "grad_norm": 2.8559371822825543, + "language_loss": 0.75326216, + "learning_rate": 3.968676577979084e-06, + "loss": 0.77008229, + "num_input_tokens_seen": 14861415, + "step": 705, + "time_per_iteration": 2.806840658187866 + }, + { + "auxiliary_loss_clip": 0.01149396, + "auxiliary_loss_mlp": 0.01045727, + "balance_loss_clip": 0.91375464, + "balance_loss_mlp": 1.03262043, + "epoch": 0.08489148079119822, + "flos": 18624495283200.0, + "grad_norm": 2.047657462475113, + "language_loss": 0.78139651, + "learning_rate": 3.968539103816605e-06, + "loss": 0.80334771, + "num_input_tokens_seen": 14879215, + "step": 706, + "time_per_iteration": 2.714547872543335 + }, + { + "auxiliary_loss_clip": 0.0120627, + "auxiliary_loss_mlp": 0.00514713, + "balance_loss_clip": 0.997967, + "balance_loss_mlp": 0.0, + "epoch": 0.0850117236818373, + "flos": 23471393725440.0, + "grad_norm": 1.8487575523321174, + "language_loss": 0.89388388, + "learning_rate": 3.9684013310283085e-06, + "loss": 0.91109371, + "num_input_tokens_seen": 14897900, + "step": 707, + "time_per_iteration": 2.722081422805786 + }, + { + "auxiliary_loss_clip": 0.01202241, + "auxiliary_loss_mlp": 0.01036895, + "balance_loss_clip": 0.99509257, + "balance_loss_mlp": 1.02494407, + "epoch": 0.0851319665724764, + "flos": 40625058896640.0, + "grad_norm": 2.227474455212696, + "language_loss": 0.64224648, + "learning_rate": 3.9682632596350956e-06, + "loss": 0.6646378, + "num_input_tokens_seen": 14919065, + "step": 708, + "time_per_iteration": 2.806365966796875 + }, + { + "auxiliary_loss_clip": 0.01224644, + "auxiliary_loss_mlp": 0.01047027, + "balance_loss_clip": 1.03236842, + "balance_loss_mlp": 1.03475404, + "epoch": 0.0852522094631155, + "flos": 15879554870400.0, + "grad_norm": 2.016214315770459, + "language_loss": 0.78456658, + "learning_rate": 3.968124889657911e-06, + "loss": 0.80728334, + "num_input_tokens_seen": 14934165, + "step": 709, + "time_per_iteration": 2.678860664367676 + }, + { + "auxiliary_loss_clip": 0.01150485, + "auxiliary_loss_mlp": 0.01050977, + "balance_loss_clip": 0.9155885, + "balance_loss_mlp": 1.03717256, + "epoch": 0.08537245235375458, + "flos": 14567091822720.0, + "grad_norm": 2.3733840421919834, + "language_loss": 0.9076786, + "learning_rate": 3.967986221117746e-06, + "loss": 0.92969322, + "num_input_tokens_seen": 14950105, + "step": 710, + "time_per_iteration": 2.740724563598633 + }, + { + "auxiliary_loss_clip": 0.01112608, + "auxiliary_loss_mlp": 0.01041741, + "balance_loss_clip": 0.83760589, + "balance_loss_mlp": 1.02878332, + "epoch": 0.08549269524439368, + "flos": 26468929555200.0, + "grad_norm": 2.3096130653384592, + "language_loss": 0.86642158, + "learning_rate": 3.967847254035635e-06, + "loss": 0.88796514, + "num_input_tokens_seen": 14969490, + "step": 711, + "time_per_iteration": 3.001957654953003 + }, + { + "auxiliary_loss_clip": 0.0118273, + "auxiliary_loss_mlp": 0.01044422, + "balance_loss_clip": 0.95589423, + "balance_loss_mlp": 1.03156567, + "epoch": 0.08561293813503276, + "flos": 13590214565760.0, + "grad_norm": 2.45897653242609, + "language_loss": 0.86548048, + "learning_rate": 3.967707988432661e-06, + "loss": 0.887752, + "num_input_tokens_seen": 14987195, + "step": 712, + "time_per_iteration": 3.11736798286438 + }, + { + "auxiliary_loss_clip": 0.01252833, + "auxiliary_loss_mlp": 0.01046504, + "balance_loss_clip": 1.06949544, + "balance_loss_mlp": 1.03332591, + "epoch": 0.08573318102567186, + "flos": 26943524979840.0, + "grad_norm": 2.025747450478754, + "language_loss": 0.87881434, + "learning_rate": 3.967568424329949e-06, + "loss": 0.90180767, + "num_input_tokens_seen": 15007620, + "step": 713, + "time_per_iteration": 2.6556360721588135 + }, + { + "auxiliary_loss_clip": 0.01125469, + "auxiliary_loss_mlp": 0.01023467, + "balance_loss_clip": 0.97584391, + "balance_loss_mlp": 1.01910424, + "epoch": 0.08585342391631094, + "flos": 67302739319040.0, + "grad_norm": 0.8293327410527604, + "language_loss": 0.55524695, + "learning_rate": 3.967428561748671e-06, + "loss": 0.57673627, + "num_input_tokens_seen": 15075590, + "step": 714, + "time_per_iteration": 3.3732564449310303 + }, + { + "auxiliary_loss_clip": 0.01143986, + "auxiliary_loss_mlp": 0.01048757, + "balance_loss_clip": 0.90911722, + "balance_loss_mlp": 1.0349766, + "epoch": 0.08597366680695004, + "flos": 22456594684800.0, + "grad_norm": 1.7966609373207028, + "language_loss": 0.87772566, + "learning_rate": 3.967288400710045e-06, + "loss": 0.89965308, + "num_input_tokens_seen": 15095055, + "step": 715, + "time_per_iteration": 2.807539701461792 + }, + { + "auxiliary_loss_clip": 0.01173493, + "auxiliary_loss_mlp": 0.01040836, + "balance_loss_clip": 0.95648301, + "balance_loss_mlp": 1.02897441, + "epoch": 0.08609390969758914, + "flos": 23550505430400.0, + "grad_norm": 1.9003598889265185, + "language_loss": 0.88330764, + "learning_rate": 3.9671479412353335e-06, + "loss": 0.90545094, + "num_input_tokens_seen": 15113520, + "step": 716, + "time_per_iteration": 2.7308106422424316 + }, + { + "auxiliary_loss_clip": 0.01226586, + "auxiliary_loss_mlp": 0.01045522, + "balance_loss_clip": 1.03175771, + "balance_loss_mlp": 1.03225446, + "epoch": 0.08621415258822822, + "flos": 25885848078720.0, + "grad_norm": 2.124310398722564, + "language_loss": 0.74346292, + "learning_rate": 3.967007183345843e-06, + "loss": 0.76618409, + "num_input_tokens_seen": 15133375, + "step": 717, + "time_per_iteration": 2.7034778594970703 + }, + { + "auxiliary_loss_clip": 0.01225445, + "auxiliary_loss_mlp": 0.01036759, + "balance_loss_clip": 1.03300273, + "balance_loss_mlp": 1.02421808, + "epoch": 0.08633439547886732, + "flos": 13589568120960.0, + "grad_norm": 6.482884964551357, + "language_loss": 0.89634806, + "learning_rate": 3.966866127062927e-06, + "loss": 0.91897011, + "num_input_tokens_seen": 15150500, + "step": 718, + "time_per_iteration": 2.639976978302002 + }, + { + "auxiliary_loss_clip": 0.01133008, + "auxiliary_loss_mlp": 0.01004119, + "balance_loss_clip": 1.01040721, + "balance_loss_mlp": 0.99985093, + "epoch": 0.0864546383695064, + "flos": 57767342434560.0, + "grad_norm": 0.9445005007085819, + "language_loss": 0.62654662, + "learning_rate": 3.966724772407982e-06, + "loss": 0.64791787, + "num_input_tokens_seen": 15208015, + "step": 719, + "time_per_iteration": 3.0864450931549072 + }, + { + "auxiliary_loss_clip": 0.01169886, + "auxiliary_loss_mlp": 0.01053273, + "balance_loss_clip": 0.95308667, + "balance_loss_mlp": 1.04093516, + "epoch": 0.0865748812601455, + "flos": 20046952753920.0, + "grad_norm": 2.06564484064913, + "language_loss": 0.88770598, + "learning_rate": 3.966583119402454e-06, + "loss": 0.90993756, + "num_input_tokens_seen": 15224780, + "step": 720, + "time_per_iteration": 2.6894216537475586 + }, + { + "auxiliary_loss_clip": 0.0122392, + "auxiliary_loss_mlp": 0.00514699, + "balance_loss_clip": 1.03153133, + "balance_loss_mlp": 0.0, + "epoch": 0.08669512415078459, + "flos": 35262446935680.0, + "grad_norm": 2.0747121758894553, + "language_loss": 0.81902623, + "learning_rate": 3.9664411680678305e-06, + "loss": 0.83641243, + "num_input_tokens_seen": 15246535, + "step": 721, + "time_per_iteration": 2.831859827041626 + }, + { + "auxiliary_loss_clip": 0.01091218, + "auxiliary_loss_mlp": 0.01006276, + "balance_loss_clip": 0.9316836, + "balance_loss_mlp": 1.001984, + "epoch": 0.08681536704142367, + "flos": 65654870048640.0, + "grad_norm": 0.8447866422159103, + "language_loss": 0.61445677, + "learning_rate": 3.966298918425644e-06, + "loss": 0.63543165, + "num_input_tokens_seen": 15304025, + "step": 722, + "time_per_iteration": 4.899695634841919 + }, + { + "auxiliary_loss_clip": 0.01229749, + "auxiliary_loss_mlp": 0.01040929, + "balance_loss_clip": 1.02979112, + "balance_loss_mlp": 1.02854371, + "epoch": 0.08693560993206277, + "flos": 34529940881280.0, + "grad_norm": 1.877106334537915, + "language_loss": 0.82650042, + "learning_rate": 3.966156370497476e-06, + "loss": 0.84920716, + "num_input_tokens_seen": 15327635, + "step": 723, + "time_per_iteration": 2.760732889175415 + }, + { + "auxiliary_loss_clip": 0.01226998, + "auxiliary_loss_mlp": 0.01049196, + "balance_loss_clip": 1.02910042, + "balance_loss_mlp": 1.03630304, + "epoch": 0.08705585282270185, + "flos": 23149419189120.0, + "grad_norm": 1.7388000823465188, + "language_loss": 0.88824022, + "learning_rate": 3.96601352430495e-06, + "loss": 0.91100216, + "num_input_tokens_seen": 15347405, + "step": 724, + "time_per_iteration": 4.50099515914917 + }, + { + "auxiliary_loss_clip": 0.01201217, + "auxiliary_loss_mlp": 0.01042141, + "balance_loss_clip": 0.99612296, + "balance_loss_mlp": 1.02946913, + "epoch": 0.08717609571334095, + "flos": 29497599498240.0, + "grad_norm": 1.4890472158780617, + "language_loss": 0.83020902, + "learning_rate": 3.965870379869735e-06, + "loss": 0.85264266, + "num_input_tokens_seen": 15369450, + "step": 725, + "time_per_iteration": 2.8430957794189453 + }, + { + "auxiliary_loss_clip": 0.01223391, + "auxiliary_loss_mlp": 0.0104538, + "balance_loss_clip": 1.02676153, + "balance_loss_mlp": 1.03287494, + "epoch": 0.08729633860398003, + "flos": 20667489137280.0, + "grad_norm": 1.9378368487280107, + "language_loss": 0.86768961, + "learning_rate": 3.965726937213547e-06, + "loss": 0.89037734, + "num_input_tokens_seen": 15388085, + "step": 726, + "time_per_iteration": 2.6502842903137207 + }, + { + "auxiliary_loss_clip": 0.01216666, + "auxiliary_loss_mlp": 0.01037026, + "balance_loss_clip": 1.02515662, + "balance_loss_mlp": 1.02406216, + "epoch": 0.08741658149461913, + "flos": 18369493655040.0, + "grad_norm": 2.0940984544566583, + "language_loss": 0.81391323, + "learning_rate": 3.965583196358144e-06, + "loss": 0.83645016, + "num_input_tokens_seen": 15407120, + "step": 727, + "time_per_iteration": 2.661538600921631 + }, + { + "auxiliary_loss_clip": 0.01249522, + "auxiliary_loss_mlp": 0.01039484, + "balance_loss_clip": 1.06827712, + "balance_loss_mlp": 1.02696705, + "epoch": 0.08753682438525823, + "flos": 18729677283840.0, + "grad_norm": 4.159390780376441, + "language_loss": 0.74565852, + "learning_rate": 3.965439157325335e-06, + "loss": 0.76854849, + "num_input_tokens_seen": 15424485, + "step": 728, + "time_per_iteration": 2.6246228218078613 + }, + { + "auxiliary_loss_clip": 0.01190846, + "auxiliary_loss_mlp": 0.01040119, + "balance_loss_clip": 0.98539114, + "balance_loss_mlp": 1.02739322, + "epoch": 0.08765706727589731, + "flos": 27776113303680.0, + "grad_norm": 1.845716188049623, + "language_loss": 0.7597664, + "learning_rate": 3.965294820136968e-06, + "loss": 0.782076, + "num_input_tokens_seen": 15446285, + "step": 729, + "time_per_iteration": 2.7858645915985107 + }, + { + "auxiliary_loss_clip": 0.01202909, + "auxiliary_loss_mlp": 0.01042357, + "balance_loss_clip": 0.99256802, + "balance_loss_mlp": 1.02892232, + "epoch": 0.08777731016653641, + "flos": 24389127239040.0, + "grad_norm": 2.0702497292353153, + "language_loss": 0.86539972, + "learning_rate": 3.965150184814938e-06, + "loss": 0.88785231, + "num_input_tokens_seen": 15465770, + "step": 730, + "time_per_iteration": 2.719048023223877 + }, + { + "auxiliary_loss_clip": 0.01187026, + "auxiliary_loss_mlp": 0.01049682, + "balance_loss_clip": 0.9897632, + "balance_loss_mlp": 1.03698659, + "epoch": 0.08789755305717549, + "flos": 21981855605760.0, + "grad_norm": 2.0803345561720605, + "language_loss": 0.76989031, + "learning_rate": 3.965005251381189e-06, + "loss": 0.79225743, + "num_input_tokens_seen": 15483705, + "step": 731, + "time_per_iteration": 2.721153736114502 + }, + { + "auxiliary_loss_clip": 0.01130079, + "auxiliary_loss_mlp": 0.01005977, + "balance_loss_clip": 1.00211263, + "balance_loss_mlp": 1.00182843, + "epoch": 0.08801779594781459, + "flos": 58360120583040.0, + "grad_norm": 0.8888046167570468, + "language_loss": 0.64582586, + "learning_rate": 3.964860019857705e-06, + "loss": 0.66718638, + "num_input_tokens_seen": 15548620, + "step": 732, + "time_per_iteration": 3.2329049110412598 + }, + { + "auxiliary_loss_clip": 0.0125074, + "auxiliary_loss_mlp": 0.01043196, + "balance_loss_clip": 1.07124281, + "balance_loss_mlp": 1.03114963, + "epoch": 0.08813803883845367, + "flos": 23294785530240.0, + "grad_norm": 1.6835600850404766, + "language_loss": 0.84429353, + "learning_rate": 3.964714490266518e-06, + "loss": 0.86723292, + "num_input_tokens_seen": 15569265, + "step": 733, + "time_per_iteration": 2.6126983165740967 + }, + { + "auxiliary_loss_clip": 0.01124532, + "auxiliary_loss_mlp": 0.01004206, + "balance_loss_clip": 1.00048804, + "balance_loss_mlp": 1.00012958, + "epoch": 0.08825828172909277, + "flos": 63424924882560.0, + "grad_norm": 0.9109610916882894, + "language_loss": 0.64596355, + "learning_rate": 3.964568662629706e-06, + "loss": 0.66725099, + "num_input_tokens_seen": 15630570, + "step": 734, + "time_per_iteration": 3.116433620452881 + }, + { + "auxiliary_loss_clip": 0.0122224, + "auxiliary_loss_mlp": 0.01039387, + "balance_loss_clip": 1.02780628, + "balance_loss_mlp": 1.02629185, + "epoch": 0.08837852461973186, + "flos": 26720986268160.0, + "grad_norm": 2.6316757553923646, + "language_loss": 0.84379566, + "learning_rate": 3.9644225369693895e-06, + "loss": 0.86641192, + "num_input_tokens_seen": 15650870, + "step": 735, + "time_per_iteration": 2.640827178955078 + }, + { + "auxiliary_loss_clip": 0.01256774, + "auxiliary_loss_mlp": 0.01052045, + "balance_loss_clip": 1.07553983, + "balance_loss_mlp": 1.03895628, + "epoch": 0.08849876751037095, + "flos": 27265427688960.0, + "grad_norm": 1.8104523602788334, + "language_loss": 0.86779147, + "learning_rate": 3.964276113307735e-06, + "loss": 0.89087969, + "num_input_tokens_seen": 15670835, + "step": 736, + "time_per_iteration": 2.67972731590271 + }, + { + "auxiliary_loss_clip": 0.01175048, + "auxiliary_loss_mlp": 0.01042243, + "balance_loss_clip": 0.95307803, + "balance_loss_mlp": 1.02765167, + "epoch": 0.08861901040101004, + "flos": 19828759587840.0, + "grad_norm": 1.8804280788891505, + "language_loss": 0.80744189, + "learning_rate": 3.9641293916669574e-06, + "loss": 0.82961482, + "num_input_tokens_seen": 15689795, + "step": 737, + "time_per_iteration": 2.724858283996582 + }, + { + "auxiliary_loss_clip": 0.01169611, + "auxiliary_loss_mlp": 0.01041406, + "balance_loss_clip": 0.95345056, + "balance_loss_mlp": 1.02864516, + "epoch": 0.08873925329164913, + "flos": 23658704173440.0, + "grad_norm": 1.8962726888333463, + "language_loss": 0.82699019, + "learning_rate": 3.9639823720693115e-06, + "loss": 0.84910035, + "num_input_tokens_seen": 15711650, + "step": 738, + "time_per_iteration": 2.778379440307617 + }, + { + "auxiliary_loss_clip": 0.01110366, + "auxiliary_loss_mlp": 0.01012923, + "balance_loss_clip": 0.95405936, + "balance_loss_mlp": 1.00827396, + "epoch": 0.08885949618228822, + "flos": 71831541893760.0, + "grad_norm": 0.8397699344578596, + "language_loss": 0.59984982, + "learning_rate": 3.963835054537102e-06, + "loss": 0.62108266, + "num_input_tokens_seen": 15780615, + "step": 739, + "time_per_iteration": 3.3767447471618652 + }, + { + "auxiliary_loss_clip": 0.01194429, + "auxiliary_loss_mlp": 0.0105386, + "balance_loss_clip": 0.98794854, + "balance_loss_mlp": 1.04089594, + "epoch": 0.08897973907292732, + "flos": 22346169298560.0, + "grad_norm": 1.9192641963371218, + "language_loss": 0.61127591, + "learning_rate": 3.963687439092676e-06, + "loss": 0.63375878, + "num_input_tokens_seen": 15801300, + "step": 740, + "time_per_iteration": 2.6633524894714355 + }, + { + "auxiliary_loss_clip": 0.0122353, + "auxiliary_loss_mlp": 0.01047332, + "balance_loss_clip": 1.03165305, + "balance_loss_mlp": 1.03431463, + "epoch": 0.0890999819635664, + "flos": 21251827589760.0, + "grad_norm": 3.2540492420296196, + "language_loss": 0.80486906, + "learning_rate": 3.963539525758427e-06, + "loss": 0.82757759, + "num_input_tokens_seen": 15820860, + "step": 741, + "time_per_iteration": 2.655348539352417 + }, + { + "auxiliary_loss_clip": 0.01204112, + "auxiliary_loss_mlp": 0.01040803, + "balance_loss_clip": 0.99405414, + "balance_loss_mlp": 1.02778578, + "epoch": 0.0892202248542055, + "flos": 25370888745600.0, + "grad_norm": 1.7644510725931972, + "language_loss": 0.67809463, + "learning_rate": 3.9633913145567925e-06, + "loss": 0.70054376, + "num_input_tokens_seen": 15841350, + "step": 742, + "time_per_iteration": 2.7120893001556396 + }, + { + "auxiliary_loss_clip": 0.0120172, + "auxiliary_loss_mlp": 0.01041113, + "balance_loss_clip": 0.99248767, + "balance_loss_mlp": 1.02913201, + "epoch": 0.08934046774484458, + "flos": 24457895827200.0, + "grad_norm": 1.844512237162603, + "language_loss": 0.81696713, + "learning_rate": 3.9632428055102575e-06, + "loss": 0.8393954, + "num_input_tokens_seen": 15861360, + "step": 743, + "time_per_iteration": 2.75834584236145 + }, + { + "auxiliary_loss_clip": 0.01230111, + "auxiliary_loss_mlp": 0.01046457, + "balance_loss_clip": 1.03481352, + "balance_loss_mlp": 1.03336191, + "epoch": 0.08946071063548368, + "flos": 35772773414400.0, + "grad_norm": 2.1828823468609553, + "language_loss": 0.66886783, + "learning_rate": 3.9630939986413495e-06, + "loss": 0.69163346, + "num_input_tokens_seen": 15883160, + "step": 744, + "time_per_iteration": 2.7359042167663574 + }, + { + "auxiliary_loss_clip": 0.01160178, + "auxiliary_loss_mlp": 0.01049628, + "balance_loss_clip": 0.95102108, + "balance_loss_mlp": 1.03687894, + "epoch": 0.08958095352612276, + "flos": 14356584167040.0, + "grad_norm": 1.8595517201761722, + "language_loss": 0.78180707, + "learning_rate": 3.962944893972643e-06, + "loss": 0.80390513, + "num_input_tokens_seen": 15901610, + "step": 745, + "time_per_iteration": 2.6809487342834473 + }, + { + "auxiliary_loss_clip": 0.01202402, + "auxiliary_loss_mlp": 0.01046229, + "balance_loss_clip": 0.99473035, + "balance_loss_mlp": 1.03322911, + "epoch": 0.08970119641676186, + "flos": 17853277345920.0, + "grad_norm": 3.7526860073957953, + "language_loss": 0.91184539, + "learning_rate": 3.962795491526756e-06, + "loss": 0.93433166, + "num_input_tokens_seen": 15918770, + "step": 746, + "time_per_iteration": 2.621166229248047 + }, + { + "auxiliary_loss_clip": 0.01255969, + "auxiliary_loss_mlp": 0.01046023, + "balance_loss_clip": 1.07211804, + "balance_loss_mlp": 1.03262949, + "epoch": 0.08982143930740095, + "flos": 20811670329600.0, + "grad_norm": 2.498580176044392, + "language_loss": 0.89462012, + "learning_rate": 3.962645791326354e-06, + "loss": 0.91764003, + "num_input_tokens_seen": 15938025, + "step": 747, + "time_per_iteration": 3.5074641704559326 + }, + { + "auxiliary_loss_clip": 0.01222929, + "auxiliary_loss_mlp": 0.010377, + "balance_loss_clip": 1.03344214, + "balance_loss_mlp": 1.0256896, + "epoch": 0.08994168219804004, + "flos": 24097712198400.0, + "grad_norm": 1.8899569177245177, + "language_loss": 0.83208454, + "learning_rate": 3.962495793394146e-06, + "loss": 0.85469079, + "num_input_tokens_seen": 15957215, + "step": 748, + "time_per_iteration": 3.621983289718628 + }, + { + "auxiliary_loss_clip": 0.01142259, + "auxiliary_loss_mlp": 0.01003298, + "balance_loss_clip": 1.03632712, + "balance_loss_mlp": 0.99929219, + "epoch": 0.09006192508867913, + "flos": 57188893812480.0, + "grad_norm": 0.7392641413770269, + "language_loss": 0.61209714, + "learning_rate": 3.9623454977528864e-06, + "loss": 0.63355267, + "num_input_tokens_seen": 16015870, + "step": 749, + "time_per_iteration": 3.0549957752227783 + }, + { + "auxiliary_loss_clip": 0.01176325, + "auxiliary_loss_mlp": 0.01052339, + "balance_loss_clip": 0.95434654, + "balance_loss_mlp": 1.03951192, + "epoch": 0.09018216797931822, + "flos": 20487505063680.0, + "grad_norm": 1.6282096239710275, + "language_loss": 0.84935975, + "learning_rate": 3.962194904425375e-06, + "loss": 0.8716464, + "num_input_tokens_seen": 16036500, + "step": 750, + "time_per_iteration": 4.52269434928894 + }, + { + "auxiliary_loss_clip": 0.01224585, + "auxiliary_loss_mlp": 0.01037966, + "balance_loss_clip": 1.03318596, + "balance_loss_mlp": 1.02581251, + "epoch": 0.09030241086995731, + "flos": 22638123043200.0, + "grad_norm": 2.1455293925336276, + "language_loss": 0.67879558, + "learning_rate": 3.9620440134344566e-06, + "loss": 0.70142114, + "num_input_tokens_seen": 16054655, + "step": 751, + "time_per_iteration": 2.6673903465270996 + }, + { + "auxiliary_loss_clip": 0.01177548, + "auxiliary_loss_mlp": 0.01041872, + "balance_loss_clip": 0.95661074, + "balance_loss_mlp": 1.02896142, + "epoch": 0.09042265376059641, + "flos": 21871502046720.0, + "grad_norm": 2.179357333142181, + "language_loss": 0.82288319, + "learning_rate": 3.9618928248030215e-06, + "loss": 0.8450774, + "num_input_tokens_seen": 16074165, + "step": 752, + "time_per_iteration": 2.7864975929260254 + }, + { + "auxiliary_loss_clip": 0.01223305, + "auxiliary_loss_mlp": 0.01037451, + "balance_loss_clip": 1.03256655, + "balance_loss_mlp": 1.02513087, + "epoch": 0.0905428966512355, + "flos": 24316192673280.0, + "grad_norm": 2.184329854329626, + "language_loss": 0.82728577, + "learning_rate": 3.961741338554005e-06, + "loss": 0.84989333, + "num_input_tokens_seen": 16092505, + "step": 753, + "time_per_iteration": 2.7250914573669434 + }, + { + "auxiliary_loss_clip": 0.01206916, + "auxiliary_loss_mlp": 0.01041034, + "balance_loss_clip": 0.9941498, + "balance_loss_mlp": 1.02753353, + "epoch": 0.09066313954187459, + "flos": 35845061535360.0, + "grad_norm": 1.765804965822078, + "language_loss": 0.75917912, + "learning_rate": 3.9615895547103865e-06, + "loss": 0.78165865, + "num_input_tokens_seen": 16116150, + "step": 754, + "time_per_iteration": 2.776718854904175 + }, + { + "auxiliary_loss_clip": 0.01199848, + "auxiliary_loss_mlp": 0.01045803, + "balance_loss_clip": 0.99068713, + "balance_loss_mlp": 1.03301811, + "epoch": 0.09078338243251367, + "flos": 29168729550720.0, + "grad_norm": 1.9801902014540669, + "language_loss": 0.78071129, + "learning_rate": 3.961437473295193e-06, + "loss": 0.80316782, + "num_input_tokens_seen": 16136295, + "step": 755, + "time_per_iteration": 2.7596359252929688 + }, + { + "auxiliary_loss_clip": 0.01134067, + "auxiliary_loss_mlp": 0.01047584, + "balance_loss_clip": 0.909033, + "balance_loss_mlp": 1.03465629, + "epoch": 0.09090362532315277, + "flos": 21907699977600.0, + "grad_norm": 2.268644005146031, + "language_loss": 0.72549891, + "learning_rate": 3.961285094331495e-06, + "loss": 0.74731547, + "num_input_tokens_seen": 16154210, + "step": 756, + "time_per_iteration": 2.8036253452301025 + }, + { + "auxiliary_loss_clip": 0.01247157, + "auxiliary_loss_mlp": 0.01049305, + "balance_loss_clip": 1.06918526, + "balance_loss_mlp": 1.03740764, + "epoch": 0.09102386821379185, + "flos": 27344503480320.0, + "grad_norm": 1.730352274665002, + "language_loss": 0.85959572, + "learning_rate": 3.961132417842406e-06, + "loss": 0.88256031, + "num_input_tokens_seen": 16173995, + "step": 757, + "time_per_iteration": 2.675474166870117 + }, + { + "auxiliary_loss_clip": 0.01217147, + "auxiliary_loss_mlp": 0.01041871, + "balance_loss_clip": 1.02987099, + "balance_loss_mlp": 1.02962792, + "epoch": 0.09114411110443095, + "flos": 20813501923200.0, + "grad_norm": 2.531327853973754, + "language_loss": 0.75279903, + "learning_rate": 3.960979443851089e-06, + "loss": 0.77538919, + "num_input_tokens_seen": 16191020, + "step": 758, + "time_per_iteration": 2.7175066471099854 + }, + { + "auxiliary_loss_clip": 0.01198709, + "auxiliary_loss_mlp": 0.01043707, + "balance_loss_clip": 0.99253845, + "balance_loss_mlp": 1.03067148, + "epoch": 0.09126435399507005, + "flos": 26145949438080.0, + "grad_norm": 1.7202370234006918, + "language_loss": 0.79375046, + "learning_rate": 3.96082617238075e-06, + "loss": 0.81617463, + "num_input_tokens_seen": 16213645, + "step": 759, + "time_per_iteration": 2.7252724170684814 + }, + { + "auxiliary_loss_clip": 0.01200417, + "auxiliary_loss_mlp": 0.01036477, + "balance_loss_clip": 0.99190074, + "balance_loss_mlp": 1.02509212, + "epoch": 0.09138459688570913, + "flos": 24388911757440.0, + "grad_norm": 2.4042903272315947, + "language_loss": 0.79974002, + "learning_rate": 3.960672603454639e-06, + "loss": 0.82210898, + "num_input_tokens_seen": 16233625, + "step": 760, + "time_per_iteration": 2.736625909805298 + }, + { + "auxiliary_loss_clip": 0.01216702, + "auxiliary_loss_mlp": 0.01045002, + "balance_loss_clip": 1.02763474, + "balance_loss_mlp": 1.03239572, + "epoch": 0.09150483977634823, + "flos": 21032664756480.0, + "grad_norm": 3.094957880429438, + "language_loss": 0.77019453, + "learning_rate": 3.960518737096054e-06, + "loss": 0.79281163, + "num_input_tokens_seen": 16253255, + "step": 761, + "time_per_iteration": 2.6238152980804443 + }, + { + "auxiliary_loss_clip": 0.01225981, + "auxiliary_loss_mlp": 0.0104272, + "balance_loss_clip": 1.03264689, + "balance_loss_mlp": 1.02939892, + "epoch": 0.09162508266698731, + "flos": 22856998567680.0, + "grad_norm": 2.215582391131215, + "language_loss": 0.73041058, + "learning_rate": 3.960364573328334e-06, + "loss": 0.75309765, + "num_input_tokens_seen": 16272580, + "step": 762, + "time_per_iteration": 2.6687676906585693 + }, + { + "auxiliary_loss_clip": 0.01172839, + "auxiliary_loss_mlp": 0.01041033, + "balance_loss_clip": 0.94934893, + "balance_loss_mlp": 1.02828336, + "epoch": 0.0917453255576264, + "flos": 21724411852800.0, + "grad_norm": 1.9755192301925624, + "language_loss": 0.88762188, + "learning_rate": 3.9602101121748675e-06, + "loss": 0.90976059, + "num_input_tokens_seen": 16293075, + "step": 763, + "time_per_iteration": 2.7147042751312256 + }, + { + "auxiliary_loss_clip": 0.0119773, + "auxiliary_loss_mlp": 0.01045172, + "balance_loss_clip": 0.99287623, + "balance_loss_mlp": 1.03260124, + "epoch": 0.0918655684482655, + "flos": 14609215497600.0, + "grad_norm": 2.1985380657314346, + "language_loss": 0.72664404, + "learning_rate": 3.960055353659085e-06, + "loss": 0.74907303, + "num_input_tokens_seen": 16310185, + "step": 764, + "time_per_iteration": 2.683475971221924 + }, + { + "auxiliary_loss_clip": 0.01175944, + "auxiliary_loss_mlp": 0.01038392, + "balance_loss_clip": 0.95363224, + "balance_loss_mlp": 1.02596426, + "epoch": 0.09198581133890459, + "flos": 23435016226560.0, + "grad_norm": 1.73383436736169, + "language_loss": 0.83988899, + "learning_rate": 3.959900297804465e-06, + "loss": 0.86203241, + "num_input_tokens_seen": 16330355, + "step": 765, + "time_per_iteration": 2.714865207672119 + }, + { + "auxiliary_loss_clip": 0.01188276, + "auxiliary_loss_mlp": 0.01042416, + "balance_loss_clip": 0.99126589, + "balance_loss_mlp": 1.03020859, + "epoch": 0.09210605422954368, + "flos": 16795887753600.0, + "grad_norm": 2.1176648139090544, + "language_loss": 0.77117014, + "learning_rate": 3.9597449446345276e-06, + "loss": 0.79347706, + "num_input_tokens_seen": 16347600, + "step": 766, + "time_per_iteration": 2.6820240020751953 + }, + { + "auxiliary_loss_clip": 0.01184225, + "auxiliary_loss_mlp": 0.01042217, + "balance_loss_clip": 0.98790944, + "balance_loss_mlp": 1.02931857, + "epoch": 0.09222629712018277, + "flos": 22674249146880.0, + "grad_norm": 2.3320019919451105, + "language_loss": 0.83683527, + "learning_rate": 3.95958929417284e-06, + "loss": 0.85909963, + "num_input_tokens_seen": 16365755, + "step": 767, + "time_per_iteration": 2.716149091720581 + }, + { + "auxiliary_loss_clip": 0.01119249, + "auxiliary_loss_mlp": 0.01004031, + "balance_loss_clip": 0.99497426, + "balance_loss_mlp": 1.00012136, + "epoch": 0.09234654001082186, + "flos": 69976756327680.0, + "grad_norm": 0.7345653278593004, + "language_loss": 0.58785307, + "learning_rate": 3.9594333464430145e-06, + "loss": 0.60908592, + "num_input_tokens_seen": 16435245, + "step": 768, + "time_per_iteration": 3.348022699356079 + }, + { + "auxiliary_loss_clip": 0.01086351, + "auxiliary_loss_mlp": 0.010407, + "balance_loss_clip": 0.82981467, + "balance_loss_mlp": 1.02939904, + "epoch": 0.09246678290146094, + "flos": 20011437181440.0, + "grad_norm": 1.8423089386729063, + "language_loss": 0.88661164, + "learning_rate": 3.959277101468709e-06, + "loss": 0.90788209, + "num_input_tokens_seen": 16454795, + "step": 769, + "time_per_iteration": 2.9644784927368164 + }, + { + "auxiliary_loss_clip": 0.01183777, + "auxiliary_loss_mlp": 0.01047416, + "balance_loss_clip": 0.9866482, + "balance_loss_mlp": 1.03550744, + "epoch": 0.09258702579210004, + "flos": 17747448900480.0, + "grad_norm": 13.637918891289262, + "language_loss": 0.78871775, + "learning_rate": 3.959120559273624e-06, + "loss": 0.81102967, + "num_input_tokens_seen": 16472580, + "step": 770, + "time_per_iteration": 3.0596063137054443 + }, + { + "auxiliary_loss_clip": 0.01185394, + "auxiliary_loss_mlp": 0.01046881, + "balance_loss_clip": 0.99014008, + "balance_loss_mlp": 1.03529382, + "epoch": 0.09270726868273914, + "flos": 20886544229760.0, + "grad_norm": 2.1889542322982494, + "language_loss": 0.83822286, + "learning_rate": 3.958963719881509e-06, + "loss": 0.86054558, + "num_input_tokens_seen": 16490670, + "step": 771, + "time_per_iteration": 2.738008975982666 + }, + { + "auxiliary_loss_clip": 0.01224285, + "auxiliary_loss_mlp": 0.01046941, + "balance_loss_clip": 1.0337919, + "balance_loss_mlp": 1.03466797, + "epoch": 0.09282751157337822, + "flos": 17015697031680.0, + "grad_norm": 1.8873646470583294, + "language_loss": 0.93699163, + "learning_rate": 3.958806583316154e-06, + "loss": 0.95970392, + "num_input_tokens_seen": 16508640, + "step": 772, + "time_per_iteration": 2.676502227783203 + }, + { + "auxiliary_loss_clip": 0.01248063, + "auxiliary_loss_mlp": 0.0104738, + "balance_loss_clip": 1.07062757, + "balance_loss_mlp": 1.03463101, + "epoch": 0.09294775446401732, + "flos": 32523647748480.0, + "grad_norm": 2.1301357463556934, + "language_loss": 0.78800535, + "learning_rate": 3.9586491496013985e-06, + "loss": 0.81095982, + "num_input_tokens_seen": 16531035, + "step": 773, + "time_per_iteration": 3.685645818710327 + }, + { + "auxiliary_loss_clip": 0.01227926, + "auxiliary_loss_mlp": 0.01044991, + "balance_loss_clip": 1.03244257, + "balance_loss_mlp": 1.03273654, + "epoch": 0.0930679973546564, + "flos": 18259750627200.0, + "grad_norm": 2.0029759352700824, + "language_loss": 0.82842541, + "learning_rate": 3.958491418761124e-06, + "loss": 0.85115457, + "num_input_tokens_seen": 16548605, + "step": 774, + "time_per_iteration": 3.5444118976593018 + }, + { + "auxiliary_loss_clip": 0.01190739, + "auxiliary_loss_mlp": 0.01042534, + "balance_loss_clip": 0.98462105, + "balance_loss_mlp": 1.02989781, + "epoch": 0.0931882402452955, + "flos": 21099745405440.0, + "grad_norm": 4.93633940569747, + "language_loss": 0.73123932, + "learning_rate": 3.958333390819258e-06, + "loss": 0.75357199, + "num_input_tokens_seen": 16565535, + "step": 775, + "time_per_iteration": 2.7477219104766846 + }, + { + "auxiliary_loss_clip": 0.01247055, + "auxiliary_loss_mlp": 0.0104216, + "balance_loss_clip": 1.06881022, + "balance_loss_mlp": 1.03053117, + "epoch": 0.0933084831359346, + "flos": 24207275658240.0, + "grad_norm": 2.0300028304329225, + "language_loss": 0.80474114, + "learning_rate": 3.9581750657997754e-06, + "loss": 0.82763326, + "num_input_tokens_seen": 16584900, + "step": 776, + "time_per_iteration": 4.456006765365601 + }, + { + "auxiliary_loss_clip": 0.0119352, + "auxiliary_loss_mlp": 0.01042129, + "balance_loss_clip": 0.98744762, + "balance_loss_mlp": 1.03025007, + "epoch": 0.09342872602657368, + "flos": 25480272637440.0, + "grad_norm": 1.8811336912704268, + "language_loss": 0.89514613, + "learning_rate": 3.95801644372669e-06, + "loss": 0.91750264, + "num_input_tokens_seen": 16604805, + "step": 777, + "time_per_iteration": 2.7802557945251465 + }, + { + "auxiliary_loss_clip": 0.01203167, + "auxiliary_loss_mlp": 0.01039447, + "balance_loss_clip": 0.9873172, + "balance_loss_mlp": 1.02717447, + "epoch": 0.09354896891721277, + "flos": 23149060053120.0, + "grad_norm": 1.7872171297491448, + "language_loss": 0.84489262, + "learning_rate": 3.957857524624068e-06, + "loss": 0.86731881, + "num_input_tokens_seen": 16623685, + "step": 778, + "time_per_iteration": 2.665618896484375 + }, + { + "auxiliary_loss_clip": 0.01197305, + "auxiliary_loss_mlp": 0.01038561, + "balance_loss_clip": 0.98852658, + "balance_loss_mlp": 1.02680743, + "epoch": 0.09366921180785186, + "flos": 24279563779200.0, + "grad_norm": 1.7931114437244455, + "language_loss": 0.89413124, + "learning_rate": 3.957698308516016e-06, + "loss": 0.91648984, + "num_input_tokens_seen": 16644985, + "step": 779, + "time_per_iteration": 2.7349157333374023 + }, + { + "auxiliary_loss_clip": 0.01220549, + "auxiliary_loss_mlp": 0.00514456, + "balance_loss_clip": 1.03335142, + "balance_loss_mlp": 0.0, + "epoch": 0.09378945469849095, + "flos": 18730036419840.0, + "grad_norm": 1.8183074102140526, + "language_loss": 0.82323575, + "learning_rate": 3.957538795426688e-06, + "loss": 0.84058577, + "num_input_tokens_seen": 16662410, + "step": 780, + "time_per_iteration": 2.7437655925750732 + }, + { + "auxiliary_loss_clip": 0.01201184, + "auxiliary_loss_mlp": 0.0104204, + "balance_loss_clip": 0.99230343, + "balance_loss_mlp": 1.0295583, + "epoch": 0.09390969758913004, + "flos": 23218834222080.0, + "grad_norm": 2.303762753397111, + "language_loss": 0.77360684, + "learning_rate": 3.9573789853802804e-06, + "loss": 0.7960391, + "num_input_tokens_seen": 16680885, + "step": 781, + "time_per_iteration": 2.711779832839966 + }, + { + "auxiliary_loss_clip": 0.01199445, + "auxiliary_loss_mlp": 0.00513823, + "balance_loss_clip": 0.99512923, + "balance_loss_mlp": 0.0, + "epoch": 0.09402994047976913, + "flos": 19646728439040.0, + "grad_norm": 2.0179963601358724, + "language_loss": 0.74682534, + "learning_rate": 3.957218878401037e-06, + "loss": 0.76395798, + "num_input_tokens_seen": 16699375, + "step": 782, + "time_per_iteration": 2.727668046951294 + }, + { + "auxiliary_loss_clip": 0.01246607, + "auxiliary_loss_mlp": 0.01044765, + "balance_loss_clip": 1.06901038, + "balance_loss_mlp": 1.03122842, + "epoch": 0.09415018337040823, + "flos": 29420463041280.0, + "grad_norm": 1.829374774374955, + "language_loss": 0.89374828, + "learning_rate": 3.957058474513246e-06, + "loss": 0.9166621, + "num_input_tokens_seen": 16719230, + "step": 783, + "time_per_iteration": 2.735694169998169 + }, + { + "auxiliary_loss_clip": 0.01221511, + "auxiliary_loss_mlp": 0.01036807, + "balance_loss_clip": 1.03387284, + "balance_loss_mlp": 1.02541637, + "epoch": 0.09427042626104731, + "flos": 24572092141440.0, + "grad_norm": 2.8655146552402067, + "language_loss": 0.78387344, + "learning_rate": 3.956897773741241e-06, + "loss": 0.80645663, + "num_input_tokens_seen": 16738220, + "step": 784, + "time_per_iteration": 2.7308969497680664 + }, + { + "auxiliary_loss_clip": 0.0118812, + "auxiliary_loss_mlp": 0.0105441, + "balance_loss_clip": 0.9897157, + "balance_loss_mlp": 1.04145765, + "epoch": 0.09439066915168641, + "flos": 26359581576960.0, + "grad_norm": 1.7260262669028967, + "language_loss": 0.7160852, + "learning_rate": 3.956736776109398e-06, + "loss": 0.73851055, + "num_input_tokens_seen": 16759395, + "step": 785, + "time_per_iteration": 2.754260778427124 + }, + { + "auxiliary_loss_clip": 0.01212725, + "auxiliary_loss_mlp": 0.00514446, + "balance_loss_clip": 1.02698183, + "balance_loss_mlp": 0.0, + "epoch": 0.09451091204232549, + "flos": 19427278296960.0, + "grad_norm": 1.780143180877762, + "language_loss": 0.83676445, + "learning_rate": 3.956575481642143e-06, + "loss": 0.85403621, + "num_input_tokens_seen": 16778285, + "step": 786, + "time_per_iteration": 2.7213947772979736 + }, + { + "auxiliary_loss_clip": 0.01142088, + "auxiliary_loss_mlp": 0.01045976, + "balance_loss_clip": 0.90934974, + "balance_loss_mlp": 1.03324461, + "epoch": 0.09463115493296459, + "flos": 25368051571200.0, + "grad_norm": 2.4950692603929228, + "language_loss": 0.7460922, + "learning_rate": 3.956413890363943e-06, + "loss": 0.76797283, + "num_input_tokens_seen": 16795265, + "step": 787, + "time_per_iteration": 2.7932779788970947 + }, + { + "auxiliary_loss_clip": 0.01220523, + "auxiliary_loss_mlp": 0.01048398, + "balance_loss_clip": 1.0299542, + "balance_loss_mlp": 1.03606582, + "epoch": 0.09475139782360369, + "flos": 10123254869760.0, + "grad_norm": 1.9930321651972172, + "language_loss": 0.81471848, + "learning_rate": 3.956252002299312e-06, + "loss": 0.83740771, + "num_input_tokens_seen": 16811165, + "step": 788, + "time_per_iteration": 2.6128592491149902 + }, + { + "auxiliary_loss_clip": 0.01249503, + "auxiliary_loss_mlp": 0.01047111, + "balance_loss_clip": 1.07140422, + "balance_loss_mlp": 1.03549373, + "epoch": 0.09487164071424277, + "flos": 17231088936960.0, + "grad_norm": 1.7966086970744073, + "language_loss": 0.90907216, + "learning_rate": 3.956089817472807e-06, + "loss": 0.93203825, + "num_input_tokens_seen": 16828470, + "step": 789, + "time_per_iteration": 2.6132264137268066 + }, + { + "auxiliary_loss_clip": 0.01200815, + "auxiliary_loss_mlp": 0.01053618, + "balance_loss_clip": 0.99580777, + "balance_loss_mlp": 1.03992689, + "epoch": 0.09499188360488187, + "flos": 30849564528000.0, + "grad_norm": 2.00478976949136, + "language_loss": 0.85456914, + "learning_rate": 3.955927335909032e-06, + "loss": 0.87711346, + "num_input_tokens_seen": 16851680, + "step": 790, + "time_per_iteration": 2.7584176063537598 + }, + { + "auxiliary_loss_clip": 0.01146265, + "auxiliary_loss_mlp": 0.01043481, + "balance_loss_clip": 0.91672873, + "balance_loss_mlp": 1.03201902, + "epoch": 0.09511212649552095, + "flos": 29351694453120.0, + "grad_norm": 2.0747400366053883, + "language_loss": 0.76186788, + "learning_rate": 3.955764557632634e-06, + "loss": 0.78376538, + "num_input_tokens_seen": 16871490, + "step": 791, + "time_per_iteration": 2.757997512817383 + }, + { + "auxiliary_loss_clip": 0.01186813, + "auxiliary_loss_mlp": 0.01040982, + "balance_loss_clip": 0.98744071, + "balance_loss_mlp": 1.02931118, + "epoch": 0.09523236938616005, + "flos": 10378687461120.0, + "grad_norm": 2.212192652032191, + "language_loss": 0.94628567, + "learning_rate": 3.955601482668309e-06, + "loss": 0.96856368, + "num_input_tokens_seen": 16889350, + "step": 792, + "time_per_iteration": 2.6738054752349854 + }, + { + "auxiliary_loss_clip": 0.01140741, + "auxiliary_loss_mlp": 0.01043386, + "balance_loss_clip": 0.90865254, + "balance_loss_mlp": 1.03086305, + "epoch": 0.09535261227679913, + "flos": 19061815368960.0, + "grad_norm": 2.2933474346305873, + "language_loss": 0.88583398, + "learning_rate": 3.955438111040794e-06, + "loss": 0.90767521, + "num_input_tokens_seen": 16907625, + "step": 793, + "time_per_iteration": 2.746819257736206 + }, + { + "auxiliary_loss_clip": 0.01138949, + "auxiliary_loss_mlp": 0.01042424, + "balance_loss_clip": 0.9086628, + "balance_loss_mlp": 1.02990103, + "epoch": 0.09547285516743823, + "flos": 20922993555840.0, + "grad_norm": 2.9809290575076535, + "language_loss": 0.79927784, + "learning_rate": 3.955274442774873e-06, + "loss": 0.82109159, + "num_input_tokens_seen": 16926205, + "step": 794, + "time_per_iteration": 2.693807601928711 + }, + { + "auxiliary_loss_clip": 0.01223275, + "auxiliary_loss_mlp": 0.01052699, + "balance_loss_clip": 1.0310483, + "balance_loss_mlp": 1.0403018, + "epoch": 0.09559309805807732, + "flos": 30154405639680.0, + "grad_norm": 2.050741416773276, + "language_loss": 0.7063579, + "learning_rate": 3.9551104778953725e-06, + "loss": 0.72911769, + "num_input_tokens_seen": 16946500, + "step": 795, + "time_per_iteration": 2.7552437782287598 + }, + { + "auxiliary_loss_clip": 0.01173014, + "auxiliary_loss_mlp": 0.0104626, + "balance_loss_clip": 0.95126659, + "balance_loss_mlp": 1.03371954, + "epoch": 0.0957133409487164, + "flos": 21066743784960.0, + "grad_norm": 1.7907788873946855, + "language_loss": 0.85551459, + "learning_rate": 3.954946216427167e-06, + "loss": 0.87770736, + "num_input_tokens_seen": 16966960, + "step": 796, + "time_per_iteration": 2.7189507484436035 + }, + { + "auxiliary_loss_clip": 0.01074692, + "auxiliary_loss_mlp": 0.01005344, + "balance_loss_clip": 0.9241178, + "balance_loss_mlp": 1.00157666, + "epoch": 0.0958335838393555, + "flos": 71297979315840.0, + "grad_norm": 0.8823380452751771, + "language_loss": 0.61560374, + "learning_rate": 3.954781658395176e-06, + "loss": 0.6364041, + "num_input_tokens_seen": 17023215, + "step": 797, + "time_per_iteration": 3.2597928047180176 + }, + { + "auxiliary_loss_clip": 0.01203965, + "auxiliary_loss_mlp": 0.01042215, + "balance_loss_clip": 0.99287486, + "balance_loss_mlp": 1.02864861, + "epoch": 0.09595382672999458, + "flos": 21872974504320.0, + "grad_norm": 1.7979854432867743, + "language_loss": 0.92443687, + "learning_rate": 3.95461680382436e-06, + "loss": 0.94689864, + "num_input_tokens_seen": 17042140, + "step": 798, + "time_per_iteration": 2.6821975708007812 + }, + { + "auxiliary_loss_clip": 0.0122565, + "auxiliary_loss_mlp": 0.01058838, + "balance_loss_clip": 1.03390861, + "balance_loss_mlp": 1.04497373, + "epoch": 0.09607406962063368, + "flos": 18695562341760.0, + "grad_norm": 2.680720028195014, + "language_loss": 0.86101079, + "learning_rate": 3.9544516527397295e-06, + "loss": 0.88385564, + "num_input_tokens_seen": 17058490, + "step": 799, + "time_per_iteration": 4.488536596298218 + }, + { + "auxiliary_loss_clip": 0.01185985, + "auxiliary_loss_mlp": 0.01037258, + "balance_loss_clip": 0.99076903, + "balance_loss_mlp": 1.02432346, + "epoch": 0.09619431251127276, + "flos": 22568456615040.0, + "grad_norm": 1.864972825318214, + "language_loss": 0.80459368, + "learning_rate": 3.954286205166338e-06, + "loss": 0.8268261, + "num_input_tokens_seen": 17079655, + "step": 800, + "time_per_iteration": 2.736246347427368 + }, + { + "auxiliary_loss_clip": 0.01227468, + "auxiliary_loss_mlp": 0.01052333, + "balance_loss_clip": 1.0353353, + "balance_loss_mlp": 1.03962517, + "epoch": 0.09631455540191186, + "flos": 14246230608000.0, + "grad_norm": 3.7388279713840507, + "language_loss": 0.84146142, + "learning_rate": 3.954120461129282e-06, + "loss": 0.86425936, + "num_input_tokens_seen": 17097065, + "step": 801, + "time_per_iteration": 2.6375057697296143 + }, + { + "auxiliary_loss_clip": 0.01252124, + "auxiliary_loss_mlp": 0.01041002, + "balance_loss_clip": 1.07436037, + "balance_loss_mlp": 1.02871132, + "epoch": 0.09643479829255096, + "flos": 20740387789440.0, + "grad_norm": 1.977905913737396, + "language_loss": 0.835823, + "learning_rate": 3.953954420653706e-06, + "loss": 0.85875422, + "num_input_tokens_seen": 17114090, + "step": 802, + "time_per_iteration": 4.500126600265503 + }, + { + "auxiliary_loss_clip": 0.01225646, + "auxiliary_loss_mlp": 0.01043745, + "balance_loss_clip": 1.03532028, + "balance_loss_mlp": 1.03036928, + "epoch": 0.09655504118319004, + "flos": 24420476833920.0, + "grad_norm": 1.8548068609442068, + "language_loss": 0.88098854, + "learning_rate": 3.953788083764798e-06, + "loss": 0.90368247, + "num_input_tokens_seen": 17133325, + "step": 803, + "time_per_iteration": 2.7346904277801514 + }, + { + "auxiliary_loss_clip": 0.0115109, + "auxiliary_loss_mlp": 0.01050444, + "balance_loss_clip": 0.91733932, + "balance_loss_mlp": 1.03836775, + "epoch": 0.09667528407382914, + "flos": 18441961344000.0, + "grad_norm": 1.8534861865554373, + "language_loss": 0.92018968, + "learning_rate": 3.953621450487792e-06, + "loss": 0.94220507, + "num_input_tokens_seen": 17151945, + "step": 804, + "time_per_iteration": 2.713721513748169 + }, + { + "auxiliary_loss_clip": 0.01148196, + "auxiliary_loss_mlp": 0.01007039, + "balance_loss_clip": 1.04325187, + "balance_loss_mlp": 1.0032717, + "epoch": 0.09679552696446822, + "flos": 70816455544320.0, + "grad_norm": 0.8347935838584212, + "language_loss": 0.61187518, + "learning_rate": 3.953454520847964e-06, + "loss": 0.63342756, + "num_input_tokens_seen": 17216790, + "step": 805, + "time_per_iteration": 3.375805377960205 + }, + { + "auxiliary_loss_clip": 0.01193153, + "auxiliary_loss_mlp": 0.01049438, + "balance_loss_clip": 0.99171221, + "balance_loss_mlp": 1.03554368, + "epoch": 0.09691576985510732, + "flos": 21945514020480.0, + "grad_norm": 2.328939067705633, + "language_loss": 0.73975956, + "learning_rate": 3.9532872948706395e-06, + "loss": 0.76218551, + "num_input_tokens_seen": 17236285, + "step": 806, + "time_per_iteration": 2.6813082695007324 + }, + { + "auxiliary_loss_clip": 0.01201337, + "auxiliary_loss_mlp": 0.01045031, + "balance_loss_clip": 0.99390507, + "balance_loss_mlp": 1.03202581, + "epoch": 0.09703601274574641, + "flos": 17965211103360.0, + "grad_norm": 2.5064591987311156, + "language_loss": 0.82961178, + "learning_rate": 3.9531197725811845e-06, + "loss": 0.85207546, + "num_input_tokens_seen": 17251670, + "step": 807, + "time_per_iteration": 2.6839849948883057 + }, + { + "auxiliary_loss_clip": 0.01248799, + "auxiliary_loss_mlp": 0.01043681, + "balance_loss_clip": 1.07398105, + "balance_loss_mlp": 1.03146791, + "epoch": 0.0971562556363855, + "flos": 22162162901760.0, + "grad_norm": 1.697405483585728, + "language_loss": 0.8778401, + "learning_rate": 3.952951954005013e-06, + "loss": 0.90076494, + "num_input_tokens_seen": 17271355, + "step": 808, + "time_per_iteration": 2.613532543182373 + }, + { + "auxiliary_loss_clip": 0.01193317, + "auxiliary_loss_mlp": 0.01036789, + "balance_loss_clip": 0.98723757, + "balance_loss_mlp": 1.0252018, + "epoch": 0.0972764985270246, + "flos": 25848716394240.0, + "grad_norm": 1.6967722640470684, + "language_loss": 0.849585, + "learning_rate": 3.952783839167584e-06, + "loss": 0.87188607, + "num_input_tokens_seen": 17291400, + "step": 809, + "time_per_iteration": 2.728121042251587 + }, + { + "auxiliary_loss_clip": 0.01219901, + "auxiliary_loss_mlp": 0.01043403, + "balance_loss_clip": 1.03126168, + "balance_loss_mlp": 1.03028369, + "epoch": 0.09739674141766368, + "flos": 20339373375360.0, + "grad_norm": 2.488432954406373, + "language_loss": 0.73908174, + "learning_rate": 3.952615428094398e-06, + "loss": 0.76171482, + "num_input_tokens_seen": 17310920, + "step": 810, + "time_per_iteration": 2.730323076248169 + }, + { + "auxiliary_loss_clip": 0.01134707, + "auxiliary_loss_mlp": 0.01039867, + "balance_loss_clip": 0.90732843, + "balance_loss_mlp": 1.02830327, + "epoch": 0.09751698430830277, + "flos": 15743059188480.0, + "grad_norm": 1.8137055780815887, + "language_loss": 0.73567092, + "learning_rate": 3.952446720811004e-06, + "loss": 0.75741673, + "num_input_tokens_seen": 17329245, + "step": 811, + "time_per_iteration": 2.73087477684021 + }, + { + "auxiliary_loss_clip": 0.01070243, + "auxiliary_loss_mlp": 0.01010479, + "balance_loss_clip": 0.92010623, + "balance_loss_mlp": 1.0064255, + "epoch": 0.09763722719894186, + "flos": 63716806800000.0, + "grad_norm": 0.8405722871542068, + "language_loss": 0.63597286, + "learning_rate": 3.952277717342995e-06, + "loss": 0.65678, + "num_input_tokens_seen": 17395680, + "step": 812, + "time_per_iteration": 3.492605447769165 + }, + { + "auxiliary_loss_clip": 0.01201768, + "auxiliary_loss_mlp": 0.01048059, + "balance_loss_clip": 0.99186826, + "balance_loss_mlp": 1.03461194, + "epoch": 0.09775747008958095, + "flos": 22090916275200.0, + "grad_norm": 1.903062971872368, + "language_loss": 0.85512984, + "learning_rate": 3.952108417716009e-06, + "loss": 0.87762809, + "num_input_tokens_seen": 17415135, + "step": 813, + "time_per_iteration": 2.977365016937256 + }, + { + "auxiliary_loss_clip": 0.01227503, + "auxiliary_loss_mlp": 0.01049014, + "balance_loss_clip": 1.03662038, + "balance_loss_mlp": 1.03625286, + "epoch": 0.09787771298022005, + "flos": 21286050272640.0, + "grad_norm": 1.8300623083944583, + "language_loss": 0.85121483, + "learning_rate": 3.951938821955727e-06, + "loss": 0.87397993, + "num_input_tokens_seen": 17434535, + "step": 814, + "time_per_iteration": 2.7245430946350098 + }, + { + "auxiliary_loss_clip": 0.01192314, + "auxiliary_loss_mlp": 0.01038421, + "balance_loss_clip": 0.99238914, + "balance_loss_mlp": 1.02577293, + "epoch": 0.09799795587085913, + "flos": 22054574689920.0, + "grad_norm": 1.58118817301882, + "language_loss": 0.76691467, + "learning_rate": 3.9517689300878786e-06, + "loss": 0.789222, + "num_input_tokens_seen": 17454270, + "step": 815, + "time_per_iteration": 2.7424967288970947 + }, + { + "auxiliary_loss_clip": 0.01247106, + "auxiliary_loss_mlp": 0.01047429, + "balance_loss_clip": 1.07052422, + "balance_loss_mlp": 1.03484058, + "epoch": 0.09811819876149823, + "flos": 22163743100160.0, + "grad_norm": 1.7490308366707392, + "language_loss": 0.7859965, + "learning_rate": 3.951598742138236e-06, + "loss": 0.80894184, + "num_input_tokens_seen": 17472995, + "step": 816, + "time_per_iteration": 2.6229591369628906 + }, + { + "auxiliary_loss_clip": 0.01200999, + "auxiliary_loss_mlp": 0.01041668, + "balance_loss_clip": 0.9893322, + "balance_loss_mlp": 1.02934718, + "epoch": 0.09823844165213731, + "flos": 22231111057920.0, + "grad_norm": 1.9301882082630786, + "language_loss": 0.79565233, + "learning_rate": 3.951428258132615e-06, + "loss": 0.81807899, + "num_input_tokens_seen": 17491115, + "step": 817, + "time_per_iteration": 2.7443902492523193 + }, + { + "auxiliary_loss_clip": 0.01203075, + "auxiliary_loss_mlp": 0.01041713, + "balance_loss_clip": 0.99622107, + "balance_loss_mlp": 1.02960134, + "epoch": 0.09835868454277641, + "flos": 22487728798080.0, + "grad_norm": 1.7479782659891998, + "language_loss": 0.8437804, + "learning_rate": 3.951257478096879e-06, + "loss": 0.86622834, + "num_input_tokens_seen": 17509480, + "step": 818, + "time_per_iteration": 2.6531786918640137 + }, + { + "auxiliary_loss_clip": 0.0119876, + "auxiliary_loss_mlp": 0.00514899, + "balance_loss_clip": 0.99298501, + "balance_loss_mlp": 0.0, + "epoch": 0.0984789274334155, + "flos": 16362554077440.0, + "grad_norm": 2.4607543097248294, + "language_loss": 0.6834783, + "learning_rate": 3.951086402056936e-06, + "loss": 0.70061493, + "num_input_tokens_seen": 17524080, + "step": 819, + "time_per_iteration": 2.715292453765869 + }, + { + "auxiliary_loss_clip": 0.01095693, + "auxiliary_loss_mlp": 0.00514243, + "balance_loss_clip": 0.8399545, + "balance_loss_mlp": 0.0, + "epoch": 0.09859917032405459, + "flos": 24243545416320.0, + "grad_norm": 1.6448596164070604, + "language_loss": 0.83483756, + "learning_rate": 3.950915030038735e-06, + "loss": 0.85093689, + "num_input_tokens_seen": 17543875, + "step": 820, + "time_per_iteration": 2.8708202838897705 + }, + { + "auxiliary_loss_clip": 0.0122376, + "auxiliary_loss_mlp": 0.01043278, + "balance_loss_clip": 1.03603399, + "balance_loss_mlp": 1.03101695, + "epoch": 0.09871941321469369, + "flos": 17420195064960.0, + "grad_norm": 2.1301819962495334, + "language_loss": 0.83320391, + "learning_rate": 3.9507433620682765e-06, + "loss": 0.8558743, + "num_input_tokens_seen": 17560810, + "step": 821, + "time_per_iteration": 3.004058599472046 + }, + { + "auxiliary_loss_clip": 0.01170495, + "auxiliary_loss_mlp": 0.01048886, + "balance_loss_clip": 0.95071906, + "balance_loss_mlp": 1.03607154, + "epoch": 0.09883965610533277, + "flos": 28477341590400.0, + "grad_norm": 1.748934880547546, + "language_loss": 0.88168371, + "learning_rate": 3.9505713981716e-06, + "loss": 0.9038775, + "num_input_tokens_seen": 17583640, + "step": 822, + "time_per_iteration": 2.7826874256134033 + }, + { + "auxiliary_loss_clip": 0.01194187, + "auxiliary_loss_mlp": 0.01036895, + "balance_loss_clip": 0.99322844, + "balance_loss_mlp": 1.02596354, + "epoch": 0.09895989899597187, + "flos": 23693932437120.0, + "grad_norm": 1.8887530719297332, + "language_loss": 0.81162453, + "learning_rate": 3.950399138374795e-06, + "loss": 0.83393526, + "num_input_tokens_seen": 17602720, + "step": 823, + "time_per_iteration": 2.6885719299316406 + }, + { + "auxiliary_loss_clip": 0.01216654, + "auxiliary_loss_mlp": 0.0104368, + "balance_loss_clip": 1.02780509, + "balance_loss_mlp": 1.03188992, + "epoch": 0.09908014188661095, + "flos": 24679608526080.0, + "grad_norm": 1.5891917536227678, + "language_loss": 0.74533772, + "learning_rate": 3.95022658270399e-06, + "loss": 0.76794106, + "num_input_tokens_seen": 17623085, + "step": 824, + "time_per_iteration": 2.7098565101623535 + }, + { + "auxiliary_loss_clip": 0.01189785, + "auxiliary_loss_mlp": 0.01043137, + "balance_loss_clip": 0.99219871, + "balance_loss_mlp": 1.03039336, + "epoch": 0.09920038477725004, + "flos": 14064307200000.0, + "grad_norm": 1.7818653195161371, + "language_loss": 0.77858126, + "learning_rate": 3.9500537311853635e-06, + "loss": 0.80091047, + "num_input_tokens_seen": 17641040, + "step": 825, + "time_per_iteration": 4.584862947463989 + }, + { + "auxiliary_loss_clip": 0.01220537, + "auxiliary_loss_mlp": 0.01046555, + "balance_loss_clip": 1.02829409, + "balance_loss_mlp": 1.03400779, + "epoch": 0.09932062766788914, + "flos": 13407070095360.0, + "grad_norm": 2.521031815182255, + "language_loss": 0.83210242, + "learning_rate": 3.949880583845136e-06, + "loss": 0.85477328, + "num_input_tokens_seen": 17659115, + "step": 826, + "time_per_iteration": 2.6594979763031006 + }, + { + "auxiliary_loss_clip": 0.01193322, + "auxiliary_loss_mlp": 0.01040799, + "balance_loss_clip": 0.99212563, + "balance_loss_mlp": 1.02880096, + "epoch": 0.09944087055852822, + "flos": 19500751566720.0, + "grad_norm": 1.6148042177832325, + "language_loss": 0.81201994, + "learning_rate": 3.949707140709575e-06, + "loss": 0.8343612, + "num_input_tokens_seen": 17678845, + "step": 827, + "time_per_iteration": 2.7427988052368164 + }, + { + "auxiliary_loss_clip": 0.0122332, + "auxiliary_loss_mlp": 0.01040535, + "balance_loss_clip": 1.03073525, + "balance_loss_mlp": 1.0285666, + "epoch": 0.09956111344916732, + "flos": 17749100926080.0, + "grad_norm": 2.28697652261087, + "language_loss": 0.83336115, + "learning_rate": 3.949533401804991e-06, + "loss": 0.85599971, + "num_input_tokens_seen": 17695750, + "step": 828, + "time_per_iteration": 3.5777339935302734 + }, + { + "auxiliary_loss_clip": 0.01215988, + "auxiliary_loss_mlp": 0.00514212, + "balance_loss_clip": 1.02870524, + "balance_loss_mlp": 0.0, + "epoch": 0.0996813563398064, + "flos": 17967581400960.0, + "grad_norm": 2.0108512985381664, + "language_loss": 0.90672505, + "learning_rate": 3.949359367157739e-06, + "loss": 0.92402709, + "num_input_tokens_seen": 17714445, + "step": 829, + "time_per_iteration": 3.540496587753296 + }, + { + "auxiliary_loss_clip": 0.01222714, + "auxiliary_loss_mlp": 0.01047261, + "balance_loss_clip": 1.03008294, + "balance_loss_mlp": 1.03408265, + "epoch": 0.0998015992304455, + "flos": 17457039440640.0, + "grad_norm": 2.3403796220614965, + "language_loss": 0.75552857, + "learning_rate": 3.949185036794222e-06, + "loss": 0.77822828, + "num_input_tokens_seen": 17732455, + "step": 830, + "time_per_iteration": 2.6817009449005127 + }, + { + "auxiliary_loss_clip": 0.01242868, + "auxiliary_loss_mlp": 0.01044818, + "balance_loss_clip": 1.07074296, + "balance_loss_mlp": 1.03339195, + "epoch": 0.0999218421210846, + "flos": 25888757080320.0, + "grad_norm": 1.5479544809016026, + "language_loss": 0.78627765, + "learning_rate": 3.949010410740884e-06, + "loss": 0.80915445, + "num_input_tokens_seen": 17755280, + "step": 831, + "time_per_iteration": 2.6796600818634033 + }, + { + "auxiliary_loss_clip": 0.01184515, + "auxiliary_loss_mlp": 0.00514449, + "balance_loss_clip": 0.98917568, + "balance_loss_mlp": 0.0, + "epoch": 0.10004208501172368, + "flos": 21215916967680.0, + "grad_norm": 1.8641290156429187, + "language_loss": 0.8656112, + "learning_rate": 3.948835489024216e-06, + "loss": 0.88260084, + "num_input_tokens_seen": 17775015, + "step": 832, + "time_per_iteration": 2.691951274871826 + }, + { + "auxiliary_loss_clip": 0.01225605, + "auxiliary_loss_mlp": 0.01051011, + "balance_loss_clip": 1.03237534, + "balance_loss_mlp": 1.03928661, + "epoch": 0.10016232790236278, + "flos": 17348409734400.0, + "grad_norm": 2.2516841561771983, + "language_loss": 0.90545678, + "learning_rate": 3.948660271670755e-06, + "loss": 0.92822295, + "num_input_tokens_seen": 17792165, + "step": 833, + "time_per_iteration": 2.63267183303833 + }, + { + "auxiliary_loss_clip": 0.01190512, + "auxiliary_loss_mlp": 0.01044518, + "balance_loss_clip": 0.99069005, + "balance_loss_mlp": 1.0318104, + "epoch": 0.10028257079300186, + "flos": 25666541591040.0, + "grad_norm": 1.9352561335555016, + "language_loss": 0.84240377, + "learning_rate": 3.948484758707079e-06, + "loss": 0.86475402, + "num_input_tokens_seen": 17811765, + "step": 834, + "time_per_iteration": 2.7209160327911377 + }, + { + "auxiliary_loss_clip": 0.01164025, + "auxiliary_loss_mlp": 0.01049837, + "balance_loss_clip": 0.95081633, + "balance_loss_mlp": 1.03807127, + "epoch": 0.10040281368364096, + "flos": 25156035544320.0, + "grad_norm": 2.014627876256335, + "language_loss": 0.83601749, + "learning_rate": 3.948308950159815e-06, + "loss": 0.85815609, + "num_input_tokens_seen": 17830445, + "step": 835, + "time_per_iteration": 2.746211290359497 + }, + { + "auxiliary_loss_clip": 0.0116442, + "auxiliary_loss_mlp": 0.01044156, + "balance_loss_clip": 0.94974136, + "balance_loss_mlp": 1.03123331, + "epoch": 0.10052305657428004, + "flos": 17603303621760.0, + "grad_norm": 2.2798923262059216, + "language_loss": 0.75999498, + "learning_rate": 3.9481328460556326e-06, + "loss": 0.78208065, + "num_input_tokens_seen": 17847665, + "step": 836, + "time_per_iteration": 2.704817533493042 + }, + { + "auxiliary_loss_clip": 0.01187706, + "auxiliary_loss_mlp": 0.0104565, + "balance_loss_clip": 0.99031621, + "balance_loss_mlp": 1.03313267, + "epoch": 0.10064329946491914, + "flos": 18660154510080.0, + "grad_norm": 2.0740709060759612, + "language_loss": 0.8959592, + "learning_rate": 3.9479564464212455e-06, + "loss": 0.91829276, + "num_input_tokens_seen": 17866825, + "step": 837, + "time_per_iteration": 2.661552667617798 + }, + { + "auxiliary_loss_clip": 0.01247478, + "auxiliary_loss_mlp": 0.01037276, + "balance_loss_clip": 1.06859326, + "balance_loss_mlp": 1.02503371, + "epoch": 0.10076354235555823, + "flos": 17199056983680.0, + "grad_norm": 4.925040441363938, + "language_loss": 0.76366603, + "learning_rate": 3.947779751283414e-06, + "loss": 0.78651357, + "num_input_tokens_seen": 17883995, + "step": 838, + "time_per_iteration": 2.6582746505737305 + }, + { + "auxiliary_loss_clip": 0.01221485, + "auxiliary_loss_mlp": 0.00515033, + "balance_loss_clip": 1.03670335, + "balance_loss_mlp": 0.0, + "epoch": 0.10088378524619732, + "flos": 22962252395520.0, + "grad_norm": 2.007031924216348, + "language_loss": 0.76214921, + "learning_rate": 3.947602760668944e-06, + "loss": 0.77951431, + "num_input_tokens_seen": 17903785, + "step": 839, + "time_per_iteration": 2.6628613471984863 + }, + { + "auxiliary_loss_clip": 0.01223417, + "auxiliary_loss_mlp": 0.01047183, + "balance_loss_clip": 1.03406775, + "balance_loss_mlp": 1.03484464, + "epoch": 0.10100402813683641, + "flos": 37885828746240.0, + "grad_norm": 1.6048752152146384, + "language_loss": 0.7157867, + "learning_rate": 3.947425474604684e-06, + "loss": 0.73849267, + "num_input_tokens_seen": 17927720, + "step": 840, + "time_per_iteration": 2.8269705772399902 + }, + { + "auxiliary_loss_clip": 0.01194146, + "auxiliary_loss_mlp": 0.01043385, + "balance_loss_clip": 0.98997843, + "balance_loss_mlp": 1.03116584, + "epoch": 0.1011242710274755, + "flos": 21543458112000.0, + "grad_norm": 2.3112739467647962, + "language_loss": 0.920385, + "learning_rate": 3.947247893117528e-06, + "loss": 0.94276035, + "num_input_tokens_seen": 17946225, + "step": 841, + "time_per_iteration": 2.708801746368408 + }, + { + "auxiliary_loss_clip": 0.01211537, + "auxiliary_loss_mlp": 0.01043485, + "balance_loss_clip": 1.0266856, + "balance_loss_mlp": 1.03056824, + "epoch": 0.10124451391811459, + "flos": 13621456419840.0, + "grad_norm": 2.746779387901683, + "language_loss": 0.6928339, + "learning_rate": 3.947070016234413e-06, + "loss": 0.71538413, + "num_input_tokens_seen": 17962015, + "step": 842, + "time_per_iteration": 2.677029848098755 + }, + { + "auxiliary_loss_clip": 0.01206246, + "auxiliary_loss_mlp": 0.01040622, + "balance_loss_clip": 0.99472356, + "balance_loss_mlp": 1.027879, + "epoch": 0.10136475680875369, + "flos": 16649228522880.0, + "grad_norm": 3.288133247130031, + "language_loss": 0.7526207, + "learning_rate": 3.946891843982326e-06, + "loss": 0.77508932, + "num_input_tokens_seen": 17979680, + "step": 843, + "time_per_iteration": 2.719975471496582 + }, + { + "auxiliary_loss_clip": 0.01222795, + "auxiliary_loss_mlp": 0.01044102, + "balance_loss_clip": 1.03444886, + "balance_loss_mlp": 1.03073871, + "epoch": 0.10148499969939277, + "flos": 19461034103040.0, + "grad_norm": 2.027102129410919, + "language_loss": 0.74330884, + "learning_rate": 3.9467133763882935e-06, + "loss": 0.7659778, + "num_input_tokens_seen": 17998145, + "step": 844, + "time_per_iteration": 2.691006898880005 + }, + { + "auxiliary_loss_clip": 0.01212914, + "auxiliary_loss_mlp": 0.01041047, + "balance_loss_clip": 1.02968359, + "balance_loss_mlp": 1.0292635, + "epoch": 0.10160524259003187, + "flos": 21104988791040.0, + "grad_norm": 2.0411540843322102, + "language_loss": 0.86155522, + "learning_rate": 3.9465346134793905e-06, + "loss": 0.88409483, + "num_input_tokens_seen": 18017955, + "step": 845, + "time_per_iteration": 2.7188527584075928 + }, + { + "auxiliary_loss_clip": 0.01171537, + "auxiliary_loss_mlp": 0.01055151, + "balance_loss_clip": 0.95507854, + "balance_loss_mlp": 1.0425626, + "epoch": 0.10172548548067095, + "flos": 17712687513600.0, + "grad_norm": 3.7880787497516986, + "language_loss": 0.79502892, + "learning_rate": 3.9463555552827335e-06, + "loss": 0.81729579, + "num_input_tokens_seen": 18035125, + "step": 846, + "time_per_iteration": 2.7330315113067627 + }, + { + "auxiliary_loss_clip": 0.01211384, + "auxiliary_loss_mlp": 0.01038237, + "balance_loss_clip": 1.0287149, + "balance_loss_mlp": 1.02636361, + "epoch": 0.10184572837131005, + "flos": 21104845136640.0, + "grad_norm": 2.8162568514444764, + "language_loss": 0.86494291, + "learning_rate": 3.946176201825487e-06, + "loss": 0.88743913, + "num_input_tokens_seen": 18053160, + "step": 847, + "time_per_iteration": 2.706669330596924 + }, + { + "auxiliary_loss_clip": 0.01197184, + "auxiliary_loss_mlp": 0.01044928, + "balance_loss_clip": 0.99529237, + "balance_loss_mlp": 1.0321548, + "epoch": 0.10196597126194913, + "flos": 26067591918720.0, + "grad_norm": 1.8937205114012081, + "language_loss": 0.83545649, + "learning_rate": 3.9459965531348575e-06, + "loss": 0.85787761, + "num_input_tokens_seen": 18072815, + "step": 848, + "time_per_iteration": 2.7241992950439453 + }, + { + "auxiliary_loss_clip": 0.01198511, + "auxiliary_loss_mlp": 0.00514512, + "balance_loss_clip": 0.99478149, + "balance_loss_mlp": 0.0, + "epoch": 0.10208621415258823, + "flos": 29314634595840.0, + "grad_norm": 1.9445636809040694, + "language_loss": 0.85880727, + "learning_rate": 3.945816609238098e-06, + "loss": 0.87593746, + "num_input_tokens_seen": 18092225, + "step": 849, + "time_per_iteration": 2.7483694553375244 + }, + { + "auxiliary_loss_clip": 0.01141591, + "auxiliary_loss_mlp": 0.01042858, + "balance_loss_clip": 0.91559762, + "balance_loss_mlp": 1.03031075, + "epoch": 0.10220645704322733, + "flos": 23805794367360.0, + "grad_norm": 1.8733430842237413, + "language_loss": 0.85234296, + "learning_rate": 3.945636370162507e-06, + "loss": 0.87418747, + "num_input_tokens_seen": 18112335, + "step": 850, + "time_per_iteration": 2.7856504917144775 + }, + { + "auxiliary_loss_clip": 0.01218116, + "auxiliary_loss_mlp": 0.01039428, + "balance_loss_clip": 1.03085256, + "balance_loss_mlp": 1.02782917, + "epoch": 0.10232669993386641, + "flos": 23218546913280.0, + "grad_norm": 1.864108312504757, + "language_loss": 0.79037714, + "learning_rate": 3.945455835935425e-06, + "loss": 0.81295264, + "num_input_tokens_seen": 18131520, + "step": 851, + "time_per_iteration": 4.598278760910034 + }, + { + "auxiliary_loss_clip": 0.01196471, + "auxiliary_loss_mlp": 0.01043026, + "balance_loss_clip": 0.99167383, + "balance_loss_mlp": 1.03008008, + "epoch": 0.1024469428245055, + "flos": 22922929981440.0, + "grad_norm": 2.4053808464845314, + "language_loss": 0.7522217, + "learning_rate": 3.94527500658424e-06, + "loss": 0.7746166, + "num_input_tokens_seen": 18149185, + "step": 852, + "time_per_iteration": 2.6844115257263184 + }, + { + "auxiliary_loss_clip": 0.01146066, + "auxiliary_loss_mlp": 0.01042411, + "balance_loss_clip": 0.91627377, + "balance_loss_mlp": 1.03102064, + "epoch": 0.10256718571514459, + "flos": 31359495957120.0, + "grad_norm": 1.791988683723413, + "language_loss": 0.81201303, + "learning_rate": 3.945093882136382e-06, + "loss": 0.83389777, + "num_input_tokens_seen": 18172960, + "step": 853, + "time_per_iteration": 3.7800168991088867 + }, + { + "auxiliary_loss_clip": 0.01201376, + "auxiliary_loss_mlp": 0.00514383, + "balance_loss_clip": 0.99766195, + "balance_loss_mlp": 0.0, + "epoch": 0.10268742860578368, + "flos": 23474877344640.0, + "grad_norm": 1.8742884394793744, + "language_loss": 0.84605813, + "learning_rate": 3.944912462619329e-06, + "loss": 0.86321574, + "num_input_tokens_seen": 18191925, + "step": 854, + "time_per_iteration": 3.6437478065490723 + }, + { + "auxiliary_loss_clip": 0.01199519, + "auxiliary_loss_mlp": 0.01048283, + "balance_loss_clip": 0.9928146, + "balance_loss_mlp": 1.03453851, + "epoch": 0.10280767149642277, + "flos": 25520313323520.0, + "grad_norm": 4.056900124954761, + "language_loss": 0.8049674, + "learning_rate": 3.9447307480606025e-06, + "loss": 0.82744539, + "num_input_tokens_seen": 18212010, + "step": 855, + "time_per_iteration": 2.723618268966675 + }, + { + "auxiliary_loss_clip": 0.01188961, + "auxiliary_loss_mlp": 0.01046676, + "balance_loss_clip": 0.99280667, + "balance_loss_mlp": 1.03463554, + "epoch": 0.10292791438706186, + "flos": 17347691462400.0, + "grad_norm": 2.293195731943655, + "language_loss": 0.90100741, + "learning_rate": 3.944548738487767e-06, + "loss": 0.92336375, + "num_input_tokens_seen": 18229525, + "step": 856, + "time_per_iteration": 2.699871301651001 + }, + { + "auxiliary_loss_clip": 0.01253992, + "auxiliary_loss_mlp": 0.01048108, + "balance_loss_clip": 1.07492721, + "balance_loss_mlp": 1.03506041, + "epoch": 0.10304815727770096, + "flos": 27052693390080.0, + "grad_norm": 2.045460808727124, + "language_loss": 0.9084819, + "learning_rate": 3.944366433928434e-06, + "loss": 0.93150288, + "num_input_tokens_seen": 18249505, + "step": 857, + "time_per_iteration": 2.761089324951172 + }, + { + "auxiliary_loss_clip": 0.01190758, + "auxiliary_loss_mlp": 0.01039387, + "balance_loss_clip": 0.99136478, + "balance_loss_mlp": 1.02781153, + "epoch": 0.10316840016834004, + "flos": 22782591544320.0, + "grad_norm": 1.457351551161125, + "language_loss": 0.83573955, + "learning_rate": 3.9441838344102594e-06, + "loss": 0.85804105, + "num_input_tokens_seen": 18269230, + "step": 858, + "time_per_iteration": 2.7597243785858154 + }, + { + "auxiliary_loss_clip": 0.01199743, + "auxiliary_loss_mlp": 0.01047544, + "balance_loss_clip": 0.99501503, + "balance_loss_mlp": 1.03506899, + "epoch": 0.10328864305897914, + "flos": 20704584908160.0, + "grad_norm": 3.2799566136814904, + "language_loss": 0.67490488, + "learning_rate": 3.944000939960943e-06, + "loss": 0.6973778, + "num_input_tokens_seen": 18287955, + "step": 859, + "time_per_iteration": 2.6624231338500977 + }, + { + "auxiliary_loss_clip": 0.01222712, + "auxiliary_loss_mlp": 0.01045959, + "balance_loss_clip": 1.02950835, + "balance_loss_mlp": 1.03402066, + "epoch": 0.10340888594961822, + "flos": 28478814048000.0, + "grad_norm": 1.4981679486304624, + "language_loss": 0.79903936, + "learning_rate": 3.943817750608229e-06, + "loss": 0.82172602, + "num_input_tokens_seen": 18310505, + "step": 860, + "time_per_iteration": 2.754289388656616 + }, + { + "auxiliary_loss_clip": 0.01223982, + "auxiliary_loss_mlp": 0.01043149, + "balance_loss_clip": 1.03415871, + "balance_loss_mlp": 1.0308764, + "epoch": 0.10352912884025732, + "flos": 13370333460480.0, + "grad_norm": 3.1281277151151974, + "language_loss": 0.82044584, + "learning_rate": 3.943634266379908e-06, + "loss": 0.84311718, + "num_input_tokens_seen": 18327400, + "step": 861, + "time_per_iteration": 2.6117873191833496 + }, + { + "auxiliary_loss_clip": 0.01223279, + "auxiliary_loss_mlp": 0.01036786, + "balance_loss_clip": 1.03005481, + "balance_loss_mlp": 1.02489519, + "epoch": 0.10364937173089642, + "flos": 25558558329600.0, + "grad_norm": 1.7920657044425523, + "language_loss": 0.84905541, + "learning_rate": 3.943450487303815e-06, + "loss": 0.87165606, + "num_input_tokens_seen": 18347895, + "step": 862, + "time_per_iteration": 2.7304630279541016 + }, + { + "auxiliary_loss_clip": 0.01220117, + "auxiliary_loss_mlp": 0.01035769, + "balance_loss_clip": 1.03330004, + "balance_loss_mlp": 1.02381861, + "epoch": 0.1037696146215355, + "flos": 21215486004480.0, + "grad_norm": 1.6898703033783873, + "language_loss": 0.85090929, + "learning_rate": 3.943266413407827e-06, + "loss": 0.87346816, + "num_input_tokens_seen": 18367170, + "step": 863, + "time_per_iteration": 2.678588390350342 + }, + { + "auxiliary_loss_clip": 0.01224986, + "auxiliary_loss_mlp": 0.01044249, + "balance_loss_clip": 1.03355336, + "balance_loss_mlp": 1.03290009, + "epoch": 0.1038898575121746, + "flos": 25807382818560.0, + "grad_norm": 1.727651391138946, + "language_loss": 0.85137975, + "learning_rate": 3.94308204471987e-06, + "loss": 0.87407213, + "num_input_tokens_seen": 18386185, + "step": 864, + "time_per_iteration": 2.6666252613067627 + }, + { + "auxiliary_loss_clip": 0.01171411, + "auxiliary_loss_mlp": 0.01045897, + "balance_loss_clip": 0.95434093, + "balance_loss_mlp": 1.03482211, + "epoch": 0.10401010040281368, + "flos": 19062425900160.0, + "grad_norm": 2.427785476640113, + "language_loss": 0.74935132, + "learning_rate": 3.942897381267912e-06, + "loss": 0.77152443, + "num_input_tokens_seen": 18402550, + "step": 865, + "time_per_iteration": 2.6651270389556885 + }, + { + "auxiliary_loss_clip": 0.01224871, + "auxiliary_loss_mlp": 0.01053237, + "balance_loss_clip": 1.03330731, + "balance_loss_mlp": 1.04083896, + "epoch": 0.10413034329345278, + "flos": 16355119962240.0, + "grad_norm": 3.021441378432964, + "language_loss": 0.66383553, + "learning_rate": 3.942712423079965e-06, + "loss": 0.6866166, + "num_input_tokens_seen": 18418940, + "step": 866, + "time_per_iteration": 2.6674118041992188 + }, + { + "auxiliary_loss_clip": 0.01156133, + "auxiliary_loss_mlp": 0.01042596, + "balance_loss_clip": 0.94581687, + "balance_loss_mlp": 1.03035343, + "epoch": 0.10425058618409186, + "flos": 17236511890560.0, + "grad_norm": 2.472317383405467, + "language_loss": 0.90182716, + "learning_rate": 3.942527170184088e-06, + "loss": 0.92381442, + "num_input_tokens_seen": 18435560, + "step": 867, + "time_per_iteration": 2.7058346271514893 + }, + { + "auxiliary_loss_clip": 0.01244363, + "auxiliary_loss_mlp": 0.01038389, + "balance_loss_clip": 1.07081044, + "balance_loss_mlp": 1.02559793, + "epoch": 0.10437082907473096, + "flos": 17967365919360.0, + "grad_norm": 2.858308544078979, + "language_loss": 0.77555788, + "learning_rate": 3.942341622608385e-06, + "loss": 0.79838532, + "num_input_tokens_seen": 18452590, + "step": 868, + "time_per_iteration": 2.669051170349121 + }, + { + "auxiliary_loss_clip": 0.01199753, + "auxiliary_loss_mlp": 0.01043857, + "balance_loss_clip": 0.99748588, + "balance_loss_mlp": 1.03199005, + "epoch": 0.10449107196537005, + "flos": 36283315374720.0, + "grad_norm": 2.502579718721618, + "language_loss": 0.77862084, + "learning_rate": 3.942155780381001e-06, + "loss": 0.80105698, + "num_input_tokens_seen": 18476325, + "step": 869, + "time_per_iteration": 2.849999189376831 + }, + { + "auxiliary_loss_clip": 0.01194471, + "auxiliary_loss_mlp": 0.01042966, + "balance_loss_clip": 0.98997492, + "balance_loss_mlp": 1.0307951, + "epoch": 0.10461131485600914, + "flos": 23802095266560.0, + "grad_norm": 1.737623684450667, + "language_loss": 0.76156121, + "learning_rate": 3.94196964353013e-06, + "loss": 0.78393555, + "num_input_tokens_seen": 18495775, + "step": 870, + "time_per_iteration": 2.6655919551849365 + }, + { + "auxiliary_loss_clip": 0.01192097, + "auxiliary_loss_mlp": 0.0051391, + "balance_loss_clip": 0.98941976, + "balance_loss_mlp": 0.0, + "epoch": 0.10473155774664823, + "flos": 18405476104320.0, + "grad_norm": 2.0207049559968744, + "language_loss": 0.80681872, + "learning_rate": 3.941783212084008e-06, + "loss": 0.82387882, + "num_input_tokens_seen": 18513530, + "step": 871, + "time_per_iteration": 2.71980619430542 + }, + { + "auxiliary_loss_clip": 0.01182328, + "auxiliary_loss_mlp": 0.01051341, + "balance_loss_clip": 0.99188453, + "balance_loss_mlp": 1.03919411, + "epoch": 0.10485180063728732, + "flos": 25592637358080.0, + "grad_norm": 2.7596066478783383, + "language_loss": 0.79134029, + "learning_rate": 3.941596486070916e-06, + "loss": 0.81367695, + "num_input_tokens_seen": 18531575, + "step": 872, + "time_per_iteration": 2.729839563369751 + }, + { + "auxiliary_loss_clip": 0.01142337, + "auxiliary_loss_mlp": 0.01043522, + "balance_loss_clip": 0.91686171, + "balance_loss_mlp": 1.03054595, + "epoch": 0.10497204352792641, + "flos": 27088747666560.0, + "grad_norm": 2.3410344784072206, + "language_loss": 0.58315873, + "learning_rate": 3.941409465519182e-06, + "loss": 0.6050173, + "num_input_tokens_seen": 18552100, + "step": 873, + "time_per_iteration": 2.8667216300964355 + }, + { + "auxiliary_loss_clip": 0.01209724, + "auxiliary_loss_mlp": 0.01043218, + "balance_loss_clip": 1.02865505, + "balance_loss_mlp": 1.03066564, + "epoch": 0.10509228641856551, + "flos": 32858479353600.0, + "grad_norm": 1.6345064256964765, + "language_loss": 0.85346806, + "learning_rate": 3.941222150457176e-06, + "loss": 0.87599748, + "num_input_tokens_seen": 18575355, + "step": 874, + "time_per_iteration": 2.820061683654785 + }, + { + "auxiliary_loss_clip": 0.01223797, + "auxiliary_loss_mlp": 0.01039148, + "balance_loss_clip": 1.03152144, + "balance_loss_mlp": 1.02735233, + "epoch": 0.10521252930920459, + "flos": 14319165173760.0, + "grad_norm": 2.4465327673095207, + "language_loss": 0.71181935, + "learning_rate": 3.941034540913311e-06, + "loss": 0.73444879, + "num_input_tokens_seen": 18592885, + "step": 875, + "time_per_iteration": 2.6907334327697754 + }, + { + "auxiliary_loss_clip": 0.01223114, + "auxiliary_loss_mlp": 0.00514439, + "balance_loss_clip": 1.03474772, + "balance_loss_mlp": 0.0, + "epoch": 0.10533277219984369, + "flos": 21687028773120.0, + "grad_norm": 1.5755175181341499, + "language_loss": 0.82470816, + "learning_rate": 3.940846636916051e-06, + "loss": 0.84208369, + "num_input_tokens_seen": 18612920, + "step": 876, + "time_per_iteration": 2.641879081726074 + }, + { + "auxiliary_loss_clip": 0.01197063, + "auxiliary_loss_mlp": 0.01043573, + "balance_loss_clip": 0.99801546, + "balance_loss_mlp": 1.03204584, + "epoch": 0.10545301509048277, + "flos": 22269787027200.0, + "grad_norm": 2.1026864712055064, + "language_loss": 0.86434674, + "learning_rate": 3.940658438493899e-06, + "loss": 0.88675308, + "num_input_tokens_seen": 18630765, + "step": 877, + "time_per_iteration": 3.6523261070251465 + }, + { + "auxiliary_loss_clip": 0.01245839, + "auxiliary_loss_mlp": 0.01037154, + "balance_loss_clip": 1.06795382, + "balance_loss_mlp": 1.0246551, + "epoch": 0.10557325798112187, + "flos": 22199725549440.0, + "grad_norm": 3.019610413578589, + "language_loss": 0.75802362, + "learning_rate": 3.940469945675405e-06, + "loss": 0.78085351, + "num_input_tokens_seen": 18649150, + "step": 878, + "time_per_iteration": 2.636066198348999 + }, + { + "auxiliary_loss_clip": 0.01125202, + "auxiliary_loss_mlp": 0.01045047, + "balance_loss_clip": 0.90907955, + "balance_loss_mlp": 1.03412771, + "epoch": 0.10569350087176095, + "flos": 25775889569280.0, + "grad_norm": 1.8563933467932983, + "language_loss": 0.91377008, + "learning_rate": 3.940281158489163e-06, + "loss": 0.93547249, + "num_input_tokens_seen": 18668380, + "step": 879, + "time_per_iteration": 3.7501413822174072 + }, + { + "auxiliary_loss_clip": 0.01120325, + "auxiliary_loss_mlp": 0.01044777, + "balance_loss_clip": 0.87174326, + "balance_loss_mlp": 1.03256369, + "epoch": 0.10581374376240005, + "flos": 17311385790720.0, + "grad_norm": 1.706487902643189, + "language_loss": 0.82516623, + "learning_rate": 3.940092076963812e-06, + "loss": 0.84681726, + "num_input_tokens_seen": 18685875, + "step": 880, + "time_per_iteration": 3.6949660778045654 + }, + { + "auxiliary_loss_clip": 0.01194621, + "auxiliary_loss_mlp": 0.01041819, + "balance_loss_clip": 0.99120265, + "balance_loss_mlp": 1.0295465, + "epoch": 0.10593398665303914, + "flos": 34349454017280.0, + "grad_norm": 1.983907372037751, + "language_loss": 0.78580928, + "learning_rate": 3.9399027011280355e-06, + "loss": 0.80817366, + "num_input_tokens_seen": 18707970, + "step": 881, + "time_per_iteration": 2.831028461456299 + }, + { + "auxiliary_loss_clip": 0.01193741, + "auxiliary_loss_mlp": 0.01040732, + "balance_loss_clip": 0.99567997, + "balance_loss_mlp": 1.02825665, + "epoch": 0.10605422954367823, + "flos": 23257977068160.0, + "grad_norm": 2.0912392977331207, + "language_loss": 0.77305472, + "learning_rate": 3.939713031010561e-06, + "loss": 0.79539943, + "num_input_tokens_seen": 18726335, + "step": 882, + "time_per_iteration": 2.6608588695526123 + }, + { + "auxiliary_loss_clip": 0.0116837, + "auxiliary_loss_mlp": 0.01044995, + "balance_loss_clip": 0.95484495, + "balance_loss_mlp": 1.03323483, + "epoch": 0.10617447243431732, + "flos": 22820118278400.0, + "grad_norm": 2.634220943205811, + "language_loss": 0.77617812, + "learning_rate": 3.939523066640163e-06, + "loss": 0.79831171, + "num_input_tokens_seen": 18745230, + "step": 883, + "time_per_iteration": 2.7495150566101074 + }, + { + "auxiliary_loss_clip": 0.01222375, + "auxiliary_loss_mlp": 0.01034627, + "balance_loss_clip": 1.03291011, + "balance_loss_mlp": 1.02274168, + "epoch": 0.10629471532495641, + "flos": 24386577373440.0, + "grad_norm": 1.8977438150680104, + "language_loss": 0.81358939, + "learning_rate": 3.939332808045657e-06, + "loss": 0.83615941, + "num_input_tokens_seen": 18764880, + "step": 884, + "time_per_iteration": 2.6555099487304688 + }, + { + "auxiliary_loss_clip": 0.01172998, + "auxiliary_loss_mlp": 0.01046282, + "balance_loss_clip": 0.9548074, + "balance_loss_mlp": 1.03437304, + "epoch": 0.1064149582155955, + "flos": 21105491581440.0, + "grad_norm": 1.6229735788376562, + "language_loss": 0.84548974, + "learning_rate": 3.939142255255906e-06, + "loss": 0.86768258, + "num_input_tokens_seen": 18785765, + "step": 885, + "time_per_iteration": 2.8233096599578857 + }, + { + "auxiliary_loss_clip": 0.01217901, + "auxiliary_loss_mlp": 0.01046585, + "balance_loss_clip": 1.03080428, + "balance_loss_mlp": 1.03495026, + "epoch": 0.1065352011062346, + "flos": 20702035042560.0, + "grad_norm": 2.426084052300571, + "language_loss": 0.86778355, + "learning_rate": 3.938951408299817e-06, + "loss": 0.89042842, + "num_input_tokens_seen": 18804605, + "step": 886, + "time_per_iteration": 2.6466493606567383 + }, + { + "auxiliary_loss_clip": 0.01068315, + "auxiliary_loss_mlp": 0.01010324, + "balance_loss_clip": 0.90712404, + "balance_loss_mlp": 1.0064615, + "epoch": 0.10665544399687368, + "flos": 62659632689280.0, + "grad_norm": 0.8274474740656013, + "language_loss": 0.54439294, + "learning_rate": 3.938760267206342e-06, + "loss": 0.56517929, + "num_input_tokens_seen": 18866425, + "step": 887, + "time_per_iteration": 3.243614912033081 + }, + { + "auxiliary_loss_clip": 0.01243789, + "auxiliary_loss_mlp": 0.01041851, + "balance_loss_clip": 1.07103074, + "balance_loss_mlp": 1.02939975, + "epoch": 0.10677568688751278, + "flos": 26140382830080.0, + "grad_norm": 2.572841061547569, + "language_loss": 0.78715491, + "learning_rate": 3.938568832004475e-06, + "loss": 0.81001127, + "num_input_tokens_seen": 18885130, + "step": 888, + "time_per_iteration": 2.7183210849761963 + }, + { + "auxiliary_loss_clip": 0.01184557, + "auxiliary_loss_mlp": 0.01033856, + "balance_loss_clip": 0.9880023, + "balance_loss_mlp": 1.02205396, + "epoch": 0.10689592977815186, + "flos": 12786533712000.0, + "grad_norm": 1.8919991013432598, + "language_loss": 0.7497853, + "learning_rate": 3.938377102723257e-06, + "loss": 0.77196944, + "num_input_tokens_seen": 18902265, + "step": 889, + "time_per_iteration": 2.654181718826294 + }, + { + "auxiliary_loss_clip": 0.0112743, + "auxiliary_loss_mlp": 0.01056134, + "balance_loss_clip": 0.90928566, + "balance_loss_mlp": 1.0429728, + "epoch": 0.10701617266879096, + "flos": 22126683242880.0, + "grad_norm": 3.95831292364909, + "language_loss": 0.83455533, + "learning_rate": 3.938185079391774e-06, + "loss": 0.85639095, + "num_input_tokens_seen": 18919310, + "step": 890, + "time_per_iteration": 2.777768611907959 + }, + { + "auxiliary_loss_clip": 0.01246782, + "auxiliary_loss_mlp": 0.01038501, + "balance_loss_clip": 1.07180929, + "balance_loss_mlp": 1.02597237, + "epoch": 0.10713641555943004, + "flos": 19745625559680.0, + "grad_norm": 2.619802745610669, + "language_loss": 1.06007981, + "learning_rate": 3.937992762039157e-06, + "loss": 1.08293259, + "num_input_tokens_seen": 18932635, + "step": 891, + "time_per_iteration": 2.565056085586548 + }, + { + "auxiliary_loss_clip": 0.01219277, + "auxiliary_loss_mlp": 0.01042404, + "balance_loss_clip": 1.03316808, + "balance_loss_mlp": 1.03078675, + "epoch": 0.10725665845006914, + "flos": 23952992302080.0, + "grad_norm": 1.9638697795681281, + "language_loss": 0.8024196, + "learning_rate": 3.937800150694577e-06, + "loss": 0.82503641, + "num_input_tokens_seen": 18953810, + "step": 892, + "time_per_iteration": 2.690608263015747 + }, + { + "auxiliary_loss_clip": 0.01149844, + "auxiliary_loss_mlp": 0.01044069, + "balance_loss_clip": 0.91626728, + "balance_loss_mlp": 1.03149819, + "epoch": 0.10737690134070824, + "flos": 18551704371840.0, + "grad_norm": 2.365723475945343, + "language_loss": 0.76221484, + "learning_rate": 3.937607245387255e-06, + "loss": 0.78415394, + "num_input_tokens_seen": 18973175, + "step": 893, + "time_per_iteration": 2.7158114910125732 + }, + { + "auxiliary_loss_clip": 0.01200748, + "auxiliary_loss_mlp": 0.01045538, + "balance_loss_clip": 0.99155879, + "balance_loss_mlp": 1.033921, + "epoch": 0.10749714423134732, + "flos": 22707609903360.0, + "grad_norm": 2.4100439327172785, + "language_loss": 0.71963215, + "learning_rate": 3.937414046146455e-06, + "loss": 0.74209499, + "num_input_tokens_seen": 18991130, + "step": 894, + "time_per_iteration": 2.7159423828125 + }, + { + "auxiliary_loss_clip": 0.012469, + "auxiliary_loss_mlp": 0.01047659, + "balance_loss_clip": 1.07222271, + "balance_loss_mlp": 1.03562438, + "epoch": 0.10761738712198642, + "flos": 21106066199040.0, + "grad_norm": 2.0037222924883813, + "language_loss": 0.75349915, + "learning_rate": 3.9372205530014845e-06, + "loss": 0.77644467, + "num_input_tokens_seen": 19009610, + "step": 895, + "time_per_iteration": 2.6462368965148926 + }, + { + "auxiliary_loss_clip": 0.01239602, + "auxiliary_loss_mlp": 0.01036555, + "balance_loss_clip": 1.06585097, + "balance_loss_mlp": 1.02415705, + "epoch": 0.1077376300126255, + "flos": 23766723348480.0, + "grad_norm": 2.8570136107071047, + "language_loss": 0.7148391, + "learning_rate": 3.937026765981696e-06, + "loss": 0.73760068, + "num_input_tokens_seen": 19029680, + "step": 896, + "time_per_iteration": 2.6234681606292725 + }, + { + "auxiliary_loss_clip": 0.01172012, + "auxiliary_loss_mlp": 0.01044421, + "balance_loss_clip": 0.95570469, + "balance_loss_mlp": 1.0323807, + "epoch": 0.1078578729032646, + "flos": 20919581763840.0, + "grad_norm": 1.7568371875999587, + "language_loss": 0.79441202, + "learning_rate": 3.936832685116488e-06, + "loss": 0.81657636, + "num_input_tokens_seen": 19047775, + "step": 897, + "time_per_iteration": 2.694279909133911 + }, + { + "auxiliary_loss_clip": 0.01243496, + "auxiliary_loss_mlp": 0.01054262, + "balance_loss_clip": 1.0707196, + "balance_loss_mlp": 1.04079163, + "epoch": 0.10797811579390369, + "flos": 14829886702080.0, + "grad_norm": 2.031868583661819, + "language_loss": 0.89987993, + "learning_rate": 3.936638310435301e-06, + "loss": 0.92285752, + "num_input_tokens_seen": 19065640, + "step": 898, + "time_per_iteration": 2.6395785808563232 + }, + { + "auxiliary_loss_clip": 0.01226059, + "auxiliary_loss_mlp": 0.01045019, + "balance_loss_clip": 1.03355956, + "balance_loss_mlp": 1.03280616, + "epoch": 0.10809835868454278, + "flos": 19536985411200.0, + "grad_norm": 2.0040115236007234, + "language_loss": 0.81511688, + "learning_rate": 3.936443641967623e-06, + "loss": 0.83782768, + "num_input_tokens_seen": 19084470, + "step": 899, + "time_per_iteration": 2.627413749694824 + }, + { + "auxiliary_loss_clip": 0.01199583, + "auxiliary_loss_mlp": 0.01046985, + "balance_loss_clip": 0.99451226, + "balance_loss_mlp": 1.03503382, + "epoch": 0.10821860157518187, + "flos": 18442320480000.0, + "grad_norm": 1.8316907241583322, + "language_loss": 0.83028519, + "learning_rate": 3.936248679742983e-06, + "loss": 0.85275078, + "num_input_tokens_seen": 19102965, + "step": 900, + "time_per_iteration": 2.7003185749053955 + }, + { + "auxiliary_loss_clip": 0.01107, + "auxiliary_loss_mlp": 0.01007782, + "balance_loss_clip": 0.95882273, + "balance_loss_mlp": 1.00396764, + "epoch": 0.10833884446582095, + "flos": 49359468447360.0, + "grad_norm": 1.0519107718345744, + "language_loss": 0.70159841, + "learning_rate": 3.936053423790959e-06, + "loss": 0.72274625, + "num_input_tokens_seen": 19151285, + "step": 901, + "time_per_iteration": 3.0868353843688965 + }, + { + "auxiliary_loss_clip": 0.01245301, + "auxiliary_loss_mlp": 0.01036745, + "balance_loss_clip": 1.0723784, + "balance_loss_mlp": 1.02498496, + "epoch": 0.10845908735646005, + "flos": 20411912891520.0, + "grad_norm": 1.9006169693274357, + "language_loss": 0.77361166, + "learning_rate": 3.935857874141168e-06, + "loss": 0.79643214, + "num_input_tokens_seen": 19170120, + "step": 902, + "time_per_iteration": 2.6565043926239014 + }, + { + "auxiliary_loss_clip": 0.01190129, + "auxiliary_loss_mlp": 0.01051251, + "balance_loss_clip": 0.99359316, + "balance_loss_mlp": 1.03894234, + "epoch": 0.10857933024709913, + "flos": 14027750133120.0, + "grad_norm": 2.3913630223210376, + "language_loss": 0.83748484, + "learning_rate": 3.935662030823279e-06, + "loss": 0.85989863, + "num_input_tokens_seen": 19186305, + "step": 903, + "time_per_iteration": 3.620297908782959 + }, + { + "auxiliary_loss_clip": 0.01223632, + "auxiliary_loss_mlp": 0.01057323, + "balance_loss_clip": 1.03284645, + "balance_loss_mlp": 1.0445323, + "epoch": 0.10869957313773823, + "flos": 13369004657280.0, + "grad_norm": 2.1191900164588358, + "language_loss": 0.71670556, + "learning_rate": 3.935465893866998e-06, + "loss": 0.73951513, + "num_input_tokens_seen": 19204530, + "step": 904, + "time_per_iteration": 3.6115670204162598 + }, + { + "auxiliary_loss_clip": 0.01194103, + "auxiliary_loss_mlp": 0.01039329, + "balance_loss_clip": 0.99429685, + "balance_loss_mlp": 1.02705634, + "epoch": 0.10881981602837733, + "flos": 25807095509760.0, + "grad_norm": 1.934945670255012, + "language_loss": 0.80196458, + "learning_rate": 3.935269463302079e-06, + "loss": 0.82429886, + "num_input_tokens_seen": 19222735, + "step": 905, + "time_per_iteration": 3.6043076515197754 + }, + { + "auxiliary_loss_clip": 0.01228208, + "auxiliary_loss_mlp": 0.0105052, + "balance_loss_clip": 1.0351634, + "balance_loss_mlp": 1.03766322, + "epoch": 0.10894005891901641, + "flos": 20777555387520.0, + "grad_norm": 1.8581628267946984, + "language_loss": 0.76574749, + "learning_rate": 3.935072739158322e-06, + "loss": 0.78853476, + "num_input_tokens_seen": 19242445, + "step": 906, + "time_per_iteration": 3.6915957927703857 + }, + { + "auxiliary_loss_clip": 0.01197144, + "auxiliary_loss_mlp": 0.01041684, + "balance_loss_clip": 0.99249524, + "balance_loss_mlp": 1.0294354, + "epoch": 0.10906030180965551, + "flos": 26649883296000.0, + "grad_norm": 1.6132124769266374, + "language_loss": 0.79933131, + "learning_rate": 3.934875721465569e-06, + "loss": 0.82171953, + "num_input_tokens_seen": 19262865, + "step": 907, + "time_per_iteration": 2.7118279933929443 + }, + { + "auxiliary_loss_clip": 0.01192562, + "auxiliary_loss_mlp": 0.01045526, + "balance_loss_clip": 0.99086231, + "balance_loss_mlp": 1.03181696, + "epoch": 0.10918054470029459, + "flos": 36534402420480.0, + "grad_norm": 2.4563326215782757, + "language_loss": 0.71754426, + "learning_rate": 3.9346784102537076e-06, + "loss": 0.73992515, + "num_input_tokens_seen": 19285000, + "step": 908, + "time_per_iteration": 2.7699482440948486 + }, + { + "auxiliary_loss_clip": 0.01241584, + "auxiliary_loss_mlp": 0.01036856, + "balance_loss_clip": 1.06830895, + "balance_loss_mlp": 1.02479219, + "epoch": 0.10930078759093369, + "flos": 21762549118080.0, + "grad_norm": 1.858357684596508, + "language_loss": 0.78012109, + "learning_rate": 3.934480805552669e-06, + "loss": 0.80290556, + "num_input_tokens_seen": 19306010, + "step": 909, + "time_per_iteration": 2.6462395191192627 + }, + { + "auxiliary_loss_clip": 0.01243004, + "auxiliary_loss_mlp": 0.00514287, + "balance_loss_clip": 1.07010293, + "balance_loss_mlp": 0.0, + "epoch": 0.10942103048157277, + "flos": 22601781457920.0, + "grad_norm": 2.4020042512629605, + "language_loss": 0.87641299, + "learning_rate": 3.93428290739243e-06, + "loss": 0.89398587, + "num_input_tokens_seen": 19325380, + "step": 910, + "time_per_iteration": 2.6017403602600098 + }, + { + "auxiliary_loss_clip": 0.0119751, + "auxiliary_loss_mlp": 0.0104426, + "balance_loss_clip": 0.99362445, + "balance_loss_mlp": 1.03299427, + "epoch": 0.10954127337221187, + "flos": 15045781397760.0, + "grad_norm": 2.4486595504305506, + "language_loss": 0.79818767, + "learning_rate": 3.9340847158030125e-06, + "loss": 0.82060534, + "num_input_tokens_seen": 19338960, + "step": 911, + "time_per_iteration": 2.625487804412842 + }, + { + "auxiliary_loss_clip": 0.01226279, + "auxiliary_loss_mlp": 0.01041787, + "balance_loss_clip": 1.03396082, + "balance_loss_mlp": 1.02977109, + "epoch": 0.10966151626285096, + "flos": 21650974496640.0, + "grad_norm": 4.395427259462161, + "language_loss": 0.7518748, + "learning_rate": 3.9338862308144814e-06, + "loss": 0.77455544, + "num_input_tokens_seen": 19357780, + "step": 912, + "time_per_iteration": 2.7841315269470215 + }, + { + "auxiliary_loss_clip": 0.01245708, + "auxiliary_loss_mlp": 0.01046847, + "balance_loss_clip": 1.07371569, + "balance_loss_mlp": 1.03533721, + "epoch": 0.10978175915349005, + "flos": 20121359777280.0, + "grad_norm": 1.5748870338197822, + "language_loss": 0.84456956, + "learning_rate": 3.933687452456946e-06, + "loss": 0.86749512, + "num_input_tokens_seen": 19377680, + "step": 913, + "time_per_iteration": 2.5972940921783447 + }, + { + "auxiliary_loss_clip": 0.01169161, + "auxiliary_loss_mlp": 0.01038448, + "balance_loss_clip": 0.95260262, + "balance_loss_mlp": 1.02458954, + "epoch": 0.10990200204412914, + "flos": 20412667077120.0, + "grad_norm": 2.0889347853845615, + "language_loss": 0.86394173, + "learning_rate": 3.933488380760562e-06, + "loss": 0.8860178, + "num_input_tokens_seen": 19397040, + "step": 914, + "time_per_iteration": 2.6742286682128906 + }, + { + "auxiliary_loss_clip": 0.01242979, + "auxiliary_loss_mlp": 0.00514831, + "balance_loss_clip": 1.06800461, + "balance_loss_mlp": 0.0, + "epoch": 0.11002224493476823, + "flos": 17530117660800.0, + "grad_norm": 1.8417511780831293, + "language_loss": 0.87036103, + "learning_rate": 3.9332890157555286e-06, + "loss": 0.8879391, + "num_input_tokens_seen": 19413975, + "step": 915, + "time_per_iteration": 2.653903007507324 + }, + { + "auxiliary_loss_clip": 0.01202667, + "auxiliary_loss_mlp": 0.01049148, + "balance_loss_clip": 0.9944703, + "balance_loss_mlp": 1.03699493, + "epoch": 0.11014248782540732, + "flos": 12203093099520.0, + "grad_norm": 2.0694900700171077, + "language_loss": 0.76517951, + "learning_rate": 3.933089357472088e-06, + "loss": 0.78769767, + "num_input_tokens_seen": 19432005, + "step": 916, + "time_per_iteration": 2.6557321548461914 + }, + { + "auxiliary_loss_clip": 0.01243331, + "auxiliary_loss_mlp": 0.0103638, + "balance_loss_clip": 1.07150853, + "balance_loss_mlp": 1.02496028, + "epoch": 0.11026273071604642, + "flos": 22382977760640.0, + "grad_norm": 2.5592199731741307, + "language_loss": 0.86034143, + "learning_rate": 3.932889405940529e-06, + "loss": 0.88313854, + "num_input_tokens_seen": 19450100, + "step": 917, + "time_per_iteration": 2.673710346221924 + }, + { + "auxiliary_loss_clip": 0.01194747, + "auxiliary_loss_mlp": 0.01042197, + "balance_loss_clip": 0.99570495, + "balance_loss_mlp": 1.0313009, + "epoch": 0.1103829736066855, + "flos": 19829046896640.0, + "grad_norm": 2.1096589315008485, + "language_loss": 0.80245614, + "learning_rate": 3.932689161191184e-06, + "loss": 0.82482553, + "num_input_tokens_seen": 19467805, + "step": 918, + "time_per_iteration": 2.6954236030578613 + }, + { + "auxiliary_loss_clip": 0.01218181, + "auxiliary_loss_mlp": 0.01036819, + "balance_loss_clip": 1.03074026, + "balance_loss_mlp": 1.02543426, + "epoch": 0.1105032164973246, + "flos": 22669616292480.0, + "grad_norm": 1.9258235065577245, + "language_loss": 0.88127083, + "learning_rate": 3.93248862325443e-06, + "loss": 0.90382087, + "num_input_tokens_seen": 19486710, + "step": 919, + "time_per_iteration": 2.751465320587158 + }, + { + "auxiliary_loss_clip": 0.01145129, + "auxiliary_loss_mlp": 0.01007364, + "balance_loss_clip": 1.0223372, + "balance_loss_mlp": 1.00304854, + "epoch": 0.11062345938796368, + "flos": 66483507876480.0, + "grad_norm": 0.9359291754405469, + "language_loss": 0.6450386, + "learning_rate": 3.932287792160688e-06, + "loss": 0.66656363, + "num_input_tokens_seen": 19545170, + "step": 920, + "time_per_iteration": 3.108856439590454 + }, + { + "auxiliary_loss_clip": 0.01224399, + "auxiliary_loss_mlp": 0.01046287, + "balance_loss_clip": 1.03122449, + "balance_loss_mlp": 1.03409791, + "epoch": 0.11074370227860278, + "flos": 21907771804800.0, + "grad_norm": 2.414732197004546, + "language_loss": 0.8034243, + "learning_rate": 3.932086667940424e-06, + "loss": 0.82613111, + "num_input_tokens_seen": 19561875, + "step": 921, + "time_per_iteration": 2.6917061805725098 + }, + { + "auxiliary_loss_clip": 0.01217918, + "auxiliary_loss_mlp": 0.00514611, + "balance_loss_clip": 1.03138947, + "balance_loss_mlp": 0.0, + "epoch": 0.11086394516924186, + "flos": 28658115763200.0, + "grad_norm": 1.8009396739854713, + "language_loss": 0.81722414, + "learning_rate": 3.93188525062415e-06, + "loss": 0.83454943, + "num_input_tokens_seen": 19582340, + "step": 922, + "time_per_iteration": 2.72895884513855 + }, + { + "auxiliary_loss_clip": 0.01220044, + "auxiliary_loss_mlp": 0.0105457, + "balance_loss_clip": 1.03111804, + "balance_loss_mlp": 1.04142141, + "epoch": 0.11098418805988096, + "flos": 24535247765760.0, + "grad_norm": 1.7734393431451998, + "language_loss": 0.8591029, + "learning_rate": 3.931683540242418e-06, + "loss": 0.88184905, + "num_input_tokens_seen": 19603405, + "step": 923, + "time_per_iteration": 2.690845489501953 + }, + { + "auxiliary_loss_clip": 0.01211973, + "auxiliary_loss_mlp": 0.01043839, + "balance_loss_clip": 1.02932048, + "balance_loss_mlp": 1.0310359, + "epoch": 0.11110443095052006, + "flos": 22960384888320.0, + "grad_norm": 2.5250812419123925, + "language_loss": 0.90910286, + "learning_rate": 3.9314815368258295e-06, + "loss": 0.93166101, + "num_input_tokens_seen": 19619885, + "step": 924, + "time_per_iteration": 2.7501401901245117 + }, + { + "auxiliary_loss_clip": 0.01224786, + "auxiliary_loss_mlp": 0.01050898, + "balance_loss_clip": 1.03531587, + "balance_loss_mlp": 1.03850615, + "epoch": 0.11122467384115914, + "flos": 18950025265920.0, + "grad_norm": 1.8404669823664617, + "language_loss": 0.78674203, + "learning_rate": 3.9312792404050275e-06, + "loss": 0.80949891, + "num_input_tokens_seen": 19637940, + "step": 925, + "time_per_iteration": 2.6299047470092773 + }, + { + "auxiliary_loss_clip": 0.01241345, + "auxiliary_loss_mlp": 0.01045427, + "balance_loss_clip": 1.07239509, + "balance_loss_mlp": 1.03487659, + "epoch": 0.11134491673179824, + "flos": 25082957324160.0, + "grad_norm": 1.7464291854428706, + "language_loss": 0.77116752, + "learning_rate": 3.9310766510107e-06, + "loss": 0.7940352, + "num_input_tokens_seen": 19657115, + "step": 926, + "time_per_iteration": 2.6472158432006836 + }, + { + "auxiliary_loss_clip": 0.01176257, + "auxiliary_loss_mlp": 0.01052599, + "balance_loss_clip": 0.95468509, + "balance_loss_mlp": 1.04003406, + "epoch": 0.11146515962243732, + "flos": 24499121662080.0, + "grad_norm": 2.003035400580827, + "language_loss": 0.91933537, + "learning_rate": 3.9308737686735806e-06, + "loss": 0.94162393, + "num_input_tokens_seen": 19677075, + "step": 927, + "time_per_iteration": 2.7051100730895996 + }, + { + "auxiliary_loss_clip": 0.01244033, + "auxiliary_loss_mlp": 0.0104574, + "balance_loss_clip": 1.06927323, + "balance_loss_mlp": 1.03319883, + "epoch": 0.11158540251307641, + "flos": 22343763087360.0, + "grad_norm": 2.65824091643498, + "language_loss": 0.82596791, + "learning_rate": 3.9306705934244455e-06, + "loss": 0.84886557, + "num_input_tokens_seen": 19697155, + "step": 928, + "time_per_iteration": 2.6438560485839844 + }, + { + "auxiliary_loss_clip": 0.01180476, + "auxiliary_loss_mlp": 0.01042935, + "balance_loss_clip": 0.98909247, + "balance_loss_mlp": 1.03167605, + "epoch": 0.11170564540371551, + "flos": 19902304684800.0, + "grad_norm": 1.6826529161012485, + "language_loss": 0.88017547, + "learning_rate": 3.930467125294116e-06, + "loss": 0.90240955, + "num_input_tokens_seen": 19716705, + "step": 929, + "time_per_iteration": 2.6604676246643066 + }, + { + "auxiliary_loss_clip": 0.01043012, + "auxiliary_loss_mlp": 0.01010646, + "balance_loss_clip": 0.86236084, + "balance_loss_mlp": 1.00675952, + "epoch": 0.1118258882943546, + "flos": 64586239499520.0, + "grad_norm": 0.9304099077967923, + "language_loss": 0.60473889, + "learning_rate": 3.930263364313458e-06, + "loss": 0.62527543, + "num_input_tokens_seen": 19767275, + "step": 930, + "time_per_iteration": 4.140265226364136 + }, + { + "auxiliary_loss_clip": 0.01167524, + "auxiliary_loss_mlp": 0.01048834, + "balance_loss_clip": 0.95110381, + "balance_loss_mlp": 1.03731298, + "epoch": 0.11194613118499369, + "flos": 17201965985280.0, + "grad_norm": 1.9224562403807464, + "language_loss": 0.83023614, + "learning_rate": 3.930059310513384e-06, + "loss": 0.85239971, + "num_input_tokens_seen": 19786315, + "step": 931, + "time_per_iteration": 3.65390682220459 + }, + { + "auxiliary_loss_clip": 0.01155115, + "auxiliary_loss_mlp": 0.00514764, + "balance_loss_clip": 0.94951916, + "balance_loss_mlp": 0.0, + "epoch": 0.11206637407563277, + "flos": 31863465728640.0, + "grad_norm": 1.760633049954635, + "language_loss": 0.84313542, + "learning_rate": 3.929854963924846e-06, + "loss": 0.85983419, + "num_input_tokens_seen": 19806580, + "step": 932, + "time_per_iteration": 2.8157079219818115 + }, + { + "auxiliary_loss_clip": 0.01163575, + "auxiliary_loss_mlp": 0.01041675, + "balance_loss_clip": 0.94773924, + "balance_loss_mlp": 1.03055894, + "epoch": 0.11218661696627187, + "flos": 21945621761280.0, + "grad_norm": 1.8444150705710305, + "language_loss": 0.77432138, + "learning_rate": 3.929650324578845e-06, + "loss": 0.79637384, + "num_input_tokens_seen": 19826045, + "step": 933, + "time_per_iteration": 3.6234049797058105 + }, + { + "auxiliary_loss_clip": 0.01196807, + "auxiliary_loss_mlp": 0.01048387, + "balance_loss_clip": 0.99390638, + "balance_loss_mlp": 1.03601325, + "epoch": 0.11230685985691095, + "flos": 25878198481920.0, + "grad_norm": 2.5891224649831623, + "language_loss": 0.82088518, + "learning_rate": 3.929445392506423e-06, + "loss": 0.84333706, + "num_input_tokens_seen": 19843985, + "step": 934, + "time_per_iteration": 2.7885332107543945 + }, + { + "auxiliary_loss_clip": 0.01215153, + "auxiliary_loss_mlp": 0.01041493, + "balance_loss_clip": 1.03219652, + "balance_loss_mlp": 1.03013873, + "epoch": 0.11242710274755005, + "flos": 22231506107520.0, + "grad_norm": 1.9581209691302994, + "language_loss": 0.75883865, + "learning_rate": 3.92924016773867e-06, + "loss": 0.78140521, + "num_input_tokens_seen": 19860480, + "step": 935, + "time_per_iteration": 2.640115976333618 + }, + { + "auxiliary_loss_clip": 0.01190934, + "auxiliary_loss_mlp": 0.00514485, + "balance_loss_clip": 0.98992109, + "balance_loss_mlp": 0.0, + "epoch": 0.11254734563818915, + "flos": 17712184723200.0, + "grad_norm": 2.2083000424053796, + "language_loss": 0.7341876, + "learning_rate": 3.9290346503067175e-06, + "loss": 0.7512418, + "num_input_tokens_seen": 19877145, + "step": 936, + "time_per_iteration": 2.6911585330963135 + }, + { + "auxiliary_loss_clip": 0.01219345, + "auxiliary_loss_mlp": 0.01042521, + "balance_loss_clip": 1.02935135, + "balance_loss_mlp": 1.03064179, + "epoch": 0.11266758852882823, + "flos": 54930397334400.0, + "grad_norm": 1.7714308945153805, + "language_loss": 0.79011464, + "learning_rate": 3.9288288402417415e-06, + "loss": 0.81273329, + "num_input_tokens_seen": 19903405, + "step": 937, + "time_per_iteration": 2.9546074867248535 + }, + { + "auxiliary_loss_clip": 0.01218264, + "auxiliary_loss_mlp": 0.01042352, + "balance_loss_clip": 1.03269386, + "balance_loss_mlp": 1.02971566, + "epoch": 0.11278783141946733, + "flos": 18878132194560.0, + "grad_norm": 4.613533469960797, + "language_loss": 0.70677376, + "learning_rate": 3.928622737574964e-06, + "loss": 0.72937989, + "num_input_tokens_seen": 19918740, + "step": 938, + "time_per_iteration": 2.6280784606933594 + }, + { + "auxiliary_loss_clip": 0.01190705, + "auxiliary_loss_mlp": 0.01040795, + "balance_loss_clip": 0.98955935, + "balance_loss_mlp": 1.02858222, + "epoch": 0.11290807431010641, + "flos": 26469252777600.0, + "grad_norm": 1.8560112960796615, + "language_loss": 0.91037083, + "learning_rate": 3.928416342337652e-06, + "loss": 0.93268585, + "num_input_tokens_seen": 19938475, + "step": 939, + "time_per_iteration": 2.648395299911499 + }, + { + "auxiliary_loss_clip": 0.01190898, + "auxiliary_loss_mlp": 0.01039494, + "balance_loss_clip": 0.99084687, + "balance_loss_mlp": 1.02799034, + "epoch": 0.1130283172007455, + "flos": 22710590732160.0, + "grad_norm": 1.85869417822274, + "language_loss": 0.82895738, + "learning_rate": 3.928209654561113e-06, + "loss": 0.85126132, + "num_input_tokens_seen": 19959310, + "step": 940, + "time_per_iteration": 2.5897786617279053 + }, + { + "auxiliary_loss_clip": 0.01186506, + "auxiliary_loss_mlp": 0.01034133, + "balance_loss_clip": 0.9908042, + "balance_loss_mlp": 1.02179503, + "epoch": 0.1131485600913846, + "flos": 23219911630080.0, + "grad_norm": 1.9138616581431394, + "language_loss": 0.81548727, + "learning_rate": 3.928002674276703e-06, + "loss": 0.83769369, + "num_input_tokens_seen": 19978700, + "step": 941, + "time_per_iteration": 2.5909440517425537 + }, + { + "auxiliary_loss_clip": 0.01121853, + "auxiliary_loss_mlp": 0.0104093, + "balance_loss_clip": 0.90564382, + "balance_loss_mlp": 1.02914011, + "epoch": 0.11326880298202369, + "flos": 14064271286400.0, + "grad_norm": 2.370940025029065, + "language_loss": 0.75892508, + "learning_rate": 3.92779540151582e-06, + "loss": 0.78055292, + "num_input_tokens_seen": 19995785, + "step": 942, + "time_per_iteration": 2.5859320163726807 + }, + { + "auxiliary_loss_clip": 0.01195492, + "auxiliary_loss_mlp": 0.01038371, + "balance_loss_clip": 0.9948982, + "balance_loss_mlp": 1.02593184, + "epoch": 0.11338904587266278, + "flos": 16325386479360.0, + "grad_norm": 1.7352185316621032, + "language_loss": 0.85528439, + "learning_rate": 3.927587836309907e-06, + "loss": 0.87762302, + "num_input_tokens_seen": 20013615, + "step": 943, + "time_per_iteration": 2.585164785385132 + }, + { + "auxiliary_loss_clip": 0.01185069, + "auxiliary_loss_mlp": 0.01043804, + "balance_loss_clip": 0.9873566, + "balance_loss_mlp": 1.03053617, + "epoch": 0.11350928876330187, + "flos": 24426258923520.0, + "grad_norm": 1.8281968656335794, + "language_loss": 0.7807315, + "learning_rate": 3.927379978690452e-06, + "loss": 0.80302024, + "num_input_tokens_seen": 20032880, + "step": 944, + "time_per_iteration": 2.730607509613037 + }, + { + "auxiliary_loss_clip": 0.01150962, + "auxiliary_loss_mlp": 0.010389, + "balance_loss_clip": 0.94299626, + "balance_loss_mlp": 1.02717555, + "epoch": 0.11362953165394096, + "flos": 24497074586880.0, + "grad_norm": 1.9169926297653404, + "language_loss": 0.87334871, + "learning_rate": 3.927171828688987e-06, + "loss": 0.89524734, + "num_input_tokens_seen": 20052405, + "step": 945, + "time_per_iteration": 2.745590925216675 + }, + { + "auxiliary_loss_clip": 0.01240155, + "auxiliary_loss_mlp": 0.01041577, + "balance_loss_clip": 1.07004344, + "balance_loss_mlp": 1.02919149, + "epoch": 0.11374977454458005, + "flos": 24060831909120.0, + "grad_norm": 2.814847512620495, + "language_loss": 0.82215881, + "learning_rate": 3.926963386337088e-06, + "loss": 0.84497613, + "num_input_tokens_seen": 20070635, + "step": 946, + "time_per_iteration": 2.664870500564575 + }, + { + "auxiliary_loss_clip": 0.01244586, + "auxiliary_loss_mlp": 0.01036122, + "balance_loss_clip": 1.07053399, + "balance_loss_mlp": 1.02399826, + "epoch": 0.11387001743521914, + "flos": 39457638967680.0, + "grad_norm": 2.9015068794140495, + "language_loss": 0.69815964, + "learning_rate": 3.926754651666375e-06, + "loss": 0.72096676, + "num_input_tokens_seen": 20091195, + "step": 947, + "time_per_iteration": 2.793400287628174 + }, + { + "auxiliary_loss_clip": 0.01165878, + "auxiliary_loss_mlp": 0.01048422, + "balance_loss_clip": 0.95180559, + "balance_loss_mlp": 1.03628051, + "epoch": 0.11399026032585824, + "flos": 25082454533760.0, + "grad_norm": 2.5187570517993225, + "language_loss": 0.78059185, + "learning_rate": 3.926545624708513e-06, + "loss": 0.80273479, + "num_input_tokens_seen": 20110435, + "step": 948, + "time_per_iteration": 2.6947381496429443 + }, + { + "auxiliary_loss_clip": 0.01160566, + "auxiliary_loss_mlp": 0.01045126, + "balance_loss_clip": 0.94928408, + "balance_loss_mlp": 1.03352726, + "epoch": 0.11411050321649732, + "flos": 17961835224960.0, + "grad_norm": 1.8223583996560107, + "language_loss": 0.85867178, + "learning_rate": 3.926336305495213e-06, + "loss": 0.88072872, + "num_input_tokens_seen": 20128995, + "step": 949, + "time_per_iteration": 2.7410833835601807 + }, + { + "auxiliary_loss_clip": 0.01151429, + "auxiliary_loss_mlp": 0.01037104, + "balance_loss_clip": 0.9478398, + "balance_loss_mlp": 1.02496862, + "epoch": 0.11423074610713642, + "flos": 22455409536000.0, + "grad_norm": 1.9821008245298752, + "language_loss": 0.88726872, + "learning_rate": 3.926126694058226e-06, + "loss": 0.909154, + "num_input_tokens_seen": 20148145, + "step": 950, + "time_per_iteration": 2.7045555114746094 + }, + { + "auxiliary_loss_clip": 0.01142811, + "auxiliary_loss_mlp": 0.01048767, + "balance_loss_clip": 0.91493374, + "balance_loss_mlp": 1.03762078, + "epoch": 0.1143509889977755, + "flos": 19717687756800.0, + "grad_norm": 1.4016982118804195, + "language_loss": 0.82118595, + "learning_rate": 3.92591679042935e-06, + "loss": 0.84310174, + "num_input_tokens_seen": 20168035, + "step": 951, + "time_per_iteration": 2.7268967628479004 + }, + { + "auxiliary_loss_clip": 0.01216422, + "auxiliary_loss_mlp": 0.01048591, + "balance_loss_clip": 1.03251076, + "balance_loss_mlp": 1.0359726, + "epoch": 0.1144712318884146, + "flos": 19822869757440.0, + "grad_norm": 1.5860415392175067, + "language_loss": 0.82299781, + "learning_rate": 3.92570659464043e-06, + "loss": 0.84564793, + "num_input_tokens_seen": 20186095, + "step": 952, + "time_per_iteration": 2.6559648513793945 + }, + { + "auxiliary_loss_clip": 0.01210753, + "auxiliary_loss_mlp": 0.00513997, + "balance_loss_clip": 1.02919924, + "balance_loss_mlp": 0.0, + "epoch": 0.1145914747790537, + "flos": 14939198766720.0, + "grad_norm": 1.8234331748710537, + "language_loss": 0.79373193, + "learning_rate": 3.925496106723349e-06, + "loss": 0.81097943, + "num_input_tokens_seen": 20203535, + "step": 953, + "time_per_iteration": 2.6315758228302 + }, + { + "auxiliary_loss_clip": 0.0121806, + "auxiliary_loss_mlp": 0.01043783, + "balance_loss_clip": 1.03025055, + "balance_loss_mlp": 1.03204691, + "epoch": 0.11471171766969278, + "flos": 19865029345920.0, + "grad_norm": 2.4589344234230093, + "language_loss": 0.8363176, + "learning_rate": 3.9252853267100405e-06, + "loss": 0.85893601, + "num_input_tokens_seen": 20222780, + "step": 954, + "time_per_iteration": 2.67392897605896 + }, + { + "auxiliary_loss_clip": 0.01161514, + "auxiliary_loss_mlp": 0.01040473, + "balance_loss_clip": 0.9526577, + "balance_loss_mlp": 1.02836752, + "epoch": 0.11483196056033187, + "flos": 22526476594560.0, + "grad_norm": 2.619451219364845, + "language_loss": 0.8373192, + "learning_rate": 3.9250742546324786e-06, + "loss": 0.85933912, + "num_input_tokens_seen": 20243015, + "step": 955, + "time_per_iteration": 3.700012445449829 + }, + { + "auxiliary_loss_clip": 0.01188912, + "auxiliary_loss_mlp": 0.01049332, + "balance_loss_clip": 0.98862404, + "balance_loss_mlp": 1.03781033, + "epoch": 0.11495220345097096, + "flos": 28220292887040.0, + "grad_norm": 1.6571695565834081, + "language_loss": 0.86712456, + "learning_rate": 3.924862890522683e-06, + "loss": 0.889507, + "num_input_tokens_seen": 20263025, + "step": 956, + "time_per_iteration": 2.723440647125244 + }, + { + "auxiliary_loss_clip": 0.0121471, + "auxiliary_loss_mlp": 0.01053973, + "balance_loss_clip": 1.02662683, + "balance_loss_mlp": 1.04114056, + "epoch": 0.11507244634161005, + "flos": 17492267704320.0, + "grad_norm": 2.1539493629548865, + "language_loss": 0.86097491, + "learning_rate": 3.9246512344127174e-06, + "loss": 0.88366175, + "num_input_tokens_seen": 20280685, + "step": 957, + "time_per_iteration": 3.690387010574341 + }, + { + "auxiliary_loss_clip": 0.01097604, + "auxiliary_loss_mlp": 0.01048373, + "balance_loss_clip": 0.87043554, + "balance_loss_mlp": 1.03650606, + "epoch": 0.11519268923224914, + "flos": 22564937082240.0, + "grad_norm": 1.8599046593446855, + "language_loss": 0.81989378, + "learning_rate": 3.9244392863346895e-06, + "loss": 0.84135354, + "num_input_tokens_seen": 20300090, + "step": 958, + "time_per_iteration": 2.786336898803711 + }, + { + "auxiliary_loss_clip": 0.0119225, + "auxiliary_loss_mlp": 0.01047039, + "balance_loss_clip": 0.99556935, + "balance_loss_mlp": 1.03524959, + "epoch": 0.11531293212288823, + "flos": 16982839065600.0, + "grad_norm": 1.9008936293694492, + "language_loss": 0.92157233, + "learning_rate": 3.9242270463207524e-06, + "loss": 0.9439652, + "num_input_tokens_seen": 20318480, + "step": 959, + "time_per_iteration": 3.6060123443603516 + }, + { + "auxiliary_loss_clip": 0.01135811, + "auxiliary_loss_mlp": 0.01041714, + "balance_loss_clip": 0.91219211, + "balance_loss_mlp": 1.02879739, + "epoch": 0.11543317501352733, + "flos": 12422004537600.0, + "grad_norm": 2.6298562423684655, + "language_loss": 0.84588933, + "learning_rate": 3.924014514403102e-06, + "loss": 0.86766458, + "num_input_tokens_seen": 20334635, + "step": 960, + "time_per_iteration": 2.710197925567627 + }, + { + "auxiliary_loss_clip": 0.01134324, + "auxiliary_loss_mlp": 0.01044825, + "balance_loss_clip": 0.91062653, + "balance_loss_mlp": 1.03291059, + "epoch": 0.11555341790416641, + "flos": 19821648695040.0, + "grad_norm": 1.9827024263473843, + "language_loss": 0.90832031, + "learning_rate": 3.92380169061398e-06, + "loss": 0.93011189, + "num_input_tokens_seen": 20352415, + "step": 961, + "time_per_iteration": 2.7333037853240967 + }, + { + "auxiliary_loss_clip": 0.01156357, + "auxiliary_loss_mlp": 0.00514509, + "balance_loss_clip": 0.94720572, + "balance_loss_mlp": 0.0, + "epoch": 0.11567366079480551, + "flos": 25738865625600.0, + "grad_norm": 1.875494102255943, + "language_loss": 0.83993268, + "learning_rate": 3.9235885749856705e-06, + "loss": 0.85664135, + "num_input_tokens_seen": 20371095, + "step": 962, + "time_per_iteration": 2.765547275543213 + }, + { + "auxiliary_loss_clip": 0.0119227, + "auxiliary_loss_mlp": 0.01044397, + "balance_loss_clip": 0.9931531, + "balance_loss_mlp": 1.03172517, + "epoch": 0.1157939036854446, + "flos": 18223301301120.0, + "grad_norm": 1.8163844810861762, + "language_loss": 0.82583904, + "learning_rate": 3.9233751675505035e-06, + "loss": 0.84820569, + "num_input_tokens_seen": 20389805, + "step": 963, + "time_per_iteration": 2.6450772285461426 + }, + { + "auxiliary_loss_clip": 0.01184127, + "auxiliary_loss_mlp": 0.01040757, + "balance_loss_clip": 0.98952788, + "balance_loss_mlp": 1.02872825, + "epoch": 0.11591414657608369, + "flos": 23073755189760.0, + "grad_norm": 1.9364669739203755, + "language_loss": 0.85105091, + "learning_rate": 3.923161468340853e-06, + "loss": 0.87329972, + "num_input_tokens_seen": 20409640, + "step": 964, + "time_per_iteration": 2.6734907627105713 + }, + { + "auxiliary_loss_clip": 0.01126877, + "auxiliary_loss_mlp": 0.01044339, + "balance_loss_clip": 0.90721357, + "balance_loss_mlp": 1.0317924, + "epoch": 0.11603438946672277, + "flos": 19461716461440.0, + "grad_norm": 1.8340272077550965, + "language_loss": 0.81769013, + "learning_rate": 3.9229474773891374e-06, + "loss": 0.83940226, + "num_input_tokens_seen": 20428180, + "step": 965, + "time_per_iteration": 2.7209601402282715 + }, + { + "auxiliary_loss_clip": 0.01173814, + "auxiliary_loss_mlp": 0.01042259, + "balance_loss_clip": 0.94888556, + "balance_loss_mlp": 1.02853179, + "epoch": 0.11615463235736187, + "flos": 26831986272000.0, + "grad_norm": 1.9776618206159347, + "language_loss": 0.83614707, + "learning_rate": 3.922733194727818e-06, + "loss": 0.85830784, + "num_input_tokens_seen": 20447975, + "step": 966, + "time_per_iteration": 2.8041799068450928 + }, + { + "auxiliary_loss_clip": 0.01219299, + "auxiliary_loss_mlp": 0.01043114, + "balance_loss_clip": 1.03217173, + "balance_loss_mlp": 1.03104377, + "epoch": 0.11627487524800097, + "flos": 18580324533120.0, + "grad_norm": 1.8634985871567529, + "language_loss": 0.87542224, + "learning_rate": 3.922518620389402e-06, + "loss": 0.89804637, + "num_input_tokens_seen": 20464840, + "step": 967, + "time_per_iteration": 2.6888914108276367 + }, + { + "auxiliary_loss_clip": 0.01064811, + "auxiliary_loss_mlp": 0.01046946, + "balance_loss_clip": 0.83112198, + "balance_loss_mlp": 1.03611028, + "epoch": 0.11639511813864005, + "flos": 18150474476160.0, + "grad_norm": 1.739206725089368, + "language_loss": 0.89758062, + "learning_rate": 3.922303754406439e-06, + "loss": 0.91869819, + "num_input_tokens_seen": 20482680, + "step": 968, + "time_per_iteration": 2.917133331298828 + }, + { + "auxiliary_loss_clip": 0.01145012, + "auxiliary_loss_mlp": 0.01041209, + "balance_loss_clip": 0.94420886, + "balance_loss_mlp": 1.02962804, + "epoch": 0.11651536102927915, + "flos": 20922023888640.0, + "grad_norm": 2.801068760898052, + "language_loss": 0.7894249, + "learning_rate": 3.922088596811526e-06, + "loss": 0.81128716, + "num_input_tokens_seen": 20501810, + "step": 969, + "time_per_iteration": 2.893845558166504 + }, + { + "auxiliary_loss_clip": 0.01205803, + "auxiliary_loss_mlp": 0.01040888, + "balance_loss_clip": 1.02752566, + "balance_loss_mlp": 1.0284487, + "epoch": 0.11663560391991823, + "flos": 16508602776960.0, + "grad_norm": 2.290243715006115, + "language_loss": 0.8657304, + "learning_rate": 3.9218731476373e-06, + "loss": 0.8881973, + "num_input_tokens_seen": 20517995, + "step": 970, + "time_per_iteration": 2.6585888862609863 + }, + { + "auxiliary_loss_clip": 0.01218973, + "auxiliary_loss_mlp": 0.0104348, + "balance_loss_clip": 1.03212535, + "balance_loss_mlp": 1.03106427, + "epoch": 0.11675584681055733, + "flos": 19865029345920.0, + "grad_norm": 1.96953548881016, + "language_loss": 0.84579945, + "learning_rate": 3.9216574069164455e-06, + "loss": 0.86842394, + "num_input_tokens_seen": 20536970, + "step": 971, + "time_per_iteration": 2.6378726959228516 + }, + { + "auxiliary_loss_clip": 0.01237529, + "auxiliary_loss_mlp": 0.01038877, + "balance_loss_clip": 1.06822658, + "balance_loss_mlp": 1.02702761, + "epoch": 0.11687608970119642, + "flos": 21944364785280.0, + "grad_norm": 1.4733272194142444, + "language_loss": 0.80077457, + "learning_rate": 3.921441374681691e-06, + "loss": 0.82353866, + "num_input_tokens_seen": 20557030, + "step": 972, + "time_per_iteration": 2.6795482635498047 + }, + { + "auxiliary_loss_clip": 0.01183543, + "auxiliary_loss_mlp": 0.01041925, + "balance_loss_clip": 0.99130952, + "balance_loss_mlp": 1.03102911, + "epoch": 0.1169963325918355, + "flos": 24061155131520.0, + "grad_norm": 1.8588887485444001, + "language_loss": 0.64820319, + "learning_rate": 3.921225050965808e-06, + "loss": 0.67045784, + "num_input_tokens_seen": 20576915, + "step": 973, + "time_per_iteration": 2.7191286087036133 + }, + { + "auxiliary_loss_clip": 0.01168775, + "auxiliary_loss_mlp": 0.01045914, + "balance_loss_clip": 0.95335621, + "balance_loss_mlp": 1.03293777, + "epoch": 0.1171165754824746, + "flos": 23368151059200.0, + "grad_norm": 1.9669914680116087, + "language_loss": 0.75470769, + "learning_rate": 3.921008435801612e-06, + "loss": 0.77685463, + "num_input_tokens_seen": 20596000, + "step": 974, + "time_per_iteration": 2.7374608516693115 + }, + { + "auxiliary_loss_clip": 0.01207924, + "auxiliary_loss_mlp": 0.01038326, + "balance_loss_clip": 1.0291965, + "balance_loss_mlp": 1.02593374, + "epoch": 0.11723681837311369, + "flos": 18552243075840.0, + "grad_norm": 2.7103193224408852, + "language_loss": 0.75953484, + "learning_rate": 3.920791529221963e-06, + "loss": 0.78199732, + "num_input_tokens_seen": 20614675, + "step": 975, + "time_per_iteration": 2.6532230377197266 + }, + { + "auxiliary_loss_clip": 0.01190977, + "auxiliary_loss_mlp": 0.00514148, + "balance_loss_clip": 0.9888379, + "balance_loss_mlp": 0.0, + "epoch": 0.11735706126375278, + "flos": 23550541344000.0, + "grad_norm": 1.8086876962766136, + "language_loss": 0.7640326, + "learning_rate": 3.920574331259768e-06, + "loss": 0.78108382, + "num_input_tokens_seen": 20635875, + "step": 976, + "time_per_iteration": 2.700284719467163 + }, + { + "auxiliary_loss_clip": 0.01181378, + "auxiliary_loss_mlp": 0.01051654, + "balance_loss_clip": 0.99023747, + "balance_loss_mlp": 1.03934526, + "epoch": 0.11747730415439187, + "flos": 22381541216640.0, + "grad_norm": 2.971715521244377, + "language_loss": 0.79878634, + "learning_rate": 3.9203568419479716e-06, + "loss": 0.82111669, + "num_input_tokens_seen": 20656430, + "step": 977, + "time_per_iteration": 2.72853684425354 + }, + { + "auxiliary_loss_clip": 0.01187057, + "auxiliary_loss_mlp": 0.0103991, + "balance_loss_clip": 0.98889649, + "balance_loss_mlp": 1.02770257, + "epoch": 0.11759754704503096, + "flos": 22200731130240.0, + "grad_norm": 1.855646357429764, + "language_loss": 0.7538296, + "learning_rate": 3.92013906131957e-06, + "loss": 0.77609932, + "num_input_tokens_seen": 20675360, + "step": 978, + "time_per_iteration": 2.660184383392334 + }, + { + "auxiliary_loss_clip": 0.01160882, + "auxiliary_loss_mlp": 0.01049635, + "balance_loss_clip": 0.950454, + "balance_loss_mlp": 1.03804779, + "epoch": 0.11771778993567006, + "flos": 22309755886080.0, + "grad_norm": 1.5308951368754729, + "language_loss": 0.82486463, + "learning_rate": 3.9199209894076e-06, + "loss": 0.84696984, + "num_input_tokens_seen": 20695675, + "step": 979, + "time_per_iteration": 2.738330364227295 + }, + { + "auxiliary_loss_clip": 0.01238272, + "auxiliary_loss_mlp": 0.0103827, + "balance_loss_clip": 1.06498468, + "balance_loss_mlp": 1.02571177, + "epoch": 0.11783803282630914, + "flos": 21288169175040.0, + "grad_norm": 1.9393904800426778, + "language_loss": 0.90037888, + "learning_rate": 3.919702626245142e-06, + "loss": 0.92314434, + "num_input_tokens_seen": 20715330, + "step": 980, + "time_per_iteration": 2.6168124675750732 + }, + { + "auxiliary_loss_clip": 0.01177064, + "auxiliary_loss_mlp": 0.01046568, + "balance_loss_clip": 0.98642886, + "balance_loss_mlp": 1.03507614, + "epoch": 0.11795827571694824, + "flos": 25371535190400.0, + "grad_norm": 2.146264881931756, + "language_loss": 0.66042304, + "learning_rate": 3.919483971865322e-06, + "loss": 0.68265933, + "num_input_tokens_seen": 20735325, + "step": 981, + "time_per_iteration": 3.6109306812286377 + }, + { + "auxiliary_loss_clip": 0.0118583, + "auxiliary_loss_mlp": 0.01046378, + "balance_loss_clip": 0.9912287, + "balance_loss_mlp": 1.03427863, + "epoch": 0.11807851860758732, + "flos": 23622218933760.0, + "grad_norm": 1.9294814203019037, + "language_loss": 0.88176346, + "learning_rate": 3.91926502630131e-06, + "loss": 0.90408558, + "num_input_tokens_seen": 20755940, + "step": 982, + "time_per_iteration": 3.7010703086853027 + }, + { + "auxiliary_loss_clip": 0.01214792, + "auxiliary_loss_mlp": 0.01043335, + "balance_loss_clip": 1.0316844, + "balance_loss_mlp": 1.03212345, + "epoch": 0.11819876149822642, + "flos": 24972496024320.0, + "grad_norm": 1.674886730820296, + "language_loss": 0.71778202, + "learning_rate": 3.91904578958632e-06, + "loss": 0.74036324, + "num_input_tokens_seen": 20775355, + "step": 983, + "time_per_iteration": 2.6287217140197754 + }, + { + "auxiliary_loss_clip": 0.01239246, + "auxiliary_loss_mlp": 0.010412, + "balance_loss_clip": 1.06841969, + "balance_loss_mlp": 1.02943993, + "epoch": 0.11831900438886551, + "flos": 23003226835200.0, + "grad_norm": 2.001314891513909, + "language_loss": 0.84113395, + "learning_rate": 3.918826261753608e-06, + "loss": 0.86393845, + "num_input_tokens_seen": 20794935, + "step": 984, + "time_per_iteration": 3.535167932510376 + }, + { + "auxiliary_loss_clip": 0.011892, + "auxiliary_loss_mlp": 0.01040508, + "balance_loss_clip": 0.99048889, + "balance_loss_mlp": 1.02916563, + "epoch": 0.1184392472795046, + "flos": 27965147604480.0, + "grad_norm": 2.544915967916611, + "language_loss": 0.71034181, + "learning_rate": 3.918606442836478e-06, + "loss": 0.7326389, + "num_input_tokens_seen": 20817155, + "step": 985, + "time_per_iteration": 3.6237783432006836 + }, + { + "auxiliary_loss_clip": 0.01213292, + "auxiliary_loss_mlp": 0.01040427, + "balance_loss_clip": 1.03225946, + "balance_loss_mlp": 1.02923918, + "epoch": 0.1185594901701437, + "flos": 19898497843200.0, + "grad_norm": 1.809258635029872, + "language_loss": 0.77069283, + "learning_rate": 3.918386332868277e-06, + "loss": 0.79323, + "num_input_tokens_seen": 20835125, + "step": 986, + "time_per_iteration": 2.6533467769622803 + }, + { + "auxiliary_loss_clip": 0.01207464, + "auxiliary_loss_mlp": 0.0103863, + "balance_loss_clip": 1.03006315, + "balance_loss_mlp": 1.02723932, + "epoch": 0.11867973306078278, + "flos": 18912354877440.0, + "grad_norm": 1.690282060394341, + "language_loss": 0.94310212, + "learning_rate": 3.918165931882394e-06, + "loss": 0.96556306, + "num_input_tokens_seen": 20853525, + "step": 987, + "time_per_iteration": 2.66853928565979 + }, + { + "auxiliary_loss_clip": 0.01104111, + "auxiliary_loss_mlp": 0.01043269, + "balance_loss_clip": 0.87138605, + "balance_loss_mlp": 1.03116941, + "epoch": 0.11879997595142187, + "flos": 16982803152000.0, + "grad_norm": 3.4340355679938046, + "language_loss": 0.75179118, + "learning_rate": 3.917945239912264e-06, + "loss": 0.773265, + "num_input_tokens_seen": 20871000, + "step": 988, + "time_per_iteration": 2.7274303436279297 + }, + { + "auxiliary_loss_clip": 0.01138935, + "auxiliary_loss_mlp": 0.01035738, + "balance_loss_clip": 0.91262734, + "balance_loss_mlp": 1.02441335, + "epoch": 0.11892021884206096, + "flos": 17530369056000.0, + "grad_norm": 1.9754092392879194, + "language_loss": 0.75753522, + "learning_rate": 3.917724256991367e-06, + "loss": 0.77928197, + "num_input_tokens_seen": 20889745, + "step": 989, + "time_per_iteration": 2.7716708183288574 + }, + { + "auxiliary_loss_clip": 0.01181754, + "auxiliary_loss_mlp": 0.01047749, + "balance_loss_clip": 0.99217194, + "balance_loss_mlp": 1.03580463, + "epoch": 0.11904046173270005, + "flos": 30955895763840.0, + "grad_norm": 2.138827839156231, + "language_loss": 0.81724077, + "learning_rate": 3.9175029831532245e-06, + "loss": 0.83953583, + "num_input_tokens_seen": 20909260, + "step": 990, + "time_per_iteration": 2.781658411026001 + }, + { + "auxiliary_loss_clip": 0.01165451, + "auxiliary_loss_mlp": 0.01042535, + "balance_loss_clip": 0.95381761, + "balance_loss_mlp": 1.03070903, + "epoch": 0.11916070462333915, + "flos": 20157234485760.0, + "grad_norm": 2.22182337848743, + "language_loss": 0.88423157, + "learning_rate": 3.917281418431404e-06, + "loss": 0.90631145, + "num_input_tokens_seen": 20928305, + "step": 991, + "time_per_iteration": 2.655848979949951 + }, + { + "auxiliary_loss_clip": 0.01190309, + "auxiliary_loss_mlp": 0.01038552, + "balance_loss_clip": 0.99522316, + "balance_loss_mlp": 1.02718496, + "epoch": 0.11928094751397823, + "flos": 23551115961600.0, + "grad_norm": 1.8579543167173054, + "language_loss": 0.7704972, + "learning_rate": 3.917059562859516e-06, + "loss": 0.79278576, + "num_input_tokens_seen": 20947630, + "step": 992, + "time_per_iteration": 2.699216365814209 + }, + { + "auxiliary_loss_clip": 0.01180765, + "auxiliary_loss_mlp": 0.01039401, + "balance_loss_clip": 0.9906708, + "balance_loss_mlp": 1.02816594, + "epoch": 0.11940119040461733, + "flos": 23908426502400.0, + "grad_norm": 2.186442069264035, + "language_loss": 0.88559973, + "learning_rate": 3.916837416471218e-06, + "loss": 0.90780139, + "num_input_tokens_seen": 20964250, + "step": 993, + "time_per_iteration": 2.681586742401123 + }, + { + "auxiliary_loss_clip": 0.01205312, + "auxiliary_loss_mlp": 0.0103864, + "balance_loss_clip": 1.02477527, + "balance_loss_mlp": 1.02652788, + "epoch": 0.11952143329525641, + "flos": 13844533835520.0, + "grad_norm": 2.2424887430677845, + "language_loss": 0.72124571, + "learning_rate": 3.916614979300207e-06, + "loss": 0.74368525, + "num_input_tokens_seen": 20979095, + "step": 994, + "time_per_iteration": 2.613887310028076 + }, + { + "auxiliary_loss_clip": 0.01132391, + "auxiliary_loss_mlp": 0.01045262, + "balance_loss_clip": 0.91341352, + "balance_loss_mlp": 1.03313816, + "epoch": 0.11964167618589551, + "flos": 27015525792000.0, + "grad_norm": 1.5834219944870103, + "language_loss": 0.78554428, + "learning_rate": 3.9163922513802274e-06, + "loss": 0.80732083, + "num_input_tokens_seen": 21001430, + "step": 995, + "time_per_iteration": 2.770183801651001 + }, + { + "auxiliary_loss_clip": 0.01241639, + "auxiliary_loss_mlp": 0.0104537, + "balance_loss_clip": 1.06910205, + "balance_loss_mlp": 1.03309166, + "epoch": 0.1197619190765346, + "flos": 12567622273920.0, + "grad_norm": 2.409102805041981, + "language_loss": 0.82118499, + "learning_rate": 3.916169232745067e-06, + "loss": 0.84405506, + "num_input_tokens_seen": 21019105, + "step": 996, + "time_per_iteration": 2.615283966064453 + }, + { + "auxiliary_loss_clip": 0.01182306, + "auxiliary_loss_mlp": 0.01049697, + "balance_loss_clip": 0.99049282, + "balance_loss_mlp": 1.03660154, + "epoch": 0.11988216196717369, + "flos": 16909437623040.0, + "grad_norm": 3.434895035355584, + "language_loss": 0.92157412, + "learning_rate": 3.915945923428559e-06, + "loss": 0.94389415, + "num_input_tokens_seen": 21035630, + "step": 997, + "time_per_iteration": 2.6623241901397705 + }, + { + "auxiliary_loss_clip": 0.01208978, + "auxiliary_loss_mlp": 0.01044208, + "balance_loss_clip": 1.02747762, + "balance_loss_mlp": 1.03194177, + "epoch": 0.12000240485781279, + "flos": 16216577205120.0, + "grad_norm": 2.1075499782281404, + "language_loss": 0.83225548, + "learning_rate": 3.915722323464577e-06, + "loss": 0.85478735, + "num_input_tokens_seen": 21054235, + "step": 998, + "time_per_iteration": 2.641187906265259 + }, + { + "auxiliary_loss_clip": 0.01214981, + "auxiliary_loss_mlp": 0.01040241, + "balance_loss_clip": 1.03063095, + "balance_loss_mlp": 1.02792048, + "epoch": 0.12012264774845187, + "flos": 49344887525760.0, + "grad_norm": 2.151555618331994, + "language_loss": 0.70427179, + "learning_rate": 3.91549843288704e-06, + "loss": 0.72682399, + "num_input_tokens_seen": 21077915, + "step": 999, + "time_per_iteration": 2.8774325847625732 + }, + { + "auxiliary_loss_clip": 0.0116008, + "auxiliary_loss_mlp": 0.00514069, + "balance_loss_clip": 0.94796145, + "balance_loss_mlp": 0.0, + "epoch": 0.12024289063909097, + "flos": 26979435601920.0, + "grad_norm": 2.175627518310944, + "language_loss": 0.78811729, + "learning_rate": 3.915274251729916e-06, + "loss": 0.8048588, + "num_input_tokens_seen": 21099205, + "step": 1000, + "time_per_iteration": 2.8534278869628906 + }, + { + "auxiliary_loss_clip": 0.01170566, + "auxiliary_loss_mlp": 0.01042174, + "balance_loss_clip": 0.95554149, + "balance_loss_mlp": 1.03017604, + "epoch": 0.12036313352973005, + "flos": 19537308633600.0, + "grad_norm": 1.8533940037948173, + "language_loss": 0.90269911, + "learning_rate": 3.91504978002721e-06, + "loss": 0.92482656, + "num_input_tokens_seen": 21118260, + "step": 1001, + "time_per_iteration": 2.7966110706329346 + }, + { + "auxiliary_loss_clip": 0.01188683, + "auxiliary_loss_mlp": 0.00513783, + "balance_loss_clip": 0.98823416, + "balance_loss_mlp": 0.0, + "epoch": 0.12048337642036915, + "flos": 17268256535040.0, + "grad_norm": 2.4633054826788725, + "language_loss": 0.76244235, + "learning_rate": 3.914825017812974e-06, + "loss": 0.77946705, + "num_input_tokens_seen": 21134910, + "step": 1002, + "time_per_iteration": 2.6588845252990723 + }, + { + "auxiliary_loss_clip": 0.01193114, + "auxiliary_loss_mlp": 0.01042669, + "balance_loss_clip": 0.99543786, + "balance_loss_mlp": 1.03088498, + "epoch": 0.12060361931100824, + "flos": 22856962654080.0, + "grad_norm": 2.4519836222721763, + "language_loss": 0.72316438, + "learning_rate": 3.9145999651213065e-06, + "loss": 0.7455222, + "num_input_tokens_seen": 21154150, + "step": 1003, + "time_per_iteration": 2.7355973720550537 + }, + { + "auxiliary_loss_clip": 0.01216338, + "auxiliary_loss_mlp": 0.01040044, + "balance_loss_clip": 1.03191566, + "balance_loss_mlp": 1.02775383, + "epoch": 0.12072386220164733, + "flos": 16726795943040.0, + "grad_norm": 2.2077904082756517, + "language_loss": 0.8842209, + "learning_rate": 3.9143746219863465e-06, + "loss": 0.90678471, + "num_input_tokens_seen": 21171255, + "step": 1004, + "time_per_iteration": 2.6047918796539307 + }, + { + "auxiliary_loss_clip": 0.01113695, + "auxiliary_loss_mlp": 0.01004571, + "balance_loss_clip": 0.97503483, + "balance_loss_mlp": 1.00113761, + "epoch": 0.12084410509228642, + "flos": 55144176105600.0, + "grad_norm": 0.9344146841906023, + "language_loss": 0.64734763, + "learning_rate": 3.914148988442278e-06, + "loss": 0.66853029, + "num_input_tokens_seen": 21227045, + "step": 1005, + "time_per_iteration": 3.188985586166382 + }, + { + "auxiliary_loss_clip": 0.01178818, + "auxiliary_loss_mlp": 0.01041537, + "balance_loss_clip": 0.99107093, + "balance_loss_mlp": 1.02873397, + "epoch": 0.1209643479829255, + "flos": 26760236855040.0, + "grad_norm": 3.298258488442525, + "language_loss": 0.95272768, + "learning_rate": 3.91392306452333e-06, + "loss": 0.97493124, + "num_input_tokens_seen": 21244120, + "step": 1006, + "time_per_iteration": 2.680649757385254 + }, + { + "auxiliary_loss_clip": 0.0123942, + "auxiliary_loss_mlp": 0.01043062, + "balance_loss_clip": 1.06847513, + "balance_loss_mlp": 1.02996087, + "epoch": 0.1210845908735646, + "flos": 11035026725760.0, + "grad_norm": 3.6942014640340943, + "language_loss": 0.66035879, + "learning_rate": 3.913696850263774e-06, + "loss": 0.68318361, + "num_input_tokens_seen": 21258485, + "step": 1007, + "time_per_iteration": 3.50620436668396 + }, + { + "auxiliary_loss_clip": 0.01212187, + "auxiliary_loss_mlp": 0.01047515, + "balance_loss_clip": 1.02934813, + "balance_loss_mlp": 1.03543329, + "epoch": 0.1212048337642037, + "flos": 20484631975680.0, + "grad_norm": 5.216926324463405, + "language_loss": 0.79386139, + "learning_rate": 3.913470345697929e-06, + "loss": 0.8164584, + "num_input_tokens_seen": 21277115, + "step": 1008, + "time_per_iteration": 3.593285322189331 + }, + { + "auxiliary_loss_clip": 0.0114121, + "auxiliary_loss_mlp": 0.01040084, + "balance_loss_clip": 0.91197544, + "balance_loss_mlp": 1.0276624, + "epoch": 0.12132507665484278, + "flos": 22346061557760.0, + "grad_norm": 2.316345993332439, + "language_loss": 0.85475415, + "learning_rate": 3.913243550860153e-06, + "loss": 0.87656713, + "num_input_tokens_seen": 21294880, + "step": 1009, + "time_per_iteration": 3.6586992740631104 + }, + { + "auxiliary_loss_clip": 0.01220751, + "auxiliary_loss_mlp": 0.0105103, + "balance_loss_clip": 1.03657067, + "balance_loss_mlp": 1.03858459, + "epoch": 0.12144531954548188, + "flos": 29314957818240.0, + "grad_norm": 2.26631050956916, + "language_loss": 0.76376992, + "learning_rate": 3.913016465784852e-06, + "loss": 0.7864877, + "num_input_tokens_seen": 21315555, + "step": 1010, + "time_per_iteration": 2.67836332321167 + }, + { + "auxiliary_loss_clip": 0.01142439, + "auxiliary_loss_mlp": 0.01039017, + "balance_loss_clip": 0.91233838, + "balance_loss_mlp": 1.02608848, + "epoch": 0.12156556243612096, + "flos": 20485242506880.0, + "grad_norm": 3.5643858379099935, + "language_loss": 0.71828097, + "learning_rate": 3.912789090506474e-06, + "loss": 0.7400955, + "num_input_tokens_seen": 21334815, + "step": 1011, + "time_per_iteration": 3.7396316528320312 + }, + { + "auxiliary_loss_clip": 0.01167552, + "auxiliary_loss_mlp": 0.0104467, + "balance_loss_clip": 0.94744492, + "balance_loss_mlp": 1.03311253, + "epoch": 0.12168580532676006, + "flos": 16472009796480.0, + "grad_norm": 2.212230438149602, + "language_loss": 0.71969527, + "learning_rate": 3.9125614250595114e-06, + "loss": 0.74181747, + "num_input_tokens_seen": 21351025, + "step": 1012, + "time_per_iteration": 2.756777048110962 + }, + { + "auxiliary_loss_clip": 0.01216456, + "auxiliary_loss_mlp": 0.01040723, + "balance_loss_clip": 1.02933168, + "balance_loss_mlp": 1.02913618, + "epoch": 0.12180604821739914, + "flos": 15341290588800.0, + "grad_norm": 2.7381540735705787, + "language_loss": 0.88561857, + "learning_rate": 3.912333469478502e-06, + "loss": 0.90819037, + "num_input_tokens_seen": 21368990, + "step": 1013, + "time_per_iteration": 2.693655014038086 + }, + { + "auxiliary_loss_clip": 0.01188006, + "auxiliary_loss_mlp": 0.01044829, + "balance_loss_clip": 0.99039131, + "balance_loss_mlp": 1.03332543, + "epoch": 0.12192629110803824, + "flos": 19318038059520.0, + "grad_norm": 1.9756539706105283, + "language_loss": 0.77805483, + "learning_rate": 3.912105223798025e-06, + "loss": 0.80038321, + "num_input_tokens_seen": 21388410, + "step": 1014, + "time_per_iteration": 2.6281020641326904 + }, + { + "auxiliary_loss_clip": 0.01087368, + "auxiliary_loss_mlp": 0.01007591, + "balance_loss_clip": 0.93249333, + "balance_loss_mlp": 1.00413382, + "epoch": 0.12204653399867733, + "flos": 47725354085760.0, + "grad_norm": 0.9928613857939749, + "language_loss": 0.67648971, + "learning_rate": 3.9118766880527065e-06, + "loss": 0.69743931, + "num_input_tokens_seen": 21442845, + "step": 1015, + "time_per_iteration": 3.206681251525879 + }, + { + "auxiliary_loss_clip": 0.01127862, + "auxiliary_loss_mlp": 0.01037313, + "balance_loss_clip": 0.91028666, + "balance_loss_mlp": 1.02586329, + "epoch": 0.12216677688931642, + "flos": 18221936584320.0, + "grad_norm": 1.6452607086591595, + "language_loss": 0.73684698, + "learning_rate": 3.9116478622772145e-06, + "loss": 0.75849873, + "num_input_tokens_seen": 21461420, + "step": 1016, + "time_per_iteration": 2.696275234222412 + }, + { + "auxiliary_loss_clip": 0.01211613, + "auxiliary_loss_mlp": 0.01040884, + "balance_loss_clip": 1.03085256, + "balance_loss_mlp": 1.02853954, + "epoch": 0.12228701977995551, + "flos": 27525636789120.0, + "grad_norm": 1.5672189451531067, + "language_loss": 0.87923831, + "learning_rate": 3.911418746506261e-06, + "loss": 0.9017632, + "num_input_tokens_seen": 21481550, + "step": 1017, + "time_per_iteration": 2.679307460784912 + }, + { + "auxiliary_loss_clip": 0.01217355, + "auxiliary_loss_mlp": 0.01045342, + "balance_loss_clip": 1.0355128, + "balance_loss_mlp": 1.03385675, + "epoch": 0.1224072626705946, + "flos": 21798136517760.0, + "grad_norm": 1.6411868869220771, + "language_loss": 0.78469336, + "learning_rate": 3.911189340774604e-06, + "loss": 0.80732036, + "num_input_tokens_seen": 21501680, + "step": 1018, + "time_per_iteration": 2.65295147895813 + }, + { + "auxiliary_loss_clip": 0.01197973, + "auxiliary_loss_mlp": 0.01038354, + "balance_loss_clip": 0.99205893, + "balance_loss_mlp": 1.02671278, + "epoch": 0.1225275055612337, + "flos": 20703758895360.0, + "grad_norm": 1.6137971777057254, + "language_loss": 0.79197931, + "learning_rate": 3.910959645117043e-06, + "loss": 0.81434256, + "num_input_tokens_seen": 21521015, + "step": 1019, + "time_per_iteration": 2.639066219329834 + }, + { + "auxiliary_loss_clip": 0.01100028, + "auxiliary_loss_mlp": 0.00505631, + "balance_loss_clip": 0.97038293, + "balance_loss_mlp": 0.0, + "epoch": 0.12264774845187278, + "flos": 57745294462080.0, + "grad_norm": 0.8138860341442092, + "language_loss": 0.56708634, + "learning_rate": 3.910729659568423e-06, + "loss": 0.583143, + "num_input_tokens_seen": 21578200, + "step": 1020, + "time_per_iteration": 3.229926109313965 + }, + { + "auxiliary_loss_clip": 0.01193355, + "auxiliary_loss_mlp": 0.01044825, + "balance_loss_clip": 0.99458528, + "balance_loss_mlp": 1.03366709, + "epoch": 0.12276799134251187, + "flos": 26396282298240.0, + "grad_norm": 1.7274118833755274, + "language_loss": 0.82311535, + "learning_rate": 3.9104993841636344e-06, + "loss": 0.84549713, + "num_input_tokens_seen": 21598770, + "step": 1021, + "time_per_iteration": 2.696859121322632 + }, + { + "auxiliary_loss_clip": 0.01185592, + "auxiliary_loss_mlp": 0.00512854, + "balance_loss_clip": 0.99475062, + "balance_loss_mlp": 0.0, + "epoch": 0.12288823423315097, + "flos": 21064193919360.0, + "grad_norm": 1.802371955201166, + "language_loss": 0.8099758, + "learning_rate": 3.910268818937608e-06, + "loss": 0.82696027, + "num_input_tokens_seen": 21616925, + "step": 1022, + "time_per_iteration": 2.696629762649536 + }, + { + "auxiliary_loss_clip": 0.01141203, + "auxiliary_loss_mlp": 0.01041967, + "balance_loss_clip": 0.91744292, + "balance_loss_mlp": 1.0310117, + "epoch": 0.12300847712379005, + "flos": 12312441077760.0, + "grad_norm": 2.4589075102644444, + "language_loss": 0.87487674, + "learning_rate": 3.9100379639253196e-06, + "loss": 0.89670849, + "num_input_tokens_seen": 21633645, + "step": 1023, + "time_per_iteration": 2.7222259044647217 + }, + { + "auxiliary_loss_clip": 0.01188579, + "auxiliary_loss_mlp": 0.01038397, + "balance_loss_clip": 0.98913944, + "balance_loss_mlp": 1.02688134, + "epoch": 0.12312872001442915, + "flos": 16762239688320.0, + "grad_norm": 2.847446403471357, + "language_loss": 0.86576712, + "learning_rate": 3.909806819161791e-06, + "loss": 0.88803685, + "num_input_tokens_seen": 21649120, + "step": 1024, + "time_per_iteration": 2.681826114654541 + }, + { + "auxiliary_loss_clip": 0.01167018, + "auxiliary_loss_mlp": 0.01038995, + "balance_loss_clip": 0.94987297, + "balance_loss_mlp": 1.02708054, + "epoch": 0.12324896290506823, + "flos": 18404937400320.0, + "grad_norm": 1.9612268631076526, + "language_loss": 0.8626048, + "learning_rate": 3.909575384682086e-06, + "loss": 0.88466495, + "num_input_tokens_seen": 21668000, + "step": 1025, + "time_per_iteration": 2.6717522144317627 + }, + { + "auxiliary_loss_clip": 0.01218751, + "auxiliary_loss_mlp": 0.01045279, + "balance_loss_clip": 1.03080034, + "balance_loss_mlp": 1.03384638, + "epoch": 0.12336920579570733, + "flos": 18915407533440.0, + "grad_norm": 1.7744485059607562, + "language_loss": 0.69201022, + "learning_rate": 3.9093436605213144e-06, + "loss": 0.71465057, + "num_input_tokens_seen": 21688500, + "step": 1026, + "time_per_iteration": 2.6838905811309814 + }, + { + "auxiliary_loss_clip": 0.01188966, + "auxiliary_loss_mlp": 0.01046555, + "balance_loss_clip": 0.99076247, + "balance_loss_mlp": 1.03476501, + "epoch": 0.12348944868634643, + "flos": 23878369797120.0, + "grad_norm": 1.9943863523960879, + "language_loss": 0.7961821, + "learning_rate": 3.909111646714627e-06, + "loss": 0.81853724, + "num_input_tokens_seen": 21709345, + "step": 1027, + "time_per_iteration": 2.691622257232666 + }, + { + "auxiliary_loss_clip": 0.01236771, + "auxiliary_loss_mlp": 0.01043206, + "balance_loss_clip": 1.06817698, + "balance_loss_mlp": 1.03189933, + "epoch": 0.12360969157698551, + "flos": 19026084314880.0, + "grad_norm": 1.9989227974704133, + "language_loss": 0.72506928, + "learning_rate": 3.9088793432972206e-06, + "loss": 0.74786907, + "num_input_tokens_seen": 21728165, + "step": 1028, + "time_per_iteration": 2.6268160343170166 + }, + { + "auxiliary_loss_clip": 0.01141076, + "auxiliary_loss_mlp": 0.01036942, + "balance_loss_clip": 0.91502607, + "balance_loss_mlp": 1.02526522, + "epoch": 0.1237299344676246, + "flos": 13224607983360.0, + "grad_norm": 2.1465920774215093, + "language_loss": 0.82321346, + "learning_rate": 3.908646750304336e-06, + "loss": 0.84499359, + "num_input_tokens_seen": 21745850, + "step": 1029, + "time_per_iteration": 2.735625982284546 + }, + { + "auxiliary_loss_clip": 0.01195747, + "auxiliary_loss_mlp": 0.01045608, + "balance_loss_clip": 0.99669313, + "balance_loss_mlp": 1.03358579, + "epoch": 0.12385017735826369, + "flos": 20485673470080.0, + "grad_norm": 1.7901813619100921, + "language_loss": 0.8722555, + "learning_rate": 3.908413867771257e-06, + "loss": 0.89466906, + "num_input_tokens_seen": 21764760, + "step": 1030, + "time_per_iteration": 2.6809160709381104 + }, + { + "auxiliary_loss_clip": 0.0121282, + "auxiliary_loss_mlp": 0.01049298, + "balance_loss_clip": 1.03190494, + "balance_loss_mlp": 1.03721023, + "epoch": 0.12397042024890279, + "flos": 17347835116800.0, + "grad_norm": 1.782994657873362, + "language_loss": 0.80546856, + "learning_rate": 3.908180695733311e-06, + "loss": 0.82808971, + "num_input_tokens_seen": 21784250, + "step": 1031, + "time_per_iteration": 2.6583776473999023 + }, + { + "auxiliary_loss_clip": 0.01113883, + "auxiliary_loss_mlp": 0.01038465, + "balance_loss_clip": 0.9046123, + "balance_loss_mlp": 1.02725387, + "epoch": 0.12409066313954187, + "flos": 20412343854720.0, + "grad_norm": 2.0243299046554584, + "language_loss": 0.82656443, + "learning_rate": 3.907947234225871e-06, + "loss": 0.84808791, + "num_input_tokens_seen": 21803260, + "step": 1032, + "time_per_iteration": 2.778040647506714 + }, + { + "auxiliary_loss_clip": 0.01107614, + "auxiliary_loss_mlp": 0.01041292, + "balance_loss_clip": 0.87121558, + "balance_loss_mlp": 1.02937067, + "epoch": 0.12421090603018096, + "flos": 20736688688640.0, + "grad_norm": 1.9858846708379756, + "language_loss": 0.86924982, + "learning_rate": 3.907713483284352e-06, + "loss": 0.89073884, + "num_input_tokens_seen": 21822735, + "step": 1033, + "time_per_iteration": 3.733365297317505 + }, + { + "auxiliary_loss_clip": 0.01075205, + "auxiliary_loss_mlp": 0.01035591, + "balance_loss_clip": 0.82821089, + "balance_loss_mlp": 1.0246352, + "epoch": 0.12433114892082006, + "flos": 24498834353280.0, + "grad_norm": 2.0855385335617176, + "language_loss": 0.96973592, + "learning_rate": 3.907479442944216e-06, + "loss": 0.99084389, + "num_input_tokens_seen": 21841140, + "step": 1034, + "time_per_iteration": 3.7987430095672607 + }, + { + "auxiliary_loss_clip": 0.01210166, + "auxiliary_loss_mlp": 0.01041812, + "balance_loss_clip": 1.02952349, + "balance_loss_mlp": 1.03071952, + "epoch": 0.12445139181145914, + "flos": 19682315838720.0, + "grad_norm": 1.9774067764866305, + "language_loss": 0.92419863, + "learning_rate": 3.907245113240963e-06, + "loss": 0.94671839, + "num_input_tokens_seen": 21859260, + "step": 1035, + "time_per_iteration": 3.5384881496429443 + }, + { + "auxiliary_loss_clip": 0.01160301, + "auxiliary_loss_mlp": 0.01047537, + "balance_loss_clip": 0.9469527, + "balance_loss_mlp": 1.035074, + "epoch": 0.12457163470209824, + "flos": 46423087522560.0, + "grad_norm": 1.9113373652671257, + "language_loss": 0.73736441, + "learning_rate": 3.907010494210144e-06, + "loss": 0.75944281, + "num_input_tokens_seen": 21881920, + "step": 1036, + "time_per_iteration": 2.9085123538970947 + }, + { + "auxiliary_loss_clip": 0.01213132, + "auxiliary_loss_mlp": 0.01040602, + "balance_loss_clip": 1.02875578, + "balance_loss_mlp": 1.02835298, + "epoch": 0.12469187759273732, + "flos": 20376289578240.0, + "grad_norm": 1.9122653538665766, + "language_loss": 0.91574901, + "learning_rate": 3.9067755858873495e-06, + "loss": 0.9382863, + "num_input_tokens_seen": 21898720, + "step": 1037, + "time_per_iteration": 3.5290942192077637 + }, + { + "auxiliary_loss_clip": 0.01071901, + "auxiliary_loss_mlp": 0.01007286, + "balance_loss_clip": 0.92567837, + "balance_loss_mlp": 1.00409162, + "epoch": 0.12481212048337642, + "flos": 69224641447680.0, + "grad_norm": 0.8645464970982801, + "language_loss": 0.62825269, + "learning_rate": 3.906540388308214e-06, + "loss": 0.64904457, + "num_input_tokens_seen": 21958305, + "step": 1038, + "time_per_iteration": 3.2305634021759033 + }, + { + "auxiliary_loss_clip": 0.01127161, + "auxiliary_loss_mlp": 0.01038297, + "balance_loss_clip": 0.91223502, + "balance_loss_mlp": 1.02723455, + "epoch": 0.12493236337401552, + "flos": 18223696350720.0, + "grad_norm": 2.87653714115967, + "language_loss": 0.8146469, + "learning_rate": 3.906304901508417e-06, + "loss": 0.83630145, + "num_input_tokens_seen": 21977205, + "step": 1039, + "time_per_iteration": 2.713463544845581 + }, + { + "auxiliary_loss_clip": 0.01215023, + "auxiliary_loss_mlp": 0.01050667, + "balance_loss_clip": 1.03407979, + "balance_loss_mlp": 1.03952682, + "epoch": 0.12505260626465461, + "flos": 30044375303040.0, + "grad_norm": 2.0327145202382177, + "language_loss": 0.76117033, + "learning_rate": 3.9060691255236835e-06, + "loss": 0.78382725, + "num_input_tokens_seen": 21997770, + "step": 1040, + "time_per_iteration": 2.6574020385742188 + }, + { + "auxiliary_loss_clip": 0.01208066, + "auxiliary_loss_mlp": 0.01047126, + "balance_loss_clip": 1.02738011, + "balance_loss_mlp": 1.03494298, + "epoch": 0.1251728491552937, + "flos": 24433980347520.0, + "grad_norm": 1.6696141473297237, + "language_loss": 0.8079021, + "learning_rate": 3.905833060389778e-06, + "loss": 0.83045399, + "num_input_tokens_seen": 22021890, + "step": 1041, + "time_per_iteration": 2.740739583969116 + }, + { + "auxiliary_loss_clip": 0.01234756, + "auxiliary_loss_mlp": 0.00513886, + "balance_loss_clip": 1.06858957, + "balance_loss_mlp": 0.0, + "epoch": 0.12529309204593278, + "flos": 27119809952640.0, + "grad_norm": 5.253791202537772, + "language_loss": 0.78054321, + "learning_rate": 3.905596706142513e-06, + "loss": 0.79802966, + "num_input_tokens_seen": 22043300, + "step": 1042, + "time_per_iteration": 2.6558804512023926 + }, + { + "auxiliary_loss_clip": 0.01156937, + "auxiliary_loss_mlp": 0.01034737, + "balance_loss_clip": 0.9485606, + "balance_loss_mlp": 1.02308989, + "epoch": 0.12541333493657186, + "flos": 30774151923840.0, + "grad_norm": 2.154638877118488, + "language_loss": 0.86017084, + "learning_rate": 3.9053600628177435e-06, + "loss": 0.88208759, + "num_input_tokens_seen": 22062910, + "step": 1043, + "time_per_iteration": 2.832819700241089 + }, + { + "auxiliary_loss_clip": 0.01231019, + "auxiliary_loss_mlp": 0.01035918, + "balance_loss_clip": 1.06469202, + "balance_loss_mlp": 1.02467024, + "epoch": 0.12553357782721097, + "flos": 23659566099840.0, + "grad_norm": 1.9822780283858366, + "language_loss": 0.84634483, + "learning_rate": 3.905123130451367e-06, + "loss": 0.86901414, + "num_input_tokens_seen": 22084010, + "step": 1044, + "time_per_iteration": 2.6716558933258057 + }, + { + "auxiliary_loss_clip": 0.01233267, + "auxiliary_loss_mlp": 0.01038643, + "balance_loss_clip": 1.06659579, + "balance_loss_mlp": 1.02691269, + "epoch": 0.12565382071785006, + "flos": 24863758577280.0, + "grad_norm": 1.68287524922027, + "language_loss": 0.79659057, + "learning_rate": 3.904885909079326e-06, + "loss": 0.81930965, + "num_input_tokens_seen": 22102795, + "step": 1045, + "time_per_iteration": 2.759639024734497 + }, + { + "auxiliary_loss_clip": 0.01207843, + "auxiliary_loss_mlp": 0.01043136, + "balance_loss_clip": 1.0267148, + "balance_loss_mlp": 1.03105998, + "epoch": 0.12577406360848914, + "flos": 21360780518400.0, + "grad_norm": 2.9308809258759014, + "language_loss": 0.77753967, + "learning_rate": 3.904648398737607e-06, + "loss": 0.80004948, + "num_input_tokens_seen": 22121360, + "step": 1046, + "time_per_iteration": 2.638547420501709 + }, + { + "auxiliary_loss_clip": 0.01232517, + "auxiliary_loss_mlp": 0.01041262, + "balance_loss_clip": 1.06737792, + "balance_loss_mlp": 1.03037238, + "epoch": 0.12589430649912825, + "flos": 36138056774400.0, + "grad_norm": 1.8635972294890366, + "language_loss": 0.78321606, + "learning_rate": 3.9044105994622406e-06, + "loss": 0.80595386, + "num_input_tokens_seen": 22142505, + "step": 1047, + "time_per_iteration": 2.7065680027008057 + }, + { + "auxiliary_loss_clip": 0.01183338, + "auxiliary_loss_mlp": 0.00514145, + "balance_loss_clip": 0.98661804, + "balance_loss_mlp": 0.0, + "epoch": 0.12601454938976733, + "flos": 25337671643520.0, + "grad_norm": 1.8653135910675698, + "language_loss": 0.81818402, + "learning_rate": 3.9041725112893005e-06, + "loss": 0.83515882, + "num_input_tokens_seen": 22163730, + "step": 1048, + "time_per_iteration": 2.72196888923645 + }, + { + "auxiliary_loss_clip": 0.01157023, + "auxiliary_loss_mlp": 0.01034815, + "balance_loss_clip": 0.95239639, + "balance_loss_mlp": 1.02340126, + "epoch": 0.12613479228040642, + "flos": 15560094286080.0, + "grad_norm": 1.6627392056760222, + "language_loss": 0.75312966, + "learning_rate": 3.903934134254904e-06, + "loss": 0.77504808, + "num_input_tokens_seen": 22181520, + "step": 1049, + "time_per_iteration": 2.645756483078003 + }, + { + "auxiliary_loss_clip": 0.01213406, + "auxiliary_loss_mlp": 0.01040418, + "balance_loss_clip": 1.0272907, + "balance_loss_mlp": 1.02860475, + "epoch": 0.1262550351710455, + "flos": 21470595373440.0, + "grad_norm": 2.0491279186376548, + "language_loss": 0.84744316, + "learning_rate": 3.903695468395213e-06, + "loss": 0.86998141, + "num_input_tokens_seen": 22199390, + "step": 1050, + "time_per_iteration": 2.659350633621216 + }, + { + "auxiliary_loss_clip": 0.01189356, + "auxiliary_loss_mlp": 0.01039857, + "balance_loss_clip": 0.98706937, + "balance_loss_mlp": 1.02905107, + "epoch": 0.1263752780616846, + "flos": 31576719456000.0, + "grad_norm": 2.01299822410381, + "language_loss": 0.55851203, + "learning_rate": 3.903456513746434e-06, + "loss": 0.58080417, + "num_input_tokens_seen": 22220365, + "step": 1051, + "time_per_iteration": 2.743755340576172 + }, + { + "auxiliary_loss_clip": 0.01230223, + "auxiliary_loss_mlp": 0.01036042, + "balance_loss_clip": 1.0649389, + "balance_loss_mlp": 1.02511621, + "epoch": 0.1264955209523237, + "flos": 28768217927040.0, + "grad_norm": 1.7318237462199189, + "language_loss": 0.87430906, + "learning_rate": 3.903217270344815e-06, + "loss": 0.8969717, + "num_input_tokens_seen": 22240615, + "step": 1052, + "time_per_iteration": 2.659285545349121 + }, + { + "auxiliary_loss_clip": 0.01153944, + "auxiliary_loss_mlp": 0.01046554, + "balance_loss_clip": 0.94818532, + "balance_loss_mlp": 1.03615248, + "epoch": 0.12661576384296278, + "flos": 29241125412480.0, + "grad_norm": 1.6361119641393918, + "language_loss": 0.82709098, + "learning_rate": 3.902977738226648e-06, + "loss": 0.84909594, + "num_input_tokens_seen": 22261350, + "step": 1053, + "time_per_iteration": 2.7240798473358154 + }, + { + "auxiliary_loss_clip": 0.01215012, + "auxiliary_loss_mlp": 0.01047008, + "balance_loss_clip": 1.03148639, + "balance_loss_mlp": 1.03593326, + "epoch": 0.12673600673360189, + "flos": 20850346298880.0, + "grad_norm": 5.133836806375172, + "language_loss": 0.9117626, + "learning_rate": 3.902737917428273e-06, + "loss": 0.9343828, + "num_input_tokens_seen": 22279515, + "step": 1054, + "time_per_iteration": 2.6216650009155273 + }, + { + "auxiliary_loss_clip": 0.01233471, + "auxiliary_loss_mlp": 0.01043905, + "balance_loss_clip": 1.06749487, + "balance_loss_mlp": 1.03257394, + "epoch": 0.12685624962424097, + "flos": 25263695583360.0, + "grad_norm": 1.5883055588300106, + "language_loss": 0.83795291, + "learning_rate": 3.902497807986068e-06, + "loss": 0.86072659, + "num_input_tokens_seen": 22299535, + "step": 1055, + "time_per_iteration": 2.615903854370117 + }, + { + "auxiliary_loss_clip": 0.01161026, + "auxiliary_loss_mlp": 0.01034649, + "balance_loss_clip": 0.94691473, + "balance_loss_mlp": 1.02356863, + "epoch": 0.12697649251488005, + "flos": 27527109246720.0, + "grad_norm": 1.5365347213763942, + "language_loss": 0.84075612, + "learning_rate": 3.902257409936458e-06, + "loss": 0.86271286, + "num_input_tokens_seen": 22320300, + "step": 1056, + "time_per_iteration": 2.796213150024414 + }, + { + "auxiliary_loss_clip": 0.01186664, + "auxiliary_loss_mlp": 0.01041714, + "balance_loss_clip": 0.99235982, + "balance_loss_mlp": 1.02993059, + "epoch": 0.12709673540551916, + "flos": 21251863503360.0, + "grad_norm": 1.7551362059843327, + "language_loss": 0.84098583, + "learning_rate": 3.902016723315912e-06, + "loss": 0.86326963, + "num_input_tokens_seen": 22338240, + "step": 1057, + "time_per_iteration": 2.7725284099578857 + }, + { + "auxiliary_loss_clip": 0.01206751, + "auxiliary_loss_mlp": 0.01038364, + "balance_loss_clip": 1.02584136, + "balance_loss_mlp": 1.02756929, + "epoch": 0.12721697829615825, + "flos": 25337707557120.0, + "grad_norm": 2.0286218793996693, + "language_loss": 0.69730043, + "learning_rate": 3.901775748160941e-06, + "loss": 0.7197516, + "num_input_tokens_seen": 22357420, + "step": 1058, + "time_per_iteration": 2.7640974521636963 + }, + { + "auxiliary_loss_clip": 0.01085424, + "auxiliary_loss_mlp": 0.01002175, + "balance_loss_clip": 0.96177799, + "balance_loss_mlp": 0.99925488, + "epoch": 0.12733722118679733, + "flos": 61943287754880.0, + "grad_norm": 0.7913121823588728, + "language_loss": 0.60917258, + "learning_rate": 3.901534484508101e-06, + "loss": 0.63004863, + "num_input_tokens_seen": 22420095, + "step": 1059, + "time_per_iteration": 4.2071311473846436 + }, + { + "auxiliary_loss_clip": 0.01173503, + "auxiliary_loss_mlp": 0.01036736, + "balance_loss_clip": 0.98649418, + "balance_loss_mlp": 1.02538133, + "epoch": 0.1274574640774364, + "flos": 26976742081920.0, + "grad_norm": 1.7090869641015556, + "language_loss": 0.74603504, + "learning_rate": 3.901292932393991e-06, + "loss": 0.7681374, + "num_input_tokens_seen": 22438975, + "step": 1060, + "time_per_iteration": 3.618583917617798 + }, + { + "auxiliary_loss_clip": 0.01235484, + "auxiliary_loss_mlp": 0.01043355, + "balance_loss_clip": 1.06803584, + "balance_loss_mlp": 1.03245962, + "epoch": 0.12757770696807552, + "flos": 22236318529920.0, + "grad_norm": 1.9947116237627995, + "language_loss": 0.85400176, + "learning_rate": 3.9010510918552555e-06, + "loss": 0.87679017, + "num_input_tokens_seen": 22458050, + "step": 1061, + "time_per_iteration": 3.526150941848755 + }, + { + "auxiliary_loss_clip": 0.01179885, + "auxiliary_loss_mlp": 0.01041541, + "balance_loss_clip": 0.98673141, + "balance_loss_mlp": 1.02893472, + "epoch": 0.1276979498587146, + "flos": 28547905858560.0, + "grad_norm": 2.103679290028297, + "language_loss": 0.74483597, + "learning_rate": 3.900808962928581e-06, + "loss": 0.76705027, + "num_input_tokens_seen": 22475665, + "step": 1062, + "time_per_iteration": 2.687591791152954 + }, + { + "auxiliary_loss_clip": 0.01231837, + "auxiliary_loss_mlp": 0.01039146, + "balance_loss_clip": 1.06747615, + "balance_loss_mlp": 1.02943063, + "epoch": 0.1278181927493537, + "flos": 17420338719360.0, + "grad_norm": 2.2150982203577816, + "language_loss": 0.89410233, + "learning_rate": 3.900566545650698e-06, + "loss": 0.91681206, + "num_input_tokens_seen": 22493335, + "step": 1063, + "time_per_iteration": 3.514143943786621 + }, + { + "auxiliary_loss_clip": 0.01207478, + "auxiliary_loss_mlp": 0.01040388, + "balance_loss_clip": 1.02764463, + "balance_loss_mlp": 1.02910507, + "epoch": 0.1279384356399928, + "flos": 21138636856320.0, + "grad_norm": 2.0647325861776396, + "language_loss": 0.81609643, + "learning_rate": 3.900323840058381e-06, + "loss": 0.83857512, + "num_input_tokens_seen": 22511045, + "step": 1064, + "time_per_iteration": 2.617783546447754 + }, + { + "auxiliary_loss_clip": 0.01209361, + "auxiliary_loss_mlp": 0.01037449, + "balance_loss_clip": 1.02741504, + "balance_loss_mlp": 1.0268693, + "epoch": 0.12805867853063188, + "flos": 26576733248640.0, + "grad_norm": 1.847335261655883, + "language_loss": 0.81883413, + "learning_rate": 3.900080846188449e-06, + "loss": 0.84130228, + "num_input_tokens_seen": 22529635, + "step": 1065, + "time_per_iteration": 2.6683356761932373 + }, + { + "auxiliary_loss_clip": 0.01233815, + "auxiliary_loss_mlp": 0.01045119, + "balance_loss_clip": 1.06692219, + "balance_loss_mlp": 1.03422308, + "epoch": 0.12817892142127096, + "flos": 16436206915200.0, + "grad_norm": 1.7622168846653747, + "language_loss": 0.81493706, + "learning_rate": 3.8998375640777625e-06, + "loss": 0.83772635, + "num_input_tokens_seen": 22547505, + "step": 1066, + "time_per_iteration": 2.5849993228912354 + }, + { + "auxiliary_loss_clip": 0.0109093, + "auxiliary_loss_mlp": 0.01005125, + "balance_loss_clip": 0.96368313, + "balance_loss_mlp": 1.00197816, + "epoch": 0.12829916431191005, + "flos": 60757049099520.0, + "grad_norm": 0.7000821044752599, + "language_loss": 0.52642578, + "learning_rate": 3.899593993763229e-06, + "loss": 0.54738635, + "num_input_tokens_seen": 22608465, + "step": 1067, + "time_per_iteration": 3.209787130355835 + }, + { + "auxiliary_loss_clip": 0.01150467, + "auxiliary_loss_mlp": 0.01039354, + "balance_loss_clip": 0.94754672, + "balance_loss_mlp": 1.02786207, + "epoch": 0.12841940720254916, + "flos": 29786895636480.0, + "grad_norm": 1.8825572715082823, + "language_loss": 0.81318057, + "learning_rate": 3.899350135281796e-06, + "loss": 0.83507872, + "num_input_tokens_seen": 22629465, + "step": 1068, + "time_per_iteration": 2.775735378265381 + }, + { + "auxiliary_loss_clip": 0.01164806, + "auxiliary_loss_mlp": 0.01039768, + "balance_loss_clip": 0.95287108, + "balance_loss_mlp": 1.02884817, + "epoch": 0.12853965009318824, + "flos": 25951851319680.0, + "grad_norm": 1.9194951687205803, + "language_loss": 0.79612362, + "learning_rate": 3.8991059886704585e-06, + "loss": 0.81816936, + "num_input_tokens_seen": 22648970, + "step": 1069, + "time_per_iteration": 2.7581124305725098 + }, + { + "auxiliary_loss_clip": 0.01151456, + "auxiliary_loss_mlp": 0.0104772, + "balance_loss_clip": 0.94946611, + "balance_loss_mlp": 1.03639531, + "epoch": 0.12865989298382732, + "flos": 30846871008000.0, + "grad_norm": 2.490314562118098, + "language_loss": 0.82974637, + "learning_rate": 3.898861553966252e-06, + "loss": 0.8517381, + "num_input_tokens_seen": 22668620, + "step": 1070, + "time_per_iteration": 2.7666690349578857 + }, + { + "auxiliary_loss_clip": 0.01081538, + "auxiliary_loss_mlp": 0.0103982, + "balance_loss_clip": 0.83115506, + "balance_loss_mlp": 1.02898979, + "epoch": 0.12878013587446643, + "flos": 25885776251520.0, + "grad_norm": 1.5374109676430943, + "language_loss": 0.880072, + "learning_rate": 3.898616831206257e-06, + "loss": 0.90128565, + "num_input_tokens_seen": 22689045, + "step": 1071, + "time_per_iteration": 2.8741071224212646 + }, + { + "auxiliary_loss_clip": 0.01152766, + "auxiliary_loss_mlp": 0.01036339, + "balance_loss_clip": 0.94732249, + "balance_loss_mlp": 1.02465653, + "epoch": 0.12890037876510552, + "flos": 23333138277120.0, + "grad_norm": 2.3386483635474233, + "language_loss": 0.77104723, + "learning_rate": 3.8983718204276e-06, + "loss": 0.79293835, + "num_input_tokens_seen": 22711265, + "step": 1072, + "time_per_iteration": 2.7703967094421387 + }, + { + "auxiliary_loss_clip": 0.0118263, + "auxiliary_loss_mlp": 0.0104605, + "balance_loss_clip": 0.98928863, + "balance_loss_mlp": 1.03451049, + "epoch": 0.1290206216557446, + "flos": 23587242065280.0, + "grad_norm": 1.7348606137939093, + "language_loss": 0.82497478, + "learning_rate": 3.898126521667446e-06, + "loss": 0.84726155, + "num_input_tokens_seen": 22731420, + "step": 1073, + "time_per_iteration": 2.6622674465179443 + }, + { + "auxiliary_loss_clip": 0.0120747, + "auxiliary_loss_mlp": 0.0103864, + "balance_loss_clip": 1.02658856, + "balance_loss_mlp": 1.02729166, + "epoch": 0.12914086454638368, + "flos": 24170610850560.0, + "grad_norm": 1.7715328517101478, + "language_loss": 0.83285451, + "learning_rate": 3.897880934963007e-06, + "loss": 0.85531557, + "num_input_tokens_seen": 22750970, + "step": 1074, + "time_per_iteration": 2.6612048149108887 + }, + { + "auxiliary_loss_clip": 0.01180612, + "auxiliary_loss_mlp": 0.01035564, + "balance_loss_clip": 0.98735899, + "balance_loss_mlp": 1.02385187, + "epoch": 0.1292611074370228, + "flos": 20267157081600.0, + "grad_norm": 2.046578669197017, + "language_loss": 0.78495836, + "learning_rate": 3.89763506035154e-06, + "loss": 0.80712008, + "num_input_tokens_seen": 22768820, + "step": 1075, + "time_per_iteration": 2.6741812229156494 + }, + { + "auxiliary_loss_clip": 0.01197794, + "auxiliary_loss_mlp": 0.01033507, + "balance_loss_clip": 1.02563858, + "balance_loss_mlp": 1.02272487, + "epoch": 0.12938135032766188, + "flos": 27377684668800.0, + "grad_norm": 1.7181421021167484, + "language_loss": 0.81437838, + "learning_rate": 3.897388897870343e-06, + "loss": 0.83669138, + "num_input_tokens_seen": 22789460, + "step": 1076, + "time_per_iteration": 2.692469596862793 + }, + { + "auxiliary_loss_clip": 0.0119368, + "auxiliary_loss_mlp": 0.01044066, + "balance_loss_clip": 0.98957932, + "balance_loss_mlp": 1.03179967, + "epoch": 0.12950159321830096, + "flos": 29277107861760.0, + "grad_norm": 3.7034731181090437, + "language_loss": 0.74831629, + "learning_rate": 3.89714244755676e-06, + "loss": 0.77069372, + "num_input_tokens_seen": 22810820, + "step": 1077, + "time_per_iteration": 2.7097558975219727 + }, + { + "auxiliary_loss_clip": 0.01116072, + "auxiliary_loss_mlp": 0.01033991, + "balance_loss_clip": 0.90269667, + "balance_loss_mlp": 1.02272618, + "epoch": 0.12962183610894007, + "flos": 24534888629760.0, + "grad_norm": 2.522176193912765, + "language_loss": 0.86050826, + "learning_rate": 3.896895709448175e-06, + "loss": 0.88200891, + "num_input_tokens_seen": 22830570, + "step": 1078, + "time_per_iteration": 2.7407076358795166 + }, + { + "auxiliary_loss_clip": 0.01101837, + "auxiliary_loss_mlp": 0.0103886, + "balance_loss_clip": 0.86776388, + "balance_loss_mlp": 1.02684414, + "epoch": 0.12974207899957915, + "flos": 11215944552960.0, + "grad_norm": 2.2800247310414274, + "language_loss": 0.77200937, + "learning_rate": 3.896648683582019e-06, + "loss": 0.79341632, + "num_input_tokens_seen": 22845905, + "step": 1079, + "time_per_iteration": 2.7042407989501953 + }, + { + "auxiliary_loss_clip": 0.01135429, + "auxiliary_loss_mlp": 0.01044697, + "balance_loss_clip": 0.91419721, + "balance_loss_mlp": 1.03355134, + "epoch": 0.12986232189021824, + "flos": 24717889445760.0, + "grad_norm": 1.9059517465538485, + "language_loss": 0.80931163, + "learning_rate": 3.896401369995766e-06, + "loss": 0.83111286, + "num_input_tokens_seen": 22865710, + "step": 1080, + "time_per_iteration": 2.769090175628662 + }, + { + "auxiliary_loss_clip": 0.01232252, + "auxiliary_loss_mlp": 0.01048617, + "balance_loss_clip": 1.06818223, + "balance_loss_mlp": 1.03631473, + "epoch": 0.12998256478085732, + "flos": 23915357827200.0, + "grad_norm": 1.8094888231039992, + "language_loss": 0.79706317, + "learning_rate": 3.896153768726932e-06, + "loss": 0.81987184, + "num_input_tokens_seen": 22886020, + "step": 1081, + "time_per_iteration": 2.6141507625579834 + }, + { + "auxiliary_loss_clip": 0.01209565, + "auxiliary_loss_mlp": 0.01033209, + "balance_loss_clip": 1.03062403, + "balance_loss_mlp": 1.02242696, + "epoch": 0.13010280767149643, + "flos": 18624207974400.0, + "grad_norm": 2.7555153439000297, + "language_loss": 0.87989318, + "learning_rate": 3.8959058798130806e-06, + "loss": 0.90232086, + "num_input_tokens_seen": 22903995, + "step": 1082, + "time_per_iteration": 2.6964917182922363 + }, + { + "auxiliary_loss_clip": 0.01179136, + "auxiliary_loss_mlp": 0.00513929, + "balance_loss_clip": 0.98771691, + "balance_loss_mlp": 0.0, + "epoch": 0.1302230505621355, + "flos": 22783992174720.0, + "grad_norm": 1.6189606879665739, + "language_loss": 0.7510792, + "learning_rate": 3.895657703291814e-06, + "loss": 0.76800984, + "num_input_tokens_seen": 22924100, + "step": 1083, + "time_per_iteration": 2.7408509254455566 + }, + { + "auxiliary_loss_clip": 0.01190634, + "auxiliary_loss_mlp": 0.01037945, + "balance_loss_clip": 0.98762834, + "balance_loss_mlp": 1.02629209, + "epoch": 0.1303432934527746, + "flos": 21323612920320.0, + "grad_norm": 2.6573387620170075, + "language_loss": 0.79723179, + "learning_rate": 3.895409239200781e-06, + "loss": 0.81951755, + "num_input_tokens_seen": 22939985, + "step": 1084, + "time_per_iteration": 2.6693363189697266 + }, + { + "auxiliary_loss_clip": 0.01199224, + "auxiliary_loss_mlp": 0.01040409, + "balance_loss_clip": 1.02522349, + "balance_loss_mlp": 1.02938819, + "epoch": 0.1304635363434137, + "flos": 20922490765440.0, + "grad_norm": 2.096717509593563, + "language_loss": 0.90915227, + "learning_rate": 3.895160487577673e-06, + "loss": 0.9315486, + "num_input_tokens_seen": 22957555, + "step": 1085, + "time_per_iteration": 3.562619924545288 + }, + { + "auxiliary_loss_clip": 0.01099388, + "auxiliary_loss_mlp": 0.01000793, + "balance_loss_clip": 0.99027705, + "balance_loss_mlp": 0.99800378, + "epoch": 0.1305837792340528, + "flos": 63245659080960.0, + "grad_norm": 0.7893530913046187, + "language_loss": 0.60934985, + "learning_rate": 3.894911448460226e-06, + "loss": 0.63035166, + "num_input_tokens_seen": 23016870, + "step": 1086, + "time_per_iteration": 4.0354273319244385 + }, + { + "auxiliary_loss_clip": 0.01066489, + "auxiliary_loss_mlp": 0.01047633, + "balance_loss_clip": 0.8289178, + "balance_loss_mlp": 1.03665364, + "epoch": 0.13070402212469187, + "flos": 26428852955520.0, + "grad_norm": 1.7879209716102265, + "language_loss": 0.72631609, + "learning_rate": 3.8946621218862195e-06, + "loss": 0.74745733, + "num_input_tokens_seen": 23037870, + "step": 1087, + "time_per_iteration": 3.790161609649658 + }, + { + "auxiliary_loss_clip": 0.0115524, + "auxiliary_loss_mlp": 0.01037689, + "balance_loss_clip": 0.95002699, + "balance_loss_mlp": 1.0267992, + "epoch": 0.13082426501533098, + "flos": 27673409341440.0, + "grad_norm": 2.109313081284278, + "language_loss": 0.88792169, + "learning_rate": 3.894412507893475e-06, + "loss": 0.90985096, + "num_input_tokens_seen": 23058150, + "step": 1088, + "time_per_iteration": 2.763751983642578 + }, + { + "auxiliary_loss_clip": 0.01137727, + "auxiliary_loss_mlp": 0.01044261, + "balance_loss_clip": 0.9074071, + "balance_loss_mlp": 1.03274584, + "epoch": 0.13094450790597006, + "flos": 24826770547200.0, + "grad_norm": 2.1150495800561555, + "language_loss": 0.7191627, + "learning_rate": 3.894162606519859e-06, + "loss": 0.74098265, + "num_input_tokens_seen": 23077100, + "step": 1089, + "time_per_iteration": 3.6800990104675293 + }, + { + "auxiliary_loss_clip": 0.01132682, + "auxiliary_loss_mlp": 0.01052216, + "balance_loss_clip": 0.91329592, + "balance_loss_mlp": 1.04102814, + "epoch": 0.13106475079660915, + "flos": 19062605468160.0, + "grad_norm": 1.7426740840884538, + "language_loss": 0.77139962, + "learning_rate": 3.893912417803282e-06, + "loss": 0.79324865, + "num_input_tokens_seen": 23096815, + "step": 1090, + "time_per_iteration": 2.722031831741333 + }, + { + "auxiliary_loss_clip": 0.01133105, + "auxiliary_loss_mlp": 0.01034257, + "balance_loss_clip": 0.90718496, + "balance_loss_mlp": 1.02287257, + "epoch": 0.13118499368724823, + "flos": 28913189218560.0, + "grad_norm": 1.8886299549196244, + "language_loss": 0.7699579, + "learning_rate": 3.8936619417816975e-06, + "loss": 0.79163152, + "num_input_tokens_seen": 23117145, + "step": 1091, + "time_per_iteration": 2.825171947479248 + }, + { + "auxiliary_loss_clip": 0.01155018, + "auxiliary_loss_mlp": 0.01041143, + "balance_loss_clip": 0.95213437, + "balance_loss_mlp": 1.02967501, + "epoch": 0.13130523657788734, + "flos": 14283398206080.0, + "grad_norm": 1.7590795924922302, + "language_loss": 0.71230477, + "learning_rate": 3.8934111784931015e-06, + "loss": 0.7342664, + "num_input_tokens_seen": 23134595, + "step": 1092, + "time_per_iteration": 2.688918352127075 + }, + { + "auxiliary_loss_clip": 0.01081821, + "auxiliary_loss_mlp": 0.01002498, + "balance_loss_clip": 0.95386064, + "balance_loss_mlp": 0.99948198, + "epoch": 0.13142547946852642, + "flos": 70174155519360.0, + "grad_norm": 0.9123090681137145, + "language_loss": 0.59043252, + "learning_rate": 3.893160127975535e-06, + "loss": 0.61127567, + "num_input_tokens_seen": 23195285, + "step": 1093, + "time_per_iteration": 3.3096842765808105 + }, + { + "auxiliary_loss_clip": 0.01137783, + "auxiliary_loss_mlp": 0.01034623, + "balance_loss_clip": 0.90906495, + "balance_loss_mlp": 1.0240252, + "epoch": 0.1315457223591655, + "flos": 45805998844800.0, + "grad_norm": 2.375259978117548, + "language_loss": 0.81419301, + "learning_rate": 3.8929087902670826e-06, + "loss": 0.835917, + "num_input_tokens_seen": 23216915, + "step": 1094, + "time_per_iteration": 2.9609768390655518 + }, + { + "auxiliary_loss_clip": 0.01103287, + "auxiliary_loss_mlp": 0.01004448, + "balance_loss_clip": 0.99094987, + "balance_loss_mlp": 1.00143242, + "epoch": 0.13166596524980462, + "flos": 62881165820160.0, + "grad_norm": 0.9794114631182638, + "language_loss": 0.60680759, + "learning_rate": 3.8926571654058715e-06, + "loss": 0.62788492, + "num_input_tokens_seen": 23273560, + "step": 1095, + "time_per_iteration": 3.1397790908813477 + }, + { + "auxiliary_loss_clip": 0.01155568, + "auxiliary_loss_mlp": 0.01040654, + "balance_loss_clip": 0.94952297, + "balance_loss_mlp": 1.0301038, + "epoch": 0.1317862081404437, + "flos": 23586523793280.0, + "grad_norm": 2.3753899559296463, + "language_loss": 0.76897031, + "learning_rate": 3.892405253430074e-06, + "loss": 0.79093254, + "num_input_tokens_seen": 23291080, + "step": 1096, + "time_per_iteration": 2.734705686569214 + }, + { + "auxiliary_loss_clip": 0.01186089, + "auxiliary_loss_mlp": 0.00514042, + "balance_loss_clip": 0.9899826, + "balance_loss_mlp": 0.0, + "epoch": 0.13190645103108278, + "flos": 20260764460800.0, + "grad_norm": 2.031113977964139, + "language_loss": 0.82398981, + "learning_rate": 3.892153054377904e-06, + "loss": 0.84099114, + "num_input_tokens_seen": 23308485, + "step": 1097, + "time_per_iteration": 2.669485569000244 + }, + { + "auxiliary_loss_clip": 0.01026552, + "auxiliary_loss_mlp": 0.01029119, + "balance_loss_clip": 0.85674167, + "balance_loss_mlp": 1.02594757, + "epoch": 0.13202669392172187, + "flos": 53455440136320.0, + "grad_norm": 0.934825221134072, + "language_loss": 0.59398258, + "learning_rate": 3.891900568287619e-06, + "loss": 0.61453927, + "num_input_tokens_seen": 23360870, + "step": 1098, + "time_per_iteration": 3.17142653465271 + }, + { + "auxiliary_loss_clip": 0.01163001, + "auxiliary_loss_mlp": 0.01037259, + "balance_loss_clip": 0.95108849, + "balance_loss_mlp": 1.02521896, + "epoch": 0.13214693681236098, + "flos": 15851293845120.0, + "grad_norm": 2.4836695363249452, + "language_loss": 0.72398317, + "learning_rate": 3.891647795197523e-06, + "loss": 0.74598575, + "num_input_tokens_seen": 23376910, + "step": 1099, + "time_per_iteration": 2.710998296737671 + }, + { + "auxiliary_loss_clip": 0.01162593, + "auxiliary_loss_mlp": 0.01041211, + "balance_loss_clip": 0.94763047, + "balance_loss_mlp": 1.02899194, + "epoch": 0.13226717970300006, + "flos": 19353840940800.0, + "grad_norm": 1.9587940992402515, + "language_loss": 0.68893039, + "learning_rate": 3.8913947351459605e-06, + "loss": 0.71096849, + "num_input_tokens_seen": 23394450, + "step": 1100, + "time_per_iteration": 2.7975659370422363 + }, + { + "auxiliary_loss_clip": 0.01227971, + "auxiliary_loss_mlp": 0.01036993, + "balance_loss_clip": 1.06500864, + "balance_loss_mlp": 1.02669311, + "epoch": 0.13238742259363914, + "flos": 20698084546560.0, + "grad_norm": 2.2672850679826695, + "language_loss": 0.68022799, + "learning_rate": 3.89114138817132e-06, + "loss": 0.70287758, + "num_input_tokens_seen": 23411115, + "step": 1101, + "time_per_iteration": 2.6152403354644775 + }, + { + "auxiliary_loss_clip": 0.01207947, + "auxiliary_loss_mlp": 0.01035503, + "balance_loss_clip": 1.03109169, + "balance_loss_mlp": 1.02460742, + "epoch": 0.13250766548427825, + "flos": 21032449274880.0, + "grad_norm": 1.7973642443419833, + "language_loss": 0.84411442, + "learning_rate": 3.890887754312035e-06, + "loss": 0.86654902, + "num_input_tokens_seen": 23429360, + "step": 1102, + "time_per_iteration": 2.684920310974121 + }, + { + "auxiliary_loss_clip": 0.0117529, + "auxiliary_loss_mlp": 0.01041078, + "balance_loss_clip": 0.98475158, + "balance_loss_mlp": 1.02975869, + "epoch": 0.13262790837491734, + "flos": 22637871648000.0, + "grad_norm": 2.785098924980659, + "language_loss": 0.87543899, + "learning_rate": 3.890633833606581e-06, + "loss": 0.89760268, + "num_input_tokens_seen": 23449050, + "step": 1103, + "time_per_iteration": 2.727630853652954 + }, + { + "auxiliary_loss_clip": 0.01206379, + "auxiliary_loss_mlp": 0.01047556, + "balance_loss_clip": 1.0335474, + "balance_loss_mlp": 1.03746498, + "epoch": 0.13274815126555642, + "flos": 19683141851520.0, + "grad_norm": 2.0598899883548674, + "language_loss": 0.6987884, + "learning_rate": 3.890379626093477e-06, + "loss": 0.72132778, + "num_input_tokens_seen": 23468800, + "step": 1104, + "time_per_iteration": 2.632444381713867 + }, + { + "auxiliary_loss_clip": 0.01118343, + "auxiliary_loss_mlp": 0.01044919, + "balance_loss_clip": 0.90744936, + "balance_loss_mlp": 1.03270638, + "epoch": 0.1328683941561955, + "flos": 21317687176320.0, + "grad_norm": 2.3663886423542158, + "language_loss": 0.92589211, + "learning_rate": 3.890125131811287e-06, + "loss": 0.94752473, + "num_input_tokens_seen": 23486850, + "step": 1105, + "time_per_iteration": 2.7335081100463867 + }, + { + "auxiliary_loss_clip": 0.01169888, + "auxiliary_loss_mlp": 0.01036075, + "balance_loss_clip": 0.98254383, + "balance_loss_mlp": 1.02553129, + "epoch": 0.1329886370468346, + "flos": 13699131580800.0, + "grad_norm": 2.0138486811937804, + "language_loss": 0.75378895, + "learning_rate": 3.889870350798618e-06, + "loss": 0.77584863, + "num_input_tokens_seen": 23504195, + "step": 1106, + "time_per_iteration": 2.633702278137207 + }, + { + "auxiliary_loss_clip": 0.01228995, + "auxiliary_loss_mlp": 0.0104265, + "balance_loss_clip": 1.06477118, + "balance_loss_mlp": 1.0310744, + "epoch": 0.1331088799374737, + "flos": 21032413361280.0, + "grad_norm": 1.8422241910805244, + "language_loss": 0.78372377, + "learning_rate": 3.889615283094119e-06, + "loss": 0.80644023, + "num_input_tokens_seen": 23523385, + "step": 1107, + "time_per_iteration": 2.6314728260040283 + }, + { + "auxiliary_loss_clip": 0.01234463, + "auxiliary_loss_mlp": 0.0104423, + "balance_loss_clip": 1.06606829, + "balance_loss_mlp": 1.0329051, + "epoch": 0.13322912282811278, + "flos": 18260432985600.0, + "grad_norm": 2.7704888835160273, + "language_loss": 0.84683132, + "learning_rate": 3.889359928736485e-06, + "loss": 0.86961818, + "num_input_tokens_seen": 23541330, + "step": 1108, + "time_per_iteration": 2.5637474060058594 + }, + { + "auxiliary_loss_clip": 0.01176007, + "auxiliary_loss_mlp": 0.00513452, + "balance_loss_clip": 0.98932797, + "balance_loss_mlp": 0.0, + "epoch": 0.1333493657187519, + "flos": 24460876656000.0, + "grad_norm": 2.3351911790706392, + "language_loss": 0.90963531, + "learning_rate": 3.889104287764451e-06, + "loss": 0.92652988, + "num_input_tokens_seen": 23561705, + "step": 1109, + "time_per_iteration": 2.7017786502838135 + }, + { + "auxiliary_loss_clip": 0.01181165, + "auxiliary_loss_mlp": 0.01041933, + "balance_loss_clip": 0.99086976, + "balance_loss_mlp": 1.03104949, + "epoch": 0.13346960860939097, + "flos": 22158930677760.0, + "grad_norm": 1.8854302587819372, + "language_loss": 0.90645778, + "learning_rate": 3.888848360216798e-06, + "loss": 0.92868876, + "num_input_tokens_seen": 23579350, + "step": 1110, + "time_per_iteration": 2.6532139778137207 + }, + { + "auxiliary_loss_clip": 0.01081965, + "auxiliary_loss_mlp": 0.01023095, + "balance_loss_clip": 0.95275795, + "balance_loss_mlp": 1.01980448, + "epoch": 0.13358985150003005, + "flos": 67931212608000.0, + "grad_norm": 0.9355823746310082, + "language_loss": 0.56622607, + "learning_rate": 3.888592146132351e-06, + "loss": 0.58727664, + "num_input_tokens_seen": 23640620, + "step": 1111, + "time_per_iteration": 3.372356414794922 + }, + { + "auxiliary_loss_clip": 0.01206902, + "auxiliary_loss_mlp": 0.01035532, + "balance_loss_clip": 1.03109407, + "balance_loss_mlp": 1.02509546, + "epoch": 0.13371009439066917, + "flos": 26834284742400.0, + "grad_norm": 1.9008741309139587, + "language_loss": 0.78429925, + "learning_rate": 3.888335645549978e-06, + "loss": 0.80672359, + "num_input_tokens_seen": 23661040, + "step": 1112, + "time_per_iteration": 4.620602369308472 + }, + { + "auxiliary_loss_clip": 0.01231606, + "auxiliary_loss_mlp": 0.01039264, + "balance_loss_clip": 1.06794119, + "balance_loss_mlp": 1.0272119, + "epoch": 0.13383033728130825, + "flos": 26322844942080.0, + "grad_norm": 2.2221978240848927, + "language_loss": 0.82013804, + "learning_rate": 3.888078858508588e-06, + "loss": 0.84284675, + "num_input_tokens_seen": 23680900, + "step": 1113, + "time_per_iteration": 3.6190009117126465 + }, + { + "auxiliary_loss_clip": 0.01180337, + "auxiliary_loss_mlp": 0.01041397, + "balance_loss_clip": 0.99141628, + "balance_loss_mlp": 1.03005397, + "epoch": 0.13395058017194733, + "flos": 22563931501440.0, + "grad_norm": 1.9600896278418138, + "language_loss": 0.84554482, + "learning_rate": 3.8878217850471365e-06, + "loss": 0.86776221, + "num_input_tokens_seen": 23700815, + "step": 1114, + "time_per_iteration": 2.6551096439361572 + }, + { + "auxiliary_loss_clip": 0.01236656, + "auxiliary_loss_mlp": 0.01046902, + "balance_loss_clip": 1.07098448, + "balance_loss_mlp": 1.03418255, + "epoch": 0.13407082306258641, + "flos": 25810938264960.0, + "grad_norm": 1.8710895947080926, + "language_loss": 0.73670524, + "learning_rate": 3.887564425204621e-06, + "loss": 0.7595408, + "num_input_tokens_seen": 23722500, + "step": 1115, + "time_per_iteration": 3.551236629486084 + }, + { + "auxiliary_loss_clip": 0.01051212, + "auxiliary_loss_mlp": 0.01008514, + "balance_loss_clip": 0.9129529, + "balance_loss_mlp": 1.0053196, + "epoch": 0.13419106595322552, + "flos": 68338365269760.0, + "grad_norm": 0.8426382548509532, + "language_loss": 0.54671973, + "learning_rate": 3.887306779020083e-06, + "loss": 0.56731701, + "num_input_tokens_seen": 23777155, + "step": 1116, + "time_per_iteration": 3.1950790882110596 + }, + { + "auxiliary_loss_clip": 0.01210986, + "auxiliary_loss_mlp": 0.01045259, + "balance_loss_clip": 1.03130043, + "balance_loss_mlp": 1.03421426, + "epoch": 0.1343113088438646, + "flos": 20449080489600.0, + "grad_norm": 2.176295278606443, + "language_loss": 0.70698768, + "learning_rate": 3.887048846532608e-06, + "loss": 0.72955012, + "num_input_tokens_seen": 23794130, + "step": 1117, + "time_per_iteration": 2.6236751079559326 + }, + { + "auxiliary_loss_clip": 0.01057939, + "auxiliary_loss_mlp": 0.01001264, + "balance_loss_clip": 0.91422522, + "balance_loss_mlp": 0.99806941, + "epoch": 0.1344315517345037, + "flos": 67389784951680.0, + "grad_norm": 0.7554697251537438, + "language_loss": 0.58093947, + "learning_rate": 3.8867906277813224e-06, + "loss": 0.60153145, + "num_input_tokens_seen": 23852285, + "step": 1118, + "time_per_iteration": 3.103220224380493 + }, + { + "auxiliary_loss_clip": 0.01210577, + "auxiliary_loss_mlp": 0.00513176, + "balance_loss_clip": 1.02747416, + "balance_loss_mlp": 0.0, + "epoch": 0.1345517946251428, + "flos": 40734442788480.0, + "grad_norm": 2.115891822615735, + "language_loss": 0.73404539, + "learning_rate": 3.886532122805399e-06, + "loss": 0.75128293, + "num_input_tokens_seen": 23874765, + "step": 1119, + "time_per_iteration": 2.8159067630767822 + }, + { + "auxiliary_loss_clip": 0.01101002, + "auxiliary_loss_mlp": 0.01041892, + "balance_loss_clip": 0.86872423, + "balance_loss_mlp": 1.03022718, + "epoch": 0.13467203751578188, + "flos": 22816850140800.0, + "grad_norm": 1.8653292120361427, + "language_loss": 0.898193, + "learning_rate": 3.886273331644053e-06, + "loss": 0.91962194, + "num_input_tokens_seen": 23893635, + "step": 1120, + "time_per_iteration": 2.8687245845794678 + }, + { + "auxiliary_loss_clip": 0.01130815, + "auxiliary_loss_mlp": 0.01034835, + "balance_loss_clip": 0.91326857, + "balance_loss_mlp": 1.02426088, + "epoch": 0.13479228040642097, + "flos": 17091576512640.0, + "grad_norm": 2.4549550187360882, + "language_loss": 0.82200325, + "learning_rate": 3.886014254336542e-06, + "loss": 0.84365976, + "num_input_tokens_seen": 23910110, + "step": 1121, + "time_per_iteration": 2.6889450550079346 + }, + { + "auxiliary_loss_clip": 0.01204754, + "auxiliary_loss_mlp": 0.01051264, + "balance_loss_clip": 1.02720284, + "balance_loss_mlp": 1.03998649, + "epoch": 0.13491252329706005, + "flos": 23730525417600.0, + "grad_norm": 1.6233581237021752, + "language_loss": 0.92648757, + "learning_rate": 3.885754890922168e-06, + "loss": 0.94904774, + "num_input_tokens_seen": 23930440, + "step": 1122, + "time_per_iteration": 2.6951255798339844 + }, + { + "auxiliary_loss_clip": 0.01075518, + "auxiliary_loss_mlp": 0.01053057, + "balance_loss_clip": 0.83245063, + "balance_loss_mlp": 1.04211402, + "epoch": 0.13503276618769916, + "flos": 34127058960000.0, + "grad_norm": 1.7558820610408892, + "language_loss": 0.78394079, + "learning_rate": 3.885495241440277e-06, + "loss": 0.80522656, + "num_input_tokens_seen": 23954535, + "step": 1123, + "time_per_iteration": 3.0642940998077393 + }, + { + "auxiliary_loss_clip": 0.01233869, + "auxiliary_loss_mlp": 0.01041887, + "balance_loss_clip": 1.06808686, + "balance_loss_mlp": 1.03125966, + "epoch": 0.13515300907833824, + "flos": 17712328377600.0, + "grad_norm": 1.8116977656053348, + "language_loss": 0.74132657, + "learning_rate": 3.885235305930257e-06, + "loss": 0.76408416, + "num_input_tokens_seen": 23972735, + "step": 1124, + "time_per_iteration": 2.7219924926757812 + }, + { + "auxiliary_loss_clip": 0.01156329, + "auxiliary_loss_mlp": 0.0104619, + "balance_loss_clip": 0.95290112, + "balance_loss_mlp": 1.03358948, + "epoch": 0.13527325196897733, + "flos": 20260872201600.0, + "grad_norm": 1.8425800896928266, + "language_loss": 0.85392404, + "learning_rate": 3.884975084431539e-06, + "loss": 0.8759492, + "num_input_tokens_seen": 23987685, + "step": 1125, + "time_per_iteration": 2.7124247550964355 + }, + { + "auxiliary_loss_clip": 0.01197672, + "auxiliary_loss_mlp": 0.0051354, + "balance_loss_clip": 1.02609301, + "balance_loss_mlp": 0.0, + "epoch": 0.13539349485961644, + "flos": 18186492839040.0, + "grad_norm": 2.6552504311233602, + "language_loss": 0.91687179, + "learning_rate": 3.8847145769836e-06, + "loss": 0.93398392, + "num_input_tokens_seen": 24004105, + "step": 1126, + "time_per_iteration": 2.660663604736328 + }, + { + "auxiliary_loss_clip": 0.01234617, + "auxiliary_loss_mlp": 0.0104707, + "balance_loss_clip": 1.06816292, + "balance_loss_mlp": 1.03522658, + "epoch": 0.13551373775025552, + "flos": 19317463441920.0, + "grad_norm": 2.230930027831777, + "language_loss": 0.66225904, + "learning_rate": 3.884453783625959e-06, + "loss": 0.68507588, + "num_input_tokens_seen": 24021715, + "step": 1127, + "time_per_iteration": 2.6293740272521973 + }, + { + "auxiliary_loss_clip": 0.01180011, + "auxiliary_loss_mlp": 0.01050458, + "balance_loss_clip": 0.99067247, + "balance_loss_mlp": 1.03919327, + "epoch": 0.1356339806408946, + "flos": 20850813175680.0, + "grad_norm": 3.3609853518110167, + "language_loss": 0.84907913, + "learning_rate": 3.884192704398176e-06, + "loss": 0.87138391, + "num_input_tokens_seen": 24038915, + "step": 1128, + "time_per_iteration": 2.684251546859741 + }, + { + "auxiliary_loss_clip": 0.01207961, + "auxiliary_loss_mlp": 0.01047084, + "balance_loss_clip": 1.02698898, + "balance_loss_mlp": 1.0358485, + "epoch": 0.13575422353153369, + "flos": 50476037696640.0, + "grad_norm": 1.6121291407924767, + "language_loss": 0.74665171, + "learning_rate": 3.883931339339858e-06, + "loss": 0.76920217, + "num_input_tokens_seen": 24063300, + "step": 1129, + "time_per_iteration": 2.9102399349212646 + }, + { + "auxiliary_loss_clip": 0.01213361, + "auxiliary_loss_mlp": 0.0103786, + "balance_loss_clip": 1.02992535, + "balance_loss_mlp": 1.02617109, + "epoch": 0.1358744664221728, + "flos": 18150797698560.0, + "grad_norm": 1.96695721900702, + "language_loss": 0.78931189, + "learning_rate": 3.883669688490654e-06, + "loss": 0.81182408, + "num_input_tokens_seen": 24081070, + "step": 1130, + "time_per_iteration": 2.646435022354126 + }, + { + "auxiliary_loss_clip": 0.01169786, + "auxiliary_loss_mlp": 0.00513232, + "balance_loss_clip": 0.98554349, + "balance_loss_mlp": 0.0, + "epoch": 0.13599470931281188, + "flos": 18442966924800.0, + "grad_norm": 2.169200901835317, + "language_loss": 0.85381842, + "learning_rate": 3.883407751890256e-06, + "loss": 0.87064862, + "num_input_tokens_seen": 24099675, + "step": 1131, + "time_per_iteration": 2.660033941268921 + }, + { + "auxiliary_loss_clip": 0.01149582, + "auxiliary_loss_mlp": 0.01042267, + "balance_loss_clip": 0.94401622, + "balance_loss_mlp": 1.03063798, + "epoch": 0.13611495220345096, + "flos": 26680766014080.0, + "grad_norm": 1.6753574199645493, + "language_loss": 0.85903084, + "learning_rate": 3.8831455295783994e-06, + "loss": 0.88094938, + "num_input_tokens_seen": 24118925, + "step": 1132, + "time_per_iteration": 2.746124267578125 + }, + { + "auxiliary_loss_clip": 0.01175345, + "auxiliary_loss_mlp": 0.01042438, + "balance_loss_clip": 0.98765343, + "balance_loss_mlp": 1.0318346, + "epoch": 0.13623519509409007, + "flos": 21686238673920.0, + "grad_norm": 2.6571454302489848, + "language_loss": 0.74222267, + "learning_rate": 3.882883021594864e-06, + "loss": 0.76440048, + "num_input_tokens_seen": 24137065, + "step": 1133, + "time_per_iteration": 2.706056594848633 + }, + { + "auxiliary_loss_clip": 0.01151684, + "auxiliary_loss_mlp": 0.01048371, + "balance_loss_clip": 0.95085835, + "balance_loss_mlp": 1.03740931, + "epoch": 0.13635543798472916, + "flos": 14830389492480.0, + "grad_norm": 5.630384233384955, + "language_loss": 0.86763322, + "learning_rate": 3.8826202279794705e-06, + "loss": 0.88963383, + "num_input_tokens_seen": 24154125, + "step": 1134, + "time_per_iteration": 2.705895185470581 + }, + { + "auxiliary_loss_clip": 0.0123575, + "auxiliary_loss_mlp": 0.01046347, + "balance_loss_clip": 1.07030392, + "balance_loss_mlp": 1.03523052, + "epoch": 0.13647568087536824, + "flos": 22890323410560.0, + "grad_norm": 2.2255079642023383, + "language_loss": 0.70398164, + "learning_rate": 3.882357148772085e-06, + "loss": 0.72680265, + "num_input_tokens_seen": 24171550, + "step": 1135, + "time_per_iteration": 2.6062376499176025 + }, + { + "auxiliary_loss_clip": 0.01142277, + "auxiliary_loss_mlp": 0.01042948, + "balance_loss_clip": 0.94788074, + "balance_loss_mlp": 1.03118849, + "epoch": 0.13659592376600732, + "flos": 19937927998080.0, + "grad_norm": 2.653512044198647, + "language_loss": 0.84430718, + "learning_rate": 3.882093784012617e-06, + "loss": 0.86615944, + "num_input_tokens_seen": 24190190, + "step": 1136, + "time_per_iteration": 2.742673635482788 + }, + { + "auxiliary_loss_clip": 0.01178758, + "auxiliary_loss_mlp": 0.01036641, + "balance_loss_clip": 0.98780155, + "balance_loss_mlp": 1.02530408, + "epoch": 0.13671616665664643, + "flos": 21428579439360.0, + "grad_norm": 1.7025020392955097, + "language_loss": 0.84065062, + "learning_rate": 3.881830133741019e-06, + "loss": 0.86280465, + "num_input_tokens_seen": 24209055, + "step": 1137, + "time_per_iteration": 4.503746747970581 + }, + { + "auxiliary_loss_clip": 0.01158843, + "auxiliary_loss_mlp": 0.0104793, + "balance_loss_clip": 0.95512521, + "balance_loss_mlp": 1.03678989, + "epoch": 0.13683640954728551, + "flos": 22778138257920.0, + "grad_norm": 2.4632091198527535, + "language_loss": 0.76416683, + "learning_rate": 3.881566197997285e-06, + "loss": 0.7862345, + "num_input_tokens_seen": 24225490, + "step": 1138, + "time_per_iteration": 2.7363686561584473 + }, + { + "auxiliary_loss_clip": 0.01181225, + "auxiliary_loss_mlp": 0.01039384, + "balance_loss_clip": 0.99260092, + "balance_loss_mlp": 1.0290482, + "epoch": 0.1369566524379246, + "flos": 21725884310400.0, + "grad_norm": 1.487676540347296, + "language_loss": 0.74768907, + "learning_rate": 3.881301976821456e-06, + "loss": 0.7698952, + "num_input_tokens_seen": 24245520, + "step": 1139, + "time_per_iteration": 3.5675861835479736 + }, + { + "auxiliary_loss_clip": 0.01206168, + "auxiliary_loss_mlp": 0.01045226, + "balance_loss_clip": 1.03032827, + "balance_loss_mlp": 1.03385949, + "epoch": 0.1370768953285637, + "flos": 18624459369600.0, + "grad_norm": 3.6544023117975324, + "language_loss": 0.91077459, + "learning_rate": 3.881037470253612e-06, + "loss": 0.93328851, + "num_input_tokens_seen": 24265035, + "step": 1140, + "time_per_iteration": 2.667377471923828 + }, + { + "auxiliary_loss_clip": 0.01134015, + "auxiliary_loss_mlp": 0.01051409, + "balance_loss_clip": 0.91383404, + "balance_loss_mlp": 1.04013801, + "epoch": 0.1371971382192028, + "flos": 14939521989120.0, + "grad_norm": 5.064929880726093, + "language_loss": 0.79590023, + "learning_rate": 3.88077267833388e-06, + "loss": 0.81775439, + "num_input_tokens_seen": 24281550, + "step": 1141, + "time_per_iteration": 3.629676103591919 + }, + { + "auxiliary_loss_clip": 0.01124673, + "auxiliary_loss_mlp": 0.01041479, + "balance_loss_clip": 0.90988684, + "balance_loss_mlp": 1.03069687, + "epoch": 0.13731738110984187, + "flos": 19023785844480.0, + "grad_norm": 2.00367528183043, + "language_loss": 0.83697635, + "learning_rate": 3.880507601102427e-06, + "loss": 0.85863793, + "num_input_tokens_seen": 24299485, + "step": 1142, + "time_per_iteration": 2.8511171340942383 + }, + { + "auxiliary_loss_clip": 0.01232037, + "auxiliary_loss_mlp": 0.01041736, + "balance_loss_clip": 1.06939459, + "balance_loss_mlp": 1.03131723, + "epoch": 0.13743762400048098, + "flos": 18187462506240.0, + "grad_norm": 1.7952672977016058, + "language_loss": 0.82208788, + "learning_rate": 3.880242238599467e-06, + "loss": 0.84482557, + "num_input_tokens_seen": 24316010, + "step": 1143, + "time_per_iteration": 2.61629581451416 + }, + { + "auxiliary_loss_clip": 0.01226246, + "auxiliary_loss_mlp": 0.01043329, + "balance_loss_clip": 1.0647974, + "balance_loss_mlp": 1.03182578, + "epoch": 0.13755786689112007, + "flos": 21031982398080.0, + "grad_norm": 1.6007259200456938, + "language_loss": 0.83218598, + "learning_rate": 3.879976590865254e-06, + "loss": 0.85488176, + "num_input_tokens_seen": 24335465, + "step": 1144, + "time_per_iteration": 2.655395746231079 + }, + { + "auxiliary_loss_clip": 0.01182143, + "auxiliary_loss_mlp": 0.01040635, + "balance_loss_clip": 0.99211675, + "balance_loss_mlp": 1.03015113, + "epoch": 0.13767810978175915, + "flos": 21360636864000.0, + "grad_norm": 1.9636819441456073, + "language_loss": 0.87442505, + "learning_rate": 3.879710657940087e-06, + "loss": 0.89665288, + "num_input_tokens_seen": 24354415, + "step": 1145, + "time_per_iteration": 2.6860077381134033 + }, + { + "auxiliary_loss_clip": 0.01208626, + "auxiliary_loss_mlp": 0.01040778, + "balance_loss_clip": 1.02886879, + "balance_loss_mlp": 1.02936935, + "epoch": 0.13779835267239823, + "flos": 30592084861440.0, + "grad_norm": 1.7405989412502214, + "language_loss": 0.70251155, + "learning_rate": 3.879444439864308e-06, + "loss": 0.72500563, + "num_input_tokens_seen": 24373990, + "step": 1146, + "time_per_iteration": 2.752629518508911 + }, + { + "auxiliary_loss_clip": 0.01203752, + "auxiliary_loss_mlp": 0.00513049, + "balance_loss_clip": 1.02606487, + "balance_loss_mlp": 0.0, + "epoch": 0.13791859556303734, + "flos": 22669867687680.0, + "grad_norm": 1.9046942903956157, + "language_loss": 0.86052549, + "learning_rate": 3.879177936678301e-06, + "loss": 0.87769347, + "num_input_tokens_seen": 24392995, + "step": 1147, + "time_per_iteration": 2.690502405166626 + }, + { + "auxiliary_loss_clip": 0.01184727, + "auxiliary_loss_mlp": 0.01041665, + "balance_loss_clip": 0.98916656, + "balance_loss_mlp": 1.03028655, + "epoch": 0.13803883845367643, + "flos": 35224166016000.0, + "grad_norm": 1.8473707504024266, + "language_loss": 0.77172774, + "learning_rate": 3.878911148422496e-06, + "loss": 0.79399163, + "num_input_tokens_seen": 24414470, + "step": 1148, + "time_per_iteration": 2.863798141479492 + }, + { + "auxiliary_loss_clip": 0.01203332, + "auxiliary_loss_mlp": 0.01037162, + "balance_loss_clip": 1.02584398, + "balance_loss_mlp": 1.0257715, + "epoch": 0.1381590813443155, + "flos": 32014542332160.0, + "grad_norm": 4.019455773856798, + "language_loss": 0.70361578, + "learning_rate": 3.878644075137364e-06, + "loss": 0.72602069, + "num_input_tokens_seen": 24435120, + "step": 1149, + "time_per_iteration": 2.7315547466278076 + }, + { + "auxiliary_loss_clip": 0.01134409, + "auxiliary_loss_mlp": 0.01043664, + "balance_loss_clip": 0.94292086, + "balance_loss_mlp": 1.0325774, + "epoch": 0.13827932423495462, + "flos": 17821855923840.0, + "grad_norm": 1.9070025167776514, + "language_loss": 0.79133207, + "learning_rate": 3.878376716863418e-06, + "loss": 0.8131128, + "num_input_tokens_seen": 24451420, + "step": 1150, + "time_per_iteration": 2.713020086288452 + }, + { + "auxiliary_loss_clip": 0.01174986, + "auxiliary_loss_mlp": 0.01034845, + "balance_loss_clip": 0.98569608, + "balance_loss_mlp": 1.0243603, + "epoch": 0.1383995671255937, + "flos": 19427098728960.0, + "grad_norm": 2.1657038146978227, + "language_loss": 0.71605307, + "learning_rate": 3.878109073641219e-06, + "loss": 0.73815143, + "num_input_tokens_seen": 24470450, + "step": 1151, + "time_per_iteration": 2.6097841262817383 + }, + { + "auxiliary_loss_clip": 0.01128112, + "auxiliary_loss_mlp": 0.0103968, + "balance_loss_clip": 0.91141093, + "balance_loss_mlp": 1.02904081, + "epoch": 0.13851981001623279, + "flos": 28296603331200.0, + "grad_norm": 1.4954417882064592, + "language_loss": 0.81259161, + "learning_rate": 3.877841145511366e-06, + "loss": 0.83426952, + "num_input_tokens_seen": 24493190, + "step": 1152, + "time_per_iteration": 2.8298568725585938 + }, + { + "auxiliary_loss_clip": 0.01207459, + "auxiliary_loss_mlp": 0.01040758, + "balance_loss_clip": 1.02692139, + "balance_loss_mlp": 1.02954602, + "epoch": 0.13864005290687187, + "flos": 21213079793280.0, + "grad_norm": 1.7991509032802508, + "language_loss": 0.82707232, + "learning_rate": 3.8775729325145035e-06, + "loss": 0.84955448, + "num_input_tokens_seen": 24512425, + "step": 1153, + "time_per_iteration": 2.649228811264038 + }, + { + "auxiliary_loss_clip": 0.01046669, + "auxiliary_loss_mlp": 0.0103308, + "balance_loss_clip": 0.92008048, + "balance_loss_mlp": 1.03049278, + "epoch": 0.13876029579751098, + "flos": 71653389413760.0, + "grad_norm": 0.7918885187157304, + "language_loss": 0.64779449, + "learning_rate": 3.877304434691321e-06, + "loss": 0.66859198, + "num_input_tokens_seen": 24579275, + "step": 1154, + "time_per_iteration": 3.4475536346435547 + }, + { + "auxiliary_loss_clip": 0.01154045, + "auxiliary_loss_mlp": 0.01043822, + "balance_loss_clip": 0.95058191, + "balance_loss_mlp": 1.03360546, + "epoch": 0.13888053868815006, + "flos": 21941348042880.0, + "grad_norm": 1.9613839393439372, + "language_loss": 0.79530501, + "learning_rate": 3.877035652082548e-06, + "loss": 0.81728369, + "num_input_tokens_seen": 24598720, + "step": 1155, + "time_per_iteration": 2.793015241622925 + }, + { + "auxiliary_loss_clip": 0.01174093, + "auxiliary_loss_mlp": 0.01046124, + "balance_loss_clip": 0.98852605, + "balance_loss_mlp": 1.03452539, + "epoch": 0.13900078157878915, + "flos": 19608627087360.0, + "grad_norm": 1.8061090871558598, + "language_loss": 0.85388952, + "learning_rate": 3.87676658472896e-06, + "loss": 0.87609172, + "num_input_tokens_seen": 24617530, + "step": 1156, + "time_per_iteration": 2.668195962905884 + }, + { + "auxiliary_loss_clip": 0.01203507, + "auxiliary_loss_mlp": 0.01041168, + "balance_loss_clip": 1.02433467, + "balance_loss_mlp": 1.03012371, + "epoch": 0.13912102446942826, + "flos": 22638051216000.0, + "grad_norm": 1.9374777729540185, + "language_loss": 0.85212374, + "learning_rate": 3.876497232671372e-06, + "loss": 0.87457043, + "num_input_tokens_seen": 24637485, + "step": 1157, + "time_per_iteration": 2.683865547180176 + }, + { + "auxiliary_loss_clip": 0.01131339, + "auxiliary_loss_mlp": 0.01042677, + "balance_loss_clip": 0.90928257, + "balance_loss_mlp": 1.03168643, + "epoch": 0.13924126736006734, + "flos": 29643324975360.0, + "grad_norm": 2.2720712401735343, + "language_loss": 0.83487755, + "learning_rate": 3.876227595950647e-06, + "loss": 0.85661769, + "num_input_tokens_seen": 24656915, + "step": 1158, + "time_per_iteration": 2.802950143814087 + }, + { + "auxiliary_loss_clip": 0.01227422, + "auxiliary_loss_mlp": 0.01040791, + "balance_loss_clip": 1.06517541, + "balance_loss_mlp": 1.02988362, + "epoch": 0.13936151025070642, + "flos": 27417653527680.0, + "grad_norm": 1.6506642585402977, + "language_loss": 0.78700554, + "learning_rate": 3.875957674607686e-06, + "loss": 0.80968767, + "num_input_tokens_seen": 24679190, + "step": 1159, + "time_per_iteration": 2.652301549911499 + }, + { + "auxiliary_loss_clip": 0.01197704, + "auxiliary_loss_mlp": 0.00514147, + "balance_loss_clip": 1.02471018, + "balance_loss_mlp": 0.0, + "epoch": 0.1394817531413455, + "flos": 16399326625920.0, + "grad_norm": 2.0325868768418474, + "language_loss": 0.87804663, + "learning_rate": 3.8756874686834386e-06, + "loss": 0.89516509, + "num_input_tokens_seen": 24697405, + "step": 1160, + "time_per_iteration": 2.6105756759643555 + }, + { + "auxiliary_loss_clip": 0.0120601, + "auxiliary_loss_mlp": 0.00514096, + "balance_loss_clip": 1.02568769, + "balance_loss_mlp": 0.0, + "epoch": 0.13960199603198462, + "flos": 30922319525760.0, + "grad_norm": 1.5664923564140012, + "language_loss": 0.80562711, + "learning_rate": 3.875416978218893e-06, + "loss": 0.82282817, + "num_input_tokens_seen": 24720600, + "step": 1161, + "time_per_iteration": 2.723109483718872 + }, + { + "auxiliary_loss_clip": 0.01161963, + "auxiliary_loss_mlp": 0.01045416, + "balance_loss_clip": 0.9456892, + "balance_loss_mlp": 1.03417516, + "epoch": 0.1397222389226237, + "flos": 18113773754880.0, + "grad_norm": 2.416313862778257, + "language_loss": 0.82828367, + "learning_rate": 3.8751462032550835e-06, + "loss": 0.85035741, + "num_input_tokens_seen": 24737605, + "step": 1162, + "time_per_iteration": 2.639537811279297 + }, + { + "auxiliary_loss_clip": 0.01179135, + "auxiliary_loss_mlp": 0.01039178, + "balance_loss_clip": 0.99242032, + "balance_loss_mlp": 1.02906346, + "epoch": 0.13984248181326278, + "flos": 16872772815360.0, + "grad_norm": 2.190188125517785, + "language_loss": 0.82818943, + "learning_rate": 3.874875143833085e-06, + "loss": 0.85037255, + "num_input_tokens_seen": 24755845, + "step": 1163, + "time_per_iteration": 3.5106537342071533 + }, + { + "auxiliary_loss_clip": 0.01203316, + "auxiliary_loss_mlp": 0.0104167, + "balance_loss_clip": 1.02586079, + "balance_loss_mlp": 1.02978504, + "epoch": 0.1399627247039019, + "flos": 54121401267840.0, + "grad_norm": 1.8320204201421344, + "language_loss": 0.68657899, + "learning_rate": 3.874603799994019e-06, + "loss": 0.70902884, + "num_input_tokens_seen": 24779380, + "step": 1164, + "time_per_iteration": 3.766359567642212 + }, + { + "auxiliary_loss_clip": 0.0114668, + "auxiliary_loss_mlp": 0.01039048, + "balance_loss_clip": 0.94727576, + "balance_loss_mlp": 1.02840304, + "epoch": 0.14008296759454097, + "flos": 11765521618560.0, + "grad_norm": 2.0144959141887258, + "language_loss": 0.86799854, + "learning_rate": 3.874332171779046e-06, + "loss": 0.88985586, + "num_input_tokens_seen": 24794260, + "step": 1165, + "time_per_iteration": 3.6208558082580566 + }, + { + "auxiliary_loss_clip": 0.01149899, + "auxiliary_loss_mlp": 0.01044338, + "balance_loss_clip": 0.94710904, + "balance_loss_mlp": 1.0334723, + "epoch": 0.14020321048518006, + "flos": 22017514832640.0, + "grad_norm": 1.880693782433562, + "language_loss": 0.75733471, + "learning_rate": 3.874060259229373e-06, + "loss": 0.77927715, + "num_input_tokens_seen": 24815835, + "step": 1166, + "time_per_iteration": 2.6978909969329834 + }, + { + "auxiliary_loss_clip": 0.01209895, + "auxiliary_loss_mlp": 0.01048596, + "balance_loss_clip": 1.03126812, + "balance_loss_mlp": 1.03652573, + "epoch": 0.14032345337581917, + "flos": 23404313076480.0, + "grad_norm": 1.9016752935804226, + "language_loss": 0.93423414, + "learning_rate": 3.873788062386249e-06, + "loss": 0.95681906, + "num_input_tokens_seen": 24834095, + "step": 1167, + "time_per_iteration": 3.5673117637634277 + }, + { + "auxiliary_loss_clip": 0.0116441, + "auxiliary_loss_mlp": 0.01042226, + "balance_loss_clip": 0.95536053, + "balance_loss_mlp": 1.03139567, + "epoch": 0.14044369626645825, + "flos": 29645767100160.0, + "grad_norm": 1.805656557666928, + "language_loss": 0.82183093, + "learning_rate": 3.873515581290965e-06, + "loss": 0.84389728, + "num_input_tokens_seen": 24858900, + "step": 1168, + "time_per_iteration": 2.780224323272705 + }, + { + "auxiliary_loss_clip": 0.0116098, + "auxiliary_loss_mlp": 0.01039347, + "balance_loss_clip": 0.95443738, + "balance_loss_mlp": 1.0286901, + "epoch": 0.14056393915709733, + "flos": 18332972501760.0, + "grad_norm": 2.0298886804960703, + "language_loss": 0.75596708, + "learning_rate": 3.8732428159848575e-06, + "loss": 0.77797031, + "num_input_tokens_seen": 24877875, + "step": 1169, + "time_per_iteration": 2.7051761150360107 + }, + { + "auxiliary_loss_clip": 0.01207741, + "auxiliary_loss_mlp": 0.01051007, + "balance_loss_clip": 1.03259218, + "balance_loss_mlp": 1.03809667, + "epoch": 0.14068418204773642, + "flos": 26687517770880.0, + "grad_norm": 1.8531476442018264, + "language_loss": 0.7811116, + "learning_rate": 3.872969766509304e-06, + "loss": 0.80369914, + "num_input_tokens_seen": 24898430, + "step": 1170, + "time_per_iteration": 2.696836471557617 + }, + { + "auxiliary_loss_clip": 0.01055354, + "auxiliary_loss_mlp": 0.01004405, + "balance_loss_clip": 0.92148966, + "balance_loss_mlp": 1.00143647, + "epoch": 0.14080442493837553, + "flos": 65259314501760.0, + "grad_norm": 0.7625300223045427, + "language_loss": 0.55639195, + "learning_rate": 3.872696432905726e-06, + "loss": 0.57698953, + "num_input_tokens_seen": 24959250, + "step": 1171, + "time_per_iteration": 3.2450647354125977 + }, + { + "auxiliary_loss_clip": 0.01207558, + "auxiliary_loss_mlp": 0.01038245, + "balance_loss_clip": 1.02607965, + "balance_loss_mlp": 1.02637196, + "epoch": 0.1409246678290146, + "flos": 25776715582080.0, + "grad_norm": 2.3467967252019566, + "language_loss": 0.71687353, + "learning_rate": 3.872422815215589e-06, + "loss": 0.7393316, + "num_input_tokens_seen": 24978330, + "step": 1172, + "time_per_iteration": 2.6846938133239746 + }, + { + "auxiliary_loss_clip": 0.01196926, + "auxiliary_loss_mlp": 0.01039115, + "balance_loss_clip": 1.02371836, + "balance_loss_mlp": 1.02756929, + "epoch": 0.1410449107196537, + "flos": 21868521217920.0, + "grad_norm": 2.1569729920715313, + "language_loss": 0.7455821, + "learning_rate": 3.8721489134803994e-06, + "loss": 0.76794261, + "num_input_tokens_seen": 24997120, + "step": 1173, + "time_per_iteration": 2.614215612411499 + }, + { + "auxiliary_loss_clip": 0.0120092, + "auxiliary_loss_mlp": 0.0104043, + "balance_loss_clip": 1.02646792, + "balance_loss_mlp": 1.02860451, + "epoch": 0.1411651536102928, + "flos": 16684133564160.0, + "grad_norm": 2.6360310136118006, + "language_loss": 0.72235245, + "learning_rate": 3.871874727741707e-06, + "loss": 0.74476594, + "num_input_tokens_seen": 25014350, + "step": 1174, + "time_per_iteration": 2.6162970066070557 + }, + { + "auxiliary_loss_clip": 0.01203139, + "auxiliary_loss_mlp": 0.01039285, + "balance_loss_clip": 1.03068423, + "balance_loss_mlp": 1.0285387, + "epoch": 0.1412853965009319, + "flos": 20992264934400.0, + "grad_norm": 1.7386716284974044, + "language_loss": 0.96477604, + "learning_rate": 3.871600258041108e-06, + "loss": 0.98720038, + "num_input_tokens_seen": 25033875, + "step": 1175, + "time_per_iteration": 2.6714742183685303 + }, + { + "auxiliary_loss_clip": 0.01175023, + "auxiliary_loss_mlp": 0.01038293, + "balance_loss_clip": 0.98729384, + "balance_loss_mlp": 1.02636051, + "epoch": 0.14140563939157097, + "flos": 20335279224960.0, + "grad_norm": 2.131651006639446, + "language_loss": 0.85896474, + "learning_rate": 3.871325504420238e-06, + "loss": 0.88109791, + "num_input_tokens_seen": 25052865, + "step": 1176, + "time_per_iteration": 2.679091691970825 + }, + { + "auxiliary_loss_clip": 0.01230348, + "auxiliary_loss_mlp": 0.01038708, + "balance_loss_clip": 1.06755924, + "balance_loss_mlp": 1.02791905, + "epoch": 0.14152588228221005, + "flos": 21068826773760.0, + "grad_norm": 1.7963530594205166, + "language_loss": 0.81480873, + "learning_rate": 3.871050466920776e-06, + "loss": 0.83749926, + "num_input_tokens_seen": 25072770, + "step": 1177, + "time_per_iteration": 2.6188414096832275 + }, + { + "auxiliary_loss_clip": 0.01150057, + "auxiliary_loss_mlp": 0.01038618, + "balance_loss_clip": 0.95000672, + "balance_loss_mlp": 1.02749026, + "epoch": 0.14164612517284916, + "flos": 18223157646720.0, + "grad_norm": 1.773923199073378, + "language_loss": 0.80030411, + "learning_rate": 3.870775145584447e-06, + "loss": 0.82219088, + "num_input_tokens_seen": 25090550, + "step": 1178, + "time_per_iteration": 2.697763204574585 + }, + { + "auxiliary_loss_clip": 0.01189473, + "auxiliary_loss_mlp": 0.01039919, + "balance_loss_clip": 0.98999828, + "balance_loss_mlp": 1.02746153, + "epoch": 0.14176636806348825, + "flos": 22744454279040.0, + "grad_norm": 3.042313822349166, + "language_loss": 0.65072644, + "learning_rate": 3.8704995404530145e-06, + "loss": 0.67302036, + "num_input_tokens_seen": 25106175, + "step": 1179, + "time_per_iteration": 2.6635680198669434 + }, + { + "auxiliary_loss_clip": 0.0122928, + "auxiliary_loss_mlp": 0.01043259, + "balance_loss_clip": 1.06944537, + "balance_loss_mlp": 1.0326848, + "epoch": 0.14188661095412733, + "flos": 22091095843200.0, + "grad_norm": 2.3990602760856574, + "language_loss": 0.84928203, + "learning_rate": 3.87022365156829e-06, + "loss": 0.87200749, + "num_input_tokens_seen": 25126890, + "step": 1180, + "time_per_iteration": 2.616661310195923 + }, + { + "auxiliary_loss_clip": 0.01083083, + "auxiliary_loss_mlp": 0.01045451, + "balance_loss_clip": 0.8342225, + "balance_loss_mlp": 1.0338937, + "epoch": 0.14200685384476644, + "flos": 24352390604160.0, + "grad_norm": 2.0312838972667624, + "language_loss": 0.81011266, + "learning_rate": 3.869947478972123e-06, + "loss": 0.83139801, + "num_input_tokens_seen": 25147915, + "step": 1181, + "time_per_iteration": 2.818544387817383 + }, + { + "auxiliary_loss_clip": 0.01196464, + "auxiliary_loss_mlp": 0.01041363, + "balance_loss_clip": 1.02520609, + "balance_loss_mlp": 1.03003275, + "epoch": 0.14212709673540552, + "flos": 24022048199040.0, + "grad_norm": 2.581404521827558, + "language_loss": 0.82851827, + "learning_rate": 3.869671022706412e-06, + "loss": 0.85089648, + "num_input_tokens_seen": 25166645, + "step": 1182, + "time_per_iteration": 2.6670002937316895 + }, + { + "auxiliary_loss_clip": 0.01117353, + "auxiliary_loss_mlp": 0.01037718, + "balance_loss_clip": 0.90775585, + "balance_loss_mlp": 1.02634549, + "epoch": 0.1422473396260446, + "flos": 26431797870720.0, + "grad_norm": 1.8606069225957185, + "language_loss": 0.65000558, + "learning_rate": 3.869394282813092e-06, + "loss": 0.67155623, + "num_input_tokens_seen": 25185845, + "step": 1183, + "time_per_iteration": 2.800981283187866 + }, + { + "auxiliary_loss_clip": 0.01165441, + "auxiliary_loss_mlp": 0.01043893, + "balance_loss_clip": 0.95065266, + "balance_loss_mlp": 1.03147721, + "epoch": 0.1423675825166837, + "flos": 17055306754560.0, + "grad_norm": 2.348150978138355, + "language_loss": 0.89521199, + "learning_rate": 3.869117259334147e-06, + "loss": 0.91730535, + "num_input_tokens_seen": 25203770, + "step": 1184, + "time_per_iteration": 2.6965479850769043 + }, + { + "auxiliary_loss_clip": 0.01200871, + "auxiliary_loss_mlp": 0.01041803, + "balance_loss_clip": 1.02655244, + "balance_loss_mlp": 1.03060317, + "epoch": 0.1424878254073228, + "flos": 17929480049280.0, + "grad_norm": 1.753169634786789, + "language_loss": 0.82169056, + "learning_rate": 3.868839952311599e-06, + "loss": 0.84411734, + "num_input_tokens_seen": 25221725, + "step": 1185, + "time_per_iteration": 2.6416170597076416 + }, + { + "auxiliary_loss_clip": 0.01177248, + "auxiliary_loss_mlp": 0.01039892, + "balance_loss_clip": 0.99041724, + "balance_loss_mlp": 1.02837634, + "epoch": 0.14260806829796188, + "flos": 20303606407680.0, + "grad_norm": 2.2275447998461257, + "language_loss": 0.8012042, + "learning_rate": 3.868562361787516e-06, + "loss": 0.82337564, + "num_input_tokens_seen": 25240855, + "step": 1186, + "time_per_iteration": 2.6408607959747314 + }, + { + "auxiliary_loss_clip": 0.01078184, + "auxiliary_loss_mlp": 0.01035958, + "balance_loss_clip": 0.83022034, + "balance_loss_mlp": 1.0245018, + "epoch": 0.14272831118860096, + "flos": 23185724860800.0, + "grad_norm": 2.4783229732949477, + "language_loss": 0.68859649, + "learning_rate": 3.868284487804009e-06, + "loss": 0.7097379, + "num_input_tokens_seen": 25260085, + "step": 1187, + "time_per_iteration": 2.796025276184082 + }, + { + "auxiliary_loss_clip": 0.01187469, + "auxiliary_loss_mlp": 0.01045148, + "balance_loss_clip": 0.98863363, + "balance_loss_mlp": 1.03379965, + "epoch": 0.14284855407924008, + "flos": 27232210586880.0, + "grad_norm": 1.5801736740227261, + "language_loss": 0.77934325, + "learning_rate": 3.86800633040323e-06, + "loss": 0.80166948, + "num_input_tokens_seen": 25280675, + "step": 1188, + "time_per_iteration": 2.6908280849456787 + }, + { + "auxiliary_loss_clip": 0.01180267, + "auxiliary_loss_mlp": 0.00514724, + "balance_loss_clip": 0.99198282, + "balance_loss_mlp": 0.0, + "epoch": 0.14296879696987916, + "flos": 28184202696960.0, + "grad_norm": 2.374467225746604, + "language_loss": 0.77811837, + "learning_rate": 3.867727889627376e-06, + "loss": 0.7950682, + "num_input_tokens_seen": 25300290, + "step": 1189, + "time_per_iteration": 4.473186016082764 + }, + { + "auxiliary_loss_clip": 0.01149479, + "auxiliary_loss_mlp": 0.01040388, + "balance_loss_clip": 0.95091009, + "balance_loss_mlp": 1.02853286, + "epoch": 0.14308903986051824, + "flos": 19390290266880.0, + "grad_norm": 2.5459937071692793, + "language_loss": 0.78416348, + "learning_rate": 3.867449165518687e-06, + "loss": 0.80606216, + "num_input_tokens_seen": 25316760, + "step": 1190, + "time_per_iteration": 3.589167356491089 + }, + { + "auxiliary_loss_clip": 0.01228511, + "auxiliary_loss_mlp": 0.0051371, + "balance_loss_clip": 1.06305027, + "balance_loss_mlp": 0.0, + "epoch": 0.14320928275115732, + "flos": 17457506317440.0, + "grad_norm": 1.8797841412223506, + "language_loss": 0.70868701, + "learning_rate": 3.867170158119444e-06, + "loss": 0.72610927, + "num_input_tokens_seen": 25335760, + "step": 1191, + "time_per_iteration": 2.5888822078704834 + }, + { + "auxiliary_loss_clip": 0.01234215, + "auxiliary_loss_mlp": 0.0104844, + "balance_loss_clip": 1.0698185, + "balance_loss_mlp": 1.03585792, + "epoch": 0.14332952564179643, + "flos": 21466070259840.0, + "grad_norm": 2.4558439318450715, + "language_loss": 0.75072539, + "learning_rate": 3.866890867471972e-06, + "loss": 0.773552, + "num_input_tokens_seen": 25354230, + "step": 1192, + "time_per_iteration": 2.58437180519104 + }, + { + "auxiliary_loss_clip": 0.01177592, + "auxiliary_loss_mlp": 0.01041657, + "balance_loss_clip": 0.98517466, + "balance_loss_mlp": 1.02936649, + "epoch": 0.14344976853243552, + "flos": 16396992241920.0, + "grad_norm": 2.139884135353099, + "language_loss": 0.89748585, + "learning_rate": 3.86661129361864e-06, + "loss": 0.91967833, + "num_input_tokens_seen": 25368720, + "step": 1193, + "time_per_iteration": 3.6567435264587402 + }, + { + "auxiliary_loss_clip": 0.01178938, + "auxiliary_loss_mlp": 0.01039785, + "balance_loss_clip": 0.98941952, + "balance_loss_mlp": 1.02756619, + "epoch": 0.1435700114230746, + "flos": 18916736336640.0, + "grad_norm": 1.7802734736090216, + "language_loss": 0.86186415, + "learning_rate": 3.866331436601859e-06, + "loss": 0.88405138, + "num_input_tokens_seen": 25386715, + "step": 1194, + "time_per_iteration": 2.728576898574829 + }, + { + "auxiliary_loss_clip": 0.01236566, + "auxiliary_loss_mlp": 0.01039862, + "balance_loss_clip": 1.07202673, + "balance_loss_mlp": 1.0280838, + "epoch": 0.1436902543137137, + "flos": 19755394058880.0, + "grad_norm": 2.7626181581359495, + "language_loss": 0.73647505, + "learning_rate": 3.866051296464083e-06, + "loss": 0.75923932, + "num_input_tokens_seen": 25405550, + "step": 1195, + "time_per_iteration": 2.617166757583618 + }, + { + "auxiliary_loss_clip": 0.01230845, + "auxiliary_loss_mlp": 0.0051439, + "balance_loss_clip": 1.06513298, + "balance_loss_mlp": 0.0, + "epoch": 0.1438104972043528, + "flos": 14684807669760.0, + "grad_norm": 2.702502469027008, + "language_loss": 0.85458946, + "learning_rate": 3.86577087324781e-06, + "loss": 0.87204182, + "num_input_tokens_seen": 25422040, + "step": 1196, + "time_per_iteration": 2.6293344497680664 + }, + { + "auxiliary_loss_clip": 0.01206813, + "auxiliary_loss_mlp": 0.010377, + "balance_loss_clip": 1.03091776, + "balance_loss_mlp": 1.02701831, + "epoch": 0.14393074009499188, + "flos": 17092330698240.0, + "grad_norm": 1.8158640122280896, + "language_loss": 0.77629638, + "learning_rate": 3.865490166995578e-06, + "loss": 0.79874152, + "num_input_tokens_seen": 25440270, + "step": 1197, + "time_per_iteration": 2.6945719718933105 + }, + { + "auxiliary_loss_clip": 0.01210098, + "auxiliary_loss_mlp": 0.01041409, + "balance_loss_clip": 1.03115547, + "balance_loss_mlp": 1.02873683, + "epoch": 0.144050982985631, + "flos": 30476200608000.0, + "grad_norm": 2.092560808919344, + "language_loss": 0.84412843, + "learning_rate": 3.86520917774997e-06, + "loss": 0.86664343, + "num_input_tokens_seen": 25459705, + "step": 1198, + "time_per_iteration": 2.7716147899627686 + }, + { + "auxiliary_loss_clip": 0.01202812, + "auxiliary_loss_mlp": 0.01049889, + "balance_loss_clip": 1.02837396, + "balance_loss_mlp": 1.0387789, + "epoch": 0.14417122587627007, + "flos": 17858484817920.0, + "grad_norm": 2.0121252029314416, + "language_loss": 0.75009549, + "learning_rate": 3.864927905553614e-06, + "loss": 0.77262247, + "num_input_tokens_seen": 25477615, + "step": 1199, + "time_per_iteration": 2.651686668395996 + }, + { + "auxiliary_loss_clip": 0.01154086, + "auxiliary_loss_mlp": 0.01042745, + "balance_loss_clip": 0.95083654, + "balance_loss_mlp": 1.03117013, + "epoch": 0.14429146876690915, + "flos": 21613914639360.0, + "grad_norm": 1.6599311673951616, + "language_loss": 0.88578719, + "learning_rate": 3.8646463504491765e-06, + "loss": 0.90775549, + "num_input_tokens_seen": 25497750, + "step": 1200, + "time_per_iteration": 2.6947975158691406 + }, + { + "auxiliary_loss_clip": 0.0120842, + "auxiliary_loss_mlp": 0.01037535, + "balance_loss_clip": 1.03142881, + "balance_loss_mlp": 1.02665687, + "epoch": 0.14441171165754824, + "flos": 23258120722560.0, + "grad_norm": 1.7343590129054078, + "language_loss": 0.83283341, + "learning_rate": 3.8643645124793705e-06, + "loss": 0.85529298, + "num_input_tokens_seen": 25516650, + "step": 1201, + "time_per_iteration": 2.672663688659668 + }, + { + "auxiliary_loss_clip": 0.01201565, + "auxiliary_loss_mlp": 0.0104066, + "balance_loss_clip": 1.02585292, + "balance_loss_mlp": 1.02937126, + "epoch": 0.14453195454818735, + "flos": 42854213963520.0, + "grad_norm": 1.85041007505218, + "language_loss": 0.75028223, + "learning_rate": 3.8640823916869515e-06, + "loss": 0.77270448, + "num_input_tokens_seen": 25540960, + "step": 1202, + "time_per_iteration": 2.873410701751709 + }, + { + "auxiliary_loss_clip": 0.01230427, + "auxiliary_loss_mlp": 0.01039063, + "balance_loss_clip": 1.06773019, + "balance_loss_mlp": 1.02772582, + "epoch": 0.14465219743882643, + "flos": 27235873774080.0, + "grad_norm": 1.5253103133430501, + "language_loss": 0.78341478, + "learning_rate": 3.863799988114714e-06, + "loss": 0.80610967, + "num_input_tokens_seen": 25562990, + "step": 1203, + "time_per_iteration": 2.636791229248047 + }, + { + "auxiliary_loss_clip": 0.01230784, + "auxiliary_loss_mlp": 0.01040558, + "balance_loss_clip": 1.06680465, + "balance_loss_mlp": 1.02916813, + "epoch": 0.1447724403294655, + "flos": 16690705752960.0, + "grad_norm": 2.1446333474178676, + "language_loss": 0.70411932, + "learning_rate": 3.863517301805502e-06, + "loss": 0.72683281, + "num_input_tokens_seen": 25581380, + "step": 1204, + "time_per_iteration": 2.604736328125 + }, + { + "auxiliary_loss_clip": 0.01158481, + "auxiliary_loss_mlp": 0.01040556, + "balance_loss_clip": 0.95447332, + "balance_loss_mlp": 1.0285337, + "epoch": 0.14489268322010462, + "flos": 20073741321600.0, + "grad_norm": 2.245108574436241, + "language_loss": 0.96846241, + "learning_rate": 3.863234332802196e-06, + "loss": 0.99045277, + "num_input_tokens_seen": 25593585, + "step": 1205, + "time_per_iteration": 2.647627353668213 + }, + { + "auxiliary_loss_clip": 0.01179053, + "auxiliary_loss_mlp": 0.01042552, + "balance_loss_clip": 0.98902142, + "balance_loss_mlp": 1.03048813, + "epoch": 0.1450129261107437, + "flos": 27125627955840.0, + "grad_norm": 2.0533089724158495, + "language_loss": 0.73796058, + "learning_rate": 3.862951081147723e-06, + "loss": 0.76017654, + "num_input_tokens_seen": 25613750, + "step": 1206, + "time_per_iteration": 2.752628803253174 + }, + { + "auxiliary_loss_clip": 0.01212261, + "auxiliary_loss_mlp": 0.01039582, + "balance_loss_clip": 1.03338408, + "balance_loss_mlp": 1.02921069, + "epoch": 0.1451331690013828, + "flos": 25702344472320.0, + "grad_norm": 2.153735506812231, + "language_loss": 0.78337002, + "learning_rate": 3.862667546885053e-06, + "loss": 0.80588835, + "num_input_tokens_seen": 25632300, + "step": 1207, + "time_per_iteration": 2.6659295558929443 + }, + { + "auxiliary_loss_clip": 0.01185405, + "auxiliary_loss_mlp": 0.01040266, + "balance_loss_clip": 0.98947817, + "balance_loss_mlp": 1.02834463, + "epoch": 0.14525341189202187, + "flos": 25737393168000.0, + "grad_norm": 2.0771776070894887, + "language_loss": 0.73516715, + "learning_rate": 3.8623837300571965e-06, + "loss": 0.75742382, + "num_input_tokens_seen": 25651285, + "step": 1208, + "time_per_iteration": 2.6975934505462646 + }, + { + "auxiliary_loss_clip": 0.01231743, + "auxiliary_loss_mlp": 0.0104086, + "balance_loss_clip": 1.06861448, + "balance_loss_mlp": 1.02868843, + "epoch": 0.14537365478266098, + "flos": 23073898844160.0, + "grad_norm": 1.7759533513112495, + "language_loss": 0.83879662, + "learning_rate": 3.8620996307072085e-06, + "loss": 0.86152267, + "num_input_tokens_seen": 25671990, + "step": 1209, + "time_per_iteration": 2.7067830562591553 + }, + { + "auxiliary_loss_clip": 0.01158517, + "auxiliary_loss_mlp": 0.01037953, + "balance_loss_clip": 0.9473815, + "balance_loss_mlp": 1.02653313, + "epoch": 0.14549389767330007, + "flos": 20595021448320.0, + "grad_norm": 1.6482177911023956, + "language_loss": 0.64349425, + "learning_rate": 3.861815248878188e-06, + "loss": 0.66545898, + "num_input_tokens_seen": 25689475, + "step": 1210, + "time_per_iteration": 2.7325546741485596 + }, + { + "auxiliary_loss_clip": 0.01176038, + "auxiliary_loss_mlp": 0.01037479, + "balance_loss_clip": 0.98897696, + "balance_loss_mlp": 1.02581453, + "epoch": 0.14561414056393915, + "flos": 15121804533120.0, + "grad_norm": 2.390114389354236, + "language_loss": 0.79604745, + "learning_rate": 3.861530584613274e-06, + "loss": 0.81818259, + "num_input_tokens_seen": 25707475, + "step": 1211, + "time_per_iteration": 2.6740920543670654 + }, + { + "auxiliary_loss_clip": 0.01210471, + "auxiliary_loss_mlp": 0.00514313, + "balance_loss_clip": 1.03106856, + "balance_loss_mlp": 0.0, + "epoch": 0.14573438345457826, + "flos": 19427493778560.0, + "grad_norm": 2.346260760714724, + "language_loss": 0.82396936, + "learning_rate": 3.86124563795565e-06, + "loss": 0.84121716, + "num_input_tokens_seen": 25726290, + "step": 1212, + "time_per_iteration": 2.6663196086883545 + }, + { + "auxiliary_loss_clip": 0.01234085, + "auxiliary_loss_mlp": 0.010352, + "balance_loss_clip": 1.07125711, + "balance_loss_mlp": 1.02453625, + "epoch": 0.14585462634521734, + "flos": 24828422572800.0, + "grad_norm": 1.6018176125647012, + "language_loss": 0.69931787, + "learning_rate": 3.860960408948543e-06, + "loss": 0.72201073, + "num_input_tokens_seen": 25748040, + "step": 1213, + "time_per_iteration": 2.650102138519287 + }, + { + "auxiliary_loss_clip": 0.01198402, + "auxiliary_loss_mlp": 0.01040137, + "balance_loss_clip": 1.02801919, + "balance_loss_mlp": 1.02937841, + "epoch": 0.14597486923585642, + "flos": 15448627405440.0, + "grad_norm": 2.1643427912599376, + "language_loss": 0.90025413, + "learning_rate": 3.860674897635222e-06, + "loss": 0.92263949, + "num_input_tokens_seen": 25764525, + "step": 1214, + "time_per_iteration": 2.6612305641174316 + }, + { + "auxiliary_loss_clip": 0.01207458, + "auxiliary_loss_mlp": 0.01041576, + "balance_loss_clip": 1.03140926, + "balance_loss_mlp": 1.03001881, + "epoch": 0.1460951121264955, + "flos": 16655154266880.0, + "grad_norm": 1.9492914791147007, + "language_loss": 0.83337784, + "learning_rate": 3.860389104058998e-06, + "loss": 0.8558681, + "num_input_tokens_seen": 25782755, + "step": 1215, + "time_per_iteration": 3.5458357334136963 + }, + { + "auxiliary_loss_clip": 0.0118125, + "auxiliary_loss_mlp": 0.01036939, + "balance_loss_clip": 0.98893476, + "balance_loss_mlp": 1.02572107, + "epoch": 0.14621535501713462, + "flos": 24863291700480.0, + "grad_norm": 2.2266994590855, + "language_loss": 0.7254855, + "learning_rate": 3.860103028263227e-06, + "loss": 0.74766743, + "num_input_tokens_seen": 25805860, + "step": 1216, + "time_per_iteration": 3.6543619632720947 + }, + { + "auxiliary_loss_clip": 0.01123662, + "auxiliary_loss_mlp": 0.01041031, + "balance_loss_clip": 0.90590185, + "balance_loss_mlp": 1.03008795, + "epoch": 0.1463355979077737, + "flos": 25228000442880.0, + "grad_norm": 2.22209116749153, + "language_loss": 0.70367908, + "learning_rate": 3.859816670291304e-06, + "loss": 0.72532606, + "num_input_tokens_seen": 25824955, + "step": 1217, + "time_per_iteration": 3.6572070121765137 + }, + { + "auxiliary_loss_clip": 0.01101084, + "auxiliary_loss_mlp": 0.01033055, + "balance_loss_clip": 0.87089825, + "balance_loss_mlp": 1.02168274, + "epoch": 0.14645584079841278, + "flos": 22054143726720.0, + "grad_norm": 2.0436807906213112, + "language_loss": 0.90078151, + "learning_rate": 3.859530030186672e-06, + "loss": 0.92212296, + "num_input_tokens_seen": 25841965, + "step": 1218, + "time_per_iteration": 2.7756385803222656 + }, + { + "auxiliary_loss_clip": 0.01187087, + "auxiliary_loss_mlp": 0.01042207, + "balance_loss_clip": 0.99199915, + "balance_loss_mlp": 1.03081656, + "epoch": 0.1465760836890519, + "flos": 23623870959360.0, + "grad_norm": 2.458134781691185, + "language_loss": 0.82936907, + "learning_rate": 3.859243107992813e-06, + "loss": 0.85166204, + "num_input_tokens_seen": 25860770, + "step": 1219, + "time_per_iteration": 3.6344714164733887 + }, + { + "auxiliary_loss_clip": 0.01156911, + "auxiliary_loss_mlp": 0.0104032, + "balance_loss_clip": 0.94473249, + "balance_loss_mlp": 1.02836347, + "epoch": 0.14669632657969098, + "flos": 37407893356800.0, + "grad_norm": 3.185076551371309, + "language_loss": 0.78115958, + "learning_rate": 3.858955903753252e-06, + "loss": 0.80313194, + "num_input_tokens_seen": 25879410, + "step": 1220, + "time_per_iteration": 2.823469877243042 + }, + { + "auxiliary_loss_clip": 0.01207232, + "auxiliary_loss_mlp": 0.01036722, + "balance_loss_clip": 1.02865291, + "balance_loss_mlp": 1.02599323, + "epoch": 0.14681656947033006, + "flos": 28365910623360.0, + "grad_norm": 1.5033593919954764, + "language_loss": 0.83790916, + "learning_rate": 3.858668417511559e-06, + "loss": 0.8603487, + "num_input_tokens_seen": 25902160, + "step": 1221, + "time_per_iteration": 2.735930919647217 + }, + { + "auxiliary_loss_clip": 0.0118338, + "auxiliary_loss_mlp": 0.01036011, + "balance_loss_clip": 0.9907003, + "balance_loss_mlp": 1.02522898, + "epoch": 0.14693681236096917, + "flos": 18479488078080.0, + "grad_norm": 2.076296927775526, + "language_loss": 0.76128763, + "learning_rate": 3.8583806493113445e-06, + "loss": 0.78348148, + "num_input_tokens_seen": 25920505, + "step": 1222, + "time_per_iteration": 2.641664505004883 + }, + { + "auxiliary_loss_clip": 0.01200226, + "auxiliary_loss_mlp": 0.01038061, + "balance_loss_clip": 1.02669358, + "balance_loss_mlp": 1.02709365, + "epoch": 0.14705705525160825, + "flos": 20777806782720.0, + "grad_norm": 3.9475233878038343, + "language_loss": 0.82405823, + "learning_rate": 3.858092599196263e-06, + "loss": 0.84644109, + "num_input_tokens_seen": 25938460, + "step": 1223, + "time_per_iteration": 2.656569242477417 + }, + { + "auxiliary_loss_clip": 0.01208069, + "auxiliary_loss_mlp": 0.01035931, + "balance_loss_clip": 1.0315249, + "balance_loss_mlp": 1.02465975, + "epoch": 0.14717729814224734, + "flos": 29932944336000.0, + "grad_norm": 3.983931287329699, + "language_loss": 0.82574677, + "learning_rate": 3.857804267210012e-06, + "loss": 0.84818673, + "num_input_tokens_seen": 25957760, + "step": 1224, + "time_per_iteration": 2.7007062435150146 + }, + { + "auxiliary_loss_clip": 0.01143868, + "auxiliary_loss_mlp": 0.01041976, + "balance_loss_clip": 0.94579601, + "balance_loss_mlp": 1.03085399, + "epoch": 0.14729754103288642, + "flos": 20047491457920.0, + "grad_norm": 1.9968370818119563, + "language_loss": 0.88259816, + "learning_rate": 3.857515653396331e-06, + "loss": 0.90445662, + "num_input_tokens_seen": 25974970, + "step": 1225, + "time_per_iteration": 2.7376906871795654 + }, + { + "auxiliary_loss_clip": 0.01131037, + "auxiliary_loss_mlp": 0.01032795, + "balance_loss_clip": 0.90870929, + "balance_loss_mlp": 1.02182174, + "epoch": 0.14741778392352553, + "flos": 19281516906240.0, + "grad_norm": 2.2074307167246534, + "language_loss": 0.86733717, + "learning_rate": 3.857226757799002e-06, + "loss": 0.8889755, + "num_input_tokens_seen": 25992525, + "step": 1226, + "time_per_iteration": 2.726511001586914 + }, + { + "auxiliary_loss_clip": 0.01181692, + "auxiliary_loss_mlp": 0.01046472, + "balance_loss_clip": 0.99038577, + "balance_loss_mlp": 1.03480768, + "epoch": 0.1475380268141646, + "flos": 25411108999680.0, + "grad_norm": 2.4268231932052666, + "language_loss": 0.74391741, + "learning_rate": 3.85693758046185e-06, + "loss": 0.76619911, + "num_input_tokens_seen": 26010815, + "step": 1227, + "time_per_iteration": 2.7579329013824463 + }, + { + "auxiliary_loss_clip": 0.01231031, + "auxiliary_loss_mlp": 0.01044022, + "balance_loss_clip": 1.06992698, + "balance_loss_mlp": 1.03275132, + "epoch": 0.1476582697048037, + "flos": 20847652778880.0, + "grad_norm": 1.9962387447046732, + "language_loss": 0.82774377, + "learning_rate": 3.8566481214287435e-06, + "loss": 0.85049427, + "num_input_tokens_seen": 26028935, + "step": 1228, + "time_per_iteration": 2.6005632877349854 + }, + { + "auxiliary_loss_clip": 0.01146144, + "auxiliary_loss_mlp": 0.01042586, + "balance_loss_clip": 0.94472873, + "balance_loss_mlp": 1.03160727, + "epoch": 0.1477785125954428, + "flos": 14028109269120.0, + "grad_norm": 1.9241009102436566, + "language_loss": 0.9038465, + "learning_rate": 3.8563583807435935e-06, + "loss": 0.9257338, + "num_input_tokens_seen": 26045080, + "step": 1229, + "time_per_iteration": 2.693169593811035 + }, + { + "auxiliary_loss_clip": 0.01206626, + "auxiliary_loss_mlp": 0.00513909, + "balance_loss_clip": 1.02721334, + "balance_loss_mlp": 0.0, + "epoch": 0.1478987554860819, + "flos": 20516699842560.0, + "grad_norm": 2.185449052443415, + "language_loss": 0.7776742, + "learning_rate": 3.856068358450353e-06, + "loss": 0.79487956, + "num_input_tokens_seen": 26065030, + "step": 1230, + "time_per_iteration": 2.653836727142334 + }, + { + "auxiliary_loss_clip": 0.01177512, + "auxiliary_loss_mlp": 0.01044585, + "balance_loss_clip": 0.99315298, + "balance_loss_mlp": 1.03323603, + "epoch": 0.14801899837672097, + "flos": 17857012360320.0, + "grad_norm": 1.7058766132893162, + "language_loss": 0.85733938, + "learning_rate": 3.8557780545930186e-06, + "loss": 0.87956035, + "num_input_tokens_seen": 26083445, + "step": 1231, + "time_per_iteration": 2.761322021484375 + }, + { + "auxiliary_loss_clip": 0.01180212, + "auxiliary_loss_mlp": 0.010475, + "balance_loss_clip": 0.99261034, + "balance_loss_mlp": 1.03653312, + "epoch": 0.14813924126736006, + "flos": 20881408584960.0, + "grad_norm": 1.6899053486566749, + "language_loss": 0.79219246, + "learning_rate": 3.855487469215628e-06, + "loss": 0.81446964, + "num_input_tokens_seen": 26102375, + "step": 1232, + "time_per_iteration": 2.688865900039673 + }, + { + "auxiliary_loss_clip": 0.01158344, + "auxiliary_loss_mlp": 0.01039979, + "balance_loss_clip": 0.95311069, + "balance_loss_mlp": 1.02828455, + "epoch": 0.14825948415799917, + "flos": 37414070496000.0, + "grad_norm": 2.805922594077157, + "language_loss": 0.71872383, + "learning_rate": 3.855196602362264e-06, + "loss": 0.74070704, + "num_input_tokens_seen": 26125295, + "step": 1233, + "time_per_iteration": 2.858673334121704 + }, + { + "auxiliary_loss_clip": 0.01207182, + "auxiliary_loss_mlp": 0.0104006, + "balance_loss_clip": 1.0296905, + "balance_loss_mlp": 1.02905655, + "epoch": 0.14837972704863825, + "flos": 22014641744640.0, + "grad_norm": 1.8948406778583096, + "language_loss": 0.94185209, + "learning_rate": 3.854905454077051e-06, + "loss": 0.96432447, + "num_input_tokens_seen": 26142905, + "step": 1234, + "time_per_iteration": 2.6475720405578613 + }, + { + "auxiliary_loss_clip": 0.01078292, + "auxiliary_loss_mlp": 0.01044993, + "balance_loss_clip": 0.83086777, + "balance_loss_mlp": 1.03391206, + "epoch": 0.14849996993927733, + "flos": 20996323171200.0, + "grad_norm": 1.7978215266562523, + "language_loss": 0.88567114, + "learning_rate": 3.854614024404155e-06, + "loss": 0.90690398, + "num_input_tokens_seen": 26161215, + "step": 1235, + "time_per_iteration": 2.811899423599243 + }, + { + "auxiliary_loss_clip": 0.01166731, + "auxiliary_loss_mlp": 0.01034656, + "balance_loss_clip": 0.98472404, + "balance_loss_mlp": 1.02411199, + "epoch": 0.14862021282991644, + "flos": 20047994248320.0, + "grad_norm": 1.7963859197024006, + "language_loss": 0.89418924, + "learning_rate": 3.8543223133877865e-06, + "loss": 0.91620314, + "num_input_tokens_seen": 26179810, + "step": 1236, + "time_per_iteration": 2.690669536590576 + }, + { + "auxiliary_loss_clip": 0.01163957, + "auxiliary_loss_mlp": 0.01037621, + "balance_loss_clip": 0.98519301, + "balance_loss_mlp": 1.0256052, + "epoch": 0.14874045572055553, + "flos": 22712027276160.0, + "grad_norm": 1.727533238046914, + "language_loss": 0.88632166, + "learning_rate": 3.854030321072198e-06, + "loss": 0.90833747, + "num_input_tokens_seen": 26199715, + "step": 1237, + "time_per_iteration": 2.6726315021514893 + }, + { + "auxiliary_loss_clip": 0.01159996, + "auxiliary_loss_mlp": 0.01034157, + "balance_loss_clip": 0.95032632, + "balance_loss_mlp": 1.02330875, + "epoch": 0.1488606986111946, + "flos": 25411288567680.0, + "grad_norm": 2.076401642349452, + "language_loss": 0.73642725, + "learning_rate": 3.853738047501682e-06, + "loss": 0.75836873, + "num_input_tokens_seen": 26220275, + "step": 1238, + "time_per_iteration": 2.769733428955078 + }, + { + "auxiliary_loss_clip": 0.01208733, + "auxiliary_loss_mlp": 0.01039138, + "balance_loss_clip": 1.03227985, + "balance_loss_mlp": 1.02768826, + "epoch": 0.1489809415018337, + "flos": 17018749687680.0, + "grad_norm": 1.7012352568811187, + "language_loss": 0.77623415, + "learning_rate": 3.85344549272058e-06, + "loss": 0.79871285, + "num_input_tokens_seen": 26238255, + "step": 1239, + "time_per_iteration": 2.6421735286712646 + }, + { + "auxiliary_loss_clip": 0.01201675, + "auxiliary_loss_mlp": 0.01041438, + "balance_loss_clip": 1.02804244, + "balance_loss_mlp": 1.03090024, + "epoch": 0.1491011843924728, + "flos": 33659394860160.0, + "grad_norm": 3.434398171438928, + "language_loss": 0.82435453, + "learning_rate": 3.853152656773269e-06, + "loss": 0.84678566, + "num_input_tokens_seen": 26259690, + "step": 1240, + "time_per_iteration": 2.8709115982055664 + }, + { + "auxiliary_loss_clip": 0.01177383, + "auxiliary_loss_mlp": 0.01032762, + "balance_loss_clip": 0.98894393, + "balance_loss_mlp": 1.02200365, + "epoch": 0.14922142728311188, + "flos": 21179000764800.0, + "grad_norm": 1.696385503708979, + "language_loss": 0.84699404, + "learning_rate": 3.852859539704174e-06, + "loss": 0.8690955, + "num_input_tokens_seen": 26278990, + "step": 1241, + "time_per_iteration": 3.688133478164673 + }, + { + "auxiliary_loss_clip": 0.01133605, + "auxiliary_loss_mlp": 0.01038794, + "balance_loss_clip": 0.91173184, + "balance_loss_mlp": 1.0282557, + "epoch": 0.14934167017375097, + "flos": 29860548474240.0, + "grad_norm": 2.71860205898606, + "language_loss": 0.76151538, + "learning_rate": 3.85256614155776e-06, + "loss": 0.78323936, + "num_input_tokens_seen": 26299120, + "step": 1242, + "time_per_iteration": 4.699735641479492 + }, + { + "auxiliary_loss_clip": 0.01203454, + "auxiliary_loss_mlp": 0.01037424, + "balance_loss_clip": 1.02548027, + "balance_loss_mlp": 1.02688622, + "epoch": 0.14946191306439008, + "flos": 17019216564480.0, + "grad_norm": 1.8568910407941976, + "language_loss": 0.7440058, + "learning_rate": 3.852272462378535e-06, + "loss": 0.76641452, + "num_input_tokens_seen": 26316995, + "step": 1243, + "time_per_iteration": 2.6298789978027344 + }, + { + "auxiliary_loss_clip": 0.0118118, + "auxiliary_loss_mlp": 0.01032745, + "balance_loss_clip": 0.98913467, + "balance_loss_mlp": 1.0226661, + "epoch": 0.14958215595502916, + "flos": 15669047214720.0, + "grad_norm": 1.9999858124545493, + "language_loss": 0.77889764, + "learning_rate": 3.85197850221105e-06, + "loss": 0.80103695, + "num_input_tokens_seen": 26333295, + "step": 1244, + "time_per_iteration": 2.675947427749634 + }, + { + "auxiliary_loss_clip": 0.0119891, + "auxiliary_loss_mlp": 0.01041818, + "balance_loss_clip": 1.02774763, + "balance_loss_mlp": 1.03196526, + "epoch": 0.14970239884566824, + "flos": 33108560818560.0, + "grad_norm": 1.9852763932980317, + "language_loss": 0.75908959, + "learning_rate": 3.851684261099899e-06, + "loss": 0.78149688, + "num_input_tokens_seen": 26355035, + "step": 1245, + "time_per_iteration": 3.677248239517212 + }, + { + "auxiliary_loss_clip": 0.01175113, + "auxiliary_loss_mlp": 0.01036163, + "balance_loss_clip": 0.98785198, + "balance_loss_mlp": 1.02582192, + "epoch": 0.14982264173630733, + "flos": 17821245392640.0, + "grad_norm": 2.051737140284334, + "language_loss": 0.86407852, + "learning_rate": 3.851389739089718e-06, + "loss": 0.88619131, + "num_input_tokens_seen": 26371655, + "step": 1246, + "time_per_iteration": 2.6692259311676025 + }, + { + "auxiliary_loss_clip": 0.0120683, + "auxiliary_loss_mlp": 0.01041327, + "balance_loss_clip": 1.03049386, + "balance_loss_mlp": 1.0300914, + "epoch": 0.14994288462694644, + "flos": 32409559175040.0, + "grad_norm": 2.1409124156664037, + "language_loss": 0.80531847, + "learning_rate": 3.851094936225186e-06, + "loss": 0.82780004, + "num_input_tokens_seen": 26392540, + "step": 1247, + "time_per_iteration": 2.713402509689331 + }, + { + "auxiliary_loss_clip": 0.01174794, + "auxiliary_loss_mlp": 0.01041642, + "balance_loss_clip": 0.98790812, + "balance_loss_mlp": 1.03122282, + "epoch": 0.15006312751758552, + "flos": 31794661226880.0, + "grad_norm": 1.3869536020935889, + "language_loss": 0.76356363, + "learning_rate": 3.850799852551024e-06, + "loss": 0.78572798, + "num_input_tokens_seen": 26414960, + "step": 1248, + "time_per_iteration": 2.711864948272705 + }, + { + "auxiliary_loss_clip": 0.01194497, + "auxiliary_loss_mlp": 0.01040476, + "balance_loss_clip": 1.02459407, + "balance_loss_mlp": 1.02952635, + "epoch": 0.1501833704082246, + "flos": 16618022582400.0, + "grad_norm": 2.095369955004491, + "language_loss": 0.86042297, + "learning_rate": 3.850504488111995e-06, + "loss": 0.88277268, + "num_input_tokens_seen": 26431635, + "step": 1249, + "time_per_iteration": 2.6515090465545654 + }, + { + "auxiliary_loss_clip": 0.01173593, + "auxiliary_loss_mlp": 0.01046203, + "balance_loss_clip": 0.98699534, + "balance_loss_mlp": 1.03609443, + "epoch": 0.15030361329886371, + "flos": 23471178243840.0, + "grad_norm": 1.8092666958922088, + "language_loss": 0.82340682, + "learning_rate": 3.850208842952907e-06, + "loss": 0.84560478, + "num_input_tokens_seen": 26450440, + "step": 1250, + "time_per_iteration": 2.710631847381592 + }, + { + "auxiliary_loss_clip": 0.01136842, + "auxiliary_loss_mlp": 0.01032203, + "balance_loss_clip": 0.90869462, + "balance_loss_mlp": 1.02080023, + "epoch": 0.1504238561895028, + "flos": 25629409906560.0, + "grad_norm": 1.8695950119610638, + "language_loss": 0.79202569, + "learning_rate": 3.849912917118608e-06, + "loss": 0.81371617, + "num_input_tokens_seen": 26471480, + "step": 1251, + "time_per_iteration": 2.7737374305725098 + }, + { + "auxiliary_loss_clip": 0.01132012, + "auxiliary_loss_mlp": 0.0101062, + "balance_loss_clip": 1.01853585, + "balance_loss_mlp": 1.00730634, + "epoch": 0.15054409908014188, + "flos": 52095146129280.0, + "grad_norm": 0.8862487782331829, + "language_loss": 0.59275675, + "learning_rate": 3.849616710653992e-06, + "loss": 0.61418307, + "num_input_tokens_seen": 26532950, + "step": 1252, + "time_per_iteration": 3.2143771648406982 + }, + { + "auxiliary_loss_clip": 0.0120285, + "auxiliary_loss_mlp": 0.01038957, + "balance_loss_clip": 1.02659535, + "balance_loss_mlp": 1.02704811, + "epoch": 0.150664341970781, + "flos": 18880251096960.0, + "grad_norm": 1.726914695894071, + "language_loss": 0.74592263, + "learning_rate": 3.84932022360399e-06, + "loss": 0.76834077, + "num_input_tokens_seen": 26551615, + "step": 1253, + "time_per_iteration": 2.672978401184082 + }, + { + "auxiliary_loss_clip": 0.01178844, + "auxiliary_loss_mlp": 0.0103542, + "balance_loss_clip": 0.99281275, + "balance_loss_mlp": 1.02528727, + "epoch": 0.15078458486142007, + "flos": 22163240309760.0, + "grad_norm": 2.5707211877830027, + "language_loss": 0.84419405, + "learning_rate": 3.849023456013581e-06, + "loss": 0.8663367, + "num_input_tokens_seen": 26569175, + "step": 1254, + "time_per_iteration": 2.6502573490142822 + }, + { + "auxiliary_loss_clip": 0.01208935, + "auxiliary_loss_mlp": 0.01054196, + "balance_loss_clip": 1.0271697, + "balance_loss_mlp": 1.04262638, + "epoch": 0.15090482775205916, + "flos": 26651894457600.0, + "grad_norm": 1.9421543815095843, + "language_loss": 0.62097037, + "learning_rate": 3.848726407927784e-06, + "loss": 0.64360172, + "num_input_tokens_seen": 26589560, + "step": 1255, + "time_per_iteration": 2.7359960079193115 + }, + { + "auxiliary_loss_clip": 0.01184073, + "auxiliary_loss_mlp": 0.01039091, + "balance_loss_clip": 0.99330068, + "balance_loss_mlp": 1.02708066, + "epoch": 0.15102507064269824, + "flos": 21798998444160.0, + "grad_norm": 2.3904687307713384, + "language_loss": 0.86501443, + "learning_rate": 3.84842907939166e-06, + "loss": 0.88724607, + "num_input_tokens_seen": 26608785, + "step": 1256, + "time_per_iteration": 2.654517412185669 + }, + { + "auxiliary_loss_clip": 0.01145723, + "auxiliary_loss_mlp": 0.01040673, + "balance_loss_clip": 0.94706142, + "balance_loss_mlp": 1.03033721, + "epoch": 0.15114531353333735, + "flos": 22820908377600.0, + "grad_norm": 4.251089604643897, + "language_loss": 0.7087636, + "learning_rate": 3.8481314704503146e-06, + "loss": 0.7306276, + "num_input_tokens_seen": 26628615, + "step": 1257, + "time_per_iteration": 2.7459864616394043 + }, + { + "auxiliary_loss_clip": 0.01206452, + "auxiliary_loss_mlp": 0.01038799, + "balance_loss_clip": 1.03230119, + "balance_loss_mlp": 1.02756405, + "epoch": 0.15126555642397643, + "flos": 19682674974720.0, + "grad_norm": 2.059683011861748, + "language_loss": 0.88174796, + "learning_rate": 3.847833581148895e-06, + "loss": 0.90420055, + "num_input_tokens_seen": 26647525, + "step": 1258, + "time_per_iteration": 2.612457036972046 + }, + { + "auxiliary_loss_clip": 0.01224732, + "auxiliary_loss_mlp": 0.01036855, + "balance_loss_clip": 1.0626924, + "balance_loss_mlp": 1.02596509, + "epoch": 0.15138579931461552, + "flos": 28726022424960.0, + "grad_norm": 2.3214709335099344, + "language_loss": 0.81400847, + "learning_rate": 3.84753541153259e-06, + "loss": 0.83662432, + "num_input_tokens_seen": 26667095, + "step": 1259, + "time_per_iteration": 2.653362512588501 + }, + { + "auxiliary_loss_clip": 0.01205671, + "auxiliary_loss_mlp": 0.01046032, + "balance_loss_clip": 1.03090715, + "balance_loss_mlp": 1.03521347, + "epoch": 0.15150604220525463, + "flos": 22127006465280.0, + "grad_norm": 1.5788138791847417, + "language_loss": 0.83359921, + "learning_rate": 3.847236961646633e-06, + "loss": 0.85611629, + "num_input_tokens_seen": 26686075, + "step": 1260, + "time_per_iteration": 2.637063503265381 + }, + { + "auxiliary_loss_clip": 0.01173644, + "auxiliary_loss_mlp": 0.01038474, + "balance_loss_clip": 0.98662806, + "balance_loss_mlp": 1.02782869, + "epoch": 0.1516262850958937, + "flos": 12968708515200.0, + "grad_norm": 2.744388637555293, + "language_loss": 0.77976441, + "learning_rate": 3.846938231536296e-06, + "loss": 0.80188555, + "num_input_tokens_seen": 26701695, + "step": 1261, + "time_per_iteration": 2.6745293140411377 + }, + { + "auxiliary_loss_clip": 0.01209659, + "auxiliary_loss_mlp": 0.01040983, + "balance_loss_clip": 1.03238833, + "balance_loss_mlp": 1.03030181, + "epoch": 0.1517465279865328, + "flos": 21797130936960.0, + "grad_norm": 1.891676242238641, + "language_loss": 0.80733347, + "learning_rate": 3.8466392212468995e-06, + "loss": 0.82983989, + "num_input_tokens_seen": 26721885, + "step": 1262, + "time_per_iteration": 2.681474208831787 + }, + { + "auxiliary_loss_clip": 0.01088419, + "auxiliary_loss_mlp": 0.01010241, + "balance_loss_clip": 0.93877178, + "balance_loss_mlp": 1.00673664, + "epoch": 0.15186677087717187, + "flos": 58174569901440.0, + "grad_norm": 0.8435922816146183, + "language_loss": 0.62001801, + "learning_rate": 3.8463399308238e-06, + "loss": 0.64100456, + "num_input_tokens_seen": 26780990, + "step": 1263, + "time_per_iteration": 3.1924901008605957 + }, + { + "auxiliary_loss_clip": 0.01203903, + "auxiliary_loss_mlp": 0.01040413, + "balance_loss_clip": 1.02908397, + "balance_loss_mlp": 1.03039384, + "epoch": 0.15198701376781099, + "flos": 32669696448000.0, + "grad_norm": 1.7591740307338712, + "language_loss": 0.6405623, + "learning_rate": 3.846040360312402e-06, + "loss": 0.66300541, + "num_input_tokens_seen": 26804250, + "step": 1264, + "time_per_iteration": 2.8506815433502197 + }, + { + "auxiliary_loss_clip": 0.0123032, + "auxiliary_loss_mlp": 0.01038779, + "balance_loss_clip": 1.06763804, + "balance_loss_mlp": 1.02826452, + "epoch": 0.15210725665845007, + "flos": 28402575431040.0, + "grad_norm": 2.2083315495494333, + "language_loss": 0.80803365, + "learning_rate": 3.8457405097581485e-06, + "loss": 0.83072466, + "num_input_tokens_seen": 26823240, + "step": 1265, + "time_per_iteration": 2.7406437397003174 + }, + { + "auxiliary_loss_clip": 0.01131577, + "auxiliary_loss_mlp": 0.01044549, + "balance_loss_clip": 0.90750355, + "balance_loss_mlp": 1.03306353, + "epoch": 0.15222749954908915, + "flos": 19938179393280.0, + "grad_norm": 2.015878057905289, + "language_loss": 0.7816118, + "learning_rate": 3.8454403792065275e-06, + "loss": 0.80337304, + "num_input_tokens_seen": 26842060, + "step": 1266, + "time_per_iteration": 2.7521257400512695 + }, + { + "auxiliary_loss_clip": 0.01145795, + "auxiliary_loss_mlp": 0.01045999, + "balance_loss_clip": 0.94941545, + "balance_loss_mlp": 1.03553271, + "epoch": 0.15234774243972826, + "flos": 21324223451520.0, + "grad_norm": 6.327474393795543, + "language_loss": 0.85354781, + "learning_rate": 3.845139968703068e-06, + "loss": 0.87546581, + "num_input_tokens_seen": 26859580, + "step": 1267, + "time_per_iteration": 3.6964776515960693 + }, + { + "auxiliary_loss_clip": 0.01129042, + "auxiliary_loss_mlp": 0.01041892, + "balance_loss_clip": 0.90905458, + "balance_loss_mlp": 1.03147292, + "epoch": 0.15246798533036734, + "flos": 25957812977280.0, + "grad_norm": 2.038883422968284, + "language_loss": 0.83357477, + "learning_rate": 3.844839278293342e-06, + "loss": 0.85528409, + "num_input_tokens_seen": 26880430, + "step": 1268, + "time_per_iteration": 3.7166271209716797 + }, + { + "auxiliary_loss_clip": 0.01229788, + "auxiliary_loss_mlp": 0.01042822, + "balance_loss_clip": 1.06886852, + "balance_loss_mlp": 1.03117514, + "epoch": 0.15258822822100643, + "flos": 25811907932160.0, + "grad_norm": 2.087877890412862, + "language_loss": 0.76599467, + "learning_rate": 3.8445383080229654e-06, + "loss": 0.78872085, + "num_input_tokens_seen": 26896445, + "step": 1269, + "time_per_iteration": 3.5616042613983154 + }, + { + "auxiliary_loss_clip": 0.01172427, + "auxiliary_loss_mlp": 0.01039189, + "balance_loss_clip": 0.98668945, + "balance_loss_mlp": 1.02755368, + "epoch": 0.1527084711116455, + "flos": 25265455349760.0, + "grad_norm": 2.1398855991118118, + "language_loss": 0.73745453, + "learning_rate": 3.844237057937593e-06, + "loss": 0.75957066, + "num_input_tokens_seen": 26915450, + "step": 1270, + "time_per_iteration": 2.743562698364258 + }, + { + "auxiliary_loss_clip": 0.01207351, + "auxiliary_loss_mlp": 0.01039506, + "balance_loss_clip": 1.0265913, + "balance_loss_mlp": 1.02884924, + "epoch": 0.15282871400228462, + "flos": 29240227572480.0, + "grad_norm": 2.488654222799911, + "language_loss": 0.7784549, + "learning_rate": 3.843935528082926e-06, + "loss": 0.80092347, + "num_input_tokens_seen": 26936475, + "step": 1271, + "time_per_iteration": 2.7305591106414795 + }, + { + "auxiliary_loss_clip": 0.01204898, + "auxiliary_loss_mlp": 0.01038173, + "balance_loss_clip": 1.02813447, + "balance_loss_mlp": 1.02690732, + "epoch": 0.1529489568929237, + "flos": 20882952869760.0, + "grad_norm": 1.993644431869412, + "language_loss": 0.85092807, + "learning_rate": 3.843633718504704e-06, + "loss": 0.87335873, + "num_input_tokens_seen": 26954920, + "step": 1272, + "time_per_iteration": 3.585035800933838 + }, + { + "auxiliary_loss_clip": 0.01156991, + "auxiliary_loss_mlp": 0.01044768, + "balance_loss_clip": 0.95244896, + "balance_loss_mlp": 1.03372955, + "epoch": 0.1530691997835628, + "flos": 20083833043200.0, + "grad_norm": 2.692005335869375, + "language_loss": 0.89854145, + "learning_rate": 3.843331629248715e-06, + "loss": 0.92055905, + "num_input_tokens_seen": 26972520, + "step": 1273, + "time_per_iteration": 2.6903421878814697 + }, + { + "auxiliary_loss_clip": 0.01226557, + "auxiliary_loss_mlp": 0.01048515, + "balance_loss_clip": 1.06731761, + "balance_loss_mlp": 1.03784025, + "epoch": 0.1531894426742019, + "flos": 28759814144640.0, + "grad_norm": 2.0612973520219318, + "language_loss": 0.76821762, + "learning_rate": 3.843029260360782e-06, + "loss": 0.7909683, + "num_input_tokens_seen": 26990890, + "step": 1274, + "time_per_iteration": 2.6665737628936768 + }, + { + "auxiliary_loss_clip": 0.01206383, + "auxiliary_loss_mlp": 0.01035954, + "balance_loss_clip": 1.03136766, + "balance_loss_mlp": 1.02544606, + "epoch": 0.15330968556484098, + "flos": 22236282616320.0, + "grad_norm": 1.9730381246891624, + "language_loss": 0.78464609, + "learning_rate": 3.8427266118867755e-06, + "loss": 0.80706948, + "num_input_tokens_seen": 27010640, + "step": 1275, + "time_per_iteration": 2.655461311340332 + }, + { + "auxiliary_loss_clip": 0.01175866, + "auxiliary_loss_mlp": 0.01046088, + "balance_loss_clip": 0.98903632, + "balance_loss_mlp": 1.0340004, + "epoch": 0.15342992845548006, + "flos": 27527504296320.0, + "grad_norm": 1.858019056074265, + "language_loss": 0.82769662, + "learning_rate": 3.842423683872608e-06, + "loss": 0.84991616, + "num_input_tokens_seen": 27031215, + "step": 1276, + "time_per_iteration": 2.7288851737976074 + }, + { + "auxiliary_loss_clip": 0.01207026, + "auxiliary_loss_mlp": 0.01037826, + "balance_loss_clip": 1.02914429, + "balance_loss_mlp": 1.02673972, + "epoch": 0.15355017134611917, + "flos": 19609596754560.0, + "grad_norm": 2.8667459728027347, + "language_loss": 0.78216779, + "learning_rate": 3.842120476364232e-06, + "loss": 0.80461627, + "num_input_tokens_seen": 27049665, + "step": 1277, + "time_per_iteration": 2.6183226108551025 + }, + { + "auxiliary_loss_clip": 0.01208067, + "auxiliary_loss_mlp": 0.01039995, + "balance_loss_clip": 1.02659273, + "balance_loss_mlp": 1.02892065, + "epoch": 0.15367041423675826, + "flos": 18478590238080.0, + "grad_norm": 2.754052127371112, + "language_loss": 0.83727825, + "learning_rate": 3.841816989407644e-06, + "loss": 0.85975885, + "num_input_tokens_seen": 27065155, + "step": 1278, + "time_per_iteration": 2.5784566402435303 + }, + { + "auxiliary_loss_clip": 0.0114864, + "auxiliary_loss_mlp": 0.01033022, + "balance_loss_clip": 0.94733679, + "balance_loss_mlp": 1.02331805, + "epoch": 0.15379065712739734, + "flos": 41427662342400.0, + "grad_norm": 1.9144107036772273, + "language_loss": 0.76494104, + "learning_rate": 3.841513223048884e-06, + "loss": 0.78675765, + "num_input_tokens_seen": 27085840, + "step": 1279, + "time_per_iteration": 2.896408796310425 + }, + { + "auxiliary_loss_clip": 0.01151791, + "auxiliary_loss_mlp": 0.01042886, + "balance_loss_clip": 0.94805396, + "balance_loss_mlp": 1.03181744, + "epoch": 0.15391090001803642, + "flos": 22054215553920.0, + "grad_norm": 2.3111941756082772, + "language_loss": 0.78667706, + "learning_rate": 3.841209177334031e-06, + "loss": 0.80862379, + "num_input_tokens_seen": 27104200, + "step": 1280, + "time_per_iteration": 2.710294246673584 + }, + { + "auxiliary_loss_clip": 0.01198031, + "auxiliary_loss_mlp": 0.01038052, + "balance_loss_clip": 1.02585268, + "balance_loss_mlp": 1.02754402, + "epoch": 0.15403114290867553, + "flos": 15450351258240.0, + "grad_norm": 1.7503204231186495, + "language_loss": 0.74832702, + "learning_rate": 3.84090485230921e-06, + "loss": 0.77068782, + "num_input_tokens_seen": 27122440, + "step": 1281, + "time_per_iteration": 2.6409058570861816 + }, + { + "auxiliary_loss_clip": 0.01225511, + "auxiliary_loss_mlp": 0.01039297, + "balance_loss_clip": 1.06502628, + "balance_loss_mlp": 1.02878857, + "epoch": 0.15415138579931462, + "flos": 17929156826880.0, + "grad_norm": 2.7009916075031053, + "language_loss": 0.76603335, + "learning_rate": 3.840600248020588e-06, + "loss": 0.78868145, + "num_input_tokens_seen": 27139380, + "step": 1282, + "time_per_iteration": 2.557797908782959 + }, + { + "auxiliary_loss_clip": 0.01181709, + "auxiliary_loss_mlp": 0.01040396, + "balance_loss_clip": 0.98453963, + "balance_loss_mlp": 1.0288502, + "epoch": 0.1542716286899537, + "flos": 11429325296640.0, + "grad_norm": 2.630047431316088, + "language_loss": 0.79781246, + "learning_rate": 3.840295364514371e-06, + "loss": 0.82003355, + "num_input_tokens_seen": 27156760, + "step": 1283, + "time_per_iteration": 2.7082877159118652 + }, + { + "auxiliary_loss_clip": 0.01181104, + "auxiliary_loss_mlp": 0.01037399, + "balance_loss_clip": 0.98937088, + "balance_loss_mlp": 1.02676535, + "epoch": 0.1543918715805928, + "flos": 17420338719360.0, + "grad_norm": 2.296678818863687, + "language_loss": 0.78376448, + "learning_rate": 3.83999020183681e-06, + "loss": 0.80594957, + "num_input_tokens_seen": 27175455, + "step": 1284, + "time_per_iteration": 2.658407211303711 + }, + { + "auxiliary_loss_clip": 0.01099716, + "auxiliary_loss_mlp": 0.01032478, + "balance_loss_clip": 0.87034047, + "balance_loss_mlp": 1.02273905, + "epoch": 0.1545121144712319, + "flos": 17786376264960.0, + "grad_norm": 2.0943998684628893, + "language_loss": 0.78696835, + "learning_rate": 3.839684760034199e-06, + "loss": 0.8082903, + "num_input_tokens_seen": 27193660, + "step": 1285, + "time_per_iteration": 2.841817617416382 + }, + { + "auxiliary_loss_clip": 0.01144738, + "auxiliary_loss_mlp": 0.01051928, + "balance_loss_clip": 0.94839323, + "balance_loss_mlp": 1.04099679, + "epoch": 0.15463235736187098, + "flos": 28220185146240.0, + "grad_norm": 2.321643373420258, + "language_loss": 0.65502679, + "learning_rate": 3.8393790391528716e-06, + "loss": 0.67699343, + "num_input_tokens_seen": 27214355, + "step": 1286, + "time_per_iteration": 2.804445505142212 + }, + { + "auxiliary_loss_clip": 0.01172788, + "auxiliary_loss_mlp": 0.01035365, + "balance_loss_clip": 0.98269832, + "balance_loss_mlp": 1.025298, + "epoch": 0.15475260025251006, + "flos": 22856890826880.0, + "grad_norm": 1.9375155301205598, + "language_loss": 0.88893414, + "learning_rate": 3.8390730392392075e-06, + "loss": 0.91101569, + "num_input_tokens_seen": 27234335, + "step": 1287, + "time_per_iteration": 2.7906486988067627 + }, + { + "auxiliary_loss_clip": 0.01228056, + "auxiliary_loss_mlp": 0.01042936, + "balance_loss_clip": 1.06699109, + "balance_loss_mlp": 1.03304172, + "epoch": 0.15487284314314917, + "flos": 17602872658560.0, + "grad_norm": 2.1854093052730406, + "language_loss": 0.791058, + "learning_rate": 3.838766760339626e-06, + "loss": 0.81376791, + "num_input_tokens_seen": 27252860, + "step": 1288, + "time_per_iteration": 2.6371567249298096 + }, + { + "auxiliary_loss_clip": 0.01132195, + "auxiliary_loss_mlp": 0.01043534, + "balance_loss_clip": 0.94163847, + "balance_loss_mlp": 1.0327158, + "epoch": 0.15499308603378825, + "flos": 20082037363200.0, + "grad_norm": 3.3277218054828808, + "language_loss": 0.79117876, + "learning_rate": 3.838460202500587e-06, + "loss": 0.81293607, + "num_input_tokens_seen": 27268650, + "step": 1289, + "time_per_iteration": 2.6691083908081055 + }, + { + "auxiliary_loss_clip": 0.01148026, + "auxiliary_loss_mlp": 0.01045392, + "balance_loss_clip": 0.95313936, + "balance_loss_mlp": 1.03426945, + "epoch": 0.15511332892442733, + "flos": 15918051271680.0, + "grad_norm": 2.1339399108797603, + "language_loss": 0.73899591, + "learning_rate": 3.838153365768599e-06, + "loss": 0.76093006, + "num_input_tokens_seen": 27285160, + "step": 1290, + "time_per_iteration": 2.674771785736084 + }, + { + "auxiliary_loss_clip": 0.0115257, + "auxiliary_loss_mlp": 0.0103978, + "balance_loss_clip": 0.95286256, + "balance_loss_mlp": 1.02857435, + "epoch": 0.15523357181506645, + "flos": 41282475569280.0, + "grad_norm": 2.435293357971652, + "language_loss": 0.75034285, + "learning_rate": 3.837846250190206e-06, + "loss": 0.77226627, + "num_input_tokens_seen": 27308025, + "step": 1291, + "time_per_iteration": 2.852656841278076 + }, + { + "auxiliary_loss_clip": 0.01124066, + "auxiliary_loss_mlp": 0.00512547, + "balance_loss_clip": 0.91045797, + "balance_loss_mlp": 0.0, + "epoch": 0.15535381470570553, + "flos": 18478769806080.0, + "grad_norm": 2.4630005684987726, + "language_loss": 0.7706027, + "learning_rate": 3.837538855811998e-06, + "loss": 0.78696883, + "num_input_tokens_seen": 27326200, + "step": 1292, + "time_per_iteration": 3.6936886310577393 + }, + { + "auxiliary_loss_clip": 0.0118332, + "auxiliary_loss_mlp": 0.01033651, + "balance_loss_clip": 0.9913075, + "balance_loss_mlp": 1.02288616, + "epoch": 0.1554740575963446, + "flos": 13918150759680.0, + "grad_norm": 2.0160219012991685, + "language_loss": 0.70760423, + "learning_rate": 3.837231182680606e-06, + "loss": 0.72977388, + "num_input_tokens_seen": 27344165, + "step": 1293, + "time_per_iteration": 2.665144205093384 + }, + { + "auxiliary_loss_clip": 0.01204791, + "auxiliary_loss_mlp": 0.01041676, + "balance_loss_clip": 1.02799106, + "balance_loss_mlp": 1.03132892, + "epoch": 0.1555943004869837, + "flos": 20847078161280.0, + "grad_norm": 2.610899077761557, + "language_loss": 0.7615152, + "learning_rate": 3.836923230842706e-06, + "loss": 0.78397989, + "num_input_tokens_seen": 27363280, + "step": 1294, + "time_per_iteration": 3.630713939666748 + }, + { + "auxiliary_loss_clip": 0.01129991, + "auxiliary_loss_mlp": 0.01040858, + "balance_loss_clip": 0.90764987, + "balance_loss_mlp": 1.03020108, + "epoch": 0.1557145433776228, + "flos": 22085888371200.0, + "grad_norm": 2.047115011095414, + "language_loss": 0.80959374, + "learning_rate": 3.836615000345011e-06, + "loss": 0.83130229, + "num_input_tokens_seen": 27381460, + "step": 1295, + "time_per_iteration": 2.813404083251953 + }, + { + "auxiliary_loss_clip": 0.0122306, + "auxiliary_loss_mlp": 0.01034589, + "balance_loss_clip": 1.06537938, + "balance_loss_mlp": 1.02431321, + "epoch": 0.1558347862682619, + "flos": 19791987039360.0, + "grad_norm": 2.152915020636472, + "language_loss": 0.78319037, + "learning_rate": 3.836306491234282e-06, + "loss": 0.80576682, + "num_input_tokens_seen": 27399310, + "step": 1296, + "time_per_iteration": 2.5989913940429688 + }, + { + "auxiliary_loss_clip": 0.01174143, + "auxiliary_loss_mlp": 0.0103609, + "balance_loss_clip": 0.9919427, + "balance_loss_mlp": 1.02635646, + "epoch": 0.15595502915890097, + "flos": 17237086508160.0, + "grad_norm": 2.2091283281395784, + "language_loss": 0.75473857, + "learning_rate": 3.835997703557317e-06, + "loss": 0.77684093, + "num_input_tokens_seen": 27416050, + "step": 1297, + "time_per_iteration": 3.615852117538452 + }, + { + "auxiliary_loss_clip": 0.01125501, + "auxiliary_loss_mlp": 0.01042152, + "balance_loss_clip": 0.90526974, + "balance_loss_mlp": 1.03268123, + "epoch": 0.15607527204954008, + "flos": 19719519350400.0, + "grad_norm": 2.450957379979458, + "language_loss": 0.80205178, + "learning_rate": 3.83568863736096e-06, + "loss": 0.82372832, + "num_input_tokens_seen": 27434920, + "step": 1298, + "time_per_iteration": 2.854189157485962 + }, + { + "auxiliary_loss_clip": 0.01149386, + "auxiliary_loss_mlp": 0.01035431, + "balance_loss_clip": 0.94500238, + "balance_loss_mlp": 1.02454734, + "epoch": 0.15619551494017916, + "flos": 18515650095360.0, + "grad_norm": 2.355396266953076, + "language_loss": 0.89416927, + "learning_rate": 3.8353792926920975e-06, + "loss": 0.91601741, + "num_input_tokens_seen": 27453570, + "step": 1299, + "time_per_iteration": 2.8130581378936768 + }, + { + "auxiliary_loss_clip": 0.01208367, + "auxiliary_loss_mlp": 0.0103776, + "balance_loss_clip": 1.02850795, + "balance_loss_mlp": 1.02710843, + "epoch": 0.15631575783081825, + "flos": 19902125116800.0, + "grad_norm": 2.0958227431003533, + "language_loss": 0.81775731, + "learning_rate": 3.835069669597655e-06, + "loss": 0.8402186, + "num_input_tokens_seen": 27471960, + "step": 1300, + "time_per_iteration": 2.7015578746795654 + }, + { + "auxiliary_loss_clip": 0.01209581, + "auxiliary_loss_mlp": 0.00513058, + "balance_loss_clip": 1.02839661, + "balance_loss_mlp": 0.0, + "epoch": 0.15643600072145733, + "flos": 20777663128320.0, + "grad_norm": 2.350293079625217, + "language_loss": 0.79972994, + "learning_rate": 3.834759768124603e-06, + "loss": 0.81695634, + "num_input_tokens_seen": 27490835, + "step": 1301, + "time_per_iteration": 2.7050087451934814 + }, + { + "auxiliary_loss_clip": 0.01156593, + "auxiliary_loss_mlp": 0.01048373, + "balance_loss_clip": 0.95144689, + "balance_loss_mlp": 1.03753078, + "epoch": 0.15655624361209644, + "flos": 18546389159040.0, + "grad_norm": 2.949996525656196, + "language_loss": 0.75862575, + "learning_rate": 3.834449588319953e-06, + "loss": 0.78067541, + "num_input_tokens_seen": 27508870, + "step": 1302, + "time_per_iteration": 2.702000617980957 + }, + { + "auxiliary_loss_clip": 0.0119909, + "auxiliary_loss_mlp": 0.01038837, + "balance_loss_clip": 1.02895427, + "balance_loss_mlp": 1.02872205, + "epoch": 0.15667648650273552, + "flos": 25229544727680.0, + "grad_norm": 1.7997913168430522, + "language_loss": 0.85153842, + "learning_rate": 3.834139130230758e-06, + "loss": 0.8739177, + "num_input_tokens_seen": 27528175, + "step": 1303, + "time_per_iteration": 2.660249710083008 + }, + { + "auxiliary_loss_clip": 0.01176616, + "auxiliary_loss_mlp": 0.01038486, + "balance_loss_clip": 0.98382962, + "balance_loss_mlp": 1.02802515, + "epoch": 0.1567967293933746, + "flos": 24827093769600.0, + "grad_norm": 1.6621542194033632, + "language_loss": 0.811602, + "learning_rate": 3.833828393904117e-06, + "loss": 0.83375305, + "num_input_tokens_seen": 27548455, + "step": 1304, + "time_per_iteration": 2.790940523147583 + }, + { + "auxiliary_loss_clip": 0.01122641, + "auxiliary_loss_mlp": 0.01035816, + "balance_loss_clip": 0.90796071, + "balance_loss_mlp": 1.02600503, + "epoch": 0.15691697228401372, + "flos": 19164555244800.0, + "grad_norm": 2.314843000319915, + "language_loss": 0.77591586, + "learning_rate": 3.833517379387165e-06, + "loss": 0.79750049, + "num_input_tokens_seen": 27564910, + "step": 1305, + "time_per_iteration": 2.6954708099365234 + }, + { + "auxiliary_loss_clip": 0.01203144, + "auxiliary_loss_mlp": 0.01037483, + "balance_loss_clip": 1.02679253, + "balance_loss_mlp": 1.02733207, + "epoch": 0.1570372151746528, + "flos": 24790931752320.0, + "grad_norm": 2.095225086001452, + "language_loss": 0.88719428, + "learning_rate": 3.833206086727085e-06, + "loss": 0.9096005, + "num_input_tokens_seen": 27584260, + "step": 1306, + "time_per_iteration": 2.70086407661438 + }, + { + "auxiliary_loss_clip": 0.01154815, + "auxiliary_loss_mlp": 0.01035657, + "balance_loss_clip": 0.94530475, + "balance_loss_mlp": 1.02499998, + "epoch": 0.15715745806529188, + "flos": 24863650836480.0, + "grad_norm": 1.9850673489234034, + "language_loss": 0.70351422, + "learning_rate": 3.8328945159710994e-06, + "loss": 0.72541893, + "num_input_tokens_seen": 27604440, + "step": 1307, + "time_per_iteration": 2.729243755340576 + }, + { + "auxiliary_loss_clip": 0.01211314, + "auxiliary_loss_mlp": 0.00512575, + "balance_loss_clip": 1.03226328, + "balance_loss_mlp": 0.0, + "epoch": 0.157277700955931, + "flos": 21872148491520.0, + "grad_norm": 1.981596274832926, + "language_loss": 0.8877846, + "learning_rate": 3.832582667166473e-06, + "loss": 0.90502346, + "num_input_tokens_seen": 27624250, + "step": 1308, + "time_per_iteration": 2.6740036010742188 + }, + { + "auxiliary_loss_clip": 0.01173467, + "auxiliary_loss_mlp": 0.01044578, + "balance_loss_clip": 0.98520404, + "balance_loss_mlp": 1.0340755, + "epoch": 0.15739794384657008, + "flos": 24533344344960.0, + "grad_norm": 1.953313314122312, + "language_loss": 0.81558537, + "learning_rate": 3.8322705403605125e-06, + "loss": 0.83776581, + "num_input_tokens_seen": 27644595, + "step": 1309, + "time_per_iteration": 2.806485414505005 + }, + { + "auxiliary_loss_clip": 0.01171157, + "auxiliary_loss_mlp": 0.01029985, + "balance_loss_clip": 0.986058, + "balance_loss_mlp": 1.02119982, + "epoch": 0.15751818673720916, + "flos": 17745329998080.0, + "grad_norm": 1.9495604394017432, + "language_loss": 0.81379223, + "learning_rate": 3.831958135600568e-06, + "loss": 0.83580363, + "num_input_tokens_seen": 27662145, + "step": 1310, + "time_per_iteration": 2.672649383544922 + }, + { + "auxiliary_loss_clip": 0.01205865, + "auxiliary_loss_mlp": 0.01043021, + "balance_loss_clip": 1.03004146, + "balance_loss_mlp": 1.03313279, + "epoch": 0.15763842962784824, + "flos": 17858520731520.0, + "grad_norm": 2.1048367424181076, + "language_loss": 0.79600441, + "learning_rate": 3.831645452934032e-06, + "loss": 0.81849325, + "num_input_tokens_seen": 27680575, + "step": 1311, + "time_per_iteration": 2.6511192321777344 + }, + { + "auxiliary_loss_clip": 0.01229762, + "auxiliary_loss_mlp": 0.01035993, + "balance_loss_clip": 1.07054198, + "balance_loss_mlp": 1.02579427, + "epoch": 0.15775867251848735, + "flos": 26980908059520.0, + "grad_norm": 1.724109985483186, + "language_loss": 0.79490525, + "learning_rate": 3.831332492408336e-06, + "loss": 0.81756282, + "num_input_tokens_seen": 27701985, + "step": 1312, + "time_per_iteration": 2.6470558643341064 + }, + { + "auxiliary_loss_clip": 0.01172476, + "auxiliary_loss_mlp": 0.01032919, + "balance_loss_clip": 0.98531139, + "balance_loss_mlp": 1.02206504, + "epoch": 0.15787891540912644, + "flos": 19240398812160.0, + "grad_norm": 1.9603737783887212, + "language_loss": 0.6896286, + "learning_rate": 3.831019254070957e-06, + "loss": 0.71168256, + "num_input_tokens_seen": 27719770, + "step": 1313, + "time_per_iteration": 2.7181379795074463 + }, + { + "auxiliary_loss_clip": 0.01129661, + "auxiliary_loss_mlp": 0.01036263, + "balance_loss_clip": 0.90885997, + "balance_loss_mlp": 1.02629733, + "epoch": 0.15799915829976552, + "flos": 27271102037760.0, + "grad_norm": 3.145769362056849, + "language_loss": 0.95011485, + "learning_rate": 3.8307057379694135e-06, + "loss": 0.9717741, + "num_input_tokens_seen": 27739105, + "step": 1314, + "time_per_iteration": 2.7956840991973877 + }, + { + "auxiliary_loss_clip": 0.01225889, + "auxiliary_loss_mlp": 0.01040173, + "balance_loss_clip": 1.06454921, + "balance_loss_mlp": 1.02976644, + "epoch": 0.15811940119040463, + "flos": 20405520270720.0, + "grad_norm": 4.759124202285313, + "language_loss": 0.82529992, + "learning_rate": 3.830391944151264e-06, + "loss": 0.84796047, + "num_input_tokens_seen": 27754985, + "step": 1315, + "time_per_iteration": 2.6450910568237305 + }, + { + "auxiliary_loss_clip": 0.01179872, + "auxiliary_loss_mlp": 0.01034699, + "balance_loss_clip": 0.98755157, + "balance_loss_mlp": 1.02456617, + "epoch": 0.1582396440810437, + "flos": 32599347661440.0, + "grad_norm": 2.669666386514809, + "language_loss": 0.67509663, + "learning_rate": 3.830077872664114e-06, + "loss": 0.69724232, + "num_input_tokens_seen": 27776110, + "step": 1316, + "time_per_iteration": 2.7221081256866455 + }, + { + "auxiliary_loss_clip": 0.01108915, + "auxiliary_loss_mlp": 0.01036181, + "balance_loss_clip": 0.87114024, + "balance_loss_mlp": 1.0252136, + "epoch": 0.1583598869716828, + "flos": 33800559310080.0, + "grad_norm": 1.6550874475074442, + "language_loss": 0.72978562, + "learning_rate": 3.829763523555604e-06, + "loss": 0.75123662, + "num_input_tokens_seen": 27796510, + "step": 1317, + "time_per_iteration": 2.8979389667510986 + }, + { + "auxiliary_loss_clip": 0.01198079, + "auxiliary_loss_mlp": 0.01037177, + "balance_loss_clip": 1.03062165, + "balance_loss_mlp": 1.02803373, + "epoch": 0.15848012986232188, + "flos": 24681332378880.0, + "grad_norm": 1.988387602959972, + "language_loss": 0.78066468, + "learning_rate": 3.829448896873423e-06, + "loss": 0.8030172, + "num_input_tokens_seen": 27815610, + "step": 1318, + "time_per_iteration": 3.634068012237549 + }, + { + "auxiliary_loss_clip": 0.01116187, + "auxiliary_loss_mlp": 0.00512664, + "balance_loss_clip": 0.91165543, + "balance_loss_mlp": 0.0, + "epoch": 0.158600372752961, + "flos": 22602068766720.0, + "grad_norm": 1.8245406101272919, + "language_loss": 0.78906119, + "learning_rate": 3.829133992665299e-06, + "loss": 0.80534971, + "num_input_tokens_seen": 27834735, + "step": 1319, + "time_per_iteration": 2.7156777381896973 + }, + { + "auxiliary_loss_clip": 0.01192874, + "auxiliary_loss_mlp": 0.01038082, + "balance_loss_clip": 1.02541375, + "balance_loss_mlp": 1.0282588, + "epoch": 0.15872061564360007, + "flos": 27927944092800.0, + "grad_norm": 2.728895109371179, + "language_loss": 0.88877189, + "learning_rate": 3.828818810979002e-06, + "loss": 0.91108143, + "num_input_tokens_seen": 27853065, + "step": 1320, + "time_per_iteration": 3.6816604137420654 + }, + { + "auxiliary_loss_clip": 0.0122905, + "auxiliary_loss_mlp": 0.01036908, + "balance_loss_clip": 1.07122445, + "balance_loss_mlp": 1.02646518, + "epoch": 0.15884085853423915, + "flos": 23696805525120.0, + "grad_norm": 3.6187043107331562, + "language_loss": 0.80428243, + "learning_rate": 3.8285033518623454e-06, + "loss": 0.82694203, + "num_input_tokens_seen": 27873315, + "step": 1321, + "time_per_iteration": 3.5356597900390625 + }, + { + "auxiliary_loss_clip": 0.01208031, + "auxiliary_loss_mlp": 0.01046346, + "balance_loss_clip": 1.03008902, + "balance_loss_mlp": 1.03588569, + "epoch": 0.15896110142487826, + "flos": 23112359331840.0, + "grad_norm": 3.6741583874061123, + "language_loss": 0.81282139, + "learning_rate": 3.8281876153631845e-06, + "loss": 0.83536518, + "num_input_tokens_seen": 27890070, + "step": 1322, + "time_per_iteration": 2.7654972076416016 + }, + { + "auxiliary_loss_clip": 0.01128756, + "auxiliary_loss_mlp": 0.0103715, + "balance_loss_clip": 0.9112879, + "balance_loss_mlp": 1.0268029, + "epoch": 0.15908134431551735, + "flos": 14685238632960.0, + "grad_norm": 2.283873835324681, + "language_loss": 0.64684719, + "learning_rate": 3.827871601529416e-06, + "loss": 0.66850626, + "num_input_tokens_seen": 27908590, + "step": 1323, + "time_per_iteration": 3.6751461029052734 + }, + { + "auxiliary_loss_clip": 0.01148348, + "auxiliary_loss_mlp": 0.01035744, + "balance_loss_clip": 0.94937176, + "balance_loss_mlp": 1.02552199, + "epoch": 0.15920158720615643, + "flos": 20193611984640.0, + "grad_norm": 2.1457368068038423, + "language_loss": 0.80418634, + "learning_rate": 3.827555310408979e-06, + "loss": 0.82602727, + "num_input_tokens_seen": 27927985, + "step": 1324, + "time_per_iteration": 2.7321925163269043 + }, + { + "auxiliary_loss_clip": 0.01149421, + "auxiliary_loss_mlp": 0.01034799, + "balance_loss_clip": 0.95168638, + "balance_loss_mlp": 1.02459502, + "epoch": 0.1593218300967955, + "flos": 24826626892800.0, + "grad_norm": 2.387753861037093, + "language_loss": 0.82764047, + "learning_rate": 3.827238742049854e-06, + "loss": 0.84948266, + "num_input_tokens_seen": 27948280, + "step": 1325, + "time_per_iteration": 2.778846502304077 + }, + { + "auxiliary_loss_clip": 0.01223857, + "auxiliary_loss_mlp": 0.01039604, + "balance_loss_clip": 1.06378937, + "balance_loss_mlp": 1.0291729, + "epoch": 0.15944207298743462, + "flos": 28328707111680.0, + "grad_norm": 2.523169901091213, + "language_loss": 0.51492929, + "learning_rate": 3.826921896500066e-06, + "loss": 0.53756392, + "num_input_tokens_seen": 27969565, + "step": 1326, + "time_per_iteration": 2.7187156677246094 + }, + { + "auxiliary_loss_clip": 0.011608, + "auxiliary_loss_mlp": 0.01037907, + "balance_loss_clip": 0.95133537, + "balance_loss_mlp": 1.02738678, + "epoch": 0.1595623158780737, + "flos": 22964838174720.0, + "grad_norm": 3.7035682127817564, + "language_loss": 0.78030205, + "learning_rate": 3.826604773807678e-06, + "loss": 0.80228913, + "num_input_tokens_seen": 27987540, + "step": 1327, + "time_per_iteration": 2.7418172359466553 + }, + { + "auxiliary_loss_clip": 0.01178965, + "auxiliary_loss_mlp": 0.01043568, + "balance_loss_clip": 0.98568732, + "balance_loss_mlp": 1.03221965, + "epoch": 0.1596825587687128, + "flos": 19710540950400.0, + "grad_norm": 2.5398733592583986, + "language_loss": 0.73133051, + "learning_rate": 3.826287374020798e-06, + "loss": 0.75355589, + "num_input_tokens_seen": 28002345, + "step": 1328, + "time_per_iteration": 2.8399479389190674 + }, + { + "auxiliary_loss_clip": 0.01230583, + "auxiliary_loss_mlp": 0.01043767, + "balance_loss_clip": 1.07088161, + "balance_loss_mlp": 1.03267503, + "epoch": 0.1598028016593519, + "flos": 22637727993600.0, + "grad_norm": 2.050008345979209, + "language_loss": 0.81356287, + "learning_rate": 3.825969697187575e-06, + "loss": 0.83630633, + "num_input_tokens_seen": 28021675, + "step": 1329, + "time_per_iteration": 2.6108405590057373 + }, + { + "auxiliary_loss_clip": 0.0114857, + "auxiliary_loss_mlp": 0.01034282, + "balance_loss_clip": 0.94639212, + "balance_loss_mlp": 1.02370799, + "epoch": 0.15992304454999098, + "flos": 20482908122880.0, + "grad_norm": 2.1045760636906268, + "language_loss": 0.69450676, + "learning_rate": 3.8256517433562015e-06, + "loss": 0.7163353, + "num_input_tokens_seen": 28039615, + "step": 1330, + "time_per_iteration": 2.8081045150756836 + }, + { + "auxiliary_loss_clip": 0.01225876, + "auxiliary_loss_mlp": 0.01030749, + "balance_loss_clip": 1.0664804, + "balance_loss_mlp": 1.02109337, + "epoch": 0.16004328744063007, + "flos": 17676094533120.0, + "grad_norm": 2.9351499852022824, + "language_loss": 0.9168129, + "learning_rate": 3.82533351257491e-06, + "loss": 0.9393791, + "num_input_tokens_seen": 28057565, + "step": 1331, + "time_per_iteration": 2.7340893745422363 + }, + { + "auxiliary_loss_clip": 0.01199501, + "auxiliary_loss_mlp": 0.01035657, + "balance_loss_clip": 1.02910233, + "balance_loss_mlp": 1.02539277, + "epoch": 0.16016353033126918, + "flos": 24098717779200.0, + "grad_norm": 1.8183862824216752, + "language_loss": 0.8891874, + "learning_rate": 3.825015004891975e-06, + "loss": 0.91153902, + "num_input_tokens_seen": 28076305, + "step": 1332, + "time_per_iteration": 2.7515745162963867 + }, + { + "auxiliary_loss_clip": 0.01196343, + "auxiliary_loss_mlp": 0.0104183, + "balance_loss_clip": 1.02574873, + "balance_loss_mlp": 1.03259158, + "epoch": 0.16028377322190826, + "flos": 27634841112960.0, + "grad_norm": 2.3733690683583997, + "language_loss": 0.75537658, + "learning_rate": 3.824696220355716e-06, + "loss": 0.77775836, + "num_input_tokens_seen": 28097895, + "step": 1333, + "time_per_iteration": 2.724102735519409 + }, + { + "auxiliary_loss_clip": 0.01176999, + "auxiliary_loss_mlp": 0.01031107, + "balance_loss_clip": 0.98944068, + "balance_loss_mlp": 1.02067649, + "epoch": 0.16040401611254734, + "flos": 20961202648320.0, + "grad_norm": 1.741916810758222, + "language_loss": 0.78454232, + "learning_rate": 3.824377159014491e-06, + "loss": 0.8066234, + "num_input_tokens_seen": 28118790, + "step": 1334, + "time_per_iteration": 2.7346081733703613 + }, + { + "auxiliary_loss_clip": 0.01199565, + "auxiliary_loss_mlp": 0.01035054, + "balance_loss_clip": 1.02807331, + "balance_loss_mlp": 1.02581513, + "epoch": 0.16052425900318643, + "flos": 21247051080960.0, + "grad_norm": 1.8657550774324023, + "language_loss": 0.84661633, + "learning_rate": 3.824057820916702e-06, + "loss": 0.86896253, + "num_input_tokens_seen": 28135995, + "step": 1335, + "time_per_iteration": 2.6064651012420654 + }, + { + "auxiliary_loss_clip": 0.01178093, + "auxiliary_loss_mlp": 0.01038408, + "balance_loss_clip": 0.98842967, + "balance_loss_mlp": 1.02760124, + "epoch": 0.16064450189382554, + "flos": 15524004096000.0, + "grad_norm": 2.577679768668053, + "language_loss": 0.71934056, + "learning_rate": 3.8237382061107904e-06, + "loss": 0.74150556, + "num_input_tokens_seen": 28152715, + "step": 1336, + "time_per_iteration": 2.695678472518921 + }, + { + "auxiliary_loss_clip": 0.01066027, + "auxiliary_loss_mlp": 0.01034786, + "balance_loss_clip": 0.82522428, + "balance_loss_mlp": 1.02452254, + "epoch": 0.16076474478446462, + "flos": 21178497974400.0, + "grad_norm": 1.74892580373847, + "language_loss": 0.78785408, + "learning_rate": 3.823418314645243e-06, + "loss": 0.80886215, + "num_input_tokens_seen": 28171590, + "step": 1337, + "time_per_iteration": 3.101031541824341 + }, + { + "auxiliary_loss_clip": 0.01111455, + "auxiliary_loss_mlp": 0.01038036, + "balance_loss_clip": 0.90941173, + "balance_loss_mlp": 1.02777863, + "epoch": 0.1608849876751037, + "flos": 18366476912640.0, + "grad_norm": 1.9967408431785454, + "language_loss": 0.74999905, + "learning_rate": 3.823098146568588e-06, + "loss": 0.77149397, + "num_input_tokens_seen": 28191295, + "step": 1338, + "time_per_iteration": 3.068718433380127 + }, + { + "auxiliary_loss_clip": 0.01199883, + "auxiliary_loss_mlp": 0.01037562, + "balance_loss_clip": 1.02746344, + "balance_loss_mlp": 1.02751231, + "epoch": 0.1610052305657428, + "flos": 29497024880640.0, + "grad_norm": 2.0299281150895387, + "language_loss": 0.71630627, + "learning_rate": 3.822777701929394e-06, + "loss": 0.73868072, + "num_input_tokens_seen": 28213120, + "step": 1339, + "time_per_iteration": 2.76759934425354 + }, + { + "auxiliary_loss_clip": 0.01190407, + "auxiliary_loss_mlp": 0.01028898, + "balance_loss_clip": 1.02364922, + "balance_loss_mlp": 1.01850867, + "epoch": 0.1611254734563819, + "flos": 26797871329920.0, + "grad_norm": 5.2056821899560655, + "language_loss": 0.73404741, + "learning_rate": 3.8224569807762714e-06, + "loss": 0.75624049, + "num_input_tokens_seen": 28232440, + "step": 1340, + "time_per_iteration": 2.6593432426452637 + }, + { + "auxiliary_loss_clip": 0.01105512, + "auxiliary_loss_mlp": 0.01033972, + "balance_loss_clip": 0.90313339, + "balance_loss_mlp": 1.02379775, + "epoch": 0.16124571634702098, + "flos": 22419570741120.0, + "grad_norm": 1.962838041610015, + "language_loss": 0.76292431, + "learning_rate": 3.822135983157873e-06, + "loss": 0.78431916, + "num_input_tokens_seen": 28251715, + "step": 1341, + "time_per_iteration": 2.829360246658325 + }, + { + "auxiliary_loss_clip": 0.01218078, + "auxiliary_loss_mlp": 0.00511836, + "balance_loss_clip": 1.06297159, + "balance_loss_mlp": 0.0, + "epoch": 0.16136595923766006, + "flos": 10999116103680.0, + "grad_norm": 1.8874134325337513, + "language_loss": 0.84166116, + "learning_rate": 3.821814709122896e-06, + "loss": 0.85896033, + "num_input_tokens_seen": 28269765, + "step": 1342, + "time_per_iteration": 2.6678433418273926 + }, + { + "auxiliary_loss_clip": 0.01174938, + "auxiliary_loss_mlp": 0.01030328, + "balance_loss_clip": 0.98765361, + "balance_loss_mlp": 1.02121425, + "epoch": 0.16148620212829917, + "flos": 21214983214080.0, + "grad_norm": 2.7105797611897984, + "language_loss": 0.84240186, + "learning_rate": 3.821493158720076e-06, + "loss": 0.86445451, + "num_input_tokens_seen": 28288870, + "step": 1343, + "time_per_iteration": 2.697660446166992 + }, + { + "auxiliary_loss_clip": 0.01148892, + "auxiliary_loss_mlp": 0.01028629, + "balance_loss_clip": 0.94474268, + "balance_loss_mlp": 1.01865101, + "epoch": 0.16160644501893826, + "flos": 16758468760320.0, + "grad_norm": 3.96502951962356, + "language_loss": 0.73528969, + "learning_rate": 3.821171331998191e-06, + "loss": 0.75706482, + "num_input_tokens_seen": 28305400, + "step": 1344, + "time_per_iteration": 2.694359064102173 + }, + { + "auxiliary_loss_clip": 0.01073829, + "auxiliary_loss_mlp": 0.01013676, + "balance_loss_clip": 0.92658949, + "balance_loss_mlp": 1.01076722, + "epoch": 0.16172668790957734, + "flos": 64444967308800.0, + "grad_norm": 0.7134281642087476, + "language_loss": 0.54463947, + "learning_rate": 3.820849229006064e-06, + "loss": 0.56551456, + "num_input_tokens_seen": 28373150, + "step": 1345, + "time_per_iteration": 4.341669082641602 + }, + { + "auxiliary_loss_clip": 0.01222683, + "auxiliary_loss_mlp": 0.01032535, + "balance_loss_clip": 1.06449258, + "balance_loss_mlp": 1.02289677, + "epoch": 0.16184693080021645, + "flos": 23257689759360.0, + "grad_norm": 1.8929225158072116, + "language_loss": 0.70793378, + "learning_rate": 3.8205268497925564e-06, + "loss": 0.73048604, + "num_input_tokens_seen": 28393620, + "step": 1346, + "time_per_iteration": 2.709151029586792 + }, + { + "auxiliary_loss_clip": 0.01225059, + "auxiliary_loss_mlp": 0.01037218, + "balance_loss_clip": 1.06685078, + "balance_loss_mlp": 1.02749693, + "epoch": 0.16196717369085553, + "flos": 17451113696640.0, + "grad_norm": 2.2433807653124034, + "language_loss": 0.78287518, + "learning_rate": 3.8202041944065725e-06, + "loss": 0.80549788, + "num_input_tokens_seen": 28409440, + "step": 1347, + "time_per_iteration": 4.3888099193573 + }, + { + "auxiliary_loss_clip": 0.01224972, + "auxiliary_loss_mlp": 0.01049354, + "balance_loss_clip": 1.06775987, + "balance_loss_mlp": 1.03930461, + "epoch": 0.16208741658149461, + "flos": 23873377806720.0, + "grad_norm": 2.095587134676866, + "language_loss": 0.74042475, + "learning_rate": 3.819881262897061e-06, + "loss": 0.76316798, + "num_input_tokens_seen": 28427575, + "step": 1348, + "time_per_iteration": 2.6616156101226807 + }, + { + "auxiliary_loss_clip": 0.01155111, + "auxiliary_loss_mlp": 0.01031595, + "balance_loss_clip": 0.95550942, + "balance_loss_mlp": 1.02070546, + "epoch": 0.1622076594721337, + "flos": 25884806584320.0, + "grad_norm": 1.9674407205446616, + "language_loss": 0.73118067, + "learning_rate": 3.819558055313008e-06, + "loss": 0.7530477, + "num_input_tokens_seen": 28448260, + "step": 1349, + "time_per_iteration": 3.859008550643921 + }, + { + "auxiliary_loss_clip": 0.01205327, + "auxiliary_loss_mlp": 0.01035562, + "balance_loss_clip": 1.02814221, + "balance_loss_mlp": 1.02510738, + "epoch": 0.1623279023627728, + "flos": 21539759011200.0, + "grad_norm": 1.8250355889671899, + "language_loss": 0.77137792, + "learning_rate": 3.819234571703444e-06, + "loss": 0.79378682, + "num_input_tokens_seen": 28467085, + "step": 1350, + "time_per_iteration": 2.7311325073242188 + }, + { + "auxiliary_loss_clip": 0.01196539, + "auxiliary_loss_mlp": 0.01032615, + "balance_loss_clip": 1.02585495, + "balance_loss_mlp": 1.02198124, + "epoch": 0.1624481452534119, + "flos": 22085421494400.0, + "grad_norm": 4.177353053455567, + "language_loss": 0.85450327, + "learning_rate": 3.8189108121174435e-06, + "loss": 0.87679482, + "num_input_tokens_seen": 28486850, + "step": 1351, + "time_per_iteration": 2.688704252243042 + }, + { + "auxiliary_loss_clip": 0.0114816, + "auxiliary_loss_mlp": 0.01041045, + "balance_loss_clip": 0.95183593, + "balance_loss_mlp": 1.03135979, + "epoch": 0.16256838814405097, + "flos": 27087490690560.0, + "grad_norm": 1.6588539738047434, + "language_loss": 0.8325218, + "learning_rate": 3.818586776604118e-06, + "loss": 0.85441387, + "num_input_tokens_seen": 28507490, + "step": 1352, + "time_per_iteration": 2.8357765674591064 + }, + { + "auxiliary_loss_clip": 0.01173113, + "auxiliary_loss_mlp": 0.01036146, + "balance_loss_clip": 0.98886377, + "balance_loss_mlp": 1.02658522, + "epoch": 0.16268863103469008, + "flos": 20120354196480.0, + "grad_norm": 2.435039474963051, + "language_loss": 0.61731768, + "learning_rate": 3.818262465212625e-06, + "loss": 0.63941032, + "num_input_tokens_seen": 28527615, + "step": 1353, + "time_per_iteration": 2.6923904418945312 + }, + { + "auxiliary_loss_clip": 0.0119183, + "auxiliary_loss_mlp": 0.01041048, + "balance_loss_clip": 1.02644074, + "balance_loss_mlp": 1.02981853, + "epoch": 0.16280887392532917, + "flos": 18332792933760.0, + "grad_norm": 2.2543790029165565, + "language_loss": 0.7723825, + "learning_rate": 3.817937877992161e-06, + "loss": 0.79471135, + "num_input_tokens_seen": 28544910, + "step": 1354, + "time_per_iteration": 2.6832761764526367 + }, + { + "auxiliary_loss_clip": 0.01151785, + "auxiliary_loss_mlp": 0.00513382, + "balance_loss_clip": 0.94767535, + "balance_loss_mlp": 0.0, + "epoch": 0.16292911681596825, + "flos": 11874330892800.0, + "grad_norm": 2.736183273654555, + "language_loss": 0.8593148, + "learning_rate": 3.817613014991967e-06, + "loss": 0.87596643, + "num_input_tokens_seen": 28561050, + "step": 1355, + "time_per_iteration": 2.723414421081543 + }, + { + "auxiliary_loss_clip": 0.01145418, + "auxiliary_loss_mlp": 0.01033503, + "balance_loss_clip": 0.94779253, + "balance_loss_mlp": 1.02295876, + "epoch": 0.16304935970660733, + "flos": 26103466627200.0, + "grad_norm": 2.904815283288997, + "language_loss": 0.76510632, + "learning_rate": 3.817287876261323e-06, + "loss": 0.78689551, + "num_input_tokens_seen": 28581385, + "step": 1356, + "time_per_iteration": 2.778052806854248 + }, + { + "auxiliary_loss_clip": 0.01173548, + "auxiliary_loss_mlp": 0.01034189, + "balance_loss_clip": 0.98941112, + "balance_loss_mlp": 1.02391315, + "epoch": 0.16316960259724644, + "flos": 29351945848320.0, + "grad_norm": 1.921336172079915, + "language_loss": 0.80283612, + "learning_rate": 3.816962461849553e-06, + "loss": 0.8249135, + "num_input_tokens_seen": 28603255, + "step": 1357, + "time_per_iteration": 2.7081849575042725 + }, + { + "auxiliary_loss_clip": 0.01170005, + "auxiliary_loss_mlp": 0.01046444, + "balance_loss_clip": 0.99035835, + "balance_loss_mlp": 1.03603125, + "epoch": 0.16328984548788553, + "flos": 20886759711360.0, + "grad_norm": 2.1947504871981334, + "language_loss": 0.84230983, + "learning_rate": 3.8166367718060235e-06, + "loss": 0.8644743, + "num_input_tokens_seen": 28623145, + "step": 1358, + "time_per_iteration": 2.781306266784668 + }, + { + "auxiliary_loss_clip": 0.01194329, + "auxiliary_loss_mlp": 0.01041104, + "balance_loss_clip": 1.02480876, + "balance_loss_mlp": 1.03089952, + "epoch": 0.1634100883785246, + "flos": 18041090584320.0, + "grad_norm": 3.215973221450571, + "language_loss": 0.75816429, + "learning_rate": 3.816310806180139e-06, + "loss": 0.78051865, + "num_input_tokens_seen": 28641555, + "step": 1359, + "time_per_iteration": 2.630038022994995 + }, + { + "auxiliary_loss_clip": 0.01175196, + "auxiliary_loss_mlp": 0.01040597, + "balance_loss_clip": 0.99081945, + "balance_loss_mlp": 1.02976036, + "epoch": 0.16353033126916372, + "flos": 24572128055040.0, + "grad_norm": 2.4769853023471207, + "language_loss": 0.80895495, + "learning_rate": 3.81598456502135e-06, + "loss": 0.83111292, + "num_input_tokens_seen": 28661575, + "step": 1360, + "time_per_iteration": 2.7918105125427246 + }, + { + "auxiliary_loss_clip": 0.01172336, + "auxiliary_loss_mlp": 0.01043185, + "balance_loss_clip": 0.98990321, + "balance_loss_mlp": 1.03339767, + "epoch": 0.1636505741598028, + "flos": 19892895321600.0, + "grad_norm": 2.4143799513276476, + "language_loss": 0.87117857, + "learning_rate": 3.8156580483791455e-06, + "loss": 0.89333379, + "num_input_tokens_seen": 28676765, + "step": 1361, + "time_per_iteration": 2.7079498767852783 + }, + { + "auxiliary_loss_clip": 0.01225145, + "auxiliary_loss_mlp": 0.01043375, + "balance_loss_clip": 1.06699133, + "balance_loss_mlp": 1.03351617, + "epoch": 0.16377081705044189, + "flos": 28402611344640.0, + "grad_norm": 3.2987828218081034, + "language_loss": 0.76970673, + "learning_rate": 3.815331256303059e-06, + "loss": 0.7923919, + "num_input_tokens_seen": 28696795, + "step": 1362, + "time_per_iteration": 2.6465866565704346 + }, + { + "auxiliary_loss_clip": 0.01151056, + "auxiliary_loss_mlp": 0.01035812, + "balance_loss_clip": 0.95272148, + "balance_loss_mlp": 1.02607822, + "epoch": 0.163891059941081, + "flos": 21908059113600.0, + "grad_norm": 2.24841691293018, + "language_loss": 0.77314472, + "learning_rate": 3.815004188842665e-06, + "loss": 0.79501343, + "num_input_tokens_seen": 28714835, + "step": 1363, + "time_per_iteration": 2.7577476501464844 + }, + { + "auxiliary_loss_clip": 0.01169374, + "auxiliary_loss_mlp": 0.01038831, + "balance_loss_clip": 0.98472166, + "balance_loss_mlp": 1.02800059, + "epoch": 0.16401130283172008, + "flos": 26797619934720.0, + "grad_norm": 1.5889086051956918, + "language_loss": 0.79373538, + "learning_rate": 3.814676846047578e-06, + "loss": 0.81581736, + "num_input_tokens_seen": 28735710, + "step": 1364, + "time_per_iteration": 2.6466174125671387 + }, + { + "auxiliary_loss_clip": 0.01197286, + "auxiliary_loss_mlp": 0.01036264, + "balance_loss_clip": 1.02683485, + "balance_loss_mlp": 1.02638161, + "epoch": 0.16413154572235916, + "flos": 32997417160320.0, + "grad_norm": 1.9903161547617487, + "language_loss": 0.69697499, + "learning_rate": 3.8143492279674565e-06, + "loss": 0.71931052, + "num_input_tokens_seen": 28758405, + "step": 1365, + "time_per_iteration": 2.741753578186035 + }, + { + "auxiliary_loss_clip": 0.0107277, + "auxiliary_loss_mlp": 0.01015017, + "balance_loss_clip": 0.92945015, + "balance_loss_mlp": 1.01196563, + "epoch": 0.16425178861299825, + "flos": 40113622074240.0, + "grad_norm": 0.8403724127406301, + "language_loss": 0.58444053, + "learning_rate": 3.8140213346519997e-06, + "loss": 0.60531843, + "num_input_tokens_seen": 28809000, + "step": 1366, + "time_per_iteration": 3.0189197063446045 + }, + { + "auxiliary_loss_clip": 0.01144968, + "auxiliary_loss_mlp": 0.01038732, + "balance_loss_clip": 0.94962001, + "balance_loss_mlp": 1.02799749, + "epoch": 0.16437203150363736, + "flos": 25447486498560.0, + "grad_norm": 1.6236656475091613, + "language_loss": 0.76990461, + "learning_rate": 3.813693166150948e-06, + "loss": 0.79174161, + "num_input_tokens_seen": 28829210, + "step": 1367, + "time_per_iteration": 2.75616192817688 + }, + { + "auxiliary_loss_clip": 0.01140927, + "auxiliary_loss_mlp": 0.01041078, + "balance_loss_clip": 0.94656241, + "balance_loss_mlp": 1.03073645, + "epoch": 0.16449227439427644, + "flos": 23476888506240.0, + "grad_norm": 2.3957648375082097, + "language_loss": 0.85827637, + "learning_rate": 3.813364722514086e-06, + "loss": 0.88009644, + "num_input_tokens_seen": 28847545, + "step": 1368, + "time_per_iteration": 2.731529474258423 + }, + { + "auxiliary_loss_clip": 0.0119676, + "auxiliary_loss_mlp": 0.01035601, + "balance_loss_clip": 1.0246855, + "balance_loss_mlp": 1.0252533, + "epoch": 0.16461251728491552, + "flos": 13545217802880.0, + "grad_norm": 2.1969635188897074, + "language_loss": 0.80349106, + "learning_rate": 3.8130360037912368e-06, + "loss": 0.82581466, + "num_input_tokens_seen": 28863990, + "step": 1369, + "time_per_iteration": 2.6767406463623047 + }, + { + "auxiliary_loss_clip": 0.01198295, + "auxiliary_loss_mlp": 0.01036895, + "balance_loss_clip": 1.02594948, + "balance_loss_mlp": 1.02608299, + "epoch": 0.16473276017555463, + "flos": 23003298662400.0, + "grad_norm": 3.487927604007106, + "language_loss": 0.81958044, + "learning_rate": 3.812707010032268e-06, + "loss": 0.84193236, + "num_input_tokens_seen": 28883045, + "step": 1370, + "time_per_iteration": 2.6695046424865723 + }, + { + "auxiliary_loss_clip": 0.01206354, + "auxiliary_loss_mlp": 0.01034164, + "balance_loss_clip": 1.0315392, + "balance_loss_mlp": 1.02324426, + "epoch": 0.16485300306619372, + "flos": 24790680357120.0, + "grad_norm": 1.6545445717159, + "language_loss": 0.79190958, + "learning_rate": 3.8123777412870863e-06, + "loss": 0.81431472, + "num_input_tokens_seen": 28902545, + "step": 1371, + "time_per_iteration": 3.610395669937134 + }, + { + "auxiliary_loss_clip": 0.01180912, + "auxiliary_loss_mlp": 0.01043115, + "balance_loss_clip": 0.98797774, + "balance_loss_mlp": 1.03234982, + "epoch": 0.1649732459568328, + "flos": 21106497162240.0, + "grad_norm": 2.469921635860881, + "language_loss": 0.7849226, + "learning_rate": 3.812048197605643e-06, + "loss": 0.80716288, + "num_input_tokens_seen": 28921440, + "step": 1372, + "time_per_iteration": 3.6074323654174805 + }, + { + "auxiliary_loss_clip": 0.01199672, + "auxiliary_loss_mlp": 0.01038693, + "balance_loss_clip": 1.02537227, + "balance_loss_mlp": 1.02841687, + "epoch": 0.16509348884747188, + "flos": 20266726118400.0, + "grad_norm": 2.30115802056192, + "language_loss": 0.8108393, + "learning_rate": 3.8117183790379277e-06, + "loss": 0.83322293, + "num_input_tokens_seen": 28939890, + "step": 1373, + "time_per_iteration": 3.603337287902832 + }, + { + "auxiliary_loss_clip": 0.0122449, + "auxiliary_loss_mlp": 0.0103776, + "balance_loss_clip": 1.06531429, + "balance_loss_mlp": 1.02729297, + "epoch": 0.165213731738111, + "flos": 11035493602560.0, + "grad_norm": 2.6946931105895047, + "language_loss": 0.94101357, + "learning_rate": 3.811388285633976e-06, + "loss": 0.9636361, + "num_input_tokens_seen": 28955875, + "step": 1374, + "time_per_iteration": 2.586242437362671 + }, + { + "auxiliary_loss_clip": 0.01135791, + "auxiliary_loss_mlp": 0.01041911, + "balance_loss_clip": 0.91291404, + "balance_loss_mlp": 1.03142071, + "epoch": 0.16533397462875007, + "flos": 29972051268480.0, + "grad_norm": 2.6632460288515962, + "language_loss": 0.62029874, + "learning_rate": 3.811057917443861e-06, + "loss": 0.64207578, + "num_input_tokens_seen": 28975140, + "step": 1375, + "time_per_iteration": 3.731567859649658 + }, + { + "auxiliary_loss_clip": 0.01096372, + "auxiliary_loss_mlp": 0.01007607, + "balance_loss_clip": 0.96673644, + "balance_loss_mlp": 1.00463867, + "epoch": 0.16545421751938916, + "flos": 65556763027200.0, + "grad_norm": 0.8635600519694454, + "language_loss": 0.68326902, + "learning_rate": 3.8107272745177e-06, + "loss": 0.70430875, + "num_input_tokens_seen": 29047470, + "step": 1376, + "time_per_iteration": 3.3976633548736572 + }, + { + "auxiliary_loss_clip": 0.01152493, + "auxiliary_loss_mlp": 0.01037909, + "balance_loss_clip": 0.95041496, + "balance_loss_mlp": 1.02737641, + "epoch": 0.16557446041002827, + "flos": 22492361652480.0, + "grad_norm": 1.8878752421168679, + "language_loss": 0.78548479, + "learning_rate": 3.8103963569056513e-06, + "loss": 0.8073889, + "num_input_tokens_seen": 29066605, + "step": 1377, + "time_per_iteration": 2.8779776096343994 + }, + { + "auxiliary_loss_clip": 0.01167162, + "auxiliary_loss_mlp": 0.01031977, + "balance_loss_clip": 0.98354048, + "balance_loss_mlp": 1.02173138, + "epoch": 0.16569470330066735, + "flos": 24602723464320.0, + "grad_norm": 1.4991864409543958, + "language_loss": 0.88092923, + "learning_rate": 3.8100651646579146e-06, + "loss": 0.9029206, + "num_input_tokens_seen": 29085815, + "step": 1378, + "time_per_iteration": 2.766885757446289 + }, + { + "auxiliary_loss_clip": 0.01168713, + "auxiliary_loss_mlp": 0.01043615, + "balance_loss_clip": 0.98317707, + "balance_loss_mlp": 1.03236187, + "epoch": 0.16581494619130643, + "flos": 15006207588480.0, + "grad_norm": 2.10724451050223, + "language_loss": 0.93243736, + "learning_rate": 3.8097336978247317e-06, + "loss": 0.9545607, + "num_input_tokens_seen": 29102520, + "step": 1379, + "time_per_iteration": 2.679023027420044 + }, + { + "auxiliary_loss_clip": 0.01165907, + "auxiliary_loss_mlp": 0.01036932, + "balance_loss_clip": 0.9871347, + "balance_loss_mlp": 1.02583337, + "epoch": 0.16593518908194552, + "flos": 17420338719360.0, + "grad_norm": 2.063402531299996, + "language_loss": 0.88867396, + "learning_rate": 3.8094019564563854e-06, + "loss": 0.91070235, + "num_input_tokens_seen": 29119450, + "step": 1380, + "time_per_iteration": 2.7230007648468018 + }, + { + "auxiliary_loss_clip": 0.01222978, + "auxiliary_loss_mlp": 0.00512996, + "balance_loss_clip": 1.06444144, + "balance_loss_mlp": 0.0, + "epoch": 0.16605543197258463, + "flos": 20412631163520.0, + "grad_norm": 2.690852936604157, + "language_loss": 0.74797118, + "learning_rate": 3.809069940603201e-06, + "loss": 0.76533091, + "num_input_tokens_seen": 29137405, + "step": 1381, + "time_per_iteration": 2.6895554065704346 + }, + { + "auxiliary_loss_clip": 0.01166176, + "auxiliary_loss_mlp": 0.01036373, + "balance_loss_clip": 0.98626894, + "balance_loss_mlp": 1.02624035, + "epoch": 0.1661756748632237, + "flos": 14209745368320.0, + "grad_norm": 2.734088682508372, + "language_loss": 0.77889168, + "learning_rate": 3.8087376503155452e-06, + "loss": 0.80091721, + "num_input_tokens_seen": 29154890, + "step": 1382, + "time_per_iteration": 2.7906787395477295 + }, + { + "auxiliary_loss_clip": 0.01092433, + "auxiliary_loss_mlp": 0.01007913, + "balance_loss_clip": 0.96495485, + "balance_loss_mlp": 1.00496864, + "epoch": 0.1662959177538628, + "flos": 66080877350400.0, + "grad_norm": 0.8942483157566214, + "language_loss": 0.56259048, + "learning_rate": 3.808405085643826e-06, + "loss": 0.58359396, + "num_input_tokens_seen": 29219770, + "step": 1383, + "time_per_iteration": 3.3455023765563965 + }, + { + "auxiliary_loss_clip": 0.01226055, + "auxiliary_loss_mlp": 0.00513332, + "balance_loss_clip": 1.0669775, + "balance_loss_mlp": 0.0, + "epoch": 0.1664161606445019, + "flos": 20740567357440.0, + "grad_norm": 2.305877923580696, + "language_loss": 0.88546479, + "learning_rate": 3.8080722466384925e-06, + "loss": 0.90285861, + "num_input_tokens_seen": 29237620, + "step": 1384, + "time_per_iteration": 2.6317086219787598 + }, + { + "auxiliary_loss_clip": 0.01227297, + "auxiliary_loss_mlp": 0.01037748, + "balance_loss_clip": 1.06471491, + "balance_loss_mlp": 1.02657843, + "epoch": 0.166536403535141, + "flos": 25260930236160.0, + "grad_norm": 2.553661721902918, + "language_loss": 0.70728099, + "learning_rate": 3.8077391333500376e-06, + "loss": 0.72993147, + "num_input_tokens_seen": 29256760, + "step": 1385, + "time_per_iteration": 2.695826768875122 + }, + { + "auxiliary_loss_clip": 0.0117846, + "auxiliary_loss_mlp": 0.01037077, + "balance_loss_clip": 0.99184942, + "balance_loss_mlp": 1.02714133, + "epoch": 0.16665664642578007, + "flos": 25447450584960.0, + "grad_norm": 3.909058888044514, + "language_loss": 0.7626456, + "learning_rate": 3.8074057458289934e-06, + "loss": 0.78480095, + "num_input_tokens_seen": 29277450, + "step": 1386, + "time_per_iteration": 2.6945807933807373 + }, + { + "auxiliary_loss_clip": 0.01175205, + "auxiliary_loss_mlp": 0.0104294, + "balance_loss_clip": 0.98472893, + "balance_loss_mlp": 1.03128743, + "epoch": 0.16677688931641918, + "flos": 22200767043840.0, + "grad_norm": 2.4035890649494505, + "language_loss": 0.82018149, + "learning_rate": 3.807072084125934e-06, + "loss": 0.84236288, + "num_input_tokens_seen": 29299300, + "step": 1387, + "time_per_iteration": 2.7260634899139404 + }, + { + "auxiliary_loss_clip": 0.01171764, + "auxiliary_loss_mlp": 0.010341, + "balance_loss_clip": 0.98741627, + "balance_loss_mlp": 1.02316844, + "epoch": 0.16689713220705826, + "flos": 16945958776320.0, + "grad_norm": 2.862357471139226, + "language_loss": 0.8054508, + "learning_rate": 3.806738148291477e-06, + "loss": 0.8275094, + "num_input_tokens_seen": 29316125, + "step": 1388, + "time_per_iteration": 2.7022085189819336 + }, + { + "auxiliary_loss_clip": 0.01102572, + "auxiliary_loss_mlp": 0.0103053, + "balance_loss_clip": 0.86927676, + "balance_loss_mlp": 1.01900244, + "epoch": 0.16701737509769735, + "flos": 36244423923840.0, + "grad_norm": 3.144160270138103, + "language_loss": 0.71032071, + "learning_rate": 3.8064039383762793e-06, + "loss": 0.73165172, + "num_input_tokens_seen": 29338490, + "step": 1389, + "time_per_iteration": 2.884230852127075 + }, + { + "auxiliary_loss_clip": 0.0120492, + "auxiliary_loss_mlp": 0.0105088, + "balance_loss_clip": 1.03200424, + "balance_loss_mlp": 1.03982377, + "epoch": 0.16713761798833643, + "flos": 23258659426560.0, + "grad_norm": 2.1454272498157714, + "language_loss": 0.77443141, + "learning_rate": 3.8060694544310396e-06, + "loss": 0.79698944, + "num_input_tokens_seen": 29357000, + "step": 1390, + "time_per_iteration": 2.6517200469970703 + }, + { + "auxiliary_loss_clip": 0.01228322, + "auxiliary_loss_mlp": 0.0104177, + "balance_loss_clip": 1.06688142, + "balance_loss_mlp": 1.03023648, + "epoch": 0.16725786087897554, + "flos": 25302515207040.0, + "grad_norm": 2.0383041911458157, + "language_loss": 0.78482127, + "learning_rate": 3.8057346965065006e-06, + "loss": 0.80752218, + "num_input_tokens_seen": 29378230, + "step": 1391, + "time_per_iteration": 2.671250104904175 + }, + { + "auxiliary_loss_clip": 0.01177232, + "auxiliary_loss_mlp": 0.01039682, + "balance_loss_clip": 0.99183881, + "balance_loss_mlp": 1.02860117, + "epoch": 0.16737810376961462, + "flos": 31831541516160.0, + "grad_norm": 1.5801888932684403, + "language_loss": 0.84449565, + "learning_rate": 3.805399664653443e-06, + "loss": 0.86666483, + "num_input_tokens_seen": 29400370, + "step": 1392, + "time_per_iteration": 2.7395074367523193 + }, + { + "auxiliary_loss_clip": 0.0123019, + "auxiliary_loss_mlp": 0.01038075, + "balance_loss_clip": 1.06989098, + "balance_loss_mlp": 1.02622509, + "epoch": 0.1674983466602537, + "flos": 27961843553280.0, + "grad_norm": 2.274120195430468, + "language_loss": 0.74743885, + "learning_rate": 3.805064358922692e-06, + "loss": 0.77012151, + "num_input_tokens_seen": 29418660, + "step": 1393, + "time_per_iteration": 2.766101598739624 + }, + { + "auxiliary_loss_clip": 0.01206111, + "auxiliary_loss_mlp": 0.010321, + "balance_loss_clip": 1.02795959, + "balance_loss_mlp": 1.0205245, + "epoch": 0.16761858955089282, + "flos": 21762656858880.0, + "grad_norm": 22.66668688936988, + "language_loss": 0.809403, + "learning_rate": 3.8047287793651136e-06, + "loss": 0.83178508, + "num_input_tokens_seen": 29440105, + "step": 1394, + "time_per_iteration": 2.691049098968506 + }, + { + "auxiliary_loss_clip": 0.01156223, + "auxiliary_loss_mlp": 0.01045533, + "balance_loss_clip": 0.95113599, + "balance_loss_mlp": 1.0344944, + "epoch": 0.1677388324415319, + "flos": 23805507058560.0, + "grad_norm": 1.7131021272030629, + "language_loss": 0.88445741, + "learning_rate": 3.8043929260316137e-06, + "loss": 0.90647495, + "num_input_tokens_seen": 29458260, + "step": 1395, + "time_per_iteration": 2.7362864017486572 + }, + { + "auxiliary_loss_clip": 0.01185691, + "auxiliary_loss_mlp": 0.01037721, + "balance_loss_clip": 0.99640107, + "balance_loss_mlp": 1.02640212, + "epoch": 0.16785907533217098, + "flos": 20558859431040.0, + "grad_norm": 2.0967501845434677, + "language_loss": 0.83406186, + "learning_rate": 3.8040567989731417e-06, + "loss": 0.856296, + "num_input_tokens_seen": 29476205, + "step": 1396, + "time_per_iteration": 2.7680506706237793 + }, + { + "auxiliary_loss_clip": 0.01199543, + "auxiliary_loss_mlp": 0.01040172, + "balance_loss_clip": 1.03057981, + "balance_loss_mlp": 1.02937174, + "epoch": 0.16797931822281006, + "flos": 15669657745920.0, + "grad_norm": 2.1251352729465696, + "language_loss": 0.79537284, + "learning_rate": 3.8037203982406876e-06, + "loss": 0.81777, + "num_input_tokens_seen": 29494370, + "step": 1397, + "time_per_iteration": 3.6273117065429688 + }, + { + "auxiliary_loss_clip": 0.01229812, + "auxiliary_loss_mlp": 0.01038996, + "balance_loss_clip": 1.06995606, + "balance_loss_mlp": 1.02832115, + "epoch": 0.16809956111344918, + "flos": 16541101607040.0, + "grad_norm": 2.821024566612327, + "language_loss": 0.73251808, + "learning_rate": 3.8033837238852835e-06, + "loss": 0.75520617, + "num_input_tokens_seen": 29511070, + "step": 1398, + "time_per_iteration": 3.5206897258758545 + }, + { + "auxiliary_loss_clip": 0.01166011, + "auxiliary_loss_mlp": 0.01039618, + "balance_loss_clip": 0.98498553, + "balance_loss_mlp": 1.02857876, + "epoch": 0.16821980400408826, + "flos": 23258084808960.0, + "grad_norm": 3.0094325975407523, + "language_loss": 0.6969648, + "learning_rate": 3.8030467759580017e-06, + "loss": 0.71902108, + "num_input_tokens_seen": 29531990, + "step": 1399, + "time_per_iteration": 3.5997350215911865 + }, + { + "auxiliary_loss_clip": 0.01206569, + "auxiliary_loss_mlp": 0.01036164, + "balance_loss_clip": 1.02896619, + "balance_loss_mlp": 1.02511883, + "epoch": 0.16834004689472734, + "flos": 20774754126720.0, + "grad_norm": 1.8630002765656843, + "language_loss": 0.86819911, + "learning_rate": 3.802709554509958e-06, + "loss": 0.89062643, + "num_input_tokens_seen": 29549790, + "step": 1400, + "time_per_iteration": 2.6420013904571533 + }, + { + "auxiliary_loss_clip": 0.01177181, + "auxiliary_loss_mlp": 0.01035898, + "balance_loss_clip": 0.98686767, + "balance_loss_mlp": 1.02520537, + "epoch": 0.16846028978536645, + "flos": 26687302289280.0, + "grad_norm": 2.0965180156309757, + "language_loss": 0.78979558, + "learning_rate": 3.8023720595923083e-06, + "loss": 0.81192636, + "num_input_tokens_seen": 29569045, + "step": 1401, + "time_per_iteration": 3.6308817863464355 + }, + { + "auxiliary_loss_clip": 0.01127971, + "auxiliary_loss_mlp": 0.01036157, + "balance_loss_clip": 0.90906727, + "balance_loss_mlp": 1.02487409, + "epoch": 0.16858053267600553, + "flos": 18843298980480.0, + "grad_norm": 2.094966158744771, + "language_loss": 0.87226856, + "learning_rate": 3.80203429125625e-06, + "loss": 0.89390981, + "num_input_tokens_seen": 29587220, + "step": 1402, + "time_per_iteration": 2.833799362182617 + }, + { + "auxiliary_loss_clip": 0.01092562, + "auxiliary_loss_mlp": 0.01044957, + "balance_loss_clip": 0.86892653, + "balance_loss_mlp": 1.03425753, + "epoch": 0.16870077556664462, + "flos": 27744548227200.0, + "grad_norm": 1.7423724580542872, + "language_loss": 0.70183802, + "learning_rate": 3.8016962495530225e-06, + "loss": 0.7232132, + "num_input_tokens_seen": 29606410, + "step": 1403, + "time_per_iteration": 2.9510486125946045 + }, + { + "auxiliary_loss_clip": 0.01228888, + "auxiliary_loss_mlp": 0.01048147, + "balance_loss_clip": 1.06666946, + "balance_loss_mlp": 1.03650069, + "epoch": 0.1688210184572837, + "flos": 13730768484480.0, + "grad_norm": 2.7508675625765675, + "language_loss": 0.76953971, + "learning_rate": 3.8013579345339063e-06, + "loss": 0.79231006, + "num_input_tokens_seen": 29621275, + "step": 1404, + "time_per_iteration": 2.9704792499542236 + }, + { + "auxiliary_loss_clip": 0.01153277, + "auxiliary_loss_mlp": 0.0104196, + "balance_loss_clip": 0.94990838, + "balance_loss_mlp": 1.0304265, + "epoch": 0.1689412613479228, + "flos": 26468785900800.0, + "grad_norm": 1.811527815906086, + "language_loss": 0.68891543, + "learning_rate": 3.801019346250224e-06, + "loss": 0.71086776, + "num_input_tokens_seen": 29641420, + "step": 1405, + "time_per_iteration": 2.7511074542999268 + }, + { + "auxiliary_loss_clip": 0.01197997, + "auxiliary_loss_mlp": 0.01041426, + "balance_loss_clip": 1.02697778, + "balance_loss_mlp": 1.03042293, + "epoch": 0.1690615042385619, + "flos": 21138852337920.0, + "grad_norm": 2.183373448049715, + "language_loss": 0.83420312, + "learning_rate": 3.8006804847533395e-06, + "loss": 0.8565973, + "num_input_tokens_seen": 29660935, + "step": 1406, + "time_per_iteration": 2.7344295978546143 + }, + { + "auxiliary_loss_clip": 0.01225179, + "auxiliary_loss_mlp": 0.01035881, + "balance_loss_clip": 1.0651114, + "balance_loss_mlp": 1.02503872, + "epoch": 0.16918174712920098, + "flos": 20849340718080.0, + "grad_norm": 1.9823045089133344, + "language_loss": 0.85516024, + "learning_rate": 3.8003413500946556e-06, + "loss": 0.87777084, + "num_input_tokens_seen": 29681045, + "step": 1407, + "time_per_iteration": 2.6064391136169434 + }, + { + "auxiliary_loss_clip": 0.01177435, + "auxiliary_loss_mlp": 0.01039807, + "balance_loss_clip": 0.98971063, + "balance_loss_mlp": 1.02858913, + "epoch": 0.1693019900198401, + "flos": 16983270028800.0, + "grad_norm": 2.9301298777421376, + "language_loss": 0.82947063, + "learning_rate": 3.8000019423256216e-06, + "loss": 0.85164309, + "num_input_tokens_seen": 29698810, + "step": 1408, + "time_per_iteration": 2.716766834259033 + }, + { + "auxiliary_loss_clip": 0.0116891, + "auxiliary_loss_mlp": 0.01032901, + "balance_loss_clip": 0.98715341, + "balance_loss_mlp": 1.02173138, + "epoch": 0.16942223291047917, + "flos": 26796901662720.0, + "grad_norm": 1.6643276161375455, + "language_loss": 0.87803066, + "learning_rate": 3.7996622614977234e-06, + "loss": 0.90004873, + "num_input_tokens_seen": 29720000, + "step": 1409, + "time_per_iteration": 2.7608656883239746 + }, + { + "auxiliary_loss_clip": 0.01179871, + "auxiliary_loss_mlp": 0.01033703, + "balance_loss_clip": 0.99408054, + "balance_loss_mlp": 1.02323699, + "epoch": 0.16954247580111825, + "flos": 18583700411520.0, + "grad_norm": 1.8792327974995824, + "language_loss": 0.78975183, + "learning_rate": 3.799322307662492e-06, + "loss": 0.81188762, + "num_input_tokens_seen": 29737820, + "step": 1410, + "time_per_iteration": 2.6653943061828613 + }, + { + "auxiliary_loss_clip": 0.01132829, + "auxiliary_loss_mlp": 0.01038597, + "balance_loss_clip": 0.91091198, + "balance_loss_mlp": 1.02777267, + "epoch": 0.16966271869175734, + "flos": 13983651210240.0, + "grad_norm": 2.6637490049954144, + "language_loss": 0.84153545, + "learning_rate": 3.798982080871496e-06, + "loss": 0.86324978, + "num_input_tokens_seen": 29752960, + "step": 1411, + "time_per_iteration": 2.858543634414673 + }, + { + "auxiliary_loss_clip": 0.01226181, + "auxiliary_loss_mlp": 0.01043423, + "balance_loss_clip": 1.06472135, + "balance_loss_mlp": 1.03243792, + "epoch": 0.16978296158239645, + "flos": 37487328284160.0, + "grad_norm": 1.7963658691745503, + "language_loss": 0.67623925, + "learning_rate": 3.798641581176349e-06, + "loss": 0.69893527, + "num_input_tokens_seen": 29775240, + "step": 1412, + "time_per_iteration": 2.7818212509155273 + }, + { + "auxiliary_loss_clip": 0.01178926, + "auxiliary_loss_mlp": 0.01034546, + "balance_loss_clip": 0.98849797, + "balance_loss_mlp": 1.02344799, + "epoch": 0.16990320447303553, + "flos": 28328958506880.0, + "grad_norm": 1.9321253865052566, + "language_loss": 0.7453177, + "learning_rate": 3.7983008086287044e-06, + "loss": 0.76745248, + "num_input_tokens_seen": 29796560, + "step": 1413, + "time_per_iteration": 2.888583183288574 + }, + { + "auxiliary_loss_clip": 0.01175837, + "auxiliary_loss_mlp": 0.0104072, + "balance_loss_clip": 0.98741597, + "balance_loss_mlp": 1.02917469, + "epoch": 0.1700234473636746, + "flos": 20188189031040.0, + "grad_norm": 2.104536048954657, + "language_loss": 0.79420412, + "learning_rate": 3.797959763280257e-06, + "loss": 0.81636965, + "num_input_tokens_seen": 29815245, + "step": 1414, + "time_per_iteration": 2.7548437118530273 + }, + { + "auxiliary_loss_clip": 0.01204052, + "auxiliary_loss_mlp": 0.01033027, + "balance_loss_clip": 1.02784097, + "balance_loss_mlp": 1.02232218, + "epoch": 0.17014369025431372, + "flos": 24858658846080.0, + "grad_norm": 2.0234235623028196, + "language_loss": 0.78750932, + "learning_rate": 3.797618445182743e-06, + "loss": 0.80988014, + "num_input_tokens_seen": 29836640, + "step": 1415, + "time_per_iteration": 2.7917912006378174 + }, + { + "auxiliary_loss_clip": 0.01123761, + "auxiliary_loss_mlp": 0.01037812, + "balance_loss_clip": 0.9085151, + "balance_loss_mlp": 1.0268805, + "epoch": 0.1702639331449528, + "flos": 16467233287680.0, + "grad_norm": 2.123408716301427, + "language_loss": 0.85083234, + "learning_rate": 3.79727685438794e-06, + "loss": 0.87244803, + "num_input_tokens_seen": 29850830, + "step": 1416, + "time_per_iteration": 2.7668654918670654 + }, + { + "auxiliary_loss_clip": 0.01119114, + "auxiliary_loss_mlp": 0.0101919, + "balance_loss_clip": 1.00534439, + "balance_loss_mlp": 1.01635325, + "epoch": 0.1703841760355919, + "flos": 52508870979840.0, + "grad_norm": 0.8452593685421399, + "language_loss": 0.61690223, + "learning_rate": 3.796934990947667e-06, + "loss": 0.63828528, + "num_input_tokens_seen": 29912515, + "step": 1417, + "time_per_iteration": 3.3304359912872314 + }, + { + "auxiliary_loss_clip": 0.01118169, + "auxiliary_loss_mlp": 0.01011299, + "balance_loss_clip": 1.00520253, + "balance_loss_mlp": 1.00833023, + "epoch": 0.170504418926231, + "flos": 49370637576960.0, + "grad_norm": 0.8880301249399027, + "language_loss": 0.62475717, + "learning_rate": 3.7965928549137854e-06, + "loss": 0.64605188, + "num_input_tokens_seen": 29969330, + "step": 1418, + "time_per_iteration": 3.2497291564941406 + }, + { + "auxiliary_loss_clip": 0.01155023, + "auxiliary_loss_mlp": 0.01037782, + "balance_loss_clip": 0.9445222, + "balance_loss_mlp": 1.025581, + "epoch": 0.17062466181687008, + "flos": 25849219184640.0, + "grad_norm": 2.2554956943355102, + "language_loss": 0.77602601, + "learning_rate": 3.7962504463381953e-06, + "loss": 0.79795414, + "num_input_tokens_seen": 29990820, + "step": 1419, + "time_per_iteration": 2.8293683528900146 + }, + { + "auxiliary_loss_clip": 0.01171748, + "auxiliary_loss_mlp": 0.00514531, + "balance_loss_clip": 0.99181825, + "balance_loss_mlp": 0.0, + "epoch": 0.17074490470750917, + "flos": 20960412549120.0, + "grad_norm": 1.6460654896764897, + "language_loss": 0.78753793, + "learning_rate": 3.7959077652728412e-06, + "loss": 0.80440068, + "num_input_tokens_seen": 30009275, + "step": 1420, + "time_per_iteration": 2.807758331298828 + }, + { + "auxiliary_loss_clip": 0.01179242, + "auxiliary_loss_mlp": 0.01034656, + "balance_loss_clip": 0.98725057, + "balance_loss_mlp": 1.02422547, + "epoch": 0.17086514759814825, + "flos": 20959766104320.0, + "grad_norm": 3.523663969793356, + "language_loss": 0.77102774, + "learning_rate": 3.795564811769707e-06, + "loss": 0.79316676, + "num_input_tokens_seen": 30027630, + "step": 1421, + "time_per_iteration": 2.8398704528808594 + }, + { + "auxiliary_loss_clip": 0.01180273, + "auxiliary_loss_mlp": 0.01043881, + "balance_loss_clip": 0.99241507, + "balance_loss_mlp": 1.03269291, + "epoch": 0.17098539048878736, + "flos": 28474073452800.0, + "grad_norm": 4.2845651314078825, + "language_loss": 0.77693117, + "learning_rate": 3.795221585880818e-06, + "loss": 0.7991727, + "num_input_tokens_seen": 30048310, + "step": 1422, + "time_per_iteration": 2.832432508468628 + }, + { + "auxiliary_loss_clip": 0.01156065, + "auxiliary_loss_mlp": 0.01043361, + "balance_loss_clip": 0.95455134, + "balance_loss_mlp": 1.03316247, + "epoch": 0.17110563337942644, + "flos": 16290014561280.0, + "grad_norm": 1.8788148368840485, + "language_loss": 0.91354489, + "learning_rate": 3.794878087658242e-06, + "loss": 0.93553913, + "num_input_tokens_seen": 30066080, + "step": 1423, + "time_per_iteration": 3.8106629848480225 + }, + { + "auxiliary_loss_clip": 0.0120018, + "auxiliary_loss_mlp": 0.01043355, + "balance_loss_clip": 1.02455962, + "balance_loss_mlp": 1.03192937, + "epoch": 0.17122587627006552, + "flos": 29674207693440.0, + "grad_norm": 2.064014335545876, + "language_loss": 0.78529012, + "learning_rate": 3.7945343171540873e-06, + "loss": 0.80772549, + "num_input_tokens_seen": 30086955, + "step": 1424, + "time_per_iteration": 3.6748948097229004 + }, + { + "auxiliary_loss_clip": 0.01225283, + "auxiliary_loss_mlp": 0.01040619, + "balance_loss_clip": 1.0657357, + "balance_loss_mlp": 1.02974761, + "epoch": 0.17134611916070464, + "flos": 25338389915520.0, + "grad_norm": 1.876904761324827, + "language_loss": 0.78961802, + "learning_rate": 3.7941902744205033e-06, + "loss": 0.81227702, + "num_input_tokens_seen": 30107990, + "step": 1425, + "time_per_iteration": 3.636847972869873 + }, + { + "auxiliary_loss_clip": 0.01180808, + "auxiliary_loss_mlp": 0.01048501, + "balance_loss_clip": 0.98764682, + "balance_loss_mlp": 1.03640175, + "epoch": 0.17146636205134372, + "flos": 13953845900160.0, + "grad_norm": 2.4178285585800414, + "language_loss": 0.83488053, + "learning_rate": 3.7938459595096817e-06, + "loss": 0.85717368, + "num_input_tokens_seen": 30126535, + "step": 1426, + "time_per_iteration": 2.7260162830352783 + }, + { + "auxiliary_loss_clip": 0.01208838, + "auxiliary_loss_mlp": 0.01049588, + "balance_loss_clip": 1.02816522, + "balance_loss_mlp": 1.03831649, + "epoch": 0.1715866049419828, + "flos": 23915214172800.0, + "grad_norm": 1.8371556832963232, + "language_loss": 0.85951352, + "learning_rate": 3.7935013724738545e-06, + "loss": 0.88209784, + "num_input_tokens_seen": 30147035, + "step": 1427, + "time_per_iteration": 2.6999030113220215 + }, + { + "auxiliary_loss_clip": 0.01198362, + "auxiliary_loss_mlp": 0.01038331, + "balance_loss_clip": 1.02885675, + "balance_loss_mlp": 1.0276562, + "epoch": 0.17170684783262188, + "flos": 22709369669760.0, + "grad_norm": 1.7841115132447969, + "language_loss": 0.78020763, + "learning_rate": 3.7931565133652945e-06, + "loss": 0.80257457, + "num_input_tokens_seen": 30167110, + "step": 1428, + "time_per_iteration": 3.6212048530578613 + }, + { + "auxiliary_loss_clip": 0.01226178, + "auxiliary_loss_mlp": 0.01043214, + "balance_loss_clip": 1.06572771, + "balance_loss_mlp": 1.03261685, + "epoch": 0.171827090723261, + "flos": 26613290315520.0, + "grad_norm": 2.1220361233838347, + "language_loss": 0.67734516, + "learning_rate": 3.792811382236317e-06, + "loss": 0.70003909, + "num_input_tokens_seen": 30185620, + "step": 1429, + "time_per_iteration": 2.7034380435943604 + }, + { + "auxiliary_loss_clip": 0.01202902, + "auxiliary_loss_mlp": 0.01044799, + "balance_loss_clip": 1.02567172, + "balance_loss_mlp": 1.03250825, + "epoch": 0.17194733361390008, + "flos": 28148507556480.0, + "grad_norm": 1.9969203302635559, + "language_loss": 0.77906257, + "learning_rate": 3.792465979139279e-06, + "loss": 0.8015396, + "num_input_tokens_seen": 30208225, + "step": 1430, + "time_per_iteration": 2.781289577484131 + }, + { + "auxiliary_loss_clip": 0.01066794, + "auxiliary_loss_mlp": 0.01025992, + "balance_loss_clip": 0.92473978, + "balance_loss_mlp": 1.02317858, + "epoch": 0.17206757650453916, + "flos": 65530689753600.0, + "grad_norm": 0.9182395529163935, + "language_loss": 0.65608394, + "learning_rate": 3.792120304126576e-06, + "loss": 0.67701185, + "num_input_tokens_seen": 30271600, + "step": 1431, + "time_per_iteration": 3.3546483516693115 + }, + { + "auxiliary_loss_clip": 0.0108356, + "auxiliary_loss_mlp": 0.01031992, + "balance_loss_clip": 0.83414197, + "balance_loss_mlp": 1.02131104, + "epoch": 0.17218781939517827, + "flos": 22273486128000.0, + "grad_norm": 1.7753491629477431, + "language_loss": 0.83670795, + "learning_rate": 3.791774357250649e-06, + "loss": 0.85786349, + "num_input_tokens_seen": 30290430, + "step": 1432, + "time_per_iteration": 2.8963241577148438 + }, + { + "auxiliary_loss_clip": 0.01172456, + "auxiliary_loss_mlp": 0.01044494, + "balance_loss_clip": 0.98588187, + "balance_loss_mlp": 1.03438532, + "epoch": 0.17230806228581735, + "flos": 14137313592960.0, + "grad_norm": 2.4909173675047254, + "language_loss": 0.78918099, + "learning_rate": 3.7914281385639757e-06, + "loss": 0.81135046, + "num_input_tokens_seen": 30308305, + "step": 1433, + "time_per_iteration": 3.1354689598083496 + }, + { + "auxiliary_loss_clip": 0.01199886, + "auxiliary_loss_mlp": 0.01027959, + "balance_loss_clip": 1.02324224, + "balance_loss_mlp": 1.01737881, + "epoch": 0.17242830517645644, + "flos": 20704836303360.0, + "grad_norm": 2.5577361921335053, + "language_loss": 0.7932899, + "learning_rate": 3.7910816481190784e-06, + "loss": 0.81556839, + "num_input_tokens_seen": 30328120, + "step": 1434, + "time_per_iteration": 2.757871627807617 + }, + { + "auxiliary_loss_clip": 0.01164365, + "auxiliary_loss_mlp": 0.0103717, + "balance_loss_clip": 0.98424363, + "balance_loss_mlp": 1.02645934, + "epoch": 0.17254854806709552, + "flos": 30774582887040.0, + "grad_norm": 4.807541807255067, + "language_loss": 0.74394101, + "learning_rate": 3.7907348859685193e-06, + "loss": 0.76595634, + "num_input_tokens_seen": 30349825, + "step": 1435, + "time_per_iteration": 2.7814297676086426 + }, + { + "auxiliary_loss_clip": 0.01196979, + "auxiliary_loss_mlp": 0.01042543, + "balance_loss_clip": 1.02792871, + "balance_loss_mlp": 1.0319097, + "epoch": 0.17266879095773463, + "flos": 26614726859520.0, + "grad_norm": 2.9186280852513384, + "language_loss": 0.80757022, + "learning_rate": 3.790387852164902e-06, + "loss": 0.82996547, + "num_input_tokens_seen": 30370555, + "step": 1436, + "time_per_iteration": 2.762472152709961 + }, + { + "auxiliary_loss_clip": 0.01203421, + "auxiliary_loss_mlp": 0.01033867, + "balance_loss_clip": 1.02963495, + "balance_loss_mlp": 1.02350211, + "epoch": 0.1727890338483737, + "flos": 20266295155200.0, + "grad_norm": 2.322758544492617, + "language_loss": 0.76646519, + "learning_rate": 3.7900405467608707e-06, + "loss": 0.78883815, + "num_input_tokens_seen": 30390100, + "step": 1437, + "time_per_iteration": 2.8286359310150146 + }, + { + "auxiliary_loss_clip": 0.01113072, + "auxiliary_loss_mlp": 0.01042534, + "balance_loss_clip": 0.90380079, + "balance_loss_mlp": 1.03176951, + "epoch": 0.1729092767390128, + "flos": 18179812909440.0, + "grad_norm": 3.4805953102639102, + "language_loss": 0.78619957, + "learning_rate": 3.7896929698091114e-06, + "loss": 0.80775571, + "num_input_tokens_seen": 30402915, + "step": 1438, + "time_per_iteration": 2.8334546089172363 + }, + { + "auxiliary_loss_clip": 0.01228106, + "auxiliary_loss_mlp": 0.01044439, + "balance_loss_clip": 1.0701623, + "balance_loss_mlp": 1.03396606, + "epoch": 0.1730295196296519, + "flos": 26759518583040.0, + "grad_norm": 8.047413521362607, + "language_loss": 0.68061864, + "learning_rate": 3.7893451213623518e-06, + "loss": 0.70334417, + "num_input_tokens_seen": 30420145, + "step": 1439, + "time_per_iteration": 2.7071220874786377 + }, + { + "auxiliary_loss_clip": 0.0119988, + "auxiliary_loss_mlp": 0.00513302, + "balance_loss_clip": 1.02914011, + "balance_loss_mlp": 0.0, + "epoch": 0.173149762520291, + "flos": 23842531002240.0, + "grad_norm": 2.352311688435801, + "language_loss": 0.82119548, + "learning_rate": 3.7889970014733606e-06, + "loss": 0.83832729, + "num_input_tokens_seen": 30439250, + "step": 1440, + "time_per_iteration": 2.681981325149536 + }, + { + "auxiliary_loss_clip": 0.01111483, + "auxiliary_loss_mlp": 0.0103358, + "balance_loss_clip": 0.90653938, + "balance_loss_mlp": 1.0224576, + "epoch": 0.17327000541093007, + "flos": 23368186972800.0, + "grad_norm": 1.8103426251369898, + "language_loss": 0.78279555, + "learning_rate": 3.7886486101949463e-06, + "loss": 0.80424613, + "num_input_tokens_seen": 30460430, + "step": 1441, + "time_per_iteration": 2.914259433746338 + }, + { + "auxiliary_loss_clip": 0.01113242, + "auxiliary_loss_mlp": 0.01040823, + "balance_loss_clip": 0.90452492, + "balance_loss_mlp": 1.03069639, + "epoch": 0.17339024830156918, + "flos": 18221290139520.0, + "grad_norm": 1.9300783351187654, + "language_loss": 0.87917191, + "learning_rate": 3.7882999475799594e-06, + "loss": 0.90071255, + "num_input_tokens_seen": 30478465, + "step": 1442, + "time_per_iteration": 2.739293098449707 + }, + { + "auxiliary_loss_clip": 0.01109786, + "auxiliary_loss_mlp": 0.01045594, + "balance_loss_clip": 0.90851307, + "balance_loss_mlp": 1.03512108, + "epoch": 0.17351049119220827, + "flos": 23332024955520.0, + "grad_norm": 1.9953370864430304, + "language_loss": 0.81413865, + "learning_rate": 3.787951013681293e-06, + "loss": 0.83569241, + "num_input_tokens_seen": 30496510, + "step": 1443, + "time_per_iteration": 2.804501533508301 + }, + { + "auxiliary_loss_clip": 0.01199603, + "auxiliary_loss_mlp": 0.01036723, + "balance_loss_clip": 1.025702, + "balance_loss_mlp": 1.02512431, + "epoch": 0.17363073408284735, + "flos": 23803495896960.0, + "grad_norm": 2.980128390737047, + "language_loss": 0.77432156, + "learning_rate": 3.787601808551879e-06, + "loss": 0.79668474, + "num_input_tokens_seen": 30516325, + "step": 1444, + "time_per_iteration": 2.621258497238159 + }, + { + "auxiliary_loss_clip": 0.01150633, + "auxiliary_loss_mlp": 0.01038106, + "balance_loss_clip": 0.94700438, + "balance_loss_mlp": 1.02775252, + "epoch": 0.17375097697348643, + "flos": 18515290959360.0, + "grad_norm": 2.2607887065648313, + "language_loss": 0.84397078, + "learning_rate": 3.7872523322446926e-06, + "loss": 0.8658582, + "num_input_tokens_seen": 30535210, + "step": 1445, + "time_per_iteration": 2.6806859970092773 + }, + { + "auxiliary_loss_clip": 0.01128853, + "auxiliary_loss_mlp": 0.01043489, + "balance_loss_clip": 0.90522611, + "balance_loss_mlp": 1.03228378, + "epoch": 0.17387121986412554, + "flos": 38877897456000.0, + "grad_norm": 1.7809123342784217, + "language_loss": 0.5980953, + "learning_rate": 3.7869025848127478e-06, + "loss": 0.61981869, + "num_input_tokens_seen": 30559405, + "step": 1446, + "time_per_iteration": 2.9269449710845947 + }, + { + "auxiliary_loss_clip": 0.01202065, + "auxiliary_loss_mlp": 0.01037966, + "balance_loss_clip": 1.02586102, + "balance_loss_mlp": 1.02688539, + "epoch": 0.17399146275476463, + "flos": 20375714960640.0, + "grad_norm": 3.117096040110838, + "language_loss": 0.8040812, + "learning_rate": 3.786552566309102e-06, + "loss": 0.82648158, + "num_input_tokens_seen": 30577615, + "step": 1447, + "time_per_iteration": 2.7163279056549072 + }, + { + "auxiliary_loss_clip": 0.0117279, + "auxiliary_loss_mlp": 0.0051283, + "balance_loss_clip": 0.99166471, + "balance_loss_mlp": 0.0, + "epoch": 0.1741117056454037, + "flos": 19164339763200.0, + "grad_norm": 2.199166247484167, + "language_loss": 0.86009085, + "learning_rate": 3.7862022767868517e-06, + "loss": 0.87694705, + "num_input_tokens_seen": 30595205, + "step": 1448, + "time_per_iteration": 2.6848223209381104 + }, + { + "auxiliary_loss_clip": 0.01142588, + "auxiliary_loss_mlp": 0.01034783, + "balance_loss_clip": 0.94916302, + "balance_loss_mlp": 1.02531826, + "epoch": 0.17423194853604282, + "flos": 25374300537600.0, + "grad_norm": 2.2954632175649468, + "language_loss": 0.84348398, + "learning_rate": 3.7858517162991367e-06, + "loss": 0.86525774, + "num_input_tokens_seen": 30615280, + "step": 1449, + "time_per_iteration": 4.633136510848999 + }, + { + "auxiliary_loss_clip": 0.01147777, + "auxiliary_loss_mlp": 0.01044535, + "balance_loss_clip": 0.94505131, + "balance_loss_mlp": 1.03417528, + "epoch": 0.1743521914266819, + "flos": 25191874339200.0, + "grad_norm": 2.995356298914423, + "language_loss": 0.60276783, + "learning_rate": 3.7855008848991363e-06, + "loss": 0.62469089, + "num_input_tokens_seen": 30633485, + "step": 1450, + "time_per_iteration": 2.794569730758667 + }, + { + "auxiliary_loss_clip": 0.01171987, + "auxiliary_loss_mlp": 0.01036817, + "balance_loss_clip": 0.98951465, + "balance_loss_mlp": 1.02703631, + "epoch": 0.17447243431732098, + "flos": 25666577504640.0, + "grad_norm": 2.2701537957500495, + "language_loss": 0.77384645, + "learning_rate": 3.7851497826400714e-06, + "loss": 0.7959345, + "num_input_tokens_seen": 30653625, + "step": 1451, + "time_per_iteration": 3.669771909713745 + }, + { + "auxiliary_loss_clip": 0.01225796, + "auxiliary_loss_mlp": 0.01038714, + "balance_loss_clip": 1.06794906, + "balance_loss_mlp": 1.02735984, + "epoch": 0.17459267720796007, + "flos": 36281950657920.0, + "grad_norm": 1.9764037467243134, + "language_loss": 0.76244128, + "learning_rate": 3.7847984095752034e-06, + "loss": 0.78508639, + "num_input_tokens_seen": 30677080, + "step": 1452, + "time_per_iteration": 2.7307026386260986 + }, + { + "auxiliary_loss_clip": 0.01223014, + "auxiliary_loss_mlp": 0.01033341, + "balance_loss_clip": 1.06572843, + "balance_loss_mlp": 1.02279711, + "epoch": 0.17471292009859918, + "flos": 20011113959040.0, + "grad_norm": 2.276075502912865, + "language_loss": 0.80475903, + "learning_rate": 3.784446765757836e-06, + "loss": 0.82732254, + "num_input_tokens_seen": 30695725, + "step": 1453, + "time_per_iteration": 3.577363967895508 + }, + { + "auxiliary_loss_clip": 0.01136572, + "auxiliary_loss_mlp": 0.01038511, + "balance_loss_clip": 0.94648087, + "balance_loss_mlp": 1.02763367, + "epoch": 0.17483316298923826, + "flos": 27819242559360.0, + "grad_norm": 12.148752459445955, + "language_loss": 0.77837873, + "learning_rate": 3.7840948512413133e-06, + "loss": 0.80012953, + "num_input_tokens_seen": 30713310, + "step": 1454, + "time_per_iteration": 2.7179551124572754 + }, + { + "auxiliary_loss_clip": 0.01142748, + "auxiliary_loss_mlp": 0.01035649, + "balance_loss_clip": 0.94842893, + "balance_loss_mlp": 1.02490246, + "epoch": 0.17495340587987734, + "flos": 44017934791680.0, + "grad_norm": 1.9208443389410874, + "language_loss": 0.7868939, + "learning_rate": 3.7837426660790196e-06, + "loss": 0.80867785, + "num_input_tokens_seen": 30734725, + "step": 1455, + "time_per_iteration": 2.898115396499634 + }, + { + "auxiliary_loss_clip": 0.01220814, + "auxiliary_loss_mlp": 0.01036532, + "balance_loss_clip": 1.06501842, + "balance_loss_mlp": 1.0267452, + "epoch": 0.17507364877051645, + "flos": 20885825957760.0, + "grad_norm": 6.674321453571528, + "language_loss": 0.81916732, + "learning_rate": 3.783390210324382e-06, + "loss": 0.84174085, + "num_input_tokens_seen": 30754450, + "step": 1456, + "time_per_iteration": 2.6191725730895996 + }, + { + "auxiliary_loss_clip": 0.01149205, + "auxiliary_loss_mlp": 0.01037063, + "balance_loss_clip": 0.95058715, + "balance_loss_mlp": 1.02776456, + "epoch": 0.17519389166115554, + "flos": 24717602136960.0, + "grad_norm": 2.543703460139879, + "language_loss": 0.7251094, + "learning_rate": 3.7830374840308676e-06, + "loss": 0.74697208, + "num_input_tokens_seen": 30774605, + "step": 1457, + "time_per_iteration": 2.7458062171936035 + }, + { + "auxiliary_loss_clip": 0.01203494, + "auxiliary_loss_mlp": 0.01036567, + "balance_loss_clip": 1.0303061, + "balance_loss_mlp": 1.02577233, + "epoch": 0.17531413455179462, + "flos": 23798144770560.0, + "grad_norm": 3.3498597740573337, + "language_loss": 0.82826686, + "learning_rate": 3.7826844872519842e-06, + "loss": 0.85066748, + "num_input_tokens_seen": 30792460, + "step": 1458, + "time_per_iteration": 2.680283308029175 + }, + { + "auxiliary_loss_clip": 0.01174443, + "auxiliary_loss_mlp": 0.01033781, + "balance_loss_clip": 0.99123776, + "balance_loss_mlp": 1.02413702, + "epoch": 0.1754343774424337, + "flos": 24572379450240.0, + "grad_norm": 1.9224605460663124, + "language_loss": 0.72373986, + "learning_rate": 3.782331220041282e-06, + "loss": 0.74582219, + "num_input_tokens_seen": 30812525, + "step": 1459, + "time_per_iteration": 2.707919120788574 + }, + { + "auxiliary_loss_clip": 0.01155603, + "auxiliary_loss_mlp": 0.01031963, + "balance_loss_clip": 0.94807923, + "balance_loss_mlp": 1.02178252, + "epoch": 0.17555462033307281, + "flos": 18114599767680.0, + "grad_norm": 2.251678483470549, + "language_loss": 0.82886267, + "learning_rate": 3.7819776824523504e-06, + "loss": 0.85073835, + "num_input_tokens_seen": 30830390, + "step": 1460, + "time_per_iteration": 2.788067579269409 + }, + { + "auxiliary_loss_clip": 0.01184257, + "auxiliary_loss_mlp": 0.01039458, + "balance_loss_clip": 0.9900369, + "balance_loss_mlp": 1.02969456, + "epoch": 0.1756748632237119, + "flos": 28366018364160.0, + "grad_norm": 2.274135279923083, + "language_loss": 0.83871114, + "learning_rate": 3.7816238745388213e-06, + "loss": 0.86094826, + "num_input_tokens_seen": 30849935, + "step": 1461, + "time_per_iteration": 2.766716957092285 + }, + { + "auxiliary_loss_clip": 0.01175595, + "auxiliary_loss_mlp": 0.01033925, + "balance_loss_clip": 0.98489261, + "balance_loss_mlp": 1.02443576, + "epoch": 0.17579510611435098, + "flos": 25732939881600.0, + "grad_norm": 1.9063286374369681, + "language_loss": 0.87322617, + "learning_rate": 3.781269796354367e-06, + "loss": 0.89532137, + "num_input_tokens_seen": 30869555, + "step": 1462, + "time_per_iteration": 2.786714792251587 + }, + { + "auxiliary_loss_clip": 0.01175001, + "auxiliary_loss_mlp": 0.01033383, + "balance_loss_clip": 0.98860651, + "balance_loss_mlp": 1.02354228, + "epoch": 0.1759153490049901, + "flos": 18588081870720.0, + "grad_norm": 1.7220851256988081, + "language_loss": 0.86124539, + "learning_rate": 3.7809154479527006e-06, + "loss": 0.88332921, + "num_input_tokens_seen": 30888760, + "step": 1463, + "time_per_iteration": 2.6971442699432373 + }, + { + "auxiliary_loss_clip": 0.01145051, + "auxiliary_loss_mlp": 0.01033935, + "balance_loss_clip": 0.9495039, + "balance_loss_mlp": 1.02452326, + "epoch": 0.17603559189562917, + "flos": 18619323724800.0, + "grad_norm": 2.04540277755163, + "language_loss": 0.84502792, + "learning_rate": 3.780560829387577e-06, + "loss": 0.86681783, + "num_input_tokens_seen": 30907260, + "step": 1464, + "time_per_iteration": 2.749688148498535 + }, + { + "auxiliary_loss_clip": 0.01122186, + "auxiliary_loss_mlp": 0.01010358, + "balance_loss_clip": 1.00788784, + "balance_loss_mlp": 1.00716352, + "epoch": 0.17615583478626826, + "flos": 60530775373440.0, + "grad_norm": 0.8608600377760629, + "language_loss": 0.5794872, + "learning_rate": 3.7802059407127915e-06, + "loss": 0.60081267, + "num_input_tokens_seen": 30965810, + "step": 1465, + "time_per_iteration": 3.3000898361206055 + }, + { + "auxiliary_loss_clip": 0.01165479, + "auxiliary_loss_mlp": 0.01040373, + "balance_loss_clip": 0.9859513, + "balance_loss_mlp": 1.0305028, + "epoch": 0.17627607767690734, + "flos": 23616221362560.0, + "grad_norm": 2.8407178542088474, + "language_loss": 0.85950536, + "learning_rate": 3.7798507819821797e-06, + "loss": 0.8815639, + "num_input_tokens_seen": 30982935, + "step": 1466, + "time_per_iteration": 2.7498159408569336 + }, + { + "auxiliary_loss_clip": 0.01140965, + "auxiliary_loss_mlp": 0.01025491, + "balance_loss_clip": 0.94846344, + "balance_loss_mlp": 1.01511395, + "epoch": 0.17639632056754645, + "flos": 17639070589440.0, + "grad_norm": 2.12862317081144, + "language_loss": 0.78408253, + "learning_rate": 3.7794953532496197e-06, + "loss": 0.80574703, + "num_input_tokens_seen": 30998840, + "step": 1467, + "time_per_iteration": 2.7872889041900635 + }, + { + "auxiliary_loss_clip": 0.01043797, + "auxiliary_loss_mlp": 0.00505112, + "balance_loss_clip": 0.8961848, + "balance_loss_mlp": 0.0, + "epoch": 0.17651656345818553, + "flos": 57932604910080.0, + "grad_norm": 0.8689808029658669, + "language_loss": 0.57946473, + "learning_rate": 3.7791396545690295e-06, + "loss": 0.59495384, + "num_input_tokens_seen": 31060075, + "step": 1468, + "time_per_iteration": 3.3328428268432617 + }, + { + "auxiliary_loss_clip": 0.01196867, + "auxiliary_loss_mlp": 0.01032512, + "balance_loss_clip": 1.02744627, + "balance_loss_mlp": 1.02309501, + "epoch": 0.17663680634882462, + "flos": 22929502170240.0, + "grad_norm": 4.121504023519162, + "language_loss": 0.80567658, + "learning_rate": 3.7787836859943685e-06, + "loss": 0.82797039, + "num_input_tokens_seen": 31078800, + "step": 1469, + "time_per_iteration": 2.9358811378479004 + }, + { + "auxiliary_loss_clip": 0.01197114, + "auxiliary_loss_mlp": 0.0103353, + "balance_loss_clip": 1.02682936, + "balance_loss_mlp": 1.02368295, + "epoch": 0.17675704923946373, + "flos": 22637979388800.0, + "grad_norm": 3.259092419286681, + "language_loss": 0.78774357, + "learning_rate": 3.7784274475796363e-06, + "loss": 0.81004995, + "num_input_tokens_seen": 31097430, + "step": 1470, + "time_per_iteration": 2.6576266288757324 + }, + { + "auxiliary_loss_clip": 0.01147677, + "auxiliary_loss_mlp": 0.01034041, + "balance_loss_clip": 0.94798827, + "balance_loss_mlp": 1.02391374, + "epoch": 0.1768772921301028, + "flos": 27126525795840.0, + "grad_norm": 2.888968547548288, + "language_loss": 0.7595129, + "learning_rate": 3.7780709393788745e-06, + "loss": 0.78133011, + "num_input_tokens_seen": 31117905, + "step": 1471, + "time_per_iteration": 2.8357183933258057 + }, + { + "auxiliary_loss_clip": 0.01221243, + "auxiliary_loss_mlp": 0.01034908, + "balance_loss_clip": 1.06590056, + "balance_loss_mlp": 1.02506125, + "epoch": 0.1769975350207419, + "flos": 19172133014400.0, + "grad_norm": 2.849647559592671, + "language_loss": 0.75372338, + "learning_rate": 3.777714161446165e-06, + "loss": 0.77628493, + "num_input_tokens_seen": 31137610, + "step": 1472, + "time_per_iteration": 2.6563608646392822 + }, + { + "auxiliary_loss_clip": 0.01196556, + "auxiliary_loss_mlp": 0.01030749, + "balance_loss_clip": 1.02713084, + "balance_loss_mlp": 1.02041352, + "epoch": 0.177117777911381, + "flos": 36134932291200.0, + "grad_norm": 2.266405828893193, + "language_loss": 0.69338083, + "learning_rate": 3.7773571138356304e-06, + "loss": 0.71565378, + "num_input_tokens_seen": 31157780, + "step": 1473, + "time_per_iteration": 2.742100238800049 + }, + { + "auxiliary_loss_clip": 0.01112106, + "auxiliary_loss_mlp": 0.01029604, + "balance_loss_clip": 0.91053939, + "balance_loss_mlp": 1.01978707, + "epoch": 0.17723802080202009, + "flos": 22090593052800.0, + "grad_norm": 2.3696939480628485, + "language_loss": 0.88466322, + "learning_rate": 3.776999796601435e-06, + "loss": 0.90608025, + "num_input_tokens_seen": 31176540, + "step": 1474, + "time_per_iteration": 2.7724757194519043 + }, + { + "auxiliary_loss_clip": 0.012003, + "auxiliary_loss_mlp": 0.01033235, + "balance_loss_clip": 1.02645993, + "balance_loss_mlp": 1.02281618, + "epoch": 0.17735826369265917, + "flos": 30222671437440.0, + "grad_norm": 1.7630685360910392, + "language_loss": 0.72740483, + "learning_rate": 3.776642209797783e-06, + "loss": 0.74974018, + "num_input_tokens_seen": 31198370, + "step": 1475, + "time_per_iteration": 4.779360294342041 + }, + { + "auxiliary_loss_clip": 0.01193486, + "auxiliary_loss_mlp": 0.01031407, + "balance_loss_clip": 1.02784801, + "balance_loss_mlp": 1.02144122, + "epoch": 0.17747850658329825, + "flos": 21397588980480.0, + "grad_norm": 2.1344086966802713, + "language_loss": 0.78280401, + "learning_rate": 3.7762843534789205e-06, + "loss": 0.80505294, + "num_input_tokens_seen": 31217120, + "step": 1476, + "time_per_iteration": 2.7175755500793457 + }, + { + "auxiliary_loss_clip": 0.0117908, + "auxiliary_loss_mlp": 0.0103749, + "balance_loss_clip": 0.98871958, + "balance_loss_mlp": 1.02733374, + "epoch": 0.17759874947393736, + "flos": 16983341856000.0, + "grad_norm": 2.119347419243377, + "language_loss": 0.88351882, + "learning_rate": 3.7759262276991343e-06, + "loss": 0.90568447, + "num_input_tokens_seen": 31234730, + "step": 1477, + "time_per_iteration": 3.637773275375366 + }, + { + "auxiliary_loss_clip": 0.01179662, + "auxiliary_loss_mlp": 0.01035801, + "balance_loss_clip": 0.9915185, + "balance_loss_mlp": 1.02612758, + "epoch": 0.17771899236457644, + "flos": 11546107390080.0, + "grad_norm": 2.4300125250814437, + "language_loss": 0.80312574, + "learning_rate": 3.7755678325127506e-06, + "loss": 0.82528043, + "num_input_tokens_seen": 31252410, + "step": 1478, + "time_per_iteration": 2.641000270843506 + }, + { + "auxiliary_loss_clip": 0.01117862, + "auxiliary_loss_mlp": 0.01035135, + "balance_loss_clip": 0.91162509, + "balance_loss_mlp": 1.02509725, + "epoch": 0.17783923525521553, + "flos": 18807747494400.0, + "grad_norm": 1.8675941128741511, + "language_loss": 0.75465214, + "learning_rate": 3.7752091679741393e-06, + "loss": 0.77618217, + "num_input_tokens_seen": 31270200, + "step": 1479, + "time_per_iteration": 3.608520030975342 + }, + { + "auxiliary_loss_clip": 0.01197193, + "auxiliary_loss_mlp": 0.01036302, + "balance_loss_clip": 1.02893138, + "balance_loss_mlp": 1.02593684, + "epoch": 0.17795947814585464, + "flos": 30408365773440.0, + "grad_norm": 2.7097666077682057, + "language_loss": 0.76853526, + "learning_rate": 3.774850234137708e-06, + "loss": 0.79087025, + "num_input_tokens_seen": 31287495, + "step": 1480, + "time_per_iteration": 2.7503461837768555 + }, + { + "auxiliary_loss_clip": 0.01193102, + "auxiliary_loss_mlp": 0.01030125, + "balance_loss_clip": 1.02605295, + "balance_loss_mlp": 1.02056456, + "epoch": 0.17807972103649372, + "flos": 24389055411840.0, + "grad_norm": 2.490792991432836, + "language_loss": 0.82838058, + "learning_rate": 3.7744910310579076e-06, + "loss": 0.85061288, + "num_input_tokens_seen": 31306420, + "step": 1481, + "time_per_iteration": 2.6681649684906006 + }, + { + "auxiliary_loss_clip": 0.01221072, + "auxiliary_loss_mlp": 0.01034664, + "balance_loss_clip": 1.06790543, + "balance_loss_mlp": 1.02538979, + "epoch": 0.1781999639271328, + "flos": 20301559332480.0, + "grad_norm": 2.0683674089773527, + "language_loss": 0.85049558, + "learning_rate": 3.774131558789229e-06, + "loss": 0.87305295, + "num_input_tokens_seen": 31325750, + "step": 1482, + "time_per_iteration": 2.6894025802612305 + }, + { + "auxiliary_loss_clip": 0.01222848, + "auxiliary_loss_mlp": 0.00511964, + "balance_loss_clip": 1.06816864, + "balance_loss_mlp": 0.0, + "epoch": 0.1783202068177719, + "flos": 15924479806080.0, + "grad_norm": 2.4663204757195696, + "language_loss": 0.69573188, + "learning_rate": 3.773771817386203e-06, + "loss": 0.71307999, + "num_input_tokens_seen": 31343080, + "step": 1483, + "time_per_iteration": 2.535508394241333 + }, + { + "auxiliary_loss_clip": 0.01172215, + "auxiliary_loss_mlp": 0.01039191, + "balance_loss_clip": 0.98852706, + "balance_loss_mlp": 1.02916551, + "epoch": 0.178440449708411, + "flos": 20631758083200.0, + "grad_norm": 1.685630351733381, + "language_loss": 0.79038417, + "learning_rate": 3.773411806903403e-06, + "loss": 0.81249827, + "num_input_tokens_seen": 31362160, + "step": 1484, + "time_per_iteration": 2.7055513858795166 + }, + { + "auxiliary_loss_clip": 0.01104174, + "auxiliary_loss_mlp": 0.01041236, + "balance_loss_clip": 0.87274277, + "balance_loss_mlp": 1.03034663, + "epoch": 0.17856069259905008, + "flos": 21686059105920.0, + "grad_norm": 1.8596789792929096, + "language_loss": 0.94694686, + "learning_rate": 3.7730515273954415e-06, + "loss": 0.96840096, + "num_input_tokens_seen": 31380770, + "step": 1485, + "time_per_iteration": 2.7958521842956543 + }, + { + "auxiliary_loss_clip": 0.01218913, + "auxiliary_loss_mlp": 0.01032299, + "balance_loss_clip": 1.06637394, + "balance_loss_mlp": 1.02311397, + "epoch": 0.17868093548968916, + "flos": 26572962320640.0, + "grad_norm": 1.933377838237911, + "language_loss": 0.85251951, + "learning_rate": 3.772690978916973e-06, + "loss": 0.87503165, + "num_input_tokens_seen": 31400525, + "step": 1486, + "time_per_iteration": 2.7424721717834473 + }, + { + "auxiliary_loss_clip": 0.01193022, + "auxiliary_loss_mlp": 0.01032939, + "balance_loss_clip": 1.02517021, + "balance_loss_mlp": 1.02304459, + "epoch": 0.17880117838032827, + "flos": 18581006891520.0, + "grad_norm": 2.536193213101442, + "language_loss": 0.86265242, + "learning_rate": 3.772330161522693e-06, + "loss": 0.88491207, + "num_input_tokens_seen": 31418435, + "step": 1487, + "time_per_iteration": 2.6268203258514404 + }, + { + "auxiliary_loss_clip": 0.01172168, + "auxiliary_loss_mlp": 0.0103429, + "balance_loss_clip": 0.99218464, + "balance_loss_mlp": 1.02495635, + "epoch": 0.17892142127096736, + "flos": 26541217676160.0, + "grad_norm": 2.1589456707697283, + "language_loss": 0.79842448, + "learning_rate": 3.7719690752673365e-06, + "loss": 0.82048905, + "num_input_tokens_seen": 31439230, + "step": 1488, + "time_per_iteration": 2.719970703125 + }, + { + "auxiliary_loss_clip": 0.01151324, + "auxiliary_loss_mlp": 0.0103707, + "balance_loss_clip": 0.95258516, + "balance_loss_mlp": 1.02636516, + "epoch": 0.17904166416160644, + "flos": 23872623621120.0, + "grad_norm": 2.248722991361788, + "language_loss": 0.78068537, + "learning_rate": 3.7716077202056796e-06, + "loss": 0.80256927, + "num_input_tokens_seen": 31457705, + "step": 1489, + "time_per_iteration": 2.7485568523406982 + }, + { + "auxiliary_loss_clip": 0.01159092, + "auxiliary_loss_mlp": 0.01035745, + "balance_loss_clip": 0.98618233, + "balance_loss_mlp": 1.02594626, + "epoch": 0.17916190705224552, + "flos": 19134426712320.0, + "grad_norm": 2.461772956782775, + "language_loss": 0.9318397, + "learning_rate": 3.7712460963925404e-06, + "loss": 0.95378804, + "num_input_tokens_seen": 31473645, + "step": 1490, + "time_per_iteration": 2.6984663009643555 + }, + { + "auxiliary_loss_clip": 0.0116578, + "auxiliary_loss_mlp": 0.01037297, + "balance_loss_clip": 0.98585832, + "balance_loss_mlp": 1.02744484, + "epoch": 0.17928214994288463, + "flos": 25152120961920.0, + "grad_norm": 1.8276314640028486, + "language_loss": 0.75484657, + "learning_rate": 3.7708842038827775e-06, + "loss": 0.77687734, + "num_input_tokens_seen": 31492605, + "step": 1491, + "time_per_iteration": 2.7675600051879883 + }, + { + "auxiliary_loss_clip": 0.01195642, + "auxiliary_loss_mlp": 0.01033703, + "balance_loss_clip": 1.02747679, + "balance_loss_mlp": 1.02425539, + "epoch": 0.17940239283352372, + "flos": 22384629786240.0, + "grad_norm": 2.16325065778954, + "language_loss": 0.85728717, + "learning_rate": 3.770522042731288e-06, + "loss": 0.87958062, + "num_input_tokens_seen": 31514500, + "step": 1492, + "time_per_iteration": 2.6923012733459473 + }, + { + "auxiliary_loss_clip": 0.01115387, + "auxiliary_loss_mlp": 0.01033791, + "balance_loss_clip": 0.90906072, + "balance_loss_mlp": 1.02325261, + "epoch": 0.1795226357241628, + "flos": 23178685795200.0, + "grad_norm": 2.1052918007922106, + "language_loss": 0.87870681, + "learning_rate": 3.7701596129930122e-06, + "loss": 0.90019858, + "num_input_tokens_seen": 31533225, + "step": 1493, + "time_per_iteration": 2.8476688861846924 + }, + { + "auxiliary_loss_clip": 0.01156264, + "auxiliary_loss_mlp": 0.01030678, + "balance_loss_clip": 0.94975972, + "balance_loss_mlp": 1.0206821, + "epoch": 0.1796428786148019, + "flos": 22090413484800.0, + "grad_norm": 1.9194031087891208, + "language_loss": 0.73006546, + "learning_rate": 3.7697969147229315e-06, + "loss": 0.75193483, + "num_input_tokens_seen": 31551385, + "step": 1494, + "time_per_iteration": 2.747166633605957 + }, + { + "auxiliary_loss_clip": 0.0119749, + "auxiliary_loss_mlp": 0.01037683, + "balance_loss_clip": 1.02852094, + "balance_loss_mlp": 1.02777708, + "epoch": 0.179763121505441, + "flos": 21324618501120.0, + "grad_norm": 3.2840013222490128, + "language_loss": 0.85329819, + "learning_rate": 3.7694339479760647e-06, + "loss": 0.87564987, + "num_input_tokens_seen": 31570415, + "step": 1495, + "time_per_iteration": 2.6612837314605713 + }, + { + "auxiliary_loss_clip": 0.01094587, + "auxiliary_loss_mlp": 0.01002686, + "balance_loss_clip": 0.96619809, + "balance_loss_mlp": 0.99958628, + "epoch": 0.17988336439608008, + "flos": 68161864815360.0, + "grad_norm": 0.773737647887615, + "language_loss": 0.57287514, + "learning_rate": 3.769070712807476e-06, + "loss": 0.59384787, + "num_input_tokens_seen": 31632445, + "step": 1496, + "time_per_iteration": 3.2840523719787598 + }, + { + "auxiliary_loss_clip": 0.01086165, + "auxiliary_loss_mlp": 0.01041912, + "balance_loss_clip": 0.86979806, + "balance_loss_mlp": 1.03165996, + "epoch": 0.18000360728671919, + "flos": 21945047143680.0, + "grad_norm": 2.779659642607768, + "language_loss": 0.7881763, + "learning_rate": 3.768707209272266e-06, + "loss": 0.80945706, + "num_input_tokens_seen": 31652575, + "step": 1497, + "time_per_iteration": 2.8127260208129883 + }, + { + "auxiliary_loss_clip": 0.01165985, + "auxiliary_loss_mlp": 0.01034411, + "balance_loss_clip": 0.98661405, + "balance_loss_mlp": 1.02464175, + "epoch": 0.18012385017735827, + "flos": 18986330937600.0, + "grad_norm": 2.871063381299519, + "language_loss": 0.76080048, + "learning_rate": 3.768343437425579e-06, + "loss": 0.78280449, + "num_input_tokens_seen": 31671145, + "step": 1498, + "time_per_iteration": 2.7029151916503906 + }, + { + "auxiliary_loss_clip": 0.01063566, + "auxiliary_loss_mlp": 0.01031841, + "balance_loss_clip": 0.83055383, + "balance_loss_mlp": 1.02188659, + "epoch": 0.18024409306799735, + "flos": 19748103598080.0, + "grad_norm": 2.8224923552531402, + "language_loss": 0.86131811, + "learning_rate": 3.7679793973225987e-06, + "loss": 0.88227212, + "num_input_tokens_seen": 31686955, + "step": 1499, + "time_per_iteration": 2.8202733993530273 + }, + { + "auxiliary_loss_clip": 0.01044212, + "auxiliary_loss_mlp": 0.01005285, + "balance_loss_clip": 0.88757348, + "balance_loss_mlp": 1.00201833, + "epoch": 0.18036433595863643, + "flos": 67227183060480.0, + "grad_norm": 0.8500013118411499, + "language_loss": 0.61595678, + "learning_rate": 3.767615089018549e-06, + "loss": 0.63645172, + "num_input_tokens_seen": 31749300, + "step": 1500, + "time_per_iteration": 3.2818875312805176 + }, + { + "auxiliary_loss_clip": 0.0116358, + "auxiliary_loss_mlp": 0.01042519, + "balance_loss_clip": 0.98488724, + "balance_loss_mlp": 1.03168869, + "epoch": 0.18048457884927555, + "flos": 18181464935040.0, + "grad_norm": 10.039349868809825, + "language_loss": 0.85940164, + "learning_rate": 3.7672505125686966e-06, + "loss": 0.88146269, + "num_input_tokens_seen": 31765665, + "step": 1501, + "time_per_iteration": 4.485357761383057 + }, + { + "auxiliary_loss_clip": 0.01121828, + "auxiliary_loss_mlp": 0.01038111, + "balance_loss_clip": 0.90702438, + "balance_loss_mlp": 1.02774024, + "epoch": 0.18060482173991463, + "flos": 15813767111040.0, + "grad_norm": 3.3382292894967533, + "language_loss": 0.83561224, + "learning_rate": 3.7668856680283455e-06, + "loss": 0.85721165, + "num_input_tokens_seen": 31782690, + "step": 1502, + "time_per_iteration": 2.709150552749634 + }, + { + "auxiliary_loss_clip": 0.01179714, + "auxiliary_loss_mlp": 0.01042377, + "balance_loss_clip": 0.99199313, + "balance_loss_mlp": 1.03210676, + "epoch": 0.1807250646305537, + "flos": 18587399512320.0, + "grad_norm": 1.868560510569334, + "language_loss": 0.82197857, + "learning_rate": 3.7665205554528437e-06, + "loss": 0.84419954, + "num_input_tokens_seen": 31802045, + "step": 1503, + "time_per_iteration": 2.6945641040802 + }, + { + "auxiliary_loss_clip": 0.01175336, + "auxiliary_loss_mlp": 0.01035837, + "balance_loss_clip": 0.99138802, + "balance_loss_mlp": 1.02678931, + "epoch": 0.18084530752119282, + "flos": 23149131880320.0, + "grad_norm": 1.650855503454632, + "language_loss": 0.7428847, + "learning_rate": 3.7661551748975782e-06, + "loss": 0.76499641, + "num_input_tokens_seen": 31820220, + "step": 1504, + "time_per_iteration": 3.842325448989868 + }, + { + "auxiliary_loss_clip": 0.01091506, + "auxiliary_loss_mlp": 0.01008502, + "balance_loss_clip": 0.96076059, + "balance_loss_mlp": 1.00533092, + "epoch": 0.1809655504118319, + "flos": 59803153568640.0, + "grad_norm": 0.8155072844590708, + "language_loss": 0.60433072, + "learning_rate": 3.7657895264179772e-06, + "loss": 0.62533081, + "num_input_tokens_seen": 31876195, + "step": 1505, + "time_per_iteration": 3.277031421661377 + }, + { + "auxiliary_loss_clip": 0.01163046, + "auxiliary_loss_mlp": 0.01039499, + "balance_loss_clip": 0.98454601, + "balance_loss_mlp": 1.02934241, + "epoch": 0.181085793302471, + "flos": 44201941188480.0, + "grad_norm": 1.948504388755523, + "language_loss": 0.74289072, + "learning_rate": 3.765423610069509e-06, + "loss": 0.76491618, + "num_input_tokens_seen": 31901585, + "step": 1506, + "time_per_iteration": 3.8012468814849854 + }, + { + "auxiliary_loss_clip": 0.01171604, + "auxiliary_loss_mlp": 0.01033773, + "balance_loss_clip": 0.9893986, + "balance_loss_mlp": 1.02394998, + "epoch": 0.18120603619311007, + "flos": 34898384638080.0, + "grad_norm": 1.822276290061088, + "language_loss": 0.72508413, + "learning_rate": 3.765057425907683e-06, + "loss": 0.7471379, + "num_input_tokens_seen": 31923045, + "step": 1507, + "time_per_iteration": 2.774601697921753 + }, + { + "auxiliary_loss_clip": 0.01199291, + "auxiliary_loss_mlp": 0.01036207, + "balance_loss_clip": 1.02767324, + "balance_loss_mlp": 1.02600861, + "epoch": 0.18132627908374918, + "flos": 21506757390720.0, + "grad_norm": 2.0051400525032976, + "language_loss": 0.78182161, + "learning_rate": 3.764690973988048e-06, + "loss": 0.80417657, + "num_input_tokens_seen": 31943385, + "step": 1508, + "time_per_iteration": 2.6898655891418457 + }, + { + "auxiliary_loss_clip": 0.01150853, + "auxiliary_loss_mlp": 0.010386, + "balance_loss_clip": 0.95041013, + "balance_loss_mlp": 1.02900374, + "epoch": 0.18144652197438826, + "flos": 29057693633280.0, + "grad_norm": 2.061546997009404, + "language_loss": 0.74047667, + "learning_rate": 3.7643242543661967e-06, + "loss": 0.76237118, + "num_input_tokens_seen": 31966045, + "step": 1509, + "time_per_iteration": 2.8792331218719482 + }, + { + "auxiliary_loss_clip": 0.01081401, + "auxiliary_loss_mlp": 0.01003594, + "balance_loss_clip": 0.95710069, + "balance_loss_mlp": 1.00037527, + "epoch": 0.18156676486502735, + "flos": 68675064382080.0, + "grad_norm": 0.8260229892966889, + "language_loss": 0.60461926, + "learning_rate": 3.7639572670977573e-06, + "loss": 0.62546921, + "num_input_tokens_seen": 32021540, + "step": 1510, + "time_per_iteration": 3.140489101409912 + }, + { + "auxiliary_loss_clip": 0.01150391, + "auxiliary_loss_mlp": 0.01035365, + "balance_loss_clip": 0.94924104, + "balance_loss_mlp": 1.02504754, + "epoch": 0.18168700775566646, + "flos": 26471515334400.0, + "grad_norm": 4.473412030587013, + "language_loss": 0.76365221, + "learning_rate": 3.7635900122384042e-06, + "loss": 0.78550977, + "num_input_tokens_seen": 32044535, + "step": 1511, + "time_per_iteration": 2.8400516510009766 + }, + { + "auxiliary_loss_clip": 0.01171195, + "auxiliary_loss_mlp": 0.01038505, + "balance_loss_clip": 0.98615479, + "balance_loss_mlp": 1.02840829, + "epoch": 0.18180725064630554, + "flos": 15005668884480.0, + "grad_norm": 2.535290852566282, + "language_loss": 0.86850536, + "learning_rate": 3.7632224898438477e-06, + "loss": 0.89060235, + "num_input_tokens_seen": 32061010, + "step": 1512, + "time_per_iteration": 2.6565845012664795 + }, + { + "auxiliary_loss_clip": 0.01154922, + "auxiliary_loss_mlp": 0.01037571, + "balance_loss_clip": 0.94919491, + "balance_loss_mlp": 1.02790892, + "epoch": 0.18192749353694462, + "flos": 19682387665920.0, + "grad_norm": 1.9214297326866696, + "language_loss": 0.78947163, + "learning_rate": 3.762854699969842e-06, + "loss": 0.8113966, + "num_input_tokens_seen": 32081520, + "step": 1513, + "time_per_iteration": 2.7218174934387207 + }, + { + "auxiliary_loss_clip": 0.01192303, + "auxiliary_loss_mlp": 0.01031641, + "balance_loss_clip": 1.0264405, + "balance_loss_mlp": 1.02173448, + "epoch": 0.1820477364275837, + "flos": 20702717400960.0, + "grad_norm": 3.417403589772337, + "language_loss": 0.73089433, + "learning_rate": 3.762486642672179e-06, + "loss": 0.75313377, + "num_input_tokens_seen": 32098460, + "step": 1514, + "time_per_iteration": 2.704101800918579 + }, + { + "auxiliary_loss_clip": 0.01174061, + "auxiliary_loss_mlp": 0.01042263, + "balance_loss_clip": 0.98947239, + "balance_loss_mlp": 1.03171277, + "epoch": 0.18216797931822282, + "flos": 17128708197120.0, + "grad_norm": 2.4853346240720326, + "language_loss": 0.86823213, + "learning_rate": 3.7621183180066946e-06, + "loss": 0.89039534, + "num_input_tokens_seen": 32116420, + "step": 1515, + "time_per_iteration": 2.7244224548339844 + }, + { + "auxiliary_loss_clip": 0.01167569, + "auxiliary_loss_mlp": 0.01031263, + "balance_loss_clip": 0.98433185, + "balance_loss_mlp": 1.02082038, + "epoch": 0.1822882222088619, + "flos": 29242561956480.0, + "grad_norm": 1.6723509453108985, + "language_loss": 0.73930454, + "learning_rate": 3.7617497260292625e-06, + "loss": 0.76129293, + "num_input_tokens_seen": 32138475, + "step": 1516, + "time_per_iteration": 2.761775016784668 + }, + { + "auxiliary_loss_clip": 0.01163546, + "auxiliary_loss_mlp": 0.01035693, + "balance_loss_clip": 0.98709738, + "balance_loss_mlp": 1.02476752, + "epoch": 0.18240846509950098, + "flos": 17702739446400.0, + "grad_norm": 3.0946666888516963, + "language_loss": 0.78817081, + "learning_rate": 3.7613808667957967e-06, + "loss": 0.8101632, + "num_input_tokens_seen": 32151165, + "step": 1517, + "time_per_iteration": 2.7168214321136475 + }, + { + "auxiliary_loss_clip": 0.01173661, + "auxiliary_loss_mlp": 0.01040654, + "balance_loss_clip": 0.98923659, + "balance_loss_mlp": 1.03067064, + "epoch": 0.1825287079901401, + "flos": 14790025584000.0, + "grad_norm": 3.4780702852977377, + "language_loss": 0.9100697, + "learning_rate": 3.7610117403622547e-06, + "loss": 0.93221283, + "num_input_tokens_seen": 32167725, + "step": 1518, + "time_per_iteration": 2.7757728099823 + }, + { + "auxiliary_loss_clip": 0.01139364, + "auxiliary_loss_mlp": 0.01035435, + "balance_loss_clip": 0.94634384, + "balance_loss_mlp": 1.0246768, + "epoch": 0.18264895088077918, + "flos": 21946232292480.0, + "grad_norm": 2.423284402896812, + "language_loss": 0.90031004, + "learning_rate": 3.7606423467846313e-06, + "loss": 0.92205799, + "num_input_tokens_seen": 32187330, + "step": 1519, + "time_per_iteration": 2.724864959716797 + }, + { + "auxiliary_loss_clip": 0.0115193, + "auxiliary_loss_mlp": 0.01048262, + "balance_loss_clip": 0.95099121, + "balance_loss_mlp": 1.03812361, + "epoch": 0.18276919377141826, + "flos": 20886759711360.0, + "grad_norm": 1.6146913700864114, + "language_loss": 0.79500985, + "learning_rate": 3.760272686118964e-06, + "loss": 0.81701177, + "num_input_tokens_seen": 32205550, + "step": 1520, + "time_per_iteration": 2.7362515926361084 + }, + { + "auxiliary_loss_clip": 0.01173265, + "auxiliary_loss_mlp": 0.01030271, + "balance_loss_clip": 0.98775232, + "balance_loss_mlp": 1.0205915, + "epoch": 0.18288943666205737, + "flos": 21469877101440.0, + "grad_norm": 2.553570885517848, + "language_loss": 0.92454791, + "learning_rate": 3.7599027584213297e-06, + "loss": 0.94658327, + "num_input_tokens_seen": 32224430, + "step": 1521, + "time_per_iteration": 2.6781463623046875 + }, + { + "auxiliary_loss_clip": 0.01198593, + "auxiliary_loss_mlp": 0.01033187, + "balance_loss_clip": 1.02686882, + "balance_loss_mlp": 1.02310181, + "epoch": 0.18300967955269645, + "flos": 21539363961600.0, + "grad_norm": 2.100228005283073, + "language_loss": 0.78297615, + "learning_rate": 3.7595325637478465e-06, + "loss": 0.80529392, + "num_input_tokens_seen": 32242455, + "step": 1522, + "time_per_iteration": 2.672412633895874 + }, + { + "auxiliary_loss_clip": 0.01166135, + "auxiliary_loss_mlp": 0.0104277, + "balance_loss_clip": 0.99232846, + "balance_loss_mlp": 1.03304815, + "epoch": 0.18312992244333554, + "flos": 28876237102080.0, + "grad_norm": 2.3845690427375597, + "language_loss": 0.81376004, + "learning_rate": 3.7591621021546723e-06, + "loss": 0.83584911, + "num_input_tokens_seen": 32264450, + "step": 1523, + "time_per_iteration": 2.7298760414123535 + }, + { + "auxiliary_loss_clip": 0.01189631, + "auxiliary_loss_mlp": 0.01034375, + "balance_loss_clip": 1.02556539, + "balance_loss_mlp": 1.02395022, + "epoch": 0.18325016533397462, + "flos": 20120102801280.0, + "grad_norm": 3.4905542310823265, + "language_loss": 0.81672621, + "learning_rate": 3.7587913736980062e-06, + "loss": 0.83896619, + "num_input_tokens_seen": 32284090, + "step": 1524, + "time_per_iteration": 2.7102065086364746 + }, + { + "auxiliary_loss_clip": 0.01083496, + "auxiliary_loss_mlp": 0.01039578, + "balance_loss_clip": 0.86728454, + "balance_loss_mlp": 1.03043437, + "epoch": 0.18337040822461373, + "flos": 23329187781120.0, + "grad_norm": 1.8567911083022857, + "language_loss": 0.84519875, + "learning_rate": 3.7584203784340865e-06, + "loss": 0.86642945, + "num_input_tokens_seen": 32303260, + "step": 1525, + "time_per_iteration": 2.7837936878204346 + }, + { + "auxiliary_loss_clip": 0.01165762, + "auxiliary_loss_mlp": 0.01035527, + "balance_loss_clip": 0.98449904, + "balance_loss_mlp": 1.02598381, + "epoch": 0.1834906511152528, + "flos": 25009555881600.0, + "grad_norm": 2.0107696302225158, + "language_loss": 0.85381776, + "learning_rate": 3.7580491164191938e-06, + "loss": 0.87583059, + "num_input_tokens_seen": 32321570, + "step": 1526, + "time_per_iteration": 3.6604726314544678 + }, + { + "auxiliary_loss_clip": 0.01114345, + "auxiliary_loss_mlp": 0.01002149, + "balance_loss_clip": 1.00618029, + "balance_loss_mlp": 0.99857277, + "epoch": 0.1836108940058919, + "flos": 67251493589760.0, + "grad_norm": 0.7447270473594826, + "language_loss": 0.61267596, + "learning_rate": 3.757677587709648e-06, + "loss": 0.63384092, + "num_input_tokens_seen": 32384835, + "step": 1527, + "time_per_iteration": 4.292892932891846 + }, + { + "auxiliary_loss_clip": 0.01142516, + "auxiliary_loss_mlp": 0.01041641, + "balance_loss_clip": 0.95142472, + "balance_loss_mlp": 1.03147888, + "epoch": 0.183731136896531, + "flos": 25738721971200.0, + "grad_norm": 2.294386590987908, + "language_loss": 0.7558139, + "learning_rate": 3.7573057923618095e-06, + "loss": 0.77765548, + "num_input_tokens_seen": 32404930, + "step": 1528, + "time_per_iteration": 2.7493035793304443 + }, + { + "auxiliary_loss_clip": 0.01118797, + "auxiliary_loss_mlp": 0.01045645, + "balance_loss_clip": 0.90671998, + "balance_loss_mlp": 1.03520823, + "epoch": 0.1838513797871701, + "flos": 20449403712000.0, + "grad_norm": 1.764032420420391, + "language_loss": 0.74014008, + "learning_rate": 3.7569337304320793e-06, + "loss": 0.76178449, + "num_input_tokens_seen": 32424515, + "step": 1529, + "time_per_iteration": 3.6520986557006836 + }, + { + "auxiliary_loss_clip": 0.01089081, + "auxiliary_loss_mlp": 0.010047, + "balance_loss_clip": 0.9647094, + "balance_loss_mlp": 1.0013144, + "epoch": 0.18397162267780917, + "flos": 68565141786240.0, + "grad_norm": 0.8391289040063002, + "language_loss": 0.6442824, + "learning_rate": 3.756561401976899e-06, + "loss": 0.66522026, + "num_input_tokens_seen": 32484220, + "step": 1530, + "time_per_iteration": 3.1556625366210938 + }, + { + "auxiliary_loss_clip": 0.0122383, + "auxiliary_loss_mlp": 0.0102979, + "balance_loss_clip": 1.06916213, + "balance_loss_mlp": 1.02025354, + "epoch": 0.18409186556844825, + "flos": 31941104976000.0, + "grad_norm": 2.0050270973006983, + "language_loss": 0.82225609, + "learning_rate": 3.7561888070527514e-06, + "loss": 0.84479225, + "num_input_tokens_seen": 32506260, + "step": 1531, + "time_per_iteration": 3.6664693355560303 + }, + { + "auxiliary_loss_clip": 0.01135918, + "auxiliary_loss_mlp": 0.0051234, + "balance_loss_clip": 0.94870478, + "balance_loss_mlp": 0.0, + "epoch": 0.18421210845908736, + "flos": 20120533764480.0, + "grad_norm": 2.2883346484595664, + "language_loss": 0.8047753, + "learning_rate": 3.7558159457161577e-06, + "loss": 0.82125789, + "num_input_tokens_seen": 32524225, + "step": 1532, + "time_per_iteration": 2.694347620010376 + }, + { + "auxiliary_loss_clip": 0.01175813, + "auxiliary_loss_mlp": 0.00511946, + "balance_loss_clip": 0.99122328, + "balance_loss_mlp": 0.0, + "epoch": 0.18433235134972645, + "flos": 23110491824640.0, + "grad_norm": 2.6316839023304577, + "language_loss": 0.77833283, + "learning_rate": 3.755442818023681e-06, + "loss": 0.79521048, + "num_input_tokens_seen": 32543850, + "step": 1533, + "time_per_iteration": 2.7365145683288574 + }, + { + "auxiliary_loss_clip": 0.01150052, + "auxiliary_loss_mlp": 0.01033519, + "balance_loss_clip": 0.9502781, + "balance_loss_mlp": 1.02339768, + "epoch": 0.18445259424036553, + "flos": 18291351617280.0, + "grad_norm": 1.932783225536802, + "language_loss": 0.75531411, + "learning_rate": 3.7550694240319246e-06, + "loss": 0.77714986, + "num_input_tokens_seen": 32561725, + "step": 1534, + "time_per_iteration": 2.684121608734131 + }, + { + "auxiliary_loss_clip": 0.01197071, + "auxiliary_loss_mlp": 0.0103322, + "balance_loss_clip": 1.02424002, + "balance_loss_mlp": 1.02274776, + "epoch": 0.18457283713100464, + "flos": 21324079797120.0, + "grad_norm": 3.648211302831678, + "language_loss": 0.75919574, + "learning_rate": 3.7546957637975326e-06, + "loss": 0.78149867, + "num_input_tokens_seen": 32579135, + "step": 1535, + "time_per_iteration": 2.7238898277282715 + }, + { + "auxiliary_loss_clip": 0.01089437, + "auxiliary_loss_mlp": 0.01040537, + "balance_loss_clip": 0.86551654, + "balance_loss_mlp": 1.03062451, + "epoch": 0.18469308002164372, + "flos": 20375679047040.0, + "grad_norm": 1.5387516975072355, + "language_loss": 0.74013072, + "learning_rate": 3.7543218373771873e-06, + "loss": 0.7614305, + "num_input_tokens_seen": 32598460, + "step": 1536, + "time_per_iteration": 2.787712335586548 + }, + { + "auxiliary_loss_clip": 0.01088778, + "auxiliary_loss_mlp": 0.00512209, + "balance_loss_clip": 0.86912358, + "balance_loss_mlp": 0.0, + "epoch": 0.1848133229122828, + "flos": 26435892021120.0, + "grad_norm": 1.8885612419014726, + "language_loss": 0.7801016, + "learning_rate": 3.753947644827615e-06, + "loss": 0.79611146, + "num_input_tokens_seen": 32621920, + "step": 1537, + "time_per_iteration": 2.9375336170196533 + }, + { + "auxiliary_loss_clip": 0.01085902, + "auxiliary_loss_mlp": 0.01014902, + "balance_loss_clip": 0.96375781, + "balance_loss_mlp": 1.0116353, + "epoch": 0.1849335658029219, + "flos": 70547447612160.0, + "grad_norm": 0.9435210997128533, + "language_loss": 0.5716573, + "learning_rate": 3.753573186205579e-06, + "loss": 0.59266531, + "num_input_tokens_seen": 32690040, + "step": 1538, + "time_per_iteration": 3.5220179557800293 + }, + { + "auxiliary_loss_clip": 0.01160888, + "auxiliary_loss_mlp": 0.00512155, + "balance_loss_clip": 0.98466003, + "balance_loss_mlp": 0.0, + "epoch": 0.185053808693561, + "flos": 17384140788480.0, + "grad_norm": 7.105206529678941, + "language_loss": 0.77779847, + "learning_rate": 3.753198461567885e-06, + "loss": 0.7945289, + "num_input_tokens_seen": 32707285, + "step": 1539, + "time_per_iteration": 2.6644287109375 + }, + { + "auxiliary_loss_clip": 0.0113922, + "auxiliary_loss_mlp": 0.01031143, + "balance_loss_clip": 0.95028573, + "balance_loss_mlp": 1.02224422, + "epoch": 0.18517405158420008, + "flos": 28986159697920.0, + "grad_norm": 1.879114261626939, + "language_loss": 0.92015445, + "learning_rate": 3.7528234709713783e-06, + "loss": 0.94185811, + "num_input_tokens_seen": 32730030, + "step": 1540, + "time_per_iteration": 2.795344829559326 + }, + { + "auxiliary_loss_clip": 0.01199022, + "auxiliary_loss_mlp": 0.01028792, + "balance_loss_clip": 1.02839744, + "balance_loss_mlp": 1.01920748, + "epoch": 0.18529429447483917, + "flos": 26794962328320.0, + "grad_norm": 1.8760792635420207, + "language_loss": 0.8466866, + "learning_rate": 3.7524482144729447e-06, + "loss": 0.86896479, + "num_input_tokens_seen": 32749485, + "step": 1541, + "time_per_iteration": 2.6681463718414307 + }, + { + "auxiliary_loss_clip": 0.01137954, + "auxiliary_loss_mlp": 0.01039287, + "balance_loss_clip": 0.94443452, + "balance_loss_mlp": 1.02919006, + "epoch": 0.18541453736547828, + "flos": 13581595301760.0, + "grad_norm": 2.089597018459927, + "language_loss": 0.83441079, + "learning_rate": 3.7520726921295106e-06, + "loss": 0.85618317, + "num_input_tokens_seen": 32766205, + "step": 1542, + "time_per_iteration": 2.761470079421997 + }, + { + "auxiliary_loss_clip": 0.01188353, + "auxiliary_loss_mlp": 0.01038026, + "balance_loss_clip": 1.02329123, + "balance_loss_mlp": 1.02835226, + "epoch": 0.18553478025611736, + "flos": 24025424077440.0, + "grad_norm": 2.079696208088684, + "language_loss": 0.72303665, + "learning_rate": 3.751696903998042e-06, + "loss": 0.74530041, + "num_input_tokens_seen": 32784840, + "step": 1543, + "time_per_iteration": 2.699788808822632 + }, + { + "auxiliary_loss_clip": 0.01191912, + "auxiliary_loss_mlp": 0.01044072, + "balance_loss_clip": 1.02785754, + "balance_loss_mlp": 1.03439236, + "epoch": 0.18565502314675644, + "flos": 25885165720320.0, + "grad_norm": 1.6620297419073469, + "language_loss": 0.70190209, + "learning_rate": 3.7513208501355456e-06, + "loss": 0.724262, + "num_input_tokens_seen": 32805945, + "step": 1544, + "time_per_iteration": 2.731259822845459 + }, + { + "auxiliary_loss_clip": 0.01165154, + "auxiliary_loss_mlp": 0.0103456, + "balance_loss_clip": 0.98579806, + "balance_loss_mlp": 1.02493966, + "epoch": 0.18577526603739553, + "flos": 19610063631360.0, + "grad_norm": 1.8851928968419236, + "language_loss": 0.83416402, + "learning_rate": 3.750944530599069e-06, + "loss": 0.85616112, + "num_input_tokens_seen": 32825515, + "step": 1545, + "time_per_iteration": 2.725382089614868 + }, + { + "auxiliary_loss_clip": 0.0120141, + "auxiliary_loss_mlp": 0.01036743, + "balance_loss_clip": 1.02828622, + "balance_loss_mlp": 1.02714062, + "epoch": 0.18589550892803464, + "flos": 18474891137280.0, + "grad_norm": 1.913118918611557, + "language_loss": 0.80357134, + "learning_rate": 3.7505679454456992e-06, + "loss": 0.82595295, + "num_input_tokens_seen": 32842125, + "step": 1546, + "time_per_iteration": 2.6673333644866943 + }, + { + "auxiliary_loss_clip": 0.01068537, + "auxiliary_loss_mlp": 0.01035457, + "balance_loss_clip": 0.8306154, + "balance_loss_mlp": 1.02513909, + "epoch": 0.18601575181867372, + "flos": 23549966726400.0, + "grad_norm": 2.2811081265223327, + "language_loss": 0.70393652, + "learning_rate": 3.750191094732564e-06, + "loss": 0.72497642, + "num_input_tokens_seen": 32862990, + "step": 1547, + "time_per_iteration": 2.9324586391448975 + }, + { + "auxiliary_loss_clip": 0.01066076, + "auxiliary_loss_mlp": 0.00512066, + "balance_loss_clip": 0.8275485, + "balance_loss_mlp": 0.0, + "epoch": 0.1861359947093128, + "flos": 26360192108160.0, + "grad_norm": 2.2771089260153894, + "language_loss": 0.75311887, + "learning_rate": 3.7498139785168313e-06, + "loss": 0.76890028, + "num_input_tokens_seen": 32883595, + "step": 1548, + "time_per_iteration": 2.8702504634857178 + }, + { + "auxiliary_loss_clip": 0.01190833, + "auxiliary_loss_mlp": 0.01032355, + "balance_loss_clip": 1.02505493, + "balance_loss_mlp": 1.02331257, + "epoch": 0.1862562375999519, + "flos": 23331198942720.0, + "grad_norm": 1.7226679138223389, + "language_loss": 0.77169555, + "learning_rate": 3.749436596855709e-06, + "loss": 0.79392743, + "num_input_tokens_seen": 32902895, + "step": 1549, + "time_per_iteration": 2.692446708679199 + }, + { + "auxiliary_loss_clip": 0.01186439, + "auxiliary_loss_mlp": 0.01037025, + "balance_loss_clip": 1.02255774, + "balance_loss_mlp": 1.02716088, + "epoch": 0.186376480490591, + "flos": 16648222942080.0, + "grad_norm": 2.5849282583583, + "language_loss": 0.90048981, + "learning_rate": 3.749058949806446e-06, + "loss": 0.92272449, + "num_input_tokens_seen": 32919620, + "step": 1550, + "time_per_iteration": 2.6214733123779297 + }, + { + "auxiliary_loss_clip": 0.01191571, + "auxiliary_loss_mlp": 0.01031612, + "balance_loss_clip": 1.02459908, + "balance_loss_mlp": 1.02186084, + "epoch": 0.18649672338123008, + "flos": 21468656039040.0, + "grad_norm": 2.0596977741527693, + "language_loss": 0.84124857, + "learning_rate": 3.748681037426331e-06, + "loss": 0.86348033, + "num_input_tokens_seen": 32938830, + "step": 1551, + "time_per_iteration": 2.6980862617492676 + }, + { + "auxiliary_loss_clip": 0.01220275, + "auxiliary_loss_mlp": 0.01035522, + "balance_loss_clip": 1.06699061, + "balance_loss_mlp": 1.02557349, + "epoch": 0.1866169662718692, + "flos": 12312728386560.0, + "grad_norm": 2.3523927281049466, + "language_loss": 0.91790855, + "learning_rate": 3.7483028597726936e-06, + "loss": 0.94046652, + "num_input_tokens_seen": 32955600, + "step": 1552, + "time_per_iteration": 3.507913112640381 + }, + { + "auxiliary_loss_clip": 0.01147041, + "auxiliary_loss_mlp": 0.01045433, + "balance_loss_clip": 0.95090604, + "balance_loss_mlp": 1.03478098, + "epoch": 0.18673720916250827, + "flos": 23581280407680.0, + "grad_norm": 2.963526339327097, + "language_loss": 0.62522948, + "learning_rate": 3.7479244169029017e-06, + "loss": 0.64715421, + "num_input_tokens_seen": 32975390, + "step": 1553, + "time_per_iteration": 3.7516138553619385 + }, + { + "auxiliary_loss_clip": 0.01192302, + "auxiliary_loss_mlp": 0.01036112, + "balance_loss_clip": 1.02091956, + "balance_loss_mlp": 1.0267657, + "epoch": 0.18685745205314735, + "flos": 19718370115200.0, + "grad_norm": 2.2538109117596417, + "language_loss": 0.7359916, + "learning_rate": 3.7475457088743658e-06, + "loss": 0.75827575, + "num_input_tokens_seen": 32992640, + "step": 1554, + "time_per_iteration": 2.6956534385681152 + }, + { + "auxiliary_loss_clip": 0.01160524, + "auxiliary_loss_mlp": 0.01034274, + "balance_loss_clip": 0.98623216, + "balance_loss_mlp": 1.02451634, + "epoch": 0.18697769494378644, + "flos": 34204123589760.0, + "grad_norm": 2.3284163497451225, + "language_loss": 0.74601316, + "learning_rate": 3.7471667357445348e-06, + "loss": 0.76796114, + "num_input_tokens_seen": 33012470, + "step": 1555, + "time_per_iteration": 2.877917528152466 + }, + { + "auxiliary_loss_clip": 0.01097041, + "auxiliary_loss_mlp": 0.01038512, + "balance_loss_clip": 0.87061793, + "balance_loss_mlp": 1.02867079, + "epoch": 0.18709793783442555, + "flos": 34241327101440.0, + "grad_norm": 1.7922589356532297, + "language_loss": 0.72355986, + "learning_rate": 3.7467874975709e-06, + "loss": 0.74491543, + "num_input_tokens_seen": 33033275, + "step": 1556, + "time_per_iteration": 4.7150719165802 + }, + { + "auxiliary_loss_clip": 0.01198678, + "auxiliary_loss_mlp": 0.01035006, + "balance_loss_clip": 1.02778232, + "balance_loss_mlp": 1.02543306, + "epoch": 0.18721818072506463, + "flos": 40734550529280.0, + "grad_norm": 3.2016466813417783, + "language_loss": 0.78184187, + "learning_rate": 3.7464079944109904e-06, + "loss": 0.80417871, + "num_input_tokens_seen": 33055135, + "step": 1557, + "time_per_iteration": 2.876816749572754 + }, + { + "auxiliary_loss_clip": 0.01145166, + "auxiliary_loss_mlp": 0.01027878, + "balance_loss_clip": 0.94608068, + "balance_loss_mlp": 1.01897299, + "epoch": 0.18733842361570371, + "flos": 22157386392960.0, + "grad_norm": 1.9169640045834742, + "language_loss": 0.77490735, + "learning_rate": 3.746028226322376e-06, + "loss": 0.79663777, + "num_input_tokens_seen": 33071015, + "step": 1558, + "time_per_iteration": 2.665660858154297 + }, + { + "auxiliary_loss_clip": 0.0116612, + "auxiliary_loss_mlp": 0.01033796, + "balance_loss_clip": 0.9869163, + "balance_loss_mlp": 1.02447379, + "epoch": 0.18745866650634282, + "flos": 18914940656640.0, + "grad_norm": 2.2619639901753357, + "language_loss": 0.75442362, + "learning_rate": 3.745648193362669e-06, + "loss": 0.7764228, + "num_input_tokens_seen": 33090370, + "step": 1559, + "time_per_iteration": 2.7009801864624023 + }, + { + "auxiliary_loss_clip": 0.01169606, + "auxiliary_loss_mlp": 0.01034279, + "balance_loss_clip": 0.98690891, + "balance_loss_mlp": 1.02523112, + "epoch": 0.1875789093969819, + "flos": 19314626267520.0, + "grad_norm": 2.6108934237801042, + "language_loss": 0.72359729, + "learning_rate": 3.745267895589518e-06, + "loss": 0.74563617, + "num_input_tokens_seen": 33108910, + "step": 1560, + "time_per_iteration": 2.6667239665985107 + }, + { + "auxiliary_loss_clip": 0.01172017, + "auxiliary_loss_mlp": 0.01031587, + "balance_loss_clip": 0.98894811, + "balance_loss_mlp": 1.02164519, + "epoch": 0.187699152287621, + "flos": 17018965169280.0, + "grad_norm": 1.7640960571126127, + "language_loss": 0.82017291, + "learning_rate": 3.7448873330606154e-06, + "loss": 0.84220886, + "num_input_tokens_seen": 33126680, + "step": 1561, + "time_per_iteration": 2.691485643386841 + }, + { + "auxiliary_loss_clip": 0.01136304, + "auxiliary_loss_mlp": 0.01037296, + "balance_loss_clip": 0.94749695, + "balance_loss_mlp": 1.02693701, + "epoch": 0.18781939517826007, + "flos": 22346384780160.0, + "grad_norm": 2.0778608355886727, + "language_loss": 0.87262905, + "learning_rate": 3.7445065058336914e-06, + "loss": 0.89436507, + "num_input_tokens_seen": 33145550, + "step": 1562, + "time_per_iteration": 2.7206289768218994 + }, + { + "auxiliary_loss_clip": 0.01105126, + "auxiliary_loss_mlp": 0.01035643, + "balance_loss_clip": 0.9034301, + "balance_loss_mlp": 1.02581978, + "epoch": 0.18793963806889918, + "flos": 14611478054400.0, + "grad_norm": 1.9431221393146683, + "language_loss": 0.8621521, + "learning_rate": 3.7441254139665176e-06, + "loss": 0.88355976, + "num_input_tokens_seen": 33161735, + "step": 1563, + "time_per_iteration": 2.8193788528442383 + }, + { + "auxiliary_loss_clip": 0.01221136, + "auxiliary_loss_mlp": 0.01034714, + "balance_loss_clip": 1.06914639, + "balance_loss_mlp": 1.02538598, + "epoch": 0.18805988095953827, + "flos": 17457075354240.0, + "grad_norm": 1.9103821786069457, + "language_loss": 0.82664967, + "learning_rate": 3.743744057516905e-06, + "loss": 0.84920824, + "num_input_tokens_seen": 33179795, + "step": 1564, + "time_per_iteration": 2.771101951599121 + }, + { + "auxiliary_loss_clip": 0.01120012, + "auxiliary_loss_mlp": 0.01035291, + "balance_loss_clip": 0.90974391, + "balance_loss_mlp": 1.02561152, + "epoch": 0.18818012385017735, + "flos": 15043877976960.0, + "grad_norm": 3.604640634408108, + "language_loss": 0.87400877, + "learning_rate": 3.743362436542706e-06, + "loss": 0.89556187, + "num_input_tokens_seen": 33194485, + "step": 1565, + "time_per_iteration": 2.686779260635376 + }, + { + "auxiliary_loss_clip": 0.01216513, + "auxiliary_loss_mlp": 0.01031841, + "balance_loss_clip": 1.06364655, + "balance_loss_mlp": 1.02231026, + "epoch": 0.18830036674081646, + "flos": 47551975136640.0, + "grad_norm": 2.3463774166835383, + "language_loss": 0.76808202, + "learning_rate": 3.7429805511018115e-06, + "loss": 0.79056561, + "num_input_tokens_seen": 33216145, + "step": 1566, + "time_per_iteration": 2.855513095855713 + }, + { + "auxiliary_loss_clip": 0.01141336, + "auxiliary_loss_mlp": 0.0051245, + "balance_loss_clip": 0.94776702, + "balance_loss_mlp": 0.0, + "epoch": 0.18842060963145554, + "flos": 30044626698240.0, + "grad_norm": 6.411869483935881, + "language_loss": 0.78018558, + "learning_rate": 3.7425984012521524e-06, + "loss": 0.79672348, + "num_input_tokens_seen": 33236345, + "step": 1567, + "time_per_iteration": 2.785634994506836 + }, + { + "auxiliary_loss_clip": 0.01060337, + "auxiliary_loss_mlp": 0.00504606, + "balance_loss_clip": 0.92813408, + "balance_loss_mlp": 0.0, + "epoch": 0.18854085252209463, + "flos": 70318372625280.0, + "grad_norm": 0.7793642497106883, + "language_loss": 0.60420376, + "learning_rate": 3.7422159870517025e-06, + "loss": 0.6198532, + "num_input_tokens_seen": 33301600, + "step": 1568, + "time_per_iteration": 3.3157870769500732 + }, + { + "auxiliary_loss_clip": 0.01167155, + "auxiliary_loss_mlp": 0.01029766, + "balance_loss_clip": 0.98789275, + "balance_loss_mlp": 1.01986575, + "epoch": 0.1886610954127337, + "flos": 21289318410240.0, + "grad_norm": 1.79772175316839, + "language_loss": 0.78880429, + "learning_rate": 3.7418333085584717e-06, + "loss": 0.81077349, + "num_input_tokens_seen": 33322785, + "step": 1569, + "time_per_iteration": 2.7222132682800293 + }, + { + "auxiliary_loss_clip": 0.01146694, + "auxiliary_loss_mlp": 0.01034821, + "balance_loss_clip": 0.95015323, + "balance_loss_mlp": 1.02588654, + "epoch": 0.18878133830337282, + "flos": 17266819991040.0, + "grad_norm": 2.098652227267604, + "language_loss": 0.90953875, + "learning_rate": 3.7414503658305128e-06, + "loss": 0.93135393, + "num_input_tokens_seen": 33340020, + "step": 1570, + "time_per_iteration": 2.6909472942352295 + }, + { + "auxiliary_loss_clip": 0.01133289, + "auxiliary_loss_mlp": 0.01035703, + "balance_loss_clip": 0.91177392, + "balance_loss_mlp": 1.02533746, + "epoch": 0.1889015811940119, + "flos": 25775207210880.0, + "grad_norm": 2.188135775596426, + "language_loss": 0.77052891, + "learning_rate": 3.7410671589259185e-06, + "loss": 0.7922188, + "num_input_tokens_seen": 33358620, + "step": 1571, + "time_per_iteration": 2.8360228538513184 + }, + { + "auxiliary_loss_clip": 0.0122219, + "auxiliary_loss_mlp": 0.01036073, + "balance_loss_clip": 1.06860328, + "balance_loss_mlp": 1.0264709, + "epoch": 0.18902182408465099, + "flos": 21032197879680.0, + "grad_norm": 3.920586532103234, + "language_loss": 0.79431921, + "learning_rate": 3.7406836879028205e-06, + "loss": 0.8169018, + "num_input_tokens_seen": 33378845, + "step": 1572, + "time_per_iteration": 2.7172818183898926 + }, + { + "auxiliary_loss_clip": 0.01195597, + "auxiliary_loss_mlp": 0.01035856, + "balance_loss_clip": 1.02816224, + "balance_loss_mlp": 1.02594364, + "epoch": 0.1891420669752901, + "flos": 22272121411200.0, + "grad_norm": 2.383201948656084, + "language_loss": 0.76828307, + "learning_rate": 3.7402999528193907e-06, + "loss": 0.79059762, + "num_input_tokens_seen": 33398345, + "step": 1573, + "time_per_iteration": 2.771998405456543 + }, + { + "auxiliary_loss_clip": 0.01133407, + "auxiliary_loss_mlp": 0.00512377, + "balance_loss_clip": 0.94670403, + "balance_loss_mlp": 0.0, + "epoch": 0.18926230986592918, + "flos": 22017802141440.0, + "grad_norm": 2.5813408108947895, + "language_loss": 0.85759425, + "learning_rate": 3.739915953733842e-06, + "loss": 0.87405211, + "num_input_tokens_seen": 33416390, + "step": 1574, + "time_per_iteration": 2.741605281829834 + }, + { + "auxiliary_loss_clip": 0.01216731, + "auxiliary_loss_mlp": 0.01033956, + "balance_loss_clip": 1.06482875, + "balance_loss_mlp": 1.02447307, + "epoch": 0.18938255275656826, + "flos": 24462672336000.0, + "grad_norm": 1.6413972099548642, + "language_loss": 0.81920105, + "learning_rate": 3.7395316907044264e-06, + "loss": 0.84170794, + "num_input_tokens_seen": 33437175, + "step": 1575, + "time_per_iteration": 2.685882568359375 + }, + { + "auxiliary_loss_clip": 0.01192814, + "auxiliary_loss_mlp": 0.01032279, + "balance_loss_clip": 1.02464771, + "balance_loss_mlp": 1.02248573, + "epoch": 0.18950279564720737, + "flos": 24427049022720.0, + "grad_norm": 1.6342331968348554, + "language_loss": 0.79778337, + "learning_rate": 3.7391471637894364e-06, + "loss": 0.82003433, + "num_input_tokens_seen": 33459440, + "step": 1576, + "time_per_iteration": 2.7198264598846436 + }, + { + "auxiliary_loss_clip": 0.01147711, + "auxiliary_loss_mlp": 0.0103469, + "balance_loss_clip": 0.94669092, + "balance_loss_mlp": 1.02396131, + "epoch": 0.18962303853784646, + "flos": 19756291898880.0, + "grad_norm": 1.781792045247473, + "language_loss": 0.84638298, + "learning_rate": 3.738762373047205e-06, + "loss": 0.86820692, + "num_input_tokens_seen": 33479360, + "step": 1577, + "time_per_iteration": 2.7712180614471436 + }, + { + "auxiliary_loss_clip": 0.01151149, + "auxiliary_loss_mlp": 0.01038314, + "balance_loss_clip": 0.94978446, + "balance_loss_mlp": 1.02861619, + "epoch": 0.18974328142848554, + "flos": 21032054225280.0, + "grad_norm": 1.6507226421718182, + "language_loss": 0.83788431, + "learning_rate": 3.738377318536103e-06, + "loss": 0.85977894, + "num_input_tokens_seen": 33499245, + "step": 1578, + "time_per_iteration": 3.7181193828582764 + }, + { + "auxiliary_loss_clip": 0.01215378, + "auxiliary_loss_mlp": 0.01038196, + "balance_loss_clip": 1.06574678, + "balance_loss_mlp": 1.02985179, + "epoch": 0.18986352431912462, + "flos": 12966122736000.0, + "grad_norm": 2.6092803885871, + "language_loss": 0.71014273, + "learning_rate": 3.7379920003145447e-06, + "loss": 0.73267841, + "num_input_tokens_seen": 33513520, + "step": 1579, + "time_per_iteration": 3.6009223461151123 + }, + { + "auxiliary_loss_clip": 0.01165619, + "auxiliary_loss_mlp": 0.01033881, + "balance_loss_clip": 0.9892, + "balance_loss_mlp": 1.02340269, + "epoch": 0.18998376720976373, + "flos": 23767908497280.0, + "grad_norm": 1.6714764774210433, + "language_loss": 0.83822834, + "learning_rate": 3.7376064184409817e-06, + "loss": 0.86022329, + "num_input_tokens_seen": 33533100, + "step": 1580, + "time_per_iteration": 2.7759077548980713 + }, + { + "auxiliary_loss_clip": 0.0116562, + "auxiliary_loss_mlp": 0.01039074, + "balance_loss_clip": 0.98827982, + "balance_loss_mlp": 1.02936423, + "epoch": 0.19010401010040281, + "flos": 22966023323520.0, + "grad_norm": 2.883547149984686, + "language_loss": 0.86885321, + "learning_rate": 3.7372205729739063e-06, + "loss": 0.89090008, + "num_input_tokens_seen": 33554915, + "step": 1581, + "time_per_iteration": 2.718477249145508 + }, + { + "auxiliary_loss_clip": 0.01196315, + "auxiliary_loss_mlp": 0.01044377, + "balance_loss_clip": 1.02586663, + "balance_loss_mlp": 1.03457189, + "epoch": 0.1902242529910419, + "flos": 19135647774720.0, + "grad_norm": 2.2298597057324896, + "language_loss": 0.72026336, + "learning_rate": 3.7368344639718514e-06, + "loss": 0.74267024, + "num_input_tokens_seen": 33572850, + "step": 1582, + "time_per_iteration": 4.446239948272705 + }, + { + "auxiliary_loss_clip": 0.01194574, + "auxiliary_loss_mlp": 0.01035827, + "balance_loss_clip": 1.0240804, + "balance_loss_mlp": 1.02648735, + "epoch": 0.190344495881681, + "flos": 25483935824640.0, + "grad_norm": 1.586658022208204, + "language_loss": 0.80587065, + "learning_rate": 3.7364480914933895e-06, + "loss": 0.82817471, + "num_input_tokens_seen": 33593090, + "step": 1583, + "time_per_iteration": 2.6910579204559326 + }, + { + "auxiliary_loss_clip": 0.01114424, + "auxiliary_loss_mlp": 0.00512507, + "balance_loss_clip": 0.90647733, + "balance_loss_mlp": 0.0, + "epoch": 0.1904647387723201, + "flos": 26792843425920.0, + "grad_norm": 2.1211718461940654, + "language_loss": 0.80939186, + "learning_rate": 3.7360614555971325e-06, + "loss": 0.82566112, + "num_input_tokens_seen": 33612745, + "step": 1584, + "time_per_iteration": 2.7452781200408936 + }, + { + "auxiliary_loss_clip": 0.01192515, + "auxiliary_loss_mlp": 0.0051248, + "balance_loss_clip": 1.02724457, + "balance_loss_mlp": 0.0, + "epoch": 0.19058498166295917, + "flos": 23987753688960.0, + "grad_norm": 1.8615941676213836, + "language_loss": 0.8506164, + "learning_rate": 3.735674556341733e-06, + "loss": 0.86766636, + "num_input_tokens_seen": 33632360, + "step": 1585, + "time_per_iteration": 2.723672866821289 + }, + { + "auxiliary_loss_clip": 0.01167185, + "auxiliary_loss_mlp": 0.01036391, + "balance_loss_clip": 0.98881972, + "balance_loss_mlp": 1.0268122, + "epoch": 0.19070522455359826, + "flos": 28293299280000.0, + "grad_norm": 3.300124287709631, + "language_loss": 0.82610929, + "learning_rate": 3.7352873937858835e-06, + "loss": 0.84814501, + "num_input_tokens_seen": 33653895, + "step": 1586, + "time_per_iteration": 2.7507786750793457 + }, + { + "auxiliary_loss_clip": 0.01133601, + "auxiliary_loss_mlp": 0.00512437, + "balance_loss_clip": 0.94482791, + "balance_loss_mlp": 0.0, + "epoch": 0.19082546744423737, + "flos": 25660220797440.0, + "grad_norm": 1.8796774740738162, + "language_loss": 0.72228241, + "learning_rate": 3.734899967988316e-06, + "loss": 0.73874283, + "num_input_tokens_seen": 33672075, + "step": 1587, + "time_per_iteration": 2.798760175704956 + }, + { + "auxiliary_loss_clip": 0.01137142, + "auxiliary_loss_mlp": 0.01030834, + "balance_loss_clip": 0.9465059, + "balance_loss_mlp": 1.02147007, + "epoch": 0.19094571033487645, + "flos": 19719483436800.0, + "grad_norm": 1.9456655779183725, + "language_loss": 0.83802634, + "learning_rate": 3.7345122790078026e-06, + "loss": 0.8597061, + "num_input_tokens_seen": 33689640, + "step": 1588, + "time_per_iteration": 2.945952892303467 + }, + { + "auxiliary_loss_clip": 0.01192612, + "auxiliary_loss_mlp": 0.01031614, + "balance_loss_clip": 1.02573121, + "balance_loss_mlp": 1.02139783, + "epoch": 0.19106595322551553, + "flos": 21616320850560.0, + "grad_norm": 2.417463620643456, + "language_loss": 0.92922282, + "learning_rate": 3.7341243269031556e-06, + "loss": 0.95146501, + "num_input_tokens_seen": 33708630, + "step": 1589, + "time_per_iteration": 2.6189632415771484 + }, + { + "auxiliary_loss_clip": 0.01162505, + "auxiliary_loss_mlp": 0.01031579, + "balance_loss_clip": 0.9866395, + "balance_loss_mlp": 1.02291226, + "epoch": 0.19118619611615464, + "flos": 29896890059520.0, + "grad_norm": 1.7860950860873783, + "language_loss": 0.77280855, + "learning_rate": 3.7337361117332275e-06, + "loss": 0.7947495, + "num_input_tokens_seen": 33730370, + "step": 1590, + "time_per_iteration": 2.785566806793213 + }, + { + "auxiliary_loss_clip": 0.01146521, + "auxiliary_loss_mlp": 0.01036629, + "balance_loss_clip": 0.94564933, + "balance_loss_mlp": 1.02702105, + "epoch": 0.19130643900679373, + "flos": 17273428093440.0, + "grad_norm": 2.0371275919254432, + "language_loss": 0.7700367, + "learning_rate": 3.7333476335569087e-06, + "loss": 0.79186821, + "num_input_tokens_seen": 33748370, + "step": 1591, + "time_per_iteration": 2.6852385997772217 + }, + { + "auxiliary_loss_clip": 0.01164893, + "auxiliary_loss_mlp": 0.0103935, + "balance_loss_clip": 0.98412591, + "balance_loss_mlp": 1.02903867, + "epoch": 0.1914266818974328, + "flos": 24826339584000.0, + "grad_norm": 3.235448594664357, + "language_loss": 0.6718359, + "learning_rate": 3.7329588924331325e-06, + "loss": 0.69387829, + "num_input_tokens_seen": 33769575, + "step": 1592, + "time_per_iteration": 2.7661123275756836 + }, + { + "auxiliary_loss_clip": 0.01134021, + "auxiliary_loss_mlp": 0.01031887, + "balance_loss_clip": 0.94237804, + "balance_loss_mlp": 1.02268982, + "epoch": 0.1915469247880719, + "flos": 18952467390720.0, + "grad_norm": 3.2719777085156974, + "language_loss": 0.82272017, + "learning_rate": 3.732569888420871e-06, + "loss": 0.84437931, + "num_input_tokens_seen": 33789110, + "step": 1593, + "time_per_iteration": 2.7834842205047607 + }, + { + "auxiliary_loss_clip": 0.01214321, + "auxiliary_loss_mlp": 0.01033714, + "balance_loss_clip": 1.05944121, + "balance_loss_mlp": 1.02348626, + "epoch": 0.191667167678711, + "flos": 21032952065280.0, + "grad_norm": 2.1586660648389877, + "language_loss": 0.82527381, + "learning_rate": 3.732180621579134e-06, + "loss": 0.84775418, + "num_input_tokens_seen": 33808325, + "step": 1594, + "time_per_iteration": 2.6318893432617188 + }, + { + "auxiliary_loss_clip": 0.01156536, + "auxiliary_loss_mlp": 0.0104249, + "balance_loss_clip": 0.95110929, + "balance_loss_mlp": 1.03263152, + "epoch": 0.1917874105693501, + "flos": 34237663914240.0, + "grad_norm": 1.953278569669232, + "language_loss": 0.81227541, + "learning_rate": 3.7317910919669745e-06, + "loss": 0.83426571, + "num_input_tokens_seen": 33829520, + "step": 1595, + "time_per_iteration": 2.8467471599578857 + }, + { + "auxiliary_loss_clip": 0.01188849, + "auxiliary_loss_mlp": 0.01037699, + "balance_loss_clip": 1.02449107, + "balance_loss_mlp": 1.02787566, + "epoch": 0.19190765345998917, + "flos": 23550613171200.0, + "grad_norm": 2.6325018479225872, + "language_loss": 0.7627672, + "learning_rate": 3.7314012996434826e-06, + "loss": 0.78503269, + "num_input_tokens_seen": 33848250, + "step": 1596, + "time_per_iteration": 2.674809694290161 + }, + { + "auxiliary_loss_clip": 0.01164218, + "auxiliary_loss_mlp": 0.01034435, + "balance_loss_clip": 0.98428953, + "balance_loss_mlp": 1.02493405, + "epoch": 0.19202789635062828, + "flos": 19861330245120.0, + "grad_norm": 2.0154772747294696, + "language_loss": 0.80889058, + "learning_rate": 3.7310112446677907e-06, + "loss": 0.83087718, + "num_input_tokens_seen": 33866160, + "step": 1597, + "time_per_iteration": 2.730269432067871 + }, + { + "auxiliary_loss_clip": 0.01218267, + "auxiliary_loss_mlp": 0.01042662, + "balance_loss_clip": 1.06469083, + "balance_loss_mlp": 1.03303027, + "epoch": 0.19214813924126736, + "flos": 20922957642240.0, + "grad_norm": 2.2490626045636333, + "language_loss": 0.69153625, + "learning_rate": 3.7306209270990695e-06, + "loss": 0.71414554, + "num_input_tokens_seen": 33884165, + "step": 1598, + "time_per_iteration": 2.5689477920532227 + }, + { + "auxiliary_loss_clip": 0.01167657, + "auxiliary_loss_mlp": 0.01037263, + "balance_loss_clip": 0.98699665, + "balance_loss_mlp": 1.02823853, + "epoch": 0.19226838213190645, + "flos": 26359725231360.0, + "grad_norm": 1.8551101295431338, + "language_loss": 0.8667134, + "learning_rate": 3.7302303469965292e-06, + "loss": 0.88876259, + "num_input_tokens_seen": 33903705, + "step": 1599, + "time_per_iteration": 2.759932279586792 + }, + { + "auxiliary_loss_clip": 0.01191824, + "auxiliary_loss_mlp": 0.01038741, + "balance_loss_clip": 1.02651477, + "balance_loss_mlp": 1.02925229, + "epoch": 0.19238862502254553, + "flos": 20850525866880.0, + "grad_norm": 1.7095263120095023, + "language_loss": 0.71207893, + "learning_rate": 3.7298395044194206e-06, + "loss": 0.7343846, + "num_input_tokens_seen": 33922515, + "step": 1600, + "time_per_iteration": 2.6575424671173096 + }, + { + "auxiliary_loss_clip": 0.01217876, + "auxiliary_loss_mlp": 0.01031582, + "balance_loss_clip": 1.06708789, + "balance_loss_mlp": 1.02252769, + "epoch": 0.19250886791318464, + "flos": 21726063878400.0, + "grad_norm": 2.7267587999543244, + "language_loss": 0.94094259, + "learning_rate": 3.7294483994270356e-06, + "loss": 0.96343714, + "num_input_tokens_seen": 33940840, + "step": 1601, + "time_per_iteration": 2.665445327758789 + }, + { + "auxiliary_loss_clip": 0.01108538, + "auxiliary_loss_mlp": 0.0103604, + "balance_loss_clip": 0.9077549, + "balance_loss_mlp": 1.02709293, + "epoch": 0.19262911080382372, + "flos": 23367827836800.0, + "grad_norm": 3.218085413047365, + "language_loss": 0.77881622, + "learning_rate": 3.7290570320787033e-06, + "loss": 0.80026197, + "num_input_tokens_seen": 33960420, + "step": 1602, + "time_per_iteration": 2.8575496673583984 + }, + { + "auxiliary_loss_clip": 0.01190887, + "auxiliary_loss_mlp": 0.01033878, + "balance_loss_clip": 1.02657402, + "balance_loss_mlp": 1.02460384, + "epoch": 0.1927493536944628, + "flos": 21943502858880.0, + "grad_norm": 1.876667440190265, + "language_loss": 0.71266752, + "learning_rate": 3.728665402433793e-06, + "loss": 0.7349152, + "num_input_tokens_seen": 33978990, + "step": 1603, + "time_per_iteration": 2.698681354522705 + }, + { + "auxiliary_loss_clip": 0.01165997, + "auxiliary_loss_mlp": 0.01031535, + "balance_loss_clip": 0.9873544, + "balance_loss_mlp": 1.0229516, + "epoch": 0.19286959658510192, + "flos": 16545590807040.0, + "grad_norm": 2.600543570845529, + "language_loss": 0.85981274, + "learning_rate": 3.7282735105517164e-06, + "loss": 0.88178802, + "num_input_tokens_seen": 33997115, + "step": 1604, + "time_per_iteration": 3.546617031097412 + }, + { + "auxiliary_loss_clip": 0.011269, + "auxiliary_loss_mlp": 0.01037049, + "balance_loss_clip": 0.90743172, + "balance_loss_mlp": 1.02745223, + "epoch": 0.192989839475741, + "flos": 21616967295360.0, + "grad_norm": 8.088406572071428, + "language_loss": 0.67018533, + "learning_rate": 3.727881356491922e-06, + "loss": 0.69182485, + "num_input_tokens_seen": 34015525, + "step": 1605, + "time_per_iteration": 3.7452869415283203 + }, + { + "auxiliary_loss_clip": 0.01216265, + "auxiliary_loss_mlp": 0.01037238, + "balance_loss_clip": 1.06713653, + "balance_loss_mlp": 1.02842879, + "epoch": 0.19311008236638008, + "flos": 19281516906240.0, + "grad_norm": 1.8879887692341053, + "language_loss": 0.7542873, + "learning_rate": 3.7274889403139002e-06, + "loss": 0.77682233, + "num_input_tokens_seen": 34033150, + "step": 1606, + "time_per_iteration": 2.6214711666107178 + }, + { + "auxiliary_loss_clip": 0.0111513, + "auxiliary_loss_mlp": 0.01037043, + "balance_loss_clip": 0.90945721, + "balance_loss_mlp": 1.0279119, + "epoch": 0.1932303252570192, + "flos": 28652369587200.0, + "grad_norm": 2.224392051830764, + "language_loss": 0.78940773, + "learning_rate": 3.727096262077179e-06, + "loss": 0.81092942, + "num_input_tokens_seen": 34052145, + "step": 1607, + "time_per_iteration": 3.720242500305176 + }, + { + "auxiliary_loss_clip": 0.01191944, + "auxiliary_loss_mlp": 0.01038141, + "balance_loss_clip": 1.0253377, + "balance_loss_mlp": 1.02917659, + "epoch": 0.19335056814765827, + "flos": 18368990864640.0, + "grad_norm": 1.9249447172412686, + "language_loss": 0.85163999, + "learning_rate": 3.7267033218413285e-06, + "loss": 0.87394083, + "num_input_tokens_seen": 34069940, + "step": 1608, + "time_per_iteration": 3.5962960720062256 + }, + { + "auxiliary_loss_clip": 0.01091567, + "auxiliary_loss_mlp": 0.01039243, + "balance_loss_clip": 0.8640244, + "balance_loss_mlp": 1.02978969, + "epoch": 0.19347081103829736, + "flos": 13260877741440.0, + "grad_norm": 2.9541892166675954, + "language_loss": 0.81403071, + "learning_rate": 3.726310119665957e-06, + "loss": 0.83533883, + "num_input_tokens_seen": 34086275, + "step": 1609, + "time_per_iteration": 2.831411123275757 + }, + { + "auxiliary_loss_clip": 0.01191565, + "auxiliary_loss_mlp": 0.01037249, + "balance_loss_clip": 1.02501094, + "balance_loss_mlp": 1.02740788, + "epoch": 0.19359105392893644, + "flos": 20300122788480.0, + "grad_norm": 2.271485143410774, + "language_loss": 0.8543821, + "learning_rate": 3.725916655610713e-06, + "loss": 0.87667024, + "num_input_tokens_seen": 34105605, + "step": 1610, + "time_per_iteration": 2.5991880893707275 + }, + { + "auxiliary_loss_clip": 0.01160049, + "auxiliary_loss_mlp": 0.01044695, + "balance_loss_clip": 0.98488033, + "balance_loss_mlp": 1.0344367, + "epoch": 0.19371129681957555, + "flos": 20484596062080.0, + "grad_norm": 2.704239859682156, + "language_loss": 0.7522375, + "learning_rate": 3.725522929735284e-06, + "loss": 0.77428496, + "num_input_tokens_seen": 34122540, + "step": 1611, + "time_per_iteration": 2.693960428237915 + }, + { + "auxiliary_loss_clip": 0.01169557, + "auxiliary_loss_mlp": 0.01036004, + "balance_loss_clip": 0.98250169, + "balance_loss_mlp": 1.02697408, + "epoch": 0.19383153971021463, + "flos": 30445497457920.0, + "grad_norm": 2.1574945621702026, + "language_loss": 0.73792148, + "learning_rate": 3.725128942099399e-06, + "loss": 0.7599771, + "num_input_tokens_seen": 34142940, + "step": 1612, + "time_per_iteration": 2.7907118797302246 + }, + { + "auxiliary_loss_clip": 0.01157348, + "auxiliary_loss_mlp": 0.01029875, + "balance_loss_clip": 0.98290277, + "balance_loss_mlp": 1.02058899, + "epoch": 0.19395178260085372, + "flos": 24569937325440.0, + "grad_norm": 1.7542664270610493, + "language_loss": 0.79943848, + "learning_rate": 3.7247346927628245e-06, + "loss": 0.82131076, + "num_input_tokens_seen": 34162875, + "step": 1613, + "time_per_iteration": 2.721674680709839 + }, + { + "auxiliary_loss_clip": 0.0116287, + "auxiliary_loss_mlp": 0.00512631, + "balance_loss_clip": 0.9842869, + "balance_loss_mlp": 0.0, + "epoch": 0.19407202549149283, + "flos": 28950608211840.0, + "grad_norm": 2.080492921778135, + "language_loss": 0.79134762, + "learning_rate": 3.7243401817853694e-06, + "loss": 0.80810273, + "num_input_tokens_seen": 34183565, + "step": 1614, + "time_per_iteration": 2.7597239017486572 + }, + { + "auxiliary_loss_clip": 0.01183182, + "auxiliary_loss_mlp": 0.01034202, + "balance_loss_clip": 1.02306271, + "balance_loss_mlp": 1.02561331, + "epoch": 0.1941922683821319, + "flos": 18004497603840.0, + "grad_norm": 1.823064041995414, + "language_loss": 0.71503615, + "learning_rate": 3.723945409226879e-06, + "loss": 0.73721004, + "num_input_tokens_seen": 34202055, + "step": 1615, + "time_per_iteration": 2.5910484790802 + }, + { + "auxiliary_loss_clip": 0.01189224, + "auxiliary_loss_mlp": 0.01038508, + "balance_loss_clip": 1.02325463, + "balance_loss_mlp": 1.02934098, + "epoch": 0.194312511272771, + "flos": 9720337034880.0, + "grad_norm": 2.5329610528703244, + "language_loss": 0.79900044, + "learning_rate": 3.723550375147241e-06, + "loss": 0.82127774, + "num_input_tokens_seen": 34216830, + "step": 1616, + "time_per_iteration": 2.5430545806884766 + }, + { + "auxiliary_loss_clip": 0.01129153, + "auxiliary_loss_mlp": 0.01035473, + "balance_loss_clip": 0.94138169, + "balance_loss_mlp": 1.02594256, + "epoch": 0.19443275416341008, + "flos": 27016208150400.0, + "grad_norm": 2.7289309972259495, + "language_loss": 0.79910755, + "learning_rate": 3.7231550796063816e-06, + "loss": 0.82075375, + "num_input_tokens_seen": 34236840, + "step": 1617, + "time_per_iteration": 2.6425983905792236 + }, + { + "auxiliary_loss_clip": 0.01174062, + "auxiliary_loss_mlp": 0.01033708, + "balance_loss_clip": 0.98735917, + "balance_loss_mlp": 1.02355695, + "epoch": 0.1945529970540492, + "flos": 15846625077120.0, + "grad_norm": 1.943025053180353, + "language_loss": 0.65080357, + "learning_rate": 3.722759522664266e-06, + "loss": 0.67288125, + "num_input_tokens_seen": 34254140, + "step": 1618, + "time_per_iteration": 2.5883350372314453 + }, + { + "auxiliary_loss_clip": 0.01117392, + "auxiliary_loss_mlp": 0.01041686, + "balance_loss_clip": 0.90607989, + "balance_loss_mlp": 1.03263783, + "epoch": 0.19467323994468827, + "flos": 19314985403520.0, + "grad_norm": 2.0724949035457976, + "language_loss": 0.82000589, + "learning_rate": 3.7223637043809016e-06, + "loss": 0.84159672, + "num_input_tokens_seen": 34273120, + "step": 1619, + "time_per_iteration": 2.850294589996338 + }, + { + "auxiliary_loss_clip": 0.01145267, + "auxiliary_loss_mlp": 0.01039537, + "balance_loss_clip": 0.9489637, + "balance_loss_mlp": 1.02970219, + "epoch": 0.19479348283532735, + "flos": 24133227770880.0, + "grad_norm": 2.406537242812584, + "language_loss": 0.8615852, + "learning_rate": 3.7219676248163322e-06, + "loss": 0.88343322, + "num_input_tokens_seen": 34290285, + "step": 1620, + "time_per_iteration": 2.7139008045196533 + }, + { + "auxiliary_loss_clip": 0.01195788, + "auxiliary_loss_mlp": 0.01037683, + "balance_loss_clip": 1.02738047, + "balance_loss_mlp": 1.02742505, + "epoch": 0.19491372572596646, + "flos": 25775638174080.0, + "grad_norm": 1.627842434382829, + "language_loss": 0.93217099, + "learning_rate": 3.721571284030643e-06, + "loss": 0.95450574, + "num_input_tokens_seen": 34310095, + "step": 1621, + "time_per_iteration": 2.750674247741699 + }, + { + "auxiliary_loss_clip": 0.01196804, + "auxiliary_loss_mlp": 0.01039243, + "balance_loss_clip": 1.02689767, + "balance_loss_mlp": 1.02981937, + "epoch": 0.19503396861660555, + "flos": 19645220067840.0, + "grad_norm": 2.202065179629091, + "language_loss": 0.78856784, + "learning_rate": 3.7211746820839587e-06, + "loss": 0.81092834, + "num_input_tokens_seen": 34327190, + "step": 1622, + "time_per_iteration": 2.713569164276123 + }, + { + "auxiliary_loss_clip": 0.01053252, + "auxiliary_loss_mlp": 0.01039585, + "balance_loss_clip": 0.8266108, + "balance_loss_mlp": 1.03000081, + "epoch": 0.19515421150724463, + "flos": 21033023892480.0, + "grad_norm": 1.9282932682939664, + "language_loss": 0.81059366, + "learning_rate": 3.7207778190364437e-06, + "loss": 0.83152199, + "num_input_tokens_seen": 34345615, + "step": 1623, + "time_per_iteration": 2.854759931564331 + }, + { + "auxiliary_loss_clip": 0.01080755, + "auxiliary_loss_mlp": 0.01036767, + "balance_loss_clip": 0.86768311, + "balance_loss_mlp": 1.02709889, + "epoch": 0.1952744543978837, + "flos": 32961255143040.0, + "grad_norm": 1.552262910401838, + "language_loss": 0.73707217, + "learning_rate": 3.720380694948302e-06, + "loss": 0.75824738, + "num_input_tokens_seen": 34368500, + "step": 1624, + "time_per_iteration": 2.954641580581665 + }, + { + "auxiliary_loss_clip": 0.01047429, + "auxiliary_loss_mlp": 0.01027212, + "balance_loss_clip": 0.890513, + "balance_loss_mlp": 1.02510202, + "epoch": 0.19539469728852282, + "flos": 64044312030720.0, + "grad_norm": 1.0414299462945156, + "language_loss": 0.71253467, + "learning_rate": 3.719983309879777e-06, + "loss": 0.73328102, + "num_input_tokens_seen": 34428280, + "step": 1625, + "time_per_iteration": 3.2983779907226562 + }, + { + "auxiliary_loss_clip": 0.01133652, + "auxiliary_loss_mlp": 0.01037577, + "balance_loss_clip": 0.94563508, + "balance_loss_mlp": 1.02838552, + "epoch": 0.1955149401791619, + "flos": 13370908078080.0, + "grad_norm": 1.5911956032679986, + "language_loss": 0.77148014, + "learning_rate": 3.719585663891151e-06, + "loss": 0.79319239, + "num_input_tokens_seen": 34445815, + "step": 1626, + "time_per_iteration": 2.6499152183532715 + }, + { + "auxiliary_loss_clip": 0.01110556, + "auxiliary_loss_mlp": 0.01034816, + "balance_loss_clip": 0.9089272, + "balance_loss_mlp": 1.02459335, + "epoch": 0.195635183069801, + "flos": 18728887184640.0, + "grad_norm": 8.300282574733593, + "language_loss": 0.78412247, + "learning_rate": 3.719187757042747e-06, + "loss": 0.80557615, + "num_input_tokens_seen": 34463635, + "step": 1627, + "time_per_iteration": 2.773684024810791 + }, + { + "auxiliary_loss_clip": 0.0109097, + "auxiliary_loss_mlp": 0.01009858, + "balance_loss_clip": 0.97292531, + "balance_loss_mlp": 1.00774765, + "epoch": 0.1957554259604401, + "flos": 69313952615040.0, + "grad_norm": 0.7559632083325418, + "language_loss": 0.54944372, + "learning_rate": 3.7187895893949275e-06, + "loss": 0.57045203, + "num_input_tokens_seen": 34530105, + "step": 1628, + "time_per_iteration": 3.3177363872528076 + }, + { + "auxiliary_loss_clip": 0.01103631, + "auxiliary_loss_mlp": 0.01035715, + "balance_loss_clip": 0.90453613, + "balance_loss_mlp": 1.02571344, + "epoch": 0.19587566885107918, + "flos": 21069257736960.0, + "grad_norm": 2.387051901391992, + "language_loss": 0.76156616, + "learning_rate": 3.7183911610080937e-06, + "loss": 0.78295958, + "num_input_tokens_seen": 34546970, + "step": 1629, + "time_per_iteration": 2.7853457927703857 + }, + { + "auxiliary_loss_clip": 0.01142172, + "auxiliary_loss_mlp": 0.0103782, + "balance_loss_clip": 0.94724637, + "balance_loss_mlp": 1.02795577, + "epoch": 0.19599591174171827, + "flos": 22194661731840.0, + "grad_norm": 2.412245937431065, + "language_loss": 0.74955922, + "learning_rate": 3.7179924719426872e-06, + "loss": 0.77135915, + "num_input_tokens_seen": 34564865, + "step": 1630, + "time_per_iteration": 3.6585886478424072 + }, + { + "auxiliary_loss_clip": 0.01195073, + "auxiliary_loss_mlp": 0.01028908, + "balance_loss_clip": 1.02832437, + "balance_loss_mlp": 1.01929939, + "epoch": 0.19611615463235738, + "flos": 23768375374080.0, + "grad_norm": 3.429868645652259, + "language_loss": 0.76150393, + "learning_rate": 3.7175935222591885e-06, + "loss": 0.78374374, + "num_input_tokens_seen": 34584165, + "step": 1631, + "time_per_iteration": 3.62199330329895 + }, + { + "auxiliary_loss_clip": 0.01172957, + "auxiliary_loss_mlp": 0.01039497, + "balance_loss_clip": 0.99208832, + "balance_loss_mlp": 1.02962077, + "epoch": 0.19623639752299646, + "flos": 28618218731520.0, + "grad_norm": 1.7950834152259625, + "language_loss": 0.74177259, + "learning_rate": 3.717194312018118e-06, + "loss": 0.76389712, + "num_input_tokens_seen": 34603150, + "step": 1632, + "time_per_iteration": 2.721092462539673 + }, + { + "auxiliary_loss_clip": 0.01192071, + "auxiliary_loss_mlp": 0.01041234, + "balance_loss_clip": 1.02495241, + "balance_loss_mlp": 1.0307914, + "epoch": 0.19635664041363554, + "flos": 21032700670080.0, + "grad_norm": 5.837357367154608, + "language_loss": 0.76067644, + "learning_rate": 3.716794841280036e-06, + "loss": 0.78300947, + "num_input_tokens_seen": 34621855, + "step": 1633, + "time_per_iteration": 4.563371419906616 + }, + { + "auxiliary_loss_clip": 0.01198965, + "auxiliary_loss_mlp": 0.01040286, + "balance_loss_clip": 1.02744448, + "balance_loss_mlp": 1.02981961, + "epoch": 0.19647688330427462, + "flos": 18879748306560.0, + "grad_norm": 2.603456720645393, + "language_loss": 0.77501917, + "learning_rate": 3.7163951101055407e-06, + "loss": 0.79741168, + "num_input_tokens_seen": 34639915, + "step": 1634, + "time_per_iteration": 2.6992318630218506 + }, + { + "auxiliary_loss_clip": 0.01163716, + "auxiliary_loss_mlp": 0.0104307, + "balance_loss_clip": 0.98581064, + "balance_loss_mlp": 1.03302121, + "epoch": 0.19659712619491373, + "flos": 24242503921920.0, + "grad_norm": 2.2363301295370945, + "language_loss": 0.78889441, + "learning_rate": 3.715995118555273e-06, + "loss": 0.81096232, + "num_input_tokens_seen": 34659890, + "step": 1635, + "time_per_iteration": 2.8470733165740967 + }, + { + "auxiliary_loss_clip": 0.01117329, + "auxiliary_loss_mlp": 0.01035831, + "balance_loss_clip": 0.90853333, + "balance_loss_mlp": 1.02448177, + "epoch": 0.19671736908555282, + "flos": 24717422568960.0, + "grad_norm": 2.009414240412389, + "language_loss": 0.86075211, + "learning_rate": 3.71559486668991e-06, + "loss": 0.88228369, + "num_input_tokens_seen": 34678750, + "step": 1636, + "time_per_iteration": 2.738138437271118 + }, + { + "auxiliary_loss_clip": 0.01200022, + "auxiliary_loss_mlp": 0.00511504, + "balance_loss_clip": 1.03001714, + "balance_loss_mlp": 0.0, + "epoch": 0.1968376119761919, + "flos": 23842279607040.0, + "grad_norm": 2.722355734812083, + "language_loss": 0.77335602, + "learning_rate": 3.715194354570169e-06, + "loss": 0.79047132, + "num_input_tokens_seen": 34698755, + "step": 1637, + "time_per_iteration": 2.6772801876068115 + }, + { + "auxiliary_loss_clip": 0.01194178, + "auxiliary_loss_mlp": 0.01035021, + "balance_loss_clip": 1.03021407, + "balance_loss_mlp": 1.02548385, + "epoch": 0.196957854866831, + "flos": 18113917409280.0, + "grad_norm": 2.58319480336508, + "language_loss": 0.83031988, + "learning_rate": 3.714793582256809e-06, + "loss": 0.8526119, + "num_input_tokens_seen": 34715820, + "step": 1638, + "time_per_iteration": 2.7187798023223877 + }, + { + "auxiliary_loss_clip": 0.01217614, + "auxiliary_loss_mlp": 0.01039089, + "balance_loss_clip": 1.06690896, + "balance_loss_mlp": 1.02982068, + "epoch": 0.1970780977574701, + "flos": 21653129312640.0, + "grad_norm": 5.855996752537947, + "language_loss": 0.8521449, + "learning_rate": 3.7143925498106253e-06, + "loss": 0.87471199, + "num_input_tokens_seen": 34734360, + "step": 1639, + "time_per_iteration": 2.586733341217041 + }, + { + "auxiliary_loss_clip": 0.01166512, + "auxiliary_loss_mlp": 0.01032123, + "balance_loss_clip": 0.98358464, + "balance_loss_mlp": 1.02244329, + "epoch": 0.19719834064810918, + "flos": 20811813984000.0, + "grad_norm": 1.9196768091815983, + "language_loss": 0.78660643, + "learning_rate": 3.7139912572924558e-06, + "loss": 0.8085928, + "num_input_tokens_seen": 34753390, + "step": 1640, + "time_per_iteration": 2.714266538619995 + }, + { + "auxiliary_loss_clip": 0.01185165, + "auxiliary_loss_mlp": 0.0103975, + "balance_loss_clip": 1.02016687, + "balance_loss_mlp": 1.03057671, + "epoch": 0.19731858353874826, + "flos": 23434800744960.0, + "grad_norm": 3.3834868588432445, + "language_loss": 0.80524218, + "learning_rate": 3.7135897047631744e-06, + "loss": 0.82749134, + "num_input_tokens_seen": 34771275, + "step": 1641, + "time_per_iteration": 2.63014554977417 + }, + { + "auxiliary_loss_clip": 0.01171971, + "auxiliary_loss_mlp": 0.01038054, + "balance_loss_clip": 0.98856604, + "balance_loss_mlp": 1.0276649, + "epoch": 0.19743882642938737, + "flos": 23988184652160.0, + "grad_norm": 2.1200841287402112, + "language_loss": 0.76126015, + "learning_rate": 3.713187892283698e-06, + "loss": 0.78336036, + "num_input_tokens_seen": 34790885, + "step": 1642, + "time_per_iteration": 2.7082014083862305 + }, + { + "auxiliary_loss_clip": 0.01120112, + "auxiliary_loss_mlp": 0.01036981, + "balance_loss_clip": 0.91016018, + "balance_loss_mlp": 1.02716374, + "epoch": 0.19755906932002645, + "flos": 15004340081280.0, + "grad_norm": 2.347841012846165, + "language_loss": 0.86937726, + "learning_rate": 3.71278581991498e-06, + "loss": 0.89094818, + "num_input_tokens_seen": 34806745, + "step": 1643, + "time_per_iteration": 2.6820223331451416 + }, + { + "auxiliary_loss_clip": 0.01150419, + "auxiliary_loss_mlp": 0.00512687, + "balance_loss_clip": 0.95438975, + "balance_loss_mlp": 0.0, + "epoch": 0.19767931221066554, + "flos": 19494466686720.0, + "grad_norm": 1.636799755199973, + "language_loss": 0.79099715, + "learning_rate": 3.712383487718015e-06, + "loss": 0.80762815, + "num_input_tokens_seen": 34824985, + "step": 1644, + "time_per_iteration": 2.766242504119873 + }, + { + "auxiliary_loss_clip": 0.01089787, + "auxiliary_loss_mlp": 0.01037931, + "balance_loss_clip": 0.8699888, + "balance_loss_mlp": 1.02818561, + "epoch": 0.19779955510130465, + "flos": 25737895958400.0, + "grad_norm": 2.0204442433809673, + "language_loss": 0.86555719, + "learning_rate": 3.7119808957538365e-06, + "loss": 0.88683438, + "num_input_tokens_seen": 34843980, + "step": 1645, + "time_per_iteration": 2.8877289295196533 + }, + { + "auxiliary_loss_clip": 0.01161205, + "auxiliary_loss_mlp": 0.01033012, + "balance_loss_clip": 0.98303807, + "balance_loss_mlp": 1.02348769, + "epoch": 0.19791979799194373, + "flos": 20777699041920.0, + "grad_norm": 1.971225334098459, + "language_loss": 0.80143213, + "learning_rate": 3.711578044083517e-06, + "loss": 0.82337427, + "num_input_tokens_seen": 34860780, + "step": 1646, + "time_per_iteration": 2.713834047317505 + }, + { + "auxiliary_loss_clip": 0.01168749, + "auxiliary_loss_mlp": 0.01040012, + "balance_loss_clip": 0.98627639, + "balance_loss_mlp": 1.02993298, + "epoch": 0.1980400408825828, + "flos": 25589010084480.0, + "grad_norm": 2.52942054406065, + "language_loss": 0.74244475, + "learning_rate": 3.7111749327681698e-06, + "loss": 0.76453233, + "num_input_tokens_seen": 34880815, + "step": 1647, + "time_per_iteration": 2.7752327919006348 + }, + { + "auxiliary_loss_clip": 0.01198011, + "auxiliary_loss_mlp": 0.01037266, + "balance_loss_clip": 1.02869296, + "balance_loss_mlp": 1.02713346, + "epoch": 0.1981602837732219, + "flos": 23513840622720.0, + "grad_norm": 2.632622507483196, + "language_loss": 0.86236095, + "learning_rate": 3.7107715618689455e-06, + "loss": 0.88471365, + "num_input_tokens_seen": 34899790, + "step": 1648, + "time_per_iteration": 2.686276912689209 + }, + { + "auxiliary_loss_clip": 0.01192811, + "auxiliary_loss_mlp": 0.01038449, + "balance_loss_clip": 1.02806318, + "balance_loss_mlp": 1.02826834, + "epoch": 0.198280526663861, + "flos": 23185365724800.0, + "grad_norm": 1.447458178234958, + "language_loss": 0.83536536, + "learning_rate": 3.710367931447035e-06, + "loss": 0.85767794, + "num_input_tokens_seen": 34921570, + "step": 1649, + "time_per_iteration": 2.7776288986206055 + }, + { + "auxiliary_loss_clip": 0.01201649, + "auxiliary_loss_mlp": 0.01038756, + "balance_loss_clip": 1.02811229, + "balance_loss_mlp": 1.02864671, + "epoch": 0.1984007695545001, + "flos": 21689470897920.0, + "grad_norm": 4.6032556222529175, + "language_loss": 0.86409891, + "learning_rate": 3.70996404156367e-06, + "loss": 0.88650298, + "num_input_tokens_seen": 34941205, + "step": 1650, + "time_per_iteration": 2.6311709880828857 + }, + { + "auxiliary_loss_clip": 0.0111203, + "auxiliary_loss_mlp": 0.0103588, + "balance_loss_clip": 0.90792751, + "balance_loss_mlp": 1.02578926, + "epoch": 0.19852101244513917, + "flos": 36064008887040.0, + "grad_norm": 1.987161160316253, + "language_loss": 0.72662044, + "learning_rate": 3.7095598922801187e-06, + "loss": 0.74809951, + "num_input_tokens_seen": 34963280, + "step": 1651, + "time_per_iteration": 2.855131149291992 + }, + { + "auxiliary_loss_clip": 0.01216438, + "auxiliary_loss_mlp": 0.01033544, + "balance_loss_clip": 1.06489801, + "balance_loss_mlp": 1.02416849, + "epoch": 0.19864125533577828, + "flos": 23105894883840.0, + "grad_norm": 2.766281344591431, + "language_loss": 0.762348, + "learning_rate": 3.7091554836576914e-06, + "loss": 0.7848478, + "num_input_tokens_seen": 34979955, + "step": 1652, + "time_per_iteration": 2.590780735015869 + }, + { + "auxiliary_loss_clip": 0.0119068, + "auxiliary_loss_mlp": 0.00512246, + "balance_loss_clip": 1.0257206, + "balance_loss_mlp": 0.0, + "epoch": 0.19876149822641737, + "flos": 24608505553920.0, + "grad_norm": 1.7963590170367238, + "language_loss": 0.82577688, + "learning_rate": 3.708750815757736e-06, + "loss": 0.8428061, + "num_input_tokens_seen": 35000725, + "step": 1653, + "time_per_iteration": 2.7204220294952393 + }, + { + "auxiliary_loss_clip": 0.01196639, + "auxiliary_loss_mlp": 0.01039364, + "balance_loss_clip": 1.02731538, + "balance_loss_mlp": 1.02895701, + "epoch": 0.19888174111705645, + "flos": 32196645308160.0, + "grad_norm": 2.3903474662837687, + "language_loss": 0.73471999, + "learning_rate": 3.7083458886416407e-06, + "loss": 0.75707996, + "num_input_tokens_seen": 35019920, + "step": 1654, + "time_per_iteration": 2.692584991455078 + }, + { + "auxiliary_loss_clip": 0.01101349, + "auxiliary_loss_mlp": 0.01035731, + "balance_loss_clip": 0.87431288, + "balance_loss_mlp": 1.02615213, + "epoch": 0.19900198400769553, + "flos": 24608469640320.0, + "grad_norm": 2.9023456575548696, + "language_loss": 0.87846446, + "learning_rate": 3.707940702370832e-06, + "loss": 0.89983529, + "num_input_tokens_seen": 35040765, + "step": 1655, + "time_per_iteration": 2.847832202911377 + }, + { + "auxiliary_loss_clip": 0.01102089, + "auxiliary_loss_mlp": 0.01012906, + "balance_loss_clip": 1.00217032, + "balance_loss_mlp": 1.01093853, + "epoch": 0.19912222689833464, + "flos": 67915805673600.0, + "grad_norm": 0.7677407374655726, + "language_loss": 0.58270931, + "learning_rate": 3.707535257006777e-06, + "loss": 0.60385925, + "num_input_tokens_seen": 35106390, + "step": 1656, + "time_per_iteration": 4.22888445854187 + }, + { + "auxiliary_loss_clip": 0.0116846, + "auxiliary_loss_mlp": 0.01028315, + "balance_loss_clip": 0.98649132, + "balance_loss_mlp": 1.01795614, + "epoch": 0.19924246978897373, + "flos": 15742340916480.0, + "grad_norm": 2.3953341128959402, + "language_loss": 0.88376665, + "learning_rate": 3.707129552610981e-06, + "loss": 0.90573442, + "num_input_tokens_seen": 35125040, + "step": 1657, + "time_per_iteration": 2.6512861251831055 + }, + { + "auxiliary_loss_clip": 0.01158328, + "auxiliary_loss_mlp": 0.01035475, + "balance_loss_clip": 0.98466051, + "balance_loss_mlp": 1.02515745, + "epoch": 0.1993627126796128, + "flos": 17566566986880.0, + "grad_norm": 1.8084511344598695, + "language_loss": 0.73880017, + "learning_rate": 3.70672358924499e-06, + "loss": 0.76073819, + "num_input_tokens_seen": 35144280, + "step": 1658, + "time_per_iteration": 2.6735525131225586 + }, + { + "auxiliary_loss_clip": 0.01144536, + "auxiliary_loss_mlp": 0.0104008, + "balance_loss_clip": 0.95165354, + "balance_loss_mlp": 1.03042448, + "epoch": 0.19948295557025192, + "flos": 40843826680320.0, + "grad_norm": 1.8610676183516088, + "language_loss": 0.7815249, + "learning_rate": 3.706317366970386e-06, + "loss": 0.80337107, + "num_input_tokens_seen": 35165280, + "step": 1659, + "time_per_iteration": 3.776402711868286 + }, + { + "auxiliary_loss_clip": 0.01218278, + "auxiliary_loss_mlp": 0.00513023, + "balance_loss_clip": 1.06389046, + "balance_loss_mlp": 0.0, + "epoch": 0.199603198460891, + "flos": 25082418620160.0, + "grad_norm": 1.8958667085503613, + "language_loss": 0.83914983, + "learning_rate": 3.705910885848795e-06, + "loss": 0.85646284, + "num_input_tokens_seen": 35183655, + "step": 1660, + "time_per_iteration": 3.571920156478882 + }, + { + "auxiliary_loss_clip": 0.01192755, + "auxiliary_loss_mlp": 0.01037155, + "balance_loss_clip": 1.0277096, + "balance_loss_mlp": 1.02726614, + "epoch": 0.19972344135153008, + "flos": 20084120352000.0, + "grad_norm": 2.7266230131458706, + "language_loss": 0.84474242, + "learning_rate": 3.705504145941879e-06, + "loss": 0.86704147, + "num_input_tokens_seen": 35201825, + "step": 1661, + "time_per_iteration": 2.6915054321289062 + }, + { + "auxiliary_loss_clip": 0.01213656, + "auxiliary_loss_mlp": 0.0103967, + "balance_loss_clip": 1.06347203, + "balance_loss_mlp": 1.0299542, + "epoch": 0.1998436842421692, + "flos": 23727472761600.0, + "grad_norm": 1.9231771177124442, + "language_loss": 0.79087549, + "learning_rate": 3.7050971473113403e-06, + "loss": 0.81340873, + "num_input_tokens_seen": 35221600, + "step": 1662, + "time_per_iteration": 2.62434983253479 + }, + { + "auxiliary_loss_clip": 0.01190392, + "auxiliary_loss_mlp": 0.00511586, + "balance_loss_clip": 1.0256418, + "balance_loss_mlp": 0.0, + "epoch": 0.19996392713280828, + "flos": 36102361633920.0, + "grad_norm": 1.6472569621574547, + "language_loss": 0.7976653, + "learning_rate": 3.7046898900189196e-06, + "loss": 0.81468505, + "num_input_tokens_seen": 35245935, + "step": 1663, + "time_per_iteration": 2.78914475440979 + }, + { + "auxiliary_loss_clip": 0.01151594, + "auxiliary_loss_mlp": 0.0103761, + "balance_loss_clip": 0.95265245, + "balance_loss_mlp": 1.02695239, + "epoch": 0.20008417002344736, + "flos": 23657662679040.0, + "grad_norm": 1.7573342298357852, + "language_loss": 0.83051342, + "learning_rate": 3.704282374126398e-06, + "loss": 0.85240543, + "num_input_tokens_seen": 35265615, + "step": 1664, + "time_per_iteration": 2.711301803588867 + }, + { + "auxiliary_loss_clip": 0.01140937, + "auxiliary_loss_mlp": 0.01037191, + "balance_loss_clip": 0.9470942, + "balance_loss_mlp": 1.0272789, + "epoch": 0.20020441291408644, + "flos": 21872076664320.0, + "grad_norm": 1.804338439123602, + "language_loss": 0.87449771, + "learning_rate": 3.7038745996955954e-06, + "loss": 0.89627898, + "num_input_tokens_seen": 35284960, + "step": 1665, + "time_per_iteration": 238.27931690216064 + }, + { + "auxiliary_loss_clip": 0.01148391, + "auxiliary_loss_mlp": 0.01039307, + "balance_loss_clip": 0.94753826, + "balance_loss_mlp": 1.02921033, + "epoch": 0.20032465580472555, + "flos": 23179691376000.0, + "grad_norm": 3.5873459717444245, + "language_loss": 0.72019905, + "learning_rate": 3.703466566788371e-06, + "loss": 0.74207598, + "num_input_tokens_seen": 35304090, + "step": 1666, + "time_per_iteration": 2.7369096279144287 + }, + { + "auxiliary_loss_clip": 0.01157881, + "auxiliary_loss_mlp": 0.01036978, + "balance_loss_clip": 0.98536193, + "balance_loss_mlp": 1.02726889, + "epoch": 0.20044489869536464, + "flos": 23873521461120.0, + "grad_norm": 2.054584505237254, + "language_loss": 0.7460596, + "learning_rate": 3.703058275466622e-06, + "loss": 0.76800823, + "num_input_tokens_seen": 35323325, + "step": 1667, + "time_per_iteration": 2.6678569316864014 + }, + { + "auxiliary_loss_clip": 0.0116522, + "auxiliary_loss_mlp": 0.01042423, + "balance_loss_clip": 0.98427212, + "balance_loss_mlp": 1.03227234, + "epoch": 0.20056514158600372, + "flos": 21945226711680.0, + "grad_norm": 1.6561194900919014, + "language_loss": 0.77545762, + "learning_rate": 3.7026497257922877e-06, + "loss": 0.79753411, + "num_input_tokens_seen": 35343635, + "step": 1668, + "time_per_iteration": 2.691164493560791 + }, + { + "auxiliary_loss_clip": 0.01115912, + "auxiliary_loss_mlp": 0.01033599, + "balance_loss_clip": 0.90930778, + "balance_loss_mlp": 1.02318597, + "epoch": 0.20068538447664283, + "flos": 23879159896320.0, + "grad_norm": 1.724285658206541, + "language_loss": 0.85122001, + "learning_rate": 3.7022409178273436e-06, + "loss": 0.87271518, + "num_input_tokens_seen": 35364615, + "step": 1669, + "time_per_iteration": 2.70910906791687 + }, + { + "auxiliary_loss_clip": 0.01188254, + "auxiliary_loss_mlp": 0.01027696, + "balance_loss_clip": 1.0240612, + "balance_loss_mlp": 1.01875508, + "epoch": 0.2008056273672819, + "flos": 18442823270400.0, + "grad_norm": 2.0673817786749504, + "language_loss": 0.78442454, + "learning_rate": 3.7018318516338054e-06, + "loss": 0.806584, + "num_input_tokens_seen": 35383775, + "step": 1670, + "time_per_iteration": 2.5746917724609375 + }, + { + "auxiliary_loss_clip": 0.01196049, + "auxiliary_loss_mlp": 0.01042939, + "balance_loss_clip": 1.026021, + "balance_loss_mlp": 1.03318155, + "epoch": 0.200925870257921, + "flos": 23659530186240.0, + "grad_norm": 4.987451596470119, + "language_loss": 0.81498909, + "learning_rate": 3.7014225272737284e-06, + "loss": 0.83737898, + "num_input_tokens_seen": 35403000, + "step": 1671, + "time_per_iteration": 2.8076517581939697 + }, + { + "auxiliary_loss_clip": 0.01183069, + "auxiliary_loss_mlp": 0.01041955, + "balance_loss_clip": 1.0220418, + "balance_loss_mlp": 1.03210258, + "epoch": 0.20104611314856008, + "flos": 16217115909120.0, + "grad_norm": 2.229947331953207, + "language_loss": 0.74383336, + "learning_rate": 3.701012944809207e-06, + "loss": 0.76608354, + "num_input_tokens_seen": 35420115, + "step": 1672, + "time_per_iteration": 2.8017401695251465 + }, + { + "auxiliary_loss_clip": 0.01167077, + "auxiliary_loss_mlp": 0.00512037, + "balance_loss_clip": 0.9876399, + "balance_loss_mlp": 0.0, + "epoch": 0.2011663560391992, + "flos": 21397373498880.0, + "grad_norm": 1.9732518984691825, + "language_loss": 0.78520584, + "learning_rate": 3.700603104302374e-06, + "loss": 0.80199695, + "num_input_tokens_seen": 35439925, + "step": 1673, + "time_per_iteration": 2.679718255996704 + }, + { + "auxiliary_loss_clip": 0.01047821, + "auxiliary_loss_mlp": 0.01007132, + "balance_loss_clip": 0.92773032, + "balance_loss_mlp": 1.00508165, + "epoch": 0.20128659892983827, + "flos": 62229459409920.0, + "grad_norm": 0.904111688509511, + "language_loss": 0.55971766, + "learning_rate": 3.7001930058154027e-06, + "loss": 0.58026719, + "num_input_tokens_seen": 35504885, + "step": 1674, + "time_per_iteration": 3.380438804626465 + }, + { + "auxiliary_loss_clip": 0.01142544, + "auxiliary_loss_mlp": 0.01042624, + "balance_loss_clip": 0.9481585, + "balance_loss_mlp": 1.03238952, + "epoch": 0.20140684182047736, + "flos": 28438737448320.0, + "grad_norm": 2.4407103396858916, + "language_loss": 0.79612434, + "learning_rate": 3.6997826494105037e-06, + "loss": 0.817976, + "num_input_tokens_seen": 35525330, + "step": 1675, + "time_per_iteration": 2.798696517944336 + }, + { + "auxiliary_loss_clip": 0.01167199, + "auxiliary_loss_mlp": 0.01032066, + "balance_loss_clip": 0.98475575, + "balance_loss_mlp": 1.02280903, + "epoch": 0.20152708471111647, + "flos": 28074064619520.0, + "grad_norm": 2.0366615843190066, + "language_loss": 0.69400799, + "learning_rate": 3.6993720351499286e-06, + "loss": 0.71600062, + "num_input_tokens_seen": 35546455, + "step": 1676, + "time_per_iteration": 2.706235647201538 + }, + { + "auxiliary_loss_clip": 0.01163077, + "auxiliary_loss_mlp": 0.01036254, + "balance_loss_clip": 0.98966205, + "balance_loss_mlp": 1.02736688, + "epoch": 0.20164732760175555, + "flos": 23549751244800.0, + "grad_norm": 1.8921710656923267, + "language_loss": 0.76729274, + "learning_rate": 3.6989611630959666e-06, + "loss": 0.78928602, + "num_input_tokens_seen": 35565010, + "step": 1677, + "time_per_iteration": 2.7425265312194824 + }, + { + "auxiliary_loss_clip": 0.01100347, + "auxiliary_loss_mlp": 0.01004891, + "balance_loss_clip": 1.00017333, + "balance_loss_mlp": 1.00281644, + "epoch": 0.20176757049239463, + "flos": 71100616037760.0, + "grad_norm": 0.6824470073052303, + "language_loss": 0.5828886, + "learning_rate": 3.6985500333109474e-06, + "loss": 0.60394102, + "num_input_tokens_seen": 35633340, + "step": 1678, + "time_per_iteration": 3.3052899837493896 + }, + { + "auxiliary_loss_clip": 0.01133854, + "auxiliary_loss_mlp": 0.01034453, + "balance_loss_clip": 0.94286752, + "balance_loss_mlp": 1.02502346, + "epoch": 0.20188781338303372, + "flos": 21430159637760.0, + "grad_norm": 2.66380301320088, + "language_loss": 0.75984162, + "learning_rate": 3.6981386458572385e-06, + "loss": 0.78152466, + "num_input_tokens_seen": 35651315, + "step": 1679, + "time_per_iteration": 2.790921688079834 + }, + { + "auxiliary_loss_clip": 0.01134882, + "auxiliary_loss_mlp": 0.01035644, + "balance_loss_clip": 0.94461203, + "balance_loss_mlp": 1.02522528, + "epoch": 0.20200805627367283, + "flos": 11546215130880.0, + "grad_norm": 2.2712022535161567, + "language_loss": 0.76336879, + "learning_rate": 3.6977270007972468e-06, + "loss": 0.78507406, + "num_input_tokens_seen": 35668850, + "step": 1680, + "time_per_iteration": 2.6330630779266357 + }, + { + "auxiliary_loss_clip": 0.0116885, + "auxiliary_loss_mlp": 0.01036167, + "balance_loss_clip": 0.98676586, + "balance_loss_mlp": 1.02708292, + "epoch": 0.2021282991643119, + "flos": 28545391906560.0, + "grad_norm": 2.358203271850336, + "language_loss": 0.72187138, + "learning_rate": 3.6973150981934196e-06, + "loss": 0.74392152, + "num_input_tokens_seen": 35690080, + "step": 1681, + "time_per_iteration": 4.535408973693848 + }, + { + "auxiliary_loss_clip": 0.01219034, + "auxiliary_loss_mlp": 0.010389, + "balance_loss_clip": 1.06448603, + "balance_loss_mlp": 1.02929175, + "epoch": 0.202248542054951, + "flos": 17923446564480.0, + "grad_norm": 2.3420227311854993, + "language_loss": 0.83427334, + "learning_rate": 3.6969029381082415e-06, + "loss": 0.85685265, + "num_input_tokens_seen": 35706075, + "step": 1682, + "time_per_iteration": 2.6910457611083984 + }, + { + "auxiliary_loss_clip": 0.01165536, + "auxiliary_loss_mlp": 0.01033066, + "balance_loss_clip": 0.98630345, + "balance_loss_mlp": 1.02448344, + "epoch": 0.2023687849455901, + "flos": 19864634296320.0, + "grad_norm": 1.697773188661186, + "language_loss": 0.79385608, + "learning_rate": 3.696490520604237e-06, + "loss": 0.81584209, + "num_input_tokens_seen": 35724765, + "step": 1683, + "time_per_iteration": 2.632603168487549 + }, + { + "auxiliary_loss_clip": 0.01187429, + "auxiliary_loss_mlp": 0.01033165, + "balance_loss_clip": 1.02414918, + "balance_loss_mlp": 1.02465916, + "epoch": 0.20248902783622919, + "flos": 22564721600640.0, + "grad_norm": 1.9145704435993014, + "language_loss": 0.80558085, + "learning_rate": 3.696077845743968e-06, + "loss": 0.82778674, + "num_input_tokens_seen": 35744355, + "step": 1684, + "time_per_iteration": 2.667296886444092 + }, + { + "auxiliary_loss_clip": 0.01215305, + "auxiliary_loss_mlp": 0.01039917, + "balance_loss_clip": 1.06254053, + "balance_loss_mlp": 1.0296948, + "epoch": 0.20260927072686827, + "flos": 22709728805760.0, + "grad_norm": 2.4639943513392417, + "language_loss": 0.7351687, + "learning_rate": 3.69566491359004e-06, + "loss": 0.75772095, + "num_input_tokens_seen": 35761000, + "step": 1685, + "time_per_iteration": 4.435232639312744 + }, + { + "auxiliary_loss_clip": 0.01161887, + "auxiliary_loss_mlp": 0.0103758, + "balance_loss_clip": 0.98340535, + "balance_loss_mlp": 1.02778101, + "epoch": 0.20272951361750738, + "flos": 51023998650240.0, + "grad_norm": 1.8616376206527836, + "language_loss": 0.69344485, + "learning_rate": 3.695251724205092e-06, + "loss": 0.71543956, + "num_input_tokens_seen": 35785360, + "step": 1686, + "time_per_iteration": 2.8498175144195557 + }, + { + "auxiliary_loss_clip": 0.01213291, + "auxiliary_loss_mlp": 0.01034791, + "balance_loss_clip": 1.06350636, + "balance_loss_mlp": 1.0248307, + "epoch": 0.20284975650814646, + "flos": 26578133879040.0, + "grad_norm": 1.794549952655838, + "language_loss": 0.86516237, + "learning_rate": 3.6948382776518054e-06, + "loss": 0.8876431, + "num_input_tokens_seen": 35806065, + "step": 1687, + "time_per_iteration": 2.6586337089538574 + }, + { + "auxiliary_loss_clip": 0.01149374, + "auxiliary_loss_mlp": 0.01039631, + "balance_loss_clip": 0.94725037, + "balance_loss_mlp": 1.02945709, + "epoch": 0.20296999939878554, + "flos": 16034222833920.0, + "grad_norm": 2.0495214245336335, + "language_loss": 0.79536521, + "learning_rate": 3.6944245739929e-06, + "loss": 0.81725526, + "num_input_tokens_seen": 35822225, + "step": 1688, + "time_per_iteration": 2.6846063137054443 + }, + { + "auxiliary_loss_clip": 0.01191484, + "auxiliary_loss_mlp": 0.01034414, + "balance_loss_clip": 1.02634144, + "balance_loss_mlp": 1.02472842, + "epoch": 0.20309024228942463, + "flos": 19203374868480.0, + "grad_norm": 2.1772379696876154, + "language_loss": 0.72025597, + "learning_rate": 3.6940106132911332e-06, + "loss": 0.74251497, + "num_input_tokens_seen": 35839410, + "step": 1689, + "time_per_iteration": 2.624595880508423 + }, + { + "auxiliary_loss_clip": 0.01193176, + "auxiliary_loss_mlp": 0.01035959, + "balance_loss_clip": 1.02590322, + "balance_loss_mlp": 1.02626693, + "epoch": 0.20321048518006374, + "flos": 22821087945600.0, + "grad_norm": 1.9152483161317442, + "language_loss": 0.8861931, + "learning_rate": 3.6935963956093037e-06, + "loss": 0.90848446, + "num_input_tokens_seen": 35859495, + "step": 1690, + "time_per_iteration": 2.7023775577545166 + }, + { + "auxiliary_loss_clip": 0.0118247, + "auxiliary_loss_mlp": 0.01032202, + "balance_loss_clip": 1.02369046, + "balance_loss_mlp": 1.02282, + "epoch": 0.20333072807070282, + "flos": 19096397187840.0, + "grad_norm": 4.902385527694705, + "language_loss": 0.69297022, + "learning_rate": 3.6931819210102474e-06, + "loss": 0.71511686, + "num_input_tokens_seen": 35878890, + "step": 1691, + "time_per_iteration": 2.778205156326294 + }, + { + "auxiliary_loss_clip": 0.0121604, + "auxiliary_loss_mlp": 0.01039883, + "balance_loss_clip": 1.06322646, + "balance_loss_mlp": 1.02933335, + "epoch": 0.2034509709613419, + "flos": 18180962144640.0, + "grad_norm": 1.7681391695965634, + "language_loss": 0.84313005, + "learning_rate": 3.6927671895568402e-06, + "loss": 0.86568928, + "num_input_tokens_seen": 35897950, + "step": 1692, + "time_per_iteration": 2.635432720184326 + }, + { + "auxiliary_loss_clip": 0.01215589, + "auxiliary_loss_mlp": 0.01041965, + "balance_loss_clip": 1.06503356, + "balance_loss_mlp": 1.0322082, + "epoch": 0.20357121385198101, + "flos": 22923899648640.0, + "grad_norm": 4.620506000199919, + "language_loss": 0.87187153, + "learning_rate": 3.692352201311996e-06, + "loss": 0.89444709, + "num_input_tokens_seen": 35916800, + "step": 1693, + "time_per_iteration": 2.686387062072754 + }, + { + "auxiliary_loss_clip": 0.01138829, + "auxiliary_loss_mlp": 0.0104247, + "balance_loss_clip": 0.94779885, + "balance_loss_mlp": 1.03246808, + "epoch": 0.2036914567426201, + "flos": 20922131629440.0, + "grad_norm": 1.974455858525869, + "language_loss": 0.76912755, + "learning_rate": 3.6919369563386687e-06, + "loss": 0.79094052, + "num_input_tokens_seen": 35936600, + "step": 1694, + "time_per_iteration": 2.7409613132476807 + }, + { + "auxiliary_loss_clip": 0.01161593, + "auxiliary_loss_mlp": 0.01037209, + "balance_loss_clip": 0.98673677, + "balance_loss_mlp": 1.02802348, + "epoch": 0.20381169963325918, + "flos": 15519155760000.0, + "grad_norm": 2.717494128299555, + "language_loss": 0.78548074, + "learning_rate": 3.69152145469985e-06, + "loss": 0.80746877, + "num_input_tokens_seen": 35953645, + "step": 1695, + "time_per_iteration": 2.672762393951416 + }, + { + "auxiliary_loss_clip": 0.01125034, + "auxiliary_loss_mlp": 0.01034305, + "balance_loss_clip": 0.9082467, + "balance_loss_mlp": 1.02418435, + "epoch": 0.20393194252389826, + "flos": 28833143760000.0, + "grad_norm": 4.277372823717247, + "language_loss": 0.81852412, + "learning_rate": 3.691105696458572e-06, + "loss": 0.84011751, + "num_input_tokens_seen": 35970940, + "step": 1696, + "time_per_iteration": 2.7773590087890625 + }, + { + "auxiliary_loss_clip": 0.01217437, + "auxiliary_loss_mlp": 0.01035975, + "balance_loss_clip": 1.06754112, + "balance_loss_mlp": 1.02632523, + "epoch": 0.20405218541453737, + "flos": 22488554810880.0, + "grad_norm": 2.288982824568413, + "language_loss": 0.67606843, + "learning_rate": 3.690689681677904e-06, + "loss": 0.69860256, + "num_input_tokens_seen": 35989410, + "step": 1697, + "time_per_iteration": 2.5527148246765137 + }, + { + "auxiliary_loss_clip": 0.0117092, + "auxiliary_loss_mlp": 0.01034235, + "balance_loss_clip": 0.98637915, + "balance_loss_mlp": 1.02462649, + "epoch": 0.20417242830517646, + "flos": 25374408278400.0, + "grad_norm": 2.12222547040891, + "language_loss": 0.88750702, + "learning_rate": 3.690273410420956e-06, + "loss": 0.90955853, + "num_input_tokens_seen": 36009175, + "step": 1698, + "time_per_iteration": 2.739593982696533 + }, + { + "auxiliary_loss_clip": 0.01188267, + "auxiliary_loss_mlp": 0.0102848, + "balance_loss_clip": 1.02294421, + "balance_loss_mlp": 1.01899695, + "epoch": 0.20429267119581554, + "flos": 14793078240000.0, + "grad_norm": 2.6906938460757726, + "language_loss": 0.76338375, + "learning_rate": 3.689856882750875e-06, + "loss": 0.78555125, + "num_input_tokens_seen": 36024375, + "step": 1699, + "time_per_iteration": 2.576319456100464 + }, + { + "auxiliary_loss_clip": 0.01191951, + "auxiliary_loss_mlp": 0.01029877, + "balance_loss_clip": 1.02838898, + "balance_loss_mlp": 1.02096057, + "epoch": 0.20441291408645465, + "flos": 17781851151360.0, + "grad_norm": 1.7687957410818183, + "language_loss": 0.78767574, + "learning_rate": 3.6894400987308486e-06, + "loss": 0.80989408, + "num_input_tokens_seen": 36041895, + "step": 1700, + "time_per_iteration": 2.6656367778778076 + }, + { + "auxiliary_loss_clip": 0.01196226, + "auxiliary_loss_mlp": 0.01038518, + "balance_loss_clip": 1.02536166, + "balance_loss_mlp": 1.02799749, + "epoch": 0.20453315697709373, + "flos": 16435668211200.0, + "grad_norm": 1.9238327979416534, + "language_loss": 0.85108507, + "learning_rate": 3.6890230584241024e-06, + "loss": 0.87343258, + "num_input_tokens_seen": 36058825, + "step": 1701, + "time_per_iteration": 2.628002166748047 + }, + { + "auxiliary_loss_clip": 0.01119481, + "auxiliary_loss_mlp": 0.01009765, + "balance_loss_clip": 1.03595185, + "balance_loss_mlp": 1.007285, + "epoch": 0.20465339986773282, + "flos": 66713085653760.0, + "grad_norm": 1.0898212865647225, + "language_loss": 0.66389126, + "learning_rate": 3.6886057618939016e-06, + "loss": 0.6851837, + "num_input_tokens_seen": 36121645, + "step": 1702, + "time_per_iteration": 3.2655458450317383 + }, + { + "auxiliary_loss_clip": 0.01137252, + "auxiliary_loss_mlp": 0.01042373, + "balance_loss_clip": 0.94698226, + "balance_loss_mlp": 1.03230572, + "epoch": 0.2047736427583719, + "flos": 41974114924800.0, + "grad_norm": 1.9946499386227101, + "language_loss": 0.69323903, + "learning_rate": 3.6881882092035492e-06, + "loss": 0.71503532, + "num_input_tokens_seen": 36143030, + "step": 1703, + "time_per_iteration": 2.9309723377227783 + }, + { + "auxiliary_loss_clip": 0.01061756, + "auxiliary_loss_mlp": 0.00504446, + "balance_loss_clip": 0.92538649, + "balance_loss_mlp": 0.0, + "epoch": 0.204893885649011, + "flos": 69940878641280.0, + "grad_norm": 0.9379534733100697, + "language_loss": 0.61246336, + "learning_rate": 3.6877704004163873e-06, + "loss": 0.62812537, + "num_input_tokens_seen": 36203435, + "step": 1704, + "time_per_iteration": 3.4164226055145264 + }, + { + "auxiliary_loss_clip": 0.01219217, + "auxiliary_loss_mlp": 0.01031343, + "balance_loss_clip": 1.06699657, + "balance_loss_mlp": 1.02157927, + "epoch": 0.2050141285396501, + "flos": 22200012858240.0, + "grad_norm": 2.4567038695731473, + "language_loss": 0.77471042, + "learning_rate": 3.6873523355957984e-06, + "loss": 0.79721606, + "num_input_tokens_seen": 36222435, + "step": 1705, + "time_per_iteration": 2.724816083908081 + }, + { + "auxiliary_loss_clip": 0.01116711, + "auxiliary_loss_mlp": 0.01005861, + "balance_loss_clip": 1.0342505, + "balance_loss_mlp": 1.00336921, + "epoch": 0.20513437143028918, + "flos": 46283721730560.0, + "grad_norm": 0.9885123810541747, + "language_loss": 0.64084411, + "learning_rate": 3.686934014805201e-06, + "loss": 0.6620698, + "num_input_tokens_seen": 36273065, + "step": 1706, + "time_per_iteration": 3.0459437370300293 + }, + { + "auxiliary_loss_clip": 0.0119098, + "auxiliary_loss_mlp": 0.01046844, + "balance_loss_clip": 1.02569747, + "balance_loss_mlp": 1.03628767, + "epoch": 0.20525461432092829, + "flos": 21904324099200.0, + "grad_norm": 1.9331488364830076, + "language_loss": 0.81042558, + "learning_rate": 3.6865154381080552e-06, + "loss": 0.83280385, + "num_input_tokens_seen": 36293750, + "step": 1707, + "time_per_iteration": 3.6662490367889404 + }, + { + "auxiliary_loss_clip": 0.01074816, + "auxiliary_loss_mlp": 0.01029346, + "balance_loss_clip": 0.83137226, + "balance_loss_mlp": 1.01994038, + "epoch": 0.20537485721156737, + "flos": 21214264942080.0, + "grad_norm": 1.8313568600863, + "language_loss": 0.82285899, + "learning_rate": 3.6860966055678585e-06, + "loss": 0.84390068, + "num_input_tokens_seen": 36310105, + "step": 1708, + "time_per_iteration": 2.8174233436584473 + }, + { + "auxiliary_loss_clip": 0.01191607, + "auxiliary_loss_mlp": 0.01038971, + "balance_loss_clip": 1.02604294, + "balance_loss_mlp": 1.02879024, + "epoch": 0.20549510010220645, + "flos": 20191205773440.0, + "grad_norm": 1.7731084705514784, + "language_loss": 0.86384475, + "learning_rate": 3.685677517248147e-06, + "loss": 0.88615054, + "num_input_tokens_seen": 36328995, + "step": 1709, + "time_per_iteration": 2.6459240913391113 + }, + { + "auxiliary_loss_clip": 0.01172516, + "auxiliary_loss_mlp": 0.00511921, + "balance_loss_clip": 0.99352813, + "balance_loss_mlp": 0.0, + "epoch": 0.20561534299284553, + "flos": 17016702612480.0, + "grad_norm": 2.2650547181153846, + "language_loss": 0.80617821, + "learning_rate": 3.6852581732124967e-06, + "loss": 0.8230226, + "num_input_tokens_seen": 36346340, + "step": 1710, + "time_per_iteration": 3.6140177249908447 + }, + { + "auxiliary_loss_clip": 0.0119918, + "auxiliary_loss_mlp": 0.01046443, + "balance_loss_clip": 1.02995217, + "balance_loss_mlp": 1.03603613, + "epoch": 0.20573558588348465, + "flos": 22890467064960.0, + "grad_norm": 2.8512573790535805, + "language_loss": 0.76105011, + "learning_rate": 3.6848385735245213e-06, + "loss": 0.78350633, + "num_input_tokens_seen": 36365430, + "step": 1711, + "time_per_iteration": 3.568014621734619 + }, + { + "auxiliary_loss_clip": 0.01180312, + "auxiliary_loss_mlp": 0.01034002, + "balance_loss_clip": 1.02251506, + "balance_loss_mlp": 1.02459037, + "epoch": 0.20585582877412373, + "flos": 24643123286400.0, + "grad_norm": 1.8928854536605615, + "language_loss": 0.85939562, + "learning_rate": 3.6844187182478734e-06, + "loss": 0.88153875, + "num_input_tokens_seen": 36386285, + "step": 1712, + "time_per_iteration": 2.6687560081481934 + }, + { + "auxiliary_loss_clip": 0.0115963, + "auxiliary_loss_mlp": 0.01031991, + "balance_loss_clip": 0.98535371, + "balance_loss_mlp": 1.02218604, + "epoch": 0.2059760716647628, + "flos": 24206952435840.0, + "grad_norm": 1.6049468499705126, + "language_loss": 0.74994302, + "learning_rate": 3.683998607446246e-06, + "loss": 0.77185923, + "num_input_tokens_seen": 36404935, + "step": 1713, + "time_per_iteration": 2.7380447387695312 + }, + { + "auxiliary_loss_clip": 0.01194145, + "auxiliary_loss_mlp": 0.01041662, + "balance_loss_clip": 1.02756798, + "balance_loss_mlp": 1.03246522, + "epoch": 0.20609631455540192, + "flos": 20229522606720.0, + "grad_norm": 1.9145321245632907, + "language_loss": 0.75101781, + "learning_rate": 3.6835782411833686e-06, + "loss": 0.77337587, + "num_input_tokens_seen": 36424455, + "step": 1714, + "time_per_iteration": 2.625387191772461 + }, + { + "auxiliary_loss_clip": 0.01132479, + "auxiliary_loss_mlp": 0.01038964, + "balance_loss_clip": 0.94601178, + "balance_loss_mlp": 1.02856922, + "epoch": 0.206216557446041, + "flos": 19864957518720.0, + "grad_norm": 1.818408952162722, + "language_loss": 0.73976648, + "learning_rate": 3.68315761952301e-06, + "loss": 0.76148087, + "num_input_tokens_seen": 36441685, + "step": 1715, + "time_per_iteration": 2.7270803451538086 + }, + { + "auxiliary_loss_clip": 0.01219816, + "auxiliary_loss_mlp": 0.01037243, + "balance_loss_clip": 1.06773138, + "balance_loss_mlp": 1.02733099, + "epoch": 0.2063368003366801, + "flos": 24096311568000.0, + "grad_norm": 2.0632329407380094, + "language_loss": 0.82819462, + "learning_rate": 3.6827367425289797e-06, + "loss": 0.85076523, + "num_input_tokens_seen": 36461460, + "step": 1716, + "time_per_iteration": 2.612651824951172 + }, + { + "auxiliary_loss_clip": 0.01173607, + "auxiliary_loss_mlp": 0.01036598, + "balance_loss_clip": 0.99018145, + "balance_loss_mlp": 1.02630997, + "epoch": 0.2064570432273192, + "flos": 20340163474560.0, + "grad_norm": 2.521991925957483, + "language_loss": 0.72693104, + "learning_rate": 3.6823156102651225e-06, + "loss": 0.74903309, + "num_input_tokens_seen": 36479615, + "step": 1717, + "time_per_iteration": 2.769141674041748 + }, + { + "auxiliary_loss_clip": 0.01086507, + "auxiliary_loss_mlp": 0.01037575, + "balance_loss_clip": 0.87242687, + "balance_loss_mlp": 1.02722216, + "epoch": 0.20657728611795828, + "flos": 20520363029760.0, + "grad_norm": 2.011682214168445, + "language_loss": 0.71225822, + "learning_rate": 3.6818942227953257e-06, + "loss": 0.73349905, + "num_input_tokens_seen": 36500160, + "step": 1718, + "time_per_iteration": 2.929427146911621 + }, + { + "auxiliary_loss_clip": 0.01145836, + "auxiliary_loss_mlp": 0.01040375, + "balance_loss_clip": 0.94967663, + "balance_loss_mlp": 1.03095722, + "epoch": 0.20669752900859736, + "flos": 21799285752960.0, + "grad_norm": 1.8552635568726812, + "language_loss": 0.68939555, + "learning_rate": 3.681472580183512e-06, + "loss": 0.7112577, + "num_input_tokens_seen": 36518810, + "step": 1719, + "time_per_iteration": 2.734294891357422 + }, + { + "auxiliary_loss_clip": 0.0118847, + "auxiliary_loss_mlp": 0.01035361, + "balance_loss_clip": 1.02639937, + "balance_loss_mlp": 1.02533579, + "epoch": 0.20681777189923645, + "flos": 15122020014720.0, + "grad_norm": 1.790830418757836, + "language_loss": 0.86309987, + "learning_rate": 3.6810506824936455e-06, + "loss": 0.88533819, + "num_input_tokens_seen": 36536890, + "step": 1720, + "time_per_iteration": 2.676379680633545 + }, + { + "auxiliary_loss_clip": 0.01073078, + "auxiliary_loss_mlp": 0.01003926, + "balance_loss_clip": 0.95735735, + "balance_loss_mlp": 1.00142241, + "epoch": 0.20693801478987556, + "flos": 56481021509760.0, + "grad_norm": 1.044614744561015, + "language_loss": 0.62608814, + "learning_rate": 3.680628529789726e-06, + "loss": 0.64685816, + "num_input_tokens_seen": 36589300, + "step": 1721, + "time_per_iteration": 3.088542938232422 + }, + { + "auxiliary_loss_clip": 0.01221342, + "auxiliary_loss_mlp": 0.01040769, + "balance_loss_clip": 1.06591129, + "balance_loss_mlp": 1.0299921, + "epoch": 0.20705825768051464, + "flos": 21614201948160.0, + "grad_norm": 1.8178677086969812, + "language_loss": 0.86481261, + "learning_rate": 3.680206122135796e-06, + "loss": 0.88743377, + "num_input_tokens_seen": 36609905, + "step": 1722, + "time_per_iteration": 2.6774895191192627 + }, + { + "auxiliary_loss_clip": 0.01128475, + "auxiliary_loss_mlp": 0.01044453, + "balance_loss_clip": 0.91334498, + "balance_loss_mlp": 1.03526235, + "epoch": 0.20717850057115372, + "flos": 25848895962240.0, + "grad_norm": 2.0518281840076216, + "language_loss": 0.78449637, + "learning_rate": 3.6797834595959323e-06, + "loss": 0.80622566, + "num_input_tokens_seen": 36629805, + "step": 1723, + "time_per_iteration": 2.7751874923706055 + }, + { + "auxiliary_loss_clip": 0.01108912, + "auxiliary_loss_mlp": 0.01037809, + "balance_loss_clip": 0.90599889, + "balance_loss_mlp": 1.02724135, + "epoch": 0.20729874346179283, + "flos": 29130807767040.0, + "grad_norm": 4.802385637943972, + "language_loss": 0.78058362, + "learning_rate": 3.679360542234254e-06, + "loss": 0.80205089, + "num_input_tokens_seen": 36649150, + "step": 1724, + "time_per_iteration": 2.801703691482544 + }, + { + "auxiliary_loss_clip": 0.01159604, + "auxiliary_loss_mlp": 0.00511447, + "balance_loss_clip": 0.98316097, + "balance_loss_mlp": 0.0, + "epoch": 0.20741898635243192, + "flos": 29023363209600.0, + "grad_norm": 1.9864262637804007, + "language_loss": 0.72620451, + "learning_rate": 3.678937370114916e-06, + "loss": 0.74291503, + "num_input_tokens_seen": 36668955, + "step": 1725, + "time_per_iteration": 2.7529609203338623 + }, + { + "auxiliary_loss_clip": 0.01165202, + "auxiliary_loss_mlp": 0.01027627, + "balance_loss_clip": 0.98908895, + "balance_loss_mlp": 1.01792884, + "epoch": 0.207539229243071, + "flos": 15559447841280.0, + "grad_norm": 1.9784719170089264, + "language_loss": 0.78788567, + "learning_rate": 3.678513943302114e-06, + "loss": 0.80981398, + "num_input_tokens_seen": 36685730, + "step": 1726, + "time_per_iteration": 2.7148220539093018 + }, + { + "auxiliary_loss_clip": 0.0121749, + "auxiliary_loss_mlp": 0.01034594, + "balance_loss_clip": 1.06591034, + "balance_loss_mlp": 1.02444327, + "epoch": 0.20765947213371008, + "flos": 20521081301760.0, + "grad_norm": 1.699473645838995, + "language_loss": 0.85000443, + "learning_rate": 3.678090261860082e-06, + "loss": 0.87252527, + "num_input_tokens_seen": 36705460, + "step": 1727, + "time_per_iteration": 2.6550979614257812 + }, + { + "auxiliary_loss_clip": 0.01143569, + "auxiliary_loss_mlp": 0.01035258, + "balance_loss_clip": 0.94377697, + "balance_loss_mlp": 1.02603722, + "epoch": 0.2077797150243492, + "flos": 19354415558400.0, + "grad_norm": 2.113903286717198, + "language_loss": 0.77497566, + "learning_rate": 3.6776663258530906e-06, + "loss": 0.7967639, + "num_input_tokens_seen": 36724110, + "step": 1728, + "time_per_iteration": 2.7187159061431885 + }, + { + "auxiliary_loss_clip": 0.01194016, + "auxiliary_loss_mlp": 0.01040728, + "balance_loss_clip": 1.02618372, + "balance_loss_mlp": 1.03117919, + "epoch": 0.20789995791498828, + "flos": 21829952989440.0, + "grad_norm": 1.7196504382544626, + "language_loss": 0.71230131, + "learning_rate": 3.6772421353454516e-06, + "loss": 0.73464876, + "num_input_tokens_seen": 36742705, + "step": 1729, + "time_per_iteration": 2.7024028301239014 + }, + { + "auxiliary_loss_clip": 0.01188012, + "auxiliary_loss_mlp": 0.01035902, + "balance_loss_clip": 1.02541029, + "balance_loss_mlp": 1.02614474, + "epoch": 0.20802020080562736, + "flos": 23148844571520.0, + "grad_norm": 3.1816993196256256, + "language_loss": 0.88499808, + "learning_rate": 3.6768176904015153e-06, + "loss": 0.90723729, + "num_input_tokens_seen": 36762510, + "step": 1730, + "time_per_iteration": 2.6935620307922363 + }, + { + "auxiliary_loss_clip": 0.01189592, + "auxiliary_loss_mlp": 0.01032816, + "balance_loss_clip": 1.02238262, + "balance_loss_mlp": 1.02311206, + "epoch": 0.20814044369626647, + "flos": 23072677781760.0, + "grad_norm": 2.0769806124228993, + "language_loss": 0.60393947, + "learning_rate": 3.6763929910856674e-06, + "loss": 0.62616348, + "num_input_tokens_seen": 36780960, + "step": 1731, + "time_per_iteration": 2.7678914070129395 + }, + { + "auxiliary_loss_clip": 0.01189979, + "auxiliary_loss_mlp": 0.0103363, + "balance_loss_clip": 1.02572656, + "balance_loss_mlp": 1.02439761, + "epoch": 0.20826068658690555, + "flos": 19608016556160.0, + "grad_norm": 2.5980089655601337, + "language_loss": 0.77539551, + "learning_rate": 3.6759680374623365e-06, + "loss": 0.7976315, + "num_input_tokens_seen": 36798875, + "step": 1732, + "time_per_iteration": 2.6030120849609375 + }, + { + "auxiliary_loss_clip": 0.0121413, + "auxiliary_loss_mlp": 0.01037975, + "balance_loss_clip": 1.06523323, + "balance_loss_mlp": 1.02773476, + "epoch": 0.20838092947754464, + "flos": 25374049142400.0, + "grad_norm": 2.135864558654562, + "language_loss": 0.75313365, + "learning_rate": 3.675542829595986e-06, + "loss": 0.77565467, + "num_input_tokens_seen": 36818540, + "step": 1733, + "time_per_iteration": 3.63076114654541 + }, + { + "auxiliary_loss_clip": 0.01169535, + "auxiliary_loss_mlp": 0.01034138, + "balance_loss_clip": 0.9879775, + "balance_loss_mlp": 1.02497625, + "epoch": 0.20850117236818372, + "flos": 24061729749120.0, + "grad_norm": 1.4986491871876482, + "language_loss": 0.796417, + "learning_rate": 3.6751173675511213e-06, + "loss": 0.81845379, + "num_input_tokens_seen": 36840585, + "step": 1734, + "time_per_iteration": 2.751126527786255 + }, + { + "auxiliary_loss_clip": 0.01162272, + "auxiliary_loss_mlp": 0.01034416, + "balance_loss_clip": 0.9819963, + "balance_loss_mlp": 1.02454543, + "epoch": 0.20862141525882283, + "flos": 20077799558400.0, + "grad_norm": 2.0890161327679806, + "language_loss": 0.87487215, + "learning_rate": 3.674691651392283e-06, + "loss": 0.89683902, + "num_input_tokens_seen": 36858255, + "step": 1735, + "time_per_iteration": 2.733452796936035 + }, + { + "auxiliary_loss_clip": 0.01172368, + "auxiliary_loss_mlp": 0.01038833, + "balance_loss_clip": 0.98915035, + "balance_loss_mlp": 1.02964795, + "epoch": 0.2087416581494619, + "flos": 39015183237120.0, + "grad_norm": 2.01114466559919, + "language_loss": 0.75442851, + "learning_rate": 3.674265681184053e-06, + "loss": 0.77654052, + "num_input_tokens_seen": 36881515, + "step": 1736, + "time_per_iteration": 3.830491304397583 + }, + { + "auxiliary_loss_clip": 0.01167886, + "auxiliary_loss_mlp": 0.01035762, + "balance_loss_clip": 0.98625398, + "balance_loss_mlp": 1.02599239, + "epoch": 0.208861901040101, + "flos": 26101994169600.0, + "grad_norm": 1.6908370648264321, + "language_loss": 0.86588579, + "learning_rate": 3.6738394569910504e-06, + "loss": 0.88792229, + "num_input_tokens_seen": 36902055, + "step": 1737, + "time_per_iteration": 3.787935733795166 + }, + { + "auxiliary_loss_clip": 0.01192658, + "auxiliary_loss_mlp": 0.01030185, + "balance_loss_clip": 1.02770281, + "balance_loss_mlp": 1.0203681, + "epoch": 0.2089821439307401, + "flos": 28398732675840.0, + "grad_norm": 2.7630741048567753, + "language_loss": 0.82739556, + "learning_rate": 3.6734129788779333e-06, + "loss": 0.84962404, + "num_input_tokens_seen": 36921230, + "step": 1738, + "time_per_iteration": 2.6993725299835205 + }, + { + "auxiliary_loss_clip": 0.0114425, + "auxiliary_loss_mlp": 0.01034201, + "balance_loss_clip": 0.95104986, + "balance_loss_mlp": 1.02523637, + "epoch": 0.2091023868213792, + "flos": 21069616872960.0, + "grad_norm": 1.8481140948170474, + "language_loss": 0.90165973, + "learning_rate": 3.6729862469093976e-06, + "loss": 0.92344421, + "num_input_tokens_seen": 36940325, + "step": 1739, + "time_per_iteration": 2.6994988918304443 + }, + { + "auxiliary_loss_clip": 0.01154207, + "auxiliary_loss_mlp": 0.01033906, + "balance_loss_clip": 0.98477209, + "balance_loss_mlp": 1.02464986, + "epoch": 0.20922262971201827, + "flos": 22455481363200.0, + "grad_norm": 2.3754036503195928, + "language_loss": 0.82592058, + "learning_rate": 3.6725592611501782e-06, + "loss": 0.84780169, + "num_input_tokens_seen": 36959000, + "step": 1740, + "time_per_iteration": 2.705933094024658 + }, + { + "auxiliary_loss_clip": 0.01191167, + "auxiliary_loss_mlp": 0.01038639, + "balance_loss_clip": 1.02485526, + "balance_loss_mlp": 1.02932286, + "epoch": 0.20934287260265738, + "flos": 27852244179840.0, + "grad_norm": 2.1011297394554034, + "language_loss": 0.76509506, + "learning_rate": 3.6721320216650496e-06, + "loss": 0.78739309, + "num_input_tokens_seen": 36979615, + "step": 1741, + "time_per_iteration": 2.7632248401641846 + }, + { + "auxiliary_loss_clip": 0.011705, + "auxiliary_loss_mlp": 0.01032794, + "balance_loss_clip": 0.98919594, + "balance_loss_mlp": 1.0227387, + "epoch": 0.20946311549329646, + "flos": 16435309075200.0, + "grad_norm": 1.9201279710660004, + "language_loss": 0.83650255, + "learning_rate": 3.6717045285188215e-06, + "loss": 0.85853553, + "num_input_tokens_seen": 36997310, + "step": 1742, + "time_per_iteration": 2.70310115814209 + }, + { + "auxiliary_loss_clip": 0.01105875, + "auxiliary_loss_mlp": 0.01043635, + "balance_loss_clip": 0.90538228, + "balance_loss_mlp": 1.03324032, + "epoch": 0.20958335838393555, + "flos": 22492720788480.0, + "grad_norm": 2.077645395322963, + "language_loss": 0.87174886, + "learning_rate": 3.671276781776346e-06, + "loss": 0.89324397, + "num_input_tokens_seen": 37015965, + "step": 1743, + "time_per_iteration": 2.8153748512268066 + }, + { + "auxiliary_loss_clip": 0.01149023, + "auxiliary_loss_mlp": 0.01031338, + "balance_loss_clip": 0.9459331, + "balance_loss_mlp": 1.02186048, + "epoch": 0.20970360127457463, + "flos": 25224768218880.0, + "grad_norm": 2.2422676397648966, + "language_loss": 0.67501855, + "learning_rate": 3.6708487815025128e-06, + "loss": 0.69682217, + "num_input_tokens_seen": 37036545, + "step": 1744, + "time_per_iteration": 2.7208118438720703 + }, + { + "auxiliary_loss_clip": 0.01143876, + "auxiliary_loss_mlp": 0.01035886, + "balance_loss_clip": 0.94731659, + "balance_loss_mlp": 1.02596748, + "epoch": 0.20982384416521374, + "flos": 18479164855680.0, + "grad_norm": 2.1872859247956327, + "language_loss": 0.74156684, + "learning_rate": 3.6704205277622463e-06, + "loss": 0.76336443, + "num_input_tokens_seen": 37054985, + "step": 1745, + "time_per_iteration": 2.751131057739258 + }, + { + "auxiliary_loss_clip": 0.01168852, + "auxiliary_loss_mlp": 0.0103052, + "balance_loss_clip": 0.98456454, + "balance_loss_mlp": 1.02208018, + "epoch": 0.20994408705585282, + "flos": 25373546352000.0, + "grad_norm": 1.7378673273008645, + "language_loss": 0.80815029, + "learning_rate": 3.6699920206205146e-06, + "loss": 0.83014399, + "num_input_tokens_seen": 37075725, + "step": 1746, + "time_per_iteration": 2.7711238861083984 + }, + { + "auxiliary_loss_clip": 0.01189714, + "auxiliary_loss_mlp": 0.01032624, + "balance_loss_clip": 1.0222733, + "balance_loss_mlp": 1.02375519, + "epoch": 0.2100643299464919, + "flos": 21320955313920.0, + "grad_norm": 1.6997862226156328, + "language_loss": 0.8180635, + "learning_rate": 3.669563260142321e-06, + "loss": 0.84028691, + "num_input_tokens_seen": 37094615, + "step": 1747, + "time_per_iteration": 2.721102237701416 + }, + { + "auxiliary_loss_clip": 0.01160068, + "auxiliary_loss_mlp": 0.01037721, + "balance_loss_clip": 0.98555148, + "balance_loss_mlp": 1.02770114, + "epoch": 0.21018457283713102, + "flos": 19354379644800.0, + "grad_norm": 2.6351976408749573, + "language_loss": 0.8433339, + "learning_rate": 3.6691342463927083e-06, + "loss": 0.86531174, + "num_input_tokens_seen": 37113610, + "step": 1748, + "time_per_iteration": 2.755178928375244 + }, + { + "auxiliary_loss_clip": 0.01142054, + "auxiliary_loss_mlp": 0.01037967, + "balance_loss_clip": 0.94504046, + "balance_loss_mlp": 1.02849555, + "epoch": 0.2103048157277701, + "flos": 28330035914880.0, + "grad_norm": 1.8286500476538565, + "language_loss": 0.8161422, + "learning_rate": 3.668704979436758e-06, + "loss": 0.83794242, + "num_input_tokens_seen": 37133705, + "step": 1749, + "time_per_iteration": 2.7373557090759277 + }, + { + "auxiliary_loss_clip": 0.01161447, + "auxiliary_loss_mlp": 0.0104527, + "balance_loss_clip": 0.98449683, + "balance_loss_mlp": 1.03538775, + "epoch": 0.21042505861840918, + "flos": 17457290835840.0, + "grad_norm": 1.8263884377861133, + "language_loss": 0.78891075, + "learning_rate": 3.668275459339588e-06, + "loss": 0.81097788, + "num_input_tokens_seen": 37152185, + "step": 1750, + "time_per_iteration": 2.7326183319091797 + }, + { + "auxiliary_loss_clip": 0.01215717, + "auxiliary_loss_mlp": 0.01042373, + "balance_loss_clip": 1.06608486, + "balance_loss_mlp": 1.03262758, + "epoch": 0.21054530150904827, + "flos": 14209817195520.0, + "grad_norm": 2.051181074863719, + "language_loss": 0.80467784, + "learning_rate": 3.667845686166358e-06, + "loss": 0.82725877, + "num_input_tokens_seen": 37169110, + "step": 1751, + "time_per_iteration": 2.59134840965271 + }, + { + "auxiliary_loss_clip": 0.01127012, + "auxiliary_loss_mlp": 0.01033717, + "balance_loss_clip": 0.94563198, + "balance_loss_mlp": 1.02391779, + "epoch": 0.21066554439968738, + "flos": 18618210403200.0, + "grad_norm": 1.77824369931516, + "language_loss": 0.85804141, + "learning_rate": 3.6674156599822634e-06, + "loss": 0.87964869, + "num_input_tokens_seen": 37184905, + "step": 1752, + "time_per_iteration": 2.7255983352661133 + }, + { + "auxiliary_loss_clip": 0.0112352, + "auxiliary_loss_mlp": 0.01036626, + "balance_loss_clip": 0.90840608, + "balance_loss_mlp": 1.02645707, + "epoch": 0.21078578729032646, + "flos": 23658883741440.0, + "grad_norm": 2.155654058929086, + "language_loss": 0.82238007, + "learning_rate": 3.666985380852539e-06, + "loss": 0.8439815, + "num_input_tokens_seen": 37203910, + "step": 1753, + "time_per_iteration": 2.7906272411346436 + }, + { + "auxiliary_loss_clip": 0.01166425, + "auxiliary_loss_mlp": 0.01033788, + "balance_loss_clip": 0.98751187, + "balance_loss_mlp": 1.02394128, + "epoch": 0.21090603018096554, + "flos": 29346379240320.0, + "grad_norm": 2.2918489730090448, + "language_loss": 0.743729, + "learning_rate": 3.6665548488424576e-06, + "loss": 0.76573116, + "num_input_tokens_seen": 37222670, + "step": 1754, + "time_per_iteration": 2.7311151027679443 + }, + { + "auxiliary_loss_clip": 0.01214612, + "auxiliary_loss_mlp": 0.01041784, + "balance_loss_clip": 1.06286979, + "balance_loss_mlp": 1.03159189, + "epoch": 0.21102627307160465, + "flos": 23261245205760.0, + "grad_norm": 2.7086319661452998, + "language_loss": 0.87836397, + "learning_rate": 3.6661240640173307e-06, + "loss": 0.9009279, + "num_input_tokens_seen": 37244140, + "step": 1755, + "time_per_iteration": 2.6248373985290527 + }, + { + "auxiliary_loss_clip": 0.01054069, + "auxiliary_loss_mlp": 0.01010417, + "balance_loss_clip": 0.92309678, + "balance_loss_mlp": 1.0079968, + "epoch": 0.21114651596224374, + "flos": 54633454577280.0, + "grad_norm": 0.8691240105644509, + "language_loss": 0.57904506, + "learning_rate": 3.6656930264425085e-06, + "loss": 0.59968984, + "num_input_tokens_seen": 37308185, + "step": 1756, + "time_per_iteration": 3.3212099075317383 + }, + { + "auxiliary_loss_clip": 0.01215302, + "auxiliary_loss_mlp": 0.01033382, + "balance_loss_clip": 1.06436312, + "balance_loss_mlp": 1.02353573, + "epoch": 0.21126675885288282, + "flos": 21543314457600.0, + "grad_norm": 1.9376460524792967, + "language_loss": 0.75367159, + "learning_rate": 3.665261736183378e-06, + "loss": 0.77615845, + "num_input_tokens_seen": 37328220, + "step": 1757, + "time_per_iteration": 2.7015492916107178 + }, + { + "auxiliary_loss_clip": 0.01143584, + "auxiliary_loss_mlp": 0.01041302, + "balance_loss_clip": 0.94935077, + "balance_loss_mlp": 1.03028095, + "epoch": 0.2113870017435219, + "flos": 10961876678400.0, + "grad_norm": 2.347732226980596, + "language_loss": 0.88644803, + "learning_rate": 3.664830193305366e-06, + "loss": 0.90829688, + "num_input_tokens_seen": 37345995, + "step": 1758, + "time_per_iteration": 2.6808152198791504 + }, + { + "auxiliary_loss_clip": 0.011389, + "auxiliary_loss_mlp": 0.0104184, + "balance_loss_clip": 0.94560146, + "balance_loss_mlp": 1.03169513, + "epoch": 0.211507244634161, + "flos": 16653825463680.0, + "grad_norm": 3.726113441944849, + "language_loss": 0.77306473, + "learning_rate": 3.6643983978739373e-06, + "loss": 0.79487216, + "num_input_tokens_seen": 37362610, + "step": 1759, + "time_per_iteration": 3.6388256549835205 + }, + { + "auxiliary_loss_clip": 0.01164386, + "auxiliary_loss_mlp": 0.01042067, + "balance_loss_clip": 0.99079788, + "balance_loss_mlp": 1.03211355, + "epoch": 0.2116274875248001, + "flos": 20954091755520.0, + "grad_norm": 2.587076529053219, + "language_loss": 0.81663167, + "learning_rate": 3.663966349954596e-06, + "loss": 0.83869618, + "num_input_tokens_seen": 37382790, + "step": 1760, + "time_per_iteration": 2.7482893466949463 + }, + { + "auxiliary_loss_clip": 0.01096943, + "auxiliary_loss_mlp": 0.01005538, + "balance_loss_clip": 0.99742329, + "balance_loss_mlp": 1.00323772, + "epoch": 0.21174773041543918, + "flos": 68196949424640.0, + "grad_norm": 0.7893740154422618, + "language_loss": 0.59701812, + "learning_rate": 3.6635340496128816e-06, + "loss": 0.61804295, + "num_input_tokens_seen": 37439720, + "step": 1761, + "time_per_iteration": 3.216573476791382 + }, + { + "auxiliary_loss_clip": 0.01119022, + "auxiliary_loss_mlp": 0.0104197, + "balance_loss_clip": 0.91086364, + "balance_loss_mlp": 1.03231382, + "epoch": 0.2118679733060783, + "flos": 20668315150080.0, + "grad_norm": 1.669013784644629, + "language_loss": 0.92757994, + "learning_rate": 3.6631014969143747e-06, + "loss": 0.9491899, + "num_input_tokens_seen": 37459410, + "step": 1762, + "time_per_iteration": 3.7048962116241455 + }, + { + "auxiliary_loss_clip": 0.01190658, + "auxiliary_loss_mlp": 0.01032812, + "balance_loss_clip": 1.02689981, + "balance_loss_mlp": 1.02267337, + "epoch": 0.21198821619671737, + "flos": 23223431162880.0, + "grad_norm": 1.75233637485541, + "language_loss": 0.88689971, + "learning_rate": 3.662668691924693e-06, + "loss": 0.90913445, + "num_input_tokens_seen": 37480460, + "step": 1763, + "time_per_iteration": 2.6781160831451416 + }, + { + "auxiliary_loss_clip": 0.01141585, + "auxiliary_loss_mlp": 0.01044235, + "balance_loss_clip": 0.94740969, + "balance_loss_mlp": 1.03444767, + "epoch": 0.21210845908735645, + "flos": 24498547044480.0, + "grad_norm": 2.0678493417443176, + "language_loss": 0.71749032, + "learning_rate": 3.6622356347094927e-06, + "loss": 0.73934853, + "num_input_tokens_seen": 37502025, + "step": 1764, + "time_per_iteration": 3.7256782054901123 + }, + { + "auxiliary_loss_clip": 0.01142931, + "auxiliary_loss_mlp": 0.01030755, + "balance_loss_clip": 0.94574827, + "balance_loss_mlp": 1.01979339, + "epoch": 0.21222870197799554, + "flos": 27089789160960.0, + "grad_norm": 1.7660109345477237, + "language_loss": 0.79057181, + "learning_rate": 3.6618023253344684e-06, + "loss": 0.81230861, + "num_input_tokens_seen": 37520885, + "step": 1765, + "time_per_iteration": 2.7743752002716064 + }, + { + "auxiliary_loss_clip": 0.0119007, + "auxiliary_loss_mlp": 0.0103419, + "balance_loss_clip": 1.02404594, + "balance_loss_mlp": 1.0234077, + "epoch": 0.21234894486863465, + "flos": 16873850223360.0, + "grad_norm": 1.5723371699148267, + "language_loss": 0.83666086, + "learning_rate": 3.6613687638653527e-06, + "loss": 0.85890341, + "num_input_tokens_seen": 37539055, + "step": 1766, + "time_per_iteration": 2.586420774459839 + }, + { + "auxiliary_loss_clip": 0.01167779, + "auxiliary_loss_mlp": 0.01031024, + "balance_loss_clip": 0.98978788, + "balance_loss_mlp": 1.02104068, + "epoch": 0.21246918775927373, + "flos": 23474949171840.0, + "grad_norm": 1.8291663995175191, + "language_loss": 0.77878052, + "learning_rate": 3.660934950367916e-06, + "loss": 0.80076855, + "num_input_tokens_seen": 37558300, + "step": 1767, + "time_per_iteration": 2.7472403049468994 + }, + { + "auxiliary_loss_clip": 0.01192729, + "auxiliary_loss_mlp": 0.01040976, + "balance_loss_clip": 1.02503538, + "balance_loss_mlp": 1.03102183, + "epoch": 0.21258943064991281, + "flos": 22382295402240.0, + "grad_norm": 1.7367577036934831, + "language_loss": 0.83545429, + "learning_rate": 3.660500884907968e-06, + "loss": 0.8577913, + "num_input_tokens_seen": 37579040, + "step": 1768, + "time_per_iteration": 2.6416449546813965 + }, + { + "auxiliary_loss_clip": 0.01030822, + "auxiliary_loss_mlp": 0.01003082, + "balance_loss_clip": 0.91206378, + "balance_loss_mlp": 1.00034034, + "epoch": 0.21270967354055192, + "flos": 59440168679040.0, + "grad_norm": 0.8209930954070639, + "language_loss": 0.60022211, + "learning_rate": 3.660066567551356e-06, + "loss": 0.62056118, + "num_input_tokens_seen": 37639185, + "step": 1769, + "time_per_iteration": 3.2371556758880615 + }, + { + "auxiliary_loss_clip": 0.01190877, + "auxiliary_loss_mlp": 0.0051281, + "balance_loss_clip": 1.02645612, + "balance_loss_mlp": 0.0, + "epoch": 0.212829916431191, + "flos": 21544032729600.0, + "grad_norm": 4.348624547548089, + "language_loss": 0.83997619, + "learning_rate": 3.6596319983639657e-06, + "loss": 0.85701305, + "num_input_tokens_seen": 37657765, + "step": 1770, + "time_per_iteration": 2.767010450363159 + }, + { + "auxiliary_loss_clip": 0.01142489, + "auxiliary_loss_mlp": 0.00513188, + "balance_loss_clip": 0.94811976, + "balance_loss_mlp": 0.0, + "epoch": 0.2129501593218301, + "flos": 28987739896320.0, + "grad_norm": 1.5699769893831443, + "language_loss": 0.86459184, + "learning_rate": 3.6591971774117214e-06, + "loss": 0.88114858, + "num_input_tokens_seen": 37680740, + "step": 1771, + "time_per_iteration": 2.7609522342681885 + }, + { + "auxiliary_loss_clip": 0.01193329, + "auxiliary_loss_mlp": 0.0104133, + "balance_loss_clip": 1.02504385, + "balance_loss_mlp": 1.0304637, + "epoch": 0.2130704022124692, + "flos": 18806993308800.0, + "grad_norm": 3.0998077049424624, + "language_loss": 0.80361223, + "learning_rate": 3.6587621047605833e-06, + "loss": 0.82595885, + "num_input_tokens_seen": 37697910, + "step": 1772, + "time_per_iteration": 2.6668429374694824 + }, + { + "auxiliary_loss_clip": 0.01191341, + "auxiliary_loss_mlp": 0.01044929, + "balance_loss_clip": 1.02645373, + "balance_loss_mlp": 1.03519559, + "epoch": 0.21319064510310828, + "flos": 13918150759680.0, + "grad_norm": 2.0709816126139273, + "language_loss": 0.86681426, + "learning_rate": 3.6583267804765542e-06, + "loss": 0.88917696, + "num_input_tokens_seen": 37712245, + "step": 1773, + "time_per_iteration": 2.6208856105804443 + }, + { + "auxiliary_loss_clip": 0.01190338, + "auxiliary_loss_mlp": 0.01027148, + "balance_loss_clip": 1.02615106, + "balance_loss_mlp": 1.01704526, + "epoch": 0.21331088799374737, + "flos": 20959694277120.0, + "grad_norm": 1.7629090546067756, + "language_loss": 0.85742527, + "learning_rate": 3.6578912046256702e-06, + "loss": 0.87960011, + "num_input_tokens_seen": 37730765, + "step": 1774, + "time_per_iteration": 2.669685125350952 + }, + { + "auxiliary_loss_clip": 0.01139468, + "auxiliary_loss_mlp": 0.01037238, + "balance_loss_clip": 0.94530612, + "balance_loss_mlp": 1.02705765, + "epoch": 0.21343113088438645, + "flos": 18624638937600.0, + "grad_norm": 1.985174516164646, + "language_loss": 0.75795746, + "learning_rate": 3.6574553772740083e-06, + "loss": 0.77972454, + "num_input_tokens_seen": 37748695, + "step": 1775, + "time_per_iteration": 2.6378839015960693 + }, + { + "auxiliary_loss_clip": 0.01105257, + "auxiliary_loss_mlp": 0.01016988, + "balance_loss_clip": 1.01213121, + "balance_loss_mlp": 1.01455593, + "epoch": 0.21355137377502556, + "flos": 67413128791680.0, + "grad_norm": 0.8469088943032013, + "language_loss": 0.61797619, + "learning_rate": 3.657019298487684e-06, + "loss": 0.63919866, + "num_input_tokens_seen": 37813705, + "step": 1776, + "time_per_iteration": 3.2481460571289062 + }, + { + "auxiliary_loss_clip": 0.01193735, + "auxiliary_loss_mlp": 0.00513609, + "balance_loss_clip": 1.02458549, + "balance_loss_mlp": 0.0, + "epoch": 0.21367161666566464, + "flos": 34532095697280.0, + "grad_norm": 1.973127442506364, + "language_loss": 0.83747768, + "learning_rate": 3.6565829683328495e-06, + "loss": 0.8545512, + "num_input_tokens_seen": 37836330, + "step": 1777, + "time_per_iteration": 2.7559080123901367 + }, + { + "auxiliary_loss_clip": 0.01185529, + "auxiliary_loss_mlp": 0.01037539, + "balance_loss_clip": 1.023628, + "balance_loss_mlp": 1.02774, + "epoch": 0.21379185955630373, + "flos": 18989347680000.0, + "grad_norm": 2.157257176209222, + "language_loss": 0.86256701, + "learning_rate": 3.6561463868756965e-06, + "loss": 0.88479769, + "num_input_tokens_seen": 37855030, + "step": 1778, + "time_per_iteration": 2.6730124950408936 + }, + { + "auxiliary_loss_clip": 0.01190842, + "auxiliary_loss_mlp": 0.01038598, + "balance_loss_clip": 1.02623785, + "balance_loss_mlp": 1.02883458, + "epoch": 0.21391210244694284, + "flos": 28218497207040.0, + "grad_norm": 1.465472882632113, + "language_loss": 0.77892035, + "learning_rate": 3.655709554182452e-06, + "loss": 0.80121481, + "num_input_tokens_seen": 37875370, + "step": 1779, + "time_per_iteration": 2.697810649871826 + }, + { + "auxiliary_loss_clip": 0.01194753, + "auxiliary_loss_mlp": 0.01036982, + "balance_loss_clip": 1.02598596, + "balance_loss_mlp": 1.02699196, + "epoch": 0.21403234533758192, + "flos": 17455064192640.0, + "grad_norm": 2.2557094985335393, + "language_loss": 0.84540838, + "learning_rate": 3.6552724703193855e-06, + "loss": 0.86772573, + "num_input_tokens_seen": 37892560, + "step": 1780, + "time_per_iteration": 2.690929889678955 + }, + { + "auxiliary_loss_clip": 0.01033218, + "auxiliary_loss_mlp": 0.01001055, + "balance_loss_clip": 0.88592172, + "balance_loss_mlp": 0.99879038, + "epoch": 0.214152588228221, + "flos": 51637606686720.0, + "grad_norm": 0.7851567889402289, + "language_loss": 0.55962962, + "learning_rate": 3.654835135352801e-06, + "loss": 0.57997233, + "num_input_tokens_seen": 37947370, + "step": 1781, + "time_per_iteration": 3.238374710083008 + }, + { + "auxiliary_loss_clip": 0.01117903, + "auxiliary_loss_mlp": 0.01034002, + "balance_loss_clip": 0.90520006, + "balance_loss_mlp": 1.02320218, + "epoch": 0.21427283111886009, + "flos": 19496154625920.0, + "grad_norm": 1.7669322022391052, + "language_loss": 0.87367749, + "learning_rate": 3.654397549349043e-06, + "loss": 0.89519656, + "num_input_tokens_seen": 37964745, + "step": 1782, + "time_per_iteration": 2.721022605895996 + }, + { + "auxiliary_loss_clip": 0.0116555, + "auxiliary_loss_mlp": 0.01038573, + "balance_loss_clip": 0.98824197, + "balance_loss_mlp": 1.02766562, + "epoch": 0.2143930740094992, + "flos": 20084802710400.0, + "grad_norm": 1.7692740869879546, + "language_loss": 0.7534945, + "learning_rate": 3.653959712374491e-06, + "loss": 0.7755357, + "num_input_tokens_seen": 37982850, + "step": 1783, + "time_per_iteration": 2.734537124633789 + }, + { + "auxiliary_loss_clip": 0.01138696, + "auxiliary_loss_mlp": 0.01041188, + "balance_loss_clip": 0.95076299, + "balance_loss_mlp": 1.03051257, + "epoch": 0.21451331690013828, + "flos": 21798603394560.0, + "grad_norm": 2.047750215883663, + "language_loss": 0.82657754, + "learning_rate": 3.6535216244955663e-06, + "loss": 0.84837639, + "num_input_tokens_seen": 38002745, + "step": 1784, + "time_per_iteration": 2.7346596717834473 + }, + { + "auxiliary_loss_clip": 0.01165313, + "auxiliary_loss_mlp": 0.01034226, + "balance_loss_clip": 0.98589766, + "balance_loss_mlp": 1.02449894, + "epoch": 0.21463355979077736, + "flos": 32853882412800.0, + "grad_norm": 2.262909643470795, + "language_loss": 0.71036327, + "learning_rate": 3.653083285778726e-06, + "loss": 0.73235869, + "num_input_tokens_seen": 38024115, + "step": 1785, + "time_per_iteration": 4.680786609649658 + }, + { + "auxiliary_loss_clip": 0.01196996, + "auxiliary_loss_mlp": 0.01039985, + "balance_loss_clip": 1.02573645, + "balance_loss_mlp": 1.02955461, + "epoch": 0.21475380268141647, + "flos": 21543817248000.0, + "grad_norm": 2.1533786323456443, + "language_loss": 0.80930042, + "learning_rate": 3.6526446962904653e-06, + "loss": 0.83167028, + "num_input_tokens_seen": 38042830, + "step": 1786, + "time_per_iteration": 2.660290479660034 + }, + { + "auxiliary_loss_clip": 0.01189388, + "auxiliary_loss_mlp": 0.01048468, + "balance_loss_clip": 1.02888775, + "balance_loss_mlp": 1.03840113, + "epoch": 0.21487404557205556, + "flos": 32159082660480.0, + "grad_norm": 1.4972819975533778, + "language_loss": 0.74310482, + "learning_rate": 3.652205856097318e-06, + "loss": 0.76548338, + "num_input_tokens_seen": 38066015, + "step": 1787, + "time_per_iteration": 2.7427456378936768 + }, + { + "auxiliary_loss_clip": 0.01153641, + "auxiliary_loss_mlp": 0.00513509, + "balance_loss_clip": 0.95001894, + "balance_loss_mlp": 0.0, + "epoch": 0.21499428846269464, + "flos": 12673091583360.0, + "grad_norm": 2.1691019993802825, + "language_loss": 0.79670227, + "learning_rate": 3.651766765265856e-06, + "loss": 0.8133738, + "num_input_tokens_seen": 38083025, + "step": 1788, + "time_per_iteration": 2.7464373111724854 + }, + { + "auxiliary_loss_clip": 0.01158391, + "auxiliary_loss_mlp": 0.01031923, + "balance_loss_clip": 0.98060966, + "balance_loss_mlp": 1.02143872, + "epoch": 0.21511453135333372, + "flos": 23471573293440.0, + "grad_norm": 2.8396936931456804, + "language_loss": 0.8146593, + "learning_rate": 3.65132742386269e-06, + "loss": 0.83656245, + "num_input_tokens_seen": 38098245, + "step": 1789, + "time_per_iteration": 4.611021995544434 + }, + { + "auxiliary_loss_clip": 0.01216195, + "auxiliary_loss_mlp": 0.01035807, + "balance_loss_clip": 1.06444609, + "balance_loss_mlp": 1.02529883, + "epoch": 0.21523477424397283, + "flos": 26943560893440.0, + "grad_norm": 1.7638703168423355, + "language_loss": 0.84699333, + "learning_rate": 3.6508878319544656e-06, + "loss": 0.86951339, + "num_input_tokens_seen": 38118460, + "step": 1790, + "time_per_iteration": 2.616900682449341 + }, + { + "auxiliary_loss_clip": 0.01164472, + "auxiliary_loss_mlp": 0.01035311, + "balance_loss_clip": 0.9913007, + "balance_loss_mlp": 1.02476716, + "epoch": 0.21535501713461191, + "flos": 18916161719040.0, + "grad_norm": 3.2332342492786443, + "language_loss": 0.81733787, + "learning_rate": 3.65044798960787e-06, + "loss": 0.83933568, + "num_input_tokens_seen": 38136800, + "step": 1791, + "time_per_iteration": 2.8411364555358887 + }, + { + "auxiliary_loss_clip": 0.01131467, + "auxiliary_loss_mlp": 0.01037072, + "balance_loss_clip": 0.94241565, + "balance_loss_mlp": 1.02675438, + "epoch": 0.215475260025251, + "flos": 17895113712000.0, + "grad_norm": 1.9770642464697403, + "language_loss": 0.7864064, + "learning_rate": 3.650007896889627e-06, + "loss": 0.80809182, + "num_input_tokens_seen": 38155380, + "step": 1792, + "time_per_iteration": 2.7830934524536133 + }, + { + "auxiliary_loss_clip": 0.01216708, + "auxiliary_loss_mlp": 0.01043694, + "balance_loss_clip": 1.06856894, + "balance_loss_mlp": 1.03378153, + "epoch": 0.2155955029158901, + "flos": 16654292340480.0, + "grad_norm": 1.9391800669720536, + "language_loss": 0.8056314, + "learning_rate": 3.6495675538664974e-06, + "loss": 0.82823539, + "num_input_tokens_seen": 38174395, + "step": 1793, + "time_per_iteration": 2.731207847595215 + }, + { + "auxiliary_loss_clip": 0.01168781, + "auxiliary_loss_mlp": 0.01037129, + "balance_loss_clip": 0.98569429, + "balance_loss_mlp": 1.02710903, + "epoch": 0.2157157458065292, + "flos": 23621213352960.0, + "grad_norm": 2.3075460093648474, + "language_loss": 0.82833427, + "learning_rate": 3.649126960605282e-06, + "loss": 0.85039335, + "num_input_tokens_seen": 38195380, + "step": 1794, + "time_per_iteration": 2.8138437271118164 + }, + { + "auxiliary_loss_clip": 0.01163116, + "auxiliary_loss_mlp": 0.01038993, + "balance_loss_clip": 0.9873004, + "balance_loss_mlp": 1.02862191, + "epoch": 0.21583598869716827, + "flos": 22127078292480.0, + "grad_norm": 2.3312051175894446, + "language_loss": 0.83602715, + "learning_rate": 3.6486861171728174e-06, + "loss": 0.8580482, + "num_input_tokens_seen": 38213775, + "step": 1795, + "time_per_iteration": 2.6928982734680176 + }, + { + "auxiliary_loss_clip": 0.01145242, + "auxiliary_loss_mlp": 0.01036684, + "balance_loss_clip": 0.94719315, + "balance_loss_mlp": 1.02600253, + "epoch": 0.21595623158780738, + "flos": 23441229279360.0, + "grad_norm": 1.5005816750836283, + "language_loss": 0.78483498, + "learning_rate": 3.6482450236359803e-06, + "loss": 0.80665421, + "num_input_tokens_seen": 38235630, + "step": 1796, + "time_per_iteration": 2.826228141784668 + }, + { + "auxiliary_loss_clip": 0.01189011, + "auxiliary_loss_mlp": 0.01039473, + "balance_loss_clip": 1.02744818, + "balance_loss_mlp": 1.02937555, + "epoch": 0.21607647447844647, + "flos": 26906501036160.0, + "grad_norm": 2.1355420612464253, + "language_loss": 0.77736002, + "learning_rate": 3.647803680061683e-06, + "loss": 0.79964489, + "num_input_tokens_seen": 38256045, + "step": 1797, + "time_per_iteration": 2.6965208053588867 + }, + { + "auxiliary_loss_clip": 0.01168205, + "auxiliary_loss_mlp": 0.01033344, + "balance_loss_clip": 0.98782617, + "balance_loss_mlp": 1.02272797, + "epoch": 0.21619671736908555, + "flos": 14495378319360.0, + "grad_norm": 2.4267070450787807, + "language_loss": 0.74529147, + "learning_rate": 3.6473620865168776e-06, + "loss": 0.76730692, + "num_input_tokens_seen": 38272915, + "step": 1798, + "time_per_iteration": 2.653581380844116 + }, + { + "auxiliary_loss_clip": 0.01171015, + "auxiliary_loss_mlp": 0.01038745, + "balance_loss_clip": 0.99299645, + "balance_loss_mlp": 1.0291065, + "epoch": 0.21631696025972463, + "flos": 17931096161280.0, + "grad_norm": 1.8064259112957382, + "language_loss": 0.81270397, + "learning_rate": 3.646920243068554e-06, + "loss": 0.83480155, + "num_input_tokens_seen": 38290810, + "step": 1799, + "time_per_iteration": 2.644907236099243 + }, + { + "auxiliary_loss_clip": 0.01153776, + "auxiliary_loss_mlp": 0.01028578, + "balance_loss_clip": 0.98381174, + "balance_loss_mlp": 1.01848125, + "epoch": 0.21643720315036374, + "flos": 24462385027200.0, + "grad_norm": 1.8529069028813936, + "language_loss": 0.74589485, + "learning_rate": 3.6464781497837384e-06, + "loss": 0.76771843, + "num_input_tokens_seen": 38312785, + "step": 1800, + "time_per_iteration": 2.743797540664673 + }, + { + "auxiliary_loss_clip": 0.01170746, + "auxiliary_loss_mlp": 0.01041329, + "balance_loss_clip": 0.98642528, + "balance_loss_mlp": 1.03091633, + "epoch": 0.21655744604100283, + "flos": 28474432588800.0, + "grad_norm": 1.635342064201475, + "language_loss": 0.72672313, + "learning_rate": 3.6460358067294965e-06, + "loss": 0.74884391, + "num_input_tokens_seen": 38334015, + "step": 1801, + "time_per_iteration": 2.7168140411376953 + }, + { + "auxiliary_loss_clip": 0.01218476, + "auxiliary_loss_mlp": 0.01037221, + "balance_loss_clip": 1.06429243, + "balance_loss_mlp": 1.02732623, + "epoch": 0.2166776889316419, + "flos": 20152960767360.0, + "grad_norm": 2.0537726980116475, + "language_loss": 0.77270842, + "learning_rate": 3.645593213972932e-06, + "loss": 0.79526532, + "num_input_tokens_seen": 38352920, + "step": 1802, + "time_per_iteration": 2.6578903198242188 + }, + { + "auxiliary_loss_clip": 0.01186489, + "auxiliary_loss_mlp": 0.01039562, + "balance_loss_clip": 1.02612352, + "balance_loss_mlp": 1.02939963, + "epoch": 0.21679793182228102, + "flos": 15193482122880.0, + "grad_norm": 2.3615510028867575, + "language_loss": 0.79695523, + "learning_rate": 3.6451503715811852e-06, + "loss": 0.81921571, + "num_input_tokens_seen": 38371230, + "step": 1803, + "time_per_iteration": 2.7096359729766846 + }, + { + "auxiliary_loss_clip": 0.01166309, + "auxiliary_loss_mlp": 0.01037229, + "balance_loss_clip": 0.99119806, + "balance_loss_mlp": 1.02804375, + "epoch": 0.2169181747129201, + "flos": 17384464010880.0, + "grad_norm": 1.9119419665386548, + "language_loss": 0.80048603, + "learning_rate": 3.6447072796214345e-06, + "loss": 0.82252139, + "num_input_tokens_seen": 38389795, + "step": 1804, + "time_per_iteration": 2.704676628112793 + }, + { + "auxiliary_loss_clip": 0.01032845, + "auxiliary_loss_mlp": 0.01012794, + "balance_loss_clip": 0.8849557, + "balance_loss_mlp": 1.01021862, + "epoch": 0.21703841760355919, + "flos": 58760955429120.0, + "grad_norm": 0.9267483734297932, + "language_loss": 0.63178396, + "learning_rate": 3.644263938160898e-06, + "loss": 0.6522404, + "num_input_tokens_seen": 38445760, + "step": 1805, + "time_per_iteration": 3.1847927570343018 + }, + { + "auxiliary_loss_clip": 0.01140276, + "auxiliary_loss_mlp": 0.0103657, + "balance_loss_clip": 0.94892883, + "balance_loss_mlp": 1.02685404, + "epoch": 0.21715866049419827, + "flos": 22418457419520.0, + "grad_norm": 1.9201751346336615, + "language_loss": 0.71796614, + "learning_rate": 3.6438203472668293e-06, + "loss": 0.73973465, + "num_input_tokens_seen": 38465405, + "step": 1806, + "time_per_iteration": 2.754828691482544 + }, + { + "auxiliary_loss_clip": 0.01168193, + "auxiliary_loss_mlp": 0.01032475, + "balance_loss_clip": 0.9883042, + "balance_loss_mlp": 1.02367163, + "epoch": 0.21727890338483738, + "flos": 17237732952960.0, + "grad_norm": 2.447102326421568, + "language_loss": 0.81761765, + "learning_rate": 3.6433765070065206e-06, + "loss": 0.83962435, + "num_input_tokens_seen": 38483195, + "step": 1807, + "time_per_iteration": 2.705918312072754 + }, + { + "auxiliary_loss_clip": 0.01214931, + "auxiliary_loss_mlp": 0.01030706, + "balance_loss_clip": 1.06446636, + "balance_loss_mlp": 1.02085984, + "epoch": 0.21739914627547646, + "flos": 13434792416640.0, + "grad_norm": 2.67071658642055, + "language_loss": 0.87488961, + "learning_rate": 3.6429324174473025e-06, + "loss": 0.89734602, + "num_input_tokens_seen": 38496735, + "step": 1808, + "time_per_iteration": 2.5612897872924805 + }, + { + "auxiliary_loss_clip": 0.01190416, + "auxiliary_loss_mlp": 0.01036357, + "balance_loss_clip": 1.0252707, + "balance_loss_mlp": 1.02671313, + "epoch": 0.21751938916611555, + "flos": 20959514709120.0, + "grad_norm": 3.3284571143652757, + "language_loss": 0.85039991, + "learning_rate": 3.6424880786565425e-06, + "loss": 0.87266767, + "num_input_tokens_seen": 38512880, + "step": 1809, + "time_per_iteration": 2.6579370498657227 + }, + { + "auxiliary_loss_clip": 0.01119281, + "auxiliary_loss_mlp": 0.01036884, + "balance_loss_clip": 0.91188687, + "balance_loss_mlp": 1.02648902, + "epoch": 0.21763963205675466, + "flos": 27599936071680.0, + "grad_norm": 2.9290925365754235, + "language_loss": 0.7964108, + "learning_rate": 3.6420434907016482e-06, + "loss": 0.81797242, + "num_input_tokens_seen": 38532570, + "step": 1810, + "time_per_iteration": 2.7802116870880127 + }, + { + "auxiliary_loss_clip": 0.0119354, + "auxiliary_loss_mlp": 0.01034208, + "balance_loss_clip": 1.02983809, + "balance_loss_mlp": 1.02451658, + "epoch": 0.21775987494739374, + "flos": 21430411032960.0, + "grad_norm": 1.5865727701931294, + "language_loss": 0.81081629, + "learning_rate": 3.6415986536500606e-06, + "loss": 0.83309382, + "num_input_tokens_seen": 38550900, + "step": 1811, + "time_per_iteration": 3.5123443603515625 + }, + { + "auxiliary_loss_clip": 0.01112259, + "auxiliary_loss_mlp": 0.01033088, + "balance_loss_clip": 0.91356391, + "balance_loss_mlp": 1.0232532, + "epoch": 0.21788011783803282, + "flos": 18332972501760.0, + "grad_norm": 1.7869501429399468, + "language_loss": 0.80331886, + "learning_rate": 3.641153567569263e-06, + "loss": 0.82477236, + "num_input_tokens_seen": 38569215, + "step": 1812, + "time_per_iteration": 2.602325916290283 + }, + { + "auxiliary_loss_clip": 0.01186098, + "auxiliary_loss_mlp": 0.01039688, + "balance_loss_clip": 1.02495217, + "balance_loss_mlp": 1.03015137, + "epoch": 0.2180003607286719, + "flos": 30262748037120.0, + "grad_norm": 2.5174146977100404, + "language_loss": 0.95195872, + "learning_rate": 3.640708232526774e-06, + "loss": 0.97421658, + "num_input_tokens_seen": 38587870, + "step": 1813, + "time_per_iteration": 2.636720895767212 + }, + { + "auxiliary_loss_clip": 0.01087424, + "auxiliary_loss_mlp": 0.01038466, + "balance_loss_clip": 0.86217362, + "balance_loss_mlp": 1.02855396, + "epoch": 0.21812060361931102, + "flos": 25480272637440.0, + "grad_norm": 1.6567706218818705, + "language_loss": 0.78353322, + "learning_rate": 3.6402626485901504e-06, + "loss": 0.80479211, + "num_input_tokens_seen": 38606965, + "step": 1814, + "time_per_iteration": 3.728670597076416 + }, + { + "auxiliary_loss_clip": 0.0118938, + "auxiliary_loss_mlp": 0.0103421, + "balance_loss_clip": 1.02819109, + "balance_loss_mlp": 1.02523935, + "epoch": 0.2182408465099501, + "flos": 21908166854400.0, + "grad_norm": 1.9290898102328016, + "language_loss": 0.78234899, + "learning_rate": 3.639816815826988e-06, + "loss": 0.80458498, + "num_input_tokens_seen": 38626290, + "step": 1815, + "time_per_iteration": 3.490337371826172 + }, + { + "auxiliary_loss_clip": 0.01166948, + "auxiliary_loss_mlp": 0.01040292, + "balance_loss_clip": 0.98945498, + "balance_loss_mlp": 1.03130364, + "epoch": 0.21836108940058918, + "flos": 23657339456640.0, + "grad_norm": 1.9215999112055457, + "language_loss": 0.77821219, + "learning_rate": 3.6393707343049176e-06, + "loss": 0.80028462, + "num_input_tokens_seen": 38646620, + "step": 1816, + "time_per_iteration": 2.7116572856903076 + }, + { + "auxiliary_loss_clip": 0.01195949, + "auxiliary_loss_mlp": 0.01034439, + "balance_loss_clip": 1.0274682, + "balance_loss_mlp": 1.02468801, + "epoch": 0.2184813322912283, + "flos": 24681009156480.0, + "grad_norm": 2.794341775740006, + "language_loss": 0.73099142, + "learning_rate": 3.6389244040916104e-06, + "loss": 0.7532953, + "num_input_tokens_seen": 38665695, + "step": 1817, + "time_per_iteration": 2.666888952255249 + }, + { + "auxiliary_loss_clip": 0.01153815, + "auxiliary_loss_mlp": 0.00512914, + "balance_loss_clip": 0.98290366, + "balance_loss_mlp": 0.0, + "epoch": 0.21860157518186737, + "flos": 26574650259840.0, + "grad_norm": 1.9036850707313464, + "language_loss": 0.7924518, + "learning_rate": 3.6384778252547747e-06, + "loss": 0.80911911, + "num_input_tokens_seen": 38681575, + "step": 1818, + "time_per_iteration": 2.7350001335144043 + }, + { + "auxiliary_loss_clip": 0.01163132, + "auxiliary_loss_mlp": 0.00512948, + "balance_loss_clip": 0.98846406, + "balance_loss_mlp": 0.0, + "epoch": 0.21872181807250646, + "flos": 20886292834560.0, + "grad_norm": 2.2125109335601145, + "language_loss": 0.77626216, + "learning_rate": 3.638030997862155e-06, + "loss": 0.79302293, + "num_input_tokens_seen": 38700510, + "step": 1819, + "time_per_iteration": 2.7295773029327393 + }, + { + "auxiliary_loss_clip": 0.01070082, + "auxiliary_loss_mlp": 0.010036, + "balance_loss_clip": 0.96172583, + "balance_loss_mlp": 1.0012989, + "epoch": 0.21884206096314554, + "flos": 61209452897280.0, + "grad_norm": 0.7673731169930139, + "language_loss": 0.59406471, + "learning_rate": 3.6375839219815356e-06, + "loss": 0.61480153, + "num_input_tokens_seen": 38758310, + "step": 1820, + "time_per_iteration": 3.1546595096588135 + }, + { + "auxiliary_loss_clip": 0.01216059, + "auxiliary_loss_mlp": 0.0104256, + "balance_loss_clip": 1.06548929, + "balance_loss_mlp": 1.03267157, + "epoch": 0.21896230385378465, + "flos": 23473835850240.0, + "grad_norm": 1.9716217578823876, + "language_loss": 0.82896274, + "learning_rate": 3.6371365976807375e-06, + "loss": 0.85154891, + "num_input_tokens_seen": 38778705, + "step": 1821, + "time_per_iteration": 2.734419345855713 + }, + { + "auxiliary_loss_clip": 0.0111332, + "auxiliary_loss_mlp": 0.01031971, + "balance_loss_clip": 0.91118824, + "balance_loss_mlp": 1.02214766, + "epoch": 0.21908254674442373, + "flos": 25081915829760.0, + "grad_norm": 1.7227611467771033, + "language_loss": 0.83826119, + "learning_rate": 3.6366890250276185e-06, + "loss": 0.85971409, + "num_input_tokens_seen": 38799660, + "step": 1822, + "time_per_iteration": 2.765018939971924 + }, + { + "auxiliary_loss_clip": 0.01212313, + "auxiliary_loss_mlp": 0.01039329, + "balance_loss_clip": 1.06487513, + "balance_loss_mlp": 1.02957773, + "epoch": 0.21920278963506282, + "flos": 23513768795520.0, + "grad_norm": 2.027941726598193, + "language_loss": 0.89987898, + "learning_rate": 3.6362412040900764e-06, + "loss": 0.92239541, + "num_input_tokens_seen": 38819450, + "step": 1823, + "time_per_iteration": 2.671598434448242 + }, + { + "auxiliary_loss_clip": 0.01193017, + "auxiliary_loss_mlp": 0.01034062, + "balance_loss_clip": 1.02558517, + "balance_loss_mlp": 1.02401233, + "epoch": 0.21932303252570193, + "flos": 29242238734080.0, + "grad_norm": 1.8640868384178646, + "language_loss": 0.80438507, + "learning_rate": 3.635793134936044e-06, + "loss": 0.82665586, + "num_input_tokens_seen": 38840460, + "step": 1824, + "time_per_iteration": 2.6985297203063965 + }, + { + "auxiliary_loss_clip": 0.01188814, + "auxiliary_loss_mlp": 0.01037868, + "balance_loss_clip": 1.0277133, + "balance_loss_mlp": 1.02814078, + "epoch": 0.219443275416341, + "flos": 20806857907200.0, + "grad_norm": 1.8058871518385222, + "language_loss": 0.7326932, + "learning_rate": 3.635344817633494e-06, + "loss": 0.75496, + "num_input_tokens_seen": 38859775, + "step": 1825, + "time_per_iteration": 2.6937644481658936 + }, + { + "auxiliary_loss_clip": 0.01185263, + "auxiliary_loss_mlp": 0.01037646, + "balance_loss_clip": 1.02505922, + "balance_loss_mlp": 1.02806187, + "epoch": 0.2195635183069801, + "flos": 14501555458560.0, + "grad_norm": 2.1547717429710085, + "language_loss": 0.75266182, + "learning_rate": 3.634896252250436e-06, + "loss": 0.77489096, + "num_input_tokens_seen": 38876540, + "step": 1826, + "time_per_iteration": 2.5801568031311035 + }, + { + "auxiliary_loss_clip": 0.01215411, + "auxiliary_loss_mlp": 0.01032167, + "balance_loss_clip": 1.06579554, + "balance_loss_mlp": 1.02318478, + "epoch": 0.2196837611976192, + "flos": 24243473589120.0, + "grad_norm": 3.282682797499977, + "language_loss": 0.8241722, + "learning_rate": 3.6344474388549157e-06, + "loss": 0.84664798, + "num_input_tokens_seen": 38896195, + "step": 1827, + "time_per_iteration": 2.655954360961914 + }, + { + "auxiliary_loss_clip": 0.01191415, + "auxiliary_loss_mlp": 0.01037584, + "balance_loss_clip": 1.027318, + "balance_loss_mlp": 1.02873313, + "epoch": 0.2198040040882583, + "flos": 18074523168000.0, + "grad_norm": 2.1280884600070498, + "language_loss": 0.79998052, + "learning_rate": 3.6339983775150183e-06, + "loss": 0.82227057, + "num_input_tokens_seen": 38912755, + "step": 1828, + "time_per_iteration": 2.6126768589019775 + }, + { + "auxiliary_loss_clip": 0.01191278, + "auxiliary_loss_mlp": 0.01035384, + "balance_loss_clip": 1.02873635, + "balance_loss_mlp": 1.02600193, + "epoch": 0.21992424697889737, + "flos": 17784185535360.0, + "grad_norm": 2.2766354148385046, + "language_loss": 0.84079224, + "learning_rate": 3.6335490682988664e-06, + "loss": 0.86305887, + "num_input_tokens_seen": 38928365, + "step": 1829, + "time_per_iteration": 2.6267566680908203 + }, + { + "auxiliary_loss_clip": 0.01084756, + "auxiliary_loss_mlp": 0.01036949, + "balance_loss_clip": 0.86591631, + "balance_loss_mlp": 1.02713847, + "epoch": 0.22004448986953645, + "flos": 17638495971840.0, + "grad_norm": 3.959495560154008, + "language_loss": 0.83311903, + "learning_rate": 3.63309951127462e-06, + "loss": 0.85433614, + "num_input_tokens_seen": 38945275, + "step": 1830, + "time_per_iteration": 2.7805609703063965 + }, + { + "auxiliary_loss_clip": 0.01144704, + "auxiliary_loss_mlp": 0.01033029, + "balance_loss_clip": 0.95138907, + "balance_loss_mlp": 1.02339697, + "epoch": 0.22016473276017556, + "flos": 22275533203200.0, + "grad_norm": 2.0035400262904095, + "language_loss": 0.75436133, + "learning_rate": 3.6326497065104757e-06, + "loss": 0.77613872, + "num_input_tokens_seen": 38965740, + "step": 1831, + "time_per_iteration": 2.709104299545288 + }, + { + "auxiliary_loss_clip": 0.01194584, + "auxiliary_loss_mlp": 0.01042732, + "balance_loss_clip": 1.02646327, + "balance_loss_mlp": 1.03303456, + "epoch": 0.22028497565081465, + "flos": 25556259859200.0, + "grad_norm": 2.000036669427333, + "language_loss": 0.7788322, + "learning_rate": 3.6321996540746697e-06, + "loss": 0.80120528, + "num_input_tokens_seen": 38984815, + "step": 1832, + "time_per_iteration": 2.736145257949829 + }, + { + "auxiliary_loss_clip": 0.01140555, + "auxiliary_loss_mlp": 0.0103816, + "balance_loss_clip": 0.949036, + "balance_loss_mlp": 1.02822948, + "epoch": 0.22040521854145373, + "flos": 36247332925440.0, + "grad_norm": 1.6856106760252516, + "language_loss": 0.80346191, + "learning_rate": 3.6317493540354733e-06, + "loss": 0.82524908, + "num_input_tokens_seen": 39008230, + "step": 1833, + "time_per_iteration": 2.821821451187134 + }, + { + "auxiliary_loss_clip": 0.01183453, + "auxiliary_loss_mlp": 0.01040145, + "balance_loss_clip": 1.02354193, + "balance_loss_mlp": 1.02951121, + "epoch": 0.22052546143209284, + "flos": 11838420270720.0, + "grad_norm": 1.8217882446156735, + "language_loss": 0.7665804, + "learning_rate": 3.6312988064611976e-06, + "loss": 0.78881633, + "num_input_tokens_seen": 39026540, + "step": 1834, + "time_per_iteration": 2.692239761352539 + }, + { + "auxiliary_loss_clip": 0.01143865, + "auxiliary_loss_mlp": 0.01035077, + "balance_loss_clip": 0.9454633, + "balance_loss_mlp": 1.02521884, + "epoch": 0.22064570432273192, + "flos": 24209250906240.0, + "grad_norm": 1.7097518150601998, + "language_loss": 0.81317717, + "learning_rate": 3.6308480114201896e-06, + "loss": 0.83496654, + "num_input_tokens_seen": 39048460, + "step": 1835, + "time_per_iteration": 2.7414753437042236 + }, + { + "auxiliary_loss_clip": 0.01219885, + "auxiliary_loss_mlp": 0.01033621, + "balance_loss_clip": 1.06976938, + "balance_loss_mlp": 1.02402425, + "epoch": 0.220765947213371, + "flos": 17931347556480.0, + "grad_norm": 1.6598702612983398, + "language_loss": 0.76578385, + "learning_rate": 3.630396968980835e-06, + "loss": 0.78831887, + "num_input_tokens_seen": 39066335, + "step": 1836, + "time_per_iteration": 2.6288251876831055 + }, + { + "auxiliary_loss_clip": 0.01166062, + "auxiliary_loss_mlp": 0.01043823, + "balance_loss_clip": 0.98565054, + "balance_loss_mlp": 1.0340476, + "epoch": 0.2208861901040101, + "flos": 26757040544640.0, + "grad_norm": 2.554360858347974, + "language_loss": 0.83252394, + "learning_rate": 3.6299456792115575e-06, + "loss": 0.85462272, + "num_input_tokens_seen": 39087590, + "step": 1837, + "time_per_iteration": 4.674769163131714 + }, + { + "auxiliary_loss_clip": 0.01054127, + "auxiliary_loss_mlp": 0.0103651, + "balance_loss_clip": 0.82648855, + "balance_loss_mlp": 1.02640152, + "epoch": 0.2210064329946492, + "flos": 17817977255040.0, + "grad_norm": 1.752879597150806, + "language_loss": 0.81127369, + "learning_rate": 3.629494142180815e-06, + "loss": 0.83218002, + "num_input_tokens_seen": 39106335, + "step": 1838, + "time_per_iteration": 2.748518943786621 + }, + { + "auxiliary_loss_clip": 0.01216957, + "auxiliary_loss_mlp": 0.0103831, + "balance_loss_clip": 1.0668124, + "balance_loss_mlp": 1.02865374, + "epoch": 0.22112667588528828, + "flos": 17967401832960.0, + "grad_norm": 2.093256430052012, + "language_loss": 0.85308409, + "learning_rate": 3.6290423579571075e-06, + "loss": 0.8756367, + "num_input_tokens_seen": 39122875, + "step": 1839, + "time_per_iteration": 2.732670307159424 + }, + { + "auxiliary_loss_clip": 0.01183744, + "auxiliary_loss_mlp": 0.0103132, + "balance_loss_clip": 1.02364492, + "balance_loss_mlp": 1.02194393, + "epoch": 0.22124691877592736, + "flos": 18369206346240.0, + "grad_norm": 1.9609566284586368, + "language_loss": 0.80204004, + "learning_rate": 3.6285903266089694e-06, + "loss": 0.82419068, + "num_input_tokens_seen": 39142150, + "step": 1840, + "time_per_iteration": 3.5025064945220947 + }, + { + "auxiliary_loss_clip": 0.01169079, + "auxiliary_loss_mlp": 0.01041204, + "balance_loss_clip": 0.98771441, + "balance_loss_mlp": 1.03130364, + "epoch": 0.22136716166656648, + "flos": 20813286441600.0, + "grad_norm": 1.8417028127542643, + "language_loss": 0.77340794, + "learning_rate": 3.628138048204974e-06, + "loss": 0.79551077, + "num_input_tokens_seen": 39162835, + "step": 1841, + "time_per_iteration": 3.5430691242218018 + }, + { + "auxiliary_loss_clip": 0.01107892, + "auxiliary_loss_mlp": 0.01037375, + "balance_loss_clip": 0.90728921, + "balance_loss_mlp": 1.02655125, + "epoch": 0.22148740455720556, + "flos": 17675699483520.0, + "grad_norm": 3.016043565820599, + "language_loss": 0.76061165, + "learning_rate": 3.6276855228137304e-06, + "loss": 0.78206432, + "num_input_tokens_seen": 39181040, + "step": 1842, + "time_per_iteration": 2.821300506591797 + }, + { + "auxiliary_loss_clip": 0.01215008, + "auxiliary_loss_mlp": 0.00512449, + "balance_loss_clip": 1.06758118, + "balance_loss_mlp": 0.0, + "epoch": 0.22160764744784464, + "flos": 21726710323200.0, + "grad_norm": 2.444634408464198, + "language_loss": 0.81861848, + "learning_rate": 3.6272327505038874e-06, + "loss": 0.83589303, + "num_input_tokens_seen": 39197505, + "step": 1843, + "time_per_iteration": 2.643641471862793 + }, + { + "auxiliary_loss_clip": 0.01121434, + "auxiliary_loss_mlp": 0.01034336, + "balance_loss_clip": 0.90894544, + "balance_loss_mlp": 1.02521086, + "epoch": 0.22172789033848372, + "flos": 23764712186880.0, + "grad_norm": 1.8230437611251866, + "language_loss": 0.78476512, + "learning_rate": 3.626779731344131e-06, + "loss": 0.80632287, + "num_input_tokens_seen": 39217295, + "step": 1844, + "time_per_iteration": 2.7739064693450928 + }, + { + "auxiliary_loss_clip": 0.01211529, + "auxiliary_loss_mlp": 0.01028101, + "balance_loss_clip": 1.06413066, + "balance_loss_mlp": 1.01870155, + "epoch": 0.22184813322912283, + "flos": 16982300361600.0, + "grad_norm": 1.977152827436162, + "language_loss": 0.85259664, + "learning_rate": 3.6263264654031814e-06, + "loss": 0.87499297, + "num_input_tokens_seen": 39234195, + "step": 1845, + "time_per_iteration": 2.6095850467681885 + }, + { + "auxiliary_loss_clip": 0.01050357, + "auxiliary_loss_mlp": 0.01006554, + "balance_loss_clip": 0.9210825, + "balance_loss_mlp": 1.00440776, + "epoch": 0.22196837611976192, + "flos": 61823740314240.0, + "grad_norm": 0.8290980634370463, + "language_loss": 0.59124422, + "learning_rate": 3.6258729527498008e-06, + "loss": 0.61181343, + "num_input_tokens_seen": 39295040, + "step": 1846, + "time_per_iteration": 3.2351999282836914 + }, + { + "auxiliary_loss_clip": 0.01171875, + "auxiliary_loss_mlp": 0.01034312, + "balance_loss_clip": 0.98918062, + "balance_loss_mlp": 1.02417278, + "epoch": 0.222088619010401, + "flos": 25558019625600.0, + "grad_norm": 3.4037629933283213, + "language_loss": 0.64366817, + "learning_rate": 3.6254191934527854e-06, + "loss": 0.66573, + "num_input_tokens_seen": 39314395, + "step": 1847, + "time_per_iteration": 2.7963430881500244 + }, + { + "auxiliary_loss_clip": 0.011403, + "auxiliary_loss_mlp": 0.01036231, + "balance_loss_clip": 0.94983232, + "balance_loss_mlp": 1.02649784, + "epoch": 0.2222088619010401, + "flos": 19318612677120.0, + "grad_norm": 2.513849217352166, + "language_loss": 0.64614785, + "learning_rate": 3.6249651875809715e-06, + "loss": 0.6679132, + "num_input_tokens_seen": 39334275, + "step": 1848, + "time_per_iteration": 2.7039432525634766 + }, + { + "auxiliary_loss_clip": 0.01159517, + "auxiliary_loss_mlp": 0.01029685, + "balance_loss_clip": 0.98665041, + "balance_loss_mlp": 1.02010632, + "epoch": 0.2223291047916792, + "flos": 19099342103040.0, + "grad_norm": 1.9326097640268824, + "language_loss": 0.89253968, + "learning_rate": 3.62451093520323e-06, + "loss": 0.91443169, + "num_input_tokens_seen": 39352180, + "step": 1849, + "time_per_iteration": 2.690131902694702 + }, + { + "auxiliary_loss_clip": 0.01109229, + "auxiliary_loss_mlp": 0.01033737, + "balance_loss_clip": 0.90445828, + "balance_loss_mlp": 1.02486777, + "epoch": 0.22244934768231828, + "flos": 20850418126080.0, + "grad_norm": 1.9545379726500294, + "language_loss": 0.90439904, + "learning_rate": 3.6240564363884714e-06, + "loss": 0.9258287, + "num_input_tokens_seen": 39372125, + "step": 1850, + "time_per_iteration": 2.7487974166870117 + }, + { + "auxiliary_loss_clip": 0.01191646, + "auxiliary_loss_mlp": 0.0103996, + "balance_loss_clip": 1.0236001, + "balance_loss_mlp": 1.03007793, + "epoch": 0.2225695905729574, + "flos": 15632921111040.0, + "grad_norm": 1.8651235659596215, + "language_loss": 0.70679104, + "learning_rate": 3.623601691205643e-06, + "loss": 0.72910708, + "num_input_tokens_seen": 39391200, + "step": 1851, + "time_per_iteration": 2.675015449523926 + }, + { + "auxiliary_loss_clip": 0.0118334, + "auxiliary_loss_mlp": 0.01040103, + "balance_loss_clip": 1.0211308, + "balance_loss_mlp": 1.03045249, + "epoch": 0.22268983346359647, + "flos": 25373582265600.0, + "grad_norm": 1.9939416366180736, + "language_loss": 0.81619048, + "learning_rate": 3.623146699723729e-06, + "loss": 0.83842492, + "num_input_tokens_seen": 39410660, + "step": 1852, + "time_per_iteration": 2.6867852210998535 + }, + { + "auxiliary_loss_clip": 0.01165727, + "auxiliary_loss_mlp": 0.01032333, + "balance_loss_clip": 0.98971045, + "balance_loss_mlp": 1.02297556, + "epoch": 0.22281007635423555, + "flos": 13261452359040.0, + "grad_norm": 2.307265628456708, + "language_loss": 0.77229822, + "learning_rate": 3.6226914620117507e-06, + "loss": 0.7942788, + "num_input_tokens_seen": 39429280, + "step": 1853, + "time_per_iteration": 2.695845603942871 + }, + { + "auxiliary_loss_clip": 0.01137954, + "auxiliary_loss_mlp": 0.01033411, + "balance_loss_clip": 0.94113809, + "balance_loss_mlp": 1.02383804, + "epoch": 0.22293031924487464, + "flos": 15340536403200.0, + "grad_norm": 2.048089338978703, + "language_loss": 0.8036533, + "learning_rate": 3.622235978138768e-06, + "loss": 0.82536691, + "num_input_tokens_seen": 39446905, + "step": 1854, + "time_per_iteration": 2.667235851287842 + }, + { + "auxiliary_loss_clip": 0.01189238, + "auxiliary_loss_mlp": 0.0103927, + "balance_loss_clip": 1.02807796, + "balance_loss_mlp": 1.02931619, + "epoch": 0.22305056213551375, + "flos": 22564649773440.0, + "grad_norm": 2.573361154057098, + "language_loss": 0.81173182, + "learning_rate": 3.621780248173877e-06, + "loss": 0.83401692, + "num_input_tokens_seen": 39465105, + "step": 1855, + "time_per_iteration": 2.6742935180664062 + }, + { + "auxiliary_loss_clip": 0.01095406, + "auxiliary_loss_mlp": 0.0100344, + "balance_loss_clip": 0.99648756, + "balance_loss_mlp": 1.00123465, + "epoch": 0.22317080502615283, + "flos": 64880419887360.0, + "grad_norm": 0.8261422681297206, + "language_loss": 0.61070359, + "learning_rate": 3.6213242721862125e-06, + "loss": 0.63169205, + "num_input_tokens_seen": 39523560, + "step": 1856, + "time_per_iteration": 3.2404298782348633 + }, + { + "auxiliary_loss_clip": 0.01156463, + "auxiliary_loss_mlp": 0.01035467, + "balance_loss_clip": 0.98591644, + "balance_loss_mlp": 1.02598977, + "epoch": 0.2232910479167919, + "flos": 25775997310080.0, + "grad_norm": 1.4796796016305152, + "language_loss": 0.75108176, + "learning_rate": 3.620868050244945e-06, + "loss": 0.77300107, + "num_input_tokens_seen": 39544040, + "step": 1857, + "time_per_iteration": 2.6736061573028564 + }, + { + "auxiliary_loss_clip": 0.01160086, + "auxiliary_loss_mlp": 0.0104016, + "balance_loss_clip": 0.98502862, + "balance_loss_mlp": 1.02977717, + "epoch": 0.22341129080743102, + "flos": 23251799928960.0, + "grad_norm": 1.6773804516175206, + "language_loss": 0.77606392, + "learning_rate": 3.6204115824192817e-06, + "loss": 0.79806638, + "num_input_tokens_seen": 39561515, + "step": 1858, + "time_per_iteration": 2.784102439880371 + }, + { + "auxiliary_loss_clip": 0.01155307, + "auxiliary_loss_mlp": 0.01034903, + "balance_loss_clip": 0.98130763, + "balance_loss_mlp": 1.02425802, + "epoch": 0.2235315336980701, + "flos": 21214552250880.0, + "grad_norm": 2.1081938978014665, + "language_loss": 0.76411223, + "learning_rate": 3.619954868778471e-06, + "loss": 0.78601432, + "num_input_tokens_seen": 39578210, + "step": 1859, + "time_per_iteration": 2.6320064067840576 + }, + { + "auxiliary_loss_clip": 0.0116183, + "auxiliary_loss_mlp": 0.01038812, + "balance_loss_clip": 0.98420393, + "balance_loss_mlp": 1.02948403, + "epoch": 0.2236517765887092, + "flos": 19901945548800.0, + "grad_norm": 1.926571841055346, + "language_loss": 0.82886505, + "learning_rate": 3.6194979093917944e-06, + "loss": 0.85087144, + "num_input_tokens_seen": 39597625, + "step": 1860, + "time_per_iteration": 2.6751182079315186 + }, + { + "auxiliary_loss_clip": 0.01158461, + "auxiliary_loss_mlp": 0.01037822, + "balance_loss_clip": 0.98463142, + "balance_loss_mlp": 1.028458, + "epoch": 0.22377201947934827, + "flos": 23214847812480.0, + "grad_norm": 1.8131703978205935, + "language_loss": 0.86703956, + "learning_rate": 3.6190407043285724e-06, + "loss": 0.88900244, + "num_input_tokens_seen": 39615360, + "step": 1861, + "time_per_iteration": 2.6676671504974365 + }, + { + "auxiliary_loss_clip": 0.01214846, + "auxiliary_loss_mlp": 0.01035366, + "balance_loss_clip": 1.06347311, + "balance_loss_mlp": 1.02497709, + "epoch": 0.22389226236998738, + "flos": 26794244056320.0, + "grad_norm": 2.0033592052203995, + "language_loss": 0.75570619, + "learning_rate": 3.618583253658163e-06, + "loss": 0.77820832, + "num_input_tokens_seen": 39635460, + "step": 1862, + "time_per_iteration": 3.6391491889953613 + }, + { + "auxiliary_loss_clip": 0.01116571, + "auxiliary_loss_mlp": 0.00512672, + "balance_loss_clip": 0.90845054, + "balance_loss_mlp": 0.0, + "epoch": 0.22401250526062647, + "flos": 24170359455360.0, + "grad_norm": 1.9208738800100629, + "language_loss": 0.86767441, + "learning_rate": 3.618125557449961e-06, + "loss": 0.88396686, + "num_input_tokens_seen": 39653515, + "step": 1863, + "time_per_iteration": 3.7153220176696777 + }, + { + "auxiliary_loss_clip": 0.01185006, + "auxiliary_loss_mlp": 0.01036166, + "balance_loss_clip": 1.02408266, + "balance_loss_mlp": 1.0260396, + "epoch": 0.22413274815126555, + "flos": 16759761649920.0, + "grad_norm": 2.247050101575428, + "language_loss": 0.82979858, + "learning_rate": 3.6176676157733983e-06, + "loss": 0.85201031, + "num_input_tokens_seen": 39668525, + "step": 1864, + "time_per_iteration": 2.5748400688171387 + }, + { + "auxiliary_loss_clip": 0.01133472, + "auxiliary_loss_mlp": 0.01040569, + "balance_loss_clip": 0.94534284, + "balance_loss_mlp": 1.03108597, + "epoch": 0.22425299104190466, + "flos": 21360205900800.0, + "grad_norm": 2.3601324395084577, + "language_loss": 0.75785583, + "learning_rate": 3.6172094286979443e-06, + "loss": 0.77959621, + "num_input_tokens_seen": 39685895, + "step": 1865, + "time_per_iteration": 2.7637462615966797 + }, + { + "auxiliary_loss_clip": 0.01162307, + "auxiliary_loss_mlp": 0.01036214, + "balance_loss_clip": 0.98333001, + "balance_loss_mlp": 1.02670717, + "epoch": 0.22437323393254374, + "flos": 32165547108480.0, + "grad_norm": 1.629832180785799, + "language_loss": 0.81347138, + "learning_rate": 3.6167509962931064e-06, + "loss": 0.83545661, + "num_input_tokens_seen": 39711595, + "step": 1866, + "time_per_iteration": 3.77738618850708 + }, + { + "auxiliary_loss_clip": 0.01123018, + "auxiliary_loss_mlp": 0.01045721, + "balance_loss_clip": 0.90977561, + "balance_loss_mlp": 1.03633356, + "epoch": 0.22449347682318282, + "flos": 18002809664640.0, + "grad_norm": 3.5560207955412135, + "language_loss": 0.76791775, + "learning_rate": 3.6162923186284276e-06, + "loss": 0.78960514, + "num_input_tokens_seen": 39727555, + "step": 1867, + "time_per_iteration": 3.727421522140503 + }, + { + "auxiliary_loss_clip": 0.01160759, + "auxiliary_loss_mlp": 0.01039584, + "balance_loss_clip": 0.98299325, + "balance_loss_mlp": 1.02993369, + "epoch": 0.2246137197138219, + "flos": 18697286194560.0, + "grad_norm": 1.8670766696963463, + "language_loss": 0.85837227, + "learning_rate": 3.6158333957734888e-06, + "loss": 0.88037568, + "num_input_tokens_seen": 39746145, + "step": 1868, + "time_per_iteration": 2.783355236053467 + }, + { + "auxiliary_loss_clip": 0.01143366, + "auxiliary_loss_mlp": 0.01043407, + "balance_loss_clip": 0.9457351, + "balance_loss_mlp": 1.03390634, + "epoch": 0.22473396260446102, + "flos": 15590653781760.0, + "grad_norm": 2.291651495023044, + "language_loss": 0.82784081, + "learning_rate": 3.6153742277979088e-06, + "loss": 0.8497085, + "num_input_tokens_seen": 39763575, + "step": 1869, + "time_per_iteration": 2.7321126461029053 + }, + { + "auxiliary_loss_clip": 0.01165628, + "auxiliary_loss_mlp": 0.01040399, + "balance_loss_clip": 0.98396152, + "balance_loss_mlp": 1.03138113, + "epoch": 0.2248542054951001, + "flos": 14465501182080.0, + "grad_norm": 2.349083344834545, + "language_loss": 0.78164864, + "learning_rate": 3.6149148147713434e-06, + "loss": 0.80370891, + "num_input_tokens_seen": 39781810, + "step": 1870, + "time_per_iteration": 2.7269575595855713 + }, + { + "auxiliary_loss_clip": 0.01195998, + "auxiliary_loss_mlp": 0.01034508, + "balance_loss_clip": 1.02898407, + "balance_loss_mlp": 1.02518559, + "epoch": 0.22497444838573918, + "flos": 19243882431360.0, + "grad_norm": 2.0007540788623337, + "language_loss": 0.86172807, + "learning_rate": 3.614455156763484e-06, + "loss": 0.88403314, + "num_input_tokens_seen": 39800115, + "step": 1871, + "time_per_iteration": 2.7615013122558594 + }, + { + "auxiliary_loss_clip": 0.01108689, + "auxiliary_loss_mlp": 0.01038017, + "balance_loss_clip": 0.90299869, + "balance_loss_mlp": 1.028337, + "epoch": 0.2250946912763783, + "flos": 16910299549440.0, + "grad_norm": 2.2833310485965272, + "language_loss": 0.7124418, + "learning_rate": 3.613995253844061e-06, + "loss": 0.73390883, + "num_input_tokens_seen": 39817795, + "step": 1872, + "time_per_iteration": 3.398311138153076 + }, + { + "auxiliary_loss_clip": 0.0118501, + "auxiliary_loss_mlp": 0.01032378, + "balance_loss_clip": 1.02434468, + "balance_loss_mlp": 1.02268028, + "epoch": 0.22521493416701738, + "flos": 24681368292480.0, + "grad_norm": 1.9850482640003195, + "language_loss": 0.80690354, + "learning_rate": 3.6135351060828414e-06, + "loss": 0.82907748, + "num_input_tokens_seen": 39838270, + "step": 1873, + "time_per_iteration": 3.2080512046813965 + }, + { + "auxiliary_loss_clip": 0.01217635, + "auxiliary_loss_mlp": 0.01043051, + "balance_loss_clip": 1.06578636, + "balance_loss_mlp": 1.03255463, + "epoch": 0.22533517705765646, + "flos": 17821963664640.0, + "grad_norm": 2.07823487399951, + "language_loss": 0.69449246, + "learning_rate": 3.6130747135496285e-06, + "loss": 0.71709931, + "num_input_tokens_seen": 39857270, + "step": 1874, + "time_per_iteration": 3.1784873008728027 + }, + { + "auxiliary_loss_clip": 0.01211558, + "auxiliary_loss_mlp": 0.01028845, + "balance_loss_clip": 1.0632, + "balance_loss_mlp": 1.0189147, + "epoch": 0.22545541994829554, + "flos": 33691390899840.0, + "grad_norm": 2.0292629241417357, + "language_loss": 0.65584642, + "learning_rate": 3.6126140763142646e-06, + "loss": 0.67825043, + "num_input_tokens_seen": 39882300, + "step": 1875, + "time_per_iteration": 3.408372163772583 + }, + { + "auxiliary_loss_clip": 0.01214885, + "auxiliary_loss_mlp": 0.0103822, + "balance_loss_clip": 1.06563294, + "balance_loss_mlp": 1.02871943, + "epoch": 0.22557566283893465, + "flos": 19171594310400.0, + "grad_norm": 2.3865515556445063, + "language_loss": 0.86066163, + "learning_rate": 3.6121531944466275e-06, + "loss": 0.88319266, + "num_input_tokens_seen": 39899625, + "step": 1876, + "time_per_iteration": 2.856015920639038 + }, + { + "auxiliary_loss_clip": 0.01185657, + "auxiliary_loss_mlp": 0.01032125, + "balance_loss_clip": 1.02458751, + "balance_loss_mlp": 1.02325583, + "epoch": 0.22569590572957374, + "flos": 20773281669120.0, + "grad_norm": 2.2830597746276866, + "language_loss": 0.78206342, + "learning_rate": 3.611692068016633e-06, + "loss": 0.80424124, + "num_input_tokens_seen": 39915955, + "step": 1877, + "time_per_iteration": 2.6269514560699463 + }, + { + "auxiliary_loss_clip": 0.01135188, + "auxiliary_loss_mlp": 0.01041687, + "balance_loss_clip": 0.94404149, + "balance_loss_mlp": 1.0309459, + "epoch": 0.22581614862021282, + "flos": 18442715529600.0, + "grad_norm": 2.367651438136522, + "language_loss": 0.74466884, + "learning_rate": 3.611230697094233e-06, + "loss": 0.76643753, + "num_input_tokens_seen": 39932655, + "step": 1878, + "time_per_iteration": 2.7197554111480713 + }, + { + "auxiliary_loss_clip": 0.01166774, + "auxiliary_loss_mlp": 0.01026798, + "balance_loss_clip": 0.98619819, + "balance_loss_mlp": 1.01766074, + "epoch": 0.22593639151085193, + "flos": 20048389297920.0, + "grad_norm": 2.0261231728165616, + "language_loss": 0.87017763, + "learning_rate": 3.6107690817494173e-06, + "loss": 0.89211333, + "num_input_tokens_seen": 39952875, + "step": 1879, + "time_per_iteration": 2.661062240600586 + }, + { + "auxiliary_loss_clip": 0.01110568, + "auxiliary_loss_mlp": 0.01030027, + "balance_loss_clip": 0.90392935, + "balance_loss_mlp": 1.02013838, + "epoch": 0.226056634401491, + "flos": 13115116350720.0, + "grad_norm": 2.1363586479719325, + "language_loss": 0.7066825, + "learning_rate": 3.6103072220522117e-06, + "loss": 0.72808844, + "num_input_tokens_seen": 39968405, + "step": 1880, + "time_per_iteration": 2.7959446907043457 + }, + { + "auxiliary_loss_clip": 0.01145227, + "auxiliary_loss_mlp": 0.01039587, + "balance_loss_clip": 0.94769502, + "balance_loss_mlp": 1.02981794, + "epoch": 0.2261768772921301, + "flos": 18988378012800.0, + "grad_norm": 1.667657663148834, + "language_loss": 0.91454244, + "learning_rate": 3.609845118072682e-06, + "loss": 0.93639064, + "num_input_tokens_seen": 39987075, + "step": 1881, + "time_per_iteration": 2.8933255672454834 + }, + { + "auxiliary_loss_clip": 0.01193801, + "auxiliary_loss_mlp": 0.00512752, + "balance_loss_clip": 1.02588558, + "balance_loss_mlp": 0.0, + "epoch": 0.2262971201827692, + "flos": 19974054101760.0, + "grad_norm": 1.7301498737802543, + "language_loss": 0.79883766, + "learning_rate": 3.6093827698809276e-06, + "loss": 0.81590319, + "num_input_tokens_seen": 40006175, + "step": 1882, + "time_per_iteration": 2.878488779067993 + }, + { + "auxiliary_loss_clip": 0.01185396, + "auxiliary_loss_mlp": 0.0103541, + "balance_loss_clip": 1.02175856, + "balance_loss_mlp": 1.02637982, + "epoch": 0.2264173630734083, + "flos": 16654543735680.0, + "grad_norm": 2.112383949037023, + "language_loss": 0.84341347, + "learning_rate": 3.6089201775470864e-06, + "loss": 0.86562157, + "num_input_tokens_seen": 40021630, + "step": 1883, + "time_per_iteration": 2.8335483074188232 + }, + { + "auxiliary_loss_clip": 0.01127547, + "auxiliary_loss_mlp": 0.0103439, + "balance_loss_clip": 0.94436598, + "balance_loss_mlp": 1.02516937, + "epoch": 0.22653760596404737, + "flos": 24389809597440.0, + "grad_norm": 2.183256431226928, + "language_loss": 0.77483165, + "learning_rate": 3.6084573411413334e-06, + "loss": 0.79645097, + "num_input_tokens_seen": 40041025, + "step": 1884, + "time_per_iteration": 2.805116653442383 + }, + { + "auxiliary_loss_clip": 0.01138181, + "auxiliary_loss_mlp": 0.01029549, + "balance_loss_clip": 0.94686353, + "balance_loss_mlp": 1.01964867, + "epoch": 0.22665784885468646, + "flos": 18332541538560.0, + "grad_norm": 1.879902635700302, + "language_loss": 0.81022078, + "learning_rate": 3.607994260733881e-06, + "loss": 0.83189809, + "num_input_tokens_seen": 40060265, + "step": 1885, + "time_per_iteration": 2.9343302249908447 + }, + { + "auxiliary_loss_clip": 0.01177377, + "auxiliary_loss_mlp": 0.01035833, + "balance_loss_clip": 1.02162862, + "balance_loss_mlp": 1.02706552, + "epoch": 0.22677809174532557, + "flos": 24058102475520.0, + "grad_norm": 1.6732285596348822, + "language_loss": 0.74928588, + "learning_rate": 3.6075309363949776e-06, + "loss": 0.77141798, + "num_input_tokens_seen": 40079435, + "step": 1886, + "time_per_iteration": 2.707692861557007 + }, + { + "auxiliary_loss_clip": 0.01213751, + "auxiliary_loss_mlp": 0.01039248, + "balance_loss_clip": 1.06480002, + "balance_loss_mlp": 1.02907348, + "epoch": 0.22689833463596465, + "flos": 20374242503040.0, + "grad_norm": 2.347199000292276, + "language_loss": 0.81332153, + "learning_rate": 3.6070673681949094e-06, + "loss": 0.83585149, + "num_input_tokens_seen": 40097800, + "step": 1887, + "time_per_iteration": 2.634211778640747 + }, + { + "auxiliary_loss_clip": 0.01168416, + "auxiliary_loss_mlp": 0.0051174, + "balance_loss_clip": 0.98880577, + "balance_loss_mlp": 0.0, + "epoch": 0.22701857752660373, + "flos": 30120398438400.0, + "grad_norm": 1.6835491318356934, + "language_loss": 0.81507009, + "learning_rate": 3.606603556203999e-06, + "loss": 0.83187163, + "num_input_tokens_seen": 40122745, + "step": 1888, + "time_per_iteration": 3.7438132762908936 + }, + { + "auxiliary_loss_clip": 0.01187366, + "auxiliary_loss_mlp": 0.01039555, + "balance_loss_clip": 1.02213967, + "balance_loss_mlp": 1.02933931, + "epoch": 0.22713882041724284, + "flos": 22492182084480.0, + "grad_norm": 1.7544417001877028, + "language_loss": 0.83372182, + "learning_rate": 3.6061395004926066e-06, + "loss": 0.85599107, + "num_input_tokens_seen": 40141680, + "step": 1889, + "time_per_iteration": 3.6195523738861084 + }, + { + "auxiliary_loss_clip": 0.01212116, + "auxiliary_loss_mlp": 0.01029481, + "balance_loss_clip": 1.06175005, + "balance_loss_mlp": 1.0197953, + "epoch": 0.22725906330788193, + "flos": 20521548178560.0, + "grad_norm": 2.108212977107228, + "language_loss": 0.84719443, + "learning_rate": 3.605675201131129e-06, + "loss": 0.86961043, + "num_input_tokens_seen": 40160140, + "step": 1890, + "time_per_iteration": 2.701949119567871 + }, + { + "auxiliary_loss_clip": 0.01193334, + "auxiliary_loss_mlp": 0.01040386, + "balance_loss_clip": 1.02756476, + "balance_loss_mlp": 1.03143346, + "epoch": 0.227379306198521, + "flos": 18989922297600.0, + "grad_norm": 2.3877397289571918, + "language_loss": 0.7945621, + "learning_rate": 3.60521065819e-06, + "loss": 0.8168993, + "num_input_tokens_seen": 40177450, + "step": 1891, + "time_per_iteration": 2.723068952560425 + }, + { + "auxiliary_loss_clip": 0.01168525, + "auxiliary_loss_mlp": 0.01035151, + "balance_loss_clip": 0.98572934, + "balance_loss_mlp": 1.0255425, + "epoch": 0.2274995490891601, + "flos": 21798351999360.0, + "grad_norm": 1.7283585962541186, + "language_loss": 0.87535965, + "learning_rate": 3.60474587173969e-06, + "loss": 0.89739639, + "num_input_tokens_seen": 40195935, + "step": 1892, + "time_per_iteration": 4.571485757827759 + }, + { + "auxiliary_loss_clip": 0.01187261, + "auxiliary_loss_mlp": 0.01028766, + "balance_loss_clip": 1.02538133, + "balance_loss_mlp": 1.01936603, + "epoch": 0.2276197919797992, + "flos": 19058654972160.0, + "grad_norm": 1.9648160476903211, + "language_loss": 0.84162915, + "learning_rate": 3.6042808418507084e-06, + "loss": 0.86378944, + "num_input_tokens_seen": 40213620, + "step": 1893, + "time_per_iteration": 2.669426918029785 + }, + { + "auxiliary_loss_clip": 0.01190794, + "auxiliary_loss_mlp": 0.0103894, + "balance_loss_clip": 1.02730227, + "balance_loss_mlp": 1.02861691, + "epoch": 0.22774003487043828, + "flos": 18806777827200.0, + "grad_norm": 2.4265390225727748, + "language_loss": 0.76968122, + "learning_rate": 3.6038155685935976e-06, + "loss": 0.7919786, + "num_input_tokens_seen": 40230190, + "step": 1894, + "time_per_iteration": 2.581418991088867 + }, + { + "auxiliary_loss_clip": 0.01184736, + "auxiliary_loss_mlp": 0.01037676, + "balance_loss_clip": 1.02334428, + "balance_loss_mlp": 1.02808571, + "epoch": 0.22786027776107737, + "flos": 23002544476800.0, + "grad_norm": 2.4983054373634497, + "language_loss": 0.70535135, + "learning_rate": 3.6033500520389404e-06, + "loss": 0.72757548, + "num_input_tokens_seen": 40246860, + "step": 1895, + "time_per_iteration": 2.6673383712768555 + }, + { + "auxiliary_loss_clip": 0.01033531, + "auxiliary_loss_mlp": 0.01000746, + "balance_loss_clip": 0.88330734, + "balance_loss_mlp": 0.99857599, + "epoch": 0.22798052065171648, + "flos": 66706872600960.0, + "grad_norm": 0.8360138431079579, + "language_loss": 0.64751816, + "learning_rate": 3.6028842922573553e-06, + "loss": 0.66786098, + "num_input_tokens_seen": 40311005, + "step": 1896, + "time_per_iteration": 3.398848295211792 + }, + { + "auxiliary_loss_clip": 0.0105469, + "auxiliary_loss_mlp": 0.0050449, + "balance_loss_clip": 0.91939378, + "balance_loss_mlp": 0.0, + "epoch": 0.22810076354235556, + "flos": 62080896758400.0, + "grad_norm": 0.8577788886771962, + "language_loss": 0.62885106, + "learning_rate": 3.602418289319497e-06, + "loss": 0.6444428, + "num_input_tokens_seen": 40369560, + "step": 1897, + "time_per_iteration": 3.224560499191284 + }, + { + "auxiliary_loss_clip": 0.01114438, + "auxiliary_loss_mlp": 0.01039106, + "balance_loss_clip": 0.90662915, + "balance_loss_mlp": 1.029176, + "epoch": 0.22822100643299464, + "flos": 23876358635520.0, + "grad_norm": 1.86580584828171, + "language_loss": 0.73265916, + "learning_rate": 3.601952043296059e-06, + "loss": 0.75419462, + "num_input_tokens_seen": 40389555, + "step": 1898, + "time_per_iteration": 2.8121612071990967 + }, + { + "auxiliary_loss_clip": 0.01166406, + "auxiliary_loss_mlp": 0.01029585, + "balance_loss_clip": 0.98233694, + "balance_loss_mlp": 1.01939249, + "epoch": 0.22834124932363373, + "flos": 20991331180800.0, + "grad_norm": 8.609869431519419, + "language_loss": 0.80898196, + "learning_rate": 3.6014855542577696e-06, + "loss": 0.83094186, + "num_input_tokens_seen": 40406765, + "step": 1899, + "time_per_iteration": 2.6317975521087646 + }, + { + "auxiliary_loss_clip": 0.01162516, + "auxiliary_loss_mlp": 0.01038627, + "balance_loss_clip": 0.98630905, + "balance_loss_mlp": 1.02926326, + "epoch": 0.22846149221427284, + "flos": 24901572620160.0, + "grad_norm": 2.1857629865155106, + "language_loss": 0.84040725, + "learning_rate": 3.6010188222753943e-06, + "loss": 0.86241865, + "num_input_tokens_seen": 40427535, + "step": 1900, + "time_per_iteration": 2.767916202545166 + }, + { + "auxiliary_loss_clip": 0.01065084, + "auxiliary_loss_mlp": 0.01002266, + "balance_loss_clip": 0.95286536, + "balance_loss_mlp": 1.00007236, + "epoch": 0.22858173510491192, + "flos": 56132294319360.0, + "grad_norm": 0.8912332176979424, + "language_loss": 0.64187121, + "learning_rate": 3.6005518474197372e-06, + "loss": 0.66254473, + "num_input_tokens_seen": 40479580, + "step": 1901, + "time_per_iteration": 3.1523518562316895 + }, + { + "auxiliary_loss_clip": 0.01188107, + "auxiliary_loss_mlp": 0.01036284, + "balance_loss_clip": 1.02678812, + "balance_loss_mlp": 1.02531695, + "epoch": 0.228701977995551, + "flos": 24170826332160.0, + "grad_norm": 2.225176565794362, + "language_loss": 0.78430951, + "learning_rate": 3.6000846297616373e-06, + "loss": 0.80655336, + "num_input_tokens_seen": 40497880, + "step": 1902, + "time_per_iteration": 2.7172930240631104 + }, + { + "auxiliary_loss_clip": 0.01217117, + "auxiliary_loss_mlp": 0.01042929, + "balance_loss_clip": 1.06773663, + "balance_loss_mlp": 1.03218246, + "epoch": 0.22882222088619011, + "flos": 21387892308480.0, + "grad_norm": 2.1703746032757283, + "language_loss": 0.72971821, + "learning_rate": 3.5996171693719717e-06, + "loss": 0.75231874, + "num_input_tokens_seen": 40513975, + "step": 1903, + "time_per_iteration": 2.6320035457611084 + }, + { + "auxiliary_loss_clip": 0.01086494, + "auxiliary_loss_mlp": 0.01005387, + "balance_loss_clip": 0.98983616, + "balance_loss_mlp": 1.00306213, + "epoch": 0.2289424637768292, + "flos": 64589615377920.0, + "grad_norm": 0.839085803375614, + "language_loss": 0.64869028, + "learning_rate": 3.5991494663216528e-06, + "loss": 0.66960907, + "num_input_tokens_seen": 40576960, + "step": 1904, + "time_per_iteration": 3.2723538875579834 + }, + { + "auxiliary_loss_clip": 0.01214425, + "auxiliary_loss_mlp": 0.01039774, + "balance_loss_clip": 1.06543005, + "balance_loss_mlp": 1.03010058, + "epoch": 0.22906270666746828, + "flos": 22163419877760.0, + "grad_norm": 2.3920075312012106, + "language_loss": 0.8765139, + "learning_rate": 3.5986815206816314e-06, + "loss": 0.89905584, + "num_input_tokens_seen": 40595780, + "step": 1905, + "time_per_iteration": 2.6736044883728027 + }, + { + "auxiliary_loss_clip": 0.01212634, + "auxiliary_loss_mlp": 0.01034125, + "balance_loss_clip": 1.06288767, + "balance_loss_mlp": 1.02436757, + "epoch": 0.2291829495581074, + "flos": 25772334122880.0, + "grad_norm": 2.2273790459748977, + "language_loss": 0.74486554, + "learning_rate": 3.598213332522895e-06, + "loss": 0.76733309, + "num_input_tokens_seen": 40615810, + "step": 1906, + "time_per_iteration": 2.606790781021118 + }, + { + "auxiliary_loss_clip": 0.0118616, + "auxiliary_loss_mlp": 0.01037357, + "balance_loss_clip": 1.02308154, + "balance_loss_mlp": 1.0279876, + "epoch": 0.22930319244874647, + "flos": 31172760126720.0, + "grad_norm": 2.082488259833898, + "language_loss": 0.77419841, + "learning_rate": 3.597744901916466e-06, + "loss": 0.79643357, + "num_input_tokens_seen": 40637095, + "step": 1907, + "time_per_iteration": 2.7426185607910156 + }, + { + "auxiliary_loss_clip": 0.01214101, + "auxiliary_loss_mlp": 0.01033347, + "balance_loss_clip": 1.06177878, + "balance_loss_mlp": 1.0227375, + "epoch": 0.22942343533938556, + "flos": 23254098399360.0, + "grad_norm": 1.8785132396995263, + "language_loss": 0.76663059, + "learning_rate": 3.5972762289334058e-06, + "loss": 0.78910506, + "num_input_tokens_seen": 40656725, + "step": 1908, + "time_per_iteration": 2.603513240814209 + }, + { + "auxiliary_loss_clip": 0.01089965, + "auxiliary_loss_mlp": 0.0103785, + "balance_loss_clip": 0.8703692, + "balance_loss_mlp": 1.02797329, + "epoch": 0.22954367823002464, + "flos": 14610903436800.0, + "grad_norm": 10.712476166362817, + "language_loss": 0.8461678, + "learning_rate": 3.5968073136448116e-06, + "loss": 0.86744595, + "num_input_tokens_seen": 40674745, + "step": 1909, + "time_per_iteration": 2.764889717102051 + }, + { + "auxiliary_loss_clip": 0.01191468, + "auxiliary_loss_mlp": 0.01037688, + "balance_loss_clip": 1.02416611, + "balance_loss_mlp": 1.02666116, + "epoch": 0.22966392112066375, + "flos": 16763604405120.0, + "grad_norm": 1.7143232419927865, + "language_loss": 0.91115862, + "learning_rate": 3.596338156121818e-06, + "loss": 0.93345016, + "num_input_tokens_seen": 40693630, + "step": 1910, + "time_per_iteration": 2.5957095623016357 + }, + { + "auxiliary_loss_clip": 0.01065055, + "auxiliary_loss_mlp": 0.01003746, + "balance_loss_clip": 0.95333076, + "balance_loss_mlp": 1.00136173, + "epoch": 0.22978416401130283, + "flos": 67474247783040.0, + "grad_norm": 0.862234326429634, + "language_loss": 0.59343231, + "learning_rate": 3.595868756435595e-06, + "loss": 0.61412036, + "num_input_tokens_seen": 40761310, + "step": 1911, + "time_per_iteration": 3.334805727005005 + }, + { + "auxiliary_loss_clip": 0.01145538, + "auxiliary_loss_mlp": 0.01036415, + "balance_loss_clip": 0.95228279, + "balance_loss_mlp": 1.02554321, + "epoch": 0.22990440690194192, + "flos": 19865137086720.0, + "grad_norm": 2.2032485427166097, + "language_loss": 0.80249542, + "learning_rate": 3.5953991146573504e-06, + "loss": 0.82431495, + "num_input_tokens_seen": 40779955, + "step": 1912, + "time_per_iteration": 2.7111012935638428 + }, + { + "auxiliary_loss_clip": 0.01188008, + "auxiliary_loss_mlp": 0.01043923, + "balance_loss_clip": 1.02038753, + "balance_loss_mlp": 1.03362381, + "epoch": 0.23002464979258103, + "flos": 13289246507520.0, + "grad_norm": 3.665661342196764, + "language_loss": 0.83609927, + "learning_rate": 3.5949292308583294e-06, + "loss": 0.85841864, + "num_input_tokens_seen": 40793200, + "step": 1913, + "time_per_iteration": 2.5117456912994385 + }, + { + "auxiliary_loss_clip": 0.01211618, + "auxiliary_loss_mlp": 0.01053398, + "balance_loss_clip": 1.06343114, + "balance_loss_mlp": 1.04351008, + "epoch": 0.2301448926832201, + "flos": 22163779013760.0, + "grad_norm": 4.996119872687659, + "language_loss": 0.80768085, + "learning_rate": 3.594459105109811e-06, + "loss": 0.83033103, + "num_input_tokens_seen": 40812380, + "step": 1914, + "time_per_iteration": 2.6065425872802734 + }, + { + "auxiliary_loss_clip": 0.01190919, + "auxiliary_loss_mlp": 0.01044437, + "balance_loss_clip": 1.02583396, + "balance_loss_mlp": 1.0355202, + "epoch": 0.2302651355738592, + "flos": 20704477167360.0, + "grad_norm": 1.8000719034563877, + "language_loss": 0.81308877, + "learning_rate": 3.593988737483115e-06, + "loss": 0.8354423, + "num_input_tokens_seen": 40832320, + "step": 1915, + "time_per_iteration": 4.466572999954224 + }, + { + "auxiliary_loss_clip": 0.01165671, + "auxiliary_loss_mlp": 0.01040498, + "balance_loss_clip": 0.98806554, + "balance_loss_mlp": 1.0305264, + "epoch": 0.23038537846449827, + "flos": 18588943797120.0, + "grad_norm": 2.140630947163473, + "language_loss": 0.78232193, + "learning_rate": 3.5935181280495947e-06, + "loss": 0.80438364, + "num_input_tokens_seen": 40850900, + "step": 1916, + "time_per_iteration": 2.633253574371338 + }, + { + "auxiliary_loss_clip": 0.01062662, + "auxiliary_loss_mlp": 0.01005575, + "balance_loss_clip": 0.95464575, + "balance_loss_mlp": 1.00336981, + "epoch": 0.23050562135513739, + "flos": 64224260190720.0, + "grad_norm": 0.8013302207596879, + "language_loss": 0.5426026, + "learning_rate": 3.5930472768806412e-06, + "loss": 0.56328499, + "num_input_tokens_seen": 40909570, + "step": 1917, + "time_per_iteration": 3.1985199451446533 + }, + { + "auxiliary_loss_clip": 0.01215651, + "auxiliary_loss_mlp": 0.01040193, + "balance_loss_clip": 1.06797588, + "balance_loss_mlp": 1.03052545, + "epoch": 0.23062586424577647, + "flos": 17313396952320.0, + "grad_norm": 1.9513476019847318, + "language_loss": 0.7730068, + "learning_rate": 3.5925761840476826e-06, + "loss": 0.79556525, + "num_input_tokens_seen": 40928180, + "step": 1918, + "time_per_iteration": 3.5347564220428467 + }, + { + "auxiliary_loss_clip": 0.01161068, + "auxiliary_loss_mlp": 0.01035318, + "balance_loss_clip": 0.98724574, + "balance_loss_mlp": 1.02535832, + "epoch": 0.23074610713641555, + "flos": 27855979194240.0, + "grad_norm": 1.8755551712978664, + "language_loss": 0.81564569, + "learning_rate": 3.592104849622183e-06, + "loss": 0.83760953, + "num_input_tokens_seen": 40950435, + "step": 1919, + "time_per_iteration": 3.6438021659851074 + }, + { + "auxiliary_loss_clip": 0.011086, + "auxiliary_loss_mlp": 0.01036668, + "balance_loss_clip": 0.90843982, + "balance_loss_mlp": 1.02731633, + "epoch": 0.23086635002705466, + "flos": 28841798937600.0, + "grad_norm": 1.5154793688021564, + "language_loss": 0.73059523, + "learning_rate": 3.591633273675644e-06, + "loss": 0.7520479, + "num_input_tokens_seen": 40972670, + "step": 1920, + "time_per_iteration": 2.837252140045166 + }, + { + "auxiliary_loss_clip": 0.01041335, + "auxiliary_loss_mlp": 0.01010837, + "balance_loss_clip": 0.92532933, + "balance_loss_mlp": 1.00861955, + "epoch": 0.23098659291769374, + "flos": 62923681566720.0, + "grad_norm": 0.9103172435929237, + "language_loss": 0.58167571, + "learning_rate": 3.591161456279602e-06, + "loss": 0.60219741, + "num_input_tokens_seen": 41018215, + "step": 1921, + "time_per_iteration": 3.1006977558135986 + }, + { + "auxiliary_loss_clip": 0.01168115, + "auxiliary_loss_mlp": 0.01030137, + "balance_loss_clip": 0.9850589, + "balance_loss_mlp": 1.0208745, + "epoch": 0.23110683580833283, + "flos": 23476816679040.0, + "grad_norm": 2.377626140866749, + "language_loss": 0.80200994, + "learning_rate": 3.590689397505633e-06, + "loss": 0.82399237, + "num_input_tokens_seen": 41039125, + "step": 1922, + "time_per_iteration": 2.7242379188537598 + }, + { + "auxiliary_loss_clip": 0.01211374, + "auxiliary_loss_mlp": 0.01029184, + "balance_loss_clip": 1.06474042, + "balance_loss_mlp": 1.02057099, + "epoch": 0.2312270786989719, + "flos": 27271066124160.0, + "grad_norm": 2.3526443191881405, + "language_loss": 0.867342, + "learning_rate": 3.590217097425347e-06, + "loss": 0.88974756, + "num_input_tokens_seen": 41059025, + "step": 1923, + "time_per_iteration": 2.657970666885376 + }, + { + "auxiliary_loss_clip": 0.01214138, + "auxiliary_loss_mlp": 0.01039907, + "balance_loss_clip": 1.06452024, + "balance_loss_mlp": 1.03012598, + "epoch": 0.23134732158961102, + "flos": 13261344618240.0, + "grad_norm": 2.205450018625263, + "language_loss": 0.70873511, + "learning_rate": 3.589744556110391e-06, + "loss": 0.73127556, + "num_input_tokens_seen": 41077015, + "step": 1924, + "time_per_iteration": 2.576838493347168 + }, + { + "auxiliary_loss_clip": 0.01160013, + "auxiliary_loss_mlp": 0.01033409, + "balance_loss_clip": 0.98443514, + "balance_loss_mlp": 1.02347255, + "epoch": 0.2314675644802501, + "flos": 36977648250240.0, + "grad_norm": 1.6651090288534318, + "language_loss": 0.84392607, + "learning_rate": 3.58927177363245e-06, + "loss": 0.86586028, + "num_input_tokens_seen": 41099840, + "step": 1925, + "time_per_iteration": 2.7582390308380127 + }, + { + "auxiliary_loss_clip": 0.01132967, + "auxiliary_loss_mlp": 0.01033981, + "balance_loss_clip": 0.94297481, + "balance_loss_mlp": 1.02319264, + "epoch": 0.2315878073708892, + "flos": 23842207779840.0, + "grad_norm": 2.285232753780566, + "language_loss": 0.72524369, + "learning_rate": 3.5887987500632447e-06, + "loss": 0.74691319, + "num_input_tokens_seen": 41117845, + "step": 1926, + "time_per_iteration": 2.692939519882202 + }, + { + "auxiliary_loss_clip": 0.01145708, + "auxiliary_loss_mlp": 0.0103741, + "balance_loss_clip": 0.94874024, + "balance_loss_mlp": 1.02843404, + "epoch": 0.2317080502615283, + "flos": 23039424766080.0, + "grad_norm": 1.7450047582418413, + "language_loss": 0.84214938, + "learning_rate": 3.5883254854745325e-06, + "loss": 0.86398053, + "num_input_tokens_seen": 41136235, + "step": 1927, + "time_per_iteration": 2.699530601501465 + }, + { + "auxiliary_loss_clip": 0.01191383, + "auxiliary_loss_mlp": 0.01031104, + "balance_loss_clip": 1.02167082, + "balance_loss_mlp": 1.02112651, + "epoch": 0.23182829315216738, + "flos": 11254656435840.0, + "grad_norm": 2.0649622620004244, + "language_loss": 0.74788719, + "learning_rate": 3.587851979938107e-06, + "loss": 0.77011204, + "num_input_tokens_seen": 41153125, + "step": 1928, + "time_per_iteration": 2.6554317474365234 + }, + { + "auxiliary_loss_clip": 0.01185849, + "auxiliary_loss_mlp": 0.01043886, + "balance_loss_clip": 1.02533889, + "balance_loss_mlp": 1.03431952, + "epoch": 0.23194853604280646, + "flos": 19828939155840.0, + "grad_norm": 2.1800138583421838, + "language_loss": 0.77595931, + "learning_rate": 3.5873782335257985e-06, + "loss": 0.7982567, + "num_input_tokens_seen": 41171290, + "step": 1929, + "time_per_iteration": 2.6602635383605957 + }, + { + "auxiliary_loss_clip": 0.01138305, + "auxiliary_loss_mlp": 0.01036464, + "balance_loss_clip": 0.95031756, + "balance_loss_mlp": 1.02742219, + "epoch": 0.23206877893344555, + "flos": 15305020830720.0, + "grad_norm": 2.4801581018829784, + "language_loss": 0.78625214, + "learning_rate": 3.5869042463094744e-06, + "loss": 0.80799985, + "num_input_tokens_seen": 41189005, + "step": 1930, + "time_per_iteration": 2.6027626991271973 + }, + { + "auxiliary_loss_clip": 0.01096946, + "auxiliary_loss_mlp": 0.0103988, + "balance_loss_clip": 0.90292621, + "balance_loss_mlp": 1.03000355, + "epoch": 0.23218902182408466, + "flos": 22711488572160.0, + "grad_norm": 2.028244527346612, + "language_loss": 0.77367395, + "learning_rate": 3.586430018361038e-06, + "loss": 0.79504222, + "num_input_tokens_seen": 41208775, + "step": 1931, + "time_per_iteration": 2.732316255569458 + }, + { + "auxiliary_loss_clip": 0.0114919, + "auxiliary_loss_mlp": 0.01030746, + "balance_loss_clip": 0.98021567, + "balance_loss_mlp": 1.02089381, + "epoch": 0.23230926471472374, + "flos": 22710734386560.0, + "grad_norm": 1.9504082572898516, + "language_loss": 0.76288962, + "learning_rate": 3.5859555497524283e-06, + "loss": 0.78468901, + "num_input_tokens_seen": 41226010, + "step": 1932, + "time_per_iteration": 2.6475634574890137 + }, + { + "auxiliary_loss_clip": 0.01185632, + "auxiliary_loss_mlp": 0.01031578, + "balance_loss_clip": 1.02529025, + "balance_loss_mlp": 1.02170777, + "epoch": 0.23242950760536282, + "flos": 20375499479040.0, + "grad_norm": 1.9029918819526384, + "language_loss": 0.92184049, + "learning_rate": 3.5854808405556237e-06, + "loss": 0.94401264, + "num_input_tokens_seen": 41245245, + "step": 1933, + "time_per_iteration": 2.6730613708496094 + }, + { + "auxiliary_loss_clip": 0.01137838, + "auxiliary_loss_mlp": 0.01034144, + "balance_loss_clip": 0.94656771, + "balance_loss_mlp": 1.02498913, + "epoch": 0.23254975049600193, + "flos": 16908324301440.0, + "grad_norm": 2.300530102701865, + "language_loss": 0.75953621, + "learning_rate": 3.5850058908426355e-06, + "loss": 0.78125602, + "num_input_tokens_seen": 41263795, + "step": 1934, + "time_per_iteration": 2.703939437866211 + }, + { + "auxiliary_loss_clip": 0.01164081, + "auxiliary_loss_mlp": 0.01040077, + "balance_loss_clip": 0.9828577, + "balance_loss_mlp": 1.03127909, + "epoch": 0.23266999338664102, + "flos": 23294821443840.0, + "grad_norm": 1.8291058210481075, + "language_loss": 0.85806638, + "learning_rate": 3.584530700685514e-06, + "loss": 0.88010794, + "num_input_tokens_seen": 41284055, + "step": 1935, + "time_per_iteration": 2.743795871734619 + }, + { + "auxiliary_loss_clip": 0.01160222, + "auxiliary_loss_mlp": 0.01033948, + "balance_loss_clip": 0.98767519, + "balance_loss_mlp": 1.02478075, + "epoch": 0.2327902362772801, + "flos": 19569987031680.0, + "grad_norm": 1.971633928597598, + "language_loss": 0.88920772, + "learning_rate": 3.5840552701563448e-06, + "loss": 0.91114944, + "num_input_tokens_seen": 41300255, + "step": 1936, + "time_per_iteration": 2.6791021823883057 + }, + { + "auxiliary_loss_clip": 0.01208895, + "auxiliary_loss_mlp": 0.01035596, + "balance_loss_clip": 1.06341219, + "balance_loss_mlp": 1.02663124, + "epoch": 0.2329104791679192, + "flos": 16727514215040.0, + "grad_norm": 2.52247845766574, + "language_loss": 0.81809914, + "learning_rate": 3.5835795993272513e-06, + "loss": 0.8405441, + "num_input_tokens_seen": 41318540, + "step": 1937, + "time_per_iteration": 2.6253862380981445 + }, + { + "auxiliary_loss_clip": 0.01036335, + "auxiliary_loss_mlp": 0.01037703, + "balance_loss_clip": 0.78983724, + "balance_loss_mlp": 1.02892363, + "epoch": 0.2330307220585583, + "flos": 22163743100160.0, + "grad_norm": 1.8163772079517306, + "language_loss": 0.71077323, + "learning_rate": 3.583103688270391e-06, + "loss": 0.73151362, + "num_input_tokens_seen": 41338320, + "step": 1938, + "time_per_iteration": 2.937152624130249 + }, + { + "auxiliary_loss_clip": 0.01147026, + "auxiliary_loss_mlp": 0.01031497, + "balance_loss_clip": 0.98160768, + "balance_loss_mlp": 1.02228165, + "epoch": 0.23315096494919738, + "flos": 19317319787520.0, + "grad_norm": 2.199155802143146, + "language_loss": 0.89229423, + "learning_rate": 3.58262753705796e-06, + "loss": 0.91407943, + "num_input_tokens_seen": 41353210, + "step": 1939, + "time_per_iteration": 2.9919075965881348 + }, + { + "auxiliary_loss_clip": 0.01064484, + "auxiliary_loss_mlp": 0.01012684, + "balance_loss_clip": 0.95518804, + "balance_loss_mlp": 1.01041925, + "epoch": 0.23327120783983646, + "flos": 53031048946560.0, + "grad_norm": 0.7610362942284067, + "language_loss": 0.55495363, + "learning_rate": 3.5821511457621902e-06, + "loss": 0.57572532, + "num_input_tokens_seen": 41410510, + "step": 1940, + "time_per_iteration": 3.169339656829834 + }, + { + "auxiliary_loss_clip": 0.01156592, + "auxiliary_loss_mlp": 0.0104221, + "balance_loss_clip": 0.98535097, + "balance_loss_mlp": 1.03264403, + "epoch": 0.23339145073047557, + "flos": 17126984344320.0, + "grad_norm": 4.523768388022929, + "language_loss": 0.808465, + "learning_rate": 3.5816745144553497e-06, + "loss": 0.83045304, + "num_input_tokens_seen": 41425830, + "step": 1941, + "time_per_iteration": 4.689767360687256 + }, + { + "auxiliary_loss_clip": 0.0110716, + "auxiliary_loss_mlp": 0.01033575, + "balance_loss_clip": 0.90768081, + "balance_loss_mlp": 1.02411008, + "epoch": 0.23351169362111465, + "flos": 13078918419840.0, + "grad_norm": 2.5339487598328208, + "language_loss": 0.7579059, + "learning_rate": 3.5811976432097424e-06, + "loss": 0.77931321, + "num_input_tokens_seen": 41443500, + "step": 1942, + "time_per_iteration": 2.7199013233184814 + }, + { + "auxiliary_loss_clip": 0.01187083, + "auxiliary_loss_mlp": 0.00512382, + "balance_loss_clip": 1.02733898, + "balance_loss_mlp": 0.0, + "epoch": 0.23363193651175373, + "flos": 15851257931520.0, + "grad_norm": 4.352922209506453, + "language_loss": 0.85111493, + "learning_rate": 3.58072053209771e-06, + "loss": 0.86810958, + "num_input_tokens_seen": 41460055, + "step": 1943, + "time_per_iteration": 2.646533489227295 + }, + { + "auxiliary_loss_clip": 0.01152255, + "auxiliary_loss_mlp": 0.01030874, + "balance_loss_clip": 0.97954828, + "balance_loss_mlp": 1.02099752, + "epoch": 0.23375217940239285, + "flos": 21025769345280.0, + "grad_norm": 2.2443448628909293, + "language_loss": 0.79214025, + "learning_rate": 3.5802431811916296e-06, + "loss": 0.81397152, + "num_input_tokens_seen": 41476665, + "step": 1944, + "time_per_iteration": 4.471642017364502 + }, + { + "auxiliary_loss_clip": 0.01158501, + "auxiliary_loss_mlp": 0.01035408, + "balance_loss_clip": 0.98585176, + "balance_loss_mlp": 1.02649117, + "epoch": 0.23387242229303193, + "flos": 20594698225920.0, + "grad_norm": 3.139014001007825, + "language_loss": 0.80478764, + "learning_rate": 3.579765590563916e-06, + "loss": 0.82672673, + "num_input_tokens_seen": 41496065, + "step": 1945, + "time_per_iteration": 2.656511068344116 + }, + { + "auxiliary_loss_clip": 0.0117861, + "auxiliary_loss_mlp": 0.01032829, + "balance_loss_clip": 1.02441525, + "balance_loss_mlp": 1.0235548, + "epoch": 0.233992665183671, + "flos": 24279491952000.0, + "grad_norm": 1.9555563602468027, + "language_loss": 0.81738508, + "learning_rate": 3.579287760287017e-06, + "loss": 0.83949947, + "num_input_tokens_seen": 41516815, + "step": 1946, + "time_per_iteration": 2.742608070373535 + }, + { + "auxiliary_loss_clip": 0.01182477, + "auxiliary_loss_mlp": 0.01033391, + "balance_loss_clip": 1.0239768, + "balance_loss_mlp": 1.02440894, + "epoch": 0.2341129080743101, + "flos": 30154621121280.0, + "grad_norm": 1.6771471062418362, + "language_loss": 0.7315647, + "learning_rate": 3.578809690433421e-06, + "loss": 0.75372338, + "num_input_tokens_seen": 41538525, + "step": 1947, + "time_per_iteration": 2.730304002761841 + }, + { + "auxiliary_loss_clip": 0.01213312, + "auxiliary_loss_mlp": 0.01030793, + "balance_loss_clip": 1.06420362, + "balance_loss_mlp": 1.02187037, + "epoch": 0.2342331509649492, + "flos": 22784135829120.0, + "grad_norm": 3.338573962656077, + "language_loss": 0.81303251, + "learning_rate": 3.578331381075651e-06, + "loss": 0.8354736, + "num_input_tokens_seen": 41559025, + "step": 1948, + "time_per_iteration": 2.7387611865997314 + }, + { + "auxiliary_loss_clip": 0.01183742, + "auxiliary_loss_mlp": 0.01034572, + "balance_loss_clip": 1.02175748, + "balance_loss_mlp": 1.02502954, + "epoch": 0.2343533938555883, + "flos": 23623152687360.0, + "grad_norm": 2.065349224958512, + "language_loss": 0.69835591, + "learning_rate": 3.5778528322862646e-06, + "loss": 0.72053909, + "num_input_tokens_seen": 41577845, + "step": 1949, + "time_per_iteration": 2.7723944187164307 + }, + { + "auxiliary_loss_clip": 0.01186315, + "auxiliary_loss_mlp": 0.01032238, + "balance_loss_clip": 1.02351928, + "balance_loss_mlp": 1.02325571, + "epoch": 0.23447363674622737, + "flos": 24570332375040.0, + "grad_norm": 1.4925937107986331, + "language_loss": 0.86411583, + "learning_rate": 3.5773740441378585e-06, + "loss": 0.8863014, + "num_input_tokens_seen": 41598600, + "step": 1950, + "time_per_iteration": 2.6561169624328613 + }, + { + "auxiliary_loss_clip": 0.01183031, + "auxiliary_loss_mlp": 0.01039243, + "balance_loss_clip": 1.02423728, + "balance_loss_mlp": 1.02955747, + "epoch": 0.23459387963686648, + "flos": 53140322119680.0, + "grad_norm": 2.79476744587042, + "language_loss": 0.73584402, + "learning_rate": 3.5768950167030633e-06, + "loss": 0.75806677, + "num_input_tokens_seen": 41623300, + "step": 1951, + "time_per_iteration": 2.9346911907196045 + }, + { + "auxiliary_loss_clip": 0.01150151, + "auxiliary_loss_mlp": 0.01034109, + "balance_loss_clip": 0.98056102, + "balance_loss_mlp": 1.02394605, + "epoch": 0.23471412252750556, + "flos": 23951412103680.0, + "grad_norm": 1.9426562189586765, + "language_loss": 0.78651232, + "learning_rate": 3.576415750054548e-06, + "loss": 0.80835485, + "num_input_tokens_seen": 41643420, + "step": 1952, + "time_per_iteration": 2.725059747695923 + }, + { + "auxiliary_loss_clip": 0.01152549, + "auxiliary_loss_mlp": 0.01033988, + "balance_loss_clip": 0.98395914, + "balance_loss_mlp": 1.02510142, + "epoch": 0.23483436541814465, + "flos": 15706573948800.0, + "grad_norm": 2.6686865022165627, + "language_loss": 0.85501027, + "learning_rate": 3.5759362442650172e-06, + "loss": 0.87687564, + "num_input_tokens_seen": 41660170, + "step": 1953, + "time_per_iteration": 2.7270913124084473 + }, + { + "auxiliary_loss_clip": 0.01182503, + "auxiliary_loss_mlp": 0.01029574, + "balance_loss_clip": 1.02418685, + "balance_loss_mlp": 1.01949477, + "epoch": 0.23495460830878373, + "flos": 24936262179840.0, + "grad_norm": 2.5733816497526107, + "language_loss": 0.8497256, + "learning_rate": 3.5754564994072113e-06, + "loss": 0.87184644, + "num_input_tokens_seen": 41679010, + "step": 1954, + "time_per_iteration": 2.677966356277466 + }, + { + "auxiliary_loss_clip": 0.01154793, + "auxiliary_loss_mlp": 0.01030643, + "balance_loss_clip": 0.98076975, + "balance_loss_mlp": 1.02087331, + "epoch": 0.23507485119942284, + "flos": 30482665056000.0, + "grad_norm": 2.771013446740172, + "language_loss": 0.59857792, + "learning_rate": 3.5749765155539067e-06, + "loss": 0.62043226, + "num_input_tokens_seen": 41699495, + "step": 1955, + "time_per_iteration": 2.8504319190979004 + }, + { + "auxiliary_loss_clip": 0.01131665, + "auxiliary_loss_mlp": 0.01036632, + "balance_loss_clip": 0.94290137, + "balance_loss_mlp": 1.02738094, + "epoch": 0.23519509409006192, + "flos": 18329129746560.0, + "grad_norm": 2.0568449972385334, + "language_loss": 0.92227447, + "learning_rate": 3.574496292777917e-06, + "loss": 0.94395739, + "num_input_tokens_seen": 41717705, + "step": 1956, + "time_per_iteration": 2.6747853755950928 + }, + { + "auxiliary_loss_clip": 0.01163131, + "auxiliary_loss_mlp": 0.01044489, + "balance_loss_clip": 0.98402232, + "balance_loss_mlp": 1.03408206, + "epoch": 0.235315336980701, + "flos": 29643217234560.0, + "grad_norm": 1.8594792362191968, + "language_loss": 0.71743381, + "learning_rate": 3.574015831152092e-06, + "loss": 0.73951006, + "num_input_tokens_seen": 41738120, + "step": 1957, + "time_per_iteration": 2.7548534870147705 + }, + { + "auxiliary_loss_clip": 0.01148891, + "auxiliary_loss_mlp": 0.01033147, + "balance_loss_clip": 0.98388135, + "balance_loss_mlp": 1.02418852, + "epoch": 0.23543557987134012, + "flos": 18551704371840.0, + "grad_norm": 2.5763435807603337, + "language_loss": 0.83249873, + "learning_rate": 3.573535130749316e-06, + "loss": 0.85431916, + "num_input_tokens_seen": 41756070, + "step": 1958, + "time_per_iteration": 3.3661625385284424 + }, + { + "auxiliary_loss_clip": 0.01152148, + "auxiliary_loss_mlp": 0.01032659, + "balance_loss_clip": 0.98619795, + "balance_loss_mlp": 1.02305698, + "epoch": 0.2355558227619792, + "flos": 24679033908480.0, + "grad_norm": 1.8432614838651182, + "language_loss": 0.73832977, + "learning_rate": 3.5730541916425127e-06, + "loss": 0.76017785, + "num_input_tokens_seen": 41777550, + "step": 1959, + "time_per_iteration": 2.7118701934814453 + }, + { + "auxiliary_loss_clip": 0.01139913, + "auxiliary_loss_mlp": 0.01039757, + "balance_loss_clip": 0.94638908, + "balance_loss_mlp": 1.02995837, + "epoch": 0.23567606565261828, + "flos": 21944795748480.0, + "grad_norm": 4.643458964632559, + "language_loss": 0.86606848, + "learning_rate": 3.572573013904639e-06, + "loss": 0.88786519, + "num_input_tokens_seen": 41797460, + "step": 1960, + "time_per_iteration": 2.698312520980835 + }, + { + "auxiliary_loss_clip": 0.01206257, + "auxiliary_loss_mlp": 0.01030955, + "balance_loss_clip": 1.060148, + "balance_loss_mlp": 1.0219543, + "epoch": 0.2357963085432574, + "flos": 13589352639360.0, + "grad_norm": 2.007090187444169, + "language_loss": 0.92285669, + "learning_rate": 3.572091597608689e-06, + "loss": 0.94522882, + "num_input_tokens_seen": 41815585, + "step": 1961, + "time_per_iteration": 2.6353583335876465 + }, + { + "auxiliary_loss_clip": 0.0116805, + "auxiliary_loss_mlp": 0.01035974, + "balance_loss_clip": 0.98739922, + "balance_loss_mlp": 1.02652073, + "epoch": 0.23591655143389648, + "flos": 22088689632000.0, + "grad_norm": 2.140347112494078, + "language_loss": 0.73600405, + "learning_rate": 3.571609942827694e-06, + "loss": 0.75804436, + "num_input_tokens_seen": 41834700, + "step": 1962, + "time_per_iteration": 2.727491617202759 + }, + { + "auxiliary_loss_clip": 0.01156435, + "auxiliary_loss_mlp": 0.0103174, + "balance_loss_clip": 0.98161435, + "balance_loss_mlp": 1.02256703, + "epoch": 0.23603679432453556, + "flos": 17017349057280.0, + "grad_norm": 1.750461914278071, + "language_loss": 0.88191509, + "learning_rate": 3.57112804963472e-06, + "loss": 0.90379691, + "num_input_tokens_seen": 41852915, + "step": 1963, + "time_per_iteration": 2.6671864986419678 + }, + { + "auxiliary_loss_clip": 0.01131343, + "auxiliary_loss_mlp": 0.01036229, + "balance_loss_clip": 0.94663453, + "balance_loss_mlp": 1.02758634, + "epoch": 0.23615703721517464, + "flos": 19171307001600.0, + "grad_norm": 2.0671419658481955, + "language_loss": 0.76303041, + "learning_rate": 3.57064591810287e-06, + "loss": 0.78470612, + "num_input_tokens_seen": 41870415, + "step": 1964, + "time_per_iteration": 2.83003568649292 + }, + { + "auxiliary_loss_clip": 0.01203979, + "auxiliary_loss_mlp": 0.00510876, + "balance_loss_clip": 1.06004143, + "balance_loss_mlp": 0.0, + "epoch": 0.23627728010581375, + "flos": 19098803399040.0, + "grad_norm": 2.157012484944567, + "language_loss": 0.80449295, + "learning_rate": 3.570163548305284e-06, + "loss": 0.82164156, + "num_input_tokens_seen": 41889345, + "step": 1965, + "time_per_iteration": 2.6553072929382324 + }, + { + "auxiliary_loss_clip": 0.01157112, + "auxiliary_loss_mlp": 0.01035126, + "balance_loss_clip": 0.9837594, + "balance_loss_mlp": 1.02540433, + "epoch": 0.23639752299645284, + "flos": 14282213057280.0, + "grad_norm": 2.1441376839577324, + "language_loss": 0.69988286, + "learning_rate": 3.569680940315135e-06, + "loss": 0.72180527, + "num_input_tokens_seen": 41905745, + "step": 1966, + "time_per_iteration": 3.5639214515686035 + }, + { + "auxiliary_loss_clip": 0.01137119, + "auxiliary_loss_mlp": 0.01037869, + "balance_loss_clip": 0.9417752, + "balance_loss_mlp": 1.02797484, + "epoch": 0.23651776588709192, + "flos": 22893411980160.0, + "grad_norm": 1.6811095149950601, + "language_loss": 0.8176446, + "learning_rate": 3.5691980942056356e-06, + "loss": 0.83939451, + "num_input_tokens_seen": 41925115, + "step": 1967, + "time_per_iteration": 3.839555263519287 + }, + { + "auxiliary_loss_clip": 0.01181224, + "auxiliary_loss_mlp": 0.0103517, + "balance_loss_clip": 1.01796043, + "balance_loss_mlp": 1.02587748, + "epoch": 0.23663800877773103, + "flos": 18624531196800.0, + "grad_norm": 1.6081828279544002, + "language_loss": 0.79213935, + "learning_rate": 3.5687150100500332e-06, + "loss": 0.81430328, + "num_input_tokens_seen": 41944815, + "step": 1968, + "time_per_iteration": 2.6002163887023926 + }, + { + "auxiliary_loss_clip": 0.01186372, + "auxiliary_loss_mlp": 0.01030081, + "balance_loss_clip": 1.02382302, + "balance_loss_mlp": 1.02104449, + "epoch": 0.2367582516683701, + "flos": 25555828896000.0, + "grad_norm": 3.7476534483023904, + "language_loss": 0.7442286, + "learning_rate": 3.568231687921611e-06, + "loss": 0.76639318, + "num_input_tokens_seen": 41964990, + "step": 1969, + "time_per_iteration": 2.7303688526153564 + }, + { + "auxiliary_loss_clip": 0.01206611, + "auxiliary_loss_mlp": 0.01036591, + "balance_loss_clip": 1.06167281, + "balance_loss_mlp": 1.0279727, + "epoch": 0.2368784945590092, + "flos": 23295072839040.0, + "grad_norm": 1.9197677887396583, + "language_loss": 0.80496818, + "learning_rate": 3.5677481278936883e-06, + "loss": 0.82740021, + "num_input_tokens_seen": 41984570, + "step": 1970, + "time_per_iteration": 3.607945680618286 + }, + { + "auxiliary_loss_clip": 0.01056361, + "auxiliary_loss_mlp": 0.01003227, + "balance_loss_clip": 0.94897234, + "balance_loss_mlp": 1.00104523, + "epoch": 0.23699873744964828, + "flos": 69859291875840.0, + "grad_norm": 0.8285428444144054, + "language_loss": 0.57810485, + "learning_rate": 3.5672643300396214e-06, + "loss": 0.59870076, + "num_input_tokens_seen": 42053715, + "step": 1971, + "time_per_iteration": 3.337249994277954 + }, + { + "auxiliary_loss_clip": 0.01134984, + "auxiliary_loss_mlp": 0.01038028, + "balance_loss_clip": 0.9442609, + "balance_loss_mlp": 1.02873015, + "epoch": 0.2371189803402874, + "flos": 21835052720640.0, + "grad_norm": 2.14769997093729, + "language_loss": 0.68044269, + "learning_rate": 3.566780294432802e-06, + "loss": 0.70217282, + "num_input_tokens_seen": 42070890, + "step": 1972, + "time_per_iteration": 2.7301690578460693 + }, + { + "auxiliary_loss_clip": 0.01206894, + "auxiliary_loss_mlp": 0.01036379, + "balance_loss_clip": 1.06211889, + "balance_loss_mlp": 1.02668154, + "epoch": 0.23723922323092647, + "flos": 21908490076800.0, + "grad_norm": 2.131550320371792, + "language_loss": 0.74584508, + "learning_rate": 3.566296021146657e-06, + "loss": 0.76827776, + "num_input_tokens_seen": 42090270, + "step": 1973, + "time_per_iteration": 2.6790354251861572 + }, + { + "auxiliary_loss_clip": 0.01209506, + "auxiliary_loss_mlp": 0.01028526, + "balance_loss_clip": 1.06177402, + "balance_loss_mlp": 1.0188936, + "epoch": 0.23735946612156555, + "flos": 32708803380480.0, + "grad_norm": 1.6426106475232856, + "language_loss": 0.73313439, + "learning_rate": 3.565811510254652e-06, + "loss": 0.75551474, + "num_input_tokens_seen": 42111150, + "step": 1974, + "time_per_iteration": 2.725097179412842 + }, + { + "auxiliary_loss_clip": 0.01088996, + "auxiliary_loss_mlp": 0.01015709, + "balance_loss_clip": 0.99944699, + "balance_loss_mlp": 1.01362252, + "epoch": 0.23747970901220466, + "flos": 70546944821760.0, + "grad_norm": 0.835703817700593, + "language_loss": 0.58234429, + "learning_rate": 3.5653267618302845e-06, + "loss": 0.60339129, + "num_input_tokens_seen": 42178730, + "step": 1975, + "time_per_iteration": 3.2438089847564697 + }, + { + "auxiliary_loss_clip": 0.01206882, + "auxiliary_loss_mlp": 0.01032546, + "balance_loss_clip": 1.06104338, + "balance_loss_mlp": 1.02290773, + "epoch": 0.23759995190284375, + "flos": 20849807594880.0, + "grad_norm": 1.7885319946664642, + "language_loss": 0.85635924, + "learning_rate": 3.564841775947093e-06, + "loss": 0.87875348, + "num_input_tokens_seen": 42199620, + "step": 1976, + "time_per_iteration": 2.6368188858032227 + }, + { + "auxiliary_loss_clip": 0.01131093, + "auxiliary_loss_mlp": 0.01034091, + "balance_loss_clip": 0.94139427, + "balance_loss_mlp": 1.02442312, + "epoch": 0.23772019479348283, + "flos": 32921645420160.0, + "grad_norm": 2.3530785076120777, + "language_loss": 0.7641145, + "learning_rate": 3.5643565526786475e-06, + "loss": 0.78576636, + "num_input_tokens_seen": 42219560, + "step": 1977, + "time_per_iteration": 2.894943952560425 + }, + { + "auxiliary_loss_clip": 0.01207169, + "auxiliary_loss_mlp": 0.01035512, + "balance_loss_clip": 1.06144416, + "balance_loss_mlp": 1.02657175, + "epoch": 0.2378404376841219, + "flos": 32342765834880.0, + "grad_norm": 1.7831903774294897, + "language_loss": 0.77382052, + "learning_rate": 3.5638710920985574e-06, + "loss": 0.79624736, + "num_input_tokens_seen": 42241020, + "step": 1978, + "time_per_iteration": 2.8325514793395996 + }, + { + "auxiliary_loss_clip": 0.01190654, + "auxiliary_loss_mlp": 0.00511919, + "balance_loss_clip": 1.02279544, + "balance_loss_mlp": 0.0, + "epoch": 0.23796068057476102, + "flos": 22997624313600.0, + "grad_norm": 2.979635823125866, + "language_loss": 0.82231641, + "learning_rate": 3.5633853942804655e-06, + "loss": 0.83934212, + "num_input_tokens_seen": 42259345, + "step": 1979, + "time_per_iteration": 2.727118730545044 + }, + { + "auxiliary_loss_clip": 0.01134243, + "auxiliary_loss_mlp": 0.01039091, + "balance_loss_clip": 0.94184726, + "balance_loss_mlp": 1.02958381, + "epoch": 0.2380809234654001, + "flos": 13480938414720.0, + "grad_norm": 2.805813185234173, + "language_loss": 0.76735961, + "learning_rate": 3.5628994592980527e-06, + "loss": 0.7890929, + "num_input_tokens_seen": 42277250, + "step": 1980, + "time_per_iteration": 2.8127756118774414 + }, + { + "auxiliary_loss_clip": 0.01206235, + "auxiliary_loss_mlp": 0.01036576, + "balance_loss_clip": 1.06056118, + "balance_loss_mlp": 1.02705097, + "epoch": 0.2382011663560392, + "flos": 16871803148160.0, + "grad_norm": 1.65761175453434, + "language_loss": 0.70288575, + "learning_rate": 3.562413287225034e-06, + "loss": 0.72531384, + "num_input_tokens_seen": 42295360, + "step": 1981, + "time_per_iteration": 2.693713903427124 + }, + { + "auxiliary_loss_clip": 0.01177588, + "auxiliary_loss_mlp": 0.01037834, + "balance_loss_clip": 1.02308226, + "balance_loss_mlp": 1.02801108, + "epoch": 0.2383214092466783, + "flos": 18441135331200.0, + "grad_norm": 2.8769756445688466, + "language_loss": 0.89369226, + "learning_rate": 3.5619268781351623e-06, + "loss": 0.91584647, + "num_input_tokens_seen": 42313430, + "step": 1982, + "time_per_iteration": 2.606677532196045 + }, + { + "auxiliary_loss_clip": 0.01152573, + "auxiliary_loss_mlp": 0.0103614, + "balance_loss_clip": 0.98509091, + "balance_loss_mlp": 1.02752674, + "epoch": 0.23844165213731738, + "flos": 19755717281280.0, + "grad_norm": 2.483263471121761, + "language_loss": 0.7647391, + "learning_rate": 3.5614402321022256e-06, + "loss": 0.78662622, + "num_input_tokens_seen": 42331260, + "step": 1983, + "time_per_iteration": 2.7216975688934326 + }, + { + "auxiliary_loss_clip": 0.0110535, + "auxiliary_loss_mlp": 0.01029331, + "balance_loss_clip": 0.90588582, + "balance_loss_mlp": 1.02100396, + "epoch": 0.23856189502795647, + "flos": 23367360960000.0, + "grad_norm": 1.686701377178328, + "language_loss": 0.87053752, + "learning_rate": 3.5609533492000463e-06, + "loss": 0.89188433, + "num_input_tokens_seen": 42350150, + "step": 1984, + "time_per_iteration": 2.7735865116119385 + }, + { + "auxiliary_loss_clip": 0.01154909, + "auxiliary_loss_mlp": 0.01030906, + "balance_loss_clip": 0.98501199, + "balance_loss_mlp": 1.02203131, + "epoch": 0.23868213791859555, + "flos": 23475056912640.0, + "grad_norm": 2.13403697889267, + "language_loss": 0.78138024, + "learning_rate": 3.560466229502485e-06, + "loss": 0.80323839, + "num_input_tokens_seen": 42369495, + "step": 1985, + "time_per_iteration": 2.7061920166015625 + }, + { + "auxiliary_loss_clip": 0.01157088, + "auxiliary_loss_mlp": 0.00511772, + "balance_loss_clip": 0.98726106, + "balance_loss_mlp": 0.0, + "epoch": 0.23880238080923466, + "flos": 16617340224000.0, + "grad_norm": 2.4290729206686197, + "language_loss": 0.90059841, + "learning_rate": 3.5599788730834384e-06, + "loss": 0.91728699, + "num_input_tokens_seen": 42387455, + "step": 1986, + "time_per_iteration": 2.6493144035339355 + }, + { + "auxiliary_loss_clip": 0.01184257, + "auxiliary_loss_mlp": 0.0103766, + "balance_loss_clip": 1.02069235, + "balance_loss_mlp": 1.02821827, + "epoch": 0.23892262369987374, + "flos": 17348409734400.0, + "grad_norm": 2.4273316241800043, + "language_loss": 0.78961557, + "learning_rate": 3.559491280016836e-06, + "loss": 0.81183469, + "num_input_tokens_seen": 42405400, + "step": 1987, + "time_per_iteration": 2.6480395793914795 + }, + { + "auxiliary_loss_clip": 0.0115872, + "auxiliary_loss_mlp": 0.01032353, + "balance_loss_clip": 0.98547018, + "balance_loss_mlp": 1.02341783, + "epoch": 0.23904286659051283, + "flos": 22309899540480.0, + "grad_norm": 1.8950881955589793, + "language_loss": 0.70672041, + "learning_rate": 3.5590034503766465e-06, + "loss": 0.72863114, + "num_input_tokens_seen": 42425065, + "step": 1988, + "time_per_iteration": 2.682471990585327 + }, + { + "auxiliary_loss_clip": 0.01207131, + "auxiliary_loss_mlp": 0.01025625, + "balance_loss_clip": 1.06181586, + "balance_loss_mlp": 1.01686335, + "epoch": 0.23916310948115194, + "flos": 21178246579200.0, + "grad_norm": 2.422810694244343, + "language_loss": 0.8089143, + "learning_rate": 3.558515384236874e-06, + "loss": 0.83124191, + "num_input_tokens_seen": 42442495, + "step": 1989, + "time_per_iteration": 2.757528781890869 + }, + { + "auxiliary_loss_clip": 0.0112662, + "auxiliary_loss_mlp": 0.00511666, + "balance_loss_clip": 0.94717932, + "balance_loss_mlp": 0.0, + "epoch": 0.23928335237179102, + "flos": 14137349506560.0, + "grad_norm": 3.4340867485981024, + "language_loss": 0.83646989, + "learning_rate": 3.558027081671556e-06, + "loss": 0.8528527, + "num_input_tokens_seen": 42459480, + "step": 1990, + "time_per_iteration": 2.7555878162384033 + }, + { + "auxiliary_loss_clip": 0.01185174, + "auxiliary_loss_mlp": 0.01035174, + "balance_loss_clip": 1.02121139, + "balance_loss_mlp": 1.02675772, + "epoch": 0.2394035952624301, + "flos": 23769596436480.0, + "grad_norm": 2.0356519784186493, + "language_loss": 0.68543386, + "learning_rate": 3.557538542754769e-06, + "loss": 0.70763737, + "num_input_tokens_seen": 42479175, + "step": 1991, + "time_per_iteration": 2.6606922149658203 + }, + { + "auxiliary_loss_clip": 0.01207524, + "auxiliary_loss_mlp": 0.01036124, + "balance_loss_clip": 1.06271851, + "balance_loss_mlp": 1.02701068, + "epoch": 0.2395238381530692, + "flos": 24206198250240.0, + "grad_norm": 2.303812820371131, + "language_loss": 0.66689324, + "learning_rate": 3.557049767560623e-06, + "loss": 0.68932974, + "num_input_tokens_seen": 42498090, + "step": 1992, + "time_per_iteration": 3.57424259185791 + }, + { + "auxiliary_loss_clip": 0.01112542, + "auxiliary_loss_mlp": 0.01034044, + "balance_loss_clip": 0.9093222, + "balance_loss_mlp": 1.02481723, + "epoch": 0.2396440810437083, + "flos": 25295763450240.0, + "grad_norm": 2.1949520361141097, + "language_loss": 0.85782576, + "learning_rate": 3.5565607561632655e-06, + "loss": 0.87929165, + "num_input_tokens_seen": 42516930, + "step": 1993, + "time_per_iteration": 4.007151365280151 + }, + { + "auxiliary_loss_clip": 0.01154069, + "auxiliary_loss_mlp": 0.01027429, + "balance_loss_clip": 0.98394227, + "balance_loss_mlp": 1.01814878, + "epoch": 0.23976432393434738, + "flos": 28543093436160.0, + "grad_norm": 2.3245181524424248, + "language_loss": 0.7925179, + "learning_rate": 3.5560715086368787e-06, + "loss": 0.81433284, + "num_input_tokens_seen": 42534800, + "step": 1994, + "time_per_iteration": 2.9748470783233643 + }, + { + "auxiliary_loss_clip": 0.01151063, + "auxiliary_loss_mlp": 0.01034277, + "balance_loss_clip": 0.98265427, + "balance_loss_mlp": 1.02544343, + "epoch": 0.23988456682498646, + "flos": 19494358945920.0, + "grad_norm": 2.437669167818466, + "language_loss": 0.82141, + "learning_rate": 3.5555820250556816e-06, + "loss": 0.84326339, + "num_input_tokens_seen": 42552000, + "step": 1995, + "time_per_iteration": 2.662651777267456 + }, + { + "auxiliary_loss_clip": 0.01161938, + "auxiliary_loss_mlp": 0.0103394, + "balance_loss_clip": 0.98481816, + "balance_loss_mlp": 1.02470136, + "epoch": 0.24000480971562557, + "flos": 20266331068800.0, + "grad_norm": 2.1118886473925276, + "language_loss": 0.69386148, + "learning_rate": 3.5550923054939278e-06, + "loss": 0.71582031, + "num_input_tokens_seen": 42571455, + "step": 1996, + "time_per_iteration": 3.60054874420166 + }, + { + "auxiliary_loss_clip": 0.01106791, + "auxiliary_loss_mlp": 0.01031193, + "balance_loss_clip": 0.90409863, + "balance_loss_mlp": 1.02234805, + "epoch": 0.24012505260626466, + "flos": 25443176866560.0, + "grad_norm": 1.9244638580364137, + "language_loss": 0.74585879, + "learning_rate": 3.5546023500259083e-06, + "loss": 0.76723862, + "num_input_tokens_seen": 42592550, + "step": 1997, + "time_per_iteration": 2.794090747833252 + }, + { + "auxiliary_loss_clip": 0.011152, + "auxiliary_loss_mlp": 0.01028086, + "balance_loss_clip": 0.90561825, + "balance_loss_mlp": 1.01917505, + "epoch": 0.24024529549690374, + "flos": 15553342529280.0, + "grad_norm": 2.5777118885566765, + "language_loss": 0.80196679, + "learning_rate": 3.5541121587259477e-06, + "loss": 0.82339966, + "num_input_tokens_seen": 42610385, + "step": 1998, + "time_per_iteration": 2.7412514686584473 + }, + { + "auxiliary_loss_clip": 0.01080405, + "auxiliary_loss_mlp": 0.0101919, + "balance_loss_clip": 0.99215269, + "balance_loss_mlp": 1.01736641, + "epoch": 0.24036553838754285, + "flos": 57122351867520.0, + "grad_norm": 0.8236703229339211, + "language_loss": 0.57883644, + "learning_rate": 3.553621731668408e-06, + "loss": 0.59983242, + "num_input_tokens_seen": 42673595, + "step": 1999, + "time_per_iteration": 3.2196459770202637 + }, + { + "auxiliary_loss_clip": 0.01174306, + "auxiliary_loss_mlp": 0.01024711, + "balance_loss_clip": 1.01765323, + "balance_loss_mlp": 1.01635456, + "epoch": 0.24048578127818193, + "flos": 24969946158720.0, + "grad_norm": 1.680818488210191, + "language_loss": 0.83051938, + "learning_rate": 3.553131068927688e-06, + "loss": 0.85250956, + "num_input_tokens_seen": 42692000, + "step": 2000, + "time_per_iteration": 2.728201150894165 + }, + { + "auxiliary_loss_clip": 0.01136677, + "auxiliary_loss_mlp": 0.010363, + "balance_loss_clip": 0.94641578, + "balance_loss_mlp": 1.02777088, + "epoch": 0.24060602416882101, + "flos": 23330947547520.0, + "grad_norm": 2.763005391861714, + "language_loss": 0.80407035, + "learning_rate": 3.552640170578219e-06, + "loss": 0.82580018, + "num_input_tokens_seen": 42712250, + "step": 2001, + "time_per_iteration": 2.702899932861328 + }, + { + "auxiliary_loss_clip": 0.01160032, + "auxiliary_loss_mlp": 0.01039229, + "balance_loss_clip": 0.98446214, + "balance_loss_mlp": 1.02951956, + "epoch": 0.2407262670594601, + "flos": 14173260128640.0, + "grad_norm": 1.9556473723438264, + "language_loss": 0.77934951, + "learning_rate": 3.5521490366944703e-06, + "loss": 0.80134213, + "num_input_tokens_seen": 42729900, + "step": 2002, + "time_per_iteration": 2.732964515686035 + }, + { + "auxiliary_loss_clip": 0.01136469, + "auxiliary_loss_mlp": 0.01036043, + "balance_loss_clip": 0.94566131, + "balance_loss_mlp": 1.02616656, + "epoch": 0.2408465099500992, + "flos": 13663113217920.0, + "grad_norm": 2.4378273513909314, + "language_loss": 0.79928327, + "learning_rate": 3.5516576673509474e-06, + "loss": 0.82100844, + "num_input_tokens_seen": 42747900, + "step": 2003, + "time_per_iteration": 2.674403667449951 + }, + { + "auxiliary_loss_clip": 0.01210664, + "auxiliary_loss_mlp": 0.01030217, + "balance_loss_clip": 1.06283855, + "balance_loss_mlp": 1.02051353, + "epoch": 0.2409667528407383, + "flos": 31248029076480.0, + "grad_norm": 1.9046365900901283, + "language_loss": 0.86204457, + "learning_rate": 3.5511660626221896e-06, + "loss": 0.88445342, + "num_input_tokens_seen": 42768540, + "step": 2004, + "time_per_iteration": 2.640427350997925 + }, + { + "auxiliary_loss_clip": 0.01158102, + "auxiliary_loss_mlp": 0.00512116, + "balance_loss_clip": 0.98500317, + "balance_loss_mlp": 0.0, + "epoch": 0.24108699573137737, + "flos": 22199941031040.0, + "grad_norm": 2.6527084261120693, + "language_loss": 0.89409983, + "learning_rate": 3.5506742225827744e-06, + "loss": 0.91080201, + "num_input_tokens_seen": 42785395, + "step": 2005, + "time_per_iteration": 2.657975196838379 + }, + { + "auxiliary_loss_clip": 0.01135986, + "auxiliary_loss_mlp": 0.01032198, + "balance_loss_clip": 0.94726986, + "balance_loss_mlp": 1.02255416, + "epoch": 0.24120723862201648, + "flos": 26103035664000.0, + "grad_norm": 2.3768503657851645, + "language_loss": 0.89820141, + "learning_rate": 3.5501821473073116e-06, + "loss": 0.91988325, + "num_input_tokens_seen": 42801980, + "step": 2006, + "time_per_iteration": 2.7474589347839355 + }, + { + "auxiliary_loss_clip": 0.01130558, + "auxiliary_loss_mlp": 0.01036329, + "balance_loss_clip": 0.94534886, + "balance_loss_mlp": 1.02680373, + "epoch": 0.24132748151265557, + "flos": 18624926246400.0, + "grad_norm": 2.0235737866314434, + "language_loss": 0.86630934, + "learning_rate": 3.54968983687045e-06, + "loss": 0.8879782, + "num_input_tokens_seen": 42818850, + "step": 2007, + "time_per_iteration": 2.697291612625122 + }, + { + "auxiliary_loss_clip": 0.01164355, + "auxiliary_loss_mlp": 0.01033405, + "balance_loss_clip": 0.98733974, + "balance_loss_mlp": 1.02389812, + "epoch": 0.24144772440329465, + "flos": 15267673664640.0, + "grad_norm": 2.5574215655447157, + "language_loss": 0.89830184, + "learning_rate": 3.549197291346872e-06, + "loss": 0.92027944, + "num_input_tokens_seen": 42835375, + "step": 2008, + "time_per_iteration": 2.7075302600860596 + }, + { + "auxiliary_loss_clip": 0.01188197, + "auxiliary_loss_mlp": 0.01037005, + "balance_loss_clip": 1.02343321, + "balance_loss_mlp": 1.02774858, + "epoch": 0.24156796729393373, + "flos": 24024274842240.0, + "grad_norm": 1.89117856449794, + "language_loss": 0.79083669, + "learning_rate": 3.548704510811297e-06, + "loss": 0.81308866, + "num_input_tokens_seen": 42854570, + "step": 2009, + "time_per_iteration": 2.7329812049865723 + }, + { + "auxiliary_loss_clip": 0.01113672, + "auxiliary_loss_mlp": 0.01039346, + "balance_loss_clip": 0.90587842, + "balance_loss_mlp": 1.02955329, + "epoch": 0.24168821018457284, + "flos": 26286790665600.0, + "grad_norm": 2.4459681528000283, + "language_loss": 0.74283934, + "learning_rate": 3.5482114953384787e-06, + "loss": 0.76436949, + "num_input_tokens_seen": 42873800, + "step": 2010, + "time_per_iteration": 2.794320583343506 + }, + { + "auxiliary_loss_clip": 0.0118805, + "auxiliary_loss_mlp": 0.01036529, + "balance_loss_clip": 1.02442777, + "balance_loss_mlp": 1.02788019, + "epoch": 0.24180845307521193, + "flos": 18223193560320.0, + "grad_norm": 1.9182867453235817, + "language_loss": 0.84252489, + "learning_rate": 3.5477182450032077e-06, + "loss": 0.86477071, + "num_input_tokens_seen": 42892400, + "step": 2011, + "time_per_iteration": 2.848605155944824 + }, + { + "auxiliary_loss_clip": 0.01186187, + "auxiliary_loss_mlp": 0.01031569, + "balance_loss_clip": 1.02720118, + "balance_loss_mlp": 1.02203834, + "epoch": 0.241928695965851, + "flos": 20449260057600.0, + "grad_norm": 14.690846002801564, + "language_loss": 0.83190334, + "learning_rate": 3.5472247598803097e-06, + "loss": 0.8540808, + "num_input_tokens_seen": 42911745, + "step": 2012, + "time_per_iteration": 2.6124508380889893 + }, + { + "auxiliary_loss_clip": 0.01207979, + "auxiliary_loss_mlp": 0.0104528, + "balance_loss_clip": 1.06178689, + "balance_loss_mlp": 1.03465211, + "epoch": 0.24204893885649012, + "flos": 25556475340800.0, + "grad_norm": 2.5338691816062755, + "language_loss": 0.85225952, + "learning_rate": 3.546731040044645e-06, + "loss": 0.8747921, + "num_input_tokens_seen": 42926915, + "step": 2013, + "time_per_iteration": 2.68284010887146 + }, + { + "auxiliary_loss_clip": 0.01211143, + "auxiliary_loss_mlp": 0.01040132, + "balance_loss_clip": 1.06490767, + "balance_loss_mlp": 1.03089976, + "epoch": 0.2421691817471292, + "flos": 30660207004800.0, + "grad_norm": 2.020671100075274, + "language_loss": 0.75185788, + "learning_rate": 3.546237085571112e-06, + "loss": 0.77437067, + "num_input_tokens_seen": 42945350, + "step": 2014, + "time_per_iteration": 2.627168655395508 + }, + { + "auxiliary_loss_clip": 0.01184384, + "auxiliary_loss_mlp": 0.01044118, + "balance_loss_clip": 1.02450001, + "balance_loss_mlp": 1.03472424, + "epoch": 0.24228942463776829, + "flos": 21945011230080.0, + "grad_norm": 2.389326297191811, + "language_loss": 0.72815073, + "learning_rate": 3.5457428965346425e-06, + "loss": 0.75043571, + "num_input_tokens_seen": 42964290, + "step": 2015, + "time_per_iteration": 2.6398773193359375 + }, + { + "auxiliary_loss_clip": 0.01085059, + "auxiliary_loss_mlp": 0.01038955, + "balance_loss_clip": 0.86755198, + "balance_loss_mlp": 1.02965713, + "epoch": 0.2424096675284074, + "flos": 33984493879680.0, + "grad_norm": 1.9463486588839267, + "language_loss": 0.74623525, + "learning_rate": 3.545248473010205e-06, + "loss": 0.76747543, + "num_input_tokens_seen": 42987095, + "step": 2016, + "time_per_iteration": 2.8670942783355713 + }, + { + "auxiliary_loss_clip": 0.01213691, + "auxiliary_loss_mlp": 0.00512533, + "balance_loss_clip": 1.06438291, + "balance_loss_mlp": 0.0, + "epoch": 0.24252991041904648, + "flos": 21653416621440.0, + "grad_norm": 1.6816518524979214, + "language_loss": 0.87720156, + "learning_rate": 3.544753815072802e-06, + "loss": 0.89446378, + "num_input_tokens_seen": 43005750, + "step": 2017, + "time_per_iteration": 2.6353514194488525 + }, + { + "auxiliary_loss_clip": 0.0105732, + "auxiliary_loss_mlp": 0.0103332, + "balance_loss_clip": 0.8270753, + "balance_loss_mlp": 1.02386665, + "epoch": 0.24265015330968556, + "flos": 21870065502720.0, + "grad_norm": 1.9891529651550943, + "language_loss": 0.87991136, + "learning_rate": 3.544258922797474e-06, + "loss": 0.90081781, + "num_input_tokens_seen": 43023870, + "step": 2018, + "time_per_iteration": 3.6954495906829834 + }, + { + "auxiliary_loss_clip": 0.01209537, + "auxiliary_loss_mlp": 0.0103924, + "balance_loss_clip": 1.06432748, + "balance_loss_mlp": 1.03068686, + "epoch": 0.24277039620032465, + "flos": 25628260671360.0, + "grad_norm": 1.6153967858474638, + "language_loss": 0.77987385, + "learning_rate": 3.543763796259295e-06, + "loss": 0.80236173, + "num_input_tokens_seen": 43043825, + "step": 2019, + "time_per_iteration": 3.6187472343444824 + }, + { + "auxiliary_loss_clip": 0.01185347, + "auxiliary_loss_mlp": 0.01035611, + "balance_loss_clip": 1.02444649, + "balance_loss_mlp": 1.02629471, + "epoch": 0.24289063909096376, + "flos": 26286575184000.0, + "grad_norm": 2.4305611430746374, + "language_loss": 0.90699422, + "learning_rate": 3.5432684355333754e-06, + "loss": 0.92920375, + "num_input_tokens_seen": 43062480, + "step": 2020, + "time_per_iteration": 2.6988840103149414 + }, + { + "auxiliary_loss_clip": 0.01183178, + "auxiliary_loss_mlp": 0.01031853, + "balance_loss_clip": 1.02204287, + "balance_loss_mlp": 1.02229214, + "epoch": 0.24301088198160284, + "flos": 25075056332160.0, + "grad_norm": 1.848277106632349, + "language_loss": 0.76231825, + "learning_rate": 3.5427728406948613e-06, + "loss": 0.78446853, + "num_input_tokens_seen": 43081595, + "step": 2021, + "time_per_iteration": 2.6482458114624023 + }, + { + "auxiliary_loss_clip": 0.0107493, + "auxiliary_loss_mlp": 0.01011704, + "balance_loss_clip": 0.96155065, + "balance_loss_mlp": 1.00948703, + "epoch": 0.24313112487224192, + "flos": 69900948673920.0, + "grad_norm": 0.7568785195779394, + "language_loss": 0.57868463, + "learning_rate": 3.542277011818934e-06, + "loss": 0.59955096, + "num_input_tokens_seen": 43145430, + "step": 2022, + "time_per_iteration": 4.207285642623901 + }, + { + "auxiliary_loss_clip": 0.01162802, + "auxiliary_loss_mlp": 0.01040205, + "balance_loss_clip": 0.98892409, + "balance_loss_mlp": 1.03109121, + "epoch": 0.24325136776288103, + "flos": 40662334235520.0, + "grad_norm": 2.1154908458638335, + "language_loss": 0.74115038, + "learning_rate": 3.5417809489808104e-06, + "loss": 0.76318049, + "num_input_tokens_seen": 43167040, + "step": 2023, + "time_per_iteration": 3.958505392074585 + }, + { + "auxiliary_loss_clip": 0.01187983, + "auxiliary_loss_mlp": 0.01032507, + "balance_loss_clip": 1.02468681, + "balance_loss_mlp": 1.02332151, + "epoch": 0.24337161065352012, + "flos": 25046400257280.0, + "grad_norm": 1.7779183631416553, + "language_loss": 0.72747481, + "learning_rate": 3.5412846522557422e-06, + "loss": 0.74967968, + "num_input_tokens_seen": 43187930, + "step": 2024, + "time_per_iteration": 2.8390252590179443 + }, + { + "auxiliary_loss_clip": 0.01212657, + "auxiliary_loss_mlp": 0.0104469, + "balance_loss_clip": 1.06579494, + "balance_loss_mlp": 1.0343132, + "epoch": 0.2434918535441592, + "flos": 18661160090880.0, + "grad_norm": 2.018063992640429, + "language_loss": 0.737867, + "learning_rate": 3.540788121719018e-06, + "loss": 0.76044047, + "num_input_tokens_seen": 43206350, + "step": 2025, + "time_per_iteration": 2.6554512977600098 + }, + { + "auxiliary_loss_clip": 0.01131813, + "auxiliary_loss_mlp": 0.01034446, + "balance_loss_clip": 0.94817746, + "balance_loss_mlp": 1.02547514, + "epoch": 0.24361209643479828, + "flos": 23915142345600.0, + "grad_norm": 1.7857801914292601, + "language_loss": 0.82036626, + "learning_rate": 3.5402913574459604e-06, + "loss": 0.84202886, + "num_input_tokens_seen": 43226255, + "step": 2026, + "time_per_iteration": 2.796354055404663 + }, + { + "auxiliary_loss_clip": 0.01085925, + "auxiliary_loss_mlp": 0.01033254, + "balance_loss_clip": 0.86676252, + "balance_loss_mlp": 1.02424145, + "epoch": 0.2437323393254374, + "flos": 28657505232000.0, + "grad_norm": 1.839507279537781, + "language_loss": 0.86020112, + "learning_rate": 3.5397943595119297e-06, + "loss": 0.8813929, + "num_input_tokens_seen": 43247675, + "step": 2027, + "time_per_iteration": 2.8481171131134033 + }, + { + "auxiliary_loss_clip": 0.01158227, + "auxiliary_loss_mlp": 0.01033358, + "balance_loss_clip": 0.98633224, + "balance_loss_mlp": 1.02405977, + "epoch": 0.24385258221607647, + "flos": 23550325862400.0, + "grad_norm": 3.2551058816360015, + "language_loss": 0.77822244, + "learning_rate": 3.5392971279923177e-06, + "loss": 0.80013829, + "num_input_tokens_seen": 43265895, + "step": 2028, + "time_per_iteration": 2.763719081878662 + }, + { + "auxiliary_loss_clip": 0.01129824, + "auxiliary_loss_mlp": 0.01033428, + "balance_loss_clip": 0.94339776, + "balance_loss_mlp": 1.02362943, + "epoch": 0.24397282510671556, + "flos": 25336091445120.0, + "grad_norm": 2.2536396149186415, + "language_loss": 0.82708716, + "learning_rate": 3.5387996629625557e-06, + "loss": 0.84871972, + "num_input_tokens_seen": 43283485, + "step": 2029, + "time_per_iteration": 2.7532856464385986 + }, + { + "auxiliary_loss_clip": 0.01113762, + "auxiliary_loss_mlp": 0.01003582, + "balance_loss_clip": 1.03541732, + "balance_loss_mlp": 1.00146055, + "epoch": 0.24409306799735467, + "flos": 65187421430400.0, + "grad_norm": 0.8066609638510489, + "language_loss": 0.55023658, + "learning_rate": 3.5383019644981083e-06, + "loss": 0.57141, + "num_input_tokens_seen": 43347180, + "step": 2030, + "time_per_iteration": 3.2676150798797607 + }, + { + "auxiliary_loss_clip": 0.01158965, + "auxiliary_loss_mlp": 0.01038073, + "balance_loss_clip": 0.985686, + "balance_loss_mlp": 1.02830386, + "epoch": 0.24421331088799375, + "flos": 19537093152000.0, + "grad_norm": 2.4238988587536903, + "language_loss": 0.72928488, + "learning_rate": 3.5378040326744763e-06, + "loss": 0.75125527, + "num_input_tokens_seen": 43366665, + "step": 2031, + "time_per_iteration": 2.7639222145080566 + }, + { + "auxiliary_loss_clip": 0.01140665, + "auxiliary_loss_mlp": 0.01031888, + "balance_loss_clip": 0.95103014, + "balance_loss_mlp": 1.02292347, + "epoch": 0.24433355377863283, + "flos": 21068575378560.0, + "grad_norm": 2.1023667465899374, + "language_loss": 0.85554171, + "learning_rate": 3.5373058675671946e-06, + "loss": 0.8772673, + "num_input_tokens_seen": 43384670, + "step": 2032, + "time_per_iteration": 2.7334036827087402 + }, + { + "auxiliary_loss_clip": 0.01104192, + "auxiliary_loss_mlp": 0.01039188, + "balance_loss_clip": 0.90365946, + "balance_loss_mlp": 1.02959716, + "epoch": 0.24445379666927192, + "flos": 22637189289600.0, + "grad_norm": 2.6879571535877504, + "language_loss": 0.72498393, + "learning_rate": 3.536807469251836e-06, + "loss": 0.74641764, + "num_input_tokens_seen": 43403825, + "step": 2033, + "time_per_iteration": 2.8355095386505127 + }, + { + "auxiliary_loss_clip": 0.0114077, + "auxiliary_loss_mlp": 0.01038584, + "balance_loss_clip": 0.94427097, + "balance_loss_mlp": 1.02877891, + "epoch": 0.24457403955991103, + "flos": 21251612108160.0, + "grad_norm": 2.126981226596408, + "language_loss": 0.83008075, + "learning_rate": 3.5363088378040055e-06, + "loss": 0.85187423, + "num_input_tokens_seen": 43422715, + "step": 2034, + "time_per_iteration": 2.70409893989563 + }, + { + "auxiliary_loss_clip": 0.01112706, + "auxiliary_loss_mlp": 0.00503699, + "balance_loss_clip": 1.03493977, + "balance_loss_mlp": 0.0, + "epoch": 0.2446942824505501, + "flos": 66997820764800.0, + "grad_norm": 0.7518614186343585, + "language_loss": 0.64330018, + "learning_rate": 3.5358099732993463e-06, + "loss": 0.65946424, + "num_input_tokens_seen": 43481825, + "step": 2035, + "time_per_iteration": 3.163702964782715 + }, + { + "auxiliary_loss_clip": 0.01168508, + "auxiliary_loss_mlp": 0.01031303, + "balance_loss_clip": 0.98735058, + "balance_loss_mlp": 1.02228451, + "epoch": 0.2448145253411892, + "flos": 20411122792320.0, + "grad_norm": 2.453044366941547, + "language_loss": 0.8963877, + "learning_rate": 3.535310875813535e-06, + "loss": 0.91838586, + "num_input_tokens_seen": 43500220, + "step": 2036, + "time_per_iteration": 2.727975368499756 + }, + { + "auxiliary_loss_clip": 0.01182554, + "auxiliary_loss_mlp": 0.01036981, + "balance_loss_clip": 1.02399707, + "balance_loss_mlp": 1.02748609, + "epoch": 0.2449347682318283, + "flos": 28804739080320.0, + "grad_norm": 1.5874931086842097, + "language_loss": 0.81747866, + "learning_rate": 3.5348115454222843e-06, + "loss": 0.839674, + "num_input_tokens_seen": 43522805, + "step": 2037, + "time_per_iteration": 2.700160503387451 + }, + { + "auxiliary_loss_clip": 0.01156274, + "auxiliary_loss_mlp": 0.01033868, + "balance_loss_clip": 0.9823041, + "balance_loss_mlp": 1.02405095, + "epoch": 0.2450550111224674, + "flos": 22528990546560.0, + "grad_norm": 1.8628976843964709, + "language_loss": 0.86320293, + "learning_rate": 3.5343119822013425e-06, + "loss": 0.88510436, + "num_input_tokens_seen": 43541915, + "step": 2038, + "time_per_iteration": 2.7215044498443604 + }, + { + "auxiliary_loss_clip": 0.01190755, + "auxiliary_loss_mlp": 0.01034593, + "balance_loss_clip": 1.02512574, + "balance_loss_mlp": 1.0245676, + "epoch": 0.24517525401310647, + "flos": 21759137326080.0, + "grad_norm": 1.8387901514873752, + "language_loss": 0.77504033, + "learning_rate": 3.533812186226493e-06, + "loss": 0.79729378, + "num_input_tokens_seen": 43562625, + "step": 2039, + "time_per_iteration": 2.704582452774048 + }, + { + "auxiliary_loss_clip": 0.01206284, + "auxiliary_loss_mlp": 0.01027805, + "balance_loss_clip": 1.06181896, + "balance_loss_mlp": 1.01916838, + "epoch": 0.24529549690374555, + "flos": 25043311687680.0, + "grad_norm": 1.7724425313423664, + "language_loss": 0.7578311, + "learning_rate": 3.5333121575735545e-06, + "loss": 0.78017193, + "num_input_tokens_seen": 43582265, + "step": 2040, + "time_per_iteration": 2.8457703590393066 + }, + { + "auxiliary_loss_clip": 0.01160331, + "auxiliary_loss_mlp": 0.01027646, + "balance_loss_clip": 0.9866817, + "balance_loss_mlp": 1.01899123, + "epoch": 0.24541573979438466, + "flos": 32123638915200.0, + "grad_norm": 1.7787051483906595, + "language_loss": 0.75588727, + "learning_rate": 3.532811896318381e-06, + "loss": 0.77776694, + "num_input_tokens_seen": 43604335, + "step": 2041, + "time_per_iteration": 2.8031368255615234 + }, + { + "auxiliary_loss_clip": 0.01137576, + "auxiliary_loss_mlp": 0.01032664, + "balance_loss_clip": 0.94656384, + "balance_loss_mlp": 1.02284122, + "epoch": 0.24553598268502375, + "flos": 31357556622720.0, + "grad_norm": 3.218548877039455, + "language_loss": 0.81419921, + "learning_rate": 3.5323114025368615e-06, + "loss": 0.83590156, + "num_input_tokens_seen": 43619400, + "step": 2042, + "time_per_iteration": 2.8162543773651123 + }, + { + "auxiliary_loss_clip": 0.01178411, + "auxiliary_loss_mlp": 0.01038662, + "balance_loss_clip": 1.01993418, + "balance_loss_mlp": 1.02969158, + "epoch": 0.24565622557566283, + "flos": 14027462824320.0, + "grad_norm": 1.9955432291528732, + "language_loss": 0.81633878, + "learning_rate": 3.53181067630492e-06, + "loss": 0.8385095, + "num_input_tokens_seen": 43636870, + "step": 2043, + "time_per_iteration": 2.620617151260376 + }, + { + "auxiliary_loss_clip": 0.01151138, + "auxiliary_loss_mlp": 0.01031513, + "balance_loss_clip": 0.98275673, + "balance_loss_mlp": 1.02316868, + "epoch": 0.24577646846630194, + "flos": 16581465515520.0, + "grad_norm": 1.6742608513238029, + "language_loss": 0.75607014, + "learning_rate": 3.5313097176985175e-06, + "loss": 0.77789664, + "num_input_tokens_seen": 43655180, + "step": 2044, + "time_per_iteration": 3.6722002029418945 + }, + { + "auxiliary_loss_clip": 0.01187286, + "auxiliary_loss_mlp": 0.01034813, + "balance_loss_clip": 1.028386, + "balance_loss_mlp": 1.02585411, + "epoch": 0.24589671135694102, + "flos": 18807424272000.0, + "grad_norm": 2.109770052821565, + "language_loss": 0.81014454, + "learning_rate": 3.5308085267936482e-06, + "loss": 0.83236551, + "num_input_tokens_seen": 43672895, + "step": 2045, + "time_per_iteration": 3.5645639896392822 + }, + { + "auxiliary_loss_clip": 0.0108628, + "auxiliary_loss_mlp": 0.00510854, + "balance_loss_clip": 0.86830783, + "balance_loss_mlp": 0.0, + "epoch": 0.2460169542475801, + "flos": 19938538529280.0, + "grad_norm": 1.6671887438120014, + "language_loss": 0.89829063, + "learning_rate": 3.530307103666342e-06, + "loss": 0.914262, + "num_input_tokens_seen": 43691975, + "step": 2046, + "time_per_iteration": 2.7417380809783936 + }, + { + "auxiliary_loss_clip": 0.01139719, + "auxiliary_loss_mlp": 0.0103285, + "balance_loss_clip": 0.94651228, + "balance_loss_mlp": 1.0240407, + "epoch": 0.24613719713821922, + "flos": 24171221381760.0, + "grad_norm": 1.7848152302318105, + "language_loss": 0.80040413, + "learning_rate": 3.5298054483926658e-06, + "loss": 0.82212985, + "num_input_tokens_seen": 43712670, + "step": 2047, + "time_per_iteration": 2.73877215385437 + }, + { + "auxiliary_loss_clip": 0.01192092, + "auxiliary_loss_mlp": 0.01028633, + "balance_loss_clip": 1.02656138, + "balance_loss_mlp": 1.01975155, + "epoch": 0.2462574400288583, + "flos": 30221055325440.0, + "grad_norm": 2.2125336862453984, + "language_loss": 0.83007425, + "learning_rate": 3.5293035610487187e-06, + "loss": 0.85228145, + "num_input_tokens_seen": 43732035, + "step": 2048, + "time_per_iteration": 3.6154768466949463 + }, + { + "auxiliary_loss_clip": 0.01064951, + "auxiliary_loss_mlp": 0.01005845, + "balance_loss_clip": 0.95951849, + "balance_loss_mlp": 1.00394976, + "epoch": 0.24637768291949738, + "flos": 68943030819840.0, + "grad_norm": 0.7358719556799475, + "language_loss": 0.61989582, + "learning_rate": 3.5288014417106374e-06, + "loss": 0.64060378, + "num_input_tokens_seen": 43798055, + "step": 2049, + "time_per_iteration": 4.172605037689209 + }, + { + "auxiliary_loss_clip": 0.01131991, + "auxiliary_loss_mlp": 0.01031962, + "balance_loss_clip": 0.94428957, + "balance_loss_mlp": 1.02242482, + "epoch": 0.24649792581013646, + "flos": 34383999922560.0, + "grad_norm": 2.840194654205181, + "language_loss": 0.75756282, + "learning_rate": 3.528299090454593e-06, + "loss": 0.77920234, + "num_input_tokens_seen": 43818590, + "step": 2050, + "time_per_iteration": 2.848954916000366 + }, + { + "auxiliary_loss_clip": 0.01183944, + "auxiliary_loss_mlp": 0.01035894, + "balance_loss_clip": 1.02140653, + "balance_loss_mlp": 1.02579117, + "epoch": 0.24661816870077558, + "flos": 19680448331520.0, + "grad_norm": 2.2135566308100074, + "language_loss": 0.82649654, + "learning_rate": 3.527796507356792e-06, + "loss": 0.84869492, + "num_input_tokens_seen": 43832480, + "step": 2051, + "time_per_iteration": 2.6623222827911377 + }, + { + "auxiliary_loss_clip": 0.01186997, + "auxiliary_loss_mlp": 0.01035904, + "balance_loss_clip": 1.02324831, + "balance_loss_mlp": 1.02653992, + "epoch": 0.24673841159141466, + "flos": 20002279213440.0, + "grad_norm": 2.9382989419568073, + "language_loss": 0.89531815, + "learning_rate": 3.527293692493475e-06, + "loss": 0.91754723, + "num_input_tokens_seen": 43848345, + "step": 2052, + "time_per_iteration": 2.6407220363616943 + }, + { + "auxiliary_loss_clip": 0.01186751, + "auxiliary_loss_mlp": 0.01037212, + "balance_loss_clip": 1.02306104, + "balance_loss_mlp": 1.02770495, + "epoch": 0.24685865448205374, + "flos": 21646593037440.0, + "grad_norm": 3.1208116707811735, + "language_loss": 0.73161983, + "learning_rate": 3.52679064594092e-06, + "loss": 0.7538594, + "num_input_tokens_seen": 43865685, + "step": 2053, + "time_per_iteration": 2.7829880714416504 + }, + { + "auxiliary_loss_clip": 0.01096887, + "auxiliary_loss_mlp": 0.01036978, + "balance_loss_clip": 0.89918828, + "balance_loss_mlp": 1.02829933, + "epoch": 0.24697889737269285, + "flos": 17960470508160.0, + "grad_norm": 2.2757101988657937, + "language_loss": 0.7503196, + "learning_rate": 3.5262873677754375e-06, + "loss": 0.7716583, + "num_input_tokens_seen": 43883690, + "step": 2054, + "time_per_iteration": 2.734684944152832 + }, + { + "auxiliary_loss_clip": 0.01207335, + "auxiliary_loss_mlp": 0.01036157, + "balance_loss_clip": 1.06339383, + "balance_loss_mlp": 1.02713931, + "epoch": 0.24709914026333193, + "flos": 27344611221120.0, + "grad_norm": 1.8132178454872627, + "language_loss": 0.80155659, + "learning_rate": 3.5257838580733745e-06, + "loss": 0.82399154, + "num_input_tokens_seen": 43903295, + "step": 2055, + "time_per_iteration": 2.68743634223938 + }, + { + "auxiliary_loss_clip": 0.01189225, + "auxiliary_loss_mlp": 0.01033568, + "balance_loss_clip": 1.02614093, + "balance_loss_mlp": 1.02466893, + "epoch": 0.24721938315397102, + "flos": 19275519335040.0, + "grad_norm": 1.9316846951825535, + "language_loss": 0.87154341, + "learning_rate": 3.5252801169111138e-06, + "loss": 0.89377141, + "num_input_tokens_seen": 43920960, + "step": 2056, + "time_per_iteration": 2.612957715988159 + }, + { + "auxiliary_loss_clip": 0.01160519, + "auxiliary_loss_mlp": 0.01030716, + "balance_loss_clip": 0.98880148, + "balance_loss_mlp": 1.02194226, + "epoch": 0.2473396260446101, + "flos": 23185796688000.0, + "grad_norm": 1.8077358898733427, + "language_loss": 0.7958284, + "learning_rate": 3.524776144365072e-06, + "loss": 0.8177408, + "num_input_tokens_seen": 43939415, + "step": 2057, + "time_per_iteration": 2.716123104095459 + }, + { + "auxiliary_loss_clip": 0.01151824, + "auxiliary_loss_mlp": 0.01036095, + "balance_loss_clip": 0.9851228, + "balance_loss_mlp": 1.02713001, + "epoch": 0.2474598689352492, + "flos": 21142443697920.0, + "grad_norm": 1.6622235503553329, + "language_loss": 0.79411292, + "learning_rate": 3.5242719405117016e-06, + "loss": 0.81599212, + "num_input_tokens_seen": 43959220, + "step": 2058, + "time_per_iteration": 2.688420534133911 + }, + { + "auxiliary_loss_clip": 0.01163199, + "auxiliary_loss_mlp": 0.00511383, + "balance_loss_clip": 0.98714948, + "balance_loss_mlp": 0.0, + "epoch": 0.2475801118258883, + "flos": 21648352803840.0, + "grad_norm": 2.5576770807108318, + "language_loss": 0.74567199, + "learning_rate": 3.5237675054274893e-06, + "loss": 0.76241779, + "num_input_tokens_seen": 43978420, + "step": 2059, + "time_per_iteration": 2.6874403953552246 + }, + { + "auxiliary_loss_clip": 0.0118247, + "auxiliary_loss_mlp": 0.0103272, + "balance_loss_clip": 1.02306819, + "balance_loss_mlp": 1.02324259, + "epoch": 0.24770035471652738, + "flos": 22674500542080.0, + "grad_norm": 2.1641041440127724, + "language_loss": 0.79919386, + "learning_rate": 3.5232628391889584e-06, + "loss": 0.82134581, + "num_input_tokens_seen": 43996710, + "step": 2060, + "time_per_iteration": 2.6440906524658203 + }, + { + "auxiliary_loss_clip": 0.01106406, + "auxiliary_loss_mlp": 0.01032666, + "balance_loss_clip": 0.90711296, + "balance_loss_mlp": 1.02376068, + "epoch": 0.2478205976071665, + "flos": 22163814927360.0, + "grad_norm": 3.993878903242693, + "language_loss": 0.64444911, + "learning_rate": 3.522757941872666e-06, + "loss": 0.66583979, + "num_input_tokens_seen": 44014865, + "step": 2061, + "time_per_iteration": 2.758127450942993 + }, + { + "auxiliary_loss_clip": 0.01209229, + "auxiliary_loss_mlp": 0.00511679, + "balance_loss_clip": 1.06494725, + "balance_loss_mlp": 0.0, + "epoch": 0.24794084049780557, + "flos": 24973106555520.0, + "grad_norm": 1.635775086849135, + "language_loss": 0.82837325, + "learning_rate": 3.5222528135552042e-06, + "loss": 0.84558231, + "num_input_tokens_seen": 44036325, + "step": 2062, + "time_per_iteration": 2.662923574447632 + }, + { + "auxiliary_loss_clip": 0.01179179, + "auxiliary_loss_mlp": 0.01028535, + "balance_loss_clip": 1.02423608, + "balance_loss_mlp": 1.01922488, + "epoch": 0.24806108338844465, + "flos": 18296379521280.0, + "grad_norm": 1.7036192020790915, + "language_loss": 0.80255401, + "learning_rate": 3.521747454313201e-06, + "loss": 0.82463121, + "num_input_tokens_seen": 44055005, + "step": 2063, + "time_per_iteration": 2.660763740539551 + }, + { + "auxiliary_loss_clip": 0.01126653, + "auxiliary_loss_mlp": 0.01035431, + "balance_loss_clip": 0.94266093, + "balance_loss_mlp": 1.02585316, + "epoch": 0.24818132627908374, + "flos": 19282163351040.0, + "grad_norm": 2.176460044819165, + "language_loss": 0.66705644, + "learning_rate": 3.521241864223319e-06, + "loss": 0.68867731, + "num_input_tokens_seen": 44073965, + "step": 2064, + "time_per_iteration": 2.7619543075561523 + }, + { + "auxiliary_loss_clip": 0.01076759, + "auxiliary_loss_mlp": 0.01019928, + "balance_loss_clip": 0.96275842, + "balance_loss_mlp": 1.01765084, + "epoch": 0.24830156916972285, + "flos": 70285837881600.0, + "grad_norm": 0.7915977432185811, + "language_loss": 0.62082922, + "learning_rate": 3.5207360433622552e-06, + "loss": 0.64179599, + "num_input_tokens_seen": 44135965, + "step": 2065, + "time_per_iteration": 3.303513526916504 + }, + { + "auxiliary_loss_clip": 0.01154609, + "auxiliary_loss_mlp": 0.01036971, + "balance_loss_clip": 0.98633814, + "balance_loss_mlp": 1.02823901, + "epoch": 0.24842181206036193, + "flos": 40409128287360.0, + "grad_norm": 2.6160796666102906, + "language_loss": 0.74439806, + "learning_rate": 3.5202299918067437e-06, + "loss": 0.76631391, + "num_input_tokens_seen": 44159560, + "step": 2066, + "time_per_iteration": 2.9218475818634033 + }, + { + "auxiliary_loss_clip": 0.01180213, + "auxiliary_loss_mlp": 0.01030052, + "balance_loss_clip": 1.02392459, + "balance_loss_mlp": 1.02143955, + "epoch": 0.248542054951001, + "flos": 20082432412800.0, + "grad_norm": 2.2081636677082845, + "language_loss": 0.69338083, + "learning_rate": 3.519723709633551e-06, + "loss": 0.71548343, + "num_input_tokens_seen": 44178320, + "step": 2067, + "time_per_iteration": 2.641209840774536 + }, + { + "auxiliary_loss_clip": 0.01154189, + "auxiliary_loss_mlp": 0.01036752, + "balance_loss_clip": 0.98564804, + "balance_loss_mlp": 1.02772164, + "epoch": 0.24866229784164012, + "flos": 23513948363520.0, + "grad_norm": 1.9545451145576835, + "language_loss": 0.83398807, + "learning_rate": 3.519217196919479e-06, + "loss": 0.85589743, + "num_input_tokens_seen": 44197305, + "step": 2068, + "time_per_iteration": 2.7651596069335938 + }, + { + "auxiliary_loss_clip": 0.01160482, + "auxiliary_loss_mlp": 0.01034917, + "balance_loss_clip": 0.98655462, + "balance_loss_mlp": 1.02585125, + "epoch": 0.2487825407322792, + "flos": 19865101173120.0, + "grad_norm": 1.7911352002659706, + "language_loss": 0.73610646, + "learning_rate": 3.518710453741367e-06, + "loss": 0.75806046, + "num_input_tokens_seen": 44216505, + "step": 2069, + "time_per_iteration": 3.58835768699646 + }, + { + "auxiliary_loss_clip": 0.01153105, + "auxiliary_loss_mlp": 0.00511732, + "balance_loss_clip": 0.98411238, + "balance_loss_mlp": 0.0, + "epoch": 0.2489027836229183, + "flos": 22017622573440.0, + "grad_norm": 2.0616787407125488, + "language_loss": 0.67683357, + "learning_rate": 3.518203480176086e-06, + "loss": 0.69348198, + "num_input_tokens_seen": 44235435, + "step": 2070, + "time_per_iteration": 2.651175022125244 + }, + { + "auxiliary_loss_clip": 0.01063199, + "auxiliary_loss_mlp": 0.01035865, + "balance_loss_clip": 0.82657897, + "balance_loss_mlp": 1.02721071, + "epoch": 0.2490230265135574, + "flos": 23294354567040.0, + "grad_norm": 1.7389228302511783, + "language_loss": 0.80669808, + "learning_rate": 3.517696276300545e-06, + "loss": 0.82768869, + "num_input_tokens_seen": 44256975, + "step": 2071, + "time_per_iteration": 3.7860987186431885 + }, + { + "auxiliary_loss_clip": 0.01182113, + "auxiliary_loss_mlp": 0.01038977, + "balance_loss_clip": 1.02571809, + "balance_loss_mlp": 1.02955317, + "epoch": 0.24914326940419648, + "flos": 19826784339840.0, + "grad_norm": 9.31946342432641, + "language_loss": 0.69755512, + "learning_rate": 3.517188842191685e-06, + "loss": 0.71976602, + "num_input_tokens_seen": 44275125, + "step": 2072, + "time_per_iteration": 2.749282121658325 + }, + { + "auxiliary_loss_clip": 0.01180948, + "auxiliary_loss_mlp": 0.01032363, + "balance_loss_clip": 1.02431607, + "balance_loss_mlp": 1.02321315, + "epoch": 0.24926351229483557, + "flos": 20229271211520.0, + "grad_norm": 2.006381058174746, + "language_loss": 0.74003196, + "learning_rate": 3.5166811779264837e-06, + "loss": 0.76216507, + "num_input_tokens_seen": 44295445, + "step": 2073, + "time_per_iteration": 3.58414626121521 + }, + { + "auxiliary_loss_clip": 0.01205777, + "auxiliary_loss_mlp": 0.01030508, + "balance_loss_clip": 1.0608989, + "balance_loss_mlp": 1.0212872, + "epoch": 0.24938375518547465, + "flos": 23294570048640.0, + "grad_norm": 2.279457707540772, + "language_loss": 0.78434777, + "learning_rate": 3.5161732835819545e-06, + "loss": 0.8067106, + "num_input_tokens_seen": 44314755, + "step": 2074, + "time_per_iteration": 2.689035654067993 + }, + { + "auxiliary_loss_clip": 0.0120955, + "auxiliary_loss_mlp": 0.01035978, + "balance_loss_clip": 1.06654847, + "balance_loss_mlp": 1.02710831, + "epoch": 0.24950399807611376, + "flos": 17311673099520.0, + "grad_norm": 2.1820011559884076, + "language_loss": 0.83057076, + "learning_rate": 3.515665159235143e-06, + "loss": 0.85302603, + "num_input_tokens_seen": 44333640, + "step": 2075, + "time_per_iteration": 3.6542675495147705 + }, + { + "auxiliary_loss_clip": 0.01156525, + "auxiliary_loss_mlp": 0.01037969, + "balance_loss_clip": 0.98186201, + "balance_loss_mlp": 1.02960098, + "epoch": 0.24962424096675284, + "flos": 19024863252480.0, + "grad_norm": 1.722743254487062, + "language_loss": 0.75148636, + "learning_rate": 3.5151568049631318e-06, + "loss": 0.7734313, + "num_input_tokens_seen": 44352355, + "step": 2076, + "time_per_iteration": 2.7377939224243164 + }, + { + "auxiliary_loss_clip": 0.01208777, + "auxiliary_loss_mlp": 0.01036872, + "balance_loss_clip": 1.06372952, + "balance_loss_mlp": 1.02781212, + "epoch": 0.24974448385739192, + "flos": 33398790710400.0, + "grad_norm": 1.8019115860719186, + "language_loss": 0.80316412, + "learning_rate": 3.5146482208430385e-06, + "loss": 0.82562053, + "num_input_tokens_seen": 44374185, + "step": 2077, + "time_per_iteration": 2.7499423027038574 + }, + { + "auxiliary_loss_clip": 0.01073817, + "auxiliary_loss_mlp": 0.01030402, + "balance_loss_clip": 0.86389685, + "balance_loss_mlp": 1.02107418, + "epoch": 0.24986472674803104, + "flos": 30007279532160.0, + "grad_norm": 7.09115172429214, + "language_loss": 0.67821914, + "learning_rate": 3.514139406952014e-06, + "loss": 0.69926131, + "num_input_tokens_seen": 44396210, + "step": 2078, + "time_per_iteration": 2.859299421310425 + }, + { + "auxiliary_loss_clip": 0.01182624, + "auxiliary_loss_mlp": 0.0103402, + "balance_loss_clip": 1.02508616, + "balance_loss_mlp": 1.02504933, + "epoch": 0.24998496963867012, + "flos": 26613074833920.0, + "grad_norm": 2.379591785434107, + "language_loss": 0.83569479, + "learning_rate": 3.5136303633672454e-06, + "loss": 0.85786122, + "num_input_tokens_seen": 44416340, + "step": 2079, + "time_per_iteration": 2.763166666030884 + }, + { + "auxiliary_loss_clip": 0.01140344, + "auxiliary_loss_mlp": 0.00511162, + "balance_loss_clip": 0.94647789, + "balance_loss_mlp": 0.0, + "epoch": 0.25010521252930923, + "flos": 23553989049600.0, + "grad_norm": 1.8492058112480174, + "language_loss": 0.7467742, + "learning_rate": 3.5131210901659544e-06, + "loss": 0.76328921, + "num_input_tokens_seen": 44438095, + "step": 2080, + "time_per_iteration": 2.8324618339538574 + }, + { + "auxiliary_loss_clip": 0.01126981, + "auxiliary_loss_mlp": 0.0102931, + "balance_loss_clip": 0.94309306, + "balance_loss_mlp": 1.02056646, + "epoch": 0.2502254554199483, + "flos": 23441193365760.0, + "grad_norm": 2.6484841020590935, + "language_loss": 0.82213968, + "learning_rate": 3.5126115874253967e-06, + "loss": 0.84370255, + "num_input_tokens_seen": 44457650, + "step": 2081, + "time_per_iteration": 2.7996764183044434 + }, + { + "auxiliary_loss_clip": 0.0113592, + "auxiliary_loss_mlp": 0.01029182, + "balance_loss_clip": 0.9497757, + "balance_loss_mlp": 1.02070057, + "epoch": 0.2503456983105874, + "flos": 28761681651840.0, + "grad_norm": 1.9912838405892315, + "language_loss": 0.80816841, + "learning_rate": 3.5121018552228644e-06, + "loss": 0.82981944, + "num_input_tokens_seen": 44476155, + "step": 2082, + "time_per_iteration": 2.795077085494995 + }, + { + "auxiliary_loss_clip": 0.01131553, + "auxiliary_loss_mlp": 0.01028766, + "balance_loss_clip": 0.94399095, + "balance_loss_mlp": 1.01946199, + "epoch": 0.2504659412012265, + "flos": 18770256673920.0, + "grad_norm": 5.937716757250502, + "language_loss": 0.76060402, + "learning_rate": 3.5115918936356827e-06, + "loss": 0.78220725, + "num_input_tokens_seen": 44492910, + "step": 2083, + "time_per_iteration": 2.761518955230713 + }, + { + "auxiliary_loss_clip": 0.01112079, + "auxiliary_loss_mlp": 0.01034296, + "balance_loss_clip": 0.94058675, + "balance_loss_mlp": 1.02622557, + "epoch": 0.25058618409186556, + "flos": 16873383346560.0, + "grad_norm": 1.931110060304116, + "language_loss": 0.78658777, + "learning_rate": 3.5110817027412123e-06, + "loss": 0.80805159, + "num_input_tokens_seen": 44512000, + "step": 2084, + "time_per_iteration": 2.812471628189087 + }, + { + "auxiliary_loss_clip": 0.01126746, + "auxiliary_loss_mlp": 0.01035167, + "balance_loss_clip": 0.94304425, + "balance_loss_mlp": 1.02624393, + "epoch": 0.25070642698250467, + "flos": 24425540651520.0, + "grad_norm": 3.760832915257642, + "language_loss": 0.68787277, + "learning_rate": 3.5105712826168493e-06, + "loss": 0.70949191, + "num_input_tokens_seen": 44531650, + "step": 2085, + "time_per_iteration": 2.8870508670806885 + }, + { + "auxiliary_loss_clip": 0.01181213, + "auxiliary_loss_mlp": 0.00510467, + "balance_loss_clip": 1.02323854, + "balance_loss_mlp": 0.0, + "epoch": 0.2508266698731437, + "flos": 20260944028800.0, + "grad_norm": 2.1628968136648288, + "language_loss": 0.71100521, + "learning_rate": 3.5100606333400235e-06, + "loss": 0.72792202, + "num_input_tokens_seen": 44548785, + "step": 2086, + "time_per_iteration": 2.7082760334014893 + }, + { + "auxiliary_loss_clip": 0.01167053, + "auxiliary_loss_mlp": 0.0103472, + "balance_loss_clip": 0.98557037, + "balance_loss_mlp": 1.0254755, + "epoch": 0.25094691276378284, + "flos": 19245318975360.0, + "grad_norm": 2.2674885294928506, + "language_loss": 0.77013737, + "learning_rate": 3.5095497549882006e-06, + "loss": 0.79215515, + "num_input_tokens_seen": 44567230, + "step": 2087, + "time_per_iteration": 2.8476545810699463 + }, + { + "auxiliary_loss_clip": 0.01184795, + "auxiliary_loss_mlp": 0.01032127, + "balance_loss_clip": 1.02752328, + "balance_loss_mlp": 1.02303755, + "epoch": 0.25106715565442195, + "flos": 26943237671040.0, + "grad_norm": 2.6551999292046506, + "language_loss": 0.72439855, + "learning_rate": 3.50903864763888e-06, + "loss": 0.74656773, + "num_input_tokens_seen": 44588020, + "step": 2088, + "time_per_iteration": 2.749335289001465 + }, + { + "auxiliary_loss_clip": 0.01185026, + "auxiliary_loss_mlp": 0.01034424, + "balance_loss_clip": 1.0236454, + "balance_loss_mlp": 1.02493525, + "epoch": 0.251187398545061, + "flos": 48359570572800.0, + "grad_norm": 2.377431770230896, + "language_loss": 0.76666856, + "learning_rate": 3.5085273113695965e-06, + "loss": 0.78886306, + "num_input_tokens_seen": 44612590, + "step": 2089, + "time_per_iteration": 2.87873911857605 + }, + { + "auxiliary_loss_clip": 0.01208314, + "auxiliary_loss_mlp": 0.01039523, + "balance_loss_clip": 1.06388998, + "balance_loss_mlp": 1.03040349, + "epoch": 0.2513076414357001, + "flos": 27016100409600.0, + "grad_norm": 2.140447068493039, + "language_loss": 0.78274035, + "learning_rate": 3.508015746257919e-06, + "loss": 0.8052187, + "num_input_tokens_seen": 44631630, + "step": 2090, + "time_per_iteration": 2.738525152206421 + }, + { + "auxiliary_loss_clip": 0.01136887, + "auxiliary_loss_mlp": 0.01034222, + "balance_loss_clip": 0.94772929, + "balance_loss_mlp": 1.02547181, + "epoch": 0.2514278843263392, + "flos": 19463619882240.0, + "grad_norm": 2.0136482477468736, + "language_loss": 0.83350307, + "learning_rate": 3.5075039523814518e-06, + "loss": 0.85521412, + "num_input_tokens_seen": 44650820, + "step": 2091, + "time_per_iteration": 2.7232563495635986 + }, + { + "auxiliary_loss_clip": 0.01186353, + "auxiliary_loss_mlp": 0.01033145, + "balance_loss_clip": 1.02234542, + "balance_loss_mlp": 1.02367973, + "epoch": 0.2515481272169783, + "flos": 16866092885760.0, + "grad_norm": 2.4481016993454197, + "language_loss": 0.81739235, + "learning_rate": 3.506991929817834e-06, + "loss": 0.83958733, + "num_input_tokens_seen": 44667540, + "step": 2092, + "time_per_iteration": 2.6883862018585205 + }, + { + "auxiliary_loss_clip": 0.01205373, + "auxiliary_loss_mlp": 0.01029224, + "balance_loss_clip": 1.06542015, + "balance_loss_mlp": 1.02052164, + "epoch": 0.2516683701076174, + "flos": 23732464752000.0, + "grad_norm": 1.8142118198813344, + "language_loss": 0.82581186, + "learning_rate": 3.506479678644738e-06, + "loss": 0.84815788, + "num_input_tokens_seen": 44687935, + "step": 2093, + "time_per_iteration": 3.0100371837615967 + }, + { + "auxiliary_loss_clip": 0.01100298, + "auxiliary_loss_mlp": 0.01030497, + "balance_loss_clip": 0.9066186, + "balance_loss_mlp": 1.0217526, + "epoch": 0.2517886129982565, + "flos": 27635954434560.0, + "grad_norm": 2.434967422160404, + "language_loss": 0.73613966, + "learning_rate": 3.505967198939873e-06, + "loss": 0.7574476, + "num_input_tokens_seen": 44704975, + "step": 2094, + "time_per_iteration": 2.8471362590789795 + }, + { + "auxiliary_loss_clip": 0.01152142, + "auxiliary_loss_mlp": 0.01028581, + "balance_loss_clip": 0.98058552, + "balance_loss_mlp": 1.01956832, + "epoch": 0.25190885588889556, + "flos": 38104596529920.0, + "grad_norm": 1.9157722547504497, + "language_loss": 0.7831403, + "learning_rate": 3.5054544907809813e-06, + "loss": 0.80494756, + "num_input_tokens_seen": 44725475, + "step": 2095, + "time_per_iteration": 3.7459475994110107 + }, + { + "auxiliary_loss_clip": 0.01156371, + "auxiliary_loss_mlp": 0.00511389, + "balance_loss_clip": 0.98794395, + "balance_loss_mlp": 0.0, + "epoch": 0.25202909877953467, + "flos": 22269894768000.0, + "grad_norm": 3.1311246665955106, + "language_loss": 0.81037474, + "learning_rate": 3.50494155424584e-06, + "loss": 0.82705235, + "num_input_tokens_seen": 44744380, + "step": 2096, + "time_per_iteration": 2.7733154296875 + }, + { + "auxiliary_loss_clip": 0.01185779, + "auxiliary_loss_mlp": 0.0102525, + "balance_loss_clip": 1.02471101, + "balance_loss_mlp": 1.01633954, + "epoch": 0.2521493416701738, + "flos": 21761759018880.0, + "grad_norm": 2.0141264972912056, + "language_loss": 0.83144438, + "learning_rate": 3.504428389412262e-06, + "loss": 0.85355461, + "num_input_tokens_seen": 44765190, + "step": 2097, + "time_per_iteration": 3.6882057189941406 + }, + { + "auxiliary_loss_clip": 0.01179737, + "auxiliary_loss_mlp": 0.0103285, + "balance_loss_clip": 1.02261448, + "balance_loss_mlp": 1.02327168, + "epoch": 0.25226958456081283, + "flos": 27746738956800.0, + "grad_norm": 2.2672146470315653, + "language_loss": 0.73064375, + "learning_rate": 3.5039149963580927e-06, + "loss": 0.75276959, + "num_input_tokens_seen": 44785210, + "step": 2098, + "time_per_iteration": 2.76507568359375 + }, + { + "auxiliary_loss_clip": 0.01153089, + "auxiliary_loss_mlp": 0.01027266, + "balance_loss_clip": 0.98721993, + "balance_loss_mlp": 1.01868927, + "epoch": 0.25238982745145194, + "flos": 30732171903360.0, + "grad_norm": 3.646056281336119, + "language_loss": 0.70085323, + "learning_rate": 3.503401375161215e-06, + "loss": 0.72265679, + "num_input_tokens_seen": 44804955, + "step": 2099, + "time_per_iteration": 3.7376880645751953 + }, + { + "auxiliary_loss_clip": 0.01204562, + "auxiliary_loss_mlp": 0.01032291, + "balance_loss_clip": 1.06195641, + "balance_loss_mlp": 1.02283144, + "epoch": 0.252510070342091, + "flos": 20266331068800.0, + "grad_norm": 1.5966974278249693, + "language_loss": 0.83578634, + "learning_rate": 3.502887525899544e-06, + "loss": 0.85815489, + "num_input_tokens_seen": 44823935, + "step": 2100, + "time_per_iteration": 2.6531982421875 + }, + { + "auxiliary_loss_clip": 0.01158847, + "auxiliary_loss_mlp": 0.01035605, + "balance_loss_clip": 0.98534936, + "balance_loss_mlp": 1.02546048, + "epoch": 0.2526303132327301, + "flos": 22747399194240.0, + "grad_norm": 2.1457273185837837, + "language_loss": 0.82903206, + "learning_rate": 3.50237344865103e-06, + "loss": 0.85097659, + "num_input_tokens_seen": 44844935, + "step": 2101, + "time_per_iteration": 3.685882568359375 + }, + { + "auxiliary_loss_clip": 0.01207727, + "auxiliary_loss_mlp": 0.01031756, + "balance_loss_clip": 1.06269264, + "balance_loss_mlp": 1.02265418, + "epoch": 0.2527505561233692, + "flos": 30263466309120.0, + "grad_norm": 2.101665543297448, + "language_loss": 0.76343501, + "learning_rate": 3.501859143493658e-06, + "loss": 0.7858299, + "num_input_tokens_seen": 44865565, + "step": 2102, + "time_per_iteration": 2.668585777282715 + }, + { + "auxiliary_loss_clip": 0.0111129, + "auxiliary_loss_mlp": 0.0100715, + "balance_loss_clip": 1.03581131, + "balance_loss_mlp": 1.00508738, + "epoch": 0.2528707990140083, + "flos": 58492917164160.0, + "grad_norm": 0.9565840001332645, + "language_loss": 0.60584366, + "learning_rate": 3.5013446105054488e-06, + "loss": 0.62702811, + "num_input_tokens_seen": 44918485, + "step": 2103, + "time_per_iteration": 3.0106582641601562 + }, + { + "auxiliary_loss_clip": 0.01120796, + "auxiliary_loss_mlp": 0.01033054, + "balance_loss_clip": 0.94500196, + "balance_loss_mlp": 1.02398825, + "epoch": 0.2529910419046474, + "flos": 24645134448000.0, + "grad_norm": 1.7564002564410919, + "language_loss": 0.74780834, + "learning_rate": 3.5008298497644555e-06, + "loss": 0.76934683, + "num_input_tokens_seen": 44937530, + "step": 2104, + "time_per_iteration": 2.745490789413452 + }, + { + "auxiliary_loss_clip": 0.01134684, + "auxiliary_loss_mlp": 0.0103253, + "balance_loss_clip": 0.94549727, + "balance_loss_mlp": 1.02356577, + "epoch": 0.2531112847952865, + "flos": 23842135952640.0, + "grad_norm": 1.8694335406141156, + "language_loss": 0.88207936, + "learning_rate": 3.500314861348767e-06, + "loss": 0.90375149, + "num_input_tokens_seen": 44958165, + "step": 2105, + "time_per_iteration": 2.73934268951416 + }, + { + "auxiliary_loss_clip": 0.01121651, + "auxiliary_loss_mlp": 0.01038023, + "balance_loss_clip": 0.94445109, + "balance_loss_mlp": 1.0288794, + "epoch": 0.25323152768592555, + "flos": 16143822207360.0, + "grad_norm": 2.006323857808613, + "language_loss": 0.76963878, + "learning_rate": 3.499799645336507e-06, + "loss": 0.79123551, + "num_input_tokens_seen": 44975060, + "step": 2106, + "time_per_iteration": 2.6532552242279053 + }, + { + "auxiliary_loss_clip": 0.01184027, + "auxiliary_loss_mlp": 0.01035351, + "balance_loss_clip": 1.02604842, + "balance_loss_mlp": 1.02682185, + "epoch": 0.25335177057656466, + "flos": 28405161210240.0, + "grad_norm": 1.8012627171461773, + "language_loss": 0.87006432, + "learning_rate": 3.4992842018058336e-06, + "loss": 0.89225811, + "num_input_tokens_seen": 44997960, + "step": 2107, + "time_per_iteration": 2.7242679595947266 + }, + { + "auxiliary_loss_clip": 0.01137564, + "auxiliary_loss_mlp": 0.01032989, + "balance_loss_clip": 0.94466496, + "balance_loss_mlp": 1.02348232, + "epoch": 0.25347201346720377, + "flos": 18799666934400.0, + "grad_norm": 3.0280991843709972, + "language_loss": 0.88843548, + "learning_rate": 3.4987685308349384e-06, + "loss": 0.91014099, + "num_input_tokens_seen": 45015690, + "step": 2108, + "time_per_iteration": 2.785169839859009 + }, + { + "auxiliary_loss_clip": 0.01127318, + "auxiliary_loss_mlp": 0.01032228, + "balance_loss_clip": 0.94024813, + "balance_loss_mlp": 1.02322185, + "epoch": 0.2535922563578428, + "flos": 15815490963840.0, + "grad_norm": 2.1773935755249267, + "language_loss": 0.61191571, + "learning_rate": 3.4982526325020497e-06, + "loss": 0.63351119, + "num_input_tokens_seen": 45032660, + "step": 2109, + "time_per_iteration": 2.7829525470733643 + }, + { + "auxiliary_loss_clip": 0.01156882, + "auxiliary_loss_mlp": 0.01034869, + "balance_loss_clip": 0.98304021, + "balance_loss_mlp": 1.02523077, + "epoch": 0.25371249924848194, + "flos": 16318922031360.0, + "grad_norm": 2.21025507759908, + "language_loss": 0.81795239, + "learning_rate": 3.4977365068854273e-06, + "loss": 0.83986992, + "num_input_tokens_seen": 45048280, + "step": 2110, + "time_per_iteration": 2.703979015350342 + }, + { + "auxiliary_loss_clip": 0.01151451, + "auxiliary_loss_mlp": 0.01033913, + "balance_loss_clip": 0.98533088, + "balance_loss_mlp": 1.02442431, + "epoch": 0.25383274213912105, + "flos": 21761615364480.0, + "grad_norm": 1.7050711323872139, + "language_loss": 0.73385358, + "learning_rate": 3.4972201540633676e-06, + "loss": 0.75570714, + "num_input_tokens_seen": 45067635, + "step": 2111, + "time_per_iteration": 2.720226287841797 + }, + { + "auxiliary_loss_clip": 0.01151534, + "auxiliary_loss_mlp": 0.0103528, + "balance_loss_clip": 0.98498434, + "balance_loss_mlp": 1.02614295, + "epoch": 0.2539529850297601, + "flos": 21396870708480.0, + "grad_norm": 1.8186322505676933, + "language_loss": 0.85135543, + "learning_rate": 3.4967035741142008e-06, + "loss": 0.87322366, + "num_input_tokens_seen": 45086455, + "step": 2112, + "time_per_iteration": 2.730797529220581 + }, + { + "auxiliary_loss_clip": 0.01148229, + "auxiliary_loss_mlp": 0.01035887, + "balance_loss_clip": 0.98726606, + "balance_loss_mlp": 1.02731633, + "epoch": 0.2540732279203992, + "flos": 25228467319680.0, + "grad_norm": 1.9444600099539053, + "language_loss": 0.82098264, + "learning_rate": 3.4961867671162917e-06, + "loss": 0.84282386, + "num_input_tokens_seen": 45106385, + "step": 2113, + "time_per_iteration": 2.779120683670044 + }, + { + "auxiliary_loss_clip": 0.01206894, + "auxiliary_loss_mlp": 0.01033092, + "balance_loss_clip": 1.06181943, + "balance_loss_mlp": 1.02372265, + "epoch": 0.2541934708110383, + "flos": 19427386037760.0, + "grad_norm": 2.8791058826205322, + "language_loss": 0.77251637, + "learning_rate": 3.4956697331480402e-06, + "loss": 0.79491621, + "num_input_tokens_seen": 45124955, + "step": 2114, + "time_per_iteration": 2.6456995010375977 + }, + { + "auxiliary_loss_clip": 0.01139001, + "auxiliary_loss_mlp": 0.01032622, + "balance_loss_clip": 0.94515097, + "balance_loss_mlp": 1.02336526, + "epoch": 0.2543137137016774, + "flos": 23949436855680.0, + "grad_norm": 1.6781110256528537, + "language_loss": 0.79970926, + "learning_rate": 3.495152472287879e-06, + "loss": 0.8214255, + "num_input_tokens_seen": 45145665, + "step": 2115, + "time_per_iteration": 2.703261375427246 + }, + { + "auxiliary_loss_clip": 0.01135376, + "auxiliary_loss_mlp": 0.01033274, + "balance_loss_clip": 0.94834018, + "balance_loss_mlp": 1.02398205, + "epoch": 0.2544339565923165, + "flos": 25593283802880.0, + "grad_norm": 2.3111186429318704, + "language_loss": 0.73825651, + "learning_rate": 3.4946349846142766e-06, + "loss": 0.75994301, + "num_input_tokens_seen": 45164805, + "step": 2116, + "time_per_iteration": 2.7693228721618652 + }, + { + "auxiliary_loss_clip": 0.01204545, + "auxiliary_loss_mlp": 0.01034107, + "balance_loss_clip": 1.06029224, + "balance_loss_mlp": 1.02470696, + "epoch": 0.25455419948295555, + "flos": 21689470897920.0, + "grad_norm": 3.049533954876156, + "language_loss": 0.75520468, + "learning_rate": 3.4941172702057353e-06, + "loss": 0.77759117, + "num_input_tokens_seen": 45184865, + "step": 2117, + "time_per_iteration": 2.6133909225463867 + }, + { + "auxiliary_loss_clip": 0.0115772, + "auxiliary_loss_mlp": 0.01038013, + "balance_loss_clip": 0.98602116, + "balance_loss_mlp": 1.02891779, + "epoch": 0.25467444237359466, + "flos": 26250341339520.0, + "grad_norm": 3.055828923130426, + "language_loss": 0.80808127, + "learning_rate": 3.4935993291407924e-06, + "loss": 0.83003867, + "num_input_tokens_seen": 45203690, + "step": 2118, + "time_per_iteration": 2.6549079418182373 + }, + { + "auxiliary_loss_clip": 0.01153498, + "auxiliary_loss_mlp": 0.010315, + "balance_loss_clip": 0.98313147, + "balance_loss_mlp": 1.02240419, + "epoch": 0.25479468526423377, + "flos": 26979686997120.0, + "grad_norm": 2.6387251542523393, + "language_loss": 0.71160316, + "learning_rate": 3.4930811614980183e-06, + "loss": 0.73345315, + "num_input_tokens_seen": 45225385, + "step": 2119, + "time_per_iteration": 2.701573371887207 + }, + { + "auxiliary_loss_clip": 0.01173435, + "auxiliary_loss_mlp": 0.01037933, + "balance_loss_clip": 1.02079356, + "balance_loss_mlp": 1.02790129, + "epoch": 0.2549149281548728, + "flos": 23475811098240.0, + "grad_norm": 1.7164947629586627, + "language_loss": 0.79493415, + "learning_rate": 3.4925627673560198e-06, + "loss": 0.81704789, + "num_input_tokens_seen": 45246045, + "step": 2120, + "time_per_iteration": 2.746208667755127 + }, + { + "auxiliary_loss_clip": 0.0113432, + "auxiliary_loss_mlp": 0.01034378, + "balance_loss_clip": 0.94741708, + "balance_loss_mlp": 1.0247221, + "epoch": 0.25503517104551193, + "flos": 25812302981760.0, + "grad_norm": 2.5216732457693403, + "language_loss": 0.88408488, + "learning_rate": 3.4920441467934357e-06, + "loss": 0.90577191, + "num_input_tokens_seen": 45266560, + "step": 2121, + "time_per_iteration": 2.793632984161377 + }, + { + "auxiliary_loss_clip": 0.01122948, + "auxiliary_loss_mlp": 0.01031644, + "balance_loss_clip": 0.94414443, + "balance_loss_mlp": 1.02223206, + "epoch": 0.25515541393615104, + "flos": 26645106787200.0, + "grad_norm": 2.8976326652846054, + "language_loss": 0.83190513, + "learning_rate": 3.491525299888941e-06, + "loss": 0.85345101, + "num_input_tokens_seen": 45285405, + "step": 2122, + "time_per_iteration": 3.6624531745910645 + }, + { + "auxiliary_loss_clip": 0.01053561, + "auxiliary_loss_mlp": 0.00503702, + "balance_loss_clip": 0.95701885, + "balance_loss_mlp": 0.0, + "epoch": 0.2552756568267901, + "flos": 65955945847680.0, + "grad_norm": 0.885388491954509, + "language_loss": 0.62699175, + "learning_rate": 3.491006226721244e-06, + "loss": 0.64256442, + "num_input_tokens_seen": 45349615, + "step": 2123, + "time_per_iteration": 4.2466936111450195 + }, + { + "auxiliary_loss_clip": 0.01159498, + "auxiliary_loss_mlp": 0.00511081, + "balance_loss_clip": 0.98647976, + "balance_loss_mlp": 0.0, + "epoch": 0.2553958997174292, + "flos": 17931096161280.0, + "grad_norm": 2.05004352850379, + "language_loss": 0.77665246, + "learning_rate": 3.4904869273690882e-06, + "loss": 0.79335821, + "num_input_tokens_seen": 45367505, + "step": 2124, + "time_per_iteration": 2.6694607734680176 + }, + { + "auxiliary_loss_clip": 0.01182748, + "auxiliary_loss_mlp": 0.01035319, + "balance_loss_clip": 1.02130818, + "balance_loss_mlp": 1.02674806, + "epoch": 0.2555161426080683, + "flos": 23367791923200.0, + "grad_norm": 15.274960108086967, + "language_loss": 0.89019978, + "learning_rate": 3.489967401911251e-06, + "loss": 0.91238052, + "num_input_tokens_seen": 45386805, + "step": 2125, + "time_per_iteration": 3.6356818675994873 + }, + { + "auxiliary_loss_clip": 0.01209237, + "auxiliary_loss_mlp": 0.01034969, + "balance_loss_clip": 1.06376696, + "balance_loss_mlp": 1.02539062, + "epoch": 0.2556363854987074, + "flos": 40625130723840.0, + "grad_norm": 1.7045213274372477, + "language_loss": 0.69111633, + "learning_rate": 3.4894476504265428e-06, + "loss": 0.71355838, + "num_input_tokens_seen": 45411045, + "step": 2126, + "time_per_iteration": 2.7781219482421875 + }, + { + "auxiliary_loss_clip": 0.01086094, + "auxiliary_loss_mlp": 0.0100426, + "balance_loss_clip": 0.99861848, + "balance_loss_mlp": 1.00175619, + "epoch": 0.2557566283893465, + "flos": 68019443389440.0, + "grad_norm": 0.7857175936039048, + "language_loss": 0.54508358, + "learning_rate": 3.4889276729938104e-06, + "loss": 0.56598711, + "num_input_tokens_seen": 45469575, + "step": 2127, + "time_per_iteration": 3.98923659324646 + }, + { + "auxiliary_loss_clip": 0.01155428, + "auxiliary_loss_mlp": 0.01027523, + "balance_loss_clip": 0.98666632, + "balance_loss_mlp": 1.01836801, + "epoch": 0.2558768712799856, + "flos": 22635645004800.0, + "grad_norm": 2.29034732805644, + "language_loss": 0.80639309, + "learning_rate": 3.488407469691934e-06, + "loss": 0.82822263, + "num_input_tokens_seen": 45490270, + "step": 2128, + "time_per_iteration": 2.7567784786224365 + }, + { + "auxiliary_loss_clip": 0.01156075, + "auxiliary_loss_mlp": 0.01036377, + "balance_loss_clip": 0.9837203, + "balance_loss_mlp": 1.02692413, + "epoch": 0.25599711417062465, + "flos": 26396354125440.0, + "grad_norm": 2.3258985194243187, + "language_loss": 0.80974245, + "learning_rate": 3.487887040599828e-06, + "loss": 0.83166701, + "num_input_tokens_seen": 45510070, + "step": 2129, + "time_per_iteration": 2.873894453048706 + }, + { + "auxiliary_loss_clip": 0.01208154, + "auxiliary_loss_mlp": 0.01035977, + "balance_loss_clip": 1.0632838, + "balance_loss_mlp": 1.02679777, + "epoch": 0.25611735706126376, + "flos": 22852042490880.0, + "grad_norm": 2.34013034468376, + "language_loss": 0.76073956, + "learning_rate": 3.4873663857964407e-06, + "loss": 0.78318083, + "num_input_tokens_seen": 45527285, + "step": 2130, + "time_per_iteration": 2.627514600753784 + }, + { + "auxiliary_loss_clip": 0.01108818, + "auxiliary_loss_mlp": 0.01031606, + "balance_loss_clip": 0.9080565, + "balance_loss_mlp": 1.02227747, + "epoch": 0.2562375999519028, + "flos": 23367863750400.0, + "grad_norm": 1.8550102897845433, + "language_loss": 0.66602004, + "learning_rate": 3.4868455053607556e-06, + "loss": 0.6874243, + "num_input_tokens_seen": 45546900, + "step": 2131, + "time_per_iteration": 2.766878366470337 + }, + { + "auxiliary_loss_clip": 0.01185212, + "auxiliary_loss_mlp": 0.0103717, + "balance_loss_clip": 1.02265561, + "balance_loss_mlp": 1.02735949, + "epoch": 0.2563578428425419, + "flos": 22856962654080.0, + "grad_norm": 2.2596162675325795, + "language_loss": 0.71546614, + "learning_rate": 3.486324399371789e-06, + "loss": 0.73768997, + "num_input_tokens_seen": 45566200, + "step": 2132, + "time_per_iteration": 2.674921989440918 + }, + { + "auxiliary_loss_clip": 0.0112753, + "auxiliary_loss_mlp": 0.01033481, + "balance_loss_clip": 0.94655025, + "balance_loss_mlp": 1.02471316, + "epoch": 0.25647808573318104, + "flos": 21653883498240.0, + "grad_norm": 2.655791092381093, + "language_loss": 0.78432435, + "learning_rate": 3.485803067908593e-06, + "loss": 0.80593443, + "num_input_tokens_seen": 45585710, + "step": 2133, + "time_per_iteration": 2.751978635787964 + }, + { + "auxiliary_loss_clip": 0.01062081, + "auxiliary_loss_mlp": 0.01032578, + "balance_loss_clip": 0.86060619, + "balance_loss_mlp": 1.02317846, + "epoch": 0.2565983286238201, + "flos": 33730569659520.0, + "grad_norm": 2.2862444379033655, + "language_loss": 0.79676503, + "learning_rate": 3.485281511050253e-06, + "loss": 0.81771159, + "num_input_tokens_seen": 45607845, + "step": 2134, + "time_per_iteration": 2.905693769454956 + }, + { + "auxiliary_loss_clip": 0.011836, + "auxiliary_loss_mlp": 0.01034291, + "balance_loss_clip": 1.02362132, + "balance_loss_mlp": 1.0249331, + "epoch": 0.2567185715144592, + "flos": 16216002587520.0, + "grad_norm": 2.602086095735063, + "language_loss": 0.89851511, + "learning_rate": 3.484759728875889e-06, + "loss": 0.92069399, + "num_input_tokens_seen": 45623210, + "step": 2135, + "time_per_iteration": 2.5915300846099854 + }, + { + "auxiliary_loss_clip": 0.01096491, + "auxiliary_loss_mlp": 0.0103366, + "balance_loss_clip": 0.90363193, + "balance_loss_mlp": 1.0249753, + "epoch": 0.2568388144050983, + "flos": 17458475984640.0, + "grad_norm": 2.3290925031777263, + "language_loss": 0.80888355, + "learning_rate": 3.4842377214646543e-06, + "loss": 0.830185, + "num_input_tokens_seen": 45641505, + "step": 2136, + "time_per_iteration": 2.7432518005371094 + }, + { + "auxiliary_loss_clip": 0.01202813, + "auxiliary_loss_mlp": 0.01028747, + "balance_loss_clip": 1.06140471, + "balance_loss_mlp": 1.019943, + "epoch": 0.25695905729573737, + "flos": 20887442069760.0, + "grad_norm": 1.7967404650918888, + "language_loss": 0.66700429, + "learning_rate": 3.483715488895737e-06, + "loss": 0.68931991, + "num_input_tokens_seen": 45661835, + "step": 2137, + "time_per_iteration": 2.607570171356201 + }, + { + "auxiliary_loss_clip": 0.01107561, + "auxiliary_loss_mlp": 0.01028902, + "balance_loss_clip": 0.90373284, + "balance_loss_mlp": 1.01973438, + "epoch": 0.2570793001863765, + "flos": 24717278914560.0, + "grad_norm": 1.8421351697960393, + "language_loss": 0.7829991, + "learning_rate": 3.48319303124836e-06, + "loss": 0.80436373, + "num_input_tokens_seen": 45682215, + "step": 2138, + "time_per_iteration": 2.826317310333252 + }, + { + "auxiliary_loss_clip": 0.01154615, + "auxiliary_loss_mlp": 0.01033062, + "balance_loss_clip": 0.98834121, + "balance_loss_mlp": 1.02396083, + "epoch": 0.2571995430770156, + "flos": 26906896085760.0, + "grad_norm": 6.339299415947147, + "language_loss": 0.66601926, + "learning_rate": 3.4826703486017798e-06, + "loss": 0.68789607, + "num_input_tokens_seen": 45701840, + "step": 2139, + "time_per_iteration": 2.721404552459717 + }, + { + "auxiliary_loss_clip": 0.01184034, + "auxiliary_loss_mlp": 0.01032077, + "balance_loss_clip": 1.02846718, + "balance_loss_mlp": 1.02258825, + "epoch": 0.25731978596765465, + "flos": 19792561656960.0, + "grad_norm": 2.6555884788778443, + "language_loss": 0.76869202, + "learning_rate": 3.4821474410352867e-06, + "loss": 0.79085314, + "num_input_tokens_seen": 45720500, + "step": 2140, + "time_per_iteration": 2.724407434463501 + }, + { + "auxiliary_loss_clip": 0.01054772, + "auxiliary_loss_mlp": 0.01006835, + "balance_loss_clip": 0.93166798, + "balance_loss_mlp": 1.00415254, + "epoch": 0.25744002885829376, + "flos": 70564970471040.0, + "grad_norm": 0.8888905316772349, + "language_loss": 0.62611115, + "learning_rate": 3.481624308628205e-06, + "loss": 0.6467272, + "num_input_tokens_seen": 45781870, + "step": 2141, + "time_per_iteration": 3.473367214202881 + }, + { + "auxiliary_loss_clip": 0.01158697, + "auxiliary_loss_mlp": 0.01041479, + "balance_loss_clip": 0.98644412, + "balance_loss_mlp": 1.03169799, + "epoch": 0.25756027174893287, + "flos": 18038181582720.0, + "grad_norm": 4.071408349034683, + "language_loss": 0.99869776, + "learning_rate": 3.481100951459893e-06, + "loss": 1.0206995, + "num_input_tokens_seen": 45794890, + "step": 2142, + "time_per_iteration": 2.6480531692504883 + }, + { + "auxiliary_loss_clip": 0.01177649, + "auxiliary_loss_mlp": 0.01039365, + "balance_loss_clip": 1.02346432, + "balance_loss_mlp": 1.02982855, + "epoch": 0.2576805146395719, + "flos": 22674069578880.0, + "grad_norm": 1.733989841545562, + "language_loss": 0.78864139, + "learning_rate": 3.4805773696097453e-06, + "loss": 0.81081152, + "num_input_tokens_seen": 45815780, + "step": 2143, + "time_per_iteration": 2.773097276687622 + }, + { + "auxiliary_loss_clip": 0.01154385, + "auxiliary_loss_mlp": 0.01031633, + "balance_loss_clip": 0.98925865, + "balance_loss_mlp": 1.02275193, + "epoch": 0.25780075753021103, + "flos": 16472225278080.0, + "grad_norm": 2.0105062446025306, + "language_loss": 0.87578928, + "learning_rate": 3.4800535631571874e-06, + "loss": 0.89764953, + "num_input_tokens_seen": 45831310, + "step": 2144, + "time_per_iteration": 2.6363778114318848 + }, + { + "auxiliary_loss_clip": 0.01160804, + "auxiliary_loss_mlp": 0.01040719, + "balance_loss_clip": 0.98558462, + "balance_loss_mlp": 1.03135562, + "epoch": 0.25792100042085014, + "flos": 22820297846400.0, + "grad_norm": 2.577271159871161, + "language_loss": 0.75804281, + "learning_rate": 3.4795295321816804e-06, + "loss": 0.78005803, + "num_input_tokens_seen": 45850135, + "step": 2145, + "time_per_iteration": 2.7519047260284424 + }, + { + "auxiliary_loss_clip": 0.01147614, + "auxiliary_loss_mlp": 0.01026532, + "balance_loss_clip": 0.98327506, + "balance_loss_mlp": 1.01726317, + "epoch": 0.2580412433114892, + "flos": 18697286194560.0, + "grad_norm": 2.3941805847854987, + "language_loss": 0.91079938, + "learning_rate": 3.47900527676272e-06, + "loss": 0.93254083, + "num_input_tokens_seen": 45868470, + "step": 2146, + "time_per_iteration": 2.7515106201171875 + }, + { + "auxiliary_loss_clip": 0.01208817, + "auxiliary_loss_mlp": 0.01037463, + "balance_loss_clip": 1.066082, + "balance_loss_mlp": 1.02849281, + "epoch": 0.2581614862021283, + "flos": 14283146810880.0, + "grad_norm": 1.8941657041044249, + "language_loss": 0.88549691, + "learning_rate": 3.478480796979835e-06, + "loss": 0.9079597, + "num_input_tokens_seen": 45886355, + "step": 2147, + "time_per_iteration": 3.636007785797119 + }, + { + "auxiliary_loss_clip": 0.01155799, + "auxiliary_loss_mlp": 0.01031466, + "balance_loss_clip": 0.98695779, + "balance_loss_mlp": 1.02168489, + "epoch": 0.25828172909276736, + "flos": 29498281856640.0, + "grad_norm": 1.597629280997296, + "language_loss": 0.77903318, + "learning_rate": 3.4779560929125894e-06, + "loss": 0.80090582, + "num_input_tokens_seen": 45907900, + "step": 2148, + "time_per_iteration": 2.859417200088501 + }, + { + "auxiliary_loss_clip": 0.01051553, + "auxiliary_loss_mlp": 0.01009012, + "balance_loss_clip": 0.93029791, + "balance_loss_mlp": 1.00678253, + "epoch": 0.2584019719834065, + "flos": 67114387376640.0, + "grad_norm": 0.6612487922532267, + "language_loss": 0.56904638, + "learning_rate": 3.4774311646405783e-06, + "loss": 0.589652, + "num_input_tokens_seen": 45977805, + "step": 2149, + "time_per_iteration": 4.342427968978882 + }, + { + "auxiliary_loss_clip": 0.01124416, + "auxiliary_loss_mlp": 0.01033236, + "balance_loss_clip": 0.94388044, + "balance_loss_mlp": 1.02459908, + "epoch": 0.2585222148740456, + "flos": 22893555634560.0, + "grad_norm": 2.169457601353912, + "language_loss": 0.83642143, + "learning_rate": 3.476906012243435e-06, + "loss": 0.85799795, + "num_input_tokens_seen": 45996715, + "step": 2150, + "time_per_iteration": 2.786722421646118 + }, + { + "auxiliary_loss_clip": 0.01174461, + "auxiliary_loss_mlp": 0.01036542, + "balance_loss_clip": 1.02399397, + "balance_loss_mlp": 1.02658772, + "epoch": 0.25864245776468464, + "flos": 28909202808960.0, + "grad_norm": 1.8744239793941369, + "language_loss": 0.81054479, + "learning_rate": 3.476380635800824e-06, + "loss": 0.83265483, + "num_input_tokens_seen": 46017915, + "step": 2151, + "time_per_iteration": 3.234665870666504 + }, + { + "auxiliary_loss_clip": 0.01156062, + "auxiliary_loss_mlp": 0.01030279, + "balance_loss_clip": 0.98823357, + "balance_loss_mlp": 1.02090955, + "epoch": 0.25876270065532375, + "flos": 14793185980800.0, + "grad_norm": 4.772814667654128, + "language_loss": 0.86253816, + "learning_rate": 3.475855035392444e-06, + "loss": 0.88440156, + "num_input_tokens_seen": 46033235, + "step": 2152, + "time_per_iteration": 3.668830394744873 + }, + { + "auxiliary_loss_clip": 0.01080874, + "auxiliary_loss_mlp": 0.01032506, + "balance_loss_clip": 0.86876094, + "balance_loss_mlp": 1.02327883, + "epoch": 0.25888294354596286, + "flos": 60467821810560.0, + "grad_norm": 1.9550475859871241, + "language_loss": 0.71517444, + "learning_rate": 3.475329211098029e-06, + "loss": 0.73630822, + "num_input_tokens_seen": 46056390, + "step": 2153, + "time_per_iteration": 4.003739356994629 + }, + { + "auxiliary_loss_clip": 0.01114537, + "auxiliary_loss_mlp": 0.01036485, + "balance_loss_clip": 0.90985012, + "balance_loss_mlp": 1.0273416, + "epoch": 0.2590031864366019, + "flos": 27851166771840.0, + "grad_norm": 1.696826783307935, + "language_loss": 0.82567316, + "learning_rate": 3.4748031629973453e-06, + "loss": 0.84718335, + "num_input_tokens_seen": 46077120, + "step": 2154, + "time_per_iteration": 3.131757974624634 + }, + { + "auxiliary_loss_clip": 0.0102331, + "auxiliary_loss_mlp": 0.01030566, + "balance_loss_clip": 0.88814378, + "balance_loss_mlp": 1.0285033, + "epoch": 0.25912342932724103, + "flos": 62422444206720.0, + "grad_norm": 0.9076776272529489, + "language_loss": 0.56592083, + "learning_rate": 3.4742768911701944e-06, + "loss": 0.58645958, + "num_input_tokens_seen": 46139815, + "step": 2155, + "time_per_iteration": 3.4292030334472656 + }, + { + "auxiliary_loss_clip": 0.01187791, + "auxiliary_loss_mlp": 0.01038572, + "balance_loss_clip": 1.02801919, + "balance_loss_mlp": 1.02759314, + "epoch": 0.25924367221788014, + "flos": 12378839368320.0, + "grad_norm": 3.363376313354111, + "language_loss": 0.70581782, + "learning_rate": 3.4737503956964113e-06, + "loss": 0.72808135, + "num_input_tokens_seen": 46152120, + "step": 2156, + "time_per_iteration": 2.7675180435180664 + }, + { + "auxiliary_loss_clip": 0.0114884, + "auxiliary_loss_mlp": 0.01032366, + "balance_loss_clip": 0.98133481, + "balance_loss_mlp": 1.02250719, + "epoch": 0.2593639151085192, + "flos": 14575208296320.0, + "grad_norm": 2.509780879574923, + "language_loss": 0.67243171, + "learning_rate": 3.473223676655865e-06, + "loss": 0.69424379, + "num_input_tokens_seen": 46170120, + "step": 2157, + "time_per_iteration": 2.7493937015533447 + }, + { + "auxiliary_loss_clip": 0.01150945, + "auxiliary_loss_mlp": 0.01036754, + "balance_loss_clip": 0.98162019, + "balance_loss_mlp": 1.02671647, + "epoch": 0.2594841579991583, + "flos": 15230937029760.0, + "grad_norm": 2.365489026466343, + "language_loss": 0.79624569, + "learning_rate": 3.472696734128459e-06, + "loss": 0.81812274, + "num_input_tokens_seen": 46187985, + "step": 2158, + "time_per_iteration": 2.6488397121429443 + }, + { + "auxiliary_loss_clip": 0.01184179, + "auxiliary_loss_mlp": 0.01030503, + "balance_loss_clip": 1.026577, + "balance_loss_mlp": 1.02101433, + "epoch": 0.2596044008897974, + "flos": 23623583650560.0, + "grad_norm": 2.0886074259645016, + "language_loss": 0.75527513, + "learning_rate": 3.4721695681941286e-06, + "loss": 0.77742195, + "num_input_tokens_seen": 46207025, + "step": 2159, + "time_per_iteration": 2.719635248184204 + }, + { + "auxiliary_loss_clip": 0.01153507, + "auxiliary_loss_mlp": 0.00511587, + "balance_loss_clip": 0.98345947, + "balance_loss_mlp": 0.0, + "epoch": 0.25972464378043647, + "flos": 13772281628160.0, + "grad_norm": 2.198497454093833, + "language_loss": 0.82211351, + "learning_rate": 3.471642178932845e-06, + "loss": 0.83876443, + "num_input_tokens_seen": 46225670, + "step": 2160, + "time_per_iteration": 2.673125982284546 + }, + { + "auxiliary_loss_clip": 0.01157837, + "auxiliary_loss_mlp": 0.01031008, + "balance_loss_clip": 0.98259646, + "balance_loss_mlp": 1.021698, + "epoch": 0.2598448866710756, + "flos": 19573578391680.0, + "grad_norm": 2.1821276387901682, + "language_loss": 0.89461201, + "learning_rate": 3.471114566424613e-06, + "loss": 0.91650045, + "num_input_tokens_seen": 46244130, + "step": 2161, + "time_per_iteration": 2.712725877761841 + }, + { + "auxiliary_loss_clip": 0.01157833, + "auxiliary_loss_mlp": 0.01040115, + "balance_loss_clip": 0.98661768, + "balance_loss_mlp": 1.0303638, + "epoch": 0.25996512956171464, + "flos": 21653237053440.0, + "grad_norm": 2.1441933264132653, + "language_loss": 0.75514865, + "learning_rate": 3.4705867307494715e-06, + "loss": 0.7771281, + "num_input_tokens_seen": 46263200, + "step": 2162, + "time_per_iteration": 2.745269775390625 + }, + { + "auxiliary_loss_clip": 0.01181335, + "auxiliary_loss_mlp": 0.01041037, + "balance_loss_clip": 1.02137768, + "balance_loss_mlp": 1.03095746, + "epoch": 0.26008537245235375, + "flos": 18223480869120.0, + "grad_norm": 2.074908728644957, + "language_loss": 0.84511262, + "learning_rate": 3.470058671987492e-06, + "loss": 0.86733627, + "num_input_tokens_seen": 46281465, + "step": 2163, + "time_per_iteration": 2.670530080795288 + }, + { + "auxiliary_loss_clip": 0.01186413, + "auxiliary_loss_mlp": 0.0103577, + "balance_loss_clip": 1.02379572, + "balance_loss_mlp": 1.02637064, + "epoch": 0.26020561534299286, + "flos": 24645385843200.0, + "grad_norm": 1.9368850116168872, + "language_loss": 0.8419041, + "learning_rate": 3.4695303902187805e-06, + "loss": 0.86412597, + "num_input_tokens_seen": 46301020, + "step": 2164, + "time_per_iteration": 2.6619889736175537 + }, + { + "auxiliary_loss_clip": 0.01128579, + "auxiliary_loss_mlp": 0.01032226, + "balance_loss_clip": 0.94265902, + "balance_loss_mlp": 1.02204561, + "epoch": 0.2603258582336319, + "flos": 25773662926080.0, + "grad_norm": 4.320747392830471, + "language_loss": 0.78259933, + "learning_rate": 3.469001885523478e-06, + "loss": 0.80420744, + "num_input_tokens_seen": 46321740, + "step": 2165, + "time_per_iteration": 2.745249032974243 + }, + { + "auxiliary_loss_clip": 0.01201775, + "auxiliary_loss_mlp": 0.01030539, + "balance_loss_clip": 1.06007838, + "balance_loss_mlp": 1.02088892, + "epoch": 0.260446101124271, + "flos": 28766314506240.0, + "grad_norm": 1.9479187323026308, + "language_loss": 0.8092584, + "learning_rate": 3.4684731579817568e-06, + "loss": 0.83158153, + "num_input_tokens_seen": 46342730, + "step": 2166, + "time_per_iteration": 2.6426620483398438 + }, + { + "auxiliary_loss_clip": 0.01082706, + "auxiliary_loss_mlp": 0.01034277, + "balance_loss_clip": 0.87081808, + "balance_loss_mlp": 1.02560413, + "epoch": 0.26056634401491013, + "flos": 25666757072640.0, + "grad_norm": 1.803160527673106, + "language_loss": 0.76378655, + "learning_rate": 3.4679442076738247e-06, + "loss": 0.78495634, + "num_input_tokens_seen": 46362445, + "step": 2167, + "time_per_iteration": 2.8550193309783936 + }, + { + "auxiliary_loss_clip": 0.01206855, + "auxiliary_loss_mlp": 0.01032457, + "balance_loss_clip": 1.06366026, + "balance_loss_mlp": 1.02212179, + "epoch": 0.2606865869055492, + "flos": 27052765217280.0, + "grad_norm": 1.9047732451817794, + "language_loss": 0.83193642, + "learning_rate": 3.4674150346799245e-06, + "loss": 0.85432953, + "num_input_tokens_seen": 46382145, + "step": 2168, + "time_per_iteration": 2.6703014373779297 + }, + { + "auxiliary_loss_clip": 0.01158509, + "auxiliary_loss_mlp": 0.010333, + "balance_loss_clip": 0.98751557, + "balance_loss_mlp": 1.02355504, + "epoch": 0.2608068297961883, + "flos": 17712615686400.0, + "grad_norm": 11.38499337327835, + "language_loss": 0.799191, + "learning_rate": 3.4668856390803295e-06, + "loss": 0.82110912, + "num_input_tokens_seen": 46400025, + "step": 2169, + "time_per_iteration": 2.7789015769958496 + }, + { + "auxiliary_loss_clip": 0.01167169, + "auxiliary_loss_mlp": 0.01034749, + "balance_loss_clip": 1.01899099, + "balance_loss_mlp": 1.02558732, + "epoch": 0.2609270726868274, + "flos": 18551632544640.0, + "grad_norm": 2.190548583240468, + "language_loss": 0.89597785, + "learning_rate": 3.4663560209553495e-06, + "loss": 0.917997, + "num_input_tokens_seen": 46418090, + "step": 2170, + "time_per_iteration": 2.629927396774292 + }, + { + "auxiliary_loss_clip": 0.01149092, + "auxiliary_loss_mlp": 0.01039062, + "balance_loss_clip": 0.98412532, + "balance_loss_mlp": 1.02988911, + "epoch": 0.26104731557746647, + "flos": 21835699165440.0, + "grad_norm": 1.9654722875038384, + "language_loss": 0.79266596, + "learning_rate": 3.4658261803853267e-06, + "loss": 0.81454748, + "num_input_tokens_seen": 46436015, + "step": 2171, + "time_per_iteration": 2.817551612854004 + }, + { + "auxiliary_loss_clip": 0.01152451, + "auxiliary_loss_mlp": 0.01028674, + "balance_loss_clip": 0.98480463, + "balance_loss_mlp": 1.0199002, + "epoch": 0.2611675584681056, + "flos": 21689650465920.0, + "grad_norm": 1.9749161630923653, + "language_loss": 0.80554402, + "learning_rate": 3.4652961174506383e-06, + "loss": 0.82735527, + "num_input_tokens_seen": 46455885, + "step": 2172, + "time_per_iteration": 2.7176735401153564 + }, + { + "auxiliary_loss_clip": 0.01082222, + "auxiliary_loss_mlp": 0.01007417, + "balance_loss_clip": 0.99773979, + "balance_loss_mlp": 1.00543845, + "epoch": 0.2612878013587447, + "flos": 71862101389440.0, + "grad_norm": 0.9673900983913511, + "language_loss": 0.58141303, + "learning_rate": 3.464765832231694e-06, + "loss": 0.60230947, + "num_input_tokens_seen": 46510050, + "step": 2173, + "time_per_iteration": 3.241257905960083 + }, + { + "auxiliary_loss_clip": 0.0118279, + "auxiliary_loss_mlp": 0.01029873, + "balance_loss_clip": 1.02645087, + "balance_loss_mlp": 1.02107573, + "epoch": 0.26140804424938374, + "flos": 20227511445120.0, + "grad_norm": 2.5549617708307952, + "language_loss": 0.70648891, + "learning_rate": 3.4642353248089373e-06, + "loss": 0.72861552, + "num_input_tokens_seen": 46528810, + "step": 2174, + "time_per_iteration": 3.592536211013794 + }, + { + "auxiliary_loss_clip": 0.01151002, + "auxiliary_loss_mlp": 0.01034806, + "balance_loss_clip": 0.98455471, + "balance_loss_mlp": 1.02436924, + "epoch": 0.26152828714002285, + "flos": 25557085872000.0, + "grad_norm": 2.15005889574836, + "language_loss": 0.80074215, + "learning_rate": 3.463704595262846e-06, + "loss": 0.82260025, + "num_input_tokens_seen": 46549690, + "step": 2175, + "time_per_iteration": 3.689110517501831 + }, + { + "auxiliary_loss_clip": 0.01128065, + "auxiliary_loss_mlp": 0.0103372, + "balance_loss_clip": 0.94600308, + "balance_loss_mlp": 1.0244875, + "epoch": 0.26164853003066196, + "flos": 25446516831360.0, + "grad_norm": 10.268888131583166, + "language_loss": 0.70202607, + "learning_rate": 3.463173643673931e-06, + "loss": 0.72364396, + "num_input_tokens_seen": 46572215, + "step": 2176, + "time_per_iteration": 2.804927110671997 + }, + { + "auxiliary_loss_clip": 0.01092853, + "auxiliary_loss_mlp": 0.01016809, + "balance_loss_clip": 0.99915612, + "balance_loss_mlp": 1.01474714, + "epoch": 0.261768772921301, + "flos": 53944580568960.0, + "grad_norm": 0.9200835564726967, + "language_loss": 0.63445497, + "learning_rate": 3.4626424701227387e-06, + "loss": 0.65555155, + "num_input_tokens_seen": 46627275, + "step": 2177, + "time_per_iteration": 4.061023712158203 + }, + { + "auxiliary_loss_clip": 0.01111737, + "auxiliary_loss_mlp": 0.01006168, + "balance_loss_clip": 1.03671646, + "balance_loss_mlp": 1.00396276, + "epoch": 0.26188901581194013, + "flos": 70687606481280.0, + "grad_norm": 0.8331361238836019, + "language_loss": 0.55836022, + "learning_rate": 3.4621110746898452e-06, + "loss": 0.5795393, + "num_input_tokens_seen": 46695135, + "step": 2178, + "time_per_iteration": 3.307249069213867 + }, + { + "auxiliary_loss_clip": 0.01182178, + "auxiliary_loss_mlp": 0.01030302, + "balance_loss_clip": 1.02436435, + "balance_loss_mlp": 1.02111137, + "epoch": 0.2620092587025792, + "flos": 21069580959360.0, + "grad_norm": 1.8144364530191623, + "language_loss": 0.74777228, + "learning_rate": 3.4615794574558654e-06, + "loss": 0.76989704, + "num_input_tokens_seen": 46714145, + "step": 2179, + "time_per_iteration": 2.6012353897094727 + }, + { + "auxiliary_loss_clip": 0.01154202, + "auxiliary_loss_mlp": 0.01027216, + "balance_loss_clip": 0.98395169, + "balance_loss_mlp": 1.01826894, + "epoch": 0.2621295015932183, + "flos": 18369601395840.0, + "grad_norm": 3.750058020903763, + "language_loss": 0.83880687, + "learning_rate": 3.4610476185014436e-06, + "loss": 0.86062098, + "num_input_tokens_seen": 46731405, + "step": 2180, + "time_per_iteration": 3.6236138343811035 + }, + { + "auxiliary_loss_clip": 0.01204908, + "auxiliary_loss_mlp": 0.01036341, + "balance_loss_clip": 1.0612551, + "balance_loss_mlp": 1.02586854, + "epoch": 0.2622497444838574, + "flos": 23659997063040.0, + "grad_norm": 18.61770078106629, + "language_loss": 0.79417169, + "learning_rate": 3.4605155579072597e-06, + "loss": 0.81658417, + "num_input_tokens_seen": 46751260, + "step": 2181, + "time_per_iteration": 2.650341272354126 + }, + { + "auxiliary_loss_clip": 0.01099053, + "auxiliary_loss_mlp": 0.01033115, + "balance_loss_clip": 0.90429664, + "balance_loss_mlp": 1.02432346, + "epoch": 0.26236998737449646, + "flos": 22123810154880.0, + "grad_norm": 1.7245758339167632, + "language_loss": 0.70922923, + "learning_rate": 3.459983275754027e-06, + "loss": 0.73055089, + "num_input_tokens_seen": 46770155, + "step": 2182, + "time_per_iteration": 2.753000497817993 + }, + { + "auxiliary_loss_clip": 0.01204872, + "auxiliary_loss_mlp": 0.01036687, + "balance_loss_clip": 1.06330836, + "balance_loss_mlp": 1.02782416, + "epoch": 0.26249023026513557, + "flos": 17895185539200.0, + "grad_norm": 5.014197726590416, + "language_loss": 0.79466069, + "learning_rate": 3.4594507721224918e-06, + "loss": 0.81707633, + "num_input_tokens_seen": 46788805, + "step": 2183, + "time_per_iteration": 2.6359288692474365 + }, + { + "auxiliary_loss_clip": 0.01157846, + "auxiliary_loss_mlp": 0.01038932, + "balance_loss_clip": 0.98532772, + "balance_loss_mlp": 1.03076673, + "epoch": 0.2626104731557747, + "flos": 18332936588160.0, + "grad_norm": 2.579871703037666, + "language_loss": 0.81626493, + "learning_rate": 3.4589180470934353e-06, + "loss": 0.8382327, + "num_input_tokens_seen": 46808670, + "step": 2184, + "time_per_iteration": 2.74722957611084 + }, + { + "auxiliary_loss_clip": 0.01181194, + "auxiliary_loss_mlp": 0.01036848, + "balance_loss_clip": 1.01800799, + "balance_loss_mlp": 1.02729905, + "epoch": 0.26273071604641374, + "flos": 19317714837120.0, + "grad_norm": 2.053672855210778, + "language_loss": 0.76508212, + "learning_rate": 3.4583851007476713e-06, + "loss": 0.7872625, + "num_input_tokens_seen": 46827140, + "step": 2185, + "time_per_iteration": 2.7955334186553955 + }, + { + "auxiliary_loss_clip": 0.01131592, + "auxiliary_loss_mlp": 0.01040949, + "balance_loss_clip": 0.94510794, + "balance_loss_mlp": 1.0308876, + "epoch": 0.26285095893705285, + "flos": 18327477720960.0, + "grad_norm": 6.497873803292735, + "language_loss": 0.68625319, + "learning_rate": 3.4578519331660464e-06, + "loss": 0.70797849, + "num_input_tokens_seen": 46844135, + "step": 2186, + "time_per_iteration": 2.776902198791504 + }, + { + "auxiliary_loss_clip": 0.0117485, + "auxiliary_loss_mlp": 0.01036419, + "balance_loss_clip": 1.02307439, + "balance_loss_mlp": 1.02800846, + "epoch": 0.26297120182769196, + "flos": 20193827466240.0, + "grad_norm": 1.9546670542298092, + "language_loss": 0.81739157, + "learning_rate": 3.4573185444294426e-06, + "loss": 0.8395043, + "num_input_tokens_seen": 46862500, + "step": 2187, + "time_per_iteration": 2.6486754417419434 + }, + { + "auxiliary_loss_clip": 0.01155468, + "auxiliary_loss_mlp": 0.00511803, + "balance_loss_clip": 0.98460156, + "balance_loss_mlp": 0.0, + "epoch": 0.263091444718331, + "flos": 22418421505920.0, + "grad_norm": 2.6734940430550314, + "language_loss": 0.78487623, + "learning_rate": 3.456784934618774e-06, + "loss": 0.80154896, + "num_input_tokens_seen": 46883665, + "step": 2188, + "time_per_iteration": 2.7143502235412598 + }, + { + "auxiliary_loss_clip": 0.01153093, + "auxiliary_loss_mlp": 0.01035137, + "balance_loss_clip": 0.98264527, + "balance_loss_mlp": 1.02563596, + "epoch": 0.2632116876089701, + "flos": 19024827338880.0, + "grad_norm": 2.0519987678611953, + "language_loss": 0.79891896, + "learning_rate": 3.4562511038149897e-06, + "loss": 0.82080126, + "num_input_tokens_seen": 46899160, + "step": 2189, + "time_per_iteration": 2.7850422859191895 + }, + { + "auxiliary_loss_clip": 0.01020372, + "auxiliary_loss_mlp": 0.01051031, + "balance_loss_clip": 0.88410819, + "balance_loss_mlp": 1.04901683, + "epoch": 0.26333193049960923, + "flos": 67308054531840.0, + "grad_norm": 0.8656316510642768, + "language_loss": 0.57705325, + "learning_rate": 3.4557170520990705e-06, + "loss": 0.59776729, + "num_input_tokens_seen": 46959835, + "step": 2190, + "time_per_iteration": 3.3091869354248047 + }, + { + "auxiliary_loss_clip": 0.01176465, + "auxiliary_loss_mlp": 0.01038858, + "balance_loss_clip": 1.0232687, + "balance_loss_mlp": 1.03005993, + "epoch": 0.2634521733902483, + "flos": 25048806468480.0, + "grad_norm": 1.750045575555868, + "language_loss": 0.86439544, + "learning_rate": 3.4551827795520324e-06, + "loss": 0.88654876, + "num_input_tokens_seen": 46982720, + "step": 2191, + "time_per_iteration": 2.6635072231292725 + }, + { + "auxiliary_loss_clip": 0.01179017, + "auxiliary_loss_mlp": 0.01038653, + "balance_loss_clip": 1.02219343, + "balance_loss_mlp": 1.02990961, + "epoch": 0.2635724162808874, + "flos": 20594985534720.0, + "grad_norm": 1.802570761794364, + "language_loss": 0.84879911, + "learning_rate": 3.4546482862549226e-06, + "loss": 0.87097585, + "num_input_tokens_seen": 47003035, + "step": 2192, + "time_per_iteration": 2.675647497177124 + }, + { + "auxiliary_loss_clip": 0.0112721, + "auxiliary_loss_mlp": 0.01033224, + "balance_loss_clip": 0.9453423, + "balance_loss_mlp": 1.02410471, + "epoch": 0.2636926591715265, + "flos": 19244636616960.0, + "grad_norm": 2.3437719160098602, + "language_loss": 0.78216672, + "learning_rate": 3.4541135722888253e-06, + "loss": 0.80377102, + "num_input_tokens_seen": 47019625, + "step": 2193, + "time_per_iteration": 2.6677141189575195 + }, + { + "auxiliary_loss_clip": 0.01202397, + "auxiliary_loss_mlp": 0.01033225, + "balance_loss_clip": 1.06000626, + "balance_loss_mlp": 1.02396786, + "epoch": 0.26381290206216557, + "flos": 28804882734720.0, + "grad_norm": 1.9380435611331883, + "language_loss": 0.80309355, + "learning_rate": 3.453578637734854e-06, + "loss": 0.82544976, + "num_input_tokens_seen": 47040815, + "step": 2194, + "time_per_iteration": 2.7047760486602783 + }, + { + "auxiliary_loss_clip": 0.01206804, + "auxiliary_loss_mlp": 0.01032259, + "balance_loss_clip": 1.06522894, + "balance_loss_mlp": 1.02288914, + "epoch": 0.2639331449528047, + "flos": 25008909436800.0, + "grad_norm": 1.9886802274969864, + "language_loss": 0.78744614, + "learning_rate": 3.4530434826741605e-06, + "loss": 0.80983675, + "num_input_tokens_seen": 47061755, + "step": 2195, + "time_per_iteration": 2.656754732131958 + }, + { + "auxiliary_loss_clip": 0.0115288, + "auxiliary_loss_mlp": 0.01032708, + "balance_loss_clip": 0.98486835, + "balance_loss_mlp": 1.02371979, + "epoch": 0.26405338784344373, + "flos": 46535775465600.0, + "grad_norm": 2.091345369129617, + "language_loss": 0.68972135, + "learning_rate": 3.452508107187926e-06, + "loss": 0.71157724, + "num_input_tokens_seen": 47085130, + "step": 2196, + "time_per_iteration": 2.8857345581054688 + }, + { + "auxiliary_loss_clip": 0.0108469, + "auxiliary_loss_mlp": 0.01032295, + "balance_loss_clip": 0.86753297, + "balance_loss_mlp": 1.02306247, + "epoch": 0.26417363073408284, + "flos": 21179467641600.0, + "grad_norm": 2.0370437738704683, + "language_loss": 0.77077997, + "learning_rate": 3.451972511357366e-06, + "loss": 0.79194981, + "num_input_tokens_seen": 47104675, + "step": 2197, + "time_per_iteration": 2.7875659465789795 + }, + { + "auxiliary_loss_clip": 0.01179154, + "auxiliary_loss_mlp": 0.01027786, + "balance_loss_clip": 1.02616036, + "balance_loss_mlp": 1.01874948, + "epoch": 0.26429387362472195, + "flos": 22674751937280.0, + "grad_norm": 1.7527093377483143, + "language_loss": 0.84950519, + "learning_rate": 3.45143669526373e-06, + "loss": 0.87157464, + "num_input_tokens_seen": 47124435, + "step": 2198, + "time_per_iteration": 2.692382335662842 + }, + { + "auxiliary_loss_clip": 0.010688, + "auxiliary_loss_mlp": 0.01001371, + "balance_loss_clip": 0.95735121, + "balance_loss_mlp": 0.99921364, + "epoch": 0.264414116515361, + "flos": 67180534272000.0, + "grad_norm": 0.7890336305717001, + "language_loss": 0.63148868, + "learning_rate": 3.450900658988302e-06, + "loss": 0.65219033, + "num_input_tokens_seen": 47185985, + "step": 2199, + "time_per_iteration": 4.15111780166626 + }, + { + "auxiliary_loss_clip": 0.0114818, + "auxiliary_loss_mlp": 0.01031426, + "balance_loss_clip": 0.98387688, + "balance_loss_mlp": 1.02223516, + "epoch": 0.2645343594060001, + "flos": 25664709997440.0, + "grad_norm": 2.059719472347569, + "language_loss": 0.77636385, + "learning_rate": 3.450364402612397e-06, + "loss": 0.79815996, + "num_input_tokens_seen": 47203140, + "step": 2200, + "time_per_iteration": 2.676957130432129 + }, + { + "auxiliary_loss_clip": 0.01152245, + "auxiliary_loss_mlp": 0.01028181, + "balance_loss_clip": 0.98407924, + "balance_loss_mlp": 1.01866198, + "epoch": 0.26465460229663923, + "flos": 22491822948480.0, + "grad_norm": 2.0599982181843335, + "language_loss": 0.8377763, + "learning_rate": 3.449827926217366e-06, + "loss": 0.85958052, + "num_input_tokens_seen": 47222575, + "step": 2201, + "time_per_iteration": 3.687969207763672 + }, + { + "auxiliary_loss_clip": 0.01156364, + "auxiliary_loss_mlp": 0.01030623, + "balance_loss_clip": 0.9799847, + "balance_loss_mlp": 1.0209254, + "epoch": 0.2647748451872783, + "flos": 29388036038400.0, + "grad_norm": 2.002452237897907, + "language_loss": 0.80336082, + "learning_rate": 3.449291229884591e-06, + "loss": 0.82523072, + "num_input_tokens_seen": 47243815, + "step": 2202, + "time_per_iteration": 2.8267457485198975 + }, + { + "auxiliary_loss_clip": 0.01135638, + "auxiliary_loss_mlp": 0.01028273, + "balance_loss_clip": 0.94471335, + "balance_loss_mlp": 1.01911807, + "epoch": 0.2648950880779174, + "flos": 26797799502720.0, + "grad_norm": 1.999793375954014, + "language_loss": 0.86556923, + "learning_rate": 3.4487543136954887e-06, + "loss": 0.88720834, + "num_input_tokens_seen": 47263435, + "step": 2203, + "time_per_iteration": 3.816973924636841 + }, + { + "auxiliary_loss_clip": 0.0113231, + "auxiliary_loss_mlp": 0.01029105, + "balance_loss_clip": 0.94758189, + "balance_loss_mlp": 1.02053428, + "epoch": 0.2650153309685565, + "flos": 28841008838400.0, + "grad_norm": 1.7990646075973906, + "language_loss": 0.90897202, + "learning_rate": 3.448217177731509e-06, + "loss": 0.93058622, + "num_input_tokens_seen": 47283920, + "step": 2204, + "time_per_iteration": 2.8590829372406006 + }, + { + "auxiliary_loss_clip": 0.01152847, + "auxiliary_loss_mlp": 0.01034936, + "balance_loss_clip": 0.98747694, + "balance_loss_mlp": 1.0260911, + "epoch": 0.26513557385919556, + "flos": 20303247271680.0, + "grad_norm": 2.3479889456878964, + "language_loss": 0.77787256, + "learning_rate": 3.4476798220741348e-06, + "loss": 0.79975033, + "num_input_tokens_seen": 47302800, + "step": 2205, + "time_per_iteration": 3.665888547897339 + }, + { + "auxiliary_loss_clip": 0.0120492, + "auxiliary_loss_mlp": 0.01029836, + "balance_loss_clip": 1.06471038, + "balance_loss_mlp": 1.02119315, + "epoch": 0.26525581674983467, + "flos": 17676274101120.0, + "grad_norm": 3.28564428306453, + "language_loss": 0.78456473, + "learning_rate": 3.4471422468048826e-06, + "loss": 0.8069123, + "num_input_tokens_seen": 47321525, + "step": 2206, + "time_per_iteration": 2.6306352615356445 + }, + { + "auxiliary_loss_clip": 0.01174752, + "auxiliary_loss_mlp": 0.0103363, + "balance_loss_clip": 1.0242964, + "balance_loss_mlp": 1.02395642, + "epoch": 0.2653760596404738, + "flos": 26833746038400.0, + "grad_norm": 2.541017637847838, + "language_loss": 0.72656214, + "learning_rate": 3.4466044520053022e-06, + "loss": 0.74864596, + "num_input_tokens_seen": 47340530, + "step": 2207, + "time_per_iteration": 2.767115831375122 + }, + { + "auxiliary_loss_clip": 0.0114192, + "auxiliary_loss_mlp": 0.01027898, + "balance_loss_clip": 0.97958732, + "balance_loss_mlp": 1.01908302, + "epoch": 0.26549630253111284, + "flos": 22782160581120.0, + "grad_norm": 2.6931972093854237, + "language_loss": 0.604074, + "learning_rate": 3.446066437756977e-06, + "loss": 0.62577212, + "num_input_tokens_seen": 47359735, + "step": 2208, + "time_per_iteration": 2.7630300521850586 + }, + { + "auxiliary_loss_clip": 0.01152215, + "auxiliary_loss_mlp": 0.01035807, + "balance_loss_clip": 0.9829582, + "balance_loss_mlp": 1.02652109, + "epoch": 0.26561654542175195, + "flos": 23550002640000.0, + "grad_norm": 2.369685728659998, + "language_loss": 0.75168729, + "learning_rate": 3.4455282041415224e-06, + "loss": 0.77356756, + "num_input_tokens_seen": 47378945, + "step": 2209, + "time_per_iteration": 2.741224527359009 + }, + { + "auxiliary_loss_clip": 0.01133427, + "auxiliary_loss_mlp": 0.01037332, + "balance_loss_clip": 0.94542319, + "balance_loss_mlp": 1.02788424, + "epoch": 0.265736788312391, + "flos": 26906680604160.0, + "grad_norm": 2.4645617093109564, + "language_loss": 0.87273175, + "learning_rate": 3.4449897512405894e-06, + "loss": 0.89443934, + "num_input_tokens_seen": 47398095, + "step": 2210, + "time_per_iteration": 2.7321481704711914 + }, + { + "auxiliary_loss_clip": 0.01077278, + "auxiliary_loss_mlp": 0.0051067, + "balance_loss_clip": 0.86549479, + "balance_loss_mlp": 0.0, + "epoch": 0.2658570312030301, + "flos": 23477139901440.0, + "grad_norm": 2.2865267330980195, + "language_loss": 0.75053644, + "learning_rate": 3.444451079135859e-06, + "loss": 0.76641589, + "num_input_tokens_seen": 47417605, + "step": 2211, + "time_per_iteration": 2.752371311187744 + }, + { + "auxiliary_loss_clip": 0.01097336, + "auxiliary_loss_mlp": 0.00511586, + "balance_loss_clip": 0.90107322, + "balance_loss_mlp": 0.0, + "epoch": 0.2659772740936692, + "flos": 21866402315520.0, + "grad_norm": 2.2152857781005832, + "language_loss": 0.73895538, + "learning_rate": 3.4439121879090493e-06, + "loss": 0.75504458, + "num_input_tokens_seen": 47435385, + "step": 2212, + "time_per_iteration": 2.689359664916992 + }, + { + "auxiliary_loss_clip": 0.01155927, + "auxiliary_loss_mlp": 0.01034782, + "balance_loss_clip": 0.98213315, + "balance_loss_mlp": 1.02528119, + "epoch": 0.2660975169843083, + "flos": 19793100360960.0, + "grad_norm": 2.2982553676610493, + "language_loss": 0.83318686, + "learning_rate": 3.4433730776419082e-06, + "loss": 0.85509396, + "num_input_tokens_seen": 47454310, + "step": 2213, + "time_per_iteration": 2.5832486152648926 + }, + { + "auxiliary_loss_clip": 0.01179179, + "auxiliary_loss_mlp": 0.00511764, + "balance_loss_clip": 1.01953948, + "balance_loss_mlp": 0.0, + "epoch": 0.2662177598749474, + "flos": 29018981750400.0, + "grad_norm": 3.8872498280009484, + "language_loss": 0.8058095, + "learning_rate": 3.4428337484162183e-06, + "loss": 0.82271898, + "num_input_tokens_seen": 47475120, + "step": 2214, + "time_per_iteration": 3.138831853866577 + }, + { + "auxiliary_loss_clip": 0.01152558, + "auxiliary_loss_mlp": 0.01031455, + "balance_loss_clip": 0.98334867, + "balance_loss_mlp": 1.02241898, + "epoch": 0.2663380027655865, + "flos": 21762549118080.0, + "grad_norm": 1.9637688188179874, + "language_loss": 0.84487593, + "learning_rate": 3.442294200313797e-06, + "loss": 0.86671609, + "num_input_tokens_seen": 47493150, + "step": 2215, + "time_per_iteration": 2.9767532348632812 + }, + { + "auxiliary_loss_clip": 0.01103523, + "auxiliary_loss_mlp": 0.01002865, + "balance_loss_clip": 1.02909112, + "balance_loss_mlp": 1.00081491, + "epoch": 0.26645824565622556, + "flos": 66980333819520.0, + "grad_norm": 0.7734471369104697, + "language_loss": 0.52671927, + "learning_rate": 3.4417544334164916e-06, + "loss": 0.54778314, + "num_input_tokens_seen": 47557295, + "step": 2216, + "time_per_iteration": 3.304392099380493 + }, + { + "auxiliary_loss_clip": 0.01130078, + "auxiliary_loss_mlp": 0.01032845, + "balance_loss_clip": 0.94589239, + "balance_loss_mlp": 1.02379131, + "epoch": 0.26657848854686467, + "flos": 25264198373760.0, + "grad_norm": 1.8060552586276555, + "language_loss": 0.7727648, + "learning_rate": 3.4412144478061854e-06, + "loss": 0.79439402, + "num_input_tokens_seen": 47579705, + "step": 2217, + "time_per_iteration": 2.781939744949341 + }, + { + "auxiliary_loss_clip": 0.01033921, + "auxiliary_loss_mlp": 0.01030324, + "balance_loss_clip": 0.79055285, + "balance_loss_mlp": 1.02013206, + "epoch": 0.2666987314375038, + "flos": 23696769611520.0, + "grad_norm": 1.8498661110793972, + "language_loss": 0.75311744, + "learning_rate": 3.4406742435647925e-06, + "loss": 0.77375984, + "num_input_tokens_seen": 47599770, + "step": 2218, + "time_per_iteration": 2.837575912475586 + }, + { + "auxiliary_loss_clip": 0.01179341, + "auxiliary_loss_mlp": 0.01031986, + "balance_loss_clip": 1.02481902, + "balance_loss_mlp": 1.02251482, + "epoch": 0.26681897432814283, + "flos": 27048958375680.0, + "grad_norm": 1.865432699921534, + "language_loss": 0.78837818, + "learning_rate": 3.440133820774263e-06, + "loss": 0.81049144, + "num_input_tokens_seen": 47619580, + "step": 2219, + "time_per_iteration": 2.717782497406006 + }, + { + "auxiliary_loss_clip": 0.01159284, + "auxiliary_loss_mlp": 0.01037153, + "balance_loss_clip": 0.98432255, + "balance_loss_mlp": 1.0268594, + "epoch": 0.26693921721878194, + "flos": 28985944216320.0, + "grad_norm": 2.433480918902928, + "language_loss": 0.82010448, + "learning_rate": 3.439593179516578e-06, + "loss": 0.84206879, + "num_input_tokens_seen": 47639490, + "step": 2220, + "time_per_iteration": 2.7213480472564697 + }, + { + "auxiliary_loss_clip": 0.01159542, + "auxiliary_loss_mlp": 0.01030961, + "balance_loss_clip": 0.98535371, + "balance_loss_mlp": 1.0218538, + "epoch": 0.26705946010942105, + "flos": 21507834798720.0, + "grad_norm": 2.602156533092866, + "language_loss": 0.80584705, + "learning_rate": 3.4390523198737524e-06, + "loss": 0.82775205, + "num_input_tokens_seen": 47658650, + "step": 2221, + "time_per_iteration": 2.7032878398895264 + }, + { + "auxiliary_loss_clip": 0.0120256, + "auxiliary_loss_mlp": 0.00510838, + "balance_loss_clip": 1.06099558, + "balance_loss_mlp": 0.0, + "epoch": 0.2671797030000601, + "flos": 21471277731840.0, + "grad_norm": 2.102426837769671, + "language_loss": 0.73977619, + "learning_rate": 3.4385112419278333e-06, + "loss": 0.7569102, + "num_input_tokens_seen": 47679875, + "step": 2222, + "time_per_iteration": 2.6243937015533447 + }, + { + "auxiliary_loss_clip": 0.01081311, + "auxiliary_loss_mlp": 0.01004042, + "balance_loss_clip": 0.98855913, + "balance_loss_mlp": 1.00211084, + "epoch": 0.2672999458906992, + "flos": 64189929767040.0, + "grad_norm": 0.7885900031024409, + "language_loss": 0.64838207, + "learning_rate": 3.4379699457609033e-06, + "loss": 0.66923559, + "num_input_tokens_seen": 47737700, + "step": 2223, + "time_per_iteration": 3.0522267818450928 + }, + { + "auxiliary_loss_clip": 0.01143944, + "auxiliary_loss_mlp": 0.01030743, + "balance_loss_clip": 0.97897291, + "balance_loss_mlp": 1.02171326, + "epoch": 0.26742018878133833, + "flos": 16909042573440.0, + "grad_norm": 1.761287569231845, + "language_loss": 0.89991748, + "learning_rate": 3.4374284314550755e-06, + "loss": 0.92166436, + "num_input_tokens_seen": 47756740, + "step": 2224, + "time_per_iteration": 2.6983377933502197 + }, + { + "auxiliary_loss_clip": 0.01202494, + "auxiliary_loss_mlp": 0.01027855, + "balance_loss_clip": 1.06134081, + "balance_loss_mlp": 1.01872897, + "epoch": 0.2675404316719774, + "flos": 20667560964480.0, + "grad_norm": 1.9312574417762356, + "language_loss": 0.80993772, + "learning_rate": 3.436886699092498e-06, + "loss": 0.83224124, + "num_input_tokens_seen": 47775255, + "step": 2225, + "time_per_iteration": 2.6263208389282227 + }, + { + "auxiliary_loss_clip": 0.01205605, + "auxiliary_loss_mlp": 0.01030296, + "balance_loss_clip": 1.0615077, + "balance_loss_mlp": 1.02045488, + "epoch": 0.2676606745626165, + "flos": 17485013157120.0, + "grad_norm": 2.576877396015363, + "language_loss": 0.71290743, + "learning_rate": 3.4363447487553502e-06, + "loss": 0.73526645, + "num_input_tokens_seen": 47788570, + "step": 2226, + "time_per_iteration": 3.4808382987976074 + }, + { + "auxiliary_loss_clip": 0.0115073, + "auxiliary_loss_mlp": 0.01037784, + "balance_loss_clip": 0.9841063, + "balance_loss_mlp": 1.02882504, + "epoch": 0.26778091745325555, + "flos": 27852675143040.0, + "grad_norm": 2.335744808007983, + "language_loss": 0.77743638, + "learning_rate": 3.4358025805258455e-06, + "loss": 0.79932153, + "num_input_tokens_seen": 47808275, + "step": 2227, + "time_per_iteration": 3.6548776626586914 + }, + { + "auxiliary_loss_clip": 0.01107468, + "auxiliary_loss_mlp": 0.01029529, + "balance_loss_clip": 0.90323985, + "balance_loss_mlp": 1.02003992, + "epoch": 0.26790116034389466, + "flos": 20955995176320.0, + "grad_norm": 1.9209529111574208, + "language_loss": 0.83113241, + "learning_rate": 3.435260194486232e-06, + "loss": 0.85250241, + "num_input_tokens_seen": 47826245, + "step": 2228, + "time_per_iteration": 2.74114727973938 + }, + { + "auxiliary_loss_clip": 0.01155946, + "auxiliary_loss_mlp": 0.01030769, + "balance_loss_clip": 0.9840436, + "balance_loss_mlp": 1.02127385, + "epoch": 0.2680214032345338, + "flos": 18040659621120.0, + "grad_norm": 3.171236568714889, + "language_loss": 0.82138801, + "learning_rate": 3.4347175907187875e-06, + "loss": 0.84325516, + "num_input_tokens_seen": 47843235, + "step": 2229, + "time_per_iteration": 3.5493695735931396 + }, + { + "auxiliary_loss_clip": 0.01175406, + "auxiliary_loss_mlp": 0.01034707, + "balance_loss_clip": 1.02154553, + "balance_loss_mlp": 1.02618384, + "epoch": 0.26814164612517283, + "flos": 22419427086720.0, + "grad_norm": 1.8475115070658958, + "language_loss": 0.87848878, + "learning_rate": 3.4341747693058254e-06, + "loss": 0.90058994, + "num_input_tokens_seen": 47861710, + "step": 2230, + "time_per_iteration": 2.707658052444458 + }, + { + "auxiliary_loss_clip": 0.01029122, + "auxiliary_loss_mlp": 0.01032662, + "balance_loss_clip": 0.78894097, + "balance_loss_mlp": 1.0235188, + "epoch": 0.26826188901581194, + "flos": 35627371159680.0, + "grad_norm": 1.9502886267900994, + "language_loss": 0.77000511, + "learning_rate": 3.4336317303296916e-06, + "loss": 0.79062295, + "num_input_tokens_seen": 47882685, + "step": 2231, + "time_per_iteration": 2.98057222366333 + }, + { + "auxiliary_loss_clip": 0.01173597, + "auxiliary_loss_mlp": 0.0102838, + "balance_loss_clip": 1.02155924, + "balance_loss_mlp": 1.01929069, + "epoch": 0.26838213190645105, + "flos": 17639788861440.0, + "grad_norm": 2.256324614865542, + "language_loss": 0.75081956, + "learning_rate": 3.4330884738727635e-06, + "loss": 0.77283937, + "num_input_tokens_seen": 47900860, + "step": 2232, + "time_per_iteration": 3.564699411392212 + }, + { + "auxiliary_loss_clip": 0.01100095, + "auxiliary_loss_mlp": 0.01028107, + "balance_loss_clip": 0.90604568, + "balance_loss_mlp": 1.01892757, + "epoch": 0.2685023747970901, + "flos": 22674823764480.0, + "grad_norm": 1.9432818717626839, + "language_loss": 0.70515764, + "learning_rate": 3.4325450000174535e-06, + "loss": 0.7264396, + "num_input_tokens_seen": 47917500, + "step": 2233, + "time_per_iteration": 2.7669858932495117 + }, + { + "auxiliary_loss_clip": 0.01098288, + "auxiliary_loss_mlp": 0.01029749, + "balance_loss_clip": 0.90468007, + "balance_loss_mlp": 1.02022386, + "epoch": 0.2686226176877292, + "flos": 20120533764480.0, + "grad_norm": 1.837593481418025, + "language_loss": 0.74272501, + "learning_rate": 3.4320013088462067e-06, + "loss": 0.76400542, + "num_input_tokens_seen": 47934860, + "step": 2234, + "time_per_iteration": 2.76301646232605 + }, + { + "auxiliary_loss_clip": 0.01136981, + "auxiliary_loss_mlp": 0.01030573, + "balance_loss_clip": 0.94557559, + "balance_loss_mlp": 1.02155483, + "epoch": 0.2687428605783683, + "flos": 21872040750720.0, + "grad_norm": 8.94865431323795, + "language_loss": 0.81630683, + "learning_rate": 3.431457400441499e-06, + "loss": 0.83798242, + "num_input_tokens_seen": 47955255, + "step": 2235, + "time_per_iteration": 2.738241672515869 + }, + { + "auxiliary_loss_clip": 0.00991914, + "auxiliary_loss_mlp": 0.01004791, + "balance_loss_clip": 0.84277999, + "balance_loss_mlp": 1.00290799, + "epoch": 0.2688631034690074, + "flos": 69943320766080.0, + "grad_norm": 0.9123550189862814, + "language_loss": 0.60874164, + "learning_rate": 3.4309132748858424e-06, + "loss": 0.62870872, + "num_input_tokens_seen": 48016245, + "step": 2236, + "time_per_iteration": 3.3783745765686035 + }, + { + "auxiliary_loss_clip": 0.01178959, + "auxiliary_loss_mlp": 0.01029, + "balance_loss_clip": 1.02714884, + "balance_loss_mlp": 1.02128136, + "epoch": 0.2689833463596465, + "flos": 22856639431680.0, + "grad_norm": 1.8483548875877809, + "language_loss": 0.83570063, + "learning_rate": 3.430368932261779e-06, + "loss": 0.85778022, + "num_input_tokens_seen": 48036600, + "step": 2237, + "time_per_iteration": 2.8426451683044434 + }, + { + "auxiliary_loss_clip": 0.01152619, + "auxiliary_loss_mlp": 0.01033992, + "balance_loss_clip": 0.98353428, + "balance_loss_mlp": 1.0253315, + "epoch": 0.2691035892502856, + "flos": 17200242132480.0, + "grad_norm": 2.070477616333361, + "language_loss": 0.75034672, + "learning_rate": 3.429824372651886e-06, + "loss": 0.7722128, + "num_input_tokens_seen": 48054750, + "step": 2238, + "time_per_iteration": 2.6795408725738525 + }, + { + "auxiliary_loss_clip": 0.01113023, + "auxiliary_loss_mlp": 0.01027886, + "balance_loss_clip": 0.90863544, + "balance_loss_mlp": 1.01809275, + "epoch": 0.26922383214092466, + "flos": 17747484814080.0, + "grad_norm": 2.294506085029805, + "language_loss": 0.83100814, + "learning_rate": 3.4292795961387732e-06, + "loss": 0.85241723, + "num_input_tokens_seen": 48072650, + "step": 2239, + "time_per_iteration": 2.7510597705841064 + }, + { + "auxiliary_loss_clip": 0.01201093, + "auxiliary_loss_mlp": 0.01027404, + "balance_loss_clip": 1.0602715, + "balance_loss_mlp": 1.01870131, + "epoch": 0.26934407503156377, + "flos": 16173376122240.0, + "grad_norm": 3.6273136617178, + "language_loss": 0.87656021, + "learning_rate": 3.4287346028050818e-06, + "loss": 0.8988452, + "num_input_tokens_seen": 48088720, + "step": 2240, + "time_per_iteration": 2.649033546447754 + }, + { + "auxiliary_loss_clip": 0.01152979, + "auxiliary_loss_mlp": 0.01031452, + "balance_loss_clip": 0.98375273, + "balance_loss_mlp": 1.02246368, + "epoch": 0.2694643179222028, + "flos": 23732895715200.0, + "grad_norm": 2.0054169195576717, + "language_loss": 0.79790843, + "learning_rate": 3.4281893927334866e-06, + "loss": 0.81975281, + "num_input_tokens_seen": 48108630, + "step": 2241, + "time_per_iteration": 2.7237491607666016 + }, + { + "auxiliary_loss_clip": 0.01178058, + "auxiliary_loss_mlp": 0.01031377, + "balance_loss_clip": 1.02351582, + "balance_loss_mlp": 1.02255583, + "epoch": 0.26958456081284193, + "flos": 24718140840960.0, + "grad_norm": 1.9878561555792462, + "language_loss": 0.74965686, + "learning_rate": 3.4276439660066963e-06, + "loss": 0.77175122, + "num_input_tokens_seen": 48128330, + "step": 2242, + "time_per_iteration": 2.9426894187927246 + }, + { + "auxiliary_loss_clip": 0.01201023, + "auxiliary_loss_mlp": 0.01032145, + "balance_loss_clip": 1.06150138, + "balance_loss_mlp": 1.02329385, + "epoch": 0.26970480370348104, + "flos": 18112588606080.0, + "grad_norm": 2.0621285636045767, + "language_loss": 0.84046286, + "learning_rate": 3.427098322707452e-06, + "loss": 0.86279458, + "num_input_tokens_seen": 48144295, + "step": 2243, + "time_per_iteration": 2.7251758575439453 + }, + { + "auxiliary_loss_clip": 0.01178056, + "auxiliary_loss_mlp": 0.01036868, + "balance_loss_clip": 1.02567172, + "balance_loss_mlp": 1.02740836, + "epoch": 0.2698250465941201, + "flos": 10816546250880.0, + "grad_norm": 2.9638820967689403, + "language_loss": 0.89817065, + "learning_rate": 3.426552462918526e-06, + "loss": 0.92031991, + "num_input_tokens_seen": 48162230, + "step": 2244, + "time_per_iteration": 2.7883124351501465 + }, + { + "auxiliary_loss_clip": 0.01203345, + "auxiliary_loss_mlp": 0.01028788, + "balance_loss_clip": 1.06454611, + "balance_loss_mlp": 1.01985323, + "epoch": 0.2699452894847592, + "flos": 17308117653120.0, + "grad_norm": 2.338856544711972, + "language_loss": 0.72543335, + "learning_rate": 3.426006386722726e-06, + "loss": 0.74775469, + "num_input_tokens_seen": 48180290, + "step": 2245, + "time_per_iteration": 2.600127696990967 + }, + { + "auxiliary_loss_clip": 0.01139544, + "auxiliary_loss_mlp": 0.01036545, + "balance_loss_clip": 0.9522475, + "balance_loss_mlp": 1.02732372, + "epoch": 0.2700655323753983, + "flos": 18078150441600.0, + "grad_norm": 1.9921307070922813, + "language_loss": 0.92409712, + "learning_rate": 3.4254600942028914e-06, + "loss": 0.945858, + "num_input_tokens_seen": 48198165, + "step": 2246, + "time_per_iteration": 2.6367475986480713 + }, + { + "auxiliary_loss_clip": 0.01151558, + "auxiliary_loss_mlp": 0.01028933, + "balance_loss_clip": 0.9846704, + "balance_loss_mlp": 1.02048707, + "epoch": 0.2701857752660374, + "flos": 18186636493440.0, + "grad_norm": 2.2872030650788537, + "language_loss": 0.82775241, + "learning_rate": 3.424913585441893e-06, + "loss": 0.84955728, + "num_input_tokens_seen": 48216000, + "step": 2247, + "time_per_iteration": 2.694944381713867 + }, + { + "auxiliary_loss_clip": 0.01175404, + "auxiliary_loss_mlp": 0.01032151, + "balance_loss_clip": 1.02294707, + "balance_loss_mlp": 1.02283478, + "epoch": 0.2703060181566765, + "flos": 16319496648960.0, + "grad_norm": 1.9756772963598153, + "language_loss": 0.87089676, + "learning_rate": 3.4243668605226374e-06, + "loss": 0.89297229, + "num_input_tokens_seen": 48233025, + "step": 2248, + "time_per_iteration": 2.5858006477355957 + }, + { + "auxiliary_loss_clip": 0.01128583, + "auxiliary_loss_mlp": 0.01031789, + "balance_loss_clip": 0.94592607, + "balance_loss_mlp": 1.02161455, + "epoch": 0.2704262610473156, + "flos": 19572357329280.0, + "grad_norm": 2.4033971115970205, + "language_loss": 0.82506835, + "learning_rate": 3.423819919528061e-06, + "loss": 0.84667212, + "num_input_tokens_seen": 48251110, + "step": 2249, + "time_per_iteration": 2.703909397125244 + }, + { + "auxiliary_loss_clip": 0.01109233, + "auxiliary_loss_mlp": 0.01030013, + "balance_loss_clip": 0.90234888, + "balance_loss_mlp": 1.0209713, + "epoch": 0.27054650393795465, + "flos": 20740746925440.0, + "grad_norm": 1.9543636206596828, + "language_loss": 0.78559184, + "learning_rate": 3.4232727625411355e-06, + "loss": 0.80698431, + "num_input_tokens_seen": 48270215, + "step": 2250, + "time_per_iteration": 2.766371726989746 + }, + { + "auxiliary_loss_clip": 0.01073959, + "auxiliary_loss_mlp": 0.01029455, + "balance_loss_clip": 0.8660931, + "balance_loss_mlp": 1.02078879, + "epoch": 0.27066674682859376, + "flos": 18658322916480.0, + "grad_norm": 1.8521636506931884, + "language_loss": 0.86348033, + "learning_rate": 3.4227253896448626e-06, + "loss": 0.88451451, + "num_input_tokens_seen": 48288075, + "step": 2251, + "time_per_iteration": 2.7639877796173096 + }, + { + "auxiliary_loss_clip": 0.01199565, + "auxiliary_loss_mlp": 0.01032978, + "balance_loss_clip": 1.05956125, + "balance_loss_mlp": 1.02361465, + "epoch": 0.2707869897192329, + "flos": 23002759958400.0, + "grad_norm": 2.271861537486547, + "language_loss": 0.81892145, + "learning_rate": 3.42217780092228e-06, + "loss": 0.84124684, + "num_input_tokens_seen": 48306415, + "step": 2252, + "time_per_iteration": 3.5705068111419678 + }, + { + "auxiliary_loss_clip": 0.01040683, + "auxiliary_loss_mlp": 0.01000582, + "balance_loss_clip": 0.91314048, + "balance_loss_mlp": 0.99878198, + "epoch": 0.27090723260987193, + "flos": 58323240293760.0, + "grad_norm": 0.7822281513763869, + "language_loss": 0.60283077, + "learning_rate": 3.421629996456456e-06, + "loss": 0.62324345, + "num_input_tokens_seen": 48365035, + "step": 2253, + "time_per_iteration": 4.105445623397827 + }, + { + "auxiliary_loss_clip": 0.01173761, + "auxiliary_loss_mlp": 0.01027578, + "balance_loss_clip": 1.02125657, + "balance_loss_mlp": 1.01802301, + "epoch": 0.27102747550051104, + "flos": 11984540797440.0, + "grad_norm": 2.0787006210101056, + "language_loss": 0.82294834, + "learning_rate": 3.421081976330491e-06, + "loss": 0.84496176, + "num_input_tokens_seen": 48383550, + "step": 2254, + "time_per_iteration": 2.57816481590271 + }, + { + "auxiliary_loss_clip": 0.01148742, + "auxiliary_loss_mlp": 0.0103135, + "balance_loss_clip": 0.98178029, + "balance_loss_mlp": 1.02236152, + "epoch": 0.27114771839115015, + "flos": 19900401264000.0, + "grad_norm": 1.7853728024749085, + "language_loss": 0.88133568, + "learning_rate": 3.4205337406275207e-06, + "loss": 0.90313661, + "num_input_tokens_seen": 48403670, + "step": 2255, + "time_per_iteration": 3.6182501316070557 + }, + { + "auxiliary_loss_clip": 0.01197963, + "auxiliary_loss_mlp": 0.01027354, + "balance_loss_clip": 1.05761743, + "balance_loss_mlp": 1.01833558, + "epoch": 0.2712679612817892, + "flos": 18331966920960.0, + "grad_norm": 2.276712848086626, + "language_loss": 0.75355506, + "learning_rate": 3.4199852894307114e-06, + "loss": 0.77580822, + "num_input_tokens_seen": 48420420, + "step": 2256, + "time_per_iteration": 2.5944955348968506 + }, + { + "auxiliary_loss_clip": 0.01084371, + "auxiliary_loss_mlp": 0.0103066, + "balance_loss_clip": 0.86824089, + "balance_loss_mlp": 1.02056921, + "epoch": 0.2713882041724283, + "flos": 24460302038400.0, + "grad_norm": 2.7256579535604164, + "language_loss": 0.79113543, + "learning_rate": 3.419436622823262e-06, + "loss": 0.81228578, + "num_input_tokens_seen": 48441140, + "step": 2257, + "time_per_iteration": 3.7509002685546875 + }, + { + "auxiliary_loss_clip": 0.0114797, + "auxiliary_loss_mlp": 0.01030826, + "balance_loss_clip": 0.98229581, + "balance_loss_mlp": 1.02269578, + "epoch": 0.27150844706306737, + "flos": 23039317025280.0, + "grad_norm": 1.7541201481092639, + "language_loss": 0.73980248, + "learning_rate": 3.4188877408884063e-06, + "loss": 0.76159048, + "num_input_tokens_seen": 48461845, + "step": 2258, + "time_per_iteration": 2.6949479579925537 + }, + { + "auxiliary_loss_clip": 0.01150258, + "auxiliary_loss_mlp": 0.01030168, + "balance_loss_clip": 0.9831543, + "balance_loss_mlp": 1.02135253, + "epoch": 0.2716286899537065, + "flos": 22563644192640.0, + "grad_norm": 3.134971573272853, + "language_loss": 0.65580249, + "learning_rate": 3.4183386437094088e-06, + "loss": 0.67760676, + "num_input_tokens_seen": 48478510, + "step": 2259, + "time_per_iteration": 2.6842496395111084 + }, + { + "auxiliary_loss_clip": 0.01154072, + "auxiliary_loss_mlp": 0.01027449, + "balance_loss_clip": 0.98181474, + "balance_loss_mlp": 1.01832962, + "epoch": 0.2717489328443456, + "flos": 13115044523520.0, + "grad_norm": 2.293031780870284, + "language_loss": 0.82232773, + "learning_rate": 3.417789331369565e-06, + "loss": 0.84414291, + "num_input_tokens_seen": 48494300, + "step": 2260, + "time_per_iteration": 2.6907598972320557 + }, + { + "auxiliary_loss_clip": 0.01206409, + "auxiliary_loss_mlp": 0.01031111, + "balance_loss_clip": 1.06465852, + "balance_loss_mlp": 1.02240896, + "epoch": 0.27186917573498465, + "flos": 29278688060160.0, + "grad_norm": 2.589603345334333, + "language_loss": 0.90986443, + "learning_rate": 3.4172398039522088e-06, + "loss": 0.93223965, + "num_input_tokens_seen": 48515585, + "step": 2261, + "time_per_iteration": 2.649540424346924 + }, + { + "auxiliary_loss_clip": 0.01174349, + "auxiliary_loss_mlp": 0.01026686, + "balance_loss_clip": 1.01947868, + "balance_loss_mlp": 1.01764965, + "epoch": 0.27198941862562376, + "flos": 26032220000640.0, + "grad_norm": 1.7250936642986663, + "language_loss": 0.79888725, + "learning_rate": 3.4166900615407e-06, + "loss": 0.82089764, + "num_input_tokens_seen": 48533500, + "step": 2262, + "time_per_iteration": 2.745497226715088 + }, + { + "auxiliary_loss_clip": 0.01174594, + "auxiliary_loss_mlp": 0.01034059, + "balance_loss_clip": 1.02291059, + "balance_loss_mlp": 1.02560067, + "epoch": 0.27210966151626287, + "flos": 32780983760640.0, + "grad_norm": 2.2403625309085027, + "language_loss": 0.74487698, + "learning_rate": 3.416140104218436e-06, + "loss": 0.76696348, + "num_input_tokens_seen": 48552865, + "step": 2263, + "time_per_iteration": 2.733386993408203 + }, + { + "auxiliary_loss_clip": 0.01051471, + "auxiliary_loss_mlp": 0.00502888, + "balance_loss_clip": 0.94764334, + "balance_loss_mlp": 0.0, + "epoch": 0.2722299044069019, + "flos": 65471043219840.0, + "grad_norm": 0.8415427431428589, + "language_loss": 0.69670981, + "learning_rate": 3.4155899320688437e-06, + "loss": 0.71225345, + "num_input_tokens_seen": 48618940, + "step": 2264, + "time_per_iteration": 3.311062812805176 + }, + { + "auxiliary_loss_clip": 0.0108169, + "auxiliary_loss_mlp": 0.01032402, + "balance_loss_clip": 0.86908728, + "balance_loss_mlp": 1.02380669, + "epoch": 0.27235014729754103, + "flos": 15334143782400.0, + "grad_norm": 2.214011066083453, + "language_loss": 0.74093521, + "learning_rate": 3.415039545175384e-06, + "loss": 0.76207614, + "num_input_tokens_seen": 48634665, + "step": 2265, + "time_per_iteration": 2.727426528930664 + }, + { + "auxiliary_loss_clip": 0.0117778, + "auxiliary_loss_mlp": 0.01033028, + "balance_loss_clip": 1.02178049, + "balance_loss_mlp": 1.02430177, + "epoch": 0.27247039018818014, + "flos": 21872363973120.0, + "grad_norm": 2.314644912178037, + "language_loss": 0.64836764, + "learning_rate": 3.414488943621551e-06, + "loss": 0.67047572, + "num_input_tokens_seen": 48653330, + "step": 2266, + "time_per_iteration": 2.652907371520996 + }, + { + "auxiliary_loss_clip": 0.01174911, + "auxiliary_loss_mlp": 0.0103216, + "balance_loss_clip": 1.02368999, + "balance_loss_mlp": 1.02279043, + "epoch": 0.2725906330788192, + "flos": 18695490514560.0, + "grad_norm": 2.180954709901695, + "language_loss": 0.73584729, + "learning_rate": 3.41393812749087e-06, + "loss": 0.757918, + "num_input_tokens_seen": 48671375, + "step": 2267, + "time_per_iteration": 2.648418664932251 + }, + { + "auxiliary_loss_clip": 0.01152154, + "auxiliary_loss_mlp": 0.01031749, + "balance_loss_clip": 0.98671079, + "balance_loss_mlp": 1.02355361, + "epoch": 0.2727108759694583, + "flos": 17886099398400.0, + "grad_norm": 2.924218957599069, + "language_loss": 0.71923023, + "learning_rate": 3.4133870968668984e-06, + "loss": 0.7410692, + "num_input_tokens_seen": 48686175, + "step": 2268, + "time_per_iteration": 2.635108709335327 + }, + { + "auxiliary_loss_clip": 0.01156231, + "auxiliary_loss_mlp": 0.01031198, + "balance_loss_clip": 0.98513985, + "balance_loss_mlp": 1.02209044, + "epoch": 0.2728311188600974, + "flos": 24461666755200.0, + "grad_norm": 1.829860608315325, + "language_loss": 0.78513521, + "learning_rate": 3.412835851833229e-06, + "loss": 0.80700946, + "num_input_tokens_seen": 48708370, + "step": 2269, + "time_per_iteration": 2.766636610031128 + }, + { + "auxiliary_loss_clip": 0.01176065, + "auxiliary_loss_mlp": 0.01034628, + "balance_loss_clip": 1.02546406, + "balance_loss_mlp": 1.02628362, + "epoch": 0.2729513617507365, + "flos": 30993314757120.0, + "grad_norm": 2.0896139051311255, + "language_loss": 0.7770716, + "learning_rate": 3.4122843924734834e-06, + "loss": 0.79917848, + "num_input_tokens_seen": 48730670, + "step": 2270, + "time_per_iteration": 2.7924771308898926 + }, + { + "auxiliary_loss_clip": 0.01147847, + "auxiliary_loss_mlp": 0.01028641, + "balance_loss_clip": 0.98171508, + "balance_loss_mlp": 1.01951551, + "epoch": 0.2730716046413756, + "flos": 19094637421440.0, + "grad_norm": 1.9514171854912985, + "language_loss": 0.87955278, + "learning_rate": 3.411732718871319e-06, + "loss": 0.90131766, + "num_input_tokens_seen": 48746510, + "step": 2271, + "time_per_iteration": 2.816511392593384 + }, + { + "auxiliary_loss_clip": 0.01199959, + "auxiliary_loss_mlp": 0.01032031, + "balance_loss_clip": 1.06420755, + "balance_loss_mlp": 1.02376974, + "epoch": 0.27319184753201464, + "flos": 26944566474240.0, + "grad_norm": 1.672439097267273, + "language_loss": 0.78571308, + "learning_rate": 3.4111808311104227e-06, + "loss": 0.80803293, + "num_input_tokens_seen": 48768825, + "step": 2272, + "time_per_iteration": 2.7399423122406006 + }, + { + "auxiliary_loss_clip": 0.01160199, + "auxiliary_loss_mlp": 0.01034546, + "balance_loss_clip": 0.98285747, + "balance_loss_mlp": 1.02485418, + "epoch": 0.27331209042265375, + "flos": 31759828012800.0, + "grad_norm": 2.0499107618290515, + "language_loss": 0.69109225, + "learning_rate": 3.410628729274517e-06, + "loss": 0.71303976, + "num_input_tokens_seen": 48790345, + "step": 2273, + "time_per_iteration": 2.798837661743164 + }, + { + "auxiliary_loss_clip": 0.0114741, + "auxiliary_loss_mlp": 0.00511326, + "balance_loss_clip": 0.98149502, + "balance_loss_mlp": 0.0, + "epoch": 0.27343233331329286, + "flos": 25739081107200.0, + "grad_norm": 1.843383765942278, + "language_loss": 0.8233639, + "learning_rate": 3.4100764134473546e-06, + "loss": 0.83995128, + "num_input_tokens_seen": 48809630, + "step": 2274, + "time_per_iteration": 2.745917320251465 + }, + { + "auxiliary_loss_clip": 0.01199865, + "auxiliary_loss_mlp": 0.01026547, + "balance_loss_clip": 1.06248522, + "balance_loss_mlp": 1.01727891, + "epoch": 0.2735525762039319, + "flos": 24389414547840.0, + "grad_norm": 2.2110488812504747, + "language_loss": 0.85109496, + "learning_rate": 3.4095238837127215e-06, + "loss": 0.87335908, + "num_input_tokens_seen": 48828770, + "step": 2275, + "time_per_iteration": 2.772203207015991 + }, + { + "auxiliary_loss_clip": 0.01123433, + "auxiliary_loss_mlp": 0.01036063, + "balance_loss_clip": 0.94247103, + "balance_loss_mlp": 1.02728319, + "epoch": 0.27367281909457103, + "flos": 14465357527680.0, + "grad_norm": 2.1525786935887474, + "language_loss": 0.79099727, + "learning_rate": 3.4089711401544355e-06, + "loss": 0.81259221, + "num_input_tokens_seen": 48846365, + "step": 2276, + "time_per_iteration": 2.6750059127807617 + }, + { + "auxiliary_loss_clip": 0.01172145, + "auxiliary_loss_mlp": 0.01031161, + "balance_loss_clip": 1.01686537, + "balance_loss_mlp": 1.02244043, + "epoch": 0.27379306198521014, + "flos": 23476996247040.0, + "grad_norm": 2.43854759617744, + "language_loss": 0.68115473, + "learning_rate": 3.4084181828563486e-06, + "loss": 0.70318782, + "num_input_tokens_seen": 48863085, + "step": 2277, + "time_per_iteration": 2.6670072078704834 + }, + { + "auxiliary_loss_clip": 0.01102926, + "auxiliary_loss_mlp": 0.01030235, + "balance_loss_clip": 0.90444356, + "balance_loss_mlp": 1.02113938, + "epoch": 0.2739133048758492, + "flos": 17458152762240.0, + "grad_norm": 1.7751983128103161, + "language_loss": 0.70463026, + "learning_rate": 3.4078650119023428e-06, + "loss": 0.72596192, + "num_input_tokens_seen": 48881400, + "step": 2278, + "time_per_iteration": 3.8123483657836914 + }, + { + "auxiliary_loss_clip": 0.01079185, + "auxiliary_loss_mlp": 0.01033606, + "balance_loss_clip": 0.86199552, + "balance_loss_mlp": 1.02483201, + "epoch": 0.2740335477664883, + "flos": 19273113123840.0, + "grad_norm": 2.4122238197908734, + "language_loss": 0.7430557, + "learning_rate": 3.4073116273763337e-06, + "loss": 0.76418364, + "num_input_tokens_seen": 48895845, + "step": 2279, + "time_per_iteration": 3.6776163578033447 + }, + { + "auxiliary_loss_clip": 0.01153738, + "auxiliary_loss_mlp": 0.01031279, + "balance_loss_clip": 0.98036993, + "balance_loss_mlp": 1.02157533, + "epoch": 0.2741537906571274, + "flos": 26104723603200.0, + "grad_norm": 2.210289372353223, + "language_loss": 0.81065619, + "learning_rate": 3.40675802936227e-06, + "loss": 0.8325063, + "num_input_tokens_seen": 48916630, + "step": 2280, + "time_per_iteration": 2.8003408908843994 + }, + { + "auxiliary_loss_clip": 0.01143348, + "auxiliary_loss_mlp": 0.01033293, + "balance_loss_clip": 0.97984135, + "balance_loss_mlp": 1.0236553, + "epoch": 0.27427403354776647, + "flos": 34164190644480.0, + "grad_norm": 1.9490405892604126, + "language_loss": 0.72213918, + "learning_rate": 3.4062042179441318e-06, + "loss": 0.74390554, + "num_input_tokens_seen": 48937100, + "step": 2281, + "time_per_iteration": 3.7564101219177246 + }, + { + "auxiliary_loss_clip": 0.01171788, + "auxiliary_loss_mlp": 0.01032872, + "balance_loss_clip": 1.02214503, + "balance_loss_mlp": 1.02459908, + "epoch": 0.2743942764384056, + "flos": 18766988536320.0, + "grad_norm": 1.8372070427081142, + "language_loss": 0.80339247, + "learning_rate": 3.4056501932059314e-06, + "loss": 0.8254391, + "num_input_tokens_seen": 48955175, + "step": 2282, + "time_per_iteration": 2.667471170425415 + }, + { + "auxiliary_loss_clip": 0.01097109, + "auxiliary_loss_mlp": 0.01003375, + "balance_loss_clip": 1.02563787, + "balance_loss_mlp": 1.00152683, + "epoch": 0.2745145193290447, + "flos": 64904048058240.0, + "grad_norm": 0.7729077591161317, + "language_loss": 0.58046567, + "learning_rate": 3.405095955231715e-06, + "loss": 0.60147047, + "num_input_tokens_seen": 49006830, + "step": 2283, + "time_per_iteration": 3.1405179500579834 + }, + { + "auxiliary_loss_clip": 0.01178498, + "auxiliary_loss_mlp": 0.0102722, + "balance_loss_clip": 1.0200665, + "balance_loss_mlp": 1.01866078, + "epoch": 0.27463476221968375, + "flos": 16136926796160.0, + "grad_norm": 2.6207776302179036, + "language_loss": 0.9429332, + "learning_rate": 3.4045415041055585e-06, + "loss": 0.96499038, + "num_input_tokens_seen": 49022470, + "step": 2284, + "time_per_iteration": 3.6034910678863525 + }, + { + "auxiliary_loss_clip": 0.01155743, + "auxiliary_loss_mlp": 0.01029064, + "balance_loss_clip": 0.98317587, + "balance_loss_mlp": 1.01953316, + "epoch": 0.27475500511032286, + "flos": 10376712213120.0, + "grad_norm": 2.2905975015509643, + "language_loss": 0.78387052, + "learning_rate": 3.4039868399115728e-06, + "loss": 0.8057186, + "num_input_tokens_seen": 49037110, + "step": 2285, + "time_per_iteration": 2.5914230346679688 + }, + { + "auxiliary_loss_clip": 0.01085015, + "auxiliary_loss_mlp": 0.01033332, + "balance_loss_clip": 0.87062037, + "balance_loss_mlp": 1.02438569, + "epoch": 0.27487524800096197, + "flos": 17311062568320.0, + "grad_norm": 2.2977585701068515, + "language_loss": 0.80418289, + "learning_rate": 3.4034319627339003e-06, + "loss": 0.82536638, + "num_input_tokens_seen": 49053975, + "step": 2286, + "time_per_iteration": 2.809285879135132 + }, + { + "auxiliary_loss_clip": 0.01155792, + "auxiliary_loss_mlp": 0.0102994, + "balance_loss_clip": 0.98547959, + "balance_loss_mlp": 1.0210712, + "epoch": 0.274995490891601, + "flos": 27120205002240.0, + "grad_norm": 4.507745165114525, + "language_loss": 0.69571352, + "learning_rate": 3.402876872656715e-06, + "loss": 0.7175709, + "num_input_tokens_seen": 49072295, + "step": 2287, + "time_per_iteration": 2.677394151687622 + }, + { + "auxiliary_loss_clip": 0.0115285, + "auxiliary_loss_mlp": 0.01032966, + "balance_loss_clip": 0.98502165, + "balance_loss_mlp": 1.02411461, + "epoch": 0.27511573378224013, + "flos": 23436093634560.0, + "grad_norm": 2.0612594790544323, + "language_loss": 0.89582515, + "learning_rate": 3.402321569764223e-06, + "loss": 0.9176833, + "num_input_tokens_seen": 49091600, + "step": 2288, + "time_per_iteration": 2.805405616760254 + }, + { + "auxiliary_loss_clip": 0.01107313, + "auxiliary_loss_mlp": 0.00511916, + "balance_loss_clip": 0.90385151, + "balance_loss_mlp": 0.0, + "epoch": 0.2752359766728792, + "flos": 16722019434240.0, + "grad_norm": 2.679978072373453, + "language_loss": 0.83491248, + "learning_rate": 3.4017660541406635e-06, + "loss": 0.8511048, + "num_input_tokens_seen": 49107665, + "step": 2289, + "time_per_iteration": 2.7033700942993164 + }, + { + "auxiliary_loss_clip": 0.01159597, + "auxiliary_loss_mlp": 0.01032201, + "balance_loss_clip": 0.98232681, + "balance_loss_mlp": 1.02269399, + "epoch": 0.2753562195635183, + "flos": 25297738698240.0, + "grad_norm": 1.6246729292421465, + "language_loss": 0.73627901, + "learning_rate": 3.4012103258703092e-06, + "loss": 0.75819695, + "num_input_tokens_seen": 49126420, + "step": 2290, + "time_per_iteration": 2.6974234580993652 + }, + { + "auxiliary_loss_clip": 0.0112989, + "auxiliary_loss_mlp": 0.01031765, + "balance_loss_clip": 0.9454307, + "balance_loss_mlp": 1.02283013, + "epoch": 0.2754764624541574, + "flos": 27338972785920.0, + "grad_norm": 2.0329783694920165, + "language_loss": 0.82758921, + "learning_rate": 3.4006543850374616e-06, + "loss": 0.84920579, + "num_input_tokens_seen": 49141470, + "step": 2291, + "time_per_iteration": 2.7582297325134277 + }, + { + "auxiliary_loss_clip": 0.01181254, + "auxiliary_loss_mlp": 0.0103601, + "balance_loss_clip": 1.02297461, + "balance_loss_mlp": 1.0262711, + "epoch": 0.27559670534479647, + "flos": 17238379397760.0, + "grad_norm": 2.047259439404056, + "language_loss": 0.75016785, + "learning_rate": 3.400098231726458e-06, + "loss": 0.77234054, + "num_input_tokens_seen": 49158570, + "step": 2292, + "time_per_iteration": 2.646892547607422 + }, + { + "auxiliary_loss_clip": 0.01130961, + "auxiliary_loss_mlp": 0.01033086, + "balance_loss_clip": 0.94117951, + "balance_loss_mlp": 1.0238595, + "epoch": 0.2757169482354356, + "flos": 21939085486080.0, + "grad_norm": 2.3825042277159096, + "language_loss": 0.86835182, + "learning_rate": 3.3995418660216657e-06, + "loss": 0.88999224, + "num_input_tokens_seen": 49176025, + "step": 2293, + "time_per_iteration": 2.8224198818206787 + }, + { + "auxiliary_loss_clip": 0.0120794, + "auxiliary_loss_mlp": 0.01029744, + "balance_loss_clip": 1.06342494, + "balance_loss_mlp": 1.02006471, + "epoch": 0.2758371911260747, + "flos": 20850669521280.0, + "grad_norm": 2.6019632874748826, + "language_loss": 0.8067404, + "learning_rate": 3.3989852880074848e-06, + "loss": 0.8291173, + "num_input_tokens_seen": 49197455, + "step": 2294, + "time_per_iteration": 2.6057512760162354 + }, + { + "auxiliary_loss_clip": 0.01059311, + "auxiliary_loss_mlp": 0.01007944, + "balance_loss_clip": 0.95836198, + "balance_loss_mlp": 1.00603616, + "epoch": 0.27595743401671374, + "flos": 69269063592960.0, + "grad_norm": 0.7416775823748297, + "language_loss": 0.60599291, + "learning_rate": 3.398428497768348e-06, + "loss": 0.62666547, + "num_input_tokens_seen": 49262625, + "step": 2295, + "time_per_iteration": 3.34500789642334 + }, + { + "auxiliary_loss_clip": 0.01134076, + "auxiliary_loss_mlp": 0.01028583, + "balance_loss_clip": 0.94236469, + "balance_loss_mlp": 1.01938009, + "epoch": 0.27607767690735285, + "flos": 21215019127680.0, + "grad_norm": 2.2710651623009084, + "language_loss": 0.722363, + "learning_rate": 3.3978714953887205e-06, + "loss": 0.74398959, + "num_input_tokens_seen": 49282380, + "step": 2296, + "time_per_iteration": 2.7035486698150635 + }, + { + "auxiliary_loss_clip": 0.01091903, + "auxiliary_loss_mlp": 0.01035656, + "balance_loss_clip": 0.90050292, + "balance_loss_mlp": 1.02728176, + "epoch": 0.27619791979799196, + "flos": 24825334003200.0, + "grad_norm": 1.8591900125746939, + "language_loss": 0.85776395, + "learning_rate": 3.397314280953098e-06, + "loss": 0.87903953, + "num_input_tokens_seen": 49303205, + "step": 2297, + "time_per_iteration": 2.780933141708374 + }, + { + "auxiliary_loss_clip": 0.01147285, + "auxiliary_loss_mlp": 0.01035161, + "balance_loss_clip": 0.98294348, + "balance_loss_mlp": 1.02654767, + "epoch": 0.276318162688631, + "flos": 24753548672640.0, + "grad_norm": 2.1734716999847117, + "language_loss": 0.80420369, + "learning_rate": 3.3967568545460108e-06, + "loss": 0.82602817, + "num_input_tokens_seen": 49322745, + "step": 2298, + "time_per_iteration": 2.6995952129364014 + }, + { + "auxiliary_loss_clip": 0.0117406, + "auxiliary_loss_mlp": 0.01033302, + "balance_loss_clip": 1.02162349, + "balance_loss_mlp": 1.0246954, + "epoch": 0.27643840557927013, + "flos": 18150007599360.0, + "grad_norm": 1.895884880042651, + "language_loss": 0.80633521, + "learning_rate": 3.3961992162520185e-06, + "loss": 0.82840878, + "num_input_tokens_seen": 49341370, + "step": 2299, + "time_per_iteration": 2.6869418621063232 + }, + { + "auxiliary_loss_clip": 0.01175709, + "auxiliary_loss_mlp": 0.01027969, + "balance_loss_clip": 1.02024221, + "balance_loss_mlp": 1.01877236, + "epoch": 0.27655864846990924, + "flos": 24823933372800.0, + "grad_norm": 2.4806724631789647, + "language_loss": 0.7168678, + "learning_rate": 3.3956413661557156e-06, + "loss": 0.7389046, + "num_input_tokens_seen": 49361545, + "step": 2300, + "time_per_iteration": 2.7214725017547607 + }, + { + "auxiliary_loss_clip": 0.01135926, + "auxiliary_loss_mlp": 0.01035871, + "balance_loss_clip": 0.94376379, + "balance_loss_mlp": 1.02660239, + "epoch": 0.2766788913605483, + "flos": 20266582464000.0, + "grad_norm": 2.282432493966539, + "language_loss": 0.66789007, + "learning_rate": 3.3950833043417273e-06, + "loss": 0.6896081, + "num_input_tokens_seen": 49379690, + "step": 2301, + "time_per_iteration": 2.703679084777832 + }, + { + "auxiliary_loss_clip": 0.01179891, + "auxiliary_loss_mlp": 0.01035823, + "balance_loss_clip": 1.02526724, + "balance_loss_mlp": 1.0261972, + "epoch": 0.2767991342511874, + "flos": 21470272151040.0, + "grad_norm": 2.5894161537712, + "language_loss": 0.73057318, + "learning_rate": 3.3945250308947105e-06, + "loss": 0.75273025, + "num_input_tokens_seen": 49395995, + "step": 2302, + "time_per_iteration": 2.638427972793579 + }, + { + "auxiliary_loss_clip": 0.01079812, + "auxiliary_loss_mlp": 0.01008181, + "balance_loss_clip": 0.98997593, + "balance_loss_mlp": 1.00634515, + "epoch": 0.2769193771418265, + "flos": 66002627571840.0, + "grad_norm": 1.2995291514712117, + "language_loss": 0.68386018, + "learning_rate": 3.3939665458993556e-06, + "loss": 0.70474017, + "num_input_tokens_seen": 49450415, + "step": 2303, + "time_per_iteration": 3.14447283744812 + }, + { + "auxiliary_loss_clip": 0.01131975, + "auxiliary_loss_mlp": 0.01030882, + "balance_loss_clip": 0.94271648, + "balance_loss_mlp": 1.02221513, + "epoch": 0.27703962003246557, + "flos": 20704441253760.0, + "grad_norm": 2.2269634197423525, + "language_loss": 0.76846194, + "learning_rate": 3.3934078494403843e-06, + "loss": 0.79009056, + "num_input_tokens_seen": 49469990, + "step": 2304, + "time_per_iteration": 3.624527931213379 + }, + { + "auxiliary_loss_clip": 0.01051294, + "auxiliary_loss_mlp": 0.00510982, + "balance_loss_clip": 0.82606339, + "balance_loss_mlp": 0.0, + "epoch": 0.2771598629231047, + "flos": 22929897219840.0, + "grad_norm": 1.790759288778379, + "language_loss": 0.81071341, + "learning_rate": 3.3928489416025495e-06, + "loss": 0.82633621, + "num_input_tokens_seen": 49490835, + "step": 2305, + "time_per_iteration": 3.8127310276031494 + }, + { + "auxiliary_loss_clip": 0.01151966, + "auxiliary_loss_mlp": 0.0102767, + "balance_loss_clip": 0.98297244, + "balance_loss_mlp": 1.01789451, + "epoch": 0.27728010581374374, + "flos": 18369457741440.0, + "grad_norm": 2.3236340692784903, + "language_loss": 0.79005039, + "learning_rate": 3.392289822470638e-06, + "loss": 0.81184679, + "num_input_tokens_seen": 49508815, + "step": 2306, + "time_per_iteration": 2.8238847255706787 + }, + { + "auxiliary_loss_clip": 0.01149457, + "auxiliary_loss_mlp": 0.01029726, + "balance_loss_clip": 0.98100758, + "balance_loss_mlp": 1.02021933, + "epoch": 0.27740034870438285, + "flos": 19427637432960.0, + "grad_norm": 2.155984585253295, + "language_loss": 0.76108295, + "learning_rate": 3.3917304921294674e-06, + "loss": 0.78287476, + "num_input_tokens_seen": 49526980, + "step": 2307, + "time_per_iteration": 3.57962703704834 + }, + { + "auxiliary_loss_clip": 0.01176069, + "auxiliary_loss_mlp": 0.01032396, + "balance_loss_clip": 1.01965666, + "balance_loss_mlp": 1.02315712, + "epoch": 0.27752059159502196, + "flos": 21614776565760.0, + "grad_norm": 1.7794844452096965, + "language_loss": 0.80546379, + "learning_rate": 3.3911709506638876e-06, + "loss": 0.82754838, + "num_input_tokens_seen": 49546290, + "step": 2308, + "time_per_iteration": 2.705569267272949 + }, + { + "auxiliary_loss_clip": 0.01115675, + "auxiliary_loss_mlp": 0.00511293, + "balance_loss_clip": 0.9379006, + "balance_loss_mlp": 0.0, + "epoch": 0.277640834485661, + "flos": 26608011016320.0, + "grad_norm": 2.411652690138517, + "language_loss": 0.81380242, + "learning_rate": 3.390611198158781e-06, + "loss": 0.8300721, + "num_input_tokens_seen": 49564165, + "step": 2309, + "time_per_iteration": 3.6558382511138916 + }, + { + "auxiliary_loss_clip": 0.01205002, + "auxiliary_loss_mlp": 0.01035941, + "balance_loss_clip": 1.06304204, + "balance_loss_mlp": 1.02585053, + "epoch": 0.2777610773763001, + "flos": 19492814661120.0, + "grad_norm": 2.1614676070126597, + "language_loss": 0.89671075, + "learning_rate": 3.3900512346990612e-06, + "loss": 0.91912019, + "num_input_tokens_seen": 49580155, + "step": 2310, + "time_per_iteration": 2.6291770935058594 + }, + { + "auxiliary_loss_clip": 0.01102593, + "auxiliary_loss_mlp": 0.01029643, + "balance_loss_clip": 0.90068507, + "balance_loss_mlp": 1.01966536, + "epoch": 0.27788132026693924, + "flos": 38290650001920.0, + "grad_norm": 2.3740178981609357, + "language_loss": 0.65772772, + "learning_rate": 3.389491060369674e-06, + "loss": 0.67905015, + "num_input_tokens_seen": 49605830, + "step": 2311, + "time_per_iteration": 2.871166944503784 + }, + { + "auxiliary_loss_clip": 0.01096511, + "auxiliary_loss_mlp": 0.01036497, + "balance_loss_clip": 0.90268099, + "balance_loss_mlp": 1.02792597, + "epoch": 0.2780015631575783, + "flos": 22382546797440.0, + "grad_norm": 1.8851416093323787, + "language_loss": 0.89514542, + "learning_rate": 3.388930675255598e-06, + "loss": 0.91647542, + "num_input_tokens_seen": 49625680, + "step": 2312, + "time_per_iteration": 2.7628087997436523 + }, + { + "auxiliary_loss_clip": 0.01156151, + "auxiliary_loss_mlp": 0.01028015, + "balance_loss_clip": 0.98143125, + "balance_loss_mlp": 1.018538, + "epoch": 0.2781218060482174, + "flos": 12203200840320.0, + "grad_norm": 2.7383854185176295, + "language_loss": 0.79294956, + "learning_rate": 3.388370079441843e-06, + "loss": 0.8147912, + "num_input_tokens_seen": 49641195, + "step": 2313, + "time_per_iteration": 2.6691668033599854 + }, + { + "auxiliary_loss_clip": 0.01127392, + "auxiliary_loss_mlp": 0.0103219, + "balance_loss_clip": 0.94398546, + "balance_loss_mlp": 1.0226239, + "epoch": 0.2782420489388565, + "flos": 18107632529280.0, + "grad_norm": 1.8984320470206013, + "language_loss": 0.92581159, + "learning_rate": 3.3878092730134505e-06, + "loss": 0.94740742, + "num_input_tokens_seen": 49659180, + "step": 2314, + "time_per_iteration": 2.8690521717071533 + }, + { + "auxiliary_loss_clip": 0.01169395, + "auxiliary_loss_mlp": 0.01034827, + "balance_loss_clip": 1.01873553, + "balance_loss_mlp": 1.02605295, + "epoch": 0.27836229182949557, + "flos": 18514752255360.0, + "grad_norm": 1.7302946424526888, + "language_loss": 0.80581367, + "learning_rate": 3.3872482560554947e-06, + "loss": 0.82785594, + "num_input_tokens_seen": 49677955, + "step": 2315, + "time_per_iteration": 2.7148525714874268 + }, + { + "auxiliary_loss_clip": 0.01077002, + "auxiliary_loss_mlp": 0.01004426, + "balance_loss_clip": 0.98656225, + "balance_loss_mlp": 1.00256634, + "epoch": 0.2784825347201347, + "flos": 67079230940160.0, + "grad_norm": 0.8053100854684638, + "language_loss": 0.56949043, + "learning_rate": 3.386687028653082e-06, + "loss": 0.59030473, + "num_input_tokens_seen": 49740800, + "step": 2316, + "time_per_iteration": 3.279787302017212 + }, + { + "auxiliary_loss_clip": 0.01108053, + "auxiliary_loss_mlp": 0.01033379, + "balance_loss_clip": 0.90760678, + "balance_loss_mlp": 1.02399731, + "epoch": 0.2786027776107738, + "flos": 22631119891200.0, + "grad_norm": 1.8709518758611778, + "language_loss": 0.85002947, + "learning_rate": 3.386125590891349e-06, + "loss": 0.87144375, + "num_input_tokens_seen": 49757675, + "step": 2317, + "time_per_iteration": 2.794264793395996 + }, + { + "auxiliary_loss_clip": 0.01139858, + "auxiliary_loss_mlp": 0.01032873, + "balance_loss_clip": 0.9781428, + "balance_loss_mlp": 1.02411151, + "epoch": 0.27872302050141284, + "flos": 15778826156160.0, + "grad_norm": 2.4085116722304507, + "language_loss": 0.83226144, + "learning_rate": 3.3855639428554657e-06, + "loss": 0.85398877, + "num_input_tokens_seen": 49775205, + "step": 2318, + "time_per_iteration": 3.0868730545043945 + }, + { + "auxiliary_loss_clip": 0.01118946, + "auxiliary_loss_mlp": 0.01029955, + "balance_loss_clip": 0.94148386, + "balance_loss_mlp": 1.02127075, + "epoch": 0.27884326339205195, + "flos": 22126970551680.0, + "grad_norm": 1.7597512143035794, + "language_loss": 0.80212563, + "learning_rate": 3.385002084630635e-06, + "loss": 0.82361472, + "num_input_tokens_seen": 49794175, + "step": 2319, + "time_per_iteration": 2.9501328468322754 + }, + { + "auxiliary_loss_clip": 0.01181531, + "auxiliary_loss_mlp": 0.01035875, + "balance_loss_clip": 1.02128911, + "balance_loss_mlp": 1.02637386, + "epoch": 0.278963506282691, + "flos": 20558715776640.0, + "grad_norm": 2.1265780225388595, + "language_loss": 0.85240901, + "learning_rate": 3.384440016302088e-06, + "loss": 0.87458313, + "num_input_tokens_seen": 49812850, + "step": 2320, + "time_per_iteration": 2.693298101425171 + }, + { + "auxiliary_loss_clip": 0.01173706, + "auxiliary_loss_mlp": 0.01029063, + "balance_loss_clip": 1.02125788, + "balance_loss_mlp": 1.01991987, + "epoch": 0.2790837491733301, + "flos": 21942928241280.0, + "grad_norm": 2.6227340600205973, + "language_loss": 0.6242469, + "learning_rate": 3.3838777379550923e-06, + "loss": 0.64627457, + "num_input_tokens_seen": 49832295, + "step": 2321, + "time_per_iteration": 2.70815110206604 + }, + { + "auxiliary_loss_clip": 0.01157115, + "auxiliary_loss_mlp": 0.01031272, + "balance_loss_clip": 0.98403275, + "balance_loss_mlp": 1.02243245, + "epoch": 0.27920399206396923, + "flos": 26286790665600.0, + "grad_norm": 2.728580939775295, + "language_loss": 0.78465134, + "learning_rate": 3.383315249674944e-06, + "loss": 0.80653518, + "num_input_tokens_seen": 49850860, + "step": 2322, + "time_per_iteration": 2.7867774963378906 + }, + { + "auxiliary_loss_clip": 0.01130845, + "auxiliary_loss_mlp": 0.01035275, + "balance_loss_clip": 0.94432622, + "balance_loss_mlp": 1.02679336, + "epoch": 0.2793242349546083, + "flos": 25400981364480.0, + "grad_norm": 1.9750597840422297, + "language_loss": 0.86412674, + "learning_rate": 3.3827525515469715e-06, + "loss": 0.88578796, + "num_input_tokens_seen": 49865765, + "step": 2323, + "time_per_iteration": 2.7424702644348145 + }, + { + "auxiliary_loss_clip": 0.01118982, + "auxiliary_loss_mlp": 0.01028269, + "balance_loss_clip": 0.9402436, + "balance_loss_mlp": 1.01824331, + "epoch": 0.2794444778452474, + "flos": 20850346298880.0, + "grad_norm": 2.31601360216012, + "language_loss": 0.70933223, + "learning_rate": 3.3821896436565367e-06, + "loss": 0.7308048, + "num_input_tokens_seen": 49885425, + "step": 2324, + "time_per_iteration": 2.7004857063293457 + }, + { + "auxiliary_loss_clip": 0.01177433, + "auxiliary_loss_mlp": 0.01031198, + "balance_loss_clip": 1.02338552, + "balance_loss_mlp": 1.02197099, + "epoch": 0.2795647207358865, + "flos": 21576244250880.0, + "grad_norm": 1.667843317037137, + "language_loss": 0.70348918, + "learning_rate": 3.381626526089032e-06, + "loss": 0.72557545, + "num_input_tokens_seen": 49904990, + "step": 2325, + "time_per_iteration": 2.7624027729034424 + }, + { + "auxiliary_loss_clip": 0.01149269, + "auxiliary_loss_mlp": 0.01030988, + "balance_loss_clip": 0.97934198, + "balance_loss_mlp": 1.02113509, + "epoch": 0.27968496362652556, + "flos": 21471744608640.0, + "grad_norm": 2.225169994943228, + "language_loss": 0.79014945, + "learning_rate": 3.3810631989298815e-06, + "loss": 0.81195194, + "num_input_tokens_seen": 49924600, + "step": 2326, + "time_per_iteration": 2.828517198562622 + }, + { + "auxiliary_loss_clip": 0.0111109, + "auxiliary_loss_mlp": 0.01038414, + "balance_loss_clip": 0.9065733, + "balance_loss_mlp": 1.02844274, + "epoch": 0.2798052065171647, + "flos": 23258695340160.0, + "grad_norm": 2.311079362406218, + "language_loss": 0.84463662, + "learning_rate": 3.3804996622645423e-06, + "loss": 0.86613166, + "num_input_tokens_seen": 49942600, + "step": 2327, + "time_per_iteration": 2.7823140621185303 + }, + { + "auxiliary_loss_clip": 0.01199692, + "auxiliary_loss_mlp": 0.01031457, + "balance_loss_clip": 1.0595541, + "balance_loss_mlp": 1.02316618, + "epoch": 0.2799254494078038, + "flos": 21539328048000.0, + "grad_norm": 2.58423208312565, + "language_loss": 0.89260191, + "learning_rate": 3.3799359161785015e-06, + "loss": 0.91491342, + "num_input_tokens_seen": 49962250, + "step": 2328, + "time_per_iteration": 2.740861415863037 + }, + { + "auxiliary_loss_clip": 0.01176306, + "auxiliary_loss_mlp": 0.01035765, + "balance_loss_clip": 1.02167439, + "balance_loss_mlp": 1.02641892, + "epoch": 0.28004569229844284, + "flos": 26393912000640.0, + "grad_norm": 1.7433424824637909, + "language_loss": 0.85669148, + "learning_rate": 3.3793719607572798e-06, + "loss": 0.87881225, + "num_input_tokens_seen": 49983215, + "step": 2329, + "time_per_iteration": 3.6516947746276855 + }, + { + "auxiliary_loss_clip": 0.01136654, + "auxiliary_loss_mlp": 0.01031514, + "balance_loss_clip": 0.97715205, + "balance_loss_mlp": 1.02234685, + "epoch": 0.28016593518908195, + "flos": 33547676584320.0, + "grad_norm": 3.495571464721565, + "language_loss": 0.77324283, + "learning_rate": 3.378807796086428e-06, + "loss": 0.79492456, + "num_input_tokens_seen": 50006075, + "step": 2330, + "time_per_iteration": 2.772650957107544 + }, + { + "auxiliary_loss_clip": 0.01200219, + "auxiliary_loss_mlp": 0.01027395, + "balance_loss_clip": 1.05933619, + "balance_loss_mlp": 1.01844847, + "epoch": 0.28028617807972106, + "flos": 15340823712000.0, + "grad_norm": 2.0014414275089827, + "language_loss": 0.7698046, + "learning_rate": 3.37824342225153e-06, + "loss": 0.79208076, + "num_input_tokens_seen": 50022495, + "step": 2331, + "time_per_iteration": 3.561471939086914 + }, + { + "auxiliary_loss_clip": 0.01108406, + "auxiliary_loss_mlp": 0.01031361, + "balance_loss_clip": 0.90778095, + "balance_loss_mlp": 1.02216434, + "epoch": 0.2804064209703601, + "flos": 25520277409920.0, + "grad_norm": 1.8582913677952186, + "language_loss": 0.77932113, + "learning_rate": 3.3776788393382006e-06, + "loss": 0.80071878, + "num_input_tokens_seen": 50041975, + "step": 2332, + "time_per_iteration": 2.7546544075012207 + }, + { + "auxiliary_loss_clip": 0.01204158, + "auxiliary_loss_mlp": 0.01036504, + "balance_loss_clip": 1.06218696, + "balance_loss_mlp": 1.02693748, + "epoch": 0.2805266638609992, + "flos": 29351766280320.0, + "grad_norm": 3.10627930963823, + "language_loss": 0.76527417, + "learning_rate": 3.3771140474320872e-06, + "loss": 0.7876808, + "num_input_tokens_seen": 50061925, + "step": 2333, + "time_per_iteration": 3.647465467453003 + }, + { + "auxiliary_loss_clip": 0.01137167, + "auxiliary_loss_mlp": 0.01036348, + "balance_loss_clip": 0.94628406, + "balance_loss_mlp": 1.02704942, + "epoch": 0.28064690675163834, + "flos": 21463735875840.0, + "grad_norm": 4.271457105211236, + "language_loss": 0.79338616, + "learning_rate": 3.3765490466188664e-06, + "loss": 0.81512135, + "num_input_tokens_seen": 50079325, + "step": 2334, + "time_per_iteration": 2.7029519081115723 + }, + { + "auxiliary_loss_clip": 0.01126494, + "auxiliary_loss_mlp": 0.01027825, + "balance_loss_clip": 0.94241738, + "balance_loss_mlp": 1.01887882, + "epoch": 0.2807671496422774, + "flos": 20995640812800.0, + "grad_norm": 5.180043448092013, + "language_loss": 0.73743773, + "learning_rate": 3.3759838369842508e-06, + "loss": 0.75898093, + "num_input_tokens_seen": 50097400, + "step": 2335, + "time_per_iteration": 3.5750789642333984 + }, + { + "auxiliary_loss_clip": 0.01131362, + "auxiliary_loss_mlp": 0.01031877, + "balance_loss_clip": 0.94421256, + "balance_loss_mlp": 1.02250099, + "epoch": 0.2808873925329165, + "flos": 21506577822720.0, + "grad_norm": 1.8112858756823542, + "language_loss": 0.72548395, + "learning_rate": 3.375418418613981e-06, + "loss": 0.74711633, + "num_input_tokens_seen": 50116425, + "step": 2336, + "time_per_iteration": 2.7159178256988525 + }, + { + "auxiliary_loss_clip": 0.01154983, + "auxiliary_loss_mlp": 0.01032611, + "balance_loss_clip": 0.98381543, + "balance_loss_mlp": 1.02269268, + "epoch": 0.28100763542355556, + "flos": 16070815814400.0, + "grad_norm": 2.381916202523656, + "language_loss": 0.83549726, + "learning_rate": 3.374852791593831e-06, + "loss": 0.85737312, + "num_input_tokens_seen": 50132625, + "step": 2337, + "time_per_iteration": 2.649933099746704 + }, + { + "auxiliary_loss_clip": 0.01114893, + "auxiliary_loss_mlp": 0.01033933, + "balance_loss_clip": 0.90514284, + "balance_loss_mlp": 1.02379405, + "epoch": 0.28112787831419467, + "flos": 19062605468160.0, + "grad_norm": 2.851391068951889, + "language_loss": 0.54232162, + "learning_rate": 3.374286956009605e-06, + "loss": 0.56380993, + "num_input_tokens_seen": 50151190, + "step": 2338, + "time_per_iteration": 2.654139518737793 + }, + { + "auxiliary_loss_clip": 0.01178108, + "auxiliary_loss_mlp": 0.01033991, + "balance_loss_clip": 1.02513373, + "balance_loss_mlp": 1.02445412, + "epoch": 0.2812481212048338, + "flos": 12823629482880.0, + "grad_norm": 2.169640106505723, + "language_loss": 0.75025231, + "learning_rate": 3.3737209119471405e-06, + "loss": 0.77237332, + "num_input_tokens_seen": 50167700, + "step": 2339, + "time_per_iteration": 2.627837896347046 + }, + { + "auxiliary_loss_clip": 0.01185147, + "auxiliary_loss_mlp": 0.01036419, + "balance_loss_clip": 1.02434468, + "balance_loss_mlp": 1.02615547, + "epoch": 0.28136836409547283, + "flos": 15633064765440.0, + "grad_norm": 2.224085999193327, + "language_loss": 0.63325995, + "learning_rate": 3.373154659492306e-06, + "loss": 0.65547562, + "num_input_tokens_seen": 50185840, + "step": 2340, + "time_per_iteration": 2.6350910663604736 + }, + { + "auxiliary_loss_clip": 0.01157759, + "auxiliary_loss_mlp": 0.01035891, + "balance_loss_clip": 0.98405731, + "balance_loss_mlp": 1.02675915, + "epoch": 0.28148860698611194, + "flos": 19933726106880.0, + "grad_norm": 1.9283645785268697, + "language_loss": 0.85130715, + "learning_rate": 3.3725881987310016e-06, + "loss": 0.87324369, + "num_input_tokens_seen": 50203375, + "step": 2341, + "time_per_iteration": 2.669426202774048 + }, + { + "auxiliary_loss_clip": 0.01152032, + "auxiliary_loss_mlp": 0.01033366, + "balance_loss_clip": 0.98200268, + "balance_loss_mlp": 1.02416849, + "epoch": 0.28160884987675106, + "flos": 17457219008640.0, + "grad_norm": 1.9309601027063448, + "language_loss": 0.87849909, + "learning_rate": 3.372021529749159e-06, + "loss": 0.90035307, + "num_input_tokens_seen": 50222435, + "step": 2342, + "time_per_iteration": 2.6810996532440186 + }, + { + "auxiliary_loss_clip": 0.0108465, + "auxiliary_loss_mlp": 0.01033179, + "balance_loss_clip": 0.86865115, + "balance_loss_mlp": 1.02458358, + "epoch": 0.2817290927673901, + "flos": 16834743290880.0, + "grad_norm": 1.9131503214991898, + "language_loss": 0.92143202, + "learning_rate": 3.3714546526327405e-06, + "loss": 0.94261026, + "num_input_tokens_seen": 50240435, + "step": 2343, + "time_per_iteration": 2.719780921936035 + }, + { + "auxiliary_loss_clip": 0.01135547, + "auxiliary_loss_mlp": 0.01035453, + "balance_loss_clip": 0.94312024, + "balance_loss_mlp": 1.02585661, + "epoch": 0.2818493356580292, + "flos": 15414081500160.0, + "grad_norm": 2.0925704170624155, + "language_loss": 0.87729192, + "learning_rate": 3.3708875674677423e-06, + "loss": 0.89900196, + "num_input_tokens_seen": 50258410, + "step": 2344, + "time_per_iteration": 2.664454936981201 + }, + { + "auxiliary_loss_clip": 0.01162986, + "auxiliary_loss_mlp": 0.01035436, + "balance_loss_clip": 0.98602581, + "balance_loss_mlp": 1.02537453, + "epoch": 0.28196957854866833, + "flos": 20412451595520.0, + "grad_norm": 2.073748686802549, + "language_loss": 0.83600092, + "learning_rate": 3.37032027434019e-06, + "loss": 0.8579852, + "num_input_tokens_seen": 50277930, + "step": 2345, + "time_per_iteration": 2.720736503601074 + }, + { + "auxiliary_loss_clip": 0.0118656, + "auxiliary_loss_mlp": 0.0103824, + "balance_loss_clip": 1.02125227, + "balance_loss_mlp": 1.02787471, + "epoch": 0.2820898214393074, + "flos": 19973120348160.0, + "grad_norm": 2.0824019221219574, + "language_loss": 0.83195257, + "learning_rate": 3.369752773336141e-06, + "loss": 0.8542006, + "num_input_tokens_seen": 50297410, + "step": 2346, + "time_per_iteration": 2.7543981075286865 + }, + { + "auxiliary_loss_clip": 0.01153192, + "auxiliary_loss_mlp": 0.01037115, + "balance_loss_clip": 0.98094249, + "balance_loss_mlp": 1.02686262, + "epoch": 0.2822100643299465, + "flos": 22528308188160.0, + "grad_norm": 2.1844362189049584, + "language_loss": 0.77978379, + "learning_rate": 3.3691850645416864e-06, + "loss": 0.80168682, + "num_input_tokens_seen": 50317120, + "step": 2347, + "time_per_iteration": 2.804042339324951 + }, + { + "auxiliary_loss_clip": 0.01183373, + "auxiliary_loss_mlp": 0.01034993, + "balance_loss_clip": 1.02252746, + "balance_loss_mlp": 1.02534282, + "epoch": 0.2823303072205856, + "flos": 11546682007680.0, + "grad_norm": 2.06889063574699, + "language_loss": 0.82557631, + "learning_rate": 3.368617148042945e-06, + "loss": 0.84775996, + "num_input_tokens_seen": 50334790, + "step": 2348, + "time_per_iteration": 2.6379876136779785 + }, + { + "auxiliary_loss_clip": 0.01147941, + "auxiliary_loss_mlp": 0.01030113, + "balance_loss_clip": 0.97862059, + "balance_loss_mlp": 1.02045095, + "epoch": 0.28245055011122466, + "flos": 18259894281600.0, + "grad_norm": 1.767070860236349, + "language_loss": 0.84506953, + "learning_rate": 3.368049023926071e-06, + "loss": 0.86685002, + "num_input_tokens_seen": 50353785, + "step": 2349, + "time_per_iteration": 2.6618106365203857 + }, + { + "auxiliary_loss_clip": 0.0118081, + "auxiliary_loss_mlp": 0.01036728, + "balance_loss_clip": 1.02584302, + "balance_loss_mlp": 1.02720881, + "epoch": 0.2825707930018638, + "flos": 24608110504320.0, + "grad_norm": 2.251996404765416, + "language_loss": 0.83574033, + "learning_rate": 3.3674806922772476e-06, + "loss": 0.85791576, + "num_input_tokens_seen": 50374670, + "step": 2350, + "time_per_iteration": 2.7145884037017822 + }, + { + "auxiliary_loss_clip": 0.01136513, + "auxiliary_loss_mlp": 0.01032379, + "balance_loss_clip": 0.94430804, + "balance_loss_mlp": 1.02323532, + "epoch": 0.28269103589250283, + "flos": 25226994862080.0, + "grad_norm": 1.6948935154264422, + "language_loss": 0.74872684, + "learning_rate": 3.3669121531826904e-06, + "loss": 0.77041578, + "num_input_tokens_seen": 50395650, + "step": 2351, + "time_per_iteration": 2.7491559982299805 + }, + { + "auxiliary_loss_clip": 0.01127816, + "auxiliary_loss_mlp": 0.01032454, + "balance_loss_clip": 0.94712353, + "balance_loss_mlp": 1.02291083, + "epoch": 0.28281127878314194, + "flos": 19281552819840.0, + "grad_norm": 2.061739337448905, + "language_loss": 0.83178532, + "learning_rate": 3.366343406728647e-06, + "loss": 0.85338807, + "num_input_tokens_seen": 50415100, + "step": 2352, + "time_per_iteration": 2.739589214324951 + }, + { + "auxiliary_loss_clip": 0.01169353, + "auxiliary_loss_mlp": 0.01042379, + "balance_loss_clip": 1.01745009, + "balance_loss_mlp": 1.0326755, + "epoch": 0.28293152167378105, + "flos": 23878405710720.0, + "grad_norm": 1.7065981640126846, + "language_loss": 0.68709129, + "learning_rate": 3.3657744530013946e-06, + "loss": 0.70920861, + "num_input_tokens_seen": 50434335, + "step": 2353, + "time_per_iteration": 2.7139225006103516 + }, + { + "auxiliary_loss_clip": 0.01183254, + "auxiliary_loss_mlp": 0.01037926, + "balance_loss_clip": 1.02338517, + "balance_loss_mlp": 1.02837193, + "epoch": 0.2830517645644201, + "flos": 43866965928960.0, + "grad_norm": 2.4142294553122965, + "language_loss": 0.71216738, + "learning_rate": 3.3652052920872437e-06, + "loss": 0.73437917, + "num_input_tokens_seen": 50457200, + "step": 2354, + "time_per_iteration": 2.871835708618164 + }, + { + "auxiliary_loss_clip": 0.01155441, + "auxiliary_loss_mlp": 0.01041192, + "balance_loss_clip": 0.98292303, + "balance_loss_mlp": 1.03168488, + "epoch": 0.2831720074550592, + "flos": 26651750803200.0, + "grad_norm": 2.4866698161364456, + "language_loss": 0.85701823, + "learning_rate": 3.3646359240725355e-06, + "loss": 0.87898457, + "num_input_tokens_seen": 50476390, + "step": 2355, + "time_per_iteration": 3.7421798706054688 + }, + { + "auxiliary_loss_clip": 0.01174757, + "auxiliary_loss_mlp": 0.00512385, + "balance_loss_clip": 1.02288222, + "balance_loss_mlp": 0.0, + "epoch": 0.2832922503456983, + "flos": 31029979564800.0, + "grad_norm": 2.6632088078841365, + "language_loss": 0.67263812, + "learning_rate": 3.364066349043643e-06, + "loss": 0.68950951, + "num_input_tokens_seen": 50497595, + "step": 2356, + "time_per_iteration": 2.774895429611206 + }, + { + "auxiliary_loss_clip": 0.01150633, + "auxiliary_loss_mlp": 0.01032901, + "balance_loss_clip": 0.98144335, + "balance_loss_mlp": 1.02416301, + "epoch": 0.2834124932363374, + "flos": 20405699838720.0, + "grad_norm": 1.9053379598167306, + "language_loss": 0.82259536, + "learning_rate": 3.363496567086969e-06, + "loss": 0.8444308, + "num_input_tokens_seen": 50514690, + "step": 2357, + "time_per_iteration": 3.974924325942993 + }, + { + "auxiliary_loss_clip": 0.01202422, + "auxiliary_loss_mlp": 0.01028714, + "balance_loss_clip": 1.06110358, + "balance_loss_mlp": 1.01889753, + "epoch": 0.2835327361269765, + "flos": 39384848056320.0, + "grad_norm": 2.21002641464807, + "language_loss": 0.75568599, + "learning_rate": 3.3629265782889506e-06, + "loss": 0.77799737, + "num_input_tokens_seen": 50536515, + "step": 2358, + "time_per_iteration": 2.866523504257202 + }, + { + "auxiliary_loss_clip": 0.01126732, + "auxiliary_loss_mlp": 0.01036688, + "balance_loss_clip": 0.94095606, + "balance_loss_mlp": 1.02716947, + "epoch": 0.2836529790176156, + "flos": 30261598801920.0, + "grad_norm": 1.9963334485016986, + "language_loss": 0.71857071, + "learning_rate": 3.362356382736054e-06, + "loss": 0.74020493, + "num_input_tokens_seen": 50557120, + "step": 2359, + "time_per_iteration": 3.885031223297119 + }, + { + "auxiliary_loss_clip": 0.01127947, + "auxiliary_loss_mlp": 0.01031942, + "balance_loss_clip": 0.93953198, + "balance_loss_mlp": 1.0222863, + "epoch": 0.28377322190825466, + "flos": 12677796264960.0, + "grad_norm": 2.2204579690972626, + "language_loss": 0.91234249, + "learning_rate": 3.361785980514777e-06, + "loss": 0.93394136, + "num_input_tokens_seen": 50573320, + "step": 2360, + "time_per_iteration": 2.7237327098846436 + }, + { + "auxiliary_loss_clip": 0.01074586, + "auxiliary_loss_mlp": 0.01033025, + "balance_loss_clip": 0.86636847, + "balance_loss_mlp": 1.02316689, + "epoch": 0.28389346479889377, + "flos": 18296666830080.0, + "grad_norm": 2.1895917142384036, + "language_loss": 0.76509881, + "learning_rate": 3.361215371711649e-06, + "loss": 0.78617489, + "num_input_tokens_seen": 50592415, + "step": 2361, + "time_per_iteration": 2.7394609451293945 + }, + { + "auxiliary_loss_clip": 0.01126773, + "auxiliary_loss_mlp": 0.0102785, + "balance_loss_clip": 0.9438383, + "balance_loss_mlp": 1.01824117, + "epoch": 0.2840137076895329, + "flos": 20406992728320.0, + "grad_norm": 2.6357564838563743, + "language_loss": 0.83263868, + "learning_rate": 3.3606445564132326e-06, + "loss": 0.85418487, + "num_input_tokens_seen": 50609710, + "step": 2362, + "time_per_iteration": 3.616109609603882 + }, + { + "auxiliary_loss_clip": 0.01205254, + "auxiliary_loss_mlp": 0.00511289, + "balance_loss_clip": 1.06321979, + "balance_loss_mlp": 0.0, + "epoch": 0.28413395058017193, + "flos": 20048030161920.0, + "grad_norm": 2.4093625874640985, + "language_loss": 0.81845999, + "learning_rate": 3.360073534706118e-06, + "loss": 0.83562541, + "num_input_tokens_seen": 50626865, + "step": 2363, + "time_per_iteration": 2.6614091396331787 + }, + { + "auxiliary_loss_clip": 0.01155672, + "auxiliary_loss_mlp": 0.01043067, + "balance_loss_clip": 0.98424494, + "balance_loss_mlp": 1.03365588, + "epoch": 0.28425419347081105, + "flos": 37663613256960.0, + "grad_norm": 1.9314149220344305, + "language_loss": 0.7558229, + "learning_rate": 3.35950230667693e-06, + "loss": 0.77781028, + "num_input_tokens_seen": 50648560, + "step": 2364, + "time_per_iteration": 2.8169498443603516 + }, + { + "auxiliary_loss_clip": 0.01180513, + "auxiliary_loss_mlp": 0.01032741, + "balance_loss_clip": 1.02129281, + "balance_loss_mlp": 1.02324009, + "epoch": 0.28437443636145016, + "flos": 13845072539520.0, + "grad_norm": 2.587034830915436, + "language_loss": 0.85527855, + "learning_rate": 3.358930872412323e-06, + "loss": 0.87741107, + "num_input_tokens_seen": 50665725, + "step": 2365, + "time_per_iteration": 2.644144058227539 + }, + { + "auxiliary_loss_clip": 0.01178395, + "auxiliary_loss_mlp": 0.01030625, + "balance_loss_clip": 1.02331853, + "balance_loss_mlp": 1.0212549, + "epoch": 0.2844946792520892, + "flos": 22747794243840.0, + "grad_norm": 2.7985024364376723, + "language_loss": 0.80978185, + "learning_rate": 3.3583592319989825e-06, + "loss": 0.83187205, + "num_input_tokens_seen": 50685095, + "step": 2366, + "time_per_iteration": 2.711735486984253 + }, + { + "auxiliary_loss_clip": 0.01186425, + "auxiliary_loss_mlp": 0.01036498, + "balance_loss_clip": 1.02365184, + "balance_loss_mlp": 1.02649665, + "epoch": 0.2846149221427283, + "flos": 32415987709440.0, + "grad_norm": 2.3794881294466963, + "language_loss": 0.69070792, + "learning_rate": 3.357787385523627e-06, + "loss": 0.71293724, + "num_input_tokens_seen": 50706500, + "step": 2367, + "time_per_iteration": 2.720097780227661 + }, + { + "auxiliary_loss_clip": 0.01090458, + "auxiliary_loss_mlp": 0.01029128, + "balance_loss_clip": 0.86760139, + "balance_loss_mlp": 1.01979375, + "epoch": 0.2847351650333674, + "flos": 28475976873600.0, + "grad_norm": 3.9364011412038096, + "language_loss": 0.83069837, + "learning_rate": 3.3572153330730048e-06, + "loss": 0.8518942, + "num_input_tokens_seen": 50727595, + "step": 2368, + "time_per_iteration": 2.795271396636963 + }, + { + "auxiliary_loss_clip": 0.01039514, + "auxiliary_loss_mlp": 0.01008073, + "balance_loss_clip": 0.91408223, + "balance_loss_mlp": 1.00634408, + "epoch": 0.2848554079240065, + "flos": 55753399704960.0, + "grad_norm": 0.8242316045996081, + "language_loss": 0.64702749, + "learning_rate": 3.3566430747338956e-06, + "loss": 0.66750348, + "num_input_tokens_seen": 50782800, + "step": 2369, + "time_per_iteration": 3.140369415283203 + }, + { + "auxiliary_loss_clip": 0.01181837, + "auxiliary_loss_mlp": 0.01030566, + "balance_loss_clip": 1.02204609, + "balance_loss_mlp": 1.02127361, + "epoch": 0.2849756508146456, + "flos": 11836875985920.0, + "grad_norm": 1.9546569195031152, + "language_loss": 0.86143881, + "learning_rate": 3.35607061059311e-06, + "loss": 0.8835628, + "num_input_tokens_seen": 50797730, + "step": 2370, + "time_per_iteration": 2.655968189239502 + }, + { + "auxiliary_loss_clip": 0.01204039, + "auxiliary_loss_mlp": 0.01041707, + "balance_loss_clip": 1.06325555, + "balance_loss_mlp": 1.03242016, + "epoch": 0.28509589370528465, + "flos": 25155209531520.0, + "grad_norm": 2.7671930612677986, + "language_loss": 0.74940199, + "learning_rate": 3.3554979407374917e-06, + "loss": 0.77185947, + "num_input_tokens_seen": 50819840, + "step": 2371, + "time_per_iteration": 2.7242093086242676 + }, + { + "auxiliary_loss_clip": 0.01180611, + "auxiliary_loss_mlp": 0.01028539, + "balance_loss_clip": 1.02223921, + "balance_loss_mlp": 1.01950347, + "epoch": 0.28521613659592376, + "flos": 19974808287360.0, + "grad_norm": 1.6286878862687293, + "language_loss": 0.73586047, + "learning_rate": 3.3549250652539134e-06, + "loss": 0.75795203, + "num_input_tokens_seen": 50838935, + "step": 2372, + "time_per_iteration": 2.674057960510254 + }, + { + "auxiliary_loss_clip": 0.01154339, + "auxiliary_loss_mlp": 0.01035872, + "balance_loss_clip": 0.98129642, + "balance_loss_mlp": 1.025769, + "epoch": 0.2853363794865629, + "flos": 23367971491200.0, + "grad_norm": 2.483937859877285, + "language_loss": 0.81965035, + "learning_rate": 3.3543519842292794e-06, + "loss": 0.8415525, + "num_input_tokens_seen": 50858590, + "step": 2373, + "time_per_iteration": 2.7063872814178467 + }, + { + "auxiliary_loss_clip": 0.01200589, + "auxiliary_loss_mlp": 0.00512273, + "balance_loss_clip": 1.05925131, + "balance_loss_mlp": 0.0, + "epoch": 0.28545662237720193, + "flos": 19861940776320.0, + "grad_norm": 2.142069578234902, + "language_loss": 0.83742344, + "learning_rate": 3.353778697750527e-06, + "loss": 0.85455209, + "num_input_tokens_seen": 50876995, + "step": 2374, + "time_per_iteration": 2.6215786933898926 + }, + { + "auxiliary_loss_clip": 0.01145902, + "auxiliary_loss_mlp": 0.01033552, + "balance_loss_clip": 0.98034149, + "balance_loss_mlp": 1.02452826, + "epoch": 0.28557686526784104, + "flos": 23879016241920.0, + "grad_norm": 1.715021709625712, + "language_loss": 0.89108175, + "learning_rate": 3.353205205904622e-06, + "loss": 0.91287625, + "num_input_tokens_seen": 50896105, + "step": 2375, + "time_per_iteration": 2.6787586212158203 + }, + { + "auxiliary_loss_clip": 0.01158715, + "auxiliary_loss_mlp": 0.010292, + "balance_loss_clip": 0.98679477, + "balance_loss_mlp": 1.0196867, + "epoch": 0.28569710815848015, + "flos": 44890384233600.0, + "grad_norm": 1.7783213412423688, + "language_loss": 0.71782553, + "learning_rate": 3.3526315087785637e-06, + "loss": 0.73970461, + "num_input_tokens_seen": 50917220, + "step": 2376, + "time_per_iteration": 2.8827221393585205 + }, + { + "auxiliary_loss_clip": 0.01091807, + "auxiliary_loss_mlp": 0.01031983, + "balance_loss_clip": 0.90477371, + "balance_loss_mlp": 1.02235079, + "epoch": 0.2858173510491192, + "flos": 26829759628800.0, + "grad_norm": 1.6646661934208196, + "language_loss": 0.80669945, + "learning_rate": 3.3520576064593805e-06, + "loss": 0.82793736, + "num_input_tokens_seen": 50937175, + "step": 2377, + "time_per_iteration": 2.761261224746704 + }, + { + "auxiliary_loss_clip": 0.01185628, + "auxiliary_loss_mlp": 0.0103722, + "balance_loss_clip": 1.02516735, + "balance_loss_mlp": 1.02739072, + "epoch": 0.2859375939397583, + "flos": 23148916398720.0, + "grad_norm": 1.4903043860340377, + "language_loss": 0.81551909, + "learning_rate": 3.3514834990341337e-06, + "loss": 0.83774751, + "num_input_tokens_seen": 50957500, + "step": 2378, + "time_per_iteration": 2.680523633956909 + }, + { + "auxiliary_loss_clip": 0.01163543, + "auxiliary_loss_mlp": 0.01029746, + "balance_loss_clip": 0.98579615, + "balance_loss_mlp": 1.02088284, + "epoch": 0.2860578368303974, + "flos": 12129799397760.0, + "grad_norm": 3.5587243192101297, + "language_loss": 0.93088984, + "learning_rate": 3.3509091865899144e-06, + "loss": 0.9528228, + "num_input_tokens_seen": 50972690, + "step": 2379, + "time_per_iteration": 2.6977627277374268 + }, + { + "auxiliary_loss_clip": 0.01202537, + "auxiliary_loss_mlp": 0.01031727, + "balance_loss_clip": 1.05901933, + "balance_loss_mlp": 1.02198768, + "epoch": 0.2861780797210365, + "flos": 19938035738880.0, + "grad_norm": 7.725883932867667, + "language_loss": 0.70707619, + "learning_rate": 3.350334669213846e-06, + "loss": 0.72941881, + "num_input_tokens_seen": 50990095, + "step": 2380, + "time_per_iteration": 2.6621785163879395 + }, + { + "auxiliary_loss_clip": 0.01181662, + "auxiliary_loss_mlp": 0.01031909, + "balance_loss_clip": 1.02639151, + "balance_loss_mlp": 1.02267075, + "epoch": 0.2862983226116756, + "flos": 27563127609600.0, + "grad_norm": 2.394328495288441, + "language_loss": 0.75737154, + "learning_rate": 3.3497599469930816e-06, + "loss": 0.77950728, + "num_input_tokens_seen": 51008305, + "step": 2381, + "time_per_iteration": 3.614126682281494 + }, + { + "auxiliary_loss_clip": 0.01208818, + "auxiliary_loss_mlp": 0.01035686, + "balance_loss_clip": 1.06382287, + "balance_loss_mlp": 1.02616751, + "epoch": 0.28641856550231465, + "flos": 22053964158720.0, + "grad_norm": 4.102286031963183, + "language_loss": 0.83294368, + "learning_rate": 3.349185020014807e-06, + "loss": 0.85538876, + "num_input_tokens_seen": 51025570, + "step": 2382, + "time_per_iteration": 2.600759267807007 + }, + { + "auxiliary_loss_clip": 0.01181146, + "auxiliary_loss_mlp": 0.01030746, + "balance_loss_clip": 1.02175403, + "balance_loss_mlp": 1.02073288, + "epoch": 0.28653880839295376, + "flos": 22378775869440.0, + "grad_norm": 3.479313924353353, + "language_loss": 0.74839276, + "learning_rate": 3.348609888366237e-06, + "loss": 0.77051169, + "num_input_tokens_seen": 51044585, + "step": 2383, + "time_per_iteration": 3.638139486312866 + }, + { + "auxiliary_loss_clip": 0.01078157, + "auxiliary_loss_mlp": 0.01026553, + "balance_loss_clip": 0.86681449, + "balance_loss_mlp": 1.01736176, + "epoch": 0.28665905128359287, + "flos": 23367971491200.0, + "grad_norm": 2.9920693900267845, + "language_loss": 0.63079369, + "learning_rate": 3.348034552134619e-06, + "loss": 0.65184075, + "num_input_tokens_seen": 51063990, + "step": 2384, + "time_per_iteration": 2.781399965286255 + }, + { + "auxiliary_loss_clip": 0.01102129, + "auxiliary_loss_mlp": 0.0103617, + "balance_loss_clip": 0.90702802, + "balance_loss_mlp": 1.02737856, + "epoch": 0.2867792941742319, + "flos": 20881695893760.0, + "grad_norm": 2.005928007294476, + "language_loss": 0.84982121, + "learning_rate": 3.3474590114072316e-06, + "loss": 0.87120414, + "num_input_tokens_seen": 51081990, + "step": 2385, + "time_per_iteration": 2.7470991611480713 + }, + { + "auxiliary_loss_clip": 0.01128656, + "auxiliary_loss_mlp": 0.01029365, + "balance_loss_clip": 0.94688451, + "balance_loss_mlp": 1.01959634, + "epoch": 0.28689953706487104, + "flos": 20664005518080.0, + "grad_norm": 2.2537320153396245, + "language_loss": 0.83165514, + "learning_rate": 3.3468832662713836e-06, + "loss": 0.85323536, + "num_input_tokens_seen": 51100235, + "step": 2386, + "time_per_iteration": 3.6752829551696777 + }, + { + "auxiliary_loss_clip": 0.0112968, + "auxiliary_loss_mlp": 0.01035967, + "balance_loss_clip": 0.94549179, + "balance_loss_mlp": 1.02648354, + "epoch": 0.28701977995551015, + "flos": 12675533708160.0, + "grad_norm": 2.5456478928532444, + "language_loss": 0.8419224, + "learning_rate": 3.346307316814415e-06, + "loss": 0.8635788, + "num_input_tokens_seen": 51115405, + "step": 2387, + "time_per_iteration": 2.704244375228882 + }, + { + "auxiliary_loss_clip": 0.01182534, + "auxiliary_loss_mlp": 0.01031836, + "balance_loss_clip": 1.02644312, + "balance_loss_mlp": 1.02194118, + "epoch": 0.2871400228461492, + "flos": 21252366293760.0, + "grad_norm": 2.9534107139090144, + "language_loss": 0.75501949, + "learning_rate": 3.3457311631236965e-06, + "loss": 0.77716321, + "num_input_tokens_seen": 51136390, + "step": 2388, + "time_per_iteration": 3.5654332637786865 + }, + { + "auxiliary_loss_clip": 0.01145904, + "auxiliary_loss_mlp": 0.0103012, + "balance_loss_clip": 0.98233664, + "balance_loss_mlp": 1.02001095, + "epoch": 0.2872602657367883, + "flos": 25119262995840.0, + "grad_norm": 2.4469743871517813, + "language_loss": 0.84484816, + "learning_rate": 3.345154805286631e-06, + "loss": 0.86660838, + "num_input_tokens_seen": 51156650, + "step": 2389, + "time_per_iteration": 2.7079966068267822 + }, + { + "auxiliary_loss_clip": 0.01175218, + "auxiliary_loss_mlp": 0.01036195, + "balance_loss_clip": 1.02240252, + "balance_loss_mlp": 1.02647352, + "epoch": 0.2873805086274274, + "flos": 16646606830080.0, + "grad_norm": 5.406570685614938, + "language_loss": 0.76474178, + "learning_rate": 3.344578243390651e-06, + "loss": 0.78685594, + "num_input_tokens_seen": 51172210, + "step": 2390, + "time_per_iteration": 2.6387243270874023 + }, + { + "auxiliary_loss_clip": 0.01155307, + "auxiliary_loss_mlp": 0.0103208, + "balance_loss_clip": 0.9867695, + "balance_loss_mlp": 1.02274036, + "epoch": 0.2875007515180665, + "flos": 17420123237760.0, + "grad_norm": 2.2845869133531247, + "language_loss": 0.78658295, + "learning_rate": 3.3440014775232206e-06, + "loss": 0.8084569, + "num_input_tokens_seen": 51190265, + "step": 2391, + "time_per_iteration": 2.682898759841919 + }, + { + "auxiliary_loss_clip": 0.01135378, + "auxiliary_loss_mlp": 0.01030034, + "balance_loss_clip": 0.94630784, + "balance_loss_mlp": 1.02101624, + "epoch": 0.2876209944087056, + "flos": 23434190213760.0, + "grad_norm": 2.120110837322712, + "language_loss": 0.71036804, + "learning_rate": 3.343424507771834e-06, + "loss": 0.73202217, + "num_input_tokens_seen": 51208475, + "step": 2392, + "time_per_iteration": 2.7108848094940186 + }, + { + "auxiliary_loss_clip": 0.01132626, + "auxiliary_loss_mlp": 0.01033423, + "balance_loss_clip": 0.94583058, + "balance_loss_mlp": 1.02374911, + "epoch": 0.2877412372993447, + "flos": 13735509079680.0, + "grad_norm": 1.8432919428188792, + "language_loss": 0.86378276, + "learning_rate": 3.342847334224018e-06, + "loss": 0.88544321, + "num_input_tokens_seen": 51225875, + "step": 2393, + "time_per_iteration": 2.728879451751709 + }, + { + "auxiliary_loss_clip": 0.01082795, + "auxiliary_loss_mlp": 0.01002902, + "balance_loss_clip": 0.99423063, + "balance_loss_mlp": 1.00116146, + "epoch": 0.28786148018998375, + "flos": 58079695104000.0, + "grad_norm": 0.9750932226481607, + "language_loss": 0.62374115, + "learning_rate": 3.342269956967329e-06, + "loss": 0.64459813, + "num_input_tokens_seen": 51287780, + "step": 2394, + "time_per_iteration": 3.460402488708496 + }, + { + "auxiliary_loss_clip": 0.01182581, + "auxiliary_loss_mlp": 0.01028888, + "balance_loss_clip": 1.02347612, + "balance_loss_mlp": 1.01823115, + "epoch": 0.28798172308062286, + "flos": 23435052140160.0, + "grad_norm": 2.688543557088662, + "language_loss": 0.71569669, + "learning_rate": 3.341692376089355e-06, + "loss": 0.73781139, + "num_input_tokens_seen": 51303335, + "step": 2395, + "time_per_iteration": 2.68639874458313 + }, + { + "auxiliary_loss_clip": 0.01179704, + "auxiliary_loss_mlp": 0.01034274, + "balance_loss_clip": 1.02443838, + "balance_loss_mlp": 1.0244807, + "epoch": 0.288101965971262, + "flos": 25110033200640.0, + "grad_norm": 3.173080138471174, + "language_loss": 0.84331036, + "learning_rate": 3.3411145916777146e-06, + "loss": 0.86545014, + "num_input_tokens_seen": 51317495, + "step": 2396, + "time_per_iteration": 0.03810405731201172 + }, + { + "auxiliary_loss_clip": 0.01150707, + "auxiliary_loss_mlp": 0.01031105, + "balance_loss_clip": 0.982283, + "balance_loss_mlp": 1.02117538, + "epoch": 0.28822220886190103, + "flos": 16252559654400.0, + "grad_norm": 2.238512771030755, + "language_loss": 0.91233027, + "learning_rate": 3.3405366038200566e-06, + "loss": 0.93414831, + "num_input_tokens_seen": 51336430, + "step": 2397, + "time_per_iteration": 2.8137245178222656 + }, + { + "auxiliary_loss_clip": 0.01158636, + "auxiliary_loss_mlp": 0.01041182, + "balance_loss_clip": 0.98844635, + "balance_loss_mlp": 1.03157961, + "epoch": 0.28834245175254014, + "flos": 24535642815360.0, + "grad_norm": 2.9116147876025167, + "language_loss": 0.84650934, + "learning_rate": 3.3399584126040617e-06, + "loss": 0.8685075, + "num_input_tokens_seen": 51355930, + "step": 2398, + "time_per_iteration": 2.7871017456054688 + }, + { + "auxiliary_loss_clip": 0.01206633, + "auxiliary_loss_mlp": 0.0051229, + "balance_loss_clip": 1.0630089, + "balance_loss_mlp": 0.0, + "epoch": 0.2884626946431792, + "flos": 24571445696640.0, + "grad_norm": 3.6888854469309877, + "language_loss": 0.90575457, + "learning_rate": 3.339380018117441e-06, + "loss": 0.92294383, + "num_input_tokens_seen": 51376765, + "step": 2399, + "time_per_iteration": 3.0787997245788574 + }, + { + "auxiliary_loss_clip": 0.01178683, + "auxiliary_loss_mlp": 0.01035896, + "balance_loss_clip": 1.02533209, + "balance_loss_mlp": 1.02638316, + "epoch": 0.2885829375338183, + "flos": 16544657053440.0, + "grad_norm": 4.527884769317718, + "language_loss": 0.7815088, + "learning_rate": 3.3388014204479366e-06, + "loss": 0.80365461, + "num_input_tokens_seen": 51394570, + "step": 2400, + "time_per_iteration": 2.7131307125091553 + }, + { + "auxiliary_loss_clip": 0.01207645, + "auxiliary_loss_mlp": 0.01033891, + "balance_loss_clip": 1.06222522, + "balance_loss_mlp": 1.0233531, + "epoch": 0.2887031804244574, + "flos": 24061226958720.0, + "grad_norm": 2.014607682562286, + "language_loss": 0.9135201, + "learning_rate": 3.338222619683321e-06, + "loss": 0.9359355, + "num_input_tokens_seen": 51414535, + "step": 2401, + "time_per_iteration": 2.631462335586548 + }, + { + "auxiliary_loss_clip": 0.01158912, + "auxiliary_loss_mlp": 0.01039449, + "balance_loss_clip": 0.98514068, + "balance_loss_mlp": 1.02976274, + "epoch": 0.2888234233150965, + "flos": 23330696152320.0, + "grad_norm": 2.4399785397909555, + "language_loss": 0.73364973, + "learning_rate": 3.337643615911398e-06, + "loss": 0.75563335, + "num_input_tokens_seen": 51434160, + "step": 2402, + "time_per_iteration": 2.891157388687134 + }, + { + "auxiliary_loss_clip": 0.01181364, + "auxiliary_loss_mlp": 0.01028579, + "balance_loss_clip": 1.02166724, + "balance_loss_mlp": 1.01805246, + "epoch": 0.2889436662057356, + "flos": 22272767856000.0, + "grad_norm": 1.8711069226120904, + "language_loss": 0.78792608, + "learning_rate": 3.3370644092200026e-06, + "loss": 0.81002545, + "num_input_tokens_seen": 51451435, + "step": 2403, + "time_per_iteration": 2.728400707244873 + }, + { + "auxiliary_loss_clip": 0.01119394, + "auxiliary_loss_mlp": 0.01030982, + "balance_loss_clip": 0.94062203, + "balance_loss_mlp": 1.02142704, + "epoch": 0.2890639090963747, + "flos": 21616931381760.0, + "grad_norm": 2.100920810480868, + "language_loss": 0.77925444, + "learning_rate": 3.3364849996969985e-06, + "loss": 0.80075812, + "num_input_tokens_seen": 51471455, + "step": 2404, + "time_per_iteration": 2.774729013442993 + }, + { + "auxiliary_loss_clip": 0.01181277, + "auxiliary_loss_mlp": 0.01036792, + "balance_loss_clip": 1.02590621, + "balance_loss_mlp": 1.02766013, + "epoch": 0.28918415198701375, + "flos": 28585540333440.0, + "grad_norm": 2.59403452566761, + "language_loss": 0.85307139, + "learning_rate": 3.335905387430283e-06, + "loss": 0.87525207, + "num_input_tokens_seen": 51492890, + "step": 2405, + "time_per_iteration": 2.7706503868103027 + }, + { + "auxiliary_loss_clip": 0.01158806, + "auxiliary_loss_mlp": 0.01036223, + "balance_loss_clip": 0.98220938, + "balance_loss_mlp": 1.0257268, + "epoch": 0.28930439487765286, + "flos": 21944688007680.0, + "grad_norm": 1.9433717546793068, + "language_loss": 0.82963371, + "learning_rate": 3.335325572507782e-06, + "loss": 0.85158396, + "num_input_tokens_seen": 51513390, + "step": 2406, + "time_per_iteration": 2.811802864074707 + }, + { + "auxiliary_loss_clip": 0.01209068, + "auxiliary_loss_mlp": 0.00512414, + "balance_loss_clip": 1.06591856, + "balance_loss_mlp": 0.0, + "epoch": 0.28942463776829197, + "flos": 19281911955840.0, + "grad_norm": 1.6401634510900975, + "language_loss": 0.73983967, + "learning_rate": 3.3347455550174537e-06, + "loss": 0.75705451, + "num_input_tokens_seen": 51532730, + "step": 2407, + "time_per_iteration": 3.610093116760254 + }, + { + "auxiliary_loss_clip": 0.01127941, + "auxiliary_loss_mlp": 0.01039634, + "balance_loss_clip": 0.94249499, + "balance_loss_mlp": 1.02960837, + "epoch": 0.289544880658931, + "flos": 14645700737280.0, + "grad_norm": 1.9619729725337545, + "language_loss": 0.68246651, + "learning_rate": 3.3341653350472864e-06, + "loss": 0.70414221, + "num_input_tokens_seen": 51549560, + "step": 2408, + "time_per_iteration": 2.6912732124328613 + }, + { + "auxiliary_loss_clip": 0.01211447, + "auxiliary_loss_mlp": 0.01036599, + "balance_loss_clip": 1.06316936, + "balance_loss_mlp": 1.02556634, + "epoch": 0.28966512354957014, + "flos": 28621881918720.0, + "grad_norm": 5.073772855321429, + "language_loss": 0.69075501, + "learning_rate": 3.333584912685298e-06, + "loss": 0.7132355, + "num_input_tokens_seen": 51568180, + "step": 2409, + "time_per_iteration": 3.585017681121826 + }, + { + "auxiliary_loss_clip": 0.01040653, + "auxiliary_loss_mlp": 0.01001609, + "balance_loss_clip": 0.91793299, + "balance_loss_mlp": 0.99951124, + "epoch": 0.28978536644020925, + "flos": 64711784511360.0, + "grad_norm": 0.8668142048181812, + "language_loss": 0.5546453, + "learning_rate": 3.3330042880195385e-06, + "loss": 0.57506788, + "num_input_tokens_seen": 51622530, + "step": 2410, + "time_per_iteration": 3.264509916305542 + }, + { + "auxiliary_loss_clip": 0.01153842, + "auxiliary_loss_mlp": 0.01035405, + "balance_loss_clip": 0.98169816, + "balance_loss_mlp": 1.02592766, + "epoch": 0.2899056093308483, + "flos": 18624638937600.0, + "grad_norm": 1.8343823989890777, + "language_loss": 0.78627795, + "learning_rate": 3.3324234611380888e-06, + "loss": 0.80817044, + "num_input_tokens_seen": 51641260, + "step": 2411, + "time_per_iteration": 3.681809425354004 + }, + { + "auxiliary_loss_clip": 0.0112764, + "auxiliary_loss_mlp": 0.01035589, + "balance_loss_clip": 0.9464823, + "balance_loss_mlp": 1.02613568, + "epoch": 0.2900258522214874, + "flos": 22893735202560.0, + "grad_norm": 2.13932600653874, + "language_loss": 0.8156836, + "learning_rate": 3.3318424321290596e-06, + "loss": 0.83731592, + "num_input_tokens_seen": 51660975, + "step": 2412, + "time_per_iteration": 2.7892117500305176 + }, + { + "auxiliary_loss_clip": 0.01040086, + "auxiliary_loss_mlp": 0.00999608, + "balance_loss_clip": 0.92003572, + "balance_loss_mlp": 0.9978199, + "epoch": 0.2901460951121265, + "flos": 71106036013440.0, + "grad_norm": 0.8320781766770011, + "language_loss": 0.59898889, + "learning_rate": 3.3312612010805917e-06, + "loss": 0.61938572, + "num_input_tokens_seen": 51720550, + "step": 2413, + "time_per_iteration": 3.3663406372070312 + }, + { + "auxiliary_loss_clip": 0.01144902, + "auxiliary_loss_mlp": 0.01031637, + "balance_loss_clip": 0.98200309, + "balance_loss_mlp": 1.02208805, + "epoch": 0.2902663380027656, + "flos": 32160986081280.0, + "grad_norm": 1.8314248783897906, + "language_loss": 0.70372915, + "learning_rate": 3.330679768080858e-06, + "loss": 0.72549456, + "num_input_tokens_seen": 51744435, + "step": 2414, + "time_per_iteration": 2.7459909915924072 + }, + { + "auxiliary_loss_clip": 0.01180625, + "auxiliary_loss_mlp": 0.01034351, + "balance_loss_clip": 1.02626801, + "balance_loss_mlp": 1.02490973, + "epoch": 0.2903865808934047, + "flos": 29351658539520.0, + "grad_norm": 2.58807208386507, + "language_loss": 0.8366338, + "learning_rate": 3.3300981332180627e-06, + "loss": 0.85878348, + "num_input_tokens_seen": 51763640, + "step": 2415, + "time_per_iteration": 3.6102774143218994 + }, + { + "auxiliary_loss_clip": 0.0113901, + "auxiliary_loss_mlp": 0.01028851, + "balance_loss_clip": 0.94647157, + "balance_loss_mlp": 1.01927853, + "epoch": 0.29050682378404374, + "flos": 17089026647040.0, + "grad_norm": 2.227875668495346, + "language_loss": 0.80038536, + "learning_rate": 3.3295162965804373e-06, + "loss": 0.82206392, + "num_input_tokens_seen": 51782135, + "step": 2416, + "time_per_iteration": 2.8725593090057373 + }, + { + "auxiliary_loss_clip": 0.01130475, + "auxiliary_loss_mlp": 0.01034907, + "balance_loss_clip": 0.94730949, + "balance_loss_mlp": 1.02570391, + "epoch": 0.29062706667468285, + "flos": 17858233422720.0, + "grad_norm": 3.938651644168407, + "language_loss": 0.78998488, + "learning_rate": 3.328934258256247e-06, + "loss": 0.81163871, + "num_input_tokens_seen": 51800200, + "step": 2417, + "time_per_iteration": 2.7422168254852295 + }, + { + "auxiliary_loss_clip": 0.01179209, + "auxiliary_loss_mlp": 0.01031614, + "balance_loss_clip": 1.02134848, + "balance_loss_mlp": 1.02195787, + "epoch": 0.29074730956532197, + "flos": 24279815174400.0, + "grad_norm": 2.4369302415104657, + "language_loss": 0.67503703, + "learning_rate": 3.3283520183337856e-06, + "loss": 0.69714522, + "num_input_tokens_seen": 51819905, + "step": 2418, + "time_per_iteration": 2.676454782485962 + }, + { + "auxiliary_loss_clip": 0.01154032, + "auxiliary_loss_mlp": 0.01033512, + "balance_loss_clip": 0.98368752, + "balance_loss_mlp": 1.02379608, + "epoch": 0.290867552455961, + "flos": 22340961826560.0, + "grad_norm": 2.1544692634848808, + "language_loss": 0.6899358, + "learning_rate": 3.3277695769013797e-06, + "loss": 0.7118113, + "num_input_tokens_seen": 51839350, + "step": 2419, + "time_per_iteration": 2.6708929538726807 + }, + { + "auxiliary_loss_clip": 0.0117945, + "auxiliary_loss_mlp": 0.01032516, + "balance_loss_clip": 1.02251911, + "balance_loss_mlp": 1.02241921, + "epoch": 0.29098779534660013, + "flos": 23186155824000.0, + "grad_norm": 1.954527825951996, + "language_loss": 0.77273494, + "learning_rate": 3.327186934047385e-06, + "loss": 0.79485458, + "num_input_tokens_seen": 51858045, + "step": 2420, + "time_per_iteration": 2.6739604473114014 + }, + { + "auxiliary_loss_clip": 0.01143977, + "auxiliary_loss_mlp": 0.01032427, + "balance_loss_clip": 0.97826117, + "balance_loss_mlp": 1.02222824, + "epoch": 0.29110803823723924, + "flos": 15304194817920.0, + "grad_norm": 7.27826025572425, + "language_loss": 0.65418851, + "learning_rate": 3.3266040898601877e-06, + "loss": 0.67595255, + "num_input_tokens_seen": 51875880, + "step": 2421, + "time_per_iteration": 2.6269876956939697 + }, + { + "auxiliary_loss_clip": 0.01101513, + "auxiliary_loss_mlp": 0.01046382, + "balance_loss_clip": 0.9046306, + "balance_loss_mlp": 1.03660703, + "epoch": 0.2912282811278783, + "flos": 22595352923520.0, + "grad_norm": 1.9395285571969438, + "language_loss": 0.77881646, + "learning_rate": 3.3260210444282045e-06, + "loss": 0.80029541, + "num_input_tokens_seen": 51893835, + "step": 2422, + "time_per_iteration": 2.7538905143737793 + }, + { + "auxiliary_loss_clip": 0.01173248, + "auxiliary_loss_mlp": 0.01030894, + "balance_loss_clip": 1.02215612, + "balance_loss_mlp": 1.02099371, + "epoch": 0.2913485240185174, + "flos": 24497900599680.0, + "grad_norm": 2.0759598586271304, + "language_loss": 0.73135853, + "learning_rate": 3.325437797839883e-06, + "loss": 0.75339991, + "num_input_tokens_seen": 51912205, + "step": 2423, + "time_per_iteration": 2.681594133377075 + }, + { + "auxiliary_loss_clip": 0.01202452, + "auxiliary_loss_mlp": 0.01030168, + "balance_loss_clip": 1.05992818, + "balance_loss_mlp": 1.02032721, + "epoch": 0.2914687669091565, + "flos": 17931024334080.0, + "grad_norm": 2.5807648647477492, + "language_loss": 0.74785852, + "learning_rate": 3.3248543501837015e-06, + "loss": 0.7701847, + "num_input_tokens_seen": 51929410, + "step": 2424, + "time_per_iteration": 2.8329920768737793 + }, + { + "auxiliary_loss_clip": 0.01114776, + "auxiliary_loss_mlp": 0.01031963, + "balance_loss_clip": 0.91116452, + "balance_loss_mlp": 1.02241433, + "epoch": 0.2915890097997956, + "flos": 22529313768960.0, + "grad_norm": 2.0736715093954383, + "language_loss": 0.77045631, + "learning_rate": 3.3242707015481684e-06, + "loss": 0.79192376, + "num_input_tokens_seen": 51949345, + "step": 2425, + "time_per_iteration": 2.7708494663238525 + }, + { + "auxiliary_loss_clip": 0.01151815, + "auxiliary_loss_mlp": 0.01034688, + "balance_loss_clip": 0.97834921, + "balance_loss_mlp": 1.02513337, + "epoch": 0.2917092526904347, + "flos": 13845216193920.0, + "grad_norm": 2.1964657306947357, + "language_loss": 0.80988514, + "learning_rate": 3.323686852021823e-06, + "loss": 0.83175021, + "num_input_tokens_seen": 51966855, + "step": 2426, + "time_per_iteration": 2.713322639465332 + }, + { + "auxiliary_loss_clip": 0.01136059, + "auxiliary_loss_mlp": 0.01033008, + "balance_loss_clip": 0.94266337, + "balance_loss_mlp": 1.02373314, + "epoch": 0.2918294955810738, + "flos": 22674859678080.0, + "grad_norm": 2.554289823247638, + "language_loss": 0.79496568, + "learning_rate": 3.323102801693235e-06, + "loss": 0.81665635, + "num_input_tokens_seen": 51985620, + "step": 2427, + "time_per_iteration": 2.7150795459747314 + }, + { + "auxiliary_loss_clip": 0.01174238, + "auxiliary_loss_mlp": 0.01030559, + "balance_loss_clip": 1.02208543, + "balance_loss_mlp": 1.02077842, + "epoch": 0.29194973847171285, + "flos": 23438284364160.0, + "grad_norm": 2.102946459704594, + "language_loss": 0.80226362, + "learning_rate": 3.322518550651003e-06, + "loss": 0.82431155, + "num_input_tokens_seen": 52004930, + "step": 2428, + "time_per_iteration": 2.6961116790771484 + }, + { + "auxiliary_loss_clip": 0.01160467, + "auxiliary_loss_mlp": 0.01034002, + "balance_loss_clip": 0.98364151, + "balance_loss_mlp": 1.02441144, + "epoch": 0.29206998136235196, + "flos": 21909064694400.0, + "grad_norm": 3.344388377528139, + "language_loss": 0.81258053, + "learning_rate": 3.3219340989837586e-06, + "loss": 0.83452523, + "num_input_tokens_seen": 52024920, + "step": 2429, + "time_per_iteration": 2.7187387943267822 + }, + { + "auxiliary_loss_clip": 0.01157225, + "auxiliary_loss_mlp": 0.01036908, + "balance_loss_clip": 0.98527026, + "balance_loss_mlp": 1.02811635, + "epoch": 0.292190224252991, + "flos": 23215925220480.0, + "grad_norm": 2.441267629544485, + "language_loss": 0.8060075, + "learning_rate": 3.3213494467801625e-06, + "loss": 0.82794881, + "num_input_tokens_seen": 52044095, + "step": 2430, + "time_per_iteration": 2.6830086708068848 + }, + { + "auxiliary_loss_clip": 0.01059306, + "auxiliary_loss_mlp": 0.01026828, + "balance_loss_clip": 0.82513714, + "balance_loss_mlp": 1.01781547, + "epoch": 0.2923104671436301, + "flos": 20740818752640.0, + "grad_norm": 2.205443645388656, + "language_loss": 0.71745241, + "learning_rate": 3.3207645941289063e-06, + "loss": 0.73831373, + "num_input_tokens_seen": 52062440, + "step": 2431, + "time_per_iteration": 2.8240227699279785 + }, + { + "auxiliary_loss_clip": 0.01178415, + "auxiliary_loss_mlp": 0.00512266, + "balance_loss_clip": 1.02330184, + "balance_loss_mlp": 0.0, + "epoch": 0.29243071003426924, + "flos": 35809114999680.0, + "grad_norm": 1.909820406818937, + "language_loss": 0.80438703, + "learning_rate": 3.320179541118711e-06, + "loss": 0.82129383, + "num_input_tokens_seen": 52084940, + "step": 2432, + "time_per_iteration": 2.760932683944702 + }, + { + "auxiliary_loss_clip": 0.01084934, + "auxiliary_loss_mlp": 0.01005627, + "balance_loss_clip": 0.99444205, + "balance_loss_mlp": 1.00375557, + "epoch": 0.2925509529249083, + "flos": 58081598524800.0, + "grad_norm": 0.9895563965946836, + "language_loss": 0.60329771, + "learning_rate": 3.3195942878383293e-06, + "loss": 0.62420332, + "num_input_tokens_seen": 52141040, + "step": 2433, + "time_per_iteration": 4.405946493148804 + }, + { + "auxiliary_loss_clip": 0.0118163, + "auxiliary_loss_mlp": 0.01034776, + "balance_loss_clip": 1.02197301, + "balance_loss_mlp": 1.02402902, + "epoch": 0.2926711958155474, + "flos": 21397122103680.0, + "grad_norm": 2.1035867662199954, + "language_loss": 0.77796423, + "learning_rate": 3.319008834376543e-06, + "loss": 0.80012834, + "num_input_tokens_seen": 52160730, + "step": 2434, + "time_per_iteration": 2.712639093399048 + }, + { + "auxiliary_loss_clip": 0.01135372, + "auxiliary_loss_mlp": 0.01032016, + "balance_loss_clip": 0.94060361, + "balance_loss_mlp": 1.02275372, + "epoch": 0.2927914387061865, + "flos": 23185796688000.0, + "grad_norm": 2.2517064318404874, + "language_loss": 0.88618022, + "learning_rate": 3.3184231808221654e-06, + "loss": 0.90785408, + "num_input_tokens_seen": 52175055, + "step": 2435, + "time_per_iteration": 3.535738229751587 + }, + { + "auxiliary_loss_clip": 0.01130881, + "auxiliary_loss_mlp": 0.01028553, + "balance_loss_clip": 0.94438338, + "balance_loss_mlp": 1.01931977, + "epoch": 0.29291168159682557, + "flos": 22455553190400.0, + "grad_norm": 2.486956251568388, + "language_loss": 0.62788039, + "learning_rate": 3.3178373272640394e-06, + "loss": 0.64947474, + "num_input_tokens_seen": 52194150, + "step": 2436, + "time_per_iteration": 2.749164342880249 + }, + { + "auxiliary_loss_clip": 0.01200243, + "auxiliary_loss_mlp": 0.01030113, + "balance_loss_clip": 1.06034148, + "balance_loss_mlp": 1.02099323, + "epoch": 0.2930319244874647, + "flos": 21170632896000.0, + "grad_norm": 7.878850199829447, + "language_loss": 0.84633547, + "learning_rate": 3.3172512737910387e-06, + "loss": 0.86863899, + "num_input_tokens_seen": 52211660, + "step": 2437, + "time_per_iteration": 3.5848212242126465 + }, + { + "auxiliary_loss_clip": 0.01178628, + "auxiliary_loss_mlp": 0.01038796, + "balance_loss_clip": 1.02143347, + "balance_loss_mlp": 1.02928925, + "epoch": 0.2931521673781038, + "flos": 31357843931520.0, + "grad_norm": 2.4492195648071604, + "language_loss": 0.88449335, + "learning_rate": 3.3166650204920674e-06, + "loss": 0.90666759, + "num_input_tokens_seen": 52232830, + "step": 2438, + "time_per_iteration": 2.7319931983947754 + }, + { + "auxiliary_loss_clip": 0.01179153, + "auxiliary_loss_mlp": 0.01036439, + "balance_loss_clip": 1.02346921, + "balance_loss_mlp": 1.0272541, + "epoch": 0.29327241026874284, + "flos": 24200990778240.0, + "grad_norm": 1.8215276452056561, + "language_loss": 0.81648648, + "learning_rate": 3.316078567456059e-06, + "loss": 0.83864236, + "num_input_tokens_seen": 52250670, + "step": 2439, + "time_per_iteration": 2.702758312225342 + }, + { + "auxiliary_loss_clip": 0.01088255, + "auxiliary_loss_mlp": 0.01032294, + "balance_loss_clip": 0.86770058, + "balance_loss_mlp": 1.02257872, + "epoch": 0.29339265315938196, + "flos": 24242611662720.0, + "grad_norm": 2.187914761528462, + "language_loss": 0.75769007, + "learning_rate": 3.3154919147719786e-06, + "loss": 0.77889556, + "num_input_tokens_seen": 52271685, + "step": 2440, + "time_per_iteration": 3.713162660598755 + }, + { + "auxiliary_loss_clip": 0.01177405, + "auxiliary_loss_mlp": 0.01033204, + "balance_loss_clip": 1.0213263, + "balance_loss_mlp": 1.02373242, + "epoch": 0.29351289605002107, + "flos": 16946641134720.0, + "grad_norm": 2.468649825385536, + "language_loss": 0.85868603, + "learning_rate": 3.31490506252882e-06, + "loss": 0.88079208, + "num_input_tokens_seen": 52291065, + "step": 2441, + "time_per_iteration": 2.649817705154419 + }, + { + "auxiliary_loss_clip": 0.01120827, + "auxiliary_loss_mlp": 0.01032654, + "balance_loss_clip": 0.93876338, + "balance_loss_mlp": 1.02346253, + "epoch": 0.2936331389406601, + "flos": 19829082810240.0, + "grad_norm": 1.7716808573338516, + "language_loss": 0.84073853, + "learning_rate": 3.31431801081561e-06, + "loss": 0.86227328, + "num_input_tokens_seen": 52310000, + "step": 2442, + "time_per_iteration": 2.6594653129577637 + }, + { + "auxiliary_loss_clip": 0.01050383, + "auxiliary_loss_mlp": 0.01003917, + "balance_loss_clip": 0.94794613, + "balance_loss_mlp": 1.00204527, + "epoch": 0.29375338183129923, + "flos": 71416844398080.0, + "grad_norm": 0.9087443088673959, + "language_loss": 0.67914772, + "learning_rate": 3.313730759721402e-06, + "loss": 0.6996907, + "num_input_tokens_seen": 52372930, + "step": 2443, + "time_per_iteration": 3.3772807121276855 + }, + { + "auxiliary_loss_clip": 0.01156918, + "auxiliary_loss_mlp": 0.01039165, + "balance_loss_clip": 0.98693877, + "balance_loss_mlp": 1.03036785, + "epoch": 0.29387362472193834, + "flos": 22054502862720.0, + "grad_norm": 2.0168156006100197, + "language_loss": 0.8603909, + "learning_rate": 3.313143309335282e-06, + "loss": 0.88235176, + "num_input_tokens_seen": 52391420, + "step": 2444, + "time_per_iteration": 2.7501027584075928 + }, + { + "auxiliary_loss_clip": 0.01130244, + "auxiliary_loss_mlp": 0.01028697, + "balance_loss_clip": 0.94491732, + "balance_loss_mlp": 1.01910639, + "epoch": 0.2939938676125774, + "flos": 22966418373120.0, + "grad_norm": 1.8863317722888924, + "language_loss": 0.84875786, + "learning_rate": 3.3125556597463665e-06, + "loss": 0.87034726, + "num_input_tokens_seen": 52410725, + "step": 2445, + "time_per_iteration": 2.758312463760376 + }, + { + "auxiliary_loss_clip": 0.0117582, + "auxiliary_loss_mlp": 0.01040089, + "balance_loss_clip": 1.02247405, + "balance_loss_mlp": 1.03161931, + "epoch": 0.2941141105032165, + "flos": 31358705857920.0, + "grad_norm": 1.6703003721004326, + "language_loss": 0.66071224, + "learning_rate": 3.311967811043801e-06, + "loss": 0.68287134, + "num_input_tokens_seen": 52432645, + "step": 2446, + "time_per_iteration": 2.7624759674072266 + }, + { + "auxiliary_loss_clip": 0.01180925, + "auxiliary_loss_mlp": 0.01033045, + "balance_loss_clip": 1.02520025, + "balance_loss_mlp": 1.02388954, + "epoch": 0.29423435339385556, + "flos": 23222138273280.0, + "grad_norm": 5.369118585284361, + "language_loss": 0.81885362, + "learning_rate": 3.3113797633167617e-06, + "loss": 0.84099329, + "num_input_tokens_seen": 52450940, + "step": 2447, + "time_per_iteration": 2.635021686553955 + }, + { + "auxiliary_loss_clip": 0.01200022, + "auxiliary_loss_mlp": 0.01040046, + "balance_loss_clip": 1.05933309, + "balance_loss_mlp": 1.03003836, + "epoch": 0.2943545962844947, + "flos": 26864054138880.0, + "grad_norm": 2.4462558155465297, + "language_loss": 0.69403386, + "learning_rate": 3.310791516654455e-06, + "loss": 0.71643454, + "num_input_tokens_seen": 52468000, + "step": 2448, + "time_per_iteration": 2.6831748485565186 + }, + { + "auxiliary_loss_clip": 0.01136927, + "auxiliary_loss_mlp": 0.01034657, + "balance_loss_clip": 0.94308418, + "balance_loss_mlp": 1.02472091, + "epoch": 0.2944748391751338, + "flos": 20231677422720.0, + "grad_norm": 2.0025868867123737, + "language_loss": 0.79167289, + "learning_rate": 3.3102030711461177e-06, + "loss": 0.81338871, + "num_input_tokens_seen": 52487575, + "step": 2449, + "time_per_iteration": 2.683507204055786 + }, + { + "auxiliary_loss_clip": 0.01131037, + "auxiliary_loss_mlp": 0.01028517, + "balance_loss_clip": 0.94335705, + "balance_loss_mlp": 1.01915288, + "epoch": 0.29459508206577284, + "flos": 15960965045760.0, + "grad_norm": 2.096924818509848, + "language_loss": 0.67824316, + "learning_rate": 3.3096144268810156e-06, + "loss": 0.6998387, + "num_input_tokens_seen": 52506335, + "step": 2450, + "time_per_iteration": 2.6793785095214844 + }, + { + "auxiliary_loss_clip": 0.01168218, + "auxiliary_loss_mlp": 0.01034157, + "balance_loss_clip": 1.01735032, + "balance_loss_mlp": 1.02455509, + "epoch": 0.29471532495641195, + "flos": 20412882558720.0, + "grad_norm": 2.086842493096396, + "language_loss": 0.7316432, + "learning_rate": 3.3090255839484462e-06, + "loss": 0.75366694, + "num_input_tokens_seen": 52524330, + "step": 2451, + "time_per_iteration": 2.629181146621704 + }, + { + "auxiliary_loss_clip": 0.01153087, + "auxiliary_loss_mlp": 0.01030093, + "balance_loss_clip": 0.97976255, + "balance_loss_mlp": 1.01990008, + "epoch": 0.29483556784705106, + "flos": 20376576887040.0, + "grad_norm": 2.0035640458761894, + "language_loss": 0.85161257, + "learning_rate": 3.3084365424377366e-06, + "loss": 0.87344432, + "num_input_tokens_seen": 52543095, + "step": 2452, + "time_per_iteration": 2.6807122230529785 + }, + { + "auxiliary_loss_clip": 0.01020588, + "auxiliary_loss_mlp": 0.0100968, + "balance_loss_clip": 0.88270533, + "balance_loss_mlp": 1.00762939, + "epoch": 0.2949558107376901, + "flos": 68555660595840.0, + "grad_norm": 0.7286267706034509, + "language_loss": 0.55976593, + "learning_rate": 3.307847302438245e-06, + "loss": 0.58006859, + "num_input_tokens_seen": 52597075, + "step": 2453, + "time_per_iteration": 3.166109800338745 + }, + { + "auxiliary_loss_clip": 0.01089454, + "auxiliary_loss_mlp": 0.01030545, + "balance_loss_clip": 0.8988179, + "balance_loss_mlp": 1.02095461, + "epoch": 0.2950760536283292, + "flos": 16107085572480.0, + "grad_norm": 2.068257786483038, + "language_loss": 0.7735762, + "learning_rate": 3.3072578640393562e-06, + "loss": 0.7947762, + "num_input_tokens_seen": 52614410, + "step": 2454, + "time_per_iteration": 2.7052223682403564 + }, + { + "auxiliary_loss_clip": 0.01152366, + "auxiliary_loss_mlp": 0.01033849, + "balance_loss_clip": 0.98149407, + "balance_loss_mlp": 1.0244081, + "epoch": 0.29519629651896834, + "flos": 20483626394880.0, + "grad_norm": 1.9374537340671825, + "language_loss": 0.7987349, + "learning_rate": 3.3066682273304886e-06, + "loss": 0.82059705, + "num_input_tokens_seen": 52632055, + "step": 2455, + "time_per_iteration": 2.7216641902923584 + }, + { + "auxiliary_loss_clip": 0.011814, + "auxiliary_loss_mlp": 0.00512626, + "balance_loss_clip": 1.02037406, + "balance_loss_mlp": 0.0, + "epoch": 0.2953165394096074, + "flos": 18916484941440.0, + "grad_norm": 2.1106136415877135, + "language_loss": 0.79047364, + "learning_rate": 3.3060783924010904e-06, + "loss": 0.80741388, + "num_input_tokens_seen": 52649980, + "step": 2456, + "time_per_iteration": 2.667710304260254 + }, + { + "auxiliary_loss_clip": 0.0113122, + "auxiliary_loss_mlp": 0.01029946, + "balance_loss_clip": 0.94364119, + "balance_loss_mlp": 1.02040958, + "epoch": 0.2954367823002465, + "flos": 20624467622400.0, + "grad_norm": 2.26469678232969, + "language_loss": 0.84931672, + "learning_rate": 3.3054883593406387e-06, + "loss": 0.87092835, + "num_input_tokens_seen": 52664730, + "step": 2457, + "time_per_iteration": 2.7361369132995605 + }, + { + "auxiliary_loss_clip": 0.01155642, + "auxiliary_loss_mlp": 0.0103156, + "balance_loss_clip": 0.98203051, + "balance_loss_mlp": 1.02243459, + "epoch": 0.2955570251908856, + "flos": 31175525473920.0, + "grad_norm": 2.7608317426516593, + "language_loss": 0.65110171, + "learning_rate": 3.3048981282386404e-06, + "loss": 0.67297369, + "num_input_tokens_seen": 52686040, + "step": 2458, + "time_per_iteration": 3.68514347076416 + }, + { + "auxiliary_loss_clip": 0.01116181, + "auxiliary_loss_mlp": 0.01033182, + "balance_loss_clip": 0.94090569, + "balance_loss_mlp": 1.02396727, + "epoch": 0.29567726808152467, + "flos": 21650328051840.0, + "grad_norm": 2.10320704476126, + "language_loss": 0.82497412, + "learning_rate": 3.304307699184634e-06, + "loss": 0.84646773, + "num_input_tokens_seen": 52704630, + "step": 2459, + "time_per_iteration": 2.717372179031372 + }, + { + "auxiliary_loss_clip": 0.01155021, + "auxiliary_loss_mlp": 0.01038966, + "balance_loss_clip": 0.98363459, + "balance_loss_mlp": 1.03004324, + "epoch": 0.2957975109721638, + "flos": 24243868638720.0, + "grad_norm": 1.8552352246624313, + "language_loss": 0.78686076, + "learning_rate": 3.3037170722681866e-06, + "loss": 0.8088007, + "num_input_tokens_seen": 52725465, + "step": 2460, + "time_per_iteration": 2.8267624378204346 + }, + { + "auxiliary_loss_clip": 0.01120353, + "auxiliary_loss_mlp": 0.01029341, + "balance_loss_clip": 0.94204724, + "balance_loss_mlp": 1.02060258, + "epoch": 0.29591775386280283, + "flos": 13479717352320.0, + "grad_norm": 7.806263133043633, + "language_loss": 0.67810297, + "learning_rate": 3.3031262475788956e-06, + "loss": 0.69959992, + "num_input_tokens_seen": 52742405, + "step": 2461, + "time_per_iteration": 3.586996555328369 + }, + { + "auxiliary_loss_clip": 0.01148895, + "auxiliary_loss_mlp": 0.01032801, + "balance_loss_clip": 0.98030639, + "balance_loss_mlp": 1.02358627, + "epoch": 0.29603799675344195, + "flos": 17749783284480.0, + "grad_norm": 1.9916853344123036, + "language_loss": 0.73004472, + "learning_rate": 3.3025352252063897e-06, + "loss": 0.75186169, + "num_input_tokens_seen": 52761100, + "step": 2462, + "time_per_iteration": 2.7055962085723877 + }, + { + "auxiliary_loss_clip": 0.01177351, + "auxiliary_loss_mlp": 0.01029312, + "balance_loss_clip": 1.02421308, + "balance_loss_mlp": 1.02037787, + "epoch": 0.29615823964408106, + "flos": 22783920347520.0, + "grad_norm": 1.7677876899553218, + "language_loss": 0.75065756, + "learning_rate": 3.3019440052403252e-06, + "loss": 0.77272421, + "num_input_tokens_seen": 52780965, + "step": 2463, + "time_per_iteration": 3.651653289794922 + }, + { + "auxiliary_loss_clip": 0.0115589, + "auxiliary_loss_mlp": 0.01027317, + "balance_loss_clip": 0.98142356, + "balance_loss_mlp": 1.0185554, + "epoch": 0.2962784825347201, + "flos": 23514199758720.0, + "grad_norm": 2.729240099853004, + "language_loss": 0.7085247, + "learning_rate": 3.30135258777039e-06, + "loss": 0.73035681, + "num_input_tokens_seen": 52800335, + "step": 2464, + "time_per_iteration": 2.672389268875122 + }, + { + "auxiliary_loss_clip": 0.01179567, + "auxiliary_loss_mlp": 0.00511811, + "balance_loss_clip": 1.01877999, + "balance_loss_mlp": 0.0, + "epoch": 0.2963987254253592, + "flos": 16362769559040.0, + "grad_norm": 3.18589344018408, + "language_loss": 0.69803411, + "learning_rate": 3.3007609728863024e-06, + "loss": 0.71494788, + "num_input_tokens_seen": 52818425, + "step": 2465, + "time_per_iteration": 2.6383719444274902 + }, + { + "auxiliary_loss_clip": 0.01075731, + "auxiliary_loss_mlp": 0.01029885, + "balance_loss_clip": 0.86978543, + "balance_loss_mlp": 1.02039015, + "epoch": 0.29651896831599833, + "flos": 33472263980160.0, + "grad_norm": 2.52126764251793, + "language_loss": 0.73000616, + "learning_rate": 3.300169160677809e-06, + "loss": 0.75106233, + "num_input_tokens_seen": 52842340, + "step": 2466, + "time_per_iteration": 3.782547950744629 + }, + { + "auxiliary_loss_clip": 0.0113613, + "auxiliary_loss_mlp": 0.01034288, + "balance_loss_clip": 0.94563359, + "balance_loss_mlp": 1.02411306, + "epoch": 0.2966392112066374, + "flos": 23805363404160.0, + "grad_norm": 2.4935765476647767, + "language_loss": 0.77295846, + "learning_rate": 3.2995771512346878e-06, + "loss": 0.79466259, + "num_input_tokens_seen": 52860690, + "step": 2467, + "time_per_iteration": 2.829360008239746 + }, + { + "auxiliary_loss_clip": 0.01205179, + "auxiliary_loss_mlp": 0.00511863, + "balance_loss_clip": 1.06251097, + "balance_loss_mlp": 0.0, + "epoch": 0.2967594540972765, + "flos": 19938466702080.0, + "grad_norm": 2.3346962515141727, + "language_loss": 0.72845471, + "learning_rate": 3.298984944646746e-06, + "loss": 0.74562514, + "num_input_tokens_seen": 52879370, + "step": 2468, + "time_per_iteration": 2.747868776321411 + }, + { + "auxiliary_loss_clip": 0.01179855, + "auxiliary_loss_mlp": 0.00510509, + "balance_loss_clip": 1.02293992, + "balance_loss_mlp": 0.0, + "epoch": 0.2968796969879156, + "flos": 23732823888000.0, + "grad_norm": 2.452427776471274, + "language_loss": 0.81788492, + "learning_rate": 3.298392541003822e-06, + "loss": 0.83478862, + "num_input_tokens_seen": 52898775, + "step": 2469, + "time_per_iteration": 2.6959617137908936 + }, + { + "auxiliary_loss_clip": 0.01153073, + "auxiliary_loss_mlp": 0.0103176, + "balance_loss_clip": 0.98217237, + "balance_loss_mlp": 1.0231595, + "epoch": 0.29699993987855466, + "flos": 22893699288960.0, + "grad_norm": 1.5857525732508093, + "language_loss": 0.89862865, + "learning_rate": 3.2977999403957806e-06, + "loss": 0.92047703, + "num_input_tokens_seen": 52917535, + "step": 2470, + "time_per_iteration": 2.720010280609131 + }, + { + "auxiliary_loss_clip": 0.01204459, + "auxiliary_loss_mlp": 0.01035248, + "balance_loss_clip": 1.06329691, + "balance_loss_mlp": 1.0255506, + "epoch": 0.2971201827691938, + "flos": 33832555349760.0, + "grad_norm": 2.0895748131969074, + "language_loss": 0.67287409, + "learning_rate": 3.2972071429125207e-06, + "loss": 0.69527113, + "num_input_tokens_seen": 52938755, + "step": 2471, + "time_per_iteration": 2.6956233978271484 + }, + { + "auxiliary_loss_clip": 0.0112527, + "auxiliary_loss_mlp": 0.01035666, + "balance_loss_clip": 0.94429094, + "balance_loss_mlp": 1.02617073, + "epoch": 0.2972404256598329, + "flos": 22054359208320.0, + "grad_norm": 3.042594198026108, + "language_loss": 0.88597131, + "learning_rate": 3.2966141486439682e-06, + "loss": 0.90758061, + "num_input_tokens_seen": 52957945, + "step": 2472, + "time_per_iteration": 2.7864339351654053 + }, + { + "auxiliary_loss_clip": 0.01081262, + "auxiliary_loss_mlp": 0.01036694, + "balance_loss_clip": 0.86300057, + "balance_loss_mlp": 1.02717543, + "epoch": 0.29736066855047194, + "flos": 31978595796480.0, + "grad_norm": 4.75782319245674, + "language_loss": 0.64289409, + "learning_rate": 3.29602095768008e-06, + "loss": 0.66407371, + "num_input_tokens_seen": 52978460, + "step": 2473, + "time_per_iteration": 2.8525757789611816 + }, + { + "auxiliary_loss_clip": 0.01148525, + "auxiliary_loss_mlp": 0.01028341, + "balance_loss_clip": 0.98377073, + "balance_loss_mlp": 1.02000213, + "epoch": 0.29748091144111105, + "flos": 33510401245440.0, + "grad_norm": 1.916229571700585, + "language_loss": 0.63269961, + "learning_rate": 3.2954275701108437e-06, + "loss": 0.6544683, + "num_input_tokens_seen": 52999640, + "step": 2474, + "time_per_iteration": 2.806187152862549 + }, + { + "auxiliary_loss_clip": 0.01102935, + "auxiliary_loss_mlp": 0.01029497, + "balance_loss_clip": 0.90439081, + "balance_loss_mlp": 1.01975727, + "epoch": 0.29760115433175016, + "flos": 41283373409280.0, + "grad_norm": 2.7559407761328805, + "language_loss": 0.68691677, + "learning_rate": 3.294833986026275e-06, + "loss": 0.70824111, + "num_input_tokens_seen": 53022880, + "step": 2475, + "time_per_iteration": 2.952035427093506 + }, + { + "auxiliary_loss_clip": 0.01120155, + "auxiliary_loss_mlp": 0.01030492, + "balance_loss_clip": 0.94209504, + "balance_loss_mlp": 1.02150345, + "epoch": 0.2977213972223892, + "flos": 24493339572480.0, + "grad_norm": 1.9384616273153246, + "language_loss": 0.857095, + "learning_rate": 3.29424020551642e-06, + "loss": 0.87860143, + "num_input_tokens_seen": 53041515, + "step": 2476, + "time_per_iteration": 2.7273144721984863 + }, + { + "auxiliary_loss_clip": 0.01207499, + "auxiliary_loss_mlp": 0.01032468, + "balance_loss_clip": 1.06309867, + "balance_loss_mlp": 1.02312231, + "epoch": 0.2978416401130283, + "flos": 21285116519040.0, + "grad_norm": 1.808255247657696, + "language_loss": 0.72154903, + "learning_rate": 3.2936462286713546e-06, + "loss": 0.74394876, + "num_input_tokens_seen": 53059865, + "step": 2477, + "time_per_iteration": 2.641120433807373 + }, + { + "auxiliary_loss_clip": 0.01178352, + "auxiliary_loss_mlp": 0.01032754, + "balance_loss_clip": 1.02180815, + "balance_loss_mlp": 1.02360487, + "epoch": 0.2979618830036674, + "flos": 25772154554880.0, + "grad_norm": 2.3197634253104784, + "language_loss": 0.77184451, + "learning_rate": 3.2930520555811846e-06, + "loss": 0.79395556, + "num_input_tokens_seen": 53079490, + "step": 2478, + "time_per_iteration": 2.7596919536590576 + }, + { + "auxiliary_loss_clip": 0.01030892, + "auxiliary_loss_mlp": 0.00512673, + "balance_loss_clip": 0.82182801, + "balance_loss_mlp": 0.0, + "epoch": 0.2980821258943065, + "flos": 23476996247040.0, + "grad_norm": 2.434745778425535, + "language_loss": 0.80274355, + "learning_rate": 3.292457686336046e-06, + "loss": 0.81817925, + "num_input_tokens_seen": 53098810, + "step": 2479, + "time_per_iteration": 2.916269063949585 + }, + { + "auxiliary_loss_clip": 0.01068202, + "auxiliary_loss_mlp": 0.01013571, + "balance_loss_clip": 0.95924628, + "balance_loss_mlp": 1.01184201, + "epoch": 0.2982023687849456, + "flos": 69752314195200.0, + "grad_norm": 0.8551352981499625, + "language_loss": 0.61254895, + "learning_rate": 3.291863121026105e-06, + "loss": 0.6333667, + "num_input_tokens_seen": 53162590, + "step": 2480, + "time_per_iteration": 3.523150682449341 + }, + { + "auxiliary_loss_clip": 0.01180511, + "auxiliary_loss_mlp": 0.01026278, + "balance_loss_clip": 1.02373445, + "balance_loss_mlp": 1.01676536, + "epoch": 0.29832261167558466, + "flos": 29825930741760.0, + "grad_norm": 5.895696206029092, + "language_loss": 0.76844931, + "learning_rate": 3.2912683597415547e-06, + "loss": 0.79051721, + "num_input_tokens_seen": 53186675, + "step": 2481, + "time_per_iteration": 2.732212543487549 + }, + { + "auxiliary_loss_clip": 0.01131628, + "auxiliary_loss_mlp": 0.0103733, + "balance_loss_clip": 0.94418252, + "balance_loss_mlp": 1.02787685, + "epoch": 0.29844285456622377, + "flos": 33910158683520.0, + "grad_norm": 2.1683841798285894, + "language_loss": 0.78139937, + "learning_rate": 3.2906734025726213e-06, + "loss": 0.8030889, + "num_input_tokens_seen": 53205940, + "step": 2482, + "time_per_iteration": 2.8374552726745605 + }, + { + "auxiliary_loss_clip": 0.01184721, + "auxiliary_loss_mlp": 0.01037349, + "balance_loss_clip": 1.02377701, + "balance_loss_mlp": 1.02805078, + "epoch": 0.2985630974568629, + "flos": 23876933253120.0, + "grad_norm": 2.601081037314794, + "language_loss": 0.87828058, + "learning_rate": 3.290078249609559e-06, + "loss": 0.90050125, + "num_input_tokens_seen": 53225360, + "step": 2483, + "time_per_iteration": 2.7036917209625244 + }, + { + "auxiliary_loss_clip": 0.01176307, + "auxiliary_loss_mlp": 0.01036781, + "balance_loss_clip": 1.02432704, + "balance_loss_mlp": 1.02812076, + "epoch": 0.29868334034750194, + "flos": 21799106184960.0, + "grad_norm": 2.2829230954806614, + "language_loss": 0.88052326, + "learning_rate": 3.2894829009426514e-06, + "loss": 0.90265411, + "num_input_tokens_seen": 53243195, + "step": 2484, + "time_per_iteration": 3.7378644943237305 + }, + { + "auxiliary_loss_clip": 0.01175128, + "auxiliary_loss_mlp": 0.01033017, + "balance_loss_clip": 1.02133656, + "balance_loss_mlp": 1.02361774, + "epoch": 0.29880358323814105, + "flos": 25666649331840.0, + "grad_norm": 2.1907294937854003, + "language_loss": 0.77626812, + "learning_rate": 3.288887356662213e-06, + "loss": 0.79834956, + "num_input_tokens_seen": 53264530, + "step": 2485, + "time_per_iteration": 2.6840851306915283 + }, + { + "auxiliary_loss_clip": 0.01078538, + "auxiliary_loss_mlp": 0.01001893, + "balance_loss_clip": 0.99330997, + "balance_loss_mlp": 1.00008094, + "epoch": 0.29892382612878016, + "flos": 71005846003200.0, + "grad_norm": 0.7641796160199221, + "language_loss": 0.59672391, + "learning_rate": 3.288291616858588e-06, + "loss": 0.61752826, + "num_input_tokens_seen": 53319920, + "step": 2486, + "time_per_iteration": 3.0672621726989746 + }, + { + "auxiliary_loss_clip": 0.0110378, + "auxiliary_loss_mlp": 0.01042194, + "balance_loss_clip": 0.90954441, + "balance_loss_mlp": 1.03305662, + "epoch": 0.2990440690194192, + "flos": 25481134563840.0, + "grad_norm": 1.919191705655436, + "language_loss": 0.76341712, + "learning_rate": 3.287695681622149e-06, + "loss": 0.78487682, + "num_input_tokens_seen": 53339270, + "step": 2487, + "time_per_iteration": 3.7432899475097656 + }, + { + "auxiliary_loss_clip": 0.01159528, + "auxiliary_loss_mlp": 0.01030316, + "balance_loss_clip": 0.984797, + "balance_loss_mlp": 1.02070785, + "epoch": 0.2991643119100583, + "flos": 23732357011200.0, + "grad_norm": 2.341520750515584, + "language_loss": 0.8113867, + "learning_rate": 3.2870995510432982e-06, + "loss": 0.83328509, + "num_input_tokens_seen": 53357750, + "step": 2488, + "time_per_iteration": 2.8091280460357666 + }, + { + "auxiliary_loss_clip": 0.01170455, + "auxiliary_loss_mlp": 0.01026318, + "balance_loss_clip": 1.02047515, + "balance_loss_mlp": 1.01791358, + "epoch": 0.29928455480069743, + "flos": 27417545786880.0, + "grad_norm": 1.8894072373651072, + "language_loss": 0.76685739, + "learning_rate": 3.2865032252124697e-06, + "loss": 0.78882509, + "num_input_tokens_seen": 53378265, + "step": 2489, + "time_per_iteration": 3.6534409523010254 + }, + { + "auxiliary_loss_clip": 0.01149141, + "auxiliary_loss_mlp": 0.01033126, + "balance_loss_clip": 0.97983664, + "balance_loss_mlp": 1.02462661, + "epoch": 0.2994047976913365, + "flos": 33692935184640.0, + "grad_norm": 4.2326681120429654, + "language_loss": 0.77631164, + "learning_rate": 3.2859067042201243e-06, + "loss": 0.79813433, + "num_input_tokens_seen": 53400305, + "step": 2490, + "time_per_iteration": 2.793156147003174 + }, + { + "auxiliary_loss_clip": 0.01060777, + "auxiliary_loss_mlp": 0.01027286, + "balance_loss_clip": 0.86327088, + "balance_loss_mlp": 1.01844704, + "epoch": 0.2995250405819756, + "flos": 16763963541120.0, + "grad_norm": 1.982504389054976, + "language_loss": 0.78150582, + "learning_rate": 3.2853099881567544e-06, + "loss": 0.80238646, + "num_input_tokens_seen": 53418705, + "step": 2491, + "time_per_iteration": 2.7497057914733887 + }, + { + "auxiliary_loss_clip": 0.01198666, + "auxiliary_loss_mlp": 0.01038266, + "balance_loss_clip": 1.06130457, + "balance_loss_mlp": 1.03005219, + "epoch": 0.29964528347261465, + "flos": 22963976248320.0, + "grad_norm": 1.6462707877084337, + "language_loss": 0.79230261, + "learning_rate": 3.284713077112881e-06, + "loss": 0.81467193, + "num_input_tokens_seen": 53438135, + "step": 2492, + "time_per_iteration": 3.5147335529327393 + }, + { + "auxiliary_loss_clip": 0.01135321, + "auxiliary_loss_mlp": 0.01033673, + "balance_loss_clip": 0.94721037, + "balance_loss_mlp": 1.0241127, + "epoch": 0.29976552636325376, + "flos": 16938021870720.0, + "grad_norm": 10.446056785180993, + "language_loss": 0.86206013, + "learning_rate": 3.284115971179056e-06, + "loss": 0.88375008, + "num_input_tokens_seen": 53452165, + "step": 2493, + "time_per_iteration": 2.7450709342956543 + }, + { + "auxiliary_loss_clip": 0.01083117, + "auxiliary_loss_mlp": 0.01034314, + "balance_loss_clip": 0.86767596, + "balance_loss_mlp": 1.02599883, + "epoch": 0.2998857692538929, + "flos": 17056455989760.0, + "grad_norm": 1.8337694070693433, + "language_loss": 0.78279531, + "learning_rate": 3.283518670445859e-06, + "loss": 0.80396962, + "num_input_tokens_seen": 53470075, + "step": 2494, + "time_per_iteration": 2.7821803092956543 + }, + { + "auxiliary_loss_clip": 0.01055582, + "auxiliary_loss_mlp": 0.00503408, + "balance_loss_clip": 0.95672679, + "balance_loss_mlp": 0.0, + "epoch": 0.30000601214453193, + "flos": 68831528025600.0, + "grad_norm": 0.6845032363250511, + "language_loss": 0.54325736, + "learning_rate": 3.2829211750038995e-06, + "loss": 0.55884719, + "num_input_tokens_seen": 53538705, + "step": 2495, + "time_per_iteration": 3.325725793838501 + }, + { + "auxiliary_loss_clip": 0.01126346, + "auxiliary_loss_mlp": 0.01036857, + "balance_loss_clip": 0.94441658, + "balance_loss_mlp": 1.0274334, + "epoch": 0.30012625503517104, + "flos": 17603267708160.0, + "grad_norm": 2.186996704923717, + "language_loss": 0.89071208, + "learning_rate": 3.2823234849438183e-06, + "loss": 0.9123441, + "num_input_tokens_seen": 53556740, + "step": 2496, + "time_per_iteration": 2.725714921951294 + }, + { + "auxiliary_loss_clip": 0.01156828, + "auxiliary_loss_mlp": 0.0102598, + "balance_loss_clip": 0.98555636, + "balance_loss_mlp": 1.01741457, + "epoch": 0.30024649792581015, + "flos": 21252581775360.0, + "grad_norm": 3.440169700296732, + "language_loss": 0.75892889, + "learning_rate": 3.2817256003562836e-06, + "loss": 0.78075701, + "num_input_tokens_seen": 53577115, + "step": 2497, + "time_per_iteration": 2.689777374267578 + }, + { + "auxiliary_loss_clip": 0.01085541, + "auxiliary_loss_mlp": 0.01029816, + "balance_loss_clip": 0.86757392, + "balance_loss_mlp": 1.02132225, + "epoch": 0.3003667408164492, + "flos": 23003262748800.0, + "grad_norm": 2.1360978832493625, + "language_loss": 0.65802264, + "learning_rate": 3.281127521331995e-06, + "loss": 0.67917621, + "num_input_tokens_seen": 53598295, + "step": 2498, + "time_per_iteration": 2.8593363761901855 + }, + { + "auxiliary_loss_clip": 0.01097364, + "auxiliary_loss_mlp": 0.01003329, + "balance_loss_clip": 1.02521193, + "balance_loss_mlp": 1.00130224, + "epoch": 0.3004869837070883, + "flos": 64232340750720.0, + "grad_norm": 0.8935401260264748, + "language_loss": 0.60668969, + "learning_rate": 3.2805292479616798e-06, + "loss": 0.62769663, + "num_input_tokens_seen": 53657160, + "step": 2499, + "time_per_iteration": 3.203421115875244 + }, + { + "auxiliary_loss_clip": 0.01154677, + "auxiliary_loss_mlp": 0.01037157, + "balance_loss_clip": 0.98386097, + "balance_loss_mlp": 1.02845502, + "epoch": 0.30060722659772743, + "flos": 26248653400320.0, + "grad_norm": 3.993998913062641, + "language_loss": 0.92369986, + "learning_rate": 3.2799307803360955e-06, + "loss": 0.94561821, + "num_input_tokens_seen": 53673090, + "step": 2500, + "time_per_iteration": 2.6857523918151855 + }, + { + "auxiliary_loss_clip": 0.0119957, + "auxiliary_loss_mlp": 0.01026951, + "balance_loss_clip": 1.05995011, + "balance_loss_mlp": 1.01817131, + "epoch": 0.3007274694883665, + "flos": 24970879912320.0, + "grad_norm": 2.7466731602252135, + "language_loss": 0.81593734, + "learning_rate": 3.27933211854603e-06, + "loss": 0.8382026, + "num_input_tokens_seen": 53692145, + "step": 2501, + "time_per_iteration": 2.6473309993743896 + }, + { + "auxiliary_loss_clip": 0.01153098, + "auxiliary_loss_mlp": 0.01034992, + "balance_loss_clip": 0.98324984, + "balance_loss_mlp": 1.0250138, + "epoch": 0.3008477123790056, + "flos": 17055845458560.0, + "grad_norm": 2.054729858271887, + "language_loss": 0.86804831, + "learning_rate": 3.278733262682299e-06, + "loss": 0.88992923, + "num_input_tokens_seen": 53710000, + "step": 2502, + "time_per_iteration": 2.6100409030914307 + }, + { + "auxiliary_loss_clip": 0.01199272, + "auxiliary_loss_mlp": 0.01032879, + "balance_loss_clip": 1.05867314, + "balance_loss_mlp": 1.02378285, + "epoch": 0.3009679552696447, + "flos": 21506398254720.0, + "grad_norm": 3.213604191256047, + "language_loss": 0.82732177, + "learning_rate": 3.2781342128357484e-06, + "loss": 0.84964329, + "num_input_tokens_seen": 53729355, + "step": 2503, + "time_per_iteration": 2.6303117275238037 + }, + { + "auxiliary_loss_clip": 0.01128148, + "auxiliary_loss_mlp": 0.01031353, + "balance_loss_clip": 0.94197786, + "balance_loss_mlp": 1.02269244, + "epoch": 0.30108819816028376, + "flos": 21134004001920.0, + "grad_norm": 2.573618951620429, + "language_loss": 0.80249578, + "learning_rate": 3.2775349690972547e-06, + "loss": 0.82409078, + "num_input_tokens_seen": 53743505, + "step": 2504, + "time_per_iteration": 2.6854934692382812 + }, + { + "auxiliary_loss_clip": 0.01072033, + "auxiliary_loss_mlp": 0.0100871, + "balance_loss_clip": 0.98786247, + "balance_loss_mlp": 1.00696909, + "epoch": 0.30120844105092287, + "flos": 71126434938240.0, + "grad_norm": 0.7662684209053745, + "language_loss": 0.51836634, + "learning_rate": 3.276935531557722e-06, + "loss": 0.53917372, + "num_input_tokens_seen": 53808725, + "step": 2505, + "time_per_iteration": 3.309422731399536 + }, + { + "auxiliary_loss_clip": 0.01107505, + "auxiliary_loss_mlp": 0.01032784, + "balance_loss_clip": 0.90357798, + "balance_loss_mlp": 1.02312231, + "epoch": 0.301328683941562, + "flos": 20264571302400.0, + "grad_norm": 2.1207437662666933, + "language_loss": 0.79674733, + "learning_rate": 3.2763359003080837e-06, + "loss": 0.81815028, + "num_input_tokens_seen": 53825680, + "step": 2506, + "time_per_iteration": 2.768047571182251 + }, + { + "auxiliary_loss_clip": 0.01057608, + "auxiliary_loss_mlp": 0.01009037, + "balance_loss_clip": 0.9511407, + "balance_loss_mlp": 1.0071888, + "epoch": 0.30144892683220104, + "flos": 70648212240000.0, + "grad_norm": 0.8109399385058067, + "language_loss": 0.62422985, + "learning_rate": 3.2757360754393047e-06, + "loss": 0.64489627, + "num_input_tokens_seen": 53889750, + "step": 2507, + "time_per_iteration": 3.3466265201568604 + }, + { + "auxiliary_loss_clip": 0.01174221, + "auxiliary_loss_mlp": 0.01035893, + "balance_loss_clip": 1.01995015, + "balance_loss_mlp": 1.02602816, + "epoch": 0.30156916972284015, + "flos": 22820549241600.0, + "grad_norm": 2.3954796563497327, + "language_loss": 0.63559562, + "learning_rate": 3.2751360570423767e-06, + "loss": 0.65769672, + "num_input_tokens_seen": 53908135, + "step": 2508, + "time_per_iteration": 2.689652442932129 + }, + { + "auxiliary_loss_clip": 0.01150473, + "auxiliary_loss_mlp": 0.010316, + "balance_loss_clip": 0.98186743, + "balance_loss_mlp": 1.02224255, + "epoch": 0.3016894126134792, + "flos": 29899188529920.0, + "grad_norm": 1.9962840028533893, + "language_loss": 0.7576412, + "learning_rate": 3.2745358452083236e-06, + "loss": 0.77946192, + "num_input_tokens_seen": 53931035, + "step": 2509, + "time_per_iteration": 2.754096031188965 + }, + { + "auxiliary_loss_clip": 0.01173976, + "auxiliary_loss_mlp": 0.01031903, + "balance_loss_clip": 1.02068269, + "balance_loss_mlp": 1.02333176, + "epoch": 0.3018096555041183, + "flos": 21546331200000.0, + "grad_norm": 1.5818808246144793, + "language_loss": 0.82330084, + "learning_rate": 3.2739354400281955e-06, + "loss": 0.84535968, + "num_input_tokens_seen": 53952255, + "step": 2510, + "time_per_iteration": 3.6236493587493896 + }, + { + "auxiliary_loss_clip": 0.01040292, + "auxiliary_loss_mlp": 0.005028, + "balance_loss_clip": 0.91504097, + "balance_loss_mlp": 0.0, + "epoch": 0.3019298983947574, + "flos": 59136294597120.0, + "grad_norm": 0.8665127054876681, + "language_loss": 0.63710916, + "learning_rate": 3.2733348415930744e-06, + "loss": 0.65254009, + "num_input_tokens_seen": 54014125, + "step": 2511, + "time_per_iteration": 3.288475751876831 + }, + { + "auxiliary_loss_clip": 0.01122751, + "auxiliary_loss_mlp": 0.0103251, + "balance_loss_clip": 0.94360483, + "balance_loss_mlp": 1.02278233, + "epoch": 0.3020501412853965, + "flos": 34423070941440.0, + "grad_norm": 2.2274689880824505, + "language_loss": 0.80668986, + "learning_rate": 3.27273404999407e-06, + "loss": 0.82824248, + "num_input_tokens_seen": 54036345, + "step": 2512, + "time_per_iteration": 2.8413889408111572 + }, + { + "auxiliary_loss_clip": 0.01057149, + "auxiliary_loss_mlp": 0.01005345, + "balance_loss_clip": 0.94972843, + "balance_loss_mlp": 1.00344992, + "epoch": 0.3021703841760356, + "flos": 71008288128000.0, + "grad_norm": 0.8025520510634196, + "language_loss": 0.60452473, + "learning_rate": 3.272133065322322e-06, + "loss": 0.62514973, + "num_input_tokens_seen": 54094615, + "step": 2513, + "time_per_iteration": 4.151487827301025 + }, + { + "auxiliary_loss_clip": 0.01193681, + "auxiliary_loss_mlp": 0.01029452, + "balance_loss_clip": 1.05507028, + "balance_loss_mlp": 1.02156079, + "epoch": 0.3022906270666747, + "flos": 21510528318720.0, + "grad_norm": 1.6967008950839138, + "language_loss": 0.79487681, + "learning_rate": 3.271531887669e-06, + "loss": 0.81710815, + "num_input_tokens_seen": 54114675, + "step": 2514, + "time_per_iteration": 2.633218288421631 + }, + { + "auxiliary_loss_clip": 0.0110905, + "auxiliary_loss_mlp": 0.01029062, + "balance_loss_clip": 0.90345013, + "balance_loss_mlp": 1.0202229, + "epoch": 0.30241086995731375, + "flos": 31132001168640.0, + "grad_norm": 2.69920834716714, + "language_loss": 0.63103318, + "learning_rate": 3.2709305171253015e-06, + "loss": 0.65241432, + "num_input_tokens_seen": 54134795, + "step": 2515, + "time_per_iteration": 3.7246227264404297 + }, + { + "auxiliary_loss_clip": 0.01173524, + "auxiliary_loss_mlp": 0.01026125, + "balance_loss_clip": 1.02090073, + "balance_loss_mlp": 1.01811373, + "epoch": 0.30253111284795287, + "flos": 23511542152320.0, + "grad_norm": 2.5925466203647947, + "language_loss": 0.77822518, + "learning_rate": 3.2703289537824536e-06, + "loss": 0.80022168, + "num_input_tokens_seen": 54154595, + "step": 2516, + "time_per_iteration": 2.682703971862793 + }, + { + "auxiliary_loss_clip": 0.0110543, + "auxiliary_loss_mlp": 0.01029708, + "balance_loss_clip": 0.90419412, + "balance_loss_mlp": 1.021173, + "epoch": 0.302651355738592, + "flos": 18725367651840.0, + "grad_norm": 4.135956092137473, + "language_loss": 0.78725517, + "learning_rate": 3.269727197731714e-06, + "loss": 0.80860656, + "num_input_tokens_seen": 54167360, + "step": 2517, + "time_per_iteration": 2.7160420417785645 + }, + { + "auxiliary_loss_clip": 0.01098929, + "auxiliary_loss_mlp": 0.01030402, + "balance_loss_clip": 0.9060396, + "balance_loss_mlp": 1.02139592, + "epoch": 0.30277159862923103, + "flos": 22418888382720.0, + "grad_norm": 4.634192965864323, + "language_loss": 0.77612555, + "learning_rate": 3.269125249064367e-06, + "loss": 0.79741883, + "num_input_tokens_seen": 54187055, + "step": 2518, + "time_per_iteration": 3.6227333545684814 + }, + { + "auxiliary_loss_clip": 0.01202764, + "auxiliary_loss_mlp": 0.01027962, + "balance_loss_clip": 1.06044018, + "balance_loss_mlp": 1.01900959, + "epoch": 0.30289184151987014, + "flos": 22273126992000.0, + "grad_norm": 1.7667589189175024, + "language_loss": 0.83368832, + "learning_rate": 3.2685231078717297e-06, + "loss": 0.85599554, + "num_input_tokens_seen": 54207245, + "step": 2519, + "time_per_iteration": 2.6718785762786865 + }, + { + "auxiliary_loss_clip": 0.01117357, + "auxiliary_loss_mlp": 0.00512146, + "balance_loss_clip": 0.94391716, + "balance_loss_mlp": 0.0, + "epoch": 0.30301208441050925, + "flos": 25225594231680.0, + "grad_norm": 3.961228171887227, + "language_loss": 0.75795817, + "learning_rate": 3.267920774245145e-06, + "loss": 0.77425319, + "num_input_tokens_seen": 54226650, + "step": 2520, + "time_per_iteration": 2.7740225791931152 + }, + { + "auxiliary_loss_clip": 0.01176079, + "auxiliary_loss_mlp": 0.01041471, + "balance_loss_clip": 1.02190077, + "balance_loss_mlp": 1.03152883, + "epoch": 0.3031323273011483, + "flos": 23039245198080.0, + "grad_norm": 1.9001370549429042, + "language_loss": 0.84792578, + "learning_rate": 3.2673182482759876e-06, + "loss": 0.87010121, + "num_input_tokens_seen": 54245765, + "step": 2521, + "time_per_iteration": 2.8875527381896973 + }, + { + "auxiliary_loss_clip": 0.01174016, + "auxiliary_loss_mlp": 0.01033914, + "balance_loss_clip": 1.02100122, + "balance_loss_mlp": 1.02445519, + "epoch": 0.3032525701917874, + "flos": 18876695650560.0, + "grad_norm": 3.08687236807007, + "language_loss": 0.65722299, + "learning_rate": 3.266715530055659e-06, + "loss": 0.67930233, + "num_input_tokens_seen": 54263915, + "step": 2522, + "time_per_iteration": 2.638690710067749 + }, + { + "auxiliary_loss_clip": 0.01167156, + "auxiliary_loss_mlp": 0.01031993, + "balance_loss_clip": 1.01819992, + "balance_loss_mlp": 1.02255142, + "epoch": 0.30337281308242653, + "flos": 17782641250560.0, + "grad_norm": 2.0295436037686736, + "language_loss": 0.8049444, + "learning_rate": 3.2661126196755927e-06, + "loss": 0.82693589, + "num_input_tokens_seen": 54283025, + "step": 2523, + "time_per_iteration": 2.646498203277588 + }, + { + "auxiliary_loss_clip": 0.01096812, + "auxiliary_loss_mlp": 0.01005003, + "balance_loss_clip": 1.02500522, + "balance_loss_mlp": 1.00320268, + "epoch": 0.3034930559730656, + "flos": 57824298426240.0, + "grad_norm": 0.7838482874305966, + "language_loss": 0.55969042, + "learning_rate": 3.265509517227248e-06, + "loss": 0.5807085, + "num_input_tokens_seen": 54339840, + "step": 2524, + "time_per_iteration": 3.164202928543091 + }, + { + "auxiliary_loss_clip": 0.01150335, + "auxiliary_loss_mlp": 0.01032301, + "balance_loss_clip": 0.97912723, + "balance_loss_mlp": 1.02402818, + "epoch": 0.3036132988637047, + "flos": 14755587419520.0, + "grad_norm": 2.0246795634289954, + "language_loss": 0.81071299, + "learning_rate": 3.264906222802115e-06, + "loss": 0.83253932, + "num_input_tokens_seen": 54357690, + "step": 2525, + "time_per_iteration": 2.627727746963501 + }, + { + "auxiliary_loss_clip": 0.01198921, + "auxiliary_loss_mlp": 0.01030907, + "balance_loss_clip": 1.05839133, + "balance_loss_mlp": 1.02113223, + "epoch": 0.30373354175434375, + "flos": 21033203460480.0, + "grad_norm": 2.2079767114230493, + "language_loss": 0.7815522, + "learning_rate": 3.264302736491715e-06, + "loss": 0.80385047, + "num_input_tokens_seen": 54377810, + "step": 2526, + "time_per_iteration": 2.645803451538086 + }, + { + "auxiliary_loss_clip": 0.0116929, + "auxiliary_loss_mlp": 0.0102843, + "balance_loss_clip": 1.02072787, + "balance_loss_mlp": 1.01900625, + "epoch": 0.30385378464498286, + "flos": 21143233797120.0, + "grad_norm": 1.9040268021610147, + "language_loss": 0.87675428, + "learning_rate": 3.263699058387594e-06, + "loss": 0.89873147, + "num_input_tokens_seen": 54395245, + "step": 2527, + "time_per_iteration": 2.6375632286071777 + }, + { + "auxiliary_loss_clip": 0.01120222, + "auxiliary_loss_mlp": 0.01035722, + "balance_loss_clip": 0.93763375, + "balance_loss_mlp": 1.02665615, + "epoch": 0.30397402753562197, + "flos": 20629244131200.0, + "grad_norm": 2.3925604041185404, + "language_loss": 0.90678757, + "learning_rate": 3.2630951885813315e-06, + "loss": 0.92834705, + "num_input_tokens_seen": 54412640, + "step": 2528, + "time_per_iteration": 2.7320399284362793 + }, + { + "auxiliary_loss_clip": 0.01151264, + "auxiliary_loss_mlp": 0.01029451, + "balance_loss_clip": 0.97913873, + "balance_loss_mlp": 1.02038527, + "epoch": 0.304094270426261, + "flos": 15085678429440.0, + "grad_norm": 1.868536601986616, + "language_loss": 0.77656287, + "learning_rate": 3.262491127164533e-06, + "loss": 0.79837, + "num_input_tokens_seen": 54431455, + "step": 2529, + "time_per_iteration": 2.646764039993286 + }, + { + "auxiliary_loss_clip": 0.01155508, + "auxiliary_loss_mlp": 0.00511604, + "balance_loss_clip": 0.98014438, + "balance_loss_mlp": 0.0, + "epoch": 0.30421451331690014, + "flos": 13845216193920.0, + "grad_norm": 2.3495904790737185, + "language_loss": 0.80221462, + "learning_rate": 3.2618868742288337e-06, + "loss": 0.8188858, + "num_input_tokens_seen": 54448380, + "step": 2530, + "time_per_iteration": 2.6782095432281494 + }, + { + "auxiliary_loss_clip": 0.01171858, + "auxiliary_loss_mlp": 0.01034273, + "balance_loss_clip": 1.01903582, + "balance_loss_mlp": 1.02555895, + "epoch": 0.30433475620753925, + "flos": 17384212615680.0, + "grad_norm": 1.8117577606505586, + "language_loss": 0.72278655, + "learning_rate": 3.261282429865899e-06, + "loss": 0.74484789, + "num_input_tokens_seen": 54466385, + "step": 2531, + "time_per_iteration": 2.6212353706359863 + }, + { + "auxiliary_loss_clip": 0.01155695, + "auxiliary_loss_mlp": 0.00511077, + "balance_loss_clip": 0.98276967, + "balance_loss_mlp": 0.0, + "epoch": 0.3044549990981783, + "flos": 18916951818240.0, + "grad_norm": 1.5254953632194055, + "language_loss": 0.72628236, + "learning_rate": 3.2606777941674225e-06, + "loss": 0.74295008, + "num_input_tokens_seen": 54485040, + "step": 2532, + "time_per_iteration": 2.7477457523345947 + }, + { + "auxiliary_loss_clip": 0.01095235, + "auxiliary_loss_mlp": 0.01025894, + "balance_loss_clip": 0.90479743, + "balance_loss_mlp": 1.01718533, + "epoch": 0.3045752419888174, + "flos": 21068431724160.0, + "grad_norm": 2.5113445567338837, + "language_loss": 0.8455649, + "learning_rate": 3.2600729672251276e-06, + "loss": 0.86677623, + "num_input_tokens_seen": 54502755, + "step": 2533, + "time_per_iteration": 2.7590925693511963 + }, + { + "auxiliary_loss_clip": 0.01198936, + "auxiliary_loss_mlp": 0.00511085, + "balance_loss_clip": 1.06021202, + "balance_loss_mlp": 0.0, + "epoch": 0.3046954848794565, + "flos": 29096405516160.0, + "grad_norm": 2.2613196780980416, + "language_loss": 0.65423888, + "learning_rate": 3.259467949130765e-06, + "loss": 0.67133904, + "num_input_tokens_seen": 54524165, + "step": 2534, + "time_per_iteration": 2.7306933403015137 + }, + { + "auxiliary_loss_clip": 0.01151916, + "auxiliary_loss_mlp": 0.01038706, + "balance_loss_clip": 0.98320019, + "balance_loss_mlp": 1.02978349, + "epoch": 0.3048157277700956, + "flos": 20295346279680.0, + "grad_norm": 2.1004637008822815, + "language_loss": 0.82587993, + "learning_rate": 3.2588627399761164e-06, + "loss": 0.84778607, + "num_input_tokens_seen": 54540160, + "step": 2535, + "time_per_iteration": 2.6820247173309326 + }, + { + "auxiliary_loss_clip": 0.01148282, + "auxiliary_loss_mlp": 0.01031093, + "balance_loss_clip": 0.98081911, + "balance_loss_mlp": 1.02270079, + "epoch": 0.3049359706607347, + "flos": 22739929165440.0, + "grad_norm": 2.6638384193372437, + "language_loss": 0.70950627, + "learning_rate": 3.2582573398529903e-06, + "loss": 0.7313, + "num_input_tokens_seen": 54557515, + "step": 2536, + "time_per_iteration": 3.6301207542419434 + }, + { + "auxiliary_loss_clip": 0.01127954, + "auxiliary_loss_mlp": 0.01034568, + "balance_loss_clip": 0.94248784, + "balance_loss_mlp": 1.02515042, + "epoch": 0.3050562135513738, + "flos": 18434634969600.0, + "grad_norm": 2.244478271050851, + "language_loss": 0.73956609, + "learning_rate": 3.2576517488532265e-06, + "loss": 0.76119131, + "num_input_tokens_seen": 54573865, + "step": 2537, + "time_per_iteration": 2.715399742126465 + }, + { + "auxiliary_loss_clip": 0.01171964, + "auxiliary_loss_mlp": 0.01028162, + "balance_loss_clip": 1.01801872, + "balance_loss_mlp": 1.01960874, + "epoch": 0.30517645644201286, + "flos": 20370327920640.0, + "grad_norm": 1.9294267843525945, + "language_loss": 0.87329388, + "learning_rate": 3.257045967068692e-06, + "loss": 0.8952952, + "num_input_tokens_seen": 54593120, + "step": 2538, + "time_per_iteration": 2.620925188064575 + }, + { + "auxiliary_loss_clip": 0.0119803, + "auxiliary_loss_mlp": 0.01029556, + "balance_loss_clip": 1.05858076, + "balance_loss_mlp": 1.02117002, + "epoch": 0.30529669933265197, + "flos": 21945118970880.0, + "grad_norm": 1.5908475567046658, + "language_loss": 0.82206368, + "learning_rate": 3.2564399945912848e-06, + "loss": 0.84433961, + "num_input_tokens_seen": 54612910, + "step": 2539, + "time_per_iteration": 3.5534377098083496 + }, + { + "auxiliary_loss_clip": 0.01106084, + "auxiliary_loss_mlp": 0.01031322, + "balance_loss_clip": 0.9036634, + "balance_loss_mlp": 1.02319217, + "epoch": 0.305416942223291, + "flos": 21835411856640.0, + "grad_norm": 2.288661196915687, + "language_loss": 0.81912541, + "learning_rate": 3.2558338315129287e-06, + "loss": 0.8404994, + "num_input_tokens_seen": 54631055, + "step": 2540, + "time_per_iteration": 2.754542827606201 + }, + { + "auxiliary_loss_clip": 0.01166727, + "auxiliary_loss_mlp": 0.01031729, + "balance_loss_clip": 1.01923633, + "balance_loss_mlp": 1.02279449, + "epoch": 0.30553718511393013, + "flos": 33911810709120.0, + "grad_norm": 6.784343889562147, + "language_loss": 0.75995308, + "learning_rate": 3.2552274779255785e-06, + "loss": 0.78193766, + "num_input_tokens_seen": 54651985, + "step": 2541, + "time_per_iteration": 3.6648573875427246 + }, + { + "auxiliary_loss_clip": 0.01173402, + "auxiliary_loss_mlp": 0.01034882, + "balance_loss_clip": 1.02027702, + "balance_loss_mlp": 1.02545893, + "epoch": 0.30565742800456924, + "flos": 22268530051200.0, + "grad_norm": 5.260161968133568, + "language_loss": 0.76881611, + "learning_rate": 3.2546209339212184e-06, + "loss": 0.79089898, + "num_input_tokens_seen": 54671005, + "step": 2542, + "time_per_iteration": 2.7510828971862793 + }, + { + "auxiliary_loss_clip": 0.0115112, + "auxiliary_loss_mlp": 0.01032395, + "balance_loss_clip": 0.97973287, + "balance_loss_mlp": 1.02342427, + "epoch": 0.3057776708952083, + "flos": 22565044823040.0, + "grad_norm": 2.4676435636070964, + "language_loss": 0.77669561, + "learning_rate": 3.25401419959186e-06, + "loss": 0.79853076, + "num_input_tokens_seen": 54691615, + "step": 2543, + "time_per_iteration": 2.7108476161956787 + }, + { + "auxiliary_loss_clip": 0.01163058, + "auxiliary_loss_mlp": 0.01037544, + "balance_loss_clip": 0.98652661, + "balance_loss_mlp": 1.02810812, + "epoch": 0.3058979137858474, + "flos": 21799213925760.0, + "grad_norm": 2.070530478586079, + "language_loss": 0.76283181, + "learning_rate": 3.253407275029545e-06, + "loss": 0.7848379, + "num_input_tokens_seen": 54710520, + "step": 2544, + "time_per_iteration": 2.7050042152404785 + }, + { + "auxiliary_loss_clip": 0.0113055, + "auxiliary_loss_mlp": 0.01034559, + "balance_loss_clip": 0.94589204, + "balance_loss_mlp": 1.0259881, + "epoch": 0.3060181566764865, + "flos": 26979435601920.0, + "grad_norm": 2.0285406586330748, + "language_loss": 0.80186266, + "learning_rate": 3.2528001603263425e-06, + "loss": 0.82351375, + "num_input_tokens_seen": 54732590, + "step": 2545, + "time_per_iteration": 3.6875154972076416 + }, + { + "auxiliary_loss_clip": 0.01176558, + "auxiliary_loss_mlp": 0.01032998, + "balance_loss_clip": 1.02462721, + "balance_loss_mlp": 1.02452874, + "epoch": 0.3061383995671256, + "flos": 19865101173120.0, + "grad_norm": 1.9173976811805251, + "language_loss": 0.81601733, + "learning_rate": 3.2521928555743514e-06, + "loss": 0.83811295, + "num_input_tokens_seen": 54749935, + "step": 2546, + "time_per_iteration": 2.649447441101074 + }, + { + "auxiliary_loss_clip": 0.01145222, + "auxiliary_loss_mlp": 0.00511284, + "balance_loss_clip": 0.98140895, + "balance_loss_mlp": 0.0, + "epoch": 0.3062586424577647, + "flos": 22127509255680.0, + "grad_norm": 2.9768423073995116, + "language_loss": 0.6753335, + "learning_rate": 3.2515853608657e-06, + "loss": 0.69189858, + "num_input_tokens_seen": 54767935, + "step": 2547, + "time_per_iteration": 2.7528042793273926 + }, + { + "auxiliary_loss_clip": 0.01170288, + "auxiliary_loss_mlp": 0.01032099, + "balance_loss_clip": 1.02018523, + "balance_loss_mlp": 1.02312887, + "epoch": 0.3063788853484038, + "flos": 20845497962880.0, + "grad_norm": 2.2538050500869016, + "language_loss": 0.74678707, + "learning_rate": 3.250977676292545e-06, + "loss": 0.76881093, + "num_input_tokens_seen": 54786175, + "step": 2548, + "time_per_iteration": 2.6490561962127686 + }, + { + "auxiliary_loss_clip": 0.01152773, + "auxiliary_loss_mlp": 0.01026306, + "balance_loss_clip": 0.98235184, + "balance_loss_mlp": 1.01668537, + "epoch": 0.30649912823904285, + "flos": 16209717707520.0, + "grad_norm": 2.0136308768153253, + "language_loss": 0.78908449, + "learning_rate": 3.2503698019470712e-06, + "loss": 0.8108753, + "num_input_tokens_seen": 54801945, + "step": 2549, + "time_per_iteration": 2.683152198791504 + }, + { + "auxiliary_loss_clip": 0.01171932, + "auxiliary_loss_mlp": 0.01036302, + "balance_loss_clip": 1.01791871, + "balance_loss_mlp": 1.02736151, + "epoch": 0.30661937112968196, + "flos": 18617815353600.0, + "grad_norm": 2.2770711780284967, + "language_loss": 0.78440154, + "learning_rate": 3.249761737921492e-06, + "loss": 0.80648386, + "num_input_tokens_seen": 54818475, + "step": 2550, + "time_per_iteration": 2.6190438270568848 + }, + { + "auxiliary_loss_clip": 0.01149914, + "auxiliary_loss_mlp": 0.01028465, + "balance_loss_clip": 0.98548645, + "balance_loss_mlp": 1.01967382, + "epoch": 0.30673961402032107, + "flos": 31390809638400.0, + "grad_norm": 1.9940353169514924, + "language_loss": 0.74217379, + "learning_rate": 3.249153484308051e-06, + "loss": 0.76395762, + "num_input_tokens_seen": 54837090, + "step": 2551, + "time_per_iteration": 2.7006802558898926 + }, + { + "auxiliary_loss_clip": 0.01091617, + "auxiliary_loss_mlp": 0.01030526, + "balance_loss_clip": 0.90234864, + "balance_loss_mlp": 1.0214597, + "epoch": 0.3068598569109601, + "flos": 20229809915520.0, + "grad_norm": 2.0144475010091174, + "language_loss": 0.77821791, + "learning_rate": 3.2485450411990194e-06, + "loss": 0.79943931, + "num_input_tokens_seen": 54856445, + "step": 2552, + "time_per_iteration": 2.7286555767059326 + }, + { + "auxiliary_loss_clip": 0.01197085, + "auxiliary_loss_mlp": 0.01029749, + "balance_loss_clip": 1.05638766, + "balance_loss_mlp": 1.02087379, + "epoch": 0.30698009980159924, + "flos": 29601991399680.0, + "grad_norm": 1.6349179805826095, + "language_loss": 0.82379544, + "learning_rate": 3.2479364086866983e-06, + "loss": 0.84606373, + "num_input_tokens_seen": 54876700, + "step": 2553, + "time_per_iteration": 2.7251670360565186 + }, + { + "auxiliary_loss_clip": 0.01153838, + "auxiliary_loss_mlp": 0.00511439, + "balance_loss_clip": 0.98465157, + "balance_loss_mlp": 0.0, + "epoch": 0.30710034269223835, + "flos": 23842423261440.0, + "grad_norm": 1.750361680552121, + "language_loss": 0.81488526, + "learning_rate": 3.247327586863416e-06, + "loss": 0.83153802, + "num_input_tokens_seen": 54897580, + "step": 2554, + "time_per_iteration": 2.721520185470581 + }, + { + "auxiliary_loss_clip": 0.01134031, + "auxiliary_loss_mlp": 0.01030022, + "balance_loss_clip": 0.94459277, + "balance_loss_mlp": 1.02099228, + "epoch": 0.3072205855828774, + "flos": 25884986152320.0, + "grad_norm": 2.0950811018451017, + "language_loss": 0.77042973, + "learning_rate": 3.2467185758215304e-06, + "loss": 0.79207027, + "num_input_tokens_seen": 54917320, + "step": 2555, + "time_per_iteration": 2.787914752960205 + }, + { + "auxiliary_loss_clip": 0.01131286, + "auxiliary_loss_mlp": 0.00511341, + "balance_loss_clip": 0.94583082, + "balance_loss_mlp": 0.0, + "epoch": 0.3073408284735165, + "flos": 22236390357120.0, + "grad_norm": 7.789852317762196, + "language_loss": 0.8556053, + "learning_rate": 3.246109375653428e-06, + "loss": 0.87203157, + "num_input_tokens_seen": 54934085, + "step": 2556, + "time_per_iteration": 2.7037301063537598 + }, + { + "auxiliary_loss_clip": 0.0119815, + "auxiliary_loss_mlp": 0.0103299, + "balance_loss_clip": 1.05857122, + "balance_loss_mlp": 1.02344191, + "epoch": 0.30746107136415557, + "flos": 19500284689920.0, + "grad_norm": 2.4010466877218275, + "language_loss": 0.78339005, + "learning_rate": 3.2454999864515243e-06, + "loss": 0.80570149, + "num_input_tokens_seen": 54953460, + "step": 2557, + "time_per_iteration": 2.5691776275634766 + }, + { + "auxiliary_loss_clip": 0.01143287, + "auxiliary_loss_mlp": 0.0051109, + "balance_loss_clip": 0.97993004, + "balance_loss_mlp": 0.0, + "epoch": 0.3075813142547947, + "flos": 21724806902400.0, + "grad_norm": 1.7940836946975882, + "language_loss": 0.69608021, + "learning_rate": 3.244890408308263e-06, + "loss": 0.71262389, + "num_input_tokens_seen": 54974165, + "step": 2558, + "time_per_iteration": 2.683511257171631 + }, + { + "auxiliary_loss_clip": 0.01105323, + "auxiliary_loss_mlp": 0.01030601, + "balance_loss_clip": 0.90430486, + "balance_loss_mlp": 1.0216186, + "epoch": 0.3077015571454338, + "flos": 24097963593600.0, + "grad_norm": 1.8947030436181882, + "language_loss": 0.61281407, + "learning_rate": 3.2442806413161165e-06, + "loss": 0.63417327, + "num_input_tokens_seen": 54993810, + "step": 2559, + "time_per_iteration": 2.7688136100769043 + }, + { + "auxiliary_loss_clip": 0.01106814, + "auxiliary_loss_mlp": 0.01031492, + "balance_loss_clip": 0.90327948, + "balance_loss_mlp": 1.02271867, + "epoch": 0.30782180003607285, + "flos": 18405476104320.0, + "grad_norm": 2.081540899847383, + "language_loss": 0.76041037, + "learning_rate": 3.243670685567586e-06, + "loss": 0.78179342, + "num_input_tokens_seen": 55011210, + "step": 2560, + "time_per_iteration": 2.7186567783355713 + }, + { + "auxiliary_loss_clip": 0.01151017, + "auxiliary_loss_mlp": 0.00510668, + "balance_loss_clip": 0.9832477, + "balance_loss_mlp": 0.0, + "epoch": 0.30794204292671196, + "flos": 23878549365120.0, + "grad_norm": 2.274893314154954, + "language_loss": 0.80152261, + "learning_rate": 3.2430605411552012e-06, + "loss": 0.81813943, + "num_input_tokens_seen": 55031325, + "step": 2561, + "time_per_iteration": 2.7026095390319824 + }, + { + "auxiliary_loss_clip": 0.01037358, + "auxiliary_loss_mlp": 0.01002611, + "balance_loss_clip": 0.91249526, + "balance_loss_mlp": 1.00068021, + "epoch": 0.30806228581735107, + "flos": 67927800816000.0, + "grad_norm": 0.9834068275532241, + "language_loss": 0.70595068, + "learning_rate": 3.2424502081715205e-06, + "loss": 0.72635037, + "num_input_tokens_seen": 55094440, + "step": 2562, + "time_per_iteration": 4.2339026927948 + }, + { + "auxiliary_loss_clip": 0.01150177, + "auxiliary_loss_mlp": 0.01029654, + "balance_loss_clip": 0.97996557, + "balance_loss_mlp": 1.02010512, + "epoch": 0.3081825287079901, + "flos": 23843213360640.0, + "grad_norm": 1.6361450770132469, + "language_loss": 0.77829838, + "learning_rate": 3.241839686709132e-06, + "loss": 0.80009663, + "num_input_tokens_seen": 55115375, + "step": 2563, + "time_per_iteration": 2.6906163692474365 + }, + { + "auxiliary_loss_clip": 0.01169115, + "auxiliary_loss_mlp": 0.01034494, + "balance_loss_clip": 1.01711905, + "balance_loss_mlp": 1.02491593, + "epoch": 0.30830277159862923, + "flos": 16209969102720.0, + "grad_norm": 2.519121562014963, + "language_loss": 0.81904036, + "learning_rate": 3.2412289768606495e-06, + "loss": 0.84107643, + "num_input_tokens_seen": 55131945, + "step": 2564, + "time_per_iteration": 2.702601671218872 + }, + { + "auxiliary_loss_clip": 0.0117613, + "auxiliary_loss_mlp": 0.01029754, + "balance_loss_clip": 1.02003777, + "balance_loss_mlp": 1.0207715, + "epoch": 0.30842301448926834, + "flos": 29349503723520.0, + "grad_norm": 2.045587838146282, + "language_loss": 0.82476354, + "learning_rate": 3.240618078718718e-06, + "loss": 0.84682238, + "num_input_tokens_seen": 55153405, + "step": 2565, + "time_per_iteration": 3.6392476558685303 + }, + { + "auxiliary_loss_clip": 0.0112357, + "auxiliary_loss_mlp": 0.01032203, + "balance_loss_clip": 0.94050586, + "balance_loss_mlp": 1.02291644, + "epoch": 0.3085432573799074, + "flos": 21945190798080.0, + "grad_norm": 5.414775285812832, + "language_loss": 0.74048066, + "learning_rate": 3.240006992376011e-06, + "loss": 0.76203841, + "num_input_tokens_seen": 55173030, + "step": 2566, + "time_per_iteration": 3.6268866062164307 + }, + { + "auxiliary_loss_clip": 0.01155294, + "auxiliary_loss_mlp": 0.01031924, + "balance_loss_clip": 0.98300278, + "balance_loss_mlp": 1.02288198, + "epoch": 0.3086635002705465, + "flos": 22054718344320.0, + "grad_norm": 2.1775996543712375, + "language_loss": 0.76032925, + "learning_rate": 3.2393957179252284e-06, + "loss": 0.78220141, + "num_input_tokens_seen": 55189565, + "step": 2567, + "time_per_iteration": 2.664773941040039 + }, + { + "auxiliary_loss_clip": 0.01201407, + "auxiliary_loss_mlp": 0.01031943, + "balance_loss_clip": 1.06144369, + "balance_loss_mlp": 1.02156591, + "epoch": 0.3087837431611856, + "flos": 32665925520000.0, + "grad_norm": 1.8403655291570884, + "language_loss": 0.80554509, + "learning_rate": 3.2387842554591016e-06, + "loss": 0.82787859, + "num_input_tokens_seen": 55210380, + "step": 2568, + "time_per_iteration": 2.6974403858184814 + }, + { + "auxiliary_loss_clip": 0.01201781, + "auxiliary_loss_mlp": 0.01031019, + "balance_loss_clip": 1.06179285, + "balance_loss_mlp": 1.0224421, + "epoch": 0.3089039860518247, + "flos": 17599245384960.0, + "grad_norm": 2.307887954901662, + "language_loss": 0.87767178, + "learning_rate": 3.238172605070388e-06, + "loss": 0.89999986, + "num_input_tokens_seen": 55225795, + "step": 2569, + "time_per_iteration": 2.5816121101379395 + }, + { + "auxiliary_loss_clip": 0.0117455, + "auxiliary_loss_mlp": 0.00512108, + "balance_loss_clip": 1.0215106, + "balance_loss_mlp": 0.0, + "epoch": 0.3090242289424638, + "flos": 14383839611520.0, + "grad_norm": 3.5696649716182964, + "language_loss": 0.7833665, + "learning_rate": 3.2375607668518745e-06, + "loss": 0.80023307, + "num_input_tokens_seen": 55238830, + "step": 2570, + "time_per_iteration": 2.627117872238159 + }, + { + "auxiliary_loss_clip": 0.01142206, + "auxiliary_loss_mlp": 0.01028039, + "balance_loss_clip": 0.9815411, + "balance_loss_mlp": 1.01897931, + "epoch": 0.30914447183310284, + "flos": 16068625084800.0, + "grad_norm": 2.865640680243947, + "language_loss": 0.89834201, + "learning_rate": 3.236948740896377e-06, + "loss": 0.92004448, + "num_input_tokens_seen": 55253630, + "step": 2571, + "time_per_iteration": 3.619610548019409 + }, + { + "auxiliary_loss_clip": 0.01175996, + "auxiliary_loss_mlp": 0.01032997, + "balance_loss_clip": 1.02183616, + "balance_loss_mlp": 1.02397323, + "epoch": 0.30926471472374195, + "flos": 32230221546240.0, + "grad_norm": 1.6241092659522036, + "language_loss": 0.84143651, + "learning_rate": 3.2363365272967384e-06, + "loss": 0.86352646, + "num_input_tokens_seen": 55276200, + "step": 2572, + "time_per_iteration": 2.723806381225586 + }, + { + "auxiliary_loss_clip": 0.01175512, + "auxiliary_loss_mlp": 0.01036281, + "balance_loss_clip": 1.02335644, + "balance_loss_mlp": 1.02682185, + "epoch": 0.30938495761438106, + "flos": 20370722970240.0, + "grad_norm": 1.9530920436744066, + "language_loss": 0.81597054, + "learning_rate": 3.235724126145832e-06, + "loss": 0.83808851, + "num_input_tokens_seen": 55292235, + "step": 2573, + "time_per_iteration": 2.6337804794311523 + }, + { + "auxiliary_loss_clip": 0.01166389, + "auxiliary_loss_mlp": 0.01033936, + "balance_loss_clip": 1.01786661, + "balance_loss_mlp": 1.02500105, + "epoch": 0.3095052005050201, + "flos": 24061155131520.0, + "grad_norm": 1.5742532854292153, + "language_loss": 0.7737487, + "learning_rate": 3.235111537536558e-06, + "loss": 0.79575193, + "num_input_tokens_seen": 55313050, + "step": 2574, + "time_per_iteration": 2.7274763584136963 + }, + { + "auxiliary_loss_clip": 0.01175116, + "auxiliary_loss_mlp": 0.0103206, + "balance_loss_clip": 1.02157903, + "balance_loss_mlp": 1.02350116, + "epoch": 0.30962544339565923, + "flos": 23401547729280.0, + "grad_norm": 1.821918656151207, + "language_loss": 0.82989687, + "learning_rate": 3.2344987615618456e-06, + "loss": 0.85196865, + "num_input_tokens_seen": 55332885, + "step": 2575, + "time_per_iteration": 2.74831223487854 + }, + { + "auxiliary_loss_clip": 0.01131172, + "auxiliary_loss_mlp": 0.01027637, + "balance_loss_clip": 0.94886196, + "balance_loss_mlp": 1.01895857, + "epoch": 0.30974568628629834, + "flos": 33799984692480.0, + "grad_norm": 1.702955641876482, + "language_loss": 0.78637725, + "learning_rate": 3.2338857983146533e-06, + "loss": 0.80796534, + "num_input_tokens_seen": 55354385, + "step": 2576, + "time_per_iteration": 2.8213577270507812 + }, + { + "auxiliary_loss_clip": 0.01143779, + "auxiliary_loss_mlp": 0.01033275, + "balance_loss_clip": 0.9830249, + "balance_loss_mlp": 1.02431631, + "epoch": 0.3098659291769374, + "flos": 20229594433920.0, + "grad_norm": 2.0903433603964334, + "language_loss": 0.76336682, + "learning_rate": 3.233272647887966e-06, + "loss": 0.7851373, + "num_input_tokens_seen": 55373275, + "step": 2577, + "time_per_iteration": 2.7550928592681885 + }, + { + "auxiliary_loss_clip": 0.01203294, + "auxiliary_loss_mlp": 0.01029193, + "balance_loss_clip": 1.06305242, + "balance_loss_mlp": 1.02025843, + "epoch": 0.3099861720675765, + "flos": 24748556682240.0, + "grad_norm": 1.7753797596175618, + "language_loss": 0.90157127, + "learning_rate": 3.2326593103747985e-06, + "loss": 0.92389613, + "num_input_tokens_seen": 55392290, + "step": 2578, + "time_per_iteration": 2.6796393394470215 + }, + { + "auxiliary_loss_clip": 0.0117787, + "auxiliary_loss_mlp": 0.01025364, + "balance_loss_clip": 1.02572346, + "balance_loss_mlp": 1.01639986, + "epoch": 0.3101064149582156, + "flos": 11765485704960.0, + "grad_norm": 2.9548640774621844, + "language_loss": 0.84869212, + "learning_rate": 3.2320457858681936e-06, + "loss": 0.87072444, + "num_input_tokens_seen": 55410680, + "step": 2579, + "time_per_iteration": 2.6170401573181152 + }, + { + "auxiliary_loss_clip": 0.01156139, + "auxiliary_loss_mlp": 0.01030074, + "balance_loss_clip": 0.98553896, + "balance_loss_mlp": 1.02138376, + "epoch": 0.31022665784885467, + "flos": 23033247626880.0, + "grad_norm": 2.77249140421397, + "language_loss": 0.85100412, + "learning_rate": 3.2314320744612228e-06, + "loss": 0.87286627, + "num_input_tokens_seen": 55425980, + "step": 2580, + "time_per_iteration": 2.6635265350341797 + }, + { + "auxiliary_loss_clip": 0.01173261, + "auxiliary_loss_mlp": 0.01031176, + "balance_loss_clip": 1.02201331, + "balance_loss_mlp": 1.02262878, + "epoch": 0.3103469007394938, + "flos": 16289188548480.0, + "grad_norm": 2.0698448283022435, + "language_loss": 0.76450312, + "learning_rate": 3.2308181762469854e-06, + "loss": 0.78654748, + "num_input_tokens_seen": 55443925, + "step": 2581, + "time_per_iteration": 2.63572096824646 + }, + { + "auxiliary_loss_clip": 0.01203433, + "auxiliary_loss_mlp": 0.01030517, + "balance_loss_clip": 1.06037045, + "balance_loss_mlp": 1.02184463, + "epoch": 0.3104671436301329, + "flos": 30515271626880.0, + "grad_norm": 6.109770480994256, + "language_loss": 0.78237242, + "learning_rate": 3.230204091318609e-06, + "loss": 0.80471188, + "num_input_tokens_seen": 55464465, + "step": 2582, + "time_per_iteration": 2.644615650177002 + }, + { + "auxiliary_loss_clip": 0.01196612, + "auxiliary_loss_mlp": 0.00511525, + "balance_loss_clip": 1.05840957, + "balance_loss_mlp": 0.0, + "epoch": 0.31058738652077195, + "flos": 20047240062720.0, + "grad_norm": 3.743376541709189, + "language_loss": 0.84256405, + "learning_rate": 3.2295898197692503e-06, + "loss": 0.85964543, + "num_input_tokens_seen": 55483425, + "step": 2583, + "time_per_iteration": 2.6124448776245117 + }, + { + "auxiliary_loss_clip": 0.01195882, + "auxiliary_loss_mlp": 0.01030167, + "balance_loss_clip": 1.05831099, + "balance_loss_mlp": 1.02176332, + "epoch": 0.31070762941141106, + "flos": 28074639237120.0, + "grad_norm": 2.1814039063759534, + "language_loss": 0.79295278, + "learning_rate": 3.228975361692094e-06, + "loss": 0.81521326, + "num_input_tokens_seen": 55504445, + "step": 2584, + "time_per_iteration": 2.683105707168579 + }, + { + "auxiliary_loss_clip": 0.01179732, + "auxiliary_loss_mlp": 0.0051217, + "balance_loss_clip": 1.0200634, + "balance_loss_mlp": 0.0, + "epoch": 0.31082787230205017, + "flos": 20521907314560.0, + "grad_norm": 2.597028787424963, + "language_loss": 0.79889709, + "learning_rate": 3.228360717180352e-06, + "loss": 0.81581604, + "num_input_tokens_seen": 55521970, + "step": 2585, + "time_per_iteration": 2.7354509830474854 + }, + { + "auxiliary_loss_clip": 0.01097965, + "auxiliary_loss_mlp": 0.00502403, + "balance_loss_clip": 1.02889037, + "balance_loss_mlp": 0.0, + "epoch": 0.3109481151926892, + "flos": 62445928723200.0, + "grad_norm": 0.8107688570202121, + "language_loss": 0.59379339, + "learning_rate": 3.227745886327266e-06, + "loss": 0.60979706, + "num_input_tokens_seen": 55580665, + "step": 2586, + "time_per_iteration": 3.1241016387939453 + }, + { + "auxiliary_loss_clip": 0.0109719, + "auxiliary_loss_mlp": 0.01002714, + "balance_loss_clip": 1.02796423, + "balance_loss_mlp": 1.00090158, + "epoch": 0.31106835808332833, + "flos": 44746744723200.0, + "grad_norm": 0.8046957276929353, + "language_loss": 0.55818486, + "learning_rate": 3.227130869226105e-06, + "loss": 0.57918382, + "num_input_tokens_seen": 55637825, + "step": 2587, + "time_per_iteration": 3.156639337539673 + }, + { + "auxiliary_loss_clip": 0.01175016, + "auxiliary_loss_mlp": 0.0103408, + "balance_loss_clip": 1.01916695, + "balance_loss_mlp": 1.02426314, + "epoch": 0.3111886009739674, + "flos": 23403056100480.0, + "grad_norm": 2.4851316182572254, + "language_loss": 0.82414579, + "learning_rate": 3.226515665970167e-06, + "loss": 0.84623677, + "num_input_tokens_seen": 55655365, + "step": 2588, + "time_per_iteration": 3.5519204139709473 + }, + { + "auxiliary_loss_clip": 0.01174158, + "auxiliary_loss_mlp": 0.01034834, + "balance_loss_clip": 1.02211618, + "balance_loss_mlp": 1.02536917, + "epoch": 0.3113088438646065, + "flos": 17530728192000.0, + "grad_norm": 2.5946419330596675, + "language_loss": 0.86619771, + "learning_rate": 3.225900276652777e-06, + "loss": 0.88828766, + "num_input_tokens_seen": 55672140, + "step": 2589, + "time_per_iteration": 2.6531643867492676 + }, + { + "auxiliary_loss_clip": 0.01158048, + "auxiliary_loss_mlp": 0.01030303, + "balance_loss_clip": 0.98267829, + "balance_loss_mlp": 1.02156472, + "epoch": 0.3114290867552456, + "flos": 28365802882560.0, + "grad_norm": 1.669608663141129, + "language_loss": 0.75236511, + "learning_rate": 3.2252847013672906e-06, + "loss": 0.77424866, + "num_input_tokens_seen": 55694800, + "step": 2590, + "time_per_iteration": 2.8031556606292725 + }, + { + "auxiliary_loss_clip": 0.01118925, + "auxiliary_loss_mlp": 0.01027355, + "balance_loss_clip": 0.93992126, + "balance_loss_mlp": 1.01781249, + "epoch": 0.31154932964588467, + "flos": 27379157126400.0, + "grad_norm": 2.375292278759193, + "language_loss": 0.75686252, + "learning_rate": 3.224668940207089e-06, + "loss": 0.77832532, + "num_input_tokens_seen": 55713785, + "step": 2591, + "time_per_iteration": 3.684377670288086 + }, + { + "auxiliary_loss_clip": 0.01096941, + "auxiliary_loss_mlp": 0.0102897, + "balance_loss_clip": 0.90120298, + "balance_loss_mlp": 1.019171, + "epoch": 0.3116695725365238, + "flos": 26541864120960.0, + "grad_norm": 3.1830640060757065, + "language_loss": 0.86749196, + "learning_rate": 3.2240529932655828e-06, + "loss": 0.88875115, + "num_input_tokens_seen": 55733050, + "step": 2592, + "time_per_iteration": 3.709407329559326 + }, + { + "auxiliary_loss_clip": 0.01149669, + "auxiliary_loss_mlp": 0.01030231, + "balance_loss_clip": 0.98403478, + "balance_loss_mlp": 1.02095687, + "epoch": 0.3117898154271629, + "flos": 21177600134400.0, + "grad_norm": 7.371033213442879, + "language_loss": 0.88653934, + "learning_rate": 3.223436860636211e-06, + "loss": 0.90833831, + "num_input_tokens_seen": 55748685, + "step": 2593, + "time_per_iteration": 2.6225810050964355 + }, + { + "auxiliary_loss_clip": 0.01199141, + "auxiliary_loss_mlp": 0.0102782, + "balance_loss_clip": 1.06052017, + "balance_loss_mlp": 1.01917136, + "epoch": 0.31191005831780194, + "flos": 27272430840960.0, + "grad_norm": 1.6603026355500525, + "language_loss": 0.74126327, + "learning_rate": 3.2228205424124403e-06, + "loss": 0.76353288, + "num_input_tokens_seen": 55771840, + "step": 2594, + "time_per_iteration": 2.6852285861968994 + }, + { + "auxiliary_loss_clip": 0.01138444, + "auxiliary_loss_mlp": 0.01032084, + "balance_loss_clip": 0.97942924, + "balance_loss_mlp": 1.02316678, + "epoch": 0.31203030120844105, + "flos": 12963501043200.0, + "grad_norm": 2.5494582587343695, + "language_loss": 0.74897349, + "learning_rate": 3.222204038687765e-06, + "loss": 0.77067882, + "num_input_tokens_seen": 55784975, + "step": 2595, + "time_per_iteration": 2.5984582901000977 + }, + { + "auxiliary_loss_clip": 0.01172492, + "auxiliary_loss_mlp": 0.01028515, + "balance_loss_clip": 1.02141142, + "balance_loss_mlp": 1.01986051, + "epoch": 0.31215054409908016, + "flos": 27562014288000.0, + "grad_norm": 1.6720791675491449, + "language_loss": 0.88160372, + "learning_rate": 3.221587349555709e-06, + "loss": 0.90361375, + "num_input_tokens_seen": 55805235, + "step": 2596, + "time_per_iteration": 2.680220603942871 + }, + { + "auxiliary_loss_clip": 0.01151384, + "auxiliary_loss_mlp": 0.01027166, + "balance_loss_clip": 0.98091674, + "balance_loss_mlp": 1.01795697, + "epoch": 0.3122707869897192, + "flos": 21506326427520.0, + "grad_norm": 2.3198314198766545, + "language_loss": 0.68981528, + "learning_rate": 3.2209704751098236e-06, + "loss": 0.71160078, + "num_input_tokens_seen": 55824265, + "step": 2597, + "time_per_iteration": 3.640049934387207 + }, + { + "auxiliary_loss_clip": 0.0115367, + "auxiliary_loss_mlp": 0.01029018, + "balance_loss_clip": 0.98426884, + "balance_loss_mlp": 1.01988029, + "epoch": 0.31239102988035833, + "flos": 15187017674880.0, + "grad_norm": 2.003952141113087, + "language_loss": 0.82358241, + "learning_rate": 3.2203534154436875e-06, + "loss": 0.84540939, + "num_input_tokens_seen": 55838620, + "step": 2598, + "time_per_iteration": 2.6537880897521973 + }, + { + "auxiliary_loss_clip": 0.01075483, + "auxiliary_loss_mlp": 0.01031795, + "balance_loss_clip": 0.8637985, + "balance_loss_mlp": 1.0229857, + "epoch": 0.31251127277099744, + "flos": 22053712763520.0, + "grad_norm": 4.2309890653160105, + "language_loss": 0.75528669, + "learning_rate": 3.2197361706509084e-06, + "loss": 0.7763595, + "num_input_tokens_seen": 55859375, + "step": 2599, + "time_per_iteration": 2.801297187805176 + }, + { + "auxiliary_loss_clip": 0.01200078, + "auxiliary_loss_mlp": 0.0102996, + "balance_loss_clip": 1.05917645, + "balance_loss_mlp": 1.02068591, + "epoch": 0.3126315156616365, + "flos": 15193984913280.0, + "grad_norm": 2.849494138926706, + "language_loss": 0.8389523, + "learning_rate": 3.2191187408251228e-06, + "loss": 0.86125267, + "num_input_tokens_seen": 55876535, + "step": 2600, + "time_per_iteration": 2.5664048194885254 + }, + { + "auxiliary_loss_clip": 0.01179745, + "auxiliary_loss_mlp": 0.01037876, + "balance_loss_clip": 1.01964593, + "balance_loss_mlp": 1.02808833, + "epoch": 0.3127517585522756, + "flos": 18145338831360.0, + "grad_norm": 2.595607014763045, + "language_loss": 0.78700918, + "learning_rate": 3.218501126059993e-06, + "loss": 0.80918539, + "num_input_tokens_seen": 55891930, + "step": 2601, + "time_per_iteration": 2.6453773975372314 + }, + { + "auxiliary_loss_clip": 0.01172445, + "auxiliary_loss_mlp": 0.01031358, + "balance_loss_clip": 1.01645052, + "balance_loss_mlp": 1.02239978, + "epoch": 0.31287200144291466, + "flos": 21908633731200.0, + "grad_norm": 2.61734506588288, + "language_loss": 0.8125177, + "learning_rate": 3.2178833264492116e-06, + "loss": 0.83455575, + "num_input_tokens_seen": 55910635, + "step": 2602, + "time_per_iteration": 2.6889052391052246 + }, + { + "auxiliary_loss_clip": 0.01181507, + "auxiliary_loss_mlp": 0.01034762, + "balance_loss_clip": 1.021384, + "balance_loss_mlp": 1.02523685, + "epoch": 0.31299224433355377, + "flos": 29896997800320.0, + "grad_norm": 3.7265961722083274, + "language_loss": 0.7612288, + "learning_rate": 3.217265342086498e-06, + "loss": 0.78339148, + "num_input_tokens_seen": 55931125, + "step": 2603, + "time_per_iteration": 2.7187602519989014 + }, + { + "auxiliary_loss_clip": 0.01127836, + "auxiliary_loss_mlp": 0.00512415, + "balance_loss_clip": 0.94371194, + "balance_loss_mlp": 0.0, + "epoch": 0.3131124872241929, + "flos": 11655886331520.0, + "grad_norm": 2.3731037506686796, + "language_loss": 0.73144555, + "learning_rate": 3.216647173065599e-06, + "loss": 0.74784803, + "num_input_tokens_seen": 55946590, + "step": 2604, + "time_per_iteration": 2.78143572807312 + }, + { + "auxiliary_loss_clip": 0.01146736, + "auxiliary_loss_mlp": 0.01027874, + "balance_loss_clip": 0.98267925, + "balance_loss_mlp": 1.0192548, + "epoch": 0.31323273011483194, + "flos": 49848785470080.0, + "grad_norm": 1.9826933683770869, + "language_loss": 0.73904908, + "learning_rate": 3.216028819480292e-06, + "loss": 0.76079524, + "num_input_tokens_seen": 55967930, + "step": 2605, + "time_per_iteration": 2.965912103652954 + }, + { + "auxiliary_loss_clip": 0.01139847, + "auxiliary_loss_mlp": 0.01026762, + "balance_loss_clip": 0.9798218, + "balance_loss_mlp": 1.01708198, + "epoch": 0.31335297300547105, + "flos": 22601278667520.0, + "grad_norm": 2.061937356487575, + "language_loss": 0.75325912, + "learning_rate": 3.2154102814243793e-06, + "loss": 0.77492523, + "num_input_tokens_seen": 55987070, + "step": 2606, + "time_per_iteration": 2.7703311443328857 + }, + { + "auxiliary_loss_clip": 0.01130328, + "auxiliary_loss_mlp": 0.01031057, + "balance_loss_clip": 0.94305682, + "balance_loss_mlp": 1.02259922, + "epoch": 0.31347321589611016, + "flos": 34710858708480.0, + "grad_norm": 2.5153113039863877, + "language_loss": 0.66880071, + "learning_rate": 3.2147915589916937e-06, + "loss": 0.69041455, + "num_input_tokens_seen": 56008630, + "step": 2607, + "time_per_iteration": 2.8664722442626953 + }, + { + "auxiliary_loss_clip": 0.01144133, + "auxiliary_loss_mlp": 0.01037875, + "balance_loss_clip": 0.97863686, + "balance_loss_mlp": 1.02844572, + "epoch": 0.3135934587867492, + "flos": 19755789108480.0, + "grad_norm": 4.3960652873689074, + "language_loss": 0.82820213, + "learning_rate": 3.2141726522760938e-06, + "loss": 0.85002226, + "num_input_tokens_seen": 56026690, + "step": 2608, + "time_per_iteration": 2.723442554473877 + }, + { + "auxiliary_loss_clip": 0.01072854, + "auxiliary_loss_mlp": 0.01004831, + "balance_loss_clip": 0.98832029, + "balance_loss_mlp": 1.00312638, + "epoch": 0.3137137016773883, + "flos": 65815535583360.0, + "grad_norm": 0.7033727815726663, + "language_loss": 0.5263592, + "learning_rate": 3.213553561371469e-06, + "loss": 0.54713607, + "num_input_tokens_seen": 56090425, + "step": 2609, + "time_per_iteration": 3.33528208732605 + }, + { + "auxiliary_loss_clip": 0.01101523, + "auxiliary_loss_mlp": 0.01033419, + "balance_loss_clip": 0.90672803, + "balance_loss_mlp": 1.02481198, + "epoch": 0.31383394456802743, + "flos": 16252739222400.0, + "grad_norm": 2.7646489392074676, + "language_loss": 0.95999455, + "learning_rate": 3.212934286371733e-06, + "loss": 0.98134392, + "num_input_tokens_seen": 56107135, + "step": 2610, + "time_per_iteration": 2.7355170249938965 + }, + { + "auxiliary_loss_clip": 0.01171687, + "auxiliary_loss_mlp": 0.01036685, + "balance_loss_clip": 1.021384, + "balance_loss_mlp": 1.02760744, + "epoch": 0.3139541874586665, + "flos": 38795517613440.0, + "grad_norm": 2.27683241151414, + "language_loss": 0.83272505, + "learning_rate": 3.2123148273708304e-06, + "loss": 0.85480881, + "num_input_tokens_seen": 56127325, + "step": 2611, + "time_per_iteration": 2.7681427001953125 + }, + { + "auxiliary_loss_clip": 0.01195698, + "auxiliary_loss_mlp": 0.01030301, + "balance_loss_clip": 1.05788171, + "balance_loss_mlp": 1.02189088, + "epoch": 0.3140744303493056, + "flos": 25046328430080.0, + "grad_norm": 1.8224110563621088, + "language_loss": 0.76681852, + "learning_rate": 3.211695184462733e-06, + "loss": 0.78907859, + "num_input_tokens_seen": 56148500, + "step": 2612, + "time_per_iteration": 2.6710879802703857 + }, + { + "auxiliary_loss_clip": 0.0103499, + "auxiliary_loss_mlp": 0.01002967, + "balance_loss_clip": 0.91438496, + "balance_loss_mlp": 1.00109565, + "epoch": 0.3141946732399447, + "flos": 72504254782080.0, + "grad_norm": 0.9040248668053219, + "language_loss": 0.60429168, + "learning_rate": 3.2110753577414383e-06, + "loss": 0.62467122, + "num_input_tokens_seen": 56210080, + "step": 2613, + "time_per_iteration": 3.26128888130188 + }, + { + "auxiliary_loss_clip": 0.01148964, + "auxiliary_loss_mlp": 0.0102969, + "balance_loss_clip": 0.97810525, + "balance_loss_mlp": 1.02075481, + "epoch": 0.31431491613058377, + "flos": 19239788280960.0, + "grad_norm": 2.580376658144217, + "language_loss": 0.78664243, + "learning_rate": 3.2104553473009757e-06, + "loss": 0.80842888, + "num_input_tokens_seen": 56228200, + "step": 2614, + "time_per_iteration": 3.593764543533325 + }, + { + "auxiliary_loss_clip": 0.01096507, + "auxiliary_loss_mlp": 0.01033532, + "balance_loss_clip": 0.9024843, + "balance_loss_mlp": 1.02446651, + "epoch": 0.3144351590212229, + "flos": 36210596290560.0, + "grad_norm": 2.072318764587854, + "language_loss": 0.67559505, + "learning_rate": 3.209835153235399e-06, + "loss": 0.69689548, + "num_input_tokens_seen": 56249755, + "step": 2615, + "time_per_iteration": 2.860692024230957 + }, + { + "auxiliary_loss_clip": 0.01117729, + "auxiliary_loss_mlp": 0.01034111, + "balance_loss_clip": 0.94115263, + "balance_loss_mlp": 1.02562904, + "epoch": 0.314555401911862, + "flos": 18551740285440.0, + "grad_norm": 1.9142046199408003, + "language_loss": 0.67951369, + "learning_rate": 3.2092147756387916e-06, + "loss": 0.70103216, + "num_input_tokens_seen": 56270080, + "step": 2616, + "time_per_iteration": 3.654019832611084 + }, + { + "auxiliary_loss_clip": 0.01140652, + "auxiliary_loss_mlp": 0.0102952, + "balance_loss_clip": 0.9768129, + "balance_loss_mlp": 1.02016211, + "epoch": 0.31467564480250104, + "flos": 16362877299840.0, + "grad_norm": 1.8860281073759118, + "language_loss": 0.83346653, + "learning_rate": 3.208594214605264e-06, + "loss": 0.85516822, + "num_input_tokens_seen": 56288625, + "step": 2617, + "time_per_iteration": 2.6900715827941895 + }, + { + "auxiliary_loss_clip": 0.01139632, + "auxiliary_loss_mlp": 0.01029092, + "balance_loss_clip": 0.97894192, + "balance_loss_mlp": 1.02061582, + "epoch": 0.31479588769314015, + "flos": 21652375127040.0, + "grad_norm": 2.26751713064026, + "language_loss": 0.77509934, + "learning_rate": 3.2079734702289553e-06, + "loss": 0.79678661, + "num_input_tokens_seen": 56307520, + "step": 2618, + "time_per_iteration": 3.5570178031921387 + }, + { + "auxiliary_loss_clip": 0.01070146, + "auxiliary_loss_mlp": 0.00502454, + "balance_loss_clip": 0.98798335, + "balance_loss_mlp": 0.0, + "epoch": 0.3149161305837792, + "flos": 66051072040320.0, + "grad_norm": 0.806023918747152, + "language_loss": 0.60448951, + "learning_rate": 3.207352542604031e-06, + "loss": 0.62021554, + "num_input_tokens_seen": 56369855, + "step": 2619, + "time_per_iteration": 3.3307297229766846 + }, + { + "auxiliary_loss_clip": 0.01117809, + "auxiliary_loss_mlp": 0.0102898, + "balance_loss_clip": 0.94200647, + "balance_loss_mlp": 1.02064145, + "epoch": 0.3150363734744183, + "flos": 28987201192320.0, + "grad_norm": 1.6697683901147546, + "language_loss": 0.78363043, + "learning_rate": 3.2067314318246864e-06, + "loss": 0.80509835, + "num_input_tokens_seen": 56390570, + "step": 2620, + "time_per_iteration": 2.7728073596954346 + }, + { + "auxiliary_loss_clip": 0.01125885, + "auxiliary_loss_mlp": 0.01031869, + "balance_loss_clip": 0.9438448, + "balance_loss_mlp": 1.02335739, + "epoch": 0.31515661636505743, + "flos": 27636600879360.0, + "grad_norm": 2.176685854007313, + "language_loss": 0.77954566, + "learning_rate": 3.206110137985143e-06, + "loss": 0.80112326, + "num_input_tokens_seen": 56410775, + "step": 2621, + "time_per_iteration": 2.772374391555786 + }, + { + "auxiliary_loss_clip": 0.0111684, + "auxiliary_loss_mlp": 0.01035091, + "balance_loss_clip": 0.9423337, + "balance_loss_mlp": 1.02607298, + "epoch": 0.3152768592556965, + "flos": 24605632465920.0, + "grad_norm": 1.9936185598419671, + "language_loss": 0.92001617, + "learning_rate": 3.2054886611796505e-06, + "loss": 0.94153547, + "num_input_tokens_seen": 56429770, + "step": 2622, + "time_per_iteration": 2.726017713546753 + }, + { + "auxiliary_loss_clip": 0.01096316, + "auxiliary_loss_mlp": 0.01000776, + "balance_loss_clip": 1.0269227, + "balance_loss_mlp": 0.99904758, + "epoch": 0.3153971021463356, + "flos": 68476908026880.0, + "grad_norm": 0.9672236885330643, + "language_loss": 0.63551319, + "learning_rate": 3.204867001502487e-06, + "loss": 0.65648407, + "num_input_tokens_seen": 56488425, + "step": 2623, + "time_per_iteration": 4.123698949813843 + }, + { + "auxiliary_loss_clip": 0.01196821, + "auxiliary_loss_mlp": 0.01034714, + "balance_loss_clip": 1.0588665, + "balance_loss_mlp": 1.02621412, + "epoch": 0.3155173450369747, + "flos": 25593714766080.0, + "grad_norm": 1.8579802950875515, + "language_loss": 0.80657208, + "learning_rate": 3.2042451590479567e-06, + "loss": 0.8288874, + "num_input_tokens_seen": 56508940, + "step": 2624, + "time_per_iteration": 2.675797462463379 + }, + { + "auxiliary_loss_clip": 0.01191556, + "auxiliary_loss_mlp": 0.01028976, + "balance_loss_clip": 1.0569011, + "balance_loss_mlp": 1.0202682, + "epoch": 0.31563758792761376, + "flos": 24309333175680.0, + "grad_norm": 2.1233431132709324, + "language_loss": 0.87076008, + "learning_rate": 3.203623133910394e-06, + "loss": 0.89296544, + "num_input_tokens_seen": 56527245, + "step": 2625, + "time_per_iteration": 2.620300054550171 + }, + { + "auxiliary_loss_clip": 0.0108074, + "auxiliary_loss_mlp": 0.01027665, + "balance_loss_clip": 0.86335671, + "balance_loss_mlp": 1.01935577, + "epoch": 0.31575783081825287, + "flos": 31903865550720.0, + "grad_norm": 2.295273898859659, + "language_loss": 0.77306664, + "learning_rate": 3.203000926184158e-06, + "loss": 0.79415071, + "num_input_tokens_seen": 56546170, + "step": 2626, + "time_per_iteration": 2.8612403869628906 + }, + { + "auxiliary_loss_clip": 0.01194707, + "auxiliary_loss_mlp": 0.01026667, + "balance_loss_clip": 1.05825162, + "balance_loss_mlp": 1.01860261, + "epoch": 0.315878073708892, + "flos": 30810960385920.0, + "grad_norm": 2.9671885138335305, + "language_loss": 0.77474934, + "learning_rate": 3.202378535963639e-06, + "loss": 0.7969631, + "num_input_tokens_seen": 56567085, + "step": 2627, + "time_per_iteration": 2.7002415657043457 + }, + { + "auxiliary_loss_clip": 0.01139744, + "auxiliary_loss_mlp": 0.00511837, + "balance_loss_clip": 0.97908354, + "balance_loss_mlp": 0.0, + "epoch": 0.31599831659953104, + "flos": 22200264253440.0, + "grad_norm": 1.7286435619459855, + "language_loss": 0.83877468, + "learning_rate": 3.2017559633432516e-06, + "loss": 0.85529047, + "num_input_tokens_seen": 56586715, + "step": 2628, + "time_per_iteration": 2.704434633255005 + }, + { + "auxiliary_loss_clip": 0.01157079, + "auxiliary_loss_mlp": 0.01030312, + "balance_loss_clip": 0.98220891, + "balance_loss_mlp": 1.02178884, + "epoch": 0.31611855949017015, + "flos": 25593463370880.0, + "grad_norm": 2.0350132128799134, + "language_loss": 0.66379726, + "learning_rate": 3.2011332084174398e-06, + "loss": 0.68567115, + "num_input_tokens_seen": 56607585, + "step": 2629, + "time_per_iteration": 2.6972732543945312 + }, + { + "auxiliary_loss_clip": 0.01168504, + "auxiliary_loss_mlp": 0.01029287, + "balance_loss_clip": 1.01844335, + "balance_loss_mlp": 1.02032208, + "epoch": 0.31623880238080926, + "flos": 20594087694720.0, + "grad_norm": 2.227974106243562, + "language_loss": 0.89309913, + "learning_rate": 3.2005102712806756e-06, + "loss": 0.91507703, + "num_input_tokens_seen": 56626415, + "step": 2630, + "time_per_iteration": 2.696563959121704 + }, + { + "auxiliary_loss_clip": 0.01175376, + "auxiliary_loss_mlp": 0.0103278, + "balance_loss_clip": 1.01914942, + "balance_loss_mlp": 1.02347004, + "epoch": 0.3163590452714483, + "flos": 12784917600000.0, + "grad_norm": 2.190642985423434, + "language_loss": 0.73012352, + "learning_rate": 3.1998871520274575e-06, + "loss": 0.75220513, + "num_input_tokens_seen": 56641750, + "step": 2631, + "time_per_iteration": 2.6015260219573975 + }, + { + "auxiliary_loss_clip": 0.01148964, + "auxiliary_loss_mlp": 0.01028458, + "balance_loss_clip": 0.97828382, + "balance_loss_mlp": 1.0197264, + "epoch": 0.3164792881620874, + "flos": 23041292273280.0, + "grad_norm": 2.1840664548030286, + "language_loss": 0.84819543, + "learning_rate": 3.199263850752312e-06, + "loss": 0.86996973, + "num_input_tokens_seen": 56662585, + "step": 2632, + "time_per_iteration": 2.7307891845703125 + }, + { + "auxiliary_loss_clip": 0.01174566, + "auxiliary_loss_mlp": 0.0103398, + "balance_loss_clip": 1.01964808, + "balance_loss_mlp": 1.02546263, + "epoch": 0.31659953105272653, + "flos": 18296271780480.0, + "grad_norm": 2.3764679292577453, + "language_loss": 0.85630548, + "learning_rate": 3.198640367549795e-06, + "loss": 0.87839091, + "num_input_tokens_seen": 56681480, + "step": 2633, + "time_per_iteration": 2.6676928997039795 + }, + { + "auxiliary_loss_clip": 0.01174888, + "auxiliary_loss_mlp": 0.00511197, + "balance_loss_clip": 1.02004361, + "balance_loss_mlp": 0.0, + "epoch": 0.3167197739433656, + "flos": 25703421880320.0, + "grad_norm": 1.6731829524342183, + "language_loss": 0.85700893, + "learning_rate": 3.198016702514487e-06, + "loss": 0.87386984, + "num_input_tokens_seen": 56701760, + "step": 2634, + "time_per_iteration": 2.72318434715271 + }, + { + "auxiliary_loss_clip": 0.01193684, + "auxiliary_loss_mlp": 0.01028865, + "balance_loss_clip": 1.05789077, + "balance_loss_mlp": 1.0206635, + "epoch": 0.3168400168340047, + "flos": 23546016230400.0, + "grad_norm": 2.28021087448095, + "language_loss": 0.84314162, + "learning_rate": 3.1973928557409972e-06, + "loss": 0.86536717, + "num_input_tokens_seen": 56719800, + "step": 2635, + "time_per_iteration": 2.666900873184204 + }, + { + "auxiliary_loss_clip": 0.01194179, + "auxiliary_loss_mlp": 0.01031563, + "balance_loss_clip": 1.05850744, + "balance_loss_mlp": 1.02337372, + "epoch": 0.31696025972464376, + "flos": 28366449327360.0, + "grad_norm": 2.0468707237201187, + "language_loss": 0.71184766, + "learning_rate": 3.1967688273239636e-06, + "loss": 0.73410505, + "num_input_tokens_seen": 56739605, + "step": 2636, + "time_per_iteration": 2.644200325012207 + }, + { + "auxiliary_loss_clip": 0.01123828, + "auxiliary_loss_mlp": 0.0103136, + "balance_loss_clip": 0.94214499, + "balance_loss_mlp": 1.02273512, + "epoch": 0.31708050261528287, + "flos": 16399111144320.0, + "grad_norm": 2.3042343575841056, + "language_loss": 0.82388902, + "learning_rate": 3.1961446173580503e-06, + "loss": 0.84544086, + "num_input_tokens_seen": 56756545, + "step": 2637, + "time_per_iteration": 2.72413969039917 + }, + { + "auxiliary_loss_clip": 0.01144434, + "auxiliary_loss_mlp": 0.01026776, + "balance_loss_clip": 0.98132533, + "balance_loss_mlp": 1.01834202, + "epoch": 0.317200745505922, + "flos": 26212347728640.0, + "grad_norm": 2.1646050215329358, + "language_loss": 0.77011013, + "learning_rate": 3.1955202259379502e-06, + "loss": 0.79182225, + "num_input_tokens_seen": 56778275, + "step": 2638, + "time_per_iteration": 2.718620538711548 + }, + { + "auxiliary_loss_clip": 0.01167621, + "auxiliary_loss_mlp": 0.01032784, + "balance_loss_clip": 1.01616573, + "balance_loss_mlp": 1.02318192, + "epoch": 0.31732098839656103, + "flos": 31350876693120.0, + "grad_norm": 1.8512054647899203, + "language_loss": 0.83044851, + "learning_rate": 3.194895653158381e-06, + "loss": 0.85245252, + "num_input_tokens_seen": 56797215, + "step": 2639, + "time_per_iteration": 2.724851131439209 + }, + { + "auxiliary_loss_clip": 0.01095891, + "auxiliary_loss_mlp": 0.01012385, + "balance_loss_clip": 1.02721, + "balance_loss_mlp": 1.01058471, + "epoch": 0.31744123128720014, + "flos": 58989024835200.0, + "grad_norm": 0.7611479698462948, + "language_loss": 0.55491292, + "learning_rate": 3.194270899114093e-06, + "loss": 0.57599568, + "num_input_tokens_seen": 56863010, + "step": 2640, + "time_per_iteration": 4.174339771270752 + }, + { + "auxiliary_loss_clip": 0.01179114, + "auxiliary_loss_mlp": 0.01039776, + "balance_loss_clip": 1.02070713, + "balance_loss_mlp": 1.03031683, + "epoch": 0.31756147417783925, + "flos": 17417573372160.0, + "grad_norm": 2.612793731101667, + "language_loss": 0.82516527, + "learning_rate": 3.193645963899858e-06, + "loss": 0.84735417, + "num_input_tokens_seen": 56880625, + "step": 2641, + "time_per_iteration": 2.711838722229004 + }, + { + "auxiliary_loss_clip": 0.01141079, + "auxiliary_loss_mlp": 0.01032997, + "balance_loss_clip": 0.97976595, + "balance_loss_mlp": 1.02396727, + "epoch": 0.3176817170684783, + "flos": 25481673267840.0, + "grad_norm": 1.7922099339044486, + "language_loss": 0.84445632, + "learning_rate": 3.193020847610479e-06, + "loss": 0.86619705, + "num_input_tokens_seen": 56900945, + "step": 2642, + "time_per_iteration": 3.725400447845459 + }, + { + "auxiliary_loss_clip": 0.01144302, + "auxiliary_loss_mlp": 0.01025315, + "balance_loss_clip": 0.98281729, + "balance_loss_mlp": 1.0168035, + "epoch": 0.3178019599591174, + "flos": 24972603765120.0, + "grad_norm": 2.098576155542727, + "language_loss": 0.71039993, + "learning_rate": 3.192395550340787e-06, + "loss": 0.73209614, + "num_input_tokens_seen": 56918895, + "step": 2643, + "time_per_iteration": 2.7572195529937744 + }, + { + "auxiliary_loss_clip": 0.01172064, + "auxiliary_loss_mlp": 0.01029309, + "balance_loss_clip": 1.0204345, + "balance_loss_mlp": 1.02056503, + "epoch": 0.31792220284975653, + "flos": 12422220019200.0, + "grad_norm": 2.0599734924367445, + "language_loss": 0.7632488, + "learning_rate": 3.191770072185638e-06, + "loss": 0.78526258, + "num_input_tokens_seen": 56935890, + "step": 2644, + "time_per_iteration": 3.608635663986206 + }, + { + "auxiliary_loss_clip": 0.01167784, + "auxiliary_loss_mlp": 0.01032768, + "balance_loss_clip": 1.01778793, + "balance_loss_mlp": 1.02381587, + "epoch": 0.3180424457403956, + "flos": 15485759089920.0, + "grad_norm": 2.371689882523918, + "language_loss": 0.72179729, + "learning_rate": 3.191144413239916e-06, + "loss": 0.74380279, + "num_input_tokens_seen": 56952460, + "step": 2645, + "time_per_iteration": 2.6416537761688232 + }, + { + "auxiliary_loss_clip": 0.01150335, + "auxiliary_loss_mlp": 0.01031746, + "balance_loss_clip": 0.98239565, + "balance_loss_mlp": 1.02349091, + "epoch": 0.3181626886310347, + "flos": 26174964648960.0, + "grad_norm": 2.3512748464104103, + "language_loss": 0.8837024, + "learning_rate": 3.190518573598534e-06, + "loss": 0.90552324, + "num_input_tokens_seen": 56969065, + "step": 2646, + "time_per_iteration": 2.7488391399383545 + }, + { + "auxiliary_loss_clip": 0.01128261, + "auxiliary_loss_mlp": 0.01029859, + "balance_loss_clip": 0.94007427, + "balance_loss_mlp": 1.02130532, + "epoch": 0.3182829315216738, + "flos": 25483109811840.0, + "grad_norm": 1.5164382423923213, + "language_loss": 0.77291179, + "learning_rate": 3.1898925533564308e-06, + "loss": 0.79449296, + "num_input_tokens_seen": 56990535, + "step": 2647, + "time_per_iteration": 2.8405709266662598 + }, + { + "auxiliary_loss_clip": 0.01113536, + "auxiliary_loss_mlp": 0.01033622, + "balance_loss_clip": 0.94030607, + "balance_loss_mlp": 1.02523541, + "epoch": 0.31840317441231286, + "flos": 18113701927680.0, + "grad_norm": 2.224896083998906, + "language_loss": 0.6395697, + "learning_rate": 3.1892663526085733e-06, + "loss": 0.66104126, + "num_input_tokens_seen": 57008910, + "step": 2648, + "time_per_iteration": 2.70914363861084 + }, + { + "auxiliary_loss_clip": 0.01097553, + "auxiliary_loss_mlp": 0.01002696, + "balance_loss_clip": 1.0282042, + "balance_loss_mlp": 1.00096726, + "epoch": 0.31852341730295197, + "flos": 64741948957440.0, + "grad_norm": 0.7557106451141433, + "language_loss": 0.56951582, + "learning_rate": 3.188639971449956e-06, + "loss": 0.5905183, + "num_input_tokens_seen": 57074960, + "step": 2649, + "time_per_iteration": 4.125219821929932 + }, + { + "auxiliary_loss_clip": 0.01199211, + "auxiliary_loss_mlp": 0.01037478, + "balance_loss_clip": 1.06017685, + "balance_loss_mlp": 1.02832842, + "epoch": 0.318643660193591, + "flos": 20668135582080.0, + "grad_norm": 2.769925760877708, + "language_loss": 0.72070616, + "learning_rate": 3.1880134099756e-06, + "loss": 0.74307311, + "num_input_tokens_seen": 57094595, + "step": 2650, + "time_per_iteration": 2.59920334815979 + }, + { + "auxiliary_loss_clip": 0.01167997, + "auxiliary_loss_mlp": 0.01034543, + "balance_loss_clip": 1.01682878, + "balance_loss_mlp": 1.02619839, + "epoch": 0.31876390308423014, + "flos": 26943345411840.0, + "grad_norm": 2.4770295919791403, + "language_loss": 0.69457304, + "learning_rate": 3.1873866682805535e-06, + "loss": 0.71659845, + "num_input_tokens_seen": 57115290, + "step": 2651, + "time_per_iteration": 2.751605272293091 + }, + { + "auxiliary_loss_clip": 0.01151545, + "auxiliary_loss_mlp": 0.01032866, + "balance_loss_clip": 0.9820838, + "balance_loss_mlp": 1.02436614, + "epoch": 0.31888414597486925, + "flos": 18041916597120.0, + "grad_norm": 1.8280381947603441, + "language_loss": 0.88351464, + "learning_rate": 3.186759746459894e-06, + "loss": 0.90535867, + "num_input_tokens_seen": 57134400, + "step": 2652, + "time_per_iteration": 2.6984312534332275 + }, + { + "auxiliary_loss_clip": 0.01144855, + "auxiliary_loss_mlp": 0.01027036, + "balance_loss_clip": 0.97941881, + "balance_loss_mlp": 1.01830947, + "epoch": 0.3190043888655083, + "flos": 25149319701120.0, + "grad_norm": 1.886143785585006, + "language_loss": 0.79421598, + "learning_rate": 3.1861326446087246e-06, + "loss": 0.8159349, + "num_input_tokens_seen": 57153140, + "step": 2653, + "time_per_iteration": 2.7193069458007812 + }, + { + "auxiliary_loss_clip": 0.0117508, + "auxiliary_loss_mlp": 0.01028204, + "balance_loss_clip": 1.02025723, + "balance_loss_mlp": 1.01937628, + "epoch": 0.3191246317561474, + "flos": 22053892331520.0, + "grad_norm": 2.031594882834147, + "language_loss": 0.71641761, + "learning_rate": 3.1855053628221763e-06, + "loss": 0.73845041, + "num_input_tokens_seen": 57172395, + "step": 2654, + "time_per_iteration": 2.6751487255096436 + }, + { + "auxiliary_loss_clip": 0.01115034, + "auxiliary_loss_mlp": 0.01030387, + "balance_loss_clip": 0.93681794, + "balance_loss_mlp": 1.0210824, + "epoch": 0.3192448746467865, + "flos": 14901815687040.0, + "grad_norm": 2.4810065868997633, + "language_loss": 0.89533734, + "learning_rate": 3.184877901195407e-06, + "loss": 0.91679156, + "num_input_tokens_seen": 57189090, + "step": 2655, + "time_per_iteration": 2.655113697052002 + }, + { + "auxiliary_loss_clip": 0.01052989, + "auxiliary_loss_mlp": 0.01004068, + "balance_loss_clip": 0.95456183, + "balance_loss_mlp": 1.00244665, + "epoch": 0.3193651175374256, + "flos": 67234832657280.0, + "grad_norm": 0.7938683739904822, + "language_loss": 0.62844342, + "learning_rate": 3.184250259823602e-06, + "loss": 0.649014, + "num_input_tokens_seen": 57251620, + "step": 2656, + "time_per_iteration": 3.342573642730713 + }, + { + "auxiliary_loss_clip": 0.01122291, + "auxiliary_loss_mlp": 0.01031838, + "balance_loss_clip": 0.93973958, + "balance_loss_mlp": 1.02288532, + "epoch": 0.3194853604280647, + "flos": 12233077977600.0, + "grad_norm": 3.2179672484991384, + "language_loss": 0.81774211, + "learning_rate": 3.183622438801974e-06, + "loss": 0.83928335, + "num_input_tokens_seen": 57266910, + "step": 2657, + "time_per_iteration": 2.6918468475341797 + }, + { + "auxiliary_loss_clip": 0.0119725, + "auxiliary_loss_mlp": 0.01034537, + "balance_loss_clip": 1.06065047, + "balance_loss_mlp": 1.02578712, + "epoch": 0.3196056033187038, + "flos": 14939917038720.0, + "grad_norm": 1.9325448747337897, + "language_loss": 0.75116569, + "learning_rate": 3.1829944382257637e-06, + "loss": 0.77348351, + "num_input_tokens_seen": 57285040, + "step": 2658, + "time_per_iteration": 2.6595680713653564 + }, + { + "auxiliary_loss_clip": 0.01169974, + "auxiliary_loss_mlp": 0.01033963, + "balance_loss_clip": 1.0213542, + "balance_loss_mlp": 1.02539229, + "epoch": 0.31972584620934286, + "flos": 23768878164480.0, + "grad_norm": 2.114692796261693, + "language_loss": 0.81450212, + "learning_rate": 3.1823662581902373e-06, + "loss": 0.83654141, + "num_input_tokens_seen": 57302725, + "step": 2659, + "time_per_iteration": 2.6962926387786865 + }, + { + "auxiliary_loss_clip": 0.01108895, + "auxiliary_loss_mlp": 0.01030872, + "balance_loss_clip": 0.93457615, + "balance_loss_mlp": 1.02255726, + "epoch": 0.31984608909998197, + "flos": 21251540280960.0, + "grad_norm": 3.0902403993807965, + "language_loss": 0.74481082, + "learning_rate": 3.1817378987906896e-06, + "loss": 0.76620853, + "num_input_tokens_seen": 57322230, + "step": 2660, + "time_per_iteration": 2.7303993701934814 + }, + { + "auxiliary_loss_clip": 0.01096819, + "auxiliary_loss_mlp": 0.01032401, + "balance_loss_clip": 0.90618169, + "balance_loss_mlp": 1.02315688, + "epoch": 0.3199663319906211, + "flos": 18296235866880.0, + "grad_norm": 2.128647815784181, + "language_loss": 0.79409659, + "learning_rate": 3.181109360122442e-06, + "loss": 0.8153888, + "num_input_tokens_seen": 57339820, + "step": 2661, + "time_per_iteration": 2.777775764465332 + }, + { + "auxiliary_loss_clip": 0.01120223, + "auxiliary_loss_mlp": 0.01028865, + "balance_loss_clip": 0.94106483, + "balance_loss_mlp": 1.01996589, + "epoch": 0.32008657488126013, + "flos": 18733627779840.0, + "grad_norm": 2.3484043829622547, + "language_loss": 0.78018963, + "learning_rate": 3.1804806422808445e-06, + "loss": 0.80168051, + "num_input_tokens_seen": 57356955, + "step": 2662, + "time_per_iteration": 2.6796793937683105 + }, + { + "auxiliary_loss_clip": 0.01137897, + "auxiliary_loss_mlp": 0.01037273, + "balance_loss_clip": 0.97876567, + "balance_loss_mlp": 1.02886915, + "epoch": 0.32020681777189924, + "flos": 20595344670720.0, + "grad_norm": 2.22760603878076, + "language_loss": 0.7310406, + "learning_rate": 3.1798517453612714e-06, + "loss": 0.75279224, + "num_input_tokens_seen": 57376760, + "step": 2663, + "time_per_iteration": 2.81699800491333 + }, + { + "auxiliary_loss_clip": 0.01167978, + "auxiliary_loss_mlp": 0.01035639, + "balance_loss_clip": 1.02116489, + "balance_loss_mlp": 1.0267942, + "epoch": 0.32032706066253835, + "flos": 35261692750080.0, + "grad_norm": 2.3263091326910836, + "language_loss": 0.75458634, + "learning_rate": 3.1792226694591265e-06, + "loss": 0.77662253, + "num_input_tokens_seen": 57398145, + "step": 2664, + "time_per_iteration": 2.7383060455322266 + }, + { + "auxiliary_loss_clip": 0.01123834, + "auxiliary_loss_mlp": 0.01032528, + "balance_loss_clip": 0.9445706, + "balance_loss_mlp": 1.02377844, + "epoch": 0.3204473035531774, + "flos": 15304230731520.0, + "grad_norm": 3.588635459583051, + "language_loss": 0.80477536, + "learning_rate": 3.178593414669841e-06, + "loss": 0.82633895, + "num_input_tokens_seen": 57416730, + "step": 2665, + "time_per_iteration": 2.7563672065734863 + }, + { + "auxiliary_loss_clip": 0.01173794, + "auxiliary_loss_mlp": 0.01029521, + "balance_loss_clip": 1.02101398, + "balance_loss_mlp": 1.02030015, + "epoch": 0.3205675464438165, + "flos": 24462564595200.0, + "grad_norm": 3.4765496426367783, + "language_loss": 0.70824277, + "learning_rate": 3.1779639810888707e-06, + "loss": 0.73027599, + "num_input_tokens_seen": 57436325, + "step": 2666, + "time_per_iteration": 3.6094439029693604 + }, + { + "auxiliary_loss_clip": 0.01168447, + "auxiliary_loss_mlp": 0.01034351, + "balance_loss_clip": 1.01948905, + "balance_loss_mlp": 1.02527285, + "epoch": 0.3206877893344556, + "flos": 22456235548800.0, + "grad_norm": 2.079854340974218, + "language_loss": 0.76135862, + "learning_rate": 3.1773343688117013e-06, + "loss": 0.78338659, + "num_input_tokens_seen": 57457235, + "step": 2667, + "time_per_iteration": 2.679410934448242 + }, + { + "auxiliary_loss_clip": 0.01149848, + "auxiliary_loss_mlp": 0.00510962, + "balance_loss_clip": 0.98076129, + "balance_loss_mlp": 0.0, + "epoch": 0.3208080322250947, + "flos": 20412236113920.0, + "grad_norm": 2.0337889929264894, + "language_loss": 0.8368454, + "learning_rate": 3.1767045779338445e-06, + "loss": 0.85345352, + "num_input_tokens_seen": 57474895, + "step": 2668, + "time_per_iteration": 2.664478302001953 + }, + { + "auxiliary_loss_clip": 0.01167045, + "auxiliary_loss_mlp": 0.01029656, + "balance_loss_clip": 1.01536727, + "balance_loss_mlp": 1.02132332, + "epoch": 0.3209282751157338, + "flos": 21762118154880.0, + "grad_norm": 2.49806661677856, + "language_loss": 0.91473329, + "learning_rate": 3.176074608550839e-06, + "loss": 0.93670022, + "num_input_tokens_seen": 57490715, + "step": 2669, + "time_per_iteration": 3.577115058898926 + }, + { + "auxiliary_loss_clip": 0.01074148, + "auxiliary_loss_mlp": 0.01028399, + "balance_loss_clip": 0.86403096, + "balance_loss_mlp": 1.02025139, + "epoch": 0.32104851800637285, + "flos": 22055041566720.0, + "grad_norm": 2.655843753695653, + "language_loss": 0.82392162, + "learning_rate": 3.17544446075825e-06, + "loss": 0.8449471, + "num_input_tokens_seen": 57509880, + "step": 2670, + "time_per_iteration": 3.7080581188201904 + }, + { + "auxiliary_loss_clip": 0.0114975, + "auxiliary_loss_mlp": 0.01028612, + "balance_loss_clip": 0.97787297, + "balance_loss_mlp": 1.02043974, + "epoch": 0.32116876089701196, + "flos": 37012301896320.0, + "grad_norm": 1.662928118252429, + "language_loss": 0.71133602, + "learning_rate": 3.174814134651671e-06, + "loss": 0.73311961, + "num_input_tokens_seen": 57532430, + "step": 2671, + "time_per_iteration": 2.842162609100342 + }, + { + "auxiliary_loss_clip": 0.01192554, + "auxiliary_loss_mlp": 0.01029324, + "balance_loss_clip": 1.05889571, + "balance_loss_mlp": 1.02085447, + "epoch": 0.3212890037876511, + "flos": 21979233912960.0, + "grad_norm": 1.9285956915304312, + "language_loss": 0.8042928, + "learning_rate": 3.1741836303267215e-06, + "loss": 0.82651162, + "num_input_tokens_seen": 57551965, + "step": 2672, + "time_per_iteration": 2.593946695327759 + }, + { + "auxiliary_loss_clip": 0.01193527, + "auxiliary_loss_mlp": 0.01026247, + "balance_loss_clip": 1.0585103, + "balance_loss_mlp": 1.01801252, + "epoch": 0.32140924667829013, + "flos": 10342345875840.0, + "grad_norm": 2.9742866015768983, + "language_loss": 0.74594867, + "learning_rate": 3.1735529478790496e-06, + "loss": 0.7681464, + "num_input_tokens_seen": 57569955, + "step": 2673, + "time_per_iteration": 2.633857250213623 + }, + { + "auxiliary_loss_clip": 0.01172324, + "auxiliary_loss_mlp": 0.0103, + "balance_loss_clip": 1.02060044, + "balance_loss_mlp": 1.02168536, + "epoch": 0.32152948956892924, + "flos": 50798910072960.0, + "grad_norm": 54.00145319023206, + "language_loss": 0.79122508, + "learning_rate": 3.172922087404328e-06, + "loss": 0.8132484, + "num_input_tokens_seen": 57592215, + "step": 2674, + "time_per_iteration": 2.8915843963623047 + }, + { + "auxiliary_loss_clip": 0.01102896, + "auxiliary_loss_mlp": 0.01003276, + "balance_loss_clip": 1.03405404, + "balance_loss_mlp": 1.0015471, + "epoch": 0.32164973245956835, + "flos": 63863250549120.0, + "grad_norm": 0.7848419589780609, + "language_loss": 0.55258071, + "learning_rate": 3.1722910489982586e-06, + "loss": 0.57364237, + "num_input_tokens_seen": 57652575, + "step": 2675, + "time_per_iteration": 4.214853048324585 + }, + { + "auxiliary_loss_clip": 0.01144176, + "auxiliary_loss_mlp": 0.01028936, + "balance_loss_clip": 0.97977746, + "balance_loss_mlp": 1.02076411, + "epoch": 0.3217699753502074, + "flos": 23513948363520.0, + "grad_norm": 1.824176700400858, + "language_loss": 0.7997365, + "learning_rate": 3.1716598327565694e-06, + "loss": 0.82146764, + "num_input_tokens_seen": 57672215, + "step": 2676, + "time_per_iteration": 2.8238468170166016 + }, + { + "auxiliary_loss_clip": 0.01194599, + "auxiliary_loss_mlp": 0.01029616, + "balance_loss_clip": 1.05851984, + "balance_loss_mlp": 1.02124763, + "epoch": 0.3218902182408465, + "flos": 19062533640960.0, + "grad_norm": 1.6219853046391957, + "language_loss": 0.84163809, + "learning_rate": 3.171028438775015e-06, + "loss": 0.86388016, + "num_input_tokens_seen": 57691410, + "step": 2677, + "time_per_iteration": 2.5957438945770264 + }, + { + "auxiliary_loss_clip": 0.01190734, + "auxiliary_loss_mlp": 0.0102651, + "balance_loss_clip": 1.05605352, + "balance_loss_mlp": 1.01862395, + "epoch": 0.3220104611314856, + "flos": 20375571306240.0, + "grad_norm": 1.873488879476241, + "language_loss": 0.84136659, + "learning_rate": 3.170396867149377e-06, + "loss": 0.86353904, + "num_input_tokens_seen": 57709415, + "step": 2678, + "time_per_iteration": 2.6143381595611572 + }, + { + "auxiliary_loss_clip": 0.01091388, + "auxiliary_loss_mlp": 0.01026492, + "balance_loss_clip": 0.90311313, + "balance_loss_mlp": 1.01812959, + "epoch": 0.3221307040221247, + "flos": 20117014231680.0, + "grad_norm": 1.8092947815991578, + "language_loss": 0.86342335, + "learning_rate": 3.1697651179754653e-06, + "loss": 0.88460219, + "num_input_tokens_seen": 57728075, + "step": 2679, + "time_per_iteration": 2.747682571411133 + }, + { + "auxiliary_loss_clip": 0.01125469, + "auxiliary_loss_mlp": 0.01029748, + "balance_loss_clip": 0.94601518, + "balance_loss_mlp": 1.02194583, + "epoch": 0.3222509469127638, + "flos": 23987789602560.0, + "grad_norm": 2.4728558132358986, + "language_loss": 0.73083454, + "learning_rate": 3.1691331913491153e-06, + "loss": 0.75238669, + "num_input_tokens_seen": 57750645, + "step": 2680, + "time_per_iteration": 2.7683024406433105 + }, + { + "auxiliary_loss_clip": 0.01192335, + "auxiliary_loss_mlp": 0.01026263, + "balance_loss_clip": 1.05544579, + "balance_loss_mlp": 1.01805508, + "epoch": 0.32237118980340285, + "flos": 17675735397120.0, + "grad_norm": 3.059772301960278, + "language_loss": 0.84633791, + "learning_rate": 3.1685010873661898e-06, + "loss": 0.86852384, + "num_input_tokens_seen": 57769820, + "step": 2681, + "time_per_iteration": 2.613640308380127 + }, + { + "auxiliary_loss_clip": 0.01166798, + "auxiliary_loss_mlp": 0.01033371, + "balance_loss_clip": 1.01902819, + "balance_loss_mlp": 1.02398324, + "epoch": 0.32249143269404196, + "flos": 23147982645120.0, + "grad_norm": 2.04809029788193, + "language_loss": 0.79872048, + "learning_rate": 3.167868806122578e-06, + "loss": 0.82072222, + "num_input_tokens_seen": 57788870, + "step": 2682, + "time_per_iteration": 2.6857290267944336 + }, + { + "auxiliary_loss_clip": 0.01150626, + "auxiliary_loss_mlp": 0.01033703, + "balance_loss_clip": 0.98259306, + "balance_loss_mlp": 1.02492356, + "epoch": 0.32261167558468107, + "flos": 24422308427520.0, + "grad_norm": 1.7932932776590882, + "language_loss": 0.66582245, + "learning_rate": 3.1672363477141968e-06, + "loss": 0.6876657, + "num_input_tokens_seen": 57808165, + "step": 2683, + "time_per_iteration": 2.753735065460205 + }, + { + "auxiliary_loss_clip": 0.01148524, + "auxiliary_loss_mlp": 0.01031305, + "balance_loss_clip": 0.9786365, + "balance_loss_mlp": 1.0220958, + "epoch": 0.3227319184753201, + "flos": 30367175852160.0, + "grad_norm": 2.5260799063018253, + "language_loss": 0.84838933, + "learning_rate": 3.1666037122369903e-06, + "loss": 0.87018758, + "num_input_tokens_seen": 57828825, + "step": 2684, + "time_per_iteration": 2.763369083404541 + }, + { + "auxiliary_loss_clip": 0.01163167, + "auxiliary_loss_mlp": 0.01028605, + "balance_loss_clip": 1.01370299, + "balance_loss_mlp": 1.0198071, + "epoch": 0.32285216136595923, + "flos": 16946174257920.0, + "grad_norm": 3.009030720298068, + "language_loss": 0.86855578, + "learning_rate": 3.165970899786928e-06, + "loss": 0.89047348, + "num_input_tokens_seen": 57846740, + "step": 2685, + "time_per_iteration": 2.6298840045928955 + }, + { + "auxiliary_loss_clip": 0.01124876, + "auxiliary_loss_mlp": 0.01029658, + "balance_loss_clip": 0.94127536, + "balance_loss_mlp": 1.02086639, + "epoch": 0.32297240425659834, + "flos": 21981532383360.0, + "grad_norm": 1.6972944274726416, + "language_loss": 0.75065321, + "learning_rate": 3.1653379104600067e-06, + "loss": 0.77219856, + "num_input_tokens_seen": 57866885, + "step": 2686, + "time_per_iteration": 2.735243082046509 + }, + { + "auxiliary_loss_clip": 0.0116986, + "auxiliary_loss_mlp": 0.01030328, + "balance_loss_clip": 1.01851618, + "balance_loss_mlp": 1.02183998, + "epoch": 0.3230926471472374, + "flos": 22748045639040.0, + "grad_norm": 1.5846124622889166, + "language_loss": 0.69665289, + "learning_rate": 3.164704744352251e-06, + "loss": 0.71865475, + "num_input_tokens_seen": 57887690, + "step": 2687, + "time_per_iteration": 2.6325433254241943 + }, + { + "auxiliary_loss_clip": 0.01167205, + "auxiliary_loss_mlp": 0.01025956, + "balance_loss_clip": 1.01852262, + "balance_loss_mlp": 1.01808178, + "epoch": 0.3232128900378765, + "flos": 16942977947520.0, + "grad_norm": 2.2345678248056204, + "language_loss": 0.80547369, + "learning_rate": 3.164071401559713e-06, + "loss": 0.82740527, + "num_input_tokens_seen": 57905090, + "step": 2688, + "time_per_iteration": 2.634166717529297 + }, + { + "auxiliary_loss_clip": 0.01149348, + "auxiliary_loss_mlp": 0.01027747, + "balance_loss_clip": 0.98007596, + "balance_loss_mlp": 1.01967955, + "epoch": 0.3233331329285156, + "flos": 24023736138240.0, + "grad_norm": 1.7819651138604795, + "language_loss": 0.70962375, + "learning_rate": 3.1634378821784674e-06, + "loss": 0.73139471, + "num_input_tokens_seen": 57925305, + "step": 2689, + "time_per_iteration": 2.6996264457702637 + }, + { + "auxiliary_loss_clip": 0.01121815, + "auxiliary_loss_mlp": 0.01026773, + "balance_loss_clip": 0.9412806, + "balance_loss_mlp": 1.0181545, + "epoch": 0.3234533758191547, + "flos": 18113845582080.0, + "grad_norm": 7.215478599510512, + "language_loss": 0.73587632, + "learning_rate": 3.1628041863046208e-06, + "loss": 0.75736225, + "num_input_tokens_seen": 57942720, + "step": 2690, + "time_per_iteration": 2.682908535003662 + }, + { + "auxiliary_loss_clip": 0.01194366, + "auxiliary_loss_mlp": 0.01025669, + "balance_loss_clip": 1.05594993, + "balance_loss_mlp": 1.01734841, + "epoch": 0.3235736187097938, + "flos": 16946138344320.0, + "grad_norm": 2.1975054578674436, + "language_loss": 0.90943164, + "learning_rate": 3.162170314034304e-06, + "loss": 0.93163198, + "num_input_tokens_seen": 57960135, + "step": 2691, + "time_per_iteration": 2.6322367191314697 + }, + { + "auxiliary_loss_clip": 0.01193972, + "auxiliary_loss_mlp": 0.01034056, + "balance_loss_clip": 1.05603421, + "balance_loss_mlp": 1.02507353, + "epoch": 0.3236938616004329, + "flos": 22127150119680.0, + "grad_norm": 1.9038147285160274, + "language_loss": 0.81265163, + "learning_rate": 3.1615362654636738e-06, + "loss": 0.83493191, + "num_input_tokens_seen": 57980875, + "step": 2692, + "time_per_iteration": 2.669978141784668 + }, + { + "auxiliary_loss_clip": 0.01118133, + "auxiliary_loss_mlp": 0.01026138, + "balance_loss_clip": 0.94484079, + "balance_loss_mlp": 1.01853871, + "epoch": 0.32381410449107195, + "flos": 17164618819200.0, + "grad_norm": 1.7555241894261877, + "language_loss": 0.87338662, + "learning_rate": 3.1609020406889163e-06, + "loss": 0.89482939, + "num_input_tokens_seen": 57998310, + "step": 2693, + "time_per_iteration": 3.565922975540161 + }, + { + "auxiliary_loss_clip": 0.01148356, + "auxiliary_loss_mlp": 0.01034977, + "balance_loss_clip": 0.97983611, + "balance_loss_mlp": 1.02662063, + "epoch": 0.32393434738171106, + "flos": 16578125550720.0, + "grad_norm": 2.4710577640418667, + "language_loss": 0.85235322, + "learning_rate": 3.1602676398062416e-06, + "loss": 0.87418652, + "num_input_tokens_seen": 58017220, + "step": 2694, + "time_per_iteration": 2.653012752532959 + }, + { + "auxiliary_loss_clip": 0.01169552, + "auxiliary_loss_mlp": 0.0103111, + "balance_loss_clip": 1.02115417, + "balance_loss_mlp": 1.02259851, + "epoch": 0.3240545902723502, + "flos": 25483612602240.0, + "grad_norm": 2.081042969086207, + "language_loss": 0.61524069, + "learning_rate": 3.1596330629118886e-06, + "loss": 0.63724732, + "num_input_tokens_seen": 58037190, + "step": 2695, + "time_per_iteration": 3.6146581172943115 + }, + { + "auxiliary_loss_clip": 0.01089857, + "auxiliary_loss_mlp": 0.01025143, + "balance_loss_clip": 0.90150392, + "balance_loss_mlp": 1.01710844, + "epoch": 0.32417483316298923, + "flos": 35845851634560.0, + "grad_norm": 1.9555043374994763, + "language_loss": 0.73490262, + "learning_rate": 3.1589983101021223e-06, + "loss": 0.75605261, + "num_input_tokens_seen": 58055820, + "step": 2696, + "time_per_iteration": 3.72983455657959 + }, + { + "auxiliary_loss_clip": 0.01146891, + "auxiliary_loss_mlp": 0.01028774, + "balance_loss_clip": 0.98041761, + "balance_loss_mlp": 1.02036381, + "epoch": 0.32429507605362834, + "flos": 30080501406720.0, + "grad_norm": 2.2381238719953407, + "language_loss": 0.8493796, + "learning_rate": 3.1583633814732337e-06, + "loss": 0.87113619, + "num_input_tokens_seen": 58075340, + "step": 2697, + "time_per_iteration": 2.7057290077209473 + }, + { + "auxiliary_loss_clip": 0.01193306, + "auxiliary_loss_mlp": 0.01028112, + "balance_loss_clip": 1.05658174, + "balance_loss_mlp": 1.01910591, + "epoch": 0.3244153189442674, + "flos": 18223265387520.0, + "grad_norm": 2.5407606132349905, + "language_loss": 0.71829224, + "learning_rate": 3.157728277121541e-06, + "loss": 0.74050641, + "num_input_tokens_seen": 58093515, + "step": 2698, + "time_per_iteration": 2.6056418418884277 + }, + { + "auxiliary_loss_clip": 0.01194094, + "auxiliary_loss_mlp": 0.0102927, + "balance_loss_clip": 1.05561244, + "balance_loss_mlp": 1.02077615, + "epoch": 0.3245355618349065, + "flos": 17710317216000.0, + "grad_norm": 3.2531097350961025, + "language_loss": 0.78294134, + "learning_rate": 3.1570929971433897e-06, + "loss": 0.80517495, + "num_input_tokens_seen": 58109300, + "step": 2699, + "time_per_iteration": 2.560279130935669 + }, + { + "auxiliary_loss_clip": 0.01171841, + "auxiliary_loss_mlp": 0.01031437, + "balance_loss_clip": 1.02218366, + "balance_loss_mlp": 1.0235157, + "epoch": 0.3246558047255456, + "flos": 23440798316160.0, + "grad_norm": 2.268225069117236, + "language_loss": 0.83693016, + "learning_rate": 3.1564575416351504e-06, + "loss": 0.85896295, + "num_input_tokens_seen": 58128000, + "step": 2700, + "time_per_iteration": 2.6278278827667236 + }, + { + "auxiliary_loss_clip": 0.01194124, + "auxiliary_loss_mlp": 0.01034032, + "balance_loss_clip": 1.0578804, + "balance_loss_mlp": 1.02548456, + "epoch": 0.32477604761618467, + "flos": 21760861178880.0, + "grad_norm": 1.9243667443514576, + "language_loss": 0.7420783, + "learning_rate": 3.155821910693221e-06, + "loss": 0.76435983, + "num_input_tokens_seen": 58147415, + "step": 2701, + "time_per_iteration": 3.5360987186431885 + }, + { + "auxiliary_loss_clip": 0.01143477, + "auxiliary_loss_mlp": 0.01032566, + "balance_loss_clip": 0.97795719, + "balance_loss_mlp": 1.02447724, + "epoch": 0.3248962905068238, + "flos": 19828328624640.0, + "grad_norm": 1.8386498990339057, + "language_loss": 0.85896522, + "learning_rate": 3.1551861044140275e-06, + "loss": 0.88072562, + "num_input_tokens_seen": 58167050, + "step": 2702, + "time_per_iteration": 2.6584911346435547 + }, + { + "auxiliary_loss_clip": 0.01092154, + "auxiliary_loss_mlp": 0.01027918, + "balance_loss_clip": 0.90320665, + "balance_loss_mlp": 1.01949632, + "epoch": 0.3250165333974629, + "flos": 23948215793280.0, + "grad_norm": 1.6626841612939118, + "language_loss": 0.77554047, + "learning_rate": 3.15455012289402e-06, + "loss": 0.79674119, + "num_input_tokens_seen": 58186695, + "step": 2703, + "time_per_iteration": 2.7969577312469482 + }, + { + "auxiliary_loss_clip": 0.01172714, + "auxiliary_loss_mlp": 0.01030386, + "balance_loss_clip": 1.02313709, + "balance_loss_mlp": 1.0215466, + "epoch": 0.32513677628810195, + "flos": 23989333887360.0, + "grad_norm": 5.577302550441904, + "language_loss": 0.84169364, + "learning_rate": 3.153913966229677e-06, + "loss": 0.86372471, + "num_input_tokens_seen": 58205815, + "step": 2704, + "time_per_iteration": 2.677494525909424 + }, + { + "auxiliary_loss_clip": 0.01079948, + "auxiliary_loss_mlp": 0.01015498, + "balance_loss_clip": 0.99434221, + "balance_loss_mlp": 1.01382959, + "epoch": 0.32525701917874106, + "flos": 70655790009600.0, + "grad_norm": 0.6770443661615152, + "language_loss": 0.50250018, + "learning_rate": 3.1532776345175027e-06, + "loss": 0.52345467, + "num_input_tokens_seen": 58270960, + "step": 2705, + "time_per_iteration": 3.2016682624816895 + }, + { + "auxiliary_loss_clip": 0.01192762, + "auxiliary_loss_mlp": 0.01029083, + "balance_loss_clip": 1.05863845, + "balance_loss_mlp": 1.02092314, + "epoch": 0.32537726206938017, + "flos": 19682639061120.0, + "grad_norm": 1.8053013400922304, + "language_loss": 0.78478521, + "learning_rate": 3.1526411278540285e-06, + "loss": 0.80700374, + "num_input_tokens_seen": 58289390, + "step": 2706, + "time_per_iteration": 2.6859467029571533 + }, + { + "auxiliary_loss_clip": 0.01151492, + "auxiliary_loss_mlp": 0.01025232, + "balance_loss_clip": 0.97820449, + "balance_loss_mlp": 1.01702487, + "epoch": 0.3254975049600192, + "flos": 28760999293440.0, + "grad_norm": 2.621602548781594, + "language_loss": 0.81448698, + "learning_rate": 3.1520044463358116e-06, + "loss": 0.83625424, + "num_input_tokens_seen": 58306120, + "step": 2707, + "time_per_iteration": 2.7826032638549805 + }, + { + "auxiliary_loss_clip": 0.01169456, + "auxiliary_loss_mlp": 0.0102489, + "balance_loss_clip": 1.01952481, + "balance_loss_mlp": 1.01611567, + "epoch": 0.32561774785065833, + "flos": 18877378008960.0, + "grad_norm": 1.5905571340283615, + "language_loss": 0.8052541, + "learning_rate": 3.151367590059436e-06, + "loss": 0.82719749, + "num_input_tokens_seen": 58324545, + "step": 2708, + "time_per_iteration": 2.6566202640533447 + }, + { + "auxiliary_loss_clip": 0.01195426, + "auxiliary_loss_mlp": 0.00510761, + "balance_loss_clip": 1.05889416, + "balance_loss_mlp": 0.0, + "epoch": 0.32573799074129745, + "flos": 23112107936640.0, + "grad_norm": 2.2187519476858744, + "language_loss": 0.8640272, + "learning_rate": 3.1507305591215117e-06, + "loss": 0.88108903, + "num_input_tokens_seen": 58342455, + "step": 2709, + "time_per_iteration": 2.642444610595703 + }, + { + "auxiliary_loss_clip": 0.01078321, + "auxiliary_loss_mlp": 0.01001426, + "balance_loss_clip": 0.99337292, + "balance_loss_mlp": 0.99973315, + "epoch": 0.3258582336319365, + "flos": 71237650423680.0, + "grad_norm": 0.6654370567351724, + "language_loss": 0.55718154, + "learning_rate": 3.150093353618677e-06, + "loss": 0.57797903, + "num_input_tokens_seen": 58407185, + "step": 2710, + "time_per_iteration": 3.2584571838378906 + }, + { + "auxiliary_loss_clip": 0.01175036, + "auxiliary_loss_mlp": 0.01031637, + "balance_loss_clip": 1.01785803, + "balance_loss_mlp": 1.02264321, + "epoch": 0.3259784765225756, + "flos": 22456020067200.0, + "grad_norm": 2.273794688095461, + "language_loss": 0.88426888, + "learning_rate": 3.149455973647596e-06, + "loss": 0.90633559, + "num_input_tokens_seen": 58425245, + "step": 2711, + "time_per_iteration": 2.624793767929077 + }, + { + "auxiliary_loss_clip": 0.01112942, + "auxiliary_loss_mlp": 0.010309, + "balance_loss_clip": 0.93509305, + "balance_loss_mlp": 1.02191186, + "epoch": 0.32609871941321467, + "flos": 20484811543680.0, + "grad_norm": 2.1018216796866835, + "language_loss": 0.76921725, + "learning_rate": 3.1488184193049563e-06, + "loss": 0.79065561, + "num_input_tokens_seen": 58444780, + "step": 2712, + "time_per_iteration": 2.738502025604248 + }, + { + "auxiliary_loss_clip": 0.01193854, + "auxiliary_loss_mlp": 0.01026135, + "balance_loss_clip": 1.05868077, + "balance_loss_mlp": 1.01751578, + "epoch": 0.3262189623038538, + "flos": 22416805393920.0, + "grad_norm": 4.652473871565208, + "language_loss": 0.72231251, + "learning_rate": 3.1481806906874767e-06, + "loss": 0.74451238, + "num_input_tokens_seen": 58466090, + "step": 2713, + "time_per_iteration": 2.638127565383911 + }, + { + "auxiliary_loss_clip": 0.01192637, + "auxiliary_loss_mlp": 0.01032815, + "balance_loss_clip": 1.0572269, + "balance_loss_mlp": 1.02496517, + "epoch": 0.3263392051944929, + "flos": 20923496346240.0, + "grad_norm": 1.8082515570325104, + "language_loss": 0.87775457, + "learning_rate": 3.147542787891899e-06, + "loss": 0.9000091, + "num_input_tokens_seen": 58485435, + "step": 2714, + "time_per_iteration": 2.644523859024048 + }, + { + "auxiliary_loss_clip": 0.0114818, + "auxiliary_loss_mlp": 0.01033899, + "balance_loss_clip": 0.9825356, + "balance_loss_mlp": 1.02537525, + "epoch": 0.32645944808513194, + "flos": 24025172682240.0, + "grad_norm": 1.9331240305321273, + "language_loss": 0.7554431, + "learning_rate": 3.1469047110149926e-06, + "loss": 0.77726388, + "num_input_tokens_seen": 58504175, + "step": 2715, + "time_per_iteration": 2.730928897857666 + }, + { + "auxiliary_loss_clip": 0.01091588, + "auxiliary_loss_mlp": 0.01028249, + "balance_loss_clip": 0.90237546, + "balance_loss_mlp": 1.01983261, + "epoch": 0.32657969097577105, + "flos": 21032413361280.0, + "grad_norm": 2.3733767595452377, + "language_loss": 0.84729087, + "learning_rate": 3.146266460153554e-06, + "loss": 0.86848927, + "num_input_tokens_seen": 58523885, + "step": 2716, + "time_per_iteration": 2.7685134410858154 + }, + { + "auxiliary_loss_clip": 0.01142983, + "auxiliary_loss_mlp": 0.00511827, + "balance_loss_clip": 0.97973222, + "balance_loss_mlp": 0.0, + "epoch": 0.32669993386641016, + "flos": 22710267509760.0, + "grad_norm": 1.9022081533686412, + "language_loss": 0.80100811, + "learning_rate": 3.145628035404404e-06, + "loss": 0.8175562, + "num_input_tokens_seen": 58543085, + "step": 2717, + "time_per_iteration": 2.760435104370117 + }, + { + "auxiliary_loss_clip": 0.01073063, + "auxiliary_loss_mlp": 0.01008311, + "balance_loss_clip": 0.99017358, + "balance_loss_mlp": 1.00639141, + "epoch": 0.3268201767570492, + "flos": 72105718406400.0, + "grad_norm": 0.886149606669382, + "language_loss": 0.57538301, + "learning_rate": 3.1449894368643922e-06, + "loss": 0.59619671, + "num_input_tokens_seen": 58605400, + "step": 2718, + "time_per_iteration": 3.310676336288452 + }, + { + "auxiliary_loss_clip": 0.01123779, + "auxiliary_loss_mlp": 0.01039513, + "balance_loss_clip": 0.94530356, + "balance_loss_mlp": 1.03128207, + "epoch": 0.32694041964768833, + "flos": 24535175938560.0, + "grad_norm": 1.5362630856651769, + "language_loss": 0.71621728, + "learning_rate": 3.1443506646303934e-06, + "loss": 0.73785019, + "num_input_tokens_seen": 58626700, + "step": 2719, + "time_per_iteration": 3.6634182929992676 + }, + { + "auxiliary_loss_clip": 0.0117511, + "auxiliary_loss_mlp": 0.01031889, + "balance_loss_clip": 1.01954794, + "balance_loss_mlp": 1.02406263, + "epoch": 0.32706066253832744, + "flos": 33183003755520.0, + "grad_norm": 1.8579092721027513, + "language_loss": 0.67234087, + "learning_rate": 3.1437117187993086e-06, + "loss": 0.69441092, + "num_input_tokens_seen": 58649020, + "step": 2720, + "time_per_iteration": 2.728026866912842 + }, + { + "auxiliary_loss_clip": 0.01117064, + "auxiliary_loss_mlp": 0.0103304, + "balance_loss_clip": 0.93851453, + "balance_loss_mlp": 1.02441549, + "epoch": 0.3271809054289665, + "flos": 24061622008320.0, + "grad_norm": 1.6526909787573494, + "language_loss": 0.79560649, + "learning_rate": 3.143072599468065e-06, + "loss": 0.81710756, + "num_input_tokens_seen": 58668845, + "step": 2721, + "time_per_iteration": 3.6980104446411133 + }, + { + "auxiliary_loss_clip": 0.01146684, + "auxiliary_loss_mlp": 0.01030672, + "balance_loss_clip": 0.98440444, + "balance_loss_mlp": 1.02222586, + "epoch": 0.3273011483196056, + "flos": 38253769712640.0, + "grad_norm": 1.4405984362197126, + "language_loss": 0.75891954, + "learning_rate": 3.1424333067336174e-06, + "loss": 0.78069305, + "num_input_tokens_seen": 58691610, + "step": 2722, + "time_per_iteration": 3.822972059249878 + }, + { + "auxiliary_loss_clip": 0.01175771, + "auxiliary_loss_mlp": 0.01029093, + "balance_loss_clip": 1.01976347, + "balance_loss_mlp": 1.02017593, + "epoch": 0.3274213912102447, + "flos": 29054389582080.0, + "grad_norm": 1.732077672738972, + "language_loss": 0.77919781, + "learning_rate": 3.141793840692945e-06, + "loss": 0.8012464, + "num_input_tokens_seen": 58712360, + "step": 2723, + "time_per_iteration": 2.6898789405822754 + }, + { + "auxiliary_loss_clip": 0.01142746, + "auxiliary_loss_mlp": 0.01033017, + "balance_loss_clip": 0.98165649, + "balance_loss_mlp": 1.02373087, + "epoch": 0.32754163410088377, + "flos": 29133249891840.0, + "grad_norm": 2.3783829913597665, + "language_loss": 0.614968, + "learning_rate": 3.1411542014430553e-06, + "loss": 0.6367256, + "num_input_tokens_seen": 58733440, + "step": 2724, + "time_per_iteration": 2.906001567840576 + }, + { + "auxiliary_loss_clip": 0.01120715, + "auxiliary_loss_mlp": 0.01030437, + "balance_loss_clip": 0.93891174, + "balance_loss_mlp": 1.02237272, + "epoch": 0.3276618769915229, + "flos": 20631075724800.0, + "grad_norm": 2.3928934122005865, + "language_loss": 0.81711066, + "learning_rate": 3.1405143890809804e-06, + "loss": 0.83862221, + "num_input_tokens_seen": 58752735, + "step": 2725, + "time_per_iteration": 2.737955331802368 + }, + { + "auxiliary_loss_clip": 0.0114134, + "auxiliary_loss_mlp": 0.01032005, + "balance_loss_clip": 0.97941989, + "balance_loss_mlp": 1.02407753, + "epoch": 0.327782119882162, + "flos": 18657425076480.0, + "grad_norm": 1.715836911972806, + "language_loss": 0.6983155, + "learning_rate": 3.1398744037037796e-06, + "loss": 0.7200489, + "num_input_tokens_seen": 58772070, + "step": 2726, + "time_per_iteration": 2.6486613750457764 + }, + { + "auxiliary_loss_clip": 0.01144811, + "auxiliary_loss_mlp": 0.01033515, + "balance_loss_clip": 0.98167384, + "balance_loss_mlp": 1.02542686, + "epoch": 0.32790236277280105, + "flos": 21795802133760.0, + "grad_norm": 2.1530289424313818, + "language_loss": 0.84011412, + "learning_rate": 3.139234245408538e-06, + "loss": 0.86189735, + "num_input_tokens_seen": 58790950, + "step": 2727, + "time_per_iteration": 3.6838788986206055 + }, + { + "auxiliary_loss_clip": 0.01123238, + "auxiliary_loss_mlp": 0.0051084, + "balance_loss_clip": 0.94323772, + "balance_loss_mlp": 0.0, + "epoch": 0.32802260566344016, + "flos": 23331414424320.0, + "grad_norm": 1.6908039178181167, + "language_loss": 0.76045841, + "learning_rate": 3.1385939142923666e-06, + "loss": 0.7767992, + "num_input_tokens_seen": 58813340, + "step": 2728, + "time_per_iteration": 2.750662088394165 + }, + { + "auxiliary_loss_clip": 0.0114711, + "auxiliary_loss_mlp": 0.01032193, + "balance_loss_clip": 0.97852373, + "balance_loss_mlp": 1.02321696, + "epoch": 0.3281428485540792, + "flos": 24206988349440.0, + "grad_norm": 2.134573082369391, + "language_loss": 0.78716159, + "learning_rate": 3.137953410452405e-06, + "loss": 0.8089546, + "num_input_tokens_seen": 58833610, + "step": 2729, + "time_per_iteration": 2.7382724285125732 + }, + { + "auxiliary_loss_clip": 0.01141614, + "auxiliary_loss_mlp": 0.0103106, + "balance_loss_clip": 0.97829759, + "balance_loss_mlp": 1.02201223, + "epoch": 0.3282630914447183, + "flos": 34128962380800.0, + "grad_norm": 1.695067179558879, + "language_loss": 0.74334025, + "learning_rate": 3.1373127339858146e-06, + "loss": 0.76506698, + "num_input_tokens_seen": 58856210, + "step": 2730, + "time_per_iteration": 2.8009867668151855 + }, + { + "auxiliary_loss_clip": 0.01116812, + "auxiliary_loss_mlp": 0.01025377, + "balance_loss_clip": 0.93905312, + "balance_loss_mlp": 1.01792693, + "epoch": 0.32838333433535744, + "flos": 27600726170880.0, + "grad_norm": 2.1732468125892757, + "language_loss": 0.74481702, + "learning_rate": 3.136671884989787e-06, + "loss": 0.76623893, + "num_input_tokens_seen": 58876120, + "step": 2731, + "time_per_iteration": 2.76954984664917 + }, + { + "auxiliary_loss_clip": 0.01077484, + "auxiliary_loss_mlp": 0.01032561, + "balance_loss_clip": 0.86502302, + "balance_loss_mlp": 1.02387023, + "epoch": 0.3285035772259965, + "flos": 12349500935040.0, + "grad_norm": 2.4934387498219492, + "language_loss": 0.87118924, + "learning_rate": 3.1360308635615383e-06, + "loss": 0.8922897, + "num_input_tokens_seen": 58894660, + "step": 2732, + "time_per_iteration": 2.75203275680542 + }, + { + "auxiliary_loss_clip": 0.01155121, + "auxiliary_loss_mlp": 0.01034796, + "balance_loss_clip": 0.98333979, + "balance_loss_mlp": 1.02515185, + "epoch": 0.3286238201166356, + "flos": 24316084932480.0, + "grad_norm": 2.235420759591003, + "language_loss": 0.78663832, + "learning_rate": 3.135389669798311e-06, + "loss": 0.80853748, + "num_input_tokens_seen": 58912720, + "step": 2733, + "time_per_iteration": 2.648494005203247 + }, + { + "auxiliary_loss_clip": 0.01170518, + "auxiliary_loss_mlp": 0.00510819, + "balance_loss_clip": 1.01865649, + "balance_loss_mlp": 0.0, + "epoch": 0.3287440630072747, + "flos": 21392812471680.0, + "grad_norm": 2.121026960442847, + "language_loss": 0.79989487, + "learning_rate": 3.134748303797373e-06, + "loss": 0.81670821, + "num_input_tokens_seen": 58930090, + "step": 2734, + "time_per_iteration": 2.6752512454986572 + }, + { + "auxiliary_loss_clip": 0.01098351, + "auxiliary_loss_mlp": 0.01028028, + "balance_loss_clip": 0.9025445, + "balance_loss_mlp": 1.01872993, + "epoch": 0.32886430589791377, + "flos": 23732536579200.0, + "grad_norm": 3.7152034752769536, + "language_loss": 0.80770266, + "learning_rate": 3.1341067656560203e-06, + "loss": 0.8289665, + "num_input_tokens_seen": 58947935, + "step": 2735, + "time_per_iteration": 2.7490687370300293 + }, + { + "auxiliary_loss_clip": 0.01157937, + "auxiliary_loss_mlp": 0.01029558, + "balance_loss_clip": 0.98216629, + "balance_loss_mlp": 1.02145803, + "epoch": 0.3289845487885529, + "flos": 22418708814720.0, + "grad_norm": 1.803115068373966, + "language_loss": 0.86470795, + "learning_rate": 3.133465055471572e-06, + "loss": 0.88658285, + "num_input_tokens_seen": 58967720, + "step": 2736, + "time_per_iteration": 2.704688549041748 + }, + { + "auxiliary_loss_clip": 0.0111981, + "auxiliary_loss_mlp": 0.01034188, + "balance_loss_clip": 0.9419027, + "balance_loss_mlp": 1.02528274, + "epoch": 0.329104791679192, + "flos": 19682603147520.0, + "grad_norm": 2.4259943980551895, + "language_loss": 0.66079795, + "learning_rate": 3.1328231733413767e-06, + "loss": 0.68233788, + "num_input_tokens_seen": 58984360, + "step": 2737, + "time_per_iteration": 2.7056174278259277 + }, + { + "auxiliary_loss_clip": 0.01165682, + "auxiliary_loss_mlp": 0.01030145, + "balance_loss_clip": 1.01834631, + "balance_loss_mlp": 1.02168167, + "epoch": 0.32922503456983104, + "flos": 15997234803840.0, + "grad_norm": 2.5330049180583236, + "language_loss": 0.90975904, + "learning_rate": 3.1321811193628067e-06, + "loss": 0.93171728, + "num_input_tokens_seen": 59002505, + "step": 2738, + "time_per_iteration": 2.6387693881988525 + }, + { + "auxiliary_loss_clip": 0.01173015, + "auxiliary_loss_mlp": 0.00511878, + "balance_loss_clip": 1.02153051, + "balance_loss_mlp": 0.0, + "epoch": 0.32934527746047015, + "flos": 26834069260800.0, + "grad_norm": 2.256479826319418, + "language_loss": 0.69926703, + "learning_rate": 3.131538893633261e-06, + "loss": 0.71611601, + "num_input_tokens_seen": 59022065, + "step": 2739, + "time_per_iteration": 2.7304670810699463 + }, + { + "auxiliary_loss_clip": 0.01196843, + "auxiliary_loss_mlp": 0.01031051, + "balance_loss_clip": 1.05982876, + "balance_loss_mlp": 1.02264702, + "epoch": 0.32946552035110926, + "flos": 23403774372480.0, + "grad_norm": 2.432498564876427, + "language_loss": 0.77965391, + "learning_rate": 3.130896496250165e-06, + "loss": 0.80193293, + "num_input_tokens_seen": 59041890, + "step": 2740, + "time_per_iteration": 2.700230598449707 + }, + { + "auxiliary_loss_clip": 0.01195471, + "auxiliary_loss_mlp": 0.01028754, + "balance_loss_clip": 1.0579133, + "balance_loss_mlp": 1.02015924, + "epoch": 0.3295857632417483, + "flos": 14172470029440.0, + "grad_norm": 4.4709006261253, + "language_loss": 0.86710441, + "learning_rate": 3.1302539273109693e-06, + "loss": 0.88934672, + "num_input_tokens_seen": 59058715, + "step": 2741, + "time_per_iteration": 2.557319402694702 + }, + { + "auxiliary_loss_clip": 0.01140961, + "auxiliary_loss_mlp": 0.01035541, + "balance_loss_clip": 0.98085827, + "balance_loss_mlp": 1.02673161, + "epoch": 0.32970600613238743, + "flos": 22196708807040.0, + "grad_norm": 1.6568811618624315, + "language_loss": 0.80386949, + "learning_rate": 3.1296111869131513e-06, + "loss": 0.82563448, + "num_input_tokens_seen": 59076140, + "step": 2742, + "time_per_iteration": 2.673118829727173 + }, + { + "auxiliary_loss_clip": 0.01195773, + "auxiliary_loss_mlp": 0.01028766, + "balance_loss_clip": 1.05971897, + "balance_loss_mlp": 1.01984334, + "epoch": 0.32982624902302654, + "flos": 22053784590720.0, + "grad_norm": 2.4358596063110034, + "language_loss": 0.85860997, + "learning_rate": 3.1289682751542153e-06, + "loss": 0.88085538, + "num_input_tokens_seen": 59095700, + "step": 2743, + "time_per_iteration": 2.641315460205078 + }, + { + "auxiliary_loss_clip": 0.01169781, + "auxiliary_loss_mlp": 0.0103203, + "balance_loss_clip": 1.0192554, + "balance_loss_mlp": 1.02340519, + "epoch": 0.3299464919136656, + "flos": 18661626967680.0, + "grad_norm": 1.8979333540594776, + "language_loss": 0.70998371, + "learning_rate": 3.1283251921316883e-06, + "loss": 0.73200184, + "num_input_tokens_seen": 59113445, + "step": 2744, + "time_per_iteration": 3.5018868446350098 + }, + { + "auxiliary_loss_clip": 0.01099796, + "auxiliary_loss_mlp": 0.01032802, + "balance_loss_clip": 0.90634298, + "balance_loss_mlp": 1.02419519, + "epoch": 0.3300667348043047, + "flos": 13407357404160.0, + "grad_norm": 2.3616078060611576, + "language_loss": 0.80458003, + "learning_rate": 3.1276819379431277e-06, + "loss": 0.82590604, + "num_input_tokens_seen": 59131535, + "step": 2745, + "time_per_iteration": 2.77675461769104 + }, + { + "auxiliary_loss_clip": 0.01155422, + "auxiliary_loss_mlp": 0.00511373, + "balance_loss_clip": 0.98370665, + "balance_loss_mlp": 0.0, + "epoch": 0.33018697769494376, + "flos": 15742556398080.0, + "grad_norm": 1.854226062076134, + "language_loss": 0.75091141, + "learning_rate": 3.1270385126861134e-06, + "loss": 0.76757932, + "num_input_tokens_seen": 59149520, + "step": 2746, + "time_per_iteration": 2.6269257068634033 + }, + { + "auxiliary_loss_clip": 0.01198001, + "auxiliary_loss_mlp": 0.01035514, + "balance_loss_clip": 1.05905128, + "balance_loss_mlp": 1.02647805, + "epoch": 0.3303072205855829, + "flos": 18258601392000.0, + "grad_norm": 3.3584157138501207, + "language_loss": 0.82109857, + "learning_rate": 3.1263949164582533e-06, + "loss": 0.84343368, + "num_input_tokens_seen": 59169170, + "step": 2747, + "time_per_iteration": 3.562481641769409 + }, + { + "auxiliary_loss_clip": 0.01195852, + "auxiliary_loss_mlp": 0.01027368, + "balance_loss_clip": 1.05717635, + "balance_loss_mlp": 1.01775956, + "epoch": 0.330427463476222, + "flos": 17749424148480.0, + "grad_norm": 2.043424176422003, + "language_loss": 0.77957046, + "learning_rate": 3.1257511493571797e-06, + "loss": 0.80180264, + "num_input_tokens_seen": 59187675, + "step": 2748, + "time_per_iteration": 3.545828342437744 + }, + { + "auxiliary_loss_clip": 0.01131582, + "auxiliary_loss_mlp": 0.01030993, + "balance_loss_clip": 0.94690859, + "balance_loss_mlp": 1.02248788, + "epoch": 0.33054770636686104, + "flos": 27162580072320.0, + "grad_norm": 2.00177326120584, + "language_loss": 0.78574032, + "learning_rate": 3.125107211480552e-06, + "loss": 0.80736607, + "num_input_tokens_seen": 59207610, + "step": 2749, + "time_per_iteration": 2.7541399002075195 + }, + { + "auxiliary_loss_clip": 0.01074764, + "auxiliary_loss_mlp": 0.01031055, + "balance_loss_clip": 0.86517203, + "balance_loss_mlp": 1.02259755, + "epoch": 0.33066794925750015, + "flos": 20117193799680.0, + "grad_norm": 1.5695222762256182, + "language_loss": 0.80036163, + "learning_rate": 3.124463102926054e-06, + "loss": 0.82141984, + "num_input_tokens_seen": 59226945, + "step": 2750, + "time_per_iteration": 2.840388059616089 + }, + { + "auxiliary_loss_clip": 0.01077942, + "auxiliary_loss_mlp": 0.01009976, + "balance_loss_clip": 0.99713075, + "balance_loss_mlp": 1.00828314, + "epoch": 0.33078819214813926, + "flos": 70642609718400.0, + "grad_norm": 0.7928519322890083, + "language_loss": 0.61572331, + "learning_rate": 3.1238188237913984e-06, + "loss": 0.63660252, + "num_input_tokens_seen": 59291485, + "step": 2751, + "time_per_iteration": 3.28607177734375 + }, + { + "auxiliary_loss_clip": 0.01202711, + "auxiliary_loss_mlp": 0.01037401, + "balance_loss_clip": 1.06169629, + "balance_loss_mlp": 1.02760756, + "epoch": 0.3309084350387783, + "flos": 21141940907520.0, + "grad_norm": 3.434968442406112, + "language_loss": 0.76690644, + "learning_rate": 3.1231743741743202e-06, + "loss": 0.78930759, + "num_input_tokens_seen": 59310990, + "step": 2752, + "time_per_iteration": 2.6280593872070312 + }, + { + "auxiliary_loss_clip": 0.0116613, + "auxiliary_loss_mlp": 0.01024107, + "balance_loss_clip": 1.01675677, + "balance_loss_mlp": 1.01539278, + "epoch": 0.3310286779294174, + "flos": 14209350318720.0, + "grad_norm": 2.673095947628867, + "language_loss": 0.83750153, + "learning_rate": 3.122529754172582e-06, + "loss": 0.85940397, + "num_input_tokens_seen": 59327875, + "step": 2753, + "time_per_iteration": 3.5252647399902344 + }, + { + "auxiliary_loss_clip": 0.01173284, + "auxiliary_loss_mlp": 0.01031842, + "balance_loss_clip": 1.02188241, + "balance_loss_mlp": 1.02322292, + "epoch": 0.33114892082005654, + "flos": 20778130005120.0, + "grad_norm": 2.01540617418051, + "language_loss": 0.7247625, + "learning_rate": 3.1218849638839736e-06, + "loss": 0.74681377, + "num_input_tokens_seen": 59347135, + "step": 2754, + "time_per_iteration": 2.6760048866271973 + }, + { + "auxiliary_loss_clip": 0.01111969, + "auxiliary_loss_mlp": 0.01028891, + "balance_loss_clip": 0.93738317, + "balance_loss_mlp": 1.02042699, + "epoch": 0.3312691637106956, + "flos": 17090750499840.0, + "grad_norm": 1.8103918334018436, + "language_loss": 0.78079957, + "learning_rate": 3.121240003406307e-06, + "loss": 0.80220819, + "num_input_tokens_seen": 59365985, + "step": 2755, + "time_per_iteration": 2.7235829830169678 + }, + { + "auxiliary_loss_clip": 0.01133311, + "auxiliary_loss_mlp": 0.01031381, + "balance_loss_clip": 0.9468739, + "balance_loss_mlp": 1.02219033, + "epoch": 0.3313894066013347, + "flos": 29456230008960.0, + "grad_norm": 1.9897271756279662, + "language_loss": 0.72167683, + "learning_rate": 3.120594872837425e-06, + "loss": 0.74332368, + "num_input_tokens_seen": 59384655, + "step": 2756, + "time_per_iteration": 2.7219529151916504 + }, + { + "auxiliary_loss_clip": 0.01078575, + "auxiliary_loss_mlp": 0.00502992, + "balance_loss_clip": 0.99545109, + "balance_loss_mlp": 0.0, + "epoch": 0.3315096494919738, + "flos": 61419242280960.0, + "grad_norm": 0.8286665531176847, + "language_loss": 0.62360597, + "learning_rate": 3.1199495722751906e-06, + "loss": 0.63942158, + "num_input_tokens_seen": 59444185, + "step": 2757, + "time_per_iteration": 3.3219449520111084 + }, + { + "auxiliary_loss_clip": 0.01099985, + "auxiliary_loss_mlp": 0.01032606, + "balance_loss_clip": 0.9018513, + "balance_loss_mlp": 1.02397561, + "epoch": 0.33162989238261287, + "flos": 21653057485440.0, + "grad_norm": 1.8692635346966209, + "language_loss": 0.83593822, + "learning_rate": 3.1193041018174972e-06, + "loss": 0.85726416, + "num_input_tokens_seen": 59464900, + "step": 2758, + "time_per_iteration": 2.751648426055908 + }, + { + "auxiliary_loss_clip": 0.01175893, + "auxiliary_loss_mlp": 0.0103016, + "balance_loss_clip": 1.02159393, + "balance_loss_mlp": 1.02150536, + "epoch": 0.331750135273252, + "flos": 22674787850880.0, + "grad_norm": 2.209430738771797, + "language_loss": 0.94476771, + "learning_rate": 3.118658461562261e-06, + "loss": 0.96682811, + "num_input_tokens_seen": 59481000, + "step": 2759, + "time_per_iteration": 2.6522305011749268 + }, + { + "auxiliary_loss_clip": 0.01148282, + "auxiliary_loss_mlp": 0.01026151, + "balance_loss_clip": 0.98473459, + "balance_loss_mlp": 1.01734185, + "epoch": 0.33187037816389103, + "flos": 22746896403840.0, + "grad_norm": 1.7105970063230054, + "language_loss": 0.84986269, + "learning_rate": 3.118012651607426e-06, + "loss": 0.87160701, + "num_input_tokens_seen": 59502605, + "step": 2760, + "time_per_iteration": 2.7217724323272705 + }, + { + "auxiliary_loss_clip": 0.01197095, + "auxiliary_loss_mlp": 0.01030758, + "balance_loss_clip": 1.06000805, + "balance_loss_mlp": 1.02129316, + "epoch": 0.33199062105453014, + "flos": 19203769918080.0, + "grad_norm": 1.9710077919763116, + "language_loss": 0.83736014, + "learning_rate": 3.1173666720509603e-06, + "loss": 0.85963869, + "num_input_tokens_seen": 59519540, + "step": 2761, + "time_per_iteration": 2.6947295665740967 + }, + { + "auxiliary_loss_clip": 0.0115563, + "auxiliary_loss_mlp": 0.01031189, + "balance_loss_clip": 0.98398954, + "balance_loss_mlp": 1.02191424, + "epoch": 0.33211086394516925, + "flos": 31577006764800.0, + "grad_norm": 1.675106612888117, + "language_loss": 0.67892677, + "learning_rate": 3.116720522990859e-06, + "loss": 0.70079494, + "num_input_tokens_seen": 59540415, + "step": 2762, + "time_per_iteration": 2.7820234298706055 + }, + { + "auxiliary_loss_clip": 0.010779, + "auxiliary_loss_mlp": 0.0103509, + "balance_loss_clip": 0.8669796, + "balance_loss_mlp": 1.02657819, + "epoch": 0.3322311068358083, + "flos": 17932496791680.0, + "grad_norm": 4.064054405694194, + "language_loss": 0.62334514, + "learning_rate": 3.116074204525142e-06, + "loss": 0.6444751, + "num_input_tokens_seen": 59558590, + "step": 2763, + "time_per_iteration": 2.74786639213562 + }, + { + "auxiliary_loss_clip": 0.01163503, + "auxiliary_loss_mlp": 0.01029622, + "balance_loss_clip": 1.01847756, + "balance_loss_mlp": 1.02043104, + "epoch": 0.3323513497264474, + "flos": 32269831269120.0, + "grad_norm": 2.312223918861448, + "language_loss": 0.83556128, + "learning_rate": 3.1154277167518553e-06, + "loss": 0.85749251, + "num_input_tokens_seen": 59580205, + "step": 2764, + "time_per_iteration": 2.863128900527954 + }, + { + "auxiliary_loss_clip": 0.01048023, + "auxiliary_loss_mlp": 0.01007273, + "balance_loss_clip": 0.951738, + "balance_loss_mlp": 1.00548482, + "epoch": 0.33247159261708653, + "flos": 52668674588160.0, + "grad_norm": 0.7788242085728072, + "language_loss": 0.59439421, + "learning_rate": 3.114781059769072e-06, + "loss": 0.6149472, + "num_input_tokens_seen": 59631530, + "step": 2765, + "time_per_iteration": 3.0846776962280273 + }, + { + "auxiliary_loss_clip": 0.01147463, + "auxiliary_loss_mlp": 0.01029378, + "balance_loss_clip": 0.98115242, + "balance_loss_mlp": 1.02016878, + "epoch": 0.3325918355077256, + "flos": 27125232906240.0, + "grad_norm": 3.0609341502761778, + "language_loss": 0.6791575, + "learning_rate": 3.1141342336748874e-06, + "loss": 0.70092595, + "num_input_tokens_seen": 59651090, + "step": 2766, + "time_per_iteration": 2.75924015045166 + }, + { + "auxiliary_loss_clip": 0.01168448, + "auxiliary_loss_mlp": 0.01027911, + "balance_loss_clip": 1.02116644, + "balance_loss_mlp": 1.01969171, + "epoch": 0.3327120783983647, + "flos": 23664414435840.0, + "grad_norm": 3.196449058646047, + "language_loss": 0.82094955, + "learning_rate": 3.1134872385674253e-06, + "loss": 0.84291321, + "num_input_tokens_seen": 59675245, + "step": 2767, + "time_per_iteration": 2.7289841175079346 + }, + { + "auxiliary_loss_clip": 0.01151241, + "auxiliary_loss_mlp": 0.01031238, + "balance_loss_clip": 0.97994757, + "balance_loss_mlp": 1.02315581, + "epoch": 0.3328323212890038, + "flos": 19171378828800.0, + "grad_norm": 1.8565747661831138, + "language_loss": 0.86137331, + "learning_rate": 3.1128400745448353e-06, + "loss": 0.88319814, + "num_input_tokens_seen": 59694625, + "step": 2768, + "time_per_iteration": 2.6615898609161377 + }, + { + "auxiliary_loss_clip": 0.01176971, + "auxiliary_loss_mlp": 0.01033489, + "balance_loss_clip": 1.02250266, + "balance_loss_mlp": 1.02441716, + "epoch": 0.33295256417964286, + "flos": 37706347463040.0, + "grad_norm": 2.111098696086957, + "language_loss": 0.62673724, + "learning_rate": 3.11219274170529e-06, + "loss": 0.64884186, + "num_input_tokens_seen": 59716435, + "step": 2769, + "time_per_iteration": 2.826403856277466 + }, + { + "auxiliary_loss_clip": 0.01143726, + "auxiliary_loss_mlp": 0.01028983, + "balance_loss_clip": 0.98085022, + "balance_loss_mlp": 1.02074552, + "epoch": 0.333072807070282, + "flos": 26505989412480.0, + "grad_norm": 2.515858884347187, + "language_loss": 0.81749552, + "learning_rate": 3.1115452401469903e-06, + "loss": 0.83922261, + "num_input_tokens_seen": 59736835, + "step": 2770, + "time_per_iteration": 3.6759989261627197 + }, + { + "auxiliary_loss_clip": 0.01093242, + "auxiliary_loss_mlp": 0.01032364, + "balance_loss_clip": 0.90032971, + "balance_loss_mlp": 1.02358985, + "epoch": 0.3331930499609211, + "flos": 21430913823360.0, + "grad_norm": 2.0096226423721086, + "language_loss": 0.86268675, + "learning_rate": 3.1108975699681613e-06, + "loss": 0.88394278, + "num_input_tokens_seen": 59754230, + "step": 2771, + "time_per_iteration": 2.7978196144104004 + }, + { + "auxiliary_loss_clip": 0.01124161, + "auxiliary_loss_mlp": 0.01032739, + "balance_loss_clip": 0.94470912, + "balance_loss_mlp": 1.02364278, + "epoch": 0.33331329285156014, + "flos": 20659947281280.0, + "grad_norm": 2.708509214388986, + "language_loss": 0.71415687, + "learning_rate": 3.1102497312670542e-06, + "loss": 0.73572582, + "num_input_tokens_seen": 59772235, + "step": 2772, + "time_per_iteration": 2.8298494815826416 + }, + { + "auxiliary_loss_clip": 0.01137021, + "auxiliary_loss_mlp": 0.01030581, + "balance_loss_clip": 0.98012292, + "balance_loss_mlp": 1.02199173, + "epoch": 0.33343353574219925, + "flos": 28001596930560.0, + "grad_norm": 2.6716105562992283, + "language_loss": 0.80319703, + "learning_rate": 3.109601724141946e-06, + "loss": 0.82487297, + "num_input_tokens_seen": 59791230, + "step": 2773, + "time_per_iteration": 3.628039598464966 + }, + { + "auxiliary_loss_clip": 0.01145269, + "auxiliary_loss_mlp": 0.01037114, + "balance_loss_clip": 0.98107934, + "balance_loss_mlp": 1.02811313, + "epoch": 0.33355377863283836, + "flos": 23764963582080.0, + "grad_norm": 1.9811005036890215, + "language_loss": 0.6797955, + "learning_rate": 3.108953548691138e-06, + "loss": 0.70161933, + "num_input_tokens_seen": 59811315, + "step": 2774, + "time_per_iteration": 3.61875581741333 + }, + { + "auxiliary_loss_clip": 0.01198513, + "auxiliary_loss_mlp": 0.01032876, + "balance_loss_clip": 1.0612402, + "balance_loss_mlp": 1.02350593, + "epoch": 0.3336740215234774, + "flos": 37779677078400.0, + "grad_norm": 4.130632000897245, + "language_loss": 0.72484487, + "learning_rate": 3.108305205012959e-06, + "loss": 0.74715877, + "num_input_tokens_seen": 59832010, + "step": 2775, + "time_per_iteration": 2.775219202041626 + }, + { + "auxiliary_loss_clip": 0.0114642, + "auxiliary_loss_mlp": 0.0103117, + "balance_loss_clip": 0.98157966, + "balance_loss_mlp": 1.02239001, + "epoch": 0.3337942644141165, + "flos": 25519056347520.0, + "grad_norm": 2.0251584682159596, + "language_loss": 0.87550032, + "learning_rate": 3.107656693205761e-06, + "loss": 0.89727622, + "num_input_tokens_seen": 59851450, + "step": 2776, + "time_per_iteration": 2.7422072887420654 + }, + { + "auxiliary_loss_clip": 0.01204073, + "auxiliary_loss_mlp": 0.01032956, + "balance_loss_clip": 1.0630374, + "balance_loss_mlp": 1.02377057, + "epoch": 0.3339145073047556, + "flos": 25989844930560.0, + "grad_norm": 3.8004666321664105, + "language_loss": 0.70352483, + "learning_rate": 3.107008013367924e-06, + "loss": 0.72589517, + "num_input_tokens_seen": 59870245, + "step": 2777, + "time_per_iteration": 2.6782381534576416 + }, + { + "auxiliary_loss_clip": 0.01124572, + "auxiliary_loss_mlp": 0.01027611, + "balance_loss_clip": 0.94325912, + "balance_loss_mlp": 1.01796746, + "epoch": 0.3340347501953947, + "flos": 19062569554560.0, + "grad_norm": 1.9778416701919541, + "language_loss": 0.86900818, + "learning_rate": 3.1063591655978507e-06, + "loss": 0.89052999, + "num_input_tokens_seen": 59886195, + "step": 2778, + "time_per_iteration": 2.685800790786743 + }, + { + "auxiliary_loss_clip": 0.01090623, + "auxiliary_loss_mlp": 0.01029161, + "balance_loss_clip": 0.9018442, + "balance_loss_mlp": 1.02044106, + "epoch": 0.3341549930860338, + "flos": 18109715518080.0, + "grad_norm": 1.9285014958227216, + "language_loss": 0.80057895, + "learning_rate": 3.105710149993972e-06, + "loss": 0.82177681, + "num_input_tokens_seen": 59905525, + "step": 2779, + "time_per_iteration": 3.690804958343506 + }, + { + "auxiliary_loss_clip": 0.01197751, + "auxiliary_loss_mlp": 0.0103052, + "balance_loss_clip": 1.05927682, + "balance_loss_mlp": 1.02178216, + "epoch": 0.33427523597667286, + "flos": 22674967418880.0, + "grad_norm": 2.110084156287277, + "language_loss": 0.85779834, + "learning_rate": 3.1050609666547427e-06, + "loss": 0.88008106, + "num_input_tokens_seen": 59925085, + "step": 2780, + "time_per_iteration": 2.615633964538574 + }, + { + "auxiliary_loss_clip": 0.01134667, + "auxiliary_loss_mlp": 0.01034349, + "balance_loss_clip": 0.94662321, + "balance_loss_mlp": 1.02545643, + "epoch": 0.33439547886731197, + "flos": 22638338524800.0, + "grad_norm": 1.8974209658644499, + "language_loss": 0.77637798, + "learning_rate": 3.104411615678644e-06, + "loss": 0.79806811, + "num_input_tokens_seen": 59943935, + "step": 2781, + "time_per_iteration": 2.801809310913086 + }, + { + "auxiliary_loss_clip": 0.01143255, + "auxiliary_loss_mlp": 0.01038281, + "balance_loss_clip": 0.98176557, + "balance_loss_mlp": 1.02917922, + "epoch": 0.3345157217579511, + "flos": 24096383395200.0, + "grad_norm": 2.6306632545997175, + "language_loss": 0.73555303, + "learning_rate": 3.1037620971641803e-06, + "loss": 0.75736833, + "num_input_tokens_seen": 59963725, + "step": 2782, + "time_per_iteration": 2.684736728668213 + }, + { + "auxiliary_loss_clip": 0.01197194, + "auxiliary_loss_mlp": 0.01030949, + "balance_loss_clip": 1.05865812, + "balance_loss_mlp": 1.02193654, + "epoch": 0.33463596464859013, + "flos": 18989491334400.0, + "grad_norm": 5.8871030028468505, + "language_loss": 0.64672017, + "learning_rate": 3.1031124112098844e-06, + "loss": 0.66900152, + "num_input_tokens_seen": 59981935, + "step": 2783, + "time_per_iteration": 2.6837024688720703 + }, + { + "auxiliary_loss_clip": 0.01152976, + "auxiliary_loss_mlp": 0.01034156, + "balance_loss_clip": 0.98460597, + "balance_loss_mlp": 1.02577591, + "epoch": 0.33475620753922924, + "flos": 20375607219840.0, + "grad_norm": 1.8534733927907052, + "language_loss": 0.72260958, + "learning_rate": 3.1024625579143127e-06, + "loss": 0.74448097, + "num_input_tokens_seen": 59999455, + "step": 2784, + "time_per_iteration": 2.696476697921753 + }, + { + "auxiliary_loss_clip": 0.01198322, + "auxiliary_loss_mlp": 0.01027482, + "balance_loss_clip": 1.0609076, + "balance_loss_mlp": 1.01868439, + "epoch": 0.33487645042986836, + "flos": 18182578256640.0, + "grad_norm": 1.7201044389902587, + "language_loss": 0.729913, + "learning_rate": 3.101812537376048e-06, + "loss": 0.75217104, + "num_input_tokens_seen": 60018475, + "step": 2785, + "time_per_iteration": 2.677490472793579 + }, + { + "auxiliary_loss_clip": 0.01141148, + "auxiliary_loss_mlp": 0.00511236, + "balance_loss_clip": 0.98025322, + "balance_loss_mlp": 0.0, + "epoch": 0.3349966933205074, + "flos": 25848824135040.0, + "grad_norm": 3.389532444301939, + "language_loss": 0.84364891, + "learning_rate": 3.1011623496936973e-06, + "loss": 0.86017275, + "num_input_tokens_seen": 60036770, + "step": 2786, + "time_per_iteration": 2.754316568374634 + }, + { + "auxiliary_loss_clip": 0.01200919, + "auxiliary_loss_mlp": 0.01028122, + "balance_loss_clip": 1.06366682, + "balance_loss_mlp": 1.01940227, + "epoch": 0.3351169362111465, + "flos": 28111447699200.0, + "grad_norm": 2.011902136479837, + "language_loss": 0.69618863, + "learning_rate": 3.100511994965893e-06, + "loss": 0.71847904, + "num_input_tokens_seen": 60056725, + "step": 2787, + "time_per_iteration": 2.675123929977417 + }, + { + "auxiliary_loss_clip": 0.01169389, + "auxiliary_loss_mlp": 0.01024331, + "balance_loss_clip": 1.02273881, + "balance_loss_mlp": 1.01590836, + "epoch": 0.33523717910178563, + "flos": 22673315393280.0, + "grad_norm": 2.850712742247936, + "language_loss": 0.84197509, + "learning_rate": 3.0998614732912947e-06, + "loss": 0.86391234, + "num_input_tokens_seen": 60076100, + "step": 2788, + "time_per_iteration": 2.6623599529266357 + }, + { + "auxiliary_loss_clip": 0.01174783, + "auxiliary_loss_mlp": 0.01032313, + "balance_loss_clip": 1.02377892, + "balance_loss_mlp": 1.02234125, + "epoch": 0.3353574219924247, + "flos": 15669801400320.0, + "grad_norm": 1.9565942907681668, + "language_loss": 0.67597282, + "learning_rate": 3.0992107847685855e-06, + "loss": 0.69804382, + "num_input_tokens_seen": 60093815, + "step": 2789, + "time_per_iteration": 2.5897507667541504 + }, + { + "auxiliary_loss_clip": 0.01154515, + "auxiliary_loss_mlp": 0.01037396, + "balance_loss_clip": 0.98803127, + "balance_loss_mlp": 1.02824044, + "epoch": 0.3354776648830638, + "flos": 24790644443520.0, + "grad_norm": 1.9643570343076888, + "language_loss": 0.79315329, + "learning_rate": 3.0985599294964736e-06, + "loss": 0.81507242, + "num_input_tokens_seen": 60113370, + "step": 2790, + "time_per_iteration": 2.707261085510254 + }, + { + "auxiliary_loss_clip": 0.01138032, + "auxiliary_loss_mlp": 0.01030877, + "balance_loss_clip": 0.9452858, + "balance_loss_mlp": 1.02182925, + "epoch": 0.33559790777370285, + "flos": 28694852398080.0, + "grad_norm": 1.8940292013588484, + "language_loss": 0.70042074, + "learning_rate": 3.097908907573695e-06, + "loss": 0.72210985, + "num_input_tokens_seen": 60131350, + "step": 2791, + "time_per_iteration": 2.7299723625183105 + }, + { + "auxiliary_loss_clip": 0.01078429, + "auxiliary_loss_mlp": 0.01036014, + "balance_loss_clip": 0.87092638, + "balance_loss_mlp": 1.02741933, + "epoch": 0.33571815066434196, + "flos": 22235779825920.0, + "grad_norm": 2.3517315086639092, + "language_loss": 0.89336634, + "learning_rate": 3.0972577190990067e-06, + "loss": 0.91451079, + "num_input_tokens_seen": 60149830, + "step": 2792, + "time_per_iteration": 2.7733001708984375 + }, + { + "auxiliary_loss_clip": 0.01133837, + "auxiliary_loss_mlp": 0.01031774, + "balance_loss_clip": 0.94656479, + "balance_loss_mlp": 1.02274942, + "epoch": 0.3358383935549811, + "flos": 23842279607040.0, + "grad_norm": 1.855361820923723, + "language_loss": 0.79712492, + "learning_rate": 3.096606364171196e-06, + "loss": 0.81878102, + "num_input_tokens_seen": 60169620, + "step": 2793, + "time_per_iteration": 2.6816651821136475 + }, + { + "auxiliary_loss_clip": 0.01110819, + "auxiliary_loss_mlp": 0.0103421, + "balance_loss_clip": 0.94219732, + "balance_loss_mlp": 1.02578759, + "epoch": 0.33595863644562013, + "flos": 22267308988800.0, + "grad_norm": 2.2491154460406113, + "language_loss": 0.85137558, + "learning_rate": 3.0959548428890703e-06, + "loss": 0.87282586, + "num_input_tokens_seen": 60188490, + "step": 2794, + "time_per_iteration": 2.8367366790771484 + }, + { + "auxiliary_loss_clip": 0.01171957, + "auxiliary_loss_mlp": 0.01027155, + "balance_loss_clip": 1.02597356, + "balance_loss_mlp": 1.01873326, + "epoch": 0.33607887933625924, + "flos": 20119779578880.0, + "grad_norm": 1.9727099551252476, + "language_loss": 0.84101588, + "learning_rate": 3.095303155351468e-06, + "loss": 0.86300701, + "num_input_tokens_seen": 60208695, + "step": 2795, + "time_per_iteration": 2.7518362998962402 + }, + { + "auxiliary_loss_clip": 0.01092108, + "auxiliary_loss_mlp": 0.01034552, + "balance_loss_clip": 0.90402722, + "balance_loss_mlp": 1.02595663, + "epoch": 0.33619912222689835, + "flos": 19318109886720.0, + "grad_norm": 2.7319129918248146, + "language_loss": 0.79451895, + "learning_rate": 3.0946513016572464e-06, + "loss": 0.81578553, + "num_input_tokens_seen": 60227600, + "step": 2796, + "time_per_iteration": 3.637193202972412 + }, + { + "auxiliary_loss_clip": 0.011765, + "auxiliary_loss_mlp": 0.01030496, + "balance_loss_clip": 1.02184987, + "balance_loss_mlp": 1.02122784, + "epoch": 0.3363193651175374, + "flos": 16800664262400.0, + "grad_norm": 1.9602523261772684, + "language_loss": 0.77041948, + "learning_rate": 3.0939992819052938e-06, + "loss": 0.79248935, + "num_input_tokens_seen": 60245110, + "step": 2797, + "time_per_iteration": 2.643490791320801 + }, + { + "auxiliary_loss_clip": 0.01150804, + "auxiliary_loss_mlp": 0.01034381, + "balance_loss_clip": 0.98534143, + "balance_loss_mlp": 1.02519643, + "epoch": 0.3364396080081765, + "flos": 23550289948800.0, + "grad_norm": 2.1944646936178818, + "language_loss": 0.81153673, + "learning_rate": 3.0933470961945193e-06, + "loss": 0.83338857, + "num_input_tokens_seen": 60263405, + "step": 2798, + "time_per_iteration": 2.693260908126831 + }, + { + "auxiliary_loss_clip": 0.01147545, + "auxiliary_loss_mlp": 0.01036419, + "balance_loss_clip": 0.98460126, + "balance_loss_mlp": 1.02747846, + "epoch": 0.3365598508988156, + "flos": 28037902602240.0, + "grad_norm": 3.46218996994493, + "language_loss": 0.68308139, + "learning_rate": 3.0926947446238597e-06, + "loss": 0.70492101, + "num_input_tokens_seen": 60282975, + "step": 2799, + "time_per_iteration": 3.678524971008301 + }, + { + "auxiliary_loss_clip": 0.01173885, + "auxiliary_loss_mlp": 0.01029124, + "balance_loss_clip": 1.016873, + "balance_loss_mlp": 1.01984346, + "epoch": 0.3366800937894547, + "flos": 16982767238400.0, + "grad_norm": 3.0956629136761427, + "language_loss": 0.82459217, + "learning_rate": 3.092042227292276e-06, + "loss": 0.84662223, + "num_input_tokens_seen": 60299810, + "step": 2800, + "time_per_iteration": 3.766993522644043 + }, + { + "auxiliary_loss_clip": 0.01194265, + "auxiliary_loss_mlp": 0.01025775, + "balance_loss_clip": 1.05945206, + "balance_loss_mlp": 1.01732874, + "epoch": 0.3368003366800938, + "flos": 23915321913600.0, + "grad_norm": 2.165380978407434, + "language_loss": 0.88397008, + "learning_rate": 3.0913895442987557e-06, + "loss": 0.90617049, + "num_input_tokens_seen": 60320775, + "step": 2801, + "time_per_iteration": 2.6758642196655273 + }, + { + "auxiliary_loss_clip": 0.01126884, + "auxiliary_loss_mlp": 0.00511637, + "balance_loss_clip": 0.94521576, + "balance_loss_mlp": 0.0, + "epoch": 0.3369205795707329, + "flos": 24791219061120.0, + "grad_norm": 1.7743387445624481, + "language_loss": 0.85880637, + "learning_rate": 3.090736695742308e-06, + "loss": 0.87519157, + "num_input_tokens_seen": 60341905, + "step": 2802, + "time_per_iteration": 2.78052020072937 + }, + { + "auxiliary_loss_clip": 0.01092872, + "auxiliary_loss_mlp": 0.01027688, + "balance_loss_clip": 0.9020654, + "balance_loss_mlp": 1.0183717, + "epoch": 0.33704082246137196, + "flos": 17931096161280.0, + "grad_norm": 2.586343875627259, + "language_loss": 0.52246201, + "learning_rate": 3.0900836817219713e-06, + "loss": 0.54366755, + "num_input_tokens_seen": 60358335, + "step": 2803, + "time_per_iteration": 2.7247064113616943 + }, + { + "auxiliary_loss_clip": 0.01195814, + "auxiliary_loss_mlp": 0.01030423, + "balance_loss_clip": 1.05835736, + "balance_loss_mlp": 1.02143478, + "epoch": 0.33716106535201107, + "flos": 21286517149440.0, + "grad_norm": 1.9640845285393675, + "language_loss": 0.83538377, + "learning_rate": 3.089430502336807e-06, + "loss": 0.85764617, + "num_input_tokens_seen": 60378305, + "step": 2804, + "time_per_iteration": 2.609968423843384 + }, + { + "auxiliary_loss_clip": 0.01178572, + "auxiliary_loss_mlp": 0.01032696, + "balance_loss_clip": 1.02396297, + "balance_loss_mlp": 1.02350497, + "epoch": 0.3372813082426502, + "flos": 18402962152320.0, + "grad_norm": 2.5181456541597065, + "language_loss": 0.9007448, + "learning_rate": 3.088777157685902e-06, + "loss": 0.92285752, + "num_input_tokens_seen": 60393895, + "step": 2805, + "time_per_iteration": 3.619097948074341 + }, + { + "auxiliary_loss_clip": 0.01144863, + "auxiliary_loss_mlp": 0.01028331, + "balance_loss_clip": 0.98247814, + "balance_loss_mlp": 1.01946807, + "epoch": 0.33740155113328923, + "flos": 17201391367680.0, + "grad_norm": 3.0699890939669703, + "language_loss": 0.85554004, + "learning_rate": 3.088123647868367e-06, + "loss": 0.87727195, + "num_input_tokens_seen": 60410445, + "step": 2806, + "time_per_iteration": 2.6589667797088623 + }, + { + "auxiliary_loss_clip": 0.01174231, + "auxiliary_loss_mlp": 0.01029957, + "balance_loss_clip": 1.01899076, + "balance_loss_mlp": 1.02107024, + "epoch": 0.33752179402392835, + "flos": 29058950609280.0, + "grad_norm": 1.9211962719305034, + "language_loss": 0.81353456, + "learning_rate": 3.0874699729833405e-06, + "loss": 0.83557647, + "num_input_tokens_seen": 60431815, + "step": 2807, + "time_per_iteration": 2.788699150085449 + }, + { + "auxiliary_loss_clip": 0.01146759, + "auxiliary_loss_mlp": 0.01031088, + "balance_loss_clip": 0.98257089, + "balance_loss_mlp": 1.02212882, + "epoch": 0.3376420369145674, + "flos": 25080730680960.0, + "grad_norm": 3.432859675476527, + "language_loss": 0.79865503, + "learning_rate": 3.086816133129983e-06, + "loss": 0.82043344, + "num_input_tokens_seen": 60452075, + "step": 2808, + "time_per_iteration": 2.7026243209838867 + }, + { + "auxiliary_loss_clip": 0.01198349, + "auxiliary_loss_mlp": 0.01033698, + "balance_loss_clip": 1.06241298, + "balance_loss_mlp": 1.02532947, + "epoch": 0.3377622798052065, + "flos": 27490624007040.0, + "grad_norm": 1.8678507839088463, + "language_loss": 0.7613892, + "learning_rate": 3.0861621284074826e-06, + "loss": 0.78370965, + "num_input_tokens_seen": 60472600, + "step": 2809, + "time_per_iteration": 2.666002035140991 + }, + { + "auxiliary_loss_clip": 0.01156149, + "auxiliary_loss_mlp": 0.01030981, + "balance_loss_clip": 0.98421681, + "balance_loss_mlp": 1.02203465, + "epoch": 0.3378825226958456, + "flos": 21975211589760.0, + "grad_norm": 1.574153607844322, + "language_loss": 0.73108459, + "learning_rate": 3.085507958915051e-06, + "loss": 0.75295591, + "num_input_tokens_seen": 60491030, + "step": 2810, + "time_per_iteration": 2.659644842147827 + }, + { + "auxiliary_loss_clip": 0.01142548, + "auxiliary_loss_mlp": 0.0102581, + "balance_loss_clip": 0.97978991, + "balance_loss_mlp": 1.01682746, + "epoch": 0.3380027655864847, + "flos": 42523189200000.0, + "grad_norm": 2.0611989606762893, + "language_loss": 0.71354604, + "learning_rate": 3.084853624751925e-06, + "loss": 0.73522961, + "num_input_tokens_seen": 60512615, + "step": 2811, + "time_per_iteration": 2.8856606483459473 + }, + { + "auxiliary_loss_clip": 0.0112859, + "auxiliary_loss_mlp": 0.01030221, + "balance_loss_clip": 0.94594061, + "balance_loss_mlp": 1.02122688, + "epoch": 0.3381230084771238, + "flos": 26725080418560.0, + "grad_norm": 1.8513442930597859, + "language_loss": 0.85757953, + "learning_rate": 3.0841991260173668e-06, + "loss": 0.87916768, + "num_input_tokens_seen": 60532520, + "step": 2812, + "time_per_iteration": 2.789208173751831 + }, + { + "auxiliary_loss_clip": 0.01199899, + "auxiliary_loss_mlp": 0.01027947, + "balance_loss_clip": 1.06182623, + "balance_loss_mlp": 1.01927447, + "epoch": 0.3382432513677629, + "flos": 22710375250560.0, + "grad_norm": 1.9467499765685397, + "language_loss": 0.80654144, + "learning_rate": 3.0835444628106634e-06, + "loss": 0.82881993, + "num_input_tokens_seen": 60551500, + "step": 2813, + "time_per_iteration": 2.711855411529541 + }, + { + "auxiliary_loss_clip": 0.01194983, + "auxiliary_loss_mlp": 0.00511093, + "balance_loss_clip": 1.0582552, + "balance_loss_mlp": 0.0, + "epoch": 0.33836349425840195, + "flos": 22122409524480.0, + "grad_norm": 1.9212469008939426, + "language_loss": 0.8308655, + "learning_rate": 3.082889635231126e-06, + "loss": 0.84792632, + "num_input_tokens_seen": 60570160, + "step": 2814, + "time_per_iteration": 2.63155198097229 + }, + { + "auxiliary_loss_clip": 0.01147473, + "auxiliary_loss_mlp": 0.01033907, + "balance_loss_clip": 0.97992635, + "balance_loss_mlp": 1.02449512, + "epoch": 0.33848373714904106, + "flos": 27308090067840.0, + "grad_norm": 2.365339612287938, + "language_loss": 0.76470661, + "learning_rate": 3.0822346433780925e-06, + "loss": 0.78652036, + "num_input_tokens_seen": 60590885, + "step": 2815, + "time_per_iteration": 2.672520399093628 + }, + { + "auxiliary_loss_clip": 0.01171336, + "auxiliary_loss_mlp": 0.010299, + "balance_loss_clip": 1.01847124, + "balance_loss_mlp": 1.02091193, + "epoch": 0.3386039800396802, + "flos": 25848716394240.0, + "grad_norm": 2.0885829792415884, + "language_loss": 0.86927676, + "learning_rate": 3.0815794873509237e-06, + "loss": 0.89128911, + "num_input_tokens_seen": 60609170, + "step": 2816, + "time_per_iteration": 2.7654497623443604 + }, + { + "auxiliary_loss_clip": 0.01195355, + "auxiliary_loss_mlp": 0.01026859, + "balance_loss_clip": 1.05878031, + "balance_loss_mlp": 1.01866972, + "epoch": 0.33872422293031923, + "flos": 18880646146560.0, + "grad_norm": 2.0262723081843097, + "language_loss": 0.72921085, + "learning_rate": 3.0809241672490066e-06, + "loss": 0.75143296, + "num_input_tokens_seen": 60627340, + "step": 2817, + "time_per_iteration": 2.605905532836914 + }, + { + "auxiliary_loss_clip": 0.01151541, + "auxiliary_loss_mlp": 0.01036075, + "balance_loss_clip": 0.98372877, + "balance_loss_mlp": 1.02791476, + "epoch": 0.33884446582095834, + "flos": 23146977064320.0, + "grad_norm": 1.7347236212469537, + "language_loss": 0.85239476, + "learning_rate": 3.080268683171753e-06, + "loss": 0.87427092, + "num_input_tokens_seen": 60647630, + "step": 2818, + "time_per_iteration": 2.673673629760742 + }, + { + "auxiliary_loss_clip": 0.01170253, + "auxiliary_loss_mlp": 0.01033243, + "balance_loss_clip": 1.01908922, + "balance_loss_mlp": 1.02494001, + "epoch": 0.33896470871159745, + "flos": 15997342544640.0, + "grad_norm": 2.166716219965086, + "language_loss": 0.88807011, + "learning_rate": 3.0796130352185985e-06, + "loss": 0.91010511, + "num_input_tokens_seen": 60664485, + "step": 2819, + "time_per_iteration": 2.627605676651001 + }, + { + "auxiliary_loss_clip": 0.01133784, + "auxiliary_loss_mlp": 0.00512054, + "balance_loss_clip": 0.97592664, + "balance_loss_mlp": 0.0, + "epoch": 0.3390849516022365, + "flos": 34495754112000.0, + "grad_norm": 1.9420631401919455, + "language_loss": 0.66853547, + "learning_rate": 3.0789572234890057e-06, + "loss": 0.68499386, + "num_input_tokens_seen": 60686125, + "step": 2820, + "time_per_iteration": 2.786991834640503 + }, + { + "auxiliary_loss_clip": 0.01147596, + "auxiliary_loss_mlp": 0.01031581, + "balance_loss_clip": 0.98339009, + "balance_loss_mlp": 1.02261055, + "epoch": 0.3392051944928756, + "flos": 16180307447040.0, + "grad_norm": 1.63939323389515, + "language_loss": 0.76937675, + "learning_rate": 3.0783012480824596e-06, + "loss": 0.79116857, + "num_input_tokens_seen": 60705270, + "step": 2821, + "time_per_iteration": 2.629301071166992 + }, + { + "auxiliary_loss_clip": 0.01193846, + "auxiliary_loss_mlp": 0.01029423, + "balance_loss_clip": 1.05708766, + "balance_loss_mlp": 1.02004695, + "epoch": 0.33932543738351467, + "flos": 17086656349440.0, + "grad_norm": 2.144224167977085, + "language_loss": 0.74542522, + "learning_rate": 3.077645109098471e-06, + "loss": 0.76765788, + "num_input_tokens_seen": 60721540, + "step": 2822, + "time_per_iteration": 3.512735605239868 + }, + { + "auxiliary_loss_clip": 0.01112016, + "auxiliary_loss_mlp": 0.01026284, + "balance_loss_clip": 0.94013989, + "balance_loss_mlp": 1.0180999, + "epoch": 0.3394456802741538, + "flos": 22126970551680.0, + "grad_norm": 1.6711173197880478, + "language_loss": 0.72220504, + "learning_rate": 3.076988806636577e-06, + "loss": 0.74358803, + "num_input_tokens_seen": 60739300, + "step": 2823, + "time_per_iteration": 2.885930299758911 + }, + { + "auxiliary_loss_clip": 0.01154409, + "auxiliary_loss_mlp": 0.00511893, + "balance_loss_clip": 0.98504519, + "balance_loss_mlp": 0.0, + "epoch": 0.3395659231647929, + "flos": 25226887121280.0, + "grad_norm": 1.962353740691291, + "language_loss": 0.89001161, + "learning_rate": 3.0763323407963377e-06, + "loss": 0.90667462, + "num_input_tokens_seen": 60758910, + "step": 2824, + "time_per_iteration": 2.7787039279937744 + }, + { + "auxiliary_loss_clip": 0.01169477, + "auxiliary_loss_mlp": 0.01029702, + "balance_loss_clip": 1.01945376, + "balance_loss_mlp": 1.02174485, + "epoch": 0.33968616605543195, + "flos": 29096477343360.0, + "grad_norm": 1.7676042405798171, + "language_loss": 0.80305147, + "learning_rate": 3.075675711677337e-06, + "loss": 0.82504326, + "num_input_tokens_seen": 60779005, + "step": 2825, + "time_per_iteration": 3.5948214530944824 + }, + { + "auxiliary_loss_clip": 0.01145187, + "auxiliary_loss_mlp": 0.01032277, + "balance_loss_clip": 0.98543251, + "balance_loss_mlp": 1.02355087, + "epoch": 0.33980640894607106, + "flos": 21433966479360.0, + "grad_norm": 2.0359700075531704, + "language_loss": 0.78436315, + "learning_rate": 3.0750189193791865e-06, + "loss": 0.80613774, + "num_input_tokens_seen": 60798590, + "step": 2826, + "time_per_iteration": 3.550309658050537 + }, + { + "auxiliary_loss_clip": 0.01168456, + "auxiliary_loss_mlp": 0.01031608, + "balance_loss_clip": 1.01908422, + "balance_loss_mlp": 1.02296579, + "epoch": 0.33992665183671017, + "flos": 32490035596800.0, + "grad_norm": 1.9267623687828217, + "language_loss": 0.69911003, + "learning_rate": 3.0743619640015203e-06, + "loss": 0.7211107, + "num_input_tokens_seen": 60818840, + "step": 2827, + "time_per_iteration": 2.810943365097046 + }, + { + "auxiliary_loss_clip": 0.01155798, + "auxiliary_loss_mlp": 0.01030702, + "balance_loss_clip": 0.98218095, + "balance_loss_mlp": 1.02165425, + "epoch": 0.3400468947273492, + "flos": 17055414495360.0, + "grad_norm": 6.955394932181298, + "language_loss": 0.92343622, + "learning_rate": 3.073704845643999e-06, + "loss": 0.94530123, + "num_input_tokens_seen": 60835965, + "step": 2828, + "time_per_iteration": 2.697096109390259 + }, + { + "auxiliary_loss_clip": 0.01174781, + "auxiliary_loss_mlp": 0.01031657, + "balance_loss_clip": 1.01931453, + "balance_loss_mlp": 1.02243042, + "epoch": 0.34016713761798834, + "flos": 16872988296960.0, + "grad_norm": 3.9261174543346073, + "language_loss": 0.78011668, + "learning_rate": 3.0730475644063063e-06, + "loss": 0.80218112, + "num_input_tokens_seen": 60851065, + "step": 2829, + "time_per_iteration": 2.565114974975586 + }, + { + "auxiliary_loss_clip": 0.01142731, + "auxiliary_loss_mlp": 0.00510738, + "balance_loss_clip": 0.97944903, + "balance_loss_mlp": 0.0, + "epoch": 0.34028738050862745, + "flos": 21907161273600.0, + "grad_norm": 1.7965548043800892, + "language_loss": 0.64991862, + "learning_rate": 3.072390120388151e-06, + "loss": 0.66645336, + "num_input_tokens_seen": 60869390, + "step": 2830, + "time_per_iteration": 2.7968387603759766 + }, + { + "auxiliary_loss_clip": 0.01172062, + "auxiliary_loss_mlp": 0.01031763, + "balance_loss_clip": 1.02038193, + "balance_loss_mlp": 1.02281094, + "epoch": 0.3404076233992665, + "flos": 22746034477440.0, + "grad_norm": 2.588313695604798, + "language_loss": 0.71763349, + "learning_rate": 3.071732513689267e-06, + "loss": 0.73967177, + "num_input_tokens_seen": 60887925, + "step": 2831, + "time_per_iteration": 3.6495091915130615 + }, + { + "auxiliary_loss_clip": 0.01173636, + "auxiliary_loss_mlp": 0.01035815, + "balance_loss_clip": 1.02355027, + "balance_loss_mlp": 1.02615929, + "epoch": 0.3405278662899056, + "flos": 17052361839360.0, + "grad_norm": 2.2465328983523793, + "language_loss": 0.67494035, + "learning_rate": 3.0710747444094134e-06, + "loss": 0.69703484, + "num_input_tokens_seen": 60905955, + "step": 2832, + "time_per_iteration": 2.6199257373809814 + }, + { + "auxiliary_loss_clip": 0.01150425, + "auxiliary_loss_mlp": 0.01029119, + "balance_loss_clip": 0.9828099, + "balance_loss_mlp": 1.02026165, + "epoch": 0.3406481091805447, + "flos": 42813131783040.0, + "grad_norm": 2.1480172784021185, + "language_loss": 0.64896482, + "learning_rate": 3.070416812648372e-06, + "loss": 0.67076027, + "num_input_tokens_seen": 60929405, + "step": 2833, + "time_per_iteration": 2.8761913776397705 + }, + { + "auxiliary_loss_clip": 0.01121506, + "auxiliary_loss_mlp": 0.01032366, + "balance_loss_clip": 0.93946487, + "balance_loss_mlp": 1.02298999, + "epoch": 0.3407683520711838, + "flos": 26761457917440.0, + "grad_norm": 2.616926322996136, + "language_loss": 0.65483278, + "learning_rate": 3.069758718505951e-06, + "loss": 0.67637151, + "num_input_tokens_seen": 60951145, + "step": 2834, + "time_per_iteration": 2.743941307067871 + }, + { + "auxiliary_loss_clip": 0.01197973, + "auxiliary_loss_mlp": 0.01035674, + "balance_loss_clip": 1.06207204, + "balance_loss_mlp": 1.02687669, + "epoch": 0.3408885949618229, + "flos": 28767643309440.0, + "grad_norm": 1.6540753380531574, + "language_loss": 0.79505491, + "learning_rate": 3.0691004620819836e-06, + "loss": 0.8173914, + "num_input_tokens_seen": 60971275, + "step": 2835, + "time_per_iteration": 2.7275071144104004 + }, + { + "auxiliary_loss_clip": 0.0101205, + "auxiliary_loss_mlp": 0.01001026, + "balance_loss_clip": 0.87925828, + "balance_loss_mlp": 0.99910688, + "epoch": 0.341008837852462, + "flos": 63576252881280.0, + "grad_norm": 0.8064102581438185, + "language_loss": 0.60150194, + "learning_rate": 3.0684420434763254e-06, + "loss": 0.6216327, + "num_input_tokens_seen": 61037460, + "step": 2836, + "time_per_iteration": 3.290886640548706 + }, + { + "auxiliary_loss_clip": 0.01116553, + "auxiliary_loss_mlp": 0.01028991, + "balance_loss_clip": 0.94356108, + "balance_loss_mlp": 1.02010977, + "epoch": 0.34112908074310105, + "flos": 20812173120000.0, + "grad_norm": 10.206769770938985, + "language_loss": 0.76750314, + "learning_rate": 3.06778346278886e-06, + "loss": 0.78895855, + "num_input_tokens_seen": 61056295, + "step": 2837, + "time_per_iteration": 2.704249858856201 + }, + { + "auxiliary_loss_clip": 0.01196979, + "auxiliary_loss_mlp": 0.0102999, + "balance_loss_clip": 1.0611546, + "balance_loss_mlp": 1.02099562, + "epoch": 0.34124932363374016, + "flos": 24976446520320.0, + "grad_norm": 1.880498562583777, + "language_loss": 0.78742671, + "learning_rate": 3.0671247201194906e-06, + "loss": 0.80969644, + "num_input_tokens_seen": 61078430, + "step": 2838, + "time_per_iteration": 2.7376060485839844 + }, + { + "auxiliary_loss_clip": 0.01124977, + "auxiliary_loss_mlp": 0.01029246, + "balance_loss_clip": 0.94241047, + "balance_loss_mlp": 1.01971555, + "epoch": 0.3413695665243792, + "flos": 28402970480640.0, + "grad_norm": 2.6573524818545424, + "language_loss": 0.75206572, + "learning_rate": 3.066465815568151e-06, + "loss": 0.77360797, + "num_input_tokens_seen": 61099260, + "step": 2839, + "time_per_iteration": 2.7664356231689453 + }, + { + "auxiliary_loss_clip": 0.01172291, + "auxiliary_loss_mlp": 0.01029063, + "balance_loss_clip": 1.01883435, + "balance_loss_mlp": 1.02063465, + "epoch": 0.34148980941501833, + "flos": 25302012416640.0, + "grad_norm": 3.837318521688591, + "language_loss": 0.6918906, + "learning_rate": 3.0658067492347947e-06, + "loss": 0.71390414, + "num_input_tokens_seen": 61121900, + "step": 2840, + "time_per_iteration": 2.7384214401245117 + }, + { + "auxiliary_loss_clip": 0.01043594, + "auxiliary_loss_mlp": 0.01031264, + "balance_loss_clip": 0.82386148, + "balance_loss_mlp": 1.02234077, + "epoch": 0.34161005230565744, + "flos": 17530081747200.0, + "grad_norm": 4.022529320555666, + "language_loss": 0.66796768, + "learning_rate": 3.065147521219402e-06, + "loss": 0.68871623, + "num_input_tokens_seen": 61141155, + "step": 2841, + "time_per_iteration": 2.760836362838745 + }, + { + "auxiliary_loss_clip": 0.01139896, + "auxiliary_loss_mlp": 0.01031967, + "balance_loss_clip": 0.98226047, + "balance_loss_mlp": 1.02316904, + "epoch": 0.3417302951962965, + "flos": 43650101566080.0, + "grad_norm": 1.8633991924182112, + "language_loss": 0.74595785, + "learning_rate": 3.064488131621977e-06, + "loss": 0.76767647, + "num_input_tokens_seen": 61164480, + "step": 2842, + "time_per_iteration": 2.873974561691284 + }, + { + "auxiliary_loss_clip": 0.01164986, + "auxiliary_loss_mlp": 0.01039127, + "balance_loss_clip": 1.01911938, + "balance_loss_mlp": 1.02960265, + "epoch": 0.3418505380869356, + "flos": 30882207012480.0, + "grad_norm": 1.8363899327988769, + "language_loss": 0.73734152, + "learning_rate": 3.063828580542549e-06, + "loss": 0.75938272, + "num_input_tokens_seen": 61185675, + "step": 2843, + "time_per_iteration": 2.687135934829712 + }, + { + "auxiliary_loss_clip": 0.01146591, + "auxiliary_loss_mlp": 0.01031017, + "balance_loss_clip": 0.98030126, + "balance_loss_mlp": 1.02223122, + "epoch": 0.3419707809775747, + "flos": 19463871277440.0, + "grad_norm": 1.958104881008478, + "language_loss": 0.73352993, + "learning_rate": 3.0631688680811706e-06, + "loss": 0.75530595, + "num_input_tokens_seen": 61205300, + "step": 2844, + "time_per_iteration": 2.7125790119171143 + }, + { + "auxiliary_loss_clip": 0.01195506, + "auxiliary_loss_mlp": 0.0103353, + "balance_loss_clip": 1.05756068, + "balance_loss_mlp": 1.0251075, + "epoch": 0.3420910238682138, + "flos": 28727818104960.0, + "grad_norm": 2.2519017421074046, + "language_loss": 0.75457752, + "learning_rate": 3.062508994337921e-06, + "loss": 0.77686787, + "num_input_tokens_seen": 61224905, + "step": 2845, + "time_per_iteration": 2.63812518119812 + }, + { + "auxiliary_loss_clip": 0.01170994, + "auxiliary_loss_mlp": 0.01033148, + "balance_loss_clip": 1.01714432, + "balance_loss_mlp": 1.02417731, + "epoch": 0.3422112667588529, + "flos": 21397265758080.0, + "grad_norm": 2.3135282061059073, + "language_loss": 0.7957679, + "learning_rate": 3.0618489594129013e-06, + "loss": 0.81780934, + "num_input_tokens_seen": 61243045, + "step": 2846, + "time_per_iteration": 2.6502468585968018 + }, + { + "auxiliary_loss_clip": 0.01130781, + "auxiliary_loss_mlp": 0.01032306, + "balance_loss_clip": 0.94641483, + "balance_loss_mlp": 1.02300215, + "epoch": 0.342331509649492, + "flos": 13881450038400.0, + "grad_norm": 2.223519707754982, + "language_loss": 0.71437287, + "learning_rate": 3.061188763406239e-06, + "loss": 0.73600376, + "num_input_tokens_seen": 61259190, + "step": 2847, + "time_per_iteration": 2.683342456817627 + }, + { + "auxiliary_loss_clip": 0.01145006, + "auxiliary_loss_mlp": 0.01030246, + "balance_loss_clip": 0.98141849, + "balance_loss_mlp": 1.0210309, + "epoch": 0.34245175254013105, + "flos": 28621450955520.0, + "grad_norm": 4.082125764467943, + "language_loss": 0.81931794, + "learning_rate": 3.060528406418085e-06, + "loss": 0.84107047, + "num_input_tokens_seen": 61279040, + "step": 2848, + "time_per_iteration": 3.653886318206787 + }, + { + "auxiliary_loss_clip": 0.01141953, + "auxiliary_loss_mlp": 0.01030979, + "balance_loss_clip": 0.98099017, + "balance_loss_mlp": 1.02261615, + "epoch": 0.34257199543077016, + "flos": 34127058960000.0, + "grad_norm": 1.950731988140852, + "language_loss": 0.61985993, + "learning_rate": 3.0598678885486145e-06, + "loss": 0.64158922, + "num_input_tokens_seen": 61301580, + "step": 2849, + "time_per_iteration": 2.8452675342559814 + }, + { + "auxiliary_loss_clip": 0.01124908, + "auxiliary_loss_mlp": 0.00511193, + "balance_loss_clip": 0.9406547, + "balance_loss_mlp": 0.0, + "epoch": 0.34269223832140927, + "flos": 19974018188160.0, + "grad_norm": 4.811222561043578, + "language_loss": 0.74133635, + "learning_rate": 3.0592072098980282e-06, + "loss": 0.75769734, + "num_input_tokens_seen": 61321240, + "step": 2850, + "time_per_iteration": 2.676787853240967 + }, + { + "auxiliary_loss_clip": 0.01141823, + "auxiliary_loss_mlp": 0.01035441, + "balance_loss_clip": 0.98010027, + "balance_loss_mlp": 1.02598763, + "epoch": 0.3428124812120483, + "flos": 27235658292480.0, + "grad_norm": 2.283603394277905, + "language_loss": 0.72479582, + "learning_rate": 3.0585463705665514e-06, + "loss": 0.74656844, + "num_input_tokens_seen": 61341615, + "step": 2851, + "time_per_iteration": 3.668896198272705 + }, + { + "auxiliary_loss_clip": 0.0112099, + "auxiliary_loss_mlp": 0.01031727, + "balance_loss_clip": 0.94076538, + "balance_loss_mlp": 1.02209508, + "epoch": 0.34293272410268744, + "flos": 24570871079040.0, + "grad_norm": 1.8780098915555261, + "language_loss": 0.70415449, + "learning_rate": 3.0578853706544304e-06, + "loss": 0.72568172, + "num_input_tokens_seen": 61359005, + "step": 2852, + "time_per_iteration": 3.6715123653411865 + }, + { + "auxiliary_loss_clip": 0.01126906, + "auxiliary_loss_mlp": 0.005111, + "balance_loss_clip": 0.94291294, + "balance_loss_mlp": 0.0, + "epoch": 0.34305296699332655, + "flos": 21506865131520.0, + "grad_norm": 2.3936882919097737, + "language_loss": 0.65997928, + "learning_rate": 3.0572242102619404e-06, + "loss": 0.67635936, + "num_input_tokens_seen": 61376160, + "step": 2853, + "time_per_iteration": 2.7246952056884766 + }, + { + "auxiliary_loss_clip": 0.01147088, + "auxiliary_loss_mlp": 0.01030214, + "balance_loss_clip": 0.98366207, + "balance_loss_mlp": 1.02145791, + "epoch": 0.3431732098839656, + "flos": 24056665931520.0, + "grad_norm": 1.9636929923132402, + "language_loss": 0.80590999, + "learning_rate": 3.0565628894893784e-06, + "loss": 0.82768297, + "num_input_tokens_seen": 61396795, + "step": 2854, + "time_per_iteration": 2.791889190673828 + }, + { + "auxiliary_loss_clip": 0.0116527, + "auxiliary_loss_mlp": 0.01033201, + "balance_loss_clip": 1.01958609, + "balance_loss_mlp": 1.02428985, + "epoch": 0.3432934527746047, + "flos": 16800879744000.0, + "grad_norm": 1.8658530161810556, + "language_loss": 0.74734992, + "learning_rate": 3.0559014084370655e-06, + "loss": 0.76933461, + "num_input_tokens_seen": 61415320, + "step": 2855, + "time_per_iteration": 2.6759567260742188 + }, + { + "auxiliary_loss_clip": 0.01156434, + "auxiliary_loss_mlp": 0.01033573, + "balance_loss_clip": 0.98249602, + "balance_loss_mlp": 1.02477539, + "epoch": 0.34341369566524377, + "flos": 23439720908160.0, + "grad_norm": 2.0738533807080413, + "language_loss": 0.78447938, + "learning_rate": 3.055239767205349e-06, + "loss": 0.80637944, + "num_input_tokens_seen": 61437070, + "step": 2856, + "time_per_iteration": 2.7319719791412354 + }, + { + "auxiliary_loss_clip": 0.01168716, + "auxiliary_loss_mlp": 0.01024787, + "balance_loss_clip": 1.02110004, + "balance_loss_mlp": 1.01621628, + "epoch": 0.3435339385558829, + "flos": 17267466435840.0, + "grad_norm": 2.30900610244694, + "language_loss": 0.78325474, + "learning_rate": 3.054577965894599e-06, + "loss": 0.80518973, + "num_input_tokens_seen": 61453215, + "step": 2857, + "time_per_iteration": 3.5412564277648926 + }, + { + "auxiliary_loss_clip": 0.0115575, + "auxiliary_loss_mlp": 0.01034703, + "balance_loss_clip": 0.98485172, + "balance_loss_mlp": 1.02479625, + "epoch": 0.343654181446522, + "flos": 22199366413440.0, + "grad_norm": 3.3952494138088283, + "language_loss": 0.70514178, + "learning_rate": 3.0539160046052094e-06, + "loss": 0.72704637, + "num_input_tokens_seen": 61472915, + "step": 2858, + "time_per_iteration": 2.686807155609131 + }, + { + "auxiliary_loss_clip": 0.01141649, + "auxiliary_loss_mlp": 0.01030411, + "balance_loss_clip": 0.98032469, + "balance_loss_mlp": 1.02079725, + "epoch": 0.34377442433716104, + "flos": 19901801894400.0, + "grad_norm": 3.081773574006026, + "language_loss": 0.70070469, + "learning_rate": 3.0532538834376003e-06, + "loss": 0.72242528, + "num_input_tokens_seen": 61492475, + "step": 2859, + "time_per_iteration": 2.6887550354003906 + }, + { + "auxiliary_loss_clip": 0.01177447, + "auxiliary_loss_mlp": 0.01036461, + "balance_loss_clip": 1.02151787, + "balance_loss_mlp": 1.02777088, + "epoch": 0.34389466722780015, + "flos": 22197678474240.0, + "grad_norm": 2.210775834898582, + "language_loss": 0.78186613, + "learning_rate": 3.0525916024922143e-06, + "loss": 0.80400521, + "num_input_tokens_seen": 61511660, + "step": 2860, + "time_per_iteration": 2.7045884132385254 + }, + { + "auxiliary_loss_clip": 0.01145232, + "auxiliary_loss_mlp": 0.01026618, + "balance_loss_clip": 0.97949111, + "balance_loss_mlp": 1.01758814, + "epoch": 0.34401491011843927, + "flos": 18624567110400.0, + "grad_norm": 3.4634980259022092, + "language_loss": 0.8399514, + "learning_rate": 3.0519291618695193e-06, + "loss": 0.86166996, + "num_input_tokens_seen": 61529060, + "step": 2861, + "time_per_iteration": 2.6479010581970215 + }, + { + "auxiliary_loss_clip": 0.01115686, + "auxiliary_loss_mlp": 0.01029464, + "balance_loss_clip": 0.93943781, + "balance_loss_mlp": 1.02102399, + "epoch": 0.3441351530090783, + "flos": 17858197509120.0, + "grad_norm": 1.5963698475833847, + "language_loss": 0.75826848, + "learning_rate": 3.0512665616700065e-06, + "loss": 0.77971995, + "num_input_tokens_seen": 61548125, + "step": 2862, + "time_per_iteration": 2.6609764099121094 + }, + { + "auxiliary_loss_clip": 0.01091552, + "auxiliary_loss_mlp": 0.01029931, + "balance_loss_clip": 0.90043455, + "balance_loss_mlp": 1.02131844, + "epoch": 0.34425539589971743, + "flos": 23112754381440.0, + "grad_norm": 2.1316655988202196, + "language_loss": 0.88888848, + "learning_rate": 3.0506038019941933e-06, + "loss": 0.91010332, + "num_input_tokens_seen": 61568135, + "step": 2863, + "time_per_iteration": 2.7741942405700684 + }, + { + "auxiliary_loss_clip": 0.0112256, + "auxiliary_loss_mlp": 0.01033862, + "balance_loss_clip": 0.94312418, + "balance_loss_mlp": 1.02385497, + "epoch": 0.34437563879035654, + "flos": 21907699977600.0, + "grad_norm": 2.8034669608117904, + "language_loss": 0.67063087, + "learning_rate": 3.049940882942617e-06, + "loss": 0.69219518, + "num_input_tokens_seen": 61586920, + "step": 2864, + "time_per_iteration": 2.7218289375305176 + }, + { + "auxiliary_loss_clip": 0.01196206, + "auxiliary_loss_mlp": 0.01037123, + "balance_loss_clip": 1.0584147, + "balance_loss_mlp": 1.02809918, + "epoch": 0.3444958816809956, + "flos": 23076915586560.0, + "grad_norm": 2.286864481846869, + "language_loss": 0.80404502, + "learning_rate": 3.0492778046158448e-06, + "loss": 0.82637829, + "num_input_tokens_seen": 61608340, + "step": 2865, + "time_per_iteration": 2.696945905685425 + }, + { + "auxiliary_loss_clip": 0.0116898, + "auxiliary_loss_mlp": 0.01036444, + "balance_loss_clip": 1.02116144, + "balance_loss_mlp": 1.02770591, + "epoch": 0.3446161245716347, + "flos": 21908633731200.0, + "grad_norm": 2.601137021686163, + "language_loss": 0.76844239, + "learning_rate": 3.0486145671144633e-06, + "loss": 0.79049659, + "num_input_tokens_seen": 61628130, + "step": 2866, + "time_per_iteration": 2.6206161975860596 + }, + { + "auxiliary_loss_clip": 0.01063643, + "auxiliary_loss_mlp": 0.01045392, + "balance_loss_clip": 0.86574626, + "balance_loss_mlp": 1.0362246, + "epoch": 0.3447363674622738, + "flos": 25112834461440.0, + "grad_norm": 2.266586166128152, + "language_loss": 0.77110517, + "learning_rate": 3.047951170539086e-06, + "loss": 0.79219556, + "num_input_tokens_seen": 61647755, + "step": 2867, + "time_per_iteration": 2.734687328338623 + }, + { + "auxiliary_loss_clip": 0.01118994, + "auxiliary_loss_mlp": 0.01031972, + "balance_loss_clip": 0.94374037, + "balance_loss_mlp": 1.02365744, + "epoch": 0.3448566103529129, + "flos": 11984684451840.0, + "grad_norm": 3.13785354622982, + "language_loss": 0.83719164, + "learning_rate": 3.047287614990349e-06, + "loss": 0.85870129, + "num_input_tokens_seen": 61665675, + "step": 2868, + "time_per_iteration": 2.595043659210205 + }, + { + "auxiliary_loss_clip": 0.01140606, + "auxiliary_loss_mlp": 0.01034633, + "balance_loss_clip": 0.98092079, + "balance_loss_mlp": 1.02560937, + "epoch": 0.344976853243552, + "flos": 40187882465280.0, + "grad_norm": 2.782508329104387, + "language_loss": 0.61543357, + "learning_rate": 3.046623900568914e-06, + "loss": 0.63718599, + "num_input_tokens_seen": 61688240, + "step": 2869, + "time_per_iteration": 2.7239127159118652 + }, + { + "auxiliary_loss_clip": 0.01143681, + "auxiliary_loss_mlp": 0.010394, + "balance_loss_clip": 0.98082972, + "balance_loss_mlp": 1.02995229, + "epoch": 0.34509709613419104, + "flos": 28723652127360.0, + "grad_norm": 5.126711510052835, + "language_loss": 0.70613796, + "learning_rate": 3.045960027375465e-06, + "loss": 0.72796881, + "num_input_tokens_seen": 61706075, + "step": 2870, + "time_per_iteration": 2.61306095123291 + }, + { + "auxiliary_loss_clip": 0.01174112, + "auxiliary_loss_mlp": 0.01036991, + "balance_loss_clip": 1.01846981, + "balance_loss_mlp": 1.0276866, + "epoch": 0.34521733902483015, + "flos": 29967597982080.0, + "grad_norm": 5.835128383218566, + "language_loss": 0.82616568, + "learning_rate": 3.045295995510711e-06, + "loss": 0.84827667, + "num_input_tokens_seen": 61723045, + "step": 2871, + "time_per_iteration": 2.7477283477783203 + }, + { + "auxiliary_loss_clip": 0.01144039, + "auxiliary_loss_mlp": 0.01025209, + "balance_loss_clip": 0.98163557, + "balance_loss_mlp": 1.01700735, + "epoch": 0.34533758191546926, + "flos": 27923059843200.0, + "grad_norm": 2.9718920902758095, + "language_loss": 0.73496157, + "learning_rate": 3.0446318050753865e-06, + "loss": 0.75665402, + "num_input_tokens_seen": 61743525, + "step": 2872, + "time_per_iteration": 2.7328836917877197 + }, + { + "auxiliary_loss_clip": 0.01165643, + "auxiliary_loss_mlp": 0.01031022, + "balance_loss_clip": 1.0192939, + "balance_loss_mlp": 1.02261174, + "epoch": 0.3454578248061083, + "flos": 27125879351040.0, + "grad_norm": 2.1474249139734596, + "language_loss": 0.77458864, + "learning_rate": 3.0439674561702474e-06, + "loss": 0.79655528, + "num_input_tokens_seen": 61763025, + "step": 2873, + "time_per_iteration": 2.6749699115753174 + }, + { + "auxiliary_loss_clip": 0.01170798, + "auxiliary_loss_mlp": 0.01031332, + "balance_loss_clip": 1.0222888, + "balance_loss_mlp": 1.02239704, + "epoch": 0.3455780676967474, + "flos": 19024899166080.0, + "grad_norm": 5.140238652759668, + "language_loss": 0.87717175, + "learning_rate": 3.043302948896076e-06, + "loss": 0.89919305, + "num_input_tokens_seen": 61781630, + "step": 2874, + "time_per_iteration": 3.5965089797973633 + }, + { + "auxiliary_loss_clip": 0.01090152, + "auxiliary_loss_mlp": 0.01030233, + "balance_loss_clip": 0.90364349, + "balance_loss_mlp": 1.02142918, + "epoch": 0.34569831058738654, + "flos": 34496005507200.0, + "grad_norm": 2.9074720266575165, + "language_loss": 0.60559404, + "learning_rate": 3.0426382833536756e-06, + "loss": 0.62679785, + "num_input_tokens_seen": 61804985, + "step": 2875, + "time_per_iteration": 2.794696569442749 + }, + { + "auxiliary_loss_clip": 0.01118925, + "auxiliary_loss_mlp": 0.01027502, + "balance_loss_clip": 0.94066852, + "balance_loss_mlp": 1.01856184, + "epoch": 0.3458185534780256, + "flos": 31138681098240.0, + "grad_norm": 2.5978286146790954, + "language_loss": 0.77790511, + "learning_rate": 3.041973459643877e-06, + "loss": 0.79936939, + "num_input_tokens_seen": 61824440, + "step": 2876, + "time_per_iteration": 2.8481674194335938 + }, + { + "auxiliary_loss_clip": 0.01093665, + "auxiliary_loss_mlp": 0.01029476, + "balance_loss_clip": 0.90152705, + "balance_loss_mlp": 1.02051115, + "epoch": 0.3459387963686647, + "flos": 32452508862720.0, + "grad_norm": 2.7362649355629274, + "language_loss": 0.66182518, + "learning_rate": 3.0413084778675334e-06, + "loss": 0.68305659, + "num_input_tokens_seen": 61845690, + "step": 2877, + "time_per_iteration": 4.72149133682251 + }, + { + "auxiliary_loss_clip": 0.01142161, + "auxiliary_loss_mlp": 0.00510592, + "balance_loss_clip": 0.97985238, + "balance_loss_mlp": 0.0, + "epoch": 0.3460590392593038, + "flos": 24675658030080.0, + "grad_norm": 3.841810120513787, + "language_loss": 0.8412903, + "learning_rate": 3.0406433381255214e-06, + "loss": 0.85781783, + "num_input_tokens_seen": 61863725, + "step": 2878, + "time_per_iteration": 2.689251184463501 + }, + { + "auxiliary_loss_clip": 0.01172612, + "auxiliary_loss_mlp": 0.01028645, + "balance_loss_clip": 1.02541399, + "balance_loss_mlp": 1.01973987, + "epoch": 0.34617928214994287, + "flos": 18807316531200.0, + "grad_norm": 2.635817035402164, + "language_loss": 0.8180936, + "learning_rate": 3.0399780405187425e-06, + "loss": 0.84010613, + "num_input_tokens_seen": 61882720, + "step": 2879, + "time_per_iteration": 2.638293504714966 + }, + { + "auxiliary_loss_clip": 0.01164927, + "auxiliary_loss_mlp": 0.01028276, + "balance_loss_clip": 1.01835215, + "balance_loss_mlp": 1.02050388, + "epoch": 0.346299525040582, + "flos": 24857653265280.0, + "grad_norm": 2.4820696413290144, + "language_loss": 0.78444076, + "learning_rate": 3.0393125851481216e-06, + "loss": 0.80637276, + "num_input_tokens_seen": 61902595, + "step": 2880, + "time_per_iteration": 2.7009174823760986 + }, + { + "auxiliary_loss_clip": 0.01118392, + "auxiliary_loss_mlp": 0.01027087, + "balance_loss_clip": 0.94363761, + "balance_loss_mlp": 1.01824152, + "epoch": 0.3464197679312211, + "flos": 16434914025600.0, + "grad_norm": 2.4292978920309327, + "language_loss": 0.86419147, + "learning_rate": 3.038646972114608e-06, + "loss": 0.88564622, + "num_input_tokens_seen": 61918920, + "step": 2881, + "time_per_iteration": 2.6565792560577393 + }, + { + "auxiliary_loss_clip": 0.01123055, + "auxiliary_loss_mlp": 0.01031191, + "balance_loss_clip": 0.94727778, + "balance_loss_mlp": 1.02283168, + "epoch": 0.34654001082186014, + "flos": 22382474970240.0, + "grad_norm": 1.7785024269788672, + "language_loss": 0.67434895, + "learning_rate": 3.037981201519174e-06, + "loss": 0.69589138, + "num_input_tokens_seen": 61939520, + "step": 2882, + "time_per_iteration": 2.820141077041626 + }, + { + "auxiliary_loss_clip": 0.0117344, + "auxiliary_loss_mlp": 0.01037533, + "balance_loss_clip": 1.02410018, + "balance_loss_mlp": 1.02915895, + "epoch": 0.34666025371249926, + "flos": 19573901614080.0, + "grad_norm": 2.445608851708481, + "language_loss": 0.71360338, + "learning_rate": 3.0373152734628175e-06, + "loss": 0.73571312, + "num_input_tokens_seen": 61957800, + "step": 2883, + "time_per_iteration": 3.5388782024383545 + }, + { + "auxiliary_loss_clip": 0.01163803, + "auxiliary_loss_mlp": 0.01031007, + "balance_loss_clip": 1.01786554, + "balance_loss_mlp": 1.02244782, + "epoch": 0.34678049660313837, + "flos": 15267637751040.0, + "grad_norm": 1.9593928801056866, + "language_loss": 0.7601794, + "learning_rate": 3.0366491880465584e-06, + "loss": 0.7821275, + "num_input_tokens_seen": 61975820, + "step": 2884, + "time_per_iteration": 2.6865925788879395 + }, + { + "auxiliary_loss_clip": 0.0119986, + "auxiliary_loss_mlp": 0.0103519, + "balance_loss_clip": 1.06387103, + "balance_loss_mlp": 1.02604032, + "epoch": 0.3469007394937774, + "flos": 21181550630400.0, + "grad_norm": 1.7650130483487803, + "language_loss": 0.82196462, + "learning_rate": 3.035982945371443e-06, + "loss": 0.84431517, + "num_input_tokens_seen": 61997515, + "step": 2885, + "time_per_iteration": 2.6529836654663086 + }, + { + "auxiliary_loss_clip": 0.01157104, + "auxiliary_loss_mlp": 0.01035304, + "balance_loss_clip": 0.98471546, + "balance_loss_mlp": 1.02707243, + "epoch": 0.34702098238441653, + "flos": 22375471818240.0, + "grad_norm": 2.49063127119276, + "language_loss": 0.85769176, + "learning_rate": 3.035316545538537e-06, + "loss": 0.87961578, + "num_input_tokens_seen": 62016310, + "step": 2886, + "time_per_iteration": 2.679356098175049 + }, + { + "auxiliary_loss_clip": 0.01144202, + "auxiliary_loss_mlp": 0.01032647, + "balance_loss_clip": 0.985039, + "balance_loss_mlp": 1.02423692, + "epoch": 0.3471412252750556, + "flos": 22929430343040.0, + "grad_norm": 1.9319911980972273, + "language_loss": 0.79152101, + "learning_rate": 3.034649988648935e-06, + "loss": 0.81328952, + "num_input_tokens_seen": 62036075, + "step": 2887, + "time_per_iteration": 2.716952323913574 + }, + { + "auxiliary_loss_clip": 0.01146998, + "auxiliary_loss_mlp": 0.01029563, + "balance_loss_clip": 0.98089826, + "balance_loss_mlp": 1.02150416, + "epoch": 0.3472614681656947, + "flos": 21324259365120.0, + "grad_norm": 1.930846506558772, + "language_loss": 0.80194485, + "learning_rate": 3.033983274803752e-06, + "loss": 0.82371044, + "num_input_tokens_seen": 62055865, + "step": 2888, + "time_per_iteration": 2.689847946166992 + }, + { + "auxiliary_loss_clip": 0.01142321, + "auxiliary_loss_mlp": 0.01031096, + "balance_loss_clip": 0.98034388, + "balance_loss_mlp": 1.02222633, + "epoch": 0.3473817110563338, + "flos": 23475739271040.0, + "grad_norm": 2.9653460644666083, + "language_loss": 0.7198016, + "learning_rate": 3.0333164041041283e-06, + "loss": 0.74153578, + "num_input_tokens_seen": 62072180, + "step": 2889, + "time_per_iteration": 2.7154788970947266 + }, + { + "auxiliary_loss_clip": 0.01075451, + "auxiliary_loss_mlp": 0.01026615, + "balance_loss_clip": 0.86606383, + "balance_loss_mlp": 1.01815152, + "epoch": 0.34750195394697286, + "flos": 22346025644160.0, + "grad_norm": 2.094086433122474, + "language_loss": 0.72129774, + "learning_rate": 3.032649376651228e-06, + "loss": 0.74231839, + "num_input_tokens_seen": 62091600, + "step": 2890, + "time_per_iteration": 2.8190231323242188 + }, + { + "auxiliary_loss_clip": 0.01119842, + "auxiliary_loss_mlp": 0.0103322, + "balance_loss_clip": 0.94057608, + "balance_loss_mlp": 1.02464831, + "epoch": 0.347622196837612, + "flos": 29095004885760.0, + "grad_norm": 1.9991905603819586, + "language_loss": 0.75457883, + "learning_rate": 3.031982192546238e-06, + "loss": 0.77610934, + "num_input_tokens_seen": 62114695, + "step": 2891, + "time_per_iteration": 2.787747383117676 + }, + { + "auxiliary_loss_clip": 0.01170512, + "auxiliary_loss_mlp": 0.01037572, + "balance_loss_clip": 1.01947212, + "balance_loss_mlp": 1.02910829, + "epoch": 0.3477424397282511, + "flos": 22455732758400.0, + "grad_norm": 2.7516243632848143, + "language_loss": 0.94842577, + "learning_rate": 3.0313148518903696e-06, + "loss": 0.97050661, + "num_input_tokens_seen": 62134520, + "step": 2892, + "time_per_iteration": 2.6474857330322266 + }, + { + "auxiliary_loss_clip": 0.01147271, + "auxiliary_loss_mlp": 0.01034755, + "balance_loss_clip": 0.98268712, + "balance_loss_mlp": 1.02686381, + "epoch": 0.34786268261889014, + "flos": 15778790242560.0, + "grad_norm": 2.6134921148669865, + "language_loss": 0.81339359, + "learning_rate": 3.030647354784859e-06, + "loss": 0.83521384, + "num_input_tokens_seen": 62151560, + "step": 2893, + "time_per_iteration": 2.7507903575897217 + }, + { + "auxiliary_loss_clip": 0.01117255, + "auxiliary_loss_mlp": 0.01026866, + "balance_loss_clip": 0.94134057, + "balance_loss_mlp": 1.01878357, + "epoch": 0.34798292550952925, + "flos": 20777627214720.0, + "grad_norm": 2.5037160811550825, + "language_loss": 0.77598894, + "learning_rate": 3.029979701330964e-06, + "loss": 0.79743016, + "num_input_tokens_seen": 62170985, + "step": 2894, + "time_per_iteration": 2.7119646072387695 + }, + { + "auxiliary_loss_clip": 0.01153674, + "auxiliary_loss_mlp": 0.01030512, + "balance_loss_clip": 0.98372126, + "balance_loss_mlp": 1.0217979, + "epoch": 0.34810316840016836, + "flos": 19937820257280.0, + "grad_norm": 2.400774583062165, + "language_loss": 0.80137181, + "learning_rate": 3.029311891629966e-06, + "loss": 0.8232137, + "num_input_tokens_seen": 62189440, + "step": 2895, + "time_per_iteration": 2.70636248588562 + }, + { + "auxiliary_loss_clip": 0.01144039, + "auxiliary_loss_mlp": 0.01029992, + "balance_loss_clip": 0.98256403, + "balance_loss_mlp": 1.02154553, + "epoch": 0.3482234112908074, + "flos": 23623296341760.0, + "grad_norm": 10.502372722583898, + "language_loss": 0.74414349, + "learning_rate": 3.0286439257831744e-06, + "loss": 0.7658838, + "num_input_tokens_seen": 62208910, + "step": 2896, + "time_per_iteration": 2.7167389392852783 + }, + { + "auxiliary_loss_clip": 0.01196342, + "auxiliary_loss_mlp": 0.01034597, + "balance_loss_clip": 1.05906725, + "balance_loss_mlp": 1.02479839, + "epoch": 0.3483436541814465, + "flos": 23986712194560.0, + "grad_norm": 2.441090959602217, + "language_loss": 0.71736956, + "learning_rate": 3.0279758038919156e-06, + "loss": 0.73967892, + "num_input_tokens_seen": 62227135, + "step": 2897, + "time_per_iteration": 2.718156099319458 + }, + { + "auxiliary_loss_clip": 0.01168994, + "auxiliary_loss_mlp": 0.0102654, + "balance_loss_clip": 1.02064908, + "balance_loss_mlp": 1.01816535, + "epoch": 0.34846389707208564, + "flos": 22638338524800.0, + "grad_norm": 3.075241096968893, + "language_loss": 0.7854315, + "learning_rate": 3.0273075260575455e-06, + "loss": 0.80738688, + "num_input_tokens_seen": 62246035, + "step": 2898, + "time_per_iteration": 2.6720640659332275 + }, + { + "auxiliary_loss_clip": 0.01148056, + "auxiliary_loss_mlp": 0.01028755, + "balance_loss_clip": 0.98113221, + "balance_loss_mlp": 1.01984417, + "epoch": 0.3485841399627247, + "flos": 21792857218560.0, + "grad_norm": 1.788827334713969, + "language_loss": 0.80863613, + "learning_rate": 3.0266390923814396e-06, + "loss": 0.83040422, + "num_input_tokens_seen": 62264095, + "step": 2899, + "time_per_iteration": 2.6904375553131104 + }, + { + "auxiliary_loss_clip": 0.01150201, + "auxiliary_loss_mlp": 0.01032941, + "balance_loss_clip": 0.98592794, + "balance_loss_mlp": 1.02430439, + "epoch": 0.3487043828533638, + "flos": 17019036996480.0, + "grad_norm": 1.8425268752033586, + "language_loss": 0.8200022, + "learning_rate": 3.0259705029650008e-06, + "loss": 0.84183359, + "num_input_tokens_seen": 62282025, + "step": 2900, + "time_per_iteration": 3.6458489894866943 + }, + { + "auxiliary_loss_clip": 0.01171966, + "auxiliary_loss_mlp": 0.01031112, + "balance_loss_clip": 1.02058959, + "balance_loss_mlp": 1.0225054, + "epoch": 0.34882462574400286, + "flos": 22601135013120.0, + "grad_norm": 1.6511742796553714, + "language_loss": 0.72763824, + "learning_rate": 3.025301757909652e-06, + "loss": 0.74966908, + "num_input_tokens_seen": 62302220, + "step": 2901, + "time_per_iteration": 2.6019341945648193 + }, + { + "auxiliary_loss_clip": 0.01126722, + "auxiliary_loss_mlp": 0.00510735, + "balance_loss_clip": 0.94488621, + "balance_loss_mlp": 0.0, + "epoch": 0.34894486863464197, + "flos": 29861518141440.0, + "grad_norm": 1.8591756504367587, + "language_loss": 0.80481929, + "learning_rate": 3.024632857316842e-06, + "loss": 0.82119381, + "num_input_tokens_seen": 62323535, + "step": 2902, + "time_per_iteration": 2.7685744762420654 + }, + { + "auxiliary_loss_clip": 0.01170756, + "auxiliary_loss_mlp": 0.01028849, + "balance_loss_clip": 1.02145338, + "balance_loss_mlp": 1.02055812, + "epoch": 0.3490651115252811, + "flos": 22122265870080.0, + "grad_norm": 2.651465128557675, + "language_loss": 0.77398723, + "learning_rate": 3.0239638012880412e-06, + "loss": 0.79598325, + "num_input_tokens_seen": 62343430, + "step": 2903, + "time_per_iteration": 3.613083600997925 + }, + { + "auxiliary_loss_clip": 0.01090964, + "auxiliary_loss_mlp": 0.01033803, + "balance_loss_clip": 0.90188795, + "balance_loss_mlp": 1.0247252, + "epoch": 0.34918535441592014, + "flos": 12676682943360.0, + "grad_norm": 2.704114900585967, + "language_loss": 0.81755579, + "learning_rate": 3.0232945899247466e-06, + "loss": 0.83880341, + "num_input_tokens_seen": 62360365, + "step": 2904, + "time_per_iteration": 2.708237409591675 + }, + { + "auxiliary_loss_clip": 0.01169387, + "auxiliary_loss_mlp": 0.01029794, + "balance_loss_clip": 1.01803768, + "balance_loss_mlp": 1.02091861, + "epoch": 0.34930559730655925, + "flos": 23185617120000.0, + "grad_norm": 2.1083362697423125, + "language_loss": 0.77281296, + "learning_rate": 3.022625223328476e-06, + "loss": 0.79480469, + "num_input_tokens_seen": 62382105, + "step": 2905, + "time_per_iteration": 2.7028167247772217 + }, + { + "auxiliary_loss_clip": 0.01175286, + "auxiliary_loss_mlp": 0.01027046, + "balance_loss_clip": 1.02122939, + "balance_loss_mlp": 1.01845121, + "epoch": 0.34942584019719836, + "flos": 22855023319680.0, + "grad_norm": 1.5947174892504759, + "language_loss": 0.68930584, + "learning_rate": 3.0219557016007723e-06, + "loss": 0.71132922, + "num_input_tokens_seen": 62402235, + "step": 2906, + "time_per_iteration": 2.653859853744507 + }, + { + "auxiliary_loss_clip": 0.01166397, + "auxiliary_loss_mlp": 0.01026707, + "balance_loss_clip": 1.02150893, + "balance_loss_mlp": 1.0173018, + "epoch": 0.3495460830878374, + "flos": 24426043441920.0, + "grad_norm": 2.0821582107345473, + "language_loss": 0.70010102, + "learning_rate": 3.021286024843202e-06, + "loss": 0.72203207, + "num_input_tokens_seen": 62420430, + "step": 2907, + "time_per_iteration": 2.679858684539795 + }, + { + "auxiliary_loss_clip": 0.01101247, + "auxiliary_loss_mlp": 0.01013555, + "balance_loss_clip": 1.03356051, + "balance_loss_mlp": 1.0118264, + "epoch": 0.3496663259784765, + "flos": 70008749389440.0, + "grad_norm": 1.0759822520503612, + "language_loss": 0.64709169, + "learning_rate": 3.0206161931573526e-06, + "loss": 0.66823971, + "num_input_tokens_seen": 62472980, + "step": 2908, + "time_per_iteration": 3.1052286624908447 + }, + { + "auxiliary_loss_clip": 0.01143514, + "auxiliary_loss_mlp": 0.01026533, + "balance_loss_clip": 0.97925663, + "balance_loss_mlp": 1.0175271, + "epoch": 0.34978656886911563, + "flos": 28692805322880.0, + "grad_norm": 3.2407190923886566, + "language_loss": 0.93043047, + "learning_rate": 3.0199462066448388e-06, + "loss": 0.95213091, + "num_input_tokens_seen": 62495175, + "step": 2909, + "time_per_iteration": 3.704749584197998 + }, + { + "auxiliary_loss_clip": 0.01171452, + "auxiliary_loss_mlp": 0.01026039, + "balance_loss_clip": 1.02169311, + "balance_loss_mlp": 1.01747954, + "epoch": 0.3499068117597547, + "flos": 21142156389120.0, + "grad_norm": 1.9039814589274178, + "language_loss": 0.69215894, + "learning_rate": 3.019276065407296e-06, + "loss": 0.71413386, + "num_input_tokens_seen": 62514295, + "step": 2910, + "time_per_iteration": 2.6866891384124756 + }, + { + "auxiliary_loss_clip": 0.01098305, + "auxiliary_loss_mlp": 0.01035963, + "balance_loss_clip": 0.9032824, + "balance_loss_mlp": 1.02714777, + "epoch": 0.3500270546503938, + "flos": 22782699285120.0, + "grad_norm": 1.7025563183902104, + "language_loss": 0.8061862, + "learning_rate": 3.018605769546385e-06, + "loss": 0.82752889, + "num_input_tokens_seen": 62534850, + "step": 2911, + "time_per_iteration": 2.81760311126709 + }, + { + "auxiliary_loss_clip": 0.01167726, + "auxiliary_loss_mlp": 0.01028788, + "balance_loss_clip": 1.01857674, + "balance_loss_mlp": 1.02029467, + "epoch": 0.3501472975410329, + "flos": 22894058424960.0, + "grad_norm": 1.6778151415122406, + "language_loss": 0.79767197, + "learning_rate": 3.017935319163788e-06, + "loss": 0.81963706, + "num_input_tokens_seen": 62553810, + "step": 2912, + "time_per_iteration": 2.6946229934692383 + }, + { + "auxiliary_loss_clip": 0.0117034, + "auxiliary_loss_mlp": 0.01029675, + "balance_loss_clip": 1.01926112, + "balance_loss_mlp": 1.02097893, + "epoch": 0.35026754043167196, + "flos": 25446588658560.0, + "grad_norm": 1.9148177136661382, + "language_loss": 0.71026123, + "learning_rate": 3.017264714361213e-06, + "loss": 0.73226142, + "num_input_tokens_seen": 62573460, + "step": 2913, + "time_per_iteration": 2.6867735385894775 + }, + { + "auxiliary_loss_clip": 0.01146142, + "auxiliary_loss_mlp": 0.00511528, + "balance_loss_clip": 0.98165381, + "balance_loss_mlp": 0.0, + "epoch": 0.3503877833223111, + "flos": 19573757959680.0, + "grad_norm": 2.5837007349516727, + "language_loss": 0.82428479, + "learning_rate": 3.016593955240389e-06, + "loss": 0.8408615, + "num_input_tokens_seen": 62592150, + "step": 2914, + "time_per_iteration": 2.7224998474121094 + }, + { + "auxiliary_loss_clip": 0.01076378, + "auxiliary_loss_mlp": 0.01000706, + "balance_loss_clip": 0.99265802, + "balance_loss_mlp": 0.99890625, + "epoch": 0.3505080262129502, + "flos": 65072075880960.0, + "grad_norm": 0.8209857180607216, + "language_loss": 0.63732719, + "learning_rate": 3.015923041903071e-06, + "loss": 0.6580981, + "num_input_tokens_seen": 62658275, + "step": 2915, + "time_per_iteration": 3.2502870559692383 + }, + { + "auxiliary_loss_clip": 0.01169435, + "auxiliary_loss_mlp": 0.01025632, + "balance_loss_clip": 1.02129698, + "balance_loss_mlp": 1.01755595, + "epoch": 0.35062826910358924, + "flos": 29314562768640.0, + "grad_norm": 1.9828905807588262, + "language_loss": 0.8368752, + "learning_rate": 3.0152519744510347e-06, + "loss": 0.8588258, + "num_input_tokens_seen": 62678075, + "step": 2916, + "time_per_iteration": 2.7735488414764404 + }, + { + "auxiliary_loss_clip": 0.01123551, + "auxiliary_loss_mlp": 0.01029282, + "balance_loss_clip": 0.94218343, + "balance_loss_mlp": 1.02068138, + "epoch": 0.35074851199422835, + "flos": 23987717775360.0, + "grad_norm": 2.208602576801514, + "language_loss": 0.82946277, + "learning_rate": 3.014580752986081e-06, + "loss": 0.85099113, + "num_input_tokens_seen": 62696950, + "step": 2917, + "time_per_iteration": 2.749913215637207 + }, + { + "auxiliary_loss_clip": 0.01099925, + "auxiliary_loss_mlp": 0.0103099, + "balance_loss_clip": 0.90459192, + "balance_loss_mlp": 1.02257943, + "epoch": 0.3508687548848674, + "flos": 15224436668160.0, + "grad_norm": 2.0320605605076376, + "language_loss": 0.78509331, + "learning_rate": 3.0139093776100345e-06, + "loss": 0.80640244, + "num_input_tokens_seen": 62713540, + "step": 2918, + "time_per_iteration": 2.7536911964416504 + }, + { + "auxiliary_loss_clip": 0.01192669, + "auxiliary_loss_mlp": 0.01024029, + "balance_loss_clip": 1.05945444, + "balance_loss_mlp": 1.0158217, + "epoch": 0.3509889977755065, + "flos": 21361750185600.0, + "grad_norm": 2.3638671339445065, + "language_loss": 0.75363284, + "learning_rate": 3.013237848424741e-06, + "loss": 0.77579981, + "num_input_tokens_seen": 62732925, + "step": 2919, + "time_per_iteration": 2.6543350219726562 + }, + { + "auxiliary_loss_clip": 0.01147787, + "auxiliary_loss_mlp": 0.01033646, + "balance_loss_clip": 0.9831171, + "balance_loss_mlp": 1.02475286, + "epoch": 0.35110924066614563, + "flos": 19135360465920.0, + "grad_norm": 2.0356242244742284, + "language_loss": 0.75333965, + "learning_rate": 3.012566165532072e-06, + "loss": 0.77515388, + "num_input_tokens_seen": 62751715, + "step": 2920, + "time_per_iteration": 2.6207356452941895 + }, + { + "auxiliary_loss_clip": 0.01079176, + "auxiliary_loss_mlp": 0.01031593, + "balance_loss_clip": 0.86533463, + "balance_loss_mlp": 1.02345133, + "epoch": 0.3512294835567847, + "flos": 21980885938560.0, + "grad_norm": 2.0481868089538713, + "language_loss": 0.76607931, + "learning_rate": 3.0118943290339207e-06, + "loss": 0.78718698, + "num_input_tokens_seen": 62771925, + "step": 2921, + "time_per_iteration": 2.76350736618042 + }, + { + "auxiliary_loss_clip": 0.01110289, + "auxiliary_loss_mlp": 0.01028293, + "balance_loss_clip": 0.93815625, + "balance_loss_mlp": 1.01921463, + "epoch": 0.3513497264474238, + "flos": 17817294896640.0, + "grad_norm": 2.846075363308693, + "language_loss": 0.67807221, + "learning_rate": 3.011222339032204e-06, + "loss": 0.699458, + "num_input_tokens_seen": 62790075, + "step": 2922, + "time_per_iteration": 2.65291690826416 + }, + { + "auxiliary_loss_clip": 0.01194776, + "auxiliary_loss_mlp": 0.0103015, + "balance_loss_clip": 1.06105137, + "balance_loss_mlp": 1.02200198, + "epoch": 0.3514699693380629, + "flos": 26943417239040.0, + "grad_norm": 1.9219187692975088, + "language_loss": 0.69756263, + "learning_rate": 3.0105501956288626e-06, + "loss": 0.71981192, + "num_input_tokens_seen": 62810545, + "step": 2923, + "time_per_iteration": 2.7337000370025635 + }, + { + "auxiliary_loss_clip": 0.01178464, + "auxiliary_loss_mlp": 0.01030961, + "balance_loss_clip": 1.0227766, + "balance_loss_mlp": 1.02187681, + "epoch": 0.35159021222870196, + "flos": 15267565923840.0, + "grad_norm": 14.758928655580075, + "language_loss": 0.73046869, + "learning_rate": 3.0098778989258602e-06, + "loss": 0.752563, + "num_input_tokens_seen": 62829155, + "step": 2924, + "time_per_iteration": 2.6045947074890137 + }, + { + "auxiliary_loss_clip": 0.0112251, + "auxiliary_loss_mlp": 0.01035033, + "balance_loss_clip": 0.94723392, + "balance_loss_mlp": 1.02614617, + "epoch": 0.35171045511934107, + "flos": 13984154000640.0, + "grad_norm": 1.9246861096363734, + "language_loss": 0.87898409, + "learning_rate": 3.009205449025183e-06, + "loss": 0.90055954, + "num_input_tokens_seen": 62845350, + "step": 2925, + "time_per_iteration": 2.7870922088623047 + }, + { + "auxiliary_loss_clip": 0.01117689, + "auxiliary_loss_mlp": 0.01036747, + "balance_loss_clip": 0.94031018, + "balance_loss_mlp": 1.02807403, + "epoch": 0.3518306980099802, + "flos": 14283434119680.0, + "grad_norm": 2.1226759829755317, + "language_loss": 0.63253695, + "learning_rate": 3.008532846028842e-06, + "loss": 0.65408134, + "num_input_tokens_seen": 62862110, + "step": 2926, + "time_per_iteration": 3.686084032058716 + }, + { + "auxiliary_loss_clip": 0.01198472, + "auxiliary_loss_mlp": 0.01034922, + "balance_loss_clip": 1.0625186, + "balance_loss_mlp": 1.02551675, + "epoch": 0.35195094090061924, + "flos": 27052872958080.0, + "grad_norm": 2.3476699495774356, + "language_loss": 0.71589828, + "learning_rate": 3.0078600900388694e-06, + "loss": 0.73823225, + "num_input_tokens_seen": 62882415, + "step": 2927, + "time_per_iteration": 2.751405715942383 + }, + { + "auxiliary_loss_clip": 0.01113619, + "auxiliary_loss_mlp": 0.01029218, + "balance_loss_clip": 0.94005072, + "balance_loss_mlp": 1.02041459, + "epoch": 0.35207118379125835, + "flos": 25629266252160.0, + "grad_norm": 1.8933166705844846, + "language_loss": 0.73891473, + "learning_rate": 3.007187181157323e-06, + "loss": 0.76034307, + "num_input_tokens_seen": 62902425, + "step": 2928, + "time_per_iteration": 2.7471413612365723 + }, + { + "auxiliary_loss_clip": 0.01070793, + "auxiliary_loss_mlp": 0.01031937, + "balance_loss_clip": 0.86735111, + "balance_loss_mlp": 1.02296627, + "epoch": 0.35219142668189746, + "flos": 18004713085440.0, + "grad_norm": 4.09046051587973, + "language_loss": 0.68293917, + "learning_rate": 3.006514119486282e-06, + "loss": 0.7039665, + "num_input_tokens_seen": 62919255, + "step": 2929, + "time_per_iteration": 3.6677777767181396 + }, + { + "auxiliary_loss_clip": 0.01116518, + "auxiliary_loss_mlp": 0.01028959, + "balance_loss_clip": 0.94377851, + "balance_loss_mlp": 1.01994729, + "epoch": 0.3523116695725365, + "flos": 14028109269120.0, + "grad_norm": 1.7960489448901944, + "language_loss": 0.69806123, + "learning_rate": 3.005840905127849e-06, + "loss": 0.71951598, + "num_input_tokens_seen": 62936160, + "step": 2930, + "time_per_iteration": 2.68224835395813 + }, + { + "auxiliary_loss_clip": 0.01195184, + "auxiliary_loss_mlp": 0.01034315, + "balance_loss_clip": 1.06154656, + "balance_loss_mlp": 1.02510571, + "epoch": 0.3524319124631756, + "flos": 21433966479360.0, + "grad_norm": 2.1650863035334695, + "language_loss": 0.86957157, + "learning_rate": 3.0051675381841516e-06, + "loss": 0.89186656, + "num_input_tokens_seen": 62953470, + "step": 2931, + "time_per_iteration": 2.580113649368286 + }, + { + "auxiliary_loss_clip": 0.01056238, + "auxiliary_loss_mlp": 0.00511333, + "balance_loss_clip": 0.82930994, + "balance_loss_mlp": 0.0, + "epoch": 0.3525521553538147, + "flos": 26322773114880.0, + "grad_norm": 1.5867947039821497, + "language_loss": 0.76806515, + "learning_rate": 3.0044940187573363e-06, + "loss": 0.78374082, + "num_input_tokens_seen": 62974480, + "step": 2932, + "time_per_iteration": 2.852957010269165 + }, + { + "auxiliary_loss_clip": 0.01173115, + "auxiliary_loss_mlp": 0.01035469, + "balance_loss_clip": 1.02156401, + "balance_loss_mlp": 1.02655172, + "epoch": 0.3526723982444538, + "flos": 21543314457600.0, + "grad_norm": 1.9589578000017436, + "language_loss": 0.64980102, + "learning_rate": 3.003820346949578e-06, + "loss": 0.6718868, + "num_input_tokens_seen": 62992560, + "step": 2933, + "time_per_iteration": 2.6611645221710205 + }, + { + "auxiliary_loss_clip": 0.01196542, + "auxiliary_loss_mlp": 0.01022963, + "balance_loss_clip": 1.06031156, + "balance_loss_mlp": 1.01460028, + "epoch": 0.3527926411350929, + "flos": 23733649900800.0, + "grad_norm": 2.2022338142847793, + "language_loss": 0.79403496, + "learning_rate": 3.003146522863071e-06, + "loss": 0.81623006, + "num_input_tokens_seen": 63013445, + "step": 2934, + "time_per_iteration": 2.610170841217041 + }, + { + "auxiliary_loss_clip": 0.01147995, + "auxiliary_loss_mlp": 0.01031814, + "balance_loss_clip": 0.98381788, + "balance_loss_mlp": 1.02343345, + "epoch": 0.35291288402573195, + "flos": 30445461544320.0, + "grad_norm": 2.341998103575074, + "language_loss": 0.86214167, + "learning_rate": 3.0024725466000345e-06, + "loss": 0.88393974, + "num_input_tokens_seen": 63033400, + "step": 2935, + "time_per_iteration": 3.651071786880493 + }, + { + "auxiliary_loss_clip": 0.0117281, + "auxiliary_loss_mlp": 0.01030819, + "balance_loss_clip": 1.02453363, + "balance_loss_mlp": 1.02252841, + "epoch": 0.35303312691637107, + "flos": 23112179763840.0, + "grad_norm": 2.0417785925362195, + "language_loss": 0.78844571, + "learning_rate": 3.0017984182627087e-06, + "loss": 0.81048203, + "num_input_tokens_seen": 63052725, + "step": 2936, + "time_per_iteration": 2.6134517192840576 + }, + { + "auxiliary_loss_clip": 0.01122516, + "auxiliary_loss_mlp": 0.0051164, + "balance_loss_clip": 0.94225496, + "balance_loss_mlp": 0.0, + "epoch": 0.3531533698070102, + "flos": 21835699165440.0, + "grad_norm": 1.9619833209045847, + "language_loss": 0.82452971, + "learning_rate": 3.00112413795336e-06, + "loss": 0.84087121, + "num_input_tokens_seen": 63072560, + "step": 2937, + "time_per_iteration": 2.7200372219085693 + }, + { + "auxiliary_loss_clip": 0.01143846, + "auxiliary_loss_mlp": 0.0102688, + "balance_loss_clip": 0.97803688, + "balance_loss_mlp": 1.01809454, + "epoch": 0.35327361269764923, + "flos": 15778969810560.0, + "grad_norm": 3.3259119848914422, + "language_loss": 0.7973994, + "learning_rate": 3.000449705774275e-06, + "loss": 0.81910658, + "num_input_tokens_seen": 63090800, + "step": 2938, + "time_per_iteration": 2.720891237258911 + }, + { + "auxiliary_loss_clip": 0.01174091, + "auxiliary_loss_mlp": 0.01027694, + "balance_loss_clip": 1.02391684, + "balance_loss_mlp": 1.01905763, + "epoch": 0.35339385558828834, + "flos": 22090413484800.0, + "grad_norm": 1.930954197385112, + "language_loss": 0.71593994, + "learning_rate": 2.9997751218277654e-06, + "loss": 0.73795784, + "num_input_tokens_seen": 63108955, + "step": 2939, + "time_per_iteration": 2.6573853492736816 + }, + { + "auxiliary_loss_clip": 0.0119608, + "auxiliary_loss_mlp": 0.01030151, + "balance_loss_clip": 1.06094539, + "balance_loss_mlp": 1.02140117, + "epoch": 0.35351409847892745, + "flos": 24165008328960.0, + "grad_norm": 2.013870476519237, + "language_loss": 0.77392662, + "learning_rate": 2.999100386216166e-06, + "loss": 0.79618895, + "num_input_tokens_seen": 63127895, + "step": 2940, + "time_per_iteration": 2.6676933765411377 + }, + { + "auxiliary_loss_clip": 0.0114966, + "auxiliary_loss_mlp": 0.0103312, + "balance_loss_clip": 0.98556894, + "balance_loss_mlp": 1.02445912, + "epoch": 0.3536343413695665, + "flos": 27052298340480.0, + "grad_norm": 1.7829245087792744, + "language_loss": 0.74478364, + "learning_rate": 2.998425499041831e-06, + "loss": 0.7666114, + "num_input_tokens_seen": 63148410, + "step": 2941, + "time_per_iteration": 2.73915696144104 + }, + { + "auxiliary_loss_clip": 0.01073323, + "auxiliary_loss_mlp": 0.01003599, + "balance_loss_clip": 0.98928583, + "balance_loss_mlp": 1.00177479, + "epoch": 0.3537545842602056, + "flos": 65991066370560.0, + "grad_norm": 1.3508423380166357, + "language_loss": 0.64616191, + "learning_rate": 2.997750460407142e-06, + "loss": 0.66693115, + "num_input_tokens_seen": 63209765, + "step": 2942, + "time_per_iteration": 3.308239459991455 + }, + { + "auxiliary_loss_clip": 0.01129364, + "auxiliary_loss_mlp": 0.01028194, + "balance_loss_clip": 0.9437077, + "balance_loss_mlp": 1.01898468, + "epoch": 0.35387482715084473, + "flos": 18436897526400.0, + "grad_norm": 2.1612194610369566, + "language_loss": 0.69869733, + "learning_rate": 2.997075270414501e-06, + "loss": 0.7202729, + "num_input_tokens_seen": 63226980, + "step": 2943, + "time_per_iteration": 2.7485437393188477 + }, + { + "auxiliary_loss_clip": 0.01053684, + "auxiliary_loss_mlp": 0.01002671, + "balance_loss_clip": 0.95474374, + "balance_loss_mlp": 1.00098991, + "epoch": 0.3539950700414838, + "flos": 65588579498880.0, + "grad_norm": 0.7069379901083408, + "language_loss": 0.57689893, + "learning_rate": 2.9963999291663347e-06, + "loss": 0.59746242, + "num_input_tokens_seen": 63292760, + "step": 2944, + "time_per_iteration": 3.27067494392395 + }, + { + "auxiliary_loss_clip": 0.01104, + "auxiliary_loss_mlp": 0.01028641, + "balance_loss_clip": 0.90967351, + "balance_loss_mlp": 1.02018273, + "epoch": 0.3541153129321229, + "flos": 20521655919360.0, + "grad_norm": 2.2207617913031386, + "language_loss": 0.74290276, + "learning_rate": 2.9957244367650915e-06, + "loss": 0.76422918, + "num_input_tokens_seen": 63309005, + "step": 2945, + "time_per_iteration": 2.854280948638916 + }, + { + "auxiliary_loss_clip": 0.01094463, + "auxiliary_loss_mlp": 0.01035316, + "balance_loss_clip": 0.90493309, + "balance_loss_mlp": 1.02609551, + "epoch": 0.354235555822762, + "flos": 19573578391680.0, + "grad_norm": 2.0943787915808993, + "language_loss": 0.8394016, + "learning_rate": 2.9950487933132425e-06, + "loss": 0.86069936, + "num_input_tokens_seen": 63326420, + "step": 2946, + "time_per_iteration": 2.7217514514923096 + }, + { + "auxiliary_loss_clip": 0.01176838, + "auxiliary_loss_mlp": 0.01029228, + "balance_loss_clip": 1.02321517, + "balance_loss_mlp": 1.0204004, + "epoch": 0.35435579871340106, + "flos": 20777268078720.0, + "grad_norm": 6.858948883783676, + "language_loss": 0.71152043, + "learning_rate": 2.994372998913283e-06, + "loss": 0.73358107, + "num_input_tokens_seen": 63344925, + "step": 2947, + "time_per_iteration": 2.7031099796295166 + }, + { + "auxiliary_loss_clip": 0.01152172, + "auxiliary_loss_mlp": 0.01032926, + "balance_loss_clip": 0.98763096, + "balance_loss_mlp": 1.02424788, + "epoch": 0.35447604160404017, + "flos": 23951807153280.0, + "grad_norm": 5.641301587679659, + "language_loss": 0.62551963, + "learning_rate": 2.99369705366773e-06, + "loss": 0.6473707, + "num_input_tokens_seen": 63365170, + "step": 2948, + "time_per_iteration": 2.671616315841675 + }, + { + "auxiliary_loss_clip": 0.01145653, + "auxiliary_loss_mlp": 0.01030712, + "balance_loss_clip": 0.98439169, + "balance_loss_mlp": 1.02195621, + "epoch": 0.3545962844946792, + "flos": 23435662671360.0, + "grad_norm": 3.813142822920064, + "language_loss": 0.81522781, + "learning_rate": 2.9930209576791244e-06, + "loss": 0.83699143, + "num_input_tokens_seen": 63383645, + "step": 2949, + "time_per_iteration": 2.797323226928711 + }, + { + "auxiliary_loss_clip": 0.01166487, + "auxiliary_loss_mlp": 0.01032277, + "balance_loss_clip": 1.01846838, + "balance_loss_mlp": 1.02432024, + "epoch": 0.35471652738531834, + "flos": 22085134185600.0, + "grad_norm": 2.910887004914826, + "language_loss": 0.63744295, + "learning_rate": 2.9923447110500285e-06, + "loss": 0.65943062, + "num_input_tokens_seen": 63402390, + "step": 2950, + "time_per_iteration": 2.6197125911712646 + }, + { + "auxiliary_loss_clip": 0.01158555, + "auxiliary_loss_mlp": 0.01034907, + "balance_loss_clip": 1.0173173, + "balance_loss_mlp": 1.02641273, + "epoch": 0.35483677027595745, + "flos": 27341881787520.0, + "grad_norm": 3.3055635038409488, + "language_loss": 0.75450104, + "learning_rate": 2.9916683138830295e-06, + "loss": 0.77643561, + "num_input_tokens_seen": 63423055, + "step": 2951, + "time_per_iteration": 2.717599868774414 + }, + { + "auxiliary_loss_clip": 0.01144767, + "auxiliary_loss_mlp": 0.01027793, + "balance_loss_clip": 0.98308444, + "balance_loss_mlp": 1.01900136, + "epoch": 0.3549570131665965, + "flos": 13516166678400.0, + "grad_norm": 2.626183066998394, + "language_loss": 0.80846918, + "learning_rate": 2.9909917662807353e-06, + "loss": 0.83019477, + "num_input_tokens_seen": 63440855, + "step": 2952, + "time_per_iteration": 3.537769079208374 + }, + { + "auxiliary_loss_clip": 0.01169644, + "auxiliary_loss_mlp": 0.01035535, + "balance_loss_clip": 1.02134919, + "balance_loss_mlp": 1.02702928, + "epoch": 0.3550772560572356, + "flos": 20887549810560.0, + "grad_norm": 15.961319841436286, + "language_loss": 0.68993104, + "learning_rate": 2.9903150683457783e-06, + "loss": 0.71198285, + "num_input_tokens_seen": 63459400, + "step": 2953, + "time_per_iteration": 2.6508841514587402 + }, + { + "auxiliary_loss_clip": 0.01147698, + "auxiliary_loss_mlp": 0.01026128, + "balance_loss_clip": 0.98053658, + "balance_loss_mlp": 1.01714563, + "epoch": 0.3551974989478747, + "flos": 20194042947840.0, + "grad_norm": 2.321860445565349, + "language_loss": 0.64968157, + "learning_rate": 2.9896382201808126e-06, + "loss": 0.6714198, + "num_input_tokens_seen": 63476800, + "step": 2954, + "time_per_iteration": 2.6666409969329834 + }, + { + "auxiliary_loss_clip": 0.01198549, + "auxiliary_loss_mlp": 0.01033537, + "balance_loss_clip": 1.06245863, + "balance_loss_mlp": 1.02384567, + "epoch": 0.3553177418385138, + "flos": 19828831415040.0, + "grad_norm": 1.9893933506044648, + "language_loss": 0.81264454, + "learning_rate": 2.988961221888516e-06, + "loss": 0.83496535, + "num_input_tokens_seen": 63493475, + "step": 2955, + "time_per_iteration": 3.5539989471435547 + }, + { + "auxiliary_loss_clip": 0.01112441, + "auxiliary_loss_mlp": 0.01027472, + "balance_loss_clip": 0.94028342, + "balance_loss_mlp": 1.01903796, + "epoch": 0.3554379847291529, + "flos": 14829132516480.0, + "grad_norm": 2.2638791994254155, + "language_loss": 0.78565049, + "learning_rate": 2.988284073571589e-06, + "loss": 0.80704963, + "num_input_tokens_seen": 63509560, + "step": 2956, + "time_per_iteration": 2.6972246170043945 + }, + { + "auxiliary_loss_clip": 0.01173159, + "auxiliary_loss_mlp": 0.00510758, + "balance_loss_clip": 1.02118695, + "balance_loss_mlp": 0.0, + "epoch": 0.355558227619792, + "flos": 20485350247680.0, + "grad_norm": 2.633677481081484, + "language_loss": 0.73020703, + "learning_rate": 2.9876067753327528e-06, + "loss": 0.74704623, + "num_input_tokens_seen": 63527290, + "step": 2957, + "time_per_iteration": 2.610063076019287 + }, + { + "auxiliary_loss_clip": 0.0117659, + "auxiliary_loss_mlp": 0.01029926, + "balance_loss_clip": 1.02132702, + "balance_loss_mlp": 1.02155197, + "epoch": 0.35567847051043106, + "flos": 37663613256960.0, + "grad_norm": 1.746160920472897, + "language_loss": 0.80258036, + "learning_rate": 2.986929327274754e-06, + "loss": 0.82464552, + "num_input_tokens_seen": 63547870, + "step": 2958, + "time_per_iteration": 2.7702081203460693 + }, + { + "auxiliary_loss_clip": 0.01170321, + "auxiliary_loss_mlp": 0.01029881, + "balance_loss_clip": 1.02262032, + "balance_loss_mlp": 1.02141738, + "epoch": 0.35579871340107017, + "flos": 26943058103040.0, + "grad_norm": 4.159862993666919, + "language_loss": 0.78916931, + "learning_rate": 2.9862517295003617e-06, + "loss": 0.81117135, + "num_input_tokens_seen": 63568285, + "step": 2959, + "time_per_iteration": 2.673765182495117 + }, + { + "auxiliary_loss_clip": 0.01118292, + "auxiliary_loss_mlp": 0.01028913, + "balance_loss_clip": 0.93937963, + "balance_loss_mlp": 1.02004409, + "epoch": 0.3559189562917093, + "flos": 28293335193600.0, + "grad_norm": 1.5002899954579152, + "language_loss": 0.72258967, + "learning_rate": 2.9855739821123654e-06, + "loss": 0.74406171, + "num_input_tokens_seen": 63589865, + "step": 2960, + "time_per_iteration": 2.7960944175720215 + }, + { + "auxiliary_loss_clip": 0.01167491, + "auxiliary_loss_mlp": 0.01031082, + "balance_loss_clip": 1.02116108, + "balance_loss_mlp": 1.02317238, + "epoch": 0.35603919918234833, + "flos": 25664063552640.0, + "grad_norm": 4.090851630769214, + "language_loss": 0.82387543, + "learning_rate": 2.98489608521358e-06, + "loss": 0.84586108, + "num_input_tokens_seen": 63609805, + "step": 2961, + "time_per_iteration": 3.5556061267852783 + }, + { + "auxiliary_loss_clip": 0.01176099, + "auxiliary_loss_mlp": 0.00511191, + "balance_loss_clip": 1.02182579, + "balance_loss_mlp": 0.0, + "epoch": 0.35615944207298744, + "flos": 23000856537600.0, + "grad_norm": 2.039398940850265, + "language_loss": 0.79786503, + "learning_rate": 2.9842180389068425e-06, + "loss": 0.81473792, + "num_input_tokens_seen": 63627115, + "step": 2962, + "time_per_iteration": 2.6094233989715576 + }, + { + "auxiliary_loss_clip": 0.010216, + "auxiliary_loss_mlp": 0.01001602, + "balance_loss_clip": 0.91180933, + "balance_loss_mlp": 0.99990886, + "epoch": 0.35627968496362655, + "flos": 68251283723520.0, + "grad_norm": 0.761033055278284, + "language_loss": 0.59199107, + "learning_rate": 2.98353984329501e-06, + "loss": 0.61222315, + "num_input_tokens_seen": 63691460, + "step": 2963, + "time_per_iteration": 3.2484562397003174 + }, + { + "auxiliary_loss_clip": 0.01145901, + "auxiliary_loss_mlp": 0.01036831, + "balance_loss_clip": 0.9820627, + "balance_loss_mlp": 1.02805722, + "epoch": 0.3563999278542656, + "flos": 22641714403200.0, + "grad_norm": 1.8924593829687706, + "language_loss": 0.70289409, + "learning_rate": 2.982861498480965e-06, + "loss": 0.72472143, + "num_input_tokens_seen": 63713840, + "step": 2964, + "time_per_iteration": 2.6876726150512695 + }, + { + "auxiliary_loss_clip": 0.01116795, + "auxiliary_loss_mlp": 0.0103421, + "balance_loss_clip": 0.9394539, + "balance_loss_mlp": 1.02539444, + "epoch": 0.3565201707449047, + "flos": 25952533678080.0, + "grad_norm": 1.8510921069893211, + "language_loss": 0.82582247, + "learning_rate": 2.9821830045676122e-06, + "loss": 0.84733248, + "num_input_tokens_seen": 63733540, + "step": 2965, + "time_per_iteration": 2.6913530826568604 + }, + { + "auxiliary_loss_clip": 0.01198343, + "auxiliary_loss_mlp": 0.01029027, + "balance_loss_clip": 1.0626179, + "balance_loss_mlp": 1.02026546, + "epoch": 0.3566404136355438, + "flos": 28475725478400.0, + "grad_norm": 1.8368638973527245, + "language_loss": 0.73025107, + "learning_rate": 2.9815043616578793e-06, + "loss": 0.75252473, + "num_input_tokens_seen": 63754335, + "step": 2966, + "time_per_iteration": 2.6449244022369385 + }, + { + "auxiliary_loss_clip": 0.01118204, + "auxiliary_loss_mlp": 0.01030611, + "balance_loss_clip": 0.94020528, + "balance_loss_mlp": 1.0216403, + "epoch": 0.3567606565261829, + "flos": 38363117690880.0, + "grad_norm": 1.9849405792132453, + "language_loss": 0.76436949, + "learning_rate": 2.9808255698547145e-06, + "loss": 0.78585768, + "num_input_tokens_seen": 63777135, + "step": 2967, + "time_per_iteration": 2.870572805404663 + }, + { + "auxiliary_loss_clip": 0.01170086, + "auxiliary_loss_mlp": 0.01033281, + "balance_loss_clip": 1.02360773, + "balance_loss_mlp": 1.02528763, + "epoch": 0.356880899416822, + "flos": 21981029592960.0, + "grad_norm": 1.9340234020147824, + "language_loss": 0.7933383, + "learning_rate": 2.9801466292610913e-06, + "loss": 0.81537199, + "num_input_tokens_seen": 63797020, + "step": 2968, + "time_per_iteration": 2.692387580871582 + }, + { + "auxiliary_loss_clip": 0.01169375, + "auxiliary_loss_mlp": 0.01029131, + "balance_loss_clip": 1.02075803, + "balance_loss_mlp": 1.02102494, + "epoch": 0.35700114230746105, + "flos": 18989132198400.0, + "grad_norm": 2.5644907350428467, + "language_loss": 0.81165332, + "learning_rate": 2.979467539980003e-06, + "loss": 0.83363837, + "num_input_tokens_seen": 63813810, + "step": 2969, + "time_per_iteration": 2.805812120437622 + }, + { + "auxiliary_loss_clip": 0.01172017, + "auxiliary_loss_mlp": 0.01029764, + "balance_loss_clip": 1.02103925, + "balance_loss_mlp": 1.02113342, + "epoch": 0.35712138519810016, + "flos": 19756112330880.0, + "grad_norm": 1.986734179077117, + "language_loss": 0.76927322, + "learning_rate": 2.978788302114468e-06, + "loss": 0.79129112, + "num_input_tokens_seen": 63830925, + "step": 2970, + "time_per_iteration": 2.6663405895233154 + }, + { + "auxiliary_loss_clip": 0.01171814, + "auxiliary_loss_mlp": 0.01037095, + "balance_loss_clip": 1.02266741, + "balance_loss_mlp": 1.02833319, + "epoch": 0.35724162808873927, + "flos": 35183012008320.0, + "grad_norm": 4.101582469167228, + "language_loss": 0.83002621, + "learning_rate": 2.9781089157675255e-06, + "loss": 0.85211527, + "num_input_tokens_seen": 63849385, + "step": 2971, + "time_per_iteration": 2.8156936168670654 + }, + { + "auxiliary_loss_clip": 0.01167976, + "auxiliary_loss_mlp": 0.01026041, + "balance_loss_clip": 1.02236962, + "balance_loss_mlp": 1.01786935, + "epoch": 0.3573618709793783, + "flos": 25556726736000.0, + "grad_norm": 1.6231749343310944, + "language_loss": 0.8836416, + "learning_rate": 2.977429381042238e-06, + "loss": 0.90558183, + "num_input_tokens_seen": 63870060, + "step": 2972, + "time_per_iteration": 2.71044921875 + }, + { + "auxiliary_loss_clip": 0.01147107, + "auxiliary_loss_mlp": 0.0102907, + "balance_loss_clip": 0.98248708, + "balance_loss_mlp": 1.02126145, + "epoch": 0.35748211387001744, + "flos": 29132352051840.0, + "grad_norm": 2.0633023941966733, + "language_loss": 0.88801801, + "learning_rate": 2.9767496980416913e-06, + "loss": 0.90977973, + "num_input_tokens_seen": 63889355, + "step": 2973, + "time_per_iteration": 2.7230610847473145 + }, + { + "auxiliary_loss_clip": 0.01139435, + "auxiliary_loss_mlp": 0.01028473, + "balance_loss_clip": 0.97898495, + "balance_loss_mlp": 1.01992559, + "epoch": 0.35760235676065655, + "flos": 13954169122560.0, + "grad_norm": 2.2543561444768048, + "language_loss": 0.81148618, + "learning_rate": 2.9760698668689914e-06, + "loss": 0.83316523, + "num_input_tokens_seen": 63905580, + "step": 2974, + "time_per_iteration": 2.715001344680786 + }, + { + "auxiliary_loss_clip": 0.01169648, + "auxiliary_loss_mlp": 0.01029581, + "balance_loss_clip": 1.01945508, + "balance_loss_mlp": 1.02068162, + "epoch": 0.3577225996512956, + "flos": 44018688977280.0, + "grad_norm": 2.0222352054851194, + "language_loss": 0.71075344, + "learning_rate": 2.975389887627269e-06, + "loss": 0.73274571, + "num_input_tokens_seen": 63928180, + "step": 2975, + "time_per_iteration": 2.8451685905456543 + }, + { + "auxiliary_loss_clip": 0.01127692, + "auxiliary_loss_mlp": 0.01029523, + "balance_loss_clip": 0.94444001, + "balance_loss_mlp": 1.02132702, + "epoch": 0.3578428425419347, + "flos": 17055199013760.0, + "grad_norm": 2.090948746649218, + "language_loss": 0.89935374, + "learning_rate": 2.9747097604196764e-06, + "loss": 0.92092592, + "num_input_tokens_seen": 63944825, + "step": 2976, + "time_per_iteration": 2.8452157974243164 + }, + { + "auxiliary_loss_clip": 0.01012976, + "auxiliary_loss_mlp": 0.01006165, + "balance_loss_clip": 0.87859666, + "balance_loss_mlp": 1.00437665, + "epoch": 0.3579630854325738, + "flos": 71676550707840.0, + "grad_norm": 0.6755976802657794, + "language_loss": 0.56660104, + "learning_rate": 2.9740294853493875e-06, + "loss": 0.58679235, + "num_input_tokens_seen": 64016385, + "step": 2977, + "time_per_iteration": 3.4789414405822754 + }, + { + "auxiliary_loss_clip": 0.01105441, + "auxiliary_loss_mlp": 0.01028392, + "balance_loss_clip": 0.90497297, + "balance_loss_mlp": 1.01929641, + "epoch": 0.3580833283232129, + "flos": 25046651652480.0, + "grad_norm": 2.179320869739504, + "language_loss": 0.67292434, + "learning_rate": 2.9733490625196008e-06, + "loss": 0.69426268, + "num_input_tokens_seen": 64036245, + "step": 2978, + "time_per_iteration": 3.7236509323120117 + }, + { + "auxiliary_loss_clip": 0.01113544, + "auxiliary_loss_mlp": 0.01032954, + "balance_loss_clip": 0.94252509, + "balance_loss_mlp": 1.02446651, + "epoch": 0.358203571213852, + "flos": 13953127628160.0, + "grad_norm": 3.9322012202591887, + "language_loss": 0.75347078, + "learning_rate": 2.9726684920335353e-06, + "loss": 0.77493578, + "num_input_tokens_seen": 64054110, + "step": 2979, + "time_per_iteration": 2.7179408073425293 + }, + { + "auxiliary_loss_clip": 0.01196222, + "auxiliary_loss_mlp": 0.00511582, + "balance_loss_clip": 1.05806863, + "balance_loss_mlp": 0.0, + "epoch": 0.35832381410449105, + "flos": 20302457172480.0, + "grad_norm": 2.774428780523488, + "language_loss": 0.81986922, + "learning_rate": 2.971987773994432e-06, + "loss": 0.83694726, + "num_input_tokens_seen": 64070295, + "step": 2980, + "time_per_iteration": 2.6164956092834473 + }, + { + "auxiliary_loss_clip": 0.01161096, + "auxiliary_loss_mlp": 0.01030078, + "balance_loss_clip": 1.01684391, + "balance_loss_mlp": 1.02135205, + "epoch": 0.35844405699513016, + "flos": 16983234115200.0, + "grad_norm": 8.374408396939563, + "language_loss": 0.8324604, + "learning_rate": 2.9713069085055566e-06, + "loss": 0.85437214, + "num_input_tokens_seen": 64088605, + "step": 2981, + "time_per_iteration": 4.557384490966797 + }, + { + "auxiliary_loss_clip": 0.01124453, + "auxiliary_loss_mlp": 0.01035416, + "balance_loss_clip": 0.94497436, + "balance_loss_mlp": 1.02674961, + "epoch": 0.35856429988576927, + "flos": 23216858974080.0, + "grad_norm": 1.6546521078346363, + "language_loss": 0.78838605, + "learning_rate": 2.9706258956701958e-06, + "loss": 0.80998474, + "num_input_tokens_seen": 64108595, + "step": 2982, + "time_per_iteration": 2.8218746185302734 + }, + { + "auxiliary_loss_clip": 0.01174256, + "auxiliary_loss_mlp": 0.01026523, + "balance_loss_clip": 1.02225423, + "balance_loss_mlp": 1.01732612, + "epoch": 0.3586845427764083, + "flos": 23034576430080.0, + "grad_norm": 2.168005799122691, + "language_loss": 0.776977, + "learning_rate": 2.9699447355916575e-06, + "loss": 0.79898483, + "num_input_tokens_seen": 64127405, + "step": 2983, + "time_per_iteration": 2.7727346420288086 + }, + { + "auxiliary_loss_clip": 0.01192639, + "auxiliary_loss_mlp": 0.00511128, + "balance_loss_clip": 1.0581013, + "balance_loss_mlp": 0.0, + "epoch": 0.35880478566704743, + "flos": 20010682995840.0, + "grad_norm": 1.9689084218052664, + "language_loss": 0.73985672, + "learning_rate": 2.969263428373275e-06, + "loss": 0.75689435, + "num_input_tokens_seen": 64145755, + "step": 2984, + "time_per_iteration": 2.5991311073303223 + }, + { + "auxiliary_loss_clip": 0.01146719, + "auxiliary_loss_mlp": 0.01029753, + "balance_loss_clip": 0.98073661, + "balance_loss_mlp": 1.0212239, + "epoch": 0.35892502855768654, + "flos": 13699095667200.0, + "grad_norm": 2.0978359310963492, + "language_loss": 0.79461443, + "learning_rate": 2.9685819741184007e-06, + "loss": 0.81637919, + "num_input_tokens_seen": 64164195, + "step": 2985, + "time_per_iteration": 2.6668076515197754 + }, + { + "auxiliary_loss_clip": 0.0111867, + "auxiliary_loss_mlp": 0.0102908, + "balance_loss_clip": 0.94292033, + "balance_loss_mlp": 1.02058065, + "epoch": 0.3590452714483256, + "flos": 18114096977280.0, + "grad_norm": 3.012477636111384, + "language_loss": 0.68657583, + "learning_rate": 2.967900372930411e-06, + "loss": 0.70805329, + "num_input_tokens_seen": 64182705, + "step": 2986, + "time_per_iteration": 2.6683521270751953 + }, + { + "auxiliary_loss_clip": 0.01138579, + "auxiliary_loss_mlp": 0.01035679, + "balance_loss_clip": 0.9780066, + "balance_loss_mlp": 1.02650571, + "epoch": 0.3591655143389647, + "flos": 17749352321280.0, + "grad_norm": 1.979337241434647, + "language_loss": 0.78935015, + "learning_rate": 2.9672186249127046e-06, + "loss": 0.81109273, + "num_input_tokens_seen": 64202170, + "step": 2987, + "time_per_iteration": 3.5992956161499023 + }, + { + "auxiliary_loss_clip": 0.0114444, + "auxiliary_loss_mlp": 0.01030399, + "balance_loss_clip": 0.98060292, + "balance_loss_mlp": 1.02179193, + "epoch": 0.3592857572296038, + "flos": 25224409082880.0, + "grad_norm": 1.8612506555533606, + "language_loss": 0.78741258, + "learning_rate": 2.9665367301687014e-06, + "loss": 0.80916095, + "num_input_tokens_seen": 64220415, + "step": 2988, + "time_per_iteration": 2.730132579803467 + }, + { + "auxiliary_loss_clip": 0.01135111, + "auxiliary_loss_mlp": 0.01027631, + "balance_loss_clip": 0.97639072, + "balance_loss_mlp": 1.01919699, + "epoch": 0.3594060001202429, + "flos": 29384408764800.0, + "grad_norm": 1.8609892255107057, + "language_loss": 0.76100361, + "learning_rate": 2.965854688801845e-06, + "loss": 0.78263098, + "num_input_tokens_seen": 64242475, + "step": 2989, + "time_per_iteration": 2.7274904251098633 + }, + { + "auxiliary_loss_clip": 0.01165939, + "auxiliary_loss_mlp": 0.01024179, + "balance_loss_clip": 1.01559639, + "balance_loss_mlp": 1.01573908, + "epoch": 0.359526243010882, + "flos": 17052900543360.0, + "grad_norm": 1.9927194821631449, + "language_loss": 0.76311362, + "learning_rate": 2.9651725009156005e-06, + "loss": 0.78501475, + "num_input_tokens_seen": 64260220, + "step": 2990, + "time_per_iteration": 2.6140098571777344 + }, + { + "auxiliary_loss_clip": 0.01138839, + "auxiliary_loss_mlp": 0.01032143, + "balance_loss_clip": 0.97886264, + "balance_loss_mlp": 1.02363145, + "epoch": 0.3596464859015211, + "flos": 22965089569920.0, + "grad_norm": 1.7499219565307744, + "language_loss": 0.74214154, + "learning_rate": 2.964490166613454e-06, + "loss": 0.7638514, + "num_input_tokens_seen": 64280145, + "step": 2991, + "time_per_iteration": 2.7155444622039795 + }, + { + "auxiliary_loss_clip": 0.01096261, + "auxiliary_loss_mlp": 0.01000417, + "balance_loss_clip": 1.03053951, + "balance_loss_mlp": 0.99865228, + "epoch": 0.35976672879216015, + "flos": 54739462590720.0, + "grad_norm": 0.7571535559189796, + "language_loss": 0.57736397, + "learning_rate": 2.963807685998917e-06, + "loss": 0.59833074, + "num_input_tokens_seen": 64336010, + "step": 2992, + "time_per_iteration": 3.065138578414917 + }, + { + "auxiliary_loss_clip": 0.01103932, + "auxiliary_loss_mlp": 0.0103113, + "balance_loss_clip": 0.90430522, + "balance_loss_mlp": 1.0227077, + "epoch": 0.35988697168279926, + "flos": 43139020901760.0, + "grad_norm": 4.043063152638128, + "language_loss": 0.78181905, + "learning_rate": 2.9631250591755196e-06, + "loss": 0.80316961, + "num_input_tokens_seen": 64358725, + "step": 2993, + "time_per_iteration": 2.9521539211273193 + }, + { + "auxiliary_loss_clip": 0.01142738, + "auxiliary_loss_mlp": 0.01025721, + "balance_loss_clip": 0.98347837, + "balance_loss_mlp": 1.01705456, + "epoch": 0.36000721457343837, + "flos": 35845600239360.0, + "grad_norm": 1.6938157064651236, + "language_loss": 0.5730173, + "learning_rate": 2.962442286246817e-06, + "loss": 0.59470189, + "num_input_tokens_seen": 64381555, + "step": 2994, + "time_per_iteration": 2.795797109603882 + }, + { + "auxiliary_loss_clip": 0.01147143, + "auxiliary_loss_mlp": 0.01027658, + "balance_loss_clip": 0.98019886, + "balance_loss_mlp": 1.01931953, + "epoch": 0.3601274574640774, + "flos": 18291100222080.0, + "grad_norm": 1.6436064279051184, + "language_loss": 0.69923532, + "learning_rate": 2.9617593673163853e-06, + "loss": 0.72098327, + "num_input_tokens_seen": 64400375, + "step": 2995, + "time_per_iteration": 2.6879725456237793 + }, + { + "auxiliary_loss_clip": 0.0114683, + "auxiliary_loss_mlp": 0.01032851, + "balance_loss_clip": 0.97860038, + "balance_loss_mlp": 1.02450609, + "epoch": 0.36024770035471654, + "flos": 13333955961600.0, + "grad_norm": 2.2019509634730334, + "language_loss": 0.77546549, + "learning_rate": 2.9610763024878216e-06, + "loss": 0.79726225, + "num_input_tokens_seen": 64415880, + "step": 2996, + "time_per_iteration": 2.7552711963653564 + }, + { + "auxiliary_loss_clip": 0.01140925, + "auxiliary_loss_mlp": 0.01038019, + "balance_loss_clip": 0.97878408, + "balance_loss_mlp": 1.02914429, + "epoch": 0.3603679432453556, + "flos": 20267013427200.0, + "grad_norm": 1.7903492629663504, + "language_loss": 0.91858399, + "learning_rate": 2.960393091864747e-06, + "loss": 0.94037348, + "num_input_tokens_seen": 64434260, + "step": 2997, + "time_per_iteration": 2.723456621170044 + }, + { + "auxiliary_loss_clip": 0.01146949, + "auxiliary_loss_mlp": 0.01032117, + "balance_loss_clip": 0.9824326, + "balance_loss_mlp": 1.02424359, + "epoch": 0.3604881861359947, + "flos": 22451135817600.0, + "grad_norm": 1.8354166409670944, + "language_loss": 0.75114632, + "learning_rate": 2.959709735550804e-06, + "loss": 0.77293694, + "num_input_tokens_seen": 64453855, + "step": 2998, + "time_per_iteration": 2.7425479888916016 + }, + { + "auxiliary_loss_clip": 0.01102325, + "auxiliary_loss_mlp": 0.01028706, + "balance_loss_clip": 0.90342224, + "balance_loss_mlp": 1.02027798, + "epoch": 0.3606084290266338, + "flos": 22054251467520.0, + "grad_norm": 2.2938213669557754, + "language_loss": 0.76315415, + "learning_rate": 2.9590262336496575e-06, + "loss": 0.78446448, + "num_input_tokens_seen": 64473585, + "step": 2999, + "time_per_iteration": 2.9270904064178467 + }, + { + "auxiliary_loss_clip": 0.01116801, + "auxiliary_loss_mlp": 0.01026887, + "balance_loss_clip": 0.94362456, + "balance_loss_mlp": 1.0179584, + "epoch": 0.36072867191727287, + "flos": 15632921111040.0, + "grad_norm": 3.2333771344558757, + "language_loss": 0.85729891, + "learning_rate": 2.9583425862649936e-06, + "loss": 0.87873578, + "num_input_tokens_seen": 64491720, + "step": 3000, + "time_per_iteration": 2.6834826469421387 + }, + { + "auxiliary_loss_clip": 0.01196342, + "auxiliary_loss_mlp": 0.01031924, + "balance_loss_clip": 1.06037748, + "balance_loss_mlp": 1.02382338, + "epoch": 0.360848914807912, + "flos": 19677000625920.0, + "grad_norm": 3.0941982812229614, + "language_loss": 0.74363256, + "learning_rate": 2.9576587935005215e-06, + "loss": 0.76591516, + "num_input_tokens_seen": 64509800, + "step": 3001, + "time_per_iteration": 2.636747121810913 + }, + { + "auxiliary_loss_clip": 0.01170817, + "auxiliary_loss_mlp": 0.0103075, + "balance_loss_clip": 1.01875508, + "balance_loss_mlp": 1.02144599, + "epoch": 0.3609691576985511, + "flos": 18877808972160.0, + "grad_norm": 2.423988709023543, + "language_loss": 0.72021466, + "learning_rate": 2.9569748554599713e-06, + "loss": 0.74223036, + "num_input_tokens_seen": 64525410, + "step": 3002, + "time_per_iteration": 2.6521458625793457 + }, + { + "auxiliary_loss_clip": 0.01146207, + "auxiliary_loss_mlp": 0.010425, + "balance_loss_clip": 0.98209429, + "balance_loss_mlp": 1.03361893, + "epoch": 0.36108940058919015, + "flos": 42224088648960.0, + "grad_norm": 2.3115201616159236, + "language_loss": 0.73173147, + "learning_rate": 2.956290772247097e-06, + "loss": 0.75361854, + "num_input_tokens_seen": 64544085, + "step": 3003, + "time_per_iteration": 2.8645410537719727 + }, + { + "auxiliary_loss_clip": 0.01090341, + "auxiliary_loss_mlp": 0.01031955, + "balance_loss_clip": 0.90320235, + "balance_loss_mlp": 1.02354479, + "epoch": 0.36120964347982926, + "flos": 23185150243200.0, + "grad_norm": 1.7070591426562496, + "language_loss": 0.73389947, + "learning_rate": 2.9556065439656724e-06, + "loss": 0.75512248, + "num_input_tokens_seen": 64563135, + "step": 3004, + "time_per_iteration": 3.725416660308838 + }, + { + "auxiliary_loss_clip": 0.01063311, + "auxiliary_loss_mlp": 0.01029852, + "balance_loss_clip": 0.85820293, + "balance_loss_mlp": 1.02160287, + "epoch": 0.36132988637046837, + "flos": 18113055482880.0, + "grad_norm": 1.6466491795031741, + "language_loss": 0.8146677, + "learning_rate": 2.9549221707194952e-06, + "loss": 0.83559936, + "num_input_tokens_seen": 64581985, + "step": 3005, + "time_per_iteration": 2.843982696533203 + }, + { + "auxiliary_loss_clip": 0.01170776, + "auxiliary_loss_mlp": 0.01029672, + "balance_loss_clip": 1.02018642, + "balance_loss_mlp": 1.02082705, + "epoch": 0.3614501292611074, + "flos": 27813101333760.0, + "grad_norm": 2.2247004165330226, + "language_loss": 0.72680461, + "learning_rate": 2.954237652612384e-06, + "loss": 0.7488091, + "num_input_tokens_seen": 64601035, + "step": 3006, + "time_per_iteration": 2.722116231918335 + }, + { + "auxiliary_loss_clip": 0.01142541, + "auxiliary_loss_mlp": 0.01031101, + "balance_loss_clip": 0.97971582, + "balance_loss_mlp": 1.02257168, + "epoch": 0.36157037215174653, + "flos": 22634926732800.0, + "grad_norm": 1.9284456706321043, + "language_loss": 0.84354949, + "learning_rate": 2.9535529897481796e-06, + "loss": 0.86528587, + "num_input_tokens_seen": 64618580, + "step": 3007, + "time_per_iteration": 4.609402656555176 + }, + { + "auxiliary_loss_clip": 0.01193047, + "auxiliary_loss_mlp": 0.01028952, + "balance_loss_clip": 1.05753231, + "balance_loss_mlp": 1.02007103, + "epoch": 0.36169061504238564, + "flos": 12600839376000.0, + "grad_norm": 2.6575590927906725, + "language_loss": 0.7708596, + "learning_rate": 2.9528681822307446e-06, + "loss": 0.79307961, + "num_input_tokens_seen": 64635430, + "step": 3008, + "time_per_iteration": 2.7066261768341064 + }, + { + "auxiliary_loss_clip": 0.01166553, + "auxiliary_loss_mlp": 0.00510431, + "balance_loss_clip": 1.01984191, + "balance_loss_mlp": 0.0, + "epoch": 0.3618108579330247, + "flos": 26684644682880.0, + "grad_norm": 2.1039794477970215, + "language_loss": 0.81947577, + "learning_rate": 2.952183230163964e-06, + "loss": 0.8362456, + "num_input_tokens_seen": 64655005, + "step": 3009, + "time_per_iteration": 2.7459003925323486 + }, + { + "auxiliary_loss_clip": 0.01115856, + "auxiliary_loss_mlp": 0.01026365, + "balance_loss_clip": 0.94015884, + "balance_loss_mlp": 1.01803803, + "epoch": 0.3619311008236638, + "flos": 22817029708800.0, + "grad_norm": 2.7575829822930267, + "language_loss": 0.72736436, + "learning_rate": 2.9514981336517448e-06, + "loss": 0.74878651, + "num_input_tokens_seen": 64674775, + "step": 3010, + "time_per_iteration": 2.860410690307617 + }, + { + "auxiliary_loss_clip": 0.01169886, + "auxiliary_loss_mlp": 0.01024451, + "balance_loss_clip": 1.02156782, + "balance_loss_mlp": 1.01630878, + "epoch": 0.36205134371430286, + "flos": 25919603884800.0, + "grad_norm": 2.014713475784561, + "language_loss": 0.81603789, + "learning_rate": 2.950812892798015e-06, + "loss": 0.83798122, + "num_input_tokens_seen": 64695670, + "step": 3011, + "time_per_iteration": 2.720512866973877 + }, + { + "auxiliary_loss_clip": 0.01093622, + "auxiliary_loss_mlp": 0.00510584, + "balance_loss_clip": 0.9047662, + "balance_loss_mlp": 0.0, + "epoch": 0.362171586604942, + "flos": 26139592730880.0, + "grad_norm": 1.815492793745663, + "language_loss": 0.8764298, + "learning_rate": 2.9501275077067256e-06, + "loss": 0.89247191, + "num_input_tokens_seen": 64716290, + "step": 3012, + "time_per_iteration": 2.790860176086426 + }, + { + "auxiliary_loss_clip": 0.01055715, + "auxiliary_loss_mlp": 0.01029363, + "balance_loss_clip": 0.85891509, + "balance_loss_mlp": 1.02129841, + "epoch": 0.3622918294955811, + "flos": 28074208273920.0, + "grad_norm": 1.8667877141933609, + "language_loss": 0.88345039, + "learning_rate": 2.949441978481848e-06, + "loss": 0.90430117, + "num_input_tokens_seen": 64737190, + "step": 3013, + "time_per_iteration": 3.7865772247314453 + }, + { + "auxiliary_loss_clip": 0.01124956, + "auxiliary_loss_mlp": 0.01029539, + "balance_loss_clip": 0.94111955, + "balance_loss_mlp": 1.02099752, + "epoch": 0.36241207238622014, + "flos": 19828005402240.0, + "grad_norm": 2.437467870535229, + "language_loss": 0.80407637, + "learning_rate": 2.9487563052273778e-06, + "loss": 0.82562137, + "num_input_tokens_seen": 64753950, + "step": 3014, + "time_per_iteration": 2.798693895339966 + }, + { + "auxiliary_loss_clip": 0.01165492, + "auxiliary_loss_mlp": 0.01026096, + "balance_loss_clip": 1.02120614, + "balance_loss_mlp": 1.01751876, + "epoch": 0.36253231527685925, + "flos": 21397158017280.0, + "grad_norm": 2.4851382143632885, + "language_loss": 0.85579491, + "learning_rate": 2.94807048804733e-06, + "loss": 0.87771082, + "num_input_tokens_seen": 64773570, + "step": 3015, + "time_per_iteration": 2.634854316711426 + }, + { + "auxiliary_loss_clip": 0.01124462, + "auxiliary_loss_mlp": 0.01035128, + "balance_loss_clip": 0.93953335, + "balance_loss_mlp": 1.02652168, + "epoch": 0.36265255816749836, + "flos": 18362885552640.0, + "grad_norm": 1.9281846197262287, + "language_loss": 0.90080082, + "learning_rate": 2.9473845270457434e-06, + "loss": 0.92239666, + "num_input_tokens_seen": 64790385, + "step": 3016, + "time_per_iteration": 2.716909408569336 + }, + { + "auxiliary_loss_clip": 0.01137573, + "auxiliary_loss_mlp": 0.01025935, + "balance_loss_clip": 0.97779059, + "balance_loss_mlp": 1.01825476, + "epoch": 0.3627728010581374, + "flos": 18660046769280.0, + "grad_norm": 2.8900284858432888, + "language_loss": 0.6955809, + "learning_rate": 2.946698422326677e-06, + "loss": 0.71721596, + "num_input_tokens_seen": 64807845, + "step": 3017, + "time_per_iteration": 2.6539955139160156 + }, + { + "auxiliary_loss_clip": 0.01098292, + "auxiliary_loss_mlp": 0.01028858, + "balance_loss_clip": 0.90212321, + "balance_loss_mlp": 1.02018595, + "epoch": 0.36289304394877653, + "flos": 27524272072320.0, + "grad_norm": 2.904167654382013, + "language_loss": 0.79963416, + "learning_rate": 2.946012173994213e-06, + "loss": 0.82090563, + "num_input_tokens_seen": 64827630, + "step": 3018, + "time_per_iteration": 2.7788171768188477 + }, + { + "auxiliary_loss_clip": 0.01162645, + "auxiliary_loss_mlp": 0.01030153, + "balance_loss_clip": 1.02051222, + "balance_loss_mlp": 1.02189755, + "epoch": 0.36301328683941564, + "flos": 34533244932480.0, + "grad_norm": 1.4153765256707738, + "language_loss": 0.67780465, + "learning_rate": 2.945325782152454e-06, + "loss": 0.69973266, + "num_input_tokens_seen": 64850665, + "step": 3019, + "time_per_iteration": 2.740993022918701 + }, + { + "auxiliary_loss_clip": 0.01145591, + "auxiliary_loss_mlp": 0.01027986, + "balance_loss_clip": 0.97924113, + "balance_loss_mlp": 1.0204221, + "epoch": 0.3631335297300547, + "flos": 19025976574080.0, + "grad_norm": 2.5204873560410235, + "language_loss": 0.78841627, + "learning_rate": 2.9446392469055257e-06, + "loss": 0.81015205, + "num_input_tokens_seen": 64868700, + "step": 3020, + "time_per_iteration": 2.7129383087158203 + }, + { + "auxiliary_loss_clip": 0.01118355, + "auxiliary_loss_mlp": 0.01034474, + "balance_loss_clip": 0.947088, + "balance_loss_mlp": 1.02643335, + "epoch": 0.3632537726206938, + "flos": 19536769929600.0, + "grad_norm": 2.785590306610927, + "language_loss": 0.80014235, + "learning_rate": 2.9439525683575745e-06, + "loss": 0.82167065, + "num_input_tokens_seen": 64887620, + "step": 3021, + "time_per_iteration": 2.6566808223724365 + }, + { + "auxiliary_loss_clip": 0.01198018, + "auxiliary_loss_mlp": 0.01027254, + "balance_loss_clip": 1.06253338, + "balance_loss_mlp": 1.01910615, + "epoch": 0.3633740155113329, + "flos": 21068611292160.0, + "grad_norm": 9.16345228242359, + "language_loss": 0.75296533, + "learning_rate": 2.9432657466127694e-06, + "loss": 0.77521801, + "num_input_tokens_seen": 64907190, + "step": 3022, + "time_per_iteration": 2.6325790882110596 + }, + { + "auxiliary_loss_clip": 0.01099266, + "auxiliary_loss_mlp": 0.01030896, + "balance_loss_clip": 0.90856469, + "balance_loss_mlp": 1.02311802, + "epoch": 0.36349425840197197, + "flos": 20298722158080.0, + "grad_norm": 1.725860490334845, + "language_loss": 0.76758462, + "learning_rate": 2.9425787817753007e-06, + "loss": 0.78888625, + "num_input_tokens_seen": 64925850, + "step": 3023, + "time_per_iteration": 2.702282428741455 + }, + { + "auxiliary_loss_clip": 0.01122715, + "auxiliary_loss_mlp": 0.0103163, + "balance_loss_clip": 0.94365788, + "balance_loss_mlp": 1.02322578, + "epoch": 0.3636145012926111, + "flos": 29716762331520.0, + "grad_norm": 1.5624934197373308, + "language_loss": 0.71925884, + "learning_rate": 2.94189167394938e-06, + "loss": 0.74080229, + "num_input_tokens_seen": 64948285, + "step": 3024, + "time_per_iteration": 2.740090847015381 + }, + { + "auxiliary_loss_clip": 0.01195656, + "auxiliary_loss_mlp": 0.01034531, + "balance_loss_clip": 1.06179023, + "balance_loss_mlp": 1.02585816, + "epoch": 0.3637347441832502, + "flos": 21431847576960.0, + "grad_norm": 2.5522742336024598, + "language_loss": 0.81079113, + "learning_rate": 2.941204423239241e-06, + "loss": 0.83309305, + "num_input_tokens_seen": 64967160, + "step": 3025, + "time_per_iteration": 2.704836845397949 + }, + { + "auxiliary_loss_clip": 0.01166855, + "auxiliary_loss_mlp": 0.01033696, + "balance_loss_clip": 1.02022791, + "balance_loss_mlp": 1.02530384, + "epoch": 0.36385498707388925, + "flos": 29533941083520.0, + "grad_norm": 1.7556729093334675, + "language_loss": 0.7591027, + "learning_rate": 2.9405170297491395e-06, + "loss": 0.78110814, + "num_input_tokens_seen": 64987155, + "step": 3026, + "time_per_iteration": 2.7150275707244873 + }, + { + "auxiliary_loss_clip": 0.01066109, + "auxiliary_loss_mlp": 0.00510643, + "balance_loss_clip": 0.86977905, + "balance_loss_mlp": 0.0, + "epoch": 0.36397522996452836, + "flos": 22236569925120.0, + "grad_norm": 2.055373179934766, + "language_loss": 0.8013978, + "learning_rate": 2.939829493583353e-06, + "loss": 0.81716526, + "num_input_tokens_seen": 65003800, + "step": 3027, + "time_per_iteration": 2.764575958251953 + }, + { + "auxiliary_loss_clip": 0.01111585, + "auxiliary_loss_mlp": 0.01025248, + "balance_loss_clip": 0.93783718, + "balance_loss_mlp": 1.01702845, + "epoch": 0.3640954728551674, + "flos": 21506505995520.0, + "grad_norm": 5.09001598814201, + "language_loss": 0.82909989, + "learning_rate": 2.939141814846179e-06, + "loss": 0.85046822, + "num_input_tokens_seen": 65021215, + "step": 3028, + "time_per_iteration": 2.724329710006714 + }, + { + "auxiliary_loss_clip": 0.01146342, + "auxiliary_loss_mlp": 0.01028383, + "balance_loss_clip": 0.98113322, + "balance_loss_mlp": 1.02059257, + "epoch": 0.3642157157458065, + "flos": 17712867081600.0, + "grad_norm": 1.6222917275733428, + "language_loss": 0.82455456, + "learning_rate": 2.938453993641938e-06, + "loss": 0.84630179, + "num_input_tokens_seen": 65039590, + "step": 3029, + "time_per_iteration": 2.7020528316497803 + }, + { + "auxiliary_loss_clip": 0.01144873, + "auxiliary_loss_mlp": 0.01029238, + "balance_loss_clip": 0.98531818, + "balance_loss_mlp": 1.02125657, + "epoch": 0.36433595863644563, + "flos": 17639537466240.0, + "grad_norm": 3.3340857968613977, + "language_loss": 0.70304823, + "learning_rate": 2.937766030074973e-06, + "loss": 0.72478938, + "num_input_tokens_seen": 65056845, + "step": 3030, + "time_per_iteration": 3.573742151260376 + }, + { + "auxiliary_loss_clip": 0.01126057, + "auxiliary_loss_mlp": 0.01026972, + "balance_loss_clip": 0.94515055, + "balance_loss_mlp": 1.01858521, + "epoch": 0.3644562015270847, + "flos": 26833279161600.0, + "grad_norm": 1.9629353434536359, + "language_loss": 0.82315141, + "learning_rate": 2.937077924249646e-06, + "loss": 0.84468174, + "num_input_tokens_seen": 65079435, + "step": 3031, + "time_per_iteration": 2.759737491607666 + }, + { + "auxiliary_loss_clip": 0.01151036, + "auxiliary_loss_mlp": 0.01028058, + "balance_loss_clip": 0.98141706, + "balance_loss_mlp": 1.01996374, + "epoch": 0.3645764444177238, + "flos": 14282715847680.0, + "grad_norm": 2.249701980275825, + "language_loss": 0.75373006, + "learning_rate": 2.9363896762703443e-06, + "loss": 0.77552104, + "num_input_tokens_seen": 65096500, + "step": 3032, + "time_per_iteration": 3.421656608581543 + }, + { + "auxiliary_loss_clip": 0.01192882, + "auxiliary_loss_mlp": 0.01031686, + "balance_loss_clip": 1.05872655, + "balance_loss_mlp": 1.02313864, + "epoch": 0.3646966873083629, + "flos": 20667489137280.0, + "grad_norm": 1.8930249293561432, + "language_loss": 0.84499729, + "learning_rate": 2.9357012862414725e-06, + "loss": 0.86724293, + "num_input_tokens_seen": 65115860, + "step": 3033, + "time_per_iteration": 3.608034610748291 + }, + { + "auxiliary_loss_clip": 0.01172285, + "auxiliary_loss_mlp": 0.01028188, + "balance_loss_clip": 1.02323437, + "balance_loss_mlp": 1.01989734, + "epoch": 0.36481693019900197, + "flos": 27782613665280.0, + "grad_norm": 2.427403655707818, + "language_loss": 0.71652657, + "learning_rate": 2.9350127542674593e-06, + "loss": 0.73853135, + "num_input_tokens_seen": 65138070, + "step": 3034, + "time_per_iteration": 2.807615280151367 + }, + { + "auxiliary_loss_clip": 0.01152859, + "auxiliary_loss_mlp": 0.01031094, + "balance_loss_clip": 0.98454964, + "balance_loss_mlp": 1.02243328, + "epoch": 0.3649371730896411, + "flos": 19712588025600.0, + "grad_norm": 2.276883400122572, + "language_loss": 0.7669332, + "learning_rate": 2.934324080452755e-06, + "loss": 0.78877276, + "num_input_tokens_seen": 65155860, + "step": 3035, + "time_per_iteration": 2.649848699569702 + }, + { + "auxiliary_loss_clip": 0.01111654, + "auxiliary_loss_mlp": 0.00511056, + "balance_loss_clip": 0.94079101, + "balance_loss_mlp": 0.0, + "epoch": 0.3650574159802802, + "flos": 24750496016640.0, + "grad_norm": 1.7576539941829024, + "language_loss": 0.78160733, + "learning_rate": 2.9336352649018307e-06, + "loss": 0.7978344, + "num_input_tokens_seen": 65175930, + "step": 3036, + "time_per_iteration": 2.752443313598633 + }, + { + "auxiliary_loss_clip": 0.01146729, + "auxiliary_loss_mlp": 0.01036383, + "balance_loss_clip": 0.98274946, + "balance_loss_mlp": 1.02803862, + "epoch": 0.36517765887091924, + "flos": 32853487363200.0, + "grad_norm": 1.9090493197947478, + "language_loss": 0.70006859, + "learning_rate": 2.9329463077191783e-06, + "loss": 0.72189969, + "num_input_tokens_seen": 65199305, + "step": 3037, + "time_per_iteration": 2.8293707370758057 + }, + { + "auxiliary_loss_clip": 0.01098006, + "auxiliary_loss_mlp": 0.01030484, + "balance_loss_clip": 0.90531993, + "balance_loss_mlp": 1.02213931, + "epoch": 0.36529790176155835, + "flos": 20120318282880.0, + "grad_norm": 3.6943378593561578, + "language_loss": 0.64659262, + "learning_rate": 2.9322572090093135e-06, + "loss": 0.6678775, + "num_input_tokens_seen": 65218010, + "step": 3038, + "time_per_iteration": 2.734482526779175 + }, + { + "auxiliary_loss_clip": 0.01098245, + "auxiliary_loss_mlp": 0.01031669, + "balance_loss_clip": 0.90320182, + "balance_loss_mlp": 1.02244782, + "epoch": 0.36541814465219746, + "flos": 17639573379840.0, + "grad_norm": 2.736022759808145, + "language_loss": 0.76265776, + "learning_rate": 2.9315679688767713e-06, + "loss": 0.78395689, + "num_input_tokens_seen": 65236020, + "step": 3039, + "time_per_iteration": 3.704946517944336 + }, + { + "auxiliary_loss_clip": 0.01142907, + "auxiliary_loss_mlp": 0.01028986, + "balance_loss_clip": 0.98175716, + "balance_loss_mlp": 1.02107906, + "epoch": 0.3655383875428365, + "flos": 22674356887680.0, + "grad_norm": 1.6170021336090807, + "language_loss": 0.65974736, + "learning_rate": 2.9308785874261085e-06, + "loss": 0.68146628, + "num_input_tokens_seen": 65256210, + "step": 3040, + "time_per_iteration": 2.7236058712005615 + }, + { + "auxiliary_loss_clip": 0.01196501, + "auxiliary_loss_mlp": 0.01029001, + "balance_loss_clip": 1.06268525, + "balance_loss_mlp": 1.0200845, + "epoch": 0.36565863043347563, + "flos": 21981173247360.0, + "grad_norm": 1.7608202644261166, + "language_loss": 0.814363, + "learning_rate": 2.9301890647619045e-06, + "loss": 0.83661807, + "num_input_tokens_seen": 65275505, + "step": 3041, + "time_per_iteration": 2.6359691619873047 + }, + { + "auxiliary_loss_clip": 0.01155596, + "auxiliary_loss_mlp": 0.01032189, + "balance_loss_clip": 0.98534584, + "balance_loss_mlp": 1.02350414, + "epoch": 0.36577887332411474, + "flos": 24827632473600.0, + "grad_norm": 2.2234286049108443, + "language_loss": 0.7998755, + "learning_rate": 2.929499400988759e-06, + "loss": 0.82175332, + "num_input_tokens_seen": 65296665, + "step": 3042, + "time_per_iteration": 2.701127529144287 + }, + { + "auxiliary_loss_clip": 0.01167156, + "auxiliary_loss_mlp": 0.01031786, + "balance_loss_clip": 1.02151191, + "balance_loss_mlp": 1.02322721, + "epoch": 0.3658991162147538, + "flos": 28293191539200.0, + "grad_norm": 1.73128575951989, + "language_loss": 0.65244973, + "learning_rate": 2.9288095962112927e-06, + "loss": 0.67443919, + "num_input_tokens_seen": 65317370, + "step": 3043, + "time_per_iteration": 2.7231693267822266 + }, + { + "auxiliary_loss_clip": 0.01192732, + "auxiliary_loss_mlp": 0.01025722, + "balance_loss_clip": 1.06027889, + "balance_loss_mlp": 1.0178448, + "epoch": 0.3660193591053929, + "flos": 17785550252160.0, + "grad_norm": 1.9172161710681124, + "language_loss": 0.85254407, + "learning_rate": 2.9281196505341503e-06, + "loss": 0.87472856, + "num_input_tokens_seen": 65334540, + "step": 3044, + "time_per_iteration": 2.617769479751587 + }, + { + "auxiliary_loss_clip": 0.01092941, + "auxiliary_loss_mlp": 0.00509821, + "balance_loss_clip": 0.90581906, + "balance_loss_mlp": 0.0, + "epoch": 0.36613960199603196, + "flos": 10342776839040.0, + "grad_norm": 4.012082511458121, + "language_loss": 0.78720665, + "learning_rate": 2.9274295640619946e-06, + "loss": 0.80323428, + "num_input_tokens_seen": 65351670, + "step": 3045, + "time_per_iteration": 2.727470874786377 + }, + { + "auxiliary_loss_clip": 0.01120723, + "auxiliary_loss_mlp": 0.01025764, + "balance_loss_clip": 0.94134653, + "balance_loss_mlp": 1.01804566, + "epoch": 0.36625984488667107, + "flos": 19755609540480.0, + "grad_norm": 6.421099562776929, + "language_loss": 0.78987098, + "learning_rate": 2.9267393368995103e-06, + "loss": 0.81133592, + "num_input_tokens_seen": 65370900, + "step": 3046, + "time_per_iteration": 2.736820697784424 + }, + { + "auxiliary_loss_clip": 0.01195081, + "auxiliary_loss_mlp": 0.01032494, + "balance_loss_clip": 1.06102049, + "balance_loss_mlp": 1.02422059, + "epoch": 0.3663800877773102, + "flos": 17674262939520.0, + "grad_norm": 2.737055827083298, + "language_loss": 0.74521852, + "learning_rate": 2.926048969151407e-06, + "loss": 0.76749426, + "num_input_tokens_seen": 65388185, + "step": 3047, + "time_per_iteration": 2.575868606567383 + }, + { + "auxiliary_loss_clip": 0.0109084, + "auxiliary_loss_mlp": 0.01030675, + "balance_loss_clip": 0.90624982, + "balance_loss_mlp": 1.02284908, + "epoch": 0.36650033066794924, + "flos": 20303606407680.0, + "grad_norm": 1.7657230251058533, + "language_loss": 0.68411863, + "learning_rate": 2.92535846092241e-06, + "loss": 0.70533383, + "num_input_tokens_seen": 65407200, + "step": 3048, + "time_per_iteration": 2.799201250076294 + }, + { + "auxiliary_loss_clip": 0.01148112, + "auxiliary_loss_mlp": 0.01028185, + "balance_loss_clip": 0.98413718, + "balance_loss_mlp": 1.01981616, + "epoch": 0.36662057355858835, + "flos": 24716237420160.0, + "grad_norm": 1.8844510274351896, + "language_loss": 0.82753676, + "learning_rate": 2.9246678123172704e-06, + "loss": 0.84929973, + "num_input_tokens_seen": 65427290, + "step": 3049, + "time_per_iteration": 2.722107172012329 + }, + { + "auxiliary_loss_clip": 0.01194382, + "auxiliary_loss_mlp": 0.01031986, + "balance_loss_clip": 1.05961418, + "balance_loss_mlp": 1.02376032, + "epoch": 0.36674081644922746, + "flos": 12385267902720.0, + "grad_norm": 9.271593243463169, + "language_loss": 0.73957658, + "learning_rate": 2.9239770234407596e-06, + "loss": 0.76184022, + "num_input_tokens_seen": 65445595, + "step": 3050, + "time_per_iteration": 2.6239817142486572 + }, + { + "auxiliary_loss_clip": 0.01168067, + "auxiliary_loss_mlp": 0.0102399, + "balance_loss_clip": 1.01947188, + "balance_loss_mlp": 1.01617026, + "epoch": 0.3668610593398665, + "flos": 21105922544640.0, + "grad_norm": 1.9000557432966971, + "language_loss": 0.68589503, + "learning_rate": 2.9232860943976686e-06, + "loss": 0.70781565, + "num_input_tokens_seen": 65466330, + "step": 3051, + "time_per_iteration": 2.6844260692596436 + }, + { + "auxiliary_loss_clip": 0.01145471, + "auxiliary_loss_mlp": 0.01030049, + "balance_loss_clip": 0.98440111, + "balance_loss_mlp": 1.02206802, + "epoch": 0.3669813022305056, + "flos": 26758082039040.0, + "grad_norm": 1.7359742592897192, + "language_loss": 0.84379697, + "learning_rate": 2.9225950252928115e-06, + "loss": 0.86555213, + "num_input_tokens_seen": 65487180, + "step": 3052, + "time_per_iteration": 2.730717420578003 + }, + { + "auxiliary_loss_clip": 0.01167848, + "auxiliary_loss_mlp": 0.01033075, + "balance_loss_clip": 1.02162933, + "balance_loss_mlp": 1.02419376, + "epoch": 0.36710154512114473, + "flos": 19099521671040.0, + "grad_norm": 3.0611820943453862, + "language_loss": 0.81721628, + "learning_rate": 2.9219038162310217e-06, + "loss": 0.83922553, + "num_input_tokens_seen": 65505380, + "step": 3053, + "time_per_iteration": 2.634885787963867 + }, + { + "auxiliary_loss_clip": 0.01045835, + "auxiliary_loss_mlp": 0.0051044, + "balance_loss_clip": 0.8253994, + "balance_loss_mlp": 0.0, + "epoch": 0.3672217880117838, + "flos": 20812029465600.0, + "grad_norm": 2.506990692614765, + "language_loss": 0.82330036, + "learning_rate": 2.921212467317157e-06, + "loss": 0.83886313, + "num_input_tokens_seen": 65524825, + "step": 3054, + "time_per_iteration": 2.8126323223114014 + }, + { + "auxiliary_loss_clip": 0.01129274, + "auxiliary_loss_mlp": 0.01032325, + "balance_loss_clip": 0.97572863, + "balance_loss_mlp": 1.0231216, + "epoch": 0.3673420309024229, + "flos": 13590394133760.0, + "grad_norm": 2.0381229548907256, + "language_loss": 0.80262959, + "learning_rate": 2.920520978656093e-06, + "loss": 0.82424551, + "num_input_tokens_seen": 65541790, + "step": 3055, + "time_per_iteration": 2.6187148094177246 + }, + { + "auxiliary_loss_clip": 0.01189606, + "auxiliary_loss_mlp": 0.00510025, + "balance_loss_clip": 1.05810738, + "balance_loss_mlp": 0.0, + "epoch": 0.367462273793062, + "flos": 28986877969920.0, + "grad_norm": 2.0669020971714103, + "language_loss": 0.77558744, + "learning_rate": 2.919829350352729e-06, + "loss": 0.7925837, + "num_input_tokens_seen": 65563395, + "step": 3056, + "time_per_iteration": 3.616201400756836 + }, + { + "auxiliary_loss_clip": 0.0109836, + "auxiliary_loss_mlp": 0.01008808, + "balance_loss_clip": 1.03286374, + "balance_loss_mlp": 1.00717449, + "epoch": 0.36758251668370107, + "flos": 62643148346880.0, + "grad_norm": 0.7710052792018247, + "language_loss": 0.60032082, + "learning_rate": 2.919137582511983e-06, + "loss": 0.62139249, + "num_input_tokens_seen": 65619835, + "step": 3057, + "time_per_iteration": 3.11606764793396 + }, + { + "auxiliary_loss_clip": 0.01125947, + "auxiliary_loss_mlp": 0.0103247, + "balance_loss_clip": 0.94581318, + "balance_loss_mlp": 1.02409577, + "epoch": 0.3677027595743402, + "flos": 12713886455040.0, + "grad_norm": 2.114498699634309, + "language_loss": 0.63877314, + "learning_rate": 2.918445675238797e-06, + "loss": 0.66035736, + "num_input_tokens_seen": 65636760, + "step": 3058, + "time_per_iteration": 2.664320230484009 + }, + { + "auxiliary_loss_clip": 0.01190829, + "auxiliary_loss_mlp": 0.01028023, + "balance_loss_clip": 1.05727029, + "balance_loss_mlp": 1.01961327, + "epoch": 0.36782300246497923, + "flos": 25046579825280.0, + "grad_norm": 1.7716454956575238, + "language_loss": 0.69794208, + "learning_rate": 2.917753628638132e-06, + "loss": 0.72013068, + "num_input_tokens_seen": 65657065, + "step": 3059, + "time_per_iteration": 3.573431968688965 + }, + { + "auxiliary_loss_clip": 0.01146007, + "auxiliary_loss_mlp": 0.0103034, + "balance_loss_clip": 0.98453665, + "balance_loss_mlp": 1.02212048, + "epoch": 0.36794324535561834, + "flos": 17419512706560.0, + "grad_norm": 2.305471411144305, + "language_loss": 0.70282847, + "learning_rate": 2.9170614428149716e-06, + "loss": 0.72459197, + "num_input_tokens_seen": 65675400, + "step": 3060, + "time_per_iteration": 3.6470119953155518 + }, + { + "auxiliary_loss_clip": 0.01116123, + "auxiliary_loss_mlp": 0.01027811, + "balance_loss_clip": 0.94372934, + "balance_loss_mlp": 1.01953781, + "epoch": 0.36806348824625745, + "flos": 24089128848000.0, + "grad_norm": 2.2310245794780714, + "language_loss": 0.86865008, + "learning_rate": 2.9163691178743195e-06, + "loss": 0.89008939, + "num_input_tokens_seen": 65694050, + "step": 3061, + "time_per_iteration": 2.7129225730895996 + }, + { + "auxiliary_loss_clip": 0.01163946, + "auxiliary_loss_mlp": 0.01034108, + "balance_loss_clip": 1.01958323, + "balance_loss_mlp": 1.0260613, + "epoch": 0.3681837311368965, + "flos": 20521871400960.0, + "grad_norm": 2.1593808557214076, + "language_loss": 0.77281642, + "learning_rate": 2.9156766539212006e-06, + "loss": 0.79479694, + "num_input_tokens_seen": 65711695, + "step": 3062, + "time_per_iteration": 2.6835944652557373 + }, + { + "auxiliary_loss_clip": 0.01173549, + "auxiliary_loss_mlp": 0.01035166, + "balance_loss_clip": 1.02084136, + "balance_loss_mlp": 1.0273931, + "epoch": 0.3683039740275356, + "flos": 21466644877440.0, + "grad_norm": 1.968880990200166, + "language_loss": 0.71893239, + "learning_rate": 2.9149840510606614e-06, + "loss": 0.74101961, + "num_input_tokens_seen": 65730350, + "step": 3063, + "time_per_iteration": 2.6389834880828857 + }, + { + "auxiliary_loss_clip": 0.01070984, + "auxiliary_loss_mlp": 0.0050165, + "balance_loss_clip": 0.99053621, + "balance_loss_mlp": 0.0, + "epoch": 0.36842421691817473, + "flos": 70380999987840.0, + "grad_norm": 1.0209029283636304, + "language_loss": 0.64137095, + "learning_rate": 2.914291309397769e-06, + "loss": 0.65709728, + "num_input_tokens_seen": 65787820, + "step": 3064, + "time_per_iteration": 3.2876412868499756 + }, + { + "auxiliary_loss_clip": 0.01059994, + "auxiliary_loss_mlp": 0.01035222, + "balance_loss_clip": 0.86150312, + "balance_loss_mlp": 1.02673459, + "epoch": 0.3685444598088138, + "flos": 23331378510720.0, + "grad_norm": 1.903849652251133, + "language_loss": 0.78339434, + "learning_rate": 2.9135984290376117e-06, + "loss": 0.8043465, + "num_input_tokens_seen": 65806685, + "step": 3065, + "time_per_iteration": 3.7241642475128174 + }, + { + "auxiliary_loss_clip": 0.01071982, + "auxiliary_loss_mlp": 0.01023829, + "balance_loss_clip": 0.86364758, + "balance_loss_mlp": 1.01545441, + "epoch": 0.3686647026994529, + "flos": 23070271570560.0, + "grad_norm": 1.871662325928617, + "language_loss": 0.82479936, + "learning_rate": 2.9129054100853e-06, + "loss": 0.84575742, + "num_input_tokens_seen": 65825525, + "step": 3066, + "time_per_iteration": 2.7380611896514893 + }, + { + "auxiliary_loss_clip": 0.01143759, + "auxiliary_loss_mlp": 0.01029677, + "balance_loss_clip": 0.9802891, + "balance_loss_mlp": 1.02124882, + "epoch": 0.368784945590092, + "flos": 25119909440640.0, + "grad_norm": 2.502501414691623, + "language_loss": 0.7618587, + "learning_rate": 2.912212252645963e-06, + "loss": 0.78359306, + "num_input_tokens_seen": 65848110, + "step": 3067, + "time_per_iteration": 2.757411003112793 + }, + { + "auxiliary_loss_clip": 0.0117303, + "auxiliary_loss_mlp": 0.01032031, + "balance_loss_clip": 1.0202446, + "balance_loss_mlp": 1.02270341, + "epoch": 0.36890518848073106, + "flos": 18442284566400.0, + "grad_norm": 3.8233646199003015, + "language_loss": 0.76480681, + "learning_rate": 2.9115189568247523e-06, + "loss": 0.78685743, + "num_input_tokens_seen": 65865670, + "step": 3068, + "time_per_iteration": 2.610055446624756 + }, + { + "auxiliary_loss_clip": 0.01087794, + "auxiliary_loss_mlp": 0.01033939, + "balance_loss_clip": 0.90699112, + "balance_loss_mlp": 1.02590454, + "epoch": 0.36902543137137017, + "flos": 16362446336640.0, + "grad_norm": 2.6721868049761586, + "language_loss": 0.91942149, + "learning_rate": 2.910825522726841e-06, + "loss": 0.94063884, + "num_input_tokens_seen": 65883195, + "step": 3069, + "time_per_iteration": 2.7506775856018066 + }, + { + "auxiliary_loss_clip": 0.01092795, + "auxiliary_loss_mlp": 0.01029564, + "balance_loss_clip": 0.90218461, + "balance_loss_mlp": 1.02146423, + "epoch": 0.3691456742620093, + "flos": 12275596702080.0, + "grad_norm": 2.241073824915194, + "language_loss": 0.77321386, + "learning_rate": 2.9101319504574215e-06, + "loss": 0.79443753, + "num_input_tokens_seen": 65899635, + "step": 3070, + "time_per_iteration": 2.7352404594421387 + }, + { + "auxiliary_loss_clip": 0.01160064, + "auxiliary_loss_mlp": 0.01035414, + "balance_loss_clip": 1.01671624, + "balance_loss_mlp": 1.02658081, + "epoch": 0.36926591715264834, + "flos": 17786412178560.0, + "grad_norm": 2.886082895485909, + "language_loss": 0.76352704, + "learning_rate": 2.909438240121709e-06, + "loss": 0.78548187, + "num_input_tokens_seen": 65919910, + "step": 3071, + "time_per_iteration": 2.653346538543701 + }, + { + "auxiliary_loss_clip": 0.01140527, + "auxiliary_loss_mlp": 0.01032905, + "balance_loss_clip": 0.98298228, + "balance_loss_mlp": 1.02427447, + "epoch": 0.36938616004328745, + "flos": 28948309741440.0, + "grad_norm": 1.6669031927273394, + "language_loss": 0.70536089, + "learning_rate": 2.908744391824939e-06, + "loss": 0.72709525, + "num_input_tokens_seen": 65940930, + "step": 3072, + "time_per_iteration": 2.7416629791259766 + }, + { + "auxiliary_loss_clip": 0.01069511, + "auxiliary_loss_mlp": 0.01030519, + "balance_loss_clip": 0.86127746, + "balance_loss_mlp": 1.02214456, + "epoch": 0.36950640293392656, + "flos": 29205394358400.0, + "grad_norm": 1.7603751923783435, + "language_loss": 0.79291058, + "learning_rate": 2.908050405672367e-06, + "loss": 0.81391084, + "num_input_tokens_seen": 65960475, + "step": 3073, + "time_per_iteration": 2.8387532234191895 + }, + { + "auxiliary_loss_clip": 0.01149097, + "auxiliary_loss_mlp": 0.01033241, + "balance_loss_clip": 0.97781146, + "balance_loss_mlp": 1.0245477, + "epoch": 0.3696266458245656, + "flos": 24827776128000.0, + "grad_norm": 2.0812029025440233, + "language_loss": 0.79387677, + "learning_rate": 2.9073562817692703e-06, + "loss": 0.81570011, + "num_input_tokens_seen": 65979160, + "step": 3074, + "time_per_iteration": 2.719740152359009 + }, + { + "auxiliary_loss_clip": 0.01015063, + "auxiliary_loss_mlp": 0.01006136, + "balance_loss_clip": 0.87700546, + "balance_loss_mlp": 1.00465155, + "epoch": 0.3697468887152047, + "flos": 59887257264000.0, + "grad_norm": 0.7290130947964488, + "language_loss": 0.56555784, + "learning_rate": 2.9066620202209468e-06, + "loss": 0.58576983, + "num_input_tokens_seen": 66041650, + "step": 3075, + "time_per_iteration": 3.262483835220337 + }, + { + "auxiliary_loss_clip": 0.01109607, + "auxiliary_loss_mlp": 0.01031293, + "balance_loss_clip": 0.93932247, + "balance_loss_mlp": 1.02371144, + "epoch": 0.3698671316058438, + "flos": 26137581569280.0, + "grad_norm": 2.2067059647593825, + "language_loss": 0.77919328, + "learning_rate": 2.905967621132716e-06, + "loss": 0.80060226, + "num_input_tokens_seen": 66059260, + "step": 3076, + "time_per_iteration": 2.7449753284454346 + }, + { + "auxiliary_loss_clip": 0.01148799, + "auxiliary_loss_mlp": 0.01028348, + "balance_loss_clip": 0.98156828, + "balance_loss_mlp": 1.01983082, + "epoch": 0.3699873744964829, + "flos": 24607464059520.0, + "grad_norm": 3.265963071775153, + "language_loss": 0.75152504, + "learning_rate": 2.9052730846099172e-06, + "loss": 0.77329654, + "num_input_tokens_seen": 66080605, + "step": 3077, + "time_per_iteration": 2.70061993598938 + }, + { + "auxiliary_loss_clip": 0.010484, + "auxiliary_loss_mlp": 0.01000764, + "balance_loss_clip": 0.94872713, + "balance_loss_mlp": 0.9993332, + "epoch": 0.370107617387122, + "flos": 64885340050560.0, + "grad_norm": 0.875015716771339, + "language_loss": 0.60910296, + "learning_rate": 2.9045784107579123e-06, + "loss": 0.62959456, + "num_input_tokens_seen": 66140710, + "step": 3078, + "time_per_iteration": 3.2708098888397217 + }, + { + "auxiliary_loss_clip": 0.01191193, + "auxiliary_loss_mlp": 0.01036162, + "balance_loss_clip": 1.05809808, + "balance_loss_mlp": 1.02815688, + "epoch": 0.37022786027776106, + "flos": 15961683317760.0, + "grad_norm": 1.9126172869478975, + "language_loss": 0.67295861, + "learning_rate": 2.9038835996820807e-06, + "loss": 0.69523215, + "num_input_tokens_seen": 66158320, + "step": 3079, + "time_per_iteration": 2.68973445892334 + }, + { + "auxiliary_loss_clip": 0.01123238, + "auxiliary_loss_mlp": 0.01029985, + "balance_loss_clip": 0.93889606, + "balance_loss_mlp": 1.02159262, + "epoch": 0.37034810316840017, + "flos": 18546927863040.0, + "grad_norm": 1.9521038408302787, + "language_loss": 0.79659742, + "learning_rate": 2.903188651487826e-06, + "loss": 0.81812966, + "num_input_tokens_seen": 66176875, + "step": 3080, + "time_per_iteration": 3.0080833435058594 + }, + { + "auxiliary_loss_clip": 0.01172817, + "auxiliary_loss_mlp": 0.01028791, + "balance_loss_clip": 1.02195501, + "balance_loss_mlp": 1.02104235, + "epoch": 0.3704683460590393, + "flos": 17821927751040.0, + "grad_norm": 2.335131948611504, + "language_loss": 0.86951959, + "learning_rate": 2.902493566280571e-06, + "loss": 0.89153564, + "num_input_tokens_seen": 66194980, + "step": 3081, + "time_per_iteration": 2.6339378356933594 + }, + { + "auxiliary_loss_clip": 0.01143962, + "auxiliary_loss_mlp": 0.01035156, + "balance_loss_clip": 0.98189497, + "balance_loss_mlp": 1.0260545, + "epoch": 0.37058858894967833, + "flos": 14134081368960.0, + "grad_norm": 2.1813798540300753, + "language_loss": 0.81443393, + "learning_rate": 2.9017983441657595e-06, + "loss": 0.83622509, + "num_input_tokens_seen": 66212310, + "step": 3082, + "time_per_iteration": 3.553403854370117 + }, + { + "auxiliary_loss_clip": 0.01095478, + "auxiliary_loss_mlp": 0.01028494, + "balance_loss_clip": 0.90083706, + "balance_loss_mlp": 1.02062011, + "epoch": 0.37070883184031744, + "flos": 13954492344960.0, + "grad_norm": 2.6212817368255314, + "language_loss": 0.75487709, + "learning_rate": 2.9011029852488564e-06, + "loss": 0.77611679, + "num_input_tokens_seen": 66229545, + "step": 3083, + "time_per_iteration": 2.768341302871704 + }, + { + "auxiliary_loss_clip": 0.01088226, + "auxiliary_loss_mlp": 0.01003594, + "balance_loss_clip": 1.02354789, + "balance_loss_mlp": 1.00234187, + "epoch": 0.37082907473095655, + "flos": 52315419306240.0, + "grad_norm": 0.9994952658877642, + "language_loss": 0.62397408, + "learning_rate": 2.9004074896353465e-06, + "loss": 0.64489233, + "num_input_tokens_seen": 66283545, + "step": 3084, + "time_per_iteration": 4.005073547363281 + }, + { + "auxiliary_loss_clip": 0.01192592, + "auxiliary_loss_mlp": 0.0102921, + "balance_loss_clip": 1.06230807, + "balance_loss_mlp": 1.02102077, + "epoch": 0.3709493176215956, + "flos": 15998096730240.0, + "grad_norm": 1.7615292175360713, + "language_loss": 0.8108412, + "learning_rate": 2.8997118574307362e-06, + "loss": 0.83305925, + "num_input_tokens_seen": 66300500, + "step": 3085, + "time_per_iteration": 3.509791135787964 + }, + { + "auxiliary_loss_clip": 0.01128761, + "auxiliary_loss_mlp": 0.01034314, + "balance_loss_clip": 0.94484353, + "balance_loss_mlp": 1.02604055, + "epoch": 0.3710695605122347, + "flos": 20959837931520.0, + "grad_norm": 2.0038502438331323, + "language_loss": 0.74771535, + "learning_rate": 2.899016088740553e-06, + "loss": 0.76934606, + "num_input_tokens_seen": 66318610, + "step": 3086, + "time_per_iteration": 2.7324702739715576 + }, + { + "auxiliary_loss_clip": 0.01096265, + "auxiliary_loss_mlp": 0.01034804, + "balance_loss_clip": 0.90317488, + "balance_loss_mlp": 1.02642322, + "epoch": 0.37118980340287383, + "flos": 14355578586240.0, + "grad_norm": 1.9087684510979437, + "language_loss": 0.79473388, + "learning_rate": 2.898320183670344e-06, + "loss": 0.81604457, + "num_input_tokens_seen": 66336025, + "step": 3087, + "time_per_iteration": 2.7262215614318848 + }, + { + "auxiliary_loss_clip": 0.01092743, + "auxiliary_loss_mlp": 0.01028254, + "balance_loss_clip": 0.90450978, + "balance_loss_mlp": 1.02018642, + "epoch": 0.3713100462935129, + "flos": 25885381201920.0, + "grad_norm": 2.436908037122517, + "language_loss": 0.8895489, + "learning_rate": 2.8976241423256767e-06, + "loss": 0.91075897, + "num_input_tokens_seen": 66356120, + "step": 3088, + "time_per_iteration": 2.7284669876098633 + }, + { + "auxiliary_loss_clip": 0.01139935, + "auxiliary_loss_mlp": 0.01025082, + "balance_loss_clip": 0.98153591, + "balance_loss_mlp": 1.01738715, + "epoch": 0.371430289184152, + "flos": 30518934814080.0, + "grad_norm": 2.3775944726718885, + "language_loss": 0.68119788, + "learning_rate": 2.896927964812142e-06, + "loss": 0.70284802, + "num_input_tokens_seen": 66376685, + "step": 3089, + "time_per_iteration": 2.7775044441223145 + }, + { + "auxiliary_loss_clip": 0.0114034, + "auxiliary_loss_mlp": 0.01029764, + "balance_loss_clip": 0.98299325, + "balance_loss_mlp": 1.02144885, + "epoch": 0.37155053207479105, + "flos": 15742233175680.0, + "grad_norm": 3.154260904981123, + "language_loss": 0.75259328, + "learning_rate": 2.8962316512353465e-06, + "loss": 0.77429426, + "num_input_tokens_seen": 66394230, + "step": 3090, + "time_per_iteration": 2.6168222427368164 + }, + { + "auxiliary_loss_clip": 0.01071109, + "auxiliary_loss_mlp": 0.01026584, + "balance_loss_clip": 0.86392623, + "balance_loss_mlp": 1.01843572, + "epoch": 0.37167077496543016, + "flos": 23404061681280.0, + "grad_norm": 1.6636347996587884, + "language_loss": 0.74983013, + "learning_rate": 2.8955352017009233e-06, + "loss": 0.77080703, + "num_input_tokens_seen": 66413475, + "step": 3091, + "time_per_iteration": 3.7649128437042236 + }, + { + "auxiliary_loss_clip": 0.01139511, + "auxiliary_loss_mlp": 0.01026043, + "balance_loss_clip": 0.98159671, + "balance_loss_mlp": 1.01802015, + "epoch": 0.3717910178560693, + "flos": 22088653718400.0, + "grad_norm": 3.061642520195137, + "language_loss": 0.77376604, + "learning_rate": 2.8948386163145212e-06, + "loss": 0.7954216, + "num_input_tokens_seen": 66432685, + "step": 3092, + "time_per_iteration": 2.678791046142578 + }, + { + "auxiliary_loss_clip": 0.01173159, + "auxiliary_loss_mlp": 0.01026644, + "balance_loss_clip": 1.02232885, + "balance_loss_mlp": 1.01887131, + "epoch": 0.3719112607467083, + "flos": 26939969533440.0, + "grad_norm": 1.9058671714239142, + "language_loss": 0.79672557, + "learning_rate": 2.8941418951818135e-06, + "loss": 0.81872356, + "num_input_tokens_seen": 66452245, + "step": 3093, + "time_per_iteration": 2.7221519947052 + }, + { + "auxiliary_loss_clip": 0.01119746, + "auxiliary_loss_mlp": 0.01030441, + "balance_loss_clip": 0.94236231, + "balance_loss_mlp": 1.02244806, + "epoch": 0.37203150363734744, + "flos": 12166500119040.0, + "grad_norm": 2.1500647390424876, + "language_loss": 0.71463299, + "learning_rate": 2.8934450384084903e-06, + "loss": 0.73613483, + "num_input_tokens_seen": 66469760, + "step": 3094, + "time_per_iteration": 2.702927350997925 + }, + { + "auxiliary_loss_clip": 0.01134046, + "auxiliary_loss_mlp": 0.01031975, + "balance_loss_clip": 0.97753537, + "balance_loss_mlp": 1.02320671, + "epoch": 0.37215174652798655, + "flos": 23697595624320.0, + "grad_norm": 2.002773003952055, + "language_loss": 0.69597006, + "learning_rate": 2.8927480461002653e-06, + "loss": 0.71763027, + "num_input_tokens_seen": 66489730, + "step": 3095, + "time_per_iteration": 2.755084753036499 + }, + { + "auxiliary_loss_clip": 0.01141556, + "auxiliary_loss_mlp": 0.01035576, + "balance_loss_clip": 0.97834063, + "balance_loss_mlp": 1.02668273, + "epoch": 0.3722719894186256, + "flos": 17887751424000.0, + "grad_norm": 3.1608088355425017, + "language_loss": 0.86149812, + "learning_rate": 2.892050918362872e-06, + "loss": 0.88326949, + "num_input_tokens_seen": 66504785, + "step": 3096, + "time_per_iteration": 2.803983688354492 + }, + { + "auxiliary_loss_clip": 0.00969602, + "auxiliary_loss_mlp": 0.01003907, + "balance_loss_clip": 0.80548513, + "balance_loss_mlp": 1.0025897, + "epoch": 0.3723922323092647, + "flos": 62419891363200.0, + "grad_norm": 0.8653352368798477, + "language_loss": 0.55881405, + "learning_rate": 2.8913536553020626e-06, + "loss": 0.57854915, + "num_input_tokens_seen": 66558840, + "step": 3097, + "time_per_iteration": 3.5223047733306885 + }, + { + "auxiliary_loss_clip": 0.01084271, + "auxiliary_loss_mlp": 0.01032117, + "balance_loss_clip": 0.89959168, + "balance_loss_mlp": 1.02296197, + "epoch": 0.3725124751999038, + "flos": 23039747988480.0, + "grad_norm": 3.097477966057115, + "language_loss": 0.84596604, + "learning_rate": 2.8906562570236137e-06, + "loss": 0.86712992, + "num_input_tokens_seen": 66576750, + "step": 3098, + "time_per_iteration": 3.058774471282959 + }, + { + "auxiliary_loss_clip": 0.01067451, + "auxiliary_loss_mlp": 0.01029538, + "balance_loss_clip": 0.86077803, + "balance_loss_mlp": 1.02167594, + "epoch": 0.3726327180905429, + "flos": 20920551431040.0, + "grad_norm": 1.7370960090074739, + "language_loss": 0.76345122, + "learning_rate": 2.889958723633318e-06, + "loss": 0.78442103, + "num_input_tokens_seen": 66595690, + "step": 3099, + "time_per_iteration": 2.8691792488098145 + }, + { + "auxiliary_loss_clip": 0.01123086, + "auxiliary_loss_mlp": 0.01033122, + "balance_loss_clip": 0.94318324, + "balance_loss_mlp": 1.02444911, + "epoch": 0.372752960981182, + "flos": 30592156688640.0, + "grad_norm": 1.8253526434848486, + "language_loss": 0.74028993, + "learning_rate": 2.889261055236992e-06, + "loss": 0.76185203, + "num_input_tokens_seen": 66617905, + "step": 3100, + "time_per_iteration": 2.762697458267212 + }, + { + "auxiliary_loss_clip": 0.01138757, + "auxiliary_loss_mlp": 0.01029724, + "balance_loss_clip": 0.98234379, + "balance_loss_mlp": 1.02138495, + "epoch": 0.3728732038718211, + "flos": 25116749043840.0, + "grad_norm": 1.8161016859654984, + "language_loss": 0.8265636, + "learning_rate": 2.8885632519404704e-06, + "loss": 0.84824836, + "num_input_tokens_seen": 66638175, + "step": 3101, + "time_per_iteration": 2.7808351516723633 + }, + { + "auxiliary_loss_clip": 0.01138747, + "auxiliary_loss_mlp": 0.01027759, + "balance_loss_clip": 0.97963524, + "balance_loss_mlp": 1.01912868, + "epoch": 0.37299344676246016, + "flos": 25302048330240.0, + "grad_norm": 3.547524514248087, + "language_loss": 0.75746459, + "learning_rate": 2.8878653138496107e-06, + "loss": 0.77912962, + "num_input_tokens_seen": 66658670, + "step": 3102, + "time_per_iteration": 2.7071218490600586 + }, + { + "auxiliary_loss_clip": 0.01065105, + "auxiliary_loss_mlp": 0.01033467, + "balance_loss_clip": 0.85974491, + "balance_loss_mlp": 1.02538455, + "epoch": 0.37311368965309927, + "flos": 23842531002240.0, + "grad_norm": 3.0000480091371404, + "language_loss": 0.76884639, + "learning_rate": 2.8871672410702878e-06, + "loss": 0.78983212, + "num_input_tokens_seen": 66676030, + "step": 3103, + "time_per_iteration": 2.800814628601074 + }, + { + "auxiliary_loss_clip": 0.01126572, + "auxiliary_loss_mlp": 0.01029303, + "balance_loss_clip": 0.94125962, + "balance_loss_mlp": 1.02049947, + "epoch": 0.3732339325437384, + "flos": 25811943845760.0, + "grad_norm": 1.797476593633957, + "language_loss": 0.81821918, + "learning_rate": 2.8864690337084008e-06, + "loss": 0.83977795, + "num_input_tokens_seen": 66695305, + "step": 3104, + "time_per_iteration": 2.790480852127075 + }, + { + "auxiliary_loss_clip": 0.01161224, + "auxiliary_loss_mlp": 0.01028232, + "balance_loss_clip": 1.01787543, + "balance_loss_mlp": 1.0197022, + "epoch": 0.37335417543437743, + "flos": 26208433146240.0, + "grad_norm": 1.8603720806613067, + "language_loss": 0.77845371, + "learning_rate": 2.885770691869866e-06, + "loss": 0.80034822, + "num_input_tokens_seen": 66716185, + "step": 3105, + "time_per_iteration": 2.7346246242523193 + }, + { + "auxiliary_loss_clip": 0.01164213, + "auxiliary_loss_mlp": 0.01030147, + "balance_loss_clip": 1.02081609, + "balance_loss_mlp": 1.02260065, + "epoch": 0.37347441832501654, + "flos": 24023879792640.0, + "grad_norm": 2.197022531703171, + "language_loss": 0.7449199, + "learning_rate": 2.8850722156606207e-06, + "loss": 0.76686347, + "num_input_tokens_seen": 66734575, + "step": 3106, + "time_per_iteration": 2.654181957244873 + }, + { + "auxiliary_loss_clip": 0.01159145, + "auxiliary_loss_mlp": 0.0102752, + "balance_loss_clip": 1.01749873, + "balance_loss_mlp": 1.01947987, + "epoch": 0.3735946612156556, + "flos": 19714922409600.0, + "grad_norm": 4.218545319302089, + "language_loss": 0.66896272, + "learning_rate": 2.8843736051866252e-06, + "loss": 0.6908294, + "num_input_tokens_seen": 66753500, + "step": 3107, + "time_per_iteration": 2.717172384262085 + }, + { + "auxiliary_loss_clip": 0.01091832, + "auxiliary_loss_mlp": 0.0051013, + "balance_loss_clip": 0.90352029, + "balance_loss_mlp": 0.0, + "epoch": 0.3737149041062947, + "flos": 23039604334080.0, + "grad_norm": 1.8645477899097231, + "language_loss": 0.69615155, + "learning_rate": 2.8836748605538557e-06, + "loss": 0.7121712, + "num_input_tokens_seen": 66775140, + "step": 3108, + "time_per_iteration": 3.771718740463257 + }, + { + "auxiliary_loss_clip": 0.01145748, + "auxiliary_loss_mlp": 0.01034277, + "balance_loss_clip": 0.97919911, + "balance_loss_mlp": 1.02630234, + "epoch": 0.3738351469969338, + "flos": 34678108483200.0, + "grad_norm": 2.096258538224706, + "language_loss": 0.63176787, + "learning_rate": 2.882975981868313e-06, + "loss": 0.65356809, + "num_input_tokens_seen": 66795525, + "step": 3109, + "time_per_iteration": 3.711573839187622 + }, + { + "auxiliary_loss_clip": 0.01170639, + "auxiliary_loss_mlp": 0.01029055, + "balance_loss_clip": 1.02485847, + "balance_loss_mlp": 1.02070475, + "epoch": 0.3739553898875729, + "flos": 43507967448960.0, + "grad_norm": 2.4633775005101155, + "language_loss": 0.68752241, + "learning_rate": 2.882276969236016e-06, + "loss": 0.70951927, + "num_input_tokens_seen": 66816885, + "step": 3110, + "time_per_iteration": 2.8364977836608887 + }, + { + "auxiliary_loss_clip": 0.01138589, + "auxiliary_loss_mlp": 0.0103368, + "balance_loss_clip": 0.98063111, + "balance_loss_mlp": 1.02548409, + "epoch": 0.374075632778212, + "flos": 12856487448960.0, + "grad_norm": 1.956235890407814, + "language_loss": 0.76219338, + "learning_rate": 2.881577822763005e-06, + "loss": 0.78391606, + "num_input_tokens_seen": 66834835, + "step": 3111, + "time_per_iteration": 3.565218925476074 + }, + { + "auxiliary_loss_clip": 0.01165833, + "auxiliary_loss_mlp": 0.01030427, + "balance_loss_clip": 1.01837337, + "balance_loss_mlp": 1.02210617, + "epoch": 0.3741958756688511, + "flos": 26024031699840.0, + "grad_norm": 1.8666684556572248, + "language_loss": 0.87113392, + "learning_rate": 2.880878542555338e-06, + "loss": 0.89309657, + "num_input_tokens_seen": 66852600, + "step": 3112, + "time_per_iteration": 2.7263877391815186 + }, + { + "auxiliary_loss_clip": 0.01191041, + "auxiliary_loss_mlp": 0.01032193, + "balance_loss_clip": 1.05792797, + "balance_loss_mlp": 1.02373552, + "epoch": 0.37431611855949015, + "flos": 21433894652160.0, + "grad_norm": 2.0032306222578953, + "language_loss": 0.80276191, + "learning_rate": 2.8801791287190976e-06, + "loss": 0.82499427, + "num_input_tokens_seen": 66870595, + "step": 3113, + "time_per_iteration": 2.6459240913391113 + }, + { + "auxiliary_loss_clip": 0.01170375, + "auxiliary_loss_mlp": 0.01033619, + "balance_loss_clip": 1.02038336, + "balance_loss_mlp": 1.02535772, + "epoch": 0.37443636145012926, + "flos": 24207096090240.0, + "grad_norm": 2.725310217325944, + "language_loss": 0.86601484, + "learning_rate": 2.8794795813603817e-06, + "loss": 0.88805473, + "num_input_tokens_seen": 66886060, + "step": 3114, + "time_per_iteration": 2.6475307941436768 + }, + { + "auxiliary_loss_clip": 0.01169733, + "auxiliary_loss_mlp": 0.01032238, + "balance_loss_clip": 1.01774669, + "balance_loss_mlp": 1.02382791, + "epoch": 0.3745566043407684, + "flos": 15378601841280.0, + "grad_norm": 9.710807729353496, + "language_loss": 0.81208324, + "learning_rate": 2.878779900585314e-06, + "loss": 0.83410287, + "num_input_tokens_seen": 66903900, + "step": 3115, + "time_per_iteration": 2.6593925952911377 + }, + { + "auxiliary_loss_clip": 0.01149621, + "auxiliary_loss_mlp": 0.01032335, + "balance_loss_clip": 0.98321831, + "balance_loss_mlp": 1.02430069, + "epoch": 0.37467684723140743, + "flos": 24608218245120.0, + "grad_norm": 2.314914013721427, + "language_loss": 0.75478065, + "learning_rate": 2.8780800865000336e-06, + "loss": 0.77660018, + "num_input_tokens_seen": 66925210, + "step": 3116, + "time_per_iteration": 2.7171802520751953 + }, + { + "auxiliary_loss_clip": 0.01071039, + "auxiliary_loss_mlp": 0.01003131, + "balance_loss_clip": 0.98841512, + "balance_loss_mlp": 1.00184369, + "epoch": 0.37479709012204654, + "flos": 64377491610240.0, + "grad_norm": 0.9896736220150333, + "language_loss": 0.59206432, + "learning_rate": 2.877380139210702e-06, + "loss": 0.61280596, + "num_input_tokens_seen": 66983880, + "step": 3117, + "time_per_iteration": 4.0749499797821045 + }, + { + "auxiliary_loss_clip": 0.01126091, + "auxiliary_loss_mlp": 0.01027629, + "balance_loss_clip": 0.94673097, + "balance_loss_mlp": 1.01932013, + "epoch": 0.37491733301268565, + "flos": 23803962773760.0, + "grad_norm": 1.7536644889211357, + "language_loss": 0.76571888, + "learning_rate": 2.876680058823501e-06, + "loss": 0.78725606, + "num_input_tokens_seen": 67004280, + "step": 3118, + "time_per_iteration": 2.722578525543213 + }, + { + "auxiliary_loss_clip": 0.01136377, + "auxiliary_loss_mlp": 0.01031157, + "balance_loss_clip": 0.98127913, + "balance_loss_mlp": 1.02250242, + "epoch": 0.3750375759033247, + "flos": 32160950167680.0, + "grad_norm": 2.2363468019128843, + "language_loss": 0.66273409, + "learning_rate": 2.8759798454446314e-06, + "loss": 0.68440944, + "num_input_tokens_seen": 67027445, + "step": 3119, + "time_per_iteration": 2.759766101837158 + }, + { + "auxiliary_loss_clip": 0.01171343, + "auxiliary_loss_mlp": 0.01034968, + "balance_loss_clip": 1.02099729, + "balance_loss_mlp": 1.02661729, + "epoch": 0.3751578187939638, + "flos": 23367791923200.0, + "grad_norm": 1.9749363119725587, + "language_loss": 0.81482661, + "learning_rate": 2.8752794991803173e-06, + "loss": 0.83688974, + "num_input_tokens_seen": 67045130, + "step": 3120, + "time_per_iteration": 2.69677996635437 + }, + { + "auxiliary_loss_clip": 0.0114508, + "auxiliary_loss_mlp": 0.01031437, + "balance_loss_clip": 0.9839952, + "balance_loss_mlp": 1.02356887, + "epoch": 0.37527806168460287, + "flos": 14605731878400.0, + "grad_norm": 5.3253507990555455, + "language_loss": 0.75383258, + "learning_rate": 2.8745790201367976e-06, + "loss": 0.77559781, + "num_input_tokens_seen": 67060885, + "step": 3121, + "time_per_iteration": 2.6815221309661865 + }, + { + "auxiliary_loss_clip": 0.01197097, + "auxiliary_loss_mlp": 0.01026578, + "balance_loss_clip": 1.06208086, + "balance_loss_mlp": 1.01786399, + "epoch": 0.375398304575242, + "flos": 26390823431040.0, + "grad_norm": 2.7151077199305984, + "language_loss": 0.84572899, + "learning_rate": 2.8738784084203373e-06, + "loss": 0.8679657, + "num_input_tokens_seen": 67080960, + "step": 3122, + "time_per_iteration": 2.6988894939422607 + }, + { + "auxiliary_loss_clip": 0.0113759, + "auxiliary_loss_mlp": 0.01030779, + "balance_loss_clip": 0.97906256, + "balance_loss_mlp": 1.02291656, + "epoch": 0.3755185474658811, + "flos": 22236605838720.0, + "grad_norm": 1.7093113613550954, + "language_loss": 0.78774631, + "learning_rate": 2.873177664137216e-06, + "loss": 0.80942994, + "num_input_tokens_seen": 67101890, + "step": 3123, + "time_per_iteration": 2.8197383880615234 + }, + { + "auxiliary_loss_clip": 0.01122042, + "auxiliary_loss_mlp": 0.01028115, + "balance_loss_clip": 0.94981253, + "balance_loss_mlp": 1.0200026, + "epoch": 0.37563879035652015, + "flos": 30812935633920.0, + "grad_norm": 1.753340919624652, + "language_loss": 0.69211149, + "learning_rate": 2.8724767873937384e-06, + "loss": 0.71361303, + "num_input_tokens_seen": 67126010, + "step": 3124, + "time_per_iteration": 2.833324670791626 + }, + { + "auxiliary_loss_clip": 0.01144631, + "auxiliary_loss_mlp": 0.01037768, + "balance_loss_clip": 0.98291779, + "balance_loss_mlp": 1.0296495, + "epoch": 0.37575903324715926, + "flos": 20773533064320.0, + "grad_norm": 2.2650222625003837, + "language_loss": 0.87140119, + "learning_rate": 2.871775778296225e-06, + "loss": 0.89322507, + "num_input_tokens_seen": 67143100, + "step": 3125, + "time_per_iteration": 2.70459246635437 + }, + { + "auxiliary_loss_clip": 0.01169999, + "auxiliary_loss_mlp": 0.01031593, + "balance_loss_clip": 1.0235579, + "balance_loss_mlp": 1.02336144, + "epoch": 0.37587927613779837, + "flos": 18697681244160.0, + "grad_norm": 2.6824156145260454, + "language_loss": 0.78572619, + "learning_rate": 2.8710746369510196e-06, + "loss": 0.80774212, + "num_input_tokens_seen": 67161085, + "step": 3126, + "time_per_iteration": 2.608654499053955 + }, + { + "auxiliary_loss_clip": 0.01145261, + "auxiliary_loss_mlp": 0.01031463, + "balance_loss_clip": 0.98511398, + "balance_loss_mlp": 1.02344942, + "epoch": 0.3759995190284374, + "flos": 13624796384640.0, + "grad_norm": 2.391456210060633, + "language_loss": 0.83013511, + "learning_rate": 2.8703733634644846e-06, + "loss": 0.85190237, + "num_input_tokens_seen": 67175840, + "step": 3127, + "time_per_iteration": 2.6616625785827637 + }, + { + "auxiliary_loss_clip": 0.01190946, + "auxiliary_loss_mlp": 0.01032265, + "balance_loss_clip": 1.06072474, + "balance_loss_mlp": 1.02392626, + "epoch": 0.37611976191907653, + "flos": 20484847457280.0, + "grad_norm": 1.6003477512271618, + "language_loss": 0.79354489, + "learning_rate": 2.869671957943002e-06, + "loss": 0.815777, + "num_input_tokens_seen": 67194995, + "step": 3128, + "time_per_iteration": 2.597862482070923 + }, + { + "auxiliary_loss_clip": 0.01139471, + "auxiliary_loss_mlp": 0.01025829, + "balance_loss_clip": 0.9853127, + "balance_loss_mlp": 1.01715672, + "epoch": 0.37624000480971564, + "flos": 21141797253120.0, + "grad_norm": 1.714107336863014, + "language_loss": 0.74466765, + "learning_rate": 2.8689704204929747e-06, + "loss": 0.76632059, + "num_input_tokens_seen": 67214175, + "step": 3129, + "time_per_iteration": 2.66424560546875 + }, + { + "auxiliary_loss_clip": 0.01194281, + "auxiliary_loss_mlp": 0.01023186, + "balance_loss_clip": 1.06098783, + "balance_loss_mlp": 1.01495433, + "epoch": 0.3763602477003547, + "flos": 22564470205440.0, + "grad_norm": 2.7067324433011524, + "language_loss": 0.81097597, + "learning_rate": 2.8682687512208253e-06, + "loss": 0.83315063, + "num_input_tokens_seen": 67233185, + "step": 3130, + "time_per_iteration": 2.5814573764801025 + }, + { + "auxiliary_loss_clip": 0.01176738, + "auxiliary_loss_mlp": 0.01034144, + "balance_loss_clip": 1.02254725, + "balance_loss_mlp": 1.02528048, + "epoch": 0.3764804905909938, + "flos": 27526857851520.0, + "grad_norm": 2.219041615851089, + "language_loss": 0.80498618, + "learning_rate": 2.8675669502329972e-06, + "loss": 0.82709497, + "num_input_tokens_seen": 67254715, + "step": 3131, + "time_per_iteration": 2.7187891006469727 + }, + { + "auxiliary_loss_clip": 0.01168894, + "auxiliary_loss_mlp": 0.00510647, + "balance_loss_clip": 1.02262127, + "balance_loss_mlp": 0.0, + "epoch": 0.3766007334816329, + "flos": 22528092706560.0, + "grad_norm": 2.2449502302112094, + "language_loss": 0.86011499, + "learning_rate": 2.866865017635952e-06, + "loss": 0.87691039, + "num_input_tokens_seen": 67272535, + "step": 3132, + "time_per_iteration": 2.64048171043396 + }, + { + "auxiliary_loss_clip": 0.01117808, + "auxiliary_loss_mlp": 0.01033133, + "balance_loss_clip": 0.94592094, + "balance_loss_mlp": 1.02511597, + "epoch": 0.376720976372272, + "flos": 25957166532480.0, + "grad_norm": 1.4435562551013008, + "language_loss": 0.79333323, + "learning_rate": 2.866162953536174e-06, + "loss": 0.81484264, + "num_input_tokens_seen": 67293505, + "step": 3133, + "time_per_iteration": 2.7180206775665283 + }, + { + "auxiliary_loss_clip": 0.01145142, + "auxiliary_loss_mlp": 0.00510221, + "balance_loss_clip": 0.98314112, + "balance_loss_mlp": 0.0, + "epoch": 0.3768412192629111, + "flos": 18041162411520.0, + "grad_norm": 1.608584903842279, + "language_loss": 0.75119197, + "learning_rate": 2.8654607580401634e-06, + "loss": 0.76774561, + "num_input_tokens_seen": 67313240, + "step": 3134, + "time_per_iteration": 3.671560049057007 + }, + { + "auxiliary_loss_clip": 0.01074848, + "auxiliary_loss_mlp": 0.01006602, + "balance_loss_clip": 0.99327165, + "balance_loss_mlp": 1.00513005, + "epoch": 0.3769614621535502, + "flos": 62989472304000.0, + "grad_norm": 1.1282801975324557, + "language_loss": 0.65173352, + "learning_rate": 2.8647584312544446e-06, + "loss": 0.67254806, + "num_input_tokens_seen": 67378445, + "step": 3135, + "time_per_iteration": 3.2633676528930664 + }, + { + "auxiliary_loss_clip": 0.01119393, + "auxiliary_loss_mlp": 0.00510143, + "balance_loss_clip": 0.94339162, + "balance_loss_mlp": 0.0, + "epoch": 0.37708170504418925, + "flos": 23661685002240.0, + "grad_norm": 1.5476577037242736, + "language_loss": 0.85514987, + "learning_rate": 2.864055973285559e-06, + "loss": 0.87144524, + "num_input_tokens_seen": 67400445, + "step": 3136, + "time_per_iteration": 3.676584243774414 + }, + { + "auxiliary_loss_clip": 0.01133782, + "auxiliary_loss_mlp": 0.01028794, + "balance_loss_clip": 0.97933167, + "balance_loss_mlp": 1.02112257, + "epoch": 0.37720194793482836, + "flos": 24423170353920.0, + "grad_norm": 2.001922062751193, + "language_loss": 0.86513644, + "learning_rate": 2.8633533842400698e-06, + "loss": 0.88676226, + "num_input_tokens_seen": 67420645, + "step": 3137, + "time_per_iteration": 3.6729161739349365 + }, + { + "auxiliary_loss_clip": 0.01169932, + "auxiliary_loss_mlp": 0.00511091, + "balance_loss_clip": 1.02330422, + "balance_loss_mlp": 0.0, + "epoch": 0.3773221908254674, + "flos": 20996502739200.0, + "grad_norm": 1.8662468348272085, + "language_loss": 0.7752794, + "learning_rate": 2.862650664224558e-06, + "loss": 0.79208964, + "num_input_tokens_seen": 67439495, + "step": 3138, + "time_per_iteration": 2.619079828262329 + }, + { + "auxiliary_loss_clip": 0.01166034, + "auxiliary_loss_mlp": 0.01026905, + "balance_loss_clip": 1.02507114, + "balance_loss_mlp": 1.01880491, + "epoch": 0.37744243371610653, + "flos": 37631724958080.0, + "grad_norm": 1.446765061945622, + "language_loss": 0.69832289, + "learning_rate": 2.861947813345627e-06, + "loss": 0.72025228, + "num_input_tokens_seen": 67462195, + "step": 3139, + "time_per_iteration": 2.818251609802246 + }, + { + "auxiliary_loss_clip": 0.01196231, + "auxiliary_loss_mlp": 0.00510959, + "balance_loss_clip": 1.06365478, + "balance_loss_mlp": 0.0, + "epoch": 0.37756267660674564, + "flos": 26140526484480.0, + "grad_norm": 1.8137935466916748, + "language_loss": 0.72740579, + "learning_rate": 2.8612448317098974e-06, + "loss": 0.74447775, + "num_input_tokens_seen": 67482530, + "step": 3140, + "time_per_iteration": 2.640659809112549 + }, + { + "auxiliary_loss_clip": 0.01121991, + "auxiliary_loss_mlp": 0.00511276, + "balance_loss_clip": 0.94353265, + "balance_loss_mlp": 0.0, + "epoch": 0.3776829194973847, + "flos": 19427888828160.0, + "grad_norm": 2.443006546256734, + "language_loss": 0.82891607, + "learning_rate": 2.8605417194240114e-06, + "loss": 0.8452487, + "num_input_tokens_seen": 67500890, + "step": 3141, + "time_per_iteration": 2.8122804164886475 + }, + { + "auxiliary_loss_clip": 0.01162411, + "auxiliary_loss_mlp": 0.01024493, + "balance_loss_clip": 1.02046347, + "balance_loss_mlp": 1.01697683, + "epoch": 0.3778031623880238, + "flos": 17382309194880.0, + "grad_norm": 1.784371130809307, + "language_loss": 0.7883122, + "learning_rate": 2.8598384765946315e-06, + "loss": 0.81018126, + "num_input_tokens_seen": 67519545, + "step": 3142, + "time_per_iteration": 2.606640100479126 + }, + { + "auxiliary_loss_clip": 0.01191445, + "auxiliary_loss_mlp": 0.01029569, + "balance_loss_clip": 1.05940223, + "balance_loss_mlp": 1.02125454, + "epoch": 0.3779234052786629, + "flos": 27125843437440.0, + "grad_norm": 2.2490650918359116, + "language_loss": 0.71308517, + "learning_rate": 2.8591351033284377e-06, + "loss": 0.7352953, + "num_input_tokens_seen": 67539275, + "step": 3143, + "time_per_iteration": 2.677027702331543 + }, + { + "auxiliary_loss_clip": 0.01170831, + "auxiliary_loss_mlp": 0.01032579, + "balance_loss_clip": 1.02029908, + "balance_loss_mlp": 1.02440119, + "epoch": 0.37804364816930197, + "flos": 19682639061120.0, + "grad_norm": 2.1423097145496826, + "language_loss": 0.83135879, + "learning_rate": 2.8584315997321325e-06, + "loss": 0.8533929, + "num_input_tokens_seen": 67558280, + "step": 3144, + "time_per_iteration": 3.540210008621216 + }, + { + "auxiliary_loss_clip": 0.01192738, + "auxiliary_loss_mlp": 0.01028507, + "balance_loss_clip": 1.05995822, + "balance_loss_mlp": 1.02010918, + "epoch": 0.3781638910599411, + "flos": 22702905221760.0, + "grad_norm": 2.3870527782475652, + "language_loss": 0.78694272, + "learning_rate": 2.8577279659124356e-06, + "loss": 0.80915517, + "num_input_tokens_seen": 67575955, + "step": 3145, + "time_per_iteration": 2.6136443614959717 + }, + { + "auxiliary_loss_clip": 0.01165783, + "auxiliary_loss_mlp": 0.0102639, + "balance_loss_clip": 1.02202439, + "balance_loss_mlp": 1.01871276, + "epoch": 0.3782841339505802, + "flos": 14647604158080.0, + "grad_norm": 1.745096523729129, + "language_loss": 0.83538735, + "learning_rate": 2.857024201976089e-06, + "loss": 0.8573091, + "num_input_tokens_seen": 67593515, + "step": 3146, + "time_per_iteration": 2.635476589202881 + }, + { + "auxiliary_loss_clip": 0.01143267, + "auxiliary_loss_mlp": 0.01027624, + "balance_loss_clip": 0.98394322, + "balance_loss_mlp": 1.01900482, + "epoch": 0.37840437684121925, + "flos": 32818223185920.0, + "grad_norm": 1.9350828152417687, + "language_loss": 0.72992033, + "learning_rate": 2.8563203080298516e-06, + "loss": 0.75162923, + "num_input_tokens_seen": 67614290, + "step": 3147, + "time_per_iteration": 2.753000259399414 + }, + { + "auxiliary_loss_clip": 0.01145791, + "auxiliary_loss_mlp": 0.00510607, + "balance_loss_clip": 0.98464561, + "balance_loss_mlp": 0.0, + "epoch": 0.37852461973185836, + "flos": 18369206346240.0, + "grad_norm": 2.0402579541645887, + "language_loss": 0.89415169, + "learning_rate": 2.855616284180505e-06, + "loss": 0.9107157, + "num_input_tokens_seen": 67631340, + "step": 3148, + "time_per_iteration": 2.7046761512756348 + }, + { + "auxiliary_loss_clip": 0.0108545, + "auxiliary_loss_mlp": 0.0100188, + "balance_loss_clip": 1.00044107, + "balance_loss_mlp": 1.00028253, + "epoch": 0.37864486262249747, + "flos": 59500680117120.0, + "grad_norm": 0.8797784726074774, + "language_loss": 0.66051149, + "learning_rate": 2.8549121305348477e-06, + "loss": 0.6813848, + "num_input_tokens_seen": 67691125, + "step": 3149, + "time_per_iteration": 3.1618130207061768 + }, + { + "auxiliary_loss_clip": 0.01169862, + "auxiliary_loss_mlp": 0.0102737, + "balance_loss_clip": 1.02334213, + "balance_loss_mlp": 1.0199132, + "epoch": 0.3787651055131365, + "flos": 23363015414400.0, + "grad_norm": 3.8913782162590205, + "language_loss": 0.83192152, + "learning_rate": 2.8542078471997006e-06, + "loss": 0.85389388, + "num_input_tokens_seen": 67708740, + "step": 3150, + "time_per_iteration": 2.733893632888794 + }, + { + "auxiliary_loss_clip": 0.0116828, + "auxiliary_loss_mlp": 0.01029087, + "balance_loss_clip": 1.02096438, + "balance_loss_mlp": 1.02124345, + "epoch": 0.37888534840377563, + "flos": 24601394661120.0, + "grad_norm": 1.725582264049851, + "language_loss": 0.75996387, + "learning_rate": 2.8535034342819013e-06, + "loss": 0.78193748, + "num_input_tokens_seen": 67726150, + "step": 3151, + "time_per_iteration": 2.712932825088501 + }, + { + "auxiliary_loss_clip": 0.01189019, + "auxiliary_loss_mlp": 0.01025346, + "balance_loss_clip": 1.05991197, + "balance_loss_mlp": 1.01717401, + "epoch": 0.37900559129441475, + "flos": 23986891762560.0, + "grad_norm": 1.5730076665942148, + "language_loss": 0.72443211, + "learning_rate": 2.85279889188831e-06, + "loss": 0.74657571, + "num_input_tokens_seen": 67746525, + "step": 3152, + "time_per_iteration": 2.6479578018188477 + }, + { + "auxiliary_loss_clip": 0.01119556, + "auxiliary_loss_mlp": 0.01027411, + "balance_loss_clip": 0.94213426, + "balance_loss_mlp": 1.01907778, + "epoch": 0.3791258341850538, + "flos": 24644667571200.0, + "grad_norm": 1.7332773458406843, + "language_loss": 0.81211245, + "learning_rate": 2.852094220125805e-06, + "loss": 0.8335821, + "num_input_tokens_seen": 67766035, + "step": 3153, + "time_per_iteration": 2.8062212467193604 + }, + { + "auxiliary_loss_clip": 0.01172593, + "auxiliary_loss_mlp": 0.01030526, + "balance_loss_clip": 1.02599955, + "balance_loss_mlp": 1.02234268, + "epoch": 0.3792460770756929, + "flos": 17420841509760.0, + "grad_norm": 2.091661889662417, + "language_loss": 0.71074736, + "learning_rate": 2.8513894191012846e-06, + "loss": 0.73277855, + "num_input_tokens_seen": 67785015, + "step": 3154, + "time_per_iteration": 2.6516873836517334 + }, + { + "auxiliary_loss_clip": 0.01195737, + "auxiliary_loss_mlp": 0.01029457, + "balance_loss_clip": 1.06276917, + "balance_loss_mlp": 1.02160084, + "epoch": 0.37936631996633197, + "flos": 24206557386240.0, + "grad_norm": 1.4734009737350056, + "language_loss": 0.78600681, + "learning_rate": 2.8506844889216664e-06, + "loss": 0.80825877, + "num_input_tokens_seen": 67804400, + "step": 3155, + "time_per_iteration": 2.651031732559204 + }, + { + "auxiliary_loss_clip": 0.01081973, + "auxiliary_loss_mlp": 0.01004555, + "balance_loss_clip": 1.0031898, + "balance_loss_mlp": 1.00294518, + "epoch": 0.3794865628569711, + "flos": 70297114752000.0, + "grad_norm": 0.8643159958388636, + "language_loss": 0.62855083, + "learning_rate": 2.849979429693887e-06, + "loss": 0.64941603, + "num_input_tokens_seen": 67865385, + "step": 3156, + "time_per_iteration": 3.2642931938171387 + }, + { + "auxiliary_loss_clip": 0.01193952, + "auxiliary_loss_mlp": 0.01029016, + "balance_loss_clip": 1.06140292, + "balance_loss_mlp": 1.02028441, + "epoch": 0.3796068057476102, + "flos": 15779364860160.0, + "grad_norm": 1.8988219334661538, + "language_loss": 0.74141431, + "learning_rate": 2.8492742415249042e-06, + "loss": 0.76364398, + "num_input_tokens_seen": 67883030, + "step": 3157, + "time_per_iteration": 2.6664366722106934 + }, + { + "auxiliary_loss_clip": 0.01188809, + "auxiliary_loss_mlp": 0.01025188, + "balance_loss_clip": 1.05863237, + "balance_loss_mlp": 1.01713562, + "epoch": 0.37972704863824924, + "flos": 25191694771200.0, + "grad_norm": 1.6393880335625823, + "language_loss": 0.7630409, + "learning_rate": 2.848568924521694e-06, + "loss": 0.78518093, + "num_input_tokens_seen": 67903810, + "step": 3158, + "time_per_iteration": 2.6417489051818848 + }, + { + "auxiliary_loss_clip": 0.01159732, + "auxiliary_loss_mlp": 0.01028308, + "balance_loss_clip": 1.01715016, + "balance_loss_mlp": 1.02010608, + "epoch": 0.37984729152888835, + "flos": 26210372480640.0, + "grad_norm": 1.8844174931055766, + "language_loss": 0.73720521, + "learning_rate": 2.8478634787912526e-06, + "loss": 0.7590856, + "num_input_tokens_seen": 67921865, + "step": 3159, + "time_per_iteration": 2.703878402709961 + }, + { + "auxiliary_loss_clip": 0.01168573, + "auxiliary_loss_mlp": 0.01031145, + "balance_loss_clip": 1.02159667, + "balance_loss_mlp": 1.02342081, + "epoch": 0.37996753441952746, + "flos": 25629302165760.0, + "grad_norm": 2.023403381144134, + "language_loss": 0.76767254, + "learning_rate": 2.847157904440596e-06, + "loss": 0.78966975, + "num_input_tokens_seen": 67941595, + "step": 3160, + "time_per_iteration": 3.5926783084869385 + }, + { + "auxiliary_loss_clip": 0.0116863, + "auxiliary_loss_mlp": 0.01024932, + "balance_loss_clip": 1.02190983, + "balance_loss_mlp": 1.01704681, + "epoch": 0.3800877773101665, + "flos": 20118414862080.0, + "grad_norm": 1.53837486499331, + "language_loss": 0.74119425, + "learning_rate": 2.846452201576759e-06, + "loss": 0.76312983, + "num_input_tokens_seen": 67960970, + "step": 3161, + "time_per_iteration": 2.6432220935821533 + }, + { + "auxiliary_loss_clip": 0.01064253, + "auxiliary_loss_mlp": 0.01013302, + "balance_loss_clip": 0.9615792, + "balance_loss_mlp": 1.01184165, + "epoch": 0.38020802020080563, + "flos": 63053608037760.0, + "grad_norm": 0.8736970683436812, + "language_loss": 0.62802875, + "learning_rate": 2.845746370306795e-06, + "loss": 0.64880431, + "num_input_tokens_seen": 68026160, + "step": 3162, + "time_per_iteration": 4.244068145751953 + }, + { + "auxiliary_loss_clip": 0.01169698, + "auxiliary_loss_mlp": 0.01030627, + "balance_loss_clip": 1.02164006, + "balance_loss_mlp": 1.02292573, + "epoch": 0.38032826309144474, + "flos": 21288420570240.0, + "grad_norm": 1.8464186731944439, + "language_loss": 0.78303355, + "learning_rate": 2.84504041073778e-06, + "loss": 0.80503678, + "num_input_tokens_seen": 68044575, + "step": 3163, + "time_per_iteration": 3.6181726455688477 + }, + { + "auxiliary_loss_clip": 0.01138376, + "auxiliary_loss_mlp": 0.01031615, + "balance_loss_clip": 0.98327601, + "balance_loss_mlp": 1.02245975, + "epoch": 0.3804485059820838, + "flos": 18954119416320.0, + "grad_norm": 1.7267330373448553, + "language_loss": 0.79132813, + "learning_rate": 2.844334322976806e-06, + "loss": 0.81302804, + "num_input_tokens_seen": 68064790, + "step": 3164, + "time_per_iteration": 2.699547290802002 + }, + { + "auxiliary_loss_clip": 0.01100617, + "auxiliary_loss_mlp": 0.01033364, + "balance_loss_clip": 0.90509194, + "balance_loss_mlp": 1.02449477, + "epoch": 0.3805687488727229, + "flos": 21833759831040.0, + "grad_norm": 1.7384910293625773, + "language_loss": 0.8340044, + "learning_rate": 2.8436281071309866e-06, + "loss": 0.85534418, + "num_input_tokens_seen": 68083330, + "step": 3165, + "time_per_iteration": 2.7907400131225586 + }, + { + "auxiliary_loss_clip": 0.01020101, + "auxiliary_loss_mlp": 0.01010583, + "balance_loss_clip": 0.8803829, + "balance_loss_mlp": 1.00899744, + "epoch": 0.380688991763362, + "flos": 58546209968640.0, + "grad_norm": 0.7301277183857348, + "language_loss": 0.52983797, + "learning_rate": 2.842921763307455e-06, + "loss": 0.55014479, + "num_input_tokens_seen": 68146140, + "step": 3166, + "time_per_iteration": 3.3168258666992188 + }, + { + "auxiliary_loss_clip": 0.01139008, + "auxiliary_loss_mlp": 0.01025033, + "balance_loss_clip": 0.98023552, + "balance_loss_mlp": 1.01700461, + "epoch": 0.38080923465400107, + "flos": 23799509487360.0, + "grad_norm": 1.9732777501994436, + "language_loss": 0.82651681, + "learning_rate": 2.842215291613361e-06, + "loss": 0.84815723, + "num_input_tokens_seen": 68164520, + "step": 3167, + "time_per_iteration": 2.7244653701782227 + }, + { + "auxiliary_loss_clip": 0.00974746, + "auxiliary_loss_mlp": 0.01009532, + "balance_loss_clip": 0.81737757, + "balance_loss_mlp": 1.00810134, + "epoch": 0.3809294775446402, + "flos": 54969866380800.0, + "grad_norm": 0.7890120747183534, + "language_loss": 0.59110951, + "learning_rate": 2.8415086921558774e-06, + "loss": 0.61095226, + "num_input_tokens_seen": 68227945, + "step": 3168, + "time_per_iteration": 3.473992347717285 + }, + { + "auxiliary_loss_clip": 0.01128189, + "auxiliary_loss_mlp": 0.01022589, + "balance_loss_clip": 0.97494507, + "balance_loss_mlp": 1.01436388, + "epoch": 0.38104972043527924, + "flos": 24643697904000.0, + "grad_norm": 1.5128078576195616, + "language_loss": 0.78553104, + "learning_rate": 2.840801965042194e-06, + "loss": 0.80703884, + "num_input_tokens_seen": 68247405, + "step": 3169, + "time_per_iteration": 2.8417036533355713 + }, + { + "auxiliary_loss_clip": 0.01132344, + "auxiliary_loss_mlp": 0.01026616, + "balance_loss_clip": 0.97814238, + "balance_loss_mlp": 1.01843858, + "epoch": 0.38116996332591835, + "flos": 22856783086080.0, + "grad_norm": 1.9613161068604335, + "language_loss": 0.84060633, + "learning_rate": 2.840095110379521e-06, + "loss": 0.86219591, + "num_input_tokens_seen": 68266925, + "step": 3170, + "time_per_iteration": 3.579390287399292 + }, + { + "auxiliary_loss_clip": 0.01009049, + "auxiliary_loss_mlp": 0.01001852, + "balance_loss_clip": 0.87779748, + "balance_loss_mlp": 1.00041008, + "epoch": 0.38129020621655746, + "flos": 60836160804480.0, + "grad_norm": 0.7300284582735741, + "language_loss": 0.53916919, + "learning_rate": 2.8393881282750884e-06, + "loss": 0.55927825, + "num_input_tokens_seen": 68329755, + "step": 3171, + "time_per_iteration": 3.21701979637146 + }, + { + "auxiliary_loss_clip": 0.01142701, + "auxiliary_loss_mlp": 0.01025186, + "balance_loss_clip": 0.98242629, + "balance_loss_mlp": 1.01696002, + "epoch": 0.3814104491071965, + "flos": 21648101408640.0, + "grad_norm": 3.012452210661034, + "language_loss": 0.78979027, + "learning_rate": 2.838681018836144e-06, + "loss": 0.81146908, + "num_input_tokens_seen": 68347075, + "step": 3172, + "time_per_iteration": 2.700009822845459 + }, + { + "auxiliary_loss_clip": 0.01117679, + "auxiliary_loss_mlp": 0.00510432, + "balance_loss_clip": 0.93995774, + "balance_loss_mlp": 0.0, + "epoch": 0.3815306919978356, + "flos": 19099090707840.0, + "grad_norm": 4.905299851868451, + "language_loss": 0.78671223, + "learning_rate": 2.837973782169955e-06, + "loss": 0.8029933, + "num_input_tokens_seen": 68365450, + "step": 3173, + "time_per_iteration": 2.733004093170166 + }, + { + "auxiliary_loss_clip": 0.01091719, + "auxiliary_loss_mlp": 0.01000691, + "balance_loss_clip": 1.02627063, + "balance_loss_mlp": 0.99918282, + "epoch": 0.38165093488847474, + "flos": 67067918156160.0, + "grad_norm": 0.8059283140177607, + "language_loss": 0.59150839, + "learning_rate": 2.8372664183838096e-06, + "loss": 0.61243248, + "num_input_tokens_seen": 68428470, + "step": 3174, + "time_per_iteration": 3.2415473461151123 + }, + { + "auxiliary_loss_clip": 0.01191287, + "auxiliary_loss_mlp": 0.01027598, + "balance_loss_clip": 1.06000769, + "balance_loss_mlp": 1.01922941, + "epoch": 0.3817711777791138, + "flos": 22341105480960.0, + "grad_norm": 2.0841308104050245, + "language_loss": 0.68167132, + "learning_rate": 2.836558927585015e-06, + "loss": 0.70386016, + "num_input_tokens_seen": 68445440, + "step": 3175, + "time_per_iteration": 2.6301066875457764 + }, + { + "auxiliary_loss_clip": 0.01170655, + "auxiliary_loss_mlp": 0.01027767, + "balance_loss_clip": 1.02235019, + "balance_loss_mlp": 1.01965821, + "epoch": 0.3818914206697529, + "flos": 22820621068800.0, + "grad_norm": 1.9148434887785177, + "language_loss": 0.8255958, + "learning_rate": 2.8358513098808957e-06, + "loss": 0.84758002, + "num_input_tokens_seen": 68465755, + "step": 3176, + "time_per_iteration": 2.695979118347168 + }, + { + "auxiliary_loss_clip": 0.01088833, + "auxiliary_loss_mlp": 0.01027869, + "balance_loss_clip": 0.90477711, + "balance_loss_mlp": 1.01935148, + "epoch": 0.382011663560392, + "flos": 24386074583040.0, + "grad_norm": 1.833634450887168, + "language_loss": 0.771698, + "learning_rate": 2.835143565378798e-06, + "loss": 0.79286504, + "num_input_tokens_seen": 68486220, + "step": 3177, + "time_per_iteration": 2.8525025844573975 + }, + { + "auxiliary_loss_clip": 0.01067234, + "auxiliary_loss_mlp": 0.01026164, + "balance_loss_clip": 0.86416161, + "balance_loss_mlp": 1.01814771, + "epoch": 0.38213190645103107, + "flos": 21981568296960.0, + "grad_norm": 1.813994992521813, + "language_loss": 0.78569186, + "learning_rate": 2.8344356941860847e-06, + "loss": 0.80662584, + "num_input_tokens_seen": 68505850, + "step": 3178, + "time_per_iteration": 2.864251136779785 + }, + { + "auxiliary_loss_clip": 0.01115674, + "auxiliary_loss_mlp": 0.010291, + "balance_loss_clip": 0.94333887, + "balance_loss_mlp": 1.02071655, + "epoch": 0.3822521493416702, + "flos": 35516945773440.0, + "grad_norm": 6.190177749398116, + "language_loss": 0.66755813, + "learning_rate": 2.8337276964101403e-06, + "loss": 0.68900585, + "num_input_tokens_seen": 68526290, + "step": 3179, + "time_per_iteration": 2.8617308139801025 + }, + { + "auxiliary_loss_clip": 0.01169036, + "auxiliary_loss_mlp": 0.01033608, + "balance_loss_clip": 1.02072382, + "balance_loss_mlp": 1.02543604, + "epoch": 0.3823723922323093, + "flos": 21069904181760.0, + "grad_norm": 1.8125716091324162, + "language_loss": 0.76056874, + "learning_rate": 2.833019572158367e-06, + "loss": 0.78259516, + "num_input_tokens_seen": 68544725, + "step": 3180, + "time_per_iteration": 2.676746368408203 + }, + { + "auxiliary_loss_clip": 0.01138035, + "auxiliary_loss_mlp": 0.01033012, + "balance_loss_clip": 0.97978222, + "balance_loss_mlp": 1.02473259, + "epoch": 0.38249263512294834, + "flos": 19789149864960.0, + "grad_norm": 1.723512406867232, + "language_loss": 0.79964429, + "learning_rate": 2.8323113215381872e-06, + "loss": 0.82135475, + "num_input_tokens_seen": 68563070, + "step": 3181, + "time_per_iteration": 2.695591688156128 + }, + { + "auxiliary_loss_clip": 0.01115185, + "auxiliary_loss_mlp": 0.01031592, + "balance_loss_clip": 0.94156098, + "balance_loss_mlp": 1.02327669, + "epoch": 0.38261287801358745, + "flos": 21433930565760.0, + "grad_norm": 2.049687152161928, + "language_loss": 0.76253295, + "learning_rate": 2.831602944657042e-06, + "loss": 0.78400075, + "num_input_tokens_seen": 68581150, + "step": 3182, + "time_per_iteration": 2.7278921604156494 + }, + { + "auxiliary_loss_clip": 0.01148424, + "auxiliary_loss_mlp": 0.01026511, + "balance_loss_clip": 0.97911012, + "balance_loss_mlp": 1.01808918, + "epoch": 0.38273312090422656, + "flos": 21981568296960.0, + "grad_norm": 2.5510870356335973, + "language_loss": 0.74052787, + "learning_rate": 2.830894441622391e-06, + "loss": 0.76227725, + "num_input_tokens_seen": 68597800, + "step": 3183, + "time_per_iteration": 2.6595535278320312 + }, + { + "auxiliary_loss_clip": 0.0111645, + "auxiliary_loss_mlp": 0.00510689, + "balance_loss_clip": 0.93891251, + "balance_loss_mlp": 0.0, + "epoch": 0.3828533637948656, + "flos": 24790895838720.0, + "grad_norm": 3.303860452813148, + "language_loss": 0.80213475, + "learning_rate": 2.8301858125417134e-06, + "loss": 0.81840611, + "num_input_tokens_seen": 68617640, + "step": 3184, + "time_per_iteration": 2.756413459777832 + }, + { + "auxiliary_loss_clip": 0.01143013, + "auxiliary_loss_mlp": 0.01028901, + "balance_loss_clip": 0.98327804, + "balance_loss_mlp": 1.02141476, + "epoch": 0.38297360668550473, + "flos": 22455445449600.0, + "grad_norm": 1.7917499740644383, + "language_loss": 0.739097, + "learning_rate": 2.8294770575225082e-06, + "loss": 0.7608161, + "num_input_tokens_seen": 68637770, + "step": 3185, + "time_per_iteration": 2.683793306350708 + }, + { + "auxiliary_loss_clip": 0.01168139, + "auxiliary_loss_mlp": 0.01030493, + "balance_loss_clip": 1.02228594, + "balance_loss_mlp": 1.02213073, + "epoch": 0.3830938495761438, + "flos": 24896903852160.0, + "grad_norm": 1.9423658631737815, + "language_loss": 0.83550525, + "learning_rate": 2.828768176672293e-06, + "loss": 0.85749161, + "num_input_tokens_seen": 68656885, + "step": 3186, + "time_per_iteration": 3.6191914081573486 + }, + { + "auxiliary_loss_clip": 0.01118333, + "auxiliary_loss_mlp": 0.01034371, + "balance_loss_clip": 0.93990242, + "balance_loss_mlp": 1.02537644, + "epoch": 0.3832140924667829, + "flos": 33036236784000.0, + "grad_norm": 1.7073062755002073, + "language_loss": 0.71715105, + "learning_rate": 2.8280591700986044e-06, + "loss": 0.7386781, + "num_input_tokens_seen": 68678750, + "step": 3187, + "time_per_iteration": 2.8296658992767334 + }, + { + "auxiliary_loss_clip": 0.01147253, + "auxiliary_loss_mlp": 0.01029383, + "balance_loss_clip": 0.98139155, + "balance_loss_mlp": 1.02140212, + "epoch": 0.383334335357422, + "flos": 31903721896320.0, + "grad_norm": 1.9420510233522301, + "language_loss": 0.75000501, + "learning_rate": 2.827350037908999e-06, + "loss": 0.77177143, + "num_input_tokens_seen": 68698190, + "step": 3188, + "time_per_iteration": 3.6631696224212646 + }, + { + "auxiliary_loss_clip": 0.011241, + "auxiliary_loss_mlp": 0.01032072, + "balance_loss_clip": 0.94219816, + "balance_loss_mlp": 1.0229404, + "epoch": 0.38345457824806106, + "flos": 19791915212160.0, + "grad_norm": 2.2380568521349415, + "language_loss": 0.79745579, + "learning_rate": 2.8266407802110496e-06, + "loss": 0.81901753, + "num_input_tokens_seen": 68716445, + "step": 3189, + "time_per_iteration": 3.738022804260254 + }, + { + "auxiliary_loss_clip": 0.01054698, + "auxiliary_loss_mlp": 0.01028545, + "balance_loss_clip": 0.82760632, + "balance_loss_mlp": 1.01932383, + "epoch": 0.3835748211387002, + "flos": 22419391173120.0, + "grad_norm": 1.6944785698634208, + "language_loss": 0.76075709, + "learning_rate": 2.8259313971123515e-06, + "loss": 0.78158957, + "num_input_tokens_seen": 68737565, + "step": 3190, + "time_per_iteration": 2.8665273189544678 + }, + { + "auxiliary_loss_clip": 0.01164937, + "auxiliary_loss_mlp": 0.0102504, + "balance_loss_clip": 1.02071548, + "balance_loss_mlp": 1.01690364, + "epoch": 0.3836950640293393, + "flos": 25118436983040.0, + "grad_norm": 1.4723423609865223, + "language_loss": 0.78478503, + "learning_rate": 2.8252218887205166e-06, + "loss": 0.80668479, + "num_input_tokens_seen": 68758255, + "step": 3191, + "time_per_iteration": 2.716646671295166 + }, + { + "auxiliary_loss_clip": 0.01073306, + "auxiliary_loss_mlp": 0.01028139, + "balance_loss_clip": 0.86483425, + "balance_loss_mlp": 1.02023578, + "epoch": 0.38381530691997834, + "flos": 21799213925760.0, + "grad_norm": 1.8271423969433818, + "language_loss": 0.80343997, + "learning_rate": 2.824512255143178e-06, + "loss": 0.82445443, + "num_input_tokens_seen": 68777490, + "step": 3192, + "time_per_iteration": 2.7914910316467285 + }, + { + "auxiliary_loss_clip": 0.01126122, + "auxiliary_loss_mlp": 0.01027713, + "balance_loss_clip": 0.94518864, + "balance_loss_mlp": 1.02011943, + "epoch": 0.38393554981061745, + "flos": 21252689516160.0, + "grad_norm": 2.4457087183500814, + "language_loss": 0.79493427, + "learning_rate": 2.8238024964879855e-06, + "loss": 0.81647265, + "num_input_tokens_seen": 68798385, + "step": 3193, + "time_per_iteration": 2.708385944366455 + }, + { + "auxiliary_loss_clip": 0.01194472, + "auxiliary_loss_mlp": 0.01033522, + "balance_loss_clip": 1.0594244, + "balance_loss_mlp": 1.02544582, + "epoch": 0.38405579270125656, + "flos": 17019360218880.0, + "grad_norm": 2.195382255788416, + "language_loss": 0.77069616, + "learning_rate": 2.8230926128626095e-06, + "loss": 0.79297614, + "num_input_tokens_seen": 68816880, + "step": 3194, + "time_per_iteration": 2.7420947551727295 + }, + { + "auxiliary_loss_clip": 0.0113693, + "auxiliary_loss_mlp": 0.01033755, + "balance_loss_clip": 0.97900224, + "balance_loss_mlp": 1.02544022, + "epoch": 0.3841760355918956, + "flos": 21835375943040.0, + "grad_norm": 1.8188222564330707, + "language_loss": 0.79395324, + "learning_rate": 2.822382604374738e-06, + "loss": 0.81566012, + "num_input_tokens_seen": 68835805, + "step": 3195, + "time_per_iteration": 2.708512544631958 + }, + { + "auxiliary_loss_clip": 0.01145636, + "auxiliary_loss_mlp": 0.0103599, + "balance_loss_clip": 0.98561561, + "balance_loss_mlp": 1.02824783, + "epoch": 0.3842962784825347, + "flos": 25915114684800.0, + "grad_norm": 3.7884252923931787, + "language_loss": 0.65728247, + "learning_rate": 2.8216724711320793e-06, + "loss": 0.67909884, + "num_input_tokens_seen": 68854930, + "step": 3196, + "time_per_iteration": 3.6238815784454346 + }, + { + "auxiliary_loss_clip": 0.01192218, + "auxiliary_loss_mlp": 0.00510137, + "balance_loss_clip": 1.05996561, + "balance_loss_mlp": 0.0, + "epoch": 0.38441652137317384, + "flos": 25337492075520.0, + "grad_norm": 1.6793628938121339, + "language_loss": 0.79808474, + "learning_rate": 2.820962213242361e-06, + "loss": 0.81510824, + "num_input_tokens_seen": 68874260, + "step": 3197, + "time_per_iteration": 2.732348918914795 + }, + { + "auxiliary_loss_clip": 0.01161595, + "auxiliary_loss_mlp": 0.01028061, + "balance_loss_clip": 1.02075243, + "balance_loss_mlp": 1.0204916, + "epoch": 0.3845367642638129, + "flos": 18113486446080.0, + "grad_norm": 2.2493714720650546, + "language_loss": 0.84378999, + "learning_rate": 2.8202518308133264e-06, + "loss": 0.86568654, + "num_input_tokens_seen": 68891535, + "step": 3198, + "time_per_iteration": 2.5997509956359863 + }, + { + "auxiliary_loss_clip": 0.01196004, + "auxiliary_loss_mlp": 0.01028612, + "balance_loss_clip": 1.06144667, + "balance_loss_mlp": 1.02008259, + "epoch": 0.384657007154452, + "flos": 25228395492480.0, + "grad_norm": 1.799097109315872, + "language_loss": 0.72947305, + "learning_rate": 2.8195413239527426e-06, + "loss": 0.75171924, + "num_input_tokens_seen": 68911275, + "step": 3199, + "time_per_iteration": 2.6423399448394775 + }, + { + "auxiliary_loss_clip": 0.01160408, + "auxiliary_loss_mlp": 0.01026121, + "balance_loss_clip": 1.01672435, + "balance_loss_mlp": 1.01824749, + "epoch": 0.38477725004509106, + "flos": 19865855358720.0, + "grad_norm": 2.0115324571423425, + "language_loss": 0.80846959, + "learning_rate": 2.8188306927683906e-06, + "loss": 0.83033484, + "num_input_tokens_seen": 68930745, + "step": 3200, + "time_per_iteration": 2.659195899963379 + }, + { + "auxiliary_loss_clip": 0.01146652, + "auxiliary_loss_mlp": 0.01026479, + "balance_loss_clip": 0.98381639, + "balance_loss_mlp": 1.01850367, + "epoch": 0.38489749293573017, + "flos": 18259391491200.0, + "grad_norm": 3.2023694383314174, + "language_loss": 0.74848318, + "learning_rate": 2.818119937368074e-06, + "loss": 0.77021444, + "num_input_tokens_seen": 68949380, + "step": 3201, + "time_per_iteration": 2.688812017440796 + }, + { + "auxiliary_loss_clip": 0.01171601, + "auxiliary_loss_mlp": 0.01033396, + "balance_loss_clip": 1.01854956, + "balance_loss_mlp": 1.02468824, + "epoch": 0.3850177358263693, + "flos": 24389163152640.0, + "grad_norm": 2.2089612813872352, + "language_loss": 0.65799785, + "learning_rate": 2.817409057859613e-06, + "loss": 0.68004781, + "num_input_tokens_seen": 68968370, + "step": 3202, + "time_per_iteration": 2.6643424034118652 + }, + { + "auxiliary_loss_clip": 0.01093574, + "auxiliary_loss_mlp": 0.01037191, + "balance_loss_clip": 0.90619457, + "balance_loss_mlp": 1.02824402, + "epoch": 0.38513797871700833, + "flos": 17671533505920.0, + "grad_norm": 1.8176916684983342, + "language_loss": 0.79521024, + "learning_rate": 2.8166980543508482e-06, + "loss": 0.81651795, + "num_input_tokens_seen": 68984260, + "step": 3203, + "time_per_iteration": 2.7684922218322754 + }, + { + "auxiliary_loss_clip": 0.01192724, + "auxiliary_loss_mlp": 0.01032035, + "balance_loss_clip": 1.05981445, + "balance_loss_mlp": 1.02399421, + "epoch": 0.38525822160764744, + "flos": 25739583897600.0, + "grad_norm": 2.2701130395694387, + "language_loss": 0.79981035, + "learning_rate": 2.815986926949638e-06, + "loss": 0.82205796, + "num_input_tokens_seen": 69002760, + "step": 3204, + "time_per_iteration": 2.704421281814575 + }, + { + "auxiliary_loss_clip": 0.01165858, + "auxiliary_loss_mlp": 0.01025393, + "balance_loss_clip": 1.02129722, + "balance_loss_mlp": 1.01737022, + "epoch": 0.38537846449828655, + "flos": 20193647898240.0, + "grad_norm": 1.6677034747161397, + "language_loss": 0.800753, + "learning_rate": 2.8152756757638597e-06, + "loss": 0.82266557, + "num_input_tokens_seen": 69021260, + "step": 3205, + "time_per_iteration": 2.6665761470794678 + }, + { + "auxiliary_loss_clip": 0.01169896, + "auxiliary_loss_mlp": 0.0102975, + "balance_loss_clip": 1.02438974, + "balance_loss_mlp": 1.02128661, + "epoch": 0.3854987073889256, + "flos": 23039352938880.0, + "grad_norm": 4.412133596726631, + "language_loss": 0.84791148, + "learning_rate": 2.8145643009014093e-06, + "loss": 0.86990792, + "num_input_tokens_seen": 69039755, + "step": 3206, + "time_per_iteration": 2.719675064086914 + }, + { + "auxiliary_loss_clip": 0.01170484, + "auxiliary_loss_mlp": 0.01034065, + "balance_loss_clip": 1.02272105, + "balance_loss_mlp": 1.0260067, + "epoch": 0.3856189502795647, + "flos": 20190631155840.0, + "grad_norm": 3.352590749325485, + "language_loss": 0.79303116, + "learning_rate": 2.813852802470202e-06, + "loss": 0.81507665, + "num_input_tokens_seen": 69057650, + "step": 3207, + "time_per_iteration": 2.670374631881714 + }, + { + "auxiliary_loss_clip": 0.01136382, + "auxiliary_loss_mlp": 0.01033008, + "balance_loss_clip": 0.9785434, + "balance_loss_mlp": 1.02425194, + "epoch": 0.38573919317020383, + "flos": 25702631781120.0, + "grad_norm": 1.7708217572967435, + "language_loss": 0.72083223, + "learning_rate": 2.8131411805781717e-06, + "loss": 0.74252617, + "num_input_tokens_seen": 69077775, + "step": 3208, + "time_per_iteration": 2.740169048309326 + }, + { + "auxiliary_loss_clip": 0.01144956, + "auxiliary_loss_mlp": 0.01030603, + "balance_loss_clip": 0.98395467, + "balance_loss_mlp": 1.02184725, + "epoch": 0.3858594360608429, + "flos": 29821405628160.0, + "grad_norm": 2.098366222005897, + "language_loss": 0.63975346, + "learning_rate": 2.8124294353332707e-06, + "loss": 0.66150904, + "num_input_tokens_seen": 69096450, + "step": 3209, + "time_per_iteration": 2.7258405685424805 + }, + { + "auxiliary_loss_clip": 0.01130062, + "auxiliary_loss_mlp": 0.01030226, + "balance_loss_clip": 0.94590318, + "balance_loss_mlp": 1.02172661, + "epoch": 0.385979678951482, + "flos": 24790428961920.0, + "grad_norm": 1.6232054966596003, + "language_loss": 0.77617228, + "learning_rate": 2.8117175668434713e-06, + "loss": 0.79777515, + "num_input_tokens_seen": 69116110, + "step": 3210, + "time_per_iteration": 2.8193960189819336 + }, + { + "auxiliary_loss_clip": 0.01192654, + "auxiliary_loss_mlp": 0.0102941, + "balance_loss_clip": 1.05978179, + "balance_loss_mlp": 1.02117229, + "epoch": 0.3860999218421211, + "flos": 21287881866240.0, + "grad_norm": 2.2244402543075825, + "language_loss": 0.70640385, + "learning_rate": 2.811005575216762e-06, + "loss": 0.72862446, + "num_input_tokens_seen": 69134825, + "step": 3211, + "time_per_iteration": 2.6128759384155273 + }, + { + "auxiliary_loss_clip": 0.0111136, + "auxiliary_loss_mlp": 0.01031588, + "balance_loss_clip": 0.94079608, + "balance_loss_mlp": 1.02293956, + "epoch": 0.38622016473276016, + "flos": 24536720223360.0, + "grad_norm": 1.6102462757570914, + "language_loss": 0.79067808, + "learning_rate": 2.8102934605611513e-06, + "loss": 0.81210762, + "num_input_tokens_seen": 69156460, + "step": 3212, + "time_per_iteration": 3.7536256313323975 + }, + { + "auxiliary_loss_clip": 0.01149989, + "auxiliary_loss_mlp": 0.01029737, + "balance_loss_clip": 0.98501301, + "balance_loss_mlp": 1.0218277, + "epoch": 0.3863404076233993, + "flos": 20558212986240.0, + "grad_norm": 2.0744687212024715, + "language_loss": 0.67585027, + "learning_rate": 2.8095812229846665e-06, + "loss": 0.69764757, + "num_input_tokens_seen": 69176420, + "step": 3213, + "time_per_iteration": 2.6916189193725586 + }, + { + "auxiliary_loss_clip": 0.01145663, + "auxiliary_loss_mlp": 0.01034007, + "balance_loss_clip": 0.9823904, + "balance_loss_mlp": 1.02526939, + "epoch": 0.3864606505140384, + "flos": 22346277039360.0, + "grad_norm": 2.2195611488281344, + "language_loss": 0.69307888, + "learning_rate": 2.808868862595355e-06, + "loss": 0.71487558, + "num_input_tokens_seen": 69196665, + "step": 3214, + "time_per_iteration": 3.652496814727783 + }, + { + "auxiliary_loss_clip": 0.01172355, + "auxiliary_loss_mlp": 0.01025608, + "balance_loss_clip": 1.02311587, + "balance_loss_mlp": 1.01759088, + "epoch": 0.38658089340467744, + "flos": 25703601448320.0, + "grad_norm": 1.9757714996023934, + "language_loss": 0.79349536, + "learning_rate": 2.8081563795012795e-06, + "loss": 0.81547499, + "num_input_tokens_seen": 69216290, + "step": 3215, + "time_per_iteration": 3.608198404312134 + }, + { + "auxiliary_loss_clip": 0.01151785, + "auxiliary_loss_mlp": 0.01031513, + "balance_loss_clip": 0.98206592, + "balance_loss_mlp": 1.02274513, + "epoch": 0.38670113629531655, + "flos": 33802534558080.0, + "grad_norm": 2.0975803987394825, + "language_loss": 0.74204481, + "learning_rate": 2.807443773810524e-06, + "loss": 0.76387781, + "num_input_tokens_seen": 69237550, + "step": 3216, + "time_per_iteration": 2.795090436935425 + }, + { + "auxiliary_loss_clip": 0.01124013, + "auxiliary_loss_mlp": 0.01039411, + "balance_loss_clip": 0.94821614, + "balance_loss_mlp": 1.03095317, + "epoch": 0.3868213791859556, + "flos": 23331522165120.0, + "grad_norm": 1.8026143211153924, + "language_loss": 0.89691228, + "learning_rate": 2.80673104563119e-06, + "loss": 0.91854656, + "num_input_tokens_seen": 69258175, + "step": 3217, + "time_per_iteration": 2.7220497131347656 + }, + { + "auxiliary_loss_clip": 0.01166234, + "auxiliary_loss_mlp": 0.01026763, + "balance_loss_clip": 1.02193904, + "balance_loss_mlp": 1.01870489, + "epoch": 0.3869416220765947, + "flos": 18441530380800.0, + "grad_norm": 1.9013107910026674, + "language_loss": 0.78763092, + "learning_rate": 2.8060181950713976e-06, + "loss": 0.80956089, + "num_input_tokens_seen": 69274965, + "step": 3218, + "time_per_iteration": 2.6849868297576904 + }, + { + "auxiliary_loss_clip": 0.0111846, + "auxiliary_loss_mlp": 0.01034585, + "balance_loss_clip": 0.94189507, + "balance_loss_mlp": 1.02549529, + "epoch": 0.3870618649672338, + "flos": 15632992938240.0, + "grad_norm": 2.024359208190718, + "language_loss": 0.80973476, + "learning_rate": 2.805305222239286e-06, + "loss": 0.83126521, + "num_input_tokens_seen": 69292220, + "step": 3219, + "time_per_iteration": 2.7828550338745117 + }, + { + "auxiliary_loss_clip": 0.01145758, + "auxiliary_loss_mlp": 0.01030273, + "balance_loss_clip": 0.98633718, + "balance_loss_mlp": 1.02160668, + "epoch": 0.3871821078578729, + "flos": 23513804709120.0, + "grad_norm": 1.8516091324429484, + "language_loss": 0.73800898, + "learning_rate": 2.8045921272430118e-06, + "loss": 0.75976926, + "num_input_tokens_seen": 69311900, + "step": 3220, + "time_per_iteration": 2.774848461151123 + }, + { + "auxiliary_loss_clip": 0.01174299, + "auxiliary_loss_mlp": 0.01029142, + "balance_loss_clip": 1.01966703, + "balance_loss_mlp": 1.02048087, + "epoch": 0.387302350748512, + "flos": 17778259791360.0, + "grad_norm": 2.0897217880425005, + "language_loss": 0.76959193, + "learning_rate": 2.803878910190753e-06, + "loss": 0.79162633, + "num_input_tokens_seen": 69328820, + "step": 3221, + "time_per_iteration": 2.6645090579986572 + }, + { + "auxiliary_loss_clip": 0.01174132, + "auxiliary_loss_mlp": 0.01028675, + "balance_loss_clip": 1.02203107, + "balance_loss_mlp": 1.02043772, + "epoch": 0.3874225936391511, + "flos": 11503409097600.0, + "grad_norm": 2.5785865993331183, + "language_loss": 0.81966293, + "learning_rate": 2.8031655711907017e-06, + "loss": 0.84169102, + "num_input_tokens_seen": 69342525, + "step": 3222, + "time_per_iteration": 3.533282995223999 + }, + { + "auxiliary_loss_clip": 0.01173651, + "auxiliary_loss_mlp": 0.01031158, + "balance_loss_clip": 1.02415001, + "balance_loss_mlp": 1.0230577, + "epoch": 0.38754283652979016, + "flos": 21945154884480.0, + "grad_norm": 7.004050555079, + "language_loss": 0.80935889, + "learning_rate": 2.8024521103510723e-06, + "loss": 0.83140695, + "num_input_tokens_seen": 69359295, + "step": 3223, + "time_per_iteration": 2.6959948539733887 + }, + { + "auxiliary_loss_clip": 0.01165711, + "auxiliary_loss_mlp": 0.01030857, + "balance_loss_clip": 1.01786244, + "balance_loss_mlp": 1.02216673, + "epoch": 0.38766307942042927, + "flos": 21175984022400.0, + "grad_norm": 2.0967755353665956, + "language_loss": 0.75279117, + "learning_rate": 2.8017385277800952e-06, + "loss": 0.77475685, + "num_input_tokens_seen": 69377650, + "step": 3224, + "time_per_iteration": 2.6384477615356445 + }, + { + "auxiliary_loss_clip": 0.01124758, + "auxiliary_loss_mlp": 0.01030669, + "balance_loss_clip": 0.9445712, + "balance_loss_mlp": 1.02196097, + "epoch": 0.3877833223110684, + "flos": 27417294391680.0, + "grad_norm": 2.3734160740656165, + "language_loss": 0.75269341, + "learning_rate": 2.8010248235860213e-06, + "loss": 0.77424771, + "num_input_tokens_seen": 69397765, + "step": 3225, + "time_per_iteration": 2.774060010910034 + }, + { + "auxiliary_loss_clip": 0.0105567, + "auxiliary_loss_mlp": 0.00501566, + "balance_loss_clip": 0.95294797, + "balance_loss_mlp": 0.0, + "epoch": 0.38790356520170743, + "flos": 64500019879680.0, + "grad_norm": 0.8329387593855645, + "language_loss": 0.62807679, + "learning_rate": 2.8003109978771192e-06, + "loss": 0.64364922, + "num_input_tokens_seen": 69458930, + "step": 3226, + "time_per_iteration": 3.3150033950805664 + }, + { + "auxiliary_loss_clip": 0.01110508, + "auxiliary_loss_mlp": 0.01023733, + "balance_loss_clip": 0.93884313, + "balance_loss_mlp": 1.01542997, + "epoch": 0.38802380809234654, + "flos": 22345415112960.0, + "grad_norm": 1.9935573785292287, + "language_loss": 0.78875101, + "learning_rate": 2.799597050761674e-06, + "loss": 0.81009346, + "num_input_tokens_seen": 69475135, + "step": 3227, + "time_per_iteration": 2.736877679824829 + }, + { + "auxiliary_loss_clip": 0.01194519, + "auxiliary_loss_mlp": 0.01031507, + "balance_loss_clip": 1.06056619, + "balance_loss_mlp": 1.02284026, + "epoch": 0.38814405098298566, + "flos": 25261361199360.0, + "grad_norm": 1.8542207664835277, + "language_loss": 0.7923193, + "learning_rate": 2.7988829823479924e-06, + "loss": 0.81457949, + "num_input_tokens_seen": 69493525, + "step": 3228, + "time_per_iteration": 2.633843183517456 + }, + { + "auxiliary_loss_clip": 0.01137843, + "auxiliary_loss_mlp": 0.01031004, + "balance_loss_clip": 0.97918046, + "balance_loss_mlp": 1.02228379, + "epoch": 0.3882642938736247, + "flos": 18841180078080.0, + "grad_norm": 2.958543627126311, + "language_loss": 0.63917387, + "learning_rate": 2.7981687927443976e-06, + "loss": 0.66086233, + "num_input_tokens_seen": 69510325, + "step": 3229, + "time_per_iteration": 2.687547206878662 + }, + { + "auxiliary_loss_clip": 0.01166802, + "auxiliary_loss_mlp": 0.01029266, + "balance_loss_clip": 1.0182476, + "balance_loss_mlp": 1.02076602, + "epoch": 0.3883845367642638, + "flos": 21652806090240.0, + "grad_norm": 2.880142928443323, + "language_loss": 0.85565108, + "learning_rate": 2.797454482059231e-06, + "loss": 0.87761176, + "num_input_tokens_seen": 69530480, + "step": 3230, + "time_per_iteration": 2.6908211708068848 + }, + { + "auxiliary_loss_clip": 0.01195723, + "auxiliary_loss_mlp": 0.01028445, + "balance_loss_clip": 1.0621736, + "balance_loss_mlp": 1.01986241, + "epoch": 0.3885047796549029, + "flos": 20557530627840.0, + "grad_norm": 1.8322944966487047, + "language_loss": 0.84294224, + "learning_rate": 2.7967400504008537e-06, + "loss": 0.86518395, + "num_input_tokens_seen": 69549780, + "step": 3231, + "time_per_iteration": 2.630692720413208 + }, + { + "auxiliary_loss_clip": 0.01009851, + "auxiliary_loss_mlp": 0.01015165, + "balance_loss_clip": 0.87886924, + "balance_loss_mlp": 1.01365709, + "epoch": 0.388625022545542, + "flos": 64325491695360.0, + "grad_norm": 0.7931042260148091, + "language_loss": 0.57483625, + "learning_rate": 2.7960254978776456e-06, + "loss": 0.59508634, + "num_input_tokens_seen": 69611870, + "step": 3232, + "time_per_iteration": 3.3235552310943604 + }, + { + "auxiliary_loss_clip": 0.01194265, + "auxiliary_loss_mlp": 0.01031109, + "balance_loss_clip": 1.06077611, + "balance_loss_mlp": 1.0227108, + "epoch": 0.3887452654361811, + "flos": 18113881495680.0, + "grad_norm": 2.1826361596829384, + "language_loss": 0.81476778, + "learning_rate": 2.7953108245980006e-06, + "loss": 0.83702147, + "num_input_tokens_seen": 69630385, + "step": 3233, + "time_per_iteration": 2.5628550052642822 + }, + { + "auxiliary_loss_clip": 0.01143084, + "auxiliary_loss_mlp": 0.01031867, + "balance_loss_clip": 0.98560113, + "balance_loss_mlp": 1.02382612, + "epoch": 0.38886550832682015, + "flos": 24975261371520.0, + "grad_norm": 1.6398437053176986, + "language_loss": 0.73983735, + "learning_rate": 2.7945960306703365e-06, + "loss": 0.7615869, + "num_input_tokens_seen": 69653370, + "step": 3234, + "time_per_iteration": 2.797668695449829 + }, + { + "auxiliary_loss_clip": 0.01172245, + "auxiliary_loss_mlp": 0.01030839, + "balance_loss_clip": 1.02138233, + "balance_loss_mlp": 1.02270937, + "epoch": 0.38898575121745926, + "flos": 27199496275200.0, + "grad_norm": 2.214555258381758, + "language_loss": 0.65872341, + "learning_rate": 2.7938811162030865e-06, + "loss": 0.68075424, + "num_input_tokens_seen": 69673635, + "step": 3235, + "time_per_iteration": 2.7180402278900146 + }, + { + "auxiliary_loss_clip": 0.01166528, + "auxiliary_loss_mlp": 0.01025631, + "balance_loss_clip": 1.02139032, + "balance_loss_mlp": 1.01769137, + "epoch": 0.3891059941080984, + "flos": 28763728727040.0, + "grad_norm": 2.087590612647001, + "language_loss": 0.82366037, + "learning_rate": 2.793166081304702e-06, + "loss": 0.84558201, + "num_input_tokens_seen": 69694130, + "step": 3236, + "time_per_iteration": 2.708270788192749 + }, + { + "auxiliary_loss_clip": 0.01124713, + "auxiliary_loss_mlp": 0.01028998, + "balance_loss_clip": 0.94218457, + "balance_loss_mlp": 1.02016473, + "epoch": 0.38922623699873743, + "flos": 22893447893760.0, + "grad_norm": 1.8543352366665447, + "language_loss": 0.82448411, + "learning_rate": 2.7924509260836543e-06, + "loss": 0.84602129, + "num_input_tokens_seen": 69713255, + "step": 3237, + "time_per_iteration": 2.72711181640625 + }, + { + "auxiliary_loss_clip": 0.01118433, + "auxiliary_loss_mlp": 0.01027941, + "balance_loss_clip": 0.94097227, + "balance_loss_mlp": 1.0196861, + "epoch": 0.38934647988937654, + "flos": 19792418002560.0, + "grad_norm": 2.1698667119941013, + "language_loss": 0.68338156, + "learning_rate": 2.791735650648431e-06, + "loss": 0.70484531, + "num_input_tokens_seen": 69732375, + "step": 3238, + "time_per_iteration": 2.7089364528656006 + }, + { + "auxiliary_loss_clip": 0.01140582, + "auxiliary_loss_mlp": 0.01028215, + "balance_loss_clip": 0.98023665, + "balance_loss_mlp": 1.02005506, + "epoch": 0.38946672278001565, + "flos": 19202081978880.0, + "grad_norm": 1.8387395571759935, + "language_loss": 0.74617362, + "learning_rate": 2.791020255107538e-06, + "loss": 0.76786155, + "num_input_tokens_seen": 69749745, + "step": 3239, + "time_per_iteration": 3.5848968029022217 + }, + { + "auxiliary_loss_clip": 0.01113571, + "auxiliary_loss_mlp": 0.01032424, + "balance_loss_clip": 0.93845761, + "balance_loss_mlp": 1.02385259, + "epoch": 0.3895869656706547, + "flos": 24936477661440.0, + "grad_norm": 1.5322988478850785, + "language_loss": 0.80809093, + "learning_rate": 2.7903047395695023e-06, + "loss": 0.82955086, + "num_input_tokens_seen": 69769645, + "step": 3240, + "time_per_iteration": 3.667451858520508 + }, + { + "auxiliary_loss_clip": 0.01166591, + "auxiliary_loss_mlp": 0.0051071, + "balance_loss_clip": 1.02129889, + "balance_loss_mlp": 0.0, + "epoch": 0.3897072085612938, + "flos": 24133622820480.0, + "grad_norm": 2.056092149670572, + "language_loss": 0.90098977, + "learning_rate": 2.789589104142865e-06, + "loss": 0.91776276, + "num_input_tokens_seen": 69787270, + "step": 3241, + "time_per_iteration": 3.66977596282959 + }, + { + "auxiliary_loss_clip": 0.01119388, + "auxiliary_loss_mlp": 0.01028659, + "balance_loss_clip": 0.94249982, + "balance_loss_mlp": 1.02018905, + "epoch": 0.3898274514519329, + "flos": 17166342672000.0, + "grad_norm": 1.7779984380742058, + "language_loss": 0.76629806, + "learning_rate": 2.7888733489361895e-06, + "loss": 0.7877785, + "num_input_tokens_seen": 69805685, + "step": 3242, + "time_per_iteration": 2.698704481124878 + }, + { + "auxiliary_loss_clip": 0.01091034, + "auxiliary_loss_mlp": 0.01006638, + "balance_loss_clip": 1.02692556, + "balance_loss_mlp": 1.00518346, + "epoch": 0.389947694342572, + "flos": 66074807952000.0, + "grad_norm": 0.7276481999164447, + "language_loss": 0.58725011, + "learning_rate": 2.788157474058054e-06, + "loss": 0.6082269, + "num_input_tokens_seen": 69867960, + "step": 3243, + "time_per_iteration": 3.2961952686309814 + }, + { + "auxiliary_loss_clip": 0.01185111, + "auxiliary_loss_mlp": 0.01028286, + "balance_loss_clip": 1.05570292, + "balance_loss_mlp": 1.02071595, + "epoch": 0.3900679372332111, + "flos": 25740912700800.0, + "grad_norm": 1.5315450191038131, + "language_loss": 0.70039082, + "learning_rate": 2.7874414796170555e-06, + "loss": 0.72252482, + "num_input_tokens_seen": 69889450, + "step": 3244, + "time_per_iteration": 2.6457488536834717 + }, + { + "auxiliary_loss_clip": 0.01162613, + "auxiliary_loss_mlp": 0.01033917, + "balance_loss_clip": 1.01820397, + "balance_loss_mlp": 1.02470207, + "epoch": 0.3901881801238502, + "flos": 11801611808640.0, + "grad_norm": 2.4369793531990576, + "language_loss": 0.83934188, + "learning_rate": 2.7867253657218113e-06, + "loss": 0.8613072, + "num_input_tokens_seen": 69903340, + "step": 3245, + "time_per_iteration": 2.6091182231903076 + }, + { + "auxiliary_loss_clip": 0.01140851, + "auxiliary_loss_mlp": 0.00510473, + "balance_loss_clip": 0.97878587, + "balance_loss_mlp": 0.0, + "epoch": 0.39030842301448926, + "flos": 27308951994240.0, + "grad_norm": 1.6958072235323447, + "language_loss": 0.73173851, + "learning_rate": 2.7860091324809544e-06, + "loss": 0.7482518, + "num_input_tokens_seen": 69924400, + "step": 3246, + "time_per_iteration": 2.7466681003570557 + }, + { + "auxiliary_loss_clip": 0.01163282, + "auxiliary_loss_mlp": 0.01028315, + "balance_loss_clip": 1.02053452, + "balance_loss_mlp": 1.01994658, + "epoch": 0.39042866590512837, + "flos": 27163334257920.0, + "grad_norm": 1.7296594787592272, + "language_loss": 0.80925643, + "learning_rate": 2.7852927800031377e-06, + "loss": 0.83117241, + "num_input_tokens_seen": 69944565, + "step": 3247, + "time_per_iteration": 2.700453519821167 + }, + { + "auxiliary_loss_clip": 0.01141418, + "auxiliary_loss_mlp": 0.01026991, + "balance_loss_clip": 0.9791072, + "balance_loss_mlp": 1.01906335, + "epoch": 0.3905489087957674, + "flos": 29716115886720.0, + "grad_norm": 1.6595406122480458, + "language_loss": 0.82893324, + "learning_rate": 2.7845763083970298e-06, + "loss": 0.85061741, + "num_input_tokens_seen": 69964965, + "step": 3248, + "time_per_iteration": 3.66330623626709 + }, + { + "auxiliary_loss_clip": 0.01157484, + "auxiliary_loss_mlp": 0.0102549, + "balance_loss_clip": 1.01638794, + "balance_loss_mlp": 1.01684463, + "epoch": 0.39066915168640653, + "flos": 24498618871680.0, + "grad_norm": 1.8075278992342745, + "language_loss": 0.82544738, + "learning_rate": 2.7838597177713205e-06, + "loss": 0.84727716, + "num_input_tokens_seen": 69986055, + "step": 3249, + "time_per_iteration": 2.7221860885620117 + }, + { + "auxiliary_loss_clip": 0.01062199, + "auxiliary_loss_mlp": 0.01031365, + "balance_loss_clip": 0.86842936, + "balance_loss_mlp": 1.02305019, + "epoch": 0.39078939457704565, + "flos": 20558572122240.0, + "grad_norm": 2.7961217425112883, + "language_loss": 0.73281401, + "learning_rate": 2.7831430082347143e-06, + "loss": 0.75374961, + "num_input_tokens_seen": 70005260, + "step": 3250, + "time_per_iteration": 2.7740108966827393 + }, + { + "auxiliary_loss_clip": 0.01167755, + "auxiliary_loss_mlp": 0.00509671, + "balance_loss_clip": 1.02095926, + "balance_loss_mlp": 0.0, + "epoch": 0.3909096374676847, + "flos": 22783417557120.0, + "grad_norm": 2.086702179735773, + "language_loss": 0.82547092, + "learning_rate": 2.7824261798959373e-06, + "loss": 0.84224522, + "num_input_tokens_seen": 70023440, + "step": 3251, + "time_per_iteration": 2.681685209274292 + }, + { + "auxiliary_loss_clip": 0.01142262, + "auxiliary_loss_mlp": 0.01033757, + "balance_loss_clip": 0.97671163, + "balance_loss_mlp": 1.02527547, + "epoch": 0.3910298803583238, + "flos": 23003119094400.0, + "grad_norm": 2.1145817833930334, + "language_loss": 0.79690731, + "learning_rate": 2.78170923286373e-06, + "loss": 0.81866747, + "num_input_tokens_seen": 70043040, + "step": 3252, + "time_per_iteration": 2.6658928394317627 + }, + { + "auxiliary_loss_clip": 0.01042997, + "auxiliary_loss_mlp": 0.01027037, + "balance_loss_clip": 0.82976228, + "balance_loss_mlp": 1.01868653, + "epoch": 0.3911501232489629, + "flos": 24316264500480.0, + "grad_norm": 2.2939147738029853, + "language_loss": 0.83920574, + "learning_rate": 2.780992167246854e-06, + "loss": 0.85990602, + "num_input_tokens_seen": 70060565, + "step": 3253, + "time_per_iteration": 2.834799289703369 + }, + { + "auxiliary_loss_clip": 0.01053997, + "auxiliary_loss_mlp": 0.01009353, + "balance_loss_clip": 0.95261073, + "balance_loss_mlp": 1.00793481, + "epoch": 0.391270366139602, + "flos": 60869054684160.0, + "grad_norm": 0.98995613928561, + "language_loss": 0.7215957, + "learning_rate": 2.7802749831540883e-06, + "loss": 0.74222922, + "num_input_tokens_seen": 70119465, + "step": 3254, + "time_per_iteration": 3.284729480743408 + }, + { + "auxiliary_loss_clip": 0.01098952, + "auxiliary_loss_mlp": 0.01028082, + "balance_loss_clip": 0.90294689, + "balance_loss_mlp": 1.02063739, + "epoch": 0.3913906090302411, + "flos": 21543494025600.0, + "grad_norm": 1.8746726664524223, + "language_loss": 0.81915694, + "learning_rate": 2.7795576806942268e-06, + "loss": 0.84042728, + "num_input_tokens_seen": 70138270, + "step": 3255, + "time_per_iteration": 2.6737470626831055 + }, + { + "auxiliary_loss_clip": 0.0104411, + "auxiliary_loss_mlp": 0.01011178, + "balance_loss_clip": 0.95443672, + "balance_loss_mlp": 1.00979543, + "epoch": 0.3915108519208802, + "flos": 49839953702400.0, + "grad_norm": 0.7672546158585051, + "language_loss": 0.54938811, + "learning_rate": 2.778840259976085e-06, + "loss": 0.56994098, + "num_input_tokens_seen": 70193500, + "step": 3256, + "time_per_iteration": 3.2247276306152344 + }, + { + "auxiliary_loss_clip": 0.01165097, + "auxiliary_loss_mlp": 0.01033846, + "balance_loss_clip": 1.01874948, + "balance_loss_mlp": 1.02553093, + "epoch": 0.39163109481151925, + "flos": 16506447960960.0, + "grad_norm": 2.143425642265949, + "language_loss": 0.76947975, + "learning_rate": 2.778122721108495e-06, + "loss": 0.79146922, + "num_input_tokens_seen": 70211730, + "step": 3257, + "time_per_iteration": 2.6370065212249756 + }, + { + "auxiliary_loss_clip": 0.01164123, + "auxiliary_loss_mlp": 0.01034041, + "balance_loss_clip": 1.02034593, + "balance_loss_mlp": 1.02621818, + "epoch": 0.39175133770215836, + "flos": 26067484177920.0, + "grad_norm": 2.022617691672486, + "language_loss": 0.88651061, + "learning_rate": 2.7774050642003076e-06, + "loss": 0.90849233, + "num_input_tokens_seen": 70232540, + "step": 3258, + "time_per_iteration": 2.6987152099609375 + }, + { + "auxiliary_loss_clip": 0.01190893, + "auxiliary_loss_mlp": 0.01031378, + "balance_loss_clip": 1.05719411, + "balance_loss_mlp": 1.02296185, + "epoch": 0.3918715805927975, + "flos": 21872076664320.0, + "grad_norm": 2.2353380488413634, + "language_loss": 0.93594527, + "learning_rate": 2.7766872893603896e-06, + "loss": 0.95816797, + "num_input_tokens_seen": 70252515, + "step": 3259, + "time_per_iteration": 2.652977705001831 + }, + { + "auxiliary_loss_clip": 0.01165924, + "auxiliary_loss_mlp": 0.01029755, + "balance_loss_clip": 1.01919675, + "balance_loss_mlp": 1.02213788, + "epoch": 0.39199182348343653, + "flos": 20376181837440.0, + "grad_norm": 1.6335983241429857, + "language_loss": 0.73321623, + "learning_rate": 2.7759693966976275e-06, + "loss": 0.75517303, + "num_input_tokens_seen": 70271020, + "step": 3260, + "time_per_iteration": 2.6749556064605713 + }, + { + "auxiliary_loss_clip": 0.01116315, + "auxiliary_loss_mlp": 0.01035353, + "balance_loss_clip": 0.94139761, + "balance_loss_mlp": 1.0265435, + "epoch": 0.39211206637407564, + "flos": 21683545153920.0, + "grad_norm": 1.868431612250564, + "language_loss": 0.85232115, + "learning_rate": 2.7752513863209242e-06, + "loss": 0.87383777, + "num_input_tokens_seen": 70289600, + "step": 3261, + "time_per_iteration": 2.728348731994629 + }, + { + "auxiliary_loss_clip": 0.01138344, + "auxiliary_loss_mlp": 0.00509705, + "balance_loss_clip": 0.98421115, + "balance_loss_mlp": 0.0, + "epoch": 0.39223230926471475, + "flos": 21066276908160.0, + "grad_norm": 1.6747390533549014, + "language_loss": 0.84655917, + "learning_rate": 2.774533258339203e-06, + "loss": 0.86303961, + "num_input_tokens_seen": 70307060, + "step": 3262, + "time_per_iteration": 2.7088427543640137 + }, + { + "auxiliary_loss_clip": 0.01093282, + "auxiliary_loss_mlp": 0.01031183, + "balance_loss_clip": 0.89854395, + "balance_loss_mlp": 1.0226239, + "epoch": 0.3923525521553538, + "flos": 17603016312960.0, + "grad_norm": 2.5591813248240616, + "language_loss": 0.79736459, + "learning_rate": 2.7738150128614014e-06, + "loss": 0.81860924, + "num_input_tokens_seen": 70324465, + "step": 3263, + "time_per_iteration": 2.7571003437042236 + }, + { + "auxiliary_loss_clip": 0.01107919, + "auxiliary_loss_mlp": 0.01031005, + "balance_loss_clip": 0.94124091, + "balance_loss_mlp": 1.02314353, + "epoch": 0.3924727950459929, + "flos": 20558284813440.0, + "grad_norm": 1.8554857908778335, + "language_loss": 0.89548469, + "learning_rate": 2.7730966499964777e-06, + "loss": 0.91687393, + "num_input_tokens_seen": 70341415, + "step": 3264, + "time_per_iteration": 3.643648624420166 + }, + { + "auxiliary_loss_clip": 0.01188143, + "auxiliary_loss_mlp": 0.01033127, + "balance_loss_clip": 1.05502152, + "balance_loss_mlp": 1.02511001, + "epoch": 0.39259303793663197, + "flos": 16216110328320.0, + "grad_norm": 2.5204414621909423, + "language_loss": 0.80661881, + "learning_rate": 2.772378169853408e-06, + "loss": 0.82883149, + "num_input_tokens_seen": 70358985, + "step": 3265, + "time_per_iteration": 2.653811454772949 + }, + { + "auxiliary_loss_clip": 0.0111789, + "auxiliary_loss_mlp": 0.01035695, + "balance_loss_clip": 0.94259375, + "balance_loss_mlp": 1.02821493, + "epoch": 0.3927132808272711, + "flos": 16797001075200.0, + "grad_norm": 2.0589041669256005, + "language_loss": 0.74479795, + "learning_rate": 2.771659572541183e-06, + "loss": 0.76633376, + "num_input_tokens_seen": 70376915, + "step": 3266, + "time_per_iteration": 3.624028205871582 + }, + { + "auxiliary_loss_clip": 0.01171114, + "auxiliary_loss_mlp": 0.01026916, + "balance_loss_clip": 1.02327132, + "balance_loss_mlp": 1.0186727, + "epoch": 0.3928335237179102, + "flos": 20267228908800.0, + "grad_norm": 1.9362056166230552, + "language_loss": 0.86785614, + "learning_rate": 2.7709408581688143e-06, + "loss": 0.88983643, + "num_input_tokens_seen": 70396900, + "step": 3267, + "time_per_iteration": 3.564037799835205 + }, + { + "auxiliary_loss_clip": 0.01123375, + "auxiliary_loss_mlp": 0.01036716, + "balance_loss_clip": 0.94200003, + "balance_loss_mlp": 1.02884793, + "epoch": 0.39295376660854925, + "flos": 24973250209920.0, + "grad_norm": 1.7020944070552824, + "language_loss": 0.88023847, + "learning_rate": 2.7702220268453307e-06, + "loss": 0.90183938, + "num_input_tokens_seen": 70417260, + "step": 3268, + "time_per_iteration": 2.7787981033325195 + }, + { + "auxiliary_loss_clip": 0.01143162, + "auxiliary_loss_mlp": 0.01028693, + "balance_loss_clip": 0.98016518, + "balance_loss_mlp": 1.02087283, + "epoch": 0.39307400949918836, + "flos": 18697788984960.0, + "grad_norm": 2.406024989718521, + "language_loss": 0.84664047, + "learning_rate": 2.7695030786797785e-06, + "loss": 0.86835903, + "num_input_tokens_seen": 70433155, + "step": 3269, + "time_per_iteration": 2.653409004211426 + }, + { + "auxiliary_loss_clip": 0.01091187, + "auxiliary_loss_mlp": 0.01027066, + "balance_loss_clip": 0.90192711, + "balance_loss_mlp": 1.01929951, + "epoch": 0.39319425238982747, + "flos": 22415476590720.0, + "grad_norm": 2.094235104378519, + "language_loss": 0.74616331, + "learning_rate": 2.7687840137812206e-06, + "loss": 0.76734585, + "num_input_tokens_seen": 70451240, + "step": 3270, + "time_per_iteration": 2.7746939659118652 + }, + { + "auxiliary_loss_clip": 0.01064069, + "auxiliary_loss_mlp": 0.01011978, + "balance_loss_clip": 0.98917162, + "balance_loss_mlp": 1.01060069, + "epoch": 0.3933144952804665, + "flos": 66192954762240.0, + "grad_norm": 0.802640545533099, + "language_loss": 0.62064224, + "learning_rate": 2.7680648322587395e-06, + "loss": 0.64140266, + "num_input_tokens_seen": 70516115, + "step": 3271, + "time_per_iteration": 3.2513511180877686 + }, + { + "auxiliary_loss_clip": 0.01186251, + "auxiliary_loss_mlp": 0.01026661, + "balance_loss_clip": 1.05633438, + "balance_loss_mlp": 1.01849484, + "epoch": 0.39343473817110564, + "flos": 15487159720320.0, + "grad_norm": 2.813113922781051, + "language_loss": 0.8085568, + "learning_rate": 2.7673455342214334e-06, + "loss": 0.83068591, + "num_input_tokens_seen": 70533105, + "step": 3272, + "time_per_iteration": 2.583488941192627 + }, + { + "auxiliary_loss_clip": 0.01164721, + "auxiliary_loss_mlp": 0.01028913, + "balance_loss_clip": 1.0197289, + "balance_loss_mlp": 1.02053893, + "epoch": 0.39355498106174475, + "flos": 21324905809920.0, + "grad_norm": 1.9667837155546808, + "language_loss": 0.76191044, + "learning_rate": 2.7666261197784198e-06, + "loss": 0.78384674, + "num_input_tokens_seen": 70551920, + "step": 3273, + "time_per_iteration": 3.5513617992401123 + }, + { + "auxiliary_loss_clip": 0.0114025, + "auxiliary_loss_mlp": 0.01030227, + "balance_loss_clip": 0.9837417, + "balance_loss_mlp": 1.02282977, + "epoch": 0.3936752239523838, + "flos": 13296357400320.0, + "grad_norm": 1.9161862762620252, + "language_loss": 0.76561582, + "learning_rate": 2.7659065890388336e-06, + "loss": 0.78732061, + "num_input_tokens_seen": 70567920, + "step": 3274, + "time_per_iteration": 2.658512830734253 + }, + { + "auxiliary_loss_clip": 0.01142241, + "auxiliary_loss_mlp": 0.01030364, + "balance_loss_clip": 0.97943997, + "balance_loss_mlp": 1.0218997, + "epoch": 0.3937954668430229, + "flos": 16800161472000.0, + "grad_norm": 1.9611250042767643, + "language_loss": 0.84793538, + "learning_rate": 2.7651869421118266e-06, + "loss": 0.86966145, + "num_input_tokens_seen": 70584530, + "step": 3275, + "time_per_iteration": 2.775085926055908 + }, + { + "auxiliary_loss_clip": 0.01170288, + "auxiliary_loss_mlp": 0.01027998, + "balance_loss_clip": 1.0226891, + "balance_loss_mlp": 1.02001953, + "epoch": 0.393915709733662, + "flos": 21064229832960.0, + "grad_norm": 1.702047806278515, + "language_loss": 0.83028018, + "learning_rate": 2.76446717910657e-06, + "loss": 0.85226303, + "num_input_tokens_seen": 70605235, + "step": 3276, + "time_per_iteration": 2.6724116802215576 + }, + { + "auxiliary_loss_clip": 0.01161021, + "auxiliary_loss_mlp": 0.01030578, + "balance_loss_clip": 1.01749468, + "balance_loss_mlp": 1.02210844, + "epoch": 0.3940359526243011, + "flos": 17165265264000.0, + "grad_norm": 3.40659542886455, + "language_loss": 0.76903665, + "learning_rate": 2.763747300132249e-06, + "loss": 0.79095256, + "num_input_tokens_seen": 70622675, + "step": 3277, + "time_per_iteration": 2.665009021759033 + }, + { + "auxiliary_loss_clip": 0.01189359, + "auxiliary_loss_mlp": 0.01026569, + "balance_loss_clip": 1.05844629, + "balance_loss_mlp": 1.01864767, + "epoch": 0.3941561955149402, + "flos": 20995856294400.0, + "grad_norm": 1.8603939391925048, + "language_loss": 0.86685985, + "learning_rate": 2.7630273052980704e-06, + "loss": 0.88901913, + "num_input_tokens_seen": 70643265, + "step": 3278, + "time_per_iteration": 2.644563913345337 + }, + { + "auxiliary_loss_clip": 0.01133484, + "auxiliary_loss_mlp": 0.01031885, + "balance_loss_clip": 0.98091424, + "balance_loss_mlp": 1.02422285, + "epoch": 0.39427643840557924, + "flos": 18843406721280.0, + "grad_norm": 2.768790615512998, + "language_loss": 0.67001742, + "learning_rate": 2.7623071947132554e-06, + "loss": 0.69167113, + "num_input_tokens_seen": 70660295, + "step": 3279, + "time_per_iteration": 2.6545886993408203 + }, + { + "auxiliary_loss_clip": 0.01149634, + "auxiliary_loss_mlp": 0.01031131, + "balance_loss_clip": 0.98105085, + "balance_loss_mlp": 1.02210116, + "epoch": 0.39439668129621835, + "flos": 23258659426560.0, + "grad_norm": 2.2683812656506763, + "language_loss": 0.78769875, + "learning_rate": 2.7615869684870458e-06, + "loss": 0.80950636, + "num_input_tokens_seen": 70679605, + "step": 3280, + "time_per_iteration": 2.6930792331695557 + }, + { + "auxiliary_loss_clip": 0.01162986, + "auxiliary_loss_mlp": 0.01023969, + "balance_loss_clip": 1.02069855, + "balance_loss_mlp": 1.01601219, + "epoch": 0.39451692418685746, + "flos": 26652289507200.0, + "grad_norm": 1.7706339060960985, + "language_loss": 0.84485173, + "learning_rate": 2.7608666267286986e-06, + "loss": 0.86672127, + "num_input_tokens_seen": 70699835, + "step": 3281, + "time_per_iteration": 2.7209203243255615 + }, + { + "auxiliary_loss_clip": 0.01062762, + "auxiliary_loss_mlp": 0.01029043, + "balance_loss_clip": 0.8586272, + "balance_loss_mlp": 1.02071011, + "epoch": 0.3946371670774965, + "flos": 18258709132800.0, + "grad_norm": 2.323797415585736, + "language_loss": 0.86699522, + "learning_rate": 2.760146169547489e-06, + "loss": 0.88791323, + "num_input_tokens_seen": 70716600, + "step": 3282, + "time_per_iteration": 2.7896082401275635 + }, + { + "auxiliary_loss_clip": 0.01144377, + "auxiliary_loss_mlp": 0.01032471, + "balance_loss_clip": 0.98422343, + "balance_loss_mlp": 1.02454376, + "epoch": 0.39475740996813563, + "flos": 24206126423040.0, + "grad_norm": 1.4075871825605606, + "language_loss": 0.76467741, + "learning_rate": 2.75942559705271e-06, + "loss": 0.78644586, + "num_input_tokens_seen": 70736335, + "step": 3283, + "time_per_iteration": 2.691218852996826 + }, + { + "auxiliary_loss_clip": 0.01163581, + "auxiliary_loss_mlp": 0.01027387, + "balance_loss_clip": 1.01971388, + "balance_loss_mlp": 1.01919091, + "epoch": 0.39487765285877474, + "flos": 19317858491520.0, + "grad_norm": 1.961318063116863, + "language_loss": 0.89291275, + "learning_rate": 2.7587049093536713e-06, + "loss": 0.91482246, + "num_input_tokens_seen": 70752665, + "step": 3284, + "time_per_iteration": 2.6424763202667236 + }, + { + "auxiliary_loss_clip": 0.0116903, + "auxiliary_loss_mlp": 0.0102839, + "balance_loss_clip": 1.0195688, + "balance_loss_mlp": 1.02024233, + "epoch": 0.3949978957494138, + "flos": 17311744926720.0, + "grad_norm": 1.9874622611534698, + "language_loss": 0.80309343, + "learning_rate": 2.757984106559701e-06, + "loss": 0.82506764, + "num_input_tokens_seen": 70771650, + "step": 3285, + "time_per_iteration": 2.6441075801849365 + }, + { + "auxiliary_loss_clip": 0.01134162, + "auxiliary_loss_mlp": 0.01023634, + "balance_loss_clip": 0.97954631, + "balance_loss_mlp": 1.01614177, + "epoch": 0.3951181386400529, + "flos": 36317861280000.0, + "grad_norm": 2.3269921347076234, + "language_loss": 0.7135601, + "learning_rate": 2.7572631887801446e-06, + "loss": 0.73513806, + "num_input_tokens_seen": 70793275, + "step": 3286, + "time_per_iteration": 2.8494815826416016 + }, + { + "auxiliary_loss_clip": 0.01163722, + "auxiliary_loss_mlp": 0.01031092, + "balance_loss_clip": 1.01757789, + "balance_loss_mlp": 1.02319479, + "epoch": 0.395238381530692, + "flos": 23110348170240.0, + "grad_norm": 1.8586417045204673, + "language_loss": 0.76504582, + "learning_rate": 2.7565421561243654e-06, + "loss": 0.78699398, + "num_input_tokens_seen": 70811440, + "step": 3287, + "time_per_iteration": 2.695673942565918 + }, + { + "auxiliary_loss_clip": 0.01110165, + "auxiliary_loss_mlp": 0.01034779, + "balance_loss_clip": 0.93904042, + "balance_loss_mlp": 1.02654219, + "epoch": 0.3953586244213311, + "flos": 24347614095360.0, + "grad_norm": 2.2132127667978185, + "language_loss": 0.82237029, + "learning_rate": 2.7558210087017413e-06, + "loss": 0.84381974, + "num_input_tokens_seen": 70831375, + "step": 3288, + "time_per_iteration": 2.733725070953369 + }, + { + "auxiliary_loss_clip": 0.01106888, + "auxiliary_loss_mlp": 0.01029424, + "balance_loss_clip": 0.94058609, + "balance_loss_mlp": 1.02084708, + "epoch": 0.3954788673119702, + "flos": 23440080044160.0, + "grad_norm": 1.7471017401391367, + "language_loss": 0.73403329, + "learning_rate": 2.7550997466216724e-06, + "loss": 0.75539637, + "num_input_tokens_seen": 70849170, + "step": 3289, + "time_per_iteration": 2.7172560691833496 + }, + { + "auxiliary_loss_clip": 0.01136297, + "auxiliary_loss_mlp": 0.01026592, + "balance_loss_clip": 0.98155296, + "balance_loss_mlp": 1.0183785, + "epoch": 0.3955991102026093, + "flos": 17494063384320.0, + "grad_norm": 2.217932887038095, + "language_loss": 0.81382197, + "learning_rate": 2.7543783699935714e-06, + "loss": 0.83545089, + "num_input_tokens_seen": 70867200, + "step": 3290, + "time_per_iteration": 2.6949684619903564 + }, + { + "auxiliary_loss_clip": 0.01164747, + "auxiliary_loss_mlp": 0.01030923, + "balance_loss_clip": 1.02143729, + "balance_loss_mlp": 1.02270341, + "epoch": 0.39571935309324835, + "flos": 18221326053120.0, + "grad_norm": 2.7945326102608994, + "language_loss": 0.86479157, + "learning_rate": 2.753656878926872e-06, + "loss": 0.88674831, + "num_input_tokens_seen": 70883080, + "step": 3291, + "time_per_iteration": 3.6553192138671875 + }, + { + "auxiliary_loss_clip": 0.01129186, + "auxiliary_loss_mlp": 0.01035886, + "balance_loss_clip": 0.97693324, + "balance_loss_mlp": 1.02827477, + "epoch": 0.39583959598388746, + "flos": 17748813617280.0, + "grad_norm": 2.0672615091328534, + "language_loss": 0.74205816, + "learning_rate": 2.752935273531023e-06, + "loss": 0.76370883, + "num_input_tokens_seen": 70901230, + "step": 3292, + "time_per_iteration": 3.6164727210998535 + }, + { + "auxiliary_loss_clip": 0.01166732, + "auxiliary_loss_mlp": 0.01028663, + "balance_loss_clip": 1.0213511, + "balance_loss_mlp": 1.02012134, + "epoch": 0.39595983887452657, + "flos": 19352368483200.0, + "grad_norm": 3.3088991343030734, + "language_loss": 0.78610772, + "learning_rate": 2.752213553915492e-06, + "loss": 0.8080616, + "num_input_tokens_seen": 70919585, + "step": 3293, + "time_per_iteration": 2.717576265335083 + }, + { + "auxiliary_loss_clip": 0.01047364, + "auxiliary_loss_mlp": 0.01011997, + "balance_loss_clip": 0.95133722, + "balance_loss_mlp": 1.01062036, + "epoch": 0.3960800817651656, + "flos": 60682282940160.0, + "grad_norm": 0.8910346146629996, + "language_loss": 0.66116297, + "learning_rate": 2.751491720189762e-06, + "loss": 0.68175656, + "num_input_tokens_seen": 70977695, + "step": 3294, + "time_per_iteration": 4.104049444198608 + }, + { + "auxiliary_loss_clip": 0.01140588, + "auxiliary_loss_mlp": 0.00509902, + "balance_loss_clip": 0.98155659, + "balance_loss_mlp": 0.0, + "epoch": 0.39620032465580474, + "flos": 16836718538880.0, + "grad_norm": 2.035373767801596, + "language_loss": 0.91480827, + "learning_rate": 2.7507697724633364e-06, + "loss": 0.93131316, + "num_input_tokens_seen": 70994455, + "step": 3295, + "time_per_iteration": 2.6258020401000977 + }, + { + "auxiliary_loss_clip": 0.0101972, + "auxiliary_loss_mlp": 0.01003424, + "balance_loss_clip": 0.9131965, + "balance_loss_mlp": 1.00196362, + "epoch": 0.3963205675464438, + "flos": 69071445941760.0, + "grad_norm": 0.8264901741329882, + "language_loss": 0.54661596, + "learning_rate": 2.7500477108457327e-06, + "loss": 0.56684744, + "num_input_tokens_seen": 71046465, + "step": 3296, + "time_per_iteration": 3.0541083812713623 + }, + { + "auxiliary_loss_clip": 0.01162379, + "auxiliary_loss_mlp": 0.01028947, + "balance_loss_clip": 1.01861644, + "balance_loss_mlp": 1.02065027, + "epoch": 0.3964408104370829, + "flos": 25667439431040.0, + "grad_norm": 2.0521250441805248, + "language_loss": 0.80390304, + "learning_rate": 2.7493255354464877e-06, + "loss": 0.82581633, + "num_input_tokens_seen": 71064275, + "step": 3297, + "time_per_iteration": 2.7360928058624268 + }, + { + "auxiliary_loss_clip": 0.00982683, + "auxiliary_loss_mlp": 0.01029354, + "balance_loss_clip": 0.74458957, + "balance_loss_mlp": 1.02171266, + "epoch": 0.396561053327722, + "flos": 24277480790400.0, + "grad_norm": 1.929414529587256, + "language_loss": 0.76045036, + "learning_rate": 2.748603246375156e-06, + "loss": 0.78057069, + "num_input_tokens_seen": 71082290, + "step": 3298, + "time_per_iteration": 3.0212411880493164 + }, + { + "auxiliary_loss_clip": 0.01190069, + "auxiliary_loss_mlp": 0.01032692, + "balance_loss_clip": 1.06050038, + "balance_loss_mlp": 1.02456236, + "epoch": 0.39668129621836107, + "flos": 20522302364160.0, + "grad_norm": 1.9540364113834934, + "language_loss": 0.69691324, + "learning_rate": 2.7478808437413055e-06, + "loss": 0.71914089, + "num_input_tokens_seen": 71101700, + "step": 3299, + "time_per_iteration": 4.189533472061157 + }, + { + "auxiliary_loss_clip": 0.01089677, + "auxiliary_loss_mlp": 0.01030899, + "balance_loss_clip": 0.90473151, + "balance_loss_mlp": 1.02295399, + "epoch": 0.3968015391090002, + "flos": 27052585649280.0, + "grad_norm": 2.129481461569243, + "language_loss": 0.65839624, + "learning_rate": 2.7471583276545263e-06, + "loss": 0.67960203, + "num_input_tokens_seen": 71122360, + "step": 3300, + "time_per_iteration": 2.83768630027771 + }, + { + "auxiliary_loss_clip": 0.01139751, + "auxiliary_loss_mlp": 0.01030802, + "balance_loss_clip": 0.97867846, + "balance_loss_mlp": 1.02281475, + "epoch": 0.3969217819996393, + "flos": 12531819392640.0, + "grad_norm": 1.8398178479826863, + "language_loss": 0.70403874, + "learning_rate": 2.7464356982244224e-06, + "loss": 0.72574431, + "num_input_tokens_seen": 71140360, + "step": 3301, + "time_per_iteration": 2.639514684677124 + }, + { + "auxiliary_loss_clip": 0.01061089, + "auxiliary_loss_mlp": 0.01000349, + "balance_loss_clip": 0.98795956, + "balance_loss_mlp": 0.99893039, + "epoch": 0.39704202489027834, + "flos": 66241399230720.0, + "grad_norm": 0.7917600874056129, + "language_loss": 0.61738342, + "learning_rate": 2.745712955560617e-06, + "loss": 0.63799781, + "num_input_tokens_seen": 71196565, + "step": 3302, + "time_per_iteration": 3.1853649616241455 + }, + { + "auxiliary_loss_clip": 0.01068159, + "auxiliary_loss_mlp": 0.01035793, + "balance_loss_clip": 0.86535645, + "balance_loss_mlp": 1.0273056, + "epoch": 0.39716226778091746, + "flos": 16982982720000.0, + "grad_norm": 2.6675667969458647, + "language_loss": 0.76652157, + "learning_rate": 2.7449900997727496e-06, + "loss": 0.78756106, + "num_input_tokens_seen": 71214675, + "step": 3303, + "time_per_iteration": 2.749413013458252 + }, + { + "auxiliary_loss_clip": 0.01137294, + "auxiliary_loss_mlp": 0.01031035, + "balance_loss_clip": 0.98245192, + "balance_loss_mlp": 1.02299988, + "epoch": 0.39728251067155657, + "flos": 23477139901440.0, + "grad_norm": 1.9304952734248892, + "language_loss": 0.84293699, + "learning_rate": 2.744267130970476e-06, + "loss": 0.86462027, + "num_input_tokens_seen": 71234400, + "step": 3304, + "time_per_iteration": 2.736765146255493 + }, + { + "auxiliary_loss_clip": 0.01134541, + "auxiliary_loss_mlp": 0.01031516, + "balance_loss_clip": 0.98080194, + "balance_loss_mlp": 1.02353466, + "epoch": 0.3974027535621956, + "flos": 20704441253760.0, + "grad_norm": 1.8483859729795742, + "language_loss": 0.76919347, + "learning_rate": 2.7435440492634697e-06, + "loss": 0.79085404, + "num_input_tokens_seen": 71253725, + "step": 3305, + "time_per_iteration": 2.7045633792877197 + }, + { + "auxiliary_loss_clip": 0.01145364, + "auxiliary_loss_mlp": 0.01032064, + "balance_loss_clip": 0.98390627, + "balance_loss_mlp": 1.02361155, + "epoch": 0.39752299645283473, + "flos": 21543278544000.0, + "grad_norm": 2.1358360543605883, + "language_loss": 0.67294014, + "learning_rate": 2.7428208547614228e-06, + "loss": 0.69471443, + "num_input_tokens_seen": 71273220, + "step": 3306, + "time_per_iteration": 2.7737927436828613 + }, + { + "auxiliary_loss_clip": 0.01163841, + "auxiliary_loss_mlp": 0.01028957, + "balance_loss_clip": 1.01878619, + "balance_loss_mlp": 1.02131033, + "epoch": 0.39764323934347384, + "flos": 19208295031680.0, + "grad_norm": 1.9847254733732238, + "language_loss": 0.77078223, + "learning_rate": 2.742097547574043e-06, + "loss": 0.79271019, + "num_input_tokens_seen": 71291445, + "step": 3307, + "time_per_iteration": 2.621401786804199 + }, + { + "auxiliary_loss_clip": 0.01146922, + "auxiliary_loss_mlp": 0.00510818, + "balance_loss_clip": 0.98114771, + "balance_loss_mlp": 0.0, + "epoch": 0.3977634822341129, + "flos": 20850202644480.0, + "grad_norm": 1.9448575487285176, + "language_loss": 0.78203344, + "learning_rate": 2.7413741278110544e-06, + "loss": 0.79861081, + "num_input_tokens_seen": 71310135, + "step": 3308, + "time_per_iteration": 2.7123677730560303 + }, + { + "auxiliary_loss_clip": 0.01148022, + "auxiliary_loss_mlp": 0.01028759, + "balance_loss_clip": 0.98456043, + "balance_loss_mlp": 1.0204742, + "epoch": 0.397883725124752, + "flos": 39786042038400.0, + "grad_norm": 2.2200926371566156, + "language_loss": 0.6943332, + "learning_rate": 2.7406505955822016e-06, + "loss": 0.71610099, + "num_input_tokens_seen": 71331160, + "step": 3309, + "time_per_iteration": 2.83040189743042 + }, + { + "auxiliary_loss_clip": 0.01138773, + "auxiliary_loss_mlp": 0.01034753, + "balance_loss_clip": 0.97837418, + "balance_loss_mlp": 1.02633047, + "epoch": 0.39800396801539106, + "flos": 17379507934080.0, + "grad_norm": 3.0005024027861853, + "language_loss": 0.65751427, + "learning_rate": 2.7399269509972415e-06, + "loss": 0.67924953, + "num_input_tokens_seen": 71345315, + "step": 3310, + "time_per_iteration": 2.6763832569122314 + }, + { + "auxiliary_loss_clip": 0.01134429, + "auxiliary_loss_mlp": 0.0102881, + "balance_loss_clip": 0.97515994, + "balance_loss_mlp": 1.02014339, + "epoch": 0.3981242109060302, + "flos": 19202764337280.0, + "grad_norm": 2.2808977443402543, + "language_loss": 0.85330772, + "learning_rate": 2.7392031941659514e-06, + "loss": 0.87494016, + "num_input_tokens_seen": 71363160, + "step": 3311, + "time_per_iteration": 2.6801347732543945 + }, + { + "auxiliary_loss_clip": 0.01146486, + "auxiliary_loss_mlp": 0.0102864, + "balance_loss_clip": 0.98687112, + "balance_loss_mlp": 1.02085614, + "epoch": 0.3982444537966693, + "flos": 24565124903040.0, + "grad_norm": 1.8468774964341765, + "language_loss": 0.86106372, + "learning_rate": 2.7384793251981244e-06, + "loss": 0.88281494, + "num_input_tokens_seen": 71382145, + "step": 3312, + "time_per_iteration": 2.7279226779937744 + }, + { + "auxiliary_loss_clip": 0.01170225, + "auxiliary_loss_mlp": 0.01028788, + "balance_loss_clip": 1.02045131, + "balance_loss_mlp": 1.02071786, + "epoch": 0.39836469668730834, + "flos": 26213856099840.0, + "grad_norm": 8.801933239495872, + "language_loss": 0.80567896, + "learning_rate": 2.737755344203571e-06, + "loss": 0.82766902, + "num_input_tokens_seen": 71402095, + "step": 3313, + "time_per_iteration": 2.759037971496582 + }, + { + "auxiliary_loss_clip": 0.01168794, + "auxiliary_loss_mlp": 0.01028831, + "balance_loss_clip": 1.02299798, + "balance_loss_mlp": 1.02083254, + "epoch": 0.39848493957794745, + "flos": 27636134002560.0, + "grad_norm": 1.5920421925890602, + "language_loss": 0.80088913, + "learning_rate": 2.7370312512921186e-06, + "loss": 0.82286537, + "num_input_tokens_seen": 71423875, + "step": 3314, + "time_per_iteration": 2.6774911880493164 + }, + { + "auxiliary_loss_clip": 0.01144213, + "auxiliary_loss_mlp": 0.01031166, + "balance_loss_clip": 0.97983623, + "balance_loss_mlp": 1.02323866, + "epoch": 0.39860518246858656, + "flos": 12239326944000.0, + "grad_norm": 2.4564292951392677, + "language_loss": 0.76484716, + "learning_rate": 2.736307046573611e-06, + "loss": 0.78660095, + "num_input_tokens_seen": 71439745, + "step": 3315, + "time_per_iteration": 2.654978036880493 + }, + { + "auxiliary_loss_clip": 0.01188267, + "auxiliary_loss_mlp": 0.01027265, + "balance_loss_clip": 1.05824387, + "balance_loss_mlp": 1.01912618, + "epoch": 0.3987254253592256, + "flos": 22379135005440.0, + "grad_norm": 1.7424540851910861, + "language_loss": 0.81524169, + "learning_rate": 2.73558273015791e-06, + "loss": 0.83739704, + "num_input_tokens_seen": 71459575, + "step": 3316, + "time_per_iteration": 2.5792036056518555 + }, + { + "auxiliary_loss_clip": 0.01189184, + "auxiliary_loss_mlp": 0.01039186, + "balance_loss_clip": 1.05773592, + "balance_loss_mlp": 1.02998269, + "epoch": 0.3988456682498647, + "flos": 23514020190720.0, + "grad_norm": 2.221855266202815, + "language_loss": 0.70802128, + "learning_rate": 2.734858302154894e-06, + "loss": 0.73030496, + "num_input_tokens_seen": 71481075, + "step": 3317, + "time_per_iteration": 3.549685001373291 + }, + { + "auxiliary_loss_clip": 0.01136403, + "auxiliary_loss_mlp": 0.01028666, + "balance_loss_clip": 0.98186767, + "balance_loss_mlp": 1.02004743, + "epoch": 0.39896591114050384, + "flos": 19208761908480.0, + "grad_norm": 2.3605692869408976, + "language_loss": 0.76809216, + "learning_rate": 2.734133762674457e-06, + "loss": 0.78974289, + "num_input_tokens_seen": 71500665, + "step": 3318, + "time_per_iteration": 3.672323226928711 + }, + { + "auxiliary_loss_clip": 0.01141678, + "auxiliary_loss_mlp": 0.01027571, + "balance_loss_clip": 0.98173553, + "balance_loss_mlp": 1.01879728, + "epoch": 0.3990861540311429, + "flos": 28401031146240.0, + "grad_norm": 4.682550482266794, + "language_loss": 0.70682812, + "learning_rate": 2.7334091118265124e-06, + "loss": 0.72852063, + "num_input_tokens_seen": 71522560, + "step": 3319, + "time_per_iteration": 2.8226943016052246 + }, + { + "auxiliary_loss_clip": 0.01070585, + "auxiliary_loss_mlp": 0.01000423, + "balance_loss_clip": 0.98912334, + "balance_loss_mlp": 0.9989869, + "epoch": 0.399206396921782, + "flos": 61758563086080.0, + "grad_norm": 0.7010734833385776, + "language_loss": 0.578435, + "learning_rate": 2.732684349720989e-06, + "loss": 0.59914505, + "num_input_tokens_seen": 71590520, + "step": 3320, + "time_per_iteration": 4.16244649887085 + }, + { + "auxiliary_loss_clip": 0.0111882, + "auxiliary_loss_mlp": 0.01039186, + "balance_loss_clip": 0.94090474, + "balance_loss_mlp": 1.03078198, + "epoch": 0.3993266398124211, + "flos": 28074567409920.0, + "grad_norm": 1.8225243764231511, + "language_loss": 0.75405753, + "learning_rate": 2.7319594764678318e-06, + "loss": 0.77563763, + "num_input_tokens_seen": 71612620, + "step": 3321, + "time_per_iteration": 2.7612040042877197 + }, + { + "auxiliary_loss_clip": 0.01095842, + "auxiliary_loss_mlp": 0.01040365, + "balance_loss_clip": 0.90363526, + "balance_loss_mlp": 1.03175759, + "epoch": 0.39944688270306017, + "flos": 23225083188480.0, + "grad_norm": 2.006947915480317, + "language_loss": 0.83209127, + "learning_rate": 2.7312344921770044e-06, + "loss": 0.85345334, + "num_input_tokens_seen": 71634320, + "step": 3322, + "time_per_iteration": 2.885953187942505 + }, + { + "auxiliary_loss_clip": 0.01141, + "auxiliary_loss_mlp": 0.01031689, + "balance_loss_clip": 0.9792732, + "balance_loss_mlp": 1.0236423, + "epoch": 0.3995671255936993, + "flos": 19390433921280.0, + "grad_norm": 1.8223199529144112, + "language_loss": 0.78333253, + "learning_rate": 2.7305093969584857e-06, + "loss": 0.80505943, + "num_input_tokens_seen": 71653145, + "step": 3323, + "time_per_iteration": 2.7190351486206055 + }, + { + "auxiliary_loss_clip": 0.01158783, + "auxiliary_loss_mlp": 0.0104281, + "balance_loss_clip": 1.01722479, + "balance_loss_mlp": 1.03408992, + "epoch": 0.3996873684843384, + "flos": 23842638743040.0, + "grad_norm": 1.6279884753016847, + "language_loss": 0.79821903, + "learning_rate": 2.729784190922272e-06, + "loss": 0.82023495, + "num_input_tokens_seen": 71674580, + "step": 3324, + "time_per_iteration": 2.6219818592071533 + }, + { + "auxiliary_loss_clip": 0.01044986, + "auxiliary_loss_mlp": 0.01000652, + "balance_loss_clip": 0.94876456, + "balance_loss_mlp": 0.99914402, + "epoch": 0.39980761137497745, + "flos": 66576877280640.0, + "grad_norm": 0.956984908684091, + "language_loss": 0.57245171, + "learning_rate": 2.729058874178378e-06, + "loss": 0.59290814, + "num_input_tokens_seen": 71745260, + "step": 3325, + "time_per_iteration": 4.2652552127838135 + }, + { + "auxiliary_loss_clip": 0.01149027, + "auxiliary_loss_mlp": 0.01029631, + "balance_loss_clip": 0.98559672, + "balance_loss_mlp": 1.02109528, + "epoch": 0.39992785426561656, + "flos": 28549162834560.0, + "grad_norm": 2.1634472048537052, + "language_loss": 0.69191724, + "learning_rate": 2.7283334468368315e-06, + "loss": 0.71370381, + "num_input_tokens_seen": 71766540, + "step": 3326, + "time_per_iteration": 2.7290196418762207 + }, + { + "auxiliary_loss_clip": 0.01016996, + "auxiliary_loss_mlp": 0.01030882, + "balance_loss_clip": 0.78067338, + "balance_loss_mlp": 1.02219784, + "epoch": 0.4000480971562556, + "flos": 15049408671360.0, + "grad_norm": 1.6878867451573183, + "language_loss": 0.72468376, + "learning_rate": 2.72760790900768e-06, + "loss": 0.74516261, + "num_input_tokens_seen": 71783125, + "step": 3327, + "time_per_iteration": 2.7760679721832275 + }, + { + "auxiliary_loss_clip": 0.01192405, + "auxiliary_loss_mlp": 0.01039531, + "balance_loss_clip": 1.06104529, + "balance_loss_mlp": 1.03145504, + "epoch": 0.4001683400468947, + "flos": 23915609222400.0, + "grad_norm": 1.7281790440395342, + "language_loss": 0.7897867, + "learning_rate": 2.7268822608009875e-06, + "loss": 0.81210601, + "num_input_tokens_seen": 71802500, + "step": 3328, + "time_per_iteration": 2.652658700942993 + }, + { + "auxiliary_loss_clip": 0.01125723, + "auxiliary_loss_mlp": 0.01029687, + "balance_loss_clip": 0.94489557, + "balance_loss_mlp": 1.0208416, + "epoch": 0.40028858293753383, + "flos": 24352677912960.0, + "grad_norm": 2.143115904328062, + "language_loss": 0.78219843, + "learning_rate": 2.726156502326834e-06, + "loss": 0.80375254, + "num_input_tokens_seen": 71823800, + "step": 3329, + "time_per_iteration": 246.85328602790833 + }, + { + "auxiliary_loss_clip": 0.00987731, + "auxiliary_loss_mlp": 0.0101928, + "balance_loss_clip": 0.83896995, + "balance_loss_mlp": 1.01751578, + "epoch": 0.4004088258281729, + "flos": 66787025800320.0, + "grad_norm": 0.6977390586151072, + "language_loss": 0.60236001, + "learning_rate": 2.725430633695316e-06, + "loss": 0.62243015, + "num_input_tokens_seen": 71886880, + "step": 3330, + "time_per_iteration": 3.364123582839966 + }, + { + "auxiliary_loss_clip": 0.01089816, + "auxiliary_loss_mlp": 0.01001666, + "balance_loss_clip": 1.02645028, + "balance_loss_mlp": 1.00018167, + "epoch": 0.400529068718812, + "flos": 58598386473600.0, + "grad_norm": 0.8819347978084489, + "language_loss": 0.57907367, + "learning_rate": 2.7247046550165485e-06, + "loss": 0.59998858, + "num_input_tokens_seen": 71939005, + "step": 3331, + "time_per_iteration": 3.493711471557617 + }, + { + "auxiliary_loss_clip": 0.0119341, + "auxiliary_loss_mlp": 0.01047129, + "balance_loss_clip": 1.06174505, + "balance_loss_mlp": 1.03877234, + "epoch": 0.4006493116094511, + "flos": 25377460934400.0, + "grad_norm": 1.463418454190203, + "language_loss": 0.75833941, + "learning_rate": 2.7239785664006606e-06, + "loss": 0.78074479, + "num_input_tokens_seen": 71962545, + "step": 3332, + "time_per_iteration": 2.7578866481781006 + }, + { + "auxiliary_loss_clip": 0.010703, + "auxiliary_loss_mlp": 0.01002056, + "balance_loss_clip": 0.98803663, + "balance_loss_mlp": 1.00059021, + "epoch": 0.40076955450009016, + "flos": 60280729822080.0, + "grad_norm": 0.7717792668958678, + "language_loss": 0.61820924, + "learning_rate": 2.7232523679578002e-06, + "loss": 0.63893282, + "num_input_tokens_seen": 72025625, + "step": 3333, + "time_per_iteration": 3.2704386711120605 + }, + { + "auxiliary_loss_clip": 0.01165084, + "auxiliary_loss_mlp": 0.01026826, + "balance_loss_clip": 1.02147555, + "balance_loss_mlp": 1.01824307, + "epoch": 0.4008897973907293, + "flos": 16617268396800.0, + "grad_norm": 2.030903142711188, + "language_loss": 0.79447347, + "learning_rate": 2.7225260597981295e-06, + "loss": 0.81639254, + "num_input_tokens_seen": 72043330, + "step": 3334, + "time_per_iteration": 2.5977511405944824 + }, + { + "auxiliary_loss_clip": 0.01114189, + "auxiliary_loss_mlp": 0.00511375, + "balance_loss_clip": 0.94139129, + "balance_loss_mlp": 0.0, + "epoch": 0.4010100402813684, + "flos": 15377344865280.0, + "grad_norm": 2.501393370119259, + "language_loss": 0.78517491, + "learning_rate": 2.721799642031831e-06, + "loss": 0.80143058, + "num_input_tokens_seen": 72059500, + "step": 3335, + "time_per_iteration": 2.722351312637329 + }, + { + "auxiliary_loss_clip": 0.01147526, + "auxiliary_loss_mlp": 0.0103071, + "balance_loss_clip": 0.97996032, + "balance_loss_mlp": 1.02160263, + "epoch": 0.40113028317200744, + "flos": 13298835438720.0, + "grad_norm": 2.2531546423901565, + "language_loss": 0.7783711, + "learning_rate": 2.721073114769101e-06, + "loss": 0.80015337, + "num_input_tokens_seen": 72077175, + "step": 3336, + "time_per_iteration": 2.705479860305786 + }, + { + "auxiliary_loss_clip": 0.01118549, + "auxiliary_loss_mlp": 0.01033434, + "balance_loss_clip": 0.94656754, + "balance_loss_mlp": 1.02524471, + "epoch": 0.40125052606264655, + "flos": 20668027841280.0, + "grad_norm": 2.3660909655327895, + "language_loss": 0.75202626, + "learning_rate": 2.7203464781201523e-06, + "loss": 0.7735461, + "num_input_tokens_seen": 72096490, + "step": 3337, + "time_per_iteration": 2.731451988220215 + }, + { + "auxiliary_loss_clip": 0.01194267, + "auxiliary_loss_mlp": 0.01032046, + "balance_loss_clip": 1.06145048, + "balance_loss_mlp": 1.02338529, + "epoch": 0.40137076895328566, + "flos": 24607679541120.0, + "grad_norm": 2.7874846560311806, + "language_loss": 0.78238153, + "learning_rate": 2.719619732195215e-06, + "loss": 0.8046447, + "num_input_tokens_seen": 72118130, + "step": 3338, + "time_per_iteration": 2.676114082336426 + }, + { + "auxiliary_loss_clip": 0.01122393, + "auxiliary_loss_mlp": 0.01026724, + "balance_loss_clip": 0.94384575, + "balance_loss_mlp": 1.01851606, + "epoch": 0.4014910118439247, + "flos": 24206593299840.0, + "grad_norm": 1.6677731170180352, + "language_loss": 0.72555256, + "learning_rate": 2.7188928771045377e-06, + "loss": 0.74704367, + "num_input_tokens_seen": 72139450, + "step": 3339, + "time_per_iteration": 2.7661538124084473 + }, + { + "auxiliary_loss_clip": 0.01112666, + "auxiliary_loss_mlp": 0.01031696, + "balance_loss_clip": 0.94080931, + "balance_loss_mlp": 1.02233803, + "epoch": 0.4016112547345638, + "flos": 26725080418560.0, + "grad_norm": 1.6285780232854885, + "language_loss": 0.79707289, + "learning_rate": 2.7181659129583815e-06, + "loss": 0.81851649, + "num_input_tokens_seen": 72159040, + "step": 3340, + "time_per_iteration": 2.7164969444274902 + }, + { + "auxiliary_loss_clip": 0.01132398, + "auxiliary_loss_mlp": 0.0102887, + "balance_loss_clip": 0.97430521, + "balance_loss_mlp": 1.01977396, + "epoch": 0.4017314976252029, + "flos": 21288025520640.0, + "grad_norm": 1.8683672672784122, + "language_loss": 0.76163507, + "learning_rate": 2.7174388398670276e-06, + "loss": 0.78324771, + "num_input_tokens_seen": 72178220, + "step": 3341, + "time_per_iteration": 2.7186334133148193 + }, + { + "auxiliary_loss_clip": 0.01191667, + "auxiliary_loss_mlp": 0.01034932, + "balance_loss_clip": 1.05655658, + "balance_loss_mlp": 1.02561569, + "epoch": 0.401851740515842, + "flos": 25484690010240.0, + "grad_norm": 1.9257383282090208, + "language_loss": 0.92045748, + "learning_rate": 2.716711657940773e-06, + "loss": 0.94272339, + "num_input_tokens_seen": 72199230, + "step": 3342, + "time_per_iteration": 3.5644757747650146 + }, + { + "auxiliary_loss_clip": 0.01023944, + "auxiliary_loss_mlp": 0.01002105, + "balance_loss_clip": 0.90899003, + "balance_loss_mlp": 1.00066304, + "epoch": 0.4019719834064811, + "flos": 55395334978560.0, + "grad_norm": 0.8091223552345603, + "language_loss": 0.56483197, + "learning_rate": 2.7159843672899284e-06, + "loss": 0.58509248, + "num_input_tokens_seen": 72263430, + "step": 3343, + "time_per_iteration": 4.368829727172852 + }, + { + "auxiliary_loss_clip": 0.01170909, + "auxiliary_loss_mlp": 0.01032385, + "balance_loss_clip": 1.0247066, + "balance_loss_mlp": 1.02355766, + "epoch": 0.40209222629712016, + "flos": 18180100218240.0, + "grad_norm": 4.231197221520971, + "language_loss": 0.81413752, + "learning_rate": 2.715256968024825e-06, + "loss": 0.83617043, + "num_input_tokens_seen": 72280505, + "step": 3344, + "time_per_iteration": 2.641183853149414 + }, + { + "auxiliary_loss_clip": 0.01147851, + "auxiliary_loss_mlp": 0.01028186, + "balance_loss_clip": 0.98014712, + "balance_loss_mlp": 1.01982355, + "epoch": 0.40221246918775927, + "flos": 25961009287680.0, + "grad_norm": 1.8856230109181722, + "language_loss": 0.81935799, + "learning_rate": 2.7145294602558083e-06, + "loss": 0.8411184, + "num_input_tokens_seen": 72301215, + "step": 3345, + "time_per_iteration": 3.800785779953003 + }, + { + "auxiliary_loss_clip": 0.01170452, + "auxiliary_loss_mlp": 0.0102719, + "balance_loss_clip": 1.02323008, + "balance_loss_mlp": 1.01884556, + "epoch": 0.4023327120783984, + "flos": 33838912056960.0, + "grad_norm": 1.8877969227846982, + "language_loss": 0.70859748, + "learning_rate": 2.713801844093241e-06, + "loss": 0.73057395, + "num_input_tokens_seen": 72322365, + "step": 3346, + "time_per_iteration": 2.737067222595215 + }, + { + "auxiliary_loss_clip": 0.011679, + "auxiliary_loss_mlp": 0.01029515, + "balance_loss_clip": 1.02012348, + "balance_loss_mlp": 1.02081299, + "epoch": 0.40245295496903744, + "flos": 26900252069760.0, + "grad_norm": 2.3062517000361744, + "language_loss": 0.88618636, + "learning_rate": 2.7130741196475014e-06, + "loss": 0.90816057, + "num_input_tokens_seen": 72340495, + "step": 3347, + "time_per_iteration": 2.7689104080200195 + }, + { + "auxiliary_loss_clip": 0.01145864, + "auxiliary_loss_mlp": 0.01033655, + "balance_loss_clip": 0.98353809, + "balance_loss_mlp": 1.0243566, + "epoch": 0.40257319785967655, + "flos": 36902738436480.0, + "grad_norm": 1.7986069866713823, + "language_loss": 0.79320878, + "learning_rate": 2.7123462870289848e-06, + "loss": 0.81500399, + "num_input_tokens_seen": 72360545, + "step": 3348, + "time_per_iteration": 2.87709903717041 + }, + { + "auxiliary_loss_clip": 0.01144492, + "auxiliary_loss_mlp": 0.01022873, + "balance_loss_clip": 0.9791351, + "balance_loss_mlp": 1.01353252, + "epoch": 0.40269344075031566, + "flos": 24353180703360.0, + "grad_norm": 1.7406100163609823, + "language_loss": 0.80769932, + "learning_rate": 2.711618346348102e-06, + "loss": 0.82937294, + "num_input_tokens_seen": 72381070, + "step": 3349, + "time_per_iteration": 2.7283880710601807 + }, + { + "auxiliary_loss_clip": 0.01137734, + "auxiliary_loss_mlp": 0.01026447, + "balance_loss_clip": 0.9812184, + "balance_loss_mlp": 1.01868641, + "epoch": 0.4028136836409547, + "flos": 14389657614720.0, + "grad_norm": 1.64090067102154, + "language_loss": 0.63336194, + "learning_rate": 2.7108902977152825e-06, + "loss": 0.65500373, + "num_input_tokens_seen": 72398970, + "step": 3350, + "time_per_iteration": 3.9473700523376465 + }, + { + "auxiliary_loss_clip": 0.01163996, + "auxiliary_loss_mlp": 0.01029215, + "balance_loss_clip": 1.01813436, + "balance_loss_mlp": 1.0202862, + "epoch": 0.4029339265315938, + "flos": 26136037284480.0, + "grad_norm": 2.9456199100576907, + "language_loss": 0.75396729, + "learning_rate": 2.7101621412409704e-06, + "loss": 0.77589941, + "num_input_tokens_seen": 72418455, + "step": 3351, + "time_per_iteration": 2.7032017707824707 + }, + { + "auxiliary_loss_clip": 0.01191326, + "auxiliary_loss_mlp": 0.01029547, + "balance_loss_clip": 1.05762148, + "balance_loss_mlp": 1.02039802, + "epoch": 0.40305416942223293, + "flos": 23256325042560.0, + "grad_norm": 2.1657915104447354, + "language_loss": 0.86080974, + "learning_rate": 2.7094338770356256e-06, + "loss": 0.88301849, + "num_input_tokens_seen": 72437540, + "step": 3352, + "time_per_iteration": 2.674004554748535 + }, + { + "auxiliary_loss_clip": 0.01138695, + "auxiliary_loss_mlp": 0.010316, + "balance_loss_clip": 0.98033631, + "balance_loss_mlp": 1.02328575, + "epoch": 0.403174412312872, + "flos": 27089645506560.0, + "grad_norm": 1.9738236945114125, + "language_loss": 0.64205706, + "learning_rate": 2.708705505209726e-06, + "loss": 0.66376001, + "num_input_tokens_seen": 72458315, + "step": 3353, + "time_per_iteration": 2.7729923725128174 + }, + { + "auxiliary_loss_clip": 0.01086246, + "auxiliary_loss_mlp": 0.01027739, + "balance_loss_clip": 0.90117949, + "balance_loss_mlp": 1.01932931, + "epoch": 0.4032946552035111, + "flos": 21756336065280.0, + "grad_norm": 2.731043562092086, + "language_loss": 0.91306019, + "learning_rate": 2.7079770258737646e-06, + "loss": 0.93420005, + "num_input_tokens_seen": 72476225, + "step": 3354, + "time_per_iteration": 2.7585103511810303 + }, + { + "auxiliary_loss_clip": 0.01115336, + "auxiliary_loss_mlp": 0.01035321, + "balance_loss_clip": 0.94004506, + "balance_loss_mlp": 1.02547455, + "epoch": 0.4034148980941502, + "flos": 17343956448000.0, + "grad_norm": 5.232106379173105, + "language_loss": 0.74932706, + "learning_rate": 2.707248439138251e-06, + "loss": 0.77083367, + "num_input_tokens_seen": 72492460, + "step": 3355, + "time_per_iteration": 2.7724788188934326 + }, + { + "auxiliary_loss_clip": 0.01139271, + "auxiliary_loss_mlp": 0.01028653, + "balance_loss_clip": 0.98347759, + "balance_loss_mlp": 1.02012408, + "epoch": 0.40353514098478926, + "flos": 22017838055040.0, + "grad_norm": 1.8312224697827988, + "language_loss": 0.65304238, + "learning_rate": 2.7065197451137114e-06, + "loss": 0.6747216, + "num_input_tokens_seen": 72513840, + "step": 3356, + "time_per_iteration": 2.747385025024414 + }, + { + "auxiliary_loss_clip": 0.01138975, + "auxiliary_loss_mlp": 0.01026761, + "balance_loss_clip": 0.98042369, + "balance_loss_mlp": 1.01810002, + "epoch": 0.4036553838754284, + "flos": 14246446089600.0, + "grad_norm": 2.2326526329711744, + "language_loss": 0.67438942, + "learning_rate": 2.7057909439106894e-06, + "loss": 0.69604677, + "num_input_tokens_seen": 72531695, + "step": 3357, + "time_per_iteration": 2.6524899005889893 + }, + { + "auxiliary_loss_clip": 0.01160384, + "auxiliary_loss_mlp": 0.0051108, + "balance_loss_clip": 1.01808095, + "balance_loss_mlp": 0.0, + "epoch": 0.40377562676606743, + "flos": 24790644443520.0, + "grad_norm": 2.0235012980512566, + "language_loss": 0.78412271, + "learning_rate": 2.7050620356397417e-06, + "loss": 0.80083734, + "num_input_tokens_seen": 72550645, + "step": 3358, + "time_per_iteration": 2.6984314918518066 + }, + { + "auxiliary_loss_clip": 0.01191353, + "auxiliary_loss_mlp": 0.01029241, + "balance_loss_clip": 1.06086969, + "balance_loss_mlp": 1.02095604, + "epoch": 0.40389586965670654, + "flos": 24061226958720.0, + "grad_norm": 2.3805714320280087, + "language_loss": 0.72187614, + "learning_rate": 2.7043330204114437e-06, + "loss": 0.74408209, + "num_input_tokens_seen": 72569355, + "step": 3359, + "time_per_iteration": 2.687837839126587 + }, + { + "auxiliary_loss_clip": 0.01184961, + "auxiliary_loss_mlp": 0.01032892, + "balance_loss_clip": 1.05535495, + "balance_loss_mlp": 1.0240941, + "epoch": 0.40401611254734565, + "flos": 16399613934720.0, + "grad_norm": 1.953488347086092, + "language_loss": 0.85771143, + "learning_rate": 2.7036038983363862e-06, + "loss": 0.87988997, + "num_input_tokens_seen": 72585960, + "step": 3360, + "time_per_iteration": 2.584622621536255 + }, + { + "auxiliary_loss_clip": 0.01161615, + "auxiliary_loss_mlp": 0.01027842, + "balance_loss_clip": 1.01948607, + "balance_loss_mlp": 1.01972365, + "epoch": 0.4041363554379847, + "flos": 23988220565760.0, + "grad_norm": 1.6460603845255306, + "language_loss": 0.84191728, + "learning_rate": 2.702874669525177e-06, + "loss": 0.86381185, + "num_input_tokens_seen": 72604440, + "step": 3361, + "time_per_iteration": 2.658069133758545 + }, + { + "auxiliary_loss_clip": 0.01125605, + "auxiliary_loss_mlp": 0.01032902, + "balance_loss_clip": 0.94872022, + "balance_loss_mlp": 1.02446759, + "epoch": 0.4042565983286238, + "flos": 28401964899840.0, + "grad_norm": 2.1850073177621727, + "language_loss": 0.6981653, + "learning_rate": 2.7021453340884394e-06, + "loss": 0.71975034, + "num_input_tokens_seen": 72622165, + "step": 3362, + "time_per_iteration": 2.793921709060669 + }, + { + "auxiliary_loss_clip": 0.01131645, + "auxiliary_loss_mlp": 0.0051071, + "balance_loss_clip": 0.97938144, + "balance_loss_mlp": 0.0, + "epoch": 0.40437684121926293, + "flos": 17710963660800.0, + "grad_norm": 2.3993984787952254, + "language_loss": 0.72996807, + "learning_rate": 2.7014158921368125e-06, + "loss": 0.74639153, + "num_input_tokens_seen": 72640490, + "step": 3363, + "time_per_iteration": 2.6823692321777344 + }, + { + "auxiliary_loss_clip": 0.01189765, + "auxiliary_loss_mlp": 0.01030347, + "balance_loss_clip": 1.05760908, + "balance_loss_mlp": 1.02135229, + "epoch": 0.404497084109902, + "flos": 24018959629440.0, + "grad_norm": 2.006271787269419, + "language_loss": 0.85530019, + "learning_rate": 2.700686343780953e-06, + "loss": 0.87750137, + "num_input_tokens_seen": 72660360, + "step": 3364, + "time_per_iteration": 2.7085790634155273 + }, + { + "auxiliary_loss_clip": 0.01140675, + "auxiliary_loss_mlp": 0.01031087, + "balance_loss_clip": 0.97836852, + "balance_loss_mlp": 1.02295089, + "epoch": 0.4046173270005411, + "flos": 22929861306240.0, + "grad_norm": 1.8134586206004006, + "language_loss": 0.88576698, + "learning_rate": 2.699956689131532e-06, + "loss": 0.90748465, + "num_input_tokens_seen": 72680345, + "step": 3365, + "time_per_iteration": 2.735234022140503 + }, + { + "auxiliary_loss_clip": 0.01144521, + "auxiliary_loss_mlp": 0.01029683, + "balance_loss_clip": 0.98020792, + "balance_loss_mlp": 1.02127314, + "epoch": 0.4047375698911802, + "flos": 20668135582080.0, + "grad_norm": 2.1146734453420475, + "language_loss": 0.85315776, + "learning_rate": 2.699226928299238e-06, + "loss": 0.87489974, + "num_input_tokens_seen": 72698365, + "step": 3366, + "time_per_iteration": 2.735440731048584 + }, + { + "auxiliary_loss_clip": 0.01167338, + "auxiliary_loss_mlp": 0.01026687, + "balance_loss_clip": 1.02073431, + "balance_loss_mlp": 1.01800251, + "epoch": 0.40485781278181926, + "flos": 28912865996160.0, + "grad_norm": 2.2818566146184702, + "language_loss": 0.79201901, + "learning_rate": 2.698497061394774e-06, + "loss": 0.81395936, + "num_input_tokens_seen": 72716850, + "step": 3367, + "time_per_iteration": 2.722165822982788 + }, + { + "auxiliary_loss_clip": 0.01126302, + "auxiliary_loss_mlp": 0.00511943, + "balance_loss_clip": 0.94431221, + "balance_loss_mlp": 0.0, + "epoch": 0.40497805567245837, + "flos": 23148377694720.0, + "grad_norm": 1.6211357987997692, + "language_loss": 0.80619437, + "learning_rate": 2.6977670885288627e-06, + "loss": 0.82257676, + "num_input_tokens_seen": 72738250, + "step": 3368, + "time_per_iteration": 3.6558384895324707 + }, + { + "auxiliary_loss_clip": 0.01132339, + "auxiliary_loss_mlp": 0.01026577, + "balance_loss_clip": 0.97802716, + "balance_loss_mlp": 1.01719546, + "epoch": 0.4050982985630975, + "flos": 16289404030080.0, + "grad_norm": 1.720970800953354, + "language_loss": 0.74967098, + "learning_rate": 2.6970370098122378e-06, + "loss": 0.77126008, + "num_input_tokens_seen": 72755235, + "step": 3369, + "time_per_iteration": 3.590160369873047 + }, + { + "auxiliary_loss_clip": 0.01189911, + "auxiliary_loss_mlp": 0.01033009, + "balance_loss_clip": 1.05759668, + "balance_loss_mlp": 1.02419388, + "epoch": 0.40521854145373654, + "flos": 34459484353920.0, + "grad_norm": 1.5529107607151664, + "language_loss": 0.8654983, + "learning_rate": 2.6963068253556535e-06, + "loss": 0.8877275, + "num_input_tokens_seen": 72776620, + "step": 3370, + "time_per_iteration": 2.751836061477661 + }, + { + "auxiliary_loss_clip": 0.01172992, + "auxiliary_loss_mlp": 0.01033165, + "balance_loss_clip": 1.01825094, + "balance_loss_mlp": 1.0237298, + "epoch": 0.40533878434437565, + "flos": 25331099454720.0, + "grad_norm": 1.892919980665916, + "language_loss": 0.85476989, + "learning_rate": 2.6955765352698763e-06, + "loss": 0.87683141, + "num_input_tokens_seen": 72796765, + "step": 3371, + "time_per_iteration": 3.720254421234131 + }, + { + "auxiliary_loss_clip": 0.01191031, + "auxiliary_loss_mlp": 0.01032478, + "balance_loss_clip": 1.05554104, + "balance_loss_mlp": 1.02403784, + "epoch": 0.40545902723501476, + "flos": 15012061505280.0, + "grad_norm": 2.1174640893498755, + "language_loss": 0.73437381, + "learning_rate": 2.6948461396656923e-06, + "loss": 0.7566089, + "num_input_tokens_seen": 72814175, + "step": 3372, + "time_per_iteration": 2.5882396697998047 + }, + { + "auxiliary_loss_clip": 0.0117397, + "auxiliary_loss_mlp": 0.01037864, + "balance_loss_clip": 1.02075112, + "balance_loss_mlp": 1.02826142, + "epoch": 0.4055792701256538, + "flos": 25521103422720.0, + "grad_norm": 3.4350259965407264, + "language_loss": 0.74269712, + "learning_rate": 2.6941156386539013e-06, + "loss": 0.76481545, + "num_input_tokens_seen": 72834125, + "step": 3373, + "time_per_iteration": 2.693863868713379 + }, + { + "auxiliary_loss_clip": 0.01135801, + "auxiliary_loss_mlp": 0.01034728, + "balance_loss_clip": 0.97920519, + "balance_loss_mlp": 1.02602553, + "epoch": 0.4056995130162929, + "flos": 19574583972480.0, + "grad_norm": 2.7252427120025464, + "language_loss": 0.8117004, + "learning_rate": 2.6933850323453203e-06, + "loss": 0.83340567, + "num_input_tokens_seen": 72852570, + "step": 3374, + "time_per_iteration": 2.6266298294067383 + }, + { + "auxiliary_loss_clip": 0.01191291, + "auxiliary_loss_mlp": 0.01030539, + "balance_loss_clip": 1.06119132, + "balance_loss_mlp": 1.02280784, + "epoch": 0.405819755906932, + "flos": 15413794191360.0, + "grad_norm": 2.130966680314845, + "language_loss": 0.74970615, + "learning_rate": 2.6926543208507806e-06, + "loss": 0.7719245, + "num_input_tokens_seen": 72871250, + "step": 3375, + "time_per_iteration": 3.519662857055664 + }, + { + "auxiliary_loss_clip": 0.01168462, + "auxiliary_loss_mlp": 0.01031644, + "balance_loss_clip": 1.02178824, + "balance_loss_mlp": 1.0228641, + "epoch": 0.4059399987975711, + "flos": 21433930565760.0, + "grad_norm": 4.020049609057898, + "language_loss": 0.80128455, + "learning_rate": 2.6919235042811316e-06, + "loss": 0.82328558, + "num_input_tokens_seen": 72890035, + "step": 3376, + "time_per_iteration": 2.7123610973358154 + }, + { + "auxiliary_loss_clip": 0.01119416, + "auxiliary_loss_mlp": 0.01033338, + "balance_loss_clip": 0.94447267, + "balance_loss_mlp": 1.02440894, + "epoch": 0.4060602416882102, + "flos": 25556942217600.0, + "grad_norm": 3.004276003565281, + "language_loss": 0.76524496, + "learning_rate": 2.691192582747237e-06, + "loss": 0.78677249, + "num_input_tokens_seen": 72909665, + "step": 3377, + "time_per_iteration": 2.7669718265533447 + }, + { + "auxiliary_loss_clip": 0.01190468, + "auxiliary_loss_mlp": 0.01027452, + "balance_loss_clip": 1.05785012, + "balance_loss_mlp": 1.01870239, + "epoch": 0.40618048457884925, + "flos": 23766759262080.0, + "grad_norm": 1.835280033542117, + "language_loss": 0.74222457, + "learning_rate": 2.6904615563599765e-06, + "loss": 0.76440376, + "num_input_tokens_seen": 72929465, + "step": 3378, + "time_per_iteration": 2.6362040042877197 + }, + { + "auxiliary_loss_clip": 0.01117961, + "auxiliary_loss_mlp": 0.01029975, + "balance_loss_clip": 0.94375646, + "balance_loss_mlp": 1.02131426, + "epoch": 0.40630072746948837, + "flos": 17639681120640.0, + "grad_norm": 1.661891700887831, + "language_loss": 0.831604, + "learning_rate": 2.6897304252302477e-06, + "loss": 0.85308337, + "num_input_tokens_seen": 72946785, + "step": 3379, + "time_per_iteration": 2.720813751220703 + }, + { + "auxiliary_loss_clip": 0.01028129, + "auxiliary_loss_mlp": 0.01004755, + "balance_loss_clip": 0.91498637, + "balance_loss_mlp": 1.0030148, + "epoch": 0.4064209703601275, + "flos": 60836053063680.0, + "grad_norm": 0.7945982463792404, + "language_loss": 0.54778802, + "learning_rate": 2.688999189468962e-06, + "loss": 0.5681169, + "num_input_tokens_seen": 73003215, + "step": 3380, + "time_per_iteration": 3.164567232131958 + }, + { + "auxiliary_loss_clip": 0.01165115, + "auxiliary_loss_mlp": 0.01025119, + "balance_loss_clip": 1.02078485, + "balance_loss_mlp": 1.01638055, + "epoch": 0.40654121325076653, + "flos": 24024346669440.0, + "grad_norm": 2.327375338266724, + "language_loss": 0.76167279, + "learning_rate": 2.6882678491870464e-06, + "loss": 0.78357512, + "num_input_tokens_seen": 73023650, + "step": 3381, + "time_per_iteration": 2.703049659729004 + }, + { + "auxiliary_loss_clip": 0.01171211, + "auxiliary_loss_mlp": 0.01033447, + "balance_loss_clip": 1.02127755, + "balance_loss_mlp": 1.02465522, + "epoch": 0.40666145614140564, + "flos": 27344252085120.0, + "grad_norm": 1.7375357932001037, + "language_loss": 0.71820837, + "learning_rate": 2.6875364044954453e-06, + "loss": 0.74025488, + "num_input_tokens_seen": 73043880, + "step": 3382, + "time_per_iteration": 2.6808738708496094 + }, + { + "auxiliary_loss_clip": 0.01141764, + "auxiliary_loss_mlp": 0.01030272, + "balance_loss_clip": 0.97765875, + "balance_loss_mlp": 1.02207613, + "epoch": 0.40678169903204475, + "flos": 26176724415360.0, + "grad_norm": 2.0049431053055593, + "language_loss": 0.82323873, + "learning_rate": 2.6868048555051185e-06, + "loss": 0.84495914, + "num_input_tokens_seen": 73065410, + "step": 3383, + "time_per_iteration": 2.762194871902466 + }, + { + "auxiliary_loss_clip": 0.01150349, + "auxiliary_loss_mlp": 0.01027459, + "balance_loss_clip": 0.98097348, + "balance_loss_mlp": 1.01858926, + "epoch": 0.4069019419226838, + "flos": 28622420622720.0, + "grad_norm": 2.600728141339647, + "language_loss": 0.86416596, + "learning_rate": 2.686073202327041e-06, + "loss": 0.88594401, + "num_input_tokens_seen": 73084410, + "step": 3384, + "time_per_iteration": 2.7075514793395996 + }, + { + "auxiliary_loss_clip": 0.01131594, + "auxiliary_loss_mlp": 0.01031615, + "balance_loss_clip": 0.97610068, + "balance_loss_mlp": 1.02264476, + "epoch": 0.4070221848133229, + "flos": 25229006023680.0, + "grad_norm": 1.819989872720455, + "language_loss": 0.73499942, + "learning_rate": 2.6853414450722043e-06, + "loss": 0.75663149, + "num_input_tokens_seen": 73104075, + "step": 3385, + "time_per_iteration": 2.805588483810425 + }, + { + "auxiliary_loss_clip": 0.01160824, + "auxiliary_loss_mlp": 0.01035624, + "balance_loss_clip": 1.01573062, + "balance_loss_mlp": 1.02711248, + "epoch": 0.40714242770396203, + "flos": 18405224709120.0, + "grad_norm": 1.8208784793167248, + "language_loss": 0.85375828, + "learning_rate": 2.684609583851616e-06, + "loss": 0.87572277, + "num_input_tokens_seen": 73122250, + "step": 3386, + "time_per_iteration": 2.732717514038086 + }, + { + "auxiliary_loss_clip": 0.01095068, + "auxiliary_loss_mlp": 0.01034728, + "balance_loss_clip": 0.90295345, + "balance_loss_mlp": 1.02627635, + "epoch": 0.4072626705946011, + "flos": 30228920403840.0, + "grad_norm": 2.285743995968929, + "language_loss": 0.80778599, + "learning_rate": 2.683877618776297e-06, + "loss": 0.82908392, + "num_input_tokens_seen": 73144505, + "step": 3387, + "time_per_iteration": 2.962249755859375 + }, + { + "auxiliary_loss_clip": 0.0113461, + "auxiliary_loss_mlp": 0.01035818, + "balance_loss_clip": 0.9759419, + "balance_loss_mlp": 1.02671981, + "epoch": 0.4073829134852402, + "flos": 21834549930240.0, + "grad_norm": 2.9405552464806135, + "language_loss": 0.7411446, + "learning_rate": 2.6831455499572876e-06, + "loss": 0.76284885, + "num_input_tokens_seen": 73162440, + "step": 3388, + "time_per_iteration": 2.7943263053894043 + }, + { + "auxiliary_loss_clip": 0.01188722, + "auxiliary_loss_mlp": 0.01027837, + "balance_loss_clip": 1.05572224, + "balance_loss_mlp": 1.01864004, + "epoch": 0.40750315637587925, + "flos": 25260211964160.0, + "grad_norm": 2.0223648182702316, + "language_loss": 0.77696168, + "learning_rate": 2.682413377505641e-06, + "loss": 0.79912734, + "num_input_tokens_seen": 73181245, + "step": 3389, + "time_per_iteration": 2.6972837448120117 + }, + { + "auxiliary_loss_clip": 0.01162624, + "auxiliary_loss_mlp": 0.01029708, + "balance_loss_clip": 1.01532423, + "balance_loss_mlp": 1.02136374, + "epoch": 0.40762339926651836, + "flos": 19712767593600.0, + "grad_norm": 1.831927649067414, + "language_loss": 0.76820207, + "learning_rate": 2.6816811015324284e-06, + "loss": 0.79012537, + "num_input_tokens_seen": 73199295, + "step": 3390, + "time_per_iteration": 2.710456132888794 + }, + { + "auxiliary_loss_clip": 0.01094621, + "auxiliary_loss_mlp": 0.01006938, + "balance_loss_clip": 1.03108287, + "balance_loss_mlp": 1.00548363, + "epoch": 0.40774364215715747, + "flos": 71449307314560.0, + "grad_norm": 0.7290662546405614, + "language_loss": 0.56674111, + "learning_rate": 2.6809487221487343e-06, + "loss": 0.58775675, + "num_input_tokens_seen": 73258780, + "step": 3391, + "time_per_iteration": 3.1813747882843018 + }, + { + "auxiliary_loss_clip": 0.01154436, + "auxiliary_loss_mlp": 0.01032574, + "balance_loss_clip": 1.01330364, + "balance_loss_mlp": 1.02344871, + "epoch": 0.4078638850477965, + "flos": 15084134144640.0, + "grad_norm": 2.7212469238906483, + "language_loss": 0.82089233, + "learning_rate": 2.6802162394656605e-06, + "loss": 0.84276247, + "num_input_tokens_seen": 73275490, + "step": 3392, + "time_per_iteration": 2.7363762855529785 + }, + { + "auxiliary_loss_clip": 0.01136777, + "auxiliary_loss_mlp": 0.01032281, + "balance_loss_clip": 0.97735995, + "balance_loss_mlp": 1.0235486, + "epoch": 0.40798412793843564, + "flos": 23842890138240.0, + "grad_norm": 1.7245801017201485, + "language_loss": 0.71579319, + "learning_rate": 2.679483653594324e-06, + "loss": 0.73748374, + "num_input_tokens_seen": 73297260, + "step": 3393, + "time_per_iteration": 2.7225592136383057 + }, + { + "auxiliary_loss_clip": 0.01165982, + "auxiliary_loss_mlp": 0.01034873, + "balance_loss_clip": 1.0180192, + "balance_loss_mlp": 1.02656364, + "epoch": 0.40810437082907475, + "flos": 21065774117760.0, + "grad_norm": 2.299380692595503, + "language_loss": 0.77230078, + "learning_rate": 2.678750964645857e-06, + "loss": 0.79430938, + "num_input_tokens_seen": 73316340, + "step": 3394, + "time_per_iteration": 2.6374223232269287 + }, + { + "auxiliary_loss_clip": 0.01167034, + "auxiliary_loss_mlp": 0.01033843, + "balance_loss_clip": 1.02203989, + "balance_loss_mlp": 1.02466989, + "epoch": 0.4082246137197138, + "flos": 11321377948800.0, + "grad_norm": 2.429530338473143, + "language_loss": 0.83695495, + "learning_rate": 2.6780181727314094e-06, + "loss": 0.85896373, + "num_input_tokens_seen": 73331245, + "step": 3395, + "time_per_iteration": 4.420885324478149 + }, + { + "auxiliary_loss_clip": 0.0112131, + "auxiliary_loss_mlp": 0.00510478, + "balance_loss_clip": 0.94219053, + "balance_loss_mlp": 0.0, + "epoch": 0.4083448566103529, + "flos": 19062569554560.0, + "grad_norm": 1.8380528033367771, + "language_loss": 0.78029519, + "learning_rate": 2.6772852779621435e-06, + "loss": 0.7966131, + "num_input_tokens_seen": 73349105, + "step": 3396, + "time_per_iteration": 2.6681907176971436 + }, + { + "auxiliary_loss_clip": 0.01159835, + "auxiliary_loss_mlp": 0.00510686, + "balance_loss_clip": 1.01997674, + "balance_loss_mlp": 0.0, + "epoch": 0.408465099500992, + "flos": 23550254035200.0, + "grad_norm": 1.795044840553996, + "language_loss": 0.8646251, + "learning_rate": 2.676552280449239e-06, + "loss": 0.88133025, + "num_input_tokens_seen": 73368990, + "step": 3397, + "time_per_iteration": 3.58202862739563 + }, + { + "auxiliary_loss_clip": 0.01157369, + "auxiliary_loss_mlp": 0.01030504, + "balance_loss_clip": 1.01626492, + "balance_loss_mlp": 1.02114022, + "epoch": 0.4085853423916311, + "flos": 12750012558720.0, + "grad_norm": 2.2831623945517765, + "language_loss": 0.75648069, + "learning_rate": 2.6758191803038917e-06, + "loss": 0.77835941, + "num_input_tokens_seen": 73387485, + "step": 3398, + "time_per_iteration": 2.673022985458374 + }, + { + "auxiliary_loss_clip": 0.01068056, + "auxiliary_loss_mlp": 0.01035574, + "balance_loss_clip": 0.86914331, + "balance_loss_mlp": 1.02688944, + "epoch": 0.4087055852822702, + "flos": 24353072962560.0, + "grad_norm": 2.1388222262793226, + "language_loss": 0.82850516, + "learning_rate": 2.6750859776373125e-06, + "loss": 0.84954143, + "num_input_tokens_seen": 73406940, + "step": 3399, + "time_per_iteration": 2.836843490600586 + }, + { + "auxiliary_loss_clip": 0.00992658, + "auxiliary_loss_mlp": 0.01000981, + "balance_loss_clip": 0.83887851, + "balance_loss_mlp": 0.99952644, + "epoch": 0.4088258281729093, + "flos": 66387950720640.0, + "grad_norm": 0.770241654594553, + "language_loss": 0.60446763, + "learning_rate": 2.674352672560727e-06, + "loss": 0.62440407, + "num_input_tokens_seen": 73468385, + "step": 3400, + "time_per_iteration": 3.418128728866577 + }, + { + "auxiliary_loss_clip": 0.01113652, + "auxiliary_loss_mlp": 0.01031725, + "balance_loss_clip": 0.93774289, + "balance_loss_mlp": 1.02310658, + "epoch": 0.40894607106354836, + "flos": 20449260057600.0, + "grad_norm": 2.0396330769571005, + "language_loss": 0.7684257, + "learning_rate": 2.673619265185377e-06, + "loss": 0.78987944, + "num_input_tokens_seen": 73488225, + "step": 3401, + "time_per_iteration": 3.9053196907043457 + }, + { + "auxiliary_loss_clip": 0.01167952, + "auxiliary_loss_mlp": 0.0103819, + "balance_loss_clip": 1.01802981, + "balance_loss_mlp": 1.02853382, + "epoch": 0.40906631395418747, + "flos": 27053627143680.0, + "grad_norm": 1.652015699220305, + "language_loss": 0.78058094, + "learning_rate": 2.672885755622521e-06, + "loss": 0.80264235, + "num_input_tokens_seen": 73510640, + "step": 3402, + "time_per_iteration": 2.654262065887451 + }, + { + "auxiliary_loss_clip": 0.01088765, + "auxiliary_loss_mlp": 0.01028371, + "balance_loss_clip": 0.90012443, + "balance_loss_mlp": 1.02075696, + "epoch": 0.4091865568448266, + "flos": 25484151306240.0, + "grad_norm": 1.941417782240784, + "language_loss": 0.70062101, + "learning_rate": 2.67215214398343e-06, + "loss": 0.72179234, + "num_input_tokens_seen": 73530655, + "step": 3403, + "time_per_iteration": 2.7448503971099854 + }, + { + "auxiliary_loss_clip": 0.01098505, + "auxiliary_loss_mlp": 0.01027615, + "balance_loss_clip": 0.90190822, + "balance_loss_mlp": 1.0191865, + "epoch": 0.40930679973546563, + "flos": 28657864368000.0, + "grad_norm": 2.2496625471015674, + "language_loss": 0.78396368, + "learning_rate": 2.671418430379393e-06, + "loss": 0.8052249, + "num_input_tokens_seen": 73549340, + "step": 3404, + "time_per_iteration": 2.8404476642608643 + }, + { + "auxiliary_loss_clip": 0.01187291, + "auxiliary_loss_mlp": 0.01031309, + "balance_loss_clip": 1.05576444, + "balance_loss_mlp": 1.02305377, + "epoch": 0.40942704262610474, + "flos": 20886292834560.0, + "grad_norm": 1.8495257716138074, + "language_loss": 0.8321929, + "learning_rate": 2.670684614921715e-06, + "loss": 0.85437894, + "num_input_tokens_seen": 73568315, + "step": 3405, + "time_per_iteration": 2.631112813949585 + }, + { + "auxiliary_loss_clip": 0.01141564, + "auxiliary_loss_mlp": 0.01030303, + "balance_loss_clip": 0.97833824, + "balance_loss_mlp": 1.02199996, + "epoch": 0.4095472855167438, + "flos": 21618080616960.0, + "grad_norm": 2.1793510897547708, + "language_loss": 0.69116235, + "learning_rate": 2.6699506977217128e-06, + "loss": 0.71288097, + "num_input_tokens_seen": 73588490, + "step": 3406, + "time_per_iteration": 2.677408218383789 + }, + { + "auxiliary_loss_clip": 0.01162589, + "auxiliary_loss_mlp": 0.01034614, + "balance_loss_clip": 1.01991153, + "balance_loss_mlp": 1.02638865, + "epoch": 0.4096675284073829, + "flos": 27926112499200.0, + "grad_norm": 2.046378340798083, + "language_loss": 0.70062184, + "learning_rate": 2.6692166788907233e-06, + "loss": 0.72259384, + "num_input_tokens_seen": 73608685, + "step": 3407, + "time_per_iteration": 2.730438709259033 + }, + { + "auxiliary_loss_clip": 0.01144872, + "auxiliary_loss_mlp": 0.0102937, + "balance_loss_clip": 0.98162585, + "balance_loss_mlp": 1.02085209, + "epoch": 0.409787771298022, + "flos": 19206607092480.0, + "grad_norm": 1.8202925268031682, + "language_loss": 0.76948947, + "learning_rate": 2.6684825585400957e-06, + "loss": 0.79123187, + "num_input_tokens_seen": 73627630, + "step": 3408, + "time_per_iteration": 2.6525278091430664 + }, + { + "auxiliary_loss_clip": 0.01047054, + "auxiliary_loss_mlp": 0.01001204, + "balance_loss_clip": 0.95263433, + "balance_loss_mlp": 0.99977922, + "epoch": 0.4099080141886611, + "flos": 59269234832640.0, + "grad_norm": 2.333416309003039, + "language_loss": 0.65122819, + "learning_rate": 2.6677483367811947e-06, + "loss": 0.67171073, + "num_input_tokens_seen": 73687670, + "step": 3409, + "time_per_iteration": 3.354959011077881 + }, + { + "auxiliary_loss_clip": 0.01165424, + "auxiliary_loss_mlp": 0.01029522, + "balance_loss_clip": 1.01710343, + "balance_loss_mlp": 1.02118373, + "epoch": 0.4100282570793002, + "flos": 21906443001600.0, + "grad_norm": 1.8946243872858048, + "language_loss": 0.75385153, + "learning_rate": 2.6670140137254028e-06, + "loss": 0.77580094, + "num_input_tokens_seen": 73707145, + "step": 3410, + "time_per_iteration": 2.6411988735198975 + }, + { + "auxiliary_loss_clip": 0.01088538, + "auxiliary_loss_mlp": 0.0103064, + "balance_loss_clip": 0.9004817, + "balance_loss_mlp": 1.02203929, + "epoch": 0.4101484999699393, + "flos": 18551596631040.0, + "grad_norm": 2.212626661508188, + "language_loss": 0.89785171, + "learning_rate": 2.666279589484115e-06, + "loss": 0.91904342, + "num_input_tokens_seen": 73725045, + "step": 3411, + "time_per_iteration": 2.7349965572357178 + }, + { + "auxiliary_loss_clip": 0.01094381, + "auxiliary_loss_mlp": 0.01027578, + "balance_loss_clip": 0.90172297, + "balance_loss_mlp": 1.0190959, + "epoch": 0.41026874286057835, + "flos": 19094529680640.0, + "grad_norm": 2.0242915108224926, + "language_loss": 0.80812138, + "learning_rate": 2.6655450641687435e-06, + "loss": 0.82934105, + "num_input_tokens_seen": 73742610, + "step": 3412, + "time_per_iteration": 2.810927629470825 + }, + { + "auxiliary_loss_clip": 0.01186505, + "auxiliary_loss_mlp": 0.01029408, + "balance_loss_clip": 1.05684543, + "balance_loss_mlp": 1.02051449, + "epoch": 0.41038898575121746, + "flos": 31209568588800.0, + "grad_norm": 1.66323079903158, + "language_loss": 0.69343913, + "learning_rate": 2.664810437890715e-06, + "loss": 0.71559823, + "num_input_tokens_seen": 73764280, + "step": 3413, + "time_per_iteration": 2.6968183517456055 + }, + { + "auxiliary_loss_clip": 0.01064469, + "auxiliary_loss_mlp": 0.01030907, + "balance_loss_clip": 0.87042546, + "balance_loss_mlp": 1.02314937, + "epoch": 0.41050922864185657, + "flos": 14355865895040.0, + "grad_norm": 1.8653171152787744, + "language_loss": 0.79465395, + "learning_rate": 2.6640757107614714e-06, + "loss": 0.81560767, + "num_input_tokens_seen": 73782375, + "step": 3414, + "time_per_iteration": 2.739098072052002 + }, + { + "auxiliary_loss_clip": 0.01111389, + "auxiliary_loss_mlp": 0.01030543, + "balance_loss_clip": 0.9426834, + "balance_loss_mlp": 1.02193666, + "epoch": 0.4106294715324956, + "flos": 30956290813440.0, + "grad_norm": 2.5259985479838427, + "language_loss": 0.69088095, + "learning_rate": 2.6633408828924697e-06, + "loss": 0.7123003, + "num_input_tokens_seen": 73801240, + "step": 3415, + "time_per_iteration": 2.7771079540252686 + }, + { + "auxiliary_loss_clip": 0.01122943, + "auxiliary_loss_mlp": 0.01030458, + "balance_loss_clip": 0.94156933, + "balance_loss_mlp": 1.02273345, + "epoch": 0.41074971442313474, + "flos": 24457321209600.0, + "grad_norm": 1.7663017604626585, + "language_loss": 0.69995081, + "learning_rate": 2.662605954395185e-06, + "loss": 0.7214849, + "num_input_tokens_seen": 73821200, + "step": 3416, + "time_per_iteration": 2.7363622188568115 + }, + { + "auxiliary_loss_clip": 0.01168476, + "auxiliary_loss_mlp": 0.0103072, + "balance_loss_clip": 1.01920891, + "balance_loss_mlp": 1.02170134, + "epoch": 0.41086995731377385, + "flos": 21542991235200.0, + "grad_norm": 1.7611135208751758, + "language_loss": 0.8340739, + "learning_rate": 2.6618709253811027e-06, + "loss": 0.85606587, + "num_input_tokens_seen": 73840655, + "step": 3417, + "time_per_iteration": 2.784750461578369 + }, + { + "auxiliary_loss_clip": 0.01186528, + "auxiliary_loss_mlp": 0.01024891, + "balance_loss_clip": 1.05947828, + "balance_loss_mlp": 1.01708889, + "epoch": 0.4109902002044129, + "flos": 20702753314560.0, + "grad_norm": 1.6002536878344884, + "language_loss": 0.87731111, + "learning_rate": 2.6611357959617277e-06, + "loss": 0.89942527, + "num_input_tokens_seen": 73860275, + "step": 3418, + "time_per_iteration": 2.6413707733154297 + }, + { + "auxiliary_loss_clip": 0.01108708, + "auxiliary_loss_mlp": 0.01025354, + "balance_loss_clip": 0.94040179, + "balance_loss_mlp": 1.01700318, + "epoch": 0.411110443095052, + "flos": 18179992477440.0, + "grad_norm": 1.818721210343612, + "language_loss": 0.9125272, + "learning_rate": 2.660400566248578e-06, + "loss": 0.93386781, + "num_input_tokens_seen": 73878400, + "step": 3419, + "time_per_iteration": 2.6745598316192627 + }, + { + "auxiliary_loss_clip": 0.01116927, + "auxiliary_loss_mlp": 0.01026832, + "balance_loss_clip": 0.93937886, + "balance_loss_mlp": 1.01784396, + "epoch": 0.41123068598569107, + "flos": 14575244209920.0, + "grad_norm": 2.825988437501278, + "language_loss": 0.66848791, + "learning_rate": 2.6596652363531876e-06, + "loss": 0.68992555, + "num_input_tokens_seen": 73894275, + "step": 3420, + "time_per_iteration": 2.753251075744629 + }, + { + "auxiliary_loss_clip": 0.01188437, + "auxiliary_loss_mlp": 0.01030795, + "balance_loss_clip": 1.05946672, + "balance_loss_mlp": 1.02274215, + "epoch": 0.4113509288763302, + "flos": 21177995184000.0, + "grad_norm": 1.5825684782576288, + "language_loss": 0.78336018, + "learning_rate": 2.6589298063871055e-06, + "loss": 0.80555248, + "num_input_tokens_seen": 73914450, + "step": 3421, + "time_per_iteration": 4.560885906219482 + }, + { + "auxiliary_loss_clip": 0.0118615, + "auxiliary_loss_mlp": 0.01031017, + "balance_loss_clip": 1.05777907, + "balance_loss_mlp": 1.02261305, + "epoch": 0.4114711717669693, + "flos": 18442212739200.0, + "grad_norm": 1.8546328408270991, + "language_loss": 0.69733518, + "learning_rate": 2.658194276461895e-06, + "loss": 0.7195068, + "num_input_tokens_seen": 73932375, + "step": 3422, + "time_per_iteration": 2.634904623031616 + }, + { + "auxiliary_loss_clip": 0.01137672, + "auxiliary_loss_mlp": 0.01034014, + "balance_loss_clip": 0.97540528, + "balance_loss_mlp": 1.02486527, + "epoch": 0.41159141465760835, + "flos": 27233395735680.0, + "grad_norm": 1.887433218130983, + "language_loss": 0.674366, + "learning_rate": 2.6574586466891368e-06, + "loss": 0.69608289, + "num_input_tokens_seen": 73952850, + "step": 3423, + "time_per_iteration": 3.711721420288086 + }, + { + "auxiliary_loss_clip": 0.01140367, + "auxiliary_loss_mlp": 0.00509683, + "balance_loss_clip": 0.97977787, + "balance_loss_mlp": 0.0, + "epoch": 0.41171165754824746, + "flos": 20006876154240.0, + "grad_norm": 2.2702180306328326, + "language_loss": 0.64983952, + "learning_rate": 2.6567229171804247e-06, + "loss": 0.66633999, + "num_input_tokens_seen": 73970735, + "step": 3424, + "time_per_iteration": 2.682624340057373 + }, + { + "auxiliary_loss_clip": 0.01119639, + "auxiliary_loss_mlp": 0.01029301, + "balance_loss_clip": 0.93725109, + "balance_loss_mlp": 1.02059913, + "epoch": 0.41183190043888657, + "flos": 18004318035840.0, + "grad_norm": 2.3500662799604113, + "language_loss": 0.87975466, + "learning_rate": 2.655987088047368e-06, + "loss": 0.90124404, + "num_input_tokens_seen": 73989080, + "step": 3425, + "time_per_iteration": 2.706106662750244 + }, + { + "auxiliary_loss_clip": 0.01134697, + "auxiliary_loss_mlp": 0.01025641, + "balance_loss_clip": 0.97666317, + "balance_loss_mlp": 1.0173614, + "epoch": 0.4119521433295256, + "flos": 27163370171520.0, + "grad_norm": 2.50723924506989, + "language_loss": 0.78333354, + "learning_rate": 2.6552511594015912e-06, + "loss": 0.80493689, + "num_input_tokens_seen": 74009470, + "step": 3426, + "time_per_iteration": 2.7675893306732178 + }, + { + "auxiliary_loss_clip": 0.01136244, + "auxiliary_loss_mlp": 0.01031428, + "balance_loss_clip": 0.97506869, + "balance_loss_mlp": 1.02241564, + "epoch": 0.41207238622016473, + "flos": 15122020014720.0, + "grad_norm": 2.149837130407334, + "language_loss": 0.8521812, + "learning_rate": 2.654515131354735e-06, + "loss": 0.87385786, + "num_input_tokens_seen": 74027735, + "step": 3427, + "time_per_iteration": 2.7220466136932373 + }, + { + "auxiliary_loss_clip": 0.01116392, + "auxiliary_loss_mlp": 0.01026448, + "balance_loss_clip": 0.94090271, + "balance_loss_mlp": 1.01789522, + "epoch": 0.41219262911080384, + "flos": 27052872958080.0, + "grad_norm": 2.2103352809233074, + "language_loss": 0.85277879, + "learning_rate": 2.653779004018453e-06, + "loss": 0.8742072, + "num_input_tokens_seen": 74048300, + "step": 3428, + "time_per_iteration": 3.6972527503967285 + }, + { + "auxiliary_loss_clip": 0.01132236, + "auxiliary_loss_mlp": 0.0102538, + "balance_loss_clip": 0.97730696, + "balance_loss_mlp": 1.0175066, + "epoch": 0.4123128720014429, + "flos": 24686360282880.0, + "grad_norm": 1.7366129427160795, + "language_loss": 0.82309866, + "learning_rate": 2.653042777504417e-06, + "loss": 0.84467483, + "num_input_tokens_seen": 74070890, + "step": 3429, + "time_per_iteration": 2.743795394897461 + }, + { + "auxiliary_loss_clip": 0.01145386, + "auxiliary_loss_mlp": 0.01027775, + "balance_loss_clip": 0.97872615, + "balance_loss_mlp": 1.01971626, + "epoch": 0.412433114892082, + "flos": 26244774731520.0, + "grad_norm": 1.7802604726034295, + "language_loss": 0.80071014, + "learning_rate": 2.6523064519243105e-06, + "loss": 0.82244182, + "num_input_tokens_seen": 74090460, + "step": 3430, + "time_per_iteration": 2.678849935531616 + }, + { + "auxiliary_loss_clip": 0.01165499, + "auxiliary_loss_mlp": 0.01028797, + "balance_loss_clip": 1.02015293, + "balance_loss_mlp": 1.02031565, + "epoch": 0.4125533577827211, + "flos": 21361031913600.0, + "grad_norm": 2.2987307799480874, + "language_loss": 0.79029429, + "learning_rate": 2.6515700273898333e-06, + "loss": 0.81223726, + "num_input_tokens_seen": 74108335, + "step": 3431, + "time_per_iteration": 2.7466938495635986 + }, + { + "auxiliary_loss_clip": 0.01129436, + "auxiliary_loss_mlp": 0.01030297, + "balance_loss_clip": 0.97965389, + "balance_loss_mlp": 1.0217917, + "epoch": 0.4126736006733602, + "flos": 26067556005120.0, + "grad_norm": 2.8864879403325063, + "language_loss": 0.68898964, + "learning_rate": 2.6508335040127018e-06, + "loss": 0.71058691, + "num_input_tokens_seen": 74128030, + "step": 3432, + "time_per_iteration": 2.695249319076538 + }, + { + "auxiliary_loss_clip": 0.01168033, + "auxiliary_loss_mlp": 0.01031845, + "balance_loss_clip": 1.01956391, + "balance_loss_mlp": 1.02366686, + "epoch": 0.4127938435639993, + "flos": 25666146541440.0, + "grad_norm": 1.439895561618955, + "language_loss": 0.77125978, + "learning_rate": 2.6500968819046446e-06, + "loss": 0.79325855, + "num_input_tokens_seen": 74148330, + "step": 3433, + "time_per_iteration": 2.686575412750244 + }, + { + "auxiliary_loss_clip": 0.01102465, + "auxiliary_loss_mlp": 0.01029716, + "balance_loss_clip": 0.93614429, + "balance_loss_mlp": 1.02113914, + "epoch": 0.4129140864546384, + "flos": 17995914253440.0, + "grad_norm": 2.3888771842764, + "language_loss": 0.58881962, + "learning_rate": 2.649360161177408e-06, + "loss": 0.61014146, + "num_input_tokens_seen": 74163390, + "step": 3434, + "time_per_iteration": 2.6597626209259033 + }, + { + "auxiliary_loss_clip": 0.01170956, + "auxiliary_loss_mlp": 0.01030615, + "balance_loss_clip": 1.01964045, + "balance_loss_mlp": 1.02229989, + "epoch": 0.41303432934527745, + "flos": 23732895715200.0, + "grad_norm": 1.8518516946409977, + "language_loss": 0.73613447, + "learning_rate": 2.6486233419427504e-06, + "loss": 0.75815016, + "num_input_tokens_seen": 74183205, + "step": 3435, + "time_per_iteration": 2.698552370071411 + }, + { + "auxiliary_loss_clip": 0.01111289, + "auxiliary_loss_mlp": 0.0103341, + "balance_loss_clip": 0.94198692, + "balance_loss_mlp": 1.0247854, + "epoch": 0.41315457223591656, + "flos": 19755286318080.0, + "grad_norm": 5.551023073839574, + "language_loss": 0.75190723, + "learning_rate": 2.6478864243124484e-06, + "loss": 0.77335423, + "num_input_tokens_seen": 74202870, + "step": 3436, + "time_per_iteration": 2.7284598350524902 + }, + { + "auxiliary_loss_clip": 0.01164122, + "auxiliary_loss_mlp": 0.0103014, + "balance_loss_clip": 1.01574266, + "balance_loss_mlp": 1.02193248, + "epoch": 0.4132748151265556, + "flos": 20923316778240.0, + "grad_norm": 6.854830019031289, + "language_loss": 0.85162628, + "learning_rate": 2.6471494083982903e-06, + "loss": 0.87356889, + "num_input_tokens_seen": 74222255, + "step": 3437, + "time_per_iteration": 2.6916377544403076 + }, + { + "auxiliary_loss_clip": 0.01118708, + "auxiliary_loss_mlp": 0.01031453, + "balance_loss_clip": 0.93892241, + "balance_loss_mlp": 1.02264929, + "epoch": 0.4133950580171947, + "flos": 32232520016640.0, + "grad_norm": 2.103854970647903, + "language_loss": 0.74857306, + "learning_rate": 2.6464122943120818e-06, + "loss": 0.77007461, + "num_input_tokens_seen": 74242480, + "step": 3438, + "time_per_iteration": 2.779881000518799 + }, + { + "auxiliary_loss_clip": 0.01116863, + "auxiliary_loss_mlp": 0.01029245, + "balance_loss_clip": 0.94423091, + "balance_loss_mlp": 1.02097189, + "epoch": 0.41351530090783384, + "flos": 23292487059840.0, + "grad_norm": 2.955945730989014, + "language_loss": 0.81890559, + "learning_rate": 2.645675082165642e-06, + "loss": 0.84036672, + "num_input_tokens_seen": 74258690, + "step": 3439, + "time_per_iteration": 2.7177910804748535 + }, + { + "auxiliary_loss_clip": 0.01144486, + "auxiliary_loss_mlp": 0.01031762, + "balance_loss_clip": 0.98299879, + "balance_loss_mlp": 1.02358449, + "epoch": 0.4136355437984729, + "flos": 25593571111680.0, + "grad_norm": 2.224046660789319, + "language_loss": 0.75128078, + "learning_rate": 2.644937772070806e-06, + "loss": 0.77304327, + "num_input_tokens_seen": 74277135, + "step": 3440, + "time_per_iteration": 2.7262630462646484 + }, + { + "auxiliary_loss_clip": 0.01189988, + "auxiliary_loss_mlp": 0.01023142, + "balance_loss_clip": 1.0593555, + "balance_loss_mlp": 1.01511288, + "epoch": 0.413755786689112, + "flos": 19828615933440.0, + "grad_norm": 2.637804509440135, + "language_loss": 0.83695132, + "learning_rate": 2.6442003641394225e-06, + "loss": 0.85908258, + "num_input_tokens_seen": 74294730, + "step": 3441, + "time_per_iteration": 2.6713991165161133 + }, + { + "auxiliary_loss_clip": 0.01136752, + "auxiliary_loss_mlp": 0.0102597, + "balance_loss_clip": 0.97754037, + "balance_loss_mlp": 1.01776886, + "epoch": 0.4138760295797511, + "flos": 26870446759680.0, + "grad_norm": 1.6387524307128658, + "language_loss": 0.83993649, + "learning_rate": 2.643462858483356e-06, + "loss": 0.86156374, + "num_input_tokens_seen": 74315015, + "step": 3442, + "time_per_iteration": 2.7755932807922363 + }, + { + "auxiliary_loss_clip": 0.01086977, + "auxiliary_loss_mlp": 0.01027966, + "balance_loss_clip": 0.90064561, + "balance_loss_mlp": 1.01937723, + "epoch": 0.41399627247039017, + "flos": 16399254798720.0, + "grad_norm": 1.8649608409249177, + "language_loss": 0.72773677, + "learning_rate": 2.6427252552144856e-06, + "loss": 0.74888623, + "num_input_tokens_seen": 74333665, + "step": 3443, + "time_per_iteration": 2.7879397869110107 + }, + { + "auxiliary_loss_clip": 0.01188537, + "auxiliary_loss_mlp": 0.01028896, + "balance_loss_clip": 1.0577389, + "balance_loss_mlp": 1.02044976, + "epoch": 0.4141165153610293, + "flos": 22930220442240.0, + "grad_norm": 2.343884697470656, + "language_loss": 0.75114155, + "learning_rate": 2.6419875544447044e-06, + "loss": 0.77331591, + "num_input_tokens_seen": 74355065, + "step": 3444, + "time_per_iteration": 2.7136247158050537 + }, + { + "auxiliary_loss_clip": 0.01188354, + "auxiliary_loss_mlp": 0.01036328, + "balance_loss_clip": 1.05672097, + "balance_loss_mlp": 1.02834082, + "epoch": 0.4142367582516684, + "flos": 25192556697600.0, + "grad_norm": 1.7886225690422701, + "language_loss": 0.71463495, + "learning_rate": 2.6412497562859218e-06, + "loss": 0.73688173, + "num_input_tokens_seen": 74376345, + "step": 3445, + "time_per_iteration": 2.6330249309539795 + }, + { + "auxiliary_loss_clip": 0.01168851, + "auxiliary_loss_mlp": 0.01032252, + "balance_loss_clip": 1.0182445, + "balance_loss_mlp": 1.02357984, + "epoch": 0.41435700114230745, + "flos": 21690476478720.0, + "grad_norm": 2.211425060330564, + "language_loss": 0.76392883, + "learning_rate": 2.6405118608500617e-06, + "loss": 0.78593987, + "num_input_tokens_seen": 74395170, + "step": 3446, + "time_per_iteration": 3.613607883453369 + }, + { + "auxiliary_loss_clip": 0.011088, + "auxiliary_loss_mlp": 0.01029113, + "balance_loss_clip": 0.94269407, + "balance_loss_mlp": 1.02165091, + "epoch": 0.41447724403294656, + "flos": 25995160143360.0, + "grad_norm": 1.7124123186366906, + "language_loss": 0.81197548, + "learning_rate": 2.6397738682490613e-06, + "loss": 0.83335459, + "num_input_tokens_seen": 74416070, + "step": 3447, + "time_per_iteration": 3.7237024307250977 + }, + { + "auxiliary_loss_clip": 0.01188543, + "auxiliary_loss_mlp": 0.01028757, + "balance_loss_clip": 1.05770767, + "balance_loss_mlp": 1.0200249, + "epoch": 0.41459748692358567, + "flos": 18259678800000.0, + "grad_norm": 1.7984073356740216, + "language_loss": 0.75219208, + "learning_rate": 2.6390357785948734e-06, + "loss": 0.77436507, + "num_input_tokens_seen": 74433185, + "step": 3448, + "time_per_iteration": 2.5947721004486084 + }, + { + "auxiliary_loss_clip": 0.01167226, + "auxiliary_loss_mlp": 0.01028311, + "balance_loss_clip": 1.02033496, + "balance_loss_mlp": 1.01968062, + "epoch": 0.4147177298142247, + "flos": 24168456034560.0, + "grad_norm": 1.6681895326952245, + "language_loss": 0.8069905, + "learning_rate": 2.6382975919994667e-06, + "loss": 0.82894588, + "num_input_tokens_seen": 74453760, + "step": 3449, + "time_per_iteration": 3.6162006855010986 + }, + { + "auxiliary_loss_clip": 0.01142365, + "auxiliary_loss_mlp": 0.01030025, + "balance_loss_clip": 0.97887909, + "balance_loss_mlp": 1.02236533, + "epoch": 0.41483797270486383, + "flos": 20084659056000.0, + "grad_norm": 1.8347423857437297, + "language_loss": 0.73229468, + "learning_rate": 2.637559308574822e-06, + "loss": 0.7540186, + "num_input_tokens_seen": 74473505, + "step": 3450, + "time_per_iteration": 2.6237432956695557 + }, + { + "auxiliary_loss_clip": 0.01189457, + "auxiliary_loss_mlp": 0.01028911, + "balance_loss_clip": 1.05849075, + "balance_loss_mlp": 1.02049518, + "epoch": 0.4149582155955029, + "flos": 30081040110720.0, + "grad_norm": 2.728222097562976, + "language_loss": 0.71265113, + "learning_rate": 2.6368209284329376e-06, + "loss": 0.73483485, + "num_input_tokens_seen": 74494135, + "step": 3451, + "time_per_iteration": 2.654784679412842 + }, + { + "auxiliary_loss_clip": 0.01164212, + "auxiliary_loss_mlp": 0.01031173, + "balance_loss_clip": 1.0176878, + "balance_loss_mlp": 1.02283442, + "epoch": 0.415078458486142, + "flos": 16764394504320.0, + "grad_norm": 2.34024631933297, + "language_loss": 0.75474727, + "learning_rate": 2.636082451685825e-06, + "loss": 0.77670115, + "num_input_tokens_seen": 74512335, + "step": 3452, + "time_per_iteration": 2.5700836181640625 + }, + { + "auxiliary_loss_clip": 0.01147801, + "auxiliary_loss_mlp": 0.01030343, + "balance_loss_clip": 0.98496473, + "balance_loss_mlp": 1.02189136, + "epoch": 0.4151987013767811, + "flos": 26033692458240.0, + "grad_norm": 1.5439189014204047, + "language_loss": 0.86303341, + "learning_rate": 2.6353438784455094e-06, + "loss": 0.88481486, + "num_input_tokens_seen": 74535620, + "step": 3453, + "time_per_iteration": 3.541104316711426 + }, + { + "auxiliary_loss_clip": 0.01135115, + "auxiliary_loss_mlp": 0.01036022, + "balance_loss_clip": 0.97935551, + "balance_loss_mlp": 1.02681875, + "epoch": 0.41531894426742016, + "flos": 24608002763520.0, + "grad_norm": 2.1365274584867504, + "language_loss": 0.71524936, + "learning_rate": 2.6346052088240326e-06, + "loss": 0.73696077, + "num_input_tokens_seen": 74555140, + "step": 3454, + "time_per_iteration": 2.6037099361419678 + }, + { + "auxiliary_loss_clip": 0.01143169, + "auxiliary_loss_mlp": 0.0102981, + "balance_loss_clip": 0.98105961, + "balance_loss_mlp": 1.02129841, + "epoch": 0.4154391871580593, + "flos": 14975791747200.0, + "grad_norm": 2.275471857146762, + "language_loss": 0.77385503, + "learning_rate": 2.63386644293345e-06, + "loss": 0.7955848, + "num_input_tokens_seen": 74571485, + "step": 3455, + "time_per_iteration": 2.5539515018463135 + }, + { + "auxiliary_loss_clip": 0.0111321, + "auxiliary_loss_mlp": 0.01032187, + "balance_loss_clip": 0.93759823, + "balance_loss_mlp": 1.02433729, + "epoch": 0.4155594300486984, + "flos": 14647173194880.0, + "grad_norm": 3.1087613454895413, + "language_loss": 0.82579136, + "learning_rate": 2.633127580885833e-06, + "loss": 0.84724528, + "num_input_tokens_seen": 74585985, + "step": 3456, + "time_per_iteration": 2.7261698246002197 + }, + { + "auxiliary_loss_clip": 0.01191714, + "auxiliary_loss_mlp": 0.01027051, + "balance_loss_clip": 1.06207633, + "balance_loss_mlp": 1.01917684, + "epoch": 0.41567967293933744, + "flos": 29497276275840.0, + "grad_norm": 5.889477357650989, + "language_loss": 0.65433967, + "learning_rate": 2.632388622793265e-06, + "loss": 0.67652732, + "num_input_tokens_seen": 74605140, + "step": 3457, + "time_per_iteration": 2.6553328037261963 + }, + { + "auxiliary_loss_clip": 0.01166016, + "auxiliary_loss_mlp": 0.01031814, + "balance_loss_clip": 1.02179348, + "balance_loss_mlp": 1.02376771, + "epoch": 0.41579991582997655, + "flos": 19238387650560.0, + "grad_norm": 1.7497580174820746, + "language_loss": 0.67822886, + "learning_rate": 2.6316495687678457e-06, + "loss": 0.70020711, + "num_input_tokens_seen": 74623790, + "step": 3458, + "time_per_iteration": 2.640658378601074 + }, + { + "auxiliary_loss_clip": 0.01086274, + "auxiliary_loss_mlp": 0.01028555, + "balance_loss_clip": 0.90065849, + "balance_loss_mlp": 1.02030528, + "epoch": 0.41592015872061566, + "flos": 24462061804800.0, + "grad_norm": 6.602090755490563, + "language_loss": 0.76567197, + "learning_rate": 2.6309104189216887e-06, + "loss": 0.78682023, + "num_input_tokens_seen": 74641355, + "step": 3459, + "time_per_iteration": 2.755277395248413 + }, + { + "auxiliary_loss_clip": 0.01101059, + "auxiliary_loss_mlp": 0.00511205, + "balance_loss_clip": 0.93604147, + "balance_loss_mlp": 0.0, + "epoch": 0.4160404016112547, + "flos": 20775651966720.0, + "grad_norm": 2.346524151526365, + "language_loss": 0.74608278, + "learning_rate": 2.630171173366923e-06, + "loss": 0.76220536, + "num_input_tokens_seen": 74657155, + "step": 3460, + "time_per_iteration": 2.6927592754364014 + }, + { + "auxiliary_loss_clip": 0.01092042, + "auxiliary_loss_mlp": 0.01035337, + "balance_loss_clip": 0.90221953, + "balance_loss_mlp": 1.02717137, + "epoch": 0.41616064450189383, + "flos": 13916462820480.0, + "grad_norm": 2.4221595594220404, + "language_loss": 0.74203628, + "learning_rate": 2.629431832215691e-06, + "loss": 0.76331007, + "num_input_tokens_seen": 74671960, + "step": 3461, + "time_per_iteration": 2.7029037475585938 + }, + { + "auxiliary_loss_clip": 0.01136983, + "auxiliary_loss_mlp": 0.01024681, + "balance_loss_clip": 0.97909558, + "balance_loss_mlp": 1.01661038, + "epoch": 0.41628088739253294, + "flos": 20010826650240.0, + "grad_norm": 1.6037023713761418, + "language_loss": 0.8716132, + "learning_rate": 2.628692395580151e-06, + "loss": 0.89322984, + "num_input_tokens_seen": 74692050, + "step": 3462, + "time_per_iteration": 2.665820598602295 + }, + { + "auxiliary_loss_clip": 0.01051442, + "auxiliary_loss_mlp": 0.01030648, + "balance_loss_clip": 0.85941648, + "balance_loss_mlp": 1.0221777, + "epoch": 0.416401130283172, + "flos": 29168801377920.0, + "grad_norm": 1.7454209118049882, + "language_loss": 0.79368865, + "learning_rate": 2.6279528635724747e-06, + "loss": 0.81450957, + "num_input_tokens_seen": 74712205, + "step": 3463, + "time_per_iteration": 2.8373796939849854 + }, + { + "auxiliary_loss_clip": 0.01163152, + "auxiliary_loss_mlp": 0.01035877, + "balance_loss_clip": 1.01709306, + "balance_loss_mlp": 1.02579141, + "epoch": 0.4165213731738111, + "flos": 16246813478400.0, + "grad_norm": 2.3286366494086765, + "language_loss": 0.7819382, + "learning_rate": 2.627213236304848e-06, + "loss": 0.80392849, + "num_input_tokens_seen": 74729005, + "step": 3464, + "time_per_iteration": 2.615957498550415 + }, + { + "auxiliary_loss_clip": 0.01167318, + "auxiliary_loss_mlp": 0.01030257, + "balance_loss_clip": 1.01943278, + "balance_loss_mlp": 1.02185833, + "epoch": 0.4166416160644502, + "flos": 33765438787200.0, + "grad_norm": 2.3930428433683826, + "language_loss": 0.70485121, + "learning_rate": 2.626473513889472e-06, + "loss": 0.72682697, + "num_input_tokens_seen": 74751385, + "step": 3465, + "time_per_iteration": 2.7680633068084717 + }, + { + "auxiliary_loss_clip": 0.01158423, + "auxiliary_loss_mlp": 0.01035743, + "balance_loss_clip": 1.01740742, + "balance_loss_mlp": 1.0273447, + "epoch": 0.41676185895508927, + "flos": 20917498775040.0, + "grad_norm": 2.028900497654701, + "language_loss": 0.82598305, + "learning_rate": 2.625733696438562e-06, + "loss": 0.84792477, + "num_input_tokens_seen": 74768890, + "step": 3466, + "time_per_iteration": 2.6274240016937256 + }, + { + "auxiliary_loss_clip": 0.01138527, + "auxiliary_loss_mlp": 0.01030579, + "balance_loss_clip": 0.98111165, + "balance_loss_mlp": 1.02219307, + "epoch": 0.4168821018457284, + "flos": 18406122549120.0, + "grad_norm": 1.6432274202061796, + "language_loss": 0.74872035, + "learning_rate": 2.6249937840643476e-06, + "loss": 0.77041137, + "num_input_tokens_seen": 74787195, + "step": 3467, + "time_per_iteration": 2.684450626373291 + }, + { + "auxiliary_loss_clip": 0.0119398, + "auxiliary_loss_mlp": 0.00510303, + "balance_loss_clip": 1.06395721, + "balance_loss_mlp": 0.0, + "epoch": 0.41700234473636744, + "flos": 18698399516160.0, + "grad_norm": 1.7849171562590143, + "language_loss": 0.6684922, + "learning_rate": 2.6242537768790733e-06, + "loss": 0.68553501, + "num_input_tokens_seen": 74806350, + "step": 3468, + "time_per_iteration": 2.6819558143615723 + }, + { + "auxiliary_loss_clip": 0.01166325, + "auxiliary_loss_mlp": 0.01031102, + "balance_loss_clip": 1.02119994, + "balance_loss_mlp": 1.0226202, + "epoch": 0.41712258762700655, + "flos": 31033283616000.0, + "grad_norm": 1.8692827709204696, + "language_loss": 0.68834996, + "learning_rate": 2.6235136749949975e-06, + "loss": 0.71032423, + "num_input_tokens_seen": 74829800, + "step": 3469, + "time_per_iteration": 2.7748281955718994 + }, + { + "auxiliary_loss_clip": 0.01187645, + "auxiliary_loss_mlp": 0.0102881, + "balance_loss_clip": 1.05762148, + "balance_loss_mlp": 1.02042377, + "epoch": 0.41724283051764566, + "flos": 35914763877120.0, + "grad_norm": 1.9448704027950556, + "language_loss": 0.60927105, + "learning_rate": 2.6227734785243924e-06, + "loss": 0.63143557, + "num_input_tokens_seen": 74849760, + "step": 3470, + "time_per_iteration": 2.7515745162963867 + }, + { + "auxiliary_loss_clip": 0.01062589, + "auxiliary_loss_mlp": 0.01023298, + "balance_loss_clip": 0.8623746, + "balance_loss_mlp": 1.01551926, + "epoch": 0.4173630734082847, + "flos": 25333649320320.0, + "grad_norm": 1.8067157386054074, + "language_loss": 0.79366392, + "learning_rate": 2.6220331875795466e-06, + "loss": 0.81452286, + "num_input_tokens_seen": 74869110, + "step": 3471, + "time_per_iteration": 2.816701889038086 + }, + { + "auxiliary_loss_clip": 0.01162156, + "auxiliary_loss_mlp": 0.01037077, + "balance_loss_clip": 1.01959324, + "balance_loss_mlp": 1.02872062, + "epoch": 0.4174833162989238, + "flos": 26685398868480.0, + "grad_norm": 1.664423106544628, + "language_loss": 0.75337011, + "learning_rate": 2.62129280227276e-06, + "loss": 0.77536249, + "num_input_tokens_seen": 74889110, + "step": 3472, + "time_per_iteration": 3.6062350273132324 + }, + { + "auxiliary_loss_clip": 0.01167949, + "auxiliary_loss_mlp": 0.01035025, + "balance_loss_clip": 1.01908553, + "balance_loss_mlp": 1.02692795, + "epoch": 0.41760355918956293, + "flos": 74739584010240.0, + "grad_norm": 2.325315170570273, + "language_loss": 0.6818794, + "learning_rate": 2.62055232271635e-06, + "loss": 0.70390916, + "num_input_tokens_seen": 74916260, + "step": 3473, + "time_per_iteration": 3.820650815963745 + }, + { + "auxiliary_loss_clip": 0.01107761, + "auxiliary_loss_mlp": 0.01028811, + "balance_loss_clip": 0.93892205, + "balance_loss_mlp": 1.02052593, + "epoch": 0.417723802080202, + "flos": 14317513148160.0, + "grad_norm": 2.18487360486596, + "language_loss": 0.87663805, + "learning_rate": 2.619811749022646e-06, + "loss": 0.8980037, + "num_input_tokens_seen": 74931570, + "step": 3474, + "time_per_iteration": 2.6726560592651367 + }, + { + "auxiliary_loss_clip": 0.01168576, + "auxiliary_loss_mlp": 0.01030977, + "balance_loss_clip": 1.0217551, + "balance_loss_mlp": 1.022686, + "epoch": 0.4178440449708411, + "flos": 14643797316480.0, + "grad_norm": 2.194538385025859, + "language_loss": 0.71361351, + "learning_rate": 2.6190710813039917e-06, + "loss": 0.73560905, + "num_input_tokens_seen": 74944695, + "step": 3475, + "time_per_iteration": 3.628016471862793 + }, + { + "auxiliary_loss_clip": 0.0109097, + "auxiliary_loss_mlp": 0.00511038, + "balance_loss_clip": 0.90099478, + "balance_loss_mlp": 0.0, + "epoch": 0.4179642878614802, + "flos": 21507296094720.0, + "grad_norm": 4.059002122982918, + "language_loss": 0.83986074, + "learning_rate": 2.618330319672747e-06, + "loss": 0.85588086, + "num_input_tokens_seen": 74964115, + "step": 3476, + "time_per_iteration": 2.774772882461548 + }, + { + "auxiliary_loss_clip": 0.01188403, + "auxiliary_loss_mlp": 0.0102993, + "balance_loss_clip": 1.0583216, + "balance_loss_mlp": 1.02222276, + "epoch": 0.41808453075211927, + "flos": 18441997257600.0, + "grad_norm": 2.101683908100602, + "language_loss": 0.91618454, + "learning_rate": 2.617589464241284e-06, + "loss": 0.93836784, + "num_input_tokens_seen": 74978515, + "step": 3477, + "time_per_iteration": 2.632897138595581 + }, + { + "auxiliary_loss_clip": 0.01121268, + "auxiliary_loss_mlp": 0.0102655, + "balance_loss_clip": 0.94250071, + "balance_loss_mlp": 1.01842618, + "epoch": 0.4182047736427584, + "flos": 20301020628480.0, + "grad_norm": 2.1945831987293456, + "language_loss": 0.74431866, + "learning_rate": 2.6168485151219914e-06, + "loss": 0.76579684, + "num_input_tokens_seen": 74998135, + "step": 3478, + "time_per_iteration": 2.6915533542633057 + }, + { + "auxiliary_loss_clip": 0.01165726, + "auxiliary_loss_mlp": 0.01032013, + "balance_loss_clip": 1.02098405, + "balance_loss_mlp": 1.02382302, + "epoch": 0.4183250165333975, + "flos": 18876623823360.0, + "grad_norm": 2.6513181771580996, + "language_loss": 0.71931678, + "learning_rate": 2.616107472427269e-06, + "loss": 0.74129415, + "num_input_tokens_seen": 75012830, + "step": 3479, + "time_per_iteration": 3.5362322330474854 + }, + { + "auxiliary_loss_clip": 0.01167719, + "auxiliary_loss_mlp": 0.01028321, + "balance_loss_clip": 1.01762891, + "balance_loss_mlp": 1.0200417, + "epoch": 0.41844525942403654, + "flos": 17740050698880.0, + "grad_norm": 2.29970230467891, + "language_loss": 0.75859404, + "learning_rate": 2.615366336269533e-06, + "loss": 0.78055441, + "num_input_tokens_seen": 75026495, + "step": 3480, + "time_per_iteration": 2.616602897644043 + }, + { + "auxiliary_loss_clip": 0.01192018, + "auxiliary_loss_mlp": 0.01032097, + "balance_loss_clip": 1.05906701, + "balance_loss_mlp": 1.02331734, + "epoch": 0.41856550231467565, + "flos": 18361377181440.0, + "grad_norm": 2.284566178090276, + "language_loss": 0.80082977, + "learning_rate": 2.6146251067612126e-06, + "loss": 0.82307088, + "num_input_tokens_seen": 75041970, + "step": 3481, + "time_per_iteration": 2.5699665546417236 + }, + { + "auxiliary_loss_clip": 0.01165451, + "auxiliary_loss_mlp": 0.01027231, + "balance_loss_clip": 1.02318335, + "balance_loss_mlp": 1.01912427, + "epoch": 0.41868574520531476, + "flos": 22781801445120.0, + "grad_norm": 1.7381967047421, + "language_loss": 0.82631654, + "learning_rate": 2.6138837840147525e-06, + "loss": 0.84824336, + "num_input_tokens_seen": 75061005, + "step": 3482, + "time_per_iteration": 2.6428964138031006 + }, + { + "auxiliary_loss_clip": 0.0111305, + "auxiliary_loss_mlp": 0.01030428, + "balance_loss_clip": 0.93962169, + "balance_loss_mlp": 1.02259612, + "epoch": 0.4188059880959538, + "flos": 13699167494400.0, + "grad_norm": 3.1367311015621433, + "language_loss": 0.76238525, + "learning_rate": 2.6131423681426103e-06, + "loss": 0.78381997, + "num_input_tokens_seen": 75076920, + "step": 3483, + "time_per_iteration": 2.6917150020599365 + }, + { + "auxiliary_loss_clip": 0.01187309, + "auxiliary_loss_mlp": 0.01029996, + "balance_loss_clip": 1.05899239, + "balance_loss_mlp": 1.02252793, + "epoch": 0.41892623098659293, + "flos": 37818281220480.0, + "grad_norm": 5.197514136139711, + "language_loss": 0.7269178, + "learning_rate": 2.6124008592572587e-06, + "loss": 0.74909091, + "num_input_tokens_seen": 75100905, + "step": 3484, + "time_per_iteration": 2.790553092956543 + }, + { + "auxiliary_loss_clip": 0.01191566, + "auxiliary_loss_mlp": 0.01028818, + "balance_loss_clip": 1.05798256, + "balance_loss_mlp": 1.02058661, + "epoch": 0.419046473877232, + "flos": 23258874908160.0, + "grad_norm": 2.185959937628124, + "language_loss": 0.81417072, + "learning_rate": 2.6116592574711835e-06, + "loss": 0.83637458, + "num_input_tokens_seen": 75119205, + "step": 3485, + "time_per_iteration": 2.6592352390289307 + }, + { + "auxiliary_loss_clip": 0.01191339, + "auxiliary_loss_mlp": 0.01029679, + "balance_loss_clip": 1.05909181, + "balance_loss_mlp": 1.02107811, + "epoch": 0.4191667167678711, + "flos": 20741034234240.0, + "grad_norm": 2.2217603571892353, + "language_loss": 0.84151971, + "learning_rate": 2.6109175628968853e-06, + "loss": 0.86372983, + "num_input_tokens_seen": 75138970, + "step": 3486, + "time_per_iteration": 2.6654789447784424 + }, + { + "auxiliary_loss_clip": 0.01153476, + "auxiliary_loss_mlp": 0.01025791, + "balance_loss_clip": 1.01484764, + "balance_loss_mlp": 1.01781034, + "epoch": 0.4192869596585102, + "flos": 23586416052480.0, + "grad_norm": 2.553054648391607, + "language_loss": 0.82885015, + "learning_rate": 2.610175775646878e-06, + "loss": 0.8506428, + "num_input_tokens_seen": 75157550, + "step": 3487, + "time_per_iteration": 2.6655983924865723 + }, + { + "auxiliary_loss_clip": 0.01135783, + "auxiliary_loss_mlp": 0.01022394, + "balance_loss_clip": 0.97765654, + "balance_loss_mlp": 1.0144192, + "epoch": 0.41940720254914926, + "flos": 25081269384960.0, + "grad_norm": 2.827491927517205, + "language_loss": 0.72990632, + "learning_rate": 2.6094338958336907e-06, + "loss": 0.75148809, + "num_input_tokens_seen": 75176220, + "step": 3488, + "time_per_iteration": 2.761880397796631 + }, + { + "auxiliary_loss_clip": 0.01142605, + "auxiliary_loss_mlp": 0.01024889, + "balance_loss_clip": 0.98423582, + "balance_loss_mlp": 1.01717639, + "epoch": 0.41952744543978837, + "flos": 15554132628480.0, + "grad_norm": 2.217802787424652, + "language_loss": 0.82170326, + "learning_rate": 2.608691923569867e-06, + "loss": 0.84337819, + "num_input_tokens_seen": 75193095, + "step": 3489, + "time_per_iteration": 2.6393208503723145 + }, + { + "auxiliary_loss_clip": 0.01167384, + "auxiliary_loss_mlp": 0.01027538, + "balance_loss_clip": 1.02091908, + "balance_loss_mlp": 1.01940227, + "epoch": 0.4196476883304275, + "flos": 24644775312000.0, + "grad_norm": 1.751548686901338, + "language_loss": 0.75945449, + "learning_rate": 2.6079498589679616e-06, + "loss": 0.78140372, + "num_input_tokens_seen": 75214185, + "step": 3490, + "time_per_iteration": 2.663773536682129 + }, + { + "auxiliary_loss_clip": 0.01064311, + "auxiliary_loss_mlp": 0.01031149, + "balance_loss_clip": 0.86112815, + "balance_loss_mlp": 1.02228618, + "epoch": 0.41976793122106654, + "flos": 24531333183360.0, + "grad_norm": 1.6556835404541275, + "language_loss": 0.75993478, + "learning_rate": 2.6072077021405465e-06, + "loss": 0.78088939, + "num_input_tokens_seen": 75233020, + "step": 3491, + "time_per_iteration": 2.7675397396087646 + }, + { + "auxiliary_loss_clip": 0.01123288, + "auxiliary_loss_mlp": 0.01035692, + "balance_loss_clip": 0.94041795, + "balance_loss_mlp": 1.02753818, + "epoch": 0.41988817411170565, + "flos": 21175301664000.0, + "grad_norm": 1.7082284945801836, + "language_loss": 0.69233119, + "learning_rate": 2.6064654532002054e-06, + "loss": 0.71392101, + "num_input_tokens_seen": 75252030, + "step": 3492, + "time_per_iteration": 2.7249157428741455 + }, + { + "auxiliary_loss_clip": 0.01188517, + "auxiliary_loss_mlp": 0.01033572, + "balance_loss_clip": 1.05826402, + "balance_loss_mlp": 1.02588296, + "epoch": 0.42000841700234476, + "flos": 31649402626560.0, + "grad_norm": 1.5149741450917347, + "language_loss": 0.76076102, + "learning_rate": 2.6057231122595375e-06, + "loss": 0.78298187, + "num_input_tokens_seen": 75273340, + "step": 3493, + "time_per_iteration": 2.6462488174438477 + }, + { + "auxiliary_loss_clip": 0.0114011, + "auxiliary_loss_mlp": 0.01028483, + "balance_loss_clip": 0.97877622, + "balance_loss_mlp": 1.02038884, + "epoch": 0.4201286598929838, + "flos": 21281525159040.0, + "grad_norm": 1.771339990576808, + "language_loss": 0.73019207, + "learning_rate": 2.604980679431154e-06, + "loss": 0.75187796, + "num_input_tokens_seen": 75291580, + "step": 3494, + "time_per_iteration": 2.676959276199341 + }, + { + "auxiliary_loss_clip": 0.01167083, + "auxiliary_loss_mlp": 0.01028796, + "balance_loss_clip": 1.01885939, + "balance_loss_mlp": 1.02084506, + "epoch": 0.4202489027836229, + "flos": 18546532813440.0, + "grad_norm": 2.318320293767807, + "language_loss": 0.74867105, + "learning_rate": 2.604238154827684e-06, + "loss": 0.77062988, + "num_input_tokens_seen": 75308205, + "step": 3495, + "time_per_iteration": 2.605729579925537 + }, + { + "auxiliary_loss_clip": 0.01167985, + "auxiliary_loss_mlp": 0.01029533, + "balance_loss_clip": 1.02072144, + "balance_loss_mlp": 1.0217011, + "epoch": 0.42036914567426203, + "flos": 19317643009920.0, + "grad_norm": 1.9028673173937625, + "language_loss": 0.72393328, + "learning_rate": 2.6034955385617656e-06, + "loss": 0.74590844, + "num_input_tokens_seen": 75326535, + "step": 3496, + "time_per_iteration": 2.63557767868042 + }, + { + "auxiliary_loss_clip": 0.01041168, + "auxiliary_loss_mlp": 0.01003116, + "balance_loss_clip": 0.92480016, + "balance_loss_mlp": 1.00172091, + "epoch": 0.4204893885649011, + "flos": 67842942935040.0, + "grad_norm": 0.7363296575934926, + "language_loss": 0.61689222, + "learning_rate": 2.6027528307460544e-06, + "loss": 0.63733506, + "num_input_tokens_seen": 75390540, + "step": 3497, + "time_per_iteration": 3.333606481552124 + }, + { + "auxiliary_loss_clip": 0.01187887, + "auxiliary_loss_mlp": 0.01032964, + "balance_loss_clip": 1.05703306, + "balance_loss_mlp": 1.0251441, + "epoch": 0.4206096314555402, + "flos": 21908777385600.0, + "grad_norm": 1.9035087581503622, + "language_loss": 0.86399728, + "learning_rate": 2.602010031493217e-06, + "loss": 0.88620579, + "num_input_tokens_seen": 75408770, + "step": 3498, + "time_per_iteration": 2.6405558586120605 + }, + { + "auxiliary_loss_clip": 0.011122, + "auxiliary_loss_mlp": 0.01030801, + "balance_loss_clip": 0.94187355, + "balance_loss_mlp": 1.02272499, + "epoch": 0.42072987434617926, + "flos": 29278185269760.0, + "grad_norm": 2.176103972624882, + "language_loss": 0.86605275, + "learning_rate": 2.6012671409159367e-06, + "loss": 0.88748276, + "num_input_tokens_seen": 75430105, + "step": 3499, + "time_per_iteration": 4.585332155227661 + }, + { + "auxiliary_loss_clip": 0.01134394, + "auxiliary_loss_mlp": 0.01027619, + "balance_loss_clip": 0.97831333, + "balance_loss_mlp": 1.01945364, + "epoch": 0.42085011723681837, + "flos": 27600726170880.0, + "grad_norm": 1.7305512485540548, + "language_loss": 0.81554216, + "learning_rate": 2.6005241591269097e-06, + "loss": 0.83716238, + "num_input_tokens_seen": 75449475, + "step": 3500, + "time_per_iteration": 2.775332450866699 + }, + { + "auxiliary_loss_clip": 0.01119401, + "auxiliary_loss_mlp": 0.01028933, + "balance_loss_clip": 0.94771075, + "balance_loss_mlp": 1.02132118, + "epoch": 0.4209703601274575, + "flos": 27818632028160.0, + "grad_norm": 2.044775819451704, + "language_loss": 0.79436058, + "learning_rate": 2.5997810862388454e-06, + "loss": 0.81584388, + "num_input_tokens_seen": 75469315, + "step": 3501, + "time_per_iteration": 3.6734578609466553 + }, + { + "auxiliary_loss_clip": 0.01141953, + "auxiliary_loss_mlp": 0.01033918, + "balance_loss_clip": 0.97816348, + "balance_loss_mlp": 1.02617288, + "epoch": 0.42109060301809653, + "flos": 27525529048320.0, + "grad_norm": 2.089021081867031, + "language_loss": 0.75352752, + "learning_rate": 2.599037922364467e-06, + "loss": 0.77528632, + "num_input_tokens_seen": 75488215, + "step": 3502, + "time_per_iteration": 2.7580018043518066 + }, + { + "auxiliary_loss_clip": 0.01108318, + "auxiliary_loss_mlp": 0.01032265, + "balance_loss_clip": 0.94232953, + "balance_loss_mlp": 1.02428949, + "epoch": 0.42121084590873564, + "flos": 29314275459840.0, + "grad_norm": 2.0468371636931755, + "language_loss": 0.75343621, + "learning_rate": 2.5982946676165112e-06, + "loss": 0.77484202, + "num_input_tokens_seen": 75507985, + "step": 3503, + "time_per_iteration": 2.761828660964966 + }, + { + "auxiliary_loss_clip": 0.01040155, + "auxiliary_loss_mlp": 0.01012522, + "balance_loss_clip": 0.93128413, + "balance_loss_mlp": 1.01104343, + "epoch": 0.42133108879937475, + "flos": 67398835178880.0, + "grad_norm": 0.7303773417133836, + "language_loss": 0.57619083, + "learning_rate": 2.5975513221077313e-06, + "loss": 0.5967176, + "num_input_tokens_seen": 75571955, + "step": 3504, + "time_per_iteration": 3.3413949012756348 + }, + { + "auxiliary_loss_clip": 0.01128833, + "auxiliary_loss_mlp": 0.01028107, + "balance_loss_clip": 0.97613096, + "balance_loss_mlp": 1.02040052, + "epoch": 0.4214513316900138, + "flos": 23106038538240.0, + "grad_norm": 2.229611712456164, + "language_loss": 0.8825357, + "learning_rate": 2.5968078859508897e-06, + "loss": 0.90410507, + "num_input_tokens_seen": 75589155, + "step": 3505, + "time_per_iteration": 3.5956273078918457 + }, + { + "auxiliary_loss_clip": 0.01163228, + "auxiliary_loss_mlp": 0.01032604, + "balance_loss_clip": 1.01774812, + "balance_loss_mlp": 1.02526414, + "epoch": 0.4215715745806529, + "flos": 15336190857600.0, + "grad_norm": 1.966167110473905, + "language_loss": 0.80302203, + "learning_rate": 2.5960643592587673e-06, + "loss": 0.82498038, + "num_input_tokens_seen": 75606565, + "step": 3506, + "time_per_iteration": 2.5927486419677734 + }, + { + "auxiliary_loss_clip": 0.01115368, + "auxiliary_loss_mlp": 0.01028844, + "balance_loss_clip": 0.94074327, + "balance_loss_mlp": 1.02092838, + "epoch": 0.42169181747129203, + "flos": 22127257860480.0, + "grad_norm": 2.0814010968040497, + "language_loss": 0.81421041, + "learning_rate": 2.5953207421441553e-06, + "loss": 0.83565253, + "num_input_tokens_seen": 75625165, + "step": 3507, + "time_per_iteration": 2.744422197341919 + }, + { + "auxiliary_loss_clip": 0.01118471, + "auxiliary_loss_mlp": 0.01030546, + "balance_loss_clip": 0.94350994, + "balance_loss_mlp": 1.02328074, + "epoch": 0.4218120603619311, + "flos": 22630724841600.0, + "grad_norm": 2.110951077261096, + "language_loss": 0.75298238, + "learning_rate": 2.5945770347198603e-06, + "loss": 0.77447253, + "num_input_tokens_seen": 75643320, + "step": 3508, + "time_per_iteration": 2.727648973464966 + }, + { + "auxiliary_loss_clip": 0.01132328, + "auxiliary_loss_mlp": 0.0102405, + "balance_loss_clip": 0.97499883, + "balance_loss_mlp": 1.01642704, + "epoch": 0.4219323032525702, + "flos": 19682818629120.0, + "grad_norm": 2.2165901523148723, + "language_loss": 0.821787, + "learning_rate": 2.593833237098701e-06, + "loss": 0.84335077, + "num_input_tokens_seen": 75660920, + "step": 3509, + "time_per_iteration": 2.67781925201416 + }, + { + "auxiliary_loss_clip": 0.01158799, + "auxiliary_loss_mlp": 0.01030729, + "balance_loss_clip": 1.01375091, + "balance_loss_mlp": 1.02243185, + "epoch": 0.4220525461432093, + "flos": 30190747224960.0, + "grad_norm": 1.8065314981877765, + "language_loss": 0.62699342, + "learning_rate": 2.593089349393512e-06, + "loss": 0.64888871, + "num_input_tokens_seen": 75681410, + "step": 3510, + "time_per_iteration": 2.774500608444214 + }, + { + "auxiliary_loss_clip": 0.0116191, + "auxiliary_loss_mlp": 0.01035259, + "balance_loss_clip": 1.02070725, + "balance_loss_mlp": 1.02747464, + "epoch": 0.42217278903384836, + "flos": 24315941278080.0, + "grad_norm": 2.2192122441327347, + "language_loss": 0.83496553, + "learning_rate": 2.592345371717141e-06, + "loss": 0.85693717, + "num_input_tokens_seen": 75700940, + "step": 3511, + "time_per_iteration": 2.6356847286224365 + }, + { + "auxiliary_loss_clip": 0.01161406, + "auxiliary_loss_mlp": 0.01031145, + "balance_loss_clip": 1.02053905, + "balance_loss_mlp": 1.02339613, + "epoch": 0.42229303192448747, + "flos": 17092474352640.0, + "grad_norm": 2.0722284267136253, + "language_loss": 0.72332615, + "learning_rate": 2.591601304182448e-06, + "loss": 0.74525166, + "num_input_tokens_seen": 75718910, + "step": 3512, + "time_per_iteration": 2.6710500717163086 + }, + { + "auxiliary_loss_clip": 0.0114037, + "auxiliary_loss_mlp": 0.01028639, + "balance_loss_clip": 0.98176181, + "balance_loss_mlp": 1.02117085, + "epoch": 0.4224132748151266, + "flos": 22784530878720.0, + "grad_norm": 2.033264279743976, + "language_loss": 0.7915073, + "learning_rate": 2.5908571469023067e-06, + "loss": 0.81319737, + "num_input_tokens_seen": 75738395, + "step": 3513, + "time_per_iteration": 2.721278190612793 + }, + { + "auxiliary_loss_clip": 0.01184644, + "auxiliary_loss_mlp": 0.01029703, + "balance_loss_clip": 1.05545771, + "balance_loss_mlp": 1.02170372, + "epoch": 0.42253351770576564, + "flos": 17819090576640.0, + "grad_norm": 2.352599402449345, + "language_loss": 0.75610793, + "learning_rate": 2.5901128999896067e-06, + "loss": 0.77825141, + "num_input_tokens_seen": 75753825, + "step": 3514, + "time_per_iteration": 2.6030092239379883 + }, + { + "auxiliary_loss_clip": 0.01160757, + "auxiliary_loss_mlp": 0.01025068, + "balance_loss_clip": 1.01840675, + "balance_loss_mlp": 1.01783788, + "epoch": 0.42265376059640475, + "flos": 28512390286080.0, + "grad_norm": 1.6790510550289692, + "language_loss": 0.68232065, + "learning_rate": 2.5893685635572487e-06, + "loss": 0.70417887, + "num_input_tokens_seen": 75774675, + "step": 3515, + "time_per_iteration": 2.8024065494537354 + }, + { + "auxiliary_loss_clip": 0.0113579, + "auxiliary_loss_mlp": 0.01026214, + "balance_loss_clip": 0.98055625, + "balance_loss_mlp": 1.01819181, + "epoch": 0.4227740034870438, + "flos": 16253349753600.0, + "grad_norm": 2.18106655592649, + "language_loss": 0.69099444, + "learning_rate": 2.5886241377181483e-06, + "loss": 0.71261448, + "num_input_tokens_seen": 75793545, + "step": 3516, + "time_per_iteration": 2.668318510055542 + }, + { + "auxiliary_loss_clip": 0.0116421, + "auxiliary_loss_mlp": 0.01031566, + "balance_loss_clip": 1.01890445, + "balance_loss_mlp": 1.02295315, + "epoch": 0.4228942463776829, + "flos": 25295691623040.0, + "grad_norm": 1.7955088931915777, + "language_loss": 0.81326568, + "learning_rate": 2.587879622585234e-06, + "loss": 0.83522344, + "num_input_tokens_seen": 75812145, + "step": 3517, + "time_per_iteration": 2.726020097732544 + }, + { + "auxiliary_loss_clip": 0.01160327, + "auxiliary_loss_mlp": 0.01023398, + "balance_loss_clip": 1.0182879, + "balance_loss_mlp": 1.01581061, + "epoch": 0.423014489268322, + "flos": 26395779507840.0, + "grad_norm": 1.986746494238622, + "language_loss": 0.76258337, + "learning_rate": 2.5871350182714486e-06, + "loss": 0.78442061, + "num_input_tokens_seen": 75833025, + "step": 3518, + "time_per_iteration": 2.7127485275268555 + }, + { + "auxiliary_loss_clip": 0.01182985, + "auxiliary_loss_mlp": 0.01030059, + "balance_loss_clip": 1.05438387, + "balance_loss_mlp": 1.02227449, + "epoch": 0.4231347321589611, + "flos": 17274002711040.0, + "grad_norm": 2.039905970632002, + "language_loss": 0.80844867, + "learning_rate": 2.586390324889748e-06, + "loss": 0.8305791, + "num_input_tokens_seen": 75848925, + "step": 3519, + "time_per_iteration": 2.5577311515808105 + }, + { + "auxiliary_loss_clip": 0.01163302, + "auxiliary_loss_mlp": 0.0102693, + "balance_loss_clip": 1.02135682, + "balance_loss_mlp": 1.01914597, + "epoch": 0.4232549750496002, + "flos": 22999635475200.0, + "grad_norm": 1.9092691662013552, + "language_loss": 0.67486024, + "learning_rate": 2.5856455425531003e-06, + "loss": 0.69676256, + "num_input_tokens_seen": 75870400, + "step": 3520, + "time_per_iteration": 2.641768455505371 + }, + { + "auxiliary_loss_clip": 0.0115803, + "auxiliary_loss_mlp": 0.01027409, + "balance_loss_clip": 1.01749516, + "balance_loss_mlp": 1.0203402, + "epoch": 0.4233752179402393, + "flos": 21248343970560.0, + "grad_norm": 1.7824696828906605, + "language_loss": 0.80873942, + "learning_rate": 2.5849006713744902e-06, + "loss": 0.83059382, + "num_input_tokens_seen": 75889195, + "step": 3521, + "time_per_iteration": 2.633281707763672 + }, + { + "auxiliary_loss_clip": 0.01136539, + "auxiliary_loss_mlp": 0.0102456, + "balance_loss_clip": 0.97914368, + "balance_loss_mlp": 1.01668072, + "epoch": 0.42349546083087836, + "flos": 20704297599360.0, + "grad_norm": 1.9782984558367127, + "language_loss": 0.72501606, + "learning_rate": 2.5841557114669135e-06, + "loss": 0.74662703, + "num_input_tokens_seen": 75906055, + "step": 3522, + "time_per_iteration": 2.673581600189209 + }, + { + "auxiliary_loss_clip": 0.01188498, + "auxiliary_loss_mlp": 0.0103087, + "balance_loss_clip": 1.05564356, + "balance_loss_mlp": 1.0226388, + "epoch": 0.42361570372151747, + "flos": 18585065128320.0, + "grad_norm": 2.5739935601428754, + "language_loss": 0.67208397, + "learning_rate": 2.58341066294338e-06, + "loss": 0.69427764, + "num_input_tokens_seen": 75922720, + "step": 3523, + "time_per_iteration": 2.580981492996216 + }, + { + "auxiliary_loss_clip": 0.01097385, + "auxiliary_loss_mlp": 0.00510614, + "balance_loss_clip": 0.902349, + "balance_loss_mlp": 0.0, + "epoch": 0.4237359466121566, + "flos": 20959478795520.0, + "grad_norm": 4.0958668881126234, + "language_loss": 0.8532058, + "learning_rate": 2.5826655259169124e-06, + "loss": 0.86928582, + "num_input_tokens_seen": 75941375, + "step": 3524, + "time_per_iteration": 3.6591553688049316 + }, + { + "auxiliary_loss_clip": 0.0118502, + "auxiliary_loss_mlp": 0.01027097, + "balance_loss_clip": 1.0574528, + "balance_loss_mlp": 1.01939034, + "epoch": 0.42385618950279563, + "flos": 18038181582720.0, + "grad_norm": 2.021094720695153, + "language_loss": 0.90692556, + "learning_rate": 2.5819203005005475e-06, + "loss": 0.92904675, + "num_input_tokens_seen": 75958710, + "step": 3525, + "time_per_iteration": 3.4908714294433594 + }, + { + "auxiliary_loss_clip": 0.0113487, + "auxiliary_loss_mlp": 0.0102812, + "balance_loss_clip": 0.98166609, + "balance_loss_mlp": 1.02020442, + "epoch": 0.42397643239343474, + "flos": 23769129559680.0, + "grad_norm": 1.699815594501102, + "language_loss": 0.7858699, + "learning_rate": 2.581174986807336e-06, + "loss": 0.80749977, + "num_input_tokens_seen": 75978945, + "step": 3526, + "time_per_iteration": 3.532228469848633 + }, + { + "auxiliary_loss_clip": 0.01152239, + "auxiliary_loss_mlp": 0.00509875, + "balance_loss_clip": 1.01406479, + "balance_loss_mlp": 0.0, + "epoch": 0.42409667528407385, + "flos": 16545088016640.0, + "grad_norm": 2.402072553524358, + "language_loss": 0.90948796, + "learning_rate": 2.580429584950341e-06, + "loss": 0.92610908, + "num_input_tokens_seen": 75994695, + "step": 3527, + "time_per_iteration": 2.6249022483825684 + }, + { + "auxiliary_loss_clip": 0.01120431, + "auxiliary_loss_mlp": 0.01026042, + "balance_loss_clip": 0.94110173, + "balance_loss_mlp": 1.01831174, + "epoch": 0.4242169181747129, + "flos": 16034186920320.0, + "grad_norm": 2.0288152160659942, + "language_loss": 0.66175997, + "learning_rate": 2.5796840950426397e-06, + "loss": 0.68322474, + "num_input_tokens_seen": 76011780, + "step": 3528, + "time_per_iteration": 2.736724853515625 + }, + { + "auxiliary_loss_clip": 0.01152443, + "auxiliary_loss_mlp": 0.01024377, + "balance_loss_clip": 1.01430368, + "balance_loss_mlp": 1.01677704, + "epoch": 0.424337161065352, + "flos": 20084012611200.0, + "grad_norm": 2.1265496146876046, + "language_loss": 0.66104203, + "learning_rate": 2.578938517197322e-06, + "loss": 0.68281025, + "num_input_tokens_seen": 76029875, + "step": 3529, + "time_per_iteration": 2.6491124629974365 + }, + { + "auxiliary_loss_clip": 0.011311, + "auxiliary_loss_mlp": 0.01024159, + "balance_loss_clip": 0.97846121, + "balance_loss_mlp": 1.01651788, + "epoch": 0.4244574039559911, + "flos": 23878369797120.0, + "grad_norm": 2.3072078590887397, + "language_loss": 0.62713575, + "learning_rate": 2.5781928515274916e-06, + "loss": 0.64868832, + "num_input_tokens_seen": 76048595, + "step": 3530, + "time_per_iteration": 2.7087719440460205 + }, + { + "auxiliary_loss_clip": 0.01166718, + "auxiliary_loss_mlp": 0.01028238, + "balance_loss_clip": 1.02045071, + "balance_loss_mlp": 1.0201019, + "epoch": 0.4245776468466302, + "flos": 17565920542080.0, + "grad_norm": 2.8597614614322255, + "language_loss": 0.67863047, + "learning_rate": 2.577447098146265e-06, + "loss": 0.70058, + "num_input_tokens_seen": 76065770, + "step": 3531, + "time_per_iteration": 3.6107587814331055 + }, + { + "auxiliary_loss_clip": 0.01116354, + "auxiliary_loss_mlp": 0.0103499, + "balance_loss_clip": 0.94194257, + "balance_loss_mlp": 1.02721786, + "epoch": 0.4246978897372693, + "flos": 27776256958080.0, + "grad_norm": 1.6586603373460707, + "language_loss": 0.79179585, + "learning_rate": 2.5767012571667724e-06, + "loss": 0.81330931, + "num_input_tokens_seen": 76085250, + "step": 3532, + "time_per_iteration": 2.736154794692993 + }, + { + "auxiliary_loss_clip": 0.01161832, + "auxiliary_loss_mlp": 0.01032328, + "balance_loss_clip": 1.01534295, + "balance_loss_mlp": 1.0247941, + "epoch": 0.42481813262790835, + "flos": 15596615439360.0, + "grad_norm": 1.732013446575136, + "language_loss": 0.68302554, + "learning_rate": 2.5759553287021587e-06, + "loss": 0.70496708, + "num_input_tokens_seen": 76103580, + "step": 3533, + "time_per_iteration": 2.6311962604522705 + }, + { + "auxiliary_loss_clip": 0.01139409, + "auxiliary_loss_mlp": 0.01027876, + "balance_loss_clip": 0.98382568, + "balance_loss_mlp": 1.02007341, + "epoch": 0.42493837551854746, + "flos": 23951088881280.0, + "grad_norm": 2.0029360359646193, + "language_loss": 0.77190524, + "learning_rate": 2.5752093128655786e-06, + "loss": 0.79357803, + "num_input_tokens_seen": 76121825, + "step": 3534, + "time_per_iteration": 2.7202415466308594 + }, + { + "auxiliary_loss_clip": 0.01130318, + "auxiliary_loss_mlp": 0.01025822, + "balance_loss_clip": 0.97628063, + "balance_loss_mlp": 1.0183953, + "epoch": 0.4250586184091866, + "flos": 20813466009600.0, + "grad_norm": 2.431150749245906, + "language_loss": 0.74074697, + "learning_rate": 2.574463209770204e-06, + "loss": 0.76230836, + "num_input_tokens_seen": 76141140, + "step": 3535, + "time_per_iteration": 2.727804183959961 + }, + { + "auxiliary_loss_clip": 0.01110634, + "auxiliary_loss_mlp": 0.01026047, + "balance_loss_clip": 0.93923616, + "balance_loss_mlp": 1.01863241, + "epoch": 0.42517886129982563, + "flos": 30371018607360.0, + "grad_norm": 1.725165261155979, + "language_loss": 0.79557669, + "learning_rate": 2.5737170195292165e-06, + "loss": 0.81694353, + "num_input_tokens_seen": 76164475, + "step": 3536, + "time_per_iteration": 2.7604053020477295 + }, + { + "auxiliary_loss_clip": 0.01117671, + "auxiliary_loss_mlp": 0.01029877, + "balance_loss_clip": 0.94169128, + "balance_loss_mlp": 1.0220629, + "epoch": 0.42529910419046474, + "flos": 20080636732800.0, + "grad_norm": 1.9827599157017672, + "language_loss": 0.7788164, + "learning_rate": 2.572970742255814e-06, + "loss": 0.8002919, + "num_input_tokens_seen": 76182965, + "step": 3537, + "time_per_iteration": 2.720912456512451 + }, + { + "auxiliary_loss_clip": 0.01160398, + "auxiliary_loss_mlp": 0.0102303, + "balance_loss_clip": 1.01907015, + "balance_loss_mlp": 1.01568079, + "epoch": 0.42541934708110385, + "flos": 22632448694400.0, + "grad_norm": 1.6607258640458231, + "language_loss": 0.81654942, + "learning_rate": 2.5722243780632046e-06, + "loss": 0.83838373, + "num_input_tokens_seen": 76201230, + "step": 3538, + "time_per_iteration": 2.6509034633636475 + }, + { + "auxiliary_loss_clip": 0.01027652, + "auxiliary_loss_mlp": 0.01006685, + "balance_loss_clip": 0.89314628, + "balance_loss_mlp": 1.00524902, + "epoch": 0.4255395899717429, + "flos": 66200676186240.0, + "grad_norm": 0.7560143749230387, + "language_loss": 0.60451466, + "learning_rate": 2.5714779270646125e-06, + "loss": 0.62485802, + "num_input_tokens_seen": 76262000, + "step": 3539, + "time_per_iteration": 3.2577085494995117 + }, + { + "auxiliary_loss_clip": 0.01141202, + "auxiliary_loss_mlp": 0.00509933, + "balance_loss_clip": 0.98052084, + "balance_loss_mlp": 0.0, + "epoch": 0.425659832862382, + "flos": 17931814433280.0, + "grad_norm": 3.0252198719861507, + "language_loss": 0.78109205, + "learning_rate": 2.5707313893732735e-06, + "loss": 0.79760343, + "num_input_tokens_seen": 76280540, + "step": 3540, + "time_per_iteration": 2.6528873443603516 + }, + { + "auxiliary_loss_clip": 0.01035469, + "auxiliary_loss_mlp": 0.01026207, + "balance_loss_clip": 0.81973064, + "balance_loss_mlp": 1.01839888, + "epoch": 0.4257800757530211, + "flos": 24022550989440.0, + "grad_norm": 2.1069453721716243, + "language_loss": 0.7691732, + "learning_rate": 2.5699847651024364e-06, + "loss": 0.78978992, + "num_input_tokens_seen": 76301180, + "step": 3541, + "time_per_iteration": 2.868044853210449 + }, + { + "auxiliary_loss_clip": 0.01159222, + "auxiliary_loss_mlp": 0.01027306, + "balance_loss_clip": 1.01958025, + "balance_loss_mlp": 1.01970601, + "epoch": 0.4259003186436602, + "flos": 23696015425920.0, + "grad_norm": 2.127091212244459, + "language_loss": 0.77271754, + "learning_rate": 2.5692380543653627e-06, + "loss": 0.79458278, + "num_input_tokens_seen": 76319335, + "step": 3542, + "time_per_iteration": 2.7921929359436035 + }, + { + "auxiliary_loss_clip": 0.01167077, + "auxiliary_loss_mlp": 0.00509957, + "balance_loss_clip": 1.01932943, + "balance_loss_mlp": 0.0, + "epoch": 0.4260205615342993, + "flos": 15259772672640.0, + "grad_norm": 2.4417270536667774, + "language_loss": 0.70058674, + "learning_rate": 2.5684912572753293e-06, + "loss": 0.71735716, + "num_input_tokens_seen": 76335010, + "step": 3543, + "time_per_iteration": 2.742690086364746 + }, + { + "auxiliary_loss_clip": 0.01181946, + "auxiliary_loss_mlp": 0.01027746, + "balance_loss_clip": 1.05529642, + "balance_loss_mlp": 1.02028704, + "epoch": 0.4261408044249384, + "flos": 30665306736000.0, + "grad_norm": 1.8649620939558074, + "language_loss": 0.84296596, + "learning_rate": 2.5677443739456245e-06, + "loss": 0.86506295, + "num_input_tokens_seen": 76356670, + "step": 3544, + "time_per_iteration": 2.721708297729492 + }, + { + "auxiliary_loss_clip": 0.01142594, + "auxiliary_loss_mlp": 0.01029812, + "balance_loss_clip": 0.98407435, + "balance_loss_mlp": 1.02208102, + "epoch": 0.42626104731557746, + "flos": 23257905240960.0, + "grad_norm": 2.665563364156944, + "language_loss": 0.79358226, + "learning_rate": 2.5669974044895495e-06, + "loss": 0.81530631, + "num_input_tokens_seen": 76373065, + "step": 3545, + "time_per_iteration": 2.6949379444122314 + }, + { + "auxiliary_loss_clip": 0.01121379, + "auxiliary_loss_mlp": 0.01027658, + "balance_loss_clip": 0.94051898, + "balance_loss_mlp": 1.01939654, + "epoch": 0.42638129020621657, + "flos": 25884770670720.0, + "grad_norm": 1.7463993110703337, + "language_loss": 0.79340655, + "learning_rate": 2.5662503490204187e-06, + "loss": 0.81489694, + "num_input_tokens_seen": 76393230, + "step": 3546, + "time_per_iteration": 2.769458293914795 + }, + { + "auxiliary_loss_clip": 0.01133114, + "auxiliary_loss_mlp": 0.01022641, + "balance_loss_clip": 0.97616613, + "balance_loss_mlp": 1.01480269, + "epoch": 0.4265015330968556, + "flos": 26502362138880.0, + "grad_norm": 1.7758364106323994, + "language_loss": 0.762514, + "learning_rate": 2.5655032076515603e-06, + "loss": 0.78407151, + "num_input_tokens_seen": 76412555, + "step": 3547, + "time_per_iteration": 2.7070837020874023 + }, + { + "auxiliary_loss_clip": 0.01139146, + "auxiliary_loss_mlp": 0.01029706, + "balance_loss_clip": 0.97947121, + "balance_loss_mlp": 1.02150416, + "epoch": 0.42662177598749473, + "flos": 24389522288640.0, + "grad_norm": 2.394215971284047, + "language_loss": 0.8221817, + "learning_rate": 2.5647559804963155e-06, + "loss": 0.84387028, + "num_input_tokens_seen": 76432485, + "step": 3548, + "time_per_iteration": 2.679839849472046 + }, + { + "auxiliary_loss_clip": 0.01098912, + "auxiliary_loss_mlp": 0.01029653, + "balance_loss_clip": 0.90287209, + "balance_loss_mlp": 1.02139211, + "epoch": 0.42674201887813384, + "flos": 23148629089920.0, + "grad_norm": 2.229089767287603, + "language_loss": 0.79197454, + "learning_rate": 2.5640086676680364e-06, + "loss": 0.8132602, + "num_input_tokens_seen": 76453980, + "step": 3549, + "time_per_iteration": 2.7089264392852783 + }, + { + "auxiliary_loss_clip": 0.01163371, + "auxiliary_loss_mlp": 0.01030845, + "balance_loss_clip": 1.01919365, + "balance_loss_mlp": 1.02261364, + "epoch": 0.4268622617687729, + "flos": 21689614552320.0, + "grad_norm": 2.195093599763408, + "language_loss": 0.80723226, + "learning_rate": 2.5632612692800923e-06, + "loss": 0.8291744, + "num_input_tokens_seen": 76473045, + "step": 3550, + "time_per_iteration": 2.6871800422668457 + }, + { + "auxiliary_loss_clip": 0.01115444, + "auxiliary_loss_mlp": 0.01033676, + "balance_loss_clip": 0.94114137, + "balance_loss_mlp": 1.02512836, + "epoch": 0.426982504659412, + "flos": 23440151871360.0, + "grad_norm": 2.1072431312391724, + "language_loss": 0.75158155, + "learning_rate": 2.5625137854458603e-06, + "loss": 0.77307278, + "num_input_tokens_seen": 76492060, + "step": 3551, + "time_per_iteration": 4.57518458366394 + }, + { + "auxiliary_loss_clip": 0.01137543, + "auxiliary_loss_mlp": 0.0102322, + "balance_loss_clip": 0.9766531, + "balance_loss_mlp": 1.01638651, + "epoch": 0.4271027475500511, + "flos": 18916556768640.0, + "grad_norm": 2.31331498963363, + "language_loss": 0.79803967, + "learning_rate": 2.561766216278735e-06, + "loss": 0.81964731, + "num_input_tokens_seen": 76509655, + "step": 3552, + "time_per_iteration": 3.6186745166778564 + }, + { + "auxiliary_loss_clip": 0.01092725, + "auxiliary_loss_mlp": 0.01034668, + "balance_loss_clip": 0.90627646, + "balance_loss_mlp": 1.02646005, + "epoch": 0.4272229904406902, + "flos": 26870554500480.0, + "grad_norm": 2.831463181447684, + "language_loss": 0.81595433, + "learning_rate": 2.561018561892121e-06, + "loss": 0.8372283, + "num_input_tokens_seen": 76528795, + "step": 3553, + "time_per_iteration": 2.7554867267608643 + }, + { + "auxiliary_loss_clip": 0.01136761, + "auxiliary_loss_mlp": 0.0102765, + "balance_loss_clip": 0.97690356, + "balance_loss_mlp": 1.02002072, + "epoch": 0.4273432333313293, + "flos": 23951376190080.0, + "grad_norm": 1.8629142844939985, + "language_loss": 0.76636732, + "learning_rate": 2.5602708223994363e-06, + "loss": 0.78801143, + "num_input_tokens_seen": 76550660, + "step": 3554, + "time_per_iteration": 2.7911906242370605 + }, + { + "auxiliary_loss_clip": 0.01114459, + "auxiliary_loss_mlp": 0.01030572, + "balance_loss_clip": 0.93597025, + "balance_loss_mlp": 1.02265644, + "epoch": 0.4274634762219684, + "flos": 29570354496000.0, + "grad_norm": 2.317358621182336, + "language_loss": 0.677935, + "learning_rate": 2.559522997914115e-06, + "loss": 0.69938529, + "num_input_tokens_seen": 76570240, + "step": 3555, + "time_per_iteration": 2.8168818950653076 + }, + { + "auxiliary_loss_clip": 0.01182762, + "auxiliary_loss_mlp": 0.010273, + "balance_loss_clip": 1.05672264, + "balance_loss_mlp": 1.01996255, + "epoch": 0.42758371911260745, + "flos": 21434146047360.0, + "grad_norm": 2.053248670730348, + "language_loss": 0.84829843, + "learning_rate": 2.558775088549599e-06, + "loss": 0.87039906, + "num_input_tokens_seen": 76589820, + "step": 3556, + "time_per_iteration": 2.6726722717285156 + }, + { + "auxiliary_loss_clip": 0.01167765, + "auxiliary_loss_mlp": 0.01025758, + "balance_loss_clip": 1.0180645, + "balance_loss_mlp": 1.01783073, + "epoch": 0.42770396200324656, + "flos": 14752822072320.0, + "grad_norm": 2.2716747386285077, + "language_loss": 0.66046154, + "learning_rate": 2.5580270944193467e-06, + "loss": 0.68239677, + "num_input_tokens_seen": 76606640, + "step": 3557, + "time_per_iteration": 3.600748300552368 + }, + { + "auxiliary_loss_clip": 0.01104001, + "auxiliary_loss_mlp": 0.0100126, + "balance_loss_clip": 1.04078412, + "balance_loss_mlp": 0.99982339, + "epoch": 0.4278242048938857, + "flos": 70654712601600.0, + "grad_norm": 0.7411868967451702, + "language_loss": 0.55516922, + "learning_rate": 2.557279015636827e-06, + "loss": 0.57622182, + "num_input_tokens_seen": 76667050, + "step": 3558, + "time_per_iteration": 3.1743316650390625 + }, + { + "auxiliary_loss_clip": 0.01079671, + "auxiliary_loss_mlp": 0.01003267, + "balance_loss_clip": 1.00163603, + "balance_loss_mlp": 1.0017705, + "epoch": 0.42794444778452473, + "flos": 69366165033600.0, + "grad_norm": 0.7684005951648423, + "language_loss": 0.61262023, + "learning_rate": 2.5565308523155245e-06, + "loss": 0.63344955, + "num_input_tokens_seen": 76726650, + "step": 3559, + "time_per_iteration": 3.194894552230835 + }, + { + "auxiliary_loss_clip": 0.01086207, + "auxiliary_loss_mlp": 0.01026294, + "balance_loss_clip": 0.90424418, + "balance_loss_mlp": 1.01844418, + "epoch": 0.42806469067516384, + "flos": 18215328481920.0, + "grad_norm": 2.561029301898324, + "language_loss": 0.81985164, + "learning_rate": 2.5557826045689336e-06, + "loss": 0.84097666, + "num_input_tokens_seen": 76742890, + "step": 3560, + "time_per_iteration": 2.7138118743896484 + }, + { + "auxiliary_loss_clip": 0.01042741, + "auxiliary_loss_mlp": 0.01005793, + "balance_loss_clip": 0.93363762, + "balance_loss_mlp": 1.0043745, + "epoch": 0.4281849335658029, + "flos": 54535814432640.0, + "grad_norm": 0.827897619043126, + "language_loss": 0.58849466, + "learning_rate": 2.5550342725105643e-06, + "loss": 0.60898006, + "num_input_tokens_seen": 76801055, + "step": 3561, + "time_per_iteration": 3.205026626586914 + }, + { + "auxiliary_loss_clip": 0.01164576, + "auxiliary_loss_mlp": 0.01036791, + "balance_loss_clip": 1.02289057, + "balance_loss_mlp": 1.02885747, + "epoch": 0.428305176456442, + "flos": 17274828723840.0, + "grad_norm": 1.9657614539977968, + "language_loss": 0.81121135, + "learning_rate": 2.554285856253937e-06, + "loss": 0.83322501, + "num_input_tokens_seen": 76819890, + "step": 3562, + "time_per_iteration": 2.6876885890960693 + }, + { + "auxiliary_loss_clip": 0.01138831, + "auxiliary_loss_mlp": 0.01024752, + "balance_loss_clip": 0.98220718, + "balance_loss_mlp": 1.01698565, + "epoch": 0.4284254193470811, + "flos": 26359509749760.0, + "grad_norm": 2.2252506541127337, + "language_loss": 0.77253717, + "learning_rate": 2.5535373559125855e-06, + "loss": 0.794173, + "num_input_tokens_seen": 76840255, + "step": 3563, + "time_per_iteration": 2.787168025970459 + }, + { + "auxiliary_loss_clip": 0.01044599, + "auxiliary_loss_mlp": 0.01025223, + "balance_loss_clip": 0.82300758, + "balance_loss_mlp": 1.01682472, + "epoch": 0.42854566223772017, + "flos": 29714248379520.0, + "grad_norm": 1.8323811836556447, + "language_loss": 0.81744027, + "learning_rate": 2.552788771600057e-06, + "loss": 0.83813852, + "num_input_tokens_seen": 76860565, + "step": 3564, + "time_per_iteration": 2.9129767417907715 + }, + { + "auxiliary_loss_clip": 0.01119872, + "auxiliary_loss_mlp": 0.01025927, + "balance_loss_clip": 0.94442064, + "balance_loss_mlp": 1.01821399, + "epoch": 0.4286659051283593, + "flos": 22018161277440.0, + "grad_norm": 1.994953259550139, + "language_loss": 0.81742883, + "learning_rate": 2.5520401034299118e-06, + "loss": 0.83888686, + "num_input_tokens_seen": 76878325, + "step": 3565, + "time_per_iteration": 2.711275100708008 + }, + { + "auxiliary_loss_clip": 0.01164951, + "auxiliary_loss_mlp": 0.01026003, + "balance_loss_clip": 1.0186615, + "balance_loss_mlp": 1.01752079, + "epoch": 0.4287861480189984, + "flos": 13334422838400.0, + "grad_norm": 2.0871180513881638, + "language_loss": 0.87750745, + "learning_rate": 2.551291351515722e-06, + "loss": 0.89941698, + "num_input_tokens_seen": 76895340, + "step": 3566, + "time_per_iteration": 2.651643991470337 + }, + { + "auxiliary_loss_clip": 0.01111436, + "auxiliary_loss_mlp": 0.00510431, + "balance_loss_clip": 0.93758929, + "balance_loss_mlp": 0.0, + "epoch": 0.42890639090963745, + "flos": 26651535321600.0, + "grad_norm": 1.6747374198248914, + "language_loss": 0.85879493, + "learning_rate": 2.5505425159710726e-06, + "loss": 0.87501359, + "num_input_tokens_seen": 76915150, + "step": 3567, + "time_per_iteration": 2.75925612449646 + }, + { + "auxiliary_loss_clip": 0.01145744, + "auxiliary_loss_mlp": 0.0051011, + "balance_loss_clip": 0.9784537, + "balance_loss_mlp": 0.0, + "epoch": 0.42902663380027656, + "flos": 24055768091520.0, + "grad_norm": 1.8017282644173074, + "language_loss": 0.83110607, + "learning_rate": 2.549793596909561e-06, + "loss": 0.84766459, + "num_input_tokens_seen": 76933770, + "step": 3568, + "time_per_iteration": 2.716304063796997 + }, + { + "auxiliary_loss_clip": 0.01135514, + "auxiliary_loss_mlp": 0.01030362, + "balance_loss_clip": 0.9805671, + "balance_loss_mlp": 1.02256584, + "epoch": 0.42914687669091567, + "flos": 15632561975040.0, + "grad_norm": 2.190647434085146, + "language_loss": 0.66276407, + "learning_rate": 2.5490445944447976e-06, + "loss": 0.68442279, + "num_input_tokens_seen": 76952265, + "step": 3569, + "time_per_iteration": 2.712155342102051 + }, + { + "auxiliary_loss_clip": 0.01163283, + "auxiliary_loss_mlp": 0.01032327, + "balance_loss_clip": 1.01958275, + "balance_loss_mlp": 1.02394664, + "epoch": 0.4292671195815547, + "flos": 31467802440960.0, + "grad_norm": 2.6301124712163637, + "language_loss": 0.65509135, + "learning_rate": 2.548295508690406e-06, + "loss": 0.67704743, + "num_input_tokens_seen": 76973560, + "step": 3570, + "time_per_iteration": 2.730593204498291 + }, + { + "auxiliary_loss_clip": 0.01165124, + "auxiliary_loss_mlp": 0.01030774, + "balance_loss_clip": 1.01748466, + "balance_loss_mlp": 1.02224469, + "epoch": 0.42938736247219383, + "flos": 30257756046720.0, + "grad_norm": 3.487417029511331, + "language_loss": 0.76577735, + "learning_rate": 2.5475463397600217e-06, + "loss": 0.7877363, + "num_input_tokens_seen": 76993640, + "step": 3571, + "time_per_iteration": 2.683683156967163 + }, + { + "auxiliary_loss_clip": 0.01190175, + "auxiliary_loss_mlp": 0.0103223, + "balance_loss_clip": 1.05882537, + "balance_loss_mlp": 1.02435017, + "epoch": 0.42950760536283294, + "flos": 29349683291520.0, + "grad_norm": 2.1521017607173545, + "language_loss": 0.77172828, + "learning_rate": 2.546797087767293e-06, + "loss": 0.79395229, + "num_input_tokens_seen": 77013765, + "step": 3572, + "time_per_iteration": 2.6681385040283203 + }, + { + "auxiliary_loss_clip": 0.0109136, + "auxiliary_loss_mlp": 0.01029927, + "balance_loss_clip": 0.9057771, + "balance_loss_mlp": 1.02212465, + "epoch": 0.429627848253472, + "flos": 26869943969280.0, + "grad_norm": 2.242211705630539, + "language_loss": 0.87122881, + "learning_rate": 2.546047752825881e-06, + "loss": 0.89244169, + "num_input_tokens_seen": 77034370, + "step": 3573, + "time_per_iteration": 2.762572765350342 + }, + { + "auxiliary_loss_clip": 0.01095642, + "auxiliary_loss_mlp": 0.01031967, + "balance_loss_clip": 0.90246189, + "balance_loss_mlp": 1.02415299, + "epoch": 0.4297480911441111, + "flos": 13881270470400.0, + "grad_norm": 2.2666594181390898, + "language_loss": 0.93068773, + "learning_rate": 2.5452983350494595e-06, + "loss": 0.95196378, + "num_input_tokens_seen": 77049925, + "step": 3574, + "time_per_iteration": 2.666644811630249 + }, + { + "auxiliary_loss_clip": 0.01164887, + "auxiliary_loss_mlp": 0.00509844, + "balance_loss_clip": 1.01953769, + "balance_loss_mlp": 0.0, + "epoch": 0.4298683340347502, + "flos": 20741141975040.0, + "grad_norm": 2.850428993982625, + "language_loss": 0.656762, + "learning_rate": 2.544548834551713e-06, + "loss": 0.67350936, + "num_input_tokens_seen": 77068930, + "step": 3575, + "time_per_iteration": 2.6463234424591064 + }, + { + "auxiliary_loss_clip": 0.01115743, + "auxiliary_loss_mlp": 0.00510556, + "balance_loss_clip": 0.94432884, + "balance_loss_mlp": 0.0, + "epoch": 0.4299885769253893, + "flos": 20882126856960.0, + "grad_norm": 2.3475389672347364, + "language_loss": 0.94782627, + "learning_rate": 2.5437992514463424e-06, + "loss": 0.96408927, + "num_input_tokens_seen": 77082255, + "step": 3576, + "time_per_iteration": 2.6859147548675537 + }, + { + "auxiliary_loss_clip": 0.01158338, + "auxiliary_loss_mlp": 0.01027276, + "balance_loss_clip": 1.01609743, + "balance_loss_mlp": 1.01862133, + "epoch": 0.4301088198160284, + "flos": 25484618183040.0, + "grad_norm": 1.7437632388566155, + "language_loss": 0.88194907, + "learning_rate": 2.5430495858470565e-06, + "loss": 0.9038052, + "num_input_tokens_seen": 77101725, + "step": 3577, + "time_per_iteration": 4.531313896179199 + }, + { + "auxiliary_loss_clip": 0.01163523, + "auxiliary_loss_mlp": 0.0102501, + "balance_loss_clip": 1.02171969, + "balance_loss_mlp": 1.01656389, + "epoch": 0.43022906270666744, + "flos": 18259427404800.0, + "grad_norm": 3.078164830201533, + "language_loss": 0.77594459, + "learning_rate": 2.54229983786758e-06, + "loss": 0.79782999, + "num_input_tokens_seen": 77119670, + "step": 3578, + "time_per_iteration": 2.6264126300811768 + }, + { + "auxiliary_loss_clip": 0.01137294, + "auxiliary_loss_mlp": 0.01029155, + "balance_loss_clip": 0.97651613, + "balance_loss_mlp": 1.02090013, + "epoch": 0.43034930559730655, + "flos": 23399536567680.0, + "grad_norm": 2.1501192166503555, + "language_loss": 0.85306311, + "learning_rate": 2.541550007621651e-06, + "loss": 0.87472761, + "num_input_tokens_seen": 77138160, + "step": 3579, + "time_per_iteration": 3.605562448501587 + }, + { + "auxiliary_loss_clip": 0.01162161, + "auxiliary_loss_mlp": 0.01027004, + "balance_loss_clip": 1.01932287, + "balance_loss_mlp": 1.01979744, + "epoch": 0.43046954848794566, + "flos": 28184382264960.0, + "grad_norm": 2.43444631561838, + "language_loss": 0.80210245, + "learning_rate": 2.5408000952230156e-06, + "loss": 0.82399416, + "num_input_tokens_seen": 77156950, + "step": 3580, + "time_per_iteration": 2.7226719856262207 + }, + { + "auxiliary_loss_clip": 0.01126707, + "auxiliary_loss_mlp": 0.01029974, + "balance_loss_clip": 0.94225931, + "balance_loss_mlp": 1.02231526, + "epoch": 0.4305897913785847, + "flos": 28580476515840.0, + "grad_norm": 2.2179231745879315, + "language_loss": 0.90611994, + "learning_rate": 2.5400501007854357e-06, + "loss": 0.92768669, + "num_input_tokens_seen": 77176395, + "step": 3581, + "time_per_iteration": 2.7524306774139404 + }, + { + "auxiliary_loss_clip": 0.01092558, + "auxiliary_loss_mlp": 0.01027433, + "balance_loss_clip": 0.90002787, + "balance_loss_mlp": 1.01929688, + "epoch": 0.43071003426922383, + "flos": 20448721353600.0, + "grad_norm": 2.206037587802157, + "language_loss": 0.75292403, + "learning_rate": 2.539300024422685e-06, + "loss": 0.77412397, + "num_input_tokens_seen": 77194340, + "step": 3582, + "time_per_iteration": 3.6318962574005127 + }, + { + "auxiliary_loss_clip": 0.0103938, + "auxiliary_loss_mlp": 0.01003802, + "balance_loss_clip": 0.92555928, + "balance_loss_mlp": 1.00243115, + "epoch": 0.43083027715986294, + "flos": 51997969883520.0, + "grad_norm": 0.7866620905667095, + "language_loss": 0.60884732, + "learning_rate": 2.538549866248549e-06, + "loss": 0.62927914, + "num_input_tokens_seen": 77249320, + "step": 3583, + "time_per_iteration": 3.099241256713867 + }, + { + "auxiliary_loss_clip": 0.011632, + "auxiliary_loss_mlp": 0.01030074, + "balance_loss_clip": 1.01879668, + "balance_loss_mlp": 1.02161336, + "epoch": 0.430950520050502, + "flos": 16690885320960.0, + "grad_norm": 1.8493778165032286, + "language_loss": 0.80952752, + "learning_rate": 2.5377996263768274e-06, + "loss": 0.83146024, + "num_input_tokens_seen": 77267400, + "step": 3584, + "time_per_iteration": 2.6247851848602295 + }, + { + "auxiliary_loss_clip": 0.01159995, + "auxiliary_loss_mlp": 0.01029778, + "balance_loss_clip": 1.01722598, + "balance_loss_mlp": 1.02145696, + "epoch": 0.4310707629411411, + "flos": 24608433726720.0, + "grad_norm": 1.7601838941525412, + "language_loss": 0.68392956, + "learning_rate": 2.5370493049213293e-06, + "loss": 0.70582736, + "num_input_tokens_seen": 77287045, + "step": 3585, + "time_per_iteration": 2.713787317276001 + }, + { + "auxiliary_loss_clip": 0.01015671, + "auxiliary_loss_mlp": 0.010276, + "balance_loss_clip": 0.78669107, + "balance_loss_mlp": 1.01894486, + "epoch": 0.4311910058317802, + "flos": 26432983019520.0, + "grad_norm": 1.9412353053965854, + "language_loss": 0.80337477, + "learning_rate": 2.536298901995878e-06, + "loss": 0.82380748, + "num_input_tokens_seen": 77306255, + "step": 3586, + "time_per_iteration": 2.936169385910034 + }, + { + "auxiliary_loss_clip": 0.01140921, + "auxiliary_loss_mlp": 0.01025886, + "balance_loss_clip": 0.9799161, + "balance_loss_mlp": 1.01745248, + "epoch": 0.43131124872241927, + "flos": 25155891889920.0, + "grad_norm": 1.6474092828217708, + "language_loss": 0.79666173, + "learning_rate": 2.535548417714311e-06, + "loss": 0.81832981, + "num_input_tokens_seen": 77325555, + "step": 3587, + "time_per_iteration": 3.1108970642089844 + }, + { + "auxiliary_loss_clip": 0.01166258, + "auxiliary_loss_mlp": 0.01024505, + "balance_loss_clip": 1.01608634, + "balance_loss_mlp": 1.01659584, + "epoch": 0.4314314916130584, + "flos": 21614812479360.0, + "grad_norm": 1.5406170527628316, + "language_loss": 0.87008172, + "learning_rate": 2.534797852190474e-06, + "loss": 0.89198935, + "num_input_tokens_seen": 77345735, + "step": 3588, + "time_per_iteration": 2.674320697784424 + }, + { + "auxiliary_loss_clip": 0.01157033, + "auxiliary_loss_mlp": 0.01031846, + "balance_loss_clip": 1.01510477, + "balance_loss_mlp": 1.023543, + "epoch": 0.4315517345036975, + "flos": 19275016544640.0, + "grad_norm": 2.291581246643419, + "language_loss": 0.81811881, + "learning_rate": 2.5340472055382283e-06, + "loss": 0.84000766, + "num_input_tokens_seen": 77361765, + "step": 3589, + "time_per_iteration": 2.6050217151641846 + }, + { + "auxiliary_loss_clip": 0.0111639, + "auxiliary_loss_mlp": 0.01032626, + "balance_loss_clip": 0.940081, + "balance_loss_mlp": 1.02451372, + "epoch": 0.43167197739433655, + "flos": 24273853516800.0, + "grad_norm": 2.1723945562552003, + "language_loss": 0.81081736, + "learning_rate": 2.5332964778714468e-06, + "loss": 0.83230758, + "num_input_tokens_seen": 77378950, + "step": 3590, + "time_per_iteration": 2.740805149078369 + }, + { + "auxiliary_loss_clip": 0.01114034, + "auxiliary_loss_mlp": 0.01028892, + "balance_loss_clip": 0.94295204, + "balance_loss_mlp": 1.0214951, + "epoch": 0.43179222028497566, + "flos": 16867816738560.0, + "grad_norm": 1.6298596316347027, + "language_loss": 0.66359854, + "learning_rate": 2.5325456693040123e-06, + "loss": 0.68502784, + "num_input_tokens_seen": 77396145, + "step": 3591, + "time_per_iteration": 2.6823370456695557 + }, + { + "auxiliary_loss_clip": 0.01167224, + "auxiliary_loss_mlp": 0.010294, + "balance_loss_clip": 1.01612163, + "balance_loss_mlp": 1.02091241, + "epoch": 0.43191246317561477, + "flos": 17639214243840.0, + "grad_norm": 2.398793869856921, + "language_loss": 0.75046718, + "learning_rate": 2.531794779949824e-06, + "loss": 0.7724334, + "num_input_tokens_seen": 77414045, + "step": 3592, + "time_per_iteration": 2.6428160667419434 + }, + { + "auxiliary_loss_clip": 0.01106587, + "auxiliary_loss_mlp": 0.01027232, + "balance_loss_clip": 0.93914711, + "balance_loss_mlp": 1.0197401, + "epoch": 0.4320327060662538, + "flos": 23878800760320.0, + "grad_norm": 1.7845817315458539, + "language_loss": 0.88109559, + "learning_rate": 2.5310438099227903e-06, + "loss": 0.90243375, + "num_input_tokens_seen": 77431310, + "step": 3593, + "time_per_iteration": 2.767960786819458 + }, + { + "auxiliary_loss_clip": 0.01079491, + "auxiliary_loss_mlp": 0.01007205, + "balance_loss_clip": 1.00019109, + "balance_loss_mlp": 1.00586939, + "epoch": 0.43215294895689293, + "flos": 66394917959040.0, + "grad_norm": 0.8190229953188961, + "language_loss": 0.53400445, + "learning_rate": 2.530292759336833e-06, + "loss": 0.55487144, + "num_input_tokens_seen": 77492045, + "step": 3594, + "time_per_iteration": 3.2699084281921387 + }, + { + "auxiliary_loss_clip": 0.01132866, + "auxiliary_loss_mlp": 0.01030433, + "balance_loss_clip": 0.97772712, + "balance_loss_mlp": 1.02121258, + "epoch": 0.432273191847532, + "flos": 20594267262720.0, + "grad_norm": 1.983793076931941, + "language_loss": 0.6990546, + "learning_rate": 2.5295416283058855e-06, + "loss": 0.72068763, + "num_input_tokens_seen": 77510910, + "step": 3595, + "time_per_iteration": 2.641561269760132 + }, + { + "auxiliary_loss_clip": 0.01135353, + "auxiliary_loss_mlp": 0.00509947, + "balance_loss_clip": 0.97948563, + "balance_loss_mlp": 0.0, + "epoch": 0.4323934347381711, + "flos": 19282127437440.0, + "grad_norm": 1.6262149072533936, + "language_loss": 0.65940464, + "learning_rate": 2.5287904169438943e-06, + "loss": 0.6758576, + "num_input_tokens_seen": 77530115, + "step": 3596, + "time_per_iteration": 2.6831321716308594 + }, + { + "auxiliary_loss_clip": 0.01051781, + "auxiliary_loss_mlp": 0.01025473, + "balance_loss_clip": 0.82682198, + "balance_loss_mlp": 1.01747406, + "epoch": 0.4325136776288102, + "flos": 21726315273600.0, + "grad_norm": 2.7540545899356226, + "language_loss": 0.63895547, + "learning_rate": 2.528039125364817e-06, + "loss": 0.65972805, + "num_input_tokens_seen": 77548920, + "step": 3597, + "time_per_iteration": 2.8398404121398926 + }, + { + "auxiliary_loss_clip": 0.01118128, + "auxiliary_loss_mlp": 0.01028092, + "balance_loss_clip": 0.94277889, + "balance_loss_mlp": 1.01967561, + "epoch": 0.43263392051944927, + "flos": 22340746344960.0, + "grad_norm": 2.1930708064176474, + "language_loss": 0.75684464, + "learning_rate": 2.5272877536826246e-06, + "loss": 0.77830684, + "num_input_tokens_seen": 77567715, + "step": 3598, + "time_per_iteration": 2.7904012203216553 + }, + { + "auxiliary_loss_clip": 0.01094884, + "auxiliary_loss_mlp": 0.0103329, + "balance_loss_clip": 0.89985394, + "balance_loss_mlp": 1.02456975, + "epoch": 0.4327541634100884, + "flos": 29168406328320.0, + "grad_norm": 2.6645662532784002, + "language_loss": 0.70439732, + "learning_rate": 2.5265363020112986e-06, + "loss": 0.7256791, + "num_input_tokens_seen": 77588035, + "step": 3599, + "time_per_iteration": 2.8615381717681885 + }, + { + "auxiliary_loss_clip": 0.01161784, + "auxiliary_loss_mlp": 0.01028159, + "balance_loss_clip": 1.01816618, + "balance_loss_mlp": 1.02035117, + "epoch": 0.4328744063007275, + "flos": 26067448264320.0, + "grad_norm": 1.896766325051288, + "language_loss": 0.84297007, + "learning_rate": 2.5257847704648344e-06, + "loss": 0.86486948, + "num_input_tokens_seen": 77609265, + "step": 3600, + "time_per_iteration": 2.6737706661224365 + }, + { + "auxiliary_loss_clip": 0.01184724, + "auxiliary_loss_mlp": 0.0102616, + "balance_loss_clip": 1.05708778, + "balance_loss_mlp": 1.01877785, + "epoch": 0.43299464919136654, + "flos": 16581357774720.0, + "grad_norm": 1.7702363783596997, + "language_loss": 0.75570452, + "learning_rate": 2.525033159157239e-06, + "loss": 0.77781332, + "num_input_tokens_seen": 77625580, + "step": 3601, + "time_per_iteration": 2.5934431552886963 + }, + { + "auxiliary_loss_clip": 0.01162717, + "auxiliary_loss_mlp": 0.01036078, + "balance_loss_clip": 1.01854467, + "balance_loss_mlp": 1.02711368, + "epoch": 0.43311489208200565, + "flos": 16107265140480.0, + "grad_norm": 1.905737128637197, + "language_loss": 0.77069068, + "learning_rate": 2.52428146820253e-06, + "loss": 0.79267859, + "num_input_tokens_seen": 77643835, + "step": 3602, + "time_per_iteration": 3.7443835735321045 + }, + { + "auxiliary_loss_clip": 0.01117493, + "auxiliary_loss_mlp": 0.01028059, + "balance_loss_clip": 0.94238871, + "balance_loss_mlp": 1.0194701, + "epoch": 0.43323513497264476, + "flos": 22930220442240.0, + "grad_norm": 2.0287693744676116, + "language_loss": 0.81463307, + "learning_rate": 2.52352969771474e-06, + "loss": 0.8360886, + "num_input_tokens_seen": 77663060, + "step": 3603, + "time_per_iteration": 3.744835376739502 + }, + { + "auxiliary_loss_clip": 0.01140532, + "auxiliary_loss_mlp": 0.01026488, + "balance_loss_clip": 0.9796474, + "balance_loss_mlp": 1.01898372, + "epoch": 0.4333553778632838, + "flos": 25299031587840.0, + "grad_norm": 1.8367459060369815, + "language_loss": 0.88540155, + "learning_rate": 2.5227778478079106e-06, + "loss": 0.90707171, + "num_input_tokens_seen": 77682470, + "step": 3604, + "time_per_iteration": 3.637761354446411 + }, + { + "auxiliary_loss_clip": 0.01156131, + "auxiliary_loss_mlp": 0.01030691, + "balance_loss_clip": 1.01486707, + "balance_loss_mlp": 1.02291203, + "epoch": 0.43347562075392293, + "flos": 19387165783680.0, + "grad_norm": 1.7099637640664849, + "language_loss": 0.76962137, + "learning_rate": 2.522025918596098e-06, + "loss": 0.7914896, + "num_input_tokens_seen": 77700770, + "step": 3605, + "time_per_iteration": 2.626309633255005 + }, + { + "auxiliary_loss_clip": 0.01165858, + "auxiliary_loss_mlp": 0.01022569, + "balance_loss_clip": 1.01928055, + "balance_loss_mlp": 1.01473665, + "epoch": 0.43359586364456204, + "flos": 26325969425280.0, + "grad_norm": 2.0622652440020888, + "language_loss": 0.65516043, + "learning_rate": 2.521273910193368e-06, + "loss": 0.67704469, + "num_input_tokens_seen": 77723950, + "step": 3606, + "time_per_iteration": 2.6911251544952393 + }, + { + "auxiliary_loss_clip": 0.01169829, + "auxiliary_loss_mlp": 0.01031707, + "balance_loss_clip": 1.01965475, + "balance_loss_mlp": 1.02344024, + "epoch": 0.4337161065352011, + "flos": 15989261984640.0, + "grad_norm": 2.8352055264719858, + "language_loss": 0.8690182, + "learning_rate": 2.5205218227138006e-06, + "loss": 0.89103359, + "num_input_tokens_seen": 77736905, + "step": 3607, + "time_per_iteration": 2.579920530319214 + }, + { + "auxiliary_loss_clip": 0.01188349, + "auxiliary_loss_mlp": 0.01027863, + "balance_loss_clip": 1.0572319, + "balance_loss_mlp": 1.0197984, + "epoch": 0.4338363494258402, + "flos": 20224710184320.0, + "grad_norm": 2.0476349641188376, + "language_loss": 0.79302722, + "learning_rate": 2.519769656271486e-06, + "loss": 0.81518936, + "num_input_tokens_seen": 77754325, + "step": 3608, + "time_per_iteration": 2.6074206829071045 + }, + { + "auxiliary_loss_clip": 0.0108584, + "auxiliary_loss_mlp": 0.01027527, + "balance_loss_clip": 0.90532196, + "balance_loss_mlp": 1.01971269, + "epoch": 0.43395659231647926, + "flos": 20083904870400.0, + "grad_norm": 1.8203002943624986, + "language_loss": 0.67590809, + "learning_rate": 2.5190174109805285e-06, + "loss": 0.69704175, + "num_input_tokens_seen": 77774150, + "step": 3609, + "time_per_iteration": 3.69950008392334 + }, + { + "auxiliary_loss_clip": 0.01128795, + "auxiliary_loss_mlp": 0.01030677, + "balance_loss_clip": 0.97798264, + "balance_loss_mlp": 1.0222013, + "epoch": 0.43407683520711837, + "flos": 19901801894400.0, + "grad_norm": 1.8035944985112697, + "language_loss": 0.63833463, + "learning_rate": 2.518265086955042e-06, + "loss": 0.65992928, + "num_input_tokens_seen": 77791870, + "step": 3610, + "time_per_iteration": 2.657318115234375 + }, + { + "auxiliary_loss_clip": 0.01186399, + "auxiliary_loss_mlp": 0.01025572, + "balance_loss_clip": 1.05645704, + "balance_loss_mlp": 1.01755548, + "epoch": 0.4341970780977575, + "flos": 23108732058240.0, + "grad_norm": 3.4765264774491773, + "language_loss": 0.83713156, + "learning_rate": 2.5175126843091534e-06, + "loss": 0.85925126, + "num_input_tokens_seen": 77811240, + "step": 3611, + "time_per_iteration": 2.6779091358184814 + }, + { + "auxiliary_loss_clip": 0.0114473, + "auxiliary_loss_mlp": 0.01027287, + "balance_loss_clip": 0.98147023, + "balance_loss_mlp": 1.01929426, + "epoch": 0.43431732098839654, + "flos": 37408288406400.0, + "grad_norm": 2.46230340254047, + "language_loss": 0.75207603, + "learning_rate": 2.5167602031570034e-06, + "loss": 0.77379626, + "num_input_tokens_seen": 77831425, + "step": 3612, + "time_per_iteration": 2.785245895385742 + }, + { + "auxiliary_loss_clip": 0.01187567, + "auxiliary_loss_mlp": 0.01031862, + "balance_loss_clip": 1.05755186, + "balance_loss_mlp": 1.02385104, + "epoch": 0.43443756387903565, + "flos": 31868206323840.0, + "grad_norm": 2.417446316195048, + "language_loss": 0.73310256, + "learning_rate": 2.51600764361274e-06, + "loss": 0.75529683, + "num_input_tokens_seen": 77852950, + "step": 3613, + "time_per_iteration": 2.6732208728790283 + }, + { + "auxiliary_loss_clip": 0.01186726, + "auxiliary_loss_mlp": 0.01030941, + "balance_loss_clip": 1.05604136, + "balance_loss_mlp": 1.0220722, + "epoch": 0.43455780676967476, + "flos": 23477139901440.0, + "grad_norm": 2.536252721006479, + "language_loss": 0.79042518, + "learning_rate": 2.5152550057905283e-06, + "loss": 0.81260186, + "num_input_tokens_seen": 77872840, + "step": 3614, + "time_per_iteration": 2.6209888458251953 + }, + { + "auxiliary_loss_clip": 0.01161925, + "auxiliary_loss_mlp": 0.0051071, + "balance_loss_clip": 1.01890445, + "balance_loss_mlp": 0.0, + "epoch": 0.4346780496603138, + "flos": 24207060176640.0, + "grad_norm": 2.10149995008999, + "language_loss": 0.7729013, + "learning_rate": 2.5145022898045415e-06, + "loss": 0.78962767, + "num_input_tokens_seen": 77892025, + "step": 3615, + "time_per_iteration": 2.676168441772461 + }, + { + "auxiliary_loss_clip": 0.01139116, + "auxiliary_loss_mlp": 0.010323, + "balance_loss_clip": 0.97637314, + "balance_loss_mlp": 1.02401483, + "epoch": 0.4347982925509529, + "flos": 17092366611840.0, + "grad_norm": 2.697079154691399, + "language_loss": 0.89886189, + "learning_rate": 2.5137494957689664e-06, + "loss": 0.92057598, + "num_input_tokens_seen": 77907635, + "step": 3616, + "time_per_iteration": 2.6460120677948 + }, + { + "auxiliary_loss_clip": 0.01050785, + "auxiliary_loss_mlp": 0.01009625, + "balance_loss_clip": 0.95710748, + "balance_loss_mlp": 1.00819433, + "epoch": 0.43491853544159204, + "flos": 60945544696320.0, + "grad_norm": 0.7650467333376874, + "language_loss": 0.5734061, + "learning_rate": 2.5129966237980016e-06, + "loss": 0.59401011, + "num_input_tokens_seen": 77970630, + "step": 3617, + "time_per_iteration": 3.216549873352051 + }, + { + "auxiliary_loss_clip": 0.01118151, + "auxiliary_loss_mlp": 0.01023613, + "balance_loss_clip": 0.94083488, + "balance_loss_mlp": 1.01591158, + "epoch": 0.4350387783322311, + "flos": 21944652094080.0, + "grad_norm": 2.136093034616159, + "language_loss": 0.78423131, + "learning_rate": 2.512243674005857e-06, + "loss": 0.80564898, + "num_input_tokens_seen": 77989995, + "step": 3618, + "time_per_iteration": 2.752502918243408 + }, + { + "auxiliary_loss_clip": 0.01065452, + "auxiliary_loss_mlp": 0.0102903, + "balance_loss_clip": 0.86227369, + "balance_loss_mlp": 1.02078068, + "epoch": 0.4351590212228702, + "flos": 25082705928960.0, + "grad_norm": 1.803853526142086, + "language_loss": 0.86141062, + "learning_rate": 2.5114906465067537e-06, + "loss": 0.88235545, + "num_input_tokens_seen": 78010980, + "step": 3619, + "time_per_iteration": 2.8356759548187256 + }, + { + "auxiliary_loss_clip": 0.01165329, + "auxiliary_loss_mlp": 0.01028749, + "balance_loss_clip": 1.01828539, + "balance_loss_mlp": 1.02077413, + "epoch": 0.4352792641135093, + "flos": 21506541909120.0, + "grad_norm": 4.147047160033438, + "language_loss": 0.74703717, + "learning_rate": 2.5107375414149264e-06, + "loss": 0.76897794, + "num_input_tokens_seen": 78030225, + "step": 3620, + "time_per_iteration": 2.7081148624420166 + }, + { + "auxiliary_loss_clip": 0.01085804, + "auxiliary_loss_mlp": 0.01029548, + "balance_loss_clip": 0.89838874, + "balance_loss_mlp": 1.02163839, + "epoch": 0.43539950700414837, + "flos": 16253457494400.0, + "grad_norm": 2.272402151736182, + "language_loss": 0.71603358, + "learning_rate": 2.5099843588446197e-06, + "loss": 0.73718709, + "num_input_tokens_seen": 78048545, + "step": 3621, + "time_per_iteration": 2.7227365970611572 + }, + { + "auxiliary_loss_clip": 0.01096583, + "auxiliary_loss_mlp": 0.01030397, + "balance_loss_clip": 0.90387619, + "balance_loss_mlp": 1.02263069, + "epoch": 0.4355197498947875, + "flos": 16691819074560.0, + "grad_norm": 1.6287180057364514, + "language_loss": 0.61607224, + "learning_rate": 2.509231098910091e-06, + "loss": 0.63734204, + "num_input_tokens_seen": 78068415, + "step": 3622, + "time_per_iteration": 2.7807412147521973 + }, + { + "auxiliary_loss_clip": 0.01135546, + "auxiliary_loss_mlp": 0.01026051, + "balance_loss_clip": 0.98153436, + "balance_loss_mlp": 1.01758122, + "epoch": 0.4356399927854266, + "flos": 16362733645440.0, + "grad_norm": 4.069598549524226, + "language_loss": 0.74670112, + "learning_rate": 2.508477761725611e-06, + "loss": 0.7683171, + "num_input_tokens_seen": 78086690, + "step": 3623, + "time_per_iteration": 2.6241812705993652 + }, + { + "auxiliary_loss_clip": 0.01166663, + "auxiliary_loss_mlp": 0.01031665, + "balance_loss_clip": 1.01785874, + "balance_loss_mlp": 1.02305174, + "epoch": 0.43576023567606564, + "flos": 17202037812480.0, + "grad_norm": 2.029403853621968, + "language_loss": 0.8073622, + "learning_rate": 2.507724347405458e-06, + "loss": 0.82934541, + "num_input_tokens_seen": 78104640, + "step": 3624, + "time_per_iteration": 2.6560890674591064 + }, + { + "auxiliary_loss_clip": 0.01083323, + "auxiliary_loss_mlp": 0.01023713, + "balance_loss_clip": 0.89689416, + "balance_loss_mlp": 1.01543999, + "epoch": 0.43588047856670475, + "flos": 15917656222080.0, + "grad_norm": 1.949374856854881, + "language_loss": 0.81870794, + "learning_rate": 2.5069708560639243e-06, + "loss": 0.8397783, + "num_input_tokens_seen": 78122550, + "step": 3625, + "time_per_iteration": 2.74027156829834 + }, + { + "auxiliary_loss_clip": 0.01114996, + "auxiliary_loss_mlp": 0.01029924, + "balance_loss_clip": 0.93978018, + "balance_loss_mlp": 1.02180576, + "epoch": 0.4360007214573438, + "flos": 23659566099840.0, + "grad_norm": 2.4347768911830303, + "language_loss": 0.61537766, + "learning_rate": 2.5062172878153158e-06, + "loss": 0.63682693, + "num_input_tokens_seen": 78141825, + "step": 3626, + "time_per_iteration": 2.7145354747772217 + }, + { + "auxiliary_loss_clip": 0.01076357, + "auxiliary_loss_mlp": 0.01028696, + "balance_loss_clip": 0.86500984, + "balance_loss_mlp": 1.02017307, + "epoch": 0.4361209643479829, + "flos": 21978767036160.0, + "grad_norm": 1.8792820154032723, + "language_loss": 0.87335968, + "learning_rate": 2.505463642773947e-06, + "loss": 0.89441013, + "num_input_tokens_seen": 78161790, + "step": 3627, + "time_per_iteration": 2.792428731918335 + }, + { + "auxiliary_loss_clip": 0.01113859, + "auxiliary_loss_mlp": 0.00510166, + "balance_loss_clip": 0.94085032, + "balance_loss_mlp": 0.0, + "epoch": 0.43624120723862203, + "flos": 17420159151360.0, + "grad_norm": 2.3949111569972725, + "language_loss": 0.7549451, + "learning_rate": 2.504709921054146e-06, + "loss": 0.77118528, + "num_input_tokens_seen": 78178605, + "step": 3628, + "time_per_iteration": 3.6714799404144287 + }, + { + "auxiliary_loss_clip": 0.01108576, + "auxiliary_loss_mlp": 0.01028916, + "balance_loss_clip": 0.93579197, + "balance_loss_mlp": 1.0207206, + "epoch": 0.4363614501292611, + "flos": 17895293280000.0, + "grad_norm": 2.1136472161372386, + "language_loss": 0.83596301, + "learning_rate": 2.50395612277025e-06, + "loss": 0.85733795, + "num_input_tokens_seen": 78194460, + "step": 3629, + "time_per_iteration": 2.7085916996002197 + }, + { + "auxiliary_loss_clip": 0.01143712, + "auxiliary_loss_mlp": 0.01032722, + "balance_loss_clip": 0.97800374, + "balance_loss_mlp": 1.02472353, + "epoch": 0.4364816930199002, + "flos": 20302888135680.0, + "grad_norm": 6.632808669003986, + "language_loss": 0.72719318, + "learning_rate": 2.503202248036612e-06, + "loss": 0.74895751, + "num_input_tokens_seen": 78213315, + "step": 3630, + "time_per_iteration": 3.6231143474578857 + }, + { + "auxiliary_loss_clip": 0.0118534, + "auxiliary_loss_mlp": 0.01026451, + "balance_loss_clip": 1.05681002, + "balance_loss_mlp": 1.01867902, + "epoch": 0.4366019359105393, + "flos": 24061334699520.0, + "grad_norm": 1.717404029848951, + "language_loss": 0.73379493, + "learning_rate": 2.5024482969675927e-06, + "loss": 0.75591278, + "num_input_tokens_seen": 78233270, + "step": 3631, + "time_per_iteration": 3.564908027648926 + }, + { + "auxiliary_loss_clip": 0.01095225, + "auxiliary_loss_mlp": 0.0102678, + "balance_loss_clip": 0.90091622, + "balance_loss_mlp": 1.01951456, + "epoch": 0.43672217880117836, + "flos": 21754109422080.0, + "grad_norm": 2.1124356587176356, + "language_loss": 0.84519839, + "learning_rate": 2.501694269677566e-06, + "loss": 0.86641848, + "num_input_tokens_seen": 78251040, + "step": 3632, + "time_per_iteration": 2.7096734046936035 + }, + { + "auxiliary_loss_clip": 0.01165771, + "auxiliary_loss_mlp": 0.01031978, + "balance_loss_clip": 1.01687169, + "balance_loss_mlp": 1.02328157, + "epoch": 0.4368424216918175, + "flos": 18035200753920.0, + "grad_norm": 2.3522251937666456, + "language_loss": 0.80486238, + "learning_rate": 2.500940166280918e-06, + "loss": 0.82683992, + "num_input_tokens_seen": 78269470, + "step": 3633, + "time_per_iteration": 2.621454954147339 + }, + { + "auxiliary_loss_clip": 0.01157789, + "auxiliary_loss_mlp": 0.01026118, + "balance_loss_clip": 1.01588762, + "balance_loss_mlp": 1.01857805, + "epoch": 0.4369626645824566, + "flos": 25447127362560.0, + "grad_norm": 2.43252164553354, + "language_loss": 0.79112512, + "learning_rate": 2.500185986892045e-06, + "loss": 0.8129642, + "num_input_tokens_seen": 78288955, + "step": 3634, + "time_per_iteration": 2.660583972930908 + }, + { + "auxiliary_loss_clip": 0.01156356, + "auxiliary_loss_mlp": 0.01029656, + "balance_loss_clip": 1.01383901, + "balance_loss_mlp": 1.02143061, + "epoch": 0.43708290747309564, + "flos": 25302694775040.0, + "grad_norm": 2.442270634039535, + "language_loss": 0.77501059, + "learning_rate": 2.499431731625355e-06, + "loss": 0.79687071, + "num_input_tokens_seen": 78307980, + "step": 3635, + "time_per_iteration": 3.5601727962493896 + }, + { + "auxiliary_loss_clip": 0.01186378, + "auxiliary_loss_mlp": 0.01026534, + "balance_loss_clip": 1.05509639, + "balance_loss_mlp": 1.01821899, + "epoch": 0.43720315036373475, + "flos": 31575103344000.0, + "grad_norm": 1.834256255029392, + "language_loss": 0.79640555, + "learning_rate": 2.4986774005952686e-06, + "loss": 0.81853473, + "num_input_tokens_seen": 78330355, + "step": 3636, + "time_per_iteration": 2.686356782913208 + }, + { + "auxiliary_loss_clip": 0.01161072, + "auxiliary_loss_mlp": 0.01027082, + "balance_loss_clip": 1.02148962, + "balance_loss_mlp": 1.01919627, + "epoch": 0.43732339325437386, + "flos": 23112000195840.0, + "grad_norm": 2.229672453099712, + "language_loss": 0.84479815, + "learning_rate": 2.4979229939162166e-06, + "loss": 0.86667967, + "num_input_tokens_seen": 78349135, + "step": 3637, + "time_per_iteration": 2.6285383701324463 + }, + { + "auxiliary_loss_clip": 0.01158267, + "auxiliary_loss_mlp": 0.01029033, + "balance_loss_clip": 1.01883054, + "balance_loss_mlp": 1.02169549, + "epoch": 0.4374436361450129, + "flos": 27746272080000.0, + "grad_norm": 1.9909924168133124, + "language_loss": 0.8060568, + "learning_rate": 2.4971685117026433e-06, + "loss": 0.82792974, + "num_input_tokens_seen": 78368900, + "step": 3638, + "time_per_iteration": 2.661912441253662 + }, + { + "auxiliary_loss_clip": 0.0116675, + "auxiliary_loss_mlp": 0.01032171, + "balance_loss_clip": 1.01923454, + "balance_loss_mlp": 1.02411258, + "epoch": 0.437563879035652, + "flos": 24172370616960.0, + "grad_norm": 1.8461238226647025, + "language_loss": 0.77048481, + "learning_rate": 2.4964139540690018e-06, + "loss": 0.79247403, + "num_input_tokens_seen": 78392235, + "step": 3639, + "time_per_iteration": 2.740582227706909 + }, + { + "auxiliary_loss_clip": 0.01123218, + "auxiliary_loss_mlp": 0.01031904, + "balance_loss_clip": 0.94612646, + "balance_loss_mlp": 1.02397096, + "epoch": 0.4376841219262911, + "flos": 23477211728640.0, + "grad_norm": 1.8563962438356487, + "language_loss": 0.72711182, + "learning_rate": 2.495659321129758e-06, + "loss": 0.74866307, + "num_input_tokens_seen": 78409980, + "step": 3640, + "time_per_iteration": 2.748072385787964 + }, + { + "auxiliary_loss_clip": 0.01157663, + "auxiliary_loss_mlp": 0.01030568, + "balance_loss_clip": 1.01577258, + "balance_loss_mlp": 1.02239656, + "epoch": 0.4378043648169302, + "flos": 25447809720960.0, + "grad_norm": 5.864263783163867, + "language_loss": 0.75247943, + "learning_rate": 2.494904612999389e-06, + "loss": 0.77436173, + "num_input_tokens_seen": 78428690, + "step": 3641, + "time_per_iteration": 2.754516363143921 + }, + { + "auxiliary_loss_clip": 0.01069933, + "auxiliary_loss_mlp": 0.01002581, + "balance_loss_clip": 0.99533272, + "balance_loss_mlp": 1.00137091, + "epoch": 0.4379246077075693, + "flos": 53914056986880.0, + "grad_norm": 0.7499332230825089, + "language_loss": 0.56512332, + "learning_rate": 2.4941498297923843e-06, + "loss": 0.58584845, + "num_input_tokens_seen": 78489260, + "step": 3642, + "time_per_iteration": 3.215003728866577 + }, + { + "auxiliary_loss_clip": 0.01159939, + "auxiliary_loss_mlp": 0.0102671, + "balance_loss_clip": 1.01936507, + "balance_loss_mlp": 1.01890147, + "epoch": 0.43804485059820836, + "flos": 20588305605120.0, + "grad_norm": 1.9260000234105121, + "language_loss": 0.6969434, + "learning_rate": 2.4933949716232424e-06, + "loss": 0.71880996, + "num_input_tokens_seen": 78506785, + "step": 3643, + "time_per_iteration": 2.5782999992370605 + }, + { + "auxiliary_loss_clip": 0.01116594, + "auxiliary_loss_mlp": 0.01027513, + "balance_loss_clip": 0.94547737, + "balance_loss_mlp": 1.01999068, + "epoch": 0.43816509348884747, + "flos": 23876214981120.0, + "grad_norm": 2.2590934502606608, + "language_loss": 0.74207056, + "learning_rate": 2.492640038606476e-06, + "loss": 0.7635116, + "num_input_tokens_seen": 78525150, + "step": 3644, + "time_per_iteration": 2.7168679237365723 + }, + { + "auxiliary_loss_clip": 0.01162569, + "auxiliary_loss_mlp": 0.01028715, + "balance_loss_clip": 1.01822114, + "balance_loss_mlp": 1.02062094, + "epoch": 0.4382853363794866, + "flos": 14684448533760.0, + "grad_norm": 3.1228536150499213, + "language_loss": 0.78266346, + "learning_rate": 2.491885030856608e-06, + "loss": 0.80457628, + "num_input_tokens_seen": 78543245, + "step": 3645, + "time_per_iteration": 2.59226393699646 + }, + { + "auxiliary_loss_clip": 0.01138012, + "auxiliary_loss_mlp": 0.01031251, + "balance_loss_clip": 0.97883469, + "balance_loss_mlp": 1.02267396, + "epoch": 0.43840557927012563, + "flos": 17165301177600.0, + "grad_norm": 2.224864384212508, + "language_loss": 0.82493174, + "learning_rate": 2.4911299484881713e-06, + "loss": 0.84662437, + "num_input_tokens_seen": 78560775, + "step": 3646, + "time_per_iteration": 2.646660566329956 + }, + { + "auxiliary_loss_clip": 0.01133245, + "auxiliary_loss_mlp": 0.01029246, + "balance_loss_clip": 0.97758031, + "balance_loss_mlp": 1.02208781, + "epoch": 0.43852582216076474, + "flos": 19390685316480.0, + "grad_norm": 2.3151752152061036, + "language_loss": 0.81063396, + "learning_rate": 2.490374791615712e-06, + "loss": 0.83225888, + "num_input_tokens_seen": 78580800, + "step": 3647, + "time_per_iteration": 2.6714186668395996 + }, + { + "auxiliary_loss_clip": 0.01190466, + "auxiliary_loss_mlp": 0.00510802, + "balance_loss_clip": 1.05887723, + "balance_loss_mlp": 0.0, + "epoch": 0.43864606505140386, + "flos": 18075133699200.0, + "grad_norm": 2.7832497512211773, + "language_loss": 0.7765305, + "learning_rate": 2.4896195603537867e-06, + "loss": 0.79354322, + "num_input_tokens_seen": 78595410, + "step": 3648, + "time_per_iteration": 2.5716986656188965 + }, + { + "auxiliary_loss_clip": 0.01089354, + "auxiliary_loss_mlp": 0.01026247, + "balance_loss_clip": 0.90881765, + "balance_loss_mlp": 1.01849866, + "epoch": 0.4387663079420429, + "flos": 19644896845440.0, + "grad_norm": 1.854013782571472, + "language_loss": 0.73888749, + "learning_rate": 2.488864254816964e-06, + "loss": 0.7600435, + "num_input_tokens_seen": 78614100, + "step": 3649, + "time_per_iteration": 2.7114145755767822 + }, + { + "auxiliary_loss_clip": 0.01165999, + "auxiliary_loss_mlp": 0.01030667, + "balance_loss_clip": 1.02116084, + "balance_loss_mlp": 1.02267385, + "epoch": 0.438886550832682, + "flos": 19719339782400.0, + "grad_norm": 1.9691540262700302, + "language_loss": 0.67699909, + "learning_rate": 2.4881088751198218e-06, + "loss": 0.69896579, + "num_input_tokens_seen": 78632260, + "step": 3650, + "time_per_iteration": 2.6268508434295654 + }, + { + "auxiliary_loss_clip": 0.01140205, + "auxiliary_loss_mlp": 0.01030032, + "balance_loss_clip": 0.97928345, + "balance_loss_mlp": 1.02134156, + "epoch": 0.43900679372332113, + "flos": 14536675981440.0, + "grad_norm": 3.1311152638601265, + "language_loss": 0.64302474, + "learning_rate": 2.4873534213769517e-06, + "loss": 0.66472709, + "num_input_tokens_seen": 78647490, + "step": 3651, + "time_per_iteration": 2.693862199783325 + }, + { + "auxiliary_loss_clip": 0.01111105, + "auxiliary_loss_mlp": 0.01026468, + "balance_loss_clip": 0.94529748, + "balance_loss_mlp": 1.01879084, + "epoch": 0.4391270366139602, + "flos": 24056234968320.0, + "grad_norm": 1.9445693071376386, + "language_loss": 0.72014964, + "learning_rate": 2.4865978937029547e-06, + "loss": 0.74152541, + "num_input_tokens_seen": 78666470, + "step": 3652, + "time_per_iteration": 2.7648215293884277 + }, + { + "auxiliary_loss_clip": 0.01086766, + "auxiliary_loss_mlp": 0.01029348, + "balance_loss_clip": 0.9042055, + "balance_loss_mlp": 1.02146804, + "epoch": 0.4392472795045993, + "flos": 31538510363520.0, + "grad_norm": 1.6755279714258722, + "language_loss": 0.66499257, + "learning_rate": 2.485842292212445e-06, + "loss": 0.68615365, + "num_input_tokens_seen": 78687685, + "step": 3653, + "time_per_iteration": 2.8578622341156006 + }, + { + "auxiliary_loss_clip": 0.01193579, + "auxiliary_loss_mlp": 0.01032248, + "balance_loss_clip": 1.06266308, + "balance_loss_mlp": 1.02443993, + "epoch": 0.4393675223952384, + "flos": 14866300114560.0, + "grad_norm": 2.2409249505132625, + "language_loss": 0.80256128, + "learning_rate": 2.485086617020045e-06, + "loss": 0.82481956, + "num_input_tokens_seen": 78706180, + "step": 3654, + "time_per_iteration": 3.5396969318389893 + }, + { + "auxiliary_loss_clip": 0.01131511, + "auxiliary_loss_mlp": 0.01026658, + "balance_loss_clip": 0.97705835, + "balance_loss_mlp": 1.01830721, + "epoch": 0.43948776528587746, + "flos": 14825900292480.0, + "grad_norm": 3.20817251776435, + "language_loss": 0.82098019, + "learning_rate": 2.4843308682403903e-06, + "loss": 0.8425619, + "num_input_tokens_seen": 78723095, + "step": 3655, + "time_per_iteration": 2.719125747680664 + }, + { + "auxiliary_loss_clip": 0.01186391, + "auxiliary_loss_mlp": 0.01026385, + "balance_loss_clip": 1.05694675, + "balance_loss_mlp": 1.01877964, + "epoch": 0.4396080081765166, + "flos": 13914523486080.0, + "grad_norm": 1.6029241436244697, + "language_loss": 0.82704782, + "learning_rate": 2.4835750459881294e-06, + "loss": 0.84917557, + "num_input_tokens_seen": 78739720, + "step": 3656, + "time_per_iteration": 3.523782968521118 + }, + { + "auxiliary_loss_clip": 0.0113225, + "auxiliary_loss_mlp": 0.01034258, + "balance_loss_clip": 0.97824943, + "balance_loss_mlp": 1.02588987, + "epoch": 0.43972825106715563, + "flos": 18222978078720.0, + "grad_norm": 2.9091223691847556, + "language_loss": 0.81847191, + "learning_rate": 2.4828191503779177e-06, + "loss": 0.840137, + "num_input_tokens_seen": 78757820, + "step": 3657, + "time_per_iteration": 3.522244691848755 + }, + { + "auxiliary_loss_clip": 0.01122487, + "auxiliary_loss_mlp": 0.01036746, + "balance_loss_clip": 0.94702935, + "balance_loss_mlp": 1.02889562, + "epoch": 0.43984849395779474, + "flos": 16873239692160.0, + "grad_norm": 2.0669323746985495, + "language_loss": 0.8991245, + "learning_rate": 2.482063181524425e-06, + "loss": 0.92071688, + "num_input_tokens_seen": 78773720, + "step": 3658, + "time_per_iteration": 2.702958106994629 + }, + { + "auxiliary_loss_clip": 0.0119107, + "auxiliary_loss_mlp": 0.0103459, + "balance_loss_clip": 1.06119609, + "balance_loss_mlp": 1.02609658, + "epoch": 0.43996873684843385, + "flos": 18691504104960.0, + "grad_norm": 4.745450168128921, + "language_loss": 0.80998498, + "learning_rate": 2.4813071395423307e-06, + "loss": 0.83224154, + "num_input_tokens_seen": 78791285, + "step": 3659, + "time_per_iteration": 2.5970826148986816 + }, + { + "auxiliary_loss_clip": 0.01164642, + "auxiliary_loss_mlp": 0.01029847, + "balance_loss_clip": 1.01930773, + "balance_loss_mlp": 1.0213232, + "epoch": 0.4400889797390729, + "flos": 23653460787840.0, + "grad_norm": 1.8241257882791502, + "language_loss": 0.64186507, + "learning_rate": 2.4805510245463263e-06, + "loss": 0.66380996, + "num_input_tokens_seen": 78811440, + "step": 3660, + "time_per_iteration": 3.5744149684906006 + }, + { + "auxiliary_loss_clip": 0.01165038, + "auxiliary_loss_mlp": 0.01027946, + "balance_loss_clip": 1.01908553, + "balance_loss_mlp": 1.02019763, + "epoch": 0.440209222629712, + "flos": 23149203707520.0, + "grad_norm": 2.613122072519736, + "language_loss": 0.60609645, + "learning_rate": 2.4797948366511137e-06, + "loss": 0.62802631, + "num_input_tokens_seen": 78831150, + "step": 3661, + "time_per_iteration": 2.683887481689453 + }, + { + "auxiliary_loss_clip": 0.01112154, + "auxiliary_loss_mlp": 0.01026688, + "balance_loss_clip": 0.93807113, + "balance_loss_mlp": 1.01836753, + "epoch": 0.4403294655203511, + "flos": 24823394668800.0, + "grad_norm": 1.8838946744353746, + "language_loss": 0.75734806, + "learning_rate": 2.4790385759714055e-06, + "loss": 0.77873647, + "num_input_tokens_seen": 78850215, + "step": 3662, + "time_per_iteration": 2.746992826461792 + }, + { + "auxiliary_loss_clip": 0.01161354, + "auxiliary_loss_mlp": 0.01032505, + "balance_loss_clip": 1.02113545, + "balance_loss_mlp": 1.02492285, + "epoch": 0.4404497084109902, + "flos": 22565080736640.0, + "grad_norm": 1.9201022402575227, + "language_loss": 0.71272099, + "learning_rate": 2.478282242621926e-06, + "loss": 0.73465955, + "num_input_tokens_seen": 78870675, + "step": 3663, + "time_per_iteration": 2.6763041019439697 + }, + { + "auxiliary_loss_clip": 0.01032035, + "auxiliary_loss_mlp": 0.01003003, + "balance_loss_clip": 0.91826236, + "balance_loss_mlp": 1.00175154, + "epoch": 0.4405699513016293, + "flos": 64967073448320.0, + "grad_norm": 0.8535615692257688, + "language_loss": 0.59559262, + "learning_rate": 2.477525836717411e-06, + "loss": 0.61594301, + "num_input_tokens_seen": 78938440, + "step": 3664, + "time_per_iteration": 3.4503252506256104 + }, + { + "auxiliary_loss_clip": 0.01161901, + "auxiliary_loss_mlp": 0.01030957, + "balance_loss_clip": 1.01822317, + "balance_loss_mlp": 1.02260673, + "epoch": 0.4406901941922684, + "flos": 35661952978560.0, + "grad_norm": 1.999139588432906, + "language_loss": 0.79783189, + "learning_rate": 2.476769358372606e-06, + "loss": 0.81976056, + "num_input_tokens_seen": 78960090, + "step": 3665, + "time_per_iteration": 2.790907621383667 + }, + { + "auxiliary_loss_clip": 0.01120134, + "auxiliary_loss_mlp": 0.01027929, + "balance_loss_clip": 0.94978762, + "balance_loss_mlp": 1.02006757, + "epoch": 0.44081043708290746, + "flos": 18040767361920.0, + "grad_norm": 2.3137844163242414, + "language_loss": 0.74630129, + "learning_rate": 2.4760128077022683e-06, + "loss": 0.76778197, + "num_input_tokens_seen": 78978225, + "step": 3666, + "time_per_iteration": 2.6837425231933594 + }, + { + "auxiliary_loss_clip": 0.01087918, + "auxiliary_loss_mlp": 0.01031016, + "balance_loss_clip": 0.90500188, + "balance_loss_mlp": 1.02318978, + "epoch": 0.44093067997354657, + "flos": 30153507799680.0, + "grad_norm": 1.5521345494447578, + "language_loss": 0.68576229, + "learning_rate": 2.4752561848211672e-06, + "loss": 0.70695162, + "num_input_tokens_seen": 79000625, + "step": 3667, + "time_per_iteration": 2.860326051712036 + }, + { + "auxiliary_loss_clip": 0.01162797, + "auxiliary_loss_mlp": 0.01027553, + "balance_loss_clip": 1.02375913, + "balance_loss_mlp": 1.02017426, + "epoch": 0.4410509228641857, + "flos": 23255068066560.0, + "grad_norm": 2.0403448754592866, + "language_loss": 0.71393847, + "learning_rate": 2.4744994898440797e-06, + "loss": 0.73584199, + "num_input_tokens_seen": 79019415, + "step": 3668, + "time_per_iteration": 2.6226484775543213 + }, + { + "auxiliary_loss_clip": 0.01122041, + "auxiliary_loss_mlp": 0.01029305, + "balance_loss_clip": 0.94342434, + "balance_loss_mlp": 1.02115059, + "epoch": 0.44117116575482473, + "flos": 19500571998720.0, + "grad_norm": 1.996395963617284, + "language_loss": 0.83574986, + "learning_rate": 2.473742722885797e-06, + "loss": 0.85726333, + "num_input_tokens_seen": 79038435, + "step": 3669, + "time_per_iteration": 2.687676429748535 + }, + { + "auxiliary_loss_clip": 0.01167, + "auxiliary_loss_mlp": 0.00510116, + "balance_loss_clip": 1.0242691, + "balance_loss_mlp": 0.0, + "epoch": 0.44129140864546385, + "flos": 27053124353280.0, + "grad_norm": 3.2777933693504377, + "language_loss": 0.655788, + "learning_rate": 2.4729858840611197e-06, + "loss": 0.6725592, + "num_input_tokens_seen": 79057345, + "step": 3670, + "time_per_iteration": 2.6682262420654297 + }, + { + "auxiliary_loss_clip": 0.01186233, + "auxiliary_loss_mlp": 0.01029481, + "balance_loss_clip": 1.05848002, + "balance_loss_mlp": 1.02176261, + "epoch": 0.4414116515361029, + "flos": 26102101910400.0, + "grad_norm": 2.259559548381101, + "language_loss": 0.72715807, + "learning_rate": 2.4722289734848605e-06, + "loss": 0.74931514, + "num_input_tokens_seen": 79077810, + "step": 3671, + "time_per_iteration": 2.6309754848480225 + }, + { + "auxiliary_loss_clip": 0.01112799, + "auxiliary_loss_mlp": 0.01030442, + "balance_loss_clip": 0.94313049, + "balance_loss_mlp": 1.02260995, + "epoch": 0.441531894426742, + "flos": 21906083865600.0, + "grad_norm": 2.157945949796001, + "language_loss": 0.78082073, + "learning_rate": 2.471471991271841e-06, + "loss": 0.80225313, + "num_input_tokens_seen": 79094935, + "step": 3672, + "time_per_iteration": 2.7174041271209717 + }, + { + "auxiliary_loss_clip": 0.01153877, + "auxiliary_loss_mlp": 0.01025565, + "balance_loss_clip": 1.01492774, + "balance_loss_mlp": 1.01736927, + "epoch": 0.4416521373173811, + "flos": 23437099215360.0, + "grad_norm": 2.052312651213339, + "language_loss": 0.79695868, + "learning_rate": 2.470714937536896e-06, + "loss": 0.81875312, + "num_input_tokens_seen": 79113660, + "step": 3673, + "time_per_iteration": 2.6615121364593506 + }, + { + "auxiliary_loss_clip": 0.01090096, + "auxiliary_loss_mlp": 0.01025087, + "balance_loss_clip": 0.90445566, + "balance_loss_mlp": 1.01718318, + "epoch": 0.4417723802080202, + "flos": 20334345471360.0, + "grad_norm": 2.122326159401547, + "language_loss": 0.70652664, + "learning_rate": 2.469957812394868e-06, + "loss": 0.72767848, + "num_input_tokens_seen": 79132470, + "step": 3674, + "time_per_iteration": 2.715877056121826 + }, + { + "auxiliary_loss_clip": 0.01185925, + "auxiliary_loss_mlp": 0.010309, + "balance_loss_clip": 1.0590899, + "balance_loss_mlp": 1.02313924, + "epoch": 0.4418926230986593, + "flos": 18880682060160.0, + "grad_norm": 2.1387287628063203, + "language_loss": 0.75942516, + "learning_rate": 2.4692006159606148e-06, + "loss": 0.78159332, + "num_input_tokens_seen": 79150000, + "step": 3675, + "time_per_iteration": 2.625814914703369 + }, + { + "auxiliary_loss_clip": 0.01186209, + "auxiliary_loss_mlp": 0.01030958, + "balance_loss_clip": 1.05799222, + "balance_loss_mlp": 1.02319098, + "epoch": 0.4420128659892984, + "flos": 19464409981440.0, + "grad_norm": 2.267120385091157, + "language_loss": 0.78609955, + "learning_rate": 2.468443348349e-06, + "loss": 0.80827123, + "num_input_tokens_seen": 79167875, + "step": 3676, + "time_per_iteration": 2.660010814666748 + }, + { + "auxiliary_loss_clip": 0.01088519, + "auxiliary_loss_mlp": 0.01029432, + "balance_loss_clip": 0.90180868, + "balance_loss_mlp": 1.0210278, + "epoch": 0.44213310887993745, + "flos": 17894359526400.0, + "grad_norm": 2.4591129408855914, + "language_loss": 0.82502311, + "learning_rate": 2.467686009674902e-06, + "loss": 0.84620261, + "num_input_tokens_seen": 79182325, + "step": 3677, + "time_per_iteration": 2.725470542907715 + }, + { + "auxiliary_loss_clip": 0.01159142, + "auxiliary_loss_mlp": 0.01035224, + "balance_loss_clip": 1.01774907, + "balance_loss_mlp": 1.02687383, + "epoch": 0.44225335177057656, + "flos": 19204667758080.0, + "grad_norm": 2.312273042305567, + "language_loss": 0.84902012, + "learning_rate": 2.466928600053209e-06, + "loss": 0.87096375, + "num_input_tokens_seen": 79197630, + "step": 3678, + "time_per_iteration": 2.6368331909179688 + }, + { + "auxiliary_loss_clip": 0.01136693, + "auxiliary_loss_mlp": 0.01019944, + "balance_loss_clip": 0.97948921, + "balance_loss_mlp": 1.0129168, + "epoch": 0.4423735946612157, + "flos": 23471321898240.0, + "grad_norm": 4.131339262483984, + "language_loss": 0.71339524, + "learning_rate": 2.466171119598818e-06, + "loss": 0.73496163, + "num_input_tokens_seen": 79217600, + "step": 3679, + "time_per_iteration": 2.683112859725952 + }, + { + "auxiliary_loss_clip": 0.01166623, + "auxiliary_loss_mlp": 0.01029199, + "balance_loss_clip": 1.01679838, + "balance_loss_mlp": 1.02110517, + "epoch": 0.44249383755185473, + "flos": 26685398868480.0, + "grad_norm": 2.0328503596823064, + "language_loss": 0.77004611, + "learning_rate": 2.465413568426639e-06, + "loss": 0.79200435, + "num_input_tokens_seen": 79238550, + "step": 3680, + "time_per_iteration": 3.6389265060424805 + }, + { + "auxiliary_loss_clip": 0.01156642, + "auxiliary_loss_mlp": 0.01024371, + "balance_loss_clip": 1.01697981, + "balance_loss_mlp": 1.01705122, + "epoch": 0.44261408044249384, + "flos": 23147659422720.0, + "grad_norm": 1.675788342199919, + "language_loss": 0.81303567, + "learning_rate": 2.464655946651591e-06, + "loss": 0.83484578, + "num_input_tokens_seen": 79257555, + "step": 3681, + "time_per_iteration": 2.619745969772339 + }, + { + "auxiliary_loss_clip": 0.01165606, + "auxiliary_loss_mlp": 0.0102868, + "balance_loss_clip": 1.01994216, + "balance_loss_mlp": 1.02014458, + "epoch": 0.44273432333313295, + "flos": 24462564595200.0, + "grad_norm": 2.1299698142958627, + "language_loss": 0.80841982, + "learning_rate": 2.4638982543886065e-06, + "loss": 0.83036268, + "num_input_tokens_seen": 79277595, + "step": 3682, + "time_per_iteration": 3.565436363220215 + }, + { + "auxiliary_loss_clip": 0.01167136, + "auxiliary_loss_mlp": 0.01030879, + "balance_loss_clip": 1.02193594, + "balance_loss_mlp": 1.02321362, + "epoch": 0.442854566223772, + "flos": 17528932512000.0, + "grad_norm": 3.6111423116326704, + "language_loss": 0.87087035, + "learning_rate": 2.4631404917526254e-06, + "loss": 0.89285052, + "num_input_tokens_seen": 79294550, + "step": 3683, + "time_per_iteration": 3.474053382873535 + }, + { + "auxiliary_loss_clip": 0.01156865, + "auxiliary_loss_mlp": 0.01033682, + "balance_loss_clip": 1.01784825, + "balance_loss_mlp": 1.02527213, + "epoch": 0.4429748091144111, + "flos": 24896293320960.0, + "grad_norm": 1.6089994176119549, + "language_loss": 0.79329193, + "learning_rate": 2.4623826588586e-06, + "loss": 0.81519735, + "num_input_tokens_seen": 79314820, + "step": 3684, + "time_per_iteration": 2.627333641052246 + }, + { + "auxiliary_loss_clip": 0.01131394, + "auxiliary_loss_mlp": 0.01029287, + "balance_loss_clip": 0.97430086, + "balance_loss_mlp": 1.02156186, + "epoch": 0.4430950520050502, + "flos": 21614704738560.0, + "grad_norm": 2.210103716165822, + "language_loss": 0.82460475, + "learning_rate": 2.461624755821492e-06, + "loss": 0.84621155, + "num_input_tokens_seen": 79334300, + "step": 3685, + "time_per_iteration": 2.669515371322632 + }, + { + "auxiliary_loss_clip": 0.01121667, + "auxiliary_loss_mlp": 0.01034305, + "balance_loss_clip": 0.94623756, + "balance_loss_mlp": 1.02644324, + "epoch": 0.4432152948956893, + "flos": 24572271709440.0, + "grad_norm": 3.20226810095858, + "language_loss": 0.76187462, + "learning_rate": 2.4608667827562763e-06, + "loss": 0.78343427, + "num_input_tokens_seen": 79353630, + "step": 3686, + "time_per_iteration": 2.7451021671295166 + }, + { + "auxiliary_loss_clip": 0.01167577, + "auxiliary_loss_mlp": 0.01026247, + "balance_loss_clip": 1.01964653, + "balance_loss_mlp": 1.01836085, + "epoch": 0.4433355377863284, + "flos": 21762261809280.0, + "grad_norm": 5.195533804453531, + "language_loss": 0.89905429, + "learning_rate": 2.460108739777936e-06, + "loss": 0.92099255, + "num_input_tokens_seen": 79372765, + "step": 3687, + "time_per_iteration": 3.5583152770996094 + }, + { + "auxiliary_loss_clip": 0.0114156, + "auxiliary_loss_mlp": 0.01025063, + "balance_loss_clip": 0.98508561, + "balance_loss_mlp": 1.0173862, + "epoch": 0.44345578067696745, + "flos": 20084479488000.0, + "grad_norm": 1.545152257115885, + "language_loss": 0.76321602, + "learning_rate": 2.4593506270014656e-06, + "loss": 0.78488219, + "num_input_tokens_seen": 79391735, + "step": 3688, + "time_per_iteration": 2.681244134902954 + }, + { + "auxiliary_loss_clip": 0.01141647, + "auxiliary_loss_mlp": 0.01033568, + "balance_loss_clip": 0.9786247, + "balance_loss_mlp": 1.02566457, + "epoch": 0.44357602356760656, + "flos": 24169497528960.0, + "grad_norm": 1.6312388026844282, + "language_loss": 0.81944859, + "learning_rate": 2.45859244454187e-06, + "loss": 0.84120071, + "num_input_tokens_seen": 79411525, + "step": 3689, + "time_per_iteration": 2.683800458908081 + }, + { + "auxiliary_loss_clip": 0.01160245, + "auxiliary_loss_mlp": 0.01027238, + "balance_loss_clip": 1.01949286, + "balance_loss_mlp": 1.01962698, + "epoch": 0.44369626645824567, + "flos": 22707717644160.0, + "grad_norm": 1.6570624248487993, + "language_loss": 0.6629144, + "learning_rate": 2.4578341925141655e-06, + "loss": 0.68478924, + "num_input_tokens_seen": 79430740, + "step": 3690, + "time_per_iteration": 2.6937015056610107 + }, + { + "auxiliary_loss_clip": 0.01167254, + "auxiliary_loss_mlp": 0.01027291, + "balance_loss_clip": 1.01812279, + "balance_loss_mlp": 1.01923871, + "epoch": 0.4438165093488847, + "flos": 38030225420160.0, + "grad_norm": 4.03968073633358, + "language_loss": 0.72667331, + "learning_rate": 2.457075871033378e-06, + "loss": 0.74861884, + "num_input_tokens_seen": 79452615, + "step": 3691, + "time_per_iteration": 2.8467416763305664 + }, + { + "auxiliary_loss_clip": 0.01112978, + "auxiliary_loss_mlp": 0.01024568, + "balance_loss_clip": 0.93978316, + "balance_loss_mlp": 1.01717663, + "epoch": 0.44393675223952384, + "flos": 15523213996800.0, + "grad_norm": 2.2910455230260904, + "language_loss": 0.88742119, + "learning_rate": 2.4563174802145445e-06, + "loss": 0.90879667, + "num_input_tokens_seen": 79469865, + "step": 3692, + "time_per_iteration": 2.693925142288208 + }, + { + "auxiliary_loss_clip": 0.01048541, + "auxiliary_loss_mlp": 0.01003228, + "balance_loss_clip": 0.95313358, + "balance_loss_mlp": 1.00194073, + "epoch": 0.44405699513016295, + "flos": 64574893779840.0, + "grad_norm": 0.6335604707817063, + "language_loss": 0.48590985, + "learning_rate": 2.455559020172712e-06, + "loss": 0.50642753, + "num_input_tokens_seen": 79537220, + "step": 3693, + "time_per_iteration": 3.2939229011535645 + }, + { + "auxiliary_loss_clip": 0.01093272, + "auxiliary_loss_mlp": 0.01028455, + "balance_loss_clip": 0.90701205, + "balance_loss_mlp": 1.02013469, + "epoch": 0.444177238020802, + "flos": 23987394552960.0, + "grad_norm": 2.2803982312632196, + "language_loss": 0.89691019, + "learning_rate": 2.4548004910229385e-06, + "loss": 0.91812742, + "num_input_tokens_seen": 79554795, + "step": 3694, + "time_per_iteration": 2.7445709705352783 + }, + { + "auxiliary_loss_clip": 0.01164662, + "auxiliary_loss_mlp": 0.00509384, + "balance_loss_clip": 1.01997197, + "balance_loss_mlp": 0.0, + "epoch": 0.4442974809114411, + "flos": 22563069575040.0, + "grad_norm": 2.0534581130821254, + "language_loss": 0.86704576, + "learning_rate": 2.4540418928802913e-06, + "loss": 0.88378626, + "num_input_tokens_seen": 79573530, + "step": 3695, + "time_per_iteration": 2.670161008834839 + }, + { + "auxiliary_loss_clip": 0.01138324, + "auxiliary_loss_mlp": 0.01030399, + "balance_loss_clip": 0.97980434, + "balance_loss_mlp": 1.0212853, + "epoch": 0.4444177238020802, + "flos": 17675699483520.0, + "grad_norm": 3.3582429100062994, + "language_loss": 0.65663773, + "learning_rate": 2.4532832258598506e-06, + "loss": 0.678325, + "num_input_tokens_seen": 79591360, + "step": 3696, + "time_per_iteration": 2.677093267440796 + }, + { + "auxiliary_loss_clip": 0.01182437, + "auxiliary_loss_mlp": 0.01027117, + "balance_loss_clip": 1.05600917, + "balance_loss_mlp": 1.01943421, + "epoch": 0.4445379666927193, + "flos": 28621594609920.0, + "grad_norm": 1.7855651616193253, + "language_loss": 0.80895257, + "learning_rate": 2.4525244900767047e-06, + "loss": 0.83104813, + "num_input_tokens_seen": 79612175, + "step": 3697, + "time_per_iteration": 2.7317512035369873 + }, + { + "auxiliary_loss_clip": 0.01063966, + "auxiliary_loss_mlp": 0.01004961, + "balance_loss_clip": 0.99085945, + "balance_loss_mlp": 1.00336397, + "epoch": 0.4446582095833584, + "flos": 70487370115200.0, + "grad_norm": 0.7666023517758304, + "language_loss": 0.60471737, + "learning_rate": 2.4517656856459536e-06, + "loss": 0.62540668, + "num_input_tokens_seen": 79678020, + "step": 3698, + "time_per_iteration": 3.348397970199585 + }, + { + "auxiliary_loss_clip": 0.01162057, + "auxiliary_loss_mlp": 0.01031421, + "balance_loss_clip": 1.01868916, + "balance_loss_mlp": 1.02375555, + "epoch": 0.4447784524739975, + "flos": 26505199313280.0, + "grad_norm": 1.6214047572459895, + "language_loss": 0.68122995, + "learning_rate": 2.4510068126827073e-06, + "loss": 0.70316476, + "num_input_tokens_seen": 79699020, + "step": 3699, + "time_per_iteration": 2.6494643688201904 + }, + { + "auxiliary_loss_clip": 0.01135037, + "auxiliary_loss_mlp": 0.01029183, + "balance_loss_clip": 0.97753978, + "balance_loss_mlp": 1.02157128, + "epoch": 0.44489869536463655, + "flos": 11656209553920.0, + "grad_norm": 2.222224221908839, + "language_loss": 0.81287587, + "learning_rate": 2.450247871302086e-06, + "loss": 0.83451807, + "num_input_tokens_seen": 79716795, + "step": 3700, + "time_per_iteration": 2.6736550331115723 + }, + { + "auxiliary_loss_clip": 0.01167335, + "auxiliary_loss_mlp": 0.01030679, + "balance_loss_clip": 1.01890659, + "balance_loss_mlp": 1.02294183, + "epoch": 0.44501893825527566, + "flos": 20448469958400.0, + "grad_norm": 2.414845633797962, + "language_loss": 0.83231425, + "learning_rate": 2.44948886161922e-06, + "loss": 0.85429442, + "num_input_tokens_seen": 79735810, + "step": 3701, + "time_per_iteration": 2.670011520385742 + }, + { + "auxiliary_loss_clip": 0.01165229, + "auxiliary_loss_mlp": 0.01025999, + "balance_loss_clip": 1.0186398, + "balance_loss_mlp": 1.01824486, + "epoch": 0.4451391811459148, + "flos": 18261079430400.0, + "grad_norm": 1.586960856139011, + "language_loss": 0.84735519, + "learning_rate": 2.4487297837492524e-06, + "loss": 0.86926746, + "num_input_tokens_seen": 79754975, + "step": 3702, + "time_per_iteration": 2.65261173248291 + }, + { + "auxiliary_loss_clip": 0.01115678, + "auxiliary_loss_mlp": 0.01023963, + "balance_loss_clip": 0.9437983, + "balance_loss_mlp": 1.01544547, + "epoch": 0.44525942403655383, + "flos": 16910155895040.0, + "grad_norm": 1.945498861001896, + "language_loss": 0.61914396, + "learning_rate": 2.4479706378073323e-06, + "loss": 0.64054036, + "num_input_tokens_seen": 79773515, + "step": 3703, + "time_per_iteration": 2.668605089187622 + }, + { + "auxiliary_loss_clip": 0.01105612, + "auxiliary_loss_mlp": 0.01031351, + "balance_loss_clip": 0.9371233, + "balance_loss_mlp": 1.02340293, + "epoch": 0.44537966692719294, + "flos": 23258838994560.0, + "grad_norm": 1.7237825794144683, + "language_loss": 0.83975732, + "learning_rate": 2.447211423908623e-06, + "loss": 0.8611269, + "num_input_tokens_seen": 79793560, + "step": 3704, + "time_per_iteration": 2.7413578033447266 + }, + { + "auxiliary_loss_clip": 0.01163947, + "auxiliary_loss_mlp": 0.01024859, + "balance_loss_clip": 1.01855826, + "balance_loss_mlp": 1.01685429, + "epoch": 0.445499909817832, + "flos": 21724160457600.0, + "grad_norm": 3.131933224706125, + "language_loss": 0.74946672, + "learning_rate": 2.4464521421682966e-06, + "loss": 0.77135479, + "num_input_tokens_seen": 79811150, + "step": 3705, + "time_per_iteration": 2.622802495956421 + }, + { + "auxiliary_loss_clip": 0.01157068, + "auxiliary_loss_mlp": 0.01026387, + "balance_loss_clip": 1.01754141, + "balance_loss_mlp": 1.018996, + "epoch": 0.4456201527084711, + "flos": 23987969170560.0, + "grad_norm": 2.0593445439255387, + "language_loss": 0.87702334, + "learning_rate": 2.4456927927015345e-06, + "loss": 0.89885783, + "num_input_tokens_seen": 79832190, + "step": 3706, + "time_per_iteration": 2.707369327545166 + }, + { + "auxiliary_loss_clip": 0.01138234, + "auxiliary_loss_mlp": 0.01035031, + "balance_loss_clip": 0.97743452, + "balance_loss_mlp": 1.02562499, + "epoch": 0.4457403955991102, + "flos": 18807065136000.0, + "grad_norm": 2.0017825052002736, + "language_loss": 0.76028603, + "learning_rate": 2.4449333756235307e-06, + "loss": 0.78201866, + "num_input_tokens_seen": 79848905, + "step": 3707, + "time_per_iteration": 3.524343490600586 + }, + { + "auxiliary_loss_clip": 0.01164995, + "auxiliary_loss_mlp": 0.01026966, + "balance_loss_clip": 1.01891267, + "balance_loss_mlp": 1.0190742, + "epoch": 0.4458606384897493, + "flos": 19207756327680.0, + "grad_norm": 2.1380151285456277, + "language_loss": 0.78857231, + "learning_rate": 2.4441738910494876e-06, + "loss": 0.81049192, + "num_input_tokens_seen": 79863640, + "step": 3708, + "time_per_iteration": 3.5851171016693115 + }, + { + "auxiliary_loss_clip": 0.01144436, + "auxiliary_loss_mlp": 0.01026802, + "balance_loss_clip": 0.97955513, + "balance_loss_mlp": 1.018857, + "epoch": 0.4459808813803884, + "flos": 21361283308800.0, + "grad_norm": 2.2746814844131023, + "language_loss": 0.82306582, + "learning_rate": 2.4434143390946176e-06, + "loss": 0.84477818, + "num_input_tokens_seen": 79882450, + "step": 3709, + "time_per_iteration": 3.6198534965515137 + }, + { + "auxiliary_loss_clip": 0.01113489, + "auxiliary_loss_mlp": 0.01024971, + "balance_loss_clip": 0.94074619, + "balance_loss_mlp": 1.01740098, + "epoch": 0.4461011242710275, + "flos": 23288967527040.0, + "grad_norm": 2.04726340945454, + "language_loss": 0.85377324, + "learning_rate": 2.4426547198741457e-06, + "loss": 0.87515783, + "num_input_tokens_seen": 79900655, + "step": 3710, + "time_per_iteration": 2.693223476409912 + }, + { + "auxiliary_loss_clip": 0.01096126, + "auxiliary_loss_mlp": 0.01030257, + "balance_loss_clip": 0.90977168, + "balance_loss_mlp": 1.02181077, + "epoch": 0.44622136716166655, + "flos": 20193001453440.0, + "grad_norm": 2.3052552639985104, + "language_loss": 0.74760348, + "learning_rate": 2.441895033503305e-06, + "loss": 0.76886731, + "num_input_tokens_seen": 79918575, + "step": 3711, + "time_per_iteration": 2.8133394718170166 + }, + { + "auxiliary_loss_clip": 0.01160761, + "auxiliary_loss_mlp": 0.01026793, + "balance_loss_clip": 1.01904821, + "balance_loss_mlp": 1.0188303, + "epoch": 0.44634161005230566, + "flos": 21283033530240.0, + "grad_norm": 1.810840503688116, + "language_loss": 0.81968808, + "learning_rate": 2.4411352800973375e-06, + "loss": 0.8415637, + "num_input_tokens_seen": 79937010, + "step": 3712, + "time_per_iteration": 2.7056961059570312 + }, + { + "auxiliary_loss_clip": 0.01109838, + "auxiliary_loss_mlp": 0.0103155, + "balance_loss_clip": 0.93924367, + "balance_loss_mlp": 1.02334833, + "epoch": 0.44646185294294477, + "flos": 22929358515840.0, + "grad_norm": 2.3794091554998342, + "language_loss": 0.75482762, + "learning_rate": 2.4403754597715005e-06, + "loss": 0.77624142, + "num_input_tokens_seen": 79956455, + "step": 3713, + "time_per_iteration": 3.5996267795562744 + }, + { + "auxiliary_loss_clip": 0.01140583, + "auxiliary_loss_mlp": 0.01027807, + "balance_loss_clip": 0.97674727, + "balance_loss_mlp": 1.01940894, + "epoch": 0.4465820958335838, + "flos": 22637692080000.0, + "grad_norm": 2.1743564378623774, + "language_loss": 0.92768645, + "learning_rate": 2.4396155726410553e-06, + "loss": 0.94937038, + "num_input_tokens_seen": 79975065, + "step": 3714, + "time_per_iteration": 2.709306478500366 + }, + { + "auxiliary_loss_clip": 0.01168711, + "auxiliary_loss_mlp": 0.01024477, + "balance_loss_clip": 1.01930439, + "balance_loss_mlp": 1.01609707, + "epoch": 0.44670233872422294, + "flos": 22672525294080.0, + "grad_norm": 2.6578404744540003, + "language_loss": 0.90620822, + "learning_rate": 2.438855618821278e-06, + "loss": 0.9281401, + "num_input_tokens_seen": 79990865, + "step": 3715, + "time_per_iteration": 2.6290969848632812 + }, + { + "auxiliary_loss_clip": 0.01152961, + "auxiliary_loss_mlp": 0.01027375, + "balance_loss_clip": 1.01394773, + "balance_loss_mlp": 1.01971555, + "epoch": 0.44682258161486205, + "flos": 23582178247680.0, + "grad_norm": 1.6374320581193806, + "language_loss": 0.67635047, + "learning_rate": 2.4380955984274517e-06, + "loss": 0.69815379, + "num_input_tokens_seen": 80009520, + "step": 3716, + "time_per_iteration": 2.659066677093506 + }, + { + "auxiliary_loss_clip": 0.0116178, + "auxiliary_loss_mlp": 0.01030051, + "balance_loss_clip": 1.01786351, + "balance_loss_mlp": 1.0217185, + "epoch": 0.4469428245055011, + "flos": 26501356558080.0, + "grad_norm": 14.559835342548649, + "language_loss": 0.77076083, + "learning_rate": 2.4373355115748716e-06, + "loss": 0.79267913, + "num_input_tokens_seen": 80030350, + "step": 3717, + "time_per_iteration": 2.686976909637451 + }, + { + "auxiliary_loss_clip": 0.01130898, + "auxiliary_loss_mlp": 0.01028719, + "balance_loss_clip": 0.97834229, + "balance_loss_mlp": 1.02105999, + "epoch": 0.4470630673961402, + "flos": 21504925797120.0, + "grad_norm": 1.7737016487339379, + "language_loss": 0.72232658, + "learning_rate": 2.436575358378842e-06, + "loss": 0.74392271, + "num_input_tokens_seen": 80049840, + "step": 3718, + "time_per_iteration": 2.683420419692993 + }, + { + "auxiliary_loss_clip": 0.01147136, + "auxiliary_loss_mlp": 0.01028608, + "balance_loss_clip": 0.98078424, + "balance_loss_mlp": 1.02003646, + "epoch": 0.44718331028677927, + "flos": 16173986653440.0, + "grad_norm": 3.103788956463239, + "language_loss": 0.82786632, + "learning_rate": 2.4358151389546782e-06, + "loss": 0.84962368, + "num_input_tokens_seen": 80066525, + "step": 3719, + "time_per_iteration": 2.62312912940979 + }, + { + "auxiliary_loss_clip": 0.0118835, + "auxiliary_loss_mlp": 0.01026781, + "balance_loss_clip": 1.05848706, + "balance_loss_mlp": 1.01940215, + "epoch": 0.4473035531774184, + "flos": 19681238430720.0, + "grad_norm": 2.316546968480545, + "language_loss": 0.75910819, + "learning_rate": 2.4350548534177035e-06, + "loss": 0.78125948, + "num_input_tokens_seen": 80083355, + "step": 3720, + "time_per_iteration": 2.6144042015075684 + }, + { + "auxiliary_loss_clip": 0.01114217, + "auxiliary_loss_mlp": 0.01032044, + "balance_loss_clip": 0.9402554, + "balance_loss_mlp": 1.02430773, + "epoch": 0.4474237960680575, + "flos": 41427590515200.0, + "grad_norm": 1.5488162504905392, + "language_loss": 0.6704278, + "learning_rate": 2.434294501883254e-06, + "loss": 0.69189036, + "num_input_tokens_seen": 80106450, + "step": 3721, + "time_per_iteration": 2.895146131515503 + }, + { + "auxiliary_loss_clip": 0.01129275, + "auxiliary_loss_mlp": 0.01024656, + "balance_loss_clip": 0.97710675, + "balance_loss_mlp": 1.01699686, + "epoch": 0.44754403895869654, + "flos": 22891328991360.0, + "grad_norm": 1.6381534712932873, + "language_loss": 0.65866435, + "learning_rate": 2.433534084466674e-06, + "loss": 0.68020368, + "num_input_tokens_seen": 80125670, + "step": 3722, + "time_per_iteration": 2.6888906955718994 + }, + { + "auxiliary_loss_clip": 0.01183748, + "auxiliary_loss_mlp": 0.01030944, + "balance_loss_clip": 1.05665541, + "balance_loss_mlp": 1.02327895, + "epoch": 0.44766428184933565, + "flos": 25630271832960.0, + "grad_norm": 1.4640217433583582, + "language_loss": 0.7079469, + "learning_rate": 2.4327736012833178e-06, + "loss": 0.73009384, + "num_input_tokens_seen": 80147390, + "step": 3723, + "time_per_iteration": 2.6833438873291016 + }, + { + "auxiliary_loss_clip": 0.01160494, + "auxiliary_loss_mlp": 0.0103151, + "balance_loss_clip": 1.01897871, + "balance_loss_mlp": 1.02270031, + "epoch": 0.44778452473997477, + "flos": 20448972748800.0, + "grad_norm": 2.0373646995214836, + "language_loss": 0.7646113, + "learning_rate": 2.4320130524485506e-06, + "loss": 0.78653133, + "num_input_tokens_seen": 80166185, + "step": 3724, + "time_per_iteration": 2.61226224899292 + }, + { + "auxiliary_loss_clip": 0.01135285, + "auxiliary_loss_mlp": 0.01032574, + "balance_loss_clip": 0.98556983, + "balance_loss_mlp": 1.02493894, + "epoch": 0.4479047676306138, + "flos": 21975462984960.0, + "grad_norm": 1.7093926077346455, + "language_loss": 0.7969594, + "learning_rate": 2.431252438077746e-06, + "loss": 0.81863797, + "num_input_tokens_seen": 80185685, + "step": 3725, + "time_per_iteration": 2.67952561378479 + }, + { + "auxiliary_loss_clip": 0.01165836, + "auxiliary_loss_mlp": 0.005108, + "balance_loss_clip": 1.01921427, + "balance_loss_mlp": 0.0, + "epoch": 0.44802501052125293, + "flos": 21467219495040.0, + "grad_norm": 2.6324993733047877, + "language_loss": 0.77225757, + "learning_rate": 2.4304917582862906e-06, + "loss": 0.78902388, + "num_input_tokens_seen": 80204865, + "step": 3726, + "time_per_iteration": 2.6249399185180664 + }, + { + "auxiliary_loss_clip": 0.01186582, + "auxiliary_loss_mlp": 0.01026802, + "balance_loss_clip": 1.05764771, + "balance_loss_mlp": 1.01896429, + "epoch": 0.44814525341189204, + "flos": 22126970551680.0, + "grad_norm": 2.09518884041632, + "language_loss": 0.87860024, + "learning_rate": 2.4297310131895774e-06, + "loss": 0.90073407, + "num_input_tokens_seen": 80223410, + "step": 3727, + "time_per_iteration": 2.64437198638916 + }, + { + "auxiliary_loss_clip": 0.01162872, + "auxiliary_loss_mlp": 0.0102794, + "balance_loss_clip": 1.0193429, + "balance_loss_mlp": 1.01952422, + "epoch": 0.4482654963025311, + "flos": 16653933204480.0, + "grad_norm": 2.3605483736444928, + "language_loss": 0.74541497, + "learning_rate": 2.4289702029030113e-06, + "loss": 0.76732314, + "num_input_tokens_seen": 80240880, + "step": 3728, + "time_per_iteration": 2.582261800765991 + }, + { + "auxiliary_loss_clip": 0.01163032, + "auxiliary_loss_mlp": 0.01031947, + "balance_loss_clip": 1.02054882, + "balance_loss_mlp": 1.02369213, + "epoch": 0.4483857391931702, + "flos": 18841251905280.0, + "grad_norm": 2.112462549112213, + "language_loss": 0.83356649, + "learning_rate": 2.4282093275420057e-06, + "loss": 0.85551631, + "num_input_tokens_seen": 80259910, + "step": 3729, + "time_per_iteration": 2.621079683303833 + }, + { + "auxiliary_loss_clip": 0.0116853, + "auxiliary_loss_mlp": 0.0102268, + "balance_loss_clip": 1.02109742, + "balance_loss_mlp": 1.01456809, + "epoch": 0.4485059820838093, + "flos": 20372590477440.0, + "grad_norm": 10.023136945732796, + "language_loss": 0.70895827, + "learning_rate": 2.4274483872219863e-06, + "loss": 0.73087043, + "num_input_tokens_seen": 80277270, + "step": 3730, + "time_per_iteration": 2.5781092643737793 + }, + { + "auxiliary_loss_clip": 0.01159934, + "auxiliary_loss_mlp": 0.01029195, + "balance_loss_clip": 1.01797962, + "balance_loss_mlp": 1.02136922, + "epoch": 0.4486262249744484, + "flos": 20047742853120.0, + "grad_norm": 1.6826489624392202, + "language_loss": 0.93895954, + "learning_rate": 2.426687382058386e-06, + "loss": 0.96085089, + "num_input_tokens_seen": 80295550, + "step": 3731, + "time_per_iteration": 2.692671537399292 + }, + { + "auxiliary_loss_clip": 0.01064323, + "auxiliary_loss_mlp": 0.01003498, + "balance_loss_clip": 0.99136072, + "balance_loss_mlp": 1.002038, + "epoch": 0.4487464678650875, + "flos": 64595684776320.0, + "grad_norm": 0.8648913530762741, + "language_loss": 0.59836781, + "learning_rate": 2.425926312166649e-06, + "loss": 0.61904597, + "num_input_tokens_seen": 80348425, + "step": 3732, + "time_per_iteration": 3.067180633544922 + }, + { + "auxiliary_loss_clip": 0.01143146, + "auxiliary_loss_mlp": 0.0103423, + "balance_loss_clip": 0.98162043, + "balance_loss_mlp": 1.02553988, + "epoch": 0.4488667107557266, + "flos": 20769798049920.0, + "grad_norm": 2.0461571503419322, + "language_loss": 0.73250526, + "learning_rate": 2.42516517766223e-06, + "loss": 0.75427902, + "num_input_tokens_seen": 80366505, + "step": 3733, + "time_per_iteration": 3.5169994831085205 + }, + { + "auxiliary_loss_clip": 0.01187736, + "auxiliary_loss_mlp": 0.01030021, + "balance_loss_clip": 1.06011903, + "balance_loss_mlp": 1.021909, + "epoch": 0.44898695364636565, + "flos": 23951735326080.0, + "grad_norm": 2.0864348632024132, + "language_loss": 0.67895055, + "learning_rate": 2.4244039786605907e-06, + "loss": 0.70112818, + "num_input_tokens_seen": 80387510, + "step": 3734, + "time_per_iteration": 3.59462571144104 + }, + { + "auxiliary_loss_clip": 0.01094616, + "auxiliary_loss_mlp": 0.0102927, + "balance_loss_clip": 0.90337098, + "balance_loss_mlp": 1.0211221, + "epoch": 0.44910719653700476, + "flos": 18624351628800.0, + "grad_norm": 2.567976754516925, + "language_loss": 0.82833683, + "learning_rate": 2.4236427152772055e-06, + "loss": 0.84957576, + "num_input_tokens_seen": 80405915, + "step": 3735, + "time_per_iteration": 2.712400197982788 + }, + { + "auxiliary_loss_clip": 0.01009229, + "auxiliary_loss_mlp": 0.01001993, + "balance_loss_clip": 0.91112173, + "balance_loss_mlp": 1.00038421, + "epoch": 0.4492274394276438, + "flos": 57033435749760.0, + "grad_norm": 0.819989163943731, + "language_loss": 0.57289732, + "learning_rate": 2.422881387627557e-06, + "loss": 0.59300959, + "num_input_tokens_seen": 80458365, + "step": 3736, + "time_per_iteration": 3.9049153327941895 + }, + { + "auxiliary_loss_clip": 0.01145035, + "auxiliary_loss_mlp": 0.01026155, + "balance_loss_clip": 0.98381811, + "balance_loss_mlp": 1.01825786, + "epoch": 0.4493476823182829, + "flos": 23254888498560.0, + "grad_norm": 1.7658855232392314, + "language_loss": 0.773808, + "learning_rate": 2.422119995827139e-06, + "loss": 0.79551995, + "num_input_tokens_seen": 80478490, + "step": 3737, + "time_per_iteration": 2.7220664024353027 + }, + { + "auxiliary_loss_clip": 0.01168937, + "auxiliary_loss_mlp": 0.01028672, + "balance_loss_clip": 1.02271807, + "balance_loss_mlp": 1.02034569, + "epoch": 0.44946792520892204, + "flos": 15815131827840.0, + "grad_norm": 3.3848010313052104, + "language_loss": 0.73996115, + "learning_rate": 2.4213585399914528e-06, + "loss": 0.76193726, + "num_input_tokens_seen": 80495695, + "step": 3738, + "time_per_iteration": 2.6069202423095703 + }, + { + "auxiliary_loss_clip": 0.01162565, + "auxiliary_loss_mlp": 0.01028914, + "balance_loss_clip": 1.02164948, + "balance_loss_mlp": 1.02092123, + "epoch": 0.4495881680995611, + "flos": 19610063631360.0, + "grad_norm": 2.112030774571606, + "language_loss": 0.85738629, + "learning_rate": 2.4205970202360113e-06, + "loss": 0.87930107, + "num_input_tokens_seen": 80515260, + "step": 3739, + "time_per_iteration": 3.520059585571289 + }, + { + "auxiliary_loss_clip": 0.01079127, + "auxiliary_loss_mlp": 0.0103157, + "balance_loss_clip": 0.89935201, + "balance_loss_mlp": 1.02332687, + "epoch": 0.4497084109902002, + "flos": 26031465815040.0, + "grad_norm": 2.7920055736742615, + "language_loss": 0.77883482, + "learning_rate": 2.4198354366763354e-06, + "loss": 0.79994178, + "num_input_tokens_seen": 80533900, + "step": 3740, + "time_per_iteration": 2.768030881881714 + }, + { + "auxiliary_loss_clip": 0.01143561, + "auxiliary_loss_mlp": 0.01031991, + "balance_loss_clip": 0.98310596, + "balance_loss_mlp": 1.02354562, + "epoch": 0.4498286538808393, + "flos": 14793688771200.0, + "grad_norm": 1.9953600131980742, + "language_loss": 0.78691959, + "learning_rate": 2.4190737894279587e-06, + "loss": 0.80867517, + "num_input_tokens_seen": 80551270, + "step": 3741, + "time_per_iteration": 2.6589457988739014 + }, + { + "auxiliary_loss_clip": 0.01103275, + "auxiliary_loss_mlp": 0.01026404, + "balance_loss_clip": 0.9373185, + "balance_loss_mlp": 1.01829767, + "epoch": 0.44994889677147837, + "flos": 15450171690240.0, + "grad_norm": 2.1464826682339533, + "language_loss": 0.80578828, + "learning_rate": 2.4183120786064203e-06, + "loss": 0.82708508, + "num_input_tokens_seen": 80568145, + "step": 3742, + "time_per_iteration": 2.8174901008605957 + }, + { + "auxiliary_loss_clip": 0.01166373, + "auxiliary_loss_mlp": 0.00509659, + "balance_loss_clip": 1.0222652, + "balance_loss_mlp": 0.0, + "epoch": 0.4500691396621175, + "flos": 21798316085760.0, + "grad_norm": 6.505202714185579, + "language_loss": 0.85452056, + "learning_rate": 2.417550304327273e-06, + "loss": 0.87128091, + "num_input_tokens_seen": 80586185, + "step": 3743, + "time_per_iteration": 2.628105878829956 + }, + { + "auxiliary_loss_clip": 0.01190968, + "auxiliary_loss_mlp": 0.0103024, + "balance_loss_clip": 1.05997086, + "balance_loss_mlp": 1.02179432, + "epoch": 0.4501893825527566, + "flos": 32382016421760.0, + "grad_norm": 1.6327972031502826, + "language_loss": 0.75494921, + "learning_rate": 2.4167884667060763e-06, + "loss": 0.77716136, + "num_input_tokens_seen": 80608895, + "step": 3744, + "time_per_iteration": 2.7238075733184814 + }, + { + "auxiliary_loss_clip": 0.01142404, + "auxiliary_loss_mlp": 0.01036751, + "balance_loss_clip": 0.98147953, + "balance_loss_mlp": 1.02844787, + "epoch": 0.45030962544339564, + "flos": 16544944362240.0, + "grad_norm": 2.389670790767816, + "language_loss": 0.87038779, + "learning_rate": 2.4160265658584e-06, + "loss": 0.89217931, + "num_input_tokens_seen": 80623785, + "step": 3745, + "time_per_iteration": 2.627521276473999 + }, + { + "auxiliary_loss_clip": 0.01169039, + "auxiliary_loss_mlp": 0.01029778, + "balance_loss_clip": 1.02285516, + "balance_loss_mlp": 1.0215199, + "epoch": 0.45042986833403476, + "flos": 19573039687680.0, + "grad_norm": 1.9803879866940772, + "language_loss": 0.68456572, + "learning_rate": 2.4152646018998253e-06, + "loss": 0.70655394, + "num_input_tokens_seen": 80642735, + "step": 3746, + "time_per_iteration": 2.6353306770324707 + }, + { + "auxiliary_loss_clip": 0.01163321, + "auxiliary_loss_mlp": 0.01028656, + "balance_loss_clip": 1.02240491, + "balance_loss_mlp": 1.0200671, + "epoch": 0.45055011122467387, + "flos": 23112467072640.0, + "grad_norm": 2.146979757877053, + "language_loss": 0.71956533, + "learning_rate": 2.4145025749459403e-06, + "loss": 0.74148512, + "num_input_tokens_seen": 80663760, + "step": 3747, + "time_per_iteration": 2.6580405235290527 + }, + { + "auxiliary_loss_clip": 0.01048038, + "auxiliary_loss_mlp": 0.01038318, + "balance_loss_clip": 0.83266366, + "balance_loss_mlp": 1.03018165, + "epoch": 0.4506703541153129, + "flos": 19934623946880.0, + "grad_norm": 2.2322711627223883, + "language_loss": 0.69935477, + "learning_rate": 2.413740485112344e-06, + "loss": 0.72021842, + "num_input_tokens_seen": 80682100, + "step": 3748, + "time_per_iteration": 2.792149543762207 + }, + { + "auxiliary_loss_clip": 0.01136948, + "auxiliary_loss_mlp": 0.01033478, + "balance_loss_clip": 0.98542309, + "balance_loss_mlp": 1.02574754, + "epoch": 0.45079059700595203, + "flos": 19499530504320.0, + "grad_norm": 1.6053959963313937, + "language_loss": 0.82409751, + "learning_rate": 2.412978332514646e-06, + "loss": 0.84580183, + "num_input_tokens_seen": 80700880, + "step": 3749, + "time_per_iteration": 2.6416711807250977 + }, + { + "auxiliary_loss_clip": 0.01143697, + "auxiliary_loss_mlp": 0.01029937, + "balance_loss_clip": 0.98269832, + "balance_loss_mlp": 1.0218184, + "epoch": 0.4509108398965911, + "flos": 27636313570560.0, + "grad_norm": 2.651469262328266, + "language_loss": 0.72164261, + "learning_rate": 2.4122161172684623e-06, + "loss": 0.74337894, + "num_input_tokens_seen": 80721675, + "step": 3750, + "time_per_iteration": 2.778202533721924 + }, + { + "auxiliary_loss_clip": 0.0114077, + "auxiliary_loss_mlp": 0.01033468, + "balance_loss_clip": 0.97996771, + "balance_loss_mlp": 1.02450967, + "epoch": 0.4510310827872302, + "flos": 20995712640000.0, + "grad_norm": 2.385020108530067, + "language_loss": 0.84344715, + "learning_rate": 2.4114538394894216e-06, + "loss": 0.86518955, + "num_input_tokens_seen": 80739315, + "step": 3751, + "time_per_iteration": 2.6414196491241455 + }, + { + "auxiliary_loss_clip": 0.01135392, + "auxiliary_loss_mlp": 0.01025709, + "balance_loss_clip": 0.97609639, + "balance_loss_mlp": 1.01838088, + "epoch": 0.4511513256778693, + "flos": 16216684945920.0, + "grad_norm": 2.2559631881646194, + "language_loss": 0.83144569, + "learning_rate": 2.410691499293161e-06, + "loss": 0.85305667, + "num_input_tokens_seen": 80757470, + "step": 3752, + "time_per_iteration": 2.688598394393921 + }, + { + "auxiliary_loss_clip": 0.01161582, + "auxiliary_loss_mlp": 0.01031133, + "balance_loss_clip": 1.01864707, + "balance_loss_mlp": 1.02309227, + "epoch": 0.45127156856850836, + "flos": 25186702780800.0, + "grad_norm": 2.7110997255716685, + "language_loss": 0.74282682, + "learning_rate": 2.409929096795326e-06, + "loss": 0.764754, + "num_input_tokens_seen": 80777840, + "step": 3753, + "time_per_iteration": 2.703594923019409 + }, + { + "auxiliary_loss_clip": 0.01162386, + "auxiliary_loss_mlp": 0.01031344, + "balance_loss_clip": 1.01769567, + "balance_loss_mlp": 1.0221653, + "epoch": 0.4513918114591475, + "flos": 20412523422720.0, + "grad_norm": 2.529068410205332, + "language_loss": 0.79120529, + "learning_rate": 2.409166632111573e-06, + "loss": 0.81314254, + "num_input_tokens_seen": 80795975, + "step": 3754, + "time_per_iteration": 2.6706223487854004 + }, + { + "auxiliary_loss_clip": 0.01170632, + "auxiliary_loss_mlp": 0.0102719, + "balance_loss_clip": 1.01995468, + "balance_loss_mlp": 1.01869678, + "epoch": 0.4515120543497866, + "flos": 26648482665600.0, + "grad_norm": 1.6992756571525245, + "language_loss": 0.8013019, + "learning_rate": 2.4084041053575674e-06, + "loss": 0.82328016, + "num_input_tokens_seen": 80815395, + "step": 3755, + "time_per_iteration": 2.6830267906188965 + }, + { + "auxiliary_loss_clip": 0.01146397, + "auxiliary_loss_mlp": 0.01030162, + "balance_loss_clip": 0.98392916, + "balance_loss_mlp": 1.0223, + "epoch": 0.45163229724042564, + "flos": 20595093275520.0, + "grad_norm": 2.1535473858993663, + "language_loss": 0.72323895, + "learning_rate": 2.4076415166489834e-06, + "loss": 0.74500453, + "num_input_tokens_seen": 80834805, + "step": 3756, + "time_per_iteration": 2.6420276165008545 + }, + { + "auxiliary_loss_clip": 0.01096856, + "auxiliary_loss_mlp": 0.01025286, + "balance_loss_clip": 0.90357846, + "balance_loss_mlp": 1.01783526, + "epoch": 0.45175254013106475, + "flos": 21689004021120.0, + "grad_norm": 1.622300817734096, + "language_loss": 0.7912032, + "learning_rate": 2.406878866101506e-06, + "loss": 0.81242466, + "num_input_tokens_seen": 80853770, + "step": 3757, + "time_per_iteration": 2.7456588745117188 + }, + { + "auxiliary_loss_clip": 0.01186498, + "auxiliary_loss_mlp": 0.01033506, + "balance_loss_clip": 1.05952179, + "balance_loss_mlp": 1.02570391, + "epoch": 0.45187278302170386, + "flos": 18878850466560.0, + "grad_norm": 1.871274333092765, + "language_loss": 0.78589314, + "learning_rate": 2.4061161538308273e-06, + "loss": 0.80809319, + "num_input_tokens_seen": 80870615, + "step": 3758, + "time_per_iteration": 2.5701076984405518 + }, + { + "auxiliary_loss_clip": 0.01162034, + "auxiliary_loss_mlp": 0.0102157, + "balance_loss_clip": 1.02035761, + "balance_loss_mlp": 1.01363027, + "epoch": 0.4519930259123429, + "flos": 18582479349120.0, + "grad_norm": 1.8343275443272564, + "language_loss": 0.88898861, + "learning_rate": 2.4053533799526523e-06, + "loss": 0.9108246, + "num_input_tokens_seen": 80886335, + "step": 3759, + "time_per_iteration": 3.5123562812805176 + }, + { + "auxiliary_loss_clip": 0.01135204, + "auxiliary_loss_mlp": 0.01026976, + "balance_loss_clip": 0.98210305, + "balance_loss_mlp": 1.01876807, + "epoch": 0.452113268802982, + "flos": 25192377129600.0, + "grad_norm": 1.996972355788818, + "language_loss": 0.86409479, + "learning_rate": 2.404590544582691e-06, + "loss": 0.88571656, + "num_input_tokens_seen": 80904570, + "step": 3760, + "time_per_iteration": 3.621382713317871 + }, + { + "auxiliary_loss_clip": 0.01094954, + "auxiliary_loss_mlp": 0.0102889, + "balance_loss_clip": 0.90083337, + "balance_loss_mlp": 1.02042627, + "epoch": 0.45223351169362114, + "flos": 39378922312320.0, + "grad_norm": 1.781681298067398, + "language_loss": 0.81474781, + "learning_rate": 2.403827647836666e-06, + "loss": 0.83598626, + "num_input_tokens_seen": 80925125, + "step": 3761, + "time_per_iteration": 3.798086404800415 + }, + { + "auxiliary_loss_clip": 0.0118632, + "auxiliary_loss_mlp": 0.0102989, + "balance_loss_clip": 1.05564976, + "balance_loss_mlp": 1.02155757, + "epoch": 0.4523537545842602, + "flos": 21582169994880.0, + "grad_norm": 2.1175145573420973, + "language_loss": 0.69479167, + "learning_rate": 2.4030646898303075e-06, + "loss": 0.71695387, + "num_input_tokens_seen": 80946615, + "step": 3762, + "time_per_iteration": 2.618457317352295 + }, + { + "auxiliary_loss_clip": 0.01145291, + "auxiliary_loss_mlp": 0.01032287, + "balance_loss_clip": 0.98200345, + "balance_loss_mlp": 1.02393055, + "epoch": 0.4524739974748993, + "flos": 28439527547520.0, + "grad_norm": 2.101672207669138, + "language_loss": 0.8194406, + "learning_rate": 2.4023016706793566e-06, + "loss": 0.84121639, + "num_input_tokens_seen": 80966410, + "step": 3763, + "time_per_iteration": 2.7249748706817627 + }, + { + "auxiliary_loss_clip": 0.01038332, + "auxiliary_loss_mlp": 0.01021322, + "balance_loss_clip": 0.92230618, + "balance_loss_mlp": 1.01977217, + "epoch": 0.4525942403655384, + "flos": 61556492148480.0, + "grad_norm": 0.792941591191313, + "language_loss": 0.56856298, + "learning_rate": 2.401538590499561e-06, + "loss": 0.58915949, + "num_input_tokens_seen": 81026865, + "step": 3764, + "time_per_iteration": 3.359478235244751 + }, + { + "auxiliary_loss_clip": 0.01167478, + "auxiliary_loss_mlp": 0.0051031, + "balance_loss_clip": 1.02135694, + "balance_loss_mlp": 0.0, + "epoch": 0.45271448325617747, + "flos": 27529838680320.0, + "grad_norm": 1.7821807696629584, + "language_loss": 0.7163592, + "learning_rate": 2.400775449406682e-06, + "loss": 0.73313707, + "num_input_tokens_seen": 81050060, + "step": 3765, + "time_per_iteration": 3.6221776008605957 + }, + { + "auxiliary_loss_clip": 0.01161986, + "auxiliary_loss_mlp": 0.01026557, + "balance_loss_clip": 1.01627779, + "balance_loss_mlp": 1.01854873, + "epoch": 0.4528347261468166, + "flos": 22452608275200.0, + "grad_norm": 23.12104457532337, + "language_loss": 0.72735012, + "learning_rate": 2.400012247516485e-06, + "loss": 0.74923557, + "num_input_tokens_seen": 81070625, + "step": 3766, + "time_per_iteration": 2.7032816410064697 + }, + { + "auxiliary_loss_clip": 0.01121387, + "auxiliary_loss_mlp": 0.01031258, + "balance_loss_clip": 0.94308043, + "balance_loss_mlp": 1.02354527, + "epoch": 0.45295496903745563, + "flos": 21103875469440.0, + "grad_norm": 5.396514876842095, + "language_loss": 0.90334094, + "learning_rate": 2.3992489849447484e-06, + "loss": 0.92486739, + "num_input_tokens_seen": 81089080, + "step": 3767, + "time_per_iteration": 2.7251858711242676 + }, + { + "auxiliary_loss_clip": 0.01126054, + "auxiliary_loss_mlp": 0.01032022, + "balance_loss_clip": 0.94489586, + "balance_loss_mlp": 1.02395785, + "epoch": 0.45307521192809475, + "flos": 23221168606080.0, + "grad_norm": 1.5576407975818687, + "language_loss": 0.79146791, + "learning_rate": 2.3984856618072584e-06, + "loss": 0.81304872, + "num_input_tokens_seen": 81109115, + "step": 3768, + "time_per_iteration": 2.6788856983184814 + }, + { + "auxiliary_loss_clip": 0.01122386, + "auxiliary_loss_mlp": 0.01030791, + "balance_loss_clip": 0.94308478, + "balance_loss_mlp": 1.02255917, + "epoch": 0.45319545481873386, + "flos": 15560094286080.0, + "grad_norm": 3.656282715326512, + "language_loss": 0.74129057, + "learning_rate": 2.3977222782198098e-06, + "loss": 0.76282233, + "num_input_tokens_seen": 81127750, + "step": 3769, + "time_per_iteration": 2.6791892051696777 + }, + { + "auxiliary_loss_clip": 0.01108698, + "auxiliary_loss_mlp": 0.01026335, + "balance_loss_clip": 0.94096887, + "balance_loss_mlp": 1.0181458, + "epoch": 0.4533156977093729, + "flos": 21944759834880.0, + "grad_norm": 1.7818321681494098, + "language_loss": 0.75550872, + "learning_rate": 2.3969588342982077e-06, + "loss": 0.77685905, + "num_input_tokens_seen": 81147125, + "step": 3770, + "time_per_iteration": 2.710920810699463 + }, + { + "auxiliary_loss_clip": 0.01164466, + "auxiliary_loss_mlp": 0.01027867, + "balance_loss_clip": 1.02265346, + "balance_loss_mlp": 1.01937938, + "epoch": 0.453435940600012, + "flos": 24242180699520.0, + "grad_norm": 1.9314053154516504, + "language_loss": 0.72621119, + "learning_rate": 2.396195330158267e-06, + "loss": 0.74813449, + "num_input_tokens_seen": 81167015, + "step": 3771, + "time_per_iteration": 2.650973320007324 + }, + { + "auxiliary_loss_clip": 0.01189536, + "auxiliary_loss_mlp": 0.01028028, + "balance_loss_clip": 1.05926991, + "balance_loss_mlp": 1.01957631, + "epoch": 0.45355618349065113, + "flos": 23440367352960.0, + "grad_norm": 1.9229851602771157, + "language_loss": 0.79319191, + "learning_rate": 2.3954317659158094e-06, + "loss": 0.81536758, + "num_input_tokens_seen": 81187350, + "step": 3772, + "time_per_iteration": 2.6171157360076904 + }, + { + "auxiliary_loss_clip": 0.01093283, + "auxiliary_loss_mlp": 0.0100369, + "balance_loss_clip": 1.03334498, + "balance_loss_mlp": 1.00237906, + "epoch": 0.4536764263812902, + "flos": 66903161448960.0, + "grad_norm": 0.8935432051818305, + "language_loss": 0.56924403, + "learning_rate": 2.394668141686667e-06, + "loss": 0.59021378, + "num_input_tokens_seen": 81249315, + "step": 3773, + "time_per_iteration": 3.2036709785461426 + }, + { + "auxiliary_loss_clip": 0.01159193, + "auxiliary_loss_mlp": 0.01025534, + "balance_loss_clip": 1.01783776, + "balance_loss_mlp": 1.01725459, + "epoch": 0.4537966692719293, + "flos": 42739766254080.0, + "grad_norm": 2.3933307652596247, + "language_loss": 0.70307589, + "learning_rate": 2.3939044575866813e-06, + "loss": 0.72492313, + "num_input_tokens_seen": 81272065, + "step": 3774, + "time_per_iteration": 2.8495306968688965 + }, + { + "auxiliary_loss_clip": 0.01137538, + "auxiliary_loss_mlp": 0.00510437, + "balance_loss_clip": 0.98083496, + "balance_loss_mlp": 0.0, + "epoch": 0.4539169121625684, + "flos": 35549480517120.0, + "grad_norm": 7.550249913985942, + "language_loss": 0.75314593, + "learning_rate": 2.3931407137317024e-06, + "loss": 0.76962566, + "num_input_tokens_seen": 81292220, + "step": 3775, + "time_per_iteration": 2.8139240741729736 + }, + { + "auxiliary_loss_clip": 0.01117558, + "auxiliary_loss_mlp": 0.0103106, + "balance_loss_clip": 0.94212639, + "balance_loss_mlp": 1.02252412, + "epoch": 0.45403715505320746, + "flos": 18514716341760.0, + "grad_norm": 1.8850864898633624, + "language_loss": 0.84814721, + "learning_rate": 2.3923769102375907e-06, + "loss": 0.86963338, + "num_input_tokens_seen": 81311085, + "step": 3776, + "time_per_iteration": 2.6872196197509766 + }, + { + "auxiliary_loss_clip": 0.01117374, + "auxiliary_loss_mlp": 0.01032217, + "balance_loss_clip": 0.94271445, + "balance_loss_mlp": 1.02358592, + "epoch": 0.4541573979438466, + "flos": 25045825639680.0, + "grad_norm": 1.9316207207235692, + "language_loss": 0.78669113, + "learning_rate": 2.391613047220213e-06, + "loss": 0.80818701, + "num_input_tokens_seen": 81330985, + "step": 3777, + "time_per_iteration": 2.7185661792755127 + }, + { + "auxiliary_loss_clip": 0.01099776, + "auxiliary_loss_mlp": 0.01033675, + "balance_loss_clip": 0.90352869, + "balance_loss_mlp": 1.0242697, + "epoch": 0.4542776408344857, + "flos": 18332397884160.0, + "grad_norm": 1.9096252956348778, + "language_loss": 0.78715861, + "learning_rate": 2.390849124795447e-06, + "loss": 0.80849308, + "num_input_tokens_seen": 81346985, + "step": 3778, + "time_per_iteration": 2.7815446853637695 + }, + { + "auxiliary_loss_clip": 0.01192212, + "auxiliary_loss_mlp": 0.01030967, + "balance_loss_clip": 1.0607183, + "balance_loss_mlp": 1.02308726, + "epoch": 0.45439788372512474, + "flos": 20701173116160.0, + "grad_norm": 2.286524807173601, + "language_loss": 0.83927417, + "learning_rate": 2.3900851430791804e-06, + "loss": 0.86150593, + "num_input_tokens_seen": 81365005, + "step": 3779, + "time_per_iteration": 2.626739263534546 + }, + { + "auxiliary_loss_clip": 0.01191461, + "auxiliary_loss_mlp": 0.0103979, + "balance_loss_clip": 1.05942523, + "balance_loss_mlp": 1.03067005, + "epoch": 0.45451812661576385, + "flos": 22309432663680.0, + "grad_norm": 2.268520012656758, + "language_loss": 0.84910345, + "learning_rate": 2.389321102187307e-06, + "loss": 0.87141591, + "num_input_tokens_seen": 81383785, + "step": 3780, + "time_per_iteration": 2.5988283157348633 + }, + { + "auxiliary_loss_clip": 0.01142935, + "auxiliary_loss_mlp": 0.00510995, + "balance_loss_clip": 0.98253518, + "balance_loss_mlp": 0.0, + "epoch": 0.4546383695064029, + "flos": 21763303303680.0, + "grad_norm": 2.847992396294831, + "language_loss": 0.81566453, + "learning_rate": 2.3885570022357326e-06, + "loss": 0.83220381, + "num_input_tokens_seen": 81402915, + "step": 3781, + "time_per_iteration": 2.682328224182129 + }, + { + "auxiliary_loss_clip": 0.01039302, + "auxiliary_loss_mlp": 0.01002187, + "balance_loss_clip": 0.92184126, + "balance_loss_mlp": 1.00082803, + "epoch": 0.454758612397042, + "flos": 64242755694720.0, + "grad_norm": 0.7972021451133241, + "language_loss": 0.6092484, + "learning_rate": 2.38779284334037e-06, + "loss": 0.62966323, + "num_input_tokens_seen": 81467890, + "step": 3782, + "time_per_iteration": 3.313845634460449 + }, + { + "auxiliary_loss_clip": 0.01077192, + "auxiliary_loss_mlp": 0.01026487, + "balance_loss_clip": 0.89923763, + "balance_loss_mlp": 1.01847601, + "epoch": 0.4548788552876811, + "flos": 27304175485440.0, + "grad_norm": 3.567682700538114, + "language_loss": 0.79072189, + "learning_rate": 2.387028625617141e-06, + "loss": 0.81175864, + "num_input_tokens_seen": 81487105, + "step": 3783, + "time_per_iteration": 2.8019142150878906 + }, + { + "auxiliary_loss_clip": 0.01129958, + "auxiliary_loss_mlp": 0.01030591, + "balance_loss_clip": 0.97888005, + "balance_loss_mlp": 1.02245474, + "epoch": 0.4549990981783202, + "flos": 22857142222080.0, + "grad_norm": 1.8880119669971003, + "language_loss": 0.84852201, + "learning_rate": 2.3862643491819766e-06, + "loss": 0.8701275, + "num_input_tokens_seen": 81505670, + "step": 3784, + "time_per_iteration": 2.6827876567840576 + }, + { + "auxiliary_loss_clip": 0.01159702, + "auxiliary_loss_mlp": 0.01029153, + "balance_loss_clip": 1.01634741, + "balance_loss_mlp": 1.02101684, + "epoch": 0.4551193410689593, + "flos": 23258587599360.0, + "grad_norm": 1.8588367312031013, + "language_loss": 0.8427155, + "learning_rate": 2.3855000141508186e-06, + "loss": 0.86460406, + "num_input_tokens_seen": 81525825, + "step": 3785, + "time_per_iteration": 3.5940072536468506 + }, + { + "auxiliary_loss_clip": 0.01146105, + "auxiliary_loss_mlp": 0.01025509, + "balance_loss_clip": 0.98684812, + "balance_loss_mlp": 1.0172596, + "epoch": 0.4552395839595984, + "flos": 20777519473920.0, + "grad_norm": 2.0732227907454917, + "language_loss": 0.83866251, + "learning_rate": 2.3847356206396143e-06, + "loss": 0.86037862, + "num_input_tokens_seen": 81543135, + "step": 3786, + "time_per_iteration": 3.659609794616699 + }, + { + "auxiliary_loss_clip": 0.01188057, + "auxiliary_loss_mlp": 0.01032019, + "balance_loss_clip": 1.05815697, + "balance_loss_mlp": 1.02408004, + "epoch": 0.45535982685023746, + "flos": 23257510191360.0, + "grad_norm": 1.5289224390839944, + "language_loss": 0.78799999, + "learning_rate": 2.3839711687643227e-06, + "loss": 0.81020069, + "num_input_tokens_seen": 81564360, + "step": 3787, + "time_per_iteration": 2.61242413520813 + }, + { + "auxiliary_loss_clip": 0.01161006, + "auxiliary_loss_mlp": 0.01035728, + "balance_loss_clip": 1.01741922, + "balance_loss_mlp": 1.02599478, + "epoch": 0.45548006974087657, + "flos": 19646117907840.0, + "grad_norm": 2.420780801068834, + "language_loss": 0.74493146, + "learning_rate": 2.38320665864091e-06, + "loss": 0.76689875, + "num_input_tokens_seen": 81583710, + "step": 3788, + "time_per_iteration": 3.535019874572754 + }, + { + "auxiliary_loss_clip": 0.01070056, + "auxiliary_loss_mlp": 0.01037964, + "balance_loss_clip": 0.86357546, + "balance_loss_mlp": 1.029935, + "epoch": 0.4556003126315157, + "flos": 20047778766720.0, + "grad_norm": 1.9065087402472314, + "language_loss": 0.81718183, + "learning_rate": 2.3824420903853516e-06, + "loss": 0.83826202, + "num_input_tokens_seen": 81602175, + "step": 3789, + "time_per_iteration": 2.760061025619507 + }, + { + "auxiliary_loss_clip": 0.01164136, + "auxiliary_loss_mlp": 0.01029044, + "balance_loss_clip": 1.02131045, + "balance_loss_mlp": 1.02013886, + "epoch": 0.45572055552215474, + "flos": 22959738443520.0, + "grad_norm": 2.7037376536332878, + "language_loss": 0.8208673, + "learning_rate": 2.3816774641136324e-06, + "loss": 0.84279913, + "num_input_tokens_seen": 81619430, + "step": 3790, + "time_per_iteration": 2.631650924682617 + }, + { + "auxiliary_loss_clip": 0.01163819, + "auxiliary_loss_mlp": 0.00510148, + "balance_loss_clip": 1.02196252, + "balance_loss_mlp": 0.0, + "epoch": 0.45584079841279385, + "flos": 33109925535360.0, + "grad_norm": 1.7447271986318296, + "language_loss": 0.71514302, + "learning_rate": 2.380912779941745e-06, + "loss": 0.73188269, + "num_input_tokens_seen": 81642550, + "step": 3791, + "time_per_iteration": 3.7468278408050537 + }, + { + "auxiliary_loss_clip": 0.01164356, + "auxiliary_loss_mlp": 0.0103177, + "balance_loss_clip": 1.01667666, + "balance_loss_mlp": 1.02192378, + "epoch": 0.45596104130343296, + "flos": 27272179445760.0, + "grad_norm": 1.841423867299112, + "language_loss": 0.83029407, + "learning_rate": 2.3801480379856918e-06, + "loss": 0.85225534, + "num_input_tokens_seen": 81664260, + "step": 3792, + "time_per_iteration": 2.767974615097046 + }, + { + "auxiliary_loss_clip": 0.01145123, + "auxiliary_loss_mlp": 0.01028827, + "balance_loss_clip": 0.98395813, + "balance_loss_mlp": 1.02009511, + "epoch": 0.456081284194072, + "flos": 21579799697280.0, + "grad_norm": 1.6285775805572187, + "language_loss": 0.836761, + "learning_rate": 2.379383238361484e-06, + "loss": 0.85850048, + "num_input_tokens_seen": 81683620, + "step": 3793, + "time_per_iteration": 2.7259714603424072 + }, + { + "auxiliary_loss_clip": 0.01159885, + "auxiliary_loss_mlp": 0.0103012, + "balance_loss_clip": 1.01863933, + "balance_loss_mlp": 1.02210355, + "epoch": 0.4562015270847111, + "flos": 35918822113920.0, + "grad_norm": 2.1378797987399873, + "language_loss": 0.79810029, + "learning_rate": 2.3786183811851407e-06, + "loss": 0.82000041, + "num_input_tokens_seen": 81704325, + "step": 3794, + "time_per_iteration": 2.76432466506958 + }, + { + "auxiliary_loss_clip": 0.01188587, + "auxiliary_loss_mlp": 0.01033544, + "balance_loss_clip": 1.05932999, + "balance_loss_mlp": 1.02476466, + "epoch": 0.45632176997535023, + "flos": 13589783602560.0, + "grad_norm": 2.1384700376957495, + "language_loss": 0.80364549, + "learning_rate": 2.3778534665726892e-06, + "loss": 0.82586682, + "num_input_tokens_seen": 81721155, + "step": 3795, + "time_per_iteration": 2.5688107013702393 + }, + { + "auxiliary_loss_clip": 0.01156303, + "auxiliary_loss_mlp": 0.01031523, + "balance_loss_clip": 1.02019131, + "balance_loss_mlp": 1.02335668, + "epoch": 0.4564420128659893, + "flos": 32635401937920.0, + "grad_norm": 2.045920916024667, + "language_loss": 0.72453141, + "learning_rate": 2.377088494640168e-06, + "loss": 0.74640965, + "num_input_tokens_seen": 81742905, + "step": 3796, + "time_per_iteration": 2.708437442779541 + }, + { + "auxiliary_loss_clip": 0.01166246, + "auxiliary_loss_mlp": 0.0103111, + "balance_loss_clip": 1.02581143, + "balance_loss_mlp": 1.02302766, + "epoch": 0.4565622557566284, + "flos": 20377690208640.0, + "grad_norm": 2.166212306046291, + "language_loss": 0.78294939, + "learning_rate": 2.3763234655036216e-06, + "loss": 0.80492294, + "num_input_tokens_seen": 81762105, + "step": 3797, + "time_per_iteration": 2.6881303787231445 + }, + { + "auxiliary_loss_clip": 0.01115424, + "auxiliary_loss_mlp": 0.01029369, + "balance_loss_clip": 0.94003803, + "balance_loss_mlp": 1.02212071, + "epoch": 0.45668249864726745, + "flos": 25374372364800.0, + "grad_norm": 1.9496145349647325, + "language_loss": 0.8741563, + "learning_rate": 2.3755583792791046e-06, + "loss": 0.89560425, + "num_input_tokens_seen": 81781975, + "step": 3798, + "time_per_iteration": 2.729099750518799 + }, + { + "auxiliary_loss_clip": 0.01167588, + "auxiliary_loss_mlp": 0.01033646, + "balance_loss_clip": 1.02185297, + "balance_loss_mlp": 1.02492023, + "epoch": 0.45680274153790656, + "flos": 15559806977280.0, + "grad_norm": 2.1530390746919377, + "language_loss": 0.74632454, + "learning_rate": 2.3747932360826803e-06, + "loss": 0.76833689, + "num_input_tokens_seen": 81798905, + "step": 3799, + "time_per_iteration": 2.5874998569488525 + }, + { + "auxiliary_loss_clip": 0.01161116, + "auxiliary_loss_mlp": 0.01030957, + "balance_loss_clip": 1.01814222, + "balance_loss_mlp": 1.02142048, + "epoch": 0.4569229844285457, + "flos": 19792884879360.0, + "grad_norm": 14.462042216467506, + "language_loss": 0.82085806, + "learning_rate": 2.3740280360304205e-06, + "loss": 0.8427788, + "num_input_tokens_seen": 81816630, + "step": 3800, + "time_per_iteration": 2.6345343589782715 + }, + { + "auxiliary_loss_clip": 0.01117156, + "auxiliary_loss_mlp": 0.01030244, + "balance_loss_clip": 0.94675577, + "balance_loss_mlp": 1.02220964, + "epoch": 0.45704322731918473, + "flos": 24093941270400.0, + "grad_norm": 1.6211959637169042, + "language_loss": 0.68208432, + "learning_rate": 2.3732627792384038e-06, + "loss": 0.70355833, + "num_input_tokens_seen": 81837700, + "step": 3801, + "time_per_iteration": 2.6966073513031006 + }, + { + "auxiliary_loss_clip": 0.01186959, + "auxiliary_loss_mlp": 0.01028668, + "balance_loss_clip": 1.05745745, + "balance_loss_mlp": 1.020365, + "epoch": 0.45716347020982384, + "flos": 31317803245440.0, + "grad_norm": 2.067134638248906, + "language_loss": 0.75105846, + "learning_rate": 2.3724974658227207e-06, + "loss": 0.77321476, + "num_input_tokens_seen": 81858490, + "step": 3802, + "time_per_iteration": 2.6896939277648926 + }, + { + "auxiliary_loss_clip": 0.01138681, + "auxiliary_loss_mlp": 0.00510347, + "balance_loss_clip": 0.98142678, + "balance_loss_mlp": 0.0, + "epoch": 0.45728371310046295, + "flos": 26501392471680.0, + "grad_norm": 1.8996306358083797, + "language_loss": 0.7126019, + "learning_rate": 2.3717320958994687e-06, + "loss": 0.72909218, + "num_input_tokens_seen": 81876050, + "step": 3803, + "time_per_iteration": 2.673177719116211 + }, + { + "auxiliary_loss_clip": 0.01116103, + "auxiliary_loss_mlp": 0.01032306, + "balance_loss_clip": 0.93848145, + "balance_loss_mlp": 1.0240624, + "epoch": 0.457403955991102, + "flos": 17929408222080.0, + "grad_norm": 1.9579263027378713, + "language_loss": 0.70020878, + "learning_rate": 2.3709666695847534e-06, + "loss": 0.72169286, + "num_input_tokens_seen": 81894230, + "step": 3804, + "time_per_iteration": 2.6992812156677246 + }, + { + "auxiliary_loss_clip": 0.010873, + "auxiliary_loss_mlp": 0.01025989, + "balance_loss_clip": 0.90323967, + "balance_loss_mlp": 1.01761484, + "epoch": 0.4575241988817411, + "flos": 42230660837760.0, + "grad_norm": 1.6355128436095816, + "language_loss": 0.70148432, + "learning_rate": 2.370201186994689e-06, + "loss": 0.72261727, + "num_input_tokens_seen": 81917915, + "step": 3805, + "time_per_iteration": 2.8724894523620605 + }, + { + "auxiliary_loss_clip": 0.01134355, + "auxiliary_loss_mlp": 0.01024796, + "balance_loss_clip": 0.982512, + "balance_loss_mlp": 1.01667166, + "epoch": 0.45764444177238023, + "flos": 30117309868800.0, + "grad_norm": 1.8406053776956242, + "language_loss": 0.70291805, + "learning_rate": 2.369435648245399e-06, + "loss": 0.72450954, + "num_input_tokens_seen": 81938130, + "step": 3806, + "time_per_iteration": 2.754772901535034 + }, + { + "auxiliary_loss_clip": 0.01140874, + "auxiliary_loss_mlp": 0.0103108, + "balance_loss_clip": 0.98261809, + "balance_loss_mlp": 1.02347445, + "epoch": 0.4577646846630193, + "flos": 24060293205120.0, + "grad_norm": 4.6583680074483365, + "language_loss": 0.8495841, + "learning_rate": 2.368670053453015e-06, + "loss": 0.87130362, + "num_input_tokens_seen": 81959820, + "step": 3807, + "time_per_iteration": 2.7160370349884033 + }, + { + "auxiliary_loss_clip": 0.01174577, + "auxiliary_loss_mlp": 0.01031643, + "balance_loss_clip": 1.0254432, + "balance_loss_mlp": 1.02343512, + "epoch": 0.4578849275536584, + "flos": 17418578952960.0, + "grad_norm": 3.1178872968309284, + "language_loss": 0.74563706, + "learning_rate": 2.3679044027336757e-06, + "loss": 0.76769924, + "num_input_tokens_seen": 81975710, + "step": 3808, + "time_per_iteration": 2.778978109359741 + }, + { + "auxiliary_loss_clip": 0.01190247, + "auxiliary_loss_mlp": 0.01037428, + "balance_loss_clip": 1.05982113, + "balance_loss_mlp": 1.02812386, + "epoch": 0.4580051704442975, + "flos": 13510169107200.0, + "grad_norm": 4.624615141790103, + "language_loss": 0.69524682, + "learning_rate": 2.3671386962035326e-06, + "loss": 0.71752357, + "num_input_tokens_seen": 81993180, + "step": 3809, + "time_per_iteration": 2.7153971195220947 + }, + { + "auxiliary_loss_clip": 0.01165935, + "auxiliary_loss_mlp": 0.01028009, + "balance_loss_clip": 1.02229309, + "balance_loss_mlp": 1.01911604, + "epoch": 0.45812541333493656, + "flos": 18037606965120.0, + "grad_norm": 1.9628914096957946, + "language_loss": 0.68761718, + "learning_rate": 2.3663729339787405e-06, + "loss": 0.70955664, + "num_input_tokens_seen": 82010115, + "step": 3810, + "time_per_iteration": 2.616090774536133 + }, + { + "auxiliary_loss_clip": 0.01189619, + "auxiliary_loss_mlp": 0.01029544, + "balance_loss_clip": 1.0615257, + "balance_loss_mlp": 1.02101433, + "epoch": 0.45824565622557567, + "flos": 20222196232320.0, + "grad_norm": 2.5760602132385926, + "language_loss": 0.73454523, + "learning_rate": 2.365607116175466e-06, + "loss": 0.75673687, + "num_input_tokens_seen": 82025540, + "step": 3811, + "time_per_iteration": 3.4934775829315186 + }, + { + "auxiliary_loss_clip": 0.01187871, + "auxiliary_loss_mlp": 0.01032858, + "balance_loss_clip": 1.05984616, + "balance_loss_mlp": 1.02453172, + "epoch": 0.4583658991162148, + "flos": 19864885691520.0, + "grad_norm": 2.930058732853804, + "language_loss": 0.67059648, + "learning_rate": 2.3648412429098825e-06, + "loss": 0.69280374, + "num_input_tokens_seen": 82043890, + "step": 3812, + "time_per_iteration": 3.545844554901123 + }, + { + "auxiliary_loss_clip": 0.01109341, + "auxiliary_loss_mlp": 0.01037329, + "balance_loss_clip": 0.93985927, + "balance_loss_mlp": 1.0281322, + "epoch": 0.45848614200685384, + "flos": 21029935322880.0, + "grad_norm": 2.072052114308023, + "language_loss": 0.8202616, + "learning_rate": 2.364075314298172e-06, + "loss": 0.84172833, + "num_input_tokens_seen": 82061345, + "step": 3813, + "time_per_iteration": 3.5054030418395996 + }, + { + "auxiliary_loss_clip": 0.01167958, + "auxiliary_loss_mlp": 0.00510895, + "balance_loss_clip": 1.02175367, + "balance_loss_mlp": 0.0, + "epoch": 0.45860638489749295, + "flos": 21069293650560.0, + "grad_norm": 1.7777366397173469, + "language_loss": 0.70029157, + "learning_rate": 2.3633093304565267e-06, + "loss": 0.71708012, + "num_input_tokens_seen": 82080400, + "step": 3814, + "time_per_iteration": 2.6489200592041016 + }, + { + "auxiliary_loss_clip": 0.01194374, + "auxiliary_loss_mlp": 0.01027455, + "balance_loss_clip": 1.06267476, + "balance_loss_mlp": 1.01905692, + "epoch": 0.458726627788132, + "flos": 26833889692800.0, + "grad_norm": 1.7765128630081224, + "language_loss": 0.63413298, + "learning_rate": 2.3625432915011443e-06, + "loss": 0.65635121, + "num_input_tokens_seen": 82102310, + "step": 3815, + "time_per_iteration": 2.6336846351623535 + }, + { + "auxiliary_loss_clip": 0.01133898, + "auxiliary_loss_mlp": 0.01029929, + "balance_loss_clip": 0.98110086, + "balance_loss_mlp": 1.02105367, + "epoch": 0.4588468706787711, + "flos": 24097927680000.0, + "grad_norm": 1.9912096626767275, + "language_loss": 0.65402627, + "learning_rate": 2.3617771975482334e-06, + "loss": 0.67566448, + "num_input_tokens_seen": 82121140, + "step": 3816, + "time_per_iteration": 2.743959903717041 + }, + { + "auxiliary_loss_clip": 0.01087948, + "auxiliary_loss_mlp": 0.01031908, + "balance_loss_clip": 0.90326858, + "balance_loss_mlp": 1.02379012, + "epoch": 0.4589671135694102, + "flos": 17889331622400.0, + "grad_norm": 1.5963961976806917, + "language_loss": 0.74618834, + "learning_rate": 2.3610110487140083e-06, + "loss": 0.76738691, + "num_input_tokens_seen": 82139575, + "step": 3817, + "time_per_iteration": 3.706568956375122 + }, + { + "auxiliary_loss_clip": 0.01145372, + "auxiliary_loss_mlp": 0.01029936, + "balance_loss_clip": 0.98869693, + "balance_loss_mlp": 1.02200913, + "epoch": 0.4590873564600493, + "flos": 25626967781760.0, + "grad_norm": 1.6352463425903048, + "language_loss": 0.80960506, + "learning_rate": 2.360244845114695e-06, + "loss": 0.83135813, + "num_input_tokens_seen": 82159195, + "step": 3818, + "time_per_iteration": 2.6879522800445557 + }, + { + "auxiliary_loss_clip": 0.0113699, + "auxiliary_loss_mlp": 0.0103238, + "balance_loss_clip": 0.98571628, + "balance_loss_mlp": 1.02393365, + "epoch": 0.4592075993506884, + "flos": 18514788168960.0, + "grad_norm": 1.9966366305482235, + "language_loss": 0.68492961, + "learning_rate": 2.3594785868665245e-06, + "loss": 0.70662332, + "num_input_tokens_seen": 82175500, + "step": 3819, + "time_per_iteration": 2.7458114624023438 + }, + { + "auxiliary_loss_clip": 0.01119155, + "auxiliary_loss_mlp": 0.00511056, + "balance_loss_clip": 0.9445771, + "balance_loss_mlp": 0.0, + "epoch": 0.4593278422413275, + "flos": 20631111638400.0, + "grad_norm": 3.2802783837437928, + "language_loss": 0.80669034, + "learning_rate": 2.3587122740857386e-06, + "loss": 0.82299244, + "num_input_tokens_seen": 82192600, + "step": 3820, + "time_per_iteration": 2.7279775142669678 + }, + { + "auxiliary_loss_clip": 0.01165128, + "auxiliary_loss_mlp": 0.01026276, + "balance_loss_clip": 1.02228379, + "balance_loss_mlp": 1.01837826, + "epoch": 0.45944808513196655, + "flos": 21358517961600.0, + "grad_norm": 1.9539105565666754, + "language_loss": 0.77983558, + "learning_rate": 2.357945906888586e-06, + "loss": 0.80174971, + "num_input_tokens_seen": 82212040, + "step": 3821, + "time_per_iteration": 2.666456699371338 + }, + { + "auxiliary_loss_clip": 0.01166643, + "auxiliary_loss_mlp": 0.01029621, + "balance_loss_clip": 1.02274406, + "balance_loss_mlp": 1.02056682, + "epoch": 0.45956832802260567, + "flos": 21427789340160.0, + "grad_norm": 2.6530130269246976, + "language_loss": 0.80243707, + "learning_rate": 2.357179485391324e-06, + "loss": 0.82439971, + "num_input_tokens_seen": 82229895, + "step": 3822, + "time_per_iteration": 2.648679256439209 + }, + { + "auxiliary_loss_clip": 0.01185449, + "auxiliary_loss_mlp": 0.0102642, + "balance_loss_clip": 1.05979729, + "balance_loss_mlp": 1.01847434, + "epoch": 0.4596885709132448, + "flos": 22382654538240.0, + "grad_norm": 1.9037671414491641, + "language_loss": 0.86133558, + "learning_rate": 2.3564130097102173e-06, + "loss": 0.8834542, + "num_input_tokens_seen": 82249550, + "step": 3823, + "time_per_iteration": 2.6535634994506836 + }, + { + "auxiliary_loss_clip": 0.01137199, + "auxiliary_loss_mlp": 0.01029927, + "balance_loss_clip": 0.9882226, + "balance_loss_mlp": 1.02146316, + "epoch": 0.45980881380388383, + "flos": 28981957806720.0, + "grad_norm": 2.238662336161341, + "language_loss": 0.75066519, + "learning_rate": 2.355646479961541e-06, + "loss": 0.77233648, + "num_input_tokens_seen": 82268860, + "step": 3824, + "time_per_iteration": 2.726954460144043 + }, + { + "auxiliary_loss_clip": 0.01189105, + "auxiliary_loss_mlp": 0.01027991, + "balance_loss_clip": 1.06058526, + "balance_loss_mlp": 1.0200516, + "epoch": 0.45992905669452294, + "flos": 33396599980800.0, + "grad_norm": 2.169341466963655, + "language_loss": 0.71423697, + "learning_rate": 2.354879896261576e-06, + "loss": 0.736408, + "num_input_tokens_seen": 82289070, + "step": 3825, + "time_per_iteration": 2.647068977355957 + }, + { + "auxiliary_loss_clip": 0.011148, + "auxiliary_loss_mlp": 0.01028452, + "balance_loss_clip": 0.94653481, + "balance_loss_mlp": 1.0201726, + "epoch": 0.46004929958516205, + "flos": 36318184502400.0, + "grad_norm": 1.8870384513131764, + "language_loss": 0.56709737, + "learning_rate": 2.3541132587266133e-06, + "loss": 0.58852988, + "num_input_tokens_seen": 82311790, + "step": 3826, + "time_per_iteration": 2.7959883213043213 + }, + { + "auxiliary_loss_clip": 0.01124411, + "auxiliary_loss_mlp": 0.0103022, + "balance_loss_clip": 0.94538063, + "balance_loss_mlp": 1.02223909, + "epoch": 0.4601695424758011, + "flos": 17238451224960.0, + "grad_norm": 1.8074204651124304, + "language_loss": 0.69538814, + "learning_rate": 2.3533465674729515e-06, + "loss": 0.7169345, + "num_input_tokens_seen": 82329020, + "step": 3827, + "time_per_iteration": 2.706059455871582 + }, + { + "auxiliary_loss_clip": 0.01190573, + "auxiliary_loss_mlp": 0.01031146, + "balance_loss_clip": 1.06235373, + "balance_loss_mlp": 1.02281952, + "epoch": 0.4602897853664402, + "flos": 15888425529600.0, + "grad_norm": 1.8712998649769987, + "language_loss": 0.72774208, + "learning_rate": 2.352579822616895e-06, + "loss": 0.74995929, + "num_input_tokens_seen": 82346455, + "step": 3828, + "time_per_iteration": 2.576389789581299 + }, + { + "auxiliary_loss_clip": 0.01143064, + "auxiliary_loss_mlp": 0.01025054, + "balance_loss_clip": 0.98258293, + "balance_loss_mlp": 1.01642323, + "epoch": 0.4604100282570793, + "flos": 25412617370880.0, + "grad_norm": 2.208796598457133, + "language_loss": 0.78106105, + "learning_rate": 2.351813024274761e-06, + "loss": 0.80274218, + "num_input_tokens_seen": 82367810, + "step": 3829, + "time_per_iteration": 2.7317891120910645 + }, + { + "auxiliary_loss_clip": 0.01125257, + "auxiliary_loss_mlp": 0.01032138, + "balance_loss_clip": 0.94736952, + "balance_loss_mlp": 1.02388287, + "epoch": 0.4605302711477184, + "flos": 27630711048960.0, + "grad_norm": 1.7819110745087432, + "language_loss": 0.73477113, + "learning_rate": 2.3510461725628693e-06, + "loss": 0.75634515, + "num_input_tokens_seen": 82388275, + "step": 3830, + "time_per_iteration": 2.7640011310577393 + }, + { + "auxiliary_loss_clip": 0.01121177, + "auxiliary_loss_mlp": 0.01029875, + "balance_loss_clip": 0.94560993, + "balance_loss_mlp": 1.02153659, + "epoch": 0.4606505140383575, + "flos": 23839657914240.0, + "grad_norm": 1.8656962919896296, + "language_loss": 0.70930797, + "learning_rate": 2.350279267597554e-06, + "loss": 0.73081845, + "num_input_tokens_seen": 82408915, + "step": 3831, + "time_per_iteration": 2.7628703117370605 + }, + { + "auxiliary_loss_clip": 0.01163987, + "auxiliary_loss_mlp": 0.01031959, + "balance_loss_clip": 1.0213728, + "balance_loss_mlp": 1.02385283, + "epoch": 0.46077075692899655, + "flos": 16107013745280.0, + "grad_norm": 2.0786487747020463, + "language_loss": 0.82770193, + "learning_rate": 2.3495123094951515e-06, + "loss": 0.84966141, + "num_input_tokens_seen": 82427260, + "step": 3832, + "time_per_iteration": 2.662182092666626 + }, + { + "auxiliary_loss_clip": 0.0113431, + "auxiliary_loss_mlp": 0.0102948, + "balance_loss_clip": 0.98226607, + "balance_loss_mlp": 1.0212841, + "epoch": 0.46089099981963566, + "flos": 48798147634560.0, + "grad_norm": 1.9011097781671382, + "language_loss": 0.76296508, + "learning_rate": 2.34874529837201e-06, + "loss": 0.784603, + "num_input_tokens_seen": 82450805, + "step": 3833, + "time_per_iteration": 2.8714559078216553 + }, + { + "auxiliary_loss_clip": 0.01066637, + "auxiliary_loss_mlp": 0.01028684, + "balance_loss_clip": 0.86524075, + "balance_loss_mlp": 1.02077436, + "epoch": 0.46101124271027477, + "flos": 19099234362240.0, + "grad_norm": 2.2065300667896377, + "language_loss": 0.78963387, + "learning_rate": 2.347978234344483e-06, + "loss": 0.81058705, + "num_input_tokens_seen": 82467010, + "step": 3834, + "time_per_iteration": 2.8212051391601562 + }, + { + "auxiliary_loss_clip": 0.01174193, + "auxiliary_loss_mlp": 0.01030399, + "balance_loss_clip": 1.02686453, + "balance_loss_mlp": 1.0218513, + "epoch": 0.4611314856009138, + "flos": 39347931853440.0, + "grad_norm": 3.2376625388395075, + "language_loss": 0.69048405, + "learning_rate": 2.347211117528935e-06, + "loss": 0.71253002, + "num_input_tokens_seen": 82489310, + "step": 3835, + "time_per_iteration": 2.781897783279419 + }, + { + "auxiliary_loss_clip": 0.01131831, + "auxiliary_loss_mlp": 0.0102653, + "balance_loss_clip": 0.95204008, + "balance_loss_mlp": 1.01835811, + "epoch": 0.46125172849155294, + "flos": 20810772489600.0, + "grad_norm": 1.545336969160767, + "language_loss": 0.71839088, + "learning_rate": 2.3464439480417374e-06, + "loss": 0.7399745, + "num_input_tokens_seen": 82508830, + "step": 3836, + "time_per_iteration": 2.7164580821990967 + }, + { + "auxiliary_loss_clip": 0.01170735, + "auxiliary_loss_mlp": 0.01029275, + "balance_loss_clip": 1.02477002, + "balance_loss_mlp": 1.02079296, + "epoch": 0.46137197138219205, + "flos": 17930808852480.0, + "grad_norm": 4.750125208173275, + "language_loss": 0.76614243, + "learning_rate": 2.3456767259992676e-06, + "loss": 0.78814256, + "num_input_tokens_seen": 82526475, + "step": 3837, + "time_per_iteration": 3.6448702812194824 + }, + { + "auxiliary_loss_clip": 0.01190646, + "auxiliary_loss_mlp": 0.00510809, + "balance_loss_clip": 1.05961549, + "balance_loss_mlp": 0.0, + "epoch": 0.4614922142728311, + "flos": 16836610798080.0, + "grad_norm": 2.1559150629480213, + "language_loss": 0.8870728, + "learning_rate": 2.3449094515179135e-06, + "loss": 0.90408731, + "num_input_tokens_seen": 82543935, + "step": 3838, + "time_per_iteration": 3.624203681945801 + }, + { + "auxiliary_loss_clip": 0.01148541, + "auxiliary_loss_mlp": 0.01030703, + "balance_loss_clip": 0.98437154, + "balance_loss_mlp": 1.02226305, + "epoch": 0.4616124571634702, + "flos": 26614906427520.0, + "grad_norm": 1.6067459278653247, + "language_loss": 0.81725097, + "learning_rate": 2.34414212471407e-06, + "loss": 0.83904338, + "num_input_tokens_seen": 82563730, + "step": 3839, + "time_per_iteration": 3.6683881282806396 + }, + { + "auxiliary_loss_clip": 0.01172395, + "auxiliary_loss_mlp": 0.01026655, + "balance_loss_clip": 1.02282238, + "balance_loss_mlp": 1.01818514, + "epoch": 0.4617327000541093, + "flos": 20340127560960.0, + "grad_norm": 2.162125778672055, + "language_loss": 0.73108304, + "learning_rate": 2.3433747457041394e-06, + "loss": 0.75307345, + "num_input_tokens_seen": 82582435, + "step": 3840, + "time_per_iteration": 2.6473469734191895 + }, + { + "auxiliary_loss_clip": 0.0112248, + "auxiliary_loss_mlp": 0.01027323, + "balance_loss_clip": 0.95069516, + "balance_loss_mlp": 1.01844835, + "epoch": 0.4618529429447484, + "flos": 29570749545600.0, + "grad_norm": 2.2620516950532985, + "language_loss": 0.84860098, + "learning_rate": 2.342607314604533e-06, + "loss": 0.87009907, + "num_input_tokens_seen": 82602185, + "step": 3841, + "time_per_iteration": 2.7959916591644287 + }, + { + "auxiliary_loss_clip": 0.01166698, + "auxiliary_loss_mlp": 0.01031858, + "balance_loss_clip": 1.02679205, + "balance_loss_mlp": 1.02351284, + "epoch": 0.4619731858353875, + "flos": 19787030962560.0, + "grad_norm": 1.8015243958735008, + "language_loss": 0.84463376, + "learning_rate": 2.3418398315316694e-06, + "loss": 0.86661935, + "num_input_tokens_seen": 82620005, + "step": 3842, + "time_per_iteration": 2.6376185417175293 + }, + { + "auxiliary_loss_clip": 0.01191988, + "auxiliary_loss_mlp": 0.01032791, + "balance_loss_clip": 1.06273437, + "balance_loss_mlp": 1.02429724, + "epoch": 0.4620934287260266, + "flos": 18951138587520.0, + "grad_norm": 2.482960673269094, + "language_loss": 0.78749847, + "learning_rate": 2.3410722966019755e-06, + "loss": 0.80974627, + "num_input_tokens_seen": 82635120, + "step": 3843, + "time_per_iteration": 3.480142116546631 + }, + { + "auxiliary_loss_clip": 0.01165594, + "auxiliary_loss_mlp": 0.01030762, + "balance_loss_clip": 1.02226269, + "balance_loss_mlp": 1.02229822, + "epoch": 0.46221367161666566, + "flos": 37341674634240.0, + "grad_norm": 1.7578060180557837, + "language_loss": 0.65866667, + "learning_rate": 2.3403047099318848e-06, + "loss": 0.68063021, + "num_input_tokens_seen": 82659190, + "step": 3844, + "time_per_iteration": 2.765272378921509 + }, + { + "auxiliary_loss_clip": 0.01089895, + "auxiliary_loss_mlp": 0.01027349, + "balance_loss_clip": 0.90370041, + "balance_loss_mlp": 1.01907611, + "epoch": 0.46233391450730477, + "flos": 14428549065600.0, + "grad_norm": 2.8838876871761676, + "language_loss": 0.75625533, + "learning_rate": 2.3395370716378405e-06, + "loss": 0.77742779, + "num_input_tokens_seen": 82676635, + "step": 3845, + "time_per_iteration": 2.6850478649139404 + }, + { + "auxiliary_loss_clip": 0.01170031, + "auxiliary_loss_mlp": 0.01026614, + "balance_loss_clip": 1.02167761, + "balance_loss_mlp": 1.01808429, + "epoch": 0.4624541573979438, + "flos": 22493044010880.0, + "grad_norm": 7.9076322739582885, + "language_loss": 0.72727543, + "learning_rate": 2.338769381836292e-06, + "loss": 0.74924183, + "num_input_tokens_seen": 82696245, + "step": 3846, + "time_per_iteration": 2.659679651260376 + }, + { + "auxiliary_loss_clip": 0.01111421, + "auxiliary_loss_mlp": 0.01031938, + "balance_loss_clip": 0.94250667, + "balance_loss_mlp": 1.0231111, + "epoch": 0.46257440028858293, + "flos": 14465070218880.0, + "grad_norm": 2.6142017223057468, + "language_loss": 0.73128182, + "learning_rate": 2.3380016406436984e-06, + "loss": 0.75271541, + "num_input_tokens_seen": 82713725, + "step": 3847, + "time_per_iteration": 2.6700010299682617 + }, + { + "auxiliary_loss_clip": 0.01087431, + "auxiliary_loss_mlp": 0.0103755, + "balance_loss_clip": 0.90250874, + "balance_loss_mlp": 1.02887762, + "epoch": 0.46269464317922204, + "flos": 23332204523520.0, + "grad_norm": 1.8570221586222642, + "language_loss": 0.81499177, + "learning_rate": 2.337233848176524e-06, + "loss": 0.8362416, + "num_input_tokens_seen": 82731495, + "step": 3848, + "time_per_iteration": 2.7412986755371094 + }, + { + "auxiliary_loss_clip": 0.01086007, + "auxiliary_loss_mlp": 0.0102583, + "balance_loss_clip": 0.90351266, + "balance_loss_mlp": 1.01682329, + "epoch": 0.4628148860698611, + "flos": 18552027594240.0, + "grad_norm": 1.9014983681456783, + "language_loss": 0.83367801, + "learning_rate": 2.3364660045512435e-06, + "loss": 0.85479635, + "num_input_tokens_seen": 82750255, + "step": 3849, + "time_per_iteration": 2.724062204360962 + }, + { + "auxiliary_loss_clip": 0.01055483, + "auxiliary_loss_mlp": 0.01005167, + "balance_loss_clip": 0.96185315, + "balance_loss_mlp": 1.0037961, + "epoch": 0.4629351289605002, + "flos": 70667569670400.0, + "grad_norm": 0.7403512921110489, + "language_loss": 0.58217895, + "learning_rate": 2.335698109884337e-06, + "loss": 0.60278547, + "num_input_tokens_seen": 82815460, + "step": 3850, + "time_per_iteration": 3.387108087539673 + }, + { + "auxiliary_loss_clip": 0.01028709, + "auxiliary_loss_mlp": 0.01025652, + "balance_loss_clip": 0.90363598, + "balance_loss_mlp": 1.02403092, + "epoch": 0.4630553718511393, + "flos": 59687200465920.0, + "grad_norm": 0.8271793274193638, + "language_loss": 0.59896165, + "learning_rate": 2.334930164292294e-06, + "loss": 0.61950529, + "num_input_tokens_seen": 82878010, + "step": 3851, + "time_per_iteration": 3.4393091201782227 + }, + { + "auxiliary_loss_clip": 0.01086153, + "auxiliary_loss_mlp": 0.01030037, + "balance_loss_clip": 0.90294838, + "balance_loss_mlp": 1.0219065, + "epoch": 0.4631756147417784, + "flos": 15960605909760.0, + "grad_norm": 1.8660148569979853, + "language_loss": 0.80325961, + "learning_rate": 2.334162167891612e-06, + "loss": 0.82442147, + "num_input_tokens_seen": 82895275, + "step": 3852, + "time_per_iteration": 2.7688143253326416 + }, + { + "auxiliary_loss_clip": 0.01142911, + "auxiliary_loss_mlp": 0.01024174, + "balance_loss_clip": 0.98192018, + "balance_loss_mlp": 1.01569855, + "epoch": 0.4632958576324175, + "flos": 16472907636480.0, + "grad_norm": 2.547491126396598, + "language_loss": 0.75072169, + "learning_rate": 2.333394120798795e-06, + "loss": 0.77239257, + "num_input_tokens_seen": 82914010, + "step": 3853, + "time_per_iteration": 2.666715383529663 + }, + { + "auxiliary_loss_clip": 0.01142579, + "auxiliary_loss_mlp": 0.01025268, + "balance_loss_clip": 0.98117423, + "balance_loss_mlp": 1.01699483, + "epoch": 0.4634161005230566, + "flos": 22346492520960.0, + "grad_norm": 2.4393392179042346, + "language_loss": 0.7247864, + "learning_rate": 2.3326260231303545e-06, + "loss": 0.74646485, + "num_input_tokens_seen": 82932610, + "step": 3854, + "time_per_iteration": 2.708007574081421 + }, + { + "auxiliary_loss_clip": 0.01188354, + "auxiliary_loss_mlp": 0.01033097, + "balance_loss_clip": 1.06159949, + "balance_loss_mlp": 1.0253036, + "epoch": 0.46353634341369565, + "flos": 15742233175680.0, + "grad_norm": 1.677527005157155, + "language_loss": 0.86857808, + "learning_rate": 2.331857875002811e-06, + "loss": 0.89079261, + "num_input_tokens_seen": 82951210, + "step": 3855, + "time_per_iteration": 2.5551133155822754 + }, + { + "auxiliary_loss_clip": 0.01142863, + "auxiliary_loss_mlp": 0.01036976, + "balance_loss_clip": 0.98596251, + "balance_loss_mlp": 1.02836883, + "epoch": 0.46365658630433476, + "flos": 28329820433280.0, + "grad_norm": 2.080884991228566, + "language_loss": 0.76392651, + "learning_rate": 2.3310896765326916e-06, + "loss": 0.78572488, + "num_input_tokens_seen": 82972210, + "step": 3856, + "time_per_iteration": 2.743114948272705 + }, + { + "auxiliary_loss_clip": 0.011175, + "auxiliary_loss_mlp": 0.01028172, + "balance_loss_clip": 0.94771874, + "balance_loss_mlp": 1.01936865, + "epoch": 0.46377682919497387, + "flos": 24608074590720.0, + "grad_norm": 1.847662364128649, + "language_loss": 0.8422941, + "learning_rate": 2.330321427836531e-06, + "loss": 0.86375082, + "num_input_tokens_seen": 82994080, + "step": 3857, + "time_per_iteration": 2.7148280143737793 + }, + { + "auxiliary_loss_clip": 0.01163431, + "auxiliary_loss_mlp": 0.01035063, + "balance_loss_clip": 1.02240789, + "balance_loss_mlp": 1.02712393, + "epoch": 0.4638970720856129, + "flos": 19060953442560.0, + "grad_norm": 1.8645141015963054, + "language_loss": 0.82766736, + "learning_rate": 2.3295531290308733e-06, + "loss": 0.84965229, + "num_input_tokens_seen": 83012230, + "step": 3858, + "time_per_iteration": 2.6567466259002686 + }, + { + "auxiliary_loss_clip": 0.01192366, + "auxiliary_loss_mlp": 0.00510995, + "balance_loss_clip": 1.06260371, + "balance_loss_mlp": 0.0, + "epoch": 0.46401731497625204, + "flos": 18471012468480.0, + "grad_norm": 2.7109228871367, + "language_loss": 0.76399004, + "learning_rate": 2.3287847802322678e-06, + "loss": 0.78102368, + "num_input_tokens_seen": 83027800, + "step": 3859, + "time_per_iteration": 2.5697624683380127 + }, + { + "auxiliary_loss_clip": 0.01149404, + "auxiliary_loss_mlp": 0.01032172, + "balance_loss_clip": 0.98674858, + "balance_loss_mlp": 1.02371407, + "epoch": 0.4641375578668911, + "flos": 26067053214720.0, + "grad_norm": 1.8373232824462746, + "language_loss": 0.84607005, + "learning_rate": 2.3280163815572723e-06, + "loss": 0.86788583, + "num_input_tokens_seen": 83048395, + "step": 3860, + "time_per_iteration": 2.7572293281555176 + }, + { + "auxiliary_loss_clip": 0.01136403, + "auxiliary_loss_mlp": 0.01032281, + "balance_loss_clip": 0.98175925, + "balance_loss_mlp": 1.0240078, + "epoch": 0.4642578007575302, + "flos": 19570382081280.0, + "grad_norm": 1.9389513940326202, + "language_loss": 0.77102774, + "learning_rate": 2.3272479331224522e-06, + "loss": 0.7927146, + "num_input_tokens_seen": 83065825, + "step": 3861, + "time_per_iteration": 2.66453218460083 + }, + { + "auxiliary_loss_clip": 0.01190565, + "auxiliary_loss_mlp": 0.01030686, + "balance_loss_clip": 1.06051159, + "balance_loss_mlp": 1.02240109, + "epoch": 0.4643780436481693, + "flos": 28186249772160.0, + "grad_norm": 1.6401896915458072, + "language_loss": 0.78038907, + "learning_rate": 2.3264794350443817e-06, + "loss": 0.80260158, + "num_input_tokens_seen": 83087920, + "step": 3862, + "time_per_iteration": 2.6501753330230713 + }, + { + "auxiliary_loss_clip": 0.0116504, + "auxiliary_loss_mlp": 0.0103299, + "balance_loss_clip": 1.01785421, + "balance_loss_mlp": 1.0248127, + "epoch": 0.46449828653880837, + "flos": 25375270204800.0, + "grad_norm": 1.7799059749111725, + "language_loss": 0.78403205, + "learning_rate": 2.3257108874396396e-06, + "loss": 0.80601239, + "num_input_tokens_seen": 83109015, + "step": 3863, + "time_per_iteration": 2.700594663619995 + }, + { + "auxiliary_loss_clip": 0.01140915, + "auxiliary_loss_mlp": 0.01030628, + "balance_loss_clip": 0.98200202, + "balance_loss_mlp": 1.0229423, + "epoch": 0.4646185294294475, + "flos": 16034330574720.0, + "grad_norm": 1.8573805030464667, + "language_loss": 0.73978794, + "learning_rate": 2.3249422904248152e-06, + "loss": 0.76150334, + "num_input_tokens_seen": 83127450, + "step": 3864, + "time_per_iteration": 4.587393283843994 + }, + { + "auxiliary_loss_clip": 0.01168652, + "auxiliary_loss_mlp": 0.01033852, + "balance_loss_clip": 1.02079773, + "balance_loss_mlp": 1.02565598, + "epoch": 0.4647387723200866, + "flos": 26363101109760.0, + "grad_norm": 1.5412835997205017, + "language_loss": 0.87103713, + "learning_rate": 2.324173644116504e-06, + "loss": 0.89306211, + "num_input_tokens_seen": 83150300, + "step": 3865, + "time_per_iteration": 2.692974090576172 + }, + { + "auxiliary_loss_clip": 0.01164037, + "auxiliary_loss_mlp": 0.01032302, + "balance_loss_clip": 1.02366495, + "balance_loss_mlp": 1.02425575, + "epoch": 0.46485901521072565, + "flos": 27160209774720.0, + "grad_norm": 1.718020805374166, + "language_loss": 0.81478691, + "learning_rate": 2.3234049486313087e-06, + "loss": 0.83675027, + "num_input_tokens_seen": 83171750, + "step": 3866, + "time_per_iteration": 3.568389415740967 + }, + { + "auxiliary_loss_clip": 0.01163039, + "auxiliary_loss_mlp": 0.01029282, + "balance_loss_clip": 1.02107978, + "balance_loss_mlp": 1.0213064, + "epoch": 0.46497925810136476, + "flos": 24279851088000.0, + "grad_norm": 3.2459443867997444, + "language_loss": 0.75960076, + "learning_rate": 2.322636204085839e-06, + "loss": 0.78152406, + "num_input_tokens_seen": 83191820, + "step": 3867, + "time_per_iteration": 2.6248600482940674 + }, + { + "auxiliary_loss_clip": 0.01132116, + "auxiliary_loss_mlp": 0.01029435, + "balance_loss_clip": 0.9778955, + "balance_loss_mlp": 1.0212636, + "epoch": 0.46509950099200387, + "flos": 16253134272000.0, + "grad_norm": 2.312640000485499, + "language_loss": 0.78837013, + "learning_rate": 2.3218674105967143e-06, + "loss": 0.80998564, + "num_input_tokens_seen": 83210085, + "step": 3868, + "time_per_iteration": 2.6769604682922363 + }, + { + "auxiliary_loss_clip": 0.0113324, + "auxiliary_loss_mlp": 0.0102844, + "balance_loss_clip": 0.97964025, + "balance_loss_mlp": 1.01974344, + "epoch": 0.4652197438826429, + "flos": 23442270773760.0, + "grad_norm": 1.6799510887844145, + "language_loss": 0.8367058, + "learning_rate": 2.3210985682805593e-06, + "loss": 0.85832262, + "num_input_tokens_seen": 83231865, + "step": 3869, + "time_per_iteration": 3.6320788860321045 + }, + { + "auxiliary_loss_clip": 0.01192951, + "auxiliary_loss_mlp": 0.01029012, + "balance_loss_clip": 1.06381774, + "balance_loss_mlp": 1.02087045, + "epoch": 0.46533998677328203, + "flos": 16216397637120.0, + "grad_norm": 2.475860650055944, + "language_loss": 0.68402123, + "learning_rate": 2.320329677254007e-06, + "loss": 0.70624083, + "num_input_tokens_seen": 83249195, + "step": 3870, + "time_per_iteration": 2.7126011848449707 + }, + { + "auxiliary_loss_clip": 0.01189522, + "auxiliary_loss_mlp": 0.01029966, + "balance_loss_clip": 1.06148171, + "balance_loss_mlp": 1.02193153, + "epoch": 0.46546022966392114, + "flos": 21141869080320.0, + "grad_norm": 2.6662595036805694, + "language_loss": 0.72740662, + "learning_rate": 2.319560737633697e-06, + "loss": 0.74960148, + "num_input_tokens_seen": 83267915, + "step": 3871, + "time_per_iteration": 2.5866775512695312 + }, + { + "auxiliary_loss_clip": 0.01121049, + "auxiliary_loss_mlp": 0.01030518, + "balance_loss_clip": 0.94065243, + "balance_loss_mlp": 1.02219105, + "epoch": 0.4655804725545602, + "flos": 41171942442240.0, + "grad_norm": 1.5674486227773152, + "language_loss": 0.68022466, + "learning_rate": 2.3187917495362775e-06, + "loss": 0.70174038, + "num_input_tokens_seen": 83292325, + "step": 3872, + "time_per_iteration": 2.895320177078247 + }, + { + "auxiliary_loss_clip": 0.01095542, + "auxiliary_loss_mlp": 0.01031366, + "balance_loss_clip": 0.90832847, + "balance_loss_mlp": 1.02297425, + "epoch": 0.4657007154451993, + "flos": 19570956698880.0, + "grad_norm": 2.8207131623161854, + "language_loss": 0.76553071, + "learning_rate": 2.318022713078403e-06, + "loss": 0.78679979, + "num_input_tokens_seen": 83306905, + "step": 3873, + "time_per_iteration": 2.7630622386932373 + }, + { + "auxiliary_loss_clip": 0.01143736, + "auxiliary_loss_mlp": 0.01032623, + "balance_loss_clip": 0.98444176, + "balance_loss_mlp": 1.02408206, + "epoch": 0.4658209583358384, + "flos": 15517826956800.0, + "grad_norm": 3.588578693491177, + "language_loss": 0.85494995, + "learning_rate": 2.3172536283767354e-06, + "loss": 0.87671351, + "num_input_tokens_seen": 83320665, + "step": 3874, + "time_per_iteration": 2.7224037647247314 + }, + { + "auxiliary_loss_clip": 0.01114178, + "auxiliary_loss_mlp": 0.01026506, + "balance_loss_clip": 0.94473213, + "balance_loss_mlp": 1.0181967, + "epoch": 0.4659412012264775, + "flos": 14903180403840.0, + "grad_norm": 2.892678691226127, + "language_loss": 0.80778193, + "learning_rate": 2.3164844955479447e-06, + "loss": 0.82918876, + "num_input_tokens_seen": 83336475, + "step": 3875, + "time_per_iteration": 2.6746742725372314 + }, + { + "auxiliary_loss_clip": 0.01100621, + "auxiliary_loss_mlp": 0.01029881, + "balance_loss_clip": 0.90450752, + "balance_loss_mlp": 1.02172065, + "epoch": 0.4660614441171166, + "flos": 24425612478720.0, + "grad_norm": 2.1799784994407156, + "language_loss": 0.70514333, + "learning_rate": 2.3157153147087082e-06, + "loss": 0.7264483, + "num_input_tokens_seen": 83358365, + "step": 3876, + "time_per_iteration": 2.786813735961914 + }, + { + "auxiliary_loss_clip": 0.011009, + "auxiliary_loss_mlp": 0.01033173, + "balance_loss_clip": 0.90705311, + "balance_loss_mlp": 1.02500129, + "epoch": 0.46618168700775564, + "flos": 22091095843200.0, + "grad_norm": 2.744266376217735, + "language_loss": 0.83125317, + "learning_rate": 2.314946085975709e-06, + "loss": 0.8525939, + "num_input_tokens_seen": 83377345, + "step": 3877, + "time_per_iteration": 2.7610902786254883 + }, + { + "auxiliary_loss_clip": 0.01102314, + "auxiliary_loss_mlp": 0.01036186, + "balance_loss_clip": 0.939632, + "balance_loss_mlp": 1.02865243, + "epoch": 0.46630192989839475, + "flos": 26176975810560.0, + "grad_norm": 2.4058447614475114, + "language_loss": 0.82407278, + "learning_rate": 2.3141768094656393e-06, + "loss": 0.84545779, + "num_input_tokens_seen": 83395920, + "step": 3878, + "time_per_iteration": 2.7437198162078857 + }, + { + "auxiliary_loss_clip": 0.01054108, + "auxiliary_loss_mlp": 0.01031926, + "balance_loss_clip": 0.82784557, + "balance_loss_mlp": 1.02297878, + "epoch": 0.46642217278903386, + "flos": 11509622150400.0, + "grad_norm": 2.6472712874215434, + "language_loss": 0.83074188, + "learning_rate": 2.3134074852951966e-06, + "loss": 0.8516022, + "num_input_tokens_seen": 83412510, + "step": 3879, + "time_per_iteration": 2.8838133811950684 + }, + { + "auxiliary_loss_clip": 0.01083745, + "auxiliary_loss_mlp": 0.01024194, + "balance_loss_clip": 0.89876676, + "balance_loss_mlp": 1.01623666, + "epoch": 0.4665424156796729, + "flos": 32306819299200.0, + "grad_norm": 1.6080467712858388, + "language_loss": 0.77752328, + "learning_rate": 2.312638113581088e-06, + "loss": 0.7986027, + "num_input_tokens_seen": 83432995, + "step": 3880, + "time_per_iteration": 3.0122249126434326 + }, + { + "auxiliary_loss_clip": 0.01162598, + "auxiliary_loss_mlp": 0.01030131, + "balance_loss_clip": 1.01814759, + "balance_loss_mlp": 1.02163172, + "epoch": 0.46666265857031203, + "flos": 18436179254400.0, + "grad_norm": 2.7358945354222914, + "language_loss": 0.77761328, + "learning_rate": 2.311868694440027e-06, + "loss": 0.79954052, + "num_input_tokens_seen": 83447415, + "step": 3881, + "time_per_iteration": 2.652575969696045 + }, + { + "auxiliary_loss_clip": 0.01099831, + "auxiliary_loss_mlp": 0.01007049, + "balance_loss_clip": 1.03912711, + "balance_loss_mlp": 1.0057261, + "epoch": 0.46678290146095114, + "flos": 68438989221120.0, + "grad_norm": 0.7344840880572455, + "language_loss": 0.62449116, + "learning_rate": 2.3110992279887323e-06, + "loss": 0.64555997, + "num_input_tokens_seen": 83519340, + "step": 3882, + "time_per_iteration": 3.2911152839660645 + }, + { + "auxiliary_loss_clip": 0.01122698, + "auxiliary_loss_mlp": 0.01027797, + "balance_loss_clip": 0.94583917, + "balance_loss_mlp": 1.01983964, + "epoch": 0.4669031443515902, + "flos": 17712507945600.0, + "grad_norm": 2.3049590901552985, + "language_loss": 0.84351087, + "learning_rate": 2.310329714343932e-06, + "loss": 0.86501575, + "num_input_tokens_seen": 83535490, + "step": 3883, + "time_per_iteration": 2.683561325073242 + }, + { + "auxiliary_loss_clip": 0.01138943, + "auxiliary_loss_mlp": 0.01026668, + "balance_loss_clip": 0.98462152, + "balance_loss_mlp": 1.01911592, + "epoch": 0.4670233872422293, + "flos": 23947748916480.0, + "grad_norm": 1.9244456217354635, + "language_loss": 0.81972373, + "learning_rate": 2.309560153622361e-06, + "loss": 0.84137982, + "num_input_tokens_seen": 83552400, + "step": 3884, + "time_per_iteration": 2.738241195678711 + }, + { + "auxiliary_loss_clip": 0.01115836, + "auxiliary_loss_mlp": 0.01023908, + "balance_loss_clip": 0.94447005, + "balance_loss_mlp": 1.01537561, + "epoch": 0.4671436301328684, + "flos": 28111268131200.0, + "grad_norm": 3.7890732758283945, + "language_loss": 0.74746001, + "learning_rate": 2.3087905459407602e-06, + "loss": 0.76885748, + "num_input_tokens_seen": 83571340, + "step": 3885, + "time_per_iteration": 2.737546682357788 + }, + { + "auxiliary_loss_clip": 0.01081256, + "auxiliary_loss_mlp": 0.01003906, + "balance_loss_clip": 1.00171804, + "balance_loss_mlp": 1.00254667, + "epoch": 0.46726387302350747, + "flos": 69369684566400.0, + "grad_norm": 0.80156302202344, + "language_loss": 0.6288116, + "learning_rate": 2.3080208914158795e-06, + "loss": 0.64966321, + "num_input_tokens_seen": 83634340, + "step": 3886, + "time_per_iteration": 3.2694127559661865 + }, + { + "auxiliary_loss_clip": 0.01145789, + "auxiliary_loss_mlp": 0.01028149, + "balance_loss_clip": 0.98957771, + "balance_loss_mlp": 1.01952398, + "epoch": 0.4673841159141466, + "flos": 25519666878720.0, + "grad_norm": 2.4173497523150407, + "language_loss": 0.72488415, + "learning_rate": 2.3072511901644753e-06, + "loss": 0.74662352, + "num_input_tokens_seen": 83653410, + "step": 3887, + "time_per_iteration": 2.7300310134887695 + }, + { + "auxiliary_loss_clip": 0.01188531, + "auxiliary_loss_mlp": 0.01028764, + "balance_loss_clip": 1.06217515, + "balance_loss_mlp": 1.02089572, + "epoch": 0.4675043588047857, + "flos": 24499265316480.0, + "grad_norm": 2.0658855244576646, + "language_loss": 0.8065486, + "learning_rate": 2.306481442303309e-06, + "loss": 0.82872164, + "num_input_tokens_seen": 83672985, + "step": 3888, + "time_per_iteration": 2.6454503536224365 + }, + { + "auxiliary_loss_clip": 0.01166696, + "auxiliary_loss_mlp": 0.01028142, + "balance_loss_clip": 1.0206064, + "balance_loss_mlp": 1.01985741, + "epoch": 0.46762460169542475, + "flos": 20960771685120.0, + "grad_norm": 2.5571955937671307, + "language_loss": 0.73356247, + "learning_rate": 2.3057116479491515e-06, + "loss": 0.75551087, + "num_input_tokens_seen": 83692395, + "step": 3889, + "time_per_iteration": 3.586402654647827 + }, + { + "auxiliary_loss_clip": 0.01157826, + "auxiliary_loss_mlp": 0.01029754, + "balance_loss_clip": 1.016469, + "balance_loss_mlp": 1.02188039, + "epoch": 0.46774484458606386, + "flos": 19171666137600.0, + "grad_norm": 1.8929344021186725, + "language_loss": 0.76091886, + "learning_rate": 2.30494180721878e-06, + "loss": 0.78279459, + "num_input_tokens_seen": 83709735, + "step": 3890, + "time_per_iteration": 3.5751700401306152 + }, + { + "auxiliary_loss_clip": 0.01163133, + "auxiliary_loss_mlp": 0.01026948, + "balance_loss_clip": 1.02136707, + "balance_loss_mlp": 1.01999497, + "epoch": 0.4678650874767029, + "flos": 17967689141760.0, + "grad_norm": 4.071120653844792, + "language_loss": 0.89913124, + "learning_rate": 2.3041719202289794e-06, + "loss": 0.92103207, + "num_input_tokens_seen": 83725910, + "step": 3891, + "time_per_iteration": 2.6239562034606934 + }, + { + "auxiliary_loss_clip": 0.01164554, + "auxiliary_loss_mlp": 0.01023785, + "balance_loss_clip": 1.02286339, + "balance_loss_mlp": 1.01655459, + "epoch": 0.467985330367342, + "flos": 21360816432000.0, + "grad_norm": 2.1435255900563077, + "language_loss": 0.80720448, + "learning_rate": 2.30340198709654e-06, + "loss": 0.82908785, + "num_input_tokens_seen": 83745745, + "step": 3892, + "time_per_iteration": 3.6052286624908447 + }, + { + "auxiliary_loss_clip": 0.01143779, + "auxiliary_loss_mlp": 0.01034965, + "balance_loss_clip": 0.98050916, + "balance_loss_mlp": 1.02697825, + "epoch": 0.46810557325798113, + "flos": 20521835487360.0, + "grad_norm": 1.8991648172598388, + "language_loss": 0.74547875, + "learning_rate": 2.3026320079382605e-06, + "loss": 0.76726615, + "num_input_tokens_seen": 83762680, + "step": 3893, + "time_per_iteration": 2.6656386852264404 + }, + { + "auxiliary_loss_clip": 0.011858, + "auxiliary_loss_mlp": 0.01024769, + "balance_loss_clip": 1.06038761, + "balance_loss_mlp": 1.0169127, + "epoch": 0.4682258161486202, + "flos": 30117848572800.0, + "grad_norm": 2.2805318418249305, + "language_loss": 0.76351213, + "learning_rate": 2.3018619828709454e-06, + "loss": 0.78561783, + "num_input_tokens_seen": 83784220, + "step": 3894, + "time_per_iteration": 2.7320902347564697 + }, + { + "auxiliary_loss_clip": 0.01166515, + "auxiliary_loss_mlp": 0.00510053, + "balance_loss_clip": 1.02749801, + "balance_loss_mlp": 0.0, + "epoch": 0.4683460590392593, + "flos": 25293357239040.0, + "grad_norm": 1.8625242174588428, + "language_loss": 0.82572365, + "learning_rate": 2.3010919120114084e-06, + "loss": 0.8424893, + "num_input_tokens_seen": 83800750, + "step": 3895, + "time_per_iteration": 3.632716417312622 + }, + { + "auxiliary_loss_clip": 0.01157648, + "auxiliary_loss_mlp": 0.01035577, + "balance_loss_clip": 1.01635396, + "balance_loss_mlp": 1.02736914, + "epoch": 0.4684663019298984, + "flos": 15368330551680.0, + "grad_norm": 2.416762880399394, + "language_loss": 0.66288388, + "learning_rate": 2.3003217954764672e-06, + "loss": 0.68481612, + "num_input_tokens_seen": 83815455, + "step": 3896, + "time_per_iteration": 2.663156747817993 + }, + { + "auxiliary_loss_clip": 0.01165228, + "auxiliary_loss_mlp": 0.01026675, + "balance_loss_clip": 1.01903057, + "balance_loss_mlp": 1.01803184, + "epoch": 0.46858654482053747, + "flos": 27778842737280.0, + "grad_norm": 1.6622883018504548, + "language_loss": 0.79460716, + "learning_rate": 2.299551633382949e-06, + "loss": 0.81652617, + "num_input_tokens_seen": 83835765, + "step": 3897, + "time_per_iteration": 2.7520487308502197 + }, + { + "auxiliary_loss_clip": 0.01133171, + "auxiliary_loss_mlp": 0.01032913, + "balance_loss_clip": 0.98079145, + "balance_loss_mlp": 1.02441978, + "epoch": 0.4687067877111766, + "flos": 18040623707520.0, + "grad_norm": 1.8829737219941531, + "language_loss": 0.86032367, + "learning_rate": 2.2987814258476854e-06, + "loss": 0.88198453, + "num_input_tokens_seen": 83853565, + "step": 3898, + "time_per_iteration": 2.6130125522613525 + }, + { + "auxiliary_loss_clip": 0.01088289, + "auxiliary_loss_mlp": 0.0103175, + "balance_loss_clip": 0.89700902, + "balance_loss_mlp": 1.02329826, + "epoch": 0.4688270306018157, + "flos": 16977380198400.0, + "grad_norm": 2.651582944404567, + "language_loss": 0.68001974, + "learning_rate": 2.2980111729875177e-06, + "loss": 0.70122015, + "num_input_tokens_seen": 83869815, + "step": 3899, + "time_per_iteration": 2.7256221771240234 + }, + { + "auxiliary_loss_clip": 0.01133615, + "auxiliary_loss_mlp": 0.01028084, + "balance_loss_clip": 0.98133349, + "balance_loss_mlp": 1.02054369, + "epoch": 0.46894727349245474, + "flos": 17821640442240.0, + "grad_norm": 3.049773706068019, + "language_loss": 0.81851578, + "learning_rate": 2.2972408749192917e-06, + "loss": 0.84013283, + "num_input_tokens_seen": 83887545, + "step": 3900, + "time_per_iteration": 2.649367332458496 + }, + { + "auxiliary_loss_clip": 0.01159871, + "auxiliary_loss_mlp": 0.00509898, + "balance_loss_clip": 1.0210129, + "balance_loss_mlp": 0.0, + "epoch": 0.46906751638309385, + "flos": 21471349559040.0, + "grad_norm": 1.8442442385683155, + "language_loss": 0.66871995, + "learning_rate": 2.296470531759861e-06, + "loss": 0.68541765, + "num_input_tokens_seen": 83905645, + "step": 3901, + "time_per_iteration": 2.769787311553955 + }, + { + "auxiliary_loss_clip": 0.01110065, + "auxiliary_loss_mlp": 0.01031854, + "balance_loss_clip": 0.94161963, + "balance_loss_mlp": 1.02411115, + "epoch": 0.46918775927373296, + "flos": 20337829090560.0, + "grad_norm": 7.756979015532881, + "language_loss": 0.79312807, + "learning_rate": 2.2957001436260866e-06, + "loss": 0.8145473, + "num_input_tokens_seen": 83922705, + "step": 3902, + "time_per_iteration": 2.726672887802124 + }, + { + "auxiliary_loss_clip": 0.01138593, + "auxiliary_loss_mlp": 0.01026744, + "balance_loss_clip": 0.98413658, + "balance_loss_mlp": 1.01906705, + "epoch": 0.469308002164372, + "flos": 18403249461120.0, + "grad_norm": 1.5727569758992321, + "language_loss": 0.73112947, + "learning_rate": 2.294929710634836e-06, + "loss": 0.75278282, + "num_input_tokens_seen": 83940795, + "step": 3903, + "time_per_iteration": 2.67569637298584 + }, + { + "auxiliary_loss_clip": 0.01161899, + "auxiliary_loss_mlp": 0.01026966, + "balance_loss_clip": 1.01904154, + "balance_loss_mlp": 1.01924717, + "epoch": 0.46942824505501113, + "flos": 37962067363200.0, + "grad_norm": 1.6924861708215178, + "language_loss": 0.60984564, + "learning_rate": 2.2941592329029823e-06, + "loss": 0.63173431, + "num_input_tokens_seen": 83961900, + "step": 3904, + "time_per_iteration": 2.784085273742676 + }, + { + "auxiliary_loss_clip": 0.01158984, + "auxiliary_loss_mlp": 0.01032648, + "balance_loss_clip": 1.02038121, + "balance_loss_mlp": 1.02427959, + "epoch": 0.46954848794565024, + "flos": 21872507627520.0, + "grad_norm": 2.3070767720595313, + "language_loss": 0.78922045, + "learning_rate": 2.2933887105474067e-06, + "loss": 0.81113672, + "num_input_tokens_seen": 83980075, + "step": 3905, + "time_per_iteration": 2.712634325027466 + }, + { + "auxiliary_loss_clip": 0.01159425, + "auxiliary_loss_mlp": 0.01030732, + "balance_loss_clip": 1.02194571, + "balance_loss_mlp": 1.02280474, + "epoch": 0.4696687308362893, + "flos": 22016545165440.0, + "grad_norm": 1.5968766012735782, + "language_loss": 0.81401575, + "learning_rate": 2.2926181436849974e-06, + "loss": 0.83591723, + "num_input_tokens_seen": 83999430, + "step": 3906, + "time_per_iteration": 2.6732659339904785 + }, + { + "auxiliary_loss_clip": 0.01161303, + "auxiliary_loss_mlp": 0.01030739, + "balance_loss_clip": 1.02126014, + "balance_loss_mlp": 1.02241194, + "epoch": 0.4697889737269284, + "flos": 21613663244160.0, + "grad_norm": 1.7827454123801643, + "language_loss": 0.72877181, + "learning_rate": 2.2918475324326478e-06, + "loss": 0.75069225, + "num_input_tokens_seen": 84019150, + "step": 3907, + "time_per_iteration": 2.7015204429626465 + }, + { + "auxiliary_loss_clip": 0.01166132, + "auxiliary_loss_mlp": 0.00510462, + "balance_loss_clip": 1.02206969, + "balance_loss_mlp": 0.0, + "epoch": 0.46990921661756746, + "flos": 25228323665280.0, + "grad_norm": 2.190141279573483, + "language_loss": 0.91192627, + "learning_rate": 2.2910768769072603e-06, + "loss": 0.92869222, + "num_input_tokens_seen": 84037930, + "step": 3908, + "time_per_iteration": 2.7654221057891846 + }, + { + "auxiliary_loss_clip": 0.01156346, + "auxiliary_loss_mlp": 0.01023761, + "balance_loss_clip": 1.01887798, + "balance_loss_mlp": 1.0163877, + "epoch": 0.47002945950820657, + "flos": 13844031045120.0, + "grad_norm": 2.4052490859198787, + "language_loss": 0.76292616, + "learning_rate": 2.2903061772257417e-06, + "loss": 0.78472722, + "num_input_tokens_seen": 84055915, + "step": 3909, + "time_per_iteration": 2.6700539588928223 + }, + { + "auxiliary_loss_clip": 0.01165581, + "auxiliary_loss_mlp": 0.01025156, + "balance_loss_clip": 1.02402449, + "balance_loss_mlp": 1.01771736, + "epoch": 0.4701497023988457, + "flos": 26247001374720.0, + "grad_norm": 1.6035389061024374, + "language_loss": 0.78637409, + "learning_rate": 2.289535433505007e-06, + "loss": 0.80828142, + "num_input_tokens_seen": 84077270, + "step": 3910, + "time_per_iteration": 2.719921588897705 + }, + { + "auxiliary_loss_clip": 0.01142835, + "auxiliary_loss_mlp": 0.01028977, + "balance_loss_clip": 0.98226535, + "balance_loss_mlp": 1.02110946, + "epoch": 0.47026994528948474, + "flos": 25629517647360.0, + "grad_norm": 1.8077390632295618, + "language_loss": 0.63659859, + "learning_rate": 2.2887646458619767e-06, + "loss": 0.65831673, + "num_input_tokens_seen": 84098635, + "step": 3911, + "time_per_iteration": 2.7053914070129395 + }, + { + "auxiliary_loss_clip": 0.01123808, + "auxiliary_loss_mlp": 0.01027637, + "balance_loss_clip": 0.9470259, + "balance_loss_mlp": 1.01919067, + "epoch": 0.47039018818012385, + "flos": 20554406144640.0, + "grad_norm": 1.8950122497321449, + "language_loss": 0.767506, + "learning_rate": 2.2879938144135797e-06, + "loss": 0.78902048, + "num_input_tokens_seen": 84114740, + "step": 3912, + "time_per_iteration": 2.7285397052764893 + }, + { + "auxiliary_loss_clip": 0.01115083, + "auxiliary_loss_mlp": 0.00509476, + "balance_loss_clip": 0.94432628, + "balance_loss_mlp": 0.0, + "epoch": 0.47051043107076296, + "flos": 21577249831680.0, + "grad_norm": 1.7624048783470936, + "language_loss": 0.75452149, + "learning_rate": 2.2872229392767496e-06, + "loss": 0.77076709, + "num_input_tokens_seen": 84134845, + "step": 3913, + "time_per_iteration": 2.7362728118896484 + }, + { + "auxiliary_loss_clip": 0.01167796, + "auxiliary_loss_mlp": 0.01034626, + "balance_loss_clip": 1.02248442, + "balance_loss_mlp": 1.026842, + "epoch": 0.470630673961402, + "flos": 18953185662720.0, + "grad_norm": 1.6610976014318755, + "language_loss": 0.7496562, + "learning_rate": 2.286452020568428e-06, + "loss": 0.77168041, + "num_input_tokens_seen": 84152920, + "step": 3914, + "time_per_iteration": 2.718353271484375 + }, + { + "auxiliary_loss_clip": 0.01190763, + "auxiliary_loss_mlp": 0.01025401, + "balance_loss_clip": 1.06027937, + "balance_loss_mlp": 1.01675236, + "epoch": 0.4707509168520411, + "flos": 19938969492480.0, + "grad_norm": 2.5544384108967364, + "language_loss": 0.73126447, + "learning_rate": 2.2856810584055637e-06, + "loss": 0.75342613, + "num_input_tokens_seen": 84170455, + "step": 3915, + "time_per_iteration": 3.57853627204895 + }, + { + "auxiliary_loss_clip": 0.01165017, + "auxiliary_loss_mlp": 0.01025407, + "balance_loss_clip": 1.02203465, + "balance_loss_mlp": 1.01792037, + "epoch": 0.47087115974268023, + "flos": 40118754741120.0, + "grad_norm": 1.4804000241121416, + "language_loss": 0.67737055, + "learning_rate": 2.2849100529051085e-06, + "loss": 0.69927478, + "num_input_tokens_seen": 84197390, + "step": 3916, + "time_per_iteration": 2.8052964210510254 + }, + { + "auxiliary_loss_clip": 0.01186057, + "auxiliary_loss_mlp": 0.01022801, + "balance_loss_clip": 1.06077147, + "balance_loss_mlp": 1.01504612, + "epoch": 0.4709914026333193, + "flos": 13552723745280.0, + "grad_norm": 2.1931654320386476, + "language_loss": 0.79980636, + "learning_rate": 2.284139004184026e-06, + "loss": 0.82189494, + "num_input_tokens_seen": 84214620, + "step": 3917, + "time_per_iteration": 3.5236213207244873 + }, + { + "auxiliary_loss_clip": 0.01187315, + "auxiliary_loss_mlp": 0.01028878, + "balance_loss_clip": 1.06045151, + "balance_loss_mlp": 1.0207777, + "epoch": 0.4711116455239584, + "flos": 19974628719360.0, + "grad_norm": 2.198175531057112, + "language_loss": 0.743976, + "learning_rate": 2.2833679123592814e-06, + "loss": 0.7661379, + "num_input_tokens_seen": 84231880, + "step": 3918, + "time_per_iteration": 2.5860843658447266 + }, + { + "auxiliary_loss_clip": 0.01137023, + "auxiliary_loss_mlp": 0.01029802, + "balance_loss_clip": 0.98255229, + "balance_loss_mlp": 1.02199411, + "epoch": 0.4712318884145975, + "flos": 32124824064000.0, + "grad_norm": 1.6392333364634233, + "language_loss": 0.63394535, + "learning_rate": 2.2825967775478508e-06, + "loss": 0.6556136, + "num_input_tokens_seen": 84252980, + "step": 3919, + "time_per_iteration": 2.802968978881836 + }, + { + "auxiliary_loss_clip": 0.01187663, + "auxiliary_loss_mlp": 0.01029615, + "balance_loss_clip": 1.06014276, + "balance_loss_mlp": 1.0218426, + "epoch": 0.47135213130523657, + "flos": 20047850593920.0, + "grad_norm": 3.037662388243686, + "language_loss": 0.83698386, + "learning_rate": 2.2818255998667135e-06, + "loss": 0.85915667, + "num_input_tokens_seen": 84271490, + "step": 3920, + "time_per_iteration": 2.5494515895843506 + }, + { + "auxiliary_loss_clip": 0.01162963, + "auxiliary_loss_mlp": 0.01030553, + "balance_loss_clip": 1.02211726, + "balance_loss_mlp": 1.02297711, + "epoch": 0.4714723741958757, + "flos": 19426990988160.0, + "grad_norm": 1.6520234301019951, + "language_loss": 0.78943908, + "learning_rate": 2.2810543794328566e-06, + "loss": 0.81137425, + "num_input_tokens_seen": 84290525, + "step": 3921, + "time_per_iteration": 3.59260892868042 + }, + { + "auxiliary_loss_clip": 0.01164787, + "auxiliary_loss_mlp": 0.01028856, + "balance_loss_clip": 1.02070987, + "balance_loss_mlp": 1.0203681, + "epoch": 0.4715926170865148, + "flos": 20373883367040.0, + "grad_norm": 2.0570115905829027, + "language_loss": 0.82672048, + "learning_rate": 2.2802831163632735e-06, + "loss": 0.84865689, + "num_input_tokens_seen": 84309245, + "step": 3922, + "time_per_iteration": 2.696458101272583 + }, + { + "auxiliary_loss_clip": 0.01063224, + "auxiliary_loss_mlp": 0.01030944, + "balance_loss_clip": 0.86169112, + "balance_loss_mlp": 1.02234888, + "epoch": 0.47171285997715384, + "flos": 22672884430080.0, + "grad_norm": 1.8809890512156884, + "language_loss": 0.74556679, + "learning_rate": 2.279511810774965e-06, + "loss": 0.76650846, + "num_input_tokens_seen": 84330775, + "step": 3923, + "time_per_iteration": 2.771249294281006 + }, + { + "auxiliary_loss_clip": 0.01186445, + "auxiliary_loss_mlp": 0.0102437, + "balance_loss_clip": 1.05921233, + "balance_loss_mlp": 1.01683617, + "epoch": 0.47183310286779295, + "flos": 21105419754240.0, + "grad_norm": 2.0888024377380923, + "language_loss": 0.71478593, + "learning_rate": 2.2787404627849364e-06, + "loss": 0.73689413, + "num_input_tokens_seen": 84349985, + "step": 3924, + "time_per_iteration": 2.6334996223449707 + }, + { + "auxiliary_loss_clip": 0.0113767, + "auxiliary_loss_mlp": 0.01025038, + "balance_loss_clip": 0.98172146, + "balance_loss_mlp": 1.01729202, + "epoch": 0.471953345758432, + "flos": 21726566668800.0, + "grad_norm": 1.755854293267306, + "language_loss": 0.79050094, + "learning_rate": 2.277969072510202e-06, + "loss": 0.81212807, + "num_input_tokens_seen": 84368965, + "step": 3925, + "time_per_iteration": 2.7000792026519775 + }, + { + "auxiliary_loss_clip": 0.01136472, + "auxiliary_loss_mlp": 0.01032839, + "balance_loss_clip": 0.98153573, + "balance_loss_mlp": 1.02527857, + "epoch": 0.4720735886490711, + "flos": 19861078849920.0, + "grad_norm": 2.191763743045183, + "language_loss": 0.81596529, + "learning_rate": 2.2771976400677803e-06, + "loss": 0.83765841, + "num_input_tokens_seen": 84387795, + "step": 3926, + "time_per_iteration": 2.712458372116089 + }, + { + "auxiliary_loss_clip": 0.01074392, + "auxiliary_loss_mlp": 0.0102821, + "balance_loss_clip": 0.89647138, + "balance_loss_mlp": 1.02058077, + "epoch": 0.47219383153971023, + "flos": 19171809792000.0, + "grad_norm": 1.6794533105790117, + "language_loss": 0.7886368, + "learning_rate": 2.2764261655746965e-06, + "loss": 0.80966288, + "num_input_tokens_seen": 84405290, + "step": 3927, + "time_per_iteration": 2.7005035877227783 + }, + { + "auxiliary_loss_clip": 0.01113406, + "auxiliary_loss_mlp": 0.01031747, + "balance_loss_clip": 0.94397449, + "balance_loss_mlp": 1.02257955, + "epoch": 0.4723140744303493, + "flos": 23224005780480.0, + "grad_norm": 1.7785449527453083, + "language_loss": 0.76174045, + "learning_rate": 2.2756546491479832e-06, + "loss": 0.78319198, + "num_input_tokens_seen": 84426205, + "step": 3928, + "time_per_iteration": 2.7181894779205322 + }, + { + "auxiliary_loss_clip": 0.01186964, + "auxiliary_loss_mlp": 0.00510221, + "balance_loss_clip": 1.05817509, + "balance_loss_mlp": 0.0, + "epoch": 0.4724343173209884, + "flos": 18223265387520.0, + "grad_norm": 2.333858620352356, + "language_loss": 0.80386794, + "learning_rate": 2.274883090904679e-06, + "loss": 0.82083976, + "num_input_tokens_seen": 84443970, + "step": 3929, + "time_per_iteration": 2.6074695587158203 + }, + { + "auxiliary_loss_clip": 0.01188289, + "auxiliary_loss_mlp": 0.01035491, + "balance_loss_clip": 1.06099343, + "balance_loss_mlp": 1.02709866, + "epoch": 0.4725545602116275, + "flos": 21251037490560.0, + "grad_norm": 2.113191271346983, + "language_loss": 0.68139815, + "learning_rate": 2.2741114909618283e-06, + "loss": 0.70363593, + "num_input_tokens_seen": 84459865, + "step": 3930, + "time_per_iteration": 2.6561765670776367 + }, + { + "auxiliary_loss_clip": 0.01114695, + "auxiliary_loss_mlp": 0.01025201, + "balance_loss_clip": 0.94343734, + "balance_loss_mlp": 1.01808405, + "epoch": 0.47267480310226656, + "flos": 21434002392960.0, + "grad_norm": 1.7302935694504111, + "language_loss": 0.72000724, + "learning_rate": 2.2733398494364828e-06, + "loss": 0.7414062, + "num_input_tokens_seen": 84479110, + "step": 3931, + "time_per_iteration": 2.7560834884643555 + }, + { + "auxiliary_loss_clip": 0.01126718, + "auxiliary_loss_mlp": 0.01028646, + "balance_loss_clip": 0.97781539, + "balance_loss_mlp": 1.02067113, + "epoch": 0.47279504599290567, + "flos": 18770508069120.0, + "grad_norm": 2.1049935782692173, + "language_loss": 0.84547663, + "learning_rate": 2.272568166445699e-06, + "loss": 0.86703026, + "num_input_tokens_seen": 84497675, + "step": 3932, + "time_per_iteration": 2.7072484493255615 + }, + { + "auxiliary_loss_clip": 0.01163042, + "auxiliary_loss_mlp": 0.0102697, + "balance_loss_clip": 1.02093887, + "balance_loss_mlp": 1.01886392, + "epoch": 0.4729152888835448, + "flos": 21105742976640.0, + "grad_norm": 2.5121224565477087, + "language_loss": 0.64260793, + "learning_rate": 2.271796442106541e-06, + "loss": 0.66450804, + "num_input_tokens_seen": 84517030, + "step": 3933, + "time_per_iteration": 2.6917879581451416 + }, + { + "auxiliary_loss_clip": 0.01032453, + "auxiliary_loss_mlp": 0.01003815, + "balance_loss_clip": 0.92323112, + "balance_loss_mlp": 1.00236082, + "epoch": 0.47303553177418384, + "flos": 70201877840640.0, + "grad_norm": 0.8052386852478928, + "language_loss": 0.56525373, + "learning_rate": 2.271024676536079e-06, + "loss": 0.58561647, + "num_input_tokens_seen": 84577290, + "step": 3934, + "time_per_iteration": 3.2422053813934326 + }, + { + "auxiliary_loss_clip": 0.01144512, + "auxiliary_loss_mlp": 0.01030696, + "balance_loss_clip": 0.98471349, + "balance_loss_mlp": 1.02216613, + "epoch": 0.47315577466482295, + "flos": 22455122227200.0, + "grad_norm": 1.9069195638903156, + "language_loss": 0.73439789, + "learning_rate": 2.2702528698513894e-06, + "loss": 0.75615001, + "num_input_tokens_seen": 84598415, + "step": 3935, + "time_per_iteration": 2.7337992191314697 + }, + { + "auxiliary_loss_clip": 0.01136648, + "auxiliary_loss_mlp": 0.01029208, + "balance_loss_clip": 0.97751284, + "balance_loss_mlp": 1.02065432, + "epoch": 0.47327601755546206, + "flos": 24352857480960.0, + "grad_norm": 2.1531001928743145, + "language_loss": 0.78655779, + "learning_rate": 2.269481022169554e-06, + "loss": 0.80821639, + "num_input_tokens_seen": 84617010, + "step": 3936, + "time_per_iteration": 2.700035572052002 + }, + { + "auxiliary_loss_clip": 0.01145533, + "auxiliary_loss_mlp": 0.01026161, + "balance_loss_clip": 0.98123187, + "balance_loss_mlp": 1.01785207, + "epoch": 0.4733962604461011, + "flos": 22926772736640.0, + "grad_norm": 4.9230959264453125, + "language_loss": 0.80004275, + "learning_rate": 2.2687091336076614e-06, + "loss": 0.82175964, + "num_input_tokens_seen": 84636350, + "step": 3937, + "time_per_iteration": 2.7535619735717773 + }, + { + "auxiliary_loss_clip": 0.01159117, + "auxiliary_loss_mlp": 0.01026816, + "balance_loss_clip": 1.01925826, + "balance_loss_mlp": 1.01933265, + "epoch": 0.4735165033367402, + "flos": 18327369980160.0, + "grad_norm": 1.829574054435468, + "language_loss": 0.79957533, + "learning_rate": 2.267937204282807e-06, + "loss": 0.82143462, + "num_input_tokens_seen": 84653490, + "step": 3938, + "time_per_iteration": 2.613262176513672 + }, + { + "auxiliary_loss_clip": 0.01168564, + "auxiliary_loss_mlp": 0.01030478, + "balance_loss_clip": 1.02192521, + "balance_loss_mlp": 1.02190113, + "epoch": 0.4736367462273793, + "flos": 23037018554880.0, + "grad_norm": 2.3779310911618166, + "language_loss": 0.79077232, + "learning_rate": 2.2671652343120926e-06, + "loss": 0.81276274, + "num_input_tokens_seen": 84673965, + "step": 3939, + "time_per_iteration": 2.6914710998535156 + }, + { + "auxiliary_loss_clip": 0.01186303, + "auxiliary_loss_mlp": 0.01024024, + "balance_loss_clip": 1.06093502, + "balance_loss_mlp": 1.0157274, + "epoch": 0.4737569891180184, + "flos": 25374336451200.0, + "grad_norm": 1.7218802683387984, + "language_loss": 0.80633175, + "learning_rate": 2.2663932238126236e-06, + "loss": 0.82843494, + "num_input_tokens_seen": 84692525, + "step": 3940, + "time_per_iteration": 2.6297667026519775 + }, + { + "auxiliary_loss_clip": 0.01161627, + "auxiliary_loss_mlp": 0.01027324, + "balance_loss_clip": 1.01884031, + "balance_loss_mlp": 1.0193367, + "epoch": 0.4738772320086575, + "flos": 25849326925440.0, + "grad_norm": 2.5507380477076205, + "language_loss": 0.80360699, + "learning_rate": 2.265621172901515e-06, + "loss": 0.82549655, + "num_input_tokens_seen": 84715640, + "step": 3941, + "time_per_iteration": 3.5631048679351807 + }, + { + "auxiliary_loss_clip": 0.01188785, + "auxiliary_loss_mlp": 0.01034496, + "balance_loss_clip": 1.0612011, + "balance_loss_mlp": 1.02644944, + "epoch": 0.47399747489929656, + "flos": 27564420499200.0, + "grad_norm": 2.1675281405514157, + "language_loss": 0.71287489, + "learning_rate": 2.2648490816958854e-06, + "loss": 0.73510766, + "num_input_tokens_seen": 84736635, + "step": 3942, + "time_per_iteration": 3.5997354984283447 + }, + { + "auxiliary_loss_clip": 0.01158497, + "auxiliary_loss_mlp": 0.01025644, + "balance_loss_clip": 1.01654637, + "balance_loss_mlp": 1.01703131, + "epoch": 0.47411771778993567, + "flos": 24863650836480.0, + "grad_norm": 2.492488029618905, + "language_loss": 0.73380244, + "learning_rate": 2.264076950312861e-06, + "loss": 0.7556439, + "num_input_tokens_seen": 84755445, + "step": 3943, + "time_per_iteration": 2.627070903778076 + }, + { + "auxiliary_loss_clip": 0.01142165, + "auxiliary_loss_mlp": 0.01032356, + "balance_loss_clip": 0.98143679, + "balance_loss_mlp": 1.02410674, + "epoch": 0.4742379606805748, + "flos": 22748009725440.0, + "grad_norm": 2.6399530617387756, + "language_loss": 0.8228516, + "learning_rate": 2.2633047788695727e-06, + "loss": 0.8445968, + "num_input_tokens_seen": 84775750, + "step": 3944, + "time_per_iteration": 3.6111443042755127 + }, + { + "auxiliary_loss_clip": 0.01138022, + "auxiliary_loss_mlp": 0.01026103, + "balance_loss_clip": 0.98174423, + "balance_loss_mlp": 1.01833391, + "epoch": 0.47435820357121383, + "flos": 19681130689920.0, + "grad_norm": 1.9373178863317697, + "language_loss": 0.63980758, + "learning_rate": 2.262532567483159e-06, + "loss": 0.66144884, + "num_input_tokens_seen": 84794310, + "step": 3945, + "time_per_iteration": 2.7058234214782715 + }, + { + "auxiliary_loss_clip": 0.01187752, + "auxiliary_loss_mlp": 0.00509762, + "balance_loss_clip": 1.05953074, + "balance_loss_mlp": 0.0, + "epoch": 0.47447844646185294, + "flos": 25228718714880.0, + "grad_norm": 1.931805463566412, + "language_loss": 0.80349243, + "learning_rate": 2.2617603162707635e-06, + "loss": 0.82046759, + "num_input_tokens_seen": 84814720, + "step": 3946, + "time_per_iteration": 2.6062169075012207 + }, + { + "auxiliary_loss_clip": 0.01184879, + "auxiliary_loss_mlp": 0.01026458, + "balance_loss_clip": 1.05845094, + "balance_loss_mlp": 1.0186919, + "epoch": 0.47459868935249205, + "flos": 24570619683840.0, + "grad_norm": 1.7561327149942658, + "language_loss": 0.82924038, + "learning_rate": 2.2609880253495363e-06, + "loss": 0.85135376, + "num_input_tokens_seen": 84834355, + "step": 3947, + "time_per_iteration": 2.6360456943511963 + }, + { + "auxiliary_loss_clip": 0.01122674, + "auxiliary_loss_mlp": 0.01025139, + "balance_loss_clip": 0.94100404, + "balance_loss_mlp": 1.0170213, + "epoch": 0.4747189322431311, + "flos": 20558500295040.0, + "grad_norm": 2.02440888717144, + "language_loss": 0.86580354, + "learning_rate": 2.260215694836633e-06, + "loss": 0.88728166, + "num_input_tokens_seen": 84853530, + "step": 3948, + "time_per_iteration": 3.6107800006866455 + }, + { + "auxiliary_loss_clip": 0.01093766, + "auxiliary_loss_mlp": 0.00510183, + "balance_loss_clip": 0.90312451, + "balance_loss_mlp": 0.0, + "epoch": 0.4748391751337702, + "flos": 25995231970560.0, + "grad_norm": 2.428946255735966, + "language_loss": 0.6478951, + "learning_rate": 2.2594433248492157e-06, + "loss": 0.66393459, + "num_input_tokens_seen": 84872505, + "step": 3949, + "time_per_iteration": 2.8122384548187256 + }, + { + "auxiliary_loss_clip": 0.01168506, + "auxiliary_loss_mlp": 0.01031666, + "balance_loss_clip": 1.0197618, + "balance_loss_mlp": 1.02397072, + "epoch": 0.47495941802440933, + "flos": 22821052032000.0, + "grad_norm": 1.634469931689416, + "language_loss": 0.80328262, + "learning_rate": 2.2586709155044527e-06, + "loss": 0.82528436, + "num_input_tokens_seen": 84893105, + "step": 3950, + "time_per_iteration": 2.6819071769714355 + }, + { + "auxiliary_loss_clip": 0.01185042, + "auxiliary_loss_mlp": 0.01023475, + "balance_loss_clip": 1.05879545, + "balance_loss_mlp": 1.01589942, + "epoch": 0.4750796609150484, + "flos": 27891782075520.0, + "grad_norm": 1.473607850557119, + "language_loss": 0.76222217, + "learning_rate": 2.2578984669195167e-06, + "loss": 0.7843073, + "num_input_tokens_seen": 84914070, + "step": 3951, + "time_per_iteration": 2.6278281211853027 + }, + { + "auxiliary_loss_clip": 0.01158322, + "auxiliary_loss_mlp": 0.01024658, + "balance_loss_clip": 1.01618564, + "balance_loss_mlp": 1.01702309, + "epoch": 0.4751999038056875, + "flos": 35660085471360.0, + "grad_norm": 1.8491812659479085, + "language_loss": 0.67752981, + "learning_rate": 2.2571259792115887e-06, + "loss": 0.69935966, + "num_input_tokens_seen": 84935290, + "step": 3952, + "time_per_iteration": 2.825082540512085 + }, + { + "auxiliary_loss_clip": 0.01159255, + "auxiliary_loss_mlp": 0.01023677, + "balance_loss_clip": 1.02168083, + "balance_loss_mlp": 1.01582968, + "epoch": 0.4753201466963266, + "flos": 22090880361600.0, + "grad_norm": 2.0326395034666636, + "language_loss": 0.79024005, + "learning_rate": 2.2563534524978544e-06, + "loss": 0.81206942, + "num_input_tokens_seen": 84952760, + "step": 3953, + "time_per_iteration": 2.6516075134277344 + }, + { + "auxiliary_loss_clip": 0.01113035, + "auxiliary_loss_mlp": 0.01024342, + "balance_loss_clip": 0.94720763, + "balance_loss_mlp": 1.01693332, + "epoch": 0.47544038958696566, + "flos": 30190854965760.0, + "grad_norm": 1.5523636567435448, + "language_loss": 0.70622438, + "learning_rate": 2.2555808868955052e-06, + "loss": 0.72759819, + "num_input_tokens_seen": 84974890, + "step": 3954, + "time_per_iteration": 2.7604503631591797 + }, + { + "auxiliary_loss_clip": 0.01095026, + "auxiliary_loss_mlp": 0.01026697, + "balance_loss_clip": 0.90544617, + "balance_loss_mlp": 1.01905572, + "epoch": 0.47556063247760477, + "flos": 23472219738240.0, + "grad_norm": 2.135334402526831, + "language_loss": 0.73072696, + "learning_rate": 2.254808282521738e-06, + "loss": 0.75194418, + "num_input_tokens_seen": 84993640, + "step": 3955, + "time_per_iteration": 2.7797837257385254 + }, + { + "auxiliary_loss_clip": 0.01115836, + "auxiliary_loss_mlp": 0.005092, + "balance_loss_clip": 0.94123912, + "balance_loss_mlp": 0.0, + "epoch": 0.4756808753682438, + "flos": 25155209531520.0, + "grad_norm": 1.8213604585203949, + "language_loss": 0.81440353, + "learning_rate": 2.2540356394937573e-06, + "loss": 0.83065391, + "num_input_tokens_seen": 85012340, + "step": 3956, + "time_per_iteration": 2.725393295288086 + }, + { + "auxiliary_loss_clip": 0.0111835, + "auxiliary_loss_mlp": 0.01033042, + "balance_loss_clip": 0.94132745, + "balance_loss_mlp": 1.02462018, + "epoch": 0.47580111825888294, + "flos": 15669729573120.0, + "grad_norm": 2.1287476635290004, + "language_loss": 0.83884478, + "learning_rate": 2.253262957928772e-06, + "loss": 0.8603586, + "num_input_tokens_seen": 85029225, + "step": 3957, + "time_per_iteration": 2.6610922813415527 + }, + { + "auxiliary_loss_clip": 0.0113021, + "auxiliary_loss_mlp": 0.01031227, + "balance_loss_clip": 0.97578484, + "balance_loss_mlp": 1.02346087, + "epoch": 0.47592136114952205, + "flos": 17636556637440.0, + "grad_norm": 2.6095411741369188, + "language_loss": 0.71960938, + "learning_rate": 2.2524902379439976e-06, + "loss": 0.74122381, + "num_input_tokens_seen": 85047895, + "step": 3958, + "time_per_iteration": 2.6768224239349365 + }, + { + "auxiliary_loss_clip": 0.0099576, + "auxiliary_loss_mlp": 0.01010389, + "balance_loss_clip": 0.85723108, + "balance_loss_mlp": 1.008708, + "epoch": 0.4760416040401611, + "flos": 61417159292160.0, + "grad_norm": 0.7389521243312972, + "language_loss": 0.63664591, + "learning_rate": 2.251717479656655e-06, + "loss": 0.65670741, + "num_input_tokens_seen": 85112690, + "step": 3959, + "time_per_iteration": 3.375824451446533 + }, + { + "auxiliary_loss_clip": 0.01186312, + "auxiliary_loss_mlp": 0.01032582, + "balance_loss_clip": 1.05827332, + "balance_loss_mlp": 1.02486968, + "epoch": 0.4761618469308002, + "flos": 18405871153920.0, + "grad_norm": 1.9992441717893445, + "language_loss": 0.76719129, + "learning_rate": 2.2509446831839704e-06, + "loss": 0.78938019, + "num_input_tokens_seen": 85132130, + "step": 3960, + "time_per_iteration": 2.5891706943511963 + }, + { + "auxiliary_loss_clip": 0.01139353, + "auxiliary_loss_mlp": 0.01025537, + "balance_loss_clip": 0.97984409, + "balance_loss_mlp": 1.01816356, + "epoch": 0.4762820898214393, + "flos": 18040911016320.0, + "grad_norm": 2.2311803225207254, + "language_loss": 0.82081223, + "learning_rate": 2.250171848643177e-06, + "loss": 0.84246111, + "num_input_tokens_seen": 85149420, + "step": 3961, + "time_per_iteration": 2.6974141597747803 + }, + { + "auxiliary_loss_clip": 0.0113464, + "auxiliary_loss_mlp": 0.01027636, + "balance_loss_clip": 0.98206043, + "balance_loss_mlp": 1.01981568, + "epoch": 0.4764023327120784, + "flos": 19318253541120.0, + "grad_norm": 2.0338270100187077, + "language_loss": 0.85586095, + "learning_rate": 2.249398976151513e-06, + "loss": 0.87748373, + "num_input_tokens_seen": 85166970, + "step": 3962, + "time_per_iteration": 2.6520793437957764 + }, + { + "auxiliary_loss_clip": 0.01183802, + "auxiliary_loss_mlp": 0.01026994, + "balance_loss_clip": 1.05830979, + "balance_loss_mlp": 1.01964164, + "epoch": 0.4765225756027175, + "flos": 22747255539840.0, + "grad_norm": 3.1223018889304686, + "language_loss": 0.79245299, + "learning_rate": 2.248626065826223e-06, + "loss": 0.81456101, + "num_input_tokens_seen": 85185175, + "step": 3963, + "time_per_iteration": 2.676785469055176 + }, + { + "auxiliary_loss_clip": 0.01093213, + "auxiliary_loss_mlp": 0.01000675, + "balance_loss_clip": 1.03301549, + "balance_loss_mlp": 0.99937516, + "epoch": 0.4766428184933566, + "flos": 65933392106880.0, + "grad_norm": 0.7570975280777948, + "language_loss": 0.62554246, + "learning_rate": 2.2478531177845564e-06, + "loss": 0.64648134, + "num_input_tokens_seen": 85246170, + "step": 3964, + "time_per_iteration": 3.1943023204803467 + }, + { + "auxiliary_loss_clip": 0.0114194, + "auxiliary_loss_mlp": 0.01025353, + "balance_loss_clip": 0.9832496, + "balance_loss_mlp": 1.01804876, + "epoch": 0.47676306138399566, + "flos": 24136495908480.0, + "grad_norm": 2.0016708390722555, + "language_loss": 0.84851277, + "learning_rate": 2.247080132143769e-06, + "loss": 0.87018573, + "num_input_tokens_seen": 85268525, + "step": 3965, + "time_per_iteration": 2.7402334213256836 + }, + { + "auxiliary_loss_clip": 0.01109526, + "auxiliary_loss_mlp": 0.0102828, + "balance_loss_clip": 0.93575549, + "balance_loss_mlp": 1.01992965, + "epoch": 0.47688330427463477, + "flos": 12604322995200.0, + "grad_norm": 2.3653078793598925, + "language_loss": 0.69365239, + "learning_rate": 2.246307109021121e-06, + "loss": 0.71503049, + "num_input_tokens_seen": 85285930, + "step": 3966, + "time_per_iteration": 2.6919755935668945 + }, + { + "auxiliary_loss_clip": 0.01133536, + "auxiliary_loss_mlp": 0.01026824, + "balance_loss_clip": 0.97772509, + "balance_loss_mlp": 1.01907587, + "epoch": 0.4770035471652739, + "flos": 21390585828480.0, + "grad_norm": 1.7028972495278103, + "language_loss": 0.8258298, + "learning_rate": 2.2455340485338817e-06, + "loss": 0.84743345, + "num_input_tokens_seen": 85303565, + "step": 3967, + "time_per_iteration": 3.6213948726654053 + }, + { + "auxiliary_loss_clip": 0.01161241, + "auxiliary_loss_mlp": 0.0102479, + "balance_loss_clip": 1.01834393, + "balance_loss_mlp": 1.01708269, + "epoch": 0.47712379005591293, + "flos": 25156251025920.0, + "grad_norm": 2.778540107584892, + "language_loss": 0.6751442, + "learning_rate": 2.244760950799322e-06, + "loss": 0.69700444, + "num_input_tokens_seen": 85321835, + "step": 3968, + "time_per_iteration": 2.693790912628174 + }, + { + "auxiliary_loss_clip": 0.01100475, + "auxiliary_loss_mlp": 0.01024985, + "balance_loss_clip": 0.94062877, + "balance_loss_mlp": 1.01725411, + "epoch": 0.47724403294655204, + "flos": 22054323294720.0, + "grad_norm": 1.7757147140577796, + "language_loss": 0.72614336, + "learning_rate": 2.2439878159347203e-06, + "loss": 0.74739802, + "num_input_tokens_seen": 85341260, + "step": 3969, + "time_per_iteration": 2.699737548828125 + }, + { + "auxiliary_loss_clip": 0.01091825, + "auxiliary_loss_mlp": 0.00999494, + "balance_loss_clip": 1.03199887, + "balance_loss_mlp": 0.99823016, + "epoch": 0.4773642758371911, + "flos": 70229387658240.0, + "grad_norm": 0.7249084012226626, + "language_loss": 0.55240166, + "learning_rate": 2.2432146440573616e-06, + "loss": 0.57331485, + "num_input_tokens_seen": 85407220, + "step": 3970, + "time_per_iteration": 4.247226715087891 + }, + { + "auxiliary_loss_clip": 0.011344, + "auxiliary_loss_mlp": 0.01023308, + "balance_loss_clip": 0.97833711, + "balance_loss_mlp": 1.0161078, + "epoch": 0.4774845187278302, + "flos": 23548602009600.0, + "grad_norm": 3.044952622944068, + "language_loss": 0.66321307, + "learning_rate": 2.242441435284534e-06, + "loss": 0.68479013, + "num_input_tokens_seen": 85426095, + "step": 3971, + "time_per_iteration": 2.72454833984375 + }, + { + "auxiliary_loss_clip": 0.01161865, + "auxiliary_loss_mlp": 0.01026347, + "balance_loss_clip": 1.02037621, + "balance_loss_mlp": 1.01868773, + "epoch": 0.4776047616184693, + "flos": 23075371301760.0, + "grad_norm": 5.08027980487961, + "language_loss": 0.85341644, + "learning_rate": 2.2416681897335337e-06, + "loss": 0.87529862, + "num_input_tokens_seen": 85444245, + "step": 3972, + "time_per_iteration": 2.63702654838562 + }, + { + "auxiliary_loss_clip": 0.01087038, + "auxiliary_loss_mlp": 0.0102253, + "balance_loss_clip": 0.90207028, + "balance_loss_mlp": 1.01446509, + "epoch": 0.4777250045091084, + "flos": 31898119374720.0, + "grad_norm": 2.7340881298263917, + "language_loss": 0.66799122, + "learning_rate": 2.240894907521661e-06, + "loss": 0.68908691, + "num_input_tokens_seen": 85463325, + "step": 3973, + "time_per_iteration": 2.831592559814453 + }, + { + "auxiliary_loss_clip": 0.01137261, + "auxiliary_loss_mlp": 0.01024658, + "balance_loss_clip": 0.97927868, + "balance_loss_mlp": 1.01685572, + "epoch": 0.4778452473997475, + "flos": 24278163148800.0, + "grad_norm": 1.728990971736566, + "language_loss": 0.63753295, + "learning_rate": 2.240121588766223e-06, + "loss": 0.65915215, + "num_input_tokens_seen": 85483375, + "step": 3974, + "time_per_iteration": 3.5925447940826416 + }, + { + "auxiliary_loss_clip": 0.01130635, + "auxiliary_loss_mlp": 0.01025161, + "balance_loss_clip": 0.9789927, + "balance_loss_mlp": 1.01764822, + "epoch": 0.4779654902903866, + "flos": 31575031516800.0, + "grad_norm": 3.3081145497811186, + "language_loss": 0.71517038, + "learning_rate": 2.239348233584531e-06, + "loss": 0.73672831, + "num_input_tokens_seen": 85504230, + "step": 3975, + "time_per_iteration": 2.8416748046875 + }, + { + "auxiliary_loss_clip": 0.01162788, + "auxiliary_loss_mlp": 0.01027873, + "balance_loss_clip": 1.01949406, + "balance_loss_mlp": 1.02014184, + "epoch": 0.47808573318102565, + "flos": 19500428344320.0, + "grad_norm": 1.9625802410763187, + "language_loss": 0.80801249, + "learning_rate": 2.2385748420939013e-06, + "loss": 0.8299191, + "num_input_tokens_seen": 85523425, + "step": 3976, + "time_per_iteration": 2.647688388824463 + }, + { + "auxiliary_loss_clip": 0.0118304, + "auxiliary_loss_mlp": 0.01032187, + "balance_loss_clip": 1.06050551, + "balance_loss_mlp": 1.02406931, + "epoch": 0.47820597607166476, + "flos": 22601135013120.0, + "grad_norm": 2.7371378127705808, + "language_loss": 0.72664267, + "learning_rate": 2.2378014144116583e-06, + "loss": 0.74879497, + "num_input_tokens_seen": 85542235, + "step": 3977, + "time_per_iteration": 2.6005263328552246 + }, + { + "auxiliary_loss_clip": 0.01186702, + "auxiliary_loss_mlp": 0.01030114, + "balance_loss_clip": 1.05804074, + "balance_loss_mlp": 1.0221864, + "epoch": 0.4783262189623039, + "flos": 23003011353600.0, + "grad_norm": 2.011049962120243, + "language_loss": 0.80021775, + "learning_rate": 2.23702795065513e-06, + "loss": 0.82238591, + "num_input_tokens_seen": 85561815, + "step": 3978, + "time_per_iteration": 2.705852746963501 + }, + { + "auxiliary_loss_clip": 0.01069474, + "auxiliary_loss_mlp": 0.01002285, + "balance_loss_clip": 0.99066496, + "balance_loss_mlp": 1.0009501, + "epoch": 0.47844646185294293, + "flos": 49772801226240.0, + "grad_norm": 0.9890361333517881, + "language_loss": 0.67534882, + "learning_rate": 2.2362544509416493e-06, + "loss": 0.69606644, + "num_input_tokens_seen": 85613930, + "step": 3979, + "time_per_iteration": 3.078338146209717 + }, + { + "auxiliary_loss_clip": 0.01130496, + "auxiliary_loss_mlp": 0.0102303, + "balance_loss_clip": 0.97769964, + "balance_loss_mlp": 1.01503062, + "epoch": 0.47856670474358204, + "flos": 20229558520320.0, + "grad_norm": 3.253171665696434, + "language_loss": 0.83183986, + "learning_rate": 2.2354809153885572e-06, + "loss": 0.85337508, + "num_input_tokens_seen": 85631000, + "step": 3980, + "time_per_iteration": 2.6752734184265137 + }, + { + "auxiliary_loss_clip": 0.01158242, + "auxiliary_loss_mlp": 0.01028417, + "balance_loss_clip": 1.01699662, + "balance_loss_mlp": 1.02003622, + "epoch": 0.47868694763422115, + "flos": 20990936131200.0, + "grad_norm": 1.9149397527637186, + "language_loss": 0.83377904, + "learning_rate": 2.234707344113197e-06, + "loss": 0.85564566, + "num_input_tokens_seen": 85649095, + "step": 3981, + "time_per_iteration": 2.5750277042388916 + }, + { + "auxiliary_loss_clip": 0.01181596, + "auxiliary_loss_mlp": 0.01029099, + "balance_loss_clip": 1.05667591, + "balance_loss_mlp": 1.02196455, + "epoch": 0.4788071905248602, + "flos": 19026551191680.0, + "grad_norm": 2.044901206080723, + "language_loss": 0.77277529, + "learning_rate": 2.233933737232919e-06, + "loss": 0.79488218, + "num_input_tokens_seen": 85666875, + "step": 3982, + "time_per_iteration": 2.660083532333374 + }, + { + "auxiliary_loss_clip": 0.01084637, + "auxiliary_loss_mlp": 0.00509897, + "balance_loss_clip": 0.90362459, + "balance_loss_mlp": 0.0, + "epoch": 0.4789274334154993, + "flos": 23002221254400.0, + "grad_norm": 1.9154371637954044, + "language_loss": 0.78488171, + "learning_rate": 2.2331600948650793e-06, + "loss": 0.80082703, + "num_input_tokens_seen": 85687020, + "step": 3983, + "time_per_iteration": 2.749619722366333 + }, + { + "auxiliary_loss_clip": 0.01104531, + "auxiliary_loss_mlp": 0.00510747, + "balance_loss_clip": 0.94324422, + "balance_loss_mlp": 0.0, + "epoch": 0.4790476763061384, + "flos": 23075586783360.0, + "grad_norm": 1.5324021919954736, + "language_loss": 0.80000043, + "learning_rate": 2.2323864171270386e-06, + "loss": 0.81615317, + "num_input_tokens_seen": 85708290, + "step": 3984, + "time_per_iteration": 2.731647491455078 + }, + { + "auxiliary_loss_clip": 0.01114846, + "auxiliary_loss_mlp": 0.01036313, + "balance_loss_clip": 0.9397155, + "balance_loss_mlp": 1.02806377, + "epoch": 0.4791679191967775, + "flos": 21179288073600.0, + "grad_norm": 1.9737828428757067, + "language_loss": 0.72877026, + "learning_rate": 2.231612704136164e-06, + "loss": 0.75028181, + "num_input_tokens_seen": 85728660, + "step": 3985, + "time_per_iteration": 2.7298409938812256 + }, + { + "auxiliary_loss_clip": 0.01155234, + "auxiliary_loss_mlp": 0.01025516, + "balance_loss_clip": 1.01676548, + "balance_loss_mlp": 1.01733804, + "epoch": 0.4792881620874166, + "flos": 22301495758080.0, + "grad_norm": 2.205711289753928, + "language_loss": 0.74825394, + "learning_rate": 2.2308389560098253e-06, + "loss": 0.77006137, + "num_input_tokens_seen": 85745035, + "step": 3986, + "time_per_iteration": 2.651979684829712 + }, + { + "auxiliary_loss_clip": 0.01118512, + "auxiliary_loss_mlp": 0.01030744, + "balance_loss_clip": 0.9446224, + "balance_loss_mlp": 1.02199376, + "epoch": 0.47940840497805565, + "flos": 17420877423360.0, + "grad_norm": 2.163431363734808, + "language_loss": 0.7693283, + "learning_rate": 2.2300651728654008e-06, + "loss": 0.79082084, + "num_input_tokens_seen": 85760295, + "step": 3987, + "time_per_iteration": 2.6542539596557617 + }, + { + "auxiliary_loss_clip": 0.01061487, + "auxiliary_loss_mlp": 0.00501272, + "balance_loss_clip": 0.9888804, + "balance_loss_mlp": 0.0, + "epoch": 0.47952864786869476, + "flos": 65358175708800.0, + "grad_norm": 0.7267762469476058, + "language_loss": 0.60205257, + "learning_rate": 2.229291354820272e-06, + "loss": 0.61768019, + "num_input_tokens_seen": 85821305, + "step": 3988, + "time_per_iteration": 3.241661787033081 + }, + { + "auxiliary_loss_clip": 0.01160328, + "auxiliary_loss_mlp": 0.01032114, + "balance_loss_clip": 1.01777351, + "balance_loss_mlp": 1.02357912, + "epoch": 0.47964889075933387, + "flos": 16799802336000.0, + "grad_norm": 2.1369986742168825, + "language_loss": 0.75664175, + "learning_rate": 2.228517501991828e-06, + "loss": 0.77856612, + "num_input_tokens_seen": 85840105, + "step": 3989, + "time_per_iteration": 2.640517473220825 + }, + { + "auxiliary_loss_clip": 0.01042626, + "auxiliary_loss_mlp": 0.01022125, + "balance_loss_clip": 0.95201015, + "balance_loss_mlp": 1.02075386, + "epoch": 0.4797691336499729, + "flos": 70079244808320.0, + "grad_norm": 0.8173160410249543, + "language_loss": 0.61079705, + "learning_rate": 2.22774361449746e-06, + "loss": 0.63144451, + "num_input_tokens_seen": 85896585, + "step": 3990, + "time_per_iteration": 3.2917652130126953 + }, + { + "auxiliary_loss_clip": 0.01070044, + "auxiliary_loss_mlp": 0.01028745, + "balance_loss_clip": 0.86850637, + "balance_loss_mlp": 1.02034116, + "epoch": 0.47988937654061203, + "flos": 18953329317120.0, + "grad_norm": 2.8112807705198417, + "language_loss": 0.70563549, + "learning_rate": 2.2269696924545668e-06, + "loss": 0.72662342, + "num_input_tokens_seen": 85914415, + "step": 3991, + "time_per_iteration": 2.759918451309204 + }, + { + "auxiliary_loss_clip": 0.01113462, + "auxiliary_loss_mlp": 0.01031636, + "balance_loss_clip": 0.94411385, + "balance_loss_mlp": 1.02416754, + "epoch": 0.48000961943125114, + "flos": 14461981649280.0, + "grad_norm": 2.316163740901138, + "language_loss": 0.77893221, + "learning_rate": 2.2261957359805523e-06, + "loss": 0.80038321, + "num_input_tokens_seen": 85931650, + "step": 3992, + "time_per_iteration": 2.694147825241089 + }, + { + "auxiliary_loss_clip": 0.0118446, + "auxiliary_loss_mlp": 0.01026648, + "balance_loss_clip": 1.05792439, + "balance_loss_mlp": 1.01928067, + "epoch": 0.4801298623218902, + "flos": 27051149105280.0, + "grad_norm": 1.7926156237623283, + "language_loss": 0.73971105, + "learning_rate": 2.225421745192823e-06, + "loss": 0.7618221, + "num_input_tokens_seen": 85951805, + "step": 3993, + "time_per_iteration": 3.642909288406372 + }, + { + "auxiliary_loss_clip": 0.01157911, + "auxiliary_loss_mlp": 0.01032101, + "balance_loss_clip": 1.01729727, + "balance_loss_mlp": 1.02338099, + "epoch": 0.4802501052125293, + "flos": 26355236031360.0, + "grad_norm": 2.246217573617373, + "language_loss": 0.78108019, + "learning_rate": 2.2246477202087955e-06, + "loss": 0.8029803, + "num_input_tokens_seen": 85972485, + "step": 3994, + "time_per_iteration": 3.568723201751709 + }, + { + "auxiliary_loss_clip": 0.01137183, + "auxiliary_loss_mlp": 0.01030995, + "balance_loss_clip": 0.97831762, + "balance_loss_mlp": 1.02301395, + "epoch": 0.4803703481031684, + "flos": 20993916960000.0, + "grad_norm": 2.3260280317444084, + "language_loss": 0.82993615, + "learning_rate": 2.223873661145887e-06, + "loss": 0.85161793, + "num_input_tokens_seen": 85992540, + "step": 3995, + "time_per_iteration": 3.5838422775268555 + }, + { + "auxiliary_loss_clip": 0.01136408, + "auxiliary_loss_mlp": 0.00510182, + "balance_loss_clip": 0.98430288, + "balance_loss_mlp": 0.0, + "epoch": 0.4804905909938075, + "flos": 20703722981760.0, + "grad_norm": 12.773843416348399, + "language_loss": 0.70996857, + "learning_rate": 2.2230995681215226e-06, + "loss": 0.72643447, + "num_input_tokens_seen": 86012065, + "step": 3996, + "time_per_iteration": 2.702681541442871 + }, + { + "auxiliary_loss_clip": 0.01109476, + "auxiliary_loss_mlp": 0.01027886, + "balance_loss_clip": 0.93966401, + "balance_loss_mlp": 1.02032781, + "epoch": 0.4806108338844466, + "flos": 16654831044480.0, + "grad_norm": 1.7742004805419618, + "language_loss": 0.78037858, + "learning_rate": 2.2223254412531305e-06, + "loss": 0.80175221, + "num_input_tokens_seen": 86029435, + "step": 3997, + "time_per_iteration": 2.694399118423462 + }, + { + "auxiliary_loss_clip": 0.01122128, + "auxiliary_loss_mlp": 0.01028854, + "balance_loss_clip": 0.9744696, + "balance_loss_mlp": 1.02112317, + "epoch": 0.4807310767750857, + "flos": 20011329440640.0, + "grad_norm": 1.9731911171931382, + "language_loss": 0.82223266, + "learning_rate": 2.221551280658146e-06, + "loss": 0.84374249, + "num_input_tokens_seen": 86048495, + "step": 3998, + "time_per_iteration": 2.6617252826690674 + }, + { + "auxiliary_loss_clip": 0.01086101, + "auxiliary_loss_mlp": 0.01027864, + "balance_loss_clip": 0.9030019, + "balance_loss_mlp": 1.01979947, + "epoch": 0.48085131966572475, + "flos": 23185257984000.0, + "grad_norm": 1.705227092268778, + "language_loss": 0.74465513, + "learning_rate": 2.2207770864540085e-06, + "loss": 0.76579481, + "num_input_tokens_seen": 86067470, + "step": 3999, + "time_per_iteration": 3.6597931385040283 + }, + { + "auxiliary_loss_clip": 0.01131605, + "auxiliary_loss_mlp": 0.01026289, + "balance_loss_clip": 0.97918808, + "balance_loss_mlp": 1.01884437, + "epoch": 0.48097156255636386, + "flos": 20558643949440.0, + "grad_norm": 1.970271562416494, + "language_loss": 0.72738814, + "learning_rate": 2.220002858758162e-06, + "loss": 0.74896705, + "num_input_tokens_seen": 86085460, + "step": 4000, + "time_per_iteration": 2.658038377761841 + }, + { + "auxiliary_loss_clip": 0.01067303, + "auxiliary_loss_mlp": 0.01000007, + "balance_loss_clip": 0.98934299, + "balance_loss_mlp": 0.99869615, + "epoch": 0.481091805447003, + "flos": 70511608817280.0, + "grad_norm": 0.8766535816655875, + "language_loss": 0.60813582, + "learning_rate": 2.2192285976880573e-06, + "loss": 0.62880898, + "num_input_tokens_seen": 86149715, + "step": 4001, + "time_per_iteration": 3.230165958404541 + }, + { + "auxiliary_loss_clip": 0.01113587, + "auxiliary_loss_mlp": 0.00509236, + "balance_loss_clip": 0.9395287, + "balance_loss_mlp": 0.0, + "epoch": 0.48121204833764203, + "flos": 36428214839040.0, + "grad_norm": 1.5530002447181357, + "language_loss": 0.80637562, + "learning_rate": 2.2184543033611485e-06, + "loss": 0.82260382, + "num_input_tokens_seen": 86170795, + "step": 4002, + "time_per_iteration": 2.84306263923645 + }, + { + "auxiliary_loss_clip": 0.01163842, + "auxiliary_loss_mlp": 0.01026652, + "balance_loss_clip": 1.01776147, + "balance_loss_mlp": 1.01856422, + "epoch": 0.48133229122828114, + "flos": 27490264871040.0, + "grad_norm": 2.764594693654514, + "language_loss": 0.81823874, + "learning_rate": 2.2176799758948957e-06, + "loss": 0.84014368, + "num_input_tokens_seen": 86190955, + "step": 4003, + "time_per_iteration": 2.7058675289154053 + }, + { + "auxiliary_loss_clip": 0.01132141, + "auxiliary_loss_mlp": 0.01032904, + "balance_loss_clip": 0.9798553, + "balance_loss_mlp": 1.02529287, + "epoch": 0.4814525341189202, + "flos": 43072802179200.0, + "grad_norm": 2.732786514258052, + "language_loss": 0.73430252, + "learning_rate": 2.2169056154067635e-06, + "loss": 0.75595295, + "num_input_tokens_seen": 86214875, + "step": 4004, + "time_per_iteration": 2.89656400680542 + }, + { + "auxiliary_loss_clip": 0.01158928, + "auxiliary_loss_mlp": 0.00510158, + "balance_loss_clip": 1.01807213, + "balance_loss_mlp": 0.0, + "epoch": 0.4815727770095593, + "flos": 24236901400320.0, + "grad_norm": 2.076254231213092, + "language_loss": 0.82328868, + "learning_rate": 2.216131222014222e-06, + "loss": 0.83997953, + "num_input_tokens_seen": 86232950, + "step": 4005, + "time_per_iteration": 2.6654319763183594 + }, + { + "auxiliary_loss_clip": 0.01102889, + "auxiliary_loss_mlp": 0.01025389, + "balance_loss_clip": 0.93644798, + "balance_loss_mlp": 1.01728261, + "epoch": 0.4816930199001984, + "flos": 18113630100480.0, + "grad_norm": 2.182894418003976, + "language_loss": 0.80844778, + "learning_rate": 2.2153567958347455e-06, + "loss": 0.82973057, + "num_input_tokens_seen": 86249160, + "step": 4006, + "time_per_iteration": 2.8148155212402344 + }, + { + "auxiliary_loss_clip": 0.01135408, + "auxiliary_loss_mlp": 0.01031136, + "balance_loss_clip": 0.98072851, + "balance_loss_mlp": 1.02230263, + "epoch": 0.48181326279083747, + "flos": 17274720983040.0, + "grad_norm": 1.828420251030258, + "language_loss": 0.79882801, + "learning_rate": 2.214582336985815e-06, + "loss": 0.82049346, + "num_input_tokens_seen": 86267060, + "step": 4007, + "time_per_iteration": 2.6221139430999756 + }, + { + "auxiliary_loss_clip": 0.01130334, + "auxiliary_loss_mlp": 0.01025184, + "balance_loss_clip": 0.97977471, + "balance_loss_mlp": 1.0171138, + "epoch": 0.4819335056814766, + "flos": 14903252231040.0, + "grad_norm": 2.1442231661474627, + "language_loss": 0.66374528, + "learning_rate": 2.2138078455849142e-06, + "loss": 0.68530041, + "num_input_tokens_seen": 86285055, + "step": 4008, + "time_per_iteration": 2.6395368576049805 + }, + { + "auxiliary_loss_clip": 0.01166301, + "auxiliary_loss_mlp": 0.0103077, + "balance_loss_clip": 1.01966953, + "balance_loss_mlp": 1.02315819, + "epoch": 0.4820537485721157, + "flos": 19244888012160.0, + "grad_norm": 2.181323979392781, + "language_loss": 0.7864964, + "learning_rate": 2.2130333217495334e-06, + "loss": 0.80846715, + "num_input_tokens_seen": 86304225, + "step": 4009, + "time_per_iteration": 2.601727247238159 + }, + { + "auxiliary_loss_clip": 0.01134671, + "auxiliary_loss_mlp": 0.01022648, + "balance_loss_clip": 0.98101068, + "balance_loss_mlp": 1.01478004, + "epoch": 0.48217399146275475, + "flos": 16033791870720.0, + "grad_norm": 2.906243482667321, + "language_loss": 0.67713732, + "learning_rate": 2.2122587655971665e-06, + "loss": 0.6987105, + "num_input_tokens_seen": 86319170, + "step": 4010, + "time_per_iteration": 2.6239469051361084 + }, + { + "auxiliary_loss_clip": 0.01136286, + "auxiliary_loss_mlp": 0.01034697, + "balance_loss_clip": 0.97809017, + "balance_loss_mlp": 1.02699053, + "epoch": 0.48229423435339386, + "flos": 24134197438080.0, + "grad_norm": 1.8202601335919673, + "language_loss": 0.63994652, + "learning_rate": 2.211484177245314e-06, + "loss": 0.66165638, + "num_input_tokens_seen": 86338760, + "step": 4011, + "time_per_iteration": 2.674016237258911 + }, + { + "auxiliary_loss_clip": 0.01185291, + "auxiliary_loss_mlp": 0.01027462, + "balance_loss_clip": 1.05734468, + "balance_loss_mlp": 1.01948106, + "epoch": 0.48241447724403297, + "flos": 23805435231360.0, + "grad_norm": 1.9426010373935405, + "language_loss": 0.721843, + "learning_rate": 2.21070955681148e-06, + "loss": 0.74397057, + "num_input_tokens_seen": 86357865, + "step": 4012, + "time_per_iteration": 2.6126608848571777 + }, + { + "auxiliary_loss_clip": 0.01100439, + "auxiliary_loss_mlp": 0.01028127, + "balance_loss_clip": 0.93654317, + "balance_loss_mlp": 1.02022982, + "epoch": 0.482534720134672, + "flos": 23110312256640.0, + "grad_norm": 1.6924282354488023, + "language_loss": 0.77966464, + "learning_rate": 2.209934904413174e-06, + "loss": 0.80095023, + "num_input_tokens_seen": 86379470, + "step": 4013, + "time_per_iteration": 2.754570722579956 + }, + { + "auxiliary_loss_clip": 0.01065369, + "auxiliary_loss_mlp": 0.01031134, + "balance_loss_clip": 0.86043143, + "balance_loss_mlp": 1.02309322, + "epoch": 0.48265496302531113, + "flos": 20923819568640.0, + "grad_norm": 2.5925667838920377, + "language_loss": 0.72033554, + "learning_rate": 2.2091602201679095e-06, + "loss": 0.74130058, + "num_input_tokens_seen": 86399080, + "step": 4014, + "time_per_iteration": 2.8509671688079834 + }, + { + "auxiliary_loss_clip": 0.0111642, + "auxiliary_loss_mlp": 0.01031831, + "balance_loss_clip": 0.94253087, + "balance_loss_mlp": 1.02452326, + "epoch": 0.48277520591595025, + "flos": 15231152511360.0, + "grad_norm": 2.5213376312920524, + "language_loss": 0.8340165, + "learning_rate": 2.208385504193206e-06, + "loss": 0.85549897, + "num_input_tokens_seen": 86416580, + "step": 4015, + "time_per_iteration": 2.6725964546203613 + }, + { + "auxiliary_loss_clip": 0.01181232, + "auxiliary_loss_mlp": 0.01026401, + "balance_loss_clip": 1.05430984, + "balance_loss_mlp": 1.01858675, + "epoch": 0.4828954488065893, + "flos": 17858664385920.0, + "grad_norm": 2.193308918951091, + "language_loss": 0.80774736, + "learning_rate": 2.2076107566065873e-06, + "loss": 0.82982361, + "num_input_tokens_seen": 86434365, + "step": 4016, + "time_per_iteration": 2.591461658477783 + }, + { + "auxiliary_loss_clip": 0.01164166, + "auxiliary_loss_mlp": 0.01028045, + "balance_loss_clip": 1.01948082, + "balance_loss_mlp": 1.02035618, + "epoch": 0.4830156916972284, + "flos": 32087405070720.0, + "grad_norm": 2.1224644541631634, + "language_loss": 0.75153673, + "learning_rate": 2.2068359775255816e-06, + "loss": 0.77345884, + "num_input_tokens_seen": 86452675, + "step": 4017, + "time_per_iteration": 2.8017914295196533 + }, + { + "auxiliary_loss_clip": 0.01087681, + "auxiliary_loss_mlp": 0.01021247, + "balance_loss_clip": 0.90383661, + "balance_loss_mlp": 1.01346827, + "epoch": 0.48313593458786747, + "flos": 21871717528320.0, + "grad_norm": 2.3827267195950737, + "language_loss": 0.78300786, + "learning_rate": 2.206061167067723e-06, + "loss": 0.80409718, + "num_input_tokens_seen": 86470785, + "step": 4018, + "time_per_iteration": 2.712045907974243 + }, + { + "auxiliary_loss_clip": 0.01107517, + "auxiliary_loss_mlp": 0.01031386, + "balance_loss_clip": 0.93636727, + "balance_loss_mlp": 1.02317214, + "epoch": 0.4832561774785066, + "flos": 22601206840320.0, + "grad_norm": 2.2756767282907466, + "language_loss": 0.80103797, + "learning_rate": 2.205286325350549e-06, + "loss": 0.82242697, + "num_input_tokens_seen": 86489850, + "step": 4019, + "time_per_iteration": 4.595645427703857 + }, + { + "auxiliary_loss_clip": 0.01087556, + "auxiliary_loss_mlp": 0.01023334, + "balance_loss_clip": 0.9007529, + "balance_loss_mlp": 1.01538873, + "epoch": 0.4833764203691457, + "flos": 13437342282240.0, + "grad_norm": 2.1271959498536632, + "language_loss": 0.72533226, + "learning_rate": 2.204511452491603e-06, + "loss": 0.74644119, + "num_input_tokens_seen": 86506475, + "step": 4020, + "time_per_iteration": 2.7296791076660156 + }, + { + "auxiliary_loss_clip": 0.01181088, + "auxiliary_loss_mlp": 0.01027421, + "balance_loss_clip": 1.05861366, + "balance_loss_mlp": 1.02009511, + "epoch": 0.48349666325978474, + "flos": 44128036955520.0, + "grad_norm": 1.5601260265918104, + "language_loss": 0.74679917, + "learning_rate": 2.2037365486084316e-06, + "loss": 0.7688843, + "num_input_tokens_seen": 86529715, + "step": 4021, + "time_per_iteration": 3.7440662384033203 + }, + { + "auxiliary_loss_clip": 0.01121599, + "auxiliary_loss_mlp": 0.0103243, + "balance_loss_clip": 0.93898535, + "balance_loss_mlp": 1.02382374, + "epoch": 0.48361690615042385, + "flos": 26028377245440.0, + "grad_norm": 1.9199352336190687, + "language_loss": 0.78138971, + "learning_rate": 2.2029616138185886e-06, + "loss": 0.80293, + "num_input_tokens_seen": 86548715, + "step": 4022, + "time_per_iteration": 2.7551589012145996 + }, + { + "auxiliary_loss_clip": 0.01105727, + "auxiliary_loss_mlp": 0.01032077, + "balance_loss_clip": 0.94103307, + "balance_loss_mlp": 1.02414918, + "epoch": 0.48373714904106296, + "flos": 22273306560000.0, + "grad_norm": 1.6776145484750211, + "language_loss": 0.82634747, + "learning_rate": 2.202186648239629e-06, + "loss": 0.84772551, + "num_input_tokens_seen": 86568650, + "step": 4023, + "time_per_iteration": 2.747019052505493 + }, + { + "auxiliary_loss_clip": 0.01156415, + "auxiliary_loss_mlp": 0.01025628, + "balance_loss_clip": 1.0168879, + "balance_loss_mlp": 1.01820183, + "epoch": 0.483857391931702, + "flos": 28292293699200.0, + "grad_norm": 2.1239625796699344, + "language_loss": 0.71546447, + "learning_rate": 2.201411651989117e-06, + "loss": 0.7372849, + "num_input_tokens_seen": 86590630, + "step": 4024, + "time_per_iteration": 2.6764214038848877 + }, + { + "auxiliary_loss_clip": 0.01132559, + "auxiliary_loss_mlp": 0.00509923, + "balance_loss_clip": 0.97997451, + "balance_loss_mlp": 0.0, + "epoch": 0.48397763482234113, + "flos": 27418048577280.0, + "grad_norm": 1.9047340156103996, + "language_loss": 0.78406018, + "learning_rate": 2.2006366251846167e-06, + "loss": 0.80048501, + "num_input_tokens_seen": 86611270, + "step": 4025, + "time_per_iteration": 2.7219626903533936 + }, + { + "auxiliary_loss_clip": 0.01130837, + "auxiliary_loss_mlp": 0.01021795, + "balance_loss_clip": 0.97698247, + "balance_loss_mlp": 1.01485729, + "epoch": 0.48409787771298024, + "flos": 16797252470400.0, + "grad_norm": 1.944680948644828, + "language_loss": 0.75358212, + "learning_rate": 2.1998615679436997e-06, + "loss": 0.7751084, + "num_input_tokens_seen": 86628810, + "step": 4026, + "time_per_iteration": 3.515737771987915 + }, + { + "auxiliary_loss_clip": 0.01139833, + "auxiliary_loss_mlp": 0.01029877, + "balance_loss_clip": 0.97804618, + "balance_loss_mlp": 1.02191937, + "epoch": 0.4842181206036193, + "flos": 25083496028160.0, + "grad_norm": 2.262625894235673, + "language_loss": 0.77469409, + "learning_rate": 2.199086480383942e-06, + "loss": 0.79639125, + "num_input_tokens_seen": 86648185, + "step": 4027, + "time_per_iteration": 2.7113983631134033 + }, + { + "auxiliary_loss_clip": 0.01147293, + "auxiliary_loss_mlp": 0.01031194, + "balance_loss_clip": 0.98119199, + "balance_loss_mlp": 1.02278399, + "epoch": 0.4843383634942584, + "flos": 30372311496960.0, + "grad_norm": 2.7463565579928306, + "language_loss": 0.67810488, + "learning_rate": 2.1983113626229234e-06, + "loss": 0.69988978, + "num_input_tokens_seen": 86667435, + "step": 4028, + "time_per_iteration": 2.709639310836792 + }, + { + "auxiliary_loss_clip": 0.0110438, + "auxiliary_loss_mlp": 0.00509755, + "balance_loss_clip": 0.9366926, + "balance_loss_mlp": 0.0, + "epoch": 0.4844586063848975, + "flos": 20413564917120.0, + "grad_norm": 1.6865611816751105, + "language_loss": 0.78801578, + "learning_rate": 2.1975362147782293e-06, + "loss": 0.80415714, + "num_input_tokens_seen": 86686630, + "step": 4029, + "time_per_iteration": 2.8006749153137207 + }, + { + "auxiliary_loss_clip": 0.01055167, + "auxiliary_loss_mlp": 0.01010765, + "balance_loss_clip": 0.96570873, + "balance_loss_mlp": 1.00917995, + "epoch": 0.48457884927553657, + "flos": 70303722854400.0, + "grad_norm": 0.6890437639661713, + "language_loss": 0.540959, + "learning_rate": 2.196761036967448e-06, + "loss": 0.56161833, + "num_input_tokens_seen": 86754595, + "step": 4030, + "time_per_iteration": 3.383754014968872 + }, + { + "auxiliary_loss_clip": 0.01154652, + "auxiliary_loss_mlp": 0.01027723, + "balance_loss_clip": 1.01735461, + "balance_loss_mlp": 1.02052617, + "epoch": 0.4846990921661757, + "flos": 19934516206080.0, + "grad_norm": 1.6847383409582803, + "language_loss": 0.77693379, + "learning_rate": 2.1959858293081743e-06, + "loss": 0.79875755, + "num_input_tokens_seen": 86773730, + "step": 4031, + "time_per_iteration": 2.6026406288146973 + }, + { + "auxiliary_loss_clip": 0.0110583, + "auxiliary_loss_mlp": 0.01028565, + "balance_loss_clip": 0.93871117, + "balance_loss_mlp": 1.02135611, + "epoch": 0.4848193350568148, + "flos": 23075945919360.0, + "grad_norm": 1.846615020881557, + "language_loss": 0.76025891, + "learning_rate": 2.1952105919180056e-06, + "loss": 0.7816028, + "num_input_tokens_seen": 86792985, + "step": 4032, + "time_per_iteration": 2.771711587905884 + }, + { + "auxiliary_loss_clip": 0.0113374, + "auxiliary_loss_mlp": 0.01027317, + "balance_loss_clip": 0.97758561, + "balance_loss_mlp": 1.01964569, + "epoch": 0.48493957794745385, + "flos": 22455481363200.0, + "grad_norm": 2.0526241600150614, + "language_loss": 0.6792171, + "learning_rate": 2.1944353249145456e-06, + "loss": 0.70082766, + "num_input_tokens_seen": 86812095, + "step": 4033, + "time_per_iteration": 2.6500208377838135 + }, + { + "auxiliary_loss_clip": 0.01184875, + "auxiliary_loss_mlp": 0.01025205, + "balance_loss_clip": 1.05904245, + "balance_loss_mlp": 1.01794243, + "epoch": 0.48505982083809296, + "flos": 25046112948480.0, + "grad_norm": 1.5831654313246821, + "language_loss": 0.74457669, + "learning_rate": 2.193660028415401e-06, + "loss": 0.76667756, + "num_input_tokens_seen": 86832875, + "step": 4034, + "time_per_iteration": 2.649397611618042 + }, + { + "auxiliary_loss_clip": 0.01127917, + "auxiliary_loss_mlp": 0.01021918, + "balance_loss_clip": 0.97705311, + "balance_loss_mlp": 1.01421762, + "epoch": 0.485180063728732, + "flos": 26761386090240.0, + "grad_norm": 1.7328713779691396, + "language_loss": 0.8215239, + "learning_rate": 2.1928847025381852e-06, + "loss": 0.84302223, + "num_input_tokens_seen": 86853480, + "step": 4035, + "time_per_iteration": 2.697472095489502 + }, + { + "auxiliary_loss_clip": 0.01153701, + "auxiliary_loss_mlp": 0.01031198, + "balance_loss_clip": 1.01360857, + "balance_loss_mlp": 1.02320492, + "epoch": 0.4853003066193711, + "flos": 24059143969920.0, + "grad_norm": 1.7062421067153641, + "language_loss": 0.83741701, + "learning_rate": 2.192109347400512e-06, + "loss": 0.85926604, + "num_input_tokens_seen": 86873695, + "step": 4036, + "time_per_iteration": 2.6366465091705322 + }, + { + "auxiliary_loss_clip": 0.01135342, + "auxiliary_loss_mlp": 0.01023473, + "balance_loss_clip": 0.97778118, + "balance_loss_mlp": 1.01580763, + "epoch": 0.48542054951001024, + "flos": 23076376882560.0, + "grad_norm": 1.7139203449025262, + "language_loss": 0.78997648, + "learning_rate": 2.191333963120004e-06, + "loss": 0.81156462, + "num_input_tokens_seen": 86892675, + "step": 4037, + "time_per_iteration": 2.6826717853546143 + }, + { + "auxiliary_loss_clip": 0.01133087, + "auxiliary_loss_mlp": 0.01029883, + "balance_loss_clip": 0.97710264, + "balance_loss_mlp": 1.02201521, + "epoch": 0.4855407924006493, + "flos": 25664889565440.0, + "grad_norm": 4.721649117899793, + "language_loss": 0.70700336, + "learning_rate": 2.190558549814286e-06, + "loss": 0.72863305, + "num_input_tokens_seen": 86912835, + "step": 4038, + "time_per_iteration": 2.728398084640503 + }, + { + "auxiliary_loss_clip": 0.01132549, + "auxiliary_loss_mlp": 0.01029423, + "balance_loss_clip": 0.97740412, + "balance_loss_mlp": 1.02136421, + "epoch": 0.4856610352912884, + "flos": 23987933256960.0, + "grad_norm": 2.487128151599106, + "language_loss": 0.79691809, + "learning_rate": 2.1897831076009872e-06, + "loss": 0.81853783, + "num_input_tokens_seen": 86932475, + "step": 4039, + "time_per_iteration": 2.7426817417144775 + }, + { + "auxiliary_loss_clip": 0.01158388, + "auxiliary_loss_mlp": 0.0102443, + "balance_loss_clip": 1.01831532, + "balance_loss_mlp": 1.01741803, + "epoch": 0.4857812781819275, + "flos": 24096814358400.0, + "grad_norm": 2.1709990283241547, + "language_loss": 0.7996136, + "learning_rate": 2.1890076365977426e-06, + "loss": 0.82144177, + "num_input_tokens_seen": 86952300, + "step": 4040, + "time_per_iteration": 2.6847872734069824 + }, + { + "auxiliary_loss_clip": 0.01045752, + "auxiliary_loss_mlp": 0.01000094, + "balance_loss_clip": 0.95397818, + "balance_loss_mlp": 0.99893743, + "epoch": 0.48590152107256657, + "flos": 56266635185280.0, + "grad_norm": 0.8512996643982264, + "language_loss": 0.52786767, + "learning_rate": 2.188232136922189e-06, + "loss": 0.54832613, + "num_input_tokens_seen": 87010420, + "step": 4041, + "time_per_iteration": 3.163475275039673 + }, + { + "auxiliary_loss_clip": 0.01053138, + "auxiliary_loss_mlp": 0.01029685, + "balance_loss_clip": 0.85677469, + "balance_loss_mlp": 1.02176404, + "epoch": 0.4860217639632057, + "flos": 20046988667520.0, + "grad_norm": 1.769714930455426, + "language_loss": 0.76315403, + "learning_rate": 2.187456608691971e-06, + "loss": 0.78398228, + "num_input_tokens_seen": 87029295, + "step": 4042, + "time_per_iteration": 2.763601303100586 + }, + { + "auxiliary_loss_clip": 0.01112826, + "auxiliary_loss_mlp": 0.01027227, + "balance_loss_clip": 0.94036174, + "balance_loss_mlp": 1.01966894, + "epoch": 0.4861420068538448, + "flos": 17822143232640.0, + "grad_norm": 2.015787809613771, + "language_loss": 0.87157619, + "learning_rate": 2.1866810520247334e-06, + "loss": 0.8929767, + "num_input_tokens_seen": 87048165, + "step": 4043, + "time_per_iteration": 2.704893112182617 + }, + { + "auxiliary_loss_clip": 0.01160915, + "auxiliary_loss_mlp": 0.01026486, + "balance_loss_clip": 1.01592088, + "balance_loss_mlp": 1.01857615, + "epoch": 0.48626224974448384, + "flos": 26250125857920.0, + "grad_norm": 1.842286981430077, + "language_loss": 0.65151626, + "learning_rate": 2.185905467038129e-06, + "loss": 0.67339027, + "num_input_tokens_seen": 87067070, + "step": 4044, + "time_per_iteration": 3.589142084121704 + }, + { + "auxiliary_loss_clip": 0.01180978, + "auxiliary_loss_mlp": 0.01021364, + "balance_loss_clip": 1.05866528, + "balance_loss_mlp": 1.01424086, + "epoch": 0.48638249263512295, + "flos": 22054502862720.0, + "grad_norm": 1.785599967586238, + "language_loss": 0.77772814, + "learning_rate": 2.1851298538498127e-06, + "loss": 0.79975152, + "num_input_tokens_seen": 87086785, + "step": 4045, + "time_per_iteration": 3.557260751724243 + }, + { + "auxiliary_loss_clip": 0.01164239, + "auxiliary_loss_mlp": 0.00510093, + "balance_loss_clip": 1.01944852, + "balance_loss_mlp": 0.0, + "epoch": 0.48650273552576206, + "flos": 25119945354240.0, + "grad_norm": 8.70736984946016, + "language_loss": 0.80032206, + "learning_rate": 2.184354212577446e-06, + "loss": 0.81706542, + "num_input_tokens_seen": 87107090, + "step": 4046, + "time_per_iteration": 2.7083303928375244 + }, + { + "auxiliary_loss_clip": 0.01182953, + "auxiliary_loss_mlp": 0.01030369, + "balance_loss_clip": 1.05621982, + "balance_loss_mlp": 1.02217293, + "epoch": 0.4866229784164011, + "flos": 17456931699840.0, + "grad_norm": 2.8164362248806656, + "language_loss": 0.63220417, + "learning_rate": 2.1835785433386907e-06, + "loss": 0.65433741, + "num_input_tokens_seen": 87125905, + "step": 4047, + "time_per_iteration": 3.5037548542022705 + }, + { + "auxiliary_loss_clip": 0.01101227, + "auxiliary_loss_mlp": 0.01025687, + "balance_loss_clip": 0.93929315, + "balance_loss_mlp": 1.01792645, + "epoch": 0.48674322130704023, + "flos": 23331127115520.0, + "grad_norm": 2.0014164973357866, + "language_loss": 0.66042823, + "learning_rate": 2.182802846251216e-06, + "loss": 0.68169737, + "num_input_tokens_seen": 87146175, + "step": 4048, + "time_per_iteration": 2.6997616291046143 + }, + { + "auxiliary_loss_clip": 0.01114526, + "auxiliary_loss_mlp": 0.01030052, + "balance_loss_clip": 0.93883222, + "balance_loss_mlp": 1.02237463, + "epoch": 0.4868634641976793, + "flos": 28804344030720.0, + "grad_norm": 1.7889167469500402, + "language_loss": 0.72149324, + "learning_rate": 2.182027121432696e-06, + "loss": 0.742939, + "num_input_tokens_seen": 87166800, + "step": 4049, + "time_per_iteration": 2.787308931350708 + }, + { + "auxiliary_loss_clip": 0.01186774, + "auxiliary_loss_mlp": 0.01025603, + "balance_loss_clip": 1.05882478, + "balance_loss_mlp": 1.01808059, + "epoch": 0.4869837070883184, + "flos": 19025976574080.0, + "grad_norm": 1.9078536155861712, + "language_loss": 0.82101077, + "learning_rate": 2.1812513690008054e-06, + "loss": 0.84313446, + "num_input_tokens_seen": 87185920, + "step": 4050, + "time_per_iteration": 2.601072311401367 + }, + { + "auxiliary_loss_clip": 0.01166271, + "auxiliary_loss_mlp": 0.01030724, + "balance_loss_clip": 1.02033973, + "balance_loss_mlp": 1.0226301, + "epoch": 0.4871039499789575, + "flos": 15121409483520.0, + "grad_norm": 2.0460246225599024, + "language_loss": 0.79770887, + "learning_rate": 2.180475589073227e-06, + "loss": 0.81967878, + "num_input_tokens_seen": 87203620, + "step": 4051, + "time_per_iteration": 2.6560728549957275 + }, + { + "auxiliary_loss_clip": 0.01150274, + "auxiliary_loss_mlp": 0.01024432, + "balance_loss_clip": 1.01573467, + "balance_loss_mlp": 1.01721382, + "epoch": 0.48722419286959656, + "flos": 26174066808960.0, + "grad_norm": 1.703729485107295, + "language_loss": 0.7331028, + "learning_rate": 2.1796997817676456e-06, + "loss": 0.75484985, + "num_input_tokens_seen": 87224630, + "step": 4052, + "time_per_iteration": 3.511791467666626 + }, + { + "auxiliary_loss_clip": 0.0116089, + "auxiliary_loss_mlp": 0.00508833, + "balance_loss_clip": 1.01896942, + "balance_loss_mlp": 0.0, + "epoch": 0.4873444357602357, + "flos": 24026142349440.0, + "grad_norm": 1.5254384255789608, + "language_loss": 0.67568594, + "learning_rate": 2.1789239472017494e-06, + "loss": 0.69238317, + "num_input_tokens_seen": 87246280, + "step": 4053, + "time_per_iteration": 2.679988145828247 + }, + { + "auxiliary_loss_clip": 0.01106716, + "auxiliary_loss_mlp": 0.01024124, + "balance_loss_clip": 0.93676198, + "balance_loss_mlp": 1.01558256, + "epoch": 0.4874646786508748, + "flos": 22820441500800.0, + "grad_norm": 2.247896313737273, + "language_loss": 0.73217845, + "learning_rate": 2.1781480854932326e-06, + "loss": 0.75348687, + "num_input_tokens_seen": 87266045, + "step": 4054, + "time_per_iteration": 2.672574281692505 + }, + { + "auxiliary_loss_clip": 0.01088846, + "auxiliary_loss_mlp": 0.01025189, + "balance_loss_clip": 0.90351427, + "balance_loss_mlp": 1.01720238, + "epoch": 0.48758492154151384, + "flos": 21287594557440.0, + "grad_norm": 2.1571902106711165, + "language_loss": 0.7931357, + "learning_rate": 2.1773721967597933e-06, + "loss": 0.81427604, + "num_input_tokens_seen": 87284495, + "step": 4055, + "time_per_iteration": 2.7281782627105713 + }, + { + "auxiliary_loss_clip": 0.01033839, + "auxiliary_loss_mlp": 0.01002403, + "balance_loss_clip": 0.9472025, + "balance_loss_mlp": 1.00116277, + "epoch": 0.48770516443215295, + "flos": 62244109180800.0, + "grad_norm": 0.8525954551293425, + "language_loss": 0.57412875, + "learning_rate": 2.1765962811191322e-06, + "loss": 0.59449124, + "num_input_tokens_seen": 87338960, + "step": 4056, + "time_per_iteration": 3.1447112560272217 + }, + { + "auxiliary_loss_clip": 0.00998109, + "auxiliary_loss_mlp": 0.01004474, + "balance_loss_clip": 0.87529433, + "balance_loss_mlp": 1.00310898, + "epoch": 0.48782540732279206, + "flos": 66133451882880.0, + "grad_norm": 0.8341773436277395, + "language_loss": 0.62074471, + "learning_rate": 2.1758203386889566e-06, + "loss": 0.64077055, + "num_input_tokens_seen": 87401730, + "step": 4057, + "time_per_iteration": 3.303899049758911 + }, + { + "auxiliary_loss_clip": 0.01111635, + "auxiliary_loss_mlp": 0.00510282, + "balance_loss_clip": 0.94122684, + "balance_loss_mlp": 0.0, + "epoch": 0.4879456502134311, + "flos": 14607922608000.0, + "grad_norm": 2.402928021716419, + "language_loss": 0.84437716, + "learning_rate": 2.1750443695869746e-06, + "loss": 0.8605963, + "num_input_tokens_seen": 87417300, + "step": 4058, + "time_per_iteration": 2.707212448120117 + }, + { + "auxiliary_loss_clip": 0.01159625, + "auxiliary_loss_mlp": 0.01026353, + "balance_loss_clip": 1.01652086, + "balance_loss_mlp": 1.01878023, + "epoch": 0.4880658931040702, + "flos": 19500464257920.0, + "grad_norm": 2.1680117195211204, + "language_loss": 0.85389566, + "learning_rate": 2.174268373930901e-06, + "loss": 0.87575549, + "num_input_tokens_seen": 87434815, + "step": 4059, + "time_per_iteration": 2.604074239730835 + }, + { + "auxiliary_loss_clip": 0.01100855, + "auxiliary_loss_mlp": 0.00509893, + "balance_loss_clip": 0.93846679, + "balance_loss_mlp": 0.0, + "epoch": 0.48818613599470934, + "flos": 16723060928640.0, + "grad_norm": 3.3631062302014447, + "language_loss": 0.79995024, + "learning_rate": 2.1734923518384537e-06, + "loss": 0.81605768, + "num_input_tokens_seen": 87451420, + "step": 4060, + "time_per_iteration": 2.690776824951172 + }, + { + "auxiliary_loss_clip": 0.01096177, + "auxiliary_loss_mlp": 0.01030402, + "balance_loss_clip": 0.93819749, + "balance_loss_mlp": 1.02270126, + "epoch": 0.4883063788853484, + "flos": 26756932803840.0, + "grad_norm": 1.8142139399964896, + "language_loss": 0.82389641, + "learning_rate": 2.1727163034273547e-06, + "loss": 0.84516221, + "num_input_tokens_seen": 87469585, + "step": 4061, + "time_per_iteration": 2.7145931720733643 + }, + { + "auxiliary_loss_clip": 0.01158529, + "auxiliary_loss_mlp": 0.01026043, + "balance_loss_clip": 1.01550436, + "balance_loss_mlp": 1.01827621, + "epoch": 0.4884266217759875, + "flos": 16763388923520.0, + "grad_norm": 2.4357178776374044, + "language_loss": 0.79329646, + "learning_rate": 2.17194022881533e-06, + "loss": 0.81514215, + "num_input_tokens_seen": 87485675, + "step": 4062, + "time_per_iteration": 2.694794178009033 + }, + { + "auxiliary_loss_clip": 0.01136844, + "auxiliary_loss_mlp": 0.0102797, + "balance_loss_clip": 0.97749388, + "balance_loss_mlp": 1.01969039, + "epoch": 0.4885468646666266, + "flos": 24207132003840.0, + "grad_norm": 2.3905402603371924, + "language_loss": 0.67839515, + "learning_rate": 2.1711641281201092e-06, + "loss": 0.70004332, + "num_input_tokens_seen": 87505605, + "step": 4063, + "time_per_iteration": 2.7437214851379395 + }, + { + "auxiliary_loss_clip": 0.01159445, + "auxiliary_loss_mlp": 0.01025022, + "balance_loss_clip": 1.0202055, + "balance_loss_mlp": 1.01782179, + "epoch": 0.48866710755726567, + "flos": 14610795696000.0, + "grad_norm": 2.1565381885095847, + "language_loss": 0.79701853, + "learning_rate": 2.1703880014594264e-06, + "loss": 0.81886315, + "num_input_tokens_seen": 87523195, + "step": 4064, + "time_per_iteration": 2.6371073722839355 + }, + { + "auxiliary_loss_clip": 0.01088411, + "auxiliary_loss_mlp": 0.0102897, + "balance_loss_clip": 0.90919673, + "balance_loss_mlp": 1.0215348, + "epoch": 0.4887873504479048, + "flos": 28804451771520.0, + "grad_norm": 1.6158913081271806, + "language_loss": 0.74039859, + "learning_rate": 2.1696118489510182e-06, + "loss": 0.76157242, + "num_input_tokens_seen": 87544125, + "step": 4065, + "time_per_iteration": 2.7940993309020996 + }, + { + "auxiliary_loss_clip": 0.01116359, + "auxiliary_loss_mlp": 0.00510089, + "balance_loss_clip": 0.94135892, + "balance_loss_mlp": 0.0, + "epoch": 0.48890759333854383, + "flos": 22784387224320.0, + "grad_norm": 2.3673567501253228, + "language_loss": 0.7227608, + "learning_rate": 2.1688356707126286e-06, + "loss": 0.73902529, + "num_input_tokens_seen": 87563745, + "step": 4066, + "time_per_iteration": 2.68623423576355 + }, + { + "auxiliary_loss_clip": 0.01109282, + "auxiliary_loss_mlp": 0.01029317, + "balance_loss_clip": 0.94157171, + "balance_loss_mlp": 1.02071619, + "epoch": 0.48902783622918294, + "flos": 17786088956160.0, + "grad_norm": 2.219979352775843, + "language_loss": 0.70111966, + "learning_rate": 2.168059466862001e-06, + "loss": 0.72250569, + "num_input_tokens_seen": 87581895, + "step": 4067, + "time_per_iteration": 2.7686071395874023 + }, + { + "auxiliary_loss_clip": 0.0113412, + "auxiliary_loss_mlp": 0.01028631, + "balance_loss_clip": 0.97561157, + "balance_loss_mlp": 1.0210315, + "epoch": 0.48914807911982205, + "flos": 22310294590080.0, + "grad_norm": 2.04107811894973, + "language_loss": 0.81571871, + "learning_rate": 2.167283237516887e-06, + "loss": 0.8373462, + "num_input_tokens_seen": 87600170, + "step": 4068, + "time_per_iteration": 2.6674652099609375 + }, + { + "auxiliary_loss_clip": 0.01136737, + "auxiliary_loss_mlp": 0.01024758, + "balance_loss_clip": 0.97738945, + "balance_loss_mlp": 1.01656199, + "epoch": 0.4892683220104611, + "flos": 16363020954240.0, + "grad_norm": 2.0199985170634553, + "language_loss": 0.74848604, + "learning_rate": 2.1665069827950383e-06, + "loss": 0.77010095, + "num_input_tokens_seen": 87617455, + "step": 4069, + "time_per_iteration": 2.6466891765594482 + }, + { + "auxiliary_loss_clip": 0.0113402, + "auxiliary_loss_mlp": 0.01026326, + "balance_loss_clip": 0.97804916, + "balance_loss_mlp": 1.0188458, + "epoch": 0.4893885649011002, + "flos": 15739144606080.0, + "grad_norm": 3.3524753549001387, + "language_loss": 0.86756176, + "learning_rate": 2.1657307028142126e-06, + "loss": 0.88916522, + "num_input_tokens_seen": 87634995, + "step": 4070, + "time_per_iteration": 2.612438917160034 + }, + { + "auxiliary_loss_clip": 0.01134956, + "auxiliary_loss_mlp": 0.01028902, + "balance_loss_clip": 0.98029351, + "balance_loss_mlp": 1.02068305, + "epoch": 0.48950880779173933, + "flos": 28581984887040.0, + "grad_norm": 2.5684087114659913, + "language_loss": 0.67885053, + "learning_rate": 2.164954397692171e-06, + "loss": 0.7004891, + "num_input_tokens_seen": 87654420, + "step": 4071, + "time_per_iteration": 3.62798810005188 + }, + { + "auxiliary_loss_clip": 0.01049935, + "auxiliary_loss_mlp": 0.0100559, + "balance_loss_clip": 0.95438218, + "balance_loss_mlp": 1.00438643, + "epoch": 0.4896290506823784, + "flos": 66186310746240.0, + "grad_norm": 1.085759405643498, + "language_loss": 0.77271414, + "learning_rate": 2.164178067546678e-06, + "loss": 0.7932694, + "num_input_tokens_seen": 87713585, + "step": 4072, + "time_per_iteration": 3.3030736446380615 + }, + { + "auxiliary_loss_clip": 0.01139641, + "auxiliary_loss_mlp": 0.01023762, + "balance_loss_clip": 0.9766866, + "balance_loss_mlp": 1.0162344, + "epoch": 0.4897492935730175, + "flos": 12531065207040.0, + "grad_norm": 1.879771434880534, + "language_loss": 0.91102624, + "learning_rate": 2.163401712495504e-06, + "loss": 0.93266022, + "num_input_tokens_seen": 87731280, + "step": 4073, + "time_per_iteration": 3.6469297409057617 + }, + { + "auxiliary_loss_clip": 0.01087015, + "auxiliary_loss_mlp": 0.01025258, + "balance_loss_clip": 0.89979744, + "balance_loss_mlp": 1.0169971, + "epoch": 0.4898695364636566, + "flos": 23476816679040.0, + "grad_norm": 3.510990253384213, + "language_loss": 0.79187721, + "learning_rate": 2.1626253326564194e-06, + "loss": 0.81299996, + "num_input_tokens_seen": 87750230, + "step": 4074, + "time_per_iteration": 2.760862350463867 + }, + { + "auxiliary_loss_clip": 0.01132819, + "auxiliary_loss_mlp": 0.01027541, + "balance_loss_clip": 0.97587144, + "balance_loss_mlp": 1.01927352, + "epoch": 0.48998977935429566, + "flos": 27160209774720.0, + "grad_norm": 1.8615809368619203, + "language_loss": 0.77129763, + "learning_rate": 2.161848928147201e-06, + "loss": 0.79290128, + "num_input_tokens_seen": 87770500, + "step": 4075, + "time_per_iteration": 2.7009363174438477 + }, + { + "auxiliary_loss_clip": 0.01159749, + "auxiliary_loss_mlp": 0.01027434, + "balance_loss_clip": 1.01908541, + "balance_loss_mlp": 1.01913691, + "epoch": 0.4901100222449348, + "flos": 20339588856960.0, + "grad_norm": 1.9539634488827018, + "language_loss": 0.80734241, + "learning_rate": 2.161072499085629e-06, + "loss": 0.82921427, + "num_input_tokens_seen": 87789495, + "step": 4076, + "time_per_iteration": 2.6667487621307373 + }, + { + "auxiliary_loss_clip": 0.01116914, + "auxiliary_loss_mlp": 0.01033826, + "balance_loss_clip": 0.94228733, + "balance_loss_mlp": 1.0263989, + "epoch": 0.4902302651355739, + "flos": 30446359384320.0, + "grad_norm": 1.7008472196737507, + "language_loss": 0.83012354, + "learning_rate": 2.160296045589487e-06, + "loss": 0.85163093, + "num_input_tokens_seen": 87812955, + "step": 4077, + "time_per_iteration": 2.765544891357422 + }, + { + "auxiliary_loss_clip": 0.01160469, + "auxiliary_loss_mlp": 0.01023507, + "balance_loss_clip": 1.02051497, + "balance_loss_mlp": 1.01562107, + "epoch": 0.49035050802621294, + "flos": 19174180089600.0, + "grad_norm": 1.7351597550415383, + "language_loss": 0.70060474, + "learning_rate": 2.159519567776562e-06, + "loss": 0.72244447, + "num_input_tokens_seen": 87832605, + "step": 4078, + "time_per_iteration": 3.5609583854675293 + }, + { + "auxiliary_loss_clip": 0.01090838, + "auxiliary_loss_mlp": 0.01024177, + "balance_loss_clip": 0.89854413, + "balance_loss_mlp": 1.01672077, + "epoch": 0.49047075091685205, + "flos": 22228489365120.0, + "grad_norm": 2.672464392982511, + "language_loss": 0.71268117, + "learning_rate": 2.1587430657646463e-06, + "loss": 0.73383129, + "num_input_tokens_seen": 87846040, + "step": 4079, + "time_per_iteration": 2.753709077835083 + }, + { + "auxiliary_loss_clip": 0.011317, + "auxiliary_loss_mlp": 0.01028575, + "balance_loss_clip": 0.97867858, + "balance_loss_mlp": 1.02126729, + "epoch": 0.4905909938074911, + "flos": 20156516213760.0, + "grad_norm": 1.9736533923342203, + "language_loss": 0.78023607, + "learning_rate": 2.157966539671533e-06, + "loss": 0.80183887, + "num_input_tokens_seen": 87865680, + "step": 4080, + "time_per_iteration": 2.6538662910461426 + }, + { + "auxiliary_loss_clip": 0.01109787, + "auxiliary_loss_mlp": 0.0102858, + "balance_loss_clip": 0.93810636, + "balance_loss_mlp": 1.02135587, + "epoch": 0.4907112366981302, + "flos": 17202217380480.0, + "grad_norm": 2.077234080787375, + "language_loss": 0.67688501, + "learning_rate": 2.157189989615021e-06, + "loss": 0.69826871, + "num_input_tokens_seen": 87884270, + "step": 4081, + "time_per_iteration": 2.6827540397644043 + }, + { + "auxiliary_loss_clip": 0.01157716, + "auxiliary_loss_mlp": 0.00510961, + "balance_loss_clip": 1.01595271, + "balance_loss_mlp": 0.0, + "epoch": 0.4908314795887693, + "flos": 21688968107520.0, + "grad_norm": 1.7429334241496492, + "language_loss": 0.75203252, + "learning_rate": 2.156413415712913e-06, + "loss": 0.76871932, + "num_input_tokens_seen": 87906320, + "step": 4082, + "time_per_iteration": 2.656543493270874 + }, + { + "auxiliary_loss_clip": 0.01138453, + "auxiliary_loss_mlp": 0.00509899, + "balance_loss_clip": 0.97853929, + "balance_loss_mlp": 0.0, + "epoch": 0.4909517224794084, + "flos": 26213676531840.0, + "grad_norm": 1.6834443122032392, + "language_loss": 0.78776187, + "learning_rate": 2.155636818083014e-06, + "loss": 0.80424535, + "num_input_tokens_seen": 87927690, + "step": 4083, + "time_per_iteration": 2.726742744445801 + }, + { + "auxiliary_loss_clip": 0.01129436, + "auxiliary_loss_mlp": 0.01023415, + "balance_loss_clip": 0.97944301, + "balance_loss_mlp": 1.01631594, + "epoch": 0.4910719653700475, + "flos": 23148377694720.0, + "grad_norm": 1.8691023198330456, + "language_loss": 0.84637088, + "learning_rate": 2.154860196843134e-06, + "loss": 0.86789936, + "num_input_tokens_seen": 87946885, + "step": 4084, + "time_per_iteration": 2.6992485523223877 + }, + { + "auxiliary_loss_clip": 0.01182749, + "auxiliary_loss_mlp": 0.01033769, + "balance_loss_clip": 1.05564129, + "balance_loss_mlp": 1.02574062, + "epoch": 0.4911922082606866, + "flos": 23331845387520.0, + "grad_norm": 2.1668963816152327, + "language_loss": 0.77030379, + "learning_rate": 2.154083552111085e-06, + "loss": 0.79246897, + "num_input_tokens_seen": 87966055, + "step": 4085, + "time_per_iteration": 2.6043753623962402 + }, + { + "auxiliary_loss_clip": 0.01181523, + "auxiliary_loss_mlp": 0.01024004, + "balance_loss_clip": 1.0537287, + "balance_loss_mlp": 1.01577306, + "epoch": 0.49131245115132566, + "flos": 29203239542400.0, + "grad_norm": 1.7274821104476379, + "language_loss": 0.82103276, + "learning_rate": 2.1533068840046834e-06, + "loss": 0.84308803, + "num_input_tokens_seen": 87986320, + "step": 4086, + "time_per_iteration": 2.6611175537109375 + }, + { + "auxiliary_loss_clip": 0.01128188, + "auxiliary_loss_mlp": 0.00510426, + "balance_loss_clip": 0.97776854, + "balance_loss_mlp": 0.0, + "epoch": 0.49143269404196477, + "flos": 20147465986560.0, + "grad_norm": 2.4174642215599316, + "language_loss": 0.60966504, + "learning_rate": 2.152530192641749e-06, + "loss": 0.62605113, + "num_input_tokens_seen": 88001230, + "step": 4087, + "time_per_iteration": 2.624140739440918 + }, + { + "auxiliary_loss_clip": 0.01160842, + "auxiliary_loss_mlp": 0.01026783, + "balance_loss_clip": 1.01736093, + "balance_loss_mlp": 1.01969004, + "epoch": 0.4915529369326039, + "flos": 24389809597440.0, + "grad_norm": 2.3841918283198833, + "language_loss": 0.72576177, + "learning_rate": 2.1517534781401068e-06, + "loss": 0.74763799, + "num_input_tokens_seen": 88019110, + "step": 4088, + "time_per_iteration": 2.656325578689575 + }, + { + "auxiliary_loss_clip": 0.0115659, + "auxiliary_loss_mlp": 0.0102499, + "balance_loss_clip": 1.01563239, + "balance_loss_mlp": 1.01729512, + "epoch": 0.49167317982324293, + "flos": 10524305197440.0, + "grad_norm": 2.5268602369241573, + "language_loss": 0.694175, + "learning_rate": 2.150976740617581e-06, + "loss": 0.71599078, + "num_input_tokens_seen": 88035670, + "step": 4089, + "time_per_iteration": 2.589298963546753 + }, + { + "auxiliary_loss_clip": 0.01138081, + "auxiliary_loss_mlp": 0.0102986, + "balance_loss_clip": 0.97944814, + "balance_loss_mlp": 1.02247775, + "epoch": 0.49179342271388204, + "flos": 25593427457280.0, + "grad_norm": 2.8824813163749377, + "language_loss": 0.712825, + "learning_rate": 2.150199980192006e-06, + "loss": 0.73450434, + "num_input_tokens_seen": 88054790, + "step": 4090, + "time_per_iteration": 2.699211835861206 + }, + { + "auxiliary_loss_clip": 0.01132289, + "auxiliary_loss_mlp": 0.01024381, + "balance_loss_clip": 0.97935891, + "balance_loss_mlp": 1.01655507, + "epoch": 0.49191366560452116, + "flos": 21102043875840.0, + "grad_norm": 1.6775545641087553, + "language_loss": 0.81026977, + "learning_rate": 2.1494231969812114e-06, + "loss": 0.83183646, + "num_input_tokens_seen": 88073780, + "step": 4091, + "time_per_iteration": 2.6184751987457275 + }, + { + "auxiliary_loss_clip": 0.01117629, + "auxiliary_loss_mlp": 0.01030322, + "balance_loss_clip": 0.9427672, + "balance_loss_mlp": 1.02299654, + "epoch": 0.4920339084951602, + "flos": 26067520091520.0, + "grad_norm": 4.718160769253759, + "language_loss": 0.81009746, + "learning_rate": 2.1486463911030372e-06, + "loss": 0.83157694, + "num_input_tokens_seen": 88094430, + "step": 4092, + "time_per_iteration": 2.7451748847961426 + }, + { + "auxiliary_loss_clip": 0.01130438, + "auxiliary_loss_mlp": 0.01023724, + "balance_loss_clip": 0.97580069, + "balance_loss_mlp": 1.01624918, + "epoch": 0.4921541513857993, + "flos": 25081269384960.0, + "grad_norm": 2.439624583988409, + "language_loss": 0.74422705, + "learning_rate": 2.147869562675324e-06, + "loss": 0.76576871, + "num_input_tokens_seen": 88113400, + "step": 4093, + "time_per_iteration": 2.781546115875244 + }, + { + "auxiliary_loss_clip": 0.01158416, + "auxiliary_loss_mlp": 0.01029338, + "balance_loss_clip": 1.01801085, + "balance_loss_mlp": 1.02136922, + "epoch": 0.49227439427643843, + "flos": 24389809597440.0, + "grad_norm": 2.070950069802507, + "language_loss": 0.72787511, + "learning_rate": 2.147092711815915e-06, + "loss": 0.7497527, + "num_input_tokens_seen": 88132750, + "step": 4094, + "time_per_iteration": 2.8037166595458984 + }, + { + "auxiliary_loss_clip": 0.01109212, + "auxiliary_loss_mlp": 0.01028752, + "balance_loss_clip": 0.94050068, + "balance_loss_mlp": 1.02100921, + "epoch": 0.4923946371670775, + "flos": 11363753018880.0, + "grad_norm": 3.4606088207084196, + "language_loss": 0.86458427, + "learning_rate": 2.1463158386426593e-06, + "loss": 0.88596392, + "num_input_tokens_seen": 88150560, + "step": 4095, + "time_per_iteration": 2.6951847076416016 + }, + { + "auxiliary_loss_clip": 0.01139185, + "auxiliary_loss_mlp": 0.01031147, + "balance_loss_clip": 0.98044288, + "balance_loss_mlp": 1.0233748, + "epoch": 0.4925148800577166, + "flos": 30445964334720.0, + "grad_norm": 1.8622852105387881, + "language_loss": 0.77421218, + "learning_rate": 2.145538943273407e-06, + "loss": 0.79591548, + "num_input_tokens_seen": 88170835, + "step": 4096, + "time_per_iteration": 3.6286940574645996 + }, + { + "auxiliary_loss_clip": 0.01185316, + "auxiliary_loss_mlp": 0.01029513, + "balance_loss_clip": 1.05955815, + "balance_loss_mlp": 1.02163899, + "epoch": 0.49263512294835565, + "flos": 20850454039680.0, + "grad_norm": 1.768621720922363, + "language_loss": 0.71963781, + "learning_rate": 2.144762025826013e-06, + "loss": 0.74178612, + "num_input_tokens_seen": 88189925, + "step": 4097, + "time_per_iteration": 3.481782913208008 + }, + { + "auxiliary_loss_clip": 0.01161507, + "auxiliary_loss_mlp": 0.01033748, + "balance_loss_clip": 1.01666367, + "balance_loss_mlp": 1.02589202, + "epoch": 0.49275536583899476, + "flos": 23767477534080.0, + "grad_norm": 2.052879463646717, + "language_loss": 0.87186456, + "learning_rate": 2.143985086418334e-06, + "loss": 0.89381719, + "num_input_tokens_seen": 88205105, + "step": 4098, + "time_per_iteration": 2.658201217651367 + }, + { + "auxiliary_loss_clip": 0.01136824, + "auxiliary_loss_mlp": 0.01024621, + "balance_loss_clip": 0.97973955, + "balance_loss_mlp": 1.01744497, + "epoch": 0.4928756087296339, + "flos": 22273522041600.0, + "grad_norm": 1.7232099223502828, + "language_loss": 0.76708341, + "learning_rate": 2.1432081251682324e-06, + "loss": 0.78869784, + "num_input_tokens_seen": 88225475, + "step": 4099, + "time_per_iteration": 3.632908582687378 + }, + { + "auxiliary_loss_clip": 0.01160763, + "auxiliary_loss_mlp": 0.01031411, + "balance_loss_clip": 1.02330256, + "balance_loss_mlp": 1.02318525, + "epoch": 0.49299585162027293, + "flos": 19645471463040.0, + "grad_norm": 1.7827768609667942, + "language_loss": 0.87298185, + "learning_rate": 2.142431142193572e-06, + "loss": 0.89490354, + "num_input_tokens_seen": 88243255, + "step": 4100, + "time_per_iteration": 2.660266399383545 + }, + { + "auxiliary_loss_clip": 0.01179715, + "auxiliary_loss_mlp": 0.01031624, + "balance_loss_clip": 1.05600262, + "balance_loss_mlp": 1.02411687, + "epoch": 0.49311609451091204, + "flos": 38837138497920.0, + "grad_norm": 2.3677111958877326, + "language_loss": 0.71545696, + "learning_rate": 2.1416541376122207e-06, + "loss": 0.73757035, + "num_input_tokens_seen": 88263435, + "step": 4101, + "time_per_iteration": 2.740380048751831 + }, + { + "auxiliary_loss_clip": 0.01181165, + "auxiliary_loss_mlp": 0.01026959, + "balance_loss_clip": 1.0551641, + "balance_loss_mlp": 1.01850104, + "epoch": 0.49323633740155115, + "flos": 28329102161280.0, + "grad_norm": 1.9452377228873994, + "language_loss": 0.73145807, + "learning_rate": 2.1408771115420496e-06, + "loss": 0.75353932, + "num_input_tokens_seen": 88283295, + "step": 4102, + "time_per_iteration": 2.712284564971924 + }, + { + "auxiliary_loss_clip": 0.01095347, + "auxiliary_loss_mlp": 0.01027097, + "balance_loss_clip": 0.90981954, + "balance_loss_mlp": 1.01897895, + "epoch": 0.4933565802921902, + "flos": 21135584200320.0, + "grad_norm": 2.63611505567524, + "language_loss": 0.65003121, + "learning_rate": 2.140100064100932e-06, + "loss": 0.67125559, + "num_input_tokens_seen": 88299270, + "step": 4103, + "time_per_iteration": 2.725717067718506 + }, + { + "auxiliary_loss_clip": 0.01151969, + "auxiliary_loss_mlp": 0.01023171, + "balance_loss_clip": 1.0156796, + "balance_loss_mlp": 1.01556516, + "epoch": 0.4934768231828293, + "flos": 18039007595520.0, + "grad_norm": 1.8964009926082595, + "language_loss": 0.75589657, + "learning_rate": 2.139322995406746e-06, + "loss": 0.77764797, + "num_input_tokens_seen": 88316905, + "step": 4104, + "time_per_iteration": 3.554745674133301 + }, + { + "auxiliary_loss_clip": 0.01182211, + "auxiliary_loss_mlp": 0.01032237, + "balance_loss_clip": 1.05727983, + "balance_loss_mlp": 1.0246315, + "epoch": 0.4935970660734684, + "flos": 23469957181440.0, + "grad_norm": 2.119496662268102, + "language_loss": 0.79743719, + "learning_rate": 2.1385459055773727e-06, + "loss": 0.81958169, + "num_input_tokens_seen": 88335095, + "step": 4105, + "time_per_iteration": 2.6017518043518066 + }, + { + "auxiliary_loss_clip": 0.01072392, + "auxiliary_loss_mlp": 0.00508736, + "balance_loss_clip": 0.89827204, + "balance_loss_mlp": 0.0, + "epoch": 0.4937173089641075, + "flos": 64479258840960.0, + "grad_norm": 2.0571964378432366, + "language_loss": 0.73913801, + "learning_rate": 2.137768794730696e-06, + "loss": 0.75494933, + "num_input_tokens_seen": 88358545, + "step": 4106, + "time_per_iteration": 3.184290885925293 + }, + { + "auxiliary_loss_clip": 0.01138248, + "auxiliary_loss_mlp": 0.01029402, + "balance_loss_clip": 0.98305959, + "balance_loss_mlp": 1.02139115, + "epoch": 0.4938375518547466, + "flos": 22346025644160.0, + "grad_norm": 1.6738830608369797, + "language_loss": 0.80612397, + "learning_rate": 2.1369916629846026e-06, + "loss": 0.82780051, + "num_input_tokens_seen": 88378295, + "step": 4107, + "time_per_iteration": 2.733976125717163 + }, + { + "auxiliary_loss_clip": 0.01131333, + "auxiliary_loss_mlp": 0.0102289, + "balance_loss_clip": 0.97587055, + "balance_loss_mlp": 1.01535034, + "epoch": 0.4939577947453857, + "flos": 17858700299520.0, + "grad_norm": 1.8059867949389767, + "language_loss": 0.74911189, + "learning_rate": 2.136214510456983e-06, + "loss": 0.77065414, + "num_input_tokens_seen": 88396750, + "step": 4108, + "time_per_iteration": 2.7420766353607178 + }, + { + "auxiliary_loss_clip": 0.01015327, + "auxiliary_loss_mlp": 0.00500494, + "balance_loss_clip": 0.88210559, + "balance_loss_mlp": 0.0, + "epoch": 0.49407803763602476, + "flos": 70066746875520.0, + "grad_norm": 0.893206189857498, + "language_loss": 0.63160825, + "learning_rate": 2.1354373372657296e-06, + "loss": 0.64676654, + "num_input_tokens_seen": 88455190, + "step": 4109, + "time_per_iteration": 3.314390182495117 + }, + { + "auxiliary_loss_clip": 0.01181912, + "auxiliary_loss_mlp": 0.0103096, + "balance_loss_clip": 1.05853438, + "balance_loss_mlp": 1.02340853, + "epoch": 0.49419828052666387, + "flos": 24317485562880.0, + "grad_norm": 1.6381413753011047, + "language_loss": 0.71047521, + "learning_rate": 2.1346601435287404e-06, + "loss": 0.73260391, + "num_input_tokens_seen": 88477460, + "step": 4110, + "time_per_iteration": 2.6273860931396484 + }, + { + "auxiliary_loss_clip": 0.01131308, + "auxiliary_loss_mlp": 0.01027384, + "balance_loss_clip": 0.97662032, + "balance_loss_mlp": 1.02033865, + "epoch": 0.494318523417303, + "flos": 29386060790400.0, + "grad_norm": 1.7641047721008158, + "language_loss": 0.79962802, + "learning_rate": 2.1338829293639144e-06, + "loss": 0.82121491, + "num_input_tokens_seen": 88497820, + "step": 4111, + "time_per_iteration": 2.7039902210235596 + }, + { + "auxiliary_loss_clip": 0.01085942, + "auxiliary_loss_mlp": 0.01026272, + "balance_loss_clip": 0.90381706, + "balance_loss_mlp": 1.01844037, + "epoch": 0.49443876630794203, + "flos": 15268284195840.0, + "grad_norm": 2.1683591060744516, + "language_loss": 0.82998687, + "learning_rate": 2.1331056948891547e-06, + "loss": 0.85110903, + "num_input_tokens_seen": 88514920, + "step": 4112, + "time_per_iteration": 2.750758409500122 + }, + { + "auxiliary_loss_clip": 0.01128128, + "auxiliary_loss_mlp": 0.0103072, + "balance_loss_clip": 0.97591591, + "balance_loss_mlp": 1.02302516, + "epoch": 0.49455900919858115, + "flos": 12347453859840.0, + "grad_norm": 3.634640409041838, + "language_loss": 0.7647385, + "learning_rate": 2.1323284402223666e-06, + "loss": 0.786327, + "num_input_tokens_seen": 88530910, + "step": 4113, + "time_per_iteration": 2.690028429031372 + }, + { + "auxiliary_loss_clip": 0.01180951, + "auxiliary_loss_mlp": 0.00508575, + "balance_loss_clip": 1.06003881, + "balance_loss_mlp": 0.0, + "epoch": 0.4946792520892202, + "flos": 22779610715520.0, + "grad_norm": 1.8492897026951576, + "language_loss": 0.88323951, + "learning_rate": 2.1315511654814597e-06, + "loss": 0.90013474, + "num_input_tokens_seen": 88549320, + "step": 4114, + "time_per_iteration": 2.6569125652313232 + }, + { + "auxiliary_loss_clip": 0.01124799, + "auxiliary_loss_mlp": 0.01023913, + "balance_loss_clip": 0.97893149, + "balance_loss_mlp": 1.01721358, + "epoch": 0.4947994949798593, + "flos": 23148126299520.0, + "grad_norm": 1.8882811129557198, + "language_loss": 0.78423721, + "learning_rate": 2.1307738707843456e-06, + "loss": 0.80572438, + "num_input_tokens_seen": 88568985, + "step": 4115, + "time_per_iteration": 2.6777501106262207 + }, + { + "auxiliary_loss_clip": 0.0116522, + "auxiliary_loss_mlp": 0.01027825, + "balance_loss_clip": 1.02053547, + "balance_loss_mlp": 1.01946795, + "epoch": 0.4949197378704984, + "flos": 23659997063040.0, + "grad_norm": 2.0359984042352814, + "language_loss": 0.69071198, + "learning_rate": 2.1299965562489385e-06, + "loss": 0.71264243, + "num_input_tokens_seen": 88588790, + "step": 4116, + "time_per_iteration": 2.6871323585510254 + }, + { + "auxiliary_loss_clip": 0.01151101, + "auxiliary_loss_mlp": 0.01026782, + "balance_loss_clip": 1.0125463, + "balance_loss_mlp": 1.0196054, + "epoch": 0.4950399807611375, + "flos": 26911493026560.0, + "grad_norm": 1.5592354673653628, + "language_loss": 0.78828585, + "learning_rate": 2.129219221993158e-06, + "loss": 0.81006467, + "num_input_tokens_seen": 88613575, + "step": 4117, + "time_per_iteration": 2.7030553817749023 + }, + { + "auxiliary_loss_clip": 0.010207, + "auxiliary_loss_mlp": 0.01001316, + "balance_loss_clip": 0.91126871, + "balance_loss_mlp": 0.99974293, + "epoch": 0.4951602236517766, + "flos": 67315270187520.0, + "grad_norm": 0.7806681327192192, + "language_loss": 0.59885395, + "learning_rate": 2.128441868134924e-06, + "loss": 0.61907411, + "num_input_tokens_seen": 88675510, + "step": 4118, + "time_per_iteration": 3.3428752422332764 + }, + { + "auxiliary_loss_clip": 0.01113323, + "auxiliary_loss_mlp": 0.01029702, + "balance_loss_clip": 0.93826878, + "balance_loss_mlp": 1.02225137, + "epoch": 0.4952804665424157, + "flos": 19901442758400.0, + "grad_norm": 2.316196126420282, + "language_loss": 0.82651871, + "learning_rate": 2.1276644947921606e-06, + "loss": 0.84794903, + "num_input_tokens_seen": 88694425, + "step": 4119, + "time_per_iteration": 2.6380844116210938 + }, + { + "auxiliary_loss_clip": 0.01155852, + "auxiliary_loss_mlp": 0.01030459, + "balance_loss_clip": 1.01542652, + "balance_loss_mlp": 1.0230267, + "epoch": 0.49540070943305475, + "flos": 18806813740800.0, + "grad_norm": 1.974331337433516, + "language_loss": 0.82575452, + "learning_rate": 2.126887102082795e-06, + "loss": 0.84761763, + "num_input_tokens_seen": 88714450, + "step": 4120, + "time_per_iteration": 2.5758705139160156 + }, + { + "auxiliary_loss_clip": 0.01106294, + "auxiliary_loss_mlp": 0.01027507, + "balance_loss_clip": 0.93662214, + "balance_loss_mlp": 1.02001452, + "epoch": 0.49552095232369386, + "flos": 24934179191040.0, + "grad_norm": 2.0089895469097168, + "language_loss": 0.70023406, + "learning_rate": 2.126109690124757e-06, + "loss": 0.72157204, + "num_input_tokens_seen": 88735265, + "step": 4121, + "time_per_iteration": 2.6111860275268555 + }, + { + "auxiliary_loss_clip": 0.01085096, + "auxiliary_loss_mlp": 0.01027367, + "balance_loss_clip": 0.89903694, + "balance_loss_mlp": 1.01972532, + "epoch": 0.495641195214333, + "flos": 22857249962880.0, + "grad_norm": 1.8321107127208633, + "language_loss": 0.71563256, + "learning_rate": 2.1253322590359786e-06, + "loss": 0.73675716, + "num_input_tokens_seen": 88754600, + "step": 4122, + "time_per_iteration": 3.5007028579711914 + }, + { + "auxiliary_loss_clip": 0.01152193, + "auxiliary_loss_mlp": 0.01026104, + "balance_loss_clip": 1.0136044, + "balance_loss_mlp": 1.01884961, + "epoch": 0.49576143810497203, + "flos": 25769748343680.0, + "grad_norm": 2.3440334665480798, + "language_loss": 0.74506581, + "learning_rate": 2.124554808934397e-06, + "loss": 0.7668488, + "num_input_tokens_seen": 88775180, + "step": 4123, + "time_per_iteration": 3.41247296333313 + }, + { + "auxiliary_loss_clip": 0.01080809, + "auxiliary_loss_mlp": 0.01029653, + "balance_loss_clip": 0.89820951, + "balance_loss_mlp": 1.02238142, + "epoch": 0.49588168099561114, + "flos": 22128838058880.0, + "grad_norm": 1.9779562744060977, + "language_loss": 0.72875977, + "learning_rate": 2.1237773399379496e-06, + "loss": 0.74986446, + "num_input_tokens_seen": 88796145, + "step": 4124, + "time_per_iteration": 2.7190639972686768 + }, + { + "auxiliary_loss_clip": 0.01133018, + "auxiliary_loss_mlp": 0.01023359, + "balance_loss_clip": 0.97230607, + "balance_loss_mlp": 1.01594377, + "epoch": 0.49600192388625025, + "flos": 24387331559040.0, + "grad_norm": 2.587737989603024, + "language_loss": 0.86975503, + "learning_rate": 2.122999852164578e-06, + "loss": 0.8913188, + "num_input_tokens_seen": 88816765, + "step": 4125, + "time_per_iteration": 3.8319027423858643 + }, + { + "auxiliary_loss_clip": 0.01083294, + "auxiliary_loss_mlp": 0.01024686, + "balance_loss_clip": 0.90147877, + "balance_loss_mlp": 1.0170151, + "epoch": 0.4961221667768893, + "flos": 22857429530880.0, + "grad_norm": 2.61339303009678, + "language_loss": 0.58274955, + "learning_rate": 2.122222345732227e-06, + "loss": 0.60382938, + "num_input_tokens_seen": 88836680, + "step": 4126, + "time_per_iteration": 2.751278877258301 + }, + { + "auxiliary_loss_clip": 0.01106306, + "auxiliary_loss_mlp": 0.0102367, + "balance_loss_clip": 0.93579608, + "balance_loss_mlp": 1.01584411, + "epoch": 0.4962424096675284, + "flos": 17858089768320.0, + "grad_norm": 1.837470703327666, + "language_loss": 0.82950586, + "learning_rate": 2.121444820758843e-06, + "loss": 0.85080564, + "num_input_tokens_seen": 88855320, + "step": 4127, + "time_per_iteration": 2.641707420349121 + }, + { + "auxiliary_loss_clip": 0.01083698, + "auxiliary_loss_mlp": 0.01030956, + "balance_loss_clip": 0.89982438, + "balance_loss_mlp": 1.02290893, + "epoch": 0.49636265255816747, + "flos": 21793611404160.0, + "grad_norm": 2.027413987405693, + "language_loss": 0.78655386, + "learning_rate": 2.120667277362376e-06, + "loss": 0.8077004, + "num_input_tokens_seen": 88874035, + "step": 4128, + "time_per_iteration": 2.7904016971588135 + }, + { + "auxiliary_loss_clip": 0.01182773, + "auxiliary_loss_mlp": 0.01031126, + "balance_loss_clip": 1.05679417, + "balance_loss_mlp": 1.02310336, + "epoch": 0.4964828954488066, + "flos": 16358603581440.0, + "grad_norm": 2.058168137392483, + "language_loss": 0.84921139, + "learning_rate": 2.1198897156607796e-06, + "loss": 0.87135041, + "num_input_tokens_seen": 88891390, + "step": 4129, + "time_per_iteration": 2.587313652038574 + }, + { + "auxiliary_loss_clip": 0.01160729, + "auxiliary_loss_mlp": 0.01025415, + "balance_loss_clip": 1.01628768, + "balance_loss_mlp": 1.01806581, + "epoch": 0.4966031383394457, + "flos": 24711101775360.0, + "grad_norm": 2.4932368222757906, + "language_loss": 0.73722041, + "learning_rate": 2.1191121357720085e-06, + "loss": 0.75908184, + "num_input_tokens_seen": 88909450, + "step": 4130, + "time_per_iteration": 3.616581678390503 + }, + { + "auxiliary_loss_clip": 0.01077721, + "auxiliary_loss_mlp": 0.01026675, + "balance_loss_clip": 0.90005529, + "balance_loss_mlp": 1.0191474, + "epoch": 0.49672338123008475, + "flos": 22930615491840.0, + "grad_norm": 1.9506486010150212, + "language_loss": 0.74425626, + "learning_rate": 2.1183345378140206e-06, + "loss": 0.76530015, + "num_input_tokens_seen": 88929195, + "step": 4131, + "time_per_iteration": 2.767815351486206 + }, + { + "auxiliary_loss_clip": 0.01065019, + "auxiliary_loss_mlp": 0.01002831, + "balance_loss_clip": 0.98727477, + "balance_loss_mlp": 1.00131691, + "epoch": 0.49684362412072386, + "flos": 65976736844160.0, + "grad_norm": 0.8552047356498014, + "language_loss": 0.62004793, + "learning_rate": 2.1175569219047783e-06, + "loss": 0.64072645, + "num_input_tokens_seen": 88990635, + "step": 4132, + "time_per_iteration": 3.3082351684570312 + }, + { + "auxiliary_loss_clip": 0.01177907, + "auxiliary_loss_mlp": 0.01025743, + "balance_loss_clip": 1.05390489, + "balance_loss_mlp": 1.01780999, + "epoch": 0.49696386701136297, + "flos": 19971288754560.0, + "grad_norm": 1.7272569097335975, + "language_loss": 0.73267043, + "learning_rate": 2.1167792881622437e-06, + "loss": 0.75470686, + "num_input_tokens_seen": 89009655, + "step": 4133, + "time_per_iteration": 2.638909101486206 + }, + { + "auxiliary_loss_clip": 0.01127921, + "auxiliary_loss_mlp": 0.01026114, + "balance_loss_clip": 0.977301, + "balance_loss_mlp": 1.01886606, + "epoch": 0.497084109902002, + "flos": 24750819239040.0, + "grad_norm": 1.6124968275786355, + "language_loss": 0.80876267, + "learning_rate": 2.116001636704384e-06, + "loss": 0.83030295, + "num_input_tokens_seen": 89030040, + "step": 4134, + "time_per_iteration": 2.657081127166748 + }, + { + "auxiliary_loss_clip": 0.01093993, + "auxiliary_loss_mlp": 0.01026411, + "balance_loss_clip": 0.90114939, + "balance_loss_mlp": 1.01901448, + "epoch": 0.49720435279264114, + "flos": 21871825269120.0, + "grad_norm": 1.811136123207159, + "language_loss": 0.80316025, + "learning_rate": 2.1152239676491685e-06, + "loss": 0.82436424, + "num_input_tokens_seen": 89048145, + "step": 4135, + "time_per_iteration": 2.736506462097168 + }, + { + "auxiliary_loss_clip": 0.01136546, + "auxiliary_loss_mlp": 0.01031012, + "balance_loss_clip": 0.97567147, + "balance_loss_mlp": 1.02391315, + "epoch": 0.49732459568328025, + "flos": 23805794367360.0, + "grad_norm": 3.2319635563779854, + "language_loss": 0.73557651, + "learning_rate": 2.114446281114569e-06, + "loss": 0.7572521, + "num_input_tokens_seen": 89067165, + "step": 4136, + "time_per_iteration": 2.6660549640655518 + }, + { + "auxiliary_loss_clip": 0.01121856, + "auxiliary_loss_mlp": 0.01029155, + "balance_loss_clip": 0.97479832, + "balance_loss_mlp": 1.02193046, + "epoch": 0.4974448385739193, + "flos": 20047742853120.0, + "grad_norm": 1.986819532850631, + "language_loss": 0.75997221, + "learning_rate": 2.1136685772185587e-06, + "loss": 0.78148234, + "num_input_tokens_seen": 89086190, + "step": 4137, + "time_per_iteration": 2.6918482780456543 + }, + { + "auxiliary_loss_clip": 0.01130075, + "auxiliary_loss_mlp": 0.0050992, + "balance_loss_clip": 0.97176611, + "balance_loss_mlp": 0.0, + "epoch": 0.4975650814645584, + "flos": 24821347593600.0, + "grad_norm": 2.0591116850532374, + "language_loss": 0.78070217, + "learning_rate": 2.1128908560791163e-06, + "loss": 0.79710215, + "num_input_tokens_seen": 89106020, + "step": 4138, + "time_per_iteration": 2.692721128463745 + }, + { + "auxiliary_loss_clip": 0.01177078, + "auxiliary_loss_mlp": 0.01024192, + "balance_loss_clip": 1.0537082, + "balance_loss_mlp": 1.01655102, + "epoch": 0.4976853243551975, + "flos": 19829477859840.0, + "grad_norm": 2.962319039937861, + "language_loss": 0.78014117, + "learning_rate": 2.1121131178142203e-06, + "loss": 0.80215389, + "num_input_tokens_seen": 89125385, + "step": 4139, + "time_per_iteration": 2.635925531387329 + }, + { + "auxiliary_loss_clip": 0.01130999, + "auxiliary_loss_mlp": 0.01024637, + "balance_loss_clip": 0.97616816, + "balance_loss_mlp": 1.01741838, + "epoch": 0.4978055672458366, + "flos": 23142990654720.0, + "grad_norm": 1.5565041265632122, + "language_loss": 0.82625419, + "learning_rate": 2.1113353625418544e-06, + "loss": 0.84781051, + "num_input_tokens_seen": 89143935, + "step": 4140, + "time_per_iteration": 2.6415624618530273 + }, + { + "auxiliary_loss_clip": 0.01150607, + "auxiliary_loss_mlp": 0.01032878, + "balance_loss_clip": 1.0185374, + "balance_loss_mlp": 1.02563298, + "epoch": 0.4979258101364757, + "flos": 15559914718080.0, + "grad_norm": 1.6387507003971202, + "language_loss": 0.79191101, + "learning_rate": 2.1105575903800017e-06, + "loss": 0.8137458, + "num_input_tokens_seen": 89162655, + "step": 4141, + "time_per_iteration": 2.649446964263916 + }, + { + "auxiliary_loss_clip": 0.01161222, + "auxiliary_loss_mlp": 0.01026008, + "balance_loss_clip": 1.0165118, + "balance_loss_mlp": 1.01840806, + "epoch": 0.4980460530271148, + "flos": 26356169784960.0, + "grad_norm": 1.7546155110030954, + "language_loss": 0.85218328, + "learning_rate": 2.1097798014466502e-06, + "loss": 0.87405556, + "num_input_tokens_seen": 89182255, + "step": 4142, + "time_per_iteration": 2.638462781906128 + }, + { + "auxiliary_loss_clip": 0.01162268, + "auxiliary_loss_mlp": 0.01027241, + "balance_loss_clip": 1.01816988, + "balance_loss_mlp": 1.01958752, + "epoch": 0.49816629591775385, + "flos": 17274541415040.0, + "grad_norm": 2.494166581562554, + "language_loss": 0.59240913, + "learning_rate": 2.109001995859791e-06, + "loss": 0.61430418, + "num_input_tokens_seen": 89201155, + "step": 4143, + "time_per_iteration": 2.634525775909424 + }, + { + "auxiliary_loss_clip": 0.0103927, + "auxiliary_loss_mlp": 0.01002887, + "balance_loss_clip": 0.94831288, + "balance_loss_mlp": 1.00152767, + "epoch": 0.49828653880839296, + "flos": 64930947344640.0, + "grad_norm": 0.802011153044224, + "language_loss": 0.60111576, + "learning_rate": 2.108224173737415e-06, + "loss": 0.62153733, + "num_input_tokens_seen": 89264455, + "step": 4144, + "time_per_iteration": 3.2075068950653076 + }, + { + "auxiliary_loss_clip": 0.01133279, + "auxiliary_loss_mlp": 0.0102927, + "balance_loss_clip": 0.97817576, + "balance_loss_mlp": 1.02036524, + "epoch": 0.498406781699032, + "flos": 27484806003840.0, + "grad_norm": 1.7413912599245032, + "language_loss": 0.76038671, + "learning_rate": 2.1074463351975183e-06, + "loss": 0.78201216, + "num_input_tokens_seen": 89283340, + "step": 4145, + "time_per_iteration": 2.7070257663726807 + }, + { + "auxiliary_loss_clip": 0.01114888, + "auxiliary_loss_mlp": 0.01028689, + "balance_loss_clip": 0.94009399, + "balance_loss_mlp": 1.02149451, + "epoch": 0.49852702458967113, + "flos": 31499870307840.0, + "grad_norm": 1.7012198987899367, + "language_loss": 0.71189743, + "learning_rate": 2.106668480358098e-06, + "loss": 0.73333323, + "num_input_tokens_seen": 89303565, + "step": 4146, + "time_per_iteration": 2.778109312057495 + }, + { + "auxiliary_loss_clip": 0.01119578, + "auxiliary_loss_mlp": 0.01028566, + "balance_loss_clip": 0.93781829, + "balance_loss_mlp": 1.02112436, + "epoch": 0.49864726748031024, + "flos": 22852868503680.0, + "grad_norm": 1.8213151956675544, + "language_loss": 0.70862031, + "learning_rate": 2.105890609337154e-06, + "loss": 0.73010182, + "num_input_tokens_seen": 89322080, + "step": 4147, + "time_per_iteration": 2.6876533031463623 + }, + { + "auxiliary_loss_clip": 0.01080188, + "auxiliary_loss_mlp": 0.01000214, + "balance_loss_clip": 1.02225447, + "balance_loss_mlp": 0.99879569, + "epoch": 0.4987675103709493, + "flos": 70405708544640.0, + "grad_norm": 0.6900785954632289, + "language_loss": 0.63795274, + "learning_rate": 2.1051127222526883e-06, + "loss": 0.65875673, + "num_input_tokens_seen": 89394195, + "step": 4148, + "time_per_iteration": 4.197906732559204 + }, + { + "auxiliary_loss_clip": 0.01152987, + "auxiliary_loss_mlp": 0.01030558, + "balance_loss_clip": 1.01710975, + "balance_loss_mlp": 1.0231725, + "epoch": 0.4988877532615884, + "flos": 28767571482240.0, + "grad_norm": 2.1519204877292832, + "language_loss": 0.80679059, + "learning_rate": 2.1043348192227067e-06, + "loss": 0.82862604, + "num_input_tokens_seen": 89414565, + "step": 4149, + "time_per_iteration": 3.57411527633667 + }, + { + "auxiliary_loss_clip": 0.01099951, + "auxiliary_loss_mlp": 0.01031074, + "balance_loss_clip": 0.93806946, + "balance_loss_mlp": 1.02371287, + "epoch": 0.4990079961522275, + "flos": 16872700988160.0, + "grad_norm": 1.8604837135644476, + "language_loss": 0.61815178, + "learning_rate": 2.1035569003652156e-06, + "loss": 0.63946199, + "num_input_tokens_seen": 89433195, + "step": 4150, + "time_per_iteration": 2.7351911067962646 + }, + { + "auxiliary_loss_clip": 0.01082959, + "auxiliary_loss_mlp": 0.01026521, + "balance_loss_clip": 0.90040714, + "balance_loss_mlp": 1.01772392, + "epoch": 0.4991282390428666, + "flos": 13291042187520.0, + "grad_norm": 2.1193011220746185, + "language_loss": 0.81662738, + "learning_rate": 2.1027789657982255e-06, + "loss": 0.83772218, + "num_input_tokens_seen": 89447410, + "step": 4151, + "time_per_iteration": 3.6072025299072266 + }, + { + "auxiliary_loss_clip": 0.01085575, + "auxiliary_loss_mlp": 0.0102051, + "balance_loss_clip": 0.90079552, + "balance_loss_mlp": 1.01320851, + "epoch": 0.4992484819335057, + "flos": 21537496454400.0, + "grad_norm": 2.179452023200161, + "language_loss": 0.77358401, + "learning_rate": 2.1020010156397482e-06, + "loss": 0.79464483, + "num_input_tokens_seen": 89464630, + "step": 4152, + "time_per_iteration": 2.7803142070770264 + }, + { + "auxiliary_loss_clip": 0.01160193, + "auxiliary_loss_mlp": 0.01024802, + "balance_loss_clip": 1.01748145, + "balance_loss_mlp": 1.01713085, + "epoch": 0.4993687248241448, + "flos": 24860095390080.0, + "grad_norm": 2.1928318040031862, + "language_loss": 0.77398396, + "learning_rate": 2.101223050007797e-06, + "loss": 0.79583395, + "num_input_tokens_seen": 89483180, + "step": 4153, + "time_per_iteration": 2.665576219558716 + }, + { + "auxiliary_loss_clip": 0.01080782, + "auxiliary_loss_mlp": 0.01002664, + "balance_loss_clip": 1.02303457, + "balance_loss_mlp": 1.00131691, + "epoch": 0.49948896771478385, + "flos": 62941602453120.0, + "grad_norm": 0.8145841390328538, + "language_loss": 0.53782153, + "learning_rate": 2.1004450690203904e-06, + "loss": 0.55865598, + "num_input_tokens_seen": 89539260, + "step": 4154, + "time_per_iteration": 3.265080451965332 + }, + { + "auxiliary_loss_clip": 0.01080262, + "auxiliary_loss_mlp": 0.01001842, + "balance_loss_clip": 1.02235222, + "balance_loss_mlp": 1.00044715, + "epoch": 0.49960921060542296, + "flos": 68284213516800.0, + "grad_norm": 0.8583107032309429, + "language_loss": 0.63369298, + "learning_rate": 2.099667072795546e-06, + "loss": 0.65451401, + "num_input_tokens_seen": 89601380, + "step": 4155, + "time_per_iteration": 3.253976345062256 + }, + { + "auxiliary_loss_clip": 0.01155968, + "auxiliary_loss_mlp": 0.01023253, + "balance_loss_clip": 1.01634002, + "balance_loss_mlp": 1.01571631, + "epoch": 0.49972945349606207, + "flos": 23659350618240.0, + "grad_norm": 1.858141528794452, + "language_loss": 0.79832155, + "learning_rate": 2.0988890614512864e-06, + "loss": 0.82011378, + "num_input_tokens_seen": 89621270, + "step": 4156, + "time_per_iteration": 2.6628520488739014 + }, + { + "auxiliary_loss_clip": 0.01134129, + "auxiliary_loss_mlp": 0.01028024, + "balance_loss_clip": 0.97930932, + "balance_loss_mlp": 1.02045119, + "epoch": 0.4998496963867011, + "flos": 19755825022080.0, + "grad_norm": 4.665442504903461, + "language_loss": 0.84137499, + "learning_rate": 2.098111035105635e-06, + "loss": 0.86299658, + "num_input_tokens_seen": 89639695, + "step": 4157, + "time_per_iteration": 3.657289743423462 + }, + { + "auxiliary_loss_clip": 0.01086755, + "auxiliary_loss_mlp": 0.01025858, + "balance_loss_clip": 0.90508693, + "balance_loss_mlp": 1.01857471, + "epoch": 0.49996993927734024, + "flos": 22265728790400.0, + "grad_norm": 1.6631493944061682, + "language_loss": 0.73198557, + "learning_rate": 2.0973329938766176e-06, + "loss": 0.75311172, + "num_input_tokens_seen": 89657125, + "step": 4158, + "time_per_iteration": 2.7227413654327393 + }, + { + "auxiliary_loss_clip": 0.01161784, + "auxiliary_loss_mlp": 0.01028225, + "balance_loss_clip": 1.01734734, + "balance_loss_mlp": 1.02034545, + "epoch": 0.5000901821679793, + "flos": 23327212533120.0, + "grad_norm": 4.213186493428555, + "language_loss": 0.78899312, + "learning_rate": 2.0965549378822618e-06, + "loss": 0.81089318, + "num_input_tokens_seen": 89678415, + "step": 4159, + "time_per_iteration": 2.674312114715576 + }, + { + "auxiliary_loss_clip": 0.01013749, + "auxiliary_loss_mlp": 0.0102968, + "balance_loss_clip": 0.78437471, + "balance_loss_mlp": 1.02193189, + "epoch": 0.5002104250586185, + "flos": 20339014239360.0, + "grad_norm": 1.9283214854261497, + "language_loss": 0.8396672, + "learning_rate": 2.095776867240599e-06, + "loss": 0.86010152, + "num_input_tokens_seen": 89695405, + "step": 4160, + "time_per_iteration": 2.975090980529785 + }, + { + "auxiliary_loss_clip": 0.01105421, + "auxiliary_loss_mlp": 0.01031564, + "balance_loss_clip": 0.93905449, + "balance_loss_mlp": 1.02400959, + "epoch": 0.5003306679492575, + "flos": 13991372634240.0, + "grad_norm": 2.042125139705398, + "language_loss": 0.82766688, + "learning_rate": 2.094998782069661e-06, + "loss": 0.84903681, + "num_input_tokens_seen": 89713110, + "step": 4161, + "time_per_iteration": 2.9314119815826416 + }, + { + "auxiliary_loss_clip": 0.0117853, + "auxiliary_loss_mlp": 0.01027685, + "balance_loss_clip": 1.0553869, + "balance_loss_mlp": 1.01988292, + "epoch": 0.5004509108398966, + "flos": 27672762896640.0, + "grad_norm": 2.0208932007187297, + "language_loss": 0.75554037, + "learning_rate": 2.0942206824874845e-06, + "loss": 0.77760255, + "num_input_tokens_seen": 89735885, + "step": 4162, + "time_per_iteration": 2.6787633895874023 + }, + { + "auxiliary_loss_clip": 0.01155581, + "auxiliary_loss_mlp": 0.01033002, + "balance_loss_clip": 1.01869524, + "balance_loss_mlp": 1.02484798, + "epoch": 0.5005711537305357, + "flos": 14976186796800.0, + "grad_norm": 2.048112001581355, + "language_loss": 0.78907603, + "learning_rate": 2.093442568612105e-06, + "loss": 0.81096184, + "num_input_tokens_seen": 89753690, + "step": 4163, + "time_per_iteration": 2.5905985832214355 + }, + { + "auxiliary_loss_clip": 0.01178269, + "auxiliary_loss_mlp": 0.0103083, + "balance_loss_clip": 1.05230498, + "balance_loss_mlp": 1.02263427, + "epoch": 0.5006913966211748, + "flos": 26503259978880.0, + "grad_norm": 1.8777605194704143, + "language_loss": 0.85114902, + "learning_rate": 2.0926644405615613e-06, + "loss": 0.87323999, + "num_input_tokens_seen": 89774590, + "step": 4164, + "time_per_iteration": 2.6646194458007812 + }, + { + "auxiliary_loss_clip": 0.0110808, + "auxiliary_loss_mlp": 0.01032883, + "balance_loss_clip": 0.94117939, + "balance_loss_mlp": 1.02530742, + "epoch": 0.5008116395118138, + "flos": 20449295971200.0, + "grad_norm": 1.9877921109149848, + "language_loss": 0.81276059, + "learning_rate": 2.091886298453897e-06, + "loss": 0.83417034, + "num_input_tokens_seen": 89792775, + "step": 4165, + "time_per_iteration": 2.639660120010376 + }, + { + "auxiliary_loss_clip": 0.01153415, + "auxiliary_loss_mlp": 0.01025494, + "balance_loss_clip": 1.01520348, + "balance_loss_mlp": 1.01820469, + "epoch": 0.500931882402453, + "flos": 21579871524480.0, + "grad_norm": 1.9518629744164857, + "language_loss": 0.72576237, + "learning_rate": 2.091108142407153e-06, + "loss": 0.7475515, + "num_input_tokens_seen": 89811515, + "step": 4166, + "time_per_iteration": 2.6348743438720703 + }, + { + "auxiliary_loss_clip": 0.01044474, + "auxiliary_loss_mlp": 0.01004433, + "balance_loss_clip": 0.95572245, + "balance_loss_mlp": 1.00296116, + "epoch": 0.5010521252930921, + "flos": 57785011925760.0, + "grad_norm": 0.8326703305677671, + "language_loss": 0.62419206, + "learning_rate": 2.090329972539377e-06, + "loss": 0.64468116, + "num_input_tokens_seen": 89870080, + "step": 4167, + "time_per_iteration": 3.265245199203491 + }, + { + "auxiliary_loss_clip": 0.01002792, + "auxiliary_loss_mlp": 0.01023294, + "balance_loss_clip": 0.78209943, + "balance_loss_mlp": 1.01592135, + "epoch": 0.5011723681837311, + "flos": 18625500864000.0, + "grad_norm": 1.8591144036876057, + "language_loss": 0.68444884, + "learning_rate": 2.089551788968616e-06, + "loss": 0.70470965, + "num_input_tokens_seen": 89888045, + "step": 4168, + "time_per_iteration": 2.808824300765991 + }, + { + "auxiliary_loss_clip": 0.01080584, + "auxiliary_loss_mlp": 0.01000859, + "balance_loss_clip": 1.02270865, + "balance_loss_mlp": 0.99952406, + "epoch": 0.5012926110743702, + "flos": 55883146608000.0, + "grad_norm": 0.8315249403082364, + "language_loss": 0.60831898, + "learning_rate": 2.08877359181292e-06, + "loss": 0.6291334, + "num_input_tokens_seen": 89944610, + "step": 4169, + "time_per_iteration": 3.141533374786377 + }, + { + "auxiliary_loss_clip": 0.01113332, + "auxiliary_loss_mlp": 0.01025924, + "balance_loss_clip": 0.93714827, + "balance_loss_mlp": 1.01804376, + "epoch": 0.5014128539650093, + "flos": 24238266117120.0, + "grad_norm": 4.539129460679834, + "language_loss": 0.85737979, + "learning_rate": 2.0879953811903396e-06, + "loss": 0.87877232, + "num_input_tokens_seen": 89959495, + "step": 4170, + "time_per_iteration": 2.725149393081665 + }, + { + "auxiliary_loss_clip": 0.01155946, + "auxiliary_loss_mlp": 0.01029607, + "balance_loss_clip": 1.01904273, + "balance_loss_mlp": 1.02157795, + "epoch": 0.5015330968556484, + "flos": 27527468382720.0, + "grad_norm": 2.367059420552353, + "language_loss": 0.78542912, + "learning_rate": 2.08721715721893e-06, + "loss": 0.80728465, + "num_input_tokens_seen": 89978820, + "step": 4171, + "time_per_iteration": 2.7294485569000244 + }, + { + "auxiliary_loss_clip": 0.01153995, + "auxiliary_loss_mlp": 0.01024357, + "balance_loss_clip": 1.01517498, + "balance_loss_mlp": 1.01702595, + "epoch": 0.5016533397462875, + "flos": 23800802376960.0, + "grad_norm": 2.156117708613022, + "language_loss": 0.7716403, + "learning_rate": 2.0864389200167477e-06, + "loss": 0.79342389, + "num_input_tokens_seen": 89997075, + "step": 4172, + "time_per_iteration": 2.6853420734405518 + }, + { + "auxiliary_loss_clip": 0.01161687, + "auxiliary_loss_mlp": 0.00509582, + "balance_loss_clip": 1.01961112, + "balance_loss_mlp": 0.0, + "epoch": 0.5017735826369266, + "flos": 25295009264640.0, + "grad_norm": 1.8251858377641748, + "language_loss": 0.79121375, + "learning_rate": 2.0856606697018504e-06, + "loss": 0.80792648, + "num_input_tokens_seen": 90015085, + "step": 4173, + "time_per_iteration": 2.7245516777038574 + }, + { + "auxiliary_loss_clip": 0.01125967, + "auxiliary_loss_mlp": 0.01025279, + "balance_loss_clip": 0.97451031, + "balance_loss_mlp": 1.01682067, + "epoch": 0.5018938255275657, + "flos": 16873203778560.0, + "grad_norm": 2.378946425152112, + "language_loss": 0.73326755, + "learning_rate": 2.084882406392297e-06, + "loss": 0.75478005, + "num_input_tokens_seen": 90033045, + "step": 4174, + "time_per_iteration": 3.5195109844207764 + }, + { + "auxiliary_loss_clip": 0.0115672, + "auxiliary_loss_mlp": 0.01033732, + "balance_loss_clip": 1.01683927, + "balance_loss_mlp": 1.02640057, + "epoch": 0.5020140684182047, + "flos": 25515429073920.0, + "grad_norm": 2.259221011866847, + "language_loss": 0.71085429, + "learning_rate": 2.0841041302061496e-06, + "loss": 0.73275888, + "num_input_tokens_seen": 90052505, + "step": 4175, + "time_per_iteration": 3.812962293624878 + }, + { + "auxiliary_loss_clip": 0.01122974, + "auxiliary_loss_mlp": 0.01030739, + "balance_loss_clip": 0.97356498, + "balance_loss_mlp": 1.02336025, + "epoch": 0.5021343113088439, + "flos": 23659278791040.0, + "grad_norm": 1.8850746472476467, + "language_loss": 0.75546533, + "learning_rate": 2.083325841261473e-06, + "loss": 0.77700245, + "num_input_tokens_seen": 90071565, + "step": 4176, + "time_per_iteration": 2.7284488677978516 + }, + { + "auxiliary_loss_clip": 0.01128122, + "auxiliary_loss_mlp": 0.01024409, + "balance_loss_clip": 0.97513103, + "balance_loss_mlp": 1.01679158, + "epoch": 0.502254554199483, + "flos": 24534673148160.0, + "grad_norm": 1.8700577821127604, + "language_loss": 0.65871429, + "learning_rate": 2.0825475396763322e-06, + "loss": 0.68023956, + "num_input_tokens_seen": 90092215, + "step": 4177, + "time_per_iteration": 3.5433733463287354 + }, + { + "auxiliary_loss_clip": 0.01009513, + "auxiliary_loss_mlp": 0.01030604, + "balance_loss_clip": 0.78509957, + "balance_loss_mlp": 1.02288508, + "epoch": 0.502374797090122, + "flos": 34240285607040.0, + "grad_norm": 1.5845340448475698, + "language_loss": 0.65746045, + "learning_rate": 2.081769225568796e-06, + "loss": 0.67786163, + "num_input_tokens_seen": 90114665, + "step": 4178, + "time_per_iteration": 3.0556414127349854 + }, + { + "auxiliary_loss_clip": 0.01157779, + "auxiliary_loss_mlp": 0.01032856, + "balance_loss_clip": 1.01506102, + "balance_loss_mlp": 1.02521467, + "epoch": 0.5024950399807612, + "flos": 26031106679040.0, + "grad_norm": 1.4807619662790557, + "language_loss": 0.75928712, + "learning_rate": 2.0809908990569327e-06, + "loss": 0.78119349, + "num_input_tokens_seen": 90136445, + "step": 4179, + "time_per_iteration": 2.9010488986968994 + }, + { + "auxiliary_loss_clip": 0.01134255, + "auxiliary_loss_mlp": 0.01028744, + "balance_loss_clip": 0.97926277, + "balance_loss_mlp": 1.02088213, + "epoch": 0.5026152828714002, + "flos": 21252438120960.0, + "grad_norm": 1.7186837755021176, + "language_loss": 0.79082143, + "learning_rate": 2.0802125602588146e-06, + "loss": 0.81245142, + "num_input_tokens_seen": 90155710, + "step": 4180, + "time_per_iteration": 2.647531032562256 + }, + { + "auxiliary_loss_clip": 0.01180997, + "auxiliary_loss_mlp": 0.01029344, + "balance_loss_clip": 1.05649257, + "balance_loss_mlp": 1.02109432, + "epoch": 0.5027355257620393, + "flos": 30956111245440.0, + "grad_norm": 2.186645021216639, + "language_loss": 0.66774952, + "learning_rate": 2.0794342092925146e-06, + "loss": 0.68985295, + "num_input_tokens_seen": 90176845, + "step": 4181, + "time_per_iteration": 2.675248861312866 + }, + { + "auxiliary_loss_clip": 0.01160728, + "auxiliary_loss_mlp": 0.01024143, + "balance_loss_clip": 1.01954436, + "balance_loss_mlp": 1.01651359, + "epoch": 0.5028557686526784, + "flos": 24791147233920.0, + "grad_norm": 2.1896036139528574, + "language_loss": 0.67953372, + "learning_rate": 2.078655846276108e-06, + "loss": 0.7013824, + "num_input_tokens_seen": 90197175, + "step": 4182, + "time_per_iteration": 2.70400071144104 + }, + { + "auxiliary_loss_clip": 0.0112883, + "auxiliary_loss_mlp": 0.01025571, + "balance_loss_clip": 0.97726005, + "balance_loss_mlp": 1.01791179, + "epoch": 0.5029760115433175, + "flos": 22966992990720.0, + "grad_norm": 2.082830140033913, + "language_loss": 0.69070888, + "learning_rate": 2.0778774713276727e-06, + "loss": 0.71225291, + "num_input_tokens_seen": 90216650, + "step": 4183, + "time_per_iteration": 3.631169080734253 + }, + { + "auxiliary_loss_clip": 0.01153379, + "auxiliary_loss_mlp": 0.01024064, + "balance_loss_clip": 1.01375401, + "balance_loss_mlp": 1.01504028, + "epoch": 0.5030962544339566, + "flos": 15305164485120.0, + "grad_norm": 2.0924831404451916, + "language_loss": 0.6809063, + "learning_rate": 2.077099084565287e-06, + "loss": 0.70268077, + "num_input_tokens_seen": 90234055, + "step": 4184, + "time_per_iteration": 2.6076254844665527 + }, + { + "auxiliary_loss_clip": 0.01124807, + "auxiliary_loss_mlp": 0.01023583, + "balance_loss_clip": 0.97269714, + "balance_loss_mlp": 1.01516652, + "epoch": 0.5032164973245957, + "flos": 24494847943680.0, + "grad_norm": 1.9856179235699631, + "language_loss": 0.65248817, + "learning_rate": 2.0763206861070313e-06, + "loss": 0.67397213, + "num_input_tokens_seen": 90253115, + "step": 4185, + "time_per_iteration": 2.7262473106384277 + }, + { + "auxiliary_loss_clip": 0.01179557, + "auxiliary_loss_mlp": 0.01028833, + "balance_loss_clip": 1.05472398, + "balance_loss_mlp": 1.02075052, + "epoch": 0.5033367402152348, + "flos": 16213452721920.0, + "grad_norm": 2.0665113505742383, + "language_loss": 0.75351655, + "learning_rate": 2.0755422760709876e-06, + "loss": 0.77560043, + "num_input_tokens_seen": 90270515, + "step": 4186, + "time_per_iteration": 2.650855302810669 + }, + { + "auxiliary_loss_clip": 0.01077957, + "auxiliary_loss_mlp": 0.01024354, + "balance_loss_clip": 0.89796627, + "balance_loss_mlp": 1.01675439, + "epoch": 0.5034569831058738, + "flos": 21391375927680.0, + "grad_norm": 1.9744317366693607, + "language_loss": 0.76887095, + "learning_rate": 2.0747638545752417e-06, + "loss": 0.78989404, + "num_input_tokens_seen": 90289075, + "step": 4187, + "time_per_iteration": 2.720167636871338 + }, + { + "auxiliary_loss_clip": 0.01131794, + "auxiliary_loss_mlp": 0.01025294, + "balance_loss_clip": 0.97945064, + "balance_loss_mlp": 1.01742065, + "epoch": 0.503577225996513, + "flos": 20558751690240.0, + "grad_norm": 2.2472729805871823, + "language_loss": 0.83231205, + "learning_rate": 2.073985421737878e-06, + "loss": 0.85388291, + "num_input_tokens_seen": 90306385, + "step": 4188, + "time_per_iteration": 2.7070813179016113 + }, + { + "auxiliary_loss_clip": 0.01158366, + "auxiliary_loss_mlp": 0.01025815, + "balance_loss_clip": 1.01738477, + "balance_loss_mlp": 1.01789927, + "epoch": 0.5036974688871521, + "flos": 27229157930880.0, + "grad_norm": 2.9834176542041226, + "language_loss": 0.73893261, + "learning_rate": 2.0732069776769844e-06, + "loss": 0.76077449, + "num_input_tokens_seen": 90323795, + "step": 4189, + "time_per_iteration": 2.671497344970703 + }, + { + "auxiliary_loss_clip": 0.01180466, + "auxiliary_loss_mlp": 0.01025525, + "balance_loss_clip": 1.0561204, + "balance_loss_mlp": 1.0176934, + "epoch": 0.5038177117777911, + "flos": 20412164286720.0, + "grad_norm": 2.4071634674994207, + "language_loss": 0.73470026, + "learning_rate": 2.072428522510651e-06, + "loss": 0.75676024, + "num_input_tokens_seen": 90340360, + "step": 4190, + "time_per_iteration": 2.6294374465942383 + }, + { + "auxiliary_loss_clip": 0.01100904, + "auxiliary_loss_mlp": 0.01028971, + "balance_loss_clip": 0.93714976, + "balance_loss_mlp": 1.02171969, + "epoch": 0.5039379546684303, + "flos": 21907987286400.0, + "grad_norm": 2.101391892654678, + "language_loss": 0.76291633, + "learning_rate": 2.071650056356968e-06, + "loss": 0.78421503, + "num_input_tokens_seen": 90357900, + "step": 4191, + "time_per_iteration": 2.6639506816864014 + }, + { + "auxiliary_loss_clip": 0.01180805, + "auxiliary_loss_mlp": 0.01027723, + "balance_loss_clip": 1.0558933, + "balance_loss_mlp": 1.02051425, + "epoch": 0.5040581975590693, + "flos": 20010718909440.0, + "grad_norm": 3.6800967097342427, + "language_loss": 0.79898304, + "learning_rate": 2.070871579334028e-06, + "loss": 0.82106829, + "num_input_tokens_seen": 90377010, + "step": 4192, + "time_per_iteration": 2.631546974182129 + }, + { + "auxiliary_loss_clip": 0.01177956, + "auxiliary_loss_mlp": 0.01023621, + "balance_loss_clip": 1.0538882, + "balance_loss_mlp": 1.01614332, + "epoch": 0.5041784404497084, + "flos": 20959837931520.0, + "grad_norm": 1.6149268784573871, + "language_loss": 0.71927655, + "learning_rate": 2.0700930915599264e-06, + "loss": 0.7412923, + "num_input_tokens_seen": 90396740, + "step": 4193, + "time_per_iteration": 2.6058406829833984 + }, + { + "auxiliary_loss_clip": 0.01177571, + "auxiliary_loss_mlp": 0.01026419, + "balance_loss_clip": 1.05542827, + "balance_loss_mlp": 1.01943338, + "epoch": 0.5042986833403476, + "flos": 12495082757760.0, + "grad_norm": 2.1702139621144445, + "language_loss": 0.78614557, + "learning_rate": 2.0693145931527583e-06, + "loss": 0.80818546, + "num_input_tokens_seen": 90413220, + "step": 4194, + "time_per_iteration": 2.6419055461883545 + }, + { + "auxiliary_loss_clip": 0.01131493, + "auxiliary_loss_mlp": 0.0103092, + "balance_loss_clip": 0.98133731, + "balance_loss_mlp": 1.02344573, + "epoch": 0.5044189262309866, + "flos": 29202305788800.0, + "grad_norm": 1.548825013352934, + "language_loss": 0.77821398, + "learning_rate": 2.068536084230622e-06, + "loss": 0.79983807, + "num_input_tokens_seen": 90435085, + "step": 4195, + "time_per_iteration": 2.7195072174072266 + }, + { + "auxiliary_loss_clip": 0.01152351, + "auxiliary_loss_mlp": 0.01025799, + "balance_loss_clip": 1.0148015, + "balance_loss_mlp": 1.01747835, + "epoch": 0.5045391691216257, + "flos": 23873198238720.0, + "grad_norm": 2.5411186195064275, + "language_loss": 0.88732529, + "learning_rate": 2.067757564911616e-06, + "loss": 0.90910679, + "num_input_tokens_seen": 90453660, + "step": 4196, + "time_per_iteration": 2.690484046936035 + }, + { + "auxiliary_loss_clip": 0.01139811, + "auxiliary_loss_mlp": 0.00508982, + "balance_loss_clip": 0.97870791, + "balance_loss_mlp": 0.0, + "epoch": 0.5046594120122648, + "flos": 24644990793600.0, + "grad_norm": 1.8184083554878692, + "language_loss": 0.92755342, + "learning_rate": 2.0669790353138407e-06, + "loss": 0.94404131, + "num_input_tokens_seen": 90472625, + "step": 4197, + "time_per_iteration": 2.6899890899658203 + }, + { + "auxiliary_loss_clip": 0.01104709, + "auxiliary_loss_mlp": 0.00509432, + "balance_loss_clip": 0.94258416, + "balance_loss_mlp": 0.0, + "epoch": 0.5047796549029039, + "flos": 23362835846400.0, + "grad_norm": 2.5925642067866796, + "language_loss": 0.73297322, + "learning_rate": 2.0662004955553995e-06, + "loss": 0.74911463, + "num_input_tokens_seen": 90492325, + "step": 4198, + "time_per_iteration": 2.737924337387085 + }, + { + "auxiliary_loss_clip": 0.01126169, + "auxiliary_loss_mlp": 0.01025275, + "balance_loss_clip": 0.97529602, + "balance_loss_mlp": 1.01852214, + "epoch": 0.5048998977935429, + "flos": 17304095329920.0, + "grad_norm": 1.8684398904450588, + "language_loss": 0.76905334, + "learning_rate": 2.065421945754395e-06, + "loss": 0.79056776, + "num_input_tokens_seen": 90510055, + "step": 4199, + "time_per_iteration": 2.6928324699401855 + }, + { + "auxiliary_loss_clip": 0.01084483, + "auxiliary_loss_mlp": 0.01022014, + "balance_loss_clip": 0.90063578, + "balance_loss_mlp": 1.01454568, + "epoch": 0.505020140684182, + "flos": 34856979235200.0, + "grad_norm": 1.6218260275733876, + "language_loss": 0.77951545, + "learning_rate": 2.0646433860289344e-06, + "loss": 0.80058038, + "num_input_tokens_seen": 90528980, + "step": 4200, + "time_per_iteration": 3.943054437637329 + }, + { + "auxiliary_loss_clip": 0.01158266, + "auxiliary_loss_mlp": 0.00509641, + "balance_loss_clip": 1.0153017, + "balance_loss_mlp": 0.0, + "epoch": 0.5051403835748212, + "flos": 24863974058880.0, + "grad_norm": 1.919129554059446, + "language_loss": 0.82960862, + "learning_rate": 2.0638648164971233e-06, + "loss": 0.84628767, + "num_input_tokens_seen": 90547445, + "step": 4201, + "time_per_iteration": 3.59244441986084 + }, + { + "auxiliary_loss_clip": 0.01132772, + "auxiliary_loss_mlp": 0.01025626, + "balance_loss_clip": 0.98031831, + "balance_loss_mlp": 1.01831269, + "epoch": 0.5052606264654602, + "flos": 20959694277120.0, + "grad_norm": 1.935517913547723, + "language_loss": 0.88584971, + "learning_rate": 2.06308623727707e-06, + "loss": 0.90743363, + "num_input_tokens_seen": 90567545, + "step": 4202, + "time_per_iteration": 2.702523708343506 + }, + { + "auxiliary_loss_clip": 0.01156691, + "auxiliary_loss_mlp": 0.01024002, + "balance_loss_clip": 1.01926208, + "balance_loss_mlp": 1.01563656, + "epoch": 0.5053808693560993, + "flos": 19642382893440.0, + "grad_norm": 4.444189131426065, + "language_loss": 0.76337707, + "learning_rate": 2.0623076484868846e-06, + "loss": 0.78518403, + "num_input_tokens_seen": 90585000, + "step": 4203, + "time_per_iteration": 3.631798505783081 + }, + { + "auxiliary_loss_clip": 0.01039193, + "auxiliary_loss_mlp": 0.01002271, + "balance_loss_clip": 0.95092463, + "balance_loss_mlp": 1.00094211, + "epoch": 0.5055011122467384, + "flos": 67504915019520.0, + "grad_norm": 0.8302097689562908, + "language_loss": 0.60646021, + "learning_rate": 2.061529050244679e-06, + "loss": 0.62687492, + "num_input_tokens_seen": 90644745, + "step": 4204, + "time_per_iteration": 3.2348270416259766 + }, + { + "auxiliary_loss_clip": 0.01114326, + "auxiliary_loss_mlp": 0.01025853, + "balance_loss_clip": 0.93993998, + "balance_loss_mlp": 1.0184679, + "epoch": 0.5056213551373775, + "flos": 16872952383360.0, + "grad_norm": 1.8958015240611656, + "language_loss": 0.74538338, + "learning_rate": 2.060750442668565e-06, + "loss": 0.76678526, + "num_input_tokens_seen": 90662500, + "step": 4205, + "time_per_iteration": 2.6602063179016113 + }, + { + "auxiliary_loss_clip": 0.01158336, + "auxiliary_loss_mlp": 0.01027967, + "balance_loss_clip": 1.01977491, + "balance_loss_mlp": 1.02037299, + "epoch": 0.5057415980280165, + "flos": 15334179696000.0, + "grad_norm": 2.5982316751316925, + "language_loss": 0.63585526, + "learning_rate": 2.059971825876657e-06, + "loss": 0.6577183, + "num_input_tokens_seen": 90677010, + "step": 4206, + "time_per_iteration": 2.611464023590088 + }, + { + "auxiliary_loss_clip": 0.01157444, + "auxiliary_loss_mlp": 0.01025869, + "balance_loss_clip": 1.01721346, + "balance_loss_mlp": 1.01841283, + "epoch": 0.5058618409186557, + "flos": 19025976574080.0, + "grad_norm": 2.009501955416511, + "language_loss": 0.76386178, + "learning_rate": 2.0591931999870713e-06, + "loss": 0.78569496, + "num_input_tokens_seen": 90695935, + "step": 4207, + "time_per_iteration": 2.6256418228149414 + }, + { + "auxiliary_loss_clip": 0.01056197, + "auxiliary_loss_mlp": 0.01002591, + "balance_loss_clip": 0.98612416, + "balance_loss_mlp": 1.00127399, + "epoch": 0.5059820838092948, + "flos": 63453114080640.0, + "grad_norm": 0.8358426964932164, + "language_loss": 0.57590663, + "learning_rate": 2.0584145651179234e-06, + "loss": 0.59649444, + "num_input_tokens_seen": 90751645, + "step": 4208, + "time_per_iteration": 3.2315430641174316 + }, + { + "auxiliary_loss_clip": 0.01130908, + "auxiliary_loss_mlp": 0.00509192, + "balance_loss_clip": 0.97937393, + "balance_loss_mlp": 0.0, + "epoch": 0.5061023266999338, + "flos": 15441803821440.0, + "grad_norm": 2.4616324813021913, + "language_loss": 0.79823905, + "learning_rate": 2.0576359213873327e-06, + "loss": 0.81464005, + "num_input_tokens_seen": 90766795, + "step": 4209, + "time_per_iteration": 3.4913032054901123 + }, + { + "auxiliary_loss_clip": 0.01137002, + "auxiliary_loss_mlp": 0.0102774, + "balance_loss_clip": 0.97497559, + "balance_loss_mlp": 1.0199641, + "epoch": 0.506222569590573, + "flos": 22451063990400.0, + "grad_norm": 23.925376226466877, + "language_loss": 0.7010715, + "learning_rate": 2.056857268913419e-06, + "loss": 0.72271889, + "num_input_tokens_seen": 90786845, + "step": 4210, + "time_per_iteration": 2.680917501449585 + }, + { + "auxiliary_loss_clip": 0.0115469, + "auxiliary_loss_mlp": 0.01024087, + "balance_loss_clip": 1.01699877, + "balance_loss_mlp": 1.01659477, + "epoch": 0.506342812481212, + "flos": 17558665994880.0, + "grad_norm": 2.2494051590881736, + "language_loss": 0.84232455, + "learning_rate": 2.056078607814303e-06, + "loss": 0.86411232, + "num_input_tokens_seen": 90802630, + "step": 4211, + "time_per_iteration": 2.7519850730895996 + }, + { + "auxiliary_loss_clip": 0.01154447, + "auxiliary_loss_mlp": 0.01027855, + "balance_loss_clip": 1.01720428, + "balance_loss_mlp": 1.02010894, + "epoch": 0.5064630553718511, + "flos": 23402050519680.0, + "grad_norm": 2.1126645235863237, + "language_loss": 0.78265584, + "learning_rate": 2.055299938208106e-06, + "loss": 0.80447888, + "num_input_tokens_seen": 90823620, + "step": 4212, + "time_per_iteration": 2.6443185806274414 + }, + { + "auxiliary_loss_clip": 0.0115877, + "auxiliary_loss_mlp": 0.01023856, + "balance_loss_clip": 1.01771951, + "balance_loss_mlp": 1.01585102, + "epoch": 0.5065832982624903, + "flos": 23987035416960.0, + "grad_norm": 2.4896725226616807, + "language_loss": 0.86299008, + "learning_rate": 2.0545212602129526e-06, + "loss": 0.88481635, + "num_input_tokens_seen": 90843475, + "step": 4213, + "time_per_iteration": 2.6947007179260254 + }, + { + "auxiliary_loss_clip": 0.01127179, + "auxiliary_loss_mlp": 0.01029621, + "balance_loss_clip": 0.97757685, + "balance_loss_mlp": 1.02113962, + "epoch": 0.5067035411531293, + "flos": 21503058289920.0, + "grad_norm": 2.235511851781552, + "language_loss": 0.6671322, + "learning_rate": 2.0537425739469673e-06, + "loss": 0.6887002, + "num_input_tokens_seen": 90862410, + "step": 4214, + "time_per_iteration": 2.6961722373962402 + }, + { + "auxiliary_loss_clip": 0.0106227, + "auxiliary_loss_mlp": 0.01003526, + "balance_loss_clip": 0.98681962, + "balance_loss_mlp": 1.00218475, + "epoch": 0.5068237840437684, + "flos": 65934397687680.0, + "grad_norm": 0.8393063122443853, + "language_loss": 0.59487879, + "learning_rate": 2.052963879528276e-06, + "loss": 0.61553675, + "num_input_tokens_seen": 90922280, + "step": 4215, + "time_per_iteration": 3.2220113277435303 + }, + { + "auxiliary_loss_clip": 0.01155858, + "auxiliary_loss_mlp": 0.01026258, + "balance_loss_clip": 1.01648092, + "balance_loss_mlp": 1.01859021, + "epoch": 0.5069440269344075, + "flos": 27264206626560.0, + "grad_norm": 3.4994482964974494, + "language_loss": 0.76304787, + "learning_rate": 2.052185177075007e-06, + "loss": 0.78486902, + "num_input_tokens_seen": 90941850, + "step": 4216, + "time_per_iteration": 2.7012040615081787 + }, + { + "auxiliary_loss_clip": 0.01154997, + "auxiliary_loss_mlp": 0.01029212, + "balance_loss_clip": 1.01447129, + "balance_loss_mlp": 1.02095127, + "epoch": 0.5070642698250466, + "flos": 23366319465600.0, + "grad_norm": 2.033706476684059, + "language_loss": 0.82932001, + "learning_rate": 2.051406466705288e-06, + "loss": 0.85116208, + "num_input_tokens_seen": 90961390, + "step": 4217, + "time_per_iteration": 2.6662983894348145 + }, + { + "auxiliary_loss_clip": 0.01177918, + "auxiliary_loss_mlp": 0.01031271, + "balance_loss_clip": 1.05398607, + "balance_loss_mlp": 1.02385592, + "epoch": 0.5071845127156857, + "flos": 20340127560960.0, + "grad_norm": 1.9848124617691545, + "language_loss": 0.81067324, + "learning_rate": 2.0506277485372486e-06, + "loss": 0.8327651, + "num_input_tokens_seen": 90980215, + "step": 4218, + "time_per_iteration": 2.595817804336548 + }, + { + "auxiliary_loss_clip": 0.01148981, + "auxiliary_loss_mlp": 0.01032596, + "balance_loss_clip": 1.0138222, + "balance_loss_mlp": 1.02481759, + "epoch": 0.5073047556063248, + "flos": 12092955022080.0, + "grad_norm": 1.832970918948848, + "language_loss": 0.67460537, + "learning_rate": 2.04984902268902e-06, + "loss": 0.69642109, + "num_input_tokens_seen": 90997415, + "step": 4219, + "time_per_iteration": 2.576429843902588 + }, + { + "auxiliary_loss_clip": 0.01163577, + "auxiliary_loss_mlp": 0.01028253, + "balance_loss_clip": 1.01690841, + "balance_loss_mlp": 1.01996779, + "epoch": 0.5074249984969639, + "flos": 19682854542720.0, + "grad_norm": 7.371384031602809, + "language_loss": 0.75800765, + "learning_rate": 2.0490702892787345e-06, + "loss": 0.77992594, + "num_input_tokens_seen": 91016475, + "step": 4220, + "time_per_iteration": 2.656843423843384 + }, + { + "auxiliary_loss_clip": 0.01145058, + "auxiliary_loss_mlp": 0.01023525, + "balance_loss_clip": 1.0124712, + "balance_loss_mlp": 1.01660156, + "epoch": 0.5075452413876029, + "flos": 28765703975040.0, + "grad_norm": 1.9179278516025045, + "language_loss": 0.62005919, + "learning_rate": 2.0482915484245246e-06, + "loss": 0.64174503, + "num_input_tokens_seen": 91038095, + "step": 4221, + "time_per_iteration": 2.8412158489227295 + }, + { + "auxiliary_loss_clip": 0.01081923, + "auxiliary_loss_mlp": 0.01031924, + "balance_loss_clip": 0.90136576, + "balance_loss_mlp": 1.02370453, + "epoch": 0.5076654842782421, + "flos": 20339445202560.0, + "grad_norm": 7.635918218517697, + "language_loss": 0.84377873, + "learning_rate": 2.047512800244526e-06, + "loss": 0.86491728, + "num_input_tokens_seen": 91053360, + "step": 4222, + "time_per_iteration": 2.782432794570923 + }, + { + "auxiliary_loss_clip": 0.01155314, + "auxiliary_loss_mlp": 0.01035772, + "balance_loss_clip": 1.01662958, + "balance_loss_mlp": 1.02757037, + "epoch": 0.5077857271688812, + "flos": 26359653404160.0, + "grad_norm": 2.6658743215926046, + "language_loss": 0.790676, + "learning_rate": 2.046734044856873e-06, + "loss": 0.8125869, + "num_input_tokens_seen": 91072770, + "step": 4223, + "time_per_iteration": 2.6428394317626953 + }, + { + "auxiliary_loss_clip": 0.01152621, + "auxiliary_loss_mlp": 0.01030022, + "balance_loss_clip": 1.01561964, + "balance_loss_mlp": 1.0228281, + "epoch": 0.5079059700595202, + "flos": 21798962530560.0, + "grad_norm": 2.086827147973825, + "language_loss": 0.81442565, + "learning_rate": 2.045955282379702e-06, + "loss": 0.83625209, + "num_input_tokens_seen": 91091430, + "step": 4224, + "time_per_iteration": 2.63199520111084 + }, + { + "auxiliary_loss_clip": 0.01151321, + "auxiliary_loss_mlp": 0.01031757, + "balance_loss_clip": 1.01353729, + "balance_loss_mlp": 1.02379084, + "epoch": 0.5080262129501594, + "flos": 13187943175680.0, + "grad_norm": 3.4534562367269577, + "language_loss": 0.75528777, + "learning_rate": 2.045176512931152e-06, + "loss": 0.77711856, + "num_input_tokens_seen": 91106060, + "step": 4225, + "time_per_iteration": 2.6122629642486572 + }, + { + "auxiliary_loss_clip": 0.01111798, + "auxiliary_loss_mlp": 0.01025427, + "balance_loss_clip": 0.93955392, + "balance_loss_mlp": 1.0177139, + "epoch": 0.5081464558407984, + "flos": 25301473712640.0, + "grad_norm": 1.8616737230058555, + "language_loss": 0.75976145, + "learning_rate": 2.0443977366293604e-06, + "loss": 0.78113377, + "num_input_tokens_seen": 91124100, + "step": 4226, + "time_per_iteration": 3.630542278289795 + }, + { + "auxiliary_loss_clip": 0.01060371, + "auxiliary_loss_mlp": 0.01027369, + "balance_loss_clip": 0.86039078, + "balance_loss_mlp": 1.01921535, + "epoch": 0.5082666987314375, + "flos": 30951226995840.0, + "grad_norm": 1.68656886343442, + "language_loss": 0.76741028, + "learning_rate": 2.043618953592468e-06, + "loss": 0.7882877, + "num_input_tokens_seen": 91146555, + "step": 4227, + "time_per_iteration": 3.775533437728882 + }, + { + "auxiliary_loss_clip": 0.01132793, + "auxiliary_loss_mlp": 0.01031857, + "balance_loss_clip": 0.98005438, + "balance_loss_mlp": 1.02403128, + "epoch": 0.5083869416220766, + "flos": 19682495406720.0, + "grad_norm": 1.6206455039189454, + "language_loss": 0.80983925, + "learning_rate": 2.0428401639386144e-06, + "loss": 0.83148575, + "num_input_tokens_seen": 91167120, + "step": 4228, + "time_per_iteration": 2.7048511505126953 + }, + { + "auxiliary_loss_clip": 0.01034047, + "auxiliary_loss_mlp": 0.01000137, + "balance_loss_clip": 0.94615042, + "balance_loss_mlp": 0.99886149, + "epoch": 0.5085071845127157, + "flos": 71817535589760.0, + "grad_norm": 0.8240794926103661, + "language_loss": 0.58112842, + "learning_rate": 2.042061367785943e-06, + "loss": 0.60147029, + "num_input_tokens_seen": 91220260, + "step": 4229, + "time_per_iteration": 4.1328582763671875 + }, + { + "auxiliary_loss_clip": 0.0110778, + "auxiliary_loss_mlp": 0.0102447, + "balance_loss_clip": 0.93574834, + "balance_loss_mlp": 1.01660872, + "epoch": 0.5086274274033548, + "flos": 35951608252800.0, + "grad_norm": 2.3467460499649024, + "language_loss": 0.75162476, + "learning_rate": 2.041282565252594e-06, + "loss": 0.77294719, + "num_input_tokens_seen": 91240425, + "step": 4230, + "time_per_iteration": 2.800226926803589 + }, + { + "auxiliary_loss_clip": 0.0110452, + "auxiliary_loss_mlp": 0.01023857, + "balance_loss_clip": 0.93528247, + "balance_loss_mlp": 1.01621878, + "epoch": 0.5087476702939938, + "flos": 23513732881920.0, + "grad_norm": 4.417442685007326, + "language_loss": 0.7694003, + "learning_rate": 2.040503756456714e-06, + "loss": 0.7906841, + "num_input_tokens_seen": 91259635, + "step": 4231, + "time_per_iteration": 2.7178597450256348 + }, + { + "auxiliary_loss_clip": 0.01148441, + "auxiliary_loss_mlp": 0.01025563, + "balance_loss_clip": 1.01337051, + "balance_loss_mlp": 1.01820779, + "epoch": 0.508867913184633, + "flos": 15122091841920.0, + "grad_norm": 1.9423172399615793, + "language_loss": 0.78872615, + "learning_rate": 2.0397249415164456e-06, + "loss": 0.81046617, + "num_input_tokens_seen": 91276990, + "step": 4232, + "time_per_iteration": 2.592738628387451 + }, + { + "auxiliary_loss_clip": 0.01123056, + "auxiliary_loss_mlp": 0.01029709, + "balance_loss_clip": 0.9728508, + "balance_loss_mlp": 1.02118587, + "epoch": 0.508988156075272, + "flos": 25885309374720.0, + "grad_norm": 1.6027614799317598, + "language_loss": 0.80362946, + "learning_rate": 2.0389461205499354e-06, + "loss": 0.82515711, + "num_input_tokens_seen": 91296125, + "step": 4233, + "time_per_iteration": 2.6886048316955566 + }, + { + "auxiliary_loss_clip": 0.01107228, + "auxiliary_loss_mlp": 0.01025128, + "balance_loss_clip": 0.93692642, + "balance_loss_mlp": 1.01834464, + "epoch": 0.5091083989659111, + "flos": 13844857057920.0, + "grad_norm": 2.0993871983294063, + "language_loss": 0.73329753, + "learning_rate": 2.03816729367533e-06, + "loss": 0.75462109, + "num_input_tokens_seen": 91314280, + "step": 4234, + "time_per_iteration": 2.6645374298095703 + }, + { + "auxiliary_loss_clip": 0.01136668, + "auxiliary_loss_mlp": 0.01031044, + "balance_loss_clip": 0.98182142, + "balance_loss_mlp": 1.02314067, + "epoch": 0.5092286418565503, + "flos": 21104881050240.0, + "grad_norm": 4.809076703651084, + "language_loss": 0.71779358, + "learning_rate": 2.0373884610107765e-06, + "loss": 0.73947072, + "num_input_tokens_seen": 91334595, + "step": 4235, + "time_per_iteration": 2.729299783706665 + }, + { + "auxiliary_loss_clip": 0.01154694, + "auxiliary_loss_mlp": 0.01024972, + "balance_loss_clip": 1.01235485, + "balance_loss_mlp": 1.0169847, + "epoch": 0.5093488847471893, + "flos": 18621298972800.0, + "grad_norm": 2.903664334903831, + "language_loss": 0.6932472, + "learning_rate": 2.0366096226744225e-06, + "loss": 0.71504378, + "num_input_tokens_seen": 91349790, + "step": 4236, + "time_per_iteration": 3.5996031761169434 + }, + { + "auxiliary_loss_clip": 0.01147749, + "auxiliary_loss_mlp": 0.01023242, + "balance_loss_clip": 1.01442206, + "balance_loss_mlp": 1.01527262, + "epoch": 0.5094691276378284, + "flos": 23803783205760.0, + "grad_norm": 1.6918728433915917, + "language_loss": 0.76894724, + "learning_rate": 2.035830778784418e-06, + "loss": 0.79065716, + "num_input_tokens_seen": 91370465, + "step": 4237, + "time_per_iteration": 2.704472780227661 + }, + { + "auxiliary_loss_clip": 0.01133726, + "auxiliary_loss_mlp": 0.01030639, + "balance_loss_clip": 0.97804523, + "balance_loss_mlp": 1.02291453, + "epoch": 0.5095893705284675, + "flos": 17420410546560.0, + "grad_norm": 3.6113438830970193, + "language_loss": 0.80363369, + "learning_rate": 2.0350519294589134e-06, + "loss": 0.82527733, + "num_input_tokens_seen": 91388505, + "step": 4238, + "time_per_iteration": 2.6679728031158447 + }, + { + "auxiliary_loss_clip": 0.01081576, + "auxiliary_loss_mlp": 0.01024527, + "balance_loss_clip": 0.89842117, + "balance_loss_mlp": 1.01595032, + "epoch": 0.5097096134191066, + "flos": 25849362839040.0, + "grad_norm": 1.7734911832035791, + "language_loss": 0.8288607, + "learning_rate": 2.0342730748160588e-06, + "loss": 0.8499217, + "num_input_tokens_seen": 91408970, + "step": 4239, + "time_per_iteration": 2.7655038833618164 + }, + { + "auxiliary_loss_clip": 0.01130043, + "auxiliary_loss_mlp": 0.01024329, + "balance_loss_clip": 0.97528058, + "balance_loss_mlp": 1.01569867, + "epoch": 0.5098298563097456, + "flos": 27745122844800.0, + "grad_norm": 2.1549657828130773, + "language_loss": 0.70325416, + "learning_rate": 2.033494214974006e-06, + "loss": 0.72479796, + "num_input_tokens_seen": 91430115, + "step": 4240, + "time_per_iteration": 2.711535692214966 + }, + { + "auxiliary_loss_clip": 0.01123398, + "auxiliary_loss_mlp": 0.01028066, + "balance_loss_clip": 0.97587669, + "balance_loss_mlp": 1.02094364, + "epoch": 0.5099500992003848, + "flos": 21358913011200.0, + "grad_norm": 1.7781909707842563, + "language_loss": 0.83987927, + "learning_rate": 2.0327153500509067e-06, + "loss": 0.86139393, + "num_input_tokens_seen": 91449140, + "step": 4241, + "time_per_iteration": 2.704472541809082 + }, + { + "auxiliary_loss_clip": 0.01132942, + "auxiliary_loss_mlp": 0.01029057, + "balance_loss_clip": 0.97875923, + "balance_loss_mlp": 1.02120709, + "epoch": 0.5100703420910239, + "flos": 19865999013120.0, + "grad_norm": 1.93903795639711, + "language_loss": 0.84708345, + "learning_rate": 2.031936480164916e-06, + "loss": 0.86870348, + "num_input_tokens_seen": 91466880, + "step": 4242, + "time_per_iteration": 2.732736587524414 + }, + { + "auxiliary_loss_clip": 0.01133541, + "auxiliary_loss_mlp": 0.01031402, + "balance_loss_clip": 0.98304939, + "balance_loss_mlp": 1.02442825, + "epoch": 0.5101905849816629, + "flos": 24648797635200.0, + "grad_norm": 3.8534092967868188, + "language_loss": 0.80036092, + "learning_rate": 2.0311576054341857e-06, + "loss": 0.82201028, + "num_input_tokens_seen": 91487495, + "step": 4243, + "time_per_iteration": 2.7144715785980225 + }, + { + "auxiliary_loss_clip": 0.0118149, + "auxiliary_loss_mlp": 0.01026449, + "balance_loss_clip": 1.05649042, + "balance_loss_mlp": 1.01909351, + "epoch": 0.5103108278723021, + "flos": 22930076787840.0, + "grad_norm": 1.9093367034484658, + "language_loss": 0.62656677, + "learning_rate": 2.0303787259768715e-06, + "loss": 0.64864612, + "num_input_tokens_seen": 91508395, + "step": 4244, + "time_per_iteration": 2.698291301727295 + }, + { + "auxiliary_loss_clip": 0.01133599, + "auxiliary_loss_mlp": 0.01022689, + "balance_loss_clip": 0.9797622, + "balance_loss_mlp": 1.01528299, + "epoch": 0.5104310707629411, + "flos": 21506613736320.0, + "grad_norm": 2.484329129605914, + "language_loss": 0.68783027, + "learning_rate": 2.0295998419111294e-06, + "loss": 0.70939314, + "num_input_tokens_seen": 91525685, + "step": 4245, + "time_per_iteration": 2.6807327270507812 + }, + { + "auxiliary_loss_clip": 0.01063059, + "auxiliary_loss_mlp": 0.01026, + "balance_loss_clip": 0.86093223, + "balance_loss_mlp": 1.01824546, + "epoch": 0.5105513136535802, + "flos": 14903180403840.0, + "grad_norm": 2.378450475525515, + "language_loss": 0.7407847, + "learning_rate": 2.028820953355115e-06, + "loss": 0.7616753, + "num_input_tokens_seen": 91543785, + "step": 4246, + "time_per_iteration": 2.7779159545898438 + }, + { + "auxiliary_loss_clip": 0.01140251, + "auxiliary_loss_mlp": 0.0103311, + "balance_loss_clip": 0.97778338, + "balance_loss_mlp": 1.02521276, + "epoch": 0.5106715565442194, + "flos": 22602212421120.0, + "grad_norm": 9.66947822064828, + "language_loss": 0.78729129, + "learning_rate": 2.0280420604269834e-06, + "loss": 0.80902493, + "num_input_tokens_seen": 91563325, + "step": 4247, + "time_per_iteration": 2.659196376800537 + }, + { + "auxiliary_loss_clip": 0.01055864, + "auxiliary_loss_mlp": 0.01008845, + "balance_loss_clip": 0.98211592, + "balance_loss_mlp": 1.00759912, + "epoch": 0.5107917994348584, + "flos": 71027645558400.0, + "grad_norm": 0.7055669204158083, + "language_loss": 0.58927, + "learning_rate": 2.027263163244895e-06, + "loss": 0.6099171, + "num_input_tokens_seen": 91632450, + "step": 4248, + "time_per_iteration": 3.370739698410034 + }, + { + "auxiliary_loss_clip": 0.01155187, + "auxiliary_loss_mlp": 0.01026756, + "balance_loss_clip": 1.01787424, + "balance_loss_mlp": 1.01942718, + "epoch": 0.5109120423254975, + "flos": 24827416992000.0, + "grad_norm": 1.9469809101843738, + "language_loss": 0.74604714, + "learning_rate": 2.026484261927005e-06, + "loss": 0.76786661, + "num_input_tokens_seen": 91651945, + "step": 4249, + "time_per_iteration": 2.717055559158325 + }, + { + "auxiliary_loss_clip": 0.01167826, + "auxiliary_loss_mlp": 0.01029002, + "balance_loss_clip": 1.02235794, + "balance_loss_mlp": 1.02143788, + "epoch": 0.5110322852161366, + "flos": 21247661612160.0, + "grad_norm": 2.5648501539155246, + "language_loss": 0.74050927, + "learning_rate": 2.025705356591475e-06, + "loss": 0.76247758, + "num_input_tokens_seen": 91669635, + "step": 4250, + "time_per_iteration": 2.6491618156433105 + }, + { + "auxiliary_loss_clip": 0.01013858, + "auxiliary_loss_mlp": 0.00500539, + "balance_loss_clip": 0.90657413, + "balance_loss_mlp": 0.0, + "epoch": 0.5111525281067757, + "flos": 66457114358400.0, + "grad_norm": 0.7597231397540688, + "language_loss": 0.57969797, + "learning_rate": 2.024926447356462e-06, + "loss": 0.59484196, + "num_input_tokens_seen": 91731920, + "step": 4251, + "time_per_iteration": 4.112889528274536 + }, + { + "auxiliary_loss_clip": 0.01159433, + "auxiliary_loss_mlp": 0.01027924, + "balance_loss_clip": 1.01941609, + "balance_loss_mlp": 1.01966882, + "epoch": 0.5112727709974147, + "flos": 14866731077760.0, + "grad_norm": 2.4544672737378384, + "language_loss": 0.78493482, + "learning_rate": 2.024147534340127e-06, + "loss": 0.80680841, + "num_input_tokens_seen": 91749780, + "step": 4252, + "time_per_iteration": 2.627737045288086 + }, + { + "auxiliary_loss_clip": 0.01129039, + "auxiliary_loss_mlp": 0.01029131, + "balance_loss_clip": 0.9745611, + "balance_loss_mlp": 1.02148926, + "epoch": 0.5113930138880539, + "flos": 21177600134400.0, + "grad_norm": 2.559314848764683, + "language_loss": 0.80047929, + "learning_rate": 2.02336861766063e-06, + "loss": 0.82206094, + "num_input_tokens_seen": 91768840, + "step": 4253, + "time_per_iteration": 3.6011464595794678 + }, + { + "auxiliary_loss_clip": 0.01163254, + "auxiliary_loss_mlp": 0.01033869, + "balance_loss_clip": 1.01748896, + "balance_loss_mlp": 1.02488661, + "epoch": 0.511513256778693, + "flos": 20409111630720.0, + "grad_norm": 2.2053667213822896, + "language_loss": 0.79124641, + "learning_rate": 2.0225896974361327e-06, + "loss": 0.81321764, + "num_input_tokens_seen": 91788945, + "step": 4254, + "time_per_iteration": 2.6799635887145996 + }, + { + "auxiliary_loss_clip": 0.01014624, + "auxiliary_loss_mlp": 0.01004135, + "balance_loss_clip": 0.90628582, + "balance_loss_mlp": 1.0026989, + "epoch": 0.511633499669332, + "flos": 69879975131520.0, + "grad_norm": 0.8572790487455019, + "language_loss": 0.59975976, + "learning_rate": 2.0218107737847962e-06, + "loss": 0.61994737, + "num_input_tokens_seen": 91850990, + "step": 4255, + "time_per_iteration": 4.2045159339904785 + }, + { + "auxiliary_loss_clip": 0.01180123, + "auxiliary_loss_mlp": 0.0102766, + "balance_loss_clip": 1.05617905, + "balance_loss_mlp": 1.01967323, + "epoch": 0.5117537425599712, + "flos": 24097855852800.0, + "grad_norm": 1.875314896194411, + "language_loss": 0.74528575, + "learning_rate": 2.0210318468247826e-06, + "loss": 0.76736355, + "num_input_tokens_seen": 91869960, + "step": 4256, + "time_per_iteration": 2.651378870010376 + }, + { + "auxiliary_loss_clip": 0.01132737, + "auxiliary_loss_mlp": 0.0102573, + "balance_loss_clip": 0.97724587, + "balance_loss_mlp": 1.01836252, + "epoch": 0.5118739854506102, + "flos": 20959550622720.0, + "grad_norm": 2.434369240021604, + "language_loss": 0.82141566, + "learning_rate": 2.020252916674255e-06, + "loss": 0.84300029, + "num_input_tokens_seen": 91889075, + "step": 4257, + "time_per_iteration": 2.709266185760498 + }, + { + "auxiliary_loss_clip": 0.01156808, + "auxiliary_loss_mlp": 0.01032777, + "balance_loss_clip": 1.0144999, + "balance_loss_mlp": 1.02477837, + "epoch": 0.5119942283412493, + "flos": 17457326749440.0, + "grad_norm": 2.0547568301608283, + "language_loss": 0.80787432, + "learning_rate": 2.019473983451375e-06, + "loss": 0.82977015, + "num_input_tokens_seen": 91907495, + "step": 4258, + "time_per_iteration": 2.680152177810669 + }, + { + "auxiliary_loss_clip": 0.01118791, + "auxiliary_loss_mlp": 0.01025192, + "balance_loss_clip": 0.9406389, + "balance_loss_mlp": 1.01718426, + "epoch": 0.5121144712318885, + "flos": 21066743784960.0, + "grad_norm": 1.7789352023616904, + "language_loss": 0.71441245, + "learning_rate": 2.0186950472743076e-06, + "loss": 0.73585224, + "num_input_tokens_seen": 91927400, + "step": 4259, + "time_per_iteration": 2.7107503414154053 + }, + { + "auxiliary_loss_clip": 0.01178327, + "auxiliary_loss_mlp": 0.01023009, + "balance_loss_clip": 1.05355859, + "balance_loss_mlp": 1.01526356, + "epoch": 0.5122347141225275, + "flos": 19860791541120.0, + "grad_norm": 1.9292712948322976, + "language_loss": 0.74332017, + "learning_rate": 2.0179161082612162e-06, + "loss": 0.76533353, + "num_input_tokens_seen": 91946790, + "step": 4260, + "time_per_iteration": 2.584010362625122 + }, + { + "auxiliary_loss_clip": 0.01126819, + "auxiliary_loss_mlp": 0.01026919, + "balance_loss_clip": 0.97452706, + "balance_loss_mlp": 1.01850915, + "epoch": 0.5123549570131666, + "flos": 22528487756160.0, + "grad_norm": 1.8076966701127597, + "language_loss": 0.72559917, + "learning_rate": 2.017137166530266e-06, + "loss": 0.74713653, + "num_input_tokens_seen": 91966325, + "step": 4261, + "time_per_iteration": 2.6733715534210205 + }, + { + "auxiliary_loss_clip": 0.01140121, + "auxiliary_loss_mlp": 0.01030449, + "balance_loss_clip": 0.97962028, + "balance_loss_mlp": 1.02267706, + "epoch": 0.5124751999038056, + "flos": 20333375804160.0, + "grad_norm": 1.9590454124969998, + "language_loss": 0.8004775, + "learning_rate": 2.0163582221996213e-06, + "loss": 0.82218325, + "num_input_tokens_seen": 91984700, + "step": 4262, + "time_per_iteration": 3.5656075477600098 + }, + { + "auxiliary_loss_clip": 0.01134146, + "auxiliary_loss_mlp": 0.01028842, + "balance_loss_clip": 0.97846913, + "balance_loss_mlp": 1.02049184, + "epoch": 0.5125954427944448, + "flos": 39785970211200.0, + "grad_norm": 2.5619620933624865, + "language_loss": 0.67910051, + "learning_rate": 2.015579275387446e-06, + "loss": 0.70073044, + "num_input_tokens_seen": 92010020, + "step": 4263, + "time_per_iteration": 2.8095977306365967 + }, + { + "auxiliary_loss_clip": 0.01125499, + "auxiliary_loss_mlp": 0.01025312, + "balance_loss_clip": 0.97761714, + "balance_loss_mlp": 1.01750422, + "epoch": 0.5127156856850839, + "flos": 29205394358400.0, + "grad_norm": 2.927502461261388, + "language_loss": 0.69088888, + "learning_rate": 2.0148003262119085e-06, + "loss": 0.71239698, + "num_input_tokens_seen": 92030990, + "step": 4264, + "time_per_iteration": 2.7359261512756348 + }, + { + "auxiliary_loss_clip": 0.01110011, + "auxiliary_loss_mlp": 0.01027817, + "balance_loss_clip": 0.94066882, + "balance_loss_mlp": 1.01959729, + "epoch": 0.5128359285757229, + "flos": 13553693412480.0, + "grad_norm": 1.8868202352441352, + "language_loss": 0.76476943, + "learning_rate": 2.0140213747911728e-06, + "loss": 0.78614771, + "num_input_tokens_seen": 92049525, + "step": 4265, + "time_per_iteration": 2.6642866134643555 + }, + { + "auxiliary_loss_clip": 0.01112846, + "auxiliary_loss_mlp": 0.01031908, + "balance_loss_clip": 0.94405967, + "balance_loss_mlp": 1.02354515, + "epoch": 0.5129561714663621, + "flos": 25192089820800.0, + "grad_norm": 2.134153728824406, + "language_loss": 0.80740368, + "learning_rate": 2.013242421243406e-06, + "loss": 0.82885122, + "num_input_tokens_seen": 92068430, + "step": 4266, + "time_per_iteration": 2.7088775634765625 + }, + { + "auxiliary_loss_clip": 0.01089379, + "auxiliary_loss_mlp": 0.01031376, + "balance_loss_clip": 0.90392101, + "balance_loss_mlp": 1.02401483, + "epoch": 0.5130764143570011, + "flos": 18150223080960.0, + "grad_norm": 2.760941557670921, + "language_loss": 0.79172468, + "learning_rate": 2.012463465686774e-06, + "loss": 0.81293225, + "num_input_tokens_seen": 92088180, + "step": 4267, + "time_per_iteration": 2.731274366378784 + }, + { + "auxiliary_loss_clip": 0.01002266, + "auxiliary_loss_mlp": 0.01003154, + "balance_loss_clip": 0.87789464, + "balance_loss_mlp": 1.00168765, + "epoch": 0.5131966572476402, + "flos": 59794896418560.0, + "grad_norm": 0.7627273187672287, + "language_loss": 0.54747832, + "learning_rate": 2.0116845082394446e-06, + "loss": 0.56753242, + "num_input_tokens_seen": 92153015, + "step": 4268, + "time_per_iteration": 3.2725777626037598 + }, + { + "auxiliary_loss_clip": 0.01158107, + "auxiliary_loss_mlp": 0.01032596, + "balance_loss_clip": 1.01523495, + "balance_loss_mlp": 1.02451396, + "epoch": 0.5133169001382794, + "flos": 18515219132160.0, + "grad_norm": 2.2231453257794076, + "language_loss": 0.79072946, + "learning_rate": 2.0109055490195836e-06, + "loss": 0.81263649, + "num_input_tokens_seen": 92171470, + "step": 4269, + "time_per_iteration": 2.7112696170806885 + }, + { + "auxiliary_loss_clip": 0.01065962, + "auxiliary_loss_mlp": 0.01029015, + "balance_loss_clip": 0.85683882, + "balance_loss_mlp": 1.02105784, + "epoch": 0.5134371430289184, + "flos": 15523537219200.0, + "grad_norm": 2.1263492645672066, + "language_loss": 0.64581758, + "learning_rate": 2.0101265881453605e-06, + "loss": 0.66676736, + "num_input_tokens_seen": 92189945, + "step": 4270, + "time_per_iteration": 2.8413925170898438 + }, + { + "auxiliary_loss_clip": 0.01130775, + "auxiliary_loss_mlp": 0.0102748, + "balance_loss_clip": 0.98226202, + "balance_loss_mlp": 1.02042842, + "epoch": 0.5135573859195575, + "flos": 21433786911360.0, + "grad_norm": 2.088296641650012, + "language_loss": 0.787709, + "learning_rate": 2.009347625734941e-06, + "loss": 0.80929148, + "num_input_tokens_seen": 92209855, + "step": 4271, + "time_per_iteration": 2.718761920928955 + }, + { + "auxiliary_loss_clip": 0.01188346, + "auxiliary_loss_mlp": 0.01027277, + "balance_loss_clip": 1.06152523, + "balance_loss_mlp": 1.01930141, + "epoch": 0.5136776288101966, + "flos": 17712651600000.0, + "grad_norm": 2.1437410794912295, + "language_loss": 0.75312084, + "learning_rate": 2.0085686619064954e-06, + "loss": 0.77527702, + "num_input_tokens_seen": 92226295, + "step": 4272, + "time_per_iteration": 2.6149168014526367 + }, + { + "auxiliary_loss_clip": 0.01163863, + "auxiliary_loss_mlp": 0.01043011, + "balance_loss_clip": 1.02022445, + "balance_loss_mlp": 1.03484559, + "epoch": 0.5137978717008357, + "flos": 16581680997120.0, + "grad_norm": 2.377852464354372, + "language_loss": 0.8276155, + "learning_rate": 2.00778969677819e-06, + "loss": 0.84968424, + "num_input_tokens_seen": 92243330, + "step": 4273, + "time_per_iteration": 2.610198974609375 + }, + { + "auxiliary_loss_clip": 0.01132614, + "auxiliary_loss_mlp": 0.01030896, + "balance_loss_clip": 0.97852772, + "balance_loss_mlp": 1.0233562, + "epoch": 0.5139181145914747, + "flos": 20668243322880.0, + "grad_norm": 1.7388507922810816, + "language_loss": 0.63908815, + "learning_rate": 2.0070107304681934e-06, + "loss": 0.66072327, + "num_input_tokens_seen": 92262285, + "step": 4274, + "time_per_iteration": 2.7048816680908203 + }, + { + "auxiliary_loss_clip": 0.0111106, + "auxiliary_loss_mlp": 0.01026929, + "balance_loss_clip": 0.94216609, + "balance_loss_mlp": 1.01957965, + "epoch": 0.5140383574821139, + "flos": 32926996546560.0, + "grad_norm": 1.9949015456840142, + "language_loss": 0.78244388, + "learning_rate": 2.006231763094675e-06, + "loss": 0.80382383, + "num_input_tokens_seen": 92283305, + "step": 4275, + "time_per_iteration": 2.8111624717712402 + }, + { + "auxiliary_loss_clip": 0.01130216, + "auxiliary_loss_mlp": 0.01025443, + "balance_loss_clip": 0.98146302, + "balance_loss_mlp": 1.01772463, + "epoch": 0.514158600372753, + "flos": 19537093152000.0, + "grad_norm": 2.1824898643050674, + "language_loss": 0.87388736, + "learning_rate": 2.0054527947758027e-06, + "loss": 0.89544398, + "num_input_tokens_seen": 92302105, + "step": 4276, + "time_per_iteration": 2.649338722229004 + }, + { + "auxiliary_loss_clip": 0.01056806, + "auxiliary_loss_mlp": 0.01005182, + "balance_loss_clip": 0.98290759, + "balance_loss_mlp": 1.00387692, + "epoch": 0.514278843263392, + "flos": 62523855279360.0, + "grad_norm": 0.7304444251119547, + "language_loss": 0.55894679, + "learning_rate": 2.004673825629746e-06, + "loss": 0.57956672, + "num_input_tokens_seen": 92362885, + "step": 4277, + "time_per_iteration": 4.06997275352478 + }, + { + "auxiliary_loss_clip": 0.01128619, + "auxiliary_loss_mlp": 0.01031853, + "balance_loss_clip": 0.97548676, + "balance_loss_mlp": 1.02364552, + "epoch": 0.5143990861540312, + "flos": 25882328545920.0, + "grad_norm": 2.0203566725609168, + "language_loss": 0.72274035, + "learning_rate": 2.0038948557746744e-06, + "loss": 0.74434507, + "num_input_tokens_seen": 92384740, + "step": 4278, + "time_per_iteration": 2.689120054244995 + }, + { + "auxiliary_loss_clip": 0.01154935, + "auxiliary_loss_mlp": 0.01028568, + "balance_loss_clip": 1.01902354, + "balance_loss_mlp": 1.02053928, + "epoch": 0.5145193290446702, + "flos": 23330660238720.0, + "grad_norm": 2.0520687020936386, + "language_loss": 0.75144887, + "learning_rate": 2.0031158853287558e-06, + "loss": 0.7732839, + "num_input_tokens_seen": 92405175, + "step": 4279, + "time_per_iteration": 3.5780975818634033 + }, + { + "auxiliary_loss_clip": 0.01136556, + "auxiliary_loss_mlp": 0.0102847, + "balance_loss_clip": 0.98233652, + "balance_loss_mlp": 1.02028012, + "epoch": 0.5146395719353093, + "flos": 22856603518080.0, + "grad_norm": 1.9395231915791384, + "language_loss": 0.70421016, + "learning_rate": 2.0023369144101593e-06, + "loss": 0.72586036, + "num_input_tokens_seen": 92423345, + "step": 4280, + "time_per_iteration": 2.7063708305358887 + }, + { + "auxiliary_loss_clip": 0.01125403, + "auxiliary_loss_mlp": 0.01026301, + "balance_loss_clip": 0.97708654, + "balance_loss_mlp": 1.01837337, + "epoch": 0.5147598148259485, + "flos": 26391577616640.0, + "grad_norm": 2.248593555685358, + "language_loss": 0.76945293, + "learning_rate": 2.0015579431370555e-06, + "loss": 0.79096997, + "num_input_tokens_seen": 92445025, + "step": 4281, + "time_per_iteration": 3.649122476577759 + }, + { + "auxiliary_loss_clip": 0.01154938, + "auxiliary_loss_mlp": 0.01026841, + "balance_loss_clip": 1.01832211, + "balance_loss_mlp": 1.01929522, + "epoch": 0.5148800577165875, + "flos": 29965694561280.0, + "grad_norm": 2.1718893027536867, + "language_loss": 0.69893628, + "learning_rate": 2.000778971627612e-06, + "loss": 0.72075409, + "num_input_tokens_seen": 92464490, + "step": 4282, + "time_per_iteration": 2.6659493446350098 + }, + { + "auxiliary_loss_clip": 0.01127919, + "auxiliary_loss_mlp": 0.01029142, + "balance_loss_clip": 0.97646022, + "balance_loss_mlp": 1.02167082, + "epoch": 0.5150003006072266, + "flos": 17931383470080.0, + "grad_norm": 1.8616762079190199, + "language_loss": 0.9029628, + "learning_rate": 2e-06, + "loss": 0.92453337, + "num_input_tokens_seen": 92482085, + "step": 4283, + "time_per_iteration": 2.673510789871216 + }, + { + "auxiliary_loss_clip": 0.01181348, + "auxiliary_loss_mlp": 0.01029469, + "balance_loss_clip": 1.05650353, + "balance_loss_mlp": 1.02185106, + "epoch": 0.5151205434978657, + "flos": 18478733892480.0, + "grad_norm": 1.6760144228087308, + "language_loss": 0.85602438, + "learning_rate": 1.9992210283723878e-06, + "loss": 0.87813258, + "num_input_tokens_seen": 92499325, + "step": 4284, + "time_per_iteration": 2.599613904953003 + }, + { + "auxiliary_loss_clip": 0.01180513, + "auxiliary_loss_mlp": 0.01030082, + "balance_loss_clip": 1.05763257, + "balance_loss_mlp": 1.02275109, + "epoch": 0.5152407863885048, + "flos": 25341263003520.0, + "grad_norm": 1.5109230741402198, + "language_loss": 0.79424638, + "learning_rate": 1.9984420568629448e-06, + "loss": 0.81635231, + "num_input_tokens_seen": 92522090, + "step": 4285, + "time_per_iteration": 2.663188934326172 + }, + { + "auxiliary_loss_clip": 0.01157973, + "auxiliary_loss_mlp": 0.01021779, + "balance_loss_clip": 1.01817739, + "balance_loss_mlp": 1.01456964, + "epoch": 0.5153610292791438, + "flos": 18329740277760.0, + "grad_norm": 2.7001180499577693, + "language_loss": 0.78223991, + "learning_rate": 1.9976630855898405e-06, + "loss": 0.80403745, + "num_input_tokens_seen": 92539845, + "step": 4286, + "time_per_iteration": 2.6193172931671143 + }, + { + "auxiliary_loss_clip": 0.01127501, + "auxiliary_loss_mlp": 0.01031386, + "balance_loss_clip": 0.97445673, + "balance_loss_mlp": 1.02378082, + "epoch": 0.515481272169783, + "flos": 30409945971840.0, + "grad_norm": 1.9509182460430328, + "language_loss": 0.74844182, + "learning_rate": 1.9968841146712445e-06, + "loss": 0.77003074, + "num_input_tokens_seen": 92559460, + "step": 4287, + "time_per_iteration": 2.743269443511963 + }, + { + "auxiliary_loss_clip": 0.01057475, + "auxiliary_loss_mlp": 0.00510224, + "balance_loss_clip": 0.85933113, + "balance_loss_mlp": 0.0, + "epoch": 0.5156015150604221, + "flos": 23037305863680.0, + "grad_norm": 2.818983376788997, + "language_loss": 0.71330392, + "learning_rate": 1.996105144225326e-06, + "loss": 0.7289809, + "num_input_tokens_seen": 92579695, + "step": 4288, + "time_per_iteration": 3.6752710342407227 + }, + { + "auxiliary_loss_clip": 0.01159024, + "auxiliary_loss_mlp": 0.01024178, + "balance_loss_clip": 1.01917887, + "balance_loss_mlp": 1.01665044, + "epoch": 0.5157217579510611, + "flos": 17858556645120.0, + "grad_norm": 1.969231932833613, + "language_loss": 0.78692997, + "learning_rate": 1.995326174370254e-06, + "loss": 0.80876201, + "num_input_tokens_seen": 92598795, + "step": 4289, + "time_per_iteration": 2.6977434158325195 + }, + { + "auxiliary_loss_clip": 0.011539, + "auxiliary_loss_mlp": 0.00509007, + "balance_loss_clip": 1.0156008, + "balance_loss_mlp": 0.0, + "epoch": 0.5158420008417003, + "flos": 19171486569600.0, + "grad_norm": 1.6252469253753676, + "language_loss": 0.73303002, + "learning_rate": 1.994547205224197e-06, + "loss": 0.74965906, + "num_input_tokens_seen": 92617700, + "step": 4290, + "time_per_iteration": 2.6444225311279297 + }, + { + "auxiliary_loss_clip": 0.01133279, + "auxiliary_loss_mlp": 0.01027719, + "balance_loss_clip": 0.98343033, + "balance_loss_mlp": 1.02014279, + "epoch": 0.5159622437323393, + "flos": 22419534827520.0, + "grad_norm": 2.259917598903427, + "language_loss": 0.67868912, + "learning_rate": 1.993768236905325e-06, + "loss": 0.70029902, + "num_input_tokens_seen": 92638370, + "step": 4291, + "time_per_iteration": 2.820357084274292 + }, + { + "auxiliary_loss_clip": 0.01128719, + "auxiliary_loss_mlp": 0.01028835, + "balance_loss_clip": 0.97609121, + "balance_loss_mlp": 1.02101517, + "epoch": 0.5160824866229784, + "flos": 24603010773120.0, + "grad_norm": 2.735543368124695, + "language_loss": 0.66235602, + "learning_rate": 1.992989269531807e-06, + "loss": 0.68393153, + "num_input_tokens_seen": 92657180, + "step": 4292, + "time_per_iteration": 2.709911823272705 + }, + { + "auxiliary_loss_clip": 0.0113331, + "auxiliary_loss_mlp": 0.01024928, + "balance_loss_clip": 0.97640342, + "balance_loss_mlp": 1.01717365, + "epoch": 0.5162027295136175, + "flos": 18002737837440.0, + "grad_norm": 3.4205207979232304, + "language_loss": 0.68783081, + "learning_rate": 1.99221030322181e-06, + "loss": 0.70941317, + "num_input_tokens_seen": 92673985, + "step": 4293, + "time_per_iteration": 2.660099744796753 + }, + { + "auxiliary_loss_clip": 0.01136625, + "auxiliary_loss_mlp": 0.01031523, + "balance_loss_clip": 0.97883129, + "balance_loss_mlp": 1.02408457, + "epoch": 0.5163229724042566, + "flos": 27344611221120.0, + "grad_norm": 1.4378963514117418, + "language_loss": 0.80594623, + "learning_rate": 1.991431338093505e-06, + "loss": 0.82762766, + "num_input_tokens_seen": 92696340, + "step": 4294, + "time_per_iteration": 2.832752227783203 + }, + { + "auxiliary_loss_clip": 0.01134201, + "auxiliary_loss_mlp": 0.0103383, + "balance_loss_clip": 0.98320401, + "balance_loss_mlp": 1.02654958, + "epoch": 0.5164432152948957, + "flos": 21762764599680.0, + "grad_norm": 1.812545058180647, + "language_loss": 0.79487288, + "learning_rate": 1.9906523742650587e-06, + "loss": 0.81655312, + "num_input_tokens_seen": 92715200, + "step": 4295, + "time_per_iteration": 2.6947505474090576 + }, + { + "auxiliary_loss_clip": 0.01181505, + "auxiliary_loss_mlp": 0.01022586, + "balance_loss_clip": 1.05545485, + "balance_loss_mlp": 1.0147959, + "epoch": 0.5165634581855347, + "flos": 25550334115200.0, + "grad_norm": 1.79846942634024, + "language_loss": 0.77596265, + "learning_rate": 1.9898734118546397e-06, + "loss": 0.79800355, + "num_input_tokens_seen": 92735150, + "step": 4296, + "time_per_iteration": 2.655029773712158 + }, + { + "auxiliary_loss_clip": 0.01045825, + "auxiliary_loss_mlp": 0.01027377, + "balance_loss_clip": 0.82557809, + "balance_loss_mlp": 1.01958704, + "epoch": 0.5166837010761739, + "flos": 19901191363200.0, + "grad_norm": 1.4489241408128934, + "language_loss": 0.80181062, + "learning_rate": 1.989094450980416e-06, + "loss": 0.82254267, + "num_input_tokens_seen": 92755250, + "step": 4297, + "time_per_iteration": 3.057720184326172 + }, + { + "auxiliary_loss_clip": 0.01152564, + "auxiliary_loss_mlp": 0.01026528, + "balance_loss_clip": 1.01600003, + "balance_loss_mlp": 1.01860619, + "epoch": 0.516803943966813, + "flos": 26646076454400.0, + "grad_norm": 1.9154500148130265, + "language_loss": 0.76489758, + "learning_rate": 1.9883154917605556e-06, + "loss": 0.78668845, + "num_input_tokens_seen": 92774460, + "step": 4298, + "time_per_iteration": 3.0219151973724365 + }, + { + "auxiliary_loss_clip": 0.0118151, + "auxiliary_loss_mlp": 0.01024608, + "balance_loss_clip": 1.05697799, + "balance_loss_mlp": 1.01719046, + "epoch": 0.516924186857452, + "flos": 19682854542720.0, + "grad_norm": 1.8042132643967148, + "language_loss": 0.83296275, + "learning_rate": 1.9875365343132262e-06, + "loss": 0.85502398, + "num_input_tokens_seen": 92791580, + "step": 4299, + "time_per_iteration": 2.603159189224243 + }, + { + "auxiliary_loss_clip": 0.01158475, + "auxiliary_loss_mlp": 0.00509257, + "balance_loss_clip": 1.01957297, + "balance_loss_mlp": 0.0, + "epoch": 0.5170444297480912, + "flos": 15956583586560.0, + "grad_norm": 2.0501269250070875, + "language_loss": 0.84841633, + "learning_rate": 1.9867575787565946e-06, + "loss": 0.86509371, + "num_input_tokens_seen": 92806240, + "step": 4300, + "time_per_iteration": 2.6466939449310303 + }, + { + "auxiliary_loss_clip": 0.01162029, + "auxiliary_loss_mlp": 0.01032397, + "balance_loss_clip": 1.02090847, + "balance_loss_mlp": 1.02443981, + "epoch": 0.5171646726387302, + "flos": 14174157968640.0, + "grad_norm": 1.9848556251827918, + "language_loss": 0.85626215, + "learning_rate": 1.9859786252088275e-06, + "loss": 0.87820637, + "num_input_tokens_seen": 92823420, + "step": 4301, + "time_per_iteration": 2.6272635459899902 + }, + { + "auxiliary_loss_clip": 0.01111182, + "auxiliary_loss_mlp": 0.01029838, + "balance_loss_clip": 0.94026464, + "balance_loss_mlp": 1.02217269, + "epoch": 0.5172849155293693, + "flos": 23578550974080.0, + "grad_norm": 3.687636248522191, + "language_loss": 0.66221076, + "learning_rate": 1.9851996737880914e-06, + "loss": 0.68362093, + "num_input_tokens_seen": 92838605, + "step": 4302, + "time_per_iteration": 2.7373809814453125 + }, + { + "auxiliary_loss_clip": 0.01162869, + "auxiliary_loss_mlp": 0.01032623, + "balance_loss_clip": 1.01864767, + "balance_loss_mlp": 1.02376533, + "epoch": 0.5174051584200084, + "flos": 14283541860480.0, + "grad_norm": 2.306366384905195, + "language_loss": 0.7412672, + "learning_rate": 1.9844207246125537e-06, + "loss": 0.7632221, + "num_input_tokens_seen": 92855185, + "step": 4303, + "time_per_iteration": 3.5455257892608643 + }, + { + "auxiliary_loss_clip": 0.01131273, + "auxiliary_loss_mlp": 0.01028838, + "balance_loss_clip": 0.97902459, + "balance_loss_mlp": 1.02116096, + "epoch": 0.5175254013106475, + "flos": 37889384192640.0, + "grad_norm": 1.743311525371231, + "language_loss": 0.68391049, + "learning_rate": 1.983641777800379e-06, + "loss": 0.70551157, + "num_input_tokens_seen": 92877830, + "step": 4304, + "time_per_iteration": 2.787951707839966 + }, + { + "auxiliary_loss_clip": 0.01040281, + "auxiliary_loss_mlp": 0.01007444, + "balance_loss_clip": 0.94459558, + "balance_loss_mlp": 1.00615633, + "epoch": 0.5176456442012866, + "flos": 68549737829760.0, + "grad_norm": 0.7521849088514939, + "language_loss": 0.5876202, + "learning_rate": 1.9828628334697343e-06, + "loss": 0.60809743, + "num_input_tokens_seen": 92945040, + "step": 4305, + "time_per_iteration": 4.34137749671936 + }, + { + "auxiliary_loss_clip": 0.01042528, + "auxiliary_loss_mlp": 0.01003238, + "balance_loss_clip": 0.94716644, + "balance_loss_mlp": 1.0019623, + "epoch": 0.5177658870919257, + "flos": 64084137235200.0, + "grad_norm": 1.6214034463756042, + "language_loss": 0.54670084, + "learning_rate": 1.982083891738784e-06, + "loss": 0.56715846, + "num_input_tokens_seen": 93005910, + "step": 4306, + "time_per_iteration": 3.3092527389526367 + }, + { + "auxiliary_loss_clip": 0.01128277, + "auxiliary_loss_mlp": 0.01022148, + "balance_loss_clip": 0.97993177, + "balance_loss_mlp": 1.0147723, + "epoch": 0.5178861299825648, + "flos": 26651248012800.0, + "grad_norm": 1.509008805971816, + "language_loss": 0.82864523, + "learning_rate": 1.9813049527256923e-06, + "loss": 0.85014951, + "num_input_tokens_seen": 93026305, + "step": 4307, + "time_per_iteration": 3.64286208152771 + }, + { + "auxiliary_loss_clip": 0.01107307, + "auxiliary_loss_mlp": 0.01028349, + "balance_loss_clip": 0.93706346, + "balance_loss_mlp": 1.0206542, + "epoch": 0.5180063728732038, + "flos": 17931886260480.0, + "grad_norm": 3.6040903391085637, + "language_loss": 0.82736194, + "learning_rate": 1.9805260165486252e-06, + "loss": 0.84871852, + "num_input_tokens_seen": 93045675, + "step": 4308, + "time_per_iteration": 2.726879596710205 + }, + { + "auxiliary_loss_clip": 0.01158346, + "auxiliary_loss_mlp": 0.0102753, + "balance_loss_clip": 1.02048779, + "balance_loss_mlp": 1.01997221, + "epoch": 0.518126615763843, + "flos": 19500895221120.0, + "grad_norm": 3.171011437910005, + "language_loss": 0.86355221, + "learning_rate": 1.9797470833257457e-06, + "loss": 0.88541102, + "num_input_tokens_seen": 93065375, + "step": 4309, + "time_per_iteration": 2.659817934036255 + }, + { + "auxiliary_loss_clip": 0.01159658, + "auxiliary_loss_mlp": 0.01028926, + "balance_loss_clip": 1.02048135, + "balance_loss_mlp": 1.02136481, + "epoch": 0.5182468586544821, + "flos": 20704082117760.0, + "grad_norm": 2.5752986646749028, + "language_loss": 0.77065629, + "learning_rate": 1.9789681531752177e-06, + "loss": 0.79254216, + "num_input_tokens_seen": 93085595, + "step": 4310, + "time_per_iteration": 2.6565840244293213 + }, + { + "auxiliary_loss_clip": 0.01084836, + "auxiliary_loss_mlp": 0.01023872, + "balance_loss_clip": 0.90230531, + "balance_loss_mlp": 1.01627827, + "epoch": 0.5183671015451211, + "flos": 23112107936640.0, + "grad_norm": 1.465511047136931, + "language_loss": 0.72622955, + "learning_rate": 1.978189226215204e-06, + "loss": 0.7473166, + "num_input_tokens_seen": 93106140, + "step": 4311, + "time_per_iteration": 2.7827422618865967 + }, + { + "auxiliary_loss_clip": 0.01178772, + "auxiliary_loss_mlp": 0.01022623, + "balance_loss_clip": 1.05387008, + "balance_loss_mlp": 1.01514292, + "epoch": 0.5184873444357603, + "flos": 17597090568960.0, + "grad_norm": 2.250065042874652, + "language_loss": 0.77225757, + "learning_rate": 1.9774103025638675e-06, + "loss": 0.79427153, + "num_input_tokens_seen": 93124265, + "step": 4312, + "time_per_iteration": 2.5594279766082764 + }, + { + "auxiliary_loss_clip": 0.0108888, + "auxiliary_loss_mlp": 0.01036914, + "balance_loss_clip": 0.90540332, + "balance_loss_mlp": 1.02868235, + "epoch": 0.5186075873263993, + "flos": 24936800883840.0, + "grad_norm": 1.915762661199746, + "language_loss": 0.76602536, + "learning_rate": 1.9766313823393696e-06, + "loss": 0.7872833, + "num_input_tokens_seen": 93145130, + "step": 4313, + "time_per_iteration": 2.8036210536956787 + }, + { + "auxiliary_loss_clip": 0.01079236, + "auxiliary_loss_mlp": 0.01026776, + "balance_loss_clip": 0.89671278, + "balance_loss_mlp": 1.0193193, + "epoch": 0.5187278302170384, + "flos": 15190106244480.0, + "grad_norm": 2.065753585772698, + "language_loss": 0.69288045, + "learning_rate": 1.975852465659873e-06, + "loss": 0.71394056, + "num_input_tokens_seen": 93161110, + "step": 4314, + "time_per_iteration": 3.648472309112549 + }, + { + "auxiliary_loss_clip": 0.01156872, + "auxiliary_loss_mlp": 0.01027047, + "balance_loss_clip": 1.01731586, + "balance_loss_mlp": 1.01900005, + "epoch": 0.5188480731076776, + "flos": 25009412227200.0, + "grad_norm": 2.0999846336506454, + "language_loss": 0.7031787, + "learning_rate": 1.9750735526435377e-06, + "loss": 0.72501791, + "num_input_tokens_seen": 93178055, + "step": 4315, + "time_per_iteration": 2.671198606491089 + }, + { + "auxiliary_loss_clip": 0.01134741, + "auxiliary_loss_mlp": 0.01024108, + "balance_loss_clip": 0.98415279, + "balance_loss_mlp": 1.01604986, + "epoch": 0.5189683159983166, + "flos": 24790141653120.0, + "grad_norm": 2.139671904614432, + "language_loss": 0.79532492, + "learning_rate": 1.974294643408525e-06, + "loss": 0.81691337, + "num_input_tokens_seen": 93195850, + "step": 4316, + "time_per_iteration": 2.7126739025115967 + }, + { + "auxiliary_loss_clip": 0.01160062, + "auxiliary_loss_mlp": 0.01026836, + "balance_loss_clip": 1.01617455, + "balance_loss_mlp": 1.0186646, + "epoch": 0.5190885588889557, + "flos": 24754266944640.0, + "grad_norm": 2.0572352577290838, + "language_loss": 0.6710698, + "learning_rate": 1.9735157380729947e-06, + "loss": 0.6929388, + "num_input_tokens_seen": 93216260, + "step": 4317, + "time_per_iteration": 2.665907859802246 + }, + { + "auxiliary_loss_clip": 0.01136714, + "auxiliary_loss_mlp": 0.01034087, + "balance_loss_clip": 0.97953457, + "balance_loss_mlp": 1.0263803, + "epoch": 0.5192088017795948, + "flos": 24712646060160.0, + "grad_norm": 1.9489099517069315, + "language_loss": 0.84545445, + "learning_rate": 1.9727368367551053e-06, + "loss": 0.86716241, + "num_input_tokens_seen": 93234810, + "step": 4318, + "time_per_iteration": 2.7276761531829834 + }, + { + "auxiliary_loss_clip": 0.01125023, + "auxiliary_loss_mlp": 0.01032286, + "balance_loss_clip": 0.9772014, + "balance_loss_mlp": 1.02437067, + "epoch": 0.5193290446702339, + "flos": 27229588894080.0, + "grad_norm": 2.3572746065640215, + "language_loss": 0.68463588, + "learning_rate": 1.9719579395730164e-06, + "loss": 0.70620906, + "num_input_tokens_seen": 93254185, + "step": 4319, + "time_per_iteration": 2.7326819896698 + }, + { + "auxiliary_loss_clip": 0.01182778, + "auxiliary_loss_mlp": 0.0103279, + "balance_loss_clip": 1.05771017, + "balance_loss_mlp": 1.02474976, + "epoch": 0.5194492875608729, + "flos": 11473352392320.0, + "grad_norm": 2.1399501298046815, + "language_loss": 0.93609262, + "learning_rate": 1.9711790466448854e-06, + "loss": 0.95824826, + "num_input_tokens_seen": 93268205, + "step": 4320, + "time_per_iteration": 2.5652599334716797 + }, + { + "auxiliary_loss_clip": 0.01088686, + "auxiliary_loss_mlp": 0.01027685, + "balance_loss_clip": 0.90283531, + "balance_loss_mlp": 1.01991904, + "epoch": 0.5195695304515121, + "flos": 20338906498560.0, + "grad_norm": 2.0812505682107916, + "language_loss": 0.70859063, + "learning_rate": 1.9704001580888704e-06, + "loss": 0.72975433, + "num_input_tokens_seen": 93286945, + "step": 4321, + "time_per_iteration": 2.7654547691345215 + }, + { + "auxiliary_loss_clip": 0.01127034, + "auxiliary_loss_mlp": 0.00510128, + "balance_loss_clip": 0.9753992, + "balance_loss_mlp": 0.0, + "epoch": 0.5196897733421512, + "flos": 20048317470720.0, + "grad_norm": 1.8664638877460622, + "language_loss": 0.86488503, + "learning_rate": 1.9696212740231283e-06, + "loss": 0.8812567, + "num_input_tokens_seen": 93305595, + "step": 4322, + "time_per_iteration": 2.6574087142944336 + }, + { + "auxiliary_loss_clip": 0.01164459, + "auxiliary_loss_mlp": 0.01027101, + "balance_loss_clip": 1.01738548, + "balance_loss_mlp": 1.01869082, + "epoch": 0.5198100162327902, + "flos": 23805507058560.0, + "grad_norm": 2.1592640006559614, + "language_loss": 0.82189614, + "learning_rate": 1.9688423945658146e-06, + "loss": 0.84381175, + "num_input_tokens_seen": 93326460, + "step": 4323, + "time_per_iteration": 2.6461644172668457 + }, + { + "auxiliary_loss_clip": 0.01072659, + "auxiliary_loss_mlp": 0.01026106, + "balance_loss_clip": 0.89321947, + "balance_loss_mlp": 1.01794624, + "epoch": 0.5199302591234293, + "flos": 24023951619840.0, + "grad_norm": 2.0336550084918863, + "language_loss": 0.71860039, + "learning_rate": 1.9680635198350845e-06, + "loss": 0.73958802, + "num_input_tokens_seen": 93346170, + "step": 4324, + "time_per_iteration": 2.833077907562256 + }, + { + "auxiliary_loss_clip": 0.01156966, + "auxiliary_loss_mlp": 0.01026641, + "balance_loss_clip": 1.01537609, + "balance_loss_mlp": 1.01839733, + "epoch": 0.5200505020140684, + "flos": 26359366095360.0, + "grad_norm": 1.8262086063180523, + "language_loss": 0.72597289, + "learning_rate": 1.967284649949093e-06, + "loss": 0.74780905, + "num_input_tokens_seen": 93365380, + "step": 4325, + "time_per_iteration": 2.697875738143921 + }, + { + "auxiliary_loss_clip": 0.01106696, + "auxiliary_loss_mlp": 0.0103112, + "balance_loss_clip": 0.93792516, + "balance_loss_mlp": 1.02344847, + "epoch": 0.5201707449047075, + "flos": 39604262284800.0, + "grad_norm": 2.3077610003930196, + "language_loss": 0.72820854, + "learning_rate": 1.966505785025994e-06, + "loss": 0.7495867, + "num_input_tokens_seen": 93387285, + "step": 4326, + "time_per_iteration": 2.8886349201202393 + }, + { + "auxiliary_loss_clip": 0.01111111, + "auxiliary_loss_mlp": 0.01026525, + "balance_loss_clip": 0.94150198, + "balance_loss_mlp": 1.01816273, + "epoch": 0.5202909877953465, + "flos": 53682788292480.0, + "grad_norm": 1.7225647983650698, + "language_loss": 0.76254845, + "learning_rate": 1.965726925183941e-06, + "loss": 0.78392482, + "num_input_tokens_seen": 93410390, + "step": 4327, + "time_per_iteration": 2.997232675552368 + }, + { + "auxiliary_loss_clip": 0.011802, + "auxiliary_loss_mlp": 0.01027369, + "balance_loss_clip": 1.05558634, + "balance_loss_mlp": 1.01925111, + "epoch": 0.5204112306859857, + "flos": 19537021324800.0, + "grad_norm": 2.7493458056021294, + "language_loss": 0.85077477, + "learning_rate": 1.964948070541087e-06, + "loss": 0.87285048, + "num_input_tokens_seen": 93429050, + "step": 4328, + "time_per_iteration": 2.6136724948883057 + }, + { + "auxiliary_loss_clip": 0.01147208, + "auxiliary_loss_mlp": 0.01030096, + "balance_loss_clip": 1.01323354, + "balance_loss_mlp": 1.02199602, + "epoch": 0.5205314735766248, + "flos": 15304697608320.0, + "grad_norm": 2.2816945155738684, + "language_loss": 0.69677126, + "learning_rate": 1.9641692212155816e-06, + "loss": 0.71854436, + "num_input_tokens_seen": 93446815, + "step": 4329, + "time_per_iteration": 3.5547990798950195 + }, + { + "auxiliary_loss_clip": 0.01085624, + "auxiliary_loss_mlp": 0.01031335, + "balance_loss_clip": 0.90583873, + "balance_loss_mlp": 1.02331805, + "epoch": 0.5206517164672638, + "flos": 59263701160320.0, + "grad_norm": 1.753176264331249, + "language_loss": 0.72419691, + "learning_rate": 1.9633903773255777e-06, + "loss": 0.74536645, + "num_input_tokens_seen": 93469130, + "step": 4330, + "time_per_iteration": 4.012321710586548 + }, + { + "auxiliary_loss_clip": 0.01178367, + "auxiliary_loss_mlp": 0.01026347, + "balance_loss_clip": 1.05383515, + "balance_loss_mlp": 1.01794314, + "epoch": 0.520771959357903, + "flos": 26871129118080.0, + "grad_norm": 1.6927666127343182, + "language_loss": 0.74931645, + "learning_rate": 1.9626115389892237e-06, + "loss": 0.77136362, + "num_input_tokens_seen": 93489920, + "step": 4331, + "time_per_iteration": 2.6342697143554688 + }, + { + "auxiliary_loss_clip": 0.01115459, + "auxiliary_loss_mlp": 0.01028126, + "balance_loss_clip": 0.93950653, + "balance_loss_mlp": 1.02001441, + "epoch": 0.520892202248542, + "flos": 26907075653760.0, + "grad_norm": 1.9086200898344845, + "language_loss": 0.85778999, + "learning_rate": 1.96183270632467e-06, + "loss": 0.87922585, + "num_input_tokens_seen": 93509770, + "step": 4332, + "time_per_iteration": 3.5866293907165527 + }, + { + "auxiliary_loss_clip": 0.01105882, + "auxiliary_loss_mlp": 0.00510862, + "balance_loss_clip": 0.93994159, + "balance_loss_mlp": 0.0, + "epoch": 0.5210124451391811, + "flos": 25849434666240.0, + "grad_norm": 1.6147828656635475, + "language_loss": 0.79039156, + "learning_rate": 1.9610538794500644e-06, + "loss": 0.80655897, + "num_input_tokens_seen": 93529320, + "step": 4333, + "time_per_iteration": 2.7278735637664795 + }, + { + "auxiliary_loss_clip": 0.01016548, + "auxiliary_loss_mlp": 0.01002512, + "balance_loss_clip": 0.90520501, + "balance_loss_mlp": 1.00107574, + "epoch": 0.5211326880298203, + "flos": 70553804319360.0, + "grad_norm": 0.8602310124507829, + "language_loss": 0.59444296, + "learning_rate": 1.9602750584835542e-06, + "loss": 0.61463356, + "num_input_tokens_seen": 93595255, + "step": 4334, + "time_per_iteration": 3.3929696083068848 + }, + { + "auxiliary_loss_clip": 0.01128772, + "auxiliary_loss_mlp": 0.01028247, + "balance_loss_clip": 0.97572857, + "balance_loss_mlp": 1.02022433, + "epoch": 0.5212529309204593, + "flos": 15628898787840.0, + "grad_norm": 2.26083173141587, + "language_loss": 0.82556748, + "learning_rate": 1.959496243543286e-06, + "loss": 0.84713769, + "num_input_tokens_seen": 93613135, + "step": 4335, + "time_per_iteration": 2.716416358947754 + }, + { + "auxiliary_loss_clip": 0.01160461, + "auxiliary_loss_mlp": 0.01027822, + "balance_loss_clip": 1.02044487, + "balance_loss_mlp": 1.01906586, + "epoch": 0.5213731738110984, + "flos": 26242655829120.0, + "grad_norm": 2.1231970911066775, + "language_loss": 0.79740763, + "learning_rate": 1.9587174347474057e-06, + "loss": 0.81929046, + "num_input_tokens_seen": 93629645, + "step": 4336, + "time_per_iteration": 2.61511492729187 + }, + { + "auxiliary_loss_clip": 0.01070727, + "auxiliary_loss_mlp": 0.01031903, + "balance_loss_clip": 0.89709979, + "balance_loss_mlp": 1.02340317, + "epoch": 0.5214934167017375, + "flos": 19418407637760.0, + "grad_norm": 2.3962791874909914, + "language_loss": 0.82054555, + "learning_rate": 1.9579386322140574e-06, + "loss": 0.84157181, + "num_input_tokens_seen": 93645325, + "step": 4337, + "time_per_iteration": 2.720012903213501 + }, + { + "auxiliary_loss_clip": 0.01182707, + "auxiliary_loss_mlp": 0.00510241, + "balance_loss_clip": 1.05591226, + "balance_loss_mlp": 0.0, + "epoch": 0.5216136595923766, + "flos": 30955788023040.0, + "grad_norm": 3.469741397138686, + "language_loss": 0.81204355, + "learning_rate": 1.9571598360613854e-06, + "loss": 0.82897305, + "num_input_tokens_seen": 93668200, + "step": 4338, + "time_per_iteration": 2.6847856044769287 + }, + { + "auxiliary_loss_clip": 0.01116957, + "auxiliary_loss_mlp": 0.01028714, + "balance_loss_clip": 0.97095525, + "balance_loss_mlp": 1.02043498, + "epoch": 0.5217339024830157, + "flos": 21945047143680.0, + "grad_norm": 2.627314146844681, + "language_loss": 0.69824135, + "learning_rate": 1.956381046407532e-06, + "loss": 0.71969807, + "num_input_tokens_seen": 93688495, + "step": 4339, + "time_per_iteration": 2.770698308944702 + }, + { + "auxiliary_loss_clip": 0.01103554, + "auxiliary_loss_mlp": 0.01026861, + "balance_loss_clip": 0.93620265, + "balance_loss_mlp": 1.01901102, + "epoch": 0.5218541453736548, + "flos": 20923209037440.0, + "grad_norm": 1.7069853778189121, + "language_loss": 0.86435771, + "learning_rate": 1.9556022633706394e-06, + "loss": 0.88566184, + "num_input_tokens_seen": 93707285, + "step": 4340, + "time_per_iteration": 3.677839756011963 + }, + { + "auxiliary_loss_clip": 0.01125209, + "auxiliary_loss_mlp": 0.01031072, + "balance_loss_clip": 0.9760164, + "balance_loss_mlp": 1.02275681, + "epoch": 0.5219743882642939, + "flos": 23951663498880.0, + "grad_norm": 1.8923480288807617, + "language_loss": 0.79735041, + "learning_rate": 1.954823487068848e-06, + "loss": 0.81891322, + "num_input_tokens_seen": 93727495, + "step": 4341, + "time_per_iteration": 2.676717758178711 + }, + { + "auxiliary_loss_clip": 0.01160253, + "auxiliary_loss_mlp": 0.01027056, + "balance_loss_clip": 1.02132964, + "balance_loss_mlp": 1.01958776, + "epoch": 0.5220946311549329, + "flos": 28799280213120.0, + "grad_norm": 1.6241966901640135, + "language_loss": 0.81394345, + "learning_rate": 1.9540447176202976e-06, + "loss": 0.83581656, + "num_input_tokens_seen": 93748740, + "step": 4342, + "time_per_iteration": 2.767413854598999 + }, + { + "auxiliary_loss_clip": 0.01051416, + "auxiliary_loss_mlp": 0.01003363, + "balance_loss_clip": 0.9792859, + "balance_loss_mlp": 1.00201607, + "epoch": 0.5222148740455721, + "flos": 67189369017600.0, + "grad_norm": 0.8698042132085897, + "language_loss": 0.60677743, + "learning_rate": 1.9532659551431272e-06, + "loss": 0.62732518, + "num_input_tokens_seen": 93815770, + "step": 4343, + "time_per_iteration": 3.362051486968994 + }, + { + "auxiliary_loss_clip": 0.0115838, + "auxiliary_loss_mlp": 0.01028912, + "balance_loss_clip": 1.01550698, + "balance_loss_mlp": 1.02090669, + "epoch": 0.5223351169362112, + "flos": 61856164339200.0, + "grad_norm": 1.5478673238248175, + "language_loss": 0.6752193, + "learning_rate": 1.9524871997554744e-06, + "loss": 0.69709224, + "num_input_tokens_seen": 93843530, + "step": 4344, + "time_per_iteration": 3.002243757247925 + }, + { + "auxiliary_loss_clip": 0.01157634, + "auxiliary_loss_mlp": 0.01025486, + "balance_loss_clip": 1.01595926, + "balance_loss_mlp": 1.01723099, + "epoch": 0.5224553598268502, + "flos": 14647388676480.0, + "grad_norm": 2.662402248211012, + "language_loss": 0.806934, + "learning_rate": 1.951708451575475e-06, + "loss": 0.82876521, + "num_input_tokens_seen": 93860595, + "step": 4345, + "time_per_iteration": 2.626399517059326 + }, + { + "auxiliary_loss_clip": 0.01115358, + "auxiliary_loss_mlp": 0.01034487, + "balance_loss_clip": 0.93616676, + "balance_loss_mlp": 1.02602935, + "epoch": 0.5225756027174894, + "flos": 14826043946880.0, + "grad_norm": 1.8923233062372276, + "language_loss": 0.82092232, + "learning_rate": 1.9509297107212657e-06, + "loss": 0.84242076, + "num_input_tokens_seen": 93877365, + "step": 4346, + "time_per_iteration": 2.6439740657806396 + }, + { + "auxiliary_loss_clip": 0.01178904, + "auxiliary_loss_mlp": 0.01027589, + "balance_loss_clip": 1.05461264, + "balance_loss_mlp": 1.01994753, + "epoch": 0.5226958456081284, + "flos": 23512009029120.0, + "grad_norm": 6.538817698346683, + "language_loss": 0.79544485, + "learning_rate": 1.95015097731098e-06, + "loss": 0.81750983, + "num_input_tokens_seen": 93896855, + "step": 4347, + "time_per_iteration": 2.613826036453247 + }, + { + "auxiliary_loss_clip": 0.01179631, + "auxiliary_loss_mlp": 0.01029809, + "balance_loss_clip": 1.0549953, + "balance_loss_mlp": 1.02145815, + "epoch": 0.5228160884987675, + "flos": 19062928690560.0, + "grad_norm": 3.670788789383871, + "language_loss": 0.81960082, + "learning_rate": 1.949372251462751e-06, + "loss": 0.84169525, + "num_input_tokens_seen": 93914270, + "step": 4348, + "time_per_iteration": 2.6040446758270264 + }, + { + "auxiliary_loss_clip": 0.01109714, + "auxiliary_loss_mlp": 0.00509338, + "balance_loss_clip": 0.94002962, + "balance_loss_mlp": 0.0, + "epoch": 0.5229363313894067, + "flos": 21063224252160.0, + "grad_norm": 2.0313936360581324, + "language_loss": 0.83029068, + "learning_rate": 1.9485935332947124e-06, + "loss": 0.84648114, + "num_input_tokens_seen": 93932180, + "step": 4349, + "time_per_iteration": 2.7267255783081055 + }, + { + "auxiliary_loss_clip": 0.01127499, + "auxiliary_loss_mlp": 0.01024334, + "balance_loss_clip": 0.9775452, + "balance_loss_mlp": 1.01681197, + "epoch": 0.5230565742800457, + "flos": 14830389492480.0, + "grad_norm": 2.197513962587583, + "language_loss": 0.83200169, + "learning_rate": 1.947814822924993e-06, + "loss": 0.85351998, + "num_input_tokens_seen": 93949690, + "step": 4350, + "time_per_iteration": 2.608577013015747 + }, + { + "auxiliary_loss_clip": 0.0118195, + "auxiliary_loss_mlp": 0.01028532, + "balance_loss_clip": 1.05788469, + "balance_loss_mlp": 1.02080703, + "epoch": 0.5231768171706848, + "flos": 25813021253760.0, + "grad_norm": 1.997465512182751, + "language_loss": 0.82980919, + "learning_rate": 1.9470361204717236e-06, + "loss": 0.85191405, + "num_input_tokens_seen": 93968830, + "step": 4351, + "time_per_iteration": 2.647904396057129 + }, + { + "auxiliary_loss_clip": 0.01108245, + "auxiliary_loss_mlp": 0.0051014, + "balance_loss_clip": 0.93593609, + "balance_loss_mlp": 0.0, + "epoch": 0.5232970600613239, + "flos": 22743807834240.0, + "grad_norm": 2.3114752902857294, + "language_loss": 0.80929863, + "learning_rate": 1.9462574260530326e-06, + "loss": 0.82548249, + "num_input_tokens_seen": 93989110, + "step": 4352, + "time_per_iteration": 2.7008652687072754 + }, + { + "auxiliary_loss_clip": 0.01144968, + "auxiliary_loss_mlp": 0.01031394, + "balance_loss_clip": 1.01152658, + "balance_loss_mlp": 1.02331209, + "epoch": 0.523417302951963, + "flos": 17310703432320.0, + "grad_norm": 1.7356825901110329, + "language_loss": 0.812491, + "learning_rate": 1.9454787397870472e-06, + "loss": 0.83425456, + "num_input_tokens_seen": 94006430, + "step": 4353, + "time_per_iteration": 2.6457743644714355 + }, + { + "auxiliary_loss_clip": 0.01057061, + "auxiliary_loss_mlp": 0.01028125, + "balance_loss_clip": 0.8648181, + "balance_loss_mlp": 1.02020359, + "epoch": 0.523537545842602, + "flos": 18551740285440.0, + "grad_norm": 1.846562116606502, + "language_loss": 0.72071993, + "learning_rate": 1.944700061791894e-06, + "loss": 0.74157172, + "num_input_tokens_seen": 94024825, + "step": 4354, + "time_per_iteration": 2.7706403732299805 + }, + { + "auxiliary_loss_clip": 0.01153555, + "auxiliary_loss_mlp": 0.0102963, + "balance_loss_clip": 1.01508713, + "balance_loss_mlp": 1.02180362, + "epoch": 0.5236577887332411, + "flos": 19719267955200.0, + "grad_norm": 2.712337326241669, + "language_loss": 0.65255296, + "learning_rate": 1.943921392185698e-06, + "loss": 0.67438477, + "num_input_tokens_seen": 94043450, + "step": 4355, + "time_per_iteration": 3.595790147781372 + }, + { + "auxiliary_loss_clip": 0.01135736, + "auxiliary_loss_mlp": 0.01030329, + "balance_loss_clip": 0.97662711, + "balance_loss_mlp": 1.02258658, + "epoch": 0.5237780316238803, + "flos": 23550218121600.0, + "grad_norm": 1.9499612939028363, + "language_loss": 0.77514029, + "learning_rate": 1.9431427310865814e-06, + "loss": 0.79680097, + "num_input_tokens_seen": 94063055, + "step": 4356, + "time_per_iteration": 3.6642894744873047 + }, + { + "auxiliary_loss_clip": 0.01095135, + "auxiliary_loss_mlp": 0.01030941, + "balance_loss_clip": 0.93992949, + "balance_loss_mlp": 1.02299571, + "epoch": 0.5238982745145193, + "flos": 22491894775680.0, + "grad_norm": 2.010499075898618, + "language_loss": 0.78187525, + "learning_rate": 1.942364078612667e-06, + "loss": 0.80313599, + "num_input_tokens_seen": 94081785, + "step": 4357, + "time_per_iteration": 2.7515671253204346 + }, + { + "auxiliary_loss_clip": 0.01114768, + "auxiliary_loss_mlp": 0.01029193, + "balance_loss_clip": 0.93773115, + "balance_loss_mlp": 1.02137852, + "epoch": 0.5240185174051584, + "flos": 27088927234560.0, + "grad_norm": 1.929727742295738, + "language_loss": 0.75721693, + "learning_rate": 1.9415854348820765e-06, + "loss": 0.77865648, + "num_input_tokens_seen": 94101635, + "step": 4358, + "time_per_iteration": 3.6722450256347656 + }, + { + "auxiliary_loss_clip": 0.01161213, + "auxiliary_loss_mlp": 0.01022906, + "balance_loss_clip": 1.01517832, + "balance_loss_mlp": 1.01502013, + "epoch": 0.5241387602957975, + "flos": 22674680110080.0, + "grad_norm": 3.951983389459086, + "language_loss": 0.68163574, + "learning_rate": 1.940806800012929e-06, + "loss": 0.70347691, + "num_input_tokens_seen": 94121705, + "step": 4359, + "time_per_iteration": 2.636817216873169 + }, + { + "auxiliary_loss_clip": 0.01081961, + "auxiliary_loss_mlp": 0.00510065, + "balance_loss_clip": 0.90129501, + "balance_loss_mlp": 0.0, + "epoch": 0.5242590031864366, + "flos": 40553453134080.0, + "grad_norm": 1.454265041121604, + "language_loss": 0.6342957, + "learning_rate": 1.9400281741233432e-06, + "loss": 0.65021598, + "num_input_tokens_seen": 94146595, + "step": 4360, + "time_per_iteration": 2.8907370567321777 + }, + { + "auxiliary_loss_clip": 0.01014282, + "auxiliary_loss_mlp": 0.01005596, + "balance_loss_clip": 0.91043758, + "balance_loss_mlp": 1.00436234, + "epoch": 0.5243792460770756, + "flos": 66676313105280.0, + "grad_norm": 0.6557374520320626, + "language_loss": 0.52532619, + "learning_rate": 1.939249557331435e-06, + "loss": 0.54552495, + "num_input_tokens_seen": 94212410, + "step": 4361, + "time_per_iteration": 3.2882943153381348 + }, + { + "auxiliary_loss_clip": 0.01116834, + "auxiliary_loss_mlp": 0.0102544, + "balance_loss_clip": 0.93931186, + "balance_loss_mlp": 1.0180881, + "epoch": 0.5244994889677148, + "flos": 28183663992960.0, + "grad_norm": 1.7746791635518813, + "language_loss": 0.73236865, + "learning_rate": 1.938470949755321e-06, + "loss": 0.75379145, + "num_input_tokens_seen": 94232290, + "step": 4362, + "time_per_iteration": 2.748748302459717 + }, + { + "auxiliary_loss_clip": 0.01015238, + "auxiliary_loss_mlp": 0.01004139, + "balance_loss_clip": 0.90718246, + "balance_loss_mlp": 1.00283992, + "epoch": 0.5246197318583539, + "flos": 65950379239680.0, + "grad_norm": 0.8043285679264356, + "language_loss": 0.55660844, + "learning_rate": 1.937692351513115e-06, + "loss": 0.57680225, + "num_input_tokens_seen": 94291285, + "step": 4363, + "time_per_iteration": 3.2108142375946045 + }, + { + "auxiliary_loss_clip": 0.01159033, + "auxiliary_loss_mlp": 0.01027764, + "balance_loss_clip": 1.01443315, + "balance_loss_mlp": 1.01985478, + "epoch": 0.5247399747489929, + "flos": 21033490769280.0, + "grad_norm": 1.6882292330341466, + "language_loss": 0.8062892, + "learning_rate": 1.9369137627229297e-06, + "loss": 0.82815719, + "num_input_tokens_seen": 94309685, + "step": 4364, + "time_per_iteration": 2.65409779548645 + }, + { + "auxiliary_loss_clip": 0.01154225, + "auxiliary_loss_mlp": 0.01030786, + "balance_loss_clip": 1.01604581, + "balance_loss_mlp": 1.02250147, + "epoch": 0.5248602176396321, + "flos": 19025940660480.0, + "grad_norm": 2.1522969435745414, + "language_loss": 0.88524401, + "learning_rate": 1.936135183502877e-06, + "loss": 0.90709412, + "num_input_tokens_seen": 94326985, + "step": 4365, + "time_per_iteration": 2.6378297805786133 + }, + { + "auxiliary_loss_clip": 0.01111753, + "auxiliary_loss_mlp": 0.01026439, + "balance_loss_clip": 0.93823379, + "balance_loss_mlp": 1.0188725, + "epoch": 0.5249804605302711, + "flos": 22200084685440.0, + "grad_norm": 2.0689052126647427, + "language_loss": 0.81005174, + "learning_rate": 1.935356613971066e-06, + "loss": 0.83143365, + "num_input_tokens_seen": 94347645, + "step": 4366, + "time_per_iteration": 3.646693229675293 + }, + { + "auxiliary_loss_clip": 0.01131325, + "auxiliary_loss_mlp": 0.00510108, + "balance_loss_clip": 0.97818637, + "balance_loss_mlp": 0.0, + "epoch": 0.5251007034209102, + "flos": 23805686626560.0, + "grad_norm": 2.1381550935748708, + "language_loss": 0.76798892, + "learning_rate": 1.9345780542456047e-06, + "loss": 0.7844032, + "num_input_tokens_seen": 94367020, + "step": 4367, + "time_per_iteration": 2.7578253746032715 + }, + { + "auxiliary_loss_clip": 0.01144489, + "auxiliary_loss_mlp": 0.01030799, + "balance_loss_clip": 1.01259542, + "balance_loss_mlp": 1.02306211, + "epoch": 0.5252209463115494, + "flos": 23294605962240.0, + "grad_norm": 1.9951213726246988, + "language_loss": 0.71785146, + "learning_rate": 1.9337995044446007e-06, + "loss": 0.73960435, + "num_input_tokens_seen": 94385860, + "step": 4368, + "time_per_iteration": 2.7076733112335205 + }, + { + "auxiliary_loss_clip": 0.01159336, + "auxiliary_loss_mlp": 0.01026925, + "balance_loss_clip": 1.01539636, + "balance_loss_mlp": 1.01914024, + "epoch": 0.5253411892021884, + "flos": 19828687760640.0, + "grad_norm": 2.236305502791564, + "language_loss": 0.80024672, + "learning_rate": 1.9330209646861596e-06, + "loss": 0.82210934, + "num_input_tokens_seen": 94405010, + "step": 4369, + "time_per_iteration": 2.6464757919311523 + }, + { + "auxiliary_loss_clip": 0.01131369, + "auxiliary_loss_mlp": 0.01027882, + "balance_loss_clip": 0.9779135, + "balance_loss_mlp": 1.02040744, + "epoch": 0.5254614320928275, + "flos": 24133730561280.0, + "grad_norm": 21.474052258797556, + "language_loss": 0.77731729, + "learning_rate": 1.9322424350883843e-06, + "loss": 0.79890984, + "num_input_tokens_seen": 94426845, + "step": 4370, + "time_per_iteration": 2.7086849212646484 + }, + { + "auxiliary_loss_clip": 0.0113097, + "auxiliary_loss_mlp": 0.01030809, + "balance_loss_clip": 0.97486782, + "balance_loss_mlp": 1.02351892, + "epoch": 0.5255816749834666, + "flos": 24644954880000.0, + "grad_norm": 1.9078447922079869, + "language_loss": 0.78869802, + "learning_rate": 1.931463915769379e-06, + "loss": 0.81031579, + "num_input_tokens_seen": 94446960, + "step": 4371, + "time_per_iteration": 2.7018861770629883 + }, + { + "auxiliary_loss_clip": 0.0107964, + "auxiliary_loss_mlp": 0.0102536, + "balance_loss_clip": 0.89738822, + "balance_loss_mlp": 1.0172534, + "epoch": 0.5257019178741057, + "flos": 14136595320960.0, + "grad_norm": 2.187154328099801, + "language_loss": 0.73963058, + "learning_rate": 1.930685406847242e-06, + "loss": 0.76068056, + "num_input_tokens_seen": 94461535, + "step": 4372, + "time_per_iteration": 2.745818853378296 + }, + { + "auxiliary_loss_clip": 0.0113043, + "auxiliary_loss_mlp": 0.01029655, + "balance_loss_clip": 0.97740459, + "balance_loss_mlp": 1.02192998, + "epoch": 0.5258221607647448, + "flos": 23548961145600.0, + "grad_norm": 1.5410604644567274, + "language_loss": 0.81854451, + "learning_rate": 1.9299069084400734e-06, + "loss": 0.84014535, + "num_input_tokens_seen": 94482395, + "step": 4373, + "time_per_iteration": 2.756991147994995 + }, + { + "auxiliary_loss_clip": 0.01105866, + "auxiliary_loss_mlp": 0.01026675, + "balance_loss_clip": 0.94063973, + "balance_loss_mlp": 1.01833045, + "epoch": 0.5259424036553839, + "flos": 24966103403520.0, + "grad_norm": 2.177928509383953, + "language_loss": 0.69618905, + "learning_rate": 1.9291284206659717e-06, + "loss": 0.7175144, + "num_input_tokens_seen": 94500580, + "step": 4374, + "time_per_iteration": 2.784297466278076 + }, + { + "auxiliary_loss_clip": 0.01180883, + "auxiliary_loss_mlp": 0.0102702, + "balance_loss_clip": 1.05684471, + "balance_loss_mlp": 1.01916468, + "epoch": 0.526062646546023, + "flos": 28763908295040.0, + "grad_norm": 1.8392056471003713, + "language_loss": 0.71336818, + "learning_rate": 1.928349943643032e-06, + "loss": 0.73544723, + "num_input_tokens_seen": 94519680, + "step": 4375, + "time_per_iteration": 2.6426637172698975 + }, + { + "auxiliary_loss_clip": 0.01155297, + "auxiliary_loss_mlp": 0.01028013, + "balance_loss_clip": 1.02004373, + "balance_loss_mlp": 1.01965618, + "epoch": 0.526182889436662, + "flos": 22821375254400.0, + "grad_norm": 1.9241656637320235, + "language_loss": 0.82062203, + "learning_rate": 1.9275714774893493e-06, + "loss": 0.84245515, + "num_input_tokens_seen": 94539135, + "step": 4376, + "time_per_iteration": 2.581437349319458 + }, + { + "auxiliary_loss_clip": 0.01099471, + "auxiliary_loss_mlp": 0.01028216, + "balance_loss_clip": 0.93613601, + "balance_loss_mlp": 1.02012181, + "epoch": 0.5263031323273012, + "flos": 22929466256640.0, + "grad_norm": 2.3156495403617443, + "language_loss": 0.72830725, + "learning_rate": 1.9267930223230154e-06, + "loss": 0.74958408, + "num_input_tokens_seen": 94557610, + "step": 4377, + "time_per_iteration": 2.7236950397491455 + }, + { + "auxiliary_loss_clip": 0.01135297, + "auxiliary_loss_mlp": 0.01034902, + "balance_loss_clip": 0.9787516, + "balance_loss_mlp": 1.02717757, + "epoch": 0.5264233752179402, + "flos": 17748634049280.0, + "grad_norm": 2.2833941248684373, + "language_loss": 0.78710949, + "learning_rate": 1.9260145782621224e-06, + "loss": 0.80881149, + "num_input_tokens_seen": 94575390, + "step": 4378, + "time_per_iteration": 2.6390089988708496 + }, + { + "auxiliary_loss_clip": 0.01129327, + "auxiliary_loss_mlp": 0.01033083, + "balance_loss_clip": 0.97960734, + "balance_loss_mlp": 1.02567434, + "epoch": 0.5265436181085793, + "flos": 24421626069120.0, + "grad_norm": 1.9812766416208416, + "language_loss": 0.88054776, + "learning_rate": 1.925236145424758e-06, + "loss": 0.90217185, + "num_input_tokens_seen": 94594210, + "step": 4379, + "time_per_iteration": 2.8000006675720215 + }, + { + "auxiliary_loss_clip": 0.01057492, + "auxiliary_loss_mlp": 0.01000535, + "balance_loss_clip": 0.98175275, + "balance_loss_mlp": 0.99919939, + "epoch": 0.5266638609992185, + "flos": 69207298156800.0, + "grad_norm": 0.6951432231681842, + "language_loss": 0.57589591, + "learning_rate": 1.924457723929012e-06, + "loss": 0.5964762, + "num_input_tokens_seen": 94665020, + "step": 4380, + "time_per_iteration": 3.3515825271606445 + }, + { + "auxiliary_loss_clip": 0.01156698, + "auxiliary_loss_mlp": 0.0102505, + "balance_loss_clip": 1.01642394, + "balance_loss_mlp": 1.01680028, + "epoch": 0.5267841038898575, + "flos": 20738699850240.0, + "grad_norm": 1.436668250120487, + "language_loss": 0.82498932, + "learning_rate": 1.9236793138929685e-06, + "loss": 0.84680682, + "num_input_tokens_seen": 94684290, + "step": 4381, + "time_per_iteration": 2.6623001098632812 + }, + { + "auxiliary_loss_clip": 0.01158116, + "auxiliary_loss_mlp": 0.01029313, + "balance_loss_clip": 1.01419544, + "balance_loss_mlp": 1.0212785, + "epoch": 0.5269043467804966, + "flos": 17234392988160.0, + "grad_norm": 2.0335411722525634, + "language_loss": 0.81425005, + "learning_rate": 1.9229009154347133e-06, + "loss": 0.83612436, + "num_input_tokens_seen": 94701880, + "step": 4382, + "time_per_iteration": 3.6330435276031494 + }, + { + "auxiliary_loss_clip": 0.01067658, + "auxiliary_loss_mlp": 0.00509576, + "balance_loss_clip": 0.89469373, + "balance_loss_mlp": 0.0, + "epoch": 0.5270245896711357, + "flos": 18223157646720.0, + "grad_norm": 2.087155291682326, + "language_loss": 0.80768824, + "learning_rate": 1.922122528672327e-06, + "loss": 0.82346052, + "num_input_tokens_seen": 94720545, + "step": 4383, + "time_per_iteration": 2.831118583679199 + }, + { + "auxiliary_loss_clip": 0.01174628, + "auxiliary_loss_mlp": 0.01026871, + "balance_loss_clip": 1.05280745, + "balance_loss_mlp": 1.01947451, + "epoch": 0.5271448325617748, + "flos": 21287558643840.0, + "grad_norm": 2.7395105199696044, + "language_loss": 0.7898562, + "learning_rate": 1.9213441537238914e-06, + "loss": 0.81187117, + "num_input_tokens_seen": 94737420, + "step": 4384, + "time_per_iteration": 3.538973331451416 + }, + { + "auxiliary_loss_clip": 0.00991296, + "auxiliary_loss_mlp": 0.01006063, + "balance_loss_clip": 0.87373757, + "balance_loss_mlp": 1.0047282, + "epoch": 0.5272650754524139, + "flos": 65495497403520.0, + "grad_norm": 0.8302742077812465, + "language_loss": 0.57357192, + "learning_rate": 1.920565790707485e-06, + "loss": 0.59354556, + "num_input_tokens_seen": 94802810, + "step": 4385, + "time_per_iteration": 3.441314458847046 + }, + { + "auxiliary_loss_clip": 0.01091741, + "auxiliary_loss_mlp": 0.01032701, + "balance_loss_clip": 0.89938295, + "balance_loss_mlp": 1.02475023, + "epoch": 0.527385318343053, + "flos": 19676426008320.0, + "grad_norm": 2.854483507091906, + "language_loss": 0.66419208, + "learning_rate": 1.9197874397411853e-06, + "loss": 0.68543649, + "num_input_tokens_seen": 94819440, + "step": 4386, + "time_per_iteration": 2.6645615100860596 + }, + { + "auxiliary_loss_clip": 0.01104719, + "auxiliary_loss_mlp": 0.01029247, + "balance_loss_clip": 0.93308568, + "balance_loss_mlp": 1.02136779, + "epoch": 0.5275055612336921, + "flos": 12712018947840.0, + "grad_norm": 4.353240245794308, + "language_loss": 0.66912967, + "learning_rate": 1.919009100943067e-06, + "loss": 0.69046926, + "num_input_tokens_seen": 94835130, + "step": 4387, + "time_per_iteration": 2.684131383895874 + }, + { + "auxiliary_loss_clip": 0.01096012, + "auxiliary_loss_mlp": 0.01030838, + "balance_loss_clip": 0.90307212, + "balance_loss_mlp": 1.02299356, + "epoch": 0.5276258041243311, + "flos": 17749029098880.0, + "grad_norm": 3.7531587597373166, + "language_loss": 0.65178913, + "learning_rate": 1.9182307744312043e-06, + "loss": 0.67305762, + "num_input_tokens_seen": 94852235, + "step": 4388, + "time_per_iteration": 2.695230007171631 + }, + { + "auxiliary_loss_clip": 0.01134598, + "auxiliary_loss_mlp": 0.01027412, + "balance_loss_clip": 0.97647369, + "balance_loss_mlp": 1.01997924, + "epoch": 0.5277460470149702, + "flos": 22710447077760.0, + "grad_norm": 2.0661173720509662, + "language_loss": 0.76926321, + "learning_rate": 1.9174524603236676e-06, + "loss": 0.79088336, + "num_input_tokens_seen": 94871185, + "step": 4389, + "time_per_iteration": 2.7118711471557617 + }, + { + "auxiliary_loss_clip": 0.01132463, + "auxiliary_loss_mlp": 0.01030828, + "balance_loss_clip": 0.97751343, + "balance_loss_mlp": 1.02259624, + "epoch": 0.5278662899056094, + "flos": 19902699734400.0, + "grad_norm": 2.0405950252258815, + "language_loss": 0.76389778, + "learning_rate": 1.916674158738527e-06, + "loss": 0.78553075, + "num_input_tokens_seen": 94890090, + "step": 4390, + "time_per_iteration": 2.7101242542266846 + }, + { + "auxiliary_loss_clip": 0.0110245, + "auxiliary_loss_mlp": 0.00510147, + "balance_loss_clip": 0.93902957, + "balance_loss_mlp": 0.0, + "epoch": 0.5279865327962484, + "flos": 18005215875840.0, + "grad_norm": 1.9359126991607258, + "language_loss": 0.60429806, + "learning_rate": 1.9158958697938506e-06, + "loss": 0.62042397, + "num_input_tokens_seen": 94908470, + "step": 4391, + "time_per_iteration": 3.634042263031006 + }, + { + "auxiliary_loss_clip": 0.01128035, + "auxiliary_loss_mlp": 0.01028327, + "balance_loss_clip": 0.97701555, + "balance_loss_mlp": 1.02021503, + "epoch": 0.5281067756868875, + "flos": 15924443892480.0, + "grad_norm": 2.6177769123940604, + "language_loss": 0.85955501, + "learning_rate": 1.9151175936077032e-06, + "loss": 0.8811186, + "num_input_tokens_seen": 94923440, + "step": 4392, + "time_per_iteration": 2.656942367553711 + }, + { + "auxiliary_loss_clip": 0.01148541, + "auxiliary_loss_mlp": 0.01023467, + "balance_loss_clip": 1.0145483, + "balance_loss_mlp": 1.01605177, + "epoch": 0.5282270185775266, + "flos": 19426488197760.0, + "grad_norm": 1.7603030482880464, + "language_loss": 0.79395509, + "learning_rate": 1.9143393302981507e-06, + "loss": 0.81567514, + "num_input_tokens_seen": 94941125, + "step": 4393, + "time_per_iteration": 2.624399185180664 + }, + { + "auxiliary_loss_clip": 0.01134491, + "auxiliary_loss_mlp": 0.01032515, + "balance_loss_clip": 0.97701442, + "balance_loss_mlp": 1.02476346, + "epoch": 0.5283472614681657, + "flos": 16399613934720.0, + "grad_norm": 3.4761678565027485, + "language_loss": 0.83486611, + "learning_rate": 1.913561079983252e-06, + "loss": 0.85653615, + "num_input_tokens_seen": 94959950, + "step": 4394, + "time_per_iteration": 2.6693978309631348 + }, + { + "auxiliary_loss_clip": 0.01134711, + "auxiliary_loss_mlp": 0.0102631, + "balance_loss_clip": 0.97473001, + "balance_loss_mlp": 1.01773882, + "epoch": 0.5284675043588047, + "flos": 26760524163840.0, + "grad_norm": 2.1740051388018644, + "language_loss": 0.74890351, + "learning_rate": 1.9127828427810693e-06, + "loss": 0.77051365, + "num_input_tokens_seen": 94980515, + "step": 4395, + "time_per_iteration": 2.6708590984344482 + }, + { + "auxiliary_loss_clip": 0.01112391, + "auxiliary_loss_mlp": 0.01030119, + "balance_loss_clip": 0.93597424, + "balance_loss_mlp": 1.02294278, + "epoch": 0.5285877472494439, + "flos": 19899898473600.0, + "grad_norm": 2.1770006440392824, + "language_loss": 0.8106603, + "learning_rate": 1.9120046188096607e-06, + "loss": 0.83208543, + "num_input_tokens_seen": 94998560, + "step": 4396, + "time_per_iteration": 2.705698013305664 + }, + { + "auxiliary_loss_clip": 0.01132526, + "auxiliary_loss_mlp": 0.01023819, + "balance_loss_clip": 0.98158985, + "balance_loss_mlp": 1.01618409, + "epoch": 0.528707990140083, + "flos": 20011257613440.0, + "grad_norm": 1.9118467490267428, + "language_loss": 0.74156314, + "learning_rate": 1.9112264081870804e-06, + "loss": 0.76312661, + "num_input_tokens_seen": 95016950, + "step": 4397, + "time_per_iteration": 2.6920435428619385 + }, + { + "auxiliary_loss_clip": 0.01108195, + "auxiliary_loss_mlp": 0.01027535, + "balance_loss_clip": 0.94330817, + "balance_loss_mlp": 1.01922059, + "epoch": 0.528828233030722, + "flos": 20667956014080.0, + "grad_norm": 2.0061389724495586, + "language_loss": 0.75624543, + "learning_rate": 1.9104482110313843e-06, + "loss": 0.77760267, + "num_input_tokens_seen": 95036540, + "step": 4398, + "time_per_iteration": 2.7482783794403076 + }, + { + "auxiliary_loss_clip": 0.0115353, + "auxiliary_loss_mlp": 0.01032288, + "balance_loss_clip": 1.01551712, + "balance_loss_mlp": 1.02440262, + "epoch": 0.5289484759213612, + "flos": 25192448956800.0, + "grad_norm": 2.0219389774885235, + "language_loss": 0.74572885, + "learning_rate": 1.909670027460623e-06, + "loss": 0.76758695, + "num_input_tokens_seen": 95053840, + "step": 4399, + "time_per_iteration": 2.639680862426758 + }, + { + "auxiliary_loss_clip": 0.01156415, + "auxiliary_loss_mlp": 0.01030443, + "balance_loss_clip": 1.017012, + "balance_loss_mlp": 1.02264118, + "epoch": 0.5290687188120002, + "flos": 31139255715840.0, + "grad_norm": 1.7549439295644749, + "language_loss": 0.71624404, + "learning_rate": 1.908891857592847e-06, + "loss": 0.73811257, + "num_input_tokens_seen": 95074910, + "step": 4400, + "time_per_iteration": 2.6979615688323975 + }, + { + "auxiliary_loss_clip": 0.01102737, + "auxiliary_loss_mlp": 0.01028784, + "balance_loss_clip": 0.94217241, + "balance_loss_mlp": 1.02100587, + "epoch": 0.5291889617026393, + "flos": 20119851406080.0, + "grad_norm": 2.8275909055751764, + "language_loss": 0.90203857, + "learning_rate": 1.9081137015461034e-06, + "loss": 0.92335379, + "num_input_tokens_seen": 95090985, + "step": 4401, + "time_per_iteration": 2.7135837078094482 + }, + { + "auxiliary_loss_clip": 0.01074164, + "auxiliary_loss_mlp": 0.01026469, + "balance_loss_clip": 0.90004969, + "balance_loss_mlp": 1.01882768, + "epoch": 0.5293092045932785, + "flos": 19643747610240.0, + "grad_norm": 2.491453529917083, + "language_loss": 0.90475357, + "learning_rate": 1.9073355594384383e-06, + "loss": 0.92575991, + "num_input_tokens_seen": 95109225, + "step": 4402, + "time_per_iteration": 2.679380416870117 + }, + { + "auxiliary_loss_clip": 0.01102529, + "auxiliary_loss_mlp": 0.01030969, + "balance_loss_clip": 0.93949234, + "balance_loss_mlp": 1.02249348, + "epoch": 0.5294294474839175, + "flos": 24317736958080.0, + "grad_norm": 1.8231622689756086, + "language_loss": 0.8076371, + "learning_rate": 1.906557431387895e-06, + "loss": 0.8289721, + "num_input_tokens_seen": 95128215, + "step": 4403, + "time_per_iteration": 2.747182607650757 + }, + { + "auxiliary_loss_clip": 0.01105477, + "auxiliary_loss_mlp": 0.0102693, + "balance_loss_clip": 0.94369715, + "balance_loss_mlp": 1.0192554, + "epoch": 0.5295496903745566, + "flos": 18875941464960.0, + "grad_norm": 1.8927122428744132, + "language_loss": 0.78876215, + "learning_rate": 1.905779317512516e-06, + "loss": 0.81008619, + "num_input_tokens_seen": 95145760, + "step": 4404, + "time_per_iteration": 2.6794793605804443 + }, + { + "auxiliary_loss_clip": 0.01152708, + "auxiliary_loss_mlp": 0.01028357, + "balance_loss_clip": 1.01550126, + "balance_loss_mlp": 1.02070355, + "epoch": 0.5296699332651957, + "flos": 20923101296640.0, + "grad_norm": 2.0107479879873615, + "language_loss": 0.80279684, + "learning_rate": 1.9050012179303385e-06, + "loss": 0.82460749, + "num_input_tokens_seen": 95164270, + "step": 4405, + "time_per_iteration": 2.6479904651641846 + }, + { + "auxiliary_loss_clip": 0.01153855, + "auxiliary_loss_mlp": 0.01028062, + "balance_loss_clip": 1.01318002, + "balance_loss_mlp": 1.02045691, + "epoch": 0.5297901761558348, + "flos": 22046745525120.0, + "grad_norm": 2.320124215393011, + "language_loss": 0.68777287, + "learning_rate": 1.904223132759401e-06, + "loss": 0.70959198, + "num_input_tokens_seen": 95182870, + "step": 4406, + "time_per_iteration": 2.652587652206421 + }, + { + "auxiliary_loss_clip": 0.01153637, + "auxiliary_loss_mlp": 0.0102891, + "balance_loss_clip": 1.01435745, + "balance_loss_mlp": 1.02087569, + "epoch": 0.5299104190464738, + "flos": 21798495653760.0, + "grad_norm": 3.1749930228276764, + "language_loss": 0.69359934, + "learning_rate": 1.9034450621177383e-06, + "loss": 0.71542478, + "num_input_tokens_seen": 95201190, + "step": 4407, + "time_per_iteration": 2.6493167877197266 + }, + { + "auxiliary_loss_clip": 0.01153633, + "auxiliary_loss_mlp": 0.01034934, + "balance_loss_clip": 1.0166223, + "balance_loss_mlp": 1.02697968, + "epoch": 0.530030661937113, + "flos": 14720790119040.0, + "grad_norm": 1.8047831327614683, + "language_loss": 0.69704109, + "learning_rate": 1.9026670061233824e-06, + "loss": 0.71892679, + "num_input_tokens_seen": 95218625, + "step": 4408, + "time_per_iteration": 4.429666996002197 + }, + { + "auxiliary_loss_clip": 0.01130277, + "auxiliary_loss_mlp": 0.01031811, + "balance_loss_clip": 0.98015702, + "balance_loss_mlp": 1.02420211, + "epoch": 0.5301509048277521, + "flos": 21251504367360.0, + "grad_norm": 1.6936652357440298, + "language_loss": 0.80472469, + "learning_rate": 1.901888964894365e-06, + "loss": 0.82634556, + "num_input_tokens_seen": 95237665, + "step": 4409, + "time_per_iteration": 2.6364102363586426 + }, + { + "auxiliary_loss_clip": 0.01177695, + "auxiliary_loss_mlp": 0.01031205, + "balance_loss_clip": 1.05258834, + "balance_loss_mlp": 1.02270508, + "epoch": 0.5302711477183911, + "flos": 25957058791680.0, + "grad_norm": 2.2617026776843123, + "language_loss": 0.67657536, + "learning_rate": 1.9011109385487134e-06, + "loss": 0.69866437, + "num_input_tokens_seen": 95258915, + "step": 4410, + "time_per_iteration": 3.6204826831817627 + }, + { + "auxiliary_loss_clip": 0.01177847, + "auxiliary_loss_mlp": 0.0102566, + "balance_loss_clip": 1.05283153, + "balance_loss_mlp": 1.01793242, + "epoch": 0.5303913906090303, + "flos": 22273126992000.0, + "grad_norm": 2.2698444835513767, + "language_loss": 0.66469848, + "learning_rate": 1.900332927204454e-06, + "loss": 0.6867336, + "num_input_tokens_seen": 95277365, + "step": 4411, + "time_per_iteration": 2.5954887866973877 + }, + { + "auxiliary_loss_clip": 0.01134784, + "auxiliary_loss_mlp": 0.01027885, + "balance_loss_clip": 0.97524852, + "balance_loss_mlp": 1.01995802, + "epoch": 0.5305116334996693, + "flos": 24936010784640.0, + "grad_norm": 2.157345176209904, + "language_loss": 0.7684437, + "learning_rate": 1.8995549309796097e-06, + "loss": 0.79007041, + "num_input_tokens_seen": 95296670, + "step": 4412, + "time_per_iteration": 2.751995086669922 + }, + { + "auxiliary_loss_clip": 0.01160407, + "auxiliary_loss_mlp": 0.01024983, + "balance_loss_clip": 1.01641655, + "balance_loss_mlp": 1.01721621, + "epoch": 0.5306318763903084, + "flos": 20189338266240.0, + "grad_norm": 2.1501337878868427, + "language_loss": 0.76294577, + "learning_rate": 1.8987769499922028e-06, + "loss": 0.7847997, + "num_input_tokens_seen": 95315640, + "step": 4413, + "time_per_iteration": 2.606807231903076 + }, + { + "auxiliary_loss_clip": 0.01155181, + "auxiliary_loss_mlp": 0.00509729, + "balance_loss_clip": 1.01737094, + "balance_loss_mlp": 0.0, + "epoch": 0.5307521192809476, + "flos": 20266366982400.0, + "grad_norm": 2.170246308497056, + "language_loss": 0.70588434, + "learning_rate": 1.897998984360252e-06, + "loss": 0.72253346, + "num_input_tokens_seen": 95334610, + "step": 4414, + "time_per_iteration": 2.6460328102111816 + }, + { + "auxiliary_loss_clip": 0.01126542, + "auxiliary_loss_mlp": 0.01028005, + "balance_loss_clip": 0.97610176, + "balance_loss_mlp": 1.02066731, + "epoch": 0.5308723621715866, + "flos": 28844276976000.0, + "grad_norm": 1.494797973194114, + "language_loss": 0.78405178, + "learning_rate": 1.897221034201775e-06, + "loss": 0.80559731, + "num_input_tokens_seen": 95358350, + "step": 4415, + "time_per_iteration": 2.7420971393585205 + }, + { + "auxiliary_loss_clip": 0.01106992, + "auxiliary_loss_mlp": 0.01025925, + "balance_loss_clip": 0.9379493, + "balance_loss_mlp": 1.01871336, + "epoch": 0.5309926050622257, + "flos": 27457766040960.0, + "grad_norm": 1.681095640771079, + "language_loss": 0.66664529, + "learning_rate": 1.8964430996347842e-06, + "loss": 0.68797445, + "num_input_tokens_seen": 95379900, + "step": 4416, + "time_per_iteration": 2.7802042961120605 + }, + { + "auxiliary_loss_clip": 0.01126999, + "auxiliary_loss_mlp": 0.01024025, + "balance_loss_clip": 0.97495544, + "balance_loss_mlp": 1.01609159, + "epoch": 0.5311128479528648, + "flos": 20514545026560.0, + "grad_norm": 1.5419548327799306, + "language_loss": 0.82538128, + "learning_rate": 1.8956651807772931e-06, + "loss": 0.84689152, + "num_input_tokens_seen": 95397935, + "step": 4417, + "time_per_iteration": 2.664985179901123 + }, + { + "auxiliary_loss_clip": 0.01153482, + "auxiliary_loss_mlp": 0.0102738, + "balance_loss_clip": 1.01778483, + "balance_loss_mlp": 1.01984549, + "epoch": 0.5312330908435039, + "flos": 21397660807680.0, + "grad_norm": 1.7027712292696053, + "language_loss": 0.84164894, + "learning_rate": 1.8948872777473115e-06, + "loss": 0.8634575, + "num_input_tokens_seen": 95415890, + "step": 4418, + "time_per_iteration": 3.4921226501464844 + }, + { + "auxiliary_loss_clip": 0.01133692, + "auxiliary_loss_mlp": 0.01024497, + "balance_loss_clip": 0.97985393, + "balance_loss_mlp": 1.01648033, + "epoch": 0.531353333734143, + "flos": 24717350741760.0, + "grad_norm": 1.8043014267863244, + "language_loss": 0.63460326, + "learning_rate": 1.8941093906628458e-06, + "loss": 0.65618515, + "num_input_tokens_seen": 95433675, + "step": 4419, + "time_per_iteration": 2.718550682067871 + }, + { + "auxiliary_loss_clip": 0.01124988, + "auxiliary_loss_mlp": 0.0102711, + "balance_loss_clip": 0.97655761, + "balance_loss_mlp": 1.01993656, + "epoch": 0.531473576624782, + "flos": 30480689808000.0, + "grad_norm": 1.958439934864322, + "language_loss": 0.71047938, + "learning_rate": 1.893331519641902e-06, + "loss": 0.73200035, + "num_input_tokens_seen": 95455820, + "step": 4420, + "time_per_iteration": 2.694232940673828 + }, + { + "auxiliary_loss_clip": 0.01093008, + "auxiliary_loss_mlp": 0.01031348, + "balance_loss_clip": 0.93314409, + "balance_loss_mlp": 1.02324772, + "epoch": 0.5315938195154212, + "flos": 23002975440000.0, + "grad_norm": 2.4054948524636544, + "language_loss": 0.7367211, + "learning_rate": 1.8925536648024815e-06, + "loss": 0.75796467, + "num_input_tokens_seen": 95473240, + "step": 4421, + "time_per_iteration": 2.7013983726501465 + }, + { + "auxiliary_loss_clip": 0.01177352, + "auxiliary_loss_mlp": 0.01023054, + "balance_loss_clip": 1.05278563, + "balance_loss_mlp": 1.01512694, + "epoch": 0.5317140624060602, + "flos": 22748584343040.0, + "grad_norm": 2.132598696845823, + "language_loss": 0.75960016, + "learning_rate": 1.8917758262625849e-06, + "loss": 0.78160417, + "num_input_tokens_seen": 95493480, + "step": 4422, + "time_per_iteration": 2.603055238723755 + }, + { + "auxiliary_loss_clip": 0.01123729, + "auxiliary_loss_mlp": 0.01026538, + "balance_loss_clip": 0.9763031, + "balance_loss_mlp": 1.01931357, + "epoch": 0.5318343052966993, + "flos": 22821087945600.0, + "grad_norm": 1.8456343626958898, + "language_loss": 0.80900073, + "learning_rate": 1.8909980041402089e-06, + "loss": 0.8305034, + "num_input_tokens_seen": 95512075, + "step": 4423, + "time_per_iteration": 2.6583986282348633 + }, + { + "auxiliary_loss_clip": 0.01146922, + "auxiliary_loss_mlp": 0.01024988, + "balance_loss_clip": 1.01261497, + "balance_loss_mlp": 1.0173732, + "epoch": 0.5319545481873384, + "flos": 13626089274240.0, + "grad_norm": 2.1837326471588243, + "language_loss": 0.65648752, + "learning_rate": 1.8902201985533494e-06, + "loss": 0.67820656, + "num_input_tokens_seen": 95529340, + "step": 4424, + "time_per_iteration": 2.702385425567627 + }, + { + "auxiliary_loss_clip": 0.01130335, + "auxiliary_loss_mlp": 0.01024067, + "balance_loss_clip": 0.97741568, + "balance_loss_mlp": 1.0167743, + "epoch": 0.5320747910779775, + "flos": 22162522037760.0, + "grad_norm": 2.0991248867701873, + "language_loss": 0.74668598, + "learning_rate": 1.8894424096199983e-06, + "loss": 0.76822996, + "num_input_tokens_seen": 95548545, + "step": 4425, + "time_per_iteration": 2.755483865737915 + }, + { + "auxiliary_loss_clip": 0.0115141, + "auxiliary_loss_mlp": 0.01030272, + "balance_loss_clip": 1.01393926, + "balance_loss_mlp": 1.02188587, + "epoch": 0.5321950339686166, + "flos": 18588081870720.0, + "grad_norm": 2.0579613757741657, + "language_loss": 0.85841954, + "learning_rate": 1.8886646374581463e-06, + "loss": 0.88023639, + "num_input_tokens_seen": 95567770, + "step": 4426, + "time_per_iteration": 2.7321503162384033 + }, + { + "auxiliary_loss_clip": 0.01150249, + "auxiliary_loss_mlp": 0.01030683, + "balance_loss_clip": 1.01299965, + "balance_loss_mlp": 1.02240372, + "epoch": 0.5323152768592557, + "flos": 22856818999680.0, + "grad_norm": 2.2434201780004424, + "language_loss": 0.71019346, + "learning_rate": 1.8878868821857795e-06, + "loss": 0.73200274, + "num_input_tokens_seen": 95587420, + "step": 4427, + "time_per_iteration": 2.643108606338501 + }, + { + "auxiliary_loss_clip": 0.01079929, + "auxiliary_loss_mlp": 0.01028448, + "balance_loss_clip": 0.89855397, + "balance_loss_mlp": 1.02031159, + "epoch": 0.5324355197498948, + "flos": 33948690998400.0, + "grad_norm": 4.328806612988095, + "language_loss": 0.75483048, + "learning_rate": 1.8871091439208838e-06, + "loss": 0.77591419, + "num_input_tokens_seen": 95609030, + "step": 4428, + "time_per_iteration": 2.8398215770721436 + }, + { + "auxiliary_loss_clip": 0.01078336, + "auxiliary_loss_mlp": 0.01034618, + "balance_loss_clip": 0.89824283, + "balance_loss_mlp": 1.02604651, + "epoch": 0.5325557626405338, + "flos": 23256720092160.0, + "grad_norm": 2.0276802422280062, + "language_loss": 0.77099276, + "learning_rate": 1.8863314227814414e-06, + "loss": 0.79212224, + "num_input_tokens_seen": 95627340, + "step": 4429, + "time_per_iteration": 2.7527594566345215 + }, + { + "auxiliary_loss_clip": 0.01159328, + "auxiliary_loss_mlp": 0.0102638, + "balance_loss_clip": 1.01670742, + "balance_loss_mlp": 1.01881623, + "epoch": 0.532676005531173, + "flos": 26718687797760.0, + "grad_norm": 4.219274981398595, + "language_loss": 0.4905811, + "learning_rate": 1.8855537188854313e-06, + "loss": 0.51243818, + "num_input_tokens_seen": 95646315, + "step": 4430, + "time_per_iteration": 2.664607524871826 + }, + { + "auxiliary_loss_clip": 0.01157369, + "auxiliary_loss_mlp": 0.01028571, + "balance_loss_clip": 1.0148102, + "balance_loss_mlp": 1.02123368, + "epoch": 0.5327962484218121, + "flos": 17894610921600.0, + "grad_norm": 2.0280954438636227, + "language_loss": 0.7856214, + "learning_rate": 1.8847760323508315e-06, + "loss": 0.80748075, + "num_input_tokens_seen": 95665220, + "step": 4431, + "time_per_iteration": 2.789821147918701 + }, + { + "auxiliary_loss_clip": 0.01125282, + "auxiliary_loss_mlp": 0.0102675, + "balance_loss_clip": 0.97775388, + "balance_loss_mlp": 1.01940382, + "epoch": 0.5329164913124511, + "flos": 17925385898880.0, + "grad_norm": 1.7346483632584648, + "language_loss": 0.76091194, + "learning_rate": 1.883998363295616e-06, + "loss": 0.78243226, + "num_input_tokens_seen": 95682700, + "step": 4432, + "time_per_iteration": 2.682694911956787 + }, + { + "auxiliary_loss_clip": 0.01036957, + "auxiliary_loss_mlp": 0.0100377, + "balance_loss_clip": 0.94557637, + "balance_loss_mlp": 1.0025835, + "epoch": 0.5330367342030903, + "flos": 57254178781440.0, + "grad_norm": 0.8713828790174287, + "language_loss": 0.62586391, + "learning_rate": 1.8832207118377565e-06, + "loss": 0.64627111, + "num_input_tokens_seen": 95738070, + "step": 4433, + "time_per_iteration": 4.114145517349243 + }, + { + "auxiliary_loss_clip": 0.01174648, + "auxiliary_loss_mlp": 0.01027338, + "balance_loss_clip": 1.0526917, + "balance_loss_mlp": 1.02003312, + "epoch": 0.5331569770937293, + "flos": 17420518287360.0, + "grad_norm": 1.8463281524550899, + "language_loss": 0.69460887, + "learning_rate": 1.882443078095222e-06, + "loss": 0.71662873, + "num_input_tokens_seen": 95756950, + "step": 4434, + "time_per_iteration": 3.5142552852630615 + }, + { + "auxiliary_loss_clip": 0.01000005, + "auxiliary_loss_mlp": 0.01001881, + "balance_loss_clip": 0.87081081, + "balance_loss_mlp": 1.00057602, + "epoch": 0.5332772199843684, + "flos": 56750783627520.0, + "grad_norm": 0.865242127629926, + "language_loss": 0.66802537, + "learning_rate": 1.8816654621859794e-06, + "loss": 0.68804425, + "num_input_tokens_seen": 95816615, + "step": 4435, + "time_per_iteration": 3.2839343547821045 + }, + { + "auxiliary_loss_clip": 0.01174859, + "auxiliary_loss_mlp": 0.01024497, + "balance_loss_clip": 1.053684, + "balance_loss_mlp": 1.01694465, + "epoch": 0.5333974628750076, + "flos": 18697753071360.0, + "grad_norm": 2.1783151532313894, + "language_loss": 0.72607195, + "learning_rate": 1.8808878642279915e-06, + "loss": 0.74806559, + "num_input_tokens_seen": 95832020, + "step": 4436, + "time_per_iteration": 3.9856081008911133 + }, + { + "auxiliary_loss_clip": 0.01104226, + "auxiliary_loss_mlp": 0.01033091, + "balance_loss_clip": 0.93228143, + "balance_loss_mlp": 1.02427506, + "epoch": 0.5335177057656466, + "flos": 23805507058560.0, + "grad_norm": 2.6519964729679484, + "language_loss": 0.65498209, + "learning_rate": 1.8801102843392209e-06, + "loss": 0.67635524, + "num_input_tokens_seen": 95851425, + "step": 4437, + "time_per_iteration": 2.7519538402557373 + }, + { + "auxiliary_loss_clip": 0.01103885, + "auxiliary_loss_mlp": 0.01026914, + "balance_loss_clip": 0.93577158, + "balance_loss_mlp": 1.01964831, + "epoch": 0.5336379486562857, + "flos": 25078683605760.0, + "grad_norm": 1.8387198268591614, + "language_loss": 0.85297155, + "learning_rate": 1.8793327226376238e-06, + "loss": 0.87427956, + "num_input_tokens_seen": 95870745, + "step": 4438, + "time_per_iteration": 2.770247459411621 + }, + { + "auxiliary_loss_clip": 0.01139029, + "auxiliary_loss_mlp": 0.01033087, + "balance_loss_clip": 0.97764421, + "balance_loss_mlp": 1.0252068, + "epoch": 0.5337581915469248, + "flos": 21396691140480.0, + "grad_norm": 1.727351994565066, + "language_loss": 0.80260611, + "learning_rate": 1.8785551792411569e-06, + "loss": 0.82432723, + "num_input_tokens_seen": 95889755, + "step": 4439, + "time_per_iteration": 2.664849042892456 + }, + { + "auxiliary_loss_clip": 0.01132103, + "auxiliary_loss_mlp": 0.01024366, + "balance_loss_clip": 0.97833651, + "balance_loss_mlp": 1.01644444, + "epoch": 0.5338784344375639, + "flos": 14865905064960.0, + "grad_norm": 1.960870456874232, + "language_loss": 0.82551628, + "learning_rate": 1.8777776542677733e-06, + "loss": 0.84708095, + "num_input_tokens_seen": 95907805, + "step": 4440, + "time_per_iteration": 2.6614654064178467 + }, + { + "auxiliary_loss_clip": 0.0110158, + "auxiliary_loss_mlp": 0.0102742, + "balance_loss_clip": 0.93285835, + "balance_loss_mlp": 1.01948047, + "epoch": 0.5339986773282029, + "flos": 20813501923200.0, + "grad_norm": 1.7697360677268323, + "language_loss": 0.73186123, + "learning_rate": 1.8770001478354216e-06, + "loss": 0.7531513, + "num_input_tokens_seen": 95927480, + "step": 4441, + "time_per_iteration": 2.7793257236480713 + }, + { + "auxiliary_loss_clip": 0.01148329, + "auxiliary_loss_mlp": 0.01028688, + "balance_loss_clip": 1.01255751, + "balance_loss_mlp": 1.02048039, + "epoch": 0.5341189202188421, + "flos": 17969089772160.0, + "grad_norm": 2.3334855896856697, + "language_loss": 0.84339404, + "learning_rate": 1.8762226600620504e-06, + "loss": 0.86516416, + "num_input_tokens_seen": 95946095, + "step": 4442, + "time_per_iteration": 2.6416263580322266 + }, + { + "auxiliary_loss_clip": 0.01136284, + "auxiliary_loss_mlp": 0.01028615, + "balance_loss_clip": 0.97470301, + "balance_loss_mlp": 1.01988304, + "epoch": 0.5342391631094812, + "flos": 11031866328960.0, + "grad_norm": 3.0705990265132526, + "language_loss": 0.58917195, + "learning_rate": 1.8754451910656031e-06, + "loss": 0.61082089, + "num_input_tokens_seen": 95959995, + "step": 4443, + "time_per_iteration": 3.548025131225586 + }, + { + "auxiliary_loss_clip": 0.01088371, + "auxiliary_loss_mlp": 0.01024279, + "balance_loss_clip": 0.89974916, + "balance_loss_mlp": 1.01657248, + "epoch": 0.5343594060001202, + "flos": 15339135772800.0, + "grad_norm": 3.980232593974098, + "language_loss": 0.82590175, + "learning_rate": 1.8746677409640212e-06, + "loss": 0.84702832, + "num_input_tokens_seen": 95977095, + "step": 4444, + "time_per_iteration": 2.723071575164795 + }, + { + "auxiliary_loss_clip": 0.01156666, + "auxiliary_loss_mlp": 0.01032846, + "balance_loss_clip": 1.01636088, + "balance_loss_mlp": 1.02476954, + "epoch": 0.5344796488907594, + "flos": 26900898514560.0, + "grad_norm": 1.9767076033911681, + "language_loss": 0.8457458, + "learning_rate": 1.8738903098752432e-06, + "loss": 0.86764097, + "num_input_tokens_seen": 95996225, + "step": 4445, + "time_per_iteration": 2.729570150375366 + }, + { + "auxiliary_loss_clip": 0.01129396, + "auxiliary_loss_mlp": 0.01032923, + "balance_loss_clip": 0.97599059, + "balance_loss_mlp": 1.02496243, + "epoch": 0.5345998917813984, + "flos": 25411216740480.0, + "grad_norm": 2.036988346711938, + "language_loss": 0.73289907, + "learning_rate": 1.8731128979172052e-06, + "loss": 0.75452226, + "num_input_tokens_seen": 96015425, + "step": 4446, + "time_per_iteration": 2.710907220840454 + }, + { + "auxiliary_loss_clip": 0.01125805, + "auxiliary_loss_mlp": 0.01021945, + "balance_loss_clip": 0.97526479, + "balance_loss_mlp": 1.01420856, + "epoch": 0.5347201346720375, + "flos": 32853379622400.0, + "grad_norm": 2.2086769151656394, + "language_loss": 0.67102945, + "learning_rate": 1.8723355052078394e-06, + "loss": 0.69250697, + "num_input_tokens_seen": 96035460, + "step": 4447, + "time_per_iteration": 2.8437533378601074 + }, + { + "auxiliary_loss_clip": 0.01148595, + "auxiliary_loss_mlp": 0.01026843, + "balance_loss_clip": 1.01181197, + "balance_loss_mlp": 1.01825976, + "epoch": 0.5348403775626767, + "flos": 17967940536960.0, + "grad_norm": 2.180712638907371, + "language_loss": 0.78003728, + "learning_rate": 1.8715581318650765e-06, + "loss": 0.80179167, + "num_input_tokens_seen": 96054515, + "step": 4448, + "time_per_iteration": 2.6502561569213867 + }, + { + "auxiliary_loss_clip": 0.01116654, + "auxiliary_loss_mlp": 0.01033104, + "balance_loss_clip": 0.94053447, + "balance_loss_mlp": 1.02509344, + "epoch": 0.5349606204533157, + "flos": 17603339535360.0, + "grad_norm": 3.227115883448808, + "language_loss": 0.82239366, + "learning_rate": 1.8707807780068422e-06, + "loss": 0.84389114, + "num_input_tokens_seen": 96072330, + "step": 4449, + "time_per_iteration": 2.6686699390411377 + }, + { + "auxiliary_loss_clip": 0.01131262, + "auxiliary_loss_mlp": 0.01024639, + "balance_loss_clip": 0.97831601, + "balance_loss_mlp": 1.01721835, + "epoch": 0.5350808633439548, + "flos": 29167831710720.0, + "grad_norm": 2.1418970695429365, + "language_loss": 0.66238546, + "learning_rate": 1.8700034437510611e-06, + "loss": 0.68394452, + "num_input_tokens_seen": 96092425, + "step": 4450, + "time_per_iteration": 2.7691900730133057 + }, + { + "auxiliary_loss_clip": 0.0109788, + "auxiliary_loss_mlp": 0.01032014, + "balance_loss_clip": 0.93587416, + "balance_loss_mlp": 1.02446771, + "epoch": 0.5352011062345938, + "flos": 19499997381120.0, + "grad_norm": 2.035476889639126, + "language_loss": 0.8142128, + "learning_rate": 1.8692261292156549e-06, + "loss": 0.83551168, + "num_input_tokens_seen": 96111660, + "step": 4451, + "time_per_iteration": 2.7200193405151367 + }, + { + "auxiliary_loss_clip": 0.01173938, + "auxiliary_loss_mlp": 0.01031843, + "balance_loss_clip": 1.05338931, + "balance_loss_mlp": 1.02425504, + "epoch": 0.535321349125233, + "flos": 23477642691840.0, + "grad_norm": 1.8057040663673471, + "language_loss": 0.81293607, + "learning_rate": 1.8684488345185401e-06, + "loss": 0.83499384, + "num_input_tokens_seen": 96131835, + "step": 4452, + "time_per_iteration": 2.724518299102783 + }, + { + "auxiliary_loss_clip": 0.01178969, + "auxiliary_loss_mlp": 0.01033697, + "balance_loss_clip": 1.05548704, + "balance_loss_mlp": 1.02588892, + "epoch": 0.535441592015872, + "flos": 20478059786880.0, + "grad_norm": 2.0999509337411646, + "language_loss": 0.7838943, + "learning_rate": 1.8676715597776332e-06, + "loss": 0.80602103, + "num_input_tokens_seen": 96150180, + "step": 4453, + "time_per_iteration": 2.614501953125 + }, + { + "auxiliary_loss_clip": 0.01072818, + "auxiliary_loss_mlp": 0.01027415, + "balance_loss_clip": 0.89665329, + "balance_loss_mlp": 1.02033973, + "epoch": 0.5355618349065111, + "flos": 19573147428480.0, + "grad_norm": 2.048124405635308, + "language_loss": 0.7599628, + "learning_rate": 1.8668943051108455e-06, + "loss": 0.78096515, + "num_input_tokens_seen": 96167485, + "step": 4454, + "time_per_iteration": 2.8331515789031982 + }, + { + "auxiliary_loss_clip": 0.01125681, + "auxiliary_loss_mlp": 0.01031409, + "balance_loss_clip": 0.97410321, + "balance_loss_mlp": 1.02330279, + "epoch": 0.5356820777971503, + "flos": 24024633978240.0, + "grad_norm": 1.7092130666555745, + "language_loss": 0.75854504, + "learning_rate": 1.8661170706360856e-06, + "loss": 0.7801159, + "num_input_tokens_seen": 96186650, + "step": 4455, + "time_per_iteration": 2.7301697731018066 + }, + { + "auxiliary_loss_clip": 0.011518, + "auxiliary_loss_mlp": 0.01024855, + "balance_loss_clip": 1.01588154, + "balance_loss_mlp": 1.01750541, + "epoch": 0.5358023206877893, + "flos": 20884676722560.0, + "grad_norm": 2.298899067152522, + "language_loss": 0.81572831, + "learning_rate": 1.8653398564712594e-06, + "loss": 0.83749485, + "num_input_tokens_seen": 96205595, + "step": 4456, + "time_per_iteration": 2.6999847888946533 + }, + { + "auxiliary_loss_clip": 0.01151759, + "auxiliary_loss_mlp": 0.01028228, + "balance_loss_clip": 1.01688743, + "balance_loss_mlp": 1.02108109, + "epoch": 0.5359225635784284, + "flos": 22418996123520.0, + "grad_norm": 1.5289177967884222, + "language_loss": 0.8225596, + "learning_rate": 1.8645626627342704e-06, + "loss": 0.84435952, + "num_input_tokens_seen": 96226360, + "step": 4457, + "time_per_iteration": 2.681851387023926 + }, + { + "auxiliary_loss_clip": 0.0115592, + "auxiliary_loss_mlp": 0.0102829, + "balance_loss_clip": 1.01496649, + "balance_loss_mlp": 1.02088702, + "epoch": 0.5360428064690675, + "flos": 24097784025600.0, + "grad_norm": 2.103662331194892, + "language_loss": 0.80702209, + "learning_rate": 1.8637854895430172e-06, + "loss": 0.82886422, + "num_input_tokens_seen": 96245625, + "step": 4458, + "time_per_iteration": 2.7533249855041504 + }, + { + "auxiliary_loss_clip": 0.01099867, + "auxiliary_loss_mlp": 0.01027357, + "balance_loss_clip": 0.93799412, + "balance_loss_mlp": 1.01920342, + "epoch": 0.5361630493597066, + "flos": 21434505183360.0, + "grad_norm": 2.682783406823195, + "language_loss": 0.69392204, + "learning_rate": 1.8630083370153978e-06, + "loss": 0.71519423, + "num_input_tokens_seen": 96265265, + "step": 4459, + "time_per_iteration": 3.675795555114746 + }, + { + "auxiliary_loss_clip": 0.00978151, + "auxiliary_loss_mlp": 0.01008572, + "balance_loss_clip": 0.83424938, + "balance_loss_mlp": 1.00743341, + "epoch": 0.5362832922503457, + "flos": 68888696520960.0, + "grad_norm": 0.7406154900686646, + "language_loss": 0.55393505, + "learning_rate": 1.8622312052693041e-06, + "loss": 0.57380223, + "num_input_tokens_seen": 96326445, + "step": 4460, + "time_per_iteration": 4.33380389213562 + }, + { + "auxiliary_loss_clip": 0.01143465, + "auxiliary_loss_mlp": 0.01027221, + "balance_loss_clip": 1.01027703, + "balance_loss_mlp": 1.01999426, + "epoch": 0.5364035351409848, + "flos": 9793702563840.0, + "grad_norm": 2.17460178821487, + "language_loss": 0.71423119, + "learning_rate": 1.8614540944226267e-06, + "loss": 0.73593807, + "num_input_tokens_seen": 96343115, + "step": 4461, + "time_per_iteration": 2.6295316219329834 + }, + { + "auxiliary_loss_clip": 0.01126957, + "auxiliary_loss_mlp": 0.01026115, + "balance_loss_clip": 0.97952563, + "balance_loss_mlp": 1.01912057, + "epoch": 0.5365237780316239, + "flos": 23290080848640.0, + "grad_norm": 1.7886666962366704, + "language_loss": 0.68478942, + "learning_rate": 1.8606770045932537e-06, + "loss": 0.70632005, + "num_input_tokens_seen": 96362230, + "step": 4462, + "time_per_iteration": 3.5301475524902344 + }, + { + "auxiliary_loss_clip": 0.01098019, + "auxiliary_loss_mlp": 0.01026212, + "balance_loss_clip": 0.93225455, + "balance_loss_mlp": 1.01786172, + "epoch": 0.5366440209222629, + "flos": 26578133879040.0, + "grad_norm": 1.784016864639085, + "language_loss": 0.81647629, + "learning_rate": 1.859899935899068e-06, + "loss": 0.83771861, + "num_input_tokens_seen": 96382085, + "step": 4463, + "time_per_iteration": 2.782198905944824 + }, + { + "auxiliary_loss_clip": 0.0112941, + "auxiliary_loss_mlp": 0.01028938, + "balance_loss_clip": 0.98112381, + "balance_loss_mlp": 1.0207926, + "epoch": 0.5367642638129021, + "flos": 19608052469760.0, + "grad_norm": 1.914568705686529, + "language_loss": 0.79138261, + "learning_rate": 1.8591228884579506e-06, + "loss": 0.81296611, + "num_input_tokens_seen": 96400580, + "step": 4464, + "time_per_iteration": 2.6702654361724854 + }, + { + "auxiliary_loss_clip": 0.01108682, + "auxiliary_loss_mlp": 0.01026549, + "balance_loss_clip": 0.93715763, + "balance_loss_mlp": 1.0196501, + "epoch": 0.5368845067035412, + "flos": 23915214172800.0, + "grad_norm": 3.153136692111458, + "language_loss": 0.82283276, + "learning_rate": 1.8583458623877795e-06, + "loss": 0.84418511, + "num_input_tokens_seen": 96419680, + "step": 4465, + "time_per_iteration": 2.713421583175659 + }, + { + "auxiliary_loss_clip": 0.01153161, + "auxiliary_loss_mlp": 0.01023174, + "balance_loss_clip": 1.01527429, + "balance_loss_mlp": 1.01563382, + "epoch": 0.5370047495941802, + "flos": 16873131951360.0, + "grad_norm": 9.49017266012574, + "language_loss": 0.74258041, + "learning_rate": 1.8575688578064281e-06, + "loss": 0.76434374, + "num_input_tokens_seen": 96437805, + "step": 4466, + "time_per_iteration": 2.67914080619812 + }, + { + "auxiliary_loss_clip": 0.01154069, + "auxiliary_loss_mlp": 0.01029519, + "balance_loss_clip": 1.01552796, + "balance_loss_mlp": 1.0218358, + "epoch": 0.5371249924848194, + "flos": 20740926493440.0, + "grad_norm": 1.659195105419592, + "language_loss": 0.76556468, + "learning_rate": 1.8567918748317674e-06, + "loss": 0.78740054, + "num_input_tokens_seen": 96457155, + "step": 4467, + "time_per_iteration": 2.710942268371582 + }, + { + "auxiliary_loss_clip": 0.01104902, + "auxiliary_loss_mlp": 0.0102641, + "balance_loss_clip": 0.93462777, + "balance_loss_mlp": 1.01845312, + "epoch": 0.5372452353754584, + "flos": 17968120104960.0, + "grad_norm": 2.017051751535218, + "language_loss": 0.82530051, + "learning_rate": 1.8560149135816659e-06, + "loss": 0.84661365, + "num_input_tokens_seen": 96473990, + "step": 4468, + "time_per_iteration": 2.775242567062378 + }, + { + "auxiliary_loss_clip": 0.01144914, + "auxiliary_loss_mlp": 0.01021937, + "balance_loss_clip": 1.01068771, + "balance_loss_mlp": 1.01476669, + "epoch": 0.5373654782660975, + "flos": 15377021642880.0, + "grad_norm": 2.379353050643064, + "language_loss": 0.85128641, + "learning_rate": 1.8552379741739873e-06, + "loss": 0.87295485, + "num_input_tokens_seen": 96491335, + "step": 4469, + "time_per_iteration": 3.513258695602417 + }, + { + "auxiliary_loss_clip": 0.01020016, + "auxiliary_loss_mlp": 0.00500485, + "balance_loss_clip": 0.90986735, + "balance_loss_mlp": 0.0, + "epoch": 0.5374857211567367, + "flos": 69000091574400.0, + "grad_norm": 0.8899416695368549, + "language_loss": 0.55673629, + "learning_rate": 1.8544610567265935e-06, + "loss": 0.57194132, + "num_input_tokens_seen": 96545275, + "step": 4470, + "time_per_iteration": 3.2126801013946533 + }, + { + "auxiliary_loss_clip": 0.01128152, + "auxiliary_loss_mlp": 0.00508706, + "balance_loss_clip": 0.97822905, + "balance_loss_mlp": 0.0, + "epoch": 0.5376059640473757, + "flos": 15085355207040.0, + "grad_norm": 2.522760687755357, + "language_loss": 0.83620775, + "learning_rate": 1.853684161357341e-06, + "loss": 0.85257626, + "num_input_tokens_seen": 96562935, + "step": 4471, + "time_per_iteration": 2.745877265930176 + }, + { + "auxiliary_loss_clip": 0.01149289, + "auxiliary_loss_mlp": 0.00509584, + "balance_loss_clip": 1.01547408, + "balance_loss_mlp": 0.0, + "epoch": 0.5377262069380148, + "flos": 19792597570560.0, + "grad_norm": 2.147809317842159, + "language_loss": 0.76821381, + "learning_rate": 1.852907288184085e-06, + "loss": 0.78480256, + "num_input_tokens_seen": 96581820, + "step": 4472, + "time_per_iteration": 2.669456958770752 + }, + { + "auxiliary_loss_clip": 0.01087887, + "auxiliary_loss_mlp": 0.01025808, + "balance_loss_clip": 0.90179408, + "balance_loss_mlp": 1.01792228, + "epoch": 0.5378464498286539, + "flos": 30003077640960.0, + "grad_norm": 2.061652935161495, + "language_loss": 0.70333695, + "learning_rate": 1.8521304373246762e-06, + "loss": 0.72447383, + "num_input_tokens_seen": 96602865, + "step": 4473, + "time_per_iteration": 2.8508739471435547 + }, + { + "auxiliary_loss_clip": 0.01155139, + "auxiliary_loss_mlp": 0.01024814, + "balance_loss_clip": 1.01548338, + "balance_loss_mlp": 1.01721776, + "epoch": 0.537966692719293, + "flos": 21251217058560.0, + "grad_norm": 8.635990013309359, + "language_loss": 0.89132416, + "learning_rate": 1.8513536088969626e-06, + "loss": 0.91312367, + "num_input_tokens_seen": 96620530, + "step": 4474, + "time_per_iteration": 2.6923255920410156 + }, + { + "auxiliary_loss_clip": 0.01153569, + "auxiliary_loss_mlp": 0.01029969, + "balance_loss_clip": 1.01622403, + "balance_loss_mlp": 1.02221453, + "epoch": 0.538086935609932, + "flos": 21543170803200.0, + "grad_norm": 2.7438531276301625, + "language_loss": 0.80262679, + "learning_rate": 1.8505768030187884e-06, + "loss": 0.82446212, + "num_input_tokens_seen": 96640660, + "step": 4475, + "time_per_iteration": 2.6961312294006348 + }, + { + "auxiliary_loss_clip": 0.01127406, + "auxiliary_loss_mlp": 0.01028503, + "balance_loss_clip": 0.97875321, + "balance_loss_mlp": 1.02115095, + "epoch": 0.5382071785005712, + "flos": 22747219626240.0, + "grad_norm": 1.520104645552715, + "language_loss": 0.80039889, + "learning_rate": 1.849800019807995e-06, + "loss": 0.82195795, + "num_input_tokens_seen": 96661885, + "step": 4476, + "time_per_iteration": 2.7750484943389893 + }, + { + "auxiliary_loss_clip": 0.01105412, + "auxiliary_loss_mlp": 0.01021327, + "balance_loss_clip": 0.94009119, + "balance_loss_mlp": 1.01352453, + "epoch": 0.5383274213912103, + "flos": 24934574240640.0, + "grad_norm": 1.9065898262196705, + "language_loss": 0.70777428, + "learning_rate": 1.8490232593824186e-06, + "loss": 0.72904164, + "num_input_tokens_seen": 96678340, + "step": 4477, + "time_per_iteration": 2.7297732830047607 + }, + { + "auxiliary_loss_clip": 0.01127481, + "auxiliary_loss_mlp": 0.01023779, + "balance_loss_clip": 0.97842968, + "balance_loss_mlp": 1.01645935, + "epoch": 0.5384476642818493, + "flos": 22310186849280.0, + "grad_norm": 1.8097656350799816, + "language_loss": 0.84735203, + "learning_rate": 1.8482465218598935e-06, + "loss": 0.8688646, + "num_input_tokens_seen": 96698285, + "step": 4478, + "time_per_iteration": 2.7414793968200684 + }, + { + "auxiliary_loss_clip": 0.01107342, + "auxiliary_loss_mlp": 0.01029298, + "balance_loss_clip": 0.93813539, + "balance_loss_mlp": 1.0212574, + "epoch": 0.5385679071724885, + "flos": 22711021695360.0, + "grad_norm": 1.9304262083808925, + "language_loss": 0.83547455, + "learning_rate": 1.8474698073582508e-06, + "loss": 0.85684097, + "num_input_tokens_seen": 96719655, + "step": 4479, + "time_per_iteration": 2.7438859939575195 + }, + { + "auxiliary_loss_clip": 0.01111097, + "auxiliary_loss_mlp": 0.01024497, + "balance_loss_clip": 0.93864822, + "balance_loss_mlp": 1.01677179, + "epoch": 0.5386881500631275, + "flos": 15953746412160.0, + "grad_norm": 2.850931371721223, + "language_loss": 0.87327474, + "learning_rate": 1.8466931159953166e-06, + "loss": 0.89463067, + "num_input_tokens_seen": 96736290, + "step": 4480, + "time_per_iteration": 2.752350091934204 + }, + { + "auxiliary_loss_clip": 0.01134077, + "auxiliary_loss_mlp": 0.01027595, + "balance_loss_clip": 0.97958219, + "balance_loss_mlp": 1.02007914, + "epoch": 0.5388083929537666, + "flos": 24060041809920.0, + "grad_norm": 1.626982669821825, + "language_loss": 0.83935285, + "learning_rate": 1.8459164478889158e-06, + "loss": 0.86096954, + "num_input_tokens_seen": 96757685, + "step": 4481, + "time_per_iteration": 2.7418181896209717 + }, + { + "auxiliary_loss_clip": 0.01096746, + "auxiliary_loss_mlp": 0.01025777, + "balance_loss_clip": 0.93380439, + "balance_loss_mlp": 1.01816881, + "epoch": 0.5389286358444056, + "flos": 22236893147520.0, + "grad_norm": 1.7908611494751885, + "language_loss": 0.75916636, + "learning_rate": 1.8451398031568663e-06, + "loss": 0.78039157, + "num_input_tokens_seen": 96777310, + "step": 4482, + "time_per_iteration": 2.9092814922332764 + }, + { + "auxiliary_loss_clip": 0.01106089, + "auxiliary_loss_mlp": 0.01028479, + "balance_loss_clip": 0.93902838, + "balance_loss_mlp": 1.02097833, + "epoch": 0.5390488787350448, + "flos": 24281718595200.0, + "grad_norm": 1.5501200610535988, + "language_loss": 0.7480464, + "learning_rate": 1.844363181916986e-06, + "loss": 0.76939207, + "num_input_tokens_seen": 96798035, + "step": 4483, + "time_per_iteration": 2.7512614727020264 + }, + { + "auxiliary_loss_clip": 0.01149542, + "auxiliary_loss_mlp": 0.01028876, + "balance_loss_clip": 1.01251376, + "balance_loss_mlp": 1.02119279, + "epoch": 0.5391691216256839, + "flos": 16581393688320.0, + "grad_norm": 1.9351350439595938, + "language_loss": 0.83144855, + "learning_rate": 1.8435865842870868e-06, + "loss": 0.85323274, + "num_input_tokens_seen": 96815975, + "step": 4484, + "time_per_iteration": 2.6462533473968506 + }, + { + "auxiliary_loss_clip": 0.01117929, + "auxiliary_loss_mlp": 0.00509588, + "balance_loss_clip": 0.97100055, + "balance_loss_mlp": 0.0, + "epoch": 0.5392893645163229, + "flos": 23330049707520.0, + "grad_norm": 1.8821328456326898, + "language_loss": 0.72390735, + "learning_rate": 1.8428100103849787e-06, + "loss": 0.74018258, + "num_input_tokens_seen": 96835770, + "step": 4485, + "time_per_iteration": 4.5577921867370605 + }, + { + "auxiliary_loss_clip": 0.01130623, + "auxiliary_loss_mlp": 0.01031545, + "balance_loss_clip": 0.98036098, + "balance_loss_mlp": 1.02416015, + "epoch": 0.5394096074069621, + "flos": 15669801400320.0, + "grad_norm": 2.4116153524362445, + "language_loss": 0.73370147, + "learning_rate": 1.842033460328467e-06, + "loss": 0.75532311, + "num_input_tokens_seen": 96854490, + "step": 4486, + "time_per_iteration": 2.709899425506592 + }, + { + "auxiliary_loss_clip": 0.01133947, + "auxiliary_loss_mlp": 0.00509493, + "balance_loss_clip": 0.9767822, + "balance_loss_mlp": 0.0, + "epoch": 0.5395298502976011, + "flos": 22893447893760.0, + "grad_norm": 1.807314056852377, + "language_loss": 0.75269246, + "learning_rate": 1.8412569342353541e-06, + "loss": 0.76912689, + "num_input_tokens_seen": 96874645, + "step": 4487, + "time_per_iteration": 2.6991782188415527 + }, + { + "auxiliary_loss_clip": 0.01136121, + "auxiliary_loss_mlp": 0.01029645, + "balance_loss_clip": 0.97966301, + "balance_loss_mlp": 1.02203035, + "epoch": 0.5396500931882402, + "flos": 23842135952640.0, + "grad_norm": 8.4904797440182, + "language_loss": 0.85229975, + "learning_rate": 1.840480432223438e-06, + "loss": 0.8739574, + "num_input_tokens_seen": 96893650, + "step": 4488, + "time_per_iteration": 3.5567023754119873 + }, + { + "auxiliary_loss_clip": 0.01129935, + "auxiliary_loss_mlp": 0.01021898, + "balance_loss_clip": 0.97461295, + "balance_loss_mlp": 1.01431346, + "epoch": 0.5397703360788794, + "flos": 26322988596480.0, + "grad_norm": 2.1575977137110205, + "language_loss": 0.78052664, + "learning_rate": 1.8397039544105131e-06, + "loss": 0.80204499, + "num_input_tokens_seen": 96912735, + "step": 4489, + "time_per_iteration": 2.7500176429748535 + }, + { + "auxiliary_loss_clip": 0.01123039, + "auxiliary_loss_mlp": 0.01031861, + "balance_loss_clip": 0.97182375, + "balance_loss_mlp": 1.0240767, + "epoch": 0.5398905789695184, + "flos": 21214588164480.0, + "grad_norm": 1.918230936333055, + "language_loss": 0.69964468, + "learning_rate": 1.8389275009143711e-06, + "loss": 0.72119367, + "num_input_tokens_seen": 96932475, + "step": 4490, + "time_per_iteration": 2.6968026161193848 + }, + { + "auxiliary_loss_clip": 0.01173347, + "auxiliary_loss_mlp": 0.01027647, + "balance_loss_clip": 1.05231953, + "balance_loss_mlp": 1.02007127, + "epoch": 0.5400108218601575, + "flos": 25080335631360.0, + "grad_norm": 1.8090331263787294, + "language_loss": 0.73266017, + "learning_rate": 1.8381510718527988e-06, + "loss": 0.75467014, + "num_input_tokens_seen": 96952085, + "step": 4491, + "time_per_iteration": 2.643869638442993 + }, + { + "auxiliary_loss_clip": 0.01130834, + "auxiliary_loss_mlp": 0.01030459, + "balance_loss_clip": 0.97384429, + "balance_loss_mlp": 1.02241814, + "epoch": 0.5401310647507966, + "flos": 26357498588160.0, + "grad_norm": 1.8195410726774959, + "language_loss": 0.6359756, + "learning_rate": 1.8373746673435812e-06, + "loss": 0.65758854, + "num_input_tokens_seen": 96973110, + "step": 4492, + "time_per_iteration": 2.7705721855163574 + }, + { + "auxiliary_loss_clip": 0.01177034, + "auxiliary_loss_mlp": 0.01024717, + "balance_loss_clip": 1.05438709, + "balance_loss_mlp": 1.0171715, + "epoch": 0.5402513076414357, + "flos": 27855332749440.0, + "grad_norm": 1.80530531613027, + "language_loss": 0.78941369, + "learning_rate": 1.8365982875044964e-06, + "loss": 0.81143123, + "num_input_tokens_seen": 96993420, + "step": 4493, + "time_per_iteration": 2.7057132720947266 + }, + { + "auxiliary_loss_clip": 0.01159659, + "auxiliary_loss_mlp": 0.00510167, + "balance_loss_clip": 1.01677227, + "balance_loss_mlp": 0.0, + "epoch": 0.5403715505320748, + "flos": 22893771116160.0, + "grad_norm": 57.75838912448722, + "language_loss": 0.76312292, + "learning_rate": 1.8358219324533217e-06, + "loss": 0.77982116, + "num_input_tokens_seen": 97013685, + "step": 4494, + "time_per_iteration": 2.6682987213134766 + }, + { + "auxiliary_loss_clip": 0.01123391, + "auxiliary_loss_mlp": 0.01031395, + "balance_loss_clip": 0.97367239, + "balance_loss_mlp": 1.02445364, + "epoch": 0.5404917934227139, + "flos": 30224143895040.0, + "grad_norm": 1.6280744711963955, + "language_loss": 0.69994175, + "learning_rate": 1.8350456023078292e-06, + "loss": 0.72148961, + "num_input_tokens_seen": 97036060, + "step": 4495, + "time_per_iteration": 3.6669774055480957 + }, + { + "auxiliary_loss_clip": 0.01178367, + "auxiliary_loss_mlp": 0.01034322, + "balance_loss_clip": 1.05231237, + "balance_loss_mlp": 1.02628684, + "epoch": 0.540612036313353, + "flos": 19938502615680.0, + "grad_norm": 2.7671049570891197, + "language_loss": 0.77966273, + "learning_rate": 1.8342692971857874e-06, + "loss": 0.80178964, + "num_input_tokens_seen": 97055260, + "step": 4496, + "time_per_iteration": 2.6089155673980713 + }, + { + "auxiliary_loss_clip": 0.01128721, + "auxiliary_loss_mlp": 0.01026829, + "balance_loss_clip": 0.97798848, + "balance_loss_mlp": 1.01946187, + "epoch": 0.540732279203992, + "flos": 24279599692800.0, + "grad_norm": 2.064688751654545, + "language_loss": 0.71427286, + "learning_rate": 1.833493017204962e-06, + "loss": 0.7358284, + "num_input_tokens_seen": 97075365, + "step": 4497, + "time_per_iteration": 2.715878963470459 + }, + { + "auxiliary_loss_clip": 0.01175709, + "auxiliary_loss_mlp": 0.01030451, + "balance_loss_clip": 1.05264533, + "balance_loss_mlp": 1.02298558, + "epoch": 0.5408525220946312, + "flos": 20193216935040.0, + "grad_norm": 1.9043467549033244, + "language_loss": 0.77826512, + "learning_rate": 1.8327167624831134e-06, + "loss": 0.80032665, + "num_input_tokens_seen": 97093095, + "step": 4498, + "time_per_iteration": 2.568472385406494 + }, + { + "auxiliary_loss_clip": 0.01176734, + "auxiliary_loss_mlp": 0.01026273, + "balance_loss_clip": 1.05416083, + "balance_loss_mlp": 1.01913202, + "epoch": 0.5409727649852702, + "flos": 24134448833280.0, + "grad_norm": 1.5359872478578507, + "language_loss": 0.71127135, + "learning_rate": 1.831940533137999e-06, + "loss": 0.7333014, + "num_input_tokens_seen": 97112000, + "step": 4499, + "time_per_iteration": 2.6660211086273193 + }, + { + "auxiliary_loss_clip": 0.01152212, + "auxiliary_loss_mlp": 0.01028162, + "balance_loss_clip": 1.01892734, + "balance_loss_mlp": 1.02111089, + "epoch": 0.5410930078759093, + "flos": 23912700220800.0, + "grad_norm": 1.6582712174584402, + "language_loss": 0.72150826, + "learning_rate": 1.8311643292873718e-06, + "loss": 0.743312, + "num_input_tokens_seen": 97130820, + "step": 4500, + "time_per_iteration": 2.6511645317077637 + }, + { + "auxiliary_loss_clip": 0.01148631, + "auxiliary_loss_mlp": 0.01026197, + "balance_loss_clip": 1.01483703, + "balance_loss_mlp": 1.01931882, + "epoch": 0.5412132507665485, + "flos": 21105132445440.0, + "grad_norm": 1.8384820239818414, + "language_loss": 0.87725776, + "learning_rate": 1.8303881510489818e-06, + "loss": 0.89900607, + "num_input_tokens_seen": 97149210, + "step": 4501, + "time_per_iteration": 2.710676670074463 + }, + { + "auxiliary_loss_clip": 0.01128715, + "auxiliary_loss_mlp": 0.01023864, + "balance_loss_clip": 0.97589374, + "balance_loss_mlp": 1.01693475, + "epoch": 0.5413334936571875, + "flos": 30227340205440.0, + "grad_norm": 1.9467695008020518, + "language_loss": 0.69417059, + "learning_rate": 1.829611998540574e-06, + "loss": 0.71569639, + "num_input_tokens_seen": 97170415, + "step": 4502, + "time_per_iteration": 2.822768449783325 + }, + { + "auxiliary_loss_clip": 0.01153264, + "auxiliary_loss_mlp": 0.0050992, + "balance_loss_clip": 1.01444995, + "balance_loss_mlp": 0.0, + "epoch": 0.5414537365478266, + "flos": 24279635606400.0, + "grad_norm": 1.7471924073030956, + "language_loss": 0.79738188, + "learning_rate": 1.8288358718798914e-06, + "loss": 0.81401372, + "num_input_tokens_seen": 97189605, + "step": 4503, + "time_per_iteration": 2.6649396419525146 + }, + { + "auxiliary_loss_clip": 0.0115213, + "auxiliary_loss_mlp": 0.00509052, + "balance_loss_clip": 1.01800895, + "balance_loss_mlp": 0.0, + "epoch": 0.5415739794384657, + "flos": 16654543735680.0, + "grad_norm": 1.916899826130107, + "language_loss": 0.72480464, + "learning_rate": 1.8280597711846703e-06, + "loss": 0.74141645, + "num_input_tokens_seen": 97207845, + "step": 4504, + "time_per_iteration": 2.634568929672241 + }, + { + "auxiliary_loss_clip": 0.01153044, + "auxiliary_loss_mlp": 0.01030406, + "balance_loss_clip": 1.01861453, + "balance_loss_mlp": 1.02287245, + "epoch": 0.5416942223291048, + "flos": 23185724860800.0, + "grad_norm": 1.8297018913671608, + "language_loss": 0.83657718, + "learning_rate": 1.8272836965726455e-06, + "loss": 0.85841167, + "num_input_tokens_seen": 97226780, + "step": 4505, + "time_per_iteration": 2.6185197830200195 + }, + { + "auxiliary_loss_clip": 0.01060371, + "auxiliary_loss_mlp": 0.01027619, + "balance_loss_clip": 0.85987484, + "balance_loss_mlp": 1.0196383, + "epoch": 0.5418144652197439, + "flos": 20303247271680.0, + "grad_norm": 1.6393742410360348, + "language_loss": 0.78324175, + "learning_rate": 1.8265076481615461e-06, + "loss": 0.80412161, + "num_input_tokens_seen": 97246695, + "step": 4506, + "time_per_iteration": 2.7912352085113525 + }, + { + "auxiliary_loss_clip": 0.01128259, + "auxiliary_loss_mlp": 0.0102517, + "balance_loss_clip": 0.9799577, + "balance_loss_mlp": 1.01735544, + "epoch": 0.541934708110383, + "flos": 12458633431680.0, + "grad_norm": 2.533792721640092, + "language_loss": 0.86986518, + "learning_rate": 1.8257316260690987e-06, + "loss": 0.89139938, + "num_input_tokens_seen": 97264480, + "step": 4507, + "time_per_iteration": 2.6478919982910156 + }, + { + "auxiliary_loss_clip": 0.01153798, + "auxiliary_loss_mlp": 0.01023, + "balance_loss_clip": 1.01489544, + "balance_loss_mlp": 1.01588607, + "epoch": 0.5420549510010221, + "flos": 21253802837760.0, + "grad_norm": 1.6315919568865063, + "language_loss": 0.76217771, + "learning_rate": 1.8249556304130254e-06, + "loss": 0.78394568, + "num_input_tokens_seen": 97285760, + "step": 4508, + "time_per_iteration": 2.671539068222046 + }, + { + "auxiliary_loss_clip": 0.0111948, + "auxiliary_loss_mlp": 0.01026742, + "balance_loss_clip": 0.97330713, + "balance_loss_mlp": 1.01899374, + "epoch": 0.5421751938916611, + "flos": 29490524519040.0, + "grad_norm": 2.2142988149465546, + "language_loss": 0.68700004, + "learning_rate": 1.824179661311044e-06, + "loss": 0.70846236, + "num_input_tokens_seen": 97304510, + "step": 4509, + "time_per_iteration": 2.7735064029693604 + }, + { + "auxiliary_loss_clip": 0.01082729, + "auxiliary_loss_mlp": 0.01031347, + "balance_loss_clip": 0.89671254, + "balance_loss_mlp": 1.02363992, + "epoch": 0.5422954367823003, + "flos": 18734238311040.0, + "grad_norm": 2.339162964359817, + "language_loss": 0.79745293, + "learning_rate": 1.823403718880868e-06, + "loss": 0.81859362, + "num_input_tokens_seen": 97323270, + "step": 4510, + "time_per_iteration": 2.767549991607666 + }, + { + "auxiliary_loss_clip": 0.01128763, + "auxiliary_loss_mlp": 0.01027506, + "balance_loss_clip": 0.97398192, + "balance_loss_mlp": 1.01970351, + "epoch": 0.5424156796729394, + "flos": 39969006940800.0, + "grad_norm": 1.6149552738779254, + "language_loss": 0.66239077, + "learning_rate": 1.822627803240207e-06, + "loss": 0.6839534, + "num_input_tokens_seen": 97345600, + "step": 4511, + "time_per_iteration": 3.780722141265869 + }, + { + "auxiliary_loss_clip": 0.01107624, + "auxiliary_loss_mlp": 0.01022913, + "balance_loss_clip": 0.93680066, + "balance_loss_mlp": 1.01552796, + "epoch": 0.5425359225635784, + "flos": 11546538353280.0, + "grad_norm": 2.3848576237798524, + "language_loss": 0.85280132, + "learning_rate": 1.8218519145067675e-06, + "loss": 0.87410671, + "num_input_tokens_seen": 97361220, + "step": 4512, + "time_per_iteration": 3.5967459678649902 + }, + { + "auxiliary_loss_clip": 0.01103233, + "auxiliary_loss_mlp": 0.0102705, + "balance_loss_clip": 0.93772817, + "balance_loss_mlp": 1.01945055, + "epoch": 0.5426561654542175, + "flos": 20229702174720.0, + "grad_norm": 1.816468286972738, + "language_loss": 0.89529121, + "learning_rate": 1.8210760527982508e-06, + "loss": 0.91659409, + "num_input_tokens_seen": 97381505, + "step": 4513, + "time_per_iteration": 3.6757938861846924 + }, + { + "auxiliary_loss_clip": 0.01132581, + "auxiliary_loss_mlp": 0.00508912, + "balance_loss_clip": 0.9799664, + "balance_loss_mlp": 0.0, + "epoch": 0.5427764083448566, + "flos": 21871681614720.0, + "grad_norm": 3.0961952374969903, + "language_loss": 0.75559902, + "learning_rate": 1.8203002182323552e-06, + "loss": 0.77201396, + "num_input_tokens_seen": 97399060, + "step": 4514, + "time_per_iteration": 2.686398983001709 + }, + { + "auxiliary_loss_clip": 0.01133257, + "auxiliary_loss_mlp": 0.0102846, + "balance_loss_clip": 0.9800365, + "balance_loss_mlp": 1.02036595, + "epoch": 0.5428966512354957, + "flos": 19640946349440.0, + "grad_norm": 2.036166058356429, + "language_loss": 0.7560665, + "learning_rate": 1.819524410926773e-06, + "loss": 0.77768373, + "num_input_tokens_seen": 97416740, + "step": 4515, + "time_per_iteration": 2.672321081161499 + }, + { + "auxiliary_loss_clip": 0.01054973, + "auxiliary_loss_mlp": 0.01022401, + "balance_loss_clip": 0.86169624, + "balance_loss_mlp": 1.01534712, + "epoch": 0.5430168941261347, + "flos": 22382187661440.0, + "grad_norm": 1.898919857893661, + "language_loss": 0.77558345, + "learning_rate": 1.8187486309991944e-06, + "loss": 0.79635721, + "num_input_tokens_seen": 97437620, + "step": 4516, + "time_per_iteration": 2.825524091720581 + }, + { + "auxiliary_loss_clip": 0.01162066, + "auxiliary_loss_mlp": 0.01024318, + "balance_loss_clip": 1.019454, + "balance_loss_mlp": 1.01660562, + "epoch": 0.5431371370167739, + "flos": 18764187275520.0, + "grad_norm": 2.1063354021956315, + "language_loss": 0.77598137, + "learning_rate": 1.817972878567304e-06, + "loss": 0.79784524, + "num_input_tokens_seen": 97456275, + "step": 4517, + "time_per_iteration": 2.6857526302337646 + }, + { + "auxiliary_loss_clip": 0.01134089, + "auxiliary_loss_mlp": 0.01025855, + "balance_loss_clip": 0.97556376, + "balance_loss_mlp": 1.01872289, + "epoch": 0.543257379907413, + "flos": 18806023641600.0, + "grad_norm": 1.8054071980610045, + "language_loss": 0.76099777, + "learning_rate": 1.8171971537487834e-06, + "loss": 0.78259724, + "num_input_tokens_seen": 97474925, + "step": 4518, + "time_per_iteration": 2.722341299057007 + }, + { + "auxiliary_loss_clip": 0.01176482, + "auxiliary_loss_mlp": 0.01027748, + "balance_loss_clip": 1.05267262, + "balance_loss_mlp": 1.02029729, + "epoch": 0.543377622798052, + "flos": 17493381025920.0, + "grad_norm": 1.924726923353375, + "language_loss": 0.80499512, + "learning_rate": 1.8164214566613093e-06, + "loss": 0.82703745, + "num_input_tokens_seen": 97493550, + "step": 4519, + "time_per_iteration": 2.584757089614868 + }, + { + "auxiliary_loss_clip": 0.01173594, + "auxiliary_loss_mlp": 0.01031215, + "balance_loss_clip": 1.05205417, + "balance_loss_mlp": 1.02362144, + "epoch": 0.5434978656886912, + "flos": 18989311766400.0, + "grad_norm": 3.1066277535998643, + "language_loss": 0.66547686, + "learning_rate": 1.8156457874225547e-06, + "loss": 0.68752491, + "num_input_tokens_seen": 97512010, + "step": 4520, + "time_per_iteration": 2.6488070487976074 + }, + { + "auxiliary_loss_clip": 0.01121996, + "auxiliary_loss_mlp": 0.01027854, + "balance_loss_clip": 0.97623485, + "balance_loss_mlp": 1.02023697, + "epoch": 0.5436181085793302, + "flos": 17274936464640.0, + "grad_norm": 1.7764706928037226, + "language_loss": 0.80443883, + "learning_rate": 1.814870146150187e-06, + "loss": 0.82593733, + "num_input_tokens_seen": 97530120, + "step": 4521, + "time_per_iteration": 2.6586527824401855 + }, + { + "auxiliary_loss_clip": 0.01135383, + "auxiliary_loss_mlp": 0.01028949, + "balance_loss_clip": 0.97562206, + "balance_loss_mlp": 1.02068746, + "epoch": 0.5437383514699693, + "flos": 19098587917440.0, + "grad_norm": 1.9246383861416474, + "language_loss": 0.7875967, + "learning_rate": 1.814094532961871e-06, + "loss": 0.80924004, + "num_input_tokens_seen": 97548695, + "step": 4522, + "time_per_iteration": 3.5904290676116943 + }, + { + "auxiliary_loss_clip": 0.01082777, + "auxiliary_loss_mlp": 0.01031536, + "balance_loss_clip": 0.89711583, + "balance_loss_mlp": 1.0240674, + "epoch": 0.5438585943606085, + "flos": 22602715211520.0, + "grad_norm": 1.9537666002003133, + "language_loss": 0.83283037, + "learning_rate": 1.8133189479752666e-06, + "loss": 0.85397351, + "num_input_tokens_seen": 97567625, + "step": 4523, + "time_per_iteration": 2.821532964706421 + }, + { + "auxiliary_loss_clip": 0.0117576, + "auxiliary_loss_mlp": 0.01026697, + "balance_loss_clip": 1.05393553, + "balance_loss_mlp": 1.01910079, + "epoch": 0.5439788372512475, + "flos": 21798495653760.0, + "grad_norm": 1.796353747872582, + "language_loss": 0.81886816, + "learning_rate": 1.8125433913080292e-06, + "loss": 0.84089267, + "num_input_tokens_seen": 97585325, + "step": 4524, + "time_per_iteration": 2.5984535217285156 + }, + { + "auxiliary_loss_clip": 0.00985647, + "auxiliary_loss_mlp": 0.01029384, + "balance_loss_clip": 0.74756849, + "balance_loss_mlp": 1.02219534, + "epoch": 0.5440990801418866, + "flos": 16399362539520.0, + "grad_norm": 14.271136438089172, + "language_loss": 0.82989144, + "learning_rate": 1.811767863077811e-06, + "loss": 0.85004187, + "num_input_tokens_seen": 97604275, + "step": 4525, + "time_per_iteration": 3.108550548553467 + }, + { + "auxiliary_loss_clip": 0.01051968, + "auxiliary_loss_mlp": 0.0102588, + "balance_loss_clip": 0.8616519, + "balance_loss_mlp": 1.01894224, + "epoch": 0.5442193230325257, + "flos": 21615638492160.0, + "grad_norm": 1.8045603706726032, + "language_loss": 0.78252983, + "learning_rate": 1.8109923634022577e-06, + "loss": 0.80330831, + "num_input_tokens_seen": 97624300, + "step": 4526, + "time_per_iteration": 2.9995157718658447 + }, + { + "auxiliary_loss_clip": 0.01178226, + "auxiliary_loss_mlp": 0.01026817, + "balance_loss_clip": 1.05307484, + "balance_loss_mlp": 1.01926541, + "epoch": 0.5443395659231648, + "flos": 15481198062720.0, + "grad_norm": 12.792715243312005, + "language_loss": 0.86797571, + "learning_rate": 1.8102168923990128e-06, + "loss": 0.89002621, + "num_input_tokens_seen": 97637845, + "step": 4527, + "time_per_iteration": 2.592426300048828 + }, + { + "auxiliary_loss_clip": 0.01155586, + "auxiliary_loss_mlp": 0.00508455, + "balance_loss_clip": 1.01627135, + "balance_loss_mlp": 0.0, + "epoch": 0.5444598088138038, + "flos": 18770436241920.0, + "grad_norm": 3.198330410849569, + "language_loss": 0.80072826, + "learning_rate": 1.809441450185714e-06, + "loss": 0.81736875, + "num_input_tokens_seen": 97656330, + "step": 4528, + "time_per_iteration": 2.644462823867798 + }, + { + "auxiliary_loss_clip": 0.01129983, + "auxiliary_loss_mlp": 0.01029937, + "balance_loss_clip": 0.97247064, + "balance_loss_mlp": 1.02276063, + "epoch": 0.544580051704443, + "flos": 21142335957120.0, + "grad_norm": 2.1659194669304225, + "language_loss": 0.73222089, + "learning_rate": 1.8086660368799958e-06, + "loss": 0.75382006, + "num_input_tokens_seen": 97674380, + "step": 4529, + "time_per_iteration": 2.67014217376709 + }, + { + "auxiliary_loss_clip": 0.01130461, + "auxiliary_loss_mlp": 0.01023452, + "balance_loss_clip": 0.9766506, + "balance_loss_mlp": 1.01581931, + "epoch": 0.5447002945950821, + "flos": 32491508054400.0, + "grad_norm": 1.8430752438311044, + "language_loss": 0.77220249, + "learning_rate": 1.807890652599488e-06, + "loss": 0.79374164, + "num_input_tokens_seen": 97698765, + "step": 4530, + "time_per_iteration": 2.9388506412506104 + }, + { + "auxiliary_loss_clip": 0.01173455, + "auxiliary_loss_mlp": 0.01022759, + "balance_loss_clip": 1.05396271, + "balance_loss_mlp": 1.01567781, + "epoch": 0.5448205374857211, + "flos": 11798307757440.0, + "grad_norm": 1.900941492549987, + "language_loss": 0.82779574, + "learning_rate": 1.8071152974618156e-06, + "loss": 0.84975785, + "num_input_tokens_seen": 97716565, + "step": 4531, + "time_per_iteration": 2.558220863342285 + }, + { + "auxiliary_loss_clip": 0.01104593, + "auxiliary_loss_mlp": 0.00508736, + "balance_loss_clip": 0.93527055, + "balance_loss_mlp": 0.0, + "epoch": 0.5449407803763603, + "flos": 24133766474880.0, + "grad_norm": 2.074586209532003, + "language_loss": 0.78253502, + "learning_rate": 1.806339971584599e-06, + "loss": 0.79866832, + "num_input_tokens_seen": 97733225, + "step": 4532, + "time_per_iteration": 2.7495217323303223 + }, + { + "auxiliary_loss_clip": 0.01173622, + "auxiliary_loss_mlp": 0.01025072, + "balance_loss_clip": 1.05186355, + "balance_loss_mlp": 1.01805329, + "epoch": 0.5450610232669993, + "flos": 23258551685760.0, + "grad_norm": 2.0724523300661186, + "language_loss": 0.85165083, + "learning_rate": 1.8055646750854546e-06, + "loss": 0.8736378, + "num_input_tokens_seen": 97752735, + "step": 4533, + "time_per_iteration": 2.611576557159424 + }, + { + "auxiliary_loss_clip": 0.01134026, + "auxiliary_loss_mlp": 0.01026562, + "balance_loss_clip": 0.97736681, + "balance_loss_mlp": 1.01954055, + "epoch": 0.5451812661576384, + "flos": 17785083375360.0, + "grad_norm": 2.289624129066367, + "language_loss": 0.81617045, + "learning_rate": 1.8047894080819945e-06, + "loss": 0.83777642, + "num_input_tokens_seen": 97769985, + "step": 4534, + "time_per_iteration": 2.689970016479492 + }, + { + "auxiliary_loss_clip": 0.01078832, + "auxiliary_loss_mlp": 0.00998687, + "balance_loss_clip": 1.02103877, + "balance_loss_mlp": 0.99760228, + "epoch": 0.5453015090482776, + "flos": 71062586513280.0, + "grad_norm": 0.7302367770603421, + "language_loss": 0.6323216, + "learning_rate": 1.8040141706918258e-06, + "loss": 0.6530968, + "num_input_tokens_seen": 97831225, + "step": 4535, + "time_per_iteration": 3.2855443954467773 + }, + { + "auxiliary_loss_clip": 0.01130318, + "auxiliary_loss_mlp": 0.01026129, + "balance_loss_clip": 0.9774425, + "balance_loss_mlp": 1.01928627, + "epoch": 0.5454217519389166, + "flos": 25552201622400.0, + "grad_norm": 2.3194078699481326, + "language_loss": 0.77143222, + "learning_rate": 1.8032389630325525e-06, + "loss": 0.79299664, + "num_input_tokens_seen": 97849975, + "step": 4536, + "time_per_iteration": 2.6950082778930664 + }, + { + "auxiliary_loss_clip": 0.01126661, + "auxiliary_loss_mlp": 0.01026282, + "balance_loss_clip": 0.97316492, + "balance_loss_mlp": 1.01891136, + "epoch": 0.5455419948295557, + "flos": 23658345037440.0, + "grad_norm": 1.7259295213563721, + "language_loss": 0.75820875, + "learning_rate": 1.8024637852217707e-06, + "loss": 0.77973819, + "num_input_tokens_seen": 97869700, + "step": 4537, + "time_per_iteration": 3.786513566970825 + }, + { + "auxiliary_loss_clip": 0.01132672, + "auxiliary_loss_mlp": 0.01026107, + "balance_loss_clip": 0.9792552, + "balance_loss_mlp": 1.01899004, + "epoch": 0.5456622377201948, + "flos": 23403989854080.0, + "grad_norm": 2.239232262696076, + "language_loss": 0.8433696, + "learning_rate": 1.8016886373770766e-06, + "loss": 0.86495733, + "num_input_tokens_seen": 97888215, + "step": 4538, + "time_per_iteration": 3.905066728591919 + }, + { + "auxiliary_loss_clip": 0.01128132, + "auxiliary_loss_mlp": 0.01028164, + "balance_loss_clip": 0.97660494, + "balance_loss_mlp": 1.02086806, + "epoch": 0.5457824806108339, + "flos": 23988040997760.0, + "grad_norm": 1.787233758197377, + "language_loss": 0.78851098, + "learning_rate": 1.8009135196160579e-06, + "loss": 0.81007397, + "num_input_tokens_seen": 97907090, + "step": 4539, + "time_per_iteration": 3.6628589630126953 + }, + { + "auxiliary_loss_clip": 0.01101331, + "auxiliary_loss_mlp": 0.01031412, + "balance_loss_clip": 0.93708229, + "balance_loss_mlp": 1.02399707, + "epoch": 0.545902723501473, + "flos": 22565870835840.0, + "grad_norm": 1.687453142186295, + "language_loss": 0.84445941, + "learning_rate": 1.8001384320563e-06, + "loss": 0.86578679, + "num_input_tokens_seen": 97927345, + "step": 4540, + "time_per_iteration": 2.7243781089782715 + }, + { + "auxiliary_loss_clip": 0.01079675, + "auxiliary_loss_mlp": 0.01001852, + "balance_loss_clip": 1.0216229, + "balance_loss_mlp": 1.00071323, + "epoch": 0.5460229663921121, + "flos": 55198399685760.0, + "grad_norm": 0.7705567145764256, + "language_loss": 0.5776999, + "learning_rate": 1.7993633748153833e-06, + "loss": 0.59851515, + "num_input_tokens_seen": 97981950, + "step": 4541, + "time_per_iteration": 3.0703890323638916 + }, + { + "auxiliary_loss_clip": 0.01157078, + "auxiliary_loss_mlp": 0.01026854, + "balance_loss_clip": 1.01451182, + "balance_loss_mlp": 1.01965058, + "epoch": 0.5461432092827512, + "flos": 15413866018560.0, + "grad_norm": 1.8258253166739677, + "language_loss": 0.72943151, + "learning_rate": 1.7985883480108834e-06, + "loss": 0.75127083, + "num_input_tokens_seen": 97999585, + "step": 4542, + "time_per_iteration": 2.6586530208587646 + }, + { + "auxiliary_loss_clip": 0.01149649, + "auxiliary_loss_mlp": 0.01023469, + "balance_loss_clip": 1.01428556, + "balance_loss_mlp": 1.01637614, + "epoch": 0.5462634521733902, + "flos": 24024921287040.0, + "grad_norm": 1.5552227554684048, + "language_loss": 0.72125769, + "learning_rate": 1.797813351760371e-06, + "loss": 0.74298888, + "num_input_tokens_seen": 98021290, + "step": 4543, + "time_per_iteration": 2.6630611419677734 + }, + { + "auxiliary_loss_clip": 0.01178105, + "auxiliary_loss_mlp": 0.01028724, + "balance_loss_clip": 1.05464077, + "balance_loss_mlp": 1.02148819, + "epoch": 0.5463836950640293, + "flos": 22820944291200.0, + "grad_norm": 2.190141795982674, + "language_loss": 0.78245872, + "learning_rate": 1.7970383861814116e-06, + "loss": 0.80452704, + "num_input_tokens_seen": 98041060, + "step": 4544, + "time_per_iteration": 2.7378082275390625 + }, + { + "auxiliary_loss_clip": 0.01152091, + "auxiliary_loss_mlp": 0.01031777, + "balance_loss_clip": 1.01592779, + "balance_loss_mlp": 1.0242312, + "epoch": 0.5465039379546685, + "flos": 20448290390400.0, + "grad_norm": 1.9956703608148825, + "language_loss": 0.74518263, + "learning_rate": 1.7962634513915684e-06, + "loss": 0.7670213, + "num_input_tokens_seen": 98058410, + "step": 4545, + "time_per_iteration": 2.6237189769744873 + }, + { + "auxiliary_loss_clip": 0.01172849, + "auxiliary_loss_mlp": 0.01022407, + "balance_loss_clip": 1.05262351, + "balance_loss_mlp": 1.01539123, + "epoch": 0.5466241808453075, + "flos": 17343310003200.0, + "grad_norm": 51.30884663294417, + "language_loss": 0.79329646, + "learning_rate": 1.7954885475083969e-06, + "loss": 0.81524903, + "num_input_tokens_seen": 98076080, + "step": 4546, + "time_per_iteration": 2.6074771881103516 + }, + { + "auxiliary_loss_clip": 0.01178516, + "auxiliary_loss_mlp": 0.01023331, + "balance_loss_clip": 1.05491161, + "balance_loss_mlp": 1.01617587, + "epoch": 0.5467444237359466, + "flos": 21617039122560.0, + "grad_norm": 3.0950926872679716, + "language_loss": 0.72740394, + "learning_rate": 1.7947136746494513e-06, + "loss": 0.74942243, + "num_input_tokens_seen": 98096995, + "step": 4547, + "time_per_iteration": 3.525786876678467 + }, + { + "auxiliary_loss_clip": 0.01150826, + "auxiliary_loss_mlp": 0.01024252, + "balance_loss_clip": 1.01588893, + "balance_loss_mlp": 1.01687324, + "epoch": 0.5468646666265857, + "flos": 24170467196160.0, + "grad_norm": 2.1323622267480067, + "language_loss": 0.88304102, + "learning_rate": 1.793938832932277e-06, + "loss": 0.90479171, + "num_input_tokens_seen": 98115105, + "step": 4548, + "time_per_iteration": 2.685530662536621 + }, + { + "auxiliary_loss_clip": 0.01177058, + "auxiliary_loss_mlp": 0.0102567, + "balance_loss_clip": 1.05420804, + "balance_loss_mlp": 1.01839852, + "epoch": 0.5469849095172248, + "flos": 27527001505920.0, + "grad_norm": 2.289775496339899, + "language_loss": 0.70531327, + "learning_rate": 1.7931640224744185e-06, + "loss": 0.72734058, + "num_input_tokens_seen": 98135655, + "step": 4549, + "time_per_iteration": 2.7529778480529785 + }, + { + "auxiliary_loss_clip": 0.0109347, + "auxiliary_loss_mlp": 0.0102525, + "balance_loss_clip": 0.93209088, + "balance_loss_mlp": 1.01794553, + "epoch": 0.5471051524078638, + "flos": 27964680727680.0, + "grad_norm": 1.6100243807372523, + "language_loss": 0.73642313, + "learning_rate": 1.7923892433934127e-06, + "loss": 0.75761032, + "num_input_tokens_seen": 98156730, + "step": 4550, + "time_per_iteration": 2.7524008750915527 + }, + { + "auxiliary_loss_clip": 0.01130788, + "auxiliary_loss_mlp": 0.00509054, + "balance_loss_clip": 0.97647202, + "balance_loss_mlp": 0.0, + "epoch": 0.547225395298503, + "flos": 18150510389760.0, + "grad_norm": 5.092747767943994, + "language_loss": 0.78955561, + "learning_rate": 1.7916144958067939e-06, + "loss": 0.80595398, + "num_input_tokens_seen": 98174590, + "step": 4551, + "time_per_iteration": 2.712824821472168 + }, + { + "auxiliary_loss_clip": 0.01153762, + "auxiliary_loss_mlp": 0.01025444, + "balance_loss_clip": 1.01458442, + "balance_loss_mlp": 1.01847076, + "epoch": 0.5473456381891421, + "flos": 21361498790400.0, + "grad_norm": 1.603450556454869, + "language_loss": 0.78729904, + "learning_rate": 1.7908397798320905e-06, + "loss": 0.80909115, + "num_input_tokens_seen": 98194325, + "step": 4552, + "time_per_iteration": 2.6416144371032715 + }, + { + "auxiliary_loss_clip": 0.01152141, + "auxiliary_loss_mlp": 0.00509566, + "balance_loss_clip": 1.01496625, + "balance_loss_mlp": 0.0, + "epoch": 0.5474658810797811, + "flos": 19932145908480.0, + "grad_norm": 23.5399619799701, + "language_loss": 0.75081384, + "learning_rate": 1.7900650955868265e-06, + "loss": 0.7674309, + "num_input_tokens_seen": 98213970, + "step": 4553, + "time_per_iteration": 2.740123987197876 + }, + { + "auxiliary_loss_clip": 0.01151126, + "auxiliary_loss_mlp": 0.00509286, + "balance_loss_clip": 1.01586235, + "balance_loss_mlp": 0.0, + "epoch": 0.5475861239704203, + "flos": 50476217264640.0, + "grad_norm": 1.3304152405936733, + "language_loss": 0.76459783, + "learning_rate": 1.7892904431885202e-06, + "loss": 0.78120196, + "num_input_tokens_seen": 98241145, + "step": 4554, + "time_per_iteration": 2.940459728240967 + }, + { + "auxiliary_loss_clip": 0.0108051, + "auxiliary_loss_mlp": 0.01023658, + "balance_loss_clip": 0.89557284, + "balance_loss_mlp": 1.01663971, + "epoch": 0.5477063668610593, + "flos": 20705123612160.0, + "grad_norm": 1.7215387338215076, + "language_loss": 0.75214291, + "learning_rate": 1.788515822754686e-06, + "loss": 0.77318454, + "num_input_tokens_seen": 98261565, + "step": 4555, + "time_per_iteration": 2.7425711154937744 + }, + { + "auxiliary_loss_clip": 0.01105845, + "auxiliary_loss_mlp": 0.01027405, + "balance_loss_clip": 0.935426, + "balance_loss_mlp": 1.02004147, + "epoch": 0.5478266097516984, + "flos": 19609740408960.0, + "grad_norm": 2.711890846383963, + "language_loss": 0.78021431, + "learning_rate": 1.7877412344028335e-06, + "loss": 0.80154681, + "num_input_tokens_seen": 98281370, + "step": 4556, + "time_per_iteration": 2.828493595123291 + }, + { + "auxiliary_loss_clip": 0.01155519, + "auxiliary_loss_mlp": 0.01027308, + "balance_loss_clip": 1.01591635, + "balance_loss_mlp": 1.02005982, + "epoch": 0.5479468526423376, + "flos": 12896599962240.0, + "grad_norm": 2.990304523207029, + "language_loss": 0.77300858, + "learning_rate": 1.7869666782504668e-06, + "loss": 0.79483682, + "num_input_tokens_seen": 98297950, + "step": 4557, + "time_per_iteration": 2.610595226287842 + }, + { + "auxiliary_loss_clip": 0.01117944, + "auxiliary_loss_mlp": 0.01023316, + "balance_loss_clip": 0.97248167, + "balance_loss_mlp": 1.01575863, + "epoch": 0.5480670955329766, + "flos": 18588800142720.0, + "grad_norm": 1.766063623137066, + "language_loss": 0.68534541, + "learning_rate": 1.7861921544150867e-06, + "loss": 0.70675802, + "num_input_tokens_seen": 98316800, + "step": 4558, + "time_per_iteration": 2.663691759109497 + }, + { + "auxiliary_loss_clip": 0.01049156, + "auxiliary_loss_mlp": 0.00509272, + "balance_loss_clip": 0.8614313, + "balance_loss_mlp": 0.0, + "epoch": 0.5481873384236157, + "flos": 15954608338560.0, + "grad_norm": 1.7562660400925187, + "language_loss": 0.7659961, + "learning_rate": 1.7854176630141856e-06, + "loss": 0.78158033, + "num_input_tokens_seen": 98333935, + "step": 4559, + "time_per_iteration": 2.718679666519165 + }, + { + "auxiliary_loss_clip": 0.01179109, + "auxiliary_loss_mlp": 0.01030818, + "balance_loss_clip": 1.05504799, + "balance_loss_mlp": 1.02370143, + "epoch": 0.5483075813142548, + "flos": 22783812606720.0, + "grad_norm": 2.085742703954669, + "language_loss": 0.84448934, + "learning_rate": 1.784643204165255e-06, + "loss": 0.86658859, + "num_input_tokens_seen": 98353255, + "step": 4560, + "time_per_iteration": 2.6642396450042725 + }, + { + "auxiliary_loss_clip": 0.01148414, + "auxiliary_loss_mlp": 0.01027968, + "balance_loss_clip": 1.01711726, + "balance_loss_mlp": 1.0211966, + "epoch": 0.5484278242048939, + "flos": 19317212046720.0, + "grad_norm": 2.18068416673931, + "language_loss": 0.77526838, + "learning_rate": 1.7838687779857783e-06, + "loss": 0.79703224, + "num_input_tokens_seen": 98371130, + "step": 4561, + "time_per_iteration": 2.6382737159729004 + }, + { + "auxiliary_loss_clip": 0.01122457, + "auxiliary_loss_mlp": 0.01027019, + "balance_loss_clip": 0.97516286, + "balance_loss_mlp": 1.01933002, + "epoch": 0.5485480670955329, + "flos": 22816024128000.0, + "grad_norm": 3.006586423987537, + "language_loss": 0.64062423, + "learning_rate": 1.7830943845932366e-06, + "loss": 0.66211897, + "num_input_tokens_seen": 98390455, + "step": 4562, + "time_per_iteration": 2.7289156913757324 + }, + { + "auxiliary_loss_clip": 0.01132585, + "auxiliary_loss_mlp": 0.010277, + "balance_loss_clip": 0.97744727, + "balance_loss_mlp": 1.02030873, + "epoch": 0.5486683099861721, + "flos": 22671304231680.0, + "grad_norm": 1.7234432141337295, + "language_loss": 0.7513324, + "learning_rate": 1.7823200241051044e-06, + "loss": 0.77293527, + "num_input_tokens_seen": 98409370, + "step": 4563, + "time_per_iteration": 3.5907511711120605 + }, + { + "auxiliary_loss_clip": 0.01176341, + "auxiliary_loss_mlp": 0.01030311, + "balance_loss_clip": 1.05473483, + "balance_loss_mlp": 1.02321815, + "epoch": 0.5487885528768112, + "flos": 23149383275520.0, + "grad_norm": 2.064585611274961, + "language_loss": 0.80580235, + "learning_rate": 1.7815456966388513e-06, + "loss": 0.82786882, + "num_input_tokens_seen": 98428465, + "step": 4564, + "time_per_iteration": 3.519960403442383 + }, + { + "auxiliary_loss_clip": 0.01107645, + "auxiliary_loss_mlp": 0.01026936, + "balance_loss_clip": 0.93695414, + "balance_loss_mlp": 1.01966083, + "epoch": 0.5489087957674502, + "flos": 22053928245120.0, + "grad_norm": 2.1432176105827234, + "language_loss": 0.80959964, + "learning_rate": 1.780771402311943e-06, + "loss": 0.83094549, + "num_input_tokens_seen": 98447300, + "step": 4565, + "time_per_iteration": 3.6534628868103027 + }, + { + "auxiliary_loss_clip": 0.01132847, + "auxiliary_loss_mlp": 0.0102727, + "balance_loss_clip": 0.97920227, + "balance_loss_mlp": 1.01967657, + "epoch": 0.5490290386580894, + "flos": 24315977191680.0, + "grad_norm": 2.873031248103744, + "language_loss": 0.78525251, + "learning_rate": 1.7799971412418374e-06, + "loss": 0.80685365, + "num_input_tokens_seen": 98468695, + "step": 4566, + "time_per_iteration": 2.769573450088501 + }, + { + "auxiliary_loss_clip": 0.011057, + "auxiliary_loss_mlp": 0.01030559, + "balance_loss_clip": 0.93777704, + "balance_loss_mlp": 1.02211869, + "epoch": 0.5491492815487284, + "flos": 18294942977280.0, + "grad_norm": 1.978171095684574, + "language_loss": 0.74068415, + "learning_rate": 1.7792229135459918e-06, + "loss": 0.76204669, + "num_input_tokens_seen": 98485345, + "step": 4567, + "time_per_iteration": 2.7115888595581055 + }, + { + "auxiliary_loss_clip": 0.01009606, + "auxiliary_loss_mlp": 0.01006595, + "balance_loss_clip": 0.8906045, + "balance_loss_mlp": 1.00531387, + "epoch": 0.5492695244393675, + "flos": 64550257050240.0, + "grad_norm": 0.728962345099949, + "language_loss": 0.61566126, + "learning_rate": 1.7784487193418538e-06, + "loss": 0.63582337, + "num_input_tokens_seen": 98543195, + "step": 4568, + "time_per_iteration": 3.2076029777526855 + }, + { + "auxiliary_loss_clip": 0.01089696, + "auxiliary_loss_mlp": 0.01025265, + "balance_loss_clip": 0.93067431, + "balance_loss_mlp": 1.01798081, + "epoch": 0.5493897673300067, + "flos": 17379579761280.0, + "grad_norm": 2.2722933491246953, + "language_loss": 0.61283731, + "learning_rate": 1.7776745587468698e-06, + "loss": 0.63398695, + "num_input_tokens_seen": 98560620, + "step": 4569, + "time_per_iteration": 2.6667187213897705 + }, + { + "auxiliary_loss_clip": 0.01173783, + "auxiliary_loss_mlp": 0.01026912, + "balance_loss_clip": 1.05123627, + "balance_loss_mlp": 1.01953864, + "epoch": 0.5495100102206457, + "flos": 19901765980800.0, + "grad_norm": 9.292089575838908, + "language_loss": 0.8184545, + "learning_rate": 1.7769004318784776e-06, + "loss": 0.84046143, + "num_input_tokens_seen": 98578265, + "step": 4570, + "time_per_iteration": 2.6299636363983154 + }, + { + "auxiliary_loss_clip": 0.01153334, + "auxiliary_loss_mlp": 0.01027529, + "balance_loss_clip": 1.01463389, + "balance_loss_mlp": 1.0195899, + "epoch": 0.5496302531112848, + "flos": 16727190992640.0, + "grad_norm": 1.7318086499103615, + "language_loss": 0.80938941, + "learning_rate": 1.776126338854113e-06, + "loss": 0.83119804, + "num_input_tokens_seen": 98596055, + "step": 4571, + "time_per_iteration": 2.610319137573242 + }, + { + "auxiliary_loss_clip": 0.01151562, + "auxiliary_loss_mlp": 0.01027436, + "balance_loss_clip": 1.01944971, + "balance_loss_mlp": 1.02032185, + "epoch": 0.5497504960019239, + "flos": 24572343536640.0, + "grad_norm": 1.5037748711615275, + "language_loss": 0.84458494, + "learning_rate": 1.7753522797912044e-06, + "loss": 0.86637497, + "num_input_tokens_seen": 98616140, + "step": 4572, + "time_per_iteration": 2.716999053955078 + }, + { + "auxiliary_loss_clip": 0.0113556, + "auxiliary_loss_mlp": 0.01025168, + "balance_loss_clip": 0.97641981, + "balance_loss_mlp": 1.01744366, + "epoch": 0.549870738892563, + "flos": 15450494912640.0, + "grad_norm": 2.4492590387908506, + "language_loss": 0.69922829, + "learning_rate": 1.7745782548071765e-06, + "loss": 0.72083557, + "num_input_tokens_seen": 98633035, + "step": 4573, + "time_per_iteration": 3.5575380325317383 + }, + { + "auxiliary_loss_clip": 0.01105595, + "auxiliary_loss_mlp": 0.01029202, + "balance_loss_clip": 0.94419616, + "balance_loss_mlp": 1.02182889, + "epoch": 0.549990981783202, + "flos": 21069114082560.0, + "grad_norm": 1.5897977436028463, + "language_loss": 0.74222147, + "learning_rate": 1.7738042640194482e-06, + "loss": 0.76356941, + "num_input_tokens_seen": 98652700, + "step": 4574, + "time_per_iteration": 2.6804921627044678 + }, + { + "auxiliary_loss_clip": 0.01175849, + "auxiliary_loss_mlp": 0.01030044, + "balance_loss_clip": 1.05337381, + "balance_loss_mlp": 1.02202773, + "epoch": 0.5501112246738411, + "flos": 21395901041280.0, + "grad_norm": 1.4975587094530434, + "language_loss": 0.70246708, + "learning_rate": 1.7730303075454335e-06, + "loss": 0.72452605, + "num_input_tokens_seen": 98671590, + "step": 4575, + "time_per_iteration": 2.6067681312561035 + }, + { + "auxiliary_loss_clip": 0.01109558, + "auxiliary_loss_mlp": 0.01029929, + "balance_loss_clip": 0.93765199, + "balance_loss_mlp": 1.02149534, + "epoch": 0.5502314675644803, + "flos": 17456931699840.0, + "grad_norm": 2.0622253684784337, + "language_loss": 0.84629095, + "learning_rate": 1.7722563855025402e-06, + "loss": 0.86768591, + "num_input_tokens_seen": 98689620, + "step": 4576, + "time_per_iteration": 2.6608097553253174 + }, + { + "auxiliary_loss_clip": 0.01129966, + "auxiliary_loss_mlp": 0.01028773, + "balance_loss_clip": 0.97391021, + "balance_loss_mlp": 1.02158451, + "epoch": 0.5503517104551193, + "flos": 24310410583680.0, + "grad_norm": 2.2769888893522436, + "language_loss": 0.70962912, + "learning_rate": 1.7714824980081721e-06, + "loss": 0.73121649, + "num_input_tokens_seen": 98708915, + "step": 4577, + "time_per_iteration": 2.786165237426758 + }, + { + "auxiliary_loss_clip": 0.011491, + "auxiliary_loss_mlp": 0.01027395, + "balance_loss_clip": 1.01662612, + "balance_loss_mlp": 1.01996219, + "epoch": 0.5504719533457584, + "flos": 22419427086720.0, + "grad_norm": 1.717844938028096, + "language_loss": 0.74032414, + "learning_rate": 1.7707086451797276e-06, + "loss": 0.76208907, + "num_input_tokens_seen": 98729790, + "step": 4578, + "time_per_iteration": 2.656864643096924 + }, + { + "auxiliary_loss_clip": 0.01020687, + "auxiliary_loss_mlp": 0.01000366, + "balance_loss_clip": 0.90971506, + "balance_loss_mlp": 0.99913257, + "epoch": 0.5505921962363975, + "flos": 67294155968640.0, + "grad_norm": 0.6934714174045603, + "language_loss": 0.5228827, + "learning_rate": 1.7699348271345993e-06, + "loss": 0.5430932, + "num_input_tokens_seen": 98792415, + "step": 4579, + "time_per_iteration": 3.360701560974121 + }, + { + "auxiliary_loss_clip": 0.01008311, + "auxiliary_loss_mlp": 0.01001302, + "balance_loss_clip": 0.8749162, + "balance_loss_mlp": 1.00012767, + "epoch": 0.5507124391270366, + "flos": 45685125578880.0, + "grad_norm": 0.7052374049587328, + "language_loss": 0.54421711, + "learning_rate": 1.7691610439901753e-06, + "loss": 0.56431329, + "num_input_tokens_seen": 98855350, + "step": 4580, + "time_per_iteration": 3.364222526550293 + }, + { + "auxiliary_loss_clip": 0.01152533, + "auxiliary_loss_mlp": 0.01023588, + "balance_loss_clip": 1.01364839, + "balance_loss_mlp": 1.01627469, + "epoch": 0.5508326820176757, + "flos": 22273845264000.0, + "grad_norm": 1.671186318679168, + "language_loss": 0.75705093, + "learning_rate": 1.7683872958638367e-06, + "loss": 0.77881217, + "num_input_tokens_seen": 98874230, + "step": 4581, + "time_per_iteration": 2.6590569019317627 + }, + { + "auxiliary_loss_clip": 0.01125125, + "auxiliary_loss_mlp": 0.01027059, + "balance_loss_clip": 0.97374946, + "balance_loss_mlp": 1.01954257, + "epoch": 0.5509529249083148, + "flos": 20012442762240.0, + "grad_norm": 1.9072933511517194, + "language_loss": 0.84542918, + "learning_rate": 1.7676135828729614e-06, + "loss": 0.86695111, + "num_input_tokens_seen": 98893940, + "step": 4582, + "time_per_iteration": 2.7039430141448975 + }, + { + "auxiliary_loss_clip": 0.01153089, + "auxiliary_loss_mlp": 0.01027274, + "balance_loss_clip": 1.01545, + "balance_loss_mlp": 1.01952207, + "epoch": 0.5510731677989539, + "flos": 21834801325440.0, + "grad_norm": 2.2526062434421745, + "language_loss": 0.82784772, + "learning_rate": 1.7668399051349205e-06, + "loss": 0.84965134, + "num_input_tokens_seen": 98913620, + "step": 4583, + "time_per_iteration": 2.6441376209259033 + }, + { + "auxiliary_loss_clip": 0.01103877, + "auxiliary_loss_mlp": 0.01028505, + "balance_loss_clip": 0.93819839, + "balance_loss_mlp": 1.02107263, + "epoch": 0.5511934106895929, + "flos": 21467901853440.0, + "grad_norm": 2.1122161850831125, + "language_loss": 0.83196056, + "learning_rate": 1.766066262767081e-06, + "loss": 0.85328436, + "num_input_tokens_seen": 98931460, + "step": 4584, + "time_per_iteration": 2.741868019104004 + }, + { + "auxiliary_loss_clip": 0.01121628, + "auxiliary_loss_mlp": 0.01023759, + "balance_loss_clip": 0.97527754, + "balance_loss_mlp": 1.01707733, + "epoch": 0.5513136535802321, + "flos": 21068934514560.0, + "grad_norm": 1.9506597524182818, + "language_loss": 0.77400029, + "learning_rate": 1.765292655886803e-06, + "loss": 0.79545414, + "num_input_tokens_seen": 98950105, + "step": 4585, + "time_per_iteration": 2.672057628631592 + }, + { + "auxiliary_loss_clip": 0.01114853, + "auxiliary_loss_mlp": 0.01026951, + "balance_loss_clip": 0.93975925, + "balance_loss_mlp": 1.01992977, + "epoch": 0.5514338964708712, + "flos": 27815004754560.0, + "grad_norm": 1.7703843846529776, + "language_loss": 0.70920795, + "learning_rate": 1.764519084611443e-06, + "loss": 0.73062599, + "num_input_tokens_seen": 98970560, + "step": 4586, + "time_per_iteration": 2.765424966812134 + }, + { + "auxiliary_loss_clip": 0.01128999, + "auxiliary_loss_mlp": 0.01030228, + "balance_loss_clip": 0.9751265, + "balance_loss_mlp": 1.02297735, + "epoch": 0.5515541393615102, + "flos": 21908525990400.0, + "grad_norm": 1.890025005614935, + "language_loss": 0.77957618, + "learning_rate": 1.7637455490583505e-06, + "loss": 0.80116844, + "num_input_tokens_seen": 98989885, + "step": 4587, + "time_per_iteration": 2.6615309715270996 + }, + { + "auxiliary_loss_clip": 0.01151918, + "auxiliary_loss_mlp": 0.01028421, + "balance_loss_clip": 1.01438057, + "balance_loss_mlp": 1.02169168, + "epoch": 0.5516743822521494, + "flos": 20485422074880.0, + "grad_norm": 1.9907782723554925, + "language_loss": 0.78038049, + "learning_rate": 1.7629720493448701e-06, + "loss": 0.80218393, + "num_input_tokens_seen": 99007180, + "step": 4588, + "time_per_iteration": 2.648155927658081 + }, + { + "auxiliary_loss_clip": 0.01135971, + "auxiliary_loss_mlp": 0.01031144, + "balance_loss_clip": 0.97640282, + "balance_loss_mlp": 1.023211, + "epoch": 0.5517946251427884, + "flos": 14940383915520.0, + "grad_norm": 2.4062123264360085, + "language_loss": 0.8493464, + "learning_rate": 1.7621985855883418e-06, + "loss": 0.87101758, + "num_input_tokens_seen": 99023880, + "step": 4589, + "time_per_iteration": 3.84785795211792 + }, + { + "auxiliary_loss_clip": 0.01122838, + "auxiliary_loss_mlp": 0.01027639, + "balance_loss_clip": 0.97540718, + "balance_loss_mlp": 1.01989019, + "epoch": 0.5519148680334275, + "flos": 18404865573120.0, + "grad_norm": 1.8391306521074833, + "language_loss": 0.72474706, + "learning_rate": 1.7614251579060983e-06, + "loss": 0.74625182, + "num_input_tokens_seen": 99042475, + "step": 4590, + "time_per_iteration": 3.568157911300659 + }, + { + "auxiliary_loss_clip": 0.01104287, + "auxiliary_loss_mlp": 0.0102413, + "balance_loss_clip": 0.93684435, + "balance_loss_mlp": 1.01710832, + "epoch": 0.5520351109240667, + "flos": 25113337251840.0, + "grad_norm": 1.5977565969475331, + "language_loss": 0.8474611, + "learning_rate": 1.76065176641547e-06, + "loss": 0.86874521, + "num_input_tokens_seen": 99065185, + "step": 4591, + "time_per_iteration": 3.6560165882110596 + }, + { + "auxiliary_loss_clip": 0.01150979, + "auxiliary_loss_mlp": 0.01027278, + "balance_loss_clip": 1.01284456, + "balance_loss_mlp": 1.02007747, + "epoch": 0.5521553538147057, + "flos": 21069545045760.0, + "grad_norm": 1.7425261324516967, + "language_loss": 0.77819467, + "learning_rate": 1.759878411233777e-06, + "loss": 0.79997718, + "num_input_tokens_seen": 99083645, + "step": 4592, + "time_per_iteration": 2.678755521774292 + }, + { + "auxiliary_loss_clip": 0.01150718, + "auxiliary_loss_mlp": 0.01031194, + "balance_loss_clip": 1.01482677, + "balance_loss_mlp": 1.0238266, + "epoch": 0.5522755967053448, + "flos": 18879999701760.0, + "grad_norm": 2.8347868859791907, + "language_loss": 0.75723422, + "learning_rate": 1.7591050924783388e-06, + "loss": 0.77905333, + "num_input_tokens_seen": 99100835, + "step": 4593, + "time_per_iteration": 2.6248466968536377 + }, + { + "auxiliary_loss_clip": 0.00998336, + "auxiliary_loss_mlp": 0.01012756, + "balance_loss_clip": 0.87110209, + "balance_loss_mlp": 1.01150417, + "epoch": 0.5523958395959839, + "flos": 64675622494080.0, + "grad_norm": 0.8888455962085535, + "language_loss": 0.57886028, + "learning_rate": 1.7583318102664661e-06, + "loss": 0.59897125, + "num_input_tokens_seen": 99168400, + "step": 4594, + "time_per_iteration": 3.3649404048919678 + }, + { + "auxiliary_loss_clip": 0.01150404, + "auxiliary_loss_mlp": 0.01027262, + "balance_loss_clip": 1.00964785, + "balance_loss_mlp": 1.01960301, + "epoch": 0.552516082486623, + "flos": 10889732211840.0, + "grad_norm": 1.949730458740706, + "language_loss": 0.79140365, + "learning_rate": 1.757558564715466e-06, + "loss": 0.81318033, + "num_input_tokens_seen": 99186475, + "step": 4595, + "time_per_iteration": 2.6096668243408203 + }, + { + "auxiliary_loss_clip": 0.01152983, + "auxiliary_loss_mlp": 0.01021164, + "balance_loss_clip": 1.01412058, + "balance_loss_mlp": 1.01387441, + "epoch": 0.552636325377262, + "flos": 22199797376640.0, + "grad_norm": 3.7504864633067188, + "language_loss": 0.74062657, + "learning_rate": 1.7567853559426386e-06, + "loss": 0.76236808, + "num_input_tokens_seen": 99203525, + "step": 4596, + "time_per_iteration": 2.645949125289917 + }, + { + "auxiliary_loss_clip": 0.01150661, + "auxiliary_loss_mlp": 0.01025871, + "balance_loss_clip": 1.01382065, + "balance_loss_mlp": 1.01879609, + "epoch": 0.5527565682679012, + "flos": 23988184652160.0, + "grad_norm": 2.2423877097853246, + "language_loss": 0.75647992, + "learning_rate": 1.7560121840652797e-06, + "loss": 0.77824521, + "num_input_tokens_seen": 99222910, + "step": 4597, + "time_per_iteration": 2.631866455078125 + }, + { + "auxiliary_loss_clip": 0.01094755, + "auxiliary_loss_mlp": 0.01027309, + "balance_loss_clip": 0.93603516, + "balance_loss_mlp": 1.01967978, + "epoch": 0.5528768111585403, + "flos": 19719267955200.0, + "grad_norm": 1.6593275784317905, + "language_loss": 0.6912567, + "learning_rate": 1.7552390492006782e-06, + "loss": 0.71247733, + "num_input_tokens_seen": 99241230, + "step": 4598, + "time_per_iteration": 2.720479726791382 + }, + { + "auxiliary_loss_clip": 0.01086881, + "auxiliary_loss_mlp": 0.00509386, + "balance_loss_clip": 0.89616072, + "balance_loss_mlp": 0.0, + "epoch": 0.5529970540491793, + "flos": 26215975002240.0, + "grad_norm": 1.7875473830217088, + "language_loss": 0.65156305, + "learning_rate": 1.7544659514661184e-06, + "loss": 0.66752577, + "num_input_tokens_seen": 99264320, + "step": 4599, + "time_per_iteration": 2.846583127975464 + }, + { + "auxiliary_loss_clip": 0.01125308, + "auxiliary_loss_mlp": 0.01026994, + "balance_loss_clip": 0.9749105, + "balance_loss_mlp": 1.01954973, + "epoch": 0.5531172969398185, + "flos": 24425971614720.0, + "grad_norm": 1.9493996383126688, + "language_loss": 0.79930675, + "learning_rate": 1.7536928909788786e-06, + "loss": 0.82082987, + "num_input_tokens_seen": 99283625, + "step": 4600, + "time_per_iteration": 3.6267030239105225 + }, + { + "auxiliary_loss_clip": 0.01001277, + "auxiliary_loss_mlp": 0.01001362, + "balance_loss_clip": 0.87142628, + "balance_loss_mlp": 1.00024176, + "epoch": 0.5532375398304575, + "flos": 64907316195840.0, + "grad_norm": 0.8780681654588518, + "language_loss": 0.61996126, + "learning_rate": 1.752919867856231e-06, + "loss": 0.63998765, + "num_input_tokens_seen": 99335270, + "step": 4601, + "time_per_iteration": 3.135740280151367 + }, + { + "auxiliary_loss_clip": 0.011209, + "auxiliary_loss_mlp": 0.01024318, + "balance_loss_clip": 0.97322464, + "balance_loss_mlp": 1.01720142, + "epoch": 0.5533577827210966, + "flos": 19683105937920.0, + "grad_norm": 1.7511489413454955, + "language_loss": 0.78924906, + "learning_rate": 1.7521468822154436e-06, + "loss": 0.81070125, + "num_input_tokens_seen": 99354185, + "step": 4602, + "time_per_iteration": 2.6907894611358643 + }, + { + "auxiliary_loss_clip": 0.01120211, + "auxiliary_loss_mlp": 0.0102441, + "balance_loss_clip": 0.97462714, + "balance_loss_mlp": 1.01738226, + "epoch": 0.5534780256117358, + "flos": 32306496076800.0, + "grad_norm": 2.015236070017491, + "language_loss": 0.75358212, + "learning_rate": 1.751373934173777e-06, + "loss": 0.77502835, + "num_input_tokens_seen": 99376930, + "step": 4603, + "time_per_iteration": 2.7594220638275146 + }, + { + "auxiliary_loss_clip": 0.0117561, + "auxiliary_loss_mlp": 0.01025381, + "balance_loss_clip": 1.05154967, + "balance_loss_mlp": 1.01753736, + "epoch": 0.5535982685023748, + "flos": 23222425582080.0, + "grad_norm": 1.669340977697074, + "language_loss": 0.72997522, + "learning_rate": 1.750601023848487e-06, + "loss": 0.75198507, + "num_input_tokens_seen": 99397655, + "step": 4604, + "time_per_iteration": 2.6547133922576904 + }, + { + "auxiliary_loss_clip": 0.01173817, + "auxiliary_loss_mlp": 0.00509236, + "balance_loss_clip": 1.05384386, + "balance_loss_mlp": 0.0, + "epoch": 0.5537185113930139, + "flos": 24352534258560.0, + "grad_norm": 2.0677698366593584, + "language_loss": 0.73629594, + "learning_rate": 1.749828151356823e-06, + "loss": 0.75312644, + "num_input_tokens_seen": 99417850, + "step": 4605, + "time_per_iteration": 2.6555979251861572 + }, + { + "auxiliary_loss_clip": 0.01128943, + "auxiliary_loss_mlp": 0.01021532, + "balance_loss_clip": 0.97673368, + "balance_loss_mlp": 1.01447523, + "epoch": 0.553838754283653, + "flos": 23549068886400.0, + "grad_norm": 1.53686230973368, + "language_loss": 0.75599891, + "learning_rate": 1.7490553168160297e-06, + "loss": 0.77750367, + "num_input_tokens_seen": 99438920, + "step": 4606, + "time_per_iteration": 2.7586982250213623 + }, + { + "auxiliary_loss_clip": 0.01129518, + "auxiliary_loss_mlp": 0.01026813, + "balance_loss_clip": 0.9771148, + "balance_loss_mlp": 1.02006233, + "epoch": 0.5539589971742921, + "flos": 17275044205440.0, + "grad_norm": 1.7956844038459034, + "language_loss": 0.7644496, + "learning_rate": 1.748282520343345e-06, + "loss": 0.78601289, + "num_input_tokens_seen": 99457950, + "step": 4607, + "time_per_iteration": 2.6668577194213867 + }, + { + "auxiliary_loss_clip": 0.01157851, + "auxiliary_loss_mlp": 0.01027348, + "balance_loss_clip": 1.01514328, + "balance_loss_mlp": 1.01975179, + "epoch": 0.5540792400649311, + "flos": 27564169104000.0, + "grad_norm": 1.853256657797941, + "language_loss": 0.78819323, + "learning_rate": 1.7475097620560023e-06, + "loss": 0.81004524, + "num_input_tokens_seen": 99478015, + "step": 4608, + "time_per_iteration": 2.792294979095459 + }, + { + "auxiliary_loss_clip": 0.01173647, + "auxiliary_loss_mlp": 0.01025343, + "balance_loss_clip": 1.05278802, + "balance_loss_mlp": 1.01821113, + "epoch": 0.5541994829555702, + "flos": 23878657105920.0, + "grad_norm": 1.9932320458329302, + "language_loss": 0.71257842, + "learning_rate": 1.746737042071228e-06, + "loss": 0.73456836, + "num_input_tokens_seen": 99496520, + "step": 4609, + "time_per_iteration": 2.6215007305145264 + }, + { + "auxiliary_loss_clip": 0.0112351, + "auxiliary_loss_mlp": 0.01025966, + "balance_loss_clip": 0.9763785, + "balance_loss_mlp": 1.01818764, + "epoch": 0.5543197258462094, + "flos": 20115721342080.0, + "grad_norm": 1.7896511899738705, + "language_loss": 0.79238707, + "learning_rate": 1.7459643605062424e-06, + "loss": 0.81388175, + "num_input_tokens_seen": 99513780, + "step": 4610, + "time_per_iteration": 2.715287923812866 + }, + { + "auxiliary_loss_clip": 0.0107698, + "auxiliary_loss_mlp": 0.01026199, + "balance_loss_clip": 0.89789963, + "balance_loss_mlp": 1.01783037, + "epoch": 0.5544399687368484, + "flos": 20916565021440.0, + "grad_norm": 1.957564623413701, + "language_loss": 0.80602646, + "learning_rate": 1.745191717478262e-06, + "loss": 0.82705826, + "num_input_tokens_seen": 99532360, + "step": 4611, + "time_per_iteration": 2.712721586227417 + }, + { + "auxiliary_loss_clip": 0.01125843, + "auxiliary_loss_mlp": 0.01028818, + "balance_loss_clip": 0.97897708, + "balance_loss_mlp": 1.02127159, + "epoch": 0.5545602116274875, + "flos": 25518661297920.0, + "grad_norm": 1.6314519785532269, + "language_loss": 0.79476535, + "learning_rate": 1.7444191131044948e-06, + "loss": 0.81631196, + "num_input_tokens_seen": 99552635, + "step": 4612, + "time_per_iteration": 2.8595383167266846 + }, + { + "auxiliary_loss_clip": 0.01128727, + "auxiliary_loss_mlp": 0.01029389, + "balance_loss_clip": 0.97858155, + "balance_loss_mlp": 1.02201033, + "epoch": 0.5546804545181266, + "flos": 20995568985600.0, + "grad_norm": 6.468437328900947, + "language_loss": 0.72901189, + "learning_rate": 1.7436465475021456e-06, + "loss": 0.75059301, + "num_input_tokens_seen": 99572685, + "step": 4613, + "time_per_iteration": 2.715320348739624 + }, + { + "auxiliary_loss_clip": 0.01096836, + "auxiliary_loss_mlp": 0.01026655, + "balance_loss_clip": 0.93669075, + "balance_loss_mlp": 1.01958263, + "epoch": 0.5548006974087657, + "flos": 26833638297600.0, + "grad_norm": 1.7317096936746272, + "language_loss": 0.71303087, + "learning_rate": 1.7428740207884111e-06, + "loss": 0.73426574, + "num_input_tokens_seen": 99593565, + "step": 4614, + "time_per_iteration": 2.7772128582000732 + }, + { + "auxiliary_loss_clip": 0.01086161, + "auxiliary_loss_mlp": 0.01022882, + "balance_loss_clip": 0.90110058, + "balance_loss_mlp": 1.01534808, + "epoch": 0.5549209402994048, + "flos": 33656414031360.0, + "grad_norm": 3.635175382642079, + "language_loss": 0.61127412, + "learning_rate": 1.7421015330804833e-06, + "loss": 0.63236457, + "num_input_tokens_seen": 99613485, + "step": 4615, + "time_per_iteration": 3.8871006965637207 + }, + { + "auxiliary_loss_clip": 0.0117459, + "auxiliary_loss_mlp": 0.01024112, + "balance_loss_clip": 1.05313277, + "balance_loss_mlp": 1.01686454, + "epoch": 0.5550411831900439, + "flos": 23769524609280.0, + "grad_norm": 1.9502966917320173, + "language_loss": 0.72664231, + "learning_rate": 1.7413290844955475e-06, + "loss": 0.74862933, + "num_input_tokens_seen": 99633515, + "step": 4616, + "time_per_iteration": 3.5670666694641113 + }, + { + "auxiliary_loss_clip": 0.01148162, + "auxiliary_loss_mlp": 0.01025999, + "balance_loss_clip": 1.01675463, + "balance_loss_mlp": 1.0187453, + "epoch": 0.555161426080683, + "flos": 21651189978240.0, + "grad_norm": 2.0789445455151894, + "language_loss": 0.78183866, + "learning_rate": 1.7405566751507843e-06, + "loss": 0.80358028, + "num_input_tokens_seen": 99651560, + "step": 4617, + "time_per_iteration": 3.4935948848724365 + }, + { + "auxiliary_loss_clip": 0.01107248, + "auxiliary_loss_mlp": 0.01029178, + "balance_loss_clip": 0.93799055, + "balance_loss_mlp": 1.02211499, + "epoch": 0.555281668971322, + "flos": 49563116605440.0, + "grad_norm": 1.6606256440853628, + "language_loss": 0.67508507, + "learning_rate": 1.7397843051633668e-06, + "loss": 0.69644934, + "num_input_tokens_seen": 99674255, + "step": 4618, + "time_per_iteration": 2.949425458908081 + }, + { + "auxiliary_loss_clip": 0.01152965, + "auxiliary_loss_mlp": 0.01025794, + "balance_loss_clip": 1.01788187, + "balance_loss_mlp": 1.01866531, + "epoch": 0.5554019118619612, + "flos": 20741608851840.0, + "grad_norm": 1.8059619560350553, + "language_loss": 0.71432364, + "learning_rate": 1.739011974650464e-06, + "loss": 0.73611116, + "num_input_tokens_seen": 99693585, + "step": 4619, + "time_per_iteration": 2.633260726928711 + }, + { + "auxiliary_loss_clip": 0.01083806, + "auxiliary_loss_mlp": 0.01026616, + "balance_loss_clip": 0.89726806, + "balance_loss_mlp": 1.01908815, + "epoch": 0.5555221547526003, + "flos": 25483217552640.0, + "grad_norm": 3.999934921373956, + "language_loss": 0.76914287, + "learning_rate": 1.7382396837292365e-06, + "loss": 0.79024708, + "num_input_tokens_seen": 99714045, + "step": 4620, + "time_per_iteration": 2.764658212661743 + }, + { + "auxiliary_loss_clip": 0.0117796, + "auxiliary_loss_mlp": 0.01031068, + "balance_loss_clip": 1.05483222, + "balance_loss_mlp": 1.02321243, + "epoch": 0.5556423976432393, + "flos": 21762513204480.0, + "grad_norm": 1.636969117984041, + "language_loss": 0.73351693, + "learning_rate": 1.737467432516841e-06, + "loss": 0.75560725, + "num_input_tokens_seen": 99734145, + "step": 4621, + "time_per_iteration": 2.7129430770874023 + }, + { + "auxiliary_loss_clip": 0.01127226, + "auxiliary_loss_mlp": 0.01025178, + "balance_loss_clip": 0.97296131, + "balance_loss_mlp": 1.01747131, + "epoch": 0.5557626405338785, + "flos": 24900171989760.0, + "grad_norm": 2.272124800465065, + "language_loss": 0.7457177, + "learning_rate": 1.7366952211304274e-06, + "loss": 0.76724178, + "num_input_tokens_seen": 99751990, + "step": 4622, + "time_per_iteration": 2.7537336349487305 + }, + { + "auxiliary_loss_clip": 0.01120462, + "auxiliary_loss_mlp": 0.01026456, + "balance_loss_clip": 0.97542167, + "balance_loss_mlp": 1.0193038, + "epoch": 0.5558828834245175, + "flos": 18697501676160.0, + "grad_norm": 1.9455199831929022, + "language_loss": 0.83482814, + "learning_rate": 1.735923049687139e-06, + "loss": 0.85629737, + "num_input_tokens_seen": 99768565, + "step": 4623, + "time_per_iteration": 2.663322687149048 + }, + { + "auxiliary_loss_clip": 0.01123807, + "auxiliary_loss_mlp": 0.01026912, + "balance_loss_clip": 0.97366989, + "balance_loss_mlp": 1.01887131, + "epoch": 0.5560031263151566, + "flos": 27272179445760.0, + "grad_norm": 1.4695305163853596, + "language_loss": 0.73920208, + "learning_rate": 1.7351509183041144e-06, + "loss": 0.76070929, + "num_input_tokens_seen": 99788895, + "step": 4624, + "time_per_iteration": 2.7519690990448 + }, + { + "auxiliary_loss_clip": 0.01179495, + "auxiliary_loss_mlp": 0.01032979, + "balance_loss_clip": 1.05532467, + "balance_loss_mlp": 1.02462256, + "epoch": 0.5561233692057957, + "flos": 23403738458880.0, + "grad_norm": 7.134543372568715, + "language_loss": 0.71983552, + "learning_rate": 1.7343788270984852e-06, + "loss": 0.74196029, + "num_input_tokens_seen": 99808035, + "step": 4625, + "time_per_iteration": 3.5308609008789062 + }, + { + "auxiliary_loss_clip": 0.01131586, + "auxiliary_loss_mlp": 0.01029756, + "balance_loss_clip": 0.98145819, + "balance_loss_mlp": 1.02183497, + "epoch": 0.5562436120964348, + "flos": 37670867804160.0, + "grad_norm": 1.7984677159633866, + "language_loss": 0.74759984, + "learning_rate": 1.7336067761873764e-06, + "loss": 0.7692132, + "num_input_tokens_seen": 99830460, + "step": 4626, + "time_per_iteration": 2.810147285461426 + }, + { + "auxiliary_loss_clip": 0.01157561, + "auxiliary_loss_mlp": 0.0102658, + "balance_loss_clip": 1.01405299, + "balance_loss_mlp": 1.01849198, + "epoch": 0.5563638549870739, + "flos": 25155245445120.0, + "grad_norm": 1.97992208960885, + "language_loss": 0.76291937, + "learning_rate": 1.7328347656879076e-06, + "loss": 0.78476077, + "num_input_tokens_seen": 99850320, + "step": 4627, + "time_per_iteration": 2.6653079986572266 + }, + { + "auxiliary_loss_clip": 0.01106734, + "auxiliary_loss_mlp": 0.01025912, + "balance_loss_clip": 0.93711054, + "balance_loss_mlp": 1.01841354, + "epoch": 0.556484097877713, + "flos": 13581810783360.0, + "grad_norm": 2.637250448021229, + "language_loss": 0.6804322, + "learning_rate": 1.7320627957171927e-06, + "loss": 0.70175862, + "num_input_tokens_seen": 99864980, + "step": 4628, + "time_per_iteration": 2.6520416736602783 + }, + { + "auxiliary_loss_clip": 0.01177298, + "auxiliary_loss_mlp": 0.01026888, + "balance_loss_clip": 1.05600667, + "balance_loss_mlp": 1.01993787, + "epoch": 0.5566043407683521, + "flos": 24681368292480.0, + "grad_norm": 1.678815672565956, + "language_loss": 0.81387854, + "learning_rate": 1.7312908663923382e-06, + "loss": 0.83592033, + "num_input_tokens_seen": 99881155, + "step": 4629, + "time_per_iteration": 2.6680331230163574 + }, + { + "auxiliary_loss_clip": 0.01146356, + "auxiliary_loss_mlp": 0.01026331, + "balance_loss_clip": 1.01209068, + "balance_loss_mlp": 1.01841593, + "epoch": 0.5567245836589911, + "flos": 20588161950720.0, + "grad_norm": 1.8259774456184308, + "language_loss": 0.67223275, + "learning_rate": 1.7305189778304463e-06, + "loss": 0.69395959, + "num_input_tokens_seen": 99899330, + "step": 4630, + "time_per_iteration": 2.616853952407837 + }, + { + "auxiliary_loss_clip": 0.0112874, + "auxiliary_loss_mlp": 0.01024689, + "balance_loss_clip": 0.97864747, + "balance_loss_mlp": 1.01708889, + "epoch": 0.5568448265496303, + "flos": 20704189858560.0, + "grad_norm": 1.9051023550043473, + "language_loss": 0.80361545, + "learning_rate": 1.729747130148611e-06, + "loss": 0.82514971, + "num_input_tokens_seen": 99918525, + "step": 4631, + "time_per_iteration": 2.6604228019714355 + }, + { + "auxiliary_loss_clip": 0.01112928, + "auxiliary_loss_mlp": 0.01028438, + "balance_loss_clip": 0.93927324, + "balance_loss_mlp": 1.02003336, + "epoch": 0.5569650694402694, + "flos": 25302910256640.0, + "grad_norm": 1.957715219930627, + "language_loss": 0.77331287, + "learning_rate": 1.7289753234639208e-06, + "loss": 0.79472649, + "num_input_tokens_seen": 99937500, + "step": 4632, + "time_per_iteration": 2.7196621894836426 + }, + { + "auxiliary_loss_clip": 0.01155831, + "auxiliary_loss_mlp": 0.01029964, + "balance_loss_clip": 1.01430559, + "balance_loss_mlp": 1.02179849, + "epoch": 0.5570853123309084, + "flos": 19712623939200.0, + "grad_norm": 1.761541958443923, + "language_loss": 0.76667011, + "learning_rate": 1.7282035578934592e-06, + "loss": 0.78852808, + "num_input_tokens_seen": 99955665, + "step": 4633, + "time_per_iteration": 2.6684823036193848 + }, + { + "auxiliary_loss_clip": 0.01127651, + "auxiliary_loss_mlp": 0.01025177, + "balance_loss_clip": 0.98031723, + "balance_loss_mlp": 1.01710105, + "epoch": 0.5572055552215476, + "flos": 16108091153280.0, + "grad_norm": 1.5737457524884364, + "language_loss": 0.78993434, + "learning_rate": 1.727431833554301e-06, + "loss": 0.81146258, + "num_input_tokens_seen": 99974140, + "step": 4634, + "time_per_iteration": 2.720391035079956 + }, + { + "auxiliary_loss_clip": 0.01064384, + "auxiliary_loss_mlp": 0.01028325, + "balance_loss_clip": 0.85969907, + "balance_loss_mlp": 1.02081513, + "epoch": 0.5573257981121866, + "flos": 17128815937920.0, + "grad_norm": 1.7939174020594082, + "language_loss": 0.77309996, + "learning_rate": 1.7266601505635175e-06, + "loss": 0.79402709, + "num_input_tokens_seen": 99991480, + "step": 4635, + "time_per_iteration": 2.8385281562805176 + }, + { + "auxiliary_loss_clip": 0.01153575, + "auxiliary_loss_mlp": 0.01026331, + "balance_loss_clip": 1.01603293, + "balance_loss_mlp": 1.01810515, + "epoch": 0.5574460410028257, + "flos": 18807029222400.0, + "grad_norm": 2.143578210818521, + "language_loss": 0.75667429, + "learning_rate": 1.7258885090381717e-06, + "loss": 0.77847338, + "num_input_tokens_seen": 100009520, + "step": 4636, + "time_per_iteration": 2.822070598602295 + }, + { + "auxiliary_loss_clip": 0.01130718, + "auxiliary_loss_mlp": 0.01024699, + "balance_loss_clip": 0.97561926, + "balance_loss_mlp": 1.0176146, + "epoch": 0.5575662838934649, + "flos": 29642678530560.0, + "grad_norm": 1.9295616252534689, + "language_loss": 0.78780591, + "learning_rate": 1.7251169090953213e-06, + "loss": 0.80936003, + "num_input_tokens_seen": 100029995, + "step": 4637, + "time_per_iteration": 2.7744436264038086 + }, + { + "auxiliary_loss_clip": 0.01151541, + "auxiliary_loss_mlp": 0.01032874, + "balance_loss_clip": 1.01347733, + "balance_loss_mlp": 1.02470231, + "epoch": 0.5576865267841039, + "flos": 22054466949120.0, + "grad_norm": 2.632141174402891, + "language_loss": 0.76251304, + "learning_rate": 1.7243453508520168e-06, + "loss": 0.78435719, + "num_input_tokens_seen": 100046980, + "step": 4638, + "time_per_iteration": 2.64821720123291 + }, + { + "auxiliary_loss_clip": 0.01132053, + "auxiliary_loss_mlp": 0.01027631, + "balance_loss_clip": 0.97698689, + "balance_loss_mlp": 1.01943541, + "epoch": 0.557806769674743, + "flos": 17196040241280.0, + "grad_norm": 2.1385082346212156, + "language_loss": 0.84598815, + "learning_rate": 1.7235738344253038e-06, + "loss": 0.86758506, + "num_input_tokens_seen": 100060610, + "step": 4639, + "time_per_iteration": 2.6250486373901367 + }, + { + "auxiliary_loss_clip": 0.01156191, + "auxiliary_loss_mlp": 0.01032181, + "balance_loss_clip": 1.02061546, + "balance_loss_mlp": 1.02410448, + "epoch": 0.557927012565382, + "flos": 24712717887360.0, + "grad_norm": 1.8751119247017203, + "language_loss": 0.82468969, + "learning_rate": 1.72280235993222e-06, + "loss": 0.84657335, + "num_input_tokens_seen": 100078915, + "step": 4640, + "time_per_iteration": 3.6711831092834473 + }, + { + "auxiliary_loss_clip": 0.01152871, + "auxiliary_loss_mlp": 0.00510971, + "balance_loss_clip": 1.01715648, + "balance_loss_mlp": 0.0, + "epoch": 0.5580472554560212, + "flos": 16983090460800.0, + "grad_norm": 2.425120450546549, + "language_loss": 0.69470227, + "learning_rate": 1.722030927489798e-06, + "loss": 0.71134079, + "num_input_tokens_seen": 100096195, + "step": 4641, + "time_per_iteration": 2.629934549331665 + }, + { + "auxiliary_loss_clip": 0.01104926, + "auxiliary_loss_mlp": 0.01032022, + "balance_loss_clip": 0.93759692, + "balance_loss_mlp": 1.02428508, + "epoch": 0.5581674983466602, + "flos": 23509100027520.0, + "grad_norm": 1.852511985533989, + "language_loss": 0.74450666, + "learning_rate": 1.7212595372150634e-06, + "loss": 0.76587617, + "num_input_tokens_seen": 100116175, + "step": 4642, + "time_per_iteration": 3.5896220207214355 + }, + { + "auxiliary_loss_clip": 0.0117587, + "auxiliary_loss_mlp": 0.01028921, + "balance_loss_clip": 1.0538156, + "balance_loss_mlp": 1.02157438, + "epoch": 0.5582877412372993, + "flos": 13480291969920.0, + "grad_norm": 4.273388670128079, + "language_loss": 0.72691047, + "learning_rate": 1.720488189225035e-06, + "loss": 0.74895835, + "num_input_tokens_seen": 100133875, + "step": 4643, + "time_per_iteration": 3.4583992958068848 + }, + { + "auxiliary_loss_clip": 0.01155281, + "auxiliary_loss_mlp": 0.01023724, + "balance_loss_clip": 1.01421881, + "balance_loss_mlp": 1.01575518, + "epoch": 0.5584079841279385, + "flos": 21903605827200.0, + "grad_norm": 2.7681317418099427, + "language_loss": 0.79069006, + "learning_rate": 1.7197168836367265e-06, + "loss": 0.81248009, + "num_input_tokens_seen": 100150685, + "step": 4644, + "time_per_iteration": 2.631204605102539 + }, + { + "auxiliary_loss_clip": 0.01151101, + "auxiliary_loss_mlp": 0.00509675, + "balance_loss_clip": 1.01460111, + "balance_loss_mlp": 0.0, + "epoch": 0.5585282270185775, + "flos": 18843550375680.0, + "grad_norm": 1.9668103160597634, + "language_loss": 0.81746566, + "learning_rate": 1.7189456205671433e-06, + "loss": 0.83407342, + "num_input_tokens_seen": 100169530, + "step": 4645, + "time_per_iteration": 2.6457958221435547 + }, + { + "auxiliary_loss_clip": 0.0116169, + "auxiliary_loss_mlp": 0.01029674, + "balance_loss_clip": 1.0163666, + "balance_loss_mlp": 1.02167499, + "epoch": 0.5586484699092166, + "flos": 21868449390720.0, + "grad_norm": 2.0035061638798646, + "language_loss": 0.81989646, + "learning_rate": 1.7181744001332866e-06, + "loss": 0.84181011, + "num_input_tokens_seen": 100188140, + "step": 4646, + "time_per_iteration": 2.6197926998138428 + }, + { + "auxiliary_loss_clip": 0.01177304, + "auxiliary_loss_mlp": 0.01031925, + "balance_loss_clip": 1.05609798, + "balance_loss_mlp": 1.02485847, + "epoch": 0.5587687127998557, + "flos": 22893232412160.0, + "grad_norm": 1.9129292097360664, + "language_loss": 0.63311893, + "learning_rate": 1.7174032224521493e-06, + "loss": 0.65521121, + "num_input_tokens_seen": 100206850, + "step": 4647, + "time_per_iteration": 2.6477127075195312 + }, + { + "auxiliary_loss_clip": 0.01152722, + "auxiliary_loss_mlp": 0.0102946, + "balance_loss_clip": 1.01507354, + "balance_loss_mlp": 1.02178884, + "epoch": 0.5588889556904948, + "flos": 20303067703680.0, + "grad_norm": 1.54389013411666, + "language_loss": 0.69473416, + "learning_rate": 1.7166320876407184e-06, + "loss": 0.71655601, + "num_input_tokens_seen": 100226270, + "step": 4648, + "time_per_iteration": 2.6930394172668457 + }, + { + "auxiliary_loss_clip": 0.01178525, + "auxiliary_loss_mlp": 0.00510197, + "balance_loss_clip": 1.05382907, + "balance_loss_mlp": 0.0, + "epoch": 0.5590091985811338, + "flos": 16472153450880.0, + "grad_norm": 2.183972663327518, + "language_loss": 0.67944002, + "learning_rate": 1.7158609958159742e-06, + "loss": 0.69632727, + "num_input_tokens_seen": 100243675, + "step": 4649, + "time_per_iteration": 2.618320941925049 + }, + { + "auxiliary_loss_clip": 0.01071687, + "auxiliary_loss_mlp": 0.01027661, + "balance_loss_clip": 0.86375701, + "balance_loss_mlp": 1.01968622, + "epoch": 0.559129441471773, + "flos": 14532186781440.0, + "grad_norm": 2.1896571838110552, + "language_loss": 0.78186417, + "learning_rate": 1.7150899470948911e-06, + "loss": 0.80285764, + "num_input_tokens_seen": 100258940, + "step": 4650, + "time_per_iteration": 2.740954637527466 + }, + { + "auxiliary_loss_clip": 0.01035757, + "auxiliary_loss_mlp": 0.01001209, + "balance_loss_clip": 0.94665146, + "balance_loss_mlp": 0.99997562, + "epoch": 0.5592496843624121, + "flos": 60521009852160.0, + "grad_norm": 0.7980016630952507, + "language_loss": 0.56624073, + "learning_rate": 1.7143189415944365e-06, + "loss": 0.58661038, + "num_input_tokens_seen": 100323400, + "step": 4651, + "time_per_iteration": 4.2198545932769775 + }, + { + "auxiliary_loss_clip": 0.01151424, + "auxiliary_loss_mlp": 0.01029428, + "balance_loss_clip": 1.01642287, + "balance_loss_mlp": 1.02147639, + "epoch": 0.5593699272530511, + "flos": 20886256920960.0, + "grad_norm": 4.052498469475523, + "language_loss": 0.76281267, + "learning_rate": 1.7135479794315714e-06, + "loss": 0.78462124, + "num_input_tokens_seen": 100340355, + "step": 4652, + "time_per_iteration": 2.6293084621429443 + }, + { + "auxiliary_loss_clip": 0.01105007, + "auxiliary_loss_mlp": 0.01024292, + "balance_loss_clip": 0.93971908, + "balance_loss_mlp": 1.01682389, + "epoch": 0.5594901701436903, + "flos": 12896743616640.0, + "grad_norm": 1.9349453730871218, + "language_loss": 0.78984714, + "learning_rate": 1.7127770607232502e-06, + "loss": 0.81114012, + "num_input_tokens_seen": 100358900, + "step": 4653, + "time_per_iteration": 2.6982240676879883 + }, + { + "auxiliary_loss_clip": 0.01114355, + "auxiliary_loss_mlp": 0.01027248, + "balance_loss_clip": 0.93911511, + "balance_loss_mlp": 1.01948798, + "epoch": 0.5596104130343293, + "flos": 23112107936640.0, + "grad_norm": 2.3761778117128016, + "language_loss": 0.79659384, + "learning_rate": 1.7120061855864204e-06, + "loss": 0.81800985, + "num_input_tokens_seen": 100378910, + "step": 4654, + "time_per_iteration": 2.743525266647339 + }, + { + "auxiliary_loss_clip": 0.0115412, + "auxiliary_loss_mlp": 0.01025795, + "balance_loss_clip": 1.01750982, + "balance_loss_mlp": 1.01864862, + "epoch": 0.5597306559249684, + "flos": 25957812977280.0, + "grad_norm": 2.120924433013463, + "language_loss": 0.71244943, + "learning_rate": 1.7112353541380233e-06, + "loss": 0.73424864, + "num_input_tokens_seen": 100398770, + "step": 4655, + "time_per_iteration": 2.7247495651245117 + }, + { + "auxiliary_loss_clip": 0.01132033, + "auxiliary_loss_mlp": 0.01028138, + "balance_loss_clip": 0.97931266, + "balance_loss_mlp": 1.01974559, + "epoch": 0.5598508988156076, + "flos": 22492289825280.0, + "grad_norm": 1.564293216906053, + "language_loss": 0.72065097, + "learning_rate": 1.7104645664949931e-06, + "loss": 0.74225271, + "num_input_tokens_seen": 100421240, + "step": 4656, + "time_per_iteration": 2.7123911380767822 + }, + { + "auxiliary_loss_clip": 0.01132011, + "auxiliary_loss_mlp": 0.01029315, + "balance_loss_clip": 0.97423553, + "balance_loss_mlp": 1.02177167, + "epoch": 0.5599711417062466, + "flos": 23112538899840.0, + "grad_norm": 1.9094623938351176, + "language_loss": 0.71625531, + "learning_rate": 1.7096938227742584e-06, + "loss": 0.73786855, + "num_input_tokens_seen": 100442370, + "step": 4657, + "time_per_iteration": 2.6873273849487305 + }, + { + "auxiliary_loss_clip": 0.01178089, + "auxiliary_loss_mlp": 0.01029481, + "balance_loss_clip": 1.05402851, + "balance_loss_mlp": 1.02142882, + "epoch": 0.5600913845968857, + "flos": 22339345714560.0, + "grad_norm": 1.7681928171888313, + "language_loss": 0.8452248, + "learning_rate": 1.70892312309274e-06, + "loss": 0.86730051, + "num_input_tokens_seen": 100460260, + "step": 4658, + "time_per_iteration": 2.753474235534668 + }, + { + "auxiliary_loss_clip": 0.01129904, + "auxiliary_loss_mlp": 0.01030487, + "balance_loss_clip": 0.97297913, + "balance_loss_mlp": 1.02329898, + "epoch": 0.5602116274875248, + "flos": 17633791290240.0, + "grad_norm": 2.1817519199633524, + "language_loss": 0.68804044, + "learning_rate": 1.7081524675673523e-06, + "loss": 0.70964432, + "num_input_tokens_seen": 100475750, + "step": 4659, + "time_per_iteration": 2.658147096633911 + }, + { + "auxiliary_loss_clip": 0.0103896, + "auxiliary_loss_mlp": 0.01001485, + "balance_loss_clip": 0.94606769, + "balance_loss_mlp": 1.00020397, + "epoch": 0.5603318703781639, + "flos": 70115945529600.0, + "grad_norm": 0.7724466610532474, + "language_loss": 0.596416, + "learning_rate": 1.7073818563150026e-06, + "loss": 0.61682045, + "num_input_tokens_seen": 100537830, + "step": 4660, + "time_per_iteration": 3.3385708332061768 + }, + { + "auxiliary_loss_clip": 0.01147479, + "auxiliary_loss_mlp": 0.01025217, + "balance_loss_clip": 1.01263213, + "balance_loss_mlp": 1.0173043, + "epoch": 0.560452113268803, + "flos": 18545850455040.0, + "grad_norm": 5.594673069259543, + "language_loss": 0.86494446, + "learning_rate": 1.7066112894525935e-06, + "loss": 0.88667142, + "num_input_tokens_seen": 100555910, + "step": 4661, + "time_per_iteration": 2.645822525024414 + }, + { + "auxiliary_loss_clip": 0.01122936, + "auxiliary_loss_mlp": 0.01029366, + "balance_loss_clip": 0.97500294, + "balance_loss_mlp": 1.021415, + "epoch": 0.5605723561594421, + "flos": 25264665250560.0, + "grad_norm": 1.5374775249826333, + "language_loss": 0.7289921, + "learning_rate": 1.7058407670970177e-06, + "loss": 0.7505151, + "num_input_tokens_seen": 100577385, + "step": 4662, + "time_per_iteration": 2.7453365325927734 + }, + { + "auxiliary_loss_clip": 0.01157852, + "auxiliary_loss_mlp": 0.01027073, + "balance_loss_clip": 1.01389241, + "balance_loss_mlp": 1.01908612, + "epoch": 0.5606925990500812, + "flos": 20594949621120.0, + "grad_norm": 2.5967784426828318, + "language_loss": 0.61828554, + "learning_rate": 1.7050702893651643e-06, + "loss": 0.64013475, + "num_input_tokens_seen": 100596965, + "step": 4663, + "time_per_iteration": 2.6478524208068848 + }, + { + "auxiliary_loss_clip": 0.01156087, + "auxiliary_loss_mlp": 0.01035496, + "balance_loss_clip": 1.01760101, + "balance_loss_mlp": 1.02691936, + "epoch": 0.5608128419407202, + "flos": 35006044677120.0, + "grad_norm": 2.042726838860577, + "language_loss": 0.75289237, + "learning_rate": 1.7042998563739134e-06, + "loss": 0.77480817, + "num_input_tokens_seen": 100615315, + "step": 4664, + "time_per_iteration": 2.790419578552246 + }, + { + "auxiliary_loss_clip": 0.01138752, + "auxiliary_loss_mlp": 0.01032823, + "balance_loss_clip": 0.97666276, + "balance_loss_mlp": 1.02497339, + "epoch": 0.5609330848313594, + "flos": 24639819235200.0, + "grad_norm": 2.3435537684814345, + "language_loss": 0.71835035, + "learning_rate": 1.703529468240139e-06, + "loss": 0.74006611, + "num_input_tokens_seen": 100634185, + "step": 4665, + "time_per_iteration": 2.6857821941375732 + }, + { + "auxiliary_loss_clip": 0.01127418, + "auxiliary_loss_mlp": 0.01025062, + "balance_loss_clip": 0.97937286, + "balance_loss_mlp": 1.01720619, + "epoch": 0.5610533277219985, + "flos": 18762894385920.0, + "grad_norm": 2.168634628821493, + "language_loss": 0.73540384, + "learning_rate": 1.7027591250807088e-06, + "loss": 0.75692862, + "num_input_tokens_seen": 100651360, + "step": 4666, + "time_per_iteration": 2.67288875579834 + }, + { + "auxiliary_loss_clip": 0.01179706, + "auxiliary_loss_mlp": 0.01031138, + "balance_loss_clip": 1.05419779, + "balance_loss_mlp": 1.02345514, + "epoch": 0.5611735706126375, + "flos": 15012384727680.0, + "grad_norm": 2.81196791209325, + "language_loss": 0.84523243, + "learning_rate": 1.7019888270124825e-06, + "loss": 0.86734086, + "num_input_tokens_seen": 100668525, + "step": 4667, + "time_per_iteration": 3.571185350418091 + }, + { + "auxiliary_loss_clip": 0.01158489, + "auxiliary_loss_mlp": 0.01034154, + "balance_loss_clip": 1.01828647, + "balance_loss_mlp": 1.02637529, + "epoch": 0.5612938135032767, + "flos": 16468167041280.0, + "grad_norm": 1.678829662633479, + "language_loss": 0.81978405, + "learning_rate": 1.7012185741523147e-06, + "loss": 0.84171045, + "num_input_tokens_seen": 100684850, + "step": 4668, + "time_per_iteration": 3.519685745239258 + }, + { + "auxiliary_loss_clip": 0.01180787, + "auxiliary_loss_mlp": 0.01030904, + "balance_loss_clip": 1.05758154, + "balance_loss_mlp": 1.02320921, + "epoch": 0.5614140563939157, + "flos": 25666433850240.0, + "grad_norm": 1.9864951716246082, + "language_loss": 0.62801403, + "learning_rate": 1.7004483666170514e-06, + "loss": 0.65013099, + "num_input_tokens_seen": 100705345, + "step": 4669, + "time_per_iteration": 3.535871744155884 + }, + { + "auxiliary_loss_clip": 0.01153296, + "auxiliary_loss_mlp": 0.01029814, + "balance_loss_clip": 1.01474237, + "balance_loss_mlp": 1.02198815, + "epoch": 0.5615342992845548, + "flos": 24717566223360.0, + "grad_norm": 2.0693990982211528, + "language_loss": 0.8058961, + "learning_rate": 1.699678204523533e-06, + "loss": 0.8277272, + "num_input_tokens_seen": 100725210, + "step": 4670, + "time_per_iteration": 2.667116403579712 + }, + { + "auxiliary_loss_clip": 0.01134733, + "auxiliary_loss_mlp": 0.01031081, + "balance_loss_clip": 0.98052949, + "balance_loss_mlp": 1.02346945, + "epoch": 0.5616545421751938, + "flos": 22015934634240.0, + "grad_norm": 3.018886389625653, + "language_loss": 0.68447614, + "learning_rate": 1.6989080879885918e-06, + "loss": 0.70613426, + "num_input_tokens_seen": 100743070, + "step": 4671, + "time_per_iteration": 2.651587963104248 + }, + { + "auxiliary_loss_clip": 0.01018729, + "auxiliary_loss_mlp": 0.00998615, + "balance_loss_clip": 0.91076148, + "balance_loss_mlp": 0.99739307, + "epoch": 0.561774785065833, + "flos": 53760358690560.0, + "grad_norm": 0.9121386966744519, + "language_loss": 0.60972637, + "learning_rate": 1.6981380171290544e-06, + "loss": 0.6298998, + "num_input_tokens_seen": 100804095, + "step": 4672, + "time_per_iteration": 3.2593774795532227 + }, + { + "auxiliary_loss_clip": 0.01126505, + "auxiliary_loss_mlp": 0.0102541, + "balance_loss_clip": 0.97418118, + "balance_loss_mlp": 1.01792359, + "epoch": 0.5618950279564721, + "flos": 19750007018880.0, + "grad_norm": 1.8121682314278131, + "language_loss": 0.74763989, + "learning_rate": 1.6973679920617396e-06, + "loss": 0.76915902, + "num_input_tokens_seen": 100821630, + "step": 4673, + "time_per_iteration": 2.6087121963500977 + }, + { + "auxiliary_loss_clip": 0.01131595, + "auxiliary_loss_mlp": 0.01028232, + "balance_loss_clip": 0.98110414, + "balance_loss_mlp": 1.02066779, + "epoch": 0.5620152708471111, + "flos": 16800592435200.0, + "grad_norm": 2.093964131233088, + "language_loss": 0.85201567, + "learning_rate": 1.6965980129034603e-06, + "loss": 0.87361395, + "num_input_tokens_seen": 100839015, + "step": 4674, + "time_per_iteration": 2.698181390762329 + }, + { + "auxiliary_loss_clip": 0.01131061, + "auxiliary_loss_mlp": 0.01031865, + "balance_loss_clip": 0.97863376, + "balance_loss_mlp": 1.02396727, + "epoch": 0.5621355137377503, + "flos": 26797799502720.0, + "grad_norm": 1.7467362914455211, + "language_loss": 0.76413649, + "learning_rate": 1.6958280797710209e-06, + "loss": 0.78576577, + "num_input_tokens_seen": 100860940, + "step": 4675, + "time_per_iteration": 2.705002784729004 + }, + { + "auxiliary_loss_clip": 0.0103853, + "auxiliary_loss_mlp": 0.01001217, + "balance_loss_clip": 0.94836438, + "balance_loss_mlp": 1.00002503, + "epoch": 0.5622557566283893, + "flos": 61207046686080.0, + "grad_norm": 0.7132179585819339, + "language_loss": 0.54735911, + "learning_rate": 1.6950581927812198e-06, + "loss": 0.56775653, + "num_input_tokens_seen": 100920510, + "step": 4676, + "time_per_iteration": 3.140871524810791 + }, + { + "auxiliary_loss_clip": 0.01154622, + "auxiliary_loss_mlp": 0.01031186, + "balance_loss_clip": 1.01577449, + "balance_loss_mlp": 1.02383924, + "epoch": 0.5623759995190284, + "flos": 26468534505600.0, + "grad_norm": 2.351169626278764, + "language_loss": 0.78929234, + "learning_rate": 1.6942883520508486e-06, + "loss": 0.81115043, + "num_input_tokens_seen": 100939245, + "step": 4677, + "time_per_iteration": 3.612706422805786 + }, + { + "auxiliary_loss_clip": 0.01155057, + "auxiliary_loss_mlp": 0.01029479, + "balance_loss_clip": 1.0153749, + "balance_loss_mlp": 1.02193904, + "epoch": 0.5624962424096676, + "flos": 19390900798080.0, + "grad_norm": 1.9427386982121855, + "language_loss": 0.77040595, + "learning_rate": 1.693518557696691e-06, + "loss": 0.79225129, + "num_input_tokens_seen": 100958385, + "step": 4678, + "time_per_iteration": 2.6035189628601074 + }, + { + "auxiliary_loss_clip": 0.01152171, + "auxiliary_loss_mlp": 0.01030301, + "balance_loss_clip": 1.01333857, + "balance_loss_mlp": 1.02239728, + "epoch": 0.5626164853003066, + "flos": 20667345482880.0, + "grad_norm": 1.9786741136729924, + "language_loss": 0.89121485, + "learning_rate": 1.6927488098355252e-06, + "loss": 0.91303957, + "num_input_tokens_seen": 100976015, + "step": 4679, + "time_per_iteration": 2.6657800674438477 + }, + { + "auxiliary_loss_clip": 0.01003417, + "auxiliary_loss_mlp": 0.01000332, + "balance_loss_clip": 0.87164283, + "balance_loss_mlp": 0.99912798, + "epoch": 0.5627367281909457, + "flos": 62766071665920.0, + "grad_norm": 0.9112717073326028, + "language_loss": 0.63246024, + "learning_rate": 1.6919791085841201e-06, + "loss": 0.65249777, + "num_input_tokens_seen": 101033425, + "step": 4680, + "time_per_iteration": 3.2587761878967285 + }, + { + "auxiliary_loss_clip": 0.01148008, + "auxiliary_loss_mlp": 0.01027101, + "balance_loss_clip": 1.01315451, + "balance_loss_mlp": 1.01904893, + "epoch": 0.5628569710815848, + "flos": 12787144243200.0, + "grad_norm": 2.2689935664050487, + "language_loss": 0.79284716, + "learning_rate": 1.6912094540592396e-06, + "loss": 0.8145982, + "num_input_tokens_seen": 101048945, + "step": 4681, + "time_per_iteration": 2.5968611240386963 + }, + { + "auxiliary_loss_clip": 0.01153139, + "auxiliary_loss_mlp": 0.01026813, + "balance_loss_clip": 1.0164547, + "balance_loss_mlp": 1.01927352, + "epoch": 0.5629772139722239, + "flos": 13762082165760.0, + "grad_norm": 3.0430342515653357, + "language_loss": 0.80808443, + "learning_rate": 1.6904398463776393e-06, + "loss": 0.82988393, + "num_input_tokens_seen": 101062745, + "step": 4682, + "time_per_iteration": 2.577354907989502 + }, + { + "auxiliary_loss_clip": 0.01156238, + "auxiliary_loss_mlp": 0.01024536, + "balance_loss_clip": 1.01536393, + "balance_loss_mlp": 1.01749659, + "epoch": 0.5630974568628629, + "flos": 21467830026240.0, + "grad_norm": 1.7515776078542369, + "language_loss": 0.73155785, + "learning_rate": 1.6896702856560683e-06, + "loss": 0.75336558, + "num_input_tokens_seen": 101081840, + "step": 4683, + "time_per_iteration": 2.692153215408325 + }, + { + "auxiliary_loss_clip": 0.01104029, + "auxiliary_loss_mlp": 0.01029817, + "balance_loss_clip": 0.93570894, + "balance_loss_mlp": 1.02217591, + "epoch": 0.5632176997535021, + "flos": 14245907385600.0, + "grad_norm": 2.478390382088465, + "language_loss": 0.69154477, + "learning_rate": 1.6889007720112677e-06, + "loss": 0.71288323, + "num_input_tokens_seen": 101099585, + "step": 4684, + "time_per_iteration": 2.686367988586426 + }, + { + "auxiliary_loss_clip": 0.01157637, + "auxiliary_loss_mlp": 0.0102419, + "balance_loss_clip": 1.01766396, + "balance_loss_mlp": 1.01647091, + "epoch": 0.5633379426441412, + "flos": 20812244947200.0, + "grad_norm": 1.6819571706284602, + "language_loss": 0.77361518, + "learning_rate": 1.6881313055599734e-06, + "loss": 0.79543352, + "num_input_tokens_seen": 101119515, + "step": 4685, + "time_per_iteration": 2.681169271469116 + }, + { + "auxiliary_loss_clip": 0.0111714, + "auxiliary_loss_mlp": 0.01028109, + "balance_loss_clip": 0.97206163, + "balance_loss_mlp": 1.01994324, + "epoch": 0.5634581855347802, + "flos": 22600883617920.0, + "grad_norm": 2.1436770443424145, + "language_loss": 0.82046002, + "learning_rate": 1.6873618864189117e-06, + "loss": 0.84191251, + "num_input_tokens_seen": 101135285, + "step": 4686, + "time_per_iteration": 2.7008092403411865 + }, + { + "auxiliary_loss_clip": 0.01155204, + "auxiliary_loss_mlp": 0.01031362, + "balance_loss_clip": 1.0145191, + "balance_loss_mlp": 1.02327919, + "epoch": 0.5635784284254194, + "flos": 21506972872320.0, + "grad_norm": 1.9958747191649537, + "language_loss": 0.78000665, + "learning_rate": 1.686592514704803e-06, + "loss": 0.80187231, + "num_input_tokens_seen": 101152680, + "step": 4687, + "time_per_iteration": 2.6632614135742188 + }, + { + "auxiliary_loss_clip": 0.01128325, + "auxiliary_loss_mlp": 0.01025135, + "balance_loss_clip": 0.97858346, + "balance_loss_mlp": 1.01748157, + "epoch": 0.5636986713160584, + "flos": 19827466698240.0, + "grad_norm": 1.9996712463211535, + "language_loss": 0.70751733, + "learning_rate": 1.685823190534361e-06, + "loss": 0.72905195, + "num_input_tokens_seen": 101170920, + "step": 4688, + "time_per_iteration": 2.678908586502075 + }, + { + "auxiliary_loss_clip": 0.01178993, + "auxiliary_loss_mlp": 0.01025563, + "balance_loss_clip": 1.05309129, + "balance_loss_mlp": 1.01757026, + "epoch": 0.5638189142066975, + "flos": 19792453916160.0, + "grad_norm": 4.21215805118774, + "language_loss": 0.83988404, + "learning_rate": 1.6850539140242907e-06, + "loss": 0.8619296, + "num_input_tokens_seen": 101190180, + "step": 4689, + "time_per_iteration": 2.708300828933716 + }, + { + "auxiliary_loss_clip": 0.01157328, + "auxiliary_loss_mlp": 0.01029026, + "balance_loss_clip": 1.01537907, + "balance_loss_mlp": 1.02140296, + "epoch": 0.5639391570973367, + "flos": 22893771116160.0, + "grad_norm": 1.7495506634841178, + "language_loss": 0.8200655, + "learning_rate": 1.684284685291292e-06, + "loss": 0.84192908, + "num_input_tokens_seen": 101211825, + "step": 4690, + "time_per_iteration": 2.7320973873138428 + }, + { + "auxiliary_loss_clip": 0.01178571, + "auxiliary_loss_mlp": 0.01030027, + "balance_loss_clip": 1.05416083, + "balance_loss_mlp": 1.0219686, + "epoch": 0.5640593999879757, + "flos": 23727077712000.0, + "grad_norm": 2.009590292480757, + "language_loss": 0.81025028, + "learning_rate": 1.683515504452055e-06, + "loss": 0.83233631, + "num_input_tokens_seen": 101229200, + "step": 4691, + "time_per_iteration": 2.734281063079834 + }, + { + "auxiliary_loss_clip": 0.01098736, + "auxiliary_loss_mlp": 0.01031379, + "balance_loss_clip": 0.93447202, + "balance_loss_mlp": 1.02287984, + "epoch": 0.5641796428786148, + "flos": 22710123855360.0, + "grad_norm": 1.5342861388868505, + "language_loss": 0.66044259, + "learning_rate": 1.6827463716232648e-06, + "loss": 0.68174374, + "num_input_tokens_seen": 101249860, + "step": 4692, + "time_per_iteration": 3.661750078201294 + }, + { + "auxiliary_loss_clip": 0.01153932, + "auxiliary_loss_mlp": 0.00509957, + "balance_loss_clip": 1.01563931, + "balance_loss_mlp": 0.0, + "epoch": 0.5642998857692539, + "flos": 19791987039360.0, + "grad_norm": 1.6309980509692148, + "language_loss": 0.75934207, + "learning_rate": 1.6819772869215972e-06, + "loss": 0.77598095, + "num_input_tokens_seen": 101268940, + "step": 4693, + "time_per_iteration": 3.499746322631836 + }, + { + "auxiliary_loss_clip": 0.01137733, + "auxiliary_loss_mlp": 0.01022773, + "balance_loss_clip": 0.97930229, + "balance_loss_mlp": 1.0154537, + "epoch": 0.564420128659893, + "flos": 23185904428800.0, + "grad_norm": 1.8571875043566934, + "language_loss": 0.82210916, + "learning_rate": 1.6812082504637228e-06, + "loss": 0.84371424, + "num_input_tokens_seen": 101290260, + "step": 4694, + "time_per_iteration": 3.5769925117492676 + }, + { + "auxiliary_loss_clip": 0.01152793, + "auxiliary_loss_mlp": 0.01026303, + "balance_loss_clip": 1.01687539, + "balance_loss_mlp": 1.01916194, + "epoch": 0.564540371550532, + "flos": 23258264376960.0, + "grad_norm": 1.4816530728631438, + "language_loss": 0.74388957, + "learning_rate": 1.6804392623663025e-06, + "loss": 0.76568055, + "num_input_tokens_seen": 101311465, + "step": 4695, + "time_per_iteration": 2.6838269233703613 + }, + { + "auxiliary_loss_clip": 0.01148502, + "auxiliary_loss_mlp": 0.01025356, + "balance_loss_clip": 1.01510859, + "balance_loss_mlp": 1.01809323, + "epoch": 0.5646606144411712, + "flos": 25010058672000.0, + "grad_norm": 1.6763229590420934, + "language_loss": 0.78412378, + "learning_rate": 1.6796703227459935e-06, + "loss": 0.80586231, + "num_input_tokens_seen": 101329420, + "step": 4696, + "time_per_iteration": 2.6717774868011475 + }, + { + "auxiliary_loss_clip": 0.01073743, + "auxiliary_loss_mlp": 0.01024421, + "balance_loss_clip": 0.89605296, + "balance_loss_mlp": 1.01679158, + "epoch": 0.5647808573318103, + "flos": 36539645806080.0, + "grad_norm": 1.8386896024432091, + "language_loss": 0.76265442, + "learning_rate": 1.6789014317194407e-06, + "loss": 0.78363603, + "num_input_tokens_seen": 101350900, + "step": 4697, + "time_per_iteration": 2.8947999477386475 + }, + { + "auxiliary_loss_clip": 0.01137824, + "auxiliary_loss_mlp": 0.01028906, + "balance_loss_clip": 0.97846735, + "balance_loss_mlp": 1.02118111, + "epoch": 0.5649011002224493, + "flos": 22528451842560.0, + "grad_norm": 2.1315770016258795, + "language_loss": 0.72926974, + "learning_rate": 1.6781325894032853e-06, + "loss": 0.75093704, + "num_input_tokens_seen": 101369860, + "step": 4698, + "time_per_iteration": 2.658681869506836 + }, + { + "auxiliary_loss_clip": 0.01126101, + "auxiliary_loss_mlp": 0.01024728, + "balance_loss_clip": 0.97811925, + "balance_loss_mlp": 1.01726508, + "epoch": 0.5650213431130885, + "flos": 18515147304960.0, + "grad_norm": 1.8029253928937525, + "language_loss": 0.91748524, + "learning_rate": 1.6773637959141608e-06, + "loss": 0.93899357, + "num_input_tokens_seen": 101386835, + "step": 4699, + "time_per_iteration": 2.716480255126953 + }, + { + "auxiliary_loss_clip": 0.01123021, + "auxiliary_loss_mlp": 0.01028994, + "balance_loss_clip": 0.97642237, + "balance_loss_mlp": 1.02121603, + "epoch": 0.5651415860037275, + "flos": 17526310819200.0, + "grad_norm": 2.5943755042178025, + "language_loss": 0.66738939, + "learning_rate": 1.6765950513686915e-06, + "loss": 0.68890947, + "num_input_tokens_seen": 101404945, + "step": 4700, + "time_per_iteration": 2.7503905296325684 + }, + { + "auxiliary_loss_clip": 0.01085214, + "auxiliary_loss_mlp": 0.01030958, + "balance_loss_clip": 0.89735317, + "balance_loss_mlp": 1.02359724, + "epoch": 0.5652618288943666, + "flos": 25520026014720.0, + "grad_norm": 3.476519110883326, + "language_loss": 0.76817465, + "learning_rate": 1.675826355883496e-06, + "loss": 0.78933632, + "num_input_tokens_seen": 101424160, + "step": 4701, + "time_per_iteration": 2.748203754425049 + }, + { + "auxiliary_loss_clip": 0.01122663, + "auxiliary_loss_mlp": 0.01024111, + "balance_loss_clip": 0.97561234, + "balance_loss_mlp": 1.01648164, + "epoch": 0.5653820717850057, + "flos": 19683105937920.0, + "grad_norm": 2.027594445864541, + "language_loss": 0.79083204, + "learning_rate": 1.6750577095751848e-06, + "loss": 0.81229985, + "num_input_tokens_seen": 101443270, + "step": 4702, + "time_per_iteration": 2.7369158267974854 + }, + { + "auxiliary_loss_clip": 0.0117357, + "auxiliary_loss_mlp": 0.01031152, + "balance_loss_clip": 1.05285895, + "balance_loss_mlp": 1.02428222, + "epoch": 0.5655023146756448, + "flos": 26979722910720.0, + "grad_norm": 1.6363271771036498, + "language_loss": 0.72654319, + "learning_rate": 1.6742891125603605e-06, + "loss": 0.74859047, + "num_input_tokens_seen": 101464175, + "step": 4703, + "time_per_iteration": 3.604862689971924 + }, + { + "auxiliary_loss_clip": 0.0115315, + "auxiliary_loss_mlp": 0.01025609, + "balance_loss_clip": 1.01617837, + "balance_loss_mlp": 1.01792622, + "epoch": 0.5656225575662839, + "flos": 27669351104640.0, + "grad_norm": 1.9044216366821582, + "language_loss": 0.71975839, + "learning_rate": 1.6735205649556185e-06, + "loss": 0.74154603, + "num_input_tokens_seen": 101484045, + "step": 4704, + "time_per_iteration": 2.7041807174682617 + }, + { + "auxiliary_loss_clip": 0.01109466, + "auxiliary_loss_mlp": 0.01032011, + "balance_loss_clip": 0.93755561, + "balance_loss_mlp": 1.02459049, + "epoch": 0.5657428004569229, + "flos": 24349732997760.0, + "grad_norm": 1.9603370818265426, + "language_loss": 0.85008729, + "learning_rate": 1.6727520668775476e-06, + "loss": 0.8715021, + "num_input_tokens_seen": 101504330, + "step": 4705, + "time_per_iteration": 2.7321369647979736 + }, + { + "auxiliary_loss_clip": 0.01178264, + "auxiliary_loss_mlp": 0.01029899, + "balance_loss_clip": 1.0524013, + "balance_loss_mlp": 1.02160168, + "epoch": 0.5658630433475621, + "flos": 21944041562880.0, + "grad_norm": 1.9161295847305053, + "language_loss": 0.75543219, + "learning_rate": 1.6719836184427275e-06, + "loss": 0.77751386, + "num_input_tokens_seen": 101524635, + "step": 4706, + "time_per_iteration": 2.6295671463012695 + }, + { + "auxiliary_loss_clip": 0.01124569, + "auxiliary_loss_mlp": 0.0102577, + "balance_loss_clip": 0.97387069, + "balance_loss_mlp": 1.01871312, + "epoch": 0.5659832862382012, + "flos": 30409012218240.0, + "grad_norm": 2.0215681971799775, + "language_loss": 0.64210129, + "learning_rate": 1.671215219767733e-06, + "loss": 0.66360468, + "num_input_tokens_seen": 101544095, + "step": 4707, + "time_per_iteration": 2.73928165435791 + }, + { + "auxiliary_loss_clip": 0.0108707, + "auxiliary_loss_mlp": 0.01026981, + "balance_loss_clip": 0.90029323, + "balance_loss_mlp": 1.01963198, + "epoch": 0.5661035291288402, + "flos": 13188194570880.0, + "grad_norm": 3.7140872229496757, + "language_loss": 0.76058632, + "learning_rate": 1.670446870969127e-06, + "loss": 0.78172684, + "num_input_tokens_seen": 101561760, + "step": 4708, + "time_per_iteration": 2.7416861057281494 + }, + { + "auxiliary_loss_clip": 0.01132519, + "auxiliary_loss_mlp": 0.01032161, + "balance_loss_clip": 0.97601706, + "balance_loss_mlp": 1.02413225, + "epoch": 0.5662237720194794, + "flos": 16143032108160.0, + "grad_norm": 2.090744927189821, + "language_loss": 0.80224776, + "learning_rate": 1.6696785721634685e-06, + "loss": 0.8238945, + "num_input_tokens_seen": 101576245, + "step": 4709, + "time_per_iteration": 2.6349382400512695 + }, + { + "auxiliary_loss_clip": 0.01155948, + "auxiliary_loss_mlp": 0.01028823, + "balance_loss_clip": 1.01550615, + "balance_loss_mlp": 1.02096748, + "epoch": 0.5663440149101184, + "flos": 17676848718720.0, + "grad_norm": 1.7888428637883396, + "language_loss": 0.74401665, + "learning_rate": 1.6689103234673086e-06, + "loss": 0.76586437, + "num_input_tokens_seen": 101594565, + "step": 4710, + "time_per_iteration": 2.590665817260742 + }, + { + "auxiliary_loss_clip": 0.01130197, + "auxiliary_loss_mlp": 0.01028896, + "balance_loss_clip": 0.97734344, + "balance_loss_mlp": 1.02106965, + "epoch": 0.5664642578007575, + "flos": 23368330627200.0, + "grad_norm": 1.8145716406136976, + "language_loss": 0.77195686, + "learning_rate": 1.668142124997189e-06, + "loss": 0.79354787, + "num_input_tokens_seen": 101614225, + "step": 4711, + "time_per_iteration": 2.7304329872131348 + }, + { + "auxiliary_loss_clip": 0.0103697, + "auxiliary_loss_mlp": 0.0100291, + "balance_loss_clip": 0.95297301, + "balance_loss_mlp": 1.0016284, + "epoch": 0.5665845006913967, + "flos": 65516470945920.0, + "grad_norm": 0.7262125418555527, + "language_loss": 0.5975219, + "learning_rate": 1.6673739768696453e-06, + "loss": 0.61792076, + "num_input_tokens_seen": 101680795, + "step": 4712, + "time_per_iteration": 3.253016471862793 + }, + { + "auxiliary_loss_clip": 0.01135412, + "auxiliary_loss_mlp": 0.0103046, + "balance_loss_clip": 0.97580278, + "balance_loss_mlp": 1.02254486, + "epoch": 0.5667047435820357, + "flos": 26140885620480.0, + "grad_norm": 1.6382917711927485, + "language_loss": 0.77770841, + "learning_rate": 1.6666058792012052e-06, + "loss": 0.79936719, + "num_input_tokens_seen": 101701680, + "step": 4713, + "time_per_iteration": 2.766463279724121 + }, + { + "auxiliary_loss_clip": 0.01056463, + "auxiliary_loss_mlp": 0.0100573, + "balance_loss_clip": 0.98137057, + "balance_loss_mlp": 1.00460327, + "epoch": 0.5668249864726748, + "flos": 71866949725440.0, + "grad_norm": 0.8764138054327124, + "language_loss": 0.68792343, + "learning_rate": 1.6658378321083878e-06, + "loss": 0.70854533, + "num_input_tokens_seen": 101766010, + "step": 4714, + "time_per_iteration": 3.2724320888519287 + }, + { + "auxiliary_loss_clip": 0.01065313, + "auxiliary_loss_mlp": 0.01033442, + "balance_loss_clip": 0.86176002, + "balance_loss_mlp": 1.02580059, + "epoch": 0.5669452293633139, + "flos": 22195667312640.0, + "grad_norm": 1.698785551676891, + "language_loss": 0.82344818, + "learning_rate": 1.6650698357077055e-06, + "loss": 0.84443569, + "num_input_tokens_seen": 101783055, + "step": 4715, + "time_per_iteration": 2.7466881275177 + }, + { + "auxiliary_loss_clip": 0.0113141, + "auxiliary_loss_mlp": 0.01028955, + "balance_loss_clip": 0.97424591, + "balance_loss_mlp": 1.02164745, + "epoch": 0.567065472253953, + "flos": 18223193560320.0, + "grad_norm": 2.6490973782606178, + "language_loss": 0.80973059, + "learning_rate": 1.6643018901156632e-06, + "loss": 0.83133429, + "num_input_tokens_seen": 101802150, + "step": 4716, + "time_per_iteration": 2.715857982635498 + }, + { + "auxiliary_loss_clip": 0.01131403, + "auxiliary_loss_mlp": 0.01027677, + "balance_loss_clip": 0.97463775, + "balance_loss_mlp": 1.02033973, + "epoch": 0.567185715144592, + "flos": 20371548983040.0, + "grad_norm": 2.4417492240887655, + "language_loss": 0.79153097, + "learning_rate": 1.6635339954487566e-06, + "loss": 0.8131218, + "num_input_tokens_seen": 101818025, + "step": 4717, + "time_per_iteration": 2.693769693374634 + }, + { + "auxiliary_loss_clip": 0.01135093, + "auxiliary_loss_mlp": 0.01026445, + "balance_loss_clip": 0.97808176, + "balance_loss_mlp": 1.01873553, + "epoch": 0.5673059580352312, + "flos": 23221348174080.0, + "grad_norm": 2.621506682829713, + "language_loss": 0.82266319, + "learning_rate": 1.6627661518234765e-06, + "loss": 0.84427857, + "num_input_tokens_seen": 101837280, + "step": 4718, + "time_per_iteration": 3.6097137928009033 + }, + { + "auxiliary_loss_clip": 0.01086072, + "auxiliary_loss_mlp": 0.01033421, + "balance_loss_clip": 0.90278214, + "balance_loss_mlp": 1.02556539, + "epoch": 0.5674262009258703, + "flos": 21719599430400.0, + "grad_norm": 1.5353982930206713, + "language_loss": 0.85236722, + "learning_rate": 1.661998359356302e-06, + "loss": 0.87356216, + "num_input_tokens_seen": 101856310, + "step": 4719, + "time_per_iteration": 3.7012343406677246 + }, + { + "auxiliary_loss_clip": 0.0107435, + "auxiliary_loss_mlp": 0.01003482, + "balance_loss_clip": 1.01850176, + "balance_loss_mlp": 1.00239682, + "epoch": 0.5675464438165093, + "flos": 67470369114240.0, + "grad_norm": 0.7415977702998282, + "language_loss": 0.55763751, + "learning_rate": 1.6612306181637077e-06, + "loss": 0.57841581, + "num_input_tokens_seen": 101915635, + "step": 4720, + "time_per_iteration": 4.075162172317505 + }, + { + "auxiliary_loss_clip": 0.01104155, + "auxiliary_loss_mlp": 0.01024536, + "balance_loss_clip": 0.93659055, + "balance_loss_mlp": 1.01670432, + "epoch": 0.5676666867071485, + "flos": 18879173688960.0, + "grad_norm": 2.2418209948663628, + "language_loss": 0.6576755, + "learning_rate": 1.6604629283621598e-06, + "loss": 0.67896241, + "num_input_tokens_seen": 101933565, + "step": 4721, + "time_per_iteration": 2.7111639976501465 + }, + { + "auxiliary_loss_clip": 0.01178783, + "auxiliary_loss_mlp": 0.01034031, + "balance_loss_clip": 1.05342913, + "balance_loss_mlp": 1.02611506, + "epoch": 0.5677869295977875, + "flos": 33546778744320.0, + "grad_norm": 1.885952749222743, + "language_loss": 0.7467643, + "learning_rate": 1.6596952900681152e-06, + "loss": 0.76889241, + "num_input_tokens_seen": 101954325, + "step": 4722, + "time_per_iteration": 2.7129435539245605 + }, + { + "auxiliary_loss_clip": 0.01074095, + "auxiliary_loss_mlp": 0.01033113, + "balance_loss_clip": 0.9002558, + "balance_loss_mlp": 1.0253284, + "epoch": 0.5679071724884266, + "flos": 28037256157440.0, + "grad_norm": 2.072754470955852, + "language_loss": 0.82197416, + "learning_rate": 1.658927703398025e-06, + "loss": 0.84304619, + "num_input_tokens_seen": 101974390, + "step": 4723, + "time_per_iteration": 2.859177350997925 + }, + { + "auxiliary_loss_clip": 0.0108021, + "auxiliary_loss_mlp": 0.01028097, + "balance_loss_clip": 0.89463997, + "balance_loss_mlp": 1.02000284, + "epoch": 0.5680274153790658, + "flos": 23550110380800.0, + "grad_norm": 2.2458644655245816, + "language_loss": 0.78077513, + "learning_rate": 1.6581601684683309e-06, + "loss": 0.80185825, + "num_input_tokens_seen": 101994815, + "step": 4724, + "time_per_iteration": 2.8080220222473145 + }, + { + "auxiliary_loss_clip": 0.01152583, + "auxiliary_loss_mlp": 0.01026171, + "balance_loss_clip": 1.01494575, + "balance_loss_mlp": 1.01838398, + "epoch": 0.5681476582697048, + "flos": 22455158140800.0, + "grad_norm": 2.359754332307469, + "language_loss": 0.68974322, + "learning_rate": 1.6573926853954674e-06, + "loss": 0.71153075, + "num_input_tokens_seen": 102012400, + "step": 4725, + "time_per_iteration": 2.629697322845459 + }, + { + "auxiliary_loss_clip": 0.01122584, + "auxiliary_loss_mlp": 0.01027983, + "balance_loss_clip": 0.97284287, + "balance_loss_mlp": 1.02061582, + "epoch": 0.5682679011603439, + "flos": 19536913584000.0, + "grad_norm": 1.6520752142233923, + "language_loss": 0.82901585, + "learning_rate": 1.6566252542958608e-06, + "loss": 0.85052156, + "num_input_tokens_seen": 102031900, + "step": 4726, + "time_per_iteration": 2.6898930072784424 + }, + { + "auxiliary_loss_clip": 0.01099147, + "auxiliary_loss_mlp": 0.01028399, + "balance_loss_clip": 0.93809164, + "balance_loss_mlp": 1.02065659, + "epoch": 0.568388144050983, + "flos": 28765488493440.0, + "grad_norm": 2.0180181904886463, + "language_loss": 0.78471518, + "learning_rate": 1.6558578752859305e-06, + "loss": 0.80599064, + "num_input_tokens_seen": 102050860, + "step": 4727, + "time_per_iteration": 2.7628746032714844 + }, + { + "auxiliary_loss_clip": 0.01104102, + "auxiliary_loss_mlp": 0.01025512, + "balance_loss_clip": 0.93576181, + "balance_loss_mlp": 1.01753736, + "epoch": 0.5685083869416221, + "flos": 21209452519680.0, + "grad_norm": 1.83434536166416, + "language_loss": 0.78834748, + "learning_rate": 1.6550905484820865e-06, + "loss": 0.80964363, + "num_input_tokens_seen": 102069320, + "step": 4728, + "time_per_iteration": 2.735320568084717 + }, + { + "auxiliary_loss_clip": 0.01178169, + "auxiliary_loss_mlp": 0.0103058, + "balance_loss_clip": 1.05252361, + "balance_loss_mlp": 1.02278924, + "epoch": 0.5686286298322611, + "flos": 24827021942400.0, + "grad_norm": 3.2016925094156687, + "language_loss": 0.79088736, + "learning_rate": 1.6543232740007328e-06, + "loss": 0.81297493, + "num_input_tokens_seen": 102086435, + "step": 4729, + "time_per_iteration": 2.6325063705444336 + }, + { + "auxiliary_loss_clip": 0.01153919, + "auxiliary_loss_mlp": 0.01028121, + "balance_loss_clip": 1.01424837, + "balance_loss_mlp": 1.02012825, + "epoch": 0.5687488727229003, + "flos": 26615121909120.0, + "grad_norm": 2.0792959740806842, + "language_loss": 0.66941833, + "learning_rate": 1.653556051958263e-06, + "loss": 0.6912387, + "num_input_tokens_seen": 102106115, + "step": 4730, + "time_per_iteration": 3.616194248199463 + }, + { + "auxiliary_loss_clip": 0.0104364, + "auxiliary_loss_mlp": 0.01025369, + "balance_loss_clip": 0.85927629, + "balance_loss_mlp": 1.01772225, + "epoch": 0.5688691156135394, + "flos": 20808725414400.0, + "grad_norm": 1.5966084912661944, + "language_loss": 0.73437941, + "learning_rate": 1.6527888824710642e-06, + "loss": 0.75506949, + "num_input_tokens_seen": 102125715, + "step": 4731, + "time_per_iteration": 2.8080289363861084 + }, + { + "auxiliary_loss_clip": 0.01098353, + "auxiliary_loss_mlp": 0.01029833, + "balance_loss_clip": 0.93377173, + "balance_loss_mlp": 1.0219059, + "epoch": 0.5689893585041784, + "flos": 25880963829120.0, + "grad_norm": 2.897905864838149, + "language_loss": 0.76538873, + "learning_rate": 1.6520217656555166e-06, + "loss": 0.78667057, + "num_input_tokens_seen": 102145005, + "step": 4732, + "time_per_iteration": 2.720452308654785 + }, + { + "auxiliary_loss_clip": 0.01117665, + "auxiliary_loss_mlp": 0.01028297, + "balance_loss_clip": 0.97285908, + "balance_loss_mlp": 1.02037537, + "epoch": 0.5691096013948175, + "flos": 23477463123840.0, + "grad_norm": 1.4465780815713187, + "language_loss": 0.70967656, + "learning_rate": 1.65125470162799e-06, + "loss": 0.7311362, + "num_input_tokens_seen": 102165360, + "step": 4733, + "time_per_iteration": 2.7277164459228516 + }, + { + "auxiliary_loss_clip": 0.01109929, + "auxiliary_loss_mlp": 0.01030853, + "balance_loss_clip": 0.93756092, + "balance_loss_mlp": 1.02266383, + "epoch": 0.5692298442854566, + "flos": 18075600576000.0, + "grad_norm": 2.665161801062134, + "language_loss": 0.69675517, + "learning_rate": 1.6504876905048485e-06, + "loss": 0.71816301, + "num_input_tokens_seen": 102182320, + "step": 4734, + "time_per_iteration": 2.749859571456909 + }, + { + "auxiliary_loss_clip": 0.01176349, + "auxiliary_loss_mlp": 0.01029865, + "balance_loss_clip": 1.05472147, + "balance_loss_mlp": 1.02264678, + "epoch": 0.5693500871760957, + "flos": 23039317025280.0, + "grad_norm": 1.6425992132784593, + "language_loss": 0.72421086, + "learning_rate": 1.6497207324024464e-06, + "loss": 0.74627304, + "num_input_tokens_seen": 102201220, + "step": 4735, + "time_per_iteration": 2.6477646827697754 + }, + { + "auxiliary_loss_clip": 0.01133643, + "auxiliary_loss_mlp": 0.0103127, + "balance_loss_clip": 0.97449154, + "balance_loss_mlp": 1.02343225, + "epoch": 0.5694703300667348, + "flos": 18989670902400.0, + "grad_norm": 1.9443655343299722, + "language_loss": 0.82672179, + "learning_rate": 1.6489538274371305e-06, + "loss": 0.84837091, + "num_input_tokens_seen": 102219825, + "step": 4736, + "time_per_iteration": 2.6577184200286865 + }, + { + "auxiliary_loss_clip": 0.01146138, + "auxiliary_loss_mlp": 0.01026222, + "balance_loss_clip": 1.01308322, + "balance_loss_mlp": 1.01852751, + "epoch": 0.5695905729573739, + "flos": 21908705558400.0, + "grad_norm": 2.040959736769105, + "language_loss": 0.8296802, + "learning_rate": 1.6481869757252396e-06, + "loss": 0.85140377, + "num_input_tokens_seen": 102238160, + "step": 4737, + "time_per_iteration": 2.6428046226501465 + }, + { + "auxiliary_loss_clip": 0.01152111, + "auxiliary_loss_mlp": 0.01027056, + "balance_loss_clip": 1.01449251, + "balance_loss_mlp": 1.01952219, + "epoch": 0.569710815848013, + "flos": 28476659232000.0, + "grad_norm": 1.3677258260655791, + "language_loss": 0.71974993, + "learning_rate": 1.647420177383105e-06, + "loss": 0.74154156, + "num_input_tokens_seen": 102261030, + "step": 4738, + "time_per_iteration": 2.7001333236694336 + }, + { + "auxiliary_loss_clip": 0.01150061, + "auxiliary_loss_mlp": 0.01024562, + "balance_loss_clip": 1.01740146, + "balance_loss_mlp": 1.01678634, + "epoch": 0.569831058738652, + "flos": 28366162018560.0, + "grad_norm": 1.7361381094079376, + "language_loss": 0.72774029, + "learning_rate": 1.646653432527049e-06, + "loss": 0.74948651, + "num_input_tokens_seen": 102281670, + "step": 4739, + "time_per_iteration": 2.6788690090179443 + }, + { + "auxiliary_loss_clip": 0.01107799, + "auxiliary_loss_mlp": 0.01030099, + "balance_loss_clip": 0.93873864, + "balance_loss_mlp": 1.0226928, + "epoch": 0.5699513016292912, + "flos": 25849973370240.0, + "grad_norm": 1.4575569439463656, + "language_loss": 0.74415183, + "learning_rate": 1.645886741273387e-06, + "loss": 0.76553082, + "num_input_tokens_seen": 102303485, + "step": 4740, + "time_per_iteration": 2.715315103530884 + }, + { + "auxiliary_loss_clip": 0.01106047, + "auxiliary_loss_mlp": 0.01030013, + "balance_loss_clip": 0.94365627, + "balance_loss_mlp": 1.02240145, + "epoch": 0.5700715445199303, + "flos": 18037858360320.0, + "grad_norm": 1.9397922064066486, + "language_loss": 0.73839217, + "learning_rate": 1.645120103738424e-06, + "loss": 0.75975275, + "num_input_tokens_seen": 102320995, + "step": 4741, + "time_per_iteration": 2.677855968475342 + }, + { + "auxiliary_loss_clip": 0.0114312, + "auxiliary_loss_mlp": 0.00509477, + "balance_loss_clip": 1.01359773, + "balance_loss_mlp": 0.0, + "epoch": 0.5701917874105693, + "flos": 11473352392320.0, + "grad_norm": 2.1682356918209242, + "language_loss": 0.84093833, + "learning_rate": 1.6443535200384591e-06, + "loss": 0.85746431, + "num_input_tokens_seen": 102339170, + "step": 4742, + "time_per_iteration": 2.668478012084961 + }, + { + "auxiliary_loss_clip": 0.01179189, + "auxiliary_loss_mlp": 0.01028746, + "balance_loss_clip": 1.05703068, + "balance_loss_mlp": 1.02121842, + "epoch": 0.5703120303012085, + "flos": 21761759018880.0, + "grad_norm": 1.5484143749239636, + "language_loss": 0.71019608, + "learning_rate": 1.6435869902897827e-06, + "loss": 0.73227543, + "num_input_tokens_seen": 102357750, + "step": 4743, + "time_per_iteration": 2.6189870834350586 + }, + { + "auxiliary_loss_clip": 0.01014577, + "auxiliary_loss_mlp": 0.01005732, + "balance_loss_clip": 0.90848643, + "balance_loss_mlp": 1.0046351, + "epoch": 0.5704322731918475, + "flos": 56746258513920.0, + "grad_norm": 0.7949635174736426, + "language_loss": 0.62031662, + "learning_rate": 1.6428205146086764e-06, + "loss": 0.64051974, + "num_input_tokens_seen": 102419730, + "step": 4744, + "time_per_iteration": 4.212113380432129 + }, + { + "auxiliary_loss_clip": 0.01135165, + "auxiliary_loss_mlp": 0.01032172, + "balance_loss_clip": 0.97625452, + "balance_loss_mlp": 1.02411926, + "epoch": 0.5705525160824866, + "flos": 20741141975040.0, + "grad_norm": 1.6168951682127335, + "language_loss": 0.70830953, + "learning_rate": 1.6420540931114142e-06, + "loss": 0.72998297, + "num_input_tokens_seen": 102440320, + "step": 4745, + "time_per_iteration": 3.712674140930176 + }, + { + "auxiliary_loss_clip": 0.01135484, + "auxiliary_loss_mlp": 0.01021817, + "balance_loss_clip": 0.9788599, + "balance_loss_mlp": 1.01378512, + "epoch": 0.5706727589731257, + "flos": 18771262254720.0, + "grad_norm": 1.7185260006501295, + "language_loss": 0.79316878, + "learning_rate": 1.6412877259142616e-06, + "loss": 0.81474185, + "num_input_tokens_seen": 102460240, + "step": 4746, + "time_per_iteration": 2.6609466075897217 + }, + { + "auxiliary_loss_clip": 0.01127814, + "auxiliary_loss_mlp": 0.01029737, + "balance_loss_clip": 0.97675908, + "balance_loss_mlp": 1.02229261, + "epoch": 0.5707930018637648, + "flos": 27634733372160.0, + "grad_norm": 2.409830148006864, + "language_loss": 0.74053067, + "learning_rate": 1.6405214131334757e-06, + "loss": 0.76210618, + "num_input_tokens_seen": 102478765, + "step": 4747, + "time_per_iteration": 3.9645016193389893 + }, + { + "auxiliary_loss_clip": 0.01078863, + "auxiliary_loss_mlp": 0.01032671, + "balance_loss_clip": 0.90119153, + "balance_loss_mlp": 1.02510381, + "epoch": 0.5709132447544039, + "flos": 27597673514880.0, + "grad_norm": 2.2163202894472533, + "language_loss": 0.79530048, + "learning_rate": 1.6397551548853052e-06, + "loss": 0.81641579, + "num_input_tokens_seen": 102496930, + "step": 4748, + "time_per_iteration": 2.8816802501678467 + }, + { + "auxiliary_loss_clip": 0.01131409, + "auxiliary_loss_mlp": 0.01029571, + "balance_loss_clip": 0.97866201, + "balance_loss_mlp": 1.0219295, + "epoch": 0.571033487645043, + "flos": 21686095019520.0, + "grad_norm": 1.6851045029703244, + "language_loss": 0.70942181, + "learning_rate": 1.6389889512859917e-06, + "loss": 0.7310316, + "num_input_tokens_seen": 102516590, + "step": 4749, + "time_per_iteration": 2.684197425842285 + }, + { + "auxiliary_loss_clip": 0.01030895, + "auxiliary_loss_mlp": 0.01000892, + "balance_loss_clip": 0.94170892, + "balance_loss_mlp": 0.999825, + "epoch": 0.5711537305356821, + "flos": 70181445980160.0, + "grad_norm": 0.8305943088270441, + "language_loss": 0.60342962, + "learning_rate": 1.638222802451767e-06, + "loss": 0.62374747, + "num_input_tokens_seen": 102578070, + "step": 4750, + "time_per_iteration": 3.239224433898926 + }, + { + "auxiliary_loss_clip": 0.01143689, + "auxiliary_loss_mlp": 0.0102553, + "balance_loss_clip": 1.01309729, + "balance_loss_mlp": 1.01824379, + "epoch": 0.5712739734263211, + "flos": 24717494396160.0, + "grad_norm": 1.5715057126278236, + "language_loss": 0.75234258, + "learning_rate": 1.6374567084988561e-06, + "loss": 0.77403474, + "num_input_tokens_seen": 102599255, + "step": 4751, + "time_per_iteration": 2.669642448425293 + }, + { + "auxiliary_loss_clip": 0.01135961, + "auxiliary_loss_mlp": 0.01028325, + "balance_loss_clip": 0.97944736, + "balance_loss_mlp": 1.0203979, + "epoch": 0.5713942163169603, + "flos": 26578169792640.0, + "grad_norm": 1.9458186112022893, + "language_loss": 0.76509535, + "learning_rate": 1.6366906695434738e-06, + "loss": 0.78673828, + "num_input_tokens_seen": 102621775, + "step": 4752, + "time_per_iteration": 2.7674174308776855 + }, + { + "auxiliary_loss_clip": 0.01155292, + "auxiliary_loss_mlp": 0.01031044, + "balance_loss_clip": 1.01784456, + "balance_loss_mlp": 1.02349198, + "epoch": 0.5715144592075994, + "flos": 21142443697920.0, + "grad_norm": 2.407867609369817, + "language_loss": 0.85742742, + "learning_rate": 1.6359246857018275e-06, + "loss": 0.87929076, + "num_input_tokens_seen": 102639305, + "step": 4753, + "time_per_iteration": 2.6315369606018066 + }, + { + "auxiliary_loss_clip": 0.01079475, + "auxiliary_loss_mlp": 0.01025244, + "balance_loss_clip": 0.89629257, + "balance_loss_mlp": 1.01775742, + "epoch": 0.5716347020982384, + "flos": 23330265189120.0, + "grad_norm": 2.307335494227417, + "language_loss": 0.78237337, + "learning_rate": 1.6351587570901178e-06, + "loss": 0.80342054, + "num_input_tokens_seen": 102659430, + "step": 4754, + "time_per_iteration": 2.771331787109375 + }, + { + "auxiliary_loss_clip": 0.0110735, + "auxiliary_loss_mlp": 0.01028628, + "balance_loss_clip": 0.94072032, + "balance_loss_mlp": 1.02095699, + "epoch": 0.5717549449888776, + "flos": 17009555806080.0, + "grad_norm": 2.302261576337891, + "language_loss": 0.75808918, + "learning_rate": 1.634392883824534e-06, + "loss": 0.77944899, + "num_input_tokens_seen": 102671430, + "step": 4755, + "time_per_iteration": 3.589486598968506 + }, + { + "auxiliary_loss_clip": 0.01086737, + "auxiliary_loss_mlp": 0.01032189, + "balance_loss_clip": 0.89979506, + "balance_loss_mlp": 1.02416623, + "epoch": 0.5718751878795166, + "flos": 35518130922240.0, + "grad_norm": 1.834051243900845, + "language_loss": 0.67946148, + "learning_rate": 1.6336270660212595e-06, + "loss": 0.70065069, + "num_input_tokens_seen": 102693025, + "step": 4756, + "time_per_iteration": 2.8273754119873047 + }, + { + "auxiliary_loss_clip": 0.0113307, + "auxiliary_loss_mlp": 0.01025803, + "balance_loss_clip": 0.98443389, + "balance_loss_mlp": 1.01794696, + "epoch": 0.5719954307701557, + "flos": 38613989255040.0, + "grad_norm": 3.212774963614983, + "language_loss": 0.66054922, + "learning_rate": 1.6328613037964676e-06, + "loss": 0.68213797, + "num_input_tokens_seen": 102716090, + "step": 4757, + "time_per_iteration": 2.8096225261688232 + }, + { + "auxiliary_loss_clip": 0.01151335, + "auxiliary_loss_mlp": 0.01027637, + "balance_loss_clip": 1.0130229, + "balance_loss_mlp": 1.01954246, + "epoch": 0.5721156736607949, + "flos": 20631111638400.0, + "grad_norm": 1.8102842887297348, + "language_loss": 0.67962438, + "learning_rate": 1.6320955972663241e-06, + "loss": 0.70141411, + "num_input_tokens_seen": 102735685, + "step": 4758, + "time_per_iteration": 2.6377389430999756 + }, + { + "auxiliary_loss_clip": 0.01152589, + "auxiliary_loss_mlp": 0.01029515, + "balance_loss_clip": 1.01308513, + "balance_loss_mlp": 1.02175426, + "epoch": 0.5722359165514339, + "flos": 37415076076800.0, + "grad_norm": 4.232660756444137, + "language_loss": 0.65320563, + "learning_rate": 1.6313299465469857e-06, + "loss": 0.6750266, + "num_input_tokens_seen": 102758415, + "step": 4759, + "time_per_iteration": 2.7952282428741455 + }, + { + "auxiliary_loss_clip": 0.01147461, + "auxiliary_loss_mlp": 0.01032742, + "balance_loss_clip": 1.01324701, + "balance_loss_mlp": 1.0249393, + "epoch": 0.572356159442073, + "flos": 21972877205760.0, + "grad_norm": 2.6565131952536274, + "language_loss": 0.79761523, + "learning_rate": 1.6305643517546014e-06, + "loss": 0.81941724, + "num_input_tokens_seen": 102773795, + "step": 4760, + "time_per_iteration": 2.6195104122161865 + }, + { + "auxiliary_loss_clip": 0.01177585, + "auxiliary_loss_mlp": 0.01026741, + "balance_loss_clip": 1.05567193, + "balance_loss_mlp": 1.01946306, + "epoch": 0.5724764023327121, + "flos": 19135540033920.0, + "grad_norm": 1.8944715147007978, + "language_loss": 0.84798628, + "learning_rate": 1.629798813005311e-06, + "loss": 0.87002957, + "num_input_tokens_seen": 102793515, + "step": 4761, + "time_per_iteration": 2.5836541652679443 + }, + { + "auxiliary_loss_clip": 0.01084612, + "auxiliary_loss_mlp": 0.01028536, + "balance_loss_clip": 0.90238047, + "balance_loss_mlp": 1.02072811, + "epoch": 0.5725966452233512, + "flos": 22819759142400.0, + "grad_norm": 1.907056663719388, + "language_loss": 0.70970154, + "learning_rate": 1.6290333304152473e-06, + "loss": 0.73083305, + "num_input_tokens_seen": 102813390, + "step": 4762, + "time_per_iteration": 2.753859281539917 + }, + { + "auxiliary_loss_clip": 0.0112993, + "auxiliary_loss_mlp": 0.01031155, + "balance_loss_clip": 0.98211467, + "balance_loss_mlp": 1.0236392, + "epoch": 0.5727168881139902, + "flos": 41496610498560.0, + "grad_norm": 1.7791311203850735, + "language_loss": 0.57110429, + "learning_rate": 1.6282679041005314e-06, + "loss": 0.59271514, + "num_input_tokens_seen": 102838980, + "step": 4763, + "time_per_iteration": 2.8322079181671143 + }, + { + "auxiliary_loss_clip": 0.01121291, + "auxiliary_loss_mlp": 0.01026768, + "balance_loss_clip": 0.97426075, + "balance_loss_mlp": 1.01938629, + "epoch": 0.5728371310046293, + "flos": 14647675985280.0, + "grad_norm": 4.154875648614295, + "language_loss": 0.87245357, + "learning_rate": 1.6275025341772789e-06, + "loss": 0.89393413, + "num_input_tokens_seen": 102855285, + "step": 4764, + "time_per_iteration": 2.6161303520202637 + }, + { + "auxiliary_loss_clip": 0.01127533, + "auxiliary_loss_mlp": 0.01024266, + "balance_loss_clip": 0.97408932, + "balance_loss_mlp": 1.01673198, + "epoch": 0.5729573738952685, + "flos": 21506613736320.0, + "grad_norm": 3.098374771244714, + "language_loss": 0.8182717, + "learning_rate": 1.626737220761596e-06, + "loss": 0.83978969, + "num_input_tokens_seen": 102872750, + "step": 4765, + "time_per_iteration": 2.712226152420044 + }, + { + "auxiliary_loss_clip": 0.01152774, + "auxiliary_loss_mlp": 0.0103216, + "balance_loss_clip": 1.01741588, + "balance_loss_mlp": 1.02445912, + "epoch": 0.5730776167859075, + "flos": 23621680229760.0, + "grad_norm": 4.263933431584146, + "language_loss": 0.78848803, + "learning_rate": 1.62597196396958e-06, + "loss": 0.81033742, + "num_input_tokens_seen": 102890920, + "step": 4766, + "time_per_iteration": 2.6900811195373535 + }, + { + "auxiliary_loss_clip": 0.01149602, + "auxiliary_loss_mlp": 0.01024698, + "balance_loss_clip": 1.01525593, + "balance_loss_mlp": 1.01642525, + "epoch": 0.5731978596765466, + "flos": 25739224761600.0, + "grad_norm": 1.9599803877157953, + "language_loss": 0.85425013, + "learning_rate": 1.6252067639173197e-06, + "loss": 0.87599313, + "num_input_tokens_seen": 102912830, + "step": 4767, + "time_per_iteration": 2.7207183837890625 + }, + { + "auxiliary_loss_clip": 0.01154757, + "auxiliary_loss_mlp": 0.01026674, + "balance_loss_clip": 1.01590788, + "balance_loss_mlp": 1.01917243, + "epoch": 0.5733181025671857, + "flos": 26359509749760.0, + "grad_norm": 1.8536636441211205, + "language_loss": 0.70021772, + "learning_rate": 1.6244416207208956e-06, + "loss": 0.72203207, + "num_input_tokens_seen": 102933765, + "step": 4768, + "time_per_iteration": 2.6936588287353516 + }, + { + "auxiliary_loss_clip": 0.01108761, + "auxiliary_loss_mlp": 0.01031348, + "balance_loss_clip": 0.9396283, + "balance_loss_mlp": 1.02343297, + "epoch": 0.5734383454578248, + "flos": 29423874833280.0, + "grad_norm": 1.8015837329578235, + "language_loss": 0.73552591, + "learning_rate": 1.6236765344963787e-06, + "loss": 0.75692701, + "num_input_tokens_seen": 102955025, + "step": 4769, + "time_per_iteration": 2.7616679668426514 + }, + { + "auxiliary_loss_clip": 0.01127282, + "auxiliary_loss_mlp": 0.01027908, + "balance_loss_clip": 0.97641748, + "balance_loss_mlp": 1.02057612, + "epoch": 0.5735585883484638, + "flos": 34969954487040.0, + "grad_norm": 2.5624027752363503, + "language_loss": 0.69511998, + "learning_rate": 1.6229115053598322e-06, + "loss": 0.71667182, + "num_input_tokens_seen": 102976780, + "step": 4770, + "time_per_iteration": 3.7406270503997803 + }, + { + "auxiliary_loss_clip": 0.01156246, + "auxiliary_loss_mlp": 0.01027927, + "balance_loss_clip": 1.01903558, + "balance_loss_mlp": 1.0204761, + "epoch": 0.573678831239103, + "flos": 18770759464320.0, + "grad_norm": 4.483044266905409, + "language_loss": 0.72212005, + "learning_rate": 1.6221465334273108e-06, + "loss": 0.74396181, + "num_input_tokens_seen": 102995990, + "step": 4771, + "time_per_iteration": 3.6015851497650146 + }, + { + "auxiliary_loss_clip": 0.01110628, + "auxiliary_loss_mlp": 0.01024092, + "balance_loss_clip": 0.93740153, + "balance_loss_mlp": 1.01628947, + "epoch": 0.5737990741297421, + "flos": 25702883176320.0, + "grad_norm": 2.8372186958890766, + "language_loss": 0.61320138, + "learning_rate": 1.6213816188148593e-06, + "loss": 0.6345486, + "num_input_tokens_seen": 103014695, + "step": 4772, + "time_per_iteration": 3.6795904636383057 + }, + { + "auxiliary_loss_clip": 0.01124989, + "auxiliary_loss_mlp": 0.01025214, + "balance_loss_clip": 0.98167896, + "balance_loss_mlp": 1.01787114, + "epoch": 0.5739193170203811, + "flos": 27269234530560.0, + "grad_norm": 1.5710808113970447, + "language_loss": 0.77023149, + "learning_rate": 1.6206167616385162e-06, + "loss": 0.79173344, + "num_input_tokens_seen": 103035760, + "step": 4773, + "time_per_iteration": 2.6963772773742676 + }, + { + "auxiliary_loss_clip": 0.01136104, + "auxiliary_loss_mlp": 0.01027002, + "balance_loss_clip": 0.98041564, + "balance_loss_mlp": 1.01933742, + "epoch": 0.5740395599110203, + "flos": 12239721993600.0, + "grad_norm": 2.000639014610222, + "language_loss": 0.73599017, + "learning_rate": 1.6198519620143078e-06, + "loss": 0.75762117, + "num_input_tokens_seen": 103052915, + "step": 4774, + "time_per_iteration": 2.6569955348968506 + }, + { + "auxiliary_loss_clip": 0.01112678, + "auxiliary_loss_mlp": 0.01029501, + "balance_loss_clip": 0.94106728, + "balance_loss_mlp": 1.02166855, + "epoch": 0.5741598028016593, + "flos": 25921399564800.0, + "grad_norm": 1.5735890082496002, + "language_loss": 0.78135031, + "learning_rate": 1.6190872200582546e-06, + "loss": 0.80277216, + "num_input_tokens_seen": 103074655, + "step": 4775, + "time_per_iteration": 2.735165596008301 + }, + { + "auxiliary_loss_clip": 0.01126069, + "auxiliary_loss_mlp": 0.00509507, + "balance_loss_clip": 0.97627056, + "balance_loss_mlp": 0.0, + "epoch": 0.5742800456922984, + "flos": 19244133826560.0, + "grad_norm": 2.205976347810466, + "language_loss": 0.7795344, + "learning_rate": 1.6183225358863676e-06, + "loss": 0.79589021, + "num_input_tokens_seen": 103091550, + "step": 4776, + "time_per_iteration": 2.6930811405181885 + }, + { + "auxiliary_loss_clip": 0.01123624, + "auxiliary_loss_mlp": 0.01029375, + "balance_loss_clip": 0.97548985, + "balance_loss_mlp": 1.0217042, + "epoch": 0.5744002885829376, + "flos": 30920487932160.0, + "grad_norm": 2.213593135398581, + "language_loss": 0.71876371, + "learning_rate": 1.617557909614648e-06, + "loss": 0.74029374, + "num_input_tokens_seen": 103110985, + "step": 4777, + "time_per_iteration": 2.7999234199523926 + }, + { + "auxiliary_loss_clip": 0.01102688, + "auxiliary_loss_mlp": 0.01025949, + "balance_loss_clip": 0.93588698, + "balance_loss_mlp": 1.01880288, + "epoch": 0.5745205314735766, + "flos": 23840017050240.0, + "grad_norm": 1.7872897801361025, + "language_loss": 0.85858363, + "learning_rate": 1.6167933413590899e-06, + "loss": 0.87987, + "num_input_tokens_seen": 103129890, + "step": 4778, + "time_per_iteration": 2.7324814796447754 + }, + { + "auxiliary_loss_clip": 0.01153526, + "auxiliary_loss_mlp": 0.01031694, + "balance_loss_clip": 1.01611662, + "balance_loss_mlp": 1.02383852, + "epoch": 0.5746407743642157, + "flos": 12311902373760.0, + "grad_norm": 2.609712747389287, + "language_loss": 0.90141416, + "learning_rate": 1.6160288312356773e-06, + "loss": 0.92326641, + "num_input_tokens_seen": 103147020, + "step": 4779, + "time_per_iteration": 2.5471012592315674 + }, + { + "auxiliary_loss_clip": 0.01155713, + "auxiliary_loss_mlp": 0.01025222, + "balance_loss_clip": 1.01441443, + "balance_loss_mlp": 1.01753902, + "epoch": 0.5747610172548548, + "flos": 24133658734080.0, + "grad_norm": 1.9823347257992228, + "language_loss": 0.8152405, + "learning_rate": 1.6152643793603857e-06, + "loss": 0.8370499, + "num_input_tokens_seen": 103167370, + "step": 4780, + "time_per_iteration": 2.6014554500579834 + }, + { + "auxiliary_loss_clip": 0.01173007, + "auxiliary_loss_mlp": 0.01025384, + "balance_loss_clip": 1.05191517, + "balance_loss_mlp": 1.01756954, + "epoch": 0.5748812601454939, + "flos": 25408451393280.0, + "grad_norm": 1.6202939547714532, + "language_loss": 0.87746167, + "learning_rate": 1.6144999858491815e-06, + "loss": 0.89944565, + "num_input_tokens_seen": 103186000, + "step": 4781, + "time_per_iteration": 2.527864933013916 + }, + { + "auxiliary_loss_clip": 0.01132415, + "auxiliary_loss_mlp": 0.01026029, + "balance_loss_clip": 0.97680414, + "balance_loss_mlp": 1.01837564, + "epoch": 0.575001503036133, + "flos": 30624942827520.0, + "grad_norm": 1.5722612310417063, + "language_loss": 0.85745013, + "learning_rate": 1.6137356508180232e-06, + "loss": 0.87903464, + "num_input_tokens_seen": 103207710, + "step": 4782, + "time_per_iteration": 3.5830178260803223 + }, + { + "auxiliary_loss_clip": 0.01176919, + "auxiliary_loss_mlp": 0.00509672, + "balance_loss_clip": 1.05365992, + "balance_loss_mlp": 0.0, + "epoch": 0.5751217459267721, + "flos": 21726566668800.0, + "grad_norm": 1.7581531773950716, + "language_loss": 0.81619787, + "learning_rate": 1.6129713743828593e-06, + "loss": 0.83306378, + "num_input_tokens_seen": 103226720, + "step": 4783, + "time_per_iteration": 2.5687415599823 + }, + { + "auxiliary_loss_clip": 0.01128266, + "auxiliary_loss_mlp": 0.01027053, + "balance_loss_clip": 0.97331017, + "balance_loss_mlp": 1.0196557, + "epoch": 0.5752419888174112, + "flos": 21651620941440.0, + "grad_norm": 1.4644532036251257, + "language_loss": 0.75480086, + "learning_rate": 1.6122071566596306e-06, + "loss": 0.77635401, + "num_input_tokens_seen": 103246995, + "step": 4784, + "time_per_iteration": 2.6700797080993652 + }, + { + "auxiliary_loss_clip": 0.0115387, + "auxiliary_loss_mlp": 0.01024354, + "balance_loss_clip": 1.01640165, + "balance_loss_mlp": 1.01685262, + "epoch": 0.5753622317080502, + "flos": 17775997234560.0, + "grad_norm": 2.2720903240137953, + "language_loss": 0.83596265, + "learning_rate": 1.6114429977642674e-06, + "loss": 0.85774493, + "num_input_tokens_seen": 103261500, + "step": 4785, + "time_per_iteration": 2.568709373474121 + }, + { + "auxiliary_loss_clip": 0.01155211, + "auxiliary_loss_mlp": 0.01028591, + "balance_loss_clip": 1.01929903, + "balance_loss_mlp": 1.02075267, + "epoch": 0.5754824745986894, + "flos": 19789616741760.0, + "grad_norm": 1.7486786037946649, + "language_loss": 0.73902971, + "learning_rate": 1.6106788978126926e-06, + "loss": 0.76086771, + "num_input_tokens_seen": 103280475, + "step": 4786, + "time_per_iteration": 2.6347641944885254 + }, + { + "auxiliary_loss_clip": 0.01078979, + "auxiliary_loss_mlp": 0.01023283, + "balance_loss_clip": 0.89746273, + "balance_loss_mlp": 1.01552248, + "epoch": 0.5756027174893285, + "flos": 30985665160320.0, + "grad_norm": 7.221311412693125, + "language_loss": 0.78330791, + "learning_rate": 1.6099148569208196e-06, + "loss": 0.80433053, + "num_input_tokens_seen": 103297695, + "step": 4787, + "time_per_iteration": 2.797200918197632 + }, + { + "auxiliary_loss_clip": 0.01128821, + "auxiliary_loss_mlp": 0.01024718, + "balance_loss_clip": 0.98002172, + "balance_loss_mlp": 1.01755011, + "epoch": 0.5757229603799675, + "flos": 28546864364160.0, + "grad_norm": 1.620359737637977, + "language_loss": 0.62787461, + "learning_rate": 1.6091508752045523e-06, + "loss": 0.64941001, + "num_input_tokens_seen": 103318575, + "step": 4788, + "time_per_iteration": 2.7217445373535156 + }, + { + "auxiliary_loss_clip": 0.01097339, + "auxiliary_loss_mlp": 0.01027289, + "balance_loss_clip": 0.93504715, + "balance_loss_mlp": 1.01933146, + "epoch": 0.5758432032706067, + "flos": 22999024944000.0, + "grad_norm": 1.682006772310682, + "language_loss": 0.86430919, + "learning_rate": 1.608386952779787e-06, + "loss": 0.88555545, + "num_input_tokens_seen": 103337945, + "step": 4789, + "time_per_iteration": 2.764385461807251 + }, + { + "auxiliary_loss_clip": 0.01135479, + "auxiliary_loss_mlp": 0.01022006, + "balance_loss_clip": 0.97794116, + "balance_loss_mlp": 1.01496696, + "epoch": 0.5759634461612457, + "flos": 25739727552000.0, + "grad_norm": 1.8179006974661942, + "language_loss": 0.74498367, + "learning_rate": 1.6076230897624098e-06, + "loss": 0.76655853, + "num_input_tokens_seen": 103360150, + "step": 4790, + "time_per_iteration": 2.6901869773864746 + }, + { + "auxiliary_loss_clip": 0.01151637, + "auxiliary_loss_mlp": 0.01027434, + "balance_loss_clip": 1.01194859, + "balance_loss_mlp": 1.02000141, + "epoch": 0.5760836890518848, + "flos": 30591761639040.0, + "grad_norm": 1.9665131231277002, + "language_loss": 0.77598846, + "learning_rate": 1.6068592862682974e-06, + "loss": 0.79777908, + "num_input_tokens_seen": 103378305, + "step": 4791, + "time_per_iteration": 2.7450244426727295 + }, + { + "auxiliary_loss_clip": 0.01132159, + "auxiliary_loss_mlp": 0.01027083, + "balance_loss_clip": 0.97822285, + "balance_loss_mlp": 1.01978731, + "epoch": 0.576203931942524, + "flos": 36538963447680.0, + "grad_norm": 1.8361251753849217, + "language_loss": 0.73667693, + "learning_rate": 1.6060955424133187e-06, + "loss": 0.75826931, + "num_input_tokens_seen": 103399230, + "step": 4792, + "time_per_iteration": 2.7497665882110596 + }, + { + "auxiliary_loss_clip": 0.01151396, + "auxiliary_loss_mlp": 0.01026383, + "balance_loss_clip": 1.01710355, + "balance_loss_mlp": 1.01879573, + "epoch": 0.576324174833163, + "flos": 25516937445120.0, + "grad_norm": 1.5825753626891832, + "language_loss": 0.89842808, + "learning_rate": 1.6053318583133332e-06, + "loss": 0.92020583, + "num_input_tokens_seen": 103420100, + "step": 4793, + "time_per_iteration": 2.6575894355773926 + }, + { + "auxiliary_loss_clip": 0.01151245, + "auxiliary_loss_mlp": 0.01028939, + "balance_loss_clip": 1.01523757, + "balance_loss_mlp": 1.02081513, + "epoch": 0.5764444177238021, + "flos": 25119262995840.0, + "grad_norm": 2.333389217347488, + "language_loss": 0.74697256, + "learning_rate": 1.6045682340841907e-06, + "loss": 0.76877439, + "num_input_tokens_seen": 103439025, + "step": 4794, + "time_per_iteration": 2.625953197479248 + }, + { + "auxiliary_loss_clip": 0.0101554, + "auxiliary_loss_mlp": 0.00500636, + "balance_loss_clip": 0.90715796, + "balance_loss_mlp": 0.0, + "epoch": 0.5765646606144411, + "flos": 62212687758720.0, + "grad_norm": 0.7509954832298222, + "language_loss": 0.58026242, + "learning_rate": 1.6038046698417336e-06, + "loss": 0.59542418, + "num_input_tokens_seen": 103499920, + "step": 4795, + "time_per_iteration": 3.274681329727173 + }, + { + "auxiliary_loss_clip": 0.01150696, + "auxiliary_loss_mlp": 0.01033041, + "balance_loss_clip": 1.01288497, + "balance_loss_mlp": 1.02519083, + "epoch": 0.5766849035050803, + "flos": 25118760205440.0, + "grad_norm": 1.7383039432807816, + "language_loss": 0.69235134, + "learning_rate": 1.6030411657017919e-06, + "loss": 0.71418875, + "num_input_tokens_seen": 103519575, + "step": 4796, + "time_per_iteration": 3.6647443771362305 + }, + { + "auxiliary_loss_clip": 0.01144918, + "auxiliary_loss_mlp": 0.01030185, + "balance_loss_clip": 1.01282632, + "balance_loss_mlp": 1.02325606, + "epoch": 0.5768051463957193, + "flos": 15991093578240.0, + "grad_norm": 2.7948745867410394, + "language_loss": 0.84483677, + "learning_rate": 1.6022777217801903e-06, + "loss": 0.86658782, + "num_input_tokens_seen": 103536530, + "step": 4797, + "time_per_iteration": 2.5922367572784424 + }, + { + "auxiliary_loss_clip": 0.01107926, + "auxiliary_loss_mlp": 0.01022237, + "balance_loss_clip": 0.94201553, + "balance_loss_mlp": 1.01439285, + "epoch": 0.5769253892863584, + "flos": 22163635359360.0, + "grad_norm": 2.057127871033824, + "language_loss": 0.73621881, + "learning_rate": 1.601514338192742e-06, + "loss": 0.75752044, + "num_input_tokens_seen": 103556460, + "step": 4798, + "time_per_iteration": 4.549978017807007 + }, + { + "auxiliary_loss_clip": 0.0116897, + "auxiliary_loss_mlp": 0.01027101, + "balance_loss_clip": 1.05057538, + "balance_loss_mlp": 1.01960862, + "epoch": 0.5770456321769976, + "flos": 22856388036480.0, + "grad_norm": 2.0481914472083913, + "language_loss": 0.71584344, + "learning_rate": 1.6007510150552514e-06, + "loss": 0.73780417, + "num_input_tokens_seen": 103574520, + "step": 4799, + "time_per_iteration": 2.6632802486419678 + }, + { + "auxiliary_loss_clip": 0.01157386, + "auxiliary_loss_mlp": 0.0102636, + "balance_loss_clip": 1.01457155, + "balance_loss_mlp": 1.01839066, + "epoch": 0.5771658750676366, + "flos": 46353672489600.0, + "grad_norm": 2.0575302166010774, + "language_loss": 0.62124628, + "learning_rate": 1.599987752483515e-06, + "loss": 0.64308375, + "num_input_tokens_seen": 103598965, + "step": 4800, + "time_per_iteration": 2.818497657775879 + }, + { + "auxiliary_loss_clip": 0.01101061, + "auxiliary_loss_mlp": 0.01030372, + "balance_loss_clip": 0.93566763, + "balance_loss_mlp": 1.02278447, + "epoch": 0.5772861179582757, + "flos": 22159972172160.0, + "grad_norm": 2.2043975579109025, + "language_loss": 0.68144113, + "learning_rate": 1.5992245505933184e-06, + "loss": 0.70275545, + "num_input_tokens_seen": 103618665, + "step": 4801, + "time_per_iteration": 2.728020429611206 + }, + { + "auxiliary_loss_clip": 0.01177579, + "auxiliary_loss_mlp": 0.0102536, + "balance_loss_clip": 1.05479109, + "balance_loss_mlp": 1.01787066, + "epoch": 0.5774063608489148, + "flos": 31248926916480.0, + "grad_norm": 2.366908990731338, + "language_loss": 0.71159565, + "learning_rate": 1.5984614095004388e-06, + "loss": 0.73362505, + "num_input_tokens_seen": 103639800, + "step": 4802, + "time_per_iteration": 2.6891016960144043 + }, + { + "auxiliary_loss_clip": 0.01149182, + "auxiliary_loss_mlp": 0.01032501, + "balance_loss_clip": 1.0148176, + "balance_loss_mlp": 1.02478766, + "epoch": 0.5775266037395539, + "flos": 22527123039360.0, + "grad_norm": 2.7725093696064307, + "language_loss": 0.81084645, + "learning_rate": 1.5976983293206438e-06, + "loss": 0.83266324, + "num_input_tokens_seen": 103655605, + "step": 4803, + "time_per_iteration": 2.7013914585113525 + }, + { + "auxiliary_loss_clip": 0.01125718, + "auxiliary_loss_mlp": 0.01028335, + "balance_loss_clip": 0.9740783, + "balance_loss_mlp": 1.02059829, + "epoch": 0.577646846630193, + "flos": 21068790860160.0, + "grad_norm": 1.9883692657364318, + "language_loss": 0.71615803, + "learning_rate": 1.5969353101696928e-06, + "loss": 0.73769855, + "num_input_tokens_seen": 103674045, + "step": 4804, + "time_per_iteration": 2.6568636894226074 + }, + { + "auxiliary_loss_clip": 0.01153907, + "auxiliary_loss_mlp": 0.01022906, + "balance_loss_clip": 1.01592755, + "balance_loss_mlp": 1.01542592, + "epoch": 0.5777670895208321, + "flos": 29714284293120.0, + "grad_norm": 1.6388626494948808, + "language_loss": 0.79675442, + "learning_rate": 1.5961723521633341e-06, + "loss": 0.81852257, + "num_input_tokens_seen": 103695285, + "step": 4805, + "time_per_iteration": 2.7025914192199707 + }, + { + "auxiliary_loss_clip": 0.01126667, + "auxiliary_loss_mlp": 0.01024518, + "balance_loss_clip": 0.97503006, + "balance_loss_mlp": 1.01686227, + "epoch": 0.5778873324114712, + "flos": 19500428344320.0, + "grad_norm": 2.0068991065349886, + "language_loss": 0.9077251, + "learning_rate": 1.5954094554173097e-06, + "loss": 0.92923701, + "num_input_tokens_seen": 103713275, + "step": 4806, + "time_per_iteration": 2.709512948989868 + }, + { + "auxiliary_loss_clip": 0.01134813, + "auxiliary_loss_mlp": 0.01026563, + "balance_loss_clip": 0.97832251, + "balance_loss_mlp": 1.01924992, + "epoch": 0.5780075753021102, + "flos": 14136846716160.0, + "grad_norm": 2.2549538738386334, + "language_loss": 0.79246986, + "learning_rate": 1.5946466200473482e-06, + "loss": 0.81408358, + "num_input_tokens_seen": 103731185, + "step": 4807, + "time_per_iteration": 2.6373727321624756 + }, + { + "auxiliary_loss_clip": 0.01132834, + "auxiliary_loss_mlp": 0.01026176, + "balance_loss_clip": 0.97641826, + "balance_loss_mlp": 1.01898527, + "epoch": 0.5781278181927494, + "flos": 15262178883840.0, + "grad_norm": 2.0682220906573288, + "language_loss": 0.83709717, + "learning_rate": 1.5938838461691723e-06, + "loss": 0.85868734, + "num_input_tokens_seen": 103748095, + "step": 4808, + "time_per_iteration": 3.6094865798950195 + }, + { + "auxiliary_loss_clip": 0.01178488, + "auxiliary_loss_mlp": 0.01027633, + "balance_loss_clip": 1.05564153, + "balance_loss_mlp": 1.01914549, + "epoch": 0.5782480610833884, + "flos": 16726831856640.0, + "grad_norm": 2.6005864986872878, + "language_loss": 0.82979834, + "learning_rate": 1.593121133898494e-06, + "loss": 0.85185951, + "num_input_tokens_seen": 103765300, + "step": 4809, + "time_per_iteration": 2.5697662830352783 + }, + { + "auxiliary_loss_clip": 0.01158616, + "auxiliary_loss_mlp": 0.01026042, + "balance_loss_clip": 1.01559734, + "balance_loss_mlp": 1.018502, + "epoch": 0.5783683039740275, + "flos": 25482140144640.0, + "grad_norm": 2.0285356924244287, + "language_loss": 0.79014289, + "learning_rate": 1.592358483351016e-06, + "loss": 0.81198955, + "num_input_tokens_seen": 103785475, + "step": 4810, + "time_per_iteration": 2.6963284015655518 + }, + { + "auxiliary_loss_clip": 0.01151016, + "auxiliary_loss_mlp": 0.0102827, + "balance_loss_clip": 1.01459002, + "balance_loss_mlp": 1.02114391, + "epoch": 0.5784885468646667, + "flos": 18405835240320.0, + "grad_norm": 1.8169001665866182, + "language_loss": 0.72537374, + "learning_rate": 1.5915958946424326e-06, + "loss": 0.74716663, + "num_input_tokens_seen": 103804160, + "step": 4811, + "time_per_iteration": 2.6458210945129395 + }, + { + "auxiliary_loss_clip": 0.01109362, + "auxiliary_loss_mlp": 0.00510368, + "balance_loss_clip": 0.94012171, + "balance_loss_mlp": 0.0, + "epoch": 0.5786087897553057, + "flos": 46100717936640.0, + "grad_norm": 1.5617916394507854, + "language_loss": 0.74509275, + "learning_rate": 1.5908333678884271e-06, + "loss": 0.76129007, + "num_input_tokens_seen": 103830580, + "step": 4812, + "time_per_iteration": 2.916720390319824 + }, + { + "auxiliary_loss_clip": 0.01155809, + "auxiliary_loss_mlp": 0.01027469, + "balance_loss_clip": 1.01817536, + "balance_loss_mlp": 1.01994109, + "epoch": 0.5787290326459448, + "flos": 12385950261120.0, + "grad_norm": 2.079852124890113, + "language_loss": 0.73730171, + "learning_rate": 1.5900709032046743e-06, + "loss": 0.75913453, + "num_input_tokens_seen": 103848655, + "step": 4813, + "time_per_iteration": 2.6490471363067627 + }, + { + "auxiliary_loss_clip": 0.01129738, + "auxiliary_loss_mlp": 0.01025908, + "balance_loss_clip": 0.98022848, + "balance_loss_mlp": 1.01849365, + "epoch": 0.5788492755365839, + "flos": 23290332243840.0, + "grad_norm": 1.8091520638530096, + "language_loss": 0.78268504, + "learning_rate": 1.5893085007068391e-06, + "loss": 0.80424148, + "num_input_tokens_seen": 103866215, + "step": 4814, + "time_per_iteration": 2.6820359230041504 + }, + { + "auxiliary_loss_clip": 0.01117463, + "auxiliary_loss_mlp": 0.01031068, + "balance_loss_clip": 0.9736709, + "balance_loss_mlp": 1.02377844, + "epoch": 0.578969518427223, + "flos": 24061047390720.0, + "grad_norm": 1.8076164315744774, + "language_loss": 0.7090556, + "learning_rate": 1.5885461605105786e-06, + "loss": 0.73054087, + "num_input_tokens_seen": 103887815, + "step": 4815, + "time_per_iteration": 2.7087466716766357 + }, + { + "auxiliary_loss_clip": 0.01133004, + "auxiliary_loss_mlp": 0.01029071, + "balance_loss_clip": 0.98024279, + "balance_loss_mlp": 1.02088141, + "epoch": 0.579089761317862, + "flos": 21871825269120.0, + "grad_norm": 1.7083134854992554, + "language_loss": 0.76628685, + "learning_rate": 1.5877838827315375e-06, + "loss": 0.78790754, + "num_input_tokens_seen": 103906360, + "step": 4816, + "time_per_iteration": 2.695035696029663 + }, + { + "auxiliary_loss_clip": 0.01177304, + "auxiliary_loss_mlp": 0.01025376, + "balance_loss_clip": 1.05601394, + "balance_loss_mlp": 1.01777673, + "epoch": 0.5792100042085012, + "flos": 22929681738240.0, + "grad_norm": 1.8372198068686338, + "language_loss": 0.70062935, + "learning_rate": 1.587021667485355e-06, + "loss": 0.72265619, + "num_input_tokens_seen": 103925730, + "step": 4817, + "time_per_iteration": 2.6375865936279297 + }, + { + "auxiliary_loss_clip": 0.01133766, + "auxiliary_loss_mlp": 0.01025406, + "balance_loss_clip": 0.97725427, + "balance_loss_mlp": 1.01840878, + "epoch": 0.5793302470991403, + "flos": 21470056669440.0, + "grad_norm": 1.6356421568000656, + "language_loss": 0.7847898, + "learning_rate": 1.5862595148876559e-06, + "loss": 0.80638152, + "num_input_tokens_seen": 103945835, + "step": 4818, + "time_per_iteration": 2.6808013916015625 + }, + { + "auxiliary_loss_clip": 0.01086864, + "auxiliary_loss_mlp": 0.01029878, + "balance_loss_clip": 0.90406156, + "balance_loss_mlp": 1.02200985, + "epoch": 0.5794504899897793, + "flos": 12711013367040.0, + "grad_norm": 11.678880731543792, + "language_loss": 0.76211107, + "learning_rate": 1.58549742505406e-06, + "loss": 0.78327847, + "num_input_tokens_seen": 103960580, + "step": 4819, + "time_per_iteration": 2.6899170875549316 + }, + { + "auxiliary_loss_clip": 0.01179193, + "auxiliary_loss_mlp": 0.01026003, + "balance_loss_clip": 1.05632448, + "balance_loss_mlp": 1.01808178, + "epoch": 0.5795707328804185, + "flos": 14867054300160.0, + "grad_norm": 2.2686486576921028, + "language_loss": 0.75302601, + "learning_rate": 1.5847353981001747e-06, + "loss": 0.775078, + "num_input_tokens_seen": 103977760, + "step": 4820, + "time_per_iteration": 2.5722579956054688 + }, + { + "auxiliary_loss_clip": 0.01128226, + "auxiliary_loss_mlp": 0.01028192, + "balance_loss_clip": 0.97702575, + "balance_loss_mlp": 1.02012146, + "epoch": 0.5796909757710575, + "flos": 36430046432640.0, + "grad_norm": 1.7168555252745064, + "language_loss": 0.69937944, + "learning_rate": 1.5839734341415993e-06, + "loss": 0.72094357, + "num_input_tokens_seen": 103999960, + "step": 4821, + "time_per_iteration": 2.848949909210205 + }, + { + "auxiliary_loss_clip": 0.01153004, + "auxiliary_loss_mlp": 0.01029292, + "balance_loss_clip": 1.02154374, + "balance_loss_mlp": 1.02202022, + "epoch": 0.5798112186616966, + "flos": 23039891642880.0, + "grad_norm": 1.6729843456146918, + "language_loss": 0.76496482, + "learning_rate": 1.5832115332939238e-06, + "loss": 0.78678775, + "num_input_tokens_seen": 104018400, + "step": 4822, + "time_per_iteration": 3.60870361328125 + }, + { + "auxiliary_loss_clip": 0.01152778, + "auxiliary_loss_mlp": 0.01020395, + "balance_loss_clip": 1.01525259, + "balance_loss_mlp": 1.01312304, + "epoch": 0.5799314615523358, + "flos": 16652604401280.0, + "grad_norm": 1.8546020427844878, + "language_loss": 0.74744987, + "learning_rate": 1.5824496956727272e-06, + "loss": 0.76918161, + "num_input_tokens_seen": 104035605, + "step": 4823, + "time_per_iteration": 3.5691144466400146 + }, + { + "auxiliary_loss_clip": 0.01129208, + "auxiliary_loss_mlp": 0.01027307, + "balance_loss_clip": 0.97679883, + "balance_loss_mlp": 1.02041721, + "epoch": 0.5800517044429748, + "flos": 20485673470080.0, + "grad_norm": 1.6701192501678106, + "language_loss": 0.73302388, + "learning_rate": 1.5816879213935797e-06, + "loss": 0.75458908, + "num_input_tokens_seen": 104054415, + "step": 4824, + "time_per_iteration": 3.5745716094970703 + }, + { + "auxiliary_loss_clip": 0.01149707, + "auxiliary_loss_mlp": 0.01027873, + "balance_loss_clip": 1.01680684, + "balance_loss_mlp": 1.0203805, + "epoch": 0.5801719473336139, + "flos": 31538258968320.0, + "grad_norm": 1.467152747399936, + "language_loss": 0.79877609, + "learning_rate": 1.5809262105720416e-06, + "loss": 0.82055187, + "num_input_tokens_seen": 104075455, + "step": 4825, + "time_per_iteration": 2.681486129760742 + }, + { + "auxiliary_loss_clip": 0.01176195, + "auxiliary_loss_mlp": 0.0102605, + "balance_loss_clip": 1.05570006, + "balance_loss_mlp": 1.01862347, + "epoch": 0.580292190224253, + "flos": 20375966355840.0, + "grad_norm": 1.4581106382300713, + "language_loss": 0.79878235, + "learning_rate": 1.5801645633236644e-06, + "loss": 0.82080477, + "num_input_tokens_seen": 104096440, + "step": 4826, + "time_per_iteration": 2.6898281574249268 + }, + { + "auxiliary_loss_clip": 0.01121301, + "auxiliary_loss_mlp": 0.01032839, + "balance_loss_clip": 0.97438818, + "balance_loss_mlp": 1.02514684, + "epoch": 0.5804124331148921, + "flos": 26615373304320.0, + "grad_norm": 2.241036065829604, + "language_loss": 0.77322394, + "learning_rate": 1.579402979763989e-06, + "loss": 0.79476529, + "num_input_tokens_seen": 104116775, + "step": 4827, + "time_per_iteration": 2.7594029903411865 + }, + { + "auxiliary_loss_clip": 0.01071885, + "auxiliary_loss_mlp": 0.01027767, + "balance_loss_clip": 0.86334109, + "balance_loss_mlp": 1.02066183, + "epoch": 0.5805326760055312, + "flos": 13478496289920.0, + "grad_norm": 2.220632961359711, + "language_loss": 0.81435055, + "learning_rate": 1.578641460008548e-06, + "loss": 0.83534706, + "num_input_tokens_seen": 104134510, + "step": 4828, + "time_per_iteration": 2.758308172225952 + }, + { + "auxiliary_loss_clip": 0.01154117, + "auxiliary_loss_mlp": 0.0102275, + "balance_loss_clip": 1.01712096, + "balance_loss_mlp": 1.01542497, + "epoch": 0.5806529188961702, + "flos": 12091374823680.0, + "grad_norm": 1.8929441884107794, + "language_loss": 0.67554545, + "learning_rate": 1.5778800041728613e-06, + "loss": 0.69731414, + "num_input_tokens_seen": 104150800, + "step": 4829, + "time_per_iteration": 2.6387276649475098 + }, + { + "auxiliary_loss_clip": 0.01147423, + "auxiliary_loss_mlp": 0.01025773, + "balance_loss_clip": 1.01585686, + "balance_loss_mlp": 1.01808977, + "epoch": 0.5807731617868094, + "flos": 26214107495040.0, + "grad_norm": 1.3733580031782628, + "language_loss": 0.66310167, + "learning_rate": 1.577118612372443e-06, + "loss": 0.68483365, + "num_input_tokens_seen": 104172640, + "step": 4830, + "time_per_iteration": 2.6844847202301025 + }, + { + "auxiliary_loss_clip": 0.01124754, + "auxiliary_loss_mlp": 0.00510224, + "balance_loss_clip": 0.97461551, + "balance_loss_mlp": 0.0, + "epoch": 0.5808934046774484, + "flos": 37962139190400.0, + "grad_norm": 4.130348833549468, + "language_loss": 0.70757276, + "learning_rate": 1.5763572847227943e-06, + "loss": 0.72392261, + "num_input_tokens_seen": 104193525, + "step": 4831, + "time_per_iteration": 2.843852996826172 + }, + { + "auxiliary_loss_clip": 0.01150445, + "auxiliary_loss_mlp": 0.01023473, + "balance_loss_clip": 1.01281714, + "balance_loss_mlp": 1.01603436, + "epoch": 0.5810136475680875, + "flos": 20485853038080.0, + "grad_norm": 2.8552531085545847, + "language_loss": 0.8152777, + "learning_rate": 1.5755960213394091e-06, + "loss": 0.83701682, + "num_input_tokens_seen": 104210625, + "step": 4832, + "time_per_iteration": 2.64207124710083 + }, + { + "auxiliary_loss_clip": 0.01109788, + "auxiliary_loss_mlp": 0.01026152, + "balance_loss_clip": 0.93933344, + "balance_loss_mlp": 1.01883245, + "epoch": 0.5811338904587267, + "flos": 17530153574400.0, + "grad_norm": 3.684559023629347, + "language_loss": 0.78599191, + "learning_rate": 1.5748348223377703e-06, + "loss": 0.80735129, + "num_input_tokens_seen": 104228180, + "step": 4833, + "time_per_iteration": 2.671863317489624 + }, + { + "auxiliary_loss_clip": 0.01127381, + "auxiliary_loss_mlp": 0.01024235, + "balance_loss_clip": 0.97815806, + "balance_loss_mlp": 1.01724362, + "epoch": 0.5812541333493657, + "flos": 19458017360640.0, + "grad_norm": 1.5948834020135525, + "language_loss": 0.78066599, + "learning_rate": 1.5740736878333507e-06, + "loss": 0.80218214, + "num_input_tokens_seen": 104246020, + "step": 4834, + "time_per_iteration": 3.588871955871582 + }, + { + "auxiliary_loss_clip": 0.0113339, + "auxiliary_loss_mlp": 0.01029454, + "balance_loss_clip": 0.97704566, + "balance_loss_mlp": 1.02101362, + "epoch": 0.5813743762400048, + "flos": 20594949621120.0, + "grad_norm": 2.503164356732585, + "language_loss": 0.7746774, + "learning_rate": 1.5733126179416143e-06, + "loss": 0.79630578, + "num_input_tokens_seen": 104260505, + "step": 4835, + "time_per_iteration": 2.645960569381714 + }, + { + "auxiliary_loss_clip": 0.01153379, + "auxiliary_loss_mlp": 0.010271, + "balance_loss_clip": 1.0168215, + "balance_loss_mlp": 1.02000713, + "epoch": 0.5814946191306439, + "flos": 33178227246720.0, + "grad_norm": 2.2937555357030366, + "language_loss": 0.72666401, + "learning_rate": 1.5725516127780137e-06, + "loss": 0.74846882, + "num_input_tokens_seen": 104282640, + "step": 4836, + "time_per_iteration": 2.737348794937134 + }, + { + "auxiliary_loss_clip": 0.01158206, + "auxiliary_loss_mlp": 0.01028607, + "balance_loss_clip": 1.01563692, + "balance_loss_mlp": 1.0205369, + "epoch": 0.581614862021283, + "flos": 16143283503360.0, + "grad_norm": 2.2193255589129466, + "language_loss": 0.8884393, + "learning_rate": 1.5717906724579943e-06, + "loss": 0.91030741, + "num_input_tokens_seen": 104299700, + "step": 4837, + "time_per_iteration": 2.5952513217926025 + }, + { + "auxiliary_loss_clip": 0.01113452, + "auxiliary_loss_mlp": 0.01030174, + "balance_loss_clip": 0.93778157, + "balance_loss_mlp": 1.02255046, + "epoch": 0.581735104911922, + "flos": 33802642298880.0, + "grad_norm": 1.8759084275468403, + "language_loss": 0.68390846, + "learning_rate": 1.571029797096989e-06, + "loss": 0.70534468, + "num_input_tokens_seen": 104320805, + "step": 4838, + "time_per_iteration": 2.8506031036376953 + }, + { + "auxiliary_loss_clip": 0.01173377, + "auxiliary_loss_mlp": 0.01027884, + "balance_loss_clip": 1.0532155, + "balance_loss_mlp": 1.02066565, + "epoch": 0.5818553478025612, + "flos": 23331163029120.0, + "grad_norm": 1.7037471478863104, + "language_loss": 0.78701043, + "learning_rate": 1.570268986810423e-06, + "loss": 0.80902302, + "num_input_tokens_seen": 104340700, + "step": 4839, + "time_per_iteration": 2.6168265342712402 + }, + { + "auxiliary_loss_clip": 0.01123963, + "auxiliary_loss_mlp": 0.0102792, + "balance_loss_clip": 0.97460645, + "balance_loss_mlp": 1.02069902, + "epoch": 0.5819755906932003, + "flos": 20996143603200.0, + "grad_norm": 1.8987747777203283, + "language_loss": 0.74574733, + "learning_rate": 1.5695082417137096e-06, + "loss": 0.76726615, + "num_input_tokens_seen": 104358575, + "step": 4840, + "time_per_iteration": 2.7705557346343994 + }, + { + "auxiliary_loss_clip": 0.01126771, + "auxiliary_loss_mlp": 0.01032474, + "balance_loss_clip": 0.97510898, + "balance_loss_mlp": 1.0250355, + "epoch": 0.5820958335838393, + "flos": 21431668008960.0, + "grad_norm": 1.7451538271280265, + "language_loss": 0.75386357, + "learning_rate": 1.5687475619222539e-06, + "loss": 0.77545595, + "num_input_tokens_seen": 104378530, + "step": 4841, + "time_per_iteration": 2.7745604515075684 + }, + { + "auxiliary_loss_clip": 0.0112413, + "auxiliary_loss_mlp": 0.01025685, + "balance_loss_clip": 0.9737131, + "balance_loss_mlp": 1.0178411, + "epoch": 0.5822160764744785, + "flos": 17967473660160.0, + "grad_norm": 2.097194380993032, + "language_loss": 0.73544466, + "learning_rate": 1.5679869475514496e-06, + "loss": 0.75694287, + "num_input_tokens_seen": 104395465, + "step": 4842, + "time_per_iteration": 2.680553436279297 + }, + { + "auxiliary_loss_clip": 0.01152332, + "auxiliary_loss_mlp": 0.01029184, + "balance_loss_clip": 1.01600361, + "balance_loss_mlp": 1.02142406, + "epoch": 0.5823363193651175, + "flos": 23033858158080.0, + "grad_norm": 1.9787200425497056, + "language_loss": 0.81406456, + "learning_rate": 1.567226398716682e-06, + "loss": 0.83587968, + "num_input_tokens_seen": 104415380, + "step": 4843, + "time_per_iteration": 2.6455557346343994 + }, + { + "auxiliary_loss_clip": 0.01136892, + "auxiliary_loss_mlp": 0.01028579, + "balance_loss_clip": 0.97793686, + "balance_loss_mlp": 1.01984668, + "epoch": 0.5824565622557566, + "flos": 32891840110080.0, + "grad_norm": 1.8324317401185917, + "language_loss": 0.61956322, + "learning_rate": 1.566465915533326e-06, + "loss": 0.64121789, + "num_input_tokens_seen": 104437410, + "step": 4844, + "time_per_iteration": 2.767376184463501 + }, + { + "auxiliary_loss_clip": 0.01151512, + "auxiliary_loss_mlp": 0.01030587, + "balance_loss_clip": 1.01549649, + "balance_loss_mlp": 1.02311277, + "epoch": 0.5825768051463958, + "flos": 22229674513920.0, + "grad_norm": 2.3143243479526516, + "language_loss": 0.88098443, + "learning_rate": 1.5657054981167458e-06, + "loss": 0.90280539, + "num_input_tokens_seen": 104456305, + "step": 4845, + "time_per_iteration": 2.716747522354126 + }, + { + "auxiliary_loss_clip": 0.01148516, + "auxiliary_loss_mlp": 0.0103168, + "balance_loss_clip": 1.01413393, + "balance_loss_mlp": 1.02432466, + "epoch": 0.5826970480370348, + "flos": 28001561016960.0, + "grad_norm": 1.7544998978409065, + "language_loss": 0.67814004, + "learning_rate": 1.5649451465822965e-06, + "loss": 0.69994205, + "num_input_tokens_seen": 104477695, + "step": 4846, + "time_per_iteration": 2.7237255573272705 + }, + { + "auxiliary_loss_clip": 0.01080389, + "auxiliary_loss_mlp": 0.01027559, + "balance_loss_clip": 0.90141976, + "balance_loss_mlp": 1.02016807, + "epoch": 0.5828172909276739, + "flos": 17858053854720.0, + "grad_norm": 1.6404196386145897, + "language_loss": 0.83713615, + "learning_rate": 1.5641848610453218e-06, + "loss": 0.85821563, + "num_input_tokens_seen": 104496355, + "step": 4847, + "time_per_iteration": 2.719616174697876 + }, + { + "auxiliary_loss_clip": 0.0115174, + "auxiliary_loss_mlp": 0.01027914, + "balance_loss_clip": 1.01704228, + "balance_loss_mlp": 1.02030838, + "epoch": 0.582937533818313, + "flos": 19865244827520.0, + "grad_norm": 2.594943589755509, + "language_loss": 0.86380577, + "learning_rate": 1.563424641621158e-06, + "loss": 0.8856023, + "num_input_tokens_seen": 104515535, + "step": 4848, + "time_per_iteration": 3.5702810287475586 + }, + { + "auxiliary_loss_clip": 0.01135447, + "auxiliary_loss_mlp": 0.01029374, + "balance_loss_clip": 0.97825611, + "balance_loss_mlp": 1.02202487, + "epoch": 0.5830577767089521, + "flos": 26870734068480.0, + "grad_norm": 1.873792295659048, + "language_loss": 0.69935459, + "learning_rate": 1.5626644884251282e-06, + "loss": 0.72100282, + "num_input_tokens_seen": 104535055, + "step": 4849, + "time_per_iteration": 3.6310644149780273 + }, + { + "auxiliary_loss_clip": 0.01174576, + "auxiliary_loss_mlp": 0.01028882, + "balance_loss_clip": 1.05348468, + "balance_loss_mlp": 1.02134848, + "epoch": 0.5831780195995911, + "flos": 25298205575040.0, + "grad_norm": 1.9156515176873663, + "language_loss": 0.88237643, + "learning_rate": 1.5619044015725488e-06, + "loss": 0.90441096, + "num_input_tokens_seen": 104554745, + "step": 4850, + "time_per_iteration": 3.5733118057250977 + }, + { + "auxiliary_loss_clip": 0.01185557, + "auxiliary_loss_mlp": 0.01031639, + "balance_loss_clip": 1.05988741, + "balance_loss_mlp": 1.02360487, + "epoch": 0.5832982624902303, + "flos": 14756988049920.0, + "grad_norm": 2.505365948890727, + "language_loss": 0.86988044, + "learning_rate": 1.5611443811787224e-06, + "loss": 0.89205241, + "num_input_tokens_seen": 104568870, + "step": 4851, + "time_per_iteration": 2.539273262023926 + }, + { + "auxiliary_loss_clip": 0.01152879, + "auxiliary_loss_mlp": 0.01028799, + "balance_loss_clip": 1.01740336, + "balance_loss_mlp": 1.02140832, + "epoch": 0.5834185053808694, + "flos": 20444555376000.0, + "grad_norm": 3.3570580724957266, + "language_loss": 0.69468832, + "learning_rate": 1.560384427358945e-06, + "loss": 0.71650511, + "num_input_tokens_seen": 104588415, + "step": 4852, + "time_per_iteration": 2.6246299743652344 + }, + { + "auxiliary_loss_clip": 0.01122381, + "auxiliary_loss_mlp": 0.01027278, + "balance_loss_clip": 0.97196925, + "balance_loss_mlp": 1.01902294, + "epoch": 0.5835387482715084, + "flos": 27200394115200.0, + "grad_norm": 1.4387888255980352, + "language_loss": 0.72921753, + "learning_rate": 1.5596245402284998e-06, + "loss": 0.75071412, + "num_input_tokens_seen": 104611940, + "step": 4853, + "time_per_iteration": 2.726141929626465 + }, + { + "auxiliary_loss_clip": 0.01160972, + "auxiliary_loss_mlp": 0.01027667, + "balance_loss_clip": 1.02193964, + "balance_loss_mlp": 1.02023435, + "epoch": 0.5836589911621476, + "flos": 16654615562880.0, + "grad_norm": 1.7544971455638432, + "language_loss": 0.81859291, + "learning_rate": 1.5588647199026619e-06, + "loss": 0.84047925, + "num_input_tokens_seen": 104629675, + "step": 4854, + "time_per_iteration": 2.681224822998047 + }, + { + "auxiliary_loss_clip": 0.01183354, + "auxiliary_loss_mlp": 0.01027144, + "balance_loss_clip": 1.05813575, + "balance_loss_mlp": 1.01985478, + "epoch": 0.5837792340527866, + "flos": 20446817932800.0, + "grad_norm": 17.637941500751154, + "language_loss": 0.87841368, + "learning_rate": 1.5581049664966956e-06, + "loss": 0.9005186, + "num_input_tokens_seen": 104647435, + "step": 4855, + "time_per_iteration": 2.5596365928649902 + }, + { + "auxiliary_loss_clip": 0.00980671, + "auxiliary_loss_mlp": 0.01003506, + "balance_loss_clip": 0.8345058, + "balance_loss_mlp": 1.00225472, + "epoch": 0.5838994769434257, + "flos": 65995480765440.0, + "grad_norm": 0.9930299120110131, + "language_loss": 0.65122718, + "learning_rate": 1.5573452801258545e-06, + "loss": 0.67106891, + "num_input_tokens_seen": 104694605, + "step": 4856, + "time_per_iteration": 3.160860061645508 + }, + { + "auxiliary_loss_clip": 0.01159736, + "auxiliary_loss_mlp": 0.01030361, + "balance_loss_clip": 1.01779389, + "balance_loss_mlp": 1.02295256, + "epoch": 0.5840197198340649, + "flos": 21470523546240.0, + "grad_norm": 1.9110019127393085, + "language_loss": 0.63522768, + "learning_rate": 1.5565856609053824e-06, + "loss": 0.65712869, + "num_input_tokens_seen": 104713400, + "step": 4857, + "time_per_iteration": 2.66914439201355 + }, + { + "auxiliary_loss_clip": 0.01179434, + "auxiliary_loss_mlp": 0.01023593, + "balance_loss_clip": 1.0569073, + "balance_loss_mlp": 1.0157963, + "epoch": 0.5841399627247039, + "flos": 19135144984320.0, + "grad_norm": 1.9119542838375925, + "language_loss": 0.80253679, + "learning_rate": 1.5558261089505127e-06, + "loss": 0.82456708, + "num_input_tokens_seen": 104732130, + "step": 4858, + "time_per_iteration": 2.6430914402008057 + }, + { + "auxiliary_loss_clip": 0.0115425, + "auxiliary_loss_mlp": 0.01023815, + "balance_loss_clip": 1.01744533, + "balance_loss_mlp": 1.01643026, + "epoch": 0.584260205615343, + "flos": 26425692558720.0, + "grad_norm": 2.355990892713687, + "language_loss": 0.79798257, + "learning_rate": 1.5550666243764697e-06, + "loss": 0.8197633, + "num_input_tokens_seen": 104750290, + "step": 4859, + "time_per_iteration": 2.72301983833313 + }, + { + "auxiliary_loss_clip": 0.01155369, + "auxiliary_loss_mlp": 0.01035018, + "balance_loss_clip": 1.01714671, + "balance_loss_mlp": 1.02729881, + "epoch": 0.584380448505982, + "flos": 13881809174400.0, + "grad_norm": 2.358012048999748, + "language_loss": 0.77415586, + "learning_rate": 1.554307207298465e-06, + "loss": 0.79605973, + "num_input_tokens_seen": 104768550, + "step": 4860, + "time_per_iteration": 3.506056785583496 + }, + { + "auxiliary_loss_clip": 0.01179973, + "auxiliary_loss_mlp": 0.01027026, + "balance_loss_clip": 1.05590677, + "balance_loss_mlp": 1.01889062, + "epoch": 0.5845006913966212, + "flos": 21543709507200.0, + "grad_norm": 1.841118853186124, + "language_loss": 0.78680694, + "learning_rate": 1.553547857831704e-06, + "loss": 0.80887687, + "num_input_tokens_seen": 104785060, + "step": 4861, + "time_per_iteration": 2.5811352729797363 + }, + { + "auxiliary_loss_clip": 0.01072604, + "auxiliary_loss_mlp": 0.00999241, + "balance_loss_clip": 1.01733923, + "balance_loss_mlp": 0.99804896, + "epoch": 0.5846209342872603, + "flos": 58375452712320.0, + "grad_norm": 0.8904043623475232, + "language_loss": 0.64205986, + "learning_rate": 1.5527885760913771e-06, + "loss": 0.66277832, + "num_input_tokens_seen": 104834950, + "step": 4862, + "time_per_iteration": 2.9865989685058594 + }, + { + "auxiliary_loss_clip": 0.0112986, + "auxiliary_loss_mlp": 0.01023762, + "balance_loss_clip": 0.98003459, + "balance_loss_mlp": 1.01703882, + "epoch": 0.5847411771778993, + "flos": 18588045957120.0, + "grad_norm": 1.6088083154586452, + "language_loss": 0.76492858, + "learning_rate": 1.552029362192668e-06, + "loss": 0.78646481, + "num_input_tokens_seen": 104854210, + "step": 4863, + "time_per_iteration": 2.6610729694366455 + }, + { + "auxiliary_loss_clip": 0.01100689, + "auxiliary_loss_mlp": 0.01023642, + "balance_loss_clip": 0.93662584, + "balance_loss_mlp": 1.0159117, + "epoch": 0.5848614200685385, + "flos": 24240780069120.0, + "grad_norm": 2.037722487366739, + "language_loss": 0.7262553, + "learning_rate": 1.5512702162507478e-06, + "loss": 0.74749857, + "num_input_tokens_seen": 104874525, + "step": 4864, + "time_per_iteration": 2.6961019039154053 + }, + { + "auxiliary_loss_clip": 0.01034283, + "auxiliary_loss_mlp": 0.01002108, + "balance_loss_clip": 0.94333422, + "balance_loss_mlp": 1.00094604, + "epoch": 0.5849816629591775, + "flos": 71660245933440.0, + "grad_norm": 1.103645980106742, + "language_loss": 0.55782181, + "learning_rate": 1.5505111383807792e-06, + "loss": 0.57818574, + "num_input_tokens_seen": 104937195, + "step": 4865, + "time_per_iteration": 3.2922165393829346 + }, + { + "auxiliary_loss_clip": 0.01085667, + "auxiliary_loss_mlp": 0.0102962, + "balance_loss_clip": 0.9000473, + "balance_loss_mlp": 1.02183521, + "epoch": 0.5851019058498166, + "flos": 23802095266560.0, + "grad_norm": 1.9399452971998856, + "language_loss": 0.80762082, + "learning_rate": 1.5497521286979138e-06, + "loss": 0.82877374, + "num_input_tokens_seen": 104957435, + "step": 4866, + "time_per_iteration": 2.773164987564087 + }, + { + "auxiliary_loss_clip": 0.01109884, + "auxiliary_loss_mlp": 0.01027091, + "balance_loss_clip": 0.93855846, + "balance_loss_mlp": 1.01912189, + "epoch": 0.5852221487404557, + "flos": 24388516707840.0, + "grad_norm": 2.438981810489429, + "language_loss": 0.74133372, + "learning_rate": 1.5489931873172927e-06, + "loss": 0.76270342, + "num_input_tokens_seen": 104978755, + "step": 4867, + "time_per_iteration": 2.750986099243164 + }, + { + "auxiliary_loss_clip": 0.0104243, + "auxiliary_loss_mlp": 0.01026778, + "balance_loss_clip": 0.85694897, + "balance_loss_mlp": 1.01907694, + "epoch": 0.5853423916310948, + "flos": 27271425260160.0, + "grad_norm": 1.8145018388832381, + "language_loss": 0.79614854, + "learning_rate": 1.5482343143540467e-06, + "loss": 0.81684065, + "num_input_tokens_seen": 105000020, + "step": 4868, + "time_per_iteration": 2.8303802013397217 + }, + { + "auxiliary_loss_clip": 0.01109874, + "auxiliary_loss_mlp": 0.00509382, + "balance_loss_clip": 0.93987703, + "balance_loss_mlp": 0.0, + "epoch": 0.5854626345217339, + "flos": 11983786611840.0, + "grad_norm": 2.2801775538165368, + "language_loss": 0.83039308, + "learning_rate": 1.547475509923295e-06, + "loss": 0.84658557, + "num_input_tokens_seen": 105017060, + "step": 4869, + "time_per_iteration": 2.6744425296783447 + }, + { + "auxiliary_loss_clip": 0.00996355, + "auxiliary_loss_mlp": 0.01001876, + "balance_loss_clip": 0.86796236, + "balance_loss_mlp": 1.00067234, + "epoch": 0.585582877412373, + "flos": 64342335173760.0, + "grad_norm": 0.7667980992226737, + "language_loss": 0.56092489, + "learning_rate": 1.5467167741401495e-06, + "loss": 0.58090723, + "num_input_tokens_seen": 105078540, + "step": 4870, + "time_per_iteration": 3.289825677871704 + }, + { + "auxiliary_loss_clip": 0.01128606, + "auxiliary_loss_mlp": 0.01023594, + "balance_loss_clip": 0.97628772, + "balance_loss_mlp": 1.01581597, + "epoch": 0.5857031203030121, + "flos": 17011926103680.0, + "grad_norm": 2.0462709353322066, + "language_loss": 0.71479112, + "learning_rate": 1.5459581071197083e-06, + "loss": 0.7363131, + "num_input_tokens_seen": 105094200, + "step": 4871, + "time_per_iteration": 2.6792335510253906 + }, + { + "auxiliary_loss_clip": 0.01155827, + "auxiliary_loss_mlp": 0.01026884, + "balance_loss_clip": 1.01809192, + "balance_loss_mlp": 1.01985407, + "epoch": 0.5858233631936511, + "flos": 20885682303360.0, + "grad_norm": 2.5159398755930673, + "language_loss": 0.83143342, + "learning_rate": 1.5451995089770624e-06, + "loss": 0.85326058, + "num_input_tokens_seen": 105113985, + "step": 4872, + "time_per_iteration": 2.672938585281372 + }, + { + "auxiliary_loss_clip": 0.01175573, + "auxiliary_loss_mlp": 0.01022074, + "balance_loss_clip": 1.05497885, + "balance_loss_mlp": 1.0148921, + "epoch": 0.5859436060842903, + "flos": 23191902000000.0, + "grad_norm": 1.2677951519623252, + "language_loss": 0.71956956, + "learning_rate": 1.5444409798272885e-06, + "loss": 0.74154603, + "num_input_tokens_seen": 105138075, + "step": 4873, + "time_per_iteration": 2.7048702239990234 + }, + { + "auxiliary_loss_clip": 0.01106384, + "auxiliary_loss_mlp": 0.0102516, + "balance_loss_clip": 0.93862653, + "balance_loss_mlp": 1.01747084, + "epoch": 0.5860638489749294, + "flos": 22492648961280.0, + "grad_norm": 1.9460807643470375, + "language_loss": 0.80819052, + "learning_rate": 1.543682519785456e-06, + "loss": 0.82950592, + "num_input_tokens_seen": 105156555, + "step": 4874, + "time_per_iteration": 2.7007200717926025 + }, + { + "auxiliary_loss_clip": 0.01127089, + "auxiliary_loss_mlp": 0.01024391, + "balance_loss_clip": 0.97603571, + "balance_loss_mlp": 1.01707447, + "epoch": 0.5861840918655684, + "flos": 17566243764480.0, + "grad_norm": 2.3130415346443036, + "language_loss": 0.7978912, + "learning_rate": 1.5429241289666219e-06, + "loss": 0.81940603, + "num_input_tokens_seen": 105174055, + "step": 4875, + "time_per_iteration": 4.484738826751709 + }, + { + "auxiliary_loss_clip": 0.01121569, + "auxiliary_loss_mlp": 0.0102468, + "balance_loss_clip": 0.97563457, + "balance_loss_mlp": 1.01759315, + "epoch": 0.5863043347562076, + "flos": 25556152118400.0, + "grad_norm": 2.8173637835841494, + "language_loss": 0.69958889, + "learning_rate": 1.5421658074858342e-06, + "loss": 0.72105139, + "num_input_tokens_seen": 105192160, + "step": 4876, + "time_per_iteration": 3.623194456100464 + }, + { + "auxiliary_loss_clip": 0.01125031, + "auxiliary_loss_mlp": 0.01027933, + "balance_loss_clip": 0.9775269, + "balance_loss_mlp": 1.01962686, + "epoch": 0.5864245776468466, + "flos": 20667525050880.0, + "grad_norm": 3.062112157565505, + "language_loss": 0.6645, + "learning_rate": 1.5414075554581298e-06, + "loss": 0.68602967, + "num_input_tokens_seen": 105210205, + "step": 4877, + "time_per_iteration": 2.6966922283172607 + }, + { + "auxiliary_loss_clip": 0.01178631, + "auxiliary_loss_mlp": 0.01032409, + "balance_loss_clip": 1.05569172, + "balance_loss_mlp": 1.02419591, + "epoch": 0.5865448205374857, + "flos": 28913907490560.0, + "grad_norm": 2.4064912216727268, + "language_loss": 0.78833973, + "learning_rate": 1.5406493729985348e-06, + "loss": 0.8104502, + "num_input_tokens_seen": 105229400, + "step": 4878, + "time_per_iteration": 2.6708333492279053 + }, + { + "auxiliary_loss_clip": 0.01089819, + "auxiliary_loss_mlp": 0.00509494, + "balance_loss_clip": 0.90387714, + "balance_loss_mlp": 0.0, + "epoch": 0.5866650634281249, + "flos": 25842575168640.0, + "grad_norm": 2.133280176174061, + "language_loss": 0.72019768, + "learning_rate": 1.5398912602220644e-06, + "loss": 0.7361908, + "num_input_tokens_seen": 105248675, + "step": 4879, + "time_per_iteration": 2.7356250286102295 + }, + { + "auxiliary_loss_clip": 0.010967, + "auxiliary_loss_mlp": 0.01028441, + "balance_loss_clip": 0.9022361, + "balance_loss_mlp": 1.02042401, + "epoch": 0.5867853063187639, + "flos": 17052325925760.0, + "grad_norm": 2.025132242457735, + "language_loss": 0.78875744, + "learning_rate": 1.539133217243724e-06, + "loss": 0.81000888, + "num_input_tokens_seen": 105265695, + "step": 4880, + "time_per_iteration": 2.7141354084014893 + }, + { + "auxiliary_loss_clip": 0.01110277, + "auxiliary_loss_mlp": 0.01032317, + "balance_loss_clip": 0.93882209, + "balance_loss_mlp": 1.02368045, + "epoch": 0.586905549209403, + "flos": 24645026707200.0, + "grad_norm": 1.8906978547902993, + "language_loss": 0.7587685, + "learning_rate": 1.5383752441785081e-06, + "loss": 0.78019446, + "num_input_tokens_seen": 105284920, + "step": 4881, + "time_per_iteration": 2.723245859146118 + }, + { + "auxiliary_loss_clip": 0.01154578, + "auxiliary_loss_mlp": 0.01031084, + "balance_loss_clip": 1.01395106, + "balance_loss_mlp": 1.02340722, + "epoch": 0.5870257921000421, + "flos": 14720538723840.0, + "grad_norm": 2.2305476262288044, + "language_loss": 0.85833001, + "learning_rate": 1.5376173411414003e-06, + "loss": 0.88018656, + "num_input_tokens_seen": 105302960, + "step": 4882, + "time_per_iteration": 2.616398572921753 + }, + { + "auxiliary_loss_clip": 0.01129444, + "auxiliary_loss_mlp": 0.0102784, + "balance_loss_clip": 0.974154, + "balance_loss_mlp": 1.02006161, + "epoch": 0.5871460349906812, + "flos": 23914998691200.0, + "grad_norm": 2.275442649945292, + "language_loss": 0.7929374, + "learning_rate": 1.5368595082473753e-06, + "loss": 0.81451023, + "num_input_tokens_seen": 105321260, + "step": 4883, + "time_per_iteration": 2.7155497074127197 + }, + { + "auxiliary_loss_clip": 0.01154569, + "auxiliary_loss_mlp": 0.01026648, + "balance_loss_clip": 1.01560998, + "balance_loss_mlp": 1.01888156, + "epoch": 0.5872662778813202, + "flos": 22164174063360.0, + "grad_norm": 1.8009349144778208, + "language_loss": 0.78155088, + "learning_rate": 1.5361017456113935e-06, + "loss": 0.80336308, + "num_input_tokens_seen": 105341610, + "step": 4884, + "time_per_iteration": 2.664186477661133 + }, + { + "auxiliary_loss_clip": 0.01153099, + "auxiliary_loss_mlp": 0.01027058, + "balance_loss_clip": 1.01339233, + "balance_loss_mlp": 1.0184927, + "epoch": 0.5873865207719594, + "flos": 18441925430400.0, + "grad_norm": 1.9482249323278564, + "language_loss": 0.85840303, + "learning_rate": 1.5353440533484085e-06, + "loss": 0.88020468, + "num_input_tokens_seen": 105360465, + "step": 4885, + "time_per_iteration": 2.6868412494659424 + }, + { + "auxiliary_loss_clip": 0.0113009, + "auxiliary_loss_mlp": 0.01024753, + "balance_loss_clip": 0.9766953, + "balance_loss_mlp": 1.01702857, + "epoch": 0.5875067636625985, + "flos": 54015321427200.0, + "grad_norm": 1.7606453680627305, + "language_loss": 0.66060704, + "learning_rate": 1.534586431573361e-06, + "loss": 0.68215549, + "num_input_tokens_seen": 105385405, + "step": 4886, + "time_per_iteration": 3.8649988174438477 + }, + { + "auxiliary_loss_clip": 0.01057593, + "auxiliary_loss_mlp": 0.01032959, + "balance_loss_clip": 0.859267, + "balance_loss_mlp": 1.02462673, + "epoch": 0.5876270065532375, + "flos": 27995707100160.0, + "grad_norm": 1.9065480365318266, + "language_loss": 0.79361719, + "learning_rate": 1.5338288804011817e-06, + "loss": 0.81452274, + "num_input_tokens_seen": 105404905, + "step": 4887, + "time_per_iteration": 2.8512141704559326 + }, + { + "auxiliary_loss_clip": 0.01123767, + "auxiliary_loss_mlp": 0.01029248, + "balance_loss_clip": 0.97326177, + "balance_loss_mlp": 1.02105856, + "epoch": 0.5877472494438767, + "flos": 21361462876800.0, + "grad_norm": 1.7665250568523658, + "language_loss": 0.72029281, + "learning_rate": 1.533071399946791e-06, + "loss": 0.74182296, + "num_input_tokens_seen": 105423650, + "step": 4888, + "time_per_iteration": 2.6937859058380127 + }, + { + "auxiliary_loss_clip": 0.01131122, + "auxiliary_loss_mlp": 0.01026782, + "balance_loss_clip": 0.9761765, + "balance_loss_mlp": 1.01899803, + "epoch": 0.5878674923345157, + "flos": 22383013674240.0, + "grad_norm": 1.6870098103534712, + "language_loss": 0.57017809, + "learning_rate": 1.5323139903250977e-06, + "loss": 0.59175718, + "num_input_tokens_seen": 105444255, + "step": 4889, + "time_per_iteration": 2.7892704010009766 + }, + { + "auxiliary_loss_clip": 0.01133525, + "auxiliary_loss_mlp": 0.01027409, + "balance_loss_clip": 0.98114443, + "balance_loss_mlp": 1.01966286, + "epoch": 0.5879877352251548, + "flos": 21868664872320.0, + "grad_norm": 1.564556060818564, + "language_loss": 0.77056766, + "learning_rate": 1.5315566516510002e-06, + "loss": 0.79217696, + "num_input_tokens_seen": 105462425, + "step": 4890, + "time_per_iteration": 2.7085304260253906 + }, + { + "auxiliary_loss_clip": 0.0117856, + "auxiliary_loss_mlp": 0.01031061, + "balance_loss_clip": 1.05580616, + "balance_loss_mlp": 1.02369034, + "epoch": 0.5881079781157939, + "flos": 17493811989120.0, + "grad_norm": 2.0285351628478567, + "language_loss": 0.67545676, + "learning_rate": 1.5307993840393857e-06, + "loss": 0.69755292, + "num_input_tokens_seen": 105480505, + "step": 4891, + "time_per_iteration": 2.559962272644043 + }, + { + "auxiliary_loss_clip": 0.01175488, + "auxiliary_loss_mlp": 0.01025415, + "balance_loss_clip": 1.0526247, + "balance_loss_mlp": 1.01809549, + "epoch": 0.588228221006433, + "flos": 22601853285120.0, + "grad_norm": 1.9007748011107275, + "language_loss": 0.80238652, + "learning_rate": 1.530042187605132e-06, + "loss": 0.8243956, + "num_input_tokens_seen": 105499760, + "step": 4892, + "time_per_iteration": 2.6827404499053955 + }, + { + "auxiliary_loss_clip": 0.01153214, + "auxiliary_loss_mlp": 0.00509093, + "balance_loss_clip": 1.01612592, + "balance_loss_mlp": 0.0, + "epoch": 0.5883484638970721, + "flos": 26176939896960.0, + "grad_norm": 1.3252416806992213, + "language_loss": 0.84337616, + "learning_rate": 1.5292850624631044e-06, + "loss": 0.85999918, + "num_input_tokens_seen": 105521955, + "step": 4893, + "time_per_iteration": 2.697732925415039 + }, + { + "auxiliary_loss_clip": 0.01153585, + "auxiliary_loss_mlp": 0.01030212, + "balance_loss_clip": 1.01924205, + "balance_loss_mlp": 1.02219486, + "epoch": 0.5884687067877111, + "flos": 30443737691520.0, + "grad_norm": 2.2049170021740343, + "language_loss": 0.80437088, + "learning_rate": 1.5285280087281593e-06, + "loss": 0.82620883, + "num_input_tokens_seen": 105542685, + "step": 4894, + "time_per_iteration": 2.715907335281372 + }, + { + "auxiliary_loss_clip": 0.0103495, + "auxiliary_loss_mlp": 0.01011738, + "balance_loss_clip": 0.94325602, + "balance_loss_mlp": 1.01052845, + "epoch": 0.5885889496783503, + "flos": 70507550580480.0, + "grad_norm": 0.6508910472820905, + "language_loss": 0.56599063, + "learning_rate": 1.5277710265151398e-06, + "loss": 0.58645755, + "num_input_tokens_seen": 105612165, + "step": 4895, + "time_per_iteration": 3.379734992980957 + }, + { + "auxiliary_loss_clip": 0.01151912, + "auxiliary_loss_mlp": 0.01032939, + "balance_loss_clip": 1.01445782, + "balance_loss_mlp": 1.02490425, + "epoch": 0.5887091925689893, + "flos": 19098767485440.0, + "grad_norm": 2.575116281549512, + "language_loss": 0.77185196, + "learning_rate": 1.5270141159388803e-06, + "loss": 0.79370046, + "num_input_tokens_seen": 105629185, + "step": 4896, + "time_per_iteration": 2.640766143798828 + }, + { + "auxiliary_loss_clip": 0.01175853, + "auxiliary_loss_mlp": 0.01026556, + "balance_loss_clip": 1.05320752, + "balance_loss_mlp": 1.01905227, + "epoch": 0.5888294354596284, + "flos": 23294282739840.0, + "grad_norm": 1.721897474767402, + "language_loss": 0.8035478, + "learning_rate": 1.526257277114203e-06, + "loss": 0.82557189, + "num_input_tokens_seen": 105650260, + "step": 4897, + "time_per_iteration": 2.613607883453369 + }, + { + "auxiliary_loss_clip": 0.01125429, + "auxiliary_loss_mlp": 0.01027531, + "balance_loss_clip": 0.97729278, + "balance_loss_mlp": 1.01979482, + "epoch": 0.5889496783502676, + "flos": 21981532383360.0, + "grad_norm": 1.961402437951699, + "language_loss": 0.79509586, + "learning_rate": 1.5255005101559201e-06, + "loss": 0.81662548, + "num_input_tokens_seen": 105667870, + "step": 4898, + "time_per_iteration": 2.6685056686401367 + }, + { + "auxiliary_loss_clip": 0.01153008, + "auxiliary_loss_mlp": 0.01023849, + "balance_loss_clip": 1.01452899, + "balance_loss_mlp": 1.0164814, + "epoch": 0.5890699212409066, + "flos": 21685233093120.0, + "grad_norm": 1.9997548709622635, + "language_loss": 0.76825315, + "learning_rate": 1.524743815178833e-06, + "loss": 0.79002166, + "num_input_tokens_seen": 105685830, + "step": 4899, + "time_per_iteration": 2.603318214416504 + }, + { + "auxiliary_loss_clip": 0.01127536, + "auxiliary_loss_mlp": 0.01027089, + "balance_loss_clip": 0.97438812, + "balance_loss_mlp": 1.01919115, + "epoch": 0.5891901641315457, + "flos": 19464553635840.0, + "grad_norm": 1.652228752217458, + "language_loss": 0.80875379, + "learning_rate": 1.5239871922977315e-06, + "loss": 0.83030003, + "num_input_tokens_seen": 105705745, + "step": 4900, + "time_per_iteration": 3.6180784702301025 + }, + { + "auxiliary_loss_clip": 0.011259, + "auxiliary_loss_mlp": 0.0102456, + "balance_loss_clip": 0.97493821, + "balance_loss_mlp": 1.01657927, + "epoch": 0.5893104070221848, + "flos": 19609884063360.0, + "grad_norm": 1.6323793039940915, + "language_loss": 0.89497185, + "learning_rate": 1.523230641627394e-06, + "loss": 0.91647649, + "num_input_tokens_seen": 105724730, + "step": 4901, + "time_per_iteration": 3.6307973861694336 + }, + { + "auxiliary_loss_clip": 0.01084765, + "auxiliary_loss_mlp": 0.01030558, + "balance_loss_clip": 0.89758897, + "balance_loss_mlp": 1.02274346, + "epoch": 0.5894306499128239, + "flos": 29060063930880.0, + "grad_norm": 2.1684023748793955, + "language_loss": 0.73019665, + "learning_rate": 1.5224741632825888e-06, + "loss": 0.75134993, + "num_input_tokens_seen": 105744920, + "step": 4902, + "time_per_iteration": 3.7082254886627197 + }, + { + "auxiliary_loss_clip": 0.01177312, + "auxiliary_loss_mlp": 0.01025531, + "balance_loss_clip": 1.0538429, + "balance_loss_mlp": 1.01809263, + "epoch": 0.589550892803463, + "flos": 42298890721920.0, + "grad_norm": 1.7194785759737377, + "language_loss": 0.6949513, + "learning_rate": 1.521717757378074e-06, + "loss": 0.71697974, + "num_input_tokens_seen": 105765465, + "step": 4903, + "time_per_iteration": 2.8119895458221436 + }, + { + "auxiliary_loss_clip": 0.01154685, + "auxiliary_loss_mlp": 0.0102875, + "balance_loss_clip": 1.01482534, + "balance_loss_mlp": 1.02112675, + "epoch": 0.5896711356941021, + "flos": 14137062197760.0, + "grad_norm": 1.719444629832071, + "language_loss": 0.69126189, + "learning_rate": 1.5209614240285943e-06, + "loss": 0.71309626, + "num_input_tokens_seen": 105783120, + "step": 4904, + "time_per_iteration": 2.6725380420684814 + }, + { + "auxiliary_loss_clip": 0.01176845, + "auxiliary_loss_mlp": 0.00510516, + "balance_loss_clip": 1.05425358, + "balance_loss_mlp": 0.0, + "epoch": 0.5897913785847412, + "flos": 17201355454080.0, + "grad_norm": 2.509921970069837, + "language_loss": 0.84760129, + "learning_rate": 1.520205163348887e-06, + "loss": 0.86447489, + "num_input_tokens_seen": 105801055, + "step": 4905, + "time_per_iteration": 2.596705198287964 + }, + { + "auxiliary_loss_clip": 0.01017423, + "auxiliary_loss_mlp": 0.01000391, + "balance_loss_clip": 0.90738142, + "balance_loss_mlp": 0.99920446, + "epoch": 0.5899116214753802, + "flos": 48794164202880.0, + "grad_norm": 0.7234180390105637, + "language_loss": 0.56981695, + "learning_rate": 1.519448975453674e-06, + "loss": 0.58999515, + "num_input_tokens_seen": 105856155, + "step": 4906, + "time_per_iteration": 3.0879433155059814 + }, + { + "auxiliary_loss_clip": 0.01152709, + "auxiliary_loss_mlp": 0.00509754, + "balance_loss_clip": 1.01760018, + "balance_loss_mlp": 0.0, + "epoch": 0.5900318643660194, + "flos": 21103659987840.0, + "grad_norm": 23.88898034320856, + "language_loss": 0.76823872, + "learning_rate": 1.5186928604576696e-06, + "loss": 0.78486335, + "num_input_tokens_seen": 105873350, + "step": 4907, + "time_per_iteration": 2.6367204189300537 + }, + { + "auxiliary_loss_clip": 0.01128594, + "auxiliary_loss_mlp": 0.01026822, + "balance_loss_clip": 0.97430801, + "balance_loss_mlp": 1.01965117, + "epoch": 0.5901521072566585, + "flos": 21178390233600.0, + "grad_norm": 2.092644603165961, + "language_loss": 0.77273959, + "learning_rate": 1.5179368184755752e-06, + "loss": 0.79429376, + "num_input_tokens_seen": 105891435, + "step": 4908, + "time_per_iteration": 2.698035717010498 + }, + { + "auxiliary_loss_clip": 0.01126179, + "auxiliary_loss_mlp": 0.01025259, + "balance_loss_clip": 0.97522807, + "balance_loss_mlp": 1.0175221, + "epoch": 0.5902723501472975, + "flos": 20225967160320.0, + "grad_norm": 2.187069295424605, + "language_loss": 0.83167416, + "learning_rate": 1.5171808496220821e-06, + "loss": 0.85318851, + "num_input_tokens_seen": 105910190, + "step": 4909, + "time_per_iteration": 2.6971187591552734 + }, + { + "auxiliary_loss_clip": 0.01132705, + "auxiliary_loss_mlp": 0.01030901, + "balance_loss_clip": 0.97572756, + "balance_loss_mlp": 1.02298498, + "epoch": 0.5903925930379367, + "flos": 22964407211520.0, + "grad_norm": 1.9236784072621673, + "language_loss": 0.81388915, + "learning_rate": 1.5164249540118708e-06, + "loss": 0.83552521, + "num_input_tokens_seen": 105929315, + "step": 4910, + "time_per_iteration": 2.6846354007720947 + }, + { + "auxiliary_loss_clip": 0.01062891, + "auxiliary_loss_mlp": 0.01031506, + "balance_loss_clip": 0.86008549, + "balance_loss_mlp": 1.0237807, + "epoch": 0.5905128359285757, + "flos": 23367720096000.0, + "grad_norm": 1.7018653859492625, + "language_loss": 0.83258915, + "learning_rate": 1.5156691317596093e-06, + "loss": 0.85353309, + "num_input_tokens_seen": 105950740, + "step": 4911, + "time_per_iteration": 2.8631744384765625 + }, + { + "auxiliary_loss_clip": 0.0115411, + "auxiliary_loss_mlp": 0.00509532, + "balance_loss_clip": 1.01557374, + "balance_loss_mlp": 0.0, + "epoch": 0.5906330788192148, + "flos": 28032335994240.0, + "grad_norm": 2.066746462320928, + "language_loss": 0.67190164, + "learning_rate": 1.5149133829799556e-06, + "loss": 0.68853807, + "num_input_tokens_seen": 105968735, + "step": 4912, + "time_per_iteration": 3.5826425552368164 + }, + { + "auxiliary_loss_clip": 0.01137774, + "auxiliary_loss_mlp": 0.01029711, + "balance_loss_clip": 0.97757947, + "balance_loss_mlp": 1.02172399, + "epoch": 0.590753321709854, + "flos": 18477943793280.0, + "grad_norm": 1.9365953826936206, + "language_loss": 0.80661738, + "learning_rate": 1.5141577077875556e-06, + "loss": 0.82829225, + "num_input_tokens_seen": 105986060, + "step": 4913, + "time_per_iteration": 2.6844098567962646 + }, + { + "auxiliary_loss_clip": 0.01153689, + "auxiliary_loss_mlp": 0.01030581, + "balance_loss_clip": 1.01454186, + "balance_loss_mlp": 1.02286184, + "epoch": 0.590873564600493, + "flos": 16873706568960.0, + "grad_norm": 1.6628019546031556, + "language_loss": 0.72152972, + "learning_rate": 1.5134021062970451e-06, + "loss": 0.74337244, + "num_input_tokens_seen": 106004440, + "step": 4914, + "time_per_iteration": 2.6317999362945557 + }, + { + "auxiliary_loss_clip": 0.01093989, + "auxiliary_loss_mlp": 0.01022436, + "balance_loss_clip": 0.93717051, + "balance_loss_mlp": 1.01509857, + "epoch": 0.5909938074911321, + "flos": 13516166678400.0, + "grad_norm": 1.7534972555076502, + "language_loss": 0.81067562, + "learning_rate": 1.5126465786230483e-06, + "loss": 0.83183992, + "num_input_tokens_seen": 106021215, + "step": 4915, + "time_per_iteration": 2.70628023147583 + }, + { + "auxiliary_loss_clip": 0.01175817, + "auxiliary_loss_mlp": 0.01026424, + "balance_loss_clip": 1.0543139, + "balance_loss_mlp": 1.01918232, + "epoch": 0.5911140503817712, + "flos": 26024067613440.0, + "grad_norm": 1.7175512805677697, + "language_loss": 0.81931186, + "learning_rate": 1.5118911248801787e-06, + "loss": 0.84133422, + "num_input_tokens_seen": 106039225, + "step": 4916, + "time_per_iteration": 2.5992531776428223 + }, + { + "auxiliary_loss_clip": 0.0114912, + "auxiliary_loss_mlp": 0.01023442, + "balance_loss_clip": 1.01608658, + "balance_loss_mlp": 1.01574707, + "epoch": 0.5912342932724103, + "flos": 23258731253760.0, + "grad_norm": 2.602047286515894, + "language_loss": 0.79909468, + "learning_rate": 1.5111357451830364e-06, + "loss": 0.82082027, + "num_input_tokens_seen": 106057920, + "step": 4917, + "time_per_iteration": 2.6783218383789062 + }, + { + "auxiliary_loss_clip": 0.01154501, + "auxiliary_loss_mlp": 0.01029453, + "balance_loss_clip": 1.01612198, + "balance_loss_mlp": 1.02187753, + "epoch": 0.5913545361630493, + "flos": 19573039687680.0, + "grad_norm": 1.7115962526646622, + "language_loss": 0.70906037, + "learning_rate": 1.5103804396462131e-06, + "loss": 0.73089993, + "num_input_tokens_seen": 106077855, + "step": 4918, + "time_per_iteration": 2.6428685188293457 + }, + { + "auxiliary_loss_clip": 0.01156399, + "auxiliary_loss_mlp": 0.01029968, + "balance_loss_clip": 1.01472783, + "balance_loss_mlp": 1.02148008, + "epoch": 0.5914747790536885, + "flos": 26213532877440.0, + "grad_norm": 1.9447862297745933, + "language_loss": 0.80410272, + "learning_rate": 1.5096252083842877e-06, + "loss": 0.82596636, + "num_input_tokens_seen": 106097065, + "step": 4919, + "time_per_iteration": 2.639151096343994 + }, + { + "auxiliary_loss_clip": 0.01150676, + "auxiliary_loss_mlp": 0.01029863, + "balance_loss_clip": 1.0131284, + "balance_loss_mlp": 1.02180409, + "epoch": 0.5915950219443276, + "flos": 27417545786880.0, + "grad_norm": 1.6555656394034468, + "language_loss": 0.85397995, + "learning_rate": 1.5088700515118285e-06, + "loss": 0.87578529, + "num_input_tokens_seen": 106116385, + "step": 4920, + "time_per_iteration": 2.6890995502471924 + }, + { + "auxiliary_loss_clip": 0.01102427, + "auxiliary_loss_mlp": 0.0102721, + "balance_loss_clip": 0.93965948, + "balance_loss_mlp": 1.0195148, + "epoch": 0.5917152648349666, + "flos": 21907879545600.0, + "grad_norm": 1.6775477307997633, + "language_loss": 0.66947556, + "learning_rate": 1.508114969143392e-06, + "loss": 0.69077194, + "num_input_tokens_seen": 106136370, + "step": 4921, + "time_per_iteration": 2.705770254135132 + }, + { + "auxiliary_loss_clip": 0.01130453, + "auxiliary_loss_mlp": 0.01025617, + "balance_loss_clip": 0.97498202, + "balance_loss_mlp": 1.01826751, + "epoch": 0.5918355077256057, + "flos": 28109185142400.0, + "grad_norm": 1.484006537533296, + "language_loss": 0.77522862, + "learning_rate": 1.5073599613935238e-06, + "loss": 0.79678935, + "num_input_tokens_seen": 106158490, + "step": 4922, + "time_per_iteration": 2.7673676013946533 + }, + { + "auxiliary_loss_clip": 0.01129882, + "auxiliary_loss_mlp": 0.01025471, + "balance_loss_clip": 0.97641098, + "balance_loss_mlp": 1.01778758, + "epoch": 0.5919557506162448, + "flos": 28183807647360.0, + "grad_norm": 1.85359125438353, + "language_loss": 0.57375699, + "learning_rate": 1.5066050283767574e-06, + "loss": 0.59531051, + "num_input_tokens_seen": 106179170, + "step": 4923, + "time_per_iteration": 2.6913232803344727 + }, + { + "auxiliary_loss_clip": 0.01125396, + "auxiliary_loss_mlp": 0.01024095, + "balance_loss_clip": 0.97628319, + "balance_loss_mlp": 1.01655531, + "epoch": 0.5920759935068839, + "flos": 12094355652480.0, + "grad_norm": 2.1434677539424234, + "language_loss": 0.83084565, + "learning_rate": 1.505850170207616e-06, + "loss": 0.85234052, + "num_input_tokens_seen": 106196035, + "step": 4924, + "time_per_iteration": 2.6284210681915283 + }, + { + "auxiliary_loss_clip": 0.01126559, + "auxiliary_loss_mlp": 0.01025901, + "balance_loss_clip": 0.9753992, + "balance_loss_mlp": 1.01798892, + "epoch": 0.592196236397523, + "flos": 29424772673280.0, + "grad_norm": 2.036228762071121, + "language_loss": 0.78150225, + "learning_rate": 1.505095387000611e-06, + "loss": 0.8030268, + "num_input_tokens_seen": 106218335, + "step": 4925, + "time_per_iteration": 2.7449164390563965 + }, + { + "auxiliary_loss_clip": 0.01122075, + "auxiliary_loss_mlp": 0.0103089, + "balance_loss_clip": 0.97652525, + "balance_loss_mlp": 1.02320409, + "epoch": 0.5923164792881621, + "flos": 24384709866240.0, + "grad_norm": 1.8143431859267658, + "language_loss": 0.74391782, + "learning_rate": 1.504340678870242e-06, + "loss": 0.76544738, + "num_input_tokens_seen": 106236550, + "step": 4926, + "time_per_iteration": 3.6400668621063232 + }, + { + "auxiliary_loss_clip": 0.01150324, + "auxiliary_loss_mlp": 0.01030324, + "balance_loss_clip": 1.01517332, + "balance_loss_mlp": 1.02265894, + "epoch": 0.5924367221788012, + "flos": 24024238928640.0, + "grad_norm": 1.846813471932787, + "language_loss": 0.89827859, + "learning_rate": 1.5035860459309989e-06, + "loss": 0.92008507, + "num_input_tokens_seen": 106254265, + "step": 4927, + "time_per_iteration": 3.578352212905884 + }, + { + "auxiliary_loss_clip": 0.01121121, + "auxiliary_loss_mlp": 0.01036356, + "balance_loss_clip": 0.9732765, + "balance_loss_mlp": 1.0280236, + "epoch": 0.5925569650694402, + "flos": 26870590414080.0, + "grad_norm": 1.8144851936649575, + "language_loss": 0.63948148, + "learning_rate": 1.5028314882973568e-06, + "loss": 0.66105628, + "num_input_tokens_seen": 106274670, + "step": 4928, + "time_per_iteration": 3.6691441535949707 + }, + { + "auxiliary_loss_clip": 0.01128714, + "auxiliary_loss_mlp": 0.01026575, + "balance_loss_clip": 0.97759628, + "balance_loss_mlp": 1.01852202, + "epoch": 0.5926772079600794, + "flos": 22302788647680.0, + "grad_norm": 1.8264309029272532, + "language_loss": 0.85034597, + "learning_rate": 1.502077006083783e-06, + "loss": 0.87189883, + "num_input_tokens_seen": 106293330, + "step": 4929, + "time_per_iteration": 2.658289909362793 + }, + { + "auxiliary_loss_clip": 0.01158108, + "auxiliary_loss_mlp": 0.00509184, + "balance_loss_clip": 1.01862693, + "balance_loss_mlp": 0.0, + "epoch": 0.5927974508507184, + "flos": 19865244827520.0, + "grad_norm": 1.8466564818688274, + "language_loss": 0.76544821, + "learning_rate": 1.5013225994047315e-06, + "loss": 0.78212118, + "num_input_tokens_seen": 106310960, + "step": 4930, + "time_per_iteration": 2.6486825942993164 + }, + { + "auxiliary_loss_clip": 0.01154661, + "auxiliary_loss_mlp": 0.00508799, + "balance_loss_clip": 1.01835811, + "balance_loss_mlp": 0.0, + "epoch": 0.5929176937413575, + "flos": 15776743167360.0, + "grad_norm": 1.523678462247109, + "language_loss": 0.80485934, + "learning_rate": 1.5005682683746452e-06, + "loss": 0.82149398, + "num_input_tokens_seen": 106329475, + "step": 4931, + "time_per_iteration": 2.6152310371398926 + }, + { + "auxiliary_loss_clip": 0.0115478, + "auxiliary_loss_mlp": 0.01027957, + "balance_loss_clip": 1.0189389, + "balance_loss_mlp": 1.02063775, + "epoch": 0.5930379366319967, + "flos": 17601472028160.0, + "grad_norm": 1.936139221814039, + "language_loss": 0.72303116, + "learning_rate": 1.4998140131079553e-06, + "loss": 0.7448585, + "num_input_tokens_seen": 106345565, + "step": 4932, + "time_per_iteration": 2.715160608291626 + }, + { + "auxiliary_loss_clip": 0.01059475, + "auxiliary_loss_mlp": 0.00509329, + "balance_loss_clip": 0.86217451, + "balance_loss_mlp": 0.0, + "epoch": 0.5931581795226357, + "flos": 17704283731200.0, + "grad_norm": 2.0473983308913315, + "language_loss": 0.73303831, + "learning_rate": 1.4990598337190821e-06, + "loss": 0.74872637, + "num_input_tokens_seen": 106361920, + "step": 4933, + "time_per_iteration": 2.784611463546753 + }, + { + "auxiliary_loss_clip": 0.0117503, + "auxiliary_loss_mlp": 0.00509751, + "balance_loss_clip": 1.05304766, + "balance_loss_mlp": 0.0, + "epoch": 0.5932784224132748, + "flos": 24280102483200.0, + "grad_norm": 1.688817971343, + "language_loss": 0.68289447, + "learning_rate": 1.4983057303224338e-06, + "loss": 0.6997422, + "num_input_tokens_seen": 106381735, + "step": 4934, + "time_per_iteration": 2.6430001258850098 + }, + { + "auxiliary_loss_clip": 0.01079653, + "auxiliary_loss_mlp": 0.01034647, + "balance_loss_clip": 0.89863622, + "balance_loss_mlp": 1.02683854, + "epoch": 0.5933986653039139, + "flos": 22926700909440.0, + "grad_norm": 1.667738473115423, + "language_loss": 0.87283909, + "learning_rate": 1.4975517030324072e-06, + "loss": 0.89398217, + "num_input_tokens_seen": 106399745, + "step": 4935, + "time_per_iteration": 2.7164382934570312 + }, + { + "auxiliary_loss_clip": 0.01070294, + "auxiliary_loss_mlp": 0.00500898, + "balance_loss_clip": 1.01529515, + "balance_loss_mlp": 0.0, + "epoch": 0.593518908194553, + "flos": 71121730256640.0, + "grad_norm": 0.7822928124351619, + "language_loss": 0.61824512, + "learning_rate": 1.4967977519633882e-06, + "loss": 0.63395703, + "num_input_tokens_seen": 106457205, + "step": 4936, + "time_per_iteration": 3.315446376800537 + }, + { + "auxiliary_loss_clip": 0.01101512, + "auxiliary_loss_mlp": 0.01028113, + "balance_loss_clip": 0.9369117, + "balance_loss_mlp": 1.02023363, + "epoch": 0.593639151085192, + "flos": 20448649526400.0, + "grad_norm": 1.9741685920323622, + "language_loss": 0.78458881, + "learning_rate": 1.4960438772297494e-06, + "loss": 0.80588508, + "num_input_tokens_seen": 106474250, + "step": 4937, + "time_per_iteration": 2.661057710647583 + }, + { + "auxiliary_loss_clip": 0.01130004, + "auxiliary_loss_mlp": 0.01032881, + "balance_loss_clip": 0.97510904, + "balance_loss_mlp": 1.02472699, + "epoch": 0.5937593939758312, + "flos": 30883428074880.0, + "grad_norm": 2.204683039202731, + "language_loss": 0.7408917, + "learning_rate": 1.495290078945855e-06, + "loss": 0.76252055, + "num_input_tokens_seen": 106494015, + "step": 4938, + "time_per_iteration": 2.7512094974517822 + }, + { + "auxiliary_loss_clip": 0.01175483, + "auxiliary_loss_mlp": 0.0102953, + "balance_loss_clip": 1.05443621, + "balance_loss_mlp": 1.02188277, + "epoch": 0.5938796368664703, + "flos": 36898069668480.0, + "grad_norm": 2.0169506698146273, + "language_loss": 0.74300838, + "learning_rate": 1.4945363572260529e-06, + "loss": 0.76505852, + "num_input_tokens_seen": 106515010, + "step": 4939, + "time_per_iteration": 3.576112747192383 + }, + { + "auxiliary_loss_clip": 0.01152108, + "auxiliary_loss_mlp": 0.0102999, + "balance_loss_clip": 1.01451421, + "balance_loss_mlp": 1.02208042, + "epoch": 0.5939998797571093, + "flos": 23842926051840.0, + "grad_norm": 2.7015352993400144, + "language_loss": 0.67864048, + "learning_rate": 1.4937827121846845e-06, + "loss": 0.70046151, + "num_input_tokens_seen": 106535265, + "step": 4940, + "time_per_iteration": 2.6612658500671387 + }, + { + "auxiliary_loss_clip": 0.01102011, + "auxiliary_loss_mlp": 0.01027636, + "balance_loss_clip": 0.94166797, + "balance_loss_mlp": 1.02057004, + "epoch": 0.5941201226477485, + "flos": 25191407462400.0, + "grad_norm": 1.485171439809512, + "language_loss": 0.7337091, + "learning_rate": 1.4930291439360755e-06, + "loss": 0.7550056, + "num_input_tokens_seen": 106557830, + "step": 4941, + "time_per_iteration": 2.7695515155792236 + }, + { + "auxiliary_loss_clip": 0.01155111, + "auxiliary_loss_mlp": 0.01028467, + "balance_loss_clip": 1.01813364, + "balance_loss_mlp": 1.02011645, + "epoch": 0.5942403655383875, + "flos": 22418996123520.0, + "grad_norm": 1.876562511746902, + "language_loss": 0.79218352, + "learning_rate": 1.4922756525945427e-06, + "loss": 0.81401932, + "num_input_tokens_seen": 106577140, + "step": 4942, + "time_per_iteration": 2.6218316555023193 + }, + { + "auxiliary_loss_clip": 0.01051595, + "auxiliary_loss_mlp": 0.01003442, + "balance_loss_clip": 0.97767907, + "balance_loss_mlp": 1.00232732, + "epoch": 0.5943606084290266, + "flos": 67629310796160.0, + "grad_norm": 0.7824042488797461, + "language_loss": 0.59642386, + "learning_rate": 1.4915222382743894e-06, + "loss": 0.61697423, + "num_input_tokens_seen": 106635975, + "step": 4943, + "time_per_iteration": 3.218879461288452 + }, + { + "auxiliary_loss_clip": 0.01155498, + "auxiliary_loss_mlp": 0.01030167, + "balance_loss_clip": 1.01849437, + "balance_loss_mlp": 1.0226599, + "epoch": 0.5944808513196658, + "flos": 18223157646720.0, + "grad_norm": 2.847170443530646, + "language_loss": 0.72248578, + "learning_rate": 1.4907689010899085e-06, + "loss": 0.74434245, + "num_input_tokens_seen": 106653555, + "step": 4944, + "time_per_iteration": 2.6922075748443604 + }, + { + "auxiliary_loss_clip": 0.01130537, + "auxiliary_loss_mlp": 0.0103114, + "balance_loss_clip": 0.97797906, + "balance_loss_mlp": 1.02337313, + "epoch": 0.5946010942103048, + "flos": 24790824011520.0, + "grad_norm": 2.1504757705537396, + "language_loss": 0.62766302, + "learning_rate": 1.4900156411553804e-06, + "loss": 0.64927977, + "num_input_tokens_seen": 106673385, + "step": 4945, + "time_per_iteration": 2.679765224456787 + }, + { + "auxiliary_loss_clip": 0.01131777, + "auxiliary_loss_mlp": 0.01030335, + "balance_loss_clip": 0.98025131, + "balance_loss_mlp": 1.02264559, + "epoch": 0.5947213371009439, + "flos": 15231619388160.0, + "grad_norm": 2.2408689609888053, + "language_loss": 0.85627401, + "learning_rate": 1.4892624585850739e-06, + "loss": 0.87789512, + "num_input_tokens_seen": 106691740, + "step": 4946, + "time_per_iteration": 2.646544933319092 + }, + { + "auxiliary_loss_clip": 0.01179462, + "auxiliary_loss_mlp": 0.01027659, + "balance_loss_clip": 1.055179, + "balance_loss_mlp": 1.02021778, + "epoch": 0.594841579991583, + "flos": 25848069949440.0, + "grad_norm": 1.8717399559640044, + "language_loss": 0.79565966, + "learning_rate": 1.4885093534932465e-06, + "loss": 0.8177309, + "num_input_tokens_seen": 106709705, + "step": 4947, + "time_per_iteration": 2.6133556365966797 + }, + { + "auxiliary_loss_clip": 0.01130182, + "auxiliary_loss_mlp": 0.01027931, + "balance_loss_clip": 0.98024446, + "balance_loss_mlp": 1.01996803, + "epoch": 0.5949618228822221, + "flos": 23981109672960.0, + "grad_norm": 2.060202171837331, + "language_loss": 0.71237952, + "learning_rate": 1.4877563259941433e-06, + "loss": 0.73396063, + "num_input_tokens_seen": 106727560, + "step": 4948, + "time_per_iteration": 2.7256007194519043 + }, + { + "auxiliary_loss_clip": 0.01161223, + "auxiliary_loss_mlp": 0.01027467, + "balance_loss_clip": 1.01797378, + "balance_loss_mlp": 1.01976013, + "epoch": 0.5950820657728612, + "flos": 40547491476480.0, + "grad_norm": 3.36864620037362, + "language_loss": 0.67771715, + "learning_rate": 1.4870033762019988e-06, + "loss": 0.69960403, + "num_input_tokens_seen": 106747725, + "step": 4949, + "time_per_iteration": 2.8360745906829834 + }, + { + "auxiliary_loss_clip": 0.01126975, + "auxiliary_loss_mlp": 0.01031801, + "balance_loss_clip": 0.97666359, + "balance_loss_mlp": 1.02416909, + "epoch": 0.5952023086635003, + "flos": 23184467884800.0, + "grad_norm": 1.6323988167296164, + "language_loss": 0.73289198, + "learning_rate": 1.4862505042310334e-06, + "loss": 0.75447965, + "num_input_tokens_seen": 106767010, + "step": 4950, + "time_per_iteration": 2.662864923477173 + }, + { + "auxiliary_loss_clip": 0.01121473, + "auxiliary_loss_mlp": 0.01025443, + "balance_loss_clip": 0.97623348, + "balance_loss_mlp": 1.01844239, + "epoch": 0.5953225515541394, + "flos": 33653289548160.0, + "grad_norm": 1.702125020857614, + "language_loss": 0.69699633, + "learning_rate": 1.4854977101954587e-06, + "loss": 0.71846557, + "num_input_tokens_seen": 106789230, + "step": 4951, + "time_per_iteration": 2.7305428981781006 + }, + { + "auxiliary_loss_clip": 0.01154391, + "auxiliary_loss_mlp": 0.01028819, + "balance_loss_clip": 1.01419413, + "balance_loss_mlp": 1.02079666, + "epoch": 0.5954427944447784, + "flos": 24459619680000.0, + "grad_norm": 3.0850321115545962, + "language_loss": 0.86294103, + "learning_rate": 1.4847449942094716e-06, + "loss": 0.88477314, + "num_input_tokens_seen": 106808110, + "step": 4952, + "time_per_iteration": 3.5983502864837646 + }, + { + "auxiliary_loss_clip": 0.01125276, + "auxiliary_loss_mlp": 0.01026528, + "balance_loss_clip": 0.978136, + "balance_loss_mlp": 1.01869035, + "epoch": 0.5955630373354175, + "flos": 18551848026240.0, + "grad_norm": 1.9999593870001429, + "language_loss": 0.86476469, + "learning_rate": 1.4839923563872598e-06, + "loss": 0.88628274, + "num_input_tokens_seen": 106826650, + "step": 4953, + "time_per_iteration": 3.5868892669677734 + }, + { + "auxiliary_loss_clip": 0.01107657, + "auxiliary_loss_mlp": 0.01030966, + "balance_loss_clip": 0.9421432, + "balance_loss_mlp": 1.0227766, + "epoch": 0.5956832802260567, + "flos": 19791699730560.0, + "grad_norm": 1.8970454754447854, + "language_loss": 0.76083553, + "learning_rate": 1.483239796842997e-06, + "loss": 0.78222179, + "num_input_tokens_seen": 106844680, + "step": 4954, + "time_per_iteration": 3.6389000415802 + }, + { + "auxiliary_loss_clip": 0.01106547, + "auxiliary_loss_mlp": 0.01024512, + "balance_loss_clip": 0.94055033, + "balance_loss_mlp": 1.01702559, + "epoch": 0.5958035231166957, + "flos": 19750868945280.0, + "grad_norm": 2.081496077246437, + "language_loss": 0.83509564, + "learning_rate": 1.4824873156908462e-06, + "loss": 0.85640621, + "num_input_tokens_seen": 106862605, + "step": 4955, + "time_per_iteration": 2.669394016265869 + }, + { + "auxiliary_loss_clip": 0.01154408, + "auxiliary_loss_mlp": 0.00510444, + "balance_loss_clip": 1.01729453, + "balance_loss_mlp": 0.0, + "epoch": 0.5959237660073348, + "flos": 21652806090240.0, + "grad_norm": 1.8991335988030054, + "language_loss": 0.75619745, + "learning_rate": 1.4817349130449584e-06, + "loss": 0.77284598, + "num_input_tokens_seen": 106882325, + "step": 4956, + "time_per_iteration": 2.6968045234680176 + }, + { + "auxiliary_loss_clip": 0.0114849, + "auxiliary_loss_mlp": 0.01031087, + "balance_loss_clip": 1.01494551, + "balance_loss_mlp": 1.02355909, + "epoch": 0.5960440088979739, + "flos": 21171207513600.0, + "grad_norm": 2.67480822310119, + "language_loss": 0.83071411, + "learning_rate": 1.4809825890194717e-06, + "loss": 0.85250986, + "num_input_tokens_seen": 106900995, + "step": 4957, + "time_per_iteration": 2.6357483863830566 + }, + { + "auxiliary_loss_clip": 0.0112406, + "auxiliary_loss_mlp": 0.01029307, + "balance_loss_clip": 0.97496188, + "balance_loss_mlp": 1.02189207, + "epoch": 0.596164251788613, + "flos": 14757526753920.0, + "grad_norm": 1.7739802825166275, + "language_loss": 0.76978898, + "learning_rate": 1.4802303437285139e-06, + "loss": 0.79132271, + "num_input_tokens_seen": 106918265, + "step": 4958, + "time_per_iteration": 2.6517083644866943 + }, + { + "auxiliary_loss_clip": 0.011282, + "auxiliary_loss_mlp": 0.01027858, + "balance_loss_clip": 0.97624689, + "balance_loss_mlp": 1.02007985, + "epoch": 0.596284494679252, + "flos": 20485924865280.0, + "grad_norm": 2.615744382278025, + "language_loss": 0.80373615, + "learning_rate": 1.4794781772861994e-06, + "loss": 0.82529676, + "num_input_tokens_seen": 106934760, + "step": 4959, + "time_per_iteration": 2.612807512283325 + }, + { + "auxiliary_loss_clip": 0.01128115, + "auxiliary_loss_mlp": 0.00509595, + "balance_loss_clip": 0.97724479, + "balance_loss_mlp": 0.0, + "epoch": 0.5964047375698912, + "flos": 31212262108800.0, + "grad_norm": 2.6932770170247866, + "language_loss": 0.66977394, + "learning_rate": 1.4787260898066324e-06, + "loss": 0.68615103, + "num_input_tokens_seen": 106954760, + "step": 4960, + "time_per_iteration": 2.761488199234009 + }, + { + "auxiliary_loss_clip": 0.0117268, + "auxiliary_loss_mlp": 0.01024434, + "balance_loss_clip": 1.05355144, + "balance_loss_mlp": 1.01707911, + "epoch": 0.5965249804605303, + "flos": 27483620855040.0, + "grad_norm": 2.262469022597532, + "language_loss": 0.85964149, + "learning_rate": 1.4779740814039023e-06, + "loss": 0.88161266, + "num_input_tokens_seen": 106974845, + "step": 4961, + "time_per_iteration": 2.6128902435302734 + }, + { + "auxiliary_loss_clip": 0.01175589, + "auxiliary_loss_mlp": 0.01026082, + "balance_loss_clip": 1.05348206, + "balance_loss_mlp": 1.01852131, + "epoch": 0.5966452233511693, + "flos": 30773936442240.0, + "grad_norm": 1.9141930871974049, + "language_loss": 0.68579292, + "learning_rate": 1.4772221521920894e-06, + "loss": 0.70780969, + "num_input_tokens_seen": 106994870, + "step": 4962, + "time_per_iteration": 2.6676905155181885 + }, + { + "auxiliary_loss_clip": 0.01128748, + "auxiliary_loss_mlp": 0.0103303, + "balance_loss_clip": 0.97929502, + "balance_loss_mlp": 1.02523386, + "epoch": 0.5967654662418085, + "flos": 25481170477440.0, + "grad_norm": 1.9059412787406413, + "language_loss": 0.74041629, + "learning_rate": 1.4764703022852598e-06, + "loss": 0.76203406, + "num_input_tokens_seen": 107015390, + "step": 4963, + "time_per_iteration": 2.7004077434539795 + }, + { + "auxiliary_loss_clip": 0.01045288, + "auxiliary_loss_mlp": 0.01028378, + "balance_loss_clip": 0.85941702, + "balance_loss_mlp": 1.02124023, + "epoch": 0.5968857091324475, + "flos": 19099126621440.0, + "grad_norm": 1.6884065319152017, + "language_loss": 0.76850677, + "learning_rate": 1.4757185317974696e-06, + "loss": 0.78924346, + "num_input_tokens_seen": 107033775, + "step": 4964, + "time_per_iteration": 2.7610387802124023 + }, + { + "auxiliary_loss_clip": 0.01152068, + "auxiliary_loss_mlp": 0.01026063, + "balance_loss_clip": 1.01426935, + "balance_loss_mlp": 1.01808834, + "epoch": 0.5970059520230866, + "flos": 23692711374720.0, + "grad_norm": 2.507479468404179, + "language_loss": 0.71168453, + "learning_rate": 1.474966840842761e-06, + "loss": 0.73346579, + "num_input_tokens_seen": 107053355, + "step": 4965, + "time_per_iteration": 3.581355094909668 + }, + { + "auxiliary_loss_clip": 0.01153937, + "auxiliary_loss_mlp": 0.01026425, + "balance_loss_clip": 1.01509154, + "balance_loss_mlp": 1.01934433, + "epoch": 0.5971261949137258, + "flos": 23185545292800.0, + "grad_norm": 1.7654811317149879, + "language_loss": 0.8688069, + "learning_rate": 1.4742152295351655e-06, + "loss": 0.89061052, + "num_input_tokens_seen": 107072510, + "step": 4966, + "time_per_iteration": 2.6457247734069824 + }, + { + "auxiliary_loss_clip": 0.01152599, + "auxiliary_loss_mlp": 0.00509902, + "balance_loss_clip": 1.01473141, + "balance_loss_mlp": 0.0, + "epoch": 0.5972464378043648, + "flos": 20557710195840.0, + "grad_norm": 18.27506060254093, + "language_loss": 0.63652623, + "learning_rate": 1.4734636979887016e-06, + "loss": 0.65315121, + "num_input_tokens_seen": 107089970, + "step": 4967, + "time_per_iteration": 2.671915054321289 + }, + { + "auxiliary_loss_clip": 0.01108226, + "auxiliary_loss_mlp": 0.01028272, + "balance_loss_clip": 0.93685746, + "balance_loss_mlp": 1.02024353, + "epoch": 0.5973666806950039, + "flos": 29387030457600.0, + "grad_norm": 2.0701385688283853, + "language_loss": 0.90151471, + "learning_rate": 1.4727122463173755e-06, + "loss": 0.9228797, + "num_input_tokens_seen": 107108500, + "step": 4968, + "time_per_iteration": 2.7528793811798096 + }, + { + "auxiliary_loss_clip": 0.01125874, + "auxiliary_loss_mlp": 0.01024038, + "balance_loss_clip": 0.97673523, + "balance_loss_mlp": 1.01680529, + "epoch": 0.597486923585643, + "flos": 22273522041600.0, + "grad_norm": 1.7545212758123223, + "language_loss": 0.64194399, + "learning_rate": 1.471960874635183e-06, + "loss": 0.66344309, + "num_input_tokens_seen": 107128060, + "step": 4969, + "time_per_iteration": 2.6965458393096924 + }, + { + "auxiliary_loss_clip": 0.01122474, + "auxiliary_loss_mlp": 0.01029073, + "balance_loss_clip": 0.97472131, + "balance_loss_mlp": 1.02055526, + "epoch": 0.5976071664762821, + "flos": 13772461196160.0, + "grad_norm": 2.8097402374455043, + "language_loss": 0.7112416, + "learning_rate": 1.4712095830561055e-06, + "loss": 0.73275709, + "num_input_tokens_seen": 107146550, + "step": 4970, + "time_per_iteration": 2.6618895530700684 + }, + { + "auxiliary_loss_clip": 0.01126176, + "auxiliary_loss_mlp": 0.01028435, + "balance_loss_clip": 0.97360307, + "balance_loss_mlp": 1.02092147, + "epoch": 0.5977274093669211, + "flos": 19098623831040.0, + "grad_norm": 3.00090067188707, + "language_loss": 0.81094742, + "learning_rate": 1.4704583716941147e-06, + "loss": 0.83249354, + "num_input_tokens_seen": 107165415, + "step": 4971, + "time_per_iteration": 2.6956875324249268 + }, + { + "auxiliary_loss_clip": 0.0114525, + "auxiliary_loss_mlp": 0.01027167, + "balance_loss_clip": 1.01549625, + "balance_loss_mlp": 1.01946628, + "epoch": 0.5978476522575603, + "flos": 20376002269440.0, + "grad_norm": 2.1126943317600952, + "language_loss": 0.72734004, + "learning_rate": 1.4697072406631672e-06, + "loss": 0.74906421, + "num_input_tokens_seen": 107185320, + "step": 4972, + "time_per_iteration": 2.6559903621673584 + }, + { + "auxiliary_loss_clip": 0.01088111, + "auxiliary_loss_mlp": 0.01028017, + "balance_loss_clip": 0.9034934, + "balance_loss_mlp": 1.02067971, + "epoch": 0.5979678951481994, + "flos": 29023147728000.0, + "grad_norm": 1.5873881965915049, + "language_loss": 0.72586393, + "learning_rate": 1.4689561900772097e-06, + "loss": 0.74702513, + "num_input_tokens_seen": 107205380, + "step": 4973, + "time_per_iteration": 2.8393545150756836 + }, + { + "auxiliary_loss_clip": 0.01126758, + "auxiliary_loss_mlp": 0.01023616, + "balance_loss_clip": 0.97658437, + "balance_loss_mlp": 1.0159688, + "epoch": 0.5980881380388384, + "flos": 17967689141760.0, + "grad_norm": 2.302505650200586, + "language_loss": 0.72261035, + "learning_rate": 1.4682052200501758e-06, + "loss": 0.74411416, + "num_input_tokens_seen": 107222585, + "step": 4974, + "time_per_iteration": 2.645613431930542 + }, + { + "auxiliary_loss_clip": 0.01176727, + "auxiliary_loss_mlp": 0.01025184, + "balance_loss_clip": 1.05565286, + "balance_loss_mlp": 1.01739979, + "epoch": 0.5982083809294776, + "flos": 22962827013120.0, + "grad_norm": 1.7178398697193222, + "language_loss": 0.79824895, + "learning_rate": 1.4674543306959876e-06, + "loss": 0.82026809, + "num_input_tokens_seen": 107242055, + "step": 4975, + "time_per_iteration": 2.6406807899475098 + }, + { + "auxiliary_loss_clip": 0.01129154, + "auxiliary_loss_mlp": 0.01033314, + "balance_loss_clip": 0.97583389, + "balance_loss_mlp": 1.0252738, + "epoch": 0.5983286238201166, + "flos": 20991941712000.0, + "grad_norm": 3.859588015281101, + "language_loss": 0.84207141, + "learning_rate": 1.4667035221285535e-06, + "loss": 0.86369616, + "num_input_tokens_seen": 107259695, + "step": 4976, + "time_per_iteration": 2.6880359649658203 + }, + { + "auxiliary_loss_clip": 0.01153119, + "auxiliary_loss_mlp": 0.01025498, + "balance_loss_clip": 1.01888335, + "balance_loss_mlp": 1.01809168, + "epoch": 0.5984488667107557, + "flos": 28183448511360.0, + "grad_norm": 2.6645764515665626, + "language_loss": 0.74178398, + "learning_rate": 1.4659527944617715e-06, + "loss": 0.76357013, + "num_input_tokens_seen": 107279640, + "step": 4977, + "time_per_iteration": 2.667933464050293 + }, + { + "auxiliary_loss_clip": 0.0105371, + "auxiliary_loss_mlp": 0.01032639, + "balance_loss_clip": 0.86006081, + "balance_loss_mlp": 1.0248313, + "epoch": 0.5985691096013949, + "flos": 16471794314880.0, + "grad_norm": 1.7678535248899152, + "language_loss": 0.76523864, + "learning_rate": 1.465202147809526e-06, + "loss": 0.78610218, + "num_input_tokens_seen": 107298135, + "step": 4978, + "time_per_iteration": 3.7256906032562256 + }, + { + "auxiliary_loss_clip": 0.0117519, + "auxiliary_loss_mlp": 0.01024823, + "balance_loss_clip": 1.05388212, + "balance_loss_mlp": 1.0175215, + "epoch": 0.5986893524920339, + "flos": 26719046933760.0, + "grad_norm": 2.4056277977101295, + "language_loss": 0.76145321, + "learning_rate": 1.4644515822856888e-06, + "loss": 0.7834534, + "num_input_tokens_seen": 107316570, + "step": 4979, + "time_per_iteration": 3.562943935394287 + }, + { + "auxiliary_loss_clip": 0.01010484, + "auxiliary_loss_mlp": 0.01001041, + "balance_loss_clip": 0.90389323, + "balance_loss_mlp": 0.99989027, + "epoch": 0.598809595382673, + "flos": 61608061100160.0, + "grad_norm": 0.7888275751759067, + "language_loss": 0.56560838, + "learning_rate": 1.4637010980041215e-06, + "loss": 0.58572364, + "num_input_tokens_seen": 107378680, + "step": 4980, + "time_per_iteration": 3.2897002696990967 + }, + { + "auxiliary_loss_clip": 0.01176322, + "auxiliary_loss_mlp": 0.01028971, + "balance_loss_clip": 1.05451047, + "balance_loss_mlp": 1.02116239, + "epoch": 0.5989298382733121, + "flos": 11801719549440.0, + "grad_norm": 2.3066107131583338, + "language_loss": 0.89769769, + "learning_rate": 1.4629506950786707e-06, + "loss": 0.91975069, + "num_input_tokens_seen": 107394860, + "step": 4981, + "time_per_iteration": 3.526718854904175 + }, + { + "auxiliary_loss_clip": 0.0106864, + "auxiliary_loss_mlp": 0.00999305, + "balance_loss_clip": 1.01384544, + "balance_loss_mlp": 0.99820232, + "epoch": 0.5990500811639512, + "flos": 60025800021120.0, + "grad_norm": 0.8069760801180371, + "language_loss": 0.56103063, + "learning_rate": 1.4622003736231733e-06, + "loss": 0.5817101, + "num_input_tokens_seen": 107453850, + "step": 4982, + "time_per_iteration": 3.2264909744262695 + }, + { + "auxiliary_loss_clip": 0.01152143, + "auxiliary_loss_mlp": 0.01025544, + "balance_loss_clip": 1.01719308, + "balance_loss_mlp": 1.01770008, + "epoch": 0.5991703240545903, + "flos": 18222726683520.0, + "grad_norm": 1.988288435983495, + "language_loss": 0.80672562, + "learning_rate": 1.461450133751451e-06, + "loss": 0.82850254, + "num_input_tokens_seen": 107471920, + "step": 4983, + "time_per_iteration": 2.6796295642852783 + }, + { + "auxiliary_loss_clip": 0.01155057, + "auxiliary_loss_mlp": 0.01025766, + "balance_loss_clip": 1.01566672, + "balance_loss_mlp": 1.0182023, + "epoch": 0.5992905669452293, + "flos": 27709894581120.0, + "grad_norm": 1.7453695724782317, + "language_loss": 0.76126778, + "learning_rate": 1.4606999755773153e-06, + "loss": 0.78307599, + "num_input_tokens_seen": 107493125, + "step": 4984, + "time_per_iteration": 2.721437931060791 + }, + { + "auxiliary_loss_clip": 0.01175609, + "auxiliary_loss_mlp": 0.01022546, + "balance_loss_clip": 1.05492461, + "balance_loss_mlp": 1.01504743, + "epoch": 0.5994108098358685, + "flos": 20449008662400.0, + "grad_norm": 1.655036994581289, + "language_loss": 0.8215096, + "learning_rate": 1.4599498992145643e-06, + "loss": 0.84349114, + "num_input_tokens_seen": 107513150, + "step": 4985, + "time_per_iteration": 2.570384979248047 + }, + { + "auxiliary_loss_clip": 0.01133217, + "auxiliary_loss_mlp": 0.00508975, + "balance_loss_clip": 0.97812223, + "balance_loss_mlp": 0.0, + "epoch": 0.5995310527265075, + "flos": 22269966595200.0, + "grad_norm": 2.0744757186852922, + "language_loss": 0.70775592, + "learning_rate": 1.4591999047769846e-06, + "loss": 0.7241779, + "num_input_tokens_seen": 107532005, + "step": 4986, + "time_per_iteration": 2.6787109375 + }, + { + "auxiliary_loss_clip": 0.01052673, + "auxiliary_loss_mlp": 0.01028438, + "balance_loss_clip": 0.85816747, + "balance_loss_mlp": 1.02154136, + "epoch": 0.5996512956171466, + "flos": 18916951818240.0, + "grad_norm": 1.8592446517645498, + "language_loss": 0.75258899, + "learning_rate": 1.4584499923783486e-06, + "loss": 0.77340007, + "num_input_tokens_seen": 107550585, + "step": 4987, + "time_per_iteration": 2.7308027744293213 + }, + { + "auxiliary_loss_clip": 0.01127019, + "auxiliary_loss_mlp": 0.01028106, + "balance_loss_clip": 0.97629011, + "balance_loss_mlp": 1.02106118, + "epoch": 0.5997715385077858, + "flos": 15370916330880.0, + "grad_norm": 1.7730245991819413, + "language_loss": 0.76272339, + "learning_rate": 1.457700162132419e-06, + "loss": 0.78427464, + "num_input_tokens_seen": 107567575, + "step": 4988, + "time_per_iteration": 2.7248599529266357 + }, + { + "auxiliary_loss_clip": 0.01079653, + "auxiliary_loss_mlp": 0.01022243, + "balance_loss_clip": 0.90127993, + "balance_loss_mlp": 1.01481009, + "epoch": 0.5998917813984248, + "flos": 25264844818560.0, + "grad_norm": 2.0596150083966367, + "language_loss": 0.72728705, + "learning_rate": 1.4569504141529433e-06, + "loss": 0.74830604, + "num_input_tokens_seen": 107585410, + "step": 4989, + "time_per_iteration": 2.776432991027832 + }, + { + "auxiliary_loss_clip": 0.01153323, + "auxiliary_loss_mlp": 0.01025928, + "balance_loss_clip": 1.0190506, + "balance_loss_mlp": 1.0181911, + "epoch": 0.6000120242890639, + "flos": 22054502862720.0, + "grad_norm": 2.217830723846243, + "language_loss": 0.72199684, + "learning_rate": 1.456200748553658e-06, + "loss": 0.74378937, + "num_input_tokens_seen": 107603405, + "step": 4990, + "time_per_iteration": 2.6323869228363037 + }, + { + "auxiliary_loss_clip": 0.01178, + "auxiliary_loss_mlp": 0.01028112, + "balance_loss_clip": 1.05629611, + "balance_loss_mlp": 1.02029228, + "epoch": 0.600132267179703, + "flos": 29863421562240.0, + "grad_norm": 1.5112941857441737, + "language_loss": 0.78796232, + "learning_rate": 1.455451165448287e-06, + "loss": 0.81002343, + "num_input_tokens_seen": 107626060, + "step": 4991, + "time_per_iteration": 3.591655731201172 + }, + { + "auxiliary_loss_clip": 0.01128892, + "auxiliary_loss_mlp": 0.0102955, + "balance_loss_clip": 0.97878772, + "balance_loss_mlp": 1.02143168, + "epoch": 0.6002525100703421, + "flos": 25045358762880.0, + "grad_norm": 6.278581466736429, + "language_loss": 0.73699093, + "learning_rate": 1.4547016649505407e-06, + "loss": 0.75857532, + "num_input_tokens_seen": 107644070, + "step": 4992, + "time_per_iteration": 2.6866366863250732 + }, + { + "auxiliary_loss_clip": 0.01100604, + "auxiliary_loss_mlp": 0.01025633, + "balance_loss_clip": 0.93411416, + "balance_loss_mlp": 1.01802778, + "epoch": 0.6003727529609811, + "flos": 20849592113280.0, + "grad_norm": 2.1031950461106126, + "language_loss": 0.8534596, + "learning_rate": 1.4539522471741193e-06, + "loss": 0.874722, + "num_input_tokens_seen": 107661495, + "step": 4993, + "time_per_iteration": 243.37141823768616 + }, + { + "auxiliary_loss_clip": 0.01153338, + "auxiliary_loss_mlp": 0.01031681, + "balance_loss_clip": 1.01359653, + "balance_loss_mlp": 1.02339578, + "epoch": 0.6004929958516203, + "flos": 15594604277760.0, + "grad_norm": 1.9823463317603287, + "language_loss": 0.71003258, + "learning_rate": 1.4532029122327067e-06, + "loss": 0.73188275, + "num_input_tokens_seen": 107678280, + "step": 4994, + "time_per_iteration": 2.6890065670013428 + }, + { + "auxiliary_loss_clip": 0.01098388, + "auxiliary_loss_mlp": 0.01024385, + "balance_loss_clip": 0.94175518, + "balance_loss_mlp": 1.01655912, + "epoch": 0.6006132387422594, + "flos": 21763267390080.0, + "grad_norm": 9.238564659696387, + "language_loss": 0.75465268, + "learning_rate": 1.4524536602399783e-06, + "loss": 0.77588046, + "num_input_tokens_seen": 107697370, + "step": 4995, + "time_per_iteration": 2.868032932281494 + }, + { + "auxiliary_loss_clip": 0.01126669, + "auxiliary_loss_mlp": 0.01025056, + "balance_loss_clip": 0.9793337, + "balance_loss_mlp": 1.01732826, + "epoch": 0.6007334816328984, + "flos": 22858542852480.0, + "grad_norm": 1.6453820819086056, + "language_loss": 0.77713746, + "learning_rate": 1.4517044913095938e-06, + "loss": 0.79865468, + "num_input_tokens_seen": 107717790, + "step": 4996, + "time_per_iteration": 2.77937912940979 + }, + { + "auxiliary_loss_clip": 0.01152457, + "auxiliary_loss_mlp": 0.01034754, + "balance_loss_clip": 1.01786554, + "balance_loss_mlp": 1.02686834, + "epoch": 0.6008537245235376, + "flos": 28324577047680.0, + "grad_norm": 1.6526488787193432, + "language_loss": 0.81325424, + "learning_rate": 1.4509554055552022e-06, + "loss": 0.83512634, + "num_input_tokens_seen": 107738020, + "step": 4997, + "time_per_iteration": 2.7401576042175293 + }, + { + "auxiliary_loss_clip": 0.01128642, + "auxiliary_loss_mlp": 0.01033054, + "balance_loss_clip": 0.97776556, + "balance_loss_mlp": 1.02539527, + "epoch": 0.6009739674141766, + "flos": 20886113266560.0, + "grad_norm": 2.466074131219762, + "language_loss": 0.8457877, + "learning_rate": 1.450206403090439e-06, + "loss": 0.86740464, + "num_input_tokens_seen": 107756215, + "step": 4998, + "time_per_iteration": 2.674470901489258 + }, + { + "auxiliary_loss_clip": 0.01149372, + "auxiliary_loss_mlp": 0.01030098, + "balance_loss_clip": 1.01658416, + "balance_loss_mlp": 1.02306509, + "epoch": 0.6010942103048157, + "flos": 20481004702080.0, + "grad_norm": 1.964919108513301, + "language_loss": 0.86230338, + "learning_rate": 1.4494574840289274e-06, + "loss": 0.88409805, + "num_input_tokens_seen": 107773330, + "step": 4999, + "time_per_iteration": 2.7018980979919434 + }, + { + "auxiliary_loss_clip": 0.01154066, + "auxiliary_loss_mlp": 0.01031795, + "balance_loss_clip": 1.01315975, + "balance_loss_mlp": 1.02393925, + "epoch": 0.6012144531954549, + "flos": 23805973935360.0, + "grad_norm": 1.630736311137905, + "language_loss": 0.73778749, + "learning_rate": 1.4487086484842782e-06, + "loss": 0.75964612, + "num_input_tokens_seen": 107791975, + "step": 5000, + "time_per_iteration": 2.607576847076416 + }, + { + "auxiliary_loss_clip": 0.01177034, + "auxiliary_loss_mlp": 0.01027242, + "balance_loss_clip": 1.05486143, + "balance_loss_mlp": 1.01939833, + "epoch": 0.6013346960860939, + "flos": 18988378012800.0, + "grad_norm": 2.1744053140180486, + "language_loss": 0.60769629, + "learning_rate": 1.4479598965700878e-06, + "loss": 0.62973905, + "num_input_tokens_seen": 107809240, + "step": 5001, + "time_per_iteration": 2.664257764816284 + }, + { + "auxiliary_loss_clip": 0.01102081, + "auxiliary_loss_mlp": 0.0102904, + "balance_loss_clip": 0.93735552, + "balance_loss_mlp": 1.02146411, + "epoch": 0.601454938976733, + "flos": 24025316336640.0, + "grad_norm": 2.4687498861739576, + "language_loss": 0.6900475, + "learning_rate": 1.4472112283999427e-06, + "loss": 0.71135867, + "num_input_tokens_seen": 107827895, + "step": 5002, + "time_per_iteration": 2.742532253265381 + }, + { + "auxiliary_loss_clip": 0.01151625, + "auxiliary_loss_mlp": 0.01028145, + "balance_loss_clip": 1.01926351, + "balance_loss_mlp": 1.02095985, + "epoch": 0.6015751818673721, + "flos": 26427129102720.0, + "grad_norm": 2.0266088076469675, + "language_loss": 0.68783176, + "learning_rate": 1.4464626440874143e-06, + "loss": 0.70962954, + "num_input_tokens_seen": 107847010, + "step": 5003, + "time_per_iteration": 2.730525255203247 + }, + { + "auxiliary_loss_clip": 0.01086483, + "auxiliary_loss_mlp": 0.01027387, + "balance_loss_clip": 0.89596617, + "balance_loss_mlp": 1.01970696, + "epoch": 0.6016954247580112, + "flos": 13115260005120.0, + "grad_norm": 2.200548713504762, + "language_loss": 0.74433827, + "learning_rate": 1.4457141437460636e-06, + "loss": 0.76547694, + "num_input_tokens_seen": 107864235, + "step": 5004, + "time_per_iteration": 4.5576252937316895 + }, + { + "auxiliary_loss_clip": 0.01134278, + "auxiliary_loss_mlp": 0.01029176, + "balance_loss_clip": 0.98088813, + "balance_loss_mlp": 1.02151382, + "epoch": 0.6018156676486502, + "flos": 23768447201280.0, + "grad_norm": 1.856161743623407, + "language_loss": 0.73338556, + "learning_rate": 1.444965727489436e-06, + "loss": 0.75502014, + "num_input_tokens_seen": 107883680, + "step": 5005, + "time_per_iteration": 2.7119526863098145 + }, + { + "auxiliary_loss_clip": 0.01101198, + "auxiliary_loss_mlp": 0.01027503, + "balance_loss_clip": 0.93488407, + "balance_loss_mlp": 1.01992762, + "epoch": 0.6019359105392894, + "flos": 26469360518400.0, + "grad_norm": 1.6953328857381451, + "language_loss": 0.63384008, + "learning_rate": 1.444217395431066e-06, + "loss": 0.65512711, + "num_input_tokens_seen": 107906220, + "step": 5006, + "time_per_iteration": 3.887378692626953 + }, + { + "auxiliary_loss_clip": 0.01007298, + "auxiliary_loss_mlp": 0.01000451, + "balance_loss_clip": 0.90501672, + "balance_loss_mlp": 0.99919367, + "epoch": 0.6020561534299285, + "flos": 69190849728000.0, + "grad_norm": 0.8204979011867629, + "language_loss": 0.55847043, + "learning_rate": 1.4434691476844755e-06, + "loss": 0.57854795, + "num_input_tokens_seen": 107967195, + "step": 5007, + "time_per_iteration": 3.233795166015625 + }, + { + "auxiliary_loss_clip": 0.01124369, + "auxiliary_loss_mlp": 0.01023153, + "balance_loss_clip": 0.97813964, + "balance_loss_mlp": 1.01594114, + "epoch": 0.6021763963205675, + "flos": 21835304115840.0, + "grad_norm": 1.9943093604745183, + "language_loss": 0.67110097, + "learning_rate": 1.4427209843631729e-06, + "loss": 0.69257617, + "num_input_tokens_seen": 107984245, + "step": 5008, + "time_per_iteration": 2.7080326080322266 + }, + { + "auxiliary_loss_clip": 0.01175159, + "auxiliary_loss_mlp": 0.00509759, + "balance_loss_clip": 1.0558033, + "balance_loss_mlp": 0.0, + "epoch": 0.6022966392112067, + "flos": 26578636669440.0, + "grad_norm": 1.8805484661973808, + "language_loss": 0.81014252, + "learning_rate": 1.4419729055806534e-06, + "loss": 0.82699168, + "num_input_tokens_seen": 108003680, + "step": 5009, + "time_per_iteration": 2.71075439453125 + }, + { + "auxiliary_loss_clip": 0.01126941, + "auxiliary_loss_mlp": 0.00508936, + "balance_loss_clip": 0.98084706, + "balance_loss_mlp": 0.0, + "epoch": 0.6024168821018457, + "flos": 20703722981760.0, + "grad_norm": 2.1548181305548284, + "language_loss": 0.82047671, + "learning_rate": 1.441224911450401e-06, + "loss": 0.8368355, + "num_input_tokens_seen": 108019635, + "step": 5010, + "time_per_iteration": 2.765463352203369 + }, + { + "auxiliary_loss_clip": 0.01158931, + "auxiliary_loss_mlp": 0.01029166, + "balance_loss_clip": 1.01971078, + "balance_loss_mlp": 1.02154231, + "epoch": 0.6025371249924848, + "flos": 24680973242880.0, + "grad_norm": 1.6064602742040002, + "language_loss": 0.82419413, + "learning_rate": 1.4404770020858851e-06, + "loss": 0.84607512, + "num_input_tokens_seen": 108039120, + "step": 5011, + "time_per_iteration": 2.721202850341797 + }, + { + "auxiliary_loss_clip": 0.01148091, + "auxiliary_loss_mlp": 0.01026457, + "balance_loss_clip": 1.01688993, + "balance_loss_mlp": 1.01922095, + "epoch": 0.602657367883124, + "flos": 25955801815680.0, + "grad_norm": 1.6763142489368719, + "language_loss": 0.86250901, + "learning_rate": 1.439729177600563e-06, + "loss": 0.88425452, + "num_input_tokens_seen": 108059615, + "step": 5012, + "time_per_iteration": 2.736422300338745 + }, + { + "auxiliary_loss_clip": 0.01150083, + "auxiliary_loss_mlp": 0.01035259, + "balance_loss_clip": 1.01709294, + "balance_loss_mlp": 1.02803767, + "epoch": 0.602777610773763, + "flos": 16690633925760.0, + "grad_norm": 2.139821185145077, + "language_loss": 0.7292639, + "learning_rate": 1.4389814381078793e-06, + "loss": 0.75111735, + "num_input_tokens_seen": 108078855, + "step": 5013, + "time_per_iteration": 2.6916534900665283 + }, + { + "auxiliary_loss_clip": 0.00988238, + "auxiliary_loss_mlp": 0.01027659, + "balance_loss_clip": 0.74707609, + "balance_loss_mlp": 1.02013063, + "epoch": 0.6028978536644021, + "flos": 13334243270400.0, + "grad_norm": 2.04603932515301, + "language_loss": 0.80124432, + "learning_rate": 1.438233783721265e-06, + "loss": 0.82140326, + "num_input_tokens_seen": 108095020, + "step": 5014, + "time_per_iteration": 3.1033682823181152 + }, + { + "auxiliary_loss_clip": 0.01129271, + "auxiliary_loss_mlp": 0.01023965, + "balance_loss_clip": 0.9847585, + "balance_loss_mlp": 1.01669288, + "epoch": 0.6030180965550412, + "flos": 19644825018240.0, + "grad_norm": 2.4025668712826715, + "language_loss": 0.77892935, + "learning_rate": 1.43748621455414e-06, + "loss": 0.80046177, + "num_input_tokens_seen": 108111455, + "step": 5015, + "time_per_iteration": 2.9659314155578613 + }, + { + "auxiliary_loss_clip": 0.011269, + "auxiliary_loss_mlp": 0.01030049, + "balance_loss_clip": 0.97885048, + "balance_loss_mlp": 1.02256215, + "epoch": 0.6031383394456803, + "flos": 14458390289280.0, + "grad_norm": 2.227762808001871, + "language_loss": 0.80511224, + "learning_rate": 1.4367387307199082e-06, + "loss": 0.82668173, + "num_input_tokens_seen": 108128305, + "step": 5016, + "time_per_iteration": 2.696359872817993 + }, + { + "auxiliary_loss_clip": 0.01145474, + "auxiliary_loss_mlp": 0.01031955, + "balance_loss_clip": 1.01314223, + "balance_loss_mlp": 1.02401543, + "epoch": 0.6032585823363193, + "flos": 13917791623680.0, + "grad_norm": 1.8433991281714275, + "language_loss": 0.82452583, + "learning_rate": 1.4359913323319632e-06, + "loss": 0.84630013, + "num_input_tokens_seen": 108145475, + "step": 5017, + "time_per_iteration": 3.968677043914795 + }, + { + "auxiliary_loss_clip": 0.01046329, + "auxiliary_loss_mlp": 0.01025723, + "balance_loss_clip": 0.85731006, + "balance_loss_mlp": 1.01792073, + "epoch": 0.6033788252269584, + "flos": 24353252530560.0, + "grad_norm": 1.847563308975477, + "language_loss": 0.77351463, + "learning_rate": 1.4352440195036847e-06, + "loss": 0.79423517, + "num_input_tokens_seen": 108165650, + "step": 5018, + "time_per_iteration": 2.937298536300659 + }, + { + "auxiliary_loss_clip": 0.01037611, + "auxiliary_loss_mlp": 0.01027538, + "balance_loss_clip": 0.81950575, + "balance_loss_mlp": 1.01983082, + "epoch": 0.6034990681175976, + "flos": 25521247077120.0, + "grad_norm": 1.9254389047249871, + "language_loss": 0.79738092, + "learning_rate": 1.4344967923484395e-06, + "loss": 0.81803244, + "num_input_tokens_seen": 108187620, + "step": 5019, + "time_per_iteration": 3.06571364402771 + }, + { + "auxiliary_loss_clip": 0.01148476, + "auxiliary_loss_mlp": 0.01026188, + "balance_loss_clip": 1.01539886, + "balance_loss_mlp": 1.0187552, + "epoch": 0.6036193110082366, + "flos": 25958387594880.0, + "grad_norm": 2.0493730750278107, + "language_loss": 0.72595906, + "learning_rate": 1.433749650979581e-06, + "loss": 0.74770576, + "num_input_tokens_seen": 108207605, + "step": 5020, + "time_per_iteration": 3.33573842048645 + }, + { + "auxiliary_loss_clip": 0.01109118, + "auxiliary_loss_mlp": 0.01027282, + "balance_loss_clip": 0.93813992, + "balance_loss_mlp": 1.01973641, + "epoch": 0.6037395538988757, + "flos": 25593427457280.0, + "grad_norm": 1.7347780956661774, + "language_loss": 0.6802007, + "learning_rate": 1.433002595510451e-06, + "loss": 0.70156467, + "num_input_tokens_seen": 108226385, + "step": 5021, + "time_per_iteration": 2.738788604736328 + }, + { + "auxiliary_loss_clip": 0.01127227, + "auxiliary_loss_mlp": 0.00509764, + "balance_loss_clip": 0.97693205, + "balance_loss_mlp": 0.0, + "epoch": 0.6038597967895148, + "flos": 17816253402240.0, + "grad_norm": 1.792693040998991, + "language_loss": 0.71802986, + "learning_rate": 1.4322556260543757e-06, + "loss": 0.73439974, + "num_input_tokens_seen": 108242960, + "step": 5022, + "time_per_iteration": 2.920330762863159 + }, + { + "auxiliary_loss_clip": 0.01012009, + "auxiliary_loss_mlp": 0.0100478, + "balance_loss_clip": 0.9037205, + "balance_loss_mlp": 1.00368965, + "epoch": 0.6039800396801539, + "flos": 65169213235200.0, + "grad_norm": 0.8926130365097207, + "language_loss": 0.62694407, + "learning_rate": 1.4315087427246703e-06, + "loss": 0.64711189, + "num_input_tokens_seen": 108296785, + "step": 5023, + "time_per_iteration": 3.2103922367095947 + }, + { + "auxiliary_loss_clip": 0.0106834, + "auxiliary_loss_mlp": 0.01002525, + "balance_loss_clip": 1.01370251, + "balance_loss_mlp": 1.00141048, + "epoch": 0.604100282570793, + "flos": 67386409073280.0, + "grad_norm": 0.8637333327635649, + "language_loss": 0.58436966, + "learning_rate": 1.4307619456346372e-06, + "loss": 0.60507834, + "num_input_tokens_seen": 108341090, + "step": 5024, + "time_per_iteration": 2.9681670665740967 + }, + { + "auxiliary_loss_clip": 0.01152843, + "auxiliary_loss_mlp": 0.01026177, + "balance_loss_clip": 1.01300263, + "balance_loss_mlp": 1.01914716, + "epoch": 0.6042205254614321, + "flos": 35297495631360.0, + "grad_norm": 1.822967958738663, + "language_loss": 0.74391174, + "learning_rate": 1.430015234897564e-06, + "loss": 0.76570189, + "num_input_tokens_seen": 108364370, + "step": 5025, + "time_per_iteration": 2.8245487213134766 + }, + { + "auxiliary_loss_clip": 0.01175271, + "auxiliary_loss_mlp": 0.00509191, + "balance_loss_clip": 1.05348897, + "balance_loss_mlp": 0.0, + "epoch": 0.6043407683520712, + "flos": 45658262206080.0, + "grad_norm": 3.052608243303084, + "language_loss": 0.66628933, + "learning_rate": 1.4292686106267274e-06, + "loss": 0.68313396, + "num_input_tokens_seen": 108387220, + "step": 5026, + "time_per_iteration": 2.881580114364624 + }, + { + "auxiliary_loss_clip": 0.01157516, + "auxiliary_loss_mlp": 0.01029722, + "balance_loss_clip": 1.0167737, + "balance_loss_mlp": 1.02248001, + "epoch": 0.6044610112427102, + "flos": 16180020138240.0, + "grad_norm": 1.712318824927689, + "language_loss": 0.77128232, + "learning_rate": 1.4285220729353876e-06, + "loss": 0.79315472, + "num_input_tokens_seen": 108405760, + "step": 5027, + "time_per_iteration": 2.717918634414673 + }, + { + "auxiliary_loss_clip": 0.01126238, + "auxiliary_loss_mlp": 0.01029157, + "balance_loss_clip": 0.97415912, + "balance_loss_mlp": 1.02183175, + "epoch": 0.6045812541333494, + "flos": 13804062186240.0, + "grad_norm": 2.4902401391474545, + "language_loss": 0.77365851, + "learning_rate": 1.4277756219367957e-06, + "loss": 0.79521251, + "num_input_tokens_seen": 108422785, + "step": 5028, + "time_per_iteration": 2.6845202445983887 + }, + { + "auxiliary_loss_clip": 0.01113347, + "auxiliary_loss_mlp": 0.01032122, + "balance_loss_clip": 0.94051898, + "balance_loss_mlp": 1.02378345, + "epoch": 0.6047014970239885, + "flos": 19975059682560.0, + "grad_norm": 2.3763439577914265, + "language_loss": 0.79988098, + "learning_rate": 1.4270292577441864e-06, + "loss": 0.82133567, + "num_input_tokens_seen": 108442290, + "step": 5029, + "time_per_iteration": 2.7821178436279297 + }, + { + "auxiliary_loss_clip": 0.01152355, + "auxiliary_loss_mlp": 0.01029289, + "balance_loss_clip": 1.01207614, + "balance_loss_mlp": 1.02129626, + "epoch": 0.6048217399146275, + "flos": 25337097025920.0, + "grad_norm": 1.673384129209823, + "language_loss": 0.71714431, + "learning_rate": 1.4262829804707836e-06, + "loss": 0.73896074, + "num_input_tokens_seen": 108464280, + "step": 5030, + "time_per_iteration": 4.688919305801392 + }, + { + "auxiliary_loss_clip": 0.01154223, + "auxiliary_loss_mlp": 0.01024876, + "balance_loss_clip": 1.01466274, + "balance_loss_mlp": 1.01690054, + "epoch": 0.6049419828052667, + "flos": 26030819370240.0, + "grad_norm": 1.4634581217042888, + "language_loss": 0.69939131, + "learning_rate": 1.4255367902297958e-06, + "loss": 0.72118223, + "num_input_tokens_seen": 108485610, + "step": 5031, + "time_per_iteration": 3.6029300689697266 + }, + { + "auxiliary_loss_clip": 0.01173306, + "auxiliary_loss_mlp": 0.01033282, + "balance_loss_clip": 1.05382061, + "balance_loss_mlp": 1.02574754, + "epoch": 0.6050622256959057, + "flos": 14648106948480.0, + "grad_norm": 2.1179035534994095, + "language_loss": 0.78941584, + "learning_rate": 1.4247906871344215e-06, + "loss": 0.81148177, + "num_input_tokens_seen": 108501005, + "step": 5032, + "time_per_iteration": 2.600135326385498 + }, + { + "auxiliary_loss_clip": 0.01121327, + "auxiliary_loss_mlp": 0.01024989, + "balance_loss_clip": 0.97276151, + "balance_loss_mlp": 1.01719272, + "epoch": 0.6051824685865448, + "flos": 23331450337920.0, + "grad_norm": 1.9279020745878577, + "language_loss": 0.74840468, + "learning_rate": 1.4240446712978415e-06, + "loss": 0.7698679, + "num_input_tokens_seen": 108519990, + "step": 5033, + "time_per_iteration": 2.688859701156616 + }, + { + "auxiliary_loss_clip": 0.01155272, + "auxiliary_loss_mlp": 0.0103026, + "balance_loss_clip": 1.01509976, + "balance_loss_mlp": 1.02310705, + "epoch": 0.605302711477184, + "flos": 27563307177600.0, + "grad_norm": 1.8050562874430085, + "language_loss": 0.74424165, + "learning_rate": 1.423298742833227e-06, + "loss": 0.76609695, + "num_input_tokens_seen": 108538650, + "step": 5034, + "time_per_iteration": 2.705775022506714 + }, + { + "auxiliary_loss_clip": 0.01111186, + "auxiliary_loss_mlp": 0.01031126, + "balance_loss_clip": 0.93788236, + "balance_loss_mlp": 1.02359772, + "epoch": 0.605422954367823, + "flos": 15154698412800.0, + "grad_norm": 2.064480859713878, + "language_loss": 0.71643317, + "learning_rate": 1.4225529018537352e-06, + "loss": 0.73785627, + "num_input_tokens_seen": 108554155, + "step": 5035, + "time_per_iteration": 2.69499135017395 + }, + { + "auxiliary_loss_clip": 0.01175212, + "auxiliary_loss_mlp": 0.0103231, + "balance_loss_clip": 1.05417705, + "balance_loss_mlp": 1.02462411, + "epoch": 0.6055431972584621, + "flos": 27673912131840.0, + "grad_norm": 1.7741090356843583, + "language_loss": 0.77987969, + "learning_rate": 1.4218071484725082e-06, + "loss": 0.80195487, + "num_input_tokens_seen": 108576275, + "step": 5036, + "time_per_iteration": 2.7809619903564453 + }, + { + "auxiliary_loss_clip": 0.01126603, + "auxiliary_loss_mlp": 0.01034503, + "balance_loss_clip": 0.97973609, + "balance_loss_mlp": 1.02690053, + "epoch": 0.6056634401491012, + "flos": 19387489006080.0, + "grad_norm": 2.2768362315719135, + "language_loss": 0.76256484, + "learning_rate": 1.4210614828026786e-06, + "loss": 0.78417587, + "num_input_tokens_seen": 108594125, + "step": 5037, + "time_per_iteration": 2.7265641689300537 + }, + { + "auxiliary_loss_clip": 0.01171955, + "auxiliary_loss_mlp": 0.01029423, + "balance_loss_clip": 1.05120349, + "balance_loss_mlp": 1.02193308, + "epoch": 0.6057836830397403, + "flos": 24789459294720.0, + "grad_norm": 1.4810850852294448, + "language_loss": 0.74384731, + "learning_rate": 1.4203159049573605e-06, + "loss": 0.76586103, + "num_input_tokens_seen": 108615360, + "step": 5038, + "time_per_iteration": 2.8447744846343994 + }, + { + "auxiliary_loss_clip": 0.0113368, + "auxiliary_loss_mlp": 0.01029714, + "balance_loss_clip": 0.97522718, + "balance_loss_mlp": 1.02186394, + "epoch": 0.6059039259303793, + "flos": 20558248899840.0, + "grad_norm": 1.9118545480161206, + "language_loss": 0.87089205, + "learning_rate": 1.4195704150496593e-06, + "loss": 0.89252597, + "num_input_tokens_seen": 108633075, + "step": 5039, + "time_per_iteration": 2.7453298568725586 + }, + { + "auxiliary_loss_clip": 0.01131715, + "auxiliary_loss_mlp": 0.01028081, + "balance_loss_clip": 0.98067999, + "balance_loss_mlp": 1.0203445, + "epoch": 0.6060241688210185, + "flos": 21069724613760.0, + "grad_norm": 1.673438540443776, + "language_loss": 0.73947299, + "learning_rate": 1.4188250131926639e-06, + "loss": 0.76107097, + "num_input_tokens_seen": 108651875, + "step": 5040, + "time_per_iteration": 2.659986734390259 + }, + { + "auxiliary_loss_clip": 0.01129691, + "auxiliary_loss_mlp": 0.01028223, + "balance_loss_clip": 0.9766584, + "balance_loss_mlp": 1.01998234, + "epoch": 0.6061444117116576, + "flos": 16361081619840.0, + "grad_norm": 1.912866305248973, + "language_loss": 0.80614662, + "learning_rate": 1.4180796994994525e-06, + "loss": 0.82772577, + "num_input_tokens_seen": 108669290, + "step": 5041, + "time_per_iteration": 2.6583306789398193 + }, + { + "auxiliary_loss_clip": 0.01124995, + "auxiliary_loss_mlp": 0.01024843, + "balance_loss_clip": 0.97488892, + "balance_loss_mlp": 1.01719618, + "epoch": 0.6062646546022966, + "flos": 21507296094720.0, + "grad_norm": 2.1876036409715325, + "language_loss": 0.71803194, + "learning_rate": 1.4173344740830877e-06, + "loss": 0.73953038, + "num_input_tokens_seen": 108688420, + "step": 5042, + "time_per_iteration": 2.6409711837768555 + }, + { + "auxiliary_loss_clip": 0.01124883, + "auxiliary_loss_mlp": 0.01023167, + "balance_loss_clip": 0.98124576, + "balance_loss_mlp": 1.01572835, + "epoch": 0.6063848974929358, + "flos": 38983151283840.0, + "grad_norm": 1.8393662823931662, + "language_loss": 0.70719659, + "learning_rate": 1.4165893370566206e-06, + "loss": 0.72867703, + "num_input_tokens_seen": 108712175, + "step": 5043, + "time_per_iteration": 3.686237096786499 + }, + { + "auxiliary_loss_clip": 0.01146238, + "auxiliary_loss_mlp": 0.01025793, + "balance_loss_clip": 1.01265192, + "balance_loss_mlp": 1.01807427, + "epoch": 0.6065051403835748, + "flos": 19646584784640.0, + "grad_norm": 1.7353005414818805, + "language_loss": 0.77502763, + "learning_rate": 1.4158442885330865e-06, + "loss": 0.79674792, + "num_input_tokens_seen": 108730745, + "step": 5044, + "time_per_iteration": 2.623899459838867 + }, + { + "auxiliary_loss_clip": 0.0114449, + "auxiliary_loss_mlp": 0.01027299, + "balance_loss_clip": 1.01305008, + "balance_loss_mlp": 1.01927042, + "epoch": 0.6066253832742139, + "flos": 23513086437120.0, + "grad_norm": 2.0220756363743853, + "language_loss": 0.78875083, + "learning_rate": 1.4150993286255094e-06, + "loss": 0.81046867, + "num_input_tokens_seen": 108749995, + "step": 5045, + "time_per_iteration": 2.6398885250091553 + }, + { + "auxiliary_loss_clip": 0.0117195, + "auxiliary_loss_mlp": 0.01025144, + "balance_loss_clip": 1.05132532, + "balance_loss_mlp": 1.01791358, + "epoch": 0.6067456261648531, + "flos": 19133708440320.0, + "grad_norm": 1.906124805155503, + "language_loss": 0.8040424, + "learning_rate": 1.4143544574468993e-06, + "loss": 0.82601333, + "num_input_tokens_seen": 108768355, + "step": 5046, + "time_per_iteration": 2.5793540477752686 + }, + { + "auxiliary_loss_clip": 0.01150026, + "auxiliary_loss_mlp": 0.01025663, + "balance_loss_clip": 1.01715088, + "balance_loss_mlp": 1.0179857, + "epoch": 0.6068658690554921, + "flos": 20520614424960.0, + "grad_norm": 1.6466923458114813, + "language_loss": 0.82594299, + "learning_rate": 1.4136096751102523e-06, + "loss": 0.84769988, + "num_input_tokens_seen": 108786685, + "step": 5047, + "time_per_iteration": 2.6809208393096924 + }, + { + "auxiliary_loss_clip": 0.01131879, + "auxiliary_loss_mlp": 0.01025169, + "balance_loss_clip": 0.97877812, + "balance_loss_mlp": 1.0182488, + "epoch": 0.6069861119461312, + "flos": 27374560185600.0, + "grad_norm": 1.8707155834644091, + "language_loss": 0.83129567, + "learning_rate": 1.4128649817285516e-06, + "loss": 0.85286605, + "num_input_tokens_seen": 108804820, + "step": 5048, + "time_per_iteration": 2.7421562671661377 + }, + { + "auxiliary_loss_clip": 0.01130877, + "auxiliary_loss_mlp": 0.01034001, + "balance_loss_clip": 0.97607088, + "balance_loss_mlp": 1.02669048, + "epoch": 0.6071063548367702, + "flos": 25626500904960.0, + "grad_norm": 1.9968000220681958, + "language_loss": 0.63236022, + "learning_rate": 1.412120377414766e-06, + "loss": 0.65400898, + "num_input_tokens_seen": 108825010, + "step": 5049, + "time_per_iteration": 2.6865570545196533 + }, + { + "auxiliary_loss_clip": 0.01174511, + "auxiliary_loss_mlp": 0.01029303, + "balance_loss_clip": 1.05574274, + "balance_loss_mlp": 1.02170932, + "epoch": 0.6072265977274094, + "flos": 24460517520000.0, + "grad_norm": 1.5340303975896163, + "language_loss": 0.71429592, + "learning_rate": 1.4113758622818522e-06, + "loss": 0.73633403, + "num_input_tokens_seen": 108845075, + "step": 5050, + "time_per_iteration": 2.7216956615448 + }, + { + "auxiliary_loss_clip": 0.01136284, + "auxiliary_loss_mlp": 0.0050874, + "balance_loss_clip": 0.98068142, + "balance_loss_mlp": 0.0, + "epoch": 0.6073468406180484, + "flos": 18149253413760.0, + "grad_norm": 2.786379032354078, + "language_loss": 0.83187795, + "learning_rate": 1.410631436442751e-06, + "loss": 0.84832823, + "num_input_tokens_seen": 108863870, + "step": 5051, + "time_per_iteration": 2.7411763668060303 + }, + { + "auxiliary_loss_clip": 0.01155749, + "auxiliary_loss_mlp": 0.01028869, + "balance_loss_clip": 1.01517558, + "balance_loss_mlp": 1.02123344, + "epoch": 0.6074670835086875, + "flos": 20697617669760.0, + "grad_norm": 1.9653808557725203, + "language_loss": 0.86363006, + "learning_rate": 1.4098871000103936e-06, + "loss": 0.88547629, + "num_input_tokens_seen": 108882470, + "step": 5052, + "time_per_iteration": 2.674041509628296 + }, + { + "auxiliary_loss_clip": 0.01127509, + "auxiliary_loss_mlp": 0.01026563, + "balance_loss_clip": 0.97477955, + "balance_loss_mlp": 1.01952052, + "epoch": 0.6075873263993267, + "flos": 23769955572480.0, + "grad_norm": 1.750244129140092, + "language_loss": 0.82424724, + "learning_rate": 1.409142853097693e-06, + "loss": 0.845788, + "num_input_tokens_seen": 108902710, + "step": 5053, + "time_per_iteration": 2.692909002304077 + }, + { + "auxiliary_loss_clip": 0.01131967, + "auxiliary_loss_mlp": 0.01027644, + "balance_loss_clip": 0.9789995, + "balance_loss_mlp": 1.01953769, + "epoch": 0.6077075692899657, + "flos": 24454484035200.0, + "grad_norm": 2.1392992249701503, + "language_loss": 0.79639852, + "learning_rate": 1.408398695817553e-06, + "loss": 0.81799471, + "num_input_tokens_seen": 108919935, + "step": 5054, + "time_per_iteration": 2.7921299934387207 + }, + { + "auxiliary_loss_clip": 0.01126908, + "auxiliary_loss_mlp": 0.01033944, + "balance_loss_clip": 0.97598159, + "balance_loss_mlp": 1.02530122, + "epoch": 0.6078278121806048, + "flos": 27382102041600.0, + "grad_norm": 1.6832830407151402, + "language_loss": 0.70285034, + "learning_rate": 1.4076546282828593e-06, + "loss": 0.72445881, + "num_input_tokens_seen": 108942790, + "step": 5055, + "time_per_iteration": 2.753661870956421 + }, + { + "auxiliary_loss_clip": 0.01130936, + "auxiliary_loss_mlp": 0.01028311, + "balance_loss_clip": 0.97318852, + "balance_loss_mlp": 1.01997876, + "epoch": 0.6079480550712439, + "flos": 38436447306240.0, + "grad_norm": 3.5804796440742477, + "language_loss": 0.6624738, + "learning_rate": 1.4069106506064874e-06, + "loss": 0.6840663, + "num_input_tokens_seen": 108964215, + "step": 5056, + "time_per_iteration": 2.8014140129089355 + }, + { + "auxiliary_loss_clip": 0.01125696, + "auxiliary_loss_mlp": 0.01027782, + "balance_loss_clip": 0.979487, + "balance_loss_mlp": 1.02070642, + "epoch": 0.608068297961883, + "flos": 25336271013120.0, + "grad_norm": 1.9756489792675795, + "language_loss": 0.7828083, + "learning_rate": 1.4061667629012989e-06, + "loss": 0.8043431, + "num_input_tokens_seen": 108984885, + "step": 5057, + "time_per_iteration": 5.417322874069214 + }, + { + "auxiliary_loss_clip": 0.01119766, + "auxiliary_loss_mlp": 0.01026426, + "balance_loss_clip": 0.97644204, + "balance_loss_mlp": 1.01912999, + "epoch": 0.608188540852522, + "flos": 24202463235840.0, + "grad_norm": 1.5598701965110733, + "language_loss": 0.83069193, + "learning_rate": 1.40542296528014e-06, + "loss": 0.85215384, + "num_input_tokens_seen": 109004545, + "step": 5058, + "time_per_iteration": 2.7301218509674072 + }, + { + "auxiliary_loss_clip": 0.01150397, + "auxiliary_loss_mlp": 0.01027245, + "balance_loss_clip": 1.01320362, + "balance_loss_mlp": 1.01935351, + "epoch": 0.6083087837431612, + "flos": 21284146851840.0, + "grad_norm": 1.8171748719700345, + "language_loss": 0.76060724, + "learning_rate": 1.4046792578558452e-06, + "loss": 0.78238368, + "num_input_tokens_seen": 109022440, + "step": 5059, + "time_per_iteration": 2.613851308822632 + }, + { + "auxiliary_loss_clip": 0.01122838, + "auxiliary_loss_mlp": 0.01028657, + "balance_loss_clip": 0.97514015, + "balance_loss_mlp": 1.02111685, + "epoch": 0.6084290266338003, + "flos": 16471435178880.0, + "grad_norm": 2.2775609159487376, + "language_loss": 0.76071638, + "learning_rate": 1.4039356407412325e-06, + "loss": 0.78223133, + "num_input_tokens_seen": 109035680, + "step": 5060, + "time_per_iteration": 2.638993501663208 + }, + { + "auxiliary_loss_clip": 0.01051521, + "auxiliary_loss_mlp": 0.00998985, + "balance_loss_clip": 0.97871482, + "balance_loss_mlp": 0.99773294, + "epoch": 0.6085492695244393, + "flos": 66443574931200.0, + "grad_norm": 0.7808818648463434, + "language_loss": 0.57148826, + "learning_rate": 1.40319211404911e-06, + "loss": 0.59199333, + "num_input_tokens_seen": 109090680, + "step": 5061, + "time_per_iteration": 3.1934733390808105 + }, + { + "auxiliary_loss_clip": 0.01174053, + "auxiliary_loss_mlp": 0.01028683, + "balance_loss_clip": 1.05310154, + "balance_loss_mlp": 1.0214051, + "epoch": 0.6086695124150785, + "flos": 23618986709760.0, + "grad_norm": 1.6209941629563358, + "language_loss": 0.90434933, + "learning_rate": 1.4024486778922691e-06, + "loss": 0.92637664, + "num_input_tokens_seen": 109108995, + "step": 5062, + "time_per_iteration": 2.5890536308288574 + }, + { + "auxiliary_loss_clip": 0.01132578, + "auxiliary_loss_mlp": 0.01024893, + "balance_loss_clip": 0.97468728, + "balance_loss_mlp": 1.01744831, + "epoch": 0.6087897553057176, + "flos": 20157054917760.0, + "grad_norm": 1.7720906012530842, + "language_loss": 0.77690208, + "learning_rate": 1.4017053323834884e-06, + "loss": 0.79847682, + "num_input_tokens_seen": 109128825, + "step": 5063, + "time_per_iteration": 2.6761386394500732 + }, + { + "auxiliary_loss_clip": 0.01127933, + "auxiliary_loss_mlp": 0.0102792, + "balance_loss_clip": 0.97410399, + "balance_loss_mlp": 1.02096987, + "epoch": 0.6089099981963566, + "flos": 25482535194240.0, + "grad_norm": 2.596654674964115, + "language_loss": 0.75673556, + "learning_rate": 1.4009620776355333e-06, + "loss": 0.77829409, + "num_input_tokens_seen": 109150425, + "step": 5064, + "time_per_iteration": 2.689953327178955 + }, + { + "auxiliary_loss_clip": 0.0114703, + "auxiliary_loss_mlp": 0.0103135, + "balance_loss_clip": 1.01340747, + "balance_loss_mlp": 1.02388191, + "epoch": 0.6090302410869958, + "flos": 25332895134720.0, + "grad_norm": 2.192788951509825, + "language_loss": 0.79394662, + "learning_rate": 1.4002189137611553e-06, + "loss": 0.81573045, + "num_input_tokens_seen": 109169765, + "step": 5065, + "time_per_iteration": 2.6895008087158203 + }, + { + "auxiliary_loss_clip": 0.01150289, + "auxiliary_loss_mlp": 0.01026878, + "balance_loss_clip": 1.01489747, + "balance_loss_mlp": 1.01893854, + "epoch": 0.6091504839776348, + "flos": 23987358639360.0, + "grad_norm": 1.8916446556188247, + "language_loss": 0.69544697, + "learning_rate": 1.3994758408730901e-06, + "loss": 0.71721864, + "num_input_tokens_seen": 109188950, + "step": 5066, + "time_per_iteration": 2.6566195487976074 + }, + { + "auxiliary_loss_clip": 0.01130684, + "auxiliary_loss_mlp": 0.01029445, + "balance_loss_clip": 0.97954601, + "balance_loss_mlp": 1.02169895, + "epoch": 0.6092707268682739, + "flos": 29643037666560.0, + "grad_norm": 2.010419167600001, + "language_loss": 0.76501185, + "learning_rate": 1.3987328590840629e-06, + "loss": 0.78661317, + "num_input_tokens_seen": 109209895, + "step": 5067, + "time_per_iteration": 2.749265432357788 + }, + { + "auxiliary_loss_clip": 0.01146608, + "auxiliary_loss_mlp": 0.01027712, + "balance_loss_clip": 1.01267278, + "balance_loss_mlp": 1.02035952, + "epoch": 0.609390969758913, + "flos": 24024957200640.0, + "grad_norm": 1.7960165887063433, + "language_loss": 0.85997283, + "learning_rate": 1.397989968506783e-06, + "loss": 0.88171601, + "num_input_tokens_seen": 109228905, + "step": 5068, + "time_per_iteration": 2.6294217109680176 + }, + { + "auxiliary_loss_clip": 0.01180548, + "auxiliary_loss_mlp": 0.01027501, + "balance_loss_clip": 1.05623078, + "balance_loss_mlp": 1.01989555, + "epoch": 0.6095112126495521, + "flos": 11102143288320.0, + "grad_norm": 2.196060887404315, + "language_loss": 0.72406512, + "learning_rate": 1.3972471692539458e-06, + "loss": 0.74614561, + "num_input_tokens_seen": 109243620, + "step": 5069, + "time_per_iteration": 3.4800362586975098 + }, + { + "auxiliary_loss_clip": 0.01123112, + "auxiliary_loss_mlp": 0.01025263, + "balance_loss_clip": 0.97697204, + "balance_loss_mlp": 1.01827741, + "epoch": 0.6096314555401912, + "flos": 17265491187840.0, + "grad_norm": 3.499297186435257, + "language_loss": 0.75582564, + "learning_rate": 1.3965044614382348e-06, + "loss": 0.77730942, + "num_input_tokens_seen": 109259070, + "step": 5070, + "time_per_iteration": 2.5933220386505127 + }, + { + "auxiliary_loss_clip": 0.01178538, + "auxiliary_loss_mlp": 0.0102874, + "balance_loss_clip": 1.05581748, + "balance_loss_mlp": 1.02101493, + "epoch": 0.6097516984308303, + "flos": 21645910679040.0, + "grad_norm": 4.4061291256114155, + "language_loss": 0.75589323, + "learning_rate": 1.3957618451723162e-06, + "loss": 0.77796602, + "num_input_tokens_seen": 109275100, + "step": 5071, + "time_per_iteration": 2.684133768081665 + }, + { + "auxiliary_loss_clip": 0.01132274, + "auxiliary_loss_mlp": 0.01026941, + "balance_loss_clip": 0.97806656, + "balance_loss_mlp": 1.01956153, + "epoch": 0.6098719413214694, + "flos": 27199208966400.0, + "grad_norm": 2.2157191058693173, + "language_loss": 0.71557498, + "learning_rate": 1.3950193205688457e-06, + "loss": 0.73716706, + "num_input_tokens_seen": 109294825, + "step": 5072, + "time_per_iteration": 2.790900230407715 + }, + { + "auxiliary_loss_clip": 0.0112752, + "auxiliary_loss_mlp": 0.01026043, + "balance_loss_clip": 0.97868544, + "balance_loss_mlp": 1.01896822, + "epoch": 0.6099921842121084, + "flos": 20412954385920.0, + "grad_norm": 1.7172460956796203, + "language_loss": 0.83239174, + "learning_rate": 1.3942768877404627e-06, + "loss": 0.85392737, + "num_input_tokens_seen": 109313790, + "step": 5073, + "time_per_iteration": 2.6997220516204834 + }, + { + "auxiliary_loss_clip": 0.01173562, + "auxiliary_loss_mlp": 0.01027532, + "balance_loss_clip": 1.05329287, + "balance_loss_mlp": 1.01986647, + "epoch": 0.6101124271027476, + "flos": 23366139897600.0, + "grad_norm": 1.5965803359091515, + "language_loss": 0.73422563, + "learning_rate": 1.393534546799795e-06, + "loss": 0.75623661, + "num_input_tokens_seen": 109333490, + "step": 5074, + "time_per_iteration": 2.6535634994506836 + }, + { + "auxiliary_loss_clip": 0.01117686, + "auxiliary_loss_mlp": 0.01024653, + "balance_loss_clip": 0.97535437, + "balance_loss_mlp": 1.01743519, + "epoch": 0.6102326699933867, + "flos": 26687840993280.0, + "grad_norm": 1.575183725496196, + "language_loss": 0.6801542, + "learning_rate": 1.3927922978594536e-06, + "loss": 0.70157754, + "num_input_tokens_seen": 109354575, + "step": 5075, + "time_per_iteration": 2.700956344604492 + }, + { + "auxiliary_loss_clip": 0.01046933, + "auxiliary_loss_mlp": 0.01006671, + "balance_loss_clip": 0.97706902, + "balance_loss_mlp": 1.00547302, + "epoch": 0.6103529128840257, + "flos": 60644612551680.0, + "grad_norm": 0.7702562614480563, + "language_loss": 0.57410967, + "learning_rate": 1.3920501410320387e-06, + "loss": 0.59464562, + "num_input_tokens_seen": 109410690, + "step": 5076, + "time_per_iteration": 3.227313280105591 + }, + { + "auxiliary_loss_clip": 0.01125277, + "auxiliary_loss_mlp": 0.01032946, + "balance_loss_clip": 0.97384441, + "balance_loss_mlp": 1.0248754, + "epoch": 0.6104731557746649, + "flos": 19021307806080.0, + "grad_norm": 4.416874618678978, + "language_loss": 0.76165277, + "learning_rate": 1.3913080764301333e-06, + "loss": 0.78323501, + "num_input_tokens_seen": 109427650, + "step": 5077, + "time_per_iteration": 2.667130470275879 + }, + { + "auxiliary_loss_clip": 0.01086462, + "auxiliary_loss_mlp": 0.01021072, + "balance_loss_clip": 0.89635193, + "balance_loss_mlp": 1.0134728, + "epoch": 0.6105933986653039, + "flos": 23366894083200.0, + "grad_norm": 1.8802521471666103, + "language_loss": 0.71159387, + "learning_rate": 1.3905661041663085e-06, + "loss": 0.73266917, + "num_input_tokens_seen": 109448835, + "step": 5078, + "time_per_iteration": 2.7874906063079834 + }, + { + "auxiliary_loss_clip": 0.01150174, + "auxiliary_loss_mlp": 0.01030717, + "balance_loss_clip": 1.01486647, + "balance_loss_mlp": 1.02236044, + "epoch": 0.610713641555943, + "flos": 34637565006720.0, + "grad_norm": 2.151058110308781, + "language_loss": 0.65467703, + "learning_rate": 1.389824224353122e-06, + "loss": 0.6764859, + "num_input_tokens_seen": 109470425, + "step": 5079, + "time_per_iteration": 2.7332804203033447 + }, + { + "auxiliary_loss_clip": 0.01152475, + "auxiliary_loss_mlp": 0.01028549, + "balance_loss_clip": 1.01733649, + "balance_loss_mlp": 1.02100301, + "epoch": 0.610833884446582, + "flos": 26646471504000.0, + "grad_norm": 1.5157347399530476, + "language_loss": 0.76824725, + "learning_rate": 1.389082437103115e-06, + "loss": 0.79005754, + "num_input_tokens_seen": 109489695, + "step": 5080, + "time_per_iteration": 2.744234800338745 + }, + { + "auxiliary_loss_clip": 0.01104036, + "auxiliary_loss_mlp": 0.01030074, + "balance_loss_clip": 0.93719012, + "balance_loss_mlp": 1.02172971, + "epoch": 0.6109541273372212, + "flos": 21215126868480.0, + "grad_norm": 1.8627938181651744, + "language_loss": 0.77947342, + "learning_rate": 1.3883407425288172e-06, + "loss": 0.80081451, + "num_input_tokens_seen": 109510030, + "step": 5081, + "time_per_iteration": 2.736915111541748 + }, + { + "auxiliary_loss_clip": 0.01123619, + "auxiliary_loss_mlp": 0.0102881, + "balance_loss_clip": 0.97366232, + "balance_loss_mlp": 1.02084088, + "epoch": 0.6110743702278603, + "flos": 20084084438400.0, + "grad_norm": 1.8538408295279658, + "language_loss": 0.7937023, + "learning_rate": 1.3875991407427417e-06, + "loss": 0.81522655, + "num_input_tokens_seen": 109528255, + "step": 5082, + "time_per_iteration": 3.6360599994659424 + }, + { + "auxiliary_loss_clip": 0.01013523, + "auxiliary_loss_mlp": 0.01002506, + "balance_loss_clip": 0.90343118, + "balance_loss_mlp": 1.00130844, + "epoch": 0.6111946131184993, + "flos": 68302957438080.0, + "grad_norm": 0.7671246794628769, + "language_loss": 0.58095527, + "learning_rate": 1.38685763185739e-06, + "loss": 0.60111558, + "num_input_tokens_seen": 109581915, + "step": 5083, + "time_per_iteration": 4.221924543380737 + }, + { + "auxiliary_loss_clip": 0.01174884, + "auxiliary_loss_mlp": 0.01030152, + "balance_loss_clip": 1.05331731, + "balance_loss_mlp": 1.02264214, + "epoch": 0.6113148560091385, + "flos": 19937676602880.0, + "grad_norm": 3.1993394429899245, + "language_loss": 0.67728412, + "learning_rate": 1.3861162159852476e-06, + "loss": 0.6993345, + "num_input_tokens_seen": 109600050, + "step": 5084, + "time_per_iteration": 3.551723003387451 + }, + { + "auxiliary_loss_clip": 0.01131495, + "auxiliary_loss_mlp": 0.01027098, + "balance_loss_clip": 0.9775672, + "balance_loss_mlp": 1.01924872, + "epoch": 0.6114350988997775, + "flos": 23731854220800.0, + "grad_norm": 1.6572989555631623, + "language_loss": 0.80252475, + "learning_rate": 1.3853748932387875e-06, + "loss": 0.82411075, + "num_input_tokens_seen": 109620690, + "step": 5085, + "time_per_iteration": 2.677888870239258 + }, + { + "auxiliary_loss_clip": 0.0112051, + "auxiliary_loss_mlp": 0.01024193, + "balance_loss_clip": 0.97638094, + "balance_loss_mlp": 1.01621795, + "epoch": 0.6115553417904166, + "flos": 24023700224640.0, + "grad_norm": 10.098750949540575, + "language_loss": 0.75155771, + "learning_rate": 1.3846336637304671e-06, + "loss": 0.77300477, + "num_input_tokens_seen": 109638960, + "step": 5086, + "time_per_iteration": 2.7064871788024902 + }, + { + "auxiliary_loss_clip": 0.01121892, + "auxiliary_loss_mlp": 0.01026341, + "balance_loss_clip": 0.97732103, + "balance_loss_mlp": 1.01900077, + "epoch": 0.6116755846810558, + "flos": 23733542160000.0, + "grad_norm": 1.9163847583584661, + "language_loss": 0.83064061, + "learning_rate": 1.3838925275727316e-06, + "loss": 0.85212302, + "num_input_tokens_seen": 109659700, + "step": 5087, + "time_per_iteration": 2.678922414779663 + }, + { + "auxiliary_loss_clip": 0.01175742, + "auxiliary_loss_mlp": 0.01023735, + "balance_loss_clip": 1.05497086, + "balance_loss_mlp": 1.01622474, + "epoch": 0.6117958275716948, + "flos": 18661626967680.0, + "grad_norm": 2.3537398740162234, + "language_loss": 0.79018044, + "learning_rate": 1.3831514848780089e-06, + "loss": 0.81217521, + "num_input_tokens_seen": 109679275, + "step": 5088, + "time_per_iteration": 2.5940608978271484 + }, + { + "auxiliary_loss_clip": 0.01146848, + "auxiliary_loss_mlp": 0.01026415, + "balance_loss_clip": 1.01529503, + "balance_loss_mlp": 1.01912522, + "epoch": 0.6119160704623339, + "flos": 16471183783680.0, + "grad_norm": 2.353610831449358, + "language_loss": 0.91814661, + "learning_rate": 1.3824105357587152e-06, + "loss": 0.9398793, + "num_input_tokens_seen": 109696380, + "step": 5089, + "time_per_iteration": 2.5853867530822754 + }, + { + "auxiliary_loss_clip": 0.01123769, + "auxiliary_loss_mlp": 0.01025946, + "balance_loss_clip": 0.97513682, + "balance_loss_mlp": 1.01853728, + "epoch": 0.612036313352973, + "flos": 23915465568000.0, + "grad_norm": 1.4807955678124587, + "language_loss": 0.82587922, + "learning_rate": 1.381669680327253e-06, + "loss": 0.84737647, + "num_input_tokens_seen": 109718060, + "step": 5090, + "time_per_iteration": 2.693054437637329 + }, + { + "auxiliary_loss_clip": 0.01126173, + "auxiliary_loss_mlp": 0.01026872, + "balance_loss_clip": 0.97922939, + "balance_loss_mlp": 1.01973772, + "epoch": 0.6121565562436121, + "flos": 26974766833920.0, + "grad_norm": 1.926336543148388, + "language_loss": 0.71496642, + "learning_rate": 1.380928918696008e-06, + "loss": 0.73649687, + "num_input_tokens_seen": 109736830, + "step": 5091, + "time_per_iteration": 2.700835704803467 + }, + { + "auxiliary_loss_clip": 0.01150355, + "auxiliary_loss_mlp": 0.01028324, + "balance_loss_clip": 1.01356375, + "balance_loss_mlp": 1.02090311, + "epoch": 0.6122767991342511, + "flos": 15668867646720.0, + "grad_norm": 2.3502404848177627, + "language_loss": 0.7178036, + "learning_rate": 1.3801882509773548e-06, + "loss": 0.73959041, + "num_input_tokens_seen": 109754690, + "step": 5092, + "time_per_iteration": 2.608442783355713 + }, + { + "auxiliary_loss_clip": 0.01147439, + "auxiliary_loss_mlp": 0.01027516, + "balance_loss_clip": 1.01435947, + "balance_loss_mlp": 1.0204829, + "epoch": 0.6123970420248903, + "flos": 27964321591680.0, + "grad_norm": 1.6937345979774237, + "language_loss": 0.81620753, + "learning_rate": 1.3794476772836503e-06, + "loss": 0.83795702, + "num_input_tokens_seen": 109775790, + "step": 5093, + "time_per_iteration": 2.7072274684906006 + }, + { + "auxiliary_loss_clip": 0.01099773, + "auxiliary_loss_mlp": 0.01028321, + "balance_loss_clip": 0.94017178, + "balance_loss_mlp": 1.02068555, + "epoch": 0.6125172849155294, + "flos": 21468727866240.0, + "grad_norm": 1.7007304021282266, + "language_loss": 0.84698755, + "learning_rate": 1.3787071977272402e-06, + "loss": 0.86826849, + "num_input_tokens_seen": 109795050, + "step": 5094, + "time_per_iteration": 2.6838462352752686 + }, + { + "auxiliary_loss_clip": 0.01077239, + "auxiliary_loss_mlp": 0.01029901, + "balance_loss_clip": 0.90109283, + "balance_loss_mlp": 1.02160454, + "epoch": 0.6126375278061684, + "flos": 16248321849600.0, + "grad_norm": 2.6616731861865204, + "language_loss": 0.7172178, + "learning_rate": 1.3779668124204535e-06, + "loss": 0.73828924, + "num_input_tokens_seen": 109811465, + "step": 5095, + "time_per_iteration": 3.6492888927459717 + }, + { + "auxiliary_loss_clip": 0.0112049, + "auxiliary_loss_mlp": 0.01028507, + "balance_loss_clip": 0.9763056, + "balance_loss_mlp": 1.02137232, + "epoch": 0.6127577706968076, + "flos": 20448865008000.0, + "grad_norm": 2.269455575467454, + "language_loss": 0.80671978, + "learning_rate": 1.3772265214756074e-06, + "loss": 0.82820976, + "num_input_tokens_seen": 109831225, + "step": 5096, + "time_per_iteration": 2.688570261001587 + }, + { + "auxiliary_loss_clip": 0.01153713, + "auxiliary_loss_mlp": 0.01025378, + "balance_loss_clip": 1.01293993, + "balance_loss_mlp": 1.01828504, + "epoch": 0.6128780135874466, + "flos": 18260397072000.0, + "grad_norm": 1.7032867296969778, + "language_loss": 0.75399256, + "learning_rate": 1.3764863250050025e-06, + "loss": 0.77578342, + "num_input_tokens_seen": 109849465, + "step": 5097, + "time_per_iteration": 2.6132140159606934 + }, + { + "auxiliary_loss_clip": 0.01104627, + "auxiliary_loss_mlp": 0.01032476, + "balance_loss_clip": 0.93753338, + "balance_loss_mlp": 1.02505565, + "epoch": 0.6129982564780857, + "flos": 24937088192640.0, + "grad_norm": 2.102424660882896, + "language_loss": 0.80852437, + "learning_rate": 1.3757462231209272e-06, + "loss": 0.82989544, + "num_input_tokens_seen": 109869770, + "step": 5098, + "time_per_iteration": 2.7468111515045166 + }, + { + "auxiliary_loss_clip": 0.01120439, + "auxiliary_loss_mlp": 0.0102849, + "balance_loss_clip": 0.97509515, + "balance_loss_mlp": 1.02083671, + "epoch": 0.6131184993687249, + "flos": 22492038430080.0, + "grad_norm": 1.9998270765075652, + "language_loss": 0.89073265, + "learning_rate": 1.3750062159356525e-06, + "loss": 0.91222191, + "num_input_tokens_seen": 109889120, + "step": 5099, + "time_per_iteration": 2.7011561393737793 + }, + { + "auxiliary_loss_clip": 0.01094617, + "auxiliary_loss_mlp": 0.01026077, + "balance_loss_clip": 0.93655396, + "balance_loss_mlp": 1.01868641, + "epoch": 0.6132387422593639, + "flos": 15885839750400.0, + "grad_norm": 1.92569961242723, + "language_loss": 0.83374554, + "learning_rate": 1.3742663035614382e-06, + "loss": 0.85495245, + "num_input_tokens_seen": 109906490, + "step": 5100, + "time_per_iteration": 2.679603099822998 + }, + { + "auxiliary_loss_clip": 0.01176875, + "auxiliary_loss_mlp": 0.01024546, + "balance_loss_clip": 1.05550981, + "balance_loss_mlp": 1.01688635, + "epoch": 0.613358985150003, + "flos": 25411539962880.0, + "grad_norm": 1.7093832429236697, + "language_loss": 0.80141115, + "learning_rate": 1.3735264861105283e-06, + "loss": 0.82342541, + "num_input_tokens_seen": 109927130, + "step": 5101, + "time_per_iteration": 2.6281375885009766 + }, + { + "auxiliary_loss_clip": 0.01105208, + "auxiliary_loss_mlp": 0.01022644, + "balance_loss_clip": 0.93696487, + "balance_loss_mlp": 1.0151751, + "epoch": 0.6134792280406421, + "flos": 21361283308800.0, + "grad_norm": 1.8739434447715762, + "language_loss": 0.78117418, + "learning_rate": 1.372786763695152e-06, + "loss": 0.8024528, + "num_input_tokens_seen": 109945890, + "step": 5102, + "time_per_iteration": 2.708158254623413 + }, + { + "auxiliary_loss_clip": 0.01153472, + "auxiliary_loss_mlp": 0.01030562, + "balance_loss_clip": 1.01497364, + "balance_loss_mlp": 1.02271771, + "epoch": 0.6135994709312812, + "flos": 21211248199680.0, + "grad_norm": 1.8649257398057983, + "language_loss": 0.77270854, + "learning_rate": 1.3720471364275257e-06, + "loss": 0.79454887, + "num_input_tokens_seen": 109965535, + "step": 5103, + "time_per_iteration": 2.6123206615448 + }, + { + "auxiliary_loss_clip": 0.01101042, + "auxiliary_loss_mlp": 0.00509691, + "balance_loss_clip": 0.93637037, + "balance_loss_mlp": 0.0, + "epoch": 0.6137197138219203, + "flos": 14794047907200.0, + "grad_norm": 2.1442163519088315, + "language_loss": 0.78192121, + "learning_rate": 1.3713076044198486e-06, + "loss": 0.79802853, + "num_input_tokens_seen": 109982345, + "step": 5104, + "time_per_iteration": 2.6928160190582275 + }, + { + "auxiliary_loss_clip": 0.0112294, + "auxiliary_loss_mlp": 0.01025243, + "balance_loss_clip": 0.9756912, + "balance_loss_mlp": 1.01754165, + "epoch": 0.6138399567125594, + "flos": 20084515401600.0, + "grad_norm": 2.1052074600943413, + "language_loss": 0.81157291, + "learning_rate": 1.3705681677843086e-06, + "loss": 0.83305466, + "num_input_tokens_seen": 110000940, + "step": 5105, + "time_per_iteration": 2.656712532043457 + }, + { + "auxiliary_loss_clip": 0.0106907, + "auxiliary_loss_mlp": 0.0099982, + "balance_loss_clip": 1.0144577, + "balance_loss_mlp": 0.99861044, + "epoch": 0.6139601996031985, + "flos": 60123838193280.0, + "grad_norm": 0.76714690669851, + "language_loss": 0.60533518, + "learning_rate": 1.3698288266330768e-06, + "loss": 0.62602407, + "num_input_tokens_seen": 110061565, + "step": 5106, + "time_per_iteration": 3.279374599456787 + }, + { + "auxiliary_loss_clip": 0.01129476, + "auxiliary_loss_mlp": 0.01027216, + "balance_loss_clip": 0.9829905, + "balance_loss_mlp": 1.01955128, + "epoch": 0.6140804424938375, + "flos": 23586703361280.0, + "grad_norm": 2.710268919214934, + "language_loss": 0.73213702, + "learning_rate": 1.3690895810783113e-06, + "loss": 0.75370395, + "num_input_tokens_seen": 110080360, + "step": 5107, + "time_per_iteration": 2.7074661254882812 + }, + { + "auxiliary_loss_clip": 0.01064556, + "auxiliary_loss_mlp": 0.00509718, + "balance_loss_clip": 0.85847163, + "balance_loss_mlp": 0.0, + "epoch": 0.6142006853844767, + "flos": 21398199511680.0, + "grad_norm": 2.008538414347548, + "language_loss": 0.71516681, + "learning_rate": 1.3683504312321543e-06, + "loss": 0.73090947, + "num_input_tokens_seen": 110100695, + "step": 5108, + "time_per_iteration": 2.7805988788604736 + }, + { + "auxiliary_loss_clip": 0.01153722, + "auxiliary_loss_mlp": 0.01025574, + "balance_loss_clip": 1.01419342, + "balance_loss_mlp": 1.01768291, + "epoch": 0.6143209282751158, + "flos": 12057367622400.0, + "grad_norm": 6.120912632674781, + "language_loss": 0.79975134, + "learning_rate": 1.3676113772067355e-06, + "loss": 0.82154429, + "num_input_tokens_seen": 110117750, + "step": 5109, + "time_per_iteration": 5.380131959915161 + }, + { + "auxiliary_loss_clip": 0.01084959, + "auxiliary_loss_mlp": 0.01024754, + "balance_loss_clip": 0.89940035, + "balance_loss_mlp": 1.01738095, + "epoch": 0.6144411711657548, + "flos": 25082274965760.0, + "grad_norm": 1.9936908534722364, + "language_loss": 0.72215378, + "learning_rate": 1.3668724191141671e-06, + "loss": 0.74325091, + "num_input_tokens_seen": 110137020, + "step": 5110, + "time_per_iteration": 2.7944211959838867 + }, + { + "auxiliary_loss_clip": 0.0109965, + "auxiliary_loss_mlp": 0.01027286, + "balance_loss_clip": 0.94275796, + "balance_loss_mlp": 1.01963878, + "epoch": 0.6145614140563939, + "flos": 20114069316480.0, + "grad_norm": 2.0988416060604487, + "language_loss": 0.66268146, + "learning_rate": 1.3661335570665493e-06, + "loss": 0.68395078, + "num_input_tokens_seen": 110154930, + "step": 5111, + "time_per_iteration": 2.6909101009368896 + }, + { + "auxiliary_loss_clip": 0.01134679, + "auxiliary_loss_mlp": 0.01032808, + "balance_loss_clip": 0.98136961, + "balance_loss_mlp": 1.02550054, + "epoch": 0.614681656947033, + "flos": 16800376953600.0, + "grad_norm": 2.25756994811921, + "language_loss": 0.69615769, + "learning_rate": 1.3653947911759676e-06, + "loss": 0.71783262, + "num_input_tokens_seen": 110172480, + "step": 5112, + "time_per_iteration": 2.6553285121917725 + }, + { + "auxiliary_loss_clip": 0.01075746, + "auxiliary_loss_mlp": 0.01031764, + "balance_loss_clip": 0.89880842, + "balance_loss_mlp": 1.02462673, + "epoch": 0.6148018998376721, + "flos": 38801587011840.0, + "grad_norm": 1.6229820643335273, + "language_loss": 0.74307215, + "learning_rate": 1.3646561215544904e-06, + "loss": 0.76414722, + "num_input_tokens_seen": 110197120, + "step": 5113, + "time_per_iteration": 2.909431219100952 + }, + { + "auxiliary_loss_clip": 0.01149658, + "auxiliary_loss_mlp": 0.0102824, + "balance_loss_clip": 1.01465392, + "balance_loss_mlp": 1.02018738, + "epoch": 0.6149221427283111, + "flos": 23327032965120.0, + "grad_norm": 2.4156503969346232, + "language_loss": 0.7948072, + "learning_rate": 1.363917548314176e-06, + "loss": 0.81658614, + "num_input_tokens_seen": 110216385, + "step": 5114, + "time_per_iteration": 2.6779227256774902 + }, + { + "auxiliary_loss_clip": 0.01158039, + "auxiliary_loss_mlp": 0.01027305, + "balance_loss_clip": 1.0162909, + "balance_loss_mlp": 1.01891875, + "epoch": 0.6150423856189503, + "flos": 22379494141440.0, + "grad_norm": 1.6804027915173543, + "language_loss": 0.73137432, + "learning_rate": 1.3631790715670626e-06, + "loss": 0.75322771, + "num_input_tokens_seen": 110234790, + "step": 5115, + "time_per_iteration": 2.613706588745117 + }, + { + "auxiliary_loss_clip": 0.01021512, + "auxiliary_loss_mlp": 0.01029276, + "balance_loss_clip": 0.82235432, + "balance_loss_mlp": 1.02301443, + "epoch": 0.6151626285095894, + "flos": 18692078722560.0, + "grad_norm": 1.7354680928367372, + "language_loss": 0.85727268, + "learning_rate": 1.3624406914251783e-06, + "loss": 0.87778062, + "num_input_tokens_seen": 110251910, + "step": 5116, + "time_per_iteration": 2.836552381515503 + }, + { + "auxiliary_loss_clip": 0.01150123, + "auxiliary_loss_mlp": 0.01027589, + "balance_loss_clip": 1.01246536, + "balance_loss_mlp": 1.0201149, + "epoch": 0.6152828714002284, + "flos": 15851688894720.0, + "grad_norm": 1.8041134448439637, + "language_loss": 0.88624454, + "learning_rate": 1.3617024080005335e-06, + "loss": 0.90802163, + "num_input_tokens_seen": 110268810, + "step": 5117, + "time_per_iteration": 2.657418966293335 + }, + { + "auxiliary_loss_clip": 0.01131112, + "auxiliary_loss_mlp": 0.00509289, + "balance_loss_clip": 0.97573054, + "balance_loss_mlp": 0.0, + "epoch": 0.6154031142908676, + "flos": 24869792062080.0, + "grad_norm": 1.5253667576219274, + "language_loss": 0.74681735, + "learning_rate": 1.3609642214051266e-06, + "loss": 0.76322132, + "num_input_tokens_seen": 110293035, + "step": 5118, + "time_per_iteration": 2.7718560695648193 + }, + { + "auxiliary_loss_clip": 0.01120865, + "auxiliary_loss_mlp": 0.01028638, + "balance_loss_clip": 0.97847176, + "balance_loss_mlp": 1.0208596, + "epoch": 0.6155233571815066, + "flos": 19244744357760.0, + "grad_norm": 2.23907911268355, + "language_loss": 0.66143847, + "learning_rate": 1.3602261317509385e-06, + "loss": 0.68293351, + "num_input_tokens_seen": 110309695, + "step": 5119, + "time_per_iteration": 2.6200766563415527 + }, + { + "auxiliary_loss_clip": 0.01152098, + "auxiliary_loss_mlp": 0.01030171, + "balance_loss_clip": 1.01431632, + "balance_loss_mlp": 1.02206469, + "epoch": 0.6156436000721457, + "flos": 18770077105920.0, + "grad_norm": 2.698254654425029, + "language_loss": 0.82782745, + "learning_rate": 1.3594881391499387e-06, + "loss": 0.84965008, + "num_input_tokens_seen": 110328610, + "step": 5120, + "time_per_iteration": 2.6176788806915283 + }, + { + "auxiliary_loss_clip": 0.0112879, + "auxiliary_loss_mlp": 0.01028991, + "balance_loss_clip": 0.97616416, + "balance_loss_mlp": 1.02094495, + "epoch": 0.6157638429627849, + "flos": 18041198325120.0, + "grad_norm": 7.6916060406402105, + "language_loss": 0.79419339, + "learning_rate": 1.3587502437140778e-06, + "loss": 0.81577122, + "num_input_tokens_seen": 110346775, + "step": 5121, + "time_per_iteration": 3.6506922245025635 + }, + { + "auxiliary_loss_clip": 0.01131061, + "auxiliary_loss_mlp": 0.01026576, + "balance_loss_clip": 0.97571129, + "balance_loss_mlp": 1.01829076, + "epoch": 0.6158840858534239, + "flos": 25556726736000.0, + "grad_norm": 2.0093470541675695, + "language_loss": 0.85558927, + "learning_rate": 1.3580124455552952e-06, + "loss": 0.87716568, + "num_input_tokens_seen": 110366140, + "step": 5122, + "time_per_iteration": 2.6907601356506348 + }, + { + "auxiliary_loss_clip": 0.01153232, + "auxiliary_loss_mlp": 0.00509182, + "balance_loss_clip": 1.01697373, + "balance_loss_mlp": 0.0, + "epoch": 0.616004328744063, + "flos": 24640788902400.0, + "grad_norm": 1.7637024978963265, + "language_loss": 0.87271476, + "learning_rate": 1.3572747447855148e-06, + "loss": 0.88933885, + "num_input_tokens_seen": 110386550, + "step": 5123, + "time_per_iteration": 2.6863112449645996 + }, + { + "auxiliary_loss_clip": 0.01174361, + "auxiliary_loss_mlp": 0.01028063, + "balance_loss_clip": 1.05279469, + "balance_loss_mlp": 1.02061844, + "epoch": 0.6161245716347021, + "flos": 21689686379520.0, + "grad_norm": 1.7686018264718384, + "language_loss": 0.69343215, + "learning_rate": 1.356537141516644e-06, + "loss": 0.71545637, + "num_input_tokens_seen": 110403970, + "step": 5124, + "time_per_iteration": 2.60087513923645 + }, + { + "auxiliary_loss_clip": 0.0114833, + "auxiliary_loss_mlp": 0.01026223, + "balance_loss_clip": 1.01490164, + "balance_loss_mlp": 1.01911211, + "epoch": 0.6162448145253412, + "flos": 35189225061120.0, + "grad_norm": 1.975341426315994, + "language_loss": 0.61631507, + "learning_rate": 1.3557996358605775e-06, + "loss": 0.63806063, + "num_input_tokens_seen": 110423890, + "step": 5125, + "time_per_iteration": 2.724250078201294 + }, + { + "auxiliary_loss_clip": 0.01149824, + "auxiliary_loss_mlp": 0.01024538, + "balance_loss_clip": 1.01394928, + "balance_loss_mlp": 1.01728082, + "epoch": 0.6163650574159802, + "flos": 21615279356160.0, + "grad_norm": 2.473639411699305, + "language_loss": 0.70356524, + "learning_rate": 1.3550622279291941e-06, + "loss": 0.7253089, + "num_input_tokens_seen": 110442035, + "step": 5126, + "time_per_iteration": 2.6255111694335938 + }, + { + "auxiliary_loss_clip": 0.01071158, + "auxiliary_loss_mlp": 0.01025145, + "balance_loss_clip": 0.89691192, + "balance_loss_mlp": 1.017802, + "epoch": 0.6164853003066194, + "flos": 24572163968640.0, + "grad_norm": 1.4024474385731311, + "language_loss": 0.83251667, + "learning_rate": 1.354324917834358e-06, + "loss": 0.85347968, + "num_input_tokens_seen": 110463280, + "step": 5127, + "time_per_iteration": 2.778869867324829 + }, + { + "auxiliary_loss_clip": 0.01053918, + "auxiliary_loss_mlp": 0.00509307, + "balance_loss_clip": 0.86245179, + "balance_loss_mlp": 0.0, + "epoch": 0.6166055431972585, + "flos": 21835986474240.0, + "grad_norm": 1.6363172103189678, + "language_loss": 0.77142811, + "learning_rate": 1.353587705687918e-06, + "loss": 0.78706038, + "num_input_tokens_seen": 110481455, + "step": 5128, + "time_per_iteration": 2.860902786254883 + }, + { + "auxiliary_loss_clip": 0.01131066, + "auxiliary_loss_mlp": 0.01026351, + "balance_loss_clip": 0.97798169, + "balance_loss_mlp": 1.01897216, + "epoch": 0.6167257860878975, + "flos": 17785262943360.0, + "grad_norm": 3.545665273757247, + "language_loss": 0.71815813, + "learning_rate": 1.3528505916017096e-06, + "loss": 0.73973227, + "num_input_tokens_seen": 110499155, + "step": 5129, + "time_per_iteration": 2.6484856605529785 + }, + { + "auxiliary_loss_clip": 0.01150035, + "auxiliary_loss_mlp": 0.01034984, + "balance_loss_clip": 1.01299047, + "balance_loss_mlp": 1.02713966, + "epoch": 0.6168460289785367, + "flos": 23214811898880.0, + "grad_norm": 1.9265818343190908, + "language_loss": 0.88770294, + "learning_rate": 1.3521135756875514e-06, + "loss": 0.90955305, + "num_input_tokens_seen": 110515470, + "step": 5130, + "time_per_iteration": 2.64778995513916 + }, + { + "auxiliary_loss_clip": 0.01050496, + "auxiliary_loss_mlp": 0.01029164, + "balance_loss_clip": 0.86236751, + "balance_loss_mlp": 1.02180886, + "epoch": 0.6169662718691757, + "flos": 26213281482240.0, + "grad_norm": 1.4034390898595666, + "language_loss": 0.86209929, + "learning_rate": 1.3513766580572496e-06, + "loss": 0.88289583, + "num_input_tokens_seen": 110538290, + "step": 5131, + "time_per_iteration": 3.018311023712158 + }, + { + "auxiliary_loss_clip": 0.01147022, + "auxiliary_loss_mlp": 0.01028346, + "balance_loss_clip": 1.01357698, + "balance_loss_mlp": 1.02080643, + "epoch": 0.6170865147598148, + "flos": 19026120228480.0, + "grad_norm": 2.009259409204186, + "language_loss": 0.76641762, + "learning_rate": 1.3506398388225924e-06, + "loss": 0.78817129, + "num_input_tokens_seen": 110555610, + "step": 5132, + "time_per_iteration": 2.716249704360962 + }, + { + "auxiliary_loss_clip": 0.01172951, + "auxiliary_loss_mlp": 0.01028, + "balance_loss_clip": 1.05353069, + "balance_loss_mlp": 1.02070487, + "epoch": 0.617206757650454, + "flos": 18260361158400.0, + "grad_norm": 1.6436548824027588, + "language_loss": 0.71817303, + "learning_rate": 1.349903118095355e-06, + "loss": 0.74018252, + "num_input_tokens_seen": 110574745, + "step": 5133, + "time_per_iteration": 2.625487804412842 + }, + { + "auxiliary_loss_clip": 0.01156235, + "auxiliary_loss_mlp": 0.01026942, + "balance_loss_clip": 1.01635468, + "balance_loss_mlp": 1.01929438, + "epoch": 0.617327000541093, + "flos": 18186959715840.0, + "grad_norm": 1.6789861729182145, + "language_loss": 0.73338044, + "learning_rate": 1.349166495987298e-06, + "loss": 0.75521219, + "num_input_tokens_seen": 110593310, + "step": 5134, + "time_per_iteration": 2.6670022010803223 + }, + { + "auxiliary_loss_clip": 0.01048015, + "auxiliary_loss_mlp": 0.01000312, + "balance_loss_clip": 0.96314412, + "balance_loss_mlp": 0.99900705, + "epoch": 0.6174472434317321, + "flos": 61833796122240.0, + "grad_norm": 0.8262832707695629, + "language_loss": 0.60870922, + "learning_rate": 1.348429972610166e-06, + "loss": 0.62919247, + "num_input_tokens_seen": 110657615, + "step": 5135, + "time_per_iteration": 6.183510780334473 + }, + { + "auxiliary_loss_clip": 0.01002456, + "auxiliary_loss_mlp": 0.01001643, + "balance_loss_clip": 0.88852966, + "balance_loss_mlp": 1.0003314, + "epoch": 0.6175674863223712, + "flos": 71230970494080.0, + "grad_norm": 0.8383971007728365, + "language_loss": 0.57821131, + "learning_rate": 1.3476935480756897e-06, + "loss": 0.5982523, + "num_input_tokens_seen": 110714365, + "step": 5136, + "time_per_iteration": 3.1814839839935303 + }, + { + "auxiliary_loss_clip": 0.01096405, + "auxiliary_loss_mlp": 0.01025304, + "balance_loss_clip": 0.93597376, + "balance_loss_mlp": 1.01753497, + "epoch": 0.6176877292130103, + "flos": 21835447770240.0, + "grad_norm": 2.123445918423543, + "language_loss": 0.75415778, + "learning_rate": 1.346957222495583e-06, + "loss": 0.77537483, + "num_input_tokens_seen": 110732160, + "step": 5137, + "time_per_iteration": 2.720057964324951 + }, + { + "auxiliary_loss_clip": 0.01136038, + "auxiliary_loss_mlp": 0.00509676, + "balance_loss_clip": 0.9807179, + "balance_loss_mlp": 0.0, + "epoch": 0.6178079721036493, + "flos": 17741738638080.0, + "grad_norm": 12.605905180618354, + "language_loss": 0.71069402, + "learning_rate": 1.3462209959815466e-06, + "loss": 0.72715116, + "num_input_tokens_seen": 110746900, + "step": 5138, + "time_per_iteration": 2.7019362449645996 + }, + { + "auxiliary_loss_clip": 0.01131166, + "auxiliary_loss_mlp": 0.01031812, + "balance_loss_clip": 0.97803789, + "balance_loss_mlp": 1.02465343, + "epoch": 0.6179282149942885, + "flos": 22633131052800.0, + "grad_norm": 1.8715442600112089, + "language_loss": 0.74665248, + "learning_rate": 1.345484868645265e-06, + "loss": 0.76828223, + "num_input_tokens_seen": 110765710, + "step": 5139, + "time_per_iteration": 2.6856157779693604 + }, + { + "auxiliary_loss_clip": 0.01110745, + "auxiliary_loss_mlp": 0.01026963, + "balance_loss_clip": 0.93786675, + "balance_loss_mlp": 1.01964343, + "epoch": 0.6180484578849276, + "flos": 22310330503680.0, + "grad_norm": 1.8786414908958151, + "language_loss": 0.78528988, + "learning_rate": 1.3447488405984088e-06, + "loss": 0.80666691, + "num_input_tokens_seen": 110783970, + "step": 5140, + "time_per_iteration": 2.7435715198516846 + }, + { + "auxiliary_loss_clip": 0.01128167, + "auxiliary_loss_mlp": 0.01024572, + "balance_loss_clip": 0.97765338, + "balance_loss_mlp": 1.01660824, + "epoch": 0.6181687007755666, + "flos": 35225458905600.0, + "grad_norm": 3.2523336266072356, + "language_loss": 0.69833744, + "learning_rate": 1.3440129119526322e-06, + "loss": 0.71986485, + "num_input_tokens_seen": 110806395, + "step": 5141, + "time_per_iteration": 2.7445921897888184 + }, + { + "auxiliary_loss_clip": 0.01070779, + "auxiliary_loss_mlp": 0.0100393, + "balance_loss_clip": 1.01655459, + "balance_loss_mlp": 1.00293446, + "epoch": 0.6182889436662057, + "flos": 61547370094080.0, + "grad_norm": 0.8056321632889489, + "language_loss": 0.51212335, + "learning_rate": 1.3432770828195762e-06, + "loss": 0.53287047, + "num_input_tokens_seen": 110867380, + "step": 5142, + "time_per_iteration": 3.307842254638672 + }, + { + "auxiliary_loss_clip": 0.01098631, + "auxiliary_loss_mlp": 0.01027626, + "balance_loss_clip": 0.93689758, + "balance_loss_mlp": 1.02050376, + "epoch": 0.6184091865568448, + "flos": 19609991804160.0, + "grad_norm": 2.2369649618505623, + "language_loss": 0.70188391, + "learning_rate": 1.3425413533108635e-06, + "loss": 0.7231465, + "num_input_tokens_seen": 110885980, + "step": 5143, + "time_per_iteration": 2.7376532554626465 + }, + { + "auxiliary_loss_clip": 0.01082344, + "auxiliary_loss_mlp": 0.01031039, + "balance_loss_clip": 0.90348321, + "balance_loss_mlp": 1.02344561, + "epoch": 0.6185294294474839, + "flos": 23586882929280.0, + "grad_norm": 2.3857986934102793, + "language_loss": 0.70551991, + "learning_rate": 1.341805723538105e-06, + "loss": 0.72665381, + "num_input_tokens_seen": 110906085, + "step": 5144, + "time_per_iteration": 2.927048683166504 + }, + { + "auxiliary_loss_clip": 0.0113403, + "auxiliary_loss_mlp": 0.01028747, + "balance_loss_clip": 0.97718018, + "balance_loss_mlp": 1.02096272, + "epoch": 0.618649672338123, + "flos": 26762032535040.0, + "grad_norm": 1.6516136604356892, + "language_loss": 0.77824765, + "learning_rate": 1.3410701936128948e-06, + "loss": 0.79987538, + "num_input_tokens_seen": 110928865, + "step": 5145, + "time_per_iteration": 2.765662431716919 + }, + { + "auxiliary_loss_clip": 0.01148933, + "auxiliary_loss_mlp": 0.01029965, + "balance_loss_clip": 1.01508832, + "balance_loss_mlp": 1.02236521, + "epoch": 0.6187699152287621, + "flos": 14456630522880.0, + "grad_norm": 2.710533201501538, + "language_loss": 0.84990966, + "learning_rate": 1.340334763646812e-06, + "loss": 0.87169862, + "num_input_tokens_seen": 110943000, + "step": 5146, + "time_per_iteration": 2.605767011642456 + }, + { + "auxiliary_loss_clip": 0.0117761, + "auxiliary_loss_mlp": 0.01027351, + "balance_loss_clip": 1.05417252, + "balance_loss_mlp": 1.01957297, + "epoch": 0.6188901581194012, + "flos": 20084766796800.0, + "grad_norm": 1.6800455009761495, + "language_loss": 0.74687034, + "learning_rate": 1.3395994337514218e-06, + "loss": 0.76891994, + "num_input_tokens_seen": 110963170, + "step": 5147, + "time_per_iteration": 3.480820655822754 + }, + { + "auxiliary_loss_clip": 0.01142908, + "auxiliary_loss_mlp": 0.0102512, + "balance_loss_clip": 1.01227129, + "balance_loss_mlp": 1.01772881, + "epoch": 0.6190104010100402, + "flos": 25700728360320.0, + "grad_norm": 1.7714529744621523, + "language_loss": 0.78602982, + "learning_rate": 1.3388642040382725e-06, + "loss": 0.80771005, + "num_input_tokens_seen": 110983595, + "step": 5148, + "time_per_iteration": 2.67315936088562 + }, + { + "auxiliary_loss_clip": 0.01107471, + "auxiliary_loss_mlp": 0.01023991, + "balance_loss_clip": 0.9351449, + "balance_loss_mlp": 1.01674271, + "epoch": 0.6191306439006794, + "flos": 30442372974720.0, + "grad_norm": 2.1115226322367455, + "language_loss": 0.84119785, + "learning_rate": 1.3381290746188975e-06, + "loss": 0.86251247, + "num_input_tokens_seen": 111002965, + "step": 5149, + "time_per_iteration": 2.819558620452881 + }, + { + "auxiliary_loss_clip": 0.01154154, + "auxiliary_loss_mlp": 0.01032282, + "balance_loss_clip": 1.01924467, + "balance_loss_mlp": 1.02472425, + "epoch": 0.6192508867913185, + "flos": 26685793918080.0, + "grad_norm": 2.3836312804588853, + "language_loss": 0.67146641, + "learning_rate": 1.3373940456048152e-06, + "loss": 0.69333076, + "num_input_tokens_seen": 111022990, + "step": 5150, + "time_per_iteration": 2.696963310241699 + }, + { + "auxiliary_loss_clip": 0.01173979, + "auxiliary_loss_mlp": 0.01026816, + "balance_loss_clip": 1.05444598, + "balance_loss_mlp": 1.01960421, + "epoch": 0.6193711296819575, + "flos": 36722036090880.0, + "grad_norm": 1.5102577204955179, + "language_loss": 0.59199941, + "learning_rate": 1.3366591171075299e-06, + "loss": 0.61400735, + "num_input_tokens_seen": 111046495, + "step": 5151, + "time_per_iteration": 2.733609199523926 + }, + { + "auxiliary_loss_clip": 0.01127822, + "auxiliary_loss_mlp": 0.01028032, + "balance_loss_clip": 0.97691661, + "balance_loss_mlp": 1.02050066, + "epoch": 0.6194913725725967, + "flos": 25192556697600.0, + "grad_norm": 2.3030796680496954, + "language_loss": 0.91060972, + "learning_rate": 1.335924289238529e-06, + "loss": 0.93216825, + "num_input_tokens_seen": 111065705, + "step": 5152, + "time_per_iteration": 2.7077040672302246 + }, + { + "auxiliary_loss_clip": 0.01157785, + "auxiliary_loss_mlp": 0.00509714, + "balance_loss_clip": 1.02334464, + "balance_loss_mlp": 0.0, + "epoch": 0.6196116154632357, + "flos": 21178821196800.0, + "grad_norm": 1.6143382333918026, + "language_loss": 0.76558775, + "learning_rate": 1.3351895621092859e-06, + "loss": 0.78226274, + "num_input_tokens_seen": 111086050, + "step": 5153, + "time_per_iteration": 2.6670384407043457 + }, + { + "auxiliary_loss_clip": 0.00996555, + "auxiliary_loss_mlp": 0.01031085, + "balance_loss_clip": 0.77752876, + "balance_loss_mlp": 1.02400959, + "epoch": 0.6197318583538748, + "flos": 16253744803200.0, + "grad_norm": 1.8809608858914282, + "language_loss": 0.77015573, + "learning_rate": 1.3344549358312567e-06, + "loss": 0.7904321, + "num_input_tokens_seen": 111104450, + "step": 5154, + "time_per_iteration": 3.0317726135253906 + }, + { + "auxiliary_loss_clip": 0.01157763, + "auxiliary_loss_mlp": 0.01024716, + "balance_loss_clip": 1.01649106, + "balance_loss_mlp": 1.01742053, + "epoch": 0.619852101244514, + "flos": 24425612478720.0, + "grad_norm": 1.910883518559551, + "language_loss": 0.7887361, + "learning_rate": 1.3337204105158852e-06, + "loss": 0.81056094, + "num_input_tokens_seen": 111123320, + "step": 5155, + "time_per_iteration": 2.939964532852173 + }, + { + "auxiliary_loss_clip": 0.01089844, + "auxiliary_loss_mlp": 0.01024585, + "balance_loss_clip": 0.92939508, + "balance_loss_mlp": 1.01653206, + "epoch": 0.619972344135153, + "flos": 16727298733440.0, + "grad_norm": 5.129770284570844, + "language_loss": 0.73210955, + "learning_rate": 1.332985986274597e-06, + "loss": 0.75325382, + "num_input_tokens_seen": 111140950, + "step": 5156, + "time_per_iteration": 2.6915202140808105 + }, + { + "auxiliary_loss_clip": 0.01052517, + "auxiliary_loss_mlp": 0.00508616, + "balance_loss_clip": 0.86372447, + "balance_loss_mlp": 0.0, + "epoch": 0.6200925870257921, + "flos": 12495190498560.0, + "grad_norm": 2.0850727275353345, + "language_loss": 0.75176108, + "learning_rate": 1.3322516632188047e-06, + "loss": 0.76737249, + "num_input_tokens_seen": 111157845, + "step": 5157, + "time_per_iteration": 2.863272190093994 + }, + { + "auxiliary_loss_clip": 0.01101209, + "auxiliary_loss_mlp": 0.01027889, + "balance_loss_clip": 0.93655705, + "balance_loss_mlp": 1.02000332, + "epoch": 0.6202128299164312, + "flos": 26539350168960.0, + "grad_norm": 1.8206459339860939, + "language_loss": 0.6723929, + "learning_rate": 1.3315174414599045e-06, + "loss": 0.69368386, + "num_input_tokens_seen": 111179165, + "step": 5158, + "time_per_iteration": 2.77070951461792 + }, + { + "auxiliary_loss_clip": 0.01145066, + "auxiliary_loss_mlp": 0.01033669, + "balance_loss_clip": 1.01256299, + "balance_loss_mlp": 1.02540147, + "epoch": 0.6203330728070703, + "flos": 18770508069120.0, + "grad_norm": 1.6238949683523594, + "language_loss": 0.7562483, + "learning_rate": 1.3307833211092768e-06, + "loss": 0.77803564, + "num_input_tokens_seen": 111197830, + "step": 5159, + "time_per_iteration": 2.6493759155273438 + }, + { + "auxiliary_loss_clip": 0.0117884, + "auxiliary_loss_mlp": 0.01032637, + "balance_loss_clip": 1.05734897, + "balance_loss_mlp": 1.02527618, + "epoch": 0.6204533156977093, + "flos": 20629782835200.0, + "grad_norm": 1.768616291932803, + "language_loss": 0.75086409, + "learning_rate": 1.3300493022782873e-06, + "loss": 0.77297884, + "num_input_tokens_seen": 111218400, + "step": 5160, + "time_per_iteration": 2.6944961547851562 + }, + { + "auxiliary_loss_clip": 0.0107106, + "auxiliary_loss_mlp": 0.00509964, + "balance_loss_clip": 0.8973453, + "balance_loss_mlp": 0.0, + "epoch": 0.6205735585883485, + "flos": 17348050598400.0, + "grad_norm": 1.7652863832761998, + "language_loss": 0.72522426, + "learning_rate": 1.3293153850782855e-06, + "loss": 0.74103457, + "num_input_tokens_seen": 111236720, + "step": 5161, + "time_per_iteration": 4.777070760726929 + }, + { + "auxiliary_loss_clip": 0.01093974, + "auxiliary_loss_mlp": 0.01036058, + "balance_loss_clip": 0.93529004, + "balance_loss_mlp": 1.02736211, + "epoch": 0.6206938014789876, + "flos": 22965017742720.0, + "grad_norm": 2.5068922912569924, + "language_loss": 0.70920926, + "learning_rate": 1.3285815696206069e-06, + "loss": 0.73050958, + "num_input_tokens_seen": 111258265, + "step": 5162, + "time_per_iteration": 2.8388407230377197 + }, + { + "auxiliary_loss_clip": 0.01104257, + "auxiliary_loss_mlp": 0.01023079, + "balance_loss_clip": 0.93500787, + "balance_loss_mlp": 1.01534843, + "epoch": 0.6208140443696266, + "flos": 23983192661760.0, + "grad_norm": 2.47174128460668, + "language_loss": 0.77391577, + "learning_rate": 1.32784785601657e-06, + "loss": 0.79518914, + "num_input_tokens_seen": 111277675, + "step": 5163, + "time_per_iteration": 2.7272021770477295 + }, + { + "auxiliary_loss_clip": 0.01127567, + "auxiliary_loss_mlp": 0.01026003, + "balance_loss_clip": 0.97353321, + "balance_loss_mlp": 1.01907408, + "epoch": 0.6209342872602658, + "flos": 35077291303680.0, + "grad_norm": 2.224540606627959, + "language_loss": 0.73767734, + "learning_rate": 1.3271142443774798e-06, + "loss": 0.75921309, + "num_input_tokens_seen": 111299910, + "step": 5164, + "time_per_iteration": 2.844346761703491 + }, + { + "auxiliary_loss_clip": 0.01124778, + "auxiliary_loss_mlp": 0.01022237, + "balance_loss_clip": 0.9776144, + "balance_loss_mlp": 1.01557338, + "epoch": 0.6210545301509048, + "flos": 26979327861120.0, + "grad_norm": 2.0852299974480215, + "language_loss": 0.81886005, + "learning_rate": 1.3263807348146228e-06, + "loss": 0.84033024, + "num_input_tokens_seen": 111319765, + "step": 5165, + "time_per_iteration": 2.778346061706543 + }, + { + "auxiliary_loss_clip": 0.01123703, + "auxiliary_loss_mlp": 0.01030133, + "balance_loss_clip": 0.97289908, + "balance_loss_mlp": 1.02218211, + "epoch": 0.6211747730415439, + "flos": 33618240852480.0, + "grad_norm": 3.0016748695804183, + "language_loss": 0.73318362, + "learning_rate": 1.3256473274392733e-06, + "loss": 0.75472194, + "num_input_tokens_seen": 111341110, + "step": 5166, + "time_per_iteration": 2.813225507736206 + }, + { + "auxiliary_loss_clip": 0.0117465, + "auxiliary_loss_mlp": 0.01027698, + "balance_loss_clip": 1.05406046, + "balance_loss_mlp": 1.02013421, + "epoch": 0.6212950159321831, + "flos": 34167099646080.0, + "grad_norm": 1.6827090547217771, + "language_loss": 0.70184159, + "learning_rate": 1.3249140223626873e-06, + "loss": 0.72386515, + "num_input_tokens_seen": 111362730, + "step": 5167, + "time_per_iteration": 2.6685569286346436 + }, + { + "auxiliary_loss_clip": 0.01147841, + "auxiliary_loss_mlp": 0.0102722, + "balance_loss_clip": 1.01584899, + "balance_loss_mlp": 1.01931012, + "epoch": 0.6214152588228221, + "flos": 27965758135680.0, + "grad_norm": 2.02002045701899, + "language_loss": 0.75312603, + "learning_rate": 1.3241808196961077e-06, + "loss": 0.77487671, + "num_input_tokens_seen": 111383855, + "step": 5168, + "time_per_iteration": 2.692883253097534 + }, + { + "auxiliary_loss_clip": 0.01114394, + "auxiliary_loss_mlp": 0.01022573, + "balance_loss_clip": 0.97310734, + "balance_loss_mlp": 1.01476479, + "epoch": 0.6215355017134612, + "flos": 20230204965120.0, + "grad_norm": 1.655751928580043, + "language_loss": 0.70951343, + "learning_rate": 1.3234477195507608e-06, + "loss": 0.73088306, + "num_input_tokens_seen": 111402685, + "step": 5169, + "time_per_iteration": 2.672100305557251 + }, + { + "auxiliary_loss_clip": 0.01105554, + "auxiliary_loss_mlp": 0.01022829, + "balance_loss_clip": 0.94101858, + "balance_loss_mlp": 1.01518774, + "epoch": 0.6216557446041003, + "flos": 41428129219200.0, + "grad_norm": 2.1740900670073677, + "language_loss": 0.62742984, + "learning_rate": 1.322714722037857e-06, + "loss": 0.64871371, + "num_input_tokens_seen": 111424130, + "step": 5170, + "time_per_iteration": 2.8603315353393555 + }, + { + "auxiliary_loss_clip": 0.01110774, + "auxiliary_loss_mlp": 0.01032159, + "balance_loss_clip": 0.93778884, + "balance_loss_mlp": 1.02455306, + "epoch": 0.6217759874947394, + "flos": 27928770105600.0, + "grad_norm": 2.4744878578491267, + "language_loss": 0.77124894, + "learning_rate": 1.321981827268591e-06, + "loss": 0.79267824, + "num_input_tokens_seen": 111444785, + "step": 5171, + "time_per_iteration": 2.7744178771972656 + }, + { + "auxiliary_loss_clip": 0.01131499, + "auxiliary_loss_mlp": 0.01027354, + "balance_loss_clip": 0.97687972, + "balance_loss_mlp": 1.02055573, + "epoch": 0.6218962303853784, + "flos": 21765673601280.0, + "grad_norm": 1.6649247201307094, + "language_loss": 0.81407195, + "learning_rate": 1.3212490353541426e-06, + "loss": 0.83566052, + "num_input_tokens_seen": 111467045, + "step": 5172, + "time_per_iteration": 3.6281661987304688 + }, + { + "auxiliary_loss_clip": 0.01175811, + "auxiliary_loss_mlp": 0.0102552, + "balance_loss_clip": 1.0532999, + "balance_loss_mlp": 1.01764023, + "epoch": 0.6220164732760175, + "flos": 21246260981760.0, + "grad_norm": 2.9987321142547847, + "language_loss": 0.8045783, + "learning_rate": 1.3205163464056762e-06, + "loss": 0.82659161, + "num_input_tokens_seen": 111483650, + "step": 5173, + "time_per_iteration": 2.640821933746338 + }, + { + "auxiliary_loss_clip": 0.01150519, + "auxiliary_loss_mlp": 0.010268, + "balance_loss_clip": 1.01588202, + "balance_loss_mlp": 1.01900947, + "epoch": 0.6221367161666567, + "flos": 26136360506880.0, + "grad_norm": 2.0342126531875966, + "language_loss": 0.72705436, + "learning_rate": 1.319783760534339e-06, + "loss": 0.74882752, + "num_input_tokens_seen": 111502895, + "step": 5174, + "time_per_iteration": 2.634085178375244 + }, + { + "auxiliary_loss_clip": 0.01153332, + "auxiliary_loss_mlp": 0.01029397, + "balance_loss_clip": 1.01674676, + "balance_loss_mlp": 1.02151132, + "epoch": 0.6222569590572957, + "flos": 16284196558080.0, + "grad_norm": 2.0249329921591355, + "language_loss": 0.75238955, + "learning_rate": 1.319051277851266e-06, + "loss": 0.77421683, + "num_input_tokens_seen": 111519180, + "step": 5175, + "time_per_iteration": 2.625645875930786 + }, + { + "auxiliary_loss_clip": 0.01154792, + "auxiliary_loss_mlp": 0.01020477, + "balance_loss_clip": 1.01607752, + "balance_loss_mlp": 1.01311016, + "epoch": 0.6223772019479348, + "flos": 18223840005120.0, + "grad_norm": 1.8689582622783882, + "language_loss": 0.84245408, + "learning_rate": 1.3183188984675716e-06, + "loss": 0.86420679, + "num_input_tokens_seen": 111537545, + "step": 5176, + "time_per_iteration": 2.6122562885284424 + }, + { + "auxiliary_loss_clip": 0.01130221, + "auxiliary_loss_mlp": 0.01030371, + "balance_loss_clip": 0.98082, + "balance_loss_mlp": 1.02278638, + "epoch": 0.6224974448385739, + "flos": 27489797994240.0, + "grad_norm": 2.5195245012141414, + "language_loss": 0.71653914, + "learning_rate": 1.3175866224943586e-06, + "loss": 0.73814499, + "num_input_tokens_seen": 111556265, + "step": 5177, + "time_per_iteration": 2.729553461074829 + }, + { + "auxiliary_loss_clip": 0.01133832, + "auxiliary_loss_mlp": 0.01030224, + "balance_loss_clip": 0.97853458, + "balance_loss_mlp": 1.02246308, + "epoch": 0.622617687729213, + "flos": 19791951125760.0, + "grad_norm": 8.318885369206892, + "language_loss": 0.73210949, + "learning_rate": 1.316854450042712e-06, + "loss": 0.75375003, + "num_input_tokens_seen": 111574205, + "step": 5178, + "time_per_iteration": 2.6788008213043213 + }, + { + "auxiliary_loss_clip": 0.01154753, + "auxiliary_loss_mlp": 0.0103071, + "balance_loss_clip": 1.01596045, + "balance_loss_mlp": 1.0232234, + "epoch": 0.622737930619852, + "flos": 23038886062080.0, + "grad_norm": 1.746195356318051, + "language_loss": 0.74376917, + "learning_rate": 1.3161223812237024e-06, + "loss": 0.76562381, + "num_input_tokens_seen": 111593560, + "step": 5179, + "time_per_iteration": 2.6309802532196045 + }, + { + "auxiliary_loss_clip": 0.01174579, + "auxiliary_loss_mlp": 0.01027305, + "balance_loss_clip": 1.05301487, + "balance_loss_mlp": 1.01915741, + "epoch": 0.6228581735104912, + "flos": 12634271959680.0, + "grad_norm": 2.536736225169227, + "language_loss": 0.85183859, + "learning_rate": 1.3153904161483842e-06, + "loss": 0.87385738, + "num_input_tokens_seen": 111608860, + "step": 5180, + "time_per_iteration": 2.5830068588256836 + }, + { + "auxiliary_loss_clip": 0.01102468, + "auxiliary_loss_mlp": 0.01031136, + "balance_loss_clip": 0.93719071, + "balance_loss_mlp": 1.02360225, + "epoch": 0.6229784164011303, + "flos": 23802813538560.0, + "grad_norm": 2.114829414980718, + "language_loss": 0.85605693, + "learning_rate": 1.3146585549277953e-06, + "loss": 0.87739289, + "num_input_tokens_seen": 111627500, + "step": 5181, + "time_per_iteration": 2.72890567779541 + }, + { + "auxiliary_loss_clip": 0.01138006, + "auxiliary_loss_mlp": 0.01033423, + "balance_loss_clip": 0.97835374, + "balance_loss_mlp": 1.02572846, + "epoch": 0.6230986592917693, + "flos": 22414219614720.0, + "grad_norm": 1.947063641040649, + "language_loss": 0.78367692, + "learning_rate": 1.3139267976729591e-06, + "loss": 0.80539119, + "num_input_tokens_seen": 111647690, + "step": 5182, + "time_per_iteration": 2.6947743892669678 + }, + { + "auxiliary_loss_clip": 0.01154733, + "auxiliary_loss_mlp": 0.01023904, + "balance_loss_clip": 1.01651263, + "balance_loss_mlp": 1.01666188, + "epoch": 0.6232189021824085, + "flos": 34528217028480.0, + "grad_norm": 1.932015587665872, + "language_loss": 0.71905297, + "learning_rate": 1.3131951444948815e-06, + "loss": 0.74083936, + "num_input_tokens_seen": 111667090, + "step": 5183, + "time_per_iteration": 2.760577917098999 + }, + { + "auxiliary_loss_clip": 0.01133322, + "auxiliary_loss_mlp": 0.01023726, + "balance_loss_clip": 0.98071992, + "balance_loss_mlp": 1.01555467, + "epoch": 0.6233391450730476, + "flos": 22237000888320.0, + "grad_norm": 2.063791226401665, + "language_loss": 0.76232541, + "learning_rate": 1.3124635955045546e-06, + "loss": 0.78389591, + "num_input_tokens_seen": 111686905, + "step": 5184, + "time_per_iteration": 2.733905076980591 + }, + { + "auxiliary_loss_clip": 0.0107161, + "auxiliary_loss_mlp": 0.00509464, + "balance_loss_clip": 0.8969714, + "balance_loss_mlp": 0.0, + "epoch": 0.6234593879636866, + "flos": 20332693445760.0, + "grad_norm": 1.8416301917677569, + "language_loss": 0.83933556, + "learning_rate": 1.3117321508129537e-06, + "loss": 0.85514641, + "num_input_tokens_seen": 111704985, + "step": 5185, + "time_per_iteration": 2.7508225440979004 + }, + { + "auxiliary_loss_clip": 0.01131212, + "auxiliary_loss_mlp": 0.01026719, + "balance_loss_clip": 0.97949606, + "balance_loss_mlp": 1.01941776, + "epoch": 0.6235796308543258, + "flos": 20664903358080.0, + "grad_norm": 1.48877898589379, + "language_loss": 0.76480937, + "learning_rate": 1.3110008105310388e-06, + "loss": 0.7863887, + "num_input_tokens_seen": 111724805, + "step": 5186, + "time_per_iteration": 3.586534261703491 + }, + { + "auxiliary_loss_clip": 0.01175576, + "auxiliary_loss_mlp": 0.01030918, + "balance_loss_clip": 1.05263972, + "balance_loss_mlp": 1.02299094, + "epoch": 0.6236998737449648, + "flos": 26618641441920.0, + "grad_norm": 2.5129371616161564, + "language_loss": 0.77870846, + "learning_rate": 1.3102695747697526e-06, + "loss": 0.80077338, + "num_input_tokens_seen": 111747675, + "step": 5187, + "time_per_iteration": 4.476925611495972 + }, + { + "auxiliary_loss_clip": 0.01058246, + "auxiliary_loss_mlp": 0.01034951, + "balance_loss_clip": 0.86348361, + "balance_loss_mlp": 1.02739286, + "epoch": 0.6238201166356039, + "flos": 12674599954560.0, + "grad_norm": 2.892913225387728, + "language_loss": 0.90736115, + "learning_rate": 1.3095384436400237e-06, + "loss": 0.92829317, + "num_input_tokens_seen": 111759205, + "step": 5188, + "time_per_iteration": 2.783332347869873 + }, + { + "auxiliary_loss_clip": 0.01133745, + "auxiliary_loss_mlp": 0.0102726, + "balance_loss_clip": 0.97624224, + "balance_loss_mlp": 1.01979756, + "epoch": 0.623940359526243, + "flos": 10452160730880.0, + "grad_norm": 5.058625124466091, + "language_loss": 0.82365453, + "learning_rate": 1.3088074172527633e-06, + "loss": 0.84526455, + "num_input_tokens_seen": 111776335, + "step": 5189, + "time_per_iteration": 2.689924955368042 + }, + { + "auxiliary_loss_clip": 0.0113188, + "auxiliary_loss_mlp": 0.01027425, + "balance_loss_clip": 0.97565913, + "balance_loss_mlp": 1.01979578, + "epoch": 0.6240606024168821, + "flos": 29059525226880.0, + "grad_norm": 2.0074414228556745, + "language_loss": 0.71143079, + "learning_rate": 1.3080764957188684e-06, + "loss": 0.73302388, + "num_input_tokens_seen": 111796580, + "step": 5190, + "time_per_iteration": 2.759399652481079 + }, + { + "auxiliary_loss_clip": 0.01084782, + "auxiliary_loss_mlp": 0.01026118, + "balance_loss_clip": 0.89819193, + "balance_loss_mlp": 1.01833367, + "epoch": 0.6241808453075212, + "flos": 22018089450240.0, + "grad_norm": 2.3753140007939475, + "language_loss": 0.70491779, + "learning_rate": 1.3073456791492192e-06, + "loss": 0.72602677, + "num_input_tokens_seen": 111816290, + "step": 5191, + "time_per_iteration": 2.755647897720337 + }, + { + "auxiliary_loss_clip": 0.01127714, + "auxiliary_loss_mlp": 0.010235, + "balance_loss_clip": 0.97446305, + "balance_loss_mlp": 1.01603127, + "epoch": 0.6243010881981603, + "flos": 21138708683520.0, + "grad_norm": 2.51393855335458, + "language_loss": 0.78137785, + "learning_rate": 1.3066149676546801e-06, + "loss": 0.80289, + "num_input_tokens_seen": 111834470, + "step": 5192, + "time_per_iteration": 2.8077430725097656 + }, + { + "auxiliary_loss_clip": 0.01129226, + "auxiliary_loss_mlp": 0.01025295, + "balance_loss_clip": 0.9822948, + "balance_loss_mlp": 1.01742101, + "epoch": 0.6244213310887994, + "flos": 22344948236160.0, + "grad_norm": 2.5466119302307346, + "language_loss": 0.65916497, + "learning_rate": 1.3058843613460985e-06, + "loss": 0.68071008, + "num_input_tokens_seen": 111852410, + "step": 5193, + "time_per_iteration": 2.8129730224609375 + }, + { + "auxiliary_loss_clip": 0.01111869, + "auxiliary_loss_mlp": 0.0102719, + "balance_loss_clip": 0.93746519, + "balance_loss_mlp": 1.0190897, + "epoch": 0.6245415739794384, + "flos": 15231978524160.0, + "grad_norm": 1.9093239572338059, + "language_loss": 0.74162555, + "learning_rate": 1.3051538603343075e-06, + "loss": 0.7630161, + "num_input_tokens_seen": 111870340, + "step": 5194, + "time_per_iteration": 2.726787567138672 + }, + { + "auxiliary_loss_clip": 0.01152264, + "auxiliary_loss_mlp": 0.01028583, + "balance_loss_clip": 1.01812708, + "balance_loss_mlp": 1.02107835, + "epoch": 0.6246618168700776, + "flos": 18879891960960.0, + "grad_norm": 2.1474535063242812, + "language_loss": 0.67914385, + "learning_rate": 1.3044234647301235e-06, + "loss": 0.70095229, + "num_input_tokens_seen": 111888365, + "step": 5195, + "time_per_iteration": 2.637169599533081 + }, + { + "auxiliary_loss_clip": 0.01146998, + "auxiliary_loss_mlp": 0.0102236, + "balance_loss_clip": 1.015625, + "balance_loss_mlp": 1.01536584, + "epoch": 0.6247820597607167, + "flos": 14319201087360.0, + "grad_norm": 1.83822537211589, + "language_loss": 0.72713709, + "learning_rate": 1.303693174644347e-06, + "loss": 0.74883068, + "num_input_tokens_seen": 111905840, + "step": 5196, + "time_per_iteration": 2.6380608081817627 + }, + { + "auxiliary_loss_clip": 0.0112297, + "auxiliary_loss_mlp": 0.01024026, + "balance_loss_clip": 0.97640669, + "balance_loss_mlp": 1.01629519, + "epoch": 0.6249023026513557, + "flos": 22637979388800.0, + "grad_norm": 1.9673725565272229, + "language_loss": 0.80759394, + "learning_rate": 1.3029629901877625e-06, + "loss": 0.82906389, + "num_input_tokens_seen": 111925215, + "step": 5197, + "time_per_iteration": 2.7007851600646973 + }, + { + "auxiliary_loss_clip": 0.01160417, + "auxiliary_loss_mlp": 0.01029859, + "balance_loss_clip": 1.01859748, + "balance_loss_mlp": 1.02206302, + "epoch": 0.6250225455419949, + "flos": 20266690204800.0, + "grad_norm": 2.329897651439332, + "language_loss": 0.77116752, + "learning_rate": 1.3022329114711376e-06, + "loss": 0.79307026, + "num_input_tokens_seen": 111943925, + "step": 5198, + "time_per_iteration": 3.563321113586426 + }, + { + "auxiliary_loss_clip": 0.01127074, + "auxiliary_loss_mlp": 0.01028076, + "balance_loss_clip": 0.97958434, + "balance_loss_mlp": 1.02060783, + "epoch": 0.6251427884326339, + "flos": 23437853400960.0, + "grad_norm": 2.2401505217254654, + "language_loss": 0.69643974, + "learning_rate": 1.3015029386052256e-06, + "loss": 0.71799123, + "num_input_tokens_seen": 111964095, + "step": 5199, + "time_per_iteration": 2.7775981426239014 + }, + { + "auxiliary_loss_clip": 0.0111255, + "auxiliary_loss_mlp": 0.01027729, + "balance_loss_clip": 0.93763864, + "balance_loss_mlp": 1.01983714, + "epoch": 0.625263031323273, + "flos": 31723055464320.0, + "grad_norm": 2.1234474200883633, + "language_loss": 0.73443913, + "learning_rate": 1.3007730717007622e-06, + "loss": 0.75584191, + "num_input_tokens_seen": 111984910, + "step": 5200, + "time_per_iteration": 2.7893693447113037 + }, + { + "auxiliary_loss_clip": 0.01176184, + "auxiliary_loss_mlp": 0.01031312, + "balance_loss_clip": 1.0540154, + "balance_loss_mlp": 1.02331913, + "epoch": 0.6253832742139122, + "flos": 24134341092480.0, + "grad_norm": 1.6557271120708563, + "language_loss": 0.75248563, + "learning_rate": 1.3000433108684676e-06, + "loss": 0.77456063, + "num_input_tokens_seen": 112005410, + "step": 5201, + "time_per_iteration": 2.659496307373047 + }, + { + "auxiliary_loss_clip": 0.0114841, + "auxiliary_loss_mlp": 0.01026899, + "balance_loss_clip": 1.0154798, + "balance_loss_mlp": 1.01922178, + "epoch": 0.6255035171045512, + "flos": 27668812400640.0, + "grad_norm": 2.0957598508550817, + "language_loss": 0.80572331, + "learning_rate": 1.2993136562190467e-06, + "loss": 0.82747644, + "num_input_tokens_seen": 112024530, + "step": 5202, + "time_per_iteration": 2.65592098236084 + }, + { + "auxiliary_loss_clip": 0.01130007, + "auxiliary_loss_mlp": 0.0103435, + "balance_loss_clip": 0.97580075, + "balance_loss_mlp": 1.02657461, + "epoch": 0.6256237599951903, + "flos": 20227798753920.0, + "grad_norm": 1.7490464195741822, + "language_loss": 0.70463955, + "learning_rate": 1.2985841078631871e-06, + "loss": 0.72628307, + "num_input_tokens_seen": 112043850, + "step": 5203, + "time_per_iteration": 2.7025139331817627 + }, + { + "auxiliary_loss_clip": 0.01053353, + "auxiliary_loss_mlp": 0.01025129, + "balance_loss_clip": 0.85684597, + "balance_loss_mlp": 1.01757121, + "epoch": 0.6257440028858293, + "flos": 24170574936960.0, + "grad_norm": 1.9748750835074516, + "language_loss": 0.78422832, + "learning_rate": 1.2978546659115608e-06, + "loss": 0.80501318, + "num_input_tokens_seen": 112061930, + "step": 5204, + "time_per_iteration": 2.799456834793091 + }, + { + "auxiliary_loss_clip": 0.01132225, + "auxiliary_loss_mlp": 0.01030044, + "balance_loss_clip": 0.97921383, + "balance_loss_mlp": 1.02296865, + "epoch": 0.6258642457764685, + "flos": 15851940289920.0, + "grad_norm": 1.8823734155119145, + "language_loss": 0.85467845, + "learning_rate": 1.2971253304748228e-06, + "loss": 0.87630111, + "num_input_tokens_seen": 112079645, + "step": 5205, + "time_per_iteration": 2.7465121746063232 + }, + { + "auxiliary_loss_clip": 0.01157701, + "auxiliary_loss_mlp": 0.01028648, + "balance_loss_clip": 1.01965332, + "balance_loss_mlp": 1.02102458, + "epoch": 0.6259844886671075, + "flos": 11911354836480.0, + "grad_norm": 1.5758750648822402, + "language_loss": 0.74799144, + "learning_rate": 1.296396101663614e-06, + "loss": 0.7698549, + "num_input_tokens_seen": 112096205, + "step": 5206, + "time_per_iteration": 2.6142423152923584 + }, + { + "auxiliary_loss_clip": 0.01156904, + "auxiliary_loss_mlp": 0.01027351, + "balance_loss_clip": 1.01778316, + "balance_loss_mlp": 1.01959634, + "epoch": 0.6261047315577466, + "flos": 15887958652800.0, + "grad_norm": 2.406784595265743, + "language_loss": 0.83885181, + "learning_rate": 1.2956669795885565e-06, + "loss": 0.86069441, + "num_input_tokens_seen": 112112835, + "step": 5207, + "time_per_iteration": 2.636347770690918 + }, + { + "auxiliary_loss_clip": 0.01098019, + "auxiliary_loss_mlp": 0.01035316, + "balance_loss_clip": 0.9382062, + "balance_loss_mlp": 1.02707303, + "epoch": 0.6262249744483858, + "flos": 31248926916480.0, + "grad_norm": 1.833338164640424, + "language_loss": 0.6800468, + "learning_rate": 1.294937964360259e-06, + "loss": 0.70138013, + "num_input_tokens_seen": 112133105, + "step": 5208, + "time_per_iteration": 2.81697154045105 + }, + { + "auxiliary_loss_clip": 0.01134934, + "auxiliary_loss_mlp": 0.01030022, + "balance_loss_clip": 0.97719157, + "balance_loss_mlp": 1.02209425, + "epoch": 0.6263452173390248, + "flos": 27198598435200.0, + "grad_norm": 2.303824772267209, + "language_loss": 0.71314156, + "learning_rate": 1.2942090560893108e-06, + "loss": 0.73479104, + "num_input_tokens_seen": 112152510, + "step": 5209, + "time_per_iteration": 2.7471656799316406 + }, + { + "auxiliary_loss_clip": 0.01177327, + "auxiliary_loss_mlp": 0.0103117, + "balance_loss_clip": 1.05644464, + "balance_loss_mlp": 1.023839, + "epoch": 0.6264654602296639, + "flos": 37342069683840.0, + "grad_norm": 1.675834069006174, + "language_loss": 0.6010952, + "learning_rate": 1.2934802548862882e-06, + "loss": 0.62318021, + "num_input_tokens_seen": 112175295, + "step": 5210, + "time_per_iteration": 2.7517950534820557 + }, + { + "auxiliary_loss_clip": 0.01124945, + "auxiliary_loss_mlp": 0.01025943, + "balance_loss_clip": 0.97582024, + "balance_loss_mlp": 1.01792026, + "epoch": 0.626585703120303, + "flos": 14756952136320.0, + "grad_norm": 1.9538291359313076, + "language_loss": 0.82583725, + "learning_rate": 1.292751560861749e-06, + "loss": 0.84734619, + "num_input_tokens_seen": 112190200, + "step": 5211, + "time_per_iteration": 2.7001633644104004 + }, + { + "auxiliary_loss_clip": 0.01178269, + "auxiliary_loss_mlp": 0.01027111, + "balance_loss_clip": 1.05528009, + "balance_loss_mlp": 1.01924324, + "epoch": 0.6267059460109421, + "flos": 22347318533760.0, + "grad_norm": 1.9541298596116803, + "language_loss": 0.79329818, + "learning_rate": 1.2920229741262354e-06, + "loss": 0.81535196, + "num_input_tokens_seen": 112208205, + "step": 5212, + "time_per_iteration": 3.582151412963867 + }, + { + "auxiliary_loss_clip": 0.01128723, + "auxiliary_loss_mlp": 0.01023731, + "balance_loss_clip": 0.97711003, + "balance_loss_mlp": 1.01653028, + "epoch": 0.6268261889015811, + "flos": 17748813617280.0, + "grad_norm": 2.371465323062644, + "language_loss": 0.75289047, + "learning_rate": 1.2912944947902739e-06, + "loss": 0.77441508, + "num_input_tokens_seen": 112224690, + "step": 5213, + "time_per_iteration": 4.620922565460205 + }, + { + "auxiliary_loss_clip": 0.01135078, + "auxiliary_loss_mlp": 0.01025055, + "balance_loss_clip": 0.97835207, + "balance_loss_mlp": 1.01759887, + "epoch": 0.6269464317922203, + "flos": 32846484211200.0, + "grad_norm": 2.647032697608616, + "language_loss": 0.72129607, + "learning_rate": 1.2905661229643742e-06, + "loss": 0.74289739, + "num_input_tokens_seen": 112244450, + "step": 5214, + "time_per_iteration": 2.811789035797119 + }, + { + "auxiliary_loss_clip": 0.01175517, + "auxiliary_loss_mlp": 0.01029496, + "balance_loss_clip": 1.05377948, + "balance_loss_mlp": 1.02151465, + "epoch": 0.6270666746828594, + "flos": 17929192740480.0, + "grad_norm": 1.9475086340597731, + "language_loss": 0.84461319, + "learning_rate": 1.2898378587590299e-06, + "loss": 0.86666334, + "num_input_tokens_seen": 112261050, + "step": 5215, + "time_per_iteration": 2.677187204360962 + }, + { + "auxiliary_loss_clip": 0.01147508, + "auxiliary_loss_mlp": 0.01026378, + "balance_loss_clip": 1.01526797, + "balance_loss_mlp": 1.01885331, + "epoch": 0.6271869175734984, + "flos": 17457326749440.0, + "grad_norm": 2.9272060073419612, + "language_loss": 0.87756503, + "learning_rate": 1.2891097022847173e-06, + "loss": 0.89930391, + "num_input_tokens_seen": 112278395, + "step": 5216, + "time_per_iteration": 2.6752538681030273 + }, + { + "auxiliary_loss_clip": 0.01128971, + "auxiliary_loss_mlp": 0.01030587, + "balance_loss_clip": 0.97746539, + "balance_loss_mlp": 1.02160418, + "epoch": 0.6273071604641376, + "flos": 26868615166080.0, + "grad_norm": 1.9727751268746303, + "language_loss": 0.66621315, + "learning_rate": 1.2883816536518978e-06, + "loss": 0.68780869, + "num_input_tokens_seen": 112299535, + "step": 5217, + "time_per_iteration": 2.773341417312622 + }, + { + "auxiliary_loss_clip": 0.01149927, + "auxiliary_loss_mlp": 0.01028058, + "balance_loss_clip": 1.01398265, + "balance_loss_mlp": 1.02052391, + "epoch": 0.6274274033547766, + "flos": 26062384446720.0, + "grad_norm": 2.280280990753688, + "language_loss": 0.82018536, + "learning_rate": 1.2876537129710155e-06, + "loss": 0.84196526, + "num_input_tokens_seen": 112317265, + "step": 5218, + "time_per_iteration": 2.6800122261047363 + }, + { + "auxiliary_loss_clip": 0.01129687, + "auxiliary_loss_mlp": 0.0103043, + "balance_loss_clip": 0.98325765, + "balance_loss_mlp": 1.02190661, + "epoch": 0.6275476462454157, + "flos": 20266259241600.0, + "grad_norm": 1.9911331983696743, + "language_loss": 0.75285685, + "learning_rate": 1.286925880352499e-06, + "loss": 0.77445805, + "num_input_tokens_seen": 112336125, + "step": 5219, + "time_per_iteration": 2.6678059101104736 + }, + { + "auxiliary_loss_clip": 0.01128246, + "auxiliary_loss_mlp": 0.0102731, + "balance_loss_clip": 0.97846413, + "balance_loss_mlp": 1.01971006, + "epoch": 0.6276678891360549, + "flos": 26320402817280.0, + "grad_norm": 1.650431381905528, + "language_loss": 0.71295792, + "learning_rate": 1.2861981559067592e-06, + "loss": 0.73451346, + "num_input_tokens_seen": 112356730, + "step": 5220, + "time_per_iteration": 2.7434427738189697 + }, + { + "auxiliary_loss_clip": 0.01073291, + "auxiliary_loss_mlp": 0.01024028, + "balance_loss_clip": 0.89863992, + "balance_loss_mlp": 1.01626158, + "epoch": 0.6277881320266939, + "flos": 13912512324480.0, + "grad_norm": 1.9713698588800148, + "language_loss": 0.80376953, + "learning_rate": 1.2854705397441917e-06, + "loss": 0.82474273, + "num_input_tokens_seen": 112372270, + "step": 5221, + "time_per_iteration": 2.702085256576538 + }, + { + "auxiliary_loss_clip": 0.01100918, + "auxiliary_loss_mlp": 0.0102618, + "balance_loss_clip": 0.93743694, + "balance_loss_mlp": 1.0186522, + "epoch": 0.627908374917333, + "flos": 27048922462080.0, + "grad_norm": 2.171504759658396, + "language_loss": 0.78119266, + "learning_rate": 1.2847430319751747e-06, + "loss": 0.80246365, + "num_input_tokens_seen": 112390365, + "step": 5222, + "time_per_iteration": 2.8312606811523438 + }, + { + "auxiliary_loss_clip": 0.01149405, + "auxiliary_loss_mlp": 0.01032244, + "balance_loss_clip": 1.01798797, + "balance_loss_mlp": 1.02435231, + "epoch": 0.6280286178079721, + "flos": 23769201386880.0, + "grad_norm": 2.2524997486863385, + "language_loss": 0.6820032, + "learning_rate": 1.2840156327100712e-06, + "loss": 0.70381975, + "num_input_tokens_seen": 112407490, + "step": 5223, + "time_per_iteration": 2.6666769981384277 + }, + { + "auxiliary_loss_clip": 0.01175589, + "auxiliary_loss_mlp": 0.01029717, + "balance_loss_clip": 1.05559516, + "balance_loss_mlp": 1.0213635, + "epoch": 0.6281488606986112, + "flos": 26359150613760.0, + "grad_norm": 1.6163226593946642, + "language_loss": 0.72574103, + "learning_rate": 1.2832883420592272e-06, + "loss": 0.74779415, + "num_input_tokens_seen": 112426385, + "step": 5224, + "time_per_iteration": 3.62642240524292 + }, + { + "auxiliary_loss_clip": 0.01123928, + "auxiliary_loss_mlp": 0.01026149, + "balance_loss_clip": 0.97677565, + "balance_loss_mlp": 1.01832294, + "epoch": 0.6282691035892503, + "flos": 36137194848000.0, + "grad_norm": 1.9316888752885526, + "language_loss": 0.63924861, + "learning_rate": 1.282561160132972e-06, + "loss": 0.66074944, + "num_input_tokens_seen": 112446905, + "step": 5225, + "time_per_iteration": 2.8454675674438477 + }, + { + "auxiliary_loss_clip": 0.01131926, + "auxiliary_loss_mlp": 0.01029837, + "balance_loss_clip": 0.97286463, + "balance_loss_mlp": 1.02149868, + "epoch": 0.6283893464798894, + "flos": 26537231266560.0, + "grad_norm": 1.6811103423300628, + "language_loss": 0.80811626, + "learning_rate": 1.2818340870416186e-06, + "loss": 0.82973391, + "num_input_tokens_seen": 112468040, + "step": 5226, + "time_per_iteration": 2.699143171310425 + }, + { + "auxiliary_loss_clip": 0.01111222, + "auxiliary_loss_mlp": 0.01027574, + "balance_loss_clip": 0.93563318, + "balance_loss_mlp": 1.01978385, + "epoch": 0.6285095893705285, + "flos": 22237216369920.0, + "grad_norm": 1.876496269678031, + "language_loss": 0.76002395, + "learning_rate": 1.2811071228954626e-06, + "loss": 0.78141195, + "num_input_tokens_seen": 112486675, + "step": 5227, + "time_per_iteration": 2.7794573307037354 + }, + { + "auxiliary_loss_clip": 0.01128929, + "auxiliary_loss_mlp": 0.01029585, + "balance_loss_clip": 0.97916269, + "balance_loss_mlp": 1.02188993, + "epoch": 0.6286298322611675, + "flos": 26542259170560.0, + "grad_norm": 1.9806002179794033, + "language_loss": 0.80536246, + "learning_rate": 1.2803802678047846e-06, + "loss": 0.82694757, + "num_input_tokens_seen": 112506825, + "step": 5228, + "time_per_iteration": 2.743443727493286 + }, + { + "auxiliary_loss_clip": 0.01132106, + "auxiliary_loss_mlp": 0.01032369, + "balance_loss_clip": 0.9783349, + "balance_loss_mlp": 1.02364254, + "epoch": 0.6287500751518067, + "flos": 21795227516160.0, + "grad_norm": 3.4483303609061324, + "language_loss": 0.74350512, + "learning_rate": 1.279653521879848e-06, + "loss": 0.76514983, + "num_input_tokens_seen": 112526890, + "step": 5229, + "time_per_iteration": 2.700155735015869 + }, + { + "auxiliary_loss_clip": 0.01024879, + "auxiliary_loss_mlp": 0.01029525, + "balance_loss_clip": 0.82086086, + "balance_loss_mlp": 1.02191949, + "epoch": 0.6288703180424458, + "flos": 20009605587840.0, + "grad_norm": 2.3017580193919307, + "language_loss": 0.8418448, + "learning_rate": 1.2789268852308997e-06, + "loss": 0.86238885, + "num_input_tokens_seen": 112542100, + "step": 5230, + "time_per_iteration": 2.8265585899353027 + }, + { + "auxiliary_loss_clip": 0.01145769, + "auxiliary_loss_mlp": 0.01024653, + "balance_loss_clip": 1.01518512, + "balance_loss_mlp": 1.0170238, + "epoch": 0.6289905609330848, + "flos": 22124923476480.0, + "grad_norm": 2.1184481871853835, + "language_loss": 0.70909941, + "learning_rate": 1.2782003579681688e-06, + "loss": 0.73080361, + "num_input_tokens_seen": 112561630, + "step": 5231, + "time_per_iteration": 2.6875076293945312 + }, + { + "auxiliary_loss_clip": 0.01177762, + "auxiliary_loss_mlp": 0.01028568, + "balance_loss_clip": 1.05453753, + "balance_loss_mlp": 1.02068806, + "epoch": 0.629110803823724, + "flos": 25518481729920.0, + "grad_norm": 1.6665654120247722, + "language_loss": 0.74203587, + "learning_rate": 1.2774739402018701e-06, + "loss": 0.76409912, + "num_input_tokens_seen": 112582465, + "step": 5232, + "time_per_iteration": 2.6245322227478027 + }, + { + "auxiliary_loss_clip": 0.01151222, + "auxiliary_loss_mlp": 0.01027999, + "balance_loss_clip": 1.01827598, + "balance_loss_mlp": 1.01959491, + "epoch": 0.629231046714363, + "flos": 20886616056960.0, + "grad_norm": 2.0145341864262947, + "language_loss": 0.7292524, + "learning_rate": 1.2767476320422002e-06, + "loss": 0.75104463, + "num_input_tokens_seen": 112602390, + "step": 5233, + "time_per_iteration": 2.6555259227752686 + }, + { + "auxiliary_loss_clip": 0.01019628, + "auxiliary_loss_mlp": 0.010015, + "balance_loss_clip": 0.90989482, + "balance_loss_mlp": 1.0003618, + "epoch": 0.6293512896050021, + "flos": 65050027908480.0, + "grad_norm": 0.6743417320389223, + "language_loss": 0.57150912, + "learning_rate": 1.2760214335993392e-06, + "loss": 0.5917204, + "num_input_tokens_seen": 112669035, + "step": 5234, + "time_per_iteration": 3.358428478240967 + }, + { + "auxiliary_loss_clip": 0.01143522, + "auxiliary_loss_mlp": 0.01025556, + "balance_loss_clip": 1.01216066, + "balance_loss_mlp": 1.0179621, + "epoch": 0.6294715324956413, + "flos": 34677857088000.0, + "grad_norm": 1.8834473490870138, + "language_loss": 0.58965111, + "learning_rate": 1.2752953449834514e-06, + "loss": 0.61134189, + "num_input_tokens_seen": 112691485, + "step": 5235, + "time_per_iteration": 2.7479279041290283 + }, + { + "auxiliary_loss_clip": 0.01176767, + "auxiliary_loss_mlp": 0.0103033, + "balance_loss_clip": 1.05554008, + "balance_loss_mlp": 1.02256954, + "epoch": 0.6295917753862803, + "flos": 22784207656320.0, + "grad_norm": 1.8582586418808749, + "language_loss": 0.80392659, + "learning_rate": 1.2745693663046836e-06, + "loss": 0.82599759, + "num_input_tokens_seen": 112710555, + "step": 5236, + "time_per_iteration": 2.6487555503845215 + }, + { + "auxiliary_loss_clip": 0.01148402, + "auxiliary_loss_mlp": 0.01024811, + "balance_loss_clip": 1.01547384, + "balance_loss_mlp": 1.01722324, + "epoch": 0.6297120182769194, + "flos": 20850454039680.0, + "grad_norm": 2.545173385852473, + "language_loss": 0.8069855, + "learning_rate": 1.2738434976731662e-06, + "loss": 0.82871759, + "num_input_tokens_seen": 112728740, + "step": 5237, + "time_per_iteration": 2.6121978759765625 + }, + { + "auxiliary_loss_clip": 0.01130768, + "auxiliary_loss_mlp": 0.01029686, + "balance_loss_clip": 0.97936898, + "balance_loss_mlp": 1.02144229, + "epoch": 0.6298322611675584, + "flos": 19497662997120.0, + "grad_norm": 1.837698655767334, + "language_loss": 0.75127399, + "learning_rate": 1.2731177391990125e-06, + "loss": 0.77287853, + "num_input_tokens_seen": 112748665, + "step": 5238, + "time_per_iteration": 3.6214239597320557 + }, + { + "auxiliary_loss_clip": 0.01129421, + "auxiliary_loss_mlp": 0.01031287, + "balance_loss_clip": 0.97617751, + "balance_loss_mlp": 1.02320492, + "epoch": 0.6299525040581976, + "flos": 12604466649600.0, + "grad_norm": 1.912507188155095, + "language_loss": 0.81863785, + "learning_rate": 1.2723920909923203e-06, + "loss": 0.84024489, + "num_input_tokens_seen": 112764410, + "step": 5239, + "time_per_iteration": 3.553302764892578 + }, + { + "auxiliary_loss_clip": 0.01073143, + "auxiliary_loss_mlp": 0.01000857, + "balance_loss_clip": 1.0190599, + "balance_loss_mlp": 0.99973011, + "epoch": 0.6300727469488366, + "flos": 57725685636480.0, + "grad_norm": 0.8436579794495265, + "language_loss": 0.60469693, + "learning_rate": 1.2716665531631688e-06, + "loss": 0.6254369, + "num_input_tokens_seen": 112818695, + "step": 5240, + "time_per_iteration": 4.094912767410278 + }, + { + "auxiliary_loss_clip": 0.01158245, + "auxiliary_loss_mlp": 0.01028183, + "balance_loss_clip": 1.01562119, + "balance_loss_mlp": 1.02055407, + "epoch": 0.6301929898394757, + "flos": 22527302607360.0, + "grad_norm": 1.8623957842855101, + "language_loss": 0.77390659, + "learning_rate": 1.270941125821623e-06, + "loss": 0.79577088, + "num_input_tokens_seen": 112839120, + "step": 5241, + "time_per_iteration": 2.6876299381256104 + }, + { + "auxiliary_loss_clip": 0.0114695, + "auxiliary_loss_mlp": 0.01024981, + "balance_loss_clip": 1.01252949, + "balance_loss_mlp": 1.01770318, + "epoch": 0.6303132327301149, + "flos": 28293550675200.0, + "grad_norm": 1.6512685231125703, + "language_loss": 0.75796854, + "learning_rate": 1.2702158090777278e-06, + "loss": 0.77968788, + "num_input_tokens_seen": 112860210, + "step": 5242, + "time_per_iteration": 2.7442190647125244 + }, + { + "auxiliary_loss_clip": 0.0110192, + "auxiliary_loss_mlp": 0.01030222, + "balance_loss_clip": 0.93647981, + "balance_loss_mlp": 1.02329898, + "epoch": 0.6304334756207539, + "flos": 25264521596160.0, + "grad_norm": 1.8539199094599985, + "language_loss": 0.74822909, + "learning_rate": 1.2694906030415148e-06, + "loss": 0.7695505, + "num_input_tokens_seen": 112877955, + "step": 5243, + "time_per_iteration": 2.7495839595794678 + }, + { + "auxiliary_loss_clip": 0.01137843, + "auxiliary_loss_mlp": 0.01027137, + "balance_loss_clip": 0.97855949, + "balance_loss_mlp": 1.0182085, + "epoch": 0.630553718511393, + "flos": 18033548728320.0, + "grad_norm": 2.87537302000776, + "language_loss": 0.82103074, + "learning_rate": 1.2687655078229958e-06, + "loss": 0.84268051, + "num_input_tokens_seen": 112892285, + "step": 5244, + "time_per_iteration": 2.6766324043273926 + }, + { + "auxiliary_loss_clip": 0.01128225, + "auxiliary_loss_mlp": 0.01028615, + "balance_loss_clip": 0.97993553, + "balance_loss_mlp": 1.02072358, + "epoch": 0.6306739614020321, + "flos": 27304103658240.0, + "grad_norm": 3.0673937420832904, + "language_loss": 0.6952216, + "learning_rate": 1.2680405235321678e-06, + "loss": 0.71679002, + "num_input_tokens_seen": 112913620, + "step": 5245, + "time_per_iteration": 2.737061023712158 + }, + { + "auxiliary_loss_clip": 0.01133746, + "auxiliary_loss_mlp": 0.00510343, + "balance_loss_clip": 0.98175776, + "balance_loss_mlp": 0.0, + "epoch": 0.6307942042926712, + "flos": 15341434243200.0, + "grad_norm": 2.747950706304742, + "language_loss": 0.79108673, + "learning_rate": 1.267315650279011e-06, + "loss": 0.8075276, + "num_input_tokens_seen": 112932090, + "step": 5246, + "time_per_iteration": 2.740020751953125 + }, + { + "auxiliary_loss_clip": 0.01099669, + "auxiliary_loss_mlp": 0.010324, + "balance_loss_clip": 0.94060403, + "balance_loss_mlp": 1.02418673, + "epoch": 0.6309144471833102, + "flos": 19606400444160.0, + "grad_norm": 1.8346448335255139, + "language_loss": 0.74094707, + "learning_rate": 1.2665908881734874e-06, + "loss": 0.76226777, + "num_input_tokens_seen": 112950925, + "step": 5247, + "time_per_iteration": 2.6924071311950684 + }, + { + "auxiliary_loss_clip": 0.0115139, + "auxiliary_loss_mlp": 0.01025188, + "balance_loss_clip": 1.0159657, + "balance_loss_mlp": 1.01754642, + "epoch": 0.6310346900739494, + "flos": 17493345112320.0, + "grad_norm": 2.1210423182585902, + "language_loss": 0.85303116, + "learning_rate": 1.2658662373255432e-06, + "loss": 0.87479699, + "num_input_tokens_seen": 112969315, + "step": 5248, + "time_per_iteration": 2.6757638454437256 + }, + { + "auxiliary_loss_clip": 0.01031315, + "auxiliary_loss_mlp": 0.00999218, + "balance_loss_clip": 0.94271106, + "balance_loss_mlp": 0.99810964, + "epoch": 0.6311549329645885, + "flos": 55070164131840.0, + "grad_norm": 0.7264955719765097, + "language_loss": 0.52245039, + "learning_rate": 1.2651416978451063e-06, + "loss": 0.54275572, + "num_input_tokens_seen": 113034700, + "step": 5249, + "time_per_iteration": 3.3579607009887695 + }, + { + "auxiliary_loss_clip": 0.01181406, + "auxiliary_loss_mlp": 0.01033824, + "balance_loss_clip": 1.057868, + "balance_loss_mlp": 1.02562225, + "epoch": 0.6312751758552275, + "flos": 41902545075840.0, + "grad_norm": 1.8029329624029096, + "language_loss": 0.65295768, + "learning_rate": 1.2644172698420903e-06, + "loss": 0.67510998, + "num_input_tokens_seen": 113056805, + "step": 5250, + "time_per_iteration": 3.6972575187683105 + }, + { + "auxiliary_loss_clip": 0.01107632, + "auxiliary_loss_mlp": 0.01036584, + "balance_loss_clip": 0.9394331, + "balance_loss_mlp": 1.02834678, + "epoch": 0.6313954187458667, + "flos": 19646800266240.0, + "grad_norm": 1.7303202957239416, + "language_loss": 0.84497076, + "learning_rate": 1.2636929534263892e-06, + "loss": 0.86641294, + "num_input_tokens_seen": 113075790, + "step": 5251, + "time_per_iteration": 2.7037487030029297 + }, + { + "auxiliary_loss_clip": 0.01105931, + "auxiliary_loss_mlp": 0.01026847, + "balance_loss_clip": 0.93441474, + "balance_loss_mlp": 1.01917005, + "epoch": 0.6315156616365057, + "flos": 22894273906560.0, + "grad_norm": 1.6734569113018183, + "language_loss": 0.78092992, + "learning_rate": 1.2629687487078821e-06, + "loss": 0.80225766, + "num_input_tokens_seen": 113094600, + "step": 5252, + "time_per_iteration": 2.72702693939209 + }, + { + "auxiliary_loss_clip": 0.0115438, + "auxiliary_loss_mlp": 0.01030058, + "balance_loss_clip": 1.01429999, + "balance_loss_mlp": 1.02168369, + "epoch": 0.6316359045271448, + "flos": 23726251699200.0, + "grad_norm": 2.0280004736868844, + "language_loss": 0.76604414, + "learning_rate": 1.2622446557964293e-06, + "loss": 0.78788853, + "num_input_tokens_seen": 113112605, + "step": 5253, + "time_per_iteration": 2.7044763565063477 + }, + { + "auxiliary_loss_clip": 0.01127966, + "auxiliary_loss_mlp": 0.01030627, + "balance_loss_clip": 0.97422802, + "balance_loss_mlp": 1.02331638, + "epoch": 0.631756147417784, + "flos": 33108417164160.0, + "grad_norm": 1.630936750772785, + "language_loss": 0.7142576, + "learning_rate": 1.261520674801876e-06, + "loss": 0.7358436, + "num_input_tokens_seen": 113133200, + "step": 5254, + "time_per_iteration": 2.7845773696899414 + }, + { + "auxiliary_loss_clip": 0.0112981, + "auxiliary_loss_mlp": 0.01027519, + "balance_loss_clip": 0.982723, + "balance_loss_mlp": 1.01944876, + "epoch": 0.631876390308423, + "flos": 31248424126080.0, + "grad_norm": 2.169288517255622, + "language_loss": 0.72312385, + "learning_rate": 1.2607968058340488e-06, + "loss": 0.74469709, + "num_input_tokens_seen": 113152895, + "step": 5255, + "time_per_iteration": 2.73934268951416 + }, + { + "auxiliary_loss_clip": 0.01125977, + "auxiliary_loss_mlp": 0.01029723, + "balance_loss_clip": 0.97654116, + "balance_loss_mlp": 1.02166402, + "epoch": 0.6319966331990621, + "flos": 24681152810880.0, + "grad_norm": 2.015533728116038, + "language_loss": 0.73299313, + "learning_rate": 1.2600730490027583e-06, + "loss": 0.75455016, + "num_input_tokens_seen": 113173135, + "step": 5256, + "time_per_iteration": 2.723235607147217 + }, + { + "auxiliary_loss_clip": 0.0110157, + "auxiliary_loss_mlp": 0.01025928, + "balance_loss_clip": 0.93688345, + "balance_loss_mlp": 1.0181973, + "epoch": 0.6321168760897012, + "flos": 17491764913920.0, + "grad_norm": 1.6288339295264462, + "language_loss": 0.80465031, + "learning_rate": 1.2593494044177984e-06, + "loss": 0.82592523, + "num_input_tokens_seen": 113191440, + "step": 5257, + "time_per_iteration": 2.77848219871521 + }, + { + "auxiliary_loss_clip": 0.01177795, + "auxiliary_loss_mlp": 0.01031136, + "balance_loss_clip": 1.05194402, + "balance_loss_mlp": 1.02238643, + "epoch": 0.6322371189803403, + "flos": 18295373940480.0, + "grad_norm": 2.180871660182301, + "language_loss": 0.80856305, + "learning_rate": 1.2586258721889448e-06, + "loss": 0.83065236, + "num_input_tokens_seen": 113208790, + "step": 5258, + "time_per_iteration": 2.662569522857666 + }, + { + "auxiliary_loss_clip": 0.01076902, + "auxiliary_loss_mlp": 0.01031203, + "balance_loss_clip": 0.90202302, + "balance_loss_mlp": 1.02377033, + "epoch": 0.6323573618709794, + "flos": 20157270399360.0, + "grad_norm": 1.8443718838841368, + "language_loss": 0.81965625, + "learning_rate": 1.2579024524259573e-06, + "loss": 0.84073734, + "num_input_tokens_seen": 113225050, + "step": 5259, + "time_per_iteration": 2.7372896671295166 + }, + { + "auxiliary_loss_clip": 0.01122731, + "auxiliary_loss_mlp": 0.0103181, + "balance_loss_clip": 0.97305727, + "balance_loss_mlp": 1.02388549, + "epoch": 0.6324776047616185, + "flos": 20042391726720.0, + "grad_norm": 2.106252109623463, + "language_loss": 0.91385937, + "learning_rate": 1.2571791452385768e-06, + "loss": 0.93540478, + "num_input_tokens_seen": 113242315, + "step": 5260, + "time_per_iteration": 2.7359511852264404 + }, + { + "auxiliary_loss_clip": 0.01133402, + "auxiliary_loss_mlp": 0.01026464, + "balance_loss_clip": 0.98057711, + "balance_loss_mlp": 1.01835155, + "epoch": 0.6325978476522576, + "flos": 30848235724800.0, + "grad_norm": 1.6884201785051725, + "language_loss": 0.77235949, + "learning_rate": 1.2564559507365301e-06, + "loss": 0.79395813, + "num_input_tokens_seen": 113264720, + "step": 5261, + "time_per_iteration": 2.8599960803985596 + }, + { + "auxiliary_loss_clip": 0.01133, + "auxiliary_loss_mlp": 0.01025848, + "balance_loss_clip": 0.9800902, + "balance_loss_mlp": 1.01797462, + "epoch": 0.6327180905428966, + "flos": 24535104111360.0, + "grad_norm": 1.9156802618033346, + "language_loss": 0.78921747, + "learning_rate": 1.2557328690295244e-06, + "loss": 0.81080592, + "num_input_tokens_seen": 113282910, + "step": 5262, + "time_per_iteration": 2.8317947387695312 + }, + { + "auxiliary_loss_clip": 0.01109117, + "auxiliary_loss_mlp": 0.01029103, + "balance_loss_clip": 0.93853819, + "balance_loss_mlp": 1.02185512, + "epoch": 0.6328383334335358, + "flos": 21575274583680.0, + "grad_norm": 2.4904291736012483, + "language_loss": 0.76454204, + "learning_rate": 1.255009900227251e-06, + "loss": 0.78592426, + "num_input_tokens_seen": 113301935, + "step": 5263, + "time_per_iteration": 2.713348388671875 + }, + { + "auxiliary_loss_clip": 0.01176635, + "auxiliary_loss_mlp": 0.0102875, + "balance_loss_clip": 1.05730581, + "balance_loss_mlp": 1.02098966, + "epoch": 0.6329585763241748, + "flos": 22929861306240.0, + "grad_norm": 4.528636243022546, + "language_loss": 0.79610527, + "learning_rate": 1.254287044439383e-06, + "loss": 0.81815916, + "num_input_tokens_seen": 113321540, + "step": 5264, + "time_per_iteration": 3.650679111480713 + }, + { + "auxiliary_loss_clip": 0.01070067, + "auxiliary_loss_mlp": 0.00998405, + "balance_loss_clip": 1.01630211, + "balance_loss_mlp": 0.99737376, + "epoch": 0.6330788192148139, + "flos": 70936897847040.0, + "grad_norm": 0.77164627416537, + "language_loss": 0.54395092, + "learning_rate": 1.2535643017755776e-06, + "loss": 0.56463569, + "num_input_tokens_seen": 113383730, + "step": 5265, + "time_per_iteration": 4.212857961654663 + }, + { + "auxiliary_loss_clip": 0.01104424, + "auxiliary_loss_mlp": 0.01027847, + "balance_loss_clip": 0.93593937, + "balance_loss_mlp": 1.0200386, + "epoch": 0.6331990621054531, + "flos": 21244501215360.0, + "grad_norm": 2.825520804450941, + "language_loss": 0.71507943, + "learning_rate": 1.2528416723454737e-06, + "loss": 0.73640215, + "num_input_tokens_seen": 113400400, + "step": 5266, + "time_per_iteration": 3.6698572635650635 + }, + { + "auxiliary_loss_clip": 0.01173774, + "auxiliary_loss_mlp": 0.01025467, + "balance_loss_clip": 1.05505037, + "balance_loss_mlp": 1.01779556, + "epoch": 0.6333193049960921, + "flos": 34459412526720.0, + "grad_norm": 1.47269537308198, + "language_loss": 0.71058834, + "learning_rate": 1.2521191562586945e-06, + "loss": 0.73258072, + "num_input_tokens_seen": 113424050, + "step": 5267, + "time_per_iteration": 2.7294678688049316 + }, + { + "auxiliary_loss_clip": 0.01176085, + "auxiliary_loss_mlp": 0.00510369, + "balance_loss_clip": 1.05512512, + "balance_loss_mlp": 0.0, + "epoch": 0.6334395478867312, + "flos": 18329883932160.0, + "grad_norm": 2.0969685155453006, + "language_loss": 0.76791722, + "learning_rate": 1.2513967536248445e-06, + "loss": 0.78478169, + "num_input_tokens_seen": 113440370, + "step": 5268, + "time_per_iteration": 2.631124496459961 + }, + { + "auxiliary_loss_clip": 0.01149945, + "auxiliary_loss_mlp": 0.01027879, + "balance_loss_clip": 1.01774395, + "balance_loss_mlp": 1.02027392, + "epoch": 0.6335597907773702, + "flos": 23623152687360.0, + "grad_norm": 1.6481188079346272, + "language_loss": 0.8084802, + "learning_rate": 1.2506744645535117e-06, + "loss": 0.83025843, + "num_input_tokens_seen": 113460800, + "step": 5269, + "time_per_iteration": 2.678013324737549 + }, + { + "auxiliary_loss_clip": 0.01120439, + "auxiliary_loss_mlp": 0.01027345, + "balance_loss_clip": 0.97176582, + "balance_loss_mlp": 1.01855993, + "epoch": 0.6336800336680094, + "flos": 22710913954560.0, + "grad_norm": 2.543432781287327, + "language_loss": 0.60635126, + "learning_rate": 1.249952289154267e-06, + "loss": 0.62782913, + "num_input_tokens_seen": 113480840, + "step": 5270, + "time_per_iteration": 2.786086320877075 + }, + { + "auxiliary_loss_clip": 0.01041107, + "auxiliary_loss_mlp": 0.01026863, + "balance_loss_clip": 0.85691965, + "balance_loss_mlp": 1.01864326, + "epoch": 0.6338002765586485, + "flos": 23622757637760.0, + "grad_norm": 1.6324622478805826, + "language_loss": 0.76858646, + "learning_rate": 1.2492302275366635e-06, + "loss": 0.78926617, + "num_input_tokens_seen": 113500515, + "step": 5271, + "time_per_iteration": 2.802985429763794 + }, + { + "auxiliary_loss_clip": 0.01148121, + "auxiliary_loss_mlp": 0.01031358, + "balance_loss_clip": 1.01452661, + "balance_loss_mlp": 1.02303743, + "epoch": 0.6339205194492875, + "flos": 26505450708480.0, + "grad_norm": 2.2277322737302674, + "language_loss": 0.65664363, + "learning_rate": 1.2485082798102377e-06, + "loss": 0.67843843, + "num_input_tokens_seen": 113520930, + "step": 5272, + "time_per_iteration": 2.698160171508789 + }, + { + "auxiliary_loss_clip": 0.0110995, + "auxiliary_loss_mlp": 0.01029423, + "balance_loss_clip": 0.93881929, + "balance_loss_mlp": 1.02087593, + "epoch": 0.6340407623399267, + "flos": 18544306170240.0, + "grad_norm": 2.0486133381366933, + "language_loss": 0.6885677, + "learning_rate": 1.2477864460845084e-06, + "loss": 0.70996141, + "num_input_tokens_seen": 113537330, + "step": 5273, + "time_per_iteration": 2.678149461746216 + }, + { + "auxiliary_loss_clip": 0.0112514, + "auxiliary_loss_mlp": 0.01032672, + "balance_loss_clip": 0.9749577, + "balance_loss_mlp": 1.02451849, + "epoch": 0.6341610052305657, + "flos": 17712579772800.0, + "grad_norm": 2.5511500644080907, + "language_loss": 0.73800433, + "learning_rate": 1.2470647264689776e-06, + "loss": 0.75958246, + "num_input_tokens_seen": 113555810, + "step": 5274, + "time_per_iteration": 2.7259695529937744 + }, + { + "auxiliary_loss_clip": 0.0106281, + "auxiliary_loss_mlp": 0.01025203, + "balance_loss_clip": 0.85866857, + "balance_loss_mlp": 1.01751423, + "epoch": 0.6342812481212048, + "flos": 23587026583680.0, + "grad_norm": 1.9452695659380255, + "language_loss": 0.71529919, + "learning_rate": 1.2463431210731282e-06, + "loss": 0.73617935, + "num_input_tokens_seen": 113575395, + "step": 5275, + "time_per_iteration": 3.0061194896698 + }, + { + "auxiliary_loss_clip": 0.01089394, + "auxiliary_loss_mlp": 0.01024928, + "balance_loss_clip": 0.89782846, + "balance_loss_mlp": 1.01720893, + "epoch": 0.634401491011844, + "flos": 17821927751040.0, + "grad_norm": 2.243196486561024, + "language_loss": 0.76354206, + "learning_rate": 1.2456216300064289e-06, + "loss": 0.78468525, + "num_input_tokens_seen": 113592945, + "step": 5276, + "time_per_iteration": 3.5924899578094482 + }, + { + "auxiliary_loss_clip": 0.01121553, + "auxiliary_loss_mlp": 0.01025543, + "balance_loss_clip": 0.97409487, + "balance_loss_mlp": 1.01735044, + "epoch": 0.634521733902483, + "flos": 21358158825600.0, + "grad_norm": 1.6468482867830048, + "language_loss": 0.78607738, + "learning_rate": 1.244900253378328e-06, + "loss": 0.80754828, + "num_input_tokens_seen": 113613000, + "step": 5277, + "time_per_iteration": 2.7830400466918945 + }, + { + "auxiliary_loss_clip": 0.01014085, + "auxiliary_loss_mlp": 0.0102045, + "balance_loss_clip": 0.78649688, + "balance_loss_mlp": 1.01329768, + "epoch": 0.6346419767931221, + "flos": 16545052103040.0, + "grad_norm": 2.059812110667967, + "language_loss": 0.69625032, + "learning_rate": 1.2441789912982583e-06, + "loss": 0.71659565, + "num_input_tokens_seen": 113630085, + "step": 5278, + "time_per_iteration": 2.923891544342041 + }, + { + "auxiliary_loss_clip": 0.01152274, + "auxiliary_loss_mlp": 0.01028309, + "balance_loss_clip": 1.01483393, + "balance_loss_mlp": 1.01998842, + "epoch": 0.6347622196837612, + "flos": 24350989973760.0, + "grad_norm": 2.036808742754648, + "language_loss": 0.64944994, + "learning_rate": 1.2434578438756346e-06, + "loss": 0.67125583, + "num_input_tokens_seen": 113650515, + "step": 5279, + "time_per_iteration": 3.2184600830078125 + }, + { + "auxiliary_loss_clip": 0.01152569, + "auxiliary_loss_mlp": 0.01026694, + "balance_loss_clip": 1.01312876, + "balance_loss_mlp": 1.0189153, + "epoch": 0.6348824625744003, + "flos": 64523178195840.0, + "grad_norm": 1.7957555273223305, + "language_loss": 0.7808907, + "learning_rate": 1.242736811219855e-06, + "loss": 0.80268323, + "num_input_tokens_seen": 113676475, + "step": 5280, + "time_per_iteration": 3.088150978088379 + }, + { + "auxiliary_loss_clip": 0.01143721, + "auxiliary_loss_mlp": 0.01030121, + "balance_loss_clip": 1.01204562, + "balance_loss_mlp": 1.02285552, + "epoch": 0.6350027054650393, + "flos": 28622133313920.0, + "grad_norm": 1.7208624300114845, + "language_loss": 0.81638932, + "learning_rate": 1.2420158934402988e-06, + "loss": 0.83812773, + "num_input_tokens_seen": 113697090, + "step": 5281, + "time_per_iteration": 2.786111831665039 + }, + { + "auxiliary_loss_clip": 0.01089754, + "auxiliary_loss_mlp": 0.01025994, + "balance_loss_clip": 0.93299747, + "balance_loss_mlp": 1.0182929, + "epoch": 0.6351229483556785, + "flos": 23002544476800.0, + "grad_norm": 2.098477364003776, + "language_loss": 0.84825307, + "learning_rate": 1.2412950906463286e-06, + "loss": 0.86941051, + "num_input_tokens_seen": 113714395, + "step": 5282, + "time_per_iteration": 2.8109538555145264 + }, + { + "auxiliary_loss_clip": 0.01081576, + "auxiliary_loss_mlp": 0.01026695, + "balance_loss_clip": 0.90087467, + "balance_loss_mlp": 1.01919127, + "epoch": 0.6352431912463176, + "flos": 21939300967680.0, + "grad_norm": 1.6624166592206209, + "language_loss": 0.90063083, + "learning_rate": 1.2405744029472902e-06, + "loss": 0.92171359, + "num_input_tokens_seen": 113733880, + "step": 5283, + "time_per_iteration": 2.8548498153686523 + }, + { + "auxiliary_loss_clip": 0.01124217, + "auxiliary_loss_mlp": 0.01025762, + "balance_loss_clip": 0.97437263, + "balance_loss_mlp": 1.01845765, + "epoch": 0.6353634341369566, + "flos": 13735257684480.0, + "grad_norm": 1.8737368918796904, + "language_loss": 0.76199389, + "learning_rate": 1.2398538304525108e-06, + "loss": 0.7834937, + "num_input_tokens_seen": 113752505, + "step": 5284, + "time_per_iteration": 2.6914408206939697 + }, + { + "auxiliary_loss_clip": 0.01104035, + "auxiliary_loss_mlp": 0.01035255, + "balance_loss_clip": 0.9423697, + "balance_loss_mlp": 1.02685046, + "epoch": 0.6354836770275958, + "flos": 19316170552320.0, + "grad_norm": 2.085373352930252, + "language_loss": 0.75525081, + "learning_rate": 1.2391333732713016e-06, + "loss": 0.77664375, + "num_input_tokens_seen": 113770310, + "step": 5285, + "time_per_iteration": 2.6888175010681152 + }, + { + "auxiliary_loss_clip": 0.01104492, + "auxiliary_loss_mlp": 0.01033398, + "balance_loss_clip": 0.93731987, + "balance_loss_mlp": 1.02489829, + "epoch": 0.6356039199182348, + "flos": 21613375935360.0, + "grad_norm": 2.367150204454153, + "language_loss": 0.78642786, + "learning_rate": 1.2384130315129543e-06, + "loss": 0.80780673, + "num_input_tokens_seen": 113788635, + "step": 5286, + "time_per_iteration": 2.7771122455596924 + }, + { + "auxiliary_loss_clip": 0.0099538, + "auxiliary_loss_mlp": 0.01032252, + "balance_loss_clip": 0.74521154, + "balance_loss_mlp": 1.02397275, + "epoch": 0.6357241628088739, + "flos": 18111978074880.0, + "grad_norm": 2.3184910333529, + "language_loss": 0.73951417, + "learning_rate": 1.2376928052867447e-06, + "loss": 0.75979054, + "num_input_tokens_seen": 113807755, + "step": 5287, + "time_per_iteration": 3.0115370750427246 + }, + { + "auxiliary_loss_clip": 0.01128723, + "auxiliary_loss_mlp": 0.01031549, + "balance_loss_clip": 0.97797596, + "balance_loss_mlp": 1.02382731, + "epoch": 0.6358444056995131, + "flos": 24935256599040.0, + "grad_norm": 2.0740748933505784, + "language_loss": 0.77616417, + "learning_rate": 1.2369726947019299e-06, + "loss": 0.79776692, + "num_input_tokens_seen": 113828230, + "step": 5288, + "time_per_iteration": 3.4689464569091797 + }, + { + "auxiliary_loss_clip": 0.01153971, + "auxiliary_loss_mlp": 0.01031025, + "balance_loss_clip": 1.0170635, + "balance_loss_mlp": 1.02329731, + "epoch": 0.6359646485901521, + "flos": 23293348986240.0, + "grad_norm": 2.5082148721828337, + "language_loss": 0.67225045, + "learning_rate": 1.2362526998677511e-06, + "loss": 0.69410038, + "num_input_tokens_seen": 113844595, + "step": 5289, + "time_per_iteration": 2.715670108795166 + }, + { + "auxiliary_loss_clip": 0.01131764, + "auxiliary_loss_mlp": 0.01025862, + "balance_loss_clip": 0.97675312, + "balance_loss_mlp": 1.01857781, + "epoch": 0.6360848914807912, + "flos": 20887442069760.0, + "grad_norm": 1.6643828373937088, + "language_loss": 0.84262747, + "learning_rate": 1.2355328208934301e-06, + "loss": 0.86420381, + "num_input_tokens_seen": 113863470, + "step": 5290, + "time_per_iteration": 3.6368205547332764 + }, + { + "auxiliary_loss_clip": 0.0115005, + "auxiliary_loss_mlp": 0.00509745, + "balance_loss_clip": 1.0121392, + "balance_loss_mlp": 0.0, + "epoch": 0.6362051343714303, + "flos": 18479775386880.0, + "grad_norm": 1.5911893069629566, + "language_loss": 0.7260294, + "learning_rate": 1.2348130578881728e-06, + "loss": 0.74262738, + "num_input_tokens_seen": 113881690, + "step": 5291, + "time_per_iteration": 3.60919189453125 + }, + { + "auxiliary_loss_clip": 0.01180509, + "auxiliary_loss_mlp": 0.01031923, + "balance_loss_clip": 1.05798078, + "balance_loss_mlp": 1.02354312, + "epoch": 0.6363253772620694, + "flos": 24389594115840.0, + "grad_norm": 1.9915331275498314, + "language_loss": 0.76449776, + "learning_rate": 1.2340934109611664e-06, + "loss": 0.78662205, + "num_input_tokens_seen": 113902450, + "step": 5292, + "time_per_iteration": 3.7368013858795166 + }, + { + "auxiliary_loss_clip": 0.0113217, + "auxiliary_loss_mlp": 0.01026562, + "balance_loss_clip": 0.97889584, + "balance_loss_mlp": 1.01842642, + "epoch": 0.6364456201527084, + "flos": 25958243940480.0, + "grad_norm": 2.852917126207278, + "language_loss": 0.69128704, + "learning_rate": 1.2333738802215798e-06, + "loss": 0.71287435, + "num_input_tokens_seen": 113922670, + "step": 5293, + "time_per_iteration": 2.823599338531494 + }, + { + "auxiliary_loss_clip": 0.01075106, + "auxiliary_loss_mlp": 0.01029551, + "balance_loss_clip": 0.89695609, + "balance_loss_mlp": 1.0219394, + "epoch": 0.6365658630433476, + "flos": 20740711011840.0, + "grad_norm": 2.141179331841782, + "language_loss": 0.81054127, + "learning_rate": 1.2326544657785668e-06, + "loss": 0.83158779, + "num_input_tokens_seen": 113942360, + "step": 5294, + "time_per_iteration": 2.7659919261932373 + }, + { + "auxiliary_loss_clip": 0.01095102, + "auxiliary_loss_mlp": 0.01032323, + "balance_loss_clip": 0.93522346, + "balance_loss_mlp": 1.02443111, + "epoch": 0.6366861059339867, + "flos": 21434146047360.0, + "grad_norm": 2.26661172952152, + "language_loss": 0.74546111, + "learning_rate": 1.2319351677412608e-06, + "loss": 0.76673532, + "num_input_tokens_seen": 113959405, + "step": 5295, + "time_per_iteration": 2.7387263774871826 + }, + { + "auxiliary_loss_clip": 0.01113216, + "auxiliary_loss_mlp": 0.0103431, + "balance_loss_clip": 0.93998551, + "balance_loss_mlp": 1.02639437, + "epoch": 0.6368063488246257, + "flos": 22267093507200.0, + "grad_norm": 1.7211722649265142, + "language_loss": 0.73966718, + "learning_rate": 1.2312159862187796e-06, + "loss": 0.76114249, + "num_input_tokens_seen": 113977815, + "step": 5296, + "time_per_iteration": 2.7235960960388184 + }, + { + "auxiliary_loss_clip": 0.01179693, + "auxiliary_loss_mlp": 0.01028975, + "balance_loss_clip": 1.05666041, + "balance_loss_mlp": 1.02067852, + "epoch": 0.6369265917152649, + "flos": 22420719976320.0, + "grad_norm": 6.576402286440781, + "language_loss": 0.76084387, + "learning_rate": 1.2304969213202217e-06, + "loss": 0.78293061, + "num_input_tokens_seen": 113999075, + "step": 5297, + "time_per_iteration": 2.6794681549072266 + }, + { + "auxiliary_loss_clip": 0.01122931, + "auxiliary_loss_mlp": 0.01026689, + "balance_loss_clip": 0.97391057, + "balance_loss_mlp": 1.01911926, + "epoch": 0.6370468346059039, + "flos": 24718176754560.0, + "grad_norm": 2.361304207977151, + "language_loss": 0.78932476, + "learning_rate": 1.2297779731546692e-06, + "loss": 0.810821, + "num_input_tokens_seen": 114018170, + "step": 5298, + "time_per_iteration": 2.775784969329834 + }, + { + "auxiliary_loss_clip": 0.01126605, + "auxiliary_loss_mlp": 0.01027972, + "balance_loss_clip": 0.97855467, + "balance_loss_mlp": 1.02036083, + "epoch": 0.637167077496543, + "flos": 25296589463040.0, + "grad_norm": 1.9375684347452478, + "language_loss": 0.78110379, + "learning_rate": 1.2290591418311853e-06, + "loss": 0.80264956, + "num_input_tokens_seen": 114035565, + "step": 5299, + "time_per_iteration": 2.7367806434631348 + }, + { + "auxiliary_loss_clip": 0.01151584, + "auxiliary_loss_mlp": 0.01031417, + "balance_loss_clip": 1.01681376, + "balance_loss_mlp": 1.02351964, + "epoch": 0.637287320387182, + "flos": 27671110871040.0, + "grad_norm": 1.7512642355365173, + "language_loss": 0.72253436, + "learning_rate": 1.2283404274588172e-06, + "loss": 0.74436438, + "num_input_tokens_seen": 114054510, + "step": 5300, + "time_per_iteration": 2.687898635864258 + }, + { + "auxiliary_loss_clip": 0.00935392, + "auxiliary_loss_mlp": 0.01010436, + "balance_loss_clip": 0.79104662, + "balance_loss_mlp": 1.00920856, + "epoch": 0.6374075632778212, + "flos": 63173406873600.0, + "grad_norm": 0.7407677281978051, + "language_loss": 0.52791846, + "learning_rate": 1.227621830146592e-06, + "loss": 0.54737675, + "num_input_tokens_seen": 114109875, + "step": 5301, + "time_per_iteration": 3.346795082092285 + }, + { + "auxiliary_loss_clip": 0.01108718, + "auxiliary_loss_mlp": 0.0103261, + "balance_loss_clip": 0.94141322, + "balance_loss_mlp": 1.02464056, + "epoch": 0.6375278061684603, + "flos": 25558127366400.0, + "grad_norm": 1.7077769774622855, + "language_loss": 0.79268438, + "learning_rate": 1.2269033500035217e-06, + "loss": 0.81409764, + "num_input_tokens_seen": 114130010, + "step": 5302, + "time_per_iteration": 3.9077847003936768 + }, + { + "auxiliary_loss_clip": 0.01109256, + "auxiliary_loss_mlp": 0.01033554, + "balance_loss_clip": 0.94146508, + "balance_loss_mlp": 1.02561474, + "epoch": 0.6376480490590993, + "flos": 25666362023040.0, + "grad_norm": 4.182132159308995, + "language_loss": 0.73755395, + "learning_rate": 1.2261849871385988e-06, + "loss": 0.75898206, + "num_input_tokens_seen": 114151115, + "step": 5303, + "time_per_iteration": 2.725306510925293 + }, + { + "auxiliary_loss_clip": 0.01174223, + "auxiliary_loss_mlp": 0.01030321, + "balance_loss_clip": 1.05305243, + "balance_loss_mlp": 1.02208364, + "epoch": 0.6377682919497385, + "flos": 31537684350720.0, + "grad_norm": 2.272292293184006, + "language_loss": 0.62239319, + "learning_rate": 1.2254667416607972e-06, + "loss": 0.64443868, + "num_input_tokens_seen": 114172715, + "step": 5304, + "time_per_iteration": 2.769834041595459 + }, + { + "auxiliary_loss_clip": 0.0114847, + "auxiliary_loss_mlp": 0.01030039, + "balance_loss_clip": 1.01399171, + "balance_loss_mlp": 1.02205777, + "epoch": 0.6378885348403776, + "flos": 23039209284480.0, + "grad_norm": 1.7493551792034547, + "language_loss": 0.83228129, + "learning_rate": 1.2247486136790756e-06, + "loss": 0.85406637, + "num_input_tokens_seen": 114192195, + "step": 5305, + "time_per_iteration": 2.7479865550994873 + }, + { + "auxiliary_loss_clip": 0.01151351, + "auxiliary_loss_mlp": 0.01029113, + "balance_loss_clip": 1.01497984, + "balance_loss_mlp": 1.02125156, + "epoch": 0.6380087777310166, + "flos": 18697070712960.0, + "grad_norm": 2.190345810329433, + "language_loss": 0.80719429, + "learning_rate": 1.2240306033023726e-06, + "loss": 0.82899898, + "num_input_tokens_seen": 114210020, + "step": 5306, + "time_per_iteration": 2.6487128734588623 + }, + { + "auxiliary_loss_clip": 0.01103415, + "auxiliary_loss_mlp": 0.01026527, + "balance_loss_clip": 0.93294704, + "balance_loss_mlp": 1.01870704, + "epoch": 0.6381290206216558, + "flos": 23331558078720.0, + "grad_norm": 1.914761012967988, + "language_loss": 0.71936196, + "learning_rate": 1.223312710639611e-06, + "loss": 0.74066138, + "num_input_tokens_seen": 114228740, + "step": 5307, + "time_per_iteration": 2.798067808151245 + }, + { + "auxiliary_loss_clip": 0.01125176, + "auxiliary_loss_mlp": 0.01027374, + "balance_loss_clip": 0.97705722, + "balance_loss_mlp": 1.01915431, + "epoch": 0.6382492635122948, + "flos": 18880466578560.0, + "grad_norm": 3.4526041731100436, + "language_loss": 0.87016743, + "learning_rate": 1.2225949357996928e-06, + "loss": 0.89169288, + "num_input_tokens_seen": 114246865, + "step": 5308, + "time_per_iteration": 2.6670098304748535 + }, + { + "auxiliary_loss_clip": 0.0114655, + "auxiliary_loss_mlp": 0.01023115, + "balance_loss_clip": 1.01571298, + "balance_loss_mlp": 1.01608777, + "epoch": 0.6383695064029339, + "flos": 27819134818560.0, + "grad_norm": 1.6339730205562708, + "language_loss": 0.80536199, + "learning_rate": 1.221877278891505e-06, + "loss": 0.82705867, + "num_input_tokens_seen": 114266120, + "step": 5309, + "time_per_iteration": 2.8207836151123047 + }, + { + "auxiliary_loss_clip": 0.01156916, + "auxiliary_loss_mlp": 0.01027521, + "balance_loss_clip": 1.01615632, + "balance_loss_mlp": 1.01985538, + "epoch": 0.638489749293573, + "flos": 26395635853440.0, + "grad_norm": 2.355874417986054, + "language_loss": 0.71411145, + "learning_rate": 1.221159740023915e-06, + "loss": 0.73595572, + "num_input_tokens_seen": 114285950, + "step": 5310, + "time_per_iteration": 2.706387519836426 + }, + { + "auxiliary_loss_clip": 0.01112991, + "auxiliary_loss_mlp": 0.00510198, + "balance_loss_clip": 0.94042957, + "balance_loss_mlp": 0.0, + "epoch": 0.6386099921842121, + "flos": 23988328306560.0, + "grad_norm": 1.9457268502744156, + "language_loss": 0.72807276, + "learning_rate": 1.2204423193057735e-06, + "loss": 0.7443046, + "num_input_tokens_seen": 114304780, + "step": 5311, + "time_per_iteration": 2.8492021560668945 + }, + { + "auxiliary_loss_clip": 0.01029157, + "auxiliary_loss_mlp": 0.01004301, + "balance_loss_clip": 0.93945348, + "balance_loss_mlp": 1.00313282, + "epoch": 0.6387302350748512, + "flos": 71731169337600.0, + "grad_norm": 0.8504534803289208, + "language_loss": 0.63360071, + "learning_rate": 1.2197250168459122e-06, + "loss": 0.65393531, + "num_input_tokens_seen": 114361180, + "step": 5312, + "time_per_iteration": 3.364144802093506 + }, + { + "auxiliary_loss_clip": 0.01155438, + "auxiliary_loss_mlp": 0.01030685, + "balance_loss_clip": 1.0169785, + "balance_loss_mlp": 1.02331495, + "epoch": 0.6388504779654903, + "flos": 14535778141440.0, + "grad_norm": 1.9676553410551172, + "language_loss": 0.7425108, + "learning_rate": 1.2190078327531454e-06, + "loss": 0.76437199, + "num_input_tokens_seen": 114377425, + "step": 5313, + "time_per_iteration": 2.7809042930603027 + }, + { + "auxiliary_loss_clip": 0.01152248, + "auxiliary_loss_mlp": 0.01026516, + "balance_loss_clip": 1.01490128, + "balance_loss_mlp": 1.01964331, + "epoch": 0.6389707208561294, + "flos": 22346133384960.0, + "grad_norm": 1.5383806232166968, + "language_loss": 0.73025656, + "learning_rate": 1.2182907671362697e-06, + "loss": 0.7520442, + "num_input_tokens_seen": 114398120, + "step": 5314, + "time_per_iteration": 2.7528655529022217 + }, + { + "auxiliary_loss_clip": 0.01151361, + "auxiliary_loss_mlp": 0.01027649, + "balance_loss_clip": 1.01714802, + "balance_loss_mlp": 1.0196228, + "epoch": 0.6390909637467684, + "flos": 19426883247360.0, + "grad_norm": 6.190389862114863, + "language_loss": 0.78849232, + "learning_rate": 1.2175738201040626e-06, + "loss": 0.81028241, + "num_input_tokens_seen": 114415160, + "step": 5315, + "time_per_iteration": 2.6824653148651123 + }, + { + "auxiliary_loss_clip": 0.01150217, + "auxiliary_loss_mlp": 0.01027966, + "balance_loss_clip": 1.01433206, + "balance_loss_mlp": 1.01992798, + "epoch": 0.6392112066374076, + "flos": 24090852700800.0, + "grad_norm": 2.1127519949815183, + "language_loss": 0.78912777, + "learning_rate": 1.2168569917652855e-06, + "loss": 0.81090963, + "num_input_tokens_seen": 114435015, + "step": 5316, + "time_per_iteration": 3.6740307807922363 + }, + { + "auxiliary_loss_clip": 0.01151033, + "auxiliary_loss_mlp": 0.01029459, + "balance_loss_clip": 1.01512492, + "balance_loss_mlp": 1.02216959, + "epoch": 0.6393314495280467, + "flos": 26795141896320.0, + "grad_norm": 1.7088701802994424, + "language_loss": 0.63926905, + "learning_rate": 1.2161402822286797e-06, + "loss": 0.66107398, + "num_input_tokens_seen": 114455700, + "step": 5317, + "time_per_iteration": 3.6657838821411133 + }, + { + "auxiliary_loss_clip": 0.01103401, + "auxiliary_loss_mlp": 0.01028006, + "balance_loss_clip": 0.9378013, + "balance_loss_mlp": 1.02029586, + "epoch": 0.6394516924186857, + "flos": 20260692633600.0, + "grad_norm": 2.229032739323879, + "language_loss": 0.79061389, + "learning_rate": 1.2154236916029703e-06, + "loss": 0.81192803, + "num_input_tokens_seen": 114473675, + "step": 5318, + "time_per_iteration": 3.725210189819336 + }, + { + "auxiliary_loss_clip": 0.01079634, + "auxiliary_loss_mlp": 0.01031252, + "balance_loss_clip": 0.89537686, + "balance_loss_mlp": 1.02328897, + "epoch": 0.6395719353093249, + "flos": 18368847210240.0, + "grad_norm": 4.365529333295447, + "language_loss": 0.73467898, + "learning_rate": 1.2147072199968627e-06, + "loss": 0.75578785, + "num_input_tokens_seen": 114492310, + "step": 5319, + "time_per_iteration": 2.8256797790527344 + }, + { + "auxiliary_loss_clip": 0.0114759, + "auxiliary_loss_mlp": 0.01025958, + "balance_loss_clip": 1.01407158, + "balance_loss_mlp": 1.01865649, + "epoch": 0.6396921781999639, + "flos": 17566315591680.0, + "grad_norm": 2.1299152004797564, + "language_loss": 0.71776396, + "learning_rate": 1.2139908675190454e-06, + "loss": 0.73949945, + "num_input_tokens_seen": 114511520, + "step": 5320, + "time_per_iteration": 2.6423211097717285 + }, + { + "auxiliary_loss_clip": 0.01046933, + "auxiliary_loss_mlp": 0.0102916, + "balance_loss_clip": 0.85692459, + "balance_loss_mlp": 1.02160239, + "epoch": 0.639812421090603, + "flos": 21251252972160.0, + "grad_norm": 1.8277195767016972, + "language_loss": 0.75001025, + "learning_rate": 1.2132746342781883e-06, + "loss": 0.77077121, + "num_input_tokens_seen": 114532680, + "step": 5321, + "time_per_iteration": 2.8823118209838867 + }, + { + "auxiliary_loss_clip": 0.01174868, + "auxiliary_loss_mlp": 0.01029418, + "balance_loss_clip": 1.05375981, + "balance_loss_mlp": 1.02203929, + "epoch": 0.6399326639812422, + "flos": 11180967684480.0, + "grad_norm": 3.799350252471818, + "language_loss": 0.80291611, + "learning_rate": 1.2125585203829442e-06, + "loss": 0.82495892, + "num_input_tokens_seen": 114548320, + "step": 5322, + "time_per_iteration": 2.627890110015869 + }, + { + "auxiliary_loss_clip": 0.0109319, + "auxiliary_loss_mlp": 0.01027137, + "balance_loss_clip": 0.93770516, + "balance_loss_mlp": 1.01919794, + "epoch": 0.6400529068718812, + "flos": 23911048195200.0, + "grad_norm": 1.7562533177875925, + "language_loss": 0.73795509, + "learning_rate": 1.211842525941946e-06, + "loss": 0.75915837, + "num_input_tokens_seen": 114568115, + "step": 5323, + "time_per_iteration": 2.7909464836120605 + }, + { + "auxiliary_loss_clip": 0.01073951, + "auxiliary_loss_mlp": 0.01025053, + "balance_loss_clip": 0.89820051, + "balance_loss_mlp": 1.0172627, + "epoch": 0.6401731497625203, + "flos": 44018724890880.0, + "grad_norm": 2.3958545431923786, + "language_loss": 0.7921474, + "learning_rate": 1.2111266510638105e-06, + "loss": 0.81313741, + "num_input_tokens_seen": 114591040, + "step": 5324, + "time_per_iteration": 2.982996940612793 + }, + { + "auxiliary_loss_clip": 0.01051995, + "auxiliary_loss_mlp": 0.01030838, + "balance_loss_clip": 0.86022097, + "balance_loss_mlp": 1.02364647, + "epoch": 0.6402933926531594, + "flos": 20662209838080.0, + "grad_norm": 1.9745310362043835, + "language_loss": 0.80068636, + "learning_rate": 1.2104108958571346e-06, + "loss": 0.82151473, + "num_input_tokens_seen": 114609310, + "step": 5325, + "time_per_iteration": 2.83385968208313 + }, + { + "auxiliary_loss_clip": 0.01147229, + "auxiliary_loss_mlp": 0.01030757, + "balance_loss_clip": 1.01588988, + "balance_loss_mlp": 1.02378321, + "epoch": 0.6404136355437985, + "flos": 24863327614080.0, + "grad_norm": 1.8751634843769454, + "language_loss": 0.76156163, + "learning_rate": 1.2096952604304975e-06, + "loss": 0.78334153, + "num_input_tokens_seen": 114629740, + "step": 5326, + "time_per_iteration": 2.805809736251831 + }, + { + "auxiliary_loss_clip": 0.01152326, + "auxiliary_loss_mlp": 0.01027757, + "balance_loss_clip": 1.01455998, + "balance_loss_mlp": 1.01995516, + "epoch": 0.6405338784344375, + "flos": 40479548901120.0, + "grad_norm": 1.9893737513103782, + "language_loss": 0.70220858, + "learning_rate": 1.2089797448924616e-06, + "loss": 0.72400939, + "num_input_tokens_seen": 114653615, + "step": 5327, + "time_per_iteration": 2.8489904403686523 + }, + { + "auxiliary_loss_clip": 0.01084435, + "auxiliary_loss_mlp": 0.01028376, + "balance_loss_clip": 0.89638436, + "balance_loss_mlp": 1.02019823, + "epoch": 0.6406541213250767, + "flos": 20886041439360.0, + "grad_norm": 1.9336554224893978, + "language_loss": 0.66533601, + "learning_rate": 1.2082643493515692e-06, + "loss": 0.68646413, + "num_input_tokens_seen": 114671935, + "step": 5328, + "time_per_iteration": 3.6815807819366455 + }, + { + "auxiliary_loss_clip": 0.01148641, + "auxiliary_loss_mlp": 0.01024632, + "balance_loss_clip": 1.0138582, + "balance_loss_mlp": 1.01715732, + "epoch": 0.6407743642157158, + "flos": 23295970679040.0, + "grad_norm": 1.7734784730473432, + "language_loss": 0.81632322, + "learning_rate": 1.207549073916346e-06, + "loss": 0.83805597, + "num_input_tokens_seen": 114692870, + "step": 5329, + "time_per_iteration": 2.761834144592285 + }, + { + "auxiliary_loss_clip": 0.01118944, + "auxiliary_loss_mlp": 0.01026769, + "balance_loss_clip": 0.97508508, + "balance_loss_mlp": 1.01975608, + "epoch": 0.6408946071063548, + "flos": 15012636122880.0, + "grad_norm": 1.9627370194632765, + "language_loss": 0.78004837, + "learning_rate": 1.2068339186952976e-06, + "loss": 0.80150545, + "num_input_tokens_seen": 114710410, + "step": 5330, + "time_per_iteration": 2.7154693603515625 + }, + { + "auxiliary_loss_clip": 0.01152842, + "auxiliary_loss_mlp": 0.0102849, + "balance_loss_clip": 1.01516151, + "balance_loss_mlp": 1.02077127, + "epoch": 0.6410148499969939, + "flos": 22528595496960.0, + "grad_norm": 1.7084360716941085, + "language_loss": 0.73165512, + "learning_rate": 1.2061188837969136e-06, + "loss": 0.75346851, + "num_input_tokens_seen": 114730020, + "step": 5331, + "time_per_iteration": 2.74722957611084 + }, + { + "auxiliary_loss_clip": 0.01097793, + "auxiliary_loss_mlp": 0.01024663, + "balance_loss_clip": 0.93344098, + "balance_loss_mlp": 1.01652086, + "epoch": 0.641135092887633, + "flos": 12422004537600.0, + "grad_norm": 2.5889240933450948, + "language_loss": 0.83725524, + "learning_rate": 1.2054039693296631e-06, + "loss": 0.85847986, + "num_input_tokens_seen": 114748015, + "step": 5332, + "time_per_iteration": 2.705540418624878 + }, + { + "auxiliary_loss_clip": 0.01099849, + "auxiliary_loss_mlp": 0.01025648, + "balance_loss_clip": 0.93777966, + "balance_loss_mlp": 1.01802456, + "epoch": 0.6412553357782721, + "flos": 22127329687680.0, + "grad_norm": 1.7858307794231973, + "language_loss": 0.81488335, + "learning_rate": 1.2046891754019992e-06, + "loss": 0.83613825, + "num_input_tokens_seen": 114768625, + "step": 5333, + "time_per_iteration": 2.8105390071868896 + }, + { + "auxiliary_loss_clip": 0.01152335, + "auxiliary_loss_mlp": 0.01028353, + "balance_loss_clip": 1.01532543, + "balance_loss_mlp": 1.01990402, + "epoch": 0.6413755786689112, + "flos": 15888605097600.0, + "grad_norm": 1.763312349260578, + "language_loss": 0.82833976, + "learning_rate": 1.2039745021223548e-06, + "loss": 0.85014665, + "num_input_tokens_seen": 114786045, + "step": 5334, + "time_per_iteration": 2.6712052822113037 + }, + { + "auxiliary_loss_clip": 0.00994974, + "auxiliary_loss_mlp": 0.01003955, + "balance_loss_clip": 0.87685066, + "balance_loss_mlp": 1.00266802, + "epoch": 0.6414958215595503, + "flos": 68039159955840.0, + "grad_norm": 0.801274265151911, + "language_loss": 0.57050163, + "learning_rate": 1.2032599495991456e-06, + "loss": 0.59049094, + "num_input_tokens_seen": 114850785, + "step": 5335, + "time_per_iteration": 3.5004889965057373 + }, + { + "auxiliary_loss_clip": 0.01149979, + "auxiliary_loss_mlp": 0.01028305, + "balance_loss_clip": 1.01504588, + "balance_loss_mlp": 1.02036524, + "epoch": 0.6416160644501894, + "flos": 44091300320640.0, + "grad_norm": 2.8528897683805554, + "language_loss": 0.69984692, + "learning_rate": 1.2025455179407685e-06, + "loss": 0.72162974, + "num_input_tokens_seen": 114871945, + "step": 5336, + "time_per_iteration": 2.858804225921631 + }, + { + "auxiliary_loss_clip": 0.01147289, + "auxiliary_loss_mlp": 0.00510255, + "balance_loss_clip": 1.01420534, + "balance_loss_mlp": 0.0, + "epoch": 0.6417363073408284, + "flos": 20959837931520.0, + "grad_norm": 1.8330353349202806, + "language_loss": 0.73664176, + "learning_rate": 1.2018312072556022e-06, + "loss": 0.75321722, + "num_input_tokens_seen": 114890445, + "step": 5337, + "time_per_iteration": 2.675083637237549 + }, + { + "auxiliary_loss_clip": 0.01172958, + "auxiliary_loss_mlp": 0.00510359, + "balance_loss_clip": 1.0540359, + "balance_loss_mlp": 0.0, + "epoch": 0.6418565502314676, + "flos": 22455122227200.0, + "grad_norm": 1.9475527371415868, + "language_loss": 0.74608254, + "learning_rate": 1.2011170176520077e-06, + "loss": 0.76291573, + "num_input_tokens_seen": 114911360, + "step": 5338, + "time_per_iteration": 2.6596755981445312 + }, + { + "auxiliary_loss_clip": 0.01038501, + "auxiliary_loss_mlp": 0.0102803, + "balance_loss_clip": 0.85615051, + "balance_loss_mlp": 1.02105069, + "epoch": 0.6419767931221066, + "flos": 25045502417280.0, + "grad_norm": 1.8662283354126532, + "language_loss": 0.81165934, + "learning_rate": 1.2004029492383256e-06, + "loss": 0.83232456, + "num_input_tokens_seen": 114932700, + "step": 5339, + "time_per_iteration": 2.8223960399627686 + }, + { + "auxiliary_loss_clip": 0.01151441, + "auxiliary_loss_mlp": 0.01025229, + "balance_loss_clip": 1.01795602, + "balance_loss_mlp": 1.01765895, + "epoch": 0.6420970360127457, + "flos": 19463691709440.0, + "grad_norm": 2.4083401307972965, + "language_loss": 0.74002469, + "learning_rate": 1.1996890021228814e-06, + "loss": 0.76179141, + "num_input_tokens_seen": 114949475, + "step": 5340, + "time_per_iteration": 2.720851421356201 + }, + { + "auxiliary_loss_clip": 0.01123163, + "auxiliary_loss_mlp": 0.0102557, + "balance_loss_clip": 0.97476298, + "balance_loss_mlp": 1.01808989, + "epoch": 0.6422172789033849, + "flos": 40406147458560.0, + "grad_norm": 1.5134035684222271, + "language_loss": 0.69935465, + "learning_rate": 1.1989751764139785e-06, + "loss": 0.720842, + "num_input_tokens_seen": 114973125, + "step": 5341, + "time_per_iteration": 2.895507335662842 + }, + { + "auxiliary_loss_clip": 0.01073113, + "auxiliary_loss_mlp": 0.01028999, + "balance_loss_clip": 0.8916437, + "balance_loss_mlp": 1.02149761, + "epoch": 0.6423375217940239, + "flos": 27672870637440.0, + "grad_norm": 1.527912776575971, + "language_loss": 0.83382374, + "learning_rate": 1.1982614722199044e-06, + "loss": 0.85484481, + "num_input_tokens_seen": 114994300, + "step": 5342, + "time_per_iteration": 3.841571092605591 + }, + { + "auxiliary_loss_clip": 0.01129383, + "auxiliary_loss_mlp": 0.01024821, + "balance_loss_clip": 0.9735595, + "balance_loss_mlp": 1.01765692, + "epoch": 0.642457764684663, + "flos": 18369242259840.0, + "grad_norm": 1.9816900770633339, + "language_loss": 0.77855468, + "learning_rate": 1.1975478896489276e-06, + "loss": 0.80009663, + "num_input_tokens_seen": 115012135, + "step": 5343, + "time_per_iteration": 3.616013288497925 + }, + { + "auxiliary_loss_clip": 0.01170565, + "auxiliary_loss_mlp": 0.01028064, + "balance_loss_clip": 1.05247402, + "balance_loss_mlp": 1.02060771, + "epoch": 0.6425780075753021, + "flos": 19750509809280.0, + "grad_norm": 1.9741201744022394, + "language_loss": 0.76655948, + "learning_rate": 1.1968344288092981e-06, + "loss": 0.78854573, + "num_input_tokens_seen": 115028715, + "step": 5344, + "time_per_iteration": 3.9255104064941406 + }, + { + "auxiliary_loss_clip": 0.01152737, + "auxiliary_loss_mlp": 0.00509597, + "balance_loss_clip": 1.01678085, + "balance_loss_mlp": 0.0, + "epoch": 0.6426982504659412, + "flos": 20558536208640.0, + "grad_norm": 1.80084364031226, + "language_loss": 0.64617586, + "learning_rate": 1.1961210898092468e-06, + "loss": 0.66279924, + "num_input_tokens_seen": 115047665, + "step": 5345, + "time_per_iteration": 2.74025821685791 + }, + { + "auxiliary_loss_clip": 0.01131762, + "auxiliary_loss_mlp": 0.01032976, + "balance_loss_clip": 0.97718483, + "balance_loss_mlp": 1.02531934, + "epoch": 0.6428184933565803, + "flos": 17851984456320.0, + "grad_norm": 3.430290486791262, + "language_loss": 0.79266942, + "learning_rate": 1.1954078727569874e-06, + "loss": 0.81431681, + "num_input_tokens_seen": 115064965, + "step": 5346, + "time_per_iteration": 2.779444694519043 + }, + { + "auxiliary_loss_clip": 0.01106845, + "auxiliary_loss_mlp": 0.00509598, + "balance_loss_clip": 0.9366225, + "balance_loss_mlp": 0.0, + "epoch": 0.6429387362472194, + "flos": 22456953820800.0, + "grad_norm": 1.7024604972614736, + "language_loss": 0.777964, + "learning_rate": 1.1946947777607141e-06, + "loss": 0.79412842, + "num_input_tokens_seen": 115086100, + "step": 5347, + "time_per_iteration": 2.762312650680542 + }, + { + "auxiliary_loss_clip": 0.01072501, + "auxiliary_loss_mlp": 0.0102821, + "balance_loss_clip": 0.89781952, + "balance_loss_mlp": 1.02018726, + "epoch": 0.6430589791378585, + "flos": 24752579005440.0, + "grad_norm": 1.8711335609851074, + "language_loss": 0.80109298, + "learning_rate": 1.1939818049286024e-06, + "loss": 0.82210004, + "num_input_tokens_seen": 115104260, + "step": 5348, + "time_per_iteration": 2.8426454067230225 + }, + { + "auxiliary_loss_clip": 0.01045282, + "auxiliary_loss_mlp": 0.01027635, + "balance_loss_clip": 0.85801512, + "balance_loss_mlp": 1.02036977, + "epoch": 0.6431792220284975, + "flos": 24901249397760.0, + "grad_norm": 1.5071095198228204, + "language_loss": 0.75445628, + "learning_rate": 1.1932689543688101e-06, + "loss": 0.77518541, + "num_input_tokens_seen": 115125365, + "step": 5349, + "time_per_iteration": 2.9049577713012695 + }, + { + "auxiliary_loss_clip": 0.01122889, + "auxiliary_loss_mlp": 0.01031561, + "balance_loss_clip": 0.97618157, + "balance_loss_mlp": 1.02380693, + "epoch": 0.6432994649191367, + "flos": 21032305620480.0, + "grad_norm": 1.7454247437601722, + "language_loss": 0.72420424, + "learning_rate": 1.1925562261894756e-06, + "loss": 0.74574876, + "num_input_tokens_seen": 115144445, + "step": 5350, + "time_per_iteration": 2.68815279006958 + }, + { + "auxiliary_loss_clip": 0.01121244, + "auxiliary_loss_mlp": 0.01024027, + "balance_loss_clip": 0.97364831, + "balance_loss_mlp": 1.01650476, + "epoch": 0.6434197078097758, + "flos": 30884433655680.0, + "grad_norm": 1.826475004711178, + "language_loss": 0.77344197, + "learning_rate": 1.1918436204987207e-06, + "loss": 0.7948947, + "num_input_tokens_seen": 115166305, + "step": 5351, + "time_per_iteration": 2.836421012878418 + }, + { + "auxiliary_loss_clip": 0.01150448, + "auxiliary_loss_mlp": 0.01028317, + "balance_loss_clip": 1.02031565, + "balance_loss_mlp": 1.02083373, + "epoch": 0.6435399507004148, + "flos": 15012492468480.0, + "grad_norm": 2.10785500440863, + "language_loss": 0.81378567, + "learning_rate": 1.191131137404645e-06, + "loss": 0.83557332, + "num_input_tokens_seen": 115183045, + "step": 5352, + "time_per_iteration": 2.647756338119507 + }, + { + "auxiliary_loss_clip": 0.01092121, + "auxiliary_loss_mlp": 0.01032003, + "balance_loss_clip": 0.93538415, + "balance_loss_mlp": 1.02460003, + "epoch": 0.643660193591054, + "flos": 19901981462400.0, + "grad_norm": 1.829890186836411, + "language_loss": 0.76812279, + "learning_rate": 1.190418777015333e-06, + "loss": 0.78936398, + "num_input_tokens_seen": 115201955, + "step": 5353, + "time_per_iteration": 2.7774782180786133 + }, + { + "auxiliary_loss_clip": 0.01125346, + "auxiliary_loss_mlp": 0.01026763, + "balance_loss_clip": 0.97575516, + "balance_loss_mlp": 1.01930666, + "epoch": 0.643780436481693, + "flos": 24133622820480.0, + "grad_norm": 1.4545213654323412, + "language_loss": 0.73761833, + "learning_rate": 1.1897065394388487e-06, + "loss": 0.75913942, + "num_input_tokens_seen": 115222395, + "step": 5354, + "time_per_iteration": 3.6527462005615234 + }, + { + "auxiliary_loss_clip": 0.01128212, + "auxiliary_loss_mlp": 0.01033044, + "balance_loss_clip": 0.98110688, + "balance_loss_mlp": 1.02527761, + "epoch": 0.6439006793723321, + "flos": 23148808657920.0, + "grad_norm": 1.7061860375160476, + "language_loss": 0.76658762, + "learning_rate": 1.1889944247832385e-06, + "loss": 0.78820026, + "num_input_tokens_seen": 115242635, + "step": 5355, + "time_per_iteration": 2.8766613006591797 + }, + { + "auxiliary_loss_clip": 0.01150238, + "auxiliary_loss_mlp": 0.01033344, + "balance_loss_clip": 1.01169264, + "balance_loss_mlp": 1.02486253, + "epoch": 0.6440209222629713, + "flos": 23617909301760.0, + "grad_norm": 1.9464291260318565, + "language_loss": 0.70784593, + "learning_rate": 1.1882824331565283e-06, + "loss": 0.72968173, + "num_input_tokens_seen": 115262095, + "step": 5356, + "time_per_iteration": 2.735563039779663 + }, + { + "auxiliary_loss_clip": 0.0109702, + "auxiliary_loss_mlp": 0.01025455, + "balance_loss_clip": 0.93442756, + "balance_loss_mlp": 1.01800179, + "epoch": 0.6441411651536103, + "flos": 16544872535040.0, + "grad_norm": 6.467395887643897, + "language_loss": 0.89189512, + "learning_rate": 1.1875705646667287e-06, + "loss": 0.91311979, + "num_input_tokens_seen": 115279985, + "step": 5357, + "time_per_iteration": 2.671252727508545 + }, + { + "auxiliary_loss_clip": 0.01142133, + "auxiliary_loss_mlp": 0.01023093, + "balance_loss_clip": 1.0092243, + "balance_loss_mlp": 1.01520145, + "epoch": 0.6442614080442494, + "flos": 25410965345280.0, + "grad_norm": 4.520546779802075, + "language_loss": 0.75137234, + "learning_rate": 1.1868588194218282e-06, + "loss": 0.77302468, + "num_input_tokens_seen": 115300365, + "step": 5358, + "time_per_iteration": 2.7022931575775146 + }, + { + "auxiliary_loss_clip": 0.01130659, + "auxiliary_loss_mlp": 0.0102654, + "balance_loss_clip": 0.97494358, + "balance_loss_mlp": 1.01947069, + "epoch": 0.6443816509348885, + "flos": 28294017552000.0, + "grad_norm": 1.5219786881886908, + "language_loss": 0.74223661, + "learning_rate": 1.1861471975297979e-06, + "loss": 0.76380861, + "num_input_tokens_seen": 115322060, + "step": 5359, + "time_per_iteration": 2.7251455783843994 + }, + { + "auxiliary_loss_clip": 0.01098819, + "auxiliary_loss_mlp": 0.01030339, + "balance_loss_clip": 0.93831873, + "balance_loss_mlp": 1.02253127, + "epoch": 0.6445018938255276, + "flos": 36690075964800.0, + "grad_norm": 1.6391507864210397, + "language_loss": 0.70836592, + "learning_rate": 1.185435699098591e-06, + "loss": 0.72965753, + "num_input_tokens_seen": 115348255, + "step": 5360, + "time_per_iteration": 2.9498252868652344 + }, + { + "auxiliary_loss_clip": 0.01130825, + "auxiliary_loss_mlp": 0.01033393, + "balance_loss_clip": 0.97548097, + "balance_loss_mlp": 1.02558756, + "epoch": 0.6446221367161666, + "flos": 14501411804160.0, + "grad_norm": 2.203335645737941, + "language_loss": 0.78819132, + "learning_rate": 1.1847243242361403e-06, + "loss": 0.80983347, + "num_input_tokens_seen": 115366845, + "step": 5361, + "time_per_iteration": 2.6457581520080566 + }, + { + "auxiliary_loss_clip": 0.01129575, + "auxiliary_loss_mlp": 0.01025636, + "balance_loss_clip": 0.97874701, + "balance_loss_mlp": 1.0174886, + "epoch": 0.6447423796068057, + "flos": 24609367480320.0, + "grad_norm": 1.6105165429590182, + "language_loss": 0.778629, + "learning_rate": 1.1840130730503624e-06, + "loss": 0.80018115, + "num_input_tokens_seen": 115388125, + "step": 5362, + "time_per_iteration": 2.684058427810669 + }, + { + "auxiliary_loss_clip": 0.01175986, + "auxiliary_loss_mlp": 0.01029677, + "balance_loss_clip": 1.0551548, + "balance_loss_mlp": 1.02259851, + "epoch": 0.6448626224974449, + "flos": 25047298097280.0, + "grad_norm": 2.9392489694245776, + "language_loss": 0.75101405, + "learning_rate": 1.1833019456491518e-06, + "loss": 0.77307069, + "num_input_tokens_seen": 115409655, + "step": 5363, + "time_per_iteration": 2.620473623275757 + }, + { + "auxiliary_loss_clip": 0.01149651, + "auxiliary_loss_mlp": 0.01030933, + "balance_loss_clip": 1.014714, + "balance_loss_mlp": 1.02328527, + "epoch": 0.6449828653880839, + "flos": 22530355263360.0, + "grad_norm": 8.388411723377239, + "language_loss": 0.7895239, + "learning_rate": 1.1825909421403871e-06, + "loss": 0.81132972, + "num_input_tokens_seen": 115428750, + "step": 5364, + "time_per_iteration": 2.7069432735443115 + }, + { + "auxiliary_loss_clip": 0.01149634, + "auxiliary_loss_mlp": 0.01026923, + "balance_loss_clip": 1.01531243, + "balance_loss_mlp": 1.01958001, + "epoch": 0.645103108278723, + "flos": 25695736369920.0, + "grad_norm": 1.8579906685184653, + "language_loss": 0.76228118, + "learning_rate": 1.181880062631926e-06, + "loss": 0.78404677, + "num_input_tokens_seen": 115448085, + "step": 5365, + "time_per_iteration": 2.6893410682678223 + }, + { + "auxiliary_loss_clip": 0.01122414, + "auxiliary_loss_mlp": 0.01026503, + "balance_loss_clip": 0.97709894, + "balance_loss_mlp": 1.018188, + "epoch": 0.6452233511693621, + "flos": 27450331925760.0, + "grad_norm": 2.1387243298858354, + "language_loss": 0.84801686, + "learning_rate": 1.1811693072316093e-06, + "loss": 0.86950594, + "num_input_tokens_seen": 115465765, + "step": 5366, + "time_per_iteration": 2.732421398162842 + }, + { + "auxiliary_loss_clip": 0.01172568, + "auxiliary_loss_mlp": 0.0050997, + "balance_loss_clip": 1.05289626, + "balance_loss_mlp": 0.0, + "epoch": 0.6453435940600012, + "flos": 19208618254080.0, + "grad_norm": 2.30608447581544, + "language_loss": 0.83745766, + "learning_rate": 1.1804586760472574e-06, + "loss": 0.85428309, + "num_input_tokens_seen": 115482230, + "step": 5367, + "time_per_iteration": 2.5853278636932373 + }, + { + "auxiliary_loss_clip": 0.01098471, + "auxiliary_loss_mlp": 0.01026065, + "balance_loss_clip": 0.93466675, + "balance_loss_mlp": 1.01880527, + "epoch": 0.6454638369506402, + "flos": 25737680476800.0, + "grad_norm": 2.324631591098355, + "language_loss": 0.8078959, + "learning_rate": 1.1797481691866736e-06, + "loss": 0.82914126, + "num_input_tokens_seen": 115499455, + "step": 5368, + "time_per_iteration": 3.5800676345825195 + }, + { + "auxiliary_loss_clip": 0.01120153, + "auxiliary_loss_mlp": 0.01030713, + "balance_loss_clip": 0.97955644, + "balance_loss_mlp": 1.02314305, + "epoch": 0.6455840798412794, + "flos": 20989176364800.0, + "grad_norm": 2.1751203251373634, + "language_loss": 0.83226764, + "learning_rate": 1.1790377867576393e-06, + "loss": 0.85377634, + "num_input_tokens_seen": 115517205, + "step": 5369, + "time_per_iteration": 3.3453166484832764 + }, + { + "auxiliary_loss_clip": 0.01127219, + "auxiliary_loss_mlp": 0.01021878, + "balance_loss_clip": 0.97437942, + "balance_loss_mlp": 1.01457047, + "epoch": 0.6457043227319185, + "flos": 26067556005120.0, + "grad_norm": 1.762976352674998, + "language_loss": 0.76733798, + "learning_rate": 1.1783275288679203e-06, + "loss": 0.78882897, + "num_input_tokens_seen": 115534370, + "step": 5370, + "time_per_iteration": 3.5442092418670654 + }, + { + "auxiliary_loss_clip": 0.01052086, + "auxiliary_loss_mlp": 0.0099911, + "balance_loss_clip": 0.97896636, + "balance_loss_mlp": 0.99809116, + "epoch": 0.6458245656225575, + "flos": 60370831088640.0, + "grad_norm": 0.8440457477543319, + "language_loss": 0.57136071, + "learning_rate": 1.177617395625262e-06, + "loss": 0.59187269, + "num_input_tokens_seen": 115592345, + "step": 5371, + "time_per_iteration": 3.1522557735443115 + }, + { + "auxiliary_loss_clip": 0.01149329, + "auxiliary_loss_mlp": 0.01023371, + "balance_loss_clip": 1.01448631, + "balance_loss_mlp": 1.01603055, + "epoch": 0.6459448085131967, + "flos": 23076771932160.0, + "grad_norm": 1.9451108801988857, + "language_loss": 0.75604129, + "learning_rate": 1.1769073871373908e-06, + "loss": 0.77776825, + "num_input_tokens_seen": 115612550, + "step": 5372, + "time_per_iteration": 2.7075254917144775 + }, + { + "auxiliary_loss_clip": 0.0109857, + "auxiliary_loss_mlp": 0.01026085, + "balance_loss_clip": 0.93361539, + "balance_loss_mlp": 1.01899767, + "epoch": 0.6460650514038357, + "flos": 22598190097920.0, + "grad_norm": 1.6669439147877503, + "language_loss": 0.83970022, + "learning_rate": 1.176197503512015e-06, + "loss": 0.86094677, + "num_input_tokens_seen": 115632265, + "step": 5373, + "time_per_iteration": 2.721552848815918 + }, + { + "auxiliary_loss_clip": 0.01120559, + "auxiliary_loss_mlp": 0.01033261, + "balance_loss_clip": 0.97372139, + "balance_loss_mlp": 1.02614403, + "epoch": 0.6461852942944748, + "flos": 20266726118400.0, + "grad_norm": 2.1591740086803846, + "language_loss": 0.82367671, + "learning_rate": 1.1754877448568223e-06, + "loss": 0.8452149, + "num_input_tokens_seen": 115651720, + "step": 5374, + "time_per_iteration": 2.6772806644439697 + }, + { + "auxiliary_loss_clip": 0.01127294, + "auxiliary_loss_mlp": 0.01023951, + "balance_loss_clip": 0.97618687, + "balance_loss_mlp": 1.01664901, + "epoch": 0.646305537185114, + "flos": 23367109564800.0, + "grad_norm": 1.9886741565653958, + "language_loss": 0.90101528, + "learning_rate": 1.1747781112794837e-06, + "loss": 0.92252767, + "num_input_tokens_seen": 115668215, + "step": 5375, + "time_per_iteration": 2.6794931888580322 + }, + { + "auxiliary_loss_clip": 0.01101029, + "auxiliary_loss_mlp": 0.01033659, + "balance_loss_clip": 0.93829203, + "balance_loss_mlp": 1.0264765, + "epoch": 0.646425780075753, + "flos": 24277480790400.0, + "grad_norm": 1.6341456424323544, + "language_loss": 0.83095509, + "learning_rate": 1.1740686028876487e-06, + "loss": 0.85230196, + "num_input_tokens_seen": 115687080, + "step": 5376, + "time_per_iteration": 2.7303214073181152 + }, + { + "auxiliary_loss_clip": 0.01145531, + "auxiliary_loss_mlp": 0.01030722, + "balance_loss_clip": 1.01495445, + "balance_loss_mlp": 1.02327108, + "epoch": 0.6465460229663921, + "flos": 20813968800000.0, + "grad_norm": 2.3264479820836725, + "language_loss": 0.74670899, + "learning_rate": 1.1733592197889507e-06, + "loss": 0.76847148, + "num_input_tokens_seen": 115703990, + "step": 5377, + "time_per_iteration": 2.6581406593322754 + }, + { + "auxiliary_loss_clip": 0.01145834, + "auxiliary_loss_mlp": 0.01022306, + "balance_loss_clip": 1.01695383, + "balance_loss_mlp": 1.01496315, + "epoch": 0.6466662658570312, + "flos": 22853299466880.0, + "grad_norm": 1.735985459468143, + "language_loss": 0.72708189, + "learning_rate": 1.1726499620910014e-06, + "loss": 0.74876332, + "num_input_tokens_seen": 115724270, + "step": 5378, + "time_per_iteration": 2.920701026916504 + }, + { + "auxiliary_loss_clip": 0.01144377, + "auxiliary_loss_mlp": 0.01028991, + "balance_loss_clip": 1.01294363, + "balance_loss_mlp": 1.02093863, + "epoch": 0.6467865087476703, + "flos": 15304553953920.0, + "grad_norm": 1.9921344329376085, + "language_loss": 0.77666396, + "learning_rate": 1.1719408299013955e-06, + "loss": 0.79839766, + "num_input_tokens_seen": 115742995, + "step": 5379, + "time_per_iteration": 2.709540605545044 + }, + { + "auxiliary_loss_clip": 0.01171339, + "auxiliary_loss_mlp": 0.01026137, + "balance_loss_clip": 1.05348802, + "balance_loss_mlp": 1.01884174, + "epoch": 0.6469067516383094, + "flos": 19573650218880.0, + "grad_norm": 2.4700348801911542, + "language_loss": 0.75433254, + "learning_rate": 1.1712318233277067e-06, + "loss": 0.77630734, + "num_input_tokens_seen": 115762015, + "step": 5380, + "time_per_iteration": 2.6566615104675293 + }, + { + "auxiliary_loss_clip": 0.01051653, + "auxiliary_loss_mlp": 0.01002436, + "balance_loss_clip": 0.9804033, + "balance_loss_mlp": 1.001405, + "epoch": 0.6470269945289485, + "flos": 65098002522240.0, + "grad_norm": 0.7579978960539503, + "language_loss": 0.57814783, + "learning_rate": 1.1705229424774916e-06, + "loss": 0.59868872, + "num_input_tokens_seen": 115816285, + "step": 5381, + "time_per_iteration": 3.9494760036468506 + }, + { + "auxiliary_loss_clip": 0.01122814, + "auxiliary_loss_mlp": 0.01022714, + "balance_loss_clip": 0.9758684, + "balance_loss_mlp": 1.01571703, + "epoch": 0.6471472374195876, + "flos": 30696943639680.0, + "grad_norm": 1.8872283120697941, + "language_loss": 0.64213306, + "learning_rate": 1.1698141874582867e-06, + "loss": 0.66358829, + "num_input_tokens_seen": 115837330, + "step": 5382, + "time_per_iteration": 2.7379279136657715 + }, + { + "auxiliary_loss_clip": 0.01169712, + "auxiliary_loss_mlp": 0.01025051, + "balance_loss_clip": 1.05324328, + "balance_loss_mlp": 1.01776171, + "epoch": 0.6472674803102266, + "flos": 20521835487360.0, + "grad_norm": 2.3720166621618413, + "language_loss": 0.72203219, + "learning_rate": 1.169105558377609e-06, + "loss": 0.74397981, + "num_input_tokens_seen": 115857420, + "step": 5383, + "time_per_iteration": 2.649667501449585 + }, + { + "auxiliary_loss_clip": 0.01081871, + "auxiliary_loss_mlp": 0.00509327, + "balance_loss_clip": 0.90547383, + "balance_loss_mlp": 0.0, + "epoch": 0.6473877232008658, + "flos": 24715447320960.0, + "grad_norm": 1.686111966263383, + "language_loss": 0.78740573, + "learning_rate": 1.1683970553429587e-06, + "loss": 0.80331767, + "num_input_tokens_seen": 115878875, + "step": 5384, + "time_per_iteration": 2.776341676712036 + }, + { + "auxiliary_loss_clip": 0.01105275, + "auxiliary_loss_mlp": 0.01034313, + "balance_loss_clip": 0.93669426, + "balance_loss_mlp": 1.02654362, + "epoch": 0.6475079660915048, + "flos": 15885552441600.0, + "grad_norm": 2.538014398265425, + "language_loss": 0.8204357, + "learning_rate": 1.1676886784618128e-06, + "loss": 0.84183162, + "num_input_tokens_seen": 115895540, + "step": 5385, + "time_per_iteration": 2.7370548248291016 + }, + { + "auxiliary_loss_clip": 0.01147704, + "auxiliary_loss_mlp": 0.01024275, + "balance_loss_clip": 1.01444948, + "balance_loss_mlp": 1.01624024, + "epoch": 0.6476282089821439, + "flos": 17381590922880.0, + "grad_norm": 2.208068810679463, + "language_loss": 0.8389641, + "learning_rate": 1.1669804278416332e-06, + "loss": 0.86068392, + "num_input_tokens_seen": 115910265, + "step": 5386, + "time_per_iteration": 2.664135694503784 + }, + { + "auxiliary_loss_clip": 0.01132142, + "auxiliary_loss_mlp": 0.010222, + "balance_loss_clip": 0.97782654, + "balance_loss_mlp": 1.01410556, + "epoch": 0.6477484518727831, + "flos": 20194078861440.0, + "grad_norm": 2.4885694375874605, + "language_loss": 0.716232, + "learning_rate": 1.1662723035898602e-06, + "loss": 0.73777544, + "num_input_tokens_seen": 115930025, + "step": 5387, + "time_per_iteration": 2.679854393005371 + }, + { + "auxiliary_loss_clip": 0.01148607, + "auxiliary_loss_mlp": 0.01029275, + "balance_loss_clip": 1.01549625, + "balance_loss_mlp": 1.02135921, + "epoch": 0.6478686947634221, + "flos": 25410426641280.0, + "grad_norm": 1.8483877173849819, + "language_loss": 0.8188647, + "learning_rate": 1.165564305813915e-06, + "loss": 0.84064353, + "num_input_tokens_seen": 115949025, + "step": 5388, + "time_per_iteration": 2.735970973968506 + }, + { + "auxiliary_loss_clip": 0.01147743, + "auxiliary_loss_mlp": 0.01024921, + "balance_loss_clip": 1.01491213, + "balance_loss_mlp": 1.01770329, + "epoch": 0.6479889376540612, + "flos": 20083581648000.0, + "grad_norm": 2.0040282540194054, + "language_loss": 0.81305486, + "learning_rate": 1.1648564346212019e-06, + "loss": 0.83478153, + "num_input_tokens_seen": 115968145, + "step": 5389, + "time_per_iteration": 2.6772522926330566 + }, + { + "auxiliary_loss_clip": 0.01146224, + "auxiliary_loss_mlp": 0.0103025, + "balance_loss_clip": 1.01665258, + "balance_loss_mlp": 1.02324605, + "epoch": 0.6481091805447003, + "flos": 26758082039040.0, + "grad_norm": 3.9883779298248645, + "language_loss": 0.75942886, + "learning_rate": 1.164148690119104e-06, + "loss": 0.78119361, + "num_input_tokens_seen": 115989425, + "step": 5390, + "time_per_iteration": 2.7591159343719482 + }, + { + "auxiliary_loss_clip": 0.01170098, + "auxiliary_loss_mlp": 0.01027217, + "balance_loss_clip": 1.05266917, + "balance_loss_mlp": 1.02044916, + "epoch": 0.6482294234353394, + "flos": 23952094462080.0, + "grad_norm": 1.819492480444657, + "language_loss": 0.73839444, + "learning_rate": 1.163441072414985e-06, + "loss": 0.76036763, + "num_input_tokens_seen": 116009630, + "step": 5391, + "time_per_iteration": 2.617182970046997 + }, + { + "auxiliary_loss_clip": 0.01150315, + "auxiliary_loss_mlp": 0.01025614, + "balance_loss_clip": 1.0161134, + "balance_loss_mlp": 1.01788902, + "epoch": 0.6483496663259785, + "flos": 26209833776640.0, + "grad_norm": 1.8222178617276357, + "language_loss": 0.69648767, + "learning_rate": 1.16273358161619e-06, + "loss": 0.71824694, + "num_input_tokens_seen": 116029965, + "step": 5392, + "time_per_iteration": 2.7892651557922363 + }, + { + "auxiliary_loss_clip": 0.01135043, + "auxiliary_loss_mlp": 0.01027413, + "balance_loss_clip": 0.97853315, + "balance_loss_mlp": 1.01920533, + "epoch": 0.6484699092166175, + "flos": 20922239370240.0, + "grad_norm": 1.8879824868425898, + "language_loss": 0.83446455, + "learning_rate": 1.1620262178300446e-06, + "loss": 0.85608912, + "num_input_tokens_seen": 116048580, + "step": 5393, + "time_per_iteration": 2.7089953422546387 + }, + { + "auxiliary_loss_clip": 0.01101333, + "auxiliary_loss_mlp": 0.01030883, + "balance_loss_clip": 0.93446249, + "balance_loss_mlp": 1.02378106, + "epoch": 0.6485901521072567, + "flos": 33072865678080.0, + "grad_norm": 2.1880585352633344, + "language_loss": 0.75721198, + "learning_rate": 1.1613189811638563e-06, + "loss": 0.77853417, + "num_input_tokens_seen": 116070305, + "step": 5394, + "time_per_iteration": 3.698200225830078 + }, + { + "auxiliary_loss_clip": 0.01149609, + "auxiliary_loss_mlp": 0.01026572, + "balance_loss_clip": 1.0153414, + "balance_loss_mlp": 1.01894867, + "epoch": 0.6487103949978957, + "flos": 22274060745600.0, + "grad_norm": 1.6402898782351045, + "language_loss": 0.78099144, + "learning_rate": 1.1606118717249117e-06, + "loss": 0.80275321, + "num_input_tokens_seen": 116090405, + "step": 5395, + "time_per_iteration": 3.5570242404937744 + }, + { + "auxiliary_loss_clip": 0.01173936, + "auxiliary_loss_mlp": 0.01027351, + "balance_loss_clip": 1.05170584, + "balance_loss_mlp": 1.0197041, + "epoch": 0.6488306378885348, + "flos": 22930400010240.0, + "grad_norm": 1.6443122029201664, + "language_loss": 0.67900229, + "learning_rate": 1.1599048896204787e-06, + "loss": 0.70101511, + "num_input_tokens_seen": 116110285, + "step": 5396, + "time_per_iteration": 3.6952311992645264 + }, + { + "auxiliary_loss_clip": 0.01109295, + "auxiliary_loss_mlp": 0.01032861, + "balance_loss_clip": 0.94172084, + "balance_loss_mlp": 1.02481437, + "epoch": 0.648950880779174, + "flos": 20376110010240.0, + "grad_norm": 1.7729654231582528, + "language_loss": 0.80925524, + "learning_rate": 1.1591980349578061e-06, + "loss": 0.83067679, + "num_input_tokens_seen": 116128955, + "step": 5397, + "time_per_iteration": 2.7714803218841553 + }, + { + "auxiliary_loss_clip": 0.01011866, + "auxiliary_loss_mlp": 0.01006177, + "balance_loss_clip": 0.90698767, + "balance_loss_mlp": 1.00510454, + "epoch": 0.649071123669813, + "flos": 59930889310080.0, + "grad_norm": 0.7341446319013041, + "language_loss": 0.54329312, + "learning_rate": 1.158491307844123e-06, + "loss": 0.56347358, + "num_input_tokens_seen": 116188875, + "step": 5398, + "time_per_iteration": 3.3410134315490723 + }, + { + "auxiliary_loss_clip": 0.01123936, + "auxiliary_loss_mlp": 0.0102678, + "balance_loss_clip": 0.97611308, + "balance_loss_mlp": 1.01928425, + "epoch": 0.6491913665604521, + "flos": 20446566537600.0, + "grad_norm": 1.8230193873541676, + "language_loss": 0.83845353, + "learning_rate": 1.1577847083866387e-06, + "loss": 0.85996068, + "num_input_tokens_seen": 116207910, + "step": 5399, + "time_per_iteration": 2.7592668533325195 + }, + { + "auxiliary_loss_clip": 0.01114095, + "auxiliary_loss_mlp": 0.01023024, + "balance_loss_clip": 0.97268116, + "balance_loss_mlp": 1.01513243, + "epoch": 0.6493116094510912, + "flos": 16946820702720.0, + "grad_norm": 2.2144195804017275, + "language_loss": 0.72235334, + "learning_rate": 1.1570782366925453e-06, + "loss": 0.74372452, + "num_input_tokens_seen": 116226425, + "step": 5400, + "time_per_iteration": 2.793898344039917 + }, + { + "auxiliary_loss_clip": 0.01125479, + "auxiliary_loss_mlp": 0.01031264, + "balance_loss_clip": 0.97321737, + "balance_loss_mlp": 1.02371764, + "epoch": 0.6494318523417303, + "flos": 18802935072000.0, + "grad_norm": 1.631832232668242, + "language_loss": 0.75965476, + "learning_rate": 1.1563718928690132e-06, + "loss": 0.78122222, + "num_input_tokens_seen": 116243860, + "step": 5401, + "time_per_iteration": 2.6930737495422363 + }, + { + "auxiliary_loss_clip": 0.01101349, + "auxiliary_loss_mlp": 0.01031053, + "balance_loss_clip": 0.938932, + "balance_loss_mlp": 1.02341175, + "epoch": 0.6495520952323693, + "flos": 18982847318400.0, + "grad_norm": 1.910038543694806, + "language_loss": 0.71597517, + "learning_rate": 1.1556656770231942e-06, + "loss": 0.7372992, + "num_input_tokens_seen": 116260055, + "step": 5402, + "time_per_iteration": 2.653193950653076 + }, + { + "auxiliary_loss_clip": 0.01149192, + "auxiliary_loss_mlp": 0.01025754, + "balance_loss_clip": 1.01320171, + "balance_loss_mlp": 1.01832104, + "epoch": 0.6496723381230085, + "flos": 22745388032640.0, + "grad_norm": 1.5195050951402014, + "language_loss": 0.7617653, + "learning_rate": 1.1549595892622207e-06, + "loss": 0.78351474, + "num_input_tokens_seen": 116278825, + "step": 5403, + "time_per_iteration": 2.6900453567504883 + }, + { + "auxiliary_loss_clip": 0.00990608, + "auxiliary_loss_mlp": 0.01005994, + "balance_loss_clip": 0.87376535, + "balance_loss_mlp": 1.00480199, + "epoch": 0.6497925810136476, + "flos": 62145283887360.0, + "grad_norm": 0.8259732774592079, + "language_loss": 0.59024513, + "learning_rate": 1.1542536296932047e-06, + "loss": 0.61021113, + "num_input_tokens_seen": 116342360, + "step": 5404, + "time_per_iteration": 3.2367727756500244 + }, + { + "auxiliary_loss_clip": 0.01107773, + "auxiliary_loss_mlp": 0.0102443, + "balance_loss_clip": 0.93564939, + "balance_loss_mlp": 1.0170033, + "epoch": 0.6499128239042866, + "flos": 20156731695360.0, + "grad_norm": 2.0216186479969083, + "language_loss": 0.70182383, + "learning_rate": 1.1535477984232414e-06, + "loss": 0.72314584, + "num_input_tokens_seen": 116362235, + "step": 5405, + "time_per_iteration": 2.8176958560943604 + }, + { + "auxiliary_loss_clip": 0.01081542, + "auxiliary_loss_mlp": 0.01029722, + "balance_loss_clip": 0.89444292, + "balance_loss_mlp": 1.02177048, + "epoch": 0.6500330667949258, + "flos": 24462420940800.0, + "grad_norm": 1.9137095078133892, + "language_loss": 0.77050352, + "learning_rate": 1.152842095559404e-06, + "loss": 0.79161614, + "num_input_tokens_seen": 116382895, + "step": 5406, + "time_per_iteration": 3.815377712249756 + }, + { + "auxiliary_loss_clip": 0.01129588, + "auxiliary_loss_mlp": 0.0102604, + "balance_loss_clip": 0.97533035, + "balance_loss_mlp": 1.0192095, + "epoch": 0.6501533096855648, + "flos": 25477399549440.0, + "grad_norm": 1.7002524031312303, + "language_loss": 0.7703715, + "learning_rate": 1.1521365212087474e-06, + "loss": 0.79192775, + "num_input_tokens_seen": 116402880, + "step": 5407, + "time_per_iteration": 2.6788077354431152 + }, + { + "auxiliary_loss_clip": 0.01149556, + "auxiliary_loss_mlp": 0.01031383, + "balance_loss_clip": 1.01379621, + "balance_loss_mlp": 1.02392995, + "epoch": 0.6502735525762039, + "flos": 44819245347840.0, + "grad_norm": 1.5502049915344622, + "language_loss": 0.70423061, + "learning_rate": 1.1514310754783062e-06, + "loss": 0.72604001, + "num_input_tokens_seen": 116425830, + "step": 5408, + "time_per_iteration": 2.908597230911255 + }, + { + "auxiliary_loss_clip": 0.01124285, + "auxiliary_loss_mlp": 0.01027574, + "balance_loss_clip": 0.97391111, + "balance_loss_mlp": 1.02026033, + "epoch": 0.6503937954668431, + "flos": 28658546726400.0, + "grad_norm": 2.3649882970099587, + "language_loss": 0.73567295, + "learning_rate": 1.1507257584750964e-06, + "loss": 0.75719154, + "num_input_tokens_seen": 116446010, + "step": 5409, + "time_per_iteration": 2.7554819583892822 + }, + { + "auxiliary_loss_clip": 0.01175019, + "auxiliary_loss_mlp": 0.01035166, + "balance_loss_clip": 1.05538654, + "balance_loss_mlp": 1.02775371, + "epoch": 0.6505140383574821, + "flos": 20922562592640.0, + "grad_norm": 1.9520482487922544, + "language_loss": 0.7718398, + "learning_rate": 1.150020570306113e-06, + "loss": 0.79394162, + "num_input_tokens_seen": 116465150, + "step": 5410, + "time_per_iteration": 2.684544324874878 + }, + { + "auxiliary_loss_clip": 0.01117401, + "auxiliary_loss_mlp": 0.01027765, + "balance_loss_clip": 0.9708814, + "balance_loss_mlp": 1.02018881, + "epoch": 0.6506342812481212, + "flos": 20595236929920.0, + "grad_norm": 1.64183216386692, + "language_loss": 0.75032371, + "learning_rate": 1.1493155110783338e-06, + "loss": 0.77177536, + "num_input_tokens_seen": 116483675, + "step": 5411, + "time_per_iteration": 2.7024571895599365 + }, + { + "auxiliary_loss_clip": 0.01149481, + "auxiliary_loss_mlp": 0.01025899, + "balance_loss_clip": 1.01500797, + "balance_loss_mlp": 1.01832366, + "epoch": 0.6507545241387603, + "flos": 30226478279040.0, + "grad_norm": 2.077607647329098, + "language_loss": 0.70855224, + "learning_rate": 1.1486105808987155e-06, + "loss": 0.73030603, + "num_input_tokens_seen": 116505165, + "step": 5412, + "time_per_iteration": 2.751595973968506 + }, + { + "auxiliary_loss_clip": 0.01149608, + "auxiliary_loss_mlp": 0.01025089, + "balance_loss_clip": 1.01518536, + "balance_loss_mlp": 1.01736116, + "epoch": 0.6508747670293994, + "flos": 17128241320320.0, + "grad_norm": 1.8511425161452455, + "language_loss": 0.80848253, + "learning_rate": 1.1479057798741947e-06, + "loss": 0.83022952, + "num_input_tokens_seen": 116523220, + "step": 5413, + "time_per_iteration": 2.625433921813965 + }, + { + "auxiliary_loss_clip": 0.01044388, + "auxiliary_loss_mlp": 0.01004548, + "balance_loss_clip": 0.96091008, + "balance_loss_mlp": 1.00321925, + "epoch": 0.6509950099200384, + "flos": 68559826573440.0, + "grad_norm": 0.7812843784541273, + "language_loss": 0.53315866, + "learning_rate": 1.14720110811169e-06, + "loss": 0.55364799, + "num_input_tokens_seen": 116580450, + "step": 5414, + "time_per_iteration": 3.2634973526000977 + }, + { + "auxiliary_loss_clip": 0.01153637, + "auxiliary_loss_mlp": 0.01031317, + "balance_loss_clip": 1.0154115, + "balance_loss_mlp": 1.0233593, + "epoch": 0.6511152528106776, + "flos": 22347462188160.0, + "grad_norm": 2.6883163734160274, + "language_loss": 0.76985669, + "learning_rate": 1.146496565718098e-06, + "loss": 0.7917062, + "num_input_tokens_seen": 116601020, + "step": 5415, + "time_per_iteration": 2.7198290824890137 + }, + { + "auxiliary_loss_clip": 0.01127496, + "auxiliary_loss_mlp": 0.01024562, + "balance_loss_clip": 0.98016393, + "balance_loss_mlp": 1.01731396, + "epoch": 0.6512354957013167, + "flos": 20522158709760.0, + "grad_norm": 1.915380505808932, + "language_loss": 0.7562176, + "learning_rate": 1.1457921528002996e-06, + "loss": 0.77773809, + "num_input_tokens_seen": 116619455, + "step": 5416, + "time_per_iteration": 2.7090389728546143 + }, + { + "auxiliary_loss_clip": 0.01171113, + "auxiliary_loss_mlp": 0.00509958, + "balance_loss_clip": 1.05182314, + "balance_loss_mlp": 0.0, + "epoch": 0.6513557385919557, + "flos": 32337342881280.0, + "grad_norm": 2.5205830352190883, + "language_loss": 0.72740436, + "learning_rate": 1.1450878694651522e-06, + "loss": 0.74421501, + "num_input_tokens_seen": 116640020, + "step": 5417, + "time_per_iteration": 2.7337777614593506 + }, + { + "auxiliary_loss_clip": 0.01079211, + "auxiliary_loss_mlp": 0.01029292, + "balance_loss_clip": 0.89853704, + "balance_loss_mlp": 1.02148938, + "epoch": 0.6514759814825949, + "flos": 12093206417280.0, + "grad_norm": 2.4155001154390097, + "language_loss": 0.6356045, + "learning_rate": 1.1443837158194954e-06, + "loss": 0.65668952, + "num_input_tokens_seen": 116655165, + "step": 5418, + "time_per_iteration": 2.707428455352783 + }, + { + "auxiliary_loss_clip": 0.01103379, + "auxiliary_loss_mlp": 0.01030779, + "balance_loss_clip": 0.94244736, + "balance_loss_mlp": 1.02336383, + "epoch": 0.651596224373234, + "flos": 22526907557760.0, + "grad_norm": 1.7703258074247359, + "language_loss": 0.74675089, + "learning_rate": 1.1436796919701484e-06, + "loss": 0.76809245, + "num_input_tokens_seen": 116673880, + "step": 5419, + "time_per_iteration": 2.7488720417022705 + }, + { + "auxiliary_loss_clip": 0.01127426, + "auxiliary_loss_mlp": 0.01028371, + "balance_loss_clip": 0.97991467, + "balance_loss_mlp": 1.02090859, + "epoch": 0.651716467263873, + "flos": 27818955250560.0, + "grad_norm": 1.9628214446477457, + "language_loss": 0.61558354, + "learning_rate": 1.1429757980239115e-06, + "loss": 0.63714147, + "num_input_tokens_seen": 116694305, + "step": 5420, + "time_per_iteration": 3.7277636528015137 + }, + { + "auxiliary_loss_clip": 0.0117397, + "auxiliary_loss_mlp": 0.01032432, + "balance_loss_clip": 1.05286622, + "balance_loss_mlp": 1.02454901, + "epoch": 0.6518367101545122, + "flos": 24316300414080.0, + "grad_norm": 2.3710311281991996, + "language_loss": 0.81833041, + "learning_rate": 1.1422720340875636e-06, + "loss": 0.84039438, + "num_input_tokens_seen": 116713055, + "step": 5421, + "time_per_iteration": 3.637913942337036 + }, + { + "auxiliary_loss_clip": 0.01155508, + "auxiliary_loss_mlp": 0.01023566, + "balance_loss_clip": 1.01515925, + "balance_loss_mlp": 1.01556098, + "epoch": 0.6519569530451512, + "flos": 20011939971840.0, + "grad_norm": 2.62164382629971, + "language_loss": 0.7921567, + "learning_rate": 1.1415684002678671e-06, + "loss": 0.81394744, + "num_input_tokens_seen": 116731815, + "step": 5422, + "time_per_iteration": 3.571322441101074 + }, + { + "auxiliary_loss_clip": 0.01125999, + "auxiliary_loss_mlp": 0.01031197, + "balance_loss_clip": 0.9735775, + "balance_loss_mlp": 1.02298975, + "epoch": 0.6520771959357903, + "flos": 21576064682880.0, + "grad_norm": 2.0173278685264986, + "language_loss": 0.7798993, + "learning_rate": 1.1408648966715617e-06, + "loss": 0.80147129, + "num_input_tokens_seen": 116749335, + "step": 5423, + "time_per_iteration": 2.6364917755126953 + }, + { + "auxiliary_loss_clip": 0.01124048, + "auxiliary_loss_mlp": 0.01024326, + "balance_loss_clip": 0.97221279, + "balance_loss_mlp": 1.01664925, + "epoch": 0.6521974388264293, + "flos": 22711021695360.0, + "grad_norm": 1.7660862820174301, + "language_loss": 0.72042376, + "learning_rate": 1.1401615234053683e-06, + "loss": 0.74190748, + "num_input_tokens_seen": 116768155, + "step": 5424, + "time_per_iteration": 2.7452170848846436 + }, + { + "auxiliary_loss_clip": 0.01125441, + "auxiliary_loss_mlp": 0.01026594, + "balance_loss_clip": 0.9735992, + "balance_loss_mlp": 1.01824307, + "epoch": 0.6523176817170685, + "flos": 23002939526400.0, + "grad_norm": 1.6512173465280366, + "language_loss": 0.75838912, + "learning_rate": 1.1394582805759885e-06, + "loss": 0.77990937, + "num_input_tokens_seen": 116787435, + "step": 5425, + "time_per_iteration": 2.739429473876953 + }, + { + "auxiliary_loss_clip": 0.01148812, + "auxiliary_loss_mlp": 0.01028186, + "balance_loss_clip": 1.01556468, + "balance_loss_mlp": 1.02009797, + "epoch": 0.6524379246077076, + "flos": 21688249835520.0, + "grad_norm": 1.923621654570139, + "language_loss": 0.75778413, + "learning_rate": 1.1387551682901022e-06, + "loss": 0.77955413, + "num_input_tokens_seen": 116808040, + "step": 5426, + "time_per_iteration": 2.6682307720184326 + }, + { + "auxiliary_loss_clip": 0.01096807, + "auxiliary_loss_mlp": 0.01025675, + "balance_loss_clip": 0.93510509, + "balance_loss_mlp": 1.01756835, + "epoch": 0.6525581674983466, + "flos": 19390936711680.0, + "grad_norm": 1.913072995829704, + "language_loss": 0.70658588, + "learning_rate": 1.138052186654373e-06, + "loss": 0.72781068, + "num_input_tokens_seen": 116825510, + "step": 5427, + "time_per_iteration": 2.813693046569824 + }, + { + "auxiliary_loss_clip": 0.01126733, + "auxiliary_loss_mlp": 0.01024587, + "balance_loss_clip": 0.97523117, + "balance_loss_mlp": 1.01680243, + "epoch": 0.6526784103889858, + "flos": 17165444832000.0, + "grad_norm": 1.8883399968979542, + "language_loss": 0.88041711, + "learning_rate": 1.1373493357754417e-06, + "loss": 0.90193027, + "num_input_tokens_seen": 116844415, + "step": 5428, + "time_per_iteration": 2.68818736076355 + }, + { + "auxiliary_loss_clip": 0.01171654, + "auxiliary_loss_mlp": 0.01025144, + "balance_loss_clip": 1.05139577, + "balance_loss_mlp": 1.01758635, + "epoch": 0.6527986532796248, + "flos": 18989168112000.0, + "grad_norm": 1.9151444599022847, + "language_loss": 0.77036446, + "learning_rate": 1.1366466157599303e-06, + "loss": 0.79233241, + "num_input_tokens_seen": 116863690, + "step": 5429, + "time_per_iteration": 2.583902359008789 + }, + { + "auxiliary_loss_clip": 0.01075287, + "auxiliary_loss_mlp": 0.00509883, + "balance_loss_clip": 0.8971591, + "balance_loss_mlp": 0.0, + "epoch": 0.6529188961702639, + "flos": 14238581011200.0, + "grad_norm": 1.98212608029161, + "language_loss": 0.76259667, + "learning_rate": 1.1359440267144412e-06, + "loss": 0.77844834, + "num_input_tokens_seen": 116881145, + "step": 5430, + "time_per_iteration": 2.7111587524414062 + }, + { + "auxiliary_loss_clip": 0.01151128, + "auxiliary_loss_mlp": 0.0102741, + "balance_loss_clip": 1.01385641, + "balance_loss_mlp": 1.02016199, + "epoch": 0.653039139060903, + "flos": 36682929158400.0, + "grad_norm": 1.6488376471307185, + "language_loss": 0.74385405, + "learning_rate": 1.1352415687455556e-06, + "loss": 0.76563942, + "num_input_tokens_seen": 116902405, + "step": 5431, + "time_per_iteration": 2.845475435256958 + }, + { + "auxiliary_loss_clip": 0.01151062, + "auxiliary_loss_mlp": 0.01030954, + "balance_loss_clip": 1.0161345, + "balance_loss_mlp": 1.02339888, + "epoch": 0.6531593819515421, + "flos": 25376275785600.0, + "grad_norm": 2.182026919029843, + "language_loss": 0.64069903, + "learning_rate": 1.1345392419598362e-06, + "loss": 0.66251922, + "num_input_tokens_seen": 116921285, + "step": 5432, + "time_per_iteration": 3.6253468990325928 + }, + { + "auxiliary_loss_clip": 0.01143261, + "auxiliary_loss_mlp": 0.01027971, + "balance_loss_clip": 1.01219904, + "balance_loss_mlp": 1.02068174, + "epoch": 0.6532796248421812, + "flos": 21178533888000.0, + "grad_norm": 1.686180282021817, + "language_loss": 0.71940279, + "learning_rate": 1.1338370464638263e-06, + "loss": 0.74111509, + "num_input_tokens_seen": 116940685, + "step": 5433, + "time_per_iteration": 2.643472194671631 + }, + { + "auxiliary_loss_clip": 0.01170163, + "auxiliary_loss_mlp": 0.01027613, + "balance_loss_clip": 1.04995131, + "balance_loss_mlp": 1.01959014, + "epoch": 0.6533998677328203, + "flos": 17675950878720.0, + "grad_norm": 2.701643400460071, + "language_loss": 0.64351642, + "learning_rate": 1.1331349823640474e-06, + "loss": 0.6654942, + "num_input_tokens_seen": 116958115, + "step": 5434, + "time_per_iteration": 2.634692668914795 + }, + { + "auxiliary_loss_clip": 0.01149717, + "auxiliary_loss_mlp": 0.00508955, + "balance_loss_clip": 1.01390314, + "balance_loss_mlp": 0.0, + "epoch": 0.6535201106234594, + "flos": 28400384701440.0, + "grad_norm": 2.0062952311218356, + "language_loss": 0.78041816, + "learning_rate": 1.132433049767003e-06, + "loss": 0.79700488, + "num_input_tokens_seen": 116976030, + "step": 5435, + "time_per_iteration": 2.70074725151062 + }, + { + "auxiliary_loss_clip": 0.01124011, + "auxiliary_loss_mlp": 0.01027718, + "balance_loss_clip": 0.97572196, + "balance_loss_mlp": 1.020136, + "epoch": 0.6536403535140984, + "flos": 23586667447680.0, + "grad_norm": 1.6170523019470369, + "language_loss": 0.81027162, + "learning_rate": 1.1317312487791748e-06, + "loss": 0.8317889, + "num_input_tokens_seen": 116997680, + "step": 5436, + "time_per_iteration": 2.7073304653167725 + }, + { + "auxiliary_loss_clip": 0.01142692, + "auxiliary_loss_mlp": 0.01026975, + "balance_loss_clip": 1.01177931, + "balance_loss_mlp": 1.01919103, + "epoch": 0.6537605964047376, + "flos": 21579476474880.0, + "grad_norm": 2.3318935729340344, + "language_loss": 0.73458606, + "learning_rate": 1.1310295795070253e-06, + "loss": 0.75628269, + "num_input_tokens_seen": 117017620, + "step": 5437, + "time_per_iteration": 2.751516580581665 + }, + { + "auxiliary_loss_clip": 0.01081434, + "auxiliary_loss_mlp": 0.01026632, + "balance_loss_clip": 0.8971827, + "balance_loss_mlp": 1.01941359, + "epoch": 0.6538808392953767, + "flos": 26833997433600.0, + "grad_norm": 2.2612791382543747, + "language_loss": 0.81295359, + "learning_rate": 1.1303280420569982e-06, + "loss": 0.8340342, + "num_input_tokens_seen": 117039505, + "step": 5438, + "time_per_iteration": 2.8352625370025635 + }, + { + "auxiliary_loss_clip": 0.01144169, + "auxiliary_loss_mlp": 0.01027494, + "balance_loss_clip": 1.01306486, + "balance_loss_mlp": 1.0199064, + "epoch": 0.6540010821860157, + "flos": 30738241301760.0, + "grad_norm": 1.693876834202397, + "language_loss": 0.77258253, + "learning_rate": 1.1296266365355158e-06, + "loss": 0.79429913, + "num_input_tokens_seen": 117062890, + "step": 5439, + "time_per_iteration": 2.714975357055664 + }, + { + "auxiliary_loss_clip": 0.0110309, + "auxiliary_loss_mlp": 0.01028967, + "balance_loss_clip": 0.93746156, + "balance_loss_mlp": 1.02094412, + "epoch": 0.6541213250766549, + "flos": 26907147480960.0, + "grad_norm": 1.8628280528776018, + "language_loss": 0.73823357, + "learning_rate": 1.1289253630489806e-06, + "loss": 0.75955415, + "num_input_tokens_seen": 117083940, + "step": 5440, + "time_per_iteration": 2.7751331329345703 + }, + { + "auxiliary_loss_clip": 0.01153442, + "auxiliary_loss_mlp": 0.01030581, + "balance_loss_clip": 1.01229858, + "balance_loss_mlp": 1.02186036, + "epoch": 0.6542415679672939, + "flos": 19172384409600.0, + "grad_norm": 2.1511108867079374, + "language_loss": 0.72596467, + "learning_rate": 1.1282242217037753e-06, + "loss": 0.74780488, + "num_input_tokens_seen": 117101440, + "step": 5441, + "time_per_iteration": 2.6263949871063232 + }, + { + "auxiliary_loss_clip": 0.01073151, + "auxiliary_loss_mlp": 0.01030906, + "balance_loss_clip": 0.89367926, + "balance_loss_mlp": 1.02254379, + "epoch": 0.654361810857933, + "flos": 48173517100800.0, + "grad_norm": 2.629813744456193, + "language_loss": 0.61890972, + "learning_rate": 1.127523212606262e-06, + "loss": 0.63995028, + "num_input_tokens_seen": 117124265, + "step": 5442, + "time_per_iteration": 2.991947889328003 + }, + { + "auxiliary_loss_clip": 0.01144532, + "auxiliary_loss_mlp": 0.01031178, + "balance_loss_clip": 1.0108788, + "balance_loss_mlp": 1.02382016, + "epoch": 0.6544820537485722, + "flos": 26943165843840.0, + "grad_norm": 3.923888711743631, + "language_loss": 0.73132676, + "learning_rate": 1.1268223358627835e-06, + "loss": 0.75308383, + "num_input_tokens_seen": 117146755, + "step": 5443, + "time_per_iteration": 2.8919029235839844 + }, + { + "auxiliary_loss_clip": 0.01172839, + "auxiliary_loss_mlp": 0.01022498, + "balance_loss_clip": 1.05213511, + "balance_loss_mlp": 1.01435614, + "epoch": 0.6546022966392112, + "flos": 20886328748160.0, + "grad_norm": 1.7242064908537762, + "language_loss": 0.72290778, + "learning_rate": 1.126121591579663e-06, + "loss": 0.74486113, + "num_input_tokens_seen": 117165960, + "step": 5444, + "time_per_iteration": 2.6036548614501953 + }, + { + "auxiliary_loss_clip": 0.01146573, + "auxiliary_loss_mlp": 0.01025755, + "balance_loss_clip": 1.01560068, + "balance_loss_mlp": 1.01867437, + "epoch": 0.6547225395298503, + "flos": 24936693143040.0, + "grad_norm": 1.8688025121023621, + "language_loss": 0.69124752, + "learning_rate": 1.1254209798632018e-06, + "loss": 0.71297073, + "num_input_tokens_seen": 117186980, + "step": 5445, + "time_per_iteration": 2.7321081161499023 + }, + { + "auxiliary_loss_clip": 0.01045146, + "auxiliary_loss_mlp": 0.0102766, + "balance_loss_clip": 0.85842276, + "balance_loss_mlp": 1.02033758, + "epoch": 0.6548427824204894, + "flos": 22565942663040.0, + "grad_norm": 1.6172849012436343, + "language_loss": 0.84642124, + "learning_rate": 1.124720500819683e-06, + "loss": 0.86714929, + "num_input_tokens_seen": 117205135, + "step": 5446, + "time_per_iteration": 4.718241930007935 + }, + { + "auxiliary_loss_clip": 0.01173564, + "auxiliary_loss_mlp": 0.01024712, + "balance_loss_clip": 1.05374217, + "balance_loss_mlp": 1.01661825, + "epoch": 0.6549630253111285, + "flos": 18442500048000.0, + "grad_norm": 1.7934839708383736, + "language_loss": 0.82552403, + "learning_rate": 1.1240201545553682e-06, + "loss": 0.84750676, + "num_input_tokens_seen": 117222935, + "step": 5447, + "time_per_iteration": 2.716322422027588 + }, + { + "auxiliary_loss_clip": 0.01099743, + "auxiliary_loss_mlp": 0.01026563, + "balance_loss_clip": 0.93659127, + "balance_loss_mlp": 1.01934755, + "epoch": 0.6550832682017675, + "flos": 25187313312000.0, + "grad_norm": 1.778625629158794, + "language_loss": 0.7338416, + "learning_rate": 1.1233199411764987e-06, + "loss": 0.75510466, + "num_input_tokens_seen": 117242370, + "step": 5448, + "time_per_iteration": 3.682558298110962 + }, + { + "auxiliary_loss_clip": 0.01088704, + "auxiliary_loss_mlp": 0.01028342, + "balance_loss_clip": 0.93346953, + "balance_loss_mlp": 1.02040839, + "epoch": 0.6552035110924067, + "flos": 22748153379840.0, + "grad_norm": 2.015926953323967, + "language_loss": 0.69414443, + "learning_rate": 1.1226198607892978e-06, + "loss": 0.71531487, + "num_input_tokens_seen": 117262930, + "step": 5449, + "time_per_iteration": 2.7635488510131836 + }, + { + "auxiliary_loss_clip": 0.01080688, + "auxiliary_loss_mlp": 0.01027031, + "balance_loss_clip": 0.90054661, + "balance_loss_mlp": 1.01958358, + "epoch": 0.6553237539830458, + "flos": 21799178012160.0, + "grad_norm": 1.8530237091886248, + "language_loss": 0.79857677, + "learning_rate": 1.1219199134999664e-06, + "loss": 0.81965399, + "num_input_tokens_seen": 117281430, + "step": 5450, + "time_per_iteration": 2.7926902770996094 + }, + { + "auxiliary_loss_clip": 0.01131295, + "auxiliary_loss_mlp": 0.01032177, + "balance_loss_clip": 0.97777015, + "balance_loss_mlp": 1.02404046, + "epoch": 0.6554439968736848, + "flos": 20887226588160.0, + "grad_norm": 2.1272354009777894, + "language_loss": 0.78335226, + "learning_rate": 1.1212200994146863e-06, + "loss": 0.80498695, + "num_input_tokens_seen": 117299185, + "step": 5451, + "time_per_iteration": 2.695134401321411 + }, + { + "auxiliary_loss_clip": 0.01097821, + "auxiliary_loss_mlp": 0.01026675, + "balance_loss_clip": 0.93257618, + "balance_loss_mlp": 1.01920033, + "epoch": 0.655564239764324, + "flos": 16139045698560.0, + "grad_norm": 2.3995462484266823, + "language_loss": 0.75871533, + "learning_rate": 1.120520418639618e-06, + "loss": 0.77996027, + "num_input_tokens_seen": 117317720, + "step": 5452, + "time_per_iteration": 2.7245099544525146 + }, + { + "auxiliary_loss_clip": 0.01149531, + "auxiliary_loss_mlp": 0.01028153, + "balance_loss_clip": 1.01514411, + "balance_loss_mlp": 1.02035046, + "epoch": 0.655684482654963, + "flos": 29570354496000.0, + "grad_norm": 2.5504837757166565, + "language_loss": 0.83436704, + "learning_rate": 1.119820871280903e-06, + "loss": 0.85614395, + "num_input_tokens_seen": 117338795, + "step": 5453, + "time_per_iteration": 2.802428722381592 + }, + { + "auxiliary_loss_clip": 0.01145756, + "auxiliary_loss_mlp": 0.01037656, + "balance_loss_clip": 1.01172721, + "balance_loss_mlp": 1.02925825, + "epoch": 0.6558047255456021, + "flos": 29789409588480.0, + "grad_norm": 1.7958173561729327, + "language_loss": 0.73527229, + "learning_rate": 1.1191214574446614e-06, + "loss": 0.75710648, + "num_input_tokens_seen": 117359040, + "step": 5454, + "time_per_iteration": 2.769458770751953 + }, + { + "auxiliary_loss_clip": 0.01118073, + "auxiliary_loss_mlp": 0.01027916, + "balance_loss_clip": 0.97224456, + "balance_loss_mlp": 1.02034903, + "epoch": 0.6559249684362413, + "flos": 29059166090880.0, + "grad_norm": 1.609371873766631, + "language_loss": 0.79915726, + "learning_rate": 1.118422177236995e-06, + "loss": 0.82061714, + "num_input_tokens_seen": 117380865, + "step": 5455, + "time_per_iteration": 2.793363571166992 + }, + { + "auxiliary_loss_clip": 0.01125163, + "auxiliary_loss_mlp": 0.01027194, + "balance_loss_clip": 0.97334117, + "balance_loss_mlp": 1.01915312, + "epoch": 0.6560452113268803, + "flos": 20225464369920.0, + "grad_norm": 2.1883127688984345, + "language_loss": 0.85573781, + "learning_rate": 1.1177230307639835e-06, + "loss": 0.8772614, + "num_input_tokens_seen": 117398405, + "step": 5456, + "time_per_iteration": 2.6694228649139404 + }, + { + "auxiliary_loss_clip": 0.0109623, + "auxiliary_loss_mlp": 0.01026317, + "balance_loss_clip": 0.93496639, + "balance_loss_mlp": 1.01921237, + "epoch": 0.6561654542175194, + "flos": 25045538330880.0, + "grad_norm": 1.6723825056043762, + "language_loss": 0.79035223, + "learning_rate": 1.1170240181316865e-06, + "loss": 0.81157768, + "num_input_tokens_seen": 117419850, + "step": 5457, + "time_per_iteration": 2.7925710678100586 + }, + { + "auxiliary_loss_clip": 0.01095632, + "auxiliary_loss_mlp": 0.01029545, + "balance_loss_clip": 0.93099642, + "balance_loss_mlp": 1.02173042, + "epoch": 0.6562856971081584, + "flos": 22856711258880.0, + "grad_norm": 2.1097780191051294, + "language_loss": 0.79247797, + "learning_rate": 1.1163251394461442e-06, + "loss": 0.81372976, + "num_input_tokens_seen": 117438330, + "step": 5458, + "time_per_iteration": 3.693333148956299 + }, + { + "auxiliary_loss_clip": 0.01144241, + "auxiliary_loss_mlp": 0.01033944, + "balance_loss_clip": 1.01300025, + "balance_loss_mlp": 1.02607584, + "epoch": 0.6564059399987976, + "flos": 18872565586560.0, + "grad_norm": 2.163767174647474, + "language_loss": 0.82093126, + "learning_rate": 1.1156263948133746e-06, + "loss": 0.84271312, + "num_input_tokens_seen": 117454985, + "step": 5459, + "time_per_iteration": 2.6076159477233887 + }, + { + "auxiliary_loss_clip": 0.01071558, + "auxiliary_loss_mlp": 0.0050975, + "balance_loss_clip": 0.89696062, + "balance_loss_mlp": 0.0, + "epoch": 0.6565261828894366, + "flos": 25484187219840.0, + "grad_norm": 1.853544634691372, + "language_loss": 0.77436054, + "learning_rate": 1.1149277843393787e-06, + "loss": 0.79017359, + "num_input_tokens_seen": 117476145, + "step": 5460, + "time_per_iteration": 2.8224387168884277 + }, + { + "auxiliary_loss_clip": 0.01050116, + "auxiliary_loss_mlp": 0.00509625, + "balance_loss_clip": 0.85455292, + "balance_loss_mlp": 0.0, + "epoch": 0.6566464257800757, + "flos": 19683500987520.0, + "grad_norm": 1.988209079419029, + "language_loss": 0.63192773, + "learning_rate": 1.1142293081301342e-06, + "loss": 0.64752519, + "num_input_tokens_seen": 117494025, + "step": 5461, + "time_per_iteration": 2.7874019145965576 + }, + { + "auxiliary_loss_clip": 0.01118984, + "auxiliary_loss_mlp": 0.01022117, + "balance_loss_clip": 0.97284347, + "balance_loss_mlp": 1.01423693, + "epoch": 0.6567666686707149, + "flos": 23514127931520.0, + "grad_norm": 1.8668634631483865, + "language_loss": 0.67931634, + "learning_rate": 1.1135309662915995e-06, + "loss": 0.70072734, + "num_input_tokens_seen": 117514190, + "step": 5462, + "time_per_iteration": 2.7332050800323486 + }, + { + "auxiliary_loss_clip": 0.01082433, + "auxiliary_loss_mlp": 0.01023422, + "balance_loss_clip": 0.89634383, + "balance_loss_mlp": 1.01562583, + "epoch": 0.6568869115613539, + "flos": 32781342896640.0, + "grad_norm": 1.8474248479523343, + "language_loss": 0.59815115, + "learning_rate": 1.112832758929712e-06, + "loss": 0.61920965, + "num_input_tokens_seen": 117536800, + "step": 5463, + "time_per_iteration": 2.8149757385253906 + }, + { + "auxiliary_loss_clip": 0.01147492, + "auxiliary_loss_mlp": 0.01026987, + "balance_loss_clip": 1.01462317, + "balance_loss_mlp": 1.01860106, + "epoch": 0.657007154451993, + "flos": 18442428220800.0, + "grad_norm": 3.445594500691987, + "language_loss": 0.74970591, + "learning_rate": 1.11213468615039e-06, + "loss": 0.7714507, + "num_input_tokens_seen": 117556230, + "step": 5464, + "time_per_iteration": 2.6856346130371094 + }, + { + "auxiliary_loss_clip": 0.01053267, + "auxiliary_loss_mlp": 0.01026764, + "balance_loss_clip": 0.85797489, + "balance_loss_mlp": 1.01919436, + "epoch": 0.6571273973426321, + "flos": 25156717902720.0, + "grad_norm": 2.280706899511885, + "language_loss": 0.75453484, + "learning_rate": 1.1114367480595292e-06, + "loss": 0.77533519, + "num_input_tokens_seen": 117577310, + "step": 5465, + "time_per_iteration": 2.788867712020874 + }, + { + "auxiliary_loss_clip": 0.01055423, + "auxiliary_loss_mlp": 0.01026808, + "balance_loss_clip": 0.8624866, + "balance_loss_mlp": 1.01854098, + "epoch": 0.6572476402332712, + "flos": 17529830352000.0, + "grad_norm": 1.8028678982205861, + "language_loss": 0.81361914, + "learning_rate": 1.1107389447630086e-06, + "loss": 0.83444142, + "num_input_tokens_seen": 117596010, + "step": 5466, + "time_per_iteration": 2.759735584259033 + }, + { + "auxiliary_loss_clip": 0.01118765, + "auxiliary_loss_mlp": 0.00509791, + "balance_loss_clip": 0.97073132, + "balance_loss_mlp": 0.0, + "epoch": 0.6573678831239103, + "flos": 17014260487680.0, + "grad_norm": 2.596006368408063, + "language_loss": 0.78314745, + "learning_rate": 1.1100412763666818e-06, + "loss": 0.79943299, + "num_input_tokens_seen": 117611270, + "step": 5467, + "time_per_iteration": 2.6965606212615967 + }, + { + "auxiliary_loss_clip": 0.01125166, + "auxiliary_loss_mlp": 0.01025117, + "balance_loss_clip": 0.97526693, + "balance_loss_mlp": 1.01792836, + "epoch": 0.6574881260145494, + "flos": 23910078528000.0, + "grad_norm": 1.4251829612881266, + "language_loss": 0.80190563, + "learning_rate": 1.1093437429763865e-06, + "loss": 0.82340848, + "num_input_tokens_seen": 117631535, + "step": 5468, + "time_per_iteration": 2.7384722232818604 + }, + { + "auxiliary_loss_clip": 0.01144671, + "auxiliary_loss_mlp": 0.01022798, + "balance_loss_clip": 1.01225758, + "balance_loss_mlp": 1.01554751, + "epoch": 0.6576083689051885, + "flos": 11218458504960.0, + "grad_norm": 1.8388221394694182, + "language_loss": 0.73212671, + "learning_rate": 1.1086463446979361e-06, + "loss": 0.75380147, + "num_input_tokens_seen": 117649885, + "step": 5469, + "time_per_iteration": 2.6672897338867188 + }, + { + "auxiliary_loss_clip": 0.01151657, + "auxiliary_loss_mlp": 0.01028597, + "balance_loss_clip": 1.01725781, + "balance_loss_mlp": 1.02115798, + "epoch": 0.6577286117958275, + "flos": 22455553190400.0, + "grad_norm": 2.405017757278119, + "language_loss": 0.77511746, + "learning_rate": 1.1079490816371277e-06, + "loss": 0.79691994, + "num_input_tokens_seen": 117669650, + "step": 5470, + "time_per_iteration": 2.6937780380249023 + }, + { + "auxiliary_loss_clip": 0.01147146, + "auxiliary_loss_mlp": 0.00510024, + "balance_loss_clip": 1.01222908, + "balance_loss_mlp": 0.0, + "epoch": 0.6578488546864667, + "flos": 21872184405120.0, + "grad_norm": 2.718849100053538, + "language_loss": 0.74404889, + "learning_rate": 1.1072519538997352e-06, + "loss": 0.76062059, + "num_input_tokens_seen": 117688790, + "step": 5471, + "time_per_iteration": 3.58268666267395 + }, + { + "auxiliary_loss_clip": 0.01124106, + "auxiliary_loss_mlp": 0.01024213, + "balance_loss_clip": 0.97237217, + "balance_loss_mlp": 1.01638126, + "epoch": 0.6579690975771058, + "flos": 23543753673600.0, + "grad_norm": 2.1735422426887965, + "language_loss": 0.82506275, + "learning_rate": 1.1065549615915095e-06, + "loss": 0.84654593, + "num_input_tokens_seen": 117708620, + "step": 5472, + "time_per_iteration": 3.6268460750579834 + }, + { + "auxiliary_loss_clip": 0.01150479, + "auxiliary_loss_mlp": 0.01027587, + "balance_loss_clip": 1.01912284, + "balance_loss_mlp": 1.01994562, + "epoch": 0.6580893404677448, + "flos": 32743995730560.0, + "grad_norm": 2.418475563990859, + "language_loss": 0.78115612, + "learning_rate": 1.105858104818187e-06, + "loss": 0.80293679, + "num_input_tokens_seen": 117729775, + "step": 5473, + "time_per_iteration": 2.722447156906128 + }, + { + "auxiliary_loss_clip": 0.01151416, + "auxiliary_loss_mlp": 0.01025378, + "balance_loss_clip": 1.01353955, + "balance_loss_mlp": 1.01751649, + "epoch": 0.658209583358384, + "flos": 15888138220800.0, + "grad_norm": 2.30290551781784, + "language_loss": 0.74709547, + "learning_rate": 1.105161383685478e-06, + "loss": 0.76886344, + "num_input_tokens_seen": 117746160, + "step": 5474, + "time_per_iteration": 3.5145294666290283 + }, + { + "auxiliary_loss_clip": 0.01010548, + "auxiliary_loss_mlp": 0.01001677, + "balance_loss_clip": 0.90567076, + "balance_loss_mlp": 1.00050926, + "epoch": 0.658329826249023, + "flos": 62695902447360.0, + "grad_norm": 0.7276649878606826, + "language_loss": 0.56321007, + "learning_rate": 1.1044647982990771e-06, + "loss": 0.58333236, + "num_input_tokens_seen": 117808045, + "step": 5475, + "time_per_iteration": 3.213273048400879 + }, + { + "auxiliary_loss_clip": 0.01125551, + "auxiliary_loss_mlp": 0.01025121, + "balance_loss_clip": 0.97618926, + "balance_loss_mlp": 1.01714551, + "epoch": 0.6584500691396621, + "flos": 31722624501120.0, + "grad_norm": 2.271206045326241, + "language_loss": 0.64530706, + "learning_rate": 1.1037683487646536e-06, + "loss": 0.66681379, + "num_input_tokens_seen": 117828330, + "step": 5476, + "time_per_iteration": 2.794074296951294 + }, + { + "auxiliary_loss_clip": 0.01121811, + "auxiliary_loss_mlp": 0.00509092, + "balance_loss_clip": 0.97699904, + "balance_loss_mlp": 0.0, + "epoch": 0.6585703120303013, + "flos": 18406086635520.0, + "grad_norm": 1.8625327867223347, + "language_loss": 0.77325791, + "learning_rate": 1.1030720351878583e-06, + "loss": 0.78956693, + "num_input_tokens_seen": 117846450, + "step": 5477, + "time_per_iteration": 2.682523488998413 + }, + { + "auxiliary_loss_clip": 0.01035082, + "auxiliary_loss_mlp": 0.01001692, + "balance_loss_clip": 0.94518083, + "balance_loss_mlp": 1.00057781, + "epoch": 0.6586905549209403, + "flos": 58309880434560.0, + "grad_norm": 0.8167228059105716, + "language_loss": 0.57673097, + "learning_rate": 1.102375857674323e-06, + "loss": 0.59709871, + "num_input_tokens_seen": 117908365, + "step": 5478, + "time_per_iteration": 3.192972183227539 + }, + { + "auxiliary_loss_clip": 0.01123999, + "auxiliary_loss_mlp": 0.01027077, + "balance_loss_clip": 0.97317749, + "balance_loss_mlp": 1.01965904, + "epoch": 0.6588107978115794, + "flos": 22782627457920.0, + "grad_norm": 1.7188441685971907, + "language_loss": 0.90454698, + "learning_rate": 1.1016798163296561e-06, + "loss": 0.92605776, + "num_input_tokens_seen": 117927565, + "step": 5479, + "time_per_iteration": 2.7126219272613525 + }, + { + "auxiliary_loss_clip": 0.01149447, + "auxiliary_loss_mlp": 0.01023497, + "balance_loss_clip": 1.01303601, + "balance_loss_mlp": 1.01565301, + "epoch": 0.6589310407022185, + "flos": 20667525050880.0, + "grad_norm": 1.8132195707838, + "language_loss": 0.66177583, + "learning_rate": 1.1009839112594471e-06, + "loss": 0.6835053, + "num_input_tokens_seen": 117945590, + "step": 5480, + "time_per_iteration": 2.6243109703063965 + }, + { + "auxiliary_loss_clip": 0.01150092, + "auxiliary_loss_mlp": 0.01026794, + "balance_loss_clip": 1.01378238, + "balance_loss_mlp": 1.01899457, + "epoch": 0.6590512835928576, + "flos": 25630595055360.0, + "grad_norm": 2.003317756782042, + "language_loss": 0.72364527, + "learning_rate": 1.1002881425692638e-06, + "loss": 0.74541414, + "num_input_tokens_seen": 117966020, + "step": 5481, + "time_per_iteration": 2.7570888996124268 + }, + { + "auxiliary_loss_clip": 0.01141593, + "auxiliary_loss_mlp": 0.0102792, + "balance_loss_clip": 1.01037669, + "balance_loss_mlp": 1.02003455, + "epoch": 0.6591715264834966, + "flos": 23726108044800.0, + "grad_norm": 1.6918580158360583, + "language_loss": 0.75626099, + "learning_rate": 1.0995925103646532e-06, + "loss": 0.77795613, + "num_input_tokens_seen": 117984620, + "step": 5482, + "time_per_iteration": 2.717825412750244 + }, + { + "auxiliary_loss_clip": 0.01094504, + "auxiliary_loss_mlp": 0.01029574, + "balance_loss_clip": 0.93534422, + "balance_loss_mlp": 1.02176905, + "epoch": 0.6592917693741358, + "flos": 35773850822400.0, + "grad_norm": 1.523531996689204, + "language_loss": 0.66740525, + "learning_rate": 1.0988970147511437e-06, + "loss": 0.68864608, + "num_input_tokens_seen": 118006500, + "step": 5483, + "time_per_iteration": 2.8253366947174072 + }, + { + "auxiliary_loss_clip": 0.01124534, + "auxiliary_loss_mlp": 0.01023762, + "balance_loss_clip": 0.97652787, + "balance_loss_mlp": 1.0162102, + "epoch": 0.6594120122647749, + "flos": 21396834794880.0, + "grad_norm": 2.2696660177451378, + "language_loss": 0.80673414, + "learning_rate": 1.0982016558342405e-06, + "loss": 0.82821715, + "num_input_tokens_seen": 118025470, + "step": 5484, + "time_per_iteration": 3.5968127250671387 + }, + { + "auxiliary_loss_clip": 0.01173211, + "auxiliary_loss_mlp": 0.01024399, + "balance_loss_clip": 1.0537684, + "balance_loss_mlp": 1.01706421, + "epoch": 0.6595322551554139, + "flos": 19351829779200.0, + "grad_norm": 2.2541341502025487, + "language_loss": 0.71326935, + "learning_rate": 1.0975064337194291e-06, + "loss": 0.73524541, + "num_input_tokens_seen": 118043515, + "step": 5485, + "time_per_iteration": 2.632718801498413 + }, + { + "auxiliary_loss_clip": 0.01097461, + "auxiliary_loss_mlp": 0.01030566, + "balance_loss_clip": 0.93792486, + "balance_loss_mlp": 1.02319264, + "epoch": 0.6596524980460531, + "flos": 16837113588480.0, + "grad_norm": 1.6448881663323216, + "language_loss": 0.70275414, + "learning_rate": 1.0968113485121743e-06, + "loss": 0.72403449, + "num_input_tokens_seen": 118063105, + "step": 5486, + "time_per_iteration": 2.7585713863372803 + }, + { + "auxiliary_loss_clip": 0.01148886, + "auxiliary_loss_mlp": 0.00510541, + "balance_loss_clip": 1.0122509, + "balance_loss_mlp": 0.0, + "epoch": 0.6597727409366921, + "flos": 21798567480960.0, + "grad_norm": 1.8206585708702052, + "language_loss": 0.80054784, + "learning_rate": 1.0961164003179185e-06, + "loss": 0.81714213, + "num_input_tokens_seen": 118081615, + "step": 5487, + "time_per_iteration": 2.632206678390503 + }, + { + "auxiliary_loss_clip": 0.01097975, + "auxiliary_loss_mlp": 0.01025143, + "balance_loss_clip": 0.93525511, + "balance_loss_mlp": 1.01729882, + "epoch": 0.6598929838273312, + "flos": 23730704985600.0, + "grad_norm": 2.0564409176831173, + "language_loss": 0.84396183, + "learning_rate": 1.0954215892420884e-06, + "loss": 0.86519301, + "num_input_tokens_seen": 118102315, + "step": 5488, + "time_per_iteration": 2.7580525875091553 + }, + { + "auxiliary_loss_clip": 0.01105435, + "auxiliary_loss_mlp": 0.01027129, + "balance_loss_clip": 0.93858331, + "balance_loss_mlp": 1.01888561, + "epoch": 0.6600132267179702, + "flos": 19974520978560.0, + "grad_norm": 1.9090606623215491, + "language_loss": 0.70600349, + "learning_rate": 1.094726915390082e-06, + "loss": 0.72732913, + "num_input_tokens_seen": 118120650, + "step": 5489, + "time_per_iteration": 2.6585581302642822 + }, + { + "auxiliary_loss_clip": 0.01148299, + "auxiliary_loss_mlp": 0.01027231, + "balance_loss_clip": 1.01430929, + "balance_loss_mlp": 1.01955342, + "epoch": 0.6601334696086094, + "flos": 22342649765760.0, + "grad_norm": 1.7863947266637246, + "language_loss": 0.69474787, + "learning_rate": 1.0940323788672836e-06, + "loss": 0.71650314, + "num_input_tokens_seen": 118139825, + "step": 5490, + "time_per_iteration": 2.6726555824279785 + }, + { + "auxiliary_loss_clip": 0.01144427, + "auxiliary_loss_mlp": 0.01022885, + "balance_loss_clip": 1.01238012, + "balance_loss_mlp": 1.01563156, + "epoch": 0.6602537124992485, + "flos": 25703098657920.0, + "grad_norm": 1.8162313868103037, + "language_loss": 0.73863691, + "learning_rate": 1.0933379797790522e-06, + "loss": 0.76031005, + "num_input_tokens_seen": 118159240, + "step": 5491, + "time_per_iteration": 2.6503584384918213 + }, + { + "auxiliary_loss_clip": 0.01173366, + "auxiliary_loss_mlp": 0.01028872, + "balance_loss_clip": 1.05265224, + "balance_loss_mlp": 1.02151382, + "epoch": 0.6603739553898875, + "flos": 25848572739840.0, + "grad_norm": 2.745138655063102, + "language_loss": 0.71365499, + "learning_rate": 1.0926437182307293e-06, + "loss": 0.73567736, + "num_input_tokens_seen": 118178050, + "step": 5492, + "time_per_iteration": 2.5940799713134766 + }, + { + "auxiliary_loss_clip": 0.01127947, + "auxiliary_loss_mlp": 0.01027807, + "balance_loss_clip": 0.97345489, + "balance_loss_mlp": 1.0201062, + "epoch": 0.6604941982805267, + "flos": 24570296461440.0, + "grad_norm": 2.1788631294307144, + "language_loss": 0.78405607, + "learning_rate": 1.0919495943276338e-06, + "loss": 0.80561364, + "num_input_tokens_seen": 118199070, + "step": 5493, + "time_per_iteration": 2.790630578994751 + }, + { + "auxiliary_loss_clip": 0.01101786, + "auxiliary_loss_mlp": 0.01029905, + "balance_loss_clip": 0.93264943, + "balance_loss_mlp": 1.02189469, + "epoch": 0.6606144411711657, + "flos": 13261775581440.0, + "grad_norm": 3.228917969579988, + "language_loss": 0.76353139, + "learning_rate": 1.0912556081750611e-06, + "loss": 0.78484833, + "num_input_tokens_seen": 118217000, + "step": 5494, + "time_per_iteration": 2.664591073989868 + }, + { + "auxiliary_loss_clip": 0.0112188, + "auxiliary_loss_mlp": 0.01026694, + "balance_loss_clip": 0.97613043, + "balance_loss_mlp": 1.01926756, + "epoch": 0.6607346840618048, + "flos": 25155281358720.0, + "grad_norm": 1.906194217140087, + "language_loss": 0.76284069, + "learning_rate": 1.0905617598782909e-06, + "loss": 0.78432643, + "num_input_tokens_seen": 118237205, + "step": 5495, + "time_per_iteration": 2.6714348793029785 + }, + { + "auxiliary_loss_clip": 0.01068583, + "auxiliary_loss_mlp": 0.01023697, + "balance_loss_clip": 0.8939836, + "balance_loss_mlp": 1.01591825, + "epoch": 0.660854926952444, + "flos": 17638029095040.0, + "grad_norm": 2.036877936985383, + "language_loss": 0.81682914, + "learning_rate": 1.0898680495425775e-06, + "loss": 0.83775192, + "num_input_tokens_seen": 118255495, + "step": 5496, + "time_per_iteration": 2.6952455043792725 + }, + { + "auxiliary_loss_clip": 0.01125624, + "auxiliary_loss_mlp": 0.01027046, + "balance_loss_clip": 0.97438478, + "balance_loss_mlp": 1.01923823, + "epoch": 0.660975169843083, + "flos": 16836000266880.0, + "grad_norm": 1.844139527157107, + "language_loss": 0.80333233, + "learning_rate": 1.0891744772731594e-06, + "loss": 0.82485902, + "num_input_tokens_seen": 118273310, + "step": 5497, + "time_per_iteration": 3.9335131645202637 + }, + { + "auxiliary_loss_clip": 0.01148121, + "auxiliary_loss_mlp": 0.01025956, + "balance_loss_clip": 1.01214361, + "balance_loss_mlp": 1.01864552, + "epoch": 0.6610954127337221, + "flos": 26870410846080.0, + "grad_norm": 1.5498081102896437, + "language_loss": 0.6581676, + "learning_rate": 1.088481043175248e-06, + "loss": 0.67990839, + "num_input_tokens_seen": 118293880, + "step": 5498, + "time_per_iteration": 3.6241509914398193 + }, + { + "auxiliary_loss_clip": 0.01113882, + "auxiliary_loss_mlp": 0.01025665, + "balance_loss_clip": 0.96989536, + "balance_loss_mlp": 1.01755321, + "epoch": 0.6612156556243612, + "flos": 26465697331200.0, + "grad_norm": 1.6532189090869116, + "language_loss": 0.75687581, + "learning_rate": 1.0877877473540368e-06, + "loss": 0.77827132, + "num_input_tokens_seen": 118314465, + "step": 5499, + "time_per_iteration": 2.7718591690063477 + }, + { + "auxiliary_loss_clip": 0.01174787, + "auxiliary_loss_mlp": 0.0103146, + "balance_loss_clip": 1.05370402, + "balance_loss_mlp": 1.02329993, + "epoch": 0.6613358985150003, + "flos": 19791915212160.0, + "grad_norm": 1.6569594740635962, + "language_loss": 0.7214582, + "learning_rate": 1.0870945899147002e-06, + "loss": 0.74352074, + "num_input_tokens_seen": 118331110, + "step": 5500, + "time_per_iteration": 3.533294916152954 + }, + { + "auxiliary_loss_clip": 0.01147905, + "auxiliary_loss_mlp": 0.01031255, + "balance_loss_clip": 1.01451218, + "balance_loss_mlp": 1.02289271, + "epoch": 0.6614561414056394, + "flos": 26831627136000.0, + "grad_norm": 2.2165258859968526, + "language_loss": 0.76356107, + "learning_rate": 1.0864015709623879e-06, + "loss": 0.78535271, + "num_input_tokens_seen": 118351980, + "step": 5501, + "time_per_iteration": 2.6795151233673096 + }, + { + "auxiliary_loss_clip": 0.01150484, + "auxiliary_loss_mlp": 0.01029198, + "balance_loss_clip": 1.01252544, + "balance_loss_mlp": 1.02158022, + "epoch": 0.6615763842962785, + "flos": 22894597128960.0, + "grad_norm": 2.9645233534132083, + "language_loss": 0.80036074, + "learning_rate": 1.0857086906022313e-06, + "loss": 0.8221575, + "num_input_tokens_seen": 118370315, + "step": 5502, + "time_per_iteration": 2.6800289154052734 + }, + { + "auxiliary_loss_clip": 0.01045194, + "auxiliary_loss_mlp": 0.01027438, + "balance_loss_clip": 0.86174631, + "balance_loss_mlp": 1.0194267, + "epoch": 0.6616966271869176, + "flos": 24790321221120.0, + "grad_norm": 2.0188021194315287, + "language_loss": 0.73125869, + "learning_rate": 1.0850159489393388e-06, + "loss": 0.75198501, + "num_input_tokens_seen": 118389575, + "step": 5503, + "time_per_iteration": 2.763158082962036 + }, + { + "auxiliary_loss_clip": 0.01092643, + "auxiliary_loss_mlp": 0.01024931, + "balance_loss_clip": 0.93031949, + "balance_loss_mlp": 1.0170629, + "epoch": 0.6618168700775566, + "flos": 17202109639680.0, + "grad_norm": 1.8143259858573986, + "language_loss": 0.82044363, + "learning_rate": 1.0843233460787992e-06, + "loss": 0.84161937, + "num_input_tokens_seen": 118406790, + "step": 5504, + "time_per_iteration": 2.7156381607055664 + }, + { + "auxiliary_loss_clip": 0.01091847, + "auxiliary_loss_mlp": 0.01026018, + "balance_loss_clip": 0.93865186, + "balance_loss_mlp": 1.01859701, + "epoch": 0.6619371129681958, + "flos": 25447091448960.0, + "grad_norm": 1.924894352836786, + "language_loss": 0.77793914, + "learning_rate": 1.0836308821256805e-06, + "loss": 0.7991178, + "num_input_tokens_seen": 118427590, + "step": 5505, + "time_per_iteration": 2.73781681060791 + }, + { + "auxiliary_loss_clip": 0.01145342, + "auxiliary_loss_mlp": 0.01025215, + "balance_loss_clip": 1.01299274, + "balance_loss_mlp": 1.01741254, + "epoch": 0.6620573558588349, + "flos": 18040444139520.0, + "grad_norm": 2.3477686321228317, + "language_loss": 0.78247893, + "learning_rate": 1.0829385571850282e-06, + "loss": 0.80418444, + "num_input_tokens_seen": 118444570, + "step": 5506, + "time_per_iteration": 2.6881022453308105 + }, + { + "auxiliary_loss_clip": 0.01174361, + "auxiliary_loss_mlp": 0.01026724, + "balance_loss_clip": 1.05128205, + "balance_loss_mlp": 1.01872158, + "epoch": 0.6621775987494739, + "flos": 17785586165760.0, + "grad_norm": 2.182144927909007, + "language_loss": 0.83796573, + "learning_rate": 1.0822463713618679e-06, + "loss": 0.85997659, + "num_input_tokens_seen": 118461425, + "step": 5507, + "time_per_iteration": 2.6290078163146973 + }, + { + "auxiliary_loss_clip": 0.01102267, + "auxiliary_loss_mlp": 0.01025401, + "balance_loss_clip": 0.93558669, + "balance_loss_mlp": 1.01756871, + "epoch": 0.6622978416401131, + "flos": 17492590926720.0, + "grad_norm": 2.0409596808473993, + "language_loss": 0.85040689, + "learning_rate": 1.0815543247612034e-06, + "loss": 0.8716836, + "num_input_tokens_seen": 118478495, + "step": 5508, + "time_per_iteration": 2.702983856201172 + }, + { + "auxiliary_loss_clip": 0.01120694, + "auxiliary_loss_mlp": 0.01023459, + "balance_loss_clip": 0.96898222, + "balance_loss_mlp": 1.01565683, + "epoch": 0.6624180845307521, + "flos": 21648352803840.0, + "grad_norm": 1.6316589629340381, + "language_loss": 0.82891083, + "learning_rate": 1.0808624174880168e-06, + "loss": 0.85035229, + "num_input_tokens_seen": 118499145, + "step": 5509, + "time_per_iteration": 2.711702823638916 + }, + { + "auxiliary_loss_clip": 0.01167813, + "auxiliary_loss_mlp": 0.01025172, + "balance_loss_clip": 1.05130756, + "balance_loss_mlp": 1.01820993, + "epoch": 0.6625383274213912, + "flos": 23805902108160.0, + "grad_norm": 2.0797487571959614, + "language_loss": 0.80185103, + "learning_rate": 1.080170649647272e-06, + "loss": 0.82378089, + "num_input_tokens_seen": 118518950, + "step": 5510, + "time_per_iteration": 3.4623212814331055 + }, + { + "auxiliary_loss_clip": 0.01170324, + "auxiliary_loss_mlp": 0.01030109, + "balance_loss_clip": 1.05200875, + "balance_loss_mlp": 1.02218795, + "epoch": 0.6626585703120303, + "flos": 33262941473280.0, + "grad_norm": 1.9580433957006744, + "language_loss": 0.67385566, + "learning_rate": 1.0794790213439068e-06, + "loss": 0.69586003, + "num_input_tokens_seen": 118545850, + "step": 5511, + "time_per_iteration": 2.7479872703552246 + }, + { + "auxiliary_loss_clip": 0.01077742, + "auxiliary_loss_mlp": 0.01030905, + "balance_loss_clip": 0.8963365, + "balance_loss_mlp": 1.02282214, + "epoch": 0.6627788132026694, + "flos": 22085780630400.0, + "grad_norm": 1.8976090866411701, + "language_loss": 0.78492463, + "learning_rate": 1.078787532682843e-06, + "loss": 0.80601108, + "num_input_tokens_seen": 118563325, + "step": 5512, + "time_per_iteration": 2.7456486225128174 + }, + { + "auxiliary_loss_clip": 0.01144777, + "auxiliary_loss_mlp": 0.01023037, + "balance_loss_clip": 1.01203871, + "balance_loss_mlp": 1.01543748, + "epoch": 0.6628990560933085, + "flos": 36173608260480.0, + "grad_norm": 2.1310076184618567, + "language_loss": 0.76040888, + "learning_rate": 1.0780961837689773e-06, + "loss": 0.78208703, + "num_input_tokens_seen": 118582835, + "step": 5513, + "time_per_iteration": 2.7443301677703857 + }, + { + "auxiliary_loss_clip": 0.01117028, + "auxiliary_loss_mlp": 0.01031762, + "balance_loss_clip": 0.97403574, + "balance_loss_mlp": 1.02373934, + "epoch": 0.6630192989839476, + "flos": 18513567106560.0, + "grad_norm": 1.5052895024613435, + "language_loss": 0.70164216, + "learning_rate": 1.0774049747071883e-06, + "loss": 0.72312999, + "num_input_tokens_seen": 118600715, + "step": 5514, + "time_per_iteration": 2.6853041648864746 + }, + { + "auxiliary_loss_clip": 0.01073556, + "auxiliary_loss_mlp": 0.01023284, + "balance_loss_clip": 0.89739263, + "balance_loss_mlp": 1.0160507, + "epoch": 0.6631395418745867, + "flos": 35809510049280.0, + "grad_norm": 1.590604874740414, + "language_loss": 0.68182606, + "learning_rate": 1.076713905602332e-06, + "loss": 0.70279443, + "num_input_tokens_seen": 118621290, + "step": 5515, + "time_per_iteration": 2.790616512298584 + }, + { + "auxiliary_loss_clip": 0.01153171, + "auxiliary_loss_mlp": 0.01025208, + "balance_loss_clip": 1.01670933, + "balance_loss_mlp": 1.01803207, + "epoch": 0.6632597847652257, + "flos": 20047742853120.0, + "grad_norm": 1.6366255184341123, + "language_loss": 0.81380868, + "learning_rate": 1.07602297655924e-06, + "loss": 0.83559251, + "num_input_tokens_seen": 118639610, + "step": 5516, + "time_per_iteration": 2.6179676055908203 + }, + { + "auxiliary_loss_clip": 0.01172255, + "auxiliary_loss_mlp": 0.01030975, + "balance_loss_clip": 1.05401111, + "balance_loss_mlp": 1.02355146, + "epoch": 0.6633800276558649, + "flos": 21214480423680.0, + "grad_norm": 1.8107025771741423, + "language_loss": 0.81244922, + "learning_rate": 1.0753321876827292e-06, + "loss": 0.83448148, + "num_input_tokens_seen": 118658895, + "step": 5517, + "time_per_iteration": 2.5949554443359375 + }, + { + "auxiliary_loss_clip": 0.01167549, + "auxiliary_loss_mlp": 0.01030841, + "balance_loss_clip": 1.04895759, + "balance_loss_mlp": 1.02343774, + "epoch": 0.663500270546504, + "flos": 23987753688960.0, + "grad_norm": 1.9506934013364294, + "language_loss": 0.74175847, + "learning_rate": 1.0746415390775893e-06, + "loss": 0.76374245, + "num_input_tokens_seen": 118677025, + "step": 5518, + "time_per_iteration": 2.6303086280822754 + }, + { + "auxiliary_loss_clip": 0.0116854, + "auxiliary_loss_mlp": 0.01030972, + "balance_loss_clip": 1.05270195, + "balance_loss_mlp": 1.02358675, + "epoch": 0.663620513437143, + "flos": 17932389050880.0, + "grad_norm": 1.7958372769289952, + "language_loss": 0.7654739, + "learning_rate": 1.0739510308485939e-06, + "loss": 0.78746903, + "num_input_tokens_seen": 118694240, + "step": 5519, + "time_per_iteration": 2.5915262699127197 + }, + { + "auxiliary_loss_clip": 0.01015098, + "auxiliary_loss_mlp": 0.00999513, + "balance_loss_clip": 0.90603185, + "balance_loss_mlp": 0.9983868, + "epoch": 0.6637407563277821, + "flos": 57840241086720.0, + "grad_norm": 0.8043672720593444, + "language_loss": 0.62518573, + "learning_rate": 1.07326066310049e-06, + "loss": 0.64533186, + "num_input_tokens_seen": 118758365, + "step": 5520, + "time_per_iteration": 3.27596116065979 + }, + { + "auxiliary_loss_clip": 0.01098306, + "auxiliary_loss_mlp": 0.01028676, + "balance_loss_clip": 0.93519837, + "balance_loss_mlp": 1.02045083, + "epoch": 0.6638609992184212, + "flos": 27306007079040.0, + "grad_norm": 1.8907973953474777, + "language_loss": 0.79659152, + "learning_rate": 1.0725704359380059e-06, + "loss": 0.81786144, + "num_input_tokens_seen": 118778220, + "step": 5521, + "time_per_iteration": 2.7297797203063965 + }, + { + "auxiliary_loss_clip": 0.01171029, + "auxiliary_loss_mlp": 0.01025845, + "balance_loss_clip": 1.05229294, + "balance_loss_mlp": 1.01829267, + "epoch": 0.6639812421090603, + "flos": 18624854419200.0, + "grad_norm": 1.860292034640944, + "language_loss": 0.72247863, + "learning_rate": 1.0718803494658497e-06, + "loss": 0.74444735, + "num_input_tokens_seen": 118797110, + "step": 5522, + "time_per_iteration": 2.6164755821228027 + }, + { + "auxiliary_loss_clip": 0.01000401, + "auxiliary_loss_mlp": 0.01029513, + "balance_loss_clip": 0.78258812, + "balance_loss_mlp": 1.02190161, + "epoch": 0.6641014849996993, + "flos": 15924479806080.0, + "grad_norm": 1.9613968491901883, + "language_loss": 0.83753926, + "learning_rate": 1.071190403788707e-06, + "loss": 0.85783839, + "num_input_tokens_seen": 118812415, + "step": 5523, + "time_per_iteration": 3.7260870933532715 + }, + { + "auxiliary_loss_clip": 0.01107076, + "auxiliary_loss_mlp": 0.0102378, + "balance_loss_clip": 0.93884408, + "balance_loss_mlp": 1.01597202, + "epoch": 0.6642217278903385, + "flos": 26505486622080.0, + "grad_norm": 1.8144864679621284, + "language_loss": 0.75785327, + "learning_rate": 1.0705005990112415e-06, + "loss": 0.77916181, + "num_input_tokens_seen": 118832195, + "step": 5524, + "time_per_iteration": 4.006349325180054 + }, + { + "auxiliary_loss_clip": 0.01068376, + "auxiliary_loss_mlp": 0.01038001, + "balance_loss_clip": 0.89937246, + "balance_loss_mlp": 1.03037131, + "epoch": 0.6643419707809776, + "flos": 15377308951680.0, + "grad_norm": 2.7167779757016985, + "language_loss": 0.74232817, + "learning_rate": 1.0698109352380957e-06, + "loss": 0.76339197, + "num_input_tokens_seen": 118849795, + "step": 5525, + "time_per_iteration": 2.7734975814819336 + }, + { + "auxiliary_loss_clip": 0.01168793, + "auxiliary_loss_mlp": 0.01028092, + "balance_loss_clip": 1.0510422, + "balance_loss_mlp": 1.02090716, + "epoch": 0.6644622136716166, + "flos": 25117610970240.0, + "grad_norm": 2.5158543888498426, + "language_loss": 0.77758223, + "learning_rate": 1.0691214125738909e-06, + "loss": 0.79955107, + "num_input_tokens_seen": 118870000, + "step": 5526, + "time_per_iteration": 3.5722031593322754 + }, + { + "auxiliary_loss_clip": 0.01070279, + "auxiliary_loss_mlp": 0.00999746, + "balance_loss_clip": 1.01682806, + "balance_loss_mlp": 0.99867362, + "epoch": 0.6645824565622558, + "flos": 66201717680640.0, + "grad_norm": 0.829638290002717, + "language_loss": 0.57567704, + "learning_rate": 1.0684320311232287e-06, + "loss": 0.59637737, + "num_input_tokens_seen": 118932905, + "step": 5527, + "time_per_iteration": 3.259105920791626 + }, + { + "auxiliary_loss_clip": 0.01118438, + "auxiliary_loss_mlp": 0.01025714, + "balance_loss_clip": 0.97303051, + "balance_loss_mlp": 1.0180732, + "epoch": 0.6647026994528948, + "flos": 25082131311360.0, + "grad_norm": 2.2044564572691994, + "language_loss": 0.81428772, + "learning_rate": 1.0677427909906865e-06, + "loss": 0.83572924, + "num_input_tokens_seen": 118953355, + "step": 5528, + "time_per_iteration": 2.731198787689209 + }, + { + "auxiliary_loss_clip": 0.01173032, + "auxiliary_loss_mlp": 0.01028439, + "balance_loss_clip": 1.05124128, + "balance_loss_mlp": 1.02046394, + "epoch": 0.6648229423435339, + "flos": 18222187979520.0, + "grad_norm": 2.4494123733940767, + "language_loss": 0.72284043, + "learning_rate": 1.0670536922808216e-06, + "loss": 0.74485511, + "num_input_tokens_seen": 118973480, + "step": 5529, + "time_per_iteration": 2.622896432876587 + }, + { + "auxiliary_loss_clip": 0.0112367, + "auxiliary_loss_mlp": 0.01031325, + "balance_loss_clip": 0.97465539, + "balance_loss_mlp": 1.02374041, + "epoch": 0.6649431852341731, + "flos": 18296882311680.0, + "grad_norm": 1.9060000118125395, + "language_loss": 0.7168659, + "learning_rate": 1.06636473509817e-06, + "loss": 0.73841584, + "num_input_tokens_seen": 118989860, + "step": 5530, + "time_per_iteration": 2.7070846557617188 + }, + { + "auxiliary_loss_clip": 0.01119987, + "auxiliary_loss_mlp": 0.00509865, + "balance_loss_clip": 0.97335625, + "balance_loss_mlp": 0.0, + "epoch": 0.6650634281248121, + "flos": 17019575700480.0, + "grad_norm": 1.8392288286519638, + "language_loss": 0.80726981, + "learning_rate": 1.0656759195472447e-06, + "loss": 0.82356834, + "num_input_tokens_seen": 119007150, + "step": 5531, + "time_per_iteration": 2.6634695529937744 + }, + { + "auxiliary_loss_clip": 0.0102742, + "auxiliary_loss_mlp": 0.01000653, + "balance_loss_clip": 0.94193506, + "balance_loss_mlp": 0.99957412, + "epoch": 0.6651836710154512, + "flos": 69294810666240.0, + "grad_norm": 0.7705694458659239, + "language_loss": 0.59734797, + "learning_rate": 1.0649872457325414e-06, + "loss": 0.61762869, + "num_input_tokens_seen": 119068435, + "step": 5532, + "time_per_iteration": 3.1760478019714355 + }, + { + "auxiliary_loss_clip": 0.01052755, + "auxiliary_loss_mlp": 0.01001228, + "balance_loss_clip": 0.98029572, + "balance_loss_mlp": 1.00016689, + "epoch": 0.6653039139060903, + "flos": 66883444882560.0, + "grad_norm": 0.8712727919559686, + "language_loss": 0.55117083, + "learning_rate": 1.0642987137585278e-06, + "loss": 0.57171071, + "num_input_tokens_seen": 119127960, + "step": 5533, + "time_per_iteration": 3.1810619831085205 + }, + { + "auxiliary_loss_clip": 0.01122648, + "auxiliary_loss_mlp": 0.01026104, + "balance_loss_clip": 0.97442847, + "balance_loss_mlp": 1.01799786, + "epoch": 0.6654241567967294, + "flos": 21470056669440.0, + "grad_norm": 1.659424077115174, + "language_loss": 0.82623541, + "learning_rate": 1.0636103237296561e-06, + "loss": 0.84772295, + "num_input_tokens_seen": 119146885, + "step": 5534, + "time_per_iteration": 2.662757635116577 + }, + { + "auxiliary_loss_clip": 0.01146739, + "auxiliary_loss_mlp": 0.01023578, + "balance_loss_clip": 1.01586473, + "balance_loss_mlp": 1.01625276, + "epoch": 0.6655443996873684, + "flos": 25119514391040.0, + "grad_norm": 2.867072801117843, + "language_loss": 0.84266633, + "learning_rate": 1.062922075750353e-06, + "loss": 0.86436951, + "num_input_tokens_seen": 119166900, + "step": 5535, + "time_per_iteration": 2.6711041927337646 + }, + { + "auxiliary_loss_clip": 0.01101429, + "auxiliary_loss_mlp": 0.01025095, + "balance_loss_clip": 0.93606377, + "balance_loss_mlp": 1.01803458, + "epoch": 0.6656646425780076, + "flos": 17457326749440.0, + "grad_norm": 1.881141738646765, + "language_loss": 0.72425246, + "learning_rate": 1.0622339699250267e-06, + "loss": 0.74551773, + "num_input_tokens_seen": 119184820, + "step": 5536, + "time_per_iteration": 3.6145310401916504 + }, + { + "auxiliary_loss_clip": 0.01102606, + "auxiliary_loss_mlp": 0.01023828, + "balance_loss_clip": 0.93760431, + "balance_loss_mlp": 1.01637483, + "epoch": 0.6657848854686467, + "flos": 23434190213760.0, + "grad_norm": 1.7013351661007425, + "language_loss": 0.79632658, + "learning_rate": 1.0615460063580624e-06, + "loss": 0.81759095, + "num_input_tokens_seen": 119203295, + "step": 5537, + "time_per_iteration": 2.708894968032837 + }, + { + "auxiliary_loss_clip": 0.01126835, + "auxiliary_loss_mlp": 0.01022334, + "balance_loss_clip": 0.97408909, + "balance_loss_mlp": 1.01520586, + "epoch": 0.6659051283592857, + "flos": 11509909459200.0, + "grad_norm": 1.8747787220061842, + "language_loss": 0.72882241, + "learning_rate": 1.060858185153821e-06, + "loss": 0.75031412, + "num_input_tokens_seen": 119221395, + "step": 5538, + "time_per_iteration": 2.6752614974975586 + }, + { + "auxiliary_loss_clip": 0.0112905, + "auxiliary_loss_mlp": 0.01033511, + "balance_loss_clip": 0.97592366, + "balance_loss_mlp": 1.0252676, + "epoch": 0.6660253712499249, + "flos": 20594554571520.0, + "grad_norm": 2.1886140985828417, + "language_loss": 0.76257044, + "learning_rate": 1.0601705064166474e-06, + "loss": 0.78419602, + "num_input_tokens_seen": 119239790, + "step": 5539, + "time_per_iteration": 2.6636602878570557 + }, + { + "auxiliary_loss_clip": 0.01124011, + "auxiliary_loss_mlp": 0.01027083, + "balance_loss_clip": 0.97958189, + "balance_loss_mlp": 1.01897037, + "epoch": 0.666145614140564, + "flos": 21251504367360.0, + "grad_norm": 2.280568020131499, + "language_loss": 0.73612905, + "learning_rate": 1.0594829702508596e-06, + "loss": 0.75764, + "num_input_tokens_seen": 119257505, + "step": 5540, + "time_per_iteration": 2.7335073947906494 + }, + { + "auxiliary_loss_clip": 0.01104767, + "auxiliary_loss_mlp": 0.01024831, + "balance_loss_clip": 0.93758339, + "balance_loss_mlp": 1.01749349, + "epoch": 0.666265857031203, + "flos": 33726188200320.0, + "grad_norm": 1.5881288929345885, + "language_loss": 0.54919708, + "learning_rate": 1.0587955767607592e-06, + "loss": 0.57049304, + "num_input_tokens_seen": 119279365, + "step": 5541, + "time_per_iteration": 2.8139092922210693 + }, + { + "auxiliary_loss_clip": 0.01169388, + "auxiliary_loss_mlp": 0.01026093, + "balance_loss_clip": 1.05073977, + "balance_loss_mlp": 1.0182786, + "epoch": 0.6663860999218422, + "flos": 17456644391040.0, + "grad_norm": 2.962709615565545, + "language_loss": 0.77093828, + "learning_rate": 1.0581083260506206e-06, + "loss": 0.79289305, + "num_input_tokens_seen": 119296150, + "step": 5542, + "time_per_iteration": 2.6426663398742676 + }, + { + "auxiliary_loss_clip": 0.01121499, + "auxiliary_loss_mlp": 0.01024845, + "balance_loss_clip": 0.97352886, + "balance_loss_mlp": 1.0176028, + "epoch": 0.6665063428124812, + "flos": 17676740977920.0, + "grad_norm": 2.069559522410866, + "language_loss": 0.76370788, + "learning_rate": 1.0574212182246993e-06, + "loss": 0.78517133, + "num_input_tokens_seen": 119314845, + "step": 5543, + "time_per_iteration": 2.619596481323242 + }, + { + "auxiliary_loss_clip": 0.01127787, + "auxiliary_loss_mlp": 0.01028107, + "balance_loss_clip": 0.97403532, + "balance_loss_mlp": 1.02037621, + "epoch": 0.6666265857031203, + "flos": 27673265687040.0, + "grad_norm": 2.7577812823798125, + "language_loss": 0.76164269, + "learning_rate": 1.0567342533872303e-06, + "loss": 0.78320163, + "num_input_tokens_seen": 119334875, + "step": 5544, + "time_per_iteration": 2.6508381366729736 + }, + { + "auxiliary_loss_clip": 0.01124192, + "auxiliary_loss_mlp": 0.01028082, + "balance_loss_clip": 0.97553021, + "balance_loss_mlp": 1.02022052, + "epoch": 0.6667468285937594, + "flos": 25046831220480.0, + "grad_norm": 1.894018097107749, + "language_loss": 0.81073111, + "learning_rate": 1.0560474316424255e-06, + "loss": 0.83225387, + "num_input_tokens_seen": 119354635, + "step": 5545, + "time_per_iteration": 2.6996419429779053 + }, + { + "auxiliary_loss_clip": 0.0112247, + "auxiliary_loss_mlp": 0.01027654, + "balance_loss_clip": 0.97178328, + "balance_loss_mlp": 1.01904082, + "epoch": 0.6668670714843985, + "flos": 22780472641920.0, + "grad_norm": 2.3630758251313866, + "language_loss": 0.74110508, + "learning_rate": 1.0553607530944746e-06, + "loss": 0.76260632, + "num_input_tokens_seen": 119372690, + "step": 5546, + "time_per_iteration": 2.6743316650390625 + }, + { + "auxiliary_loss_clip": 0.01100991, + "auxiliary_loss_mlp": 0.01030054, + "balance_loss_clip": 0.93354112, + "balance_loss_mlp": 1.02256155, + "epoch": 0.6669873143750376, + "flos": 22163886754560.0, + "grad_norm": 2.008428954819521, + "language_loss": 0.89790142, + "learning_rate": 1.0546742178475463e-06, + "loss": 0.91921192, + "num_input_tokens_seen": 119391685, + "step": 5547, + "time_per_iteration": 2.685417413711548 + }, + { + "auxiliary_loss_clip": 0.01080028, + "auxiliary_loss_mlp": 0.01026505, + "balance_loss_clip": 0.89748192, + "balance_loss_mlp": 1.01957345, + "epoch": 0.6671075572656767, + "flos": 20514832335360.0, + "grad_norm": 1.7656482069154178, + "language_loss": 0.86760664, + "learning_rate": 1.0539878260057868e-06, + "loss": 0.88867199, + "num_input_tokens_seen": 119410725, + "step": 5548, + "time_per_iteration": 2.724482536315918 + }, + { + "auxiliary_loss_clip": 0.01151008, + "auxiliary_loss_mlp": 0.01031873, + "balance_loss_clip": 1.0174576, + "balance_loss_mlp": 1.02385581, + "epoch": 0.6672278001563158, + "flos": 17931203902080.0, + "grad_norm": 2.6659185881072003, + "language_loss": 0.68799865, + "learning_rate": 1.0533015776733226e-06, + "loss": 0.70982742, + "num_input_tokens_seen": 119426875, + "step": 5549, + "time_per_iteration": 3.573974370956421 + }, + { + "auxiliary_loss_clip": 0.01121768, + "auxiliary_loss_mlp": 0.01029097, + "balance_loss_clip": 0.97509611, + "balance_loss_mlp": 1.02121735, + "epoch": 0.6673480430469548, + "flos": 22342146975360.0, + "grad_norm": 2.80363129968065, + "language_loss": 0.79251528, + "learning_rate": 1.0526154729542566e-06, + "loss": 0.81402397, + "num_input_tokens_seen": 119446935, + "step": 5550, + "time_per_iteration": 3.705512046813965 + }, + { + "auxiliary_loss_clip": 0.01102363, + "auxiliary_loss_mlp": 0.01029745, + "balance_loss_clip": 0.93923837, + "balance_loss_mlp": 1.02214217, + "epoch": 0.6674682859375939, + "flos": 20703830722560.0, + "grad_norm": 4.115271636187283, + "language_loss": 0.79983723, + "learning_rate": 1.0519295119526699e-06, + "loss": 0.82115835, + "num_input_tokens_seen": 119463240, + "step": 5551, + "time_per_iteration": 2.756636619567871 + }, + { + "auxiliary_loss_clip": 0.01127374, + "auxiliary_loss_mlp": 0.01023406, + "balance_loss_clip": 0.97450149, + "balance_loss_mlp": 1.01577616, + "epoch": 0.667588528828233, + "flos": 26206673379840.0, + "grad_norm": 1.7181046134509483, + "language_loss": 0.82906681, + "learning_rate": 1.0512436947726227e-06, + "loss": 0.85057461, + "num_input_tokens_seen": 119484655, + "step": 5552, + "time_per_iteration": 3.718090534210205 + }, + { + "auxiliary_loss_clip": 0.0110098, + "auxiliary_loss_mlp": 0.01031191, + "balance_loss_clip": 0.93542051, + "balance_loss_mlp": 1.02342415, + "epoch": 0.6677087717188721, + "flos": 23071025756160.0, + "grad_norm": 2.1868311877060207, + "language_loss": 0.65260911, + "learning_rate": 1.0505580215181517e-06, + "loss": 0.67393088, + "num_input_tokens_seen": 119502895, + "step": 5553, + "time_per_iteration": 2.7341532707214355 + }, + { + "auxiliary_loss_clip": 0.01002779, + "auxiliary_loss_mlp": 0.01018421, + "balance_loss_clip": 0.90646195, + "balance_loss_mlp": 1.01722932, + "epoch": 0.6678290146095112, + "flos": 70941315219840.0, + "grad_norm": 0.8020515792024673, + "language_loss": 0.56577396, + "learning_rate": 1.0498724922932753e-06, + "loss": 0.58598602, + "num_input_tokens_seen": 119561010, + "step": 5554, + "time_per_iteration": 3.1899707317352295 + }, + { + "auxiliary_loss_clip": 0.01178622, + "auxiliary_loss_mlp": 0.01034063, + "balance_loss_clip": 1.05674851, + "balance_loss_mlp": 1.02583194, + "epoch": 0.6679492575001503, + "flos": 18661088263680.0, + "grad_norm": 1.9659882824771515, + "language_loss": 0.86748606, + "learning_rate": 1.0491871072019851e-06, + "loss": 0.88961297, + "num_input_tokens_seen": 119578900, + "step": 5555, + "time_per_iteration": 2.640352249145508 + }, + { + "auxiliary_loss_clip": 0.01103134, + "auxiliary_loss_mlp": 0.01026623, + "balance_loss_clip": 0.93264568, + "balance_loss_mlp": 1.01895761, + "epoch": 0.6680695003907894, + "flos": 29711985822720.0, + "grad_norm": 1.8797902798774666, + "language_loss": 0.64196312, + "learning_rate": 1.0485018663482555e-06, + "loss": 0.6632607, + "num_input_tokens_seen": 119598920, + "step": 5556, + "time_per_iteration": 2.739605665206909 + }, + { + "auxiliary_loss_clip": 0.01144219, + "auxiliary_loss_mlp": 0.01025407, + "balance_loss_clip": 1.01349521, + "balance_loss_mlp": 1.01789045, + "epoch": 0.6681897432814284, + "flos": 28218964083840.0, + "grad_norm": 2.5042454287951394, + "language_loss": 0.71358538, + "learning_rate": 1.0478167698360354e-06, + "loss": 0.73528159, + "num_input_tokens_seen": 119618220, + "step": 5557, + "time_per_iteration": 2.704832077026367 + }, + { + "auxiliary_loss_clip": 0.01139471, + "auxiliary_loss_mlp": 0.01026733, + "balance_loss_clip": 1.0097717, + "balance_loss_mlp": 1.01940727, + "epoch": 0.6683099861720676, + "flos": 25046543911680.0, + "grad_norm": 1.994615521637999, + "language_loss": 0.7004559, + "learning_rate": 1.0471318177692556e-06, + "loss": 0.7221179, + "num_input_tokens_seen": 119638520, + "step": 5558, + "time_per_iteration": 2.678058624267578 + }, + { + "auxiliary_loss_clip": 0.01077665, + "auxiliary_loss_mlp": 0.01023314, + "balance_loss_clip": 0.89576471, + "balance_loss_mlp": 1.01580715, + "epoch": 0.6684302290627067, + "flos": 22996977868800.0, + "grad_norm": 2.268446692011797, + "language_loss": 0.75790763, + "learning_rate": 1.046447010251821e-06, + "loss": 0.77891743, + "num_input_tokens_seen": 119655850, + "step": 5559, + "time_per_iteration": 2.7692322731018066 + }, + { + "auxiliary_loss_clip": 0.01122918, + "auxiliary_loss_mlp": 0.01028263, + "balance_loss_clip": 0.97549075, + "balance_loss_mlp": 1.02074385, + "epoch": 0.6685504719533457, + "flos": 26573824247040.0, + "grad_norm": 1.7422266811635794, + "language_loss": 0.75880992, + "learning_rate": 1.0457623473876157e-06, + "loss": 0.78032172, + "num_input_tokens_seen": 119675355, + "step": 5560, + "time_per_iteration": 2.7125089168548584 + }, + { + "auxiliary_loss_clip": 0.01169118, + "auxiliary_loss_mlp": 0.01030004, + "balance_loss_clip": 1.05140209, + "balance_loss_mlp": 1.02286649, + "epoch": 0.6686707148439849, + "flos": 28986087870720.0, + "grad_norm": 2.490722744489308, + "language_loss": 0.70916623, + "learning_rate": 1.0450778292805046e-06, + "loss": 0.73115748, + "num_input_tokens_seen": 119695340, + "step": 5561, + "time_per_iteration": 2.6624269485473633 + }, + { + "auxiliary_loss_clip": 0.01151217, + "auxiliary_loss_mlp": 0.01027753, + "balance_loss_clip": 1.01323199, + "balance_loss_mlp": 1.0200758, + "epoch": 0.6687909577346239, + "flos": 23623152687360.0, + "grad_norm": 1.4779428600177529, + "language_loss": 0.78456688, + "learning_rate": 1.0443934560343267e-06, + "loss": 0.80635667, + "num_input_tokens_seen": 119716750, + "step": 5562, + "time_per_iteration": 3.6070942878723145 + }, + { + "auxiliary_loss_clip": 0.01090704, + "auxiliary_loss_mlp": 0.01024, + "balance_loss_clip": 0.93554509, + "balance_loss_mlp": 1.01635563, + "epoch": 0.668911200625263, + "flos": 23148593176320.0, + "grad_norm": 2.376343190647375, + "language_loss": 0.77636933, + "learning_rate": 1.0437092277529034e-06, + "loss": 0.79751635, + "num_input_tokens_seen": 119736005, + "step": 5563, + "time_per_iteration": 2.696761131286621 + }, + { + "auxiliary_loss_clip": 0.01119199, + "auxiliary_loss_mlp": 0.01023753, + "balance_loss_clip": 0.9725467, + "balance_loss_mlp": 1.01650476, + "epoch": 0.6690314435159022, + "flos": 18551919853440.0, + "grad_norm": 2.0748570241317914, + "language_loss": 0.74055159, + "learning_rate": 1.0430251445400292e-06, + "loss": 0.76198113, + "num_input_tokens_seen": 119754050, + "step": 5564, + "time_per_iteration": 2.721614122390747 + }, + { + "auxiliary_loss_clip": 0.01006327, + "auxiliary_loss_mlp": 0.01026321, + "balance_loss_clip": 0.78335035, + "balance_loss_mlp": 1.0185312, + "epoch": 0.6691516864065412, + "flos": 31759540704000.0, + "grad_norm": 3.1269084175798745, + "language_loss": 0.62714785, + "learning_rate": 1.0423412064994787e-06, + "loss": 0.64747435, + "num_input_tokens_seen": 119774820, + "step": 5565, + "time_per_iteration": 3.3866028785705566 + }, + { + "auxiliary_loss_clip": 0.0109931, + "auxiliary_loss_mlp": 0.01026849, + "balance_loss_clip": 0.93308252, + "balance_loss_mlp": 1.02022409, + "epoch": 0.6692719292971803, + "flos": 34933864296960.0, + "grad_norm": 1.8059862110767735, + "language_loss": 0.73997027, + "learning_rate": 1.0416574137350064e-06, + "loss": 0.7612319, + "num_input_tokens_seen": 119795525, + "step": 5566, + "time_per_iteration": 3.082972764968872 + }, + { + "auxiliary_loss_clip": 0.0114038, + "auxiliary_loss_mlp": 0.01024821, + "balance_loss_clip": 1.01071596, + "balance_loss_mlp": 1.01773071, + "epoch": 0.6693921721878194, + "flos": 20449188230400.0, + "grad_norm": 2.152071914831999, + "language_loss": 0.80973685, + "learning_rate": 1.0409737663503428e-06, + "loss": 0.83138883, + "num_input_tokens_seen": 119813905, + "step": 5567, + "time_per_iteration": 2.6464908123016357 + }, + { + "auxiliary_loss_clip": 0.01140963, + "auxiliary_loss_mlp": 0.01026342, + "balance_loss_clip": 1.00839067, + "balance_loss_mlp": 1.01877213, + "epoch": 0.6695124150784585, + "flos": 16614538963200.0, + "grad_norm": 1.8305809376980622, + "language_loss": 0.8320632, + "learning_rate": 1.040290264449196e-06, + "loss": 0.85373622, + "num_input_tokens_seen": 119832010, + "step": 5568, + "time_per_iteration": 2.679281234741211 + }, + { + "auxiliary_loss_clip": 0.01142889, + "auxiliary_loss_mlp": 0.01027492, + "balance_loss_clip": 1.01343226, + "balance_loss_mlp": 1.02021456, + "epoch": 0.6696326579690975, + "flos": 26652145852800.0, + "grad_norm": 2.117579600641754, + "language_loss": 0.63794655, + "learning_rate": 1.0396069081352532e-06, + "loss": 0.65965033, + "num_input_tokens_seen": 119851165, + "step": 5569, + "time_per_iteration": 2.705983877182007 + }, + { + "auxiliary_loss_clip": 0.01066168, + "auxiliary_loss_mlp": 0.0100023, + "balance_loss_clip": 1.01346111, + "balance_loss_mlp": 0.99913961, + "epoch": 0.6697529008597367, + "flos": 66964603662720.0, + "grad_norm": 0.7730637468336253, + "language_loss": 0.56019509, + "learning_rate": 1.0389236975121782e-06, + "loss": 0.58085907, + "num_input_tokens_seen": 119906015, + "step": 5570, + "time_per_iteration": 3.047776699066162 + }, + { + "auxiliary_loss_clip": 0.01171962, + "auxiliary_loss_mlp": 0.01026987, + "balance_loss_clip": 1.05217218, + "balance_loss_mlp": 1.01895833, + "epoch": 0.6698731437503758, + "flos": 20886939279360.0, + "grad_norm": 2.475590613493509, + "language_loss": 0.71187019, + "learning_rate": 1.0382406326836147e-06, + "loss": 0.73385966, + "num_input_tokens_seen": 119925160, + "step": 5571, + "time_per_iteration": 2.615964651107788 + }, + { + "auxiliary_loss_clip": 0.01152037, + "auxiliary_loss_mlp": 0.01023576, + "balance_loss_clip": 1.01417959, + "balance_loss_mlp": 1.01561534, + "epoch": 0.6699933866410148, + "flos": 20409470766720.0, + "grad_norm": 2.2460317641518546, + "language_loss": 0.75849485, + "learning_rate": 1.0375577137531828e-06, + "loss": 0.78025097, + "num_input_tokens_seen": 119943720, + "step": 5572, + "time_per_iteration": 2.6661856174468994 + }, + { + "auxiliary_loss_clip": 0.01126381, + "auxiliary_loss_mlp": 0.01024281, + "balance_loss_clip": 0.97545415, + "balance_loss_mlp": 1.01689339, + "epoch": 0.670113629531654, + "flos": 29023075900800.0, + "grad_norm": 2.9419983288111275, + "language_loss": 0.72228312, + "learning_rate": 1.0368749408244802e-06, + "loss": 0.74378967, + "num_input_tokens_seen": 119966640, + "step": 5573, + "time_per_iteration": 2.8388781547546387 + }, + { + "auxiliary_loss_clip": 0.01143571, + "auxiliary_loss_mlp": 0.0102803, + "balance_loss_clip": 1.01426852, + "balance_loss_mlp": 1.02071011, + "epoch": 0.670233872422293, + "flos": 19791699730560.0, + "grad_norm": 1.7265593123513834, + "language_loss": 0.7907353, + "learning_rate": 1.0361923140010836e-06, + "loss": 0.81245124, + "num_input_tokens_seen": 119985125, + "step": 5574, + "time_per_iteration": 2.778289794921875 + }, + { + "auxiliary_loss_clip": 0.01151091, + "auxiliary_loss_mlp": 0.01028305, + "balance_loss_clip": 1.01187825, + "balance_loss_mlp": 1.0203656, + "epoch": 0.6703541153129321, + "flos": 24243689070720.0, + "grad_norm": 1.8499341856179081, + "language_loss": 0.63643909, + "learning_rate": 1.0355098333865455e-06, + "loss": 0.65823305, + "num_input_tokens_seen": 120004355, + "step": 5575, + "time_per_iteration": 3.98189115524292 + }, + { + "auxiliary_loss_clip": 0.01149962, + "auxiliary_loss_mlp": 0.01028876, + "balance_loss_clip": 1.0212692, + "balance_loss_mlp": 1.02173519, + "epoch": 0.6704743582035713, + "flos": 26688523351680.0, + "grad_norm": 1.6698905646709696, + "language_loss": 0.69379145, + "learning_rate": 1.0348274990844006e-06, + "loss": 0.71557981, + "num_input_tokens_seen": 120027115, + "step": 5576, + "time_per_iteration": 3.6074557304382324 + }, + { + "auxiliary_loss_clip": 0.0114641, + "auxiliary_loss_mlp": 0.0103222, + "balance_loss_clip": 1.01387608, + "balance_loss_mlp": 1.02483499, + "epoch": 0.6705946010942103, + "flos": 23514379326720.0, + "grad_norm": 1.7808771141204724, + "language_loss": 0.72791076, + "learning_rate": 1.034145311198155e-06, + "loss": 0.74969709, + "num_input_tokens_seen": 120047130, + "step": 5577, + "time_per_iteration": 2.6634862422943115 + }, + { + "auxiliary_loss_clip": 0.01167449, + "auxiliary_loss_mlp": 0.01027968, + "balance_loss_clip": 1.05059338, + "balance_loss_mlp": 1.02091098, + "epoch": 0.6707148439848494, + "flos": 24061011477120.0, + "grad_norm": 2.4224152737128644, + "language_loss": 0.64031649, + "learning_rate": 1.0334632698312989e-06, + "loss": 0.66227067, + "num_input_tokens_seen": 120067925, + "step": 5578, + "time_per_iteration": 3.6935248374938965 + }, + { + "auxiliary_loss_clip": 0.01117955, + "auxiliary_loss_mlp": 0.01028263, + "balance_loss_clip": 0.97406483, + "balance_loss_mlp": 1.02099717, + "epoch": 0.6708350868754885, + "flos": 22528667324160.0, + "grad_norm": 2.901853443614453, + "language_loss": 0.75231433, + "learning_rate": 1.032781375087295e-06, + "loss": 0.77377653, + "num_input_tokens_seen": 120087825, + "step": 5579, + "time_per_iteration": 2.722325563430786 + }, + { + "auxiliary_loss_clip": 0.01127659, + "auxiliary_loss_mlp": 0.01028719, + "balance_loss_clip": 0.97756863, + "balance_loss_mlp": 1.02113152, + "epoch": 0.6709553297661276, + "flos": 25227749047680.0, + "grad_norm": 1.3818367066490491, + "language_loss": 0.67362332, + "learning_rate": 1.0320996270695891e-06, + "loss": 0.69518709, + "num_input_tokens_seen": 120108895, + "step": 5580, + "time_per_iteration": 2.691431760787964 + }, + { + "auxiliary_loss_clip": 0.01097661, + "auxiliary_loss_mlp": 0.01023372, + "balance_loss_clip": 0.93213046, + "balance_loss_mlp": 1.01590347, + "epoch": 0.6710755726567667, + "flos": 20448757267200.0, + "grad_norm": 1.8768319266412035, + "language_loss": 0.73575318, + "learning_rate": 1.0314180258815998e-06, + "loss": 0.75696361, + "num_input_tokens_seen": 120127535, + "step": 5581, + "time_per_iteration": 2.7371954917907715 + }, + { + "auxiliary_loss_clip": 0.01090509, + "auxiliary_loss_mlp": 0.01024016, + "balance_loss_clip": 0.93197942, + "balance_loss_mlp": 1.01657784, + "epoch": 0.6711958155474057, + "flos": 25995411538560.0, + "grad_norm": 1.5349722782074624, + "language_loss": 0.74397796, + "learning_rate": 1.0307365716267247e-06, + "loss": 0.76512325, + "num_input_tokens_seen": 120147980, + "step": 5582, + "time_per_iteration": 2.757352828979492 + }, + { + "auxiliary_loss_clip": 0.01144985, + "auxiliary_loss_mlp": 0.01022925, + "balance_loss_clip": 1.01231003, + "balance_loss_mlp": 1.01537919, + "epoch": 0.6713160584380449, + "flos": 19937712516480.0, + "grad_norm": 2.35795744709578, + "language_loss": 0.78253579, + "learning_rate": 1.0300552644083423e-06, + "loss": 0.80421484, + "num_input_tokens_seen": 120166905, + "step": 5583, + "time_per_iteration": 2.7020885944366455 + }, + { + "auxiliary_loss_clip": 0.01110186, + "auxiliary_loss_mlp": 0.01030217, + "balance_loss_clip": 0.94349992, + "balance_loss_mlp": 1.02194452, + "epoch": 0.6714363013286839, + "flos": 18223373128320.0, + "grad_norm": 6.477199426896484, + "language_loss": 0.72654784, + "learning_rate": 1.0293741043298036e-06, + "loss": 0.74795187, + "num_input_tokens_seen": 120185255, + "step": 5584, + "time_per_iteration": 2.6473350524902344 + }, + { + "auxiliary_loss_clip": 0.01111938, + "auxiliary_loss_mlp": 0.01028003, + "balance_loss_clip": 0.94612855, + "balance_loss_mlp": 1.02033186, + "epoch": 0.671556544219323, + "flos": 25812374808960.0, + "grad_norm": 2.1309220563626043, + "language_loss": 0.71065891, + "learning_rate": 1.0286930914944436e-06, + "loss": 0.73205835, + "num_input_tokens_seen": 120205070, + "step": 5585, + "time_per_iteration": 2.7569026947021484 + }, + { + "auxiliary_loss_clip": 0.01169476, + "auxiliary_loss_mlp": 0.01024596, + "balance_loss_clip": 1.04802966, + "balance_loss_mlp": 1.01705253, + "epoch": 0.6716767871099621, + "flos": 15850431918720.0, + "grad_norm": 2.3915004866562377, + "language_loss": 0.77089965, + "learning_rate": 1.0280122260055684e-06, + "loss": 0.79284036, + "num_input_tokens_seen": 120220780, + "step": 5586, + "time_per_iteration": 2.565486431121826 + }, + { + "auxiliary_loss_clip": 0.01171569, + "auxiliary_loss_mlp": 0.01025339, + "balance_loss_clip": 1.05164206, + "balance_loss_mlp": 1.01768851, + "epoch": 0.6717970300006012, + "flos": 19756112330880.0, + "grad_norm": 1.9226312559051744, + "language_loss": 0.81910145, + "learning_rate": 1.0273315079664652e-06, + "loss": 0.84107059, + "num_input_tokens_seen": 120238735, + "step": 5587, + "time_per_iteration": 2.6157045364379883 + }, + { + "auxiliary_loss_clip": 0.01146999, + "auxiliary_loss_mlp": 0.01025815, + "balance_loss_clip": 1.01203191, + "balance_loss_mlp": 1.01911819, + "epoch": 0.6719172728912403, + "flos": 25485049146240.0, + "grad_norm": 2.115257172588196, + "language_loss": 0.74245071, + "learning_rate": 1.0266509374803992e-06, + "loss": 0.76417887, + "num_input_tokens_seen": 120259895, + "step": 5588, + "time_per_iteration": 3.6321473121643066 + }, + { + "auxiliary_loss_clip": 0.01171349, + "auxiliary_loss_mlp": 0.00509832, + "balance_loss_clip": 1.0510788, + "balance_loss_mlp": 0.0, + "epoch": 0.6720375157818794, + "flos": 15880344969600.0, + "grad_norm": 2.4646556636565, + "language_loss": 0.84469056, + "learning_rate": 1.0259705146506123e-06, + "loss": 0.86150241, + "num_input_tokens_seen": 120274790, + "step": 5589, + "time_per_iteration": 2.58512806892395 + }, + { + "auxiliary_loss_clip": 0.01148021, + "auxiliary_loss_mlp": 0.01027553, + "balance_loss_clip": 1.01252687, + "balance_loss_mlp": 1.02025723, + "epoch": 0.6721577586725185, + "flos": 32010843231360.0, + "grad_norm": 1.8266371770541474, + "language_loss": 0.77454531, + "learning_rate": 1.025290239580324e-06, + "loss": 0.79630113, + "num_input_tokens_seen": 120295460, + "step": 5590, + "time_per_iteration": 2.784982204437256 + }, + { + "auxiliary_loss_clip": 0.01074149, + "auxiliary_loss_mlp": 0.01027591, + "balance_loss_clip": 0.89440477, + "balance_loss_mlp": 1.0205009, + "epoch": 0.6722780015631575, + "flos": 20737873837440.0, + "grad_norm": 1.6627458955743697, + "language_loss": 0.7536307, + "learning_rate": 1.0246101123727313e-06, + "loss": 0.77464813, + "num_input_tokens_seen": 120314440, + "step": 5591, + "time_per_iteration": 2.7270030975341797 + }, + { + "auxiliary_loss_clip": 0.01147356, + "auxiliary_loss_mlp": 0.01025172, + "balance_loss_clip": 1.01223683, + "balance_loss_mlp": 1.01797175, + "epoch": 0.6723982444537967, + "flos": 16909617191040.0, + "grad_norm": 1.7976674093758374, + "language_loss": 0.79015839, + "learning_rate": 1.0239301331310085e-06, + "loss": 0.81188369, + "num_input_tokens_seen": 120332060, + "step": 5592, + "time_per_iteration": 2.667569398880005 + }, + { + "auxiliary_loss_clip": 0.01142788, + "auxiliary_loss_mlp": 0.01028129, + "balance_loss_clip": 1.0116384, + "balance_loss_mlp": 1.0211699, + "epoch": 0.6725184873444358, + "flos": 20667812359680.0, + "grad_norm": 1.7572709129265673, + "language_loss": 0.88399601, + "learning_rate": 1.0232503019583088e-06, + "loss": 0.90570521, + "num_input_tokens_seen": 120351670, + "step": 5593, + "time_per_iteration": 2.647831916809082 + }, + { + "auxiliary_loss_clip": 0.01142397, + "auxiliary_loss_mlp": 0.01023461, + "balance_loss_clip": 1.0128994, + "balance_loss_mlp": 1.01558709, + "epoch": 0.6726387302350748, + "flos": 23727616416000.0, + "grad_norm": 1.7050345404379494, + "language_loss": 0.69800812, + "learning_rate": 1.0225706189577619e-06, + "loss": 0.71966672, + "num_input_tokens_seen": 120370195, + "step": 5594, + "time_per_iteration": 2.7142157554626465 + }, + { + "auxiliary_loss_clip": 0.01148251, + "auxiliary_loss_mlp": 0.01027972, + "balance_loss_clip": 1.01214743, + "balance_loss_mlp": 1.02073622, + "epoch": 0.672758973125714, + "flos": 15188274650880.0, + "grad_norm": 2.0658903454758972, + "language_loss": 0.75118345, + "learning_rate": 1.021891084232475e-06, + "loss": 0.7729457, + "num_input_tokens_seen": 120388130, + "step": 5595, + "time_per_iteration": 2.6803970336914062 + }, + { + "auxiliary_loss_clip": 0.01144428, + "auxiliary_loss_mlp": 0.01020842, + "balance_loss_clip": 1.01064897, + "balance_loss_mlp": 1.01354003, + "epoch": 0.672879216016353, + "flos": 18077252601600.0, + "grad_norm": 2.9489228515140598, + "language_loss": 0.80142266, + "learning_rate": 1.0212116978855325e-06, + "loss": 0.82307529, + "num_input_tokens_seen": 120406145, + "step": 5596, + "time_per_iteration": 2.6523802280426025 + }, + { + "auxiliary_loss_clip": 0.01097027, + "auxiliary_loss_mlp": 0.01027911, + "balance_loss_clip": 0.93525881, + "balance_loss_mlp": 1.02059162, + "epoch": 0.6729994589069921, + "flos": 23476349802240.0, + "grad_norm": 1.7818196474426151, + "language_loss": 0.78873873, + "learning_rate": 1.020532460019997e-06, + "loss": 0.80998814, + "num_input_tokens_seen": 120425395, + "step": 5597, + "time_per_iteration": 2.7828121185302734 + }, + { + "auxiliary_loss_clip": 0.01034808, + "auxiliary_loss_mlp": 0.01026161, + "balance_loss_clip": 0.820952, + "balance_loss_mlp": 1.01873446, + "epoch": 0.6731197017976313, + "flos": 26322018929280.0, + "grad_norm": 2.7308401537554596, + "language_loss": 0.71343911, + "learning_rate": 1.0198533707389096e-06, + "loss": 0.73404884, + "num_input_tokens_seen": 120446270, + "step": 5598, + "time_per_iteration": 2.92031192779541 + }, + { + "auxiliary_loss_clip": 0.01143596, + "auxiliary_loss_mlp": 0.00509578, + "balance_loss_clip": 1.01347721, + "balance_loss_mlp": 0.0, + "epoch": 0.6732399446882703, + "flos": 21616428591360.0, + "grad_norm": 8.77583198017679, + "language_loss": 0.7335068, + "learning_rate": 1.0191744301452853e-06, + "loss": 0.7500385, + "num_input_tokens_seen": 120465570, + "step": 5599, + "time_per_iteration": 3.1788973808288574 + }, + { + "auxiliary_loss_clip": 0.01165114, + "auxiliary_loss_mlp": 0.01028632, + "balance_loss_clip": 1.04777598, + "balance_loss_mlp": 1.02130067, + "epoch": 0.6733601875789094, + "flos": 25880173729920.0, + "grad_norm": 1.6799170798338128, + "language_loss": 0.70061231, + "learning_rate": 1.0184956383421208e-06, + "loss": 0.72254974, + "num_input_tokens_seen": 120484220, + "step": 5600, + "time_per_iteration": 2.6346077919006348 + }, + { + "auxiliary_loss_clip": 0.01151742, + "auxiliary_loss_mlp": 0.01026734, + "balance_loss_clip": 1.01521683, + "balance_loss_mlp": 1.01941466, + "epoch": 0.6734804304695485, + "flos": 22929573997440.0, + "grad_norm": 1.9619601504257194, + "language_loss": 0.65541846, + "learning_rate": 1.017816995432387e-06, + "loss": 0.67720318, + "num_input_tokens_seen": 120503320, + "step": 5601, + "time_per_iteration": 3.530447483062744 + }, + { + "auxiliary_loss_clip": 0.01121202, + "auxiliary_loss_mlp": 0.01026861, + "balance_loss_clip": 0.97364944, + "balance_loss_mlp": 1.0197506, + "epoch": 0.6736006733601876, + "flos": 18697968552960.0, + "grad_norm": 1.9305634286370623, + "language_loss": 0.74442971, + "learning_rate": 1.0171385015190353e-06, + "loss": 0.76591027, + "num_input_tokens_seen": 120523180, + "step": 5602, + "time_per_iteration": 3.60040283203125 + }, + { + "auxiliary_loss_clip": 0.01120492, + "auxiliary_loss_mlp": 0.00509282, + "balance_loss_clip": 0.97837174, + "balance_loss_mlp": 0.0, + "epoch": 0.6737209162508266, + "flos": 19427745173760.0, + "grad_norm": 1.9139176308947314, + "language_loss": 0.73335826, + "learning_rate": 1.0164601567049908e-06, + "loss": 0.74965596, + "num_input_tokens_seen": 120541710, + "step": 5603, + "time_per_iteration": 2.6711602210998535 + }, + { + "auxiliary_loss_clip": 0.0112265, + "auxiliary_loss_mlp": 0.01029259, + "balance_loss_clip": 0.97409129, + "balance_loss_mlp": 1.02159929, + "epoch": 0.6738411591414658, + "flos": 20158060498560.0, + "grad_norm": 1.9592587660269745, + "language_loss": 0.80232608, + "learning_rate": 1.015781961093158e-06, + "loss": 0.82384515, + "num_input_tokens_seen": 120561030, + "step": 5604, + "time_per_iteration": 3.60835862159729 + }, + { + "auxiliary_loss_clip": 0.01126834, + "auxiliary_loss_mlp": 0.01024941, + "balance_loss_clip": 0.97205675, + "balance_loss_mlp": 1.01770508, + "epoch": 0.6739614020321049, + "flos": 21653847584640.0, + "grad_norm": 1.637933729850825, + "language_loss": 0.77044284, + "learning_rate": 1.0151039147864197e-06, + "loss": 0.79196066, + "num_input_tokens_seen": 120581005, + "step": 5605, + "time_per_iteration": 2.669081926345825 + }, + { + "auxiliary_loss_clip": 0.01024218, + "auxiliary_loss_mlp": 0.01029958, + "balance_loss_clip": 0.82439053, + "balance_loss_mlp": 1.02200723, + "epoch": 0.6740816449227439, + "flos": 19171702051200.0, + "grad_norm": 2.7481525130905844, + "language_loss": 0.66196728, + "learning_rate": 1.0144260178876336e-06, + "loss": 0.68250906, + "num_input_tokens_seen": 120600350, + "step": 5606, + "time_per_iteration": 2.8312973976135254 + }, + { + "auxiliary_loss_clip": 0.01130377, + "auxiliary_loss_mlp": 0.01029726, + "balance_loss_clip": 0.97479308, + "balance_loss_mlp": 1.02188766, + "epoch": 0.6742018878133831, + "flos": 21097015971840.0, + "grad_norm": 2.1920708000801157, + "language_loss": 0.67172587, + "learning_rate": 1.0137482704996388e-06, + "loss": 0.69332683, + "num_input_tokens_seen": 120614700, + "step": 5607, + "time_per_iteration": 2.7686634063720703 + }, + { + "auxiliary_loss_clip": 0.01104926, + "auxiliary_loss_mlp": 0.01025777, + "balance_loss_clip": 0.93839097, + "balance_loss_mlp": 1.01829028, + "epoch": 0.6743221307040221, + "flos": 23549966726400.0, + "grad_norm": 1.836936161820758, + "language_loss": 0.79114354, + "learning_rate": 1.0130706727252461e-06, + "loss": 0.81245053, + "num_input_tokens_seen": 120631755, + "step": 5608, + "time_per_iteration": 2.7651147842407227 + }, + { + "auxiliary_loss_clip": 0.01105071, + "auxiliary_loss_mlp": 0.0102643, + "balance_loss_clip": 0.93634057, + "balance_loss_mlp": 1.01818061, + "epoch": 0.6744423735946612, + "flos": 16249542912000.0, + "grad_norm": 2.4116218691277176, + "language_loss": 0.68493104, + "learning_rate": 1.0123932246672468e-06, + "loss": 0.70624608, + "num_input_tokens_seen": 120645900, + "step": 5609, + "time_per_iteration": 2.7259361743927 + }, + { + "auxiliary_loss_clip": 0.00983546, + "auxiliary_loss_mlp": 0.00501209, + "balance_loss_clip": 0.86414695, + "balance_loss_mlp": 0.0, + "epoch": 0.6745626164853004, + "flos": 57843257829120.0, + "grad_norm": 0.749925610730636, + "language_loss": 0.5582726, + "learning_rate": 1.0117159264284114e-06, + "loss": 0.57312012, + "num_input_tokens_seen": 120709070, + "step": 5610, + "time_per_iteration": 3.2618279457092285 + }, + { + "auxiliary_loss_clip": 0.01126028, + "auxiliary_loss_mlp": 0.0102733, + "balance_loss_clip": 0.97493637, + "balance_loss_mlp": 1.01965332, + "epoch": 0.6746828593759394, + "flos": 20485027025280.0, + "grad_norm": 1.7104573534513576, + "language_loss": 0.77331436, + "learning_rate": 1.0110387781114837e-06, + "loss": 0.79484791, + "num_input_tokens_seen": 120727685, + "step": 5611, + "time_per_iteration": 2.7474679946899414 + }, + { + "auxiliary_loss_clip": 0.01168543, + "auxiliary_loss_mlp": 0.0103133, + "balance_loss_clip": 1.04964519, + "balance_loss_mlp": 1.02434468, + "epoch": 0.6748031022665785, + "flos": 19208223204480.0, + "grad_norm": 2.1080221899328198, + "language_loss": 0.77494144, + "learning_rate": 1.0103617798191872e-06, + "loss": 0.79694021, + "num_input_tokens_seen": 120747160, + "step": 5612, + "time_per_iteration": 2.750002384185791 + }, + { + "auxiliary_loss_clip": 0.01122723, + "auxiliary_loss_mlp": 0.01023861, + "balance_loss_clip": 0.97812867, + "balance_loss_mlp": 1.01570749, + "epoch": 0.6749233451572175, + "flos": 15195026407680.0, + "grad_norm": 10.477066370625066, + "language_loss": 0.83704823, + "learning_rate": 1.0096849316542217e-06, + "loss": 0.85851401, + "num_input_tokens_seen": 120763710, + "step": 5613, + "time_per_iteration": 2.7182111740112305 + }, + { + "auxiliary_loss_clip": 0.01014405, + "auxiliary_loss_mlp": 0.01028797, + "balance_loss_clip": 0.81624842, + "balance_loss_mlp": 1.02156138, + "epoch": 0.6750435880478567, + "flos": 26499489050880.0, + "grad_norm": 2.0284039152632767, + "language_loss": 0.74414366, + "learning_rate": 1.0090082337192643e-06, + "loss": 0.76457572, + "num_input_tokens_seen": 120783355, + "step": 5614, + "time_per_iteration": 3.767937183380127 + }, + { + "auxiliary_loss_clip": 0.01045793, + "auxiliary_loss_mlp": 0.01024884, + "balance_loss_clip": 0.85264331, + "balance_loss_mlp": 1.01720142, + "epoch": 0.6751638309384957, + "flos": 23404313076480.0, + "grad_norm": 2.0098650344835223, + "language_loss": 0.78411484, + "learning_rate": 1.0083316861169705e-06, + "loss": 0.80482167, + "num_input_tokens_seen": 120802090, + "step": 5615, + "time_per_iteration": 2.7925822734832764 + }, + { + "auxiliary_loss_clip": 0.01104774, + "auxiliary_loss_mlp": 0.01026514, + "balance_loss_clip": 0.93407631, + "balance_loss_mlp": 1.01824117, + "epoch": 0.6752840738291348, + "flos": 23441408847360.0, + "grad_norm": 2.4840950248112277, + "language_loss": 0.71956164, + "learning_rate": 1.0076552889499713e-06, + "loss": 0.74087453, + "num_input_tokens_seen": 120822855, + "step": 5616, + "time_per_iteration": 2.7856476306915283 + }, + { + "auxiliary_loss_clip": 0.01147864, + "auxiliary_loss_mlp": 0.0102819, + "balance_loss_clip": 1.01640725, + "balance_loss_mlp": 1.02113891, + "epoch": 0.675404316719774, + "flos": 30335826257280.0, + "grad_norm": 1.981982707706752, + "language_loss": 0.73648596, + "learning_rate": 1.006979042320876e-06, + "loss": 0.75824654, + "num_input_tokens_seen": 120843070, + "step": 5617, + "time_per_iteration": 2.7382049560546875 + }, + { + "auxiliary_loss_clip": 0.01118598, + "auxiliary_loss_mlp": 0.01031802, + "balance_loss_clip": 0.97156066, + "balance_loss_mlp": 1.02421427, + "epoch": 0.675524559610413, + "flos": 23622613983360.0, + "grad_norm": 2.123977143645754, + "language_loss": 0.63017058, + "learning_rate": 1.0063029463322702e-06, + "loss": 0.65167451, + "num_input_tokens_seen": 120863345, + "step": 5618, + "time_per_iteration": 2.919607400894165 + }, + { + "auxiliary_loss_clip": 0.01071422, + "auxiliary_loss_mlp": 0.00509951, + "balance_loss_clip": 0.89552188, + "balance_loss_mlp": 0.0, + "epoch": 0.6756448025010521, + "flos": 21248631279360.0, + "grad_norm": 2.628362028307301, + "language_loss": 0.75546348, + "learning_rate": 1.0056270010867164e-06, + "loss": 0.77127719, + "num_input_tokens_seen": 120880915, + "step": 5619, + "time_per_iteration": 2.7856485843658447 + }, + { + "auxiliary_loss_clip": 0.0112449, + "auxiliary_loss_mlp": 0.01030799, + "balance_loss_clip": 0.97069639, + "balance_loss_mlp": 1.02285397, + "epoch": 0.6757650453916912, + "flos": 21646521210240.0, + "grad_norm": 3.221277432903523, + "language_loss": 0.7840879, + "learning_rate": 1.004951206686758e-06, + "loss": 0.80564082, + "num_input_tokens_seen": 120899190, + "step": 5620, + "time_per_iteration": 2.7094223499298096 + }, + { + "auxiliary_loss_clip": 0.0113812, + "auxiliary_loss_mlp": 0.01029547, + "balance_loss_clip": 1.00877738, + "balance_loss_mlp": 1.02153039, + "epoch": 0.6758852882823303, + "flos": 21795658479360.0, + "grad_norm": 1.8931550120615122, + "language_loss": 0.720649, + "learning_rate": 1.0042755632349087e-06, + "loss": 0.74232566, + "num_input_tokens_seen": 120916080, + "step": 5621, + "time_per_iteration": 2.664651870727539 + }, + { + "auxiliary_loss_clip": 0.01098427, + "auxiliary_loss_mlp": 0.01031883, + "balance_loss_clip": 0.93526489, + "balance_loss_mlp": 1.02390146, + "epoch": 0.6760055311729694, + "flos": 27088783580160.0, + "grad_norm": 2.14660389050846, + "language_loss": 0.63383871, + "learning_rate": 1.0036000708336653e-06, + "loss": 0.65514183, + "num_input_tokens_seen": 120935210, + "step": 5622, + "time_per_iteration": 2.7505271434783936 + }, + { + "auxiliary_loss_clip": 0.01127655, + "auxiliary_loss_mlp": 0.01027055, + "balance_loss_clip": 0.9760437, + "balance_loss_mlp": 1.0197649, + "epoch": 0.6761257740636085, + "flos": 17999792922240.0, + "grad_norm": 3.304946972423049, + "language_loss": 0.79330903, + "learning_rate": 1.0029247295854984e-06, + "loss": 0.81485617, + "num_input_tokens_seen": 120951830, + "step": 5623, + "time_per_iteration": 2.675652503967285 + }, + { + "auxiliary_loss_clip": 0.01103684, + "auxiliary_loss_mlp": 0.01031116, + "balance_loss_clip": 0.93693346, + "balance_loss_mlp": 1.02433252, + "epoch": 0.6762460169542476, + "flos": 15121912273920.0, + "grad_norm": 1.893264105130326, + "language_loss": 0.71953255, + "learning_rate": 1.0022495395928588e-06, + "loss": 0.74088061, + "num_input_tokens_seen": 120970310, + "step": 5624, + "time_per_iteration": 2.691714286804199 + }, + { + "auxiliary_loss_clip": 0.01064898, + "auxiliary_loss_mlp": 0.01000234, + "balance_loss_clip": 1.01241469, + "balance_loss_mlp": 0.99919146, + "epoch": 0.6763662598448866, + "flos": 67886970030720.0, + "grad_norm": 0.7911074859345278, + "language_loss": 0.62391722, + "learning_rate": 1.0015745009581697e-06, + "loss": 0.64456856, + "num_input_tokens_seen": 121031915, + "step": 5625, + "time_per_iteration": 3.2536211013793945 + }, + { + "auxiliary_loss_clip": 0.01146456, + "auxiliary_loss_mlp": 0.0102759, + "balance_loss_clip": 1.01443231, + "balance_loss_mlp": 1.0199846, + "epoch": 0.6764865027355258, + "flos": 20631829910400.0, + "grad_norm": 1.773363569617356, + "language_loss": 0.6699701, + "learning_rate": 1.0008996137838343e-06, + "loss": 0.69171059, + "num_input_tokens_seen": 121050890, + "step": 5626, + "time_per_iteration": 2.6359074115753174 + }, + { + "auxiliary_loss_clip": 0.0117411, + "auxiliary_loss_mlp": 0.01031263, + "balance_loss_clip": 1.05106473, + "balance_loss_mlp": 1.02347875, + "epoch": 0.6766067456261649, + "flos": 21215809226880.0, + "grad_norm": 1.974890442523256, + "language_loss": 0.80161887, + "learning_rate": 1.000224878172234e-06, + "loss": 0.82367259, + "num_input_tokens_seen": 121070015, + "step": 5627, + "time_per_iteration": 3.5124452114105225 + }, + { + "auxiliary_loss_clip": 0.01151444, + "auxiliary_loss_mlp": 0.0102526, + "balance_loss_clip": 1.01475811, + "balance_loss_mlp": 1.01805949, + "epoch": 0.6767269885168039, + "flos": 19938251220480.0, + "grad_norm": 2.5935802759377475, + "language_loss": 0.72898984, + "learning_rate": 9.99550294225724e-07, + "loss": 0.75075686, + "num_input_tokens_seen": 121089170, + "step": 5628, + "time_per_iteration": 3.6082098484039307 + }, + { + "auxiliary_loss_clip": 0.01078645, + "auxiliary_loss_mlp": 0.01023109, + "balance_loss_clip": 0.89589071, + "balance_loss_mlp": 1.0157299, + "epoch": 0.6768472314074431, + "flos": 20814076540800.0, + "grad_norm": 1.944645624976119, + "language_loss": 0.72777689, + "learning_rate": 9.988758620466402e-07, + "loss": 0.7487945, + "num_input_tokens_seen": 121108040, + "step": 5629, + "time_per_iteration": 2.784719944000244 + }, + { + "auxiliary_loss_clip": 0.01060298, + "auxiliary_loss_mlp": 0.01021662, + "balance_loss_clip": 0.85904586, + "balance_loss_mlp": 1.01476538, + "epoch": 0.6769674742980821, + "flos": 23186012169600.0, + "grad_norm": 1.5466442336238362, + "language_loss": 0.76112539, + "learning_rate": 9.982015817372917e-07, + "loss": 0.78194499, + "num_input_tokens_seen": 121128480, + "step": 5630, + "time_per_iteration": 3.6825649738311768 + }, + { + "auxiliary_loss_clip": 0.01075889, + "auxiliary_loss_mlp": 0.01024441, + "balance_loss_clip": 0.89658803, + "balance_loss_mlp": 1.01687765, + "epoch": 0.6770877171887212, + "flos": 24242934885120.0, + "grad_norm": 2.9132164240894296, + "language_loss": 0.82148111, + "learning_rate": 9.975274533999657e-07, + "loss": 0.84248441, + "num_input_tokens_seen": 121148010, + "step": 5631, + "time_per_iteration": 2.7671103477478027 + }, + { + "auxiliary_loss_clip": 0.01169541, + "auxiliary_loss_mlp": 0.01027811, + "balance_loss_clip": 1.04968739, + "balance_loss_mlp": 1.01984179, + "epoch": 0.6772079600793603, + "flos": 18141567903360.0, + "grad_norm": 2.6745063988232345, + "language_loss": 0.84019113, + "learning_rate": 9.96853477136929e-07, + "loss": 0.86216468, + "num_input_tokens_seen": 121162755, + "step": 5632, + "time_per_iteration": 2.5890493392944336 + }, + { + "auxiliary_loss_clip": 0.01091127, + "auxiliary_loss_mlp": 0.01026841, + "balance_loss_clip": 0.93155116, + "balance_loss_mlp": 1.01895523, + "epoch": 0.6773282029699994, + "flos": 22452069571200.0, + "grad_norm": 2.2296366289382346, + "language_loss": 0.75343657, + "learning_rate": 9.96179653050422e-07, + "loss": 0.77461624, + "num_input_tokens_seen": 121182915, + "step": 5633, + "time_per_iteration": 2.6976521015167236 + }, + { + "auxiliary_loss_clip": 0.01091021, + "auxiliary_loss_mlp": 0.01029388, + "balance_loss_clip": 0.93311149, + "balance_loss_mlp": 1.02115119, + "epoch": 0.6774484458606385, + "flos": 18693730748160.0, + "grad_norm": 2.68813438997159, + "language_loss": 0.73718321, + "learning_rate": 9.955059812426635e-07, + "loss": 0.75838733, + "num_input_tokens_seen": 121200445, + "step": 5634, + "time_per_iteration": 2.663524627685547 + }, + { + "auxiliary_loss_clip": 0.01172321, + "auxiliary_loss_mlp": 0.01033104, + "balance_loss_clip": 1.05396152, + "balance_loss_mlp": 1.02523661, + "epoch": 0.6775686887512776, + "flos": 25994046821760.0, + "grad_norm": 1.9944256305717076, + "language_loss": 0.82828557, + "learning_rate": 9.948324618158493e-07, + "loss": 0.85033977, + "num_input_tokens_seen": 121220785, + "step": 5635, + "time_per_iteration": 2.7234017848968506 + }, + { + "auxiliary_loss_clip": 0.01145501, + "auxiliary_loss_mlp": 0.01027477, + "balance_loss_clip": 1.0083642, + "balance_loss_mlp": 1.01979399, + "epoch": 0.6776889316419167, + "flos": 13587987922560.0, + "grad_norm": 7.817873835742727, + "language_loss": 0.77752775, + "learning_rate": 9.941590948721502e-07, + "loss": 0.79925752, + "num_input_tokens_seen": 121237985, + "step": 5636, + "time_per_iteration": 2.576998472213745 + }, + { + "auxiliary_loss_clip": 0.01117023, + "auxiliary_loss_mlp": 0.01021169, + "balance_loss_clip": 0.97473437, + "balance_loss_mlp": 1.01407337, + "epoch": 0.6778091745325557, + "flos": 27601121220480.0, + "grad_norm": 1.5807096085139964, + "language_loss": 0.76447761, + "learning_rate": 9.934858805137188e-07, + "loss": 0.78585958, + "num_input_tokens_seen": 121258635, + "step": 5637, + "time_per_iteration": 2.7333438396453857 + }, + { + "auxiliary_loss_clip": 0.01142612, + "auxiliary_loss_mlp": 0.01027541, + "balance_loss_clip": 1.01289165, + "balance_loss_mlp": 1.02017951, + "epoch": 0.6779294174231949, + "flos": 18734058743040.0, + "grad_norm": 1.5802614975599716, + "language_loss": 0.81094348, + "learning_rate": 9.92812818842677e-07, + "loss": 0.832645, + "num_input_tokens_seen": 121277810, + "step": 5638, + "time_per_iteration": 2.614018440246582 + }, + { + "auxiliary_loss_clip": 0.01143954, + "auxiliary_loss_mlp": 0.01026052, + "balance_loss_clip": 1.01343179, + "balance_loss_mlp": 1.01813686, + "epoch": 0.678049660313834, + "flos": 45873797765760.0, + "grad_norm": 1.9526715120309317, + "language_loss": 0.64304942, + "learning_rate": 9.921399099611306e-07, + "loss": 0.6647495, + "num_input_tokens_seen": 121298975, + "step": 5639, + "time_per_iteration": 3.8490397930145264 + }, + { + "auxiliary_loss_clip": 0.01122755, + "auxiliary_loss_mlp": 0.01028132, + "balance_loss_clip": 0.97336113, + "balance_loss_mlp": 1.0214057, + "epoch": 0.678169903204473, + "flos": 19974556892160.0, + "grad_norm": 1.5010110529534986, + "language_loss": 0.69144404, + "learning_rate": 9.914671539711588e-07, + "loss": 0.71295291, + "num_input_tokens_seen": 121318495, + "step": 5640, + "time_per_iteration": 2.72786021232605 + }, + { + "auxiliary_loss_clip": 0.01008744, + "auxiliary_loss_mlp": 0.005097, + "balance_loss_clip": 0.78390574, + "balance_loss_mlp": 0.0, + "epoch": 0.6782901460951122, + "flos": 21395613732480.0, + "grad_norm": 3.006245663846035, + "language_loss": 0.78724998, + "learning_rate": 9.90794550974817e-07, + "loss": 0.80243444, + "num_input_tokens_seen": 121338890, + "step": 5641, + "time_per_iteration": 2.9913337230682373 + }, + { + "auxiliary_loss_clip": 0.01095976, + "auxiliary_loss_mlp": 0.0102398, + "balance_loss_clip": 0.93473864, + "balance_loss_mlp": 1.01703286, + "epoch": 0.6784103889857512, + "flos": 21434002392960.0, + "grad_norm": 1.9507334293223773, + "language_loss": 0.81216586, + "learning_rate": 9.901221010741407e-07, + "loss": 0.83336538, + "num_input_tokens_seen": 121358210, + "step": 5642, + "time_per_iteration": 3.0786194801330566 + }, + { + "auxiliary_loss_clip": 0.01148864, + "auxiliary_loss_mlp": 0.01024009, + "balance_loss_clip": 1.01148057, + "balance_loss_mlp": 1.01652217, + "epoch": 0.6785306318763903, + "flos": 32671923091200.0, + "grad_norm": 1.8975869955289697, + "language_loss": 0.74419665, + "learning_rate": 9.894498043711375e-07, + "loss": 0.76592541, + "num_input_tokens_seen": 121379955, + "step": 5643, + "time_per_iteration": 2.7527873516082764 + }, + { + "auxiliary_loss_clip": 0.01119499, + "auxiliary_loss_mlp": 0.01024721, + "balance_loss_clip": 0.97215277, + "balance_loss_mlp": 1.01707411, + "epoch": 0.6786508747670293, + "flos": 25632139340160.0, + "grad_norm": 2.1613830946140693, + "language_loss": 0.68961334, + "learning_rate": 9.887776609677962e-07, + "loss": 0.71105564, + "num_input_tokens_seen": 121401325, + "step": 5644, + "time_per_iteration": 2.7535061836242676 + }, + { + "auxiliary_loss_clip": 0.01085302, + "auxiliary_loss_mlp": 0.0102826, + "balance_loss_clip": 0.92868769, + "balance_loss_mlp": 1.02112567, + "epoch": 0.6787711176576685, + "flos": 19171881619200.0, + "grad_norm": 1.544213625295283, + "language_loss": 0.72542936, + "learning_rate": 9.88105670966079e-07, + "loss": 0.74656492, + "num_input_tokens_seen": 121419785, + "step": 5645, + "time_per_iteration": 2.71854567527771 + }, + { + "auxiliary_loss_clip": 0.01065646, + "auxiliary_loss_mlp": 0.01037063, + "balance_loss_clip": 0.89761615, + "balance_loss_mlp": 1.02926087, + "epoch": 0.6788913605483076, + "flos": 13985159581440.0, + "grad_norm": 3.0960854267225155, + "language_loss": 0.78934073, + "learning_rate": 9.874338344679283e-07, + "loss": 0.81036788, + "num_input_tokens_seen": 121435630, + "step": 5646, + "time_per_iteration": 2.704768657684326 + }, + { + "auxiliary_loss_clip": 0.01167469, + "auxiliary_loss_mlp": 0.0103063, + "balance_loss_clip": 1.05025244, + "balance_loss_mlp": 1.02317023, + "epoch": 0.6790116034389466, + "flos": 22017586659840.0, + "grad_norm": 1.5956955690053778, + "language_loss": 0.73757493, + "learning_rate": 9.86762151575259e-07, + "loss": 0.75955594, + "num_input_tokens_seen": 121455625, + "step": 5647, + "time_per_iteration": 2.6233386993408203 + }, + { + "auxiliary_loss_clip": 0.01075786, + "auxiliary_loss_mlp": 0.00508531, + "balance_loss_clip": 0.89655173, + "balance_loss_mlp": 0.0, + "epoch": 0.6791318463295858, + "flos": 20922454851840.0, + "grad_norm": 3.376511999083847, + "language_loss": 0.7987808, + "learning_rate": 9.860906223899651e-07, + "loss": 0.81462401, + "num_input_tokens_seen": 121475020, + "step": 5648, + "time_per_iteration": 2.774573802947998 + }, + { + "auxiliary_loss_clip": 0.01128231, + "auxiliary_loss_mlp": 0.01028551, + "balance_loss_clip": 0.97388196, + "balance_loss_mlp": 1.02112997, + "epoch": 0.6792520892202248, + "flos": 28512749422080.0, + "grad_norm": 1.6806675781572782, + "language_loss": 0.75767159, + "learning_rate": 9.854192470139184e-07, + "loss": 0.77923942, + "num_input_tokens_seen": 121496500, + "step": 5649, + "time_per_iteration": 2.768066167831421 + }, + { + "auxiliary_loss_clip": 0.01121735, + "auxiliary_loss_mlp": 0.01029469, + "balance_loss_clip": 0.97437191, + "balance_loss_mlp": 1.02235472, + "epoch": 0.6793723321108639, + "flos": 20011904058240.0, + "grad_norm": 2.786666673206875, + "language_loss": 0.71822166, + "learning_rate": 9.847480255489645e-07, + "loss": 0.7397337, + "num_input_tokens_seen": 121515525, + "step": 5650, + "time_per_iteration": 2.6588525772094727 + }, + { + "auxiliary_loss_clip": 0.01125718, + "auxiliary_loss_mlp": 0.01025952, + "balance_loss_clip": 0.97269738, + "balance_loss_mlp": 1.01928508, + "epoch": 0.6794925750015031, + "flos": 26649488246400.0, + "grad_norm": 2.2935433626737254, + "language_loss": 0.68898463, + "learning_rate": 9.840769580969295e-07, + "loss": 0.71050131, + "num_input_tokens_seen": 121535965, + "step": 5651, + "time_per_iteration": 2.800471067428589 + }, + { + "auxiliary_loss_clip": 0.01139922, + "auxiliary_loss_mlp": 0.01029382, + "balance_loss_clip": 1.01087379, + "balance_loss_mlp": 1.02079284, + "epoch": 0.6796128178921421, + "flos": 21580374314880.0, + "grad_norm": 1.8382236272730597, + "language_loss": 0.80055434, + "learning_rate": 9.834060447596114e-07, + "loss": 0.82224739, + "num_input_tokens_seen": 121555235, + "step": 5652, + "time_per_iteration": 2.6287264823913574 + }, + { + "auxiliary_loss_clip": 0.01146071, + "auxiliary_loss_mlp": 0.0102701, + "balance_loss_clip": 1.00924444, + "balance_loss_mlp": 1.01943421, + "epoch": 0.6797330607827812, + "flos": 22492002516480.0, + "grad_norm": 2.079558279558666, + "language_loss": 0.77968681, + "learning_rate": 9.827352856387868e-07, + "loss": 0.80141759, + "num_input_tokens_seen": 121574945, + "step": 5653, + "time_per_iteration": 3.964901924133301 + }, + { + "auxiliary_loss_clip": 0.00978102, + "auxiliary_loss_mlp": 0.01000615, + "balance_loss_clip": 0.8632586, + "balance_loss_mlp": 0.999304, + "epoch": 0.6798533036734203, + "flos": 66306648286080.0, + "grad_norm": 0.7729930646338838, + "language_loss": 0.64272797, + "learning_rate": 9.820646808362118e-07, + "loss": 0.66251516, + "num_input_tokens_seen": 121641200, + "step": 5654, + "time_per_iteration": 4.8028364181518555 + }, + { + "auxiliary_loss_clip": 0.01119659, + "auxiliary_loss_mlp": 0.0102811, + "balance_loss_clip": 0.97481966, + "balance_loss_mlp": 1.02052188, + "epoch": 0.6799735465640594, + "flos": 16180163792640.0, + "grad_norm": 2.3845236074386005, + "language_loss": 0.72857875, + "learning_rate": 9.813942304536154e-07, + "loss": 0.75005639, + "num_input_tokens_seen": 121659170, + "step": 5655, + "time_per_iteration": 2.6988749504089355 + }, + { + "auxiliary_loss_clip": 0.01121888, + "auxiliary_loss_mlp": 0.01026166, + "balance_loss_clip": 0.97304511, + "balance_loss_mlp": 1.01903701, + "epoch": 0.6800937894546984, + "flos": 22125749489280.0, + "grad_norm": 1.7622324662530031, + "language_loss": 0.64009106, + "learning_rate": 9.807239345927043e-07, + "loss": 0.66157162, + "num_input_tokens_seen": 121679180, + "step": 5656, + "time_per_iteration": 3.6059021949768066 + }, + { + "auxiliary_loss_clip": 0.01123484, + "auxiliary_loss_mlp": 0.01025295, + "balance_loss_clip": 0.97032064, + "balance_loss_mlp": 1.01781487, + "epoch": 0.6802140323453376, + "flos": 31612953300480.0, + "grad_norm": 2.9245423477552683, + "language_loss": 0.7245605, + "learning_rate": 9.80053793355162e-07, + "loss": 0.74604833, + "num_input_tokens_seen": 121697875, + "step": 5657, + "time_per_iteration": 2.698871374130249 + }, + { + "auxiliary_loss_clip": 0.01074385, + "auxiliary_loss_mlp": 0.01028628, + "balance_loss_clip": 0.89826721, + "balance_loss_mlp": 1.02079582, + "epoch": 0.6803342752359767, + "flos": 17712938908800.0, + "grad_norm": 2.1856743330387367, + "language_loss": 0.7515102, + "learning_rate": 9.793838068426472e-07, + "loss": 0.77254033, + "num_input_tokens_seen": 121715570, + "step": 5658, + "time_per_iteration": 2.720667600631714 + }, + { + "auxiliary_loss_clip": 0.01169459, + "auxiliary_loss_mlp": 0.01024879, + "balance_loss_clip": 1.05154002, + "balance_loss_mlp": 1.01754773, + "epoch": 0.6804545181266157, + "flos": 11326800902400.0, + "grad_norm": 2.0716022637558758, + "language_loss": 0.61091244, + "learning_rate": 9.78713975156799e-07, + "loss": 0.63285583, + "num_input_tokens_seen": 121731435, + "step": 5659, + "time_per_iteration": 2.5905277729034424 + }, + { + "auxiliary_loss_clip": 0.01104399, + "auxiliary_loss_mlp": 0.01029147, + "balance_loss_clip": 0.94050348, + "balance_loss_mlp": 1.02118969, + "epoch": 0.6805747610172549, + "flos": 29350976181120.0, + "grad_norm": 1.8830278445490989, + "language_loss": 0.71653247, + "learning_rate": 9.780442983992273e-07, + "loss": 0.73786789, + "num_input_tokens_seen": 121749950, + "step": 5660, + "time_per_iteration": 2.7526638507843018 + }, + { + "auxiliary_loss_clip": 0.01118071, + "auxiliary_loss_mlp": 0.01028329, + "balance_loss_clip": 0.97328889, + "balance_loss_mlp": 1.02090192, + "epoch": 0.680695003907894, + "flos": 37631868612480.0, + "grad_norm": 1.9576210083149306, + "language_loss": 0.71561152, + "learning_rate": 9.773747766715238e-07, + "loss": 0.73707551, + "num_input_tokens_seen": 121770770, + "step": 5661, + "time_per_iteration": 2.816664934158325 + }, + { + "auxiliary_loss_clip": 0.01125231, + "auxiliary_loss_mlp": 0.01026004, + "balance_loss_clip": 0.97245431, + "balance_loss_mlp": 1.01858664, + "epoch": 0.680815246798533, + "flos": 22127365601280.0, + "grad_norm": 1.6008242072395134, + "language_loss": 0.80371141, + "learning_rate": 9.767054100752536e-07, + "loss": 0.82522374, + "num_input_tokens_seen": 121790720, + "step": 5662, + "time_per_iteration": 2.6814162731170654 + }, + { + "auxiliary_loss_clip": 0.01101404, + "auxiliary_loss_mlp": 0.01021184, + "balance_loss_clip": 0.93579596, + "balance_loss_mlp": 1.01390004, + "epoch": 0.6809354896891722, + "flos": 17201822330880.0, + "grad_norm": 2.0315254484347713, + "language_loss": 0.81473351, + "learning_rate": 9.760361987119584e-07, + "loss": 0.83595932, + "num_input_tokens_seen": 121808455, + "step": 5663, + "time_per_iteration": 2.7107789516448975 + }, + { + "auxiliary_loss_clip": 0.01121528, + "auxiliary_loss_mlp": 0.01025266, + "balance_loss_clip": 0.97432816, + "balance_loss_mlp": 1.01771426, + "epoch": 0.6810557325798112, + "flos": 12458166554880.0, + "grad_norm": 2.203826150890875, + "language_loss": 0.67852956, + "learning_rate": 9.753671426831592e-07, + "loss": 0.69999754, + "num_input_tokens_seen": 121824470, + "step": 5664, + "time_per_iteration": 2.755943775177002 + }, + { + "auxiliary_loss_clip": 0.01137174, + "auxiliary_loss_mlp": 0.01026963, + "balance_loss_clip": 1.00809932, + "balance_loss_mlp": 1.01875484, + "epoch": 0.6811759754704503, + "flos": 22156165330560.0, + "grad_norm": 1.7531829526334004, + "language_loss": 0.79284412, + "learning_rate": 9.746982420903483e-07, + "loss": 0.81448549, + "num_input_tokens_seen": 121842665, + "step": 5665, + "time_per_iteration": 2.6561789512634277 + }, + { + "auxiliary_loss_clip": 0.0114571, + "auxiliary_loss_mlp": 0.0102703, + "balance_loss_clip": 1.01535821, + "balance_loss_mlp": 1.0202024, + "epoch": 0.6812962183610894, + "flos": 17525377065600.0, + "grad_norm": 1.5718802689858924, + "language_loss": 0.74460685, + "learning_rate": 9.740294970349993e-07, + "loss": 0.76633418, + "num_input_tokens_seen": 121859080, + "step": 5666, + "time_per_iteration": 3.5596683025360107 + }, + { + "auxiliary_loss_clip": 0.01024626, + "auxiliary_loss_mlp": 0.01000483, + "balance_loss_clip": 0.93705952, + "balance_loss_mlp": 0.99942172, + "epoch": 0.6814164612517285, + "flos": 60274480855680.0, + "grad_norm": 0.8888112390602946, + "language_loss": 0.60952723, + "learning_rate": 9.733609076185594e-07, + "loss": 0.62977827, + "num_input_tokens_seen": 121915485, + "step": 5667, + "time_per_iteration": 3.127501964569092 + }, + { + "auxiliary_loss_clip": 0.01143324, + "auxiliary_loss_mlp": 0.01034934, + "balance_loss_clip": 1.01092029, + "balance_loss_mlp": 1.02700925, + "epoch": 0.6815367041423676, + "flos": 19317750750720.0, + "grad_norm": 1.7620696723689615, + "language_loss": 0.83715129, + "learning_rate": 9.72692473942455e-07, + "loss": 0.85893381, + "num_input_tokens_seen": 121932710, + "step": 5668, + "time_per_iteration": 2.606125831604004 + }, + { + "auxiliary_loss_clip": 0.01081173, + "auxiliary_loss_mlp": 0.01032131, + "balance_loss_clip": 0.90171272, + "balance_loss_mlp": 1.02425098, + "epoch": 0.6816569470330067, + "flos": 22161696024960.0, + "grad_norm": 1.4888335605228964, + "language_loss": 0.77759254, + "learning_rate": 9.720241961080849e-07, + "loss": 0.79872555, + "num_input_tokens_seen": 121952025, + "step": 5669, + "time_per_iteration": 2.773291826248169 + }, + { + "auxiliary_loss_clip": 0.01168362, + "auxiliary_loss_mlp": 0.01030898, + "balance_loss_clip": 1.04906976, + "balance_loss_mlp": 1.02374589, + "epoch": 0.6817771899236458, + "flos": 41463501137280.0, + "grad_norm": 2.2633788899237524, + "language_loss": 0.73033273, + "learning_rate": 9.713560742168259e-07, + "loss": 0.75232536, + "num_input_tokens_seen": 121974650, + "step": 5670, + "time_per_iteration": 2.798236131668091 + }, + { + "auxiliary_loss_clip": 0.01097821, + "auxiliary_loss_mlp": 0.01026841, + "balance_loss_clip": 0.93409991, + "balance_loss_mlp": 1.01971805, + "epoch": 0.6818974328142848, + "flos": 21106138026240.0, + "grad_norm": 1.984405787228015, + "language_loss": 0.71372658, + "learning_rate": 9.706881083700333e-07, + "loss": 0.73497319, + "num_input_tokens_seen": 121994335, + "step": 5671, + "time_per_iteration": 2.7691385746002197 + }, + { + "auxiliary_loss_clip": 0.01056762, + "auxiliary_loss_mlp": 0.0102869, + "balance_loss_clip": 0.86597174, + "balance_loss_mlp": 1.02125144, + "epoch": 0.682017675704924, + "flos": 20441897769600.0, + "grad_norm": 4.683822878865778, + "language_loss": 0.82143354, + "learning_rate": 9.700202986690357e-07, + "loss": 0.84228802, + "num_input_tokens_seen": 122012635, + "step": 5672, + "time_per_iteration": 2.793410062789917 + }, + { + "auxiliary_loss_clip": 0.01143569, + "auxiliary_loss_mlp": 0.00510084, + "balance_loss_clip": 1.01046824, + "balance_loss_mlp": 0.0, + "epoch": 0.682137918595563, + "flos": 20044438801920.0, + "grad_norm": 2.8554086506699865, + "language_loss": 0.66699457, + "learning_rate": 9.693526452151413e-07, + "loss": 0.68353111, + "num_input_tokens_seen": 122031685, + "step": 5673, + "time_per_iteration": 2.6764028072357178 + }, + { + "auxiliary_loss_clip": 0.01101827, + "auxiliary_loss_mlp": 0.01027101, + "balance_loss_clip": 0.93343794, + "balance_loss_mlp": 1.01934981, + "epoch": 0.6822581614862021, + "flos": 31684559063040.0, + "grad_norm": 1.577112184698362, + "language_loss": 0.75662583, + "learning_rate": 9.686851481096305e-07, + "loss": 0.77791512, + "num_input_tokens_seen": 122052995, + "step": 5674, + "time_per_iteration": 2.790314197540283 + }, + { + "auxiliary_loss_clip": 0.01050564, + "auxiliary_loss_mlp": 0.01033496, + "balance_loss_clip": 0.85784292, + "balance_loss_mlp": 1.025998, + "epoch": 0.6823784043768413, + "flos": 23477570864640.0, + "grad_norm": 1.962971381838952, + "language_loss": 0.72074175, + "learning_rate": 9.68017807453762e-07, + "loss": 0.74158227, + "num_input_tokens_seen": 122071740, + "step": 5675, + "time_per_iteration": 2.8049464225769043 + }, + { + "auxiliary_loss_clip": 0.01124238, + "auxiliary_loss_mlp": 0.00509123, + "balance_loss_clip": 0.97503209, + "balance_loss_mlp": 0.0, + "epoch": 0.6824986472674803, + "flos": 14137134024960.0, + "grad_norm": 1.7700827036005238, + "language_loss": 0.73603529, + "learning_rate": 9.673506233487721e-07, + "loss": 0.75236893, + "num_input_tokens_seen": 122089705, + "step": 5676, + "time_per_iteration": 2.6953842639923096 + }, + { + "auxiliary_loss_clip": 0.0112339, + "auxiliary_loss_mlp": 0.00508668, + "balance_loss_clip": 0.97265166, + "balance_loss_mlp": 0.0, + "epoch": 0.6826188901581194, + "flos": 21504997624320.0, + "grad_norm": 1.8829680854938269, + "language_loss": 0.86080754, + "learning_rate": 9.666835958958717e-07, + "loss": 0.87712812, + "num_input_tokens_seen": 122109025, + "step": 5677, + "time_per_iteration": 2.681647539138794 + }, + { + "auxiliary_loss_clip": 0.01167646, + "auxiliary_loss_mlp": 0.01026008, + "balance_loss_clip": 1.05079257, + "balance_loss_mlp": 1.0181942, + "epoch": 0.6827391330487584, + "flos": 20810126044800.0, + "grad_norm": 1.8961117900952043, + "language_loss": 0.80559039, + "learning_rate": 9.660167251962484e-07, + "loss": 0.82752693, + "num_input_tokens_seen": 122127385, + "step": 5678, + "time_per_iteration": 2.6776833534240723 + }, + { + "auxiliary_loss_clip": 0.01100311, + "auxiliary_loss_mlp": 0.01034158, + "balance_loss_clip": 0.93308526, + "balance_loss_mlp": 1.02685654, + "epoch": 0.6828593759393976, + "flos": 21688788539520.0, + "grad_norm": 1.6872511496834282, + "language_loss": 0.77703011, + "learning_rate": 9.653500113510654e-07, + "loss": 0.79837483, + "num_input_tokens_seen": 122146500, + "step": 5679, + "time_per_iteration": 4.472065687179565 + }, + { + "auxiliary_loss_clip": 0.01118673, + "auxiliary_loss_mlp": 0.01033416, + "balance_loss_clip": 0.97363442, + "balance_loss_mlp": 1.0258882, + "epoch": 0.6829796188300367, + "flos": 25337707557120.0, + "grad_norm": 2.3436034554878833, + "language_loss": 0.67140579, + "learning_rate": 9.646834544614627e-07, + "loss": 0.69292665, + "num_input_tokens_seen": 122167000, + "step": 5680, + "time_per_iteration": 2.794544219970703 + }, + { + "auxiliary_loss_clip": 0.01117109, + "auxiliary_loss_mlp": 0.01027106, + "balance_loss_clip": 0.97534728, + "balance_loss_mlp": 1.0195775, + "epoch": 0.6830998617206757, + "flos": 20704800389760.0, + "grad_norm": 1.9260300728106368, + "language_loss": 0.76475203, + "learning_rate": 9.64017054628558e-07, + "loss": 0.78619421, + "num_input_tokens_seen": 122185825, + "step": 5681, + "time_per_iteration": 2.7403388023376465 + }, + { + "auxiliary_loss_clip": 0.01076109, + "auxiliary_loss_mlp": 0.01027959, + "balance_loss_clip": 0.89346373, + "balance_loss_mlp": 1.02012086, + "epoch": 0.6832201046113149, + "flos": 21726638496000.0, + "grad_norm": 1.649513045563631, + "language_loss": 0.78777015, + "learning_rate": 9.63350811953441e-07, + "loss": 0.80881083, + "num_input_tokens_seen": 122206200, + "step": 5682, + "time_per_iteration": 3.7756011486053467 + }, + { + "auxiliary_loss_clip": 0.01099034, + "auxiliary_loss_mlp": 0.0102538, + "balance_loss_clip": 0.93425006, + "balance_loss_mlp": 1.01822722, + "epoch": 0.6833403475019539, + "flos": 19536554448000.0, + "grad_norm": 1.9186165042881924, + "language_loss": 0.7040478, + "learning_rate": 9.626847265371826e-07, + "loss": 0.72529197, + "num_input_tokens_seen": 122225520, + "step": 5683, + "time_per_iteration": 2.746662139892578 + }, + { + "auxiliary_loss_clip": 0.01114493, + "auxiliary_loss_mlp": 0.01025872, + "balance_loss_clip": 0.97095454, + "balance_loss_mlp": 1.01823068, + "epoch": 0.683460590392593, + "flos": 19352153001600.0, + "grad_norm": 5.077134791140002, + "language_loss": 0.78526169, + "learning_rate": 9.620187984808262e-07, + "loss": 0.8066653, + "num_input_tokens_seen": 122244320, + "step": 5684, + "time_per_iteration": 2.676926374435425 + }, + { + "auxiliary_loss_clip": 0.01123359, + "auxiliary_loss_mlp": 0.00509533, + "balance_loss_clip": 0.97515541, + "balance_loss_mlp": 0.0, + "epoch": 0.6835808332832322, + "flos": 23288500650240.0, + "grad_norm": 6.452025824830686, + "language_loss": 0.86050683, + "learning_rate": 9.613530278853919e-07, + "loss": 0.87683576, + "num_input_tokens_seen": 122264295, + "step": 5685, + "time_per_iteration": 2.7489705085754395 + }, + { + "auxiliary_loss_clip": 0.01146059, + "auxiliary_loss_mlp": 0.01027571, + "balance_loss_clip": 1.01503325, + "balance_loss_mlp": 1.01975977, + "epoch": 0.6837010761738712, + "flos": 21653416621440.0, + "grad_norm": 1.8997165200265247, + "language_loss": 0.74494612, + "learning_rate": 9.60687414851879e-07, + "loss": 0.76668239, + "num_input_tokens_seen": 122285300, + "step": 5686, + "time_per_iteration": 2.6403908729553223 + }, + { + "auxiliary_loss_clip": 0.01124908, + "auxiliary_loss_mlp": 0.01027163, + "balance_loss_clip": 0.97444713, + "balance_loss_mlp": 1.01950347, + "epoch": 0.6838213190645103, + "flos": 17566387418880.0, + "grad_norm": 2.1824532093323317, + "language_loss": 0.77896392, + "learning_rate": 9.600219594812575e-07, + "loss": 0.80048466, + "num_input_tokens_seen": 122303240, + "step": 5687, + "time_per_iteration": 2.683932304382324 + }, + { + "auxiliary_loss_clip": 0.01166589, + "auxiliary_loss_mlp": 0.01021906, + "balance_loss_clip": 1.0498507, + "balance_loss_mlp": 1.01500702, + "epoch": 0.6839415619551494, + "flos": 23112538899840.0, + "grad_norm": 1.7967020243297929, + "language_loss": 0.72792065, + "learning_rate": 9.593566618744786e-07, + "loss": 0.74980557, + "num_input_tokens_seen": 122323390, + "step": 5688, + "time_per_iteration": 2.6517045497894287 + }, + { + "auxiliary_loss_clip": 0.01166493, + "auxiliary_loss_mlp": 0.01028688, + "balance_loss_clip": 1.04865432, + "balance_loss_mlp": 1.02130842, + "epoch": 0.6840618048457885, + "flos": 22127868391680.0, + "grad_norm": 3.307120985962542, + "language_loss": 0.73906386, + "learning_rate": 9.58691522132466e-07, + "loss": 0.76101565, + "num_input_tokens_seen": 122342200, + "step": 5689, + "time_per_iteration": 2.6005074977874756 + }, + { + "auxiliary_loss_clip": 0.01125974, + "auxiliary_loss_mlp": 0.01027085, + "balance_loss_clip": 0.9747231, + "balance_loss_mlp": 1.0190618, + "epoch": 0.6841820477364275, + "flos": 22015898720640.0, + "grad_norm": 2.111378094175794, + "language_loss": 0.85059476, + "learning_rate": 9.58026540356123e-07, + "loss": 0.87212533, + "num_input_tokens_seen": 122360465, + "step": 5690, + "time_per_iteration": 2.727102518081665 + }, + { + "auxiliary_loss_clip": 0.0114694, + "auxiliary_loss_mlp": 0.01025408, + "balance_loss_clip": 1.01078594, + "balance_loss_mlp": 1.01775753, + "epoch": 0.6843022906270667, + "flos": 24900531125760.0, + "grad_norm": 1.6987259529368002, + "language_loss": 0.86582303, + "learning_rate": 9.573617166463246e-07, + "loss": 0.88754654, + "num_input_tokens_seen": 122381680, + "step": 5691, + "time_per_iteration": 3.5893290042877197 + }, + { + "auxiliary_loss_clip": 0.01122084, + "auxiliary_loss_mlp": 0.0102542, + "balance_loss_clip": 0.97150254, + "balance_loss_mlp": 1.01818681, + "epoch": 0.6844225335177058, + "flos": 19969924037760.0, + "grad_norm": 1.7835140636525537, + "language_loss": 0.60147464, + "learning_rate": 9.56697051103924e-07, + "loss": 0.62294966, + "num_input_tokens_seen": 122399120, + "step": 5692, + "time_per_iteration": 2.7373461723327637 + }, + { + "auxiliary_loss_clip": 0.01121653, + "auxiliary_loss_mlp": 0.01025254, + "balance_loss_clip": 0.9740814, + "balance_loss_mlp": 1.01835203, + "epoch": 0.6845427764083448, + "flos": 25883334126720.0, + "grad_norm": 1.9187881810939798, + "language_loss": 0.80863261, + "learning_rate": 9.560325438297522e-07, + "loss": 0.83010161, + "num_input_tokens_seen": 122417430, + "step": 5693, + "time_per_iteration": 2.6486756801605225 + }, + { + "auxiliary_loss_clip": 0.01124633, + "auxiliary_loss_mlp": 0.01028805, + "balance_loss_clip": 0.97883677, + "balance_loss_mlp": 1.02145588, + "epoch": 0.684663019298984, + "flos": 18880143356160.0, + "grad_norm": 1.7553110788363284, + "language_loss": 0.86894107, + "learning_rate": 9.553681949246127e-07, + "loss": 0.89047551, + "num_input_tokens_seen": 122435055, + "step": 5694, + "time_per_iteration": 2.7339258193969727 + }, + { + "auxiliary_loss_clip": 0.0110424, + "auxiliary_loss_mlp": 0.0103105, + "balance_loss_clip": 0.93647599, + "balance_loss_mlp": 1.02362299, + "epoch": 0.684783262189623, + "flos": 54193725302400.0, + "grad_norm": 1.7584375104527858, + "language_loss": 0.7517305, + "learning_rate": 9.547040044892886e-07, + "loss": 0.77308339, + "num_input_tokens_seen": 122462570, + "step": 5695, + "time_per_iteration": 3.0207748413085938 + }, + { + "auxiliary_loss_clip": 0.01047507, + "auxiliary_loss_mlp": 0.00998337, + "balance_loss_clip": 0.9753207, + "balance_loss_mlp": 0.99724656, + "epoch": 0.6849035050802621, + "flos": 63970264143360.0, + "grad_norm": 0.8694126317717525, + "language_loss": 0.60148835, + "learning_rate": 9.540399726245354e-07, + "loss": 0.62194681, + "num_input_tokens_seen": 122519275, + "step": 5696, + "time_per_iteration": 3.056849241256714 + }, + { + "auxiliary_loss_clip": 0.01116834, + "auxiliary_loss_mlp": 0.01028313, + "balance_loss_clip": 0.97006923, + "balance_loss_mlp": 1.0205524, + "epoch": 0.6850237479709013, + "flos": 25224121774080.0, + "grad_norm": 1.8507592233722276, + "language_loss": 0.6885066, + "learning_rate": 9.533760994310859e-07, + "loss": 0.70995808, + "num_input_tokens_seen": 122539675, + "step": 5697, + "time_per_iteration": 2.7655081748962402 + }, + { + "auxiliary_loss_clip": 0.0116952, + "auxiliary_loss_mlp": 0.01030194, + "balance_loss_clip": 1.04968548, + "balance_loss_mlp": 1.02248073, + "epoch": 0.6851439908615403, + "flos": 19354128249600.0, + "grad_norm": 1.8916607872906297, + "language_loss": 0.74839908, + "learning_rate": 9.527123850096508e-07, + "loss": 0.77039623, + "num_input_tokens_seen": 122558035, + "step": 5698, + "time_per_iteration": 2.606245756149292 + }, + { + "auxiliary_loss_clip": 0.01149662, + "auxiliary_loss_mlp": 0.01018733, + "balance_loss_clip": 1.01152515, + "balance_loss_mlp": 1.0117774, + "epoch": 0.6852642337521794, + "flos": 23182133500800.0, + "grad_norm": 1.8972752439591607, + "language_loss": 0.71797091, + "learning_rate": 9.520488294609142e-07, + "loss": 0.7396549, + "num_input_tokens_seen": 122576815, + "step": 5699, + "time_per_iteration": 2.771625518798828 + }, + { + "auxiliary_loss_clip": 0.00985835, + "auxiliary_loss_mlp": 0.01003961, + "balance_loss_clip": 0.86517215, + "balance_loss_mlp": 1.00287008, + "epoch": 0.6853844766428185, + "flos": 62647206583680.0, + "grad_norm": 0.7330794106051937, + "language_loss": 0.53758752, + "learning_rate": 9.513854328855368e-07, + "loss": 0.55748552, + "num_input_tokens_seen": 122634690, + "step": 5700, + "time_per_iteration": 3.2548012733459473 + }, + { + "auxiliary_loss_clip": 0.01166642, + "auxiliary_loss_mlp": 0.01029278, + "balance_loss_clip": 1.04927611, + "balance_loss_mlp": 1.02211905, + "epoch": 0.6855047195334576, + "flos": 23437242869760.0, + "grad_norm": 1.8711768850798387, + "language_loss": 0.81169027, + "learning_rate": 9.507221953841558e-07, + "loss": 0.8336494, + "num_input_tokens_seen": 122652320, + "step": 5701, + "time_per_iteration": 2.650909185409546 + }, + { + "auxiliary_loss_clip": 0.01147252, + "auxiliary_loss_mlp": 0.01035034, + "balance_loss_clip": 1.01384318, + "balance_loss_mlp": 1.0276072, + "epoch": 0.6856249624240967, + "flos": 20664831530880.0, + "grad_norm": 1.6033751120335966, + "language_loss": 0.7812711, + "learning_rate": 9.500591170573824e-07, + "loss": 0.80309391, + "num_input_tokens_seen": 122672340, + "step": 5702, + "time_per_iteration": 2.6925346851348877 + }, + { + "auxiliary_loss_clip": 0.01069131, + "auxiliary_loss_mlp": 0.01025242, + "balance_loss_clip": 0.8953281, + "balance_loss_mlp": 1.01809251, + "epoch": 0.6857452053147358, + "flos": 17087302794240.0, + "grad_norm": 20.86483767283792, + "language_loss": 0.74359095, + "learning_rate": 9.493961980058078e-07, + "loss": 0.76453471, + "num_input_tokens_seen": 122689935, + "step": 5703, + "time_per_iteration": 2.6897366046905518 + }, + { + "auxiliary_loss_clip": 0.01037091, + "auxiliary_loss_mlp": 0.01022483, + "balance_loss_clip": 0.8557235, + "balance_loss_mlp": 1.01558423, + "epoch": 0.6858654482053749, + "flos": 30847266057600.0, + "grad_norm": 2.7942937140455464, + "language_loss": 0.67754382, + "learning_rate": 9.48733438329993e-07, + "loss": 0.69813955, + "num_input_tokens_seen": 122710200, + "step": 5704, + "time_per_iteration": 2.8664495944976807 + }, + { + "auxiliary_loss_clip": 0.01168773, + "auxiliary_loss_mlp": 0.00509087, + "balance_loss_clip": 1.05192983, + "balance_loss_mlp": 0.0, + "epoch": 0.6859856910960139, + "flos": 28877314510080.0, + "grad_norm": 1.6113459901369895, + "language_loss": 0.74456102, + "learning_rate": 9.480708381304807e-07, + "loss": 0.76133966, + "num_input_tokens_seen": 122731495, + "step": 5705, + "time_per_iteration": 3.668410539627075 + }, + { + "auxiliary_loss_clip": 0.0107481, + "auxiliary_loss_mlp": 0.01027223, + "balance_loss_clip": 0.90054816, + "balance_loss_mlp": 1.01975203, + "epoch": 0.6861059339866531, + "flos": 19354523299200.0, + "grad_norm": 3.691019539883786, + "language_loss": 0.83659732, + "learning_rate": 9.474083975077858e-07, + "loss": 0.85761762, + "num_input_tokens_seen": 122748620, + "step": 5706, + "time_per_iteration": 4.016341209411621 + }, + { + "auxiliary_loss_clip": 0.01137796, + "auxiliary_loss_mlp": 0.01023406, + "balance_loss_clip": 1.00827277, + "balance_loss_mlp": 1.01642036, + "epoch": 0.6862261768772921, + "flos": 22199976944640.0, + "grad_norm": 2.0505241991833314, + "language_loss": 0.79826832, + "learning_rate": 9.467461165623994e-07, + "loss": 0.81988037, + "num_input_tokens_seen": 122767670, + "step": 5707, + "time_per_iteration": 2.707911729812622 + }, + { + "auxiliary_loss_clip": 0.01147547, + "auxiliary_loss_mlp": 0.01027127, + "balance_loss_clip": 1.01096249, + "balance_loss_mlp": 1.01989114, + "epoch": 0.6863464197679312, + "flos": 26285677344000.0, + "grad_norm": 2.046792286294146, + "language_loss": 0.7986269, + "learning_rate": 9.46083995394791e-07, + "loss": 0.82037365, + "num_input_tokens_seen": 122785480, + "step": 5708, + "time_per_iteration": 3.692060708999634 + }, + { + "auxiliary_loss_clip": 0.01143434, + "auxiliary_loss_mlp": 0.00509424, + "balance_loss_clip": 1.01078987, + "balance_loss_mlp": 0.0, + "epoch": 0.6864666626585703, + "flos": 37815228564480.0, + "grad_norm": 2.1000582565379995, + "language_loss": 0.63148832, + "learning_rate": 9.454220341054012e-07, + "loss": 0.64801693, + "num_input_tokens_seen": 122810265, + "step": 5709, + "time_per_iteration": 2.821803569793701 + }, + { + "auxiliary_loss_clip": 0.01098319, + "auxiliary_loss_mlp": 0.01022629, + "balance_loss_clip": 0.93570262, + "balance_loss_mlp": 1.01531589, + "epoch": 0.6865869055492094, + "flos": 19391152193280.0, + "grad_norm": 2.4739759772087617, + "language_loss": 0.81013346, + "learning_rate": 9.447602327946512e-07, + "loss": 0.83134294, + "num_input_tokens_seen": 122828905, + "step": 5710, + "time_per_iteration": 2.719005584716797 + }, + { + "auxiliary_loss_clip": 0.01118662, + "auxiliary_loss_mlp": 0.01024247, + "balance_loss_clip": 0.96986836, + "balance_loss_mlp": 1.01655233, + "epoch": 0.6867071484398485, + "flos": 20375966355840.0, + "grad_norm": 1.8185790952293215, + "language_loss": 0.76571053, + "learning_rate": 9.440985915629338e-07, + "loss": 0.78713959, + "num_input_tokens_seen": 122846235, + "step": 5711, + "time_per_iteration": 2.67814564704895 + }, + { + "auxiliary_loss_clip": 0.01170931, + "auxiliary_loss_mlp": 0.01028214, + "balance_loss_clip": 1.05360961, + "balance_loss_mlp": 1.02107954, + "epoch": 0.6868273913304875, + "flos": 15889143801600.0, + "grad_norm": 2.001935656601048, + "language_loss": 0.73219568, + "learning_rate": 9.434371105106223e-07, + "loss": 0.75418711, + "num_input_tokens_seen": 122863835, + "step": 5712, + "time_per_iteration": 2.6392295360565186 + }, + { + "auxiliary_loss_clip": 0.01094189, + "auxiliary_loss_mlp": 0.0102456, + "balance_loss_clip": 0.93330753, + "balance_loss_mlp": 1.01718354, + "epoch": 0.6869476342211267, + "flos": 24462492768000.0, + "grad_norm": 2.554254338181186, + "language_loss": 0.70630544, + "learning_rate": 9.427757897380602e-07, + "loss": 0.72749293, + "num_input_tokens_seen": 122883235, + "step": 5713, + "time_per_iteration": 2.7788848876953125 + }, + { + "auxiliary_loss_clip": 0.01095408, + "auxiliary_loss_mlp": 0.0103069, + "balance_loss_clip": 0.93743068, + "balance_loss_mlp": 1.02224398, + "epoch": 0.6870678771117658, + "flos": 18442571875200.0, + "grad_norm": 2.010347109164211, + "language_loss": 0.85287046, + "learning_rate": 9.421146293455695e-07, + "loss": 0.87413138, + "num_input_tokens_seen": 122898975, + "step": 5714, + "time_per_iteration": 2.7116599082946777 + }, + { + "auxiliary_loss_clip": 0.01119487, + "auxiliary_loss_mlp": 0.01031763, + "balance_loss_clip": 0.97222394, + "balance_loss_mlp": 1.02475309, + "epoch": 0.6871881200024048, + "flos": 22200371994240.0, + "grad_norm": 1.8249228419773893, + "language_loss": 0.68537825, + "learning_rate": 9.414536294334489e-07, + "loss": 0.7068907, + "num_input_tokens_seen": 122918995, + "step": 5715, + "time_per_iteration": 2.760270357131958 + }, + { + "auxiliary_loss_clip": 0.01123124, + "auxiliary_loss_mlp": 0.01030682, + "balance_loss_clip": 0.97019887, + "balance_loss_mlp": 1.0231899, + "epoch": 0.687308362893044, + "flos": 22127724737280.0, + "grad_norm": 1.694088761191425, + "language_loss": 0.6949141, + "learning_rate": 9.407927901019708e-07, + "loss": 0.71645212, + "num_input_tokens_seen": 122938125, + "step": 5716, + "time_per_iteration": 2.7607040405273438 + }, + { + "auxiliary_loss_clip": 0.01144352, + "auxiliary_loss_mlp": 0.01030823, + "balance_loss_clip": 1.0107739, + "balance_loss_mlp": 1.02379608, + "epoch": 0.687428605783683, + "flos": 25040546340480.0, + "grad_norm": 2.0844659469223954, + "language_loss": 0.76682055, + "learning_rate": 9.401321114513854e-07, + "loss": 0.78857231, + "num_input_tokens_seen": 122957020, + "step": 5717, + "time_per_iteration": 3.5202245712280273 + }, + { + "auxiliary_loss_clip": 0.01170172, + "auxiliary_loss_mlp": 0.01032679, + "balance_loss_clip": 1.0513941, + "balance_loss_mlp": 1.02496314, + "epoch": 0.6875488486743221, + "flos": 23770063313280.0, + "grad_norm": 2.463424565957909, + "language_loss": 0.75080943, + "learning_rate": 9.394715935819155e-07, + "loss": 0.77283794, + "num_input_tokens_seen": 122977410, + "step": 5718, + "time_per_iteration": 2.658703565597534 + }, + { + "auxiliary_loss_clip": 0.01148511, + "auxiliary_loss_mlp": 0.01025219, + "balance_loss_clip": 1.01168323, + "balance_loss_mlp": 1.01748192, + "epoch": 0.6876690915649613, + "flos": 25516937445120.0, + "grad_norm": 1.8861775621871482, + "language_loss": 0.62147915, + "learning_rate": 9.388112365937608e-07, + "loss": 0.64321643, + "num_input_tokens_seen": 122996875, + "step": 5719, + "time_per_iteration": 2.6952574253082275 + }, + { + "auxiliary_loss_clip": 0.01099421, + "auxiliary_loss_mlp": 0.01032015, + "balance_loss_clip": 0.93625391, + "balance_loss_mlp": 1.02427804, + "epoch": 0.6877893344556003, + "flos": 19427996568960.0, + "grad_norm": 2.9351525062007906, + "language_loss": 0.82843792, + "learning_rate": 9.381510405870985e-07, + "loss": 0.84975231, + "num_input_tokens_seen": 123015890, + "step": 5720, + "time_per_iteration": 2.717484474182129 + }, + { + "auxiliary_loss_clip": 0.01141517, + "auxiliary_loss_mlp": 0.01026504, + "balance_loss_clip": 1.01012278, + "balance_loss_mlp": 1.01903522, + "epoch": 0.6879095773462394, + "flos": 18661303745280.0, + "grad_norm": 2.150962166138582, + "language_loss": 0.7743355, + "learning_rate": 9.374910056620791e-07, + "loss": 0.79601574, + "num_input_tokens_seen": 123034955, + "step": 5721, + "time_per_iteration": 2.6097054481506348 + }, + { + "auxiliary_loss_clip": 0.01148152, + "auxiliary_loss_mlp": 0.01027811, + "balance_loss_clip": 1.01408422, + "balance_loss_mlp": 1.02003837, + "epoch": 0.6880298202368785, + "flos": 20883132437760.0, + "grad_norm": 1.7632222362869536, + "language_loss": 0.80853033, + "learning_rate": 9.368311319188293e-07, + "loss": 0.83028996, + "num_input_tokens_seen": 123052770, + "step": 5722, + "time_per_iteration": 2.6362853050231934 + }, + { + "auxiliary_loss_clip": 0.01097634, + "auxiliary_loss_mlp": 0.0102618, + "balance_loss_clip": 0.93413258, + "balance_loss_mlp": 1.01862788, + "epoch": 0.6881500631275176, + "flos": 30153292318080.0, + "grad_norm": 1.8683965953567354, + "language_loss": 0.79637635, + "learning_rate": 9.361714194574515e-07, + "loss": 0.8176145, + "num_input_tokens_seen": 123075105, + "step": 5723, + "time_per_iteration": 2.8348429203033447 + }, + { + "auxiliary_loss_clip": 0.01065794, + "auxiliary_loss_mlp": 0.00999121, + "balance_loss_clip": 1.01258242, + "balance_loss_mlp": 0.99804795, + "epoch": 0.6882703060181566, + "flos": 66181537215360.0, + "grad_norm": 0.7317746006602213, + "language_loss": 0.58281785, + "learning_rate": 9.355118683780228e-07, + "loss": 0.60346705, + "num_input_tokens_seen": 123145175, + "step": 5724, + "time_per_iteration": 3.29288387298584 + }, + { + "auxiliary_loss_clip": 0.01166915, + "auxiliary_loss_mlp": 0.0102379, + "balance_loss_clip": 1.04877675, + "balance_loss_mlp": 1.01670563, + "epoch": 0.6883905489087958, + "flos": 18214646123520.0, + "grad_norm": 2.2012931229824653, + "language_loss": 0.78991514, + "learning_rate": 9.348524787805987e-07, + "loss": 0.81182224, + "num_input_tokens_seen": 123160365, + "step": 5725, + "time_per_iteration": 2.608470916748047 + }, + { + "auxiliary_loss_clip": 0.01098034, + "auxiliary_loss_mlp": 0.01023153, + "balance_loss_clip": 0.93081427, + "balance_loss_mlp": 1.01536286, + "epoch": 0.6885107917994349, + "flos": 14056262553600.0, + "grad_norm": 2.674022767167365, + "language_loss": 0.85827553, + "learning_rate": 9.341932507652053e-07, + "loss": 0.87948734, + "num_input_tokens_seen": 123174855, + "step": 5726, + "time_per_iteration": 2.79364013671875 + }, + { + "auxiliary_loss_clip": 0.0111798, + "auxiliary_loss_mlp": 0.01024335, + "balance_loss_clip": 0.96842223, + "balance_loss_mlp": 1.01656222, + "epoch": 0.6886310346900739, + "flos": 28690722334080.0, + "grad_norm": 2.2005510031811295, + "language_loss": 0.78656197, + "learning_rate": 9.335341844318489e-07, + "loss": 0.80798513, + "num_input_tokens_seen": 123194995, + "step": 5727, + "time_per_iteration": 2.8029837608337402 + }, + { + "auxiliary_loss_clip": 0.01119535, + "auxiliary_loss_mlp": 0.01030932, + "balance_loss_clip": 0.97169566, + "balance_loss_mlp": 1.02287316, + "epoch": 0.6887512775807131, + "flos": 24535319592960.0, + "grad_norm": 1.7581172418242363, + "language_loss": 0.7324751, + "learning_rate": 9.328752798805091e-07, + "loss": 0.7539798, + "num_input_tokens_seen": 123213465, + "step": 5728, + "time_per_iteration": 2.718127727508545 + }, + { + "auxiliary_loss_clip": 0.01146256, + "auxiliary_loss_mlp": 0.01028625, + "balance_loss_clip": 1.0126586, + "balance_loss_mlp": 1.02134752, + "epoch": 0.6888715204713521, + "flos": 22414363269120.0, + "grad_norm": 2.970049611928304, + "language_loss": 0.75989807, + "learning_rate": 9.322165372111399e-07, + "loss": 0.78164691, + "num_input_tokens_seen": 123231610, + "step": 5729, + "time_per_iteration": 2.6512413024902344 + }, + { + "auxiliary_loss_clip": 0.01095454, + "auxiliary_loss_mlp": 0.01030918, + "balance_loss_clip": 0.93694437, + "balance_loss_mlp": 1.02378309, + "epoch": 0.6889917633619912, + "flos": 22054323294720.0, + "grad_norm": 2.7240873124663487, + "language_loss": 0.75672019, + "learning_rate": 9.315579565236747e-07, + "loss": 0.7779839, + "num_input_tokens_seen": 123250715, + "step": 5730, + "time_per_iteration": 2.7185075283050537 + }, + { + "auxiliary_loss_clip": 0.0111779, + "auxiliary_loss_mlp": 0.0103015, + "balance_loss_clip": 0.9753648, + "balance_loss_mlp": 1.02210951, + "epoch": 0.6891120062526304, + "flos": 23949724164480.0, + "grad_norm": 1.7137374706899835, + "language_loss": 0.74310637, + "learning_rate": 9.308995379180162e-07, + "loss": 0.76458579, + "num_input_tokens_seen": 123270270, + "step": 5731, + "time_per_iteration": 2.6757984161376953 + }, + { + "auxiliary_loss_clip": 0.01047994, + "auxiliary_loss_mlp": 0.01003058, + "balance_loss_clip": 0.97577089, + "balance_loss_mlp": 1.00204492, + "epoch": 0.6892322491432694, + "flos": 64117354337280.0, + "grad_norm": 0.9648451505151585, + "language_loss": 0.59455168, + "learning_rate": 9.302412814940488e-07, + "loss": 0.61506212, + "num_input_tokens_seen": 123333045, + "step": 5732, + "time_per_iteration": 5.131500959396362 + }, + { + "auxiliary_loss_clip": 0.01120067, + "auxiliary_loss_mlp": 0.01027068, + "balance_loss_clip": 0.97147107, + "balance_loss_mlp": 1.01974893, + "epoch": 0.6893524920339085, + "flos": 23002436736000.0, + "grad_norm": 10.377710523122063, + "language_loss": 0.71101975, + "learning_rate": 9.295831873516276e-07, + "loss": 0.73249114, + "num_input_tokens_seen": 123352320, + "step": 5733, + "time_per_iteration": 2.7311453819274902 + }, + { + "auxiliary_loss_clip": 0.0117101, + "auxiliary_loss_mlp": 0.01028133, + "balance_loss_clip": 1.05301297, + "balance_loss_mlp": 1.02103424, + "epoch": 0.6894727349245476, + "flos": 21396260177280.0, + "grad_norm": 1.451776141415614, + "language_loss": 0.75911951, + "learning_rate": 9.289252555905873e-07, + "loss": 0.781111, + "num_input_tokens_seen": 123372400, + "step": 5734, + "time_per_iteration": 3.6075937747955322 + }, + { + "auxiliary_loss_clip": 0.01148868, + "auxiliary_loss_mlp": 0.0102726, + "balance_loss_clip": 1.01651883, + "balance_loss_mlp": 1.02005363, + "epoch": 0.6895929778151867, + "flos": 19865316654720.0, + "grad_norm": 2.1782285935087553, + "language_loss": 0.76019669, + "learning_rate": 9.282674863107334e-07, + "loss": 0.78195792, + "num_input_tokens_seen": 123390215, + "step": 5735, + "time_per_iteration": 2.6401450634002686 + }, + { + "auxiliary_loss_clip": 0.01143588, + "auxiliary_loss_mlp": 0.01031647, + "balance_loss_clip": 1.01331663, + "balance_loss_mlp": 1.02458382, + "epoch": 0.6897132207058257, + "flos": 18179166464640.0, + "grad_norm": 2.3249874998386613, + "language_loss": 0.76168036, + "learning_rate": 9.276098796118488e-07, + "loss": 0.78343272, + "num_input_tokens_seen": 123406700, + "step": 5736, + "time_per_iteration": 2.6548244953155518 + }, + { + "auxiliary_loss_clip": 0.01123418, + "auxiliary_loss_mlp": 0.01027896, + "balance_loss_clip": 0.9745512, + "balance_loss_mlp": 1.02040923, + "epoch": 0.6898334635964649, + "flos": 32561641359360.0, + "grad_norm": 2.2703455535108623, + "language_loss": 0.66419482, + "learning_rate": 9.269524355936938e-07, + "loss": 0.68570793, + "num_input_tokens_seen": 123429880, + "step": 5737, + "time_per_iteration": 2.7928833961486816 + }, + { + "auxiliary_loss_clip": 0.01114692, + "auxiliary_loss_mlp": 0.01019141, + "balance_loss_clip": 0.97050381, + "balance_loss_mlp": 1.01225114, + "epoch": 0.689953706487104, + "flos": 22819004956800.0, + "grad_norm": 2.3688173220553614, + "language_loss": 0.84848118, + "learning_rate": 9.262951543560002e-07, + "loss": 0.86981952, + "num_input_tokens_seen": 123449105, + "step": 5738, + "time_per_iteration": 2.7089028358459473 + }, + { + "auxiliary_loss_clip": 0.01123382, + "auxiliary_loss_mlp": 0.01024386, + "balance_loss_clip": 0.97818828, + "balance_loss_mlp": 1.01613021, + "epoch": 0.690073949377743, + "flos": 18515362786560.0, + "grad_norm": 2.32013019249825, + "language_loss": 0.86175179, + "learning_rate": 9.256380359984795e-07, + "loss": 0.88322943, + "num_input_tokens_seen": 123466215, + "step": 5739, + "time_per_iteration": 2.6187291145324707 + }, + { + "auxiliary_loss_clip": 0.0108009, + "auxiliary_loss_mlp": 0.01023568, + "balance_loss_clip": 0.89231658, + "balance_loss_mlp": 1.01606369, + "epoch": 0.6901941922683821, + "flos": 34857194716800.0, + "grad_norm": 1.8604294369023007, + "language_loss": 0.75141084, + "learning_rate": 9.249810806208139e-07, + "loss": 0.77244735, + "num_input_tokens_seen": 123485480, + "step": 5740, + "time_per_iteration": 2.865144729614258 + }, + { + "auxiliary_loss_clip": 0.01071435, + "auxiliary_loss_mlp": 0.00509006, + "balance_loss_clip": 0.89177686, + "balance_loss_mlp": 0.0, + "epoch": 0.6903144351590212, + "flos": 16253672976000.0, + "grad_norm": 2.2682076277491072, + "language_loss": 0.80874854, + "learning_rate": 9.243242883226627e-07, + "loss": 0.82455301, + "num_input_tokens_seen": 123504575, + "step": 5741, + "time_per_iteration": 2.7738335132598877 + }, + { + "auxiliary_loss_clip": 0.01145561, + "auxiliary_loss_mlp": 0.01027307, + "balance_loss_clip": 1.00639725, + "balance_loss_mlp": 1.01924801, + "epoch": 0.6904346780496603, + "flos": 28035137255040.0, + "grad_norm": 2.048358042283759, + "language_loss": 0.69383168, + "learning_rate": 9.236676592036628e-07, + "loss": 0.71556032, + "num_input_tokens_seen": 123524250, + "step": 5742, + "time_per_iteration": 2.6723737716674805 + }, + { + "auxiliary_loss_clip": 0.01119551, + "auxiliary_loss_mlp": 0.01021482, + "balance_loss_clip": 0.97707462, + "balance_loss_mlp": 1.01391852, + "epoch": 0.6905549209402994, + "flos": 23624266008960.0, + "grad_norm": 2.0147931695817647, + "language_loss": 0.73525715, + "learning_rate": 9.230111933634228e-07, + "loss": 0.75666749, + "num_input_tokens_seen": 123545845, + "step": 5743, + "time_per_iteration": 3.6579792499542236 + }, + { + "auxiliary_loss_clip": 0.01147674, + "auxiliary_loss_mlp": 0.0102868, + "balance_loss_clip": 1.01368225, + "balance_loss_mlp": 1.02161431, + "epoch": 0.6906751638309385, + "flos": 23114945111040.0, + "grad_norm": 1.502615733617126, + "language_loss": 0.80591071, + "learning_rate": 9.223548909015288e-07, + "loss": 0.82767427, + "num_input_tokens_seen": 123567535, + "step": 5744, + "time_per_iteration": 2.6553072929382324 + }, + { + "auxiliary_loss_clip": 0.01069177, + "auxiliary_loss_mlp": 0.01020761, + "balance_loss_clip": 0.89565802, + "balance_loss_mlp": 1.01363838, + "epoch": 0.6907954067215776, + "flos": 27305468375040.0, + "grad_norm": 1.9741381511696763, + "language_loss": 0.72414446, + "learning_rate": 9.216987519175407e-07, + "loss": 0.74504387, + "num_input_tokens_seen": 123587710, + "step": 5745, + "time_per_iteration": 2.7737271785736084 + }, + { + "auxiliary_loss_clip": 0.01142476, + "auxiliary_loss_mlp": 0.01024623, + "balance_loss_clip": 1.01395929, + "balance_loss_mlp": 1.01735771, + "epoch": 0.6909156496122166, + "flos": 21689399070720.0, + "grad_norm": 1.7433287398097337, + "language_loss": 0.68869817, + "learning_rate": 9.210427765109942e-07, + "loss": 0.71036917, + "num_input_tokens_seen": 123607385, + "step": 5746, + "time_per_iteration": 2.641893148422241 + }, + { + "auxiliary_loss_clip": 0.01122361, + "auxiliary_loss_mlp": 0.01026435, + "balance_loss_clip": 0.97206581, + "balance_loss_mlp": 1.01829338, + "epoch": 0.6910358925028558, + "flos": 22561453463040.0, + "grad_norm": 1.8893144226523273, + "language_loss": 0.81387877, + "learning_rate": 9.20386964781402e-07, + "loss": 0.83536673, + "num_input_tokens_seen": 123625405, + "step": 5747, + "time_per_iteration": 2.716653347015381 + }, + { + "auxiliary_loss_clip": 0.01118474, + "auxiliary_loss_mlp": 0.01026492, + "balance_loss_clip": 0.97268194, + "balance_loss_mlp": 1.01948881, + "epoch": 0.6911561353934949, + "flos": 22054107813120.0, + "grad_norm": 2.6510366886329537, + "language_loss": 0.84417737, + "learning_rate": 9.197313168282472e-07, + "loss": 0.86562705, + "num_input_tokens_seen": 123642850, + "step": 5748, + "time_per_iteration": 2.749892473220825 + }, + { + "auxiliary_loss_clip": 0.01138849, + "auxiliary_loss_mlp": 0.0102083, + "balance_loss_clip": 1.0085125, + "balance_loss_mlp": 1.01374304, + "epoch": 0.6912763782841339, + "flos": 24206557386240.0, + "grad_norm": 2.0981929031430004, + "language_loss": 0.72062111, + "learning_rate": 9.190758327509935e-07, + "loss": 0.7422179, + "num_input_tokens_seen": 123661595, + "step": 5749, + "time_per_iteration": 2.728789806365967 + }, + { + "auxiliary_loss_clip": 0.00991827, + "auxiliary_loss_mlp": 0.00500827, + "balance_loss_clip": 0.86512136, + "balance_loss_mlp": 0.0, + "epoch": 0.6913966211747731, + "flos": 52329641091840.0, + "grad_norm": 0.9239107061386759, + "language_loss": 0.64511108, + "learning_rate": 9.184205126490767e-07, + "loss": 0.66003764, + "num_input_tokens_seen": 123710490, + "step": 5750, + "time_per_iteration": 3.1225807666778564 + }, + { + "auxiliary_loss_clip": 0.01006202, + "auxiliary_loss_mlp": 0.00501157, + "balance_loss_clip": 0.89989471, + "balance_loss_mlp": 0.0, + "epoch": 0.6915168640654121, + "flos": 66741274851840.0, + "grad_norm": 1.1036577203025046, + "language_loss": 0.59596771, + "learning_rate": 9.177653566219075e-07, + "loss": 0.61104131, + "num_input_tokens_seen": 123765215, + "step": 5751, + "time_per_iteration": 3.1724042892456055 + }, + { + "auxiliary_loss_clip": 0.01101363, + "auxiliary_loss_mlp": 0.0102445, + "balance_loss_clip": 0.93342519, + "balance_loss_mlp": 1.0169102, + "epoch": 0.6916371069560512, + "flos": 18296523175680.0, + "grad_norm": 4.324537688253975, + "language_loss": 0.76407909, + "learning_rate": 9.171103647688744e-07, + "loss": 0.78533721, + "num_input_tokens_seen": 123783955, + "step": 5752, + "time_per_iteration": 2.696579694747925 + }, + { + "auxiliary_loss_clip": 0.01020501, + "auxiliary_loss_mlp": 0.01024733, + "balance_loss_clip": 0.81785214, + "balance_loss_mlp": 1.01758075, + "epoch": 0.6917573498466904, + "flos": 19645794685440.0, + "grad_norm": 2.3571953946378135, + "language_loss": 0.69328845, + "learning_rate": 9.164555371893367e-07, + "loss": 0.71374077, + "num_input_tokens_seen": 123803885, + "step": 5753, + "time_per_iteration": 2.8606021404266357 + }, + { + "auxiliary_loss_clip": 0.01145423, + "auxiliary_loss_mlp": 0.00509227, + "balance_loss_clip": 1.01304662, + "balance_loss_mlp": 0.0, + "epoch": 0.6918775927373294, + "flos": 14210319985920.0, + "grad_norm": 1.7775327824469132, + "language_loss": 0.7534678, + "learning_rate": 9.158008739826333e-07, + "loss": 0.77001429, + "num_input_tokens_seen": 123821485, + "step": 5754, + "time_per_iteration": 2.8981997966766357 + }, + { + "auxiliary_loss_clip": 0.01119424, + "auxiliary_loss_mlp": 0.01026887, + "balance_loss_clip": 0.97547293, + "balance_loss_mlp": 1.01919174, + "epoch": 0.6919978356279685, + "flos": 23985455218560.0, + "grad_norm": 1.669124339032961, + "language_loss": 0.86455345, + "learning_rate": 9.151463752480744e-07, + "loss": 0.88601661, + "num_input_tokens_seen": 123840215, + "step": 5755, + "time_per_iteration": 2.67126727104187 + }, + { + "auxiliary_loss_clip": 0.01088569, + "auxiliary_loss_mlp": 0.01028868, + "balance_loss_clip": 0.93257165, + "balance_loss_mlp": 1.02159882, + "epoch": 0.6921180785186076, + "flos": 23622937205760.0, + "grad_norm": 1.432629434234581, + "language_loss": 0.80087358, + "learning_rate": 9.144920410849493e-07, + "loss": 0.82204795, + "num_input_tokens_seen": 123861450, + "step": 5756, + "time_per_iteration": 2.747800588607788 + }, + { + "auxiliary_loss_clip": 0.01125832, + "auxiliary_loss_mlp": 0.01025926, + "balance_loss_clip": 0.97300386, + "balance_loss_mlp": 1.01894069, + "epoch": 0.6922383214092467, + "flos": 21142623265920.0, + "grad_norm": 2.256789083405993, + "language_loss": 0.8068018, + "learning_rate": 9.138378715925176e-07, + "loss": 0.82831943, + "num_input_tokens_seen": 123880545, + "step": 5757, + "time_per_iteration": 4.4551167488098145 + }, + { + "auxiliary_loss_clip": 0.01114095, + "auxiliary_loss_mlp": 0.01021471, + "balance_loss_clip": 0.97136617, + "balance_loss_mlp": 1.0142051, + "epoch": 0.6923585642998857, + "flos": 21470667200640.0, + "grad_norm": 2.9417363296437693, + "language_loss": 0.80757093, + "learning_rate": 9.131838668700167e-07, + "loss": 0.82892656, + "num_input_tokens_seen": 123900615, + "step": 5758, + "time_per_iteration": 2.685431480407715 + }, + { + "auxiliary_loss_clip": 0.01096742, + "auxiliary_loss_mlp": 0.01026438, + "balance_loss_clip": 0.93166256, + "balance_loss_mlp": 1.01979518, + "epoch": 0.6924788071905249, + "flos": 21105204272640.0, + "grad_norm": 1.9148458370735792, + "language_loss": 0.86497951, + "learning_rate": 9.125300270166598e-07, + "loss": 0.88621128, + "num_input_tokens_seen": 123921220, + "step": 5759, + "time_per_iteration": 2.7794346809387207 + }, + { + "auxiliary_loss_clip": 0.01103827, + "auxiliary_loss_mlp": 0.01024886, + "balance_loss_clip": 0.93283045, + "balance_loss_mlp": 1.01751256, + "epoch": 0.692599050081164, + "flos": 26250018117120.0, + "grad_norm": 1.6861307220855821, + "language_loss": 0.85886633, + "learning_rate": 9.118763521316324e-07, + "loss": 0.88015354, + "num_input_tokens_seen": 123941795, + "step": 5760, + "time_per_iteration": 2.801017999649048 + }, + { + "auxiliary_loss_clip": 0.01167049, + "auxiliary_loss_mlp": 0.00509563, + "balance_loss_clip": 1.04870892, + "balance_loss_mlp": 0.0, + "epoch": 0.692719292971803, + "flos": 20885215426560.0, + "grad_norm": 2.1544138400810158, + "language_loss": 0.76417685, + "learning_rate": 9.112228423140987e-07, + "loss": 0.78094298, + "num_input_tokens_seen": 123960715, + "step": 5761, + "time_per_iteration": 3.59798264503479 + }, + { + "auxiliary_loss_clip": 0.01126954, + "auxiliary_loss_mlp": 0.01023837, + "balance_loss_clip": 0.97497618, + "balance_loss_mlp": 1.01645803, + "epoch": 0.6928395358624422, + "flos": 25921938268800.0, + "grad_norm": 2.4584003052908314, + "language_loss": 0.86836082, + "learning_rate": 9.105694976631932e-07, + "loss": 0.88986874, + "num_input_tokens_seen": 123978625, + "step": 5762, + "time_per_iteration": 2.744628667831421 + }, + { + "auxiliary_loss_clip": 0.01147652, + "auxiliary_loss_mlp": 0.01026479, + "balance_loss_clip": 1.01415825, + "balance_loss_mlp": 1.01947272, + "epoch": 0.6929597787530812, + "flos": 23586559706880.0, + "grad_norm": 1.8777937844778032, + "language_loss": 0.72471166, + "learning_rate": 9.099163182780283e-07, + "loss": 0.74645305, + "num_input_tokens_seen": 123996780, + "step": 5763, + "time_per_iteration": 2.677149772644043 + }, + { + "auxiliary_loss_clip": 0.01119596, + "auxiliary_loss_mlp": 0.01026859, + "balance_loss_clip": 0.97412056, + "balance_loss_mlp": 1.01930094, + "epoch": 0.6930800216437203, + "flos": 18255656476800.0, + "grad_norm": 2.455415239670791, + "language_loss": 0.49814025, + "learning_rate": 9.092633042576916e-07, + "loss": 0.5196048, + "num_input_tokens_seen": 124014045, + "step": 5764, + "time_per_iteration": 2.6237199306488037 + }, + { + "auxiliary_loss_clip": 0.01116219, + "auxiliary_loss_mlp": 0.01022389, + "balance_loss_clip": 0.9734965, + "balance_loss_mlp": 1.01527214, + "epoch": 0.6932002645343595, + "flos": 29168621809920.0, + "grad_norm": 2.094684665686504, + "language_loss": 0.56407255, + "learning_rate": 9.086104557012446e-07, + "loss": 0.58545858, + "num_input_tokens_seen": 124034615, + "step": 5765, + "time_per_iteration": 2.712000846862793 + }, + { + "auxiliary_loss_clip": 0.01136265, + "auxiliary_loss_mlp": 0.01023816, + "balance_loss_clip": 1.00832677, + "balance_loss_mlp": 1.01665425, + "epoch": 0.6933205074249985, + "flos": 23842746483840.0, + "grad_norm": 1.7841629967034243, + "language_loss": 0.65693974, + "learning_rate": 9.079577727077239e-07, + "loss": 0.67854059, + "num_input_tokens_seen": 124053445, + "step": 5766, + "time_per_iteration": 2.6683013439178467 + }, + { + "auxiliary_loss_clip": 0.01144333, + "auxiliary_loss_mlp": 0.01026261, + "balance_loss_clip": 1.01154399, + "balance_loss_mlp": 1.01845932, + "epoch": 0.6934407503156376, + "flos": 24166696268160.0, + "grad_norm": 3.1160743381111566, + "language_loss": 0.72097683, + "learning_rate": 9.073052553761404e-07, + "loss": 0.74268281, + "num_input_tokens_seen": 124072810, + "step": 5767, + "time_per_iteration": 2.652632474899292 + }, + { + "auxiliary_loss_clip": 0.01076972, + "auxiliary_loss_mlp": 0.01029882, + "balance_loss_clip": 0.89621478, + "balance_loss_mlp": 1.02179945, + "epoch": 0.6935609932062767, + "flos": 20631327120000.0, + "grad_norm": 1.6450509598613672, + "language_loss": 0.77972972, + "learning_rate": 9.066529038054805e-07, + "loss": 0.8007983, + "num_input_tokens_seen": 124092875, + "step": 5768, + "time_per_iteration": 2.769788980484009 + }, + { + "auxiliary_loss_clip": 0.01121749, + "auxiliary_loss_mlp": 0.01031413, + "balance_loss_clip": 0.97438633, + "balance_loss_mlp": 1.02387929, + "epoch": 0.6936812360969158, + "flos": 18254184019200.0, + "grad_norm": 1.8384315040425465, + "language_loss": 0.73941344, + "learning_rate": 9.060007180947071e-07, + "loss": 0.76094508, + "num_input_tokens_seen": 124110930, + "step": 5769, + "time_per_iteration": 3.5550625324249268 + }, + { + "auxiliary_loss_clip": 0.01076788, + "auxiliary_loss_mlp": 0.01026205, + "balance_loss_clip": 0.89388764, + "balance_loss_mlp": 1.01833713, + "epoch": 0.6938014789875548, + "flos": 31317336368640.0, + "grad_norm": 1.7488491702661588, + "language_loss": 0.73733413, + "learning_rate": 9.053486983427534e-07, + "loss": 0.75836396, + "num_input_tokens_seen": 124132180, + "step": 5770, + "time_per_iteration": 2.855607271194458 + }, + { + "auxiliary_loss_clip": 0.01124437, + "auxiliary_loss_mlp": 0.01021792, + "balance_loss_clip": 0.97139776, + "balance_loss_mlp": 1.01407063, + "epoch": 0.6939217218781939, + "flos": 17528429721600.0, + "grad_norm": 12.227260759300647, + "language_loss": 0.70664132, + "learning_rate": 9.046968446485326e-07, + "loss": 0.72810364, + "num_input_tokens_seen": 124150585, + "step": 5771, + "time_per_iteration": 2.7102673053741455 + }, + { + "auxiliary_loss_clip": 0.01149859, + "auxiliary_loss_mlp": 0.01031183, + "balance_loss_clip": 1.01415253, + "balance_loss_mlp": 1.02345848, + "epoch": 0.6940419647688331, + "flos": 18551776199040.0, + "grad_norm": 1.983630061299044, + "language_loss": 0.70548666, + "learning_rate": 9.040451571109295e-07, + "loss": 0.72729707, + "num_input_tokens_seen": 124166205, + "step": 5772, + "time_per_iteration": 2.599730968475342 + }, + { + "auxiliary_loss_clip": 0.01019727, + "auxiliary_loss_mlp": 0.01012513, + "balance_loss_clip": 0.91931617, + "balance_loss_mlp": 1.01126695, + "epoch": 0.6941622076594721, + "flos": 66926286829440.0, + "grad_norm": 0.831894694258736, + "language_loss": 0.60446584, + "learning_rate": 9.033936358288042e-07, + "loss": 0.62478822, + "num_input_tokens_seen": 124219940, + "step": 5773, + "time_per_iteration": 3.1550846099853516 + }, + { + "auxiliary_loss_clip": 0.01171959, + "auxiliary_loss_mlp": 0.01030008, + "balance_loss_clip": 1.05105865, + "balance_loss_mlp": 1.02238488, + "epoch": 0.6942824505501112, + "flos": 26578062051840.0, + "grad_norm": 1.7439525393983333, + "language_loss": 0.8212952, + "learning_rate": 9.027422809009937e-07, + "loss": 0.84331489, + "num_input_tokens_seen": 124239885, + "step": 5774, + "time_per_iteration": 2.609630584716797 + }, + { + "auxiliary_loss_clip": 0.01146812, + "auxiliary_loss_mlp": 0.01027545, + "balance_loss_clip": 1.01003945, + "balance_loss_mlp": 1.01997519, + "epoch": 0.6944026934407503, + "flos": 21248308056960.0, + "grad_norm": 1.862729783183379, + "language_loss": 0.82847583, + "learning_rate": 9.020910924263054e-07, + "loss": 0.85021937, + "num_input_tokens_seen": 124258410, + "step": 5775, + "time_per_iteration": 2.6277575492858887 + }, + { + "auxiliary_loss_clip": 0.01015882, + "auxiliary_loss_mlp": 0.01010211, + "balance_loss_clip": 0.91641212, + "balance_loss_mlp": 1.008775, + "epoch": 0.6945229363313894, + "flos": 70677191537280.0, + "grad_norm": 0.8108116820668555, + "language_loss": 0.58129978, + "learning_rate": 9.014400705035261e-07, + "loss": 0.60156071, + "num_input_tokens_seen": 124315315, + "step": 5776, + "time_per_iteration": 3.308884382247925 + }, + { + "auxiliary_loss_clip": 0.01168648, + "auxiliary_loss_mlp": 0.01030081, + "balance_loss_clip": 1.0524931, + "balance_loss_mlp": 1.02273142, + "epoch": 0.6946431792220285, + "flos": 18952934267520.0, + "grad_norm": 2.3125844823973405, + "language_loss": 0.76553965, + "learning_rate": 9.00789215231414e-07, + "loss": 0.78752691, + "num_input_tokens_seen": 124333710, + "step": 5777, + "time_per_iteration": 2.5978076457977295 + }, + { + "auxiliary_loss_clip": 0.01095024, + "auxiliary_loss_mlp": 0.00509883, + "balance_loss_clip": 0.92996126, + "balance_loss_mlp": 0.0, + "epoch": 0.6947634221126676, + "flos": 20338834671360.0, + "grad_norm": 2.1184185439268997, + "language_loss": 0.82263249, + "learning_rate": 9.001385267087056e-07, + "loss": 0.83868158, + "num_input_tokens_seen": 124352855, + "step": 5778, + "time_per_iteration": 2.728576898574829 + }, + { + "auxiliary_loss_clip": 0.01148333, + "auxiliary_loss_mlp": 0.01025924, + "balance_loss_clip": 1.01423991, + "balance_loss_mlp": 1.01859593, + "epoch": 0.6948836650033067, + "flos": 21833723917440.0, + "grad_norm": 1.6098895891139997, + "language_loss": 0.70235294, + "learning_rate": 8.994880050341072e-07, + "loss": 0.72409552, + "num_input_tokens_seen": 124372960, + "step": 5779, + "time_per_iteration": 2.634962558746338 + }, + { + "auxiliary_loss_clip": 0.01116155, + "auxiliary_loss_mlp": 0.01026799, + "balance_loss_clip": 0.97341013, + "balance_loss_mlp": 1.0196439, + "epoch": 0.6950039078939457, + "flos": 23657519024640.0, + "grad_norm": 4.336686650883618, + "language_loss": 0.7775085, + "learning_rate": 8.988376503063026e-07, + "loss": 0.79893804, + "num_input_tokens_seen": 124394220, + "step": 5780, + "time_per_iteration": 2.7153987884521484 + }, + { + "auxiliary_loss_clip": 0.01080755, + "auxiliary_loss_mlp": 0.01026751, + "balance_loss_clip": 0.89599848, + "balance_loss_mlp": 1.01933575, + "epoch": 0.6951241507845849, + "flos": 21792462168960.0, + "grad_norm": 2.1127268180080816, + "language_loss": 0.81813097, + "learning_rate": 8.981874626239521e-07, + "loss": 0.8392061, + "num_input_tokens_seen": 124412795, + "step": 5781, + "time_per_iteration": 2.799713373184204 + }, + { + "auxiliary_loss_clip": 0.01146104, + "auxiliary_loss_mlp": 0.0102691, + "balance_loss_clip": 1.01350713, + "balance_loss_mlp": 1.01908422, + "epoch": 0.695244393675224, + "flos": 14647568244480.0, + "grad_norm": 2.2960853022815546, + "language_loss": 0.88492644, + "learning_rate": 8.975374420856872e-07, + "loss": 0.90665656, + "num_input_tokens_seen": 124429690, + "step": 5782, + "time_per_iteration": 2.6557064056396484 + }, + { + "auxiliary_loss_clip": 0.01087789, + "auxiliary_loss_mlp": 0.01025206, + "balance_loss_clip": 0.93120527, + "balance_loss_mlp": 1.01802921, + "epoch": 0.695364636565863, + "flos": 16873203778560.0, + "grad_norm": 2.2929760283136478, + "language_loss": 0.73084944, + "learning_rate": 8.968875887901157e-07, + "loss": 0.75197941, + "num_input_tokens_seen": 124447070, + "step": 5783, + "time_per_iteration": 3.6688570976257324 + }, + { + "auxiliary_loss_clip": 0.0111833, + "auxiliary_loss_mlp": 0.01028283, + "balance_loss_clip": 0.96941411, + "balance_loss_mlp": 1.0208025, + "epoch": 0.6954848794565022, + "flos": 19354523299200.0, + "grad_norm": 1.8503644300866362, + "language_loss": 0.62956011, + "learning_rate": 8.9623790283582e-07, + "loss": 0.65102625, + "num_input_tokens_seen": 124464950, + "step": 5784, + "time_per_iteration": 3.582793712615967 + }, + { + "auxiliary_loss_clip": 0.01100543, + "auxiliary_loss_mlp": 0.01033894, + "balance_loss_clip": 0.93645334, + "balance_loss_mlp": 1.02574587, + "epoch": 0.6956051223471412, + "flos": 18990209606400.0, + "grad_norm": 4.458100978519555, + "language_loss": 0.76710063, + "learning_rate": 8.955883843213561e-07, + "loss": 0.788445, + "num_input_tokens_seen": 124483965, + "step": 5785, + "time_per_iteration": 2.7027924060821533 + }, + { + "auxiliary_loss_clip": 0.01155514, + "auxiliary_loss_mlp": 0.01035244, + "balance_loss_clip": 1.01449847, + "balance_loss_mlp": 1.0276593, + "epoch": 0.6957253652377803, + "flos": 16107229226880.0, + "grad_norm": 2.5741350937631493, + "language_loss": 0.87102294, + "learning_rate": 8.949390333452569e-07, + "loss": 0.89293057, + "num_input_tokens_seen": 124501910, + "step": 5786, + "time_per_iteration": 3.5517020225524902 + }, + { + "auxiliary_loss_clip": 0.0116938, + "auxiliary_loss_mlp": 0.01026974, + "balance_loss_clip": 1.05202305, + "balance_loss_mlp": 1.01948762, + "epoch": 0.6958456081284194, + "flos": 29388646569600.0, + "grad_norm": 1.9080597529566778, + "language_loss": 0.67687869, + "learning_rate": 8.942898500060279e-07, + "loss": 0.69884223, + "num_input_tokens_seen": 124521625, + "step": 5787, + "time_per_iteration": 2.6753623485565186 + }, + { + "auxiliary_loss_clip": 0.01081693, + "auxiliary_loss_mlp": 0.01026243, + "balance_loss_clip": 0.89779031, + "balance_loss_mlp": 1.01801145, + "epoch": 0.6959658510190585, + "flos": 25154850395520.0, + "grad_norm": 2.541047009104164, + "language_loss": 0.71039879, + "learning_rate": 8.936408344021493e-07, + "loss": 0.73147815, + "num_input_tokens_seen": 124538540, + "step": 5788, + "time_per_iteration": 2.8139917850494385 + }, + { + "auxiliary_loss_clip": 0.01132657, + "auxiliary_loss_mlp": 0.01026146, + "balance_loss_clip": 0.9780516, + "balance_loss_mlp": 1.01865959, + "epoch": 0.6960860939096976, + "flos": 42814388759040.0, + "grad_norm": 2.1375051456800516, + "language_loss": 0.71309638, + "learning_rate": 8.929919866320765e-07, + "loss": 0.73468435, + "num_input_tokens_seen": 124559355, + "step": 5789, + "time_per_iteration": 2.8456666469573975 + }, + { + "auxiliary_loss_clip": 0.01104029, + "auxiliary_loss_mlp": 0.00509824, + "balance_loss_clip": 0.93452245, + "balance_loss_mlp": 0.0, + "epoch": 0.6962063368003367, + "flos": 17566566986880.0, + "grad_norm": 1.9265510065819154, + "language_loss": 0.8130545, + "learning_rate": 8.923433067942385e-07, + "loss": 0.829193, + "num_input_tokens_seen": 124577920, + "step": 5790, + "time_per_iteration": 2.7409138679504395 + }, + { + "auxiliary_loss_clip": 0.0110641, + "auxiliary_loss_mlp": 0.01026061, + "balance_loss_clip": 0.9368242, + "balance_loss_mlp": 1.01937628, + "epoch": 0.6963265796909758, + "flos": 21251648021760.0, + "grad_norm": 2.231991810960731, + "language_loss": 0.68771183, + "learning_rate": 8.916947949870417e-07, + "loss": 0.70903653, + "num_input_tokens_seen": 124597585, + "step": 5791, + "time_per_iteration": 2.705136299133301 + }, + { + "auxiliary_loss_clip": 0.01046928, + "auxiliary_loss_mlp": 0.01001662, + "balance_loss_clip": 0.97529525, + "balance_loss_mlp": 1.00054729, + "epoch": 0.6964468225816148, + "flos": 68828295801600.0, + "grad_norm": 1.0658495959354795, + "language_loss": 0.58142698, + "learning_rate": 8.910464513088615e-07, + "loss": 0.60191286, + "num_input_tokens_seen": 124661625, + "step": 5792, + "time_per_iteration": 3.3080906867980957 + }, + { + "auxiliary_loss_clip": 0.01113794, + "auxiliary_loss_mlp": 0.01026659, + "balance_loss_clip": 0.97098035, + "balance_loss_mlp": 1.01902652, + "epoch": 0.696567065472254, + "flos": 18950887192320.0, + "grad_norm": 1.8408485353555513, + "language_loss": 0.78852737, + "learning_rate": 8.903982758580542e-07, + "loss": 0.80993187, + "num_input_tokens_seen": 124680565, + "step": 5793, + "time_per_iteration": 2.7037432193756104 + }, + { + "auxiliary_loss_clip": 0.0112142, + "auxiliary_loss_mlp": 0.01026363, + "balance_loss_clip": 0.97547495, + "balance_loss_mlp": 1.01828718, + "epoch": 0.696687308362893, + "flos": 22856675345280.0, + "grad_norm": 2.5275359645943616, + "language_loss": 0.80393267, + "learning_rate": 8.897502687329457e-07, + "loss": 0.82541049, + "num_input_tokens_seen": 124700365, + "step": 5794, + "time_per_iteration": 2.686445474624634 + }, + { + "auxiliary_loss_clip": 0.01092479, + "auxiliary_loss_mlp": 0.01024344, + "balance_loss_clip": 0.93214762, + "balance_loss_mlp": 1.01735282, + "epoch": 0.6968075512535321, + "flos": 24972926987520.0, + "grad_norm": 2.2690781332823633, + "language_loss": 0.79877096, + "learning_rate": 8.891024300318382e-07, + "loss": 0.8199392, + "num_input_tokens_seen": 124718935, + "step": 5795, + "time_per_iteration": 3.6866021156311035 + }, + { + "auxiliary_loss_clip": 0.01089332, + "auxiliary_loss_mlp": 0.01029479, + "balance_loss_clip": 0.93265104, + "balance_loss_mlp": 1.02236211, + "epoch": 0.6969277941441713, + "flos": 21030438113280.0, + "grad_norm": 1.756655747324812, + "language_loss": 0.75644135, + "learning_rate": 8.884547598530103e-07, + "loss": 0.77762944, + "num_input_tokens_seen": 124739505, + "step": 5796, + "time_per_iteration": 2.7248051166534424 + }, + { + "auxiliary_loss_clip": 0.01019025, + "auxiliary_loss_mlp": 0.0102687, + "balance_loss_clip": 0.81747103, + "balance_loss_mlp": 1.01967859, + "epoch": 0.6970480370348103, + "flos": 21579404647680.0, + "grad_norm": 2.5376737034707313, + "language_loss": 0.75441074, + "learning_rate": 8.8780725829471e-07, + "loss": 0.77486974, + "num_input_tokens_seen": 124757410, + "step": 5797, + "time_per_iteration": 2.828737497329712 + }, + { + "auxiliary_loss_clip": 0.01168159, + "auxiliary_loss_mlp": 0.01031272, + "balance_loss_clip": 1.04964805, + "balance_loss_mlp": 1.02402449, + "epoch": 0.6971682799254494, + "flos": 22419175691520.0, + "grad_norm": 1.9557845877620297, + "language_loss": 0.77983189, + "learning_rate": 8.87159925455165e-07, + "loss": 0.80182624, + "num_input_tokens_seen": 124777240, + "step": 5798, + "time_per_iteration": 2.61140513420105 + }, + { + "auxiliary_loss_clip": 0.01096662, + "auxiliary_loss_mlp": 0.01028472, + "balance_loss_clip": 0.93558568, + "balance_loss_mlp": 1.0214026, + "epoch": 0.6972885228160886, + "flos": 20005834659840.0, + "grad_norm": 1.9713778997317593, + "language_loss": 0.73624027, + "learning_rate": 8.865127614325738e-07, + "loss": 0.75749159, + "num_input_tokens_seen": 124795670, + "step": 5799, + "time_per_iteration": 2.7779242992401123 + }, + { + "auxiliary_loss_clip": 0.01114285, + "auxiliary_loss_mlp": 0.01034483, + "balance_loss_clip": 0.97031617, + "balance_loss_mlp": 1.02549458, + "epoch": 0.6974087657067276, + "flos": 37853437656960.0, + "grad_norm": 1.7697100649008786, + "language_loss": 0.6697315, + "learning_rate": 8.85865766325113e-07, + "loss": 0.69121921, + "num_input_tokens_seen": 124819600, + "step": 5800, + "time_per_iteration": 2.8689823150634766 + }, + { + "auxiliary_loss_clip": 0.01119118, + "auxiliary_loss_mlp": 0.01026919, + "balance_loss_clip": 0.97251201, + "balance_loss_mlp": 1.01917601, + "epoch": 0.6975290085973667, + "flos": 29489267543040.0, + "grad_norm": 2.7803565646433666, + "language_loss": 0.72006208, + "learning_rate": 8.852189402309287e-07, + "loss": 0.74152243, + "num_input_tokens_seen": 124838785, + "step": 5801, + "time_per_iteration": 2.7163918018341064 + }, + { + "auxiliary_loss_clip": 0.01146244, + "auxiliary_loss_mlp": 0.01027571, + "balance_loss_clip": 1.01232767, + "balance_loss_mlp": 1.0198524, + "epoch": 0.6976492514880057, + "flos": 12895630295040.0, + "grad_norm": 2.1122703042610484, + "language_loss": 0.74228632, + "learning_rate": 8.845722832481441e-07, + "loss": 0.7640245, + "num_input_tokens_seen": 124854215, + "step": 5802, + "time_per_iteration": 2.645874261856079 + }, + { + "auxiliary_loss_clip": 0.01144512, + "auxiliary_loss_mlp": 0.01029534, + "balance_loss_clip": 1.01206255, + "balance_loss_mlp": 1.02211916, + "epoch": 0.6977694943786449, + "flos": 24352929308160.0, + "grad_norm": 3.313844540698486, + "language_loss": 0.77931178, + "learning_rate": 8.83925795474858e-07, + "loss": 0.80105221, + "num_input_tokens_seen": 124874340, + "step": 5803, + "time_per_iteration": 2.705620288848877 + }, + { + "auxiliary_loss_clip": 0.01096245, + "auxiliary_loss_mlp": 0.01029547, + "balance_loss_clip": 0.93604398, + "balance_loss_mlp": 1.0214293, + "epoch": 0.6978897372692839, + "flos": 29898470257920.0, + "grad_norm": 2.103254555402428, + "language_loss": 0.59309447, + "learning_rate": 8.832794770091414e-07, + "loss": 0.61435235, + "num_input_tokens_seen": 124895175, + "step": 5804, + "time_per_iteration": 2.7777469158172607 + }, + { + "auxiliary_loss_clip": 0.01125903, + "auxiliary_loss_mlp": 0.01023582, + "balance_loss_clip": 0.97292054, + "balance_loss_mlp": 1.01608396, + "epoch": 0.698009980159923, + "flos": 21761579450880.0, + "grad_norm": 1.9601036927031155, + "language_loss": 0.82299209, + "learning_rate": 8.826333279490401e-07, + "loss": 0.84448695, + "num_input_tokens_seen": 124915810, + "step": 5805, + "time_per_iteration": 2.66252064704895 + }, + { + "auxiliary_loss_clip": 0.01127364, + "auxiliary_loss_mlp": 0.0103104, + "balance_loss_clip": 0.97610283, + "balance_loss_mlp": 1.02400374, + "epoch": 0.6981302230505622, + "flos": 19857164267520.0, + "grad_norm": 2.19716980726336, + "language_loss": 0.68800676, + "learning_rate": 8.819873483925748e-07, + "loss": 0.70959079, + "num_input_tokens_seen": 124932930, + "step": 5806, + "time_per_iteration": 2.6815147399902344 + }, + { + "auxiliary_loss_clip": 0.01102277, + "auxiliary_loss_mlp": 0.00510002, + "balance_loss_clip": 0.9355551, + "balance_loss_mlp": 0.0, + "epoch": 0.6982504659412012, + "flos": 22198648141440.0, + "grad_norm": 2.493735557023335, + "language_loss": 0.74851745, + "learning_rate": 8.81341538437739e-07, + "loss": 0.76464027, + "num_input_tokens_seen": 124951220, + "step": 5807, + "time_per_iteration": 2.7238643169403076 + }, + { + "auxiliary_loss_clip": 0.01125285, + "auxiliary_loss_mlp": 0.01029373, + "balance_loss_clip": 0.97144043, + "balance_loss_mlp": 1.02242935, + "epoch": 0.6983707088318403, + "flos": 35588479708800.0, + "grad_norm": 1.6252316978218646, + "language_loss": 0.68013775, + "learning_rate": 8.80695898182503e-07, + "loss": 0.70168436, + "num_input_tokens_seen": 124972200, + "step": 5808, + "time_per_iteration": 2.839712619781494 + }, + { + "auxiliary_loss_clip": 0.01053954, + "auxiliary_loss_mlp": 0.01004084, + "balance_loss_clip": 0.99073535, + "balance_loss_mlp": 1.00281477, + "epoch": 0.6984909517224794, + "flos": 65440052760960.0, + "grad_norm": 0.8010595195416507, + "language_loss": 0.65078944, + "learning_rate": 8.800504277248093e-07, + "loss": 0.67136973, + "num_input_tokens_seen": 125036950, + "step": 5809, + "time_per_iteration": 4.177175998687744 + }, + { + "auxiliary_loss_clip": 0.01098235, + "auxiliary_loss_mlp": 0.00509219, + "balance_loss_clip": 0.9426499, + "balance_loss_mlp": 0.0, + "epoch": 0.6986111946131185, + "flos": 18546927863040.0, + "grad_norm": 1.5803223959989907, + "language_loss": 0.75250292, + "learning_rate": 8.794051271625753e-07, + "loss": 0.76857746, + "num_input_tokens_seen": 125054585, + "step": 5810, + "time_per_iteration": 2.6714956760406494 + }, + { + "auxiliary_loss_clip": 0.01120939, + "auxiliary_loss_mlp": 0.01026537, + "balance_loss_clip": 0.9729386, + "balance_loss_mlp": 1.01893687, + "epoch": 0.6987314375037575, + "flos": 23039173370880.0, + "grad_norm": 1.5591468452028112, + "language_loss": 0.83288407, + "learning_rate": 8.787599965936925e-07, + "loss": 0.85435885, + "num_input_tokens_seen": 125075515, + "step": 5811, + "time_per_iteration": 2.71928071975708 + }, + { + "auxiliary_loss_clip": 0.01093488, + "auxiliary_loss_mlp": 0.01027802, + "balance_loss_clip": 0.93646532, + "balance_loss_mlp": 1.02045918, + "epoch": 0.6988516803943967, + "flos": 38400393029760.0, + "grad_norm": 1.5693557628348997, + "language_loss": 0.72016895, + "learning_rate": 8.781150361160261e-07, + "loss": 0.74138188, + "num_input_tokens_seen": 125097425, + "step": 5812, + "time_per_iteration": 3.7868995666503906 + }, + { + "auxiliary_loss_clip": 0.01103644, + "auxiliary_loss_mlp": 0.01026616, + "balance_loss_clip": 0.9365353, + "balance_loss_mlp": 1.01913238, + "epoch": 0.6989719232850358, + "flos": 24096993926400.0, + "grad_norm": 1.8309658510373417, + "language_loss": 0.73406172, + "learning_rate": 8.774702458274181e-07, + "loss": 0.75536436, + "num_input_tokens_seen": 125117830, + "step": 5813, + "time_per_iteration": 2.7423012256622314 + }, + { + "auxiliary_loss_clip": 0.01146861, + "auxiliary_loss_mlp": 0.01024182, + "balance_loss_clip": 1.01486135, + "balance_loss_mlp": 1.01674926, + "epoch": 0.6990921661756748, + "flos": 14866838818560.0, + "grad_norm": 2.0808553037977244, + "language_loss": 0.70630169, + "learning_rate": 8.768256258256799e-07, + "loss": 0.72801208, + "num_input_tokens_seen": 125134455, + "step": 5814, + "time_per_iteration": 2.6485507488250732 + }, + { + "auxiliary_loss_clip": 0.01149446, + "auxiliary_loss_mlp": 0.01033194, + "balance_loss_clip": 1.01290882, + "balance_loss_mlp": 1.02565062, + "epoch": 0.699212409066314, + "flos": 20193719725440.0, + "grad_norm": 1.6410035640187424, + "language_loss": 0.73610663, + "learning_rate": 8.76181176208602e-07, + "loss": 0.75793302, + "num_input_tokens_seen": 125152555, + "step": 5815, + "time_per_iteration": 2.8020169734954834 + }, + { + "auxiliary_loss_clip": 0.0106562, + "auxiliary_loss_mlp": 0.01027169, + "balance_loss_clip": 0.89502454, + "balance_loss_mlp": 1.01952124, + "epoch": 0.699332651956953, + "flos": 19427888828160.0, + "grad_norm": 1.747274023170851, + "language_loss": 0.73361391, + "learning_rate": 8.755368970739461e-07, + "loss": 0.75454175, + "num_input_tokens_seen": 125171915, + "step": 5816, + "time_per_iteration": 2.8338427543640137 + }, + { + "auxiliary_loss_clip": 0.01103289, + "auxiliary_loss_mlp": 0.01028274, + "balance_loss_clip": 0.93257833, + "balance_loss_mlp": 1.02038217, + "epoch": 0.6994528948475921, + "flos": 16143714466560.0, + "grad_norm": 2.058214885976885, + "language_loss": 0.61415589, + "learning_rate": 8.748927885194479e-07, + "loss": 0.63547146, + "num_input_tokens_seen": 125190220, + "step": 5817, + "time_per_iteration": 2.6984851360321045 + }, + { + "auxiliary_loss_clip": 0.00999382, + "auxiliary_loss_mlp": 0.01001365, + "balance_loss_clip": 0.90262675, + "balance_loss_mlp": 1.00010145, + "epoch": 0.6995731377382313, + "flos": 64952420699520.0, + "grad_norm": 0.7928403569275212, + "language_loss": 0.57352537, + "learning_rate": 8.742488506428209e-07, + "loss": 0.59353286, + "num_input_tokens_seen": 125249310, + "step": 5818, + "time_per_iteration": 3.2403712272644043 + }, + { + "auxiliary_loss_clip": 0.01123437, + "auxiliary_loss_mlp": 0.00508618, + "balance_loss_clip": 0.97213906, + "balance_loss_mlp": 0.0, + "epoch": 0.6996933806288703, + "flos": 24900136076160.0, + "grad_norm": 1.6468041118877657, + "language_loss": 0.78009963, + "learning_rate": 8.736050835417466e-07, + "loss": 0.79642016, + "num_input_tokens_seen": 125269350, + "step": 5819, + "time_per_iteration": 2.759202718734741 + }, + { + "auxiliary_loss_clip": 0.01148145, + "auxiliary_loss_mlp": 0.01026232, + "balance_loss_clip": 1.0114491, + "balance_loss_mlp": 1.01863801, + "epoch": 0.6998136235195094, + "flos": 20777806782720.0, + "grad_norm": 4.120465912357785, + "language_loss": 0.61325127, + "learning_rate": 8.729614873138862e-07, + "loss": 0.63499504, + "num_input_tokens_seen": 125286985, + "step": 5820, + "time_per_iteration": 3.59261417388916 + }, + { + "auxiliary_loss_clip": 0.01082021, + "auxiliary_loss_mlp": 0.01029288, + "balance_loss_clip": 0.89970708, + "balance_loss_mlp": 1.02204347, + "epoch": 0.6999338664101485, + "flos": 23733470332800.0, + "grad_norm": 2.0389235725491592, + "language_loss": 0.77804255, + "learning_rate": 8.723180620568716e-07, + "loss": 0.79915559, + "num_input_tokens_seen": 125306240, + "step": 5821, + "time_per_iteration": 2.7758631706237793 + }, + { + "auxiliary_loss_clip": 0.0112432, + "auxiliary_loss_mlp": 0.0102554, + "balance_loss_clip": 0.97189128, + "balance_loss_mlp": 1.01788127, + "epoch": 0.7000541093007876, + "flos": 19864598382720.0, + "grad_norm": 1.8723905015995677, + "language_loss": 0.85041535, + "learning_rate": 8.716748078683116e-07, + "loss": 0.87191391, + "num_input_tokens_seen": 125323015, + "step": 5822, + "time_per_iteration": 2.6583354473114014 + }, + { + "auxiliary_loss_clip": 0.01019146, + "auxiliary_loss_mlp": 0.01027966, + "balance_loss_clip": 0.81727248, + "balance_loss_mlp": 1.01996088, + "epoch": 0.7001743521914267, + "flos": 29679056029440.0, + "grad_norm": 1.9754849788827087, + "language_loss": 0.68504006, + "learning_rate": 8.710317248457855e-07, + "loss": 0.70551121, + "num_input_tokens_seen": 125342630, + "step": 5823, + "time_per_iteration": 2.866809606552124 + }, + { + "auxiliary_loss_clip": 0.0112194, + "auxiliary_loss_mlp": 0.01026544, + "balance_loss_clip": 0.97664684, + "balance_loss_mlp": 1.0191201, + "epoch": 0.7002945950820658, + "flos": 27489762080640.0, + "grad_norm": 2.916790637298011, + "language_loss": 0.72344291, + "learning_rate": 8.703888130868482e-07, + "loss": 0.7449277, + "num_input_tokens_seen": 125364480, + "step": 5824, + "time_per_iteration": 2.7414350509643555 + }, + { + "auxiliary_loss_clip": 0.01099364, + "auxiliary_loss_mlp": 0.01027032, + "balance_loss_clip": 0.93602192, + "balance_loss_mlp": 1.0193727, + "epoch": 0.7004148379727049, + "flos": 22158463800960.0, + "grad_norm": 1.9003149067476328, + "language_loss": 0.82087463, + "learning_rate": 8.697460726890307e-07, + "loss": 0.84213859, + "num_input_tokens_seen": 125381625, + "step": 5825, + "time_per_iteration": 2.7634875774383545 + }, + { + "auxiliary_loss_clip": 0.01095029, + "auxiliary_loss_mlp": 0.00509582, + "balance_loss_clip": 0.93089271, + "balance_loss_mlp": 0.0, + "epoch": 0.7005350808633439, + "flos": 19423758764160.0, + "grad_norm": 1.9410750635203358, + "language_loss": 0.90528262, + "learning_rate": 8.691035037498354e-07, + "loss": 0.92132872, + "num_input_tokens_seen": 125397615, + "step": 5826, + "time_per_iteration": 2.697458028793335 + }, + { + "auxiliary_loss_clip": 0.01114275, + "auxiliary_loss_mlp": 0.01021984, + "balance_loss_clip": 0.96792817, + "balance_loss_mlp": 1.01522839, + "epoch": 0.7006553237539831, + "flos": 23476708938240.0, + "grad_norm": 1.6947015925256284, + "language_loss": 0.72630131, + "learning_rate": 8.684611063667391e-07, + "loss": 0.74766386, + "num_input_tokens_seen": 125418080, + "step": 5827, + "time_per_iteration": 2.6870017051696777 + }, + { + "auxiliary_loss_clip": 0.01144314, + "auxiliary_loss_mlp": 0.01027496, + "balance_loss_clip": 1.01052618, + "balance_loss_mlp": 1.01988173, + "epoch": 0.7007755666446221, + "flos": 31212872640000.0, + "grad_norm": 2.2345177525070845, + "language_loss": 0.76973367, + "learning_rate": 8.678188806371935e-07, + "loss": 0.79145181, + "num_input_tokens_seen": 125440115, + "step": 5828, + "time_per_iteration": 2.7432000637054443 + }, + { + "auxiliary_loss_clip": 0.01144137, + "auxiliary_loss_mlp": 0.01025587, + "balance_loss_clip": 1.00951529, + "balance_loss_mlp": 1.01810706, + "epoch": 0.7008958095352612, + "flos": 18149899858560.0, + "grad_norm": 2.5098313112018156, + "language_loss": 0.85101134, + "learning_rate": 8.671768266586228e-07, + "loss": 0.87270856, + "num_input_tokens_seen": 125458240, + "step": 5829, + "time_per_iteration": 2.6165611743927 + }, + { + "auxiliary_loss_clip": 0.01093793, + "auxiliary_loss_mlp": 0.01030348, + "balance_loss_clip": 0.93317866, + "balance_loss_mlp": 1.02278447, + "epoch": 0.7010160524259004, + "flos": 27452307173760.0, + "grad_norm": 2.5057701046231227, + "language_loss": 0.78109825, + "learning_rate": 8.665349445284275e-07, + "loss": 0.80233967, + "num_input_tokens_seen": 125477980, + "step": 5830, + "time_per_iteration": 2.7779808044433594 + }, + { + "auxiliary_loss_clip": 0.01104238, + "auxiliary_loss_mlp": 0.01030136, + "balance_loss_clip": 0.94128251, + "balance_loss_mlp": 1.0228045, + "epoch": 0.7011362953165394, + "flos": 23842064125440.0, + "grad_norm": 1.4933721479734383, + "language_loss": 0.80825329, + "learning_rate": 8.658932343439799e-07, + "loss": 0.82959706, + "num_input_tokens_seen": 125497765, + "step": 5831, + "time_per_iteration": 2.7232377529144287 + }, + { + "auxiliary_loss_clip": 0.0117038, + "auxiliary_loss_mlp": 0.01025992, + "balance_loss_clip": 1.05142498, + "balance_loss_mlp": 1.01858068, + "epoch": 0.7012565382071785, + "flos": 24823430582400.0, + "grad_norm": 1.8953444893645295, + "language_loss": 0.77722502, + "learning_rate": 8.65251696202627e-07, + "loss": 0.79918873, + "num_input_tokens_seen": 125514145, + "step": 5832, + "time_per_iteration": 2.6832175254821777 + }, + { + "auxiliary_loss_clip": 0.0110213, + "auxiliary_loss_mlp": 0.01031648, + "balance_loss_clip": 0.93712723, + "balance_loss_mlp": 1.02359498, + "epoch": 0.7013767810978175, + "flos": 21397445326080.0, + "grad_norm": 2.321647446701578, + "language_loss": 0.8731513, + "learning_rate": 8.646103302016896e-07, + "loss": 0.89448905, + "num_input_tokens_seen": 125533115, + "step": 5833, + "time_per_iteration": 2.719571828842163 + }, + { + "auxiliary_loss_clip": 0.01083475, + "auxiliary_loss_mlp": 0.01027907, + "balance_loss_clip": 0.89561939, + "balance_loss_mlp": 1.02032495, + "epoch": 0.7014970239884567, + "flos": 16687150306560.0, + "grad_norm": 2.088872361298674, + "language_loss": 0.88706779, + "learning_rate": 8.639691364384614e-07, + "loss": 0.90818161, + "num_input_tokens_seen": 125550740, + "step": 5834, + "time_per_iteration": 2.7463135719299316 + }, + { + "auxiliary_loss_clip": 0.01126393, + "auxiliary_loss_mlp": 0.01031375, + "balance_loss_clip": 0.9761408, + "balance_loss_mlp": 1.02366805, + "epoch": 0.7016172668790958, + "flos": 12568268718720.0, + "grad_norm": 1.8267394176439193, + "language_loss": 0.72580409, + "learning_rate": 8.633281150102136e-07, + "loss": 0.74738181, + "num_input_tokens_seen": 125567590, + "step": 5835, + "time_per_iteration": 4.557177305221558 + }, + { + "auxiliary_loss_clip": 0.01119218, + "auxiliary_loss_mlp": 0.01032765, + "balance_loss_clip": 0.97302336, + "balance_loss_mlp": 1.02580595, + "epoch": 0.7017375097697348, + "flos": 17452729808640.0, + "grad_norm": 2.1243035076368133, + "language_loss": 0.67844725, + "learning_rate": 8.626872660141855e-07, + "loss": 0.69996715, + "num_input_tokens_seen": 125585500, + "step": 5836, + "time_per_iteration": 2.734546422958374 + }, + { + "auxiliary_loss_clip": 0.01074114, + "auxiliary_loss_mlp": 0.010257, + "balance_loss_clip": 0.89719617, + "balance_loss_mlp": 1.01860988, + "epoch": 0.701857752660374, + "flos": 18513028402560.0, + "grad_norm": 1.9844061099617603, + "language_loss": 0.74742591, + "learning_rate": 8.620465895475957e-07, + "loss": 0.76842409, + "num_input_tokens_seen": 125603720, + "step": 5837, + "time_per_iteration": 2.7847096920013428 + }, + { + "auxiliary_loss_clip": 0.01071326, + "auxiliary_loss_mlp": 0.01028332, + "balance_loss_clip": 0.89880848, + "balance_loss_mlp": 1.02122724, + "epoch": 0.701977995551013, + "flos": 24425971614720.0, + "grad_norm": 1.4933448173929647, + "language_loss": 0.75181043, + "learning_rate": 8.614060857076333e-07, + "loss": 0.772807, + "num_input_tokens_seen": 125624390, + "step": 5838, + "time_per_iteration": 3.7302730083465576 + }, + { + "auxiliary_loss_clip": 0.01117219, + "auxiliary_loss_mlp": 0.01029697, + "balance_loss_clip": 0.97293794, + "balance_loss_mlp": 1.02239549, + "epoch": 0.7020982384416521, + "flos": 23002759958400.0, + "grad_norm": 2.3894247359635763, + "language_loss": 0.74920654, + "learning_rate": 8.60765754591462e-07, + "loss": 0.77067566, + "num_input_tokens_seen": 125644085, + "step": 5839, + "time_per_iteration": 2.6538586616516113 + }, + { + "auxiliary_loss_clip": 0.01164471, + "auxiliary_loss_mlp": 0.01025905, + "balance_loss_clip": 1.0482204, + "balance_loss_mlp": 1.01805556, + "epoch": 0.7022184813322913, + "flos": 20449080489600.0, + "grad_norm": 7.147969203265878, + "language_loss": 0.72786164, + "learning_rate": 8.601255962962211e-07, + "loss": 0.7497654, + "num_input_tokens_seen": 125663095, + "step": 5840, + "time_per_iteration": 2.731910228729248 + }, + { + "auxiliary_loss_clip": 0.01154121, + "auxiliary_loss_mlp": 0.01029868, + "balance_loss_clip": 1.01553106, + "balance_loss_mlp": 1.02181542, + "epoch": 0.7023387242229303, + "flos": 19790514581760.0, + "grad_norm": 2.7472488307158827, + "language_loss": 0.71869791, + "learning_rate": 8.594856109190194e-07, + "loss": 0.74053776, + "num_input_tokens_seen": 125680125, + "step": 5841, + "time_per_iteration": 2.643199920654297 + }, + { + "auxiliary_loss_clip": 0.01170413, + "auxiliary_loss_mlp": 0.01029345, + "balance_loss_clip": 1.0512569, + "balance_loss_mlp": 1.0216856, + "epoch": 0.7024589671135694, + "flos": 33259278286080.0, + "grad_norm": 1.8129023979038243, + "language_loss": 0.69343972, + "learning_rate": 8.588457985569446e-07, + "loss": 0.71543735, + "num_input_tokens_seen": 125703035, + "step": 5842, + "time_per_iteration": 2.754035234451294 + }, + { + "auxiliary_loss_clip": 0.01171017, + "auxiliary_loss_mlp": 0.01026954, + "balance_loss_clip": 1.05121768, + "balance_loss_mlp": 1.01943839, + "epoch": 0.7025792100042085, + "flos": 19098982967040.0, + "grad_norm": 1.9899960597833122, + "language_loss": 0.72029924, + "learning_rate": 8.582061593070542e-07, + "loss": 0.74227893, + "num_input_tokens_seen": 125723765, + "step": 5843, + "time_per_iteration": 2.6148502826690674 + }, + { + "auxiliary_loss_clip": 0.01166766, + "auxiliary_loss_mlp": 0.00509969, + "balance_loss_clip": 1.04911304, + "balance_loss_mlp": 0.0, + "epoch": 0.7026994528948476, + "flos": 18952611045120.0, + "grad_norm": 2.1333977965517708, + "language_loss": 0.77033639, + "learning_rate": 8.57566693266383e-07, + "loss": 0.78710377, + "num_input_tokens_seen": 125741455, + "step": 5844, + "time_per_iteration": 2.5851783752441406 + }, + { + "auxiliary_loss_clip": 0.01124016, + "auxiliary_loss_mlp": 0.00508767, + "balance_loss_clip": 0.97072285, + "balance_loss_mlp": 0.0, + "epoch": 0.7028196957854866, + "flos": 19536662188800.0, + "grad_norm": 2.908981428684474, + "language_loss": 0.69173181, + "learning_rate": 8.569274005319354e-07, + "loss": 0.70805961, + "num_input_tokens_seen": 125759855, + "step": 5845, + "time_per_iteration": 2.7986693382263184 + }, + { + "auxiliary_loss_clip": 0.01138831, + "auxiliary_loss_mlp": 0.01022909, + "balance_loss_clip": 1.00982404, + "balance_loss_mlp": 1.01563764, + "epoch": 0.7029399386761258, + "flos": 20845318394880.0, + "grad_norm": 1.705712283198472, + "language_loss": 0.79605883, + "learning_rate": 8.562882812006913e-07, + "loss": 0.81767619, + "num_input_tokens_seen": 125777345, + "step": 5846, + "time_per_iteration": 3.5551443099975586 + }, + { + "auxiliary_loss_clip": 0.01166237, + "auxiliary_loss_mlp": 0.01027519, + "balance_loss_clip": 1.04978359, + "balance_loss_mlp": 1.01979995, + "epoch": 0.7030601815667649, + "flos": 22055005653120.0, + "grad_norm": 1.8010955424708337, + "language_loss": 0.77798104, + "learning_rate": 8.556493353696066e-07, + "loss": 0.79991859, + "num_input_tokens_seen": 125796345, + "step": 5847, + "time_per_iteration": 2.711244583129883 + }, + { + "auxiliary_loss_clip": 0.01147756, + "auxiliary_loss_mlp": 0.00510147, + "balance_loss_clip": 1.0123117, + "balance_loss_mlp": 0.0, + "epoch": 0.7031804244574039, + "flos": 27198742089600.0, + "grad_norm": 1.9697416613238379, + "language_loss": 0.67997253, + "learning_rate": 8.550105631356077e-07, + "loss": 0.6965515, + "num_input_tokens_seen": 125816070, + "step": 5848, + "time_per_iteration": 2.6865789890289307 + }, + { + "auxiliary_loss_clip": 0.01092933, + "auxiliary_loss_mlp": 0.01024188, + "balance_loss_clip": 0.93297398, + "balance_loss_mlp": 1.01655304, + "epoch": 0.7033006673480431, + "flos": 22379853277440.0, + "grad_norm": 1.9438306948078363, + "language_loss": 0.77247667, + "learning_rate": 8.543719645955961e-07, + "loss": 0.79364783, + "num_input_tokens_seen": 125834400, + "step": 5849, + "time_per_iteration": 2.7457468509674072 + }, + { + "auxiliary_loss_clip": 0.01122867, + "auxiliary_loss_mlp": 0.01024562, + "balance_loss_clip": 0.97263277, + "balance_loss_mlp": 1.01723099, + "epoch": 0.7034209102386821, + "flos": 24715986024960.0, + "grad_norm": 2.899748428000443, + "language_loss": 0.74489975, + "learning_rate": 8.537335398464467e-07, + "loss": 0.76637405, + "num_input_tokens_seen": 125854720, + "step": 5850, + "time_per_iteration": 2.714414358139038 + }, + { + "auxiliary_loss_clip": 0.01120979, + "auxiliary_loss_mlp": 0.01031883, + "balance_loss_clip": 0.97076893, + "balance_loss_mlp": 1.02422428, + "epoch": 0.7035411531293212, + "flos": 22556174163840.0, + "grad_norm": 2.545575841938487, + "language_loss": 0.8583234, + "learning_rate": 8.53095288985007e-07, + "loss": 0.87985206, + "num_input_tokens_seen": 125868455, + "step": 5851, + "time_per_iteration": 2.6741631031036377 + }, + { + "auxiliary_loss_clip": 0.01165883, + "auxiliary_loss_mlp": 0.0102544, + "balance_loss_clip": 1.05018985, + "balance_loss_mlp": 1.01819229, + "epoch": 0.7036613960199604, + "flos": 22674967418880.0, + "grad_norm": 1.6050909161780047, + "language_loss": 0.82305592, + "learning_rate": 8.524572121081009e-07, + "loss": 0.84496915, + "num_input_tokens_seen": 125888555, + "step": 5852, + "time_per_iteration": 2.6267168521881104 + }, + { + "auxiliary_loss_clip": 0.01148054, + "auxiliary_loss_mlp": 0.0102426, + "balance_loss_clip": 1.01098347, + "balance_loss_mlp": 1.01722074, + "epoch": 0.7037816389105994, + "flos": 22492146170880.0, + "grad_norm": 1.9069171453572722, + "language_loss": 0.62500226, + "learning_rate": 8.518193093125232e-07, + "loss": 0.64672542, + "num_input_tokens_seen": 125907610, + "step": 5853, + "time_per_iteration": 2.7054200172424316 + }, + { + "auxiliary_loss_clip": 0.01129156, + "auxiliary_loss_mlp": 0.01027924, + "balance_loss_clip": 0.97664523, + "balance_loss_mlp": 1.02087605, + "epoch": 0.7039018818012385, + "flos": 27087490690560.0, + "grad_norm": 2.397286367662183, + "language_loss": 0.81112289, + "learning_rate": 8.511815806950436e-07, + "loss": 0.8326937, + "num_input_tokens_seen": 125928640, + "step": 5854, + "time_per_iteration": 2.6921536922454834 + }, + { + "auxiliary_loss_clip": 0.0114194, + "auxiliary_loss_mlp": 0.01024513, + "balance_loss_clip": 1.00851655, + "balance_loss_mlp": 1.01755106, + "epoch": 0.7040221246918776, + "flos": 17749819198080.0, + "grad_norm": 2.0395786091440056, + "language_loss": 0.78148437, + "learning_rate": 8.505440263524044e-07, + "loss": 0.80314887, + "num_input_tokens_seen": 125947485, + "step": 5855, + "time_per_iteration": 2.640496253967285 + }, + { + "auxiliary_loss_clip": 0.01148161, + "auxiliary_loss_mlp": 0.01031205, + "balance_loss_clip": 1.01116836, + "balance_loss_mlp": 1.02348042, + "epoch": 0.7041423675825167, + "flos": 16279851012480.0, + "grad_norm": 2.525786133237028, + "language_loss": 0.88457716, + "learning_rate": 8.49906646381322e-07, + "loss": 0.90637076, + "num_input_tokens_seen": 125960320, + "step": 5856, + "time_per_iteration": 2.584827184677124 + }, + { + "auxiliary_loss_clip": 0.01101186, + "auxiliary_loss_mlp": 0.01025803, + "balance_loss_clip": 0.93688047, + "balance_loss_mlp": 1.01905036, + "epoch": 0.7042626104731557, + "flos": 25483181639040.0, + "grad_norm": 1.7421437770505153, + "language_loss": 0.72187054, + "learning_rate": 8.492694408784884e-07, + "loss": 0.7431404, + "num_input_tokens_seen": 125980575, + "step": 5857, + "time_per_iteration": 2.7084929943084717 + }, + { + "auxiliary_loss_clip": 0.01146429, + "auxiliary_loss_mlp": 0.01034445, + "balance_loss_clip": 1.01114273, + "balance_loss_mlp": 1.0273726, + "epoch": 0.7043828533637949, + "flos": 17857622891520.0, + "grad_norm": 2.247772770986984, + "language_loss": 0.62074685, + "learning_rate": 8.486324099405642e-07, + "loss": 0.64255559, + "num_input_tokens_seen": 125997420, + "step": 5858, + "time_per_iteration": 2.6357710361480713 + }, + { + "auxiliary_loss_clip": 0.01142345, + "auxiliary_loss_mlp": 0.01021752, + "balance_loss_clip": 1.01005304, + "balance_loss_mlp": 1.01449811, + "epoch": 0.704503096254434, + "flos": 29494259533440.0, + "grad_norm": 1.618507895394773, + "language_loss": 0.74711382, + "learning_rate": 8.479955536641887e-07, + "loss": 0.76875484, + "num_input_tokens_seen": 126018915, + "step": 5859, + "time_per_iteration": 2.7970781326293945 + }, + { + "auxiliary_loss_clip": 0.01113007, + "auxiliary_loss_mlp": 0.01030766, + "balance_loss_clip": 0.96786129, + "balance_loss_mlp": 1.02310634, + "epoch": 0.704623339145073, + "flos": 30920739327360.0, + "grad_norm": 1.8839195234121628, + "language_loss": 0.66495281, + "learning_rate": 8.473588721459716e-07, + "loss": 0.68639052, + "num_input_tokens_seen": 126038825, + "step": 5860, + "time_per_iteration": 3.7015814781188965 + }, + { + "auxiliary_loss_clip": 0.011454, + "auxiliary_loss_mlp": 0.01027367, + "balance_loss_clip": 1.01377213, + "balance_loss_mlp": 1.01883793, + "epoch": 0.7047435820357122, + "flos": 23914747296000.0, + "grad_norm": 2.2344398105827765, + "language_loss": 0.70644444, + "learning_rate": 8.467223654824967e-07, + "loss": 0.72817212, + "num_input_tokens_seen": 126058280, + "step": 5861, + "time_per_iteration": 3.5866408348083496 + }, + { + "auxiliary_loss_clip": 0.0114006, + "auxiliary_loss_mlp": 0.01032481, + "balance_loss_clip": 1.01047659, + "balance_loss_mlp": 1.02442586, + "epoch": 0.7048638249263512, + "flos": 46494010926720.0, + "grad_norm": 2.062543471283196, + "language_loss": 0.62395859, + "learning_rate": 8.460860337703233e-07, + "loss": 0.645684, + "num_input_tokens_seen": 126078885, + "step": 5862, + "time_per_iteration": 2.8383805751800537 + }, + { + "auxiliary_loss_clip": 0.01085309, + "auxiliary_loss_mlp": 0.01026531, + "balance_loss_clip": 0.93257755, + "balance_loss_mlp": 1.01857972, + "epoch": 0.7049840678169903, + "flos": 21689219502720.0, + "grad_norm": 1.9768541362456133, + "language_loss": 0.7073282, + "learning_rate": 8.454498771059797e-07, + "loss": 0.7284466, + "num_input_tokens_seen": 126098260, + "step": 5863, + "time_per_iteration": 2.7206313610076904 + }, + { + "auxiliary_loss_clip": 0.01068226, + "auxiliary_loss_mlp": 0.01028858, + "balance_loss_clip": 0.89574146, + "balance_loss_mlp": 1.02150273, + "epoch": 0.7051043107076294, + "flos": 18405081054720.0, + "grad_norm": 2.097545102998374, + "language_loss": 0.83385408, + "learning_rate": 8.448138955859725e-07, + "loss": 0.8548249, + "num_input_tokens_seen": 126114845, + "step": 5864, + "time_per_iteration": 3.699817657470703 + }, + { + "auxiliary_loss_clip": 0.01122133, + "auxiliary_loss_mlp": 0.010336, + "balance_loss_clip": 0.97332507, + "balance_loss_mlp": 1.02621794, + "epoch": 0.7052245535982685, + "flos": 19319043640320.0, + "grad_norm": 1.8739912788446025, + "language_loss": 0.89883822, + "learning_rate": 8.44178089306778e-07, + "loss": 0.92039555, + "num_input_tokens_seen": 126132780, + "step": 5865, + "time_per_iteration": 2.6671953201293945 + }, + { + "auxiliary_loss_clip": 0.01168117, + "auxiliary_loss_mlp": 0.01027167, + "balance_loss_clip": 1.05122137, + "balance_loss_mlp": 1.02032721, + "epoch": 0.7053447964889076, + "flos": 19062138591360.0, + "grad_norm": 1.7785591836202848, + "language_loss": 0.76852369, + "learning_rate": 8.4354245836485e-07, + "loss": 0.79047656, + "num_input_tokens_seen": 126151225, + "step": 5866, + "time_per_iteration": 2.5848569869995117 + }, + { + "auxiliary_loss_clip": 0.01102404, + "auxiliary_loss_mlp": 0.01027188, + "balance_loss_clip": 0.93698096, + "balance_loss_mlp": 1.01852155, + "epoch": 0.7054650393795466, + "flos": 27379228953600.0, + "grad_norm": 2.6048457779559873, + "language_loss": 0.72946632, + "learning_rate": 8.429070028566108e-07, + "loss": 0.75076222, + "num_input_tokens_seen": 126172535, + "step": 5867, + "time_per_iteration": 2.760528564453125 + }, + { + "auxiliary_loss_clip": 0.01142974, + "auxiliary_loss_mlp": 0.01025211, + "balance_loss_clip": 1.01121199, + "balance_loss_mlp": 1.01795077, + "epoch": 0.7055852822701858, + "flos": 16102201322880.0, + "grad_norm": 1.964305112119734, + "language_loss": 0.74719292, + "learning_rate": 8.422717228784586e-07, + "loss": 0.76887476, + "num_input_tokens_seen": 126189410, + "step": 5868, + "time_per_iteration": 2.6383302211761475 + }, + { + "auxiliary_loss_clip": 0.01073064, + "auxiliary_loss_mlp": 0.01027201, + "balance_loss_clip": 0.89872617, + "balance_loss_mlp": 1.01982796, + "epoch": 0.7057055251608249, + "flos": 11692299744000.0, + "grad_norm": 4.152856092250974, + "language_loss": 0.69111729, + "learning_rate": 8.416366185267663e-07, + "loss": 0.71211994, + "num_input_tokens_seen": 126206910, + "step": 5869, + "time_per_iteration": 2.743861675262451 + }, + { + "auxiliary_loss_clip": 0.01142197, + "auxiliary_loss_mlp": 0.01026167, + "balance_loss_clip": 1.00847375, + "balance_loss_mlp": 1.0185256, + "epoch": 0.7058257680514639, + "flos": 22711560399360.0, + "grad_norm": 2.0030951414106455, + "language_loss": 0.77842265, + "learning_rate": 8.410016898978778e-07, + "loss": 0.80010629, + "num_input_tokens_seen": 126224385, + "step": 5870, + "time_per_iteration": 2.6875579357147217 + }, + { + "auxiliary_loss_clip": 0.01079127, + "auxiliary_loss_mlp": 0.01025557, + "balance_loss_clip": 0.90281558, + "balance_loss_mlp": 1.01852036, + "epoch": 0.7059460109421031, + "flos": 17529543043200.0, + "grad_norm": 1.6932183755848804, + "language_loss": 0.7903384, + "learning_rate": 8.403669370881115e-07, + "loss": 0.81138527, + "num_input_tokens_seen": 126243120, + "step": 5871, + "time_per_iteration": 2.741325616836548 + }, + { + "auxiliary_loss_clip": 0.0116995, + "auxiliary_loss_mlp": 0.01027187, + "balance_loss_clip": 1.05232418, + "balance_loss_mlp": 1.01984942, + "epoch": 0.7060662538327421, + "flos": 23544687427200.0, + "grad_norm": 1.570629540342994, + "language_loss": 0.78576303, + "learning_rate": 8.397323601937587e-07, + "loss": 0.80773437, + "num_input_tokens_seen": 126263020, + "step": 5872, + "time_per_iteration": 3.5503029823303223 + }, + { + "auxiliary_loss_clip": 0.01092169, + "auxiliary_loss_mlp": 0.01021878, + "balance_loss_clip": 0.93386757, + "balance_loss_mlp": 1.01433229, + "epoch": 0.7061864967233812, + "flos": 30260736875520.0, + "grad_norm": 1.9321544793272913, + "language_loss": 0.7711311, + "learning_rate": 8.390979593110838e-07, + "loss": 0.79227155, + "num_input_tokens_seen": 126285150, + "step": 5873, + "time_per_iteration": 2.7559919357299805 + }, + { + "auxiliary_loss_clip": 0.01126039, + "auxiliary_loss_mlp": 0.01025409, + "balance_loss_clip": 0.97482765, + "balance_loss_mlp": 1.01779175, + "epoch": 0.7063067396140204, + "flos": 20701460424960.0, + "grad_norm": 1.6580117819815063, + "language_loss": 0.81635404, + "learning_rate": 8.384637345363262e-07, + "loss": 0.83786845, + "num_input_tokens_seen": 126304340, + "step": 5874, + "time_per_iteration": 2.7167928218841553 + }, + { + "auxiliary_loss_clip": 0.01112327, + "auxiliary_loss_mlp": 0.0102492, + "balance_loss_clip": 0.96723962, + "balance_loss_mlp": 1.01741004, + "epoch": 0.7064269825046594, + "flos": 32266168081920.0, + "grad_norm": 1.7077371153150784, + "language_loss": 0.7637853, + "learning_rate": 8.378296859656964e-07, + "loss": 0.7851578, + "num_input_tokens_seen": 126325495, + "step": 5875, + "time_per_iteration": 2.7793996334075928 + }, + { + "auxiliary_loss_clip": 0.01125539, + "auxiliary_loss_mlp": 0.01028587, + "balance_loss_clip": 0.97543591, + "balance_loss_mlp": 1.0214045, + "epoch": 0.7065472253952985, + "flos": 30227124723840.0, + "grad_norm": 2.269703862057238, + "language_loss": 0.68578124, + "learning_rate": 8.371958136953792e-07, + "loss": 0.70732248, + "num_input_tokens_seen": 126345525, + "step": 5876, + "time_per_iteration": 2.759709119796753 + }, + { + "auxiliary_loss_clip": 0.01101929, + "auxiliary_loss_mlp": 0.0102841, + "balance_loss_clip": 0.9330374, + "balance_loss_mlp": 1.02064943, + "epoch": 0.7066674682859376, + "flos": 16216720859520.0, + "grad_norm": 3.4139302900135475, + "language_loss": 0.66515344, + "learning_rate": 8.365621178215326e-07, + "loss": 0.6864568, + "num_input_tokens_seen": 126361995, + "step": 5877, + "time_per_iteration": 2.706087589263916 + }, + { + "auxiliary_loss_clip": 0.01139879, + "auxiliary_loss_mlp": 0.01022932, + "balance_loss_clip": 1.01044106, + "balance_loss_mlp": 1.0150578, + "epoch": 0.7067877111765767, + "flos": 14830461319680.0, + "grad_norm": 2.8750445146048587, + "language_loss": 0.75408536, + "learning_rate": 8.359285984402871e-07, + "loss": 0.77571344, + "num_input_tokens_seen": 126379260, + "step": 5878, + "time_per_iteration": 2.6692819595336914 + }, + { + "auxiliary_loss_clip": 0.01118489, + "auxiliary_loss_mlp": 0.01026662, + "balance_loss_clip": 0.97422349, + "balance_loss_mlp": 1.01946819, + "epoch": 0.7069079540672157, + "flos": 25440196037760.0, + "grad_norm": 1.9079273016366405, + "language_loss": 0.74037683, + "learning_rate": 8.352952556477489e-07, + "loss": 0.76182842, + "num_input_tokens_seen": 126397170, + "step": 5879, + "time_per_iteration": 2.736116647720337 + }, + { + "auxiliary_loss_clip": 0.01145092, + "auxiliary_loss_mlp": 0.0102475, + "balance_loss_clip": 1.01282835, + "balance_loss_mlp": 1.01734734, + "epoch": 0.7070281969578549, + "flos": 24607751368320.0, + "grad_norm": 2.0311751233333717, + "language_loss": 0.76392663, + "learning_rate": 8.34662089539993e-07, + "loss": 0.78562504, + "num_input_tokens_seen": 126416680, + "step": 5880, + "time_per_iteration": 2.7097980976104736 + }, + { + "auxiliary_loss_clip": 0.01167646, + "auxiliary_loss_mlp": 0.01033315, + "balance_loss_clip": 1.05117106, + "balance_loss_mlp": 1.02636528, + "epoch": 0.707148439848494, + "flos": 26724469887360.0, + "grad_norm": 1.9680961058147233, + "language_loss": 0.79501361, + "learning_rate": 8.340291002130722e-07, + "loss": 0.81702322, + "num_input_tokens_seen": 126435870, + "step": 5881, + "time_per_iteration": 2.7242319583892822 + }, + { + "auxiliary_loss_clip": 0.01171366, + "auxiliary_loss_mlp": 0.01025456, + "balance_loss_clip": 1.05167484, + "balance_loss_mlp": 1.01776695, + "epoch": 0.707268682739133, + "flos": 15085750256640.0, + "grad_norm": 2.1702169210171984, + "language_loss": 0.80048764, + "learning_rate": 8.3339628776301e-07, + "loss": 0.82245588, + "num_input_tokens_seen": 126454010, + "step": 5882, + "time_per_iteration": 2.61369252204895 + }, + { + "auxiliary_loss_clip": 0.01169172, + "auxiliary_loss_mlp": 0.01025749, + "balance_loss_clip": 1.0512538, + "balance_loss_mlp": 1.01853132, + "epoch": 0.7073889256297722, + "flos": 34313148345600.0, + "grad_norm": 1.8390405611552207, + "language_loss": 0.57246935, + "learning_rate": 8.327636522858033e-07, + "loss": 0.59441864, + "num_input_tokens_seen": 126473615, + "step": 5883, + "time_per_iteration": 2.722869873046875 + }, + { + "auxiliary_loss_clip": 0.01076466, + "auxiliary_loss_mlp": 0.01032681, + "balance_loss_clip": 0.89956105, + "balance_loss_mlp": 1.02547216, + "epoch": 0.7075091685204112, + "flos": 20083940784000.0, + "grad_norm": 1.809334665120789, + "language_loss": 0.76981628, + "learning_rate": 8.321311938774225e-07, + "loss": 0.79090774, + "num_input_tokens_seen": 126492705, + "step": 5884, + "time_per_iteration": 2.7114884853363037 + }, + { + "auxiliary_loss_clip": 0.01172353, + "auxiliary_loss_mlp": 0.01026482, + "balance_loss_clip": 1.05170369, + "balance_loss_mlp": 1.01880479, + "epoch": 0.7076294114110503, + "flos": 20777124424320.0, + "grad_norm": 2.202487555999011, + "language_loss": 0.79178953, + "learning_rate": 8.314989126338104e-07, + "loss": 0.81377786, + "num_input_tokens_seen": 126512715, + "step": 5885, + "time_per_iteration": 2.6114654541015625 + }, + { + "auxiliary_loss_clip": 0.01146027, + "auxiliary_loss_mlp": 0.01025891, + "balance_loss_clip": 1.01095569, + "balance_loss_mlp": 1.01826465, + "epoch": 0.7077496543016895, + "flos": 17967689141760.0, + "grad_norm": 1.7473912916709236, + "language_loss": 0.84549248, + "learning_rate": 8.308668086508847e-07, + "loss": 0.86721158, + "num_input_tokens_seen": 126530795, + "step": 5886, + "time_per_iteration": 3.506570339202881 + }, + { + "auxiliary_loss_clip": 0.01097456, + "auxiliary_loss_mlp": 0.01020709, + "balance_loss_clip": 0.93106383, + "balance_loss_mlp": 1.01279378, + "epoch": 0.7078698971923285, + "flos": 45478098564480.0, + "grad_norm": 1.9921385688550863, + "language_loss": 0.73653769, + "learning_rate": 8.302348820245342e-07, + "loss": 0.7577194, + "num_input_tokens_seen": 126553360, + "step": 5887, + "time_per_iteration": 2.954531192779541 + }, + { + "auxiliary_loss_clip": 0.01099226, + "auxiliary_loss_mlp": 0.01020807, + "balance_loss_clip": 0.9351877, + "balance_loss_mlp": 1.01271868, + "epoch": 0.7079901400829676, + "flos": 26943704547840.0, + "grad_norm": 2.1733335864803136, + "language_loss": 0.70045644, + "learning_rate": 8.296031328506232e-07, + "loss": 0.72165674, + "num_input_tokens_seen": 126573110, + "step": 5888, + "time_per_iteration": 3.6522414684295654 + }, + { + "auxiliary_loss_clip": 0.01123172, + "auxiliary_loss_mlp": 0.01028934, + "balance_loss_clip": 0.97341657, + "balance_loss_mlp": 1.02125072, + "epoch": 0.7081103829736067, + "flos": 24423206267520.0, + "grad_norm": 1.6765506037973448, + "language_loss": 0.75710779, + "learning_rate": 8.289715612249857e-07, + "loss": 0.77862883, + "num_input_tokens_seen": 126593725, + "step": 5889, + "time_per_iteration": 2.7805018424987793 + }, + { + "auxiliary_loss_clip": 0.0111697, + "auxiliary_loss_mlp": 0.01033577, + "balance_loss_clip": 0.97219932, + "balance_loss_mlp": 1.02617097, + "epoch": 0.7082306258642458, + "flos": 18543300589440.0, + "grad_norm": 2.646775416377299, + "language_loss": 0.77799094, + "learning_rate": 8.283401672434305e-07, + "loss": 0.79949647, + "num_input_tokens_seen": 126608950, + "step": 5890, + "time_per_iteration": 3.66072940826416 + }, + { + "auxiliary_loss_clip": 0.01119763, + "auxiliary_loss_mlp": 0.01021457, + "balance_loss_clip": 0.97547311, + "balance_loss_mlp": 1.01427221, + "epoch": 0.7083508687548848, + "flos": 23477534951040.0, + "grad_norm": 2.6426615619542715, + "language_loss": 0.70405626, + "learning_rate": 8.277089510017412e-07, + "loss": 0.72546846, + "num_input_tokens_seen": 126629755, + "step": 5891, + "time_per_iteration": 2.810634136199951 + }, + { + "auxiliary_loss_clip": 0.01118367, + "auxiliary_loss_mlp": 0.01024927, + "balance_loss_clip": 0.97492826, + "balance_loss_mlp": 1.0176667, + "epoch": 0.708471111645524, + "flos": 22419463000320.0, + "grad_norm": 2.135473547753215, + "language_loss": 0.82474625, + "learning_rate": 8.270779125956719e-07, + "loss": 0.84617913, + "num_input_tokens_seen": 126650135, + "step": 5892, + "time_per_iteration": 2.679727554321289 + }, + { + "auxiliary_loss_clip": 0.01072298, + "auxiliary_loss_mlp": 0.01022553, + "balance_loss_clip": 0.89741719, + "balance_loss_mlp": 1.01500654, + "epoch": 0.7085913545361631, + "flos": 20922885815040.0, + "grad_norm": 2.060742763808379, + "language_loss": 0.79707021, + "learning_rate": 8.264470521209505e-07, + "loss": 0.81801879, + "num_input_tokens_seen": 126668500, + "step": 5893, + "time_per_iteration": 2.7313287258148193 + }, + { + "auxiliary_loss_clip": 0.01138937, + "auxiliary_loss_mlp": 0.01025923, + "balance_loss_clip": 1.01020992, + "balance_loss_mlp": 1.01846027, + "epoch": 0.7087115974268021, + "flos": 15012384727680.0, + "grad_norm": 2.248187231568006, + "language_loss": 0.76364458, + "learning_rate": 8.258163696732785e-07, + "loss": 0.7852931, + "num_input_tokens_seen": 126686090, + "step": 5894, + "time_per_iteration": 2.6244351863861084 + }, + { + "auxiliary_loss_clip": 0.01140585, + "auxiliary_loss_mlp": 0.01024893, + "balance_loss_clip": 1.01109052, + "balance_loss_mlp": 1.01731086, + "epoch": 0.7088318403174413, + "flos": 21539040739200.0, + "grad_norm": 2.0802911132339403, + "language_loss": 0.77086723, + "learning_rate": 8.251858653483288e-07, + "loss": 0.79252201, + "num_input_tokens_seen": 126704255, + "step": 5895, + "time_per_iteration": 2.6798245906829834 + }, + { + "auxiliary_loss_clip": 0.01147573, + "auxiliary_loss_mlp": 0.01029737, + "balance_loss_clip": 1.01410913, + "balance_loss_mlp": 1.0218935, + "epoch": 0.7089520832080803, + "flos": 15516785462400.0, + "grad_norm": 2.193230048537344, + "language_loss": 0.85645187, + "learning_rate": 8.245555392417501e-07, + "loss": 0.87822497, + "num_input_tokens_seen": 126718910, + "step": 5896, + "time_per_iteration": 2.586038112640381 + }, + { + "auxiliary_loss_clip": 0.01062917, + "auxiliary_loss_mlp": 0.01025379, + "balance_loss_clip": 0.8931784, + "balance_loss_mlp": 1.01779771, + "epoch": 0.7090723260987194, + "flos": 20412667077120.0, + "grad_norm": 1.9260940278007246, + "language_loss": 0.79064107, + "learning_rate": 8.239253914491613e-07, + "loss": 0.81152403, + "num_input_tokens_seen": 126737235, + "step": 5897, + "time_per_iteration": 2.745985984802246 + }, + { + "auxiliary_loss_clip": 0.01094214, + "auxiliary_loss_mlp": 0.01028284, + "balance_loss_clip": 0.93525052, + "balance_loss_mlp": 1.02085471, + "epoch": 0.7091925689893585, + "flos": 25668337271040.0, + "grad_norm": 1.8081947327202448, + "language_loss": 0.75080246, + "learning_rate": 8.232954220661556e-07, + "loss": 0.77202743, + "num_input_tokens_seen": 126759970, + "step": 5898, + "time_per_iteration": 3.72884202003479 + }, + { + "auxiliary_loss_clip": 0.01168717, + "auxiliary_loss_mlp": 0.01025051, + "balance_loss_clip": 1.05239475, + "balance_loss_mlp": 1.01760602, + "epoch": 0.7093128118799976, + "flos": 24206629213440.0, + "grad_norm": 2.4487119924873966, + "language_loss": 0.70990157, + "learning_rate": 8.226656311882989e-07, + "loss": 0.7318393, + "num_input_tokens_seen": 126779280, + "step": 5899, + "time_per_iteration": 2.635096549987793 + }, + { + "auxiliary_loss_clip": 0.01145067, + "auxiliary_loss_mlp": 0.01024424, + "balance_loss_clip": 1.0142746, + "balance_loss_mlp": 1.01777852, + "epoch": 0.7094330547706367, + "flos": 16646786398080.0, + "grad_norm": 2.1462534018071424, + "language_loss": 0.76679349, + "learning_rate": 8.22036018911129e-07, + "loss": 0.78848845, + "num_input_tokens_seen": 126797310, + "step": 5900, + "time_per_iteration": 2.639052629470825 + }, + { + "auxiliary_loss_clip": 0.011714, + "auxiliary_loss_mlp": 0.0102596, + "balance_loss_clip": 1.04970098, + "balance_loss_mlp": 1.01835418, + "epoch": 0.7095532976612757, + "flos": 16283370545280.0, + "grad_norm": 2.40671685438497, + "language_loss": 0.80697775, + "learning_rate": 8.214065853301599e-07, + "loss": 0.82895136, + "num_input_tokens_seen": 126812840, + "step": 5901, + "time_per_iteration": 2.5950334072113037 + }, + { + "auxiliary_loss_clip": 0.01049643, + "auxiliary_loss_mlp": 0.01002262, + "balance_loss_clip": 0.97745538, + "balance_loss_mlp": 1.00129056, + "epoch": 0.7096735405519149, + "flos": 70722080559360.0, + "grad_norm": 0.7905896881601506, + "language_loss": 0.58188999, + "learning_rate": 8.207773305408734e-07, + "loss": 0.60240901, + "num_input_tokens_seen": 126880060, + "step": 5902, + "time_per_iteration": 3.374000072479248 + }, + { + "auxiliary_loss_clip": 0.01081667, + "auxiliary_loss_mlp": 0.01024222, + "balance_loss_clip": 0.89618158, + "balance_loss_mlp": 1.01686716, + "epoch": 0.709793783442554, + "flos": 23621500661760.0, + "grad_norm": 2.5858486309319653, + "language_loss": 0.7973423, + "learning_rate": 8.201482546387288e-07, + "loss": 0.81840122, + "num_input_tokens_seen": 126899535, + "step": 5903, + "time_per_iteration": 2.814220905303955 + }, + { + "auxiliary_loss_clip": 0.01144487, + "auxiliary_loss_mlp": 0.01027567, + "balance_loss_clip": 1.01276469, + "balance_loss_mlp": 1.02005124, + "epoch": 0.709914026333193, + "flos": 25993472204160.0, + "grad_norm": 1.8237161584684178, + "language_loss": 0.91648513, + "learning_rate": 8.195193577191553e-07, + "loss": 0.9382056, + "num_input_tokens_seen": 126921365, + "step": 5904, + "time_per_iteration": 2.7219696044921875 + }, + { + "auxiliary_loss_clip": 0.01128906, + "auxiliary_loss_mlp": 0.0050949, + "balance_loss_clip": 0.97337615, + "balance_loss_mlp": 0.0, + "epoch": 0.7100342692238322, + "flos": 24861531934080.0, + "grad_norm": 1.5907498380523208, + "language_loss": 0.8433342, + "learning_rate": 8.188906398775579e-07, + "loss": 0.8597182, + "num_input_tokens_seen": 126941910, + "step": 5905, + "time_per_iteration": 2.7499284744262695 + }, + { + "auxiliary_loss_clip": 0.01168608, + "auxiliary_loss_mlp": 0.00510105, + "balance_loss_clip": 1.04864991, + "balance_loss_mlp": 0.0, + "epoch": 0.7101545121144712, + "flos": 24932203943040.0, + "grad_norm": 1.7843857003166739, + "language_loss": 0.68991041, + "learning_rate": 8.18262101209311e-07, + "loss": 0.70669758, + "num_input_tokens_seen": 126961120, + "step": 5906, + "time_per_iteration": 2.6959989070892334 + }, + { + "auxiliary_loss_clip": 0.0114848, + "auxiliary_loss_mlp": 0.01025944, + "balance_loss_clip": 1.01199841, + "balance_loss_mlp": 1.01833844, + "epoch": 0.7102747550051103, + "flos": 23768842250880.0, + "grad_norm": 1.7504914844591808, + "language_loss": 0.70107663, + "learning_rate": 8.176337418097626e-07, + "loss": 0.72282088, + "num_input_tokens_seen": 126981590, + "step": 5907, + "time_per_iteration": 2.6844875812530518 + }, + { + "auxiliary_loss_clip": 0.01142696, + "auxiliary_loss_mlp": 0.0050984, + "balance_loss_clip": 1.01232648, + "balance_loss_mlp": 0.0, + "epoch": 0.7103949978957494, + "flos": 15303907509120.0, + "grad_norm": 2.1400441450213625, + "language_loss": 0.8000772, + "learning_rate": 8.170055617742364e-07, + "loss": 0.81660259, + "num_input_tokens_seen": 126998870, + "step": 5908, + "time_per_iteration": 2.71073579788208 + }, + { + "auxiliary_loss_clip": 0.01116738, + "auxiliary_loss_mlp": 0.01023733, + "balance_loss_clip": 0.97348619, + "balance_loss_mlp": 1.01635408, + "epoch": 0.7105152407863885, + "flos": 22638805401600.0, + "grad_norm": 2.340663605141524, + "language_loss": 0.70611495, + "learning_rate": 8.163775611980252e-07, + "loss": 0.72751963, + "num_input_tokens_seen": 127017980, + "step": 5909, + "time_per_iteration": 2.7212729454040527 + }, + { + "auxiliary_loss_clip": 0.01120036, + "auxiliary_loss_mlp": 0.01031022, + "balance_loss_clip": 0.97347075, + "balance_loss_mlp": 1.0236491, + "epoch": 0.7106354836770276, + "flos": 17238594879360.0, + "grad_norm": 1.6689256483188202, + "language_loss": 0.78679979, + "learning_rate": 8.157497401763982e-07, + "loss": 0.80831039, + "num_input_tokens_seen": 127035645, + "step": 5910, + "time_per_iteration": 2.676495313644409 + }, + { + "auxiliary_loss_clip": 0.01143225, + "auxiliary_loss_mlp": 0.01032586, + "balance_loss_clip": 1.01198769, + "balance_loss_mlp": 1.02535009, + "epoch": 0.7107557265676667, + "flos": 20193647898240.0, + "grad_norm": 1.7955607328302645, + "language_loss": 0.77972424, + "learning_rate": 8.151220988045935e-07, + "loss": 0.80148238, + "num_input_tokens_seen": 127054900, + "step": 5911, + "time_per_iteration": 2.7234528064727783 + }, + { + "auxiliary_loss_clip": 0.01142824, + "auxiliary_loss_mlp": 0.01030868, + "balance_loss_clip": 1.01114094, + "balance_loss_mlp": 1.02360225, + "epoch": 0.7108759694583058, + "flos": 21507080613120.0, + "grad_norm": 1.7238656618427324, + "language_loss": 0.83146828, + "learning_rate": 8.144946371778234e-07, + "loss": 0.8532052, + "num_input_tokens_seen": 127075010, + "step": 5912, + "time_per_iteration": 3.531931161880493 + }, + { + "auxiliary_loss_clip": 0.01122466, + "auxiliary_loss_mlp": 0.00509933, + "balance_loss_clip": 0.97466713, + "balance_loss_mlp": 0.0, + "epoch": 0.7109962123489448, + "flos": 24061909317120.0, + "grad_norm": 7.199046898203354, + "language_loss": 0.78508401, + "learning_rate": 8.138673553912751e-07, + "loss": 0.80140799, + "num_input_tokens_seen": 127095570, + "step": 5913, + "time_per_iteration": 2.7735705375671387 + }, + { + "auxiliary_loss_clip": 0.01075414, + "auxiliary_loss_mlp": 0.01027084, + "balance_loss_clip": 0.89654756, + "balance_loss_mlp": 1.02005625, + "epoch": 0.711116455239584, + "flos": 30480474326400.0, + "grad_norm": 2.261729837729427, + "language_loss": 0.56835377, + "learning_rate": 8.132402535401059e-07, + "loss": 0.58937871, + "num_input_tokens_seen": 127116825, + "step": 5914, + "time_per_iteration": 3.7208023071289062 + }, + { + "auxiliary_loss_clip": 0.01146899, + "auxiliary_loss_mlp": 0.01022112, + "balance_loss_clip": 1.01718032, + "balance_loss_mlp": 1.01519477, + "epoch": 0.711236698130223, + "flos": 25045610158080.0, + "grad_norm": 1.8186835603796159, + "language_loss": 0.74072063, + "learning_rate": 8.126133317194465e-07, + "loss": 0.76241076, + "num_input_tokens_seen": 127137015, + "step": 5915, + "time_per_iteration": 2.6676290035247803 + }, + { + "auxiliary_loss_clip": 0.01064234, + "auxiliary_loss_mlp": 0.0102845, + "balance_loss_clip": 0.86190647, + "balance_loss_mlp": 1.0208447, + "epoch": 0.7113569410208621, + "flos": 24206701040640.0, + "grad_norm": 1.7396855546168, + "language_loss": 0.74276054, + "learning_rate": 8.11986590024401e-07, + "loss": 0.76368737, + "num_input_tokens_seen": 127156755, + "step": 5916, + "time_per_iteration": 2.937927007675171 + }, + { + "auxiliary_loss_clip": 0.01129279, + "auxiliary_loss_mlp": 0.01027024, + "balance_loss_clip": 0.97918189, + "balance_loss_mlp": 1.01929951, + "epoch": 0.7114771839115013, + "flos": 35439306526080.0, + "grad_norm": 1.9257430958872666, + "language_loss": 0.68932182, + "learning_rate": 8.113600285500442e-07, + "loss": 0.71088481, + "num_input_tokens_seen": 127176965, + "step": 5917, + "time_per_iteration": 3.841526508331299 + }, + { + "auxiliary_loss_clip": 0.01168292, + "auxiliary_loss_mlp": 0.01027272, + "balance_loss_clip": 1.05026722, + "balance_loss_mlp": 1.01969635, + "epoch": 0.7115974268021403, + "flos": 21099458096640.0, + "grad_norm": 1.9107144208155575, + "language_loss": 0.74434, + "learning_rate": 8.107336473914268e-07, + "loss": 0.76629561, + "num_input_tokens_seen": 127195595, + "step": 5918, + "time_per_iteration": 2.638413667678833 + }, + { + "auxiliary_loss_clip": 0.01028657, + "auxiliary_loss_mlp": 0.00999833, + "balance_loss_clip": 0.94159299, + "balance_loss_mlp": 0.99877244, + "epoch": 0.7117176696927794, + "flos": 56752866616320.0, + "grad_norm": 0.7708582988372799, + "language_loss": 0.5567919, + "learning_rate": 8.101074466435694e-07, + "loss": 0.57707679, + "num_input_tokens_seen": 127255070, + "step": 5919, + "time_per_iteration": 3.1698882579803467 + }, + { + "auxiliary_loss_clip": 0.01139275, + "auxiliary_loss_mlp": 0.01018521, + "balance_loss_clip": 1.01070809, + "balance_loss_mlp": 1.01108801, + "epoch": 0.7118379125834186, + "flos": 15925269905280.0, + "grad_norm": 1.7826205942567592, + "language_loss": 0.68009752, + "learning_rate": 8.094814264014662e-07, + "loss": 0.70167542, + "num_input_tokens_seen": 127273825, + "step": 5920, + "time_per_iteration": 2.6188759803771973 + }, + { + "auxiliary_loss_clip": 0.01172999, + "auxiliary_loss_mlp": 0.01030818, + "balance_loss_clip": 1.05202734, + "balance_loss_mlp": 1.02302742, + "epoch": 0.7119581554740576, + "flos": 20193360589440.0, + "grad_norm": 2.00998765101658, + "language_loss": 0.81734228, + "learning_rate": 8.088555867600844e-07, + "loss": 0.83938044, + "num_input_tokens_seen": 127289990, + "step": 5921, + "time_per_iteration": 2.6447536945343018 + }, + { + "auxiliary_loss_clip": 0.01095085, + "auxiliary_loss_mlp": 0.01023077, + "balance_loss_clip": 0.93413091, + "balance_loss_mlp": 1.0160495, + "epoch": 0.7120783983646967, + "flos": 34715383822080.0, + "grad_norm": 4.136619340258567, + "language_loss": 0.60229224, + "learning_rate": 8.08229927814362e-07, + "loss": 0.62347382, + "num_input_tokens_seen": 127312880, + "step": 5922, + "time_per_iteration": 2.8581912517547607 + }, + { + "auxiliary_loss_clip": 0.01094918, + "auxiliary_loss_mlp": 0.01026888, + "balance_loss_clip": 0.9333089, + "balance_loss_mlp": 1.02005184, + "epoch": 0.7121986412553358, + "flos": 26359114700160.0, + "grad_norm": 3.4882892286158573, + "language_loss": 0.65050405, + "learning_rate": 8.076044496592134e-07, + "loss": 0.67172211, + "num_input_tokens_seen": 127334730, + "step": 5923, + "time_per_iteration": 2.768967628479004 + }, + { + "auxiliary_loss_clip": 0.01120088, + "auxiliary_loss_mlp": 0.01023915, + "balance_loss_clip": 0.9737457, + "balance_loss_mlp": 1.01631546, + "epoch": 0.7123188841459749, + "flos": 11145344371200.0, + "grad_norm": 2.2675327344643383, + "language_loss": 0.78218293, + "learning_rate": 8.069791523895204e-07, + "loss": 0.8036229, + "num_input_tokens_seen": 127351180, + "step": 5924, + "time_per_iteration": 3.6954965591430664 + }, + { + "auxiliary_loss_clip": 0.010874, + "auxiliary_loss_mlp": 0.01025855, + "balance_loss_clip": 0.93119323, + "balance_loss_mlp": 1.01835966, + "epoch": 0.7124391270366139, + "flos": 20811670329600.0, + "grad_norm": 1.9807463044871974, + "language_loss": 0.77514333, + "learning_rate": 8.063540361001422e-07, + "loss": 0.79627585, + "num_input_tokens_seen": 127369750, + "step": 5925, + "time_per_iteration": 2.736798048019409 + }, + { + "auxiliary_loss_clip": 0.01093125, + "auxiliary_loss_mlp": 0.0103001, + "balance_loss_clip": 0.93433058, + "balance_loss_mlp": 1.02224314, + "epoch": 0.7125593699272531, + "flos": 17603734584960.0, + "grad_norm": 2.5325417262430157, + "language_loss": 0.79985499, + "learning_rate": 8.057291008859069e-07, + "loss": 0.82108629, + "num_input_tokens_seen": 127387910, + "step": 5926, + "time_per_iteration": 2.7252256870269775 + }, + { + "auxiliary_loss_clip": 0.01139958, + "auxiliary_loss_mlp": 0.01031195, + "balance_loss_clip": 1.00915325, + "balance_loss_mlp": 1.02398014, + "epoch": 0.7126796128178922, + "flos": 28654057526400.0, + "grad_norm": 1.900599473001487, + "language_loss": 0.68427265, + "learning_rate": 8.051043468416187e-07, + "loss": 0.70598423, + "num_input_tokens_seen": 127409160, + "step": 5927, + "time_per_iteration": 2.747286558151245 + }, + { + "auxiliary_loss_clip": 0.01167503, + "auxiliary_loss_mlp": 0.0102922, + "balance_loss_clip": 1.0519073, + "balance_loss_mlp": 1.02220178, + "epoch": 0.7127998557085312, + "flos": 16034438315520.0, + "grad_norm": 1.8204233988343457, + "language_loss": 0.82086605, + "learning_rate": 8.044797740620506e-07, + "loss": 0.84283322, + "num_input_tokens_seen": 127427765, + "step": 5928, + "time_per_iteration": 2.625638961791992 + }, + { + "auxiliary_loss_clip": 0.01069012, + "auxiliary_loss_mlp": 0.01025669, + "balance_loss_clip": 0.89791167, + "balance_loss_mlp": 1.01805413, + "epoch": 0.7129200985991703, + "flos": 23403271582080.0, + "grad_norm": 2.0613764091847524, + "language_loss": 0.78851795, + "learning_rate": 8.038553826419494e-07, + "loss": 0.80946481, + "num_input_tokens_seen": 127446475, + "step": 5929, + "time_per_iteration": 2.778960704803467 + }, + { + "auxiliary_loss_clip": 0.0116533, + "auxiliary_loss_mlp": 0.01025241, + "balance_loss_clip": 1.04731536, + "balance_loss_mlp": 1.01802254, + "epoch": 0.7130403414898094, + "flos": 21397445326080.0, + "grad_norm": 1.7169811693543378, + "language_loss": 0.81035715, + "learning_rate": 8.032311726760364e-07, + "loss": 0.83226287, + "num_input_tokens_seen": 127467695, + "step": 5930, + "time_per_iteration": 2.666313409805298 + }, + { + "auxiliary_loss_clip": 0.01090637, + "auxiliary_loss_mlp": 0.01029971, + "balance_loss_clip": 0.93379891, + "balance_loss_mlp": 1.02218699, + "epoch": 0.7131605843804485, + "flos": 74739045306240.0, + "grad_norm": 2.2140392746107027, + "language_loss": 0.6846391, + "learning_rate": 8.026071442590022e-07, + "loss": 0.70584512, + "num_input_tokens_seen": 127494590, + "step": 5931, + "time_per_iteration": 3.160552978515625 + }, + { + "auxiliary_loss_clip": 0.01147064, + "auxiliary_loss_mlp": 0.01028244, + "balance_loss_clip": 1.01549733, + "balance_loss_mlp": 1.02077568, + "epoch": 0.7132808272710875, + "flos": 18368739469440.0, + "grad_norm": 2.051913399820228, + "language_loss": 0.80377424, + "learning_rate": 8.019832974855134e-07, + "loss": 0.82552731, + "num_input_tokens_seen": 127512550, + "step": 5932, + "time_per_iteration": 2.6392202377319336 + }, + { + "auxiliary_loss_clip": 0.01095638, + "auxiliary_loss_mlp": 0.01032223, + "balance_loss_clip": 0.93418956, + "balance_loss_mlp": 1.02485573, + "epoch": 0.7134010701617267, + "flos": 23253380127360.0, + "grad_norm": 2.096934262742798, + "language_loss": 0.82646286, + "learning_rate": 8.013596324502052e-07, + "loss": 0.84774143, + "num_input_tokens_seen": 127531015, + "step": 5933, + "time_per_iteration": 2.7776055335998535 + }, + { + "auxiliary_loss_clip": 0.01140163, + "auxiliary_loss_mlp": 0.01024402, + "balance_loss_clip": 1.0131886, + "balance_loss_mlp": 1.01738095, + "epoch": 0.7135213130523658, + "flos": 23653137565440.0, + "grad_norm": 1.6907964474049546, + "language_loss": 0.78567207, + "learning_rate": 8.007361492476872e-07, + "loss": 0.80731761, + "num_input_tokens_seen": 127550340, + "step": 5934, + "time_per_iteration": 2.677441120147705 + }, + { + "auxiliary_loss_clip": 0.01105099, + "auxiliary_loss_mlp": 0.01023609, + "balance_loss_clip": 0.93417728, + "balance_loss_mlp": 1.01627731, + "epoch": 0.7136415559430048, + "flos": 24790644443520.0, + "grad_norm": 1.4367960201654473, + "language_loss": 0.78934556, + "learning_rate": 8.001128479725426e-07, + "loss": 0.81063265, + "num_input_tokens_seen": 127572245, + "step": 5935, + "time_per_iteration": 2.7944812774658203 + }, + { + "auxiliary_loss_clip": 0.01068177, + "auxiliary_loss_mlp": 0.01029916, + "balance_loss_clip": 0.89404804, + "balance_loss_mlp": 1.02287102, + "epoch": 0.713761798833644, + "flos": 18296954138880.0, + "grad_norm": 1.6086044992911477, + "language_loss": 0.80943328, + "learning_rate": 7.994897287193248e-07, + "loss": 0.83041418, + "num_input_tokens_seen": 127591625, + "step": 5936, + "time_per_iteration": 2.806513786315918 + }, + { + "auxiliary_loss_clip": 0.01149198, + "auxiliary_loss_mlp": 0.01025511, + "balance_loss_clip": 1.01289046, + "balance_loss_mlp": 1.01829553, + "epoch": 0.713882041724283, + "flos": 15558262692480.0, + "grad_norm": 2.353385644437358, + "language_loss": 0.8357327, + "learning_rate": 7.988667915825605e-07, + "loss": 0.85747987, + "num_input_tokens_seen": 127608690, + "step": 5937, + "time_per_iteration": 2.650026798248291 + }, + { + "auxiliary_loss_clip": 0.01121614, + "auxiliary_loss_mlp": 0.01030406, + "balance_loss_clip": 0.97327328, + "balance_loss_mlp": 1.0230267, + "epoch": 0.7140022846149221, + "flos": 24061011477120.0, + "grad_norm": 2.763682853089622, + "language_loss": 0.75475818, + "learning_rate": 7.982440366567491e-07, + "loss": 0.77627838, + "num_input_tokens_seen": 127627180, + "step": 5938, + "time_per_iteration": 3.6732921600341797 + }, + { + "auxiliary_loss_clip": 0.01137945, + "auxiliary_loss_mlp": 0.01023875, + "balance_loss_clip": 1.00918174, + "balance_loss_mlp": 1.01716661, + "epoch": 0.7141225275055613, + "flos": 27891710248320.0, + "grad_norm": 1.5766422838074716, + "language_loss": 0.75081193, + "learning_rate": 7.97621464036361e-07, + "loss": 0.77243018, + "num_input_tokens_seen": 127648940, + "step": 5939, + "time_per_iteration": 2.6947104930877686 + }, + { + "auxiliary_loss_clip": 0.01147266, + "auxiliary_loss_mlp": 0.01023369, + "balance_loss_clip": 1.01106358, + "balance_loss_mlp": 1.01593602, + "epoch": 0.7142427703962003, + "flos": 19682603147520.0, + "grad_norm": 1.593478350221908, + "language_loss": 0.67679262, + "learning_rate": 7.969990738158417e-07, + "loss": 0.69849896, + "num_input_tokens_seen": 127667350, + "step": 5940, + "time_per_iteration": 3.537332057952881 + }, + { + "auxiliary_loss_clip": 0.01148267, + "auxiliary_loss_mlp": 0.01025633, + "balance_loss_clip": 1.014045, + "balance_loss_mlp": 1.01808071, + "epoch": 0.7143630132868394, + "flos": 21032377447680.0, + "grad_norm": 3.6417941431882586, + "language_loss": 0.84916127, + "learning_rate": 7.963768660896062e-07, + "loss": 0.87090027, + "num_input_tokens_seen": 127685760, + "step": 5941, + "time_per_iteration": 2.709652900695801 + }, + { + "auxiliary_loss_clip": 0.01148887, + "auxiliary_loss_mlp": 0.0103128, + "balance_loss_clip": 1.01271129, + "balance_loss_mlp": 1.02404392, + "epoch": 0.7144832561774785, + "flos": 24129923719680.0, + "grad_norm": 2.008785009646027, + "language_loss": 0.82569242, + "learning_rate": 7.957548409520432e-07, + "loss": 0.84749413, + "num_input_tokens_seen": 127704985, + "step": 5942, + "time_per_iteration": 3.6278762817382812 + }, + { + "auxiliary_loss_clip": 0.01100128, + "auxiliary_loss_mlp": 0.01028235, + "balance_loss_clip": 0.93439329, + "balance_loss_mlp": 1.02102244, + "epoch": 0.7146034990681176, + "flos": 16325817442560.0, + "grad_norm": 1.7885898697326819, + "language_loss": 0.84127367, + "learning_rate": 7.951329984975135e-07, + "loss": 0.86255729, + "num_input_tokens_seen": 127721925, + "step": 5943, + "time_per_iteration": 2.720792293548584 + }, + { + "auxiliary_loss_clip": 0.01015635, + "auxiliary_loss_mlp": 0.00999832, + "balance_loss_clip": 0.9050408, + "balance_loss_mlp": 0.99881917, + "epoch": 0.7147237419587567, + "flos": 69627164232960.0, + "grad_norm": 0.7105166906236522, + "language_loss": 0.54255843, + "learning_rate": 7.94511338820349e-07, + "loss": 0.56271309, + "num_input_tokens_seen": 127784230, + "step": 5944, + "time_per_iteration": 3.2708730697631836 + }, + { + "auxiliary_loss_clip": 0.01121792, + "auxiliary_loss_mlp": 0.00509901, + "balance_loss_clip": 0.97354048, + "balance_loss_mlp": 0.0, + "epoch": 0.7148439848493958, + "flos": 22266806198400.0, + "grad_norm": 2.630696609344901, + "language_loss": 0.78604245, + "learning_rate": 7.938898620148575e-07, + "loss": 0.80235946, + "num_input_tokens_seen": 127801990, + "step": 5945, + "time_per_iteration": 2.7640511989593506 + }, + { + "auxiliary_loss_clip": 0.01120299, + "auxiliary_loss_mlp": 0.0102358, + "balance_loss_clip": 0.97390968, + "balance_loss_mlp": 1.0164665, + "epoch": 0.7149642277400349, + "flos": 17931383470080.0, + "grad_norm": 1.9143324121441507, + "language_loss": 0.71220881, + "learning_rate": 7.932685681753135e-07, + "loss": 0.73364758, + "num_input_tokens_seen": 127819270, + "step": 5946, + "time_per_iteration": 2.6886146068573 + }, + { + "auxiliary_loss_clip": 0.01166228, + "auxiliary_loss_mlp": 0.01022753, + "balance_loss_clip": 1.05124605, + "balance_loss_mlp": 1.01546335, + "epoch": 0.7150844706306739, + "flos": 31681937370240.0, + "grad_norm": 1.8970445098882973, + "language_loss": 0.62919521, + "learning_rate": 7.92647457395969e-07, + "loss": 0.65108502, + "num_input_tokens_seen": 127841095, + "step": 5947, + "time_per_iteration": 2.7388787269592285 + }, + { + "auxiliary_loss_clip": 0.01055344, + "auxiliary_loss_mlp": 0.01036001, + "balance_loss_clip": 0.85415387, + "balance_loss_mlp": 1.02806199, + "epoch": 0.7152047135213131, + "flos": 10926217451520.0, + "grad_norm": 2.0564341814684535, + "language_loss": 0.74126422, + "learning_rate": 7.920265297710444e-07, + "loss": 0.76217771, + "num_input_tokens_seen": 127858485, + "step": 5948, + "time_per_iteration": 2.749333620071411 + }, + { + "auxiliary_loss_clip": 0.01146372, + "auxiliary_loss_mlp": 0.01024233, + "balance_loss_clip": 1.01307535, + "balance_loss_mlp": 1.01713109, + "epoch": 0.7153249564119522, + "flos": 20995640812800.0, + "grad_norm": 2.691666701209496, + "language_loss": 0.73373652, + "learning_rate": 7.914057853947363e-07, + "loss": 0.7554425, + "num_input_tokens_seen": 127877665, + "step": 5949, + "time_per_iteration": 2.6783299446105957 + }, + { + "auxiliary_loss_clip": 0.01097584, + "auxiliary_loss_mlp": 0.01029013, + "balance_loss_clip": 0.93658257, + "balance_loss_mlp": 1.02156878, + "epoch": 0.7154451993025912, + "flos": 24243114453120.0, + "grad_norm": 1.75517503988951, + "language_loss": 0.62320399, + "learning_rate": 7.907852243612089e-07, + "loss": 0.64446998, + "num_input_tokens_seen": 127898070, + "step": 5950, + "time_per_iteration": 3.674748182296753 + }, + { + "auxiliary_loss_clip": 0.0111871, + "auxiliary_loss_mlp": 0.01024098, + "balance_loss_clip": 0.97176999, + "balance_loss_mlp": 1.01707673, + "epoch": 0.7155654421932304, + "flos": 23330947547520.0, + "grad_norm": 1.7125601540987208, + "language_loss": 0.7249071, + "learning_rate": 7.901648467646009e-07, + "loss": 0.74633515, + "num_input_tokens_seen": 127917010, + "step": 5951, + "time_per_iteration": 2.777231454849243 + }, + { + "auxiliary_loss_clip": 0.01170846, + "auxiliary_loss_mlp": 0.0102583, + "balance_loss_clip": 1.05250955, + "balance_loss_mlp": 1.01836741, + "epoch": 0.7156856850838694, + "flos": 22711883621760.0, + "grad_norm": 1.6030480833564604, + "language_loss": 0.7258302, + "learning_rate": 7.895446526990244e-07, + "loss": 0.74779695, + "num_input_tokens_seen": 127937025, + "step": 5952, + "time_per_iteration": 2.685830593109131 + }, + { + "auxiliary_loss_clip": 0.01082681, + "auxiliary_loss_mlp": 0.01025488, + "balance_loss_clip": 0.89616716, + "balance_loss_mlp": 1.01884246, + "epoch": 0.7158059279745085, + "flos": 19865424395520.0, + "grad_norm": 1.566533340121585, + "language_loss": 0.76007468, + "learning_rate": 7.889246422585609e-07, + "loss": 0.78115642, + "num_input_tokens_seen": 127956410, + "step": 5953, + "time_per_iteration": 2.8132388591766357 + }, + { + "auxiliary_loss_clip": 0.01167139, + "auxiliary_loss_mlp": 0.01020526, + "balance_loss_clip": 1.0489974, + "balance_loss_mlp": 1.01349831, + "epoch": 0.7159261708651476, + "flos": 24134772055680.0, + "grad_norm": 2.2792294120743497, + "language_loss": 0.73417425, + "learning_rate": 7.883048155372675e-07, + "loss": 0.75605083, + "num_input_tokens_seen": 127974925, + "step": 5954, + "time_per_iteration": 2.6755170822143555 + }, + { + "auxiliary_loss_clip": 0.01126519, + "auxiliary_loss_mlp": 0.01029308, + "balance_loss_clip": 0.97500104, + "balance_loss_mlp": 1.02205384, + "epoch": 0.7160464137557867, + "flos": 16983198201600.0, + "grad_norm": 2.2355638052562994, + "language_loss": 0.71623027, + "learning_rate": 7.876851726291698e-07, + "loss": 0.73778856, + "num_input_tokens_seen": 127993225, + "step": 5955, + "time_per_iteration": 2.688987970352173 + }, + { + "auxiliary_loss_clip": 0.0110015, + "auxiliary_loss_mlp": 0.01024392, + "balance_loss_clip": 0.9327969, + "balance_loss_mlp": 1.01741803, + "epoch": 0.7161666566464258, + "flos": 25228251838080.0, + "grad_norm": 2.0468627953098926, + "language_loss": 0.78240407, + "learning_rate": 7.870657136282666e-07, + "loss": 0.80364949, + "num_input_tokens_seen": 128012085, + "step": 5956, + "time_per_iteration": 2.817854404449463 + }, + { + "auxiliary_loss_clip": 0.01137783, + "auxiliary_loss_mlp": 0.01027375, + "balance_loss_clip": 1.00877917, + "balance_loss_mlp": 1.01970947, + "epoch": 0.7162868995370649, + "flos": 26468390851200.0, + "grad_norm": 2.166224089455068, + "language_loss": 0.81675929, + "learning_rate": 7.86446438628531e-07, + "loss": 0.83841085, + "num_input_tokens_seen": 128033155, + "step": 5957, + "time_per_iteration": 2.7245285511016846 + }, + { + "auxiliary_loss_clip": 0.0106799, + "auxiliary_loss_mlp": 0.01005615, + "balance_loss_clip": 1.01497281, + "balance_loss_mlp": 1.00464368, + "epoch": 0.716407142427704, + "flos": 69998912040960.0, + "grad_norm": 0.7702114972866971, + "language_loss": 0.56855619, + "learning_rate": 7.858273477239059e-07, + "loss": 0.58929229, + "num_input_tokens_seen": 128101575, + "step": 5958, + "time_per_iteration": 3.209266424179077 + }, + { + "auxiliary_loss_clip": 0.01063187, + "auxiliary_loss_mlp": 0.01029276, + "balance_loss_clip": 0.89360416, + "balance_loss_mlp": 1.0221293, + "epoch": 0.716527385318343, + "flos": 20740459616640.0, + "grad_norm": 1.6779313155088436, + "language_loss": 0.71445966, + "learning_rate": 7.852084410083067e-07, + "loss": 0.73538435, + "num_input_tokens_seen": 128120395, + "step": 5959, + "time_per_iteration": 2.7964539527893066 + }, + { + "auxiliary_loss_clip": 0.01116702, + "auxiliary_loss_mlp": 0.01026641, + "balance_loss_clip": 0.97321844, + "balance_loss_mlp": 1.01974773, + "epoch": 0.7166476282089821, + "flos": 25371966153600.0, + "grad_norm": 1.570926701786334, + "language_loss": 0.63593781, + "learning_rate": 7.84589718575621e-07, + "loss": 0.65737122, + "num_input_tokens_seen": 128140840, + "step": 5960, + "time_per_iteration": 2.7209630012512207 + }, + { + "auxiliary_loss_clip": 0.01119692, + "auxiliary_loss_mlp": 0.01025866, + "balance_loss_clip": 0.96810919, + "balance_loss_mlp": 1.01843381, + "epoch": 0.7167678710996213, + "flos": 24133730561280.0, + "grad_norm": 2.2032882244558443, + "language_loss": 0.68874502, + "learning_rate": 7.83971180519708e-07, + "loss": 0.71020055, + "num_input_tokens_seen": 128159695, + "step": 5961, + "time_per_iteration": 2.7548794746398926 + }, + { + "auxiliary_loss_clip": 0.01171007, + "auxiliary_loss_mlp": 0.01036388, + "balance_loss_clip": 1.05152202, + "balance_loss_mlp": 1.02794218, + "epoch": 0.7168881139902603, + "flos": 30226586019840.0, + "grad_norm": 2.372061008775362, + "language_loss": 0.7558229, + "learning_rate": 7.833528269344008e-07, + "loss": 0.77789688, + "num_input_tokens_seen": 128179600, + "step": 5962, + "time_per_iteration": 2.702732563018799 + }, + { + "auxiliary_loss_clip": 0.01100232, + "auxiliary_loss_mlp": 0.0102823, + "balance_loss_clip": 0.93979263, + "balance_loss_mlp": 1.02031493, + "epoch": 0.7170083568808994, + "flos": 14606414236800.0, + "grad_norm": 2.2001026546821993, + "language_loss": 0.774598, + "learning_rate": 7.827346579135023e-07, + "loss": 0.79588264, + "num_input_tokens_seen": 128196940, + "step": 5963, + "time_per_iteration": 2.686873197555542 + }, + { + "auxiliary_loss_clip": 0.01115332, + "auxiliary_loss_mlp": 0.01023327, + "balance_loss_clip": 0.97048914, + "balance_loss_mlp": 1.01606715, + "epoch": 0.7171285997715385, + "flos": 23331091201920.0, + "grad_norm": 1.8790173046556782, + "language_loss": 0.83106041, + "learning_rate": 7.821166735507885e-07, + "loss": 0.85244691, + "num_input_tokens_seen": 128215970, + "step": 5964, + "time_per_iteration": 3.693455219268799 + }, + { + "auxiliary_loss_clip": 0.01168174, + "auxiliary_loss_mlp": 0.01029724, + "balance_loss_clip": 1.051229, + "balance_loss_mlp": 1.02214873, + "epoch": 0.7172488426621776, + "flos": 16543543731840.0, + "grad_norm": 1.6280823516840863, + "language_loss": 0.68935913, + "learning_rate": 7.81498873940007e-07, + "loss": 0.71133816, + "num_input_tokens_seen": 128233185, + "step": 5965, + "time_per_iteration": 2.5650720596313477 + }, + { + "auxiliary_loss_clip": 0.01147885, + "auxiliary_loss_mlp": 0.01027314, + "balance_loss_clip": 1.00816703, + "balance_loss_mlp": 1.01954722, + "epoch": 0.7173690855528166, + "flos": 26541612725760.0, + "grad_norm": 3.344297969078546, + "language_loss": 0.77446765, + "learning_rate": 7.808812591748768e-07, + "loss": 0.79621965, + "num_input_tokens_seen": 128253565, + "step": 5966, + "time_per_iteration": 3.5783536434173584 + }, + { + "auxiliary_loss_clip": 0.01094771, + "auxiliary_loss_mlp": 0.01023691, + "balance_loss_clip": 0.93479443, + "balance_loss_mlp": 1.01557255, + "epoch": 0.7174893284434558, + "flos": 22784099915520.0, + "grad_norm": 1.8320794387287853, + "language_loss": 0.64611781, + "learning_rate": 7.802638293490915e-07, + "loss": 0.66730243, + "num_input_tokens_seen": 128273210, + "step": 5967, + "time_per_iteration": 2.7740590572357178 + }, + { + "auxiliary_loss_clip": 0.01124807, + "auxiliary_loss_mlp": 0.01022313, + "balance_loss_clip": 0.97327816, + "balance_loss_mlp": 1.01489556, + "epoch": 0.7176095713340949, + "flos": 23293564467840.0, + "grad_norm": 6.0740858254614905, + "language_loss": 0.77038622, + "learning_rate": 7.796465845563123e-07, + "loss": 0.79185736, + "num_input_tokens_seen": 128292085, + "step": 5968, + "time_per_iteration": 2.7173409461975098 + }, + { + "auxiliary_loss_clip": 0.01119312, + "auxiliary_loss_mlp": 0.00509784, + "balance_loss_clip": 0.97401893, + "balance_loss_mlp": 0.0, + "epoch": 0.7177298142247339, + "flos": 25591631777280.0, + "grad_norm": 2.1178875809886657, + "language_loss": 0.79587853, + "learning_rate": 7.790295248901766e-07, + "loss": 0.81216955, + "num_input_tokens_seen": 128313215, + "step": 5969, + "time_per_iteration": 3.6387767791748047 + }, + { + "auxiliary_loss_clip": 0.01145448, + "auxiliary_loss_mlp": 0.01031838, + "balance_loss_clip": 1.01213479, + "balance_loss_mlp": 1.02382755, + "epoch": 0.7178500571153731, + "flos": 31652778504960.0, + "grad_norm": 1.8472919195820119, + "language_loss": 0.62196088, + "learning_rate": 7.784126504442902e-07, + "loss": 0.64373374, + "num_input_tokens_seen": 128336445, + "step": 5970, + "time_per_iteration": 2.7753427028656006 + }, + { + "auxiliary_loss_clip": 0.01089605, + "auxiliary_loss_mlp": 0.0102658, + "balance_loss_clip": 0.93371224, + "balance_loss_mlp": 1.01891494, + "epoch": 0.7179703000060121, + "flos": 19427242383360.0, + "grad_norm": 1.3747948506153695, + "language_loss": 0.67768568, + "learning_rate": 7.777959613122351e-07, + "loss": 0.69884753, + "num_input_tokens_seen": 128356270, + "step": 5971, + "time_per_iteration": 2.7343015670776367 + }, + { + "auxiliary_loss_clip": 0.01117992, + "auxiliary_loss_mlp": 0.01027036, + "balance_loss_clip": 0.97529173, + "balance_loss_mlp": 1.0199455, + "epoch": 0.7180905428966512, + "flos": 28839249072000.0, + "grad_norm": 1.8660835365170094, + "language_loss": 0.78029513, + "learning_rate": 7.771794575875604e-07, + "loss": 0.80174541, + "num_input_tokens_seen": 128378140, + "step": 5972, + "time_per_iteration": 2.7418324947357178 + }, + { + "auxiliary_loss_clip": 0.01150161, + "auxiliary_loss_mlp": 0.01026793, + "balance_loss_clip": 1.0185411, + "balance_loss_mlp": 1.01874065, + "epoch": 0.7182107857872904, + "flos": 20047563285120.0, + "grad_norm": 2.152289574104285, + "language_loss": 0.77807206, + "learning_rate": 7.765631393637888e-07, + "loss": 0.79984164, + "num_input_tokens_seen": 128396335, + "step": 5973, + "time_per_iteration": 2.634401559829712 + }, + { + "auxiliary_loss_clip": 0.01139627, + "auxiliary_loss_mlp": 0.01028309, + "balance_loss_clip": 1.00855803, + "balance_loss_mlp": 1.02010119, + "epoch": 0.7183310286779294, + "flos": 22747686503040.0, + "grad_norm": 3.6430447401354353, + "language_loss": 0.48977447, + "learning_rate": 7.75947006734417e-07, + "loss": 0.51145381, + "num_input_tokens_seen": 128414115, + "step": 5974, + "time_per_iteration": 2.6617391109466553 + }, + { + "auxiliary_loss_clip": 0.01166735, + "auxiliary_loss_mlp": 0.01023672, + "balance_loss_clip": 1.04815793, + "balance_loss_mlp": 1.01552439, + "epoch": 0.7184512715685685, + "flos": 17158262112000.0, + "grad_norm": 2.111036281822467, + "language_loss": 0.82788342, + "learning_rate": 7.753310597929101e-07, + "loss": 0.84978747, + "num_input_tokens_seen": 128430755, + "step": 5975, + "time_per_iteration": 2.579041004180908 + }, + { + "auxiliary_loss_clip": 0.01066287, + "auxiliary_loss_mlp": 0.01002278, + "balance_loss_clip": 1.01363587, + "balance_loss_mlp": 1.00128889, + "epoch": 0.7185715144592076, + "flos": 65509611448320.0, + "grad_norm": 0.7560164168412933, + "language_loss": 0.55096734, + "learning_rate": 7.747152986327095e-07, + "loss": 0.57165289, + "num_input_tokens_seen": 128491300, + "step": 5976, + "time_per_iteration": 4.007817506790161 + }, + { + "auxiliary_loss_clip": 0.01080995, + "auxiliary_loss_mlp": 0.01033859, + "balance_loss_clip": 0.8982209, + "balance_loss_mlp": 1.02639008, + "epoch": 0.7186917573498467, + "flos": 16180522928640.0, + "grad_norm": 1.930392281619434, + "language_loss": 0.68222404, + "learning_rate": 7.740997233472228e-07, + "loss": 0.70337254, + "num_input_tokens_seen": 128508920, + "step": 5977, + "time_per_iteration": 2.729417324066162 + }, + { + "auxiliary_loss_clip": 0.01122505, + "auxiliary_loss_mlp": 0.01022057, + "balance_loss_clip": 0.97367513, + "balance_loss_mlp": 1.01534522, + "epoch": 0.7188120002404857, + "flos": 29242274647680.0, + "grad_norm": 3.7505610032150702, + "language_loss": 0.70406187, + "learning_rate": 7.734843340298329e-07, + "loss": 0.7255075, + "num_input_tokens_seen": 128528745, + "step": 5978, + "time_per_iteration": 2.7853946685791016 + }, + { + "auxiliary_loss_clip": 0.0112573, + "auxiliary_loss_mlp": 0.01032458, + "balance_loss_clip": 0.97175747, + "balance_loss_mlp": 1.02471554, + "epoch": 0.7189322431311249, + "flos": 33401161008000.0, + "grad_norm": 1.9927120579938968, + "language_loss": 0.74970174, + "learning_rate": 7.72869130773895e-07, + "loss": 0.77128363, + "num_input_tokens_seen": 128549345, + "step": 5979, + "time_per_iteration": 2.893022060394287 + }, + { + "auxiliary_loss_clip": 0.01048978, + "auxiliary_loss_mlp": 0.00999306, + "balance_loss_clip": 0.97721303, + "balance_loss_mlp": 0.99823874, + "epoch": 0.719052486021764, + "flos": 61351263792000.0, + "grad_norm": 0.7888939225711301, + "language_loss": 0.5937984, + "learning_rate": 7.722541136727343e-07, + "loss": 0.61428118, + "num_input_tokens_seen": 128605360, + "step": 5980, + "time_per_iteration": 3.106750726699829 + }, + { + "auxiliary_loss_clip": 0.0114551, + "auxiliary_loss_mlp": 0.0102811, + "balance_loss_clip": 1.01281774, + "balance_loss_mlp": 1.02099276, + "epoch": 0.719172728912403, + "flos": 15596795007360.0, + "grad_norm": 1.894401075724962, + "language_loss": 0.808918, + "learning_rate": 7.716392828196483e-07, + "loss": 0.83065426, + "num_input_tokens_seen": 128623160, + "step": 5981, + "time_per_iteration": 2.7451140880584717 + }, + { + "auxiliary_loss_clip": 0.01146781, + "auxiliary_loss_mlp": 0.01034623, + "balance_loss_clip": 1.01314116, + "balance_loss_mlp": 1.02677917, + "epoch": 0.7192929718030422, + "flos": 15553162961280.0, + "grad_norm": 2.2868075798009624, + "language_loss": 0.77450532, + "learning_rate": 7.710246383079064e-07, + "loss": 0.79631937, + "num_input_tokens_seen": 128638545, + "step": 5982, + "time_per_iteration": 2.624828815460205 + }, + { + "auxiliary_loss_clip": 0.01121269, + "auxiliary_loss_mlp": 0.01024345, + "balance_loss_clip": 0.96975607, + "balance_loss_mlp": 1.01708853, + "epoch": 0.7194132146936812, + "flos": 21862487733120.0, + "grad_norm": 2.2140372134431354, + "language_loss": 0.91906142, + "learning_rate": 7.704101802307492e-07, + "loss": 0.94051754, + "num_input_tokens_seen": 128650845, + "step": 5983, + "time_per_iteration": 2.7055954933166504 + }, + { + "auxiliary_loss_clip": 0.0109203, + "auxiliary_loss_mlp": 0.0103137, + "balance_loss_clip": 0.93447775, + "balance_loss_mlp": 1.02355599, + "epoch": 0.7195334575843203, + "flos": 27338900958720.0, + "grad_norm": 2.544209486434349, + "language_loss": 0.8710891, + "learning_rate": 7.697959086813912e-07, + "loss": 0.89232314, + "num_input_tokens_seen": 128667010, + "step": 5984, + "time_per_iteration": 2.713505268096924 + }, + { + "auxiliary_loss_clip": 0.01090722, + "auxiliary_loss_mlp": 0.01031959, + "balance_loss_clip": 0.93210918, + "balance_loss_mlp": 1.02418375, + "epoch": 0.7196537004749595, + "flos": 18770615809920.0, + "grad_norm": 1.634009583617879, + "language_loss": 0.80355239, + "learning_rate": 7.691818237530145e-07, + "loss": 0.82477921, + "num_input_tokens_seen": 128685870, + "step": 5985, + "time_per_iteration": 2.7492141723632812 + }, + { + "auxiliary_loss_clip": 0.01085199, + "auxiliary_loss_mlp": 0.0102354, + "balance_loss_clip": 0.89601797, + "balance_loss_mlp": 1.01601744, + "epoch": 0.7197739433655985, + "flos": 24531009960960.0, + "grad_norm": 2.2500185086420497, + "language_loss": 0.77299058, + "learning_rate": 7.685679255387774e-07, + "loss": 0.79407799, + "num_input_tokens_seen": 128704185, + "step": 5986, + "time_per_iteration": 2.7783234119415283 + }, + { + "auxiliary_loss_clip": 0.01119254, + "auxiliary_loss_mlp": 0.01028174, + "balance_loss_clip": 0.97401237, + "balance_loss_mlp": 1.02122378, + "epoch": 0.7198941862562376, + "flos": 18040587793920.0, + "grad_norm": 2.009465301470442, + "language_loss": 0.77147943, + "learning_rate": 7.679542141318065e-07, + "loss": 0.79295373, + "num_input_tokens_seen": 128721290, + "step": 5987, + "time_per_iteration": 2.6430575847625732 + }, + { + "auxiliary_loss_clip": 0.01108087, + "auxiliary_loss_mlp": 0.01024636, + "balance_loss_clip": 0.96905887, + "balance_loss_mlp": 1.01725745, + "epoch": 0.7200144291468767, + "flos": 29022393542400.0, + "grad_norm": 1.9839513026566882, + "language_loss": 0.757442, + "learning_rate": 7.673406896252013e-07, + "loss": 0.77876925, + "num_input_tokens_seen": 128742665, + "step": 5988, + "time_per_iteration": 2.7690165042877197 + }, + { + "auxiliary_loss_clip": 0.01097171, + "auxiliary_loss_mlp": 0.01028953, + "balance_loss_clip": 0.93323874, + "balance_loss_mlp": 1.02144277, + "epoch": 0.7201346720375158, + "flos": 25374264624000.0, + "grad_norm": 1.7431446350083049, + "language_loss": 0.78037655, + "learning_rate": 7.667273521120347e-07, + "loss": 0.80163777, + "num_input_tokens_seen": 128762225, + "step": 5989, + "time_per_iteration": 2.7988526821136475 + }, + { + "auxiliary_loss_clip": 0.01101432, + "auxiliary_loss_mlp": 0.01030999, + "balance_loss_clip": 0.93500888, + "balance_loss_mlp": 1.02382839, + "epoch": 0.7202549149281549, + "flos": 14355614499840.0, + "grad_norm": 2.081913165301464, + "language_loss": 0.79458994, + "learning_rate": 7.661142016853468e-07, + "loss": 0.81591415, + "num_input_tokens_seen": 128779585, + "step": 5990, + "time_per_iteration": 3.6142067909240723 + }, + { + "auxiliary_loss_clip": 0.01069265, + "auxiliary_loss_mlp": 0.01023305, + "balance_loss_clip": 0.89351457, + "balance_loss_mlp": 1.01605988, + "epoch": 0.7203751578187939, + "flos": 23001682550400.0, + "grad_norm": 2.8341191634064304, + "language_loss": 0.75067782, + "learning_rate": 7.655012384381543e-07, + "loss": 0.77160358, + "num_input_tokens_seen": 128799070, + "step": 5991, + "time_per_iteration": 2.766965627670288 + }, + { + "auxiliary_loss_clip": 0.01118031, + "auxiliary_loss_mlp": 0.01023984, + "balance_loss_clip": 0.97489208, + "balance_loss_mlp": 1.01624155, + "epoch": 0.7204954007094331, + "flos": 23692424065920.0, + "grad_norm": 1.9050914058315487, + "language_loss": 0.81628299, + "learning_rate": 7.648884624634415e-07, + "loss": 0.83770311, + "num_input_tokens_seen": 128817620, + "step": 5992, + "time_per_iteration": 2.705134391784668 + }, + { + "auxiliary_loss_clip": 0.01142097, + "auxiliary_loss_mlp": 0.01026929, + "balance_loss_clip": 1.01163411, + "balance_loss_mlp": 1.01996148, + "epoch": 0.7206156436000721, + "flos": 16253026531200.0, + "grad_norm": 1.9242067446251365, + "language_loss": 0.88836348, + "learning_rate": 7.642758738541683e-07, + "loss": 0.91005373, + "num_input_tokens_seen": 128834200, + "step": 5993, + "time_per_iteration": 3.5510494709014893 + }, + { + "auxiliary_loss_clip": 0.01046245, + "auxiliary_loss_mlp": 0.01003524, + "balance_loss_clip": 0.97614539, + "balance_loss_mlp": 1.00252235, + "epoch": 0.7207358864907112, + "flos": 54377806504320.0, + "grad_norm": 0.7610007968807477, + "language_loss": 0.60760337, + "learning_rate": 7.636634727032621e-07, + "loss": 0.62810105, + "num_input_tokens_seen": 128891305, + "step": 5994, + "time_per_iteration": 3.9203596115112305 + }, + { + "auxiliary_loss_clip": 0.01098845, + "auxiliary_loss_mlp": 0.0102735, + "balance_loss_clip": 0.93098253, + "balance_loss_mlp": 1.01891899, + "epoch": 0.7208561293813504, + "flos": 19135540033920.0, + "grad_norm": 1.854624205141099, + "language_loss": 0.78723156, + "learning_rate": 7.630512591036231e-07, + "loss": 0.80849355, + "num_input_tokens_seen": 128910615, + "step": 5995, + "time_per_iteration": 2.7349376678466797 + }, + { + "auxiliary_loss_clip": 0.01144243, + "auxiliary_loss_mlp": 0.01027954, + "balance_loss_clip": 1.01088881, + "balance_loss_mlp": 1.02061629, + "epoch": 0.7209763722719894, + "flos": 17748526308480.0, + "grad_norm": 3.771307107711331, + "language_loss": 0.64664322, + "learning_rate": 7.624392331481255e-07, + "loss": 0.66836524, + "num_input_tokens_seen": 128928270, + "step": 5996, + "time_per_iteration": 2.6636016368865967 + }, + { + "auxiliary_loss_clip": 0.01045023, + "auxiliary_loss_mlp": 0.01004701, + "balance_loss_clip": 0.97512829, + "balance_loss_mlp": 1.00366426, + "epoch": 0.7210966151626285, + "flos": 66819488716800.0, + "grad_norm": 0.7397081886089012, + "language_loss": 0.518314, + "learning_rate": 7.618273949296115e-07, + "loss": 0.53881121, + "num_input_tokens_seen": 128987780, + "step": 5997, + "time_per_iteration": 3.1240663528442383 + }, + { + "auxiliary_loss_clip": 0.01113868, + "auxiliary_loss_mlp": 0.01025299, + "balance_loss_clip": 0.96959949, + "balance_loss_mlp": 1.0177474, + "epoch": 0.7212168580532676, + "flos": 21141869080320.0, + "grad_norm": 1.9126376945855401, + "language_loss": 0.68441343, + "learning_rate": 7.612157445408987e-07, + "loss": 0.70580506, + "num_input_tokens_seen": 129005590, + "step": 5998, + "time_per_iteration": 2.772252082824707 + }, + { + "auxiliary_loss_clip": 0.01130143, + "auxiliary_loss_mlp": 0.01030644, + "balance_loss_clip": 0.97959208, + "balance_loss_mlp": 1.02303863, + "epoch": 0.7213371009439067, + "flos": 22345738335360.0, + "grad_norm": 1.990503038181268, + "language_loss": 0.74104488, + "learning_rate": 7.606042820747716e-07, + "loss": 0.76265275, + "num_input_tokens_seen": 129021995, + "step": 5999, + "time_per_iteration": 2.739039182662964 + }, + { + "auxiliary_loss_clip": 0.01126593, + "auxiliary_loss_mlp": 0.01034277, + "balance_loss_clip": 0.97594398, + "balance_loss_mlp": 1.02701771, + "epoch": 0.7214573438345457, + "flos": 18515901490560.0, + "grad_norm": 1.7804198487507112, + "language_loss": 0.85401034, + "learning_rate": 7.599930076239889e-07, + "loss": 0.87561905, + "num_input_tokens_seen": 129039280, + "step": 6000, + "time_per_iteration": 2.731710433959961 + }, + { + "auxiliary_loss_clip": 0.01081199, + "auxiliary_loss_mlp": 0.00509905, + "balance_loss_clip": 0.89875227, + "balance_loss_mlp": 0.0, + "epoch": 0.7215775867251849, + "flos": 35736108606720.0, + "grad_norm": 1.873864487804745, + "language_loss": 0.70868796, + "learning_rate": 7.593819212812818e-07, + "loss": 0.724599, + "num_input_tokens_seen": 129060860, + "step": 6001, + "time_per_iteration": 3.8032314777374268 + }, + { + "auxiliary_loss_clip": 0.01144108, + "auxiliary_loss_mlp": 0.01025648, + "balance_loss_clip": 1.01124537, + "balance_loss_mlp": 1.01842999, + "epoch": 0.721697829615824, + "flos": 20372410909440.0, + "grad_norm": 1.9903056630255822, + "language_loss": 0.7181105, + "learning_rate": 7.587710231393508e-07, + "loss": 0.73980808, + "num_input_tokens_seen": 129079215, + "step": 6002, + "time_per_iteration": 2.627772808074951 + }, + { + "auxiliary_loss_clip": 0.01015537, + "auxiliary_loss_mlp": 0.01024996, + "balance_loss_clip": 0.81621182, + "balance_loss_mlp": 1.01720572, + "epoch": 0.721818072506463, + "flos": 20229809915520.0, + "grad_norm": 2.472806882812666, + "language_loss": 0.8316443, + "learning_rate": 7.581603132908685e-07, + "loss": 0.85204971, + "num_input_tokens_seen": 129097185, + "step": 6003, + "time_per_iteration": 2.7460708618164062 + }, + { + "auxiliary_loss_clip": 0.01091897, + "auxiliary_loss_mlp": 0.01022758, + "balance_loss_clip": 0.93306124, + "balance_loss_mlp": 1.01515269, + "epoch": 0.7219383153971022, + "flos": 18186887888640.0, + "grad_norm": 2.39642116252288, + "language_loss": 0.78530204, + "learning_rate": 7.575497918284795e-07, + "loss": 0.80644858, + "num_input_tokens_seen": 129114730, + "step": 6004, + "time_per_iteration": 2.6901817321777344 + }, + { + "auxiliary_loss_clip": 0.01169951, + "auxiliary_loss_mlp": 0.01025954, + "balance_loss_clip": 1.04925799, + "balance_loss_mlp": 1.01777041, + "epoch": 0.7220585582877412, + "flos": 17342124854400.0, + "grad_norm": 3.8964447126466775, + "language_loss": 0.74635035, + "learning_rate": 7.569394588447984e-07, + "loss": 0.76830935, + "num_input_tokens_seen": 129131745, + "step": 6005, + "time_per_iteration": 2.589541435241699 + }, + { + "auxiliary_loss_clip": 0.01136851, + "auxiliary_loss_mlp": 0.01033542, + "balance_loss_clip": 1.00863373, + "balance_loss_mlp": 1.02576327, + "epoch": 0.7221788011783803, + "flos": 16976338704000.0, + "grad_norm": 2.299454493576833, + "language_loss": 0.78166413, + "learning_rate": 7.563293144324146e-07, + "loss": 0.80336809, + "num_input_tokens_seen": 129147295, + "step": 6006, + "time_per_iteration": 2.624046564102173 + }, + { + "auxiliary_loss_clip": 0.01167389, + "auxiliary_loss_mlp": 0.01025535, + "balance_loss_clip": 1.05090499, + "balance_loss_mlp": 1.01847184, + "epoch": 0.7222990440690195, + "flos": 26286359702400.0, + "grad_norm": 1.875735657001475, + "language_loss": 0.80107874, + "learning_rate": 7.557193586838834e-07, + "loss": 0.82300794, + "num_input_tokens_seen": 129162660, + "step": 6007, + "time_per_iteration": 2.7030279636383057 + }, + { + "auxiliary_loss_clip": 0.01125949, + "auxiliary_loss_mlp": 0.01027003, + "balance_loss_clip": 0.97244459, + "balance_loss_mlp": 1.01892662, + "epoch": 0.7224192869596585, + "flos": 17601687509760.0, + "grad_norm": 2.3606782253673, + "language_loss": 0.70985055, + "learning_rate": 7.551095916917371e-07, + "loss": 0.73138016, + "num_input_tokens_seen": 129179990, + "step": 6008, + "time_per_iteration": 2.627965211868286 + }, + { + "auxiliary_loss_clip": 0.01107698, + "auxiliary_loss_mlp": 0.01037284, + "balance_loss_clip": 0.93426764, + "balance_loss_mlp": 1.02913046, + "epoch": 0.7225395298502976, + "flos": 12932331016320.0, + "grad_norm": 2.8947790400010738, + "language_loss": 0.66910362, + "learning_rate": 7.545000135484758e-07, + "loss": 0.69055343, + "num_input_tokens_seen": 129197425, + "step": 6009, + "time_per_iteration": 2.702402353286743 + }, + { + "auxiliary_loss_clip": 0.01170264, + "auxiliary_loss_mlp": 0.00509565, + "balance_loss_clip": 1.05236995, + "balance_loss_mlp": 0.0, + "epoch": 0.7226597727409367, + "flos": 29643899592960.0, + "grad_norm": 1.9495522613344767, + "language_loss": 0.62808239, + "learning_rate": 7.538906243465714e-07, + "loss": 0.64488065, + "num_input_tokens_seen": 129217560, + "step": 6010, + "time_per_iteration": 2.672633647918701 + }, + { + "auxiliary_loss_clip": 0.01168807, + "auxiliary_loss_mlp": 0.01028969, + "balance_loss_clip": 1.0496999, + "balance_loss_mlp": 1.02102375, + "epoch": 0.7227800156315758, + "flos": 13771635183360.0, + "grad_norm": 2.010069071764713, + "language_loss": 0.78572953, + "learning_rate": 7.5328142417847e-07, + "loss": 0.80770731, + "num_input_tokens_seen": 129234325, + "step": 6011, + "time_per_iteration": 2.7042253017425537 + }, + { + "auxiliary_loss_clip": 0.01140047, + "auxiliary_loss_mlp": 0.01026488, + "balance_loss_clip": 1.00788808, + "balance_loss_mlp": 1.01970851, + "epoch": 0.7229002585222148, + "flos": 20301882554880.0, + "grad_norm": 1.7581443443957834, + "language_loss": 0.69027543, + "learning_rate": 7.526724131365838e-07, + "loss": 0.71194071, + "num_input_tokens_seen": 129255280, + "step": 6012, + "time_per_iteration": 2.6335582733154297 + }, + { + "auxiliary_loss_clip": 0.01121354, + "auxiliary_loss_mlp": 0.01033601, + "balance_loss_clip": 0.97620499, + "balance_loss_mlp": 1.02573967, + "epoch": 0.723020501412854, + "flos": 16581250033920.0, + "grad_norm": 2.013152672802028, + "language_loss": 0.7072528, + "learning_rate": 7.520635913133017e-07, + "loss": 0.72880232, + "num_input_tokens_seen": 129273910, + "step": 6013, + "time_per_iteration": 2.700230598449707 + }, + { + "auxiliary_loss_clip": 0.01148553, + "auxiliary_loss_mlp": 0.01029308, + "balance_loss_clip": 1.01075149, + "balance_loss_mlp": 1.0210942, + "epoch": 0.7231407443034931, + "flos": 28548300908160.0, + "grad_norm": 1.7419093485937147, + "language_loss": 0.82565635, + "learning_rate": 7.514549588009798e-07, + "loss": 0.847435, + "num_input_tokens_seen": 129294785, + "step": 6014, + "time_per_iteration": 2.7179672718048096 + }, + { + "auxiliary_loss_clip": 0.01125816, + "auxiliary_loss_mlp": 0.01024576, + "balance_loss_clip": 0.97508866, + "balance_loss_mlp": 1.01679802, + "epoch": 0.7232609871941321, + "flos": 30008536508160.0, + "grad_norm": 1.9697165917134392, + "language_loss": 0.70366263, + "learning_rate": 7.508465156919492e-07, + "loss": 0.72516656, + "num_input_tokens_seen": 129318295, + "step": 6015, + "time_per_iteration": 2.7587764263153076 + }, + { + "auxiliary_loss_clip": 0.01119529, + "auxiliary_loss_mlp": 0.01030861, + "balance_loss_clip": 0.97126663, + "balance_loss_mlp": 1.02347314, + "epoch": 0.7233812300847713, + "flos": 16654005031680.0, + "grad_norm": 2.394917835569151, + "language_loss": 0.61208737, + "learning_rate": 7.502382620785083e-07, + "loss": 0.63359129, + "num_input_tokens_seen": 129334845, + "step": 6016, + "time_per_iteration": 3.650913715362549 + }, + { + "auxiliary_loss_clip": 0.01004775, + "auxiliary_loss_mlp": 0.01006906, + "balance_loss_clip": 0.90559685, + "balance_loss_mlp": 1.00579762, + "epoch": 0.7235014729754103, + "flos": 67258784050560.0, + "grad_norm": 0.8236048492129814, + "language_loss": 0.62542367, + "learning_rate": 7.496301980529289e-07, + "loss": 0.64554048, + "num_input_tokens_seen": 129398055, + "step": 6017, + "time_per_iteration": 3.2851972579956055 + }, + { + "auxiliary_loss_clip": 0.01168653, + "auxiliary_loss_mlp": 0.01025207, + "balance_loss_clip": 1.04987705, + "balance_loss_mlp": 1.01809049, + "epoch": 0.7236217158660494, + "flos": 26943237671040.0, + "grad_norm": 2.4079970353397466, + "language_loss": 0.74460626, + "learning_rate": 7.490223237074547e-07, + "loss": 0.76654488, + "num_input_tokens_seen": 129417765, + "step": 6018, + "time_per_iteration": 3.635420799255371 + }, + { + "auxiliary_loss_clip": 0.01096743, + "auxiliary_loss_mlp": 0.01023261, + "balance_loss_clip": 0.93309706, + "balance_loss_mlp": 1.01615882, + "epoch": 0.7237419587566886, + "flos": 29423372042880.0, + "grad_norm": 1.8730895799040468, + "language_loss": 0.66433311, + "learning_rate": 7.484146391342989e-07, + "loss": 0.68553311, + "num_input_tokens_seen": 129437560, + "step": 6019, + "time_per_iteration": 2.8443801403045654 + }, + { + "auxiliary_loss_clip": 0.0111696, + "auxiliary_loss_mlp": 0.01026739, + "balance_loss_clip": 0.97182095, + "balance_loss_mlp": 1.01927006, + "epoch": 0.7238622016473276, + "flos": 17821496787840.0, + "grad_norm": 2.069405223832027, + "language_loss": 0.56591719, + "learning_rate": 7.478071444256484e-07, + "loss": 0.58735418, + "num_input_tokens_seen": 129455320, + "step": 6020, + "time_per_iteration": 3.6895864009857178 + }, + { + "auxiliary_loss_clip": 0.01106194, + "auxiliary_loss_mlp": 0.01033415, + "balance_loss_clip": 0.93325603, + "balance_loss_mlp": 1.02592862, + "epoch": 0.7239824445379667, + "flos": 25739117020800.0, + "grad_norm": 1.8992272497914864, + "language_loss": 0.79246104, + "learning_rate": 7.471998396736579e-07, + "loss": 0.8138572, + "num_input_tokens_seen": 129475700, + "step": 6021, + "time_per_iteration": 2.8502469062805176 + }, + { + "auxiliary_loss_clip": 0.01099779, + "auxiliary_loss_mlp": 0.01027226, + "balance_loss_clip": 0.93585664, + "balance_loss_mlp": 1.01991808, + "epoch": 0.7241026874286057, + "flos": 23148916398720.0, + "grad_norm": 1.6887735757793882, + "language_loss": 0.7620526, + "learning_rate": 7.465927249704549e-07, + "loss": 0.78332269, + "num_input_tokens_seen": 129493585, + "step": 6022, + "time_per_iteration": 2.7484328746795654 + }, + { + "auxiliary_loss_clip": 0.01144623, + "auxiliary_loss_mlp": 0.0102196, + "balance_loss_clip": 1.01155114, + "balance_loss_mlp": 1.01484942, + "epoch": 0.7242229303192449, + "flos": 20266905686400.0, + "grad_norm": 2.4526570831212124, + "language_loss": 0.77432209, + "learning_rate": 7.459858004081398e-07, + "loss": 0.7959879, + "num_input_tokens_seen": 129511555, + "step": 6023, + "time_per_iteration": 2.6187450885772705 + }, + { + "auxiliary_loss_clip": 0.01000096, + "auxiliary_loss_mlp": 0.00999878, + "balance_loss_clip": 0.90211141, + "balance_loss_mlp": 0.99868602, + "epoch": 0.724343173209884, + "flos": 62311659684480.0, + "grad_norm": 0.6526942852330236, + "language_loss": 0.57995939, + "learning_rate": 7.453790660787815e-07, + "loss": 0.59995914, + "num_input_tokens_seen": 129579650, + "step": 6024, + "time_per_iteration": 3.363387107849121 + }, + { + "auxiliary_loss_clip": 0.01126067, + "auxiliary_loss_mlp": 0.01034557, + "balance_loss_clip": 0.9754014, + "balance_loss_mlp": 1.02671874, + "epoch": 0.724463416100523, + "flos": 35006403813120.0, + "grad_norm": 1.9131856165877965, + "language_loss": 0.63348603, + "learning_rate": 7.447725220744214e-07, + "loss": 0.65509224, + "num_input_tokens_seen": 129601895, + "step": 6025, + "time_per_iteration": 2.772064208984375 + }, + { + "auxiliary_loss_clip": 0.01169437, + "auxiliary_loss_mlp": 0.01028065, + "balance_loss_clip": 1.0499661, + "balance_loss_mlp": 1.02029562, + "epoch": 0.7245836589911622, + "flos": 21871968923520.0, + "grad_norm": 2.2114864030019326, + "language_loss": 0.77082187, + "learning_rate": 7.441661684870717e-07, + "loss": 0.79279691, + "num_input_tokens_seen": 129622150, + "step": 6026, + "time_per_iteration": 2.7903151512145996 + }, + { + "auxiliary_loss_clip": 0.01168959, + "auxiliary_loss_mlp": 0.0102794, + "balance_loss_clip": 1.049209, + "balance_loss_mlp": 1.02070379, + "epoch": 0.7247039018818012, + "flos": 23006494972800.0, + "grad_norm": 1.672681646596672, + "language_loss": 0.81508851, + "learning_rate": 7.435600054087152e-07, + "loss": 0.83705753, + "num_input_tokens_seen": 129644315, + "step": 6027, + "time_per_iteration": 3.5767078399658203 + }, + { + "auxiliary_loss_clip": 0.01171248, + "auxiliary_loss_mlp": 0.01028301, + "balance_loss_clip": 1.05199361, + "balance_loss_mlp": 1.02083838, + "epoch": 0.7248241447724403, + "flos": 31722588587520.0, + "grad_norm": 2.541997694458346, + "language_loss": 0.74194431, + "learning_rate": 7.42954032931308e-07, + "loss": 0.76393974, + "num_input_tokens_seen": 129665355, + "step": 6028, + "time_per_iteration": 2.629394769668579 + }, + { + "auxiliary_loss_clip": 0.01123158, + "auxiliary_loss_mlp": 0.01028184, + "balance_loss_clip": 0.97380394, + "balance_loss_mlp": 1.02105808, + "epoch": 0.7249443876630794, + "flos": 34896984007680.0, + "grad_norm": 2.282585954446485, + "language_loss": 0.74825537, + "learning_rate": 7.423482511467733e-07, + "loss": 0.76976877, + "num_input_tokens_seen": 129686125, + "step": 6029, + "time_per_iteration": 2.7322888374328613 + }, + { + "auxiliary_loss_clip": 0.01041191, + "auxiliary_loss_mlp": 0.01032111, + "balance_loss_clip": 0.85592365, + "balance_loss_mlp": 1.02443361, + "epoch": 0.7250646305537185, + "flos": 26359294268160.0, + "grad_norm": 2.258411169305999, + "language_loss": 0.64952475, + "learning_rate": 7.417426601470099e-07, + "loss": 0.67025781, + "num_input_tokens_seen": 129706485, + "step": 6030, + "time_per_iteration": 2.7031242847442627 + }, + { + "auxiliary_loss_clip": 0.01146231, + "auxiliary_loss_mlp": 0.01028375, + "balance_loss_clip": 1.01139998, + "balance_loss_mlp": 1.02041173, + "epoch": 0.7251848734443576, + "flos": 30081614728320.0, + "grad_norm": 2.0282079852351202, + "language_loss": 0.78655362, + "learning_rate": 7.411372600238841e-07, + "loss": 0.80829966, + "num_input_tokens_seen": 129727100, + "step": 6031, + "time_per_iteration": 2.5882060527801514 + }, + { + "auxiliary_loss_clip": 0.01169994, + "auxiliary_loss_mlp": 0.01026328, + "balance_loss_clip": 1.05041802, + "balance_loss_mlp": 1.01902962, + "epoch": 0.7253051163349967, + "flos": 17785262943360.0, + "grad_norm": 2.125440487153611, + "language_loss": 0.74025244, + "learning_rate": 7.405320508692346e-07, + "loss": 0.76221567, + "num_input_tokens_seen": 129745840, + "step": 6032, + "time_per_iteration": 2.586878776550293 + }, + { + "auxiliary_loss_clip": 0.01167052, + "auxiliary_loss_mlp": 0.01026925, + "balance_loss_clip": 1.05126679, + "balance_loss_mlp": 1.01984119, + "epoch": 0.7254253592256358, + "flos": 12641346938880.0, + "grad_norm": 2.229186253497394, + "language_loss": 0.75747615, + "learning_rate": 7.399270327748727e-07, + "loss": 0.77941585, + "num_input_tokens_seen": 129763500, + "step": 6033, + "time_per_iteration": 2.577116012573242 + }, + { + "auxiliary_loss_clip": 0.01099774, + "auxiliary_loss_mlp": 0.00508561, + "balance_loss_clip": 0.9352181, + "balance_loss_mlp": 0.0, + "epoch": 0.7255456021162748, + "flos": 27199208966400.0, + "grad_norm": 2.4554535765642447, + "language_loss": 0.74432093, + "learning_rate": 7.39322205832577e-07, + "loss": 0.76040435, + "num_input_tokens_seen": 129784390, + "step": 6034, + "time_per_iteration": 2.7380142211914062 + }, + { + "auxiliary_loss_clip": 0.01116601, + "auxiliary_loss_mlp": 0.0102411, + "balance_loss_clip": 0.97192913, + "balance_loss_mlp": 1.01684976, + "epoch": 0.725665845006914, + "flos": 21288205088640.0, + "grad_norm": 1.8083414744812203, + "language_loss": 0.80799735, + "learning_rate": 7.387175701341009e-07, + "loss": 0.82940441, + "num_input_tokens_seen": 129803060, + "step": 6035, + "time_per_iteration": 2.695855140686035 + }, + { + "auxiliary_loss_clip": 0.01145456, + "auxiliary_loss_mlp": 0.01024587, + "balance_loss_clip": 1.0110563, + "balance_loss_mlp": 1.01662958, + "epoch": 0.7257860878975531, + "flos": 16033684129920.0, + "grad_norm": 14.390846810109796, + "language_loss": 0.72065949, + "learning_rate": 7.381131257711659e-07, + "loss": 0.74235994, + "num_input_tokens_seen": 129820165, + "step": 6036, + "time_per_iteration": 2.6024749279022217 + }, + { + "auxiliary_loss_clip": 0.01123062, + "auxiliary_loss_mlp": 0.01033941, + "balance_loss_clip": 0.98003531, + "balance_loss_mlp": 1.0269196, + "epoch": 0.7259063307881921, + "flos": 12129943052160.0, + "grad_norm": 1.8162352830929063, + "language_loss": 0.83641744, + "learning_rate": 7.375088728354677e-07, + "loss": 0.8579874, + "num_input_tokens_seen": 129835195, + "step": 6037, + "time_per_iteration": 2.632211208343506 + }, + { + "auxiliary_loss_clip": 0.01101553, + "auxiliary_loss_mlp": 0.0102918, + "balance_loss_clip": 0.93379891, + "balance_loss_mlp": 1.02159786, + "epoch": 0.7260265736788313, + "flos": 30443845432320.0, + "grad_norm": 1.8006710288452192, + "language_loss": 0.67197657, + "learning_rate": 7.369048114186691e-07, + "loss": 0.69328392, + "num_input_tokens_seen": 129856240, + "step": 6038, + "time_per_iteration": 2.7798895835876465 + }, + { + "auxiliary_loss_clip": 0.01106243, + "auxiliary_loss_mlp": 0.00509102, + "balance_loss_clip": 0.93555725, + "balance_loss_mlp": 0.0, + "epoch": 0.7261468165694703, + "flos": 21142264129920.0, + "grad_norm": 1.9227914181910206, + "language_loss": 0.82997698, + "learning_rate": 7.363009416124055e-07, + "loss": 0.84613037, + "num_input_tokens_seen": 129875565, + "step": 6039, + "time_per_iteration": 2.723798990249634 + }, + { + "auxiliary_loss_clip": 0.01103473, + "auxiliary_loss_mlp": 0.01029547, + "balance_loss_clip": 0.93691623, + "balance_loss_mlp": 1.02184606, + "epoch": 0.7262670594601094, + "flos": 22306308180480.0, + "grad_norm": 2.0852736134567853, + "language_loss": 0.62753195, + "learning_rate": 7.356972635082852e-07, + "loss": 0.64886212, + "num_input_tokens_seen": 129894420, + "step": 6040, + "time_per_iteration": 2.7133939266204834 + }, + { + "auxiliary_loss_clip": 0.01078281, + "auxiliary_loss_mlp": 0.01028056, + "balance_loss_clip": 0.90409732, + "balance_loss_mlp": 1.02070069, + "epoch": 0.7263873023507486, + "flos": 25335049950720.0, + "grad_norm": 1.7091112254813068, + "language_loss": 0.75139892, + "learning_rate": 7.35093777197884e-07, + "loss": 0.77246225, + "num_input_tokens_seen": 129914490, + "step": 6041, + "time_per_iteration": 2.805971622467041 + }, + { + "auxiliary_loss_clip": 0.01120205, + "auxiliary_loss_mlp": 0.01024711, + "balance_loss_clip": 0.97448087, + "balance_loss_mlp": 1.01785374, + "epoch": 0.7265075452413876, + "flos": 23878621192320.0, + "grad_norm": 2.0533093671079663, + "language_loss": 0.85779881, + "learning_rate": 7.344904827727525e-07, + "loss": 0.8792479, + "num_input_tokens_seen": 129931670, + "step": 6042, + "time_per_iteration": 3.6276278495788574 + }, + { + "auxiliary_loss_clip": 0.01097152, + "auxiliary_loss_mlp": 0.01026092, + "balance_loss_clip": 0.93253189, + "balance_loss_mlp": 1.01884389, + "epoch": 0.7266277881320267, + "flos": 28724549967360.0, + "grad_norm": 2.3471406335137703, + "language_loss": 0.73601544, + "learning_rate": 7.338873803244076e-07, + "loss": 0.75724792, + "num_input_tokens_seen": 129946905, + "step": 6043, + "time_per_iteration": 2.847869634628296 + }, + { + "auxiliary_loss_clip": 0.01119061, + "auxiliary_loss_mlp": 0.01022742, + "balance_loss_clip": 0.97457635, + "balance_loss_mlp": 1.01554501, + "epoch": 0.7267480310226658, + "flos": 24863507182080.0, + "grad_norm": 1.6711739849623972, + "language_loss": 0.80759025, + "learning_rate": 7.332844699443401e-07, + "loss": 0.82900834, + "num_input_tokens_seen": 129965505, + "step": 6044, + "time_per_iteration": 3.6766574382781982 + }, + { + "auxiliary_loss_clip": 0.01067316, + "auxiliary_loss_mlp": 0.01030028, + "balance_loss_clip": 0.89360613, + "balance_loss_mlp": 1.02282178, + "epoch": 0.7268682739133049, + "flos": 27198490694400.0, + "grad_norm": 2.004888802612326, + "language_loss": 0.75304055, + "learning_rate": 7.326817517240121e-07, + "loss": 0.774014, + "num_input_tokens_seen": 129987210, + "step": 6045, + "time_per_iteration": 2.9810571670532227 + }, + { + "auxiliary_loss_clip": 0.0114282, + "auxiliary_loss_mlp": 0.00508749, + "balance_loss_clip": 1.00849724, + "balance_loss_mlp": 0.0, + "epoch": 0.7269885168039439, + "flos": 33508138688640.0, + "grad_norm": 1.6994272401968171, + "language_loss": 0.83262354, + "learning_rate": 7.320792257548545e-07, + "loss": 0.84913921, + "num_input_tokens_seen": 130008385, + "step": 6046, + "time_per_iteration": 3.689598560333252 + }, + { + "auxiliary_loss_clip": 0.01124701, + "auxiliary_loss_mlp": 0.01025556, + "balance_loss_clip": 0.97298014, + "balance_loss_mlp": 1.01845086, + "epoch": 0.7271087596945831, + "flos": 24313750548480.0, + "grad_norm": 1.996118988915303, + "language_loss": 0.76411289, + "learning_rate": 7.314768921282704e-07, + "loss": 0.78561544, + "num_input_tokens_seen": 130029040, + "step": 6047, + "time_per_iteration": 2.7003345489501953 + }, + { + "auxiliary_loss_clip": 0.01147807, + "auxiliary_loss_mlp": 0.01027278, + "balance_loss_clip": 1.01165175, + "balance_loss_mlp": 1.0200063, + "epoch": 0.7272290025852222, + "flos": 23805147922560.0, + "grad_norm": 2.952982038799493, + "language_loss": 0.7217406, + "learning_rate": 7.30874750935633e-07, + "loss": 0.74349141, + "num_input_tokens_seen": 130048725, + "step": 6048, + "time_per_iteration": 2.7336745262145996 + }, + { + "auxiliary_loss_clip": 0.0109798, + "auxiliary_loss_mlp": 0.01027354, + "balance_loss_clip": 0.9355709, + "balance_loss_mlp": 1.02009726, + "epoch": 0.7273492454758612, + "flos": 16720367408640.0, + "grad_norm": 1.8713969859975588, + "language_loss": 0.79364246, + "learning_rate": 7.30272802268286e-07, + "loss": 0.81489581, + "num_input_tokens_seen": 130065720, + "step": 6049, + "time_per_iteration": 2.6610496044158936 + }, + { + "auxiliary_loss_clip": 0.01026037, + "auxiliary_loss_mlp": 0.01024032, + "balance_loss_clip": 0.85388696, + "balance_loss_mlp": 1.01727235, + "epoch": 0.7274694883665004, + "flos": 28031330413440.0, + "grad_norm": 1.7074797252217013, + "language_loss": 0.76439333, + "learning_rate": 7.29671046217547e-07, + "loss": 0.78489399, + "num_input_tokens_seen": 130084830, + "step": 6050, + "time_per_iteration": 2.8719987869262695 + }, + { + "auxiliary_loss_clip": 0.01099304, + "auxiliary_loss_mlp": 0.01032192, + "balance_loss_clip": 0.93558353, + "balance_loss_mlp": 1.02489901, + "epoch": 0.7275897312571394, + "flos": 30372706546560.0, + "grad_norm": 1.956860319228567, + "language_loss": 0.81717694, + "learning_rate": 7.290694828746988e-07, + "loss": 0.83849192, + "num_input_tokens_seen": 130104495, + "step": 6051, + "time_per_iteration": 2.7754569053649902 + }, + { + "auxiliary_loss_clip": 0.01102179, + "auxiliary_loss_mlp": 0.01027014, + "balance_loss_clip": 0.93452191, + "balance_loss_mlp": 1.0198493, + "epoch": 0.7277099741477785, + "flos": 19204775498880.0, + "grad_norm": 1.7948321566311323, + "language_loss": 0.85754776, + "learning_rate": 7.284681123310004e-07, + "loss": 0.87883961, + "num_input_tokens_seen": 130123210, + "step": 6052, + "time_per_iteration": 2.7570722103118896 + }, + { + "auxiliary_loss_clip": 0.0114347, + "auxiliary_loss_mlp": 0.01026593, + "balance_loss_clip": 1.01089084, + "balance_loss_mlp": 1.01911831, + "epoch": 0.7278302170384175, + "flos": 20667884186880.0, + "grad_norm": 1.7217316953892243, + "language_loss": 0.80037397, + "learning_rate": 7.27866934677678e-07, + "loss": 0.82207459, + "num_input_tokens_seen": 130142880, + "step": 6053, + "time_per_iteration": 2.6500890254974365 + }, + { + "auxiliary_loss_clip": 0.01071554, + "auxiliary_loss_mlp": 0.01024309, + "balance_loss_clip": 0.89580238, + "balance_loss_mlp": 1.01653409, + "epoch": 0.7279504599290567, + "flos": 19093200877440.0, + "grad_norm": 1.649901719888617, + "language_loss": 0.78225458, + "learning_rate": 7.272659500059297e-07, + "loss": 0.80321312, + "num_input_tokens_seen": 130160220, + "step": 6054, + "time_per_iteration": 3.6899912357330322 + }, + { + "auxiliary_loss_clip": 0.01141427, + "auxiliary_loss_mlp": 0.01028307, + "balance_loss_clip": 1.01232004, + "balance_loss_mlp": 1.02021217, + "epoch": 0.7280707028196958, + "flos": 19062174504960.0, + "grad_norm": 2.1484255436314452, + "language_loss": 0.80334687, + "learning_rate": 7.266651584069264e-07, + "loss": 0.82504421, + "num_input_tokens_seen": 130177885, + "step": 6055, + "time_per_iteration": 2.6600234508514404 + }, + { + "auxiliary_loss_clip": 0.01150636, + "auxiliary_loss_mlp": 0.01027397, + "balance_loss_clip": 1.01515961, + "balance_loss_mlp": 1.02048635, + "epoch": 0.7281909457103348, + "flos": 37196308293120.0, + "grad_norm": 1.6331366437106294, + "language_loss": 0.57086515, + "learning_rate": 7.260645599718045e-07, + "loss": 0.59264553, + "num_input_tokens_seen": 130204240, + "step": 6056, + "time_per_iteration": 2.766472816467285 + }, + { + "auxiliary_loss_clip": 0.01123075, + "auxiliary_loss_mlp": 0.01029061, + "balance_loss_clip": 0.97333246, + "balance_loss_mlp": 1.0208292, + "epoch": 0.728311188600974, + "flos": 20667094087680.0, + "grad_norm": 2.0913275457928644, + "language_loss": 0.66958785, + "learning_rate": 7.254641547916767e-07, + "loss": 0.69110924, + "num_input_tokens_seen": 130221735, + "step": 6057, + "time_per_iteration": 2.670987606048584 + }, + { + "auxiliary_loss_clip": 0.01167121, + "auxiliary_loss_mlp": 0.01027103, + "balance_loss_clip": 1.05075693, + "balance_loss_mlp": 1.01963127, + "epoch": 0.728431431491613, + "flos": 28840685616000.0, + "grad_norm": 1.6349057205473532, + "language_loss": 0.69599193, + "learning_rate": 7.248639429576226e-07, + "loss": 0.71793425, + "num_input_tokens_seen": 130241190, + "step": 6058, + "time_per_iteration": 2.660191297531128 + }, + { + "auxiliary_loss_clip": 0.01148088, + "auxiliary_loss_mlp": 0.01026344, + "balance_loss_clip": 1.01267493, + "balance_loss_mlp": 1.01875055, + "epoch": 0.7285516743822521, + "flos": 25991856092160.0, + "grad_norm": 1.6780986879421593, + "language_loss": 0.72227907, + "learning_rate": 7.242639245606959e-07, + "loss": 0.74402332, + "num_input_tokens_seen": 130260980, + "step": 6059, + "time_per_iteration": 2.6982195377349854 + }, + { + "auxiliary_loss_clip": 0.01126852, + "auxiliary_loss_mlp": 0.01029026, + "balance_loss_clip": 0.97239393, + "balance_loss_mlp": 1.02158189, + "epoch": 0.7286719172728913, + "flos": 16399721675520.0, + "grad_norm": 1.9450685012566373, + "language_loss": 0.82477838, + "learning_rate": 7.236640996919168e-07, + "loss": 0.84633714, + "num_input_tokens_seen": 130280025, + "step": 6060, + "time_per_iteration": 2.6415693759918213 + }, + { + "auxiliary_loss_clip": 0.0114748, + "auxiliary_loss_mlp": 0.01027424, + "balance_loss_clip": 1.0118084, + "balance_loss_mlp": 1.01961541, + "epoch": 0.7287921601635303, + "flos": 22018161277440.0, + "grad_norm": 1.4988176436487075, + "language_loss": 0.71019524, + "learning_rate": 7.230644684422782e-07, + "loss": 0.7319442, + "num_input_tokens_seen": 130300255, + "step": 6061, + "time_per_iteration": 2.67122483253479 + }, + { + "auxiliary_loss_clip": 0.01094977, + "auxiliary_loss_mlp": 0.01026116, + "balance_loss_clip": 0.93273205, + "balance_loss_mlp": 1.01888871, + "epoch": 0.7289124030541694, + "flos": 24600927784320.0, + "grad_norm": 1.673792565745114, + "language_loss": 0.815772, + "learning_rate": 7.224650309027451e-07, + "loss": 0.83698285, + "num_input_tokens_seen": 130320005, + "step": 6062, + "time_per_iteration": 2.749579906463623 + }, + { + "auxiliary_loss_clip": 0.01150193, + "auxiliary_loss_mlp": 0.01026816, + "balance_loss_clip": 1.01500463, + "balance_loss_mlp": 1.01965761, + "epoch": 0.7290326459448085, + "flos": 21393638484480.0, + "grad_norm": 2.5851484546411356, + "language_loss": 0.68614805, + "learning_rate": 7.218657871642506e-07, + "loss": 0.70791817, + "num_input_tokens_seen": 130338810, + "step": 6063, + "time_per_iteration": 2.650386333465576 + }, + { + "auxiliary_loss_clip": 0.01171576, + "auxiliary_loss_mlp": 0.01027144, + "balance_loss_clip": 1.05145884, + "balance_loss_mlp": 1.01956201, + "epoch": 0.7291528888354476, + "flos": 18587686821120.0, + "grad_norm": 2.2902710389109786, + "language_loss": 0.62534124, + "learning_rate": 7.212667373177012e-07, + "loss": 0.64732844, + "num_input_tokens_seen": 130353805, + "step": 6064, + "time_per_iteration": 2.604680061340332 + }, + { + "auxiliary_loss_clip": 0.01095996, + "auxiliary_loss_mlp": 0.01025569, + "balance_loss_clip": 0.93262917, + "balance_loss_mlp": 1.01807106, + "epoch": 0.7292731317260867, + "flos": 18951066760320.0, + "grad_norm": 1.6672386954543532, + "language_loss": 0.75067937, + "learning_rate": 7.206678814539704e-07, + "loss": 0.77189499, + "num_input_tokens_seen": 130372105, + "step": 6065, + "time_per_iteration": 2.72367262840271 + }, + { + "auxiliary_loss_clip": 0.01085006, + "auxiliary_loss_mlp": 0.0102753, + "balance_loss_clip": 0.89886081, + "balance_loss_mlp": 1.02072608, + "epoch": 0.7293933746167258, + "flos": 21067569797760.0, + "grad_norm": 1.5028837252542437, + "language_loss": 0.72855163, + "learning_rate": 7.20069219663904e-07, + "loss": 0.749677, + "num_input_tokens_seen": 130391990, + "step": 6066, + "time_per_iteration": 2.7265586853027344 + }, + { + "auxiliary_loss_clip": 0.01148312, + "auxiliary_loss_mlp": 0.01028004, + "balance_loss_clip": 1.01023424, + "balance_loss_mlp": 1.02009988, + "epoch": 0.7295136175073649, + "flos": 22453326547200.0, + "grad_norm": 1.8734318549515545, + "language_loss": 0.79667032, + "learning_rate": 7.1947075203832e-07, + "loss": 0.81843346, + "num_input_tokens_seen": 130411970, + "step": 6067, + "time_per_iteration": 2.664531946182251 + }, + { + "auxiliary_loss_clip": 0.01065312, + "auxiliary_loss_mlp": 0.01001183, + "balance_loss_clip": 1.01310706, + "balance_loss_mlp": 1.00015759, + "epoch": 0.7296338603980039, + "flos": 56125506648960.0, + "grad_norm": 0.8602679870887141, + "language_loss": 0.60096604, + "learning_rate": 7.188724786680049e-07, + "loss": 0.62163097, + "num_input_tokens_seen": 130472440, + "step": 6068, + "time_per_iteration": 4.1197075843811035 + }, + { + "auxiliary_loss_clip": 0.01120864, + "auxiliary_loss_mlp": 0.01025978, + "balance_loss_clip": 0.97141743, + "balance_loss_mlp": 1.01846194, + "epoch": 0.7297541032886431, + "flos": 25228287751680.0, + "grad_norm": 1.548030140718682, + "language_loss": 0.75675029, + "learning_rate": 7.182743996437162e-07, + "loss": 0.77821869, + "num_input_tokens_seen": 130491975, + "step": 6069, + "time_per_iteration": 2.719278573989868 + }, + { + "auxiliary_loss_clip": 0.01104359, + "auxiliary_loss_mlp": 0.01029297, + "balance_loss_clip": 0.93340987, + "balance_loss_mlp": 1.02057695, + "epoch": 0.7298743461792822, + "flos": 26467600752000.0, + "grad_norm": 1.8139459207710353, + "language_loss": 0.68839163, + "learning_rate": 7.176765150561819e-07, + "loss": 0.70972818, + "num_input_tokens_seen": 130510580, + "step": 6070, + "time_per_iteration": 2.7799930572509766 + }, + { + "auxiliary_loss_clip": 0.01168315, + "auxiliary_loss_mlp": 0.01032364, + "balance_loss_clip": 1.04915106, + "balance_loss_mlp": 1.02430511, + "epoch": 0.7299945890699212, + "flos": 19569053278080.0, + "grad_norm": 1.7933103754502215, + "language_loss": 0.80072522, + "learning_rate": 7.170788249961002e-07, + "loss": 0.82273197, + "num_input_tokens_seen": 130529090, + "step": 6071, + "time_per_iteration": 3.493670701980591 + }, + { + "auxiliary_loss_clip": 0.01165557, + "auxiliary_loss_mlp": 0.0102698, + "balance_loss_clip": 1.04901993, + "balance_loss_mlp": 1.0194819, + "epoch": 0.7301148319605604, + "flos": 22928963466240.0, + "grad_norm": 2.308732486488541, + "language_loss": 0.885041, + "learning_rate": 7.164813295541418e-07, + "loss": 0.90696633, + "num_input_tokens_seen": 130548655, + "step": 6072, + "time_per_iteration": 3.6016252040863037 + }, + { + "auxiliary_loss_clip": 0.0112288, + "auxiliary_loss_mlp": 0.01027426, + "balance_loss_clip": 0.97229517, + "balance_loss_mlp": 1.02007699, + "epoch": 0.7302350748511994, + "flos": 25369703596800.0, + "grad_norm": 1.6180743598533638, + "language_loss": 0.7048589, + "learning_rate": 7.15884028820944e-07, + "loss": 0.72636199, + "num_input_tokens_seen": 130567710, + "step": 6073, + "time_per_iteration": 2.6783859729766846 + }, + { + "auxiliary_loss_clip": 0.0109492, + "auxiliary_loss_mlp": 0.01022726, + "balance_loss_clip": 0.93289149, + "balance_loss_mlp": 1.01542401, + "epoch": 0.7303553177418385, + "flos": 27819170732160.0, + "grad_norm": 3.1710460490167147, + "language_loss": 0.60506546, + "learning_rate": 7.152869228871185e-07, + "loss": 0.62624192, + "num_input_tokens_seen": 130590195, + "step": 6074, + "time_per_iteration": 2.7862508296966553 + }, + { + "auxiliary_loss_clip": 0.01118232, + "auxiliary_loss_mlp": 0.01019727, + "balance_loss_clip": 0.9724195, + "balance_loss_mlp": 1.01160026, + "epoch": 0.7304755606324776, + "flos": 24426510318720.0, + "grad_norm": 2.8196732682003987, + "language_loss": 0.72044951, + "learning_rate": 7.146900118432457e-07, + "loss": 0.7418291, + "num_input_tokens_seen": 130609940, + "step": 6075, + "time_per_iteration": 2.7152862548828125 + }, + { + "auxiliary_loss_clip": 0.01023756, + "auxiliary_loss_mlp": 0.01029607, + "balance_loss_clip": 0.81287301, + "balance_loss_mlp": 1.02178049, + "epoch": 0.7305958035231167, + "flos": 23840483927040.0, + "grad_norm": 1.5577097299517342, + "language_loss": 0.86116505, + "learning_rate": 7.140932957798753e-07, + "loss": 0.88169873, + "num_input_tokens_seen": 130628380, + "step": 6076, + "time_per_iteration": 2.8355040550231934 + }, + { + "auxiliary_loss_clip": 0.01123873, + "auxiliary_loss_mlp": 0.01025518, + "balance_loss_clip": 0.9706589, + "balance_loss_mlp": 1.0185802, + "epoch": 0.7307160464137558, + "flos": 16726939597440.0, + "grad_norm": 1.9621583474386102, + "language_loss": 0.7139436, + "learning_rate": 7.134967747875309e-07, + "loss": 0.73543751, + "num_input_tokens_seen": 130646590, + "step": 6077, + "time_per_iteration": 2.7070415019989014 + }, + { + "auxiliary_loss_clip": 0.01139365, + "auxiliary_loss_mlp": 0.01029, + "balance_loss_clip": 1.00948727, + "balance_loss_mlp": 1.02136469, + "epoch": 0.7308362893043949, + "flos": 21798280172160.0, + "grad_norm": 1.8704161999942013, + "language_loss": 0.81791174, + "learning_rate": 7.129004489567014e-07, + "loss": 0.83959544, + "num_input_tokens_seen": 130664070, + "step": 6078, + "time_per_iteration": 2.6616878509521484 + }, + { + "auxiliary_loss_clip": 0.01101686, + "auxiliary_loss_mlp": 0.01027665, + "balance_loss_clip": 0.933851, + "balance_loss_mlp": 1.01980305, + "epoch": 0.730956532195034, + "flos": 10707377840640.0, + "grad_norm": 2.2664870074256873, + "language_loss": 0.77802467, + "learning_rate": 7.123043183778512e-07, + "loss": 0.79931819, + "num_input_tokens_seen": 130681400, + "step": 6079, + "time_per_iteration": 2.692450761795044 + }, + { + "auxiliary_loss_clip": 0.01106083, + "auxiliary_loss_mlp": 0.01030072, + "balance_loss_clip": 0.93769658, + "balance_loss_mlp": 1.02235341, + "epoch": 0.731076775085673, + "flos": 19791987039360.0, + "grad_norm": 1.6343443231114545, + "language_loss": 0.65406632, + "learning_rate": 7.117083831414114e-07, + "loss": 0.67542791, + "num_input_tokens_seen": 130700675, + "step": 6080, + "time_per_iteration": 3.6180613040924072 + }, + { + "auxiliary_loss_clip": 0.01167825, + "auxiliary_loss_mlp": 0.01030269, + "balance_loss_clip": 1.05158114, + "balance_loss_mlp": 1.02259195, + "epoch": 0.7311970179763122, + "flos": 20447033414400.0, + "grad_norm": 1.7137539408337863, + "language_loss": 0.69850063, + "learning_rate": 7.11112643337787e-07, + "loss": 0.72048163, + "num_input_tokens_seen": 130719720, + "step": 6081, + "time_per_iteration": 2.6364388465881348 + }, + { + "auxiliary_loss_clip": 0.01128445, + "auxiliary_loss_mlp": 0.01026771, + "balance_loss_clip": 0.97948921, + "balance_loss_mlp": 1.01904631, + "epoch": 0.7313172608669513, + "flos": 18513818501760.0, + "grad_norm": 2.3796683964661307, + "language_loss": 0.76543272, + "learning_rate": 7.10517099057349e-07, + "loss": 0.7869848, + "num_input_tokens_seen": 130736670, + "step": 6082, + "time_per_iteration": 2.703209161758423 + }, + { + "auxiliary_loss_clip": 0.01122131, + "auxiliary_loss_mlp": 0.0102647, + "balance_loss_clip": 0.97119457, + "balance_loss_mlp": 1.01847696, + "epoch": 0.7314375037575903, + "flos": 16180738410240.0, + "grad_norm": 2.1050605602830554, + "language_loss": 0.6184594, + "learning_rate": 7.099217503904411e-07, + "loss": 0.63994539, + "num_input_tokens_seen": 130754525, + "step": 6083, + "time_per_iteration": 2.7088775634765625 + }, + { + "auxiliary_loss_clip": 0.01122317, + "auxiliary_loss_mlp": 0.01025602, + "balance_loss_clip": 0.97199118, + "balance_loss_mlp": 1.01874423, + "epoch": 0.7315577466482295, + "flos": 17967940536960.0, + "grad_norm": 2.4765449785563747, + "language_loss": 0.90073669, + "learning_rate": 7.093265974273788e-07, + "loss": 0.92221582, + "num_input_tokens_seen": 130772420, + "step": 6084, + "time_per_iteration": 2.6200766563415527 + }, + { + "auxiliary_loss_clip": 0.01145957, + "auxiliary_loss_mlp": 0.01028091, + "balance_loss_clip": 1.00978327, + "balance_loss_mlp": 1.02128386, + "epoch": 0.7316779895388685, + "flos": 18405440190720.0, + "grad_norm": 1.6999289474763606, + "language_loss": 0.71649158, + "learning_rate": 7.087316402584447e-07, + "loss": 0.73823202, + "num_input_tokens_seen": 130791245, + "step": 6085, + "time_per_iteration": 2.6566214561462402 + }, + { + "auxiliary_loss_clip": 0.01168768, + "auxiliary_loss_mlp": 0.01025875, + "balance_loss_clip": 1.05055666, + "balance_loss_mlp": 1.01865435, + "epoch": 0.7317982324295076, + "flos": 17928294900480.0, + "grad_norm": 2.213083327066824, + "language_loss": 0.8670519, + "learning_rate": 7.081368789738953e-07, + "loss": 0.88899833, + "num_input_tokens_seen": 130808445, + "step": 6086, + "time_per_iteration": 2.646476984024048 + }, + { + "auxiliary_loss_clip": 0.0111451, + "auxiliary_loss_mlp": 0.01028269, + "balance_loss_clip": 0.96837294, + "balance_loss_mlp": 1.02056515, + "epoch": 0.7319184753201466, + "flos": 27229840289280.0, + "grad_norm": 2.2239994926155933, + "language_loss": 0.77842098, + "learning_rate": 7.075423136639537e-07, + "loss": 0.79984874, + "num_input_tokens_seen": 130827700, + "step": 6087, + "time_per_iteration": 2.7321927547454834 + }, + { + "auxiliary_loss_clip": 0.01091978, + "auxiliary_loss_mlp": 0.01032232, + "balance_loss_clip": 0.93196714, + "balance_loss_mlp": 1.02481091, + "epoch": 0.7320387182107858, + "flos": 37448544574080.0, + "grad_norm": 1.9002621408176237, + "language_loss": 0.75132108, + "learning_rate": 7.069479444188149e-07, + "loss": 0.77256316, + "num_input_tokens_seen": 130848290, + "step": 6088, + "time_per_iteration": 2.8365862369537354 + }, + { + "auxiliary_loss_clip": 0.01114673, + "auxiliary_loss_mlp": 0.01028314, + "balance_loss_clip": 0.97225487, + "balance_loss_mlp": 1.02106571, + "epoch": 0.7321589611014249, + "flos": 17859023521920.0, + "grad_norm": 1.8769103135739307, + "language_loss": 0.81922066, + "learning_rate": 7.063537713286453e-07, + "loss": 0.8406505, + "num_input_tokens_seen": 130865970, + "step": 6089, + "time_per_iteration": 2.7056496143341064 + }, + { + "auxiliary_loss_clip": 0.01126599, + "auxiliary_loss_mlp": 0.01029104, + "balance_loss_clip": 0.97207493, + "balance_loss_mlp": 1.02128994, + "epoch": 0.7322792039920639, + "flos": 26100593539200.0, + "grad_norm": 1.8986488878923866, + "language_loss": 0.81085706, + "learning_rate": 7.057597944835803e-07, + "loss": 0.83241409, + "num_input_tokens_seen": 130885245, + "step": 6090, + "time_per_iteration": 2.73445200920105 + }, + { + "auxiliary_loss_clip": 0.01099908, + "auxiliary_loss_mlp": 0.01024244, + "balance_loss_clip": 0.93114507, + "balance_loss_mlp": 1.01718724, + "epoch": 0.7323994468827031, + "flos": 25369093065600.0, + "grad_norm": 1.6235439724905494, + "language_loss": 0.7475251, + "learning_rate": 7.051660139737253e-07, + "loss": 0.76876664, + "num_input_tokens_seen": 130903465, + "step": 6091, + "time_per_iteration": 2.7467408180236816 + }, + { + "auxiliary_loss_clip": 0.01143246, + "auxiliary_loss_mlp": 0.0050931, + "balance_loss_clip": 1.01253533, + "balance_loss_mlp": 0.0, + "epoch": 0.7325196897733421, + "flos": 26907075653760.0, + "grad_norm": 2.071714446584922, + "language_loss": 0.76654303, + "learning_rate": 7.045724298891565e-07, + "loss": 0.78306866, + "num_input_tokens_seen": 130922935, + "step": 6092, + "time_per_iteration": 2.659179925918579 + }, + { + "auxiliary_loss_clip": 0.01144728, + "auxiliary_loss_mlp": 0.01024947, + "balance_loss_clip": 1.01257372, + "balance_loss_mlp": 1.01746416, + "epoch": 0.7326399326639812, + "flos": 25775781828480.0, + "grad_norm": 1.9233073197731432, + "language_loss": 0.69411367, + "learning_rate": 7.039790423199192e-07, + "loss": 0.71581042, + "num_input_tokens_seen": 130942575, + "step": 6093, + "time_per_iteration": 2.731048345565796 + }, + { + "auxiliary_loss_clip": 0.01124022, + "auxiliary_loss_mlp": 0.01026611, + "balance_loss_clip": 0.97378093, + "balance_loss_mlp": 1.01902318, + "epoch": 0.7327601755546204, + "flos": 21032269706880.0, + "grad_norm": 2.175133857737955, + "language_loss": 0.78305566, + "learning_rate": 7.033858513560322e-07, + "loss": 0.80456203, + "num_input_tokens_seen": 130958870, + "step": 6094, + "time_per_iteration": 3.5743961334228516 + }, + { + "auxiliary_loss_clip": 0.01145782, + "auxiliary_loss_mlp": 0.01023627, + "balance_loss_clip": 1.012712, + "balance_loss_mlp": 1.0168674, + "epoch": 0.7328804184452594, + "flos": 16289224462080.0, + "grad_norm": 2.208302108270174, + "language_loss": 0.76542211, + "learning_rate": 7.027928570874794e-07, + "loss": 0.78711617, + "num_input_tokens_seen": 130977060, + "step": 6095, + "time_per_iteration": 2.6360597610473633 + }, + { + "auxiliary_loss_clip": 0.01167429, + "auxiliary_loss_mlp": 0.0102611, + "balance_loss_clip": 1.04983652, + "balance_loss_mlp": 1.01877892, + "epoch": 0.7330006613358985, + "flos": 17858233422720.0, + "grad_norm": 1.875161061166255, + "language_loss": 0.85623765, + "learning_rate": 7.022000596042194e-07, + "loss": 0.87817305, + "num_input_tokens_seen": 130994160, + "step": 6096, + "time_per_iteration": 2.6488029956817627 + }, + { + "auxiliary_loss_clip": 0.01099406, + "auxiliary_loss_mlp": 0.01025384, + "balance_loss_clip": 0.93127197, + "balance_loss_mlp": 1.01820517, + "epoch": 0.7331209042265376, + "flos": 22492074343680.0, + "grad_norm": 2.609357572931416, + "language_loss": 0.81898427, + "learning_rate": 7.016074589961784e-07, + "loss": 0.84023225, + "num_input_tokens_seen": 131012725, + "step": 6097, + "time_per_iteration": 3.655366897583008 + }, + { + "auxiliary_loss_clip": 0.01121739, + "auxiliary_loss_mlp": 0.0102894, + "balance_loss_clip": 0.97751176, + "balance_loss_mlp": 1.02189815, + "epoch": 0.7332411471171767, + "flos": 33072757937280.0, + "grad_norm": 1.9720105078666876, + "language_loss": 0.67163444, + "learning_rate": 7.01015055353253e-07, + "loss": 0.69314122, + "num_input_tokens_seen": 131035150, + "step": 6098, + "time_per_iteration": 2.828397512435913 + }, + { + "auxiliary_loss_clip": 0.01067041, + "auxiliary_loss_mlp": 0.01032987, + "balance_loss_clip": 0.89804435, + "balance_loss_mlp": 1.02523255, + "epoch": 0.7333613900078157, + "flos": 22743017735040.0, + "grad_norm": 1.8458443544407717, + "language_loss": 0.77869445, + "learning_rate": 7.004228487653123e-07, + "loss": 0.79969466, + "num_input_tokens_seen": 131055955, + "step": 6099, + "time_per_iteration": 3.7191162109375 + }, + { + "auxiliary_loss_clip": 0.01093656, + "auxiliary_loss_mlp": 0.01027991, + "balance_loss_clip": 0.9293468, + "balance_loss_mlp": 1.02041483, + "epoch": 0.7334816328984549, + "flos": 22346133384960.0, + "grad_norm": 1.9167150401458615, + "language_loss": 0.78201413, + "learning_rate": 6.998308393221906e-07, + "loss": 0.80323064, + "num_input_tokens_seen": 131074360, + "step": 6100, + "time_per_iteration": 2.7127017974853516 + }, + { + "auxiliary_loss_clip": 0.01100414, + "auxiliary_loss_mlp": 0.01027012, + "balance_loss_clip": 0.93539673, + "balance_loss_mlp": 1.01986217, + "epoch": 0.733601875789094, + "flos": 20736149984640.0, + "grad_norm": 2.1160284293234732, + "language_loss": 0.71139657, + "learning_rate": 6.992390271136977e-07, + "loss": 0.73267078, + "num_input_tokens_seen": 131090070, + "step": 6101, + "time_per_iteration": 2.7155933380126953 + }, + { + "auxiliary_loss_clip": 0.01137215, + "auxiliary_loss_mlp": 0.01026621, + "balance_loss_clip": 1.0083878, + "balance_loss_mlp": 1.01961732, + "epoch": 0.733722118679733, + "flos": 22564362464640.0, + "grad_norm": 1.648617708596045, + "language_loss": 0.85791391, + "learning_rate": 6.986474122296094e-07, + "loss": 0.87955225, + "num_input_tokens_seen": 131109185, + "step": 6102, + "time_per_iteration": 2.6535027027130127 + }, + { + "auxiliary_loss_clip": 0.01172569, + "auxiliary_loss_mlp": 0.0102451, + "balance_loss_clip": 1.05196667, + "balance_loss_mlp": 1.01659417, + "epoch": 0.7338423615703722, + "flos": 20084192179200.0, + "grad_norm": 1.8681512660127626, + "language_loss": 0.72471094, + "learning_rate": 6.980559947596751e-07, + "loss": 0.74668169, + "num_input_tokens_seen": 131127725, + "step": 6103, + "time_per_iteration": 2.6695873737335205 + }, + { + "auxiliary_loss_clip": 0.01074076, + "auxiliary_loss_mlp": 0.01024966, + "balance_loss_clip": 0.89641416, + "balance_loss_mlp": 1.01770639, + "epoch": 0.7339626044610112, + "flos": 21687675217920.0, + "grad_norm": 2.051079271574274, + "language_loss": 0.75695992, + "learning_rate": 6.974647747936109e-07, + "loss": 0.77795041, + "num_input_tokens_seen": 131146110, + "step": 6104, + "time_per_iteration": 2.7632570266723633 + }, + { + "auxiliary_loss_clip": 0.01169741, + "auxiliary_loss_mlp": 0.00509577, + "balance_loss_clip": 1.05157495, + "balance_loss_mlp": 0.0, + "epoch": 0.7340828473516503, + "flos": 15268248282240.0, + "grad_norm": 2.1456755036380737, + "language_loss": 0.82638407, + "learning_rate": 6.968737524211039e-07, + "loss": 0.84317726, + "num_input_tokens_seen": 131162920, + "step": 6105, + "time_per_iteration": 2.6167452335357666 + }, + { + "auxiliary_loss_clip": 0.01145261, + "auxiliary_loss_mlp": 0.01024835, + "balance_loss_clip": 1.01416337, + "balance_loss_mlp": 1.01791453, + "epoch": 0.7342030902422895, + "flos": 22930112701440.0, + "grad_norm": 2.695529285346433, + "language_loss": 0.80397749, + "learning_rate": 6.962829277318132e-07, + "loss": 0.82567847, + "num_input_tokens_seen": 131182515, + "step": 6106, + "time_per_iteration": 3.657550811767578 + }, + { + "auxiliary_loss_clip": 0.01146616, + "auxiliary_loss_mlp": 0.01028788, + "balance_loss_clip": 1.01395226, + "balance_loss_mlp": 1.0217371, + "epoch": 0.7343233331329285, + "flos": 25847890381440.0, + "grad_norm": 3.790119861805121, + "language_loss": 0.83508301, + "learning_rate": 6.956923008153652e-07, + "loss": 0.85683703, + "num_input_tokens_seen": 131202280, + "step": 6107, + "time_per_iteration": 2.6792540550231934 + }, + { + "auxiliary_loss_clip": 0.01145259, + "auxiliary_loss_mlp": 0.01025411, + "balance_loss_clip": 1.00945735, + "balance_loss_mlp": 1.01810062, + "epoch": 0.7344435760235676, + "flos": 18478985287680.0, + "grad_norm": 2.5540520568441942, + "language_loss": 0.83971024, + "learning_rate": 6.951018717613593e-07, + "loss": 0.861417, + "num_input_tokens_seen": 131221295, + "step": 6108, + "time_per_iteration": 2.6732633113861084 + }, + { + "auxiliary_loss_clip": 0.01142824, + "auxiliary_loss_mlp": 0.01030211, + "balance_loss_clip": 1.01126182, + "balance_loss_mlp": 1.02280521, + "epoch": 0.7345638189142067, + "flos": 17640040256640.0, + "grad_norm": 1.8175119092850347, + "language_loss": 0.78578115, + "learning_rate": 6.945116406593614e-07, + "loss": 0.80751151, + "num_input_tokens_seen": 131240150, + "step": 6109, + "time_per_iteration": 2.6668450832366943 + }, + { + "auxiliary_loss_clip": 0.01078753, + "auxiliary_loss_mlp": 0.0102679, + "balance_loss_clip": 0.8985635, + "balance_loss_mlp": 1.01909471, + "epoch": 0.7346840618048458, + "flos": 20260225756800.0, + "grad_norm": 2.267403603109074, + "language_loss": 0.74108446, + "learning_rate": 6.939216075989089e-07, + "loss": 0.76213992, + "num_input_tokens_seen": 131258080, + "step": 6110, + "time_per_iteration": 2.6787900924682617 + }, + { + "auxiliary_loss_clip": 0.01119086, + "auxiliary_loss_mlp": 0.01022328, + "balance_loss_clip": 0.97272092, + "balance_loss_mlp": 1.01534247, + "epoch": 0.7348043046954849, + "flos": 29023183641600.0, + "grad_norm": 2.057369485018251, + "language_loss": 0.66331017, + "learning_rate": 6.933317726695109e-07, + "loss": 0.68472433, + "num_input_tokens_seen": 131279310, + "step": 6111, + "time_per_iteration": 2.67868709564209 + }, + { + "auxiliary_loss_clip": 0.01097574, + "auxiliary_loss_mlp": 0.01030293, + "balance_loss_clip": 0.93913162, + "balance_loss_mlp": 1.02265739, + "epoch": 0.734924547586124, + "flos": 17931203902080.0, + "grad_norm": 2.566970243517358, + "language_loss": 0.79854804, + "learning_rate": 6.92742135960644e-07, + "loss": 0.81982672, + "num_input_tokens_seen": 131297010, + "step": 6112, + "time_per_iteration": 2.7884867191314697 + }, + { + "auxiliary_loss_clip": 0.01045097, + "auxiliary_loss_mlp": 0.00999564, + "balance_loss_clip": 0.9746424, + "balance_loss_mlp": 0.99859858, + "epoch": 0.7350447904767631, + "flos": 63588319850880.0, + "grad_norm": 0.8186706166768355, + "language_loss": 0.55668598, + "learning_rate": 6.921526975617556e-07, + "loss": 0.57713258, + "num_input_tokens_seen": 131356470, + "step": 6113, + "time_per_iteration": 3.2681708335876465 + }, + { + "auxiliary_loss_clip": 0.01122754, + "auxiliary_loss_mlp": 0.0102882, + "balance_loss_clip": 0.970222, + "balance_loss_mlp": 1.02099097, + "epoch": 0.7351650333674021, + "flos": 21580015178880.0, + "grad_norm": 1.8005117244968682, + "language_loss": 0.75139296, + "learning_rate": 6.915634575622631e-07, + "loss": 0.77290869, + "num_input_tokens_seen": 131374985, + "step": 6114, + "time_per_iteration": 2.684793472290039 + }, + { + "auxiliary_loss_clip": 0.01166547, + "auxiliary_loss_mlp": 0.01027101, + "balance_loss_clip": 1.04920137, + "balance_loss_mlp": 1.01950169, + "epoch": 0.7352852762580413, + "flos": 18186349184640.0, + "grad_norm": 2.3023713401651986, + "language_loss": 0.70710325, + "learning_rate": 6.909744160515532e-07, + "loss": 0.72903967, + "num_input_tokens_seen": 131393125, + "step": 6115, + "time_per_iteration": 2.5915279388427734 + }, + { + "auxiliary_loss_clip": 0.01121426, + "auxiliary_loss_mlp": 0.01022785, + "balance_loss_clip": 0.97590721, + "balance_loss_mlp": 1.0154326, + "epoch": 0.7354055191486804, + "flos": 38910073063680.0, + "grad_norm": 4.76741299611063, + "language_loss": 0.69558996, + "learning_rate": 6.903855731189849e-07, + "loss": 0.71703207, + "num_input_tokens_seen": 131415760, + "step": 6116, + "time_per_iteration": 2.823133945465088 + }, + { + "auxiliary_loss_clip": 0.01124692, + "auxiliary_loss_mlp": 0.01027687, + "balance_loss_clip": 0.9719643, + "balance_loss_mlp": 1.02028155, + "epoch": 0.7355257620393194, + "flos": 16289978647680.0, + "grad_norm": 2.65510196135695, + "language_loss": 0.81891525, + "learning_rate": 6.897969288538825e-07, + "loss": 0.84043908, + "num_input_tokens_seen": 131433705, + "step": 6117, + "time_per_iteration": 2.6889021396636963 + }, + { + "auxiliary_loss_clip": 0.01116862, + "auxiliary_loss_mlp": 0.01023398, + "balance_loss_clip": 0.97178674, + "balance_loss_mlp": 1.01614976, + "epoch": 0.7356460049299585, + "flos": 18114240631680.0, + "grad_norm": 1.7268753037447735, + "language_loss": 0.8137657, + "learning_rate": 6.892084833455452e-07, + "loss": 0.83516824, + "num_input_tokens_seen": 131453275, + "step": 6118, + "time_per_iteration": 2.7145729064941406 + }, + { + "auxiliary_loss_clip": 0.01142407, + "auxiliary_loss_mlp": 0.01026265, + "balance_loss_clip": 1.01270151, + "balance_loss_mlp": 1.01903772, + "epoch": 0.7357662478205976, + "flos": 21325193118720.0, + "grad_norm": 1.403759431882962, + "language_loss": 0.83870214, + "learning_rate": 6.886202366832384e-07, + "loss": 0.86038888, + "num_input_tokens_seen": 131474960, + "step": 6119, + "time_per_iteration": 2.6489880084991455 + }, + { + "auxiliary_loss_clip": 0.01075344, + "auxiliary_loss_mlp": 0.0102709, + "balance_loss_clip": 0.89913344, + "balance_loss_mlp": 1.01943064, + "epoch": 0.7358864907112367, + "flos": 14246841139200.0, + "grad_norm": 1.757112533998284, + "language_loss": 0.73654556, + "learning_rate": 6.880321889561987e-07, + "loss": 0.75756985, + "num_input_tokens_seen": 131492935, + "step": 6120, + "time_per_iteration": 3.6828906536102295 + }, + { + "auxiliary_loss_clip": 0.01089344, + "auxiliary_loss_mlp": 0.0102745, + "balance_loss_clip": 0.93180645, + "balance_loss_mlp": 1.0188787, + "epoch": 0.7360067336018757, + "flos": 22309684058880.0, + "grad_norm": 2.3311658191599314, + "language_loss": 0.64856994, + "learning_rate": 6.874443402536338e-07, + "loss": 0.66973794, + "num_input_tokens_seen": 131512025, + "step": 6121, + "time_per_iteration": 2.7446019649505615 + }, + { + "auxiliary_loss_clip": 0.01125384, + "auxiliary_loss_mlp": 0.01028599, + "balance_loss_clip": 0.97404277, + "balance_loss_mlp": 1.02094567, + "epoch": 0.7361269764925149, + "flos": 25554607833600.0, + "grad_norm": 1.6045565391816372, + "language_loss": 0.80255461, + "learning_rate": 6.868566906647177e-07, + "loss": 0.82409453, + "num_input_tokens_seen": 131532975, + "step": 6122, + "time_per_iteration": 2.718618869781494 + }, + { + "auxiliary_loss_clip": 0.01145205, + "auxiliary_loss_mlp": 0.01030285, + "balance_loss_clip": 1.01213551, + "balance_loss_mlp": 1.02299213, + "epoch": 0.736247219383154, + "flos": 20376505059840.0, + "grad_norm": 1.8150034017659262, + "language_loss": 0.83656144, + "learning_rate": 6.862692402785984e-07, + "loss": 0.85831642, + "num_input_tokens_seen": 131553225, + "step": 6123, + "time_per_iteration": 3.624889612197876 + }, + { + "auxiliary_loss_clip": 0.0102531, + "auxiliary_loss_mlp": 0.01005453, + "balance_loss_clip": 0.92662734, + "balance_loss_mlp": 1.00414741, + "epoch": 0.736367462273793, + "flos": 70339525735680.0, + "grad_norm": 0.6885369677345473, + "language_loss": 0.4961018, + "learning_rate": 6.856819891843899e-07, + "loss": 0.5164094, + "num_input_tokens_seen": 131617930, + "step": 6124, + "time_per_iteration": 3.3643033504486084 + }, + { + "auxiliary_loss_clip": 0.01051863, + "auxiliary_loss_mlp": 0.0102409, + "balance_loss_clip": 0.85876882, + "balance_loss_mlp": 1.01683664, + "epoch": 0.7364877051644322, + "flos": 22412711243520.0, + "grad_norm": 1.954767123523509, + "language_loss": 0.72057289, + "learning_rate": 6.8509493747118e-07, + "loss": 0.74133241, + "num_input_tokens_seen": 131636740, + "step": 6125, + "time_per_iteration": 3.696303367614746 + }, + { + "auxiliary_loss_clip": 0.01169514, + "auxiliary_loss_mlp": 0.0102878, + "balance_loss_clip": 1.05244303, + "balance_loss_mlp": 1.02094233, + "epoch": 0.7366079480550712, + "flos": 12130266274560.0, + "grad_norm": 2.0321531409481404, + "language_loss": 0.88410652, + "learning_rate": 6.845080852280221e-07, + "loss": 0.90608943, + "num_input_tokens_seen": 131653810, + "step": 6126, + "time_per_iteration": 2.595865249633789 + }, + { + "auxiliary_loss_clip": 0.01097966, + "auxiliary_loss_mlp": 0.01026406, + "balance_loss_clip": 0.93388307, + "balance_loss_mlp": 1.01979625, + "epoch": 0.7367281909457103, + "flos": 15049336844160.0, + "grad_norm": 4.200018785980926, + "language_loss": 0.74402118, + "learning_rate": 6.839214325439409e-07, + "loss": 0.76526493, + "num_input_tokens_seen": 131671505, + "step": 6127, + "time_per_iteration": 2.7283859252929688 + }, + { + "auxiliary_loss_clip": 0.01116741, + "auxiliary_loss_mlp": 0.01023814, + "balance_loss_clip": 0.97517794, + "balance_loss_mlp": 1.01686728, + "epoch": 0.7368484338363495, + "flos": 23510752053120.0, + "grad_norm": 1.5860934731394203, + "language_loss": 0.7206856, + "learning_rate": 6.833349795079327e-07, + "loss": 0.74209112, + "num_input_tokens_seen": 131690615, + "step": 6128, + "time_per_iteration": 2.6752583980560303 + }, + { + "auxiliary_loss_clip": 0.01096647, + "auxiliary_loss_mlp": 0.01027212, + "balance_loss_clip": 0.9370721, + "balance_loss_mlp": 1.01986623, + "epoch": 0.7369686767269885, + "flos": 27417833095680.0, + "grad_norm": 4.2959232501625975, + "language_loss": 0.68732452, + "learning_rate": 6.827487262089613e-07, + "loss": 0.70856315, + "num_input_tokens_seen": 131711120, + "step": 6129, + "time_per_iteration": 2.7760822772979736 + }, + { + "auxiliary_loss_clip": 0.01021819, + "auxiliary_loss_mlp": 0.01002851, + "balance_loss_clip": 0.93863785, + "balance_loss_mlp": 1.00186777, + "epoch": 0.7370889196176276, + "flos": 70293343824000.0, + "grad_norm": 0.8888859023377677, + "language_loss": 0.56812876, + "learning_rate": 6.821626727359606e-07, + "loss": 0.58837551, + "num_input_tokens_seen": 131776680, + "step": 6130, + "time_per_iteration": 3.281632423400879 + }, + { + "auxiliary_loss_clip": 0.01125611, + "auxiliary_loss_mlp": 0.01028869, + "balance_loss_clip": 0.98070979, + "balance_loss_mlp": 1.0215497, + "epoch": 0.7372091625082667, + "flos": 18040839189120.0, + "grad_norm": 2.000197394686929, + "language_loss": 0.77329004, + "learning_rate": 6.815768191778348e-07, + "loss": 0.79483485, + "num_input_tokens_seen": 131794760, + "step": 6131, + "time_per_iteration": 2.6839137077331543 + }, + { + "auxiliary_loss_clip": 0.01139755, + "auxiliary_loss_mlp": 0.01028967, + "balance_loss_clip": 1.01012397, + "balance_loss_mlp": 1.02154076, + "epoch": 0.7373294053989058, + "flos": 33726331854720.0, + "grad_norm": 2.363262794422057, + "language_loss": 0.72807288, + "learning_rate": 6.809911656234569e-07, + "loss": 0.74976015, + "num_input_tokens_seen": 131816735, + "step": 6132, + "time_per_iteration": 3.6679885387420654 + }, + { + "auxiliary_loss_clip": 0.01098318, + "auxiliary_loss_mlp": 0.01027728, + "balance_loss_clip": 0.93150795, + "balance_loss_mlp": 1.02109671, + "epoch": 0.7374496482895448, + "flos": 21506326427520.0, + "grad_norm": 1.9049488846580795, + "language_loss": 0.7825194, + "learning_rate": 6.804057121616707e-07, + "loss": 0.80377984, + "num_input_tokens_seen": 131834940, + "step": 6133, + "time_per_iteration": 2.705024480819702 + }, + { + "auxiliary_loss_clip": 0.01144483, + "auxiliary_loss_mlp": 0.01023538, + "balance_loss_clip": 1.01175749, + "balance_loss_mlp": 1.01632595, + "epoch": 0.737569891180184, + "flos": 24936908624640.0, + "grad_norm": 1.7748130008968885, + "language_loss": 0.72113496, + "learning_rate": 6.798204588812888e-07, + "loss": 0.7428152, + "num_input_tokens_seen": 131854355, + "step": 6134, + "time_per_iteration": 2.6691439151763916 + }, + { + "auxiliary_loss_clip": 0.01040166, + "auxiliary_loss_mlp": 0.00509717, + "balance_loss_clip": 0.85441428, + "balance_loss_mlp": 0.0, + "epoch": 0.7376901340708231, + "flos": 20664544222080.0, + "grad_norm": 1.637711124392384, + "language_loss": 0.7566036, + "learning_rate": 6.792354058710937e-07, + "loss": 0.77210242, + "num_input_tokens_seen": 131871825, + "step": 6135, + "time_per_iteration": 2.7980868816375732 + }, + { + "auxiliary_loss_clip": 0.01161511, + "auxiliary_loss_mlp": 0.01025548, + "balance_loss_clip": 1.04775977, + "balance_loss_mlp": 1.01905966, + "epoch": 0.7378103769614621, + "flos": 23805794367360.0, + "grad_norm": 1.8197993518009132, + "language_loss": 0.65237725, + "learning_rate": 6.786505532198374e-07, + "loss": 0.67424786, + "num_input_tokens_seen": 131890770, + "step": 6136, + "time_per_iteration": 2.623934507369995 + }, + { + "auxiliary_loss_clip": 0.01169126, + "auxiliary_loss_mlp": 0.01028575, + "balance_loss_clip": 1.05096769, + "balance_loss_mlp": 1.02127671, + "epoch": 0.7379306198521013, + "flos": 22237216369920.0, + "grad_norm": 3.8328218282379196, + "language_loss": 0.851282, + "learning_rate": 6.780659010162411e-07, + "loss": 0.87325901, + "num_input_tokens_seen": 131909720, + "step": 6137, + "time_per_iteration": 2.6875898838043213 + }, + { + "auxiliary_loss_clip": 0.01101345, + "auxiliary_loss_mlp": 0.01028243, + "balance_loss_clip": 0.936261, + "balance_loss_mlp": 1.02085757, + "epoch": 0.7380508627427403, + "flos": 14903108576640.0, + "grad_norm": 1.5655083200450863, + "language_loss": 0.8323071, + "learning_rate": 6.774814493489975e-07, + "loss": 0.85360301, + "num_input_tokens_seen": 131927395, + "step": 6138, + "time_per_iteration": 2.6754324436187744 + }, + { + "auxiliary_loss_clip": 0.01139985, + "auxiliary_loss_mlp": 0.01024327, + "balance_loss_clip": 1.01057625, + "balance_loss_mlp": 1.01744866, + "epoch": 0.7381711056333794, + "flos": 21685843624320.0, + "grad_norm": 1.8803038888784385, + "language_loss": 0.66088688, + "learning_rate": 6.768971983067655e-07, + "loss": 0.68253005, + "num_input_tokens_seen": 131947725, + "step": 6139, + "time_per_iteration": 2.6572415828704834 + }, + { + "auxiliary_loss_clip": 0.01065504, + "auxiliary_loss_mlp": 0.01000146, + "balance_loss_clip": 1.01313269, + "balance_loss_mlp": 0.99914497, + "epoch": 0.7382913485240186, + "flos": 52404263596800.0, + "grad_norm": 1.0728803404667517, + "language_loss": 0.67786068, + "learning_rate": 6.763131479781772e-07, + "loss": 0.69851726, + "num_input_tokens_seen": 131997485, + "step": 6140, + "time_per_iteration": 3.0050246715545654 + }, + { + "auxiliary_loss_clip": 0.01114611, + "auxiliary_loss_mlp": 0.01022086, + "balance_loss_clip": 0.97450167, + "balance_loss_mlp": 1.01518059, + "epoch": 0.7384115914146576, + "flos": 21798818876160.0, + "grad_norm": 2.1638960827628133, + "language_loss": 0.76394475, + "learning_rate": 6.757292984518316e-07, + "loss": 0.7853117, + "num_input_tokens_seen": 132016885, + "step": 6141, + "time_per_iteration": 2.682215452194214 + }, + { + "auxiliary_loss_clip": 0.01048168, + "auxiliary_loss_mlp": 0.010003, + "balance_loss_clip": 0.97782397, + "balance_loss_mlp": 0.99928635, + "epoch": 0.7385318343052967, + "flos": 61494331662720.0, + "grad_norm": 0.7394121808023839, + "language_loss": 0.56412399, + "learning_rate": 6.751456498162981e-07, + "loss": 0.58460867, + "num_input_tokens_seen": 132075920, + "step": 6142, + "time_per_iteration": 3.122246742248535 + }, + { + "auxiliary_loss_clip": 0.01139889, + "auxiliary_loss_mlp": 0.01025911, + "balance_loss_clip": 1.00696611, + "balance_loss_mlp": 1.01891029, + "epoch": 0.7386520771959358, + "flos": 17013757697280.0, + "grad_norm": 1.7996084196218538, + "language_loss": 0.85674691, + "learning_rate": 6.745622021601174e-07, + "loss": 0.87840497, + "num_input_tokens_seen": 132092945, + "step": 6143, + "time_per_iteration": 2.6558892726898193 + }, + { + "auxiliary_loss_clip": 0.01099057, + "auxiliary_loss_mlp": 0.01027078, + "balance_loss_clip": 0.93475556, + "balance_loss_mlp": 1.0196898, + "epoch": 0.7387723200865749, + "flos": 18770759464320.0, + "grad_norm": 3.1084054665203245, + "language_loss": 0.69955128, + "learning_rate": 6.739789555717954e-07, + "loss": 0.72081262, + "num_input_tokens_seen": 132109920, + "step": 6144, + "time_per_iteration": 2.7054145336151123 + }, + { + "auxiliary_loss_clip": 0.01165281, + "auxiliary_loss_mlp": 0.01024889, + "balance_loss_clip": 1.04886258, + "balance_loss_mlp": 1.01730156, + "epoch": 0.738892562977214, + "flos": 22525542840960.0, + "grad_norm": 1.9349428669797986, + "language_loss": 0.77387726, + "learning_rate": 6.733959101398124e-07, + "loss": 0.79577893, + "num_input_tokens_seen": 132128050, + "step": 6145, + "time_per_iteration": 2.6026298999786377 + }, + { + "auxiliary_loss_clip": 0.01117792, + "auxiliary_loss_mlp": 0.01024384, + "balance_loss_clip": 0.97080874, + "balance_loss_mlp": 1.01727319, + "epoch": 0.7390128058678531, + "flos": 21501478091520.0, + "grad_norm": 2.1373925272457703, + "language_loss": 0.81656939, + "learning_rate": 6.728130659526143e-07, + "loss": 0.83799118, + "num_input_tokens_seen": 132145860, + "step": 6146, + "time_per_iteration": 3.554851531982422 + }, + { + "auxiliary_loss_clip": 0.0112418, + "auxiliary_loss_mlp": 0.01028407, + "balance_loss_clip": 0.9749195, + "balance_loss_mlp": 1.02066994, + "epoch": 0.7391330487584922, + "flos": 25776176878080.0, + "grad_norm": 2.069596823645374, + "language_loss": 0.71023434, + "learning_rate": 6.7223042309862e-07, + "loss": 0.7317602, + "num_input_tokens_seen": 132166060, + "step": 6147, + "time_per_iteration": 2.7130141258239746 + }, + { + "auxiliary_loss_clip": 0.01138348, + "auxiliary_loss_mlp": 0.01023587, + "balance_loss_clip": 1.00818932, + "balance_loss_mlp": 1.01650262, + "epoch": 0.7392532916491312, + "flos": 28366736636160.0, + "grad_norm": 1.907177765671887, + "language_loss": 0.73877525, + "learning_rate": 6.716479816662144e-07, + "loss": 0.76039457, + "num_input_tokens_seen": 132187790, + "step": 6148, + "time_per_iteration": 2.695775270462036 + }, + { + "auxiliary_loss_clip": 0.01123088, + "auxiliary_loss_mlp": 0.01025625, + "balance_loss_clip": 0.97194731, + "balance_loss_mlp": 1.01884246, + "epoch": 0.7393735345397703, + "flos": 23585877348480.0, + "grad_norm": 1.7351952423056856, + "language_loss": 0.72726589, + "learning_rate": 6.710657417437531e-07, + "loss": 0.74875307, + "num_input_tokens_seen": 132207495, + "step": 6149, + "time_per_iteration": 3.681119918823242 + }, + { + "auxiliary_loss_clip": 0.01119718, + "auxiliary_loss_mlp": 0.01023577, + "balance_loss_clip": 0.97233605, + "balance_loss_mlp": 1.01636767, + "epoch": 0.7394937774304094, + "flos": 19974772373760.0, + "grad_norm": 4.049657636203485, + "language_loss": 0.80241382, + "learning_rate": 6.704837034195628e-07, + "loss": 0.82384682, + "num_input_tokens_seen": 132225960, + "step": 6150, + "time_per_iteration": 2.636399030685425 + }, + { + "auxiliary_loss_clip": 0.01136797, + "auxiliary_loss_mlp": 0.0102683, + "balance_loss_clip": 1.0094552, + "balance_loss_mlp": 1.01911116, + "epoch": 0.7396140203210485, + "flos": 23478037741440.0, + "grad_norm": 1.8679493529582858, + "language_loss": 0.84732717, + "learning_rate": 6.699018667819376e-07, + "loss": 0.86896342, + "num_input_tokens_seen": 132245360, + "step": 6151, + "time_per_iteration": 3.640026092529297 + }, + { + "auxiliary_loss_clip": 0.01137417, + "auxiliary_loss_mlp": 0.01025916, + "balance_loss_clip": 1.00813842, + "balance_loss_mlp": 1.01833093, + "epoch": 0.7397342632116876, + "flos": 25555433846400.0, + "grad_norm": 1.7191420356278546, + "language_loss": 0.73076653, + "learning_rate": 6.693202319191415e-07, + "loss": 0.75239986, + "num_input_tokens_seen": 132267095, + "step": 6152, + "time_per_iteration": 2.6665842533111572 + }, + { + "auxiliary_loss_clip": 0.01165888, + "auxiliary_loss_mlp": 0.01025012, + "balance_loss_clip": 1.0512445, + "balance_loss_mlp": 1.01804113, + "epoch": 0.7398545061023267, + "flos": 24755021130240.0, + "grad_norm": 1.6987931794075035, + "language_loss": 0.74834788, + "learning_rate": 6.687387989194084e-07, + "loss": 0.77025688, + "num_input_tokens_seen": 132286610, + "step": 6153, + "time_per_iteration": 2.7190184593200684 + }, + { + "auxiliary_loss_clip": 0.01120888, + "auxiliary_loss_mlp": 0.01030395, + "balance_loss_clip": 0.97788262, + "balance_loss_mlp": 1.02292061, + "epoch": 0.7399747489929658, + "flos": 16508602776960.0, + "grad_norm": 1.8349255315373973, + "language_loss": 0.79401755, + "learning_rate": 6.681575678709404e-07, + "loss": 0.81553042, + "num_input_tokens_seen": 132305300, + "step": 6154, + "time_per_iteration": 2.6814253330230713 + }, + { + "auxiliary_loss_clip": 0.01141936, + "auxiliary_loss_mlp": 0.01026147, + "balance_loss_clip": 1.01083934, + "balance_loss_mlp": 1.01902986, + "epoch": 0.7400949918836048, + "flos": 24097065753600.0, + "grad_norm": 1.8731923319821604, + "language_loss": 0.70722657, + "learning_rate": 6.67576538861911e-07, + "loss": 0.72890735, + "num_input_tokens_seen": 132323875, + "step": 6155, + "time_per_iteration": 2.6767239570617676 + }, + { + "auxiliary_loss_clip": 0.01118959, + "auxiliary_loss_mlp": 0.01023649, + "balance_loss_clip": 0.97310233, + "balance_loss_mlp": 1.01655924, + "epoch": 0.740215234774244, + "flos": 21802517976960.0, + "grad_norm": 1.6412619309844005, + "language_loss": 0.82345867, + "learning_rate": 6.669957119804612e-07, + "loss": 0.84488475, + "num_input_tokens_seen": 132345510, + "step": 6156, + "time_per_iteration": 2.733687162399292 + }, + { + "auxiliary_loss_clip": 0.01124835, + "auxiliary_loss_mlp": 0.01022546, + "balance_loss_clip": 0.97207248, + "balance_loss_mlp": 1.01516986, + "epoch": 0.7403354776648831, + "flos": 18733196816640.0, + "grad_norm": 4.696862370021132, + "language_loss": 0.73027647, + "learning_rate": 6.66415087314702e-07, + "loss": 0.75175023, + "num_input_tokens_seen": 132360465, + "step": 6157, + "time_per_iteration": 2.635392189025879 + }, + { + "auxiliary_loss_clip": 0.01120396, + "auxiliary_loss_mlp": 0.01023353, + "balance_loss_clip": 0.97083956, + "balance_loss_mlp": 1.01549089, + "epoch": 0.7404557205555221, + "flos": 16909581277440.0, + "grad_norm": 2.0114159488646943, + "language_loss": 0.73307025, + "learning_rate": 6.65834664952714e-07, + "loss": 0.75450778, + "num_input_tokens_seen": 132377915, + "step": 6158, + "time_per_iteration": 3.570845603942871 + }, + { + "auxiliary_loss_clip": 0.0109777, + "auxiliary_loss_mlp": 0.01021476, + "balance_loss_clip": 0.93388665, + "balance_loss_mlp": 1.01446033, + "epoch": 0.7405759634461613, + "flos": 21214408596480.0, + "grad_norm": 1.8633874817551483, + "language_loss": 0.76064062, + "learning_rate": 6.652544449825457e-07, + "loss": 0.78183311, + "num_input_tokens_seen": 132398170, + "step": 6159, + "time_per_iteration": 2.731156349182129 + }, + { + "auxiliary_loss_clip": 0.01126595, + "auxiliary_loss_mlp": 0.01028688, + "balance_loss_clip": 0.97092324, + "balance_loss_mlp": 1.02109396, + "epoch": 0.7406962063368003, + "flos": 20480106862080.0, + "grad_norm": 1.5424286443880029, + "language_loss": 0.76459247, + "learning_rate": 6.646744274922182e-07, + "loss": 0.78614527, + "num_input_tokens_seen": 132416615, + "step": 6160, + "time_per_iteration": 2.684758186340332 + }, + { + "auxiliary_loss_clip": 0.01120469, + "auxiliary_loss_mlp": 0.01028697, + "balance_loss_clip": 0.9715122, + "balance_loss_mlp": 1.02180946, + "epoch": 0.7408164492274394, + "flos": 19791915212160.0, + "grad_norm": 2.4490430984894345, + "language_loss": 0.75429809, + "learning_rate": 6.640946125697171e-07, + "loss": 0.77578974, + "num_input_tokens_seen": 132434145, + "step": 6161, + "time_per_iteration": 2.654428243637085 + }, + { + "auxiliary_loss_clip": 0.01143785, + "auxiliary_loss_mlp": 0.01028772, + "balance_loss_clip": 1.00948, + "balance_loss_mlp": 1.02077293, + "epoch": 0.7409366921180786, + "flos": 29204855654400.0, + "grad_norm": 1.9478466337084934, + "language_loss": 0.75680399, + "learning_rate": 6.635150003030017e-07, + "loss": 0.77852952, + "num_input_tokens_seen": 132452670, + "step": 6162, + "time_per_iteration": 2.6818082332611084 + }, + { + "auxiliary_loss_clip": 0.01073245, + "auxiliary_loss_mlp": 0.01029132, + "balance_loss_clip": 0.89391196, + "balance_loss_mlp": 1.02205658, + "epoch": 0.7410569350087176, + "flos": 22930004960640.0, + "grad_norm": 2.065506073828293, + "language_loss": 0.8582319, + "learning_rate": 6.629355907799981e-07, + "loss": 0.87925571, + "num_input_tokens_seen": 132472475, + "step": 6163, + "time_per_iteration": 2.7465837001800537 + }, + { + "auxiliary_loss_clip": 0.01145359, + "auxiliary_loss_mlp": 0.01026108, + "balance_loss_clip": 1.01047242, + "balance_loss_mlp": 1.01846075, + "epoch": 0.7411771778993567, + "flos": 30440397726720.0, + "grad_norm": 1.6889262952403823, + "language_loss": 0.69136691, + "learning_rate": 6.623563840886015e-07, + "loss": 0.71308154, + "num_input_tokens_seen": 132493400, + "step": 6164, + "time_per_iteration": 2.7958619594573975 + }, + { + "auxiliary_loss_clip": 0.01137786, + "auxiliary_loss_mlp": 0.01022281, + "balance_loss_clip": 1.0086875, + "balance_loss_mlp": 1.0155189, + "epoch": 0.7412974207899958, + "flos": 20522050968960.0, + "grad_norm": 1.5711548137434355, + "language_loss": 0.6972186, + "learning_rate": 6.617773803166795e-07, + "loss": 0.71881926, + "num_input_tokens_seen": 132511725, + "step": 6165, + "time_per_iteration": 2.632084846496582 + }, + { + "auxiliary_loss_clip": 0.01124511, + "auxiliary_loss_mlp": 0.00509645, + "balance_loss_clip": 0.973333, + "balance_loss_mlp": 0.0, + "epoch": 0.7414176636806349, + "flos": 22090700793600.0, + "grad_norm": 2.5913442351467766, + "language_loss": 0.81974459, + "learning_rate": 6.611985795520634e-07, + "loss": 0.83608609, + "num_input_tokens_seen": 132530270, + "step": 6166, + "time_per_iteration": 2.7237322330474854 + }, + { + "auxiliary_loss_clip": 0.01105562, + "auxiliary_loss_mlp": 0.01020409, + "balance_loss_clip": 0.93503976, + "balance_loss_mlp": 1.01325941, + "epoch": 0.7415379065712739, + "flos": 25155245445120.0, + "grad_norm": 1.9503733084188308, + "language_loss": 0.77690613, + "learning_rate": 6.606199818825588e-07, + "loss": 0.79816586, + "num_input_tokens_seen": 132550725, + "step": 6167, + "time_per_iteration": 2.822666645050049 + }, + { + "auxiliary_loss_clip": 0.01123958, + "auxiliary_loss_mlp": 0.01029051, + "balance_loss_clip": 0.97104365, + "balance_loss_mlp": 1.02198172, + "epoch": 0.7416581494619131, + "flos": 16871731320960.0, + "grad_norm": 2.004713163920257, + "language_loss": 0.8183108, + "learning_rate": 6.600415873959377e-07, + "loss": 0.83984089, + "num_input_tokens_seen": 132568600, + "step": 6168, + "time_per_iteration": 2.609647274017334 + }, + { + "auxiliary_loss_clip": 0.01052038, + "auxiliary_loss_mlp": 0.00508559, + "balance_loss_clip": 0.85907108, + "balance_loss_mlp": 0.0, + "epoch": 0.7417783923525522, + "flos": 28438881102720.0, + "grad_norm": 1.8921346879619596, + "language_loss": 0.64688385, + "learning_rate": 6.594633961799437e-07, + "loss": 0.66248977, + "num_input_tokens_seen": 132587640, + "step": 6169, + "time_per_iteration": 2.846083641052246 + }, + { + "auxiliary_loss_clip": 0.01104538, + "auxiliary_loss_mlp": 0.01028539, + "balance_loss_clip": 0.93449426, + "balance_loss_mlp": 1.02152371, + "epoch": 0.7418986352431912, + "flos": 20084299920000.0, + "grad_norm": 1.611479310240867, + "language_loss": 0.81934392, + "learning_rate": 6.588854083222857e-07, + "loss": 0.84067464, + "num_input_tokens_seen": 132607075, + "step": 6170, + "time_per_iteration": 2.7101588249206543 + }, + { + "auxiliary_loss_clip": 0.01123924, + "auxiliary_loss_mlp": 0.01028978, + "balance_loss_clip": 0.97450799, + "balance_loss_mlp": 1.02112842, + "epoch": 0.7420188781338304, + "flos": 18259571059200.0, + "grad_norm": 1.889130906547326, + "language_loss": 0.8115983, + "learning_rate": 6.583076239106444e-07, + "loss": 0.83312738, + "num_input_tokens_seen": 132625580, + "step": 6171, + "time_per_iteration": 2.777817487716675 + }, + { + "auxiliary_loss_clip": 0.01125025, + "auxiliary_loss_mlp": 0.01023668, + "balance_loss_clip": 0.97223049, + "balance_loss_mlp": 1.01646757, + "epoch": 0.7421391210244694, + "flos": 13771994319360.0, + "grad_norm": 2.149238697548494, + "language_loss": 0.75436229, + "learning_rate": 6.577300430326707e-07, + "loss": 0.77584922, + "num_input_tokens_seen": 132640525, + "step": 6172, + "time_per_iteration": 3.569164514541626 + }, + { + "auxiliary_loss_clip": 0.01094683, + "auxiliary_loss_mlp": 0.01023609, + "balance_loss_clip": 0.93592578, + "balance_loss_mlp": 1.01704979, + "epoch": 0.7422593639151085, + "flos": 15961683317760.0, + "grad_norm": 1.958098887417122, + "language_loss": 0.72095227, + "learning_rate": 6.571526657759821e-07, + "loss": 0.74213517, + "num_input_tokens_seen": 132656265, + "step": 6173, + "time_per_iteration": 2.6747584342956543 + }, + { + "auxiliary_loss_clip": 0.01137819, + "auxiliary_loss_mlp": 0.01023271, + "balance_loss_clip": 1.00922203, + "balance_loss_mlp": 1.01571882, + "epoch": 0.7423796068057477, + "flos": 30114400867200.0, + "grad_norm": 1.633065564801813, + "language_loss": 0.70959139, + "learning_rate": 6.565754922281663e-07, + "loss": 0.73120224, + "num_input_tokens_seen": 132678510, + "step": 6174, + "time_per_iteration": 3.6688950061798096 + }, + { + "auxiliary_loss_clip": 0.01118472, + "auxiliary_loss_mlp": 0.01027683, + "balance_loss_clip": 0.97094822, + "balance_loss_mlp": 1.02003884, + "epoch": 0.7424998496963867, + "flos": 20521907314560.0, + "grad_norm": 2.0680375832130706, + "language_loss": 0.78415489, + "learning_rate": 6.559985224767801e-07, + "loss": 0.80561644, + "num_input_tokens_seen": 132696385, + "step": 6175, + "time_per_iteration": 2.7123939990997314 + }, + { + "auxiliary_loss_clip": 0.01102032, + "auxiliary_loss_mlp": 0.0102776, + "balance_loss_clip": 0.93429601, + "balance_loss_mlp": 1.02069688, + "epoch": 0.7426200925870258, + "flos": 21871573873920.0, + "grad_norm": 2.6302938383055743, + "language_loss": 0.75584221, + "learning_rate": 6.55421756609349e-07, + "loss": 0.77714014, + "num_input_tokens_seen": 132714640, + "step": 6176, + "time_per_iteration": 2.6986851692199707 + }, + { + "auxiliary_loss_clip": 0.01146011, + "auxiliary_loss_mlp": 0.01026353, + "balance_loss_clip": 1.01710415, + "balance_loss_mlp": 1.01887918, + "epoch": 0.7427403354776649, + "flos": 26432049265920.0, + "grad_norm": 2.7079498285923265, + "language_loss": 0.78825569, + "learning_rate": 6.54845194713369e-07, + "loss": 0.80997932, + "num_input_tokens_seen": 132735590, + "step": 6177, + "time_per_iteration": 3.637134075164795 + }, + { + "auxiliary_loss_clip": 0.01140186, + "auxiliary_loss_mlp": 0.01028109, + "balance_loss_clip": 1.01084304, + "balance_loss_mlp": 1.02102804, + "epoch": 0.742860578368304, + "flos": 19898390102400.0, + "grad_norm": 1.8804914880005523, + "language_loss": 0.79589319, + "learning_rate": 6.542688368763034e-07, + "loss": 0.81757617, + "num_input_tokens_seen": 132753995, + "step": 6178, + "time_per_iteration": 2.673935651779175 + }, + { + "auxiliary_loss_clip": 0.01141267, + "auxiliary_loss_mlp": 0.01028716, + "balance_loss_clip": 1.01144505, + "balance_loss_mlp": 1.02165294, + "epoch": 0.742980821258943, + "flos": 24827201510400.0, + "grad_norm": 1.6205103222778254, + "language_loss": 0.77022648, + "learning_rate": 6.536926831855854e-07, + "loss": 0.79192626, + "num_input_tokens_seen": 132773160, + "step": 6179, + "time_per_iteration": 2.720323085784912 + }, + { + "auxiliary_loss_clip": 0.01117072, + "auxiliary_loss_mlp": 0.01025451, + "balance_loss_clip": 0.97425056, + "balance_loss_mlp": 1.01901126, + "epoch": 0.7431010641495821, + "flos": 25228646887680.0, + "grad_norm": 2.4542761223338845, + "language_loss": 0.73127156, + "learning_rate": 6.531167337286165e-07, + "loss": 0.75269675, + "num_input_tokens_seen": 132793180, + "step": 6180, + "time_per_iteration": 2.708954095840454 + }, + { + "auxiliary_loss_clip": 0.01120139, + "auxiliary_loss_mlp": 0.01024665, + "balance_loss_clip": 0.97442639, + "balance_loss_mlp": 1.01664281, + "epoch": 0.7432213070402213, + "flos": 21762369550080.0, + "grad_norm": 1.5110373229334173, + "language_loss": 0.7966243, + "learning_rate": 6.52540988592768e-07, + "loss": 0.81807238, + "num_input_tokens_seen": 132814200, + "step": 6181, + "time_per_iteration": 2.7570765018463135 + }, + { + "auxiliary_loss_clip": 0.01122043, + "auxiliary_loss_mlp": 0.01022359, + "balance_loss_clip": 0.9720602, + "balance_loss_mlp": 1.01541507, + "epoch": 0.7433415499308603, + "flos": 14793832425600.0, + "grad_norm": 2.1168721895328835, + "language_loss": 0.83468711, + "learning_rate": 6.519654478653814e-07, + "loss": 0.8561312, + "num_input_tokens_seen": 132832565, + "step": 6182, + "time_per_iteration": 2.7537262439727783 + }, + { + "auxiliary_loss_clip": 0.01028989, + "auxiliary_loss_mlp": 0.00999929, + "balance_loss_clip": 0.93917167, + "balance_loss_mlp": 0.99886245, + "epoch": 0.7434617928214994, + "flos": 67155577297920.0, + "grad_norm": 0.7472788190811077, + "language_loss": 0.56155425, + "learning_rate": 6.51390111633763e-07, + "loss": 0.58184344, + "num_input_tokens_seen": 132897840, + "step": 6183, + "time_per_iteration": 3.331674814224243 + }, + { + "auxiliary_loss_clip": 0.0104996, + "auxiliary_loss_mlp": 0.01028637, + "balance_loss_clip": 0.85571778, + "balance_loss_mlp": 1.02106142, + "epoch": 0.7435820357121385, + "flos": 27377576928000.0, + "grad_norm": 1.6393586624338559, + "language_loss": 0.76128745, + "learning_rate": 6.508149799851932e-07, + "loss": 0.78207344, + "num_input_tokens_seen": 132919505, + "step": 6184, + "time_per_iteration": 3.742133378982544 + }, + { + "auxiliary_loss_clip": 0.01114512, + "auxiliary_loss_mlp": 0.01023272, + "balance_loss_clip": 0.97143692, + "balance_loss_mlp": 1.01642358, + "epoch": 0.7437022786027776, + "flos": 23987645948160.0, + "grad_norm": 2.034820672080584, + "language_loss": 0.61640227, + "learning_rate": 6.502400530069183e-07, + "loss": 0.63778013, + "num_input_tokens_seen": 132939390, + "step": 6185, + "time_per_iteration": 2.6918139457702637 + }, + { + "auxiliary_loss_clip": 0.01097346, + "auxiliary_loss_mlp": 0.01030827, + "balance_loss_clip": 0.934457, + "balance_loss_mlp": 1.02352262, + "epoch": 0.7438225214934167, + "flos": 21866761451520.0, + "grad_norm": 1.9626965900908109, + "language_loss": 0.68641078, + "learning_rate": 6.496653307861535e-07, + "loss": 0.7076925, + "num_input_tokens_seen": 132960060, + "step": 6186, + "time_per_iteration": 2.774122476577759 + }, + { + "auxiliary_loss_clip": 0.01147596, + "auxiliary_loss_mlp": 0.01031566, + "balance_loss_clip": 1.01077485, + "balance_loss_mlp": 1.02422297, + "epoch": 0.7439427643840558, + "flos": 20230097224320.0, + "grad_norm": 2.072791879092256, + "language_loss": 0.65988755, + "learning_rate": 6.490908134100857e-07, + "loss": 0.68167913, + "num_input_tokens_seen": 132978525, + "step": 6187, + "time_per_iteration": 2.682725191116333 + }, + { + "auxiliary_loss_clip": 0.01148564, + "auxiliary_loss_mlp": 0.01030355, + "balance_loss_clip": 1.01221025, + "balance_loss_mlp": 1.02291024, + "epoch": 0.7440630072746949, + "flos": 20849915335680.0, + "grad_norm": 2.090439861415386, + "language_loss": 0.69399226, + "learning_rate": 6.48516500965866e-07, + "loss": 0.71578145, + "num_input_tokens_seen": 132998460, + "step": 6188, + "time_per_iteration": 2.725876808166504 + }, + { + "auxiliary_loss_clip": 0.01146351, + "auxiliary_loss_mlp": 0.01025086, + "balance_loss_clip": 1.00905752, + "balance_loss_mlp": 1.01760209, + "epoch": 0.7441832501653339, + "flos": 26503762769280.0, + "grad_norm": 1.7639978058694852, + "language_loss": 0.81769174, + "learning_rate": 6.479423935406192e-07, + "loss": 0.83940613, + "num_input_tokens_seen": 133018445, + "step": 6189, + "time_per_iteration": 2.7075464725494385 + }, + { + "auxiliary_loss_clip": 0.01018796, + "auxiliary_loss_mlp": 0.00998726, + "balance_loss_clip": 0.93999326, + "balance_loss_mlp": 0.99747485, + "epoch": 0.7443034930559731, + "flos": 68602848088320.0, + "grad_norm": 0.8090603708678632, + "language_loss": 0.61987293, + "learning_rate": 6.473684912214357e-07, + "loss": 0.64004821, + "num_input_tokens_seen": 133082005, + "step": 6190, + "time_per_iteration": 3.425109386444092 + }, + { + "auxiliary_loss_clip": 0.01145906, + "auxiliary_loss_mlp": 0.0102412, + "balance_loss_clip": 1.01393366, + "balance_loss_mlp": 1.01654458, + "epoch": 0.7444237359466122, + "flos": 18654982951680.0, + "grad_norm": 1.90684796230036, + "language_loss": 0.69833922, + "learning_rate": 6.467947940953778e-07, + "loss": 0.72003955, + "num_input_tokens_seen": 133100530, + "step": 6191, + "time_per_iteration": 2.7905285358428955 + }, + { + "auxiliary_loss_clip": 0.01120168, + "auxiliary_loss_mlp": 0.01025282, + "balance_loss_clip": 0.97146606, + "balance_loss_mlp": 1.01843083, + "epoch": 0.7445439788372512, + "flos": 22817604326400.0, + "grad_norm": 1.7692695782841474, + "language_loss": 0.72465217, + "learning_rate": 6.462213022494732e-07, + "loss": 0.74610674, + "num_input_tokens_seen": 133119775, + "step": 6192, + "time_per_iteration": 2.7262096405029297 + }, + { + "auxiliary_loss_clip": 0.01047174, + "auxiliary_loss_mlp": 0.01002403, + "balance_loss_clip": 0.97611535, + "balance_loss_mlp": 1.0013541, + "epoch": 0.7446642217278904, + "flos": 67045690615680.0, + "grad_norm": 0.7729767676544229, + "language_loss": 0.60994303, + "learning_rate": 6.456480157707201e-07, + "loss": 0.6304388, + "num_input_tokens_seen": 133184550, + "step": 6193, + "time_per_iteration": 3.187744140625 + }, + { + "auxiliary_loss_clip": 0.01092211, + "auxiliary_loss_mlp": 0.01027263, + "balance_loss_clip": 0.93290985, + "balance_loss_mlp": 1.02030098, + "epoch": 0.7447844646185294, + "flos": 17417465631360.0, + "grad_norm": 1.9481093611834215, + "language_loss": 0.8502512, + "learning_rate": 6.450749347460866e-07, + "loss": 0.87144589, + "num_input_tokens_seen": 133201525, + "step": 6194, + "time_per_iteration": 2.6740410327911377 + }, + { + "auxiliary_loss_clip": 0.01168111, + "auxiliary_loss_mlp": 0.01026762, + "balance_loss_clip": 1.04964757, + "balance_loss_mlp": 1.01905489, + "epoch": 0.7449047075091685, + "flos": 26615876094720.0, + "grad_norm": 1.867692574006875, + "language_loss": 0.79306197, + "learning_rate": 6.445020592625083e-07, + "loss": 0.81501067, + "num_input_tokens_seen": 133222175, + "step": 6195, + "time_per_iteration": 2.746541738510132 + }, + { + "auxiliary_loss_clip": 0.0116483, + "auxiliary_loss_mlp": 0.01027692, + "balance_loss_clip": 1.04777813, + "balance_loss_mlp": 1.02073646, + "epoch": 0.7450249503998077, + "flos": 14170458867840.0, + "grad_norm": 2.541319726387554, + "language_loss": 0.80230933, + "learning_rate": 6.4392938940689e-07, + "loss": 0.8242346, + "num_input_tokens_seen": 133237590, + "step": 6196, + "time_per_iteration": 2.695756435394287 + }, + { + "auxiliary_loss_clip": 0.0107104, + "auxiliary_loss_mlp": 0.00509772, + "balance_loss_clip": 0.8959406, + "balance_loss_mlp": 0.0, + "epoch": 0.7451451932904467, + "flos": 19606687752960.0, + "grad_norm": 2.2023690405529437, + "language_loss": 0.71232593, + "learning_rate": 6.433569252661049e-07, + "loss": 0.72813404, + "num_input_tokens_seen": 133255590, + "step": 6197, + "time_per_iteration": 2.761362075805664 + }, + { + "auxiliary_loss_clip": 0.01089845, + "auxiliary_loss_mlp": 0.01022824, + "balance_loss_clip": 0.93279946, + "balance_loss_mlp": 1.01595497, + "epoch": 0.7452654361810858, + "flos": 12495405980160.0, + "grad_norm": 1.8087378080509915, + "language_loss": 0.7114805, + "learning_rate": 6.427846669269952e-07, + "loss": 0.73260719, + "num_input_tokens_seen": 133273210, + "step": 6198, + "time_per_iteration": 3.5865304470062256 + }, + { + "auxiliary_loss_clip": 0.01170164, + "auxiliary_loss_mlp": 0.01029051, + "balance_loss_clip": 1.05243599, + "balance_loss_mlp": 1.02172017, + "epoch": 0.7453856790717249, + "flos": 22127329687680.0, + "grad_norm": 2.3985346055323697, + "language_loss": 0.8239401, + "learning_rate": 6.422126144763729e-07, + "loss": 0.84593225, + "num_input_tokens_seen": 133292600, + "step": 6199, + "time_per_iteration": 2.6145308017730713 + }, + { + "auxiliary_loss_clip": 0.0109204, + "auxiliary_loss_mlp": 0.00509637, + "balance_loss_clip": 0.92994446, + "balance_loss_mlp": 0.0, + "epoch": 0.745505921962364, + "flos": 20010682995840.0, + "grad_norm": 1.9700199830514065, + "language_loss": 0.7710197, + "learning_rate": 6.416407680010174e-07, + "loss": 0.78703648, + "num_input_tokens_seen": 133306960, + "step": 6200, + "time_per_iteration": 2.7142128944396973 + }, + { + "auxiliary_loss_clip": 0.01087192, + "auxiliary_loss_mlp": 0.01025899, + "balance_loss_clip": 0.89780235, + "balance_loss_mlp": 1.01833248, + "epoch": 0.745626164853003, + "flos": 24677884673280.0, + "grad_norm": 1.7649759003875527, + "language_loss": 0.80985492, + "learning_rate": 6.410691275876774e-07, + "loss": 0.83098578, + "num_input_tokens_seen": 133326380, + "step": 6201, + "time_per_iteration": 3.6766574382781982 + }, + { + "auxiliary_loss_clip": 0.01131167, + "auxiliary_loss_mlp": 0.0102865, + "balance_loss_clip": 0.9745571, + "balance_loss_mlp": 1.02153957, + "epoch": 0.7457464077436422, + "flos": 14538830797440.0, + "grad_norm": 2.88303435138702, + "language_loss": 0.76737785, + "learning_rate": 6.404976933230704e-07, + "loss": 0.78897601, + "num_input_tokens_seen": 133342900, + "step": 6202, + "time_per_iteration": 3.575712203979492 + }, + { + "auxiliary_loss_clip": 0.01127041, + "auxiliary_loss_mlp": 0.01027643, + "balance_loss_clip": 0.97503221, + "balance_loss_mlp": 1.0199064, + "epoch": 0.7458666506342813, + "flos": 34021194600960.0, + "grad_norm": 1.7617569832207898, + "language_loss": 0.72713405, + "learning_rate": 6.399264652938813e-07, + "loss": 0.74868089, + "num_input_tokens_seen": 133363805, + "step": 6203, + "time_per_iteration": 2.7950899600982666 + }, + { + "auxiliary_loss_clip": 0.01116554, + "auxiliary_loss_mlp": 0.01028887, + "balance_loss_clip": 0.96976042, + "balance_loss_mlp": 1.02067304, + "epoch": 0.7459868935249203, + "flos": 24279025075200.0, + "grad_norm": 1.9591297742134184, + "language_loss": 0.74700129, + "learning_rate": 6.393554435867679e-07, + "loss": 0.76845574, + "num_input_tokens_seen": 133384655, + "step": 6204, + "time_per_iteration": 2.7132177352905273 + }, + { + "auxiliary_loss_clip": 0.01091629, + "auxiliary_loss_mlp": 0.01027314, + "balance_loss_clip": 0.93092, + "balance_loss_mlp": 1.02019739, + "epoch": 0.7461071364155595, + "flos": 21908777385600.0, + "grad_norm": 2.113404686311601, + "language_loss": 0.83576214, + "learning_rate": 6.387846282883502e-07, + "loss": 0.85695153, + "num_input_tokens_seen": 133401185, + "step": 6205, + "time_per_iteration": 2.730015277862549 + }, + { + "auxiliary_loss_clip": 0.01164664, + "auxiliary_loss_mlp": 0.01025952, + "balance_loss_clip": 1.04874301, + "balance_loss_mlp": 1.0188086, + "epoch": 0.7462273793061985, + "flos": 22889712879360.0, + "grad_norm": 1.8187763715687688, + "language_loss": 0.77198696, + "learning_rate": 6.38214019485223e-07, + "loss": 0.7938931, + "num_input_tokens_seen": 133420010, + "step": 6206, + "time_per_iteration": 2.5988218784332275 + }, + { + "auxiliary_loss_clip": 0.01044394, + "auxiliary_loss_mlp": 0.01022464, + "balance_loss_clip": 0.85462677, + "balance_loss_mlp": 1.01491785, + "epoch": 0.7463476221968376, + "flos": 19968451580160.0, + "grad_norm": 2.086498314167365, + "language_loss": 0.71339113, + "learning_rate": 6.376436172639461e-07, + "loss": 0.73405969, + "num_input_tokens_seen": 133437855, + "step": 6207, + "time_per_iteration": 2.8325726985931396 + }, + { + "auxiliary_loss_clip": 0.0103203, + "auxiliary_loss_mlp": 0.01027511, + "balance_loss_clip": 0.82167697, + "balance_loss_mlp": 1.02000678, + "epoch": 0.7464678650874768, + "flos": 16836610798080.0, + "grad_norm": 2.213954738782324, + "language_loss": 0.654275, + "learning_rate": 6.370734217110487e-07, + "loss": 0.67487037, + "num_input_tokens_seen": 133456600, + "step": 6208, + "time_per_iteration": 2.831324577331543 + }, + { + "auxiliary_loss_clip": 0.0112547, + "auxiliary_loss_mlp": 0.01028612, + "balance_loss_clip": 0.97679096, + "balance_loss_mlp": 1.02063048, + "epoch": 0.7465881079781158, + "flos": 48100869843840.0, + "grad_norm": 1.429157164038023, + "language_loss": 0.64246404, + "learning_rate": 6.36503432913031e-07, + "loss": 0.66400486, + "num_input_tokens_seen": 133479745, + "step": 6209, + "time_per_iteration": 2.946666955947876 + }, + { + "auxiliary_loss_clip": 0.01142131, + "auxiliary_loss_mlp": 0.01023098, + "balance_loss_clip": 1.01051736, + "balance_loss_mlp": 1.01565921, + "epoch": 0.7467083508687549, + "flos": 19677359761920.0, + "grad_norm": 1.7828260324989662, + "language_loss": 0.68606025, + "learning_rate": 6.359336509563569e-07, + "loss": 0.70771247, + "num_input_tokens_seen": 133495765, + "step": 6210, + "time_per_iteration": 2.714498281478882 + }, + { + "auxiliary_loss_clip": 0.01086506, + "auxiliary_loss_mlp": 0.01025554, + "balance_loss_clip": 0.93346292, + "balance_loss_mlp": 1.0183835, + "epoch": 0.7468285937593939, + "flos": 17895436934400.0, + "grad_norm": 1.9866699659490372, + "language_loss": 0.80748379, + "learning_rate": 6.353640759274641e-07, + "loss": 0.82860434, + "num_input_tokens_seen": 133514655, + "step": 6211, + "time_per_iteration": 3.6011648178100586 + }, + { + "auxiliary_loss_clip": 0.01140541, + "auxiliary_loss_mlp": 0.01027212, + "balance_loss_clip": 1.00831175, + "balance_loss_mlp": 1.01982725, + "epoch": 0.7469488366500331, + "flos": 23141446369920.0, + "grad_norm": 3.346116476893223, + "language_loss": 0.75366783, + "learning_rate": 6.347947079127556e-07, + "loss": 0.77534539, + "num_input_tokens_seen": 133532555, + "step": 6212, + "time_per_iteration": 2.6748251914978027 + }, + { + "auxiliary_loss_clip": 0.01115927, + "auxiliary_loss_mlp": 0.01028716, + "balance_loss_clip": 0.97261232, + "balance_loss_mlp": 1.02154565, + "epoch": 0.7470690795406721, + "flos": 16690849407360.0, + "grad_norm": 2.1425748769865094, + "language_loss": 0.77001566, + "learning_rate": 6.342255469986053e-07, + "loss": 0.79146206, + "num_input_tokens_seen": 133551300, + "step": 6213, + "time_per_iteration": 2.6939070224761963 + }, + { + "auxiliary_loss_clip": 0.01168608, + "auxiliary_loss_mlp": 0.01023455, + "balance_loss_clip": 1.05144215, + "balance_loss_mlp": 1.01585555, + "epoch": 0.7471893224313112, + "flos": 25192700352000.0, + "grad_norm": 1.860665284129303, + "language_loss": 0.76547599, + "learning_rate": 6.336565932713533e-07, + "loss": 0.78739667, + "num_input_tokens_seen": 133570725, + "step": 6214, + "time_per_iteration": 2.6465060710906982 + }, + { + "auxiliary_loss_clip": 0.01121622, + "auxiliary_loss_mlp": 0.01025688, + "balance_loss_clip": 0.97805178, + "balance_loss_mlp": 1.01817226, + "epoch": 0.7473095653219504, + "flos": 22526225199360.0, + "grad_norm": 1.7102802781989654, + "language_loss": 0.77705061, + "learning_rate": 6.330878468173088e-07, + "loss": 0.79852366, + "num_input_tokens_seen": 133590790, + "step": 6215, + "time_per_iteration": 2.7075822353363037 + }, + { + "auxiliary_loss_clip": 0.0113699, + "auxiliary_loss_mlp": 0.01024352, + "balance_loss_clip": 1.00949788, + "balance_loss_mlp": 1.01699662, + "epoch": 0.7474298082125894, + "flos": 18113989236480.0, + "grad_norm": 1.7229388057440895, + "language_loss": 0.72650892, + "learning_rate": 6.32519307722752e-07, + "loss": 0.74812233, + "num_input_tokens_seen": 133608685, + "step": 6216, + "time_per_iteration": 2.643867015838623 + }, + { + "auxiliary_loss_clip": 0.01021681, + "auxiliary_loss_mlp": 0.01002607, + "balance_loss_clip": 0.92243111, + "balance_loss_mlp": 1.00137961, + "epoch": 0.7475500511032285, + "flos": 62086535193600.0, + "grad_norm": 0.878512990641867, + "language_loss": 0.54958093, + "learning_rate": 6.31950976073929e-07, + "loss": 0.56982386, + "num_input_tokens_seen": 133662775, + "step": 6217, + "time_per_iteration": 3.235473871231079 + }, + { + "auxiliary_loss_clip": 0.01070738, + "auxiliary_loss_mlp": 0.01019821, + "balance_loss_clip": 0.8986479, + "balance_loss_mlp": 1.01232839, + "epoch": 0.7476702939938676, + "flos": 17785586165760.0, + "grad_norm": 2.464690434326151, + "language_loss": 0.81217742, + "learning_rate": 6.31382851957055e-07, + "loss": 0.83308303, + "num_input_tokens_seen": 133679595, + "step": 6218, + "time_per_iteration": 2.772632122039795 + }, + { + "auxiliary_loss_clip": 0.01092321, + "auxiliary_loss_mlp": 0.00509809, + "balance_loss_clip": 0.93474972, + "balance_loss_mlp": 0.0, + "epoch": 0.7477905368845067, + "flos": 27927944092800.0, + "grad_norm": 2.3110608442776397, + "language_loss": 0.71954083, + "learning_rate": 6.308149354583143e-07, + "loss": 0.73556221, + "num_input_tokens_seen": 133699000, + "step": 6219, + "time_per_iteration": 2.763951063156128 + }, + { + "auxiliary_loss_clip": 0.01146965, + "auxiliary_loss_mlp": 0.01029391, + "balance_loss_clip": 1.01102614, + "balance_loss_mlp": 1.02227437, + "epoch": 0.7479107797751458, + "flos": 26870374932480.0, + "grad_norm": 2.7426369712134635, + "language_loss": 0.81944877, + "learning_rate": 6.302472266638586e-07, + "loss": 0.84121239, + "num_input_tokens_seen": 133719540, + "step": 6220, + "time_per_iteration": 2.695533514022827 + }, + { + "auxiliary_loss_clip": 0.01171959, + "auxiliary_loss_mlp": 0.01032393, + "balance_loss_clip": 1.0508225, + "balance_loss_mlp": 1.02434671, + "epoch": 0.7480310226657849, + "flos": 33943375785600.0, + "grad_norm": 1.9388562556308089, + "language_loss": 0.69859004, + "learning_rate": 6.296797256598101e-07, + "loss": 0.72063357, + "num_input_tokens_seen": 133741020, + "step": 6221, + "time_per_iteration": 2.701080322265625 + }, + { + "auxiliary_loss_clip": 0.01088208, + "auxiliary_loss_mlp": 0.01021965, + "balance_loss_clip": 0.9324432, + "balance_loss_mlp": 1.01445222, + "epoch": 0.748151265556424, + "flos": 24826555065600.0, + "grad_norm": 1.8733813030704463, + "language_loss": 0.81506467, + "learning_rate": 6.291124325322576e-07, + "loss": 0.83616638, + "num_input_tokens_seen": 133761145, + "step": 6222, + "time_per_iteration": 2.790743350982666 + }, + { + "auxiliary_loss_clip": 0.01124786, + "auxiliary_loss_mlp": 0.01020366, + "balance_loss_clip": 0.97224677, + "balance_loss_mlp": 1.01323128, + "epoch": 0.748271508447063, + "flos": 38399351535360.0, + "grad_norm": 1.6450835453185504, + "language_loss": 0.62303019, + "learning_rate": 6.285453473672595e-07, + "loss": 0.64448178, + "num_input_tokens_seen": 133783715, + "step": 6223, + "time_per_iteration": 2.8669936656951904 + }, + { + "auxiliary_loss_clip": 0.01165781, + "auxiliary_loss_mlp": 0.01027758, + "balance_loss_clip": 1.04868817, + "balance_loss_mlp": 1.02032566, + "epoch": 0.7483917513377022, + "flos": 21541842000000.0, + "grad_norm": 2.687859628550168, + "language_loss": 0.7520417, + "learning_rate": 6.279784702508415e-07, + "loss": 0.77397704, + "num_input_tokens_seen": 133804465, + "step": 6224, + "time_per_iteration": 3.5842435359954834 + }, + { + "auxiliary_loss_clip": 0.01010674, + "auxiliary_loss_mlp": 0.01003044, + "balance_loss_clip": 0.90140259, + "balance_loss_mlp": 1.00198352, + "epoch": 0.7485119942283412, + "flos": 62314532772480.0, + "grad_norm": 0.7907758788637281, + "language_loss": 0.58564085, + "learning_rate": 6.274118012689979e-07, + "loss": 0.60577804, + "num_input_tokens_seen": 133866365, + "step": 6225, + "time_per_iteration": 3.4226956367492676 + }, + { + "auxiliary_loss_clip": 0.01111508, + "auxiliary_loss_mlp": 0.01026972, + "balance_loss_clip": 0.97008598, + "balance_loss_mlp": 1.01978338, + "epoch": 0.7486322371189803, + "flos": 29937613104000.0, + "grad_norm": 1.466158337572158, + "language_loss": 0.68111211, + "learning_rate": 6.268453405076943e-07, + "loss": 0.70249689, + "num_input_tokens_seen": 133888760, + "step": 6226, + "time_per_iteration": 3.694051504135132 + }, + { + "auxiliary_loss_clip": 0.01120206, + "auxiliary_loss_mlp": 0.01028699, + "balance_loss_clip": 0.97116673, + "balance_loss_mlp": 1.02166915, + "epoch": 0.7487524800096195, + "flos": 18949414734720.0, + "grad_norm": 2.484664605384633, + "language_loss": 0.82443386, + "learning_rate": 6.262790880528592e-07, + "loss": 0.84592295, + "num_input_tokens_seen": 133906380, + "step": 6227, + "time_per_iteration": 2.678929567337036 + }, + { + "auxiliary_loss_clip": 0.01104149, + "auxiliary_loss_mlp": 0.01029259, + "balance_loss_clip": 0.930233, + "balance_loss_mlp": 1.02140355, + "epoch": 0.7488727229002585, + "flos": 18697393935360.0, + "grad_norm": 2.53347081402235, + "language_loss": 0.79800248, + "learning_rate": 6.257130439903951e-07, + "loss": 0.81933653, + "num_input_tokens_seen": 133922875, + "step": 6228, + "time_per_iteration": 3.624272108078003 + }, + { + "auxiliary_loss_clip": 0.01170179, + "auxiliary_loss_mlp": 0.01023608, + "balance_loss_clip": 1.05166268, + "balance_loss_mlp": 1.01615727, + "epoch": 0.7489929657908976, + "flos": 23623368168960.0, + "grad_norm": 1.7118632388698023, + "language_loss": 0.81209731, + "learning_rate": 6.251472084061695e-07, + "loss": 0.83403516, + "num_input_tokens_seen": 133941795, + "step": 6229, + "time_per_iteration": 2.6375732421875 + }, + { + "auxiliary_loss_clip": 0.01142468, + "auxiliary_loss_mlp": 0.01023236, + "balance_loss_clip": 1.01203108, + "balance_loss_mlp": 1.01584196, + "epoch": 0.7491132086815367, + "flos": 20551533056640.0, + "grad_norm": 1.865937675136106, + "language_loss": 0.88955581, + "learning_rate": 6.245815813860191e-07, + "loss": 0.91121286, + "num_input_tokens_seen": 133957305, + "step": 6230, + "time_per_iteration": 2.6273298263549805 + }, + { + "auxiliary_loss_clip": 0.01170259, + "auxiliary_loss_mlp": 0.01027077, + "balance_loss_clip": 1.04982877, + "balance_loss_mlp": 1.0193398, + "epoch": 0.7492334515721758, + "flos": 23003011353600.0, + "grad_norm": 2.4741134355048375, + "language_loss": 0.7043575, + "learning_rate": 6.240161630157495e-07, + "loss": 0.72633088, + "num_input_tokens_seen": 133976660, + "step": 6231, + "time_per_iteration": 2.6156327724456787 + }, + { + "auxiliary_loss_clip": 0.01170004, + "auxiliary_loss_mlp": 0.01027204, + "balance_loss_clip": 1.05111575, + "balance_loss_mlp": 1.02006602, + "epoch": 0.7493536944628149, + "flos": 16398823835520.0, + "grad_norm": 2.6257411807766253, + "language_loss": 0.69996172, + "learning_rate": 6.23450953381133e-07, + "loss": 0.72193378, + "num_input_tokens_seen": 133994750, + "step": 6232, + "time_per_iteration": 2.5894594192504883 + }, + { + "auxiliary_loss_clip": 0.01114507, + "auxiliary_loss_mlp": 0.01027106, + "balance_loss_clip": 0.97085208, + "balance_loss_mlp": 1.02034342, + "epoch": 0.749473937353454, + "flos": 15338561155200.0, + "grad_norm": 2.002993677070503, + "language_loss": 0.68109834, + "learning_rate": 6.228859525679131e-07, + "loss": 0.70251441, + "num_input_tokens_seen": 134009165, + "step": 6233, + "time_per_iteration": 2.623809337615967 + }, + { + "auxiliary_loss_clip": 0.01142577, + "auxiliary_loss_mlp": 0.01022816, + "balance_loss_clip": 1.01047802, + "balance_loss_mlp": 1.01601243, + "epoch": 0.7495941802440931, + "flos": 18951138587520.0, + "grad_norm": 2.085682549604827, + "language_loss": 0.79795808, + "learning_rate": 6.223211606617986e-07, + "loss": 0.81961197, + "num_input_tokens_seen": 134027585, + "step": 6234, + "time_per_iteration": 2.666095018386841 + }, + { + "auxiliary_loss_clip": 0.01140413, + "auxiliary_loss_mlp": 0.0102742, + "balance_loss_clip": 1.01275134, + "balance_loss_mlp": 1.02095902, + "epoch": 0.7497144231347321, + "flos": 22492469393280.0, + "grad_norm": 1.7950605775721054, + "language_loss": 0.84116077, + "learning_rate": 6.217565777484701e-07, + "loss": 0.8628391, + "num_input_tokens_seen": 134046680, + "step": 6235, + "time_per_iteration": 2.6449155807495117 + }, + { + "auxiliary_loss_clip": 0.01120139, + "auxiliary_loss_mlp": 0.00509293, + "balance_loss_clip": 0.97562563, + "balance_loss_mlp": 0.0, + "epoch": 0.7498346660253713, + "flos": 24243509502720.0, + "grad_norm": 2.3587722729491127, + "language_loss": 0.80353296, + "learning_rate": 6.211922039135722e-07, + "loss": 0.81982726, + "num_input_tokens_seen": 134066825, + "step": 6236, + "time_per_iteration": 3.641669988632202 + }, + { + "auxiliary_loss_clip": 0.01167899, + "auxiliary_loss_mlp": 0.01022344, + "balance_loss_clip": 1.05060506, + "balance_loss_mlp": 1.01551294, + "epoch": 0.7499549089160104, + "flos": 24387080163840.0, + "grad_norm": 1.8506187681778583, + "language_loss": 0.81093085, + "learning_rate": 6.206280392427201e-07, + "loss": 0.83283329, + "num_input_tokens_seen": 134086410, + "step": 6237, + "time_per_iteration": 2.6477975845336914 + }, + { + "auxiliary_loss_clip": 0.01133872, + "auxiliary_loss_mlp": 0.01029665, + "balance_loss_clip": 1.00679874, + "balance_loss_mlp": 1.02288842, + "epoch": 0.7500751518066494, + "flos": 34057320704640.0, + "grad_norm": 1.5004442928230013, + "language_loss": 0.7388047, + "learning_rate": 6.200640838214983e-07, + "loss": 0.76044011, + "num_input_tokens_seen": 134109185, + "step": 6238, + "time_per_iteration": 2.761423110961914 + }, + { + "auxiliary_loss_clip": 0.01163539, + "auxiliary_loss_mlp": 0.01022399, + "balance_loss_clip": 1.04775739, + "balance_loss_mlp": 1.01580071, + "epoch": 0.7501953946972886, + "flos": 18843586289280.0, + "grad_norm": 1.8544273151974078, + "language_loss": 0.67036629, + "learning_rate": 6.195003377354578e-07, + "loss": 0.69222569, + "num_input_tokens_seen": 134128455, + "step": 6239, + "time_per_iteration": 2.6431732177734375 + }, + { + "auxiliary_loss_clip": 0.01139621, + "auxiliary_loss_mlp": 0.0102139, + "balance_loss_clip": 1.00878358, + "balance_loss_mlp": 1.01408863, + "epoch": 0.7503156375879276, + "flos": 20257675891200.0, + "grad_norm": 2.274004706014524, + "language_loss": 0.73730469, + "learning_rate": 6.189368010701183e-07, + "loss": 0.75891483, + "num_input_tokens_seen": 134145515, + "step": 6240, + "time_per_iteration": 2.6934945583343506 + }, + { + "auxiliary_loss_clip": 0.01146946, + "auxiliary_loss_mlp": 0.01025021, + "balance_loss_clip": 1.0092833, + "balance_loss_mlp": 1.01774883, + "epoch": 0.7504358804785667, + "flos": 13480040574720.0, + "grad_norm": 1.9148723472885216, + "language_loss": 0.76411605, + "learning_rate": 6.183734739109683e-07, + "loss": 0.78583574, + "num_input_tokens_seen": 134163335, + "step": 6241, + "time_per_iteration": 2.6260056495666504 + }, + { + "auxiliary_loss_clip": 0.01148825, + "auxiliary_loss_mlp": 0.01027338, + "balance_loss_clip": 1.01092732, + "balance_loss_mlp": 1.01965499, + "epoch": 0.7505561233692057, + "flos": 29461042431360.0, + "grad_norm": 1.9645068782140718, + "language_loss": 0.69152015, + "learning_rate": 6.178103563434629e-07, + "loss": 0.71328175, + "num_input_tokens_seen": 134182335, + "step": 6242, + "time_per_iteration": 2.687269687652588 + }, + { + "auxiliary_loss_clip": 0.01164929, + "auxiliary_loss_mlp": 0.01022357, + "balance_loss_clip": 1.04818344, + "balance_loss_mlp": 1.01536536, + "epoch": 0.7506763662598449, + "flos": 20302457172480.0, + "grad_norm": 1.7658123715320089, + "language_loss": 0.83937967, + "learning_rate": 6.172474484530283e-07, + "loss": 0.86125255, + "num_input_tokens_seen": 134201070, + "step": 6243, + "time_per_iteration": 2.65395450592041 + }, + { + "auxiliary_loss_clip": 0.01107994, + "auxiliary_loss_mlp": 0.01022488, + "balance_loss_clip": 0.96805215, + "balance_loss_mlp": 1.01514482, + "epoch": 0.750796609150484, + "flos": 37230961939200.0, + "grad_norm": 7.711651144875797, + "language_loss": 0.76287758, + "learning_rate": 6.166847503250563e-07, + "loss": 0.78418243, + "num_input_tokens_seen": 134223310, + "step": 6244, + "time_per_iteration": 2.783688545227051 + }, + { + "auxiliary_loss_clip": 0.01119909, + "auxiliary_loss_mlp": 0.01022455, + "balance_loss_clip": 0.97212625, + "balance_loss_mlp": 1.01539207, + "epoch": 0.750916852041123, + "flos": 19609417186560.0, + "grad_norm": 2.2768267156289084, + "language_loss": 0.79082394, + "learning_rate": 6.161222620449078e-07, + "loss": 0.81224763, + "num_input_tokens_seen": 134242085, + "step": 6245, + "time_per_iteration": 2.6447622776031494 + }, + { + "auxiliary_loss_clip": 0.01102342, + "auxiliary_loss_mlp": 0.01030053, + "balance_loss_clip": 0.93529451, + "balance_loss_mlp": 1.02257562, + "epoch": 0.7510370949317622, + "flos": 25112690807040.0, + "grad_norm": 2.52204802615436, + "language_loss": 0.80872893, + "learning_rate": 6.155599836979117e-07, + "loss": 0.83005285, + "num_input_tokens_seen": 134260770, + "step": 6246, + "time_per_iteration": 2.7774219512939453 + }, + { + "auxiliary_loss_clip": 0.01071216, + "auxiliary_loss_mlp": 0.01029141, + "balance_loss_clip": 0.89432949, + "balance_loss_mlp": 1.02159166, + "epoch": 0.7511573378224012, + "flos": 19062282245760.0, + "grad_norm": 2.1253274823232315, + "language_loss": 0.81254148, + "learning_rate": 6.149979153693649e-07, + "loss": 0.83354503, + "num_input_tokens_seen": 134278025, + "step": 6247, + "time_per_iteration": 2.7124171257019043 + }, + { + "auxiliary_loss_clip": 0.01140404, + "auxiliary_loss_mlp": 0.01029086, + "balance_loss_clip": 1.00989687, + "balance_loss_mlp": 1.0214684, + "epoch": 0.7512775807130403, + "flos": 19937676602880.0, + "grad_norm": 2.1785592455120817, + "language_loss": 0.77177489, + "learning_rate": 6.144360571445343e-07, + "loss": 0.79346979, + "num_input_tokens_seen": 134297170, + "step": 6248, + "time_per_iteration": 2.6779139041900635 + }, + { + "auxiliary_loss_clip": 0.01140161, + "auxiliary_loss_mlp": 0.01027326, + "balance_loss_clip": 1.01162148, + "balance_loss_mlp": 1.01996505, + "epoch": 0.7513978236036795, + "flos": 20739920912640.0, + "grad_norm": 1.9139393791762425, + "language_loss": 0.79921418, + "learning_rate": 6.138744091086509e-07, + "loss": 0.82088906, + "num_input_tokens_seen": 134316755, + "step": 6249, + "time_per_iteration": 2.6783225536346436 + }, + { + "auxiliary_loss_clip": 0.01104087, + "auxiliary_loss_mlp": 0.01031372, + "balance_loss_clip": 0.93780494, + "balance_loss_mlp": 1.02416289, + "epoch": 0.7515180664943185, + "flos": 27563163523200.0, + "grad_norm": 2.151248319019155, + "language_loss": 0.72699928, + "learning_rate": 6.133129713469183e-07, + "loss": 0.74835384, + "num_input_tokens_seen": 134335960, + "step": 6250, + "time_per_iteration": 3.6101205348968506 + }, + { + "auxiliary_loss_clip": 0.01101917, + "auxiliary_loss_mlp": 0.01025507, + "balance_loss_clip": 0.93111765, + "balance_loss_mlp": 1.0184859, + "epoch": 0.7516383093849576, + "flos": 33803181002880.0, + "grad_norm": 1.9085158303201557, + "language_loss": 0.63763303, + "learning_rate": 6.127517439445053e-07, + "loss": 0.65890729, + "num_input_tokens_seen": 134356805, + "step": 6251, + "time_per_iteration": 2.862614631652832 + }, + { + "auxiliary_loss_clip": 0.01067738, + "auxiliary_loss_mlp": 0.01024675, + "balance_loss_clip": 0.89529037, + "balance_loss_mlp": 1.0178982, + "epoch": 0.7517585522755967, + "flos": 29746172592000.0, + "grad_norm": 1.8057164968778745, + "language_loss": 0.81497884, + "learning_rate": 6.121907269865498e-07, + "loss": 0.83590299, + "num_input_tokens_seen": 134376295, + "step": 6252, + "time_per_iteration": 2.8513216972351074 + }, + { + "auxiliary_loss_clip": 0.01001663, + "auxiliary_loss_mlp": 0.01002134, + "balance_loss_clip": 0.90048999, + "balance_loss_mlp": 1.00106132, + "epoch": 0.7518787951662358, + "flos": 69807974319360.0, + "grad_norm": 0.9566775728829922, + "language_loss": 0.67241508, + "learning_rate": 6.116299205581577e-07, + "loss": 0.69245303, + "num_input_tokens_seen": 134431125, + "step": 6253, + "time_per_iteration": 4.29931116104126 + }, + { + "auxiliary_loss_clip": 0.0117081, + "auxiliary_loss_mlp": 0.01029202, + "balance_loss_clip": 1.04907918, + "balance_loss_mlp": 1.02135491, + "epoch": 0.7519990380568748, + "flos": 34203225749760.0, + "grad_norm": 2.1956419108318, + "language_loss": 0.68668789, + "learning_rate": 6.110693247444018e-07, + "loss": 0.70868796, + "num_input_tokens_seen": 134452960, + "step": 6254, + "time_per_iteration": 3.7072529792785645 + }, + { + "auxiliary_loss_clip": 0.01086218, + "auxiliary_loss_mlp": 0.01025974, + "balance_loss_clip": 0.93213081, + "balance_loss_mlp": 1.01885104, + "epoch": 0.752119280947514, + "flos": 21725704742400.0, + "grad_norm": 1.76036078515694, + "language_loss": 0.82698667, + "learning_rate": 6.105089396303258e-07, + "loss": 0.84810865, + "num_input_tokens_seen": 134471350, + "step": 6255, + "time_per_iteration": 2.717372179031372 + }, + { + "auxiliary_loss_clip": 0.01119287, + "auxiliary_loss_mlp": 0.01029376, + "balance_loss_clip": 0.97093987, + "balance_loss_mlp": 1.02177668, + "epoch": 0.7522395238381531, + "flos": 32742774668160.0, + "grad_norm": 2.368504317201858, + "language_loss": 0.75751197, + "learning_rate": 6.099487653009383e-07, + "loss": 0.77899861, + "num_input_tokens_seen": 134490695, + "step": 6256, + "time_per_iteration": 2.7194862365722656 + }, + { + "auxiliary_loss_clip": 0.01140364, + "auxiliary_loss_mlp": 0.01025638, + "balance_loss_clip": 1.00779247, + "balance_loss_mlp": 1.01866126, + "epoch": 0.7523597667287921, + "flos": 23476026579840.0, + "grad_norm": 2.3706254682982744, + "language_loss": 0.83338702, + "learning_rate": 6.093888018412192e-07, + "loss": 0.85504699, + "num_input_tokens_seen": 134506885, + "step": 6257, + "time_per_iteration": 2.672675609588623 + }, + { + "auxiliary_loss_clip": 0.01046931, + "auxiliary_loss_mlp": 0.01003329, + "balance_loss_clip": 0.97575319, + "balance_loss_mlp": 1.00228643, + "epoch": 0.7524800096194313, + "flos": 67346730501120.0, + "grad_norm": 0.755732575357576, + "language_loss": 0.54665911, + "learning_rate": 6.088290493361125e-07, + "loss": 0.5671618, + "num_input_tokens_seen": 134571770, + "step": 6258, + "time_per_iteration": 3.3572463989257812 + }, + { + "auxiliary_loss_clip": 0.01068937, + "auxiliary_loss_mlp": 0.01027536, + "balance_loss_clip": 0.89742863, + "balance_loss_mlp": 1.02008569, + "epoch": 0.7526002525100703, + "flos": 13006055681280.0, + "grad_norm": 2.35247594485712, + "language_loss": 0.7173326, + "learning_rate": 6.082695078705322e-07, + "loss": 0.73829728, + "num_input_tokens_seen": 134589250, + "step": 6259, + "time_per_iteration": 2.772541046142578 + }, + { + "auxiliary_loss_clip": 0.01136626, + "auxiliary_loss_mlp": 0.01031695, + "balance_loss_clip": 1.00952744, + "balance_loss_mlp": 1.02389908, + "epoch": 0.7527204954007094, + "flos": 21397229844480.0, + "grad_norm": 2.871363275227667, + "language_loss": 0.68617356, + "learning_rate": 6.077101775293618e-07, + "loss": 0.70785683, + "num_input_tokens_seen": 134608075, + "step": 6260, + "time_per_iteration": 2.651548147201538 + }, + { + "auxiliary_loss_clip": 0.01145246, + "auxiliary_loss_mlp": 0.01026437, + "balance_loss_clip": 1.01100254, + "balance_loss_mlp": 1.0184499, + "epoch": 0.7528407382913486, + "flos": 18947188091520.0, + "grad_norm": 2.6717886633625083, + "language_loss": 0.82156402, + "learning_rate": 6.071510583974504e-07, + "loss": 0.84328091, + "num_input_tokens_seen": 134623260, + "step": 6261, + "time_per_iteration": 2.5964694023132324 + }, + { + "auxiliary_loss_clip": 0.01166698, + "auxiliary_loss_mlp": 0.01028032, + "balance_loss_clip": 1.04876101, + "balance_loss_mlp": 1.0207243, + "epoch": 0.7529609811819876, + "flos": 15231798956160.0, + "grad_norm": 1.9062923499972237, + "language_loss": 0.72042167, + "learning_rate": 6.065921505596161e-07, + "loss": 0.74236894, + "num_input_tokens_seen": 134641540, + "step": 6262, + "time_per_iteration": 3.523890733718872 + }, + { + "auxiliary_loss_clip": 0.01094382, + "auxiliary_loss_mlp": 0.01024035, + "balance_loss_clip": 0.93414867, + "balance_loss_mlp": 1.01639366, + "epoch": 0.7530812240726267, + "flos": 19354487385600.0, + "grad_norm": 1.5339875945521875, + "language_loss": 0.76940113, + "learning_rate": 6.060334541006445e-07, + "loss": 0.79058528, + "num_input_tokens_seen": 134660035, + "step": 6263, + "time_per_iteration": 2.7543153762817383 + }, + { + "auxiliary_loss_clip": 0.01097855, + "auxiliary_loss_mlp": 0.01026106, + "balance_loss_clip": 0.92992079, + "balance_loss_mlp": 1.01888239, + "epoch": 0.7532014669632658, + "flos": 27748247328000.0, + "grad_norm": 1.4493256662755911, + "language_loss": 0.69099522, + "learning_rate": 6.05474969105289e-07, + "loss": 0.7122348, + "num_input_tokens_seen": 134683025, + "step": 6264, + "time_per_iteration": 2.74859881401062 + }, + { + "auxiliary_loss_clip": 0.01145242, + "auxiliary_loss_mlp": 0.01029889, + "balance_loss_clip": 1.01166201, + "balance_loss_mlp": 1.02209282, + "epoch": 0.7533217098539049, + "flos": 14137421333760.0, + "grad_norm": 2.215201242907889, + "language_loss": 0.73790872, + "learning_rate": 6.049166956582725e-07, + "loss": 0.75965995, + "num_input_tokens_seen": 134701290, + "step": 6265, + "time_per_iteration": 2.650301933288574 + }, + { + "auxiliary_loss_clip": 0.01138073, + "auxiliary_loss_mlp": 0.01028703, + "balance_loss_clip": 1.00933838, + "balance_loss_mlp": 1.02172005, + "epoch": 0.753441952744544, + "flos": 26429068437120.0, + "grad_norm": 1.8821872869054401, + "language_loss": 0.87411237, + "learning_rate": 6.043586338442841e-07, + "loss": 0.89578015, + "num_input_tokens_seen": 134720345, + "step": 6266, + "time_per_iteration": 2.674849033355713 + }, + { + "auxiliary_loss_clip": 0.01162248, + "auxiliary_loss_mlp": 0.01022403, + "balance_loss_clip": 1.04960418, + "balance_loss_mlp": 1.01575387, + "epoch": 0.7535621956351831, + "flos": 23878621192320.0, + "grad_norm": 1.3999727678055247, + "language_loss": 0.73174626, + "learning_rate": 6.038007837479815e-07, + "loss": 0.75359279, + "num_input_tokens_seen": 134741450, + "step": 6267, + "time_per_iteration": 2.635359764099121 + }, + { + "auxiliary_loss_clip": 0.01140027, + "auxiliary_loss_mlp": 0.01026602, + "balance_loss_clip": 1.00945663, + "balance_loss_mlp": 1.01930594, + "epoch": 0.7536824385258222, + "flos": 21795873960960.0, + "grad_norm": 1.81199649101973, + "language_loss": 0.63996536, + "learning_rate": 6.032431454539897e-07, + "loss": 0.6616317, + "num_input_tokens_seen": 134760295, + "step": 6268, + "time_per_iteration": 2.6058030128479004 + }, + { + "auxiliary_loss_clip": 0.01098346, + "auxiliary_loss_mlp": 0.01022068, + "balance_loss_clip": 0.9362511, + "balance_loss_mlp": 1.01532125, + "epoch": 0.7538026814164612, + "flos": 28911644933760.0, + "grad_norm": 5.302221553053828, + "language_loss": 0.81207788, + "learning_rate": 6.026857190469014e-07, + "loss": 0.83328199, + "num_input_tokens_seen": 134782050, + "step": 6269, + "time_per_iteration": 2.796739339828491 + }, + { + "auxiliary_loss_clip": 0.0111921, + "auxiliary_loss_mlp": 0.01027489, + "balance_loss_clip": 0.97089797, + "balance_loss_mlp": 1.01993418, + "epoch": 0.7539229243071004, + "flos": 21104701482240.0, + "grad_norm": 1.8421348060220164, + "language_loss": 0.73872268, + "learning_rate": 6.0212850461128e-07, + "loss": 0.76018965, + "num_input_tokens_seen": 134801170, + "step": 6270, + "time_per_iteration": 2.682002544403076 + }, + { + "auxiliary_loss_clip": 0.01117897, + "auxiliary_loss_mlp": 0.01023323, + "balance_loss_clip": 0.96903837, + "balance_loss_mlp": 1.01614094, + "epoch": 0.7540431671977395, + "flos": 15158469340800.0, + "grad_norm": 2.576726505371197, + "language_loss": 0.74997413, + "learning_rate": 6.015715022316516e-07, + "loss": 0.77138633, + "num_input_tokens_seen": 134819150, + "step": 6271, + "time_per_iteration": 2.711113214492798 + }, + { + "auxiliary_loss_clip": 0.01073329, + "auxiliary_loss_mlp": 0.010281, + "balance_loss_clip": 0.89312845, + "balance_loss_mlp": 1.02046156, + "epoch": 0.7541634100883785, + "flos": 18770579896320.0, + "grad_norm": 2.409357854552185, + "language_loss": 0.78257787, + "learning_rate": 6.010147119925154e-07, + "loss": 0.80359221, + "num_input_tokens_seen": 134836905, + "step": 6272, + "time_per_iteration": 2.709639310836792 + }, + { + "auxiliary_loss_clip": 0.01089146, + "auxiliary_loss_mlp": 0.01023042, + "balance_loss_clip": 0.9324671, + "balance_loss_mlp": 1.016083, + "epoch": 0.7542836529790176, + "flos": 20594770053120.0, + "grad_norm": 2.4165378635792796, + "language_loss": 0.66523087, + "learning_rate": 6.004581339783348e-07, + "loss": 0.68635279, + "num_input_tokens_seen": 134855225, + "step": 6273, + "time_per_iteration": 2.698003053665161 + }, + { + "auxiliary_loss_clip": 0.0114614, + "auxiliary_loss_mlp": 0.01026412, + "balance_loss_clip": 1.01069808, + "balance_loss_mlp": 1.01839542, + "epoch": 0.7544038958696567, + "flos": 19095104298240.0, + "grad_norm": 3.084796201284738, + "language_loss": 0.68462723, + "learning_rate": 5.999017682735425e-07, + "loss": 0.70635277, + "num_input_tokens_seen": 134871615, + "step": 6274, + "time_per_iteration": 2.6674182415008545 + }, + { + "auxiliary_loss_clip": 0.01053997, + "auxiliary_loss_mlp": 0.0102432, + "balance_loss_clip": 0.85558093, + "balance_loss_mlp": 1.01641059, + "epoch": 0.7545241387602958, + "flos": 31723306859520.0, + "grad_norm": 1.9393057423352897, + "language_loss": 0.66513455, + "learning_rate": 5.993456149625387e-07, + "loss": 0.68591774, + "num_input_tokens_seen": 134892765, + "step": 6275, + "time_per_iteration": 2.833991289138794 + }, + { + "auxiliary_loss_clip": 0.01087512, + "auxiliary_loss_mlp": 0.01022543, + "balance_loss_clip": 0.93209863, + "balance_loss_mlp": 1.01506257, + "epoch": 0.7546443816509348, + "flos": 20296495514880.0, + "grad_norm": 1.9428047853937753, + "language_loss": 0.82618797, + "learning_rate": 5.987896741296909e-07, + "loss": 0.84728849, + "num_input_tokens_seen": 134910505, + "step": 6276, + "time_per_iteration": 3.627152442932129 + }, + { + "auxiliary_loss_clip": 0.01118512, + "auxiliary_loss_mlp": 0.0103721, + "balance_loss_clip": 0.97378588, + "balance_loss_mlp": 1.03021598, + "epoch": 0.754764624541574, + "flos": 23696159080320.0, + "grad_norm": 1.8625661274402678, + "language_loss": 0.7823869, + "learning_rate": 5.982339458593361e-07, + "loss": 0.80394417, + "num_input_tokens_seen": 134930445, + "step": 6277, + "time_per_iteration": 2.695251226425171 + }, + { + "auxiliary_loss_clip": 0.01137727, + "auxiliary_loss_mlp": 0.005087, + "balance_loss_clip": 1.00979185, + "balance_loss_mlp": 0.0, + "epoch": 0.7548848674322131, + "flos": 25337204766720.0, + "grad_norm": 1.6425895608845031, + "language_loss": 0.84162718, + "learning_rate": 5.976784302357767e-07, + "loss": 0.85809147, + "num_input_tokens_seen": 134951010, + "step": 6278, + "time_per_iteration": 2.7275848388671875 + }, + { + "auxiliary_loss_clip": 0.01145326, + "auxiliary_loss_mlp": 0.01026571, + "balance_loss_clip": 1.01173294, + "balance_loss_mlp": 1.01881683, + "epoch": 0.7550051103228521, + "flos": 19573147428480.0, + "grad_norm": 1.6508974039681845, + "language_loss": 0.73282349, + "learning_rate": 5.971231273432855e-07, + "loss": 0.75454241, + "num_input_tokens_seen": 134970495, + "step": 6279, + "time_per_iteration": 3.554612874984741 + }, + { + "auxiliary_loss_clip": 0.01045699, + "auxiliary_loss_mlp": 0.01000924, + "balance_loss_clip": 0.97582608, + "balance_loss_mlp": 0.99986345, + "epoch": 0.7551253532134913, + "flos": 64150068648960.0, + "grad_norm": 0.8096997502595579, + "language_loss": 0.54554909, + "learning_rate": 5.965680372661e-07, + "loss": 0.5660153, + "num_input_tokens_seen": 135028060, + "step": 6280, + "time_per_iteration": 3.9808735847473145 + }, + { + "auxiliary_loss_clip": 0.01119899, + "auxiliary_loss_mlp": 0.01022748, + "balance_loss_clip": 0.97206402, + "balance_loss_mlp": 1.01626348, + "epoch": 0.7552455961041303, + "flos": 26067986968320.0, + "grad_norm": 1.7971117708275266, + "language_loss": 0.5618766, + "learning_rate": 5.960131600884266e-07, + "loss": 0.58330309, + "num_input_tokens_seen": 135047330, + "step": 6281, + "time_per_iteration": 2.720773696899414 + }, + { + "auxiliary_loss_clip": 0.0109873, + "auxiliary_loss_mlp": 0.0102302, + "balance_loss_clip": 0.93288946, + "balance_loss_mlp": 1.01601362, + "epoch": 0.7553658389947694, + "flos": 24498223822080.0, + "grad_norm": 1.7368902405160411, + "language_loss": 0.76018333, + "learning_rate": 5.954584958944413e-07, + "loss": 0.7814008, + "num_input_tokens_seen": 135065995, + "step": 6282, + "time_per_iteration": 2.7026968002319336 + }, + { + "auxiliary_loss_clip": 0.01099443, + "auxiliary_loss_mlp": 0.00509559, + "balance_loss_clip": 0.93238032, + "balance_loss_mlp": 0.0, + "epoch": 0.7554860818854086, + "flos": 21799465320960.0, + "grad_norm": 1.7416384189247174, + "language_loss": 0.81751025, + "learning_rate": 5.949040447682854e-07, + "loss": 0.83360028, + "num_input_tokens_seen": 135085820, + "step": 6283, + "time_per_iteration": 2.744870185852051 + }, + { + "auxiliary_loss_clip": 0.01125314, + "auxiliary_loss_mlp": 0.01024462, + "balance_loss_clip": 0.97278988, + "balance_loss_mlp": 1.01691604, + "epoch": 0.7556063247760476, + "flos": 16362123114240.0, + "grad_norm": 2.456365943669177, + "language_loss": 0.68526536, + "learning_rate": 5.943498067940686e-07, + "loss": 0.70676315, + "num_input_tokens_seen": 135102845, + "step": 6284, + "time_per_iteration": 2.6722216606140137 + }, + { + "auxiliary_loss_clip": 0.01118647, + "auxiliary_loss_mlp": 0.01029872, + "balance_loss_clip": 0.97976112, + "balance_loss_mlp": 1.02303791, + "epoch": 0.7557265676666867, + "flos": 27235155502080.0, + "grad_norm": 1.9509364324548266, + "language_loss": 0.81688756, + "learning_rate": 5.937957820558686e-07, + "loss": 0.83837277, + "num_input_tokens_seen": 135122190, + "step": 6285, + "time_per_iteration": 2.744436025619507 + }, + { + "auxiliary_loss_clip": 0.010286, + "auxiliary_loss_mlp": 0.01000254, + "balance_loss_clip": 0.93834758, + "balance_loss_mlp": 0.99919289, + "epoch": 0.7558468105573258, + "flos": 62189131415040.0, + "grad_norm": 0.8458841513441585, + "language_loss": 0.65350282, + "learning_rate": 5.932419706377296e-07, + "loss": 0.67379129, + "num_input_tokens_seen": 135180495, + "step": 6286, + "time_per_iteration": 3.1677818298339844 + }, + { + "auxiliary_loss_clip": 0.01094752, + "auxiliary_loss_mlp": 0.01025311, + "balance_loss_clip": 0.93719864, + "balance_loss_mlp": 1.01761615, + "epoch": 0.7559670534479649, + "flos": 33249078823680.0, + "grad_norm": 1.8998260464699568, + "language_loss": 0.74004477, + "learning_rate": 5.92688372623666e-07, + "loss": 0.76124537, + "num_input_tokens_seen": 135199200, + "step": 6287, + "time_per_iteration": 2.7403225898742676 + }, + { + "auxiliary_loss_clip": 0.01143061, + "auxiliary_loss_mlp": 0.01027681, + "balance_loss_clip": 1.00846577, + "balance_loss_mlp": 1.01995611, + "epoch": 0.7560872963386039, + "flos": 14064379027200.0, + "grad_norm": 1.8369041157969022, + "language_loss": 0.7387709, + "learning_rate": 5.921349880976574e-07, + "loss": 0.76047826, + "num_input_tokens_seen": 135217035, + "step": 6288, + "time_per_iteration": 3.5477006435394287 + }, + { + "auxiliary_loss_clip": 0.01120885, + "auxiliary_loss_mlp": 0.00510159, + "balance_loss_clip": 0.9704718, + "balance_loss_mlp": 0.0, + "epoch": 0.7562075392292431, + "flos": 20412307941120.0, + "grad_norm": 1.7140168276957686, + "language_loss": 0.81907696, + "learning_rate": 5.915818171436515e-07, + "loss": 0.83538741, + "num_input_tokens_seen": 135236370, + "step": 6289, + "time_per_iteration": 2.651667356491089 + }, + { + "auxiliary_loss_clip": 0.01116646, + "auxiliary_loss_mlp": 0.01025087, + "balance_loss_clip": 0.96768343, + "balance_loss_mlp": 1.01751685, + "epoch": 0.7563277821198822, + "flos": 20376792368640.0, + "grad_norm": 1.8268257778885413, + "language_loss": 0.74640703, + "learning_rate": 5.910288598455642e-07, + "loss": 0.76782429, + "num_input_tokens_seen": 135255720, + "step": 6290, + "time_per_iteration": 2.69891095161438 + }, + { + "auxiliary_loss_clip": 0.01149779, + "auxiliary_loss_mlp": 0.01028487, + "balance_loss_clip": 1.01207232, + "balance_loss_mlp": 1.02024341, + "epoch": 0.7564480250105212, + "flos": 18588261438720.0, + "grad_norm": 2.485910579279326, + "language_loss": 0.74595296, + "learning_rate": 5.90476116287278e-07, + "loss": 0.7677356, + "num_input_tokens_seen": 135273320, + "step": 6291, + "time_per_iteration": 2.6294612884521484 + }, + { + "auxiliary_loss_clip": 0.01119341, + "auxiliary_loss_mlp": 0.01023496, + "balance_loss_clip": 0.97368062, + "balance_loss_mlp": 1.01640856, + "epoch": 0.7565682679011604, + "flos": 21215521918080.0, + "grad_norm": 2.0483242886129034, + "language_loss": 0.68066376, + "learning_rate": 5.899235865526456e-07, + "loss": 0.70209217, + "num_input_tokens_seen": 135292615, + "step": 6292, + "time_per_iteration": 2.672096014022827 + }, + { + "auxiliary_loss_clip": 0.01087208, + "auxiliary_loss_mlp": 0.01024077, + "balance_loss_clip": 0.93129778, + "balance_loss_mlp": 1.01682663, + "epoch": 0.7566885107917994, + "flos": 20449008662400.0, + "grad_norm": 1.7336179699560679, + "language_loss": 0.82380098, + "learning_rate": 5.893712707254825e-07, + "loss": 0.84491378, + "num_input_tokens_seen": 135310075, + "step": 6293, + "time_per_iteration": 2.736236572265625 + }, + { + "auxiliary_loss_clip": 0.01068351, + "auxiliary_loss_mlp": 0.01026257, + "balance_loss_clip": 0.89281356, + "balance_loss_mlp": 1.01874661, + "epoch": 0.7568087536824385, + "flos": 19025832919680.0, + "grad_norm": 3.0777894603082454, + "language_loss": 0.6628691, + "learning_rate": 5.888191688895769e-07, + "loss": 0.68381518, + "num_input_tokens_seen": 135327335, + "step": 6294, + "time_per_iteration": 2.7707901000976562 + }, + { + "auxiliary_loss_clip": 0.01164438, + "auxiliary_loss_mlp": 0.01021574, + "balance_loss_clip": 1.04655612, + "balance_loss_mlp": 1.01404858, + "epoch": 0.7569289965730777, + "flos": 15225442248960.0, + "grad_norm": 2.070159594868205, + "language_loss": 0.61917168, + "learning_rate": 5.882672811286813e-07, + "loss": 0.64103186, + "num_input_tokens_seen": 135343615, + "step": 6295, + "time_per_iteration": 2.621735095977783 + }, + { + "auxiliary_loss_clip": 0.01167814, + "auxiliary_loss_mlp": 0.01025288, + "balance_loss_clip": 1.04803205, + "balance_loss_mlp": 1.01779604, + "epoch": 0.7570492394637167, + "flos": 20769367086720.0, + "grad_norm": 2.042228383444008, + "language_loss": 0.69444042, + "learning_rate": 5.877156075265166e-07, + "loss": 0.71637142, + "num_input_tokens_seen": 135359880, + "step": 6296, + "time_per_iteration": 2.638075351715088 + }, + { + "auxiliary_loss_clip": 0.01114741, + "auxiliary_loss_mlp": 0.01024394, + "balance_loss_clip": 0.96874702, + "balance_loss_mlp": 1.01709843, + "epoch": 0.7571694823543558, + "flos": 15664091137920.0, + "grad_norm": 2.854240310416322, + "language_loss": 0.69994044, + "learning_rate": 5.871641481667715e-07, + "loss": 0.72133183, + "num_input_tokens_seen": 135374325, + "step": 6297, + "time_per_iteration": 2.659466505050659 + }, + { + "auxiliary_loss_clip": 0.01074875, + "auxiliary_loss_mlp": 0.01027118, + "balance_loss_clip": 0.89567935, + "balance_loss_mlp": 1.02003384, + "epoch": 0.7572897252449949, + "flos": 25409241492480.0, + "grad_norm": 1.7812992921322344, + "language_loss": 0.8432039, + "learning_rate": 5.866129031331011e-07, + "loss": 0.86422384, + "num_input_tokens_seen": 135393980, + "step": 6298, + "time_per_iteration": 2.7946560382843018 + }, + { + "auxiliary_loss_clip": 0.01119147, + "auxiliary_loss_mlp": 0.0102485, + "balance_loss_clip": 0.97087431, + "balance_loss_mlp": 1.01748276, + "epoch": 0.757409968135634, + "flos": 24279348297600.0, + "grad_norm": 2.459267010950101, + "language_loss": 0.8401022, + "learning_rate": 5.8606187250913e-07, + "loss": 0.86154217, + "num_input_tokens_seen": 135412030, + "step": 6299, + "time_per_iteration": 2.7022244930267334 + }, + { + "auxiliary_loss_clip": 0.01142696, + "auxiliary_loss_mlp": 0.00509747, + "balance_loss_clip": 1.0128026, + "balance_loss_mlp": 0.0, + "epoch": 0.757530211026273, + "flos": 24133766474880.0, + "grad_norm": 2.433943640797426, + "language_loss": 0.84223294, + "learning_rate": 5.855110563784482e-07, + "loss": 0.85875738, + "num_input_tokens_seen": 135430565, + "step": 6300, + "time_per_iteration": 2.7235677242279053 + }, + { + "auxiliary_loss_clip": 0.01138134, + "auxiliary_loss_mlp": 0.00509073, + "balance_loss_clip": 1.00962043, + "balance_loss_mlp": 0.0, + "epoch": 0.7576504539169122, + "flos": 23951807153280.0, + "grad_norm": 1.6638037231120988, + "language_loss": 0.64214033, + "learning_rate": 5.849604548246156e-07, + "loss": 0.65861237, + "num_input_tokens_seen": 135451675, + "step": 6301, + "time_per_iteration": 2.7133331298828125 + }, + { + "auxiliary_loss_clip": 0.01125859, + "auxiliary_loss_mlp": 0.00509785, + "balance_loss_clip": 0.9749307, + "balance_loss_mlp": 0.0, + "epoch": 0.7577706968075513, + "flos": 21251360712960.0, + "grad_norm": 1.9728994459741456, + "language_loss": 0.80055404, + "learning_rate": 5.844100679311565e-07, + "loss": 0.81691045, + "num_input_tokens_seen": 135470635, + "step": 6302, + "time_per_iteration": 3.669867992401123 + }, + { + "auxiliary_loss_clip": 0.0112095, + "auxiliary_loss_mlp": 0.01025486, + "balance_loss_clip": 0.97434437, + "balance_loss_mlp": 1.01827657, + "epoch": 0.7578909396981903, + "flos": 18296595002880.0, + "grad_norm": 2.156255882403243, + "language_loss": 0.76285863, + "learning_rate": 5.838598957815637e-07, + "loss": 0.78432298, + "num_input_tokens_seen": 135487865, + "step": 6303, + "time_per_iteration": 2.640127182006836 + }, + { + "auxiliary_loss_clip": 0.01114737, + "auxiliary_loss_mlp": 0.01025095, + "balance_loss_clip": 0.97179973, + "balance_loss_mlp": 1.01746011, + "epoch": 0.7580111825888295, + "flos": 25373869574400.0, + "grad_norm": 1.6624340909248616, + "language_loss": 0.85234529, + "learning_rate": 5.833099384592996e-07, + "loss": 0.87374365, + "num_input_tokens_seen": 135508440, + "step": 6304, + "time_per_iteration": 2.734510660171509 + }, + { + "auxiliary_loss_clip": 0.01114461, + "auxiliary_loss_mlp": 0.01026032, + "balance_loss_clip": 0.97201931, + "balance_loss_mlp": 1.01812267, + "epoch": 0.7581314254794685, + "flos": 23768662682880.0, + "grad_norm": 3.2680266417727615, + "language_loss": 0.71599883, + "learning_rate": 5.827601960477913e-07, + "loss": 0.73740375, + "num_input_tokens_seen": 135526365, + "step": 6305, + "time_per_iteration": 3.615419387817383 + }, + { + "auxiliary_loss_clip": 0.01140246, + "auxiliary_loss_mlp": 0.01025166, + "balance_loss_clip": 1.00889862, + "balance_loss_mlp": 1.01812065, + "epoch": 0.7582516683701076, + "flos": 22054610603520.0, + "grad_norm": 1.8757972384683177, + "language_loss": 0.70744455, + "learning_rate": 5.822106686304344e-07, + "loss": 0.72909868, + "num_input_tokens_seen": 135545655, + "step": 6306, + "time_per_iteration": 3.563889265060425 + }, + { + "auxiliary_loss_clip": 0.01098977, + "auxiliary_loss_mlp": 0.01027003, + "balance_loss_clip": 0.93197405, + "balance_loss_mlp": 1.01961815, + "epoch": 0.7583719112607467, + "flos": 31649725848960.0, + "grad_norm": 1.8294019796572791, + "language_loss": 0.57974261, + "learning_rate": 5.816613562905919e-07, + "loss": 0.60100245, + "num_input_tokens_seen": 135566840, + "step": 6307, + "time_per_iteration": 2.8176708221435547 + }, + { + "auxiliary_loss_clip": 0.01092282, + "auxiliary_loss_mlp": 0.01021017, + "balance_loss_clip": 0.93607211, + "balance_loss_mlp": 1.01411188, + "epoch": 0.7584921541513858, + "flos": 33068376478080.0, + "grad_norm": 1.7298921574553772, + "language_loss": 0.69987226, + "learning_rate": 5.811122591115933e-07, + "loss": 0.7210052, + "num_input_tokens_seen": 135587825, + "step": 6308, + "time_per_iteration": 2.783341407775879 + }, + { + "auxiliary_loss_clip": 0.01098094, + "auxiliary_loss_mlp": 0.01029407, + "balance_loss_clip": 0.93991196, + "balance_loss_mlp": 1.02177215, + "epoch": 0.7586123970420249, + "flos": 23326350606720.0, + "grad_norm": 2.368996264017206, + "language_loss": 0.71764517, + "learning_rate": 5.805633771767376e-07, + "loss": 0.73892015, + "num_input_tokens_seen": 135605220, + "step": 6309, + "time_per_iteration": 2.7871053218841553 + }, + { + "auxiliary_loss_clip": 0.01117132, + "auxiliary_loss_mlp": 0.01026715, + "balance_loss_clip": 0.97158659, + "balance_loss_mlp": 1.01964569, + "epoch": 0.7587326399326639, + "flos": 18334229477760.0, + "grad_norm": 1.6883593745842724, + "language_loss": 0.77712941, + "learning_rate": 5.800147105692888e-07, + "loss": 0.79856789, + "num_input_tokens_seen": 135624795, + "step": 6310, + "time_per_iteration": 2.7042148113250732 + }, + { + "auxiliary_loss_clip": 0.01143246, + "auxiliary_loss_mlp": 0.01026426, + "balance_loss_clip": 1.00901473, + "balance_loss_mlp": 1.01935053, + "epoch": 0.7588528828233031, + "flos": 17275080119040.0, + "grad_norm": 2.7494272305417033, + "language_loss": 0.79619598, + "learning_rate": 5.794662593724795e-07, + "loss": 0.81789267, + "num_input_tokens_seen": 135643800, + "step": 6311, + "time_per_iteration": 2.659874439239502 + }, + { + "auxiliary_loss_clip": 0.01169089, + "auxiliary_loss_mlp": 0.01025907, + "balance_loss_clip": 1.05325198, + "balance_loss_mlp": 1.01838493, + "epoch": 0.7589731257139422, + "flos": 17713621267200.0, + "grad_norm": 1.803466122292302, + "language_loss": 0.75340855, + "learning_rate": 5.789180236695091e-07, + "loss": 0.77535856, + "num_input_tokens_seen": 135660655, + "step": 6312, + "time_per_iteration": 2.597639560699463 + }, + { + "auxiliary_loss_clip": 0.0114076, + "auxiliary_loss_mlp": 0.01026641, + "balance_loss_clip": 1.01259375, + "balance_loss_mlp": 1.02023315, + "epoch": 0.7590933686045812, + "flos": 15961072786560.0, + "grad_norm": 1.921466078035742, + "language_loss": 0.85194397, + "learning_rate": 5.78370003543544e-07, + "loss": 0.87361801, + "num_input_tokens_seen": 135679410, + "step": 6313, + "time_per_iteration": 2.6482009887695312 + }, + { + "auxiliary_loss_clip": 0.01144629, + "auxiliary_loss_mlp": 0.00509531, + "balance_loss_clip": 1.01097524, + "balance_loss_mlp": 0.0, + "epoch": 0.7592136114952204, + "flos": 21068072588160.0, + "grad_norm": 1.861747401927517, + "language_loss": 0.83761704, + "learning_rate": 5.778221990777203e-07, + "loss": 0.85415864, + "num_input_tokens_seen": 135697150, + "step": 6314, + "time_per_iteration": 3.588402271270752 + }, + { + "auxiliary_loss_clip": 0.01122428, + "auxiliary_loss_mlp": 0.01023493, + "balance_loss_clip": 0.97486961, + "balance_loss_mlp": 1.01622081, + "epoch": 0.7593338543858594, + "flos": 25297666871040.0, + "grad_norm": 1.9282545634622617, + "language_loss": 0.82626605, + "learning_rate": 5.772746103551372e-07, + "loss": 0.84772521, + "num_input_tokens_seen": 135712545, + "step": 6315, + "time_per_iteration": 2.7003066539764404 + }, + { + "auxiliary_loss_clip": 0.01118176, + "auxiliary_loss_mlp": 0.01024317, + "balance_loss_clip": 0.97433245, + "balance_loss_mlp": 1.01712263, + "epoch": 0.7594540972764985, + "flos": 31832367528960.0, + "grad_norm": 1.5657804100265909, + "language_loss": 0.71926391, + "learning_rate": 5.767272374588648e-07, + "loss": 0.7406888, + "num_input_tokens_seen": 135733950, + "step": 6316, + "time_per_iteration": 2.7777836322784424 + }, + { + "auxiliary_loss_clip": 0.01142291, + "auxiliary_loss_mlp": 0.01023608, + "balance_loss_clip": 1.01328921, + "balance_loss_mlp": 1.01643777, + "epoch": 0.7595743401671377, + "flos": 37597250880000.0, + "grad_norm": 1.6399563046217447, + "language_loss": 0.7815271, + "learning_rate": 5.76180080471939e-07, + "loss": 0.80318612, + "num_input_tokens_seen": 135757120, + "step": 6317, + "time_per_iteration": 2.7916433811187744 + }, + { + "auxiliary_loss_clip": 0.01171917, + "auxiliary_loss_mlp": 0.01024825, + "balance_loss_clip": 1.05132294, + "balance_loss_mlp": 1.01712465, + "epoch": 0.7596945830577767, + "flos": 18287724343680.0, + "grad_norm": 2.7851715866876368, + "language_loss": 0.72516471, + "learning_rate": 5.756331394773631e-07, + "loss": 0.74713218, + "num_input_tokens_seen": 135773335, + "step": 6318, + "time_per_iteration": 2.5914552211761475 + }, + { + "auxiliary_loss_clip": 0.01051247, + "auxiliary_loss_mlp": 0.00509435, + "balance_loss_clip": 0.85856426, + "balance_loss_mlp": 0.0, + "epoch": 0.7598148259484158, + "flos": 22233122219520.0, + "grad_norm": 2.372092721110119, + "language_loss": 0.7627188, + "learning_rate": 5.750864145581071e-07, + "loss": 0.77832556, + "num_input_tokens_seen": 135792555, + "step": 6319, + "time_per_iteration": 2.7604873180389404 + }, + { + "auxiliary_loss_clip": 0.01164707, + "auxiliary_loss_mlp": 0.01024548, + "balance_loss_clip": 1.04902577, + "balance_loss_mlp": 1.01727295, + "epoch": 0.7599350688390549, + "flos": 27161718145920.0, + "grad_norm": 1.9861702890129804, + "language_loss": 0.86185014, + "learning_rate": 5.745399057971085e-07, + "loss": 0.88374269, + "num_input_tokens_seen": 135813690, + "step": 6320, + "time_per_iteration": 2.6494851112365723 + }, + { + "auxiliary_loss_clip": 0.01147064, + "auxiliary_loss_mlp": 0.01025646, + "balance_loss_clip": 1.01131427, + "balance_loss_mlp": 1.01772487, + "epoch": 0.760055311729694, + "flos": 15560704817280.0, + "grad_norm": 2.1491080188338976, + "language_loss": 0.75347579, + "learning_rate": 5.739936132772738e-07, + "loss": 0.77520287, + "num_input_tokens_seen": 135832255, + "step": 6321, + "time_per_iteration": 2.6071057319641113 + }, + { + "auxiliary_loss_clip": 0.01164478, + "auxiliary_loss_mlp": 0.01028142, + "balance_loss_clip": 1.04794574, + "balance_loss_mlp": 1.02068543, + "epoch": 0.760175554620333, + "flos": 25155496840320.0, + "grad_norm": 2.093380411654381, + "language_loss": 0.74502301, + "learning_rate": 5.734475370814733e-07, + "loss": 0.7669493, + "num_input_tokens_seen": 135851935, + "step": 6322, + "time_per_iteration": 2.6466610431671143 + }, + { + "auxiliary_loss_clip": 0.01143235, + "auxiliary_loss_mlp": 0.01024265, + "balance_loss_clip": 1.00830698, + "balance_loss_mlp": 1.01698756, + "epoch": 0.7602957975109722, + "flos": 24353791234560.0, + "grad_norm": 1.6016919315160623, + "language_loss": 0.78266943, + "learning_rate": 5.729016772925483e-07, + "loss": 0.80434442, + "num_input_tokens_seen": 135873510, + "step": 6323, + "time_per_iteration": 2.724236011505127 + }, + { + "auxiliary_loss_clip": 0.01069949, + "auxiliary_loss_mlp": 0.01027493, + "balance_loss_clip": 0.89797574, + "balance_loss_mlp": 1.02026892, + "epoch": 0.7604160404016113, + "flos": 25192664438400.0, + "grad_norm": 1.6218902244607325, + "language_loss": 0.70786846, + "learning_rate": 5.723560339933038e-07, + "loss": 0.72884285, + "num_input_tokens_seen": 135893845, + "step": 6324, + "time_per_iteration": 2.7286293506622314 + }, + { + "auxiliary_loss_clip": 0.01142567, + "auxiliary_loss_mlp": 0.0050966, + "balance_loss_clip": 1.01077521, + "balance_loss_mlp": 0.0, + "epoch": 0.7605362832922503, + "flos": 29861841363840.0, + "grad_norm": 1.9919231251012601, + "language_loss": 0.65357566, + "learning_rate": 5.71810607266513e-07, + "loss": 0.67009795, + "num_input_tokens_seen": 135912430, + "step": 6325, + "time_per_iteration": 2.781615734100342 + }, + { + "auxiliary_loss_clip": 0.01144582, + "auxiliary_loss_mlp": 0.01026558, + "balance_loss_clip": 1.01070166, + "balance_loss_mlp": 1.01941466, + "epoch": 0.7606565261828895, + "flos": 13917935278080.0, + "grad_norm": 1.7117842826249814, + "language_loss": 0.60659719, + "learning_rate": 5.712653971949184e-07, + "loss": 0.62830859, + "num_input_tokens_seen": 135930550, + "step": 6326, + "time_per_iteration": 2.619044780731201 + }, + { + "auxiliary_loss_clip": 0.01139508, + "auxiliary_loss_mlp": 0.01028051, + "balance_loss_clip": 1.01046038, + "balance_loss_mlp": 1.02048135, + "epoch": 0.7607767690735285, + "flos": 18551273408640.0, + "grad_norm": 2.408139433321685, + "language_loss": 0.7547071, + "learning_rate": 5.707204038612268e-07, + "loss": 0.77638274, + "num_input_tokens_seen": 135947980, + "step": 6327, + "time_per_iteration": 2.6233341693878174 + }, + { + "auxiliary_loss_clip": 0.01129834, + "auxiliary_loss_mlp": 0.01027755, + "balance_loss_clip": 0.97775567, + "balance_loss_mlp": 1.0199523, + "epoch": 0.7608970119641676, + "flos": 20922993555840.0, + "grad_norm": 1.9496121016279293, + "language_loss": 0.73842072, + "learning_rate": 5.701756273481138e-07, + "loss": 0.75999659, + "num_input_tokens_seen": 135965400, + "step": 6328, + "time_per_iteration": 3.6299238204956055 + }, + { + "auxiliary_loss_clip": 0.01122387, + "auxiliary_loss_mlp": 0.01027024, + "balance_loss_clip": 0.96977913, + "balance_loss_mlp": 1.0199368, + "epoch": 0.7610172548548068, + "flos": 23807302738560.0, + "grad_norm": 1.5768064548541172, + "language_loss": 0.73918635, + "learning_rate": 5.696310677382212e-07, + "loss": 0.76068044, + "num_input_tokens_seen": 135986795, + "step": 6329, + "time_per_iteration": 2.762209892272949 + }, + { + "auxiliary_loss_clip": 0.01005722, + "auxiliary_loss_mlp": 0.01002555, + "balance_loss_clip": 0.90409392, + "balance_loss_mlp": 1.00136268, + "epoch": 0.7611374977454458, + "flos": 66496580426880.0, + "grad_norm": 0.8852710999735504, + "language_loss": 0.61742151, + "learning_rate": 5.690867251141576e-07, + "loss": 0.63750422, + "num_input_tokens_seen": 136053450, + "step": 6330, + "time_per_iteration": 3.4024970531463623 + }, + { + "auxiliary_loss_clip": 0.0114856, + "auxiliary_loss_mlp": 0.01025854, + "balance_loss_clip": 1.01006055, + "balance_loss_mlp": 1.01775432, + "epoch": 0.7612577406360849, + "flos": 15633136592640.0, + "grad_norm": 4.183289126264849, + "language_loss": 0.91487586, + "learning_rate": 5.685425995585013e-07, + "loss": 0.93662, + "num_input_tokens_seen": 136071375, + "step": 6331, + "time_per_iteration": 3.5264618396759033 + }, + { + "auxiliary_loss_clip": 0.01028681, + "auxiliary_loss_mlp": 0.01002036, + "balance_loss_clip": 0.93975896, + "balance_loss_mlp": 1.00100505, + "epoch": 0.761377983526724, + "flos": 60526253237760.0, + "grad_norm": 0.7509785671233855, + "language_loss": 0.58994466, + "learning_rate": 5.679986911537935e-07, + "loss": 0.61025184, + "num_input_tokens_seen": 136138905, + "step": 6332, + "time_per_iteration": 4.306978702545166 + }, + { + "auxiliary_loss_clip": 0.01064949, + "auxiliary_loss_mlp": 0.01022726, + "balance_loss_clip": 0.89789075, + "balance_loss_mlp": 1.01529908, + "epoch": 0.7614982264173631, + "flos": 35772522019200.0, + "grad_norm": 1.8164426940471072, + "language_loss": 0.67609274, + "learning_rate": 5.674549999825462e-07, + "loss": 0.69696951, + "num_input_tokens_seen": 136161720, + "step": 6333, + "time_per_iteration": 2.8832881450653076 + }, + { + "auxiliary_loss_clip": 0.01047154, + "auxiliary_loss_mlp": 0.01000016, + "balance_loss_clip": 0.97639459, + "balance_loss_mlp": 0.99893135, + "epoch": 0.7616184693080021, + "flos": 67925502345600.0, + "grad_norm": 0.9136683811216376, + "language_loss": 0.71391082, + "learning_rate": 5.669115261272363e-07, + "loss": 0.73438251, + "num_input_tokens_seen": 136222040, + "step": 6334, + "time_per_iteration": 3.212878942489624 + }, + { + "auxiliary_loss_clip": 0.01144201, + "auxiliary_loss_mlp": 0.01027688, + "balance_loss_clip": 1.01132214, + "balance_loss_mlp": 1.02006769, + "epoch": 0.7617387121986413, + "flos": 20521979141760.0, + "grad_norm": 2.204491394234474, + "language_loss": 0.72359884, + "learning_rate": 5.663682696703081e-07, + "loss": 0.74531782, + "num_input_tokens_seen": 136240305, + "step": 6335, + "time_per_iteration": 2.656879186630249 + }, + { + "auxiliary_loss_clip": 0.0116529, + "auxiliary_loss_mlp": 0.01026031, + "balance_loss_clip": 1.0491997, + "balance_loss_mlp": 1.01944435, + "epoch": 0.7618589550892804, + "flos": 18624495283200.0, + "grad_norm": 2.0101013135045096, + "language_loss": 0.82072496, + "learning_rate": 5.658252306941746e-07, + "loss": 0.84263819, + "num_input_tokens_seen": 136259625, + "step": 6336, + "time_per_iteration": 2.6192574501037598 + }, + { + "auxiliary_loss_clip": 0.01076053, + "auxiliary_loss_mlp": 0.01025566, + "balance_loss_clip": 0.8974508, + "balance_loss_mlp": 1.01812124, + "epoch": 0.7619791979799194, + "flos": 17453735389440.0, + "grad_norm": 2.5550494166085014, + "language_loss": 0.75332856, + "learning_rate": 5.65282409281212e-07, + "loss": 0.7743448, + "num_input_tokens_seen": 136277090, + "step": 6337, + "time_per_iteration": 2.721601724624634 + }, + { + "auxiliary_loss_clip": 0.0111712, + "auxiliary_loss_mlp": 0.01030469, + "balance_loss_clip": 0.97201526, + "balance_loss_mlp": 1.02323341, + "epoch": 0.7620994408705585, + "flos": 14137421333760.0, + "grad_norm": 1.8408534956305047, + "language_loss": 0.70052052, + "learning_rate": 5.64739805513768e-07, + "loss": 0.72199643, + "num_input_tokens_seen": 136294635, + "step": 6338, + "time_per_iteration": 2.7109737396240234 + }, + { + "auxiliary_loss_clip": 0.01041507, + "auxiliary_loss_mlp": 0.00500756, + "balance_loss_clip": 0.97459531, + "balance_loss_mlp": 0.0, + "epoch": 0.7622196837611976, + "flos": 70708792527360.0, + "grad_norm": 0.785571618024564, + "language_loss": 0.55760938, + "learning_rate": 5.641974194741541e-07, + "loss": 0.57303202, + "num_input_tokens_seen": 136350320, + "step": 6339, + "time_per_iteration": 3.1796715259552 + }, + { + "auxiliary_loss_clip": 0.0104033, + "auxiliary_loss_mlp": 0.01003644, + "balance_loss_clip": 0.96491599, + "balance_loss_mlp": 1.00242221, + "epoch": 0.7623399266518367, + "flos": 60684150447360.0, + "grad_norm": 0.7870463251283109, + "language_loss": 0.63732302, + "learning_rate": 5.636552512446502e-07, + "loss": 0.65776277, + "num_input_tokens_seen": 136411375, + "step": 6340, + "time_per_iteration": 4.1188507080078125 + }, + { + "auxiliary_loss_clip": 0.01139885, + "auxiliary_loss_mlp": 0.01024036, + "balance_loss_clip": 1.01078653, + "balance_loss_mlp": 1.01751828, + "epoch": 0.7624601695424758, + "flos": 26468893641600.0, + "grad_norm": 1.656419119977178, + "language_loss": 0.78248757, + "learning_rate": 5.631133009075027e-07, + "loss": 0.80412674, + "num_input_tokens_seen": 136430560, + "step": 6341, + "time_per_iteration": 2.6931591033935547 + }, + { + "auxiliary_loss_clip": 0.01143559, + "auxiliary_loss_mlp": 0.00508877, + "balance_loss_clip": 1.01172149, + "balance_loss_mlp": 0.0, + "epoch": 0.7625804124331149, + "flos": 19135755515520.0, + "grad_norm": 1.9238297599423093, + "language_loss": 0.68453741, + "learning_rate": 5.625715685449242e-07, + "loss": 0.70106179, + "num_input_tokens_seen": 136448665, + "step": 6342, + "time_per_iteration": 2.623960494995117 + }, + { + "auxiliary_loss_clip": 0.0109931, + "auxiliary_loss_mlp": 0.01023585, + "balance_loss_clip": 0.94124162, + "balance_loss_mlp": 1.01644754, + "epoch": 0.762700655323754, + "flos": 26213101914240.0, + "grad_norm": 2.041260503039044, + "language_loss": 0.71565539, + "learning_rate": 5.620300542390966e-07, + "loss": 0.73688436, + "num_input_tokens_seen": 136469710, + "step": 6343, + "time_per_iteration": 2.737151861190796 + }, + { + "auxiliary_loss_clip": 0.01114961, + "auxiliary_loss_mlp": 0.01024735, + "balance_loss_clip": 0.96903008, + "balance_loss_mlp": 1.01766562, + "epoch": 0.762820898214393, + "flos": 22382582711040.0, + "grad_norm": 2.176228097101722, + "language_loss": 0.85274595, + "learning_rate": 5.614887580721659e-07, + "loss": 0.87414289, + "num_input_tokens_seen": 136489855, + "step": 6344, + "time_per_iteration": 2.692056179046631 + }, + { + "auxiliary_loss_clip": 0.01090772, + "auxiliary_loss_mlp": 0.01030579, + "balance_loss_clip": 0.93668628, + "balance_loss_mlp": 1.02289641, + "epoch": 0.7629411411050322, + "flos": 15700504550400.0, + "grad_norm": 2.0254612220735897, + "language_loss": 0.73880398, + "learning_rate": 5.609476801262481e-07, + "loss": 0.76001751, + "num_input_tokens_seen": 136504715, + "step": 6345, + "time_per_iteration": 2.7210440635681152 + }, + { + "auxiliary_loss_clip": 0.0109428, + "auxiliary_loss_mlp": 0.01024927, + "balance_loss_clip": 0.93714035, + "balance_loss_mlp": 1.01741362, + "epoch": 0.7630613839956712, + "flos": 13770342293760.0, + "grad_norm": 2.1510914702998494, + "language_loss": 0.64489889, + "learning_rate": 5.604068204834223e-07, + "loss": 0.66609097, + "num_input_tokens_seen": 136521610, + "step": 6346, + "time_per_iteration": 2.6749463081359863 + }, + { + "auxiliary_loss_clip": 0.01071426, + "auxiliary_loss_mlp": 0.00509826, + "balance_loss_clip": 0.89763129, + "balance_loss_mlp": 0.0, + "epoch": 0.7631816268863103, + "flos": 14569569861120.0, + "grad_norm": 2.0190604011940483, + "language_loss": 0.76345563, + "learning_rate": 5.598661792257367e-07, + "loss": 0.77926815, + "num_input_tokens_seen": 136538655, + "step": 6347, + "time_per_iteration": 2.7541141510009766 + }, + { + "auxiliary_loss_clip": 0.01139681, + "auxiliary_loss_mlp": 0.01025699, + "balance_loss_clip": 1.0081569, + "balance_loss_mlp": 1.01821232, + "epoch": 0.7633018697769495, + "flos": 19062210418560.0, + "grad_norm": 2.3876522981655133, + "language_loss": 0.759202, + "learning_rate": 5.593257564352071e-07, + "loss": 0.78085577, + "num_input_tokens_seen": 136557095, + "step": 6348, + "time_per_iteration": 2.6511647701263428 + }, + { + "auxiliary_loss_clip": 0.01141357, + "auxiliary_loss_mlp": 0.01027262, + "balance_loss_clip": 1.01004577, + "balance_loss_mlp": 1.02026725, + "epoch": 0.7634221126675885, + "flos": 22052958577920.0, + "grad_norm": 1.481757693127448, + "language_loss": 0.75833714, + "learning_rate": 5.58785552193815e-07, + "loss": 0.78002334, + "num_input_tokens_seen": 136577340, + "step": 6349, + "time_per_iteration": 2.656571865081787 + }, + { + "auxiliary_loss_clip": 0.01166676, + "auxiliary_loss_mlp": 0.01023473, + "balance_loss_clip": 1.04988849, + "balance_loss_mlp": 1.0167948, + "epoch": 0.7635423555582276, + "flos": 29382720825600.0, + "grad_norm": 1.7711405987303979, + "language_loss": 0.75814795, + "learning_rate": 5.582455665835086e-07, + "loss": 0.78004944, + "num_input_tokens_seen": 136597635, + "step": 6350, + "time_per_iteration": 2.6622848510742188 + }, + { + "auxiliary_loss_clip": 0.01127731, + "auxiliary_loss_mlp": 0.01026186, + "balance_loss_clip": 0.97207224, + "balance_loss_mlp": 1.01880705, + "epoch": 0.7636625984488667, + "flos": 17784903807360.0, + "grad_norm": 4.165766449785253, + "language_loss": 0.73292357, + "learning_rate": 5.577057996862036e-07, + "loss": 0.75446272, + "num_input_tokens_seen": 136615260, + "step": 6351, + "time_per_iteration": 2.7825088500976562 + }, + { + "auxiliary_loss_clip": 0.0116248, + "auxiliary_loss_mlp": 0.01023878, + "balance_loss_clip": 1.04850268, + "balance_loss_mlp": 1.01654959, + "epoch": 0.7637828413395058, + "flos": 23734583654400.0, + "grad_norm": 1.613490768687119, + "language_loss": 0.76354194, + "learning_rate": 5.571662515837814e-07, + "loss": 0.78540552, + "num_input_tokens_seen": 136637220, + "step": 6352, + "time_per_iteration": 2.6525185108184814 + }, + { + "auxiliary_loss_clip": 0.01120296, + "auxiliary_loss_mlp": 0.01026228, + "balance_loss_clip": 0.97392952, + "balance_loss_mlp": 1.01945686, + "epoch": 0.7639030842301449, + "flos": 36283279461120.0, + "grad_norm": 1.5726241463206163, + "language_loss": 0.83941162, + "learning_rate": 5.566269223580926e-07, + "loss": 0.86087686, + "num_input_tokens_seen": 136658930, + "step": 6353, + "time_per_iteration": 2.8374152183532715 + }, + { + "auxiliary_loss_clip": 0.01144054, + "auxiliary_loss_mlp": 0.010231, + "balance_loss_clip": 1.01109099, + "balance_loss_mlp": 1.01646626, + "epoch": 0.764023327120784, + "flos": 28878104609280.0, + "grad_norm": 1.6231561897252085, + "language_loss": 0.75488943, + "learning_rate": 5.560878120909511e-07, + "loss": 0.77656096, + "num_input_tokens_seen": 136681530, + "step": 6354, + "time_per_iteration": 3.6005775928497314 + }, + { + "auxiliary_loss_clip": 0.01046492, + "auxiliary_loss_mlp": 0.00999623, + "balance_loss_clip": 0.97606266, + "balance_loss_mlp": 0.99854392, + "epoch": 0.7641435700114231, + "flos": 64789711067520.0, + "grad_norm": 0.8356184014959251, + "language_loss": 0.58549565, + "learning_rate": 5.55548920864141e-07, + "loss": 0.60595679, + "num_input_tokens_seen": 136742185, + "step": 6355, + "time_per_iteration": 3.2636353969573975 + }, + { + "auxiliary_loss_clip": 0.01144509, + "auxiliary_loss_mlp": 0.01025482, + "balance_loss_clip": 1.01353049, + "balance_loss_mlp": 1.01843095, + "epoch": 0.7642638129020621, + "flos": 16835784785280.0, + "grad_norm": 2.2621360524001766, + "language_loss": 0.77911329, + "learning_rate": 5.550102487594113e-07, + "loss": 0.80081326, + "num_input_tokens_seen": 136760855, + "step": 6356, + "time_per_iteration": 3.525754451751709 + }, + { + "auxiliary_loss_clip": 0.01076654, + "auxiliary_loss_mlp": 0.00508885, + "balance_loss_clip": 0.89507335, + "balance_loss_mlp": 0.0, + "epoch": 0.7643840557927013, + "flos": 30408940391040.0, + "grad_norm": 2.5010774631532824, + "language_loss": 0.71666288, + "learning_rate": 5.54471795858477e-07, + "loss": 0.7325182, + "num_input_tokens_seen": 136780925, + "step": 6357, + "time_per_iteration": 2.872729778289795 + }, + { + "auxiliary_loss_clip": 0.01094636, + "auxiliary_loss_mlp": 0.01027293, + "balance_loss_clip": 0.93064332, + "balance_loss_mlp": 1.02051914, + "epoch": 0.7645042986833404, + "flos": 16983234115200.0, + "grad_norm": 2.0366998249997024, + "language_loss": 0.83449388, + "learning_rate": 5.539335622430235e-07, + "loss": 0.85571313, + "num_input_tokens_seen": 136799545, + "step": 6358, + "time_per_iteration": 3.5703887939453125 + }, + { + "auxiliary_loss_clip": 0.01135659, + "auxiliary_loss_mlp": 0.01021922, + "balance_loss_clip": 1.00756645, + "balance_loss_mlp": 1.01488316, + "epoch": 0.7646245415739794, + "flos": 17311493531520.0, + "grad_norm": 1.9874293166763077, + "language_loss": 0.74802846, + "learning_rate": 5.533955479946975e-07, + "loss": 0.76960427, + "num_input_tokens_seen": 136818325, + "step": 6359, + "time_per_iteration": 2.631253242492676 + }, + { + "auxiliary_loss_clip": 0.01021269, + "auxiliary_loss_mlp": 0.00501623, + "balance_loss_clip": 0.92725801, + "balance_loss_mlp": 0.0, + "epoch": 0.7647447844646186, + "flos": 70402332666240.0, + "grad_norm": 0.8579343389722336, + "language_loss": 0.65783072, + "learning_rate": 5.528577531951173e-07, + "loss": 0.6730597, + "num_input_tokens_seen": 136878730, + "step": 6360, + "time_per_iteration": 3.23775315284729 + }, + { + "auxiliary_loss_clip": 0.01122284, + "auxiliary_loss_mlp": 0.0102806, + "balance_loss_clip": 0.97223127, + "balance_loss_mlp": 1.0211401, + "epoch": 0.7648650273552576, + "flos": 17675914965120.0, + "grad_norm": 1.9582751703851342, + "language_loss": 0.73617858, + "learning_rate": 5.523201779258653e-07, + "loss": 0.75768203, + "num_input_tokens_seen": 136897705, + "step": 6361, + "time_per_iteration": 2.700343370437622 + }, + { + "auxiliary_loss_clip": 0.01163765, + "auxiliary_loss_mlp": 0.01028177, + "balance_loss_clip": 1.04631948, + "balance_loss_mlp": 1.02069938, + "epoch": 0.7649852702458967, + "flos": 22162019247360.0, + "grad_norm": 2.4741384580296617, + "language_loss": 0.83846986, + "learning_rate": 5.517828222684912e-07, + "loss": 0.86038923, + "num_input_tokens_seen": 136918360, + "step": 6362, + "time_per_iteration": 2.6903719902038574 + }, + { + "auxiliary_loss_clip": 0.01022612, + "auxiliary_loss_mlp": 0.01001124, + "balance_loss_clip": 0.93778616, + "balance_loss_mlp": 1.00008106, + "epoch": 0.7651055131365359, + "flos": 69848338227840.0, + "grad_norm": 0.771892124879838, + "language_loss": 0.59064841, + "learning_rate": 5.512456863045117e-07, + "loss": 0.61088586, + "num_input_tokens_seen": 136979050, + "step": 6363, + "time_per_iteration": 3.2259535789489746 + }, + { + "auxiliary_loss_clip": 0.01165003, + "auxiliary_loss_mlp": 0.01024088, + "balance_loss_clip": 1.04731989, + "balance_loss_mlp": 1.01660728, + "epoch": 0.7652257560271749, + "flos": 19464014931840.0, + "grad_norm": 1.7370937369239432, + "language_loss": 0.74321294, + "learning_rate": 5.507087701154089e-07, + "loss": 0.76510388, + "num_input_tokens_seen": 136998970, + "step": 6364, + "time_per_iteration": 2.6374459266662598 + }, + { + "auxiliary_loss_clip": 0.01075946, + "auxiliary_loss_mlp": 0.01028364, + "balance_loss_clip": 0.89534748, + "balance_loss_mlp": 1.02139592, + "epoch": 0.765345998917814, + "flos": 15961108700160.0, + "grad_norm": 1.8188228116580785, + "language_loss": 0.75162816, + "learning_rate": 5.50172073782634e-07, + "loss": 0.77267134, + "num_input_tokens_seen": 137016950, + "step": 6365, + "time_per_iteration": 2.761918306350708 + }, + { + "auxiliary_loss_clip": 0.01094097, + "auxiliary_loss_mlp": 0.01025312, + "balance_loss_clip": 0.93644607, + "balance_loss_mlp": 1.01870513, + "epoch": 0.7654662418084531, + "flos": 23659853408640.0, + "grad_norm": 1.7240418137534177, + "language_loss": 0.87898827, + "learning_rate": 5.496355973876023e-07, + "loss": 0.90018237, + "num_input_tokens_seen": 137036205, + "step": 6366, + "time_per_iteration": 3.731826066970825 + }, + { + "auxiliary_loss_clip": 0.01091549, + "auxiliary_loss_mlp": 0.00510394, + "balance_loss_clip": 0.93022513, + "balance_loss_mlp": 0.0, + "epoch": 0.7655864846990922, + "flos": 41463608878080.0, + "grad_norm": 1.5916579139632532, + "language_loss": 0.71172196, + "learning_rate": 5.490993410116984e-07, + "loss": 0.72774136, + "num_input_tokens_seen": 137059195, + "step": 6367, + "time_per_iteration": 3.0302953720092773 + }, + { + "auxiliary_loss_clip": 0.01092354, + "auxiliary_loss_mlp": 0.01027062, + "balance_loss_clip": 0.93568337, + "balance_loss_mlp": 1.02046382, + "epoch": 0.7657067275897312, + "flos": 43142684088960.0, + "grad_norm": 1.771142023371518, + "language_loss": 0.69261116, + "learning_rate": 5.485633047362704e-07, + "loss": 0.71380532, + "num_input_tokens_seen": 137081200, + "step": 6368, + "time_per_iteration": 2.9021501541137695 + }, + { + "auxiliary_loss_clip": 0.01172397, + "auxiliary_loss_mlp": 0.0103619, + "balance_loss_clip": 1.05397046, + "balance_loss_mlp": 1.02829552, + "epoch": 0.7658269704803703, + "flos": 17311780840320.0, + "grad_norm": 2.128021361332956, + "language_loss": 0.78760421, + "learning_rate": 5.480274886426341e-07, + "loss": 0.80969, + "num_input_tokens_seen": 137097840, + "step": 6369, + "time_per_iteration": 2.603309392929077 + }, + { + "auxiliary_loss_clip": 0.01139936, + "auxiliary_loss_mlp": 0.01025852, + "balance_loss_clip": 1.01399183, + "balance_loss_mlp": 1.01928329, + "epoch": 0.7659472133710095, + "flos": 12568160977920.0, + "grad_norm": 3.4116773682451917, + "language_loss": 0.77827209, + "learning_rate": 5.474918928120744e-07, + "loss": 0.79992992, + "num_input_tokens_seen": 137114335, + "step": 6370, + "time_per_iteration": 2.6012747287750244 + }, + { + "auxiliary_loss_clip": 0.01140709, + "auxiliary_loss_mlp": 0.01023208, + "balance_loss_clip": 1.00984359, + "balance_loss_mlp": 1.01624918, + "epoch": 0.7660674562616485, + "flos": 22707430335360.0, + "grad_norm": 1.6129711159985554, + "language_loss": 0.87221801, + "learning_rate": 5.469565173258392e-07, + "loss": 0.89385712, + "num_input_tokens_seen": 137132850, + "step": 6371, + "time_per_iteration": 2.666722059249878 + }, + { + "auxiliary_loss_clip": 0.01169226, + "auxiliary_loss_mlp": 0.01029593, + "balance_loss_clip": 1.04977918, + "balance_loss_mlp": 1.02135539, + "epoch": 0.7661876991522876, + "flos": 17056455989760.0, + "grad_norm": 1.8889736124031176, + "language_loss": 0.63840771, + "learning_rate": 5.464213622651454e-07, + "loss": 0.66039592, + "num_input_tokens_seen": 137150665, + "step": 6372, + "time_per_iteration": 2.659332036972046 + }, + { + "auxiliary_loss_clip": 0.01099675, + "auxiliary_loss_mlp": 0.01025029, + "balance_loss_clip": 0.93406469, + "balance_loss_mlp": 1.0172627, + "epoch": 0.7663079420429267, + "flos": 20084228092800.0, + "grad_norm": 1.6490446988577474, + "language_loss": 0.8420608, + "learning_rate": 5.458864277111753e-07, + "loss": 0.86330783, + "num_input_tokens_seen": 137168500, + "step": 6373, + "time_per_iteration": 2.7641797065734863 + }, + { + "auxiliary_loss_clip": 0.0111158, + "auxiliary_loss_mlp": 0.00508584, + "balance_loss_clip": 0.96931863, + "balance_loss_mlp": 0.0, + "epoch": 0.7664281849335658, + "flos": 12677473042560.0, + "grad_norm": 2.3091063108071395, + "language_loss": 0.6948306, + "learning_rate": 5.453517137450769e-07, + "loss": 0.71103227, + "num_input_tokens_seen": 137185075, + "step": 6374, + "time_per_iteration": 2.684230327606201 + }, + { + "auxiliary_loss_clip": 0.01144709, + "auxiliary_loss_mlp": 0.01031427, + "balance_loss_clip": 1.01359129, + "balance_loss_mlp": 1.02370238, + "epoch": 0.7665484278242048, + "flos": 22345271458560.0, + "grad_norm": 2.197789771510199, + "language_loss": 0.76102805, + "learning_rate": 5.448172204479684e-07, + "loss": 0.78278941, + "num_input_tokens_seen": 137204355, + "step": 6375, + "time_per_iteration": 2.6316049098968506 + }, + { + "auxiliary_loss_clip": 0.01164982, + "auxiliary_loss_mlp": 0.01022713, + "balance_loss_clip": 1.0477277, + "balance_loss_mlp": 1.0157299, + "epoch": 0.766668670714844, + "flos": 23617909301760.0, + "grad_norm": 1.8415465612538804, + "language_loss": 0.74605846, + "learning_rate": 5.442829479009294e-07, + "loss": 0.7679354, + "num_input_tokens_seen": 137223135, + "step": 6376, + "time_per_iteration": 2.6343891620635986 + }, + { + "auxiliary_loss_clip": 0.01148906, + "auxiliary_loss_mlp": 0.01028912, + "balance_loss_clip": 1.01010203, + "balance_loss_mlp": 1.02174115, + "epoch": 0.7667889136054831, + "flos": 19427134642560.0, + "grad_norm": 4.006441561058482, + "language_loss": 0.71758449, + "learning_rate": 5.437488961850103e-07, + "loss": 0.7393626, + "num_input_tokens_seen": 137242935, + "step": 6377, + "time_per_iteration": 2.6131467819213867 + }, + { + "auxiliary_loss_clip": 0.01070576, + "auxiliary_loss_mlp": 0.01026965, + "balance_loss_clip": 0.89619893, + "balance_loss_mlp": 1.02009296, + "epoch": 0.7669091564961221, + "flos": 26866352609280.0, + "grad_norm": 1.9316136348404394, + "language_loss": 0.75451213, + "learning_rate": 5.432150653812258e-07, + "loss": 0.77548754, + "num_input_tokens_seen": 137262970, + "step": 6378, + "time_per_iteration": 2.791721820831299 + }, + { + "auxiliary_loss_clip": 0.01140212, + "auxiliary_loss_mlp": 0.01029275, + "balance_loss_clip": 1.01194215, + "balance_loss_mlp": 1.02222073, + "epoch": 0.7670293993867613, + "flos": 12385303816320.0, + "grad_norm": 2.233489740201665, + "language_loss": 0.82420123, + "learning_rate": 5.42681455570557e-07, + "loss": 0.84589612, + "num_input_tokens_seen": 137279500, + "step": 6379, + "time_per_iteration": 2.611743688583374 + }, + { + "auxiliary_loss_clip": 0.01161143, + "auxiliary_loss_mlp": 0.01020826, + "balance_loss_clip": 1.04598725, + "balance_loss_mlp": 1.01406956, + "epoch": 0.7671496422774003, + "flos": 21762944167680.0, + "grad_norm": 1.7872735047833626, + "language_loss": 0.64856744, + "learning_rate": 5.42148066833954e-07, + "loss": 0.67038709, + "num_input_tokens_seen": 137298745, + "step": 6380, + "time_per_iteration": 2.642528772354126 + }, + { + "auxiliary_loss_clip": 0.01165686, + "auxiliary_loss_mlp": 0.01024283, + "balance_loss_clip": 1.04861867, + "balance_loss_mlp": 1.01757455, + "epoch": 0.7672698851680394, + "flos": 21069221823360.0, + "grad_norm": 3.5204289916353204, + "language_loss": 0.75397933, + "learning_rate": 5.416148992523289e-07, + "loss": 0.77587903, + "num_input_tokens_seen": 137317320, + "step": 6381, + "time_per_iteration": 3.4607529640197754 + }, + { + "auxiliary_loss_clip": 0.01026957, + "auxiliary_loss_mlp": 0.01024927, + "balance_loss_clip": 0.82025272, + "balance_loss_mlp": 1.01781607, + "epoch": 0.7673901280586786, + "flos": 16976697840000.0, + "grad_norm": 1.810068289476618, + "language_loss": 0.78483784, + "learning_rate": 5.410819529065644e-07, + "loss": 0.80535674, + "num_input_tokens_seen": 137335275, + "step": 6382, + "time_per_iteration": 2.7810606956481934 + }, + { + "auxiliary_loss_clip": 0.01072939, + "auxiliary_loss_mlp": 0.01030907, + "balance_loss_clip": 0.8939274, + "balance_loss_mlp": 1.02366197, + "epoch": 0.7675103709493176, + "flos": 29242669697280.0, + "grad_norm": 2.2384200566732035, + "language_loss": 0.6567533, + "learning_rate": 5.405492278775079e-07, + "loss": 0.67779171, + "num_input_tokens_seen": 137355055, + "step": 6383, + "time_per_iteration": 3.7176177501678467 + }, + { + "auxiliary_loss_clip": 0.01120805, + "auxiliary_loss_mlp": 0.01026998, + "balance_loss_clip": 0.97151816, + "balance_loss_mlp": 1.01948166, + "epoch": 0.7676306138399567, + "flos": 29023004073600.0, + "grad_norm": 2.4488693517416564, + "language_loss": 0.79702085, + "learning_rate": 5.400167242459732e-07, + "loss": 0.81849891, + "num_input_tokens_seen": 137374015, + "step": 6384, + "time_per_iteration": 3.6677160263061523 + }, + { + "auxiliary_loss_clip": 0.01140341, + "auxiliary_loss_mlp": 0.01024089, + "balance_loss_clip": 1.01021051, + "balance_loss_mlp": 1.01707363, + "epoch": 0.7677508567305958, + "flos": 22565116650240.0, + "grad_norm": 1.5839997979611633, + "language_loss": 0.80721509, + "learning_rate": 5.394844420927405e-07, + "loss": 0.82885939, + "num_input_tokens_seen": 137393625, + "step": 6385, + "time_per_iteration": 2.635478973388672 + }, + { + "auxiliary_loss_clip": 0.01164906, + "auxiliary_loss_mlp": 0.01030632, + "balance_loss_clip": 1.04889762, + "balance_loss_mlp": 1.02344942, + "epoch": 0.7678710996212349, + "flos": 25411432222080.0, + "grad_norm": 2.7323806665884516, + "language_loss": 0.73668039, + "learning_rate": 5.389523814985562e-07, + "loss": 0.75863576, + "num_input_tokens_seen": 137413045, + "step": 6386, + "time_per_iteration": 2.6602463722229004 + }, + { + "auxiliary_loss_clip": 0.01072588, + "auxiliary_loss_mlp": 0.01025705, + "balance_loss_clip": 0.89572477, + "balance_loss_mlp": 1.01789713, + "epoch": 0.767991342511874, + "flos": 26756825063040.0, + "grad_norm": 1.8766553403092607, + "language_loss": 0.76663196, + "learning_rate": 5.384205425441344e-07, + "loss": 0.78761494, + "num_input_tokens_seen": 137433955, + "step": 6387, + "time_per_iteration": 2.7577526569366455 + }, + { + "auxiliary_loss_clip": 0.01120113, + "auxiliary_loss_mlp": 0.01026728, + "balance_loss_clip": 0.97120845, + "balance_loss_mlp": 1.01902115, + "epoch": 0.7681115854025131, + "flos": 26359509749760.0, + "grad_norm": 7.058112210076538, + "language_loss": 0.84358513, + "learning_rate": 5.378889253101537e-07, + "loss": 0.86505353, + "num_input_tokens_seen": 137454510, + "step": 6388, + "time_per_iteration": 2.757233142852783 + }, + { + "auxiliary_loss_clip": 0.01139682, + "auxiliary_loss_mlp": 0.01021894, + "balance_loss_clip": 1.00659728, + "balance_loss_mlp": 1.01464057, + "epoch": 0.7682318282931522, + "flos": 23257043314560.0, + "grad_norm": 1.6887433347623502, + "language_loss": 0.80979866, + "learning_rate": 5.373575298772617e-07, + "loss": 0.8314144, + "num_input_tokens_seen": 137473630, + "step": 6389, + "time_per_iteration": 2.6528913974761963 + }, + { + "auxiliary_loss_clip": 0.01045622, + "auxiliary_loss_mlp": 0.00999475, + "balance_loss_clip": 0.97474682, + "balance_loss_mlp": 0.99840814, + "epoch": 0.7683520711837912, + "flos": 70072457137920.0, + "grad_norm": 0.7606132253490901, + "language_loss": 0.6131255, + "learning_rate": 5.368263563260689e-07, + "loss": 0.63357645, + "num_input_tokens_seen": 137538765, + "step": 6390, + "time_per_iteration": 3.334986448287964 + }, + { + "auxiliary_loss_clip": 0.01140342, + "auxiliary_loss_mlp": 0.0102939, + "balance_loss_clip": 1.00841248, + "balance_loss_mlp": 1.02129614, + "epoch": 0.7684723140744304, + "flos": 18624890332800.0, + "grad_norm": 1.7114596323333167, + "language_loss": 0.6456266, + "learning_rate": 5.362954047371537e-07, + "loss": 0.66732395, + "num_input_tokens_seen": 137557875, + "step": 6391, + "time_per_iteration": 2.709932565689087 + }, + { + "auxiliary_loss_clip": 0.01099357, + "auxiliary_loss_mlp": 0.01024131, + "balance_loss_clip": 0.94086444, + "balance_loss_mlp": 1.01657867, + "epoch": 0.7685925569650695, + "flos": 27452989532160.0, + "grad_norm": 1.9465949882467035, + "language_loss": 0.72149366, + "learning_rate": 5.357646751910627e-07, + "loss": 0.74272859, + "num_input_tokens_seen": 137579055, + "step": 6392, + "time_per_iteration": 3.6840341091156006 + }, + { + "auxiliary_loss_clip": 0.01116546, + "auxiliary_loss_mlp": 0.01027359, + "balance_loss_clip": 0.97171938, + "balance_loss_mlp": 1.02014363, + "epoch": 0.7687127998557085, + "flos": 24535714642560.0, + "grad_norm": 1.9749910507262323, + "language_loss": 0.79758513, + "learning_rate": 5.352341677683061e-07, + "loss": 0.81902415, + "num_input_tokens_seen": 137600355, + "step": 6393, + "time_per_iteration": 2.7114627361297607 + }, + { + "auxiliary_loss_clip": 0.01101422, + "auxiliary_loss_mlp": 0.01021678, + "balance_loss_clip": 0.93396354, + "balance_loss_mlp": 1.01485586, + "epoch": 0.7688330427463477, + "flos": 25155963717120.0, + "grad_norm": 1.8086550187989425, + "language_loss": 0.79052645, + "learning_rate": 5.347038825493617e-07, + "loss": 0.81175745, + "num_input_tokens_seen": 137621885, + "step": 6394, + "time_per_iteration": 2.818326473236084 + }, + { + "auxiliary_loss_clip": 0.0111484, + "auxiliary_loss_mlp": 0.01017589, + "balance_loss_clip": 0.9734391, + "balance_loss_mlp": 1.01101708, + "epoch": 0.7689532856369867, + "flos": 21211284113280.0, + "grad_norm": 1.9023732654374224, + "language_loss": 0.68111926, + "learning_rate": 5.341738196146732e-07, + "loss": 0.70244348, + "num_input_tokens_seen": 137640230, + "step": 6395, + "time_per_iteration": 2.758335828781128 + }, + { + "auxiliary_loss_clip": 0.01138347, + "auxiliary_loss_mlp": 0.01029768, + "balance_loss_clip": 1.00683701, + "balance_loss_mlp": 1.02185893, + "epoch": 0.7690735285276258, + "flos": 25119083427840.0, + "grad_norm": 3.244272914653875, + "language_loss": 0.73949194, + "learning_rate": 5.336439790446503e-07, + "loss": 0.76117307, + "num_input_tokens_seen": 137659330, + "step": 6396, + "time_per_iteration": 2.69596266746521 + }, + { + "auxiliary_loss_clip": 0.01092483, + "auxiliary_loss_mlp": 0.01028529, + "balance_loss_clip": 0.9315955, + "balance_loss_mlp": 1.02156138, + "epoch": 0.769193771418265, + "flos": 54744020640000.0, + "grad_norm": 1.8547302603245723, + "language_loss": 0.6249404, + "learning_rate": 5.331143609196711e-07, + "loss": 0.64615059, + "num_input_tokens_seen": 137683145, + "step": 6397, + "time_per_iteration": 3.0119011402130127 + }, + { + "auxiliary_loss_clip": 0.01141639, + "auxiliary_loss_mlp": 0.01027259, + "balance_loss_clip": 1.01260734, + "balance_loss_mlp": 1.01962996, + "epoch": 0.769314014308904, + "flos": 37341890115840.0, + "grad_norm": 1.7057134050092546, + "language_loss": 0.77183205, + "learning_rate": 5.325849653200758e-07, + "loss": 0.79352105, + "num_input_tokens_seen": 137707095, + "step": 6398, + "time_per_iteration": 2.810112237930298 + }, + { + "auxiliary_loss_clip": 0.01166164, + "auxiliary_loss_mlp": 0.01023472, + "balance_loss_clip": 1.05029392, + "balance_loss_mlp": 1.01630127, + "epoch": 0.7694342571995431, + "flos": 20631686256000.0, + "grad_norm": 1.777780397644704, + "language_loss": 0.76605374, + "learning_rate": 5.32055792326175e-07, + "loss": 0.78795004, + "num_input_tokens_seen": 137725520, + "step": 6399, + "time_per_iteration": 2.67596435546875 + }, + { + "auxiliary_loss_clip": 0.01123291, + "auxiliary_loss_mlp": 0.01030615, + "balance_loss_clip": 0.97536623, + "balance_loss_mlp": 1.02329504, + "epoch": 0.7695545000901821, + "flos": 24207706621440.0, + "grad_norm": 1.877951347298452, + "language_loss": 0.72963053, + "learning_rate": 5.315268420182437e-07, + "loss": 0.75116956, + "num_input_tokens_seen": 137744195, + "step": 6400, + "time_per_iteration": 2.721731424331665 + }, + { + "auxiliary_loss_clip": 0.01103068, + "auxiliary_loss_mlp": 0.00509516, + "balance_loss_clip": 0.93613631, + "balance_loss_mlp": 0.0, + "epoch": 0.7696747429808213, + "flos": 28001273708160.0, + "grad_norm": 1.6372556915323917, + "language_loss": 0.76576501, + "learning_rate": 5.309981144765221e-07, + "loss": 0.78189087, + "num_input_tokens_seen": 137764340, + "step": 6401, + "time_per_iteration": 2.7731049060821533 + }, + { + "auxiliary_loss_clip": 0.01076295, + "auxiliary_loss_mlp": 0.01024695, + "balance_loss_clip": 0.89324456, + "balance_loss_mlp": 1.01756322, + "epoch": 0.7697949858714603, + "flos": 11509550323200.0, + "grad_norm": 2.346452510464409, + "language_loss": 0.76195848, + "learning_rate": 5.304696097812196e-07, + "loss": 0.78296834, + "num_input_tokens_seen": 137780940, + "step": 6402, + "time_per_iteration": 2.707777738571167 + }, + { + "auxiliary_loss_clip": 0.01118475, + "auxiliary_loss_mlp": 0.01024667, + "balance_loss_clip": 0.9710201, + "balance_loss_mlp": 1.01623893, + "epoch": 0.7699152287620994, + "flos": 26688271956480.0, + "grad_norm": 3.142508130451182, + "language_loss": 0.60185647, + "learning_rate": 5.299413280125078e-07, + "loss": 0.62328792, + "num_input_tokens_seen": 137799250, + "step": 6403, + "time_per_iteration": 2.6972274780273438 + }, + { + "auxiliary_loss_clip": 0.01118844, + "auxiliary_loss_mlp": 0.01022824, + "balance_loss_clip": 0.97142708, + "balance_loss_mlp": 1.01600504, + "epoch": 0.7700354716527386, + "flos": 16544944362240.0, + "grad_norm": 1.8177003313724849, + "language_loss": 0.72989404, + "learning_rate": 5.294132692505284e-07, + "loss": 0.75131071, + "num_input_tokens_seen": 137817660, + "step": 6404, + "time_per_iteration": 2.6821022033691406 + }, + { + "auxiliary_loss_clip": 0.01063066, + "auxiliary_loss_mlp": 0.01027683, + "balance_loss_clip": 0.89280236, + "balance_loss_mlp": 1.02053916, + "epoch": 0.7701557145433776, + "flos": 19242733196160.0, + "grad_norm": 2.13323214930762, + "language_loss": 0.79292089, + "learning_rate": 5.288854335753861e-07, + "loss": 0.81382835, + "num_input_tokens_seen": 137835920, + "step": 6405, + "time_per_iteration": 2.724047899246216 + }, + { + "auxiliary_loss_clip": 0.01144771, + "auxiliary_loss_mlp": 0.01025188, + "balance_loss_clip": 1.0101707, + "balance_loss_mlp": 1.01822627, + "epoch": 0.7702759574340167, + "flos": 31685744211840.0, + "grad_norm": 1.7560240153137479, + "language_loss": 0.75790983, + "learning_rate": 5.283578210671551e-07, + "loss": 0.77960944, + "num_input_tokens_seen": 137858160, + "step": 6406, + "time_per_iteration": 3.6150434017181396 + }, + { + "auxiliary_loss_clip": 0.01120864, + "auxiliary_loss_mlp": 0.01029404, + "balance_loss_clip": 0.97156674, + "balance_loss_mlp": 1.02181029, + "epoch": 0.7703962003246558, + "flos": 16800089644800.0, + "grad_norm": 6.298433220407872, + "language_loss": 0.76996708, + "learning_rate": 5.278304318058719e-07, + "loss": 0.79146969, + "num_input_tokens_seen": 137876015, + "step": 6407, + "time_per_iteration": 2.6887776851654053 + }, + { + "auxiliary_loss_clip": 0.0104838, + "auxiliary_loss_mlp": 0.01026476, + "balance_loss_clip": 0.85764021, + "balance_loss_mlp": 1.01892996, + "epoch": 0.7705164432152949, + "flos": 35736072693120.0, + "grad_norm": 2.039791122509331, + "language_loss": 0.78924894, + "learning_rate": 5.273032658715411e-07, + "loss": 0.80999756, + "num_input_tokens_seen": 137898825, + "step": 6408, + "time_per_iteration": 3.8144826889038086 + }, + { + "auxiliary_loss_clip": 0.01068893, + "auxiliary_loss_mlp": 0.01024814, + "balance_loss_clip": 0.89516181, + "balance_loss_mlp": 1.01778674, + "epoch": 0.7706366861059339, + "flos": 23365960329600.0, + "grad_norm": 2.035983347552893, + "language_loss": 0.76404572, + "learning_rate": 5.267763233441347e-07, + "loss": 0.7849828, + "num_input_tokens_seen": 137919455, + "step": 6409, + "time_per_iteration": 2.7671680450439453 + }, + { + "auxiliary_loss_clip": 0.01146669, + "auxiliary_loss_mlp": 0.01028484, + "balance_loss_clip": 1.011554, + "balance_loss_mlp": 1.02087224, + "epoch": 0.7707569289965731, + "flos": 22929897219840.0, + "grad_norm": 2.1886649506462352, + "language_loss": 0.69726849, + "learning_rate": 5.26249604303588e-07, + "loss": 0.71902001, + "num_input_tokens_seen": 137937960, + "step": 6410, + "time_per_iteration": 3.6413707733154297 + }, + { + "auxiliary_loss_clip": 0.01167821, + "auxiliary_loss_mlp": 0.01026728, + "balance_loss_clip": 1.05128622, + "balance_loss_mlp": 1.01921153, + "epoch": 0.7708771718872122, + "flos": 17420661941760.0, + "grad_norm": 2.656930606545348, + "language_loss": 0.78311306, + "learning_rate": 5.257231088298057e-07, + "loss": 0.80505854, + "num_input_tokens_seen": 137956370, + "step": 6411, + "time_per_iteration": 2.6990389823913574 + }, + { + "auxiliary_loss_clip": 0.01001355, + "auxiliary_loss_mlp": 0.01002188, + "balance_loss_clip": 0.90086102, + "balance_loss_mlp": 1.00114465, + "epoch": 0.7709974147778512, + "flos": 72241316248320.0, + "grad_norm": 0.7944758139012267, + "language_loss": 0.5395323, + "learning_rate": 5.25196837002655e-07, + "loss": 0.55956769, + "num_input_tokens_seen": 138016080, + "step": 6412, + "time_per_iteration": 3.315033197402954 + }, + { + "auxiliary_loss_clip": 0.01117985, + "auxiliary_loss_mlp": 0.01026103, + "balance_loss_clip": 0.97197556, + "balance_loss_mlp": 1.018363, + "epoch": 0.7711176576684904, + "flos": 39859694876160.0, + "grad_norm": 2.037306471703262, + "language_loss": 0.68420637, + "learning_rate": 5.24670788901971e-07, + "loss": 0.70564729, + "num_input_tokens_seen": 138039170, + "step": 6413, + "time_per_iteration": 2.8116416931152344 + }, + { + "auxiliary_loss_clip": 0.01120692, + "auxiliary_loss_mlp": 0.01030499, + "balance_loss_clip": 0.97458321, + "balance_loss_mlp": 1.02183247, + "epoch": 0.7712379005591294, + "flos": 36976391274240.0, + "grad_norm": 9.229470791282386, + "language_loss": 0.68774199, + "learning_rate": 5.241449646075557e-07, + "loss": 0.70925391, + "num_input_tokens_seen": 138062395, + "step": 6414, + "time_per_iteration": 2.8119208812713623 + }, + { + "auxiliary_loss_clip": 0.01148345, + "auxiliary_loss_mlp": 0.0102384, + "balance_loss_clip": 1.00985861, + "balance_loss_mlp": 1.01622868, + "epoch": 0.7713581434497685, + "flos": 22776773541120.0, + "grad_norm": 1.9116984204384095, + "language_loss": 0.72463459, + "learning_rate": 5.236193641991762e-07, + "loss": 0.74635643, + "num_input_tokens_seen": 138080325, + "step": 6415, + "time_per_iteration": 2.685436725616455 + }, + { + "auxiliary_loss_clip": 0.01119274, + "auxiliary_loss_mlp": 0.01025743, + "balance_loss_clip": 0.97456908, + "balance_loss_mlp": 1.01794934, + "epoch": 0.7714783863404077, + "flos": 24097460803200.0, + "grad_norm": 1.9602763016795608, + "language_loss": 0.69893694, + "learning_rate": 5.23093987756565e-07, + "loss": 0.72038716, + "num_input_tokens_seen": 138099020, + "step": 6416, + "time_per_iteration": 2.7253687381744385 + }, + { + "auxiliary_loss_clip": 0.01101204, + "auxiliary_loss_mlp": 0.01029543, + "balance_loss_clip": 0.93103409, + "balance_loss_mlp": 1.02165163, + "epoch": 0.7715986292310467, + "flos": 21063655215360.0, + "grad_norm": 1.8362882977645552, + "language_loss": 0.75421882, + "learning_rate": 5.225688353594217e-07, + "loss": 0.77552629, + "num_input_tokens_seen": 138118650, + "step": 6417, + "time_per_iteration": 2.765282154083252 + }, + { + "auxiliary_loss_clip": 0.0112343, + "auxiliary_loss_mlp": 0.00508732, + "balance_loss_clip": 0.97340816, + "balance_loss_mlp": 0.0, + "epoch": 0.7717188721216858, + "flos": 20594877793920.0, + "grad_norm": 3.928159353751915, + "language_loss": 0.77459288, + "learning_rate": 5.220439070874108e-07, + "loss": 0.79091454, + "num_input_tokens_seen": 138137890, + "step": 6418, + "time_per_iteration": 3.6366970539093018 + }, + { + "auxiliary_loss_clip": 0.01145192, + "auxiliary_loss_mlp": 0.01027344, + "balance_loss_clip": 1.01501966, + "balance_loss_mlp": 1.0203402, + "epoch": 0.7718391150123249, + "flos": 26250951870720.0, + "grad_norm": 1.4929860701832014, + "language_loss": 0.71158552, + "learning_rate": 5.215192030201652e-07, + "loss": 0.73331082, + "num_input_tokens_seen": 138158880, + "step": 6419, + "time_per_iteration": 2.7153096199035645 + }, + { + "auxiliary_loss_clip": 0.01083348, + "auxiliary_loss_mlp": 0.01025627, + "balance_loss_clip": 0.92909408, + "balance_loss_mlp": 1.0184741, + "epoch": 0.771959357902964, + "flos": 22049762267520.0, + "grad_norm": 1.86673207537706, + "language_loss": 0.86113131, + "learning_rate": 5.209947232372798e-07, + "loss": 0.8822211, + "num_input_tokens_seen": 138176370, + "step": 6420, + "time_per_iteration": 2.753781795501709 + }, + { + "auxiliary_loss_clip": 0.01146327, + "auxiliary_loss_mlp": 0.0050957, + "balance_loss_clip": 1.01046669, + "balance_loss_mlp": 0.0, + "epoch": 0.772079600793603, + "flos": 30446000248320.0, + "grad_norm": 1.719148205014383, + "language_loss": 0.81076169, + "learning_rate": 5.204704678183196e-07, + "loss": 0.82732069, + "num_input_tokens_seen": 138195105, + "step": 6421, + "time_per_iteration": 2.7606751918792725 + }, + { + "auxiliary_loss_clip": 0.01166654, + "auxiliary_loss_mlp": 0.0102569, + "balance_loss_clip": 1.04998064, + "balance_loss_mlp": 1.01841831, + "epoch": 0.7721998436842422, + "flos": 12969857750400.0, + "grad_norm": 2.9922077498665156, + "language_loss": 0.85162622, + "learning_rate": 5.19946436842813e-07, + "loss": 0.8735497, + "num_input_tokens_seen": 138212235, + "step": 6422, + "time_per_iteration": 2.6045753955841064 + }, + { + "auxiliary_loss_clip": 0.01099409, + "auxiliary_loss_mlp": 0.01026897, + "balance_loss_clip": 0.93748957, + "balance_loss_mlp": 1.01968825, + "epoch": 0.7723200865748813, + "flos": 32635509678720.0, + "grad_norm": 1.6771557772375452, + "language_loss": 0.68283856, + "learning_rate": 5.194226303902546e-07, + "loss": 0.70410156, + "num_input_tokens_seen": 138231970, + "step": 6423, + "time_per_iteration": 2.8111109733581543 + }, + { + "auxiliary_loss_clip": 0.01116192, + "auxiliary_loss_mlp": 0.01029342, + "balance_loss_clip": 0.97179121, + "balance_loss_mlp": 1.02152181, + "epoch": 0.7724403294655203, + "flos": 21105707063040.0, + "grad_norm": 1.7600765949876165, + "language_loss": 0.70740926, + "learning_rate": 5.188990485401072e-07, + "loss": 0.72886461, + "num_input_tokens_seen": 138251175, + "step": 6424, + "time_per_iteration": 2.702836751937866 + }, + { + "auxiliary_loss_clip": 0.01145507, + "auxiliary_loss_mlp": 0.01028948, + "balance_loss_clip": 1.01295817, + "balance_loss_mlp": 1.02140808, + "epoch": 0.7725605723561595, + "flos": 22090736707200.0, + "grad_norm": 1.8572154349478254, + "language_loss": 0.86299241, + "learning_rate": 5.183756913717954e-07, + "loss": 0.8847369, + "num_input_tokens_seen": 138270950, + "step": 6425, + "time_per_iteration": 2.6742656230926514 + }, + { + "auxiliary_loss_clip": 0.01115083, + "auxiliary_loss_mlp": 0.0103075, + "balance_loss_clip": 0.97208726, + "balance_loss_mlp": 1.02341878, + "epoch": 0.7726808152467985, + "flos": 34495610457600.0, + "grad_norm": 1.848605322514448, + "language_loss": 0.72853398, + "learning_rate": 5.178525589647136e-07, + "loss": 0.74999225, + "num_input_tokens_seen": 138292590, + "step": 6426, + "time_per_iteration": 2.765068292617798 + }, + { + "auxiliary_loss_clip": 0.01123839, + "auxiliary_loss_mlp": 0.01021207, + "balance_loss_clip": 0.97162968, + "balance_loss_mlp": 1.0143199, + "epoch": 0.7728010581374376, + "flos": 22306344094080.0, + "grad_norm": 2.0412202716357313, + "language_loss": 0.78961897, + "learning_rate": 5.173296513982197e-07, + "loss": 0.81106949, + "num_input_tokens_seen": 138311115, + "step": 6427, + "time_per_iteration": 2.6474320888519287 + }, + { + "auxiliary_loss_clip": 0.0110727, + "auxiliary_loss_mlp": 0.01031084, + "balance_loss_clip": 0.93696719, + "balance_loss_mlp": 1.02349615, + "epoch": 0.7729213010280768, + "flos": 27126453968640.0, + "grad_norm": 2.1976798523212633, + "language_loss": 0.65227556, + "learning_rate": 5.168069687516398e-07, + "loss": 0.67365909, + "num_input_tokens_seen": 138330885, + "step": 6428, + "time_per_iteration": 2.7905263900756836 + }, + { + "auxiliary_loss_clip": 0.01121475, + "auxiliary_loss_mlp": 0.01023668, + "balance_loss_clip": 0.97477543, + "balance_loss_mlp": 1.01576447, + "epoch": 0.7730415439187158, + "flos": 18150223080960.0, + "grad_norm": 3.939521499688736, + "language_loss": 0.71784544, + "learning_rate": 5.16284511104263e-07, + "loss": 0.73929679, + "num_input_tokens_seen": 138350020, + "step": 6429, + "time_per_iteration": 2.6895527839660645 + }, + { + "auxiliary_loss_clip": 0.0111847, + "auxiliary_loss_mlp": 0.01026859, + "balance_loss_clip": 0.9727385, + "balance_loss_mlp": 1.01947999, + "epoch": 0.7731617868093549, + "flos": 11947480940160.0, + "grad_norm": 2.1382252047463988, + "language_loss": 0.79699117, + "learning_rate": 5.157622785353457e-07, + "loss": 0.81844449, + "num_input_tokens_seen": 138368135, + "step": 6430, + "time_per_iteration": 2.669059991836548 + }, + { + "auxiliary_loss_clip": 0.01044857, + "auxiliary_loss_mlp": 0.01002014, + "balance_loss_clip": 0.9752835, + "balance_loss_mlp": 1.00097132, + "epoch": 0.7732820296999939, + "flos": 64201027069440.0, + "grad_norm": 0.666934479430652, + "language_loss": 0.60412335, + "learning_rate": 5.152402711241113e-07, + "loss": 0.62459207, + "num_input_tokens_seen": 138436040, + "step": 6431, + "time_per_iteration": 3.2417192459106445 + }, + { + "auxiliary_loss_clip": 0.01093487, + "auxiliary_loss_mlp": 0.0102474, + "balance_loss_clip": 0.93179929, + "balance_loss_mlp": 1.01808202, + "epoch": 0.7734022725906331, + "flos": 25302191984640.0, + "grad_norm": 1.8381261597333125, + "language_loss": 0.83039665, + "learning_rate": 5.147184889497465e-07, + "loss": 0.85157895, + "num_input_tokens_seen": 138455510, + "step": 6432, + "time_per_iteration": 3.693715810775757 + }, + { + "auxiliary_loss_clip": 0.01087411, + "auxiliary_loss_mlp": 0.01025609, + "balance_loss_clip": 0.93283057, + "balance_loss_mlp": 1.01838517, + "epoch": 0.7735225154812722, + "flos": 17347440067200.0, + "grad_norm": 2.14312273429574, + "language_loss": 0.79983574, + "learning_rate": 5.141969320914072e-07, + "loss": 0.82096595, + "num_input_tokens_seen": 138473015, + "step": 6433, + "time_per_iteration": 2.675623893737793 + }, + { + "auxiliary_loss_clip": 0.01170625, + "auxiliary_loss_mlp": 0.01026917, + "balance_loss_clip": 1.05035496, + "balance_loss_mlp": 1.01899552, + "epoch": 0.7736427583719112, + "flos": 32630086725120.0, + "grad_norm": 2.2231592088531094, + "language_loss": 0.62381631, + "learning_rate": 5.136756006282113e-07, + "loss": 0.64579177, + "num_input_tokens_seen": 138491680, + "step": 6434, + "time_per_iteration": 3.585341691970825 + }, + { + "auxiliary_loss_clip": 0.01169821, + "auxiliary_loss_mlp": 0.01028504, + "balance_loss_clip": 1.05031371, + "balance_loss_mlp": 1.02095187, + "epoch": 0.7737630012625504, + "flos": 19860073269120.0, + "grad_norm": 2.417662112340939, + "language_loss": 0.85108209, + "learning_rate": 5.131544946392446e-07, + "loss": 0.87306535, + "num_input_tokens_seen": 138506960, + "step": 6435, + "time_per_iteration": 2.6635944843292236 + }, + { + "auxiliary_loss_clip": 0.01124432, + "auxiliary_loss_mlp": 0.01029321, + "balance_loss_clip": 0.97874784, + "balance_loss_mlp": 1.02204621, + "epoch": 0.7738832441531894, + "flos": 36022639397760.0, + "grad_norm": 2.332238257585198, + "language_loss": 0.63875401, + "learning_rate": 5.126336142035592e-07, + "loss": 0.66029149, + "num_input_tokens_seen": 138526995, + "step": 6436, + "time_per_iteration": 3.680683135986328 + }, + { + "auxiliary_loss_clip": 0.01119206, + "auxiliary_loss_mlp": 0.01020966, + "balance_loss_clip": 0.97048455, + "balance_loss_mlp": 1.01395631, + "epoch": 0.7740034870438285, + "flos": 13405274415360.0, + "grad_norm": 2.352532808227451, + "language_loss": 0.71986508, + "learning_rate": 5.121129594001721e-07, + "loss": 0.74126679, + "num_input_tokens_seen": 138541260, + "step": 6437, + "time_per_iteration": 2.6379921436309814 + }, + { + "auxiliary_loss_clip": 0.01140825, + "auxiliary_loss_mlp": 0.01032447, + "balance_loss_clip": 1.01077306, + "balance_loss_mlp": 1.02554131, + "epoch": 0.7741237299344677, + "flos": 22086714384000.0, + "grad_norm": 7.11313194477119, + "language_loss": 0.81449592, + "learning_rate": 5.115925303080661e-07, + "loss": 0.83622861, + "num_input_tokens_seen": 138560970, + "step": 6438, + "time_per_iteration": 2.7502899169921875 + }, + { + "auxiliary_loss_clip": 0.011212, + "auxiliary_loss_mlp": 0.01027193, + "balance_loss_clip": 0.97318077, + "balance_loss_mlp": 1.01984978, + "epoch": 0.7742439728251067, + "flos": 19864777950720.0, + "grad_norm": 2.5616620978552502, + "language_loss": 0.79068214, + "learning_rate": 5.110723270061899e-07, + "loss": 0.81216609, + "num_input_tokens_seen": 138577460, + "step": 6439, + "time_per_iteration": 2.6571474075317383 + }, + { + "auxiliary_loss_clip": 0.01164939, + "auxiliary_loss_mlp": 0.01025606, + "balance_loss_clip": 1.04850888, + "balance_loss_mlp": 1.01867652, + "epoch": 0.7743642157157458, + "flos": 16690167048960.0, + "grad_norm": 1.7145626049174414, + "language_loss": 0.79347354, + "learning_rate": 5.105523495734572e-07, + "loss": 0.8153789, + "num_input_tokens_seen": 138594860, + "step": 6440, + "time_per_iteration": 2.5789906978607178 + }, + { + "auxiliary_loss_clip": 0.01166393, + "auxiliary_loss_mlp": 0.0103131, + "balance_loss_clip": 1.04799783, + "balance_loss_mlp": 1.02405572, + "epoch": 0.7744844586063849, + "flos": 20304360593280.0, + "grad_norm": 1.5562148872835122, + "language_loss": 0.75004661, + "learning_rate": 5.100325980887499e-07, + "loss": 0.77202356, + "num_input_tokens_seen": 138614785, + "step": 6441, + "time_per_iteration": 2.6172502040863037 + }, + { + "auxiliary_loss_clip": 0.01128964, + "auxiliary_loss_mlp": 0.01034206, + "balance_loss_clip": 0.97360992, + "balance_loss_mlp": 1.02719092, + "epoch": 0.774604701497024, + "flos": 22966705681920.0, + "grad_norm": 1.8350569981908071, + "language_loss": 0.83119911, + "learning_rate": 5.095130726309116e-07, + "loss": 0.85283077, + "num_input_tokens_seen": 138634960, + "step": 6442, + "time_per_iteration": 2.6627931594848633 + }, + { + "auxiliary_loss_clip": 0.01063364, + "auxiliary_loss_mlp": 0.01001329, + "balance_loss_clip": 1.01194692, + "balance_loss_mlp": 1.00030351, + "epoch": 0.774724944387663, + "flos": 60288523073280.0, + "grad_norm": 0.7883100824605087, + "language_loss": 0.58965713, + "learning_rate": 5.089937732787559e-07, + "loss": 0.61030406, + "num_input_tokens_seen": 138699520, + "step": 6443, + "time_per_iteration": 3.217052698135376 + }, + { + "auxiliary_loss_clip": 0.01094467, + "auxiliary_loss_mlp": 0.01024419, + "balance_loss_clip": 0.93241346, + "balance_loss_mlp": 1.01695585, + "epoch": 0.7748451872783022, + "flos": 26761026954240.0, + "grad_norm": 2.481938262547933, + "language_loss": 0.66500509, + "learning_rate": 5.084747001110592e-07, + "loss": 0.68619394, + "num_input_tokens_seen": 138719145, + "step": 6444, + "time_per_iteration": 3.7798497676849365 + }, + { + "auxiliary_loss_clip": 0.01143804, + "auxiliary_loss_mlp": 0.00509081, + "balance_loss_clip": 1.0171957, + "balance_loss_mlp": 0.0, + "epoch": 0.7749654301689413, + "flos": 30338627518080.0, + "grad_norm": 1.6710545275568727, + "language_loss": 0.69894326, + "learning_rate": 5.07955853206564e-07, + "loss": 0.7154721, + "num_input_tokens_seen": 138743850, + "step": 6445, + "time_per_iteration": 2.778719663619995 + }, + { + "auxiliary_loss_clip": 0.01146808, + "auxiliary_loss_mlp": 0.01032381, + "balance_loss_clip": 1.01138258, + "balance_loss_mlp": 1.02475142, + "epoch": 0.7750856730595803, + "flos": 43179851687040.0, + "grad_norm": 1.8190686652903147, + "language_loss": 0.70938635, + "learning_rate": 5.074372326439807e-07, + "loss": 0.73117816, + "num_input_tokens_seen": 138766860, + "step": 6446, + "time_per_iteration": 2.7972991466522217 + }, + { + "auxiliary_loss_clip": 0.01100107, + "auxiliary_loss_mlp": 0.01027312, + "balance_loss_clip": 0.93640041, + "balance_loss_mlp": 1.01956367, + "epoch": 0.7752059159502195, + "flos": 17640040256640.0, + "grad_norm": 2.286939015429062, + "language_loss": 0.73817551, + "learning_rate": 5.069188385019814e-07, + "loss": 0.75944966, + "num_input_tokens_seen": 138784560, + "step": 6447, + "time_per_iteration": 2.6927340030670166 + }, + { + "auxiliary_loss_clip": 0.01077351, + "auxiliary_loss_mlp": 0.01024498, + "balance_loss_clip": 0.89331025, + "balance_loss_mlp": 1.01726782, + "epoch": 0.7753261588408585, + "flos": 12677688524160.0, + "grad_norm": 2.675195942493612, + "language_loss": 0.61662042, + "learning_rate": 5.064006708592077e-07, + "loss": 0.63763893, + "num_input_tokens_seen": 138800805, + "step": 6448, + "time_per_iteration": 2.726083517074585 + }, + { + "auxiliary_loss_clip": 0.01116036, + "auxiliary_loss_mlp": 0.01031243, + "balance_loss_clip": 0.97343779, + "balance_loss_mlp": 1.02425098, + "epoch": 0.7754464017314976, + "flos": 16690741666560.0, + "grad_norm": 3.005143797152126, + "language_loss": 0.75129062, + "learning_rate": 5.058827297942641e-07, + "loss": 0.77276337, + "num_input_tokens_seen": 138815910, + "step": 6449, + "time_per_iteration": 2.694218873977661 + }, + { + "auxiliary_loss_clip": 0.01126942, + "auxiliary_loss_mlp": 0.0102202, + "balance_loss_clip": 0.97368574, + "balance_loss_mlp": 1.01473665, + "epoch": 0.7755666446221368, + "flos": 19718944732800.0, + "grad_norm": 2.0667274374548326, + "language_loss": 0.75301069, + "learning_rate": 5.053650153857237e-07, + "loss": 0.77450031, + "num_input_tokens_seen": 138834920, + "step": 6450, + "time_per_iteration": 2.7257578372955322 + }, + { + "auxiliary_loss_clip": 0.01144229, + "auxiliary_loss_mlp": 0.01026313, + "balance_loss_clip": 1.01186776, + "balance_loss_mlp": 1.0187881, + "epoch": 0.7756868875127758, + "flos": 18693623007360.0, + "grad_norm": 1.6809304120606439, + "language_loss": 0.69730532, + "learning_rate": 5.048475277121214e-07, + "loss": 0.71901071, + "num_input_tokens_seen": 138852135, + "step": 6451, + "time_per_iteration": 2.657362937927246 + }, + { + "auxiliary_loss_clip": 0.01142441, + "auxiliary_loss_mlp": 0.01021981, + "balance_loss_clip": 1.00943708, + "balance_loss_mlp": 1.01507604, + "epoch": 0.7758071304034149, + "flos": 28404191543040.0, + "grad_norm": 2.020885003743218, + "language_loss": 0.77249199, + "learning_rate": 5.043302668519598e-07, + "loss": 0.79413617, + "num_input_tokens_seen": 138871470, + "step": 6452, + "time_per_iteration": 2.7884459495544434 + }, + { + "auxiliary_loss_clip": 0.01145368, + "auxiliary_loss_mlp": 0.01025217, + "balance_loss_clip": 1.01028049, + "balance_loss_mlp": 1.01788545, + "epoch": 0.775927373294054, + "flos": 20595344670720.0, + "grad_norm": 1.8413897727691682, + "language_loss": 0.72156626, + "learning_rate": 5.038132328837079e-07, + "loss": 0.74327207, + "num_input_tokens_seen": 138889860, + "step": 6453, + "time_per_iteration": 2.6505515575408936 + }, + { + "auxiliary_loss_clip": 0.01143524, + "auxiliary_loss_mlp": 0.01027981, + "balance_loss_clip": 1.01053858, + "balance_loss_mlp": 1.02046537, + "epoch": 0.7760476161846931, + "flos": 22526368853760.0, + "grad_norm": 2.1814423232935383, + "language_loss": 0.74335253, + "learning_rate": 5.032964258857993e-07, + "loss": 0.76506758, + "num_input_tokens_seen": 138909955, + "step": 6454, + "time_per_iteration": 2.6970407962799072 + }, + { + "auxiliary_loss_clip": 0.01139178, + "auxiliary_loss_mlp": 0.0102481, + "balance_loss_clip": 1.00631404, + "balance_loss_mlp": 1.01716268, + "epoch": 0.7761678590753321, + "flos": 48651488403840.0, + "grad_norm": 1.7684389638803095, + "language_loss": 0.68328011, + "learning_rate": 5.027798459366329e-07, + "loss": 0.70491993, + "num_input_tokens_seen": 138935320, + "step": 6455, + "time_per_iteration": 2.856266975402832 + }, + { + "auxiliary_loss_clip": 0.0114747, + "auxiliary_loss_mlp": 0.01026037, + "balance_loss_clip": 1.01141906, + "balance_loss_mlp": 1.01839566, + "epoch": 0.7762881019659713, + "flos": 26177047637760.0, + "grad_norm": 2.6004638007176357, + "language_loss": 0.63411534, + "learning_rate": 5.02263493114573e-07, + "loss": 0.65585041, + "num_input_tokens_seen": 138957115, + "step": 6456, + "time_per_iteration": 2.6880037784576416 + }, + { + "auxiliary_loss_clip": 0.01165227, + "auxiliary_loss_mlp": 0.01027525, + "balance_loss_clip": 1.04802442, + "balance_loss_mlp": 1.02030134, + "epoch": 0.7764083448566104, + "flos": 20588341518720.0, + "grad_norm": 2.424885090742023, + "language_loss": 0.77042103, + "learning_rate": 5.017473674979502e-07, + "loss": 0.79234862, + "num_input_tokens_seen": 138973140, + "step": 6457, + "time_per_iteration": 2.560831069946289 + }, + { + "auxiliary_loss_clip": 0.00998293, + "auxiliary_loss_mlp": 0.01005458, + "balance_loss_clip": 0.90430629, + "balance_loss_mlp": 1.00421798, + "epoch": 0.7765285877472494, + "flos": 67293078560640.0, + "grad_norm": 0.7432952932766667, + "language_loss": 0.58299136, + "learning_rate": 5.01231469165061e-07, + "loss": 0.60302889, + "num_input_tokens_seen": 139028965, + "step": 6458, + "time_per_iteration": 4.042046308517456 + }, + { + "auxiliary_loss_clip": 0.01045881, + "auxiliary_loss_mlp": 0.00998085, + "balance_loss_clip": 0.97594988, + "balance_loss_mlp": 0.99703038, + "epoch": 0.7766488306378886, + "flos": 61344476121600.0, + "grad_norm": 0.827573049006621, + "language_loss": 0.56869197, + "learning_rate": 5.007157981941663e-07, + "loss": 0.58913165, + "num_input_tokens_seen": 139094325, + "step": 6459, + "time_per_iteration": 3.300374984741211 + }, + { + "auxiliary_loss_clip": 0.01025798, + "auxiliary_loss_mlp": 0.00998558, + "balance_loss_clip": 0.93796861, + "balance_loss_mlp": 0.99746764, + "epoch": 0.7767690735285276, + "flos": 62946199393920.0, + "grad_norm": 0.8818073819639712, + "language_loss": 0.67467058, + "learning_rate": 5.002003546634928e-07, + "loss": 0.6949141, + "num_input_tokens_seen": 139150425, + "step": 6460, + "time_per_iteration": 4.125439643859863 + }, + { + "auxiliary_loss_clip": 0.01077282, + "auxiliary_loss_mlp": 0.01026332, + "balance_loss_clip": 0.90226686, + "balance_loss_mlp": 1.01896799, + "epoch": 0.7768893164191667, + "flos": 20886400575360.0, + "grad_norm": 1.6315204652172404, + "language_loss": 0.75940073, + "learning_rate": 4.996851386512331e-07, + "loss": 0.78043681, + "num_input_tokens_seen": 139169130, + "step": 6461, + "time_per_iteration": 2.779444694519043 + }, + { + "auxiliary_loss_clip": 0.01121617, + "auxiliary_loss_mlp": 0.01028811, + "balance_loss_clip": 0.9746294, + "balance_loss_mlp": 1.02097297, + "epoch": 0.7770095593098058, + "flos": 20704584908160.0, + "grad_norm": 1.9104564010422684, + "language_loss": 0.83260667, + "learning_rate": 4.991701502355444e-07, + "loss": 0.85411096, + "num_input_tokens_seen": 139189595, + "step": 6462, + "time_per_iteration": 3.6190245151519775 + }, + { + "auxiliary_loss_clip": 0.01143137, + "auxiliary_loss_mlp": 0.01027765, + "balance_loss_clip": 1.0088135, + "balance_loss_mlp": 1.02079678, + "epoch": 0.7771298022004449, + "flos": 24717709877760.0, + "grad_norm": 1.6347325339995085, + "language_loss": 0.76123154, + "learning_rate": 4.986553894945518e-07, + "loss": 0.78294063, + "num_input_tokens_seen": 139210805, + "step": 6463, + "time_per_iteration": 2.920241594314575 + }, + { + "auxiliary_loss_clip": 0.01076741, + "auxiliary_loss_mlp": 0.010235, + "balance_loss_clip": 0.89506638, + "balance_loss_mlp": 1.01615667, + "epoch": 0.777250045091084, + "flos": 25009232659200.0, + "grad_norm": 2.5079786076132953, + "language_loss": 0.86131656, + "learning_rate": 4.981408565063416e-07, + "loss": 0.88231897, + "num_input_tokens_seen": 139230750, + "step": 6464, + "time_per_iteration": 2.7623369693756104 + }, + { + "auxiliary_loss_clip": 0.0116831, + "auxiliary_loss_mlp": 0.0102551, + "balance_loss_clip": 1.04956007, + "balance_loss_mlp": 1.01810753, + "epoch": 0.777370287981723, + "flos": 20119887319680.0, + "grad_norm": 4.403620253785857, + "language_loss": 0.75849497, + "learning_rate": 4.976265513489701e-07, + "loss": 0.78043318, + "num_input_tokens_seen": 139250720, + "step": 6465, + "time_per_iteration": 2.6134424209594727 + }, + { + "auxiliary_loss_clip": 0.01143711, + "auxiliary_loss_mlp": 0.01024734, + "balance_loss_clip": 1.01005125, + "balance_loss_mlp": 1.01719391, + "epoch": 0.7774905308723622, + "flos": 21718809331200.0, + "grad_norm": 1.8765821886060614, + "language_loss": 0.80408013, + "learning_rate": 4.971124741004562e-07, + "loss": 0.8257646, + "num_input_tokens_seen": 139269720, + "step": 6466, + "time_per_iteration": 2.6662065982818604 + }, + { + "auxiliary_loss_clip": 0.01141913, + "auxiliary_loss_mlp": 0.01020567, + "balance_loss_clip": 1.00951719, + "balance_loss_mlp": 1.01332474, + "epoch": 0.7776107737630013, + "flos": 16034115093120.0, + "grad_norm": 1.7675780754522668, + "language_loss": 0.76906276, + "learning_rate": 4.965986248387846e-07, + "loss": 0.79068756, + "num_input_tokens_seen": 139288035, + "step": 6467, + "time_per_iteration": 2.6227943897247314 + }, + { + "auxiliary_loss_clip": 0.01120385, + "auxiliary_loss_mlp": 0.01026754, + "balance_loss_clip": 0.970366, + "balance_loss_mlp": 1.01954198, + "epoch": 0.7777310166536403, + "flos": 24790895838720.0, + "grad_norm": 1.6552213113840553, + "language_loss": 0.77287805, + "learning_rate": 4.960850036419073e-07, + "loss": 0.79434943, + "num_input_tokens_seen": 139307135, + "step": 6468, + "time_per_iteration": 2.6890172958374023 + }, + { + "auxiliary_loss_clip": 0.01116836, + "auxiliary_loss_mlp": 0.01026323, + "balance_loss_clip": 0.9736712, + "balance_loss_mlp": 1.01916468, + "epoch": 0.7778512595442795, + "flos": 17272530253440.0, + "grad_norm": 1.8320762880773327, + "language_loss": 0.78797424, + "learning_rate": 4.955716105877378e-07, + "loss": 0.8094058, + "num_input_tokens_seen": 139325905, + "step": 6469, + "time_per_iteration": 2.6608948707580566 + }, + { + "auxiliary_loss_clip": 0.01148137, + "auxiliary_loss_mlp": 0.00509862, + "balance_loss_clip": 1.01169813, + "balance_loss_mlp": 0.0, + "epoch": 0.7779715024349185, + "flos": 17748418567680.0, + "grad_norm": 1.7741834258372595, + "language_loss": 0.83335489, + "learning_rate": 4.950584457541598e-07, + "loss": 0.84993488, + "num_input_tokens_seen": 139344370, + "step": 6470, + "time_per_iteration": 3.5727505683898926 + }, + { + "auxiliary_loss_clip": 0.0114722, + "auxiliary_loss_mlp": 0.01029276, + "balance_loss_clip": 1.01159263, + "balance_loss_mlp": 1.02181983, + "epoch": 0.7780917453255576, + "flos": 24316875031680.0, + "grad_norm": 1.362557728521974, + "language_loss": 0.81919408, + "learning_rate": 4.945455092190183e-07, + "loss": 0.84095907, + "num_input_tokens_seen": 139365625, + "step": 6471, + "time_per_iteration": 2.755084753036499 + }, + { + "auxiliary_loss_clip": 0.01063278, + "auxiliary_loss_mlp": 0.00999563, + "balance_loss_clip": 1.01189446, + "balance_loss_mlp": 0.99844819, + "epoch": 0.7782119882161967, + "flos": 56364601530240.0, + "grad_norm": 0.6849858134589645, + "language_loss": 0.55979109, + "learning_rate": 4.940328010601271e-07, + "loss": 0.58041954, + "num_input_tokens_seen": 139430540, + "step": 6472, + "time_per_iteration": 3.2309563159942627 + }, + { + "auxiliary_loss_clip": 0.01129422, + "auxiliary_loss_mlp": 0.01030708, + "balance_loss_clip": 0.97786552, + "balance_loss_mlp": 1.02287042, + "epoch": 0.7783322311068358, + "flos": 46789986994560.0, + "grad_norm": 2.267295277162185, + "language_loss": 0.76806545, + "learning_rate": 4.935203213552621e-07, + "loss": 0.78966677, + "num_input_tokens_seen": 139454280, + "step": 6473, + "time_per_iteration": 2.8646280765533447 + }, + { + "auxiliary_loss_clip": 0.01123545, + "auxiliary_loss_mlp": 0.01024724, + "balance_loss_clip": 0.97532564, + "balance_loss_mlp": 1.01760435, + "epoch": 0.7784524739974749, + "flos": 19057864872960.0, + "grad_norm": 3.055884882116718, + "language_loss": 0.6676414, + "learning_rate": 4.930080701821662e-07, + "loss": 0.68912405, + "num_input_tokens_seen": 139471745, + "step": 6474, + "time_per_iteration": 2.6656148433685303 + }, + { + "auxiliary_loss_clip": 0.01120437, + "auxiliary_loss_mlp": 0.0102339, + "balance_loss_clip": 0.97213739, + "balance_loss_mlp": 1.016536, + "epoch": 0.778572716888114, + "flos": 24791111320320.0, + "grad_norm": 4.486952471234579, + "language_loss": 0.7689755, + "learning_rate": 4.92496047618548e-07, + "loss": 0.7904138, + "num_input_tokens_seen": 139491505, + "step": 6475, + "time_per_iteration": 2.7495250701904297 + }, + { + "auxiliary_loss_clip": 0.0114461, + "auxiliary_loss_mlp": 0.01026781, + "balance_loss_clip": 1.01286745, + "balance_loss_mlp": 1.0198791, + "epoch": 0.7786929597787531, + "flos": 20078086867200.0, + "grad_norm": 2.132693557655257, + "language_loss": 0.77715629, + "learning_rate": 4.919842537420811e-07, + "loss": 0.79887021, + "num_input_tokens_seen": 139508620, + "step": 6476, + "time_per_iteration": 2.702070474624634 + }, + { + "auxiliary_loss_clip": 0.01121488, + "auxiliary_loss_mlp": 0.0102398, + "balance_loss_clip": 0.97565615, + "balance_loss_mlp": 1.01695859, + "epoch": 0.7788132026693921, + "flos": 21872220318720.0, + "grad_norm": 1.5617384879812999, + "language_loss": 0.79112542, + "learning_rate": 4.91472688630404e-07, + "loss": 0.81258011, + "num_input_tokens_seen": 139529360, + "step": 6477, + "time_per_iteration": 2.683603525161743 + }, + { + "auxiliary_loss_clip": 0.01165203, + "auxiliary_loss_mlp": 0.01024371, + "balance_loss_clip": 1.04983485, + "balance_loss_mlp": 1.01705122, + "epoch": 0.7789334455600313, + "flos": 11181937351680.0, + "grad_norm": 1.7779963538404642, + "language_loss": 0.73930788, + "learning_rate": 4.909613523611202e-07, + "loss": 0.76120365, + "num_input_tokens_seen": 139546240, + "step": 6478, + "time_per_iteration": 2.5972869396209717 + }, + { + "auxiliary_loss_clip": 0.01071632, + "auxiliary_loss_mlp": 0.00510114, + "balance_loss_clip": 0.89355528, + "balance_loss_mlp": 0.0, + "epoch": 0.7790536884506704, + "flos": 28695427015680.0, + "grad_norm": 2.688116048003215, + "language_loss": 0.75112391, + "learning_rate": 4.904502450117991e-07, + "loss": 0.76694137, + "num_input_tokens_seen": 139567200, + "step": 6479, + "time_per_iteration": 2.7299256324768066 + }, + { + "auxiliary_loss_clip": 0.01118586, + "auxiliary_loss_mlp": 0.01026703, + "balance_loss_clip": 0.97647655, + "balance_loss_mlp": 1.01934505, + "epoch": 0.7791739313413094, + "flos": 11072302064640.0, + "grad_norm": 2.223372747937097, + "language_loss": 0.72410244, + "learning_rate": 4.899393666599762e-07, + "loss": 0.7455554, + "num_input_tokens_seen": 139583775, + "step": 6480, + "time_per_iteration": 2.6530046463012695 + }, + { + "auxiliary_loss_clip": 0.01163987, + "auxiliary_loss_mlp": 0.0102234, + "balance_loss_clip": 1.04665804, + "balance_loss_mlp": 1.01522636, + "epoch": 0.7792941742319486, + "flos": 14679276975360.0, + "grad_norm": 2.679159470102534, + "language_loss": 0.72799927, + "learning_rate": 4.894287173831506e-07, + "loss": 0.74986255, + "num_input_tokens_seen": 139599735, + "step": 6481, + "time_per_iteration": 2.6084372997283936 + }, + { + "auxiliary_loss_clip": 0.01120906, + "auxiliary_loss_mlp": 0.01027554, + "balance_loss_clip": 0.97126043, + "balance_loss_mlp": 1.01987958, + "epoch": 0.7794144171225876, + "flos": 23258874908160.0, + "grad_norm": 2.1098957618864436, + "language_loss": 0.84335208, + "learning_rate": 4.889182972587877e-07, + "loss": 0.86483669, + "num_input_tokens_seen": 139619030, + "step": 6482, + "time_per_iteration": 2.7136735916137695 + }, + { + "auxiliary_loss_clip": 0.01107069, + "auxiliary_loss_mlp": 0.01030845, + "balance_loss_clip": 0.93746769, + "balance_loss_mlp": 1.02437508, + "epoch": 0.7795346600132267, + "flos": 21507080613120.0, + "grad_norm": 2.1151176172787873, + "language_loss": 0.66266775, + "learning_rate": 4.884081063643177e-07, + "loss": 0.68404686, + "num_input_tokens_seen": 139637690, + "step": 6483, + "time_per_iteration": 2.7520029544830322 + }, + { + "auxiliary_loss_clip": 0.01016999, + "auxiliary_loss_mlp": 0.01002379, + "balance_loss_clip": 0.93717891, + "balance_loss_mlp": 1.00107944, + "epoch": 0.7796549029038659, + "flos": 70052273694720.0, + "grad_norm": 0.8879163342616772, + "language_loss": 0.52532303, + "learning_rate": 4.878981447771353e-07, + "loss": 0.54551673, + "num_input_tokens_seen": 139692070, + "step": 6484, + "time_per_iteration": 3.2187087535858154 + }, + { + "auxiliary_loss_clip": 0.01092284, + "auxiliary_loss_mlp": 0.01027447, + "balance_loss_clip": 0.9342882, + "balance_loss_mlp": 1.01965678, + "epoch": 0.7797751457945049, + "flos": 23989405714560.0, + "grad_norm": 1.5702682548719764, + "language_loss": 0.7284314, + "learning_rate": 4.873884125746035e-07, + "loss": 0.74962872, + "num_input_tokens_seen": 139713745, + "step": 6485, + "time_per_iteration": 3.710150718688965 + }, + { + "auxiliary_loss_clip": 0.01115788, + "auxiliary_loss_mlp": 0.01025226, + "balance_loss_clip": 0.97211421, + "balance_loss_mlp": 1.0180881, + "epoch": 0.779895388685144, + "flos": 22674751937280.0, + "grad_norm": 2.726222409833096, + "language_loss": 0.726246, + "learning_rate": 4.868789098340456e-07, + "loss": 0.74765617, + "num_input_tokens_seen": 139731650, + "step": 6486, + "time_per_iteration": 3.610107183456421 + }, + { + "auxiliary_loss_clip": 0.01096752, + "auxiliary_loss_mlp": 0.01027699, + "balance_loss_clip": 0.93450212, + "balance_loss_mlp": 1.02032018, + "epoch": 0.7800156315757831, + "flos": 23768698596480.0, + "grad_norm": 2.2725065037457726, + "language_loss": 0.72849643, + "learning_rate": 4.863696366327543e-07, + "loss": 0.74974084, + "num_input_tokens_seen": 139750820, + "step": 6487, + "time_per_iteration": 2.7108826637268066 + }, + { + "auxiliary_loss_clip": 0.01144004, + "auxiliary_loss_mlp": 0.01028195, + "balance_loss_clip": 1.00841117, + "balance_loss_mlp": 1.02045822, + "epoch": 0.7801358744664222, + "flos": 26429714881920.0, + "grad_norm": 1.807872287811368, + "language_loss": 0.78013802, + "learning_rate": 4.85860593047986e-07, + "loss": 0.80186009, + "num_input_tokens_seen": 139770885, + "step": 6488, + "time_per_iteration": 3.6307055950164795 + }, + { + "auxiliary_loss_clip": 0.01088423, + "auxiliary_loss_mlp": 0.01025254, + "balance_loss_clip": 0.92965609, + "balance_loss_mlp": 1.01811337, + "epoch": 0.7802561173570612, + "flos": 26322162583680.0, + "grad_norm": 2.117789515514493, + "language_loss": 0.74698555, + "learning_rate": 4.853517791569613e-07, + "loss": 0.76812232, + "num_input_tokens_seen": 139793065, + "step": 6489, + "time_per_iteration": 2.7387003898620605 + }, + { + "auxiliary_loss_clip": 0.01123292, + "auxiliary_loss_mlp": 0.00509973, + "balance_loss_clip": 0.97135538, + "balance_loss_mlp": 0.0, + "epoch": 0.7803763602477004, + "flos": 40333751596800.0, + "grad_norm": 1.6106420955857637, + "language_loss": 0.65919268, + "learning_rate": 4.848431950368684e-07, + "loss": 0.67552531, + "num_input_tokens_seen": 139815625, + "step": 6490, + "time_per_iteration": 2.888005495071411 + }, + { + "auxiliary_loss_clip": 0.01063245, + "auxiliary_loss_mlp": 0.00500702, + "balance_loss_clip": 1.01164007, + "balance_loss_mlp": 0.0, + "epoch": 0.7804966031383395, + "flos": 67001448038400.0, + "grad_norm": 0.7025885301408923, + "language_loss": 0.55748284, + "learning_rate": 4.843348407648569e-07, + "loss": 0.57312238, + "num_input_tokens_seen": 139876905, + "step": 6491, + "time_per_iteration": 3.140892267227173 + }, + { + "auxiliary_loss_clip": 0.01144671, + "auxiliary_loss_mlp": 0.01027569, + "balance_loss_clip": 1.00744414, + "balance_loss_mlp": 1.02011204, + "epoch": 0.7806168460289785, + "flos": 17740733057280.0, + "grad_norm": 2.984798668178119, + "language_loss": 0.83542788, + "learning_rate": 4.838267164180457e-07, + "loss": 0.85715026, + "num_input_tokens_seen": 139892575, + "step": 6492, + "time_per_iteration": 2.570096492767334 + }, + { + "auxiliary_loss_clip": 0.01167515, + "auxiliary_loss_mlp": 0.01024225, + "balance_loss_clip": 1.04862189, + "balance_loss_mlp": 1.01700687, + "epoch": 0.7807370889196176, + "flos": 23946240545280.0, + "grad_norm": 2.3160444939259137, + "language_loss": 0.84115779, + "learning_rate": 4.833188220735156e-07, + "loss": 0.8630752, + "num_input_tokens_seen": 139912245, + "step": 6493, + "time_per_iteration": 2.691912889480591 + }, + { + "auxiliary_loss_clip": 0.01142537, + "auxiliary_loss_mlp": 0.01029198, + "balance_loss_clip": 1.01033068, + "balance_loss_mlp": 1.02242422, + "epoch": 0.7808573318102567, + "flos": 18989024457600.0, + "grad_norm": 2.0980501548184103, + "language_loss": 0.74787778, + "learning_rate": 4.828111578083152e-07, + "loss": 0.76959515, + "num_input_tokens_seen": 139929150, + "step": 6494, + "time_per_iteration": 2.6245944499969482 + }, + { + "auxiliary_loss_clip": 0.01117417, + "auxiliary_loss_mlp": 0.01024977, + "balance_loss_clip": 0.97303098, + "balance_loss_mlp": 1.0179317, + "epoch": 0.7809775747008958, + "flos": 23980750536960.0, + "grad_norm": 1.859647769420174, + "language_loss": 0.81211752, + "learning_rate": 4.823037236994556e-07, + "loss": 0.83354145, + "num_input_tokens_seen": 139947315, + "step": 6495, + "time_per_iteration": 3.666391372680664 + }, + { + "auxiliary_loss_clip": 0.01045659, + "auxiliary_loss_mlp": 0.01000253, + "balance_loss_clip": 0.97499609, + "balance_loss_mlp": 0.99909097, + "epoch": 0.7810978175915348, + "flos": 68535875180160.0, + "grad_norm": 0.7149002475650713, + "language_loss": 0.56288934, + "learning_rate": 4.817965198239136e-07, + "loss": 0.58334845, + "num_input_tokens_seen": 140013775, + "step": 6496, + "time_per_iteration": 3.1783931255340576 + }, + { + "auxiliary_loss_clip": 0.01095394, + "auxiliary_loss_mlp": 0.01032203, + "balance_loss_clip": 0.93285215, + "balance_loss_mlp": 1.02454114, + "epoch": 0.781218060482174, + "flos": 19642131498240.0, + "grad_norm": 2.1065610046631345, + "language_loss": 0.74843597, + "learning_rate": 4.812895462586331e-07, + "loss": 0.76971197, + "num_input_tokens_seen": 140031600, + "step": 6497, + "time_per_iteration": 2.6922402381896973 + }, + { + "auxiliary_loss_clip": 0.01100278, + "auxiliary_loss_mlp": 0.01032045, + "balance_loss_clip": 0.93550646, + "balance_loss_mlp": 1.02545285, + "epoch": 0.7813383033728131, + "flos": 25627865621760.0, + "grad_norm": 1.6322059299858869, + "language_loss": 0.81889892, + "learning_rate": 4.807828030805207e-07, + "loss": 0.84022212, + "num_input_tokens_seen": 140050590, + "step": 6498, + "time_per_iteration": 2.7581357955932617 + }, + { + "auxiliary_loss_clip": 0.01143399, + "auxiliary_loss_mlp": 0.01027454, + "balance_loss_clip": 1.01419568, + "balance_loss_mlp": 1.01990771, + "epoch": 0.7814585462634521, + "flos": 20485924865280.0, + "grad_norm": 1.8659364197689816, + "language_loss": 0.68019378, + "learning_rate": 4.802762903664495e-07, + "loss": 0.70190233, + "num_input_tokens_seen": 140069770, + "step": 6499, + "time_per_iteration": 2.618340492248535 + }, + { + "auxiliary_loss_clip": 0.01123483, + "auxiliary_loss_mlp": 0.01033235, + "balance_loss_clip": 0.97297442, + "balance_loss_mlp": 1.02589417, + "epoch": 0.7815787891540913, + "flos": 22304297018880.0, + "grad_norm": 2.146460998533379, + "language_loss": 0.73477489, + "learning_rate": 4.797700081932565e-07, + "loss": 0.75634205, + "num_input_tokens_seen": 140087635, + "step": 6500, + "time_per_iteration": 2.66375732421875 + }, + { + "auxiliary_loss_clip": 0.01043819, + "auxiliary_loss_mlp": 0.01026911, + "balance_loss_clip": 0.85479045, + "balance_loss_mlp": 1.01969254, + "epoch": 0.7816990320447303, + "flos": 22600668136320.0, + "grad_norm": 2.2647531596882473, + "language_loss": 0.82208413, + "learning_rate": 4.792639566377442e-07, + "loss": 0.84279144, + "num_input_tokens_seen": 140105045, + "step": 6501, + "time_per_iteration": 2.8065943717956543 + }, + { + "auxiliary_loss_clip": 0.01136304, + "auxiliary_loss_mlp": 0.01027002, + "balance_loss_clip": 1.00764263, + "balance_loss_mlp": 1.01976931, + "epoch": 0.7818192749353694, + "flos": 24935974871040.0, + "grad_norm": 1.7227274998273092, + "language_loss": 0.77289242, + "learning_rate": 4.78758135776681e-07, + "loss": 0.7945255, + "num_input_tokens_seen": 140124900, + "step": 6502, + "time_per_iteration": 2.7428977489471436 + }, + { + "auxiliary_loss_clip": 0.0112205, + "auxiliary_loss_mlp": 0.01028709, + "balance_loss_clip": 0.97385085, + "balance_loss_mlp": 1.02094269, + "epoch": 0.7819395178260086, + "flos": 23733039369600.0, + "grad_norm": 1.9063043709330922, + "language_loss": 0.79048479, + "learning_rate": 4.782525456867989e-07, + "loss": 0.81199241, + "num_input_tokens_seen": 140143755, + "step": 6503, + "time_per_iteration": 2.6121606826782227 + }, + { + "auxiliary_loss_clip": 0.01099882, + "auxiliary_loss_mlp": 0.01027121, + "balance_loss_clip": 0.93472171, + "balance_loss_mlp": 1.01885366, + "epoch": 0.7820597607166476, + "flos": 23221671396480.0, + "grad_norm": 1.5011095490719624, + "language_loss": 0.82908183, + "learning_rate": 4.777471864447959e-07, + "loss": 0.85035181, + "num_input_tokens_seen": 140164495, + "step": 6504, + "time_per_iteration": 2.721182346343994 + }, + { + "auxiliary_loss_clip": 0.01120137, + "auxiliary_loss_mlp": 0.0102707, + "balance_loss_clip": 0.96979773, + "balance_loss_mlp": 1.01969111, + "epoch": 0.7821800036072867, + "flos": 22309540404480.0, + "grad_norm": 1.9484574545022015, + "language_loss": 0.80209637, + "learning_rate": 4.772420581273344e-07, + "loss": 0.82356834, + "num_input_tokens_seen": 140181980, + "step": 6505, + "time_per_iteration": 2.6510889530181885 + }, + { + "auxiliary_loss_clip": 0.01139874, + "auxiliary_loss_mlp": 0.01025613, + "balance_loss_clip": 1.01126122, + "balance_loss_mlp": 1.01815629, + "epoch": 0.7823002464979258, + "flos": 21544176384000.0, + "grad_norm": 3.362882200460213, + "language_loss": 0.76336503, + "learning_rate": 4.7673716081104134e-07, + "loss": 0.78501987, + "num_input_tokens_seen": 140202155, + "step": 6506, + "time_per_iteration": 2.666107416152954 + }, + { + "auxiliary_loss_clip": 0.01143092, + "auxiliary_loss_mlp": 0.0102738, + "balance_loss_clip": 1.01270437, + "balance_loss_mlp": 1.02000391, + "epoch": 0.7824204893885649, + "flos": 24535642815360.0, + "grad_norm": 1.6527172864253394, + "language_loss": 0.84531319, + "learning_rate": 4.762324945725109e-07, + "loss": 0.86701792, + "num_input_tokens_seen": 140221600, + "step": 6507, + "time_per_iteration": 2.6668570041656494 + }, + { + "auxiliary_loss_clip": 0.01117917, + "auxiliary_loss_mlp": 0.01026931, + "balance_loss_clip": 0.97544837, + "balance_loss_mlp": 1.01930201, + "epoch": 0.782540732279204, + "flos": 27415211402880.0, + "grad_norm": 1.6677115252852035, + "language_loss": 0.75572759, + "learning_rate": 4.7572805948829844e-07, + "loss": 0.77717602, + "num_input_tokens_seen": 140241860, + "step": 6508, + "time_per_iteration": 2.7050278186798096 + }, + { + "auxiliary_loss_clip": 0.0107664, + "auxiliary_loss_mlp": 0.01029239, + "balance_loss_clip": 0.89475948, + "balance_loss_mlp": 1.02213383, + "epoch": 0.7826609751698431, + "flos": 24353216616960.0, + "grad_norm": 1.7045978580849823, + "language_loss": 0.71123731, + "learning_rate": 4.7522385563492795e-07, + "loss": 0.73229611, + "num_input_tokens_seen": 140262160, + "step": 6509, + "time_per_iteration": 2.8378443717956543 + }, + { + "auxiliary_loss_clip": 0.01099463, + "auxiliary_loss_mlp": 0.01029716, + "balance_loss_clip": 0.93507767, + "balance_loss_mlp": 1.0220654, + "epoch": 0.7827812180604822, + "flos": 23988543788160.0, + "grad_norm": 2.0074979675711426, + "language_loss": 0.70228362, + "learning_rate": 4.747198830888863e-07, + "loss": 0.72357541, + "num_input_tokens_seen": 140282030, + "step": 6510, + "time_per_iteration": 3.69269061088562 + }, + { + "auxiliary_loss_clip": 0.01115213, + "auxiliary_loss_mlp": 0.01027973, + "balance_loss_clip": 0.97127259, + "balance_loss_mlp": 1.02057052, + "epoch": 0.7829014609511212, + "flos": 27454318335360.0, + "grad_norm": 1.8975798173872291, + "language_loss": 0.68382263, + "learning_rate": 4.742161419266251e-07, + "loss": 0.7052545, + "num_input_tokens_seen": 140301190, + "step": 6511, + "time_per_iteration": 2.6736226081848145 + }, + { + "auxiliary_loss_clip": 0.01147032, + "auxiliary_loss_mlp": 0.01029734, + "balance_loss_clip": 1.01106119, + "balance_loss_mlp": 1.02189589, + "epoch": 0.7830217038417604, + "flos": 29204532432000.0, + "grad_norm": 3.06729464875333, + "language_loss": 0.64897048, + "learning_rate": 4.7371263222456304e-07, + "loss": 0.6707381, + "num_input_tokens_seen": 140318510, + "step": 6512, + "time_per_iteration": 3.597464084625244 + }, + { + "auxiliary_loss_clip": 0.01039687, + "auxiliary_loss_mlp": 0.01000754, + "balance_loss_clip": 0.97322237, + "balance_loss_mlp": 0.99966925, + "epoch": 0.7831419467323995, + "flos": 60950895822720.0, + "grad_norm": 0.7960056118901324, + "language_loss": 0.61363697, + "learning_rate": 4.7320935405908004e-07, + "loss": 0.63404143, + "num_input_tokens_seen": 140379380, + "step": 6513, + "time_per_iteration": 3.1498825550079346 + }, + { + "auxiliary_loss_clip": 0.011714, + "auxiliary_loss_mlp": 0.01027256, + "balance_loss_clip": 1.04967928, + "balance_loss_mlp": 1.01961803, + "epoch": 0.7832621896230385, + "flos": 19682531320320.0, + "grad_norm": 1.9757134831469763, + "language_loss": 0.84344774, + "learning_rate": 4.7270630750652475e-07, + "loss": 0.86543429, + "num_input_tokens_seen": 140395335, + "step": 6514, + "time_per_iteration": 3.475416898727417 + }, + { + "auxiliary_loss_clip": 0.01140031, + "auxiliary_loss_mlp": 0.01025716, + "balance_loss_clip": 1.0081687, + "balance_loss_mlp": 1.01875997, + "epoch": 0.7833824325136777, + "flos": 25009232659200.0, + "grad_norm": 1.773441954573592, + "language_loss": 0.80108303, + "learning_rate": 4.7220349264320746e-07, + "loss": 0.8227405, + "num_input_tokens_seen": 140414420, + "step": 6515, + "time_per_iteration": 2.66160249710083 + }, + { + "auxiliary_loss_clip": 0.01044461, + "auxiliary_loss_mlp": 0.00998752, + "balance_loss_clip": 0.97571301, + "balance_loss_mlp": 0.99766678, + "epoch": 0.7835026754043167, + "flos": 68800142517120.0, + "grad_norm": 0.7377159636698356, + "language_loss": 0.54923981, + "learning_rate": 4.71700909545407e-07, + "loss": 0.56967193, + "num_input_tokens_seen": 140477365, + "step": 6516, + "time_per_iteration": 3.223567485809326 + }, + { + "auxiliary_loss_clip": 0.01145381, + "auxiliary_loss_mlp": 0.01025443, + "balance_loss_clip": 1.01104188, + "balance_loss_mlp": 1.01775992, + "epoch": 0.7836229182949558, + "flos": 19864598382720.0, + "grad_norm": 2.5002291003436277, + "language_loss": 0.76789784, + "learning_rate": 4.711985582893627e-07, + "loss": 0.78960603, + "num_input_tokens_seen": 140495885, + "step": 6517, + "time_per_iteration": 2.6577985286712646 + }, + { + "auxiliary_loss_clip": 0.01074039, + "auxiliary_loss_mlp": 0.01027887, + "balance_loss_clip": 0.8926329, + "balance_loss_mlp": 1.02037072, + "epoch": 0.783743161185595, + "flos": 22965843755520.0, + "grad_norm": 1.6395622503011766, + "language_loss": 0.71738666, + "learning_rate": 4.706964389512811e-07, + "loss": 0.73840594, + "num_input_tokens_seen": 140515920, + "step": 6518, + "time_per_iteration": 2.7192676067352295 + }, + { + "auxiliary_loss_clip": 0.01165874, + "auxiliary_loss_mlp": 0.01020996, + "balance_loss_clip": 1.05092573, + "balance_loss_mlp": 1.01443911, + "epoch": 0.783863404076234, + "flos": 12458489777280.0, + "grad_norm": 1.7718603914075375, + "language_loss": 0.87280488, + "learning_rate": 4.701945516073345e-07, + "loss": 0.89467359, + "num_input_tokens_seen": 140533395, + "step": 6519, + "time_per_iteration": 2.614284038543701 + }, + { + "auxiliary_loss_clip": 0.01092916, + "auxiliary_loss_mlp": 0.01024238, + "balance_loss_clip": 0.93478239, + "balance_loss_mlp": 1.01656103, + "epoch": 0.7839836469668731, + "flos": 24243940465920.0, + "grad_norm": 2.1488480129811465, + "language_loss": 0.74781561, + "learning_rate": 4.696928963336577e-07, + "loss": 0.76898718, + "num_input_tokens_seen": 140552825, + "step": 6520, + "time_per_iteration": 2.7084431648254395 + }, + { + "auxiliary_loss_clip": 0.01040043, + "auxiliary_loss_mlp": 0.00999896, + "balance_loss_clip": 0.97359407, + "balance_loss_mlp": 0.99880552, + "epoch": 0.7841038898575122, + "flos": 62121978938880.0, + "grad_norm": 0.8527408475129893, + "language_loss": 0.60978538, + "learning_rate": 4.6919147320635224e-07, + "loss": 0.63018477, + "num_input_tokens_seen": 140615535, + "step": 6521, + "time_per_iteration": 4.106834650039673 + }, + { + "auxiliary_loss_clip": 0.01145614, + "auxiliary_loss_mlp": 0.01024345, + "balance_loss_clip": 1.01132929, + "balance_loss_mlp": 1.01710582, + "epoch": 0.7842241327481513, + "flos": 20193899293440.0, + "grad_norm": 6.76616716244906, + "language_loss": 0.73077017, + "learning_rate": 4.6869028230148286e-07, + "loss": 0.75246978, + "num_input_tokens_seen": 140633330, + "step": 6522, + "time_per_iteration": 2.7894015312194824 + }, + { + "auxiliary_loss_clip": 0.01089208, + "auxiliary_loss_mlp": 0.01027642, + "balance_loss_clip": 0.92960513, + "balance_loss_mlp": 1.01992369, + "epoch": 0.7843443756387903, + "flos": 28074531496320.0, + "grad_norm": 2.145031743894771, + "language_loss": 0.59814513, + "learning_rate": 4.6818932369507957e-07, + "loss": 0.6193136, + "num_input_tokens_seen": 140652830, + "step": 6523, + "time_per_iteration": 2.6975364685058594 + }, + { + "auxiliary_loss_clip": 0.01142803, + "auxiliary_loss_mlp": 0.01038426, + "balance_loss_clip": 1.01155579, + "balance_loss_mlp": 1.03097534, + "epoch": 0.7844646185294295, + "flos": 21323397438720.0, + "grad_norm": 2.6862339488947926, + "language_loss": 0.89098173, + "learning_rate": 4.676885974631386e-07, + "loss": 0.91279399, + "num_input_tokens_seen": 140671190, + "step": 6524, + "time_per_iteration": 2.6458120346069336 + }, + { + "auxiliary_loss_clip": 0.01145198, + "auxiliary_loss_mlp": 0.01021437, + "balance_loss_clip": 1.01094258, + "balance_loss_mlp": 1.01421869, + "epoch": 0.7845848614200686, + "flos": 23656585271040.0, + "grad_norm": 2.234869454021775, + "language_loss": 0.81178117, + "learning_rate": 4.67188103681619e-07, + "loss": 0.83344752, + "num_input_tokens_seen": 140690975, + "step": 6525, + "time_per_iteration": 2.7245452404022217 + }, + { + "auxiliary_loss_clip": 0.01142401, + "auxiliary_loss_mlp": 0.0050982, + "balance_loss_clip": 1.01543796, + "balance_loss_mlp": 0.0, + "epoch": 0.7847051043107076, + "flos": 23402194174080.0, + "grad_norm": 2.108912913856117, + "language_loss": 0.69064128, + "learning_rate": 4.666878424264453e-07, + "loss": 0.70716351, + "num_input_tokens_seen": 140710930, + "step": 6526, + "time_per_iteration": 2.6776528358459473 + }, + { + "auxiliary_loss_clip": 0.01112196, + "auxiliary_loss_mlp": 0.01024369, + "balance_loss_clip": 0.97187316, + "balance_loss_mlp": 1.01718998, + "epoch": 0.7848253472013467, + "flos": 19022277473280.0, + "grad_norm": 1.6046202996482728, + "language_loss": 0.73788947, + "learning_rate": 4.661878137735069e-07, + "loss": 0.75925511, + "num_input_tokens_seen": 140729120, + "step": 6527, + "time_per_iteration": 2.6432886123657227 + }, + { + "auxiliary_loss_clip": 0.01118275, + "auxiliary_loss_mlp": 0.010299, + "balance_loss_clip": 0.97140133, + "balance_loss_mlp": 1.02289045, + "epoch": 0.7849455900919858, + "flos": 21179180332800.0, + "grad_norm": 1.8668489525479885, + "language_loss": 0.74878746, + "learning_rate": 4.656880177986571e-07, + "loss": 0.77026916, + "num_input_tokens_seen": 140747665, + "step": 6528, + "time_per_iteration": 2.6979424953460693 + }, + { + "auxiliary_loss_clip": 0.01121248, + "auxiliary_loss_mlp": 0.01024856, + "balance_loss_clip": 0.97160494, + "balance_loss_mlp": 1.01745343, + "epoch": 0.7850658329826249, + "flos": 19536482620800.0, + "grad_norm": 2.1963178494481403, + "language_loss": 0.8150022, + "learning_rate": 4.6518845457771607e-07, + "loss": 0.83646321, + "num_input_tokens_seen": 140766525, + "step": 6529, + "time_per_iteration": 2.6375017166137695 + }, + { + "auxiliary_loss_clip": 0.01137274, + "auxiliary_loss_mlp": 0.00509759, + "balance_loss_clip": 1.00959635, + "balance_loss_mlp": 0.0, + "epoch": 0.7851860758732639, + "flos": 12495334152960.0, + "grad_norm": 2.0181169240252643, + "language_loss": 0.78960156, + "learning_rate": 4.646891241864652e-07, + "loss": 0.80607194, + "num_input_tokens_seen": 140785090, + "step": 6530, + "time_per_iteration": 2.645888090133667 + }, + { + "auxiliary_loss_clip": 0.01140341, + "auxiliary_loss_mlp": 0.01032722, + "balance_loss_clip": 1.00942886, + "balance_loss_mlp": 1.02464855, + "epoch": 0.7853063187639031, + "flos": 22960959505920.0, + "grad_norm": 1.7940397852531877, + "language_loss": 0.72948086, + "learning_rate": 4.6419002670065397e-07, + "loss": 0.75121152, + "num_input_tokens_seen": 140804670, + "step": 6531, + "time_per_iteration": 2.621530771255493 + }, + { + "auxiliary_loss_clip": 0.01098511, + "auxiliary_loss_mlp": 0.01020819, + "balance_loss_clip": 0.93553865, + "balance_loss_mlp": 1.01327932, + "epoch": 0.7854265616545422, + "flos": 17347260499200.0, + "grad_norm": 1.8379740106240319, + "language_loss": 0.86653244, + "learning_rate": 4.6369116219599445e-07, + "loss": 0.88772577, + "num_input_tokens_seen": 140820655, + "step": 6532, + "time_per_iteration": 2.6826839447021484 + }, + { + "auxiliary_loss_clip": 0.01096651, + "auxiliary_loss_mlp": 0.01022665, + "balance_loss_clip": 0.93385267, + "balance_loss_mlp": 1.01583409, + "epoch": 0.7855468045451812, + "flos": 23838293197440.0, + "grad_norm": 2.7591385259268852, + "language_loss": 0.79298866, + "learning_rate": 4.631925307481637e-07, + "loss": 0.8141818, + "num_input_tokens_seen": 140840470, + "step": 6533, + "time_per_iteration": 2.759941816329956 + }, + { + "auxiliary_loss_clip": 0.01120801, + "auxiliary_loss_mlp": 0.01028272, + "balance_loss_clip": 0.9751001, + "balance_loss_mlp": 1.02130401, + "epoch": 0.7856670474358204, + "flos": 25666792986240.0, + "grad_norm": 2.4375067473501995, + "language_loss": 0.75623626, + "learning_rate": 4.6269413243280533e-07, + "loss": 0.77772695, + "num_input_tokens_seen": 140859890, + "step": 6534, + "time_per_iteration": 2.7148773670196533 + }, + { + "auxiliary_loss_clip": 0.01125504, + "auxiliary_loss_mlp": 0.0102352, + "balance_loss_clip": 0.97588468, + "balance_loss_mlp": 1.01600385, + "epoch": 0.7857872903264594, + "flos": 18144656472960.0, + "grad_norm": 2.1853807368330216, + "language_loss": 0.74401164, + "learning_rate": 4.621959673255236e-07, + "loss": 0.76550186, + "num_input_tokens_seen": 140876190, + "step": 6535, + "time_per_iteration": 2.646486759185791 + }, + { + "auxiliary_loss_clip": 0.01073738, + "auxiliary_loss_mlp": 0.0102518, + "balance_loss_clip": 0.89603573, + "balance_loss_mlp": 1.01801586, + "epoch": 0.7859075332170985, + "flos": 14386138081920.0, + "grad_norm": 2.3759459928777487, + "language_loss": 0.90223974, + "learning_rate": 4.6169803550189135e-07, + "loss": 0.92322892, + "num_input_tokens_seen": 140891885, + "step": 6536, + "time_per_iteration": 2.6884841918945312 + }, + { + "auxiliary_loss_clip": 0.01069584, + "auxiliary_loss_mlp": 0.01024892, + "balance_loss_clip": 0.89915383, + "balance_loss_mlp": 1.01700044, + "epoch": 0.7860277761077377, + "flos": 19864059678720.0, + "grad_norm": 7.329166172474234, + "language_loss": 0.77583206, + "learning_rate": 4.6120033703744355e-07, + "loss": 0.79677677, + "num_input_tokens_seen": 140910780, + "step": 6537, + "time_per_iteration": 3.6808791160583496 + }, + { + "auxiliary_loss_clip": 0.01109668, + "auxiliary_loss_mlp": 0.01030645, + "balance_loss_clip": 0.9694007, + "balance_loss_mlp": 1.02331662, + "epoch": 0.7861480189983767, + "flos": 26396174557440.0, + "grad_norm": 2.077186950187351, + "language_loss": 0.78371418, + "learning_rate": 4.607028720076822e-07, + "loss": 0.80511731, + "num_input_tokens_seen": 140927460, + "step": 6538, + "time_per_iteration": 3.8402516841888428 + }, + { + "auxiliary_loss_clip": 0.0114365, + "auxiliary_loss_mlp": 0.01021505, + "balance_loss_clip": 1.01198816, + "balance_loss_mlp": 1.01433492, + "epoch": 0.7862682618890158, + "flos": 24236578177920.0, + "grad_norm": 1.8806114713965647, + "language_loss": 0.73732352, + "learning_rate": 4.6020564048807074e-07, + "loss": 0.75897503, + "num_input_tokens_seen": 140945135, + "step": 6539, + "time_per_iteration": 2.671877384185791 + }, + { + "auxiliary_loss_clip": 0.01144299, + "auxiliary_loss_mlp": 0.01026793, + "balance_loss_clip": 1.01069045, + "balance_loss_mlp": 1.01911008, + "epoch": 0.7863885047796549, + "flos": 47551508259840.0, + "grad_norm": 3.519104897461646, + "language_loss": 0.71719193, + "learning_rate": 4.5970864255403883e-07, + "loss": 0.73890287, + "num_input_tokens_seen": 140966660, + "step": 6540, + "time_per_iteration": 3.7580504417419434 + }, + { + "auxiliary_loss_clip": 0.01133914, + "auxiliary_loss_mlp": 0.01023877, + "balance_loss_clip": 1.0092994, + "balance_loss_mlp": 1.01666486, + "epoch": 0.786508747670294, + "flos": 24389234979840.0, + "grad_norm": 2.3301376418696256, + "language_loss": 0.8212586, + "learning_rate": 4.59211878280982e-07, + "loss": 0.8428365, + "num_input_tokens_seen": 140986175, + "step": 6541, + "time_per_iteration": 2.618943691253662 + }, + { + "auxiliary_loss_clip": 0.01120279, + "auxiliary_loss_mlp": 0.01025426, + "balance_loss_clip": 0.97206575, + "balance_loss_mlp": 1.01808214, + "epoch": 0.786628990560933, + "flos": 18041234238720.0, + "grad_norm": 4.274602925413464, + "language_loss": 0.70272529, + "learning_rate": 4.587153477442578e-07, + "loss": 0.72418231, + "num_input_tokens_seen": 141002490, + "step": 6542, + "time_per_iteration": 2.711305618286133 + }, + { + "auxiliary_loss_clip": 0.01171314, + "auxiliary_loss_mlp": 0.01029167, + "balance_loss_clip": 1.05169129, + "balance_loss_mlp": 1.02116227, + "epoch": 0.7867492334515722, + "flos": 25848860048640.0, + "grad_norm": 1.953192730167236, + "language_loss": 0.81695306, + "learning_rate": 4.582190510191899e-07, + "loss": 0.83895791, + "num_input_tokens_seen": 141021150, + "step": 6543, + "time_per_iteration": 2.6680119037628174 + }, + { + "auxiliary_loss_clip": 0.01092265, + "auxiliary_loss_mlp": 0.01028053, + "balance_loss_clip": 0.93408513, + "balance_loss_mlp": 1.02128816, + "epoch": 0.7868694763422113, + "flos": 16580819070720.0, + "grad_norm": 2.655980415335399, + "language_loss": 0.87118989, + "learning_rate": 4.5772298818106625e-07, + "loss": 0.89239305, + "num_input_tokens_seen": 141036940, + "step": 6544, + "time_per_iteration": 2.6768813133239746 + }, + { + "auxiliary_loss_clip": 0.01103917, + "auxiliary_loss_mlp": 0.01027835, + "balance_loss_clip": 0.93748629, + "balance_loss_mlp": 1.02032804, + "epoch": 0.7869897192328503, + "flos": 29386276272000.0, + "grad_norm": 13.924576725307778, + "language_loss": 0.72042781, + "learning_rate": 4.572271593051384e-07, + "loss": 0.74174535, + "num_input_tokens_seen": 141054295, + "step": 6545, + "time_per_iteration": 2.767151117324829 + }, + { + "auxiliary_loss_clip": 0.01065911, + "auxiliary_loss_mlp": 0.01025394, + "balance_loss_clip": 0.89593887, + "balance_loss_mlp": 1.0182476, + "epoch": 0.7871099621234895, + "flos": 17128923678720.0, + "grad_norm": 9.845857165490264, + "language_loss": 0.78324497, + "learning_rate": 4.567315644666245e-07, + "loss": 0.80415809, + "num_input_tokens_seen": 141073090, + "step": 6546, + "time_per_iteration": 2.7644107341766357 + }, + { + "auxiliary_loss_clip": 0.01090498, + "auxiliary_loss_mlp": 0.0102426, + "balance_loss_clip": 0.93484092, + "balance_loss_mlp": 1.01742578, + "epoch": 0.7872302050141285, + "flos": 23440187784960.0, + "grad_norm": 2.8094050556101213, + "language_loss": 0.84112501, + "learning_rate": 4.5623620374070507e-07, + "loss": 0.86227262, + "num_input_tokens_seen": 141092405, + "step": 6547, + "time_per_iteration": 3.695094108581543 + }, + { + "auxiliary_loss_clip": 0.01003402, + "auxiliary_loss_mlp": 0.01004785, + "balance_loss_clip": 0.90093946, + "balance_loss_mlp": 1.00369978, + "epoch": 0.7873504479047676, + "flos": 65959752689280.0, + "grad_norm": 0.7836655749674727, + "language_loss": 0.58379769, + "learning_rate": 4.557410772025263e-07, + "loss": 0.60387957, + "num_input_tokens_seen": 141154355, + "step": 6548, + "time_per_iteration": 3.3405423164367676 + }, + { + "auxiliary_loss_clip": 0.01118142, + "auxiliary_loss_mlp": 0.0102747, + "balance_loss_clip": 0.97243726, + "balance_loss_mlp": 1.01992965, + "epoch": 0.7874706907954068, + "flos": 23258336204160.0, + "grad_norm": 2.025619333191456, + "language_loss": 0.6610586, + "learning_rate": 4.5524618492719803e-07, + "loss": 0.68251467, + "num_input_tokens_seen": 141173575, + "step": 6549, + "time_per_iteration": 2.709336519241333 + }, + { + "auxiliary_loss_clip": 0.01145142, + "auxiliary_loss_mlp": 0.01021458, + "balance_loss_clip": 1.01199389, + "balance_loss_mlp": 1.01440668, + "epoch": 0.7875909336860458, + "flos": 28767786963840.0, + "grad_norm": 1.6452182423978434, + "language_loss": 0.79006147, + "learning_rate": 4.54751526989795e-07, + "loss": 0.81172746, + "num_input_tokens_seen": 141195415, + "step": 6550, + "time_per_iteration": 2.7063794136047363 + }, + { + "auxiliary_loss_clip": 0.01146267, + "auxiliary_loss_mlp": 0.01028436, + "balance_loss_clip": 1.01116931, + "balance_loss_mlp": 1.02122998, + "epoch": 0.7877111765766849, + "flos": 18697286194560.0, + "grad_norm": 2.0467288244487416, + "language_loss": 0.79260141, + "learning_rate": 4.5425710346535775e-07, + "loss": 0.81434846, + "num_input_tokens_seen": 141213360, + "step": 6551, + "time_per_iteration": 2.5941073894500732 + }, + { + "auxiliary_loss_clip": 0.01143351, + "auxiliary_loss_mlp": 0.01030789, + "balance_loss_clip": 1.00905001, + "balance_loss_mlp": 1.02330291, + "epoch": 0.787831419467324, + "flos": 27592968833280.0, + "grad_norm": 1.924723638980282, + "language_loss": 0.81763732, + "learning_rate": 4.537629144288877e-07, + "loss": 0.83937871, + "num_input_tokens_seen": 141230815, + "step": 6552, + "time_per_iteration": 2.6969168186187744 + }, + { + "auxiliary_loss_clip": 0.01077343, + "auxiliary_loss_mlp": 0.01028549, + "balance_loss_clip": 0.89211869, + "balance_loss_mlp": 1.02116346, + "epoch": 0.7879516623579631, + "flos": 18150187167360.0, + "grad_norm": 2.180092462648112, + "language_loss": 0.75328094, + "learning_rate": 4.5326895995535477e-07, + "loss": 0.77433985, + "num_input_tokens_seen": 141249715, + "step": 6553, + "time_per_iteration": 2.708195924758911 + }, + { + "auxiliary_loss_clip": 0.01139528, + "auxiliary_loss_mlp": 0.01023527, + "balance_loss_clip": 1.0098803, + "balance_loss_mlp": 1.01591539, + "epoch": 0.7880719052486022, + "flos": 20339193807360.0, + "grad_norm": 2.7338813294960627, + "language_loss": 0.84373164, + "learning_rate": 4.527752401196907e-07, + "loss": 0.86536217, + "num_input_tokens_seen": 141267730, + "step": 6554, + "time_per_iteration": 2.6704115867614746 + }, + { + "auxiliary_loss_clip": 0.0111239, + "auxiliary_loss_mlp": 0.01031509, + "balance_loss_clip": 0.97025543, + "balance_loss_mlp": 1.02346206, + "epoch": 0.7881921481392413, + "flos": 21653237053440.0, + "grad_norm": 2.3485462824574768, + "language_loss": 0.67253423, + "learning_rate": 4.5228175499679254e-07, + "loss": 0.6939733, + "num_input_tokens_seen": 141287315, + "step": 6555, + "time_per_iteration": 2.672635793685913 + }, + { + "auxiliary_loss_clip": 0.010434, + "auxiliary_loss_mlp": 0.01006671, + "balance_loss_clip": 0.97439468, + "balance_loss_mlp": 1.00570571, + "epoch": 0.7883123910298804, + "flos": 68565860058240.0, + "grad_norm": 0.8441246753845575, + "language_loss": 0.54474473, + "learning_rate": 4.5178850466152174e-07, + "loss": 0.56524545, + "num_input_tokens_seen": 141346145, + "step": 6556, + "time_per_iteration": 3.2793667316436768 + }, + { + "auxiliary_loss_clip": 0.01114285, + "auxiliary_loss_mlp": 0.01024086, + "balance_loss_clip": 0.96942997, + "balance_loss_mlp": 1.01673377, + "epoch": 0.7884326339205194, + "flos": 19318217627520.0, + "grad_norm": 1.8122810962643157, + "language_loss": 0.81611401, + "learning_rate": 4.512954891887031e-07, + "loss": 0.83749771, + "num_input_tokens_seen": 141364445, + "step": 6557, + "time_per_iteration": 2.700981616973877 + }, + { + "auxiliary_loss_clip": 0.01114616, + "auxiliary_loss_mlp": 0.01030893, + "balance_loss_clip": 0.9737643, + "balance_loss_mlp": 1.02319169, + "epoch": 0.7885528768111585, + "flos": 17784903807360.0, + "grad_norm": 1.9832136300800995, + "language_loss": 0.83346039, + "learning_rate": 4.5080270865312806e-07, + "loss": 0.8549155, + "num_input_tokens_seen": 141381640, + "step": 6558, + "time_per_iteration": 2.745770215988159 + }, + { + "auxiliary_loss_clip": 0.01143127, + "auxiliary_loss_mlp": 0.01025376, + "balance_loss_clip": 1.01086223, + "balance_loss_mlp": 1.018507, + "epoch": 0.7886731197017977, + "flos": 18807639753600.0, + "grad_norm": 1.7917457160461565, + "language_loss": 0.70920515, + "learning_rate": 4.5031016312954985e-07, + "loss": 0.73089021, + "num_input_tokens_seen": 141399955, + "step": 6559, + "time_per_iteration": 2.681705951690674 + }, + { + "auxiliary_loss_clip": 0.01152992, + "auxiliary_loss_mlp": 0.01031579, + "balance_loss_clip": 1.01479983, + "balance_loss_mlp": 1.02386618, + "epoch": 0.7887933625924367, + "flos": 33365358126720.0, + "grad_norm": 1.9766225285573424, + "language_loss": 0.74206811, + "learning_rate": 4.498178526926886e-07, + "loss": 0.76391387, + "num_input_tokens_seen": 141420820, + "step": 6560, + "time_per_iteration": 2.7770535945892334 + }, + { + "auxiliary_loss_clip": 0.01166175, + "auxiliary_loss_mlp": 0.01032759, + "balance_loss_clip": 1.0508734, + "balance_loss_mlp": 1.02545106, + "epoch": 0.7889136054830758, + "flos": 17019360218880.0, + "grad_norm": 2.047740890418663, + "language_loss": 0.72356671, + "learning_rate": 4.4932577741722635e-07, + "loss": 0.74555606, + "num_input_tokens_seen": 141439350, + "step": 6561, + "time_per_iteration": 2.620349645614624 + }, + { + "auxiliary_loss_clip": 0.01116591, + "auxiliary_loss_mlp": 0.01028679, + "balance_loss_clip": 0.97113472, + "balance_loss_mlp": 1.02050138, + "epoch": 0.7890338483737149, + "flos": 29424629018880.0, + "grad_norm": 1.7749049368124001, + "language_loss": 0.74135852, + "learning_rate": 4.4883393737780985e-07, + "loss": 0.76281124, + "num_input_tokens_seen": 141460300, + "step": 6562, + "time_per_iteration": 2.6972014904022217 + }, + { + "auxiliary_loss_clip": 0.01136824, + "auxiliary_loss_mlp": 0.01029698, + "balance_loss_clip": 1.00817847, + "balance_loss_mlp": 1.02255404, + "epoch": 0.789154091264354, + "flos": 19971576063360.0, + "grad_norm": 1.811094063340143, + "language_loss": 0.78479636, + "learning_rate": 4.4834233264905254e-07, + "loss": 0.80646157, + "num_input_tokens_seen": 141477315, + "step": 6563, + "time_per_iteration": 3.6178159713745117 + }, + { + "auxiliary_loss_clip": 0.0108878, + "auxiliary_loss_mlp": 0.0103274, + "balance_loss_clip": 0.93242157, + "balance_loss_mlp": 1.02496719, + "epoch": 0.789274334154993, + "flos": 14537825216640.0, + "grad_norm": 2.72270682668415, + "language_loss": 0.71641886, + "learning_rate": 4.478509633055294e-07, + "loss": 0.73763406, + "num_input_tokens_seen": 141495025, + "step": 6564, + "time_per_iteration": 3.6613054275512695 + }, + { + "auxiliary_loss_clip": 0.01123116, + "auxiliary_loss_mlp": 0.01023425, + "balance_loss_clip": 0.97270101, + "balance_loss_mlp": 1.01506233, + "epoch": 0.7893945770456322, + "flos": 21827403123840.0, + "grad_norm": 2.956868732661735, + "language_loss": 0.7997309, + "learning_rate": 4.473598294217813e-07, + "loss": 0.82119632, + "num_input_tokens_seen": 141510450, + "step": 6565, + "time_per_iteration": 2.6999223232269287 + }, + { + "auxiliary_loss_clip": 0.01139795, + "auxiliary_loss_mlp": 0.01027397, + "balance_loss_clip": 1.01164556, + "balance_loss_mlp": 1.02020824, + "epoch": 0.7895148199362713, + "flos": 20740639184640.0, + "grad_norm": 1.9467187799237713, + "language_loss": 0.71533734, + "learning_rate": 4.468689310723124e-07, + "loss": 0.73700929, + "num_input_tokens_seen": 141528265, + "step": 6566, + "time_per_iteration": 3.5750362873077393 + }, + { + "auxiliary_loss_clip": 0.01102275, + "auxiliary_loss_mlp": 0.0102489, + "balance_loss_clip": 0.93463111, + "balance_loss_mlp": 1.01769614, + "epoch": 0.7896350628269103, + "flos": 16690669839360.0, + "grad_norm": 1.8294387586207506, + "language_loss": 0.7822969, + "learning_rate": 4.463782683315913e-07, + "loss": 0.8035686, + "num_input_tokens_seen": 141547270, + "step": 6567, + "time_per_iteration": 2.703831434249878 + }, + { + "auxiliary_loss_clip": 0.01164872, + "auxiliary_loss_mlp": 0.0102216, + "balance_loss_clip": 1.04934955, + "balance_loss_mlp": 1.01505494, + "epoch": 0.7897553057175495, + "flos": 22638374438400.0, + "grad_norm": 1.6729675111463147, + "language_loss": 0.73468679, + "learning_rate": 4.458878412740523e-07, + "loss": 0.75655711, + "num_input_tokens_seen": 141566050, + "step": 6568, + "time_per_iteration": 2.596609354019165 + }, + { + "auxiliary_loss_clip": 0.01140633, + "auxiliary_loss_mlp": 0.01022995, + "balance_loss_clip": 1.01074815, + "balance_loss_mlp": 1.01557374, + "epoch": 0.7898755486081885, + "flos": 14537573821440.0, + "grad_norm": 2.899988252781461, + "language_loss": 0.78058136, + "learning_rate": 4.453976499740919e-07, + "loss": 0.8022176, + "num_input_tokens_seen": 141583695, + "step": 6569, + "time_per_iteration": 2.6652798652648926 + }, + { + "auxiliary_loss_clip": 0.01140461, + "auxiliary_loss_mlp": 0.01023829, + "balance_loss_clip": 1.01190281, + "balance_loss_mlp": 1.01675963, + "epoch": 0.7899957914988276, + "flos": 17238487138560.0, + "grad_norm": 1.7120551220330622, + "language_loss": 0.77842838, + "learning_rate": 4.4490769450607215e-07, + "loss": 0.80007136, + "num_input_tokens_seen": 141601320, + "step": 6570, + "time_per_iteration": 2.617518186569214 + }, + { + "auxiliary_loss_clip": 0.01091907, + "auxiliary_loss_mlp": 0.01027165, + "balance_loss_clip": 0.92973989, + "balance_loss_mlp": 1.01923108, + "epoch": 0.7901160343894668, + "flos": 41279351086080.0, + "grad_norm": 2.1987645340901785, + "language_loss": 0.72482562, + "learning_rate": 4.4441797494431845e-07, + "loss": 0.74601632, + "num_input_tokens_seen": 141623125, + "step": 6571, + "time_per_iteration": 2.950489044189453 + }, + { + "auxiliary_loss_clip": 0.01140579, + "auxiliary_loss_mlp": 0.01023552, + "balance_loss_clip": 1.0112083, + "balance_loss_mlp": 1.01648259, + "epoch": 0.7902362772801058, + "flos": 16837005847680.0, + "grad_norm": 5.792952300561198, + "language_loss": 0.77618349, + "learning_rate": 4.439284913631207e-07, + "loss": 0.79782486, + "num_input_tokens_seen": 141640335, + "step": 6572, + "time_per_iteration": 2.610690116882324 + }, + { + "auxiliary_loss_clip": 0.01100235, + "auxiliary_loss_mlp": 0.01026815, + "balance_loss_clip": 0.93726718, + "balance_loss_mlp": 1.01944184, + "epoch": 0.7903565201707449, + "flos": 27125987091840.0, + "grad_norm": 1.9954254838550483, + "language_loss": 0.83631098, + "learning_rate": 4.434392438367347e-07, + "loss": 0.8575815, + "num_input_tokens_seen": 141659760, + "step": 6573, + "time_per_iteration": 3.6093909740448 + }, + { + "auxiliary_loss_clip": 0.01147806, + "auxiliary_loss_mlp": 0.01021488, + "balance_loss_clip": 1.01094794, + "balance_loss_mlp": 1.01377249, + "epoch": 0.790476763061384, + "flos": 31025167142400.0, + "grad_norm": 1.8631816574453617, + "language_loss": 0.74337351, + "learning_rate": 4.4295023243937677e-07, + "loss": 0.76506644, + "num_input_tokens_seen": 141679965, + "step": 6574, + "time_per_iteration": 2.640293836593628 + }, + { + "auxiliary_loss_clip": 0.01146925, + "auxiliary_loss_mlp": 0.01028703, + "balance_loss_clip": 1.01499498, + "balance_loss_mlp": 1.02048898, + "epoch": 0.7905970059520231, + "flos": 22089084681600.0, + "grad_norm": 1.8304082571560791, + "language_loss": 0.8018918, + "learning_rate": 4.4246145724523123e-07, + "loss": 0.82364804, + "num_input_tokens_seen": 141697710, + "step": 6575, + "time_per_iteration": 2.75060772895813 + }, + { + "auxiliary_loss_clip": 0.01095811, + "auxiliary_loss_mlp": 0.01028055, + "balance_loss_clip": 0.93475431, + "balance_loss_mlp": 1.02061343, + "epoch": 0.7907172488426621, + "flos": 20558141159040.0, + "grad_norm": 2.7789625338528783, + "language_loss": 0.77584809, + "learning_rate": 4.41972918328444e-07, + "loss": 0.79708678, + "num_input_tokens_seen": 141715145, + "step": 6576, + "time_per_iteration": 2.785721778869629 + }, + { + "auxiliary_loss_clip": 0.01140037, + "auxiliary_loss_mlp": 0.01021456, + "balance_loss_clip": 1.01159346, + "balance_loss_mlp": 1.01429772, + "epoch": 0.7908374917333013, + "flos": 30081542901120.0, + "grad_norm": 2.0563084032162036, + "language_loss": 0.77587187, + "learning_rate": 4.4148461576312646e-07, + "loss": 0.79748678, + "num_input_tokens_seen": 141734810, + "step": 6577, + "time_per_iteration": 2.6924548149108887 + }, + { + "auxiliary_loss_clip": 0.0114537, + "auxiliary_loss_mlp": 0.01025503, + "balance_loss_clip": 1.01374364, + "balance_loss_mlp": 1.01863074, + "epoch": 0.7909577346239404, + "flos": 20996359084800.0, + "grad_norm": 1.5477737386610846, + "language_loss": 0.74370331, + "learning_rate": 4.4099654962335343e-07, + "loss": 0.76541203, + "num_input_tokens_seen": 141755260, + "step": 6578, + "time_per_iteration": 2.7189042568206787 + }, + { + "auxiliary_loss_clip": 0.01125882, + "auxiliary_loss_mlp": 0.01024847, + "balance_loss_clip": 0.97326839, + "balance_loss_mlp": 1.01790881, + "epoch": 0.7910779775145794, + "flos": 26247935128320.0, + "grad_norm": 2.318507457275571, + "language_loss": 0.7517066, + "learning_rate": 4.405087199831636e-07, + "loss": 0.77321386, + "num_input_tokens_seen": 141775500, + "step": 6579, + "time_per_iteration": 2.691427230834961 + }, + { + "auxiliary_loss_clip": 0.01121904, + "auxiliary_loss_mlp": 0.0050908, + "balance_loss_clip": 0.97128683, + "balance_loss_mlp": 0.0, + "epoch": 0.7911982204052186, + "flos": 22564434291840.0, + "grad_norm": 2.0812889442838234, + "language_loss": 0.67375106, + "learning_rate": 4.400211269165619e-07, + "loss": 0.69006085, + "num_input_tokens_seen": 141791955, + "step": 6580, + "time_per_iteration": 2.7208166122436523 + }, + { + "auxiliary_loss_clip": 0.01168321, + "auxiliary_loss_mlp": 0.01028428, + "balance_loss_clip": 1.05303597, + "balance_loss_mlp": 1.02144265, + "epoch": 0.7913184632958576, + "flos": 23112538899840.0, + "grad_norm": 1.4826545410338376, + "language_loss": 0.76651406, + "learning_rate": 4.3953377049751416e-07, + "loss": 0.78848153, + "num_input_tokens_seen": 141812380, + "step": 6581, + "time_per_iteration": 2.680013418197632 + }, + { + "auxiliary_loss_clip": 0.01123525, + "auxiliary_loss_mlp": 0.01025972, + "balance_loss_clip": 0.9735136, + "balance_loss_mlp": 1.01907539, + "epoch": 0.7914387061864967, + "flos": 12311758719360.0, + "grad_norm": 2.2199641796064524, + "language_loss": 0.77844775, + "learning_rate": 4.390466507999537e-07, + "loss": 0.79994273, + "num_input_tokens_seen": 141828130, + "step": 6582, + "time_per_iteration": 2.716188669204712 + }, + { + "auxiliary_loss_clip": 0.01095287, + "auxiliary_loss_mlp": 0.0102937, + "balance_loss_clip": 0.93362099, + "balance_loss_mlp": 1.02196097, + "epoch": 0.7915589490771359, + "flos": 17603267708160.0, + "grad_norm": 2.3485118418562454, + "language_loss": 0.76674366, + "learning_rate": 4.385597678977748e-07, + "loss": 0.78799021, + "num_input_tokens_seen": 141846965, + "step": 6583, + "time_per_iteration": 2.6838181018829346 + }, + { + "auxiliary_loss_clip": 0.01120831, + "auxiliary_loss_mlp": 0.0102314, + "balance_loss_clip": 0.9740386, + "balance_loss_mlp": 1.01543355, + "epoch": 0.7916791919677749, + "flos": 25591272641280.0, + "grad_norm": 1.7659282950140438, + "language_loss": 0.7527265, + "learning_rate": 4.3807312186483726e-07, + "loss": 0.77416623, + "num_input_tokens_seen": 141867685, + "step": 6584, + "time_per_iteration": 2.7356925010681152 + }, + { + "auxiliary_loss_clip": 0.01139688, + "auxiliary_loss_mlp": 0.01023966, + "balance_loss_clip": 1.01325107, + "balance_loss_mlp": 1.01721537, + "epoch": 0.791799434858414, + "flos": 18844340474880.0, + "grad_norm": 2.0506286717138837, + "language_loss": 0.7820918, + "learning_rate": 4.375867127749655e-07, + "loss": 0.80372834, + "num_input_tokens_seen": 141885960, + "step": 6585, + "time_per_iteration": 2.636225461959839 + }, + { + "auxiliary_loss_clip": 0.01100116, + "auxiliary_loss_mlp": 0.01027004, + "balance_loss_clip": 0.93777072, + "balance_loss_mlp": 1.01992941, + "epoch": 0.7919196777490531, + "flos": 25812015672960.0, + "grad_norm": 2.1501386340181283, + "language_loss": 0.6735099, + "learning_rate": 4.3710054070194744e-07, + "loss": 0.69478106, + "num_input_tokens_seen": 141905655, + "step": 6586, + "time_per_iteration": 2.768397331237793 + }, + { + "auxiliary_loss_clip": 0.01168101, + "auxiliary_loss_mlp": 0.00509433, + "balance_loss_clip": 1.04875171, + "balance_loss_mlp": 0.0, + "epoch": 0.7920399206396922, + "flos": 11947624594560.0, + "grad_norm": 2.748386102847964, + "language_loss": 0.66485023, + "learning_rate": 4.3661460571953455e-07, + "loss": 0.6816256, + "num_input_tokens_seen": 141922390, + "step": 6587, + "time_per_iteration": 2.648672103881836 + }, + { + "auxiliary_loss_clip": 0.01142082, + "auxiliary_loss_mlp": 0.01024591, + "balance_loss_clip": 1.0081445, + "balance_loss_mlp": 1.01769733, + "epoch": 0.7921601635303313, + "flos": 21579907438080.0, + "grad_norm": 2.8125259887747274, + "language_loss": 0.68487883, + "learning_rate": 4.36128907901443e-07, + "loss": 0.70654553, + "num_input_tokens_seen": 141941985, + "step": 6588, + "time_per_iteration": 3.6909992694854736 + }, + { + "auxiliary_loss_clip": 0.01098392, + "auxiliary_loss_mlp": 0.01022433, + "balance_loss_clip": 0.9335196, + "balance_loss_mlp": 1.01501787, + "epoch": 0.7922804064209703, + "flos": 18113989236480.0, + "grad_norm": 2.313983545067407, + "language_loss": 0.72456825, + "learning_rate": 4.356434473213519e-07, + "loss": 0.74577647, + "num_input_tokens_seen": 141959435, + "step": 6589, + "time_per_iteration": 2.734682321548462 + }, + { + "auxiliary_loss_clip": 0.0112025, + "auxiliary_loss_mlp": 0.01029817, + "balance_loss_clip": 0.97497928, + "balance_loss_mlp": 1.02214026, + "epoch": 0.7924006493116095, + "flos": 21652806090240.0, + "grad_norm": 1.6694919030372575, + "language_loss": 0.79630923, + "learning_rate": 4.351582240529068e-07, + "loss": 0.81780994, + "num_input_tokens_seen": 141980265, + "step": 6590, + "time_per_iteration": 3.61253023147583 + }, + { + "auxiliary_loss_clip": 0.01027166, + "auxiliary_loss_mlp": 0.01003426, + "balance_loss_clip": 0.93732023, + "balance_loss_mlp": 1.00237072, + "epoch": 0.7925208922022485, + "flos": 64242755694720.0, + "grad_norm": 0.7331109980087861, + "language_loss": 0.58176458, + "learning_rate": 4.346732381697149e-07, + "loss": 0.60207045, + "num_input_tokens_seen": 142044395, + "step": 6591, + "time_per_iteration": 3.36016845703125 + }, + { + "auxiliary_loss_clip": 0.01116147, + "auxiliary_loss_mlp": 0.01029647, + "balance_loss_clip": 0.97471893, + "balance_loss_mlp": 1.02205968, + "epoch": 0.7926411350928876, + "flos": 16941541403520.0, + "grad_norm": 1.8282101893526588, + "language_loss": 0.81495631, + "learning_rate": 4.3418848974534825e-07, + "loss": 0.83641428, + "num_input_tokens_seen": 142061335, + "step": 6592, + "time_per_iteration": 3.6636407375335693 + }, + { + "auxiliary_loss_clip": 0.01102373, + "auxiliary_loss_mlp": 0.01027729, + "balance_loss_clip": 0.93570179, + "balance_loss_mlp": 1.02091932, + "epoch": 0.7927613779835267, + "flos": 34459987144320.0, + "grad_norm": 1.92708032379959, + "language_loss": 0.68658531, + "learning_rate": 4.3370397885334276e-07, + "loss": 0.70788634, + "num_input_tokens_seen": 142081965, + "step": 6593, + "time_per_iteration": 2.8068790435791016 + }, + { + "auxiliary_loss_clip": 0.01138165, + "auxiliary_loss_mlp": 0.01023588, + "balance_loss_clip": 1.01091123, + "balance_loss_mlp": 1.01651239, + "epoch": 0.7928816208741658, + "flos": 18951174501120.0, + "grad_norm": 2.744665718924535, + "language_loss": 0.75446957, + "learning_rate": 4.3321970556719777e-07, + "loss": 0.77608705, + "num_input_tokens_seen": 142100260, + "step": 6594, + "time_per_iteration": 2.6765007972717285 + }, + { + "auxiliary_loss_clip": 0.01167243, + "auxiliary_loss_mlp": 0.0102932, + "balance_loss_clip": 1.05045605, + "balance_loss_mlp": 1.02179742, + "epoch": 0.7930018637648049, + "flos": 18623022825600.0, + "grad_norm": 4.827775377568693, + "language_loss": 0.72239828, + "learning_rate": 4.3273566996037856e-07, + "loss": 0.7443639, + "num_input_tokens_seen": 142116955, + "step": 6595, + "time_per_iteration": 2.6053552627563477 + }, + { + "auxiliary_loss_clip": 0.011191, + "auxiliary_loss_mlp": 0.01022241, + "balance_loss_clip": 0.97217333, + "balance_loss_mlp": 1.01552916, + "epoch": 0.793122106655444, + "flos": 24530650824960.0, + "grad_norm": 2.0619605275250636, + "language_loss": 0.80451161, + "learning_rate": 4.322518721063113e-07, + "loss": 0.82592499, + "num_input_tokens_seen": 142135505, + "step": 6596, + "time_per_iteration": 2.7272980213165283 + }, + { + "auxiliary_loss_clip": 0.01143013, + "auxiliary_loss_mlp": 0.01027075, + "balance_loss_clip": 1.0127213, + "balance_loss_mlp": 1.01962137, + "epoch": 0.7932423495460831, + "flos": 34421203434240.0, + "grad_norm": 2.1529322675587306, + "language_loss": 0.70275986, + "learning_rate": 4.3176831207838906e-07, + "loss": 0.72446078, + "num_input_tokens_seen": 142158915, + "step": 6597, + "time_per_iteration": 2.7777085304260254 + }, + { + "auxiliary_loss_clip": 0.0114257, + "auxiliary_loss_mlp": 0.01029691, + "balance_loss_clip": 1.01373887, + "balance_loss_mlp": 1.02241886, + "epoch": 0.7933625924367221, + "flos": 26980333441920.0, + "grad_norm": 1.761194519701267, + "language_loss": 0.7434656, + "learning_rate": 4.3128498994996685e-07, + "loss": 0.76518822, + "num_input_tokens_seen": 142178390, + "step": 6598, + "time_per_iteration": 2.6928281784057617 + }, + { + "auxiliary_loss_clip": 0.01148049, + "auxiliary_loss_mlp": 0.01032594, + "balance_loss_clip": 1.01264596, + "balance_loss_mlp": 1.02482808, + "epoch": 0.7934828353273613, + "flos": 29568630643200.0, + "grad_norm": 3.5153720664114134, + "language_loss": 0.71236753, + "learning_rate": 4.308019057943646e-07, + "loss": 0.73417395, + "num_input_tokens_seen": 142200115, + "step": 6599, + "time_per_iteration": 3.5965800285339355 + }, + { + "auxiliary_loss_clip": 0.01080643, + "auxiliary_loss_mlp": 0.01023774, + "balance_loss_clip": 0.89655054, + "balance_loss_mlp": 1.01637089, + "epoch": 0.7936030782180004, + "flos": 28615381557120.0, + "grad_norm": 1.6175584065759194, + "language_loss": 0.74742454, + "learning_rate": 4.3031905968486535e-07, + "loss": 0.76846874, + "num_input_tokens_seen": 142220945, + "step": 6600, + "time_per_iteration": 2.765514373779297 + }, + { + "auxiliary_loss_clip": 0.01075851, + "auxiliary_loss_mlp": 0.01026653, + "balance_loss_clip": 0.9018814, + "balance_loss_mlp": 1.01904154, + "epoch": 0.7937233211086394, + "flos": 16392574869120.0, + "grad_norm": 2.2164156064884124, + "language_loss": 0.68228662, + "learning_rate": 4.298364516947162e-07, + "loss": 0.70331168, + "num_input_tokens_seen": 142238175, + "step": 6601, + "time_per_iteration": 2.7693216800689697 + }, + { + "auxiliary_loss_clip": 0.01069961, + "auxiliary_loss_mlp": 0.01024402, + "balance_loss_clip": 0.8944, + "balance_loss_mlp": 1.01727605, + "epoch": 0.7938435639992786, + "flos": 22013420682240.0, + "grad_norm": 2.0463905864317065, + "language_loss": 0.65286034, + "learning_rate": 4.293540818971295e-07, + "loss": 0.67380393, + "num_input_tokens_seen": 142255980, + "step": 6602, + "time_per_iteration": 2.8657729625701904 + }, + { + "auxiliary_loss_clip": 0.01148957, + "auxiliary_loss_mlp": 0.0102945, + "balance_loss_clip": 1.01173496, + "balance_loss_mlp": 1.02196419, + "epoch": 0.7939638068899176, + "flos": 22197032029440.0, + "grad_norm": 2.0640040413162923, + "language_loss": 0.76562738, + "learning_rate": 4.2887195036527934e-07, + "loss": 0.78741145, + "num_input_tokens_seen": 142274785, + "step": 6603, + "time_per_iteration": 2.693875312805176 + }, + { + "auxiliary_loss_clip": 0.011343, + "auxiliary_loss_mlp": 0.01031791, + "balance_loss_clip": 1.00631201, + "balance_loss_mlp": 1.02408445, + "epoch": 0.7940840497805567, + "flos": 17745186343680.0, + "grad_norm": 2.284446056587339, + "language_loss": 0.73503846, + "learning_rate": 4.28390057172306e-07, + "loss": 0.75669932, + "num_input_tokens_seen": 142291290, + "step": 6604, + "time_per_iteration": 2.6374447345733643 + }, + { + "auxiliary_loss_clip": 0.01090743, + "auxiliary_loss_mlp": 0.01029542, + "balance_loss_clip": 0.93068528, + "balance_loss_mlp": 1.02127504, + "epoch": 0.7942042926711959, + "flos": 23805435231360.0, + "grad_norm": 2.198496869699491, + "language_loss": 0.72348452, + "learning_rate": 4.279084023913111e-07, + "loss": 0.74468744, + "num_input_tokens_seen": 142309165, + "step": 6605, + "time_per_iteration": 2.7099335193634033 + }, + { + "auxiliary_loss_clip": 0.01144857, + "auxiliary_loss_mlp": 0.0102616, + "balance_loss_clip": 1.01321208, + "balance_loss_mlp": 1.01914406, + "epoch": 0.7943245355618349, + "flos": 19244959839360.0, + "grad_norm": 2.0226663673529672, + "language_loss": 0.69243836, + "learning_rate": 4.2742698609536096e-07, + "loss": 0.71414852, + "num_input_tokens_seen": 142327475, + "step": 6606, + "time_per_iteration": 2.6564407348632812 + }, + { + "auxiliary_loss_clip": 0.01123721, + "auxiliary_loss_mlp": 0.01020838, + "balance_loss_clip": 0.97444391, + "balance_loss_mlp": 1.01324451, + "epoch": 0.794444778452474, + "flos": 25007616547200.0, + "grad_norm": 2.0811028956151723, + "language_loss": 0.78488421, + "learning_rate": 4.2694580835748706e-07, + "loss": 0.80632985, + "num_input_tokens_seen": 142347335, + "step": 6607, + "time_per_iteration": 2.7202885150909424 + }, + { + "auxiliary_loss_clip": 0.01119366, + "auxiliary_loss_mlp": 0.0102378, + "balance_loss_clip": 0.9724775, + "balance_loss_mlp": 1.01647234, + "epoch": 0.7945650213431131, + "flos": 23221491828480.0, + "grad_norm": 1.897736799676816, + "language_loss": 0.74261224, + "learning_rate": 4.264648692506836e-07, + "loss": 0.76404369, + "num_input_tokens_seen": 142366125, + "step": 6608, + "time_per_iteration": 2.6866016387939453 + }, + { + "auxiliary_loss_clip": 0.01115306, + "auxiliary_loss_mlp": 0.01029315, + "balance_loss_clip": 0.97111088, + "balance_loss_mlp": 1.02191854, + "epoch": 0.7946852642337522, + "flos": 26062887237120.0, + "grad_norm": 1.7518543361171939, + "language_loss": 0.72328675, + "learning_rate": 4.2598416884790824e-07, + "loss": 0.74473298, + "num_input_tokens_seen": 142385175, + "step": 6609, + "time_per_iteration": 2.7058675289154053 + }, + { + "auxiliary_loss_clip": 0.01128766, + "auxiliary_loss_mlp": 0.01027825, + "balance_loss_clip": 0.97223163, + "balance_loss_mlp": 1.02007675, + "epoch": 0.7948055071243912, + "flos": 23769704177280.0, + "grad_norm": 1.8107401872319386, + "language_loss": 0.80594659, + "learning_rate": 4.255037072220828e-07, + "loss": 0.82751238, + "num_input_tokens_seen": 142406545, + "step": 6610, + "time_per_iteration": 2.734405994415283 + }, + { + "auxiliary_loss_clip": 0.01163992, + "auxiliary_loss_mlp": 0.01024569, + "balance_loss_clip": 1.04782987, + "balance_loss_mlp": 1.01757741, + "epoch": 0.7949257500150304, + "flos": 21980814111360.0, + "grad_norm": 1.531306971808141, + "language_loss": 0.71803772, + "learning_rate": 4.2502348444609293e-07, + "loss": 0.73992336, + "num_input_tokens_seen": 142426165, + "step": 6611, + "time_per_iteration": 2.625368118286133 + }, + { + "auxiliary_loss_clip": 0.01069963, + "auxiliary_loss_mlp": 0.01025089, + "balance_loss_clip": 0.89224595, + "balance_loss_mlp": 1.01751316, + "epoch": 0.7950459929056695, + "flos": 25774129802880.0, + "grad_norm": 1.7024895697658158, + "language_loss": 0.69450021, + "learning_rate": 4.2454350059278844e-07, + "loss": 0.71545076, + "num_input_tokens_seen": 142447225, + "step": 6612, + "time_per_iteration": 2.7655749320983887 + }, + { + "auxiliary_loss_clip": 0.01112677, + "auxiliary_loss_mlp": 0.01023519, + "balance_loss_clip": 0.9668076, + "balance_loss_mlp": 1.01636696, + "epoch": 0.7951662357963085, + "flos": 22158068751360.0, + "grad_norm": 1.9869005998279696, + "language_loss": 0.8470428, + "learning_rate": 4.240637557349824e-07, + "loss": 0.86840475, + "num_input_tokens_seen": 142464440, + "step": 6613, + "time_per_iteration": 2.69301438331604 + }, + { + "auxiliary_loss_clip": 0.01107904, + "auxiliary_loss_mlp": 0.01026239, + "balance_loss_clip": 0.97008359, + "balance_loss_mlp": 1.01876497, + "epoch": 0.7952864786869477, + "flos": 24641938137600.0, + "grad_norm": 1.9599919245768236, + "language_loss": 0.66615021, + "learning_rate": 4.235842499454516e-07, + "loss": 0.6874916, + "num_input_tokens_seen": 142484355, + "step": 6614, + "time_per_iteration": 3.6841440200805664 + }, + { + "auxiliary_loss_clip": 0.0112092, + "auxiliary_loss_mlp": 0.01030122, + "balance_loss_clip": 0.97370529, + "balance_loss_mlp": 1.02289534, + "epoch": 0.7954067215775867, + "flos": 21830922656640.0, + "grad_norm": 1.6261965769397364, + "language_loss": 0.82584822, + "learning_rate": 4.2310498329693687e-07, + "loss": 0.84735858, + "num_input_tokens_seen": 142505255, + "step": 6615, + "time_per_iteration": 2.7086780071258545 + }, + { + "auxiliary_loss_clip": 0.01147841, + "auxiliary_loss_mlp": 0.01029377, + "balance_loss_clip": 1.01267326, + "balance_loss_mlp": 1.02172923, + "epoch": 0.7955269644682258, + "flos": 24060652341120.0, + "grad_norm": 1.7593081668263644, + "language_loss": 0.81069374, + "learning_rate": 4.2262595586214164e-07, + "loss": 0.83246595, + "num_input_tokens_seen": 142526350, + "step": 6616, + "time_per_iteration": 3.55609130859375 + }, + { + "auxiliary_loss_clip": 0.01146897, + "auxiliary_loss_mlp": 0.01024009, + "balance_loss_clip": 1.01111794, + "balance_loss_mlp": 1.01672554, + "epoch": 0.795647207358865, + "flos": 25010741030400.0, + "grad_norm": 1.61405504988325, + "language_loss": 0.76594537, + "learning_rate": 4.221471677137358e-07, + "loss": 0.7876544, + "num_input_tokens_seen": 142547165, + "step": 6617, + "time_per_iteration": 2.767138719558716 + }, + { + "auxiliary_loss_clip": 0.01109085, + "auxiliary_loss_mlp": 0.01025127, + "balance_loss_clip": 0.97121853, + "balance_loss_mlp": 1.0182662, + "epoch": 0.795767450249504, + "flos": 14648358343680.0, + "grad_norm": 1.5750883135915958, + "language_loss": 0.70214081, + "learning_rate": 4.216686189243492e-07, + "loss": 0.72348291, + "num_input_tokens_seen": 142565955, + "step": 6618, + "time_per_iteration": 3.6270055770874023 + }, + { + "auxiliary_loss_clip": 0.01095302, + "auxiliary_loss_mlp": 0.01026076, + "balance_loss_clip": 0.9362731, + "balance_loss_mlp": 1.01850069, + "epoch": 0.7958876931401431, + "flos": 18547897530240.0, + "grad_norm": 1.7148166230032282, + "language_loss": 0.72966522, + "learning_rate": 4.211903095665785e-07, + "loss": 0.75087905, + "num_input_tokens_seen": 142585340, + "step": 6619, + "time_per_iteration": 2.7134523391723633 + }, + { + "auxiliary_loss_clip": 0.01135164, + "auxiliary_loss_mlp": 0.01022048, + "balance_loss_clip": 1.00766063, + "balance_loss_mlp": 1.01456118, + "epoch": 0.7960079360307821, + "flos": 21543960902400.0, + "grad_norm": 2.105662547652821, + "language_loss": 0.75157923, + "learning_rate": 4.2071223971298277e-07, + "loss": 0.7731514, + "num_input_tokens_seen": 142602525, + "step": 6620, + "time_per_iteration": 2.6543684005737305 + }, + { + "auxiliary_loss_clip": 0.01143383, + "auxiliary_loss_mlp": 0.01028202, + "balance_loss_clip": 1.00959206, + "balance_loss_mlp": 1.02070355, + "epoch": 0.7961281789214213, + "flos": 25481745095040.0, + "grad_norm": 1.8133923077554872, + "language_loss": 0.61337906, + "learning_rate": 4.2023440943608433e-07, + "loss": 0.63509494, + "num_input_tokens_seen": 142622490, + "step": 6621, + "time_per_iteration": 2.6451828479766846 + }, + { + "auxiliary_loss_clip": 0.01141847, + "auxiliary_loss_mlp": 0.01027656, + "balance_loss_clip": 1.00838256, + "balance_loss_mlp": 1.02067065, + "epoch": 0.7962484218120603, + "flos": 21944436612480.0, + "grad_norm": 1.633044184014735, + "language_loss": 0.78108013, + "learning_rate": 4.1975681880837023e-07, + "loss": 0.8027752, + "num_input_tokens_seen": 142642495, + "step": 6622, + "time_per_iteration": 2.6620259284973145 + }, + { + "auxiliary_loss_clip": 0.01090832, + "auxiliary_loss_mlp": 0.01023233, + "balance_loss_clip": 0.93042004, + "balance_loss_mlp": 1.01613975, + "epoch": 0.7963686647026994, + "flos": 18876264687360.0, + "grad_norm": 1.8751886644033526, + "language_loss": 0.82365727, + "learning_rate": 4.192794679022895e-07, + "loss": 0.84479791, + "num_input_tokens_seen": 142660820, + "step": 6623, + "time_per_iteration": 2.6500861644744873 + }, + { + "auxiliary_loss_clip": 0.01142034, + "auxiliary_loss_mlp": 0.01024453, + "balance_loss_clip": 1.00850391, + "balance_loss_mlp": 1.01712155, + "epoch": 0.7964889075933386, + "flos": 29716582763520.0, + "grad_norm": 1.9100552090234277, + "language_loss": 0.72064376, + "learning_rate": 4.1880235679025743e-07, + "loss": 0.74230862, + "num_input_tokens_seen": 142680915, + "step": 6624, + "time_per_iteration": 3.6293559074401855 + }, + { + "auxiliary_loss_clip": 0.0105211, + "auxiliary_loss_mlp": 0.01029059, + "balance_loss_clip": 0.85738689, + "balance_loss_mlp": 1.02148318, + "epoch": 0.7966091504839776, + "flos": 29491458272640.0, + "grad_norm": 1.804395979752274, + "language_loss": 0.63604432, + "learning_rate": 4.1832548554464986e-07, + "loss": 0.656856, + "num_input_tokens_seen": 142699210, + "step": 6625, + "time_per_iteration": 2.8560047149658203 + }, + { + "auxiliary_loss_clip": 0.01037358, + "auxiliary_loss_mlp": 0.00997096, + "balance_loss_clip": 0.97264451, + "balance_loss_mlp": 0.99599928, + "epoch": 0.7967293933746167, + "flos": 67288697101440.0, + "grad_norm": 0.7457551451940891, + "language_loss": 0.58670336, + "learning_rate": 4.178488542378098e-07, + "loss": 0.60704798, + "num_input_tokens_seen": 142756790, + "step": 6626, + "time_per_iteration": 3.108901023864746 + }, + { + "auxiliary_loss_clip": 0.01167994, + "auxiliary_loss_mlp": 0.01029503, + "balance_loss_clip": 1.04863286, + "balance_loss_mlp": 1.02159929, + "epoch": 0.7968496362652558, + "flos": 25554679660800.0, + "grad_norm": 1.7747241999908516, + "language_loss": 0.89047527, + "learning_rate": 4.173724629420401e-07, + "loss": 0.91245019, + "num_input_tokens_seen": 142778150, + "step": 6627, + "time_per_iteration": 2.673234462738037 + }, + { + "auxiliary_loss_clip": 0.01121668, + "auxiliary_loss_mlp": 0.01023807, + "balance_loss_clip": 0.97159034, + "balance_loss_mlp": 1.01590347, + "epoch": 0.7969698791558949, + "flos": 14501088581760.0, + "grad_norm": 2.2913820470193054, + "language_loss": 0.68279755, + "learning_rate": 4.168963117296087e-07, + "loss": 0.70425224, + "num_input_tokens_seen": 142795485, + "step": 6628, + "time_per_iteration": 2.616100788116455 + }, + { + "auxiliary_loss_clip": 0.01167793, + "auxiliary_loss_mlp": 0.01026494, + "balance_loss_clip": 1.05081975, + "balance_loss_mlp": 1.01896322, + "epoch": 0.797090122046534, + "flos": 22127545169280.0, + "grad_norm": 2.3511989421304147, + "language_loss": 0.76735449, + "learning_rate": 4.1642040067274876e-07, + "loss": 0.78929734, + "num_input_tokens_seen": 142815155, + "step": 6629, + "time_per_iteration": 2.6550683975219727 + }, + { + "auxiliary_loss_clip": 0.01123554, + "auxiliary_loss_mlp": 0.01025868, + "balance_loss_clip": 0.97235012, + "balance_loss_mlp": 1.01867938, + "epoch": 0.7972103649371731, + "flos": 19897671830400.0, + "grad_norm": 1.6140237028772884, + "language_loss": 0.72725695, + "learning_rate": 4.1594472984365493e-07, + "loss": 0.74875116, + "num_input_tokens_seen": 142833840, + "step": 6630, + "time_per_iteration": 2.6632745265960693 + }, + { + "auxiliary_loss_clip": 0.0114045, + "auxiliary_loss_mlp": 0.01025113, + "balance_loss_clip": 1.01210856, + "balance_loss_mlp": 1.01791835, + "epoch": 0.7973306078278122, + "flos": 36058621847040.0, + "grad_norm": 1.8125093076205756, + "language_loss": 0.7808969, + "learning_rate": 4.154692993144862e-07, + "loss": 0.80255258, + "num_input_tokens_seen": 142853610, + "step": 6631, + "time_per_iteration": 2.7430663108825684 + }, + { + "auxiliary_loss_clip": 0.01165322, + "auxiliary_loss_mlp": 0.00509334, + "balance_loss_clip": 1.04864025, + "balance_loss_mlp": 0.0, + "epoch": 0.7974508507184512, + "flos": 21360600950400.0, + "grad_norm": 2.0340828782904548, + "language_loss": 0.71565342, + "learning_rate": 4.1499410915736476e-07, + "loss": 0.73239994, + "num_input_tokens_seen": 142872540, + "step": 6632, + "time_per_iteration": 2.61312198638916 + }, + { + "auxiliary_loss_clip": 0.0104249, + "auxiliary_loss_mlp": 0.01002218, + "balance_loss_clip": 0.973252, + "balance_loss_mlp": 1.00121677, + "epoch": 0.7975710936090904, + "flos": 68253115317120.0, + "grad_norm": 1.3949925709059185, + "language_loss": 0.64227116, + "learning_rate": 4.145191594443762e-07, + "loss": 0.6627183, + "num_input_tokens_seen": 142936895, + "step": 6633, + "time_per_iteration": 3.353013277053833 + }, + { + "auxiliary_loss_clip": 0.01091395, + "auxiliary_loss_mlp": 0.01029013, + "balance_loss_clip": 0.93263388, + "balance_loss_mlp": 1.02098417, + "epoch": 0.7976913364997295, + "flos": 22492433479680.0, + "grad_norm": 1.8839342227540419, + "language_loss": 0.70252204, + "learning_rate": 4.140444502475713e-07, + "loss": 0.72372615, + "num_input_tokens_seen": 142956445, + "step": 6634, + "time_per_iteration": 2.7204909324645996 + }, + { + "auxiliary_loss_clip": 0.01140109, + "auxiliary_loss_mlp": 0.01029363, + "balance_loss_clip": 1.00986099, + "balance_loss_mlp": 1.02219605, + "epoch": 0.7978115793903685, + "flos": 15263220378240.0, + "grad_norm": 2.0121478562628776, + "language_loss": 0.70200336, + "learning_rate": 4.1356998163896216e-07, + "loss": 0.72369802, + "num_input_tokens_seen": 142973495, + "step": 6635, + "time_per_iteration": 2.609166145324707 + }, + { + "auxiliary_loss_clip": 0.01100522, + "auxiliary_loss_mlp": 0.01023986, + "balance_loss_clip": 0.93423265, + "balance_loss_mlp": 1.01717651, + "epoch": 0.7979318222810077, + "flos": 19719232041600.0, + "grad_norm": 1.9957408897068292, + "language_loss": 0.74778724, + "learning_rate": 4.130957536905255e-07, + "loss": 0.76903236, + "num_input_tokens_seen": 142991510, + "step": 6636, + "time_per_iteration": 2.7062723636627197 + }, + { + "auxiliary_loss_clip": 0.0112998, + "auxiliary_loss_mlp": 0.01023712, + "balance_loss_clip": 0.97432941, + "balance_loss_mlp": 1.01617181, + "epoch": 0.7980520651716467, + "flos": 15560273854080.0, + "grad_norm": 2.194249279616035, + "language_loss": 0.71326935, + "learning_rate": 4.1262176647420134e-07, + "loss": 0.73480636, + "num_input_tokens_seen": 143009675, + "step": 6637, + "time_per_iteration": 2.6180837154388428 + }, + { + "auxiliary_loss_clip": 0.01126778, + "auxiliary_loss_mlp": 0.01024375, + "balance_loss_clip": 0.97492719, + "balance_loss_mlp": 1.01706791, + "epoch": 0.7981723080622858, + "flos": 22309432663680.0, + "grad_norm": 2.6183500244327944, + "language_loss": 0.79819, + "learning_rate": 4.121480200618923e-07, + "loss": 0.81970149, + "num_input_tokens_seen": 143029330, + "step": 6638, + "time_per_iteration": 2.6959457397460938 + }, + { + "auxiliary_loss_clip": 0.01114247, + "auxiliary_loss_mlp": 0.01027478, + "balance_loss_clip": 0.97103453, + "balance_loss_mlp": 1.01989639, + "epoch": 0.798292550952925, + "flos": 22929573997440.0, + "grad_norm": 2.270624410452798, + "language_loss": 0.79935849, + "learning_rate": 4.116745145254674e-07, + "loss": 0.82077575, + "num_input_tokens_seen": 143048865, + "step": 6639, + "time_per_iteration": 2.6912119388580322 + }, + { + "auxiliary_loss_clip": 0.01018337, + "auxiliary_loss_mlp": 0.01001279, + "balance_loss_clip": 0.93568015, + "balance_loss_mlp": 1.00018811, + "epoch": 0.798412793843564, + "flos": 64497936890880.0, + "grad_norm": 0.7778161369904729, + "language_loss": 0.58066308, + "learning_rate": 4.1120124993675476e-07, + "loss": 0.60085928, + "num_input_tokens_seen": 143113295, + "step": 6640, + "time_per_iteration": 4.074777841567993 + }, + { + "auxiliary_loss_clip": 0.01126674, + "auxiliary_loss_mlp": 0.01030801, + "balance_loss_clip": 0.97292042, + "balance_loss_mlp": 1.0231688, + "epoch": 0.7985330367342031, + "flos": 13586910514560.0, + "grad_norm": 1.8602688404709167, + "language_loss": 0.61810529, + "learning_rate": 4.107282263675498e-07, + "loss": 0.63968009, + "num_input_tokens_seen": 143130965, + "step": 6641, + "time_per_iteration": 2.6651439666748047 + }, + { + "auxiliary_loss_clip": 0.01020101, + "auxiliary_loss_mlp": 0.00501501, + "balance_loss_clip": 0.9379127, + "balance_loss_mlp": 0.0, + "epoch": 0.7986532796248422, + "flos": 67698797656320.0, + "grad_norm": 0.7627221767594871, + "language_loss": 0.52503353, + "learning_rate": 4.1025544388960907e-07, + "loss": 0.54024959, + "num_input_tokens_seen": 143192005, + "step": 6642, + "time_per_iteration": 4.097802639007568 + }, + { + "auxiliary_loss_clip": 0.0114297, + "auxiliary_loss_mlp": 0.01024621, + "balance_loss_clip": 1.01290822, + "balance_loss_mlp": 1.01689589, + "epoch": 0.7987735225154813, + "flos": 22455373622400.0, + "grad_norm": 1.8330648575079371, + "language_loss": 0.71644378, + "learning_rate": 4.097829025746538e-07, + "loss": 0.73811972, + "num_input_tokens_seen": 143213550, + "step": 6643, + "time_per_iteration": 2.643348455429077 + }, + { + "auxiliary_loss_clip": 0.01042021, + "auxiliary_loss_mlp": 0.01002258, + "balance_loss_clip": 0.9736551, + "balance_loss_mlp": 1.00117302, + "epoch": 0.7988937654061203, + "flos": 68864098682880.0, + "grad_norm": 0.6599494243278985, + "language_loss": 0.61013228, + "learning_rate": 4.0931060249436757e-07, + "loss": 0.63057506, + "num_input_tokens_seen": 143277390, + "step": 6644, + "time_per_iteration": 4.132117986679077 + }, + { + "auxiliary_loss_clip": 0.01144893, + "auxiliary_loss_mlp": 0.01026997, + "balance_loss_clip": 1.01420784, + "balance_loss_mlp": 1.01940954, + "epoch": 0.7990140082967595, + "flos": 20806893820800.0, + "grad_norm": 3.3175061212168893, + "language_loss": 0.69686103, + "learning_rate": 4.088385437203978e-07, + "loss": 0.71857989, + "num_input_tokens_seen": 143294400, + "step": 6645, + "time_per_iteration": 2.635857105255127 + }, + { + "auxiliary_loss_clip": 0.01167914, + "auxiliary_loss_mlp": 0.01029002, + "balance_loss_clip": 1.04949129, + "balance_loss_mlp": 1.02175069, + "epoch": 0.7991342511873986, + "flos": 18985289443200.0, + "grad_norm": 2.337220209332457, + "language_loss": 0.77536339, + "learning_rate": 4.083667263243564e-07, + "loss": 0.79733253, + "num_input_tokens_seen": 143312745, + "step": 6646, + "time_per_iteration": 2.690190076828003 + }, + { + "auxiliary_loss_clip": 0.01144308, + "auxiliary_loss_mlp": 0.01027068, + "balance_loss_clip": 1.01609421, + "balance_loss_mlp": 1.0199126, + "epoch": 0.7992544940780376, + "flos": 20816805974400.0, + "grad_norm": 1.8274155884004015, + "language_loss": 0.71881187, + "learning_rate": 4.0789515037781653e-07, + "loss": 0.74052566, + "num_input_tokens_seen": 143333470, + "step": 6647, + "time_per_iteration": 2.67348313331604 + }, + { + "auxiliary_loss_clip": 0.01145548, + "auxiliary_loss_mlp": 0.01031243, + "balance_loss_clip": 1.01154709, + "balance_loss_mlp": 1.02379274, + "epoch": 0.7993747369686768, + "flos": 12640772321280.0, + "grad_norm": 1.811349644801296, + "language_loss": 0.82683814, + "learning_rate": 4.0742381595231755e-07, + "loss": 0.84860611, + "num_input_tokens_seen": 143350195, + "step": 6648, + "time_per_iteration": 2.6099307537078857 + }, + { + "auxiliary_loss_clip": 0.01099843, + "auxiliary_loss_mlp": 0.01025903, + "balance_loss_clip": 0.93387806, + "balance_loss_mlp": 1.01861358, + "epoch": 0.7994949798593158, + "flos": 20078769225600.0, + "grad_norm": 1.6721238019233864, + "language_loss": 0.7846868, + "learning_rate": 4.06952723119359e-07, + "loss": 0.8059442, + "num_input_tokens_seen": 143370070, + "step": 6649, + "time_per_iteration": 2.7124812602996826 + }, + { + "auxiliary_loss_clip": 0.01113275, + "auxiliary_loss_mlp": 0.0102329, + "balance_loss_clip": 0.97315764, + "balance_loss_mlp": 1.0153718, + "epoch": 0.7996152227499549, + "flos": 38654209509120.0, + "grad_norm": 2.0725037692497676, + "language_loss": 0.67109448, + "learning_rate": 4.0648187195040504e-07, + "loss": 0.69246006, + "num_input_tokens_seen": 143392275, + "step": 6650, + "time_per_iteration": 2.7962865829467773 + }, + { + "auxiliary_loss_clip": 0.0103696, + "auxiliary_loss_mlp": 0.00998193, + "balance_loss_clip": 0.97228914, + "balance_loss_mlp": 0.99706697, + "epoch": 0.799735465640594, + "flos": 70243821947520.0, + "grad_norm": 0.8173278413209054, + "language_loss": 0.67554694, + "learning_rate": 4.060112625168848e-07, + "loss": 0.69589847, + "num_input_tokens_seen": 143457385, + "step": 6651, + "time_per_iteration": 4.168779611587524 + }, + { + "auxiliary_loss_clip": 0.01169039, + "auxiliary_loss_mlp": 0.01028843, + "balance_loss_clip": 1.051548, + "balance_loss_mlp": 1.02031326, + "epoch": 0.7998557085312331, + "flos": 24240995550720.0, + "grad_norm": 2.1720634080233725, + "language_loss": 0.74025512, + "learning_rate": 4.055408948901886e-07, + "loss": 0.76223397, + "num_input_tokens_seen": 143478785, + "step": 6652, + "time_per_iteration": 2.6301863193511963 + }, + { + "auxiliary_loss_clip": 0.01147052, + "auxiliary_loss_mlp": 0.01023905, + "balance_loss_clip": 1.01211464, + "balance_loss_mlp": 1.0164814, + "epoch": 0.7999759514218722, + "flos": 27564025449600.0, + "grad_norm": 3.7138133199865395, + "language_loss": 0.71586347, + "learning_rate": 4.050707691416708e-07, + "loss": 0.73757303, + "num_input_tokens_seen": 143500095, + "step": 6653, + "time_per_iteration": 2.7171528339385986 + }, + { + "auxiliary_loss_clip": 0.01036896, + "auxiliary_loss_mlp": 0.00998559, + "balance_loss_clip": 0.97224438, + "balance_loss_mlp": 0.99745655, + "epoch": 0.8000961943125112, + "flos": 67337428878720.0, + "grad_norm": 0.6766270247293465, + "language_loss": 0.59795117, + "learning_rate": 4.046008853426495e-07, + "loss": 0.61830568, + "num_input_tokens_seen": 143563410, + "step": 6654, + "time_per_iteration": 3.250641107559204 + }, + { + "auxiliary_loss_clip": 0.01091645, + "auxiliary_loss_mlp": 0.0102712, + "balance_loss_clip": 0.93285942, + "balance_loss_mlp": 1.0192765, + "epoch": 0.8002164372031504, + "flos": 28733815676160.0, + "grad_norm": 12.370768958819204, + "language_loss": 0.62756598, + "learning_rate": 4.0413124356440464e-07, + "loss": 0.64875364, + "num_input_tokens_seen": 143587455, + "step": 6655, + "time_per_iteration": 2.7472987174987793 + }, + { + "auxiliary_loss_clip": 0.01072851, + "auxiliary_loss_mlp": 0.01025953, + "balance_loss_clip": 0.89303315, + "balance_loss_mlp": 1.01851463, + "epoch": 0.8003366800937894, + "flos": 17639429725440.0, + "grad_norm": 2.5801830573668223, + "language_loss": 0.8239305, + "learning_rate": 4.0366184387818223e-07, + "loss": 0.84491849, + "num_input_tokens_seen": 143605915, + "step": 6656, + "time_per_iteration": 2.7262017726898193 + }, + { + "auxiliary_loss_clip": 0.01172784, + "auxiliary_loss_mlp": 0.01028161, + "balance_loss_clip": 1.051507, + "balance_loss_mlp": 1.02030468, + "epoch": 0.8004569229844285, + "flos": 25995303797760.0, + "grad_norm": 1.7174436898552625, + "language_loss": 0.85187781, + "learning_rate": 4.0319268635518797e-07, + "loss": 0.8738873, + "num_input_tokens_seen": 143626490, + "step": 6657, + "time_per_iteration": 229.9182846546173 + }, + { + "auxiliary_loss_clip": 0.01144015, + "auxiliary_loss_mlp": 0.01027091, + "balance_loss_clip": 1.01031208, + "balance_loss_mlp": 1.0193367, + "epoch": 0.8005771658750677, + "flos": 20812352688000.0, + "grad_norm": 1.5936346851504606, + "language_loss": 0.7535972, + "learning_rate": 4.027237710665943e-07, + "loss": 0.77530825, + "num_input_tokens_seen": 143644955, + "step": 6658, + "time_per_iteration": 2.714085102081299 + }, + { + "auxiliary_loss_clip": 0.01098972, + "auxiliary_loss_mlp": 0.01023641, + "balance_loss_clip": 0.93228614, + "balance_loss_mlp": 1.01608658, + "epoch": 0.8006974087657067, + "flos": 25812626204160.0, + "grad_norm": 1.7146592476172122, + "language_loss": 0.69590074, + "learning_rate": 4.022550980835344e-07, + "loss": 0.71712685, + "num_input_tokens_seen": 143667200, + "step": 6659, + "time_per_iteration": 2.852431058883667 + }, + { + "auxiliary_loss_clip": 0.01095958, + "auxiliary_loss_mlp": 0.01023716, + "balance_loss_clip": 0.93120956, + "balance_loss_mlp": 1.01650333, + "epoch": 0.8008176516563458, + "flos": 17164690646400.0, + "grad_norm": 3.826290486298246, + "language_loss": 0.79398459, + "learning_rate": 4.017866674771051e-07, + "loss": 0.81518131, + "num_input_tokens_seen": 143684685, + "step": 6660, + "time_per_iteration": 2.6901416778564453 + }, + { + "auxiliary_loss_clip": 0.01063815, + "auxiliary_loss_mlp": 0.01023178, + "balance_loss_clip": 0.89196688, + "balance_loss_mlp": 1.01538718, + "epoch": 0.8009378945469849, + "flos": 24207311571840.0, + "grad_norm": 1.9280668150771547, + "language_loss": 0.74749708, + "learning_rate": 4.013184793183688e-07, + "loss": 0.76836693, + "num_input_tokens_seen": 143706780, + "step": 6661, + "time_per_iteration": 2.7622475624084473 + }, + { + "auxiliary_loss_clip": 0.01140431, + "auxiliary_loss_mlp": 0.01023096, + "balance_loss_clip": 1.00810003, + "balance_loss_mlp": 1.01627779, + "epoch": 0.801058137437624, + "flos": 19787318271360.0, + "grad_norm": 1.8915951820817258, + "language_loss": 0.7264539, + "learning_rate": 4.008505336783472e-07, + "loss": 0.74808919, + "num_input_tokens_seen": 143724505, + "step": 6662, + "time_per_iteration": 2.6461267471313477 + }, + { + "auxiliary_loss_clip": 0.01133121, + "auxiliary_loss_mlp": 0.01025067, + "balance_loss_clip": 1.00786829, + "balance_loss_mlp": 1.01793814, + "epoch": 0.801178380328263, + "flos": 18659400324480.0, + "grad_norm": 1.8324042422330389, + "language_loss": 0.8075937, + "learning_rate": 4.003828306280284e-07, + "loss": 0.82917565, + "num_input_tokens_seen": 143742180, + "step": 6663, + "time_per_iteration": 2.6756327152252197 + }, + { + "auxiliary_loss_clip": 0.0114559, + "auxiliary_loss_mlp": 0.01024228, + "balance_loss_clip": 1.0119276, + "balance_loss_mlp": 1.01698565, + "epoch": 0.8012986232189022, + "flos": 15706573948800.0, + "grad_norm": 1.7405818750763835, + "language_loss": 0.78251809, + "learning_rate": 3.999153702383626e-07, + "loss": 0.80421627, + "num_input_tokens_seen": 143760070, + "step": 6664, + "time_per_iteration": 2.6527698040008545 + }, + { + "auxiliary_loss_clip": 0.01144166, + "auxiliary_loss_mlp": 0.01025392, + "balance_loss_clip": 1.00935674, + "balance_loss_mlp": 1.01816213, + "epoch": 0.8014188661095413, + "flos": 28584139703040.0, + "grad_norm": 1.7625082123985907, + "language_loss": 0.74021971, + "learning_rate": 3.9944815258026263e-07, + "loss": 0.76191527, + "num_input_tokens_seen": 143781890, + "step": 6665, + "time_per_iteration": 2.7347865104675293 + }, + { + "auxiliary_loss_clip": 0.01147836, + "auxiliary_loss_mlp": 0.01027955, + "balance_loss_clip": 1.01235211, + "balance_loss_mlp": 1.01994443, + "epoch": 0.8015391090001803, + "flos": 29310360877440.0, + "grad_norm": 1.7771138820325365, + "language_loss": 0.82724625, + "learning_rate": 3.989811777246057e-07, + "loss": 0.84900415, + "num_input_tokens_seen": 143802060, + "step": 6666, + "time_per_iteration": 3.7094154357910156 + }, + { + "auxiliary_loss_clip": 0.01061196, + "auxiliary_loss_mlp": 0.01000163, + "balance_loss_clip": 1.01003206, + "balance_loss_mlp": 0.99912548, + "epoch": 0.8016593518908195, + "flos": 70397340675840.0, + "grad_norm": 0.848053658613517, + "language_loss": 0.66284108, + "learning_rate": 3.985144457422305e-07, + "loss": 0.68345463, + "num_input_tokens_seen": 143856345, + "step": 6667, + "time_per_iteration": 3.982034683227539 + }, + { + "auxiliary_loss_clip": 0.01167743, + "auxiliary_loss_mlp": 0.01024591, + "balance_loss_clip": 1.05017507, + "balance_loss_mlp": 1.01751614, + "epoch": 0.8017795947814585, + "flos": 26026114688640.0, + "grad_norm": 3.327817292133707, + "language_loss": 0.7690649, + "learning_rate": 3.9804795670394096e-07, + "loss": 0.79098827, + "num_input_tokens_seen": 143876470, + "step": 6668, + "time_per_iteration": 2.6254723072052 + }, + { + "auxiliary_loss_clip": 0.0111364, + "auxiliary_loss_mlp": 0.01026478, + "balance_loss_clip": 0.97088397, + "balance_loss_mlp": 1.01943612, + "epoch": 0.8018998376720976, + "flos": 22087181260800.0, + "grad_norm": 1.6877796026969778, + "language_loss": 0.70732093, + "learning_rate": 3.975817106805022e-07, + "loss": 0.7287221, + "num_input_tokens_seen": 143895170, + "step": 6669, + "time_per_iteration": 3.9015533924102783 + }, + { + "auxiliary_loss_clip": 0.0109587, + "auxiliary_loss_mlp": 0.01026612, + "balance_loss_clip": 0.93355036, + "balance_loss_mlp": 1.01882792, + "epoch": 0.8020200805627368, + "flos": 34568545023360.0, + "grad_norm": 1.922281157989761, + "language_loss": 0.65047926, + "learning_rate": 3.97115707742645e-07, + "loss": 0.67170405, + "num_input_tokens_seen": 143915845, + "step": 6670, + "time_per_iteration": 2.8602476119995117 + }, + { + "auxiliary_loss_clip": 0.01124204, + "auxiliary_loss_mlp": 0.01025688, + "balance_loss_clip": 0.9752754, + "balance_loss_mlp": 1.01793075, + "epoch": 0.8021403234533758, + "flos": 20120354196480.0, + "grad_norm": 2.054065305282127, + "language_loss": 0.65155458, + "learning_rate": 3.966499479610599e-07, + "loss": 0.6730535, + "num_input_tokens_seen": 143933940, + "step": 6671, + "time_per_iteration": 2.6734941005706787 + }, + { + "auxiliary_loss_clip": 0.01093486, + "auxiliary_loss_mlp": 0.01021776, + "balance_loss_clip": 0.93547076, + "balance_loss_mlp": 1.01518393, + "epoch": 0.8022605663440149, + "flos": 27746200252800.0, + "grad_norm": 12.166903268310746, + "language_loss": 0.65262467, + "learning_rate": 3.9618443140640225e-07, + "loss": 0.67377728, + "num_input_tokens_seen": 143952850, + "step": 6672, + "time_per_iteration": 2.7552595138549805 + }, + { + "auxiliary_loss_clip": 0.00981292, + "auxiliary_loss_mlp": 0.01003436, + "balance_loss_clip": 0.86279047, + "balance_loss_mlp": 1.00236344, + "epoch": 0.802380809234654, + "flos": 60244998768000.0, + "grad_norm": 0.6817086764211158, + "language_loss": 0.51341188, + "learning_rate": 3.957191581492918e-07, + "loss": 0.53325915, + "num_input_tokens_seen": 144013610, + "step": 6673, + "time_per_iteration": 3.368462324142456 + }, + { + "auxiliary_loss_clip": 0.01114105, + "auxiliary_loss_mlp": 0.01025754, + "balance_loss_clip": 0.97090751, + "balance_loss_mlp": 1.01840472, + "epoch": 0.8025010521252931, + "flos": 15080722352640.0, + "grad_norm": 2.6903944259570656, + "language_loss": 0.71024269, + "learning_rate": 3.952541282603097e-07, + "loss": 0.73164129, + "num_input_tokens_seen": 144028715, + "step": 6674, + "time_per_iteration": 2.712080240249634 + }, + { + "auxiliary_loss_clip": 0.01138317, + "auxiliary_loss_mlp": 0.01029076, + "balance_loss_clip": 1.00970387, + "balance_loss_mlp": 1.02126145, + "epoch": 0.8026212950159322, + "flos": 22163527618560.0, + "grad_norm": 1.745454447102963, + "language_loss": 0.83396935, + "learning_rate": 3.9478934181000013e-07, + "loss": 0.85564327, + "num_input_tokens_seen": 144048740, + "step": 6675, + "time_per_iteration": 2.665609121322632 + }, + { + "auxiliary_loss_clip": 0.01168821, + "auxiliary_loss_mlp": 0.01029109, + "balance_loss_clip": 1.04890394, + "balance_loss_mlp": 1.02132487, + "epoch": 0.8027415379065713, + "flos": 17675986792320.0, + "grad_norm": 2.1651318595186395, + "language_loss": 0.84536642, + "learning_rate": 3.943247988688714e-07, + "loss": 0.86734569, + "num_input_tokens_seen": 144067435, + "step": 6676, + "time_per_iteration": 2.7320525646209717 + }, + { + "auxiliary_loss_clip": 0.01143449, + "auxiliary_loss_mlp": 0.01025179, + "balance_loss_clip": 1.01002216, + "balance_loss_mlp": 1.0184257, + "epoch": 0.8028617807972104, + "flos": 21979593048960.0, + "grad_norm": 1.7221587699292769, + "language_loss": 0.72149444, + "learning_rate": 3.938604995073933e-07, + "loss": 0.74318075, + "num_input_tokens_seen": 144085905, + "step": 6677, + "time_per_iteration": 3.646343469619751 + }, + { + "auxiliary_loss_clip": 0.01121731, + "auxiliary_loss_mlp": 0.01026868, + "balance_loss_clip": 0.971753, + "balance_loss_mlp": 1.01924467, + "epoch": 0.8029820236878494, + "flos": 26428457905920.0, + "grad_norm": 1.9202287023163267, + "language_loss": 0.65035486, + "learning_rate": 3.9339644379600157e-07, + "loss": 0.67184085, + "num_input_tokens_seen": 144105735, + "step": 6678, + "time_per_iteration": 2.682034969329834 + }, + { + "auxiliary_loss_clip": 0.0114813, + "auxiliary_loss_mlp": 0.01025033, + "balance_loss_clip": 1.01370859, + "balance_loss_mlp": 1.01821446, + "epoch": 0.8031022665784886, + "flos": 17676489582720.0, + "grad_norm": 1.705841898382144, + "language_loss": 0.71147573, + "learning_rate": 3.929326318050907e-07, + "loss": 0.73320735, + "num_input_tokens_seen": 144123405, + "step": 6679, + "time_per_iteration": 2.6580419540405273 + }, + { + "auxiliary_loss_clip": 0.01162132, + "auxiliary_loss_mlp": 0.01025898, + "balance_loss_clip": 1.04687667, + "balance_loss_mlp": 1.01838183, + "epoch": 0.8032225094691277, + "flos": 15450279431040.0, + "grad_norm": 1.8005492255009232, + "language_loss": 0.78958243, + "learning_rate": 3.924690636050225e-07, + "loss": 0.81146276, + "num_input_tokens_seen": 144140815, + "step": 6680, + "time_per_iteration": 2.524914026260376 + }, + { + "auxiliary_loss_clip": 0.01143848, + "auxiliary_loss_mlp": 0.01026817, + "balance_loss_clip": 1.01091123, + "balance_loss_mlp": 1.01937497, + "epoch": 0.8033427523597667, + "flos": 26179202453760.0, + "grad_norm": 2.451662665596385, + "language_loss": 0.73169684, + "learning_rate": 3.9200573926611915e-07, + "loss": 0.75340343, + "num_input_tokens_seen": 144162230, + "step": 6681, + "time_per_iteration": 2.7479846477508545 + }, + { + "auxiliary_loss_clip": 0.01142293, + "auxiliary_loss_mlp": 0.01027224, + "balance_loss_clip": 1.01391125, + "balance_loss_mlp": 1.01991916, + "epoch": 0.8034629952504058, + "flos": 21324905809920.0, + "grad_norm": 1.945054973726498, + "language_loss": 0.72751862, + "learning_rate": 3.9154265885866613e-07, + "loss": 0.74921376, + "num_input_tokens_seen": 144181540, + "step": 6682, + "time_per_iteration": 2.6222176551818848 + }, + { + "auxiliary_loss_clip": 0.01141723, + "auxiliary_loss_mlp": 0.01020983, + "balance_loss_clip": 1.01068342, + "balance_loss_mlp": 1.01342487, + "epoch": 0.8035832381410449, + "flos": 21651585027840.0, + "grad_norm": 2.639373055367413, + "language_loss": 0.74629653, + "learning_rate": 3.9107982245291394e-07, + "loss": 0.76792359, + "num_input_tokens_seen": 144199665, + "step": 6683, + "time_per_iteration": 2.7400591373443604 + }, + { + "auxiliary_loss_clip": 0.01098482, + "auxiliary_loss_mlp": 0.01024592, + "balance_loss_clip": 0.93700296, + "balance_loss_mlp": 1.01727235, + "epoch": 0.803703481031684, + "flos": 20518818744960.0, + "grad_norm": 2.1823311043916434, + "language_loss": 0.7726469, + "learning_rate": 3.9061723011907245e-07, + "loss": 0.79387766, + "num_input_tokens_seen": 144219020, + "step": 6684, + "time_per_iteration": 2.7724695205688477 + }, + { + "auxiliary_loss_clip": 0.01117436, + "auxiliary_loss_mlp": 0.01027146, + "balance_loss_clip": 0.97137356, + "balance_loss_mlp": 1.02005923, + "epoch": 0.803823723922323, + "flos": 22854807838080.0, + "grad_norm": 2.647988548552584, + "language_loss": 0.79542029, + "learning_rate": 3.901548819273179e-07, + "loss": 0.8168661, + "num_input_tokens_seen": 144239035, + "step": 6685, + "time_per_iteration": 2.761024236679077 + }, + { + "auxiliary_loss_clip": 0.01146579, + "auxiliary_loss_mlp": 0.01022759, + "balance_loss_clip": 1.01346266, + "balance_loss_mlp": 1.01563025, + "epoch": 0.8039439668129622, + "flos": 21362145235200.0, + "grad_norm": 2.098487674542611, + "language_loss": 0.69040895, + "learning_rate": 3.896927779477881e-07, + "loss": 0.71210229, + "num_input_tokens_seen": 144258295, + "step": 6686, + "time_per_iteration": 2.6028738021850586 + }, + { + "auxiliary_loss_clip": 0.01095329, + "auxiliary_loss_mlp": 0.01029638, + "balance_loss_clip": 0.93261147, + "balance_loss_mlp": 1.0225594, + "epoch": 0.8040642097036013, + "flos": 23802382575360.0, + "grad_norm": 9.475652519983091, + "language_loss": 0.67292523, + "learning_rate": 3.892309182505833e-07, + "loss": 0.69417489, + "num_input_tokens_seen": 144276110, + "step": 6687, + "time_per_iteration": 2.8040411472320557 + }, + { + "auxiliary_loss_clip": 0.01166817, + "auxiliary_loss_mlp": 0.01026379, + "balance_loss_clip": 1.04925418, + "balance_loss_mlp": 1.01942301, + "epoch": 0.8041844525942403, + "flos": 25922046009600.0, + "grad_norm": 2.5630973066278524, + "language_loss": 0.86522448, + "learning_rate": 3.887693029057675e-07, + "loss": 0.88715643, + "num_input_tokens_seen": 144295620, + "step": 6688, + "time_per_iteration": 2.656615734100342 + }, + { + "auxiliary_loss_clip": 0.01122478, + "auxiliary_loss_mlp": 0.01023729, + "balance_loss_clip": 0.97396898, + "balance_loss_mlp": 1.01710367, + "epoch": 0.8043046954848795, + "flos": 25191120153600.0, + "grad_norm": 1.9422364848985052, + "language_loss": 0.80989349, + "learning_rate": 3.8830793198336684e-07, + "loss": 0.83135557, + "num_input_tokens_seen": 144315210, + "step": 6689, + "time_per_iteration": 2.7070212364196777 + }, + { + "auxiliary_loss_clip": 0.01146036, + "auxiliary_loss_mlp": 0.0102501, + "balance_loss_clip": 1.00934315, + "balance_loss_mlp": 1.01791358, + "epoch": 0.8044249383755185, + "flos": 41719185123840.0, + "grad_norm": 1.7155543611167103, + "language_loss": 0.7050525, + "learning_rate": 3.878468055533721e-07, + "loss": 0.72676301, + "num_input_tokens_seen": 144337750, + "step": 6690, + "time_per_iteration": 2.7797470092773438 + }, + { + "auxiliary_loss_clip": 0.0110467, + "auxiliary_loss_mlp": 0.01028811, + "balance_loss_clip": 0.93715763, + "balance_loss_mlp": 1.02094316, + "epoch": 0.8045451812661576, + "flos": 20631434860800.0, + "grad_norm": 2.29465192111588, + "language_loss": 0.84357506, + "learning_rate": 3.8738592368573464e-07, + "loss": 0.86490989, + "num_input_tokens_seen": 144355305, + "step": 6691, + "time_per_iteration": 2.750666618347168 + }, + { + "auxiliary_loss_clip": 0.01090357, + "auxiliary_loss_mlp": 0.01023441, + "balance_loss_clip": 0.93438369, + "balance_loss_mlp": 1.01561785, + "epoch": 0.8046654241567968, + "flos": 29711806254720.0, + "grad_norm": 1.9850794725288685, + "language_loss": 0.87797213, + "learning_rate": 3.8692528645037137e-07, + "loss": 0.89911014, + "num_input_tokens_seen": 144374485, + "step": 6692, + "time_per_iteration": 3.721782922744751 + }, + { + "auxiliary_loss_clip": 0.01167493, + "auxiliary_loss_mlp": 0.01027075, + "balance_loss_clip": 1.04980707, + "balance_loss_mlp": 1.0199883, + "epoch": 0.8047856670474358, + "flos": 17671389851520.0, + "grad_norm": 2.3816218157465983, + "language_loss": 0.77645457, + "learning_rate": 3.8646489391715907e-07, + "loss": 0.79840028, + "num_input_tokens_seen": 144388780, + "step": 6693, + "time_per_iteration": 2.6367952823638916 + }, + { + "auxiliary_loss_clip": 0.01121565, + "auxiliary_loss_mlp": 0.01034925, + "balance_loss_clip": 0.97321093, + "balance_loss_mlp": 1.02680707, + "epoch": 0.8049059099380749, + "flos": 17120699464320.0, + "grad_norm": 3.803693828723012, + "language_loss": 0.87975687, + "learning_rate": 3.8600474615593903e-07, + "loss": 0.90132177, + "num_input_tokens_seen": 144403395, + "step": 6694, + "time_per_iteration": 3.6122677326202393 + }, + { + "auxiliary_loss_clip": 0.01003441, + "auxiliary_loss_mlp": 0.00999916, + "balance_loss_clip": 0.89877963, + "balance_loss_mlp": 0.99888438, + "epoch": 0.805026152828714, + "flos": 62212903240320.0, + "grad_norm": 0.7834712928218669, + "language_loss": 0.59645283, + "learning_rate": 3.8554484323651605e-07, + "loss": 0.61648637, + "num_input_tokens_seen": 144465265, + "step": 6695, + "time_per_iteration": 4.192009210586548 + }, + { + "auxiliary_loss_clip": 0.01141525, + "auxiliary_loss_mlp": 0.00509403, + "balance_loss_clip": 1.0109359, + "balance_loss_mlp": 0.0, + "epoch": 0.8051463957193531, + "flos": 21688608971520.0, + "grad_norm": 1.4678710566906352, + "language_loss": 0.79139626, + "learning_rate": 3.85085185228657e-07, + "loss": 0.80790555, + "num_input_tokens_seen": 144484235, + "step": 6696, + "time_per_iteration": 2.667631149291992 + }, + { + "auxiliary_loss_clip": 0.01114833, + "auxiliary_loss_mlp": 0.01032328, + "balance_loss_clip": 0.97128928, + "balance_loss_mlp": 1.02498209, + "epoch": 0.8052666386099921, + "flos": 32051458535040.0, + "grad_norm": 1.907429281917777, + "language_loss": 0.73299015, + "learning_rate": 3.8462577220209114e-07, + "loss": 0.75446177, + "num_input_tokens_seen": 144504610, + "step": 6697, + "time_per_iteration": 2.784695863723755 + }, + { + "auxiliary_loss_clip": 0.01061089, + "auxiliary_loss_mlp": 0.0100344, + "balance_loss_clip": 1.00959158, + "balance_loss_mlp": 1.00239682, + "epoch": 0.8053868815006313, + "flos": 67157875768320.0, + "grad_norm": 0.6989890033430483, + "language_loss": 0.58970439, + "learning_rate": 3.8416660422651127e-07, + "loss": 0.61034966, + "num_input_tokens_seen": 144574260, + "step": 6698, + "time_per_iteration": 3.2637221813201904 + }, + { + "auxiliary_loss_clip": 0.0109865, + "auxiliary_loss_mlp": 0.01029189, + "balance_loss_clip": 0.93207681, + "balance_loss_mlp": 1.02167296, + "epoch": 0.8055071243912704, + "flos": 23837000307840.0, + "grad_norm": 1.6872462452559664, + "language_loss": 0.68120384, + "learning_rate": 3.837076813715723e-07, + "loss": 0.70248222, + "num_input_tokens_seen": 144594145, + "step": 6699, + "time_per_iteration": 2.667208671569824 + }, + { + "auxiliary_loss_clip": 0.01095412, + "auxiliary_loss_mlp": 0.01027901, + "balance_loss_clip": 0.93262899, + "balance_loss_mlp": 1.02012253, + "epoch": 0.8056273672819094, + "flos": 21324510760320.0, + "grad_norm": 16.784811329055916, + "language_loss": 0.75442886, + "learning_rate": 3.832490037068941e-07, + "loss": 0.77566195, + "num_input_tokens_seen": 144612935, + "step": 6700, + "time_per_iteration": 2.7274887561798096 + }, + { + "auxiliary_loss_clip": 0.01043735, + "auxiliary_loss_mlp": 0.01025463, + "balance_loss_clip": 0.85466254, + "balance_loss_mlp": 1.01804554, + "epoch": 0.8057476101725486, + "flos": 25768383626880.0, + "grad_norm": 1.8966731541679793, + "language_loss": 0.76009303, + "learning_rate": 3.827905713020554e-07, + "loss": 0.78078502, + "num_input_tokens_seen": 144630580, + "step": 6701, + "time_per_iteration": 2.8153998851776123 + }, + { + "auxiliary_loss_clip": 0.01098862, + "auxiliary_loss_mlp": 0.01029263, + "balance_loss_clip": 0.93159384, + "balance_loss_mlp": 1.02081108, + "epoch": 0.8058678530631876, + "flos": 24535283679360.0, + "grad_norm": 1.828093497747688, + "language_loss": 0.68859792, + "learning_rate": 3.823323842266017e-07, + "loss": 0.70987916, + "num_input_tokens_seen": 144649975, + "step": 6702, + "time_per_iteration": 2.8077147006988525 + }, + { + "auxiliary_loss_clip": 0.01144338, + "auxiliary_loss_mlp": 0.01026978, + "balance_loss_clip": 1.00843716, + "balance_loss_mlp": 1.01967025, + "epoch": 0.8059880959538267, + "flos": 24753728240640.0, + "grad_norm": 3.2654409472689085, + "language_loss": 0.73144734, + "learning_rate": 3.818744425500393e-07, + "loss": 0.75316048, + "num_input_tokens_seen": 144667990, + "step": 6703, + "time_per_iteration": 3.602764129638672 + }, + { + "auxiliary_loss_clip": 0.01089762, + "auxiliary_loss_mlp": 0.01023712, + "balance_loss_clip": 0.93071175, + "balance_loss_mlp": 1.01649725, + "epoch": 0.8061083388444659, + "flos": 22196349671040.0, + "grad_norm": 1.835897669817471, + "language_loss": 0.80624163, + "learning_rate": 3.8141674634183675e-07, + "loss": 0.82737637, + "num_input_tokens_seen": 144687020, + "step": 6704, + "time_per_iteration": 2.7204158306121826 + }, + { + "auxiliary_loss_clip": 0.01072975, + "auxiliary_loss_mlp": 0.01027824, + "balance_loss_clip": 0.89901453, + "balance_loss_mlp": 1.02125812, + "epoch": 0.8062285817351049, + "flos": 30044195735040.0, + "grad_norm": 1.7150306306033105, + "language_loss": 0.66348875, + "learning_rate": 3.809592956714278e-07, + "loss": 0.68449676, + "num_input_tokens_seen": 144710255, + "step": 6705, + "time_per_iteration": 2.8710877895355225 + }, + { + "auxiliary_loss_clip": 0.01149667, + "auxiliary_loss_mlp": 0.01025227, + "balance_loss_clip": 1.01390839, + "balance_loss_mlp": 1.01728797, + "epoch": 0.806348824625744, + "flos": 22782591544320.0, + "grad_norm": 1.848637507824779, + "language_loss": 0.74518508, + "learning_rate": 3.805020906082057e-07, + "loss": 0.76693404, + "num_input_tokens_seen": 144728830, + "step": 6706, + "time_per_iteration": 2.6450564861297607 + }, + { + "auxiliary_loss_clip": 0.01122885, + "auxiliary_loss_mlp": 0.01030427, + "balance_loss_clip": 0.97267675, + "balance_loss_mlp": 1.02303648, + "epoch": 0.8064690675163831, + "flos": 23404600385280.0, + "grad_norm": 2.051030276805097, + "language_loss": 0.81174624, + "learning_rate": 3.8004513122152917e-07, + "loss": 0.83327937, + "num_input_tokens_seen": 144747140, + "step": 6707, + "time_per_iteration": 2.772183895111084 + }, + { + "auxiliary_loss_clip": 0.01113976, + "auxiliary_loss_mlp": 0.01026143, + "balance_loss_clip": 0.97612828, + "balance_loss_mlp": 1.01924694, + "epoch": 0.8065893104070222, + "flos": 24060903736320.0, + "grad_norm": 1.8632402609911058, + "language_loss": 0.67376518, + "learning_rate": 3.79588417580718e-07, + "loss": 0.69516641, + "num_input_tokens_seen": 144765250, + "step": 6708, + "time_per_iteration": 2.695497989654541 + }, + { + "auxiliary_loss_clip": 0.01142329, + "auxiliary_loss_mlp": 0.01025528, + "balance_loss_clip": 1.01031101, + "balance_loss_mlp": 1.01794946, + "epoch": 0.8067095532976613, + "flos": 22305410340480.0, + "grad_norm": 2.0499359459608075, + "language_loss": 0.7629993, + "learning_rate": 3.791319497550558e-07, + "loss": 0.78467786, + "num_input_tokens_seen": 144783080, + "step": 6709, + "time_per_iteration": 2.6773669719696045 + }, + { + "auxiliary_loss_clip": 0.01103628, + "auxiliary_loss_mlp": 0.00508965, + "balance_loss_clip": 0.93661809, + "balance_loss_mlp": 0.0, + "epoch": 0.8068297961883004, + "flos": 17129498296320.0, + "grad_norm": 1.8697246174190458, + "language_loss": 0.70537555, + "learning_rate": 3.78675727813788e-07, + "loss": 0.72150147, + "num_input_tokens_seen": 144800645, + "step": 6710, + "time_per_iteration": 2.7335896492004395 + }, + { + "auxiliary_loss_clip": 0.01119742, + "auxiliary_loss_mlp": 0.01027325, + "balance_loss_clip": 0.97367942, + "balance_loss_mlp": 1.01980281, + "epoch": 0.8069500390789395, + "flos": 22018843635840.0, + "grad_norm": 1.7503329133947527, + "language_loss": 0.73920918, + "learning_rate": 3.782197518261225e-07, + "loss": 0.76067984, + "num_input_tokens_seen": 144820085, + "step": 6711, + "time_per_iteration": 2.716594934463501 + }, + { + "auxiliary_loss_clip": 0.01124892, + "auxiliary_loss_mlp": 0.01027201, + "balance_loss_clip": 0.97304916, + "balance_loss_mlp": 1.01997733, + "epoch": 0.8070702819695785, + "flos": 19244241567360.0, + "grad_norm": 2.0055495834331727, + "language_loss": 0.95568275, + "learning_rate": 3.777640218612319e-07, + "loss": 0.97720373, + "num_input_tokens_seen": 144838070, + "step": 6712, + "time_per_iteration": 2.7411558628082275 + }, + { + "auxiliary_loss_clip": 0.01138401, + "auxiliary_loss_mlp": 0.01025839, + "balance_loss_clip": 1.00996923, + "balance_loss_mlp": 1.01840031, + "epoch": 0.8071905248602176, + "flos": 21544320038400.0, + "grad_norm": 2.459470175249041, + "language_loss": 0.71745509, + "learning_rate": 3.773085379882488e-07, + "loss": 0.73909748, + "num_input_tokens_seen": 144857125, + "step": 6713, + "time_per_iteration": 2.765958070755005 + }, + { + "auxiliary_loss_clip": 0.01140647, + "auxiliary_loss_mlp": 0.00509356, + "balance_loss_clip": 1.00679362, + "balance_loss_mlp": 0.0, + "epoch": 0.8073107677508568, + "flos": 37268309105280.0, + "grad_norm": 1.8409358091031847, + "language_loss": 0.7598455, + "learning_rate": 3.768533002762715e-07, + "loss": 0.77634549, + "num_input_tokens_seen": 144880660, + "step": 6714, + "time_per_iteration": 2.769035577774048 + }, + { + "auxiliary_loss_clip": 0.01118832, + "auxiliary_loss_mlp": 0.01024342, + "balance_loss_clip": 0.96931475, + "balance_loss_mlp": 1.01733804, + "epoch": 0.8074310106414958, + "flos": 28366269759360.0, + "grad_norm": 1.8491818801907385, + "language_loss": 0.77025694, + "learning_rate": 3.763983087943572e-07, + "loss": 0.79168868, + "num_input_tokens_seen": 144900050, + "step": 6715, + "time_per_iteration": 2.7624831199645996 + }, + { + "auxiliary_loss_clip": 0.01134544, + "auxiliary_loss_mlp": 0.00509323, + "balance_loss_clip": 1.0079155, + "balance_loss_mlp": 0.0, + "epoch": 0.8075512535321349, + "flos": 24281646768000.0, + "grad_norm": 1.689195907250033, + "language_loss": 0.80848634, + "learning_rate": 3.759435636115282e-07, + "loss": 0.82492507, + "num_input_tokens_seen": 144920835, + "step": 6716, + "time_per_iteration": 2.7060282230377197 + }, + { + "auxiliary_loss_clip": 0.01045858, + "auxiliary_loss_mlp": 0.00509468, + "balance_loss_clip": 0.85990798, + "balance_loss_mlp": 0.0, + "epoch": 0.807671496422774, + "flos": 26030855283840.0, + "grad_norm": 1.8675932051254613, + "language_loss": 0.73376477, + "learning_rate": 3.7548906479676967e-07, + "loss": 0.74931806, + "num_input_tokens_seen": 144940430, + "step": 6717, + "time_per_iteration": 2.7708864212036133 + }, + { + "auxiliary_loss_clip": 0.01145064, + "auxiliary_loss_mlp": 0.01026903, + "balance_loss_clip": 1.00933588, + "balance_loss_mlp": 1.0195539, + "epoch": 0.8077917393134131, + "flos": 23730740899200.0, + "grad_norm": 1.8643777901261738, + "language_loss": 0.71843868, + "learning_rate": 3.7503481241902855e-07, + "loss": 0.74015832, + "num_input_tokens_seen": 144960405, + "step": 6718, + "time_per_iteration": 3.547781467437744 + }, + { + "auxiliary_loss_clip": 0.01118166, + "auxiliary_loss_mlp": 0.00509614, + "balance_loss_clip": 0.97175944, + "balance_loss_mlp": 0.0, + "epoch": 0.8079119822040521, + "flos": 18402028398720.0, + "grad_norm": 1.6802220343239098, + "language_loss": 0.79831588, + "learning_rate": 3.745808065472145e-07, + "loss": 0.81459367, + "num_input_tokens_seen": 144977700, + "step": 6719, + "time_per_iteration": 2.688102960586548 + }, + { + "auxiliary_loss_clip": 0.01143795, + "auxiliary_loss_mlp": 0.01027844, + "balance_loss_clip": 1.01849246, + "balance_loss_mlp": 1.02102256, + "epoch": 0.8080322250946913, + "flos": 23621787970560.0, + "grad_norm": 1.586047692545909, + "language_loss": 0.76377547, + "learning_rate": 3.741270472501994e-07, + "loss": 0.78549182, + "num_input_tokens_seen": 144998340, + "step": 6720, + "time_per_iteration": 3.6191928386688232 + }, + { + "auxiliary_loss_clip": 0.01117305, + "auxiliary_loss_mlp": 0.01027003, + "balance_loss_clip": 0.9733752, + "balance_loss_mlp": 1.01994562, + "epoch": 0.8081524679853304, + "flos": 22820692896000.0, + "grad_norm": 1.6094022073722187, + "language_loss": 0.73005259, + "learning_rate": 3.736735345968183e-07, + "loss": 0.75149572, + "num_input_tokens_seen": 145017950, + "step": 6721, + "time_per_iteration": 3.5703964233398438 + }, + { + "auxiliary_loss_clip": 0.0114776, + "auxiliary_loss_mlp": 0.01028655, + "balance_loss_clip": 1.01443195, + "balance_loss_mlp": 1.02150857, + "epoch": 0.8082727108759694, + "flos": 17640004343040.0, + "grad_norm": 1.7720756246157776, + "language_loss": 0.78691363, + "learning_rate": 3.7322026865586986e-07, + "loss": 0.80867779, + "num_input_tokens_seen": 145036985, + "step": 6722, + "time_per_iteration": 2.644252300262451 + }, + { + "auxiliary_loss_clip": 0.01151968, + "auxiliary_loss_mlp": 0.01028413, + "balance_loss_clip": 1.01484144, + "balance_loss_mlp": 1.02154064, + "epoch": 0.8083929537666086, + "flos": 25958172113280.0, + "grad_norm": 1.979074818731895, + "language_loss": 0.73434246, + "learning_rate": 3.7276724949611206e-07, + "loss": 0.75614625, + "num_input_tokens_seen": 145057095, + "step": 6723, + "time_per_iteration": 2.654755115509033 + }, + { + "auxiliary_loss_clip": 0.01122464, + "auxiliary_loss_mlp": 0.01028055, + "balance_loss_clip": 0.9722265, + "balance_loss_mlp": 1.0203836, + "epoch": 0.8085131966572476, + "flos": 27089178629760.0, + "grad_norm": 1.9322386823255056, + "language_loss": 0.75158948, + "learning_rate": 3.723144771862694e-07, + "loss": 0.77309465, + "num_input_tokens_seen": 145077735, + "step": 6724, + "time_per_iteration": 2.779750108718872 + }, + { + "auxiliary_loss_clip": 0.01100122, + "auxiliary_loss_mlp": 0.0102542, + "balance_loss_clip": 0.93315506, + "balance_loss_mlp": 1.01782084, + "epoch": 0.8086334395478867, + "flos": 23988543788160.0, + "grad_norm": 1.9915797354357268, + "language_loss": 0.76792145, + "learning_rate": 3.718619517950263e-07, + "loss": 0.78917688, + "num_input_tokens_seen": 145098330, + "step": 6725, + "time_per_iteration": 2.776503086090088 + }, + { + "auxiliary_loss_clip": 0.01167303, + "auxiliary_loss_mlp": 0.01032607, + "balance_loss_clip": 1.05115199, + "balance_loss_mlp": 1.02549624, + "epoch": 0.8087536824385259, + "flos": 20405879406720.0, + "grad_norm": 1.9758588173773979, + "language_loss": 0.76765156, + "learning_rate": 3.714096733910301e-07, + "loss": 0.78965068, + "num_input_tokens_seen": 145115855, + "step": 6726, + "time_per_iteration": 2.6514148712158203 + }, + { + "auxiliary_loss_clip": 0.0114898, + "auxiliary_loss_mlp": 0.01030774, + "balance_loss_clip": 1.01146519, + "balance_loss_mlp": 1.02303421, + "epoch": 0.8088739253291649, + "flos": 25919639798400.0, + "grad_norm": 1.9812658517424684, + "language_loss": 0.69991976, + "learning_rate": 3.709576420428926e-07, + "loss": 0.72171736, + "num_input_tokens_seen": 145136655, + "step": 6727, + "time_per_iteration": 2.789743661880493 + }, + { + "auxiliary_loss_clip": 0.01119054, + "auxiliary_loss_mlp": 0.010315, + "balance_loss_clip": 0.96954513, + "balance_loss_mlp": 1.02431726, + "epoch": 0.808994168219804, + "flos": 28402072640640.0, + "grad_norm": 1.9994916373251628, + "language_loss": 0.73348141, + "learning_rate": 3.7050585781918463e-07, + "loss": 0.754987, + "num_input_tokens_seen": 145156955, + "step": 6728, + "time_per_iteration": 2.7341668605804443 + }, + { + "auxiliary_loss_clip": 0.01145018, + "auxiliary_loss_mlp": 0.01032258, + "balance_loss_clip": 1.00962174, + "balance_loss_mlp": 1.02483177, + "epoch": 0.8091144111104431, + "flos": 17421056991360.0, + "grad_norm": 2.684590534781135, + "language_loss": 0.68980926, + "learning_rate": 3.700543207884428e-07, + "loss": 0.71158206, + "num_input_tokens_seen": 145173865, + "step": 6729, + "time_per_iteration": 3.5514867305755615 + }, + { + "auxiliary_loss_clip": 0.01140323, + "auxiliary_loss_mlp": 0.01026456, + "balance_loss_clip": 1.01013517, + "balance_loss_mlp": 1.01957798, + "epoch": 0.8092346540010822, + "flos": 32153803361280.0, + "grad_norm": 2.080272893719736, + "language_loss": 0.70693195, + "learning_rate": 3.6960303101916466e-07, + "loss": 0.72859967, + "num_input_tokens_seen": 145193780, + "step": 6730, + "time_per_iteration": 2.702251434326172 + }, + { + "auxiliary_loss_clip": 0.01060796, + "auxiliary_loss_mlp": 0.00500834, + "balance_loss_clip": 1.00965071, + "balance_loss_mlp": 0.0, + "epoch": 0.8093548968917212, + "flos": 58035093390720.0, + "grad_norm": 0.750726148118989, + "language_loss": 0.55539203, + "learning_rate": 3.6915198857981047e-07, + "loss": 0.57100832, + "num_input_tokens_seen": 145258980, + "step": 6731, + "time_per_iteration": 3.290656566619873 + }, + { + "auxiliary_loss_clip": 0.01097103, + "auxiliary_loss_mlp": 0.01026921, + "balance_loss_clip": 0.93759179, + "balance_loss_mlp": 1.01983666, + "epoch": 0.8094751397823604, + "flos": 27381599251200.0, + "grad_norm": 1.744320817609139, + "language_loss": 0.68235391, + "learning_rate": 3.687011935388027e-07, + "loss": 0.70359415, + "num_input_tokens_seen": 145281875, + "step": 6732, + "time_per_iteration": 2.756100654602051 + }, + { + "auxiliary_loss_clip": 0.01142822, + "auxiliary_loss_mlp": 0.01027789, + "balance_loss_clip": 1.01100945, + "balance_loss_mlp": 1.02080333, + "epoch": 0.8095953826729995, + "flos": 24061083304320.0, + "grad_norm": 2.030970675552875, + "language_loss": 0.73199201, + "learning_rate": 3.6825064596452646e-07, + "loss": 0.75369811, + "num_input_tokens_seen": 145302220, + "step": 6733, + "time_per_iteration": 2.6482295989990234 + }, + { + "auxiliary_loss_clip": 0.01142676, + "auxiliary_loss_mlp": 0.01024765, + "balance_loss_clip": 1.00907052, + "balance_loss_mlp": 1.01824415, + "epoch": 0.8097156255636385, + "flos": 23951412103680.0, + "grad_norm": 1.5995997205840204, + "language_loss": 0.7110579, + "learning_rate": 3.678003459253305e-07, + "loss": 0.73273236, + "num_input_tokens_seen": 145323070, + "step": 6734, + "time_per_iteration": 2.6461429595947266 + }, + { + "auxiliary_loss_clip": 0.01095322, + "auxiliary_loss_mlp": 0.01027255, + "balance_loss_clip": 0.93451464, + "balance_loss_mlp": 1.02007258, + "epoch": 0.8098358684542777, + "flos": 21799142098560.0, + "grad_norm": 2.2693045596629866, + "language_loss": 0.74121749, + "learning_rate": 3.673502934895236e-07, + "loss": 0.76244324, + "num_input_tokens_seen": 145342575, + "step": 6735, + "time_per_iteration": 2.7030365467071533 + }, + { + "auxiliary_loss_clip": 0.01060854, + "auxiliary_loss_mlp": 0.00999574, + "balance_loss_clip": 1.00941229, + "balance_loss_mlp": 0.99857819, + "epoch": 0.8099561113449167, + "flos": 68809515966720.0, + "grad_norm": 0.6931911061990648, + "language_loss": 0.57976431, + "learning_rate": 3.669004887253802e-07, + "loss": 0.60036862, + "num_input_tokens_seen": 145408865, + "step": 6736, + "time_per_iteration": 3.3791544437408447 + }, + { + "auxiliary_loss_clip": 0.01121778, + "auxiliary_loss_mlp": 0.01029362, + "balance_loss_clip": 0.97328746, + "balance_loss_mlp": 1.02329183, + "epoch": 0.8100763542355558, + "flos": 23586056916480.0, + "grad_norm": 1.5191658803948842, + "language_loss": 0.78853762, + "learning_rate": 3.664509317011335e-07, + "loss": 0.81004906, + "num_input_tokens_seen": 145429200, + "step": 6737, + "time_per_iteration": 2.74320650100708 + }, + { + "auxiliary_loss_clip": 0.01142994, + "auxiliary_loss_mlp": 0.01028696, + "balance_loss_clip": 1.01337838, + "balance_loss_mlp": 1.02095938, + "epoch": 0.810196597126195, + "flos": 31650408207360.0, + "grad_norm": 1.820165733537653, + "language_loss": 0.74127454, + "learning_rate": 3.6600162248498134e-07, + "loss": 0.76299143, + "num_input_tokens_seen": 145452830, + "step": 6738, + "time_per_iteration": 2.7133915424346924 + }, + { + "auxiliary_loss_clip": 0.01032117, + "auxiliary_loss_mlp": 0.01025435, + "balance_loss_clip": 0.8540163, + "balance_loss_mlp": 1.01821375, + "epoch": 0.810316840016834, + "flos": 24900459298560.0, + "grad_norm": 1.8893234000995107, + "language_loss": 0.75807166, + "learning_rate": 3.6555256114508426e-07, + "loss": 0.77864712, + "num_input_tokens_seen": 145472625, + "step": 6739, + "time_per_iteration": 2.7707157135009766 + }, + { + "auxiliary_loss_clip": 0.01119509, + "auxiliary_loss_mlp": 0.01023209, + "balance_loss_clip": 0.96927011, + "balance_loss_mlp": 1.01557338, + "epoch": 0.8104370829074731, + "flos": 27965003950080.0, + "grad_norm": 2.1337609679819325, + "language_loss": 0.73078132, + "learning_rate": 3.651037477495642e-07, + "loss": 0.75220847, + "num_input_tokens_seen": 145494075, + "step": 6740, + "time_per_iteration": 2.768000364303589 + }, + { + "auxiliary_loss_clip": 0.0116433, + "auxiliary_loss_mlp": 0.01027732, + "balance_loss_clip": 1.04710865, + "balance_loss_mlp": 1.02033806, + "epoch": 0.8105573257981122, + "flos": 24640752988800.0, + "grad_norm": 2.07596750904738, + "language_loss": 0.67949748, + "learning_rate": 3.6465518236650584e-07, + "loss": 0.70141804, + "num_input_tokens_seen": 145514220, + "step": 6741, + "time_per_iteration": 2.6516096591949463 + }, + { + "auxiliary_loss_clip": 0.01093012, + "auxiliary_loss_mlp": 0.01028509, + "balance_loss_clip": 0.93156326, + "balance_loss_mlp": 1.0218035, + "epoch": 0.8106775686887513, + "flos": 26358935132160.0, + "grad_norm": 2.074432303325487, + "language_loss": 0.78162038, + "learning_rate": 3.642068650639558e-07, + "loss": 0.80283558, + "num_input_tokens_seen": 145533965, + "step": 6742, + "time_per_iteration": 2.906135082244873 + }, + { + "auxiliary_loss_clip": 0.01110745, + "auxiliary_loss_mlp": 0.01025898, + "balance_loss_clip": 0.9667781, + "balance_loss_mlp": 1.01870131, + "epoch": 0.8107978115793903, + "flos": 27271892136960.0, + "grad_norm": 2.13303300058268, + "language_loss": 0.64630282, + "learning_rate": 3.6375879590992334e-07, + "loss": 0.6676693, + "num_input_tokens_seen": 145554310, + "step": 6743, + "time_per_iteration": 2.787008285522461 + }, + { + "auxiliary_loss_clip": 0.01115989, + "auxiliary_loss_mlp": 0.01027348, + "balance_loss_clip": 0.97132432, + "balance_loss_mlp": 1.01948035, + "epoch": 0.8109180544700295, + "flos": 24934322845440.0, + "grad_norm": 1.770922075935293, + "language_loss": 0.80895823, + "learning_rate": 3.6331097497238173e-07, + "loss": 0.83039159, + "num_input_tokens_seen": 145573755, + "step": 6744, + "time_per_iteration": 3.5902998447418213 + }, + { + "auxiliary_loss_clip": 0.01091747, + "auxiliary_loss_mlp": 0.01022479, + "balance_loss_clip": 0.93244702, + "balance_loss_mlp": 1.0154283, + "epoch": 0.8110382973606686, + "flos": 21105383840640.0, + "grad_norm": 1.7777113277453278, + "language_loss": 0.79692072, + "learning_rate": 3.628634023192627e-07, + "loss": 0.81806296, + "num_input_tokens_seen": 145594000, + "step": 6745, + "time_per_iteration": 2.7550535202026367 + }, + { + "auxiliary_loss_clip": 0.01142662, + "auxiliary_loss_mlp": 0.01027857, + "balance_loss_clip": 1.009817, + "balance_loss_mlp": 1.02068007, + "epoch": 0.8111585402513076, + "flos": 15414081500160.0, + "grad_norm": 1.8912638639766386, + "language_loss": 0.74956828, + "learning_rate": 3.624160780184644e-07, + "loss": 0.77127349, + "num_input_tokens_seen": 145611215, + "step": 6746, + "time_per_iteration": 3.490591049194336 + }, + { + "auxiliary_loss_clip": 0.01112789, + "auxiliary_loss_mlp": 0.01023085, + "balance_loss_clip": 0.97079861, + "balance_loss_mlp": 1.01599789, + "epoch": 0.8112787831419467, + "flos": 24095736950400.0, + "grad_norm": 1.9557367986760217, + "language_loss": 0.7467227, + "learning_rate": 3.6196900213784496e-07, + "loss": 0.76808143, + "num_input_tokens_seen": 145630530, + "step": 6747, + "time_per_iteration": 3.5952556133270264 + }, + { + "auxiliary_loss_clip": 0.01142136, + "auxiliary_loss_mlp": 0.01025262, + "balance_loss_clip": 1.00962019, + "balance_loss_mlp": 1.01819313, + "epoch": 0.8113990260325858, + "flos": 20483374999680.0, + "grad_norm": 1.7849020808866651, + "language_loss": 0.86467373, + "learning_rate": 3.6152217474522527e-07, + "loss": 0.88634765, + "num_input_tokens_seen": 145647345, + "step": 6748, + "time_per_iteration": 2.6581501960754395 + }, + { + "auxiliary_loss_clip": 0.01142398, + "auxiliary_loss_mlp": 0.01026982, + "balance_loss_clip": 1.01296091, + "balance_loss_mlp": 1.02013958, + "epoch": 0.8115192689232249, + "flos": 24901141656960.0, + "grad_norm": 1.5541779475447883, + "language_loss": 0.72671568, + "learning_rate": 3.6107559590838975e-07, + "loss": 0.74840951, + "num_input_tokens_seen": 145666330, + "step": 6749, + "time_per_iteration": 2.6540613174438477 + }, + { + "auxiliary_loss_clip": 0.01045376, + "auxiliary_loss_mlp": 0.01026818, + "balance_loss_clip": 0.85639513, + "balance_loss_mlp": 1.01908112, + "epoch": 0.811639511813864, + "flos": 24057204635520.0, + "grad_norm": 2.356892861029, + "language_loss": 0.66690975, + "learning_rate": 3.606292656950822e-07, + "loss": 0.68763173, + "num_input_tokens_seen": 145684740, + "step": 6750, + "time_per_iteration": 2.81777286529541 + }, + { + "auxiliary_loss_clip": 0.01115309, + "auxiliary_loss_mlp": 0.01022074, + "balance_loss_clip": 0.96996439, + "balance_loss_mlp": 1.01443279, + "epoch": 0.8117597547045031, + "flos": 23185150243200.0, + "grad_norm": 2.2682828356374087, + "language_loss": 0.865058, + "learning_rate": 3.601831841730121e-07, + "loss": 0.88643181, + "num_input_tokens_seen": 145702660, + "step": 6751, + "time_per_iteration": 2.6282551288604736 + }, + { + "auxiliary_loss_clip": 0.01141114, + "auxiliary_loss_mlp": 0.01028067, + "balance_loss_clip": 1.01016021, + "balance_loss_mlp": 1.02143896, + "epoch": 0.8118799975951422, + "flos": 23040250778880.0, + "grad_norm": 1.6849816825570605, + "language_loss": 0.72776353, + "learning_rate": 3.5973735140984916e-07, + "loss": 0.74945533, + "num_input_tokens_seen": 145722830, + "step": 6752, + "time_per_iteration": 2.664229393005371 + }, + { + "auxiliary_loss_clip": 0.01066601, + "auxiliary_loss_mlp": 0.00509107, + "balance_loss_clip": 0.89163166, + "balance_loss_mlp": 0.0, + "epoch": 0.8120002404857812, + "flos": 24639962889600.0, + "grad_norm": 2.1824053105986034, + "language_loss": 0.79896581, + "learning_rate": 3.5929176747322607e-07, + "loss": 0.8147229, + "num_input_tokens_seen": 145741935, + "step": 6753, + "time_per_iteration": 2.769497871398926 + }, + { + "auxiliary_loss_clip": 0.01025297, + "auxiliary_loss_mlp": 0.00999635, + "balance_loss_clip": 0.93614411, + "balance_loss_mlp": 0.99852622, + "epoch": 0.8121204833764204, + "flos": 57415742156160.0, + "grad_norm": 0.8130547947239644, + "language_loss": 0.56170106, + "learning_rate": 3.588464324307372e-07, + "loss": 0.58195031, + "num_input_tokens_seen": 145805560, + "step": 6754, + "time_per_iteration": 3.2693710327148438 + }, + { + "auxiliary_loss_clip": 0.01143041, + "auxiliary_loss_mlp": 0.01022608, + "balance_loss_clip": 1.00766504, + "balance_loss_mlp": 1.01579821, + "epoch": 0.8122407262670595, + "flos": 19464589549440.0, + "grad_norm": 1.8097836525667177, + "language_loss": 0.75605285, + "learning_rate": 3.584013463499391e-07, + "loss": 0.77770936, + "num_input_tokens_seen": 145824180, + "step": 6755, + "time_per_iteration": 3.607110023498535 + }, + { + "auxiliary_loss_clip": 0.01021351, + "auxiliary_loss_mlp": 0.01005469, + "balance_loss_clip": 0.9348588, + "balance_loss_mlp": 1.00442588, + "epoch": 0.8123609691576985, + "flos": 56425325472000.0, + "grad_norm": 0.7343757876231412, + "language_loss": 0.64384103, + "learning_rate": 3.579565092983521e-07, + "loss": 0.66410923, + "num_input_tokens_seen": 145885300, + "step": 6756, + "time_per_iteration": 3.127570629119873 + }, + { + "auxiliary_loss_clip": 0.01165454, + "auxiliary_loss_mlp": 0.01023896, + "balance_loss_clip": 1.0490005, + "balance_loss_mlp": 1.01651716, + "epoch": 0.8124812120483377, + "flos": 20631973564800.0, + "grad_norm": 2.2384230748580145, + "language_loss": 0.84110242, + "learning_rate": 3.575119213434565e-07, + "loss": 0.86299592, + "num_input_tokens_seen": 145903815, + "step": 6757, + "time_per_iteration": 2.782341241836548 + }, + { + "auxiliary_loss_clip": 0.01139926, + "auxiliary_loss_mlp": 0.0102516, + "balance_loss_clip": 1.01155651, + "balance_loss_mlp": 1.01805496, + "epoch": 0.8126014549389767, + "flos": 22492397566080.0, + "grad_norm": 3.1861529224895393, + "language_loss": 0.81954169, + "learning_rate": 3.5706758255269765e-07, + "loss": 0.8411926, + "num_input_tokens_seen": 145922270, + "step": 6758, + "time_per_iteration": 2.692730188369751 + }, + { + "auxiliary_loss_clip": 0.01126066, + "auxiliary_loss_mlp": 0.01023696, + "balance_loss_clip": 0.97353733, + "balance_loss_mlp": 1.01662135, + "epoch": 0.8127216978296158, + "flos": 23287961946240.0, + "grad_norm": 1.5840885143407466, + "language_loss": 0.69672239, + "learning_rate": 3.566234929934795e-07, + "loss": 0.71822006, + "num_input_tokens_seen": 145941470, + "step": 6759, + "time_per_iteration": 2.7455992698669434 + }, + { + "auxiliary_loss_clip": 0.01141911, + "auxiliary_loss_mlp": 0.01024827, + "balance_loss_clip": 1.0146296, + "balance_loss_mlp": 1.01759124, + "epoch": 0.812841940720255, + "flos": 25154994049920.0, + "grad_norm": 1.5152900286455153, + "language_loss": 0.71383691, + "learning_rate": 3.561796527331706e-07, + "loss": 0.73550433, + "num_input_tokens_seen": 145963145, + "step": 6760, + "time_per_iteration": 2.7515997886657715 + }, + { + "auxiliary_loss_clip": 0.01096375, + "auxiliary_loss_mlp": 0.01025776, + "balance_loss_clip": 0.93272042, + "balance_loss_mlp": 1.01858771, + "epoch": 0.812962183610894, + "flos": 26648446752000.0, + "grad_norm": 2.1053492454456393, + "language_loss": 0.77444375, + "learning_rate": 3.5573606183910163e-07, + "loss": 0.79566526, + "num_input_tokens_seen": 145983150, + "step": 6761, + "time_per_iteration": 2.743783712387085 + }, + { + "auxiliary_loss_clip": 0.01146212, + "auxiliary_loss_mlp": 0.01027918, + "balance_loss_clip": 1.00832319, + "balance_loss_mlp": 1.01990414, + "epoch": 0.8130824265015331, + "flos": 24966965329920.0, + "grad_norm": 1.699491938909376, + "language_loss": 0.78438127, + "learning_rate": 3.5529272037856493e-07, + "loss": 0.8061226, + "num_input_tokens_seen": 146001365, + "step": 6762, + "time_per_iteration": 2.653681993484497 + }, + { + "auxiliary_loss_clip": 0.00968629, + "auxiliary_loss_mlp": 0.01012037, + "balance_loss_clip": 0.8239851, + "balance_loss_mlp": 1.01092243, + "epoch": 0.8132026693921722, + "flos": 67622918175360.0, + "grad_norm": 0.7172666806764928, + "language_loss": 0.53847229, + "learning_rate": 3.548496284188149e-07, + "loss": 0.55827904, + "num_input_tokens_seen": 146061570, + "step": 6763, + "time_per_iteration": 3.62907075881958 + }, + { + "auxiliary_loss_clip": 0.01067617, + "auxiliary_loss_mlp": 0.01025648, + "balance_loss_clip": 0.89644158, + "balance_loss_mlp": 1.01836419, + "epoch": 0.8133229122828113, + "flos": 19495149045120.0, + "grad_norm": 1.7533929792606133, + "language_loss": 0.79448807, + "learning_rate": 3.544067860270681e-07, + "loss": 0.81542075, + "num_input_tokens_seen": 146079145, + "step": 6764, + "time_per_iteration": 3.118243932723999 + }, + { + "auxiliary_loss_clip": 0.01098531, + "auxiliary_loss_mlp": 0.01025247, + "balance_loss_clip": 0.93272662, + "balance_loss_mlp": 1.01718843, + "epoch": 0.8134431551734503, + "flos": 20668135582080.0, + "grad_norm": 1.8437439443665953, + "language_loss": 0.70852399, + "learning_rate": 3.539641932705029e-07, + "loss": 0.72976178, + "num_input_tokens_seen": 146097625, + "step": 6765, + "time_per_iteration": 2.7813079357147217 + }, + { + "auxiliary_loss_clip": 0.0116822, + "auxiliary_loss_mlp": 0.01024308, + "balance_loss_clip": 1.0490458, + "balance_loss_mlp": 1.01661301, + "epoch": 0.8135633980640895, + "flos": 21507332008320.0, + "grad_norm": 2.5848306052634697, + "language_loss": 0.77231425, + "learning_rate": 3.53521850216262e-07, + "loss": 0.79423952, + "num_input_tokens_seen": 146117195, + "step": 6766, + "time_per_iteration": 2.6493473052978516 + }, + { + "auxiliary_loss_clip": 0.01167061, + "auxiliary_loss_mlp": 0.01028407, + "balance_loss_clip": 1.04987395, + "balance_loss_mlp": 1.02124834, + "epoch": 0.8136836409547286, + "flos": 20554442058240.0, + "grad_norm": 1.7032502656567527, + "language_loss": 0.76945734, + "learning_rate": 3.530797569314461e-07, + "loss": 0.791412, + "num_input_tokens_seen": 146136220, + "step": 6767, + "time_per_iteration": 2.734833240509033 + }, + { + "auxiliary_loss_clip": 0.01166349, + "auxiliary_loss_mlp": 0.01024076, + "balance_loss_clip": 1.04936409, + "balance_loss_mlp": 1.01638389, + "epoch": 0.8138038838453676, + "flos": 20299045380480.0, + "grad_norm": 1.693529004443467, + "language_loss": 0.77682114, + "learning_rate": 3.5263791348312235e-07, + "loss": 0.79872543, + "num_input_tokens_seen": 146155415, + "step": 6768, + "time_per_iteration": 2.6554551124572754 + }, + { + "auxiliary_loss_clip": 0.01117738, + "auxiliary_loss_mlp": 0.01020914, + "balance_loss_clip": 0.97099829, + "balance_loss_mlp": 1.01304662, + "epoch": 0.8139241267360068, + "flos": 29789840551680.0, + "grad_norm": 2.4691455369879933, + "language_loss": 0.70665497, + "learning_rate": 3.521963199383171e-07, + "loss": 0.72804147, + "num_input_tokens_seen": 146178370, + "step": 6769, + "time_per_iteration": 2.8681812286376953 + }, + { + "auxiliary_loss_clip": 0.01070752, + "auxiliary_loss_mlp": 0.01024378, + "balance_loss_clip": 0.89380264, + "balance_loss_mlp": 1.01718688, + "epoch": 0.8140443696266458, + "flos": 19713270384000.0, + "grad_norm": 2.2919627885103813, + "language_loss": 0.7684207, + "learning_rate": 3.517549763640197e-07, + "loss": 0.78937203, + "num_input_tokens_seen": 146196010, + "step": 6770, + "time_per_iteration": 3.5782270431518555 + }, + { + "auxiliary_loss_clip": 0.01139044, + "auxiliary_loss_mlp": 0.00508616, + "balance_loss_clip": 1.01128864, + "balance_loss_mlp": 0.0, + "epoch": 0.8141646125172849, + "flos": 27160568910720.0, + "grad_norm": 1.7018600398872699, + "language_loss": 0.71274698, + "learning_rate": 3.513138828271829e-07, + "loss": 0.72922361, + "num_input_tokens_seen": 146215880, + "step": 6771, + "time_per_iteration": 2.7261736392974854 + }, + { + "auxiliary_loss_clip": 0.0109281, + "auxiliary_loss_mlp": 0.0102373, + "balance_loss_clip": 0.93478531, + "balance_loss_mlp": 1.01666081, + "epoch": 0.8142848554079241, + "flos": 39673102700160.0, + "grad_norm": 1.8051857339828825, + "language_loss": 0.7028966, + "learning_rate": 3.508730393947179e-07, + "loss": 0.72406197, + "num_input_tokens_seen": 146239135, + "step": 6772, + "time_per_iteration": 3.8730480670928955 + }, + { + "auxiliary_loss_clip": 0.01094203, + "auxiliary_loss_mlp": 0.01027098, + "balance_loss_clip": 0.93426001, + "balance_loss_mlp": 1.01986265, + "epoch": 0.8144050982985631, + "flos": 22237288197120.0, + "grad_norm": 1.8637209632006326, + "language_loss": 0.72326386, + "learning_rate": 3.504324461335024e-07, + "loss": 0.7444768, + "num_input_tokens_seen": 146259245, + "step": 6773, + "time_per_iteration": 3.6747822761535645 + }, + { + "auxiliary_loss_clip": 0.01067599, + "auxiliary_loss_mlp": 0.01028741, + "balance_loss_clip": 0.89610755, + "balance_loss_mlp": 1.02114773, + "epoch": 0.8145253411892022, + "flos": 23038239617280.0, + "grad_norm": 1.7006261858885052, + "language_loss": 0.88429058, + "learning_rate": 3.499921031103732e-07, + "loss": 0.90525401, + "num_input_tokens_seen": 146280015, + "step": 6774, + "time_per_iteration": 2.7978408336639404 + }, + { + "auxiliary_loss_clip": 0.01105849, + "auxiliary_loss_mlp": 0.01026791, + "balance_loss_clip": 0.93218112, + "balance_loss_mlp": 1.01963234, + "epoch": 0.8146455840798413, + "flos": 24827668387200.0, + "grad_norm": 2.022112999233852, + "language_loss": 0.78352642, + "learning_rate": 3.4955201039212987e-07, + "loss": 0.80485284, + "num_input_tokens_seen": 146300935, + "step": 6775, + "time_per_iteration": 2.7462317943573 + }, + { + "auxiliary_loss_clip": 0.01146445, + "auxiliary_loss_mlp": 0.01025768, + "balance_loss_clip": 1.01108241, + "balance_loss_mlp": 1.01849961, + "epoch": 0.8147658269704804, + "flos": 19974520978560.0, + "grad_norm": 2.195427732608446, + "language_loss": 0.65299451, + "learning_rate": 3.4911216804553465e-07, + "loss": 0.67471659, + "num_input_tokens_seen": 146319835, + "step": 6776, + "time_per_iteration": 2.586945056915283 + }, + { + "auxiliary_loss_clip": 0.01119502, + "auxiliary_loss_mlp": 0.01029005, + "balance_loss_clip": 0.97252554, + "balance_loss_mlp": 1.02147102, + "epoch": 0.8148860698611194, + "flos": 21178031097600.0, + "grad_norm": 2.5701603697062065, + "language_loss": 0.70496714, + "learning_rate": 3.4867257613731017e-07, + "loss": 0.72645223, + "num_input_tokens_seen": 146339030, + "step": 6777, + "time_per_iteration": 2.7381083965301514 + }, + { + "auxiliary_loss_clip": 0.01120098, + "auxiliary_loss_mlp": 0.01032854, + "balance_loss_clip": 0.97066593, + "balance_loss_mlp": 1.02583897, + "epoch": 0.8150063127517585, + "flos": 19606903234560.0, + "grad_norm": 1.8238617065047025, + "language_loss": 0.85577893, + "learning_rate": 3.4823323473414343e-07, + "loss": 0.87730849, + "num_input_tokens_seen": 146358550, + "step": 6778, + "time_per_iteration": 2.6463675498962402 + }, + { + "auxiliary_loss_clip": 0.01101494, + "auxiliary_loss_mlp": 0.01027339, + "balance_loss_clip": 0.93301034, + "balance_loss_mlp": 1.01894689, + "epoch": 0.8151265556423977, + "flos": 22638374438400.0, + "grad_norm": 2.627611892189522, + "language_loss": 0.76534301, + "learning_rate": 3.477941439026812e-07, + "loss": 0.78663129, + "num_input_tokens_seen": 146376770, + "step": 6779, + "time_per_iteration": 2.6914889812469482 + }, + { + "auxiliary_loss_clip": 0.01121628, + "auxiliary_loss_mlp": 0.01028667, + "balance_loss_clip": 0.97415835, + "balance_loss_mlp": 1.02199447, + "epoch": 0.8152467985330367, + "flos": 17968048277760.0, + "grad_norm": 2.2015898039665376, + "language_loss": 0.73132879, + "learning_rate": 3.473553037095349e-07, + "loss": 0.7528317, + "num_input_tokens_seen": 146395795, + "step": 6780, + "time_per_iteration": 2.619886875152588 + }, + { + "auxiliary_loss_clip": 0.01111971, + "auxiliary_loss_mlp": 0.01027831, + "balance_loss_clip": 0.96965766, + "balance_loss_mlp": 1.02105045, + "epoch": 0.8153670414236758, + "flos": 24969012405120.0, + "grad_norm": 1.7716409481464284, + "language_loss": 0.83408475, + "learning_rate": 3.469167142212743e-07, + "loss": 0.8554827, + "num_input_tokens_seen": 146417640, + "step": 6781, + "time_per_iteration": 3.5853123664855957 + }, + { + "auxiliary_loss_clip": 0.01143101, + "auxiliary_loss_mlp": 0.01031473, + "balance_loss_clip": 1.00974298, + "balance_loss_mlp": 1.0240171, + "epoch": 0.8154872843143149, + "flos": 31066069754880.0, + "grad_norm": 2.255141691372942, + "language_loss": 0.63001764, + "learning_rate": 3.4647837550443337e-07, + "loss": 0.65176344, + "num_input_tokens_seen": 146436205, + "step": 6782, + "time_per_iteration": 2.689624547958374 + }, + { + "auxiliary_loss_clip": 0.01098361, + "auxiliary_loss_mlp": 0.01025727, + "balance_loss_clip": 0.93371052, + "balance_loss_mlp": 1.01880074, + "epoch": 0.815607527204954, + "flos": 19391654983680.0, + "grad_norm": 23.76503960260309, + "language_loss": 0.74723256, + "learning_rate": 3.460402876255086e-07, + "loss": 0.76847339, + "num_input_tokens_seen": 146453595, + "step": 6783, + "time_per_iteration": 2.6740658283233643 + }, + { + "auxiliary_loss_clip": 0.01145127, + "auxiliary_loss_mlp": 0.01031361, + "balance_loss_clip": 1.01017892, + "balance_loss_mlp": 1.02327263, + "epoch": 0.815727770095593, + "flos": 26140418743680.0, + "grad_norm": 2.1019330345179057, + "language_loss": 0.71734673, + "learning_rate": 3.456024506509574e-07, + "loss": 0.73911166, + "num_input_tokens_seen": 146474515, + "step": 6784, + "time_per_iteration": 2.6965510845184326 + }, + { + "auxiliary_loss_clip": 0.01145823, + "auxiliary_loss_mlp": 0.00509366, + "balance_loss_clip": 1.0141561, + "balance_loss_mlp": 0.0, + "epoch": 0.8158480129862322, + "flos": 25337527989120.0, + "grad_norm": 3.273389930395153, + "language_loss": 0.73995733, + "learning_rate": 3.4516486464719873e-07, + "loss": 0.75650918, + "num_input_tokens_seen": 146493905, + "step": 6785, + "time_per_iteration": 2.793201446533203 + }, + { + "auxiliary_loss_clip": 0.0106641, + "auxiliary_loss_mlp": 0.01026228, + "balance_loss_clip": 0.89282525, + "balance_loss_mlp": 1.01877773, + "epoch": 0.8159682558768713, + "flos": 34423645559040.0, + "grad_norm": 1.804643462675872, + "language_loss": 0.6184727, + "learning_rate": 3.4472752968061445e-07, + "loss": 0.63939905, + "num_input_tokens_seen": 146518335, + "step": 6786, + "time_per_iteration": 2.90409517288208 + }, + { + "auxiliary_loss_clip": 0.01142858, + "auxiliary_loss_mlp": 0.01021981, + "balance_loss_clip": 1.01024199, + "balance_loss_mlp": 1.01512957, + "epoch": 0.8160884987675103, + "flos": 18653223185280.0, + "grad_norm": 4.679957733001401, + "language_loss": 0.74000114, + "learning_rate": 3.442904458175475e-07, + "loss": 0.76164955, + "num_input_tokens_seen": 146535655, + "step": 6787, + "time_per_iteration": 2.62294602394104 + }, + { + "auxiliary_loss_clip": 0.01139573, + "auxiliary_loss_mlp": 0.01029241, + "balance_loss_clip": 1.00887454, + "balance_loss_mlp": 1.02190328, + "epoch": 0.8162087416581495, + "flos": 31430527102080.0, + "grad_norm": 2.3198972349688396, + "language_loss": 0.76122499, + "learning_rate": 3.438536131243044e-07, + "loss": 0.78291309, + "num_input_tokens_seen": 146556815, + "step": 6788, + "time_per_iteration": 2.762836217880249 + }, + { + "auxiliary_loss_clip": 0.01121529, + "auxiliary_loss_mlp": 0.01026443, + "balance_loss_clip": 0.97206718, + "balance_loss_mlp": 1.01888502, + "epoch": 0.8163289845487885, + "flos": 37593910915200.0, + "grad_norm": 2.0151150676113785, + "language_loss": 0.61888683, + "learning_rate": 3.434170316671503e-07, + "loss": 0.64036655, + "num_input_tokens_seen": 146581845, + "step": 6789, + "time_per_iteration": 2.8450465202331543 + }, + { + "auxiliary_loss_clip": 0.01095294, + "auxiliary_loss_mlp": 0.01026821, + "balance_loss_clip": 0.93895924, + "balance_loss_mlp": 1.01964498, + "epoch": 0.8164492274394276, + "flos": 13953989554560.0, + "grad_norm": 2.320836820532338, + "language_loss": 0.89778841, + "learning_rate": 3.4298070151231583e-07, + "loss": 0.91900957, + "num_input_tokens_seen": 146597245, + "step": 6790, + "time_per_iteration": 2.6990814208984375 + }, + { + "auxiliary_loss_clip": 0.01123987, + "auxiliary_loss_mlp": 0.01024084, + "balance_loss_clip": 0.97189564, + "balance_loss_mlp": 1.01690722, + "epoch": 0.8165694703300668, + "flos": 28986554747520.0, + "grad_norm": 2.5270677476747365, + "language_loss": 0.60404551, + "learning_rate": 3.425446227259916e-07, + "loss": 0.62552619, + "num_input_tokens_seen": 146618210, + "step": 6791, + "time_per_iteration": 2.6563124656677246 + }, + { + "auxiliary_loss_clip": 0.01119684, + "auxiliary_loss_mlp": 0.01024688, + "balance_loss_clip": 0.97239697, + "balance_loss_mlp": 1.0176692, + "epoch": 0.8166897132207058, + "flos": 25118365155840.0, + "grad_norm": 1.9825134444648327, + "language_loss": 0.82293427, + "learning_rate": 3.421087953743296e-07, + "loss": 0.84437799, + "num_input_tokens_seen": 146637975, + "step": 6792, + "time_per_iteration": 2.7271010875701904 + }, + { + "auxiliary_loss_clip": 0.01140085, + "auxiliary_loss_mlp": 0.01022779, + "balance_loss_clip": 1.00733733, + "balance_loss_mlp": 1.01547146, + "epoch": 0.8168099561113449, + "flos": 23148593176320.0, + "grad_norm": 1.9163322742425097, + "language_loss": 0.79832268, + "learning_rate": 3.416732195234464e-07, + "loss": 0.8199513, + "num_input_tokens_seen": 146658030, + "step": 6793, + "time_per_iteration": 2.6987602710723877 + }, + { + "auxiliary_loss_clip": 0.01145594, + "auxiliary_loss_mlp": 0.01025319, + "balance_loss_clip": 1.00967836, + "balance_loss_mlp": 1.01811862, + "epoch": 0.816930199001984, + "flos": 18407666833920.0, + "grad_norm": 2.1515444744267067, + "language_loss": 0.796597, + "learning_rate": 3.4123789523941613e-07, + "loss": 0.81830621, + "num_input_tokens_seen": 146677855, + "step": 6794, + "time_per_iteration": 2.684185743331909 + }, + { + "auxiliary_loss_clip": 0.01137404, + "auxiliary_loss_mlp": 0.01024462, + "balance_loss_clip": 1.0087347, + "balance_loss_mlp": 1.01710033, + "epoch": 0.8170504418926231, + "flos": 21251324799360.0, + "grad_norm": 1.5427294195127184, + "language_loss": 0.63421106, + "learning_rate": 3.4080282258827884e-07, + "loss": 0.65582979, + "num_input_tokens_seen": 146696230, + "step": 6795, + "time_per_iteration": 2.6687867641448975 + }, + { + "auxiliary_loss_clip": 0.01144811, + "auxiliary_loss_mlp": 0.01023559, + "balance_loss_clip": 1.01042414, + "balance_loss_mlp": 1.01671684, + "epoch": 0.8171706847832622, + "flos": 19099234362240.0, + "grad_norm": 2.2802586891457044, + "language_loss": 0.73108506, + "learning_rate": 3.403680016360342e-07, + "loss": 0.75276875, + "num_input_tokens_seen": 146714835, + "step": 6796, + "time_per_iteration": 3.479952335357666 + }, + { + "auxiliary_loss_clip": 0.01137766, + "auxiliary_loss_mlp": 0.01027437, + "balance_loss_clip": 1.01129138, + "balance_loss_mlp": 1.01987934, + "epoch": 0.8172909276739013, + "flos": 21470128496640.0, + "grad_norm": 1.9169350863440069, + "language_loss": 0.67619812, + "learning_rate": 3.3993343244864403e-07, + "loss": 0.69785023, + "num_input_tokens_seen": 146734425, + "step": 6797, + "time_per_iteration": 2.699516773223877 + }, + { + "auxiliary_loss_clip": 0.01140821, + "auxiliary_loss_mlp": 0.01023046, + "balance_loss_clip": 1.01093841, + "balance_loss_mlp": 1.01633418, + "epoch": 0.8174111705645404, + "flos": 27599792417280.0, + "grad_norm": 1.737488170377098, + "language_loss": 0.72856736, + "learning_rate": 3.394991150920323e-07, + "loss": 0.75020599, + "num_input_tokens_seen": 146757545, + "step": 6798, + "time_per_iteration": 3.6278281211853027 + }, + { + "auxiliary_loss_clip": 0.01072481, + "auxiliary_loss_mlp": 0.00509808, + "balance_loss_clip": 0.89279342, + "balance_loss_mlp": 0.0, + "epoch": 0.8175314134551794, + "flos": 14064594508800.0, + "grad_norm": 2.050188756087465, + "language_loss": 0.74374241, + "learning_rate": 3.3906504963208396e-07, + "loss": 0.75956529, + "num_input_tokens_seen": 146774240, + "step": 6799, + "time_per_iteration": 3.7087559700012207 + }, + { + "auxiliary_loss_clip": 0.01068459, + "auxiliary_loss_mlp": 0.01028542, + "balance_loss_clip": 0.89750803, + "balance_loss_mlp": 1.02110302, + "epoch": 0.8176516563458186, + "flos": 22708076780160.0, + "grad_norm": 1.8367231545554568, + "language_loss": 0.66687387, + "learning_rate": 3.3863123613464774e-07, + "loss": 0.68784392, + "num_input_tokens_seen": 146793140, + "step": 6800, + "time_per_iteration": 2.9532113075256348 + }, + { + "auxiliary_loss_clip": 0.01116839, + "auxiliary_loss_mlp": 0.01022021, + "balance_loss_clip": 0.96602452, + "balance_loss_mlp": 1.01506543, + "epoch": 0.8177718992364577, + "flos": 21945406279680.0, + "grad_norm": 1.7876704632779763, + "language_loss": 0.75152457, + "learning_rate": 3.381976746655317e-07, + "loss": 0.77291322, + "num_input_tokens_seen": 146812895, + "step": 6801, + "time_per_iteration": 2.6679835319519043 + }, + { + "auxiliary_loss_clip": 0.01068544, + "auxiliary_loss_mlp": 0.01024036, + "balance_loss_clip": 0.89856672, + "balance_loss_mlp": 1.0162096, + "epoch": 0.8178921421270967, + "flos": 22017443005440.0, + "grad_norm": 2.4392096793732407, + "language_loss": 0.67659456, + "learning_rate": 3.3776436529050756e-07, + "loss": 0.69752032, + "num_input_tokens_seen": 146832445, + "step": 6802, + "time_per_iteration": 2.6989197731018066 + }, + { + "auxiliary_loss_clip": 0.01162647, + "auxiliary_loss_mlp": 0.01021477, + "balance_loss_clip": 1.04765582, + "balance_loss_mlp": 1.01429796, + "epoch": 0.8180123850177359, + "flos": 33183111496320.0, + "grad_norm": 2.957172960537227, + "language_loss": 0.72528625, + "learning_rate": 3.373313080753073e-07, + "loss": 0.74712753, + "num_input_tokens_seen": 146856505, + "step": 6803, + "time_per_iteration": 2.720956325531006 + }, + { + "auxiliary_loss_clip": 0.0113708, + "auxiliary_loss_mlp": 0.01022913, + "balance_loss_clip": 1.00819397, + "balance_loss_mlp": 1.01531672, + "epoch": 0.8181326279083749, + "flos": 22091167670400.0, + "grad_norm": 1.5153151299183365, + "language_loss": 0.77585095, + "learning_rate": 3.3689850308562527e-07, + "loss": 0.7974509, + "num_input_tokens_seen": 146876950, + "step": 6804, + "time_per_iteration": 2.631516218185425 + }, + { + "auxiliary_loss_clip": 0.01067275, + "auxiliary_loss_mlp": 0.01025424, + "balance_loss_clip": 0.90034437, + "balance_loss_mlp": 1.01801229, + "epoch": 0.818252870799014, + "flos": 15705747936000.0, + "grad_norm": 1.8666744554247794, + "language_loss": 0.77684045, + "learning_rate": 3.364659503871183e-07, + "loss": 0.7977674, + "num_input_tokens_seen": 146894885, + "step": 6805, + "time_per_iteration": 2.709083318710327 + }, + { + "auxiliary_loss_clip": 0.01089528, + "auxiliary_loss_mlp": 0.01025383, + "balance_loss_clip": 0.92970681, + "balance_loss_mlp": 1.01886821, + "epoch": 0.8183731136896532, + "flos": 18770687637120.0, + "grad_norm": 2.0601768870008503, + "language_loss": 0.84031814, + "learning_rate": 3.3603365004540417e-07, + "loss": 0.86146724, + "num_input_tokens_seen": 146913180, + "step": 6806, + "time_per_iteration": 3.5686793327331543 + }, + { + "auxiliary_loss_clip": 0.0116603, + "auxiliary_loss_mlp": 0.01025761, + "balance_loss_clip": 1.05079842, + "balance_loss_mlp": 1.01894808, + "epoch": 0.8184933565802922, + "flos": 26541792293760.0, + "grad_norm": 2.301469798000432, + "language_loss": 0.76620543, + "learning_rate": 3.356016021260624e-07, + "loss": 0.78812331, + "num_input_tokens_seen": 146933510, + "step": 6807, + "time_per_iteration": 2.8122506141662598 + }, + { + "auxiliary_loss_clip": 0.01143131, + "auxiliary_loss_mlp": 0.01025995, + "balance_loss_clip": 1.01217246, + "balance_loss_mlp": 1.01813889, + "epoch": 0.8186135994709313, + "flos": 17530117660800.0, + "grad_norm": 2.3486822951988087, + "language_loss": 0.65782863, + "learning_rate": 3.35169806694634e-07, + "loss": 0.67951989, + "num_input_tokens_seen": 146951760, + "step": 6808, + "time_per_iteration": 2.621760845184326 + }, + { + "auxiliary_loss_clip": 0.01004069, + "auxiliary_loss_mlp": 0.01012686, + "balance_loss_clip": 0.90269828, + "balance_loss_mlp": 1.01156545, + "epoch": 0.8187338423615703, + "flos": 63480300675840.0, + "grad_norm": 0.7169615056554829, + "language_loss": 0.60642815, + "learning_rate": 3.3473826381662186e-07, + "loss": 0.62659574, + "num_input_tokens_seen": 147022900, + "step": 6809, + "time_per_iteration": 3.410768747329712 + }, + { + "auxiliary_loss_clip": 0.01138736, + "auxiliary_loss_mlp": 0.01026696, + "balance_loss_clip": 1.0117383, + "balance_loss_mlp": 1.02013302, + "epoch": 0.8188540852522095, + "flos": 17529974006400.0, + "grad_norm": 1.7599893944444385, + "language_loss": 0.81521082, + "learning_rate": 3.3430697355749216e-07, + "loss": 0.83686513, + "num_input_tokens_seen": 147040590, + "step": 6810, + "time_per_iteration": 2.649962902069092 + }, + { + "auxiliary_loss_clip": 0.01065993, + "auxiliary_loss_mlp": 0.0103026, + "balance_loss_clip": 0.89339697, + "balance_loss_mlp": 1.02284503, + "epoch": 0.8189743281428485, + "flos": 14392530702720.0, + "grad_norm": 1.8023482436504348, + "language_loss": 0.75102061, + "learning_rate": 3.3387593598266907e-07, + "loss": 0.77198315, + "num_input_tokens_seen": 147057200, + "step": 6811, + "time_per_iteration": 2.8200669288635254 + }, + { + "auxiliary_loss_clip": 0.01087384, + "auxiliary_loss_mlp": 0.0102245, + "balance_loss_clip": 0.92934525, + "balance_loss_mlp": 1.01486576, + "epoch": 0.8190945710334876, + "flos": 25080479285760.0, + "grad_norm": 1.7603587892606032, + "language_loss": 0.78170109, + "learning_rate": 3.3344515115754225e-07, + "loss": 0.80279946, + "num_input_tokens_seen": 147076180, + "step": 6812, + "time_per_iteration": 2.794726848602295 + }, + { + "auxiliary_loss_clip": 0.01099609, + "auxiliary_loss_mlp": 0.01021903, + "balance_loss_clip": 0.93198562, + "balance_loss_mlp": 1.01456869, + "epoch": 0.8192148139241268, + "flos": 21507152440320.0, + "grad_norm": 2.1680964831395313, + "language_loss": 0.80133128, + "learning_rate": 3.33014619147461e-07, + "loss": 0.82254642, + "num_input_tokens_seen": 147094205, + "step": 6813, + "time_per_iteration": 2.7287650108337402 + }, + { + "auxiliary_loss_clip": 0.01120679, + "auxiliary_loss_mlp": 0.01030288, + "balance_loss_clip": 0.97588623, + "balance_loss_mlp": 1.02294171, + "epoch": 0.8193350568147658, + "flos": 23952166289280.0, + "grad_norm": 2.0875114580875596, + "language_loss": 0.71582115, + "learning_rate": 3.325843400177362e-07, + "loss": 0.73733079, + "num_input_tokens_seen": 147115545, + "step": 6814, + "time_per_iteration": 2.89404296875 + }, + { + "auxiliary_loss_clip": 0.01146224, + "auxiliary_loss_mlp": 0.00509273, + "balance_loss_clip": 1.0100317, + "balance_loss_mlp": 0.0, + "epoch": 0.8194552997054049, + "flos": 20559469962240.0, + "grad_norm": 2.632789380007816, + "language_loss": 0.73699069, + "learning_rate": 3.32154313833642e-07, + "loss": 0.75354564, + "num_input_tokens_seen": 147135700, + "step": 6815, + "time_per_iteration": 2.769861936569214 + }, + { + "auxiliary_loss_clip": 0.01167115, + "auxiliary_loss_mlp": 0.01024519, + "balance_loss_clip": 1.04844093, + "balance_loss_mlp": 1.01717556, + "epoch": 0.819575542596044, + "flos": 26031753123840.0, + "grad_norm": 2.2288782662467472, + "language_loss": 0.59710073, + "learning_rate": 3.3172454066041164e-07, + "loss": 0.61901712, + "num_input_tokens_seen": 147155205, + "step": 6816, + "time_per_iteration": 2.7002933025360107 + }, + { + "auxiliary_loss_clip": 0.01048614, + "auxiliary_loss_mlp": 0.00508252, + "balance_loss_clip": 0.8604753, + "balance_loss_mlp": 0.0, + "epoch": 0.8196957854866831, + "flos": 29096944220160.0, + "grad_norm": 1.8282599603904117, + "language_loss": 0.75901592, + "learning_rate": 3.3129502056324234e-07, + "loss": 0.77458453, + "num_input_tokens_seen": 147176570, + "step": 6817, + "time_per_iteration": 2.8051328659057617 + }, + { + "auxiliary_loss_clip": 0.00944056, + "auxiliary_loss_mlp": 0.00999022, + "balance_loss_clip": 0.79019678, + "balance_loss_mlp": 0.99783039, + "epoch": 0.8198160283773221, + "flos": 69033631898880.0, + "grad_norm": 0.7921859363800291, + "language_loss": 0.5978142, + "learning_rate": 3.3086575360729165e-07, + "loss": 0.61724496, + "num_input_tokens_seen": 147234105, + "step": 6818, + "time_per_iteration": 3.460684299468994 + }, + { + "auxiliary_loss_clip": 0.01115251, + "auxiliary_loss_mlp": 0.0103591, + "balance_loss_clip": 0.97094131, + "balance_loss_mlp": 1.02821207, + "epoch": 0.8199362712679613, + "flos": 16618058496000.0, + "grad_norm": 16.473907179039674, + "language_loss": 0.71348107, + "learning_rate": 3.3043673985767906e-07, + "loss": 0.73499262, + "num_input_tokens_seen": 147253170, + "step": 6819, + "time_per_iteration": 2.908585786819458 + }, + { + "auxiliary_loss_clip": 0.01083944, + "auxiliary_loss_mlp": 0.01027275, + "balance_loss_clip": 0.928617, + "balance_loss_mlp": 1.02027714, + "epoch": 0.8200565141586004, + "flos": 21757664868480.0, + "grad_norm": 2.305153014368976, + "language_loss": 0.77848721, + "learning_rate": 3.3000797937948564e-07, + "loss": 0.79959941, + "num_input_tokens_seen": 147271465, + "step": 6820, + "time_per_iteration": 2.695793628692627 + }, + { + "auxiliary_loss_clip": 0.01004008, + "auxiliary_loss_mlp": 0.00996772, + "balance_loss_clip": 0.899616, + "balance_loss_mlp": 0.99577636, + "epoch": 0.8201767570492394, + "flos": 69807112392960.0, + "grad_norm": 0.9452933865198787, + "language_loss": 0.650837, + "learning_rate": 3.295794722377534e-07, + "loss": 0.67084485, + "num_input_tokens_seen": 147335070, + "step": 6821, + "time_per_iteration": 3.290015935897827 + }, + { + "auxiliary_loss_clip": 0.01160144, + "auxiliary_loss_mlp": 0.01027419, + "balance_loss_clip": 1.04548776, + "balance_loss_mlp": 1.0206058, + "epoch": 0.8202969999398786, + "flos": 23111892455040.0, + "grad_norm": 1.825149155719632, + "language_loss": 0.80396551, + "learning_rate": 3.291512184974876e-07, + "loss": 0.82584107, + "num_input_tokens_seen": 147355460, + "step": 6822, + "time_per_iteration": 2.6255342960357666 + }, + { + "auxiliary_loss_clip": 0.01115113, + "auxiliary_loss_mlp": 0.01027842, + "balance_loss_clip": 0.96838015, + "balance_loss_mlp": 1.01992917, + "epoch": 0.8204172428305176, + "flos": 28220616109440.0, + "grad_norm": 1.9446237162044264, + "language_loss": 0.66440868, + "learning_rate": 3.2872321822365346e-07, + "loss": 0.68583822, + "num_input_tokens_seen": 147375675, + "step": 6823, + "time_per_iteration": 4.127599477767944 + }, + { + "auxiliary_loss_clip": 0.01142115, + "auxiliary_loss_mlp": 0.0102666, + "balance_loss_clip": 1.01256859, + "balance_loss_mlp": 1.01922178, + "epoch": 0.8205374857211567, + "flos": 20887011106560.0, + "grad_norm": 2.048723319167131, + "language_loss": 0.73306942, + "learning_rate": 3.282954714811783e-07, + "loss": 0.75475717, + "num_input_tokens_seen": 147394580, + "step": 6824, + "time_per_iteration": 4.59493613243103 + }, + { + "auxiliary_loss_clip": 0.01106619, + "auxiliary_loss_mlp": 0.01023633, + "balance_loss_clip": 0.96693826, + "balance_loss_mlp": 1.01628339, + "epoch": 0.8206577286117959, + "flos": 13152140294400.0, + "grad_norm": 2.4986400722292434, + "language_loss": 0.70696068, + "learning_rate": 3.2786797833495093e-07, + "loss": 0.7282632, + "num_input_tokens_seen": 147409935, + "step": 6825, + "time_per_iteration": 2.6824488639831543 + }, + { + "auxiliary_loss_clip": 0.01162913, + "auxiliary_loss_mlp": 0.01022271, + "balance_loss_clip": 1.04750395, + "balance_loss_mlp": 1.01532423, + "epoch": 0.8207779715024349, + "flos": 25265634917760.0, + "grad_norm": 1.6514832402948123, + "language_loss": 0.7270298, + "learning_rate": 3.274407388498213e-07, + "loss": 0.74888164, + "num_input_tokens_seen": 147428065, + "step": 6826, + "time_per_iteration": 2.6752378940582275 + }, + { + "auxiliary_loss_clip": 0.01089346, + "auxiliary_loss_mlp": 0.01023365, + "balance_loss_clip": 0.93024486, + "balance_loss_mlp": 1.01659656, + "epoch": 0.820898214393074, + "flos": 19610243199360.0, + "grad_norm": 1.7651305729743596, + "language_loss": 0.74328887, + "learning_rate": 3.270137530906021e-07, + "loss": 0.76441598, + "num_input_tokens_seen": 147447300, + "step": 6827, + "time_per_iteration": 2.788158893585205 + }, + { + "auxiliary_loss_clip": 0.01065112, + "auxiliary_loss_mlp": 0.01023314, + "balance_loss_clip": 0.89813709, + "balance_loss_mlp": 1.01621842, + "epoch": 0.8210184572837131, + "flos": 15596615439360.0, + "grad_norm": 2.1762730062369178, + "language_loss": 0.83503509, + "learning_rate": 3.265870211220665e-07, + "loss": 0.85591936, + "num_input_tokens_seen": 147465135, + "step": 6828, + "time_per_iteration": 2.7947237491607666 + }, + { + "auxiliary_loss_clip": 0.01094934, + "auxiliary_loss_mlp": 0.01031203, + "balance_loss_clip": 0.93738711, + "balance_loss_mlp": 1.02341819, + "epoch": 0.8211387001743522, + "flos": 20813932886400.0, + "grad_norm": 1.8316895533020179, + "language_loss": 0.81948262, + "learning_rate": 3.2616054300894934e-07, + "loss": 0.84074402, + "num_input_tokens_seen": 147484585, + "step": 6829, + "time_per_iteration": 2.8134329319000244 + }, + { + "auxiliary_loss_clip": 0.01110315, + "auxiliary_loss_mlp": 0.01027963, + "balance_loss_clip": 0.9723596, + "balance_loss_mlp": 1.02061963, + "epoch": 0.8212589430649913, + "flos": 27704579368320.0, + "grad_norm": 2.7035257277703715, + "language_loss": 0.84877193, + "learning_rate": 3.2573431881594693e-07, + "loss": 0.87015468, + "num_input_tokens_seen": 147504130, + "step": 6830, + "time_per_iteration": 2.754572629928589 + }, + { + "auxiliary_loss_clip": 0.01043292, + "auxiliary_loss_mlp": 0.01025048, + "balance_loss_clip": 0.85223264, + "balance_loss_mlp": 1.01773405, + "epoch": 0.8213791859556304, + "flos": 22455625017600.0, + "grad_norm": 2.051351562266758, + "language_loss": 0.65863961, + "learning_rate": 3.2530834860771663e-07, + "loss": 0.67932302, + "num_input_tokens_seen": 147523510, + "step": 6831, + "time_per_iteration": 2.8665683269500732 + }, + { + "auxiliary_loss_clip": 0.01142933, + "auxiliary_loss_mlp": 0.01029231, + "balance_loss_clip": 1.00974965, + "balance_loss_mlp": 1.02172709, + "epoch": 0.8214994288462695, + "flos": 16654471908480.0, + "grad_norm": 6.331658107183371, + "language_loss": 0.74503011, + "learning_rate": 3.248826324488794e-07, + "loss": 0.76675177, + "num_input_tokens_seen": 147540805, + "step": 6832, + "time_per_iteration": 3.5646626949310303 + }, + { + "auxiliary_loss_clip": 0.01164757, + "auxiliary_loss_mlp": 0.01025396, + "balance_loss_clip": 1.05083561, + "balance_loss_mlp": 1.01851439, + "epoch": 0.8216196717369085, + "flos": 25221787390080.0, + "grad_norm": 1.7169171098068567, + "language_loss": 0.87999135, + "learning_rate": 3.244571704040138e-07, + "loss": 0.9018929, + "num_input_tokens_seen": 147560965, + "step": 6833, + "time_per_iteration": 2.6429333686828613 + }, + { + "auxiliary_loss_clip": 0.01137526, + "auxiliary_loss_mlp": 0.01025913, + "balance_loss_clip": 1.00700903, + "balance_loss_mlp": 1.0184679, + "epoch": 0.8217399146275477, + "flos": 25371930240000.0, + "grad_norm": 1.832746715202199, + "language_loss": 0.7355963, + "learning_rate": 3.2403196253766374e-07, + "loss": 0.75723076, + "num_input_tokens_seen": 147580045, + "step": 6834, + "time_per_iteration": 2.748215675354004 + }, + { + "auxiliary_loss_clip": 0.01140749, + "auxiliary_loss_mlp": 0.01028101, + "balance_loss_clip": 1.01022351, + "balance_loss_mlp": 1.02013791, + "epoch": 0.8218601575181868, + "flos": 25629625388160.0, + "grad_norm": 6.05226131095438, + "language_loss": 0.79362619, + "learning_rate": 3.2360700891433254e-07, + "loss": 0.81531465, + "num_input_tokens_seen": 147599070, + "step": 6835, + "time_per_iteration": 2.6565732955932617 + }, + { + "auxiliary_loss_clip": 0.00981159, + "auxiliary_loss_mlp": 0.01000383, + "balance_loss_clip": 0.86270308, + "balance_loss_mlp": 0.99948943, + "epoch": 0.8219804004088258, + "flos": 67660229427840.0, + "grad_norm": 0.7890233291719684, + "language_loss": 0.57306254, + "learning_rate": 3.231823095984847e-07, + "loss": 0.59287798, + "num_input_tokens_seen": 147653710, + "step": 6836, + "time_per_iteration": 3.2207372188568115 + }, + { + "auxiliary_loss_clip": 0.01116299, + "auxiliary_loss_mlp": 0.01021854, + "balance_loss_clip": 0.97114468, + "balance_loss_mlp": 1.01515102, + "epoch": 0.822100643299465, + "flos": 19464266327040.0, + "grad_norm": 2.0604996747924553, + "language_loss": 0.76383686, + "learning_rate": 3.2275786465454814e-07, + "loss": 0.78521836, + "num_input_tokens_seen": 147670360, + "step": 6837, + "time_per_iteration": 2.681225061416626 + }, + { + "auxiliary_loss_clip": 0.01094592, + "auxiliary_loss_mlp": 0.01024421, + "balance_loss_clip": 0.93262333, + "balance_loss_mlp": 1.01732469, + "epoch": 0.822220886190104, + "flos": 24681368292480.0, + "grad_norm": 1.7847482529278977, + "language_loss": 0.76051641, + "learning_rate": 3.2233367414690917e-07, + "loss": 0.78170651, + "num_input_tokens_seen": 147692550, + "step": 6838, + "time_per_iteration": 2.9333066940307617 + }, + { + "auxiliary_loss_clip": 0.01089151, + "auxiliary_loss_mlp": 0.01029604, + "balance_loss_clip": 0.92902786, + "balance_loss_mlp": 1.02237725, + "epoch": 0.8223411290807431, + "flos": 27819062991360.0, + "grad_norm": 2.722042335916213, + "language_loss": 0.84771335, + "learning_rate": 3.219097381399183e-07, + "loss": 0.8689009, + "num_input_tokens_seen": 147709725, + "step": 6839, + "time_per_iteration": 2.782564163208008 + }, + { + "auxiliary_loss_clip": 0.01124243, + "auxiliary_loss_mlp": 0.01028965, + "balance_loss_clip": 0.9717589, + "balance_loss_mlp": 1.02165782, + "epoch": 0.8224613719713821, + "flos": 23218546913280.0, + "grad_norm": 2.001674342632184, + "language_loss": 0.81000876, + "learning_rate": 3.2148605669788584e-07, + "loss": 0.83154082, + "num_input_tokens_seen": 147729615, + "step": 6840, + "time_per_iteration": 2.749215602874756 + }, + { + "auxiliary_loss_clip": 0.01120623, + "auxiliary_loss_mlp": 0.01022317, + "balance_loss_clip": 0.97377652, + "balance_loss_mlp": 1.01531613, + "epoch": 0.8225816148620213, + "flos": 15706250726400.0, + "grad_norm": 2.674111806008082, + "language_loss": 0.76920056, + "learning_rate": 3.2106262988508405e-07, + "loss": 0.79062998, + "num_input_tokens_seen": 147747665, + "step": 6841, + "time_per_iteration": 2.7441813945770264 + }, + { + "auxiliary_loss_clip": 0.01120886, + "auxiliary_loss_mlp": 0.01020354, + "balance_loss_clip": 0.97148257, + "balance_loss_mlp": 1.01293314, + "epoch": 0.8227018577526604, + "flos": 18515111391360.0, + "grad_norm": 1.8546680247484706, + "language_loss": 0.74067187, + "learning_rate": 3.206394577657465e-07, + "loss": 0.76208425, + "num_input_tokens_seen": 147765445, + "step": 6842, + "time_per_iteration": 2.6303958892822266 + }, + { + "auxiliary_loss_clip": 0.01147189, + "auxiliary_loss_mlp": 0.01029015, + "balance_loss_clip": 1.01295519, + "balance_loss_mlp": 1.02150226, + "epoch": 0.8228221006432994, + "flos": 22236785406720.0, + "grad_norm": 3.6873494146612575, + "language_loss": 0.72989428, + "learning_rate": 3.202165404040675e-07, + "loss": 0.75165629, + "num_input_tokens_seen": 147783365, + "step": 6843, + "time_per_iteration": 2.6741762161254883 + }, + { + "auxiliary_loss_clip": 0.01051183, + "auxiliary_loss_mlp": 0.01029531, + "balance_loss_clip": 0.8594768, + "balance_loss_mlp": 1.02162147, + "epoch": 0.8229423435339386, + "flos": 24097532630400.0, + "grad_norm": 3.96066622159134, + "language_loss": 0.7450124, + "learning_rate": 3.1979387786420396e-07, + "loss": 0.76581955, + "num_input_tokens_seen": 147803605, + "step": 6844, + "time_per_iteration": 2.767547845840454 + }, + { + "auxiliary_loss_clip": 0.01117669, + "auxiliary_loss_mlp": 0.01027296, + "balance_loss_clip": 0.96815169, + "balance_loss_mlp": 1.02006602, + "epoch": 0.8230625864245776, + "flos": 23878549365120.0, + "grad_norm": 1.782460862766872, + "language_loss": 0.82222456, + "learning_rate": 3.1937147021027346e-07, + "loss": 0.84367418, + "num_input_tokens_seen": 147822060, + "step": 6845, + "time_per_iteration": 2.7183992862701416 + }, + { + "auxiliary_loss_clip": 0.01141741, + "auxiliary_loss_mlp": 0.01017444, + "balance_loss_clip": 1.01004338, + "balance_loss_mlp": 1.01072967, + "epoch": 0.8231828293152167, + "flos": 16581106379520.0, + "grad_norm": 2.5550734598284035, + "language_loss": 0.76712471, + "learning_rate": 3.189493175063547e-07, + "loss": 0.78871655, + "num_input_tokens_seen": 147839295, + "step": 6846, + "time_per_iteration": 2.595937967300415 + }, + { + "auxiliary_loss_clip": 0.01121697, + "auxiliary_loss_mlp": 0.01026114, + "balance_loss_clip": 0.97447264, + "balance_loss_mlp": 1.01869345, + "epoch": 0.8233030722058559, + "flos": 18880071528960.0, + "grad_norm": 1.812366740802438, + "language_loss": 0.67241895, + "learning_rate": 3.1852741981648776e-07, + "loss": 0.69389707, + "num_input_tokens_seen": 147857945, + "step": 6847, + "time_per_iteration": 2.695524215698242 + }, + { + "auxiliary_loss_clip": 0.01089393, + "auxiliary_loss_mlp": 0.01025359, + "balance_loss_clip": 0.93669045, + "balance_loss_mlp": 1.01786685, + "epoch": 0.8234233150964949, + "flos": 28439024757120.0, + "grad_norm": 3.318648199002163, + "language_loss": 0.69851232, + "learning_rate": 3.1810577720467404e-07, + "loss": 0.71965981, + "num_input_tokens_seen": 147879675, + "step": 6848, + "time_per_iteration": 2.7734732627868652 + }, + { + "auxiliary_loss_clip": 0.01121308, + "auxiliary_loss_mlp": 0.01023063, + "balance_loss_clip": 0.9727571, + "balance_loss_mlp": 1.01579666, + "epoch": 0.823543557987134, + "flos": 33765941577600.0, + "grad_norm": 1.5952957254214213, + "language_loss": 0.56579745, + "learning_rate": 3.176843897348769e-07, + "loss": 0.58724117, + "num_input_tokens_seen": 147902870, + "step": 6849, + "time_per_iteration": 3.992716073989868 + }, + { + "auxiliary_loss_clip": 0.01115172, + "auxiliary_loss_mlp": 0.01028178, + "balance_loss_clip": 0.97248268, + "balance_loss_mlp": 1.0208286, + "epoch": 0.8236638008777731, + "flos": 17092366611840.0, + "grad_norm": 2.3712843881979095, + "language_loss": 0.75689077, + "learning_rate": 3.1726325747102034e-07, + "loss": 0.77832425, + "num_input_tokens_seen": 147921245, + "step": 6850, + "time_per_iteration": 2.668447732925415 + }, + { + "auxiliary_loss_clip": 0.01063787, + "auxiliary_loss_mlp": 0.01026339, + "balance_loss_clip": 0.89075327, + "balance_loss_mlp": 1.01977944, + "epoch": 0.8237840437684122, + "flos": 61639982334720.0, + "grad_norm": 1.6150846615763879, + "language_loss": 0.64169633, + "learning_rate": 3.1684238047698974e-07, + "loss": 0.66259754, + "num_input_tokens_seen": 147949515, + "step": 6851, + "time_per_iteration": 4.176229238510132 + }, + { + "auxiliary_loss_clip": 0.01121625, + "auxiliary_loss_mlp": 0.01024203, + "balance_loss_clip": 0.97231948, + "balance_loss_mlp": 1.0166924, + "epoch": 0.8239042866590512, + "flos": 27309023821440.0, + "grad_norm": 1.7791475639424525, + "language_loss": 0.52735913, + "learning_rate": 3.1642175881663155e-07, + "loss": 0.5488174, + "num_input_tokens_seen": 147969245, + "step": 6852, + "time_per_iteration": 2.7121551036834717 + }, + { + "auxiliary_loss_clip": 0.01163496, + "auxiliary_loss_mlp": 0.010275, + "balance_loss_clip": 1.04755282, + "balance_loss_mlp": 1.01998639, + "epoch": 0.8240245295496904, + "flos": 21726351187200.0, + "grad_norm": 2.6831652046187764, + "language_loss": 0.83510482, + "learning_rate": 3.160013925537537e-07, + "loss": 0.85701478, + "num_input_tokens_seen": 147990080, + "step": 6853, + "time_per_iteration": 2.6427299976348877 + }, + { + "auxiliary_loss_clip": 0.01097205, + "auxiliary_loss_mlp": 0.01021261, + "balance_loss_clip": 0.93284237, + "balance_loss_mlp": 1.01401901, + "epoch": 0.8241447724403295, + "flos": 20009318279040.0, + "grad_norm": 1.8487792848903564, + "language_loss": 0.75845665, + "learning_rate": 3.155812817521266e-07, + "loss": 0.77964139, + "num_input_tokens_seen": 148010455, + "step": 6854, + "time_per_iteration": 2.88718318939209 + }, + { + "auxiliary_loss_clip": 0.01122975, + "auxiliary_loss_mlp": 0.01027708, + "balance_loss_clip": 0.97532797, + "balance_loss_mlp": 1.02027559, + "epoch": 0.8242650153309685, + "flos": 22272983337600.0, + "grad_norm": 1.91052393975879, + "language_loss": 0.77858526, + "learning_rate": 3.151614264754787e-07, + "loss": 0.80009204, + "num_input_tokens_seen": 148028400, + "step": 6855, + "time_per_iteration": 2.879272222518921 + }, + { + "auxiliary_loss_clip": 0.01164896, + "auxiliary_loss_mlp": 0.0102896, + "balance_loss_clip": 1.04646778, + "balance_loss_mlp": 1.02147055, + "epoch": 0.8243852582216077, + "flos": 22309971367680.0, + "grad_norm": 2.0665144186522943, + "language_loss": 0.79606843, + "learning_rate": 3.147418267875035e-07, + "loss": 0.81800693, + "num_input_tokens_seen": 148046530, + "step": 6856, + "time_per_iteration": 2.8201541900634766 + }, + { + "auxiliary_loss_clip": 0.01043001, + "auxiliary_loss_mlp": 0.00509145, + "balance_loss_clip": 0.85275728, + "balance_loss_mlp": 0.0, + "epoch": 0.8245055011122467, + "flos": 24645421756800.0, + "grad_norm": 2.1175286635577644, + "language_loss": 0.65532601, + "learning_rate": 3.1432248275185315e-07, + "loss": 0.67084742, + "num_input_tokens_seen": 148067040, + "step": 6857, + "time_per_iteration": 2.918597459793091 + }, + { + "auxiliary_loss_clip": 0.01140343, + "auxiliary_loss_mlp": 0.01023837, + "balance_loss_clip": 1.01072633, + "balance_loss_mlp": 1.01659489, + "epoch": 0.8246257440028858, + "flos": 17487275713920.0, + "grad_norm": 2.2681711975558496, + "language_loss": 0.76885581, + "learning_rate": 3.139033944321412e-07, + "loss": 0.79049766, + "num_input_tokens_seen": 148084400, + "step": 6858, + "time_per_iteration": 2.7251875400543213 + }, + { + "auxiliary_loss_clip": 0.0114367, + "auxiliary_loss_mlp": 0.01029678, + "balance_loss_clip": 1.00989676, + "balance_loss_mlp": 1.02257895, + "epoch": 0.824745986893525, + "flos": 25010130499200.0, + "grad_norm": 1.6414015031843754, + "language_loss": 0.78830683, + "learning_rate": 3.1348456189194507e-07, + "loss": 0.8100403, + "num_input_tokens_seen": 148104860, + "step": 6859, + "time_per_iteration": 3.6167521476745605 + }, + { + "auxiliary_loss_clip": 0.01083836, + "auxiliary_loss_mlp": 0.01026601, + "balance_loss_clip": 0.92885363, + "balance_loss_mlp": 1.01959157, + "epoch": 0.824866229784164, + "flos": 18772698798720.0, + "grad_norm": 1.6542806562815713, + "language_loss": 0.83008736, + "learning_rate": 3.1306598519479876e-07, + "loss": 0.8511917, + "num_input_tokens_seen": 148124680, + "step": 6860, + "time_per_iteration": 2.770766258239746 + }, + { + "auxiliary_loss_clip": 0.01115383, + "auxiliary_loss_mlp": 0.01024017, + "balance_loss_clip": 0.9715929, + "balance_loss_mlp": 1.01672769, + "epoch": 0.8249864726748031, + "flos": 23842171866240.0, + "grad_norm": 2.436907469476039, + "language_loss": 0.78481179, + "learning_rate": 3.1264766440420177e-07, + "loss": 0.80620587, + "num_input_tokens_seen": 148147150, + "step": 6861, + "time_per_iteration": 2.8841869831085205 + }, + { + "auxiliary_loss_clip": 0.01137659, + "auxiliary_loss_mlp": 0.0102622, + "balance_loss_clip": 1.00942993, + "balance_loss_mlp": 1.01937151, + "epoch": 0.8251067155654422, + "flos": 20303103617280.0, + "grad_norm": 1.875451392157817, + "language_loss": 0.69175315, + "learning_rate": 3.122295995836124e-07, + "loss": 0.71339202, + "num_input_tokens_seen": 148167020, + "step": 6862, + "time_per_iteration": 2.7364308834075928 + }, + { + "auxiliary_loss_clip": 0.01142672, + "auxiliary_loss_mlp": 0.01024985, + "balance_loss_clip": 1.00721383, + "balance_loss_mlp": 1.01728094, + "epoch": 0.8252269584560813, + "flos": 25009699536000.0, + "grad_norm": 1.7761978260268858, + "language_loss": 0.77498406, + "learning_rate": 3.118117907964508e-07, + "loss": 0.7966606, + "num_input_tokens_seen": 148188965, + "step": 6863, + "time_per_iteration": 2.8474864959716797 + }, + { + "auxiliary_loss_clip": 0.01099938, + "auxiliary_loss_mlp": 0.01025764, + "balance_loss_clip": 0.93191409, + "balance_loss_mlp": 1.01867056, + "epoch": 0.8253472013467203, + "flos": 17128564542720.0, + "grad_norm": 2.0025809025568617, + "language_loss": 0.80236328, + "learning_rate": 3.1139423810609856e-07, + "loss": 0.82362032, + "num_input_tokens_seen": 148205660, + "step": 6864, + "time_per_iteration": 2.9360358715057373 + }, + { + "auxiliary_loss_clip": 0.01161778, + "auxiliary_loss_mlp": 0.01019947, + "balance_loss_clip": 1.04507053, + "balance_loss_mlp": 1.01243067, + "epoch": 0.8254674442373595, + "flos": 22414794232320.0, + "grad_norm": 2.180517921044557, + "language_loss": 0.75167549, + "learning_rate": 3.1097694157589714e-07, + "loss": 0.77349275, + "num_input_tokens_seen": 148225545, + "step": 6865, + "time_per_iteration": 2.7722175121307373 + }, + { + "auxiliary_loss_clip": 0.01139441, + "auxiliary_loss_mlp": 0.01028955, + "balance_loss_clip": 1.01199341, + "balance_loss_mlp": 1.02192795, + "epoch": 0.8255876871279986, + "flos": 24786765774720.0, + "grad_norm": 2.8135003155013885, + "language_loss": 0.76760894, + "learning_rate": 3.105599012691511e-07, + "loss": 0.78929281, + "num_input_tokens_seen": 148243975, + "step": 6866, + "time_per_iteration": 2.7883408069610596 + }, + { + "auxiliary_loss_clip": 0.01138793, + "auxiliary_loss_mlp": 0.01022101, + "balance_loss_clip": 1.00975513, + "balance_loss_mlp": 1.01529121, + "epoch": 0.8257079300186376, + "flos": 27455431656960.0, + "grad_norm": 1.6756757836430403, + "language_loss": 0.82578218, + "learning_rate": 3.101431172491249e-07, + "loss": 0.84739119, + "num_input_tokens_seen": 148265520, + "step": 6867, + "time_per_iteration": 2.8840646743774414 + }, + { + "auxiliary_loss_clip": 0.01097905, + "auxiliary_loss_mlp": 0.00509987, + "balance_loss_clip": 0.93114835, + "balance_loss_mlp": 0.0, + "epoch": 0.8258281729092768, + "flos": 16471866142080.0, + "grad_norm": 2.1789950232838775, + "language_loss": 0.72010243, + "learning_rate": 3.097265895790444e-07, + "loss": 0.73618132, + "num_input_tokens_seen": 148283730, + "step": 6868, + "time_per_iteration": 2.7676374912261963 + }, + { + "auxiliary_loss_clip": 0.01095035, + "auxiliary_loss_mlp": 0.01029045, + "balance_loss_clip": 0.93362355, + "balance_loss_mlp": 1.02198148, + "epoch": 0.8259484157999158, + "flos": 21433822824960.0, + "grad_norm": 1.9677463557579695, + "language_loss": 0.82848191, + "learning_rate": 3.093103183220962e-07, + "loss": 0.84972274, + "num_input_tokens_seen": 148303775, + "step": 6869, + "time_per_iteration": 2.9081857204437256 + }, + { + "auxiliary_loss_clip": 0.01041984, + "auxiliary_loss_mlp": 0.01001084, + "balance_loss_clip": 0.97327113, + "balance_loss_mlp": 1.00015378, + "epoch": 0.8260686586905549, + "flos": 58322342453760.0, + "grad_norm": 0.8164938270341826, + "language_loss": 0.59389371, + "learning_rate": 3.0889430354142796e-07, + "loss": 0.61432439, + "num_input_tokens_seen": 148365285, + "step": 6870, + "time_per_iteration": 3.378241777420044 + }, + { + "auxiliary_loss_clip": 0.01097277, + "auxiliary_loss_mlp": 0.01026493, + "balance_loss_clip": 0.93119895, + "balance_loss_mlp": 1.01984119, + "epoch": 0.826188901581194, + "flos": 27527288814720.0, + "grad_norm": 2.9943908664029686, + "language_loss": 0.69945228, + "learning_rate": 3.084785453001497e-07, + "loss": 0.72068995, + "num_input_tokens_seen": 148386200, + "step": 6871, + "time_per_iteration": 2.8263978958129883 + }, + { + "auxiliary_loss_clip": 0.01117704, + "auxiliary_loss_mlp": 0.00509394, + "balance_loss_clip": 0.97369123, + "balance_loss_mlp": 0.0, + "epoch": 0.8263091444718331, + "flos": 23696051339520.0, + "grad_norm": 2.050048027658988, + "language_loss": 0.81783026, + "learning_rate": 3.080630436613314e-07, + "loss": 0.83410126, + "num_input_tokens_seen": 148403970, + "step": 6872, + "time_per_iteration": 2.86712646484375 + }, + { + "auxiliary_loss_clip": 0.01133231, + "auxiliary_loss_mlp": 0.01023004, + "balance_loss_clip": 1.00695777, + "balance_loss_mlp": 1.0156014, + "epoch": 0.8264293873624722, + "flos": 17165157523200.0, + "grad_norm": 2.0556570930151836, + "language_loss": 0.86215377, + "learning_rate": 3.076477986880039e-07, + "loss": 0.88371611, + "num_input_tokens_seen": 148421765, + "step": 6873, + "time_per_iteration": 2.630086898803711 + }, + { + "auxiliary_loss_clip": 0.01119775, + "auxiliary_loss_mlp": 0.01023328, + "balance_loss_clip": 0.97327304, + "balance_loss_mlp": 1.01643777, + "epoch": 0.8265496302531112, + "flos": 24098645952000.0, + "grad_norm": 2.575730213336296, + "language_loss": 0.69516003, + "learning_rate": 3.0723281044315986e-07, + "loss": 0.716591, + "num_input_tokens_seen": 148443720, + "step": 6874, + "time_per_iteration": 2.752511978149414 + }, + { + "auxiliary_loss_clip": 0.01160308, + "auxiliary_loss_mlp": 0.0102566, + "balance_loss_clip": 1.04528522, + "balance_loss_mlp": 1.01889193, + "epoch": 0.8266698731437504, + "flos": 14099894599680.0, + "grad_norm": 2.083483899926083, + "language_loss": 0.76694137, + "learning_rate": 3.068180789897521e-07, + "loss": 0.78880101, + "num_input_tokens_seen": 148462130, + "step": 6875, + "time_per_iteration": 4.439730405807495 + }, + { + "auxiliary_loss_clip": 0.01142743, + "auxiliary_loss_mlp": 0.0102628, + "balance_loss_clip": 1.00796533, + "balance_loss_mlp": 1.01878738, + "epoch": 0.8267901160343895, + "flos": 30777563715840.0, + "grad_norm": 2.822684864742229, + "language_loss": 0.81634516, + "learning_rate": 3.064036043906966e-07, + "loss": 0.8380354, + "num_input_tokens_seen": 148485570, + "step": 6876, + "time_per_iteration": 2.782397747039795 + }, + { + "auxiliary_loss_clip": 0.01103288, + "auxiliary_loss_mlp": 0.01028238, + "balance_loss_clip": 0.93353117, + "balance_loss_mlp": 1.02052498, + "epoch": 0.8269103589250285, + "flos": 40624915242240.0, + "grad_norm": 2.1795864618064598, + "language_loss": 0.67861247, + "learning_rate": 3.059893867088668e-07, + "loss": 0.69992769, + "num_input_tokens_seen": 148509715, + "step": 6877, + "time_per_iteration": 3.9622385501861572 + }, + { + "auxiliary_loss_clip": 0.01141085, + "auxiliary_loss_mlp": 0.01022575, + "balance_loss_clip": 1.01076865, + "balance_loss_mlp": 1.0156374, + "epoch": 0.8270306018156677, + "flos": 30263645877120.0, + "grad_norm": 2.9321623783699544, + "language_loss": 0.6673761, + "learning_rate": 3.055754260071004e-07, + "loss": 0.68901265, + "num_input_tokens_seen": 148532010, + "step": 6878, + "time_per_iteration": 2.783869743347168 + }, + { + "auxiliary_loss_clip": 0.01141962, + "auxiliary_loss_mlp": 0.01020722, + "balance_loss_clip": 1.010746, + "balance_loss_mlp": 1.01397169, + "epoch": 0.8271508447063067, + "flos": 25226599812480.0, + "grad_norm": 3.119156602053112, + "language_loss": 0.73532498, + "learning_rate": 3.051617223481948e-07, + "loss": 0.75695181, + "num_input_tokens_seen": 148553330, + "step": 6879, + "time_per_iteration": 2.6647167205810547 + }, + { + "auxiliary_loss_clip": 0.01107943, + "auxiliary_loss_mlp": 0.01026952, + "balance_loss_clip": 0.93458974, + "balance_loss_mlp": 1.01933467, + "epoch": 0.8272710875969458, + "flos": 17566602900480.0, + "grad_norm": 2.3182638530664503, + "language_loss": 0.74802303, + "learning_rate": 3.047482757949078e-07, + "loss": 0.76937193, + "num_input_tokens_seen": 148570960, + "step": 6880, + "time_per_iteration": 2.6991772651672363 + }, + { + "auxiliary_loss_clip": 0.01092123, + "auxiliary_loss_mlp": 0.00509055, + "balance_loss_clip": 0.93204319, + "balance_loss_mlp": 0.0, + "epoch": 0.827391330487585, + "flos": 19755465886080.0, + "grad_norm": 3.2939925743698213, + "language_loss": 0.86056161, + "learning_rate": 3.043350864099605e-07, + "loss": 0.87657338, + "num_input_tokens_seen": 148589520, + "step": 6881, + "time_per_iteration": 2.787224769592285 + }, + { + "auxiliary_loss_clip": 0.01142424, + "auxiliary_loss_mlp": 0.0102484, + "balance_loss_clip": 1.00731778, + "balance_loss_mlp": 1.01805735, + "epoch": 0.827511573378224, + "flos": 16835174254080.0, + "grad_norm": 2.1699740908726546, + "language_loss": 0.81086564, + "learning_rate": 3.039221542560315e-07, + "loss": 0.83253831, + "num_input_tokens_seen": 148606085, + "step": 6882, + "time_per_iteration": 2.603848457336426 + }, + { + "auxiliary_loss_clip": 0.01139567, + "auxiliary_loss_mlp": 0.01021543, + "balance_loss_clip": 1.00979662, + "balance_loss_mlp": 1.01463509, + "epoch": 0.8276318162688631, + "flos": 18369242259840.0, + "grad_norm": 2.1647336675955984, + "language_loss": 0.73911786, + "learning_rate": 3.0350947939576356e-07, + "loss": 0.76072896, + "num_input_tokens_seen": 148625240, + "step": 6883, + "time_per_iteration": 2.6637094020843506 + }, + { + "auxiliary_loss_clip": 0.01146865, + "auxiliary_loss_mlp": 0.0102696, + "balance_loss_clip": 1.01079309, + "balance_loss_mlp": 1.02017713, + "epoch": 0.8277520591595022, + "flos": 19352691705600.0, + "grad_norm": 1.890880301664068, + "language_loss": 0.72313881, + "learning_rate": 3.0309706189175876e-07, + "loss": 0.7448771, + "num_input_tokens_seen": 148645075, + "step": 6884, + "time_per_iteration": 3.711489200592041 + }, + { + "auxiliary_loss_clip": 0.01023419, + "auxiliary_loss_mlp": 0.01006647, + "balance_loss_clip": 0.9356916, + "balance_loss_mlp": 1.00565743, + "epoch": 0.8278723020501413, + "flos": 67918858329600.0, + "grad_norm": 0.7730762599534461, + "language_loss": 0.57288945, + "learning_rate": 3.0268490180658045e-07, + "loss": 0.59319007, + "num_input_tokens_seen": 148707855, + "step": 6885, + "time_per_iteration": 3.231106996536255 + }, + { + "auxiliary_loss_clip": 0.01168913, + "auxiliary_loss_mlp": 0.01030444, + "balance_loss_clip": 1.05112123, + "balance_loss_mlp": 1.02269554, + "epoch": 0.8279925449407803, + "flos": 18185738653440.0, + "grad_norm": 2.112940410133204, + "language_loss": 0.79493344, + "learning_rate": 3.0227299920275305e-07, + "loss": 0.81692696, + "num_input_tokens_seen": 148724170, + "step": 6886, + "time_per_iteration": 2.721996784210205 + }, + { + "auxiliary_loss_clip": 0.01101218, + "auxiliary_loss_mlp": 0.01026609, + "balance_loss_clip": 0.93625599, + "balance_loss_mlp": 1.01903892, + "epoch": 0.8281127878314195, + "flos": 20631434860800.0, + "grad_norm": 3.6439626811384884, + "language_loss": 0.85864115, + "learning_rate": 3.018613541427613e-07, + "loss": 0.87991941, + "num_input_tokens_seen": 148743690, + "step": 6887, + "time_per_iteration": 2.880134105682373 + }, + { + "auxiliary_loss_clip": 0.01164631, + "auxiliary_loss_mlp": 0.01022196, + "balance_loss_clip": 1.04787946, + "balance_loss_mlp": 1.01510346, + "epoch": 0.8282330307220586, + "flos": 18004282122240.0, + "grad_norm": 1.741813412093608, + "language_loss": 0.73697734, + "learning_rate": 3.0144996668905243e-07, + "loss": 0.75884557, + "num_input_tokens_seen": 148761070, + "step": 6888, + "time_per_iteration": 2.5703887939453125 + }, + { + "auxiliary_loss_clip": 0.01049489, + "auxiliary_loss_mlp": 0.00509566, + "balance_loss_clip": 0.85365969, + "balance_loss_mlp": 0.0, + "epoch": 0.8283532736126976, + "flos": 20084120352000.0, + "grad_norm": 2.7580794444588186, + "language_loss": 0.81847316, + "learning_rate": 3.010388369040331e-07, + "loss": 0.83406377, + "num_input_tokens_seen": 148779730, + "step": 6889, + "time_per_iteration": 2.8076915740966797 + }, + { + "auxiliary_loss_clip": 0.01144844, + "auxiliary_loss_mlp": 0.01028834, + "balance_loss_clip": 1.01088166, + "balance_loss_mlp": 1.0220933, + "epoch": 0.8284735165033368, + "flos": 31868421805440.0, + "grad_norm": 1.4356458674010688, + "language_loss": 0.82773769, + "learning_rate": 3.0062796485007156e-07, + "loss": 0.84947449, + "num_input_tokens_seen": 148800670, + "step": 6890, + "time_per_iteration": 2.73831844329834 + }, + { + "auxiliary_loss_clip": 0.01166932, + "auxiliary_loss_mlp": 0.00509639, + "balance_loss_clip": 1.04857349, + "balance_loss_mlp": 0.0, + "epoch": 0.8285937593939758, + "flos": 26651319840000.0, + "grad_norm": 2.2939479227996933, + "language_loss": 0.6520797, + "learning_rate": 3.002173505894965e-07, + "loss": 0.66884542, + "num_input_tokens_seen": 148819820, + "step": 6891, + "time_per_iteration": 2.624572515487671 + }, + { + "auxiliary_loss_clip": 0.01143726, + "auxiliary_loss_mlp": 0.01026482, + "balance_loss_clip": 1.00807297, + "balance_loss_mlp": 1.01908207, + "epoch": 0.8287140022846149, + "flos": 20193683811840.0, + "grad_norm": 3.5458651480335903, + "language_loss": 0.6269539, + "learning_rate": 2.998069941845973e-07, + "loss": 0.64865595, + "num_input_tokens_seen": 148838890, + "step": 6892, + "time_per_iteration": 2.628312826156616 + }, + { + "auxiliary_loss_clip": 0.01060244, + "auxiliary_loss_mlp": 0.01000936, + "balance_loss_clip": 1.00901926, + "balance_loss_mlp": 0.99996483, + "epoch": 0.8288342451752541, + "flos": 70755980019840.0, + "grad_norm": 0.7194656848135629, + "language_loss": 0.57479048, + "learning_rate": 2.993968956976258e-07, + "loss": 0.59540224, + "num_input_tokens_seen": 148906635, + "step": 6893, + "time_per_iteration": 3.2941884994506836 + }, + { + "auxiliary_loss_clip": 0.01171353, + "auxiliary_loss_mlp": 0.01031136, + "balance_loss_clip": 1.05094135, + "balance_loss_mlp": 1.02341747, + "epoch": 0.8289544880658931, + "flos": 24572235795840.0, + "grad_norm": 1.8354303441192135, + "language_loss": 0.7059688, + "learning_rate": 2.9898705519079313e-07, + "loss": 0.72799373, + "num_input_tokens_seen": 148925740, + "step": 6894, + "time_per_iteration": 2.8865017890930176 + }, + { + "auxiliary_loss_clip": 0.01110777, + "auxiliary_loss_mlp": 0.01027566, + "balance_loss_clip": 0.96971184, + "balance_loss_mlp": 1.02030969, + "epoch": 0.8290747309565322, + "flos": 22273378387200.0, + "grad_norm": 4.231410350778104, + "language_loss": 0.74838358, + "learning_rate": 2.985774727262715e-07, + "loss": 0.76976705, + "num_input_tokens_seen": 148944585, + "step": 6895, + "time_per_iteration": 2.7328696250915527 + }, + { + "auxiliary_loss_clip": 0.01162872, + "auxiliary_loss_mlp": 0.01026076, + "balance_loss_clip": 1.04715824, + "balance_loss_mlp": 1.01942158, + "epoch": 0.8291949738471713, + "flos": 23255570856960.0, + "grad_norm": 1.654716111981231, + "language_loss": 0.81450588, + "learning_rate": 2.981681483661949e-07, + "loss": 0.83639538, + "num_input_tokens_seen": 148964170, + "step": 6896, + "time_per_iteration": 2.600236654281616 + }, + { + "auxiliary_loss_clip": 0.01145509, + "auxiliary_loss_mlp": 0.0102988, + "balance_loss_clip": 1.01363325, + "balance_loss_mlp": 1.02257562, + "epoch": 0.8293152167378104, + "flos": 52555768185600.0, + "grad_norm": 2.8769417522250844, + "language_loss": 0.712084, + "learning_rate": 2.9775908217265633e-07, + "loss": 0.73383784, + "num_input_tokens_seen": 148989405, + "step": 6897, + "time_per_iteration": 2.863689661026001 + }, + { + "auxiliary_loss_clip": 0.00961955, + "auxiliary_loss_mlp": 0.01004614, + "balance_loss_clip": 0.82665145, + "balance_loss_mlp": 1.00348139, + "epoch": 0.8294354596284494, + "flos": 63356156294400.0, + "grad_norm": 0.8283923011675877, + "language_loss": 0.50321698, + "learning_rate": 2.9735027420771253e-07, + "loss": 0.52288264, + "num_input_tokens_seen": 149049740, + "step": 6898, + "time_per_iteration": 3.2642040252685547 + }, + { + "auxiliary_loss_clip": 0.01113845, + "auxiliary_loss_mlp": 0.01024615, + "balance_loss_clip": 0.97380084, + "balance_loss_mlp": 1.01793039, + "epoch": 0.8295557025190886, + "flos": 24827021942400.0, + "grad_norm": 1.9489578610307978, + "language_loss": 0.71253645, + "learning_rate": 2.969417245333774e-07, + "loss": 0.73392105, + "num_input_tokens_seen": 149069120, + "step": 6899, + "time_per_iteration": 2.7934653759002686 + }, + { + "auxiliary_loss_clip": 0.01092551, + "auxiliary_loss_mlp": 0.01023229, + "balance_loss_clip": 0.93505585, + "balance_loss_mlp": 1.01619005, + "epoch": 0.8296759454097277, + "flos": 25118580637440.0, + "grad_norm": 2.0438037798167215, + "language_loss": 0.78125107, + "learning_rate": 2.9653343321162915e-07, + "loss": 0.80240887, + "num_input_tokens_seen": 149088630, + "step": 6900, + "time_per_iteration": 2.773620128631592 + }, + { + "auxiliary_loss_clip": 0.01098773, + "auxiliary_loss_mlp": 0.0101995, + "balance_loss_clip": 0.93599957, + "balance_loss_mlp": 1.01256549, + "epoch": 0.8297961883003667, + "flos": 24132581326080.0, + "grad_norm": 1.9549504767827428, + "language_loss": 0.65028095, + "learning_rate": 2.9612540030440446e-07, + "loss": 0.67146814, + "num_input_tokens_seen": 149109175, + "step": 6901, + "time_per_iteration": 4.602871417999268 + }, + { + "auxiliary_loss_clip": 0.01021326, + "auxiliary_loss_mlp": 0.01003455, + "balance_loss_clip": 0.93505675, + "balance_loss_mlp": 1.00251341, + "epoch": 0.8299164311910058, + "flos": 67446561375360.0, + "grad_norm": 0.849741507251521, + "language_loss": 0.64089131, + "learning_rate": 2.9571762587360206e-07, + "loss": 0.66113913, + "num_input_tokens_seen": 149165560, + "step": 6902, + "time_per_iteration": 3.1689445972442627 + }, + { + "auxiliary_loss_clip": 0.01068835, + "auxiliary_loss_mlp": 0.01020916, + "balance_loss_clip": 0.88902926, + "balance_loss_mlp": 1.01354289, + "epoch": 0.8300366740816449, + "flos": 25228682801280.0, + "grad_norm": 1.683201796999036, + "language_loss": 0.74137259, + "learning_rate": 2.953101099810806e-07, + "loss": 0.76227009, + "num_input_tokens_seen": 149185165, + "step": 6903, + "time_per_iteration": 2.7696962356567383 + }, + { + "auxiliary_loss_clip": 0.0113739, + "auxiliary_loss_mlp": 0.01027437, + "balance_loss_clip": 1.01115692, + "balance_loss_mlp": 1.01987648, + "epoch": 0.830156916972284, + "flos": 18041018757120.0, + "grad_norm": 2.056229046703029, + "language_loss": 0.82486129, + "learning_rate": 2.9490285268865965e-07, + "loss": 0.84650958, + "num_input_tokens_seen": 149202655, + "step": 6904, + "time_per_iteration": 3.6173646450042725 + }, + { + "auxiliary_loss_clip": 0.01146865, + "auxiliary_loss_mlp": 0.01025324, + "balance_loss_clip": 1.01234794, + "balance_loss_mlp": 1.01780152, + "epoch": 0.830277159862923, + "flos": 26322485806080.0, + "grad_norm": 3.1921903094128035, + "language_loss": 0.7981205, + "learning_rate": 2.9449585405812085e-07, + "loss": 0.8198424, + "num_input_tokens_seen": 149220035, + "step": 6905, + "time_per_iteration": 2.714747667312622 + }, + { + "auxiliary_loss_clip": 0.01099123, + "auxiliary_loss_mlp": 0.01025146, + "balance_loss_clip": 0.93356383, + "balance_loss_mlp": 1.01816583, + "epoch": 0.8303974027535622, + "flos": 19938861751680.0, + "grad_norm": 1.8103917783235417, + "language_loss": 0.74009842, + "learning_rate": 2.940891141512043e-07, + "loss": 0.76134109, + "num_input_tokens_seen": 149238055, + "step": 6906, + "time_per_iteration": 2.8056983947753906 + }, + { + "auxiliary_loss_clip": 0.01118165, + "auxiliary_loss_mlp": 0.01027068, + "balance_loss_clip": 0.97128022, + "balance_loss_mlp": 1.01943922, + "epoch": 0.8305176456442013, + "flos": 17165552572800.0, + "grad_norm": 2.233546506690344, + "language_loss": 0.72124958, + "learning_rate": 2.9368263302961385e-07, + "loss": 0.74270189, + "num_input_tokens_seen": 149256755, + "step": 6907, + "time_per_iteration": 2.67940354347229 + }, + { + "auxiliary_loss_clip": 0.01047189, + "auxiliary_loss_mlp": 0.01026073, + "balance_loss_clip": 0.85411894, + "balance_loss_mlp": 1.01870251, + "epoch": 0.8306378885348403, + "flos": 25627614226560.0, + "grad_norm": 2.0476504529292985, + "language_loss": 0.80091178, + "learning_rate": 2.9327641075501075e-07, + "loss": 0.82164448, + "num_input_tokens_seen": 149275745, + "step": 6908, + "time_per_iteration": 2.760429859161377 + }, + { + "auxiliary_loss_clip": 0.01109253, + "auxiliary_loss_mlp": 0.01026286, + "balance_loss_clip": 0.96735775, + "balance_loss_mlp": 1.01901126, + "epoch": 0.8307581314254795, + "flos": 33947864985600.0, + "grad_norm": 2.616887247804757, + "language_loss": 0.66962862, + "learning_rate": 2.9287044738901866e-07, + "loss": 0.69098401, + "num_input_tokens_seen": 149293730, + "step": 6909, + "time_per_iteration": 2.794330358505249 + }, + { + "auxiliary_loss_clip": 0.01144249, + "auxiliary_loss_mlp": 0.00509238, + "balance_loss_clip": 1.01067209, + "balance_loss_mlp": 0.0, + "epoch": 0.8308783743161186, + "flos": 17562724231680.0, + "grad_norm": 2.265761759576843, + "language_loss": 0.90857291, + "learning_rate": 2.9246474299322274e-07, + "loss": 0.92510778, + "num_input_tokens_seen": 149309290, + "step": 6910, + "time_per_iteration": 2.664583921432495 + }, + { + "auxiliary_loss_clip": 0.00999628, + "auxiliary_loss_mlp": 0.00999276, + "balance_loss_clip": 0.89872414, + "balance_loss_mlp": 0.99825633, + "epoch": 0.8309986172067576, + "flos": 69412885649280.0, + "grad_norm": 0.9021906688446711, + "language_loss": 0.63118923, + "learning_rate": 2.920592976291678e-07, + "loss": 0.65117836, + "num_input_tokens_seen": 149366620, + "step": 6911, + "time_per_iteration": 4.106064319610596 + }, + { + "auxiliary_loss_clip": 0.01139014, + "auxiliary_loss_mlp": 0.01025105, + "balance_loss_clip": 1.00885427, + "balance_loss_mlp": 1.01776743, + "epoch": 0.8311188600973968, + "flos": 22309755886080.0, + "grad_norm": 1.8991201023751987, + "language_loss": 0.80282271, + "learning_rate": 2.916541113583595e-07, + "loss": 0.8244639, + "num_input_tokens_seen": 149385120, + "step": 6912, + "time_per_iteration": 2.826300859451294 + }, + { + "auxiliary_loss_clip": 0.01102968, + "auxiliary_loss_mlp": 0.01025286, + "balance_loss_clip": 0.93792498, + "balance_loss_mlp": 1.01800847, + "epoch": 0.8312391029880358, + "flos": 18770077105920.0, + "grad_norm": 2.1465786882883724, + "language_loss": 0.66539788, + "learning_rate": 2.912491842422642e-07, + "loss": 0.68668044, + "num_input_tokens_seen": 149402825, + "step": 6913, + "time_per_iteration": 2.713177442550659 + }, + { + "auxiliary_loss_clip": 0.01142352, + "auxiliary_loss_mlp": 0.01026175, + "balance_loss_clip": 1.00976014, + "balance_loss_mlp": 1.01921332, + "epoch": 0.8313593458786749, + "flos": 20376648714240.0, + "grad_norm": 1.6068381773261504, + "language_loss": 0.70846289, + "learning_rate": 2.9084451634230857e-07, + "loss": 0.7301482, + "num_input_tokens_seen": 149422125, + "step": 6914, + "time_per_iteration": 2.7971560955047607 + }, + { + "auxiliary_loss_clip": 0.01093817, + "auxiliary_loss_mlp": 0.01026284, + "balance_loss_clip": 0.93138659, + "balance_loss_mlp": 1.01900053, + "epoch": 0.831479588769314, + "flos": 32124069878400.0, + "grad_norm": 2.283389044567569, + "language_loss": 0.70979774, + "learning_rate": 2.9044010771988125e-07, + "loss": 0.73099875, + "num_input_tokens_seen": 149441940, + "step": 6915, + "time_per_iteration": 2.839708089828491 + }, + { + "auxiliary_loss_clip": 0.01114944, + "auxiliary_loss_mlp": 0.01024994, + "balance_loss_clip": 0.97159266, + "balance_loss_mlp": 1.01744759, + "epoch": 0.8315998316599531, + "flos": 45185929338240.0, + "grad_norm": 1.8261606193641833, + "language_loss": 0.72484982, + "learning_rate": 2.900359584363303e-07, + "loss": 0.7462492, + "num_input_tokens_seen": 149465045, + "step": 6916, + "time_per_iteration": 2.8824069499969482 + }, + { + "auxiliary_loss_clip": 0.01067827, + "auxiliary_loss_mlp": 0.01039813, + "balance_loss_clip": 0.89613116, + "balance_loss_mlp": 1.03173614, + "epoch": 0.8317200745505922, + "flos": 18363747479040.0, + "grad_norm": 5.150560287899092, + "language_loss": 0.84500754, + "learning_rate": 2.8963206855296494e-07, + "loss": 0.86608398, + "num_input_tokens_seen": 149481285, + "step": 6917, + "time_per_iteration": 2.7029483318328857 + }, + { + "auxiliary_loss_clip": 0.01143651, + "auxiliary_loss_mlp": 0.01029377, + "balance_loss_clip": 1.0100944, + "balance_loss_mlp": 1.02207565, + "epoch": 0.8318403174412313, + "flos": 24206557386240.0, + "grad_norm": 2.2549851632808204, + "language_loss": 0.77308047, + "learning_rate": 2.892284381310548e-07, + "loss": 0.79481077, + "num_input_tokens_seen": 149502700, + "step": 6918, + "time_per_iteration": 2.6938698291778564 + }, + { + "auxiliary_loss_clip": 0.01116409, + "auxiliary_loss_mlp": 0.01034412, + "balance_loss_clip": 0.97082692, + "balance_loss_mlp": 1.026968, + "epoch": 0.8319605603318704, + "flos": 22418780641920.0, + "grad_norm": 2.4870154336673576, + "language_loss": 0.71949744, + "learning_rate": 2.888250672318302e-07, + "loss": 0.7410056, + "num_input_tokens_seen": 149520100, + "step": 6919, + "time_per_iteration": 2.808969736099243 + }, + { + "auxiliary_loss_clip": 0.0116889, + "auxiliary_loss_mlp": 0.01027616, + "balance_loss_clip": 1.05092239, + "balance_loss_mlp": 1.02026033, + "epoch": 0.8320808032225094, + "flos": 37414501459200.0, + "grad_norm": 1.95956326363623, + "language_loss": 0.68711793, + "learning_rate": 2.884219559164831e-07, + "loss": 0.70908296, + "num_input_tokens_seen": 149543245, + "step": 6920, + "time_per_iteration": 2.7639517784118652 + }, + { + "auxiliary_loss_clip": 0.01141511, + "auxiliary_loss_mlp": 0.0102584, + "balance_loss_clip": 1.01129591, + "balance_loss_mlp": 1.01813269, + "epoch": 0.8322010461131486, + "flos": 12787395638400.0, + "grad_norm": 1.7947601556758643, + "language_loss": 0.81199592, + "learning_rate": 2.880191042461635e-07, + "loss": 0.83366942, + "num_input_tokens_seen": 149559185, + "step": 6921, + "time_per_iteration": 2.673065662384033 + }, + { + "auxiliary_loss_clip": 0.01078263, + "auxiliary_loss_mlp": 0.0102828, + "balance_loss_clip": 0.89475536, + "balance_loss_mlp": 1.0216372, + "epoch": 0.8323212890037877, + "flos": 15815455050240.0, + "grad_norm": 2.4968248193364317, + "language_loss": 0.80438691, + "learning_rate": 2.876165122819849e-07, + "loss": 0.82545233, + "num_input_tokens_seen": 149577165, + "step": 6922, + "time_per_iteration": 2.9252798557281494 + }, + { + "auxiliary_loss_clip": 0.011635, + "auxiliary_loss_mlp": 0.01023437, + "balance_loss_clip": 1.04926956, + "balance_loss_mlp": 1.01631391, + "epoch": 0.8324415318944267, + "flos": 21719276208000.0, + "grad_norm": 1.576132346412924, + "language_loss": 0.79462588, + "learning_rate": 2.872141800850201e-07, + "loss": 0.81649518, + "num_input_tokens_seen": 149594340, + "step": 6923, + "time_per_iteration": 2.828197717666626 + }, + { + "auxiliary_loss_clip": 0.01163337, + "auxiliary_loss_mlp": 0.01028665, + "balance_loss_clip": 1.047804, + "balance_loss_mlp": 1.0211966, + "epoch": 0.8325617747850659, + "flos": 34198700636160.0, + "grad_norm": 2.0853829509541786, + "language_loss": 0.73523748, + "learning_rate": 2.868121077163024e-07, + "loss": 0.75715744, + "num_input_tokens_seen": 149613895, + "step": 6924, + "time_per_iteration": 2.756345272064209 + }, + { + "auxiliary_loss_clip": 0.01142034, + "auxiliary_loss_mlp": 0.01024384, + "balance_loss_clip": 1.00745678, + "balance_loss_mlp": 1.01741052, + "epoch": 0.8326820176757049, + "flos": 18369457741440.0, + "grad_norm": 1.8533884611910005, + "language_loss": 0.72531855, + "learning_rate": 2.864102952368257e-07, + "loss": 0.74698275, + "num_input_tokens_seen": 149631820, + "step": 6925, + "time_per_iteration": 2.6761300563812256 + }, + { + "auxiliary_loss_clip": 0.01058858, + "auxiliary_loss_mlp": 0.0102388, + "balance_loss_clip": 0.88921148, + "balance_loss_mlp": 1.01650143, + "epoch": 0.832802260566344, + "flos": 35991325716480.0, + "grad_norm": 1.273481208268698, + "language_loss": 0.59314561, + "learning_rate": 2.860087427075444e-07, + "loss": 0.61397302, + "num_input_tokens_seen": 149656070, + "step": 6926, + "time_per_iteration": 2.951859474182129 + }, + { + "auxiliary_loss_clip": 0.0111386, + "auxiliary_loss_mlp": 0.01027779, + "balance_loss_clip": 0.97057855, + "balance_loss_mlp": 1.02068591, + "epoch": 0.8329225034569832, + "flos": 14244434928000.0, + "grad_norm": 2.352463652709498, + "language_loss": 0.86246216, + "learning_rate": 2.856074501893744e-07, + "loss": 0.88387859, + "num_input_tokens_seen": 149671270, + "step": 6927, + "time_per_iteration": 3.5369482040405273 + }, + { + "auxiliary_loss_clip": 0.011482, + "auxiliary_loss_mlp": 0.01024935, + "balance_loss_clip": 1.0142746, + "balance_loss_mlp": 1.01805091, + "epoch": 0.8330427463476222, + "flos": 18077468083200.0, + "grad_norm": 2.750018964567304, + "language_loss": 0.81630504, + "learning_rate": 2.8520641774319054e-07, + "loss": 0.83803648, + "num_input_tokens_seen": 149689360, + "step": 6928, + "time_per_iteration": 2.6450133323669434 + }, + { + "auxiliary_loss_clip": 0.01116567, + "auxiliary_loss_mlp": 0.01026067, + "balance_loss_clip": 0.96705973, + "balance_loss_mlp": 1.01868236, + "epoch": 0.8331629892382613, + "flos": 18040839189120.0, + "grad_norm": 1.9102740864906842, + "language_loss": 0.75940382, + "learning_rate": 2.848056454298309e-07, + "loss": 0.78083014, + "num_input_tokens_seen": 149706685, + "step": 6929, + "time_per_iteration": 3.6310667991638184 + }, + { + "auxiliary_loss_clip": 0.0111864, + "auxiliary_loss_mlp": 0.01026242, + "balance_loss_clip": 0.97372812, + "balance_loss_mlp": 1.01861835, + "epoch": 0.8332832321289004, + "flos": 17457398576640.0, + "grad_norm": 3.4967057811555473, + "language_loss": 0.65282249, + "learning_rate": 2.844051333100905e-07, + "loss": 0.67427135, + "num_input_tokens_seen": 149724230, + "step": 6930, + "time_per_iteration": 2.695464849472046 + }, + { + "auxiliary_loss_clip": 0.01118132, + "auxiliary_loss_mlp": 0.01026081, + "balance_loss_clip": 0.97300756, + "balance_loss_mlp": 1.02004051, + "epoch": 0.8334034750195395, + "flos": 15084852416640.0, + "grad_norm": 3.1455922297860166, + "language_loss": 0.84168279, + "learning_rate": 2.840048814447269e-07, + "loss": 0.86312491, + "num_input_tokens_seen": 149742395, + "step": 6931, + "time_per_iteration": 2.7165160179138184 + }, + { + "auxiliary_loss_clip": 0.01110382, + "auxiliary_loss_mlp": 0.01029065, + "balance_loss_clip": 0.96945167, + "balance_loss_mlp": 1.02196062, + "epoch": 0.8335237179101785, + "flos": 19427170556160.0, + "grad_norm": 2.0772707619146527, + "language_loss": 0.73715067, + "learning_rate": 2.836048898944587e-07, + "loss": 0.75854516, + "num_input_tokens_seen": 149760820, + "step": 6932, + "time_per_iteration": 2.6924972534179688 + }, + { + "auxiliary_loss_clip": 0.01116012, + "auxiliary_loss_mlp": 0.01020889, + "balance_loss_clip": 0.9687897, + "balance_loss_mlp": 1.01343799, + "epoch": 0.8336439608008177, + "flos": 21762046327680.0, + "grad_norm": 2.2623225207234867, + "language_loss": 0.72664446, + "learning_rate": 2.832051587199642e-07, + "loss": 0.7480135, + "num_input_tokens_seen": 149778075, + "step": 6933, + "time_per_iteration": 2.667112112045288 + }, + { + "auxiliary_loss_clip": 0.01042856, + "auxiliary_loss_mlp": 0.00999192, + "balance_loss_clip": 0.97295755, + "balance_loss_mlp": 0.99823803, + "epoch": 0.8337642036914568, + "flos": 59702783990400.0, + "grad_norm": 0.8058297329214709, + "language_loss": 0.57708859, + "learning_rate": 2.828056879818821e-07, + "loss": 0.59750915, + "num_input_tokens_seen": 149837150, + "step": 6934, + "time_per_iteration": 3.1572651863098145 + }, + { + "auxiliary_loss_clip": 0.01091267, + "auxiliary_loss_mlp": 0.01025605, + "balance_loss_clip": 0.92809701, + "balance_loss_mlp": 1.01867557, + "epoch": 0.8338844465820958, + "flos": 27162185022720.0, + "grad_norm": 1.9783040312967024, + "language_loss": 0.83351851, + "learning_rate": 2.824064777408117e-07, + "loss": 0.85468721, + "num_input_tokens_seen": 149856940, + "step": 6935, + "time_per_iteration": 2.7442963123321533 + }, + { + "auxiliary_loss_clip": 0.01141109, + "auxiliary_loss_mlp": 0.01030164, + "balance_loss_clip": 1.01137125, + "balance_loss_mlp": 1.02286863, + "epoch": 0.8340046894727349, + "flos": 30481264425600.0, + "grad_norm": 1.639280751386022, + "language_loss": 0.75753534, + "learning_rate": 2.8200752805731263e-07, + "loss": 0.77924812, + "num_input_tokens_seen": 149879930, + "step": 6936, + "time_per_iteration": 3.608529806137085 + }, + { + "auxiliary_loss_clip": 0.01139876, + "auxiliary_loss_mlp": 0.01032749, + "balance_loss_clip": 1.01059961, + "balance_loss_mlp": 1.02559388, + "epoch": 0.834124932363374, + "flos": 27126166659840.0, + "grad_norm": 1.4526958637118033, + "language_loss": 0.80992353, + "learning_rate": 2.8160883899190625e-07, + "loss": 0.83164978, + "num_input_tokens_seen": 149903200, + "step": 6937, + "time_per_iteration": 2.8658900260925293 + }, + { + "auxiliary_loss_clip": 0.01088681, + "auxiliary_loss_mlp": 0.01022031, + "balance_loss_clip": 0.93486953, + "balance_loss_mlp": 1.01474714, + "epoch": 0.8342451752540131, + "flos": 24569865498240.0, + "grad_norm": 2.1426146696407207, + "language_loss": 0.73347348, + "learning_rate": 2.8121041060507234e-07, + "loss": 0.75458062, + "num_input_tokens_seen": 149922230, + "step": 6938, + "time_per_iteration": 2.798255681991577 + }, + { + "auxiliary_loss_clip": 0.01145515, + "auxiliary_loss_mlp": 0.01025171, + "balance_loss_clip": 1.01053739, + "balance_loss_mlp": 1.01814103, + "epoch": 0.8343654181446521, + "flos": 26615085995520.0, + "grad_norm": 1.8172699310558478, + "language_loss": 0.711007, + "learning_rate": 2.808122429572528e-07, + "loss": 0.73271388, + "num_input_tokens_seen": 149942435, + "step": 6939, + "time_per_iteration": 2.6902737617492676 + }, + { + "auxiliary_loss_clip": 0.01100095, + "auxiliary_loss_mlp": 0.01025064, + "balance_loss_clip": 0.93257439, + "balance_loss_mlp": 1.01844788, + "epoch": 0.8344856610352913, + "flos": 20777268078720.0, + "grad_norm": 3.0782011209690836, + "language_loss": 0.75917876, + "learning_rate": 2.804143361088489e-07, + "loss": 0.78043038, + "num_input_tokens_seen": 149961615, + "step": 6940, + "time_per_iteration": 2.7555699348449707 + }, + { + "auxiliary_loss_clip": 0.01112947, + "auxiliary_loss_mlp": 0.0102312, + "balance_loss_clip": 0.97124648, + "balance_loss_mlp": 1.01552653, + "epoch": 0.8346059039259304, + "flos": 26095960684800.0, + "grad_norm": 2.595804572256351, + "language_loss": 0.77899855, + "learning_rate": 2.8001669012022277e-07, + "loss": 0.80035925, + "num_input_tokens_seen": 149979585, + "step": 6941, + "time_per_iteration": 2.701629877090454 + }, + { + "auxiliary_loss_clip": 0.01141278, + "auxiliary_loss_mlp": 0.01025869, + "balance_loss_clip": 1.01269937, + "balance_loss_mlp": 1.01840687, + "epoch": 0.8347261468165694, + "flos": 29027708755200.0, + "grad_norm": 1.5336001818173364, + "language_loss": 0.69581831, + "learning_rate": 2.7961930505169795e-07, + "loss": 0.71748978, + "num_input_tokens_seen": 150003830, + "step": 6942, + "time_per_iteration": 2.723872661590576 + }, + { + "auxiliary_loss_clip": 0.0114364, + "auxiliary_loss_mlp": 0.00509516, + "balance_loss_clip": 1.01034498, + "balance_loss_mlp": 0.0, + "epoch": 0.8348463897072086, + "flos": 26396461866240.0, + "grad_norm": 1.8397086402762182, + "language_loss": 0.76588732, + "learning_rate": 2.792221809635558e-07, + "loss": 0.78241897, + "num_input_tokens_seen": 150024460, + "step": 6943, + "time_per_iteration": 2.7008399963378906 + }, + { + "auxiliary_loss_clip": 0.01027723, + "auxiliary_loss_mlp": 0.0102767, + "balance_loss_clip": 0.82002378, + "balance_loss_mlp": 1.02023172, + "epoch": 0.8349666325978476, + "flos": 23367720096000.0, + "grad_norm": 2.399453601513501, + "language_loss": 0.74894917, + "learning_rate": 2.788253179160411e-07, + "loss": 0.76950312, + "num_input_tokens_seen": 150045620, + "step": 6944, + "time_per_iteration": 2.933816909790039 + }, + { + "auxiliary_loss_clip": 0.01117436, + "auxiliary_loss_mlp": 0.0103121, + "balance_loss_clip": 0.97078145, + "balance_loss_mlp": 1.0244627, + "epoch": 0.8350868754884867, + "flos": 12896528135040.0, + "grad_norm": 2.087291957995637, + "language_loss": 0.65437293, + "learning_rate": 2.7842871596935725e-07, + "loss": 0.67585939, + "num_input_tokens_seen": 150064135, + "step": 6945, + "time_per_iteration": 3.2544984817504883 + }, + { + "auxiliary_loss_clip": 0.01147086, + "auxiliary_loss_mlp": 0.01026061, + "balance_loss_clip": 1.01000667, + "balance_loss_mlp": 1.01856256, + "epoch": 0.8352071183791259, + "flos": 26505522535680.0, + "grad_norm": 1.5638549541780493, + "language_loss": 0.69327468, + "learning_rate": 2.780323751836682e-07, + "loss": 0.71500611, + "num_input_tokens_seen": 150085350, + "step": 6946, + "time_per_iteration": 2.6581764221191406 + }, + { + "auxiliary_loss_clip": 0.01116776, + "auxiliary_loss_mlp": 0.00509239, + "balance_loss_clip": 0.96940899, + "balance_loss_mlp": 0.0, + "epoch": 0.8353273612697649, + "flos": 20668063754880.0, + "grad_norm": 2.2171203613728463, + "language_loss": 0.78445911, + "learning_rate": 2.7763629561909876e-07, + "loss": 0.80071926, + "num_input_tokens_seen": 150106180, + "step": 6947, + "time_per_iteration": 2.7890472412109375 + }, + { + "auxiliary_loss_clip": 0.01161791, + "auxiliary_loss_mlp": 0.01024183, + "balance_loss_clip": 1.0463028, + "balance_loss_mlp": 1.01681554, + "epoch": 0.835447604160404, + "flos": 19754137082880.0, + "grad_norm": 2.5134821988796117, + "language_loss": 0.77038872, + "learning_rate": 2.772404773357335e-07, + "loss": 0.79224849, + "num_input_tokens_seen": 150125585, + "step": 6948, + "time_per_iteration": 2.657285213470459 + }, + { + "auxiliary_loss_clip": 0.0108998, + "auxiliary_loss_mlp": 0.01025415, + "balance_loss_clip": 0.93289983, + "balance_loss_mlp": 1.01779759, + "epoch": 0.8355678470510431, + "flos": 23435842239360.0, + "grad_norm": 1.9447247290885785, + "language_loss": 0.78170824, + "learning_rate": 2.7684492039361853e-07, + "loss": 0.80286217, + "num_input_tokens_seen": 150144810, + "step": 6949, + "time_per_iteration": 2.8061020374298096 + }, + { + "auxiliary_loss_clip": 0.01167673, + "auxiliary_loss_mlp": 0.01025716, + "balance_loss_clip": 1.05000424, + "balance_loss_mlp": 1.01884961, + "epoch": 0.8356880899416822, + "flos": 21214588164480.0, + "grad_norm": 1.5970889553691892, + "language_loss": 0.83745676, + "learning_rate": 2.764496248527586e-07, + "loss": 0.85939068, + "num_input_tokens_seen": 150163785, + "step": 6950, + "time_per_iteration": 2.6183693408966064 + }, + { + "auxiliary_loss_clip": 0.01103057, + "auxiliary_loss_mlp": 0.01028867, + "balance_loss_clip": 0.93270332, + "balance_loss_mlp": 1.02135944, + "epoch": 0.8358083328323213, + "flos": 28037543466240.0, + "grad_norm": 2.4458449916513842, + "language_loss": 0.78855193, + "learning_rate": 2.760545907731211e-07, + "loss": 0.8098712, + "num_input_tokens_seen": 150184360, + "step": 6951, + "time_per_iteration": 2.8195791244506836 + }, + { + "auxiliary_loss_clip": 0.01141692, + "auxiliary_loss_mlp": 0.01021892, + "balance_loss_clip": 1.00799286, + "balance_loss_mlp": 1.01439965, + "epoch": 0.8359285757229604, + "flos": 27783655159680.0, + "grad_norm": 1.6830269823715953, + "language_loss": 0.68037236, + "learning_rate": 2.75659818214631e-07, + "loss": 0.70200825, + "num_input_tokens_seen": 150205465, + "step": 6952, + "time_per_iteration": 2.7537307739257812 + }, + { + "auxiliary_loss_clip": 0.01120156, + "auxiliary_loss_mlp": 0.01020673, + "balance_loss_clip": 0.97143042, + "balance_loss_mlp": 1.01402116, + "epoch": 0.8360488186135995, + "flos": 21435115714560.0, + "grad_norm": 1.8567039235014702, + "language_loss": 0.77941316, + "learning_rate": 2.752653072371749e-07, + "loss": 0.80082142, + "num_input_tokens_seen": 150224900, + "step": 6953, + "time_per_iteration": 4.8323588371276855 + }, + { + "auxiliary_loss_clip": 0.01092266, + "auxiliary_loss_mlp": 0.01024367, + "balance_loss_clip": 0.93471444, + "balance_loss_mlp": 1.0174855, + "epoch": 0.8361690615042385, + "flos": 27632327160960.0, + "grad_norm": 1.6903913354119342, + "language_loss": 0.74844402, + "learning_rate": 2.7487105790060105e-07, + "loss": 0.7696104, + "num_input_tokens_seen": 150244310, + "step": 6954, + "time_per_iteration": 2.953988552093506 + }, + { + "auxiliary_loss_clip": 0.0114357, + "auxiliary_loss_mlp": 0.01023196, + "balance_loss_clip": 1.00960112, + "balance_loss_mlp": 1.01654148, + "epoch": 0.8362893043948777, + "flos": 39202529598720.0, + "grad_norm": 1.9198779840671307, + "language_loss": 0.69452107, + "learning_rate": 2.7447707026471587e-07, + "loss": 0.71618867, + "num_input_tokens_seen": 150267285, + "step": 6955, + "time_per_iteration": 2.81756591796875 + }, + { + "auxiliary_loss_clip": 0.01093528, + "auxiliary_loss_mlp": 0.01026349, + "balance_loss_clip": 0.9316082, + "balance_loss_mlp": 1.01964331, + "epoch": 0.8364095472855168, + "flos": 24785329230720.0, + "grad_norm": 1.9998273793504961, + "language_loss": 0.79531598, + "learning_rate": 2.740833443892874e-07, + "loss": 0.81651473, + "num_input_tokens_seen": 150285455, + "step": 6956, + "time_per_iteration": 3.714895009994507 + }, + { + "auxiliary_loss_clip": 0.01116936, + "auxiliary_loss_mlp": 0.01017904, + "balance_loss_clip": 0.9713341, + "balance_loss_mlp": 1.01149344, + "epoch": 0.8365297901761558, + "flos": 22743412784640.0, + "grad_norm": 1.7580391158048159, + "language_loss": 0.79766738, + "learning_rate": 2.7368988033404327e-07, + "loss": 0.8190158, + "num_input_tokens_seen": 150302970, + "step": 6957, + "time_per_iteration": 2.682225465774536 + }, + { + "auxiliary_loss_clip": 0.01096321, + "auxiliary_loss_mlp": 0.01028967, + "balance_loss_clip": 0.93446982, + "balance_loss_mlp": 1.02220464, + "epoch": 0.836650033066795, + "flos": 28396003242240.0, + "grad_norm": 2.425130757846698, + "language_loss": 0.84367698, + "learning_rate": 2.732966781586712e-07, + "loss": 0.86492991, + "num_input_tokens_seen": 150322715, + "step": 6958, + "time_per_iteration": 2.7343099117279053 + }, + { + "auxiliary_loss_clip": 0.01134553, + "auxiliary_loss_mlp": 0.0102495, + "balance_loss_clip": 1.00673413, + "balance_loss_mlp": 1.01766396, + "epoch": 0.836770275957434, + "flos": 22236857233920.0, + "grad_norm": 1.5675709304151992, + "language_loss": 0.66660303, + "learning_rate": 2.729037379228205e-07, + "loss": 0.68819803, + "num_input_tokens_seen": 150342900, + "step": 6959, + "time_per_iteration": 2.718914747238159 + }, + { + "auxiliary_loss_clip": 0.01119771, + "auxiliary_loss_mlp": 0.01031832, + "balance_loss_clip": 0.97547084, + "balance_loss_mlp": 1.02485847, + "epoch": 0.8368905188480731, + "flos": 22491930689280.0, + "grad_norm": 1.4667163108058536, + "language_loss": 0.80466753, + "learning_rate": 2.725110596860998e-07, + "loss": 0.82618356, + "num_input_tokens_seen": 150363580, + "step": 6960, + "time_per_iteration": 2.811491012573242 + }, + { + "auxiliary_loss_clip": 0.01069416, + "auxiliary_loss_mlp": 0.01023983, + "balance_loss_clip": 0.89568096, + "balance_loss_mlp": 1.01694989, + "epoch": 0.8370107617387123, + "flos": 13370405287680.0, + "grad_norm": 1.8121454024108024, + "language_loss": 0.69935286, + "learning_rate": 2.7211864350807776e-07, + "loss": 0.72028679, + "num_input_tokens_seen": 150381780, + "step": 6961, + "time_per_iteration": 2.7106785774230957 + }, + { + "auxiliary_loss_clip": 0.01162587, + "auxiliary_loss_mlp": 0.01027006, + "balance_loss_clip": 1.04671979, + "balance_loss_mlp": 1.01931691, + "epoch": 0.8371310046293513, + "flos": 25261289372160.0, + "grad_norm": 1.6566317952206184, + "language_loss": 0.7381736, + "learning_rate": 2.717264894482836e-07, + "loss": 0.76006949, + "num_input_tokens_seen": 150402120, + "step": 6962, + "time_per_iteration": 2.661343574523926 + }, + { + "auxiliary_loss_clip": 0.01145692, + "auxiliary_loss_mlp": 0.01027357, + "balance_loss_clip": 1.01243639, + "balance_loss_mlp": 1.01994228, + "epoch": 0.8372512475199904, + "flos": 19792705311360.0, + "grad_norm": 1.89048810154233, + "language_loss": 0.81172097, + "learning_rate": 2.7133459756620646e-07, + "loss": 0.83345145, + "num_input_tokens_seen": 150419315, + "step": 6963, + "time_per_iteration": 3.5500566959381104 + }, + { + "auxiliary_loss_clip": 0.01136175, + "auxiliary_loss_mlp": 0.01030569, + "balance_loss_clip": 1.00835311, + "balance_loss_mlp": 1.02308607, + "epoch": 0.8373714904106295, + "flos": 19391224020480.0, + "grad_norm": 1.6447000074085867, + "language_loss": 0.73888433, + "learning_rate": 2.7094296792129733e-07, + "loss": 0.76055169, + "num_input_tokens_seen": 150438915, + "step": 6964, + "time_per_iteration": 2.672319173812866 + }, + { + "auxiliary_loss_clip": 0.01141715, + "auxiliary_loss_mlp": 0.01027136, + "balance_loss_clip": 1.01043844, + "balance_loss_mlp": 1.02031779, + "epoch": 0.8374917333012686, + "flos": 14975935401600.0, + "grad_norm": 1.8209593733827367, + "language_loss": 0.75718135, + "learning_rate": 2.7055160057296424e-07, + "loss": 0.77886987, + "num_input_tokens_seen": 150456155, + "step": 6965, + "time_per_iteration": 2.640765905380249 + }, + { + "auxiliary_loss_clip": 0.01095148, + "auxiliary_loss_mlp": 0.01023904, + "balance_loss_clip": 0.93255734, + "balance_loss_mlp": 1.01687992, + "epoch": 0.8376119761919076, + "flos": 30331839847680.0, + "grad_norm": 1.567510333734376, + "language_loss": 0.72318995, + "learning_rate": 2.7016049558057896e-07, + "loss": 0.74438041, + "num_input_tokens_seen": 150478115, + "step": 6966, + "time_per_iteration": 2.839409351348877 + }, + { + "auxiliary_loss_clip": 0.0114158, + "auxiliary_loss_mlp": 0.0102202, + "balance_loss_clip": 1.01121187, + "balance_loss_mlp": 1.01531732, + "epoch": 0.8377322190825467, + "flos": 29423336129280.0, + "grad_norm": 1.6368323866993755, + "language_loss": 0.70476085, + "learning_rate": 2.6976965300347074e-07, + "loss": 0.7263968, + "num_input_tokens_seen": 150500725, + "step": 6967, + "time_per_iteration": 2.715069055557251 + }, + { + "auxiliary_loss_clip": 0.01112334, + "auxiliary_loss_mlp": 0.01025247, + "balance_loss_clip": 0.96886373, + "balance_loss_mlp": 1.01832104, + "epoch": 0.8378524619731859, + "flos": 26687086807680.0, + "grad_norm": 3.0089368333867412, + "language_loss": 0.69544232, + "learning_rate": 2.693790729009309e-07, + "loss": 0.71681809, + "num_input_tokens_seen": 150522335, + "step": 6968, + "time_per_iteration": 2.7129175662994385 + }, + { + "auxiliary_loss_clip": 0.01119512, + "auxiliary_loss_mlp": 0.01023938, + "balance_loss_clip": 0.97365534, + "balance_loss_mlp": 1.01717544, + "epoch": 0.8379727048638249, + "flos": 20703866636160.0, + "grad_norm": 1.7356066614592054, + "language_loss": 0.8845619, + "learning_rate": 2.6898875533220946e-07, + "loss": 0.90599632, + "num_input_tokens_seen": 150541640, + "step": 6969, + "time_per_iteration": 2.6736931800842285 + }, + { + "auxiliary_loss_clip": 0.01160145, + "auxiliary_loss_mlp": 0.01024499, + "balance_loss_clip": 1.04765391, + "balance_loss_mlp": 1.0175612, + "epoch": 0.838092947754464, + "flos": 20084084438400.0, + "grad_norm": 1.7413664306284593, + "language_loss": 0.82008886, + "learning_rate": 2.685987003565171e-07, + "loss": 0.84193528, + "num_input_tokens_seen": 150559680, + "step": 6970, + "time_per_iteration": 2.618933916091919 + }, + { + "auxiliary_loss_clip": 0.01089044, + "auxiliary_loss_mlp": 0.01028316, + "balance_loss_clip": 0.93725485, + "balance_loss_mlp": 1.02097845, + "epoch": 0.8382131906451031, + "flos": 18113270964480.0, + "grad_norm": 2.158844931478616, + "language_loss": 0.74847811, + "learning_rate": 2.6820890803302566e-07, + "loss": 0.76965177, + "num_input_tokens_seen": 150575205, + "step": 6971, + "time_per_iteration": 2.720384120941162 + }, + { + "auxiliary_loss_clip": 0.0112004, + "auxiliary_loss_mlp": 0.01022344, + "balance_loss_clip": 0.97661495, + "balance_loss_mlp": 1.01532865, + "epoch": 0.8383334335357422, + "flos": 17092653920640.0, + "grad_norm": 2.4098422818252936, + "language_loss": 0.81807017, + "learning_rate": 2.6781937842086557e-07, + "loss": 0.83949399, + "num_input_tokens_seen": 150593995, + "step": 6972, + "time_per_iteration": 2.6275105476379395 + }, + { + "auxiliary_loss_clip": 0.01142424, + "auxiliary_loss_mlp": 0.01024212, + "balance_loss_clip": 1.00961494, + "balance_loss_mlp": 1.01654696, + "epoch": 0.8384536764263812, + "flos": 20704728562560.0, + "grad_norm": 1.850327641814525, + "language_loss": 0.67549175, + "learning_rate": 2.6743011157912933e-07, + "loss": 0.6971581, + "num_input_tokens_seen": 150613715, + "step": 6973, + "time_per_iteration": 2.6483359336853027 + }, + { + "auxiliary_loss_clip": 0.01068624, + "auxiliary_loss_mlp": 0.01025697, + "balance_loss_clip": 0.89141488, + "balance_loss_mlp": 1.01834798, + "epoch": 0.8385739193170204, + "flos": 28986842056320.0, + "grad_norm": 2.294620682328161, + "language_loss": 0.65394551, + "learning_rate": 2.6704110756686725e-07, + "loss": 0.67488873, + "num_input_tokens_seen": 150634540, + "step": 6974, + "time_per_iteration": 2.756585121154785 + }, + { + "auxiliary_loss_clip": 0.01114822, + "auxiliary_loss_mlp": 0.00509354, + "balance_loss_clip": 0.96913821, + "balance_loss_mlp": 0.0, + "epoch": 0.8386941622076595, + "flos": 23438068882560.0, + "grad_norm": 4.989153477547493, + "language_loss": 0.84088659, + "learning_rate": 2.6665236644309085e-07, + "loss": 0.85712838, + "num_input_tokens_seen": 150654850, + "step": 6975, + "time_per_iteration": 2.837299108505249 + }, + { + "auxiliary_loss_clip": 0.01139655, + "auxiliary_loss_mlp": 0.01024125, + "balance_loss_clip": 1.00801015, + "balance_loss_mlp": 1.01668084, + "epoch": 0.8388144050982985, + "flos": 23002724044800.0, + "grad_norm": 1.9265700588057497, + "language_loss": 0.79540765, + "learning_rate": 2.662638882667727e-07, + "loss": 0.81704545, + "num_input_tokens_seen": 150673790, + "step": 6976, + "time_per_iteration": 2.613839626312256 + }, + { + "auxiliary_loss_clip": 0.01167505, + "auxiliary_loss_mlp": 0.01030956, + "balance_loss_clip": 1.04769778, + "balance_loss_mlp": 1.02341008, + "epoch": 0.8389346479889377, + "flos": 24280353878400.0, + "grad_norm": 3.1798009875906414, + "language_loss": 0.73159242, + "learning_rate": 2.658756730968443e-07, + "loss": 0.75357699, + "num_input_tokens_seen": 150692255, + "step": 6977, + "time_per_iteration": 2.653949737548828 + }, + { + "auxiliary_loss_clip": 0.01122026, + "auxiliary_loss_mlp": 0.01026527, + "balance_loss_clip": 0.97363776, + "balance_loss_mlp": 1.01946652, + "epoch": 0.8390548908795767, + "flos": 21215019127680.0, + "grad_norm": 3.600507486362053, + "language_loss": 0.88506341, + "learning_rate": 2.654877209921975e-07, + "loss": 0.90654898, + "num_input_tokens_seen": 150709790, + "step": 6978, + "time_per_iteration": 3.5681097507476807 + }, + { + "auxiliary_loss_clip": 0.01078275, + "auxiliary_loss_mlp": 0.01035244, + "balance_loss_clip": 0.8936826, + "balance_loss_mlp": 1.02722132, + "epoch": 0.8391751337702158, + "flos": 35627299332480.0, + "grad_norm": 2.7203706025499264, + "language_loss": 0.62980121, + "learning_rate": 2.651000320116843e-07, + "loss": 0.65093637, + "num_input_tokens_seen": 150730675, + "step": 6979, + "time_per_iteration": 3.8242974281311035 + }, + { + "auxiliary_loss_clip": 0.01090721, + "auxiliary_loss_mlp": 0.0050953, + "balance_loss_clip": 0.9312495, + "balance_loss_mlp": 0.0, + "epoch": 0.839295376660855, + "flos": 21325229032320.0, + "grad_norm": 1.7906536355737568, + "language_loss": 0.76009315, + "learning_rate": 2.647126062141163e-07, + "loss": 0.77609569, + "num_input_tokens_seen": 150749750, + "step": 6980, + "time_per_iteration": 2.738056182861328 + }, + { + "auxiliary_loss_clip": 0.01119522, + "auxiliary_loss_mlp": 0.01026176, + "balance_loss_clip": 0.96858758, + "balance_loss_mlp": 1.01874876, + "epoch": 0.839415619551494, + "flos": 18442535961600.0, + "grad_norm": 1.814459483213339, + "language_loss": 0.8376866, + "learning_rate": 2.643254436582669e-07, + "loss": 0.85914361, + "num_input_tokens_seen": 150769240, + "step": 6981, + "time_per_iteration": 3.6379523277282715 + }, + { + "auxiliary_loss_clip": 0.01075926, + "auxiliary_loss_mlp": 0.0102647, + "balance_loss_clip": 0.89609265, + "balance_loss_mlp": 1.01904655, + "epoch": 0.8395358624421331, + "flos": 23221958705280.0, + "grad_norm": 1.7324978217423601, + "language_loss": 0.82268703, + "learning_rate": 2.6393854440286743e-07, + "loss": 0.8437109, + "num_input_tokens_seen": 150788410, + "step": 6982, + "time_per_iteration": 2.778067111968994 + }, + { + "auxiliary_loss_clip": 0.01167432, + "auxiliary_loss_mlp": 0.01023851, + "balance_loss_clip": 1.05183661, + "balance_loss_mlp": 1.01742816, + "epoch": 0.8396561053327722, + "flos": 24381657210240.0, + "grad_norm": 1.9531352573345788, + "language_loss": 0.7113657, + "learning_rate": 2.6355190850661045e-07, + "loss": 0.73327851, + "num_input_tokens_seen": 150805245, + "step": 6983, + "time_per_iteration": 2.6612327098846436 + }, + { + "auxiliary_loss_clip": 0.01118951, + "auxiliary_loss_mlp": 0.01025943, + "balance_loss_clip": 0.97354293, + "balance_loss_mlp": 1.018749, + "epoch": 0.8397763482234113, + "flos": 22237755073920.0, + "grad_norm": 1.5335343531658092, + "language_loss": 0.86475205, + "learning_rate": 2.631655360281486e-07, + "loss": 0.88620096, + "num_input_tokens_seen": 150824920, + "step": 6984, + "time_per_iteration": 2.668443441390991 + }, + { + "auxiliary_loss_clip": 0.01145682, + "auxiliary_loss_mlp": 0.00509212, + "balance_loss_clip": 1.00947869, + "balance_loss_mlp": 0.0, + "epoch": 0.8398965911140504, + "flos": 22163743100160.0, + "grad_norm": 1.948363296798426, + "language_loss": 0.65705216, + "learning_rate": 2.6277942702609323e-07, + "loss": 0.67360109, + "num_input_tokens_seen": 150844400, + "step": 6985, + "time_per_iteration": 2.7803609371185303 + }, + { + "auxiliary_loss_clip": 0.01098728, + "auxiliary_loss_mlp": 0.01028833, + "balance_loss_clip": 0.9347229, + "balance_loss_mlp": 1.02164459, + "epoch": 0.8400168340046895, + "flos": 21542775753600.0, + "grad_norm": 1.9387959897669729, + "language_loss": 0.87169671, + "learning_rate": 2.623935815590186e-07, + "loss": 0.89297235, + "num_input_tokens_seen": 150862780, + "step": 6986, + "time_per_iteration": 2.681102991104126 + }, + { + "auxiliary_loss_clip": 0.0112138, + "auxiliary_loss_mlp": 0.01029416, + "balance_loss_clip": 0.97506809, + "balance_loss_mlp": 1.0223825, + "epoch": 0.8401370768953286, + "flos": 22491966602880.0, + "grad_norm": 1.6773732881649228, + "language_loss": 0.81239188, + "learning_rate": 2.6200799968545516e-07, + "loss": 0.83389986, + "num_input_tokens_seen": 150883075, + "step": 6987, + "time_per_iteration": 2.6971867084503174 + }, + { + "auxiliary_loss_clip": 0.01019524, + "auxiliary_loss_mlp": 0.01008888, + "balance_loss_clip": 0.93871075, + "balance_loss_mlp": 1.00782156, + "epoch": 0.8402573197859676, + "flos": 59238890818560.0, + "grad_norm": 0.8062012681141842, + "language_loss": 0.5642761, + "learning_rate": 2.616226814638969e-07, + "loss": 0.58456016, + "num_input_tokens_seen": 150948180, + "step": 6988, + "time_per_iteration": 3.247701644897461 + }, + { + "auxiliary_loss_clip": 0.01120705, + "auxiliary_loss_mlp": 0.01029824, + "balance_loss_clip": 0.97304171, + "balance_loss_mlp": 1.02270103, + "epoch": 0.8403775626766068, + "flos": 22674608282880.0, + "grad_norm": 1.9574826544543835, + "language_loss": 0.77825093, + "learning_rate": 2.612376269527954e-07, + "loss": 0.79975617, + "num_input_tokens_seen": 150967885, + "step": 6989, + "time_per_iteration": 3.5497419834136963 + }, + { + "auxiliary_loss_clip": 0.01115882, + "auxiliary_loss_mlp": 0.01027287, + "balance_loss_clip": 0.97287333, + "balance_loss_mlp": 1.02017593, + "epoch": 0.8404978055672458, + "flos": 19609704495360.0, + "grad_norm": 2.4550979992252797, + "language_loss": 0.67865109, + "learning_rate": 2.608528362105635e-07, + "loss": 0.70008278, + "num_input_tokens_seen": 150987255, + "step": 6990, + "time_per_iteration": 2.6970980167388916 + }, + { + "auxiliary_loss_clip": 0.01096176, + "auxiliary_loss_mlp": 0.01023366, + "balance_loss_clip": 0.93135989, + "balance_loss_mlp": 1.01623106, + "epoch": 0.8406180484578849, + "flos": 27526929678720.0, + "grad_norm": 7.285683419653332, + "language_loss": 0.73179233, + "learning_rate": 2.6046830929557374e-07, + "loss": 0.75298774, + "num_input_tokens_seen": 151006905, + "step": 6991, + "time_per_iteration": 2.725550889968872 + }, + { + "auxiliary_loss_clip": 0.01092796, + "auxiliary_loss_mlp": 0.01028854, + "balance_loss_clip": 0.93430316, + "balance_loss_mlp": 1.0213083, + "epoch": 0.8407382913485241, + "flos": 22127473342080.0, + "grad_norm": 1.8991952447765628, + "language_loss": 0.84833175, + "learning_rate": 2.6008404626615776e-07, + "loss": 0.86954826, + "num_input_tokens_seen": 151025405, + "step": 6992, + "time_per_iteration": 2.765289545059204 + }, + { + "auxiliary_loss_clip": 0.0114468, + "auxiliary_loss_mlp": 0.01035334, + "balance_loss_clip": 1.01098752, + "balance_loss_mlp": 1.02820563, + "epoch": 0.8408585342391631, + "flos": 13918473982080.0, + "grad_norm": 2.9990811686250747, + "language_loss": 0.74603271, + "learning_rate": 2.597000471806092e-07, + "loss": 0.76783288, + "num_input_tokens_seen": 151041970, + "step": 6993, + "time_per_iteration": 2.6230273246765137 + }, + { + "auxiliary_loss_clip": 0.01115836, + "auxiliary_loss_mlp": 0.01022654, + "balance_loss_clip": 0.97412944, + "balance_loss_mlp": 1.01504207, + "epoch": 0.8409787771298022, + "flos": 20187865808640.0, + "grad_norm": 1.9138151433839805, + "language_loss": 0.73202753, + "learning_rate": 2.593163120971793e-07, + "loss": 0.75341237, + "num_input_tokens_seen": 151060835, + "step": 6994, + "time_per_iteration": 2.6672067642211914 + }, + { + "auxiliary_loss_clip": 0.0106187, + "auxiliary_loss_mlp": 0.01021731, + "balance_loss_clip": 0.89163864, + "balance_loss_mlp": 1.01495349, + "epoch": 0.8410990200204413, + "flos": 23142523777920.0, + "grad_norm": 2.1420330336805535, + "language_loss": 0.6892246, + "learning_rate": 2.5893284107408165e-07, + "loss": 0.7100606, + "num_input_tokens_seen": 151078205, + "step": 6995, + "time_per_iteration": 2.7386715412139893 + }, + { + "auxiliary_loss_clip": 0.01075741, + "auxiliary_loss_mlp": 0.01029363, + "balance_loss_clip": 0.8992129, + "balance_loss_mlp": 1.02175403, + "epoch": 0.8412192629110804, + "flos": 24027219757440.0, + "grad_norm": 1.6618781056263219, + "language_loss": 0.78081679, + "learning_rate": 2.5854963416948726e-07, + "loss": 0.80186784, + "num_input_tokens_seen": 151100470, + "step": 6996, + "time_per_iteration": 2.827310562133789 + }, + { + "auxiliary_loss_clip": 0.01068706, + "auxiliary_loss_mlp": 0.01028629, + "balance_loss_clip": 0.88909376, + "balance_loss_mlp": 1.02150583, + "epoch": 0.8413395058017195, + "flos": 25591703604480.0, + "grad_norm": 1.6120576064131547, + "language_loss": 0.69157815, + "learning_rate": 2.5816669144152816e-07, + "loss": 0.71255147, + "num_input_tokens_seen": 151121650, + "step": 6997, + "time_per_iteration": 2.7884325981140137 + }, + { + "auxiliary_loss_clip": 0.01060192, + "auxiliary_loss_mlp": 0.00998136, + "balance_loss_clip": 1.00900698, + "balance_loss_mlp": 0.99718857, + "epoch": 0.8414597486923585, + "flos": 63635396624640.0, + "grad_norm": 0.8488225032415478, + "language_loss": 0.66305137, + "learning_rate": 2.5778401294829777e-07, + "loss": 0.68363464, + "num_input_tokens_seen": 151180390, + "step": 6998, + "time_per_iteration": 3.296912908554077 + }, + { + "auxiliary_loss_clip": 0.0113725, + "auxiliary_loss_mlp": 0.00509105, + "balance_loss_clip": 1.00816751, + "balance_loss_mlp": 0.0, + "epoch": 0.8415799915829977, + "flos": 19098731571840.0, + "grad_norm": 1.9652845998344393, + "language_loss": 0.64986598, + "learning_rate": 2.574015987478473e-07, + "loss": 0.66632962, + "num_input_tokens_seen": 151198520, + "step": 6999, + "time_per_iteration": 2.635978937149048 + }, + { + "auxiliary_loss_clip": 0.01121345, + "auxiliary_loss_mlp": 0.01028247, + "balance_loss_clip": 0.97024459, + "balance_loss_mlp": 1.02051973, + "epoch": 0.8417002344736367, + "flos": 19821612781440.0, + "grad_norm": 1.94694629730928, + "language_loss": 0.86994332, + "learning_rate": 2.570194488981887e-07, + "loss": 0.89143926, + "num_input_tokens_seen": 151215065, + "step": 7000, + "time_per_iteration": 2.6595757007598877 + }, + { + "auxiliary_loss_clip": 0.01060343, + "auxiliary_loss_mlp": 0.0100122, + "balance_loss_clip": 1.00921488, + "balance_loss_mlp": 1.00026631, + "epoch": 0.8418204773642758, + "flos": 62161516834560.0, + "grad_norm": 0.8409248181622954, + "language_loss": 0.60353434, + "learning_rate": 2.566375634572939e-07, + "loss": 0.62414998, + "num_input_tokens_seen": 151275705, + "step": 7001, + "time_per_iteration": 3.122138023376465 + }, + { + "auxiliary_loss_clip": 0.01098576, + "auxiliary_loss_mlp": 0.01025453, + "balance_loss_clip": 0.93085718, + "balance_loss_mlp": 1.01806808, + "epoch": 0.841940720254915, + "flos": 17092905315840.0, + "grad_norm": 1.728543072420983, + "language_loss": 0.76278019, + "learning_rate": 2.562559424830943e-07, + "loss": 0.78402042, + "num_input_tokens_seen": 151293665, + "step": 7002, + "time_per_iteration": 2.6973319053649902 + }, + { + "auxiliary_loss_clip": 0.01113051, + "auxiliary_loss_mlp": 0.01031596, + "balance_loss_clip": 0.96884155, + "balance_loss_mlp": 1.02421451, + "epoch": 0.842060963145554, + "flos": 16283586026880.0, + "grad_norm": 3.4443482364265168, + "language_loss": 0.70131487, + "learning_rate": 2.5587458603348256e-07, + "loss": 0.72276139, + "num_input_tokens_seen": 151310955, + "step": 7003, + "time_per_iteration": 2.654416799545288 + }, + { + "auxiliary_loss_clip": 0.01090971, + "auxiliary_loss_mlp": 0.01026585, + "balance_loss_clip": 0.93268371, + "balance_loss_mlp": 1.01921153, + "epoch": 0.8421812060361931, + "flos": 21908238681600.0, + "grad_norm": 2.2816989587643826, + "language_loss": 0.84177184, + "learning_rate": 2.554934941663085e-07, + "loss": 0.86294746, + "num_input_tokens_seen": 151328490, + "step": 7004, + "time_per_iteration": 2.799095869064331 + }, + { + "auxiliary_loss_clip": 0.01094077, + "auxiliary_loss_mlp": 0.01023736, + "balance_loss_clip": 0.9327631, + "balance_loss_mlp": 1.01622558, + "epoch": 0.8423014489268322, + "flos": 27777693502080.0, + "grad_norm": 2.6589172482346752, + "language_loss": 0.73573077, + "learning_rate": 2.5511266693938484e-07, + "loss": 0.75690889, + "num_input_tokens_seen": 151346950, + "step": 7005, + "time_per_iteration": 3.613276243209839 + }, + { + "auxiliary_loss_clip": 0.01117026, + "auxiliary_loss_mlp": 0.01024109, + "balance_loss_clip": 0.97314906, + "balance_loss_mlp": 1.015872, + "epoch": 0.8424216918174713, + "flos": 25117610970240.0, + "grad_norm": 1.4475657086599851, + "language_loss": 0.778678, + "learning_rate": 2.547321044104822e-07, + "loss": 0.8000893, + "num_input_tokens_seen": 151368445, + "step": 7006, + "time_per_iteration": 2.7745003700256348 + }, + { + "auxiliary_loss_clip": 0.01169287, + "auxiliary_loss_mlp": 0.01025967, + "balance_loss_clip": 1.04999399, + "balance_loss_mlp": 1.01851642, + "epoch": 0.8425419347081103, + "flos": 24748448941440.0, + "grad_norm": 1.6491767283917418, + "language_loss": 0.76842463, + "learning_rate": 2.5435180663733113e-07, + "loss": 0.7903772, + "num_input_tokens_seen": 151388745, + "step": 7007, + "time_per_iteration": 2.6404941082000732 + }, + { + "auxiliary_loss_clip": 0.01075885, + "auxiliary_loss_mlp": 0.01022914, + "balance_loss_clip": 0.89103854, + "balance_loss_mlp": 1.01625335, + "epoch": 0.8426621775987495, + "flos": 24820916630400.0, + "grad_norm": 2.8196475301594517, + "language_loss": 0.71701956, + "learning_rate": 2.539717736776241e-07, + "loss": 0.73800755, + "num_input_tokens_seen": 151404970, + "step": 7008, + "time_per_iteration": 3.7161850929260254 + }, + { + "auxiliary_loss_clip": 0.01138857, + "auxiliary_loss_mlp": 0.01026734, + "balance_loss_clip": 1.01124215, + "balance_loss_mlp": 1.01963806, + "epoch": 0.8427824204893886, + "flos": 23550074467200.0, + "grad_norm": 2.5728012036477113, + "language_loss": 0.76405883, + "learning_rate": 2.535920055890097e-07, + "loss": 0.78571475, + "num_input_tokens_seen": 151426265, + "step": 7009, + "time_per_iteration": 2.697416067123413 + }, + { + "auxiliary_loss_clip": 0.01064615, + "auxiliary_loss_mlp": 0.01026202, + "balance_loss_clip": 0.89193285, + "balance_loss_mlp": 1.01887035, + "epoch": 0.8429026633800276, + "flos": 16143858120960.0, + "grad_norm": 2.305261340589426, + "language_loss": 0.64751697, + "learning_rate": 2.5321250242910006e-07, + "loss": 0.6684252, + "num_input_tokens_seen": 151444180, + "step": 7010, + "time_per_iteration": 2.72540545463562 + }, + { + "auxiliary_loss_clip": 0.01164693, + "auxiliary_loss_mlp": 0.01026068, + "balance_loss_clip": 1.04918385, + "balance_loss_mlp": 1.01935911, + "epoch": 0.8430229062706668, + "flos": 22198540400640.0, + "grad_norm": 1.8221815435698014, + "language_loss": 0.86395693, + "learning_rate": 2.5283326425546493e-07, + "loss": 0.8858645, + "num_input_tokens_seen": 151463290, + "step": 7011, + "time_per_iteration": 2.669152021408081 + }, + { + "auxiliary_loss_clip": 0.01092461, + "auxiliary_loss_mlp": 0.01024195, + "balance_loss_clip": 0.93618041, + "balance_loss_mlp": 1.01740289, + "epoch": 0.8431431491613058, + "flos": 35330317683840.0, + "grad_norm": 1.8995992518364053, + "language_loss": 0.69613254, + "learning_rate": 2.5245429112563443e-07, + "loss": 0.7172991, + "num_input_tokens_seen": 151483965, + "step": 7012, + "time_per_iteration": 2.794853687286377 + }, + { + "auxiliary_loss_clip": 0.0114033, + "auxiliary_loss_mlp": 0.01022219, + "balance_loss_clip": 1.01081932, + "balance_loss_mlp": 1.01509666, + "epoch": 0.8432633920519449, + "flos": 25812374808960.0, + "grad_norm": 1.8220363034420866, + "language_loss": 0.82099622, + "learning_rate": 2.5207558309709865e-07, + "loss": 0.84262174, + "num_input_tokens_seen": 151503700, + "step": 7013, + "time_per_iteration": 2.7209136486053467 + }, + { + "auxiliary_loss_clip": 0.01006758, + "auxiliary_loss_mlp": 0.00500663, + "balance_loss_clip": 0.89814222, + "balance_loss_mlp": 0.0, + "epoch": 0.8433836349425841, + "flos": 64959531592320.0, + "grad_norm": 0.6552779742035842, + "language_loss": 0.56234372, + "learning_rate": 2.516971402273065e-07, + "loss": 0.57741797, + "num_input_tokens_seen": 151569765, + "step": 7014, + "time_per_iteration": 3.3501391410827637 + }, + { + "auxiliary_loss_clip": 0.01116756, + "auxiliary_loss_mlp": 0.01029535, + "balance_loss_clip": 0.96977472, + "balance_loss_mlp": 1.02249312, + "epoch": 0.8435038778332231, + "flos": 20229989483520.0, + "grad_norm": 1.837743349865203, + "language_loss": 0.67914885, + "learning_rate": 2.513189625736687e-07, + "loss": 0.70061183, + "num_input_tokens_seen": 151586660, + "step": 7015, + "time_per_iteration": 3.50602650642395 + }, + { + "auxiliary_loss_clip": 0.010992, + "auxiliary_loss_mlp": 0.01027267, + "balance_loss_clip": 0.9330802, + "balance_loss_mlp": 1.02007246, + "epoch": 0.8436241207238622, + "flos": 20992229020800.0, + "grad_norm": 5.03533322241735, + "language_loss": 0.7192384, + "learning_rate": 2.509410501935534e-07, + "loss": 0.74050307, + "num_input_tokens_seen": 151602295, + "step": 7016, + "time_per_iteration": 2.7175118923187256 + }, + { + "auxiliary_loss_clip": 0.01122711, + "auxiliary_loss_mlp": 0.01030298, + "balance_loss_clip": 0.97339928, + "balance_loss_mlp": 1.02280903, + "epoch": 0.8437443636145013, + "flos": 14682257804160.0, + "grad_norm": 2.2079909785630583, + "language_loss": 0.75018108, + "learning_rate": 2.5056340314429116e-07, + "loss": 0.77171117, + "num_input_tokens_seen": 151619760, + "step": 7017, + "time_per_iteration": 2.6466073989868164 + }, + { + "auxiliary_loss_clip": 0.01072385, + "auxiliary_loss_mlp": 0.01024215, + "balance_loss_clip": 0.89178562, + "balance_loss_mlp": 1.01700854, + "epoch": 0.8438646065051404, + "flos": 21608814908160.0, + "grad_norm": 2.486084808575424, + "language_loss": 0.80494583, + "learning_rate": 2.5018602148316904e-07, + "loss": 0.82591188, + "num_input_tokens_seen": 151635795, + "step": 7018, + "time_per_iteration": 2.833369493484497 + }, + { + "auxiliary_loss_clip": 0.01163865, + "auxiliary_loss_mlp": 0.0102819, + "balance_loss_clip": 1.0500077, + "balance_loss_mlp": 1.02149034, + "epoch": 0.8439848493957794, + "flos": 23289937194240.0, + "grad_norm": 2.0158300678799455, + "language_loss": 0.80416107, + "learning_rate": 2.498089052674359e-07, + "loss": 0.82608157, + "num_input_tokens_seen": 151653770, + "step": 7019, + "time_per_iteration": 2.6410396099090576 + }, + { + "auxiliary_loss_clip": 0.01143855, + "auxiliary_loss_mlp": 0.01026183, + "balance_loss_clip": 1.01321542, + "balance_loss_mlp": 1.01902461, + "epoch": 0.8441050922864186, + "flos": 19719339782400.0, + "grad_norm": 2.1914369116934833, + "language_loss": 0.75515586, + "learning_rate": 2.494320545543007e-07, + "loss": 0.77685624, + "num_input_tokens_seen": 151673340, + "step": 7020, + "time_per_iteration": 2.6327502727508545 + }, + { + "auxiliary_loss_clip": 0.0116679, + "auxiliary_loss_mlp": 0.01023635, + "balance_loss_clip": 1.04823804, + "balance_loss_mlp": 1.01664937, + "epoch": 0.8442253351770577, + "flos": 21835268202240.0, + "grad_norm": 1.748582525464842, + "language_loss": 0.67005408, + "learning_rate": 2.490554694009308e-07, + "loss": 0.69195831, + "num_input_tokens_seen": 151694205, + "step": 7021, + "time_per_iteration": 2.6007840633392334 + }, + { + "auxiliary_loss_clip": 0.01144348, + "auxiliary_loss_mlp": 0.01026197, + "balance_loss_clip": 1.00858307, + "balance_loss_mlp": 1.01903844, + "epoch": 0.8443455780676967, + "flos": 34346365447680.0, + "grad_norm": 1.5796999763349813, + "language_loss": 0.78628916, + "learning_rate": 2.4867914986445426e-07, + "loss": 0.8079946, + "num_input_tokens_seen": 151716595, + "step": 7022, + "time_per_iteration": 2.7836780548095703 + }, + { + "auxiliary_loss_clip": 0.01119165, + "auxiliary_loss_mlp": 0.0102095, + "balance_loss_clip": 0.9696936, + "balance_loss_mlp": 1.01446176, + "epoch": 0.8444658209583359, + "flos": 48214599281280.0, + "grad_norm": 2.0702054197169266, + "language_loss": 0.71286863, + "learning_rate": 2.483030960019581e-07, + "loss": 0.7342698, + "num_input_tokens_seen": 151740525, + "step": 7023, + "time_per_iteration": 2.928463935852051 + }, + { + "auxiliary_loss_clip": 0.00978857, + "auxiliary_loss_mlp": 0.01000848, + "balance_loss_clip": 0.86011815, + "balance_loss_mlp": 0.99981135, + "epoch": 0.8445860638489749, + "flos": 68484773105280.0, + "grad_norm": 0.7301867543346768, + "language_loss": 0.55434042, + "learning_rate": 2.479273078704891e-07, + "loss": 0.57413751, + "num_input_tokens_seen": 151793890, + "step": 7024, + "time_per_iteration": 3.1271464824676514 + }, + { + "auxiliary_loss_clip": 0.00976201, + "auxiliary_loss_mlp": 0.01010828, + "balance_loss_clip": 0.86376882, + "balance_loss_mlp": 1.00968969, + "epoch": 0.844706306739614, + "flos": 62833331882880.0, + "grad_norm": 0.7839646849799244, + "language_loss": 0.64673162, + "learning_rate": 2.475517855270552e-07, + "loss": 0.6666019, + "num_input_tokens_seen": 151853970, + "step": 7025, + "time_per_iteration": 3.296414613723755 + }, + { + "auxiliary_loss_clip": 0.01162539, + "auxiliary_loss_mlp": 0.01030837, + "balance_loss_clip": 1.04765379, + "balance_loss_mlp": 1.02360117, + "epoch": 0.8448265496302532, + "flos": 14976114969600.0, + "grad_norm": 2.1363840554352636, + "language_loss": 0.72689968, + "learning_rate": 2.4717652902862143e-07, + "loss": 0.74883348, + "num_input_tokens_seen": 151872945, + "step": 7026, + "time_per_iteration": 2.5704057216644287 + }, + { + "auxiliary_loss_clip": 0.0112153, + "auxiliary_loss_mlp": 0.01022958, + "balance_loss_clip": 0.97154742, + "balance_loss_mlp": 1.01630032, + "epoch": 0.8449467925208922, + "flos": 23441265192960.0, + "grad_norm": 1.6682724896875385, + "language_loss": 0.81197029, + "learning_rate": 2.4680153843211495e-07, + "loss": 0.83341515, + "num_input_tokens_seen": 151892875, + "step": 7027, + "time_per_iteration": 2.7018203735351562 + }, + { + "auxiliary_loss_clip": 0.01119083, + "auxiliary_loss_mlp": 0.01024268, + "balance_loss_clip": 0.97361249, + "balance_loss_mlp": 1.01696312, + "epoch": 0.8450670354115313, + "flos": 22748045639040.0, + "grad_norm": 1.8428061488893077, + "language_loss": 0.72026086, + "learning_rate": 2.464268137944212e-07, + "loss": 0.74169439, + "num_input_tokens_seen": 151914170, + "step": 7028, + "time_per_iteration": 2.7001242637634277 + }, + { + "auxiliary_loss_clip": 0.01059909, + "auxiliary_loss_mlp": 0.01026388, + "balance_loss_clip": 0.89390516, + "balance_loss_mlp": 1.01861525, + "epoch": 0.8451872783021703, + "flos": 29825571605760.0, + "grad_norm": 1.960696874656023, + "language_loss": 0.78455961, + "learning_rate": 2.46052355172385e-07, + "loss": 0.80542254, + "num_input_tokens_seen": 151932210, + "step": 7029, + "time_per_iteration": 2.8304691314697266 + }, + { + "auxiliary_loss_clip": 0.01166096, + "auxiliary_loss_mlp": 0.01027556, + "balance_loss_clip": 1.04915261, + "balance_loss_mlp": 1.01993537, + "epoch": 0.8453075211928095, + "flos": 21870029589120.0, + "grad_norm": 2.0376718160305503, + "language_loss": 0.7498517, + "learning_rate": 2.456781626228128e-07, + "loss": 0.77178824, + "num_input_tokens_seen": 151951715, + "step": 7030, + "time_per_iteration": 2.5705513954162598 + }, + { + "auxiliary_loss_clip": 0.00984147, + "auxiliary_loss_mlp": 0.0050067, + "balance_loss_clip": 0.86109293, + "balance_loss_mlp": 0.0, + "epoch": 0.8454277640834486, + "flos": 58751869288320.0, + "grad_norm": 0.9139062321161718, + "language_loss": 0.66305709, + "learning_rate": 2.453042362024675e-07, + "loss": 0.67790532, + "num_input_tokens_seen": 152004960, + "step": 7031, + "time_per_iteration": 5.123303174972534 + }, + { + "auxiliary_loss_clip": 0.01161628, + "auxiliary_loss_mlp": 0.01024345, + "balance_loss_clip": 1.04621911, + "balance_loss_mlp": 1.01739812, + "epoch": 0.8455480069740876, + "flos": 27090076469760.0, + "grad_norm": 1.5543804355084834, + "language_loss": 0.73263037, + "learning_rate": 2.449305759680751e-07, + "loss": 0.75449014, + "num_input_tokens_seen": 152026285, + "step": 7032, + "time_per_iteration": 2.6971073150634766 + }, + { + "auxiliary_loss_clip": 0.01092623, + "auxiliary_loss_mlp": 0.01025867, + "balance_loss_clip": 0.9351874, + "balance_loss_mlp": 1.0185504, + "epoch": 0.8456682498647268, + "flos": 27198670262400.0, + "grad_norm": 1.3959925385307792, + "language_loss": 0.75275457, + "learning_rate": 2.445571819763188e-07, + "loss": 0.77393949, + "num_input_tokens_seen": 152048585, + "step": 7033, + "time_per_iteration": 2.7974600791931152 + }, + { + "auxiliary_loss_clip": 0.01164139, + "auxiliary_loss_mlp": 0.01027697, + "balance_loss_clip": 1.04995394, + "balance_loss_mlp": 1.02021646, + "epoch": 0.8457884927553658, + "flos": 20631901737600.0, + "grad_norm": 1.664164796185238, + "language_loss": 0.5813117, + "learning_rate": 2.4418405428384227e-07, + "loss": 0.60323012, + "num_input_tokens_seen": 152068795, + "step": 7034, + "time_per_iteration": 3.646955728530884 + }, + { + "auxiliary_loss_clip": 0.01162025, + "auxiliary_loss_mlp": 0.00509087, + "balance_loss_clip": 1.04696321, + "balance_loss_mlp": 0.0, + "epoch": 0.8459087356460049, + "flos": 15299023259520.0, + "grad_norm": 1.7036865334168227, + "language_loss": 0.71779191, + "learning_rate": 2.4381119294724864e-07, + "loss": 0.73450303, + "num_input_tokens_seen": 152086240, + "step": 7035, + "time_per_iteration": 2.6909689903259277 + }, + { + "auxiliary_loss_clip": 0.01164993, + "auxiliary_loss_mlp": 0.01023638, + "balance_loss_clip": 1.04841089, + "balance_loss_mlp": 1.01662564, + "epoch": 0.846028978536644, + "flos": 18843155326080.0, + "grad_norm": 2.1436408330126, + "language_loss": 0.53786421, + "learning_rate": 2.434385980231004e-07, + "loss": 0.55975056, + "num_input_tokens_seen": 152105080, + "step": 7036, + "time_per_iteration": 2.6061718463897705 + }, + { + "auxiliary_loss_clip": 0.01140226, + "auxiliary_loss_mlp": 0.01022007, + "balance_loss_clip": 1.00961852, + "balance_loss_mlp": 1.01509571, + "epoch": 0.8461492214272831, + "flos": 52661740285440.0, + "grad_norm": 1.576712380025288, + "language_loss": 0.65555811, + "learning_rate": 2.4306626956792043e-07, + "loss": 0.67718041, + "num_input_tokens_seen": 152130025, + "step": 7037, + "time_per_iteration": 2.936403274536133 + }, + { + "auxiliary_loss_clip": 0.01139748, + "auxiliary_loss_mlp": 0.01024722, + "balance_loss_clip": 1.00808227, + "balance_loss_mlp": 1.01770115, + "epoch": 0.8462694643179222, + "flos": 18588405093120.0, + "grad_norm": 2.054287158259991, + "language_loss": 0.75326502, + "learning_rate": 2.4269420763819017e-07, + "loss": 0.77490973, + "num_input_tokens_seen": 152148070, + "step": 7038, + "time_per_iteration": 2.6723439693450928 + }, + { + "auxiliary_loss_clip": 0.01137064, + "auxiliary_loss_mlp": 0.01027901, + "balance_loss_clip": 1.00814569, + "balance_loss_mlp": 1.02106988, + "epoch": 0.8463897072085613, + "flos": 24387080163840.0, + "grad_norm": 2.327938106984055, + "language_loss": 0.83709341, + "learning_rate": 2.4232241229035223e-07, + "loss": 0.85874307, + "num_input_tokens_seen": 152165825, + "step": 7039, + "time_per_iteration": 2.7122962474823 + }, + { + "auxiliary_loss_clip": 0.010422, + "auxiliary_loss_mlp": 0.01002937, + "balance_loss_clip": 0.9723711, + "balance_loss_mlp": 1.00192404, + "epoch": 0.8465099500992004, + "flos": 68702140258560.0, + "grad_norm": 0.7779571930098635, + "language_loss": 0.5670439, + "learning_rate": 2.419508835808064e-07, + "loss": 0.58749527, + "num_input_tokens_seen": 152222380, + "step": 7040, + "time_per_iteration": 3.1649672985076904 + }, + { + "auxiliary_loss_clip": 0.01116288, + "auxiliary_loss_mlp": 0.0102475, + "balance_loss_clip": 0.97066188, + "balance_loss_mlp": 1.01719224, + "epoch": 0.8466301929898394, + "flos": 13735724561280.0, + "grad_norm": 4.92231510624292, + "language_loss": 0.62683105, + "learning_rate": 2.415796215659134e-07, + "loss": 0.64824134, + "num_input_tokens_seen": 152239085, + "step": 7041, + "time_per_iteration": 3.5630576610565186 + }, + { + "auxiliary_loss_clip": 0.01096717, + "auxiliary_loss_mlp": 0.01032378, + "balance_loss_clip": 0.93012774, + "balance_loss_mlp": 1.02520776, + "epoch": 0.8467504358804786, + "flos": 19241260738560.0, + "grad_norm": 2.3278845568273656, + "language_loss": 0.77023464, + "learning_rate": 2.412086263019939e-07, + "loss": 0.7915256, + "num_input_tokens_seen": 152257110, + "step": 7042, + "time_per_iteration": 2.8073477745056152 + }, + { + "auxiliary_loss_clip": 0.0116137, + "auxiliary_loss_mlp": 0.0102327, + "balance_loss_clip": 1.05016816, + "balance_loss_mlp": 1.01695156, + "epoch": 0.8468706787711177, + "flos": 21324115710720.0, + "grad_norm": 2.0073774987340225, + "language_loss": 0.79935318, + "learning_rate": 2.408378978453276e-07, + "loss": 0.82119966, + "num_input_tokens_seen": 152277230, + "step": 7043, + "time_per_iteration": 2.643080949783325 + }, + { + "auxiliary_loss_clip": 0.01047404, + "auxiliary_loss_mlp": 0.00996945, + "balance_loss_clip": 0.97754323, + "balance_loss_mlp": 0.99602717, + "epoch": 0.8469909216617567, + "flos": 64877439058560.0, + "grad_norm": 0.8109507974430736, + "language_loss": 0.6397928, + "learning_rate": 2.404674362521533e-07, + "loss": 0.66023624, + "num_input_tokens_seen": 152335725, + "step": 7044, + "time_per_iteration": 3.1415278911590576 + }, + { + "auxiliary_loss_clip": 0.01137711, + "auxiliary_loss_mlp": 0.01024441, + "balance_loss_clip": 1.01038957, + "balance_loss_mlp": 1.01733947, + "epoch": 0.8471111645523959, + "flos": 19280583152640.0, + "grad_norm": 2.0540202896332707, + "language_loss": 0.74457932, + "learning_rate": 2.4009724157866997e-07, + "loss": 0.76620084, + "num_input_tokens_seen": 152352785, + "step": 7045, + "time_per_iteration": 2.620410680770874 + }, + { + "auxiliary_loss_clip": 0.01162995, + "auxiliary_loss_mlp": 0.01025629, + "balance_loss_clip": 1.04841745, + "balance_loss_mlp": 1.01875687, + "epoch": 0.8472314074430349, + "flos": 22015826893440.0, + "grad_norm": 1.899367128617057, + "language_loss": 0.76983976, + "learning_rate": 2.3972731388103564e-07, + "loss": 0.79172599, + "num_input_tokens_seen": 152371265, + "step": 7046, + "time_per_iteration": 2.641402006149292 + }, + { + "auxiliary_loss_clip": 0.00956902, + "auxiliary_loss_mlp": 0.01000704, + "balance_loss_clip": 0.82370806, + "balance_loss_mlp": 0.99966067, + "epoch": 0.847351650333674, + "flos": 57882580243200.0, + "grad_norm": 0.8051568894484499, + "language_loss": 0.62364817, + "learning_rate": 2.393576532153687e-07, + "loss": 0.64322418, + "num_input_tokens_seen": 152435050, + "step": 7047, + "time_per_iteration": 3.5047192573547363 + }, + { + "auxiliary_loss_clip": 0.01039193, + "auxiliary_loss_mlp": 0.00999998, + "balance_loss_clip": 0.97197765, + "balance_loss_mlp": 0.99897879, + "epoch": 0.8474718932243132, + "flos": 41284238313600.0, + "grad_norm": 0.9320342450386607, + "language_loss": 0.57842749, + "learning_rate": 2.389882596377453e-07, + "loss": 0.59881943, + "num_input_tokens_seen": 152489315, + "step": 7048, + "time_per_iteration": 3.31720232963562 + }, + { + "auxiliary_loss_clip": 0.01161989, + "auxiliary_loss_mlp": 0.01029777, + "balance_loss_clip": 1.04621172, + "balance_loss_mlp": 1.02257085, + "epoch": 0.8475921361149522, + "flos": 38180906974080.0, + "grad_norm": 1.7018851154643289, + "language_loss": 0.7630465, + "learning_rate": 2.386191332042031e-07, + "loss": 0.78496414, + "num_input_tokens_seen": 152511210, + "step": 7049, + "time_per_iteration": 2.782102108001709 + }, + { + "auxiliary_loss_clip": 0.01168768, + "auxiliary_loss_mlp": 0.01031694, + "balance_loss_clip": 1.04998016, + "balance_loss_mlp": 1.02392483, + "epoch": 0.8477123790055913, + "flos": 25375054723200.0, + "grad_norm": 1.6613678458548051, + "language_loss": 0.72865129, + "learning_rate": 2.3825027397073794e-07, + "loss": 0.75065589, + "num_input_tokens_seen": 152531685, + "step": 7050, + "time_per_iteration": 2.732602119445801 + }, + { + "auxiliary_loss_clip": 0.01138283, + "auxiliary_loss_mlp": 0.0102542, + "balance_loss_clip": 1.0115931, + "balance_loss_mlp": 1.01852715, + "epoch": 0.8478326218962304, + "flos": 30225185389440.0, + "grad_norm": 2.3879232121226655, + "language_loss": 0.66728967, + "learning_rate": 2.3788168199330515e-07, + "loss": 0.6889267, + "num_input_tokens_seen": 152553245, + "step": 7051, + "time_per_iteration": 2.69331693649292 + }, + { + "auxiliary_loss_clip": 0.01102528, + "auxiliary_loss_mlp": 0.01024112, + "balance_loss_clip": 0.9655894, + "balance_loss_mlp": 1.01707602, + "epoch": 0.8479528647868695, + "flos": 38213800853760.0, + "grad_norm": 2.089147907867653, + "language_loss": 0.7272383, + "learning_rate": 2.3751335732782074e-07, + "loss": 0.7485047, + "num_input_tokens_seen": 152574505, + "step": 7052, + "time_per_iteration": 2.866725444793701 + }, + { + "auxiliary_loss_clip": 0.01141867, + "auxiliary_loss_mlp": 0.01025796, + "balance_loss_clip": 1.01239502, + "balance_loss_mlp": 1.01897454, + "epoch": 0.8480731076775085, + "flos": 20957790856320.0, + "grad_norm": 1.9562459779943202, + "language_loss": 0.79408073, + "learning_rate": 2.371453000301582e-07, + "loss": 0.81575733, + "num_input_tokens_seen": 152593190, + "step": 7053, + "time_per_iteration": 2.609027624130249 + }, + { + "auxiliary_loss_clip": 0.01090935, + "auxiliary_loss_mlp": 0.01021381, + "balance_loss_clip": 0.93311727, + "balance_loss_mlp": 1.01462495, + "epoch": 0.8481933505681477, + "flos": 32596510487040.0, + "grad_norm": 2.0803936399199916, + "language_loss": 0.74639809, + "learning_rate": 2.3677751015615222e-07, + "loss": 0.76752126, + "num_input_tokens_seen": 152615265, + "step": 7054, + "time_per_iteration": 2.752802610397339 + }, + { + "auxiliary_loss_clip": 0.01107464, + "auxiliary_loss_mlp": 0.01026119, + "balance_loss_clip": 0.96850896, + "balance_loss_mlp": 1.01933312, + "epoch": 0.8483135934587868, + "flos": 20741177888640.0, + "grad_norm": 1.5939022062087638, + "language_loss": 0.85368299, + "learning_rate": 2.3640998776159593e-07, + "loss": 0.87501884, + "num_input_tokens_seen": 152632770, + "step": 7055, + "time_per_iteration": 2.6709845066070557 + }, + { + "auxiliary_loss_clip": 0.01120415, + "auxiliary_loss_mlp": 0.01022734, + "balance_loss_clip": 0.97362185, + "balance_loss_mlp": 1.01626122, + "epoch": 0.8484338363494258, + "flos": 21653057485440.0, + "grad_norm": 1.6882927496292366, + "language_loss": 0.81221104, + "learning_rate": 2.3604273290224253e-07, + "loss": 0.83364248, + "num_input_tokens_seen": 152653485, + "step": 7056, + "time_per_iteration": 2.6524925231933594 + }, + { + "auxiliary_loss_clip": 0.01121734, + "auxiliary_loss_mlp": 0.01028845, + "balance_loss_clip": 0.97376263, + "balance_loss_mlp": 1.02115059, + "epoch": 0.848554079240065, + "flos": 15013964926080.0, + "grad_norm": 1.73897632958131, + "language_loss": 0.74342883, + "learning_rate": 2.356757456338039e-07, + "loss": 0.76493466, + "num_input_tokens_seen": 152670970, + "step": 7057, + "time_per_iteration": 4.813460826873779 + }, + { + "auxiliary_loss_clip": 0.01017343, + "auxiliary_loss_mlp": 0.01002758, + "balance_loss_clip": 0.93729424, + "balance_loss_mlp": 1.00158966, + "epoch": 0.848674322130704, + "flos": 68060453742720.0, + "grad_norm": 0.7564571645720449, + "language_loss": 0.59015709, + "learning_rate": 2.3530902601195147e-07, + "loss": 0.61035806, + "num_input_tokens_seen": 152739460, + "step": 7058, + "time_per_iteration": 3.455054998397827 + }, + { + "auxiliary_loss_clip": 0.01140604, + "auxiliary_loss_mlp": 0.01023477, + "balance_loss_clip": 1.00958633, + "balance_loss_mlp": 1.01588058, + "epoch": 0.8487945650213431, + "flos": 18475788977280.0, + "grad_norm": 2.57783217044059, + "language_loss": 0.78973806, + "learning_rate": 2.34942574092317e-07, + "loss": 0.8113789, + "num_input_tokens_seen": 152754710, + "step": 7059, + "time_per_iteration": 2.6242377758026123 + }, + { + "auxiliary_loss_clip": 0.01147185, + "auxiliary_loss_mlp": 0.0102615, + "balance_loss_clip": 1.01177967, + "balance_loss_mlp": 1.01896787, + "epoch": 0.8489148079119821, + "flos": 23473189405440.0, + "grad_norm": 2.7462182216760596, + "language_loss": 0.7744785, + "learning_rate": 2.3457638993049045e-07, + "loss": 0.79621184, + "num_input_tokens_seen": 152772700, + "step": 7060, + "time_per_iteration": 3.622462034225464 + }, + { + "auxiliary_loss_clip": 0.01052075, + "auxiliary_loss_mlp": 0.01030634, + "balance_loss_clip": 0.85977674, + "balance_loss_mlp": 1.02278376, + "epoch": 0.8490350508026213, + "flos": 19937604775680.0, + "grad_norm": 2.2510924566726427, + "language_loss": 0.64580786, + "learning_rate": 2.3421047358202252e-07, + "loss": 0.66663504, + "num_input_tokens_seen": 152791550, + "step": 7061, + "time_per_iteration": 2.8294730186462402 + }, + { + "auxiliary_loss_clip": 0.01141428, + "auxiliary_loss_mlp": 0.01023622, + "balance_loss_clip": 1.00990152, + "balance_loss_mlp": 1.0168004, + "epoch": 0.8491552936932604, + "flos": 24279958828800.0, + "grad_norm": 2.7643416096696907, + "language_loss": 0.83468473, + "learning_rate": 2.3384482510242144e-07, + "loss": 0.85633528, + "num_input_tokens_seen": 152809410, + "step": 7062, + "time_per_iteration": 2.737515449523926 + }, + { + "auxiliary_loss_clip": 0.01163411, + "auxiliary_loss_mlp": 0.01023894, + "balance_loss_clip": 1.04557693, + "balance_loss_mlp": 1.01675296, + "epoch": 0.8492755365838994, + "flos": 22522526098560.0, + "grad_norm": 2.112460069092214, + "language_loss": 0.77440649, + "learning_rate": 2.3347944454715575e-07, + "loss": 0.79627955, + "num_input_tokens_seen": 152825800, + "step": 7063, + "time_per_iteration": 2.6451666355133057 + }, + { + "auxiliary_loss_clip": 0.01167787, + "auxiliary_loss_mlp": 0.01027988, + "balance_loss_clip": 1.04936957, + "balance_loss_mlp": 1.02044225, + "epoch": 0.8493957794745386, + "flos": 26980441182720.0, + "grad_norm": 2.0771755403694647, + "language_loss": 0.6738835, + "learning_rate": 2.331143319716542e-07, + "loss": 0.69584125, + "num_input_tokens_seen": 152845330, + "step": 7064, + "time_per_iteration": 2.745914936065674 + }, + { + "auxiliary_loss_clip": 0.01100811, + "auxiliary_loss_mlp": 0.01029247, + "balance_loss_clip": 0.93428469, + "balance_loss_mlp": 1.02157545, + "epoch": 0.8495160223651776, + "flos": 29861985018240.0, + "grad_norm": 2.2393209957379185, + "language_loss": 0.65931922, + "learning_rate": 2.3274948743130363e-07, + "loss": 0.68061984, + "num_input_tokens_seen": 152865165, + "step": 7065, + "time_per_iteration": 2.8123557567596436 + }, + { + "auxiliary_loss_clip": 0.01164017, + "auxiliary_loss_mlp": 0.01023299, + "balance_loss_clip": 1.04640245, + "balance_loss_mlp": 1.01616478, + "epoch": 0.8496362652558167, + "flos": 23075443128960.0, + "grad_norm": 2.0829945202777447, + "language_loss": 0.79471427, + "learning_rate": 2.3238491098145085e-07, + "loss": 0.81658745, + "num_input_tokens_seen": 152884695, + "step": 7066, + "time_per_iteration": 2.662079334259033 + }, + { + "auxiliary_loss_clip": 0.01140546, + "auxiliary_loss_mlp": 0.01021449, + "balance_loss_clip": 1.01097178, + "balance_loss_mlp": 1.01422238, + "epoch": 0.8497565081464559, + "flos": 14609107756800.0, + "grad_norm": 2.1380056208872573, + "language_loss": 0.73509145, + "learning_rate": 2.3202060267740141e-07, + "loss": 0.75671136, + "num_input_tokens_seen": 152902220, + "step": 7067, + "time_per_iteration": 3.6028335094451904 + }, + { + "auxiliary_loss_clip": 0.01066311, + "auxiliary_loss_mlp": 0.01021899, + "balance_loss_clip": 0.89127147, + "balance_loss_mlp": 1.01479435, + "epoch": 0.8498767510370949, + "flos": 21136446126720.0, + "grad_norm": 3.8352871697375157, + "language_loss": 0.7742058, + "learning_rate": 2.3165656257442044e-07, + "loss": 0.79508793, + "num_input_tokens_seen": 152920740, + "step": 7068, + "time_per_iteration": 2.6869583129882812 + }, + { + "auxiliary_loss_clip": 0.01136827, + "auxiliary_loss_mlp": 0.01022756, + "balance_loss_clip": 1.0096662, + "balance_loss_mlp": 1.01633978, + "epoch": 0.849996993927734, + "flos": 23654538195840.0, + "grad_norm": 1.8856579856713678, + "language_loss": 0.90272605, + "learning_rate": 2.31292790727734e-07, + "loss": 0.92432189, + "num_input_tokens_seen": 152938305, + "step": 7069, + "time_per_iteration": 2.623485565185547 + }, + { + "auxiliary_loss_clip": 0.01160846, + "auxiliary_loss_mlp": 0.01027114, + "balance_loss_clip": 1.04563665, + "balance_loss_mlp": 1.02005672, + "epoch": 0.8501172368183731, + "flos": 20558069331840.0, + "grad_norm": 2.166330991811594, + "language_loss": 0.80291021, + "learning_rate": 2.3092928719252392e-07, + "loss": 0.82478976, + "num_input_tokens_seen": 152956705, + "step": 7070, + "time_per_iteration": 2.620110034942627 + }, + { + "auxiliary_loss_clip": 0.01141063, + "auxiliary_loss_mlp": 0.01027895, + "balance_loss_clip": 1.01056135, + "balance_loss_mlp": 1.02085257, + "epoch": 0.8502374797090122, + "flos": 22272624201600.0, + "grad_norm": 3.038224900450382, + "language_loss": 0.78425896, + "learning_rate": 2.3056605202393475e-07, + "loss": 0.8059485, + "num_input_tokens_seen": 152974265, + "step": 7071, + "time_per_iteration": 2.663093328475952 + }, + { + "auxiliary_loss_clip": 0.01134121, + "auxiliary_loss_mlp": 0.00509744, + "balance_loss_clip": 1.00505424, + "balance_loss_mlp": 0.0, + "epoch": 0.8503577225996513, + "flos": 23659817495040.0, + "grad_norm": 1.9882986788759665, + "language_loss": 0.67287755, + "learning_rate": 2.3020308527706888e-07, + "loss": 0.68931621, + "num_input_tokens_seen": 152993680, + "step": 7072, + "time_per_iteration": 2.6427040100097656 + }, + { + "auxiliary_loss_clip": 0.01122463, + "auxiliary_loss_mlp": 0.01031805, + "balance_loss_clip": 0.97141469, + "balance_loss_mlp": 1.0248847, + "epoch": 0.8504779654902904, + "flos": 26758513002240.0, + "grad_norm": 2.6962345173686613, + "language_loss": 0.88906085, + "learning_rate": 2.2984038700698715e-07, + "loss": 0.91060352, + "num_input_tokens_seen": 153012990, + "step": 7073, + "time_per_iteration": 2.7054712772369385 + }, + { + "auxiliary_loss_clip": 0.01138691, + "auxiliary_loss_mlp": 0.01022422, + "balance_loss_clip": 1.0108155, + "balance_loss_mlp": 1.01546359, + "epoch": 0.8505982083809295, + "flos": 26468247196800.0, + "grad_norm": 1.603348624151481, + "language_loss": 0.79023683, + "learning_rate": 2.2947795726871222e-07, + "loss": 0.81184804, + "num_input_tokens_seen": 153034015, + "step": 7074, + "time_per_iteration": 2.6906683444976807 + }, + { + "auxiliary_loss_clip": 0.01143687, + "auxiliary_loss_mlp": 0.00509655, + "balance_loss_clip": 1.01647329, + "balance_loss_mlp": 0.0, + "epoch": 0.8507184512715685, + "flos": 20303390926080.0, + "grad_norm": 2.5446855814984817, + "language_loss": 0.85742939, + "learning_rate": 2.2911579611722253e-07, + "loss": 0.87396288, + "num_input_tokens_seen": 153053160, + "step": 7075, + "time_per_iteration": 2.5896756649017334 + }, + { + "auxiliary_loss_clip": 0.01113337, + "auxiliary_loss_mlp": 0.01029697, + "balance_loss_clip": 0.96894825, + "balance_loss_mlp": 1.02260983, + "epoch": 0.8508386941622077, + "flos": 19025186474880.0, + "grad_norm": 1.8738371941844252, + "language_loss": 0.87702137, + "learning_rate": 2.2875390360745905e-07, + "loss": 0.89845175, + "num_input_tokens_seen": 153072565, + "step": 7076, + "time_per_iteration": 2.6385385990142822 + }, + { + "auxiliary_loss_clip": 0.01098422, + "auxiliary_loss_mlp": 0.01027363, + "balance_loss_clip": 0.93160093, + "balance_loss_mlp": 1.02001667, + "epoch": 0.8509589370528468, + "flos": 16433405654400.0, + "grad_norm": 1.6410458737471405, + "language_loss": 0.7764464, + "learning_rate": 2.2839227979432008e-07, + "loss": 0.79770428, + "num_input_tokens_seen": 153090215, + "step": 7077, + "time_per_iteration": 2.6712749004364014 + }, + { + "auxiliary_loss_clip": 0.0111881, + "auxiliary_loss_mlp": 0.01022857, + "balance_loss_clip": 0.97087514, + "balance_loss_mlp": 1.01582956, + "epoch": 0.8510791799434858, + "flos": 18259714713600.0, + "grad_norm": 2.0720158660594117, + "language_loss": 0.85105443, + "learning_rate": 2.2803092473266373e-07, + "loss": 0.87247109, + "num_input_tokens_seen": 153107740, + "step": 7078, + "time_per_iteration": 2.680837631225586 + }, + { + "auxiliary_loss_clip": 0.0116456, + "auxiliary_loss_mlp": 0.0103188, + "balance_loss_clip": 1.04807854, + "balance_loss_mlp": 1.02502537, + "epoch": 0.851199422834125, + "flos": 23441372933760.0, + "grad_norm": 2.40389749023067, + "language_loss": 0.86080348, + "learning_rate": 2.2766983847730724e-07, + "loss": 0.88276786, + "num_input_tokens_seen": 153127410, + "step": 7079, + "time_per_iteration": 2.746610403060913 + }, + { + "auxiliary_loss_clip": 0.01102329, + "auxiliary_loss_mlp": 0.01030281, + "balance_loss_clip": 0.93203688, + "balance_loss_mlp": 1.02296197, + "epoch": 0.851319665724764, + "flos": 16289404030080.0, + "grad_norm": 1.8533020387258219, + "language_loss": 0.667386, + "learning_rate": 2.2730902108302663e-07, + "loss": 0.68871206, + "num_input_tokens_seen": 153144325, + "step": 7080, + "time_per_iteration": 2.699108839035034 + }, + { + "auxiliary_loss_clip": 0.0111016, + "auxiliary_loss_mlp": 0.01023443, + "balance_loss_clip": 0.96823847, + "balance_loss_mlp": 1.01640964, + "epoch": 0.8514399086154031, + "flos": 18989347680000.0, + "grad_norm": 1.9398079861181141, + "language_loss": 0.68885541, + "learning_rate": 2.269484726045583e-07, + "loss": 0.71019143, + "num_input_tokens_seen": 153163240, + "step": 7081, + "time_per_iteration": 2.6112477779388428 + }, + { + "auxiliary_loss_clip": 0.01100823, + "auxiliary_loss_mlp": 0.01027713, + "balance_loss_clip": 0.93596649, + "balance_loss_mlp": 1.02079856, + "epoch": 0.8515601515060423, + "flos": 24571194301440.0, + "grad_norm": 1.6916904591401292, + "language_loss": 0.79025608, + "learning_rate": 2.2658819309659672e-07, + "loss": 0.81154144, + "num_input_tokens_seen": 153183440, + "step": 7082, + "time_per_iteration": 2.7460083961486816 + }, + { + "auxiliary_loss_clip": 0.01114258, + "auxiliary_loss_mlp": 0.01023094, + "balance_loss_clip": 0.97349691, + "balance_loss_mlp": 1.01629329, + "epoch": 0.8516803943966813, + "flos": 19529443555200.0, + "grad_norm": 1.9101708074510482, + "language_loss": 0.84679097, + "learning_rate": 2.2622818261379706e-07, + "loss": 0.86816454, + "num_input_tokens_seen": 153200460, + "step": 7083, + "time_per_iteration": 3.5804622173309326 + }, + { + "auxiliary_loss_clip": 0.01116262, + "auxiliary_loss_mlp": 0.0102488, + "balance_loss_clip": 0.9706378, + "balance_loss_mlp": 1.01709592, + "epoch": 0.8518006372873204, + "flos": 20265792364800.0, + "grad_norm": 2.2637369396326084, + "language_loss": 0.75112414, + "learning_rate": 2.2586844121077142e-07, + "loss": 0.77253562, + "num_input_tokens_seen": 153218970, + "step": 7084, + "time_per_iteration": 2.6410634517669678 + }, + { + "auxiliary_loss_clip": 0.01072565, + "auxiliary_loss_mlp": 0.0102831, + "balance_loss_clip": 0.89240503, + "balance_loss_mlp": 1.02098513, + "epoch": 0.8519208801779595, + "flos": 24133227770880.0, + "grad_norm": 3.55470092792352, + "language_loss": 0.72139227, + "learning_rate": 2.2550896894209215e-07, + "loss": 0.742401, + "num_input_tokens_seen": 153238485, + "step": 7085, + "time_per_iteration": 2.7761330604553223 + }, + { + "auxiliary_loss_clip": 0.00963183, + "auxiliary_loss_mlp": 0.01003611, + "balance_loss_clip": 0.82457507, + "balance_loss_mlp": 1.00243652, + "epoch": 0.8520411230685986, + "flos": 63035223252480.0, + "grad_norm": 0.6782848741881086, + "language_loss": 0.56644458, + "learning_rate": 2.2514976586229184e-07, + "loss": 0.5861125, + "num_input_tokens_seen": 153306430, + "step": 7086, + "time_per_iteration": 3.5643129348754883 + }, + { + "auxiliary_loss_clip": 0.01042327, + "auxiliary_loss_mlp": 0.01001454, + "balance_loss_clip": 0.9730103, + "balance_loss_mlp": 1.00050592, + "epoch": 0.8521613659592376, + "flos": 65836865283840.0, + "grad_norm": 0.7502285096717064, + "language_loss": 0.54682022, + "learning_rate": 2.247908320258609e-07, + "loss": 0.567258, + "num_input_tokens_seen": 153366520, + "step": 7087, + "time_per_iteration": 4.254504442214966 + }, + { + "auxiliary_loss_clip": 0.01065128, + "auxiliary_loss_mlp": 0.01021972, + "balance_loss_clip": 0.89591408, + "balance_loss_mlp": 1.01381826, + "epoch": 0.8522816088498768, + "flos": 23112323418240.0, + "grad_norm": 4.012267525897293, + "language_loss": 0.79622996, + "learning_rate": 2.2443216748724914e-07, + "loss": 0.817101, + "num_input_tokens_seen": 153387230, + "step": 7088, + "time_per_iteration": 2.7317612171173096 + }, + { + "auxiliary_loss_clip": 0.01143941, + "auxiliary_loss_mlp": 0.00508886, + "balance_loss_clip": 1.01130474, + "balance_loss_mlp": 0.0, + "epoch": 0.8524018517405159, + "flos": 31758140073600.0, + "grad_norm": 3.127704328658053, + "language_loss": 0.74613386, + "learning_rate": 2.2407377230086588e-07, + "loss": 0.76266211, + "num_input_tokens_seen": 153409585, + "step": 7089, + "time_per_iteration": 2.665402889251709 + }, + { + "auxiliary_loss_clip": 0.01097398, + "auxiliary_loss_mlp": 0.01027438, + "balance_loss_clip": 0.93909383, + "balance_loss_mlp": 1.02040458, + "epoch": 0.8525220946311549, + "flos": 18690318956160.0, + "grad_norm": 2.0875341868088277, + "language_loss": 0.833884, + "learning_rate": 2.23715646521079e-07, + "loss": 0.85513234, + "num_input_tokens_seen": 153427105, + "step": 7090, + "time_per_iteration": 2.7697436809539795 + }, + { + "auxiliary_loss_clip": 0.01143945, + "auxiliary_loss_mlp": 0.00509952, + "balance_loss_clip": 1.0092181, + "balance_loss_mlp": 0.0, + "epoch": 0.852642337521794, + "flos": 21793216354560.0, + "grad_norm": 1.7905117988159884, + "language_loss": 0.84351122, + "learning_rate": 2.2335779020221724e-07, + "loss": 0.8600502, + "num_input_tokens_seen": 153443725, + "step": 7091, + "time_per_iteration": 2.697500705718994 + }, + { + "auxiliary_loss_clip": 0.01056289, + "auxiliary_loss_mlp": 0.01016544, + "balance_loss_clip": 0.9945097, + "balance_loss_mlp": 1.01529837, + "epoch": 0.8527625804124331, + "flos": 69040132260480.0, + "grad_norm": 0.8030863108690768, + "language_loss": 0.5644753, + "learning_rate": 2.2300020339856497e-07, + "loss": 0.58520365, + "num_input_tokens_seen": 153506410, + "step": 7092, + "time_per_iteration": 3.2714414596557617 + }, + { + "auxiliary_loss_clip": 0.01115419, + "auxiliary_loss_mlp": 0.0102272, + "balance_loss_clip": 0.97210097, + "balance_loss_mlp": 1.01608634, + "epoch": 0.8528828233030722, + "flos": 26979399688320.0, + "grad_norm": 2.2604163663266337, + "language_loss": 0.78246444, + "learning_rate": 2.2264288616436966e-07, + "loss": 0.80384582, + "num_input_tokens_seen": 153526665, + "step": 7093, + "time_per_iteration": 3.626866340637207 + }, + { + "auxiliary_loss_clip": 0.01113982, + "auxiliary_loss_mlp": 0.01029949, + "balance_loss_clip": 0.97087896, + "balance_loss_mlp": 1.02232599, + "epoch": 0.8530030661937112, + "flos": 17487598936320.0, + "grad_norm": 1.8805094120049828, + "language_loss": 0.72606653, + "learning_rate": 2.222858385538351e-07, + "loss": 0.74750578, + "num_input_tokens_seen": 153543465, + "step": 7094, + "time_per_iteration": 2.591820478439331 + }, + { + "auxiliary_loss_clip": 0.01138244, + "auxiliary_loss_mlp": 0.01026272, + "balance_loss_clip": 1.00907218, + "balance_loss_mlp": 1.01943874, + "epoch": 0.8531233090843504, + "flos": 22160798184960.0, + "grad_norm": 1.5869980423343037, + "language_loss": 0.68005633, + "learning_rate": 2.2192906062112527e-07, + "loss": 0.70170146, + "num_input_tokens_seen": 153563340, + "step": 7095, + "time_per_iteration": 2.6828503608703613 + }, + { + "auxiliary_loss_clip": 0.01161722, + "auxiliary_loss_mlp": 0.01025121, + "balance_loss_clip": 1.04545259, + "balance_loss_mlp": 1.01806402, + "epoch": 0.8532435519749895, + "flos": 37635388145280.0, + "grad_norm": 1.6329859956182737, + "language_loss": 0.70676088, + "learning_rate": 2.2157255242036377e-07, + "loss": 0.72862929, + "num_input_tokens_seen": 153587005, + "step": 7096, + "time_per_iteration": 2.737306833267212 + }, + { + "auxiliary_loss_clip": 0.01089241, + "auxiliary_loss_mlp": 0.01026758, + "balance_loss_clip": 0.93258578, + "balance_loss_mlp": 1.02001929, + "epoch": 0.8533637948656285, + "flos": 21398163598080.0, + "grad_norm": 1.990635916171604, + "language_loss": 0.74198824, + "learning_rate": 2.2121631400563135e-07, + "loss": 0.76314819, + "num_input_tokens_seen": 153606835, + "step": 7097, + "time_per_iteration": 2.6973555088043213 + }, + { + "auxiliary_loss_clip": 0.01035765, + "auxiliary_loss_mlp": 0.01002087, + "balance_loss_clip": 0.97363806, + "balance_loss_mlp": 1.00090659, + "epoch": 0.8534840377562677, + "flos": 53345122490880.0, + "grad_norm": 0.782123884367452, + "language_loss": 0.52993309, + "learning_rate": 2.208603454309701e-07, + "loss": 0.55031168, + "num_input_tokens_seen": 153664925, + "step": 7098, + "time_per_iteration": 3.2235653400421143 + }, + { + "auxiliary_loss_clip": 0.01073788, + "auxiliary_loss_mlp": 0.01024949, + "balance_loss_clip": 0.89591056, + "balance_loss_mlp": 1.0178802, + "epoch": 0.8536042806469067, + "flos": 20814148368000.0, + "grad_norm": 4.741168656384555, + "language_loss": 0.71056157, + "learning_rate": 2.2050464675037994e-07, + "loss": 0.73154891, + "num_input_tokens_seen": 153683550, + "step": 7099, + "time_per_iteration": 2.721975564956665 + }, + { + "auxiliary_loss_clip": 0.01118332, + "auxiliary_loss_mlp": 0.01030285, + "balance_loss_clip": 0.97251523, + "balance_loss_mlp": 1.02270293, + "epoch": 0.8537245235375458, + "flos": 24681368292480.0, + "grad_norm": 1.984122204865594, + "language_loss": 0.7295171, + "learning_rate": 2.2014921801782016e-07, + "loss": 0.75100321, + "num_input_tokens_seen": 153703040, + "step": 7100, + "time_per_iteration": 2.8217530250549316 + }, + { + "auxiliary_loss_clip": 0.01118258, + "auxiliary_loss_mlp": 0.01022844, + "balance_loss_clip": 0.96857905, + "balance_loss_mlp": 1.01609361, + "epoch": 0.853844766428185, + "flos": 24384817607040.0, + "grad_norm": 2.6162477254182797, + "language_loss": 0.73945749, + "learning_rate": 2.1979405928720872e-07, + "loss": 0.76086849, + "num_input_tokens_seen": 153722695, + "step": 7101, + "time_per_iteration": 2.6796422004699707 + }, + { + "auxiliary_loss_clip": 0.01121378, + "auxiliary_loss_mlp": 0.01026465, + "balance_loss_clip": 0.97164917, + "balance_loss_mlp": 1.01920521, + "epoch": 0.853965009318824, + "flos": 20955707867520.0, + "grad_norm": 1.3922545853931918, + "language_loss": 0.79536116, + "learning_rate": 2.1943917061242257e-07, + "loss": 0.81683958, + "num_input_tokens_seen": 153742550, + "step": 7102, + "time_per_iteration": 2.6667771339416504 + }, + { + "auxiliary_loss_clip": 0.01147411, + "auxiliary_loss_mlp": 0.00510157, + "balance_loss_clip": 1.01092327, + "balance_loss_mlp": 0.0, + "epoch": 0.8540852522094631, + "flos": 24201816791040.0, + "grad_norm": 2.08570642256442, + "language_loss": 0.66267651, + "learning_rate": 2.1908455204729903e-07, + "loss": 0.67925215, + "num_input_tokens_seen": 153761700, + "step": 7103, + "time_per_iteration": 2.670424222946167 + }, + { + "auxiliary_loss_clip": 0.01117002, + "auxiliary_loss_mlp": 0.01025276, + "balance_loss_clip": 0.97014791, + "balance_loss_mlp": 1.01811159, + "epoch": 0.8542054951001022, + "flos": 25082921410560.0, + "grad_norm": 2.178705529842407, + "language_loss": 0.78743076, + "learning_rate": 2.1873020364563265e-07, + "loss": 0.80885357, + "num_input_tokens_seen": 153780765, + "step": 7104, + "time_per_iteration": 2.728243827819824 + }, + { + "auxiliary_loss_clip": 0.01139396, + "auxiliary_loss_mlp": 0.01024967, + "balance_loss_clip": 1.01095808, + "balance_loss_mlp": 1.01838624, + "epoch": 0.8543257379907413, + "flos": 24316551809280.0, + "grad_norm": 3.0129393900424133, + "language_loss": 0.7605477, + "learning_rate": 2.183761254611789e-07, + "loss": 0.78219128, + "num_input_tokens_seen": 153801090, + "step": 7105, + "time_per_iteration": 2.718104600906372 + }, + { + "auxiliary_loss_clip": 0.01142797, + "auxiliary_loss_mlp": 0.01024039, + "balance_loss_clip": 1.0115521, + "balance_loss_mlp": 1.01736879, + "epoch": 0.8544459808813804, + "flos": 55286630467200.0, + "grad_norm": 1.9560602581043516, + "language_loss": 0.70090616, + "learning_rate": 2.1802231754764987e-07, + "loss": 0.72257447, + "num_input_tokens_seen": 153826530, + "step": 7106, + "time_per_iteration": 3.0372891426086426 + }, + { + "auxiliary_loss_clip": 0.01117388, + "auxiliary_loss_mlp": 0.01027105, + "balance_loss_clip": 0.96985579, + "balance_loss_mlp": 1.0196991, + "epoch": 0.8545662237720195, + "flos": 25776248705280.0, + "grad_norm": 4.53259738437686, + "language_loss": 0.76503885, + "learning_rate": 2.17668779958718e-07, + "loss": 0.78648376, + "num_input_tokens_seen": 153849110, + "step": 7107, + "time_per_iteration": 2.7747833728790283 + }, + { + "auxiliary_loss_clip": 0.01165403, + "auxiliary_loss_mlp": 0.0102562, + "balance_loss_clip": 1.04952109, + "balance_loss_mlp": 1.01857185, + "epoch": 0.8546864666626586, + "flos": 11108320427520.0, + "grad_norm": 2.2753295804398452, + "language_loss": 0.80326122, + "learning_rate": 2.1731551274801553e-07, + "loss": 0.82517141, + "num_input_tokens_seen": 153865550, + "step": 7108, + "time_per_iteration": 2.5998544692993164 + }, + { + "auxiliary_loss_clip": 0.0112045, + "auxiliary_loss_mlp": 0.01026709, + "balance_loss_clip": 0.97389179, + "balance_loss_mlp": 1.01921988, + "epoch": 0.8548067095532976, + "flos": 25520169669120.0, + "grad_norm": 2.111598710956447, + "language_loss": 0.61938417, + "learning_rate": 2.169625159691324e-07, + "loss": 0.64085579, + "num_input_tokens_seen": 153885425, + "step": 7109, + "time_per_iteration": 4.54714822769165 + }, + { + "auxiliary_loss_clip": 0.01076313, + "auxiliary_loss_mlp": 0.01021541, + "balance_loss_clip": 0.89367026, + "balance_loss_mlp": 1.01455212, + "epoch": 0.8549269524439368, + "flos": 24717853532160.0, + "grad_norm": 3.113977836861843, + "language_loss": 0.74323136, + "learning_rate": 2.1660978967561784e-07, + "loss": 0.76420987, + "num_input_tokens_seen": 153904760, + "step": 7110, + "time_per_iteration": 2.7971463203430176 + }, + { + "auxiliary_loss_clip": 0.01162515, + "auxiliary_loss_mlp": 0.01020433, + "balance_loss_clip": 1.04687536, + "balance_loss_mlp": 1.01362062, + "epoch": 0.8550471953345758, + "flos": 19825599191040.0, + "grad_norm": 2.329864553190937, + "language_loss": 0.78772402, + "learning_rate": 2.1625733392098035e-07, + "loss": 0.8095535, + "num_input_tokens_seen": 153920370, + "step": 7111, + "time_per_iteration": 2.6207778453826904 + }, + { + "auxiliary_loss_clip": 0.01160383, + "auxiliary_loss_mlp": 0.01021969, + "balance_loss_clip": 1.04483593, + "balance_loss_mlp": 1.01502764, + "epoch": 0.8551674382252149, + "flos": 22820441500800.0, + "grad_norm": 1.733499170561638, + "language_loss": 0.79484499, + "learning_rate": 2.159051487586867e-07, + "loss": 0.81666851, + "num_input_tokens_seen": 153940500, + "step": 7112, + "time_per_iteration": 2.6391937732696533 + }, + { + "auxiliary_loss_clip": 0.01122532, + "auxiliary_loss_mlp": 0.01022379, + "balance_loss_clip": 0.97392929, + "balance_loss_mlp": 1.01497638, + "epoch": 0.8552876811158541, + "flos": 20631255292800.0, + "grad_norm": 2.1134351769953446, + "language_loss": 0.72593558, + "learning_rate": 2.155532342421642e-07, + "loss": 0.74738467, + "num_input_tokens_seen": 153958500, + "step": 7113, + "time_per_iteration": 3.6202807426452637 + }, + { + "auxiliary_loss_clip": 0.01143224, + "auxiliary_loss_mlp": 0.01029525, + "balance_loss_clip": 1.00995243, + "balance_loss_mlp": 1.02191019, + "epoch": 0.8554079240064931, + "flos": 23112359331840.0, + "grad_norm": 1.875495143470282, + "language_loss": 0.78584576, + "learning_rate": 2.1520159042479636e-07, + "loss": 0.80757326, + "num_input_tokens_seen": 153976790, + "step": 7114, + "time_per_iteration": 2.755157470703125 + }, + { + "auxiliary_loss_clip": 0.0114197, + "auxiliary_loss_mlp": 0.01025938, + "balance_loss_clip": 1.01095855, + "balance_loss_mlp": 1.01963496, + "epoch": 0.8555281668971322, + "flos": 22128047959680.0, + "grad_norm": 2.0122879462914653, + "language_loss": 0.71218002, + "learning_rate": 2.148502173599287e-07, + "loss": 0.73385912, + "num_input_tokens_seen": 153994930, + "step": 7115, + "time_per_iteration": 2.5936052799224854 + }, + { + "auxiliary_loss_clip": 0.0111227, + "auxiliary_loss_mlp": 0.01024458, + "balance_loss_clip": 0.97036195, + "balance_loss_mlp": 1.01742125, + "epoch": 0.8556484097877713, + "flos": 31139040234240.0, + "grad_norm": 1.84807972254912, + "language_loss": 0.65785855, + "learning_rate": 2.1449911510086372e-07, + "loss": 0.6792258, + "num_input_tokens_seen": 154014400, + "step": 7116, + "time_per_iteration": 2.749743938446045 + }, + { + "auxiliary_loss_clip": 0.01140056, + "auxiliary_loss_mlp": 0.01028949, + "balance_loss_clip": 1.01012373, + "balance_loss_mlp": 1.02186203, + "epoch": 0.8557686526784104, + "flos": 24316551809280.0, + "grad_norm": 2.067649019349783, + "language_loss": 0.76962292, + "learning_rate": 2.141482837008628e-07, + "loss": 0.79131293, + "num_input_tokens_seen": 154034940, + "step": 7117, + "time_per_iteration": 2.639251470565796 + }, + { + "auxiliary_loss_clip": 0.01134917, + "auxiliary_loss_mlp": 0.01023377, + "balance_loss_clip": 1.00797439, + "balance_loss_mlp": 1.01652288, + "epoch": 0.8558888955690495, + "flos": 17712723427200.0, + "grad_norm": 1.8529607937971246, + "language_loss": 0.72192454, + "learning_rate": 2.1379772321314826e-07, + "loss": 0.74350744, + "num_input_tokens_seen": 154052985, + "step": 7118, + "time_per_iteration": 2.6822726726531982 + }, + { + "auxiliary_loss_clip": 0.01039604, + "auxiliary_loss_mlp": 0.0102335, + "balance_loss_clip": 0.85755336, + "balance_loss_mlp": 1.01650691, + "epoch": 0.8560091384596886, + "flos": 19171702051200.0, + "grad_norm": 2.321296074402951, + "language_loss": 0.81617415, + "learning_rate": 2.1344743369089802e-07, + "loss": 0.83680373, + "num_input_tokens_seen": 154068765, + "step": 7119, + "time_per_iteration": 3.71783447265625 + }, + { + "auxiliary_loss_clip": 0.01121074, + "auxiliary_loss_mlp": 0.01027295, + "balance_loss_clip": 0.97575009, + "balance_loss_mlp": 1.02035117, + "epoch": 0.8561293813503277, + "flos": 23914855036800.0, + "grad_norm": 1.915713500701142, + "language_loss": 0.82253075, + "learning_rate": 2.130974151872522e-07, + "loss": 0.84401447, + "num_input_tokens_seen": 154089100, + "step": 7120, + "time_per_iteration": 2.6815547943115234 + }, + { + "auxiliary_loss_clip": 0.01096301, + "auxiliary_loss_mlp": 0.01026074, + "balance_loss_clip": 0.93509614, + "balance_loss_mlp": 1.01900506, + "epoch": 0.8562496242409667, + "flos": 22529206028160.0, + "grad_norm": 1.723117861922089, + "language_loss": 0.78699279, + "learning_rate": 2.1274766775530773e-07, + "loss": 0.80821657, + "num_input_tokens_seen": 154108965, + "step": 7121, + "time_per_iteration": 2.701556921005249 + }, + { + "auxiliary_loss_clip": 0.0116688, + "auxiliary_loss_mlp": 0.01022889, + "balance_loss_clip": 1.04788911, + "balance_loss_mlp": 1.01551294, + "epoch": 0.8563698671316058, + "flos": 14712745472640.0, + "grad_norm": 2.1711105639121597, + "language_loss": 0.80049348, + "learning_rate": 2.1239819144812077e-07, + "loss": 0.82239115, + "num_input_tokens_seen": 154123425, + "step": 7122, + "time_per_iteration": 2.5891802310943604 + }, + { + "auxiliary_loss_clip": 0.0108648, + "auxiliary_loss_mlp": 0.01024503, + "balance_loss_clip": 0.92976522, + "balance_loss_mlp": 1.01752079, + "epoch": 0.856490110022245, + "flos": 39167768211840.0, + "grad_norm": 1.999934651122167, + "language_loss": 0.70363742, + "learning_rate": 2.1204898631870716e-07, + "loss": 0.72474724, + "num_input_tokens_seen": 154148315, + "step": 7123, + "time_per_iteration": 2.8892900943756104 + }, + { + "auxiliary_loss_clip": 0.01118502, + "auxiliary_loss_mlp": 0.01024271, + "balance_loss_clip": 0.97464991, + "balance_loss_mlp": 1.01746118, + "epoch": 0.856610352912884, + "flos": 29059345658880.0, + "grad_norm": 1.7338875005355356, + "language_loss": 0.76314384, + "learning_rate": 2.1170005242004006e-07, + "loss": 0.78457159, + "num_input_tokens_seen": 154169665, + "step": 7124, + "time_per_iteration": 2.7734477519989014 + }, + { + "auxiliary_loss_clip": 0.01121115, + "auxiliary_loss_mlp": 0.0102532, + "balance_loss_clip": 0.97004318, + "balance_loss_mlp": 1.01887107, + "epoch": 0.8567305958035231, + "flos": 23878333883520.0, + "grad_norm": 2.283821274812699, + "language_loss": 0.78199661, + "learning_rate": 2.1135138980505384e-07, + "loss": 0.80346096, + "num_input_tokens_seen": 154190335, + "step": 7125, + "time_per_iteration": 2.775430679321289 + }, + { + "auxiliary_loss_clip": 0.01115093, + "auxiliary_loss_mlp": 0.01024539, + "balance_loss_clip": 0.97351485, + "balance_loss_mlp": 1.01757991, + "epoch": 0.8568508386941622, + "flos": 22200120599040.0, + "grad_norm": 1.7875354904819796, + "language_loss": 0.72458488, + "learning_rate": 2.110029985266395e-07, + "loss": 0.74598122, + "num_input_tokens_seen": 154210040, + "step": 7126, + "time_per_iteration": 2.7926018238067627 + }, + { + "auxiliary_loss_clip": 0.01120501, + "auxiliary_loss_mlp": 0.01026275, + "balance_loss_clip": 0.9690547, + "balance_loss_mlp": 1.01918173, + "epoch": 0.8569710815848013, + "flos": 17307507121920.0, + "grad_norm": 1.5310995290979537, + "language_loss": 0.73834831, + "learning_rate": 2.1065487863764787e-07, + "loss": 0.75981605, + "num_input_tokens_seen": 154228385, + "step": 7127, + "time_per_iteration": 2.646944284439087 + }, + { + "auxiliary_loss_clip": 0.01060459, + "auxiliary_loss_mlp": 0.0103098, + "balance_loss_clip": 0.88930178, + "balance_loss_mlp": 1.02326703, + "epoch": 0.8570913244754403, + "flos": 23732285184000.0, + "grad_norm": 1.5147858773454457, + "language_loss": 0.85696596, + "learning_rate": 2.1030703019088846e-07, + "loss": 0.87788033, + "num_input_tokens_seen": 154249015, + "step": 7128, + "time_per_iteration": 2.780641794204712 + }, + { + "auxiliary_loss_clip": 0.01134258, + "auxiliary_loss_mlp": 0.01020195, + "balance_loss_clip": 1.0081259, + "balance_loss_mlp": 1.01345122, + "epoch": 0.8572115673660795, + "flos": 20048748433920.0, + "grad_norm": 1.8549194100692716, + "language_loss": 0.70991164, + "learning_rate": 2.099594532391291e-07, + "loss": 0.73145616, + "num_input_tokens_seen": 154267700, + "step": 7129, + "time_per_iteration": 2.6376070976257324 + }, + { + "auxiliary_loss_clip": 0.01130303, + "auxiliary_loss_mlp": 0.01029283, + "balance_loss_clip": 1.00555027, + "balance_loss_mlp": 1.02185905, + "epoch": 0.8573318102567186, + "flos": 27160389342720.0, + "grad_norm": 1.50962867357326, + "language_loss": 0.79234052, + "learning_rate": 2.0961214783509806e-07, + "loss": 0.81393635, + "num_input_tokens_seen": 154290580, + "step": 7130, + "time_per_iteration": 2.7045931816101074 + }, + { + "auxiliary_loss_clip": 0.01119856, + "auxiliary_loss_mlp": 0.01022855, + "balance_loss_clip": 0.96895587, + "balance_loss_mlp": 1.01567256, + "epoch": 0.8574520531473576, + "flos": 24936585402240.0, + "grad_norm": 2.2530683475321114, + "language_loss": 0.74443185, + "learning_rate": 2.0926511403148051e-07, + "loss": 0.76585901, + "num_input_tokens_seen": 154309545, + "step": 7131, + "time_per_iteration": 2.823970079421997 + }, + { + "auxiliary_loss_clip": 0.01103032, + "auxiliary_loss_mlp": 0.01026097, + "balance_loss_clip": 0.93493611, + "balance_loss_mlp": 1.01923394, + "epoch": 0.8575722960379968, + "flos": 18771154513920.0, + "grad_norm": 1.7517427513326598, + "language_loss": 0.76000619, + "learning_rate": 2.0891835188092143e-07, + "loss": 0.7812975, + "num_input_tokens_seen": 154326545, + "step": 7132, + "time_per_iteration": 2.7385170459747314 + }, + { + "auxiliary_loss_clip": 0.01101102, + "auxiliary_loss_mlp": 0.01025111, + "balance_loss_clip": 0.93154037, + "balance_loss_mlp": 1.0182327, + "epoch": 0.8576925389286358, + "flos": 22200300167040.0, + "grad_norm": 2.0971408818633344, + "language_loss": 0.81435728, + "learning_rate": 2.0857186143602434e-07, + "loss": 0.83561945, + "num_input_tokens_seen": 154345190, + "step": 7133, + "time_per_iteration": 2.795560598373413 + }, + { + "auxiliary_loss_clip": 0.01086661, + "auxiliary_loss_mlp": 0.01026405, + "balance_loss_clip": 0.93035704, + "balance_loss_mlp": 1.01963997, + "epoch": 0.8578127818192749, + "flos": 22894345733760.0, + "grad_norm": 2.9426040312638855, + "language_loss": 0.67344189, + "learning_rate": 2.0822564274935094e-07, + "loss": 0.69457251, + "num_input_tokens_seen": 154364615, + "step": 7134, + "time_per_iteration": 2.814117908477783 + }, + { + "auxiliary_loss_clip": 0.0111862, + "auxiliary_loss_mlp": 0.01027637, + "balance_loss_clip": 0.97428954, + "balance_loss_mlp": 1.0198828, + "epoch": 0.8579330247099141, + "flos": 34824839541120.0, + "grad_norm": 2.4240663938454268, + "language_loss": 0.66871679, + "learning_rate": 2.078796958734239e-07, + "loss": 0.69017935, + "num_input_tokens_seen": 154387335, + "step": 7135, + "time_per_iteration": 3.7120110988616943 + }, + { + "auxiliary_loss_clip": 0.01142467, + "auxiliary_loss_mlp": 0.01027109, + "balance_loss_clip": 1.01264954, + "balance_loss_mlp": 1.01968181, + "epoch": 0.8580532676005531, + "flos": 19755681367680.0, + "grad_norm": 1.885418944405714, + "language_loss": 0.7532863, + "learning_rate": 2.0753402086072124e-07, + "loss": 0.77498204, + "num_input_tokens_seen": 154405965, + "step": 7136, + "time_per_iteration": 2.61315655708313 + }, + { + "auxiliary_loss_clip": 0.01037269, + "auxiliary_loss_mlp": 0.01030794, + "balance_loss_clip": 0.81910348, + "balance_loss_mlp": 1.02360559, + "epoch": 0.8581735104911922, + "flos": 22739318634240.0, + "grad_norm": 2.0849457184272326, + "language_loss": 0.75560921, + "learning_rate": 2.071886177636828e-07, + "loss": 0.77628988, + "num_input_tokens_seen": 154422750, + "step": 7137, + "time_per_iteration": 2.8131043910980225 + }, + { + "auxiliary_loss_clip": 0.01137472, + "auxiliary_loss_mlp": 0.01027007, + "balance_loss_clip": 1.01027775, + "balance_loss_mlp": 1.01971185, + "epoch": 0.8582937533818313, + "flos": 23149131880320.0, + "grad_norm": 2.4684525188207718, + "language_loss": 0.83022857, + "learning_rate": 2.0684348663470575e-07, + "loss": 0.8518734, + "num_input_tokens_seen": 154442930, + "step": 7138, + "time_per_iteration": 3.6252918243408203 + }, + { + "auxiliary_loss_clip": 0.01114399, + "auxiliary_loss_mlp": 0.01023079, + "balance_loss_clip": 0.9680723, + "balance_loss_mlp": 1.01540232, + "epoch": 0.8584139962724704, + "flos": 19498668577920.0, + "grad_norm": 1.7101575533920919, + "language_loss": 0.61476552, + "learning_rate": 2.0649862752614555e-07, + "loss": 0.63614029, + "num_input_tokens_seen": 154461640, + "step": 7139, + "time_per_iteration": 2.724086284637451 + }, + { + "auxiliary_loss_clip": 0.01023463, + "auxiliary_loss_mlp": 0.01005917, + "balance_loss_clip": 0.93579453, + "balance_loss_mlp": 1.00491524, + "epoch": 0.8585342391631094, + "flos": 71276577788160.0, + "grad_norm": 0.7539750997145664, + "language_loss": 0.57048464, + "learning_rate": 2.0615404049031838e-07, + "loss": 0.59077847, + "num_input_tokens_seen": 154518610, + "step": 7140, + "time_per_iteration": 3.2223575115203857 + }, + { + "auxiliary_loss_clip": 0.01140703, + "auxiliary_loss_mlp": 0.0102485, + "balance_loss_clip": 1.01090896, + "balance_loss_mlp": 1.01731002, + "epoch": 0.8586544820537486, + "flos": 10815432929280.0, + "grad_norm": 5.749535233769337, + "language_loss": 0.77585453, + "learning_rate": 2.0580972557949616e-07, + "loss": 0.79751009, + "num_input_tokens_seen": 154533700, + "step": 7141, + "time_per_iteration": 2.670638084411621 + }, + { + "auxiliary_loss_clip": 0.01042468, + "auxiliary_loss_mlp": 0.01003096, + "balance_loss_clip": 0.97228843, + "balance_loss_mlp": 1.00205886, + "epoch": 0.8587747249443877, + "flos": 64811184422400.0, + "grad_norm": 0.8010318546060721, + "language_loss": 0.54220665, + "learning_rate": 2.054656828459125e-07, + "loss": 0.56266236, + "num_input_tokens_seen": 154597810, + "step": 7142, + "time_per_iteration": 3.2408699989318848 + }, + { + "auxiliary_loss_clip": 0.01067088, + "auxiliary_loss_mlp": 0.01031006, + "balance_loss_clip": 0.89469606, + "balance_loss_mlp": 1.02324581, + "epoch": 0.8588949678350267, + "flos": 26834607964800.0, + "grad_norm": 1.6068382739255704, + "language_loss": 0.77481312, + "learning_rate": 2.051219123417578e-07, + "loss": 0.79579401, + "num_input_tokens_seen": 154617870, + "step": 7143, + "time_per_iteration": 2.77868914604187 + }, + { + "auxiliary_loss_clip": 0.01164497, + "auxiliary_loss_mlp": 0.01025541, + "balance_loss_clip": 1.04702759, + "balance_loss_mlp": 1.01788807, + "epoch": 0.8590152107256659, + "flos": 26104256726400.0, + "grad_norm": 2.124942884284681, + "language_loss": 0.60085607, + "learning_rate": 2.0477841411918196e-07, + "loss": 0.62275648, + "num_input_tokens_seen": 154637395, + "step": 7144, + "time_per_iteration": 2.6424660682678223 + }, + { + "auxiliary_loss_clip": 0.01135594, + "auxiliary_loss_mlp": 0.01020105, + "balance_loss_clip": 1.00650811, + "balance_loss_mlp": 1.01323533, + "epoch": 0.859135453616305, + "flos": 26140885620480.0, + "grad_norm": 2.0980111001375104, + "language_loss": 0.74451005, + "learning_rate": 2.0443518823029326e-07, + "loss": 0.76606703, + "num_input_tokens_seen": 154657935, + "step": 7145, + "time_per_iteration": 3.6453702449798584 + }, + { + "auxiliary_loss_clip": 0.01087237, + "auxiliary_loss_mlp": 0.01023686, + "balance_loss_clip": 0.9301126, + "balance_loss_mlp": 1.01648545, + "epoch": 0.859255696506944, + "flos": 12969319046400.0, + "grad_norm": 2.1995998340600975, + "language_loss": 0.76661205, + "learning_rate": 2.0409223472715854e-07, + "loss": 0.78772128, + "num_input_tokens_seen": 154675080, + "step": 7146, + "time_per_iteration": 2.740410327911377 + }, + { + "auxiliary_loss_clip": 0.01093755, + "auxiliary_loss_mlp": 0.00509003, + "balance_loss_clip": 0.93262374, + "balance_loss_mlp": 0.0, + "epoch": 0.8593759393975832, + "flos": 18475753063680.0, + "grad_norm": 2.5417711817049464, + "language_loss": 0.74856043, + "learning_rate": 2.0374955366180434e-07, + "loss": 0.764588, + "num_input_tokens_seen": 154692720, + "step": 7147, + "time_per_iteration": 2.6866984367370605 + }, + { + "auxiliary_loss_clip": 0.01096084, + "auxiliary_loss_mlp": 0.01027633, + "balance_loss_clip": 0.93018079, + "balance_loss_mlp": 1.02054012, + "epoch": 0.8594961822882222, + "flos": 22200156512640.0, + "grad_norm": 2.005310655349532, + "language_loss": 0.72864377, + "learning_rate": 2.034071450862147e-07, + "loss": 0.74988097, + "num_input_tokens_seen": 154710190, + "step": 7148, + "time_per_iteration": 2.679108142852783 + }, + { + "auxiliary_loss_clip": 0.0111908, + "auxiliary_loss_mlp": 0.01029506, + "balance_loss_clip": 0.96899158, + "balance_loss_mlp": 1.02210283, + "epoch": 0.8596164251788613, + "flos": 23294749616640.0, + "grad_norm": 2.274098027679146, + "language_loss": 0.76789534, + "learning_rate": 2.030650090523327e-07, + "loss": 0.78938115, + "num_input_tokens_seen": 154729380, + "step": 7149, + "time_per_iteration": 2.6640145778656006 + }, + { + "auxiliary_loss_clip": 0.0109221, + "auxiliary_loss_mlp": 0.01027156, + "balance_loss_clip": 0.93127567, + "balance_loss_mlp": 1.02001238, + "epoch": 0.8597366680695004, + "flos": 31649905416960.0, + "grad_norm": 1.5703068584054105, + "language_loss": 0.59489059, + "learning_rate": 2.0272314561205995e-07, + "loss": 0.61608422, + "num_input_tokens_seen": 154749775, + "step": 7150, + "time_per_iteration": 2.782644271850586 + }, + { + "auxiliary_loss_clip": 0.01084107, + "auxiliary_loss_mlp": 0.01026165, + "balance_loss_clip": 0.92642844, + "balance_loss_mlp": 1.01904249, + "epoch": 0.8598569109601395, + "flos": 21287738211840.0, + "grad_norm": 1.7165951283956848, + "language_loss": 0.72941178, + "learning_rate": 2.023815548172567e-07, + "loss": 0.75051451, + "num_input_tokens_seen": 154769845, + "step": 7151, + "time_per_iteration": 2.8203964233398438 + }, + { + "auxiliary_loss_clip": 0.01138139, + "auxiliary_loss_mlp": 0.01024161, + "balance_loss_clip": 1.00716901, + "balance_loss_mlp": 1.01696038, + "epoch": 0.8599771538507786, + "flos": 25447809720960.0, + "grad_norm": 1.5879708280365499, + "language_loss": 0.66504073, + "learning_rate": 2.0204023671974267e-07, + "loss": 0.68666375, + "num_input_tokens_seen": 154789230, + "step": 7152, + "time_per_iteration": 2.6746416091918945 + }, + { + "auxiliary_loss_clip": 0.01138167, + "auxiliary_loss_mlp": 0.010208, + "balance_loss_clip": 1.00822115, + "balance_loss_mlp": 1.01391268, + "epoch": 0.8600973967414177, + "flos": 16723958768640.0, + "grad_norm": 2.1364391060715433, + "language_loss": 0.80916446, + "learning_rate": 2.0169919137129532e-07, + "loss": 0.83075416, + "num_input_tokens_seen": 154807670, + "step": 7153, + "time_per_iteration": 2.575066328048706 + }, + { + "auxiliary_loss_clip": 0.01141722, + "auxiliary_loss_mlp": 0.01026984, + "balance_loss_clip": 1.01124513, + "balance_loss_mlp": 1.01896727, + "epoch": 0.8602176396320568, + "flos": 25227928615680.0, + "grad_norm": 3.66658745289506, + "language_loss": 0.70672691, + "learning_rate": 2.013584188236508e-07, + "loss": 0.72841394, + "num_input_tokens_seen": 154825575, + "step": 7154, + "time_per_iteration": 2.7135257720947266 + }, + { + "auxiliary_loss_clip": 0.01164735, + "auxiliary_loss_mlp": 0.01028224, + "balance_loss_clip": 1.04762626, + "balance_loss_mlp": 1.02130985, + "epoch": 0.8603378825226958, + "flos": 20412236113920.0, + "grad_norm": 1.6087847272099358, + "language_loss": 0.79456162, + "learning_rate": 2.0101791912850396e-07, + "loss": 0.81649119, + "num_input_tokens_seen": 154845115, + "step": 7155, + "time_per_iteration": 2.604766368865967 + }, + { + "auxiliary_loss_clip": 0.01119366, + "auxiliary_loss_mlp": 0.01028922, + "balance_loss_clip": 0.97426838, + "balance_loss_mlp": 1.02180839, + "epoch": 0.8604581254133349, + "flos": 34930201109760.0, + "grad_norm": 2.3446553429863948, + "language_loss": 0.64081669, + "learning_rate": 2.006776923375082e-07, + "loss": 0.66229951, + "num_input_tokens_seen": 154866770, + "step": 7156, + "time_per_iteration": 2.823759078979492 + }, + { + "auxiliary_loss_clip": 0.0116352, + "auxiliary_loss_mlp": 0.01023601, + "balance_loss_clip": 1.04766512, + "balance_loss_mlp": 1.01653457, + "epoch": 0.860578368303974, + "flos": 22596538072320.0, + "grad_norm": 1.6284834876557104, + "language_loss": 0.71234739, + "learning_rate": 2.003377385022764e-07, + "loss": 0.73421866, + "num_input_tokens_seen": 154885595, + "step": 7157, + "time_per_iteration": 2.6375467777252197 + }, + { + "auxiliary_loss_clip": 0.01118358, + "auxiliary_loss_mlp": 0.01023682, + "balance_loss_clip": 0.9705705, + "balance_loss_mlp": 1.01642823, + "epoch": 0.8606986111946131, + "flos": 21324331192320.0, + "grad_norm": 2.018580778374236, + "language_loss": 0.77387595, + "learning_rate": 1.9999805767437826e-07, + "loss": 0.79529643, + "num_input_tokens_seen": 154904485, + "step": 7158, + "time_per_iteration": 2.6920249462127686 + }, + { + "auxiliary_loss_clip": 0.01108662, + "auxiliary_loss_mlp": 0.0102596, + "balance_loss_clip": 0.96745753, + "balance_loss_mlp": 1.01880193, + "epoch": 0.8608188540852522, + "flos": 28877206769280.0, + "grad_norm": 1.6518166291650571, + "language_loss": 0.71771657, + "learning_rate": 1.9965864990534386e-07, + "loss": 0.73906285, + "num_input_tokens_seen": 154925010, + "step": 7159, + "time_per_iteration": 2.7155771255493164 + }, + { + "auxiliary_loss_clip": 0.01084213, + "auxiliary_loss_mlp": 0.01023387, + "balance_loss_clip": 0.92701638, + "balance_loss_mlp": 1.01628792, + "epoch": 0.8609390969758913, + "flos": 29716187713920.0, + "grad_norm": 1.9333751221134658, + "language_loss": 0.77653337, + "learning_rate": 1.9931951524666092e-07, + "loss": 0.79760933, + "num_input_tokens_seen": 154946100, + "step": 7160, + "time_per_iteration": 2.8193228244781494 + }, + { + "auxiliary_loss_clip": 0.01146609, + "auxiliary_loss_mlp": 0.00509074, + "balance_loss_clip": 1.01098537, + "balance_loss_mlp": 0.0, + "epoch": 0.8610593398665304, + "flos": 21249349551360.0, + "grad_norm": 1.7029338412410244, + "language_loss": 0.81210148, + "learning_rate": 1.9898065374977534e-07, + "loss": 0.82865834, + "num_input_tokens_seen": 154966305, + "step": 7161, + "time_per_iteration": 3.615192413330078 + }, + { + "auxiliary_loss_clip": 0.01092467, + "auxiliary_loss_mlp": 0.01021058, + "balance_loss_clip": 0.93241352, + "balance_loss_mlp": 1.01454973, + "epoch": 0.8611795827571694, + "flos": 14830102183680.0, + "grad_norm": 1.7321415938629199, + "language_loss": 0.72983152, + "learning_rate": 1.9864206546609342e-07, + "loss": 0.75096679, + "num_input_tokens_seen": 154985145, + "step": 7162, + "time_per_iteration": 3.683990001678467 + }, + { + "auxiliary_loss_clip": 0.01162805, + "auxiliary_loss_mlp": 0.0102085, + "balance_loss_clip": 1.04589629, + "balance_loss_mlp": 1.01381671, + "epoch": 0.8612998256478086, + "flos": 24243258107520.0, + "grad_norm": 1.7404157586540665, + "language_loss": 0.84454334, + "learning_rate": 1.983037504469771e-07, + "loss": 0.86637992, + "num_input_tokens_seen": 155003855, + "step": 7163, + "time_per_iteration": 2.70552396774292 + }, + { + "auxiliary_loss_clip": 0.01143634, + "auxiliary_loss_mlp": 0.0102715, + "balance_loss_clip": 1.01062655, + "balance_loss_mlp": 1.01999712, + "epoch": 0.8614200685384477, + "flos": 21252653602560.0, + "grad_norm": 1.5206038027969133, + "language_loss": 0.66526324, + "learning_rate": 1.9796570874374984e-07, + "loss": 0.68697107, + "num_input_tokens_seen": 155023960, + "step": 7164, + "time_per_iteration": 3.6603500843048096 + }, + { + "auxiliary_loss_clip": 0.01118421, + "auxiliary_loss_mlp": 0.01024512, + "balance_loss_clip": 0.97023988, + "balance_loss_mlp": 1.01778316, + "epoch": 0.8615403114290867, + "flos": 20007738080640.0, + "grad_norm": 1.6814109804477821, + "language_loss": 0.7748844, + "learning_rate": 1.976279404076917e-07, + "loss": 0.79631376, + "num_input_tokens_seen": 155043360, + "step": 7165, + "time_per_iteration": 2.6973166465759277 + }, + { + "auxiliary_loss_clip": 0.01093293, + "auxiliary_loss_mlp": 0.01028206, + "balance_loss_clip": 0.93529975, + "balance_loss_mlp": 1.02137518, + "epoch": 0.8616605543197259, + "flos": 29789373674880.0, + "grad_norm": 2.4591022782729537, + "language_loss": 0.76383865, + "learning_rate": 1.9729044549004193e-07, + "loss": 0.78505367, + "num_input_tokens_seen": 155064745, + "step": 7166, + "time_per_iteration": 2.7628238201141357 + }, + { + "auxiliary_loss_clip": 0.01140588, + "auxiliary_loss_mlp": 0.01027724, + "balance_loss_clip": 1.01070619, + "balance_loss_mlp": 1.02073216, + "epoch": 0.8617807972103649, + "flos": 28911609020160.0, + "grad_norm": 2.275775690645358, + "language_loss": 0.70591354, + "learning_rate": 1.9695322404199822e-07, + "loss": 0.72759664, + "num_input_tokens_seen": 155086790, + "step": 7167, + "time_per_iteration": 2.710080623626709 + }, + { + "auxiliary_loss_clip": 0.01122742, + "auxiliary_loss_mlp": 0.01027331, + "balance_loss_clip": 0.97587025, + "balance_loss_mlp": 1.01985025, + "epoch": 0.861901040101004, + "flos": 27673804391040.0, + "grad_norm": 1.8184901075350877, + "language_loss": 0.82204753, + "learning_rate": 1.9661627611471654e-07, + "loss": 0.8435483, + "num_input_tokens_seen": 155106585, + "step": 7168, + "time_per_iteration": 2.720693588256836 + }, + { + "auxiliary_loss_clip": 0.01122063, + "auxiliary_loss_mlp": 0.01024163, + "balance_loss_clip": 0.96997833, + "balance_loss_mlp": 1.01697779, + "epoch": 0.8620212829916432, + "flos": 49748056755840.0, + "grad_norm": 2.533893005703259, + "language_loss": 0.70036268, + "learning_rate": 1.9627960175931246e-07, + "loss": 0.721825, + "num_input_tokens_seen": 155131285, + "step": 7169, + "time_per_iteration": 2.888350486755371 + }, + { + "auxiliary_loss_clip": 0.01142918, + "auxiliary_loss_mlp": 0.01026826, + "balance_loss_clip": 1.01304126, + "balance_loss_mlp": 1.0196619, + "epoch": 0.8621415258822822, + "flos": 21138672769920.0, + "grad_norm": 1.6916598875344662, + "language_loss": 0.7429477, + "learning_rate": 1.9594320102685847e-07, + "loss": 0.7646451, + "num_input_tokens_seen": 155150555, + "step": 7170, + "time_per_iteration": 2.759169578552246 + }, + { + "auxiliary_loss_clip": 0.0110883, + "auxiliary_loss_mlp": 0.00509077, + "balance_loss_clip": 0.97028404, + "balance_loss_mlp": 0.0, + "epoch": 0.8622617687729213, + "flos": 21689039934720.0, + "grad_norm": 2.0393235470926983, + "language_loss": 0.64004332, + "learning_rate": 1.956070739683864e-07, + "loss": 0.6562224, + "num_input_tokens_seen": 155169890, + "step": 7171, + "time_per_iteration": 3.606321334838867 + }, + { + "auxiliary_loss_clip": 0.0107883, + "auxiliary_loss_mlp": 0.01019867, + "balance_loss_clip": 0.92905295, + "balance_loss_mlp": 1.0134151, + "epoch": 0.8623820116635604, + "flos": 26250592734720.0, + "grad_norm": 1.5333459973806558, + "language_loss": 0.7403757, + "learning_rate": 1.9527122063488678e-07, + "loss": 0.76136261, + "num_input_tokens_seen": 155191005, + "step": 7172, + "time_per_iteration": 2.963667869567871 + }, + { + "auxiliary_loss_clip": 0.01114903, + "auxiliary_loss_mlp": 0.01020443, + "balance_loss_clip": 0.96772206, + "balance_loss_mlp": 1.01354098, + "epoch": 0.8625022545541995, + "flos": 19647554451840.0, + "grad_norm": 1.56144043139148, + "language_loss": 0.80461675, + "learning_rate": 1.9493564107730755e-07, + "loss": 0.82597029, + "num_input_tokens_seen": 155211005, + "step": 7173, + "time_per_iteration": 2.7304773330688477 + }, + { + "auxiliary_loss_clip": 0.01108138, + "auxiliary_loss_mlp": 0.01022161, + "balance_loss_clip": 0.96591592, + "balance_loss_mlp": 1.01579833, + "epoch": 0.8626224974448385, + "flos": 21908382336000.0, + "grad_norm": 2.1679842889936447, + "language_loss": 0.61303937, + "learning_rate": 1.9460033534655684e-07, + "loss": 0.63434237, + "num_input_tokens_seen": 155230365, + "step": 7174, + "time_per_iteration": 2.7349278926849365 + }, + { + "auxiliary_loss_clip": 0.01109774, + "auxiliary_loss_mlp": 0.01023306, + "balance_loss_clip": 0.96514487, + "balance_loss_mlp": 1.01601028, + "epoch": 0.8627427403354777, + "flos": 23331198942720.0, + "grad_norm": 1.5061760647848315, + "language_loss": 0.84037411, + "learning_rate": 1.9426530349349978e-07, + "loss": 0.86170495, + "num_input_tokens_seen": 155250815, + "step": 7175, + "time_per_iteration": 2.7237918376922607 + }, + { + "auxiliary_loss_clip": 0.0113803, + "auxiliary_loss_mlp": 0.00508647, + "balance_loss_clip": 1.00744939, + "balance_loss_mlp": 0.0, + "epoch": 0.8628629832261168, + "flos": 16362877299840.0, + "grad_norm": 1.9413636640382914, + "language_loss": 0.65380859, + "learning_rate": 1.9393054556896038e-07, + "loss": 0.67027533, + "num_input_tokens_seen": 155268515, + "step": 7176, + "time_per_iteration": 2.613675117492676 + }, + { + "auxiliary_loss_clip": 0.01090604, + "auxiliary_loss_mlp": 0.01025517, + "balance_loss_clip": 0.93015599, + "balance_loss_mlp": 1.01843596, + "epoch": 0.8629832261167558, + "flos": 28103941756800.0, + "grad_norm": 2.8134010586378335, + "language_loss": 0.68796188, + "learning_rate": 1.9359606162372133e-07, + "loss": 0.70912308, + "num_input_tokens_seen": 155290120, + "step": 7177, + "time_per_iteration": 2.8959643840789795 + }, + { + "auxiliary_loss_clip": 0.01162197, + "auxiliary_loss_mlp": 0.01021313, + "balance_loss_clip": 1.04782653, + "balance_loss_mlp": 1.01473594, + "epoch": 0.863103469007395, + "flos": 20230061310720.0, + "grad_norm": 1.5781570005258079, + "language_loss": 0.70567071, + "learning_rate": 1.9326185170852293e-07, + "loss": 0.7275058, + "num_input_tokens_seen": 155309085, + "step": 7178, + "time_per_iteration": 2.6049487590789795 + }, + { + "auxiliary_loss_clip": 0.01138452, + "auxiliary_loss_mlp": 0.01026838, + "balance_loss_clip": 1.00926447, + "balance_loss_mlp": 1.01979256, + "epoch": 0.863223711898034, + "flos": 24498547044480.0, + "grad_norm": 1.9177897179108836, + "language_loss": 0.72110307, + "learning_rate": 1.9292791587406598e-07, + "loss": 0.74275595, + "num_input_tokens_seen": 155327945, + "step": 7179, + "time_per_iteration": 2.6912765502929688 + }, + { + "auxiliary_loss_clip": 0.01136194, + "auxiliary_loss_mlp": 0.00509221, + "balance_loss_clip": 1.00555134, + "balance_loss_mlp": 0.0, + "epoch": 0.8633439547886731, + "flos": 17675376261120.0, + "grad_norm": 4.369286112032048, + "language_loss": 0.87027049, + "learning_rate": 1.9259425417100661e-07, + "loss": 0.88672459, + "num_input_tokens_seen": 155344060, + "step": 7180, + "time_per_iteration": 2.6414711475372314 + }, + { + "auxiliary_loss_clip": 0.01039832, + "auxiliary_loss_mlp": 0.01026119, + "balance_loss_clip": 0.84769225, + "balance_loss_mlp": 1.01923466, + "epoch": 0.8634641976793123, + "flos": 12895055677440.0, + "grad_norm": 2.0484648310425246, + "language_loss": 0.74993801, + "learning_rate": 1.9226086664996234e-07, + "loss": 0.77059758, + "num_input_tokens_seen": 155362305, + "step": 7181, + "time_per_iteration": 2.7741196155548096 + }, + { + "auxiliary_loss_clip": 0.0111967, + "auxiliary_loss_mlp": 0.01028993, + "balance_loss_clip": 0.97177929, + "balance_loss_mlp": 1.02211773, + "epoch": 0.8635844405699513, + "flos": 23878980328320.0, + "grad_norm": 2.4599416972505437, + "language_loss": 0.7411353, + "learning_rate": 1.9192775336150712e-07, + "loss": 0.76262188, + "num_input_tokens_seen": 155382605, + "step": 7182, + "time_per_iteration": 2.6996803283691406 + }, + { + "auxiliary_loss_clip": 0.01037995, + "auxiliary_loss_mlp": 0.01001015, + "balance_loss_clip": 0.97241765, + "balance_loss_mlp": 1.00004387, + "epoch": 0.8637046834605904, + "flos": 60453387521280.0, + "grad_norm": 0.7706080014840823, + "language_loss": 0.56247759, + "learning_rate": 1.915949143561739e-07, + "loss": 0.58286768, + "num_input_tokens_seen": 155437280, + "step": 7183, + "time_per_iteration": 3.1607065200805664 + }, + { + "auxiliary_loss_clip": 0.01141561, + "auxiliary_loss_mlp": 0.01025385, + "balance_loss_clip": 1.01144695, + "balance_loss_mlp": 1.0179348, + "epoch": 0.8638249263512295, + "flos": 20558751690240.0, + "grad_norm": 3.6905544099278904, + "language_loss": 0.78371245, + "learning_rate": 1.9126234968445498e-07, + "loss": 0.80538201, + "num_input_tokens_seen": 155456970, + "step": 7184, + "time_per_iteration": 2.6929240226745605 + }, + { + "auxiliary_loss_clip": 0.01164261, + "auxiliary_loss_mlp": 0.01029446, + "balance_loss_clip": 1.04778028, + "balance_loss_mlp": 1.02285361, + "epoch": 0.8639451692418686, + "flos": 26615768353920.0, + "grad_norm": 1.4859449626048886, + "language_loss": 0.67678833, + "learning_rate": 1.9093005939679884e-07, + "loss": 0.69872546, + "num_input_tokens_seen": 155478925, + "step": 7185, + "time_per_iteration": 2.618346691131592 + }, + { + "auxiliary_loss_clip": 0.01141228, + "auxiliary_loss_mlp": 0.01030962, + "balance_loss_clip": 1.01082897, + "balance_loss_mlp": 1.02370226, + "epoch": 0.8640654121325076, + "flos": 15122450977920.0, + "grad_norm": 1.7574384858153063, + "language_loss": 0.76150298, + "learning_rate": 1.9059804354361452e-07, + "loss": 0.78322482, + "num_input_tokens_seen": 155496700, + "step": 7186, + "time_per_iteration": 2.6307530403137207 + }, + { + "auxiliary_loss_clip": 0.0111062, + "auxiliary_loss_mlp": 0.01027442, + "balance_loss_clip": 0.96792662, + "balance_loss_mlp": 1.02017343, + "epoch": 0.8641856550231467, + "flos": 31869068250240.0, + "grad_norm": 2.5086615821924476, + "language_loss": 0.70252168, + "learning_rate": 1.902663021752684e-07, + "loss": 0.72390223, + "num_input_tokens_seen": 155518130, + "step": 7187, + "time_per_iteration": 4.543181896209717 + }, + { + "auxiliary_loss_clip": 0.01168599, + "auxiliary_loss_mlp": 0.01025657, + "balance_loss_clip": 1.05047297, + "balance_loss_mlp": 1.01818883, + "epoch": 0.8643058979137859, + "flos": 14976545932800.0, + "grad_norm": 2.0467978848754256, + "language_loss": 0.82307363, + "learning_rate": 1.8993483534208556e-07, + "loss": 0.84501624, + "num_input_tokens_seen": 155537040, + "step": 7188, + "time_per_iteration": 2.5703442096710205 + }, + { + "auxiliary_loss_clip": 0.01112604, + "auxiliary_loss_mlp": 0.01026375, + "balance_loss_clip": 0.9723773, + "balance_loss_mlp": 1.01896071, + "epoch": 0.8644261408044249, + "flos": 13115726881920.0, + "grad_norm": 2.171018661795476, + "language_loss": 0.74675822, + "learning_rate": 1.8960364309434884e-07, + "loss": 0.76814806, + "num_input_tokens_seen": 155554535, + "step": 7189, + "time_per_iteration": 2.7557032108306885 + }, + { + "auxiliary_loss_clip": 0.01042, + "auxiliary_loss_mlp": 0.00508742, + "balance_loss_clip": 0.85514486, + "balance_loss_mlp": 0.0, + "epoch": 0.864546383695064, + "flos": 20850920916480.0, + "grad_norm": 2.0907766131869243, + "language_loss": 0.78319263, + "learning_rate": 1.8927272548229967e-07, + "loss": 0.79870009, + "num_input_tokens_seen": 155574225, + "step": 7190, + "time_per_iteration": 3.777625560760498 + }, + { + "auxiliary_loss_clip": 0.01071574, + "auxiliary_loss_mlp": 0.01034524, + "balance_loss_clip": 0.89550489, + "balance_loss_mlp": 1.02729988, + "epoch": 0.8646666265857031, + "flos": 21324582587520.0, + "grad_norm": 1.5708900872818226, + "language_loss": 0.82958806, + "learning_rate": 1.8894208255613876e-07, + "loss": 0.850649, + "num_input_tokens_seen": 155593540, + "step": 7191, + "time_per_iteration": 2.7109692096710205 + }, + { + "auxiliary_loss_clip": 0.01161702, + "auxiliary_loss_mlp": 0.01024595, + "balance_loss_clip": 1.04676139, + "balance_loss_mlp": 1.0181427, + "epoch": 0.8647868694763422, + "flos": 19750833031680.0, + "grad_norm": 2.1220755833734555, + "language_loss": 0.77756029, + "learning_rate": 1.8861171436602397e-07, + "loss": 0.79942322, + "num_input_tokens_seen": 155610655, + "step": 7192, + "time_per_iteration": 2.633915901184082 + }, + { + "auxiliary_loss_clip": 0.0114054, + "auxiliary_loss_mlp": 0.01020549, + "balance_loss_clip": 1.00978398, + "balance_loss_mlp": 1.01338732, + "epoch": 0.8649071123669813, + "flos": 26176760328960.0, + "grad_norm": 2.1445049877488063, + "language_loss": 0.80476719, + "learning_rate": 1.882816209620719e-07, + "loss": 0.82637805, + "num_input_tokens_seen": 155627365, + "step": 7193, + "time_per_iteration": 2.67862868309021 + }, + { + "auxiliary_loss_clip": 0.01125691, + "auxiliary_loss_mlp": 0.01026745, + "balance_loss_clip": 0.9778989, + "balance_loss_mlp": 1.01955044, + "epoch": 0.8650273552576204, + "flos": 20302888135680.0, + "grad_norm": 1.7361248183146871, + "language_loss": 0.7691977, + "learning_rate": 1.8795180239435738e-07, + "loss": 0.79072201, + "num_input_tokens_seen": 155646220, + "step": 7194, + "time_per_iteration": 2.7048654556274414 + }, + { + "auxiliary_loss_clip": 0.01122266, + "auxiliary_loss_mlp": 0.01025083, + "balance_loss_clip": 0.97182798, + "balance_loss_mlp": 1.01821637, + "epoch": 0.8651475981482595, + "flos": 23951088881280.0, + "grad_norm": 2.4649524005001062, + "language_loss": 0.75692308, + "learning_rate": 1.8762225871291348e-07, + "loss": 0.77839661, + "num_input_tokens_seen": 155662095, + "step": 7195, + "time_per_iteration": 2.661823034286499 + }, + { + "auxiliary_loss_clip": 0.01162284, + "auxiliary_loss_mlp": 0.00508797, + "balance_loss_clip": 1.04674411, + "balance_loss_mlp": 0.0, + "epoch": 0.8652678410388985, + "flos": 21684622561920.0, + "grad_norm": 1.6557206019036348, + "language_loss": 0.80940974, + "learning_rate": 1.8729298996773201e-07, + "loss": 0.8261205, + "num_input_tokens_seen": 155680845, + "step": 7196, + "time_per_iteration": 2.613351345062256 + }, + { + "auxiliary_loss_clip": 0.0103608, + "auxiliary_loss_mlp": 0.00997984, + "balance_loss_clip": 0.97209966, + "balance_loss_mlp": 0.99692327, + "epoch": 0.8653880839295377, + "flos": 65224660855680.0, + "grad_norm": 0.8354863945909069, + "language_loss": 0.61005539, + "learning_rate": 1.8696399620876301e-07, + "loss": 0.63039601, + "num_input_tokens_seen": 155737875, + "step": 7197, + "time_per_iteration": 3.9979617595672607 + }, + { + "auxiliary_loss_clip": 0.01089237, + "auxiliary_loss_mlp": 0.01025485, + "balance_loss_clip": 0.92743742, + "balance_loss_mlp": 1.01736116, + "epoch": 0.8655083268201768, + "flos": 17749172753280.0, + "grad_norm": 2.169177504485948, + "language_loss": 0.79158354, + "learning_rate": 1.866352774859141e-07, + "loss": 0.81273079, + "num_input_tokens_seen": 155753100, + "step": 7198, + "time_per_iteration": 2.674745559692383 + }, + { + "auxiliary_loss_clip": 0.01095868, + "auxiliary_loss_mlp": 0.01021866, + "balance_loss_clip": 0.93011677, + "balance_loss_mlp": 1.01523757, + "epoch": 0.8656285697108158, + "flos": 20703974376960.0, + "grad_norm": 2.326713843336673, + "language_loss": 0.6928013, + "learning_rate": 1.8630683384905188e-07, + "loss": 0.71397865, + "num_input_tokens_seen": 155772430, + "step": 7199, + "time_per_iteration": 2.736748218536377 + }, + { + "auxiliary_loss_clip": 0.01166004, + "auxiliary_loss_mlp": 0.00509473, + "balance_loss_clip": 1.05085254, + "balance_loss_mlp": 0.0, + "epoch": 0.865748812601455, + "flos": 18653833716480.0, + "grad_norm": 2.081808793950144, + "language_loss": 0.89060605, + "learning_rate": 1.8597866534800045e-07, + "loss": 0.90736079, + "num_input_tokens_seen": 155787545, + "step": 7200, + "time_per_iteration": 2.6268513202667236 + }, + { + "auxiliary_loss_clip": 0.01146189, + "auxiliary_loss_mlp": 0.00509315, + "balance_loss_clip": 1.01216412, + "balance_loss_mlp": 0.0, + "epoch": 0.865869055492094, + "flos": 70652554807680.0, + "grad_norm": 2.258261958742309, + "language_loss": 0.74283183, + "learning_rate": 1.8565077203254398e-07, + "loss": 0.7593869, + "num_input_tokens_seen": 155813005, + "step": 7201, + "time_per_iteration": 3.1110527515411377 + }, + { + "auxiliary_loss_clip": 0.01096068, + "auxiliary_loss_mlp": 0.0101804, + "balance_loss_clip": 0.93693638, + "balance_loss_mlp": 1.01051736, + "epoch": 0.8659892983827331, + "flos": 17383961220480.0, + "grad_norm": 2.3881996213252656, + "language_loss": 0.72825611, + "learning_rate": 1.8532315395242203e-07, + "loss": 0.74939722, + "num_input_tokens_seen": 155829455, + "step": 7202, + "time_per_iteration": 2.683335542678833 + }, + { + "auxiliary_loss_clip": 0.01096968, + "auxiliary_loss_mlp": 0.01024458, + "balance_loss_clip": 0.93206763, + "balance_loss_mlp": 1.01752043, + "epoch": 0.8661095412733723, + "flos": 17895221452800.0, + "grad_norm": 6.176875583296049, + "language_loss": 0.72431111, + "learning_rate": 1.849958111573353e-07, + "loss": 0.74552536, + "num_input_tokens_seen": 155848060, + "step": 7203, + "time_per_iteration": 2.6750059127807617 + }, + { + "auxiliary_loss_clip": 0.01160928, + "auxiliary_loss_mlp": 0.01026994, + "balance_loss_clip": 1.04671407, + "balance_loss_mlp": 1.01960611, + "epoch": 0.8662297841640113, + "flos": 18224163227520.0, + "grad_norm": 1.8972243590742752, + "language_loss": 0.64135265, + "learning_rate": 1.8466874369694074e-07, + "loss": 0.66323191, + "num_input_tokens_seen": 155865755, + "step": 7204, + "time_per_iteration": 2.54830002784729 + }, + { + "auxiliary_loss_clip": 0.01093656, + "auxiliary_loss_mlp": 0.01023412, + "balance_loss_clip": 0.92923266, + "balance_loss_mlp": 1.01659918, + "epoch": 0.8663500270546504, + "flos": 16362159027840.0, + "grad_norm": 2.2477717469342577, + "language_loss": 0.70320642, + "learning_rate": 1.843419516208542e-07, + "loss": 0.7243771, + "num_input_tokens_seen": 155882680, + "step": 7205, + "time_per_iteration": 2.6739563941955566 + }, + { + "auxiliary_loss_clip": 0.01144677, + "auxiliary_loss_mlp": 0.01021772, + "balance_loss_clip": 1.0134964, + "balance_loss_mlp": 1.01430333, + "epoch": 0.8664702699452895, + "flos": 17894431353600.0, + "grad_norm": 2.1665042231251848, + "language_loss": 0.80029416, + "learning_rate": 1.8401543497865047e-07, + "loss": 0.82195866, + "num_input_tokens_seen": 155900680, + "step": 7206, + "time_per_iteration": 2.590968370437622 + }, + { + "auxiliary_loss_clip": 0.01140953, + "auxiliary_loss_mlp": 0.00508949, + "balance_loss_clip": 1.00724101, + "balance_loss_mlp": 0.0, + "epoch": 0.8665905128359286, + "flos": 30736373794560.0, + "grad_norm": 2.034062561098525, + "language_loss": 0.63646567, + "learning_rate": 1.836891938198608e-07, + "loss": 0.65296471, + "num_input_tokens_seen": 155921105, + "step": 7207, + "time_per_iteration": 2.6952426433563232 + }, + { + "auxiliary_loss_clip": 0.0111572, + "auxiliary_loss_mlp": 0.01023873, + "balance_loss_clip": 0.97082394, + "balance_loss_mlp": 1.01694643, + "epoch": 0.8667107557265676, + "flos": 18656419495680.0, + "grad_norm": 4.968706379306582, + "language_loss": 0.71444595, + "learning_rate": 1.8336322819397677e-07, + "loss": 0.73584187, + "num_input_tokens_seen": 155938640, + "step": 7208, + "time_per_iteration": 2.6147398948669434 + }, + { + "auxiliary_loss_clip": 0.01094353, + "auxiliary_loss_mlp": 0.01025881, + "balance_loss_clip": 0.92853105, + "balance_loss_mlp": 1.01821613, + "epoch": 0.8668309986172068, + "flos": 20083725302400.0, + "grad_norm": 2.2925328325320344, + "language_loss": 0.63042819, + "learning_rate": 1.8303753815044654e-07, + "loss": 0.65163058, + "num_input_tokens_seen": 155957945, + "step": 7209, + "time_per_iteration": 2.679046154022217 + }, + { + "auxiliary_loss_clip": 0.01128452, + "auxiliary_loss_mlp": 0.01029148, + "balance_loss_clip": 0.97205865, + "balance_loss_mlp": 1.02144694, + "epoch": 0.8669512415078459, + "flos": 21615099788160.0, + "grad_norm": 2.510897559981825, + "language_loss": 0.70567214, + "learning_rate": 1.827121237386773e-07, + "loss": 0.72724819, + "num_input_tokens_seen": 155975390, + "step": 7210, + "time_per_iteration": 2.651294708251953 + }, + { + "auxiliary_loss_clip": 0.01117824, + "auxiliary_loss_mlp": 0.0102816, + "balance_loss_clip": 0.96949172, + "balance_loss_mlp": 1.02065563, + "epoch": 0.8670714843984849, + "flos": 17703601372800.0, + "grad_norm": 2.7206073534417365, + "language_loss": 0.74961716, + "learning_rate": 1.8238698500803374e-07, + "loss": 0.77107704, + "num_input_tokens_seen": 155988155, + "step": 7211, + "time_per_iteration": 2.599130153656006 + }, + { + "auxiliary_loss_clip": 0.01042093, + "auxiliary_loss_mlp": 0.00999769, + "balance_loss_clip": 0.97221833, + "balance_loss_mlp": 0.99880332, + "epoch": 0.8671917272891241, + "flos": 60705483125760.0, + "grad_norm": 0.7511219577574586, + "language_loss": 0.56263644, + "learning_rate": 1.820621220078391e-07, + "loss": 0.58305502, + "num_input_tokens_seen": 156052065, + "step": 7212, + "time_per_iteration": 3.243030071258545 + }, + { + "auxiliary_loss_clip": 0.01163896, + "auxiliary_loss_mlp": 0.01027534, + "balance_loss_clip": 1.04788089, + "balance_loss_mlp": 1.02020252, + "epoch": 0.8673119701797631, + "flos": 20451881750400.0, + "grad_norm": 1.9582757944713058, + "language_loss": 0.6784966, + "learning_rate": 1.8173753478737553e-07, + "loss": 0.70041084, + "num_input_tokens_seen": 156072500, + "step": 7213, + "time_per_iteration": 3.466503620147705 + }, + { + "auxiliary_loss_clip": 0.01163825, + "auxiliary_loss_mlp": 0.01023904, + "balance_loss_clip": 1.04785216, + "balance_loss_mlp": 1.01680195, + "epoch": 0.8674322130704022, + "flos": 19647410797440.0, + "grad_norm": 2.0187000242640014, + "language_loss": 0.79959643, + "learning_rate": 1.8141322339588205e-07, + "loss": 0.82147378, + "num_input_tokens_seen": 156089840, + "step": 7214, + "time_per_iteration": 2.621060371398926 + }, + { + "auxiliary_loss_clip": 0.01162854, + "auxiliary_loss_mlp": 0.01025683, + "balance_loss_clip": 1.04825187, + "balance_loss_mlp": 1.01906133, + "epoch": 0.8675524559610414, + "flos": 26025001367040.0, + "grad_norm": 19.95816256184391, + "language_loss": 0.70414245, + "learning_rate": 1.810891878825569e-07, + "loss": 0.72602785, + "num_input_tokens_seen": 156109815, + "step": 7215, + "time_per_iteration": 2.623398780822754 + }, + { + "auxiliary_loss_clip": 0.01113883, + "auxiliary_loss_mlp": 0.0102442, + "balance_loss_clip": 0.96808147, + "balance_loss_mlp": 1.01735711, + "epoch": 0.8676726988516804, + "flos": 15049444584960.0, + "grad_norm": 2.6276237681671653, + "language_loss": 0.71811777, + "learning_rate": 1.8076542829655561e-07, + "loss": 0.73950082, + "num_input_tokens_seen": 156128620, + "step": 7216, + "time_per_iteration": 2.740483522415161 + }, + { + "auxiliary_loss_clip": 0.01119026, + "auxiliary_loss_mlp": 0.01028587, + "balance_loss_clip": 0.97440368, + "balance_loss_mlp": 1.02073741, + "epoch": 0.8677929417423195, + "flos": 16288111140480.0, + "grad_norm": 2.5665869724522934, + "language_loss": 0.79177833, + "learning_rate": 1.8044194468699203e-07, + "loss": 0.81325442, + "num_input_tokens_seen": 156145930, + "step": 7217, + "time_per_iteration": 3.601229667663574 + }, + { + "auxiliary_loss_clip": 0.01114134, + "auxiliary_loss_mlp": 0.01029664, + "balance_loss_clip": 0.9722122, + "balance_loss_mlp": 1.02240443, + "epoch": 0.8679131846329585, + "flos": 18844160906880.0, + "grad_norm": 3.294834199566424, + "language_loss": 0.75824732, + "learning_rate": 1.8011873710293912e-07, + "loss": 0.77968532, + "num_input_tokens_seen": 156164435, + "step": 7218, + "time_per_iteration": 2.7676467895507812 + }, + { + "auxiliary_loss_clip": 0.01140975, + "auxiliary_loss_mlp": 0.01023729, + "balance_loss_clip": 1.01152229, + "balance_loss_mlp": 1.01629055, + "epoch": 0.8680334275235977, + "flos": 33620718890880.0, + "grad_norm": 1.7242895751386145, + "language_loss": 0.69351494, + "learning_rate": 1.7979580559342677e-07, + "loss": 0.71516204, + "num_input_tokens_seen": 156185165, + "step": 7219, + "time_per_iteration": 2.9395272731781006 + }, + { + "auxiliary_loss_clip": 0.01116704, + "auxiliary_loss_mlp": 0.01026552, + "balance_loss_clip": 0.97143072, + "balance_loss_mlp": 1.01989436, + "epoch": 0.8681536704142367, + "flos": 24681152810880.0, + "grad_norm": 1.7067307643007357, + "language_loss": 0.66913569, + "learning_rate": 1.7947315020744358e-07, + "loss": 0.69056833, + "num_input_tokens_seen": 156206260, + "step": 7220, + "time_per_iteration": 2.6546871662139893 + }, + { + "auxiliary_loss_clip": 0.01115192, + "auxiliary_loss_mlp": 0.01021229, + "balance_loss_clip": 0.96958697, + "balance_loss_mlp": 1.01413882, + "epoch": 0.8682739133048758, + "flos": 20011042131840.0, + "grad_norm": 1.8278029834905745, + "language_loss": 0.80181772, + "learning_rate": 1.7915077099393594e-07, + "loss": 0.82318193, + "num_input_tokens_seen": 156222860, + "step": 7221, + "time_per_iteration": 2.664224624633789 + }, + { + "auxiliary_loss_clip": 0.011433, + "auxiliary_loss_mlp": 0.01027454, + "balance_loss_clip": 1.00898254, + "balance_loss_mlp": 1.02015817, + "epoch": 0.868394156195515, + "flos": 16654759217280.0, + "grad_norm": 2.7161027931970265, + "language_loss": 0.73326778, + "learning_rate": 1.788286680018083e-07, + "loss": 0.75497538, + "num_input_tokens_seen": 156241570, + "step": 7222, + "time_per_iteration": 2.632756233215332 + }, + { + "auxiliary_loss_clip": 0.01119398, + "auxiliary_loss_mlp": 0.01026357, + "balance_loss_clip": 0.97016656, + "balance_loss_mlp": 1.01932621, + "epoch": 0.868514399086154, + "flos": 28001381448960.0, + "grad_norm": 1.6082910388971237, + "language_loss": 0.72287929, + "learning_rate": 1.7850684127992443e-07, + "loss": 0.74433684, + "num_input_tokens_seen": 156261315, + "step": 7223, + "time_per_iteration": 3.553680658340454 + }, + { + "auxiliary_loss_clip": 0.01094982, + "auxiliary_loss_mlp": 0.01029311, + "balance_loss_clip": 0.93470621, + "balance_loss_mlp": 1.02215207, + "epoch": 0.8686346419767931, + "flos": 20084587228800.0, + "grad_norm": 1.6506044048610902, + "language_loss": 0.70390022, + "learning_rate": 1.7818529087710378e-07, + "loss": 0.72514307, + "num_input_tokens_seen": 156281670, + "step": 7224, + "time_per_iteration": 2.74075984954834 + }, + { + "auxiliary_loss_clip": 0.01138311, + "auxiliary_loss_mlp": 0.00509193, + "balance_loss_clip": 1.00798905, + "balance_loss_mlp": 0.0, + "epoch": 0.8687548848674322, + "flos": 18223516782720.0, + "grad_norm": 1.6858260044368196, + "language_loss": 0.8417961, + "learning_rate": 1.7786401684212637e-07, + "loss": 0.85827112, + "num_input_tokens_seen": 156300500, + "step": 7225, + "time_per_iteration": 2.627263069152832 + }, + { + "auxiliary_loss_clip": 0.0100279, + "auxiliary_loss_mlp": 0.01014231, + "balance_loss_clip": 0.90159804, + "balance_loss_mlp": 1.01312864, + "epoch": 0.8688751277580713, + "flos": 70457885049600.0, + "grad_norm": 0.7343026668610554, + "language_loss": 0.55894476, + "learning_rate": 1.7754301922372883e-07, + "loss": 0.57911497, + "num_input_tokens_seen": 156350145, + "step": 7226, + "time_per_iteration": 3.134162425994873 + }, + { + "auxiliary_loss_clip": 0.01054251, + "auxiliary_loss_mlp": 0.01025041, + "balance_loss_clip": 0.85828352, + "balance_loss_mlp": 1.01784921, + "epoch": 0.8689953706487104, + "flos": 26906788344960.0, + "grad_norm": 1.9031268517700604, + "language_loss": 0.80991787, + "learning_rate": 1.7722229807060617e-07, + "loss": 0.83071077, + "num_input_tokens_seen": 156368725, + "step": 7227, + "time_per_iteration": 2.8511252403259277 + }, + { + "auxiliary_loss_clip": 0.0108401, + "auxiliary_loss_mlp": 0.01024212, + "balance_loss_clip": 0.92817092, + "balance_loss_mlp": 1.016904, + "epoch": 0.8691156135393495, + "flos": 34637385438720.0, + "grad_norm": 2.1591699771367665, + "language_loss": 0.81873703, + "learning_rate": 1.7690185343141172e-07, + "loss": 0.83981925, + "num_input_tokens_seen": 156388640, + "step": 7228, + "time_per_iteration": 2.8078227043151855 + }, + { + "auxiliary_loss_clip": 0.01114446, + "auxiliary_loss_mlp": 0.01021564, + "balance_loss_clip": 0.96721411, + "balance_loss_mlp": 1.01470327, + "epoch": 0.8692358564299886, + "flos": 18989814556800.0, + "grad_norm": 2.048335234370079, + "language_loss": 0.70192206, + "learning_rate": 1.7658168535475615e-07, + "loss": 0.72328222, + "num_input_tokens_seen": 156406425, + "step": 7229, + "time_per_iteration": 2.693288564682007 + }, + { + "auxiliary_loss_clip": 0.01121494, + "auxiliary_loss_mlp": 0.01028396, + "balance_loss_clip": 0.97224224, + "balance_loss_mlp": 1.02094567, + "epoch": 0.8693560993206276, + "flos": 30370839039360.0, + "grad_norm": 1.754765220940865, + "language_loss": 0.64535177, + "learning_rate": 1.7626179388920948e-07, + "loss": 0.66685063, + "num_input_tokens_seen": 156427705, + "step": 7230, + "time_per_iteration": 2.7451934814453125 + }, + { + "auxiliary_loss_clip": 0.01118183, + "auxiliary_loss_mlp": 0.00509442, + "balance_loss_clip": 0.97183388, + "balance_loss_mlp": 0.0, + "epoch": 0.8694763422112668, + "flos": 27200430028800.0, + "grad_norm": 1.9479806824722945, + "language_loss": 0.8076117, + "learning_rate": 1.7594217908329866e-07, + "loss": 0.823888, + "num_input_tokens_seen": 156449890, + "step": 7231, + "time_per_iteration": 2.763690948486328 + }, + { + "auxiliary_loss_clip": 0.01110025, + "auxiliary_loss_mlp": 0.01024558, + "balance_loss_clip": 0.97024369, + "balance_loss_mlp": 1.01744401, + "epoch": 0.8695965851019059, + "flos": 26139161767680.0, + "grad_norm": 3.0835913615849995, + "language_loss": 0.73842812, + "learning_rate": 1.7562284098550895e-07, + "loss": 0.75977397, + "num_input_tokens_seen": 156469600, + "step": 7232, + "time_per_iteration": 2.6795146465301514 + }, + { + "auxiliary_loss_clip": 0.01014724, + "auxiliary_loss_mlp": 0.0100166, + "balance_loss_clip": 0.93632817, + "balance_loss_mlp": 1.00043762, + "epoch": 0.8697168279925449, + "flos": 67332616456320.0, + "grad_norm": 0.8359171665485053, + "language_loss": 0.62259597, + "learning_rate": 1.753037796442838e-07, + "loss": 0.64275986, + "num_input_tokens_seen": 156529040, + "step": 7233, + "time_per_iteration": 3.2344181537628174 + }, + { + "auxiliary_loss_clip": 0.01164491, + "auxiliary_loss_mlp": 0.01027326, + "balance_loss_clip": 1.04742324, + "balance_loss_mlp": 1.02044773, + "epoch": 0.8698370708831841, + "flos": 19718693337600.0, + "grad_norm": 2.0791964396751963, + "language_loss": 0.75291657, + "learning_rate": 1.74984995108024e-07, + "loss": 0.77483475, + "num_input_tokens_seen": 156546970, + "step": 7234, + "time_per_iteration": 2.5607945919036865 + }, + { + "auxiliary_loss_clip": 0.01143169, + "auxiliary_loss_mlp": 0.01027048, + "balance_loss_clip": 1.01049924, + "balance_loss_mlp": 1.01975846, + "epoch": 0.8699573137738231, + "flos": 12859971068160.0, + "grad_norm": 2.0104017325988828, + "language_loss": 0.83074772, + "learning_rate": 1.7466648742508981e-07, + "loss": 0.85244989, + "num_input_tokens_seen": 156563155, + "step": 7235, + "time_per_iteration": 2.6321184635162354 + }, + { + "auxiliary_loss_clip": 0.01115393, + "auxiliary_loss_mlp": 0.01025263, + "balance_loss_clip": 0.97258317, + "balance_loss_mlp": 1.01828337, + "epoch": 0.8700775566644622, + "flos": 17420733768960.0, + "grad_norm": 3.9349335882115675, + "language_loss": 0.84589994, + "learning_rate": 1.7434825664379837e-07, + "loss": 0.86730647, + "num_input_tokens_seen": 156581660, + "step": 7236, + "time_per_iteration": 2.61635160446167 + }, + { + "auxiliary_loss_clip": 0.01142972, + "auxiliary_loss_mlp": 0.01022236, + "balance_loss_clip": 1.01039767, + "balance_loss_mlp": 1.01504767, + "epoch": 0.8701977995551013, + "flos": 13735221770880.0, + "grad_norm": 2.7650563263071244, + "language_loss": 0.86037064, + "learning_rate": 1.740303028124246e-07, + "loss": 0.88202268, + "num_input_tokens_seen": 156597720, + "step": 7237, + "time_per_iteration": 2.6134657859802246 + }, + { + "auxiliary_loss_clip": 0.01040915, + "auxiliary_loss_mlp": 0.01028529, + "balance_loss_clip": 0.85609961, + "balance_loss_mlp": 1.02131927, + "epoch": 0.8703180424457404, + "flos": 30555707362560.0, + "grad_norm": 1.975169212920377, + "language_loss": 0.75864398, + "learning_rate": 1.7371262597920212e-07, + "loss": 0.77933848, + "num_input_tokens_seen": 156619780, + "step": 7238, + "time_per_iteration": 3.7103371620178223 + }, + { + "auxiliary_loss_clip": 0.01071825, + "auxiliary_loss_mlp": 0.01034258, + "balance_loss_clip": 0.89854789, + "balance_loss_mlp": 1.02689123, + "epoch": 0.8704382853363795, + "flos": 19608986223360.0, + "grad_norm": 1.5562062309080988, + "language_loss": 0.76508045, + "learning_rate": 1.7339522619232195e-07, + "loss": 0.78614128, + "num_input_tokens_seen": 156638160, + "step": 7239, + "time_per_iteration": 2.7649335861206055 + }, + { + "auxiliary_loss_clip": 0.01124317, + "auxiliary_loss_mlp": 0.01029659, + "balance_loss_clip": 0.97020036, + "balance_loss_mlp": 1.02210963, + "epoch": 0.8705585282270186, + "flos": 26613900846720.0, + "grad_norm": 2.7835642402834333, + "language_loss": 0.75692648, + "learning_rate": 1.730781034999338e-07, + "loss": 0.77846622, + "num_input_tokens_seen": 156659740, + "step": 7240, + "time_per_iteration": 3.719055652618408 + }, + { + "auxiliary_loss_clip": 0.01159776, + "auxiliary_loss_mlp": 0.01021159, + "balance_loss_clip": 1.04795051, + "balance_loss_mlp": 1.01460266, + "epoch": 0.8706787711176577, + "flos": 34090465979520.0, + "grad_norm": 2.10651736777975, + "language_loss": 0.73324519, + "learning_rate": 1.7276125795014497e-07, + "loss": 0.75505459, + "num_input_tokens_seen": 156678190, + "step": 7241, + "time_per_iteration": 2.692657947540283 + }, + { + "auxiliary_loss_clip": 0.01120467, + "auxiliary_loss_mlp": 0.01023997, + "balance_loss_clip": 0.96946627, + "balance_loss_mlp": 1.01636136, + "epoch": 0.8707990140082967, + "flos": 14611513968000.0, + "grad_norm": 1.9271190785253782, + "language_loss": 0.67469919, + "learning_rate": 1.7244468959102054e-07, + "loss": 0.69614387, + "num_input_tokens_seen": 156695245, + "step": 7242, + "time_per_iteration": 2.6770853996276855 + }, + { + "auxiliary_loss_clip": 0.01140235, + "auxiliary_loss_mlp": 0.01021223, + "balance_loss_clip": 1.01058018, + "balance_loss_mlp": 1.01453507, + "epoch": 0.8709192568989359, + "flos": 20084156265600.0, + "grad_norm": 2.445813864842841, + "language_loss": 0.85291409, + "learning_rate": 1.7212839847058348e-07, + "loss": 0.87452871, + "num_input_tokens_seen": 156710375, + "step": 7243, + "time_per_iteration": 3.664158821105957 + }, + { + "auxiliary_loss_clip": 0.01056176, + "auxiliary_loss_mlp": 0.010213, + "balance_loss_clip": 0.85537022, + "balance_loss_mlp": 1.01407278, + "epoch": 0.871039499789575, + "flos": 16727083251840.0, + "grad_norm": 1.7127016432172586, + "language_loss": 0.73589778, + "learning_rate": 1.718123846368147e-07, + "loss": 0.7566725, + "num_input_tokens_seen": 156729420, + "step": 7244, + "time_per_iteration": 2.793038845062256 + }, + { + "auxiliary_loss_clip": 0.01116497, + "auxiliary_loss_mlp": 0.00508655, + "balance_loss_clip": 0.97206426, + "balance_loss_mlp": 0.0, + "epoch": 0.871159742680214, + "flos": 21068790860160.0, + "grad_norm": 1.7328718478778238, + "language_loss": 0.71571589, + "learning_rate": 1.714966481376543e-07, + "loss": 0.73196745, + "num_input_tokens_seen": 156746100, + "step": 7245, + "time_per_iteration": 2.7180724143981934 + }, + { + "auxiliary_loss_clip": 0.01138151, + "auxiliary_loss_mlp": 0.01027512, + "balance_loss_clip": 1.00843179, + "balance_loss_mlp": 1.02056217, + "epoch": 0.8712799855708532, + "flos": 28256526731520.0, + "grad_norm": 2.5681295509556286, + "language_loss": 0.82905722, + "learning_rate": 1.7118118902099797e-07, + "loss": 0.85071385, + "num_input_tokens_seen": 156764185, + "step": 7246, + "time_per_iteration": 2.654289960861206 + }, + { + "auxiliary_loss_clip": 0.01141594, + "auxiliary_loss_mlp": 0.01030056, + "balance_loss_clip": 1.00987363, + "balance_loss_mlp": 1.02326131, + "epoch": 0.8714002284614922, + "flos": 22236677665920.0, + "grad_norm": 2.0954450211361952, + "language_loss": 0.8072409, + "learning_rate": 1.7086600733470146e-07, + "loss": 0.82895732, + "num_input_tokens_seen": 156784855, + "step": 7247, + "time_per_iteration": 2.6519088745117188 + }, + { + "auxiliary_loss_clip": 0.01137916, + "auxiliary_loss_mlp": 0.01023711, + "balance_loss_clip": 1.00914741, + "balance_loss_mlp": 1.01708627, + "epoch": 0.8715204713521313, + "flos": 21431919404160.0, + "grad_norm": 2.0109431886998252, + "language_loss": 0.76857471, + "learning_rate": 1.7055110312657738e-07, + "loss": 0.79019094, + "num_input_tokens_seen": 156804350, + "step": 7248, + "time_per_iteration": 2.6408824920654297 + }, + { + "auxiliary_loss_clip": 0.011144, + "auxiliary_loss_mlp": 0.01027197, + "balance_loss_clip": 0.97141969, + "balance_loss_mlp": 1.02050972, + "epoch": 0.8716407142427703, + "flos": 23440439180160.0, + "grad_norm": 4.190639713488522, + "language_loss": 0.74549043, + "learning_rate": 1.702364764443962e-07, + "loss": 0.76690638, + "num_input_tokens_seen": 156823425, + "step": 7249, + "time_per_iteration": 3.6112430095672607 + }, + { + "auxiliary_loss_clip": 0.01044297, + "auxiliary_loss_mlp": 0.01026245, + "balance_loss_clip": 0.85399497, + "balance_loss_mlp": 1.0186094, + "epoch": 0.8717609571334095, + "flos": 27958683156480.0, + "grad_norm": 1.9824722402172597, + "language_loss": 0.72296274, + "learning_rate": 1.6992212733588685e-07, + "loss": 0.74366814, + "num_input_tokens_seen": 156843090, + "step": 7250, + "time_per_iteration": 2.807070255279541 + }, + { + "auxiliary_loss_clip": 0.01111811, + "auxiliary_loss_mlp": 0.0102581, + "balance_loss_clip": 0.96864516, + "balance_loss_mlp": 1.0188601, + "epoch": 0.8718812000240486, + "flos": 25479482538240.0, + "grad_norm": 1.9981510006155765, + "language_loss": 0.7498188, + "learning_rate": 1.6960805584873538e-07, + "loss": 0.77119505, + "num_input_tokens_seen": 156861090, + "step": 7251, + "time_per_iteration": 2.6769371032714844 + }, + { + "auxiliary_loss_clip": 0.01071329, + "auxiliary_loss_mlp": 0.01026728, + "balance_loss_clip": 0.89236039, + "balance_loss_mlp": 1.02005851, + "epoch": 0.8720014429146876, + "flos": 23403056100480.0, + "grad_norm": 1.5804734632543875, + "language_loss": 0.7822057, + "learning_rate": 1.6929426203058684e-07, + "loss": 0.8031863, + "num_input_tokens_seen": 156881515, + "step": 7252, + "time_per_iteration": 2.8178064823150635 + }, + { + "auxiliary_loss_clip": 0.01166498, + "auxiliary_loss_mlp": 0.00509889, + "balance_loss_clip": 1.04748487, + "balance_loss_mlp": 0.0, + "epoch": 0.8721216858053268, + "flos": 24352821567360.0, + "grad_norm": 2.363974385004603, + "language_loss": 0.80354285, + "learning_rate": 1.689807459290431e-07, + "loss": 0.82030678, + "num_input_tokens_seen": 156900170, + "step": 7253, + "time_per_iteration": 2.6258554458618164 + }, + { + "auxiliary_loss_clip": 0.01120956, + "auxiliary_loss_mlp": 0.01024278, + "balance_loss_clip": 0.97402167, + "balance_loss_mlp": 1.01745033, + "epoch": 0.8722419286959658, + "flos": 33869687034240.0, + "grad_norm": 2.307206451652315, + "language_loss": 0.70785642, + "learning_rate": 1.6866750759166437e-07, + "loss": 0.72930878, + "num_input_tokens_seen": 156920150, + "step": 7254, + "time_per_iteration": 2.8259425163269043 + }, + { + "auxiliary_loss_clip": 0.01087387, + "auxiliary_loss_mlp": 0.01026, + "balance_loss_clip": 0.92773312, + "balance_loss_mlp": 1.01929164, + "epoch": 0.8723621715866049, + "flos": 18369385914240.0, + "grad_norm": 2.3618636400757347, + "language_loss": 0.77361584, + "learning_rate": 1.6835454706596865e-07, + "loss": 0.79474974, + "num_input_tokens_seen": 156937980, + "step": 7255, + "time_per_iteration": 2.708254814147949 + }, + { + "auxiliary_loss_clip": 0.01165565, + "auxiliary_loss_mlp": 0.01027529, + "balance_loss_clip": 1.0496726, + "balance_loss_mlp": 1.02062976, + "epoch": 0.8724824144772441, + "flos": 22013348855040.0, + "grad_norm": 2.230926190509214, + "language_loss": 0.73523104, + "learning_rate": 1.680418643994317e-07, + "loss": 0.75716203, + "num_input_tokens_seen": 156956550, + "step": 7256, + "time_per_iteration": 2.6362786293029785 + }, + { + "auxiliary_loss_clip": 0.01059997, + "auxiliary_loss_mlp": 0.01001373, + "balance_loss_clip": 1.00933611, + "balance_loss_mlp": 1.0004313, + "epoch": 0.8726026573678831, + "flos": 66698720213760.0, + "grad_norm": 0.8872890042777191, + "language_loss": 0.64561999, + "learning_rate": 1.6772945963948738e-07, + "loss": 0.66623366, + "num_input_tokens_seen": 157014715, + "step": 7257, + "time_per_iteration": 3.1811814308166504 + }, + { + "auxiliary_loss_clip": 0.01117431, + "auxiliary_loss_mlp": 0.01022888, + "balance_loss_clip": 0.97206998, + "balance_loss_mlp": 1.01519275, + "epoch": 0.8727229002585222, + "flos": 13370908078080.0, + "grad_norm": 2.227227340329101, + "language_loss": 0.77683985, + "learning_rate": 1.6741733283352733e-07, + "loss": 0.79824305, + "num_input_tokens_seen": 157032320, + "step": 7258, + "time_per_iteration": 2.7203540802001953 + }, + { + "auxiliary_loss_clip": 0.01076398, + "auxiliary_loss_mlp": 0.01026574, + "balance_loss_clip": 0.89433646, + "balance_loss_mlp": 1.01908159, + "epoch": 0.8728431431491613, + "flos": 21796987282560.0, + "grad_norm": 1.5737216468531896, + "language_loss": 0.83803332, + "learning_rate": 1.6710548402890102e-07, + "loss": 0.85906291, + "num_input_tokens_seen": 157052845, + "step": 7259, + "time_per_iteration": 2.7161340713500977 + }, + { + "auxiliary_loss_clip": 0.01167106, + "auxiliary_loss_mlp": 0.01021261, + "balance_loss_clip": 1.04731917, + "balance_loss_mlp": 1.01408482, + "epoch": 0.8729633860398004, + "flos": 36173823742080.0, + "grad_norm": 1.7677285580467288, + "language_loss": 0.66317576, + "learning_rate": 1.6679391327291527e-07, + "loss": 0.68505943, + "num_input_tokens_seen": 157074050, + "step": 7260, + "time_per_iteration": 2.713362455368042 + }, + { + "auxiliary_loss_clip": 0.01117476, + "auxiliary_loss_mlp": 0.01021878, + "balance_loss_clip": 0.96873027, + "balance_loss_mlp": 1.01498508, + "epoch": 0.8730836289304394, + "flos": 16359680989440.0, + "grad_norm": 2.601843641120282, + "language_loss": 0.67770231, + "learning_rate": 1.6648262061283492e-07, + "loss": 0.69909585, + "num_input_tokens_seen": 157089350, + "step": 7261, + "time_per_iteration": 2.6699976921081543 + }, + { + "auxiliary_loss_clip": 0.01094388, + "auxiliary_loss_mlp": 0.01024739, + "balance_loss_clip": 0.93015212, + "balance_loss_mlp": 1.01712132, + "epoch": 0.8732038718210786, + "flos": 21215126868480.0, + "grad_norm": 2.553463704639977, + "language_loss": 0.73469186, + "learning_rate": 1.6617160609588353e-07, + "loss": 0.75588316, + "num_input_tokens_seen": 157108525, + "step": 7262, + "time_per_iteration": 2.7069883346557617 + }, + { + "auxiliary_loss_clip": 0.01120448, + "auxiliary_loss_mlp": 0.01025373, + "balance_loss_clip": 0.97124124, + "balance_loss_mlp": 1.01857805, + "epoch": 0.8733241147117177, + "flos": 16610696208000.0, + "grad_norm": 1.9651219319437767, + "language_loss": 0.71810293, + "learning_rate": 1.6586086976924163e-07, + "loss": 0.7395612, + "num_input_tokens_seen": 157124025, + "step": 7263, + "time_per_iteration": 2.6309638023376465 + }, + { + "auxiliary_loss_clip": 0.01143526, + "auxiliary_loss_mlp": 0.010277, + "balance_loss_clip": 1.01011515, + "balance_loss_mlp": 1.02105391, + "epoch": 0.8734443576023567, + "flos": 20193935207040.0, + "grad_norm": 1.8082203030250836, + "language_loss": 0.7829895, + "learning_rate": 1.6555041168004747e-07, + "loss": 0.80470181, + "num_input_tokens_seen": 157143345, + "step": 7264, + "time_per_iteration": 3.555130958557129 + }, + { + "auxiliary_loss_clip": 0.01111139, + "auxiliary_loss_mlp": 0.01028112, + "balance_loss_clip": 0.96909547, + "balance_loss_mlp": 1.02108455, + "epoch": 0.8735646004929959, + "flos": 18041162411520.0, + "grad_norm": 1.75719762513501, + "language_loss": 0.69056177, + "learning_rate": 1.6524023187539715e-07, + "loss": 0.7119543, + "num_input_tokens_seen": 157161630, + "step": 7265, + "time_per_iteration": 2.614509344100952 + }, + { + "auxiliary_loss_clip": 0.01118095, + "auxiliary_loss_mlp": 0.0102842, + "balance_loss_clip": 0.97104132, + "balance_loss_mlp": 1.02152658, + "epoch": 0.873684843383635, + "flos": 20262344659200.0, + "grad_norm": 2.301191801182603, + "language_loss": 0.74840033, + "learning_rate": 1.649303304023446e-07, + "loss": 0.76986551, + "num_input_tokens_seen": 157181385, + "step": 7266, + "time_per_iteration": 3.5355465412139893 + }, + { + "auxiliary_loss_clip": 0.01091516, + "auxiliary_loss_mlp": 0.01030146, + "balance_loss_clip": 0.93583441, + "balance_loss_mlp": 1.02332449, + "epoch": 0.873805086274274, + "flos": 16947287579520.0, + "grad_norm": 1.7644456817357737, + "language_loss": 0.78647941, + "learning_rate": 1.6462070730790246e-07, + "loss": 0.80769604, + "num_input_tokens_seen": 157200545, + "step": 7267, + "time_per_iteration": 2.660996675491333 + }, + { + "auxiliary_loss_clip": 0.01109239, + "auxiliary_loss_mlp": 0.0102461, + "balance_loss_clip": 0.96574926, + "balance_loss_mlp": 1.01774025, + "epoch": 0.8739253291649132, + "flos": 18041270152320.0, + "grad_norm": 2.1711686972403266, + "language_loss": 0.78684711, + "learning_rate": 1.6431136263903912e-07, + "loss": 0.80818564, + "num_input_tokens_seen": 157219545, + "step": 7268, + "time_per_iteration": 2.6045262813568115 + }, + { + "auxiliary_loss_clip": 0.01142725, + "auxiliary_loss_mlp": 0.00509225, + "balance_loss_clip": 1.00756216, + "balance_loss_mlp": 0.0, + "epoch": 0.8740455720555522, + "flos": 21325085377920.0, + "grad_norm": 2.4947084559332224, + "language_loss": 0.73555124, + "learning_rate": 1.6400229644268282e-07, + "loss": 0.75207072, + "num_input_tokens_seen": 157237900, + "step": 7269, + "time_per_iteration": 3.5887465476989746 + }, + { + "auxiliary_loss_clip": 0.01093834, + "auxiliary_loss_mlp": 0.0102373, + "balance_loss_clip": 0.93866068, + "balance_loss_mlp": 1.01625574, + "epoch": 0.8741658149461913, + "flos": 15158684822400.0, + "grad_norm": 1.8564529918875619, + "language_loss": 0.81188649, + "learning_rate": 1.6369350876571852e-07, + "loss": 0.83306217, + "num_input_tokens_seen": 157256055, + "step": 7270, + "time_per_iteration": 2.6659016609191895 + }, + { + "auxiliary_loss_clip": 0.01064572, + "auxiliary_loss_mlp": 0.01025897, + "balance_loss_clip": 0.89013922, + "balance_loss_mlp": 1.01932907, + "epoch": 0.8742860578368304, + "flos": 23039855729280.0, + "grad_norm": 1.9715906515799222, + "language_loss": 0.81684101, + "learning_rate": 1.6338499965498874e-07, + "loss": 0.83774573, + "num_input_tokens_seen": 157274785, + "step": 7271, + "time_per_iteration": 2.8634157180786133 + }, + { + "auxiliary_loss_clip": 0.01091063, + "auxiliary_loss_mlp": 0.01026536, + "balance_loss_clip": 0.93162262, + "balance_loss_mlp": 1.01885939, + "epoch": 0.8744063007274695, + "flos": 28145347159680.0, + "grad_norm": 1.6445621272865283, + "language_loss": 0.77659118, + "learning_rate": 1.630767691572943e-07, + "loss": 0.79776716, + "num_input_tokens_seen": 157294805, + "step": 7272, + "time_per_iteration": 2.778129816055298 + }, + { + "auxiliary_loss_clip": 0.01024868, + "auxiliary_loss_mlp": 0.01000344, + "balance_loss_clip": 0.93694544, + "balance_loss_mlp": 0.99941999, + "epoch": 0.8745265436181086, + "flos": 64034076654720.0, + "grad_norm": 0.7398023694720216, + "language_loss": 0.53492212, + "learning_rate": 1.6276881731939306e-07, + "loss": 0.55517423, + "num_input_tokens_seen": 157356695, + "step": 7273, + "time_per_iteration": 3.3387529850006104 + }, + { + "auxiliary_loss_clip": 0.01136627, + "auxiliary_loss_mlp": 0.01023486, + "balance_loss_clip": 1.00931621, + "balance_loss_mlp": 1.01637459, + "epoch": 0.8746467865087477, + "flos": 28658618553600.0, + "grad_norm": 2.2800635343028617, + "language_loss": 0.75639278, + "learning_rate": 1.6246114418800193e-07, + "loss": 0.77799392, + "num_input_tokens_seen": 157376975, + "step": 7274, + "time_per_iteration": 2.6944119930267334 + }, + { + "auxiliary_loss_clip": 0.01134267, + "auxiliary_loss_mlp": 0.01026859, + "balance_loss_clip": 1.00711298, + "balance_loss_mlp": 1.01964068, + "epoch": 0.8747670293993868, + "flos": 23985850268160.0, + "grad_norm": 1.676233279547194, + "language_loss": 0.76810294, + "learning_rate": 1.6215374980979423e-07, + "loss": 0.78971422, + "num_input_tokens_seen": 157397385, + "step": 7275, + "time_per_iteration": 3.538564682006836 + }, + { + "auxiliary_loss_clip": 0.01138001, + "auxiliary_loss_mlp": 0.01030737, + "balance_loss_clip": 1.01119947, + "balance_loss_mlp": 1.02354884, + "epoch": 0.8748872722900258, + "flos": 45221624478720.0, + "grad_norm": 1.8532028772336364, + "language_loss": 0.68735945, + "learning_rate": 1.6184663423140133e-07, + "loss": 0.70904684, + "num_input_tokens_seen": 157417685, + "step": 7276, + "time_per_iteration": 2.8949577808380127 + }, + { + "auxiliary_loss_clip": 0.01075903, + "auxiliary_loss_mlp": 0.01022735, + "balance_loss_clip": 0.89780241, + "balance_loss_mlp": 1.01545155, + "epoch": 0.875007515180665, + "flos": 19754280737280.0, + "grad_norm": 2.187188569880515, + "language_loss": 0.641258, + "learning_rate": 1.615397974994126e-07, + "loss": 0.66224438, + "num_input_tokens_seen": 157435490, + "step": 7277, + "time_per_iteration": 2.7702529430389404 + }, + { + "auxiliary_loss_clip": 0.01162389, + "auxiliary_loss_mlp": 0.01024928, + "balance_loss_clip": 1.04792452, + "balance_loss_mlp": 1.0184902, + "epoch": 0.875127758071304, + "flos": 22710734386560.0, + "grad_norm": 1.5893903785809753, + "language_loss": 0.80947483, + "learning_rate": 1.6123323966037438e-07, + "loss": 0.83134794, + "num_input_tokens_seen": 157454010, + "step": 7278, + "time_per_iteration": 2.612703323364258 + }, + { + "auxiliary_loss_clip": 0.0116389, + "auxiliary_loss_mlp": 0.01023642, + "balance_loss_clip": 1.0494225, + "balance_loss_mlp": 1.01654649, + "epoch": 0.8752480009619431, + "flos": 23403846199680.0, + "grad_norm": 2.0719556912814006, + "language_loss": 0.78604984, + "learning_rate": 1.6092696076079216e-07, + "loss": 0.80792522, + "num_input_tokens_seen": 157472385, + "step": 7279, + "time_per_iteration": 2.5807721614837646 + }, + { + "auxiliary_loss_clip": 0.01084012, + "auxiliary_loss_mlp": 0.01026936, + "balance_loss_clip": 0.93234777, + "balance_loss_mlp": 1.01917219, + "epoch": 0.8753682438525822, + "flos": 26213101914240.0, + "grad_norm": 1.86678552887546, + "language_loss": 0.74067819, + "learning_rate": 1.6062096084712785e-07, + "loss": 0.76178765, + "num_input_tokens_seen": 157493735, + "step": 7280, + "time_per_iteration": 2.7376716136932373 + }, + { + "auxiliary_loss_clip": 0.01107717, + "auxiliary_loss_mlp": 0.00509483, + "balance_loss_clip": 0.96734768, + "balance_loss_mlp": 0.0, + "epoch": 0.8754884867432213, + "flos": 23326745656320.0, + "grad_norm": 1.7158225284048747, + "language_loss": 0.70691693, + "learning_rate": 1.6031523996580098e-07, + "loss": 0.72308898, + "num_input_tokens_seen": 157511295, + "step": 7281, + "time_per_iteration": 2.6409339904785156 + }, + { + "auxiliary_loss_clip": 0.01099917, + "auxiliary_loss_mlp": 0.01025672, + "balance_loss_clip": 0.93188322, + "balance_loss_mlp": 1.01845431, + "epoch": 0.8756087296338604, + "flos": 12495226412160.0, + "grad_norm": 2.019267417689364, + "language_loss": 0.66434938, + "learning_rate": 1.6000979816318981e-07, + "loss": 0.68560529, + "num_input_tokens_seen": 157529760, + "step": 7282, + "time_per_iteration": 2.7043886184692383 + }, + { + "auxiliary_loss_clip": 0.01138612, + "auxiliary_loss_mlp": 0.01020946, + "balance_loss_clip": 1.01043081, + "balance_loss_mlp": 1.01431513, + "epoch": 0.8757289725244994, + "flos": 18952898353920.0, + "grad_norm": 2.0173610496510808, + "language_loss": 0.7531026, + "learning_rate": 1.5970463548562886e-07, + "loss": 0.77469814, + "num_input_tokens_seen": 157548915, + "step": 7283, + "time_per_iteration": 2.6509203910827637 + }, + { + "auxiliary_loss_clip": 0.01113418, + "auxiliary_loss_mlp": 0.01024721, + "balance_loss_clip": 0.97048473, + "balance_loss_mlp": 1.01787233, + "epoch": 0.8758492154151386, + "flos": 25265958140160.0, + "grad_norm": 1.7350676795037165, + "language_loss": 0.71085668, + "learning_rate": 1.5939975197941192e-07, + "loss": 0.73223805, + "num_input_tokens_seen": 157570570, + "step": 7284, + "time_per_iteration": 2.708195209503174 + }, + { + "auxiliary_loss_clip": 0.01023982, + "auxiliary_loss_mlp": 0.00999828, + "balance_loss_clip": 0.93667066, + "balance_loss_mlp": 0.99887991, + "epoch": 0.8759694583057777, + "flos": 65571664193280.0, + "grad_norm": 0.8058829779363023, + "language_loss": 0.53307021, + "learning_rate": 1.5909514769078892e-07, + "loss": 0.55330837, + "num_input_tokens_seen": 157635675, + "step": 7285, + "time_per_iteration": 3.246025800704956 + }, + { + "auxiliary_loss_clip": 0.01088357, + "auxiliary_loss_mlp": 0.01027671, + "balance_loss_clip": 0.93723392, + "balance_loss_mlp": 1.02086163, + "epoch": 0.8760897011964167, + "flos": 25446193608960.0, + "grad_norm": 2.5045429877458303, + "language_loss": 0.77836883, + "learning_rate": 1.5879082266596867e-07, + "loss": 0.79952908, + "num_input_tokens_seen": 157657015, + "step": 7286, + "time_per_iteration": 2.706719398498535 + }, + { + "auxiliary_loss_clip": 0.01109838, + "auxiliary_loss_mlp": 0.01031076, + "balance_loss_clip": 0.96641004, + "balance_loss_mlp": 1.02399516, + "epoch": 0.8762099440870559, + "flos": 28984830894720.0, + "grad_norm": 3.4996276406086633, + "language_loss": 0.71930975, + "learning_rate": 1.5848677695111645e-07, + "loss": 0.7407189, + "num_input_tokens_seen": 157678615, + "step": 7287, + "time_per_iteration": 2.839334011077881 + }, + { + "auxiliary_loss_clip": 0.01103525, + "auxiliary_loss_mlp": 0.01025737, + "balance_loss_clip": 0.93614763, + "balance_loss_mlp": 1.01822698, + "epoch": 0.8763301869776949, + "flos": 21609461352960.0, + "grad_norm": 3.4399259403403835, + "language_loss": 0.70202661, + "learning_rate": 1.5818301059235562e-07, + "loss": 0.72331917, + "num_input_tokens_seen": 157693790, + "step": 7288, + "time_per_iteration": 2.7482426166534424 + }, + { + "auxiliary_loss_clip": 0.01116953, + "auxiliary_loss_mlp": 0.0102737, + "balance_loss_clip": 0.97103328, + "balance_loss_mlp": 1.01994348, + "epoch": 0.876450429868334, + "flos": 24644416176000.0, + "grad_norm": 1.7810192144304984, + "language_loss": 0.81621504, + "learning_rate": 1.578795236357684e-07, + "loss": 0.83765829, + "num_input_tokens_seen": 157715255, + "step": 7289, + "time_per_iteration": 2.740968704223633 + }, + { + "auxiliary_loss_clip": 0.01119356, + "auxiliary_loss_mlp": 0.01025522, + "balance_loss_clip": 0.97302783, + "balance_loss_mlp": 1.01888824, + "epoch": 0.8765706727589732, + "flos": 20260046188800.0, + "grad_norm": 4.8708890532289395, + "language_loss": 0.85475421, + "learning_rate": 1.5757631612739218e-07, + "loss": 0.876203, + "num_input_tokens_seen": 157728800, + "step": 7290, + "time_per_iteration": 2.610489845275879 + }, + { + "auxiliary_loss_clip": 0.01059411, + "auxiliary_loss_mlp": 0.0100294, + "balance_loss_clip": 1.0088644, + "balance_loss_mlp": 1.0019865, + "epoch": 0.8766909156496122, + "flos": 71371165276800.0, + "grad_norm": 0.7975163730734447, + "language_loss": 0.6145252, + "learning_rate": 1.572733881132242e-07, + "loss": 0.63514864, + "num_input_tokens_seen": 157789445, + "step": 7291, + "time_per_iteration": 4.090655326843262 + }, + { + "auxiliary_loss_clip": 0.00994778, + "auxiliary_loss_mlp": 0.00999368, + "balance_loss_clip": 0.89900541, + "balance_loss_mlp": 0.99818802, + "epoch": 0.8768111585402513, + "flos": 69523490603520.0, + "grad_norm": 0.8537849847980834, + "language_loss": 0.58516544, + "learning_rate": 1.5697073963921814e-07, + "loss": 0.60510689, + "num_input_tokens_seen": 157848685, + "step": 7292, + "time_per_iteration": 4.080475568771362 + }, + { + "auxiliary_loss_clip": 0.01140529, + "auxiliary_loss_mlp": 0.01023273, + "balance_loss_clip": 1.01054978, + "balance_loss_mlp": 1.01613832, + "epoch": 0.8769314014308904, + "flos": 18838558385280.0, + "grad_norm": 2.5623863347218885, + "language_loss": 0.85295188, + "learning_rate": 1.566683707512857e-07, + "loss": 0.87458992, + "num_input_tokens_seen": 157866360, + "step": 7293, + "time_per_iteration": 2.557582139968872 + }, + { + "auxiliary_loss_clip": 0.01112584, + "auxiliary_loss_mlp": 0.01027822, + "balance_loss_clip": 0.96878785, + "balance_loss_mlp": 1.02055311, + "epoch": 0.8770516443215295, + "flos": 14976402278400.0, + "grad_norm": 2.0248995290418352, + "language_loss": 0.79382241, + "learning_rate": 1.5636628149529553e-07, + "loss": 0.81522644, + "num_input_tokens_seen": 157884150, + "step": 7294, + "time_per_iteration": 3.5708794593811035 + }, + { + "auxiliary_loss_clip": 0.01115953, + "auxiliary_loss_mlp": 0.01024845, + "balance_loss_clip": 0.97013223, + "balance_loss_mlp": 1.01806235, + "epoch": 0.8771718872121685, + "flos": 31649654021760.0, + "grad_norm": 2.022665377607136, + "language_loss": 0.79842484, + "learning_rate": 1.560644719170743e-07, + "loss": 0.8198328, + "num_input_tokens_seen": 157905020, + "step": 7295, + "time_per_iteration": 2.7348787784576416 + }, + { + "auxiliary_loss_clip": 0.01092264, + "auxiliary_loss_mlp": 0.01025406, + "balance_loss_clip": 0.93036085, + "balance_loss_mlp": 1.01841736, + "epoch": 0.8772921301028077, + "flos": 36095466222720.0, + "grad_norm": 1.7486218715614663, + "language_loss": 0.721457, + "learning_rate": 1.5576294206240692e-07, + "loss": 0.7426337, + "num_input_tokens_seen": 157924545, + "step": 7296, + "time_per_iteration": 2.793015956878662 + }, + { + "auxiliary_loss_clip": 0.01115143, + "auxiliary_loss_mlp": 0.01027489, + "balance_loss_clip": 0.97278166, + "balance_loss_mlp": 1.02010691, + "epoch": 0.8774123729934468, + "flos": 57116961849600.0, + "grad_norm": 1.848073685632059, + "language_loss": 0.67757457, + "learning_rate": 1.5546169197703507e-07, + "loss": 0.69900084, + "num_input_tokens_seen": 157950820, + "step": 7297, + "time_per_iteration": 2.962134838104248 + }, + { + "auxiliary_loss_clip": 0.01121082, + "auxiliary_loss_mlp": 0.01025665, + "balance_loss_clip": 0.96944559, + "balance_loss_mlp": 1.01850104, + "epoch": 0.8775326158840858, + "flos": 23914495900800.0, + "grad_norm": 2.5436372404377368, + "language_loss": 0.7767961, + "learning_rate": 1.5516072170665774e-07, + "loss": 0.79826361, + "num_input_tokens_seen": 157968790, + "step": 7298, + "time_per_iteration": 2.7582576274871826 + }, + { + "auxiliary_loss_clip": 0.01142535, + "auxiliary_loss_mlp": 0.01023451, + "balance_loss_clip": 1.01130259, + "balance_loss_mlp": 1.01640558, + "epoch": 0.877652858774725, + "flos": 17123285243520.0, + "grad_norm": 2.002940271820267, + "language_loss": 0.8703261, + "learning_rate": 1.5486003129693214e-07, + "loss": 0.89198595, + "num_input_tokens_seen": 157986155, + "step": 7299, + "time_per_iteration": 2.6712701320648193 + }, + { + "auxiliary_loss_clip": 0.01143007, + "auxiliary_loss_mlp": 0.01021217, + "balance_loss_clip": 1.00953174, + "balance_loss_mlp": 1.0141387, + "epoch": 0.877773101665364, + "flos": 16508961912960.0, + "grad_norm": 1.6982976148854116, + "language_loss": 0.78296995, + "learning_rate": 1.545596207934725e-07, + "loss": 0.80461228, + "num_input_tokens_seen": 158004640, + "step": 7300, + "time_per_iteration": 2.7122819423675537 + }, + { + "auxiliary_loss_clip": 0.01111713, + "auxiliary_loss_mlp": 0.01024164, + "balance_loss_clip": 0.96953619, + "balance_loss_mlp": 1.01742625, + "epoch": 0.8778933445560031, + "flos": 22053209973120.0, + "grad_norm": 1.7013663881351655, + "language_loss": 0.7789892, + "learning_rate": 1.5425949024185147e-07, + "loss": 0.80034804, + "num_input_tokens_seen": 158024665, + "step": 7301, + "time_per_iteration": 3.5558958053588867 + }, + { + "auxiliary_loss_clip": 0.011181, + "auxiliary_loss_mlp": 0.01023218, + "balance_loss_clip": 0.96892971, + "balance_loss_mlp": 1.01631904, + "epoch": 0.8780135874466423, + "flos": 22564757514240.0, + "grad_norm": 1.9569071687372568, + "language_loss": 0.67723322, + "learning_rate": 1.5395963968759818e-07, + "loss": 0.69864643, + "num_input_tokens_seen": 158044940, + "step": 7302, + "time_per_iteration": 2.7572476863861084 + }, + { + "auxiliary_loss_clip": 0.01117803, + "auxiliary_loss_mlp": 0.01021693, + "balance_loss_clip": 0.96891999, + "balance_loss_mlp": 1.01472867, + "epoch": 0.8781338303372813, + "flos": 61531999073280.0, + "grad_norm": 1.7900773377970345, + "language_loss": 0.6455251, + "learning_rate": 1.536600691761998e-07, + "loss": 0.66692007, + "num_input_tokens_seen": 158070770, + "step": 7303, + "time_per_iteration": 3.1164824962615967 + }, + { + "auxiliary_loss_clip": 0.01096819, + "auxiliary_loss_mlp": 0.01020189, + "balance_loss_clip": 0.93446112, + "balance_loss_mlp": 1.01314402, + "epoch": 0.8782540732279204, + "flos": 22674751937280.0, + "grad_norm": 1.84536802665859, + "language_loss": 0.71559072, + "learning_rate": 1.5336077875310084e-07, + "loss": 0.7367608, + "num_input_tokens_seen": 158089995, + "step": 7304, + "time_per_iteration": 2.861142158508301 + }, + { + "auxiliary_loss_clip": 0.01076003, + "auxiliary_loss_mlp": 0.01018794, + "balance_loss_clip": 0.89432323, + "balance_loss_mlp": 1.01218939, + "epoch": 0.8783743161185595, + "flos": 16070348937600.0, + "grad_norm": 2.9212913215550773, + "language_loss": 0.74379742, + "learning_rate": 1.5306176846370321e-07, + "loss": 0.76474535, + "num_input_tokens_seen": 158108140, + "step": 7305, + "time_per_iteration": 2.891357898712158 + }, + { + "auxiliary_loss_clip": 0.01125661, + "auxiliary_loss_mlp": 0.01024964, + "balance_loss_clip": 0.97128987, + "balance_loss_mlp": 1.01763868, + "epoch": 0.8784945590091986, + "flos": 26067879227520.0, + "grad_norm": 2.130921390662998, + "language_loss": 0.74144042, + "learning_rate": 1.5276303835336712e-07, + "loss": 0.76294667, + "num_input_tokens_seen": 158128680, + "step": 7306, + "time_per_iteration": 2.751912832260132 + }, + { + "auxiliary_loss_clip": 0.01041581, + "auxiliary_loss_mlp": 0.01000373, + "balance_loss_clip": 0.97225022, + "balance_loss_mlp": 0.99940127, + "epoch": 0.8786148018998376, + "flos": 62720643939840.0, + "grad_norm": 0.7819667569623006, + "language_loss": 0.53501225, + "learning_rate": 1.524645884674094e-07, + "loss": 0.55543184, + "num_input_tokens_seen": 158185610, + "step": 7307, + "time_per_iteration": 3.246455430984497 + }, + { + "auxiliary_loss_clip": 0.01166654, + "auxiliary_loss_mlp": 0.00509711, + "balance_loss_clip": 1.04810834, + "balance_loss_mlp": 0.0, + "epoch": 0.8787350447904768, + "flos": 21652734263040.0, + "grad_norm": 2.0822942200172494, + "language_loss": 0.79334396, + "learning_rate": 1.521664188511047e-07, + "loss": 0.81010765, + "num_input_tokens_seen": 158205635, + "step": 7308, + "time_per_iteration": 2.6901886463165283 + }, + { + "auxiliary_loss_clip": 0.01117978, + "auxiliary_loss_mlp": 0.00509094, + "balance_loss_clip": 0.97582376, + "balance_loss_mlp": 0.0, + "epoch": 0.8788552876811159, + "flos": 25478476957440.0, + "grad_norm": 2.0316161593541944, + "language_loss": 0.80450255, + "learning_rate": 1.518685295496851e-07, + "loss": 0.8207733, + "num_input_tokens_seen": 158223495, + "step": 7309, + "time_per_iteration": 2.7821261882781982 + }, + { + "auxiliary_loss_clip": 0.01140236, + "auxiliary_loss_mlp": 0.01021224, + "balance_loss_clip": 1.00780797, + "balance_loss_mlp": 1.01472461, + "epoch": 0.8789755305717549, + "flos": 22310222762880.0, + "grad_norm": 1.7491490892517993, + "language_loss": 0.85428679, + "learning_rate": 1.5157092060833975e-07, + "loss": 0.87590134, + "num_input_tokens_seen": 158243145, + "step": 7310, + "time_per_iteration": 2.646714925765991 + }, + { + "auxiliary_loss_clip": 0.01113095, + "auxiliary_loss_mlp": 0.01023285, + "balance_loss_clip": 0.96748585, + "balance_loss_mlp": 1.01590276, + "epoch": 0.879095773462394, + "flos": 29310971408640.0, + "grad_norm": 1.680804368162984, + "language_loss": 0.66042781, + "learning_rate": 1.5127359207221658e-07, + "loss": 0.6817916, + "num_input_tokens_seen": 158262625, + "step": 7311, + "time_per_iteration": 2.7067995071411133 + }, + { + "auxiliary_loss_clip": 0.01035692, + "auxiliary_loss_mlp": 0.01025685, + "balance_loss_clip": 0.85038018, + "balance_loss_mlp": 1.01751971, + "epoch": 0.8792160163530331, + "flos": 16690023394560.0, + "grad_norm": 2.0312778854979747, + "language_loss": 0.73191607, + "learning_rate": 1.5097654398641923e-07, + "loss": 0.75252986, + "num_input_tokens_seen": 158280530, + "step": 7312, + "time_per_iteration": 2.765984058380127 + }, + { + "auxiliary_loss_clip": 0.01148753, + "auxiliary_loss_mlp": 0.01024308, + "balance_loss_clip": 1.01352906, + "balance_loss_mlp": 1.0170486, + "epoch": 0.8793362592436722, + "flos": 24499301230080.0, + "grad_norm": 1.385173203239773, + "language_loss": 0.73040462, + "learning_rate": 1.5067977639601014e-07, + "loss": 0.75213522, + "num_input_tokens_seen": 158303290, + "step": 7313, + "time_per_iteration": 2.7302322387695312 + }, + { + "auxiliary_loss_clip": 0.01116059, + "auxiliary_loss_mlp": 0.01022148, + "balance_loss_clip": 0.97185171, + "balance_loss_mlp": 1.01540124, + "epoch": 0.8794565021343113, + "flos": 14538399834240.0, + "grad_norm": 2.644617173270566, + "language_loss": 0.71102434, + "learning_rate": 1.5038328934600864e-07, + "loss": 0.73240644, + "num_input_tokens_seen": 158319925, + "step": 7314, + "time_per_iteration": 2.642960786819458 + }, + { + "auxiliary_loss_clip": 0.01116242, + "auxiliary_loss_mlp": 0.01027193, + "balance_loss_clip": 0.97098744, + "balance_loss_mlp": 1.02021015, + "epoch": 0.8795767450249504, + "flos": 39530286224640.0, + "grad_norm": 1.8515434536773656, + "language_loss": 0.69647372, + "learning_rate": 1.5008708288139161e-07, + "loss": 0.71790808, + "num_input_tokens_seen": 158342285, + "step": 7315, + "time_per_iteration": 2.7991888523101807 + }, + { + "auxiliary_loss_clip": 0.01141004, + "auxiliary_loss_mlp": 0.01024428, + "balance_loss_clip": 1.01239848, + "balance_loss_mlp": 1.01637506, + "epoch": 0.8796969879155895, + "flos": 22960672197120.0, + "grad_norm": 1.8683752900825157, + "language_loss": 0.73024309, + "learning_rate": 1.497911570470931e-07, + "loss": 0.75189739, + "num_input_tokens_seen": 158362290, + "step": 7316, + "time_per_iteration": 2.684035062789917 + }, + { + "auxiliary_loss_clip": 0.01085477, + "auxiliary_loss_mlp": 0.01028701, + "balance_loss_clip": 0.93168426, + "balance_loss_mlp": 1.02138448, + "epoch": 0.8798172308062285, + "flos": 28362427004160.0, + "grad_norm": 1.6168101811724636, + "language_loss": 0.86019599, + "learning_rate": 1.494955118880048e-07, + "loss": 0.88133776, + "num_input_tokens_seen": 158383275, + "step": 7317, + "time_per_iteration": 2.7434253692626953 + }, + { + "auxiliary_loss_clip": 0.01141277, + "auxiliary_loss_mlp": 0.01025349, + "balance_loss_clip": 1.00944769, + "balance_loss_mlp": 1.01870871, + "epoch": 0.8799374736968677, + "flos": 23988974751360.0, + "grad_norm": 1.723952716289604, + "language_loss": 0.73027205, + "learning_rate": 1.4920014744897634e-07, + "loss": 0.75193834, + "num_input_tokens_seen": 158402690, + "step": 7318, + "time_per_iteration": 3.5725245475769043 + }, + { + "auxiliary_loss_clip": 0.01108736, + "auxiliary_loss_mlp": 0.01023711, + "balance_loss_clip": 0.96979213, + "balance_loss_mlp": 1.01643896, + "epoch": 0.8800577165875068, + "flos": 25630271832960.0, + "grad_norm": 2.015093001861931, + "language_loss": 0.86369026, + "learning_rate": 1.4890506377481392e-07, + "loss": 0.88501477, + "num_input_tokens_seen": 158421780, + "step": 7319, + "time_per_iteration": 3.5156993865966797 + }, + { + "auxiliary_loss_clip": 0.01046344, + "auxiliary_loss_mlp": 0.01027926, + "balance_loss_clip": 0.85846418, + "balance_loss_mlp": 1.02134824, + "epoch": 0.8801779594781458, + "flos": 23440331439360.0, + "grad_norm": 1.5329953622992272, + "language_loss": 0.63896132, + "learning_rate": 1.486102609102815e-07, + "loss": 0.65970397, + "num_input_tokens_seen": 158442330, + "step": 7320, + "time_per_iteration": 3.451094627380371 + }, + { + "auxiliary_loss_clip": 0.01113002, + "auxiliary_loss_mlp": 0.01025337, + "balance_loss_clip": 0.97082871, + "balance_loss_mlp": 1.01824951, + "epoch": 0.880298202368785, + "flos": 11508580656000.0, + "grad_norm": 2.803579510092893, + "language_loss": 0.858127, + "learning_rate": 1.483157389001004e-07, + "loss": 0.8795104, + "num_input_tokens_seen": 158459890, + "step": 7321, + "time_per_iteration": 2.6500260829925537 + }, + { + "auxiliary_loss_clip": 0.01114465, + "auxiliary_loss_mlp": 0.01021426, + "balance_loss_clip": 0.96758831, + "balance_loss_mlp": 1.01356697, + "epoch": 0.880418445259424, + "flos": 22671447886080.0, + "grad_norm": 2.156763289925786, + "language_loss": 0.78721762, + "learning_rate": 1.4802149778894933e-07, + "loss": 0.80857652, + "num_input_tokens_seen": 158478680, + "step": 7322, + "time_per_iteration": 2.6541478633880615 + }, + { + "auxiliary_loss_clip": 0.01130091, + "auxiliary_loss_mlp": 0.0102051, + "balance_loss_clip": 1.00378668, + "balance_loss_mlp": 1.01403749, + "epoch": 0.8805386881500631, + "flos": 20522158709760.0, + "grad_norm": 2.3137371203201713, + "language_loss": 0.87387908, + "learning_rate": 1.4772753762146484e-07, + "loss": 0.89538509, + "num_input_tokens_seen": 158497935, + "step": 7323, + "time_per_iteration": 2.6361117362976074 + }, + { + "auxiliary_loss_clip": 0.01131433, + "auxiliary_loss_mlp": 0.01027487, + "balance_loss_clip": 1.00444627, + "balance_loss_mlp": 1.01998615, + "epoch": 0.8806589310407023, + "flos": 36538891620480.0, + "grad_norm": 1.5322552598421517, + "language_loss": 0.70864463, + "learning_rate": 1.474338584422401e-07, + "loss": 0.73023391, + "num_input_tokens_seen": 158523145, + "step": 7324, + "time_per_iteration": 2.753671884536743 + }, + { + "auxiliary_loss_clip": 0.01135478, + "auxiliary_loss_mlp": 0.01024465, + "balance_loss_clip": 1.0089004, + "balance_loss_mlp": 1.01764023, + "epoch": 0.8807791739313413, + "flos": 23440187784960.0, + "grad_norm": 1.6468914976716935, + "language_loss": 0.76172096, + "learning_rate": 1.4714046029582595e-07, + "loss": 0.78332037, + "num_input_tokens_seen": 158542210, + "step": 7325, + "time_per_iteration": 2.6749308109283447 + }, + { + "auxiliary_loss_clip": 0.01094211, + "auxiliary_loss_mlp": 0.01023334, + "balance_loss_clip": 0.93102014, + "balance_loss_mlp": 1.01672435, + "epoch": 0.8808994168219804, + "flos": 25956843310080.0, + "grad_norm": 1.7687150644365035, + "language_loss": 0.75698817, + "learning_rate": 1.46847343226731e-07, + "loss": 0.77816367, + "num_input_tokens_seen": 158563250, + "step": 7326, + "time_per_iteration": 3.678865432739258 + }, + { + "auxiliary_loss_clip": 0.01144582, + "auxiliary_loss_mlp": 0.01025675, + "balance_loss_clip": 1.00978887, + "balance_loss_mlp": 1.01855254, + "epoch": 0.8810196597126195, + "flos": 17092079303040.0, + "grad_norm": 1.7813710885977374, + "language_loss": 0.69342983, + "learning_rate": 1.465545072794203e-07, + "loss": 0.71513236, + "num_input_tokens_seen": 158581125, + "step": 7327, + "time_per_iteration": 2.673583507537842 + }, + { + "auxiliary_loss_clip": 0.01068168, + "auxiliary_loss_mlp": 0.01022968, + "balance_loss_clip": 0.89654589, + "balance_loss_mlp": 1.01600277, + "epoch": 0.8811399026032586, + "flos": 23002831785600.0, + "grad_norm": 3.422405978436276, + "language_loss": 0.75652277, + "learning_rate": 1.4626195249831774e-07, + "loss": 0.77743411, + "num_input_tokens_seen": 158602025, + "step": 7328, + "time_per_iteration": 2.7945427894592285 + }, + { + "auxiliary_loss_clip": 0.01138508, + "auxiliary_loss_mlp": 0.01021532, + "balance_loss_clip": 1.0090543, + "balance_loss_mlp": 1.0151515, + "epoch": 0.8812601454938976, + "flos": 14463813242880.0, + "grad_norm": 2.3255648577979193, + "language_loss": 0.71740043, + "learning_rate": 1.4596967892780244e-07, + "loss": 0.7390008, + "num_input_tokens_seen": 158618355, + "step": 7329, + "time_per_iteration": 2.6479408740997314 + }, + { + "auxiliary_loss_clip": 0.0116316, + "auxiliary_loss_mlp": 0.01022526, + "balance_loss_clip": 1.04866719, + "balance_loss_mlp": 1.01578188, + "epoch": 0.8813803883845368, + "flos": 22493223578880.0, + "grad_norm": 1.6315538144822699, + "language_loss": 0.74863029, + "learning_rate": 1.4567768661221314e-07, + "loss": 0.77048719, + "num_input_tokens_seen": 158638925, + "step": 7330, + "time_per_iteration": 2.6260597705841064 + }, + { + "auxiliary_loss_clip": 0.01144638, + "auxiliary_loss_mlp": 0.0050994, + "balance_loss_clip": 1.01172662, + "balance_loss_mlp": 0.0, + "epoch": 0.8815006312751759, + "flos": 21506901045120.0, + "grad_norm": 1.8369085809312733, + "language_loss": 0.74497002, + "learning_rate": 1.4538597559584442e-07, + "loss": 0.76151574, + "num_input_tokens_seen": 158656715, + "step": 7331, + "time_per_iteration": 2.7392544746398926 + }, + { + "auxiliary_loss_clip": 0.01111708, + "auxiliary_loss_mlp": 0.0102333, + "balance_loss_clip": 0.97011113, + "balance_loss_mlp": 1.01553035, + "epoch": 0.8816208741658149, + "flos": 22784566792320.0, + "grad_norm": 1.965803659015005, + "language_loss": 0.78603715, + "learning_rate": 1.4509454592294823e-07, + "loss": 0.80738759, + "num_input_tokens_seen": 158677200, + "step": 7332, + "time_per_iteration": 2.6598103046417236 + }, + { + "auxiliary_loss_clip": 0.01097082, + "auxiliary_loss_mlp": 0.00509079, + "balance_loss_clip": 0.93602574, + "balance_loss_mlp": 0.0, + "epoch": 0.8817411170564541, + "flos": 17779409026560.0, + "grad_norm": 2.3678752931193126, + "language_loss": 0.79060155, + "learning_rate": 1.448033976377354e-07, + "loss": 0.80666316, + "num_input_tokens_seen": 158692185, + "step": 7333, + "time_per_iteration": 2.71458101272583 + }, + { + "auxiliary_loss_clip": 0.01142211, + "auxiliary_loss_mlp": 0.01032383, + "balance_loss_clip": 1.00843692, + "balance_loss_mlp": 1.02533174, + "epoch": 0.8818613599470931, + "flos": 18551812112640.0, + "grad_norm": 2.5666521299474176, + "language_loss": 0.74178296, + "learning_rate": 1.445125307843713e-07, + "loss": 0.76352888, + "num_input_tokens_seen": 158710410, + "step": 7334, + "time_per_iteration": 2.618861675262451 + }, + { + "auxiliary_loss_clip": 0.011384, + "auxiliary_loss_mlp": 0.01026719, + "balance_loss_clip": 1.01042295, + "balance_loss_mlp": 1.02027011, + "epoch": 0.8819816028377322, + "flos": 27599792417280.0, + "grad_norm": 2.186779408035889, + "language_loss": 0.75753146, + "learning_rate": 1.442219454069813e-07, + "loss": 0.77918267, + "num_input_tokens_seen": 158731435, + "step": 7335, + "time_per_iteration": 2.703911066055298 + }, + { + "auxiliary_loss_clip": 0.01072365, + "auxiliary_loss_mlp": 0.01019994, + "balance_loss_clip": 0.89375114, + "balance_loss_mlp": 1.01349139, + "epoch": 0.8821018457283714, + "flos": 23404600385280.0, + "grad_norm": 1.9467374821530747, + "language_loss": 0.66623807, + "learning_rate": 1.4393164154964676e-07, + "loss": 0.68716168, + "num_input_tokens_seen": 158750965, + "step": 7336, + "time_per_iteration": 2.7392730712890625 + }, + { + "auxiliary_loss_clip": 0.01140859, + "auxiliary_loss_mlp": 0.01030001, + "balance_loss_clip": 1.01235914, + "balance_loss_mlp": 1.02313733, + "epoch": 0.8822220886190104, + "flos": 29132459792640.0, + "grad_norm": 1.7292749699165626, + "language_loss": 0.93878913, + "learning_rate": 1.4364161925640649e-07, + "loss": 0.96049774, + "num_input_tokens_seen": 158772365, + "step": 7337, + "time_per_iteration": 2.690284013748169 + }, + { + "auxiliary_loss_clip": 0.0116257, + "auxiliary_loss_mlp": 0.01027521, + "balance_loss_clip": 1.04725635, + "balance_loss_mlp": 1.02057147, + "epoch": 0.8823423315096495, + "flos": 20485422074880.0, + "grad_norm": 1.753865327581322, + "language_loss": 0.84844291, + "learning_rate": 1.4335187857125663e-07, + "loss": 0.87034392, + "num_input_tokens_seen": 158791065, + "step": 7338, + "time_per_iteration": 2.598970651626587 + }, + { + "auxiliary_loss_clip": 0.0114153, + "auxiliary_loss_mlp": 0.01023512, + "balance_loss_clip": 1.00881076, + "balance_loss_mlp": 1.01690197, + "epoch": 0.8824625744002886, + "flos": 24206377818240.0, + "grad_norm": 2.2682023714745596, + "language_loss": 0.75147486, + "learning_rate": 1.4306241953815023e-07, + "loss": 0.77312529, + "num_input_tokens_seen": 158812125, + "step": 7339, + "time_per_iteration": 2.6913037300109863 + }, + { + "auxiliary_loss_clip": 0.01139818, + "auxiliary_loss_mlp": 0.01024956, + "balance_loss_clip": 1.00797725, + "balance_loss_mlp": 1.01799679, + "epoch": 0.8825828172909277, + "flos": 24679500785280.0, + "grad_norm": 1.6351651258204583, + "language_loss": 0.71239048, + "learning_rate": 1.4277324220099862e-07, + "loss": 0.73403823, + "num_input_tokens_seen": 158834035, + "step": 7340, + "time_per_iteration": 2.642132520675659 + }, + { + "auxiliary_loss_clip": 0.01088853, + "auxiliary_loss_mlp": 0.01021622, + "balance_loss_clip": 0.9298383, + "balance_loss_mlp": 1.01514268, + "epoch": 0.8827030601815667, + "flos": 22456163721600.0, + "grad_norm": 2.5182012858328715, + "language_loss": 0.74137354, + "learning_rate": 1.4248434660366938e-07, + "loss": 0.76247823, + "num_input_tokens_seen": 158853510, + "step": 7341, + "time_per_iteration": 2.715754985809326 + }, + { + "auxiliary_loss_clip": 0.01115101, + "auxiliary_loss_mlp": 0.01028762, + "balance_loss_clip": 0.97123003, + "balance_loss_mlp": 1.02140129, + "epoch": 0.8828233030722058, + "flos": 19865639877120.0, + "grad_norm": 1.6802891849275117, + "language_loss": 0.70698601, + "learning_rate": 1.4219573278998808e-07, + "loss": 0.72842467, + "num_input_tokens_seen": 158871970, + "step": 7342, + "time_per_iteration": 2.707557439804077 + }, + { + "auxiliary_loss_clip": 0.01115674, + "auxiliary_loss_mlp": 0.01026696, + "balance_loss_clip": 0.96918613, + "balance_loss_mlp": 1.01954293, + "epoch": 0.882943545962845, + "flos": 39347213581440.0, + "grad_norm": 1.9466558722258291, + "language_loss": 0.64920723, + "learning_rate": 1.4190740080373685e-07, + "loss": 0.67063093, + "num_input_tokens_seen": 158892250, + "step": 7343, + "time_per_iteration": 3.7341697216033936 + }, + { + "auxiliary_loss_clip": 0.01070519, + "auxiliary_loss_mlp": 0.01023267, + "balance_loss_clip": 0.89474595, + "balance_loss_mlp": 1.01573861, + "epoch": 0.883063788853484, + "flos": 19054524908160.0, + "grad_norm": 1.8534705400989062, + "language_loss": 0.83862591, + "learning_rate": 1.4161935068865538e-07, + "loss": 0.85956383, + "num_input_tokens_seen": 158907395, + "step": 7344, + "time_per_iteration": 3.5969557762145996 + }, + { + "auxiliary_loss_clip": 0.01164746, + "auxiliary_loss_mlp": 0.01026256, + "balance_loss_clip": 1.04838622, + "balance_loss_mlp": 1.01897788, + "epoch": 0.8831840317441231, + "flos": 18733196816640.0, + "grad_norm": 2.47574558468724, + "language_loss": 0.75702953, + "learning_rate": 1.4133158248844113e-07, + "loss": 0.77893955, + "num_input_tokens_seen": 158926300, + "step": 7345, + "time_per_iteration": 2.6597797870635986 + }, + { + "auxiliary_loss_clip": 0.0109549, + "auxiliary_loss_mlp": 0.01021338, + "balance_loss_clip": 0.931988, + "balance_loss_mlp": 1.01394153, + "epoch": 0.8833042746347622, + "flos": 26827712553600.0, + "grad_norm": 1.736526522289649, + "language_loss": 0.73019934, + "learning_rate": 1.4104409624674785e-07, + "loss": 0.75136757, + "num_input_tokens_seen": 158946085, + "step": 7346, + "time_per_iteration": 3.6412439346313477 + }, + { + "auxiliary_loss_clip": 0.01142914, + "auxiliary_loss_mlp": 0.01025433, + "balance_loss_clip": 1.01216817, + "balance_loss_mlp": 1.01865017, + "epoch": 0.8834245175254013, + "flos": 26104077158400.0, + "grad_norm": 1.7942836860950595, + "language_loss": 0.78526032, + "learning_rate": 1.407568920071873e-07, + "loss": 0.80694377, + "num_input_tokens_seen": 158964950, + "step": 7347, + "time_per_iteration": 2.710300922393799 + }, + { + "auxiliary_loss_clip": 0.0116885, + "auxiliary_loss_mlp": 0.01027026, + "balance_loss_clip": 1.0493387, + "balance_loss_mlp": 1.01955462, + "epoch": 0.8835447604160404, + "flos": 30629036977920.0, + "grad_norm": 2.0240524481631677, + "language_loss": 0.68107504, + "learning_rate": 1.4046996981332782e-07, + "loss": 0.7030338, + "num_input_tokens_seen": 158984835, + "step": 7348, + "time_per_iteration": 2.648326873779297 + }, + { + "auxiliary_loss_clip": 0.01096291, + "auxiliary_loss_mlp": 0.01024601, + "balance_loss_clip": 0.9321003, + "balance_loss_mlp": 1.01725769, + "epoch": 0.8836650033066795, + "flos": 24718356322560.0, + "grad_norm": 3.045813994880613, + "language_loss": 0.78432995, + "learning_rate": 1.4018332970869516e-07, + "loss": 0.80553889, + "num_input_tokens_seen": 159002775, + "step": 7349, + "time_per_iteration": 2.7245826721191406 + }, + { + "auxiliary_loss_clip": 0.01112015, + "auxiliary_loss_mlp": 0.01030385, + "balance_loss_clip": 0.97341186, + "balance_loss_mlp": 1.02301788, + "epoch": 0.8837852461973186, + "flos": 25413371556480.0, + "grad_norm": 1.9061454071711794, + "language_loss": 0.85257053, + "learning_rate": 1.398969717367733e-07, + "loss": 0.87399459, + "num_input_tokens_seen": 159024100, + "step": 7350, + "time_per_iteration": 2.69917631149292 + }, + { + "auxiliary_loss_clip": 0.01069115, + "auxiliary_loss_mlp": 0.01028285, + "balance_loss_clip": 0.89711791, + "balance_loss_mlp": 1.0211916, + "epoch": 0.8839054890879576, + "flos": 17822574195840.0, + "grad_norm": 3.9456405183791268, + "language_loss": 0.76325405, + "learning_rate": 1.396108959410014e-07, + "loss": 0.78422809, + "num_input_tokens_seen": 159043315, + "step": 7351, + "time_per_iteration": 2.7474403381347656 + }, + { + "auxiliary_loss_clip": 0.0113852, + "auxiliary_loss_mlp": 0.00509087, + "balance_loss_clip": 1.01066351, + "balance_loss_mlp": 0.0, + "epoch": 0.8840257319785968, + "flos": 23769021818880.0, + "grad_norm": 1.7495157783204491, + "language_loss": 0.81366241, + "learning_rate": 1.3932510236477745e-07, + "loss": 0.83013844, + "num_input_tokens_seen": 159063985, + "step": 7352, + "time_per_iteration": 2.6500964164733887 + }, + { + "auxiliary_loss_clip": 0.01136046, + "auxiliary_loss_mlp": 0.01023341, + "balance_loss_clip": 1.00591087, + "balance_loss_mlp": 1.01555657, + "epoch": 0.8841459748692359, + "flos": 29059776622080.0, + "grad_norm": 3.6085577849322625, + "language_loss": 0.56291103, + "learning_rate": 1.3903959105145636e-07, + "loss": 0.5845049, + "num_input_tokens_seen": 159084475, + "step": 7353, + "time_per_iteration": 3.656743049621582 + }, + { + "auxiliary_loss_clip": 0.01162578, + "auxiliary_loss_mlp": 0.01023575, + "balance_loss_clip": 1.04692125, + "balance_loss_mlp": 1.01644909, + "epoch": 0.8842662177598749, + "flos": 24311523905280.0, + "grad_norm": 7.243053857725652, + "language_loss": 0.83239841, + "learning_rate": 1.387543620443492e-07, + "loss": 0.85425997, + "num_input_tokens_seen": 159101320, + "step": 7354, + "time_per_iteration": 2.630816698074341 + }, + { + "auxiliary_loss_clip": 0.01162476, + "auxiliary_loss_mlp": 0.01025925, + "balance_loss_clip": 1.04723406, + "balance_loss_mlp": 1.01889145, + "epoch": 0.8843864606505141, + "flos": 25007867942400.0, + "grad_norm": 1.5912455649190562, + "language_loss": 0.84388769, + "learning_rate": 1.3846941538672606e-07, + "loss": 0.86577177, + "num_input_tokens_seen": 159120025, + "step": 7355, + "time_per_iteration": 2.6461191177368164 + }, + { + "auxiliary_loss_clip": 0.01076819, + "auxiliary_loss_mlp": 0.01023956, + "balance_loss_clip": 0.89756918, + "balance_loss_mlp": 1.01666892, + "epoch": 0.8845067035411531, + "flos": 28183915388160.0, + "grad_norm": 2.0032352626485004, + "language_loss": 0.807271, + "learning_rate": 1.3818475112181193e-07, + "loss": 0.82827878, + "num_input_tokens_seen": 159138820, + "step": 7356, + "time_per_iteration": 2.8049421310424805 + }, + { + "auxiliary_loss_clip": 0.01115082, + "auxiliary_loss_mlp": 0.01024335, + "balance_loss_clip": 0.97035307, + "balance_loss_mlp": 1.01681948, + "epoch": 0.8846269464317922, + "flos": 12853219311360.0, + "grad_norm": 2.0543866431364446, + "language_loss": 0.79715711, + "learning_rate": 1.3790036929279091e-07, + "loss": 0.8185513, + "num_input_tokens_seen": 159155975, + "step": 7357, + "time_per_iteration": 2.5926451683044434 + }, + { + "auxiliary_loss_clip": 0.01144134, + "auxiliary_loss_mlp": 0.00508958, + "balance_loss_clip": 1.01129651, + "balance_loss_mlp": 0.0, + "epoch": 0.8847471893224313, + "flos": 18624351628800.0, + "grad_norm": 2.132107654293465, + "language_loss": 0.59077233, + "learning_rate": 1.3761626994280363e-07, + "loss": 0.6073032, + "num_input_tokens_seen": 159173445, + "step": 7358, + "time_per_iteration": 2.6250455379486084 + }, + { + "auxiliary_loss_clip": 0.01098846, + "auxiliary_loss_mlp": 0.01023788, + "balance_loss_clip": 0.93275183, + "balance_loss_mlp": 1.0169661, + "epoch": 0.8848674322130704, + "flos": 35769433449600.0, + "grad_norm": 1.5745244977630697, + "language_loss": 0.73623306, + "learning_rate": 1.3733245311494735e-07, + "loss": 0.7574594, + "num_input_tokens_seen": 159196100, + "step": 7359, + "time_per_iteration": 2.8246490955352783 + }, + { + "auxiliary_loss_clip": 0.01141719, + "auxiliary_loss_mlp": 0.01023684, + "balance_loss_clip": 1.0115999, + "balance_loss_mlp": 1.01689839, + "epoch": 0.8849876751037095, + "flos": 24243760897920.0, + "grad_norm": 1.9636666902939879, + "language_loss": 0.70745635, + "learning_rate": 1.3704891885227676e-07, + "loss": 0.72911036, + "num_input_tokens_seen": 159216145, + "step": 7360, + "time_per_iteration": 2.6380348205566406 + }, + { + "auxiliary_loss_clip": 0.01093068, + "auxiliary_loss_mlp": 0.01024793, + "balance_loss_clip": 0.92934471, + "balance_loss_mlp": 1.01732469, + "epoch": 0.8851079179943486, + "flos": 21500580251520.0, + "grad_norm": 1.8925678362840195, + "language_loss": 0.77720237, + "learning_rate": 1.367656671978037e-07, + "loss": 0.79838097, + "num_input_tokens_seen": 159233610, + "step": 7361, + "time_per_iteration": 2.6804020404815674 + }, + { + "auxiliary_loss_clip": 0.0112405, + "auxiliary_loss_mlp": 0.01023982, + "balance_loss_clip": 0.97093683, + "balance_loss_mlp": 1.01739872, + "epoch": 0.8852281608849877, + "flos": 15300711198720.0, + "grad_norm": 2.0346082947767323, + "language_loss": 0.73914856, + "learning_rate": 1.36482698194498e-07, + "loss": 0.76062882, + "num_input_tokens_seen": 159250155, + "step": 7362, + "time_per_iteration": 2.6740734577178955 + }, + { + "auxiliary_loss_clip": 0.0111602, + "auxiliary_loss_mlp": 0.01027691, + "balance_loss_clip": 0.96927738, + "balance_loss_mlp": 1.02021098, + "epoch": 0.8853484037756267, + "flos": 23295719283840.0, + "grad_norm": 1.7412366977815414, + "language_loss": 0.72273731, + "learning_rate": 1.3620001188528506e-07, + "loss": 0.74417436, + "num_input_tokens_seen": 159270875, + "step": 7363, + "time_per_iteration": 2.647442102432251 + }, + { + "auxiliary_loss_clip": 0.01142732, + "auxiliary_loss_mlp": 0.01025403, + "balance_loss_clip": 1.00800014, + "balance_loss_mlp": 1.0184952, + "epoch": 0.8854686466662659, + "flos": 25114773795840.0, + "grad_norm": 2.945338781961673, + "language_loss": 0.74224246, + "learning_rate": 1.3591760831304865e-07, + "loss": 0.76392382, + "num_input_tokens_seen": 159288565, + "step": 7364, + "time_per_iteration": 2.68462872505188 + }, + { + "auxiliary_loss_clip": 0.01163163, + "auxiliary_loss_mlp": 0.01024911, + "balance_loss_clip": 1.0476532, + "balance_loss_mlp": 1.01724005, + "epoch": 0.885588889556905, + "flos": 21390873137280.0, + "grad_norm": 1.8615751565676524, + "language_loss": 0.79335189, + "learning_rate": 1.356354875206287e-07, + "loss": 0.81523263, + "num_input_tokens_seen": 159306400, + "step": 7365, + "time_per_iteration": 2.6072466373443604 + }, + { + "auxiliary_loss_clip": 0.01095647, + "auxiliary_loss_mlp": 0.01027035, + "balance_loss_clip": 0.93693912, + "balance_loss_mlp": 1.01985216, + "epoch": 0.885709132447544, + "flos": 26906752431360.0, + "grad_norm": 1.8326541009664155, + "language_loss": 0.69771695, + "learning_rate": 1.3535364955082296e-07, + "loss": 0.71894372, + "num_input_tokens_seen": 159326250, + "step": 7366, + "time_per_iteration": 2.7256431579589844 + }, + { + "auxiliary_loss_clip": 0.01162689, + "auxiliary_loss_mlp": 0.01022655, + "balance_loss_clip": 1.0489881, + "balance_loss_mlp": 1.01551461, + "epoch": 0.8858293753381832, + "flos": 26103394800000.0, + "grad_norm": 1.7073487076419216, + "language_loss": 0.64750743, + "learning_rate": 1.3507209444638613e-07, + "loss": 0.66936094, + "num_input_tokens_seen": 159348250, + "step": 7367, + "time_per_iteration": 2.629197835922241 + }, + { + "auxiliary_loss_clip": 0.01141253, + "auxiliary_loss_mlp": 0.01030344, + "balance_loss_clip": 1.00901604, + "balance_loss_mlp": 1.02262509, + "epoch": 0.8859496182288222, + "flos": 23292810282240.0, + "grad_norm": 1.8341257103616295, + "language_loss": 0.74026424, + "learning_rate": 1.347908222500298e-07, + "loss": 0.76198018, + "num_input_tokens_seen": 159368325, + "step": 7368, + "time_per_iteration": 2.612192153930664 + }, + { + "auxiliary_loss_clip": 0.01081745, + "auxiliary_loss_mlp": 0.01023817, + "balance_loss_clip": 0.93313485, + "balance_loss_mlp": 1.01679873, + "epoch": 0.8860698611194613, + "flos": 16872916469760.0, + "grad_norm": 1.8156202992245165, + "language_loss": 0.69892222, + "learning_rate": 1.3450983300442276e-07, + "loss": 0.71997786, + "num_input_tokens_seen": 159387555, + "step": 7369, + "time_per_iteration": 2.7116265296936035 + }, + { + "auxiliary_loss_clip": 0.01140779, + "auxiliary_loss_mlp": 0.0102404, + "balance_loss_clip": 1.00858557, + "balance_loss_mlp": 1.01697075, + "epoch": 0.8861901040101005, + "flos": 24681404206080.0, + "grad_norm": 1.9437094760362708, + "language_loss": 0.73536503, + "learning_rate": 1.3422912675219068e-07, + "loss": 0.75701326, + "num_input_tokens_seen": 159407310, + "step": 7370, + "time_per_iteration": 3.5607211589813232 + }, + { + "auxiliary_loss_clip": 0.01161687, + "auxiliary_loss_mlp": 0.01025147, + "balance_loss_clip": 1.04799485, + "balance_loss_mlp": 1.01880181, + "epoch": 0.8863103469007395, + "flos": 24423026699520.0, + "grad_norm": 1.9617567648861827, + "language_loss": 0.79315817, + "learning_rate": 1.339487035359166e-07, + "loss": 0.81502652, + "num_input_tokens_seen": 159427680, + "step": 7371, + "time_per_iteration": 3.5660629272460938 + }, + { + "auxiliary_loss_clip": 0.01118874, + "auxiliary_loss_mlp": 0.00507876, + "balance_loss_clip": 0.97304386, + "balance_loss_mlp": 0.0, + "epoch": 0.8864305897913786, + "flos": 22053964158720.0, + "grad_norm": 7.663781081112261, + "language_loss": 0.84779787, + "learning_rate": 1.336685633981409e-07, + "loss": 0.86406541, + "num_input_tokens_seen": 159448765, + "step": 7372, + "time_per_iteration": 3.674288272857666 + }, + { + "auxiliary_loss_clip": 0.01140919, + "auxiliary_loss_mlp": 0.01026562, + "balance_loss_clip": 1.00878942, + "balance_loss_mlp": 1.01948428, + "epoch": 0.8865508326820177, + "flos": 19099449843840.0, + "grad_norm": 2.059458507768744, + "language_loss": 0.75108975, + "learning_rate": 1.333887063813597e-07, + "loss": 0.77276456, + "num_input_tokens_seen": 159466870, + "step": 7373, + "time_per_iteration": 2.6296863555908203 + }, + { + "auxiliary_loss_clip": 0.01118699, + "auxiliary_loss_mlp": 0.01027615, + "balance_loss_clip": 0.96959543, + "balance_loss_mlp": 1.02085567, + "epoch": 0.8866710755726568, + "flos": 15414189240960.0, + "grad_norm": 12.830544445573741, + "language_loss": 0.66522574, + "learning_rate": 1.331091325280278e-07, + "loss": 0.6866889, + "num_input_tokens_seen": 159485840, + "step": 7374, + "time_per_iteration": 2.621542453765869 + }, + { + "auxiliary_loss_clip": 0.01064656, + "auxiliary_loss_mlp": 0.01026216, + "balance_loss_clip": 0.89512545, + "balance_loss_mlp": 1.018646, + "epoch": 0.8867913184632958, + "flos": 20083689388800.0, + "grad_norm": 1.6285455637693915, + "language_loss": 0.78411132, + "learning_rate": 1.3282984188055625e-07, + "loss": 0.80502003, + "num_input_tokens_seen": 159505630, + "step": 7375, + "time_per_iteration": 2.806675672531128 + }, + { + "auxiliary_loss_clip": 0.0116624, + "auxiliary_loss_mlp": 0.01027381, + "balance_loss_clip": 1.04940224, + "balance_loss_mlp": 1.02048731, + "epoch": 0.8869115613539349, + "flos": 23365852588800.0, + "grad_norm": 1.8857047694782967, + "language_loss": 0.79803491, + "learning_rate": 1.3255083448131288e-07, + "loss": 0.81997114, + "num_input_tokens_seen": 159524675, + "step": 7376, + "time_per_iteration": 2.6325676441192627 + }, + { + "auxiliary_loss_clip": 0.01144003, + "auxiliary_loss_mlp": 0.01026186, + "balance_loss_clip": 1.00799012, + "balance_loss_mlp": 1.01842546, + "epoch": 0.8870318042445741, + "flos": 21286840371840.0, + "grad_norm": 1.958933677763797, + "language_loss": 0.78796268, + "learning_rate": 1.3227211037262365e-07, + "loss": 0.80966455, + "num_input_tokens_seen": 159541915, + "step": 7377, + "time_per_iteration": 2.622736930847168 + }, + { + "auxiliary_loss_clip": 0.0107295, + "auxiliary_loss_mlp": 0.01026742, + "balance_loss_clip": 0.892699, + "balance_loss_mlp": 1.01959848, + "epoch": 0.8871520471352131, + "flos": 20010862563840.0, + "grad_norm": 3.0543784922708297, + "language_loss": 0.85510063, + "learning_rate": 1.319936695967696e-07, + "loss": 0.87609756, + "num_input_tokens_seen": 159559740, + "step": 7378, + "time_per_iteration": 2.740985155105591 + }, + { + "auxiliary_loss_clip": 0.01167926, + "auxiliary_loss_mlp": 0.01022237, + "balance_loss_clip": 1.0471313, + "balance_loss_mlp": 1.01447678, + "epoch": 0.8872722900258522, + "flos": 22601422321920.0, + "grad_norm": 2.113354036454635, + "language_loss": 0.82135254, + "learning_rate": 1.3171551219599097e-07, + "loss": 0.84325415, + "num_input_tokens_seen": 159578265, + "step": 7379, + "time_per_iteration": 3.5885725021362305 + }, + { + "auxiliary_loss_clip": 0.0116381, + "auxiliary_loss_mlp": 0.01024073, + "balance_loss_clip": 1.05012345, + "balance_loss_mlp": 1.01711988, + "epoch": 0.8873925329164913, + "flos": 22163276223360.0, + "grad_norm": 3.126930786557576, + "language_loss": 0.77955389, + "learning_rate": 1.3143763821248377e-07, + "loss": 0.80143267, + "num_input_tokens_seen": 159595350, + "step": 7380, + "time_per_iteration": 2.634666681289673 + }, + { + "auxiliary_loss_clip": 0.01161142, + "auxiliary_loss_mlp": 0.01027344, + "balance_loss_clip": 1.04624152, + "balance_loss_mlp": 1.02075195, + "epoch": 0.8875127758071304, + "flos": 19208223204480.0, + "grad_norm": 1.9449424345793478, + "language_loss": 0.72194123, + "learning_rate": 1.3116004768840118e-07, + "loss": 0.74382615, + "num_input_tokens_seen": 159613725, + "step": 7381, + "time_per_iteration": 2.632207155227661 + }, + { + "auxiliary_loss_clip": 0.01167377, + "auxiliary_loss_mlp": 0.01026323, + "balance_loss_clip": 1.04956818, + "balance_loss_mlp": 1.01912236, + "epoch": 0.8876330186977694, + "flos": 18110900666880.0, + "grad_norm": 1.5698396327361581, + "language_loss": 0.74063003, + "learning_rate": 1.3088274066585348e-07, + "loss": 0.76256704, + "num_input_tokens_seen": 159631335, + "step": 7382, + "time_per_iteration": 2.625612735748291 + }, + { + "auxiliary_loss_clip": 0.01101096, + "auxiliary_loss_mlp": 0.01023284, + "balance_loss_clip": 0.93297541, + "balance_loss_mlp": 1.01630974, + "epoch": 0.8877532615884086, + "flos": 22009434272640.0, + "grad_norm": 2.940786456290449, + "language_loss": 0.90521044, + "learning_rate": 1.3060571718690749e-07, + "loss": 0.92645419, + "num_input_tokens_seen": 159648830, + "step": 7383, + "time_per_iteration": 2.716554880142212 + }, + { + "auxiliary_loss_clip": 0.01002938, + "auxiliary_loss_mlp": 0.00500554, + "balance_loss_clip": 0.89819205, + "balance_loss_mlp": 0.0, + "epoch": 0.8878735044790477, + "flos": 72136924346880.0, + "grad_norm": 0.745998436443947, + "language_loss": 0.56898403, + "learning_rate": 1.3032897729358805e-07, + "loss": 0.58401895, + "num_input_tokens_seen": 159709785, + "step": 7384, + "time_per_iteration": 3.2796437740325928 + }, + { + "auxiliary_loss_clip": 0.01036757, + "auxiliary_loss_mlp": 0.00509393, + "balance_loss_clip": 0.85326147, + "balance_loss_mlp": 0.0, + "epoch": 0.8879937473696867, + "flos": 27526355061120.0, + "grad_norm": 2.0313496376407625, + "language_loss": 0.79941058, + "learning_rate": 1.3005252102787645e-07, + "loss": 0.81487209, + "num_input_tokens_seen": 159728725, + "step": 7385, + "time_per_iteration": 3.003946542739868 + }, + { + "auxiliary_loss_clip": 0.01143697, + "auxiliary_loss_mlp": 0.01024016, + "balance_loss_clip": 1.00981748, + "balance_loss_mlp": 1.01673794, + "epoch": 0.8881139902603259, + "flos": 22234091886720.0, + "grad_norm": 1.5094073791281721, + "language_loss": 0.7353999, + "learning_rate": 1.297763484317105e-07, + "loss": 0.75707698, + "num_input_tokens_seen": 159747020, + "step": 7386, + "time_per_iteration": 2.802701234817505 + }, + { + "auxiliary_loss_clip": 0.01066975, + "auxiliary_loss_mlp": 0.00509589, + "balance_loss_clip": 0.89178288, + "balance_loss_mlp": 0.0, + "epoch": 0.888234233150965, + "flos": 20299548170880.0, + "grad_norm": 4.532733828212624, + "language_loss": 0.7018739, + "learning_rate": 1.2950045954698551e-07, + "loss": 0.71763945, + "num_input_tokens_seen": 159764855, + "step": 7387, + "time_per_iteration": 2.9073143005371094 + }, + { + "auxiliary_loss_clip": 0.01084519, + "auxiliary_loss_mlp": 0.01023894, + "balance_loss_clip": 0.93213868, + "balance_loss_mlp": 1.01732826, + "epoch": 0.888354476041604, + "flos": 18147996437760.0, + "grad_norm": 2.127760449340493, + "language_loss": 0.75580215, + "learning_rate": 1.2922485441555343e-07, + "loss": 0.77688622, + "num_input_tokens_seen": 159783935, + "step": 7388, + "time_per_iteration": 2.8244998455047607 + }, + { + "auxiliary_loss_clip": 0.01161323, + "auxiliary_loss_mlp": 0.01022292, + "balance_loss_clip": 1.04547763, + "balance_loss_mlp": 1.01545787, + "epoch": 0.8884747189322432, + "flos": 22014282608640.0, + "grad_norm": 1.809546381117307, + "language_loss": 0.81807876, + "learning_rate": 1.2894953307922363e-07, + "loss": 0.83991492, + "num_input_tokens_seen": 159802895, + "step": 7389, + "time_per_iteration": 2.6322216987609863 + }, + { + "auxiliary_loss_clip": 0.0108972, + "auxiliary_loss_mlp": 0.01030458, + "balance_loss_clip": 0.93338811, + "balance_loss_mlp": 1.02257204, + "epoch": 0.8885949618228822, + "flos": 19786779567360.0, + "grad_norm": 1.8576871037006641, + "language_loss": 0.8383373, + "learning_rate": 1.2867449557976208e-07, + "loss": 0.85953915, + "num_input_tokens_seen": 159820995, + "step": 7390, + "time_per_iteration": 2.820617914199829 + }, + { + "auxiliary_loss_clip": 0.01139014, + "auxiliary_loss_mlp": 0.01023131, + "balance_loss_clip": 1.01006866, + "balance_loss_mlp": 1.0162828, + "epoch": 0.8887152047135213, + "flos": 20047599198720.0, + "grad_norm": 2.7420492074706373, + "language_loss": 0.76181161, + "learning_rate": 1.283997419588916e-07, + "loss": 0.78343302, + "num_input_tokens_seen": 159840465, + "step": 7391, + "time_per_iteration": 2.7360687255859375 + }, + { + "auxiliary_loss_clip": 0.01145265, + "auxiliary_loss_mlp": 0.01026804, + "balance_loss_clip": 1.00954366, + "balance_loss_mlp": 1.0196867, + "epoch": 0.8888354476041604, + "flos": 18588117784320.0, + "grad_norm": 2.026173560926451, + "language_loss": 0.61805218, + "learning_rate": 1.2812527225829216e-07, + "loss": 0.63977289, + "num_input_tokens_seen": 159858690, + "step": 7392, + "time_per_iteration": 2.7583119869232178 + }, + { + "auxiliary_loss_clip": 0.01146367, + "auxiliary_loss_mlp": 0.01029021, + "balance_loss_clip": 1.01130199, + "balance_loss_mlp": 1.02053976, + "epoch": 0.8889556904947995, + "flos": 21689794120320.0, + "grad_norm": 1.9033399411873844, + "language_loss": 0.76547933, + "learning_rate": 1.2785108651960052e-07, + "loss": 0.78723323, + "num_input_tokens_seen": 159880325, + "step": 7393, + "time_per_iteration": 2.772860288619995 + }, + { + "auxiliary_loss_clip": 0.01142312, + "auxiliary_loss_mlp": 0.01030898, + "balance_loss_clip": 1.00794673, + "balance_loss_mlp": 1.02398705, + "epoch": 0.8890759333854386, + "flos": 27381204201600.0, + "grad_norm": 2.07112514519537, + "language_loss": 0.80740333, + "learning_rate": 1.2757718478441094e-07, + "loss": 0.82913542, + "num_input_tokens_seen": 159901070, + "step": 7394, + "time_per_iteration": 2.7259764671325684 + }, + { + "auxiliary_loss_clip": 0.01116294, + "auxiliary_loss_mlp": 0.01023348, + "balance_loss_clip": 0.96868384, + "balance_loss_mlp": 1.01662159, + "epoch": 0.8891961762760777, + "flos": 24498834353280.0, + "grad_norm": 1.6513412926475026, + "language_loss": 0.76913357, + "learning_rate": 1.2730356709427302e-07, + "loss": 0.79052997, + "num_input_tokens_seen": 159919750, + "step": 7395, + "time_per_iteration": 3.8078315258026123 + }, + { + "auxiliary_loss_clip": 0.01143463, + "auxiliary_loss_mlp": 0.01021226, + "balance_loss_clip": 1.01631498, + "balance_loss_mlp": 1.01455331, + "epoch": 0.8893164191667168, + "flos": 41499770895360.0, + "grad_norm": 6.3018654515777275, + "language_loss": 0.59837449, + "learning_rate": 1.2703023349069542e-07, + "loss": 0.62002134, + "num_input_tokens_seen": 159944600, + "step": 7396, + "time_per_iteration": 2.9066288471221924 + }, + { + "auxiliary_loss_clip": 0.01137699, + "auxiliary_loss_mlp": 0.01024714, + "balance_loss_clip": 1.01043701, + "balance_loss_mlp": 1.01781726, + "epoch": 0.8894366620573558, + "flos": 33583623120000.0, + "grad_norm": 1.9394750753913872, + "language_loss": 0.61737901, + "learning_rate": 1.2675718401514223e-07, + "loss": 0.63900316, + "num_input_tokens_seen": 159968780, + "step": 7397, + "time_per_iteration": 3.6594607830047607 + }, + { + "auxiliary_loss_clip": 0.0111661, + "auxiliary_loss_mlp": 0.0102597, + "balance_loss_clip": 0.9713558, + "balance_loss_mlp": 1.0187484, + "epoch": 0.889556904947995, + "flos": 16909832672640.0, + "grad_norm": 3.0948386482563017, + "language_loss": 0.74852622, + "learning_rate": 1.264844187090346e-07, + "loss": 0.76995206, + "num_input_tokens_seen": 159985905, + "step": 7398, + "time_per_iteration": 3.811847448348999 + }, + { + "auxiliary_loss_clip": 0.01112402, + "auxiliary_loss_mlp": 0.01023063, + "balance_loss_clip": 0.96953619, + "balance_loss_mlp": 1.01631618, + "epoch": 0.889677147838634, + "flos": 26030855283840.0, + "grad_norm": 2.7963178379220817, + "language_loss": 0.75347853, + "learning_rate": 1.262119376137516e-07, + "loss": 0.77483314, + "num_input_tokens_seen": 160006965, + "step": 7399, + "time_per_iteration": 2.8419582843780518 + }, + { + "auxiliary_loss_clip": 0.01132737, + "auxiliary_loss_mlp": 0.01025267, + "balance_loss_clip": 1.00733769, + "balance_loss_mlp": 1.01833487, + "epoch": 0.8897973907292731, + "flos": 26468283110400.0, + "grad_norm": 1.6047030022975433, + "language_loss": 0.84961593, + "learning_rate": 1.2593974077062707e-07, + "loss": 0.87119603, + "num_input_tokens_seen": 160028585, + "step": 7400, + "time_per_iteration": 2.790278911590576 + }, + { + "auxiliary_loss_clip": 0.01084335, + "auxiliary_loss_mlp": 0.01022904, + "balance_loss_clip": 0.93100536, + "balance_loss_mlp": 1.0154326, + "epoch": 0.8899176336199123, + "flos": 26249694894720.0, + "grad_norm": 1.581747999962922, + "language_loss": 0.63587582, + "learning_rate": 1.2566782822095423e-07, + "loss": 0.65694821, + "num_input_tokens_seen": 160048840, + "step": 7401, + "time_per_iteration": 2.8339717388153076 + }, + { + "auxiliary_loss_clip": 0.01096909, + "auxiliary_loss_mlp": 0.0102623, + "balance_loss_clip": 0.93416005, + "balance_loss_mlp": 1.01970351, + "epoch": 0.8900378765105513, + "flos": 20811742156800.0, + "grad_norm": 1.7418148285965795, + "language_loss": 0.71388012, + "learning_rate": 1.2539620000598162e-07, + "loss": 0.73511147, + "num_input_tokens_seen": 160068175, + "step": 7402, + "time_per_iteration": 2.7536423206329346 + }, + { + "auxiliary_loss_clip": 0.01163175, + "auxiliary_loss_mlp": 0.01027205, + "balance_loss_clip": 1.04757869, + "balance_loss_mlp": 1.01952159, + "epoch": 0.8901581194011904, + "flos": 16472333018880.0, + "grad_norm": 1.7224254604171712, + "language_loss": 0.79642588, + "learning_rate": 1.2512485616691492e-07, + "loss": 0.81832969, + "num_input_tokens_seen": 160085230, + "step": 7403, + "time_per_iteration": 2.6591289043426514 + }, + { + "auxiliary_loss_clip": 0.01095332, + "auxiliary_loss_mlp": 0.01027711, + "balance_loss_clip": 0.93241537, + "balance_loss_mlp": 1.02028155, + "epoch": 0.8902783622918296, + "flos": 35155253773440.0, + "grad_norm": 1.4508994023096897, + "language_loss": 0.81109089, + "learning_rate": 1.2485379674491681e-07, + "loss": 0.83232129, + "num_input_tokens_seen": 160111425, + "step": 7404, + "time_per_iteration": 3.7476742267608643 + }, + { + "auxiliary_loss_clip": 0.01118544, + "auxiliary_loss_mlp": 0.0102493, + "balance_loss_clip": 0.9733789, + "balance_loss_mlp": 1.01751494, + "epoch": 0.8903986051824686, + "flos": 17201068145280.0, + "grad_norm": 2.333603864958072, + "language_loss": 0.79050905, + "learning_rate": 1.2458302178110657e-07, + "loss": 0.81194377, + "num_input_tokens_seen": 160129790, + "step": 7405, + "time_per_iteration": 2.6965999603271484 + }, + { + "auxiliary_loss_clip": 0.01082435, + "auxiliary_loss_mlp": 0.01024335, + "balance_loss_clip": 0.92964369, + "balance_loss_mlp": 1.01749563, + "epoch": 0.8905188480731077, + "flos": 25483863997440.0, + "grad_norm": 2.144137785252278, + "language_loss": 0.82528162, + "learning_rate": 1.2431253131656118e-07, + "loss": 0.84634936, + "num_input_tokens_seen": 160149265, + "step": 7406, + "time_per_iteration": 2.7612123489379883 + }, + { + "auxiliary_loss_clip": 0.01110008, + "auxiliary_loss_mlp": 0.01028298, + "balance_loss_clip": 0.97079623, + "balance_loss_mlp": 1.02113318, + "epoch": 0.8906390909637467, + "flos": 23365888502400.0, + "grad_norm": 1.7024183090953648, + "language_loss": 0.76847446, + "learning_rate": 1.240423253923133e-07, + "loss": 0.78985751, + "num_input_tokens_seen": 160168870, + "step": 7407, + "time_per_iteration": 2.670480489730835 + }, + { + "auxiliary_loss_clip": 0.01142029, + "auxiliary_loss_mlp": 0.01023242, + "balance_loss_clip": 1.00972104, + "balance_loss_mlp": 1.01619041, + "epoch": 0.8907593338543859, + "flos": 21068790860160.0, + "grad_norm": 2.5781484576426674, + "language_loss": 0.69547969, + "learning_rate": 1.237724040493533e-07, + "loss": 0.71713239, + "num_input_tokens_seen": 160187495, + "step": 7408, + "time_per_iteration": 2.7075536251068115 + }, + { + "auxiliary_loss_clip": 0.01168936, + "auxiliary_loss_mlp": 0.01028666, + "balance_loss_clip": 1.05136681, + "balance_loss_mlp": 1.02082241, + "epoch": 0.8908795767450249, + "flos": 21869562712320.0, + "grad_norm": 3.4647090157701648, + "language_loss": 0.72988164, + "learning_rate": 1.2350276732862773e-07, + "loss": 0.7518577, + "num_input_tokens_seen": 160208520, + "step": 7409, + "time_per_iteration": 2.63274884223938 + }, + { + "auxiliary_loss_clip": 0.01041426, + "auxiliary_loss_mlp": 0.01000787, + "balance_loss_clip": 0.97281694, + "balance_loss_mlp": 0.99986327, + "epoch": 0.890999819635664, + "flos": 66307869348480.0, + "grad_norm": 0.8343565863821668, + "language_loss": 0.56612319, + "learning_rate": 1.2323341527103993e-07, + "loss": 0.58654529, + "num_input_tokens_seen": 160263720, + "step": 7410, + "time_per_iteration": 3.12510085105896 + }, + { + "auxiliary_loss_clip": 0.0116261, + "auxiliary_loss_mlp": 0.01023586, + "balance_loss_clip": 1.04683518, + "balance_loss_mlp": 1.0164274, + "epoch": 0.8911200625263032, + "flos": 26869908055680.0, + "grad_norm": 1.9761881200463325, + "language_loss": 0.85054004, + "learning_rate": 1.2296434791745135e-07, + "loss": 0.87240195, + "num_input_tokens_seen": 160282170, + "step": 7411, + "time_per_iteration": 2.6466047763824463 + }, + { + "auxiliary_loss_clip": 0.01142195, + "auxiliary_loss_mlp": 0.01019847, + "balance_loss_clip": 1.00887203, + "balance_loss_mlp": 1.01202703, + "epoch": 0.8912403054169422, + "flos": 20885825957760.0, + "grad_norm": 1.694630606214373, + "language_loss": 0.76853704, + "learning_rate": 1.2269556530867875e-07, + "loss": 0.79015756, + "num_input_tokens_seen": 160300725, + "step": 7412, + "time_per_iteration": 2.763704538345337 + }, + { + "auxiliary_loss_clip": 0.01169071, + "auxiliary_loss_mlp": 0.01028635, + "balance_loss_clip": 1.0496372, + "balance_loss_mlp": 1.02113986, + "epoch": 0.8913605483075813, + "flos": 27016567286400.0, + "grad_norm": 3.569070842445597, + "language_loss": 0.81785524, + "learning_rate": 1.2242706748549614e-07, + "loss": 0.83983231, + "num_input_tokens_seen": 160318720, + "step": 7413, + "time_per_iteration": 2.6463778018951416 + }, + { + "auxiliary_loss_clip": 0.01117853, + "auxiliary_loss_mlp": 0.01026872, + "balance_loss_clip": 0.96767378, + "balance_loss_mlp": 1.02016962, + "epoch": 0.8914807911982204, + "flos": 23621500661760.0, + "grad_norm": 1.7782521730962473, + "language_loss": 0.82260287, + "learning_rate": 1.2215885448863473e-07, + "loss": 0.84405017, + "num_input_tokens_seen": 160339595, + "step": 7414, + "time_per_iteration": 2.744532346725464 + }, + { + "auxiliary_loss_clip": 0.0111546, + "auxiliary_loss_mlp": 0.01025886, + "balance_loss_clip": 0.97204751, + "balance_loss_mlp": 1.01860845, + "epoch": 0.8916010340888595, + "flos": 24462277286400.0, + "grad_norm": 2.2112490083400016, + "language_loss": 0.80509919, + "learning_rate": 1.2189092635878152e-07, + "loss": 0.82651263, + "num_input_tokens_seen": 160361045, + "step": 7415, + "time_per_iteration": 2.7802517414093018 + }, + { + "auxiliary_loss_clip": 0.01085801, + "auxiliary_loss_mlp": 0.01024022, + "balance_loss_clip": 0.9321028, + "balance_loss_mlp": 1.01652932, + "epoch": 0.8917212769794985, + "flos": 21215773313280.0, + "grad_norm": 1.566231864305172, + "language_loss": 0.77233231, + "learning_rate": 1.216232831365822e-07, + "loss": 0.79343057, + "num_input_tokens_seen": 160379990, + "step": 7416, + "time_per_iteration": 2.7951996326446533 + }, + { + "auxiliary_loss_clip": 0.01121939, + "auxiliary_loss_mlp": 0.01027878, + "balance_loss_clip": 0.97120011, + "balance_loss_mlp": 1.02041841, + "epoch": 0.8918415198701377, + "flos": 25513992529920.0, + "grad_norm": 1.8453432604329882, + "language_loss": 0.80955648, + "learning_rate": 1.2135592486263678e-07, + "loss": 0.83105469, + "num_input_tokens_seen": 160399240, + "step": 7417, + "time_per_iteration": 2.77893328666687 + }, + { + "auxiliary_loss_clip": 0.01117011, + "auxiliary_loss_mlp": 0.01019385, + "balance_loss_clip": 0.97074455, + "balance_loss_mlp": 1.0125277, + "epoch": 0.8919617627607768, + "flos": 37853006693760.0, + "grad_norm": 1.7237435567121941, + "language_loss": 0.60927385, + "learning_rate": 1.2108885157750415e-07, + "loss": 0.63063782, + "num_input_tokens_seen": 160421600, + "step": 7418, + "time_per_iteration": 2.816720485687256 + }, + { + "auxiliary_loss_clip": 0.01089944, + "auxiliary_loss_mlp": 0.00508692, + "balance_loss_clip": 0.93350226, + "balance_loss_mlp": 0.0, + "epoch": 0.8920820056514158, + "flos": 26213676531840.0, + "grad_norm": 1.7681027363644541, + "language_loss": 0.80334491, + "learning_rate": 1.2082206332169897e-07, + "loss": 0.81933129, + "num_input_tokens_seen": 160441695, + "step": 7419, + "time_per_iteration": 2.756859064102173 + }, + { + "auxiliary_loss_clip": 0.01113688, + "auxiliary_loss_mlp": 0.01028687, + "balance_loss_clip": 0.97181177, + "balance_loss_mlp": 1.02166879, + "epoch": 0.892202248542055, + "flos": 17383135207680.0, + "grad_norm": 2.368813986672928, + "language_loss": 0.73573315, + "learning_rate": 1.2055556013569225e-07, + "loss": 0.75715697, + "num_input_tokens_seen": 160457205, + "step": 7420, + "time_per_iteration": 2.6523966789245605 + }, + { + "auxiliary_loss_clip": 0.01118694, + "auxiliary_loss_mlp": 0.01022575, + "balance_loss_clip": 0.97144139, + "balance_loss_mlp": 1.01587796, + "epoch": 0.892322491432694, + "flos": 21324223451520.0, + "grad_norm": 1.7621292559296866, + "language_loss": 0.82040083, + "learning_rate": 1.2028934205991315e-07, + "loss": 0.84181356, + "num_input_tokens_seen": 160476525, + "step": 7421, + "time_per_iteration": 2.7528865337371826 + }, + { + "auxiliary_loss_clip": 0.01140067, + "auxiliary_loss_mlp": 0.0102248, + "balance_loss_clip": 1.00852442, + "balance_loss_mlp": 1.01520228, + "epoch": 0.8924427343233331, + "flos": 24029374573440.0, + "grad_norm": 1.4137937023615588, + "language_loss": 0.76526582, + "learning_rate": 1.2002340913474607e-07, + "loss": 0.78689128, + "num_input_tokens_seen": 160500160, + "step": 7422, + "time_per_iteration": 4.547281980514526 + }, + { + "auxiliary_loss_clip": 0.01166162, + "auxiliary_loss_mlp": 0.01023208, + "balance_loss_clip": 1.04823077, + "balance_loss_mlp": 1.01530993, + "epoch": 0.8925629772139723, + "flos": 30008069631360.0, + "grad_norm": 2.043939130766285, + "language_loss": 0.73876178, + "learning_rate": 1.1975776140053317e-07, + "loss": 0.76065546, + "num_input_tokens_seen": 160520130, + "step": 7423, + "time_per_iteration": 2.6823067665100098 + }, + { + "auxiliary_loss_clip": 0.01072275, + "auxiliary_loss_mlp": 0.0102772, + "balance_loss_clip": 0.89535177, + "balance_loss_mlp": 1.01995969, + "epoch": 0.8926832201046113, + "flos": 22601709630720.0, + "grad_norm": 1.8842008916207595, + "language_loss": 0.7344448, + "learning_rate": 1.194923988975729e-07, + "loss": 0.75544477, + "num_input_tokens_seen": 160539730, + "step": 7424, + "time_per_iteration": 3.745990037918091 + }, + { + "auxiliary_loss_clip": 0.01091791, + "auxiliary_loss_mlp": 0.01023225, + "balance_loss_clip": 0.93299985, + "balance_loss_mlp": 1.01656723, + "epoch": 0.8928034629952504, + "flos": 13297722117120.0, + "grad_norm": 2.644553709220112, + "language_loss": 0.73615336, + "learning_rate": 1.192273216661206e-07, + "loss": 0.75730348, + "num_input_tokens_seen": 160557820, + "step": 7425, + "time_per_iteration": 2.627596378326416 + }, + { + "auxiliary_loss_clip": 0.00961966, + "auxiliary_loss_mlp": 0.01002197, + "balance_loss_clip": 0.82445657, + "balance_loss_mlp": 1.00124311, + "epoch": 0.8929237058858895, + "flos": 54854556744960.0, + "grad_norm": 0.7707128016006023, + "language_loss": 0.57494712, + "learning_rate": 1.189625297463881e-07, + "loss": 0.59458876, + "num_input_tokens_seen": 160619510, + "step": 7426, + "time_per_iteration": 3.3598477840423584 + }, + { + "auxiliary_loss_clip": 0.01041748, + "auxiliary_loss_mlp": 0.01022232, + "balance_loss_clip": 0.85182726, + "balance_loss_mlp": 1.01533318, + "epoch": 0.8930439487765286, + "flos": 28883850785280.0, + "grad_norm": 1.8213355580844426, + "language_loss": 0.79544073, + "learning_rate": 1.1869802317854394e-07, + "loss": 0.81608057, + "num_input_tokens_seen": 160643295, + "step": 7427, + "time_per_iteration": 3.1963024139404297 + }, + { + "auxiliary_loss_clip": 0.01073237, + "auxiliary_loss_mlp": 0.01026438, + "balance_loss_clip": 0.8943029, + "balance_loss_mlp": 1.0193305, + "epoch": 0.8931641916671677, + "flos": 22419283432320.0, + "grad_norm": 1.9750794579119975, + "language_loss": 0.72441518, + "learning_rate": 1.1843380200271425e-07, + "loss": 0.74541199, + "num_input_tokens_seen": 160662495, + "step": 7428, + "time_per_iteration": 2.8972113132476807 + }, + { + "auxiliary_loss_clip": 0.01089081, + "auxiliary_loss_mlp": 0.01025974, + "balance_loss_clip": 0.93182623, + "balance_loss_mlp": 1.01903868, + "epoch": 0.8932844345578068, + "flos": 25843149786240.0, + "grad_norm": 1.7703427148808286, + "language_loss": 0.80873585, + "learning_rate": 1.181698662589805e-07, + "loss": 0.82988638, + "num_input_tokens_seen": 160682080, + "step": 7429, + "time_per_iteration": 2.7545325756073 + }, + { + "auxiliary_loss_clip": 0.01139251, + "auxiliary_loss_mlp": 0.01025354, + "balance_loss_clip": 1.00719297, + "balance_loss_mlp": 1.0183444, + "epoch": 0.8934046774484459, + "flos": 22925803069440.0, + "grad_norm": 1.8745480773524612, + "language_loss": 0.7627337, + "learning_rate": 1.1790621598738249e-07, + "loss": 0.78437984, + "num_input_tokens_seen": 160700395, + "step": 7430, + "time_per_iteration": 3.837284564971924 + }, + { + "auxiliary_loss_clip": 0.01164815, + "auxiliary_loss_mlp": 0.01023651, + "balance_loss_clip": 1.05184698, + "balance_loss_mlp": 1.01690412, + "epoch": 0.8935249203390849, + "flos": 24462097718400.0, + "grad_norm": 1.9165389190493174, + "language_loss": 0.74693131, + "learning_rate": 1.1764285122791461e-07, + "loss": 0.76881593, + "num_input_tokens_seen": 160721115, + "step": 7431, + "time_per_iteration": 2.762174129486084 + }, + { + "auxiliary_loss_clip": 0.01139372, + "auxiliary_loss_mlp": 0.01019413, + "balance_loss_clip": 1.00757813, + "balance_loss_mlp": 1.01292825, + "epoch": 0.8936451632297241, + "flos": 15742735966080.0, + "grad_norm": 1.8218479710578714, + "language_loss": 0.7729497, + "learning_rate": 1.173797720205294e-07, + "loss": 0.79453754, + "num_input_tokens_seen": 160739150, + "step": 7432, + "time_per_iteration": 2.655207872390747 + }, + { + "auxiliary_loss_clip": 0.01143609, + "auxiliary_loss_mlp": 0.01030607, + "balance_loss_clip": 1.01108086, + "balance_loss_mlp": 1.02268207, + "epoch": 0.8937654061203631, + "flos": 35115500396160.0, + "grad_norm": 3.1776555806542843, + "language_loss": 0.71786278, + "learning_rate": 1.1711697840513602e-07, + "loss": 0.73960495, + "num_input_tokens_seen": 160758585, + "step": 7433, + "time_per_iteration": 2.7389516830444336 + }, + { + "auxiliary_loss_clip": 0.01134623, + "auxiliary_loss_mlp": 0.01026609, + "balance_loss_clip": 1.00870192, + "balance_loss_mlp": 1.0194056, + "epoch": 0.8938856490110022, + "flos": 16107444708480.0, + "grad_norm": 2.0593818819752863, + "language_loss": 0.70844352, + "learning_rate": 1.1685447042160012e-07, + "loss": 0.73005587, + "num_input_tokens_seen": 160776620, + "step": 7434, + "time_per_iteration": 2.645660877227783 + }, + { + "auxiliary_loss_clip": 0.01166183, + "auxiliary_loss_mlp": 0.0102156, + "balance_loss_clip": 1.04775238, + "balance_loss_mlp": 1.01406789, + "epoch": 0.8940058919016414, + "flos": 20704189858560.0, + "grad_norm": 1.557747212298703, + "language_loss": 0.71361041, + "learning_rate": 1.1659224810974367e-07, + "loss": 0.73548782, + "num_input_tokens_seen": 160796580, + "step": 7435, + "time_per_iteration": 2.606388807296753 + }, + { + "auxiliary_loss_clip": 0.01115734, + "auxiliary_loss_mlp": 0.01025251, + "balance_loss_clip": 0.97331238, + "balance_loss_mlp": 1.01811671, + "epoch": 0.8941261347922804, + "flos": 25229041937280.0, + "grad_norm": 1.438947120323784, + "language_loss": 0.6809094, + "learning_rate": 1.1633031150934591e-07, + "loss": 0.70231926, + "num_input_tokens_seen": 160819610, + "step": 7436, + "time_per_iteration": 2.7970314025878906 + }, + { + "auxiliary_loss_clip": 0.01143945, + "auxiliary_loss_mlp": 0.01024913, + "balance_loss_clip": 1.0128181, + "balance_loss_mlp": 1.01765323, + "epoch": 0.8942463776829195, + "flos": 19537236806400.0, + "grad_norm": 2.0029876978087175, + "language_loss": 0.79971397, + "learning_rate": 1.1606866066014176e-07, + "loss": 0.82140255, + "num_input_tokens_seen": 160838660, + "step": 7437, + "time_per_iteration": 2.586324453353882 + }, + { + "auxiliary_loss_clip": 0.01092117, + "auxiliary_loss_mlp": 0.01021924, + "balance_loss_clip": 0.93359238, + "balance_loss_mlp": 1.01523328, + "epoch": 0.8943666205735585, + "flos": 22301567585280.0, + "grad_norm": 2.392185690813213, + "language_loss": 0.75275815, + "learning_rate": 1.1580729560182434e-07, + "loss": 0.7738986, + "num_input_tokens_seen": 160854515, + "step": 7438, + "time_per_iteration": 2.6993327140808105 + }, + { + "auxiliary_loss_clip": 0.01162785, + "auxiliary_loss_mlp": 0.00508839, + "balance_loss_clip": 1.04762626, + "balance_loss_mlp": 0.0, + "epoch": 0.8944868634641977, + "flos": 18912893581440.0, + "grad_norm": 1.7091226390334142, + "language_loss": 0.71067882, + "learning_rate": 1.1554621637404171e-07, + "loss": 0.72739506, + "num_input_tokens_seen": 160872605, + "step": 7439, + "time_per_iteration": 2.610666036605835 + }, + { + "auxiliary_loss_clip": 0.01142985, + "auxiliary_loss_mlp": 0.01022017, + "balance_loss_clip": 1.00907922, + "balance_loss_mlp": 1.01478148, + "epoch": 0.8946071063548368, + "flos": 14460904241280.0, + "grad_norm": 2.21305300556327, + "language_loss": 0.61198843, + "learning_rate": 1.1528542301639999e-07, + "loss": 0.63363844, + "num_input_tokens_seen": 160889395, + "step": 7440, + "time_per_iteration": 2.701385498046875 + }, + { + "auxiliary_loss_clip": 0.01095479, + "auxiliary_loss_mlp": 0.01026026, + "balance_loss_clip": 0.93144691, + "balance_loss_mlp": 1.01871896, + "epoch": 0.8947273492454758, + "flos": 20084084438400.0, + "grad_norm": 2.1948635936153917, + "language_loss": 0.82429248, + "learning_rate": 1.1502491556846105e-07, + "loss": 0.8455075, + "num_input_tokens_seen": 160907890, + "step": 7441, + "time_per_iteration": 2.7815024852752686 + }, + { + "auxiliary_loss_clip": 0.01116414, + "auxiliary_loss_mlp": 0.01025282, + "balance_loss_clip": 0.97209632, + "balance_loss_mlp": 1.01801634, + "epoch": 0.894847592136115, + "flos": 18550555136640.0, + "grad_norm": 2.142145715210463, + "language_loss": 0.81344539, + "learning_rate": 1.1476469406974331e-07, + "loss": 0.83486235, + "num_input_tokens_seen": 160923490, + "step": 7442, + "time_per_iteration": 2.657600164413452 + }, + { + "auxiliary_loss_clip": 0.0116259, + "auxiliary_loss_mlp": 0.01023023, + "balance_loss_clip": 1.04839039, + "balance_loss_mlp": 1.01625776, + "epoch": 0.894967835026754, + "flos": 23478468704640.0, + "grad_norm": 1.6413086442805278, + "language_loss": 0.7677958, + "learning_rate": 1.1450475855972341e-07, + "loss": 0.78965199, + "num_input_tokens_seen": 160944280, + "step": 7443, + "time_per_iteration": 2.671661376953125 + }, + { + "auxiliary_loss_clip": 0.01115698, + "auxiliary_loss_mlp": 0.00509611, + "balance_loss_clip": 0.96885836, + "balance_loss_mlp": 0.0, + "epoch": 0.8950880779173931, + "flos": 15188310564480.0, + "grad_norm": 2.1800223523155733, + "language_loss": 0.70764637, + "learning_rate": 1.1424510907783158e-07, + "loss": 0.72389948, + "num_input_tokens_seen": 160961560, + "step": 7444, + "time_per_iteration": 2.7088968753814697 + }, + { + "auxiliary_loss_clip": 0.01120515, + "auxiliary_loss_mlp": 0.01021649, + "balance_loss_clip": 0.96850383, + "balance_loss_mlp": 1.01504493, + "epoch": 0.8952083208080323, + "flos": 22091957769600.0, + "grad_norm": 1.5464607018286047, + "language_loss": 0.82355344, + "learning_rate": 1.1398574566345787e-07, + "loss": 0.84497505, + "num_input_tokens_seen": 160982195, + "step": 7445, + "time_per_iteration": 2.6881139278411865 + }, + { + "auxiliary_loss_clip": 0.01120454, + "auxiliary_loss_mlp": 0.01027819, + "balance_loss_clip": 0.96937376, + "balance_loss_mlp": 1.02025843, + "epoch": 0.8953285636986713, + "flos": 23254026572160.0, + "grad_norm": 4.47579042805147, + "language_loss": 0.82072496, + "learning_rate": 1.1372666835594702e-07, + "loss": 0.84220767, + "num_input_tokens_seen": 161000520, + "step": 7446, + "time_per_iteration": 2.661200523376465 + }, + { + "auxiliary_loss_clip": 0.01115581, + "auxiliary_loss_mlp": 0.01025842, + "balance_loss_clip": 0.97071433, + "balance_loss_mlp": 1.01894855, + "epoch": 0.8954488065893104, + "flos": 16362661818240.0, + "grad_norm": 2.267603229280069, + "language_loss": 0.71743429, + "learning_rate": 1.1346787719460071e-07, + "loss": 0.73884851, + "num_input_tokens_seen": 161019405, + "step": 7447, + "time_per_iteration": 3.568350315093994 + }, + { + "auxiliary_loss_clip": 0.01113775, + "auxiliary_loss_mlp": 0.01022635, + "balance_loss_clip": 0.97078454, + "balance_loss_mlp": 1.01561928, + "epoch": 0.8955690494799495, + "flos": 18257883120000.0, + "grad_norm": 1.7469571462254523, + "language_loss": 0.72345817, + "learning_rate": 1.1320937221867732e-07, + "loss": 0.74482226, + "num_input_tokens_seen": 161036985, + "step": 7448, + "time_per_iteration": 2.6848244667053223 + }, + { + "auxiliary_loss_clip": 0.01114735, + "auxiliary_loss_mlp": 0.01025353, + "balance_loss_clip": 0.96858233, + "balance_loss_mlp": 1.01841784, + "epoch": 0.8956892923705886, + "flos": 25447486498560.0, + "grad_norm": 1.769067787574771, + "language_loss": 0.79381359, + "learning_rate": 1.1295115346739192e-07, + "loss": 0.8152144, + "num_input_tokens_seen": 161056985, + "step": 7449, + "time_per_iteration": 3.6182057857513428 + }, + { + "auxiliary_loss_clip": 0.01121084, + "auxiliary_loss_mlp": 0.0102266, + "balance_loss_clip": 0.97191656, + "balance_loss_mlp": 1.01557922, + "epoch": 0.8958095352612276, + "flos": 52661883939840.0, + "grad_norm": 2.1911406502369046, + "language_loss": 0.72562182, + "learning_rate": 1.1269322097991629e-07, + "loss": 0.74705935, + "num_input_tokens_seen": 161080270, + "step": 7450, + "time_per_iteration": 3.8991281986236572 + }, + { + "auxiliary_loss_clip": 0.01145983, + "auxiliary_loss_mlp": 0.01026789, + "balance_loss_clip": 1.01167798, + "balance_loss_mlp": 1.01965427, + "epoch": 0.8959297781518668, + "flos": 23186335392000.0, + "grad_norm": 2.356407702989389, + "language_loss": 0.67728353, + "learning_rate": 1.1243557479537846e-07, + "loss": 0.69901121, + "num_input_tokens_seen": 161100160, + "step": 7451, + "time_per_iteration": 2.7101573944091797 + }, + { + "auxiliary_loss_clip": 0.01161907, + "auxiliary_loss_mlp": 0.01025063, + "balance_loss_clip": 1.04535186, + "balance_loss_mlp": 1.01769626, + "epoch": 0.8960500210425059, + "flos": 20334309557760.0, + "grad_norm": 2.0192124103646525, + "language_loss": 0.68901843, + "learning_rate": 1.121782149528634e-07, + "loss": 0.71088815, + "num_input_tokens_seen": 161117260, + "step": 7452, + "time_per_iteration": 2.690133571624756 + }, + { + "auxiliary_loss_clip": 0.0112558, + "auxiliary_loss_mlp": 0.01022457, + "balance_loss_clip": 0.97695094, + "balance_loss_mlp": 1.0151372, + "epoch": 0.8961702639331449, + "flos": 19901694153600.0, + "grad_norm": 2.2572815951150775, + "language_loss": 0.78961289, + "learning_rate": 1.1192114149141208e-07, + "loss": 0.81109321, + "num_input_tokens_seen": 161136895, + "step": 7453, + "time_per_iteration": 2.7524333000183105 + }, + { + "auxiliary_loss_clip": 0.01119469, + "auxiliary_loss_mlp": 0.01024225, + "balance_loss_clip": 0.96962607, + "balance_loss_mlp": 1.01715589, + "epoch": 0.8962905068237841, + "flos": 12896348567040.0, + "grad_norm": 2.853064344952377, + "language_loss": 0.64883995, + "learning_rate": 1.1166435445002197e-07, + "loss": 0.67027688, + "num_input_tokens_seen": 161154565, + "step": 7454, + "time_per_iteration": 2.649488925933838 + }, + { + "auxiliary_loss_clip": 0.01143055, + "auxiliary_loss_mlp": 0.01026823, + "balance_loss_clip": 1.01119792, + "balance_loss_mlp": 1.01941729, + "epoch": 0.8964107497144231, + "flos": 23440331439360.0, + "grad_norm": 2.2132106229279063, + "language_loss": 0.68266308, + "learning_rate": 1.1140785386764818e-07, + "loss": 0.70436186, + "num_input_tokens_seen": 161173265, + "step": 7455, + "time_per_iteration": 2.654636859893799 + }, + { + "auxiliary_loss_clip": 0.01136772, + "auxiliary_loss_mlp": 0.01028887, + "balance_loss_clip": 1.00915205, + "balance_loss_mlp": 1.02194285, + "epoch": 0.8965309926050622, + "flos": 19500176949120.0, + "grad_norm": 1.9858422563832951, + "language_loss": 0.69730455, + "learning_rate": 1.1115163978320153e-07, + "loss": 0.71896118, + "num_input_tokens_seen": 161191995, + "step": 7456, + "time_per_iteration": 3.518434762954712 + }, + { + "auxiliary_loss_clip": 0.01145842, + "auxiliary_loss_mlp": 0.00509306, + "balance_loss_clip": 1.00938046, + "balance_loss_mlp": 0.0, + "epoch": 0.8966512354957014, + "flos": 28658008022400.0, + "grad_norm": 1.9044203432412399, + "language_loss": 0.82872748, + "learning_rate": 1.1089571223554917e-07, + "loss": 0.84527898, + "num_input_tokens_seen": 161212880, + "step": 7457, + "time_per_iteration": 2.728400707244873 + }, + { + "auxiliary_loss_clip": 0.01140013, + "auxiliary_loss_mlp": 0.010229, + "balance_loss_clip": 1.0067389, + "balance_loss_mlp": 1.01633132, + "epoch": 0.8967714783863404, + "flos": 23370916406400.0, + "grad_norm": 1.8924965307975317, + "language_loss": 0.8561126, + "learning_rate": 1.1064007126351537e-07, + "loss": 0.87774163, + "num_input_tokens_seen": 161233595, + "step": 7458, + "time_per_iteration": 2.6530730724334717 + }, + { + "auxiliary_loss_clip": 0.01113064, + "auxiliary_loss_mlp": 0.01022507, + "balance_loss_clip": 0.97247529, + "balance_loss_mlp": 1.01518798, + "epoch": 0.8968917212769795, + "flos": 24535175938560.0, + "grad_norm": 2.289071919155027, + "language_loss": 0.76291788, + "learning_rate": 1.1038471690588003e-07, + "loss": 0.78427368, + "num_input_tokens_seen": 161252740, + "step": 7459, + "time_per_iteration": 2.7124183177948 + }, + { + "auxiliary_loss_clip": 0.01072159, + "auxiliary_loss_mlp": 0.0102415, + "balance_loss_clip": 0.89700305, + "balance_loss_mlp": 1.0173552, + "epoch": 0.8970119641676186, + "flos": 23475416048640.0, + "grad_norm": 2.259981114650945, + "language_loss": 0.79781473, + "learning_rate": 1.1012964920138145e-07, + "loss": 0.8187778, + "num_input_tokens_seen": 161272325, + "step": 7460, + "time_per_iteration": 2.8368492126464844 + }, + { + "auxiliary_loss_clip": 0.01110402, + "auxiliary_loss_mlp": 0.01023788, + "balance_loss_clip": 0.96636599, + "balance_loss_mlp": 1.01643586, + "epoch": 0.8971322070582577, + "flos": 24538192680960.0, + "grad_norm": 2.6878195178717856, + "language_loss": 0.76013458, + "learning_rate": 1.0987486818871205e-07, + "loss": 0.7814765, + "num_input_tokens_seen": 161295915, + "step": 7461, + "time_per_iteration": 2.7244834899902344 + }, + { + "auxiliary_loss_clip": 0.01140297, + "auxiliary_loss_mlp": 0.0050935, + "balance_loss_clip": 1.00998056, + "balance_loss_mlp": 0.0, + "epoch": 0.8972524499488967, + "flos": 21797454159360.0, + "grad_norm": 2.1581252366171135, + "language_loss": 0.73548007, + "learning_rate": 1.0962037390652245e-07, + "loss": 0.75197649, + "num_input_tokens_seen": 161314935, + "step": 7462, + "time_per_iteration": 2.6555755138397217 + }, + { + "auxiliary_loss_clip": 0.0111858, + "auxiliary_loss_mlp": 0.01024309, + "balance_loss_clip": 0.97340125, + "balance_loss_mlp": 1.01738548, + "epoch": 0.8973726928395359, + "flos": 21726243446400.0, + "grad_norm": 1.7370460269832908, + "language_loss": 0.7194097, + "learning_rate": 1.0936616639341911e-07, + "loss": 0.74083853, + "num_input_tokens_seen": 161335225, + "step": 7463, + "time_per_iteration": 2.6919875144958496 + }, + { + "auxiliary_loss_clip": 0.01037944, + "auxiliary_loss_mlp": 0.0100916, + "balance_loss_clip": 0.97533435, + "balance_loss_mlp": 1.00807559, + "epoch": 0.897492935730175, + "flos": 53837100097920.0, + "grad_norm": 0.7328938464351001, + "language_loss": 0.54747266, + "learning_rate": 1.0911224568796473e-07, + "loss": 0.56794369, + "num_input_tokens_seen": 161393420, + "step": 7464, + "time_per_iteration": 3.2425220012664795 + }, + { + "auxiliary_loss_clip": 0.01141227, + "auxiliary_loss_mlp": 0.01025932, + "balance_loss_clip": 1.01230502, + "balance_loss_mlp": 1.01903534, + "epoch": 0.897613178620814, + "flos": 18290346036480.0, + "grad_norm": 1.9044324263441363, + "language_loss": 0.71205533, + "learning_rate": 1.0885861182867984e-07, + "loss": 0.73372698, + "num_input_tokens_seen": 161411525, + "step": 7465, + "time_per_iteration": 2.656677722930908 + }, + { + "auxiliary_loss_clip": 0.01122431, + "auxiliary_loss_mlp": 0.01026635, + "balance_loss_clip": 0.97285557, + "balance_loss_mlp": 1.01878548, + "epoch": 0.8977334215114532, + "flos": 32993718059520.0, + "grad_norm": 1.6852544531215867, + "language_loss": 0.70895046, + "learning_rate": 1.0860526485403942e-07, + "loss": 0.73044109, + "num_input_tokens_seen": 161432800, + "step": 7466, + "time_per_iteration": 2.765517473220825 + }, + { + "auxiliary_loss_clip": 0.01164311, + "auxiliary_loss_mlp": 0.0102537, + "balance_loss_clip": 1.04847968, + "balance_loss_mlp": 1.01819944, + "epoch": 0.8978536644020922, + "flos": 15195636938880.0, + "grad_norm": 1.5726759204187666, + "language_loss": 0.76921952, + "learning_rate": 1.0835220480247675e-07, + "loss": 0.79111636, + "num_input_tokens_seen": 161451295, + "step": 7467, + "time_per_iteration": 2.5619986057281494 + }, + { + "auxiliary_loss_clip": 0.01117438, + "auxiliary_loss_mlp": 0.01022049, + "balance_loss_clip": 0.9740181, + "balance_loss_mlp": 1.01512551, + "epoch": 0.8979739072927313, + "flos": 18004389863040.0, + "grad_norm": 2.154946197030815, + "language_loss": 0.8351295, + "learning_rate": 1.0809943171238067e-07, + "loss": 0.85652435, + "num_input_tokens_seen": 161469220, + "step": 7468, + "time_per_iteration": 2.675175428390503 + }, + { + "auxiliary_loss_clip": 0.01123272, + "auxiliary_loss_mlp": 0.0102706, + "balance_loss_clip": 0.97148705, + "balance_loss_mlp": 1.01937699, + "epoch": 0.8980941501833704, + "flos": 22271546793600.0, + "grad_norm": 2.2034050226507835, + "language_loss": 0.6297735, + "learning_rate": 1.078469456220965e-07, + "loss": 0.65127683, + "num_input_tokens_seen": 161489375, + "step": 7469, + "time_per_iteration": 2.717543363571167 + }, + { + "auxiliary_loss_clip": 0.01141602, + "auxiliary_loss_mlp": 0.01021724, + "balance_loss_clip": 1.00843716, + "balance_loss_mlp": 1.01486921, + "epoch": 0.8982143930740095, + "flos": 37560729726720.0, + "grad_norm": 2.3137714678345986, + "language_loss": 0.69677615, + "learning_rate": 1.0759474656992606e-07, + "loss": 0.71840942, + "num_input_tokens_seen": 161512145, + "step": 7470, + "time_per_iteration": 2.723224401473999 + }, + { + "auxiliary_loss_clip": 0.01123004, + "auxiliary_loss_mlp": 0.01026107, + "balance_loss_clip": 0.97065604, + "balance_loss_mlp": 1.01876366, + "epoch": 0.8983346359646486, + "flos": 18076893465600.0, + "grad_norm": 3.8166673711865227, + "language_loss": 0.77666122, + "learning_rate": 1.0734283459412785e-07, + "loss": 0.79815233, + "num_input_tokens_seen": 161528995, + "step": 7471, + "time_per_iteration": 2.701493501663208 + }, + { + "auxiliary_loss_clip": 0.01076637, + "auxiliary_loss_mlp": 0.01023463, + "balance_loss_clip": 0.89596915, + "balance_loss_mlp": 1.01615572, + "epoch": 0.8984548788552876, + "flos": 20558895344640.0, + "grad_norm": 1.9382145585518278, + "language_loss": 0.80461097, + "learning_rate": 1.0709120973291707e-07, + "loss": 0.82561195, + "num_input_tokens_seen": 161548775, + "step": 7472, + "time_per_iteration": 2.700972318649292 + }, + { + "auxiliary_loss_clip": 0.01166357, + "auxiliary_loss_mlp": 0.0103088, + "balance_loss_clip": 1.04841232, + "balance_loss_mlp": 1.02353692, + "epoch": 0.8985751217459268, + "flos": 17785442511360.0, + "grad_norm": 2.1975745766089223, + "language_loss": 0.77814102, + "learning_rate": 1.0683987202446475e-07, + "loss": 0.80011338, + "num_input_tokens_seen": 161566960, + "step": 7473, + "time_per_iteration": 3.557159662246704 + }, + { + "auxiliary_loss_clip": 0.01145239, + "auxiliary_loss_mlp": 0.01022175, + "balance_loss_clip": 1.01003659, + "balance_loss_mlp": 1.01500511, + "epoch": 0.8986953646365659, + "flos": 21617003208960.0, + "grad_norm": 2.4465690337130788, + "language_loss": 0.69905996, + "learning_rate": 1.0658882150689862e-07, + "loss": 0.72073412, + "num_input_tokens_seen": 161585820, + "step": 7474, + "time_per_iteration": 3.6197030544281006 + }, + { + "auxiliary_loss_clip": 0.01096643, + "auxiliary_loss_mlp": 0.01025094, + "balance_loss_clip": 0.93204314, + "balance_loss_mlp": 1.01770341, + "epoch": 0.8988156075272049, + "flos": 14027355083520.0, + "grad_norm": 2.756207520270651, + "language_loss": 0.78924382, + "learning_rate": 1.0633805821830288e-07, + "loss": 0.81046122, + "num_input_tokens_seen": 161602505, + "step": 7475, + "time_per_iteration": 2.721013307571411 + }, + { + "auxiliary_loss_clip": 0.01119323, + "auxiliary_loss_mlp": 0.01017864, + "balance_loss_clip": 0.97294652, + "balance_loss_mlp": 1.01099718, + "epoch": 0.8989358504178441, + "flos": 29059202004480.0, + "grad_norm": 6.368273907416863, + "language_loss": 0.82779497, + "learning_rate": 1.0608758219671753e-07, + "loss": 0.84916687, + "num_input_tokens_seen": 161621545, + "step": 7476, + "time_per_iteration": 3.764873504638672 + }, + { + "auxiliary_loss_clip": 0.01121226, + "auxiliary_loss_mlp": 0.01032912, + "balance_loss_clip": 0.97134864, + "balance_loss_mlp": 1.02613449, + "epoch": 0.8990560933084831, + "flos": 20230420446720.0, + "grad_norm": 1.7832409689581794, + "language_loss": 0.7100299, + "learning_rate": 1.0583739348014065e-07, + "loss": 0.73157126, + "num_input_tokens_seen": 161642630, + "step": 7477, + "time_per_iteration": 2.7008769512176514 + }, + { + "auxiliary_loss_clip": 0.01166982, + "auxiliary_loss_mlp": 0.0103019, + "balance_loss_clip": 1.05040932, + "balance_loss_mlp": 1.02285862, + "epoch": 0.8991763361991222, + "flos": 25520672459520.0, + "grad_norm": 1.9874397480679435, + "language_loss": 0.84492338, + "learning_rate": 1.0558749210652518e-07, + "loss": 0.86689508, + "num_input_tokens_seen": 161662560, + "step": 7478, + "time_per_iteration": 2.6428332328796387 + }, + { + "auxiliary_loss_clip": 0.0109871, + "auxiliary_loss_mlp": 0.0103049, + "balance_loss_clip": 0.93509346, + "balance_loss_mlp": 1.02298546, + "epoch": 0.8992965790897613, + "flos": 25119191168640.0, + "grad_norm": 1.7145769533203556, + "language_loss": 0.85452437, + "learning_rate": 1.053378781137808e-07, + "loss": 0.87581635, + "num_input_tokens_seen": 161683480, + "step": 7479, + "time_per_iteration": 2.828705310821533 + }, + { + "auxiliary_loss_clip": 0.01121141, + "auxiliary_loss_mlp": 0.01026346, + "balance_loss_clip": 0.9712314, + "balance_loss_mlp": 1.01876414, + "epoch": 0.8994168219804004, + "flos": 16070815814400.0, + "grad_norm": 1.7745960343584433, + "language_loss": 0.77759069, + "learning_rate": 1.0508855153977392e-07, + "loss": 0.79906559, + "num_input_tokens_seen": 161699945, + "step": 7480, + "time_per_iteration": 2.663206100463867 + }, + { + "auxiliary_loss_clip": 0.01137946, + "auxiliary_loss_mlp": 0.01022521, + "balance_loss_clip": 1.00540674, + "balance_loss_mlp": 1.0153625, + "epoch": 0.8995370648710395, + "flos": 24825764966400.0, + "grad_norm": 2.4717086550507092, + "language_loss": 0.66791403, + "learning_rate": 1.0483951242232669e-07, + "loss": 0.68951869, + "num_input_tokens_seen": 161720420, + "step": 7481, + "time_per_iteration": 2.6632750034332275 + }, + { + "auxiliary_loss_clip": 0.01060011, + "auxiliary_loss_mlp": 0.01002661, + "balance_loss_clip": 1.00918794, + "balance_loss_mlp": 1.0017314, + "epoch": 0.8996573077616786, + "flos": 63116238378240.0, + "grad_norm": 0.9803590850941422, + "language_loss": 0.57706684, + "learning_rate": 1.0459076079921936e-07, + "loss": 0.59769356, + "num_input_tokens_seen": 161773080, + "step": 7482, + "time_per_iteration": 4.1070640087127686 + }, + { + "auxiliary_loss_clip": 0.01112781, + "auxiliary_loss_mlp": 0.01028403, + "balance_loss_clip": 0.97221577, + "balance_loss_mlp": 1.02153325, + "epoch": 0.8997775506523177, + "flos": 18219674027520.0, + "grad_norm": 4.282303189234512, + "language_loss": 0.85490572, + "learning_rate": 1.0434229670818618e-07, + "loss": 0.8763175, + "num_input_tokens_seen": 161789755, + "step": 7483, + "time_per_iteration": 2.6323888301849365 + }, + { + "auxiliary_loss_clip": 0.01110771, + "auxiliary_loss_mlp": 0.01023962, + "balance_loss_clip": 0.97097623, + "balance_loss_mlp": 1.01708102, + "epoch": 0.8998977935429567, + "flos": 24166768095360.0, + "grad_norm": 1.656016752932849, + "language_loss": 0.80080062, + "learning_rate": 1.0409412018691944e-07, + "loss": 0.82214785, + "num_input_tokens_seen": 161810220, + "step": 7484, + "time_per_iteration": 2.679511547088623 + }, + { + "auxiliary_loss_clip": 0.01116607, + "auxiliary_loss_mlp": 0.01030075, + "balance_loss_clip": 0.97333205, + "balance_loss_mlp": 1.02305341, + "epoch": 0.9000180364335959, + "flos": 20773030273920.0, + "grad_norm": 1.9427653162019067, + "language_loss": 0.75254178, + "learning_rate": 1.0384623127306724e-07, + "loss": 0.77400863, + "num_input_tokens_seen": 161827565, + "step": 7485, + "time_per_iteration": 2.717555522918701 + }, + { + "auxiliary_loss_clip": 0.01093229, + "auxiliary_loss_mlp": 0.01027779, + "balance_loss_clip": 0.93281746, + "balance_loss_mlp": 1.02068257, + "epoch": 0.900138279324235, + "flos": 19205745166080.0, + "grad_norm": 1.9190265020745485, + "language_loss": 0.79487193, + "learning_rate": 1.0359863000423397e-07, + "loss": 0.816082, + "num_input_tokens_seen": 161845700, + "step": 7486, + "time_per_iteration": 2.705264091491699 + }, + { + "auxiliary_loss_clip": 0.01164054, + "auxiliary_loss_mlp": 0.01024641, + "balance_loss_clip": 1.04711175, + "balance_loss_mlp": 1.0174911, + "epoch": 0.900258522214874, + "flos": 28731158069760.0, + "grad_norm": 1.6002236530664147, + "language_loss": 0.71789664, + "learning_rate": 1.0335131641798112e-07, + "loss": 0.73978359, + "num_input_tokens_seen": 161867660, + "step": 7487, + "time_per_iteration": 2.70163631439209 + }, + { + "auxiliary_loss_clip": 0.01021188, + "auxiliary_loss_mlp": 0.00998865, + "balance_loss_clip": 0.93537331, + "balance_loss_mlp": 0.99792939, + "epoch": 0.9003787651055132, + "flos": 58280685655680.0, + "grad_norm": 0.8078821400947188, + "language_loss": 0.55603385, + "learning_rate": 1.0310429055182512e-07, + "loss": 0.5762344, + "num_input_tokens_seen": 161921980, + "step": 7488, + "time_per_iteration": 3.102503538131714 + }, + { + "auxiliary_loss_clip": 0.01096078, + "auxiliary_loss_mlp": 0.01027991, + "balance_loss_clip": 0.93247771, + "balance_loss_mlp": 1.02053416, + "epoch": 0.9004990079961522, + "flos": 25556475340800.0, + "grad_norm": 1.607370217836155, + "language_loss": 0.74132669, + "learning_rate": 1.0285755244324024e-07, + "loss": 0.76256734, + "num_input_tokens_seen": 161942725, + "step": 7489, + "time_per_iteration": 2.7691640853881836 + }, + { + "auxiliary_loss_clip": 0.01121297, + "auxiliary_loss_mlp": 0.00509009, + "balance_loss_clip": 0.97062582, + "balance_loss_mlp": 0.0, + "epoch": 0.9006192508867913, + "flos": 23335185352320.0, + "grad_norm": 1.6999830255979531, + "language_loss": 0.68306899, + "learning_rate": 1.0261110212965629e-07, + "loss": 0.69937205, + "num_input_tokens_seen": 161964520, + "step": 7490, + "time_per_iteration": 2.8295462131500244 + }, + { + "auxiliary_loss_clip": 0.01119275, + "auxiliary_loss_mlp": 0.01021876, + "balance_loss_clip": 0.97161138, + "balance_loss_mlp": 1.01499796, + "epoch": 0.9007394937774305, + "flos": 18040300485120.0, + "grad_norm": 2.374032097002151, + "language_loss": 0.79027945, + "learning_rate": 1.023649396484596e-07, + "loss": 0.81169093, + "num_input_tokens_seen": 161983575, + "step": 7491, + "time_per_iteration": 2.673304557800293 + }, + { + "auxiliary_loss_clip": 0.01165766, + "auxiliary_loss_mlp": 0.01023339, + "balance_loss_clip": 1.04959559, + "balance_loss_mlp": 1.01604319, + "epoch": 0.9008597366680695, + "flos": 43068456633600.0, + "grad_norm": 2.2462430876244315, + "language_loss": 0.67594242, + "learning_rate": 1.0211906503699275e-07, + "loss": 0.69783348, + "num_input_tokens_seen": 162006550, + "step": 7492, + "time_per_iteration": 2.748488187789917 + }, + { + "auxiliary_loss_clip": 0.01145193, + "auxiliary_loss_mlp": 0.01027393, + "balance_loss_clip": 1.01319075, + "balance_loss_mlp": 1.02035618, + "epoch": 0.9009799795587086, + "flos": 14939055112320.0, + "grad_norm": 2.3990357663789332, + "language_loss": 0.82659745, + "learning_rate": 1.0187347833255455e-07, + "loss": 0.84832335, + "num_input_tokens_seen": 162022455, + "step": 7493, + "time_per_iteration": 2.5854861736297607 + }, + { + "auxiliary_loss_clip": 0.01162289, + "auxiliary_loss_mlp": 0.01026239, + "balance_loss_clip": 1.0480907, + "balance_loss_mlp": 1.01927114, + "epoch": 0.9011002224493477, + "flos": 21579584215680.0, + "grad_norm": 2.3699219009486487, + "language_loss": 0.79151493, + "learning_rate": 1.0162817957240056e-07, + "loss": 0.81340015, + "num_input_tokens_seen": 162042350, + "step": 7494, + "time_per_iteration": 2.598102569580078 + }, + { + "auxiliary_loss_clip": 0.01042171, + "auxiliary_loss_mlp": 0.01001571, + "balance_loss_clip": 0.97318947, + "balance_loss_mlp": 1.00064147, + "epoch": 0.9012204653399868, + "flos": 71166367883520.0, + "grad_norm": 0.8888074025089384, + "language_loss": 0.63004839, + "learning_rate": 1.0138316879374253e-07, + "loss": 0.65048581, + "num_input_tokens_seen": 162111640, + "step": 7495, + "time_per_iteration": 3.3178181648254395 + }, + { + "auxiliary_loss_clip": 0.01121477, + "auxiliary_loss_mlp": 0.01025074, + "balance_loss_clip": 0.97455573, + "balance_loss_mlp": 1.01808858, + "epoch": 0.9013407082306258, + "flos": 15594963413760.0, + "grad_norm": 2.295109258785078, + "language_loss": 0.74746335, + "learning_rate": 1.0113844603374833e-07, + "loss": 0.76892883, + "num_input_tokens_seen": 162128165, + "step": 7496, + "time_per_iteration": 2.6960482597351074 + }, + { + "auxiliary_loss_clip": 0.01117069, + "auxiliary_loss_mlp": 0.01025613, + "balance_loss_clip": 0.9679696, + "balance_loss_mlp": 1.01795959, + "epoch": 0.901460951121265, + "flos": 15049157276160.0, + "grad_norm": 2.033717035957816, + "language_loss": 0.71925044, + "learning_rate": 1.0089401132954178e-07, + "loss": 0.74067724, + "num_input_tokens_seen": 162146145, + "step": 7497, + "time_per_iteration": 2.6485917568206787 + }, + { + "auxiliary_loss_clip": 0.01116386, + "auxiliary_loss_mlp": 0.0102468, + "balance_loss_clip": 0.9709838, + "balance_loss_mlp": 1.01781654, + "epoch": 0.9015811940119041, + "flos": 22236857233920.0, + "grad_norm": 1.5610677988075137, + "language_loss": 0.72370279, + "learning_rate": 1.006498647182037e-07, + "loss": 0.74511349, + "num_input_tokens_seen": 162164800, + "step": 7498, + "time_per_iteration": 3.5186586380004883 + }, + { + "auxiliary_loss_clip": 0.01043865, + "auxiliary_loss_mlp": 0.01024275, + "balance_loss_clip": 0.853329, + "balance_loss_mlp": 1.01714969, + "epoch": 0.9017014369025431, + "flos": 24973824827520.0, + "grad_norm": 2.1148867909194835, + "language_loss": 0.71282238, + "learning_rate": 1.004060062367713e-07, + "loss": 0.73350382, + "num_input_tokens_seen": 162185895, + "step": 7499, + "time_per_iteration": 2.9301986694335938 + }, + { + "auxiliary_loss_clip": 0.01142948, + "auxiliary_loss_mlp": 0.01029466, + "balance_loss_clip": 1.01051593, + "balance_loss_mlp": 1.02273047, + "epoch": 0.9018216797931822, + "flos": 18114168804480.0, + "grad_norm": 1.5948890892074088, + "language_loss": 0.69367892, + "learning_rate": 1.0016243592223728e-07, + "loss": 0.71540308, + "num_input_tokens_seen": 162206295, + "step": 7500, + "time_per_iteration": 2.6382741928100586 + }, + { + "auxiliary_loss_clip": 0.01044779, + "auxiliary_loss_mlp": 0.01026801, + "balance_loss_clip": 0.85617977, + "balance_loss_mlp": 1.01919532, + "epoch": 0.9019419226838213, + "flos": 37268452759680.0, + "grad_norm": 2.077242011019684, + "language_loss": 0.6555804, + "learning_rate": 9.991915381155114e-08, + "loss": 0.67629617, + "num_input_tokens_seen": 162229275, + "step": 7501, + "time_per_iteration": 3.9327235221862793 + }, + { + "auxiliary_loss_clip": 0.01146796, + "auxiliary_loss_mlp": 0.01028571, + "balance_loss_clip": 1.01167274, + "balance_loss_mlp": 1.02112353, + "epoch": 0.9020621655744604, + "flos": 23441121538560.0, + "grad_norm": 1.9986815159797404, + "language_loss": 0.74998724, + "learning_rate": 9.967615994161871e-08, + "loss": 0.77174091, + "num_input_tokens_seen": 162248935, + "step": 7502, + "time_per_iteration": 3.6332364082336426 + }, + { + "auxiliary_loss_clip": 0.01161791, + "auxiliary_loss_mlp": 0.010261, + "balance_loss_clip": 1.04719496, + "balance_loss_mlp": 1.01898324, + "epoch": 0.9021824084650995, + "flos": 22857465444480.0, + "grad_norm": 2.1148153985828846, + "language_loss": 0.77949584, + "learning_rate": 9.943345434930161e-08, + "loss": 0.80137479, + "num_input_tokens_seen": 162269185, + "step": 7503, + "time_per_iteration": 2.6554648876190186 + }, + { + "auxiliary_loss_clip": 0.01094786, + "auxiliary_loss_mlp": 0.01030018, + "balance_loss_clip": 0.93492955, + "balance_loss_mlp": 1.02248371, + "epoch": 0.9023026513557386, + "flos": 22127581082880.0, + "grad_norm": 3.619442233967649, + "language_loss": 0.69218373, + "learning_rate": 9.919103707141885e-08, + "loss": 0.71343178, + "num_input_tokens_seen": 162288065, + "step": 7504, + "time_per_iteration": 2.7898945808410645 + }, + { + "auxiliary_loss_clip": 0.01142077, + "auxiliary_loss_mlp": 0.01026475, + "balance_loss_clip": 1.01034498, + "balance_loss_mlp": 1.01877439, + "epoch": 0.9024228942463777, + "flos": 24199087357440.0, + "grad_norm": 1.9888946040241877, + "language_loss": 0.76408768, + "learning_rate": 9.89489081447441e-08, + "loss": 0.78577316, + "num_input_tokens_seen": 162305265, + "step": 7505, + "time_per_iteration": 2.7056660652160645 + }, + { + "auxiliary_loss_clip": 0.0111655, + "auxiliary_loss_mlp": 0.01026873, + "balance_loss_clip": 0.96897137, + "balance_loss_mlp": 1.01912427, + "epoch": 0.9025431371370167, + "flos": 25008262992000.0, + "grad_norm": 2.248961323011434, + "language_loss": 0.82799542, + "learning_rate": 9.870706760600844e-08, + "loss": 0.84942967, + "num_input_tokens_seen": 162325215, + "step": 7506, + "time_per_iteration": 2.76440167427063 + }, + { + "auxiliary_loss_clip": 0.01077882, + "auxiliary_loss_mlp": 0.01027404, + "balance_loss_clip": 0.89766365, + "balance_loss_mlp": 1.02036488, + "epoch": 0.9026633800276559, + "flos": 18952862440320.0, + "grad_norm": 1.813854653415032, + "language_loss": 0.73063707, + "learning_rate": 9.846551549189918e-08, + "loss": 0.75168991, + "num_input_tokens_seen": 162344820, + "step": 7507, + "time_per_iteration": 2.720703363418579 + }, + { + "auxiliary_loss_clip": 0.01117001, + "auxiliary_loss_mlp": 0.01029235, + "balance_loss_clip": 0.97228742, + "balance_loss_mlp": 1.02220488, + "epoch": 0.902783622918295, + "flos": 32416059536640.0, + "grad_norm": 1.9317998171079809, + "language_loss": 0.68533707, + "learning_rate": 9.822425183905902e-08, + "loss": 0.70679939, + "num_input_tokens_seen": 162365345, + "step": 7508, + "time_per_iteration": 2.7488596439361572 + }, + { + "auxiliary_loss_clip": 0.01003847, + "auxiliary_loss_mlp": 0.01000802, + "balance_loss_clip": 0.89953899, + "balance_loss_mlp": 0.99990219, + "epoch": 0.902903865808934, + "flos": 63717453244800.0, + "grad_norm": 0.9247261973571251, + "language_loss": 0.75142765, + "learning_rate": 9.798327668408823e-08, + "loss": 0.77147418, + "num_input_tokens_seen": 162426980, + "step": 7509, + "time_per_iteration": 4.234669923782349 + }, + { + "auxiliary_loss_clip": 0.0116876, + "auxiliary_loss_mlp": 0.01024221, + "balance_loss_clip": 1.04895473, + "balance_loss_mlp": 1.01654363, + "epoch": 0.9030241086995732, + "flos": 23804034600960.0, + "grad_norm": 2.0888171501293673, + "language_loss": 0.68679059, + "learning_rate": 9.774259006354158e-08, + "loss": 0.70872033, + "num_input_tokens_seen": 162447050, + "step": 7510, + "time_per_iteration": 2.680375337600708 + }, + { + "auxiliary_loss_clip": 0.01121502, + "auxiliary_loss_mlp": 0.01024426, + "balance_loss_clip": 0.97108185, + "balance_loss_mlp": 1.01726723, + "epoch": 0.9031443515902122, + "flos": 26395887248640.0, + "grad_norm": 3.2611439470928767, + "language_loss": 0.76165414, + "learning_rate": 9.750219201393184e-08, + "loss": 0.78311336, + "num_input_tokens_seen": 162467015, + "step": 7511, + "time_per_iteration": 2.7116992473602295 + }, + { + "auxiliary_loss_clip": 0.01138325, + "auxiliary_loss_mlp": 0.01025506, + "balance_loss_clip": 1.00829101, + "balance_loss_mlp": 1.01834154, + "epoch": 0.9032645944808513, + "flos": 24939350749440.0, + "grad_norm": 1.8473574910601342, + "language_loss": 0.77678639, + "learning_rate": 9.726208257172697e-08, + "loss": 0.79842472, + "num_input_tokens_seen": 162488710, + "step": 7512, + "time_per_iteration": 2.6723527908325195 + }, + { + "auxiliary_loss_clip": 0.01161468, + "auxiliary_loss_mlp": 0.0102381, + "balance_loss_clip": 1.04772711, + "balance_loss_mlp": 1.01731932, + "epoch": 0.9033848373714904, + "flos": 21178821196800.0, + "grad_norm": 1.8998843806642365, + "language_loss": 0.74560726, + "learning_rate": 9.702226177335115e-08, + "loss": 0.76745999, + "num_input_tokens_seen": 162507205, + "step": 7513, + "time_per_iteration": 2.576032876968384 + }, + { + "auxiliary_loss_clip": 0.01119254, + "auxiliary_loss_mlp": 0.01027561, + "balance_loss_clip": 0.97383547, + "balance_loss_mlp": 1.02053988, + "epoch": 0.9035050802621295, + "flos": 26286359702400.0, + "grad_norm": 1.476657190981271, + "language_loss": 0.72364062, + "learning_rate": 9.67827296551853e-08, + "loss": 0.74510872, + "num_input_tokens_seen": 162528490, + "step": 7514, + "time_per_iteration": 2.76417875289917 + }, + { + "auxiliary_loss_clip": 0.01109071, + "auxiliary_loss_mlp": 0.00508568, + "balance_loss_clip": 0.9681952, + "balance_loss_mlp": 0.0, + "epoch": 0.9036253231527686, + "flos": 24204546224640.0, + "grad_norm": 2.071692461409888, + "language_loss": 0.68776417, + "learning_rate": 9.65434862535659e-08, + "loss": 0.70394057, + "num_input_tokens_seen": 162547860, + "step": 7515, + "time_per_iteration": 2.810420274734497 + }, + { + "auxiliary_loss_clip": 0.01120755, + "auxiliary_loss_mlp": 0.01020219, + "balance_loss_clip": 0.97119009, + "balance_loss_mlp": 1.01300716, + "epoch": 0.9037455660434077, + "flos": 18072655660800.0, + "grad_norm": 126.09411304717806, + "language_loss": 0.65026557, + "learning_rate": 9.630453160478635e-08, + "loss": 0.67167538, + "num_input_tokens_seen": 162563215, + "step": 7516, + "time_per_iteration": 2.692500352859497 + }, + { + "auxiliary_loss_clip": 0.01069998, + "auxiliary_loss_mlp": 0.01022263, + "balance_loss_clip": 0.89328361, + "balance_loss_mlp": 1.01580811, + "epoch": 0.9038658089340468, + "flos": 24060795995520.0, + "grad_norm": 1.7657624213497622, + "language_loss": 0.82181704, + "learning_rate": 9.60658657450959e-08, + "loss": 0.84273958, + "num_input_tokens_seen": 162583515, + "step": 7517, + "time_per_iteration": 2.94446063041687 + }, + { + "auxiliary_loss_clip": 0.01107233, + "auxiliary_loss_mlp": 0.01020433, + "balance_loss_clip": 0.96837145, + "balance_loss_mlp": 1.01327443, + "epoch": 0.9039860518246858, + "flos": 21834298535040.0, + "grad_norm": 1.570282727928761, + "language_loss": 0.79395056, + "learning_rate": 9.582748871069979e-08, + "loss": 0.81522715, + "num_input_tokens_seen": 162602955, + "step": 7518, + "time_per_iteration": 2.711500644683838 + }, + { + "auxiliary_loss_clip": 0.01117954, + "auxiliary_loss_mlp": 0.00508922, + "balance_loss_clip": 0.96872377, + "balance_loss_mlp": 0.0, + "epoch": 0.904106294715325, + "flos": 26614870513920.0, + "grad_norm": 3.976856969876448, + "language_loss": 0.83475173, + "learning_rate": 9.558940053775954e-08, + "loss": 0.85102046, + "num_input_tokens_seen": 162621595, + "step": 7519, + "time_per_iteration": 2.7515368461608887 + }, + { + "auxiliary_loss_clip": 0.01142261, + "auxiliary_loss_mlp": 0.01024523, + "balance_loss_clip": 1.01115942, + "balance_loss_mlp": 1.01754045, + "epoch": 0.904226537605964, + "flos": 17785693906560.0, + "grad_norm": 2.947461065115089, + "language_loss": 0.68059635, + "learning_rate": 9.535160126239294e-08, + "loss": 0.70226419, + "num_input_tokens_seen": 162638220, + "step": 7520, + "time_per_iteration": 2.6301262378692627 + }, + { + "auxiliary_loss_clip": 0.01140521, + "auxiliary_loss_mlp": 0.01024019, + "balance_loss_clip": 1.01077557, + "balance_loss_mlp": 1.01704872, + "epoch": 0.9043467804966031, + "flos": 24790428961920.0, + "grad_norm": 1.6022371361788843, + "language_loss": 0.7080732, + "learning_rate": 9.511409092067424e-08, + "loss": 0.72971869, + "num_input_tokens_seen": 162658575, + "step": 7521, + "time_per_iteration": 2.717280387878418 + }, + { + "auxiliary_loss_clip": 0.01117375, + "auxiliary_loss_mlp": 0.01025212, + "balance_loss_clip": 0.97314972, + "balance_loss_mlp": 1.01783013, + "epoch": 0.9044670233872423, + "flos": 22632125472000.0, + "grad_norm": 1.6794857698240226, + "language_loss": 0.67617124, + "learning_rate": 9.487686954863327e-08, + "loss": 0.69759709, + "num_input_tokens_seen": 162678295, + "step": 7522, + "time_per_iteration": 2.807753562927246 + }, + { + "auxiliary_loss_clip": 0.01140601, + "auxiliary_loss_mlp": 0.01031551, + "balance_loss_clip": 1.01147699, + "balance_loss_mlp": 1.02459824, + "epoch": 0.9045872662778813, + "flos": 23771320289280.0, + "grad_norm": 1.8296347452958535, + "language_loss": 0.769575, + "learning_rate": 9.46399371822566e-08, + "loss": 0.79129648, + "num_input_tokens_seen": 162698070, + "step": 7523, + "time_per_iteration": 2.7008306980133057 + }, + { + "auxiliary_loss_clip": 0.01166228, + "auxiliary_loss_mlp": 0.01027702, + "balance_loss_clip": 1.04916203, + "balance_loss_mlp": 1.02029943, + "epoch": 0.9047075091685204, + "flos": 15191039998080.0, + "grad_norm": 2.1010626927835196, + "language_loss": 0.72141957, + "learning_rate": 9.440329385748657e-08, + "loss": 0.74335885, + "num_input_tokens_seen": 162715140, + "step": 7524, + "time_per_iteration": 3.5164144039154053 + }, + { + "auxiliary_loss_clip": 0.01093467, + "auxiliary_loss_mlp": 0.01025022, + "balance_loss_clip": 0.93364596, + "balance_loss_mlp": 1.01872802, + "epoch": 0.9048277520591596, + "flos": 18003707504640.0, + "grad_norm": 1.8231122039488352, + "language_loss": 0.70480645, + "learning_rate": 9.416693961022137e-08, + "loss": 0.72599137, + "num_input_tokens_seen": 162733390, + "step": 7525, + "time_per_iteration": 2.7431111335754395 + }, + { + "auxiliary_loss_clip": 0.01033101, + "auxiliary_loss_mlp": 0.01024258, + "balance_loss_clip": 0.85267901, + "balance_loss_mlp": 1.01753747, + "epoch": 0.9049479949497986, + "flos": 21872471713920.0, + "grad_norm": 2.1334026322955233, + "language_loss": 0.77212262, + "learning_rate": 9.393087447631654e-08, + "loss": 0.79269624, + "num_input_tokens_seen": 162751670, + "step": 7526, + "time_per_iteration": 2.790696382522583 + }, + { + "auxiliary_loss_clip": 0.01118665, + "auxiliary_loss_mlp": 0.01021606, + "balance_loss_clip": 0.97074181, + "balance_loss_mlp": 1.01448286, + "epoch": 0.9050682378404377, + "flos": 20773928113920.0, + "grad_norm": 1.564501362023424, + "language_loss": 0.72768313, + "learning_rate": 9.36950984915823e-08, + "loss": 0.74908584, + "num_input_tokens_seen": 162770025, + "step": 7527, + "time_per_iteration": 3.6562390327453613 + }, + { + "auxiliary_loss_clip": 0.01166132, + "auxiliary_loss_mlp": 0.01027358, + "balance_loss_clip": 1.04940271, + "balance_loss_mlp": 1.02024102, + "epoch": 0.9051884807310768, + "flos": 21580015178880.0, + "grad_norm": 1.785907049565379, + "language_loss": 0.6910224, + "learning_rate": 9.345961169178607e-08, + "loss": 0.71295738, + "num_input_tokens_seen": 162789710, + "step": 7528, + "time_per_iteration": 3.662391185760498 + }, + { + "auxiliary_loss_clip": 0.01088233, + "auxiliary_loss_mlp": 0.0102591, + "balance_loss_clip": 0.93896782, + "balance_loss_mlp": 1.01900804, + "epoch": 0.9053087236217159, + "flos": 21908059113600.0, + "grad_norm": 1.7358553984553053, + "language_loss": 0.72651505, + "learning_rate": 9.322441411265081e-08, + "loss": 0.74765646, + "num_input_tokens_seen": 162810695, + "step": 7529, + "time_per_iteration": 2.7048826217651367 + }, + { + "auxiliary_loss_clip": 0.01113284, + "auxiliary_loss_mlp": 0.010252, + "balance_loss_clip": 0.97081411, + "balance_loss_mlp": 1.01822603, + "epoch": 0.9054289665123549, + "flos": 17055809544960.0, + "grad_norm": 3.8568233212779996, + "language_loss": 0.73045921, + "learning_rate": 9.298950578985554e-08, + "loss": 0.75184399, + "num_input_tokens_seen": 162827770, + "step": 7530, + "time_per_iteration": 2.631464958190918 + }, + { + "auxiliary_loss_clip": 0.01138342, + "auxiliary_loss_mlp": 0.00509703, + "balance_loss_clip": 1.01093698, + "balance_loss_mlp": 0.0, + "epoch": 0.905549209402994, + "flos": 20777268078720.0, + "grad_norm": 2.397157006118175, + "language_loss": 0.70741081, + "learning_rate": 9.275488675903665e-08, + "loss": 0.72389126, + "num_input_tokens_seen": 162846715, + "step": 7531, + "time_per_iteration": 2.650404930114746 + }, + { + "auxiliary_loss_clip": 0.01070087, + "auxiliary_loss_mlp": 0.0102344, + "balance_loss_clip": 0.89686906, + "balance_loss_mlp": 1.0162282, + "epoch": 0.9056694522936332, + "flos": 21686813291520.0, + "grad_norm": 1.8928585358555146, + "language_loss": 0.7349686, + "learning_rate": 9.252055705578454e-08, + "loss": 0.75590384, + "num_input_tokens_seen": 162866215, + "step": 7532, + "time_per_iteration": 2.6951000690460205 + }, + { + "auxiliary_loss_clip": 0.01139742, + "auxiliary_loss_mlp": 0.01022924, + "balance_loss_clip": 1.00890589, + "balance_loss_mlp": 1.01578951, + "epoch": 0.9057896951842722, + "flos": 29569133433600.0, + "grad_norm": 1.781150443400286, + "language_loss": 0.72174603, + "learning_rate": 9.228651671564747e-08, + "loss": 0.74337268, + "num_input_tokens_seen": 162888245, + "step": 7533, + "time_per_iteration": 2.7555959224700928 + }, + { + "auxiliary_loss_clip": 0.01069025, + "auxiliary_loss_mlp": 0.01024148, + "balance_loss_clip": 0.8964116, + "balance_loss_mlp": 1.01748395, + "epoch": 0.9059099380749113, + "flos": 27892248952320.0, + "grad_norm": 1.4792534897528633, + "language_loss": 0.77847207, + "learning_rate": 9.205276577412901e-08, + "loss": 0.79940385, + "num_input_tokens_seen": 162911025, + "step": 7534, + "time_per_iteration": 2.8033714294433594 + }, + { + "auxiliary_loss_clip": 0.01124033, + "auxiliary_loss_mlp": 0.00509643, + "balance_loss_clip": 0.9709214, + "balance_loss_mlp": 0.0, + "epoch": 0.9060301809655504, + "flos": 17748993185280.0, + "grad_norm": 2.5722174478365867, + "language_loss": 0.76761556, + "learning_rate": 9.181930426668905e-08, + "loss": 0.78395224, + "num_input_tokens_seen": 162927820, + "step": 7535, + "time_per_iteration": 2.7070329189300537 + }, + { + "auxiliary_loss_clip": 0.01070466, + "auxiliary_loss_mlp": 0.01028616, + "balance_loss_clip": 0.89707714, + "balance_loss_mlp": 1.02200866, + "epoch": 0.9061504238561895, + "flos": 31759432963200.0, + "grad_norm": 2.4139116008679236, + "language_loss": 0.67848402, + "learning_rate": 9.158613222874346e-08, + "loss": 0.69947481, + "num_input_tokens_seen": 162949445, + "step": 7536, + "time_per_iteration": 3.682749032974243 + }, + { + "auxiliary_loss_clip": 0.01116882, + "auxiliary_loss_mlp": 0.01025438, + "balance_loss_clip": 0.970891, + "balance_loss_mlp": 1.01854753, + "epoch": 0.9062706667468285, + "flos": 20048066075520.0, + "grad_norm": 1.6094112843927764, + "language_loss": 0.82050133, + "learning_rate": 9.135324969566394e-08, + "loss": 0.84192455, + "num_input_tokens_seen": 162968945, + "step": 7537, + "time_per_iteration": 2.6457483768463135 + }, + { + "auxiliary_loss_clip": 0.01146644, + "auxiliary_loss_mlp": 0.01021059, + "balance_loss_clip": 1.01159692, + "balance_loss_mlp": 1.01425183, + "epoch": 0.9063909096374677, + "flos": 18437292576000.0, + "grad_norm": 2.118234303841104, + "language_loss": 0.75398928, + "learning_rate": 9.112065670277913e-08, + "loss": 0.77566624, + "num_input_tokens_seen": 162985310, + "step": 7538, + "time_per_iteration": 2.637674331665039 + }, + { + "auxiliary_loss_clip": 0.01110801, + "auxiliary_loss_mlp": 0.01025885, + "balance_loss_clip": 0.96676296, + "balance_loss_mlp": 1.01932824, + "epoch": 0.9065111525281068, + "flos": 33547353361920.0, + "grad_norm": 1.8312215878153377, + "language_loss": 0.73127961, + "learning_rate": 9.088835328537303e-08, + "loss": 0.75264657, + "num_input_tokens_seen": 163006900, + "step": 7539, + "time_per_iteration": 2.837228298187256 + }, + { + "auxiliary_loss_clip": 0.01121306, + "auxiliary_loss_mlp": 0.01024063, + "balance_loss_clip": 0.97209466, + "balance_loss_mlp": 1.01681769, + "epoch": 0.9066313954187458, + "flos": 23367863750400.0, + "grad_norm": 2.215968563071286, + "language_loss": 0.71411175, + "learning_rate": 9.065633947868568e-08, + "loss": 0.73556548, + "num_input_tokens_seen": 163026505, + "step": 7540, + "time_per_iteration": 2.7423248291015625 + }, + { + "auxiliary_loss_clip": 0.01092539, + "auxiliary_loss_mlp": 0.0050866, + "balance_loss_clip": 0.93349612, + "balance_loss_mlp": 0.0, + "epoch": 0.906751638309385, + "flos": 26249623067520.0, + "grad_norm": 2.1215521751714026, + "language_loss": 0.79927856, + "learning_rate": 9.042461531791379e-08, + "loss": 0.81529051, + "num_input_tokens_seen": 163044925, + "step": 7541, + "time_per_iteration": 2.813324451446533 + }, + { + "auxiliary_loss_clip": 0.01158186, + "auxiliary_loss_mlp": 0.01025025, + "balance_loss_clip": 1.04546142, + "balance_loss_mlp": 1.01871896, + "epoch": 0.906871881200024, + "flos": 16544477485440.0, + "grad_norm": 1.92252393018608, + "language_loss": 0.78274035, + "learning_rate": 9.019318083820903e-08, + "loss": 0.8045724, + "num_input_tokens_seen": 163063505, + "step": 7542, + "time_per_iteration": 2.63885498046875 + }, + { + "auxiliary_loss_clip": 0.01139088, + "auxiliary_loss_mlp": 0.01027913, + "balance_loss_clip": 1.00948882, + "balance_loss_mlp": 1.02083766, + "epoch": 0.9069921240906631, + "flos": 24605129675520.0, + "grad_norm": 1.5925701123250149, + "language_loss": 0.84981698, + "learning_rate": 8.996203607468045e-08, + "loss": 0.87148702, + "num_input_tokens_seen": 163082505, + "step": 7543, + "time_per_iteration": 2.644848346710205 + }, + { + "auxiliary_loss_clip": 0.01135368, + "auxiliary_loss_mlp": 0.01026256, + "balance_loss_clip": 1.00593138, + "balance_loss_mlp": 1.01921034, + "epoch": 0.9071123669813023, + "flos": 25374731500800.0, + "grad_norm": 1.491738826981858, + "language_loss": 0.75275367, + "learning_rate": 8.973118106239241e-08, + "loss": 0.7743699, + "num_input_tokens_seen": 163105110, + "step": 7544, + "time_per_iteration": 2.741412878036499 + }, + { + "auxiliary_loss_clip": 0.0104564, + "auxiliary_loss_mlp": 0.01023843, + "balance_loss_clip": 0.85122597, + "balance_loss_mlp": 1.01678014, + "epoch": 0.9072326098719413, + "flos": 26725798690560.0, + "grad_norm": 1.9618763932203964, + "language_loss": 0.94846821, + "learning_rate": 8.95006158363656e-08, + "loss": 0.969163, + "num_input_tokens_seen": 163125295, + "step": 7545, + "time_per_iteration": 2.849466323852539 + }, + { + "auxiliary_loss_clip": 0.01144365, + "auxiliary_loss_mlp": 0.01027269, + "balance_loss_clip": 1.01364708, + "balance_loss_mlp": 1.01986635, + "epoch": 0.9073528527625804, + "flos": 23878800760320.0, + "grad_norm": 1.6313641923153384, + "language_loss": 0.7711333, + "learning_rate": 8.9270340431576e-08, + "loss": 0.79284966, + "num_input_tokens_seen": 163144385, + "step": 7546, + "time_per_iteration": 2.5715627670288086 + }, + { + "auxiliary_loss_clip": 0.01142353, + "auxiliary_loss_mlp": 0.01024438, + "balance_loss_clip": 1.00949144, + "balance_loss_mlp": 1.01756513, + "epoch": 0.9074730956532195, + "flos": 37852144767360.0, + "grad_norm": 1.975922724436064, + "language_loss": 0.73529321, + "learning_rate": 8.904035488295658e-08, + "loss": 0.75696117, + "num_input_tokens_seen": 163163885, + "step": 7547, + "time_per_iteration": 2.7477564811706543 + }, + { + "auxiliary_loss_clip": 0.01042007, + "auxiliary_loss_mlp": 0.0050068, + "balance_loss_clip": 0.97384119, + "balance_loss_mlp": 0.0, + "epoch": 0.9075933385438586, + "flos": 65173307385600.0, + "grad_norm": 0.6653877218843588, + "language_loss": 0.53230333, + "learning_rate": 8.881065922539632e-08, + "loss": 0.54773021, + "num_input_tokens_seen": 163224325, + "step": 7548, + "time_per_iteration": 3.1142642498016357 + }, + { + "auxiliary_loss_clip": 0.01091421, + "auxiliary_loss_mlp": 0.01023095, + "balance_loss_clip": 0.93659008, + "balance_loss_mlp": 1.01601124, + "epoch": 0.9077135814344977, + "flos": 19931571290880.0, + "grad_norm": 2.207783356065874, + "language_loss": 0.73577094, + "learning_rate": 8.85812534937389e-08, + "loss": 0.75691611, + "num_input_tokens_seen": 163242425, + "step": 7549, + "time_per_iteration": 2.7075231075286865 + }, + { + "auxiliary_loss_clip": 0.0114644, + "auxiliary_loss_mlp": 0.01024357, + "balance_loss_clip": 1.0102222, + "balance_loss_mlp": 1.01706123, + "epoch": 0.9078338243251368, + "flos": 17529650784000.0, + "grad_norm": 2.635253513860539, + "language_loss": 0.68112814, + "learning_rate": 8.835213772278583e-08, + "loss": 0.70283616, + "num_input_tokens_seen": 163259280, + "step": 7550, + "time_per_iteration": 3.4878671169281006 + }, + { + "auxiliary_loss_clip": 0.01088305, + "auxiliary_loss_mlp": 0.01022783, + "balance_loss_clip": 0.93567789, + "balance_loss_mlp": 1.01573169, + "epoch": 0.9079540672157759, + "flos": 28803410277120.0, + "grad_norm": 1.758646436803053, + "language_loss": 0.7915892, + "learning_rate": 8.812331194729373e-08, + "loss": 0.81270009, + "num_input_tokens_seen": 163278925, + "step": 7551, + "time_per_iteration": 2.753875255584717 + }, + { + "auxiliary_loss_clip": 0.0116758, + "auxiliary_loss_mlp": 0.01030313, + "balance_loss_clip": 1.05050898, + "balance_loss_mlp": 1.02270198, + "epoch": 0.9080743101064149, + "flos": 23513840622720.0, + "grad_norm": 3.2736105837190523, + "language_loss": 0.71961308, + "learning_rate": 8.789477620197461e-08, + "loss": 0.74159205, + "num_input_tokens_seen": 163298450, + "step": 7552, + "time_per_iteration": 2.639772415161133 + }, + { + "auxiliary_loss_clip": 0.01116242, + "auxiliary_loss_mlp": 0.01029955, + "balance_loss_clip": 0.97138387, + "balance_loss_mlp": 1.02248633, + "epoch": 0.9081945529970541, + "flos": 22778102344320.0, + "grad_norm": 2.034498316026933, + "language_loss": 0.79086655, + "learning_rate": 8.766653052149831e-08, + "loss": 0.81232846, + "num_input_tokens_seen": 163313635, + "step": 7553, + "time_per_iteration": 3.584939479827881 + }, + { + "auxiliary_loss_clip": 0.01118959, + "auxiliary_loss_mlp": 0.01030177, + "balance_loss_clip": 0.97168732, + "balance_loss_mlp": 1.02285147, + "epoch": 0.9083147958876931, + "flos": 18873714821760.0, + "grad_norm": 2.078142119531839, + "language_loss": 0.74609387, + "learning_rate": 8.743857494048823e-08, + "loss": 0.76758528, + "num_input_tokens_seen": 163330450, + "step": 7554, + "time_per_iteration": 3.590228796005249 + }, + { + "auxiliary_loss_clip": 0.0109372, + "auxiliary_loss_mlp": 0.01025645, + "balance_loss_clip": 0.93395162, + "balance_loss_mlp": 1.01875126, + "epoch": 0.9084350387783322, + "flos": 18909374048640.0, + "grad_norm": 1.8956386861624823, + "language_loss": 0.62994385, + "learning_rate": 8.721090949352605e-08, + "loss": 0.65113753, + "num_input_tokens_seen": 163346690, + "step": 7555, + "time_per_iteration": 2.701000928878784 + }, + { + "auxiliary_loss_clip": 0.01150239, + "auxiliary_loss_mlp": 0.01029658, + "balance_loss_clip": 1.01142967, + "balance_loss_mlp": 1.02190936, + "epoch": 0.9085552816689714, + "flos": 20595488325120.0, + "grad_norm": 2.043101294215774, + "language_loss": 0.73001218, + "learning_rate": 8.698353421514793e-08, + "loss": 0.75181115, + "num_input_tokens_seen": 163365065, + "step": 7556, + "time_per_iteration": 2.5960450172424316 + }, + { + "auxiliary_loss_clip": 0.01141159, + "auxiliary_loss_mlp": 0.01025684, + "balance_loss_clip": 1.01159596, + "balance_loss_mlp": 1.01916313, + "epoch": 0.9086755245596104, + "flos": 18113163223680.0, + "grad_norm": 2.285902553580392, + "language_loss": 0.80283397, + "learning_rate": 8.67564491398467e-08, + "loss": 0.82450247, + "num_input_tokens_seen": 163382070, + "step": 7557, + "time_per_iteration": 2.6244924068450928 + }, + { + "auxiliary_loss_clip": 0.01143945, + "auxiliary_loss_mlp": 0.01026606, + "balance_loss_clip": 1.01005781, + "balance_loss_mlp": 1.01949489, + "epoch": 0.9087957674502495, + "flos": 19129793857920.0, + "grad_norm": 2.2882961868670195, + "language_loss": 0.74042666, + "learning_rate": 8.652965430207104e-08, + "loss": 0.76213223, + "num_input_tokens_seen": 163399975, + "step": 7558, + "time_per_iteration": 2.6385879516601562 + }, + { + "auxiliary_loss_clip": 0.01144761, + "auxiliary_loss_mlp": 0.01024636, + "balance_loss_clip": 1.01063466, + "balance_loss_mlp": 1.01727509, + "epoch": 0.9089160103408886, + "flos": 18109930999680.0, + "grad_norm": 2.0605665763461016, + "language_loss": 0.65790761, + "learning_rate": 8.630314973622521e-08, + "loss": 0.67960155, + "num_input_tokens_seen": 163417520, + "step": 7559, + "time_per_iteration": 2.6617543697357178 + }, + { + "auxiliary_loss_clip": 0.01139395, + "auxiliary_loss_mlp": 0.01027131, + "balance_loss_clip": 1.01191545, + "balance_loss_mlp": 1.02040207, + "epoch": 0.9090362532315277, + "flos": 33364855336320.0, + "grad_norm": 1.8927978328195318, + "language_loss": 0.70859063, + "learning_rate": 8.607693547666995e-08, + "loss": 0.73025596, + "num_input_tokens_seen": 163440060, + "step": 7560, + "time_per_iteration": 2.770784378051758 + }, + { + "auxiliary_loss_clip": 0.01002399, + "auxiliary_loss_mlp": 0.01000795, + "balance_loss_clip": 0.89830995, + "balance_loss_mlp": 0.99979967, + "epoch": 0.9091564961221668, + "flos": 71480585082240.0, + "grad_norm": 0.8856755067540258, + "language_loss": 0.57966876, + "learning_rate": 8.585101155772201e-08, + "loss": 0.59970069, + "num_input_tokens_seen": 163502180, + "step": 7561, + "time_per_iteration": 4.316712856292725 + }, + { + "auxiliary_loss_clip": 0.01109452, + "auxiliary_loss_mlp": 0.01024379, + "balance_loss_clip": 0.96616173, + "balance_loss_mlp": 1.01705313, + "epoch": 0.9092767390128058, + "flos": 24712574232960.0, + "grad_norm": 2.055829507482017, + "language_loss": 0.68443537, + "learning_rate": 8.562537801365377e-08, + "loss": 0.70577371, + "num_input_tokens_seen": 163521915, + "step": 7562, + "time_per_iteration": 2.745483160018921 + }, + { + "auxiliary_loss_clip": 0.01164075, + "auxiliary_loss_mlp": 0.01031213, + "balance_loss_clip": 1.047647, + "balance_loss_mlp": 1.0237689, + "epoch": 0.909396981903445, + "flos": 23586487879680.0, + "grad_norm": 1.8283000236997147, + "language_loss": 0.70168388, + "learning_rate": 8.540003487869362e-08, + "loss": 0.72363681, + "num_input_tokens_seen": 163543585, + "step": 7563, + "time_per_iteration": 2.674276828765869 + }, + { + "auxiliary_loss_clip": 0.01083753, + "auxiliary_loss_mlp": 0.01022884, + "balance_loss_clip": 0.93159783, + "balance_loss_mlp": 1.01617551, + "epoch": 0.909517224794084, + "flos": 23404169422080.0, + "grad_norm": 1.8229252479294829, + "language_loss": 0.79554474, + "learning_rate": 8.517498218702557e-08, + "loss": 0.81661105, + "num_input_tokens_seen": 163561515, + "step": 7564, + "time_per_iteration": 2.70184063911438 + }, + { + "auxiliary_loss_clip": 0.01087622, + "auxiliary_loss_mlp": 0.01026103, + "balance_loss_clip": 0.93008184, + "balance_loss_mlp": 1.01881373, + "epoch": 0.9096374676847231, + "flos": 19208618254080.0, + "grad_norm": 2.337258658267096, + "language_loss": 0.69744968, + "learning_rate": 8.49502199727905e-08, + "loss": 0.71858698, + "num_input_tokens_seen": 163579540, + "step": 7565, + "time_per_iteration": 2.6750869750976562 + }, + { + "auxiliary_loss_clip": 0.01135567, + "auxiliary_loss_mlp": 0.0102904, + "balance_loss_clip": 1.00729084, + "balance_loss_mlp": 1.02149725, + "epoch": 0.9097577105753623, + "flos": 33292495388160.0, + "grad_norm": 2.9225676900604256, + "language_loss": 0.66498053, + "learning_rate": 8.472574827008428e-08, + "loss": 0.68662661, + "num_input_tokens_seen": 163600425, + "step": 7566, + "time_per_iteration": 2.774531364440918 + }, + { + "auxiliary_loss_clip": 0.01139298, + "auxiliary_loss_mlp": 0.01026442, + "balance_loss_clip": 1.00786257, + "balance_loss_mlp": 1.01922393, + "epoch": 0.9098779534660013, + "flos": 21906443001600.0, + "grad_norm": 1.6578026626113387, + "language_loss": 0.83753639, + "learning_rate": 8.450156711295942e-08, + "loss": 0.8591938, + "num_input_tokens_seen": 163620595, + "step": 7567, + "time_per_iteration": 2.641451120376587 + }, + { + "auxiliary_loss_clip": 0.01115507, + "auxiliary_loss_mlp": 0.01029249, + "balance_loss_clip": 0.9725486, + "balance_loss_mlp": 1.02252817, + "epoch": 0.9099981963566404, + "flos": 25730354102400.0, + "grad_norm": 1.9590690418454877, + "language_loss": 0.86956245, + "learning_rate": 8.427767653542383e-08, + "loss": 0.89100999, + "num_input_tokens_seen": 163635765, + "step": 7568, + "time_per_iteration": 2.7042548656463623 + }, + { + "auxiliary_loss_clip": 0.0106521, + "auxiliary_loss_mlp": 0.01026612, + "balance_loss_clip": 0.89263189, + "balance_loss_mlp": 1.01985288, + "epoch": 0.9101184392472795, + "flos": 21069437304960.0, + "grad_norm": 1.7989146577521709, + "language_loss": 0.70289803, + "learning_rate": 8.405407657144125e-08, + "loss": 0.72381628, + "num_input_tokens_seen": 163654925, + "step": 7569, + "time_per_iteration": 2.731152057647705 + }, + { + "auxiliary_loss_clip": 0.01112492, + "auxiliary_loss_mlp": 0.01027459, + "balance_loss_clip": 0.97107178, + "balance_loss_mlp": 1.02101874, + "epoch": 0.9102386821379186, + "flos": 24752614919040.0, + "grad_norm": 1.8850888711536071, + "language_loss": 0.72731328, + "learning_rate": 8.383076725493232e-08, + "loss": 0.74871284, + "num_input_tokens_seen": 163672245, + "step": 7570, + "time_per_iteration": 2.7165231704711914 + }, + { + "auxiliary_loss_clip": 0.01142245, + "auxiliary_loss_mlp": 0.01025094, + "balance_loss_clip": 1.01019788, + "balance_loss_mlp": 1.01802182, + "epoch": 0.9103589250285576, + "flos": 22562818179840.0, + "grad_norm": 1.9865138583499913, + "language_loss": 0.68484432, + "learning_rate": 8.360774861977216e-08, + "loss": 0.7065177, + "num_input_tokens_seen": 163691365, + "step": 7571, + "time_per_iteration": 2.6201961040496826 + }, + { + "auxiliary_loss_clip": 0.01114349, + "auxiliary_loss_mlp": 0.01025, + "balance_loss_clip": 0.9670974, + "balance_loss_mlp": 1.01842892, + "epoch": 0.9104791679191968, + "flos": 25373474524800.0, + "grad_norm": 4.044707587517256, + "language_loss": 0.74775255, + "learning_rate": 8.338502069979281e-08, + "loss": 0.76914608, + "num_input_tokens_seen": 163711675, + "step": 7572, + "time_per_iteration": 2.7381792068481445 + }, + { + "auxiliary_loss_clip": 0.01140159, + "auxiliary_loss_mlp": 0.01021059, + "balance_loss_clip": 1.00587404, + "balance_loss_mlp": 1.01401687, + "epoch": 0.9105994108098359, + "flos": 14426681558400.0, + "grad_norm": 2.9847977093513745, + "language_loss": 0.79375774, + "learning_rate": 8.316258352878214e-08, + "loss": 0.8153699, + "num_input_tokens_seen": 163728095, + "step": 7573, + "time_per_iteration": 2.664867401123047 + }, + { + "auxiliary_loss_clip": 0.0114356, + "auxiliary_loss_mlp": 0.01026185, + "balance_loss_clip": 1.00770462, + "balance_loss_mlp": 1.0182991, + "epoch": 0.9107196537004749, + "flos": 26718292748160.0, + "grad_norm": 7.777224284245735, + "language_loss": 0.71368885, + "learning_rate": 8.294043714048338e-08, + "loss": 0.73538637, + "num_input_tokens_seen": 163747175, + "step": 7574, + "time_per_iteration": 2.6927223205566406 + }, + { + "auxiliary_loss_clip": 0.01021519, + "auxiliary_loss_mlp": 0.01000374, + "balance_loss_clip": 0.93462014, + "balance_loss_mlp": 0.99939632, + "epoch": 0.9108398965911141, + "flos": 66532634703360.0, + "grad_norm": 0.7475337029619682, + "language_loss": 0.60478771, + "learning_rate": 8.271858156859624e-08, + "loss": 0.62500668, + "num_input_tokens_seen": 163812545, + "step": 7575, + "time_per_iteration": 3.286606788635254 + }, + { + "auxiliary_loss_clip": 0.01162477, + "auxiliary_loss_mlp": 0.01022352, + "balance_loss_clip": 1.04846501, + "balance_loss_mlp": 1.01568508, + "epoch": 0.9109601394817531, + "flos": 25411073086080.0, + "grad_norm": 1.567491403803516, + "language_loss": 0.73723596, + "learning_rate": 8.249701684677557e-08, + "loss": 0.75908422, + "num_input_tokens_seen": 163833870, + "step": 7576, + "time_per_iteration": 3.591566324234009 + }, + { + "auxiliary_loss_clip": 0.01145156, + "auxiliary_loss_mlp": 0.01030544, + "balance_loss_clip": 1.01366472, + "balance_loss_mlp": 1.02331984, + "epoch": 0.9110803823723922, + "flos": 22747794243840.0, + "grad_norm": 1.7507805759164767, + "language_loss": 0.81258571, + "learning_rate": 8.227574300863294e-08, + "loss": 0.83434272, + "num_input_tokens_seen": 163854040, + "step": 7577, + "time_per_iteration": 2.658351421356201 + }, + { + "auxiliary_loss_clip": 0.0112552, + "auxiliary_loss_mlp": 0.01031155, + "balance_loss_clip": 0.97650135, + "balance_loss_mlp": 1.02399659, + "epoch": 0.9112006252630314, + "flos": 48469924131840.0, + "grad_norm": 1.5476055651674128, + "language_loss": 0.69415909, + "learning_rate": 8.205476008773548e-08, + "loss": 0.7157259, + "num_input_tokens_seen": 163878040, + "step": 7578, + "time_per_iteration": 3.8606925010681152 + }, + { + "auxiliary_loss_clip": 0.01089601, + "auxiliary_loss_mlp": 0.01027343, + "balance_loss_clip": 0.93565524, + "balance_loss_mlp": 1.02037549, + "epoch": 0.9113208681536704, + "flos": 30009649829760.0, + "grad_norm": 2.0062920715967185, + "language_loss": 0.82240397, + "learning_rate": 8.183406811760596e-08, + "loss": 0.84357345, + "num_input_tokens_seen": 163897770, + "step": 7579, + "time_per_iteration": 2.7363204956054688 + }, + { + "auxiliary_loss_clip": 0.01080844, + "auxiliary_loss_mlp": 0.01023564, + "balance_loss_clip": 0.93008387, + "balance_loss_mlp": 1.01657832, + "epoch": 0.9114411110443095, + "flos": 25594971742080.0, + "grad_norm": 1.5094873407317844, + "language_loss": 0.742212, + "learning_rate": 8.161366713172313e-08, + "loss": 0.76325607, + "num_input_tokens_seen": 163920160, + "step": 7580, + "time_per_iteration": 3.8118269443511963 + }, + { + "auxiliary_loss_clip": 0.01097144, + "auxiliary_loss_mlp": 0.01024346, + "balance_loss_clip": 0.93196356, + "balance_loss_mlp": 1.01716375, + "epoch": 0.9115613539349486, + "flos": 18399729928320.0, + "grad_norm": 2.932060541783875, + "language_loss": 0.841295, + "learning_rate": 8.139355716352137e-08, + "loss": 0.86250997, + "num_input_tokens_seen": 163935000, + "step": 7581, + "time_per_iteration": 2.697112798690796 + }, + { + "auxiliary_loss_clip": 0.01120095, + "auxiliary_loss_mlp": 0.01024297, + "balance_loss_clip": 0.97043669, + "balance_loss_mlp": 1.01695621, + "epoch": 0.9116815968255877, + "flos": 21726171619200.0, + "grad_norm": 2.279412742412915, + "language_loss": 0.70293772, + "learning_rate": 8.117373824639196e-08, + "loss": 0.72438169, + "num_input_tokens_seen": 163955265, + "step": 7582, + "time_per_iteration": 2.709439516067505 + }, + { + "auxiliary_loss_clip": 0.01059449, + "auxiliary_loss_mlp": 0.01001036, + "balance_loss_clip": 1.0089066, + "balance_loss_mlp": 1.00006437, + "epoch": 0.9118018397162267, + "flos": 65363526835200.0, + "grad_norm": 0.7233396563964899, + "language_loss": 0.59320426, + "learning_rate": 8.095421041368067e-08, + "loss": 0.61380911, + "num_input_tokens_seen": 164014680, + "step": 7583, + "time_per_iteration": 3.0873565673828125 + }, + { + "auxiliary_loss_clip": 0.01117534, + "auxiliary_loss_mlp": 0.00509246, + "balance_loss_clip": 0.9728508, + "balance_loss_mlp": 0.0, + "epoch": 0.9119220826068659, + "flos": 20922885815040.0, + "grad_norm": 2.269975331838521, + "language_loss": 0.70511723, + "learning_rate": 8.073497369868999e-08, + "loss": 0.721385, + "num_input_tokens_seen": 164033140, + "step": 7584, + "time_per_iteration": 2.694303274154663 + }, + { + "auxiliary_loss_clip": 0.01123048, + "auxiliary_loss_mlp": 0.01023783, + "balance_loss_clip": 0.97114968, + "balance_loss_mlp": 1.01656508, + "epoch": 0.912042325497505, + "flos": 28366449327360.0, + "grad_norm": 1.7166703518732296, + "language_loss": 0.75627995, + "learning_rate": 8.051602813467772e-08, + "loss": 0.77774829, + "num_input_tokens_seen": 164054995, + "step": 7585, + "time_per_iteration": 2.7804243564605713 + }, + { + "auxiliary_loss_clip": 0.01143594, + "auxiliary_loss_mlp": 0.01027426, + "balance_loss_clip": 1.01008964, + "balance_loss_mlp": 1.02060091, + "epoch": 0.912162568388144, + "flos": 17566782468480.0, + "grad_norm": 1.630184210882542, + "language_loss": 0.7076757, + "learning_rate": 8.029737375485756e-08, + "loss": 0.72938591, + "num_input_tokens_seen": 164074225, + "step": 7586, + "time_per_iteration": 2.670387029647827 + }, + { + "auxiliary_loss_clip": 0.011653, + "auxiliary_loss_mlp": 0.0102858, + "balance_loss_clip": 1.04912555, + "balance_loss_mlp": 1.02162445, + "epoch": 0.9122828112787832, + "flos": 19827897661440.0, + "grad_norm": 1.9018590912062028, + "language_loss": 0.7243942, + "learning_rate": 8.007901059239986e-08, + "loss": 0.746333, + "num_input_tokens_seen": 164093505, + "step": 7587, + "time_per_iteration": 3.5645363330841064 + }, + { + "auxiliary_loss_clip": 0.01116057, + "auxiliary_loss_mlp": 0.010187, + "balance_loss_clip": 0.96928847, + "balance_loss_mlp": 1.01203918, + "epoch": 0.9124030541694222, + "flos": 20813789232000.0, + "grad_norm": 1.6503938168918628, + "language_loss": 0.80045438, + "learning_rate": 7.986093868042964e-08, + "loss": 0.8218019, + "num_input_tokens_seen": 164113750, + "step": 7588, + "time_per_iteration": 2.7014856338500977 + }, + { + "auxiliary_loss_clip": 0.01139244, + "auxiliary_loss_mlp": 0.01025011, + "balance_loss_clip": 1.00932503, + "balance_loss_mlp": 1.0183202, + "epoch": 0.9125232970600613, + "flos": 25192305302400.0, + "grad_norm": 1.8421826361089269, + "language_loss": 0.67620134, + "learning_rate": 7.964315805202826e-08, + "loss": 0.69784385, + "num_input_tokens_seen": 164134330, + "step": 7589, + "time_per_iteration": 2.6798782348632812 + }, + { + "auxiliary_loss_clip": 0.01121075, + "auxiliary_loss_mlp": 0.01026553, + "balance_loss_clip": 0.97450149, + "balance_loss_mlp": 1.0191381, + "epoch": 0.9126435399507005, + "flos": 19719591177600.0, + "grad_norm": 1.7686698884866183, + "language_loss": 0.73346698, + "learning_rate": 7.942566874023304e-08, + "loss": 0.75494331, + "num_input_tokens_seen": 164153515, + "step": 7590, + "time_per_iteration": 2.6890487670898438 + }, + { + "auxiliary_loss_clip": 0.01114649, + "auxiliary_loss_mlp": 0.01024284, + "balance_loss_clip": 0.96863097, + "balance_loss_mlp": 1.01734352, + "epoch": 0.9127637828413395, + "flos": 19573614305280.0, + "grad_norm": 1.978674189988319, + "language_loss": 0.69684803, + "learning_rate": 7.920847077803649e-08, + "loss": 0.7182374, + "num_input_tokens_seen": 164171305, + "step": 7591, + "time_per_iteration": 2.691124439239502 + }, + { + "auxiliary_loss_clip": 0.0105846, + "auxiliary_loss_mlp": 0.01028139, + "balance_loss_clip": 0.8880645, + "balance_loss_mlp": 1.02087367, + "epoch": 0.9128840257319786, + "flos": 20230635928320.0, + "grad_norm": 1.9382835422525027, + "language_loss": 0.82135999, + "learning_rate": 7.899156419838826e-08, + "loss": 0.84222603, + "num_input_tokens_seen": 164190275, + "step": 7592, + "time_per_iteration": 2.8347530364990234 + }, + { + "auxiliary_loss_clip": 0.01091971, + "auxiliary_loss_mlp": 0.01023635, + "balance_loss_clip": 0.93200356, + "balance_loss_mlp": 1.016819, + "epoch": 0.9130042686226177, + "flos": 24858658846080.0, + "grad_norm": 1.7026244759461362, + "language_loss": 0.65674227, + "learning_rate": 7.87749490341918e-08, + "loss": 0.67789835, + "num_input_tokens_seen": 164210550, + "step": 7593, + "time_per_iteration": 2.7104744911193848 + }, + { + "auxiliary_loss_clip": 0.0116749, + "auxiliary_loss_mlp": 0.01025084, + "balance_loss_clip": 1.049981, + "balance_loss_mlp": 1.01737165, + "epoch": 0.9131245115132568, + "flos": 23581747284480.0, + "grad_norm": 2.0347151797570926, + "language_loss": 0.83806586, + "learning_rate": 7.855862531830836e-08, + "loss": 0.85999161, + "num_input_tokens_seen": 164226660, + "step": 7594, + "time_per_iteration": 2.797539234161377 + }, + { + "auxiliary_loss_clip": 0.0113743, + "auxiliary_loss_mlp": 0.01022881, + "balance_loss_clip": 1.00712228, + "balance_loss_mlp": 1.01548648, + "epoch": 0.9132447544038959, + "flos": 19931607204480.0, + "grad_norm": 1.614184133770744, + "language_loss": 0.72638822, + "learning_rate": 7.834259308355373e-08, + "loss": 0.74799132, + "num_input_tokens_seen": 164245425, + "step": 7595, + "time_per_iteration": 2.596559524536133 + }, + { + "auxiliary_loss_clip": 0.0103081, + "auxiliary_loss_mlp": 0.01023392, + "balance_loss_clip": 0.85284412, + "balance_loss_mlp": 1.01649272, + "epoch": 0.9133649972945349, + "flos": 21981747864960.0, + "grad_norm": 2.019904188604184, + "language_loss": 0.75158221, + "learning_rate": 7.812685236269989e-08, + "loss": 0.77212429, + "num_input_tokens_seen": 164264085, + "step": 7596, + "time_per_iteration": 2.8812737464904785 + }, + { + "auxiliary_loss_clip": 0.00999053, + "auxiliary_loss_mlp": 0.01012246, + "balance_loss_clip": 0.89990896, + "balance_loss_mlp": 1.01114333, + "epoch": 0.9134852401851741, + "flos": 71240523511680.0, + "grad_norm": 0.7895129535937279, + "language_loss": 0.58610016, + "learning_rate": 7.791140318847445e-08, + "loss": 0.60621315, + "num_input_tokens_seen": 164322220, + "step": 7597, + "time_per_iteration": 3.2359042167663574 + }, + { + "auxiliary_loss_clip": 0.01116835, + "auxiliary_loss_mlp": 0.01022816, + "balance_loss_clip": 0.97573709, + "balance_loss_mlp": 1.0157913, + "epoch": 0.9136054830758131, + "flos": 23626923615360.0, + "grad_norm": 1.3912593622014908, + "language_loss": 0.80317479, + "learning_rate": 7.769624559356081e-08, + "loss": 0.82457131, + "num_input_tokens_seen": 164345615, + "step": 7598, + "time_per_iteration": 2.7193217277526855 + }, + { + "auxiliary_loss_clip": 0.01139901, + "auxiliary_loss_mlp": 0.01027949, + "balance_loss_clip": 1.01027298, + "balance_loss_mlp": 1.02107918, + "epoch": 0.9137257259664522, + "flos": 23438858981760.0, + "grad_norm": 4.285381061408213, + "language_loss": 0.75209451, + "learning_rate": 7.748137961059842e-08, + "loss": 0.77377301, + "num_input_tokens_seen": 164359595, + "step": 7599, + "time_per_iteration": 2.6225388050079346 + }, + { + "auxiliary_loss_clip": 0.01160976, + "auxiliary_loss_mlp": 0.01027489, + "balance_loss_clip": 1.0487504, + "balance_loss_mlp": 1.02050591, + "epoch": 0.9138459688570914, + "flos": 19127854523520.0, + "grad_norm": 2.142613618670162, + "language_loss": 0.65737975, + "learning_rate": 7.726680527218211e-08, + "loss": 0.67926443, + "num_input_tokens_seen": 164376635, + "step": 7600, + "time_per_iteration": 2.560971975326538 + }, + { + "auxiliary_loss_clip": 0.01164735, + "auxiliary_loss_mlp": 0.01024738, + "balance_loss_clip": 1.04763794, + "balance_loss_mlp": 1.01743007, + "epoch": 0.9139662117477304, + "flos": 46281240714240.0, + "grad_norm": 1.736992068971664, + "language_loss": 0.75407732, + "learning_rate": 7.70525226108627e-08, + "loss": 0.77597207, + "num_input_tokens_seen": 164400305, + "step": 7601, + "time_per_iteration": 2.881577253341675 + }, + { + "auxiliary_loss_clip": 0.01144616, + "auxiliary_loss_mlp": 0.0102943, + "balance_loss_clip": 1.0127759, + "balance_loss_mlp": 1.02252793, + "epoch": 0.9140864546383695, + "flos": 22273198819200.0, + "grad_norm": 1.9117549135079186, + "language_loss": 0.7969715, + "learning_rate": 7.683853165914666e-08, + "loss": 0.81871194, + "num_input_tokens_seen": 164418075, + "step": 7602, + "time_per_iteration": 3.524461030960083 + }, + { + "auxiliary_loss_clip": 0.01071787, + "auxiliary_loss_mlp": 0.01027749, + "balance_loss_clip": 0.89391685, + "balance_loss_mlp": 1.0205965, + "epoch": 0.9142066975290086, + "flos": 17530009920000.0, + "grad_norm": 1.7088401983426895, + "language_loss": 0.77232569, + "learning_rate": 7.662483244949602e-08, + "loss": 0.79332101, + "num_input_tokens_seen": 164435335, + "step": 7603, + "time_per_iteration": 2.77181077003479 + }, + { + "auxiliary_loss_clip": 0.0108836, + "auxiliary_loss_mlp": 0.01026437, + "balance_loss_clip": 0.93233049, + "balance_loss_mlp": 1.01954353, + "epoch": 0.9143269404196477, + "flos": 17712148809600.0, + "grad_norm": 2.6859404111932292, + "language_loss": 0.80760682, + "learning_rate": 7.641142501432951e-08, + "loss": 0.82875478, + "num_input_tokens_seen": 164451530, + "step": 7604, + "time_per_iteration": 3.6640331745147705 + }, + { + "auxiliary_loss_clip": 0.01113269, + "auxiliary_loss_mlp": 0.010288, + "balance_loss_clip": 0.97084802, + "balance_loss_mlp": 1.02209997, + "epoch": 0.9144471833102867, + "flos": 33323414019840.0, + "grad_norm": 1.8621973268245076, + "language_loss": 0.73632848, + "learning_rate": 7.619830938602013e-08, + "loss": 0.7577492, + "num_input_tokens_seen": 164472755, + "step": 7605, + "time_per_iteration": 2.778928756713867 + }, + { + "auxiliary_loss_clip": 0.01136335, + "auxiliary_loss_mlp": 0.01035176, + "balance_loss_clip": 1.00842333, + "balance_loss_mlp": 1.02829206, + "epoch": 0.9145674262009259, + "flos": 21068970428160.0, + "grad_norm": 3.3043352951765628, + "language_loss": 0.82666987, + "learning_rate": 7.598548559689777e-08, + "loss": 0.84838498, + "num_input_tokens_seen": 164491155, + "step": 7606, + "time_per_iteration": 3.628265857696533 + }, + { + "auxiliary_loss_clip": 0.01088788, + "auxiliary_loss_mlp": 0.01027581, + "balance_loss_clip": 0.93191755, + "balance_loss_mlp": 1.02020502, + "epoch": 0.914687669091565, + "flos": 16800269212800.0, + "grad_norm": 15.791253896091806, + "language_loss": 0.81067407, + "learning_rate": 7.577295367924751e-08, + "loss": 0.83183777, + "num_input_tokens_seen": 164507555, + "step": 7607, + "time_per_iteration": 2.7659690380096436 + }, + { + "auxiliary_loss_clip": 0.01124071, + "auxiliary_loss_mlp": 0.01024141, + "balance_loss_clip": 0.9739002, + "balance_loss_mlp": 1.01679778, + "epoch": 0.914807911982204, + "flos": 25773627012480.0, + "grad_norm": 1.6216961910894878, + "language_loss": 0.81989789, + "learning_rate": 7.556071366531002e-08, + "loss": 0.84138, + "num_input_tokens_seen": 164528525, + "step": 7608, + "time_per_iteration": 2.668389081954956 + }, + { + "auxiliary_loss_clip": 0.01141669, + "auxiliary_loss_mlp": 0.01028099, + "balance_loss_clip": 1.01217198, + "balance_loss_mlp": 1.02024865, + "epoch": 0.9149281548728432, + "flos": 19208043636480.0, + "grad_norm": 1.922178493935542, + "language_loss": 0.7896226, + "learning_rate": 7.53487655872822e-08, + "loss": 0.8113203, + "num_input_tokens_seen": 164547695, + "step": 7609, + "time_per_iteration": 2.6211347579956055 + }, + { + "auxiliary_loss_clip": 0.01073062, + "auxiliary_loss_mlp": 0.01023741, + "balance_loss_clip": 0.8931582, + "balance_loss_mlp": 1.01643336, + "epoch": 0.9150483977634822, + "flos": 26870554500480.0, + "grad_norm": 1.734849461237414, + "language_loss": 0.73831803, + "learning_rate": 7.513710947731656e-08, + "loss": 0.75928605, + "num_input_tokens_seen": 164568905, + "step": 7610, + "time_per_iteration": 2.7900216579437256 + }, + { + "auxiliary_loss_clip": 0.01109567, + "auxiliary_loss_mlp": 0.01030345, + "balance_loss_clip": 0.96946073, + "balance_loss_mlp": 1.02318358, + "epoch": 0.9151686406541213, + "flos": 21908956953600.0, + "grad_norm": 1.8259915494725862, + "language_loss": 0.85290694, + "learning_rate": 7.492574536752095e-08, + "loss": 0.87430608, + "num_input_tokens_seen": 164588895, + "step": 7611, + "time_per_iteration": 2.710099935531616 + }, + { + "auxiliary_loss_clip": 0.01135787, + "auxiliary_loss_mlp": 0.01026456, + "balance_loss_clip": 1.00918484, + "balance_loss_mlp": 1.02005792, + "epoch": 0.9152888835447605, + "flos": 27308556944640.0, + "grad_norm": 2.0906056665617223, + "language_loss": 0.78266025, + "learning_rate": 7.471467328995907e-08, + "loss": 0.80428267, + "num_input_tokens_seen": 164607705, + "step": 7612, + "time_per_iteration": 2.637317419052124 + }, + { + "auxiliary_loss_clip": 0.01008047, + "auxiliary_loss_mlp": 0.01025002, + "balance_loss_clip": 0.77776134, + "balance_loss_mlp": 1.01822209, + "epoch": 0.9154091264353995, + "flos": 13370728510080.0, + "grad_norm": 3.070523119103801, + "language_loss": 0.60797644, + "learning_rate": 7.450389327665018e-08, + "loss": 0.62830687, + "num_input_tokens_seen": 164625540, + "step": 7613, + "time_per_iteration": 4.06112813949585 + }, + { + "auxiliary_loss_clip": 0.01098929, + "auxiliary_loss_mlp": 0.01025155, + "balance_loss_clip": 0.93913007, + "balance_loss_mlp": 1.01802659, + "epoch": 0.9155293693260386, + "flos": 20193037367040.0, + "grad_norm": 2.7932556338083168, + "language_loss": 0.67482978, + "learning_rate": 7.429340535957029e-08, + "loss": 0.69607061, + "num_input_tokens_seen": 164640735, + "step": 7614, + "time_per_iteration": 2.9061036109924316 + }, + { + "auxiliary_loss_clip": 0.01120406, + "auxiliary_loss_mlp": 0.01023212, + "balance_loss_clip": 0.97161126, + "balance_loss_mlp": 1.01614869, + "epoch": 0.9156496122166777, + "flos": 19354990176000.0, + "grad_norm": 5.338504371739984, + "language_loss": 0.71185833, + "learning_rate": 7.40832095706494e-08, + "loss": 0.73329455, + "num_input_tokens_seen": 164657430, + "step": 7615, + "time_per_iteration": 2.6728882789611816 + }, + { + "auxiliary_loss_clip": 0.01097166, + "auxiliary_loss_mlp": 0.01026236, + "balance_loss_clip": 0.93158269, + "balance_loss_mlp": 1.01923585, + "epoch": 0.9157698551073168, + "flos": 21107287261440.0, + "grad_norm": 1.7489595530005613, + "language_loss": 0.80434561, + "learning_rate": 7.387330594177443e-08, + "loss": 0.82557964, + "num_input_tokens_seen": 164679505, + "step": 7616, + "time_per_iteration": 2.774228811264038 + }, + { + "auxiliary_loss_clip": 0.01087776, + "auxiliary_loss_mlp": 0.01025087, + "balance_loss_clip": 0.93271601, + "balance_loss_mlp": 1.01833057, + "epoch": 0.9158900979979558, + "flos": 25193167228800.0, + "grad_norm": 11.252495618931604, + "language_loss": 0.79465872, + "learning_rate": 7.366369450478749e-08, + "loss": 0.81578737, + "num_input_tokens_seen": 164700615, + "step": 7617, + "time_per_iteration": 2.705923557281494 + }, + { + "auxiliary_loss_clip": 0.01088102, + "auxiliary_loss_mlp": 0.01023602, + "balance_loss_clip": 0.93170834, + "balance_loss_mlp": 1.01695585, + "epoch": 0.916010340888595, + "flos": 30146648302080.0, + "grad_norm": 1.676118415723998, + "language_loss": 0.66119361, + "learning_rate": 7.345437529148646e-08, + "loss": 0.68231058, + "num_input_tokens_seen": 164719625, + "step": 7618, + "time_per_iteration": 2.8005869388580322 + }, + { + "auxiliary_loss_clip": 0.010949, + "auxiliary_loss_mlp": 0.01032179, + "balance_loss_clip": 0.93294334, + "balance_loss_mlp": 1.02531815, + "epoch": 0.9161305837792341, + "flos": 17091827907840.0, + "grad_norm": 2.080821628044406, + "language_loss": 0.72730398, + "learning_rate": 7.324534833362483e-08, + "loss": 0.74857473, + "num_input_tokens_seen": 164737200, + "step": 7619, + "time_per_iteration": 2.6691441535949707 + }, + { + "auxiliary_loss_clip": 0.01113652, + "auxiliary_loss_mlp": 0.01028306, + "balance_loss_clip": 0.96994323, + "balance_loss_mlp": 1.02136242, + "epoch": 0.9162508266698731, + "flos": 22893699288960.0, + "grad_norm": 1.7219148821513668, + "language_loss": 0.68486875, + "learning_rate": 7.303661366291192e-08, + "loss": 0.70628834, + "num_input_tokens_seen": 164757870, + "step": 7620, + "time_per_iteration": 2.7045998573303223 + }, + { + "auxiliary_loss_clip": 0.01071396, + "auxiliary_loss_mlp": 0.01027845, + "balance_loss_clip": 0.89554173, + "balance_loss_mlp": 1.02104378, + "epoch": 0.9163710695605123, + "flos": 19974808287360.0, + "grad_norm": 1.7949892330896555, + "language_loss": 0.82255751, + "learning_rate": 7.28281713110126e-08, + "loss": 0.84354997, + "num_input_tokens_seen": 164775945, + "step": 7621, + "time_per_iteration": 2.7295150756835938 + }, + { + "auxiliary_loss_clip": 0.01115831, + "auxiliary_loss_mlp": 0.01025599, + "balance_loss_clip": 0.97459137, + "balance_loss_mlp": 1.01849997, + "epoch": 0.9164913124511513, + "flos": 22783812606720.0, + "grad_norm": 1.8996395399158779, + "language_loss": 0.77331924, + "learning_rate": 7.262002130954759e-08, + "loss": 0.79473352, + "num_input_tokens_seen": 164794400, + "step": 7622, + "time_per_iteration": 2.635037660598755 + }, + { + "auxiliary_loss_clip": 0.01072572, + "auxiliary_loss_mlp": 0.01021776, + "balance_loss_clip": 0.89387178, + "balance_loss_mlp": 1.01477587, + "epoch": 0.9166115553417904, + "flos": 24900854348160.0, + "grad_norm": 1.8712574566738662, + "language_loss": 0.79038405, + "learning_rate": 7.241216369009296e-08, + "loss": 0.81132752, + "num_input_tokens_seen": 164814585, + "step": 7623, + "time_per_iteration": 2.792825937271118 + }, + { + "auxiliary_loss_clip": 0.01162476, + "auxiliary_loss_mlp": 0.01023093, + "balance_loss_clip": 1.04583657, + "balance_loss_mlp": 1.01600039, + "epoch": 0.9167317982324296, + "flos": 25702919089920.0, + "grad_norm": 2.507344691309345, + "language_loss": 0.66329193, + "learning_rate": 7.220459848418037e-08, + "loss": 0.68514764, + "num_input_tokens_seen": 164834660, + "step": 7624, + "time_per_iteration": 2.6164603233337402 + }, + { + "auxiliary_loss_clip": 0.01162974, + "auxiliary_loss_mlp": 0.01020992, + "balance_loss_clip": 1.04945016, + "balance_loss_mlp": 1.01410735, + "epoch": 0.9168520411230686, + "flos": 15632813370240.0, + "grad_norm": 2.352964614184113, + "language_loss": 0.79520619, + "learning_rate": 7.199732572329708e-08, + "loss": 0.81704587, + "num_input_tokens_seen": 164852560, + "step": 7625, + "time_per_iteration": 2.6354472637176514 + }, + { + "auxiliary_loss_clip": 0.0109563, + "auxiliary_loss_mlp": 0.01027187, + "balance_loss_clip": 0.93187779, + "balance_loss_mlp": 1.0201031, + "epoch": 0.9169722840137077, + "flos": 30258151096320.0, + "grad_norm": 2.1458465563249325, + "language_loss": 0.75629151, + "learning_rate": 7.179034543888684e-08, + "loss": 0.7775197, + "num_input_tokens_seen": 164872065, + "step": 7626, + "time_per_iteration": 2.789332389831543 + }, + { + "auxiliary_loss_clip": 0.01140655, + "auxiliary_loss_mlp": 0.01024301, + "balance_loss_clip": 1.00718498, + "balance_loss_mlp": 1.01767278, + "epoch": 0.9170925269043467, + "flos": 22491643380480.0, + "grad_norm": 2.413594807097117, + "language_loss": 0.77717507, + "learning_rate": 7.158365766234808e-08, + "loss": 0.79882461, + "num_input_tokens_seen": 164890915, + "step": 7627, + "time_per_iteration": 2.6542556285858154 + }, + { + "auxiliary_loss_clip": 0.01084002, + "auxiliary_loss_mlp": 0.0102439, + "balance_loss_clip": 0.92647856, + "balance_loss_mlp": 1.0172255, + "epoch": 0.9172127697949859, + "flos": 22893914770560.0, + "grad_norm": 2.1676937343875546, + "language_loss": 0.72634292, + "learning_rate": 7.137726242503527e-08, + "loss": 0.74742675, + "num_input_tokens_seen": 164909835, + "step": 7628, + "time_per_iteration": 4.039317846298218 + }, + { + "auxiliary_loss_clip": 0.01138865, + "auxiliary_loss_mlp": 0.00509035, + "balance_loss_clip": 1.00898826, + "balance_loss_mlp": 0.0, + "epoch": 0.917333012685625, + "flos": 17451867882240.0, + "grad_norm": 2.9393512131493087, + "language_loss": 0.78333986, + "learning_rate": 7.11711597582585e-08, + "loss": 0.79981887, + "num_input_tokens_seen": 164927195, + "step": 7629, + "time_per_iteration": 2.6371006965637207 + }, + { + "auxiliary_loss_clip": 0.01093846, + "auxiliary_loss_mlp": 0.01027046, + "balance_loss_clip": 0.92964554, + "balance_loss_mlp": 1.01969683, + "epoch": 0.917453255576264, + "flos": 14318949692160.0, + "grad_norm": 2.0010613606609073, + "language_loss": 0.80172479, + "learning_rate": 7.096534969328271e-08, + "loss": 0.82293373, + "num_input_tokens_seen": 164944640, + "step": 7630, + "time_per_iteration": 3.645413875579834 + }, + { + "auxiliary_loss_clip": 0.01118218, + "auxiliary_loss_mlp": 0.01025722, + "balance_loss_clip": 0.96681118, + "balance_loss_mlp": 1.019171, + "epoch": 0.9175734984669032, + "flos": 20741177888640.0, + "grad_norm": 1.9413133655724488, + "language_loss": 0.84185255, + "learning_rate": 7.075983226132987e-08, + "loss": 0.86329198, + "num_input_tokens_seen": 164963570, + "step": 7631, + "time_per_iteration": 2.682028293609619 + }, + { + "auxiliary_loss_clip": 0.01121445, + "auxiliary_loss_mlp": 0.00509806, + "balance_loss_clip": 0.97097051, + "balance_loss_mlp": 0.0, + "epoch": 0.9176937413575422, + "flos": 14830497233280.0, + "grad_norm": 2.519910725881686, + "language_loss": 0.79410231, + "learning_rate": 7.055460749357656e-08, + "loss": 0.81041479, + "num_input_tokens_seen": 164979850, + "step": 7632, + "time_per_iteration": 3.5951099395751953 + }, + { + "auxiliary_loss_clip": 0.01116865, + "auxiliary_loss_mlp": 0.01027799, + "balance_loss_clip": 0.97167534, + "balance_loss_mlp": 1.0208075, + "epoch": 0.9178139842481813, + "flos": 18474603828480.0, + "grad_norm": 1.829839769861458, + "language_loss": 0.70613921, + "learning_rate": 7.034967542115521e-08, + "loss": 0.72758591, + "num_input_tokens_seen": 164998115, + "step": 7633, + "time_per_iteration": 2.6737799644470215 + }, + { + "auxiliary_loss_clip": 0.01130799, + "auxiliary_loss_mlp": 0.00509112, + "balance_loss_clip": 1.00647795, + "balance_loss_mlp": 0.0, + "epoch": 0.9179342271388204, + "flos": 20047455544320.0, + "grad_norm": 2.2010443727325484, + "language_loss": 0.75299382, + "learning_rate": 7.014503607515388e-08, + "loss": 0.76939303, + "num_input_tokens_seen": 165017420, + "step": 7634, + "time_per_iteration": 2.64872407913208 + }, + { + "auxiliary_loss_clip": 0.01118882, + "auxiliary_loss_mlp": 0.01034121, + "balance_loss_clip": 0.97378385, + "balance_loss_mlp": 1.02731383, + "epoch": 0.9180544700294595, + "flos": 24676232647680.0, + "grad_norm": 2.077188737695757, + "language_loss": 0.68390715, + "learning_rate": 6.994068948661592e-08, + "loss": 0.70543718, + "num_input_tokens_seen": 165035575, + "step": 7635, + "time_per_iteration": 2.722825527191162 + }, + { + "auxiliary_loss_clip": 0.01143207, + "auxiliary_loss_mlp": 0.01025158, + "balance_loss_clip": 1.0116415, + "balance_loss_mlp": 1.01767719, + "epoch": 0.9181747129200986, + "flos": 16727478301440.0, + "grad_norm": 2.035231109137917, + "language_loss": 0.77052832, + "learning_rate": 6.973663568654142e-08, + "loss": 0.79221195, + "num_input_tokens_seen": 165053280, + "step": 7636, + "time_per_iteration": 2.6081345081329346 + }, + { + "auxiliary_loss_clip": 0.01162781, + "auxiliary_loss_mlp": 0.01021667, + "balance_loss_clip": 1.04831076, + "balance_loss_mlp": 1.01479459, + "epoch": 0.9182949558107377, + "flos": 24271626873600.0, + "grad_norm": 2.279546780512856, + "language_loss": 0.65235025, + "learning_rate": 6.953287470588386e-08, + "loss": 0.67419475, + "num_input_tokens_seen": 165071235, + "step": 7637, + "time_per_iteration": 2.629446029663086 + }, + { + "auxiliary_loss_clip": 0.01143749, + "auxiliary_loss_mlp": 0.01025011, + "balance_loss_clip": 1.00848317, + "balance_loss_mlp": 1.0177933, + "epoch": 0.9184151987013768, + "flos": 22082117443200.0, + "grad_norm": 2.114286637760067, + "language_loss": 0.86047912, + "learning_rate": 6.932940657555452e-08, + "loss": 0.88216668, + "num_input_tokens_seen": 165087365, + "step": 7638, + "time_per_iteration": 2.616373062133789 + }, + { + "auxiliary_loss_clip": 0.01158936, + "auxiliary_loss_mlp": 0.01026545, + "balance_loss_clip": 1.04649031, + "balance_loss_mlp": 1.02050376, + "epoch": 0.9185354415920158, + "flos": 32166732257280.0, + "grad_norm": 1.84499161220336, + "language_loss": 0.7622931, + "learning_rate": 6.912623132641938e-08, + "loss": 0.78414786, + "num_input_tokens_seen": 165112455, + "step": 7639, + "time_per_iteration": 3.699871778488159 + }, + { + "auxiliary_loss_clip": 0.01118199, + "auxiliary_loss_mlp": 0.01022789, + "balance_loss_clip": 0.970083, + "balance_loss_mlp": 1.015481, + "epoch": 0.918655684482655, + "flos": 20997831542400.0, + "grad_norm": 2.3038900326771543, + "language_loss": 0.76959795, + "learning_rate": 6.892334898929952e-08, + "loss": 0.79100782, + "num_input_tokens_seen": 165132700, + "step": 7640, + "time_per_iteration": 2.659329652786255 + }, + { + "auxiliary_loss_clip": 0.01135988, + "auxiliary_loss_mlp": 0.01022428, + "balance_loss_clip": 1.00893891, + "balance_loss_mlp": 1.01534128, + "epoch": 0.918775927373294, + "flos": 15560704817280.0, + "grad_norm": 2.260663230472614, + "language_loss": 0.84726411, + "learning_rate": 6.872075959497236e-08, + "loss": 0.86884826, + "num_input_tokens_seen": 165151475, + "step": 7641, + "time_per_iteration": 2.6330060958862305 + }, + { + "auxiliary_loss_clip": 0.01142966, + "auxiliary_loss_mlp": 0.01024378, + "balance_loss_clip": 1.00767136, + "balance_loss_mlp": 1.01746368, + "epoch": 0.9188961702639331, + "flos": 29934057657600.0, + "grad_norm": 2.1605299766543564, + "language_loss": 0.82651979, + "learning_rate": 6.85184631741702e-08, + "loss": 0.84819329, + "num_input_tokens_seen": 165172040, + "step": 7642, + "time_per_iteration": 2.7235870361328125 + }, + { + "auxiliary_loss_clip": 0.01137392, + "auxiliary_loss_mlp": 0.01024767, + "balance_loss_clip": 1.0078584, + "balance_loss_mlp": 1.01796651, + "epoch": 0.9190164131545723, + "flos": 20701244943360.0, + "grad_norm": 1.9393852097839435, + "language_loss": 0.77168691, + "learning_rate": 6.831645975758161e-08, + "loss": 0.7933085, + "num_input_tokens_seen": 165189980, + "step": 7643, + "time_per_iteration": 2.649827718734741 + }, + { + "auxiliary_loss_clip": 0.01109879, + "auxiliary_loss_mlp": 0.01028344, + "balance_loss_clip": 0.97110653, + "balance_loss_mlp": 1.02109325, + "epoch": 0.9191366560452113, + "flos": 25629912696960.0, + "grad_norm": 2.2141495134098386, + "language_loss": 0.6773271, + "learning_rate": 6.811474937585026e-08, + "loss": 0.69870937, + "num_input_tokens_seen": 165209770, + "step": 7644, + "time_per_iteration": 2.70404052734375 + }, + { + "auxiliary_loss_clip": 0.0109118, + "auxiliary_loss_mlp": 0.01020943, + "balance_loss_clip": 0.93387544, + "balance_loss_mlp": 1.0142765, + "epoch": 0.9192568989358504, + "flos": 21434325615360.0, + "grad_norm": 1.4794856059956152, + "language_loss": 0.79018927, + "learning_rate": 6.79133320595755e-08, + "loss": 0.81131053, + "num_input_tokens_seen": 165229690, + "step": 7645, + "time_per_iteration": 2.7141830921173096 + }, + { + "auxiliary_loss_clip": 0.01124017, + "auxiliary_loss_mlp": 0.01025968, + "balance_loss_clip": 0.9764601, + "balance_loss_mlp": 1.01901782, + "epoch": 0.9193771418264896, + "flos": 23185078416000.0, + "grad_norm": 1.7445936170956595, + "language_loss": 0.7565167, + "learning_rate": 6.771220783931198e-08, + "loss": 0.77801651, + "num_input_tokens_seen": 165249850, + "step": 7646, + "time_per_iteration": 2.6965935230255127 + }, + { + "auxiliary_loss_clip": 0.00942891, + "auxiliary_loss_mlp": 0.0050146, + "balance_loss_clip": 0.77649438, + "balance_loss_mlp": 0.0, + "epoch": 0.9194973847171286, + "flos": 70582963184640.0, + "grad_norm": 0.8256493401384498, + "language_loss": 0.64568299, + "learning_rate": 6.751137674556994e-08, + "loss": 0.66012651, + "num_input_tokens_seen": 165310235, + "step": 7647, + "time_per_iteration": 3.737841844558716 + }, + { + "auxiliary_loss_clip": 0.01143761, + "auxiliary_loss_mlp": 0.0102123, + "balance_loss_clip": 1.00817394, + "balance_loss_mlp": 1.01408374, + "epoch": 0.9196176276077677, + "flos": 14720682378240.0, + "grad_norm": 2.9172727590354257, + "language_loss": 0.77687383, + "learning_rate": 6.731083880881572e-08, + "loss": 0.79852372, + "num_input_tokens_seen": 165326455, + "step": 7648, + "time_per_iteration": 3.410243034362793 + }, + { + "auxiliary_loss_clip": 0.0111766, + "auxiliary_loss_mlp": 0.0102515, + "balance_loss_clip": 0.97176862, + "balance_loss_mlp": 1.01827145, + "epoch": 0.9197378704984068, + "flos": 23294893271040.0, + "grad_norm": 2.090243024001213, + "language_loss": 0.81098402, + "learning_rate": 6.711059405947072e-08, + "loss": 0.83241212, + "num_input_tokens_seen": 165344645, + "step": 7649, + "time_per_iteration": 2.6963722705841064 + }, + { + "auxiliary_loss_clip": 0.01090184, + "auxiliary_loss_mlp": 0.01023727, + "balance_loss_clip": 0.93170112, + "balance_loss_mlp": 1.01688695, + "epoch": 0.9198581133890459, + "flos": 20302564913280.0, + "grad_norm": 1.7670308032739335, + "language_loss": 0.77103126, + "learning_rate": 6.691064252791156e-08, + "loss": 0.79217041, + "num_input_tokens_seen": 165364120, + "step": 7650, + "time_per_iteration": 2.692042112350464 + }, + { + "auxiliary_loss_clip": 0.01065695, + "auxiliary_loss_mlp": 0.01022934, + "balance_loss_clip": 0.89386982, + "balance_loss_mlp": 1.01588273, + "epoch": 0.9199783562796849, + "flos": 17675663569920.0, + "grad_norm": 4.273599776960814, + "language_loss": 0.77992249, + "learning_rate": 6.67109842444713e-08, + "loss": 0.80080879, + "num_input_tokens_seen": 165383050, + "step": 7651, + "time_per_iteration": 2.6971261501312256 + }, + { + "auxiliary_loss_clip": 0.01141582, + "auxiliary_loss_mlp": 0.00509277, + "balance_loss_clip": 1.01405978, + "balance_loss_mlp": 0.0, + "epoch": 0.9200985991703241, + "flos": 17676022705920.0, + "grad_norm": 1.8602645672779679, + "language_loss": 0.76303536, + "learning_rate": 6.651161923943704e-08, + "loss": 0.77954394, + "num_input_tokens_seen": 165400955, + "step": 7652, + "time_per_iteration": 2.5636065006256104 + }, + { + "auxiliary_loss_clip": 0.01135006, + "auxiliary_loss_mlp": 0.01026625, + "balance_loss_clip": 1.00666177, + "balance_loss_mlp": 1.01956749, + "epoch": 0.9202188420609632, + "flos": 20996574566400.0, + "grad_norm": 1.7716209170359905, + "language_loss": 0.76832551, + "learning_rate": 6.631254754305326e-08, + "loss": 0.78994179, + "num_input_tokens_seen": 165420415, + "step": 7653, + "time_per_iteration": 2.666980028152466 + }, + { + "auxiliary_loss_clip": 0.01165497, + "auxiliary_loss_mlp": 0.01026164, + "balance_loss_clip": 1.04763556, + "balance_loss_mlp": 1.01867795, + "epoch": 0.9203390849516022, + "flos": 13918222586880.0, + "grad_norm": 1.8267884782513708, + "language_loss": 0.78188431, + "learning_rate": 6.611376918551848e-08, + "loss": 0.80380094, + "num_input_tokens_seen": 165439200, + "step": 7654, + "time_per_iteration": 2.591928243637085 + }, + { + "auxiliary_loss_clip": 0.01092272, + "auxiliary_loss_mlp": 0.00509507, + "balance_loss_clip": 0.93198371, + "balance_loss_mlp": 0.0, + "epoch": 0.9204593278422414, + "flos": 21175912195200.0, + "grad_norm": 3.378694479391627, + "language_loss": 0.79819989, + "learning_rate": 6.591528419698744e-08, + "loss": 0.81421769, + "num_input_tokens_seen": 165458985, + "step": 7655, + "time_per_iteration": 3.57797908782959 + }, + { + "auxiliary_loss_clip": 0.01116759, + "auxiliary_loss_mlp": 0.0101981, + "balance_loss_clip": 0.96808398, + "balance_loss_mlp": 1.01297033, + "epoch": 0.9205795707328804, + "flos": 14501375890560.0, + "grad_norm": 2.3593697884734945, + "language_loss": 0.83739018, + "learning_rate": 6.571709260756986e-08, + "loss": 0.85875583, + "num_input_tokens_seen": 165475630, + "step": 7656, + "time_per_iteration": 3.5576021671295166 + }, + { + "auxiliary_loss_clip": 0.01143824, + "auxiliary_loss_mlp": 0.01029774, + "balance_loss_clip": 1.01299238, + "balance_loss_mlp": 1.02238321, + "epoch": 0.9206998136235195, + "flos": 22417559579520.0, + "grad_norm": 2.2469046618426276, + "language_loss": 0.76232195, + "learning_rate": 6.551919444733122e-08, + "loss": 0.78405797, + "num_input_tokens_seen": 165493445, + "step": 7657, + "time_per_iteration": 3.538369655609131 + }, + { + "auxiliary_loss_clip": 0.01116008, + "auxiliary_loss_mlp": 0.01025314, + "balance_loss_clip": 0.97202146, + "balance_loss_mlp": 1.0182302, + "epoch": 0.9208200565141585, + "flos": 53358407544960.0, + "grad_norm": 1.7365925967789608, + "language_loss": 0.65669549, + "learning_rate": 6.53215897462931e-08, + "loss": 0.67810869, + "num_input_tokens_seen": 165517200, + "step": 7658, + "time_per_iteration": 2.9707865715026855 + }, + { + "auxiliary_loss_clip": 0.01139094, + "auxiliary_loss_mlp": 0.01027563, + "balance_loss_clip": 1.00928915, + "balance_loss_mlp": 1.02094126, + "epoch": 0.9209402994047977, + "flos": 30589139946240.0, + "grad_norm": 2.0305926239400076, + "language_loss": 0.750229, + "learning_rate": 6.512427853443103e-08, + "loss": 0.77189559, + "num_input_tokens_seen": 165539280, + "step": 7659, + "time_per_iteration": 2.7315571308135986 + }, + { + "auxiliary_loss_clip": 0.01143506, + "auxiliary_loss_mlp": 0.01028401, + "balance_loss_clip": 1.00985444, + "balance_loss_mlp": 1.02160048, + "epoch": 0.9210605422954368, + "flos": 29132711187840.0, + "grad_norm": 1.5451415582509633, + "language_loss": 0.76101398, + "learning_rate": 6.492726084167799e-08, + "loss": 0.78273302, + "num_input_tokens_seen": 165561395, + "step": 7660, + "time_per_iteration": 2.730545997619629 + }, + { + "auxiliary_loss_clip": 0.01059469, + "auxiliary_loss_mlp": 0.00999487, + "balance_loss_clip": 1.00895917, + "balance_loss_mlp": 0.99855709, + "epoch": 0.9211807851860758, + "flos": 54853838472960.0, + "grad_norm": 0.7770802776525381, + "language_loss": 0.57464159, + "learning_rate": 6.473053669792072e-08, + "loss": 0.59523118, + "num_input_tokens_seen": 165616085, + "step": 7661, + "time_per_iteration": 3.081465721130371 + }, + { + "auxiliary_loss_clip": 0.01140148, + "auxiliary_loss_mlp": 0.010283, + "balance_loss_clip": 1.00906193, + "balance_loss_mlp": 1.02117443, + "epoch": 0.921301028076715, + "flos": 19201974238080.0, + "grad_norm": 2.0441531090596072, + "language_loss": 0.72927344, + "learning_rate": 6.453410613300248e-08, + "loss": 0.75095797, + "num_input_tokens_seen": 165634015, + "step": 7662, + "time_per_iteration": 2.6142640113830566 + }, + { + "auxiliary_loss_clip": 0.01048766, + "auxiliary_loss_mlp": 0.01027702, + "balance_loss_clip": 0.85658556, + "balance_loss_mlp": 1.02052283, + "epoch": 0.921421270967354, + "flos": 27526893765120.0, + "grad_norm": 1.615913086437705, + "language_loss": 0.58554006, + "learning_rate": 6.43379691767214e-08, + "loss": 0.60630476, + "num_input_tokens_seen": 165653220, + "step": 7663, + "time_per_iteration": 2.885392189025879 + }, + { + "auxiliary_loss_clip": 0.00982061, + "auxiliary_loss_mlp": 0.00999561, + "balance_loss_clip": 0.86169839, + "balance_loss_mlp": 0.9986254, + "epoch": 0.9215415138579931, + "flos": 70209311955840.0, + "grad_norm": 0.7186816780562217, + "language_loss": 0.55138773, + "learning_rate": 6.414212585883105e-08, + "loss": 0.57120395, + "num_input_tokens_seen": 165715850, + "step": 7664, + "time_per_iteration": 3.366678476333618 + }, + { + "auxiliary_loss_clip": 0.01121695, + "auxiliary_loss_mlp": 0.01028466, + "balance_loss_clip": 0.97292048, + "balance_loss_mlp": 1.02128959, + "epoch": 0.9216617567486323, + "flos": 35553107790720.0, + "grad_norm": 1.4406920886470207, + "language_loss": 0.69997919, + "learning_rate": 6.394657620904143e-08, + "loss": 0.72148085, + "num_input_tokens_seen": 165738960, + "step": 7665, + "time_per_iteration": 3.652472496032715 + }, + { + "auxiliary_loss_clip": 0.0116907, + "auxiliary_loss_mlp": 0.01026233, + "balance_loss_clip": 1.05057406, + "balance_loss_mlp": 1.01937509, + "epoch": 0.9217819996392713, + "flos": 29533330552320.0, + "grad_norm": 1.7208745993142993, + "language_loss": 0.71520257, + "learning_rate": 6.375132025701657e-08, + "loss": 0.73715556, + "num_input_tokens_seen": 165761260, + "step": 7666, + "time_per_iteration": 2.6760170459747314 + }, + { + "auxiliary_loss_clip": 0.01167143, + "auxiliary_loss_mlp": 0.01032965, + "balance_loss_clip": 1.05023575, + "balance_loss_mlp": 1.02538943, + "epoch": 0.9219022425299104, + "flos": 14574669592320.0, + "grad_norm": 2.4237254271497095, + "language_loss": 0.6946913, + "learning_rate": 6.355635803237724e-08, + "loss": 0.71669239, + "num_input_tokens_seen": 165776960, + "step": 7667, + "time_per_iteration": 2.59016489982605 + }, + { + "auxiliary_loss_clip": 0.01138622, + "auxiliary_loss_mlp": 0.01027037, + "balance_loss_clip": 1.00828314, + "balance_loss_mlp": 1.02001286, + "epoch": 0.9220224854205495, + "flos": 18077503996800.0, + "grad_norm": 4.22267780922592, + "language_loss": 0.79594016, + "learning_rate": 6.336168956469867e-08, + "loss": 0.81759673, + "num_input_tokens_seen": 165795435, + "step": 7668, + "time_per_iteration": 2.687650442123413 + }, + { + "auxiliary_loss_clip": 0.01110476, + "auxiliary_loss_mlp": 0.01024605, + "balance_loss_clip": 0.97080052, + "balance_loss_mlp": 1.01787531, + "epoch": 0.9221427283111886, + "flos": 24790464875520.0, + "grad_norm": 1.78820435780952, + "language_loss": 0.71652961, + "learning_rate": 6.316731488351168e-08, + "loss": 0.73788041, + "num_input_tokens_seen": 165816625, + "step": 7669, + "time_per_iteration": 2.7127327919006348 + }, + { + "auxiliary_loss_clip": 0.01140803, + "auxiliary_loss_mlp": 0.01023029, + "balance_loss_clip": 1.01013315, + "balance_loss_mlp": 1.01570356, + "epoch": 0.9222629712018277, + "flos": 13845036625920.0, + "grad_norm": 1.8319069845249496, + "language_loss": 0.63319993, + "learning_rate": 6.297323401830334e-08, + "loss": 0.65483832, + "num_input_tokens_seen": 165835410, + "step": 7670, + "time_per_iteration": 2.6803829669952393 + }, + { + "auxiliary_loss_clip": 0.01141436, + "auxiliary_loss_mlp": 0.01027985, + "balance_loss_clip": 1.00784254, + "balance_loss_mlp": 1.02084422, + "epoch": 0.9223832140924668, + "flos": 21616177196160.0, + "grad_norm": 1.9721228209333843, + "language_loss": 0.6896081, + "learning_rate": 6.277944699851523e-08, + "loss": 0.71130228, + "num_input_tokens_seen": 165854930, + "step": 7671, + "time_per_iteration": 2.6932828426361084 + }, + { + "auxiliary_loss_clip": 0.01164538, + "auxiliary_loss_mlp": 0.0102674, + "balance_loss_clip": 1.04799151, + "balance_loss_mlp": 1.01958132, + "epoch": 0.9225034569831059, + "flos": 21142084561920.0, + "grad_norm": 2.598941241705381, + "language_loss": 0.7320745, + "learning_rate": 6.25859538535447e-08, + "loss": 0.75398731, + "num_input_tokens_seen": 165875725, + "step": 7672, + "time_per_iteration": 2.6975104808807373 + }, + { + "auxiliary_loss_clip": 0.01116727, + "auxiliary_loss_mlp": 0.01024323, + "balance_loss_clip": 0.97224581, + "balance_loss_mlp": 1.01713133, + "epoch": 0.9226236998737449, + "flos": 12495046844160.0, + "grad_norm": 2.4635071048020043, + "language_loss": 0.78246593, + "learning_rate": 6.239275461274474e-08, + "loss": 0.8038764, + "num_input_tokens_seen": 165892100, + "step": 7673, + "time_per_iteration": 2.6604442596435547 + }, + { + "auxiliary_loss_clip": 0.01141206, + "auxiliary_loss_mlp": 0.01027361, + "balance_loss_clip": 1.00944877, + "balance_loss_mlp": 1.02054846, + "epoch": 0.9227439427643841, + "flos": 26214071581440.0, + "grad_norm": 1.8511154228918885, + "language_loss": 0.86021423, + "learning_rate": 6.219984930542299e-08, + "loss": 0.88189995, + "num_input_tokens_seen": 165912840, + "step": 7674, + "time_per_iteration": 2.6699397563934326 + }, + { + "auxiliary_loss_clip": 0.01144199, + "auxiliary_loss_mlp": 0.0102446, + "balance_loss_clip": 1.01010478, + "balance_loss_mlp": 1.01720047, + "epoch": 0.9228641856550232, + "flos": 17967581400960.0, + "grad_norm": 2.113878361095277, + "language_loss": 0.75782835, + "learning_rate": 6.200723796084383e-08, + "loss": 0.77951497, + "num_input_tokens_seen": 165930935, + "step": 7675, + "time_per_iteration": 2.6153149604797363 + }, + { + "auxiliary_loss_clip": 0.01004672, + "auxiliary_loss_mlp": 0.00999879, + "balance_loss_clip": 0.89841658, + "balance_loss_mlp": 0.99893171, + "epoch": 0.9229844285456622, + "flos": 70420609710720.0, + "grad_norm": 0.778610019327042, + "language_loss": 0.630252, + "learning_rate": 6.181492060822546e-08, + "loss": 0.65029752, + "num_input_tokens_seen": 165991110, + "step": 7676, + "time_per_iteration": 3.1513004302978516 + }, + { + "auxiliary_loss_clip": 0.01065308, + "auxiliary_loss_mlp": 0.01026153, + "balance_loss_clip": 0.89148569, + "balance_loss_mlp": 1.0191555, + "epoch": 0.9231046714363014, + "flos": 17967832796160.0, + "grad_norm": 3.1653942746574146, + "language_loss": 0.81938684, + "learning_rate": 6.162289727674274e-08, + "loss": 0.84030145, + "num_input_tokens_seen": 166008790, + "step": 7677, + "time_per_iteration": 2.723081588745117 + }, + { + "auxiliary_loss_clip": 0.01093424, + "auxiliary_loss_mlp": 0.01024507, + "balance_loss_clip": 0.93272388, + "balance_loss_mlp": 1.01773906, + "epoch": 0.9232249143269404, + "flos": 17858233422720.0, + "grad_norm": 2.1231708056021033, + "language_loss": 0.8794229, + "learning_rate": 6.143116799552527e-08, + "loss": 0.90060222, + "num_input_tokens_seen": 166025035, + "step": 7678, + "time_per_iteration": 2.7745840549468994 + }, + { + "auxiliary_loss_clip": 0.01143736, + "auxiliary_loss_mlp": 0.01031147, + "balance_loss_clip": 1.0104655, + "balance_loss_mlp": 1.02388144, + "epoch": 0.9233451572175795, + "flos": 23404384903680.0, + "grad_norm": 2.2407038277060582, + "language_loss": 0.56186259, + "learning_rate": 6.123973279365802e-08, + "loss": 0.58361143, + "num_input_tokens_seen": 166044010, + "step": 7679, + "time_per_iteration": 2.6678268909454346 + }, + { + "auxiliary_loss_clip": 0.01145048, + "auxiliary_loss_mlp": 0.01025279, + "balance_loss_clip": 1.0113821, + "balance_loss_mlp": 1.01840997, + "epoch": 0.9234654001082186, + "flos": 17999326045440.0, + "grad_norm": 1.8267869503358012, + "language_loss": 0.77772272, + "learning_rate": 6.10485917001824e-08, + "loss": 0.79942602, + "num_input_tokens_seen": 166061865, + "step": 7680, + "time_per_iteration": 2.609283208847046 + }, + { + "auxiliary_loss_clip": 0.01120053, + "auxiliary_loss_mlp": 0.01025868, + "balance_loss_clip": 0.96921122, + "balance_loss_mlp": 1.01913309, + "epoch": 0.9235856429988577, + "flos": 24750747411840.0, + "grad_norm": 1.5246422837399285, + "language_loss": 0.81023061, + "learning_rate": 6.085774474409322e-08, + "loss": 0.83168983, + "num_input_tokens_seen": 166082425, + "step": 7681, + "time_per_iteration": 3.586912155151367 + }, + { + "auxiliary_loss_clip": 0.01117656, + "auxiliary_loss_mlp": 0.01026862, + "balance_loss_clip": 0.97448069, + "balance_loss_mlp": 1.01986468, + "epoch": 0.9237058858894968, + "flos": 14099894599680.0, + "grad_norm": 1.9383137407090951, + "language_loss": 0.70104194, + "learning_rate": 6.066719195434267e-08, + "loss": 0.72248709, + "num_input_tokens_seen": 166100225, + "step": 7682, + "time_per_iteration": 3.611557960510254 + }, + { + "auxiliary_loss_clip": 0.01145048, + "auxiliary_loss_mlp": 0.01029283, + "balance_loss_clip": 1.01047051, + "balance_loss_mlp": 1.02134347, + "epoch": 0.9238261287801359, + "flos": 28694529175680.0, + "grad_norm": 2.6125565979604017, + "language_loss": 0.66756678, + "learning_rate": 6.047693335983717e-08, + "loss": 0.68931007, + "num_input_tokens_seen": 166122570, + "step": 7683, + "time_per_iteration": 3.718804359436035 + }, + { + "auxiliary_loss_clip": 0.01141792, + "auxiliary_loss_mlp": 0.01026117, + "balance_loss_clip": 1.00885034, + "balance_loss_mlp": 1.01867199, + "epoch": 0.923946371670775, + "flos": 23111856541440.0, + "grad_norm": 2.4296203950050894, + "language_loss": 0.82807362, + "learning_rate": 6.028696898943853e-08, + "loss": 0.84975278, + "num_input_tokens_seen": 166141630, + "step": 7684, + "time_per_iteration": 2.6365952491760254 + }, + { + "auxiliary_loss_clip": 0.01115535, + "auxiliary_loss_mlp": 0.00509753, + "balance_loss_clip": 0.96820045, + "balance_loss_mlp": 0.0, + "epoch": 0.924066614561414, + "flos": 21867120587520.0, + "grad_norm": 2.7696880761278146, + "language_loss": 0.71055478, + "learning_rate": 6.00972988719648e-08, + "loss": 0.72680765, + "num_input_tokens_seen": 166159865, + "step": 7685, + "time_per_iteration": 2.675607204437256 + }, + { + "auxiliary_loss_clip": 0.0109666, + "auxiliary_loss_mlp": 0.00509061, + "balance_loss_clip": 0.93300223, + "balance_loss_mlp": 0.0, + "epoch": 0.9241868574520532, + "flos": 28511887495680.0, + "grad_norm": 2.2848655612915585, + "language_loss": 0.71075004, + "learning_rate": 5.990792303618807e-08, + "loss": 0.7268073, + "num_input_tokens_seen": 166179445, + "step": 7686, + "time_per_iteration": 2.809413194656372 + }, + { + "auxiliary_loss_clip": 0.01096402, + "auxiliary_loss_mlp": 0.01029398, + "balance_loss_clip": 0.93708098, + "balance_loss_mlp": 1.02270734, + "epoch": 0.9243071003426923, + "flos": 30518324282880.0, + "grad_norm": 1.6146724657787173, + "language_loss": 0.69498491, + "learning_rate": 5.971884151083695e-08, + "loss": 0.71624291, + "num_input_tokens_seen": 166201855, + "step": 7687, + "time_per_iteration": 2.8138790130615234 + }, + { + "auxiliary_loss_clip": 0.01121304, + "auxiliary_loss_mlp": 0.01027316, + "balance_loss_clip": 0.97244728, + "balance_loss_mlp": 1.02033305, + "epoch": 0.9244273432333313, + "flos": 28658331244800.0, + "grad_norm": 1.7501528999058142, + "language_loss": 0.74416757, + "learning_rate": 5.9530054324595124e-08, + "loss": 0.76565373, + "num_input_tokens_seen": 166221970, + "step": 7688, + "time_per_iteration": 2.7756216526031494 + }, + { + "auxiliary_loss_clip": 0.01035329, + "auxiliary_loss_mlp": 0.00500775, + "balance_loss_clip": 0.97161591, + "balance_loss_mlp": 0.0, + "epoch": 0.9245475861239704, + "flos": 66230589237120.0, + "grad_norm": 0.7175613476102292, + "language_loss": 0.5752362, + "learning_rate": 5.934156150610103e-08, + "loss": 0.59059727, + "num_input_tokens_seen": 166279335, + "step": 7689, + "time_per_iteration": 3.2425537109375 + }, + { + "auxiliary_loss_clip": 0.01114607, + "auxiliary_loss_mlp": 0.01028393, + "balance_loss_clip": 0.97021568, + "balance_loss_mlp": 1.02073407, + "epoch": 0.9246678290146095, + "flos": 24239918142720.0, + "grad_norm": 2.464044884667216, + "language_loss": 0.78933108, + "learning_rate": 5.915336308394914e-08, + "loss": 0.81076109, + "num_input_tokens_seen": 166298170, + "step": 7690, + "time_per_iteration": 2.663135528564453 + }, + { + "auxiliary_loss_clip": 0.01136731, + "auxiliary_loss_mlp": 0.0102274, + "balance_loss_clip": 1.00933909, + "balance_loss_mlp": 1.01558137, + "epoch": 0.9247880719052486, + "flos": 18988808976000.0, + "grad_norm": 1.6757199182304565, + "language_loss": 0.77104759, + "learning_rate": 5.89654590866886e-08, + "loss": 0.7926423, + "num_input_tokens_seen": 166317670, + "step": 7691, + "time_per_iteration": 3.540916681289673 + }, + { + "auxiliary_loss_clip": 0.01056481, + "auxiliary_loss_mlp": 0.01026117, + "balance_loss_clip": 0.86374336, + "balance_loss_mlp": 1.01890767, + "epoch": 0.9249083147958876, + "flos": 24024095274240.0, + "grad_norm": 2.6291755832889403, + "language_loss": 0.88137335, + "learning_rate": 5.877784954282483e-08, + "loss": 0.90219939, + "num_input_tokens_seen": 166337010, + "step": 7692, + "time_per_iteration": 2.8890633583068848 + }, + { + "auxiliary_loss_clip": 0.0114322, + "auxiliary_loss_mlp": 0.0102462, + "balance_loss_clip": 1.00952458, + "balance_loss_mlp": 1.01710379, + "epoch": 0.9250285576865268, + "flos": 30773972355840.0, + "grad_norm": 2.0898099728665147, + "language_loss": 0.72562057, + "learning_rate": 5.8590534480817963e-08, + "loss": 0.74729896, + "num_input_tokens_seen": 166358735, + "step": 7693, + "time_per_iteration": 2.7635629177093506 + }, + { + "auxiliary_loss_clip": 0.0116535, + "auxiliary_loss_mlp": 0.01030214, + "balance_loss_clip": 1.04889572, + "balance_loss_mlp": 1.02345502, + "epoch": 0.9251488005771659, + "flos": 10633581348480.0, + "grad_norm": 2.853347742695027, + "language_loss": 0.72695637, + "learning_rate": 5.840351392908349e-08, + "loss": 0.7489121, + "num_input_tokens_seen": 166374455, + "step": 7694, + "time_per_iteration": 2.650831460952759 + }, + { + "auxiliary_loss_clip": 0.01124114, + "auxiliary_loss_mlp": 0.00509143, + "balance_loss_clip": 0.97135144, + "balance_loss_mlp": 0.0, + "epoch": 0.9252690434678049, + "flos": 23586416052480.0, + "grad_norm": 3.0859886800092062, + "language_loss": 0.70943451, + "learning_rate": 5.821678791599205e-08, + "loss": 0.72576708, + "num_input_tokens_seen": 166393900, + "step": 7695, + "time_per_iteration": 2.7242982387542725 + }, + { + "auxiliary_loss_clip": 0.01115165, + "auxiliary_loss_mlp": 0.01028024, + "balance_loss_clip": 0.97213143, + "balance_loss_mlp": 1.02151179, + "epoch": 0.9253892863584441, + "flos": 21469158829440.0, + "grad_norm": 1.9237780462437872, + "language_loss": 0.80962598, + "learning_rate": 5.803035646986965e-08, + "loss": 0.83105791, + "num_input_tokens_seen": 166413235, + "step": 7696, + "time_per_iteration": 2.838181495666504 + }, + { + "auxiliary_loss_clip": 0.01168522, + "auxiliary_loss_mlp": 0.01026081, + "balance_loss_clip": 1.04931593, + "balance_loss_mlp": 1.01867199, + "epoch": 0.9255095292490831, + "flos": 17456680304640.0, + "grad_norm": 2.5487406887390667, + "language_loss": 0.67284262, + "learning_rate": 5.7844219618998766e-08, + "loss": 0.69478863, + "num_input_tokens_seen": 166427560, + "step": 7697, + "time_per_iteration": 2.6089084148406982 + }, + { + "auxiliary_loss_clip": 0.01080886, + "auxiliary_loss_mlp": 0.01024208, + "balance_loss_clip": 0.92763537, + "balance_loss_mlp": 1.01777041, + "epoch": 0.9256297721397222, + "flos": 24750675584640.0, + "grad_norm": 1.8637222199433359, + "language_loss": 0.7199564, + "learning_rate": 5.765837739161505e-08, + "loss": 0.74100733, + "num_input_tokens_seen": 166446680, + "step": 7698, + "time_per_iteration": 2.7527244091033936 + }, + { + "auxiliary_loss_clip": 0.0109645, + "auxiliary_loss_mlp": 0.01024979, + "balance_loss_clip": 0.93316746, + "balance_loss_mlp": 1.01766527, + "epoch": 0.9257500150303614, + "flos": 23112215677440.0, + "grad_norm": 1.7500296008965306, + "language_loss": 0.7438094, + "learning_rate": 5.7472829815911504e-08, + "loss": 0.76502371, + "num_input_tokens_seen": 166465505, + "step": 7699, + "time_per_iteration": 2.7964067459106445 + }, + { + "auxiliary_loss_clip": 0.01112141, + "auxiliary_loss_mlp": 0.01031369, + "balance_loss_clip": 0.97022176, + "balance_loss_mlp": 1.02369738, + "epoch": 0.9258702579210004, + "flos": 22564685687040.0, + "grad_norm": 1.8095852296252626, + "language_loss": 0.8170774, + "learning_rate": 5.7287576920035164e-08, + "loss": 0.83851242, + "num_input_tokens_seen": 166484520, + "step": 7700, + "time_per_iteration": 2.6880013942718506 + }, + { + "auxiliary_loss_clip": 0.01092502, + "auxiliary_loss_mlp": 0.01025769, + "balance_loss_clip": 0.93365586, + "balance_loss_mlp": 1.01859581, + "epoch": 0.9259905008116395, + "flos": 30004298703360.0, + "grad_norm": 1.826769752647213, + "language_loss": 0.76610053, + "learning_rate": 5.7102618732088435e-08, + "loss": 0.78728324, + "num_input_tokens_seen": 166503850, + "step": 7701, + "time_per_iteration": 2.795530319213867 + }, + { + "auxiliary_loss_clip": 0.01122297, + "auxiliary_loss_mlp": 0.0102378, + "balance_loss_clip": 0.97193867, + "balance_loss_mlp": 1.01694345, + "epoch": 0.9261107437022786, + "flos": 24572128055040.0, + "grad_norm": 1.9386093302860656, + "language_loss": 0.74377346, + "learning_rate": 5.6917955280130216e-08, + "loss": 0.76523423, + "num_input_tokens_seen": 166525330, + "step": 7702, + "time_per_iteration": 2.7166576385498047 + }, + { + "auxiliary_loss_clip": 0.01139186, + "auxiliary_loss_mlp": 0.01023878, + "balance_loss_clip": 1.01069295, + "balance_loss_mlp": 1.01634431, + "epoch": 0.9262309865929177, + "flos": 22018448586240.0, + "grad_norm": 3.135237864323596, + "language_loss": 0.72250479, + "learning_rate": 5.6733586592172755e-08, + "loss": 0.74413544, + "num_input_tokens_seen": 166544825, + "step": 7703, + "time_per_iteration": 2.691969871520996 + }, + { + "auxiliary_loss_clip": 0.0111015, + "auxiliary_loss_mlp": 0.00508503, + "balance_loss_clip": 0.96761072, + "balance_loss_mlp": 0.0, + "epoch": 0.9263512294835567, + "flos": 20339481116160.0, + "grad_norm": 3.823296918454374, + "language_loss": 0.79717267, + "learning_rate": 5.6549512696185244e-08, + "loss": 0.81335926, + "num_input_tokens_seen": 166563325, + "step": 7704, + "time_per_iteration": 2.821479558944702 + }, + { + "auxiliary_loss_clip": 0.01161003, + "auxiliary_loss_mlp": 0.01028122, + "balance_loss_clip": 1.0477798, + "balance_loss_mlp": 1.0210948, + "epoch": 0.9264714723741959, + "flos": 21215378263680.0, + "grad_norm": 1.5423074502594445, + "language_loss": 0.68164212, + "learning_rate": 5.636573362009156e-08, + "loss": 0.70353335, + "num_input_tokens_seen": 166583385, + "step": 7705, + "time_per_iteration": 2.684640645980835 + }, + { + "auxiliary_loss_clip": 0.01164428, + "auxiliary_loss_mlp": 0.01030851, + "balance_loss_clip": 1.04678977, + "balance_loss_mlp": 1.02373421, + "epoch": 0.926591715264835, + "flos": 18004964480640.0, + "grad_norm": 2.1940856689617716, + "language_loss": 0.77125835, + "learning_rate": 5.618224939177074e-08, + "loss": 0.7932111, + "num_input_tokens_seen": 166601290, + "step": 7706, + "time_per_iteration": 2.618861436843872 + }, + { + "auxiliary_loss_clip": 0.01107231, + "auxiliary_loss_mlp": 0.01023178, + "balance_loss_clip": 0.96898246, + "balance_loss_mlp": 1.0159421, + "epoch": 0.926711958155474, + "flos": 36167969825280.0, + "grad_norm": 1.7128532117528013, + "language_loss": 0.7027182, + "learning_rate": 5.599906003905719e-08, + "loss": 0.72402227, + "num_input_tokens_seen": 166623835, + "step": 7707, + "time_per_iteration": 3.686952590942383 + }, + { + "auxiliary_loss_clip": 0.01139149, + "auxiliary_loss_mlp": 0.01028062, + "balance_loss_clip": 1.01329029, + "balance_loss_mlp": 1.02090657, + "epoch": 0.9268322010461132, + "flos": 21032736583680.0, + "grad_norm": 2.124296585179197, + "language_loss": 0.81592524, + "learning_rate": 5.581616558974023e-08, + "loss": 0.83759725, + "num_input_tokens_seen": 166642400, + "step": 7708, + "time_per_iteration": 2.660606861114502 + }, + { + "auxiliary_loss_clip": 0.0114896, + "auxiliary_loss_mlp": 0.00509294, + "balance_loss_clip": 1.01249862, + "balance_loss_mlp": 0.0, + "epoch": 0.9269524439367522, + "flos": 22964838174720.0, + "grad_norm": 2.4907577045311395, + "language_loss": 0.79081333, + "learning_rate": 5.5633566071565444e-08, + "loss": 0.80739582, + "num_input_tokens_seen": 166661640, + "step": 7709, + "time_per_iteration": 4.580806255340576 + }, + { + "auxiliary_loss_clip": 0.01052463, + "auxiliary_loss_mlp": 0.01020616, + "balance_loss_clip": 0.85717261, + "balance_loss_mlp": 1.01372862, + "epoch": 0.9270726868273913, + "flos": 41975551468800.0, + "grad_norm": 2.01611659683692, + "language_loss": 0.71072465, + "learning_rate": 5.5451261512232896e-08, + "loss": 0.73145545, + "num_input_tokens_seen": 166684320, + "step": 7710, + "time_per_iteration": 2.9366650581359863 + }, + { + "auxiliary_loss_clip": 0.01142753, + "auxiliary_loss_mlp": 0.01025535, + "balance_loss_clip": 1.00742435, + "balance_loss_mlp": 1.01842999, + "epoch": 0.9271929297180305, + "flos": 19791771557760.0, + "grad_norm": 1.7827677323295539, + "language_loss": 0.62338734, + "learning_rate": 5.5269251939397576e-08, + "loss": 0.64507025, + "num_input_tokens_seen": 166703835, + "step": 7711, + "time_per_iteration": 2.626945734024048 + }, + { + "auxiliary_loss_clip": 0.01094967, + "auxiliary_loss_mlp": 0.01023743, + "balance_loss_clip": 0.92982423, + "balance_loss_mlp": 1.01657903, + "epoch": 0.9273131726086695, + "flos": 19968343839360.0, + "grad_norm": 2.3415223923157797, + "language_loss": 0.76430345, + "learning_rate": 5.508753738067073e-08, + "loss": 0.78549051, + "num_input_tokens_seen": 166723375, + "step": 7712, + "time_per_iteration": 2.7180120944976807 + }, + { + "auxiliary_loss_clip": 0.011426, + "auxiliary_loss_mlp": 0.01019508, + "balance_loss_clip": 1.00848532, + "balance_loss_mlp": 1.01227212, + "epoch": 0.9274334154993086, + "flos": 23258587599360.0, + "grad_norm": 2.0080185063387384, + "language_loss": 0.79554045, + "learning_rate": 5.4906117863617875e-08, + "loss": 0.81716156, + "num_input_tokens_seen": 166742760, + "step": 7713, + "time_per_iteration": 2.6937255859375 + }, + { + "auxiliary_loss_clip": 0.01088571, + "auxiliary_loss_mlp": 0.01025594, + "balance_loss_clip": 0.93029964, + "balance_loss_mlp": 1.01919818, + "epoch": 0.9275536583899477, + "flos": 31795343585280.0, + "grad_norm": 2.3804368111340537, + "language_loss": 0.78164035, + "learning_rate": 5.4724993415760533e-08, + "loss": 0.80278194, + "num_input_tokens_seen": 166761115, + "step": 7714, + "time_per_iteration": 2.895266532897949 + }, + { + "auxiliary_loss_clip": 0.01100903, + "auxiliary_loss_mlp": 0.0050945, + "balance_loss_clip": 0.93112075, + "balance_loss_mlp": 0.0, + "epoch": 0.9276739012805868, + "flos": 18696998885760.0, + "grad_norm": 2.150301859843223, + "language_loss": 0.74803066, + "learning_rate": 5.454416406457496e-08, + "loss": 0.76413423, + "num_input_tokens_seen": 166780210, + "step": 7715, + "time_per_iteration": 2.7265031337738037 + }, + { + "auxiliary_loss_clip": 0.01137418, + "auxiliary_loss_mlp": 0.01025505, + "balance_loss_clip": 1.00633585, + "balance_loss_mlp": 1.01848328, + "epoch": 0.9277941441712259, + "flos": 13879079740800.0, + "grad_norm": 3.252737962577659, + "language_loss": 0.74250865, + "learning_rate": 5.436362983749299e-08, + "loss": 0.76413786, + "num_input_tokens_seen": 166795380, + "step": 7716, + "time_per_iteration": 2.8703672885894775 + }, + { + "auxiliary_loss_clip": 0.01089105, + "auxiliary_loss_mlp": 0.0102705, + "balance_loss_clip": 0.93496299, + "balance_loss_mlp": 1.02047586, + "epoch": 0.927914387061865, + "flos": 23258659426560.0, + "grad_norm": 1.9409134852456085, + "language_loss": 0.64419842, + "learning_rate": 5.418339076190137e-08, + "loss": 0.66535997, + "num_input_tokens_seen": 166814890, + "step": 7717, + "time_per_iteration": 2.794740915298462 + }, + { + "auxiliary_loss_clip": 0.01111337, + "auxiliary_loss_mlp": 0.01024891, + "balance_loss_clip": 0.97153771, + "balance_loss_mlp": 1.01771462, + "epoch": 0.9280346299525041, + "flos": 18073733068800.0, + "grad_norm": 1.7456790820625194, + "language_loss": 0.88827521, + "learning_rate": 5.400344686514202e-08, + "loss": 0.90963751, + "num_input_tokens_seen": 166832475, + "step": 7718, + "time_per_iteration": 3.5954813957214355 + }, + { + "auxiliary_loss_clip": 0.01140462, + "auxiliary_loss_mlp": 0.01024673, + "balance_loss_clip": 1.01077819, + "balance_loss_mlp": 1.01781869, + "epoch": 0.9281548728431431, + "flos": 22342901160960.0, + "grad_norm": 1.69947217855135, + "language_loss": 0.66217226, + "learning_rate": 5.38237981745131e-08, + "loss": 0.68382359, + "num_input_tokens_seen": 166850590, + "step": 7719, + "time_per_iteration": 2.64758038520813 + }, + { + "auxiliary_loss_clip": 0.01141268, + "auxiliary_loss_mlp": 0.00509238, + "balance_loss_clip": 1.00808275, + "balance_loss_mlp": 0.0, + "epoch": 0.9282751157337822, + "flos": 18843765857280.0, + "grad_norm": 2.007835020292658, + "language_loss": 0.81353092, + "learning_rate": 5.364444471726592e-08, + "loss": 0.83003604, + "num_input_tokens_seen": 166869795, + "step": 7720, + "time_per_iteration": 2.7634339332580566 + }, + { + "auxiliary_loss_clip": 0.01137407, + "auxiliary_loss_mlp": 0.01024697, + "balance_loss_clip": 1.00785518, + "balance_loss_mlp": 1.01744866, + "epoch": 0.9283953586244214, + "flos": 25556834476800.0, + "grad_norm": 2.026894215362561, + "language_loss": 0.79645205, + "learning_rate": 5.346538652060939e-08, + "loss": 0.81807303, + "num_input_tokens_seen": 166891150, + "step": 7721, + "time_per_iteration": 2.6652677059173584 + }, + { + "auxiliary_loss_clip": 0.01114468, + "auxiliary_loss_mlp": 0.0102571, + "balance_loss_clip": 0.97210878, + "balance_loss_mlp": 1.01894236, + "epoch": 0.9285156015150604, + "flos": 18223480869120.0, + "grad_norm": 1.8761558646861396, + "language_loss": 0.70143676, + "learning_rate": 5.3286623611705994e-08, + "loss": 0.72283852, + "num_input_tokens_seen": 166909195, + "step": 7722, + "time_per_iteration": 2.696591854095459 + }, + { + "auxiliary_loss_clip": 0.01059243, + "auxiliary_loss_mlp": 0.00999488, + "balance_loss_clip": 1.00873518, + "balance_loss_mlp": 0.99853402, + "epoch": 0.9286358444056995, + "flos": 66400017690240.0, + "grad_norm": 0.817219061229135, + "language_loss": 0.60591364, + "learning_rate": 5.3108156017673824e-08, + "loss": 0.62650096, + "num_input_tokens_seen": 166970955, + "step": 7723, + "time_per_iteration": 3.29101300239563 + }, + { + "auxiliary_loss_clip": 0.01121707, + "auxiliary_loss_mlp": 0.010289, + "balance_loss_clip": 0.97094703, + "balance_loss_mlp": 1.02166963, + "epoch": 0.9287560872963386, + "flos": 22345630594560.0, + "grad_norm": 1.6787909946284754, + "language_loss": 0.71644682, + "learning_rate": 5.2929983765586775e-08, + "loss": 0.73795295, + "num_input_tokens_seen": 166989735, + "step": 7724, + "time_per_iteration": 2.6462290287017822 + }, + { + "auxiliary_loss_clip": 0.01164093, + "auxiliary_loss_mlp": 0.01026375, + "balance_loss_clip": 1.04821372, + "balance_loss_mlp": 1.01904368, + "epoch": 0.9288763301869777, + "flos": 25700225569920.0, + "grad_norm": 6.221847149871268, + "language_loss": 0.62736452, + "learning_rate": 5.275210688247278e-08, + "loss": 0.64926916, + "num_input_tokens_seen": 167010060, + "step": 7725, + "time_per_iteration": 2.6903905868530273 + }, + { + "auxiliary_loss_clip": 0.01073399, + "auxiliary_loss_mlp": 0.01025555, + "balance_loss_clip": 0.89836216, + "balance_loss_mlp": 1.01842606, + "epoch": 0.9289965730776167, + "flos": 12312046028160.0, + "grad_norm": 2.9563035781997558, + "language_loss": 0.85056436, + "learning_rate": 5.257452539531604e-08, + "loss": 0.8715539, + "num_input_tokens_seen": 167027130, + "step": 7726, + "time_per_iteration": 2.7148008346557617 + }, + { + "auxiliary_loss_clip": 0.011379, + "auxiliary_loss_mlp": 0.01021015, + "balance_loss_clip": 1.00577521, + "balance_loss_mlp": 1.01436913, + "epoch": 0.9291168159682559, + "flos": 26685973486080.0, + "grad_norm": 1.9817552843643609, + "language_loss": 0.68630159, + "learning_rate": 5.2397239331055445e-08, + "loss": 0.70789075, + "num_input_tokens_seen": 167049130, + "step": 7727, + "time_per_iteration": 2.7088735103607178 + }, + { + "auxiliary_loss_clip": 0.01115786, + "auxiliary_loss_mlp": 0.01028736, + "balance_loss_clip": 0.9741143, + "balance_loss_mlp": 1.02120817, + "epoch": 0.929237058858895, + "flos": 14538256179840.0, + "grad_norm": 2.16008492000884, + "language_loss": 0.81288934, + "learning_rate": 5.2220248716585036e-08, + "loss": 0.83433455, + "num_input_tokens_seen": 167066810, + "step": 7728, + "time_per_iteration": 2.6795613765716553 + }, + { + "auxiliary_loss_clip": 0.01133567, + "auxiliary_loss_mlp": 0.01029696, + "balance_loss_clip": 1.00651395, + "balance_loss_mlp": 1.02237606, + "epoch": 0.929357301749534, + "flos": 23835456023040.0, + "grad_norm": 2.613204271185493, + "language_loss": 0.75448334, + "learning_rate": 5.204355357875445e-08, + "loss": 0.77611601, + "num_input_tokens_seen": 167085155, + "step": 7729, + "time_per_iteration": 2.71012020111084 + }, + { + "auxiliary_loss_clip": 0.01113136, + "auxiliary_loss_mlp": 0.01027199, + "balance_loss_clip": 0.96799719, + "balance_loss_mlp": 1.01965046, + "epoch": 0.9294775446401732, + "flos": 12969319046400.0, + "grad_norm": 2.221972804581732, + "language_loss": 0.70295221, + "learning_rate": 5.1867153944367584e-08, + "loss": 0.72435558, + "num_input_tokens_seen": 167101545, + "step": 7730, + "time_per_iteration": 2.646897077560425 + }, + { + "auxiliary_loss_clip": 0.01099086, + "auxiliary_loss_mlp": 0.01022555, + "balance_loss_clip": 0.93333912, + "balance_loss_mlp": 1.01581025, + "epoch": 0.9295977875308122, + "flos": 26211809024640.0, + "grad_norm": 1.6156115292107014, + "language_loss": 0.73651612, + "learning_rate": 5.16910498401848e-08, + "loss": 0.75773251, + "num_input_tokens_seen": 167120995, + "step": 7731, + "time_per_iteration": 2.75738263130188 + }, + { + "auxiliary_loss_clip": 0.01162776, + "auxiliary_loss_mlp": 0.01026982, + "balance_loss_clip": 1.04899907, + "balance_loss_mlp": 1.02042866, + "epoch": 0.9297180304214513, + "flos": 16472297105280.0, + "grad_norm": 2.310802401615001, + "language_loss": 0.83494234, + "learning_rate": 5.151524129292073e-08, + "loss": 0.8568399, + "num_input_tokens_seen": 167138890, + "step": 7732, + "time_per_iteration": 2.584803342819214 + }, + { + "auxiliary_loss_clip": 0.01138837, + "auxiliary_loss_mlp": 0.0102371, + "balance_loss_clip": 1.00901103, + "balance_loss_mlp": 1.016361, + "epoch": 0.9298382733120905, + "flos": 24060436859520.0, + "grad_norm": 1.91014594486703, + "language_loss": 0.66274953, + "learning_rate": 5.1339728329245155e-08, + "loss": 0.68437505, + "num_input_tokens_seen": 167159455, + "step": 7733, + "time_per_iteration": 3.5502912998199463 + }, + { + "auxiliary_loss_clip": 0.01167191, + "auxiliary_loss_mlp": 0.0102554, + "balance_loss_clip": 1.04746318, + "balance_loss_mlp": 1.01819658, + "epoch": 0.9299585162027295, + "flos": 22127652910080.0, + "grad_norm": 3.1115239441976312, + "language_loss": 0.79269141, + "learning_rate": 5.116451097578367e-08, + "loss": 0.81461871, + "num_input_tokens_seen": 167178495, + "step": 7734, + "time_per_iteration": 3.456279754638672 + }, + { + "auxiliary_loss_clip": 0.0109256, + "auxiliary_loss_mlp": 0.01032019, + "balance_loss_clip": 0.93173611, + "balance_loss_mlp": 1.02512538, + "epoch": 0.9300787590933686, + "flos": 21471780522240.0, + "grad_norm": 1.6887283054011608, + "language_loss": 0.74396932, + "learning_rate": 5.0989589259115895e-08, + "loss": 0.7652151, + "num_input_tokens_seen": 167199380, + "step": 7735, + "time_per_iteration": 3.6606321334838867 + }, + { + "auxiliary_loss_clip": 0.01135897, + "auxiliary_loss_mlp": 0.01026354, + "balance_loss_clip": 1.00636363, + "balance_loss_mlp": 1.01883435, + "epoch": 0.9301990019840077, + "flos": 17779588594560.0, + "grad_norm": 1.7398420233629859, + "language_loss": 0.71390671, + "learning_rate": 5.081496320577816e-08, + "loss": 0.73552918, + "num_input_tokens_seen": 167216500, + "step": 7736, + "time_per_iteration": 2.593033790588379 + }, + { + "auxiliary_loss_clip": 0.01035572, + "auxiliary_loss_mlp": 0.01009826, + "balance_loss_clip": 0.95553768, + "balance_loss_mlp": 1.00862825, + "epoch": 0.9303192448746468, + "flos": 58896122307840.0, + "grad_norm": 0.9131626042266919, + "language_loss": 0.61190701, + "learning_rate": 5.0640632842260835e-08, + "loss": 0.63236099, + "num_input_tokens_seen": 167276760, + "step": 7737, + "time_per_iteration": 3.305795192718506 + }, + { + "auxiliary_loss_clip": 0.01093336, + "auxiliary_loss_mlp": 0.00508837, + "balance_loss_clip": 0.93583918, + "balance_loss_mlp": 0.0, + "epoch": 0.9304394877652858, + "flos": 57663522172800.0, + "grad_norm": 3.5463698704621818, + "language_loss": 0.72553325, + "learning_rate": 5.0466598195009426e-08, + "loss": 0.74155504, + "num_input_tokens_seen": 167303630, + "step": 7738, + "time_per_iteration": 3.0430781841278076 + }, + { + "auxiliary_loss_clip": 0.01095792, + "auxiliary_loss_mlp": 0.01024587, + "balance_loss_clip": 0.93308324, + "balance_loss_mlp": 1.01751804, + "epoch": 0.930559730655925, + "flos": 20996143603200.0, + "grad_norm": 1.7859029135116191, + "language_loss": 0.70199335, + "learning_rate": 5.0292859290425036e-08, + "loss": 0.7231971, + "num_input_tokens_seen": 167321500, + "step": 7739, + "time_per_iteration": 2.7275140285491943 + }, + { + "auxiliary_loss_clip": 0.01164577, + "auxiliary_loss_mlp": 0.01026964, + "balance_loss_clip": 1.04919147, + "balance_loss_mlp": 1.02062201, + "epoch": 0.9306799735465641, + "flos": 23258264376960.0, + "grad_norm": 1.958781820029011, + "language_loss": 0.77433515, + "learning_rate": 5.011941615486348e-08, + "loss": 0.79625046, + "num_input_tokens_seen": 167340615, + "step": 7740, + "time_per_iteration": 2.618527889251709 + }, + { + "auxiliary_loss_clip": 0.01163596, + "auxiliary_loss_mlp": 0.01025108, + "balance_loss_clip": 1.04788291, + "balance_loss_mlp": 1.01812816, + "epoch": 0.9308002164372031, + "flos": 15231547560960.0, + "grad_norm": 2.013731155007137, + "language_loss": 0.84597719, + "learning_rate": 4.994626881463659e-08, + "loss": 0.86786425, + "num_input_tokens_seen": 167356870, + "step": 7741, + "time_per_iteration": 2.5391530990600586 + }, + { + "auxiliary_loss_clip": 0.01058259, + "auxiliary_loss_mlp": 0.0102687, + "balance_loss_clip": 0.89226037, + "balance_loss_mlp": 1.02018881, + "epoch": 0.9309204593278423, + "flos": 30847481539200.0, + "grad_norm": 1.686339466220093, + "language_loss": 0.71232486, + "learning_rate": 4.9773417296009814e-08, + "loss": 0.73317611, + "num_input_tokens_seen": 167378390, + "step": 7742, + "time_per_iteration": 2.870636463165283 + }, + { + "auxiliary_loss_clip": 0.01144882, + "auxiliary_loss_mlp": 0.01027717, + "balance_loss_clip": 1.0103786, + "balance_loss_mlp": 1.02078509, + "epoch": 0.9310407022184813, + "flos": 23037269950080.0, + "grad_norm": 4.82587768837772, + "language_loss": 0.65369284, + "learning_rate": 4.960086162520527e-08, + "loss": 0.67541885, + "num_input_tokens_seen": 167398480, + "step": 7743, + "time_per_iteration": 3.6629722118377686 + }, + { + "auxiliary_loss_clip": 0.01077302, + "auxiliary_loss_mlp": 0.01025231, + "balance_loss_clip": 0.89505613, + "balance_loss_mlp": 1.01841831, + "epoch": 0.9311609451091204, + "flos": 22127976132480.0, + "grad_norm": 2.1281583066361436, + "language_loss": 0.82691526, + "learning_rate": 4.942860182839936e-08, + "loss": 0.84794056, + "num_input_tokens_seen": 167416825, + "step": 7744, + "time_per_iteration": 2.7463791370391846 + }, + { + "auxiliary_loss_clip": 0.01116144, + "auxiliary_loss_mlp": 0.01027814, + "balance_loss_clip": 0.97024107, + "balance_loss_mlp": 1.02106333, + "epoch": 0.9312811879997596, + "flos": 21099206701440.0, + "grad_norm": 1.7846587413356292, + "language_loss": 0.79437578, + "learning_rate": 4.925663793172341e-08, + "loss": 0.81581533, + "num_input_tokens_seen": 167434785, + "step": 7745, + "time_per_iteration": 2.7208406925201416 + }, + { + "auxiliary_loss_clip": 0.01017316, + "auxiliary_loss_mlp": 0.00500825, + "balance_loss_clip": 0.93528509, + "balance_loss_mlp": 0.0, + "epoch": 0.9314014308903986, + "flos": 67148179096320.0, + "grad_norm": 0.7844754802769519, + "language_loss": 0.56524074, + "learning_rate": 4.908496996126477e-08, + "loss": 0.58042204, + "num_input_tokens_seen": 167498245, + "step": 7746, + "time_per_iteration": 3.2495172023773193 + }, + { + "auxiliary_loss_clip": 0.01142761, + "auxiliary_loss_mlp": 0.01023785, + "balance_loss_clip": 1.0153985, + "balance_loss_mlp": 1.01651073, + "epoch": 0.9315216737810377, + "flos": 22565583527040.0, + "grad_norm": 1.6074632614857012, + "language_loss": 0.76838166, + "learning_rate": 4.89135979430646e-08, + "loss": 0.79004711, + "num_input_tokens_seen": 167518290, + "step": 7747, + "time_per_iteration": 2.6325154304504395 + }, + { + "auxiliary_loss_clip": 0.01162786, + "auxiliary_loss_mlp": 0.01032781, + "balance_loss_clip": 1.04800236, + "balance_loss_mlp": 1.02567315, + "epoch": 0.9316419166716768, + "flos": 23984054588160.0, + "grad_norm": 1.7002151289845184, + "language_loss": 0.85639352, + "learning_rate": 4.874252190312078e-08, + "loss": 0.87834918, + "num_input_tokens_seen": 167538675, + "step": 7748, + "time_per_iteration": 2.650946855545044 + }, + { + "auxiliary_loss_clip": 0.01145034, + "auxiliary_loss_mlp": 0.01026774, + "balance_loss_clip": 1.01012635, + "balance_loss_mlp": 1.01975858, + "epoch": 0.9317621595623159, + "flos": 30230464688640.0, + "grad_norm": 2.169897323026549, + "language_loss": 0.65058589, + "learning_rate": 4.857174186738477e-08, + "loss": 0.67230392, + "num_input_tokens_seen": 167562025, + "step": 7749, + "time_per_iteration": 2.693464517593384 + }, + { + "auxiliary_loss_clip": 0.01167991, + "auxiliary_loss_mlp": 0.01025798, + "balance_loss_clip": 1.05184305, + "balance_loss_mlp": 1.01868153, + "epoch": 0.931882402452955, + "flos": 15742735966080.0, + "grad_norm": 2.1458560635920003, + "language_loss": 0.73235548, + "learning_rate": 4.840125786176408e-08, + "loss": 0.75429332, + "num_input_tokens_seen": 167578230, + "step": 7750, + "time_per_iteration": 2.579017400741577 + }, + { + "auxiliary_loss_clip": 0.0111339, + "auxiliary_loss_mlp": 0.01021204, + "balance_loss_clip": 0.97144765, + "balance_loss_mlp": 1.01445043, + "epoch": 0.932002645343594, + "flos": 28366521154560.0, + "grad_norm": 1.7295878761716255, + "language_loss": 0.77211952, + "learning_rate": 4.823106991212067e-08, + "loss": 0.79346544, + "num_input_tokens_seen": 167597470, + "step": 7751, + "time_per_iteration": 2.748002767562866 + }, + { + "auxiliary_loss_clip": 0.01143996, + "auxiliary_loss_mlp": 0.01020751, + "balance_loss_clip": 1.01034641, + "balance_loss_mlp": 1.01360726, + "epoch": 0.9321228882342332, + "flos": 15341146934400.0, + "grad_norm": 1.786356816585453, + "language_loss": 0.83518767, + "learning_rate": 4.806117804427212e-08, + "loss": 0.85683507, + "num_input_tokens_seen": 167615405, + "step": 7752, + "time_per_iteration": 2.622554063796997 + }, + { + "auxiliary_loss_clip": 0.0113447, + "auxiliary_loss_mlp": 0.01023782, + "balance_loss_clip": 1.00726426, + "balance_loss_mlp": 1.01648641, + "epoch": 0.9322431311248722, + "flos": 17895365107200.0, + "grad_norm": 2.11135933888165, + "language_loss": 0.64377403, + "learning_rate": 4.7891582283990926e-08, + "loss": 0.66535652, + "num_input_tokens_seen": 167634130, + "step": 7753, + "time_per_iteration": 2.620847463607788 + }, + { + "auxiliary_loss_clip": 0.010945, + "auxiliary_loss_mlp": 0.01025515, + "balance_loss_clip": 0.93192339, + "balance_loss_mlp": 1.01846421, + "epoch": 0.9323633740155113, + "flos": 24169713010560.0, + "grad_norm": 1.5716468529702603, + "language_loss": 0.72927594, + "learning_rate": 4.772228265700473e-08, + "loss": 0.75047606, + "num_input_tokens_seen": 167654990, + "step": 7754, + "time_per_iteration": 2.7239480018615723 + }, + { + "auxiliary_loss_clip": 0.01144508, + "auxiliary_loss_mlp": 0.01021474, + "balance_loss_clip": 1.0103296, + "balance_loss_mlp": 1.01470339, + "epoch": 0.9324836169061504, + "flos": 15043482927360.0, + "grad_norm": 2.9467617044431282, + "language_loss": 0.75847673, + "learning_rate": 4.75532791889961e-08, + "loss": 0.78013653, + "num_input_tokens_seen": 167671690, + "step": 7755, + "time_per_iteration": 2.6183371543884277 + }, + { + "auxiliary_loss_clip": 0.01139495, + "auxiliary_loss_mlp": 0.01024381, + "balance_loss_clip": 1.00852561, + "balance_loss_mlp": 1.01734495, + "epoch": 0.9326038597967895, + "flos": 18624890332800.0, + "grad_norm": 1.891693481749165, + "language_loss": 0.65511, + "learning_rate": 4.738457190560252e-08, + "loss": 0.67674875, + "num_input_tokens_seen": 167690800, + "step": 7756, + "time_per_iteration": 2.57529616355896 + }, + { + "auxiliary_loss_clip": 0.0107203, + "auxiliary_loss_mlp": 0.01025845, + "balance_loss_clip": 0.8979454, + "balance_loss_mlp": 1.01867414, + "epoch": 0.9327241026874286, + "flos": 18952646958720.0, + "grad_norm": 2.9291933241015733, + "language_loss": 0.7905668, + "learning_rate": 4.721616083241664e-08, + "loss": 0.81154555, + "num_input_tokens_seen": 167709055, + "step": 7757, + "time_per_iteration": 2.756377935409546 + }, + { + "auxiliary_loss_clip": 0.01134931, + "auxiliary_loss_mlp": 0.01031213, + "balance_loss_clip": 1.00817811, + "balance_loss_mlp": 1.02346468, + "epoch": 0.9328443455780677, + "flos": 29570282668800.0, + "grad_norm": 1.6954360398218067, + "language_loss": 0.77779543, + "learning_rate": 4.7048045994986684e-08, + "loss": 0.79945689, + "num_input_tokens_seen": 167729915, + "step": 7758, + "time_per_iteration": 2.703427791595459 + }, + { + "auxiliary_loss_clip": 0.0114963, + "auxiliary_loss_mlp": 0.01026453, + "balance_loss_clip": 1.0127387, + "balance_loss_mlp": 1.01925564, + "epoch": 0.9329645884687068, + "flos": 30081722469120.0, + "grad_norm": 1.875016494951661, + "language_loss": 0.91343933, + "learning_rate": 4.688022741881559e-08, + "loss": 0.93520021, + "num_input_tokens_seen": 167750440, + "step": 7759, + "time_per_iteration": 3.5934767723083496 + }, + { + "auxiliary_loss_clip": 0.01135429, + "auxiliary_loss_mlp": 0.01022358, + "balance_loss_clip": 1.00765073, + "balance_loss_mlp": 1.01561642, + "epoch": 0.9330848313593458, + "flos": 21867982513920.0, + "grad_norm": 1.4968201692228935, + "language_loss": 0.74885303, + "learning_rate": 4.671270512936076e-08, + "loss": 0.77043092, + "num_input_tokens_seen": 167769600, + "step": 7760, + "time_per_iteration": 3.5267271995544434 + }, + { + "auxiliary_loss_clip": 0.01086181, + "auxiliary_loss_mlp": 0.01025049, + "balance_loss_clip": 0.92982638, + "balance_loss_mlp": 1.01840591, + "epoch": 0.933205074249985, + "flos": 22127221946880.0, + "grad_norm": 1.7167469352085578, + "language_loss": 0.82648927, + "learning_rate": 4.6545479152035884e-08, + "loss": 0.84760153, + "num_input_tokens_seen": 167788770, + "step": 7761, + "time_per_iteration": 3.721881866455078 + }, + { + "auxiliary_loss_clip": 0.01139849, + "auxiliary_loss_mlp": 0.01023813, + "balance_loss_clip": 1.00836051, + "balance_loss_mlp": 1.01728308, + "epoch": 0.9333253171406241, + "flos": 15341254675200.0, + "grad_norm": 2.0502041533638162, + "language_loss": 0.76184857, + "learning_rate": 4.637854951220821e-08, + "loss": 0.78348517, + "num_input_tokens_seen": 167805555, + "step": 7762, + "time_per_iteration": 2.611614227294922 + }, + { + "auxiliary_loss_clip": 0.01089828, + "auxiliary_loss_mlp": 0.01024756, + "balance_loss_clip": 0.93288529, + "balance_loss_mlp": 1.01748407, + "epoch": 0.9334455600312631, + "flos": 15706142985600.0, + "grad_norm": 1.757433605949505, + "language_loss": 0.74846685, + "learning_rate": 4.621191623520171e-08, + "loss": 0.76961267, + "num_input_tokens_seen": 167823985, + "step": 7763, + "time_per_iteration": 2.720301866531372 + }, + { + "auxiliary_loss_clip": 0.01059609, + "auxiliary_loss_mlp": 0.0102452, + "balance_loss_clip": 0.85667068, + "balance_loss_mlp": 1.01753783, + "epoch": 0.9335658029219023, + "flos": 22163563532160.0, + "grad_norm": 2.245052691990776, + "language_loss": 0.84545863, + "learning_rate": 4.604557934629372e-08, + "loss": 0.86629987, + "num_input_tokens_seen": 167843060, + "step": 7764, + "time_per_iteration": 2.783256769180298 + }, + { + "auxiliary_loss_clip": 0.01109761, + "auxiliary_loss_mlp": 0.01032551, + "balance_loss_clip": 0.96977359, + "balance_loss_mlp": 1.02550888, + "epoch": 0.9336860458125413, + "flos": 20266833859200.0, + "grad_norm": 3.076347055813215, + "language_loss": 0.80385578, + "learning_rate": 4.587953887071805e-08, + "loss": 0.82527894, + "num_input_tokens_seen": 167862880, + "step": 7765, + "time_per_iteration": 2.6961066722869873 + }, + { + "auxiliary_loss_clip": 0.01112162, + "auxiliary_loss_mlp": 0.01019835, + "balance_loss_clip": 0.96847558, + "balance_loss_mlp": 1.01280773, + "epoch": 0.9338062887031804, + "flos": 20919689504640.0, + "grad_norm": 11.21473997716162, + "language_loss": 0.8567152, + "learning_rate": 4.5713794833662554e-08, + "loss": 0.87803507, + "num_input_tokens_seen": 167882095, + "step": 7766, + "time_per_iteration": 2.6530957221984863 + }, + { + "auxiliary_loss_clip": 0.01164427, + "auxiliary_loss_mlp": 0.01028104, + "balance_loss_clip": 1.04776001, + "balance_loss_mlp": 1.02049017, + "epoch": 0.9339265315938196, + "flos": 23221635482880.0, + "grad_norm": 1.907514864448311, + "language_loss": 0.63362479, + "learning_rate": 4.5548347260270236e-08, + "loss": 0.65555012, + "num_input_tokens_seen": 167901385, + "step": 7767, + "time_per_iteration": 2.6310880184173584 + }, + { + "auxiliary_loss_clip": 0.01088854, + "auxiliary_loss_mlp": 0.01025175, + "balance_loss_clip": 0.93159944, + "balance_loss_mlp": 1.01796567, + "epoch": 0.9340467744844586, + "flos": 22820261932800.0, + "grad_norm": 1.615664953634473, + "language_loss": 0.69546044, + "learning_rate": 4.538319617564012e-08, + "loss": 0.71660072, + "num_input_tokens_seen": 167920405, + "step": 7768, + "time_per_iteration": 2.7054038047790527 + }, + { + "auxiliary_loss_clip": 0.0111532, + "auxiliary_loss_mlp": 0.01025342, + "balance_loss_clip": 0.96969569, + "balance_loss_mlp": 1.01767993, + "epoch": 0.9341670173750977, + "flos": 23660428026240.0, + "grad_norm": 2.570978314533616, + "language_loss": 0.74974495, + "learning_rate": 4.521834160482485e-08, + "loss": 0.77115154, + "num_input_tokens_seen": 167939145, + "step": 7769, + "time_per_iteration": 2.718510866165161 + }, + { + "auxiliary_loss_clip": 0.01144838, + "auxiliary_loss_mlp": 0.01027679, + "balance_loss_clip": 1.01028347, + "balance_loss_mlp": 1.02044332, + "epoch": 0.9342872602657368, + "flos": 24824256595200.0, + "grad_norm": 1.7210354824671248, + "language_loss": 0.82119507, + "learning_rate": 4.5053783572832846e-08, + "loss": 0.8429203, + "num_input_tokens_seen": 167959325, + "step": 7770, + "time_per_iteration": 3.616830348968506 + }, + { + "auxiliary_loss_clip": 0.01141634, + "auxiliary_loss_mlp": 0.01028165, + "balance_loss_clip": 1.01174533, + "balance_loss_mlp": 1.02130461, + "epoch": 0.9344075031563759, + "flos": 25771831332480.0, + "grad_norm": 1.8165701356635717, + "language_loss": 0.76226228, + "learning_rate": 4.488952210462771e-08, + "loss": 0.78396028, + "num_input_tokens_seen": 167979530, + "step": 7771, + "time_per_iteration": 2.7365834712982178 + }, + { + "auxiliary_loss_clip": 0.01163202, + "auxiliary_loss_mlp": 0.01032025, + "balance_loss_clip": 1.04874134, + "balance_loss_mlp": 1.02478647, + "epoch": 0.9345277460470149, + "flos": 25551303782400.0, + "grad_norm": 1.930988150971591, + "language_loss": 0.86032319, + "learning_rate": 4.4725557225127495e-08, + "loss": 0.88227546, + "num_input_tokens_seen": 167997870, + "step": 7772, + "time_per_iteration": 2.636309862136841 + }, + { + "auxiliary_loss_clip": 0.0114374, + "auxiliary_loss_mlp": 0.01030022, + "balance_loss_clip": 1.01065254, + "balance_loss_mlp": 1.02336431, + "epoch": 0.9346479889376541, + "flos": 34313112432000.0, + "grad_norm": 1.6399296541505504, + "language_loss": 0.7938453, + "learning_rate": 4.456188895920565e-08, + "loss": 0.81558293, + "num_input_tokens_seen": 168019625, + "step": 7773, + "time_per_iteration": 2.6995482444763184 + }, + { + "auxiliary_loss_clip": 0.01163701, + "auxiliary_loss_mlp": 0.01028613, + "balance_loss_clip": 1.04848361, + "balance_loss_mlp": 1.02106404, + "epoch": 0.9347682318282932, + "flos": 19093739581440.0, + "grad_norm": 1.9097494287448828, + "language_loss": 0.85124362, + "learning_rate": 4.439851733169031e-08, + "loss": 0.8731668, + "num_input_tokens_seen": 168037415, + "step": 7774, + "time_per_iteration": 2.577232599258423 + }, + { + "auxiliary_loss_clip": 0.0109417, + "auxiliary_loss_mlp": 0.01021857, + "balance_loss_clip": 0.93179405, + "balance_loss_mlp": 1.01472569, + "epoch": 0.9348884747189322, + "flos": 26249587153920.0, + "grad_norm": 2.289400264274814, + "language_loss": 0.69513118, + "learning_rate": 4.4235442367365204e-08, + "loss": 0.71629143, + "num_input_tokens_seen": 168057725, + "step": 7775, + "time_per_iteration": 2.7665414810180664 + }, + { + "auxiliary_loss_clip": 0.01112807, + "auxiliary_loss_mlp": 0.01028196, + "balance_loss_clip": 0.96699572, + "balance_loss_mlp": 1.02090049, + "epoch": 0.9350087176095714, + "flos": 18333080242560.0, + "grad_norm": 2.466700430131023, + "language_loss": 0.79452097, + "learning_rate": 4.4072664090968545e-08, + "loss": 0.81593108, + "num_input_tokens_seen": 168076110, + "step": 7776, + "time_per_iteration": 2.653993844985962 + }, + { + "auxiliary_loss_clip": 0.01115552, + "auxiliary_loss_mlp": 0.01022986, + "balance_loss_clip": 0.96809602, + "balance_loss_mlp": 1.01563716, + "epoch": 0.9351289605002104, + "flos": 19318253541120.0, + "grad_norm": 1.8704428368296406, + "language_loss": 0.84730452, + "learning_rate": 4.391018252719347e-08, + "loss": 0.86868989, + "num_input_tokens_seen": 168095905, + "step": 7777, + "time_per_iteration": 2.692474365234375 + }, + { + "auxiliary_loss_clip": 0.01117963, + "auxiliary_loss_mlp": 0.01026156, + "balance_loss_clip": 0.96999955, + "balance_loss_mlp": 1.01937878, + "epoch": 0.9352492033908495, + "flos": 18799990156800.0, + "grad_norm": 2.0411126207964783, + "language_loss": 0.69312632, + "learning_rate": 4.374799770068849e-08, + "loss": 0.71456754, + "num_input_tokens_seen": 168112580, + "step": 7778, + "time_per_iteration": 2.688784122467041 + }, + { + "auxiliary_loss_clip": 0.01138781, + "auxiliary_loss_mlp": 0.01026565, + "balance_loss_clip": 1.01026297, + "balance_loss_mlp": 1.01950204, + "epoch": 0.9353694462814887, + "flos": 29530134241920.0, + "grad_norm": 2.237906504976051, + "language_loss": 0.74789011, + "learning_rate": 4.358610963605658e-08, + "loss": 0.76954359, + "num_input_tokens_seen": 168133030, + "step": 7779, + "time_per_iteration": 2.6611647605895996 + }, + { + "auxiliary_loss_clip": 0.01166944, + "auxiliary_loss_mlp": 0.01027286, + "balance_loss_clip": 1.04968643, + "balance_loss_mlp": 1.0196507, + "epoch": 0.9354896891721277, + "flos": 30665450390400.0, + "grad_norm": 3.4340740133365295, + "language_loss": 0.68608922, + "learning_rate": 4.342451835785677e-08, + "loss": 0.70803154, + "num_input_tokens_seen": 168153940, + "step": 7780, + "time_per_iteration": 2.73408842086792 + }, + { + "auxiliary_loss_clip": 0.01113751, + "auxiliary_loss_mlp": 0.01024728, + "balance_loss_clip": 0.96960306, + "balance_loss_mlp": 1.01760507, + "epoch": 0.9356099320627668, + "flos": 19463907191040.0, + "grad_norm": 1.5517095617882406, + "language_loss": 0.75168121, + "learning_rate": 4.3263223890601665e-08, + "loss": 0.77306604, + "num_input_tokens_seen": 168172650, + "step": 7781, + "time_per_iteration": 2.6131298542022705 + }, + { + "auxiliary_loss_clip": 0.01137881, + "auxiliary_loss_mlp": 0.00509018, + "balance_loss_clip": 1.0115881, + "balance_loss_mlp": 0.0, + "epoch": 0.9357301749534058, + "flos": 19098156954240.0, + "grad_norm": 1.7918477146669254, + "language_loss": 0.79607773, + "learning_rate": 4.31022262587597e-08, + "loss": 0.81254667, + "num_input_tokens_seen": 168191325, + "step": 7782, + "time_per_iteration": 2.867361307144165 + }, + { + "auxiliary_loss_clip": 0.011416, + "auxiliary_loss_mlp": 0.01028675, + "balance_loss_clip": 1.01074862, + "balance_loss_mlp": 1.02117717, + "epoch": 0.935850417844045, + "flos": 23550361776000.0, + "grad_norm": 1.7888200706086619, + "language_loss": 0.66131073, + "learning_rate": 4.2941525486754225e-08, + "loss": 0.68301356, + "num_input_tokens_seen": 168211645, + "step": 7783, + "time_per_iteration": 2.7769408226013184 + }, + { + "auxiliary_loss_clip": 0.01091912, + "auxiliary_loss_mlp": 0.01020923, + "balance_loss_clip": 0.93450248, + "balance_loss_mlp": 1.01418483, + "epoch": 0.935970660734684, + "flos": 18588333265920.0, + "grad_norm": 1.7649854035603776, + "language_loss": 0.79293418, + "learning_rate": 4.278112159896286e-08, + "loss": 0.81406254, + "num_input_tokens_seen": 168229485, + "step": 7784, + "time_per_iteration": 2.667055368423462 + }, + { + "auxiliary_loss_clip": 0.01110467, + "auxiliary_loss_mlp": 0.01020847, + "balance_loss_clip": 0.967924, + "balance_loss_mlp": 1.01380503, + "epoch": 0.9360909036253231, + "flos": 20631255292800.0, + "grad_norm": 1.8030064746681302, + "language_loss": 0.67491269, + "learning_rate": 4.2621014619719896e-08, + "loss": 0.69622582, + "num_input_tokens_seen": 168247250, + "step": 7785, + "time_per_iteration": 3.6387836933135986 + }, + { + "auxiliary_loss_clip": 0.01019191, + "auxiliary_loss_mlp": 0.01001044, + "balance_loss_clip": 0.9346801, + "balance_loss_mlp": 1.00009084, + "epoch": 0.9362111465159623, + "flos": 61791421052160.0, + "grad_norm": 0.7222964845134445, + "language_loss": 0.58610141, + "learning_rate": 4.246120457331215e-08, + "loss": 0.60630375, + "num_input_tokens_seen": 168309425, + "step": 7786, + "time_per_iteration": 4.1579742431640625 + }, + { + "auxiliary_loss_clip": 0.01110845, + "auxiliary_loss_mlp": 0.01025758, + "balance_loss_clip": 0.97174197, + "balance_loss_mlp": 1.01871312, + "epoch": 0.9363313894066013, + "flos": 24170395368960.0, + "grad_norm": 2.9114642051814044, + "language_loss": 0.71927714, + "learning_rate": 4.2301691483983325e-08, + "loss": 0.7406432, + "num_input_tokens_seen": 168329545, + "step": 7787, + "time_per_iteration": 2.8938286304473877 + }, + { + "auxiliary_loss_clip": 0.01141023, + "auxiliary_loss_mlp": 0.01033286, + "balance_loss_clip": 1.00816882, + "balance_loss_mlp": 1.02634239, + "epoch": 0.9364516322972404, + "flos": 20120354196480.0, + "grad_norm": 1.5586715180013133, + "language_loss": 0.75867116, + "learning_rate": 4.214247537593163e-08, + "loss": 0.78041422, + "num_input_tokens_seen": 168348795, + "step": 7788, + "time_per_iteration": 3.6760942935943604 + }, + { + "auxiliary_loss_clip": 0.01119326, + "auxiliary_loss_mlp": 0.01026668, + "balance_loss_clip": 0.97111976, + "balance_loss_mlp": 1.0200727, + "epoch": 0.9365718751878795, + "flos": 20703758895360.0, + "grad_norm": 1.910167672078907, + "language_loss": 0.80361521, + "learning_rate": 4.1983556273309293e-08, + "loss": 0.82507509, + "num_input_tokens_seen": 168367545, + "step": 7789, + "time_per_iteration": 2.8189117908477783 + }, + { + "auxiliary_loss_clip": 0.01165679, + "auxiliary_loss_mlp": 0.01027741, + "balance_loss_clip": 1.04870772, + "balance_loss_mlp": 1.01981926, + "epoch": 0.9366921180785186, + "flos": 18655270260480.0, + "grad_norm": 2.177943787400798, + "language_loss": 0.69096422, + "learning_rate": 4.182493420022526e-08, + "loss": 0.71289849, + "num_input_tokens_seen": 168383215, + "step": 7790, + "time_per_iteration": 2.612748861312866 + }, + { + "auxiliary_loss_clip": 0.01096301, + "auxiliary_loss_mlp": 0.01023187, + "balance_loss_clip": 0.93421984, + "balance_loss_mlp": 1.01667833, + "epoch": 0.9368123609691577, + "flos": 25774955815680.0, + "grad_norm": 1.8148514848493527, + "language_loss": 0.78509843, + "learning_rate": 4.166660918074139e-08, + "loss": 0.80629331, + "num_input_tokens_seen": 168403120, + "step": 7791, + "time_per_iteration": 2.7619917392730713 + }, + { + "auxiliary_loss_clip": 0.01088167, + "auxiliary_loss_mlp": 0.01024602, + "balance_loss_clip": 0.93156505, + "balance_loss_mlp": 1.0169847, + "epoch": 0.9369326038597968, + "flos": 25553386771200.0, + "grad_norm": 1.5550394536694907, + "language_loss": 0.73518717, + "learning_rate": 4.15085812388758e-08, + "loss": 0.75631487, + "num_input_tokens_seen": 168425340, + "step": 7792, + "time_per_iteration": 2.9506711959838867 + }, + { + "auxiliary_loss_clip": 0.01117196, + "auxiliary_loss_mlp": 0.0102773, + "balance_loss_clip": 0.97222877, + "balance_loss_mlp": 1.02054143, + "epoch": 0.9370528467504359, + "flos": 23220019370880.0, + "grad_norm": 1.696312087284186, + "language_loss": 0.78647107, + "learning_rate": 4.135085039860153e-08, + "loss": 0.80792028, + "num_input_tokens_seen": 168444740, + "step": 7793, + "time_per_iteration": 2.6875524520874023 + }, + { + "auxiliary_loss_clip": 0.01115433, + "auxiliary_loss_mlp": 0.01022824, + "balance_loss_clip": 0.97388268, + "balance_loss_mlp": 1.0154804, + "epoch": 0.9371730896410749, + "flos": 24967468120320.0, + "grad_norm": 2.8894482790376412, + "language_loss": 0.7847811, + "learning_rate": 4.1193416683845906e-08, + "loss": 0.80616367, + "num_input_tokens_seen": 168463670, + "step": 7794, + "time_per_iteration": 2.703113317489624 + }, + { + "auxiliary_loss_clip": 0.01094384, + "auxiliary_loss_mlp": 0.0102266, + "balance_loss_clip": 0.93436551, + "balance_loss_mlp": 1.01653528, + "epoch": 0.9372933325317141, + "flos": 15553091134080.0, + "grad_norm": 2.337954967147046, + "language_loss": 0.83737737, + "learning_rate": 4.103628011849136e-08, + "loss": 0.85854781, + "num_input_tokens_seen": 168479030, + "step": 7795, + "time_per_iteration": 2.6945316791534424 + }, + { + "auxiliary_loss_clip": 0.01118494, + "auxiliary_loss_mlp": 0.01026992, + "balance_loss_clip": 0.97234571, + "balance_loss_mlp": 1.01977944, + "epoch": 0.9374135754223532, + "flos": 21871861182720.0, + "grad_norm": 1.875061875269362, + "language_loss": 0.75974095, + "learning_rate": 4.0879440726375506e-08, + "loss": 0.78119582, + "num_input_tokens_seen": 168496815, + "step": 7796, + "time_per_iteration": 3.521573543548584 + }, + { + "auxiliary_loss_clip": 0.01112335, + "auxiliary_loss_mlp": 0.01023901, + "balance_loss_clip": 0.96714538, + "balance_loss_mlp": 1.01649237, + "epoch": 0.9375338183129922, + "flos": 22631048064000.0, + "grad_norm": 2.5339542989096895, + "language_loss": 0.55779439, + "learning_rate": 4.0722898531291074e-08, + "loss": 0.57915676, + "num_input_tokens_seen": 168514055, + "step": 7797, + "time_per_iteration": 2.641127109527588 + }, + { + "auxiliary_loss_clip": 0.01124242, + "auxiliary_loss_mlp": 0.01029779, + "balance_loss_clip": 0.97077692, + "balance_loss_mlp": 1.02217388, + "epoch": 0.9376540612036314, + "flos": 26104292640000.0, + "grad_norm": 2.081968039326385, + "language_loss": 0.76770949, + "learning_rate": 4.0566653556985295e-08, + "loss": 0.78924966, + "num_input_tokens_seen": 168534600, + "step": 7798, + "time_per_iteration": 2.7438039779663086 + }, + { + "auxiliary_loss_clip": 0.01027826, + "auxiliary_loss_mlp": 0.01025988, + "balance_loss_clip": 0.81776071, + "balance_loss_mlp": 1.01816821, + "epoch": 0.9377743040942704, + "flos": 19717580016000.0, + "grad_norm": 2.2482666667410465, + "language_loss": 0.8195051, + "learning_rate": 4.0410705827159886e-08, + "loss": 0.84004319, + "num_input_tokens_seen": 168551895, + "step": 7799, + "time_per_iteration": 2.7130203247070312 + }, + { + "auxiliary_loss_clip": 0.01114618, + "auxiliary_loss_mlp": 0.01032604, + "balance_loss_clip": 0.96880293, + "balance_loss_mlp": 1.02520108, + "epoch": 0.9378945469849095, + "flos": 15267530010240.0, + "grad_norm": 3.050342683821999, + "language_loss": 0.71290517, + "learning_rate": 4.0255055365472356e-08, + "loss": 0.73437738, + "num_input_tokens_seen": 168569990, + "step": 7800, + "time_per_iteration": 2.67348575592041 + }, + { + "auxiliary_loss_clip": 0.01046365, + "auxiliary_loss_mlp": 0.01025946, + "balance_loss_clip": 0.85207915, + "balance_loss_mlp": 1.01870966, + "epoch": 0.9380147898755486, + "flos": 20591394174720.0, + "grad_norm": 2.042958054150751, + "language_loss": 0.75066251, + "learning_rate": 4.009970219553471e-08, + "loss": 0.77138561, + "num_input_tokens_seen": 168586940, + "step": 7801, + "time_per_iteration": 2.8055925369262695 + }, + { + "auxiliary_loss_clip": 0.01144064, + "auxiliary_loss_mlp": 0.01030534, + "balance_loss_clip": 1.00990415, + "balance_loss_mlp": 1.02296424, + "epoch": 0.9381350327661877, + "flos": 26281116316800.0, + "grad_norm": 2.4916146634161302, + "language_loss": 0.76457345, + "learning_rate": 3.99446463409141e-08, + "loss": 0.78631949, + "num_input_tokens_seen": 168604795, + "step": 7802, + "time_per_iteration": 2.744377374649048 + }, + { + "auxiliary_loss_clip": 0.01143574, + "auxiliary_loss_mlp": 0.01029697, + "balance_loss_clip": 1.00703859, + "balance_loss_mlp": 1.02240181, + "epoch": 0.9382552756568268, + "flos": 23586344225280.0, + "grad_norm": 2.085733987774827, + "language_loss": 0.68857187, + "learning_rate": 3.978988782513215e-08, + "loss": 0.71030462, + "num_input_tokens_seen": 168622290, + "step": 7803, + "time_per_iteration": 2.8248934745788574 + }, + { + "auxiliary_loss_clip": 0.01145366, + "auxiliary_loss_mlp": 0.01029047, + "balance_loss_clip": 1.01086521, + "balance_loss_mlp": 1.02163804, + "epoch": 0.9383755185474659, + "flos": 28438809275520.0, + "grad_norm": 1.9201165681662262, + "language_loss": 0.76186776, + "learning_rate": 3.963542667166586e-08, + "loss": 0.78361183, + "num_input_tokens_seen": 168642395, + "step": 7804, + "time_per_iteration": 2.69731068611145 + }, + { + "auxiliary_loss_clip": 0.01099203, + "auxiliary_loss_mlp": 0.01025258, + "balance_loss_clip": 0.93691003, + "balance_loss_mlp": 1.01829672, + "epoch": 0.938495761438105, + "flos": 20449583280000.0, + "grad_norm": 2.505369091222347, + "language_loss": 0.68579531, + "learning_rate": 3.9481262903946486e-08, + "loss": 0.70703989, + "num_input_tokens_seen": 168661840, + "step": 7805, + "time_per_iteration": 2.621415138244629 + }, + { + "auxiliary_loss_clip": 0.00987663, + "auxiliary_loss_mlp": 0.01005224, + "balance_loss_clip": 0.86241078, + "balance_loss_mlp": 1.00417471, + "epoch": 0.938616004328744, + "flos": 69302711658240.0, + "grad_norm": 0.8051620781060518, + "language_loss": 0.54507589, + "learning_rate": 3.932739654536066e-08, + "loss": 0.56500471, + "num_input_tokens_seen": 168724540, + "step": 7806, + "time_per_iteration": 3.2586050033569336 + }, + { + "auxiliary_loss_clip": 0.0114062, + "auxiliary_loss_mlp": 0.01023617, + "balance_loss_clip": 1.01106894, + "balance_loss_mlp": 1.01674128, + "epoch": 0.9387362472193832, + "flos": 18911636605440.0, + "grad_norm": 2.395790360119612, + "language_loss": 0.74198604, + "learning_rate": 3.917382761925014e-08, + "loss": 0.76362836, + "num_input_tokens_seen": 168740375, + "step": 7807, + "time_per_iteration": 2.74080491065979 + }, + { + "auxiliary_loss_clip": 0.0113531, + "auxiliary_loss_mlp": 0.01020232, + "balance_loss_clip": 1.00948751, + "balance_loss_mlp": 1.01366019, + "epoch": 0.9388564901100223, + "flos": 26501967089280.0, + "grad_norm": 1.9492937603772174, + "language_loss": 0.7907331, + "learning_rate": 3.9020556148910754e-08, + "loss": 0.81228852, + "num_input_tokens_seen": 168759730, + "step": 7808, + "time_per_iteration": 2.9219508171081543 + }, + { + "auxiliary_loss_clip": 0.01023265, + "auxiliary_loss_mlp": 0.01000262, + "balance_loss_clip": 0.93509269, + "balance_loss_mlp": 0.99940968, + "epoch": 0.9389767330006613, + "flos": 58941083157120.0, + "grad_norm": 0.7104029558891923, + "language_loss": 0.56665128, + "learning_rate": 3.8867582157593895e-08, + "loss": 0.58688653, + "num_input_tokens_seen": 168813935, + "step": 7809, + "time_per_iteration": 3.1358110904693604 + }, + { + "auxiliary_loss_clip": 0.01137757, + "auxiliary_loss_mlp": 0.01024897, + "balance_loss_clip": 1.01111293, + "balance_loss_mlp": 1.01820326, + "epoch": 0.9390969758913005, + "flos": 31102554994560.0, + "grad_norm": 1.8592283629310056, + "language_loss": 0.76529831, + "learning_rate": 3.871490566850544e-08, + "loss": 0.78692484, + "num_input_tokens_seen": 168838145, + "step": 7810, + "time_per_iteration": 2.7623305320739746 + }, + { + "auxiliary_loss_clip": 0.01111454, + "auxiliary_loss_mlp": 0.01023396, + "balance_loss_clip": 0.97003895, + "balance_loss_mlp": 1.01652384, + "epoch": 0.9392172187819395, + "flos": 22419391173120.0, + "grad_norm": 2.546976605175136, + "language_loss": 0.70920318, + "learning_rate": 3.856252670480642e-08, + "loss": 0.73055166, + "num_input_tokens_seen": 168856805, + "step": 7811, + "time_per_iteration": 3.5792317390441895 + }, + { + "auxiliary_loss_clip": 0.0111401, + "auxiliary_loss_mlp": 0.01030262, + "balance_loss_clip": 0.96840626, + "balance_loss_mlp": 1.02280593, + "epoch": 0.9393374616725786, + "flos": 19719483436800.0, + "grad_norm": 1.81116868251622, + "language_loss": 0.81410879, + "learning_rate": 3.841044528961279e-08, + "loss": 0.8355515, + "num_input_tokens_seen": 168874600, + "step": 7812, + "time_per_iteration": 3.578618288040161 + }, + { + "auxiliary_loss_clip": 0.01164954, + "auxiliary_loss_mlp": 0.01026471, + "balance_loss_clip": 1.04700804, + "balance_loss_mlp": 1.01862693, + "epoch": 0.9394577045632178, + "flos": 24170215800960.0, + "grad_norm": 1.9344102280330275, + "language_loss": 0.78606248, + "learning_rate": 3.825866144599477e-08, + "loss": 0.80797672, + "num_input_tokens_seen": 168893655, + "step": 7813, + "time_per_iteration": 3.622910499572754 + }, + { + "auxiliary_loss_clip": 0.01115238, + "auxiliary_loss_mlp": 0.01023986, + "balance_loss_clip": 0.9691174, + "balance_loss_mlp": 1.01654768, + "epoch": 0.9395779474538568, + "flos": 19023929498880.0, + "grad_norm": 2.09267590517992, + "language_loss": 0.75329792, + "learning_rate": 3.8107175196978145e-08, + "loss": 0.77469015, + "num_input_tokens_seen": 168909960, + "step": 7814, + "time_per_iteration": 2.783902168273926 + }, + { + "auxiliary_loss_clip": 0.01096243, + "auxiliary_loss_mlp": 0.01027315, + "balance_loss_clip": 0.93776381, + "balance_loss_mlp": 1.02040958, + "epoch": 0.9396981903444959, + "flos": 14319129260160.0, + "grad_norm": 1.805308808402793, + "language_loss": 0.7659179, + "learning_rate": 3.7955986565542996e-08, + "loss": 0.78715348, + "num_input_tokens_seen": 168928040, + "step": 7815, + "time_per_iteration": 2.7297780513763428 + }, + { + "auxiliary_loss_clip": 0.01091579, + "auxiliary_loss_mlp": 0.01022049, + "balance_loss_clip": 0.93093252, + "balance_loss_mlp": 1.01514053, + "epoch": 0.9398184332351349, + "flos": 34787564202240.0, + "grad_norm": 1.9264604868110846, + "language_loss": 0.68015474, + "learning_rate": 3.780509557462497e-08, + "loss": 0.70129102, + "num_input_tokens_seen": 168948240, + "step": 7816, + "time_per_iteration": 2.798429012298584 + }, + { + "auxiliary_loss_clip": 0.01116613, + "auxiliary_loss_mlp": 0.01027022, + "balance_loss_clip": 0.97021842, + "balance_loss_mlp": 1.01967227, + "epoch": 0.9399386761257741, + "flos": 25372253462400.0, + "grad_norm": 1.9097590663190838, + "language_loss": 0.75572687, + "learning_rate": 3.765450224711375e-08, + "loss": 0.77716321, + "num_input_tokens_seen": 168968745, + "step": 7817, + "time_per_iteration": 2.714230537414551 + }, + { + "auxiliary_loss_clip": 0.01115093, + "auxiliary_loss_mlp": 0.01025093, + "balance_loss_clip": 0.97389114, + "balance_loss_mlp": 1.0184381, + "epoch": 0.9400589190164131, + "flos": 27304965584640.0, + "grad_norm": 1.6440197222015842, + "language_loss": 0.79714489, + "learning_rate": 3.750420660585396e-08, + "loss": 0.81854677, + "num_input_tokens_seen": 168990685, + "step": 7818, + "time_per_iteration": 2.7239458560943604 + }, + { + "auxiliary_loss_clip": 0.0116217, + "auxiliary_loss_mlp": 0.01026641, + "balance_loss_clip": 1.04786658, + "balance_loss_mlp": 1.01945233, + "epoch": 0.9401791619070522, + "flos": 23399859790080.0, + "grad_norm": 1.6680984605166476, + "language_loss": 0.79669327, + "learning_rate": 3.735420867364603e-08, + "loss": 0.81858134, + "num_input_tokens_seen": 169011665, + "step": 7819, + "time_per_iteration": 2.599271059036255 + }, + { + "auxiliary_loss_clip": 0.01038905, + "auxiliary_loss_mlp": 0.01019476, + "balance_loss_clip": 0.85262126, + "balance_loss_mlp": 1.01308894, + "epoch": 0.9402994047976914, + "flos": 35881403120640.0, + "grad_norm": 1.4869536991842507, + "language_loss": 0.61419904, + "learning_rate": 3.7204508473244186e-08, + "loss": 0.63478279, + "num_input_tokens_seen": 169035290, + "step": 7820, + "time_per_iteration": 2.8761813640594482 + }, + { + "auxiliary_loss_clip": 0.01022049, + "auxiliary_loss_mlp": 0.01022465, + "balance_loss_clip": 0.81790471, + "balance_loss_mlp": 1.01595926, + "epoch": 0.9404196476883304, + "flos": 22236821320320.0, + "grad_norm": 2.549484967316978, + "language_loss": 0.69136357, + "learning_rate": 3.7055106027357395e-08, + "loss": 0.71180868, + "num_input_tokens_seen": 169055155, + "step": 7821, + "time_per_iteration": 2.9149765968322754 + }, + { + "auxiliary_loss_clip": 0.01135141, + "auxiliary_loss_mlp": 0.01021928, + "balance_loss_clip": 1.00812721, + "balance_loss_mlp": 1.01483536, + "epoch": 0.9405398905789695, + "flos": 18915802583040.0, + "grad_norm": 1.9980931713576247, + "language_loss": 0.72399485, + "learning_rate": 3.690600135865063e-08, + "loss": 0.74556553, + "num_input_tokens_seen": 169072080, + "step": 7822, + "time_per_iteration": 3.5355989933013916 + }, + { + "auxiliary_loss_clip": 0.00983459, + "auxiliary_loss_mlp": 0.01000774, + "balance_loss_clip": 0.86187768, + "balance_loss_mlp": 0.99971867, + "epoch": 0.9406601334696086, + "flos": 70274130048000.0, + "grad_norm": 0.783833089018743, + "language_loss": 0.58083928, + "learning_rate": 3.675719448974246e-08, + "loss": 0.60068154, + "num_input_tokens_seen": 169137170, + "step": 7823, + "time_per_iteration": 3.3162853717803955 + }, + { + "auxiliary_loss_clip": 0.01068356, + "auxiliary_loss_mlp": 0.00508906, + "balance_loss_clip": 0.8968969, + "balance_loss_mlp": 0.0, + "epoch": 0.9407803763602477, + "flos": 22165071903360.0, + "grad_norm": 2.126447207289669, + "language_loss": 0.604415, + "learning_rate": 3.6608685443207054e-08, + "loss": 0.62018764, + "num_input_tokens_seen": 169156320, + "step": 7824, + "time_per_iteration": 2.8120856285095215 + }, + { + "auxiliary_loss_clip": 0.01093212, + "auxiliary_loss_mlp": 0.01023964, + "balance_loss_clip": 0.93148243, + "balance_loss_mlp": 1.01702571, + "epoch": 0.9409006192508867, + "flos": 18879496911360.0, + "grad_norm": 5.272453426796476, + "language_loss": 0.66712713, + "learning_rate": 3.646047424157306e-08, + "loss": 0.68829888, + "num_input_tokens_seen": 169173295, + "step": 7825, + "time_per_iteration": 2.7188920974731445 + }, + { + "auxiliary_loss_clip": 0.0111719, + "auxiliary_loss_mlp": 0.01025562, + "balance_loss_clip": 0.97243178, + "balance_loss_mlp": 1.01787019, + "epoch": 0.9410208621415259, + "flos": 23368258800000.0, + "grad_norm": 6.996038292701496, + "language_loss": 0.68433762, + "learning_rate": 3.631256090732382e-08, + "loss": 0.70576513, + "num_input_tokens_seen": 169193755, + "step": 7826, + "time_per_iteration": 2.6943442821502686 + }, + { + "auxiliary_loss_clip": 0.0109666, + "auxiliary_loss_mlp": 0.01026222, + "balance_loss_clip": 0.9359647, + "balance_loss_mlp": 1.01984167, + "epoch": 0.941141105032165, + "flos": 22742227635840.0, + "grad_norm": 1.6190441549548789, + "language_loss": 0.82543361, + "learning_rate": 3.6164945462897833e-08, + "loss": 0.84666246, + "num_input_tokens_seen": 169213045, + "step": 7827, + "time_per_iteration": 2.6814074516296387 + }, + { + "auxiliary_loss_clip": 0.01138172, + "auxiliary_loss_mlp": 0.00508613, + "balance_loss_clip": 1.00963295, + "balance_loss_mlp": 0.0, + "epoch": 0.941261347922804, + "flos": 20704908130560.0, + "grad_norm": 1.7514602553366208, + "language_loss": 0.75679684, + "learning_rate": 3.6017627930687856e-08, + "loss": 0.77326471, + "num_input_tokens_seen": 169232870, + "step": 7828, + "time_per_iteration": 2.6925864219665527 + }, + { + "auxiliary_loss_clip": 0.01066296, + "auxiliary_loss_mlp": 0.01028993, + "balance_loss_clip": 0.89039141, + "balance_loss_mlp": 1.02174544, + "epoch": 0.9413815908134432, + "flos": 19421998997760.0, + "grad_norm": 1.959479488998453, + "language_loss": 0.76713496, + "learning_rate": 3.587060833304267e-08, + "loss": 0.78808784, + "num_input_tokens_seen": 169251060, + "step": 7829, + "time_per_iteration": 2.743157148361206 + }, + { + "auxiliary_loss_clip": 0.01142471, + "auxiliary_loss_mlp": 0.01031457, + "balance_loss_clip": 1.00974441, + "balance_loss_mlp": 1.02462077, + "epoch": 0.9415018337040822, + "flos": 17493452853120.0, + "grad_norm": 2.0931324648496403, + "language_loss": 0.63764656, + "learning_rate": 3.5723886692264225e-08, + "loss": 0.65938586, + "num_input_tokens_seen": 169268600, + "step": 7830, + "time_per_iteration": 2.6427111625671387 + }, + { + "auxiliary_loss_clip": 0.01114395, + "auxiliary_loss_mlp": 0.01024914, + "balance_loss_clip": 0.96951443, + "balance_loss_mlp": 1.01834011, + "epoch": 0.9416220765947213, + "flos": 31831613343360.0, + "grad_norm": 2.2557886306136106, + "language_loss": 0.61733305, + "learning_rate": 3.557746303061071e-08, + "loss": 0.63872612, + "num_input_tokens_seen": 169290355, + "step": 7831, + "time_per_iteration": 2.784083843231201 + }, + { + "auxiliary_loss_clip": 0.01115245, + "auxiliary_loss_mlp": 0.01025422, + "balance_loss_clip": 0.97028661, + "balance_loss_mlp": 1.01814461, + "epoch": 0.9417423194853605, + "flos": 23511973115520.0, + "grad_norm": 1.7386363126955144, + "language_loss": 0.72154617, + "learning_rate": 3.543133737029391e-08, + "loss": 0.74295282, + "num_input_tokens_seen": 169310865, + "step": 7832, + "time_per_iteration": 2.7083699703216553 + }, + { + "auxiliary_loss_clip": 0.01145336, + "auxiliary_loss_mlp": 0.01027413, + "balance_loss_clip": 1.01169538, + "balance_loss_mlp": 1.02004278, + "epoch": 0.9418625623759995, + "flos": 23915106432000.0, + "grad_norm": 2.1307047785225843, + "language_loss": 0.69058883, + "learning_rate": 3.5285509733481214e-08, + "loss": 0.71231639, + "num_input_tokens_seen": 169330590, + "step": 7833, + "time_per_iteration": 2.679961919784546 + }, + { + "auxiliary_loss_clip": 0.01137233, + "auxiliary_loss_mlp": 0.01025313, + "balance_loss_clip": 1.00785232, + "balance_loss_mlp": 1.01768327, + "epoch": 0.9419828052666386, + "flos": 18076965292800.0, + "grad_norm": 1.7895146908377904, + "language_loss": 0.76372582, + "learning_rate": 3.513998014229469e-08, + "loss": 0.78535128, + "num_input_tokens_seen": 169349540, + "step": 7834, + "time_per_iteration": 2.6117725372314453 + }, + { + "auxiliary_loss_clip": 0.01121117, + "auxiliary_loss_mlp": 0.01023058, + "balance_loss_clip": 0.97400391, + "balance_loss_mlp": 1.01652479, + "epoch": 0.9421030481572777, + "flos": 17712328377600.0, + "grad_norm": 2.2297119252425293, + "language_loss": 0.86300933, + "learning_rate": 3.499474861881069e-08, + "loss": 0.88445103, + "num_input_tokens_seen": 169366765, + "step": 7835, + "time_per_iteration": 2.7294535636901855 + }, + { + "auxiliary_loss_clip": 0.0104853, + "auxiliary_loss_mlp": 0.01028402, + "balance_loss_clip": 0.85666031, + "balance_loss_mlp": 1.02094865, + "epoch": 0.9422232910479168, + "flos": 20194114775040.0, + "grad_norm": 2.013145397598368, + "language_loss": 0.67706275, + "learning_rate": 3.4849815185061136e-08, + "loss": 0.69783205, + "num_input_tokens_seen": 169386655, + "step": 7836, + "time_per_iteration": 2.9711973667144775 + }, + { + "auxiliary_loss_clip": 0.01138849, + "auxiliary_loss_mlp": 0.01025735, + "balance_loss_clip": 1.00750887, + "balance_loss_mlp": 1.01878774, + "epoch": 0.9423435339385559, + "flos": 18442571875200.0, + "grad_norm": 1.8388223607503018, + "language_loss": 0.76076365, + "learning_rate": 3.470517986303223e-08, + "loss": 0.78240943, + "num_input_tokens_seen": 169405640, + "step": 7837, + "time_per_iteration": 3.5509231090545654 + }, + { + "auxiliary_loss_clip": 0.01092703, + "auxiliary_loss_mlp": 0.0102486, + "balance_loss_clip": 0.93438393, + "balance_loss_mlp": 1.01816952, + "epoch": 0.942463776829195, + "flos": 20080636732800.0, + "grad_norm": 2.9210115850539586, + "language_loss": 0.79223329, + "learning_rate": 3.4560842674664856e-08, + "loss": 0.81340897, + "num_input_tokens_seen": 169424155, + "step": 7838, + "time_per_iteration": 3.582810878753662 + }, + { + "auxiliary_loss_clip": 0.01141162, + "auxiliary_loss_mlp": 0.01024544, + "balance_loss_clip": 1.00736475, + "balance_loss_mlp": 1.0173676, + "epoch": 0.9425840197198341, + "flos": 22636255536000.0, + "grad_norm": 1.8774265396182677, + "language_loss": 0.75051606, + "learning_rate": 3.441680364185506e-08, + "loss": 0.77217311, + "num_input_tokens_seen": 169444025, + "step": 7839, + "time_per_iteration": 3.628129243850708 + }, + { + "auxiliary_loss_clip": 0.01119396, + "auxiliary_loss_mlp": 0.0103392, + "balance_loss_clip": 0.97267348, + "balance_loss_mlp": 1.02659488, + "epoch": 0.9427042626104731, + "flos": 19937892084480.0, + "grad_norm": 2.204550795075462, + "language_loss": 0.74722987, + "learning_rate": 3.427306278645314e-08, + "loss": 0.76876301, + "num_input_tokens_seen": 169462480, + "step": 7840, + "time_per_iteration": 2.65205717086792 + }, + { + "auxiliary_loss_clip": 0.01072865, + "auxiliary_loss_mlp": 0.01022242, + "balance_loss_clip": 0.89497757, + "balance_loss_mlp": 1.01516438, + "epoch": 0.9428245055011123, + "flos": 22856998567680.0, + "grad_norm": 1.7678265141349196, + "language_loss": 0.7309233, + "learning_rate": 3.4129620130264767e-08, + "loss": 0.75187433, + "num_input_tokens_seen": 169480840, + "step": 7841, + "time_per_iteration": 2.760251045227051 + }, + { + "auxiliary_loss_clip": 0.01120349, + "auxiliary_loss_mlp": 0.00509145, + "balance_loss_clip": 0.97228265, + "balance_loss_mlp": 0.0, + "epoch": 0.9429447483917514, + "flos": 20951757371520.0, + "grad_norm": 2.1779347687023494, + "language_loss": 0.78210694, + "learning_rate": 3.398647569505009e-08, + "loss": 0.79840189, + "num_input_tokens_seen": 169498265, + "step": 7842, + "time_per_iteration": 2.6419174671173096 + }, + { + "auxiliary_loss_clip": 0.01099139, + "auxiliary_loss_mlp": 0.01023733, + "balance_loss_clip": 0.93296528, + "balance_loss_mlp": 1.01628876, + "epoch": 0.9430649912823904, + "flos": 18843658116480.0, + "grad_norm": 2.328289065739177, + "language_loss": 0.7455461, + "learning_rate": 3.384362950252373e-08, + "loss": 0.76677483, + "num_input_tokens_seen": 169515235, + "step": 7843, + "time_per_iteration": 2.718350887298584 + }, + { + "auxiliary_loss_clip": 0.01114393, + "auxiliary_loss_mlp": 0.01024869, + "balance_loss_clip": 0.96997166, + "balance_loss_mlp": 1.01801467, + "epoch": 0.9431852341730296, + "flos": 32556038837760.0, + "grad_norm": 1.8368861591835588, + "language_loss": 0.57177931, + "learning_rate": 3.3701081574355473e-08, + "loss": 0.59317195, + "num_input_tokens_seen": 169537195, + "step": 7844, + "time_per_iteration": 2.765493154525757 + }, + { + "auxiliary_loss_clip": 0.01022644, + "auxiliary_loss_mlp": 0.01001031, + "balance_loss_clip": 0.93523198, + "balance_loss_mlp": 1.00010109, + "epoch": 0.9433054770636686, + "flos": 66904490252160.0, + "grad_norm": 0.7370870270166717, + "language_loss": 0.51643658, + "learning_rate": 3.3558831932169796e-08, + "loss": 0.53667331, + "num_input_tokens_seen": 169605865, + "step": 7845, + "time_per_iteration": 3.3259897232055664 + }, + { + "auxiliary_loss_clip": 0.0113848, + "auxiliary_loss_mlp": 0.01022017, + "balance_loss_clip": 1.00921416, + "balance_loss_mlp": 1.01450109, + "epoch": 0.9434257199543077, + "flos": 26140346916480.0, + "grad_norm": 1.7418450282551667, + "language_loss": 0.88425702, + "learning_rate": 3.341688059754588e-08, + "loss": 0.90586197, + "num_input_tokens_seen": 169621520, + "step": 7846, + "time_per_iteration": 2.65655255317688 + }, + { + "auxiliary_loss_clip": 0.01100551, + "auxiliary_loss_mlp": 0.00509006, + "balance_loss_clip": 0.93179542, + "balance_loss_mlp": 0.0, + "epoch": 0.9435459628449467, + "flos": 25003486483200.0, + "grad_norm": 2.179118422779003, + "language_loss": 0.78063655, + "learning_rate": 3.327522759201762e-08, + "loss": 0.79673213, + "num_input_tokens_seen": 169641390, + "step": 7847, + "time_per_iteration": 2.8265092372894287 + }, + { + "auxiliary_loss_clip": 0.01093279, + "auxiliary_loss_mlp": 0.01028065, + "balance_loss_clip": 0.93535674, + "balance_loss_mlp": 1.02110279, + "epoch": 0.9436662057355859, + "flos": 22163240309760.0, + "grad_norm": 2.6816944766337487, + "language_loss": 0.67081046, + "learning_rate": 3.313387293707359e-08, + "loss": 0.69202387, + "num_input_tokens_seen": 169660095, + "step": 7848, + "time_per_iteration": 3.618273973464966 + }, + { + "auxiliary_loss_clip": 0.01090394, + "auxiliary_loss_mlp": 0.01025225, + "balance_loss_clip": 0.93451649, + "balance_loss_mlp": 1.01776814, + "epoch": 0.943786448626225, + "flos": 20118522602880.0, + "grad_norm": 1.822247907183975, + "language_loss": 0.68729788, + "learning_rate": 3.29928166541571e-08, + "loss": 0.70845401, + "num_input_tokens_seen": 169679050, + "step": 7849, + "time_per_iteration": 2.7822325229644775 + }, + { + "auxiliary_loss_clip": 0.01110318, + "auxiliary_loss_mlp": 0.01021369, + "balance_loss_clip": 0.97079086, + "balance_loss_mlp": 1.01422811, + "epoch": 0.943906691516864, + "flos": 22090808534400.0, + "grad_norm": 2.2811123643776168, + "language_loss": 0.80808228, + "learning_rate": 3.2852058764666346e-08, + "loss": 0.82939917, + "num_input_tokens_seen": 169698150, + "step": 7850, + "time_per_iteration": 2.727673053741455 + }, + { + "auxiliary_loss_clip": 0.01083524, + "auxiliary_loss_mlp": 0.01024728, + "balance_loss_clip": 0.93533355, + "balance_loss_mlp": 1.01782906, + "epoch": 0.9440269344075032, + "flos": 35298501212160.0, + "grad_norm": 2.742937755566747, + "language_loss": 0.68283617, + "learning_rate": 3.2711599289954264e-08, + "loss": 0.7039187, + "num_input_tokens_seen": 169722185, + "step": 7851, + "time_per_iteration": 2.761075735092163 + }, + { + "auxiliary_loss_clip": 0.01049666, + "auxiliary_loss_mlp": 0.01025638, + "balance_loss_clip": 0.85786843, + "balance_loss_mlp": 1.0184617, + "epoch": 0.9441471772981422, + "flos": 19238136255360.0, + "grad_norm": 1.9248078278267224, + "language_loss": 0.77650821, + "learning_rate": 3.257143825132847e-08, + "loss": 0.79726124, + "num_input_tokens_seen": 169740355, + "step": 7852, + "time_per_iteration": 2.9052281379699707 + }, + { + "auxiliary_loss_clip": 0.01118683, + "auxiliary_loss_mlp": 0.0102185, + "balance_loss_clip": 0.97209132, + "balance_loss_mlp": 1.01543331, + "epoch": 0.9442674201887813, + "flos": 25739799379200.0, + "grad_norm": 9.391318665517304, + "language_loss": 0.76026964, + "learning_rate": 3.243157567005106e-08, + "loss": 0.78167492, + "num_input_tokens_seen": 169758535, + "step": 7853, + "time_per_iteration": 2.687551736831665 + }, + { + "auxiliary_loss_clip": 0.01169311, + "auxiliary_loss_mlp": 0.0102518, + "balance_loss_clip": 1.05133545, + "balance_loss_mlp": 1.0179646, + "epoch": 0.9443876630794205, + "flos": 15523321737600.0, + "grad_norm": 2.368352774109953, + "language_loss": 0.63877529, + "learning_rate": 3.2292011567339296e-08, + "loss": 0.66072017, + "num_input_tokens_seen": 169776340, + "step": 7854, + "time_per_iteration": 2.571603775024414 + }, + { + "auxiliary_loss_clip": 0.01141416, + "auxiliary_loss_mlp": 0.00508699, + "balance_loss_clip": 1.00908983, + "balance_loss_mlp": 0.0, + "epoch": 0.9445079059700595, + "flos": 13400821128960.0, + "grad_norm": 2.070855234205772, + "language_loss": 0.56036723, + "learning_rate": 3.21527459643649e-08, + "loss": 0.57686841, + "num_input_tokens_seen": 169793225, + "step": 7855, + "time_per_iteration": 2.619028329849243 + }, + { + "auxiliary_loss_clip": 0.01143869, + "auxiliary_loss_mlp": 0.01026794, + "balance_loss_clip": 1.01080501, + "balance_loss_mlp": 1.01965046, + "epoch": 0.9446281488606986, + "flos": 23659242877440.0, + "grad_norm": 2.0659858850053925, + "language_loss": 0.74289, + "learning_rate": 3.2013778882254536e-08, + "loss": 0.76459658, + "num_input_tokens_seen": 169812020, + "step": 7856, + "time_per_iteration": 2.6342413425445557 + }, + { + "auxiliary_loss_clip": 0.0113365, + "auxiliary_loss_mlp": 0.01024877, + "balance_loss_clip": 1.00745046, + "balance_loss_mlp": 1.01791549, + "epoch": 0.9447483917513377, + "flos": 25557337267200.0, + "grad_norm": 1.6935941947693878, + "language_loss": 0.75982606, + "learning_rate": 3.1875110342088676e-08, + "loss": 0.78141141, + "num_input_tokens_seen": 169833470, + "step": 7857, + "time_per_iteration": 2.7493679523468018 + }, + { + "auxiliary_loss_clip": 0.0111074, + "auxiliary_loss_mlp": 0.01025479, + "balance_loss_clip": 0.97091997, + "balance_loss_mlp": 1.01854968, + "epoch": 0.9448686346419768, + "flos": 24535463247360.0, + "grad_norm": 1.6195694834982768, + "language_loss": 0.65606892, + "learning_rate": 3.1736740364904035e-08, + "loss": 0.67743111, + "num_input_tokens_seen": 169854000, + "step": 7858, + "time_per_iteration": 2.7599356174468994 + }, + { + "auxiliary_loss_clip": 0.01067785, + "auxiliary_loss_mlp": 0.00509257, + "balance_loss_clip": 0.89291859, + "balance_loss_mlp": 0.0, + "epoch": 0.9449888775326158, + "flos": 14721256995840.0, + "grad_norm": 3.3124760657168673, + "language_loss": 0.77225071, + "learning_rate": 3.159866897169094e-08, + "loss": 0.78802109, + "num_input_tokens_seen": 169872200, + "step": 7859, + "time_per_iteration": 2.71537446975708 + }, + { + "auxiliary_loss_clip": 0.01098598, + "auxiliary_loss_mlp": 0.01026452, + "balance_loss_clip": 0.93191719, + "balance_loss_mlp": 1.01965964, + "epoch": 0.945109120423255, + "flos": 15447873219840.0, + "grad_norm": 1.9021137306219646, + "language_loss": 0.75612819, + "learning_rate": 3.146089618339487e-08, + "loss": 0.77737868, + "num_input_tokens_seen": 169889055, + "step": 7860, + "time_per_iteration": 2.600919246673584 + }, + { + "auxiliary_loss_clip": 0.01095191, + "auxiliary_loss_mlp": 0.01026427, + "balance_loss_clip": 0.93340868, + "balance_loss_mlp": 1.01931, + "epoch": 0.9452293633138941, + "flos": 25448097029760.0, + "grad_norm": 1.7992729957228077, + "language_loss": 0.68244469, + "learning_rate": 3.132342202091554e-08, + "loss": 0.70366085, + "num_input_tokens_seen": 169909280, + "step": 7861, + "time_per_iteration": 2.7700366973876953 + }, + { + "auxiliary_loss_clip": 0.01167404, + "auxiliary_loss_mlp": 0.01023473, + "balance_loss_clip": 1.05007029, + "balance_loss_mlp": 1.01637411, + "epoch": 0.9453496062045331, + "flos": 21215342350080.0, + "grad_norm": 2.5309616090614986, + "language_loss": 0.68419433, + "learning_rate": 3.1186246505107595e-08, + "loss": 0.70610315, + "num_input_tokens_seen": 169928420, + "step": 7862, + "time_per_iteration": 2.6241469383239746 + }, + { + "auxiliary_loss_clip": 0.01141394, + "auxiliary_loss_mlp": 0.01022422, + "balance_loss_clip": 1.01213956, + "balance_loss_mlp": 1.01564491, + "epoch": 0.9454698490951723, + "flos": 20010898477440.0, + "grad_norm": 1.8546774408880424, + "language_loss": 0.83642292, + "learning_rate": 3.104936965678084e-08, + "loss": 0.85806108, + "num_input_tokens_seen": 169946750, + "step": 7863, + "time_per_iteration": 3.5062122344970703 + }, + { + "auxiliary_loss_clip": 0.01139905, + "auxiliary_loss_mlp": 0.01022403, + "balance_loss_clip": 1.00831342, + "balance_loss_mlp": 1.01551533, + "epoch": 0.9455900919858113, + "flos": 21069652786560.0, + "grad_norm": 2.037397700177493, + "language_loss": 0.82127345, + "learning_rate": 3.091279149669956e-08, + "loss": 0.84289646, + "num_input_tokens_seen": 169965540, + "step": 7864, + "time_per_iteration": 3.5399296283721924 + }, + { + "auxiliary_loss_clip": 0.01140011, + "auxiliary_loss_mlp": 0.00508937, + "balance_loss_clip": 1.01007855, + "balance_loss_mlp": 0.0, + "epoch": 0.9457103348764504, + "flos": 20740854666240.0, + "grad_norm": 2.1509000568417296, + "language_loss": 0.73994255, + "learning_rate": 3.0776512045581624e-08, + "loss": 0.75643194, + "num_input_tokens_seen": 169984330, + "step": 7865, + "time_per_iteration": 3.605914831161499 + }, + { + "auxiliary_loss_clip": 0.01110101, + "auxiliary_loss_mlp": 0.01025511, + "balance_loss_clip": 0.97012293, + "balance_loss_mlp": 1.01836443, + "epoch": 0.9458305777670896, + "flos": 21428363957760.0, + "grad_norm": 1.9119025483857137, + "language_loss": 0.78135979, + "learning_rate": 3.0640531324101384e-08, + "loss": 0.80271596, + "num_input_tokens_seen": 170002095, + "step": 7866, + "time_per_iteration": 2.6824727058410645 + }, + { + "auxiliary_loss_clip": 0.01143405, + "auxiliary_loss_mlp": 0.01027483, + "balance_loss_clip": 1.01393175, + "balance_loss_mlp": 1.01997328, + "epoch": 0.9459508206577286, + "flos": 20011185786240.0, + "grad_norm": 2.4387599890374254, + "language_loss": 0.76003236, + "learning_rate": 3.0504849352886554e-08, + "loss": 0.78174126, + "num_input_tokens_seen": 170020240, + "step": 7867, + "time_per_iteration": 2.6002163887023926 + }, + { + "auxiliary_loss_clip": 0.01139648, + "auxiliary_loss_mlp": 0.01025201, + "balance_loss_clip": 1.01024508, + "balance_loss_mlp": 1.01828718, + "epoch": 0.9460710635483677, + "flos": 12166428291840.0, + "grad_norm": 3.8218944407670983, + "language_loss": 0.71380961, + "learning_rate": 3.036946615252023e-08, + "loss": 0.73545808, + "num_input_tokens_seen": 170035770, + "step": 7868, + "time_per_iteration": 2.5810093879699707 + }, + { + "auxiliary_loss_clip": 0.01125414, + "auxiliary_loss_mlp": 0.01026293, + "balance_loss_clip": 0.97275269, + "balance_loss_mlp": 1.01907527, + "epoch": 0.9461913064390068, + "flos": 34276196229120.0, + "grad_norm": 2.0169880134243297, + "language_loss": 0.66697264, + "learning_rate": 3.0234381743539984e-08, + "loss": 0.68848974, + "num_input_tokens_seen": 170053385, + "step": 7869, + "time_per_iteration": 2.9197728633880615 + }, + { + "auxiliary_loss_clip": 0.01119694, + "auxiliary_loss_mlp": 0.01025029, + "balance_loss_clip": 0.96932054, + "balance_loss_mlp": 1.01839232, + "epoch": 0.9463115493296459, + "flos": 19463763536640.0, + "grad_norm": 1.955322731208352, + "language_loss": 0.79881167, + "learning_rate": 3.0099596146437863e-08, + "loss": 0.82025886, + "num_input_tokens_seen": 170070490, + "step": 7870, + "time_per_iteration": 2.6573121547698975 + }, + { + "auxiliary_loss_clip": 0.01059066, + "auxiliary_loss_mlp": 0.0100164, + "balance_loss_clip": 1.00845003, + "balance_loss_mlp": 1.00074005, + "epoch": 0.946431792220285, + "flos": 70570824387840.0, + "grad_norm": 0.7702944440473017, + "language_loss": 0.60061359, + "learning_rate": 2.996510938166086e-08, + "loss": 0.62122065, + "num_input_tokens_seen": 170133465, + "step": 7871, + "time_per_iteration": 3.2941911220550537 + }, + { + "auxiliary_loss_clip": 0.01143162, + "auxiliary_loss_mlp": 0.01024583, + "balance_loss_clip": 1.01256728, + "balance_loss_mlp": 1.01815748, + "epoch": 0.9465520351109241, + "flos": 18947906363520.0, + "grad_norm": 1.9141694718283486, + "language_loss": 0.73135722, + "learning_rate": 2.983092146960997e-08, + "loss": 0.75303465, + "num_input_tokens_seen": 170150810, + "step": 7872, + "time_per_iteration": 2.7437384128570557 + }, + { + "auxiliary_loss_clip": 0.01118509, + "auxiliary_loss_mlp": 0.01024692, + "balance_loss_clip": 0.97031856, + "balance_loss_mlp": 1.01690173, + "epoch": 0.9466722780015632, + "flos": 19135647774720.0, + "grad_norm": 1.9709843269338114, + "language_loss": 0.80067289, + "learning_rate": 2.9697032430642256e-08, + "loss": 0.82210499, + "num_input_tokens_seen": 170169025, + "step": 7873, + "time_per_iteration": 2.677929639816284 + }, + { + "auxiliary_loss_clip": 0.01158808, + "auxiliary_loss_mlp": 0.01021427, + "balance_loss_clip": 1.04619598, + "balance_loss_mlp": 1.0150255, + "epoch": 0.9467925208922022, + "flos": 17237912520960.0, + "grad_norm": 2.308599653755495, + "language_loss": 0.73831975, + "learning_rate": 2.9563442285067906e-08, + "loss": 0.76012212, + "num_input_tokens_seen": 170186070, + "step": 7874, + "time_per_iteration": 3.516911029815674 + }, + { + "auxiliary_loss_clip": 0.01144549, + "auxiliary_loss_mlp": 0.0102822, + "balance_loss_clip": 1.01084709, + "balance_loss_mlp": 1.02053046, + "epoch": 0.9469127637828414, + "flos": 29169016859520.0, + "grad_norm": 1.866009778815738, + "language_loss": 0.79831314, + "learning_rate": 2.943015105315294e-08, + "loss": 0.82004082, + "num_input_tokens_seen": 170206265, + "step": 7875, + "time_per_iteration": 2.703395128250122 + }, + { + "auxiliary_loss_clip": 0.01073185, + "auxiliary_loss_mlp": 0.01029292, + "balance_loss_clip": 0.89444292, + "balance_loss_mlp": 1.02172208, + "epoch": 0.9470330066734804, + "flos": 26030460234240.0, + "grad_norm": 2.81185739885699, + "language_loss": 0.66451478, + "learning_rate": 2.929715875511718e-08, + "loss": 0.68553948, + "num_input_tokens_seen": 170225300, + "step": 7876, + "time_per_iteration": 2.8250973224639893 + }, + { + "auxiliary_loss_clip": 0.01139639, + "auxiliary_loss_mlp": 0.01022601, + "balance_loss_clip": 1.00609648, + "balance_loss_mlp": 1.01608276, + "epoch": 0.9471532495641195, + "flos": 23440906056960.0, + "grad_norm": 1.788919648868143, + "language_loss": 0.70093006, + "learning_rate": 2.9164465411135375e-08, + "loss": 0.72255242, + "num_input_tokens_seen": 170245070, + "step": 7877, + "time_per_iteration": 2.7135517597198486 + }, + { + "auxiliary_loss_clip": 0.0114165, + "auxiliary_loss_mlp": 0.01027001, + "balance_loss_clip": 1.01225114, + "balance_loss_mlp": 1.0199883, + "epoch": 0.9472734924547586, + "flos": 15815850099840.0, + "grad_norm": 2.1020373317935737, + "language_loss": 0.80935711, + "learning_rate": 2.9032071041337426e-08, + "loss": 0.8310436, + "num_input_tokens_seen": 170263305, + "step": 7878, + "time_per_iteration": 2.6943962574005127 + }, + { + "auxiliary_loss_clip": 0.01109105, + "auxiliary_loss_mlp": 0.01023997, + "balance_loss_clip": 0.97056448, + "balance_loss_mlp": 1.01724935, + "epoch": 0.9473937353453977, + "flos": 11181793697280.0, + "grad_norm": 1.5914194929790013, + "language_loss": 0.73137856, + "learning_rate": 2.889997566580704e-08, + "loss": 0.75270957, + "num_input_tokens_seen": 170281460, + "step": 7879, + "time_per_iteration": 2.6151185035705566 + }, + { + "auxiliary_loss_clip": 0.01168564, + "auxiliary_loss_mlp": 0.01025657, + "balance_loss_clip": 1.05080748, + "balance_loss_mlp": 1.01870465, + "epoch": 0.9475139782360368, + "flos": 25775530433280.0, + "grad_norm": 1.6708599019750356, + "language_loss": 0.70320404, + "learning_rate": 2.8768179304583086e-08, + "loss": 0.72514617, + "num_input_tokens_seen": 170303515, + "step": 7880, + "time_per_iteration": 2.752342700958252 + }, + { + "auxiliary_loss_clip": 0.01094341, + "auxiliary_loss_mlp": 0.01027, + "balance_loss_clip": 0.93599606, + "balance_loss_mlp": 1.020509, + "epoch": 0.9476342211266758, + "flos": 22820046451200.0, + "grad_norm": 1.64021909974403, + "language_loss": 0.73624122, + "learning_rate": 2.8636681977659117e-08, + "loss": 0.75745463, + "num_input_tokens_seen": 170323165, + "step": 7881, + "time_per_iteration": 2.828052282333374 + }, + { + "auxiliary_loss_clip": 0.01067525, + "auxiliary_loss_mlp": 0.01028327, + "balance_loss_clip": 0.8969245, + "balance_loss_mlp": 1.02146029, + "epoch": 0.947754464017315, + "flos": 20193611984640.0, + "grad_norm": 1.9954119791981881, + "language_loss": 0.77725703, + "learning_rate": 2.850548370498318e-08, + "loss": 0.79821551, + "num_input_tokens_seen": 170341005, + "step": 7882, + "time_per_iteration": 2.7798683643341064 + }, + { + "auxiliary_loss_clip": 0.0114086, + "auxiliary_loss_mlp": 0.01023721, + "balance_loss_clip": 1.00886416, + "balance_loss_mlp": 1.01689005, + "epoch": 0.9478747069079541, + "flos": 24717925359360.0, + "grad_norm": 1.6146037334990253, + "language_loss": 0.71611249, + "learning_rate": 2.8374584506457798e-08, + "loss": 0.73775828, + "num_input_tokens_seen": 170362280, + "step": 7883, + "time_per_iteration": 2.688753366470337 + }, + { + "auxiliary_loss_clip": 0.01113099, + "auxiliary_loss_mlp": 0.01026138, + "balance_loss_clip": 0.97092688, + "balance_loss_mlp": 1.01881838, + "epoch": 0.9479949497985931, + "flos": 21361355136000.0, + "grad_norm": 2.1196173769570468, + "language_loss": 0.67422974, + "learning_rate": 2.824398440193998e-08, + "loss": 0.69562209, + "num_input_tokens_seen": 170381080, + "step": 7884, + "time_per_iteration": 2.7088370323181152 + }, + { + "auxiliary_loss_clip": 0.01067953, + "auxiliary_loss_mlp": 0.01026997, + "balance_loss_clip": 0.8952986, + "balance_loss_mlp": 1.0199697, + "epoch": 0.9481151926892323, + "flos": 18148606968960.0, + "grad_norm": 3.461165072584028, + "language_loss": 0.71477449, + "learning_rate": 2.811368341124232e-08, + "loss": 0.73572403, + "num_input_tokens_seen": 170400150, + "step": 7885, + "time_per_iteration": 2.734394073486328 + }, + { + "auxiliary_loss_clip": 0.01119679, + "auxiliary_loss_mlp": 0.01029799, + "balance_loss_clip": 0.97351706, + "balance_loss_mlp": 1.02334082, + "epoch": 0.9482354355798713, + "flos": 22128012046080.0, + "grad_norm": 2.5468720425377267, + "language_loss": 0.68727916, + "learning_rate": 2.7983681554131222e-08, + "loss": 0.70877397, + "num_input_tokens_seen": 170420410, + "step": 7886, + "time_per_iteration": 2.647919178009033 + }, + { + "auxiliary_loss_clip": 0.01116509, + "auxiliary_loss_mlp": 0.01022429, + "balance_loss_clip": 0.97023571, + "balance_loss_mlp": 1.01503813, + "epoch": 0.9483556784705104, + "flos": 19063072344960.0, + "grad_norm": 1.9530405015582275, + "language_loss": 0.69689047, + "learning_rate": 2.7853978850327365e-08, + "loss": 0.71827984, + "num_input_tokens_seen": 170439580, + "step": 7887, + "time_per_iteration": 2.6711337566375732 + }, + { + "auxiliary_loss_clip": 0.01095516, + "auxiliary_loss_mlp": 0.01027134, + "balance_loss_clip": 0.93676364, + "balance_loss_mlp": 1.0199455, + "epoch": 0.9484759213611496, + "flos": 25777110631680.0, + "grad_norm": 1.7303130254719061, + "language_loss": 0.87136531, + "learning_rate": 2.7724575319507225e-08, + "loss": 0.89259183, + "num_input_tokens_seen": 170459290, + "step": 7888, + "time_per_iteration": 2.755695104598999 + }, + { + "auxiliary_loss_clip": 0.0113831, + "auxiliary_loss_mlp": 0.01027523, + "balance_loss_clip": 1.00704908, + "balance_loss_mlp": 1.02063894, + "epoch": 0.9485961642517886, + "flos": 20667740532480.0, + "grad_norm": 1.8107024063082846, + "language_loss": 0.77041519, + "learning_rate": 2.759547098130044e-08, + "loss": 0.79207355, + "num_input_tokens_seen": 170478020, + "step": 7889, + "time_per_iteration": 4.016118288040161 + }, + { + "auxiliary_loss_clip": 0.01163259, + "auxiliary_loss_mlp": 0.01023659, + "balance_loss_clip": 1.04794014, + "balance_loss_mlp": 1.01695108, + "epoch": 0.9487164071424277, + "flos": 22674069578880.0, + "grad_norm": 1.9754830571501816, + "language_loss": 0.77023596, + "learning_rate": 2.746666585529267e-08, + "loss": 0.79210514, + "num_input_tokens_seen": 170498295, + "step": 7890, + "time_per_iteration": 3.6134774684906006 + }, + { + "auxiliary_loss_clip": 0.01134594, + "auxiliary_loss_mlp": 0.01023956, + "balance_loss_clip": 1.00860488, + "balance_loss_mlp": 1.01711345, + "epoch": 0.9488366500330668, + "flos": 38726461716480.0, + "grad_norm": 2.1110402181588532, + "language_loss": 0.74007136, + "learning_rate": 2.73381599610234e-08, + "loss": 0.76165688, + "num_input_tokens_seen": 170518695, + "step": 7891, + "time_per_iteration": 2.938246488571167 + }, + { + "auxiliary_loss_clip": 0.01134325, + "auxiliary_loss_mlp": 0.01024408, + "balance_loss_clip": 1.00680625, + "balance_loss_mlp": 1.01770568, + "epoch": 0.9489568929237059, + "flos": 27890920149120.0, + "grad_norm": 2.596638430252404, + "language_loss": 0.71323127, + "learning_rate": 2.7209953317987033e-08, + "loss": 0.73481864, + "num_input_tokens_seen": 170539735, + "step": 7892, + "time_per_iteration": 3.658203363418579 + }, + { + "auxiliary_loss_clip": 0.01141391, + "auxiliary_loss_mlp": 0.010231, + "balance_loss_clip": 1.01089716, + "balance_loss_mlp": 1.01643646, + "epoch": 0.9490771358143449, + "flos": 33580642291200.0, + "grad_norm": 1.9197929615994382, + "language_loss": 0.78267807, + "learning_rate": 2.7082045945631793e-08, + "loss": 0.80432296, + "num_input_tokens_seen": 170561950, + "step": 7893, + "time_per_iteration": 2.7823991775512695 + }, + { + "auxiliary_loss_clip": 0.01085285, + "auxiliary_loss_mlp": 0.0102292, + "balance_loss_clip": 0.93198156, + "balance_loss_mlp": 1.01586246, + "epoch": 0.9491973787049841, + "flos": 14793796512000.0, + "grad_norm": 2.1052650195864078, + "language_loss": 0.69050258, + "learning_rate": 2.6954437863361712e-08, + "loss": 0.71158469, + "num_input_tokens_seen": 170579865, + "step": 7894, + "time_per_iteration": 2.6731109619140625 + }, + { + "auxiliary_loss_clip": 0.01049523, + "auxiliary_loss_mlp": 0.01023331, + "balance_loss_clip": 0.85622311, + "balance_loss_mlp": 1.0167594, + "epoch": 0.9493176215956232, + "flos": 25332535998720.0, + "grad_norm": 2.0597638872911994, + "language_loss": 0.71057421, + "learning_rate": 2.6827129090534862e-08, + "loss": 0.73130274, + "num_input_tokens_seen": 170600165, + "step": 7895, + "time_per_iteration": 2.816192150115967 + }, + { + "auxiliary_loss_clip": 0.01116743, + "auxiliary_loss_mlp": 0.01021562, + "balance_loss_clip": 0.97377706, + "balance_loss_mlp": 1.01425457, + "epoch": 0.9494378644862622, + "flos": 21029971236480.0, + "grad_norm": 1.8491696519055745, + "language_loss": 0.7790519, + "learning_rate": 2.670011964646335e-08, + "loss": 0.80043495, + "num_input_tokens_seen": 170618845, + "step": 7896, + "time_per_iteration": 2.6849608421325684 + }, + { + "auxiliary_loss_clip": 0.01028126, + "auxiliary_loss_mlp": 0.01027653, + "balance_loss_clip": 0.81349218, + "balance_loss_mlp": 1.02064943, + "epoch": 0.9495581073769014, + "flos": 15195134148480.0, + "grad_norm": 2.335434838416177, + "language_loss": 0.68120706, + "learning_rate": 2.657340955041487e-08, + "loss": 0.70176482, + "num_input_tokens_seen": 170637620, + "step": 7897, + "time_per_iteration": 2.8998446464538574 + }, + { + "auxiliary_loss_clip": 0.01118237, + "auxiliary_loss_mlp": 0.01024614, + "balance_loss_clip": 0.97559977, + "balance_loss_mlp": 1.01731825, + "epoch": 0.9496783502675404, + "flos": 28616566705920.0, + "grad_norm": 2.0936871988648815, + "language_loss": 0.71790755, + "learning_rate": 2.6446998821611167e-08, + "loss": 0.73933601, + "num_input_tokens_seen": 170657815, + "step": 7898, + "time_per_iteration": 3.0393264293670654 + }, + { + "auxiliary_loss_clip": 0.01074377, + "auxiliary_loss_mlp": 0.01024752, + "balance_loss_clip": 0.89631587, + "balance_loss_mlp": 1.01785541, + "epoch": 0.9497985931581795, + "flos": 14866874732160.0, + "grad_norm": 2.70157840068146, + "language_loss": 0.71943176, + "learning_rate": 2.6320887479228228e-08, + "loss": 0.74042308, + "num_input_tokens_seen": 170674415, + "step": 7899, + "time_per_iteration": 2.753525495529175 + }, + { + "auxiliary_loss_clip": 0.01120146, + "auxiliary_loss_mlp": 0.01025645, + "balance_loss_clip": 0.97150779, + "balance_loss_mlp": 1.01866221, + "epoch": 0.9499188360488187, + "flos": 27193319136000.0, + "grad_norm": 2.404284454639865, + "language_loss": 0.72417152, + "learning_rate": 2.619507554239786e-08, + "loss": 0.74562937, + "num_input_tokens_seen": 170692975, + "step": 7900, + "time_per_iteration": 3.601382255554199 + }, + { + "auxiliary_loss_clip": 0.01118558, + "auxiliary_loss_mlp": 0.01022563, + "balance_loss_clip": 0.97361946, + "balance_loss_mlp": 1.01547885, + "epoch": 0.9500390789394577, + "flos": 24316479982080.0, + "grad_norm": 1.739475182571045, + "language_loss": 0.69528437, + "learning_rate": 2.606956303020502e-08, + "loss": 0.71669561, + "num_input_tokens_seen": 170713780, + "step": 7901, + "time_per_iteration": 2.762693405151367 + }, + { + "auxiliary_loss_clip": 0.01140375, + "auxiliary_loss_mlp": 0.01024451, + "balance_loss_clip": 1.00999069, + "balance_loss_mlp": 1.01784122, + "epoch": 0.9501593218300968, + "flos": 14354752573440.0, + "grad_norm": 1.7651326420212232, + "language_loss": 0.84615523, + "learning_rate": 2.5944349961690036e-08, + "loss": 0.86780345, + "num_input_tokens_seen": 170730800, + "step": 7902, + "time_per_iteration": 2.6910195350646973 + }, + { + "auxiliary_loss_clip": 0.01094593, + "auxiliary_loss_mlp": 0.01025467, + "balance_loss_clip": 0.93373591, + "balance_loss_mlp": 1.01834106, + "epoch": 0.9502795647207359, + "flos": 38728113742080.0, + "grad_norm": 2.3995148046873616, + "language_loss": 0.73457277, + "learning_rate": 2.581943635584749e-08, + "loss": 0.75577343, + "num_input_tokens_seen": 170753630, + "step": 7903, + "time_per_iteration": 2.922074794769287 + }, + { + "auxiliary_loss_clip": 0.01109893, + "auxiliary_loss_mlp": 0.01024619, + "balance_loss_clip": 0.97093308, + "balance_loss_mlp": 1.01783347, + "epoch": 0.950399807611375, + "flos": 40808023799040.0, + "grad_norm": 1.5952964876921614, + "language_loss": 0.6522249, + "learning_rate": 2.569482223162689e-08, + "loss": 0.67357004, + "num_input_tokens_seen": 170777605, + "step": 7904, + "time_per_iteration": 3.060786247253418 + }, + { + "auxiliary_loss_clip": 0.0113935, + "auxiliary_loss_mlp": 0.01025968, + "balance_loss_clip": 1.00776052, + "balance_loss_mlp": 1.01877117, + "epoch": 0.950520050502014, + "flos": 23440403266560.0, + "grad_norm": 1.5788329211344625, + "language_loss": 0.72654432, + "learning_rate": 2.5570507607932e-08, + "loss": 0.7481975, + "num_input_tokens_seen": 170797520, + "step": 7905, + "time_per_iteration": 2.664447069168091 + }, + { + "auxiliary_loss_clip": 0.01144589, + "auxiliary_loss_mlp": 0.0102696, + "balance_loss_clip": 1.0103159, + "balance_loss_mlp": 1.0198853, + "epoch": 0.9506402933926532, + "flos": 17783718658560.0, + "grad_norm": 2.800178923110433, + "language_loss": 0.63812137, + "learning_rate": 2.54464925036213e-08, + "loss": 0.65983689, + "num_input_tokens_seen": 170814810, + "step": 7906, + "time_per_iteration": 2.7576863765716553 + }, + { + "auxiliary_loss_clip": 0.01140675, + "auxiliary_loss_mlp": 0.0102227, + "balance_loss_clip": 1.01036739, + "balance_loss_mlp": 1.01548672, + "epoch": 0.9507605362832923, + "flos": 32561928668160.0, + "grad_norm": 1.8929203396997225, + "language_loss": 0.61141932, + "learning_rate": 2.532277693750773e-08, + "loss": 0.63304877, + "num_input_tokens_seen": 170835735, + "step": 7907, + "time_per_iteration": 2.859792709350586 + }, + { + "auxiliary_loss_clip": 0.01069589, + "auxiliary_loss_mlp": 0.01025673, + "balance_loss_clip": 0.89938915, + "balance_loss_mlp": 1.0188427, + "epoch": 0.9508807791739313, + "flos": 19602054898560.0, + "grad_norm": 2.3549082445827865, + "language_loss": 0.75834662, + "learning_rate": 2.5199360928358948e-08, + "loss": 0.77929926, + "num_input_tokens_seen": 170852970, + "step": 7908, + "time_per_iteration": 2.7011101245880127 + }, + { + "auxiliary_loss_clip": 0.0113114, + "auxiliary_loss_mlp": 0.00508655, + "balance_loss_clip": 1.00776148, + "balance_loss_mlp": 0.0, + "epoch": 0.9510010220645704, + "flos": 21471852349440.0, + "grad_norm": 2.2392026332440396, + "language_loss": 0.87083, + "learning_rate": 2.507624449489665e-08, + "loss": 0.88722795, + "num_input_tokens_seen": 170871600, + "step": 7909, + "time_per_iteration": 2.6720077991485596 + }, + { + "auxiliary_loss_clip": 0.01120499, + "auxiliary_loss_mlp": 0.01029576, + "balance_loss_clip": 0.97344559, + "balance_loss_mlp": 1.02253985, + "epoch": 0.9511212649552095, + "flos": 18879999701760.0, + "grad_norm": 2.149833163711499, + "language_loss": 0.65059531, + "learning_rate": 2.495342765579811e-08, + "loss": 0.67209601, + "num_input_tokens_seen": 170890260, + "step": 7910, + "time_per_iteration": 2.6488959789276123 + }, + { + "auxiliary_loss_clip": 0.01071993, + "auxiliary_loss_mlp": 0.01023309, + "balance_loss_clip": 0.899041, + "balance_loss_mlp": 1.01601982, + "epoch": 0.9512415078458486, + "flos": 20810521094400.0, + "grad_norm": 1.6502588134657903, + "language_loss": 0.7100594, + "learning_rate": 2.4830910429693984e-08, + "loss": 0.73101246, + "num_input_tokens_seen": 170910220, + "step": 7911, + "time_per_iteration": 2.7540392875671387 + }, + { + "auxiliary_loss_clip": 0.01163391, + "auxiliary_loss_mlp": 0.01021951, + "balance_loss_clip": 1.04797077, + "balance_loss_mlp": 1.01475704, + "epoch": 0.9513617507364877, + "flos": 18369565482240.0, + "grad_norm": 2.2009811308555585, + "language_loss": 0.7966662, + "learning_rate": 2.470869283517052e-08, + "loss": 0.81851959, + "num_input_tokens_seen": 170928255, + "step": 7912, + "time_per_iteration": 2.597522497177124 + }, + { + "auxiliary_loss_clip": 0.01132056, + "auxiliary_loss_mlp": 0.01022502, + "balance_loss_clip": 1.00638795, + "balance_loss_mlp": 1.01577592, + "epoch": 0.9514819936271268, + "flos": 25010166412800.0, + "grad_norm": 1.8003327973781809, + "language_loss": 0.77121675, + "learning_rate": 2.458677489076777e-08, + "loss": 0.79276234, + "num_input_tokens_seen": 170949265, + "step": 7913, + "time_per_iteration": 2.6884708404541016 + }, + { + "auxiliary_loss_clip": 0.01131828, + "auxiliary_loss_mlp": 0.01021644, + "balance_loss_clip": 1.00647235, + "balance_loss_mlp": 1.01484656, + "epoch": 0.9516022365177659, + "flos": 18662129758080.0, + "grad_norm": 1.6328037775772777, + "language_loss": 0.82920194, + "learning_rate": 2.446515661498072e-08, + "loss": 0.85073668, + "num_input_tokens_seen": 170968595, + "step": 7914, + "time_per_iteration": 3.615971088409424 + }, + { + "auxiliary_loss_clip": 0.01045538, + "auxiliary_loss_mlp": 0.01023969, + "balance_loss_clip": 0.85663104, + "balance_loss_mlp": 1.01679873, + "epoch": 0.9517224794084049, + "flos": 25372109808000.0, + "grad_norm": 2.107308316258436, + "language_loss": 0.7412923, + "learning_rate": 2.434383802625861e-08, + "loss": 0.76198739, + "num_input_tokens_seen": 170987550, + "step": 7915, + "time_per_iteration": 2.8033227920532227 + }, + { + "auxiliary_loss_clip": 0.01093639, + "auxiliary_loss_mlp": 0.01024067, + "balance_loss_clip": 0.93089777, + "balance_loss_mlp": 1.01671183, + "epoch": 0.9518427222990441, + "flos": 21470918595840.0, + "grad_norm": 1.9425444662189282, + "language_loss": 0.73755896, + "learning_rate": 2.4222819143005168e-08, + "loss": 0.75873601, + "num_input_tokens_seen": 171007145, + "step": 7916, + "time_per_iteration": 3.589448928833008 + }, + { + "auxiliary_loss_clip": 0.01163276, + "auxiliary_loss_mlp": 0.0102443, + "balance_loss_clip": 1.05012822, + "balance_loss_mlp": 1.01774216, + "epoch": 0.9519629651896832, + "flos": 21033634423680.0, + "grad_norm": 1.8101739730095918, + "language_loss": 0.81011713, + "learning_rate": 2.4102099983579706e-08, + "loss": 0.83199418, + "num_input_tokens_seen": 171026295, + "step": 7917, + "time_per_iteration": 3.6286017894744873 + }, + { + "auxiliary_loss_clip": 0.01140144, + "auxiliary_loss_mlp": 0.010187, + "balance_loss_clip": 1.00918722, + "balance_loss_mlp": 1.01103473, + "epoch": 0.9520832080803222, + "flos": 21689219502720.0, + "grad_norm": 1.6703270989491406, + "language_loss": 0.76938182, + "learning_rate": 2.3981680566294236e-08, + "loss": 0.79097021, + "num_input_tokens_seen": 171045895, + "step": 7918, + "time_per_iteration": 2.624415397644043 + }, + { + "auxiliary_loss_clip": 0.01161615, + "auxiliary_loss_mlp": 0.01024739, + "balance_loss_clip": 1.04774857, + "balance_loss_mlp": 1.01780081, + "epoch": 0.9522034509709614, + "flos": 23145289125120.0, + "grad_norm": 1.9599443470463893, + "language_loss": 0.7358855, + "learning_rate": 2.3861560909416822e-08, + "loss": 0.75774902, + "num_input_tokens_seen": 171065445, + "step": 7919, + "time_per_iteration": 2.671950340270996 + }, + { + "auxiliary_loss_clip": 0.01069914, + "auxiliary_loss_mlp": 0.01027703, + "balance_loss_clip": 0.8948077, + "balance_loss_mlp": 1.02077401, + "epoch": 0.9523236938616004, + "flos": 24679428958080.0, + "grad_norm": 2.358525289859063, + "language_loss": 0.82598644, + "learning_rate": 2.3741741031169325e-08, + "loss": 0.84696257, + "num_input_tokens_seen": 171085015, + "step": 7920, + "time_per_iteration": 2.7690060138702393 + }, + { + "auxiliary_loss_clip": 0.01066522, + "auxiliary_loss_mlp": 0.01027121, + "balance_loss_clip": 0.89221168, + "balance_loss_mlp": 1.02021313, + "epoch": 0.9524439367522395, + "flos": 22672309812480.0, + "grad_norm": 2.052511041923169, + "language_loss": 0.7152108, + "learning_rate": 2.3622220949728544e-08, + "loss": 0.73614722, + "num_input_tokens_seen": 171103900, + "step": 7921, + "time_per_iteration": 2.7362635135650635 + }, + { + "auxiliary_loss_clip": 0.01134713, + "auxiliary_loss_mlp": 0.01023395, + "balance_loss_clip": 1.00851893, + "balance_loss_mlp": 1.01656175, + "epoch": 0.9525641796428787, + "flos": 34055525024640.0, + "grad_norm": 3.0228959938711477, + "language_loss": 0.61523712, + "learning_rate": 2.3503000683225526e-08, + "loss": 0.63681817, + "num_input_tokens_seen": 171121615, + "step": 7922, + "time_per_iteration": 2.7281382083892822 + }, + { + "auxiliary_loss_clip": 0.01162799, + "auxiliary_loss_mlp": 0.0102291, + "balance_loss_clip": 1.04668283, + "balance_loss_mlp": 1.01602006, + "epoch": 0.9526844225335177, + "flos": 16727083251840.0, + "grad_norm": 2.2137059015133165, + "language_loss": 0.84342694, + "learning_rate": 2.3384080249745585e-08, + "loss": 0.86528397, + "num_input_tokens_seen": 171139505, + "step": 7923, + "time_per_iteration": 2.6179513931274414 + }, + { + "auxiliary_loss_clip": 0.01071578, + "auxiliary_loss_mlp": 0.01025082, + "balance_loss_clip": 0.89329469, + "balance_loss_mlp": 1.01826644, + "epoch": 0.9528046654241568, + "flos": 36939367330560.0, + "grad_norm": 2.3231024841836634, + "language_loss": 0.8322382, + "learning_rate": 2.3265459667329178e-08, + "loss": 0.85320485, + "num_input_tokens_seen": 171158995, + "step": 7924, + "time_per_iteration": 2.778780460357666 + }, + { + "auxiliary_loss_clip": 0.01118374, + "auxiliary_loss_mlp": 0.01024184, + "balance_loss_clip": 0.97301364, + "balance_loss_mlp": 1.0176034, + "epoch": 0.9529249083147959, + "flos": 18255010032000.0, + "grad_norm": 2.1696775653993785, + "language_loss": 0.86488867, + "learning_rate": 2.31471389539708e-08, + "loss": 0.88631421, + "num_input_tokens_seen": 171176120, + "step": 7925, + "time_per_iteration": 3.6219675540924072 + }, + { + "auxiliary_loss_clip": 0.01140725, + "auxiliary_loss_mlp": 0.00508532, + "balance_loss_clip": 1.00991249, + "balance_loss_mlp": 0.0, + "epoch": 0.953045151205435, + "flos": 28658438985600.0, + "grad_norm": 1.8784685052732388, + "language_loss": 0.72746325, + "learning_rate": 2.3029118127619872e-08, + "loss": 0.74395579, + "num_input_tokens_seen": 171195835, + "step": 7926, + "time_per_iteration": 2.7678000926971436 + }, + { + "auxiliary_loss_clip": 0.01111047, + "auxiliary_loss_mlp": 0.01024413, + "balance_loss_clip": 0.97055018, + "balance_loss_mlp": 1.01706636, + "epoch": 0.953165394096074, + "flos": 21835232288640.0, + "grad_norm": 3.059121491938359, + "language_loss": 0.86895573, + "learning_rate": 2.2911397206179628e-08, + "loss": 0.89031029, + "num_input_tokens_seen": 171212585, + "step": 7927, + "time_per_iteration": 2.7065212726593018 + }, + { + "auxiliary_loss_clip": 0.01164334, + "auxiliary_loss_mlp": 0.01026228, + "balance_loss_clip": 1.04947042, + "balance_loss_mlp": 1.01940656, + "epoch": 0.9532856369867132, + "flos": 19975059682560.0, + "grad_norm": 3.059534917739304, + "language_loss": 0.63297522, + "learning_rate": 2.279397620750845e-08, + "loss": 0.65488088, + "num_input_tokens_seen": 171231630, + "step": 7928, + "time_per_iteration": 2.6164445877075195 + }, + { + "auxiliary_loss_clip": 0.01112787, + "auxiliary_loss_mlp": 0.01027704, + "balance_loss_clip": 0.97009075, + "balance_loss_mlp": 1.02086997, + "epoch": 0.9534058798773523, + "flos": 15049588239360.0, + "grad_norm": 1.8340893429184528, + "language_loss": 0.78706282, + "learning_rate": 2.2676855149419195e-08, + "loss": 0.80846775, + "num_input_tokens_seen": 171248800, + "step": 7929, + "time_per_iteration": 2.5806832313537598 + }, + { + "auxiliary_loss_clip": 0.01118466, + "auxiliary_loss_mlp": 0.01024327, + "balance_loss_clip": 0.9770146, + "balance_loss_mlp": 1.01726055, + "epoch": 0.9535261227679913, + "flos": 17602800831360.0, + "grad_norm": 2.202508363976754, + "language_loss": 0.7518208, + "learning_rate": 2.2560034049678988e-08, + "loss": 0.77324879, + "num_input_tokens_seen": 171263150, + "step": 7930, + "time_per_iteration": 2.6370291709899902 + }, + { + "auxiliary_loss_clip": 0.01166664, + "auxiliary_loss_mlp": 0.0103032, + "balance_loss_clip": 1.04957724, + "balance_loss_mlp": 1.02300692, + "epoch": 0.9536463656586305, + "flos": 23142954741120.0, + "grad_norm": 1.989887553152041, + "language_loss": 0.75590432, + "learning_rate": 2.2443512926008988e-08, + "loss": 0.77787411, + "num_input_tokens_seen": 171282480, + "step": 7931, + "time_per_iteration": 2.6369144916534424 + }, + { + "auxiliary_loss_clip": 0.01094908, + "auxiliary_loss_mlp": 0.01026909, + "balance_loss_clip": 0.9305743, + "balance_loss_mlp": 1.0200541, + "epoch": 0.9537666085492695, + "flos": 18625033987200.0, + "grad_norm": 2.6040827846756374, + "language_loss": 0.70132923, + "learning_rate": 2.2327291796085946e-08, + "loss": 0.72254735, + "num_input_tokens_seen": 171300840, + "step": 7932, + "time_per_iteration": 2.7295432090759277 + }, + { + "auxiliary_loss_clip": 0.01164206, + "auxiliary_loss_mlp": 0.0102483, + "balance_loss_clip": 1.0480845, + "balance_loss_mlp": 1.01695585, + "epoch": 0.9538868514399086, + "flos": 18989347680000.0, + "grad_norm": 2.7840810820046533, + "language_loss": 0.77162057, + "learning_rate": 2.2211370677540197e-08, + "loss": 0.79351091, + "num_input_tokens_seen": 171317365, + "step": 7933, + "time_per_iteration": 2.648146867752075 + }, + { + "auxiliary_loss_clip": 0.01165803, + "auxiliary_loss_mlp": 0.01024437, + "balance_loss_clip": 1.04860139, + "balance_loss_mlp": 1.01730204, + "epoch": 0.9540070943305478, + "flos": 16800556521600.0, + "grad_norm": 2.326458116898756, + "language_loss": 0.78061455, + "learning_rate": 2.2095749587957012e-08, + "loss": 0.802517, + "num_input_tokens_seen": 171335270, + "step": 7934, + "time_per_iteration": 2.6854405403137207 + }, + { + "auxiliary_loss_clip": 0.01111752, + "auxiliary_loss_mlp": 0.0102319, + "balance_loss_clip": 0.96705246, + "balance_loss_mlp": 1.01631176, + "epoch": 0.9541273372211868, + "flos": 20156911263360.0, + "grad_norm": 2.234485577855617, + "language_loss": 0.69662774, + "learning_rate": 2.1980428544876138e-08, + "loss": 0.71797717, + "num_input_tokens_seen": 171353910, + "step": 7935, + "time_per_iteration": 2.7050113677978516 + }, + { + "auxiliary_loss_clip": 0.01062936, + "auxiliary_loss_mlp": 0.01030758, + "balance_loss_clip": 0.8886041, + "balance_loss_mlp": 1.02366173, + "epoch": 0.9542475801118259, + "flos": 26725511381760.0, + "grad_norm": 2.7315596671417555, + "language_loss": 0.74134326, + "learning_rate": 2.1865407565791584e-08, + "loss": 0.76228023, + "num_input_tokens_seen": 171375480, + "step": 7936, + "time_per_iteration": 2.7869865894317627 + }, + { + "auxiliary_loss_clip": 0.01114457, + "auxiliary_loss_mlp": 0.01022668, + "balance_loss_clip": 0.96775985, + "balance_loss_mlp": 1.01550317, + "epoch": 0.954367823002465, + "flos": 23330911633920.0, + "grad_norm": 2.3450937458388244, + "language_loss": 0.77517784, + "learning_rate": 2.175068666815183e-08, + "loss": 0.79654908, + "num_input_tokens_seen": 171396320, + "step": 7937, + "time_per_iteration": 2.7921836376190186 + }, + { + "auxiliary_loss_clip": 0.01093846, + "auxiliary_loss_mlp": 0.01026611, + "balance_loss_clip": 0.93114406, + "balance_loss_mlp": 1.01915479, + "epoch": 0.9544880658931041, + "flos": 14902713527040.0, + "grad_norm": 2.0209679831934357, + "language_loss": 0.78988206, + "learning_rate": 2.163626586935985e-08, + "loss": 0.8110866, + "num_input_tokens_seen": 171412860, + "step": 7938, + "time_per_iteration": 2.6928646564483643 + }, + { + "auxiliary_loss_clip": 0.01136868, + "auxiliary_loss_mlp": 0.01032908, + "balance_loss_clip": 1.00810266, + "balance_loss_mlp": 1.02559495, + "epoch": 0.9546083087837431, + "flos": 29095902725760.0, + "grad_norm": 2.261946153289324, + "language_loss": 0.62884927, + "learning_rate": 2.1522145186773755e-08, + "loss": 0.65054703, + "num_input_tokens_seen": 171431780, + "step": 7939, + "time_per_iteration": 2.714561700820923 + }, + { + "auxiliary_loss_clip": 0.01116224, + "auxiliary_loss_mlp": 0.01023773, + "balance_loss_clip": 0.97118044, + "balance_loss_mlp": 1.01688552, + "epoch": 0.9547285516743822, + "flos": 21142335957120.0, + "grad_norm": 1.9658995084138615, + "language_loss": 0.85533035, + "learning_rate": 2.140832463770481e-08, + "loss": 0.87673038, + "num_input_tokens_seen": 171450975, + "step": 7940, + "time_per_iteration": 3.5346477031707764 + }, + { + "auxiliary_loss_clip": 0.01118168, + "auxiliary_loss_mlp": 0.01024984, + "balance_loss_clip": 0.96997553, + "balance_loss_mlp": 1.01814759, + "epoch": 0.9548487945650214, + "flos": 27490157130240.0, + "grad_norm": 2.0602751847304797, + "language_loss": 0.7602011, + "learning_rate": 2.129480423941987e-08, + "loss": 0.7816326, + "num_input_tokens_seen": 171467645, + "step": 7941, + "time_per_iteration": 2.7424838542938232 + }, + { + "auxiliary_loss_clip": 0.01120572, + "auxiliary_loss_mlp": 0.01024736, + "balance_loss_clip": 0.97171754, + "balance_loss_mlp": 1.01777124, + "epoch": 0.9549690374556604, + "flos": 22273198819200.0, + "grad_norm": 1.7828746206759003, + "language_loss": 0.79843652, + "learning_rate": 2.1181584009140052e-08, + "loss": 0.81988966, + "num_input_tokens_seen": 171487185, + "step": 7942, + "time_per_iteration": 3.5496532917022705 + }, + { + "auxiliary_loss_clip": 0.0110224, + "auxiliary_loss_mlp": 0.01024451, + "balance_loss_clip": 0.93414521, + "balance_loss_mlp": 1.01815939, + "epoch": 0.9550892803462995, + "flos": 17595294888960.0, + "grad_norm": 2.2867680748814743, + "language_loss": 0.84027076, + "learning_rate": 2.10686639640405e-08, + "loss": 0.86153769, + "num_input_tokens_seen": 171501275, + "step": 7943, + "time_per_iteration": 3.640230894088745 + }, + { + "auxiliary_loss_clip": 0.01142488, + "auxiliary_loss_mlp": 0.01027318, + "balance_loss_clip": 1.0083977, + "balance_loss_mlp": 1.02042747, + "epoch": 0.9552095232369386, + "flos": 24353144789760.0, + "grad_norm": 1.6701617008272793, + "language_loss": 0.8110593, + "learning_rate": 2.0956044121251294e-08, + "loss": 0.83275735, + "num_input_tokens_seen": 171520060, + "step": 7944, + "time_per_iteration": 2.661079168319702 + }, + { + "auxiliary_loss_clip": 0.01097374, + "auxiliary_loss_mlp": 0.01028108, + "balance_loss_clip": 0.9385249, + "balance_loss_mlp": 1.02031183, + "epoch": 0.9553297661275777, + "flos": 22746860490240.0, + "grad_norm": 1.9796334980338148, + "language_loss": 0.81102479, + "learning_rate": 2.084372449785654e-08, + "loss": 0.83227968, + "num_input_tokens_seen": 171539895, + "step": 7945, + "time_per_iteration": 2.715471029281616 + }, + { + "auxiliary_loss_clip": 0.01114165, + "auxiliary_loss_mlp": 0.01027625, + "balance_loss_clip": 0.97015119, + "balance_loss_mlp": 1.0203172, + "epoch": 0.9554500090182168, + "flos": 15413866018560.0, + "grad_norm": 2.1941126570070146, + "language_loss": 0.68738329, + "learning_rate": 2.0731705110895282e-08, + "loss": 0.70880127, + "num_input_tokens_seen": 171557385, + "step": 7946, + "time_per_iteration": 2.651082754135132 + }, + { + "auxiliary_loss_clip": 0.01145902, + "auxiliary_loss_mlp": 0.01028145, + "balance_loss_clip": 1.0134697, + "balance_loss_mlp": 1.02065885, + "epoch": 0.9555702519088559, + "flos": 23513517400320.0, + "grad_norm": 1.928563421407582, + "language_loss": 0.87148905, + "learning_rate": 2.0619985977360587e-08, + "loss": 0.89322948, + "num_input_tokens_seen": 171575705, + "step": 7947, + "time_per_iteration": 2.6707029342651367 + }, + { + "auxiliary_loss_clip": 0.01069678, + "auxiliary_loss_mlp": 0.01025377, + "balance_loss_clip": 0.890212, + "balance_loss_mlp": 1.0185194, + "epoch": 0.955690494799495, + "flos": 22962072827520.0, + "grad_norm": 1.6260083403985606, + "language_loss": 0.76822919, + "learning_rate": 2.0508567114200237e-08, + "loss": 0.78917974, + "num_input_tokens_seen": 171595620, + "step": 7948, + "time_per_iteration": 2.7427947521209717 + }, + { + "auxiliary_loss_clip": 0.01117181, + "auxiliary_loss_mlp": 0.01024816, + "balance_loss_clip": 0.96923971, + "balance_loss_mlp": 1.01825655, + "epoch": 0.955810737690134, + "flos": 26031250333440.0, + "grad_norm": 2.035031047083267, + "language_loss": 0.78774297, + "learning_rate": 2.0397448538316485e-08, + "loss": 0.80916297, + "num_input_tokens_seen": 171616660, + "step": 7949, + "time_per_iteration": 2.802124500274658 + }, + { + "auxiliary_loss_clip": 0.01089819, + "auxiliary_loss_mlp": 0.01028551, + "balance_loss_clip": 0.93223727, + "balance_loss_mlp": 1.02238822, + "epoch": 0.9559309805807732, + "flos": 20849951249280.0, + "grad_norm": 2.15688764862309, + "language_loss": 0.66799533, + "learning_rate": 2.028663026656563e-08, + "loss": 0.68917906, + "num_input_tokens_seen": 171635515, + "step": 7950, + "time_per_iteration": 2.682581901550293 + }, + { + "auxiliary_loss_clip": 0.01160715, + "auxiliary_loss_mlp": 0.0050897, + "balance_loss_clip": 1.04717922, + "balance_loss_mlp": 0.0, + "epoch": 0.9560512234714122, + "flos": 21578219498880.0, + "grad_norm": 1.9316801910033536, + "language_loss": 0.71602321, + "learning_rate": 2.0176112315758885e-08, + "loss": 0.73272008, + "num_input_tokens_seen": 171653305, + "step": 7951, + "time_per_iteration": 3.4917948246002197 + }, + { + "auxiliary_loss_clip": 0.01078581, + "auxiliary_loss_mlp": 0.01021946, + "balance_loss_clip": 0.89593589, + "balance_loss_mlp": 1.01502609, + "epoch": 0.9561714663620513, + "flos": 17450144029440.0, + "grad_norm": 2.4282681663070513, + "language_loss": 0.69230902, + "learning_rate": 2.0065894702661957e-08, + "loss": 0.71331429, + "num_input_tokens_seen": 171669980, + "step": 7952, + "time_per_iteration": 2.7211921215057373 + }, + { + "auxiliary_loss_clip": 0.01085242, + "auxiliary_loss_mlp": 0.00508966, + "balance_loss_clip": 0.928608, + "balance_loss_mlp": 0.0, + "epoch": 0.9562917092526905, + "flos": 26098510550400.0, + "grad_norm": 1.5786331162111107, + "language_loss": 0.77871048, + "learning_rate": 1.9955977443994577e-08, + "loss": 0.79465258, + "num_input_tokens_seen": 171689970, + "step": 7953, + "time_per_iteration": 2.7677297592163086 + }, + { + "auxiliary_loss_clip": 0.01117821, + "auxiliary_loss_mlp": 0.01025905, + "balance_loss_clip": 0.97252178, + "balance_loss_mlp": 1.01780462, + "epoch": 0.9564119521433295, + "flos": 24096742531200.0, + "grad_norm": 2.849845951781106, + "language_loss": 0.62530279, + "learning_rate": 1.9846360556430965e-08, + "loss": 0.64674002, + "num_input_tokens_seen": 171708270, + "step": 7954, + "time_per_iteration": 2.678147077560425 + }, + { + "auxiliary_loss_clip": 0.01161227, + "auxiliary_loss_mlp": 0.01021494, + "balance_loss_clip": 1.0464704, + "balance_loss_mlp": 1.01512218, + "epoch": 0.9565321950339686, + "flos": 32008903896960.0, + "grad_norm": 2.0827638215367625, + "language_loss": 0.61737907, + "learning_rate": 1.973704405660004e-08, + "loss": 0.63920629, + "num_input_tokens_seen": 171729385, + "step": 7955, + "time_per_iteration": 2.719264507293701 + }, + { + "auxiliary_loss_clip": 0.01046076, + "auxiliary_loss_mlp": 0.01023212, + "balance_loss_clip": 0.85602582, + "balance_loss_mlp": 1.01611924, + "epoch": 0.9566524379246077, + "flos": 23588642695680.0, + "grad_norm": 5.662162397844774, + "language_loss": 0.78189981, + "learning_rate": 1.9628027961085203e-08, + "loss": 0.80259264, + "num_input_tokens_seen": 171752615, + "step": 7956, + "time_per_iteration": 2.8373212814331055 + }, + { + "auxiliary_loss_clip": 0.01081067, + "auxiliary_loss_mlp": 0.01027105, + "balance_loss_clip": 0.92848831, + "balance_loss_mlp": 1.02007806, + "epoch": 0.9567726808152468, + "flos": 38067716240640.0, + "grad_norm": 1.781548375431061, + "language_loss": 0.8403275, + "learning_rate": 1.9519312286423894e-08, + "loss": 0.86140931, + "num_input_tokens_seen": 171775810, + "step": 7957, + "time_per_iteration": 2.9116547107696533 + }, + { + "auxiliary_loss_clip": 0.01138847, + "auxiliary_loss_mlp": 0.01026247, + "balance_loss_clip": 1.01053584, + "balance_loss_mlp": 1.01936507, + "epoch": 0.9568929237058859, + "flos": 22744059229440.0, + "grad_norm": 2.2720658222188415, + "language_loss": 0.77666253, + "learning_rate": 1.9410897049108255e-08, + "loss": 0.7983135, + "num_input_tokens_seen": 171795090, + "step": 7958, + "time_per_iteration": 2.6370067596435547 + }, + { + "auxiliary_loss_clip": 0.01169884, + "auxiliary_loss_mlp": 0.01026702, + "balance_loss_clip": 1.05038941, + "balance_loss_mlp": 1.01943922, + "epoch": 0.957013166596525, + "flos": 23841633162240.0, + "grad_norm": 3.033839864066647, + "language_loss": 0.91567278, + "learning_rate": 1.9302782265584905e-08, + "loss": 0.93763864, + "num_input_tokens_seen": 171815755, + "step": 7959, + "time_per_iteration": 2.625863552093506 + }, + { + "auxiliary_loss_clip": 0.01063534, + "auxiliary_loss_mlp": 0.0102758, + "balance_loss_clip": 0.89629841, + "balance_loss_mlp": 1.02100575, + "epoch": 0.9571334094871641, + "flos": 17639286071040.0, + "grad_norm": 2.035510580342296, + "language_loss": 0.8672204, + "learning_rate": 1.9194967952254282e-08, + "loss": 0.8881315, + "num_input_tokens_seen": 171834330, + "step": 7960, + "time_per_iteration": 2.7483584880828857 + }, + { + "auxiliary_loss_clip": 0.01139825, + "auxiliary_loss_mlp": 0.01024736, + "balance_loss_clip": 1.01096773, + "balance_loss_mlp": 1.01800978, + "epoch": 0.9572536523778031, + "flos": 15369623441280.0, + "grad_norm": 3.4370117814978745, + "language_loss": 0.80689585, + "learning_rate": 1.9087454125472635e-08, + "loss": 0.8285414, + "num_input_tokens_seen": 171848805, + "step": 7961, + "time_per_iteration": 2.5736536979675293 + }, + { + "auxiliary_loss_clip": 0.0116331, + "auxiliary_loss_mlp": 0.01025922, + "balance_loss_clip": 1.04736733, + "balance_loss_mlp": 1.0190016, + "epoch": 0.9573738952684423, + "flos": 24969838417920.0, + "grad_norm": 2.610595968078905, + "language_loss": 0.78540534, + "learning_rate": 1.8980240801548696e-08, + "loss": 0.80729771, + "num_input_tokens_seen": 171867995, + "step": 7962, + "time_per_iteration": 2.6748149394989014 + }, + { + "auxiliary_loss_clip": 0.01116011, + "auxiliary_loss_mlp": 0.01025067, + "balance_loss_clip": 0.97360873, + "balance_loss_mlp": 1.01814127, + "epoch": 0.9574941381590814, + "flos": 25769461034880.0, + "grad_norm": 2.0587229899614554, + "language_loss": 0.74079096, + "learning_rate": 1.8873327996747458e-08, + "loss": 0.76220179, + "num_input_tokens_seen": 171886495, + "step": 7963, + "time_per_iteration": 2.728787660598755 + }, + { + "auxiliary_loss_clip": 0.01142012, + "auxiliary_loss_mlp": 0.01025001, + "balance_loss_clip": 1.00810468, + "balance_loss_mlp": 1.01769614, + "epoch": 0.9576143810497204, + "flos": 32307178435200.0, + "grad_norm": 1.809396647487144, + "language_loss": 0.65977705, + "learning_rate": 1.8766715727287053e-08, + "loss": 0.68144715, + "num_input_tokens_seen": 171908200, + "step": 7964, + "time_per_iteration": 2.695496082305908 + }, + { + "auxiliary_loss_clip": 0.01144056, + "auxiliary_loss_mlp": 0.00509738, + "balance_loss_clip": 1.00926375, + "balance_loss_mlp": 0.0, + "epoch": 0.9577346239403596, + "flos": 27745733376000.0, + "grad_norm": 1.6410459995076758, + "language_loss": 0.79364926, + "learning_rate": 1.8660404009340546e-08, + "loss": 0.81018722, + "num_input_tokens_seen": 171928650, + "step": 7965, + "time_per_iteration": 2.681640148162842 + }, + { + "auxiliary_loss_clip": 0.0104105, + "auxiliary_loss_mlp": 0.00999597, + "balance_loss_clip": 0.97172546, + "balance_loss_mlp": 0.99867934, + "epoch": 0.9578548668309986, + "flos": 57468313710720.0, + "grad_norm": 0.8707466002713039, + "language_loss": 0.59473348, + "learning_rate": 1.8554392859035485e-08, + "loss": 0.61513996, + "num_input_tokens_seen": 171986400, + "step": 7966, + "time_per_iteration": 3.192824125289917 + }, + { + "auxiliary_loss_clip": 0.01029206, + "auxiliary_loss_mlp": 0.0102425, + "balance_loss_clip": 0.81909096, + "balance_loss_mlp": 1.01716876, + "epoch": 0.9579751097216377, + "flos": 19756040503680.0, + "grad_norm": 1.8151012494290215, + "language_loss": 0.78813583, + "learning_rate": 1.8448682292453444e-08, + "loss": 0.8086704, + "num_input_tokens_seen": 172005475, + "step": 7967, + "time_per_iteration": 4.590431451797485 + }, + { + "auxiliary_loss_clip": 0.01166978, + "auxiliary_loss_mlp": 0.01025399, + "balance_loss_clip": 1.0511806, + "balance_loss_mlp": 1.01821709, + "epoch": 0.9580953526122769, + "flos": 18041270152320.0, + "grad_norm": 14.224010498921645, + "language_loss": 0.66340774, + "learning_rate": 1.8343272325631154e-08, + "loss": 0.68533146, + "num_input_tokens_seen": 172024420, + "step": 7968, + "time_per_iteration": 2.6237411499023438 + }, + { + "auxiliary_loss_clip": 0.01028336, + "auxiliary_loss_mlp": 0.00509542, + "balance_loss_clip": 0.81947827, + "balance_loss_mlp": 0.0, + "epoch": 0.9582155955029159, + "flos": 24270154416000.0, + "grad_norm": 2.5753092480609956, + "language_loss": 0.78172517, + "learning_rate": 1.8238162974558492e-08, + "loss": 0.79710388, + "num_input_tokens_seen": 172038350, + "step": 7969, + "time_per_iteration": 3.7396795749664307 + }, + { + "auxiliary_loss_clip": 0.01116376, + "auxiliary_loss_mlp": 0.01024171, + "balance_loss_clip": 0.97361249, + "balance_loss_mlp": 1.01712596, + "epoch": 0.958335838393555, + "flos": 22783309816320.0, + "grad_norm": 2.364407643002166, + "language_loss": 0.7471379, + "learning_rate": 1.8133354255181144e-08, + "loss": 0.76854336, + "num_input_tokens_seen": 172058665, + "step": 7970, + "time_per_iteration": 2.8864986896514893 + }, + { + "auxiliary_loss_clip": 0.01134666, + "auxiliary_loss_mlp": 0.010268, + "balance_loss_clip": 1.0070107, + "balance_loss_mlp": 1.02001417, + "epoch": 0.958456081284194, + "flos": 16911484698240.0, + "grad_norm": 1.9023607083557783, + "language_loss": 0.75016832, + "learning_rate": 1.802884618339795e-08, + "loss": 0.77178293, + "num_input_tokens_seen": 172077470, + "step": 7971, + "time_per_iteration": 2.6250789165496826 + }, + { + "auxiliary_loss_clip": 0.01140654, + "auxiliary_loss_mlp": 0.01026896, + "balance_loss_clip": 1.00961971, + "balance_loss_mlp": 1.01975584, + "epoch": 0.9585763241748332, + "flos": 19974951941760.0, + "grad_norm": 2.1409787176239257, + "language_loss": 0.80989867, + "learning_rate": 1.7924638775062894e-08, + "loss": 0.8315742, + "num_input_tokens_seen": 172096590, + "step": 7972, + "time_per_iteration": 2.6504669189453125 + }, + { + "auxiliary_loss_clip": 0.01087603, + "auxiliary_loss_mlp": 0.01029592, + "balance_loss_clip": 0.93407881, + "balance_loss_mlp": 1.02286839, + "epoch": 0.9586965670654722, + "flos": 21395649646080.0, + "grad_norm": 2.5438179513084282, + "language_loss": 0.81379402, + "learning_rate": 1.7820732045984444e-08, + "loss": 0.83496594, + "num_input_tokens_seen": 172116735, + "step": 7973, + "time_per_iteration": 2.686983108520508 + }, + { + "auxiliary_loss_clip": 0.01135418, + "auxiliary_loss_mlp": 0.01026618, + "balance_loss_clip": 1.00718045, + "balance_loss_mlp": 1.01940858, + "epoch": 0.9588168099561113, + "flos": 21435115714560.0, + "grad_norm": 1.815291330724636, + "language_loss": 0.73849344, + "learning_rate": 1.7717126011924655e-08, + "loss": 0.76011384, + "num_input_tokens_seen": 172138320, + "step": 7974, + "time_per_iteration": 2.6657512187957764 + }, + { + "auxiliary_loss_clip": 0.01061554, + "auxiliary_loss_mlp": 0.0102531, + "balance_loss_clip": 0.88863522, + "balance_loss_mlp": 1.0175252, + "epoch": 0.9589370528467505, + "flos": 11763761852160.0, + "grad_norm": 3.1154360494514326, + "language_loss": 0.76468599, + "learning_rate": 1.7613820688600957e-08, + "loss": 0.78555465, + "num_input_tokens_seen": 172154225, + "step": 7975, + "time_per_iteration": 2.695336103439331 + }, + { + "auxiliary_loss_clip": 0.0112211, + "auxiliary_loss_mlp": 0.01025201, + "balance_loss_clip": 0.96986783, + "balance_loss_mlp": 1.01853108, + "epoch": 0.9590572957373895, + "flos": 23441516588160.0, + "grad_norm": 1.8501209234952916, + "language_loss": 0.78783906, + "learning_rate": 1.7510816091684588e-08, + "loss": 0.80931216, + "num_input_tokens_seen": 172174150, + "step": 7976, + "time_per_iteration": 2.7247767448425293 + }, + { + "auxiliary_loss_clip": 0.01119093, + "auxiliary_loss_mlp": 0.01025345, + "balance_loss_clip": 0.97334433, + "balance_loss_mlp": 1.01768923, + "epoch": 0.9591775386280286, + "flos": 22528272274560.0, + "grad_norm": 3.3590760665322907, + "language_loss": 0.78311187, + "learning_rate": 1.740811223680083e-08, + "loss": 0.80455625, + "num_input_tokens_seen": 172191005, + "step": 7977, + "time_per_iteration": 3.546218156814575 + }, + { + "auxiliary_loss_clip": 0.01162323, + "auxiliary_loss_mlp": 0.01027296, + "balance_loss_clip": 1.04729128, + "balance_loss_mlp": 1.0200361, + "epoch": 0.9592977815186677, + "flos": 18186959715840.0, + "grad_norm": 4.797772038678186, + "language_loss": 0.74495852, + "learning_rate": 1.7305709139530334e-08, + "loss": 0.7668547, + "num_input_tokens_seen": 172209785, + "step": 7978, + "time_per_iteration": 2.5885677337646484 + }, + { + "auxiliary_loss_clip": 0.01132592, + "auxiliary_loss_mlp": 0.01022374, + "balance_loss_clip": 1.00590396, + "balance_loss_mlp": 1.01546896, + "epoch": 0.9594180244093068, + "flos": 16537797555840.0, + "grad_norm": 2.4299377272285616, + "language_loss": 0.74574155, + "learning_rate": 1.7203606815407334e-08, + "loss": 0.76729119, + "num_input_tokens_seen": 172224380, + "step": 7979, + "time_per_iteration": 2.7000927925109863 + }, + { + "auxiliary_loss_clip": 0.01122435, + "auxiliary_loss_mlp": 0.01023985, + "balance_loss_clip": 0.97339976, + "balance_loss_mlp": 1.01681805, + "epoch": 0.9595382672999458, + "flos": 20554334317440.0, + "grad_norm": 1.7234195102606054, + "language_loss": 0.79383326, + "learning_rate": 1.7101805279920557e-08, + "loss": 0.81529748, + "num_input_tokens_seen": 172242540, + "step": 7980, + "time_per_iteration": 2.7317757606506348 + }, + { + "auxiliary_loss_clip": 0.01167902, + "auxiliary_loss_mlp": 0.01024641, + "balance_loss_clip": 1.0510242, + "balance_loss_mlp": 1.01727664, + "epoch": 0.959658510190585, + "flos": 22638266697600.0, + "grad_norm": 2.1696765518985623, + "language_loss": 0.81240463, + "learning_rate": 1.7000304548513643e-08, + "loss": 0.83433008, + "num_input_tokens_seen": 172262645, + "step": 7981, + "time_per_iteration": 2.6453921794891357 + }, + { + "auxiliary_loss_clip": 0.01091619, + "auxiliary_loss_mlp": 0.01023386, + "balance_loss_clip": 0.9319346, + "balance_loss_mlp": 1.01650763, + "epoch": 0.9597787530812241, + "flos": 19135252725120.0, + "grad_norm": 2.1339303844957724, + "language_loss": 0.82943928, + "learning_rate": 1.6899104636583394e-08, + "loss": 0.85058928, + "num_input_tokens_seen": 172280695, + "step": 7982, + "time_per_iteration": 2.649808883666992 + }, + { + "auxiliary_loss_clip": 0.01041644, + "auxiliary_loss_mlp": 0.0100183, + "balance_loss_clip": 0.97175181, + "balance_loss_mlp": 1.00090015, + "epoch": 0.9598989959718631, + "flos": 60098124055680.0, + "grad_norm": 0.7228889557035295, + "language_loss": 0.61875427, + "learning_rate": 1.6798205559482638e-08, + "loss": 0.639189, + "num_input_tokens_seen": 172343075, + "step": 7983, + "time_per_iteration": 3.3073840141296387 + }, + { + "auxiliary_loss_clip": 0.0109939, + "auxiliary_loss_mlp": 0.01024984, + "balance_loss_clip": 0.93704146, + "balance_loss_mlp": 1.01817703, + "epoch": 0.9600192388625023, + "flos": 20886795624960.0, + "grad_norm": 1.7077177863485753, + "language_loss": 0.76652235, + "learning_rate": 1.669760733251713e-08, + "loss": 0.7877661, + "num_input_tokens_seen": 172361950, + "step": 7984, + "time_per_iteration": 2.741274118423462 + }, + { + "auxiliary_loss_clip": 0.01055133, + "auxiliary_loss_mlp": 0.01022983, + "balance_loss_clip": 0.85747457, + "balance_loss_mlp": 1.01675677, + "epoch": 0.9601394817531413, + "flos": 20445740524800.0, + "grad_norm": 1.7785959240084732, + "language_loss": 0.82754499, + "learning_rate": 1.659730997094755e-08, + "loss": 0.84832615, + "num_input_tokens_seen": 172380440, + "step": 7985, + "time_per_iteration": 2.8269755840301514 + }, + { + "auxiliary_loss_clip": 0.01132169, + "auxiliary_loss_mlp": 0.01025755, + "balance_loss_clip": 1.00722849, + "balance_loss_mlp": 1.01881123, + "epoch": 0.9602597246437804, + "flos": 21507152440320.0, + "grad_norm": 1.7576515929726324, + "language_loss": 0.62298608, + "learning_rate": 1.6497313489989283e-08, + "loss": 0.64456534, + "num_input_tokens_seen": 172400265, + "step": 7986, + "time_per_iteration": 2.6900134086608887 + }, + { + "auxiliary_loss_clip": 0.01067118, + "auxiliary_loss_mlp": 0.0102383, + "balance_loss_clip": 0.88823724, + "balance_loss_mlp": 1.01645136, + "epoch": 0.9603799675344196, + "flos": 29935099152000.0, + "grad_norm": 2.446314733182254, + "language_loss": 0.69936055, + "learning_rate": 1.639761790481131e-08, + "loss": 0.72027004, + "num_input_tokens_seen": 172421145, + "step": 7987, + "time_per_iteration": 2.87526273727417 + }, + { + "auxiliary_loss_clip": 0.01144022, + "auxiliary_loss_mlp": 0.01030498, + "balance_loss_clip": 1.01122296, + "balance_loss_mlp": 1.02400362, + "epoch": 0.9605002104250586, + "flos": 28001525103360.0, + "grad_norm": 1.921963344112247, + "language_loss": 0.79117084, + "learning_rate": 1.6298223230537754e-08, + "loss": 0.81291604, + "num_input_tokens_seen": 172438945, + "step": 7988, + "time_per_iteration": 2.726189374923706 + }, + { + "auxiliary_loss_clip": 0.01117307, + "auxiliary_loss_mlp": 0.00509469, + "balance_loss_clip": 0.97058153, + "balance_loss_mlp": 0.0, + "epoch": 0.9606204533156977, + "flos": 35590490870400.0, + "grad_norm": 2.332115249027124, + "language_loss": 0.69517875, + "learning_rate": 1.619912948224611e-08, + "loss": 0.71144652, + "num_input_tokens_seen": 172460150, + "step": 7989, + "time_per_iteration": 2.92622709274292 + }, + { + "auxiliary_loss_clip": 0.01086817, + "auxiliary_loss_mlp": 0.01024041, + "balance_loss_clip": 0.93072736, + "balance_loss_mlp": 1.01677561, + "epoch": 0.9607406962063368, + "flos": 26574614346240.0, + "grad_norm": 11.159570916703384, + "language_loss": 0.61057359, + "learning_rate": 1.6100336674969682e-08, + "loss": 0.63168222, + "num_input_tokens_seen": 172478990, + "step": 7990, + "time_per_iteration": 2.984185218811035 + }, + { + "auxiliary_loss_clip": 0.01071991, + "auxiliary_loss_mlp": 0.01032208, + "balance_loss_clip": 0.89110458, + "balance_loss_mlp": 1.02427471, + "epoch": 0.9608609390969759, + "flos": 25331781813120.0, + "grad_norm": 2.2628154989182887, + "language_loss": 0.76984107, + "learning_rate": 1.600184482369449e-08, + "loss": 0.79088306, + "num_input_tokens_seen": 172498905, + "step": 7991, + "time_per_iteration": 2.893831491470337 + }, + { + "auxiliary_loss_clip": 0.01094947, + "auxiliary_loss_mlp": 0.01023144, + "balance_loss_clip": 0.93175113, + "balance_loss_mlp": 1.01566982, + "epoch": 0.960981181987615, + "flos": 21069114082560.0, + "grad_norm": 2.380655010432598, + "language_loss": 0.89060676, + "learning_rate": 1.5903653943362126e-08, + "loss": 0.91178769, + "num_input_tokens_seen": 172517900, + "step": 7992, + "time_per_iteration": 3.6529510021209717 + }, + { + "auxiliary_loss_clip": 0.01118575, + "auxiliary_loss_mlp": 0.01023736, + "balance_loss_clip": 0.97273028, + "balance_loss_mlp": 1.01705694, + "epoch": 0.9611014248782541, + "flos": 17823256554240.0, + "grad_norm": 1.9644380190848563, + "language_loss": 0.77238375, + "learning_rate": 1.580576404886802e-08, + "loss": 0.79380691, + "num_input_tokens_seen": 172536430, + "step": 7993, + "time_per_iteration": 3.656841516494751 + }, + { + "auxiliary_loss_clip": 0.01138407, + "auxiliary_loss_mlp": 0.01022037, + "balance_loss_clip": 1.00788307, + "balance_loss_mlp": 1.0154748, + "epoch": 0.9612216677688932, + "flos": 19354631040000.0, + "grad_norm": 2.070275854328843, + "language_loss": 0.80127728, + "learning_rate": 1.570817515506162e-08, + "loss": 0.82288176, + "num_input_tokens_seen": 172555120, + "step": 7994, + "time_per_iteration": 2.724475383758545 + }, + { + "auxiliary_loss_clip": 0.01163457, + "auxiliary_loss_mlp": 0.01024964, + "balance_loss_clip": 1.04947186, + "balance_loss_mlp": 1.01790059, + "epoch": 0.9613419106595322, + "flos": 15808739207040.0, + "grad_norm": 2.384039830624153, + "language_loss": 0.81413293, + "learning_rate": 1.561088727674753e-08, + "loss": 0.83601713, + "num_input_tokens_seen": 172569330, + "step": 7995, + "time_per_iteration": 3.6324331760406494 + }, + { + "auxiliary_loss_clip": 0.01082753, + "auxiliary_loss_mlp": 0.01027039, + "balance_loss_clip": 0.89402968, + "balance_loss_mlp": 1.01951933, + "epoch": 0.9614621535501714, + "flos": 25702488126720.0, + "grad_norm": 3.1332299844855043, + "language_loss": 0.71449614, + "learning_rate": 1.551390042868417e-08, + "loss": 0.73559403, + "num_input_tokens_seen": 172591100, + "step": 7996, + "time_per_iteration": 2.8159286975860596 + }, + { + "auxiliary_loss_clip": 0.01141966, + "auxiliary_loss_mlp": 0.0102472, + "balance_loss_clip": 1.01106703, + "balance_loss_mlp": 1.01801157, + "epoch": 0.9615823964408104, + "flos": 17819054663040.0, + "grad_norm": 2.038499557400336, + "language_loss": 0.70998418, + "learning_rate": 1.5417214625584207e-08, + "loss": 0.73165101, + "num_input_tokens_seen": 172608755, + "step": 7997, + "time_per_iteration": 2.68320894241333 + }, + { + "auxiliary_loss_clip": 0.01134674, + "auxiliary_loss_mlp": 0.0102412, + "balance_loss_clip": 1.00783324, + "balance_loss_mlp": 1.01695228, + "epoch": 0.9617026393314495, + "flos": 20190020624640.0, + "grad_norm": 1.6597190303888514, + "language_loss": 0.85087562, + "learning_rate": 1.5320829882114806e-08, + "loss": 0.87246358, + "num_input_tokens_seen": 172626830, + "step": 7998, + "time_per_iteration": 2.6652166843414307 + }, + { + "auxiliary_loss_clip": 0.01163897, + "auxiliary_loss_mlp": 0.01027025, + "balance_loss_clip": 1.045946, + "balance_loss_mlp": 1.0199796, + "epoch": 0.9618228822220887, + "flos": 20267013427200.0, + "grad_norm": 2.552877921112628, + "language_loss": 0.7851513, + "learning_rate": 1.5224746212897378e-08, + "loss": 0.80706048, + "num_input_tokens_seen": 172646125, + "step": 7999, + "time_per_iteration": 2.6729745864868164 + }, + { + "auxiliary_loss_clip": 0.01160618, + "auxiliary_loss_mlp": 0.01024234, + "balance_loss_clip": 1.04673409, + "balance_loss_mlp": 1.01745367, + "epoch": 0.9619431251127277, + "flos": 21031300039680.0, + "grad_norm": 1.712395997833118, + "language_loss": 0.77500433, + "learning_rate": 1.512896363250804e-08, + "loss": 0.79685283, + "num_input_tokens_seen": 172666235, + "step": 8000, + "time_per_iteration": 2.6623141765594482 + }, + { + "auxiliary_loss_clip": 0.01144688, + "auxiliary_loss_mlp": 0.01026095, + "balance_loss_clip": 1.01039171, + "balance_loss_mlp": 1.01921105, + "epoch": 0.9620633680033668, + "flos": 22382654538240.0, + "grad_norm": 1.7894550454331397, + "language_loss": 0.75577819, + "learning_rate": 1.503348215547673e-08, + "loss": 0.77748603, + "num_input_tokens_seen": 172687325, + "step": 8001, + "time_per_iteration": 2.7659404277801514 + }, + { + "auxiliary_loss_clip": 0.01113283, + "auxiliary_loss_mlp": 0.01022758, + "balance_loss_clip": 0.96879721, + "balance_loss_mlp": 1.01577234, + "epoch": 0.962183610894006, + "flos": 18471730740480.0, + "grad_norm": 2.2697660016458823, + "language_loss": 0.80901563, + "learning_rate": 1.4938301796288078e-08, + "loss": 0.83037603, + "num_input_tokens_seen": 172703895, + "step": 8002, + "time_per_iteration": 2.6990773677825928 + }, + { + "auxiliary_loss_clip": 0.0116532, + "auxiliary_loss_mlp": 0.01023597, + "balance_loss_clip": 1.04918194, + "balance_loss_mlp": 1.01603937, + "epoch": 0.962303853784645, + "flos": 18435245500800.0, + "grad_norm": 3.273438367589917, + "language_loss": 0.82461375, + "learning_rate": 1.4843422569380537e-08, + "loss": 0.8465029, + "num_input_tokens_seen": 172720650, + "step": 8003, + "time_per_iteration": 3.5205001831054688 + }, + { + "auxiliary_loss_clip": 0.0106904, + "auxiliary_loss_mlp": 0.01024162, + "balance_loss_clip": 0.89232999, + "balance_loss_mlp": 1.0173285, + "epoch": 0.9624240966752841, + "flos": 26391074826240.0, + "grad_norm": 1.7069782140765513, + "language_loss": 0.83023053, + "learning_rate": 1.4748844489147483e-08, + "loss": 0.85116255, + "num_input_tokens_seen": 172737640, + "step": 8004, + "time_per_iteration": 2.788465976715088 + }, + { + "auxiliary_loss_clip": 0.01115759, + "auxiliary_loss_mlp": 0.01023194, + "balance_loss_clip": 0.96985263, + "balance_loss_mlp": 1.01637518, + "epoch": 0.9625443395659231, + "flos": 14647675985280.0, + "grad_norm": 1.8223282002165762, + "language_loss": 0.70883453, + "learning_rate": 1.4654567569936326e-08, + "loss": 0.73022407, + "num_input_tokens_seen": 172755215, + "step": 8005, + "time_per_iteration": 2.6967525482177734 + }, + { + "auxiliary_loss_clip": 0.01066402, + "auxiliary_loss_mlp": 0.01025496, + "balance_loss_clip": 0.89344716, + "balance_loss_mlp": 1.01812339, + "epoch": 0.9626645824565623, + "flos": 18367626147840.0, + "grad_norm": 2.019587487287537, + "language_loss": 0.83126926, + "learning_rate": 1.456059182604874e-08, + "loss": 0.85218823, + "num_input_tokens_seen": 172774020, + "step": 8006, + "time_per_iteration": 2.7586207389831543 + }, + { + "auxiliary_loss_clip": 0.01165054, + "auxiliary_loss_mlp": 0.01028907, + "balance_loss_clip": 1.04928637, + "balance_loss_mlp": 1.02138448, + "epoch": 0.9627848253472013, + "flos": 16580424021120.0, + "grad_norm": 2.966080517442119, + "language_loss": 0.76804131, + "learning_rate": 1.4466917271740653e-08, + "loss": 0.78998101, + "num_input_tokens_seen": 172792220, + "step": 8007, + "time_per_iteration": 2.631436824798584 + }, + { + "auxiliary_loss_clip": 0.01114645, + "auxiliary_loss_mlp": 0.01025252, + "balance_loss_clip": 0.97064519, + "balance_loss_mlp": 1.01792943, + "epoch": 0.9629050682378404, + "flos": 20886867452160.0, + "grad_norm": 1.8596614602817598, + "language_loss": 0.67630571, + "learning_rate": 1.4373543921222697e-08, + "loss": 0.69770467, + "num_input_tokens_seen": 172811805, + "step": 8008, + "time_per_iteration": 2.7046215534210205 + }, + { + "auxiliary_loss_clip": 0.01117075, + "auxiliary_loss_mlp": 0.01025479, + "balance_loss_clip": 0.97170383, + "balance_loss_mlp": 1.01834488, + "epoch": 0.9630253111284796, + "flos": 17019252478080.0, + "grad_norm": 2.323511214020601, + "language_loss": 0.77667725, + "learning_rate": 1.428047178865932e-08, + "loss": 0.7981028, + "num_input_tokens_seen": 172828595, + "step": 8009, + "time_per_iteration": 2.622396945953369 + }, + { + "auxiliary_loss_clip": 0.01114682, + "auxiliary_loss_mlp": 0.01023322, + "balance_loss_clip": 0.96836293, + "balance_loss_mlp": 1.0162673, + "epoch": 0.9631455540191186, + "flos": 20338942412160.0, + "grad_norm": 1.6377140851135223, + "language_loss": 0.74437749, + "learning_rate": 1.4187700888169451e-08, + "loss": 0.76575756, + "num_input_tokens_seen": 172847770, + "step": 8010, + "time_per_iteration": 2.727647542953491 + }, + { + "auxiliary_loss_clip": 0.01039762, + "auxiliary_loss_mlp": 0.00999031, + "balance_loss_clip": 0.97184813, + "balance_loss_mlp": 0.99809533, + "epoch": 0.9632657969097577, + "flos": 65956700033280.0, + "grad_norm": 0.7525182882346149, + "language_loss": 0.56974351, + "learning_rate": 1.40952312338265e-08, + "loss": 0.59013152, + "num_input_tokens_seen": 172912415, + "step": 8011, + "time_per_iteration": 3.2572076320648193 + }, + { + "auxiliary_loss_clip": 0.01094807, + "auxiliary_loss_mlp": 0.01020474, + "balance_loss_clip": 0.93103373, + "balance_loss_mlp": 1.01391411, + "epoch": 0.9633860398003968, + "flos": 44419523823360.0, + "grad_norm": 1.9292723416302386, + "language_loss": 0.6843949, + "learning_rate": 1.4003062839657909e-08, + "loss": 0.70554775, + "num_input_tokens_seen": 172934895, + "step": 8012, + "time_per_iteration": 2.9562227725982666 + }, + { + "auxiliary_loss_clip": 0.01096642, + "auxiliary_loss_mlp": 0.01022335, + "balance_loss_clip": 0.93233442, + "balance_loss_mlp": 1.01574612, + "epoch": 0.9635062826910359, + "flos": 24827704300800.0, + "grad_norm": 1.6278023924409966, + "language_loss": 0.80532217, + "learning_rate": 1.391119571964583e-08, + "loss": 0.82651198, + "num_input_tokens_seen": 172955835, + "step": 8013, + "time_per_iteration": 2.706021785736084 + }, + { + "auxiliary_loss_clip": 0.01140698, + "auxiliary_loss_mlp": 0.01022563, + "balance_loss_clip": 1.00977027, + "balance_loss_mlp": 1.01591134, + "epoch": 0.9636265255816749, + "flos": 15961360095360.0, + "grad_norm": 1.9262783386705127, + "language_loss": 0.72979093, + "learning_rate": 1.3819629887726225e-08, + "loss": 0.75142354, + "num_input_tokens_seen": 172973925, + "step": 8014, + "time_per_iteration": 2.6734230518341064 + }, + { + "auxiliary_loss_clip": 0.01121204, + "auxiliary_loss_mlp": 0.01023771, + "balance_loss_clip": 0.97166538, + "balance_loss_mlp": 1.01716697, + "epoch": 0.9637467684723141, + "flos": 22601781457920.0, + "grad_norm": 1.7760884047430443, + "language_loss": 0.76341927, + "learning_rate": 1.3728365357789317e-08, + "loss": 0.78486902, + "num_input_tokens_seen": 172993290, + "step": 8015, + "time_per_iteration": 2.685145854949951 + }, + { + "auxiliary_loss_clip": 0.01038497, + "auxiliary_loss_mlp": 0.01022691, + "balance_loss_clip": 0.85541087, + "balance_loss_mlp": 1.01515722, + "epoch": 0.9638670113629532, + "flos": 17565812801280.0, + "grad_norm": 3.804085431830418, + "language_loss": 0.76713371, + "learning_rate": 1.3637402143680254e-08, + "loss": 0.78774554, + "num_input_tokens_seen": 173008190, + "step": 8016, + "time_per_iteration": 2.7118382453918457 + }, + { + "auxiliary_loss_clip": 0.0100245, + "auxiliary_loss_mlp": 0.01009461, + "balance_loss_clip": 0.90315998, + "balance_loss_mlp": 1.00837064, + "epoch": 0.9639872542535922, + "flos": 55072139379840.0, + "grad_norm": 0.7219571576089558, + "language_loss": 0.55069196, + "learning_rate": 1.3546740259197998e-08, + "loss": 0.57081109, + "num_input_tokens_seen": 173061000, + "step": 8017, + "time_per_iteration": 3.2288753986358643 + }, + { + "auxiliary_loss_clip": 0.01118131, + "auxiliary_loss_mlp": 0.01021772, + "balance_loss_clip": 0.97201478, + "balance_loss_mlp": 1.01467276, + "epoch": 0.9641074971442314, + "flos": 24134484746880.0, + "grad_norm": 2.1983328492193817, + "language_loss": 0.69953549, + "learning_rate": 1.3456379718095989e-08, + "loss": 0.72093463, + "num_input_tokens_seen": 173081415, + "step": 8018, + "time_per_iteration": 2.71683669090271 + }, + { + "auxiliary_loss_clip": 0.01017556, + "auxiliary_loss_mlp": 0.00999904, + "balance_loss_clip": 0.93440169, + "balance_loss_mlp": 0.99882507, + "epoch": 0.9642277400348704, + "flos": 66747416077440.0, + "grad_norm": 0.8424279047006704, + "language_loss": 0.61973405, + "learning_rate": 1.3366320534081487e-08, + "loss": 0.63990867, + "num_input_tokens_seen": 173144095, + "step": 8019, + "time_per_iteration": 5.046861171722412 + }, + { + "auxiliary_loss_clip": 0.01138561, + "auxiliary_loss_mlp": 0.01028139, + "balance_loss_clip": 1.00874376, + "balance_loss_mlp": 1.02092409, + "epoch": 0.9643479829255095, + "flos": 30920272450560.0, + "grad_norm": 1.8851216905128934, + "language_loss": 0.75635689, + "learning_rate": 1.3276562720816675e-08, + "loss": 0.7780239, + "num_input_tokens_seen": 173165605, + "step": 8020, + "time_per_iteration": 2.775564432144165 + }, + { + "auxiliary_loss_clip": 0.01163536, + "auxiliary_loss_mlp": 0.01027217, + "balance_loss_clip": 1.04731822, + "balance_loss_mlp": 1.01980829, + "epoch": 0.9644682258161487, + "flos": 20048245643520.0, + "grad_norm": 2.1911137192289964, + "language_loss": 0.82751435, + "learning_rate": 1.3187106291917549e-08, + "loss": 0.84942186, + "num_input_tokens_seen": 173182595, + "step": 8021, + "time_per_iteration": 3.5556960105895996 + }, + { + "auxiliary_loss_clip": 0.01137187, + "auxiliary_loss_mlp": 0.01030815, + "balance_loss_clip": 1.00995755, + "balance_loss_mlp": 1.02316761, + "epoch": 0.9645884687067877, + "flos": 21178713456000.0, + "grad_norm": 1.7350407196378343, + "language_loss": 0.70968843, + "learning_rate": 1.309795126095503e-08, + "loss": 0.73136842, + "num_input_tokens_seen": 173200895, + "step": 8022, + "time_per_iteration": 2.679591655731201 + }, + { + "auxiliary_loss_clip": 0.01020551, + "auxiliary_loss_mlp": 0.01022088, + "balance_loss_clip": 0.81673986, + "balance_loss_mlp": 1.01531112, + "epoch": 0.9647087115974268, + "flos": 18945967029120.0, + "grad_norm": 2.239066071631626, + "language_loss": 0.80843723, + "learning_rate": 1.3009097641453192e-08, + "loss": 0.82886362, + "num_input_tokens_seen": 173218745, + "step": 8023, + "time_per_iteration": 2.757826328277588 + }, + { + "auxiliary_loss_clip": 0.01118609, + "auxiliary_loss_mlp": 0.01022147, + "balance_loss_clip": 0.97509843, + "balance_loss_mlp": 1.01489615, + "epoch": 0.9648289544880659, + "flos": 16545088016640.0, + "grad_norm": 3.6032782403758494, + "language_loss": 0.76161563, + "learning_rate": 1.2920545446891474e-08, + "loss": 0.78302324, + "num_input_tokens_seen": 173235465, + "step": 8024, + "time_per_iteration": 2.6062450408935547 + }, + { + "auxiliary_loss_clip": 0.01118834, + "auxiliary_loss_mlp": 0.0103398, + "balance_loss_clip": 0.97300905, + "balance_loss_mlp": 1.02691984, + "epoch": 0.964949197378705, + "flos": 24057527857920.0, + "grad_norm": 1.5932098360803286, + "language_loss": 0.70828414, + "learning_rate": 1.2832294690703127e-08, + "loss": 0.72981226, + "num_input_tokens_seen": 173254440, + "step": 8025, + "time_per_iteration": 2.605433464050293 + }, + { + "auxiliary_loss_clip": 0.01139864, + "auxiliary_loss_mlp": 0.01023937, + "balance_loss_clip": 1.00995505, + "balance_loss_mlp": 1.01682019, + "epoch": 0.965069440269344, + "flos": 23365565280000.0, + "grad_norm": 2.6459001553800534, + "language_loss": 0.78034908, + "learning_rate": 1.2744345386275668e-08, + "loss": 0.80198705, + "num_input_tokens_seen": 173273980, + "step": 8026, + "time_per_iteration": 2.5141358375549316 + }, + { + "auxiliary_loss_clip": 0.01126246, + "auxiliary_loss_mlp": 0.01027765, + "balance_loss_clip": 0.97566819, + "balance_loss_mlp": 1.02068996, + "epoch": 0.9651896831599832, + "flos": 25374875155200.0, + "grad_norm": 2.229869841900245, + "language_loss": 0.78790373, + "learning_rate": 1.265669754695109e-08, + "loss": 0.80944383, + "num_input_tokens_seen": 173293550, + "step": 8027, + "time_per_iteration": 2.6975743770599365 + }, + { + "auxiliary_loss_clip": 0.01050792, + "auxiliary_loss_mlp": 0.01026164, + "balance_loss_clip": 0.8563413, + "balance_loss_mlp": 1.01843333, + "epoch": 0.9653099260506223, + "flos": 22272875596800.0, + "grad_norm": 1.936838555307079, + "language_loss": 0.82016492, + "learning_rate": 1.2569351186025201e-08, + "loss": 0.84093451, + "num_input_tokens_seen": 173312005, + "step": 8028, + "time_per_iteration": 3.6560564041137695 + }, + { + "auxiliary_loss_clip": 0.0108186, + "auxiliary_loss_mlp": 0.01022893, + "balance_loss_clip": 0.92967087, + "balance_loss_mlp": 1.01635742, + "epoch": 0.9654301689412613, + "flos": 26760847386240.0, + "grad_norm": 1.514694988598872, + "language_loss": 0.75643039, + "learning_rate": 1.2482306316748737e-08, + "loss": 0.77747798, + "num_input_tokens_seen": 173332450, + "step": 8029, + "time_per_iteration": 2.6973838806152344 + }, + { + "auxiliary_loss_clip": 0.01145545, + "auxiliary_loss_mlp": 0.01023976, + "balance_loss_clip": 1.00906253, + "balance_loss_mlp": 1.01667738, + "epoch": 0.9655504118319005, + "flos": 17412689122560.0, + "grad_norm": 2.152599971415859, + "language_loss": 0.78478062, + "learning_rate": 1.2395562952326021e-08, + "loss": 0.80647582, + "num_input_tokens_seen": 173349610, + "step": 8030, + "time_per_iteration": 2.5942604541778564 + }, + { + "auxiliary_loss_clip": 0.0112401, + "auxiliary_loss_mlp": 0.01028485, + "balance_loss_clip": 0.97212124, + "balance_loss_mlp": 1.02065289, + "epoch": 0.9656706547225395, + "flos": 22126970551680.0, + "grad_norm": 2.524190750173558, + "language_loss": 0.80926007, + "learning_rate": 1.2309121105916309e-08, + "loss": 0.83078504, + "num_input_tokens_seen": 173367900, + "step": 8031, + "time_per_iteration": 2.661585569381714 + }, + { + "auxiliary_loss_clip": 0.01143412, + "auxiliary_loss_mlp": 0.01030465, + "balance_loss_clip": 1.0100944, + "balance_loss_mlp": 1.02307427, + "epoch": 0.9657908976131786, + "flos": 37049289926400.0, + "grad_norm": 1.6770392364725075, + "language_loss": 0.69304985, + "learning_rate": 1.222298079063222e-08, + "loss": 0.71478868, + "num_input_tokens_seen": 173389040, + "step": 8032, + "time_per_iteration": 2.755265712738037 + }, + { + "auxiliary_loss_clip": 0.01139743, + "auxiliary_loss_mlp": 0.01023882, + "balance_loss_clip": 1.00984669, + "balance_loss_mlp": 1.01726258, + "epoch": 0.9659111405038178, + "flos": 24389809597440.0, + "grad_norm": 2.2094445482804206, + "language_loss": 0.72731113, + "learning_rate": 1.2137142019541524e-08, + "loss": 0.74894732, + "num_input_tokens_seen": 173407595, + "step": 8033, + "time_per_iteration": 2.6877241134643555 + }, + { + "auxiliary_loss_clip": 0.01124319, + "auxiliary_loss_mlp": 0.01022277, + "balance_loss_clip": 0.97173989, + "balance_loss_mlp": 1.01517844, + "epoch": 0.9660313833944568, + "flos": 25009412227200.0, + "grad_norm": 2.7951254682405593, + "language_loss": 0.73447847, + "learning_rate": 1.2051604805666027e-08, + "loss": 0.75594443, + "num_input_tokens_seen": 173424720, + "step": 8034, + "time_per_iteration": 2.801743507385254 + }, + { + "auxiliary_loss_clip": 0.01164843, + "auxiliary_loss_mlp": 0.00509191, + "balance_loss_clip": 1.04933214, + "balance_loss_mlp": 0.0, + "epoch": 0.9661516262850959, + "flos": 11801575895040.0, + "grad_norm": 1.8321680467584665, + "language_loss": 0.78128403, + "learning_rate": 1.196636916198135e-08, + "loss": 0.79802442, + "num_input_tokens_seen": 173442260, + "step": 8035, + "time_per_iteration": 2.626540422439575 + }, + { + "auxiliary_loss_clip": 0.01165387, + "auxiliary_loss_mlp": 0.01025436, + "balance_loss_clip": 1.04842854, + "balance_loss_mlp": 1.01814044, + "epoch": 0.9662718691757349, + "flos": 20047778766720.0, + "grad_norm": 2.1041532480331084, + "language_loss": 0.76985073, + "learning_rate": 1.1881435101418036e-08, + "loss": 0.79175901, + "num_input_tokens_seen": 173461675, + "step": 8036, + "time_per_iteration": 2.6500420570373535 + }, + { + "auxiliary_loss_clip": 0.01019728, + "auxiliary_loss_mlp": 0.00999792, + "balance_loss_clip": 0.93535459, + "balance_loss_mlp": 0.99886209, + "epoch": 0.9663921120663741, + "flos": 68027703517440.0, + "grad_norm": 0.7464585276203023, + "language_loss": 0.65558386, + "learning_rate": 1.1796802636860003e-08, + "loss": 0.67577899, + "num_input_tokens_seen": 173530205, + "step": 8037, + "time_per_iteration": 3.258732795715332 + }, + { + "auxiliary_loss_clip": 0.01165216, + "auxiliary_loss_mlp": 0.01022811, + "balance_loss_clip": 1.04882824, + "balance_loss_mlp": 1.01561689, + "epoch": 0.9665123549570132, + "flos": 26322916769280.0, + "grad_norm": 1.9030355244133241, + "language_loss": 0.73939967, + "learning_rate": 1.1712471781146316e-08, + "loss": 0.76127994, + "num_input_tokens_seen": 173549540, + "step": 8038, + "time_per_iteration": 2.6552114486694336 + }, + { + "auxiliary_loss_clip": 0.01161741, + "auxiliary_loss_mlp": 0.01021053, + "balance_loss_clip": 1.04643178, + "balance_loss_mlp": 1.01444864, + "epoch": 0.9666325978476522, + "flos": 43941121557120.0, + "grad_norm": 2.2080964225321513, + "language_loss": 0.6664995, + "learning_rate": 1.1628442547069628e-08, + "loss": 0.68832743, + "num_input_tokens_seen": 173571740, + "step": 8039, + "time_per_iteration": 2.7626898288726807 + }, + { + "auxiliary_loss_clip": 0.01144454, + "auxiliary_loss_mlp": 0.00509573, + "balance_loss_clip": 1.01041794, + "balance_loss_mlp": 0.0, + "epoch": 0.9667528407382914, + "flos": 21543422198400.0, + "grad_norm": 1.9948922131849365, + "language_loss": 0.77463078, + "learning_rate": 1.1544714947377521e-08, + "loss": 0.79117107, + "num_input_tokens_seen": 173589425, + "step": 8040, + "time_per_iteration": 2.6548008918762207 + }, + { + "auxiliary_loss_clip": 0.01166184, + "auxiliary_loss_mlp": 0.01027443, + "balance_loss_clip": 1.04976606, + "balance_loss_mlp": 1.02016807, + "epoch": 0.9668730836289304, + "flos": 23878585278720.0, + "grad_norm": 1.8657154790812687, + "language_loss": 0.70104194, + "learning_rate": 1.1461288994770945e-08, + "loss": 0.72297823, + "num_input_tokens_seen": 173608500, + "step": 8041, + "time_per_iteration": 2.6206395626068115 + }, + { + "auxiliary_loss_clip": 0.01164132, + "auxiliary_loss_mlp": 0.0102653, + "balance_loss_clip": 1.04482365, + "balance_loss_mlp": 1.0195446, + "epoch": 0.9669933265195695, + "flos": 28293011971200.0, + "grad_norm": 1.7417995577264134, + "language_loss": 0.77146971, + "learning_rate": 1.1378164701906002e-08, + "loss": 0.79337633, + "num_input_tokens_seen": 173630265, + "step": 8042, + "time_per_iteration": 2.6834771633148193 + }, + { + "auxiliary_loss_clip": 0.0116672, + "auxiliary_loss_mlp": 0.01024203, + "balance_loss_clip": 1.04890203, + "balance_loss_mlp": 1.01655853, + "epoch": 0.9671135694102087, + "flos": 22454763091200.0, + "grad_norm": 1.8435768005007018, + "language_loss": 0.66976166, + "learning_rate": 1.1295342081392156e-08, + "loss": 0.69167089, + "num_input_tokens_seen": 173649625, + "step": 8043, + "time_per_iteration": 2.6048099994659424 + }, + { + "auxiliary_loss_clip": 0.01121691, + "auxiliary_loss_mlp": 0.01024304, + "balance_loss_clip": 0.97186261, + "balance_loss_mlp": 1.01738429, + "epoch": 0.9672338123008477, + "flos": 20155941596160.0, + "grad_norm": 1.839164564949704, + "language_loss": 0.70048827, + "learning_rate": 1.1212821145793804e-08, + "loss": 0.72194815, + "num_input_tokens_seen": 173669240, + "step": 8044, + "time_per_iteration": 2.6305837631225586 + }, + { + "auxiliary_loss_clip": 0.01117265, + "auxiliary_loss_mlp": 0.01029086, + "balance_loss_clip": 0.97035569, + "balance_loss_mlp": 1.02178752, + "epoch": 0.9673540551914868, + "flos": 16977487939200.0, + "grad_norm": 1.9483839476147229, + "language_loss": 0.78898048, + "learning_rate": 1.1130601907629156e-08, + "loss": 0.810444, + "num_input_tokens_seen": 173686970, + "step": 8045, + "time_per_iteration": 4.836038112640381 + }, + { + "auxiliary_loss_clip": 0.01041587, + "auxiliary_loss_mlp": 0.01001611, + "balance_loss_clip": 0.97160685, + "balance_loss_mlp": 1.00062704, + "epoch": 0.9674742980821259, + "flos": 61892903952000.0, + "grad_norm": 0.8098634125457269, + "language_loss": 0.64800858, + "learning_rate": 1.1048684379370899e-08, + "loss": 0.66844046, + "num_input_tokens_seen": 173747655, + "step": 8046, + "time_per_iteration": 3.160395622253418 + }, + { + "auxiliary_loss_clip": 0.01107558, + "auxiliary_loss_mlp": 0.01019506, + "balance_loss_clip": 0.97003055, + "balance_loss_mlp": 1.0128721, + "epoch": 0.967594540972765, + "flos": 18697824898560.0, + "grad_norm": 1.8138516721149802, + "language_loss": 0.74380374, + "learning_rate": 1.0967068573445759e-08, + "loss": 0.76507437, + "num_input_tokens_seen": 173765140, + "step": 8047, + "time_per_iteration": 3.5834078788757324 + }, + { + "auxiliary_loss_clip": 0.01112391, + "auxiliary_loss_mlp": 0.01021761, + "balance_loss_clip": 0.96865731, + "balance_loss_mlp": 1.01501417, + "epoch": 0.967714783863404, + "flos": 20777411733120.0, + "grad_norm": 2.1716464753052986, + "language_loss": 0.65187311, + "learning_rate": 1.0885754502234945e-08, + "loss": 0.67321461, + "num_input_tokens_seen": 173784800, + "step": 8048, + "time_per_iteration": 2.675633192062378 + }, + { + "auxiliary_loss_clip": 0.01091495, + "auxiliary_loss_mlp": 0.01024, + "balance_loss_clip": 0.93271995, + "balance_loss_mlp": 1.01757431, + "epoch": 0.9678350267540432, + "flos": 23185473465600.0, + "grad_norm": 3.515809411350283, + "language_loss": 0.78082263, + "learning_rate": 1.08047421780737e-08, + "loss": 0.80197752, + "num_input_tokens_seen": 173803990, + "step": 8049, + "time_per_iteration": 2.708982467651367 + }, + { + "auxiliary_loss_clip": 0.01123128, + "auxiliary_loss_mlp": 0.00509635, + "balance_loss_clip": 0.97167921, + "balance_loss_mlp": 0.0, + "epoch": 0.9679552696446823, + "flos": 21726063878400.0, + "grad_norm": 2.072221244833599, + "language_loss": 0.7410652, + "learning_rate": 1.0724031613251305e-08, + "loss": 0.75739288, + "num_input_tokens_seen": 173821890, + "step": 8050, + "time_per_iteration": 2.6736230850219727 + }, + { + "auxiliary_loss_clip": 0.01146833, + "auxiliary_loss_mlp": 0.0102903, + "balance_loss_clip": 1.01041639, + "balance_loss_mlp": 1.02121615, + "epoch": 0.9680755125353213, + "flos": 26869046129280.0, + "grad_norm": 1.8934199158615652, + "language_loss": 0.66566497, + "learning_rate": 1.0643622820011744e-08, + "loss": 0.68742359, + "num_input_tokens_seen": 173842945, + "step": 8051, + "time_per_iteration": 2.6803781986236572 + }, + { + "auxiliary_loss_clip": 0.01165892, + "auxiliary_loss_mlp": 0.01023351, + "balance_loss_clip": 1.04667151, + "balance_loss_mlp": 1.01555502, + "epoch": 0.9681957554259605, + "flos": 28325008010880.0, + "grad_norm": 2.5004590172411443, + "language_loss": 0.68424249, + "learning_rate": 1.0563515810552814e-08, + "loss": 0.70613492, + "num_input_tokens_seen": 173859915, + "step": 8052, + "time_per_iteration": 2.633096694946289 + }, + { + "auxiliary_loss_clip": 0.01166799, + "auxiliary_loss_mlp": 0.01027266, + "balance_loss_clip": 1.05185449, + "balance_loss_mlp": 1.01983309, + "epoch": 0.9683159983165995, + "flos": 20557674282240.0, + "grad_norm": 1.5494601586857448, + "language_loss": 0.73453069, + "learning_rate": 1.0483710597026795e-08, + "loss": 0.75647134, + "num_input_tokens_seen": 173879775, + "step": 8053, + "time_per_iteration": 2.6061670780181885 + }, + { + "auxiliary_loss_clip": 0.01093859, + "auxiliary_loss_mlp": 0.01024024, + "balance_loss_clip": 0.93239403, + "balance_loss_mlp": 1.01710343, + "epoch": 0.9684362412072386, + "flos": 24207958016640.0, + "grad_norm": 2.0626443080808783, + "language_loss": 0.73859501, + "learning_rate": 1.0404207191540227e-08, + "loss": 0.75977379, + "num_input_tokens_seen": 173900230, + "step": 8054, + "time_per_iteration": 3.7717838287353516 + }, + { + "auxiliary_loss_clip": 0.01160974, + "auxiliary_loss_mlp": 0.0102463, + "balance_loss_clip": 1.0466938, + "balance_loss_mlp": 1.01766479, + "epoch": 0.9685564840978778, + "flos": 22346241125760.0, + "grad_norm": 1.77382391898376, + "language_loss": 0.74924743, + "learning_rate": 1.0325005606153236e-08, + "loss": 0.7711035, + "num_input_tokens_seen": 173919690, + "step": 8055, + "time_per_iteration": 2.60487961769104 + }, + { + "auxiliary_loss_clip": 0.0107485, + "auxiliary_loss_mlp": 0.0102436, + "balance_loss_clip": 0.89447391, + "balance_loss_mlp": 1.01732934, + "epoch": 0.9686767269885168, + "flos": 14386389477120.0, + "grad_norm": 2.352611485564682, + "language_loss": 0.79636252, + "learning_rate": 1.0246105852881104e-08, + "loss": 0.81735462, + "num_input_tokens_seen": 173934790, + "step": 8056, + "time_per_iteration": 2.7147934436798096 + }, + { + "auxiliary_loss_clip": 0.01164542, + "auxiliary_loss_mlp": 0.01024209, + "balance_loss_clip": 1.04735327, + "balance_loss_mlp": 1.01690149, + "epoch": 0.9687969698791559, + "flos": 21287630471040.0, + "grad_norm": 1.8832919241037867, + "language_loss": 0.78608501, + "learning_rate": 1.0167507943692476e-08, + "loss": 0.80797255, + "num_input_tokens_seen": 173953875, + "step": 8057, + "time_per_iteration": 2.581787347793579 + }, + { + "auxiliary_loss_clip": 0.01142489, + "auxiliary_loss_mlp": 0.01026466, + "balance_loss_clip": 1.01330781, + "balance_loss_mlp": 1.01877069, + "epoch": 0.968917212769795, + "flos": 19828328624640.0, + "grad_norm": 2.0236363116722575, + "language_loss": 0.71412718, + "learning_rate": 1.008921189051093e-08, + "loss": 0.73581672, + "num_input_tokens_seen": 173971220, + "step": 8058, + "time_per_iteration": 2.5957064628601074 + }, + { + "auxiliary_loss_clip": 0.01166045, + "auxiliary_loss_mlp": 0.01027267, + "balance_loss_clip": 1.05036509, + "balance_loss_mlp": 1.02024305, + "epoch": 0.9690374556604341, + "flos": 21681749473920.0, + "grad_norm": 2.4496976616219768, + "language_loss": 0.77429879, + "learning_rate": 1.0011217705213848e-08, + "loss": 0.79623187, + "num_input_tokens_seen": 173989095, + "step": 8059, + "time_per_iteration": 2.595085620880127 + }, + { + "auxiliary_loss_clip": 0.01137432, + "auxiliary_loss_mlp": 0.01026673, + "balance_loss_clip": 1.00983167, + "balance_loss_mlp": 1.02019072, + "epoch": 0.9691576985510731, + "flos": 32635437851520.0, + "grad_norm": 1.7665997281148875, + "language_loss": 0.74907589, + "learning_rate": 9.933525399632658e-09, + "loss": 0.77071691, + "num_input_tokens_seen": 174007330, + "step": 8060, + "time_per_iteration": 2.6901044845581055 + }, + { + "auxiliary_loss_clip": 0.01115455, + "auxiliary_loss_mlp": 0.01032731, + "balance_loss_clip": 0.97214782, + "balance_loss_mlp": 1.02481294, + "epoch": 0.9692779414417123, + "flos": 35663174040960.0, + "grad_norm": 2.113748241311556, + "language_loss": 0.64817059, + "learning_rate": 9.856134985553488e-09, + "loss": 0.66965246, + "num_input_tokens_seen": 174027055, + "step": 8061, + "time_per_iteration": 2.7906785011291504 + }, + { + "auxiliary_loss_clip": 0.01162716, + "auxiliary_loss_mlp": 0.01022316, + "balance_loss_clip": 1.04764462, + "balance_loss_mlp": 1.01544642, + "epoch": 0.9693981843323514, + "flos": 28366952117760.0, + "grad_norm": 2.158225253829676, + "language_loss": 0.73727763, + "learning_rate": 9.77904647471628e-09, + "loss": 0.75912797, + "num_input_tokens_seen": 174050235, + "step": 8062, + "time_per_iteration": 2.682793140411377 + }, + { + "auxiliary_loss_clip": 0.01062421, + "auxiliary_loss_mlp": 0.01024796, + "balance_loss_clip": 0.89175498, + "balance_loss_mlp": 1.01771474, + "epoch": 0.9695184272229904, + "flos": 23622865378560.0, + "grad_norm": 1.4246348411737515, + "language_loss": 0.7420609, + "learning_rate": 9.702259878815454e-09, + "loss": 0.76293302, + "num_input_tokens_seen": 174070560, + "step": 8063, + "time_per_iteration": 2.6882052421569824 + }, + { + "auxiliary_loss_clip": 0.01144193, + "auxiliary_loss_mlp": 0.01030657, + "balance_loss_clip": 1.01257396, + "balance_loss_mlp": 1.02345645, + "epoch": 0.9696386701136296, + "flos": 23294677789440.0, + "grad_norm": 2.7375613861093933, + "language_loss": 0.74384785, + "learning_rate": 9.625775209499254e-09, + "loss": 0.76559627, + "num_input_tokens_seen": 174090565, + "step": 8064, + "time_per_iteration": 2.6190688610076904 + }, + { + "auxiliary_loss_clip": 0.01084458, + "auxiliary_loss_mlp": 0.01023711, + "balance_loss_clip": 0.92912853, + "balance_loss_mlp": 1.01685596, + "epoch": 0.9697589130042686, + "flos": 15121876360320.0, + "grad_norm": 2.337783003767951, + "language_loss": 0.7385273, + "learning_rate": 9.549592478370172e-09, + "loss": 0.75960904, + "num_input_tokens_seen": 174108745, + "step": 8065, + "time_per_iteration": 2.675015926361084 + }, + { + "auxiliary_loss_clip": 0.01140858, + "auxiliary_loss_mlp": 0.01028517, + "balance_loss_clip": 1.0078913, + "balance_loss_mlp": 1.02177596, + "epoch": 0.9698791558949077, + "flos": 18879532824960.0, + "grad_norm": 1.6993379685400436, + "language_loss": 0.79200184, + "learning_rate": 9.473711696985632e-09, + "loss": 0.81369567, + "num_input_tokens_seen": 174128075, + "step": 8066, + "time_per_iteration": 2.6527836322784424 + }, + { + "auxiliary_loss_clip": 0.01116286, + "auxiliary_loss_mlp": 0.01027776, + "balance_loss_clip": 0.97028279, + "balance_loss_mlp": 1.02055752, + "epoch": 0.9699993987855468, + "flos": 17931455297280.0, + "grad_norm": 3.0599821818014905, + "language_loss": 0.76121414, + "learning_rate": 9.398132876856201e-09, + "loss": 0.78265476, + "num_input_tokens_seen": 174147040, + "step": 8067, + "time_per_iteration": 2.6193737983703613 + }, + { + "auxiliary_loss_clip": 0.00981869, + "auxiliary_loss_mlp": 0.01002074, + "balance_loss_clip": 0.86174011, + "balance_loss_mlp": 1.00104296, + "epoch": 0.9701196416761859, + "flos": 67182186297600.0, + "grad_norm": 0.7745588892668203, + "language_loss": 0.60764444, + "learning_rate": 9.322856029447379e-09, + "loss": 0.62748384, + "num_input_tokens_seen": 174208225, + "step": 8068, + "time_per_iteration": 3.226806879043579 + }, + { + "auxiliary_loss_clip": 0.01161035, + "auxiliary_loss_mlp": 0.0102686, + "balance_loss_clip": 1.04740977, + "balance_loss_mlp": 1.01966321, + "epoch": 0.970239884566825, + "flos": 24277804012800.0, + "grad_norm": 1.7896275552099232, + "language_loss": 0.80540264, + "learning_rate": 9.247881166178695e-09, + "loss": 0.82728159, + "num_input_tokens_seen": 174226935, + "step": 8069, + "time_per_iteration": 2.6150949001312256 + }, + { + "auxiliary_loss_clip": 0.01101262, + "auxiliary_loss_mlp": 0.01028592, + "balance_loss_clip": 0.93470073, + "balance_loss_mlp": 1.02151132, + "epoch": 0.970360127457464, + "flos": 25301689194240.0, + "grad_norm": 2.06482928007303, + "language_loss": 0.76672375, + "learning_rate": 9.173208298423274e-09, + "loss": 0.78802228, + "num_input_tokens_seen": 174248140, + "step": 8070, + "time_per_iteration": 3.7889151573181152 + }, + { + "auxiliary_loss_clip": 0.01069926, + "auxiliary_loss_mlp": 0.00508732, + "balance_loss_clip": 0.89622265, + "balance_loss_mlp": 0.0, + "epoch": 0.9704803703481032, + "flos": 29572473398400.0, + "grad_norm": 1.6006326191961473, + "language_loss": 0.76257879, + "learning_rate": 9.09883743750961e-09, + "loss": 0.77836537, + "num_input_tokens_seen": 174271030, + "step": 8071, + "time_per_iteration": 3.7422211170196533 + }, + { + "auxiliary_loss_clip": 0.01115707, + "auxiliary_loss_mlp": 0.0102355, + "balance_loss_clip": 0.9710654, + "balance_loss_mlp": 1.01634383, + "epoch": 0.9706006132387422, + "flos": 17380046638080.0, + "grad_norm": 1.6156832242834451, + "language_loss": 0.84040689, + "learning_rate": 9.024768594719124e-09, + "loss": 0.86179954, + "num_input_tokens_seen": 174289410, + "step": 8072, + "time_per_iteration": 2.7042884826660156 + }, + { + "auxiliary_loss_clip": 0.01095475, + "auxiliary_loss_mlp": 0.01025099, + "balance_loss_clip": 0.93527216, + "balance_loss_mlp": 1.0184207, + "epoch": 0.9707208561293813, + "flos": 18186421011840.0, + "grad_norm": 2.4647171644402435, + "language_loss": 0.72491097, + "learning_rate": 8.95100178128816e-09, + "loss": 0.74611676, + "num_input_tokens_seen": 174308550, + "step": 8073, + "time_per_iteration": 3.74307918548584 + }, + { + "auxiliary_loss_clip": 0.01118651, + "auxiliary_loss_mlp": 0.01022215, + "balance_loss_clip": 0.97153199, + "balance_loss_mlp": 1.01562297, + "epoch": 0.9708410990200205, + "flos": 31248388212480.0, + "grad_norm": 1.9653137829293645, + "language_loss": 0.69951963, + "learning_rate": 8.877537008407321e-09, + "loss": 0.72092831, + "num_input_tokens_seen": 174328600, + "step": 8074, + "time_per_iteration": 2.7977962493896484 + }, + { + "auxiliary_loss_clip": 0.01122807, + "auxiliary_loss_mlp": 0.01024226, + "balance_loss_clip": 0.97160351, + "balance_loss_mlp": 1.0171715, + "epoch": 0.9709613419106595, + "flos": 30554450386560.0, + "grad_norm": 1.7322426652734186, + "language_loss": 0.68855304, + "learning_rate": 8.804374287221028e-09, + "loss": 0.71002328, + "num_input_tokens_seen": 174349835, + "step": 8075, + "time_per_iteration": 2.7176129817962646 + }, + { + "auxiliary_loss_clip": 0.01085662, + "auxiliary_loss_mlp": 0.0102715, + "balance_loss_clip": 0.92778254, + "balance_loss_mlp": 1.02014303, + "epoch": 0.9710815848012986, + "flos": 23730166281600.0, + "grad_norm": 1.625004411857327, + "language_loss": 0.84753418, + "learning_rate": 8.731513628827958e-09, + "loss": 0.8686623, + "num_input_tokens_seen": 174369200, + "step": 8076, + "time_per_iteration": 2.738823175430298 + }, + { + "auxiliary_loss_clip": 0.01141933, + "auxiliary_loss_mlp": 0.01026222, + "balance_loss_clip": 1.01029611, + "balance_loss_mlp": 1.01970744, + "epoch": 0.9712018276919377, + "flos": 23761875012480.0, + "grad_norm": 2.0936306092871804, + "language_loss": 0.82615447, + "learning_rate": 8.658955044280825e-09, + "loss": 0.84783602, + "num_input_tokens_seen": 174388125, + "step": 8077, + "time_per_iteration": 2.6478030681610107 + }, + { + "auxiliary_loss_clip": 0.01141432, + "auxiliary_loss_mlp": 0.01022238, + "balance_loss_clip": 1.01272953, + "balance_loss_mlp": 1.0148555, + "epoch": 0.9713220705825768, + "flos": 23330983461120.0, + "grad_norm": 1.8376794295420589, + "language_loss": 0.77579892, + "learning_rate": 8.586698544587268e-09, + "loss": 0.79743564, + "num_input_tokens_seen": 174409735, + "step": 8078, + "time_per_iteration": 2.686116933822632 + }, + { + "auxiliary_loss_clip": 0.01110457, + "auxiliary_loss_mlp": 0.01026982, + "balance_loss_clip": 0.97083825, + "balance_loss_mlp": 1.01999295, + "epoch": 0.9714423134732159, + "flos": 22200946611840.0, + "grad_norm": 1.9786785859322316, + "language_loss": 0.74516052, + "learning_rate": 8.514744140707853e-09, + "loss": 0.76653492, + "num_input_tokens_seen": 174428875, + "step": 8079, + "time_per_iteration": 2.6173927783966064 + }, + { + "auxiliary_loss_clip": 0.01161308, + "auxiliary_loss_mlp": 0.0102471, + "balance_loss_clip": 1.04744601, + "balance_loss_mlp": 1.01786709, + "epoch": 0.971562556363855, + "flos": 20229917656320.0, + "grad_norm": 1.5402183997973051, + "language_loss": 0.76497042, + "learning_rate": 8.443091843558515e-09, + "loss": 0.7868306, + "num_input_tokens_seen": 174447960, + "step": 8080, + "time_per_iteration": 2.690528154373169 + }, + { + "auxiliary_loss_clip": 0.0111103, + "auxiliary_loss_mlp": 0.01025655, + "balance_loss_clip": 0.97006011, + "balance_loss_mlp": 1.01853275, + "epoch": 0.9716827992544941, + "flos": 24970197553920.0, + "grad_norm": 2.084403665463299, + "language_loss": 0.64378619, + "learning_rate": 8.37174166400878e-09, + "loss": 0.66515303, + "num_input_tokens_seen": 174463535, + "step": 8081, + "time_per_iteration": 3.5897161960601807 + }, + { + "auxiliary_loss_clip": 0.01166428, + "auxiliary_loss_mlp": 0.01024513, + "balance_loss_clip": 1.05190468, + "balance_loss_mlp": 1.01711011, + "epoch": 0.9718030421451331, + "flos": 24681476033280.0, + "grad_norm": 3.9472566130499653, + "language_loss": 0.85075426, + "learning_rate": 8.300693612881992e-09, + "loss": 0.87266368, + "num_input_tokens_seen": 174483600, + "step": 8082, + "time_per_iteration": 2.614617347717285 + }, + { + "auxiliary_loss_clip": 0.01138429, + "auxiliary_loss_mlp": 0.00508946, + "balance_loss_clip": 1.0104816, + "balance_loss_mlp": 0.0, + "epoch": 0.9719232850357723, + "flos": 22090700793600.0, + "grad_norm": 1.8882977379243675, + "language_loss": 0.81220663, + "learning_rate": 8.22994770095664e-09, + "loss": 0.8286804, + "num_input_tokens_seen": 174502175, + "step": 8083, + "time_per_iteration": 2.6416916847229004 + }, + { + "auxiliary_loss_clip": 0.01121683, + "auxiliary_loss_mlp": 0.01024105, + "balance_loss_clip": 0.9774915, + "balance_loss_mlp": 1.01670814, + "epoch": 0.9720435279264114, + "flos": 23656908493440.0, + "grad_norm": 2.026348903959501, + "language_loss": 0.75032955, + "learning_rate": 8.159503938964585e-09, + "loss": 0.77178741, + "num_input_tokens_seen": 174519495, + "step": 8084, + "time_per_iteration": 2.660274028778076 + }, + { + "auxiliary_loss_clip": 0.01088894, + "auxiliary_loss_mlp": 0.01027896, + "balance_loss_clip": 0.93370682, + "balance_loss_mlp": 1.02152455, + "epoch": 0.9721637708170504, + "flos": 28365910623360.0, + "grad_norm": 1.9611710589703693, + "language_loss": 0.70485961, + "learning_rate": 8.089362337592164e-09, + "loss": 0.72602749, + "num_input_tokens_seen": 174543120, + "step": 8085, + "time_per_iteration": 2.7828361988067627 + }, + { + "auxiliary_loss_clip": 0.01113573, + "auxiliary_loss_mlp": 0.01023034, + "balance_loss_clip": 0.97186804, + "balance_loss_mlp": 1.01585174, + "epoch": 0.9722840137076896, + "flos": 29130807767040.0, + "grad_norm": 1.506240572946744, + "language_loss": 0.72130179, + "learning_rate": 8.019522907479536e-09, + "loss": 0.74266785, + "num_input_tokens_seen": 174563480, + "step": 8086, + "time_per_iteration": 2.6897835731506348 + }, + { + "auxiliary_loss_clip": 0.01143919, + "auxiliary_loss_mlp": 0.01030349, + "balance_loss_clip": 1.01172304, + "balance_loss_mlp": 1.02330375, + "epoch": 0.9724042565983286, + "flos": 19243954258560.0, + "grad_norm": 2.6418143217689583, + "language_loss": 0.77399766, + "learning_rate": 7.949985659221558e-09, + "loss": 0.79574037, + "num_input_tokens_seen": 174580745, + "step": 8087, + "time_per_iteration": 2.6029391288757324 + }, + { + "auxiliary_loss_clip": 0.01118537, + "auxiliary_loss_mlp": 0.01024321, + "balance_loss_clip": 0.9701978, + "balance_loss_mlp": 1.01798248, + "epoch": 0.9725244994889677, + "flos": 23039676161280.0, + "grad_norm": 2.0494817726662027, + "language_loss": 0.7895906, + "learning_rate": 7.880750603366904e-09, + "loss": 0.81101906, + "num_input_tokens_seen": 174599615, + "step": 8088, + "time_per_iteration": 2.631101608276367 + }, + { + "auxiliary_loss_clip": 0.01102255, + "auxiliary_loss_mlp": 0.01027112, + "balance_loss_clip": 0.93173814, + "balance_loss_mlp": 1.01996791, + "epoch": 0.9726447423796069, + "flos": 23367468700800.0, + "grad_norm": 3.250943093361066, + "language_loss": 0.79682046, + "learning_rate": 7.811817750418282e-09, + "loss": 0.81811416, + "num_input_tokens_seen": 174618375, + "step": 8089, + "time_per_iteration": 2.7375197410583496 + }, + { + "auxiliary_loss_clip": 0.01094603, + "auxiliary_loss_mlp": 0.01025715, + "balance_loss_clip": 0.93586123, + "balance_loss_mlp": 1.01819313, + "epoch": 0.9727649852702459, + "flos": 26541648639360.0, + "grad_norm": 2.0506945032995723, + "language_loss": 0.80136478, + "learning_rate": 7.743187110833105e-09, + "loss": 0.82256794, + "num_input_tokens_seen": 174641135, + "step": 8090, + "time_per_iteration": 2.7491800785064697 + }, + { + "auxiliary_loss_clip": 0.0111757, + "auxiliary_loss_mlp": 0.01025961, + "balance_loss_clip": 0.96882027, + "balance_loss_mlp": 1.01927924, + "epoch": 0.972885228160885, + "flos": 20522338277760.0, + "grad_norm": 1.5243278255303276, + "language_loss": 0.80663669, + "learning_rate": 7.674858695022602e-09, + "loss": 0.82807195, + "num_input_tokens_seen": 174659490, + "step": 8091, + "time_per_iteration": 2.647831916809082 + }, + { + "auxiliary_loss_clip": 0.01166514, + "auxiliary_loss_mlp": 0.01022467, + "balance_loss_clip": 1.04886162, + "balance_loss_mlp": 1.01480746, + "epoch": 0.9730054710515241, + "flos": 17566064196480.0, + "grad_norm": 2.601016079445074, + "language_loss": 0.76145768, + "learning_rate": 7.606832513351591e-09, + "loss": 0.78334749, + "num_input_tokens_seen": 174677440, + "step": 8092, + "time_per_iteration": 2.6185920238494873 + }, + { + "auxiliary_loss_clip": 0.01059114, + "auxiliary_loss_mlp": 0.00500667, + "balance_loss_clip": 1.00864577, + "balance_loss_mlp": 0.0, + "epoch": 0.9731257139421632, + "flos": 68972010117120.0, + "grad_norm": 0.8258823534502003, + "language_loss": 0.6399914, + "learning_rate": 7.539108576140264e-09, + "loss": 0.65558922, + "num_input_tokens_seen": 174741550, + "step": 8093, + "time_per_iteration": 3.2254090309143066 + }, + { + "auxiliary_loss_clip": 0.01069508, + "auxiliary_loss_mlp": 0.01023559, + "balance_loss_clip": 0.89533412, + "balance_loss_mlp": 1.01612663, + "epoch": 0.9732459568328022, + "flos": 18478841633280.0, + "grad_norm": 1.8459569249024568, + "language_loss": 0.70225024, + "learning_rate": 7.471686893661732e-09, + "loss": 0.72318101, + "num_input_tokens_seen": 174759845, + "step": 8094, + "time_per_iteration": 2.7498369216918945 + }, + { + "auxiliary_loss_clip": 0.01114801, + "auxiliary_loss_mlp": 0.0102727, + "balance_loss_clip": 0.97253424, + "balance_loss_mlp": 1.02037132, + "epoch": 0.9733661997234414, + "flos": 20883886623360.0, + "grad_norm": 1.680180415654208, + "language_loss": 0.64271736, + "learning_rate": 7.4045674761442636e-09, + "loss": 0.66413808, + "num_input_tokens_seen": 174777175, + "step": 8095, + "time_per_iteration": 2.617680311203003 + }, + { + "auxiliary_loss_clip": 0.01161613, + "auxiliary_loss_mlp": 0.00508431, + "balance_loss_clip": 1.04712844, + "balance_loss_mlp": 0.0, + "epoch": 0.9734864426140805, + "flos": 23766795175680.0, + "grad_norm": 1.6655504093112345, + "language_loss": 0.74406767, + "learning_rate": 7.337750333769488e-09, + "loss": 0.760768, + "num_input_tokens_seen": 174796980, + "step": 8096, + "time_per_iteration": 3.557244300842285 + }, + { + "auxiliary_loss_clip": 0.01120108, + "auxiliary_loss_mlp": 0.01026873, + "balance_loss_clip": 0.96691042, + "balance_loss_mlp": 1.01989293, + "epoch": 0.9736066855047195, + "flos": 35042422176000.0, + "grad_norm": 2.2363162556298697, + "language_loss": 0.73005146, + "learning_rate": 7.2712354766737425e-09, + "loss": 0.75152123, + "num_input_tokens_seen": 174817310, + "step": 8097, + "time_per_iteration": 3.7224152088165283 + }, + { + "auxiliary_loss_clip": 0.01087693, + "auxiliary_loss_mlp": 0.01028868, + "balance_loss_clip": 0.93573725, + "balance_loss_mlp": 1.02126193, + "epoch": 0.9737269283953586, + "flos": 20410620001920.0, + "grad_norm": 4.269830864848792, + "language_loss": 0.81061119, + "learning_rate": 7.2050229149469565e-09, + "loss": 0.8317768, + "num_input_tokens_seen": 174837320, + "step": 8098, + "time_per_iteration": 2.697587251663208 + }, + { + "auxiliary_loss_clip": 0.01095357, + "auxiliary_loss_mlp": 0.01021616, + "balance_loss_clip": 0.92893553, + "balance_loss_mlp": 1.01448154, + "epoch": 0.9738471712859977, + "flos": 28911680847360.0, + "grad_norm": 1.7162015327250633, + "language_loss": 0.6379776, + "learning_rate": 7.139112658633984e-09, + "loss": 0.65914738, + "num_input_tokens_seen": 174857470, + "step": 8099, + "time_per_iteration": 3.6957435607910156 + }, + { + "auxiliary_loss_clip": 0.01089316, + "auxiliary_loss_mlp": 0.01023918, + "balance_loss_clip": 0.93272442, + "balance_loss_mlp": 1.01710832, + "epoch": 0.9739674141766368, + "flos": 27782326356480.0, + "grad_norm": 2.1228322960035655, + "language_loss": 0.70391935, + "learning_rate": 7.073504717733048e-09, + "loss": 0.72505164, + "num_input_tokens_seen": 174877035, + "step": 8100, + "time_per_iteration": 2.793489694595337 + }, + { + "auxiliary_loss_clip": 0.0098436, + "auxiliary_loss_mlp": 0.01009733, + "balance_loss_clip": 0.86427867, + "balance_loss_mlp": 1.0086422, + "epoch": 0.9740876570672758, + "flos": 68863057188480.0, + "grad_norm": 0.730902769808324, + "language_loss": 0.57291406, + "learning_rate": 7.008199102196855e-09, + "loss": 0.59285498, + "num_input_tokens_seen": 174938460, + "step": 8101, + "time_per_iteration": 3.287557363510132 + }, + { + "auxiliary_loss_clip": 0.01010275, + "auxiliary_loss_mlp": 0.01001943, + "balance_loss_clip": 0.9345032, + "balance_loss_mlp": 1.00075042, + "epoch": 0.974207899957915, + "flos": 58236622646400.0, + "grad_norm": 0.7946789848705229, + "language_loss": 0.58919382, + "learning_rate": 6.9431958219321464e-09, + "loss": 0.60931599, + "num_input_tokens_seen": 174994625, + "step": 8102, + "time_per_iteration": 3.2271344661712646 + }, + { + "auxiliary_loss_clip": 0.01116973, + "auxiliary_loss_mlp": 0.01019578, + "balance_loss_clip": 0.97014976, + "balance_loss_mlp": 1.01233602, + "epoch": 0.9743281428485541, + "flos": 22600057605120.0, + "grad_norm": 1.6384326533008517, + "language_loss": 0.77880633, + "learning_rate": 6.878494886800146e-09, + "loss": 0.80017185, + "num_input_tokens_seen": 175015400, + "step": 8103, + "time_per_iteration": 2.708652973175049 + }, + { + "auxiliary_loss_clip": 0.0112059, + "auxiliary_loss_mlp": 0.01021441, + "balance_loss_clip": 0.97238731, + "balance_loss_mlp": 1.01451468, + "epoch": 0.9744483857391931, + "flos": 20008815488640.0, + "grad_norm": 3.0212540694463277, + "language_loss": 0.76684219, + "learning_rate": 6.814096306615669e-09, + "loss": 0.78826249, + "num_input_tokens_seen": 175033540, + "step": 8104, + "time_per_iteration": 2.6647820472717285 + }, + { + "auxiliary_loss_clip": 0.01120842, + "auxiliary_loss_mlp": 0.01029518, + "balance_loss_clip": 0.96893954, + "balance_loss_mlp": 1.02215981, + "epoch": 0.9745686286298323, + "flos": 17675268520320.0, + "grad_norm": 2.757990320062099, + "language_loss": 0.65720445, + "learning_rate": 6.750000091148011e-09, + "loss": 0.67870808, + "num_input_tokens_seen": 175050835, + "step": 8105, + "time_per_iteration": 2.648524045944214 + }, + { + "auxiliary_loss_clip": 0.01165834, + "auxiliary_loss_mlp": 0.01023961, + "balance_loss_clip": 1.05010545, + "balance_loss_mlp": 1.0167098, + "epoch": 0.9746888715204713, + "flos": 29460252332160.0, + "grad_norm": 3.546433103103256, + "language_loss": 0.72716147, + "learning_rate": 6.686206250120729e-09, + "loss": 0.74905944, + "num_input_tokens_seen": 175072330, + "step": 8106, + "time_per_iteration": 2.7213168144226074 + }, + { + "auxiliary_loss_clip": 0.01098057, + "auxiliary_loss_mlp": 0.01026746, + "balance_loss_clip": 0.93035746, + "balance_loss_mlp": 1.01946545, + "epoch": 0.9748091144111104, + "flos": 18479308510080.0, + "grad_norm": 2.588659037465235, + "language_loss": 0.74695772, + "learning_rate": 6.622714793210749e-09, + "loss": 0.76820576, + "num_input_tokens_seen": 175091250, + "step": 8107, + "time_per_iteration": 3.625272750854492 + }, + { + "auxiliary_loss_clip": 0.0116248, + "auxiliary_loss_mlp": 0.01025301, + "balance_loss_clip": 1.04623747, + "balance_loss_mlp": 1.01846766, + "epoch": 0.9749293573017496, + "flos": 20665154753280.0, + "grad_norm": 1.6584757369699825, + "language_loss": 0.78942823, + "learning_rate": 6.559525730050364e-09, + "loss": 0.81130606, + "num_input_tokens_seen": 175111350, + "step": 8108, + "time_per_iteration": 2.53672456741333 + }, + { + "auxiliary_loss_clip": 0.0109895, + "auxiliary_loss_mlp": 0.01026227, + "balance_loss_clip": 0.93517953, + "balance_loss_mlp": 1.01952755, + "epoch": 0.9750496001923886, + "flos": 18478590238080.0, + "grad_norm": 1.858421572653249, + "language_loss": 0.76284981, + "learning_rate": 6.496639070224574e-09, + "loss": 0.78410161, + "num_input_tokens_seen": 175129835, + "step": 8109, + "time_per_iteration": 2.664712905883789 + }, + { + "auxiliary_loss_clip": 0.0114209, + "auxiliary_loss_mlp": 0.01019615, + "balance_loss_clip": 1.01017725, + "balance_loss_mlp": 1.01299882, + "epoch": 0.9751698430830277, + "flos": 19572967860480.0, + "grad_norm": 2.22916922589322, + "language_loss": 0.8352133, + "learning_rate": 6.4340548232739714e-09, + "loss": 0.85683036, + "num_input_tokens_seen": 175146035, + "step": 8110, + "time_per_iteration": 2.6228930950164795 + }, + { + "auxiliary_loss_clip": 0.0109845, + "auxiliary_loss_mlp": 0.01031327, + "balance_loss_clip": 0.93198097, + "balance_loss_mlp": 1.02431154, + "epoch": 0.9752900859736668, + "flos": 23550325862400.0, + "grad_norm": 2.1721865171744166, + "language_loss": 0.79314327, + "learning_rate": 6.371772998692071e-09, + "loss": 0.81444108, + "num_input_tokens_seen": 175165290, + "step": 8111, + "time_per_iteration": 2.7665717601776123 + }, + { + "auxiliary_loss_clip": 0.01097077, + "auxiliary_loss_mlp": 0.0103057, + "balance_loss_clip": 0.93060362, + "balance_loss_mlp": 1.02374816, + "epoch": 0.9754103288643059, + "flos": 20303211358080.0, + "grad_norm": 9.072598603651166, + "language_loss": 0.64938688, + "learning_rate": 6.309793605927094e-09, + "loss": 0.67066336, + "num_input_tokens_seen": 175183610, + "step": 8112, + "time_per_iteration": 2.677058219909668 + }, + { + "auxiliary_loss_clip": 0.0112443, + "auxiliary_loss_mlp": 0.01023054, + "balance_loss_clip": 0.97190082, + "balance_loss_mlp": 1.01593983, + "epoch": 0.975530571754945, + "flos": 19350680544000.0, + "grad_norm": 2.0766427846521185, + "language_loss": 0.80198592, + "learning_rate": 6.248116654381297e-09, + "loss": 0.8234607, + "num_input_tokens_seen": 175202080, + "step": 8113, + "time_per_iteration": 2.825495719909668 + }, + { + "auxiliary_loss_clip": 0.0112006, + "auxiliary_loss_mlp": 0.01022751, + "balance_loss_clip": 0.96907908, + "balance_loss_mlp": 1.0161643, + "epoch": 0.9756508146455841, + "flos": 23583399310080.0, + "grad_norm": 2.0133968576918093, + "language_loss": 0.72746885, + "learning_rate": 6.186742153410751e-09, + "loss": 0.74889696, + "num_input_tokens_seen": 175221575, + "step": 8114, + "time_per_iteration": 2.673204183578491 + }, + { + "auxiliary_loss_clip": 0.01114963, + "auxiliary_loss_mlp": 0.01028388, + "balance_loss_clip": 0.97096497, + "balance_loss_mlp": 1.02098501, + "epoch": 0.9757710575362232, + "flos": 22966921163520.0, + "grad_norm": 3.0256958629035813, + "language_loss": 0.87487698, + "learning_rate": 6.125670112326453e-09, + "loss": 0.89631057, + "num_input_tokens_seen": 175240835, + "step": 8115, + "time_per_iteration": 2.816563129425049 + }, + { + "auxiliary_loss_clip": 0.01141416, + "auxiliary_loss_mlp": 0.01028295, + "balance_loss_clip": 1.00791192, + "balance_loss_mlp": 1.02119863, + "epoch": 0.9758913004268622, + "flos": 27966009530880.0, + "grad_norm": 1.5149367465595012, + "language_loss": 0.70303971, + "learning_rate": 6.064900540392548e-09, + "loss": 0.72473681, + "num_input_tokens_seen": 175262930, + "step": 8116, + "time_per_iteration": 2.8060476779937744 + }, + { + "auxiliary_loss_clip": 0.01114658, + "auxiliary_loss_mlp": 0.01021955, + "balance_loss_clip": 0.97248012, + "balance_loss_mlp": 1.0151037, + "epoch": 0.9760115433175014, + "flos": 22200156512640.0, + "grad_norm": 2.088568839070699, + "language_loss": 0.78711939, + "learning_rate": 6.0044334468278835e-09, + "loss": 0.80848557, + "num_input_tokens_seen": 175282275, + "step": 8117, + "time_per_iteration": 2.6889548301696777 + }, + { + "auxiliary_loss_clip": 0.01072694, + "auxiliary_loss_mlp": 0.01024147, + "balance_loss_clip": 0.894611, + "balance_loss_mlp": 1.0170126, + "epoch": 0.9761317862081405, + "flos": 26250736389120.0, + "grad_norm": 2.6245815854393904, + "language_loss": 0.71417046, + "learning_rate": 5.944268840805345e-09, + "loss": 0.73513883, + "num_input_tokens_seen": 175303020, + "step": 8118, + "time_per_iteration": 2.866373300552368 + }, + { + "auxiliary_loss_clip": 0.01091554, + "auxiliary_loss_mlp": 0.01022437, + "balance_loss_clip": 0.93346238, + "balance_loss_mlp": 1.01581216, + "epoch": 0.9762520290987795, + "flos": 26575440359040.0, + "grad_norm": 2.1258290624222664, + "language_loss": 0.64274645, + "learning_rate": 5.88440673145163e-09, + "loss": 0.66388631, + "num_input_tokens_seen": 175324070, + "step": 8119, + "time_per_iteration": 2.7497754096984863 + }, + { + "auxiliary_loss_clip": 0.01138721, + "auxiliary_loss_mlp": 0.01024742, + "balance_loss_clip": 1.01133609, + "balance_loss_mlp": 1.0181793, + "epoch": 0.9763722719894187, + "flos": 18005036307840.0, + "grad_norm": 2.175241420907228, + "language_loss": 0.82398748, + "learning_rate": 5.824847127848142e-09, + "loss": 0.84562206, + "num_input_tokens_seen": 175342595, + "step": 8120, + "time_per_iteration": 2.654815912246704 + }, + { + "auxiliary_loss_clip": 0.0107516, + "auxiliary_loss_mlp": 0.01026525, + "balance_loss_clip": 0.89532053, + "balance_loss_mlp": 1.01954818, + "epoch": 0.9764925148800577, + "flos": 22455660931200.0, + "grad_norm": 1.9421692930572274, + "language_loss": 0.78563857, + "learning_rate": 5.765590039029433e-09, + "loss": 0.80665541, + "num_input_tokens_seen": 175361915, + "step": 8121, + "time_per_iteration": 2.7153351306915283 + }, + { + "auxiliary_loss_clip": 0.0116263, + "auxiliary_loss_mlp": 0.01029828, + "balance_loss_clip": 1.04775882, + "balance_loss_mlp": 1.02254486, + "epoch": 0.9766127577706968, + "flos": 36757084786560.0, + "grad_norm": 3.9577957972815243, + "language_loss": 0.71292758, + "learning_rate": 5.706635473985422e-09, + "loss": 0.73485219, + "num_input_tokens_seen": 175385785, + "step": 8122, + "time_per_iteration": 3.6095869541168213 + }, + { + "auxiliary_loss_clip": 0.01137773, + "auxiliary_loss_mlp": 0.01024207, + "balance_loss_clip": 1.00850403, + "balance_loss_mlp": 1.01724207, + "epoch": 0.976733000661336, + "flos": 22309971367680.0, + "grad_norm": 5.282415533343532, + "language_loss": 0.8518694, + "learning_rate": 5.6479834416591764e-09, + "loss": 0.8734892, + "num_input_tokens_seen": 175405145, + "step": 8123, + "time_per_iteration": 2.648430585861206 + }, + { + "auxiliary_loss_clip": 0.01139013, + "auxiliary_loss_mlp": 0.00509722, + "balance_loss_clip": 1.01006055, + "balance_loss_mlp": 0.0, + "epoch": 0.976853243551975, + "flos": 25810938264960.0, + "grad_norm": 1.8461534848269612, + "language_loss": 0.68534923, + "learning_rate": 5.589633950947803e-09, + "loss": 0.70183659, + "num_input_tokens_seen": 175422645, + "step": 8124, + "time_per_iteration": 3.5914242267608643 + }, + { + "auxiliary_loss_clip": 0.01115705, + "auxiliary_loss_mlp": 0.01030946, + "balance_loss_clip": 0.97242171, + "balance_loss_mlp": 1.02324522, + "epoch": 0.9769734864426141, + "flos": 21397445326080.0, + "grad_norm": 10.488470129273562, + "language_loss": 0.6991086, + "learning_rate": 5.5315870107035535e-09, + "loss": 0.72057515, + "num_input_tokens_seen": 175440695, + "step": 8125, + "time_per_iteration": 3.6219749450683594 + }, + { + "auxiliary_loss_clip": 0.01116503, + "auxiliary_loss_mlp": 0.01026795, + "balance_loss_clip": 0.97556478, + "balance_loss_mlp": 1.01958299, + "epoch": 0.9770937293332532, + "flos": 13990977584640.0, + "grad_norm": 2.431316683493338, + "language_loss": 0.7869966, + "learning_rate": 5.473842629731607e-09, + "loss": 0.8084296, + "num_input_tokens_seen": 175459195, + "step": 8126, + "time_per_iteration": 2.6789002418518066 + }, + { + "auxiliary_loss_clip": 0.0112486, + "auxiliary_loss_mlp": 0.00509686, + "balance_loss_clip": 0.97145939, + "balance_loss_mlp": 0.0, + "epoch": 0.9772139722238923, + "flos": 17931994001280.0, + "grad_norm": 1.9789569462300918, + "language_loss": 0.7801249, + "learning_rate": 5.416400816792066e-09, + "loss": 0.7964704, + "num_input_tokens_seen": 175476710, + "step": 8127, + "time_per_iteration": 2.6250717639923096 + }, + { + "auxiliary_loss_clip": 0.01161414, + "auxiliary_loss_mlp": 0.01025411, + "balance_loss_clip": 1.04645634, + "balance_loss_mlp": 1.01829481, + "epoch": 0.9773342151145313, + "flos": 20446171488000.0, + "grad_norm": 2.3278068169074806, + "language_loss": 0.78523731, + "learning_rate": 5.359261580598407e-09, + "loss": 0.80710554, + "num_input_tokens_seen": 175492550, + "step": 8128, + "time_per_iteration": 2.6045730113983154 + }, + { + "auxiliary_loss_clip": 0.01144761, + "auxiliary_loss_mlp": 0.01027146, + "balance_loss_clip": 1.01249433, + "balance_loss_mlp": 1.0194993, + "epoch": 0.9774544580051704, + "flos": 11837306949120.0, + "grad_norm": 2.553252147227442, + "language_loss": 0.77815747, + "learning_rate": 5.302424929819027e-09, + "loss": 0.79987657, + "num_input_tokens_seen": 175506560, + "step": 8129, + "time_per_iteration": 2.5894832611083984 + }, + { + "auxiliary_loss_clip": 0.01141447, + "auxiliary_loss_mlp": 0.0102212, + "balance_loss_clip": 1.00623548, + "balance_loss_mlp": 1.01499176, + "epoch": 0.9775747008958096, + "flos": 13479932833920.0, + "grad_norm": 2.4583350692985206, + "language_loss": 0.7305519, + "learning_rate": 5.24589087307592e-09, + "loss": 0.75218761, + "num_input_tokens_seen": 175524180, + "step": 8130, + "time_per_iteration": 2.6217570304870605 + }, + { + "auxiliary_loss_clip": 0.01164802, + "auxiliary_loss_mlp": 0.01023977, + "balance_loss_clip": 1.04849184, + "balance_loss_mlp": 1.01710105, + "epoch": 0.9776949437864486, + "flos": 59532314042880.0, + "grad_norm": 1.843824663203002, + "language_loss": 0.65384257, + "learning_rate": 5.189659418944891e-09, + "loss": 0.67573035, + "num_input_tokens_seen": 175554355, + "step": 8131, + "time_per_iteration": 2.987427234649658 + }, + { + "auxiliary_loss_clip": 0.01163032, + "auxiliary_loss_mlp": 0.01020763, + "balance_loss_clip": 1.0479486, + "balance_loss_mlp": 1.01373911, + "epoch": 0.9778151866770877, + "flos": 21178605715200.0, + "grad_norm": 1.7802582759794334, + "language_loss": 0.78589427, + "learning_rate": 5.133730575956674e-09, + "loss": 0.80773222, + "num_input_tokens_seen": 175574025, + "step": 8132, + "time_per_iteration": 2.6950454711914062 + }, + { + "auxiliary_loss_clip": 0.01118484, + "auxiliary_loss_mlp": 0.01027857, + "balance_loss_clip": 0.97073299, + "balance_loss_mlp": 1.02071619, + "epoch": 0.9779354295677268, + "flos": 20886795624960.0, + "grad_norm": 2.386689472711446, + "language_loss": 0.72134399, + "learning_rate": 5.0781043525953696e-09, + "loss": 0.74280739, + "num_input_tokens_seen": 175592090, + "step": 8133, + "time_per_iteration": 3.6284730434417725 + }, + { + "auxiliary_loss_clip": 0.01112381, + "auxiliary_loss_mlp": 0.01022316, + "balance_loss_clip": 0.97244376, + "balance_loss_mlp": 1.01550293, + "epoch": 0.9780556724583659, + "flos": 23440618748160.0, + "grad_norm": 1.645193551270244, + "language_loss": 0.73611617, + "learning_rate": 5.0227807572995605e-09, + "loss": 0.75746316, + "num_input_tokens_seen": 175614065, + "step": 8134, + "time_per_iteration": 2.7392687797546387 + }, + { + "auxiliary_loss_clip": 0.01118554, + "auxiliary_loss_mlp": 0.01028695, + "balance_loss_clip": 0.97042346, + "balance_loss_mlp": 1.02190888, + "epoch": 0.9781759153490049, + "flos": 20923244951040.0, + "grad_norm": 2.126559948128804, + "language_loss": 0.67457128, + "learning_rate": 4.967759798461646e-09, + "loss": 0.69604379, + "num_input_tokens_seen": 175632410, + "step": 8135, + "time_per_iteration": 2.708028554916382 + }, + { + "auxiliary_loss_clip": 0.01162048, + "auxiliary_loss_mlp": 0.01029, + "balance_loss_clip": 1.04882002, + "balance_loss_mlp": 1.02261972, + "epoch": 0.9782961582396441, + "flos": 28293191539200.0, + "grad_norm": 2.7871578363681655, + "language_loss": 0.74947453, + "learning_rate": 4.913041484428282e-09, + "loss": 0.77138507, + "num_input_tokens_seen": 175652885, + "step": 8136, + "time_per_iteration": 2.6232972145080566 + }, + { + "auxiliary_loss_clip": 0.01145093, + "auxiliary_loss_mlp": 0.01025144, + "balance_loss_clip": 1.01183677, + "balance_loss_mlp": 1.01857841, + "epoch": 0.9784164011302832, + "flos": 25552955808000.0, + "grad_norm": 1.9177791096297618, + "language_loss": 0.74060529, + "learning_rate": 4.858625823500384e-09, + "loss": 0.76230764, + "num_input_tokens_seen": 175670585, + "step": 8137, + "time_per_iteration": 2.6795125007629395 + }, + { + "auxiliary_loss_clip": 0.01143524, + "auxiliary_loss_mlp": 0.01025715, + "balance_loss_clip": 1.00939584, + "balance_loss_mlp": 1.01836514, + "epoch": 0.9785366440209222, + "flos": 29965945956480.0, + "grad_norm": 2.1855673264057778, + "language_loss": 0.73456764, + "learning_rate": 4.80451282393246e-09, + "loss": 0.75626004, + "num_input_tokens_seen": 175690570, + "step": 8138, + "time_per_iteration": 2.6715126037597656 + }, + { + "auxiliary_loss_clip": 0.0112, + "auxiliary_loss_mlp": 0.01024983, + "balance_loss_clip": 0.97146428, + "balance_loss_mlp": 1.01774073, + "epoch": 0.9786568869115614, + "flos": 32343591847680.0, + "grad_norm": 3.9272573897887524, + "language_loss": 0.67468363, + "learning_rate": 4.750702493933722e-09, + "loss": 0.69613343, + "num_input_tokens_seen": 175710455, + "step": 8139, + "time_per_iteration": 2.749297857284546 + }, + { + "auxiliary_loss_clip": 0.01118304, + "auxiliary_loss_mlp": 0.00509312, + "balance_loss_clip": 0.97397792, + "balance_loss_mlp": 0.0, + "epoch": 0.9787771298022004, + "flos": 23331414424320.0, + "grad_norm": 1.936065064822089, + "language_loss": 0.85401154, + "learning_rate": 4.697194841666974e-09, + "loss": 0.87028766, + "num_input_tokens_seen": 175729380, + "step": 8140, + "time_per_iteration": 2.676236391067505 + }, + { + "auxiliary_loss_clip": 0.01141367, + "auxiliary_loss_mlp": 0.01029492, + "balance_loss_clip": 1.00856256, + "balance_loss_mlp": 1.02219057, + "epoch": 0.9788973726928395, + "flos": 21468548298240.0, + "grad_norm": 2.117822845463113, + "language_loss": 0.81881964, + "learning_rate": 4.6439898752492764e-09, + "loss": 0.84052825, + "num_input_tokens_seen": 175749520, + "step": 8141, + "time_per_iteration": 2.7660627365112305 + }, + { + "auxiliary_loss_clip": 0.01040823, + "auxiliary_loss_mlp": 0.00500652, + "balance_loss_clip": 0.97217995, + "balance_loss_mlp": 0.0, + "epoch": 0.9790176155834787, + "flos": 68897459439360.0, + "grad_norm": 0.7478422455140182, + "language_loss": 0.63621038, + "learning_rate": 4.591087602751731e-09, + "loss": 0.6516251, + "num_input_tokens_seen": 175811380, + "step": 8142, + "time_per_iteration": 3.267573595046997 + }, + { + "auxiliary_loss_clip": 0.01141512, + "auxiliary_loss_mlp": 0.01022521, + "balance_loss_clip": 1.00965571, + "balance_loss_mlp": 1.01593804, + "epoch": 0.9791378584741177, + "flos": 21430877909760.0, + "grad_norm": 1.672413914022118, + "language_loss": 0.71925628, + "learning_rate": 4.538488032199916e-09, + "loss": 0.74089664, + "num_input_tokens_seen": 175829480, + "step": 8143, + "time_per_iteration": 2.627002000808716 + }, + { + "auxiliary_loss_clip": 0.01142976, + "auxiliary_loss_mlp": 0.01022261, + "balance_loss_clip": 1.0073812, + "balance_loss_mlp": 1.01489997, + "epoch": 0.9792581013647568, + "flos": 20153032594560.0, + "grad_norm": 3.58138808070312, + "language_loss": 0.69350874, + "learning_rate": 4.486191171572784e-09, + "loss": 0.71516114, + "num_input_tokens_seen": 175846750, + "step": 8144, + "time_per_iteration": 2.6180124282836914 + }, + { + "auxiliary_loss_clip": 0.01145311, + "auxiliary_loss_mlp": 0.01021989, + "balance_loss_clip": 1.01134729, + "balance_loss_mlp": 1.01515484, + "epoch": 0.9793783442553959, + "flos": 23728191033600.0, + "grad_norm": 1.469344310114612, + "language_loss": 0.77511072, + "learning_rate": 4.434197028803766e-09, + "loss": 0.79678369, + "num_input_tokens_seen": 175865975, + "step": 8145, + "time_per_iteration": 2.6827445030212402 + }, + { + "auxiliary_loss_clip": 0.01101443, + "auxiliary_loss_mlp": 0.01023406, + "balance_loss_clip": 0.9335714, + "balance_loss_mlp": 1.01598811, + "epoch": 0.979498587146035, + "flos": 23038742407680.0, + "grad_norm": 2.055295828396352, + "language_loss": 0.82008398, + "learning_rate": 4.3825056117805514e-09, + "loss": 0.84133244, + "num_input_tokens_seen": 175881860, + "step": 8146, + "time_per_iteration": 2.6940999031066895 + }, + { + "auxiliary_loss_clip": 0.01161829, + "auxiliary_loss_mlp": 0.01022869, + "balance_loss_clip": 1.04554892, + "balance_loss_mlp": 1.01622629, + "epoch": 0.979618830036674, + "flos": 14318841951360.0, + "grad_norm": 3.3386964996490494, + "language_loss": 0.79005134, + "learning_rate": 4.331116928344425e-09, + "loss": 0.81189829, + "num_input_tokens_seen": 175898175, + "step": 8147, + "time_per_iteration": 2.5862014293670654 + }, + { + "auxiliary_loss_clip": 0.01122304, + "auxiliary_loss_mlp": 0.00509463, + "balance_loss_clip": 0.97062379, + "balance_loss_mlp": 0.0, + "epoch": 0.9797390729273132, + "flos": 16727514215040.0, + "grad_norm": 2.183627561786364, + "language_loss": 0.62637067, + "learning_rate": 4.28003098629115e-09, + "loss": 0.64268833, + "num_input_tokens_seen": 175914310, + "step": 8148, + "time_per_iteration": 3.541095495223999 + }, + { + "auxiliary_loss_clip": 0.01089919, + "auxiliary_loss_mlp": 0.01023042, + "balance_loss_clip": 0.9274798, + "balance_loss_mlp": 1.01538587, + "epoch": 0.9798593158179523, + "flos": 24532661986560.0, + "grad_norm": 58.394725319151874, + "language_loss": 0.78424704, + "learning_rate": 4.229247793370305e-09, + "loss": 0.80537665, + "num_input_tokens_seen": 175933435, + "step": 8149, + "time_per_iteration": 3.68022084236145 + }, + { + "auxiliary_loss_clip": 0.01167206, + "auxiliary_loss_mlp": 0.01028044, + "balance_loss_clip": 1.04987359, + "balance_loss_mlp": 1.02046871, + "epoch": 0.9799795587085913, + "flos": 27308808339840.0, + "grad_norm": 1.579572539646454, + "language_loss": 0.70312148, + "learning_rate": 4.178767357285951e-09, + "loss": 0.72507399, + "num_input_tokens_seen": 175955065, + "step": 8150, + "time_per_iteration": 2.637507200241089 + }, + { + "auxiliary_loss_clip": 0.01144074, + "auxiliary_loss_mlp": 0.00509389, + "balance_loss_clip": 1.01155376, + "balance_loss_mlp": 0.0, + "epoch": 0.9800998015992305, + "flos": 26286575184000.0, + "grad_norm": 1.8942890244859376, + "language_loss": 0.7156353, + "learning_rate": 4.128589685695516e-09, + "loss": 0.73216993, + "num_input_tokens_seen": 175975490, + "step": 8151, + "time_per_iteration": 3.6276211738586426 + }, + { + "auxiliary_loss_clip": 0.01162792, + "auxiliary_loss_mlp": 0.01024331, + "balance_loss_clip": 1.04771769, + "balance_loss_mlp": 1.0169698, + "epoch": 0.9802200444898695, + "flos": 16723635546240.0, + "grad_norm": 1.8561897207966052, + "language_loss": 0.84297919, + "learning_rate": 4.078714786211135e-09, + "loss": 0.8648504, + "num_input_tokens_seen": 175991340, + "step": 8152, + "time_per_iteration": 2.5255656242370605 + }, + { + "auxiliary_loss_clip": 0.01139744, + "auxiliary_loss_mlp": 0.01019769, + "balance_loss_clip": 1.01046753, + "balance_loss_mlp": 1.01379085, + "epoch": 0.9803402873805086, + "flos": 24900459298560.0, + "grad_norm": 1.765859283895054, + "language_loss": 0.76430863, + "learning_rate": 4.029142666398977e-09, + "loss": 0.78590375, + "num_input_tokens_seen": 176011505, + "step": 8153, + "time_per_iteration": 2.7376487255096436 + }, + { + "auxiliary_loss_clip": 0.01160581, + "auxiliary_loss_mlp": 0.01029884, + "balance_loss_clip": 1.04707897, + "balance_loss_mlp": 1.02309453, + "epoch": 0.9804605302711478, + "flos": 22564937082240.0, + "grad_norm": 1.9056585360177078, + "language_loss": 0.80180609, + "learning_rate": 3.979873333778805e-09, + "loss": 0.8237108, + "num_input_tokens_seen": 176029680, + "step": 8154, + "time_per_iteration": 2.638505697250366 + }, + { + "auxiliary_loss_clip": 0.011224, + "auxiliary_loss_mlp": 0.0102434, + "balance_loss_clip": 0.97278786, + "balance_loss_mlp": 1.01761353, + "epoch": 0.9805807731617868, + "flos": 38905368382080.0, + "grad_norm": 1.8044184234165404, + "language_loss": 0.73949617, + "learning_rate": 3.930906795824862e-09, + "loss": 0.76096356, + "num_input_tokens_seen": 176050355, + "step": 8155, + "time_per_iteration": 2.8224918842315674 + }, + { + "auxiliary_loss_clip": 0.01138352, + "auxiliary_loss_mlp": 0.01025989, + "balance_loss_clip": 1.00861979, + "balance_loss_mlp": 1.01906872, + "epoch": 0.9807010160524259, + "flos": 17821999578240.0, + "grad_norm": 2.1050929104476976, + "language_loss": 0.76416773, + "learning_rate": 3.882243059965207e-09, + "loss": 0.78581113, + "num_input_tokens_seen": 176068070, + "step": 8156, + "time_per_iteration": 2.6385743618011475 + }, + { + "auxiliary_loss_clip": 0.01133022, + "auxiliary_loss_mlp": 0.01029084, + "balance_loss_clip": 1.00630021, + "balance_loss_mlp": 1.02126956, + "epoch": 0.980821258943065, + "flos": 13552975140480.0, + "grad_norm": 2.0275338605068294, + "language_loss": 0.65599263, + "learning_rate": 3.833882133582156e-09, + "loss": 0.67761374, + "num_input_tokens_seen": 176083730, + "step": 8157, + "time_per_iteration": 2.591156005859375 + }, + { + "auxiliary_loss_clip": 0.0114322, + "auxiliary_loss_mlp": 0.01023773, + "balance_loss_clip": 1.01025558, + "balance_loss_mlp": 1.01709414, + "epoch": 0.9809415018337041, + "flos": 21689794120320.0, + "grad_norm": 2.100081353379986, + "language_loss": 0.78291857, + "learning_rate": 3.785824024012285e-09, + "loss": 0.80458844, + "num_input_tokens_seen": 176102730, + "step": 8158, + "time_per_iteration": 2.613525867462158 + }, + { + "auxiliary_loss_clip": 0.01112791, + "auxiliary_loss_mlp": 0.01030002, + "balance_loss_clip": 0.97355378, + "balance_loss_mlp": 1.02325225, + "epoch": 0.9810617447243432, + "flos": 23294857357440.0, + "grad_norm": 1.6149987548458413, + "language_loss": 0.78420371, + "learning_rate": 3.738068738545541e-09, + "loss": 0.80563164, + "num_input_tokens_seen": 176121815, + "step": 8159, + "time_per_iteration": 3.5283448696136475 + }, + { + "auxiliary_loss_clip": 0.0114295, + "auxiliary_loss_mlp": 0.01027884, + "balance_loss_clip": 1.00793338, + "balance_loss_mlp": 1.0211401, + "epoch": 0.9811819876149822, + "flos": 18332038748160.0, + "grad_norm": 2.7186178408171524, + "language_loss": 0.78786683, + "learning_rate": 3.6906162844265733e-09, + "loss": 0.8095752, + "num_input_tokens_seen": 176138900, + "step": 8160, + "time_per_iteration": 2.605321168899536 + }, + { + "auxiliary_loss_clip": 0.01112909, + "auxiliary_loss_mlp": 0.01026467, + "balance_loss_clip": 0.97028822, + "balance_loss_mlp": 1.01877165, + "epoch": 0.9813022305056214, + "flos": 22601961025920.0, + "grad_norm": 2.1849185879426805, + "language_loss": 0.70536363, + "learning_rate": 3.643466668853845e-09, + "loss": 0.72675741, + "num_input_tokens_seen": 176156925, + "step": 8161, + "time_per_iteration": 2.6156959533691406 + }, + { + "auxiliary_loss_clip": 0.01117619, + "auxiliary_loss_mlp": 0.01026164, + "balance_loss_clip": 0.96947849, + "balance_loss_mlp": 1.0189755, + "epoch": 0.9814224733962604, + "flos": 25413335642880.0, + "grad_norm": 2.34963862846635, + "language_loss": 0.75021464, + "learning_rate": 3.59661989898008e-09, + "loss": 0.77165246, + "num_input_tokens_seen": 176177980, + "step": 8162, + "time_per_iteration": 2.7644453048706055 + }, + { + "auxiliary_loss_clip": 0.01086217, + "auxiliary_loss_mlp": 0.0102328, + "balance_loss_clip": 0.9322626, + "balance_loss_mlp": 1.01614499, + "epoch": 0.9815427162868995, + "flos": 25007185584000.0, + "grad_norm": 2.678142760132396, + "language_loss": 0.76902682, + "learning_rate": 3.5500759819115934e-09, + "loss": 0.79012179, + "num_input_tokens_seen": 176198345, + "step": 8163, + "time_per_iteration": 2.698538303375244 + }, + { + "auxiliary_loss_clip": 0.01168677, + "auxiliary_loss_mlp": 0.01026153, + "balance_loss_clip": 1.05272484, + "balance_loss_mlp": 1.01919377, + "epoch": 0.9816629591775387, + "flos": 20662604887680.0, + "grad_norm": 1.9161136476012732, + "language_loss": 0.80901933, + "learning_rate": 3.5038349247094034e-09, + "loss": 0.83096766, + "num_input_tokens_seen": 176215605, + "step": 8164, + "time_per_iteration": 2.6022934913635254 + }, + { + "auxiliary_loss_clip": 0.01118883, + "auxiliary_loss_mlp": 0.01023618, + "balance_loss_clip": 0.97113419, + "balance_loss_mlp": 1.01631689, + "epoch": 0.9817832020681777, + "flos": 17712220636800.0, + "grad_norm": 2.274372538247917, + "language_loss": 0.77536654, + "learning_rate": 3.4578967343878994e-09, + "loss": 0.79679155, + "num_input_tokens_seen": 176231810, + "step": 8165, + "time_per_iteration": 2.8030776977539062 + }, + { + "auxiliary_loss_clip": 0.01117148, + "auxiliary_loss_mlp": 0.01029155, + "balance_loss_clip": 0.97189581, + "balance_loss_mlp": 1.02221394, + "epoch": 0.9819034449588168, + "flos": 22530032040960.0, + "grad_norm": 1.8407805836542042, + "language_loss": 0.80523717, + "learning_rate": 3.4122614179161733e-09, + "loss": 0.82670021, + "num_input_tokens_seen": 176251770, + "step": 8166, + "time_per_iteration": 2.7027010917663574 + }, + { + "auxiliary_loss_clip": 0.01084022, + "auxiliary_loss_mlp": 0.01024335, + "balance_loss_clip": 0.93122876, + "balance_loss_mlp": 1.01729298, + "epoch": 0.9820236878494559, + "flos": 20011221699840.0, + "grad_norm": 1.638502552032811, + "language_loss": 0.78530598, + "learning_rate": 3.36692898221691e-09, + "loss": 0.80638957, + "num_input_tokens_seen": 176270135, + "step": 8167, + "time_per_iteration": 2.8714652061462402 + }, + { + "auxiliary_loss_clip": 0.01140376, + "auxiliary_loss_mlp": 0.01026788, + "balance_loss_clip": 1.01029587, + "balance_loss_mlp": 1.02014518, + "epoch": 0.982143930740095, + "flos": 18807316531200.0, + "grad_norm": 1.6469128718799213, + "language_loss": 0.73582023, + "learning_rate": 3.3218994341668305e-09, + "loss": 0.75749183, + "num_input_tokens_seen": 176289065, + "step": 8168, + "time_per_iteration": 2.6109092235565186 + }, + { + "auxiliary_loss_clip": 0.01163106, + "auxiliary_loss_mlp": 0.01022448, + "balance_loss_clip": 1.04955471, + "balance_loss_mlp": 1.01556933, + "epoch": 0.982264173630734, + "flos": 26578026138240.0, + "grad_norm": 1.586065624597333, + "language_loss": 0.75500602, + "learning_rate": 3.2771727805971373e-09, + "loss": 0.77686155, + "num_input_tokens_seen": 176310450, + "step": 8169, + "time_per_iteration": 2.648606061935425 + }, + { + "auxiliary_loss_clip": 0.01062039, + "auxiliary_loss_mlp": 0.01024048, + "balance_loss_clip": 0.89001334, + "balance_loss_mlp": 1.01658535, + "epoch": 0.9823844165213732, + "flos": 22014462176640.0, + "grad_norm": 1.676117308879268, + "language_loss": 0.76905203, + "learning_rate": 3.232749028292847e-09, + "loss": 0.78991282, + "num_input_tokens_seen": 176327415, + "step": 8170, + "time_per_iteration": 2.697004556655884 + }, + { + "auxiliary_loss_clip": 0.01164485, + "auxiliary_loss_mlp": 0.01025052, + "balance_loss_clip": 1.04795587, + "balance_loss_mlp": 1.01806664, + "epoch": 0.9825046594120123, + "flos": 21908166854400.0, + "grad_norm": 4.304609647517901, + "language_loss": 0.88094771, + "learning_rate": 3.188628183992792e-09, + "loss": 0.90284312, + "num_input_tokens_seen": 176347680, + "step": 8171, + "time_per_iteration": 2.626643657684326 + }, + { + "auxiliary_loss_clip": 0.01041206, + "auxiliary_loss_mlp": 0.00998811, + "balance_loss_clip": 0.97160649, + "balance_loss_mlp": 0.99786913, + "epoch": 0.9826249023026513, + "flos": 59494610718720.0, + "grad_norm": 0.7396769717856599, + "language_loss": 0.62529224, + "learning_rate": 3.1448102543902844e-09, + "loss": 0.64569241, + "num_input_tokens_seen": 176411595, + "step": 8172, + "time_per_iteration": 3.2038631439208984 + }, + { + "auxiliary_loss_clip": 0.01112757, + "auxiliary_loss_mlp": 0.01020609, + "balance_loss_clip": 0.97113943, + "balance_loss_mlp": 1.01381409, + "epoch": 0.9827451451932905, + "flos": 16071031296000.0, + "grad_norm": 3.116656170200565, + "language_loss": 0.67602837, + "learning_rate": 3.1012952461324515e-09, + "loss": 0.69736207, + "num_input_tokens_seen": 176430570, + "step": 8173, + "time_per_iteration": 2.6909000873565674 + }, + { + "auxiliary_loss_clip": 0.01138449, + "auxiliary_loss_mlp": 0.01024775, + "balance_loss_clip": 1.01288414, + "balance_loss_mlp": 1.01839101, + "epoch": 0.9828653880839295, + "flos": 20262775622400.0, + "grad_norm": 2.3256378326048384, + "language_loss": 0.73489588, + "learning_rate": 3.0580831658204575e-09, + "loss": 0.75652814, + "num_input_tokens_seen": 176448150, + "step": 8174, + "time_per_iteration": 3.623889446258545 + }, + { + "auxiliary_loss_clip": 0.01141061, + "auxiliary_loss_mlp": 0.01023206, + "balance_loss_clip": 1.01139438, + "balance_loss_mlp": 1.01651525, + "epoch": 0.9829856309745686, + "flos": 21616141282560.0, + "grad_norm": 2.2167970182095957, + "language_loss": 0.78038937, + "learning_rate": 3.015174020009281e-09, + "loss": 0.80203211, + "num_input_tokens_seen": 176467475, + "step": 8175, + "time_per_iteration": 2.7018723487854004 + }, + { + "auxiliary_loss_clip": 0.01095002, + "auxiliary_loss_mlp": 0.01022993, + "balance_loss_clip": 0.93035352, + "balance_loss_mlp": 1.01621294, + "epoch": 0.9831058738652078, + "flos": 23764209396480.0, + "grad_norm": 1.7356283997439101, + "language_loss": 0.75128222, + "learning_rate": 2.9725678152086043e-09, + "loss": 0.77246213, + "num_input_tokens_seen": 176486045, + "step": 8176, + "time_per_iteration": 3.6466598510742188 + }, + { + "auxiliary_loss_clip": 0.01106694, + "auxiliary_loss_mlp": 0.01025719, + "balance_loss_clip": 0.96943676, + "balance_loss_mlp": 1.01831341, + "epoch": 0.9832261167558468, + "flos": 11320911072000.0, + "grad_norm": 2.561403526790718, + "language_loss": 0.82624543, + "learning_rate": 2.930264557881257e-09, + "loss": 0.84756958, + "num_input_tokens_seen": 176501230, + "step": 8177, + "time_per_iteration": 3.56288743019104 + }, + { + "auxiliary_loss_clip": 0.01059349, + "auxiliary_loss_mlp": 0.00999893, + "balance_loss_clip": 1.00879467, + "balance_loss_mlp": 0.99895728, + "epoch": 0.9833463596464859, + "flos": 60000304343040.0, + "grad_norm": 0.8461902622877245, + "language_loss": 0.58229327, + "learning_rate": 2.8882642544452163e-09, + "loss": 0.6028856, + "num_input_tokens_seen": 176565955, + "step": 8178, + "time_per_iteration": 3.1995439529418945 + }, + { + "auxiliary_loss_clip": 0.01107655, + "auxiliary_loss_mlp": 0.01028314, + "balance_loss_clip": 0.9686327, + "balance_loss_mlp": 1.02134943, + "epoch": 0.983466602537125, + "flos": 13626699805440.0, + "grad_norm": 2.2323997843078858, + "language_loss": 0.74499583, + "learning_rate": 2.8465669112716083e-09, + "loss": 0.76635551, + "num_input_tokens_seen": 176583480, + "step": 8179, + "time_per_iteration": 2.743216037750244 + }, + { + "auxiliary_loss_clip": 0.01141391, + "auxiliary_loss_mlp": 0.00509178, + "balance_loss_clip": 1.00775409, + "balance_loss_mlp": 0.0, + "epoch": 0.9835868454277641, + "flos": 22926844563840.0, + "grad_norm": 1.7693373103270618, + "language_loss": 0.76361644, + "learning_rate": 2.8051725346858177e-09, + "loss": 0.78012216, + "num_input_tokens_seen": 176603740, + "step": 8180, + "time_per_iteration": 2.728748321533203 + }, + { + "auxiliary_loss_clip": 0.01162986, + "auxiliary_loss_mlp": 0.01024344, + "balance_loss_clip": 1.04513669, + "balance_loss_mlp": 1.01775455, + "epoch": 0.9837070883184031, + "flos": 27673409341440.0, + "grad_norm": 2.350019301789073, + "language_loss": 0.70566517, + "learning_rate": 2.7640811309674883e-09, + "loss": 0.72753847, + "num_input_tokens_seen": 176623240, + "step": 8181, + "time_per_iteration": 2.7225208282470703 + }, + { + "auxiliary_loss_clip": 0.01085806, + "auxiliary_loss_mlp": 0.01025659, + "balance_loss_clip": 0.93273532, + "balance_loss_mlp": 1.01861942, + "epoch": 0.9838273312090423, + "flos": 29241951425280.0, + "grad_norm": 2.03948079810593, + "language_loss": 0.80726814, + "learning_rate": 2.7232927063498557e-09, + "loss": 0.82838273, + "num_input_tokens_seen": 176643615, + "step": 8182, + "time_per_iteration": 2.7390923500061035 + }, + { + "auxiliary_loss_clip": 0.01143101, + "auxiliary_loss_mlp": 0.01028141, + "balance_loss_clip": 1.00971973, + "balance_loss_mlp": 1.02096736, + "epoch": 0.9839475740996814, + "flos": 40110207304320.0, + "grad_norm": 1.9361623913539867, + "language_loss": 0.69160789, + "learning_rate": 2.682807267020859e-09, + "loss": 0.71332026, + "num_input_tokens_seen": 176666375, + "step": 8183, + "time_per_iteration": 2.7530641555786133 + }, + { + "auxiliary_loss_clip": 0.01139736, + "auxiliary_loss_mlp": 0.01029981, + "balance_loss_clip": 1.00945449, + "balance_loss_mlp": 1.02298641, + "epoch": 0.9840678169903204, + "flos": 24169389788160.0, + "grad_norm": 1.722622826146328, + "language_loss": 0.62314165, + "learning_rate": 2.642624819121808e-09, + "loss": 0.64483887, + "num_input_tokens_seen": 176686525, + "step": 8184, + "time_per_iteration": 2.642765998840332 + }, + { + "auxiliary_loss_clip": 0.01115606, + "auxiliary_loss_mlp": 0.01022637, + "balance_loss_clip": 0.97281057, + "balance_loss_mlp": 1.01587439, + "epoch": 0.9841880598809596, + "flos": 14684484447360.0, + "grad_norm": 5.049911503600911, + "language_loss": 0.61878169, + "learning_rate": 2.6027453687487154e-09, + "loss": 0.64016414, + "num_input_tokens_seen": 176703615, + "step": 8185, + "time_per_iteration": 3.5236575603485107 + }, + { + "auxiliary_loss_clip": 0.0111815, + "auxiliary_loss_mlp": 0.01023188, + "balance_loss_clip": 0.97075814, + "balance_loss_mlp": 1.01549888, + "epoch": 0.9843083027715986, + "flos": 22344768668160.0, + "grad_norm": 2.2732817852902367, + "language_loss": 0.53748149, + "learning_rate": 2.5631689219509643e-09, + "loss": 0.55889487, + "num_input_tokens_seen": 176722295, + "step": 8186, + "time_per_iteration": 2.6687228679656982 + }, + { + "auxiliary_loss_clip": 0.01118507, + "auxiliary_loss_mlp": 0.01022748, + "balance_loss_clip": 0.97345936, + "balance_loss_mlp": 1.0164566, + "epoch": 0.9844285456622377, + "flos": 21800111765760.0, + "grad_norm": 1.6574196470159714, + "language_loss": 0.83362293, + "learning_rate": 2.523895484732197e-09, + "loss": 0.85503542, + "num_input_tokens_seen": 176741750, + "step": 8187, + "time_per_iteration": 2.656658172607422 + }, + { + "auxiliary_loss_clip": 0.01147594, + "auxiliary_loss_mlp": 0.01024579, + "balance_loss_clip": 1.01061606, + "balance_loss_mlp": 1.01711619, + "epoch": 0.9845487885528769, + "flos": 18035380321920.0, + "grad_norm": 1.8353381875662125, + "language_loss": 0.74719191, + "learning_rate": 2.4849250630505357e-09, + "loss": 0.76891363, + "num_input_tokens_seen": 176759995, + "step": 8188, + "time_per_iteration": 2.6401195526123047 + }, + { + "auxiliary_loss_clip": 0.01012304, + "auxiliary_loss_mlp": 0.01025144, + "balance_loss_clip": 0.81590176, + "balance_loss_mlp": 1.01845002, + "epoch": 0.9846690314435159, + "flos": 25228610974080.0, + "grad_norm": 1.8059891169424904, + "language_loss": 0.73559946, + "learning_rate": 2.4462576628172528e-09, + "loss": 0.75597394, + "num_input_tokens_seen": 176778625, + "step": 8189, + "time_per_iteration": 2.9848246574401855 + }, + { + "auxiliary_loss_clip": 0.01137744, + "auxiliary_loss_mlp": 0.01023081, + "balance_loss_clip": 1.01134467, + "balance_loss_mlp": 1.01622605, + "epoch": 0.984789274334155, + "flos": 18552171248640.0, + "grad_norm": 1.6973086876548185, + "language_loss": 0.74148166, + "learning_rate": 2.407893289898766e-09, + "loss": 0.76308995, + "num_input_tokens_seen": 176797655, + "step": 8190, + "time_per_iteration": 2.7938120365142822 + }, + { + "auxiliary_loss_clip": 0.01084578, + "auxiliary_loss_mlp": 0.01025211, + "balance_loss_clip": 0.92945141, + "balance_loss_mlp": 1.01798999, + "epoch": 0.984909517224794, + "flos": 27345437233920.0, + "grad_norm": 1.8148050303145968, + "language_loss": 0.84065437, + "learning_rate": 2.3698319501144202e-09, + "loss": 0.86175227, + "num_input_tokens_seen": 176818640, + "step": 8191, + "time_per_iteration": 2.719287395477295 + }, + { + "auxiliary_loss_clip": 0.01145126, + "auxiliary_loss_mlp": 0.01021884, + "balance_loss_clip": 1.00856876, + "balance_loss_mlp": 1.01443291, + "epoch": 0.9850297601154332, + "flos": 18734058743040.0, + "grad_norm": 1.6477376389802596, + "language_loss": 0.73301119, + "learning_rate": 2.3320736492382644e-09, + "loss": 0.75468129, + "num_input_tokens_seen": 176837475, + "step": 8192, + "time_per_iteration": 2.5530126094818115 + }, + { + "auxiliary_loss_clip": 0.01159953, + "auxiliary_loss_mlp": 0.01023108, + "balance_loss_clip": 1.04731584, + "balance_loss_mlp": 1.01615238, + "epoch": 0.9851500030060723, + "flos": 22308247514880.0, + "grad_norm": 1.5898240572631488, + "language_loss": 0.68022758, + "learning_rate": 2.29461839299816e-09, + "loss": 0.7020582, + "num_input_tokens_seen": 176857190, + "step": 8193, + "time_per_iteration": 2.659139633178711 + }, + { + "auxiliary_loss_clip": 0.01095579, + "auxiliary_loss_mlp": 0.01021633, + "balance_loss_clip": 0.93270218, + "balance_loss_mlp": 1.01472807, + "epoch": 0.9852702458967113, + "flos": 26353691746560.0, + "grad_norm": 1.7141946652851883, + "language_loss": 0.79887831, + "learning_rate": 2.257466187076229e-09, + "loss": 0.82005042, + "num_input_tokens_seen": 176876395, + "step": 8194, + "time_per_iteration": 2.7953836917877197 + }, + { + "auxiliary_loss_clip": 0.0114449, + "auxiliary_loss_mlp": 0.00509457, + "balance_loss_clip": 1.00921285, + "balance_loss_mlp": 0.0, + "epoch": 0.9853904887873505, + "flos": 20883599314560.0, + "grad_norm": 2.023788068016984, + "language_loss": 0.71587461, + "learning_rate": 2.2206170371081854e-09, + "loss": 0.73241413, + "num_input_tokens_seen": 176894980, + "step": 8195, + "time_per_iteration": 2.6897764205932617 + }, + { + "auxiliary_loss_clip": 0.01116182, + "auxiliary_loss_mlp": 0.01032081, + "balance_loss_clip": 0.96884865, + "balance_loss_mlp": 1.02523243, + "epoch": 0.9855107316779895, + "flos": 25263444188160.0, + "grad_norm": 1.5698363408213012, + "language_loss": 0.84634668, + "learning_rate": 2.1840709486842247e-09, + "loss": 0.86782932, + "num_input_tokens_seen": 176914600, + "step": 8196, + "time_per_iteration": 2.6764912605285645 + }, + { + "auxiliary_loss_clip": 0.01109939, + "auxiliary_loss_mlp": 0.01025196, + "balance_loss_clip": 0.96897978, + "balance_loss_mlp": 1.01819861, + "epoch": 0.9856309745686286, + "flos": 19062102677760.0, + "grad_norm": 2.1017087051506844, + "language_loss": 0.7954663, + "learning_rate": 2.1478279273481335e-09, + "loss": 0.81681764, + "num_input_tokens_seen": 176933085, + "step": 8197, + "time_per_iteration": 2.621297597885132 + }, + { + "auxiliary_loss_clip": 0.01141193, + "auxiliary_loss_mlp": 0.01023481, + "balance_loss_clip": 1.01186574, + "balance_loss_mlp": 1.0169034, + "epoch": 0.9857512174592677, + "flos": 34130758060800.0, + "grad_norm": 1.9620869339812965, + "language_loss": 0.8004986, + "learning_rate": 2.1118879785981815e-09, + "loss": 0.82214534, + "num_input_tokens_seen": 176953225, + "step": 8198, + "time_per_iteration": 2.6965651512145996 + }, + { + "auxiliary_loss_clip": 0.01116649, + "auxiliary_loss_mlp": 0.01024046, + "balance_loss_clip": 0.97072852, + "balance_loss_mlp": 1.01717031, + "epoch": 0.9858714603499068, + "flos": 25994693266560.0, + "grad_norm": 4.099983453360164, + "language_loss": 0.7944966, + "learning_rate": 2.0762511078862288e-09, + "loss": 0.81590354, + "num_input_tokens_seen": 176973570, + "step": 8199, + "time_per_iteration": 2.6939735412597656 + }, + { + "auxiliary_loss_clip": 0.01125271, + "auxiliary_loss_mlp": 0.010243, + "balance_loss_clip": 0.97170788, + "balance_loss_mlp": 1.01730227, + "epoch": 0.9859917032405459, + "flos": 23696230907520.0, + "grad_norm": 1.6563791301032011, + "language_loss": 0.64926422, + "learning_rate": 2.0409173206186183e-09, + "loss": 0.67075992, + "num_input_tokens_seen": 176992810, + "step": 8200, + "time_per_iteration": 2.6260411739349365 + }, + { + "auxiliary_loss_clip": 0.01092545, + "auxiliary_loss_mlp": 0.01022851, + "balance_loss_clip": 0.93518162, + "balance_loss_mlp": 1.01594329, + "epoch": 0.986111946131185, + "flos": 19938287134080.0, + "grad_norm": 1.993645844402326, + "language_loss": 0.87399769, + "learning_rate": 2.0058866221550617e-09, + "loss": 0.89515167, + "num_input_tokens_seen": 177011050, + "step": 8201, + "time_per_iteration": 3.628269672393799 + }, + { + "auxiliary_loss_clip": 0.01163684, + "auxiliary_loss_mlp": 0.01021118, + "balance_loss_clip": 1.04669988, + "balance_loss_mlp": 1.01411176, + "epoch": 0.9862321890218241, + "flos": 19828831415040.0, + "grad_norm": 2.0270762515381766, + "language_loss": 0.75216377, + "learning_rate": 1.971159017809976e-09, + "loss": 0.77401173, + "num_input_tokens_seen": 177029340, + "step": 8202, + "time_per_iteration": 3.5642614364624023 + }, + { + "auxiliary_loss_clip": 0.01139358, + "auxiliary_loss_mlp": 0.01030848, + "balance_loss_clip": 1.00972939, + "balance_loss_mlp": 1.02366567, + "epoch": 0.9863524319124631, + "flos": 21652051904640.0, + "grad_norm": 2.4730715057802595, + "language_loss": 0.77447867, + "learning_rate": 1.93673451285159e-09, + "loss": 0.79618073, + "num_input_tokens_seen": 177048390, + "step": 8203, + "time_per_iteration": 3.508507490158081 + }, + { + "auxiliary_loss_clip": 0.01024332, + "auxiliary_loss_mlp": 0.0100163, + "balance_loss_clip": 0.9358204, + "balance_loss_mlp": 1.00067639, + "epoch": 0.9864726748031023, + "flos": 52769977920000.0, + "grad_norm": 0.7305427207872394, + "language_loss": 0.5651778, + "learning_rate": 1.9026131125019495e-09, + "loss": 0.58543742, + "num_input_tokens_seen": 177105760, + "step": 8204, + "time_per_iteration": 3.183227777481079 + }, + { + "auxiliary_loss_clip": 0.01136713, + "auxiliary_loss_mlp": 0.01023484, + "balance_loss_clip": 1.01067042, + "balance_loss_mlp": 1.01682341, + "epoch": 0.9865929176937414, + "flos": 23364631526400.0, + "grad_norm": 2.078731902192086, + "language_loss": 0.86928928, + "learning_rate": 1.8687948219371363e-09, + "loss": 0.89089119, + "num_input_tokens_seen": 177124985, + "step": 8205, + "time_per_iteration": 2.6717031002044678 + }, + { + "auxiliary_loss_clip": 0.01167271, + "auxiliary_loss_mlp": 0.01031827, + "balance_loss_clip": 1.0473156, + "balance_loss_mlp": 1.02444816, + "epoch": 0.9867131605843804, + "flos": 21616679986560.0, + "grad_norm": 1.9839614497099558, + "language_loss": 0.88546562, + "learning_rate": 1.835279646287491e-09, + "loss": 0.90745664, + "num_input_tokens_seen": 177142995, + "step": 8206, + "time_per_iteration": 2.6278843879699707 + }, + { + "auxiliary_loss_clip": 0.01151759, + "auxiliary_loss_mlp": 0.01034123, + "balance_loss_clip": 1.01398277, + "balance_loss_mlp": 1.02637398, + "epoch": 0.9868334034750196, + "flos": 22271403139200.0, + "grad_norm": 1.8020545902503744, + "language_loss": 0.76435828, + "learning_rate": 1.8020675906371685e-09, + "loss": 0.78621709, + "num_input_tokens_seen": 177162390, + "step": 8207, + "time_per_iteration": 2.620211124420166 + }, + { + "auxiliary_loss_clip": 0.01065049, + "auxiliary_loss_mlp": 0.01025593, + "balance_loss_clip": 0.89240563, + "balance_loss_mlp": 1.01930165, + "epoch": 0.9869536463656586, + "flos": 25809573548160.0, + "grad_norm": 1.9467420267413653, + "language_loss": 0.74854797, + "learning_rate": 1.7691586600243612e-09, + "loss": 0.76945436, + "num_input_tokens_seen": 177181290, + "step": 8208, + "time_per_iteration": 2.6889355182647705 + }, + { + "auxiliary_loss_clip": 0.01115646, + "auxiliary_loss_mlp": 0.01022325, + "balance_loss_clip": 0.97462523, + "balance_loss_mlp": 1.01527333, + "epoch": 0.9870738892562977, + "flos": 16398500613120.0, + "grad_norm": 4.304511940285103, + "language_loss": 0.86339259, + "learning_rate": 1.7365528594415202e-09, + "loss": 0.8847723, + "num_input_tokens_seen": 177195360, + "step": 8209, + "time_per_iteration": 2.609712839126587 + }, + { + "auxiliary_loss_clip": 0.01145034, + "auxiliary_loss_mlp": 0.00509594, + "balance_loss_clip": 1.0102309, + "balance_loss_mlp": 0.0, + "epoch": 0.9871941321469369, + "flos": 35481358373760.0, + "grad_norm": 1.5515948594940538, + "language_loss": 0.678123, + "learning_rate": 1.7042501938346888e-09, + "loss": 0.69466925, + "num_input_tokens_seen": 177218090, + "step": 8210, + "time_per_iteration": 2.7247612476348877 + }, + { + "auxiliary_loss_clip": 0.01105791, + "auxiliary_loss_mlp": 0.01028456, + "balance_loss_clip": 0.96664912, + "balance_loss_mlp": 1.02130604, + "epoch": 0.9873143750375759, + "flos": 21434217874560.0, + "grad_norm": 2.529790484038628, + "language_loss": 0.76207227, + "learning_rate": 1.6722506681043913e-09, + "loss": 0.78341472, + "num_input_tokens_seen": 177237050, + "step": 8211, + "time_per_iteration": 3.6776540279388428 + }, + { + "auxiliary_loss_clip": 0.0112072, + "auxiliary_loss_mlp": 0.01025993, + "balance_loss_clip": 0.97184372, + "balance_loss_mlp": 1.01921296, + "epoch": 0.987434617928215, + "flos": 16326499800960.0, + "grad_norm": 2.256500284503598, + "language_loss": 0.69407511, + "learning_rate": 1.640554287104745e-09, + "loss": 0.71554232, + "num_input_tokens_seen": 177255325, + "step": 8212, + "time_per_iteration": 2.666171073913574 + }, + { + "auxiliary_loss_clip": 0.01094459, + "auxiliary_loss_mlp": 0.01026125, + "balance_loss_clip": 0.92848402, + "balance_loss_mlp": 1.01863241, + "epoch": 0.9875548608188541, + "flos": 17851984456320.0, + "grad_norm": 2.1476314379141623, + "language_loss": 0.80402863, + "learning_rate": 1.609161055644348e-09, + "loss": 0.82523447, + "num_input_tokens_seen": 177271250, + "step": 8213, + "time_per_iteration": 2.641176462173462 + }, + { + "auxiliary_loss_clip": 0.01143255, + "auxiliary_loss_mlp": 0.01026456, + "balance_loss_clip": 1.00642848, + "balance_loss_mlp": 1.01860881, + "epoch": 0.9876751037094932, + "flos": 26132876887680.0, + "grad_norm": 2.563269695226291, + "language_loss": 0.68274558, + "learning_rate": 1.5780709784849467e-09, + "loss": 0.70444262, + "num_input_tokens_seen": 177288270, + "step": 8214, + "time_per_iteration": 2.674901008605957 + }, + { + "auxiliary_loss_clip": 0.01052547, + "auxiliary_loss_mlp": 0.01026179, + "balance_loss_clip": 0.85966784, + "balance_loss_mlp": 1.01900244, + "epoch": 0.9877953466001322, + "flos": 15991344973440.0, + "grad_norm": 2.0003526991735283, + "language_loss": 0.82369208, + "learning_rate": 1.5472840603436565e-09, + "loss": 0.84447932, + "num_input_tokens_seen": 177305500, + "step": 8215, + "time_per_iteration": 2.776216983795166 + }, + { + "auxiliary_loss_clip": 0.01118887, + "auxiliary_loss_mlp": 0.01022186, + "balance_loss_clip": 0.97143304, + "balance_loss_mlp": 1.01524496, + "epoch": 0.9879155894907714, + "flos": 18806777827200.0, + "grad_norm": 1.9122480846263357, + "language_loss": 0.7816487, + "learning_rate": 1.5168003058900757e-09, + "loss": 0.8030594, + "num_input_tokens_seen": 177323500, + "step": 8216, + "time_per_iteration": 2.6710362434387207 + }, + { + "auxiliary_loss_clip": 0.01090493, + "auxiliary_loss_mlp": 0.01026827, + "balance_loss_clip": 0.92968756, + "balance_loss_mlp": 1.02011883, + "epoch": 0.9880358323814105, + "flos": 22382044007040.0, + "grad_norm": 1.8453042163763644, + "language_loss": 0.91949028, + "learning_rate": 1.4866197197491715e-09, + "loss": 0.94066352, + "num_input_tokens_seen": 177342860, + "step": 8217, + "time_per_iteration": 2.7093825340270996 + }, + { + "auxiliary_loss_clip": 0.011449, + "auxiliary_loss_mlp": 0.00509418, + "balance_loss_clip": 1.01059866, + "balance_loss_mlp": 0.0, + "epoch": 0.9881560752720495, + "flos": 15668831733120.0, + "grad_norm": 4.219660819433708, + "language_loss": 0.7882874, + "learning_rate": 1.4567423064988371e-09, + "loss": 0.80483061, + "num_input_tokens_seen": 177360210, + "step": 8218, + "time_per_iteration": 2.760760545730591 + }, + { + "auxiliary_loss_clip": 0.01165605, + "auxiliary_loss_mlp": 0.01024834, + "balance_loss_clip": 1.04856062, + "balance_loss_mlp": 1.01742554, + "epoch": 0.9882763181626887, + "flos": 21500113374720.0, + "grad_norm": 2.0592968294578737, + "language_loss": 0.78035831, + "learning_rate": 1.4271680706718913e-09, + "loss": 0.80226272, + "num_input_tokens_seen": 177377885, + "step": 8219, + "time_per_iteration": 2.6140551567077637 + }, + { + "auxiliary_loss_clip": 0.01146222, + "auxiliary_loss_mlp": 0.01026637, + "balance_loss_clip": 1.01258171, + "balance_loss_mlp": 1.01907909, + "epoch": 0.9883965610533277, + "flos": 28034598551040.0, + "grad_norm": 2.089086637780297, + "language_loss": 0.82497835, + "learning_rate": 1.3978970167543013e-09, + "loss": 0.84670693, + "num_input_tokens_seen": 177398065, + "step": 8220, + "time_per_iteration": 2.692345380783081 + }, + { + "auxiliary_loss_clip": 0.01111303, + "auxiliary_loss_mlp": 0.01024034, + "balance_loss_clip": 0.97207093, + "balance_loss_mlp": 1.01742959, + "epoch": 0.9885168039439668, + "flos": 14098601710080.0, + "grad_norm": 2.495441624456559, + "language_loss": 0.77903235, + "learning_rate": 1.3689291491867372e-09, + "loss": 0.80038571, + "num_input_tokens_seen": 177416380, + "step": 8221, + "time_per_iteration": 2.5870096683502197 + }, + { + "auxiliary_loss_clip": 0.01163641, + "auxiliary_loss_mlp": 0.0102502, + "balance_loss_clip": 1.0464561, + "balance_loss_mlp": 1.01807308, + "epoch": 0.988637046834606, + "flos": 26432013352320.0, + "grad_norm": 1.998523160801931, + "language_loss": 0.73515135, + "learning_rate": 1.3402644723636836e-09, + "loss": 0.757038, + "num_input_tokens_seen": 177438410, + "step": 8222, + "time_per_iteration": 2.669780969619751 + }, + { + "auxiliary_loss_clip": 0.0111441, + "auxiliary_loss_mlp": 0.01029259, + "balance_loss_clip": 0.97341663, + "balance_loss_mlp": 1.02209473, + "epoch": 0.988757289725245, + "flos": 25229113764480.0, + "grad_norm": 1.9550967741520062, + "language_loss": 0.83506733, + "learning_rate": 1.311902990633218e-09, + "loss": 0.85650408, + "num_input_tokens_seen": 177457375, + "step": 8223, + "time_per_iteration": 2.6914587020874023 + }, + { + "auxiliary_loss_clip": 0.01108208, + "auxiliary_loss_mlp": 0.01026446, + "balance_loss_clip": 0.96537441, + "balance_loss_mlp": 1.01940036, + "epoch": 0.9888775326158841, + "flos": 26359042872960.0, + "grad_norm": 1.5143988805512087, + "language_loss": 0.71258092, + "learning_rate": 1.2838447082978987e-09, + "loss": 0.73392743, + "num_input_tokens_seen": 177478530, + "step": 8224, + "time_per_iteration": 2.7271604537963867 + }, + { + "auxiliary_loss_clip": 0.01136088, + "auxiliary_loss_mlp": 0.01024314, + "balance_loss_clip": 1.0073303, + "balance_loss_mlp": 1.01767135, + "epoch": 0.9889977755065231, + "flos": 24316120846080.0, + "grad_norm": 2.470508769142906, + "language_loss": 0.8310594, + "learning_rate": 1.2560896296143208e-09, + "loss": 0.8526634, + "num_input_tokens_seen": 177496995, + "step": 8225, + "time_per_iteration": 2.654616117477417 + }, + { + "auxiliary_loss_clip": 0.01163546, + "auxiliary_loss_mlp": 0.01023628, + "balance_loss_clip": 1.04796743, + "balance_loss_mlp": 1.01647282, + "epoch": 0.9891180183971623, + "flos": 18951066760320.0, + "grad_norm": 2.0302118438241656, + "language_loss": 0.81898546, + "learning_rate": 1.2286377587926722e-09, + "loss": 0.84085721, + "num_input_tokens_seen": 177513785, + "step": 8226, + "time_per_iteration": 2.522533655166626 + }, + { + "auxiliary_loss_clip": 0.01164175, + "auxiliary_loss_mlp": 0.01024198, + "balance_loss_clip": 1.04756546, + "balance_loss_mlp": 1.01721859, + "epoch": 0.9892382612878013, + "flos": 26176580760960.0, + "grad_norm": 1.9858609139488819, + "language_loss": 0.74775922, + "learning_rate": 1.2014890999973992e-09, + "loss": 0.76964301, + "num_input_tokens_seen": 177530705, + "step": 8227, + "time_per_iteration": 2.6475508213043213 + }, + { + "auxiliary_loss_clip": 0.0116106, + "auxiliary_loss_mlp": 0.01021018, + "balance_loss_clip": 1.04564011, + "balance_loss_mlp": 1.0139668, + "epoch": 0.9893585041784404, + "flos": 25449605400960.0, + "grad_norm": 1.5852889389685, + "language_loss": 0.78637552, + "learning_rate": 1.1746436573472073e-09, + "loss": 0.80819631, + "num_input_tokens_seen": 177552440, + "step": 8228, + "time_per_iteration": 4.435015916824341 + }, + { + "auxiliary_loss_clip": 0.01125671, + "auxiliary_loss_mlp": 0.01028202, + "balance_loss_clip": 0.97108209, + "balance_loss_mlp": 1.02127302, + "epoch": 0.9894787470690796, + "flos": 20189302352640.0, + "grad_norm": 1.9575229680488146, + "language_loss": 0.69650865, + "learning_rate": 1.1481014349141726e-09, + "loss": 0.71804738, + "num_input_tokens_seen": 177569660, + "step": 8229, + "time_per_iteration": 3.552865743637085 + }, + { + "auxiliary_loss_clip": 0.0111826, + "auxiliary_loss_mlp": 0.01032773, + "balance_loss_clip": 0.97139776, + "balance_loss_mlp": 1.0250423, + "epoch": 0.9895989899597186, + "flos": 24644308435200.0, + "grad_norm": 1.8018229883154533, + "language_loss": 0.84433341, + "learning_rate": 1.121862436724852e-09, + "loss": 0.86584371, + "num_input_tokens_seen": 177588500, + "step": 8230, + "time_per_iteration": 2.7440905570983887 + }, + { + "auxiliary_loss_clip": 0.01142047, + "auxiliary_loss_mlp": 0.01024308, + "balance_loss_clip": 1.01225877, + "balance_loss_mlp": 1.01722121, + "epoch": 0.9897192328503577, + "flos": 21799034357760.0, + "grad_norm": 1.831683781852547, + "language_loss": 0.70054615, + "learning_rate": 1.0959266667598388e-09, + "loss": 0.72220963, + "num_input_tokens_seen": 177607315, + "step": 8231, + "time_per_iteration": 2.6711230278015137 + }, + { + "auxiliary_loss_clip": 0.01095686, + "auxiliary_loss_mlp": 0.01030994, + "balance_loss_clip": 0.93345702, + "balance_loss_mlp": 1.02309072, + "epoch": 0.9898394757409968, + "flos": 21325229032320.0, + "grad_norm": 2.6184682531599917, + "language_loss": 0.74608564, + "learning_rate": 1.0702941289533196e-09, + "loss": 0.76735234, + "num_input_tokens_seen": 177625990, + "step": 8232, + "time_per_iteration": 2.7218525409698486 + }, + { + "auxiliary_loss_clip": 0.0109637, + "auxiliary_loss_mlp": 0.01024556, + "balance_loss_clip": 0.93402821, + "balance_loss_mlp": 1.01777315, + "epoch": 0.9899597186316359, + "flos": 18545024442240.0, + "grad_norm": 2.142620553519716, + "language_loss": 0.89018863, + "learning_rate": 1.0449648271939615e-09, + "loss": 0.91139793, + "num_input_tokens_seen": 177642335, + "step": 8233, + "time_per_iteration": 2.6658084392547607 + }, + { + "auxiliary_loss_clip": 0.01074572, + "auxiliary_loss_mlp": 0.0050923, + "balance_loss_clip": 0.89853001, + "balance_loss_mlp": 0.0, + "epoch": 0.990079961522275, + "flos": 23766723348480.0, + "grad_norm": 1.8711569831469457, + "language_loss": 0.72512293, + "learning_rate": 1.0199387653240243e-09, + "loss": 0.74096096, + "num_input_tokens_seen": 177662025, + "step": 8234, + "time_per_iteration": 2.7185537815093994 + }, + { + "auxiliary_loss_clip": 0.01110313, + "auxiliary_loss_mlp": 0.01027213, + "balance_loss_clip": 0.97062737, + "balance_loss_mlp": 1.02012038, + "epoch": 0.9902002044129141, + "flos": 16399182971520.0, + "grad_norm": 1.553189242249634, + "language_loss": 0.70918769, + "learning_rate": 9.952159471400267e-10, + "loss": 0.73056293, + "num_input_tokens_seen": 177679065, + "step": 8235, + "time_per_iteration": 2.598564386367798 + }, + { + "auxiliary_loss_clip": 0.01143866, + "auxiliary_loss_mlp": 0.00508895, + "balance_loss_clip": 1.009359, + "balance_loss_mlp": 0.0, + "epoch": 0.9903204473035532, + "flos": 22559657783040.0, + "grad_norm": 1.8701885980285848, + "language_loss": 0.84157062, + "learning_rate": 9.707963763923022e-10, + "loss": 0.85809815, + "num_input_tokens_seen": 177698115, + "step": 8236, + "time_per_iteration": 2.6066501140594482 + }, + { + "auxiliary_loss_clip": 0.01116037, + "auxiliary_loss_mlp": 0.0102393, + "balance_loss_clip": 0.96956563, + "balance_loss_mlp": 1.01651561, + "epoch": 0.9904406901941922, + "flos": 16144001775360.0, + "grad_norm": 1.8986308291921288, + "language_loss": 0.79040611, + "learning_rate": 9.466800567854427e-10, + "loss": 0.81180573, + "num_input_tokens_seen": 177716715, + "step": 8237, + "time_per_iteration": 3.559481620788574 + }, + { + "auxiliary_loss_clip": 0.0109332, + "auxiliary_loss_mlp": 0.01026344, + "balance_loss_clip": 0.92885798, + "balance_loss_mlp": 1.01882243, + "epoch": 0.9905609330848314, + "flos": 26651499408000.0, + "grad_norm": 2.263024825548955, + "language_loss": 0.68327117, + "learning_rate": 9.228669919778553e-10, + "loss": 0.70446777, + "num_input_tokens_seen": 177735640, + "step": 8238, + "time_per_iteration": 2.7038326263427734 + }, + { + "auxiliary_loss_clip": 0.01114443, + "auxiliary_loss_mlp": 0.01026261, + "balance_loss_clip": 0.96975243, + "balance_loss_mlp": 1.01933825, + "epoch": 0.9906811759754705, + "flos": 23111820627840.0, + "grad_norm": 2.1907477584340636, + "language_loss": 0.79272002, + "learning_rate": 8.993571855817617e-10, + "loss": 0.81412709, + "num_input_tokens_seen": 177754470, + "step": 8239, + "time_per_iteration": 2.686549186706543 + }, + { + "auxiliary_loss_clip": 0.01139587, + "auxiliary_loss_mlp": 0.01021757, + "balance_loss_clip": 1.00851488, + "balance_loss_mlp": 1.01464581, + "epoch": 0.9908014188661095, + "flos": 22090593052800.0, + "grad_norm": 1.826755356381217, + "language_loss": 0.75113177, + "learning_rate": 8.761506411638642e-10, + "loss": 0.77274513, + "num_input_tokens_seen": 177773935, + "step": 8240, + "time_per_iteration": 2.60701847076416 + }, + { + "auxiliary_loss_clip": 0.01119813, + "auxiliary_loss_mlp": 0.01029685, + "balance_loss_clip": 0.97565788, + "balance_loss_mlp": 1.02272344, + "epoch": 0.9909216617567487, + "flos": 19242948677760.0, + "grad_norm": 1.7990328282334143, + "language_loss": 0.73540425, + "learning_rate": 8.53247362244236e-10, + "loss": 0.75689924, + "num_input_tokens_seen": 177792745, + "step": 8241, + "time_per_iteration": 2.6724789142608643 + }, + { + "auxiliary_loss_clip": 0.01116272, + "auxiliary_loss_mlp": 0.0102615, + "balance_loss_clip": 0.97118849, + "balance_loss_mlp": 1.01847875, + "epoch": 0.9910419046473877, + "flos": 23621213352960.0, + "grad_norm": 1.6844829010497293, + "language_loss": 0.68232286, + "learning_rate": 8.306473522976532e-10, + "loss": 0.70374703, + "num_input_tokens_seen": 177812150, + "step": 8242, + "time_per_iteration": 2.7255046367645264 + }, + { + "auxiliary_loss_clip": 0.01165216, + "auxiliary_loss_mlp": 0.01026354, + "balance_loss_clip": 1.04936528, + "balance_loss_mlp": 1.01899898, + "epoch": 0.9911621475380268, + "flos": 22711380831360.0, + "grad_norm": 1.8867793955322587, + "language_loss": 0.71889484, + "learning_rate": 8.083506147522623e-10, + "loss": 0.74081051, + "num_input_tokens_seen": 177831545, + "step": 8243, + "time_per_iteration": 2.652863025665283 + }, + { + "auxiliary_loss_clip": 0.01134946, + "auxiliary_loss_mlp": 0.01025125, + "balance_loss_clip": 1.00786757, + "balance_loss_mlp": 1.01819015, + "epoch": 0.991282390428666, + "flos": 13516956777600.0, + "grad_norm": 2.0721848452849163, + "language_loss": 0.85374546, + "learning_rate": 7.863571529906909e-10, + "loss": 0.87534618, + "num_input_tokens_seen": 177847130, + "step": 8244, + "time_per_iteration": 2.594231367111206 + }, + { + "auxiliary_loss_clip": 0.01041424, + "auxiliary_loss_mlp": 0.01000016, + "balance_loss_clip": 0.9719547, + "balance_loss_mlp": 0.99904424, + "epoch": 0.991402633319305, + "flos": 61830492071040.0, + "grad_norm": 0.7250690399646178, + "language_loss": 0.59626347, + "learning_rate": 7.646669703489372e-10, + "loss": 0.61667788, + "num_input_tokens_seen": 177911440, + "step": 8245, + "time_per_iteration": 3.2713146209716797 + }, + { + "auxiliary_loss_clip": 0.00983538, + "auxiliary_loss_mlp": 0.01027736, + "balance_loss_clip": 0.73918378, + "balance_loss_mlp": 1.02060771, + "epoch": 0.9915228762099441, + "flos": 18770148933120.0, + "grad_norm": 2.139318702741226, + "language_loss": 0.57284844, + "learning_rate": 7.432800701177023e-10, + "loss": 0.59296119, + "num_input_tokens_seen": 177929440, + "step": 8246, + "time_per_iteration": 3.1634514331817627 + }, + { + "auxiliary_loss_clip": 0.01023099, + "auxiliary_loss_mlp": 0.00999639, + "balance_loss_clip": 0.93609607, + "balance_loss_mlp": 0.99874461, + "epoch": 0.9916431191005832, + "flos": 65936660244480.0, + "grad_norm": 0.7966062409338043, + "language_loss": 0.57740283, + "learning_rate": 7.221964555415017e-10, + "loss": 0.5976302, + "num_input_tokens_seen": 177989100, + "step": 8247, + "time_per_iteration": 3.3947975635528564 + }, + { + "auxiliary_loss_clip": 0.01116036, + "auxiliary_loss_mlp": 0.01022844, + "balance_loss_clip": 0.97251272, + "balance_loss_mlp": 1.01578355, + "epoch": 0.9917633619912223, + "flos": 16581573256320.0, + "grad_norm": 1.9737852620024754, + "language_loss": 0.75033873, + "learning_rate": 7.01416129818222e-10, + "loss": 0.7717275, + "num_input_tokens_seen": 178006720, + "step": 8248, + "time_per_iteration": 2.627028703689575 + }, + { + "auxiliary_loss_clip": 0.01101776, + "auxiliary_loss_mlp": 0.01027944, + "balance_loss_clip": 0.93370581, + "balance_loss_mlp": 1.02069664, + "epoch": 0.9918836048818613, + "flos": 25411108999680.0, + "grad_norm": 2.1499304799147336, + "language_loss": 0.58506209, + "learning_rate": 6.809390961006745e-10, + "loss": 0.60635936, + "num_input_tokens_seen": 178026850, + "step": 8249, + "time_per_iteration": 2.721306800842285 + }, + { + "auxiliary_loss_clip": 0.01119355, + "auxiliary_loss_mlp": 0.0102685, + "balance_loss_clip": 0.97236162, + "balance_loss_mlp": 1.01927459, + "epoch": 0.9920038477725005, + "flos": 25046867134080.0, + "grad_norm": 1.8151685131542117, + "language_loss": 0.68581045, + "learning_rate": 6.607653574948191e-10, + "loss": 0.70727253, + "num_input_tokens_seen": 178047630, + "step": 8250, + "time_per_iteration": 2.6826181411743164 + }, + { + "auxiliary_loss_clip": 0.01132319, + "auxiliary_loss_mlp": 0.01023179, + "balance_loss_clip": 1.00669372, + "balance_loss_mlp": 1.01619339, + "epoch": 0.9921240906631396, + "flos": 21829773421440.0, + "grad_norm": 1.8624521152836917, + "language_loss": 0.81913114, + "learning_rate": 6.408949170613187e-10, + "loss": 0.84068608, + "num_input_tokens_seen": 178066895, + "step": 8251, + "time_per_iteration": 2.596766233444214 + }, + { + "auxiliary_loss_clip": 0.01114872, + "auxiliary_loss_mlp": 0.01025564, + "balance_loss_clip": 0.97035992, + "balance_loss_mlp": 1.01816142, + "epoch": 0.9922443335537786, + "flos": 24864225454080.0, + "grad_norm": 1.7539155609722243, + "language_loss": 0.81938505, + "learning_rate": 6.213277778144288e-10, + "loss": 0.84078944, + "num_input_tokens_seen": 178088540, + "step": 8252, + "time_per_iteration": 2.689500093460083 + }, + { + "auxiliary_loss_clip": 0.01048576, + "auxiliary_loss_mlp": 0.01029932, + "balance_loss_clip": 0.85480362, + "balance_loss_mlp": 1.02257955, + "epoch": 0.9923645764444178, + "flos": 21613088626560.0, + "grad_norm": 2.2252731882700436, + "language_loss": 0.66958874, + "learning_rate": 6.020639427224416e-10, + "loss": 0.69037384, + "num_input_tokens_seen": 178106185, + "step": 8253, + "time_per_iteration": 2.7325077056884766 + }, + { + "auxiliary_loss_clip": 0.01116914, + "auxiliary_loss_mlp": 0.01025, + "balance_loss_clip": 0.97142041, + "balance_loss_mlp": 1.01806462, + "epoch": 0.9924848193350568, + "flos": 25001798544000.0, + "grad_norm": 2.3912403453819318, + "language_loss": 0.72874713, + "learning_rate": 5.831034147076864e-10, + "loss": 0.75016618, + "num_input_tokens_seen": 178123435, + "step": 8254, + "time_per_iteration": 4.427818536758423 + }, + { + "auxiliary_loss_clip": 0.01035035, + "auxiliary_loss_mlp": 0.01002055, + "balance_loss_clip": 0.97083545, + "balance_loss_mlp": 1.00113106, + "epoch": 0.9926050622256959, + "flos": 68912543151360.0, + "grad_norm": 0.6864431864352559, + "language_loss": 0.55716646, + "learning_rate": 5.644461966463065e-10, + "loss": 0.5775373, + "num_input_tokens_seen": 178191045, + "step": 8255, + "time_per_iteration": 4.225614070892334 + }, + { + "auxiliary_loss_clip": 0.01117775, + "auxiliary_loss_mlp": 0.01027223, + "balance_loss_clip": 0.97275996, + "balance_loss_mlp": 1.02009988, + "epoch": 0.9927253051163349, + "flos": 20923675914240.0, + "grad_norm": 2.0757278451384016, + "language_loss": 0.75893283, + "learning_rate": 5.460922913687049e-10, + "loss": 0.78038281, + "num_input_tokens_seen": 178210135, + "step": 8256, + "time_per_iteration": 2.7241363525390625 + }, + { + "auxiliary_loss_clip": 0.01069033, + "auxiliary_loss_mlp": 0.00510169, + "balance_loss_clip": 0.89247882, + "balance_loss_mlp": 0.0, + "epoch": 0.9928455480069741, + "flos": 22308211601280.0, + "grad_norm": 1.9664928269576016, + "language_loss": 0.75169748, + "learning_rate": 5.280417016593208e-10, + "loss": 0.76748955, + "num_input_tokens_seen": 178229925, + "step": 8257, + "time_per_iteration": 2.725895881652832 + }, + { + "auxiliary_loss_clip": 0.0113965, + "auxiliary_loss_mlp": 0.00508645, + "balance_loss_clip": 1.01207757, + "balance_loss_mlp": 0.0, + "epoch": 0.9929657908976132, + "flos": 17383889393280.0, + "grad_norm": 1.6469707641260791, + "language_loss": 0.74921507, + "learning_rate": 5.102944302559642e-10, + "loss": 0.76569796, + "num_input_tokens_seen": 178247420, + "step": 8258, + "time_per_iteration": 2.666602611541748 + }, + { + "auxiliary_loss_clip": 0.010559, + "auxiliary_loss_mlp": 0.01028896, + "balance_loss_clip": 0.85969508, + "balance_loss_mlp": 1.0215888, + "epoch": 0.9930860337882522, + "flos": 22674680110080.0, + "grad_norm": 2.1458256147052466, + "language_loss": 0.79592407, + "learning_rate": 4.9285047985137e-10, + "loss": 0.81677198, + "num_input_tokens_seen": 178266840, + "step": 8259, + "time_per_iteration": 2.7733426094055176 + }, + { + "auxiliary_loss_clip": 0.01148148, + "auxiliary_loss_mlp": 0.01029757, + "balance_loss_clip": 1.01273477, + "balance_loss_mlp": 1.02283061, + "epoch": 0.9932062766788914, + "flos": 28147789284480.0, + "grad_norm": 2.2514187746916594, + "language_loss": 0.74607599, + "learning_rate": 4.757098530916436e-10, + "loss": 0.76785505, + "num_input_tokens_seen": 178287285, + "step": 8260, + "time_per_iteration": 2.663079261779785 + }, + { + "auxiliary_loss_clip": 0.01144201, + "auxiliary_loss_mlp": 0.0103086, + "balance_loss_clip": 1.01214409, + "balance_loss_mlp": 1.02307022, + "epoch": 0.9933265195695304, + "flos": 20156659868160.0, + "grad_norm": 4.6138520114929875, + "language_loss": 0.77262938, + "learning_rate": 4.5887255257670563e-10, + "loss": 0.79437995, + "num_input_tokens_seen": 178304325, + "step": 8261, + "time_per_iteration": 2.5491538047790527 + }, + { + "auxiliary_loss_clip": 0.01162239, + "auxiliary_loss_mlp": 0.01026143, + "balance_loss_clip": 1.04655647, + "balance_loss_mlp": 1.01930308, + "epoch": 0.9934467624601695, + "flos": 21362037494400.0, + "grad_norm": 1.8714773246927572, + "language_loss": 0.76820213, + "learning_rate": 4.4233858086117906e-10, + "loss": 0.79008597, + "num_input_tokens_seen": 178322850, + "step": 8262, + "time_per_iteration": 2.591808557510376 + }, + { + "auxiliary_loss_clip": 0.01068951, + "auxiliary_loss_mlp": 0.01024223, + "balance_loss_clip": 0.89716613, + "balance_loss_mlp": 1.01751804, + "epoch": 0.9935670053508087, + "flos": 19756040503680.0, + "grad_norm": 2.1770854295439697, + "language_loss": 0.67623568, + "learning_rate": 4.261079404528356e-10, + "loss": 0.69716746, + "num_input_tokens_seen": 178342330, + "step": 8263, + "time_per_iteration": 3.9301071166992188 + }, + { + "auxiliary_loss_clip": 0.01137894, + "auxiliary_loss_mlp": 0.01025115, + "balance_loss_clip": 1.00921392, + "balance_loss_mlp": 1.01808119, + "epoch": 0.9936872482414477, + "flos": 21978838863360.0, + "grad_norm": 1.9247454070648744, + "language_loss": 0.69178283, + "learning_rate": 4.1018063381437205e-10, + "loss": 0.71341288, + "num_input_tokens_seen": 178362715, + "step": 8264, + "time_per_iteration": 2.672431707382202 + }, + { + "auxiliary_loss_clip": 0.01037166, + "auxiliary_loss_mlp": 0.01008306, + "balance_loss_clip": 0.97475618, + "balance_loss_mlp": 1.00720978, + "epoch": 0.9938074911320868, + "flos": 69810667839360.0, + "grad_norm": 0.8612003699216421, + "language_loss": 0.61088109, + "learning_rate": 3.9455666336141167e-10, + "loss": 0.63133585, + "num_input_tokens_seen": 178426495, + "step": 8265, + "time_per_iteration": 3.239495038986206 + }, + { + "auxiliary_loss_clip": 0.01164953, + "auxiliary_loss_mlp": 0.01029239, + "balance_loss_clip": 1.0504303, + "balance_loss_mlp": 1.02191377, + "epoch": 0.9939277340227259, + "flos": 15084170058240.0, + "grad_norm": 2.6703262016809157, + "language_loss": 0.83534515, + "learning_rate": 3.7923603146450267e-10, + "loss": 0.85728705, + "num_input_tokens_seen": 178442555, + "step": 8266, + "time_per_iteration": 2.5761075019836426 + }, + { + "auxiliary_loss_clip": 0.01095364, + "auxiliary_loss_mlp": 0.01022302, + "balance_loss_clip": 0.93114531, + "balance_loss_mlp": 1.01545656, + "epoch": 0.994047976913365, + "flos": 17712364291200.0, + "grad_norm": 3.1917685708336476, + "language_loss": 0.80681407, + "learning_rate": 3.642187404473418e-10, + "loss": 0.82799065, + "num_input_tokens_seen": 178460715, + "step": 8267, + "time_per_iteration": 2.7401318550109863 + }, + { + "auxiliary_loss_clip": 0.01140601, + "auxiliary_loss_mlp": 0.01021034, + "balance_loss_clip": 1.00920224, + "balance_loss_mlp": 1.01395607, + "epoch": 0.994168219804004, + "flos": 19171558396800.0, + "grad_norm": 2.6241633867638003, + "language_loss": 0.86126673, + "learning_rate": 3.495047925885508e-10, + "loss": 0.88288307, + "num_input_tokens_seen": 178479050, + "step": 8268, + "time_per_iteration": 2.631901979446411 + }, + { + "auxiliary_loss_clip": 0.01113587, + "auxiliary_loss_mlp": 0.0102893, + "balance_loss_clip": 0.96926081, + "balance_loss_mlp": 1.02156866, + "epoch": 0.9942884626946432, + "flos": 17851589406720.0, + "grad_norm": 2.034959741859419, + "language_loss": 0.82914436, + "learning_rate": 3.350941901199e-10, + "loss": 0.85056961, + "num_input_tokens_seen": 178495970, + "step": 8269, + "time_per_iteration": 2.65519380569458 + }, + { + "auxiliary_loss_clip": 0.01118902, + "auxiliary_loss_mlp": 0.01027492, + "balance_loss_clip": 0.97027135, + "balance_loss_mlp": 1.02044976, + "epoch": 0.9944087055852823, + "flos": 18796578364800.0, + "grad_norm": 2.415200552451641, + "language_loss": 0.83154869, + "learning_rate": 3.2098693522764066e-10, + "loss": 0.85301262, + "num_input_tokens_seen": 178509170, + "step": 8270, + "time_per_iteration": 2.636218547821045 + }, + { + "auxiliary_loss_clip": 0.01123599, + "auxiliary_loss_mlp": 0.00508988, + "balance_loss_clip": 0.9724493, + "balance_loss_mlp": 0.0, + "epoch": 0.9945289484759213, + "flos": 20996969616000.0, + "grad_norm": 1.889290158953081, + "language_loss": 0.81023109, + "learning_rate": 3.071830300516165e-10, + "loss": 0.82655704, + "num_input_tokens_seen": 178527000, + "step": 8271, + "time_per_iteration": 2.6833391189575195 + }, + { + "auxiliary_loss_clip": 0.01145316, + "auxiliary_loss_mlp": 0.01028276, + "balance_loss_clip": 1.01056778, + "balance_loss_mlp": 1.02080417, + "epoch": 0.9946491913665605, + "flos": 14756952136320.0, + "grad_norm": 2.7605296345080625, + "language_loss": 0.71047246, + "learning_rate": 2.9368247668615234e-10, + "loss": 0.73220837, + "num_input_tokens_seen": 178545590, + "step": 8272, + "time_per_iteration": 2.6543664932250977 + }, + { + "auxiliary_loss_clip": 0.01170741, + "auxiliary_loss_mlp": 0.01031012, + "balance_loss_clip": 1.0522517, + "balance_loss_mlp": 1.02268803, + "epoch": 0.9947694342571995, + "flos": 12669931186560.0, + "grad_norm": 3.01085524048587, + "language_loss": 0.61672038, + "learning_rate": 2.804852771789434e-10, + "loss": 0.63873792, + "num_input_tokens_seen": 178558890, + "step": 8273, + "time_per_iteration": 2.5707931518554688 + }, + { + "auxiliary_loss_clip": 0.01162677, + "auxiliary_loss_mlp": 0.01025727, + "balance_loss_clip": 1.04819906, + "balance_loss_mlp": 1.01880133, + "epoch": 0.9948896771478386, + "flos": 18843442634880.0, + "grad_norm": 1.7318458876211222, + "language_loss": 0.55173314, + "learning_rate": 2.675914335321661e-10, + "loss": 0.57361722, + "num_input_tokens_seen": 178577645, + "step": 8274, + "time_per_iteration": 2.6408276557922363 + }, + { + "auxiliary_loss_clip": 0.01147723, + "auxiliary_loss_mlp": 0.01028541, + "balance_loss_clip": 1.01131737, + "balance_loss_mlp": 1.02117372, + "epoch": 0.9950099200384778, + "flos": 24900207903360.0, + "grad_norm": 3.043568936150111, + "language_loss": 0.79968297, + "learning_rate": 2.550009477018111e-10, + "loss": 0.82144564, + "num_input_tokens_seen": 178596415, + "step": 8275, + "time_per_iteration": 2.6333839893341064 + }, + { + "auxiliary_loss_clip": 0.0111771, + "auxiliary_loss_mlp": 0.00509428, + "balance_loss_clip": 0.97330821, + "balance_loss_mlp": 0.0, + "epoch": 0.9951301629291168, + "flos": 23733613987200.0, + "grad_norm": 1.9005493765087293, + "language_loss": 0.63697255, + "learning_rate": 2.4271382159790634e-10, + "loss": 0.6532439, + "num_input_tokens_seen": 178613845, + "step": 8276, + "time_per_iteration": 2.667895555496216 + }, + { + "auxiliary_loss_clip": 0.01059674, + "auxiliary_loss_mlp": 0.01029515, + "balance_loss_clip": 0.85842735, + "balance_loss_mlp": 1.02279437, + "epoch": 0.9952504058197559, + "flos": 22236893147520.0, + "grad_norm": 2.1034973264722354, + "language_loss": 0.85652572, + "learning_rate": 2.3073005708429406e-10, + "loss": 0.87741756, + "num_input_tokens_seen": 178633490, + "step": 8277, + "time_per_iteration": 2.791806936264038 + }, + { + "auxiliary_loss_clip": 0.01092914, + "auxiliary_loss_mlp": 0.01025753, + "balance_loss_clip": 0.93587607, + "balance_loss_mlp": 1.01873481, + "epoch": 0.995370648710395, + "flos": 21211032718080.0, + "grad_norm": 1.7630622357725008, + "language_loss": 0.72371155, + "learning_rate": 2.190496559788535e-10, + "loss": 0.7448982, + "num_input_tokens_seen": 178651775, + "step": 8278, + "time_per_iteration": 2.78309965133667 + }, + { + "auxiliary_loss_clip": 0.01116161, + "auxiliary_loss_mlp": 0.01029509, + "balance_loss_clip": 0.97318566, + "balance_loss_mlp": 1.02170098, + "epoch": 0.9954908916010341, + "flos": 14866731077760.0, + "grad_norm": 2.681066507318154, + "language_loss": 0.76351023, + "learning_rate": 2.0767262005372265e-10, + "loss": 0.78496695, + "num_input_tokens_seen": 178669290, + "step": 8279, + "time_per_iteration": 2.65704083442688 + }, + { + "auxiliary_loss_clip": 0.01101686, + "auxiliary_loss_mlp": 0.01034767, + "balance_loss_clip": 0.93288529, + "balance_loss_mlp": 1.02701259, + "epoch": 0.9956111344916732, + "flos": 19208259118080.0, + "grad_norm": 1.749791293880615, + "language_loss": 0.75048172, + "learning_rate": 1.965989510346322e-10, + "loss": 0.77184618, + "num_input_tokens_seen": 178688410, + "step": 8280, + "time_per_iteration": 3.704442262649536 + }, + { + "auxiliary_loss_clip": 0.01065943, + "auxiliary_loss_mlp": 0.01025598, + "balance_loss_clip": 0.89414358, + "balance_loss_mlp": 1.01784408, + "epoch": 0.9957313773823123, + "flos": 20047060494720.0, + "grad_norm": 2.297359220521375, + "language_loss": 0.71212041, + "learning_rate": 1.8582865060134955e-10, + "loss": 0.7330358, + "num_input_tokens_seen": 178706600, + "step": 8281, + "time_per_iteration": 3.680910587310791 + }, + { + "auxiliary_loss_clip": 0.01059396, + "auxiliary_loss_mlp": 0.00999688, + "balance_loss_clip": 1.00874209, + "balance_loss_mlp": 0.99875224, + "epoch": 0.9958516202729514, + "flos": 57483253768320.0, + "grad_norm": 0.7775896964261049, + "language_loss": 0.55728984, + "learning_rate": 1.7536172038790098e-10, + "loss": 0.57788068, + "num_input_tokens_seen": 178766910, + "step": 8282, + "time_per_iteration": 3.249130964279175 + }, + { + "auxiliary_loss_clip": 0.01120271, + "auxiliary_loss_mlp": 0.01028461, + "balance_loss_clip": 0.9726305, + "balance_loss_mlp": 1.02138567, + "epoch": 0.9959718631635904, + "flos": 27782900974080.0, + "grad_norm": 3.96796896229573, + "language_loss": 0.68864608, + "learning_rate": 1.651981619819054e-10, + "loss": 0.71013331, + "num_input_tokens_seen": 178784060, + "step": 8283, + "time_per_iteration": 2.7075490951538086 + }, + { + "auxiliary_loss_clip": 0.01074246, + "auxiliary_loss_mlp": 0.01021812, + "balance_loss_clip": 0.89403683, + "balance_loss_mlp": 1.01508832, + "epoch": 0.9960921060542296, + "flos": 24024095274240.0, + "grad_norm": 7.60051081935273, + "language_loss": 0.70469755, + "learning_rate": 1.5533797692546257e-10, + "loss": 0.72565812, + "num_input_tokens_seen": 178802795, + "step": 8284, + "time_per_iteration": 2.7315452098846436 + }, + { + "auxiliary_loss_clip": 0.01136868, + "auxiliary_loss_mlp": 0.01027767, + "balance_loss_clip": 1.00728178, + "balance_loss_mlp": 1.02068901, + "epoch": 0.9962123489448687, + "flos": 18697393935360.0, + "grad_norm": 2.0483506297471865, + "language_loss": 0.84309548, + "learning_rate": 1.4578116671404296e-10, + "loss": 0.8647418, + "num_input_tokens_seen": 178821075, + "step": 8285, + "time_per_iteration": 2.682523727416992 + }, + { + "auxiliary_loss_clip": 0.01139151, + "auxiliary_loss_mlp": 0.01023286, + "balance_loss_clip": 1.01249158, + "balance_loss_mlp": 1.01670849, + "epoch": 0.9963325918355077, + "flos": 20010754823040.0, + "grad_norm": 2.217725225196971, + "language_loss": 0.71222943, + "learning_rate": 1.3652773279759777e-10, + "loss": 0.73385382, + "num_input_tokens_seen": 178837725, + "step": 8286, + "time_per_iteration": 2.6365113258361816 + }, + { + "auxiliary_loss_clip": 0.01141112, + "auxiliary_loss_mlp": 0.01023589, + "balance_loss_clip": 1.00975609, + "balance_loss_mlp": 1.01627874, + "epoch": 0.9964528347261468, + "flos": 33108488991360.0, + "grad_norm": 1.8831125292650113, + "language_loss": 0.62977052, + "learning_rate": 1.2757767657989305e-10, + "loss": 0.65141755, + "num_input_tokens_seen": 178861515, + "step": 8287, + "time_per_iteration": 2.7164864540100098 + }, + { + "auxiliary_loss_clip": 0.01138476, + "auxiliary_loss_mlp": 0.01023267, + "balance_loss_clip": 1.00997734, + "balance_loss_mlp": 1.01642466, + "epoch": 0.9965730776167859, + "flos": 23109342589440.0, + "grad_norm": 1.8965540000017145, + "language_loss": 0.87141418, + "learning_rate": 1.1893099941850948e-10, + "loss": 0.89303166, + "num_input_tokens_seen": 178880410, + "step": 8288, + "time_per_iteration": 2.697730779647827 + }, + { + "auxiliary_loss_clip": 0.01124149, + "auxiliary_loss_mlp": 0.01020848, + "balance_loss_clip": 0.97044635, + "balance_loss_mlp": 1.01379704, + "epoch": 0.996693320507425, + "flos": 22965843755520.0, + "grad_norm": 2.055821654548491, + "language_loss": 0.77263099, + "learning_rate": 1.105877026252866e-10, + "loss": 0.79408097, + "num_input_tokens_seen": 178898740, + "step": 8289, + "time_per_iteration": 3.6003365516662598 + }, + { + "auxiliary_loss_clip": 0.01164144, + "auxiliary_loss_mlp": 0.01031851, + "balance_loss_clip": 1.04680789, + "balance_loss_mlp": 1.02422118, + "epoch": 0.996813563398064, + "flos": 13222740476160.0, + "grad_norm": 1.91840955279212, + "language_loss": 0.72276938, + "learning_rate": 1.0254778746565663e-10, + "loss": 0.74472934, + "num_input_tokens_seen": 178914015, + "step": 8290, + "time_per_iteration": 2.542044162750244 + }, + { + "auxiliary_loss_clip": 0.0109159, + "auxiliary_loss_mlp": 0.01028421, + "balance_loss_clip": 0.93299937, + "balance_loss_mlp": 1.02199316, + "epoch": 0.9969338062887032, + "flos": 14647855553280.0, + "grad_norm": 1.8824169305418825, + "language_loss": 0.73343945, + "learning_rate": 9.481125515953259e-11, + "loss": 0.75463957, + "num_input_tokens_seen": 178932075, + "step": 8291, + "time_per_iteration": 2.684535264968872 + }, + { + "auxiliary_loss_clip": 0.01069297, + "auxiliary_loss_mlp": 0.01020248, + "balance_loss_clip": 0.89043796, + "balance_loss_mlp": 1.01274967, + "epoch": 0.9970540491793423, + "flos": 25735741142400.0, + "grad_norm": 1.7508393780326856, + "language_loss": 0.80125773, + "learning_rate": 8.737810688064228e-11, + "loss": 0.82215321, + "num_input_tokens_seen": 178951910, + "step": 8292, + "time_per_iteration": 2.7859508991241455 + }, + { + "auxiliary_loss_clip": 0.01089308, + "auxiliary_loss_mlp": 0.01028244, + "balance_loss_clip": 0.93512189, + "balance_loss_mlp": 1.02092457, + "epoch": 0.9971742920699813, + "flos": 21470236237440.0, + "grad_norm": 1.8214305669507367, + "language_loss": 0.79083991, + "learning_rate": 8.024834375608414e-11, + "loss": 0.81201547, + "num_input_tokens_seen": 178970500, + "step": 8293, + "time_per_iteration": 2.699547290802002 + }, + { + "auxiliary_loss_clip": 0.01059251, + "auxiliary_loss_mlp": 0.00999694, + "balance_loss_clip": 1.00875914, + "balance_loss_mlp": 0.99876422, + "epoch": 0.9972945349606205, + "flos": 72211223629440.0, + "grad_norm": 0.8219743707676126, + "language_loss": 0.62816566, + "learning_rate": 7.342196686788149e-11, + "loss": 0.64875513, + "num_input_tokens_seen": 179023665, + "step": 8294, + "time_per_iteration": 3.0799975395202637 + }, + { + "auxiliary_loss_clip": 0.0112194, + "auxiliary_loss_mlp": 0.01025229, + "balance_loss_clip": 0.97938108, + "balance_loss_mlp": 1.018121, + "epoch": 0.9974147778512595, + "flos": 19678293515520.0, + "grad_norm": 1.9437109924846916, + "language_loss": 0.69017422, + "learning_rate": 6.689897725142834e-11, + "loss": 0.71164596, + "num_input_tokens_seen": 179043140, + "step": 8295, + "time_per_iteration": 2.68523907661438 + }, + { + "auxiliary_loss_clip": 0.01119108, + "auxiliary_loss_mlp": 0.01023846, + "balance_loss_clip": 0.97056735, + "balance_loss_mlp": 1.0167768, + "epoch": 0.9975350207418986, + "flos": 15960821391360.0, + "grad_norm": 2.4088561544627196, + "language_loss": 0.88594007, + "learning_rate": 6.067937589615545e-11, + "loss": 0.90736961, + "num_input_tokens_seen": 179061215, + "step": 8296, + "time_per_iteration": 2.6149771213531494 + }, + { + "auxiliary_loss_clip": 0.01023051, + "auxiliary_loss_mlp": 0.01000589, + "balance_loss_clip": 0.93474972, + "balance_loss_mlp": 0.99961132, + "epoch": 0.9976552636325378, + "flos": 59961879768960.0, + "grad_norm": 0.7523642851220586, + "language_loss": 0.57658851, + "learning_rate": 5.476316374575241e-11, + "loss": 0.59682494, + "num_input_tokens_seen": 179124700, + "step": 8297, + "time_per_iteration": 3.221318006515503 + }, + { + "auxiliary_loss_clip": 0.01165676, + "auxiliary_loss_mlp": 0.01032112, + "balance_loss_clip": 1.04840469, + "balance_loss_mlp": 1.0244168, + "epoch": 0.9977755065231768, + "flos": 22487872452480.0, + "grad_norm": 2.73103094482899, + "language_loss": 0.72646815, + "learning_rate": 4.9150341697723476e-11, + "loss": 0.74844599, + "num_input_tokens_seen": 179144590, + "step": 8298, + "time_per_iteration": 2.6253228187561035 + }, + { + "auxiliary_loss_clip": 0.01116393, + "auxiliary_loss_mlp": 0.01028575, + "balance_loss_clip": 0.97276485, + "balance_loss_mlp": 1.02090979, + "epoch": 0.9978957494138159, + "flos": 26030280666240.0, + "grad_norm": 2.2950829949814673, + "language_loss": 0.66502064, + "learning_rate": 4.384091060338768e-11, + "loss": 0.68647027, + "num_input_tokens_seen": 179165060, + "step": 8299, + "time_per_iteration": 2.7487847805023193 + }, + { + "auxiliary_loss_clip": 0.01139347, + "auxiliary_loss_mlp": 0.01022042, + "balance_loss_clip": 1.00927579, + "balance_loss_mlp": 1.01506519, + "epoch": 0.998015992304455, + "flos": 22637835734400.0, + "grad_norm": 2.7861234339808987, + "language_loss": 0.73163104, + "learning_rate": 3.883487126810081e-11, + "loss": 0.75324488, + "num_input_tokens_seen": 179184320, + "step": 8300, + "time_per_iteration": 2.751586675643921 + }, + { + "auxiliary_loss_clip": 0.01131559, + "auxiliary_loss_mlp": 0.01024665, + "balance_loss_clip": 1.00589347, + "balance_loss_mlp": 1.01768231, + "epoch": 0.9981362351950941, + "flos": 18223444955520.0, + "grad_norm": 1.6255811703094765, + "language_loss": 0.79482758, + "learning_rate": 3.41322244516995e-11, + "loss": 0.8163898, + "num_input_tokens_seen": 179202265, + "step": 8301, + "time_per_iteration": 2.633971691131592 + }, + { + "auxiliary_loss_clip": 0.01059194, + "auxiliary_loss_mlp": 0.01026049, + "balance_loss_clip": 0.89472502, + "balance_loss_mlp": 1.01903319, + "epoch": 0.9982564780857331, + "flos": 33474095573760.0, + "grad_norm": 1.5885736616509663, + "language_loss": 0.63071281, + "learning_rate": 2.9732970866946925e-11, + "loss": 0.65156525, + "num_input_tokens_seen": 179222145, + "step": 8302, + "time_per_iteration": 2.8139536380767822 + }, + { + "auxiliary_loss_clip": 0.01082402, + "auxiliary_loss_mlp": 0.01027937, + "balance_loss_clip": 0.92917669, + "balance_loss_mlp": 1.01975369, + "epoch": 0.9983767209763723, + "flos": 15523465392000.0, + "grad_norm": 2.726330972933184, + "language_loss": 0.78106666, + "learning_rate": 2.563711118175327e-11, + "loss": 0.80217004, + "num_input_tokens_seen": 179239030, + "step": 8303, + "time_per_iteration": 2.7350945472717285 + }, + { + "auxiliary_loss_clip": 0.01087433, + "auxiliary_loss_mlp": 0.01025767, + "balance_loss_clip": 0.93225062, + "balance_loss_mlp": 1.01877546, + "epoch": 0.9984969638670114, + "flos": 19974377324160.0, + "grad_norm": 1.9448237768226535, + "language_loss": 0.84108055, + "learning_rate": 2.184464601717728e-11, + "loss": 0.86221254, + "num_input_tokens_seen": 179257345, + "step": 8304, + "time_per_iteration": 2.7621936798095703 + }, + { + "auxiliary_loss_clip": 0.01146256, + "auxiliary_loss_mlp": 0.01024833, + "balance_loss_clip": 1.01328897, + "balance_loss_mlp": 1.01761448, + "epoch": 0.9986172067576504, + "flos": 20375750874240.0, + "grad_norm": 2.3118025346001083, + "language_loss": 0.77908289, + "learning_rate": 1.8355575948758585e-11, + "loss": 0.80079383, + "num_input_tokens_seen": 179275330, + "step": 8305, + "time_per_iteration": 2.5751636028289795 + }, + { + "auxiliary_loss_clip": 0.01115773, + "auxiliary_loss_mlp": 0.0102831, + "balance_loss_clip": 0.96807569, + "balance_loss_mlp": 1.02099085, + "epoch": 0.9987374496482896, + "flos": 23727903724800.0, + "grad_norm": 2.1842940171041003, + "language_loss": 0.7333982, + "learning_rate": 1.5169901505407424e-11, + "loss": 0.754839, + "num_input_tokens_seen": 179292395, + "step": 8306, + "time_per_iteration": 3.592195510864258 + }, + { + "auxiliary_loss_clip": 0.01114098, + "auxiliary_loss_mlp": 0.01027913, + "balance_loss_clip": 0.97104943, + "balance_loss_mlp": 1.02099347, + "epoch": 0.9988576925389286, + "flos": 25044029959680.0, + "grad_norm": 1.6750762323921193, + "language_loss": 0.74601281, + "learning_rate": 1.228762317073695e-11, + "loss": 0.76743293, + "num_input_tokens_seen": 179311225, + "step": 8307, + "time_per_iteration": 2.634600877761841 + }, + { + "auxiliary_loss_clip": 0.01118351, + "auxiliary_loss_mlp": 0.01023233, + "balance_loss_clip": 0.9736315, + "balance_loss_mlp": 1.01619935, + "epoch": 0.9989779354295677, + "flos": 31285627637760.0, + "grad_norm": 2.0149170329196022, + "language_loss": 0.79472858, + "learning_rate": 9.70874138195299e-12, + "loss": 0.81614441, + "num_input_tokens_seen": 179333135, + "step": 8308, + "time_per_iteration": 3.529797077178955 + }, + { + "auxiliary_loss_clip": 0.01164325, + "auxiliary_loss_mlp": 0.01024627, + "balance_loss_clip": 1.04658127, + "balance_loss_mlp": 1.01725996, + "epoch": 0.9990981783202069, + "flos": 19573398823680.0, + "grad_norm": 1.5405389601528876, + "language_loss": 0.74713415, + "learning_rate": 7.433256530076093e-12, + "loss": 0.76902372, + "num_input_tokens_seen": 179353090, + "step": 8309, + "time_per_iteration": 2.79410982131958 + }, + { + "auxiliary_loss_clip": 0.01075312, + "auxiliary_loss_mlp": 0.01023269, + "balance_loss_clip": 0.89341414, + "balance_loss_mlp": 1.01662922, + "epoch": 0.9992184212108459, + "flos": 17199667514880.0, + "grad_norm": 2.100871424590134, + "language_loss": 0.7599318, + "learning_rate": 5.46116896038562e-12, + "loss": 0.78091758, + "num_input_tokens_seen": 179367500, + "step": 8310, + "time_per_iteration": 2.822453737258911 + }, + { + "auxiliary_loss_clip": 0.01117176, + "auxiliary_loss_mlp": 0.01028212, + "balance_loss_clip": 0.9727779, + "balance_loss_mlp": 1.02081466, + "epoch": 0.999338664101485, + "flos": 46497853681920.0, + "grad_norm": 3.205546687817824, + "language_loss": 0.61755323, + "learning_rate": 3.792478972197699e-12, + "loss": 0.63900709, + "num_input_tokens_seen": 179388085, + "step": 8311, + "time_per_iteration": 2.8591063022613525 + }, + { + "auxiliary_loss_clip": 0.01161561, + "auxiliary_loss_mlp": 0.01024574, + "balance_loss_clip": 1.04561281, + "balance_loss_mlp": 1.01794338, + "epoch": 0.9994589069921241, + "flos": 15158253859200.0, + "grad_norm": 2.2896886056771777, + "language_loss": 0.7007618, + "learning_rate": 2.4271868181990895e-12, + "loss": 0.72262311, + "num_input_tokens_seen": 179405250, + "step": 8312, + "time_per_iteration": 2.569187879562378 + }, + { + "auxiliary_loss_clip": 0.01143399, + "auxiliary_loss_mlp": 0.01025884, + "balance_loss_clip": 1.0102818, + "balance_loss_mlp": 1.01833248, + "epoch": 0.9995791498827632, + "flos": 12531460256640.0, + "grad_norm": 2.2482067679464053, + "language_loss": 0.81134689, + "learning_rate": 1.3652927060014973e-12, + "loss": 0.8330397, + "num_input_tokens_seen": 179420845, + "step": 8313, + "time_per_iteration": 2.677323818206787 + }, + { + "auxiliary_loss_clip": 0.01099648, + "auxiliary_loss_mlp": 0.01022908, + "balance_loss_clip": 0.93535769, + "balance_loss_mlp": 1.01550531, + "epoch": 0.9996993927734023, + "flos": 19245175320960.0, + "grad_norm": 2.497745550025309, + "language_loss": 0.6376729, + "learning_rate": 6.067967965872612e-13, + "loss": 0.65889847, + "num_input_tokens_seen": 179440455, + "step": 8314, + "time_per_iteration": 2.715543031692505 + }, + { + "auxiliary_loss_clip": 0.01094822, + "auxiliary_loss_mlp": 0.01025766, + "balance_loss_clip": 0.93621576, + "balance_loss_mlp": 1.01864958, + "epoch": 0.9998196356640414, + "flos": 62952804518400.0, + "grad_norm": 1.710079024742955, + "language_loss": 0.76923704, + "learning_rate": 1.5169920497548615e-13, + "loss": 0.79044294, + "num_input_tokens_seen": 179465075, + "step": 8315, + "time_per_iteration": 3.958231210708618 + }, + { + "auxiliary_loss_clip": 0.01094812, + "auxiliary_loss_mlp": 0.01013421, + "balance_loss_clip": 0.99125433, + "balance_loss_mlp": 1.00914884, + "epoch": 0.9999398785546805, + "flos": 50922375073920.0, + "grad_norm": 1.1323373732078033, + "language_loss": 0.55069089, + "learning_rate": 0.0, + "loss": 0.57177317, + "num_input_tokens_seen": 179513955, + "step": 8316, + "time_per_iteration": 3.2666239738464355 + }, + { + "epoch": 0.9999398785546805, + "num_input_tokens_seen": 179513955, + "step": 8316, + "total_flos": 6.996749092776837e+17, + "train_loss": 0.7890647420477787, + "train_runtime": 25258.8769, + "train_samples_per_second": 13.17, + "train_steps_per_second": 0.329 + } + ], + "logging_steps": 1.0, + "max_steps": 8316, + "num_input_tokens_seen": 179513955, + "num_train_epochs": 1, + "save_steps": 1664, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 6.996749092776837e+17, + "train_batch_size": 5, + "trial_name": null, + "trial_params": null +}