{ "best_metric": null, "best_model_checkpoint": null, "epoch": 3.0, "eval_steps": 500, "global_step": 5268, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0056947608200455585, "grad_norm": 5.042048454284668, "learning_rate": 4.990508731966591e-05, "loss": 1.2902, "step": 10 }, { "epoch": 0.011389521640091117, "grad_norm": 1.2201155424118042, "learning_rate": 4.981017463933182e-05, "loss": 0.6163, "step": 20 }, { "epoch": 0.017084282460136675, "grad_norm": 1.153208613395691, "learning_rate": 4.971526195899772e-05, "loss": 0.2969, "step": 30 }, { "epoch": 0.022779043280182234, "grad_norm": 1.3441071510314941, "learning_rate": 4.962034927866363e-05, "loss": 0.427, "step": 40 }, { "epoch": 0.02847380410022779, "grad_norm": 1.4184179306030273, "learning_rate": 4.9525436598329536e-05, "loss": 0.2912, "step": 50 }, { "epoch": 0.03416856492027335, "grad_norm": 1.7669013738632202, "learning_rate": 4.9430523917995447e-05, "loss": 0.2581, "step": 60 }, { "epoch": 0.03986332574031891, "grad_norm": 2.6280012130737305, "learning_rate": 4.933561123766136e-05, "loss": 0.191, "step": 70 }, { "epoch": 0.04555808656036447, "grad_norm": 2.000683069229126, "learning_rate": 4.924069855732726e-05, "loss": 0.1709, "step": 80 }, { "epoch": 0.05125284738041002, "grad_norm": 2.97003436088562, "learning_rate": 4.9145785876993165e-05, "loss": 0.1989, "step": 90 }, { "epoch": 0.05694760820045558, "grad_norm": 1.8411054611206055, "learning_rate": 4.9050873196659076e-05, "loss": 0.1736, "step": 100 }, { "epoch": 0.06264236902050115, "grad_norm": 7.188544750213623, "learning_rate": 4.895596051632499e-05, "loss": 0.2518, "step": 110 }, { "epoch": 0.0683371298405467, "grad_norm": 2.3247721195220947, "learning_rate": 4.886104783599089e-05, "loss": 0.1648, "step": 120 }, { "epoch": 0.07403189066059225, "grad_norm": 0.8464665412902832, "learning_rate": 4.87661351556568e-05, "loss": 0.1625, "step": 130 }, { "epoch": 0.07972665148063782, "grad_norm": 2.508469343185425, "learning_rate": 4.8671222475322705e-05, "loss": 0.1449, "step": 140 }, { "epoch": 0.08542141230068337, "grad_norm": 0.8359629511833191, "learning_rate": 4.857630979498861e-05, "loss": 0.1114, "step": 150 }, { "epoch": 0.09111617312072894, "grad_norm": 1.0664496421813965, "learning_rate": 4.848139711465452e-05, "loss": 0.0721, "step": 160 }, { "epoch": 0.09681093394077449, "grad_norm": 1.8065450191497803, "learning_rate": 4.838648443432043e-05, "loss": 0.1384, "step": 170 }, { "epoch": 0.10250569476082004, "grad_norm": 1.9225945472717285, "learning_rate": 4.8291571753986335e-05, "loss": 0.0997, "step": 180 }, { "epoch": 0.1082004555808656, "grad_norm": 4.79730224609375, "learning_rate": 4.8196659073652246e-05, "loss": 0.0918, "step": 190 }, { "epoch": 0.11389521640091116, "grad_norm": 0.7483096122741699, "learning_rate": 4.810174639331815e-05, "loss": 0.0932, "step": 200 }, { "epoch": 0.11958997722095673, "grad_norm": 4.0233473777771, "learning_rate": 4.8006833712984054e-05, "loss": 0.1084, "step": 210 }, { "epoch": 0.1252847380410023, "grad_norm": 3.508195638656616, "learning_rate": 4.7911921032649964e-05, "loss": 0.103, "step": 220 }, { "epoch": 0.13097949886104784, "grad_norm": 1.099330186843872, "learning_rate": 4.7817008352315875e-05, "loss": 0.1048, "step": 230 }, { "epoch": 0.1366742596810934, "grad_norm": 0.9204321503639221, "learning_rate": 4.772209567198178e-05, "loss": 0.1044, "step": 240 }, { "epoch": 0.14236902050113895, "grad_norm": 0.697127103805542, "learning_rate": 4.762718299164768e-05, "loss": 0.1003, "step": 250 }, { "epoch": 0.1480637813211845, "grad_norm": 2.3334975242614746, "learning_rate": 4.7532270311313594e-05, "loss": 0.0891, "step": 260 }, { "epoch": 0.15375854214123008, "grad_norm": 2.2273714542388916, "learning_rate": 4.74373576309795e-05, "loss": 0.1135, "step": 270 }, { "epoch": 0.15945330296127563, "grad_norm": 1.4537644386291504, "learning_rate": 4.734244495064541e-05, "loss": 0.0991, "step": 280 }, { "epoch": 0.16514806378132119, "grad_norm": 2.1925878524780273, "learning_rate": 4.724753227031132e-05, "loss": 0.0885, "step": 290 }, { "epoch": 0.17084282460136674, "grad_norm": 1.6182861328125, "learning_rate": 4.715261958997722e-05, "loss": 0.0959, "step": 300 }, { "epoch": 0.1765375854214123, "grad_norm": 0.6557925343513489, "learning_rate": 4.705770690964313e-05, "loss": 0.1518, "step": 310 }, { "epoch": 0.18223234624145787, "grad_norm": 1.7739641666412354, "learning_rate": 4.696279422930904e-05, "loss": 0.1077, "step": 320 }, { "epoch": 0.18792710706150342, "grad_norm": 0.9851402044296265, "learning_rate": 4.686788154897495e-05, "loss": 0.1562, "step": 330 }, { "epoch": 0.19362186788154898, "grad_norm": 0.23916186392307281, "learning_rate": 4.677296886864085e-05, "loss": 0.0787, "step": 340 }, { "epoch": 0.19931662870159453, "grad_norm": 1.2982560396194458, "learning_rate": 4.6678056188306763e-05, "loss": 0.0843, "step": 350 }, { "epoch": 0.20501138952164008, "grad_norm": 0.8306344747543335, "learning_rate": 4.658314350797267e-05, "loss": 0.096, "step": 360 }, { "epoch": 0.21070615034168566, "grad_norm": 1.477241039276123, "learning_rate": 4.648823082763857e-05, "loss": 0.0799, "step": 370 }, { "epoch": 0.2164009111617312, "grad_norm": 2.520557165145874, "learning_rate": 4.639331814730448e-05, "loss": 0.12, "step": 380 }, { "epoch": 0.22209567198177677, "grad_norm": 5.481024742126465, "learning_rate": 4.629840546697039e-05, "loss": 0.0756, "step": 390 }, { "epoch": 0.22779043280182232, "grad_norm": 7.298396587371826, "learning_rate": 4.62034927866363e-05, "loss": 0.0926, "step": 400 }, { "epoch": 0.23348519362186787, "grad_norm": 1.9556351900100708, "learning_rate": 4.61085801063022e-05, "loss": 0.061, "step": 410 }, { "epoch": 0.23917995444191345, "grad_norm": 1.6126965284347534, "learning_rate": 4.601366742596811e-05, "loss": 0.0913, "step": 420 }, { "epoch": 0.244874715261959, "grad_norm": 1.855068325996399, "learning_rate": 4.5918754745634016e-05, "loss": 0.0881, "step": 430 }, { "epoch": 0.2505694760820046, "grad_norm": 4.383277416229248, "learning_rate": 4.5823842065299926e-05, "loss": 0.0948, "step": 440 }, { "epoch": 0.25626423690205014, "grad_norm": 4.287411212921143, "learning_rate": 4.572892938496584e-05, "loss": 0.1398, "step": 450 }, { "epoch": 0.2619589977220957, "grad_norm": 1.1114435195922852, "learning_rate": 4.563401670463174e-05, "loss": 0.105, "step": 460 }, { "epoch": 0.26765375854214124, "grad_norm": 0.9561439156532288, "learning_rate": 4.5539104024297645e-05, "loss": 0.1039, "step": 470 }, { "epoch": 0.2733485193621868, "grad_norm": 2.2738897800445557, "learning_rate": 4.5444191343963556e-05, "loss": 0.1049, "step": 480 }, { "epoch": 0.27904328018223234, "grad_norm": 2.5520405769348145, "learning_rate": 4.5349278663629466e-05, "loss": 0.1059, "step": 490 }, { "epoch": 0.2847380410022779, "grad_norm": 1.6588718891143799, "learning_rate": 4.525436598329537e-05, "loss": 0.07, "step": 500 }, { "epoch": 0.29043280182232345, "grad_norm": 1.141545057296753, "learning_rate": 4.515945330296128e-05, "loss": 0.1057, "step": 510 }, { "epoch": 0.296127562642369, "grad_norm": 1.383521556854248, "learning_rate": 4.5064540622627185e-05, "loss": 0.041, "step": 520 }, { "epoch": 0.30182232346241455, "grad_norm": 2.6926703453063965, "learning_rate": 4.496962794229309e-05, "loss": 0.0825, "step": 530 }, { "epoch": 0.30751708428246016, "grad_norm": 5.290349006652832, "learning_rate": 4.4874715261959e-05, "loss": 0.0614, "step": 540 }, { "epoch": 0.3132118451025057, "grad_norm": 3.191392183303833, "learning_rate": 4.477980258162491e-05, "loss": 0.1721, "step": 550 }, { "epoch": 0.31890660592255127, "grad_norm": 1.4004778861999512, "learning_rate": 4.4684889901290815e-05, "loss": 0.0448, "step": 560 }, { "epoch": 0.3246013667425968, "grad_norm": 1.8395417928695679, "learning_rate": 4.4589977220956725e-05, "loss": 0.0967, "step": 570 }, { "epoch": 0.33029612756264237, "grad_norm": 2.5588178634643555, "learning_rate": 4.449506454062263e-05, "loss": 0.0741, "step": 580 }, { "epoch": 0.3359908883826879, "grad_norm": 3.3830788135528564, "learning_rate": 4.440015186028853e-05, "loss": 0.0723, "step": 590 }, { "epoch": 0.3416856492027335, "grad_norm": 2.411625862121582, "learning_rate": 4.4305239179954444e-05, "loss": 0.1199, "step": 600 }, { "epoch": 0.34738041002277903, "grad_norm": 0.10270216315984726, "learning_rate": 4.4210326499620355e-05, "loss": 0.051, "step": 610 }, { "epoch": 0.3530751708428246, "grad_norm": 3.844116687774658, "learning_rate": 4.411541381928626e-05, "loss": 0.0484, "step": 620 }, { "epoch": 0.35876993166287013, "grad_norm": 1.4396392107009888, "learning_rate": 4.402050113895216e-05, "loss": 0.0355, "step": 630 }, { "epoch": 0.36446469248291574, "grad_norm": 2.1004416942596436, "learning_rate": 4.3925588458618073e-05, "loss": 0.0854, "step": 640 }, { "epoch": 0.3701594533029613, "grad_norm": 1.7273060083389282, "learning_rate": 4.383067577828398e-05, "loss": 0.068, "step": 650 }, { "epoch": 0.37585421412300685, "grad_norm": 1.6584464311599731, "learning_rate": 4.373576309794989e-05, "loss": 0.0591, "step": 660 }, { "epoch": 0.3815489749430524, "grad_norm": 5.8685150146484375, "learning_rate": 4.36408504176158e-05, "loss": 0.0703, "step": 670 }, { "epoch": 0.38724373576309795, "grad_norm": 1.8818565607070923, "learning_rate": 4.35459377372817e-05, "loss": 0.0864, "step": 680 }, { "epoch": 0.3929384965831435, "grad_norm": 0.8174405694007874, "learning_rate": 4.345102505694761e-05, "loss": 0.0541, "step": 690 }, { "epoch": 0.39863325740318906, "grad_norm": 2.2386655807495117, "learning_rate": 4.335611237661352e-05, "loss": 0.0779, "step": 700 }, { "epoch": 0.4043280182232346, "grad_norm": 1.5718015432357788, "learning_rate": 4.326119969627943e-05, "loss": 0.0827, "step": 710 }, { "epoch": 0.41002277904328016, "grad_norm": 0.7225226759910583, "learning_rate": 4.316628701594533e-05, "loss": 0.0961, "step": 720 }, { "epoch": 0.4157175398633257, "grad_norm": 2.2464139461517334, "learning_rate": 4.307137433561124e-05, "loss": 0.0432, "step": 730 }, { "epoch": 0.4214123006833713, "grad_norm": 1.8811899423599243, "learning_rate": 4.297646165527715e-05, "loss": 0.0752, "step": 740 }, { "epoch": 0.4271070615034169, "grad_norm": 1.1304991245269775, "learning_rate": 4.288154897494305e-05, "loss": 0.0684, "step": 750 }, { "epoch": 0.4328018223234624, "grad_norm": 5.151347637176514, "learning_rate": 4.278663629460896e-05, "loss": 0.0498, "step": 760 }, { "epoch": 0.438496583143508, "grad_norm": 0.0376850925385952, "learning_rate": 4.269172361427487e-05, "loss": 0.1001, "step": 770 }, { "epoch": 0.44419134396355353, "grad_norm": 0.9273952841758728, "learning_rate": 4.2596810933940777e-05, "loss": 0.0773, "step": 780 }, { "epoch": 0.4498861047835991, "grad_norm": 2.6929280757904053, "learning_rate": 4.250189825360668e-05, "loss": 0.0555, "step": 790 }, { "epoch": 0.45558086560364464, "grad_norm": 0.7927613854408264, "learning_rate": 4.240698557327259e-05, "loss": 0.0659, "step": 800 }, { "epoch": 0.4612756264236902, "grad_norm": 0.3811888098716736, "learning_rate": 4.2312072892938495e-05, "loss": 0.0361, "step": 810 }, { "epoch": 0.46697038724373574, "grad_norm": 8.167976379394531, "learning_rate": 4.2217160212604406e-05, "loss": 0.0742, "step": 820 }, { "epoch": 0.47266514806378135, "grad_norm": 2.267798662185669, "learning_rate": 4.212224753227032e-05, "loss": 0.0655, "step": 830 }, { "epoch": 0.4783599088838269, "grad_norm": 0.3650657832622528, "learning_rate": 4.202733485193622e-05, "loss": 0.0508, "step": 840 }, { "epoch": 0.48405466970387245, "grad_norm": 2.446582078933716, "learning_rate": 4.1932422171602125e-05, "loss": 0.0824, "step": 850 }, { "epoch": 0.489749430523918, "grad_norm": 2.046130418777466, "learning_rate": 4.1837509491268035e-05, "loss": 0.0443, "step": 860 }, { "epoch": 0.49544419134396356, "grad_norm": 2.4211316108703613, "learning_rate": 4.174259681093394e-05, "loss": 0.0642, "step": 870 }, { "epoch": 0.5011389521640092, "grad_norm": 3.668614625930786, "learning_rate": 4.164768413059985e-05, "loss": 0.1067, "step": 880 }, { "epoch": 0.5068337129840547, "grad_norm": 0.860970675945282, "learning_rate": 4.155277145026576e-05, "loss": 0.0583, "step": 890 }, { "epoch": 0.5125284738041003, "grad_norm": 5.353702068328857, "learning_rate": 4.1457858769931665e-05, "loss": 0.0761, "step": 900 }, { "epoch": 0.5182232346241458, "grad_norm": 4.8318681716918945, "learning_rate": 4.136294608959757e-05, "loss": 0.0878, "step": 910 }, { "epoch": 0.5239179954441914, "grad_norm": 0.8994714617729187, "learning_rate": 4.126803340926348e-05, "loss": 0.0502, "step": 920 }, { "epoch": 0.5296127562642369, "grad_norm": 1.8920825719833374, "learning_rate": 4.117312072892939e-05, "loss": 0.0952, "step": 930 }, { "epoch": 0.5353075170842825, "grad_norm": 2.3858070373535156, "learning_rate": 4.1078208048595294e-05, "loss": 0.0797, "step": 940 }, { "epoch": 0.541002277904328, "grad_norm": 3.2179317474365234, "learning_rate": 4.0983295368261205e-05, "loss": 0.1181, "step": 950 }, { "epoch": 0.5466970387243736, "grad_norm": 0.5987422466278076, "learning_rate": 4.088838268792711e-05, "loss": 0.0765, "step": 960 }, { "epoch": 0.5523917995444191, "grad_norm": 1.7444469928741455, "learning_rate": 4.079347000759301e-05, "loss": 0.0519, "step": 970 }, { "epoch": 0.5580865603644647, "grad_norm": 0.14087039232254028, "learning_rate": 4.0698557327258924e-05, "loss": 0.0438, "step": 980 }, { "epoch": 0.5637813211845103, "grad_norm": 2.490300416946411, "learning_rate": 4.0603644646924834e-05, "loss": 0.0796, "step": 990 }, { "epoch": 0.5694760820045558, "grad_norm": 0.5218722820281982, "learning_rate": 4.050873196659074e-05, "loss": 0.0595, "step": 1000 }, { "epoch": 0.5751708428246014, "grad_norm": 2.9636070728302, "learning_rate": 4.041381928625664e-05, "loss": 0.0439, "step": 1010 }, { "epoch": 0.5808656036446469, "grad_norm": 1.2304065227508545, "learning_rate": 4.031890660592255e-05, "loss": 0.0421, "step": 1020 }, { "epoch": 0.5865603644646925, "grad_norm": 0.5061107277870178, "learning_rate": 4.022399392558846e-05, "loss": 0.0698, "step": 1030 }, { "epoch": 0.592255125284738, "grad_norm": 2.6270368099212646, "learning_rate": 4.012908124525437e-05, "loss": 0.0456, "step": 1040 }, { "epoch": 0.5979498861047836, "grad_norm": 2.7538509368896484, "learning_rate": 4.003416856492028e-05, "loss": 0.0934, "step": 1050 }, { "epoch": 0.6036446469248291, "grad_norm": 0.8339662551879883, "learning_rate": 3.993925588458618e-05, "loss": 0.0493, "step": 1060 }, { "epoch": 0.6093394077448747, "grad_norm": 2.7625882625579834, "learning_rate": 3.9844343204252087e-05, "loss": 0.0661, "step": 1070 }, { "epoch": 0.6150341685649203, "grad_norm": 0.10445995628833771, "learning_rate": 3.9749430523918e-05, "loss": 0.0565, "step": 1080 }, { "epoch": 0.6207289293849658, "grad_norm": 4.233523845672607, "learning_rate": 3.965451784358391e-05, "loss": 0.0608, "step": 1090 }, { "epoch": 0.6264236902050114, "grad_norm": 2.9015579223632812, "learning_rate": 3.955960516324981e-05, "loss": 0.0394, "step": 1100 }, { "epoch": 0.6321184510250569, "grad_norm": 1.122717261314392, "learning_rate": 3.946469248291572e-05, "loss": 0.0297, "step": 1110 }, { "epoch": 0.6378132118451025, "grad_norm": 0.9211141467094421, "learning_rate": 3.936977980258163e-05, "loss": 0.0733, "step": 1120 }, { "epoch": 0.643507972665148, "grad_norm": 1.9820640087127686, "learning_rate": 3.927486712224753e-05, "loss": 0.0814, "step": 1130 }, { "epoch": 0.6492027334851936, "grad_norm": 0.8466345071792603, "learning_rate": 3.917995444191344e-05, "loss": 0.0933, "step": 1140 }, { "epoch": 0.6548974943052391, "grad_norm": 1.1416703462600708, "learning_rate": 3.908504176157935e-05, "loss": 0.0826, "step": 1150 }, { "epoch": 0.6605922551252847, "grad_norm": 1.0590317249298096, "learning_rate": 3.8990129081245256e-05, "loss": 0.1039, "step": 1160 }, { "epoch": 0.6662870159453302, "grad_norm": 0.13635486364364624, "learning_rate": 3.889521640091117e-05, "loss": 0.0717, "step": 1170 }, { "epoch": 0.6719817767653758, "grad_norm": 0.7193166017532349, "learning_rate": 3.880030372057707e-05, "loss": 0.0554, "step": 1180 }, { "epoch": 0.6776765375854215, "grad_norm": 1.9214146137237549, "learning_rate": 3.8705391040242975e-05, "loss": 0.0405, "step": 1190 }, { "epoch": 0.683371298405467, "grad_norm": 3.638381242752075, "learning_rate": 3.8610478359908886e-05, "loss": 0.069, "step": 1200 }, { "epoch": 0.6890660592255126, "grad_norm": 0.2034996747970581, "learning_rate": 3.8515565679574796e-05, "loss": 0.0473, "step": 1210 }, { "epoch": 0.6947608200455581, "grad_norm": 2.54274582862854, "learning_rate": 3.84206529992407e-05, "loss": 0.0405, "step": 1220 }, { "epoch": 0.7004555808656037, "grad_norm": 1.0374627113342285, "learning_rate": 3.8325740318906604e-05, "loss": 0.0593, "step": 1230 }, { "epoch": 0.7061503416856492, "grad_norm": 1.1265807151794434, "learning_rate": 3.8230827638572515e-05, "loss": 0.0367, "step": 1240 }, { "epoch": 0.7118451025056948, "grad_norm": 1.830880045890808, "learning_rate": 3.813591495823842e-05, "loss": 0.0589, "step": 1250 }, { "epoch": 0.7175398633257403, "grad_norm": 1.4103747606277466, "learning_rate": 3.804100227790433e-05, "loss": 0.0602, "step": 1260 }, { "epoch": 0.7232346241457859, "grad_norm": 1.1360421180725098, "learning_rate": 3.794608959757024e-05, "loss": 0.0843, "step": 1270 }, { "epoch": 0.7289293849658315, "grad_norm": 1.648220181465149, "learning_rate": 3.7851176917236145e-05, "loss": 0.1067, "step": 1280 }, { "epoch": 0.734624145785877, "grad_norm": 2.033055305480957, "learning_rate": 3.775626423690205e-05, "loss": 0.0531, "step": 1290 }, { "epoch": 0.7403189066059226, "grad_norm": 2.301596164703369, "learning_rate": 3.766135155656796e-05, "loss": 0.0332, "step": 1300 }, { "epoch": 0.7460136674259681, "grad_norm": 0.5601398944854736, "learning_rate": 3.756643887623387e-05, "loss": 0.0452, "step": 1310 }, { "epoch": 0.7517084282460137, "grad_norm": 4.6115312576293945, "learning_rate": 3.7471526195899774e-05, "loss": 0.0737, "step": 1320 }, { "epoch": 0.7574031890660592, "grad_norm": 2.644883632659912, "learning_rate": 3.7376613515565685e-05, "loss": 0.0558, "step": 1330 }, { "epoch": 0.7630979498861048, "grad_norm": 2.106394052505493, "learning_rate": 3.728170083523159e-05, "loss": 0.033, "step": 1340 }, { "epoch": 0.7687927107061503, "grad_norm": 1.5199148654937744, "learning_rate": 3.718678815489749e-05, "loss": 0.0811, "step": 1350 }, { "epoch": 0.7744874715261959, "grad_norm": 2.3490426540374756, "learning_rate": 3.7091875474563403e-05, "loss": 0.0554, "step": 1360 }, { "epoch": 0.7801822323462415, "grad_norm": 2.133591413497925, "learning_rate": 3.6996962794229314e-05, "loss": 0.0792, "step": 1370 }, { "epoch": 0.785876993166287, "grad_norm": 0.3369489014148712, "learning_rate": 3.690205011389522e-05, "loss": 0.0436, "step": 1380 }, { "epoch": 0.7915717539863326, "grad_norm": 1.3656107187271118, "learning_rate": 3.680713743356112e-05, "loss": 0.0543, "step": 1390 }, { "epoch": 0.7972665148063781, "grad_norm": 0.7134440541267395, "learning_rate": 3.671222475322703e-05, "loss": 0.0329, "step": 1400 }, { "epoch": 0.8029612756264237, "grad_norm": 0.5661123991012573, "learning_rate": 3.661731207289294e-05, "loss": 0.0285, "step": 1410 }, { "epoch": 0.8086560364464692, "grad_norm": 1.2485991716384888, "learning_rate": 3.652239939255885e-05, "loss": 0.0908, "step": 1420 }, { "epoch": 0.8143507972665148, "grad_norm": 2.110455274581909, "learning_rate": 3.642748671222476e-05, "loss": 0.0284, "step": 1430 }, { "epoch": 0.8200455580865603, "grad_norm": 12.661782264709473, "learning_rate": 3.633257403189066e-05, "loss": 0.0719, "step": 1440 }, { "epoch": 0.8257403189066059, "grad_norm": 5.139870643615723, "learning_rate": 3.6237661351556566e-05, "loss": 0.084, "step": 1450 }, { "epoch": 0.8314350797266514, "grad_norm": 1.949857234954834, "learning_rate": 3.614274867122248e-05, "loss": 0.0396, "step": 1460 }, { "epoch": 0.837129840546697, "grad_norm": 3.2489545345306396, "learning_rate": 3.604783599088839e-05, "loss": 0.0521, "step": 1470 }, { "epoch": 0.8428246013667426, "grad_norm": 0.4927772879600525, "learning_rate": 3.595292331055429e-05, "loss": 0.0584, "step": 1480 }, { "epoch": 0.8485193621867881, "grad_norm": 0.08044728636741638, "learning_rate": 3.58580106302202e-05, "loss": 0.0478, "step": 1490 }, { "epoch": 0.8542141230068337, "grad_norm": 1.3953468799591064, "learning_rate": 3.5763097949886106e-05, "loss": 0.097, "step": 1500 }, { "epoch": 0.8599088838268792, "grad_norm": 0.017015213146805763, "learning_rate": 3.566818526955201e-05, "loss": 0.0659, "step": 1510 }, { "epoch": 0.8656036446469249, "grad_norm": 0.984900951385498, "learning_rate": 3.557327258921792e-05, "loss": 0.0674, "step": 1520 }, { "epoch": 0.8712984054669703, "grad_norm": 1.6620683670043945, "learning_rate": 3.547835990888383e-05, "loss": 0.0546, "step": 1530 }, { "epoch": 0.876993166287016, "grad_norm": 4.897571086883545, "learning_rate": 3.5383447228549736e-05, "loss": 0.0754, "step": 1540 }, { "epoch": 0.8826879271070615, "grad_norm": 1.2931139469146729, "learning_rate": 3.528853454821565e-05, "loss": 0.0508, "step": 1550 }, { "epoch": 0.8883826879271071, "grad_norm": 4.596437931060791, "learning_rate": 3.519362186788155e-05, "loss": 0.085, "step": 1560 }, { "epoch": 0.8940774487471527, "grad_norm": 2.0353763103485107, "learning_rate": 3.5098709187547455e-05, "loss": 0.0462, "step": 1570 }, { "epoch": 0.8997722095671982, "grad_norm": 1.4557256698608398, "learning_rate": 3.5003796507213365e-05, "loss": 0.067, "step": 1580 }, { "epoch": 0.9054669703872438, "grad_norm": 1.3519636392593384, "learning_rate": 3.4908883826879276e-05, "loss": 0.0541, "step": 1590 }, { "epoch": 0.9111617312072893, "grad_norm": 0.38374024629592896, "learning_rate": 3.481397114654518e-05, "loss": 0.0696, "step": 1600 }, { "epoch": 0.9168564920273349, "grad_norm": 0.17080959677696228, "learning_rate": 3.4719058466211084e-05, "loss": 0.0974, "step": 1610 }, { "epoch": 0.9225512528473804, "grad_norm": 0.33803385496139526, "learning_rate": 3.4624145785876995e-05, "loss": 0.0237, "step": 1620 }, { "epoch": 0.928246013667426, "grad_norm": 1.0675870180130005, "learning_rate": 3.45292331055429e-05, "loss": 0.1002, "step": 1630 }, { "epoch": 0.9339407744874715, "grad_norm": 0.23425699770450592, "learning_rate": 3.443432042520881e-05, "loss": 0.028, "step": 1640 }, { "epoch": 0.9396355353075171, "grad_norm": 1.695246696472168, "learning_rate": 3.433940774487472e-05, "loss": 0.0434, "step": 1650 }, { "epoch": 0.9453302961275627, "grad_norm": 1.134783148765564, "learning_rate": 3.4244495064540624e-05, "loss": 0.0608, "step": 1660 }, { "epoch": 0.9510250569476082, "grad_norm": 0.09808668494224548, "learning_rate": 3.414958238420653e-05, "loss": 0.027, "step": 1670 }, { "epoch": 0.9567198177676538, "grad_norm": 1.7201132774353027, "learning_rate": 3.405466970387244e-05, "loss": 0.0744, "step": 1680 }, { "epoch": 0.9624145785876993, "grad_norm": 0.9982559084892273, "learning_rate": 3.395975702353835e-05, "loss": 0.039, "step": 1690 }, { "epoch": 0.9681093394077449, "grad_norm": 0.573174238204956, "learning_rate": 3.3864844343204254e-05, "loss": 0.0448, "step": 1700 }, { "epoch": 0.9738041002277904, "grad_norm": 0.15374533832073212, "learning_rate": 3.3769931662870164e-05, "loss": 0.0307, "step": 1710 }, { "epoch": 0.979498861047836, "grad_norm": 0.28979602456092834, "learning_rate": 3.367501898253607e-05, "loss": 0.0312, "step": 1720 }, { "epoch": 0.9851936218678815, "grad_norm": 0.4142756760120392, "learning_rate": 3.358010630220197e-05, "loss": 0.0671, "step": 1730 }, { "epoch": 0.9908883826879271, "grad_norm": 1.2099072933197021, "learning_rate": 3.348519362186788e-05, "loss": 0.0613, "step": 1740 }, { "epoch": 0.9965831435079726, "grad_norm": 2.8961591720581055, "learning_rate": 3.3390280941533794e-05, "loss": 0.0468, "step": 1750 }, { "epoch": 1.0, "eval_accuracy": 0.9834836297060943, "eval_f1": 0.9111092563225106, "eval_loss": 0.06205601245164871, "eval_precision": 0.9033432638199271, "eval_recall": 0.9190099343323792, "eval_runtime": 13.7445, "eval_samples_per_second": 236.458, "eval_steps_per_second": 29.612, "step": 1756 }, { "epoch": 1.0022779043280183, "grad_norm": 1.1877965927124023, "learning_rate": 3.32953682611997e-05, "loss": 0.0464, "step": 1760 }, { "epoch": 1.0079726651480638, "grad_norm": 1.4486593008041382, "learning_rate": 3.32004555808656e-05, "loss": 0.0177, "step": 1770 }, { "epoch": 1.0136674259681093, "grad_norm": 0.045917004346847534, "learning_rate": 3.310554290053151e-05, "loss": 0.0293, "step": 1780 }, { "epoch": 1.0193621867881548, "grad_norm": 0.501386821269989, "learning_rate": 3.3010630220197417e-05, "loss": 0.0117, "step": 1790 }, { "epoch": 1.0250569476082005, "grad_norm": 4.940465927124023, "learning_rate": 3.291571753986333e-05, "loss": 0.0227, "step": 1800 }, { "epoch": 1.030751708428246, "grad_norm": 0.1896640509366989, "learning_rate": 3.282080485952924e-05, "loss": 0.0157, "step": 1810 }, { "epoch": 1.0364464692482915, "grad_norm": 0.2719975411891937, "learning_rate": 3.272589217919514e-05, "loss": 0.0366, "step": 1820 }, { "epoch": 1.042141230068337, "grad_norm": 1.6215999126434326, "learning_rate": 3.2630979498861046e-05, "loss": 0.0206, "step": 1830 }, { "epoch": 1.0478359908883828, "grad_norm": 0.15671390295028687, "learning_rate": 3.253606681852696e-05, "loss": 0.0142, "step": 1840 }, { "epoch": 1.0535307517084282, "grad_norm": 0.4000178575515747, "learning_rate": 3.244115413819286e-05, "loss": 0.0246, "step": 1850 }, { "epoch": 1.0592255125284737, "grad_norm": 0.03926578164100647, "learning_rate": 3.234624145785877e-05, "loss": 0.0249, "step": 1860 }, { "epoch": 1.0649202733485195, "grad_norm": 0.6636160612106323, "learning_rate": 3.225132877752468e-05, "loss": 0.0614, "step": 1870 }, { "epoch": 1.070615034168565, "grad_norm": 0.16424879431724548, "learning_rate": 3.2156416097190586e-05, "loss": 0.0411, "step": 1880 }, { "epoch": 1.0763097949886105, "grad_norm": 4.489077568054199, "learning_rate": 3.206150341685649e-05, "loss": 0.053, "step": 1890 }, { "epoch": 1.082004555808656, "grad_norm": 0.028950968757271767, "learning_rate": 3.19665907365224e-05, "loss": 0.0244, "step": 1900 }, { "epoch": 1.0876993166287017, "grad_norm": 3.7143993377685547, "learning_rate": 3.187167805618831e-05, "loss": 0.0291, "step": 1910 }, { "epoch": 1.0933940774487472, "grad_norm": 2.972627639770508, "learning_rate": 3.1776765375854216e-05, "loss": 0.0206, "step": 1920 }, { "epoch": 1.0990888382687927, "grad_norm": 1.4652332067489624, "learning_rate": 3.1681852695520126e-05, "loss": 0.0169, "step": 1930 }, { "epoch": 1.1047835990888384, "grad_norm": 1.768967866897583, "learning_rate": 3.158694001518603e-05, "loss": 0.0216, "step": 1940 }, { "epoch": 1.1104783599088839, "grad_norm": 0.3902505338191986, "learning_rate": 3.1492027334851934e-05, "loss": 0.0343, "step": 1950 }, { "epoch": 1.1161731207289294, "grad_norm": 0.0570041760802269, "learning_rate": 3.1397114654517845e-05, "loss": 0.0224, "step": 1960 }, { "epoch": 1.1218678815489749, "grad_norm": 3.5211033821105957, "learning_rate": 3.1302201974183756e-05, "loss": 0.0142, "step": 1970 }, { "epoch": 1.1275626423690206, "grad_norm": 4.0271077156066895, "learning_rate": 3.120728929384966e-05, "loss": 0.0188, "step": 1980 }, { "epoch": 1.133257403189066, "grad_norm": 0.030941655859351158, "learning_rate": 3.1112376613515564e-05, "loss": 0.0338, "step": 1990 }, { "epoch": 1.1389521640091116, "grad_norm": 0.20438140630722046, "learning_rate": 3.1017463933181475e-05, "loss": 0.0127, "step": 2000 }, { "epoch": 1.144646924829157, "grad_norm": 0.011589931324124336, "learning_rate": 3.092255125284738e-05, "loss": 0.0188, "step": 2010 }, { "epoch": 1.1503416856492028, "grad_norm": 0.036334265023469925, "learning_rate": 3.082763857251329e-05, "loss": 0.0254, "step": 2020 }, { "epoch": 1.1560364464692483, "grad_norm": 0.12188951671123505, "learning_rate": 3.07327258921792e-05, "loss": 0.0319, "step": 2030 }, { "epoch": 1.1617312072892938, "grad_norm": 3.9816770553588867, "learning_rate": 3.0637813211845104e-05, "loss": 0.0809, "step": 2040 }, { "epoch": 1.1674259681093395, "grad_norm": 2.992051124572754, "learning_rate": 3.054290053151101e-05, "loss": 0.0255, "step": 2050 }, { "epoch": 1.173120728929385, "grad_norm": 0.4495603144168854, "learning_rate": 3.044798785117692e-05, "loss": 0.0351, "step": 2060 }, { "epoch": 1.1788154897494305, "grad_norm": 2.132361888885498, "learning_rate": 3.0353075170842826e-05, "loss": 0.0317, "step": 2070 }, { "epoch": 1.184510250569476, "grad_norm": 1.9215974807739258, "learning_rate": 3.0258162490508733e-05, "loss": 0.0453, "step": 2080 }, { "epoch": 1.1902050113895217, "grad_norm": 3.614654302597046, "learning_rate": 3.0163249810174644e-05, "loss": 0.0302, "step": 2090 }, { "epoch": 1.1958997722095672, "grad_norm": 0.05617928504943848, "learning_rate": 3.0068337129840545e-05, "loss": 0.0376, "step": 2100 }, { "epoch": 1.2015945330296127, "grad_norm": 0.08208244293928146, "learning_rate": 2.9973424449506455e-05, "loss": 0.0393, "step": 2110 }, { "epoch": 1.2072892938496582, "grad_norm": 0.10671478509902954, "learning_rate": 2.9878511769172363e-05, "loss": 0.0152, "step": 2120 }, { "epoch": 1.212984054669704, "grad_norm": 0.021128954365849495, "learning_rate": 2.978359908883827e-05, "loss": 0.0091, "step": 2130 }, { "epoch": 1.2186788154897494, "grad_norm": 0.9037052392959595, "learning_rate": 2.968868640850418e-05, "loss": 0.0213, "step": 2140 }, { "epoch": 1.224373576309795, "grad_norm": 0.2880910336971283, "learning_rate": 2.959377372817008e-05, "loss": 0.0205, "step": 2150 }, { "epoch": 1.2300683371298406, "grad_norm": 0.014215152710676193, "learning_rate": 2.949886104783599e-05, "loss": 0.0177, "step": 2160 }, { "epoch": 1.2357630979498861, "grad_norm": 0.01584666594862938, "learning_rate": 2.94039483675019e-05, "loss": 0.0304, "step": 2170 }, { "epoch": 1.2414578587699316, "grad_norm": 0.027049539610743523, "learning_rate": 2.9309035687167807e-05, "loss": 0.022, "step": 2180 }, { "epoch": 1.2471526195899771, "grad_norm": 4.972292900085449, "learning_rate": 2.9214123006833714e-05, "loss": 0.0345, "step": 2190 }, { "epoch": 1.2528473804100229, "grad_norm": 1.9279311895370483, "learning_rate": 2.9119210326499625e-05, "loss": 0.0078, "step": 2200 }, { "epoch": 1.2585421412300684, "grad_norm": 4.7130255699157715, "learning_rate": 2.9024297646165526e-05, "loss": 0.0195, "step": 2210 }, { "epoch": 1.2642369020501139, "grad_norm": 0.13271582126617432, "learning_rate": 2.8929384965831436e-05, "loss": 0.0147, "step": 2220 }, { "epoch": 1.2699316628701594, "grad_norm": 2.073155641555786, "learning_rate": 2.8834472285497344e-05, "loss": 0.0295, "step": 2230 }, { "epoch": 1.275626423690205, "grad_norm": 0.2916022837162018, "learning_rate": 2.873955960516325e-05, "loss": 0.0108, "step": 2240 }, { "epoch": 1.2813211845102506, "grad_norm": 0.059762142598629, "learning_rate": 2.8644646924829162e-05, "loss": 0.051, "step": 2250 }, { "epoch": 1.287015945330296, "grad_norm": 0.08399085700511932, "learning_rate": 2.8549734244495063e-05, "loss": 0.028, "step": 2260 }, { "epoch": 1.2927107061503418, "grad_norm": 1.698492169380188, "learning_rate": 2.8454821564160973e-05, "loss": 0.0478, "step": 2270 }, { "epoch": 1.2984054669703873, "grad_norm": 0.1704496294260025, "learning_rate": 2.835990888382688e-05, "loss": 0.0403, "step": 2280 }, { "epoch": 1.3041002277904328, "grad_norm": 0.056637831032276154, "learning_rate": 2.8264996203492788e-05, "loss": 0.0215, "step": 2290 }, { "epoch": 1.3097949886104785, "grad_norm": 0.02712276019155979, "learning_rate": 2.8170083523158695e-05, "loss": 0.0216, "step": 2300 }, { "epoch": 1.315489749430524, "grad_norm": 1.20741605758667, "learning_rate": 2.8075170842824606e-05, "loss": 0.0161, "step": 2310 }, { "epoch": 1.3211845102505695, "grad_norm": 2.321183443069458, "learning_rate": 2.7980258162490507e-05, "loss": 0.0611, "step": 2320 }, { "epoch": 1.326879271070615, "grad_norm": 0.0435294434428215, "learning_rate": 2.7885345482156417e-05, "loss": 0.0203, "step": 2330 }, { "epoch": 1.3325740318906605, "grad_norm": 2.0571095943450928, "learning_rate": 2.7790432801822325e-05, "loss": 0.039, "step": 2340 }, { "epoch": 1.3382687927107062, "grad_norm": 0.1287350356578827, "learning_rate": 2.7695520121488232e-05, "loss": 0.0211, "step": 2350 }, { "epoch": 1.3439635535307517, "grad_norm": 0.2631838917732239, "learning_rate": 2.7600607441154143e-05, "loss": 0.0196, "step": 2360 }, { "epoch": 1.3496583143507972, "grad_norm": 0.11948617547750473, "learning_rate": 2.7505694760820043e-05, "loss": 0.0225, "step": 2370 }, { "epoch": 1.355353075170843, "grad_norm": 0.30161651968955994, "learning_rate": 2.7410782080485954e-05, "loss": 0.0425, "step": 2380 }, { "epoch": 1.3610478359908884, "grad_norm": 0.09938052296638489, "learning_rate": 2.731586940015186e-05, "loss": 0.02, "step": 2390 }, { "epoch": 1.366742596810934, "grad_norm": 0.12348034232854843, "learning_rate": 2.722095671981777e-05, "loss": 0.0312, "step": 2400 }, { "epoch": 1.3724373576309796, "grad_norm": 1.968189001083374, "learning_rate": 2.7126044039483676e-05, "loss": 0.0296, "step": 2410 }, { "epoch": 1.3781321184510251, "grad_norm": 2.6236233711242676, "learning_rate": 2.7031131359149587e-05, "loss": 0.026, "step": 2420 }, { "epoch": 1.3838268792710706, "grad_norm": 0.05339749529957771, "learning_rate": 2.6936218678815488e-05, "loss": 0.0235, "step": 2430 }, { "epoch": 1.3895216400911161, "grad_norm": 0.15053918957710266, "learning_rate": 2.68413059984814e-05, "loss": 0.022, "step": 2440 }, { "epoch": 1.3952164009111616, "grad_norm": 0.2259766161441803, "learning_rate": 2.6746393318147306e-05, "loss": 0.0268, "step": 2450 }, { "epoch": 1.4009111617312073, "grad_norm": 0.04061640426516533, "learning_rate": 2.6651480637813213e-05, "loss": 0.0229, "step": 2460 }, { "epoch": 1.4066059225512528, "grad_norm": 0.10259843617677689, "learning_rate": 2.6556567957479124e-05, "loss": 0.0214, "step": 2470 }, { "epoch": 1.4123006833712983, "grad_norm": 0.5361473560333252, "learning_rate": 2.6461655277145024e-05, "loss": 0.0321, "step": 2480 }, { "epoch": 1.417995444191344, "grad_norm": 1.511304497718811, "learning_rate": 2.6366742596810935e-05, "loss": 0.0083, "step": 2490 }, { "epoch": 1.4236902050113895, "grad_norm": 0.5383846759796143, "learning_rate": 2.6271829916476843e-05, "loss": 0.0255, "step": 2500 }, { "epoch": 1.429384965831435, "grad_norm": 6.834687232971191, "learning_rate": 2.617691723614275e-05, "loss": 0.0324, "step": 2510 }, { "epoch": 1.4350797266514808, "grad_norm": 0.4993496835231781, "learning_rate": 2.608200455580866e-05, "loss": 0.047, "step": 2520 }, { "epoch": 1.4407744874715263, "grad_norm": 0.32794803380966187, "learning_rate": 2.5987091875474568e-05, "loss": 0.0127, "step": 2530 }, { "epoch": 1.4464692482915718, "grad_norm": 2.7348170280456543, "learning_rate": 2.589217919514047e-05, "loss": 0.0373, "step": 2540 }, { "epoch": 1.4521640091116172, "grad_norm": 0.02981397695839405, "learning_rate": 2.579726651480638e-05, "loss": 0.0403, "step": 2550 }, { "epoch": 1.4578587699316627, "grad_norm": 0.2732754945755005, "learning_rate": 2.5702353834472287e-05, "loss": 0.0193, "step": 2560 }, { "epoch": 1.4635535307517085, "grad_norm": 0.0971960499882698, "learning_rate": 2.5607441154138194e-05, "loss": 0.015, "step": 2570 }, { "epoch": 1.469248291571754, "grad_norm": 0.6999448537826538, "learning_rate": 2.5512528473804105e-05, "loss": 0.0112, "step": 2580 }, { "epoch": 1.4749430523917995, "grad_norm": 4.170445919036865, "learning_rate": 2.5417615793470005e-05, "loss": 0.0296, "step": 2590 }, { "epoch": 1.4806378132118452, "grad_norm": 0.017274674028158188, "learning_rate": 2.5322703113135916e-05, "loss": 0.0096, "step": 2600 }, { "epoch": 1.4863325740318907, "grad_norm": 2.0988516807556152, "learning_rate": 2.5227790432801824e-05, "loss": 0.0219, "step": 2610 }, { "epoch": 1.4920273348519362, "grad_norm": 0.33241793513298035, "learning_rate": 2.513287775246773e-05, "loss": 0.0337, "step": 2620 }, { "epoch": 1.497722095671982, "grad_norm": 0.34468936920166016, "learning_rate": 2.503796507213364e-05, "loss": 0.0209, "step": 2630 }, { "epoch": 1.5034168564920274, "grad_norm": 0.026434799656271935, "learning_rate": 2.4943052391799546e-05, "loss": 0.017, "step": 2640 }, { "epoch": 1.5091116173120729, "grad_norm": 0.00831619743257761, "learning_rate": 2.4848139711465453e-05, "loss": 0.0318, "step": 2650 }, { "epoch": 1.5148063781321186, "grad_norm": 0.5614336133003235, "learning_rate": 2.475322703113136e-05, "loss": 0.0349, "step": 2660 }, { "epoch": 1.5205011389521639, "grad_norm": 0.6730136871337891, "learning_rate": 2.4658314350797268e-05, "loss": 0.0163, "step": 2670 }, { "epoch": 1.5261958997722096, "grad_norm": 0.08429904282093048, "learning_rate": 2.4563401670463175e-05, "loss": 0.0182, "step": 2680 }, { "epoch": 1.531890660592255, "grad_norm": 0.017787586897611618, "learning_rate": 2.4468488990129082e-05, "loss": 0.0746, "step": 2690 }, { "epoch": 1.5375854214123006, "grad_norm": 0.2954597771167755, "learning_rate": 2.437357630979499e-05, "loss": 0.0093, "step": 2700 }, { "epoch": 1.5432801822323463, "grad_norm": 1.5878958702087402, "learning_rate": 2.4278663629460897e-05, "loss": 0.023, "step": 2710 }, { "epoch": 1.5489749430523918, "grad_norm": 0.09025635570287704, "learning_rate": 2.4183750949126804e-05, "loss": 0.0492, "step": 2720 }, { "epoch": 1.5546697038724373, "grad_norm": 0.3361518979072571, "learning_rate": 2.4088838268792712e-05, "loss": 0.0336, "step": 2730 }, { "epoch": 1.560364464692483, "grad_norm": 1.164107322692871, "learning_rate": 2.399392558845862e-05, "loss": 0.0158, "step": 2740 }, { "epoch": 1.5660592255125285, "grad_norm": 2.972662925720215, "learning_rate": 2.3899012908124527e-05, "loss": 0.0198, "step": 2750 }, { "epoch": 1.571753986332574, "grad_norm": 0.5792711973190308, "learning_rate": 2.3804100227790434e-05, "loss": 0.0186, "step": 2760 }, { "epoch": 1.5774487471526197, "grad_norm": 3.0818066596984863, "learning_rate": 2.370918754745634e-05, "loss": 0.0402, "step": 2770 }, { "epoch": 1.583143507972665, "grad_norm": 0.48252251744270325, "learning_rate": 2.361427486712225e-05, "loss": 0.0335, "step": 2780 }, { "epoch": 1.5888382687927107, "grad_norm": 0.024195630103349686, "learning_rate": 2.3519362186788156e-05, "loss": 0.0482, "step": 2790 }, { "epoch": 1.5945330296127562, "grad_norm": 2.509986400604248, "learning_rate": 2.3424449506454063e-05, "loss": 0.0164, "step": 2800 }, { "epoch": 1.6002277904328017, "grad_norm": 0.16147683560848236, "learning_rate": 2.332953682611997e-05, "loss": 0.0172, "step": 2810 }, { "epoch": 1.6059225512528474, "grad_norm": 6.9054484367370605, "learning_rate": 2.3234624145785878e-05, "loss": 0.0249, "step": 2820 }, { "epoch": 1.611617312072893, "grad_norm": 0.02088954672217369, "learning_rate": 2.3139711465451785e-05, "loss": 0.0442, "step": 2830 }, { "epoch": 1.6173120728929384, "grad_norm": 0.334330677986145, "learning_rate": 2.3044798785117693e-05, "loss": 0.0128, "step": 2840 }, { "epoch": 1.6230068337129842, "grad_norm": 0.7511960864067078, "learning_rate": 2.29498861047836e-05, "loss": 0.0134, "step": 2850 }, { "epoch": 1.6287015945330297, "grad_norm": 0.07080753147602081, "learning_rate": 2.2854973424449508e-05, "loss": 0.015, "step": 2860 }, { "epoch": 1.6343963553530751, "grad_norm": 1.2584781646728516, "learning_rate": 2.2760060744115415e-05, "loss": 0.0277, "step": 2870 }, { "epoch": 1.6400911161731209, "grad_norm": 1.7000540494918823, "learning_rate": 2.2665148063781322e-05, "loss": 0.0361, "step": 2880 }, { "epoch": 1.6457858769931661, "grad_norm": 0.07065621763467789, "learning_rate": 2.257023538344723e-05, "loss": 0.0204, "step": 2890 }, { "epoch": 1.6514806378132119, "grad_norm": 1.1046833992004395, "learning_rate": 2.2475322703113137e-05, "loss": 0.0279, "step": 2900 }, { "epoch": 1.6571753986332574, "grad_norm": 0.12607507407665253, "learning_rate": 2.2380410022779044e-05, "loss": 0.0126, "step": 2910 }, { "epoch": 1.6628701594533029, "grad_norm": 0.016763895750045776, "learning_rate": 2.2285497342444952e-05, "loss": 0.0198, "step": 2920 }, { "epoch": 1.6685649202733486, "grad_norm": 0.35705041885375977, "learning_rate": 2.219058466211086e-05, "loss": 0.0196, "step": 2930 }, { "epoch": 1.674259681093394, "grad_norm": 2.3731415271759033, "learning_rate": 2.2095671981776766e-05, "loss": 0.0195, "step": 2940 }, { "epoch": 1.6799544419134396, "grad_norm": 0.042373090982437134, "learning_rate": 2.2000759301442674e-05, "loss": 0.0532, "step": 2950 }, { "epoch": 1.6856492027334853, "grad_norm": 0.13854794204235077, "learning_rate": 2.190584662110858e-05, "loss": 0.0093, "step": 2960 }, { "epoch": 1.6913439635535308, "grad_norm": 1.4668946266174316, "learning_rate": 2.181093394077449e-05, "loss": 0.0155, "step": 2970 }, { "epoch": 1.6970387243735763, "grad_norm": 0.01016306784003973, "learning_rate": 2.1716021260440396e-05, "loss": 0.0254, "step": 2980 }, { "epoch": 1.702733485193622, "grad_norm": 1.037170171737671, "learning_rate": 2.1621108580106303e-05, "loss": 0.0315, "step": 2990 }, { "epoch": 1.7084282460136673, "grad_norm": 1.5667529106140137, "learning_rate": 2.152619589977221e-05, "loss": 0.026, "step": 3000 }, { "epoch": 1.714123006833713, "grad_norm": 2.287351608276367, "learning_rate": 2.1431283219438118e-05, "loss": 0.0186, "step": 3010 }, { "epoch": 1.7198177676537585, "grad_norm": 0.9441989064216614, "learning_rate": 2.1336370539104025e-05, "loss": 0.029, "step": 3020 }, { "epoch": 1.725512528473804, "grad_norm": 0.032152529805898666, "learning_rate": 2.1241457858769933e-05, "loss": 0.0175, "step": 3030 }, { "epoch": 1.7312072892938497, "grad_norm": 0.23306457698345184, "learning_rate": 2.114654517843584e-05, "loss": 0.0168, "step": 3040 }, { "epoch": 1.7369020501138952, "grad_norm": 0.6797131896018982, "learning_rate": 2.1051632498101747e-05, "loss": 0.0147, "step": 3050 }, { "epoch": 1.7425968109339407, "grad_norm": 0.3807680904865265, "learning_rate": 2.0956719817767655e-05, "loss": 0.0184, "step": 3060 }, { "epoch": 1.7482915717539864, "grad_norm": 0.3467291295528412, "learning_rate": 2.0861807137433562e-05, "loss": 0.0142, "step": 3070 }, { "epoch": 1.753986332574032, "grad_norm": 1.8795088529586792, "learning_rate": 2.076689445709947e-05, "loss": 0.048, "step": 3080 }, { "epoch": 1.7596810933940774, "grad_norm": 2.4153385162353516, "learning_rate": 2.0671981776765377e-05, "loss": 0.017, "step": 3090 }, { "epoch": 1.7653758542141231, "grad_norm": 1.2200242280960083, "learning_rate": 2.0577069096431284e-05, "loss": 0.0244, "step": 3100 }, { "epoch": 1.7710706150341684, "grad_norm": 0.9201875925064087, "learning_rate": 2.048215641609719e-05, "loss": 0.0377, "step": 3110 }, { "epoch": 1.7767653758542141, "grad_norm": 0.008683672174811363, "learning_rate": 2.03872437357631e-05, "loss": 0.0292, "step": 3120 }, { "epoch": 1.7824601366742598, "grad_norm": 0.28299397230148315, "learning_rate": 2.0292331055429006e-05, "loss": 0.0309, "step": 3130 }, { "epoch": 1.7881548974943051, "grad_norm": 0.22945231199264526, "learning_rate": 2.0197418375094914e-05, "loss": 0.0249, "step": 3140 }, { "epoch": 1.7938496583143508, "grad_norm": 0.12214020639657974, "learning_rate": 2.010250569476082e-05, "loss": 0.0089, "step": 3150 }, { "epoch": 1.7995444191343963, "grad_norm": 0.06660139560699463, "learning_rate": 2.000759301442673e-05, "loss": 0.0189, "step": 3160 }, { "epoch": 1.8052391799544418, "grad_norm": 5.765466213226318, "learning_rate": 1.9912680334092636e-05, "loss": 0.0715, "step": 3170 }, { "epoch": 1.8109339407744875, "grad_norm": 0.15886840224266052, "learning_rate": 1.9817767653758543e-05, "loss": 0.0066, "step": 3180 }, { "epoch": 1.816628701594533, "grad_norm": 1.608310341835022, "learning_rate": 1.972285497342445e-05, "loss": 0.027, "step": 3190 }, { "epoch": 1.8223234624145785, "grad_norm": 2.088512420654297, "learning_rate": 1.9627942293090358e-05, "loss": 0.0348, "step": 3200 }, { "epoch": 1.8280182232346243, "grad_norm": 0.33316129446029663, "learning_rate": 1.9533029612756265e-05, "loss": 0.056, "step": 3210 }, { "epoch": 1.8337129840546698, "grad_norm": 0.11885243654251099, "learning_rate": 1.9438116932422173e-05, "loss": 0.0231, "step": 3220 }, { "epoch": 1.8394077448747153, "grad_norm": 5.114065647125244, "learning_rate": 1.934320425208808e-05, "loss": 0.0324, "step": 3230 }, { "epoch": 1.845102505694761, "grad_norm": 3.247770309448242, "learning_rate": 1.9248291571753987e-05, "loss": 0.0437, "step": 3240 }, { "epoch": 1.8507972665148062, "grad_norm": 4.07563591003418, "learning_rate": 1.9153378891419895e-05, "loss": 0.0152, "step": 3250 }, { "epoch": 1.856492027334852, "grad_norm": 0.7992623448371887, "learning_rate": 1.9058466211085802e-05, "loss": 0.04, "step": 3260 }, { "epoch": 1.8621867881548975, "grad_norm": 2.1190977096557617, "learning_rate": 1.896355353075171e-05, "loss": 0.0205, "step": 3270 }, { "epoch": 1.867881548974943, "grad_norm": 0.08607299625873566, "learning_rate": 1.8868640850417617e-05, "loss": 0.0614, "step": 3280 }, { "epoch": 1.8735763097949887, "grad_norm": 0.11223249137401581, "learning_rate": 1.8773728170083524e-05, "loss": 0.0152, "step": 3290 }, { "epoch": 1.8792710706150342, "grad_norm": 0.09186819195747375, "learning_rate": 1.867881548974943e-05, "loss": 0.0607, "step": 3300 }, { "epoch": 1.8849658314350797, "grad_norm": 1.1647542715072632, "learning_rate": 1.858390280941534e-05, "loss": 0.0312, "step": 3310 }, { "epoch": 1.8906605922551254, "grad_norm": 0.19422082602977753, "learning_rate": 1.8488990129081246e-05, "loss": 0.0085, "step": 3320 }, { "epoch": 1.896355353075171, "grad_norm": 2.4002015590667725, "learning_rate": 1.8394077448747154e-05, "loss": 0.0264, "step": 3330 }, { "epoch": 1.9020501138952164, "grad_norm": 0.8677728176116943, "learning_rate": 1.829916476841306e-05, "loss": 0.0296, "step": 3340 }, { "epoch": 1.907744874715262, "grad_norm": 2.0761377811431885, "learning_rate": 1.8204252088078968e-05, "loss": 0.0192, "step": 3350 }, { "epoch": 1.9134396355353074, "grad_norm": 0.7011535167694092, "learning_rate": 1.8109339407744876e-05, "loss": 0.0147, "step": 3360 }, { "epoch": 1.919134396355353, "grad_norm": 0.8465079069137573, "learning_rate": 1.8014426727410783e-05, "loss": 0.0287, "step": 3370 }, { "epoch": 1.9248291571753986, "grad_norm": 0.30490973591804504, "learning_rate": 1.791951404707669e-05, "loss": 0.0111, "step": 3380 }, { "epoch": 1.930523917995444, "grad_norm": 0.2505026161670685, "learning_rate": 1.7824601366742598e-05, "loss": 0.01, "step": 3390 }, { "epoch": 1.9362186788154898, "grad_norm": 0.8068025708198547, "learning_rate": 1.7729688686408505e-05, "loss": 0.0286, "step": 3400 }, { "epoch": 1.9419134396355353, "grad_norm": 2.2277448177337646, "learning_rate": 1.7634776006074412e-05, "loss": 0.0198, "step": 3410 }, { "epoch": 1.9476082004555808, "grad_norm": 0.13520954549312592, "learning_rate": 1.753986332574032e-05, "loss": 0.0232, "step": 3420 }, { "epoch": 1.9533029612756265, "grad_norm": 0.0674479603767395, "learning_rate": 1.7444950645406227e-05, "loss": 0.0176, "step": 3430 }, { "epoch": 1.958997722095672, "grad_norm": 0.018033193424344063, "learning_rate": 1.7350037965072134e-05, "loss": 0.0262, "step": 3440 }, { "epoch": 1.9646924829157175, "grad_norm": 0.09576301276683807, "learning_rate": 1.7255125284738042e-05, "loss": 0.0217, "step": 3450 }, { "epoch": 1.9703872437357632, "grad_norm": 0.9586463570594788, "learning_rate": 1.716021260440395e-05, "loss": 0.0123, "step": 3460 }, { "epoch": 1.9760820045558085, "grad_norm": 0.006530022248625755, "learning_rate": 1.7065299924069857e-05, "loss": 0.0163, "step": 3470 }, { "epoch": 1.9817767653758542, "grad_norm": 1.7253230810165405, "learning_rate": 1.6970387243735764e-05, "loss": 0.0126, "step": 3480 }, { "epoch": 1.9874715261958997, "grad_norm": 0.07670151442289352, "learning_rate": 1.687547456340167e-05, "loss": 0.0096, "step": 3490 }, { "epoch": 1.9931662870159452, "grad_norm": 0.013690901920199394, "learning_rate": 1.678056188306758e-05, "loss": 0.0168, "step": 3500 }, { "epoch": 1.998861047835991, "grad_norm": 0.4434203505516052, "learning_rate": 1.6685649202733486e-05, "loss": 0.0178, "step": 3510 }, { "epoch": 2.0, "eval_accuracy": 0.9870868472820054, "eval_f1": 0.9356108180524156, "eval_loss": 0.05325901880860329, "eval_precision": 0.9305463024650233, "eval_recall": 0.9407307627546725, "eval_runtime": 13.1682, "eval_samples_per_second": 246.807, "eval_steps_per_second": 30.908, "step": 3512 }, { "epoch": 2.0045558086560367, "grad_norm": 1.626996397972107, "learning_rate": 1.6590736522399393e-05, "loss": 0.0158, "step": 3520 }, { "epoch": 2.010250569476082, "grad_norm": 0.17399589717388153, "learning_rate": 1.64958238420653e-05, "loss": 0.0077, "step": 3530 }, { "epoch": 2.0159453302961277, "grad_norm": 0.027639882639050484, "learning_rate": 1.6400911161731208e-05, "loss": 0.0236, "step": 3540 }, { "epoch": 2.021640091116173, "grad_norm": 0.7574236392974854, "learning_rate": 1.6305998481397115e-05, "loss": 0.0153, "step": 3550 }, { "epoch": 2.0273348519362187, "grad_norm": 0.023957232013344765, "learning_rate": 1.6211085801063023e-05, "loss": 0.0046, "step": 3560 }, { "epoch": 2.0330296127562644, "grad_norm": 2.448194980621338, "learning_rate": 1.611617312072893e-05, "loss": 0.011, "step": 3570 }, { "epoch": 2.0387243735763096, "grad_norm": 0.5603412985801697, "learning_rate": 1.6021260440394838e-05, "loss": 0.0129, "step": 3580 }, { "epoch": 2.0444191343963554, "grad_norm": 0.7681556940078735, "learning_rate": 1.5926347760060745e-05, "loss": 0.0129, "step": 3590 }, { "epoch": 2.050113895216401, "grad_norm": 0.09505771845579147, "learning_rate": 1.5831435079726652e-05, "loss": 0.0093, "step": 3600 }, { "epoch": 2.0558086560364464, "grad_norm": 0.05849296599626541, "learning_rate": 1.573652239939256e-05, "loss": 0.0311, "step": 3610 }, { "epoch": 2.061503416856492, "grad_norm": 0.5222179293632507, "learning_rate": 1.5641609719058467e-05, "loss": 0.0083, "step": 3620 }, { "epoch": 2.067198177676538, "grad_norm": 1.0224918127059937, "learning_rate": 1.5546697038724374e-05, "loss": 0.0159, "step": 3630 }, { "epoch": 2.072892938496583, "grad_norm": 1.3902862071990967, "learning_rate": 1.545178435839028e-05, "loss": 0.0077, "step": 3640 }, { "epoch": 2.078587699316629, "grad_norm": 0.2441783845424652, "learning_rate": 1.535687167805619e-05, "loss": 0.0079, "step": 3650 }, { "epoch": 2.084282460136674, "grad_norm": 0.012967357411980629, "learning_rate": 1.5261958997722096e-05, "loss": 0.0068, "step": 3660 }, { "epoch": 2.08997722095672, "grad_norm": 0.5421155095100403, "learning_rate": 1.5167046317388004e-05, "loss": 0.0108, "step": 3670 }, { "epoch": 2.0956719817767655, "grad_norm": 0.24783940613269806, "learning_rate": 1.5072133637053911e-05, "loss": 0.0023, "step": 3680 }, { "epoch": 2.1013667425968108, "grad_norm": 0.03946458548307419, "learning_rate": 1.4977220956719817e-05, "loss": 0.0036, "step": 3690 }, { "epoch": 2.1070615034168565, "grad_norm": 0.009126793593168259, "learning_rate": 1.4882308276385726e-05, "loss": 0.0064, "step": 3700 }, { "epoch": 2.112756264236902, "grad_norm": 1.0558329820632935, "learning_rate": 1.4787395596051635e-05, "loss": 0.0095, "step": 3710 }, { "epoch": 2.1184510250569475, "grad_norm": 0.3603816032409668, "learning_rate": 1.469248291571754e-05, "loss": 0.0038, "step": 3720 }, { "epoch": 2.124145785876993, "grad_norm": 0.16135656833648682, "learning_rate": 1.4597570235383448e-05, "loss": 0.0019, "step": 3730 }, { "epoch": 2.129840546697039, "grad_norm": 0.006461753975600004, "learning_rate": 1.4502657555049357e-05, "loss": 0.0122, "step": 3740 }, { "epoch": 2.135535307517084, "grad_norm": 1.8980814218521118, "learning_rate": 1.4407744874715263e-05, "loss": 0.0058, "step": 3750 }, { "epoch": 2.14123006833713, "grad_norm": 1.94465970993042, "learning_rate": 1.431283219438117e-05, "loss": 0.0196, "step": 3760 }, { "epoch": 2.146924829157175, "grad_norm": 0.01323908194899559, "learning_rate": 1.4217919514047076e-05, "loss": 0.0141, "step": 3770 }, { "epoch": 2.152619589977221, "grad_norm": 0.07417018711566925, "learning_rate": 1.4123006833712985e-05, "loss": 0.0076, "step": 3780 }, { "epoch": 2.1583143507972666, "grad_norm": 0.02077096700668335, "learning_rate": 1.4028094153378892e-05, "loss": 0.0033, "step": 3790 }, { "epoch": 2.164009111617312, "grad_norm": 0.01688476838171482, "learning_rate": 1.3933181473044798e-05, "loss": 0.0198, "step": 3800 }, { "epoch": 2.1697038724373576, "grad_norm": 0.04020548611879349, "learning_rate": 1.3838268792710707e-05, "loss": 0.0214, "step": 3810 }, { "epoch": 2.1753986332574033, "grad_norm": 0.01843833364546299, "learning_rate": 1.3743356112376616e-05, "loss": 0.0049, "step": 3820 }, { "epoch": 2.1810933940774486, "grad_norm": 0.12003644555807114, "learning_rate": 1.3648443432042522e-05, "loss": 0.006, "step": 3830 }, { "epoch": 2.1867881548974943, "grad_norm": 1.9108054637908936, "learning_rate": 1.3553530751708429e-05, "loss": 0.0211, "step": 3840 }, { "epoch": 2.19248291571754, "grad_norm": 0.5751833319664001, "learning_rate": 1.3458618071374338e-05, "loss": 0.008, "step": 3850 }, { "epoch": 2.1981776765375853, "grad_norm": 0.5237908959388733, "learning_rate": 1.3363705391040244e-05, "loss": 0.0122, "step": 3860 }, { "epoch": 2.203872437357631, "grad_norm": 1.8080928325653076, "learning_rate": 1.3268792710706151e-05, "loss": 0.0074, "step": 3870 }, { "epoch": 2.2095671981776768, "grad_norm": 1.2654815912246704, "learning_rate": 1.3173880030372057e-05, "loss": 0.0098, "step": 3880 }, { "epoch": 2.215261958997722, "grad_norm": 0.4974677860736847, "learning_rate": 1.3078967350037966e-05, "loss": 0.0038, "step": 3890 }, { "epoch": 2.2209567198177678, "grad_norm": 2.2507166862487793, "learning_rate": 1.2984054669703875e-05, "loss": 0.0184, "step": 3900 }, { "epoch": 2.226651480637813, "grad_norm": 0.04915757104754448, "learning_rate": 1.2889141989369779e-05, "loss": 0.003, "step": 3910 }, { "epoch": 2.2323462414578588, "grad_norm": 0.03872552514076233, "learning_rate": 1.2794229309035688e-05, "loss": 0.0138, "step": 3920 }, { "epoch": 2.2380410022779045, "grad_norm": 0.7878251671791077, "learning_rate": 1.2699316628701597e-05, "loss": 0.0129, "step": 3930 }, { "epoch": 2.2437357630979498, "grad_norm": 0.011351537890732288, "learning_rate": 1.2604403948367503e-05, "loss": 0.0159, "step": 3940 }, { "epoch": 2.2494305239179955, "grad_norm": 0.012801927514374256, "learning_rate": 1.250949126803341e-05, "loss": 0.0084, "step": 3950 }, { "epoch": 2.255125284738041, "grad_norm": 0.8789533376693726, "learning_rate": 1.2414578587699317e-05, "loss": 0.0137, "step": 3960 }, { "epoch": 2.2608200455580865, "grad_norm": 0.008565433323383331, "learning_rate": 1.2319665907365225e-05, "loss": 0.0018, "step": 3970 }, { "epoch": 2.266514806378132, "grad_norm": 0.04799957945942879, "learning_rate": 1.2224753227031132e-05, "loss": 0.0114, "step": 3980 }, { "epoch": 2.2722095671981775, "grad_norm": 0.27682027220726013, "learning_rate": 1.212984054669704e-05, "loss": 0.0102, "step": 3990 }, { "epoch": 2.277904328018223, "grad_norm": 0.008524512872099876, "learning_rate": 1.2034927866362947e-05, "loss": 0.016, "step": 4000 }, { "epoch": 2.283599088838269, "grad_norm": 0.14340077340602875, "learning_rate": 1.1940015186028854e-05, "loss": 0.0052, "step": 4010 }, { "epoch": 2.289293849658314, "grad_norm": 0.16430233418941498, "learning_rate": 1.1845102505694761e-05, "loss": 0.0066, "step": 4020 }, { "epoch": 2.29498861047836, "grad_norm": 0.14387188851833344, "learning_rate": 1.1750189825360669e-05, "loss": 0.0066, "step": 4030 }, { "epoch": 2.3006833712984056, "grad_norm": 0.16657856106758118, "learning_rate": 1.1655277145026576e-05, "loss": 0.0172, "step": 4040 }, { "epoch": 2.306378132118451, "grad_norm": 0.9251142740249634, "learning_rate": 1.1560364464692483e-05, "loss": 0.0204, "step": 4050 }, { "epoch": 2.3120728929384966, "grad_norm": 0.04931972548365593, "learning_rate": 1.146545178435839e-05, "loss": 0.0097, "step": 4060 }, { "epoch": 2.3177676537585423, "grad_norm": 4.824743747711182, "learning_rate": 1.1370539104024298e-05, "loss": 0.0169, "step": 4070 }, { "epoch": 2.3234624145785876, "grad_norm": 0.01028421986848116, "learning_rate": 1.1275626423690206e-05, "loss": 0.0049, "step": 4080 }, { "epoch": 2.3291571753986333, "grad_norm": 0.04389754310250282, "learning_rate": 1.1180713743356113e-05, "loss": 0.0051, "step": 4090 }, { "epoch": 2.334851936218679, "grad_norm": 0.7398139238357544, "learning_rate": 1.108580106302202e-05, "loss": 0.0062, "step": 4100 }, { "epoch": 2.3405466970387243, "grad_norm": 0.2955804467201233, "learning_rate": 1.0990888382687928e-05, "loss": 0.0092, "step": 4110 }, { "epoch": 2.34624145785877, "grad_norm": 0.026279909536242485, "learning_rate": 1.0895975702353835e-05, "loss": 0.0241, "step": 4120 }, { "epoch": 2.3519362186788153, "grad_norm": 2.0768744945526123, "learning_rate": 1.0801063022019742e-05, "loss": 0.0175, "step": 4130 }, { "epoch": 2.357630979498861, "grad_norm": 1.5808554887771606, "learning_rate": 1.070615034168565e-05, "loss": 0.0137, "step": 4140 }, { "epoch": 2.3633257403189067, "grad_norm": 2.1162478923797607, "learning_rate": 1.0611237661351557e-05, "loss": 0.011, "step": 4150 }, { "epoch": 2.369020501138952, "grad_norm": 1.039797306060791, "learning_rate": 1.0516324981017464e-05, "loss": 0.026, "step": 4160 }, { "epoch": 2.3747152619589977, "grad_norm": 0.1136997640132904, "learning_rate": 1.0421412300683372e-05, "loss": 0.0123, "step": 4170 }, { "epoch": 2.3804100227790435, "grad_norm": 0.648858904838562, "learning_rate": 1.032649962034928e-05, "loss": 0.0145, "step": 4180 }, { "epoch": 2.3861047835990887, "grad_norm": 0.03447713330388069, "learning_rate": 1.0231586940015187e-05, "loss": 0.0014, "step": 4190 }, { "epoch": 2.3917995444191344, "grad_norm": 0.10330228507518768, "learning_rate": 1.0136674259681094e-05, "loss": 0.0191, "step": 4200 }, { "epoch": 2.3974943052391797, "grad_norm": 0.018602780997753143, "learning_rate": 1.0041761579347001e-05, "loss": 0.0071, "step": 4210 }, { "epoch": 2.4031890660592254, "grad_norm": 0.14088431000709534, "learning_rate": 9.946848899012909e-06, "loss": 0.0028, "step": 4220 }, { "epoch": 2.408883826879271, "grad_norm": 0.19464652240276337, "learning_rate": 9.851936218678816e-06, "loss": 0.0035, "step": 4230 }, { "epoch": 2.4145785876993164, "grad_norm": 2.1453051567077637, "learning_rate": 9.757023538344723e-06, "loss": 0.0067, "step": 4240 }, { "epoch": 2.420273348519362, "grad_norm": 0.0253062192350626, "learning_rate": 9.66211085801063e-06, "loss": 0.0128, "step": 4250 }, { "epoch": 2.425968109339408, "grad_norm": 0.06526502221822739, "learning_rate": 9.567198177676538e-06, "loss": 0.0183, "step": 4260 }, { "epoch": 2.431662870159453, "grad_norm": 0.4081474244594574, "learning_rate": 9.472285497342445e-06, "loss": 0.0013, "step": 4270 }, { "epoch": 2.437357630979499, "grad_norm": 1.4686610698699951, "learning_rate": 9.377372817008353e-06, "loss": 0.0067, "step": 4280 }, { "epoch": 2.4430523917995446, "grad_norm": 0.048870667815208435, "learning_rate": 9.28246013667426e-06, "loss": 0.0156, "step": 4290 }, { "epoch": 2.44874715261959, "grad_norm": 0.04144718497991562, "learning_rate": 9.187547456340167e-06, "loss": 0.007, "step": 4300 }, { "epoch": 2.4544419134396356, "grad_norm": 0.01028160285204649, "learning_rate": 9.092634776006075e-06, "loss": 0.004, "step": 4310 }, { "epoch": 2.4601366742596813, "grad_norm": 0.11557559669017792, "learning_rate": 8.997722095671982e-06, "loss": 0.0014, "step": 4320 }, { "epoch": 2.4658314350797266, "grad_norm": 0.014574232511222363, "learning_rate": 8.90280941533789e-06, "loss": 0.001, "step": 4330 }, { "epoch": 2.4715261958997723, "grad_norm": 0.010948601178824902, "learning_rate": 8.807896735003797e-06, "loss": 0.0381, "step": 4340 }, { "epoch": 2.477220956719818, "grad_norm": 0.011354812420904636, "learning_rate": 8.712984054669704e-06, "loss": 0.0058, "step": 4350 }, { "epoch": 2.4829157175398633, "grad_norm": 0.03828594833612442, "learning_rate": 8.618071374335612e-06, "loss": 0.0022, "step": 4360 }, { "epoch": 2.488610478359909, "grad_norm": 2.171107053756714, "learning_rate": 8.523158694001519e-06, "loss": 0.0201, "step": 4370 }, { "epoch": 2.4943052391799543, "grad_norm": 2.7769832611083984, "learning_rate": 8.428246013667426e-06, "loss": 0.03, "step": 4380 }, { "epoch": 2.5, "grad_norm": 0.3383297324180603, "learning_rate": 8.333333333333334e-06, "loss": 0.0013, "step": 4390 }, { "epoch": 2.5056947608200457, "grad_norm": 0.008414714597165585, "learning_rate": 8.238420652999241e-06, "loss": 0.0015, "step": 4400 }, { "epoch": 2.511389521640091, "grad_norm": 0.7041877508163452, "learning_rate": 8.143507972665148e-06, "loss": 0.0564, "step": 4410 }, { "epoch": 2.5170842824601367, "grad_norm": 2.449838399887085, "learning_rate": 8.048595292331056e-06, "loss": 0.0127, "step": 4420 }, { "epoch": 2.522779043280182, "grad_norm": 5.027844429016113, "learning_rate": 7.953682611996963e-06, "loss": 0.0122, "step": 4430 }, { "epoch": 2.5284738041002277, "grad_norm": 1.5973525047302246, "learning_rate": 7.85876993166287e-06, "loss": 0.0088, "step": 4440 }, { "epoch": 2.5341685649202734, "grad_norm": 0.02022445946931839, "learning_rate": 7.763857251328778e-06, "loss": 0.0011, "step": 4450 }, { "epoch": 2.5398633257403187, "grad_norm": 0.20087704062461853, "learning_rate": 7.668944570994685e-06, "loss": 0.0077, "step": 4460 }, { "epoch": 2.5455580865603644, "grad_norm": 0.006311175879091024, "learning_rate": 7.574031890660592e-06, "loss": 0.0023, "step": 4470 }, { "epoch": 2.55125284738041, "grad_norm": 0.057606592774391174, "learning_rate": 7.4791192103265e-06, "loss": 0.0036, "step": 4480 }, { "epoch": 2.5569476082004554, "grad_norm": 0.029715605080127716, "learning_rate": 7.384206529992407e-06, "loss": 0.0085, "step": 4490 }, { "epoch": 2.562642369020501, "grad_norm": 0.17613981664180756, "learning_rate": 7.289293849658315e-06, "loss": 0.0073, "step": 4500 }, { "epoch": 2.568337129840547, "grad_norm": 0.09365126490592957, "learning_rate": 7.194381169324223e-06, "loss": 0.0119, "step": 4510 }, { "epoch": 2.574031890660592, "grad_norm": 0.19684773683547974, "learning_rate": 7.099468488990129e-06, "loss": 0.0064, "step": 4520 }, { "epoch": 2.579726651480638, "grad_norm": 1.106364130973816, "learning_rate": 7.004555808656037e-06, "loss": 0.0025, "step": 4530 }, { "epoch": 2.5854214123006836, "grad_norm": 0.040543317794799805, "learning_rate": 6.909643128321943e-06, "loss": 0.0065, "step": 4540 }, { "epoch": 2.591116173120729, "grad_norm": 0.00573074072599411, "learning_rate": 6.814730447987852e-06, "loss": 0.0195, "step": 4550 }, { "epoch": 2.5968109339407746, "grad_norm": 0.04998818784952164, "learning_rate": 6.719817767653759e-06, "loss": 0.0012, "step": 4560 }, { "epoch": 2.6025056947608203, "grad_norm": 0.502973198890686, "learning_rate": 6.624905087319666e-06, "loss": 0.0172, "step": 4570 }, { "epoch": 2.6082004555808656, "grad_norm": 0.06730928272008896, "learning_rate": 6.529992406985573e-06, "loss": 0.003, "step": 4580 }, { "epoch": 2.6138952164009113, "grad_norm": 0.007552881259471178, "learning_rate": 6.435079726651482e-06, "loss": 0.0045, "step": 4590 }, { "epoch": 2.619589977220957, "grad_norm": 0.015098507516086102, "learning_rate": 6.340167046317388e-06, "loss": 0.0039, "step": 4600 }, { "epoch": 2.6252847380410023, "grad_norm": 1.1533013582229614, "learning_rate": 6.245254365983296e-06, "loss": 0.0024, "step": 4610 }, { "epoch": 2.630979498861048, "grad_norm": 1.244232177734375, "learning_rate": 6.150341685649203e-06, "loss": 0.0234, "step": 4620 }, { "epoch": 2.6366742596810933, "grad_norm": 0.011134306900203228, "learning_rate": 6.05542900531511e-06, "loss": 0.0126, "step": 4630 }, { "epoch": 2.642369020501139, "grad_norm": 0.008693058975040913, "learning_rate": 5.960516324981018e-06, "loss": 0.0013, "step": 4640 }, { "epoch": 2.6480637813211843, "grad_norm": 0.4049939513206482, "learning_rate": 5.865603644646925e-06, "loss": 0.0045, "step": 4650 }, { "epoch": 2.65375854214123, "grad_norm": 0.01917268894612789, "learning_rate": 5.7706909643128325e-06, "loss": 0.0011, "step": 4660 }, { "epoch": 2.6594533029612757, "grad_norm": 0.032446008175611496, "learning_rate": 5.67577828397874e-06, "loss": 0.0034, "step": 4670 }, { "epoch": 2.665148063781321, "grad_norm": 0.3410187065601349, "learning_rate": 5.580865603644647e-06, "loss": 0.0179, "step": 4680 }, { "epoch": 2.6708428246013667, "grad_norm": 0.021253902465105057, "learning_rate": 5.4859529233105546e-06, "loss": 0.0147, "step": 4690 }, { "epoch": 2.6765375854214124, "grad_norm": 0.011832083575427532, "learning_rate": 5.391040242976462e-06, "loss": 0.0157, "step": 4700 }, { "epoch": 2.6822323462414577, "grad_norm": 0.011258887127041817, "learning_rate": 5.296127562642369e-06, "loss": 0.0054, "step": 4710 }, { "epoch": 2.6879271070615034, "grad_norm": 0.6816171407699585, "learning_rate": 5.201214882308277e-06, "loss": 0.0103, "step": 4720 }, { "epoch": 2.693621867881549, "grad_norm": 0.015361886471509933, "learning_rate": 5.106302201974184e-06, "loss": 0.0177, "step": 4730 }, { "epoch": 2.6993166287015944, "grad_norm": 1.1560934782028198, "learning_rate": 5.011389521640091e-06, "loss": 0.0121, "step": 4740 }, { "epoch": 2.70501138952164, "grad_norm": 0.2196737378835678, "learning_rate": 4.916476841305999e-06, "loss": 0.0174, "step": 4750 }, { "epoch": 2.710706150341686, "grad_norm": 0.00927172601222992, "learning_rate": 4.821564160971906e-06, "loss": 0.0069, "step": 4760 }, { "epoch": 2.716400911161731, "grad_norm": 0.0045639206655323505, "learning_rate": 4.7266514806378134e-06, "loss": 0.0041, "step": 4770 }, { "epoch": 2.722095671981777, "grad_norm": 0.026646023616194725, "learning_rate": 4.631738800303721e-06, "loss": 0.0138, "step": 4780 }, { "epoch": 2.7277904328018225, "grad_norm": 0.021237283945083618, "learning_rate": 4.536826119969628e-06, "loss": 0.0036, "step": 4790 }, { "epoch": 2.733485193621868, "grad_norm": 0.008988427929580212, "learning_rate": 4.4419134396355355e-06, "loss": 0.0048, "step": 4800 }, { "epoch": 2.7391799544419135, "grad_norm": 0.04954913258552551, "learning_rate": 4.347000759301443e-06, "loss": 0.0085, "step": 4810 }, { "epoch": 2.7448747152619593, "grad_norm": 1.486095666885376, "learning_rate": 4.25208807896735e-06, "loss": 0.0072, "step": 4820 }, { "epoch": 2.7505694760820045, "grad_norm": 0.6007639169692993, "learning_rate": 4.157175398633258e-06, "loss": 0.0045, "step": 4830 }, { "epoch": 2.7562642369020502, "grad_norm": 0.008939745835959911, "learning_rate": 4.062262718299165e-06, "loss": 0.0034, "step": 4840 }, { "epoch": 2.7619589977220955, "grad_norm": 1.5877684354782104, "learning_rate": 3.967350037965072e-06, "loss": 0.0211, "step": 4850 }, { "epoch": 2.7676537585421412, "grad_norm": 0.029445238411426544, "learning_rate": 3.87243735763098e-06, "loss": 0.0165, "step": 4860 }, { "epoch": 2.7733485193621865, "grad_norm": 0.021546538919210434, "learning_rate": 3.777524677296887e-06, "loss": 0.0023, "step": 4870 }, { "epoch": 2.7790432801822322, "grad_norm": 0.044492632150650024, "learning_rate": 3.682611996962795e-06, "loss": 0.0013, "step": 4880 }, { "epoch": 2.784738041002278, "grad_norm": 0.24219924211502075, "learning_rate": 3.5876993166287018e-06, "loss": 0.0231, "step": 4890 }, { "epoch": 2.7904328018223232, "grad_norm": 0.03521392121911049, "learning_rate": 3.492786636294609e-06, "loss": 0.0014, "step": 4900 }, { "epoch": 2.796127562642369, "grad_norm": 1.6615145206451416, "learning_rate": 3.3978739559605165e-06, "loss": 0.007, "step": 4910 }, { "epoch": 2.8018223234624147, "grad_norm": 0.3814733028411865, "learning_rate": 3.302961275626424e-06, "loss": 0.0028, "step": 4920 }, { "epoch": 2.80751708428246, "grad_norm": 1.6600357294082642, "learning_rate": 3.208048595292331e-06, "loss": 0.0098, "step": 4930 }, { "epoch": 2.8132118451025057, "grad_norm": 0.09767989814281464, "learning_rate": 3.1131359149582386e-06, "loss": 0.0125, "step": 4940 }, { "epoch": 2.8189066059225514, "grad_norm": 0.004134484566748142, "learning_rate": 3.018223234624146e-06, "loss": 0.001, "step": 4950 }, { "epoch": 2.8246013667425967, "grad_norm": 0.3089199960231781, "learning_rate": 2.9233105542900533e-06, "loss": 0.0153, "step": 4960 }, { "epoch": 2.8302961275626424, "grad_norm": 1.1428207159042358, "learning_rate": 2.8283978739559607e-06, "loss": 0.0217, "step": 4970 }, { "epoch": 2.835990888382688, "grad_norm": 1.542417049407959, "learning_rate": 2.733485193621868e-06, "loss": 0.0077, "step": 4980 }, { "epoch": 2.8416856492027334, "grad_norm": 0.7124744653701782, "learning_rate": 2.6385725132877754e-06, "loss": 0.025, "step": 4990 }, { "epoch": 2.847380410022779, "grad_norm": 0.045164983719587326, "learning_rate": 2.5436598329536827e-06, "loss": 0.0071, "step": 5000 }, { "epoch": 2.853075170842825, "grad_norm": 0.14568312466144562, "learning_rate": 2.44874715261959e-06, "loss": 0.0009, "step": 5010 }, { "epoch": 2.85876993166287, "grad_norm": 1.8170379400253296, "learning_rate": 2.3538344722854975e-06, "loss": 0.0071, "step": 5020 }, { "epoch": 2.864464692482916, "grad_norm": 0.018464110791683197, "learning_rate": 2.258921791951405e-06, "loss": 0.018, "step": 5030 }, { "epoch": 2.8701594533029615, "grad_norm": 0.01644555851817131, "learning_rate": 2.164009111617312e-06, "loss": 0.0014, "step": 5040 }, { "epoch": 2.875854214123007, "grad_norm": 1.8252747058868408, "learning_rate": 2.0690964312832195e-06, "loss": 0.0063, "step": 5050 }, { "epoch": 2.8815489749430525, "grad_norm": 2.68369197845459, "learning_rate": 1.974183750949127e-06, "loss": 0.0068, "step": 5060 }, { "epoch": 2.887243735763098, "grad_norm": 0.1895146667957306, "learning_rate": 1.879271070615034e-06, "loss": 0.0013, "step": 5070 }, { "epoch": 2.8929384965831435, "grad_norm": 0.9527952671051025, "learning_rate": 1.7843583902809414e-06, "loss": 0.0214, "step": 5080 }, { "epoch": 2.8986332574031892, "grad_norm": 0.3007192313671112, "learning_rate": 1.6894457099468492e-06, "loss": 0.0059, "step": 5090 }, { "epoch": 2.9043280182232345, "grad_norm": 1.1584299802780151, "learning_rate": 1.5945330296127566e-06, "loss": 0.0323, "step": 5100 }, { "epoch": 2.91002277904328, "grad_norm": 0.17047913372516632, "learning_rate": 1.4996203492786637e-06, "loss": 0.0075, "step": 5110 }, { "epoch": 2.9157175398633255, "grad_norm": 0.019094351679086685, "learning_rate": 1.404707668944571e-06, "loss": 0.0111, "step": 5120 }, { "epoch": 2.921412300683371, "grad_norm": 0.030679482966661453, "learning_rate": 1.3097949886104784e-06, "loss": 0.0066, "step": 5130 }, { "epoch": 2.927107061503417, "grad_norm": 2.1198318004608154, "learning_rate": 1.2148823082763858e-06, "loss": 0.0182, "step": 5140 }, { "epoch": 2.932801822323462, "grad_norm": 0.2060464471578598, "learning_rate": 1.1199696279422931e-06, "loss": 0.0084, "step": 5150 }, { "epoch": 2.938496583143508, "grad_norm": 0.4984743893146515, "learning_rate": 1.0250569476082005e-06, "loss": 0.0106, "step": 5160 }, { "epoch": 2.9441913439635536, "grad_norm": 2.510512351989746, "learning_rate": 9.301442672741079e-07, "loss": 0.0032, "step": 5170 }, { "epoch": 2.949886104783599, "grad_norm": 1.7853169441223145, "learning_rate": 8.352315869400152e-07, "loss": 0.0278, "step": 5180 }, { "epoch": 2.9555808656036446, "grad_norm": 0.020138094201683998, "learning_rate": 7.403189066059226e-07, "loss": 0.0105, "step": 5190 }, { "epoch": 2.9612756264236904, "grad_norm": 0.009497715160250664, "learning_rate": 6.4540622627183e-07, "loss": 0.0011, "step": 5200 }, { "epoch": 2.9669703872437356, "grad_norm": 0.5586445331573486, "learning_rate": 5.504935459377373e-07, "loss": 0.0179, "step": 5210 }, { "epoch": 2.9726651480637813, "grad_norm": 0.4381959140300751, "learning_rate": 4.5558086560364467e-07, "loss": 0.0251, "step": 5220 }, { "epoch": 2.978359908883827, "grad_norm": 0.004315070807933807, "learning_rate": 3.6066818526955203e-07, "loss": 0.0243, "step": 5230 }, { "epoch": 2.9840546697038723, "grad_norm": 0.5453450083732605, "learning_rate": 2.657555049354594e-07, "loss": 0.0085, "step": 5240 }, { "epoch": 2.989749430523918, "grad_norm": 0.2711639404296875, "learning_rate": 1.7084282460136675e-07, "loss": 0.0167, "step": 5250 }, { "epoch": 2.995444191343964, "grad_norm": 2.088967800140381, "learning_rate": 7.593014426727411e-08, "loss": 0.0021, "step": 5260 } ], "logging_steps": 10, "max_steps": 5268, "num_input_tokens_seen": 0, "num_train_epochs": 3, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 1376049275709696.0, "train_batch_size": 8, "trial_name": null, "trial_params": null }