bitnet-1bitllm / vm_backup /logs /fp32_ref.stdout
hidude562's picture
1bitllm code (checkpoints to follow)
4754707 verified
{"event": "config", "args": {"data_dir": "/root/bitnet1/data", "ckpt_dir": "/root/bitnet1/ckpt", "log_dir": "/root/bitnet1/logs", "tag": "fp32_ref", "vocab_size": 128, "d_model": 256, "n_layers": 8, "n_heads": 8, "d_ff": 512, "seq_len": 256, "batch_size": 64, "total_steps": 30000, "eval_interval": 500, "eval_iters": 40, "lr": 0.0003, "weight_decay": 0.1, "warmup": 500, "seed": 42, "device": "cuda", "grad_clip": 1.0}, "n_params": 5345536}
FP32 params: 5,345,536 (5.35M)
{"event": "train", "step": 50, "loss": 163.4659225463867, "bpc": 235.8314758120115, "lr": 2.9999999999999997e-05, "elapsed_s": 2.4371449947357178}
{"event": "train", "step": 100, "loss": 71.65171993255615, "bpc": 103.37158101786365, "lr": 5.9999999999999995e-05, "elapsed_s": 4.937875986099243}
{"event": "train", "step": 150, "loss": 12.32454535484314, "bpc": 17.780560464643308, "lr": 8.999999999999999e-05, "elapsed_s": 7.6249165534973145}
{"event": "train", "step": 200, "loss": 3.1582606077194213, "bpc": 4.556406916591773, "lr": 0.00011999999999999999, "elapsed_s": 9.789016246795654}
{"event": "train", "step": 250, "loss": 2.5878134870529177, "bpc": 3.73342568451682, "lr": 0.00015, "elapsed_s": 11.986709833145142}
{"event": "train", "step": 300, "loss": 2.472269449234009, "bpc": 3.566730874151194, "lr": 0.00017999999999999998, "elapsed_s": 14.19581651687622}
{"event": "train", "step": 350, "loss": 2.420706830024719, "bpc": 3.4923417391227054, "lr": 0.00020999999999999998, "elapsed_s": 16.40384030342102}
{"event": "train", "step": 400, "loss": 2.3962778759002688, "bpc": 3.4570982081532566, "lr": 0.00023999999999999998, "elapsed_s": 18.61300301551819}
{"event": "train", "step": 450, "loss": 2.3901847267150877, "bpc": 3.4483076520403992, "lr": 0.00026999999999999995, "elapsed_s": 20.82123875617981}
{"event": "train", "step": 500, "loss": 2.3645445489883423, "bpc": 3.411316694786512, "lr": 0.0003, "elapsed_s": 23.031250715255737}
{"event": "eval", "step": 500, "val_loss": 2.364356428384781, "val_bpc": 3.411045294124665}
{"event": "train", "step": 550, "loss": 2.3676443481445313, "bpc": 3.415788759656898, "lr": 0.0002999978735461616, "elapsed_s": 25.77823519706726}
{"event": "train", "step": 600, "loss": 2.358785619735718, "bpc": 3.4030083161129205, "lr": 0.0002999914942449372, "elapsed_s": 28.003056049346924}
{"event": "train", "step": 650, "loss": 2.3436150217056273, "bpc": 3.381121769567589, "lr": 0.0002999808622771973, "elapsed_s": 30.21974229812622}
{"event": "train", "step": 700, "loss": 2.340182313919067, "bpc": 3.376169419067098, "lr": 0.00029996597794438706, "elapsed_s": 32.415406227111816}
{"event": "train", "step": 750, "loss": 2.318672571182251, "bpc": 3.345137419889896, "lr": 0.00029994684166851786, "elapsed_s": 34.63958263397217}
{"event": "train", "step": 800, "loss": 2.2935378074646, "bpc": 3.3088756209205243, "lr": 0.000299923453992155, "elapsed_s": 36.83741497993469}
{"event": "train", "step": 850, "loss": 2.272192988395691, "bpc": 3.2780815563011374, "lr": 0.00029989581557840283, "elapsed_s": 39.07463312149048}
{"event": "train", "step": 900, "loss": 2.245728778839111, "bpc": 3.2399017724128134, "lr": 0.00029986392721088537, "elapsed_s": 41.27412796020508}
{"event": "train", "step": 950, "loss": 2.1993922424316406, "bpc": 3.1730522811257846, "lr": 0.00029982778979372447, "elapsed_s": 43.501365423202515}
{"event": "train", "step": 1000, "loss": 2.1743079376220704, "bpc": 3.1368632789728705, "lr": 0.0002997874043515143, "elapsed_s": 45.73177194595337}
{"event": "eval", "step": 1000, "val_loss": 2.1628299236297606, "val_bpc": 3.1203040051069113}
{"event": "train", "step": 1050, "loss": 2.1251106882095336, "bpc": 3.0658866512200262, "lr": 0.0002997427720292917, "elapsed_s": 48.48419785499573}
{"event": "train", "step": 1100, "loss": 2.077954888343811, "bpc": 2.997855212604596, "lr": 0.0002996938940925045, "elapsed_s": 50.70262289047241}
{"event": "train", "step": 1150, "loss": 2.021740372180939, "bpc": 2.9167548089104476, "lr": 0.00029964077192697487, "elapsed_s": 52.92057752609253}
{"event": "train", "step": 1200, "loss": 1.9756231427192688, "bpc": 2.850221710666558, "lr": 0.0002995834070388607, "elapsed_s": 55.164682388305664}
{"event": "train", "step": 1250, "loss": 1.9222112250328065, "bpc": 2.7731646018959295, "lr": 0.0002995218010546125, "elapsed_s": 57.37662363052368}
{"event": "train", "step": 1300, "loss": 1.8768472146987916, "bpc": 2.7077181691522103, "lr": 0.00029945595572092727, "elapsed_s": 59.59226727485657}
{"event": "train", "step": 1350, "loss": 1.834372479915619, "bpc": 2.6464400799174532, "lr": 0.0002993858729046992, "elapsed_s": 61.80652213096619}
{"event": "train", "step": 1400, "loss": 1.7937206315994263, "bpc": 2.5877918599487115, "lr": 0.0002993115545929667, "elapsed_s": 64.0237078666687}
{"event": "train", "step": 1450, "loss": 1.7451805520057677, "bpc": 2.517763327834585, "lr": 0.0002992330028928557, "elapsed_s": 66.23980140686035}
{"event": "train", "step": 1500, "loss": 1.7052776837348937, "bpc": 2.4601956576629496, "lr": 0.00029915022003152054, "elapsed_s": 68.4555275440216}
{"event": "eval", "step": 1500, "val_loss": 1.6789317369461059, "val_bpc": 2.4221864908832407}
{"event": "train", "step": 1550, "loss": 1.6691168427467347, "bpc": 2.408026591694958, "lr": 0.0002990632083560803, "elapsed_s": 71.23088455200195}
{"event": "train", "step": 1600, "loss": 1.634403820037842, "bpc": 2.3579462859785725, "lr": 0.0002989719703335525, "elapsed_s": 73.46529793739319}
{"event": "train", "step": 1650, "loss": 1.601617205142975, "bpc": 2.3106451992622117, "lr": 0.00029887650855078287, "elapsed_s": 75.6921694278717}
{"event": "train", "step": 1700, "loss": 1.5679509019851685, "bpc": 2.2620749906513797, "lr": 0.00029877682571437264, "elapsed_s": 77.92502689361572}
{"event": "train", "step": 1750, "loss": 1.5340871286392213, "bpc": 2.213219892779394, "lr": 0.00029867292465060103, "elapsed_s": 80.15193104743958}
{"event": "train", "step": 1800, "loss": 1.5069920802116394, "bpc": 2.1741300007802753, "lr": 0.00029856480830534564, "elapsed_s": 82.36718034744263}
{"event": "train", "step": 1850, "loss": 1.4778501272201539, "bpc": 2.1320870497176396, "lr": 0.00029845247974399854, "elapsed_s": 84.58375287055969}
{"event": "train", "step": 1900, "loss": 1.4683940696716309, "bpc": 2.118444842386025, "lr": 0.00029833594215137997, "elapsed_s": 86.79923748970032}
{"event": "train", "step": 1950, "loss": 1.4325750350952149, "bpc": 2.066768898833199, "lr": 0.0002982151988316472, "elapsed_s": 89.01745462417603}
{"event": "train", "step": 2000, "loss": 1.4263888001441956, "bpc": 2.0578440483475897, "lr": 0.0002980902532082017, "elapsed_s": 91.24165058135986}
{"event": "eval", "step": 2000, "val_loss": 1.4163457036018372, "val_bpc": 2.04335492277076}
{"event": "train", "step": 2050, "loss": 1.3928663635253906, "bpc": 2.009481395279125, "lr": 0.00029796110882359127, "elapsed_s": 93.99497556686401}
{"event": "train", "step": 2100, "loss": 1.3763801813125611, "bpc": 1.9856968619574844, "lr": 0.0002978277693394104, "elapsed_s": 96.21410202980042}
{"event": "train", "step": 2150, "loss": 1.355199944972992, "bpc": 1.9551402400255316, "lr": 0.0002976902385361957, "elapsed_s": 98.44300818443298}
{"event": "train", "step": 2200, "loss": 1.3479076719284058, "bpc": 1.944619713867299, "lr": 0.0002975485203133194, "elapsed_s": 100.67296171188354}
{"event": "train", "step": 2250, "loss": 1.324949758052826, "bpc": 1.911498445369844, "lr": 0.0002974026186888781, "elapsed_s": 102.90808653831482}
{"event": "train", "step": 2300, "loss": 1.3111907267570495, "bpc": 1.8916483591519913, "lr": 0.0002972525377995796, "elapsed_s": 105.15161442756653}
{"event": "train", "step": 2350, "loss": 1.2990328001976013, "bpc": 1.874108178797183, "lr": 0.0002970982819006248, "elapsed_s": 107.36965894699097}
{"event": "train", "step": 2400, "loss": 1.2804954028129578, "bpc": 1.8473643675193698, "lr": 0.00029693985536558774, "elapsed_s": 109.58640193939209}
{"event": "train", "step": 2450, "loss": 1.2667696738243104, "bpc": 1.8275623263748624, "lr": 0.00029677726268629136, "elapsed_s": 111.80429100990295}
{"event": "train", "step": 2500, "loss": 1.2485546278953552, "bpc": 1.801283569943594, "lr": 0.00029661050847267997, "elapsed_s": 114.01921963691711}
{"event": "eval", "step": 2500, "val_loss": 1.2396256089210511, "val_bpc": 1.7884017185493621}
{"event": "train", "step": 2550, "loss": 1.237751326560974, "bpc": 1.7856977006832533, "lr": 0.00029643959745268873, "elapsed_s": 116.79649138450623}
{"event": "train", "step": 2600, "loss": 1.2299505066871643, "bpc": 1.77444349653644, "lr": 0.0002962645344721095, "elapsed_s": 119.02595114707947}
{"event": "train", "step": 2650, "loss": 1.2140543961524963, "bpc": 1.7515102566986516, "lr": 0.00029608532449445374, "elapsed_s": 121.2540590763092}
{"event": "train", "step": 2700, "loss": 1.2004994249343872, "bpc": 1.7319545669428928, "lr": 0.0002959019726008112, "elapsed_s": 123.48013353347778}
{"event": "train", "step": 2750, "loss": 1.1888506841659545, "bpc": 1.715148986403674, "lr": 0.0002957144839897065, "elapsed_s": 125.72106504440308}
{"event": "train", "step": 2800, "loss": 1.1732279920578004, "bpc": 1.6926102059739048, "lr": 0.0002955228639769513, "elapsed_s": 127.93673300743103}
{"event": "train", "step": 2850, "loss": 1.1711602520942688, "bpc": 1.6896270877826698, "lr": 0.00029532711799549365, "elapsed_s": 130.68140602111816}
{"event": "train", "step": 2900, "loss": 1.1559786868095399, "bpc": 1.6677247188334594, "lr": 0.0002951272515952643, "elapsed_s": 133.2275333404541}
{"event": "train", "step": 2950, "loss": 1.1527617073059082, "bpc": 1.6630835984569285, "lr": 0.0002949232704430189, "elapsed_s": 135.47186064720154}
{"event": "train", "step": 3000, "loss": 1.135857448577881, "bpc": 1.6386959082200996, "lr": 0.00029471518032217736, "elapsed_s": 137.68757605552673}
{"event": "eval", "step": 3000, "val_loss": 1.1301247000694274, "val_bpc": 1.6304253003762903}
{"event": "train", "step": 3050, "loss": 1.1226125764846802, "bpc": 1.6195875969340303, "lr": 0.00029450298713266026, "elapsed_s": 140.46228694915771}
{"event": "train", "step": 3100, "loss": 1.1245490622520447, "bpc": 1.6223813553473592, "lr": 0.00029428669689072123, "elapsed_s": 142.6821677684784}
{"event": "train", "step": 3150, "loss": 1.1065255165100099, "bpc": 1.5963788752860901, "lr": 0.00029406631572877646, "elapsed_s": 144.8976526260376}
{"event": "train", "step": 3200, "loss": 1.1038329911231994, "bpc": 1.592494382263071, "lr": 0.00029384184989523083, "elapsed_s": 147.12749552726746}
{"event": "train", "step": 3250, "loss": 1.098079104423523, "bpc": 1.5841932784556108, "lr": 0.00029361330575430075, "elapsed_s": 149.3616554737091}
{"event": "train", "step": 3300, "loss": 1.092913911342621, "bpc": 1.5767414800125594, "lr": 0.0002933806897858339, "elapsed_s": 151.56196331977844}
{"event": "train", "step": 3350, "loss": 1.0778805947303771, "bpc": 1.5550529886879616, "lr": 0.0002931440085851251, "elapsed_s": 153.79621529579163}
{"event": "train", "step": 3400, "loss": 1.0717615818977355, "bpc": 1.5462251192191736, "lr": 0.00029290326886272967, "elapsed_s": 156.02560305595398}
{"event": "train", "step": 3450, "loss": 1.0647810196876526, "bpc": 1.5361542967360702, "lr": 0.00029265847744427303, "elapsed_s": 158.25447845458984}
{"event": "train", "step": 3500, "loss": 1.0576676273345946, "bpc": 1.525891840864416, "lr": 0.00029240964127025715, "elapsed_s": 160.47252678871155}
{"event": "eval", "step": 3500, "val_loss": 1.0524882823228836, "val_bpc": 1.5184196255009674}
{"event": "train", "step": 3550, "loss": 1.0528270769119263, "bpc": 1.5189084027744595, "lr": 0.000292156767395864, "elapsed_s": 162.66703820228577}
{"event": "train", "step": 3600, "loss": 1.045215322971344, "bpc": 1.507926963111914, "lr": 0.000291899862990755, "elapsed_s": 164.05802583694458}
{"event": "train", "step": 3650, "loss": 1.0320848548412322, "bpc": 1.4889837018560514, "lr": 0.00029163893533886847, "elapsed_s": 165.44872975349426}
{"event": "train", "step": 3700, "loss": 1.0287956523895263, "bpc": 1.4842383857904955, "lr": 0.0002913739918382123, "elapsed_s": 166.83993339538574}
{"event": "train", "step": 3750, "loss": 1.0222343075275422, "bpc": 1.4747723660965488, "lr": 0.00029110504000065497, "elapsed_s": 168.2312581539154}
{"event": "train", "step": 3800, "loss": 1.0212281215190888, "bpc": 1.4733207465319411, "lr": 0.0002908320874517119, "elapsed_s": 169.6212875843048}
{"event": "train", "step": 3850, "loss": 1.0173701000213624, "bpc": 1.4677547980495282, "lr": 0.00029055514193032986, "elapsed_s": 171.01237201690674}
{"event": "train", "step": 3900, "loss": 1.0077540743350983, "bpc": 1.4538818054788942, "lr": 0.000290274211288667, "elapsed_s": 172.40345525741577}
{"event": "train", "step": 3950, "loss": 1.0017607963085176, "bpc": 1.4452353329912773, "lr": 0.0002899893034918705, "elapsed_s": 173.79506540298462}
{"event": "train", "step": 4000, "loss": 0.9893337047100067, "bpc": 1.4273068295694329, "lr": 0.00028970042661785074, "elapsed_s": 175.18660354614258}
{"event": "eval", "step": 4000, "val_loss": 0.9845516040921212, "val_bpc": 1.4204077167229774}
{"event": "train", "step": 4050, "loss": 0.9825972950458527, "bpc": 1.4175882447535613, "lr": 0.00028940758885705225, "elapsed_s": 176.87801790237427}
{"event": "train", "step": 4100, "loss": 0.9885420262813568, "bpc": 1.4261646790264408, "lr": 0.00028911079851222143, "elapsed_s": 178.26883840560913}
{"event": "train", "step": 4150, "loss": 0.9816023921966552, "bpc": 1.4161529033468578, "lr": 0.0002888100639981712, "elapsed_s": 179.65994024276733}
{"event": "train", "step": 4200, "loss": 0.976797616481781, "bpc": 1.409221077250425, "lr": 0.0002885053938415424, "elapsed_s": 181.05097031593323}
{"event": "train", "step": 4250, "loss": 0.9694454598426819, "bpc": 1.398614157327358, "lr": 0.0002881967966805619, "elapsed_s": 182.44189882278442}
{"event": "train", "step": 4300, "loss": 0.9711025750637055, "bpc": 1.4010048692389103, "lr": 0.0002878842812647981, "elapsed_s": 183.83326077461243}
{"event": "train", "step": 4350, "loss": 0.9538211047649383, "bpc": 1.376072977739609, "lr": 0.0002875678564549123, "elapsed_s": 185.22535467147827}
{"event": "train", "step": 4400, "loss": 0.9581913387775421, "bpc": 1.3823778926771169, "lr": 0.0002872475312224079, "elapsed_s": 186.61692643165588}
{"event": "train", "step": 4450, "loss": 0.9412487661838531, "bpc": 1.3579349272163004, "lr": 0.00028692331464937575, "elapsed_s": 188.008686542511}
{"event": "train", "step": 4500, "loss": 0.9362939131259919, "bpc": 1.3507865852813905, "lr": 0.000286595215928237, "elapsed_s": 189.4007625579834}
{"event": "eval", "step": 4500, "val_loss": 0.928053443133831, "val_bpc": 1.3388981000891056}
{"event": "train", "step": 4550, "loss": 0.9396201801300049, "bpc": 1.3555853741927526, "lr": 0.000286263244361482, "elapsed_s": 191.0919623374939}
{"event": "train", "step": 4600, "loss": 0.9379837608337402, "bpc": 1.3532245201892166, "lr": 0.00028592740936140697, "elapsed_s": 192.48409628868103}
{"event": "train", "step": 4650, "loss": 0.9322925508022308, "bpc": 1.3450138397001004, "lr": 0.0002855877204498469, "elapsed_s": 193.8760471343994}
{"event": "train", "step": 4700, "loss": 0.9259245896339416, "bpc": 1.3358268137020362, "lr": 0.0002852441872579056, "elapsed_s": 195.26790070533752}
{"event": "train", "step": 4750, "loss": 0.9227092552185059, "bpc": 1.3311880666860874, "lr": 0.00028489681952568286, "elapsed_s": 196.6599826812744}
{"event": "train", "step": 4800, "loss": 0.9206184005737305, "bpc": 1.3281716010588502, "lr": 0.0002845456271019979, "elapsed_s": 198.05262637138367}
{"event": "train", "step": 4850, "loss": 0.9148526608943939, "bpc": 1.3198533970164146, "lr": 0.0002841906199441104, "elapsed_s": 199.44501304626465}
{"event": "train", "step": 4900, "loss": 0.90702667593956, "bpc": 1.308562887332004, "lr": 0.0002838318081174382, "elapsed_s": 200.83704781532288}
{"event": "train", "step": 4950, "loss": 0.9060748445987702, "bpc": 1.307189684976884, "lr": 0.0002834692017952717, "elapsed_s": 202.23016238212585}
{"event": "train", "step": 5000, "loss": 0.904851666688919, "bpc": 1.3054250122722169, "lr": 0.0002831028112584857, "elapsed_s": 203.62241864204407}
{"event": "eval", "step": 5000, "val_loss": 0.9014824867248535, "val_bpc": 1.300564313046197}
{"event": "train", "step": 5050, "loss": 0.8983526277542114, "bpc": 1.29604888103057, "lr": 0.0002827326468952477, "elapsed_s": 205.3148069381714}
{"event": "train", "step": 5100, "loss": 0.9058929741382599, "bpc": 1.3069273013654215, "lr": 0.00028235871920072347, "elapsed_s": 206.70748734474182}
{"event": "train", "step": 5150, "loss": 0.8936972546577454, "bpc": 1.2893325973508103, "lr": 0.00028198103877677936, "elapsed_s": 208.09988284111023}
{"event": "train", "step": 5200, "loss": 0.8861483657360076, "bpc": 1.2784418527391976, "lr": 0.000281599616331682, "elapsed_s": 209.49239873886108}
{"event": "train", "step": 5250, "loss": 0.8812303984165192, "bpc": 1.2713467256761177, "lr": 0.00028121446267979417, "elapsed_s": 210.8850963115692}
{"event": "train", "step": 5300, "loss": 0.884240733385086, "bpc": 1.2756897210066838, "lr": 0.0002808255887412688, "elapsed_s": 212.27818894386292}
{"event": "train", "step": 5350, "loss": 0.8805515551567078, "bpc": 1.270367361871647, "lr": 0.00028043300554173896, "elapsed_s": 213.6714584827423}
{"event": "train", "step": 5400, "loss": 0.8790141725540161, "bpc": 1.2681493876147947, "lr": 0.0002800367242120053, "elapsed_s": 215.0681767463684}
{"event": "train", "step": 5450, "loss": 0.8785793209075927, "bpc": 1.2675220293009772, "lr": 0.0002796367559877205, "elapsed_s": 216.46446824073792}
{"event": "train", "step": 5500, "loss": 0.8721084070205688, "bpc": 1.2581864739261484, "lr": 0.0002792331122090709, "elapsed_s": 217.85795640945435}
{"event": "eval", "step": 5500, "val_loss": 0.8698394030332566, "val_bpc": 1.2549129931258955}
{"event": "train", "step": 5550, "loss": 0.8702216219902038, "bpc": 1.255464418519617, "lr": 0.00027882580432045455, "elapsed_s": 219.55085635185242}
{"event": "train", "step": 5600, "loss": 0.8634885442256928, "bpc": 1.2457506406188374, "lr": 0.0002784148438701571, "elapsed_s": 220.94401741027832}
{"event": "train", "step": 5650, "loss": 0.8638731503486633, "bpc": 1.2463055099651426, "lr": 0.00027800024251002436, "elapsed_s": 222.33777356147766}
{"event": "train", "step": 5700, "loss": 0.8629485666751862, "bpc": 1.24497161768453, "lr": 0.0002775820119951316, "elapsed_s": 223.73430800437927}
{"event": "train", "step": 5750, "loss": 0.857204383611679, "bpc": 1.23668451326485, "lr": 0.0002771601641834506, "elapsed_s": 225.52064418792725}
{"event": "train", "step": 5800, "loss": 0.8551169240474701, "bpc": 1.2336729457035096, "lr": 0.00027673471103551345, "elapsed_s": 227.7429256439209}
{"event": "train", "step": 5850, "loss": 0.856341460943222, "bpc": 1.2354395790103965, "lr": 0.00027630566461407305, "elapsed_s": 229.96463918685913}
{"event": "train", "step": 5900, "loss": 0.8534545075893402, "bpc": 1.2312745857234733, "lr": 0.00027587303708376156, "elapsed_s": 232.18774366378784}
{"event": "train", "step": 5950, "loss": 0.847775970697403, "bpc": 1.2230821887099705, "lr": 0.00027543684071074543, "elapsed_s": 234.41101598739624}
{"event": "train", "step": 6000, "loss": 0.8458605301380158, "bpc": 1.220318792113825, "lr": 0.00027499708786237723, "elapsed_s": 236.63152956962585}
{"event": "eval", "step": 6000, "val_loss": 0.8429894506931305, "val_bpc": 1.2161767000366908}
{"event": "train", "step": 6050, "loss": 0.8421157288551331, "bpc": 1.2149161858738955, "lr": 0.00027455379100684534, "elapsed_s": 239.394278049469}
{"event": "train", "step": 6100, "loss": 0.8454306948184968, "bpc": 1.219698670829956, "lr": 0.0002741069627128206, "elapsed_s": 241.61737775802612}
{"event": "train", "step": 6150, "loss": 0.8442359721660614, "bpc": 1.2179750503840499, "lr": 0.0002736566156490994, "elapsed_s": 243.89374589920044}
{"event": "train", "step": 6200, "loss": 0.8360489344596863, "bpc": 1.2061636516854914, "lr": 0.00027320276258424505, "elapsed_s": 246.14685988426208}
{"event": "train", "step": 6250, "loss": 0.8313269281387329, "bpc": 1.1993512365832055, "lr": 0.0002727454163862253, "elapsed_s": 248.36833095550537}
{"event": "train", "step": 6300, "loss": 0.8347238373756408, "bpc": 1.2042519406936427, "lr": 0.00027228459002204806, "elapsed_s": 250.57824063301086}
{"event": "train", "step": 6350, "loss": 0.8367522621154785, "bpc": 1.2071783390066229, "lr": 0.0002718202965573931, "elapsed_s": 252.76334285736084}
{"event": "train", "step": 6400, "loss": 0.835223000049591, "bpc": 1.2049720802079475, "lr": 0.0002713525491562421, "elapsed_s": 254.96612691879272}
{"event": "train", "step": 6450, "loss": 0.8286229956150055, "bpc": 1.1954502865403258, "lr": 0.000270881361080505, "elapsed_s": 257.1732804775238}
{"event": "train", "step": 6500, "loss": 0.8272208774089813, "bpc": 1.1934274575577546, "lr": 0.0002704067456896445, "elapsed_s": 259.39191365242004}
{"event": "eval", "step": 6500, "val_loss": 0.8222301706671715, "val_bpc": 1.1862273896908144}
{"event": "train", "step": 6550, "loss": 0.8270091056823731, "bpc": 1.1931219355379763, "lr": 0.00026992871644029685, "elapsed_s": 262.1521031856537}
{"event": "train", "step": 6600, "loss": 0.8213538467884064, "bpc": 1.1849631215767074, "lr": 0.0002694472868858904, "elapsed_s": 264.3729679584503}
{"event": "train", "step": 6650, "loss": 0.8185470283031464, "bpc": 1.1809137384673474, "lr": 0.0002689624706762615, "elapsed_s": 266.5937223434448}
{"event": "train", "step": 6700, "loss": 0.820749636888504, "bpc": 1.1840914309504622, "lr": 0.0002684742815572674, "elapsed_s": 268.8130102157593}
{"event": "train", "step": 6750, "loss": 0.8144229245185852, "bpc": 1.1749639143892494, "lr": 0.00026798273337039636, "elapsed_s": 271.0341463088989}
{"event": "train", "step": 6800, "loss": 0.8197625505924225, "bpc": 1.182667366446176, "lr": 0.0002674878400523755, "elapsed_s": 273.25510454177856}
{"event": "train", "step": 6850, "loss": 0.8142920851707458, "bpc": 1.1747751531109685, "lr": 0.0002669896156347754, "elapsed_s": 275.4739603996277}
{"event": "train", "step": 6900, "loss": 0.8090411293506622, "bpc": 1.1671996251894068, "lr": 0.0002664880742436124, "elapsed_s": 277.6888997554779}
{"event": "train", "step": 6950, "loss": 0.8127654433250427, "bpc": 1.172572674490959, "lr": 0.00026598323009894805, "elapsed_s": 279.8937795162201}
{"event": "train", "step": 7000, "loss": 0.8078807270526887, "bpc": 1.1655255185486841, "lr": 0.0002654750975144859, "elapsed_s": 282.10534024238586}
{"event": "eval", "step": 7000, "val_loss": 0.8013624206185341, "val_bpc": 1.1561215901811348}
{"event": "train", "step": 7050, "loss": 0.8072891354560852, "bpc": 1.1646720322860327, "lr": 0.0002649636908971658, "elapsed_s": 284.8301856517792}
{"event": "train", "step": 7100, "loss": 0.8078305399417878, "bpc": 1.165453113852671, "lr": 0.00026444902474675536, "elapsed_s": 287.02595233917236}
{"event": "train", "step": 7150, "loss": 0.8048895335197449, "bpc": 1.1612101384723672, "lr": 0.00026393111365543854, "elapsed_s": 289.24515080451965}
{"event": "train", "step": 7200, "loss": 0.8006132340431213, "bpc": 1.1550407424240863, "lr": 0.00026340997230740244, "elapsed_s": 291.4652976989746}
{"event": "train", "step": 7250, "loss": 0.8013011848926545, "bpc": 1.156033245703083, "lr": 0.00026288561547842075, "elapsed_s": 293.6850266456604}
{"event": "train", "step": 7300, "loss": 0.8005909812450409, "bpc": 1.1550086384226497, "lr": 0.00026235805803543466, "elapsed_s": 295.9047372341156}
{"event": "train", "step": 7350, "loss": 0.79781489610672, "bpc": 1.1510035941605086, "lr": 0.0002618273149361314, "elapsed_s": 298.12339997291565}
{"event": "train", "step": 7400, "loss": 0.7971168553829193, "bpc": 1.1499965342699427, "lr": 0.00026129340122852036, "elapsed_s": 300.3300998210907}
{"event": "train", "step": 7450, "loss": 0.7992326021194458, "bpc": 1.1530489115945066, "lr": 0.0002607563320505063, "elapsed_s": 302.5639476776123}
{"event": "train", "step": 7500, "loss": 0.7885252201557159, "bpc": 1.1376014247345294, "lr": 0.00026021612262946004, "elapsed_s": 304.78355288505554}
{"event": "eval", "step": 7500, "val_loss": 0.7933776348829269, "val_bpc": 1.1446019793978133}
{"event": "train", "step": 7550, "loss": 0.7899175786972046, "bpc": 1.1396101734974746, "lr": 0.0002596727882817869, "elapsed_s": 307.5510399341583}
{"event": "train", "step": 7600, "loss": 0.7899502992630005, "bpc": 1.1396573792954834, "lr": 0.0002591263444124923, "elapsed_s": 309.7623248100281}
{"event": "train", "step": 7650, "loss": 0.7882441759109498, "bpc": 1.137195963596335, "lr": 0.0002585768065147452, "elapsed_s": 311.9842321872711}
{"event": "train", "step": 7700, "loss": 0.7871417391300202, "bpc": 1.1356054835195943, "lr": 0.00025802419016943834, "elapsed_s": 314.20220255851746}
{"event": "train", "step": 7750, "loss": 0.7845488798618316, "bpc": 1.1318647783116558, "lr": 0.0002574685110447472, "elapsed_s": 316.4210274219513}
{"event": "train", "step": 7800, "loss": 0.7844453048706055, "bpc": 1.1317153511854536, "lr": 0.0002569097848956852, "elapsed_s": 318.63999032974243}
{"event": "train", "step": 7850, "loss": 0.785537942647934, "bpc": 1.1332916942882933, "lr": 0.0002563480275636572, "elapsed_s": 320.85941672325134}
{"event": "train", "step": 7900, "loss": 0.7855139768123627, "bpc": 1.1332571188961638, "lr": 0.0002557832549760103, "elapsed_s": 323.0790412425995}
{"event": "train", "step": 7950, "loss": 0.7834314119815826, "bpc": 1.1302526129424677, "lr": 0.00025521548314558223, "elapsed_s": 325.2968683242798}
{"event": "train", "step": 8000, "loss": 0.7834827744960785, "bpc": 1.1303267133874184, "lr": 0.00025464472817024773, "elapsed_s": 327.49710273742676}
{"event": "eval", "step": 8000, "val_loss": 0.777826452255249, "val_bpc": 1.122166365340904}
{"event": "train", "step": 8050, "loss": 0.784827527999878, "bpc": 1.1322667825985682, "lr": 0.00025407100623246135, "elapsed_s": 330.2336149215698}
{"event": "train", "step": 8100, "loss": 0.782369589805603, "bpc": 1.128720727354876, "lr": 0.0002534943335987994, "elapsed_s": 332.4352939128876}
{"event": "train", "step": 8150, "loss": 0.7800487148761749, "bpc": 1.1253724126036666, "lr": 0.0002529147266194984, "elapsed_s": 334.639511346817}
{"event": "train", "step": 8200, "loss": 0.7768098759651184, "bpc": 1.120699755768447, "lr": 0.00025233220172799143, "elapsed_s": 336.84960865974426}
{"event": "train", "step": 8250, "loss": 0.770458517074585, "bpc": 1.1115366817941685, "lr": 0.0002517467754404424, "elapsed_s": 339.07009196281433}
{"event": "train", "step": 8300, "loss": 0.7699203884601593, "bpc": 1.1107603263107761, "lr": 0.00025115846435527767, "elapsed_s": 341.2898871898651}
{"event": "train", "step": 8350, "loss": 0.7735969769954681, "bpc": 1.1160645223580554, "lr": 0.0002505672851527154, "elapsed_s": 343.50892090797424}
{"event": "train", "step": 8400, "loss": 0.7777754187583923, "bpc": 1.1220927395680695, "lr": 0.00024997325459429276, "elapsed_s": 345.7277133464813}
{"event": "train", "step": 8450, "loss": 0.7711277782917023, "bpc": 1.112502221633163, "lr": 0.0002493763895223905, "elapsed_s": 348.1439759731293}
{"event": "train", "step": 8500, "loss": 0.7675447964668274, "bpc": 1.1073330715228207, "lr": 0.0002487767068597558, "elapsed_s": 351.013619184494}
{"event": "eval", "step": 8500, "val_loss": 0.7689894124865532, "val_bpc": 1.1094172118904677}
{"event": "train", "step": 8550, "loss": 0.7678859496116638, "bpc": 1.1078252514730598, "lr": 0.00024817422360902175, "elapsed_s": 353.7503750324249}
{"event": "train", "step": 8600, "loss": 0.7727355182170867, "bpc": 1.1148217000505543, "lr": 0.000247568956852226, "elapsed_s": 355.9521954059601}
{"event": "train", "step": 8650, "loss": 0.7715871381759644, "bpc": 1.1131649378601711, "lr": 0.00024696092375032624, "elapsed_s": 358.16577410697937}
{"event": "train", "step": 8700, "loss": 0.7692658412456512, "bpc": 1.1098160142903777, "lr": 0.00024635014154271337, "elapsed_s": 360.38657450675964}
{"event": "train", "step": 8750, "loss": 0.763764135837555, "bpc": 1.1018787311816851, "lr": 0.000245736627546723, "elapsed_s": 362.60721921920776}
{"event": "train", "step": 8800, "loss": 0.7671374869346619, "bpc": 1.1067454480806587, "lr": 0.0002451203991571444, "elapsed_s": 364.82780742645264}
{"event": "train", "step": 8850, "loss": 0.7651527190208435, "bpc": 1.1038820332540773, "lr": 0.0002445014738457271, "elapsed_s": 367.047744512558}
{"event": "train", "step": 8900, "loss": 0.7611311006546021, "bpc": 1.098080064380753, "lr": 0.00024387986916068596, "elapsed_s": 369.26587748527527}
{"event": "train", "step": 8950, "loss": 0.7600489842891693, "bpc": 1.0965189004666782, "lr": 0.00024325560272620315, "elapsed_s": 371.4855275154114}
{"event": "train", "step": 9000, "loss": 0.7633783698081971, "bpc": 1.101322188444187, "lr": 0.00024262869224192876, "elapsed_s": 373.70388865470886}
{"event": "eval", "step": 9000, "val_loss": 0.7597151532769203, "val_bpc": 1.0960372841208117}
{"event": "train", "step": 9050, "loss": 0.7570063245296478, "bpc": 1.092129270320504, "lr": 0.00024199915548247893, "elapsed_s": 376.4630079269409}
{"event": "train", "step": 9100, "loss": 0.7625630331039429, "bpc": 1.1001459062243049, "lr": 0.00024136701029693166, "elapsed_s": 378.68208932876587}
{"event": "train", "step": 9150, "loss": 0.7580311977863312, "bpc": 1.0936078498854611, "lr": 0.00024073227460832112, "elapsed_s": 380.90080070495605}
{"event": "train", "step": 9200, "loss": 0.7556259000301361, "bpc": 1.090137738740737, "lr": 0.00024009496641312912, "elapsed_s": 383.11805272102356}
{"event": "train", "step": 9250, "loss": 0.7527533042430877, "bpc": 1.085993459044284, "lr": 0.0002394551037807752, "elapsed_s": 385.32781291007996}
{"event": "train", "step": 9300, "loss": 0.7492916333675385, "bpc": 1.080999323638939, "lr": 0.00023881270485310395, "elapsed_s": 387.5384123325348}
{"event": "train", "step": 9350, "loss": 0.7575507283210754, "bpc": 1.0929146789706379, "lr": 0.00023816778784387094, "elapsed_s": 389.74356341362}
{"event": "train", "step": 9400, "loss": 0.753875777721405, "bpc": 1.0876128459649816, "lr": 0.00023752037103822617, "elapsed_s": 391.9276180267334}
{"event": "train", "step": 9450, "loss": 0.7515290248394012, "bpc": 1.0842271972199227, "lr": 0.00023687047279219567, "elapsed_s": 394.1312417984009}
{"event": "train", "step": 9500, "loss": 0.7517926394939423, "bpc": 1.0846075127747348, "lr": 0.00023621811153216104, "elapsed_s": 396.35083079338074}
{"event": "eval", "step": 9500, "val_loss": 0.7505874082446098, "val_bpc": 1.0828687316281984}
{"event": "train", "step": 9550, "loss": 0.7488784301280975, "bpc": 1.0804031973745185, "lr": 0.00023556330575433696, "elapsed_s": 399.10921597480774}
{"event": "train", "step": 9600, "loss": 0.7543759775161744, "bpc": 1.088334481728349, "lr": 0.00023490607402424691, "elapsed_s": 401.32899832725525}
{"event": "train", "step": 9650, "loss": 0.753570955991745, "bpc": 1.0871730811672458, "lr": 0.0002342464349761968, "elapsed_s": 403.5479824542999}
{"event": "train", "step": 9700, "loss": 0.7471367335319519, "bpc": 1.077890460332526, "lr": 0.00023358440731274625, "elapsed_s": 405.76727056503296}
{"event": "train", "step": 9750, "loss": 0.7490184056758881, "bpc": 1.0806051394031615, "lr": 0.00023292000980417897, "elapsed_s": 407.9863531589508}
{"event": "train", "step": 9800, "loss": 0.744343147277832, "bpc": 1.0738601672974115, "lr": 0.00023225326128797, "elapsed_s": 410.20459151268005}
{"event": "train", "step": 9850, "loss": 0.7441506648063659, "bpc": 1.0735824737903694, "lr": 0.0002315841806682519, "elapsed_s": 412.4154441356659}
{"event": "train", "step": 9900, "loss": 0.7451907968521119, "bpc": 1.0750830671346367, "lr": 0.00023091278691527868, "elapsed_s": 414.6082193851471}
{"event": "train", "step": 9950, "loss": 0.7440228462219238, "bpc": 1.0733980705524613, "lr": 0.000230239099064888, "elapsed_s": 416.802533864975}
{"event": "train", "step": 10000, "loss": 0.7378709733486175, "bpc": 1.064522794065963, "lr": 0.00022956313621796135, "elapsed_s": 418.9914228916168}
{"event": "eval", "step": 10000, "val_loss": 0.7359327107667923, "val_bpc": 1.0617264722512232}
{"event": "train", "step": 10050, "loss": 0.7413612163066864, "bpc": 1.0695581502730667, "lr": 0.00022888491753988264, "elapsed_s": 421.7475848197937}
{"event": "train", "step": 10100, "loss": 0.7381374251842499, "bpc": 1.0649072028078654, "lr": 0.00022820446225999474, "elapsed_s": 423.96475863456726}
{"event": "train", "step": 10150, "loss": 0.7452303981781006, "bpc": 1.0751401997712533, "lr": 0.0002275217896710541, "elapsed_s": 426.1821885108948}
{"event": "train", "step": 10200, "loss": 0.737699693441391, "bpc": 1.0642756893932033, "lr": 0.00022683691912868411, "elapsed_s": 428.3988654613495}
{"event": "train", "step": 10250, "loss": 0.7375219821929931, "bpc": 1.0640193062564296, "lr": 0.00022614987005082596, "elapsed_s": 430.6175534725189}
{"event": "train", "step": 10300, "loss": 0.7374185025691986, "bpc": 1.0638700167163482, "lr": 0.00022546066191718835, "elapsed_s": 432.83532881736755}
{"event": "train", "step": 10350, "loss": 0.7342497825622558, "bpc": 1.0592985200763663, "lr": 0.00022476931426869496, "elapsed_s": 435.05255126953125}
{"event": "train", "step": 10400, "loss": 0.737616798877716, "bpc": 1.0641560978172728, "lr": 0.00022407584670693063, "elapsed_s": 437.26942801475525}
{"event": "train", "step": 10450, "loss": 0.7334369575977325, "bpc": 1.0581258615309377, "lr": 0.0002233802788935855, "elapsed_s": 439.48742294311523}
{"event": "train", "step": 10500, "loss": 0.7326447784900665, "bpc": 1.056982988660812, "lr": 0.0002226826305498975, "elapsed_s": 441.70550632476807}
{"event": "eval", "step": 10500, "val_loss": 0.7372956916689872, "val_bpc": 1.0636928380396462}
{"event": "train", "step": 10550, "loss": 0.7317908191680909, "bpc": 1.055750985781877, "lr": 0.00022198292145609328, "elapsed_s": 444.4643075466156}
{"event": "train", "step": 10600, "loss": 0.7306818628311157, "bpc": 1.0541510999739605, "lr": 0.00022128117145082737, "elapsed_s": 446.685035943985}
{"event": "train", "step": 10650, "loss": 0.7343186259269714, "bpc": 1.0593978400572395, "lr": 0.0002205774004306196, "elapsed_s": 448.9029266834259}
{"event": "train", "step": 10700, "loss": 0.7329383170604706, "bpc": 1.0574064753006438, "lr": 0.00021987162834929123, "elapsed_s": 451.1214828491211}
{"event": "train", "step": 10750, "loss": 0.7333087372779846, "bpc": 1.0579408787114961, "lr": 0.00021916387521739886, "elapsed_s": 453.3407266139984}
{"event": "train", "step": 10800, "loss": 0.7322798943519593, "bpc": 1.0564565721242658, "lr": 0.0002184541611016674, "elapsed_s": 455.5586082935333}
{"event": "train", "step": 10850, "loss": 0.7268372821807861, "bpc": 1.0486045425354322, "lr": 0.00021774250612442087, "elapsed_s": 457.7694606781006}
{"event": "train", "step": 10900, "loss": 0.7275426352024078, "bpc": 1.0496221518418019, "lr": 0.00021702893046301208, "elapsed_s": 459.95273876190186}
{"event": "train", "step": 10950, "loss": 0.7296254694461822, "bpc": 1.052627046476289, "lr": 0.00021631345434925046, "elapsed_s": 462.1493721008301}
{"event": "train", "step": 11000, "loss": 0.7288564026355744, "bpc": 1.0515175176025129, "lr": 0.00021559609806882831, "elapsed_s": 464.35266160964966}
{"event": "eval", "step": 11000, "val_loss": 0.7217931687831879, "val_bpc": 1.0413274251510358}
{"event": "train", "step": 11050, "loss": 0.7205509471893311, "bpc": 1.0395352782178935, "lr": 0.00021487688196074583, "elapsed_s": 467.1082444190979}
{"event": "train", "step": 11100, "loss": 0.7285466849803924, "bpc": 1.051070689477306, "lr": 0.00021415582641673452, "elapsed_s": 469.32628893852234}
{"event": "train", "step": 11150, "loss": 0.7279361641407013, "bpc": 1.0501898940895242, "lr": 0.0002134329518806787, "elapsed_s": 471.5453495979309}
{"event": "train", "step": 11200, "loss": 0.7239388132095337, "bpc": 1.044422935724436, "lr": 0.00021270827884803606, "elapsed_s": 473.8811056613922}
{"event": "train", "step": 11250, "loss": 0.725095899105072, "bpc": 1.0460922578078116, "lr": 0.0002119818278652567, "elapsed_s": 476.8183345794678}
{"event": "train", "step": 11300, "loss": 0.7258107721805572, "bpc": 1.0471236016486791, "lr": 0.00021125361952920038, "elapsed_s": 479.0186114311218}
{"event": "train", "step": 11350, "loss": 0.725304194688797, "bpc": 1.0463927648134907, "lr": 0.00021052367448655258, "elapsed_s": 481.22243094444275}
{"event": "train", "step": 11400, "loss": 0.7170698893070221, "bpc": 1.0345131732740387, "lr": 0.00020979201343323907, "elapsed_s": 483.4423849582672}
{"event": "train", "step": 11450, "loss": 0.7241705429553985, "bpc": 1.0447572510796215, "lr": 0.00020905865711383934, "elapsed_s": 485.6618025302887}
{"event": "train", "step": 11500, "loss": 0.7212023794651031, "bpc": 1.0404750963316247, "lr": 0.0002083236263209981, "elapsed_s": 487.8801646232605}
{"event": "eval", "step": 11500, "val_loss": 0.721884223818779, "val_bpc": 1.041458789799331}
{"event": "train", "step": 11550, "loss": 0.7189889764785766, "bpc": 1.037281830819474, "lr": 0.00020758694189483608, "elapsed_s": 490.66240549087524}
{"event": "train", "step": 11600, "loss": 0.7201290726661682, "bpc": 1.0389266419354488, "lr": 0.00020684862472235895, "elapsed_s": 492.8560745716095}
{"event": "train", "step": 11650, "loss": 0.7175399446487427, "bpc": 1.0351913197844824, "lr": 0.00020610869573686524, "elapsed_s": 495.0748701095581}
{"event": "train", "step": 11700, "loss": 0.719339655637741, "bpc": 1.0377877539033438, "lr": 0.0002053671759173526, "elapsed_s": 497.2938132286072}
{"event": "train", "step": 11750, "loss": 0.7170610749721527, "bpc": 1.034500456876834, "lr": 0.00020462408628792332, "elapsed_s": 499.5130708217621}
{"event": "train", "step": 11800, "loss": 0.7155180013179779, "bpc": 1.0322742721682296, "lr": 0.00020387944791718795, "elapsed_s": 501.7314054965973}
{"event": "train", "step": 11850, "loss": 0.7188120698928833, "bpc": 1.0370266085655937, "lr": 0.00020313328191766815, "elapsed_s": 503.949205160141}
{"event": "train", "step": 11900, "loss": 0.7188682770729065, "bpc": 1.0371076983854755, "lr": 0.00020238560944519794, "elapsed_s": 506.15888142585754}
{"event": "train", "step": 11950, "loss": 0.7170502853393554, "bpc": 1.0344848907271043, "lr": 0.000201636451698324, "elapsed_s": 508.37220191955566}
{"event": "train", "step": 12000, "loss": 0.7128110432624817, "bpc": 1.0283689572056707, "lr": 0.0002008858299177045, "elapsed_s": 510.5655951499939}
{"event": "eval", "step": 12000, "val_loss": 0.7193452656269074, "val_bpc": 1.0377958474068933}
{"event": "train", "step": 12050, "loss": 0.7195314264297485, "bpc": 1.0380644206739602, "lr": 0.00020013376538550708, "elapsed_s": 513.3073136806488}
{"event": "train", "step": 12100, "loss": 0.7148631048202515, "bpc": 1.0313294562386641, "lr": 0.00019938027942480516, "elapsed_s": 515.5206482410431}
{"event": "train", "step": 12150, "loss": 0.7142130088806152, "bpc": 1.0303915660504488, "lr": 0.0001986253933989736, "elapsed_s": 517.7223644256592}
{"event": "train", "step": 12200, "loss": 0.7113348543643951, "bpc": 1.0262392668029858, "lr": 0.00019786912871108285, "elapsed_s": 519.941456079483}
{"event": "train", "step": 12250, "loss": 0.7156886911392212, "bpc": 1.0325205255268675, "lr": 0.00019711150680329232, "elapsed_s": 522.1604034900665}
{"event": "train", "step": 12300, "loss": 0.7105096101760864, "bpc": 1.0250486911049905, "lr": 0.0001963525491562421, "elapsed_s": 524.3785772323608}
{"event": "train", "step": 12350, "loss": 0.7078179442882537, "bpc": 1.0211654380768842, "lr": 0.00019559227728844423, "elapsed_s": 526.5982975959778}
{"event": "train", "step": 12400, "loss": 0.7077540898323059, "bpc": 1.0210733155699496, "lr": 0.00019483071275567252, "elapsed_s": 528.8180422782898}
{"event": "train", "step": 12450, "loss": 0.7128656113147736, "bpc": 1.028447682264103, "lr": 0.0001940678771503513, "elapsed_s": 531.0390801429749}
{"event": "train", "step": 12500, "loss": 0.7129365801811218, "bpc": 1.0285500686956413, "lr": 0.00019330379210094313, "elapsed_s": 533.2586417198181}
{"event": "eval", "step": 12500, "val_loss": 0.7101022094488144, "val_bpc": 1.0244609360961006}
{"event": "train", "step": 12550, "loss": 0.7066700649261475, "bpc": 1.0195093982136347, "lr": 0.00019253847927133593, "elapsed_s": 536.0163152217865}
{"event": "train", "step": 12600, "loss": 0.7122836530208587, "bpc": 1.027608093919468, "lr": 0.00019177196036022838, "elapsed_s": 538.2329723834991}
{"event": "train", "step": 12650, "loss": 0.7100833916664123, "bpc": 1.0244337877747485, "lr": 0.0001910042571005149, "elapsed_s": 540.4528465270996}
{"event": "train", "step": 12700, "loss": 0.7079520082473755, "bpc": 1.0213588514858711, "lr": 0.00019023539125866942, "elapsed_s": 542.6731474399567}
{"event": "train", "step": 12750, "loss": 0.705379912853241, "bpc": 1.01764810221606, "lr": 0.00018946538463412814, "elapsed_s": 544.8946666717529}
{"event": "train", "step": 12800, "loss": 0.7049469268321991, "bpc": 1.0170234354307286, "lr": 0.00018869425905867175, "elapsed_s": 547.1150884628296}
{"event": "train", "step": 12850, "loss": 0.7093304896354675, "bpc": 1.0233475797484293, "lr": 0.00018792203639580603, "elapsed_s": 549.3327882289886}
{"event": "train", "step": 12900, "loss": 0.7048114442825317, "bpc": 1.0168279754281966, "lr": 0.00018714873854014225, "elapsed_s": 551.5512113571167}
{"event": "train", "step": 12950, "loss": 0.7039293611049652, "bpc": 1.0155553984022696, "lr": 0.00018637438741677631, "elapsed_s": 553.7687392234802}
{"event": "train", "step": 13000, "loss": 0.7047434628009797, "bpc": 1.0167298988818891, "lr": 0.00018559900498066724, "elapsed_s": 555.9860436916351}
{"event": "eval", "step": 13000, "val_loss": 0.7006173297762871, "val_bpc": 1.010777147229117}
{"event": "train", "step": 13050, "loss": 0.7077050852775574, "bpc": 1.021002616941833, "lr": 0.00018482261321601433, "elapsed_s": 558.8123300075531}
{"event": "train", "step": 13100, "loss": 0.6987000477313995, "bpc": 1.008011093930972, "lr": 0.00018404523413563425, "elapsed_s": 561.0218462944031}
{"event": "train", "step": 13150, "loss": 0.7037429201602936, "bpc": 1.0152864209759733, "lr": 0.00018326688978033678, "elapsed_s": 563.2095572948456}
{"event": "train", "step": 13200, "loss": 0.7023246681690216, "bpc": 1.0132403158614343, "lr": 0.0001824876022182998, "elapsed_s": 565.4382774829865}
{"event": "train", "step": 13250, "loss": 0.7028184401988983, "bpc": 1.013952678320267, "lr": 0.00018170739354444364, "elapsed_s": 567.6273536682129}
{"event": "train", "step": 13300, "loss": 0.704573575258255, "bpc": 1.0164848029664915, "lr": 0.0001809262858798047, "elapsed_s": 569.8456773757935}
{"event": "train", "step": 13350, "loss": 0.7024375867843627, "bpc": 1.013403222987811, "lr": 0.00018014430137090824, "elapsed_s": 572.064101934433}
{"event": "train", "step": 13400, "loss": 0.7014937126636505, "bpc": 1.012041500474636, "lr": 0.00017936146218914037, "elapsed_s": 574.2832746505737}
{"event": "train", "step": 13450, "loss": 0.7061855316162109, "bpc": 1.0188103644102438, "lr": 0.00017857779053011952, "elapsed_s": 576.4891984462738}
{"event": "train", "step": 13500, "loss": 0.6988632512092591, "bpc": 1.008246546779136, "lr": 0.00017779330861306713, "elapsed_s": 578.6794307231903}
{"event": "eval", "step": 13500, "val_loss": 0.7056420505046844, "val_bpc": 1.0180262869058276}
{"event": "train", "step": 13550, "loss": 0.6994288980960846, "bpc": 1.0090626027376532, "lr": 0.00017700803868017764, "elapsed_s": 581.4143290519714}
{"event": "train", "step": 13600, "loss": 0.6967456352710724, "bpc": 1.0051914727666065, "lr": 0.00017622200299598792, "elapsed_s": 583.6117870807648}
{"event": "train", "step": 13650, "loss": 0.6999799799919129, "bpc": 1.0098576458558886, "lr": 0.00017543522384674595, "elapsed_s": 585.8309578895569}
{"event": "train", "step": 13700, "loss": 0.6983976912498474, "bpc": 1.0075748857344562, "lr": 0.0001746477235397789, "elapsed_s": 588.0512909889221}
{"event": "train", "step": 13750, "loss": 0.7010157155990601, "bpc": 1.011351896479992, "lr": 0.0001738595244028608, "elapsed_s": 590.2707307338715}
{"event": "train", "step": 13800, "loss": 0.6986307597160339, "bpc": 1.0079111324548111, "lr": 0.0001730706487835794, "elapsed_s": 592.4914541244507}
{"event": "train", "step": 13850, "loss": 0.6943267500400543, "bpc": 1.0017017590393373, "lr": 0.0001722811190487025, "elapsed_s": 594.7112379074097}
{"event": "train", "step": 13900, "loss": 0.6963407075405121, "bpc": 1.004607285537809, "lr": 0.00017149095758354395, "elapsed_s": 596.905523777008}
{"event": "train", "step": 13950, "loss": 0.6935796201229095, "bpc": 1.0006238784129726, "lr": 0.00017070018679132885, "elapsed_s": 599.1061570644379}
{"event": "train", "step": 14000, "loss": 0.6972282326221466, "bpc": 1.0058877135717474, "lr": 0.00016990882909255828, "elapsed_s": 601.9003636837006}
{"event": "eval", "step": 14000, "val_loss": 0.6953574046492577, "val_bpc": 1.0031886793329043}
{"event": "train", "step": 14050, "loss": 0.693741946220398, "bpc": 1.0008580654688262, "lr": 0.00016911690692437384, "elapsed_s": 604.8396258354187}
{"event": "train", "step": 14100, "loss": 0.6951301991939545, "bpc": 1.0028608911492753, "lr": 0.00016832444273992127, "elapsed_s": 607.0583777427673}
{"event": "train", "step": 14150, "loss": 0.6945300567150116, "bpc": 1.0019950685710777, "lr": 0.00016753145900771409, "elapsed_s": 609.2745962142944}
{"event": "train", "step": 14200, "loss": 0.6922418916225433, "bpc": 0.9986939441394386, "lr": 0.0001667379782109962, "elapsed_s": 611.4881608486176}
{"event": "train", "step": 14250, "loss": 0.6932108986377716, "bpc": 1.000091925754895, "lr": 0.0001659440228471048, "elapsed_s": 613.67506980896}
{"event": "train", "step": 14300, "loss": 0.6942260313034058, "bpc": 1.0015564526174499, "lr": 0.0001651496154268323, "elapsed_s": 615.8645217418671}
{"event": "train", "step": 14350, "loss": 0.6898158633708954, "bpc": 0.9951939252117296, "lr": 0.00016435477847378816, "elapsed_s": 618.0673310756683}
{"event": "train", "step": 14400, "loss": 0.6916474711894989, "bpc": 0.9978363767284822, "lr": 0.0001635595345237602, "elapsed_s": 620.2860713005066}
{"event": "train", "step": 14450, "loss": 0.6898594522476196, "bpc": 0.9952568106680174, "lr": 0.0001627639061240758, "elapsed_s": 622.502560377121}
{"event": "train", "step": 14500, "loss": 0.691207115650177, "bpc": 0.9972010779756746, "lr": 0.00016196791583296245, "elapsed_s": 624.7071566581726}
{"event": "eval", "step": 14500, "val_loss": 0.6923780485987663, "val_bpc": 0.9988903771338179}
{"event": "train", "step": 14550, "loss": 0.6945219218730927, "bpc": 1.001983332474983, "lr": 0.00016117158621890832, "elapsed_s": 627.4376351833344}
{"event": "train", "step": 14600, "loss": 0.6919592761993408, "bpc": 0.9982862162699055, "lr": 0.0001603749398600222, "elapsed_s": 629.6324374675751}
{"event": "train", "step": 14650, "loss": 0.6945308220386505, "bpc": 1.0019961726996962, "lr": 0.0001595779993433936, "elapsed_s": 631.843977689743}
{"event": "train", "step": 14700, "loss": 0.6922227740287781, "bpc": 0.9986663632817196, "lr": 0.00015878078726445206, "elapsed_s": 634.0507185459137}
{"event": "train", "step": 14750, "loss": 0.6884810554981232, "bpc": 0.9932682045131416, "lr": 0.0001579833262263268, "elapsed_s": 636.2685289382935}
{"event": "train", "step": 14800, "loss": 0.6865290832519532, "bpc": 0.9904521038336391, "lr": 0.00015718563883920553, "elapsed_s": 638.4864988327026}
{"event": "train", "step": 14850, "loss": 0.6857313406467438, "bpc": 0.9893012045331978, "lr": 0.00015638774771969377, "elapsed_s": 640.7010593414307}
{"event": "train", "step": 14900, "loss": 0.6832743525505066, "bpc": 0.9857565199912331, "lr": 0.00015558967549017319, "elapsed_s": 642.8994829654694}
{"event": "train", "step": 14950, "loss": 0.6868806481361389, "bpc": 0.9909593047486046, "lr": 0.00015479144477816066, "elapsed_s": 645.0870914459229}
{"event": "train", "step": 15000, "loss": 0.6829348301887512, "bpc": 0.9852666927636577, "lr": 0.0001539930782156662, "elapsed_s": 647.2802119255066}
{"event": "eval", "step": 15000, "val_loss": 0.6859046280384063, "val_bpc": 0.9895512053937979}
{"event": "train", "step": 15050, "loss": 0.688316662311554, "bpc": 0.9930310352781222, "lr": 0.00015319459843855178, "elapsed_s": 650.0370886325836}
{"event": "train", "step": 15100, "loss": 0.6854967963695526, "bpc": 0.9889628286676252, "lr": 0.0001523960280858892, "elapsed_s": 652.2574772834778}
{"event": "train", "step": 15150, "loss": 0.686109836101532, "bpc": 0.9898472580488197, "lr": 0.00015159738979931837, "elapsed_s": 654.4735023975372}
{"event": "train", "step": 15200, "loss": 0.6871518194675446, "bpc": 0.9913505222836548, "lr": 0.00015079870622240524, "elapsed_s": 656.6901249885559}
{"event": "train", "step": 15250, "loss": 0.6824443542957306, "bpc": 0.9845590856251213, "lr": 0.00015, "elapsed_s": 658.909316778183}
{"event": "train", "step": 15300, "loss": 0.6817438614368438, "bpc": 0.9835484880514272, "lr": 0.00014920129377759474, "elapsed_s": 661.127813577652}
{"event": "train", "step": 15350, "loss": 0.6802359688282013, "bpc": 0.9813730588627455, "lr": 0.00014840261020068166, "elapsed_s": 663.3452496528625}
{"event": "train", "step": 15400, "loss": 0.6803019726276398, "bpc": 0.9814682822168753, "lr": 0.00014760397191411077, "elapsed_s": 665.5490484237671}
{"event": "train", "step": 15450, "loss": 0.6815784847736359, "bpc": 0.9833098999595384, "lr": 0.00014680540156144823, "elapsed_s": 667.7567200660706}
{"event": "train", "step": 15500, "loss": 0.6846741676330567, "bpc": 0.9877760262689897, "lr": 0.00014600692178433377, "elapsed_s": 669.9577333927155}
{"event": "eval", "step": 15500, "val_loss": 0.6830173075199127, "val_bpc": 0.9853856824003102}
{"event": "train", "step": 15550, "loss": 0.6832853353023529, "bpc": 0.9857723647528571, "lr": 0.00014520855522183937, "elapsed_s": 672.7305936813354}
{"event": "train", "step": 15600, "loss": 0.681345636844635, "bpc": 0.9829739714070875, "lr": 0.00014441032450982676, "elapsed_s": 674.9180345535278}
{"event": "train", "step": 15650, "loss": 0.6805417370796204, "bpc": 0.9818141892027292, "lr": 0.0001436122522803062, "elapsed_s": 677.1118943691254}
{"event": "train", "step": 15700, "loss": 0.6843041062355042, "bpc": 0.9872421405259163, "lr": 0.00014281436116079447, "elapsed_s": 679.3314390182495}
{"event": "train", "step": 15750, "loss": 0.6809757220745086, "bpc": 0.9824402972026747, "lr": 0.00014201667377367324, "elapsed_s": 681.5506646633148}
{"event": "train", "step": 15800, "loss": 0.6792441964149475, "bpc": 0.9799422337204537, "lr": 0.00014121921273554792, "elapsed_s": 683.7691731452942}
{"event": "train", "step": 15850, "loss": 0.6830677342414856, "bpc": 0.9854584327814517, "lr": 0.0001404220006566064, "elapsed_s": 685.9878735542297}
{"event": "train", "step": 15900, "loss": 0.6805819642543792, "bpc": 0.9818722247482627, "lr": 0.0001396250601399778, "elapsed_s": 688.2066314220428}
{"event": "train", "step": 15950, "loss": 0.6793185377120972, "bpc": 0.9800494855411849, "lr": 0.0001388284137810917, "elapsed_s": 690.4196677207947}
{"event": "train", "step": 16000, "loss": 0.683238549232483, "bpc": 0.9857048667218731, "lr": 0.0001380320841670375, "elapsed_s": 692.6075716018677}
{"event": "eval", "step": 16000, "val_loss": 0.6842514246702194, "val_bpc": 0.9871661370929337}
{"event": "train", "step": 16050, "loss": 0.676483862400055, "bpc": 0.9759599135259712, "lr": 0.00013723609387592418, "elapsed_s": 695.35209608078}
{"event": "train", "step": 16100, "loss": 0.6765139126777648, "bpc": 0.9760032669126006, "lr": 0.0001364404654762398, "elapsed_s": 697.5354986190796}
{"event": "train", "step": 16150, "loss": 0.6791763865947723, "bpc": 0.9798444048291635, "lr": 0.00013564522152621186, "elapsed_s": 699.7522554397583}
{"event": "train", "step": 16200, "loss": 0.6762365067005157, "bpc": 0.9756030546849104, "lr": 0.00013485038457316766, "elapsed_s": 701.9721829891205}
{"event": "train", "step": 16250, "loss": 0.6735934066772461, "bpc": 0.9717898673887657, "lr": 0.0001340559771528952, "elapsed_s": 704.1900219917297}
{"event": "train", "step": 16300, "loss": 0.6760942029953003, "bpc": 0.975397753835096, "lr": 0.00013326202178900377, "elapsed_s": 706.4089648723602}
{"event": "train", "step": 16350, "loss": 0.6748483550548553, "bpc": 0.9736003751897142, "lr": 0.00013246854099228591, "elapsed_s": 708.6280567646027}
{"event": "train", "step": 16400, "loss": 0.6752672028541565, "bpc": 0.9742046448326533, "lr": 0.00013167555726007865, "elapsed_s": 710.8465912342072}
{"event": "train", "step": 16450, "loss": 0.6797698009014129, "bpc": 0.9807005207065465, "lr": 0.00013088309307562613, "elapsed_s": 713.0602557659149}
{"event": "train", "step": 16500, "loss": 0.6785700511932373, "bpc": 0.9789696477522536, "lr": 0.0001300911709074417, "elapsed_s": 715.2492253780365}
{"event": "eval", "step": 16500, "val_loss": 0.6763619661331177, "val_bpc": 0.975784054386158}
{"event": "train", "step": 16550, "loss": 0.6710870206356049, "bpc": 0.9681739166759367, "lr": 0.00012929981320867117, "elapsed_s": 717.9964730739594}
{"event": "train", "step": 16600, "loss": 0.675784386396408, "bpc": 0.9749507829642889, "lr": 0.00012850904241645602, "elapsed_s": 720.1958997249603}
{"event": "train", "step": 16650, "loss": 0.6750325858592987, "bpc": 0.9738661640576637, "lr": 0.00012771888095129748, "elapsed_s": 722.3955993652344}
{"event": "train", "step": 16700, "loss": 0.6754707324504853, "bpc": 0.9744982759719509, "lr": 0.0001269293512164206, "elapsed_s": 724.6146168708801}
{"event": "train", "step": 16750, "loss": 0.6719318127632141, "bpc": 0.9693926940890205, "lr": 0.0001261404755971392, "elapsed_s": 727.0392260551453}
{"event": "train", "step": 16800, "loss": 0.674993394613266, "bpc": 0.9738096230413661, "lr": 0.0001253522764602211, "elapsed_s": 729.8986365795135}
{"event": "train", "step": 16850, "loss": 0.6749497950077057, "bpc": 0.9737467221066395, "lr": 0.00012456477615325402, "elapsed_s": 732.0930454730988}
{"event": "train", "step": 16900, "loss": 0.6735267448425293, "bpc": 0.9716936946904032, "lr": 0.00012377799700401205, "elapsed_s": 734.2943465709686}
{"event": "train", "step": 16950, "loss": 0.67251948595047, "bpc": 0.9702405272819379, "lr": 0.00012299196131982236, "elapsed_s": 736.5115036964417}
{"event": "train", "step": 17000, "loss": 0.672699875831604, "bpc": 0.9705007748688765, "lr": 0.00012220669138693287, "elapsed_s": 738.7307095527649}
{"event": "eval", "step": 17000, "val_loss": 0.6764336049556732, "val_bpc": 0.9758874073601939}
{"event": "train", "step": 17050, "loss": 0.6721290028095246, "bpc": 0.9696771791909453, "lr": 0.00012142220946988046, "elapsed_s": 741.4859783649445}
{"event": "train", "step": 17100, "loss": 0.6730441749095917, "bpc": 0.9709974934412721, "lr": 0.00012063853781085961, "elapsed_s": 743.7041358947754}
{"event": "train", "step": 17150, "loss": 0.6718435859680176, "bpc": 0.969265409729117, "lr": 0.00011985569862909176, "elapsed_s": 745.9222111701965}
{"event": "train", "step": 17200, "loss": 0.6743604302406311, "bpc": 0.9728964484799062, "lr": 0.0001190737141201953, "elapsed_s": 748.1391522884369}
{"event": "train", "step": 17250, "loss": 0.670499906539917, "bpc": 0.9673268900816517, "lr": 0.00011829260645555634, "elapsed_s": 750.3572218418121}
{"event": "train", "step": 17300, "loss": 0.6674233341217041, "bpc": 0.9628883343109602, "lr": 0.0001175123977817002, "elapsed_s": 752.5712177753448}
{"event": "train", "step": 17350, "loss": 0.6688513636589051, "bpc": 0.9649485454425231, "lr": 0.00011673311021966323, "elapsed_s": 754.7605226039886}
{"event": "train", "step": 17400, "loss": 0.6712428283691406, "bpc": 0.9683986997204408, "lr": 0.00011595476586436575, "elapsed_s": 756.9502563476562}
{"event": "train", "step": 17450, "loss": 0.6764183127880097, "bpc": 0.9758653454257413, "lr": 0.00011517738678398568, "elapsed_s": 759.1512181758881}
{"event": "train", "step": 17500, "loss": 0.6686432194709778, "bpc": 0.9646482568548105, "lr": 0.00011440099501933276, "elapsed_s": 761.3696491718292}
{"event": "eval", "step": 17500, "val_loss": 0.6687010243535042, "val_bpc": 0.9647316516721705}
{"event": "train", "step": 17550, "loss": 0.6700019407272338, "bpc": 0.9666084772731615, "lr": 0.00011362561258322365, "elapsed_s": 764.1680424213409}
{"event": "train", "step": 17600, "loss": 0.6673480188846588, "bpc": 0.9627796773919716, "lr": 0.00011285126145985777, "elapsed_s": 766.3678398132324}
{"event": "train", "step": 17650, "loss": 0.6663969492912293, "bpc": 0.9614075740059905, "lr": 0.00011207796360419396, "elapsed_s": 768.5696816444397}
{"event": "train", "step": 17700, "loss": 0.6692521846294404, "bpc": 0.9655268078689986, "lr": 0.00011130574094132824, "elapsed_s": 770.7842013835907}
{"event": "train", "step": 17750, "loss": 0.6719021964073181, "bpc": 0.9693499667192401, "lr": 0.00011053461536587182, "elapsed_s": 773.0034294128418}
{"event": "train", "step": 17800, "loss": 0.6708154308795929, "bpc": 0.967782095481782, "lr": 0.00010976460874133058, "elapsed_s": 775.2219722270966}
{"event": "train", "step": 17850, "loss": 0.6703490960597992, "bpc": 0.9671093165498716, "lr": 0.00010899574289948508, "elapsed_s": 777.4394464492798}
{"event": "train", "step": 17900, "loss": 0.6639877080917358, "bpc": 0.9579317736751759, "lr": 0.00010822803963977158, "elapsed_s": 779.6121685504913}
{"event": "train", "step": 17950, "loss": 0.6667386567592621, "bpc": 0.9619005536755563, "lr": 0.00010746152072866405, "elapsed_s": 781.8134696483612}
{"event": "train", "step": 18000, "loss": 0.6691559052467346, "bpc": 0.9653879060810292, "lr": 0.00010669620789905688, "elapsed_s": 784.0164024829865}
{"event": "eval", "step": 18000, "val_loss": 0.6694979101419449, "val_bpc": 0.9658813148473088}
{"event": "train", "step": 18050, "loss": 0.6662026000022888, "bpc": 0.9611271872506358, "lr": 0.00010593212284964875, "elapsed_s": 786.7726106643677}
{"event": "train", "step": 18100, "loss": 0.6627536368370056, "bpc": 0.9561513851958731, "lr": 0.00010516928724432743, "elapsed_s": 788.9892482757568}
{"event": "train", "step": 18150, "loss": 0.6657631659507751, "bpc": 0.9604932179237192, "lr": 0.00010440772271155573, "elapsed_s": 791.2086486816406}
{"event": "train", "step": 18200, "loss": 0.6618948662281037, "bpc": 0.954912441097149, "lr": 0.0001036474508437579, "elapsed_s": 793.4112532138824}
{"event": "train", "step": 18250, "loss": 0.666173528432846, "bpc": 0.9610852458415698, "lr": 0.00010288849319670771, "elapsed_s": 795.6141452789307}
{"event": "train", "step": 18300, "loss": 0.664857667684555, "bpc": 0.95918686006551, "lr": 0.0001021308712889171, "elapsed_s": 797.7952926158905}
{"event": "train", "step": 18350, "loss": 0.6651471316814422, "bpc": 0.959604468338335, "lr": 0.00010137460660102639, "elapsed_s": 800.0034484863281}
{"event": "train", "step": 18400, "loss": 0.6616902613639831, "bpc": 0.9546172586743406, "lr": 0.00010061972057519484, "elapsed_s": 802.2212383747101}
{"event": "train", "step": 18450, "loss": 0.6635430371761322, "bpc": 0.957290249150407, "lr": 9.986623461449292e-05, "elapsed_s": 804.4414618015289}
{"event": "train", "step": 18500, "loss": 0.6631872403621674, "bpc": 0.9567769428513359, "lr": 9.911417008229544e-05, "elapsed_s": 806.6586456298828}
{"event": "eval", "step": 18500, "val_loss": 0.6684446185827255, "val_bpc": 0.9643617363382128}
{"event": "train", "step": 18550, "loss": 0.6651015675067902, "bpc": 0.9595387331295223, "lr": 9.8363548301676e-05, "elapsed_s": 809.3809282779694}
{"event": "train", "step": 18600, "loss": 0.6602735388278961, "bpc": 0.9525733600972122, "lr": 9.761439055480204e-05, "elapsed_s": 811.5587549209595}
{"event": "train", "step": 18650, "loss": 0.6615681028366089, "bpc": 0.9544410211726955, "lr": 9.686671808233185e-05, "elapsed_s": 813.7697401046753}
{"event": "train", "step": 18700, "loss": 0.6660811233520508, "bpc": 0.9609519334897537, "lr": 9.6120552082812e-05, "elapsed_s": 815.9894850254059}
{"event": "train", "step": 18750, "loss": 0.662702386379242, "bpc": 0.9560774464146141, "lr": 9.537591371207666e-05, "elapsed_s": 818.2083117961884}
{"event": "train", "step": 18800, "loss": 0.6620621633529663, "bpc": 0.9551537998295433, "lr": 9.46328240826474e-05, "elapsed_s": 820.4016063213348}
{"event": "train", "step": 18850, "loss": 0.6603107762336731, "bpc": 0.9526270823178622, "lr": 9.389130426313479e-05, "elapsed_s": 822.5872311592102}