{ "best_metric": null, "best_model_checkpoint": null, "epoch": 0.034399944960088066, "eval_steps": 200000, "global_step": 21500, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0001599997440004096, "grad_norm": 84.32501983642578, "learning_rate": 3.103950336794611e-08, "loss": 10.8792, "step": 100 }, { "epoch": 0.0003199994880008192, "grad_norm": 60.63747024536133, "learning_rate": 6.303899137613798e-08, "loss": 10.9284, "step": 200 }, { "epoch": 0.00047999923200122877, "grad_norm": 55.71075439453125, "learning_rate": 9.503847938432986e-08, "loss": 10.6466, "step": 300 }, { "epoch": 0.0006399989760016384, "grad_norm": 57.63307189941406, "learning_rate": 1.2703796739252173e-07, "loss": 10.841, "step": 400 }, { "epoch": 0.000799998720002048, "grad_norm": 89.1032485961914, "learning_rate": 1.590374554007136e-07, "loss": 10.8094, "step": 500 }, { "epoch": 0.0009599984640024575, "grad_norm": 57.2479362487793, "learning_rate": 1.9103694340890547e-07, "loss": 10.4323, "step": 600 }, { "epoch": 0.0011199982080028672, "grad_norm": 51.17530059814453, "learning_rate": 2.2303643141709733e-07, "loss": 10.3032, "step": 700 }, { "epoch": 0.0012799979520032767, "grad_norm": 60.76409912109375, "learning_rate": 2.550359194252892e-07, "loss": 10.4006, "step": 800 }, { "epoch": 0.0014399976960036865, "grad_norm": 67.00859069824219, "learning_rate": 2.870354074334811e-07, "loss": 10.4743, "step": 900 }, { "epoch": 0.001599997440004096, "grad_norm": 68.4343032836914, "learning_rate": 3.19034895441673e-07, "loss": 10.2334, "step": 1000 }, { "epoch": 0.0017599971840045055, "grad_norm": 48.704105377197266, "learning_rate": 3.510343834498648e-07, "loss": 10.0135, "step": 1100 }, { "epoch": 0.001919996928004915, "grad_norm": 45.30134963989258, "learning_rate": 3.830338714580567e-07, "loss": 9.7874, "step": 1200 }, { "epoch": 0.002079996672005325, "grad_norm": 84.56024169921875, "learning_rate": 4.150333594662486e-07, "loss": 9.7419, "step": 1300 }, { "epoch": 0.0022399964160057344, "grad_norm": 45.73213195800781, "learning_rate": 4.470328474744404e-07, "loss": 9.7412, "step": 1400 }, { "epoch": 0.002399996160006144, "grad_norm": 50.21996307373047, "learning_rate": 4.790323354826324e-07, "loss": 9.4585, "step": 1500 }, { "epoch": 0.0025599959040065534, "grad_norm": 59.475799560546875, "learning_rate": 5.110318234908241e-07, "loss": 9.5339, "step": 1600 }, { "epoch": 0.002719995648006963, "grad_norm": 82.53620910644531, "learning_rate": 5.43031311499016e-07, "loss": 9.4345, "step": 1700 }, { "epoch": 0.002879995392007373, "grad_norm": 39.44235610961914, "learning_rate": 5.750307995072079e-07, "loss": 9.1733, "step": 1800 }, { "epoch": 0.0030399951360077825, "grad_norm": 37.58698654174805, "learning_rate": 6.070302875153998e-07, "loss": 8.9952, "step": 1900 }, { "epoch": 0.003199994880008192, "grad_norm": 40.35204315185547, "learning_rate": 6.390297755235917e-07, "loss": 8.9669, "step": 2000 }, { "epoch": 0.0033599946240086016, "grad_norm": 57.84451675415039, "learning_rate": 6.707092686517017e-07, "loss": 8.8152, "step": 2100 }, { "epoch": 0.003519994368009011, "grad_norm": 40.126953125, "learning_rate": 7.027087566598935e-07, "loss": 8.7936, "step": 2200 }, { "epoch": 0.0036799941120094206, "grad_norm": 35.435707092285156, "learning_rate": 7.347082446680854e-07, "loss": 8.6771, "step": 2300 }, { "epoch": 0.00383999385600983, "grad_norm": 42.3509635925293, "learning_rate": 7.667077326762773e-07, "loss": 8.4648, "step": 2400 }, { "epoch": 0.00399999360001024, "grad_norm": 33.58556365966797, "learning_rate": 7.987072206844691e-07, "loss": 8.5764, "step": 2500 }, { "epoch": 0.00415999334401065, "grad_norm": 34.014678955078125, "learning_rate": 8.30706708692661e-07, "loss": 8.4587, "step": 2600 }, { "epoch": 0.004319993088011059, "grad_norm": 36.43831253051758, "learning_rate": 8.627061967008528e-07, "loss": 8.2966, "step": 2700 }, { "epoch": 0.004479992832011469, "grad_norm": 31.411684036254883, "learning_rate": 8.947056847090448e-07, "loss": 8.2329, "step": 2800 }, { "epoch": 0.004639992576011879, "grad_norm": 47.570125579833984, "learning_rate": 9.267051727172366e-07, "loss": 8.1415, "step": 2900 }, { "epoch": 0.004799992320012288, "grad_norm": 30.771928787231445, "learning_rate": 9.587046607254284e-07, "loss": 8.0404, "step": 3000 }, { "epoch": 0.004959992064012698, "grad_norm": 26.92803955078125, "learning_rate": 9.907041487336204e-07, "loss": 7.9698, "step": 3100 }, { "epoch": 0.005119991808013107, "grad_norm": 31.121917724609375, "learning_rate": 1.0227036367418122e-06, "loss": 7.9205, "step": 3200 }, { "epoch": 0.005279991552013517, "grad_norm": 33.991416931152344, "learning_rate": 1.054703124750004e-06, "loss": 7.8314, "step": 3300 }, { "epoch": 0.005439991296013926, "grad_norm": 31.278030395507812, "learning_rate": 1.086702612758196e-06, "loss": 7.8369, "step": 3400 }, { "epoch": 0.005599991040014336, "grad_norm": 28.116140365600586, "learning_rate": 1.1187021007663878e-06, "loss": 7.6403, "step": 3500 }, { "epoch": 0.005759990784014746, "grad_norm": 30.954113006591797, "learning_rate": 1.1507015887745798e-06, "loss": 7.5842, "step": 3600 }, { "epoch": 0.005919990528015155, "grad_norm": 36.53567886352539, "learning_rate": 1.1827010767827715e-06, "loss": 7.5812, "step": 3700 }, { "epoch": 0.006079990272015565, "grad_norm": 36.81153106689453, "learning_rate": 1.2147005647909635e-06, "loss": 7.4335, "step": 3800 }, { "epoch": 0.006239990016015974, "grad_norm": 22.556833267211914, "learning_rate": 1.2467000527991553e-06, "loss": 7.4917, "step": 3900 }, { "epoch": 0.006399989760016384, "grad_norm": 40.195579528808594, "learning_rate": 1.278699540807347e-06, "loss": 7.3204, "step": 4000 }, { "epoch": 0.006559989504016793, "grad_norm": 21.862642288208008, "learning_rate": 1.310699028815539e-06, "loss": 7.2971, "step": 4100 }, { "epoch": 0.006719989248017203, "grad_norm": 29.61161231994629, "learning_rate": 1.3426985168237308e-06, "loss": 7.2233, "step": 4200 }, { "epoch": 0.006879988992017613, "grad_norm": 22.342451095581055, "learning_rate": 1.3746980048319228e-06, "loss": 7.2081, "step": 4300 }, { "epoch": 0.007039988736018022, "grad_norm": 36.36684799194336, "learning_rate": 1.4066974928401148e-06, "loss": 7.1364, "step": 4400 }, { "epoch": 0.007199988480018432, "grad_norm": 25.563953399658203, "learning_rate": 1.4386969808483064e-06, "loss": 7.0663, "step": 4500 }, { "epoch": 0.007359988224018841, "grad_norm": 22.50385856628418, "learning_rate": 1.4706964688564984e-06, "loss": 6.9601, "step": 4600 }, { "epoch": 0.007519987968019251, "grad_norm": 31.61231231689453, "learning_rate": 1.5026959568646904e-06, "loss": 6.9546, "step": 4700 }, { "epoch": 0.00767998771201966, "grad_norm": 18.862520217895508, "learning_rate": 1.5346954448728822e-06, "loss": 6.9019, "step": 4800 }, { "epoch": 0.00783998745602007, "grad_norm": 32.594539642333984, "learning_rate": 1.5666949328810741e-06, "loss": 6.8801, "step": 4900 }, { "epoch": 0.00799998720002048, "grad_norm": 21.06804084777832, "learning_rate": 1.598694420889266e-06, "loss": 6.7734, "step": 5000 }, { "epoch": 0.00815998694402089, "grad_norm": 31.783803939819336, "learning_rate": 1.6303739140173757e-06, "loss": 6.7648, "step": 5100 }, { "epoch": 0.0083199866880213, "grad_norm": 49.79084777832031, "learning_rate": 1.6623734020255677e-06, "loss": 6.7498, "step": 5200 }, { "epoch": 0.008479986432021708, "grad_norm": 26.1977481842041, "learning_rate": 1.6943728900337597e-06, "loss": 6.6872, "step": 5300 }, { "epoch": 0.008639986176022118, "grad_norm": 21.942001342773438, "learning_rate": 1.7263723780419515e-06, "loss": 6.6264, "step": 5400 }, { "epoch": 0.008799985920022528, "grad_norm": 32.572959899902344, "learning_rate": 1.7583718660501433e-06, "loss": 6.579, "step": 5500 }, { "epoch": 0.008959985664022938, "grad_norm": 20.728240966796875, "learning_rate": 1.7903713540583353e-06, "loss": 6.6001, "step": 5600 }, { "epoch": 0.009119985408023347, "grad_norm": 24.334205627441406, "learning_rate": 1.822370842066527e-06, "loss": 6.5971, "step": 5700 }, { "epoch": 0.009279985152023757, "grad_norm": 27.025753021240234, "learning_rate": 1.854370330074719e-06, "loss": 6.4694, "step": 5800 }, { "epoch": 0.009439984896024167, "grad_norm": 23.506013870239258, "learning_rate": 1.8863698180829106e-06, "loss": 6.3983, "step": 5900 }, { "epoch": 0.009599984640024576, "grad_norm": 35.65713882446289, "learning_rate": 1.9183693060911026e-06, "loss": 6.4477, "step": 6000 }, { "epoch": 0.009759984384024985, "grad_norm": 22.977373123168945, "learning_rate": 1.950368794099295e-06, "loss": 6.4308, "step": 6100 }, { "epoch": 0.009919984128025396, "grad_norm": 22.127635955810547, "learning_rate": 1.982368282107486e-06, "loss": 6.4248, "step": 6200 }, { "epoch": 0.010079983872025805, "grad_norm": 33.53960418701172, "learning_rate": 2.0143677701156784e-06, "loss": 6.2642, "step": 6300 }, { "epoch": 0.010239983616026214, "grad_norm": 24.39597511291504, "learning_rate": 2.04636725812387e-06, "loss": 6.2763, "step": 6400 }, { "epoch": 0.010399983360026625, "grad_norm": 24.471288681030273, "learning_rate": 2.078366746132062e-06, "loss": 6.3878, "step": 6500 }, { "epoch": 0.010559983104027034, "grad_norm": 34.05498123168945, "learning_rate": 2.110366234140254e-06, "loss": 6.2601, "step": 6600 }, { "epoch": 0.010719982848027443, "grad_norm": 30.60455322265625, "learning_rate": 2.142365722148446e-06, "loss": 6.1789, "step": 6700 }, { "epoch": 0.010879982592027852, "grad_norm": 27.737686157226562, "learning_rate": 2.1743652101566377e-06, "loss": 6.1773, "step": 6800 }, { "epoch": 0.011039982336028263, "grad_norm": 24.246810913085938, "learning_rate": 2.2063646981648294e-06, "loss": 6.1439, "step": 6900 }, { "epoch": 0.011199982080028672, "grad_norm": 27.53533363342285, "learning_rate": 2.2383641861730217e-06, "loss": 6.1863, "step": 7000 }, { "epoch": 0.011359981824029081, "grad_norm": 27.81687355041504, "learning_rate": 2.2703636741812134e-06, "loss": 6.0513, "step": 7100 }, { "epoch": 0.011519981568029492, "grad_norm": 28.00519371032715, "learning_rate": 2.3020431673093234e-06, "loss": 6.0671, "step": 7200 }, { "epoch": 0.011679981312029901, "grad_norm": 29.347061157226562, "learning_rate": 2.3340426553175152e-06, "loss": 6.0212, "step": 7300 }, { "epoch": 0.01183998105603031, "grad_norm": 29.621200561523438, "learning_rate": 2.365722148445625e-06, "loss": 6.0043, "step": 7400 }, { "epoch": 0.011999980800030719, "grad_norm": 31.689117431640625, "learning_rate": 2.397721636453817e-06, "loss": 6.0166, "step": 7500 }, { "epoch": 0.01215998054403113, "grad_norm": 46.79508972167969, "learning_rate": 2.429721124462009e-06, "loss": 5.9754, "step": 7600 }, { "epoch": 0.012319980288031539, "grad_norm": 28.857833862304688, "learning_rate": 2.4617206124702006e-06, "loss": 5.9211, "step": 7700 }, { "epoch": 0.012479980032031948, "grad_norm": 58.34132766723633, "learning_rate": 2.4937201004783928e-06, "loss": 5.7867, "step": 7800 }, { "epoch": 0.012639979776032359, "grad_norm": 49.33425521850586, "learning_rate": 2.525719588486584e-06, "loss": 5.8534, "step": 7900 }, { "epoch": 0.012799979520032768, "grad_norm": 39.17392349243164, "learning_rate": 2.5577190764947763e-06, "loss": 5.7708, "step": 8000 }, { "epoch": 0.012959979264033177, "grad_norm": 45.94136428833008, "learning_rate": 2.589718564502968e-06, "loss": 5.8328, "step": 8100 }, { "epoch": 0.013119979008033586, "grad_norm": 36.19196319580078, "learning_rate": 2.6217180525111603e-06, "loss": 5.7417, "step": 8200 }, { "epoch": 0.013279978752033997, "grad_norm": 37.051658630371094, "learning_rate": 2.653717540519352e-06, "loss": 5.8097, "step": 8300 }, { "epoch": 0.013439978496034406, "grad_norm": 90.0757064819336, "learning_rate": 2.6857170285275435e-06, "loss": 5.7578, "step": 8400 }, { "epoch": 0.013599978240034815, "grad_norm": 92.7857894897461, "learning_rate": 2.7177165165357357e-06, "loss": 5.643, "step": 8500 }, { "epoch": 0.013759977984035226, "grad_norm": 26.648149490356445, "learning_rate": 2.7497160045439274e-06, "loss": 5.6401, "step": 8600 }, { "epoch": 0.013919977728035635, "grad_norm": 45.42919158935547, "learning_rate": 2.7817154925521196e-06, "loss": 5.6627, "step": 8700 }, { "epoch": 0.014079977472036044, "grad_norm": 48.3182487487793, "learning_rate": 2.8137149805603114e-06, "loss": 5.6167, "step": 8800 }, { "epoch": 0.014239977216036454, "grad_norm": 51.463653564453125, "learning_rate": 2.8457144685685028e-06, "loss": 5.6539, "step": 8900 }, { "epoch": 0.014399976960036864, "grad_norm": 47.81680679321289, "learning_rate": 2.877713956576695e-06, "loss": 5.4513, "step": 9000 }, { "epoch": 0.014559976704037273, "grad_norm": 42.410667419433594, "learning_rate": 2.9097134445848868e-06, "loss": 5.4132, "step": 9100 }, { "epoch": 0.014719976448037683, "grad_norm": 55.33562088012695, "learning_rate": 2.941712932593079e-06, "loss": 5.4714, "step": 9200 }, { "epoch": 0.014879976192038093, "grad_norm": 38.538246154785156, "learning_rate": 2.9737124206012707e-06, "loss": 5.4786, "step": 9300 }, { "epoch": 0.015039975936038502, "grad_norm": 43.42023468017578, "learning_rate": 3.0057119086094625e-06, "loss": 5.3928, "step": 9400 }, { "epoch": 0.015199975680038912, "grad_norm": 24.861467361450195, "learning_rate": 3.037391401737572e-06, "loss": 5.4774, "step": 9500 }, { "epoch": 0.01535997542403932, "grad_norm": 98.92141723632812, "learning_rate": 3.0693908897457643e-06, "loss": 5.2881, "step": 9600 }, { "epoch": 0.015519975168039732, "grad_norm": 62.839866638183594, "learning_rate": 3.101390377753956e-06, "loss": 5.3699, "step": 9700 }, { "epoch": 0.01567997491204014, "grad_norm": 46.006065368652344, "learning_rate": 3.133069870882066e-06, "loss": 5.1483, "step": 9800 }, { "epoch": 0.01583997465604055, "grad_norm": 89.62445068359375, "learning_rate": 3.1650693588902583e-06, "loss": 5.3051, "step": 9900 }, { "epoch": 0.01599997440004096, "grad_norm": 41.113609313964844, "learning_rate": 3.19706884689845e-06, "loss": 5.2546, "step": 10000 }, { "epoch": 0.01615997414404137, "grad_norm": 46.37376403808594, "learning_rate": 3.2290683349066414e-06, "loss": 5.2314, "step": 10100 }, { "epoch": 0.01631997388804178, "grad_norm": 60.3846321105957, "learning_rate": 3.2610678229148337e-06, "loss": 5.1783, "step": 10200 }, { "epoch": 0.016479973632042188, "grad_norm": 145.4359130859375, "learning_rate": 3.2930673109230254e-06, "loss": 5.2074, "step": 10300 }, { "epoch": 0.0166399733760426, "grad_norm": 69.00183868408203, "learning_rate": 3.325066798931217e-06, "loss": 5.2825, "step": 10400 }, { "epoch": 0.01679997312004301, "grad_norm": 48.03580093383789, "learning_rate": 3.3570662869394094e-06, "loss": 5.1715, "step": 10500 }, { "epoch": 0.016959972864043417, "grad_norm": 58.56736755371094, "learning_rate": 3.389065774947601e-06, "loss": 5.087, "step": 10600 }, { "epoch": 0.017119972608043828, "grad_norm": 54.484527587890625, "learning_rate": 3.421065262955793e-06, "loss": 5.082, "step": 10700 }, { "epoch": 0.017279972352044235, "grad_norm": 74.30866241455078, "learning_rate": 3.4530647509639847e-06, "loss": 4.9111, "step": 10800 }, { "epoch": 0.017439972096044646, "grad_norm": 60.489505767822266, "learning_rate": 3.4850642389721765e-06, "loss": 5.0213, "step": 10900 }, { "epoch": 0.017599971840045057, "grad_norm": 61.25093460083008, "learning_rate": 3.5170637269803687e-06, "loss": 4.9898, "step": 11000 }, { "epoch": 0.017759971584045464, "grad_norm": 51.98568344116211, "learning_rate": 3.5490632149885605e-06, "loss": 4.7734, "step": 11100 }, { "epoch": 0.017919971328045875, "grad_norm": 64.08167266845703, "learning_rate": 3.581062702996752e-06, "loss": 4.9511, "step": 11200 }, { "epoch": 0.018079971072046286, "grad_norm": 61.8354606628418, "learning_rate": 3.613062191004944e-06, "loss": 5.0481, "step": 11300 }, { "epoch": 0.018239970816046693, "grad_norm": 97.53675842285156, "learning_rate": 3.645061679013136e-06, "loss": 4.8441, "step": 11400 }, { "epoch": 0.018399970560047104, "grad_norm": 49.35017013549805, "learning_rate": 3.677061167021328e-06, "loss": 4.873, "step": 11500 }, { "epoch": 0.018559970304047515, "grad_norm": 44.33409118652344, "learning_rate": 3.70906065502952e-06, "loss": 4.9988, "step": 11600 }, { "epoch": 0.018719970048047922, "grad_norm": 140.5505828857422, "learning_rate": 3.741060143037712e-06, "loss": 4.7653, "step": 11700 }, { "epoch": 0.018879969792048333, "grad_norm": 68.21163177490234, "learning_rate": 3.7730596310459034e-06, "loss": 4.804, "step": 11800 }, { "epoch": 0.019039969536048744, "grad_norm": 48.678226470947266, "learning_rate": 3.805059119054095e-06, "loss": 4.8288, "step": 11900 }, { "epoch": 0.01919996928004915, "grad_norm": 76.32611083984375, "learning_rate": 3.837058607062287e-06, "loss": 4.7053, "step": 12000 }, { "epoch": 0.019359969024049562, "grad_norm": 70.85586547851562, "learning_rate": 3.869058095070479e-06, "loss": 4.6887, "step": 12100 }, { "epoch": 0.01951996876804997, "grad_norm": 66.46036529541016, "learning_rate": 3.901057583078671e-06, "loss": 4.7832, "step": 12200 }, { "epoch": 0.01967996851205038, "grad_norm": 165.13221740722656, "learning_rate": 3.9330570710868636e-06, "loss": 4.6817, "step": 12300 }, { "epoch": 0.01983996825605079, "grad_norm": 118.48895263671875, "learning_rate": 3.965056559095055e-06, "loss": 4.6252, "step": 12400 }, { "epoch": 0.0199999680000512, "grad_norm": 64.3436050415039, "learning_rate": 3.997056047103246e-06, "loss": 4.5936, "step": 12500 }, { "epoch": 0.02015996774405161, "grad_norm": 42.27592468261719, "learning_rate": 4.0290555351114385e-06, "loss": 4.7452, "step": 12600 }, { "epoch": 0.02031996748805202, "grad_norm": 60.829036712646484, "learning_rate": 4.061055023119631e-06, "loss": 4.5321, "step": 12700 }, { "epoch": 0.020479967232052428, "grad_norm": 161.975830078125, "learning_rate": 4.093054511127823e-06, "loss": 4.4964, "step": 12800 }, { "epoch": 0.02063996697605284, "grad_norm": 99.2963638305664, "learning_rate": 4.125053999136014e-06, "loss": 4.4421, "step": 12900 }, { "epoch": 0.02079996672005325, "grad_norm": 68.78880310058594, "learning_rate": 4.156733492264124e-06, "loss": 4.3782, "step": 13000 }, { "epoch": 0.020959966464053657, "grad_norm": 80.74951171875, "learning_rate": 4.188732980272316e-06, "loss": 4.5169, "step": 13100 }, { "epoch": 0.021119966208054067, "grad_norm": 157.87254333496094, "learning_rate": 4.220412473400426e-06, "loss": 4.533, "step": 13200 }, { "epoch": 0.02127996595205448, "grad_norm": 148.68331909179688, "learning_rate": 4.252411961408618e-06, "loss": 4.3725, "step": 13300 }, { "epoch": 0.021439965696054886, "grad_norm": 72.9531021118164, "learning_rate": 4.28441144941681e-06, "loss": 4.2911, "step": 13400 }, { "epoch": 0.021599965440055297, "grad_norm": 73.24847412109375, "learning_rate": 4.316410937425001e-06, "loss": 4.2261, "step": 13500 }, { "epoch": 0.021759965184055704, "grad_norm": 94.57313537597656, "learning_rate": 4.348410425433194e-06, "loss": 4.2467, "step": 13600 }, { "epoch": 0.021919964928056115, "grad_norm": 105.674560546875, "learning_rate": 4.380409913441385e-06, "loss": 4.1558, "step": 13700 }, { "epoch": 0.022079964672056526, "grad_norm": 63.658287048339844, "learning_rate": 4.412409401449577e-06, "loss": 4.2794, "step": 13800 }, { "epoch": 0.022239964416056933, "grad_norm": 77.69287109375, "learning_rate": 4.444408889457769e-06, "loss": 4.2383, "step": 13900 }, { "epoch": 0.022399964160057344, "grad_norm": 82.83360290527344, "learning_rate": 4.4764083774659615e-06, "loss": 4.1654, "step": 14000 }, { "epoch": 0.022559963904057755, "grad_norm": 47.373531341552734, "learning_rate": 4.508407865474153e-06, "loss": 4.158, "step": 14100 }, { "epoch": 0.022719963648058162, "grad_norm": 97.64757537841797, "learning_rate": 4.540407353482344e-06, "loss": 4.1299, "step": 14200 }, { "epoch": 0.022879963392058573, "grad_norm": 54.75618362426758, "learning_rate": 4.5724068414905365e-06, "loss": 4.1902, "step": 14300 }, { "epoch": 0.023039963136058984, "grad_norm": 258.4887390136719, "learning_rate": 4.604406329498729e-06, "loss": 3.7853, "step": 14400 }, { "epoch": 0.02319996288005939, "grad_norm": 104.63798522949219, "learning_rate": 4.63640581750692e-06, "loss": 4.0514, "step": 14500 }, { "epoch": 0.023359962624059802, "grad_norm": 60.090843200683594, "learning_rate": 4.668405305515112e-06, "loss": 4.1655, "step": 14600 }, { "epoch": 0.023519962368060213, "grad_norm": 44.36670684814453, "learning_rate": 4.7004047935233036e-06, "loss": 4.051, "step": 14700 }, { "epoch": 0.02367996211206062, "grad_norm": 41.61213302612305, "learning_rate": 4.732404281531496e-06, "loss": 4.078, "step": 14800 }, { "epoch": 0.02383996185606103, "grad_norm": 73.2448501586914, "learning_rate": 4.764403769539688e-06, "loss": 4.1193, "step": 14900 }, { "epoch": 0.023999961600061438, "grad_norm": 77.30301666259766, "learning_rate": 4.796403257547879e-06, "loss": 4.1536, "step": 15000 }, { "epoch": 0.02415996134406185, "grad_norm": 48.1458854675293, "learning_rate": 4.8284027455560715e-06, "loss": 3.935, "step": 15100 }, { "epoch": 0.02431996108806226, "grad_norm": 129.59295654296875, "learning_rate": 4.860402233564263e-06, "loss": 3.9535, "step": 15200 }, { "epoch": 0.024479960832062667, "grad_norm": 163.0813751220703, "learning_rate": 4.892401721572455e-06, "loss": 3.7051, "step": 15300 }, { "epoch": 0.024639960576063078, "grad_norm": 102.2786865234375, "learning_rate": 4.924401209580647e-06, "loss": 3.8329, "step": 15400 }, { "epoch": 0.02479996032006349, "grad_norm": 160.66392517089844, "learning_rate": 4.956400697588839e-06, "loss": 3.9412, "step": 15500 }, { "epoch": 0.024959960064063896, "grad_norm": 136.77218627929688, "learning_rate": 4.988400185597031e-06, "loss": 3.6668, "step": 15600 }, { "epoch": 0.025119959808064307, "grad_norm": 63.87991714477539, "learning_rate": 5.0200796787251404e-06, "loss": 3.7758, "step": 15700 }, { "epoch": 0.025279959552064718, "grad_norm": 352.977294921875, "learning_rate": 5.052079166733333e-06, "loss": 3.8805, "step": 15800 }, { "epoch": 0.025439959296065125, "grad_norm": 148.54776000976562, "learning_rate": 5.084078654741524e-06, "loss": 3.8848, "step": 15900 }, { "epoch": 0.025599959040065536, "grad_norm": 105.01113891601562, "learning_rate": 5.116078142749716e-06, "loss": 3.75, "step": 16000 }, { "epoch": 0.025759958784065947, "grad_norm": 170.62828063964844, "learning_rate": 5.148077630757908e-06, "loss": 3.5685, "step": 16100 }, { "epoch": 0.025919958528066354, "grad_norm": 164.85324096679688, "learning_rate": 5.180077118766101e-06, "loss": 3.7016, "step": 16200 }, { "epoch": 0.026079958272066765, "grad_norm": 79.85810852050781, "learning_rate": 5.212076606774292e-06, "loss": 4.0955, "step": 16300 }, { "epoch": 0.026239958016067173, "grad_norm": 109.73529815673828, "learning_rate": 5.244076094782484e-06, "loss": 3.7577, "step": 16400 }, { "epoch": 0.026399957760067583, "grad_norm": 105.98066711425781, "learning_rate": 5.276075582790676e-06, "loss": 3.7485, "step": 16500 }, { "epoch": 0.026559957504067994, "grad_norm": 71.02545166015625, "learning_rate": 5.3080750707988686e-06, "loss": 3.8263, "step": 16600 }, { "epoch": 0.0267199572480684, "grad_norm": 245.44224548339844, "learning_rate": 5.340074558807059e-06, "loss": 3.6922, "step": 16700 }, { "epoch": 0.026879956992068813, "grad_norm": 42.178157806396484, "learning_rate": 5.372074046815251e-06, "loss": 3.6568, "step": 16800 }, { "epoch": 0.027039956736069223, "grad_norm": 114.55894470214844, "learning_rate": 5.404073534823443e-06, "loss": 3.7317, "step": 16900 }, { "epoch": 0.02719995648006963, "grad_norm": 86.70626831054688, "learning_rate": 5.436073022831635e-06, "loss": 3.5089, "step": 17000 }, { "epoch": 0.02735995622407004, "grad_norm": 202.02505493164062, "learning_rate": 5.468072510839827e-06, "loss": 3.7377, "step": 17100 }, { "epoch": 0.027519955968070452, "grad_norm": 114.00701141357422, "learning_rate": 5.500071998848019e-06, "loss": 3.6206, "step": 17200 }, { "epoch": 0.02767995571207086, "grad_norm": 152.38311767578125, "learning_rate": 5.532071486856211e-06, "loss": 3.3702, "step": 17300 }, { "epoch": 0.02783995545607127, "grad_norm": 156.1048126220703, "learning_rate": 5.564070974864403e-06, "loss": 3.5126, "step": 17400 }, { "epoch": 0.02799995520007168, "grad_norm": 117.87386322021484, "learning_rate": 5.596070462872595e-06, "loss": 3.4841, "step": 17500 }, { "epoch": 0.02815995494407209, "grad_norm": 616.7991333007812, "learning_rate": 5.628069950880786e-06, "loss": 3.1464, "step": 17600 }, { "epoch": 0.0283199546880725, "grad_norm": 131.32760620117188, "learning_rate": 5.6600694388889786e-06, "loss": 3.7012, "step": 17700 }, { "epoch": 0.028479954432072907, "grad_norm": 60.172969818115234, "learning_rate": 5.69206892689717e-06, "loss": 3.5802, "step": 17800 }, { "epoch": 0.028639954176073318, "grad_norm": 169.24374389648438, "learning_rate": 5.724068414905361e-06, "loss": 3.4952, "step": 17900 }, { "epoch": 0.02879995392007373, "grad_norm": 158.77391052246094, "learning_rate": 5.7560679029135535e-06, "loss": 3.1174, "step": 18000 }, { "epoch": 0.028959953664074136, "grad_norm": 218.98867797851562, "learning_rate": 5.787747396041664e-06, "loss": 3.3134, "step": 18100 }, { "epoch": 0.029119953408074547, "grad_norm": 185.3249053955078, "learning_rate": 5.819746884049856e-06, "loss": 3.3578, "step": 18200 }, { "epoch": 0.029279953152074958, "grad_norm": 93.69242858886719, "learning_rate": 5.851746372058048e-06, "loss": 3.0209, "step": 18300 }, { "epoch": 0.029439952896075365, "grad_norm": 85.82784271240234, "learning_rate": 5.883745860066239e-06, "loss": 3.3796, "step": 18400 }, { "epoch": 0.029599952640075776, "grad_norm": 125.96697998046875, "learning_rate": 5.915745348074431e-06, "loss": 3.2287, "step": 18500 }, { "epoch": 0.029759952384076187, "grad_norm": 235.71075439453125, "learning_rate": 5.947744836082623e-06, "loss": 3.1537, "step": 18600 }, { "epoch": 0.029919952128076594, "grad_norm": 139.5558319091797, "learning_rate": 5.979744324090815e-06, "loss": 2.9073, "step": 18700 }, { "epoch": 0.030079951872077005, "grad_norm": 204.2928924560547, "learning_rate": 6.011743812099007e-06, "loss": 3.3444, "step": 18800 }, { "epoch": 0.030239951616077416, "grad_norm": 165.4457244873047, "learning_rate": 6.043743300107199e-06, "loss": 3.1341, "step": 18900 }, { "epoch": 0.030399951360077823, "grad_norm": 66.5983657836914, "learning_rate": 6.07574278811539e-06, "loss": 2.8862, "step": 19000 }, { "epoch": 0.030559951104078234, "grad_norm": 219.95774841308594, "learning_rate": 6.1077422761235826e-06, "loss": 3.2033, "step": 19100 }, { "epoch": 0.03071995084807864, "grad_norm": 125.15766906738281, "learning_rate": 6.139741764131775e-06, "loss": 3.2764, "step": 19200 }, { "epoch": 0.030879950592079052, "grad_norm": 207.95970153808594, "learning_rate": 6.171741252139967e-06, "loss": 3.0725, "step": 19300 }, { "epoch": 0.031039950336079463, "grad_norm": 368.32781982421875, "learning_rate": 6.203740740148158e-06, "loss": 3.0436, "step": 19400 }, { "epoch": 0.03119995008007987, "grad_norm": 412.2764587402344, "learning_rate": 6.23574022815635e-06, "loss": 3.3493, "step": 19500 }, { "epoch": 0.03135994982408028, "grad_norm": 155.46766662597656, "learning_rate": 6.267739716164542e-06, "loss": 3.0141, "step": 19600 }, { "epoch": 0.03151994956808069, "grad_norm": 89.32569885253906, "learning_rate": 6.299739204172733e-06, "loss": 2.779, "step": 19700 }, { "epoch": 0.0316799493120811, "grad_norm": 241.4378204345703, "learning_rate": 6.3317386921809254e-06, "loss": 3.3543, "step": 19800 }, { "epoch": 0.03183994905608151, "grad_norm": 13.20569133758545, "learning_rate": 6.363738180189118e-06, "loss": 3.1526, "step": 19900 }, { "epoch": 0.03199994880008192, "grad_norm": 270.6402893066406, "learning_rate": 6.395737668197309e-06, "loss": 2.7896, "step": 20000 }, { "epoch": 0.03215994854408233, "grad_norm": 106.38632202148438, "learning_rate": 6.427737156205501e-06, "loss": 2.9398, "step": 20100 }, { "epoch": 0.03231994828808274, "grad_norm": 191.7210693359375, "learning_rate": 6.459416649333611e-06, "loss": 3.1254, "step": 20200 }, { "epoch": 0.03247994803208315, "grad_norm": 143.96151733398438, "learning_rate": 6.491416137341803e-06, "loss": 2.8832, "step": 20300 }, { "epoch": 0.03263994777608356, "grad_norm": 150.26368713378906, "learning_rate": 6.523415625349994e-06, "loss": 3.0542, "step": 20400 }, { "epoch": 0.032799947520083965, "grad_norm": 178.11705017089844, "learning_rate": 6.5554151133581865e-06, "loss": 2.9722, "step": 20500 }, { "epoch": 0.032959947264084376, "grad_norm": 222.4794921875, "learning_rate": 6.587414601366379e-06, "loss": 2.9321, "step": 20600 }, { "epoch": 0.03311994700808479, "grad_norm": 155.37796020507812, "learning_rate": 6.619414089374571e-06, "loss": 2.6448, "step": 20700 }, { "epoch": 0.0332799467520852, "grad_norm": 155.5786590576172, "learning_rate": 6.651413577382762e-06, "loss": 3.4006, "step": 20800 }, { "epoch": 0.03343994649608561, "grad_norm": 684.525146484375, "learning_rate": 6.6834130653909545e-06, "loss": 3.0022, "step": 20900 }, { "epoch": 0.03359994624008602, "grad_norm": 545.5623168945312, "learning_rate": 6.715412553399147e-06, "loss": 2.6366, "step": 21000 }, { "epoch": 0.03375994598408642, "grad_norm": 292.9093017578125, "learning_rate": 6.747412041407339e-06, "loss": 3.0112, "step": 21100 }, { "epoch": 0.033919945728086834, "grad_norm": 2.531680107116699, "learning_rate": 6.7794115294155294e-06, "loss": 2.7856, "step": 21200 }, { "epoch": 0.034079945472087245, "grad_norm": 216.7860565185547, "learning_rate": 6.811411017423722e-06, "loss": 3.0967, "step": 21300 }, { "epoch": 0.034239945216087656, "grad_norm": 138.73028564453125, "learning_rate": 6.843410505431913e-06, "loss": 2.8754, "step": 21400 }, { "epoch": 0.034399944960088066, "grad_norm": 78.2362060546875, "learning_rate": 6.875409993440105e-06, "loss": 3.1269, "step": 21500 } ], "logging_steps": 100, "max_steps": 625001, "num_input_tokens_seen": 0, "num_train_epochs": 1, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 128, "trial_name": null, "trial_params": null }