| { | |
| "best_metric": 0.11092506349086761, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-700", | |
| "epoch": 0.7632282202031, | |
| "eval_steps": 100, | |
| "global_step": 714, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0010689470871191875, | |
| "grad_norm": 1.8254921436309814, | |
| "learning_rate": 2e-05, | |
| "loss": 3.9246, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0010689470871191875, | |
| "eval_loss": 3.9061317443847656, | |
| "eval_runtime": 286.2861, | |
| "eval_samples_per_second": 4.359, | |
| "eval_steps_per_second": 1.09, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.002137894174238375, | |
| "grad_norm": 1.6388949155807495, | |
| "learning_rate": 4e-05, | |
| "loss": 3.7244, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.003206841261357563, | |
| "grad_norm": 1.8553436994552612, | |
| "learning_rate": 6e-05, | |
| "loss": 3.8132, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.00427578834847675, | |
| "grad_norm": 1.73699152469635, | |
| "learning_rate": 8e-05, | |
| "loss": 3.5206, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.005344735435595938, | |
| "grad_norm": 2.0092246532440186, | |
| "learning_rate": 0.0001, | |
| "loss": 3.2891, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.006413682522715126, | |
| "grad_norm": 2.0765933990478516, | |
| "learning_rate": 0.00012, | |
| "loss": 2.8629, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0074826296098343134, | |
| "grad_norm": 1.9038790464401245, | |
| "learning_rate": 0.00014, | |
| "loss": 2.1354, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0085515766969535, | |
| "grad_norm": 1.854448914527893, | |
| "learning_rate": 0.00016, | |
| "loss": 1.6619, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.009620523784072688, | |
| "grad_norm": 2.755218267440796, | |
| "learning_rate": 0.00018, | |
| "loss": 1.3892, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.010689470871191877, | |
| "grad_norm": 2.0359535217285156, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9684, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.011758417958311064, | |
| "grad_norm": 1.6178686618804932, | |
| "learning_rate": 0.00019999900431099657, | |
| "loss": 0.7566, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.012827365045430252, | |
| "grad_norm": 1.7248117923736572, | |
| "learning_rate": 0.00019999601726381413, | |
| "loss": 0.3901, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.01389631213254944, | |
| "grad_norm": 1.2264153957366943, | |
| "learning_rate": 0.00019999103891793618, | |
| "loss": 0.4864, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.014965259219668627, | |
| "grad_norm": 1.0095995664596558, | |
| "learning_rate": 0.00019998406937250034, | |
| "loss": 0.2396, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.016034206306787813, | |
| "grad_norm": 0.9053572416305542, | |
| "learning_rate": 0.0001999751087662966, | |
| "loss": 0.3481, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.017103153393907, | |
| "grad_norm": 0.7639911770820618, | |
| "learning_rate": 0.00019996415727776455, | |
| "loss": 0.2529, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.018172100481026188, | |
| "grad_norm": 0.9801614880561829, | |
| "learning_rate": 0.00019995121512498968, | |
| "loss": 0.3283, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.019241047568145375, | |
| "grad_norm": 3.5335612297058105, | |
| "learning_rate": 0.0001999362825656992, | |
| "loss": 0.2665, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.020309994655264563, | |
| "grad_norm": 0.8772406578063965, | |
| "learning_rate": 0.00019991935989725675, | |
| "loss": 0.3129, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.021378941742383754, | |
| "grad_norm": 0.4901948571205139, | |
| "learning_rate": 0.00019990044745665672, | |
| "loss": 0.2087, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.02244788882950294, | |
| "grad_norm": 0.5716347694396973, | |
| "learning_rate": 0.00019987954562051725, | |
| "loss": 0.2954, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.02351683591662213, | |
| "grad_norm": 0.5213301777839661, | |
| "learning_rate": 0.0001998566548050729, | |
| "loss": 0.309, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.024585783003741316, | |
| "grad_norm": 0.7063998579978943, | |
| "learning_rate": 0.00019983177546616633, | |
| "loss": 0.2885, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.025654730090860504, | |
| "grad_norm": 0.6609776020050049, | |
| "learning_rate": 0.00019980490809923926, | |
| "loss": 0.3596, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.02672367717797969, | |
| "grad_norm": 0.6100655794143677, | |
| "learning_rate": 0.0001997760532393225, | |
| "loss": 0.2666, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.02779262426509888, | |
| "grad_norm": 0.4118332266807556, | |
| "learning_rate": 0.00019974521146102537, | |
| "loss": 0.2019, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.028861571352218066, | |
| "grad_norm": 0.431892067193985, | |
| "learning_rate": 0.00019971238337852427, | |
| "loss": 0.2165, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.029930518439337254, | |
| "grad_norm": 0.4727337956428528, | |
| "learning_rate": 0.00019967756964555045, | |
| "loss": 0.2206, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.03099946552645644, | |
| "grad_norm": 0.40187329053878784, | |
| "learning_rate": 0.00019964077095537688, | |
| "loss": 0.2338, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.032068412613575625, | |
| "grad_norm": 0.3421157896518707, | |
| "learning_rate": 0.0001996019880408046, | |
| "loss": 0.1973, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.03313735970069481, | |
| "grad_norm": 0.451432466506958, | |
| "learning_rate": 0.00019956122167414807, | |
| "loss": 0.225, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.034206306787814, | |
| "grad_norm": 0.3993541896343231, | |
| "learning_rate": 0.0001995184726672197, | |
| "loss": 0.2093, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.03527525387493319, | |
| "grad_norm": 0.40808215737342834, | |
| "learning_rate": 0.00019947374187131384, | |
| "loss": 0.2254, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.036344200962052375, | |
| "grad_norm": 0.4794451892375946, | |
| "learning_rate": 0.00019942703017718975, | |
| "loss": 0.2226, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.03741314804917156, | |
| "grad_norm": 0.4402725398540497, | |
| "learning_rate": 0.0001993783385150538, | |
| "loss": 0.2356, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.03848209513629075, | |
| "grad_norm": 0.3642333447933197, | |
| "learning_rate": 0.000199327667854541, | |
| "loss": 0.1812, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.03955104222340994, | |
| "grad_norm": 0.3686297833919525, | |
| "learning_rate": 0.00019927501920469584, | |
| "loss": 0.2304, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.040619989310529125, | |
| "grad_norm": 0.31539255380630493, | |
| "learning_rate": 0.00019922039361395185, | |
| "loss": 0.1428, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.04168893639764832, | |
| "grad_norm": 0.36080244183540344, | |
| "learning_rate": 0.0001991637921701111, | |
| "loss": 0.1526, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.04275788348476751, | |
| "grad_norm": 0.44204625487327576, | |
| "learning_rate": 0.00019910521600032227, | |
| "loss": 0.2179, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.043826830571886695, | |
| "grad_norm": 0.323862761259079, | |
| "learning_rate": 0.00019904466627105834, | |
| "loss": 0.1372, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.04489577765900588, | |
| "grad_norm": 0.3098718225955963, | |
| "learning_rate": 0.0001989821441880933, | |
| "loss": 0.1665, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.04596472474612507, | |
| "grad_norm": 0.3269341289997101, | |
| "learning_rate": 0.0001989176509964781, | |
| "loss": 0.1824, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.04703367183324426, | |
| "grad_norm": 0.2592598795890808, | |
| "learning_rate": 0.00019885118798051605, | |
| "loss": 0.1704, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.048102618920363445, | |
| "grad_norm": 0.4089893698692322, | |
| "learning_rate": 0.000198782756463737, | |
| "loss": 0.1892, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.04917156600748263, | |
| "grad_norm": 0.2590075731277466, | |
| "learning_rate": 0.00019871235780887113, | |
| "loss": 0.1545, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.05024051309460182, | |
| "grad_norm": 0.28687378764152527, | |
| "learning_rate": 0.00019863999341782177, | |
| "loss": 0.1854, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.05130946018172101, | |
| "grad_norm": 0.24908822774887085, | |
| "learning_rate": 0.00019856566473163746, | |
| "loss": 0.16, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.052378407268840195, | |
| "grad_norm": 0.28324803709983826, | |
| "learning_rate": 0.00019848937323048337, | |
| "loss": 0.158, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.05344735435595938, | |
| "grad_norm": 0.33342692255973816, | |
| "learning_rate": 0.0001984111204336116, | |
| "loss": 0.2, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.05451630144307857, | |
| "grad_norm": 0.3774797320365906, | |
| "learning_rate": 0.00019833090789933122, | |
| "loss": 0.2204, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.05558524853019776, | |
| "grad_norm": 0.33370718359947205, | |
| "learning_rate": 0.00019824873722497694, | |
| "loss": 0.1605, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.056654195617316945, | |
| "grad_norm": 0.2836022973060608, | |
| "learning_rate": 0.0001981646100468775, | |
| "loss": 0.1768, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.05772314270443613, | |
| "grad_norm": 0.3592361807823181, | |
| "learning_rate": 0.00019807852804032305, | |
| "loss": 0.1903, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.05879208979155532, | |
| "grad_norm": 0.19539162516593933, | |
| "learning_rate": 0.0001979904929195317, | |
| "loss": 0.1543, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.05986103687867451, | |
| "grad_norm": 0.29716816544532776, | |
| "learning_rate": 0.00019790050643761552, | |
| "loss": 0.2106, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.060929983965793695, | |
| "grad_norm": 0.3658367097377777, | |
| "learning_rate": 0.0001978085703865455, | |
| "loss": 0.2197, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.06199893105291288, | |
| "grad_norm": 0.30680355429649353, | |
| "learning_rate": 0.00019771468659711595, | |
| "loss": 0.1856, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.06306787814003206, | |
| "grad_norm": 0.22984617948532104, | |
| "learning_rate": 0.000197618856938908, | |
| "loss": 0.1519, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.06413682522715125, | |
| "grad_norm": 0.23498129844665527, | |
| "learning_rate": 0.0001975210833202524, | |
| "loss": 0.1692, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.06520577231427044, | |
| "grad_norm": 0.3555554449558258, | |
| "learning_rate": 0.00019742136768819145, | |
| "loss": 0.179, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.06627471940138963, | |
| "grad_norm": 0.3502557575702667, | |
| "learning_rate": 0.00019731971202844036, | |
| "loss": 0.2044, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.06734366648850881, | |
| "grad_norm": 0.2852421998977661, | |
| "learning_rate": 0.00019721611836534756, | |
| "loss": 0.1737, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.068412613575628, | |
| "grad_norm": 0.27989548444747925, | |
| "learning_rate": 0.00019711058876185447, | |
| "loss": 0.184, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.06948156066274719, | |
| "grad_norm": 0.26199406385421753, | |
| "learning_rate": 0.00019700312531945442, | |
| "loss": 0.1867, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.07055050774986638, | |
| "grad_norm": 0.21757884323596954, | |
| "learning_rate": 0.00019689373017815073, | |
| "loss": 0.1509, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.07161945483698556, | |
| "grad_norm": 0.23988686501979828, | |
| "learning_rate": 0.00019678240551641426, | |
| "loss": 0.1667, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.07268840192410475, | |
| "grad_norm": 0.25003519654273987, | |
| "learning_rate": 0.00019666915355113975, | |
| "loss": 0.1412, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.07375734901122394, | |
| "grad_norm": 0.19979026913642883, | |
| "learning_rate": 0.000196553976537602, | |
| "loss": 0.1422, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.07482629609834313, | |
| "grad_norm": 0.2231016457080841, | |
| "learning_rate": 0.00019643687676941068, | |
| "loss": 0.1447, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.07589524318546231, | |
| "grad_norm": 0.22386683523654938, | |
| "learning_rate": 0.00019631785657846487, | |
| "loss": 0.1405, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.0769641902725815, | |
| "grad_norm": 0.24956126511096954, | |
| "learning_rate": 0.00019619691833490643, | |
| "loss": 0.1741, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.07803313735970069, | |
| "grad_norm": 0.5022568702697754, | |
| "learning_rate": 0.00019607406444707294, | |
| "loss": 0.3543, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.07910208444681988, | |
| "grad_norm": 0.3087567985057831, | |
| "learning_rate": 0.00019594929736144976, | |
| "loss": 0.1787, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.08017103153393906, | |
| "grad_norm": 0.27388593554496765, | |
| "learning_rate": 0.0001958226195626211, | |
| "loss": 0.1552, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.08123997862105825, | |
| "grad_norm": 0.3442430794239044, | |
| "learning_rate": 0.0001956940335732209, | |
| "loss": 0.1789, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.08230892570817744, | |
| "grad_norm": 0.2176661193370819, | |
| "learning_rate": 0.00019556354195388221, | |
| "loss": 0.1834, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.08337787279529664, | |
| "grad_norm": 0.3691951632499695, | |
| "learning_rate": 0.0001954311473031864, | |
| "loss": 0.1758, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.08444681988241583, | |
| "grad_norm": 0.2579557001590729, | |
| "learning_rate": 0.00019529685225761148, | |
| "loss": 0.1875, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.08551576696953501, | |
| "grad_norm": 0.2234066128730774, | |
| "learning_rate": 0.00019516065949147943, | |
| "loss": 0.144, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.0865847140566542, | |
| "grad_norm": 0.2428237646818161, | |
| "learning_rate": 0.0001950225717169031, | |
| "loss": 0.1616, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.08765366114377339, | |
| "grad_norm": 0.2643418610095978, | |
| "learning_rate": 0.00019488259168373197, | |
| "loss": 0.186, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.08872260823089258, | |
| "grad_norm": 0.27387118339538574, | |
| "learning_rate": 0.00019474072217949768, | |
| "loss": 0.1985, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.08979155531801176, | |
| "grad_norm": 0.16810058057308197, | |
| "learning_rate": 0.00019459696602935837, | |
| "loss": 0.1341, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.09086050240513095, | |
| "grad_norm": 0.2106161266565323, | |
| "learning_rate": 0.00019445132609604234, | |
| "loss": 0.146, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.09192944949225014, | |
| "grad_norm": 0.21163029968738556, | |
| "learning_rate": 0.00019430380527979123, | |
| "loss": 0.1906, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.09299839657936933, | |
| "grad_norm": 0.22357632219791412, | |
| "learning_rate": 0.00019415440651830208, | |
| "loss": 0.1809, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.09406734366648851, | |
| "grad_norm": 0.24073223769664764, | |
| "learning_rate": 0.00019400313278666902, | |
| "loss": 0.1523, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.0951362907536077, | |
| "grad_norm": 0.2266823649406433, | |
| "learning_rate": 0.00019384998709732386, | |
| "loss": 0.1661, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.09620523784072689, | |
| "grad_norm": 0.1823054999113083, | |
| "learning_rate": 0.0001936949724999762, | |
| "loss": 0.141, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.09727418492784608, | |
| "grad_norm": 0.19450423121452332, | |
| "learning_rate": 0.00019353809208155255, | |
| "loss": 0.1409, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.09834313201496526, | |
| "grad_norm": 0.36040905117988586, | |
| "learning_rate": 0.00019337934896613516, | |
| "loss": 0.2028, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.09941207910208445, | |
| "grad_norm": 0.21639132499694824, | |
| "learning_rate": 0.00019321874631489947, | |
| "loss": 0.1518, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.10048102618920364, | |
| "grad_norm": 0.20086891949176788, | |
| "learning_rate": 0.00019305628732605137, | |
| "loss": 0.1514, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.10154997327632283, | |
| "grad_norm": 0.26318323612213135, | |
| "learning_rate": 0.0001928919752347634, | |
| "loss": 0.1552, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.10261892036344201, | |
| "grad_norm": 0.4378178119659424, | |
| "learning_rate": 0.00019272581331311047, | |
| "loss": 0.2046, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.1036878674505612, | |
| "grad_norm": 0.2704489529132843, | |
| "learning_rate": 0.0001925578048700045, | |
| "loss": 0.1898, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.10475681453768039, | |
| "grad_norm": 0.22563497722148895, | |
| "learning_rate": 0.0001923879532511287, | |
| "loss": 0.1709, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.10582576162479958, | |
| "grad_norm": 0.23802538216114044, | |
| "learning_rate": 0.0001922162618388708, | |
| "loss": 0.1713, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.10689470871191876, | |
| "grad_norm": 0.36641374230384827, | |
| "learning_rate": 0.00019204273405225587, | |
| "loss": 0.1519, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.10689470871191876, | |
| "eval_loss": 0.15769338607788086, | |
| "eval_runtime": 288.7962, | |
| "eval_samples_per_second": 4.321, | |
| "eval_steps_per_second": 1.08, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.10796365579903795, | |
| "grad_norm": 0.2594605088233948, | |
| "learning_rate": 0.0001918673733468781, | |
| "loss": 0.1791, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.10903260288615714, | |
| "grad_norm": 0.23555320501327515, | |
| "learning_rate": 0.00019169018321483198, | |
| "loss": 0.1331, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.11010154997327633, | |
| "grad_norm": 0.26566898822784424, | |
| "learning_rate": 0.00019151116718464284, | |
| "loss": 0.1662, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.11117049706039552, | |
| "grad_norm": 0.1984795778989792, | |
| "learning_rate": 0.00019133032882119653, | |
| "loss": 0.1417, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.1122394441475147, | |
| "grad_norm": 0.2048155814409256, | |
| "learning_rate": 0.0001911476717256685, | |
| "loss": 0.155, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.11330839123463389, | |
| "grad_norm": 0.2680719792842865, | |
| "learning_rate": 0.00019096319953545185, | |
| "loss": 0.1726, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.11437733832175308, | |
| "grad_norm": 0.25940361618995667, | |
| "learning_rate": 0.0001907769159240853, | |
| "loss": 0.189, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.11544628540887227, | |
| "grad_norm": 0.23745258152484894, | |
| "learning_rate": 0.00019058882460117974, | |
| "loss": 0.1602, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.11651523249599145, | |
| "grad_norm": 0.1798817664384842, | |
| "learning_rate": 0.00019039892931234435, | |
| "loss": 0.1397, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.11758417958311064, | |
| "grad_norm": 0.23396556079387665, | |
| "learning_rate": 0.00019020723383911215, | |
| "loss": 0.1431, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.11865312667022983, | |
| "grad_norm": 0.2253292351961136, | |
| "learning_rate": 0.00019001374199886467, | |
| "loss": 0.1432, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.11972207375734902, | |
| "grad_norm": 0.18429440259933472, | |
| "learning_rate": 0.00018981845764475582, | |
| "loss": 0.125, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.1207910208444682, | |
| "grad_norm": 0.20196135342121124, | |
| "learning_rate": 0.00018962138466563534, | |
| "loss": 0.1381, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.12185996793158739, | |
| "grad_norm": 0.2396540343761444, | |
| "learning_rate": 0.00018942252698597113, | |
| "loss": 0.1618, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.12292891501870658, | |
| "grad_norm": 0.2343413084745407, | |
| "learning_rate": 0.00018922188856577133, | |
| "loss": 0.1874, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.12399786210582577, | |
| "grad_norm": 0.23756933212280273, | |
| "learning_rate": 0.00018901947340050528, | |
| "loss": 0.1758, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.12506680919294494, | |
| "grad_norm": 0.22977083921432495, | |
| "learning_rate": 0.0001888152855210241, | |
| "loss": 0.1464, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.12613575628006413, | |
| "grad_norm": 0.28528642654418945, | |
| "learning_rate": 0.00018860932899348028, | |
| "loss": 0.235, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.1272047033671833, | |
| "grad_norm": 0.2515600025653839, | |
| "learning_rate": 0.0001884016079192468, | |
| "loss": 0.1562, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.1282736504543025, | |
| "grad_norm": 0.2021314799785614, | |
| "learning_rate": 0.0001881921264348355, | |
| "loss": 0.1463, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.1293425975414217, | |
| "grad_norm": 0.21556632220745087, | |
| "learning_rate": 0.0001879808887118146, | |
| "loss": 0.1441, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.13041154462854088, | |
| "grad_norm": 0.17659486830234528, | |
| "learning_rate": 0.00018776789895672558, | |
| "loss": 0.1228, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.13148049171566006, | |
| "grad_norm": 0.21652792394161224, | |
| "learning_rate": 0.00018755316141099964, | |
| "loss": 0.1225, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.13254943880277925, | |
| "grad_norm": 0.23109202086925507, | |
| "learning_rate": 0.00018733668035087302, | |
| "loss": 0.1547, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.13361838588989844, | |
| "grad_norm": 0.20818117260932922, | |
| "learning_rate": 0.00018711846008730192, | |
| "loss": 0.1387, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.13468733297701763, | |
| "grad_norm": 0.25625041127204895, | |
| "learning_rate": 0.00018689850496587674, | |
| "loss": 0.1411, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.13575628006413681, | |
| "grad_norm": 0.1805257648229599, | |
| "learning_rate": 0.00018667681936673528, | |
| "loss": 0.1198, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.136825227151256, | |
| "grad_norm": 0.23043088614940643, | |
| "learning_rate": 0.00018645340770447595, | |
| "loss": 0.1705, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.1378941742383752, | |
| "grad_norm": 0.28206974267959595, | |
| "learning_rate": 0.00018622827442806933, | |
| "loss": 0.2061, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.13896312132549438, | |
| "grad_norm": 0.43259602785110474, | |
| "learning_rate": 0.00018600142402077006, | |
| "loss": 0.1901, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.14003206841261356, | |
| "grad_norm": 0.2347777932882309, | |
| "learning_rate": 0.00018577286100002723, | |
| "loss": 0.1512, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.14110101549973275, | |
| "grad_norm": 0.22234401106834412, | |
| "learning_rate": 0.00018554258991739452, | |
| "loss": 0.1599, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.14216996258685194, | |
| "grad_norm": 0.1948978453874588, | |
| "learning_rate": 0.0001853106153584397, | |
| "loss": 0.1369, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.14323890967397113, | |
| "grad_norm": 0.286856472492218, | |
| "learning_rate": 0.0001850769419426531, | |
| "loss": 0.1469, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.14430785676109031, | |
| "grad_norm": 0.23832087218761444, | |
| "learning_rate": 0.0001848415743233557, | |
| "loss": 0.1647, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.1453768038482095, | |
| "grad_norm": 0.20166629552841187, | |
| "learning_rate": 0.0001846045171876065, | |
| "loss": 0.1376, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.1464457509353287, | |
| "grad_norm": 0.21285155415534973, | |
| "learning_rate": 0.0001843657752561092, | |
| "loss": 0.17, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.14751469802244788, | |
| "grad_norm": 0.2186753898859024, | |
| "learning_rate": 0.00018412535328311814, | |
| "loss": 0.1369, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.14858364510956706, | |
| "grad_norm": 0.23166581988334656, | |
| "learning_rate": 0.00018388325605634353, | |
| "loss": 0.1986, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.14965259219668625, | |
| "grad_norm": 0.17363017797470093, | |
| "learning_rate": 0.00018363948839685636, | |
| "loss": 0.1275, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.15072153928380544, | |
| "grad_norm": 0.30469027161598206, | |
| "learning_rate": 0.0001833940551589922, | |
| "loss": 0.1953, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.15179048637092463, | |
| "grad_norm": 0.17805826663970947, | |
| "learning_rate": 0.00018314696123025454, | |
| "loss": 0.1276, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.15285943345804381, | |
| "grad_norm": 0.20481310784816742, | |
| "learning_rate": 0.00018289821153121753, | |
| "loss": 0.1551, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.153928380545163, | |
| "grad_norm": 0.22326795756816864, | |
| "learning_rate": 0.000182647811015428, | |
| "loss": 0.1669, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.1549973276322822, | |
| "grad_norm": 0.21309794485569, | |
| "learning_rate": 0.00018239576466930668, | |
| "loss": 0.1652, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.15606627471940138, | |
| "grad_norm": 0.2599788010120392, | |
| "learning_rate": 0.00018214207751204918, | |
| "loss": 0.181, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.15713522180652056, | |
| "grad_norm": 0.20984572172164917, | |
| "learning_rate": 0.00018188675459552566, | |
| "loss": 0.1296, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.15820416889363975, | |
| "grad_norm": 0.22253574430942535, | |
| "learning_rate": 0.0001816298010041806, | |
| "loss": 0.159, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.15927311598075894, | |
| "grad_norm": 0.1939833164215088, | |
| "learning_rate": 0.00018137122185493123, | |
| "loss": 0.1497, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.16034206306787813, | |
| "grad_norm": 0.18843698501586914, | |
| "learning_rate": 0.0001811110222970659, | |
| "loss": 0.1286, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.16141101015499731, | |
| "grad_norm": 0.2895407974720001, | |
| "learning_rate": 0.00018084920751214144, | |
| "loss": 0.1761, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.1624799572421165, | |
| "grad_norm": 0.23353318870067596, | |
| "learning_rate": 0.0001805857827138798, | |
| "loss": 0.1605, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.1635489043292357, | |
| "grad_norm": 0.2674696445465088, | |
| "learning_rate": 0.00018032075314806448, | |
| "loss": 0.1676, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.16461785141635488, | |
| "grad_norm": 0.2173846811056137, | |
| "learning_rate": 0.00018005412409243606, | |
| "loss": 0.1391, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.16568679850347406, | |
| "grad_norm": 0.20100249350070953, | |
| "learning_rate": 0.0001797859008565868, | |
| "loss": 0.1254, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.16675574559059328, | |
| "grad_norm": 0.22731080651283264, | |
| "learning_rate": 0.0001795160887818553, | |
| "loss": 0.1659, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.16782469267771247, | |
| "grad_norm": 0.18832936882972717, | |
| "learning_rate": 0.00017924469324121986, | |
| "loss": 0.1329, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.16889363976483165, | |
| "grad_norm": 0.23300229012966156, | |
| "learning_rate": 0.0001789717196391916, | |
| "loss": 0.123, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.16996258685195084, | |
| "grad_norm": 0.21164602041244507, | |
| "learning_rate": 0.00017869717341170675, | |
| "loss": 0.1337, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.17103153393907003, | |
| "grad_norm": 0.1995665431022644, | |
| "learning_rate": 0.00017842106002601856, | |
| "loss": 0.1265, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.17210048102618922, | |
| "grad_norm": 0.20902606844902039, | |
| "learning_rate": 0.0001781433849805882, | |
| "loss": 0.1392, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.1731694281133084, | |
| "grad_norm": 0.18716943264007568, | |
| "learning_rate": 0.00017786415380497553, | |
| "loss": 0.1187, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.1742383752004276, | |
| "grad_norm": 0.22572395205497742, | |
| "learning_rate": 0.0001775833720597287, | |
| "loss": 0.1428, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.17530732228754678, | |
| "grad_norm": 0.17674115300178528, | |
| "learning_rate": 0.0001773010453362737, | |
| "loss": 0.1292, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.17637626937466597, | |
| "grad_norm": 0.2090616673231125, | |
| "learning_rate": 0.00017701717925680278, | |
| "loss": 0.1726, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.17744521646178515, | |
| "grad_norm": 0.16593225300312042, | |
| "learning_rate": 0.00017673177947416258, | |
| "loss": 0.1201, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.17851416354890434, | |
| "grad_norm": 0.19040323793888092, | |
| "learning_rate": 0.00017644485167174164, | |
| "loss": 0.1255, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.17958311063602353, | |
| "grad_norm": 0.20446278154850006, | |
| "learning_rate": 0.00017615640156335712, | |
| "loss": 0.1433, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.18065205772314272, | |
| "grad_norm": 0.154877707362175, | |
| "learning_rate": 0.000175866434893141, | |
| "loss": 0.1032, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.1817210048102619, | |
| "grad_norm": 0.39736929535865784, | |
| "learning_rate": 0.00017557495743542585, | |
| "loss": 0.1302, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.1827899518973811, | |
| "grad_norm": 0.2045171856880188, | |
| "learning_rate": 0.00017528197499462958, | |
| "loss": 0.1431, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.18385889898450028, | |
| "grad_norm": 0.19055134057998657, | |
| "learning_rate": 0.0001749874934051401, | |
| "loss": 0.1465, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.18492784607161947, | |
| "grad_norm": 0.2106601893901825, | |
| "learning_rate": 0.00017469151853119906, | |
| "loss": 0.1355, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.18599679315873865, | |
| "grad_norm": 0.27130061388015747, | |
| "learning_rate": 0.00017439405626678496, | |
| "loss": 0.1614, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.18706574024585784, | |
| "grad_norm": 0.2109244167804718, | |
| "learning_rate": 0.00017409511253549593, | |
| "loss": 0.1509, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.18813468733297703, | |
| "grad_norm": 0.17472617328166962, | |
| "learning_rate": 0.00017379469329043167, | |
| "loss": 0.1117, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.18920363442009622, | |
| "grad_norm": 0.2262858897447586, | |
| "learning_rate": 0.00017349280451407499, | |
| "loss": 0.1509, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.1902725815072154, | |
| "grad_norm": 0.18161039054393768, | |
| "learning_rate": 0.00017318945221817255, | |
| "loss": 0.1182, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.1913415285943346, | |
| "grad_norm": 0.17918163537979126, | |
| "learning_rate": 0.00017288464244361528, | |
| "loss": 0.1477, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.19241047568145378, | |
| "grad_norm": 0.22044941782951355, | |
| "learning_rate": 0.00017257838126031797, | |
| "loss": 0.164, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.19347942276857297, | |
| "grad_norm": 0.2787477970123291, | |
| "learning_rate": 0.00017227067476709854, | |
| "loss": 0.1934, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.19454836985569215, | |
| "grad_norm": 0.2308286875486374, | |
| "learning_rate": 0.00017196152909155628, | |
| "loss": 0.1649, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.19561731694281134, | |
| "grad_norm": 0.20782075822353363, | |
| "learning_rate": 0.00017165095038995033, | |
| "loss": 0.1595, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.19668626402993053, | |
| "grad_norm": 0.17654070258140564, | |
| "learning_rate": 0.00017133894484707655, | |
| "loss": 0.1111, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.19775521111704972, | |
| "grad_norm": 0.1830965131521225, | |
| "learning_rate": 0.00017102551867614478, | |
| "loss": 0.1184, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.1988241582041689, | |
| "grad_norm": 0.151173397898674, | |
| "learning_rate": 0.00017071067811865476, | |
| "loss": 0.1262, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.1998931052912881, | |
| "grad_norm": 0.26543188095092773, | |
| "learning_rate": 0.00017039442944427217, | |
| "loss": 0.1657, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.20096205237840728, | |
| "grad_norm": 0.190846785902977, | |
| "learning_rate": 0.00017007677895070357, | |
| "loss": 0.1426, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.20203099946552647, | |
| "grad_norm": 0.21965020895004272, | |
| "learning_rate": 0.00016975773296357098, | |
| "loss": 0.1327, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.20309994655264565, | |
| "grad_norm": 0.17773884534835815, | |
| "learning_rate": 0.00016943729783628608, | |
| "loss": 0.1629, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.20416889363976484, | |
| "grad_norm": 2.5340681076049805, | |
| "learning_rate": 0.00016911547994992348, | |
| "loss": 0.1637, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.20523784072688403, | |
| "grad_norm": 0.21551626920700073, | |
| "learning_rate": 0.00016879228571309377, | |
| "loss": 0.1599, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.20630678781400322, | |
| "grad_norm": 0.15846356749534607, | |
| "learning_rate": 0.00016846772156181594, | |
| "loss": 0.1375, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.2073757349011224, | |
| "grad_norm": 0.14692065119743347, | |
| "learning_rate": 0.00016814179395938913, | |
| "loss": 0.1117, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.2084446819882416, | |
| "grad_norm": 0.3390064537525177, | |
| "learning_rate": 0.00016781450939626387, | |
| "loss": 0.1271, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.20951362907536078, | |
| "grad_norm": 0.17453472316265106, | |
| "learning_rate": 0.00016748587438991303, | |
| "loss": 0.1396, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.21058257616247997, | |
| "grad_norm": 0.19709929823875427, | |
| "learning_rate": 0.00016715589548470185, | |
| "loss": 0.13, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.21165152324959915, | |
| "grad_norm": 0.24145397543907166, | |
| "learning_rate": 0.00016682457925175763, | |
| "loss": 0.1634, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.21272047033671834, | |
| "grad_norm": 0.17630192637443542, | |
| "learning_rate": 0.000166491932288839, | |
| "loss": 0.1299, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.21378941742383753, | |
| "grad_norm": 0.2846260368824005, | |
| "learning_rate": 0.00016615796122020443, | |
| "loss": 0.164, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.21378941742383753, | |
| "eval_loss": 0.1369618922472, | |
| "eval_runtime": 288.8956, | |
| "eval_samples_per_second": 4.32, | |
| "eval_steps_per_second": 1.08, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.21485836451095672, | |
| "grad_norm": 0.18763378262519836, | |
| "learning_rate": 0.0001658226726964803, | |
| "loss": 0.1481, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.2159273115980759, | |
| "grad_norm": 0.18099422752857208, | |
| "learning_rate": 0.00016548607339452853, | |
| "loss": 0.1275, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.2169962586851951, | |
| "grad_norm": 0.22163395583629608, | |
| "learning_rate": 0.0001651481700173136, | |
| "loss": 0.1657, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.21806520577231428, | |
| "grad_norm": 0.1837584376335144, | |
| "learning_rate": 0.00016480896929376907, | |
| "loss": 0.133, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.21913415285943347, | |
| "grad_norm": 0.23913772404193878, | |
| "learning_rate": 0.0001644684779786635, | |
| "loss": 0.1439, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.22020309994655266, | |
| "grad_norm": 0.24784860014915466, | |
| "learning_rate": 0.0001641267028524661, | |
| "loss": 0.1588, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.22127204703367184, | |
| "grad_norm": 0.189157634973526, | |
| "learning_rate": 0.00016378365072121156, | |
| "loss": 0.154, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.22234099412079103, | |
| "grad_norm": 0.18977932631969452, | |
| "learning_rate": 0.00016343932841636456, | |
| "loss": 0.1338, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.22340994120791022, | |
| "grad_norm": 0.1805882751941681, | |
| "learning_rate": 0.00016309374279468376, | |
| "loss": 0.1333, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.2244788882950294, | |
| "grad_norm": 0.17292280495166779, | |
| "learning_rate": 0.0001627469007380852, | |
| "loss": 0.1269, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.2255478353821486, | |
| "grad_norm": 0.20466209948062897, | |
| "learning_rate": 0.00016239880915350536, | |
| "loss": 0.1548, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.22661678246926778, | |
| "grad_norm": 0.18790069222450256, | |
| "learning_rate": 0.00016204947497276345, | |
| "loss": 0.1228, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.22768572955638697, | |
| "grad_norm": 0.17624615132808685, | |
| "learning_rate": 0.00016169890515242355, | |
| "loss": 0.1152, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.22875467664350616, | |
| "grad_norm": 0.18744930624961853, | |
| "learning_rate": 0.00016134710667365596, | |
| "loss": 0.1446, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.22982362373062534, | |
| "grad_norm": 0.18070746958255768, | |
| "learning_rate": 0.0001609940865420982, | |
| "loss": 0.1391, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.23089257081774453, | |
| "grad_norm": 0.16776110231876373, | |
| "learning_rate": 0.00016063985178771555, | |
| "loss": 0.1237, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.23196151790486372, | |
| "grad_norm": 0.2151951640844345, | |
| "learning_rate": 0.00016028440946466095, | |
| "loss": 0.1541, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.2330304649919829, | |
| "grad_norm": 0.17011284828186035, | |
| "learning_rate": 0.0001599277666511347, | |
| "loss": 0.1127, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.2340994120791021, | |
| "grad_norm": 0.15342286229133606, | |
| "learning_rate": 0.00015956993044924334, | |
| "loss": 0.12, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.23516835916622128, | |
| "grad_norm": 0.15606354176998138, | |
| "learning_rate": 0.00015921090798485832, | |
| "loss": 0.1177, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.23623730625334047, | |
| "grad_norm": 0.18545080721378326, | |
| "learning_rate": 0.00015885070640747395, | |
| "loss": 0.1296, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.23730625334045966, | |
| "grad_norm": 0.21306568384170532, | |
| "learning_rate": 0.0001584893328900653, | |
| "loss": 0.1224, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.23837520042757884, | |
| "grad_norm": 0.19317717850208282, | |
| "learning_rate": 0.00015812679462894513, | |
| "loss": 0.144, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.23944414751469803, | |
| "grad_norm": 0.14700594544410706, | |
| "learning_rate": 0.0001577630988436206, | |
| "loss": 0.1212, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.24051309460181722, | |
| "grad_norm": 0.20297332108020782, | |
| "learning_rate": 0.00015739825277664955, | |
| "loss": 0.1477, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.2415820416889364, | |
| "grad_norm": 0.21095982193946838, | |
| "learning_rate": 0.0001570322636934964, | |
| "loss": 0.1258, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.2426509887760556, | |
| "grad_norm": 0.24124930799007416, | |
| "learning_rate": 0.00015666513888238726, | |
| "loss": 0.1861, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.24371993586317478, | |
| "grad_norm": 0.18464890122413635, | |
| "learning_rate": 0.00015629688565416478, | |
| "loss": 0.1393, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.24478888295029397, | |
| "grad_norm": 0.1888304352760315, | |
| "learning_rate": 0.00015592751134214282, | |
| "loss": 0.1408, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.24585783003741316, | |
| "grad_norm": 0.16777703166007996, | |
| "learning_rate": 0.00015555702330196023, | |
| "loss": 0.113, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.24692677712453234, | |
| "grad_norm": 0.21285921335220337, | |
| "learning_rate": 0.0001551854289114343, | |
| "loss": 0.1451, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.24799572421165153, | |
| "grad_norm": 0.2046755701303482, | |
| "learning_rate": 0.000154812735570414, | |
| "loss": 0.1059, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.24906467129877072, | |
| "grad_norm": 0.20832888782024384, | |
| "learning_rate": 0.0001544389507006326, | |
| "loss": 0.1333, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.2501336183858899, | |
| "grad_norm": 0.15345333516597748, | |
| "learning_rate": 0.00015406408174555976, | |
| "loss": 0.1083, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.25120256547300907, | |
| "grad_norm": 0.18688637018203735, | |
| "learning_rate": 0.00015368813617025343, | |
| "loss": 0.1225, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.25227151256012825, | |
| "grad_norm": 0.22301477193832397, | |
| "learning_rate": 0.00015331112146121104, | |
| "loss": 0.1515, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.25334045964724744, | |
| "grad_norm": 0.21012060344219208, | |
| "learning_rate": 0.00015293304512622072, | |
| "loss": 0.1566, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.2544094067343666, | |
| "grad_norm": 0.17001067101955414, | |
| "learning_rate": 0.00015255391469421128, | |
| "loss": 0.1249, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.2554783538214858, | |
| "grad_norm": 0.23301930725574493, | |
| "learning_rate": 0.00015217373771510292, | |
| "loss": 0.1624, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.256547300908605, | |
| "grad_norm": 0.18844428658485413, | |
| "learning_rate": 0.00015179252175965633, | |
| "loss": 0.1299, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.2576162479957242, | |
| "grad_norm": 0.20432229340076447, | |
| "learning_rate": 0.00015141027441932216, | |
| "loss": 0.1499, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.2586851950828434, | |
| "grad_norm": 0.21431198716163635, | |
| "learning_rate": 0.00015102700330609, | |
| "loss": 0.1414, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.25975414216996257, | |
| "grad_norm": 0.185688316822052, | |
| "learning_rate": 0.00015064271605233636, | |
| "loss": 0.1243, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.26082308925708175, | |
| "grad_norm": 0.1799941509962082, | |
| "learning_rate": 0.00015025742031067317, | |
| "loss": 0.1382, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.26189203634420094, | |
| "grad_norm": 0.1698986291885376, | |
| "learning_rate": 0.0001498711237537951, | |
| "loss": 0.132, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.26296098343132013, | |
| "grad_norm": 0.17504070699214935, | |
| "learning_rate": 0.00014948383407432678, | |
| "loss": 0.1386, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.2640299305184393, | |
| "grad_norm": 0.28007063269615173, | |
| "learning_rate": 0.00014909555898466974, | |
| "loss": 0.1891, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.2650988776055585, | |
| "grad_norm": 0.17352139949798584, | |
| "learning_rate": 0.00014870630621684872, | |
| "loss": 0.1375, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.2661678246926777, | |
| "grad_norm": 0.16270796954631805, | |
| "learning_rate": 0.00014831608352235774, | |
| "loss": 0.1329, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.2672367717797969, | |
| "grad_norm": 0.1661270558834076, | |
| "learning_rate": 0.0001479248986720057, | |
| "loss": 0.129, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.26830571886691607, | |
| "grad_norm": 0.15105728805065155, | |
| "learning_rate": 0.00014753275945576165, | |
| "loss": 0.1299, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.26937466595403525, | |
| "grad_norm": 0.21838223934173584, | |
| "learning_rate": 0.0001471396736825998, | |
| "loss": 0.1563, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.27044361304115444, | |
| "grad_norm": 0.16052576899528503, | |
| "learning_rate": 0.00014674564918034367, | |
| "loss": 0.1301, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.27151256012827363, | |
| "grad_norm": 0.1738349050283432, | |
| "learning_rate": 0.00014635069379551055, | |
| "loss": 0.1432, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.2725815072153928, | |
| "grad_norm": 0.1612621247768402, | |
| "learning_rate": 0.00014595481539315518, | |
| "loss": 0.1129, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.273650454302512, | |
| "grad_norm": 0.17328664660453796, | |
| "learning_rate": 0.00014555802185671298, | |
| "loss": 0.1414, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.2747194013896312, | |
| "grad_norm": 0.21222056448459625, | |
| "learning_rate": 0.0001451603210878431, | |
| "loss": 0.1222, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.2757883484767504, | |
| "grad_norm": 0.23580165207386017, | |
| "learning_rate": 0.00014476172100627127, | |
| "loss": 0.1539, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.27685729556386957, | |
| "grad_norm": 0.19674788415431976, | |
| "learning_rate": 0.00014436222954963178, | |
| "loss": 0.1318, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.27792624265098875, | |
| "grad_norm": 0.180185005068779, | |
| "learning_rate": 0.00014396185467330973, | |
| "loss": 0.1174, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.27899518973810794, | |
| "grad_norm": 0.17812158167362213, | |
| "learning_rate": 0.00014356060435028226, | |
| "loss": 0.1169, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.28006413682522713, | |
| "grad_norm": 0.17591345310211182, | |
| "learning_rate": 0.00014315848657096004, | |
| "loss": 0.1132, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.2811330839123463, | |
| "grad_norm": 0.20523187518119812, | |
| "learning_rate": 0.00014275550934302823, | |
| "loss": 0.1358, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.2822020309994655, | |
| "grad_norm": 0.18519556522369385, | |
| "learning_rate": 0.00014235168069128657, | |
| "loss": 0.1382, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.2832709780865847, | |
| "grad_norm": 0.2633715271949768, | |
| "learning_rate": 0.00014194700865749008, | |
| "loss": 0.143, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.2843399251737039, | |
| "grad_norm": 0.30586305260658264, | |
| "learning_rate": 0.00014154150130018866, | |
| "loss": 0.1788, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.28540887226082307, | |
| "grad_norm": 0.14305737614631653, | |
| "learning_rate": 0.0001411351666945666, | |
| "loss": 0.0905, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.28647781934794225, | |
| "grad_norm": 0.19511297345161438, | |
| "learning_rate": 0.00014072801293228188, | |
| "loss": 0.1424, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.28754676643506144, | |
| "grad_norm": 0.3783949911594391, | |
| "learning_rate": 0.00014032004812130497, | |
| "loss": 0.1346, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.28861571352218063, | |
| "grad_norm": 0.19596309959888458, | |
| "learning_rate": 0.00013991128038575741, | |
| "loss": 0.1384, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.2896846606092998, | |
| "grad_norm": 0.204111248254776, | |
| "learning_rate": 0.00013950171786575002, | |
| "loss": 0.1577, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.290753607696419, | |
| "grad_norm": 0.19096602499485016, | |
| "learning_rate": 0.00013909136871722067, | |
| "loss": 0.1333, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.2918225547835382, | |
| "grad_norm": 0.21785670518875122, | |
| "learning_rate": 0.00013868024111177206, | |
| "loss": 0.1247, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.2928915018706574, | |
| "grad_norm": 0.21137268841266632, | |
| "learning_rate": 0.000138268343236509, | |
| "loss": 0.1231, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.29396044895777657, | |
| "grad_norm": 0.21565985679626465, | |
| "learning_rate": 0.00013785568329387508, | |
| "loss": 0.1514, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.29502939604489575, | |
| "grad_norm": 0.1645229011774063, | |
| "learning_rate": 0.00013744226950148972, | |
| "loss": 0.1204, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.29609834313201494, | |
| "grad_norm": 0.18113082647323608, | |
| "learning_rate": 0.0001370281100919842, | |
| "loss": 0.1374, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.29716729021913413, | |
| "grad_norm": 0.15095637738704681, | |
| "learning_rate": 0.00013661321331283796, | |
| "loss": 0.0931, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.2982362373062533, | |
| "grad_norm": 0.17453165352344513, | |
| "learning_rate": 0.00013619758742621418, | |
| "loss": 0.1336, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.2993051843933725, | |
| "grad_norm": 0.1741105020046234, | |
| "learning_rate": 0.00013578124070879534, | |
| "loss": 0.1263, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.3003741314804917, | |
| "grad_norm": 0.17308761179447174, | |
| "learning_rate": 0.0001353641814516185, | |
| "loss": 0.1254, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.3014430785676109, | |
| "grad_norm": 0.15845873951911926, | |
| "learning_rate": 0.00013494641795990986, | |
| "loss": 0.1217, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.30251202565473007, | |
| "grad_norm": 0.16994373500347137, | |
| "learning_rate": 0.00013452795855291977, | |
| "loss": 0.1282, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.30358097274184925, | |
| "grad_norm": 0.1903841644525528, | |
| "learning_rate": 0.00013410881156375684, | |
| "loss": 0.1397, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.30464991982896844, | |
| "grad_norm": 0.17168985307216644, | |
| "learning_rate": 0.000133688985339222, | |
| "loss": 0.1322, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.30571886691608763, | |
| "grad_norm": 0.17627929151058197, | |
| "learning_rate": 0.00013326848823964243, | |
| "loss": 0.1262, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.3067878140032068, | |
| "grad_norm": 0.18915101885795593, | |
| "learning_rate": 0.00013284732863870475, | |
| "loss": 0.1422, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.307856761090326, | |
| "grad_norm": 0.15876813232898712, | |
| "learning_rate": 0.00013242551492328875, | |
| "loss": 0.1197, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.3089257081774452, | |
| "grad_norm": 0.15107540786266327, | |
| "learning_rate": 0.00013200305549329995, | |
| "loss": 0.1146, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.3099946552645644, | |
| "grad_norm": 0.16001242399215698, | |
| "learning_rate": 0.0001315799587615025, | |
| "loss": 0.1027, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.31106360235168357, | |
| "grad_norm": 0.166344553232193, | |
| "learning_rate": 0.00013115623315335172, | |
| "loss": 0.1282, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.31213254943880275, | |
| "grad_norm": 0.17358538508415222, | |
| "learning_rate": 0.0001307318871068261, | |
| "loss": 0.1266, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.31320149652592194, | |
| "grad_norm": 0.1970827579498291, | |
| "learning_rate": 0.00013030692907225956, | |
| "loss": 0.1049, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.31427044361304113, | |
| "grad_norm": 0.19751529395580292, | |
| "learning_rate": 0.00012988136751217291, | |
| "loss": 0.1383, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.3153393907001603, | |
| "grad_norm": 0.24741291999816895, | |
| "learning_rate": 0.00012945521090110547, | |
| "loss": 0.158, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.3164083377872795, | |
| "grad_norm": 0.16440382599830627, | |
| "learning_rate": 0.00012902846772544624, | |
| "loss": 0.1189, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.3174772848743987, | |
| "grad_norm": 0.2023131549358368, | |
| "learning_rate": 0.00012860114648326502, | |
| "loss": 0.1633, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.3185462319615179, | |
| "grad_norm": 0.2175939977169037, | |
| "learning_rate": 0.00012817325568414297, | |
| "loss": 0.1672, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.31961517904863707, | |
| "grad_norm": 0.17920686304569244, | |
| "learning_rate": 0.00012774480384900345, | |
| "loss": 0.1236, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.32068412613575625, | |
| "grad_norm": 0.18356698751449585, | |
| "learning_rate": 0.000127315799509942, | |
| "loss": 0.1199, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.32068412613575625, | |
| "eval_loss": 0.13152572512626648, | |
| "eval_runtime": 289.0306, | |
| "eval_samples_per_second": 4.318, | |
| "eval_steps_per_second": 1.079, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.32175307322287544, | |
| "grad_norm": 0.17195501923561096, | |
| "learning_rate": 0.00012688625121005668, | |
| "loss": 0.1167, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.32282202030999463, | |
| "grad_norm": 0.22263203561306, | |
| "learning_rate": 0.0001264561675032779, | |
| "loss": 0.1406, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.3238909673971138, | |
| "grad_norm": 0.18973477184772491, | |
| "learning_rate": 0.000126025556954198, | |
| "loss": 0.1294, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.324959914484233, | |
| "grad_norm": 0.15628819167613983, | |
| "learning_rate": 0.00012559442813790076, | |
| "loss": 0.1193, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.3260288615713522, | |
| "grad_norm": 0.16901010274887085, | |
| "learning_rate": 0.00012516278963979057, | |
| "loss": 0.1426, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.3270978086584714, | |
| "grad_norm": 0.1784602552652359, | |
| "learning_rate": 0.00012473065005542155, | |
| "loss": 0.1152, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.32816675574559057, | |
| "grad_norm": 0.2367001473903656, | |
| "learning_rate": 0.0001242980179903264, | |
| "loss": 0.1611, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.32923570283270975, | |
| "grad_norm": 0.21481367945671082, | |
| "learning_rate": 0.00012386490205984488, | |
| "loss": 0.1334, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.33030464991982894, | |
| "grad_norm": 0.32029592990875244, | |
| "learning_rate": 0.00012343131088895235, | |
| "loss": 0.1074, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.33137359700694813, | |
| "grad_norm": 0.1740630567073822, | |
| "learning_rate": 0.00012299725311208808, | |
| "loss": 0.0963, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.3324425440940673, | |
| "grad_norm": 0.17548616230487823, | |
| "learning_rate": 0.00012256273737298312, | |
| "loss": 0.1164, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.33351149118118656, | |
| "grad_norm": 0.18481452763080597, | |
| "learning_rate": 0.00012212777232448838, | |
| "loss": 0.1345, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.33458043826830575, | |
| "grad_norm": 0.18825702369213104, | |
| "learning_rate": 0.00012169236662840212, | |
| "loss": 0.1528, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.33564938535542493, | |
| "grad_norm": 0.17717814445495605, | |
| "learning_rate": 0.00012125652895529766, | |
| "loss": 0.1194, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.3367183324425441, | |
| "grad_norm": 0.25791436433792114, | |
| "learning_rate": 0.0001208202679843506, | |
| "loss": 0.1718, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.3377872795296633, | |
| "grad_norm": 0.1630530208349228, | |
| "learning_rate": 0.00012038359240316589, | |
| "loss": 0.1178, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.3388562266167825, | |
| "grad_norm": 0.16724711656570435, | |
| "learning_rate": 0.00011994651090760505, | |
| "loss": 0.147, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.3399251737039017, | |
| "grad_norm": 0.1993224173784256, | |
| "learning_rate": 0.00011950903220161285, | |
| "loss": 0.1492, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.34099412079102087, | |
| "grad_norm": 0.15769435465335846, | |
| "learning_rate": 0.00011907116499704401, | |
| "loss": 0.1246, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.34206306787814006, | |
| "grad_norm": 0.1998382806777954, | |
| "learning_rate": 0.0001186329180134898, | |
| "loss": 0.1245, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.34313201496525925, | |
| "grad_norm": 0.20432628691196442, | |
| "learning_rate": 0.00011819429997810418, | |
| "loss": 0.164, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.34420096205237843, | |
| "grad_norm": 0.315298467874527, | |
| "learning_rate": 0.00011775531962543036, | |
| "loss": 0.1371, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.3452699091394976, | |
| "grad_norm": 0.18633940815925598, | |
| "learning_rate": 0.00011731598569722642, | |
| "loss": 0.1138, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.3463388562266168, | |
| "grad_norm": 0.1944500356912613, | |
| "learning_rate": 0.0001168763069422916, | |
| "loss": 0.1379, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.347407803313736, | |
| "grad_norm": 0.22939671576023102, | |
| "learning_rate": 0.000116436292116292, | |
| "loss": 0.1119, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.3484767504008552, | |
| "grad_norm": 0.175564244389534, | |
| "learning_rate": 0.00011599594998158602, | |
| "loss": 0.1322, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.34954569748797437, | |
| "grad_norm": 0.17719417810440063, | |
| "learning_rate": 0.00011555528930705013, | |
| "loss": 0.158, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.35061464457509356, | |
| "grad_norm": 0.17542250454425812, | |
| "learning_rate": 0.00011511431886790407, | |
| "loss": 0.1422, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.35168359166221275, | |
| "grad_norm": 0.16461284458637238, | |
| "learning_rate": 0.00011467304744553618, | |
| "loss": 0.1319, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.35275253874933193, | |
| "grad_norm": 0.2131093442440033, | |
| "learning_rate": 0.00011423148382732853, | |
| "loss": 0.1429, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.3538214858364511, | |
| "grad_norm": 0.17762120068073273, | |
| "learning_rate": 0.00011378963680648187, | |
| "loss": 0.1444, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.3548904329235703, | |
| "grad_norm": 0.1586885303258896, | |
| "learning_rate": 0.00011334751518184061, | |
| "loss": 0.1067, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.3559593800106895, | |
| "grad_norm": 0.1545344591140747, | |
| "learning_rate": 0.00011290512775771758, | |
| "loss": 0.1382, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.3570283270978087, | |
| "grad_norm": 0.15905602276325226, | |
| "learning_rate": 0.0001124624833437186, | |
| "loss": 0.1121, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.35809727418492787, | |
| "grad_norm": 0.22019295394420624, | |
| "learning_rate": 0.0001120195907545672, | |
| "loss": 0.1273, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.35916622127204706, | |
| "grad_norm": 0.15988142788410187, | |
| "learning_rate": 0.00011157645880992902, | |
| "loss": 0.1064, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.36023516835916625, | |
| "grad_norm": 0.17529579997062683, | |
| "learning_rate": 0.0001111330963342361, | |
| "loss": 0.12, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.36130411544628543, | |
| "grad_norm": 0.15964984893798828, | |
| "learning_rate": 0.00011068951215651132, | |
| "loss": 0.124, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.3623730625334046, | |
| "grad_norm": 0.14573580026626587, | |
| "learning_rate": 0.00011024571511019235, | |
| "loss": 0.1018, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.3634420096205238, | |
| "grad_norm": 0.1717827022075653, | |
| "learning_rate": 0.0001098017140329561, | |
| "loss": 0.1214, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.364510956707643, | |
| "grad_norm": 0.1566823571920395, | |
| "learning_rate": 0.00010935751776654224, | |
| "loss": 0.1022, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.3655799037947622, | |
| "grad_norm": 0.16118775308132172, | |
| "learning_rate": 0.0001089131351565776, | |
| "loss": 0.1235, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.36664885088188137, | |
| "grad_norm": 0.16470201313495636, | |
| "learning_rate": 0.00010846857505239973, | |
| "loss": 0.112, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.36771779796900056, | |
| "grad_norm": 0.14973486959934235, | |
| "learning_rate": 0.00010802384630688078, | |
| "loss": 0.125, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.36878674505611975, | |
| "grad_norm": 0.18552237749099731, | |
| "learning_rate": 0.00010757895777625118, | |
| "loss": 0.1203, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.36985569214323893, | |
| "grad_norm": 0.2491609752178192, | |
| "learning_rate": 0.00010713391831992323, | |
| "loss": 0.1705, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.3709246392303581, | |
| "grad_norm": 0.20768113434314728, | |
| "learning_rate": 0.00010668873680031486, | |
| "loss": 0.1483, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.3719935863174773, | |
| "grad_norm": 0.20317085087299347, | |
| "learning_rate": 0.00010624342208267292, | |
| "loss": 0.1364, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.3730625334045965, | |
| "grad_norm": 0.16608084738254547, | |
| "learning_rate": 0.00010579798303489675, | |
| "loss": 0.1046, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.3741314804917157, | |
| "grad_norm": 0.2229534536600113, | |
| "learning_rate": 0.00010535242852736151, | |
| "loss": 0.1065, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.37520042757883487, | |
| "grad_norm": 0.15245893597602844, | |
| "learning_rate": 0.00010490676743274181, | |
| "loss": 0.1015, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.37626937466595406, | |
| "grad_norm": 0.17506876587867737, | |
| "learning_rate": 0.00010446100862583459, | |
| "loss": 0.1406, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.37733832175307325, | |
| "grad_norm": 0.15572670102119446, | |
| "learning_rate": 0.00010401516098338271, | |
| "loss": 0.1031, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.37840726884019243, | |
| "grad_norm": 0.17110490798950195, | |
| "learning_rate": 0.00010356923338389806, | |
| "loss": 0.1194, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.3794762159273116, | |
| "grad_norm": 0.20592443645000458, | |
| "learning_rate": 0.00010312323470748478, | |
| "loss": 0.1529, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.3805451630144308, | |
| "grad_norm": 0.18145787715911865, | |
| "learning_rate": 0.00010267717383566246, | |
| "loss": 0.1253, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.38161411010155, | |
| "grad_norm": 0.15736672282218933, | |
| "learning_rate": 0.00010223105965118913, | |
| "loss": 0.1104, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.3826830571886692, | |
| "grad_norm": 0.17708028852939606, | |
| "learning_rate": 0.0001017849010378846, | |
| "loss": 0.1253, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.38375200427578837, | |
| "grad_norm": 0.20931871235370636, | |
| "learning_rate": 0.0001013387068804534, | |
| "loss": 0.1364, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.38482095136290756, | |
| "grad_norm": 0.200309619307518, | |
| "learning_rate": 0.00010089248606430775, | |
| "loss": 0.1235, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.38588989845002675, | |
| "grad_norm": 0.19272620975971222, | |
| "learning_rate": 0.00010044624747539094, | |
| "loss": 0.1662, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.38695884553714593, | |
| "grad_norm": 0.1405670940876007, | |
| "learning_rate": 0.0001, | |
| "loss": 0.1104, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.3880277926242651, | |
| "grad_norm": 0.15820738673210144, | |
| "learning_rate": 9.95537525246091e-05, | |
| "loss": 0.1158, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.3890967397113843, | |
| "grad_norm": 0.17021127045154572, | |
| "learning_rate": 9.910751393569229e-05, | |
| "loss": 0.1366, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.3901656867985035, | |
| "grad_norm": 0.19209320843219757, | |
| "learning_rate": 9.866129311954664e-05, | |
| "loss": 0.1362, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.3912346338856227, | |
| "grad_norm": 0.23325751721858978, | |
| "learning_rate": 9.821509896211539e-05, | |
| "loss": 0.1575, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.39230358097274187, | |
| "grad_norm": 0.1879902184009552, | |
| "learning_rate": 9.776894034881089e-05, | |
| "loss": 0.153, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.39337252805986106, | |
| "grad_norm": 0.15710167586803436, | |
| "learning_rate": 9.732282616433756e-05, | |
| "loss": 0.1192, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.39444147514698025, | |
| "grad_norm": 0.13521148264408112, | |
| "learning_rate": 9.687676529251526e-05, | |
| "loss": 0.1033, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.39551042223409943, | |
| "grad_norm": 0.16447614133358002, | |
| "learning_rate": 9.643076661610196e-05, | |
| "loss": 0.1361, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.3965793693212186, | |
| "grad_norm": 0.17937707901000977, | |
| "learning_rate": 9.598483901661731e-05, | |
| "loss": 0.0982, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.3976483164083378, | |
| "grad_norm": 0.16494877636432648, | |
| "learning_rate": 9.553899137416545e-05, | |
| "loss": 0.1331, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.398717263495457, | |
| "grad_norm": 0.12207219749689102, | |
| "learning_rate": 9.509323256725821e-05, | |
| "loss": 0.0951, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.3997862105825762, | |
| "grad_norm": 0.15613038837909698, | |
| "learning_rate": 9.464757147263849e-05, | |
| "loss": 0.1296, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.40085515766969537, | |
| "grad_norm": 0.15942755341529846, | |
| "learning_rate": 9.420201696510332e-05, | |
| "loss": 0.1099, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.40192410475681456, | |
| "grad_norm": 0.13835251331329346, | |
| "learning_rate": 9.37565779173271e-05, | |
| "loss": 0.0892, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.40299305184393375, | |
| "grad_norm": 0.17169111967086792, | |
| "learning_rate": 9.331126319968514e-05, | |
| "loss": 0.1262, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.40406199893105293, | |
| "grad_norm": 0.16025681793689728, | |
| "learning_rate": 9.286608168007678e-05, | |
| "loss": 0.1191, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.4051309460181721, | |
| "grad_norm": 0.14073841273784637, | |
| "learning_rate": 9.242104222374886e-05, | |
| "loss": 0.0932, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.4061998931052913, | |
| "grad_norm": 0.15486131608486176, | |
| "learning_rate": 9.197615369311925e-05, | |
| "loss": 0.0997, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.4072688401924105, | |
| "grad_norm": 0.1486068218946457, | |
| "learning_rate": 9.153142494760028e-05, | |
| "loss": 0.1164, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.4083377872795297, | |
| "grad_norm": 0.14497411251068115, | |
| "learning_rate": 9.108686484342241e-05, | |
| "loss": 0.0991, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.4094067343666489, | |
| "grad_norm": 0.18235650658607483, | |
| "learning_rate": 9.06424822334578e-05, | |
| "loss": 0.1332, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.41047568145376806, | |
| "grad_norm": 0.17887920141220093, | |
| "learning_rate": 9.019828596704394e-05, | |
| "loss": 0.1429, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.41154462854088725, | |
| "grad_norm": 0.16443820297718048, | |
| "learning_rate": 8.975428488980765e-05, | |
| "loss": 0.1039, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.41261357562800643, | |
| "grad_norm": 0.17346209287643433, | |
| "learning_rate": 8.931048784348875e-05, | |
| "loss": 0.1193, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.4136825227151256, | |
| "grad_norm": 0.14269988238811493, | |
| "learning_rate": 8.886690366576394e-05, | |
| "loss": 0.0907, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.4147514698022448, | |
| "grad_norm": 0.18494229018688202, | |
| "learning_rate": 8.842354119007099e-05, | |
| "loss": 0.1241, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.415820416889364, | |
| "grad_norm": 0.14606322348117828, | |
| "learning_rate": 8.798040924543281e-05, | |
| "loss": 0.0939, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.4168893639764832, | |
| "grad_norm": 0.1913435161113739, | |
| "learning_rate": 8.753751665628141e-05, | |
| "loss": 0.1224, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.4179583110636024, | |
| "grad_norm": 0.16582347452640533, | |
| "learning_rate": 8.709487224228246e-05, | |
| "loss": 0.1047, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.41902725815072156, | |
| "grad_norm": 0.1860746145248413, | |
| "learning_rate": 8.665248481815941e-05, | |
| "loss": 0.1451, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.42009620523784075, | |
| "grad_norm": 0.18656174838542938, | |
| "learning_rate": 8.621036319351814e-05, | |
| "loss": 0.1295, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.42116515232495993, | |
| "grad_norm": 0.16289152204990387, | |
| "learning_rate": 8.57685161726715e-05, | |
| "loss": 0.1104, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.4222340994120791, | |
| "grad_norm": 0.1636841893196106, | |
| "learning_rate": 8.532695255446383e-05, | |
| "loss": 0.1181, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.4233030464991983, | |
| "grad_norm": 0.1686917096376419, | |
| "learning_rate": 8.488568113209593e-05, | |
| "loss": 0.1181, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.4243719935863175, | |
| "grad_norm": 0.14516602456569672, | |
| "learning_rate": 8.444471069294988e-05, | |
| "loss": 0.0933, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.4254409406734367, | |
| "grad_norm": 0.16165108978748322, | |
| "learning_rate": 8.400405001841399e-05, | |
| "loss": 0.1376, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.4265098877605559, | |
| "grad_norm": 0.2126784473657608, | |
| "learning_rate": 8.3563707883708e-05, | |
| "loss": 0.1541, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.42757883484767506, | |
| "grad_norm": 0.2085818201303482, | |
| "learning_rate": 8.312369305770843e-05, | |
| "loss": 0.1493, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.42757883484767506, | |
| "eval_loss": 0.1267417073249817, | |
| "eval_runtime": 288.9107, | |
| "eval_samples_per_second": 4.32, | |
| "eval_steps_per_second": 1.08, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.42864778193479425, | |
| "grad_norm": 0.15687406063079834, | |
| "learning_rate": 8.268401430277362e-05, | |
| "loss": 0.1135, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.42971672902191344, | |
| "grad_norm": 0.233999103307724, | |
| "learning_rate": 8.224468037456969e-05, | |
| "loss": 0.1453, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.4307856761090326, | |
| "grad_norm": 0.18340054154396057, | |
| "learning_rate": 8.180570002189583e-05, | |
| "loss": 0.1268, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.4318546231961518, | |
| "grad_norm": 0.16533297300338745, | |
| "learning_rate": 8.136708198651021e-05, | |
| "loss": 0.1216, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.432923570283271, | |
| "grad_norm": 0.1564950942993164, | |
| "learning_rate": 8.0928835002956e-05, | |
| "loss": 0.1088, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.4339925173703902, | |
| "grad_norm": 0.17598888278007507, | |
| "learning_rate": 8.049096779838719e-05, | |
| "loss": 0.1117, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.4350614644575094, | |
| "grad_norm": 0.16911205649375916, | |
| "learning_rate": 8.005348909239499e-05, | |
| "loss": 0.1226, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.43613041154462856, | |
| "grad_norm": 0.17374524474143982, | |
| "learning_rate": 7.961640759683416e-05, | |
| "loss": 0.1104, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.43719935863174775, | |
| "grad_norm": 0.14513778686523438, | |
| "learning_rate": 7.917973201564943e-05, | |
| "loss": 0.0912, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.43826830571886694, | |
| "grad_norm": 0.16398054361343384, | |
| "learning_rate": 7.874347104470234e-05, | |
| "loss": 0.1062, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.4393372528059861, | |
| "grad_norm": 0.17156217992305756, | |
| "learning_rate": 7.830763337159789e-05, | |
| "loss": 0.1221, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.4404061998931053, | |
| "grad_norm": 0.15586666762828827, | |
| "learning_rate": 7.787222767551164e-05, | |
| "loss": 0.1188, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.4414751469802245, | |
| "grad_norm": 0.20987504720687866, | |
| "learning_rate": 7.743726262701693e-05, | |
| "loss": 0.1527, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.4425440940673437, | |
| "grad_norm": 0.18425814807415009, | |
| "learning_rate": 7.700274688791196e-05, | |
| "loss": 0.1495, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.4436130411544629, | |
| "grad_norm": 0.1313970386981964, | |
| "learning_rate": 7.656868911104766e-05, | |
| "loss": 0.0865, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.44468198824158206, | |
| "grad_norm": 0.12412843853235245, | |
| "learning_rate": 7.613509794015517e-05, | |
| "loss": 0.0849, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.44575093532870125, | |
| "grad_norm": 0.18094252049922943, | |
| "learning_rate": 7.570198200967362e-05, | |
| "loss": 0.1186, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.44681988241582044, | |
| "grad_norm": 0.1528427004814148, | |
| "learning_rate": 7.526934994457844e-05, | |
| "loss": 0.127, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.4478888295029396, | |
| "grad_norm": 0.16211071610450745, | |
| "learning_rate": 7.483721036020948e-05, | |
| "loss": 0.1268, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.4489577765900588, | |
| "grad_norm": 0.1652361899614334, | |
| "learning_rate": 7.440557186209926e-05, | |
| "loss": 0.11, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.450026723677178, | |
| "grad_norm": 0.17334988713264465, | |
| "learning_rate": 7.3974443045802e-05, | |
| "loss": 0.1316, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.4510956707642972, | |
| "grad_norm": 0.1573891043663025, | |
| "learning_rate": 7.354383249672212e-05, | |
| "loss": 0.129, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.4521646178514164, | |
| "grad_norm": 0.17436981201171875, | |
| "learning_rate": 7.311374878994334e-05, | |
| "loss": 0.1294, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.45323356493853556, | |
| "grad_norm": 0.22541721165180206, | |
| "learning_rate": 7.268420049005807e-05, | |
| "loss": 0.1435, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.45430251202565475, | |
| "grad_norm": 0.19348768889904022, | |
| "learning_rate": 7.225519615099659e-05, | |
| "loss": 0.117, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.45537145911277394, | |
| "grad_norm": 0.1548422873020172, | |
| "learning_rate": 7.182674431585704e-05, | |
| "loss": 0.1094, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.4564404061998931, | |
| "grad_norm": 0.17469367384910583, | |
| "learning_rate": 7.139885351673503e-05, | |
| "loss": 0.1235, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.4575093532870123, | |
| "grad_norm": 0.20104283094406128, | |
| "learning_rate": 7.097153227455379e-05, | |
| "loss": 0.1051, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.4585783003741315, | |
| "grad_norm": 0.15787599980831146, | |
| "learning_rate": 7.054478909889455e-05, | |
| "loss": 0.1212, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.4596472474612507, | |
| "grad_norm": 0.1746998131275177, | |
| "learning_rate": 7.011863248782711e-05, | |
| "loss": 0.1163, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.4607161945483699, | |
| "grad_norm": 0.15664790570735931, | |
| "learning_rate": 6.969307092774047e-05, | |
| "loss": 0.1072, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.46178514163548906, | |
| "grad_norm": 0.18459376692771912, | |
| "learning_rate": 6.92681128931739e-05, | |
| "loss": 0.1225, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.46285408872260825, | |
| "grad_norm": 0.18352118134498596, | |
| "learning_rate": 6.884376684664832e-05, | |
| "loss": 0.125, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.46392303580972744, | |
| "grad_norm": 0.16230523586273193, | |
| "learning_rate": 6.842004123849752e-05, | |
| "loss": 0.1244, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.4649919828968466, | |
| "grad_norm": 0.15808574855327606, | |
| "learning_rate": 6.799694450670008e-05, | |
| "loss": 0.1047, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.4660609299839658, | |
| "grad_norm": 0.1523110568523407, | |
| "learning_rate": 6.757448507671128e-05, | |
| "loss": 0.1243, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.467129877071085, | |
| "grad_norm": 0.15416131913661957, | |
| "learning_rate": 6.715267136129525e-05, | |
| "loss": 0.0946, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 0.4681988241582042, | |
| "grad_norm": 0.170943483710289, | |
| "learning_rate": 6.673151176035762e-05, | |
| "loss": 0.1148, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.4692677712453234, | |
| "grad_norm": 0.13973015546798706, | |
| "learning_rate": 6.6311014660778e-05, | |
| "loss": 0.0978, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 0.47033671833244256, | |
| "grad_norm": 0.1479904055595398, | |
| "learning_rate": 6.589118843624315e-05, | |
| "loss": 0.1199, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.47140566541956175, | |
| "grad_norm": 0.14838270843029022, | |
| "learning_rate": 6.547204144708025e-05, | |
| "loss": 0.1009, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.47247461250668094, | |
| "grad_norm": 0.18144488334655762, | |
| "learning_rate": 6.505358204009017e-05, | |
| "loss": 0.0992, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 0.4735435595938001, | |
| "grad_norm": 0.15086080133914948, | |
| "learning_rate": 6.463581854838152e-05, | |
| "loss": 0.0959, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 0.4746125066809193, | |
| "grad_norm": 0.15263791382312775, | |
| "learning_rate": 6.421875929120469e-05, | |
| "loss": 0.1061, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.4756814537680385, | |
| "grad_norm": 0.18711231648921967, | |
| "learning_rate": 6.380241257378585e-05, | |
| "loss": 0.1395, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.4767504008551577, | |
| "grad_norm": 0.16735321283340454, | |
| "learning_rate": 6.338678668716209e-05, | |
| "loss": 0.1099, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 0.4778193479422769, | |
| "grad_norm": 0.17839953303337097, | |
| "learning_rate": 6.297188990801584e-05, | |
| "loss": 0.1316, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 0.47888829502939606, | |
| "grad_norm": 0.14725211262702942, | |
| "learning_rate": 6.25577304985103e-05, | |
| "loss": 0.1034, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.47995724211651525, | |
| "grad_norm": 0.1480267494916916, | |
| "learning_rate": 6.214431670612494e-05, | |
| "loss": 0.109, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 0.48102618920363444, | |
| "grad_norm": 0.17610524594783783, | |
| "learning_rate": 6.173165676349103e-05, | |
| "loss": 0.1159, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.4820951362907536, | |
| "grad_norm": 0.17769086360931396, | |
| "learning_rate": 6.131975888822793e-05, | |
| "loss": 0.1002, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 0.4831640833778728, | |
| "grad_norm": 0.1791263222694397, | |
| "learning_rate": 6.090863128277938e-05, | |
| "loss": 0.1291, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 0.484233030464992, | |
| "grad_norm": 0.20171315968036652, | |
| "learning_rate": 6.049828213425002e-05, | |
| "loss": 0.1383, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 0.4853019775521112, | |
| "grad_norm": 0.182616725564003, | |
| "learning_rate": 6.008871961424258e-05, | |
| "loss": 0.1145, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 0.4863709246392304, | |
| "grad_norm": 0.21832366287708282, | |
| "learning_rate": 5.967995187869505e-05, | |
| "loss": 0.1562, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.48743987172634956, | |
| "grad_norm": 0.1896723210811615, | |
| "learning_rate": 5.9271987067718125e-05, | |
| "loss": 0.1156, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 0.48850881881346875, | |
| "grad_norm": 0.1710599958896637, | |
| "learning_rate": 5.8864833305433457e-05, | |
| "loss": 0.1077, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 0.48957776590058794, | |
| "grad_norm": 0.17611974477767944, | |
| "learning_rate": 5.845849869981137e-05, | |
| "loss": 0.0891, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 0.4906467129877071, | |
| "grad_norm": 0.20914289355278015, | |
| "learning_rate": 5.805299134250991e-05, | |
| "loss": 0.1424, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 0.4917156600748263, | |
| "grad_norm": 0.17185574769973755, | |
| "learning_rate": 5.7648319308713464e-05, | |
| "loss": 0.1043, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.4927846071619455, | |
| "grad_norm": 0.17580777406692505, | |
| "learning_rate": 5.7244490656971815e-05, | |
| "loss": 0.0928, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 0.4938535542490647, | |
| "grad_norm": 0.13630133867263794, | |
| "learning_rate": 5.684151342903992e-05, | |
| "loss": 0.096, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 0.4949225013361839, | |
| "grad_norm": 0.13813966512680054, | |
| "learning_rate": 5.643939564971779e-05, | |
| "loss": 0.0951, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 0.49599144842330306, | |
| "grad_norm": 0.15008163452148438, | |
| "learning_rate": 5.603814532669032e-05, | |
| "loss": 0.0785, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 0.49706039551042225, | |
| "grad_norm": 0.17430590093135834, | |
| "learning_rate": 5.5637770450368186e-05, | |
| "loss": 0.0941, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.49812934259754144, | |
| "grad_norm": 0.131693497300148, | |
| "learning_rate": 5.5238278993728756e-05, | |
| "loss": 0.0869, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 0.4991982896846606, | |
| "grad_norm": 0.14627690613269806, | |
| "learning_rate": 5.483967891215691e-05, | |
| "loss": 0.1071, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 0.5002672367717798, | |
| "grad_norm": 0.20751060545444489, | |
| "learning_rate": 5.4441978143287066e-05, | |
| "loss": 0.113, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 0.501336183858899, | |
| "grad_norm": 0.16088207066059113, | |
| "learning_rate": 5.4045184606844834e-05, | |
| "loss": 0.1283, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 0.5024051309460181, | |
| "grad_norm": 0.15302957594394684, | |
| "learning_rate": 5.364930620448946e-05, | |
| "loss": 0.1173, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.5034740780331374, | |
| "grad_norm": 0.16366977989673615, | |
| "learning_rate": 5.3254350819656374e-05, | |
| "loss": 0.1184, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 0.5045430251202565, | |
| "grad_norm": 0.18926464021205902, | |
| "learning_rate": 5.286032631740023e-05, | |
| "loss": 0.1403, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 0.5056119722073757, | |
| "grad_norm": 0.21518391370773315, | |
| "learning_rate": 5.246724054423834e-05, | |
| "loss": 0.1348, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 0.5066809192944949, | |
| "grad_norm": 0.19751004874706268, | |
| "learning_rate": 5.207510132799436e-05, | |
| "loss": 0.127, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 0.5077498663816141, | |
| "grad_norm": 0.23470665514469147, | |
| "learning_rate": 5.168391647764227e-05, | |
| "loss": 0.139, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.5088188134687333, | |
| "grad_norm": 0.18137036263942719, | |
| "learning_rate": 5.1293693783151275e-05, | |
| "loss": 0.1213, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 0.5098877605558525, | |
| "grad_norm": 0.16304810345172882, | |
| "learning_rate": 5.090444101533029e-05, | |
| "loss": 0.1206, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 0.5109567076429716, | |
| "grad_norm": 0.16015806794166565, | |
| "learning_rate": 5.051616592567323e-05, | |
| "loss": 0.114, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 0.5120256547300909, | |
| "grad_norm": 0.18064187467098236, | |
| "learning_rate": 5.012887624620496e-05, | |
| "loss": 0.1103, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 0.51309460181721, | |
| "grad_norm": 0.18429364264011383, | |
| "learning_rate": 4.9742579689326874e-05, | |
| "loss": 0.1344, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.5141635489043292, | |
| "grad_norm": 0.13880470395088196, | |
| "learning_rate": 4.9357283947663676e-05, | |
| "loss": 0.097, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 0.5152324959914484, | |
| "grad_norm": 0.15282510221004486, | |
| "learning_rate": 4.8972996693910054e-05, | |
| "loss": 0.0952, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 0.5163014430785676, | |
| "grad_norm": 0.1321898102760315, | |
| "learning_rate": 4.8589725580677835e-05, | |
| "loss": 0.0896, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 0.5173703901656868, | |
| "grad_norm": 0.15336064994335175, | |
| "learning_rate": 4.82074782403437e-05, | |
| "loss": 0.1185, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 0.518439337252806, | |
| "grad_norm": 0.14170211553573608, | |
| "learning_rate": 4.7826262284897095e-05, | |
| "loss": 0.1037, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.5195082843399251, | |
| "grad_norm": 0.16489258408546448, | |
| "learning_rate": 4.7446085305788725e-05, | |
| "loss": 0.1238, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 0.5205772314270444, | |
| "grad_norm": 0.16994819045066833, | |
| "learning_rate": 4.706695487377932e-05, | |
| "loss": 0.1086, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 0.5216461785141635, | |
| "grad_norm": 0.17084138095378876, | |
| "learning_rate": 4.668887853878896e-05, | |
| "loss": 0.1088, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 0.5227151256012827, | |
| "grad_norm": 0.169681578874588, | |
| "learning_rate": 4.6311863829746596e-05, | |
| "loss": 0.1055, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 0.5237840726884019, | |
| "grad_norm": 0.16820192337036133, | |
| "learning_rate": 4.593591825444028e-05, | |
| "loss": 0.0947, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.5248530197755211, | |
| "grad_norm": 0.1566527783870697, | |
| "learning_rate": 4.55610492993674e-05, | |
| "loss": 0.0913, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 0.5259219668626403, | |
| "grad_norm": 0.20646369457244873, | |
| "learning_rate": 4.518726442958599e-05, | |
| "loss": 0.1707, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 0.5269909139497595, | |
| "grad_norm": 0.17909252643585205, | |
| "learning_rate": 4.4814571088565735e-05, | |
| "loss": 0.1444, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 0.5280598610368786, | |
| "grad_norm": 0.15825694799423218, | |
| "learning_rate": 4.444297669803981e-05, | |
| "loss": 0.1026, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 0.5291288081239979, | |
| "grad_norm": 0.16258923709392548, | |
| "learning_rate": 4.407248865785716e-05, | |
| "loss": 0.1118, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.530197755211117, | |
| "grad_norm": 0.3853148818016052, | |
| "learning_rate": 4.370311434583525e-05, | |
| "loss": 0.1833, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 0.5312667022982362, | |
| "grad_norm": 0.1683734655380249, | |
| "learning_rate": 4.333486111761279e-05, | |
| "loss": 0.1154, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 0.5323356493853554, | |
| "grad_norm": 0.12579625844955444, | |
| "learning_rate": 4.296773630650358e-05, | |
| "loss": 0.0924, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 0.5334045964724746, | |
| "grad_norm": 0.15746884047985077, | |
| "learning_rate": 4.260174722335046e-05, | |
| "loss": 0.1052, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 0.5344735435595938, | |
| "grad_norm": 0.15484918653964996, | |
| "learning_rate": 4.223690115637944e-05, | |
| "loss": 0.1023, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.5344735435595938, | |
| "eval_loss": 0.1166808158159256, | |
| "eval_runtime": 288.9178, | |
| "eval_samples_per_second": 4.32, | |
| "eval_steps_per_second": 1.08, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.535542490646713, | |
| "grad_norm": 0.16423402726650238, | |
| "learning_rate": 4.187320537105489e-05, | |
| "loss": 0.1262, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 0.5366114377338321, | |
| "grad_norm": 0.19593016803264618, | |
| "learning_rate": 4.15106671099347e-05, | |
| "loss": 0.1401, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 0.5376803848209514, | |
| "grad_norm": 0.13396374881267548, | |
| "learning_rate": 4.114929359252606e-05, | |
| "loss": 0.1002, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 0.5387493319080705, | |
| "grad_norm": 0.18291138112545013, | |
| "learning_rate": 4.0789092015141714e-05, | |
| "loss": 0.1332, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 0.5398182789951897, | |
| "grad_norm": 0.16943904757499695, | |
| "learning_rate": 4.0430069550756665e-05, | |
| "loss": 0.1018, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.5408872260823089, | |
| "grad_norm": 0.1741015464067459, | |
| "learning_rate": 4.007223334886531e-05, | |
| "loss": 0.1224, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 0.5419561731694281, | |
| "grad_norm": 0.13967184722423553, | |
| "learning_rate": 3.971559053533906e-05, | |
| "loss": 0.105, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 0.5430251202565473, | |
| "grad_norm": 0.1619994342327118, | |
| "learning_rate": 3.9360148212284475e-05, | |
| "loss": 0.0997, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 0.5440940673436665, | |
| "grad_norm": 0.1519555002450943, | |
| "learning_rate": 3.900591345790181e-05, | |
| "loss": 0.1045, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 0.5451630144307856, | |
| "grad_norm": 0.16195788979530334, | |
| "learning_rate": 3.865289332634407e-05, | |
| "loss": 0.1041, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.5462319615179049, | |
| "grad_norm": 0.1591130644083023, | |
| "learning_rate": 3.8301094847576446e-05, | |
| "loss": 0.1128, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 0.547300908605024, | |
| "grad_norm": 0.15072457492351532, | |
| "learning_rate": 3.7950525027236585e-05, | |
| "loss": 0.1148, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 0.5483698556921432, | |
| "grad_norm": 0.16123512387275696, | |
| "learning_rate": 3.760119084649468e-05, | |
| "loss": 0.1088, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 0.5494388027792624, | |
| "grad_norm": 0.152877539396286, | |
| "learning_rate": 3.725309926191479e-05, | |
| "loss": 0.1065, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 0.5505077498663816, | |
| "grad_norm": 0.23673762381076813, | |
| "learning_rate": 3.690625720531627e-05, | |
| "loss": 0.1095, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.5515766969535008, | |
| "grad_norm": 0.13103598356246948, | |
| "learning_rate": 3.6560671583635467e-05, | |
| "loss": 0.095, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 0.55264564404062, | |
| "grad_norm": 0.1740058958530426, | |
| "learning_rate": 3.621634927878846e-05, | |
| "loss": 0.1247, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 0.5537145911277391, | |
| "grad_norm": 0.20583245158195496, | |
| "learning_rate": 3.5873297147533915e-05, | |
| "loss": 0.1513, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 0.5547835382148584, | |
| "grad_norm": 0.1914791762828827, | |
| "learning_rate": 3.553152202133651e-05, | |
| "loss": 0.1267, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 0.5558524853019775, | |
| "grad_norm": 0.2231132984161377, | |
| "learning_rate": 3.519103070623096e-05, | |
| "loss": 0.148, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.5569214323890967, | |
| "grad_norm": 0.14965134859085083, | |
| "learning_rate": 3.485182998268642e-05, | |
| "loss": 0.1011, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 0.5579903794762159, | |
| "grad_norm": 0.1701858639717102, | |
| "learning_rate": 3.45139266054715e-05, | |
| "loss": 0.1145, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 0.5590593265633351, | |
| "grad_norm": 0.17237675189971924, | |
| "learning_rate": 3.4177327303519765e-05, | |
| "loss": 0.1403, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 0.5601282736504543, | |
| "grad_norm": 0.17186829447746277, | |
| "learning_rate": 3.384203877979559e-05, | |
| "loss": 0.1114, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 0.5611972207375735, | |
| "grad_norm": 0.17262551188468933, | |
| "learning_rate": 3.3508067711161004e-05, | |
| "loss": 0.0962, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.5622661678246926, | |
| "grad_norm": 0.1815585047006607, | |
| "learning_rate": 3.3175420748242406e-05, | |
| "loss": 0.1222, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 0.5633351149118119, | |
| "grad_norm": 0.20134767889976501, | |
| "learning_rate": 3.2844104515298155e-05, | |
| "loss": 0.1498, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 0.564404061998931, | |
| "grad_norm": 0.12754414975643158, | |
| "learning_rate": 3.2514125610086955e-05, | |
| "loss": 0.0829, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 0.5654730090860502, | |
| "grad_norm": 0.15976329147815704, | |
| "learning_rate": 3.218549060373615e-05, | |
| "loss": 0.0966, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 0.5665419561731694, | |
| "grad_norm": 0.17927618324756622, | |
| "learning_rate": 3.185820604061088e-05, | |
| "loss": 0.1234, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.5676109032602886, | |
| "grad_norm": 0.14203406870365143, | |
| "learning_rate": 3.153227843818405e-05, | |
| "loss": 0.0785, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 0.5686798503474078, | |
| "grad_norm": 0.18028829991817474, | |
| "learning_rate": 3.1207714286906256e-05, | |
| "loss": 0.1068, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 0.569748797434527, | |
| "grad_norm": 0.1877748668193817, | |
| "learning_rate": 3.088452005007656e-05, | |
| "loss": 0.1479, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 0.5708177445216461, | |
| "grad_norm": 0.20719939470291138, | |
| "learning_rate": 3.056270216371395e-05, | |
| "loss": 0.164, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 0.5718866916087654, | |
| "grad_norm": 0.16265811026096344, | |
| "learning_rate": 3.0242267036429027e-05, | |
| "loss": 0.0998, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.5729556386958845, | |
| "grad_norm": 0.18794025480747223, | |
| "learning_rate": 2.9923221049296446e-05, | |
| "loss": 0.142, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 0.5740245857830037, | |
| "grad_norm": 0.142957404255867, | |
| "learning_rate": 2.9605570555727835e-05, | |
| "loss": 0.1258, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 0.5750935328701229, | |
| "grad_norm": 0.16433711349964142, | |
| "learning_rate": 2.9289321881345254e-05, | |
| "loss": 0.1155, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 0.5761624799572421, | |
| "grad_norm": 0.16358301043510437, | |
| "learning_rate": 2.897448132385525e-05, | |
| "loss": 0.1072, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 0.5772314270443613, | |
| "grad_norm": 0.14673899114131927, | |
| "learning_rate": 2.8661055152923456e-05, | |
| "loss": 0.1165, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.5783003741314805, | |
| "grad_norm": 0.15978941321372986, | |
| "learning_rate": 2.8349049610049683e-05, | |
| "loss": 0.0882, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 0.5793693212185996, | |
| "grad_norm": 0.16689325869083405, | |
| "learning_rate": 2.8038470908443714e-05, | |
| "loss": 0.1183, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 0.5804382683057189, | |
| "grad_norm": 0.19823519885540009, | |
| "learning_rate": 2.7729325232901494e-05, | |
| "loss": 0.1188, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 0.581507215392838, | |
| "grad_norm": 0.14368073642253876, | |
| "learning_rate": 2.742161873968202e-05, | |
| "loss": 0.0981, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 0.5825761624799572, | |
| "grad_norm": 0.16314393281936646, | |
| "learning_rate": 2.7115357556384756e-05, | |
| "loss": 0.0958, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.5836451095670764, | |
| "grad_norm": 0.15624821186065674, | |
| "learning_rate": 2.681054778182748e-05, | |
| "loss": 0.1096, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 0.5847140566541956, | |
| "grad_norm": 0.1835729330778122, | |
| "learning_rate": 2.650719548592502e-05, | |
| "loss": 0.1118, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 0.5857830037413148, | |
| "grad_norm": 0.18949463963508606, | |
| "learning_rate": 2.6205306709568354e-05, | |
| "loss": 0.1127, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 0.586851950828434, | |
| "grad_norm": 0.18184354901313782, | |
| "learning_rate": 2.5904887464504114e-05, | |
| "loss": 0.0934, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 0.5879208979155531, | |
| "grad_norm": 0.2154887169599533, | |
| "learning_rate": 2.5605943733215042e-05, | |
| "loss": 0.1444, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.5889898450026724, | |
| "grad_norm": 0.14278879761695862, | |
| "learning_rate": 2.530848146880097e-05, | |
| "loss": 0.1012, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 0.5900587920897915, | |
| "grad_norm": 0.15597015619277954, | |
| "learning_rate": 2.501250659485992e-05, | |
| "loss": 0.1025, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 0.5911277391769107, | |
| "grad_norm": 0.1862352341413498, | |
| "learning_rate": 2.4718025005370414e-05, | |
| "loss": 0.1163, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 0.5921966862640299, | |
| "grad_norm": 0.15955589711666107, | |
| "learning_rate": 2.4425042564574184e-05, | |
| "loss": 0.1337, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 0.5932656333511491, | |
| "grad_norm": 0.1371564120054245, | |
| "learning_rate": 2.4133565106859003e-05, | |
| "loss": 0.0946, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.5943345804382683, | |
| "grad_norm": 0.13667546212673187, | |
| "learning_rate": 2.38435984366429e-05, | |
| "loss": 0.0941, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 0.5954035275253875, | |
| "grad_norm": 0.12957549095153809, | |
| "learning_rate": 2.3555148328258368e-05, | |
| "loss": 0.0998, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 0.5964724746125066, | |
| "grad_norm": 0.1276206374168396, | |
| "learning_rate": 2.3268220525837437e-05, | |
| "loss": 0.0895, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 0.5975414216996259, | |
| "grad_norm": 0.15202417969703674, | |
| "learning_rate": 2.2982820743197243e-05, | |
| "loss": 0.0923, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 0.598610368786745, | |
| "grad_norm": 0.17828796803951263, | |
| "learning_rate": 2.26989546637263e-05, | |
| "loss": 0.12, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.5996793158738642, | |
| "grad_norm": 0.17945897579193115, | |
| "learning_rate": 2.2416627940271297e-05, | |
| "loss": 0.1285, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 0.6007482629609834, | |
| "grad_norm": 0.18568158149719238, | |
| "learning_rate": 2.2135846195024513e-05, | |
| "loss": 0.15, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 0.6018172100481026, | |
| "grad_norm": 0.1976700872182846, | |
| "learning_rate": 2.1856615019411797e-05, | |
| "loss": 0.1337, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 0.6028861571352218, | |
| "grad_norm": 0.20092736184597015, | |
| "learning_rate": 2.157893997398146e-05, | |
| "loss": 0.1435, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 0.603955104222341, | |
| "grad_norm": 0.16839931905269623, | |
| "learning_rate": 2.130282658829328e-05, | |
| "loss": 0.0981, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.6050240513094601, | |
| "grad_norm": 0.16764996945858002, | |
| "learning_rate": 2.1028280360808407e-05, | |
| "loss": 0.0963, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 0.6060929983965794, | |
| "grad_norm": 0.13916774094104767, | |
| "learning_rate": 2.075530675878016e-05, | |
| "loss": 0.0937, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 0.6071619454836985, | |
| "grad_norm": 0.1933029145002365, | |
| "learning_rate": 2.0483911218144715e-05, | |
| "loss": 0.1366, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 0.6082308925708177, | |
| "grad_norm": 0.15825247764587402, | |
| "learning_rate": 2.0214099143413212e-05, | |
| "loss": 0.0996, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 0.6092998396579369, | |
| "grad_norm": 0.1718830019235611, | |
| "learning_rate": 1.994587590756397e-05, | |
| "loss": 0.102, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.6103687867450561, | |
| "grad_norm": 0.18970075249671936, | |
| "learning_rate": 1.967924685193552e-05, | |
| "loss": 0.1178, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 0.6114377338321753, | |
| "grad_norm": 0.16414959728717804, | |
| "learning_rate": 1.941421728612023e-05, | |
| "loss": 0.1052, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 0.6125066809192945, | |
| "grad_norm": 0.1575910896062851, | |
| "learning_rate": 1.915079248785858e-05, | |
| "loss": 0.1141, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 0.6135756280064136, | |
| "grad_norm": 0.18138492107391357, | |
| "learning_rate": 1.8888977702934085e-05, | |
| "loss": 0.1054, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 0.6146445750935329, | |
| "grad_norm": 0.15937894582748413, | |
| "learning_rate": 1.862877814506878e-05, | |
| "loss": 0.104, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.615713522180652, | |
| "grad_norm": 0.2073594182729721, | |
| "learning_rate": 1.837019899581943e-05, | |
| "loss": 0.1216, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 0.6167824692677712, | |
| "grad_norm": 0.14365796744823456, | |
| "learning_rate": 1.8113245404474354e-05, | |
| "loss": 0.0961, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 0.6178514163548904, | |
| "grad_norm": 0.18027973175048828, | |
| "learning_rate": 1.7857922487950874e-05, | |
| "loss": 0.1319, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 0.6189203634420096, | |
| "grad_norm": 0.18247497081756592, | |
| "learning_rate": 1.760423533069332e-05, | |
| "loss": 0.1246, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 0.6199893105291288, | |
| "grad_norm": 0.21188899874687195, | |
| "learning_rate": 1.7352188984572026e-05, | |
| "loss": 0.1493, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.621058257616248, | |
| "grad_norm": 0.17170637845993042, | |
| "learning_rate": 1.710178846878251e-05, | |
| "loss": 0.0808, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 0.6221272047033671, | |
| "grad_norm": 0.13367417454719543, | |
| "learning_rate": 1.6853038769745467e-05, | |
| "loss": 0.0853, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 0.6231961517904864, | |
| "grad_norm": 0.15833276510238647, | |
| "learning_rate": 1.6605944841007802e-05, | |
| "loss": 0.1133, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 0.6242650988776055, | |
| "grad_norm": 0.14467398822307587, | |
| "learning_rate": 1.6360511603143648e-05, | |
| "loss": 0.0959, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 0.6253340459647247, | |
| "grad_norm": 0.15642966330051422, | |
| "learning_rate": 1.6116743943656488e-05, | |
| "loss": 0.1015, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.6264029930518439, | |
| "grad_norm": 0.18850955367088318, | |
| "learning_rate": 1.587464671688187e-05, | |
| "loss": 0.1258, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 0.6274719401389631, | |
| "grad_norm": 0.16557101905345917, | |
| "learning_rate": 1.56342247438908e-05, | |
| "loss": 0.119, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 0.6285408872260823, | |
| "grad_norm": 0.13995720446109772, | |
| "learning_rate": 1.5395482812393514e-05, | |
| "loss": 0.1172, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 0.6296098343132015, | |
| "grad_norm": 0.21847642958164215, | |
| "learning_rate": 1.5158425676644329e-05, | |
| "loss": 0.1504, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 0.6306787814003206, | |
| "grad_norm": 0.14314381778240204, | |
| "learning_rate": 1.4923058057346929e-05, | |
| "loss": 0.0987, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.6317477284874399, | |
| "grad_norm": 0.17072826623916626, | |
| "learning_rate": 1.4689384641560311e-05, | |
| "loss": 0.1182, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 0.632816675574559, | |
| "grad_norm": 0.18338800966739655, | |
| "learning_rate": 1.4457410082605483e-05, | |
| "loss": 0.1188, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 0.6338856226616782, | |
| "grad_norm": 0.13059154152870178, | |
| "learning_rate": 1.42271389999728e-05, | |
| "loss": 0.0729, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 0.6349545697487974, | |
| "grad_norm": 0.17279797792434692, | |
| "learning_rate": 1.3998575979229944e-05, | |
| "loss": 0.0968, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 0.6360235168359166, | |
| "grad_norm": 0.17134664952754974, | |
| "learning_rate": 1.3771725571930672e-05, | |
| "loss": 0.1149, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.6370924639230358, | |
| "grad_norm": 0.3472153842449188, | |
| "learning_rate": 1.3546592295524074e-05, | |
| "loss": 0.1259, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 0.638161411010155, | |
| "grad_norm": 0.1609698385000229, | |
| "learning_rate": 1.332318063326471e-05, | |
| "loss": 0.1091, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 0.6392303580972741, | |
| "grad_norm": 0.16347070038318634, | |
| "learning_rate": 1.3101495034123313e-05, | |
| "loss": 0.121, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 0.6402993051843934, | |
| "grad_norm": 0.14996081590652466, | |
| "learning_rate": 1.2881539912698082e-05, | |
| "loss": 0.0921, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 0.6413682522715125, | |
| "grad_norm": 0.15875977277755737, | |
| "learning_rate": 1.2663319649127026e-05, | |
| "loss": 0.1015, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.6413682522715125, | |
| "eval_loss": 0.11237961053848267, | |
| "eval_runtime": 288.8569, | |
| "eval_samples_per_second": 4.32, | |
| "eval_steps_per_second": 1.08, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.6424371993586317, | |
| "grad_norm": 0.19052091240882874, | |
| "learning_rate": 1.2446838589000397e-05, | |
| "loss": 0.1107, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 0.6435061464457509, | |
| "grad_norm": 0.16737191379070282, | |
| "learning_rate": 1.2232101043274436e-05, | |
| "loss": 0.1163, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 0.6445750935328701, | |
| "grad_norm": 0.14812670648097992, | |
| "learning_rate": 1.2019111288185437e-05, | |
| "loss": 0.0909, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 0.6456440406199893, | |
| "grad_norm": 0.143129363656044, | |
| "learning_rate": 1.1807873565164506e-05, | |
| "loss": 0.0955, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 0.6467129877071085, | |
| "grad_norm": 0.1692836433649063, | |
| "learning_rate": 1.1598392080753206e-05, | |
| "loss": 0.1016, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.6477819347942276, | |
| "grad_norm": 0.13897959887981415, | |
| "learning_rate": 1.139067100651976e-05, | |
| "loss": 0.0999, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 0.6488508818813469, | |
| "grad_norm": 0.18241965770721436, | |
| "learning_rate": 1.1184714478975934e-05, | |
| "loss": 0.1129, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 0.649919828968466, | |
| "grad_norm": 0.1324148029088974, | |
| "learning_rate": 1.0980526599494733e-05, | |
| "loss": 0.0854, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 0.6509887760555852, | |
| "grad_norm": 0.17263776063919067, | |
| "learning_rate": 1.0778111434228689e-05, | |
| "loss": 0.1312, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 0.6520577231427044, | |
| "grad_norm": 0.14975489675998688, | |
| "learning_rate": 1.057747301402887e-05, | |
| "loss": 0.1075, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.6531266702298236, | |
| "grad_norm": 0.13988271355628967, | |
| "learning_rate": 1.0378615334364694e-05, | |
| "loss": 0.0856, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 0.6541956173169428, | |
| "grad_norm": 0.2114977240562439, | |
| "learning_rate": 1.0181542355244167e-05, | |
| "loss": 0.1277, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 0.655264564404062, | |
| "grad_norm": 0.1697227954864502, | |
| "learning_rate": 9.98625800113534e-06, | |
| "loss": 0.1142, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 0.6563335114911811, | |
| "grad_norm": 0.17489199340343475, | |
| "learning_rate": 9.792766160887868e-06, | |
| "loss": 0.1246, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 0.6574024585783004, | |
| "grad_norm": 0.17671118676662445, | |
| "learning_rate": 9.601070687655667e-06, | |
| "loss": 0.1258, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.6584714056654195, | |
| "grad_norm": 0.15085704624652863, | |
| "learning_rate": 9.411175398820271e-06, | |
| "loss": 0.1179, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 0.6595403527525387, | |
| "grad_norm": 0.18436650931835175, | |
| "learning_rate": 9.223084075914712e-06, | |
| "loss": 0.1252, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 0.6606092998396579, | |
| "grad_norm": 0.16538633406162262, | |
| "learning_rate": 9.036800464548157e-06, | |
| "loss": 0.1066, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 0.6616782469267771, | |
| "grad_norm": 0.14048168063163757, | |
| "learning_rate": 8.852328274331545e-06, | |
| "loss": 0.0866, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 0.6627471940138963, | |
| "grad_norm": 0.22049780189990997, | |
| "learning_rate": 8.669671178803485e-06, | |
| "loss": 0.1782, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.6638161411010155, | |
| "grad_norm": 0.15474286675453186, | |
| "learning_rate": 8.488832815357173e-06, | |
| "loss": 0.1086, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 0.6648850881881346, | |
| "grad_norm": 0.14860907196998596, | |
| "learning_rate": 8.309816785168034e-06, | |
| "loss": 0.0887, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 0.6659540352752539, | |
| "grad_norm": 0.1767965406179428, | |
| "learning_rate": 8.132626653121912e-06, | |
| "loss": 0.1028, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 0.6670229823623731, | |
| "grad_norm": 0.1610213965177536, | |
| "learning_rate": 7.95726594774413e-06, | |
| "loss": 0.1179, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 0.6680919294494923, | |
| "grad_norm": 0.1768268495798111, | |
| "learning_rate": 7.783738161129206e-06, | |
| "loss": 0.1308, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.6691608765366115, | |
| "grad_norm": 0.1573677510023117, | |
| "learning_rate": 7.612046748871327e-06, | |
| "loss": 0.1099, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 0.6702298236237306, | |
| "grad_norm": 0.1720394343137741, | |
| "learning_rate": 7.44219512999551e-06, | |
| "loss": 0.1102, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 0.6712987707108499, | |
| "grad_norm": 0.13623464107513428, | |
| "learning_rate": 7.2741866868895395e-06, | |
| "loss": 0.0867, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 0.672367717797969, | |
| "grad_norm": 0.18556751310825348, | |
| "learning_rate": 7.108024765236588e-06, | |
| "loss": 0.0872, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 0.6734366648850882, | |
| "grad_norm": 0.1814941018819809, | |
| "learning_rate": 6.943712673948644e-06, | |
| "loss": 0.1179, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.6745056119722074, | |
| "grad_norm": 0.1475355327129364, | |
| "learning_rate": 6.78125368510053e-06, | |
| "loss": 0.0925, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 0.6755745590593266, | |
| "grad_norm": 0.15053223073482513, | |
| "learning_rate": 6.620651033864844e-06, | |
| "loss": 0.1029, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 0.6766435061464458, | |
| "grad_norm": 0.17404505610466003, | |
| "learning_rate": 6.461907918447463e-06, | |
| "loss": 0.1339, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 0.677712453233565, | |
| "grad_norm": 0.165951207280159, | |
| "learning_rate": 6.3050275000238414e-06, | |
| "loss": 0.12, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 0.6787814003206841, | |
| "grad_norm": 0.1419910490512848, | |
| "learning_rate": 6.150012902676139e-06, | |
| "loss": 0.0969, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.6798503474078034, | |
| "grad_norm": 0.17707513272762299, | |
| "learning_rate": 5.996867213330992e-06, | |
| "loss": 0.1325, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 0.6809192944949225, | |
| "grad_norm": 0.15850454568862915, | |
| "learning_rate": 5.8455934816979305e-06, | |
| "loss": 0.0944, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 0.6819882415820417, | |
| "grad_norm": 0.17115013301372528, | |
| "learning_rate": 5.696194720208792e-06, | |
| "loss": 0.1229, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 0.6830571886691609, | |
| "grad_norm": 0.1440781056880951, | |
| "learning_rate": 5.54867390395768e-06, | |
| "loss": 0.093, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 0.6841261357562801, | |
| "grad_norm": 0.18968208134174347, | |
| "learning_rate": 5.403033970641647e-06, | |
| "loss": 0.1386, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.6851950828433993, | |
| "grad_norm": 0.14848604798316956, | |
| "learning_rate": 5.259277820502306e-06, | |
| "loss": 0.0903, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 0.6862640299305185, | |
| "grad_norm": 0.16736701130867004, | |
| "learning_rate": 5.1174083162680465e-06, | |
| "loss": 0.1156, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 0.6873329770176376, | |
| "grad_norm": 0.15795965492725372, | |
| "learning_rate": 4.9774282830969296e-06, | |
| "loss": 0.1229, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 0.6884019241047569, | |
| "grad_norm": 0.16274042427539825, | |
| "learning_rate": 4.839340508520562e-06, | |
| "loss": 0.1149, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 0.689470871191876, | |
| "grad_norm": 0.14476804435253143, | |
| "learning_rate": 4.703147742388536e-06, | |
| "loss": 0.0914, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.6905398182789952, | |
| "grad_norm": 0.15442495048046112, | |
| "learning_rate": 4.568852696813619e-06, | |
| "loss": 0.0897, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 0.6916087653661144, | |
| "grad_norm": 0.15678077936172485, | |
| "learning_rate": 4.436458046117831e-06, | |
| "loss": 0.0956, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 0.6926777124532336, | |
| "grad_norm": 0.15000145137310028, | |
| "learning_rate": 4.305966426779118e-06, | |
| "loss": 0.1062, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 0.6937466595403528, | |
| "grad_norm": 0.1456059366464615, | |
| "learning_rate": 4.177380437378886e-06, | |
| "loss": 0.0894, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 0.694815606627472, | |
| "grad_norm": 0.14858050644397736, | |
| "learning_rate": 4.050702638550275e-06, | |
| "loss": 0.1073, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.6958845537145911, | |
| "grad_norm": 0.12687133252620697, | |
| "learning_rate": 3.9259355529270625e-06, | |
| "loss": 0.1037, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 0.6969535008017104, | |
| "grad_norm": 0.16253836452960968, | |
| "learning_rate": 3.8030816650935776e-06, | |
| "loss": 0.1209, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 0.6980224478888295, | |
| "grad_norm": 0.13301688432693481, | |
| "learning_rate": 3.682143421535156e-06, | |
| "loss": 0.094, | |
| "step": 653 | |
| }, | |
| { | |
| "epoch": 0.6990913949759487, | |
| "grad_norm": 0.15833823382854462, | |
| "learning_rate": 3.5631232305893046e-06, | |
| "loss": 0.1151, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 0.7001603420630679, | |
| "grad_norm": 0.1909790337085724, | |
| "learning_rate": 3.446023462398018e-06, | |
| "loss": 0.1443, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.7012292891501871, | |
| "grad_norm": 0.14769120514392853, | |
| "learning_rate": 3.330846448860259e-06, | |
| "loss": 0.1063, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 0.7022982362373063, | |
| "grad_norm": 0.14119353890419006, | |
| "learning_rate": 3.2175944835857617e-06, | |
| "loss": 0.0852, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 0.7033671833244255, | |
| "grad_norm": 0.16330832242965698, | |
| "learning_rate": 3.1062698218492724e-06, | |
| "loss": 0.1177, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 0.7044361304115446, | |
| "grad_norm": 0.15180334448814392, | |
| "learning_rate": 2.996874680545603e-06, | |
| "loss": 0.1117, | |
| "step": 659 | |
| }, | |
| { | |
| "epoch": 0.7055050774986639, | |
| "grad_norm": 0.13674339652061462, | |
| "learning_rate": 2.889411238145545e-06, | |
| "loss": 0.0722, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.706574024585783, | |
| "grad_norm": 0.15879607200622559, | |
| "learning_rate": 2.7838816346524634e-06, | |
| "loss": 0.0888, | |
| "step": 661 | |
| }, | |
| { | |
| "epoch": 0.7076429716729022, | |
| "grad_norm": 0.1543118804693222, | |
| "learning_rate": 2.6802879715596585e-06, | |
| "loss": 0.1114, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 0.7087119187600214, | |
| "grad_norm": 0.17156855762004852, | |
| "learning_rate": 2.5786323118085596e-06, | |
| "loss": 0.1223, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 0.7097808658471406, | |
| "grad_norm": 0.18760466575622559, | |
| "learning_rate": 2.4789166797476228e-06, | |
| "loss": 0.1395, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 0.7108498129342598, | |
| "grad_norm": 0.18010297417640686, | |
| "learning_rate": 2.381143061092006e-06, | |
| "loss": 0.1258, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.711918760021379, | |
| "grad_norm": 0.18845847249031067, | |
| "learning_rate": 2.2853134028840594e-06, | |
| "loss": 0.1218, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 0.7129877071084981, | |
| "grad_norm": 0.1728360801935196, | |
| "learning_rate": 2.191429613454499e-06, | |
| "loss": 0.1337, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 0.7140566541956174, | |
| "grad_norm": 0.11702633649110794, | |
| "learning_rate": 2.0994935623844692e-06, | |
| "loss": 0.084, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 0.7151256012827365, | |
| "grad_norm": 0.14172986149787903, | |
| "learning_rate": 2.009507080468298e-06, | |
| "loss": 0.0913, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 0.7161945483698557, | |
| "grad_norm": 0.17234764993190765, | |
| "learning_rate": 1.921471959676957e-06, | |
| "loss": 0.12, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.7172634954569749, | |
| "grad_norm": 0.17920809984207153, | |
| "learning_rate": 1.8353899531224906e-06, | |
| "loss": 0.1267, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 0.7183324425440941, | |
| "grad_norm": 0.14638297259807587, | |
| "learning_rate": 1.751262775023077e-06, | |
| "loss": 0.111, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 0.7194013896312133, | |
| "grad_norm": 0.13949543237686157, | |
| "learning_rate": 1.6690921006687899e-06, | |
| "loss": 0.0887, | |
| "step": 673 | |
| }, | |
| { | |
| "epoch": 0.7204703367183325, | |
| "grad_norm": 0.19536994397640228, | |
| "learning_rate": 1.5888795663883904e-06, | |
| "loss": 0.1467, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 0.7215392838054516, | |
| "grad_norm": 0.15846571326255798, | |
| "learning_rate": 1.5106267695166543e-06, | |
| "loss": 0.1128, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.7226082308925709, | |
| "grad_norm": 0.13756738603115082, | |
| "learning_rate": 1.4343352683625411e-06, | |
| "loss": 0.0804, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 0.72367717797969, | |
| "grad_norm": 0.1360546052455902, | |
| "learning_rate": 1.3600065821782526e-06, | |
| "loss": 0.0979, | |
| "step": 677 | |
| }, | |
| { | |
| "epoch": 0.7247461250668092, | |
| "grad_norm": 0.15744271874427795, | |
| "learning_rate": 1.2876421911288905e-06, | |
| "loss": 0.1013, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 0.7258150721539284, | |
| "grad_norm": 0.13687777519226074, | |
| "learning_rate": 1.217243536263013e-06, | |
| "loss": 0.0861, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 0.7268840192410476, | |
| "grad_norm": 0.1810121387243271, | |
| "learning_rate": 1.148812019483958e-06, | |
| "loss": 0.1093, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.7279529663281668, | |
| "grad_norm": 0.1584986299276352, | |
| "learning_rate": 1.0823490035218987e-06, | |
| "loss": 0.1093, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 0.729021913415286, | |
| "grad_norm": 0.15906885266304016, | |
| "learning_rate": 1.0178558119067315e-06, | |
| "loss": 0.1173, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 0.7300908605024051, | |
| "grad_norm": 0.18237455189228058, | |
| "learning_rate": 9.553337289416653e-07, | |
| "loss": 0.1355, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 0.7311598075895244, | |
| "grad_norm": 0.12682956457138062, | |
| "learning_rate": 8.947839996777285e-07, | |
| "loss": 0.0717, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 0.7322287546766435, | |
| "grad_norm": 0.160029336810112, | |
| "learning_rate": 8.362078298889021e-07, | |
| "loss": 0.1009, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.7332977017637627, | |
| "grad_norm": 0.15527096390724182, | |
| "learning_rate": 7.796063860481595e-07, | |
| "loss": 0.115, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 0.7343666488508819, | |
| "grad_norm": 0.15923753380775452, | |
| "learning_rate": 7.249807953041865e-07, | |
| "loss": 0.0966, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 0.7354355959380011, | |
| "grad_norm": 0.17644928395748138, | |
| "learning_rate": 6.723321454590092e-07, | |
| "loss": 0.1159, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 0.7365045430251203, | |
| "grad_norm": 0.1382478028535843, | |
| "learning_rate": 6.216614849462343e-07, | |
| "loss": 0.1078, | |
| "step": 689 | |
| }, | |
| { | |
| "epoch": 0.7375734901122395, | |
| "grad_norm": 0.17270506918430328, | |
| "learning_rate": 5.729698228102653e-07, | |
| "loss": 0.0972, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.7386424371993586, | |
| "grad_norm": 0.14153295755386353, | |
| "learning_rate": 5.262581286861634e-07, | |
| "loss": 0.0839, | |
| "step": 691 | |
| }, | |
| { | |
| "epoch": 0.7397113842864779, | |
| "grad_norm": 0.13959170877933502, | |
| "learning_rate": 4.815273327803182e-07, | |
| "loss": 0.0961, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 0.740780331373597, | |
| "grad_norm": 0.17043890058994293, | |
| "learning_rate": 4.387783258519518e-07, | |
| "loss": 0.1042, | |
| "step": 693 | |
| }, | |
| { | |
| "epoch": 0.7418492784607162, | |
| "grad_norm": 0.1593395173549652, | |
| "learning_rate": 3.9801195919541014e-07, | |
| "loss": 0.1103, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 0.7429182255478354, | |
| "grad_norm": 0.14094416797161102, | |
| "learning_rate": 3.5922904462314387e-07, | |
| "loss": 0.1123, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 0.7439871726349546, | |
| "grad_norm": 0.15978014469146729, | |
| "learning_rate": 3.224303544495766e-07, | |
| "loss": 0.1089, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 0.7450561197220738, | |
| "grad_norm": 0.14494763314723969, | |
| "learning_rate": 2.876166214757392e-07, | |
| "loss": 0.0909, | |
| "step": 697 | |
| }, | |
| { | |
| "epoch": 0.746125066809193, | |
| "grad_norm": 0.14322949945926666, | |
| "learning_rate": 2.547885389746485e-07, | |
| "loss": 0.0928, | |
| "step": 698 | |
| }, | |
| { | |
| "epoch": 0.7471940138963121, | |
| "grad_norm": 0.13338381052017212, | |
| "learning_rate": 2.2394676067752918e-07, | |
| "loss": 0.0905, | |
| "step": 699 | |
| }, | |
| { | |
| "epoch": 0.7482629609834314, | |
| "grad_norm": 0.1393570601940155, | |
| "learning_rate": 1.9509190076074657e-07, | |
| "loss": 0.1039, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.7482629609834314, | |
| "eval_loss": 0.11092506349086761, | |
| "eval_runtime": 289.0283, | |
| "eval_samples_per_second": 4.318, | |
| "eval_steps_per_second": 1.079, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.7493319080705505, | |
| "grad_norm": 0.16759201884269714, | |
| "learning_rate": 1.6822453383367186e-07, | |
| "loss": 0.1136, | |
| "step": 701 | |
| }, | |
| { | |
| "epoch": 0.7504008551576697, | |
| "grad_norm": 0.149668887257576, | |
| "learning_rate": 1.4334519492711362e-07, | |
| "loss": 0.1026, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 0.7514698022447889, | |
| "grad_norm": 0.14984771609306335, | |
| "learning_rate": 1.204543794827595e-07, | |
| "loss": 0.0942, | |
| "step": 703 | |
| }, | |
| { | |
| "epoch": 0.7525387493319081, | |
| "grad_norm": 0.15921737253665924, | |
| "learning_rate": 9.955254334328423e-08, | |
| "loss": 0.1044, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 0.7536076964190273, | |
| "grad_norm": 0.14484407007694244, | |
| "learning_rate": 8.064010274324573e-08, | |
| "loss": 0.0849, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.7546766435061465, | |
| "grad_norm": 0.1964549571275711, | |
| "learning_rate": 6.37174343008251e-08, | |
| "loss": 0.125, | |
| "step": 706 | |
| }, | |
| { | |
| "epoch": 0.7557455905932656, | |
| "grad_norm": 0.1271902620792389, | |
| "learning_rate": 4.878487501033258e-08, | |
| "loss": 0.0783, | |
| "step": 707 | |
| }, | |
| { | |
| "epoch": 0.7568145376803849, | |
| "grad_norm": 0.155805766582489, | |
| "learning_rate": 3.584272223546847e-08, | |
| "loss": 0.0972, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 0.757883484767504, | |
| "grad_norm": 0.15836969017982483, | |
| "learning_rate": 2.4891233703394634e-08, | |
| "loss": 0.095, | |
| "step": 709 | |
| }, | |
| { | |
| "epoch": 0.7589524318546232, | |
| "grad_norm": 0.1573915183544159, | |
| "learning_rate": 1.593062749967178e-08, | |
| "loss": 0.1093, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.7600213789417424, | |
| "grad_norm": 0.15478433668613434, | |
| "learning_rate": 8.961082063829729e-09, | |
| "loss": 0.0977, | |
| "step": 711 | |
| }, | |
| { | |
| "epoch": 0.7610903260288616, | |
| "grad_norm": 0.1593605875968933, | |
| "learning_rate": 3.982736185859093e-09, | |
| "loss": 0.1136, | |
| "step": 712 | |
| }, | |
| { | |
| "epoch": 0.7621592731159808, | |
| "grad_norm": 0.1565905064344406, | |
| "learning_rate": 9.956890034468202e-10, | |
| "loss": 0.0978, | |
| "step": 713 | |
| }, | |
| { | |
| "epoch": 0.7632282202031, | |
| "grad_norm": 0.1467771828174591, | |
| "learning_rate": 0.0, | |
| "loss": 0.1067, | |
| "step": 714 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 714, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 2, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 3.6866682847966003e+18, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |