| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.5006075334143378, | |
| "eval_steps": 103, | |
| "global_step": 618, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.002430133657351154, | |
| "grad_norm": 715.4923219036787, | |
| "learning_rate": 0.0, | |
| "loss": 1.3541, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.002430133657351154, | |
| "eval_loss": 1.3335719108581543, | |
| "eval_runtime": 53.4883, | |
| "eval_samples_per_second": 13.91, | |
| "eval_steps_per_second": 1.739, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.004860267314702308, | |
| "grad_norm": 614.6970578314867, | |
| "learning_rate": 5e-06, | |
| "loss": 1.3775, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.007290400972053463, | |
| "grad_norm": 471.59017991123795, | |
| "learning_rate": 1e-05, | |
| "loss": 1.339, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.009720534629404616, | |
| "grad_norm": 238.72216262259653, | |
| "learning_rate": 1.5e-05, | |
| "loss": 1.3829, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.012150668286755772, | |
| "grad_norm": 355.68955726709873, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3597, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.014580801944106925, | |
| "grad_norm": 414.5627284272111, | |
| "learning_rate": 2.5e-05, | |
| "loss": 1.3862, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.01701093560145808, | |
| "grad_norm": 534.9877222052693, | |
| "learning_rate": 3e-05, | |
| "loss": 1.2784, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.019441069258809233, | |
| "grad_norm": 153.38895635666677, | |
| "learning_rate": 3.5e-05, | |
| "loss": 1.3521, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.02187120291616039, | |
| "grad_norm": 858.293734138087, | |
| "learning_rate": 4e-05, | |
| "loss": 1.2461, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.024301336573511544, | |
| "grad_norm": 255.81989388533376, | |
| "learning_rate": 4.5e-05, | |
| "loss": 1.2778, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.026731470230862697, | |
| "grad_norm": 368.91949003479226, | |
| "learning_rate": 5e-05, | |
| "loss": 1.3412, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.02916160388821385, | |
| "grad_norm": 176.49481799555898, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 1.3437, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.031591737545565005, | |
| "grad_norm": 208.57742104974147, | |
| "learning_rate": 6e-05, | |
| "loss": 1.2859, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.03402187120291616, | |
| "grad_norm": 93.26742036471734, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 1.1843, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.03645200486026731, | |
| "grad_norm": 145.53380444622215, | |
| "learning_rate": 7e-05, | |
| "loss": 1.4281, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.038882138517618466, | |
| "grad_norm": 126.56724937430516, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 1.3908, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.041312272174969626, | |
| "grad_norm": 106.19246390662754, | |
| "learning_rate": 8e-05, | |
| "loss": 1.344, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.04374240583232078, | |
| "grad_norm": 289.348178084847, | |
| "learning_rate": 8.5e-05, | |
| "loss": 1.2708, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.046172539489671933, | |
| "grad_norm": 286.63676887065634, | |
| "learning_rate": 9e-05, | |
| "loss": 1.3564, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.04860267314702309, | |
| "grad_norm": 269.6096299101413, | |
| "learning_rate": 9.5e-05, | |
| "loss": 1.2184, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.05103280680437424, | |
| "grad_norm": 151.28678796160915, | |
| "learning_rate": 0.0001, | |
| "loss": 1.2974, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.053462940461725394, | |
| "grad_norm": 265.5625538646362, | |
| "learning_rate": 0.000105, | |
| "loss": 1.2703, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.05589307411907655, | |
| "grad_norm": 724.7157187586193, | |
| "learning_rate": 0.00011000000000000002, | |
| "loss": 1.2691, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.0583232077764277, | |
| "grad_norm": 425.3768239347252, | |
| "learning_rate": 0.00011499999999999999, | |
| "loss": 1.375, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.060753341433778855, | |
| "grad_norm": 314.5119318308783, | |
| "learning_rate": 0.00012, | |
| "loss": 1.2952, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.06318347509113001, | |
| "grad_norm": 557.519173033834, | |
| "learning_rate": 0.000125, | |
| "loss": 1.2923, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.06561360874848117, | |
| "grad_norm": 211.4069356529637, | |
| "learning_rate": 0.00013000000000000002, | |
| "loss": 1.2629, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.06804374240583232, | |
| "grad_norm": 299.7742653722713, | |
| "learning_rate": 0.00013500000000000003, | |
| "loss": 1.3099, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.07047387606318348, | |
| "grad_norm": 182.18551965886013, | |
| "learning_rate": 0.00014, | |
| "loss": 1.2215, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.07290400972053462, | |
| "grad_norm": 153.38300520125887, | |
| "learning_rate": 0.000145, | |
| "loss": 1.2799, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07533414337788578, | |
| "grad_norm": 849.4472853252786, | |
| "learning_rate": 0.00015000000000000001, | |
| "loss": 1.2012, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.07776427703523693, | |
| "grad_norm": 179.94814586965418, | |
| "learning_rate": 0.000155, | |
| "loss": 1.2103, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.08019441069258809, | |
| "grad_norm": 180.36681057956048, | |
| "learning_rate": 0.00016, | |
| "loss": 1.2414, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.08262454434993925, | |
| "grad_norm": 113.72852454032189, | |
| "learning_rate": 0.000165, | |
| "loss": 1.2508, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.0850546780072904, | |
| "grad_norm": 150.53415363213057, | |
| "learning_rate": 0.00017, | |
| "loss": 1.2528, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.08748481166464156, | |
| "grad_norm": 156.19567878683574, | |
| "learning_rate": 0.000175, | |
| "loss": 1.2016, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.0899149453219927, | |
| "grad_norm": 416.34884765145057, | |
| "learning_rate": 0.00018, | |
| "loss": 1.254, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.09234507897934387, | |
| "grad_norm": 269.7105025581372, | |
| "learning_rate": 0.00018500000000000002, | |
| "loss": 1.2215, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.09477521263669501, | |
| "grad_norm": 249.35069047655023, | |
| "learning_rate": 0.00019, | |
| "loss": 1.2078, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.09720534629404617, | |
| "grad_norm": 167.16896045613478, | |
| "learning_rate": 0.000195, | |
| "loss": 1.1866, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.09963547995139732, | |
| "grad_norm": 248.22240554128427, | |
| "learning_rate": 0.0002, | |
| "loss": 1.252, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.10206561360874848, | |
| "grad_norm": 180.89520841022969, | |
| "learning_rate": 0.0001999991930332148, | |
| "loss": 1.2251, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.10449574726609964, | |
| "grad_norm": 614.4291375430485, | |
| "learning_rate": 0.00019999677214588312, | |
| "loss": 1.2563, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.10692588092345079, | |
| "grad_norm": 211.7523427355369, | |
| "learning_rate": 0.00019999273737707646, | |
| "loss": 1.193, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.10935601458080195, | |
| "grad_norm": 181.56788458769344, | |
| "learning_rate": 0.00019998708879191335, | |
| "loss": 1.2598, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.1117861482381531, | |
| "grad_norm": 157.5783414916277, | |
| "learning_rate": 0.00019997982648155814, | |
| "loss": 1.2663, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.11421628189550426, | |
| "grad_norm": 155.78006251192625, | |
| "learning_rate": 0.00019997095056321971, | |
| "loss": 1.1637, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.1166464155528554, | |
| "grad_norm": 202.0253360488958, | |
| "learning_rate": 0.00019996046118014955, | |
| "loss": 1.2508, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.11907654921020656, | |
| "grad_norm": 192.7576297264874, | |
| "learning_rate": 0.00019994835850163924, | |
| "loss": 1.2014, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.12150668286755771, | |
| "grad_norm": 132.5484871621418, | |
| "learning_rate": 0.00019993464272301804, | |
| "loss": 1.2279, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.12393681652490887, | |
| "grad_norm": 128.32285438248965, | |
| "learning_rate": 0.00019991931406564944, | |
| "loss": 1.2179, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.12636695018226002, | |
| "grad_norm": 552.3669463716512, | |
| "learning_rate": 0.00019990237277692788, | |
| "loss": 1.1498, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.12879708383961117, | |
| "grad_norm": 86.17911790260192, | |
| "learning_rate": 0.00019988381913027442, | |
| "loss": 1.2784, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.13122721749696234, | |
| "grad_norm": 70.83294605515782, | |
| "learning_rate": 0.00019986365342513265, | |
| "loss": 1.2224, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.1336573511543135, | |
| "grad_norm": 45.23624563299466, | |
| "learning_rate": 0.00019984187598696363, | |
| "loss": 1.1746, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.13608748481166463, | |
| "grad_norm": 57.67645735585192, | |
| "learning_rate": 0.00019981848716724073, | |
| "loss": 1.2154, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.1385176184690158, | |
| "grad_norm": 45.661268047129674, | |
| "learning_rate": 0.00019979348734344398, | |
| "loss": 1.1411, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.14094775212636695, | |
| "grad_norm": 53.10628399970359, | |
| "learning_rate": 0.00019976687691905393, | |
| "loss": 1.2029, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.1433778857837181, | |
| "grad_norm": 38.71353325803162, | |
| "learning_rate": 0.00019973865632354516, | |
| "loss": 1.1976, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.14580801944106925, | |
| "grad_norm": 42.789208063581114, | |
| "learning_rate": 0.0001997088260123793, | |
| "loss": 1.1477, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.14823815309842042, | |
| "grad_norm": 37.613194740192164, | |
| "learning_rate": 0.0001996773864669978, | |
| "loss": 1.2529, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.15066828675577157, | |
| "grad_norm": 47.96813084127655, | |
| "learning_rate": 0.00019964433819481405, | |
| "loss": 1.2328, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.15309842041312272, | |
| "grad_norm": 55.30483872428545, | |
| "learning_rate": 0.00019960968172920516, | |
| "loss": 1.1996, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.15552855407047386, | |
| "grad_norm": 35.58995799070749, | |
| "learning_rate": 0.00019957341762950344, | |
| "loss": 1.1248, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.15795868772782504, | |
| "grad_norm": 58.86131222300149, | |
| "learning_rate": 0.00019953554648098748, | |
| "loss": 1.3017, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.16038882138517618, | |
| "grad_norm": 32.12091331878439, | |
| "learning_rate": 0.00019949606889487233, | |
| "loss": 1.1961, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.16281895504252733, | |
| "grad_norm": 167.27433996357928, | |
| "learning_rate": 0.0001994549855083001, | |
| "loss": 1.1768, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.1652490886998785, | |
| "grad_norm": 32.3328494297432, | |
| "learning_rate": 0.0001994122969843293, | |
| "loss": 1.1802, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.16767922235722965, | |
| "grad_norm": 39.92530074438497, | |
| "learning_rate": 0.0001993680040119244, | |
| "loss": 1.2098, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.1701093560145808, | |
| "grad_norm": 45.60830517129956, | |
| "learning_rate": 0.0001993221073059445, | |
| "loss": 1.2159, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.17253948967193194, | |
| "grad_norm": 35.462695032736335, | |
| "learning_rate": 0.00019927460760713197, | |
| "loss": 1.1818, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.17496962332928312, | |
| "grad_norm": 43.05751624597826, | |
| "learning_rate": 0.0001992255056821004, | |
| "loss": 1.2011, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.17739975698663427, | |
| "grad_norm": 47.13143404969894, | |
| "learning_rate": 0.00019917480232332224, | |
| "loss": 1.1669, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.1798298906439854, | |
| "grad_norm": 72.07146401418987, | |
| "learning_rate": 0.000199122498349116, | |
| "loss": 1.181, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.1822600243013366, | |
| "grad_norm": 36.289202348834955, | |
| "learning_rate": 0.00019906859460363307, | |
| "loss": 1.1787, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.18469015795868773, | |
| "grad_norm": 46.92636167228936, | |
| "learning_rate": 0.00019901309195684416, | |
| "loss": 1.2316, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.18712029161603888, | |
| "grad_norm": 31.71425340357504, | |
| "learning_rate": 0.00019895599130452505, | |
| "loss": 1.1607, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.18955042527339003, | |
| "grad_norm": 43.94199928621344, | |
| "learning_rate": 0.00019889729356824235, | |
| "loss": 1.1919, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.1919805589307412, | |
| "grad_norm": 45.33073791860179, | |
| "learning_rate": 0.0001988369996953386, | |
| "loss": 1.2237, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.19441069258809235, | |
| "grad_norm": 135.89980489661897, | |
| "learning_rate": 0.00019877511065891673, | |
| "loss": 1.1822, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.1968408262454435, | |
| "grad_norm": 439.6770852212966, | |
| "learning_rate": 0.00019871162745782478, | |
| "loss": 1.1441, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.19927095990279464, | |
| "grad_norm": 80.73319798776026, | |
| "learning_rate": 0.0001986465511166394, | |
| "loss": 1.1709, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.20170109356014582, | |
| "grad_norm": 87.76515297497458, | |
| "learning_rate": 0.00019857988268564953, | |
| "loss": 1.1549, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.20413122721749696, | |
| "grad_norm": 70.08754986406095, | |
| "learning_rate": 0.00019851162324083932, | |
| "loss": 1.1771, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.2065613608748481, | |
| "grad_norm": 187.8198997057664, | |
| "learning_rate": 0.0001984417738838709, | |
| "loss": 1.2068, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.20899149453219928, | |
| "grad_norm": 127.78818684755072, | |
| "learning_rate": 0.00019837033574206646, | |
| "loss": 1.1974, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.21142162818955043, | |
| "grad_norm": 127.82979216871074, | |
| "learning_rate": 0.0001982973099683902, | |
| "loss": 1.185, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.21385176184690158, | |
| "grad_norm": 142.35425084857746, | |
| "learning_rate": 0.00019822269774142954, | |
| "loss": 1.2225, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.21628189550425272, | |
| "grad_norm": 246.64019353564817, | |
| "learning_rate": 0.0001981465002653763, | |
| "loss": 1.2574, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.2187120291616039, | |
| "grad_norm": 189.88471076285524, | |
| "learning_rate": 0.0001980687187700071, | |
| "loss": 1.1635, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.22114216281895505, | |
| "grad_norm": 116.65693373141701, | |
| "learning_rate": 0.00019798935451066361, | |
| "loss": 1.1457, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.2235722964763062, | |
| "grad_norm": 71.76422539970217, | |
| "learning_rate": 0.00019790840876823232, | |
| "loss": 1.2354, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.22600243013365734, | |
| "grad_norm": 139.42330509386431, | |
| "learning_rate": 0.0001978258828491236, | |
| "loss": 1.18, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.2284325637910085, | |
| "grad_norm": 131.88308820601443, | |
| "learning_rate": 0.00019774177808525113, | |
| "loss": 1.1868, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.23086269744835966, | |
| "grad_norm": 85.81071125615291, | |
| "learning_rate": 0.00019765609583400977, | |
| "loss": 1.1814, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.2332928311057108, | |
| "grad_norm": 84.43756298541064, | |
| "learning_rate": 0.00019756883747825424, | |
| "loss": 1.1658, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.23572296476306198, | |
| "grad_norm": 114.24245545143974, | |
| "learning_rate": 0.0001974800044262764, | |
| "loss": 1.2497, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.23815309842041313, | |
| "grad_norm": 76.577511222722, | |
| "learning_rate": 0.00019738959811178272, | |
| "loss": 1.1414, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.24058323207776428, | |
| "grad_norm": 171.8084830895381, | |
| "learning_rate": 0.00019729761999387103, | |
| "loss": 1.1619, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.24301336573511542, | |
| "grad_norm": 221.87752250936416, | |
| "learning_rate": 0.00019720407155700707, | |
| "loss": 1.2718, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.2454434993924666, | |
| "grad_norm": 205.64943975370608, | |
| "learning_rate": 0.00019710895431100046, | |
| "loss": 1.1786, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.24787363304981774, | |
| "grad_norm": 160.16582903260615, | |
| "learning_rate": 0.00019701226979098037, | |
| "loss": 1.1426, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.2503037667071689, | |
| "grad_norm": 82.85031394537334, | |
| "learning_rate": 0.00019691401955737072, | |
| "loss": 1.1718, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.2503037667071689, | |
| "eval_loss": 1.1633374691009521, | |
| "eval_runtime": 52.6182, | |
| "eval_samples_per_second": 14.14, | |
| "eval_steps_per_second": 1.767, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.25273390036452004, | |
| "grad_norm": 94.74469296109082, | |
| "learning_rate": 0.000196814205195865, | |
| "loss": 1.2255, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.2551640340218712, | |
| "grad_norm": 126.15797466756656, | |
| "learning_rate": 0.00019671282831740076, | |
| "loss": 1.1623, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.25759416767922233, | |
| "grad_norm": 79.41156434272008, | |
| "learning_rate": 0.0001966098905581334, | |
| "loss": 1.1606, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.2600243013365735, | |
| "grad_norm": 70.33104031058372, | |
| "learning_rate": 0.00019650539357941003, | |
| "loss": 1.196, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.2624544349939247, | |
| "grad_norm": 69.57260733822498, | |
| "learning_rate": 0.0001963993390677424, | |
| "loss": 1.1939, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.2648845686512758, | |
| "grad_norm": 81.78820691772725, | |
| "learning_rate": 0.00019629172873477995, | |
| "loss": 1.2553, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.267314702308627, | |
| "grad_norm": 117.06324110268656, | |
| "learning_rate": 0.00019618256431728194, | |
| "loss": 1.2535, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.26974483596597815, | |
| "grad_norm": 83.26993317104247, | |
| "learning_rate": 0.00019607184757708951, | |
| "loss": 1.157, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.27217496962332927, | |
| "grad_norm": 51.990829456422375, | |
| "learning_rate": 0.00019595958030109735, | |
| "loss": 1.1274, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.27460510328068044, | |
| "grad_norm": 119.7487160875729, | |
| "learning_rate": 0.00019584576430122473, | |
| "loss": 1.1422, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.2770352369380316, | |
| "grad_norm": 88.15636932272304, | |
| "learning_rate": 0.00019573040141438624, | |
| "loss": 1.1599, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.27946537059538273, | |
| "grad_norm": 62.346402225534774, | |
| "learning_rate": 0.00019561349350246226, | |
| "loss": 1.1909, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.2818955042527339, | |
| "grad_norm": 76.40612150653034, | |
| "learning_rate": 0.0001954950424522688, | |
| "loss": 1.1646, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.284325637910085, | |
| "grad_norm": 94.8711613055073, | |
| "learning_rate": 0.00019537505017552716, | |
| "loss": 1.1547, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.2867557715674362, | |
| "grad_norm": 63.86961661796314, | |
| "learning_rate": 0.00019525351860883293, | |
| "loss": 1.1841, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.2891859052247874, | |
| "grad_norm": 133.2417924150684, | |
| "learning_rate": 0.00019513044971362494, | |
| "loss": 1.1365, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.2916160388821385, | |
| "grad_norm": 133.44891510996445, | |
| "learning_rate": 0.00019500584547615333, | |
| "loss": 1.1696, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.29404617253948967, | |
| "grad_norm": 58.51701768739601, | |
| "learning_rate": 0.00019487970790744774, | |
| "loss": 1.1874, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.29647630619684084, | |
| "grad_norm": 49.536158238056196, | |
| "learning_rate": 0.00019475203904328474, | |
| "loss": 1.1798, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.29890643985419196, | |
| "grad_norm": 94.27608706983857, | |
| "learning_rate": 0.000194622840944155, | |
| "loss": 1.2443, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.30133657351154314, | |
| "grad_norm": 103.868243202843, | |
| "learning_rate": 0.00019449211569523, | |
| "loss": 1.1759, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.3037667071688943, | |
| "grad_norm": 73.31536435980003, | |
| "learning_rate": 0.00019435986540632843, | |
| "loss": 1.1885, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.30619684082624543, | |
| "grad_norm": 64.91149114745738, | |
| "learning_rate": 0.00019422609221188207, | |
| "loss": 1.1864, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.3086269744835966, | |
| "grad_norm": 95.34449184763653, | |
| "learning_rate": 0.00019409079827090145, | |
| "loss": 1.1339, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.3110571081409477, | |
| "grad_norm": 67.36156159754226, | |
| "learning_rate": 0.00019395398576694086, | |
| "loss": 1.1845, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.3134872417982989, | |
| "grad_norm": 36.94913176821407, | |
| "learning_rate": 0.00019381565690806328, | |
| "loss": 1.2154, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.3159173754556501, | |
| "grad_norm": 69.05265214547647, | |
| "learning_rate": 0.00019367581392680457, | |
| "loss": 1.1642, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.3183475091130012, | |
| "grad_norm": 38.974761165559855, | |
| "learning_rate": 0.00019353445908013755, | |
| "loss": 1.1508, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.32077764277035237, | |
| "grad_norm": 48.47215142199794, | |
| "learning_rate": 0.00019339159464943557, | |
| "loss": 1.2011, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.32320777642770354, | |
| "grad_norm": 41.88512063342574, | |
| "learning_rate": 0.00019324722294043558, | |
| "loss": 1.1643, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.32563791008505466, | |
| "grad_norm": 25.59403215229145, | |
| "learning_rate": 0.00019310134628320114, | |
| "loss": 1.1954, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.32806804374240583, | |
| "grad_norm": 58.02634988046396, | |
| "learning_rate": 0.00019295396703208453, | |
| "loss": 1.1544, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.330498177399757, | |
| "grad_norm": 31.26218977398251, | |
| "learning_rate": 0.00019280508756568896, | |
| "loss": 1.1613, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.33292831105710813, | |
| "grad_norm": 31.81234539284103, | |
| "learning_rate": 0.00019265471028683014, | |
| "loss": 1.1892, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.3353584447144593, | |
| "grad_norm": 54.44930114675527, | |
| "learning_rate": 0.00019250283762249748, | |
| "loss": 1.2801, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.3377885783718105, | |
| "grad_norm": 30.320486287732734, | |
| "learning_rate": 0.00019234947202381486, | |
| "loss": 1.1934, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.3402187120291616, | |
| "grad_norm": 32.76175001943503, | |
| "learning_rate": 0.00019219461596600113, | |
| "loss": 1.1436, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.34264884568651277, | |
| "grad_norm": 36.802264122697316, | |
| "learning_rate": 0.00019203827194833026, | |
| "loss": 1.1418, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.3450789793438639, | |
| "grad_norm": 35.03898729580271, | |
| "learning_rate": 0.0001918804424940908, | |
| "loss": 1.2479, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.34750911300121506, | |
| "grad_norm": 89.58068030461165, | |
| "learning_rate": 0.00019172113015054532, | |
| "loss": 1.2504, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.34993924665856624, | |
| "grad_norm": 30.05799668441019, | |
| "learning_rate": 0.00019156033748888917, | |
| "loss": 1.1662, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.35236938031591736, | |
| "grad_norm": 33.80121199203598, | |
| "learning_rate": 0.00019139806710420914, | |
| "loss": 1.1862, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.35479951397326853, | |
| "grad_norm": 31.510896023067872, | |
| "learning_rate": 0.00019123432161544142, | |
| "loss": 1.147, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.3572296476306197, | |
| "grad_norm": 32.92613286618093, | |
| "learning_rate": 0.00019106910366532942, | |
| "loss": 1.1421, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.3596597812879708, | |
| "grad_norm": 245.36013493823395, | |
| "learning_rate": 0.00019090241592038113, | |
| "loss": 1.1306, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.362089914945322, | |
| "grad_norm": 72.3061625644275, | |
| "learning_rate": 0.000190734261070826, | |
| "loss": 1.1144, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.3645200486026732, | |
| "grad_norm": 63.77748866336388, | |
| "learning_rate": 0.00019056464183057157, | |
| "loss": 1.1249, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.3669501822600243, | |
| "grad_norm": 633.2421324308109, | |
| "learning_rate": 0.00019039356093715975, | |
| "loss": 1.1359, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.36938031591737547, | |
| "grad_norm": 34.456657555313704, | |
| "learning_rate": 0.00019022102115172248, | |
| "loss": 1.1397, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.3718104495747266, | |
| "grad_norm": 35.21328820959324, | |
| "learning_rate": 0.00019004702525893732, | |
| "loss": 1.1741, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.37424058323207776, | |
| "grad_norm": 90.32405227187036, | |
| "learning_rate": 0.00018987157606698235, | |
| "loss": 1.1844, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.37667071688942894, | |
| "grad_norm": 39.348755664527914, | |
| "learning_rate": 0.000189694676407491, | |
| "loss": 1.1216, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.37910085054678005, | |
| "grad_norm": 58.85540744859834, | |
| "learning_rate": 0.00018951632913550626, | |
| "loss": 1.115, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.38153098420413123, | |
| "grad_norm": 39.849945227365325, | |
| "learning_rate": 0.0001893365371294346, | |
| "loss": 1.1705, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.3839611178614824, | |
| "grad_norm": 40.300954908722304, | |
| "learning_rate": 0.0001891553032909996, | |
| "loss": 1.1831, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.3863912515188335, | |
| "grad_norm": 53.72009888405355, | |
| "learning_rate": 0.00018897263054519498, | |
| "loss": 1.1613, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.3888213851761847, | |
| "grad_norm": 142.22686975859034, | |
| "learning_rate": 0.0001887885218402375, | |
| "loss": 1.1639, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.39125151883353587, | |
| "grad_norm": 50.141889086717356, | |
| "learning_rate": 0.00018860298014751944, | |
| "loss": 1.1659, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.393681652490887, | |
| "grad_norm": 63.25519968311113, | |
| "learning_rate": 0.0001884160084615604, | |
| "loss": 1.168, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.39611178614823817, | |
| "grad_norm": 50.59325246324073, | |
| "learning_rate": 0.0001882276097999592, | |
| "loss": 1.1202, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.3985419198055893, | |
| "grad_norm": 58.32587879810431, | |
| "learning_rate": 0.0001880377872033451, | |
| "loss": 1.1587, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.40097205346294046, | |
| "grad_norm": 211.50882688314653, | |
| "learning_rate": 0.00018784654373532866, | |
| "loss": 1.1551, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.40340218712029163, | |
| "grad_norm": 47.82888424614203, | |
| "learning_rate": 0.00018765388248245246, | |
| "loss": 1.2274, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.40583232077764275, | |
| "grad_norm": 97.94922685274778, | |
| "learning_rate": 0.00018745980655414114, | |
| "loss": 1.0872, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.4082624544349939, | |
| "grad_norm": 44.74994721544976, | |
| "learning_rate": 0.0001872643190826512, | |
| "loss": 1.1244, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.4106925880923451, | |
| "grad_norm": 53.84692426866845, | |
| "learning_rate": 0.00018706742322302064, | |
| "loss": 1.1576, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.4131227217496962, | |
| "grad_norm": 54.43599132185614, | |
| "learning_rate": 0.0001868691221530178, | |
| "loss": 1.0957, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.4155528554070474, | |
| "grad_norm": 39.21766518089018, | |
| "learning_rate": 0.00018666941907309026, | |
| "loss": 1.1625, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.41798298906439857, | |
| "grad_norm": 49.40030697752548, | |
| "learning_rate": 0.000186468317206313, | |
| "loss": 1.1556, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.4204131227217497, | |
| "grad_norm": 101.50309647820374, | |
| "learning_rate": 0.0001862658197983366, | |
| "loss": 1.1687, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.42284325637910086, | |
| "grad_norm": 105.41233861946563, | |
| "learning_rate": 0.0001860619301173347, | |
| "loss": 1.1687, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.425273390036452, | |
| "grad_norm": 103.99749987770305, | |
| "learning_rate": 0.0001858566514539513, | |
| "loss": 1.144, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.42770352369380316, | |
| "grad_norm": 78.83490301242213, | |
| "learning_rate": 0.0001856499871212477, | |
| "loss": 1.2318, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.43013365735115433, | |
| "grad_norm": 62.325757489859335, | |
| "learning_rate": 0.00018544194045464886, | |
| "loss": 1.1092, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.43256379100850545, | |
| "grad_norm": 81.32804926878099, | |
| "learning_rate": 0.00018523251481188986, | |
| "loss": 1.2233, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.4349939246658566, | |
| "grad_norm": 38.97928032166606, | |
| "learning_rate": 0.00018502171357296144, | |
| "loss": 1.2371, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.4374240583232078, | |
| "grad_norm": 82.62345361244209, | |
| "learning_rate": 0.0001848095401400555, | |
| "loss": 1.1562, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.4398541919805589, | |
| "grad_norm": 47.793381366401626, | |
| "learning_rate": 0.0001845959979375104, | |
| "loss": 1.1249, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.4422843256379101, | |
| "grad_norm": 53.6022948471739, | |
| "learning_rate": 0.00018438109041175532, | |
| "loss": 1.1415, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.44471445929526127, | |
| "grad_norm": 65.92717051568573, | |
| "learning_rate": 0.00018416482103125506, | |
| "loss": 1.1748, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.4471445929526124, | |
| "grad_norm": 59.410481167619494, | |
| "learning_rate": 0.0001839471932864537, | |
| "loss": 1.1399, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.44957472660996356, | |
| "grad_norm": 64.22740395872977, | |
| "learning_rate": 0.0001837282106897185, | |
| "loss": 1.2193, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.4520048602673147, | |
| "grad_norm": 54.63497168787729, | |
| "learning_rate": 0.00018350787677528306, | |
| "loss": 1.153, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.45443499392466585, | |
| "grad_norm": 49.60676029637355, | |
| "learning_rate": 0.00018328619509919044, | |
| "loss": 1.1509, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.456865127582017, | |
| "grad_norm": 32.29074835877607, | |
| "learning_rate": 0.00018306316923923563, | |
| "loss": 1.1851, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.45929526123936815, | |
| "grad_norm": 61.13632454163589, | |
| "learning_rate": 0.0001828388027949078, | |
| "loss": 1.1323, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.4617253948967193, | |
| "grad_norm": 67.48617660835801, | |
| "learning_rate": 0.00018261309938733238, | |
| "loss": 1.1956, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.4641555285540705, | |
| "grad_norm": 38.31182257784929, | |
| "learning_rate": 0.00018238606265921238, | |
| "loss": 1.1379, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.4665856622114216, | |
| "grad_norm": 47.30995766708629, | |
| "learning_rate": 0.00018215769627476984, | |
| "loss": 1.1462, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.4690157958687728, | |
| "grad_norm": 34.57093925891121, | |
| "learning_rate": 0.00018192800391968642, | |
| "loss": 1.1979, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.47144592952612396, | |
| "grad_norm": 34.45645740457662, | |
| "learning_rate": 0.0001816969893010442, | |
| "loss": 1.1763, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.4738760631834751, | |
| "grad_norm": 39.21862152859671, | |
| "learning_rate": 0.00018146465614726567, | |
| "loss": 1.1514, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.47630619684082626, | |
| "grad_norm": 34.765347344568106, | |
| "learning_rate": 0.00018123100820805355, | |
| "loss": 1.1426, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.4787363304981774, | |
| "grad_norm": 35.04245362239315, | |
| "learning_rate": 0.00018099604925433043, | |
| "loss": 1.143, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.48116646415552855, | |
| "grad_norm": 103.45636476066032, | |
| "learning_rate": 0.00018075978307817764, | |
| "loss": 1.1713, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.4835965978128797, | |
| "grad_norm": 43.0297373660821, | |
| "learning_rate": 0.00018052221349277442, | |
| "loss": 1.2226, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.48602673147023084, | |
| "grad_norm": 32.80474372048966, | |
| "learning_rate": 0.000180283344332336, | |
| "loss": 1.1556, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.488456865127582, | |
| "grad_norm": 59.42688731224296, | |
| "learning_rate": 0.00018004317945205197, | |
| "loss": 1.1411, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.4908869987849332, | |
| "grad_norm": 102.0917822407188, | |
| "learning_rate": 0.000179801722728024, | |
| "loss": 1.1309, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.4933171324422843, | |
| "grad_norm": 309.9346821950787, | |
| "learning_rate": 0.0001795589780572031, | |
| "loss": 1.1953, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.4957472660996355, | |
| "grad_norm": 344.5019267346993, | |
| "learning_rate": 0.0001793149493573271, | |
| "loss": 1.1524, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.49817739975698666, | |
| "grad_norm": 50.075205946207085, | |
| "learning_rate": 0.00017906964056685706, | |
| "loss": 1.1495, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.5006075334143378, | |
| "grad_norm": 132.32227258331488, | |
| "learning_rate": 0.00017882305564491396, | |
| "loss": 1.1976, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.5006075334143378, | |
| "eval_loss": 1.146019458770752, | |
| "eval_runtime": 52.7816, | |
| "eval_samples_per_second": 14.096, | |
| "eval_steps_per_second": 1.762, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.503037667071689, | |
| "grad_norm": 138.57200377669218, | |
| "learning_rate": 0.00017857519857121458, | |
| "loss": 1.2159, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.5054678007290401, | |
| "grad_norm": 268.41109734161546, | |
| "learning_rate": 0.00017832607334600746, | |
| "loss": 1.1748, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.5078979343863913, | |
| "grad_norm": 72.44153953442401, | |
| "learning_rate": 0.00017807568399000822, | |
| "loss": 1.1758, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.5103280680437424, | |
| "grad_norm": 97.75400124096738, | |
| "learning_rate": 0.00017782403454433477, | |
| "loss": 1.1004, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.5127582017010935, | |
| "grad_norm": 84.19522802756285, | |
| "learning_rate": 0.000177571129070442, | |
| "loss": 1.1397, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.5151883353584447, | |
| "grad_norm": 132.95081835535706, | |
| "learning_rate": 0.00017731697165005618, | |
| "loss": 1.146, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.5176184690157959, | |
| "grad_norm": 560.3351292126325, | |
| "learning_rate": 0.0001770615663851093, | |
| "loss": 1.1937, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.520048602673147, | |
| "grad_norm": 252.72862614645885, | |
| "learning_rate": 0.0001768049173976727, | |
| "loss": 1.1213, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.5224787363304981, | |
| "grad_norm": 356.2985211032981, | |
| "learning_rate": 0.0001765470288298905, | |
| "loss": 1.22, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.5249088699878494, | |
| "grad_norm": 952.600672502031, | |
| "learning_rate": 0.00017628790484391284, | |
| "loss": 1.1321, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.5273390036452005, | |
| "grad_norm": 289.9357041930161, | |
| "learning_rate": 0.0001760275496218288, | |
| "loss": 1.1688, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.5297691373025516, | |
| "grad_norm": 48.69445264741508, | |
| "learning_rate": 0.0001757659673655986, | |
| "loss": 1.1551, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.5321992709599028, | |
| "grad_norm": 40.15160247154335, | |
| "learning_rate": 0.0001755031622969862, | |
| "loss": 1.1459, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.534629404617254, | |
| "grad_norm": 44.59390817019205, | |
| "learning_rate": 0.00017523913865749078, | |
| "loss": 1.2012, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.5370595382746051, | |
| "grad_norm": 30.189717624412484, | |
| "learning_rate": 0.00017497390070827848, | |
| "loss": 1.15, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.5394896719319563, | |
| "grad_norm": 27.185608574176108, | |
| "learning_rate": 0.00017470745273011362, | |
| "loss": 1.0763, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.5419198055893074, | |
| "grad_norm": 99.44121390806423, | |
| "learning_rate": 0.00017443979902328956, | |
| "loss": 1.1478, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.5443499392466585, | |
| "grad_norm": 29.684499344634585, | |
| "learning_rate": 0.00017417094390755934, | |
| "loss": 1.1123, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.5467800729040098, | |
| "grad_norm": 26.788847114635054, | |
| "learning_rate": 0.00017390089172206592, | |
| "loss": 1.1169, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.5492102065613609, | |
| "grad_norm": 31.84817878214798, | |
| "learning_rate": 0.00017362964682527218, | |
| "loss": 1.1524, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.551640340218712, | |
| "grad_norm": 34.834632993822424, | |
| "learning_rate": 0.00017335721359489057, | |
| "loss": 1.1761, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.5540704738760632, | |
| "grad_norm": 66.6084234453716, | |
| "learning_rate": 0.00017308359642781242, | |
| "loss": 1.1175, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.5565006075334143, | |
| "grad_norm": 35.15720180142773, | |
| "learning_rate": 0.00017280879974003707, | |
| "loss": 1.2012, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.5589307411907655, | |
| "grad_norm": 35.975450782756226, | |
| "learning_rate": 0.00017253282796660056, | |
| "loss": 1.1801, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.5613608748481167, | |
| "grad_norm": 83.49050230764925, | |
| "learning_rate": 0.0001722556855615039, | |
| "loss": 1.1576, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.5637910085054678, | |
| "grad_norm": 150.44630441002784, | |
| "learning_rate": 0.00017197737699764146, | |
| "loss": 1.1826, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.5662211421628189, | |
| "grad_norm": 31.322382197739042, | |
| "learning_rate": 0.00017169790676672858, | |
| "loss": 1.1784, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.56865127582017, | |
| "grad_norm": 33.15983653687515, | |
| "learning_rate": 0.0001714172793792291, | |
| "loss": 1.1411, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.5710814094775213, | |
| "grad_norm": 22.206850165103052, | |
| "learning_rate": 0.0001711354993642827, | |
| "loss": 1.1772, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.5735115431348724, | |
| "grad_norm": 43.35721272668955, | |
| "learning_rate": 0.00017085257126963152, | |
| "loss": 1.0915, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.5759416767922235, | |
| "grad_norm": 29.57234737116712, | |
| "learning_rate": 0.0001705684996615472, | |
| "loss": 1.0977, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.5783718104495748, | |
| "grad_norm": 42.929644875053214, | |
| "learning_rate": 0.00017028328912475668, | |
| "loss": 1.1782, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.5808019441069259, | |
| "grad_norm": 32.15711272871687, | |
| "learning_rate": 0.0001699969442623686, | |
| "loss": 1.1855, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.583232077764277, | |
| "grad_norm": 43.64453730184205, | |
| "learning_rate": 0.00016970946969579887, | |
| "loss": 1.1171, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5856622114216282, | |
| "grad_norm": 26.145541544112593, | |
| "learning_rate": 0.00016942087006469592, | |
| "loss": 1.1656, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.5880923450789793, | |
| "grad_norm": 53.98173886095731, | |
| "learning_rate": 0.00016913115002686616, | |
| "loss": 1.1378, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.5905224787363305, | |
| "grad_norm": 50.851193586801195, | |
| "learning_rate": 0.00016884031425819853, | |
| "loss": 1.1338, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.5929526123936817, | |
| "grad_norm": 30.166674036386443, | |
| "learning_rate": 0.0001685483674525891, | |
| "loss": 1.1732, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.5953827460510328, | |
| "grad_norm": 32.580505176392656, | |
| "learning_rate": 0.00016825531432186543, | |
| "loss": 1.143, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.5978128797083839, | |
| "grad_norm": 35.087231952662634, | |
| "learning_rate": 0.0001679611595957103, | |
| "loss": 1.212, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.6002430133657352, | |
| "grad_norm": 44.69578306542608, | |
| "learning_rate": 0.00016766590802158566, | |
| "loss": 1.1527, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.6026731470230863, | |
| "grad_norm": 39.8378839133733, | |
| "learning_rate": 0.00016736956436465573, | |
| "loss": 1.2174, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.6051032806804374, | |
| "grad_norm": 25.571860004032857, | |
| "learning_rate": 0.0001670721334077103, | |
| "loss": 1.1031, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.6075334143377886, | |
| "grad_norm": 27.626061413643438, | |
| "learning_rate": 0.00016677361995108743, | |
| "loss": 1.107, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.6099635479951397, | |
| "grad_norm": 47.405627339857176, | |
| "learning_rate": 0.00016647402881259598, | |
| "loss": 1.1521, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.6123936816524909, | |
| "grad_norm": 31.951762409660272, | |
| "learning_rate": 0.00016617336482743794, | |
| "loss": 1.174, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.6148238153098421, | |
| "grad_norm": 44.304437144236104, | |
| "learning_rate": 0.00016587163284813032, | |
| "loss": 1.1286, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.6172539489671932, | |
| "grad_norm": 21.990501251879344, | |
| "learning_rate": 0.00016556883774442675, | |
| "loss": 1.1927, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.6196840826245443, | |
| "grad_norm": 43.91119350789936, | |
| "learning_rate": 0.00016526498440323914, | |
| "loss": 1.1399, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.6221142162818954, | |
| "grad_norm": 28.064569132249982, | |
| "learning_rate": 0.00016496007772855853, | |
| "loss": 1.1913, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.6245443499392467, | |
| "grad_norm": 99.97142272243896, | |
| "learning_rate": 0.0001646541226413761, | |
| "loss": 1.1694, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.6269744835965978, | |
| "grad_norm": 27.12524206817854, | |
| "learning_rate": 0.00016434712407960373, | |
| "loss": 1.2398, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.6294046172539489, | |
| "grad_norm": 42.99171796479219, | |
| "learning_rate": 0.00016403908699799425, | |
| "loss": 1.145, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.6318347509113001, | |
| "grad_norm": 24.064938768293658, | |
| "learning_rate": 0.00016373001636806153, | |
| "loss": 1.098, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.6342648845686513, | |
| "grad_norm": 31.72232981247621, | |
| "learning_rate": 0.00016341991717800023, | |
| "loss": 1.1779, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.6366950182260024, | |
| "grad_norm": 39.97326887390835, | |
| "learning_rate": 0.00016310879443260528, | |
| "loss": 1.3142, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.6391251518833536, | |
| "grad_norm": 27.519208072826963, | |
| "learning_rate": 0.00016279665315319114, | |
| "loss": 1.2039, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.6415552855407047, | |
| "grad_norm": 52.94895557810481, | |
| "learning_rate": 0.00016248349837751062, | |
| "loss": 1.1718, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.6439854191980559, | |
| "grad_norm": 23.603047222747566, | |
| "learning_rate": 0.0001621693351596739, | |
| "loss": 1.1155, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.6464155528554071, | |
| "grad_norm": 21.400341520569807, | |
| "learning_rate": 0.00016185416857006647, | |
| "loss": 1.1242, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.6488456865127582, | |
| "grad_norm": 51.167335508822276, | |
| "learning_rate": 0.00016153800369526788, | |
| "loss": 1.1746, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.6512758201701093, | |
| "grad_norm": 26.219581065473573, | |
| "learning_rate": 0.00016122084563796905, | |
| "loss": 1.0836, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.6537059538274606, | |
| "grad_norm": 56.820249886600706, | |
| "learning_rate": 0.0001609026995168904, | |
| "loss": 1.1625, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.6561360874848117, | |
| "grad_norm": 37.43384869992443, | |
| "learning_rate": 0.00016058357046669898, | |
| "loss": 1.2143, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.6585662211421628, | |
| "grad_norm": 31.885237168871473, | |
| "learning_rate": 0.00016026346363792567, | |
| "loss": 1.1536, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.660996354799514, | |
| "grad_norm": 34.66147983279251, | |
| "learning_rate": 0.00015994238419688199, | |
| "loss": 1.2095, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.6634264884568651, | |
| "grad_norm": 86.90365354594917, | |
| "learning_rate": 0.00015962033732557686, | |
| "loss": 1.1149, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.6658566221142163, | |
| "grad_norm": 52.21177462889067, | |
| "learning_rate": 0.00015929732822163287, | |
| "loss": 1.1861, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.6682867557715675, | |
| "grad_norm": 92.11184701145604, | |
| "learning_rate": 0.00015897336209820239, | |
| "loss": 1.1853, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.6707168894289186, | |
| "grad_norm": 30.662475573811115, | |
| "learning_rate": 0.00015864844418388342, | |
| "loss": 1.0912, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.6731470230862697, | |
| "grad_norm": 26.15855468837027, | |
| "learning_rate": 0.00015832257972263523, | |
| "loss": 1.1618, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.675577156743621, | |
| "grad_norm": 41.14250673970726, | |
| "learning_rate": 0.00015799577397369375, | |
| "loss": 1.1499, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.6780072904009721, | |
| "grad_norm": 31.93253644773631, | |
| "learning_rate": 0.00015766803221148673, | |
| "loss": 1.1229, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.6804374240583232, | |
| "grad_norm": 39.87120131585165, | |
| "learning_rate": 0.00015733935972554844, | |
| "loss": 1.1647, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.6828675577156743, | |
| "grad_norm": 52.741654062271124, | |
| "learning_rate": 0.0001570097618204345, | |
| "loss": 1.1362, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.6852976913730255, | |
| "grad_norm": 33.13137686002526, | |
| "learning_rate": 0.0001566792438156362, | |
| "loss": 1.1825, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.6877278250303767, | |
| "grad_norm": 20.284041564566042, | |
| "learning_rate": 0.00015634781104549442, | |
| "loss": 1.1439, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.6901579586877278, | |
| "grad_norm": 164.9222932471453, | |
| "learning_rate": 0.00015601546885911404, | |
| "loss": 1.122, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.692588092345079, | |
| "grad_norm": 27.092346730158148, | |
| "learning_rate": 0.00015568222262027717, | |
| "loss": 1.157, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.6950182260024301, | |
| "grad_norm": 39.46898996008012, | |
| "learning_rate": 0.00015534807770735664, | |
| "loss": 1.1092, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.6974483596597812, | |
| "grad_norm": 30.00942949300714, | |
| "learning_rate": 0.00015501303951322943, | |
| "loss": 1.243, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.6998784933171325, | |
| "grad_norm": 31.435817418038887, | |
| "learning_rate": 0.00015467711344518942, | |
| "loss": 1.1034, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.7023086269744836, | |
| "grad_norm": 54.53572773177548, | |
| "learning_rate": 0.00015434030492486023, | |
| "loss": 1.2216, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.7047387606318347, | |
| "grad_norm": 24.51082708234768, | |
| "learning_rate": 0.00015400261938810757, | |
| "loss": 1.1532, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.707168894289186, | |
| "grad_norm": 104.85480514443172, | |
| "learning_rate": 0.00015366406228495172, | |
| "loss": 1.1156, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.7095990279465371, | |
| "grad_norm": 26.398830117870997, | |
| "learning_rate": 0.0001533246390794794, | |
| "loss": 1.0934, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.7120291616038882, | |
| "grad_norm": 25.062392373037707, | |
| "learning_rate": 0.00015298435524975572, | |
| "loss": 1.1453, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.7144592952612394, | |
| "grad_norm": 25.385505352027444, | |
| "learning_rate": 0.0001526432162877356, | |
| "loss": 1.1359, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.7168894289185905, | |
| "grad_norm": 18.00146943000571, | |
| "learning_rate": 0.00015230122769917527, | |
| "loss": 1.1129, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.7193195625759417, | |
| "grad_norm": 22.55383473288135, | |
| "learning_rate": 0.00015195839500354335, | |
| "loss": 1.142, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.7217496962332929, | |
| "grad_norm": 30.013723395820165, | |
| "learning_rate": 0.00015161472373393186, | |
| "loss": 1.1379, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.724179829890644, | |
| "grad_norm": 40.566201545240425, | |
| "learning_rate": 0.0001512702194369668, | |
| "loss": 1.1326, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.7266099635479951, | |
| "grad_norm": 27.34716639907029, | |
| "learning_rate": 0.00015092488767271857, | |
| "loss": 1.0782, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.7290400972053463, | |
| "grad_norm": 45.0837594669075, | |
| "learning_rate": 0.00015057873401461253, | |
| "loss": 1.2054, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.7314702308626975, | |
| "grad_norm": 22.39794101270309, | |
| "learning_rate": 0.00015023176404933874, | |
| "loss": 1.1052, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.7339003645200486, | |
| "grad_norm": 21.818512025585306, | |
| "learning_rate": 0.00014988398337676198, | |
| "loss": 1.1664, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.7363304981773997, | |
| "grad_norm": 33.09386163968815, | |
| "learning_rate": 0.00014953539760983122, | |
| "loss": 1.1364, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.7387606318347509, | |
| "grad_norm": 26.3253592215911, | |
| "learning_rate": 0.00014918601237448923, | |
| "loss": 1.1093, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.741190765492102, | |
| "grad_norm": 32.54878723405212, | |
| "learning_rate": 0.0001488358333095816, | |
| "loss": 1.182, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.7436208991494532, | |
| "grad_norm": 28.645473311846015, | |
| "learning_rate": 0.0001484848660667658, | |
| "loss": 1.2064, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.7460510328068044, | |
| "grad_norm": 29.02693042820854, | |
| "learning_rate": 0.00014813311631041995, | |
| "loss": 1.1545, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.7484811664641555, | |
| "grad_norm": 20.28193033099828, | |
| "learning_rate": 0.00014778058971755154, | |
| "loss": 1.1885, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.7509113001215066, | |
| "grad_norm": 121.86121371804961, | |
| "learning_rate": 0.00014742729197770552, | |
| "loss": 1.095, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.7509113001215066, | |
| "eval_loss": 1.133868932723999, | |
| "eval_runtime": 52.6711, | |
| "eval_samples_per_second": 14.125, | |
| "eval_steps_per_second": 1.766, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.7533414337788579, | |
| "grad_norm": 50.1793074315811, | |
| "learning_rate": 0.00014707322879287276, | |
| "loss": 1.1679, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.755771567436209, | |
| "grad_norm": 31.791309498678103, | |
| "learning_rate": 0.00014671840587739783, | |
| "loss": 1.1277, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.7582017010935601, | |
| "grad_norm": 56.88911226488106, | |
| "learning_rate": 0.00014636282895788688, | |
| "loss": 1.1492, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.7606318347509113, | |
| "grad_norm": 117.29437608667352, | |
| "learning_rate": 0.00014600650377311522, | |
| "loss": 1.1123, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.7630619684082625, | |
| "grad_norm": 107.56728772749254, | |
| "learning_rate": 0.00014564943607393459, | |
| "loss": 1.171, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.7654921020656136, | |
| "grad_norm": 34.085830256919685, | |
| "learning_rate": 0.0001452916316231805, | |
| "loss": 1.1854, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.7679222357229648, | |
| "grad_norm": 23.625747202851176, | |
| "learning_rate": 0.000144933096195579, | |
| "loss": 1.1622, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.7703523693803159, | |
| "grad_norm": 56.9917185309248, | |
| "learning_rate": 0.00014457383557765386, | |
| "loss": 1.2037, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.772782503037667, | |
| "grad_norm": 34.55554043725056, | |
| "learning_rate": 0.00014421385556763266, | |
| "loss": 1.1273, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.7752126366950183, | |
| "grad_norm": 34.205286759913115, | |
| "learning_rate": 0.00014385316197535372, | |
| "loss": 1.2039, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.7776427703523694, | |
| "grad_norm": 27.30015395778206, | |
| "learning_rate": 0.00014349176062217195, | |
| "loss": 1.1903, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.7800729040097205, | |
| "grad_norm": 23.077745147127867, | |
| "learning_rate": 0.00014312965734086518, | |
| "loss": 1.1539, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.7825030376670717, | |
| "grad_norm": 26.22112568156326, | |
| "learning_rate": 0.00014276685797553977, | |
| "loss": 1.1807, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.7849331713244229, | |
| "grad_norm": 34.813719314948514, | |
| "learning_rate": 0.0001424033683815365, | |
| "loss": 1.1247, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.787363304981774, | |
| "grad_norm": 27.109609629038324, | |
| "learning_rate": 0.00014203919442533597, | |
| "loss": 1.1735, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.7897934386391251, | |
| "grad_norm": 144.91672798575476, | |
| "learning_rate": 0.00014167434198446383, | |
| "loss": 1.1007, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.7922235722964763, | |
| "grad_norm": 42.19042828736382, | |
| "learning_rate": 0.00014130881694739616, | |
| "loss": 1.1398, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.7946537059538274, | |
| "grad_norm": 43.00144921766715, | |
| "learning_rate": 0.00014094262521346427, | |
| "loss": 1.1712, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.7970838396111786, | |
| "grad_norm": 26.343159670729925, | |
| "learning_rate": 0.0001405757726927595, | |
| "loss": 1.2103, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.7995139732685298, | |
| "grad_norm": 31.68271222195729, | |
| "learning_rate": 0.00014020826530603776, | |
| "loss": 1.1578, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.8019441069258809, | |
| "grad_norm": 39.08920292536896, | |
| "learning_rate": 0.00013984010898462416, | |
| "loss": 1.1377, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.804374240583232, | |
| "grad_norm": 34.56898084569197, | |
| "learning_rate": 0.00013947130967031717, | |
| "loss": 1.1886, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.8068043742405833, | |
| "grad_norm": 42.016356369933895, | |
| "learning_rate": 0.00013910187331529276, | |
| "loss": 1.1577, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.8092345078979344, | |
| "grad_norm": 21.25953597879822, | |
| "learning_rate": 0.00013873180588200827, | |
| "loss": 1.1259, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.8116646415552855, | |
| "grad_norm": 39.49634140985428, | |
| "learning_rate": 0.0001383611133431062, | |
| "loss": 1.173, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.8140947752126367, | |
| "grad_norm": 29.837690582268863, | |
| "learning_rate": 0.00013798980168131794, | |
| "loss": 1.1322, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.8165249088699879, | |
| "grad_norm": 23.510451396240928, | |
| "learning_rate": 0.000137617876889367, | |
| "loss": 1.1392, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.818955042527339, | |
| "grad_norm": 19.183017199526635, | |
| "learning_rate": 0.00013724534496987247, | |
| "loss": 1.157, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.8213851761846902, | |
| "grad_norm": 51.85037647612581, | |
| "learning_rate": 0.0001368722119352521, | |
| "loss": 1.1255, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.8238153098420413, | |
| "grad_norm": 31.635699477838273, | |
| "learning_rate": 0.00013649848380762513, | |
| "loss": 1.1429, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.8262454434993924, | |
| "grad_norm": 39.6479124739029, | |
| "learning_rate": 0.00013612416661871533, | |
| "loss": 1.1609, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.8286755771567437, | |
| "grad_norm": 21.453228401011238, | |
| "learning_rate": 0.0001357492664097534, | |
| "loss": 1.1247, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.8311057108140948, | |
| "grad_norm": 28.514958428145494, | |
| "learning_rate": 0.00013537378923137973, | |
| "loss": 1.0845, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.8335358444714459, | |
| "grad_norm": 26.98663985253516, | |
| "learning_rate": 0.00013499774114354655, | |
| "loss": 1.1092, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.8359659781287971, | |
| "grad_norm": 30.76143424141064, | |
| "learning_rate": 0.00013462112821542016, | |
| "loss": 1.1759, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.8383961117861483, | |
| "grad_norm": 39.023771167108656, | |
| "learning_rate": 0.0001342439565252831, | |
| "loss": 1.1024, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.8408262454434994, | |
| "grad_norm": 29.787639099820225, | |
| "learning_rate": 0.0001338662321604358, | |
| "loss": 1.2141, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.8432563791008505, | |
| "grad_norm": 25.60634301240642, | |
| "learning_rate": 0.00013348796121709862, | |
| "loss": 1.1244, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.8456865127582017, | |
| "grad_norm": 76.98542857181108, | |
| "learning_rate": 0.00013310914980031334, | |
| "loss": 1.19, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.8481166464155528, | |
| "grad_norm": 110.28982985071892, | |
| "learning_rate": 0.0001327298040238446, | |
| "loss": 1.1295, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.850546780072904, | |
| "grad_norm": 22.610631125609732, | |
| "learning_rate": 0.0001323499300100811, | |
| "loss": 1.1445, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.8529769137302552, | |
| "grad_norm": 29.958515973723888, | |
| "learning_rate": 0.00013196953388993726, | |
| "loss": 1.2048, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.8554070473876063, | |
| "grad_norm": 30.691798031468103, | |
| "learning_rate": 0.00013158862180275363, | |
| "loss": 1.1628, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.8578371810449574, | |
| "grad_norm": 28.568576369680258, | |
| "learning_rate": 0.00013120719989619833, | |
| "loss": 1.0899, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.8602673147023087, | |
| "grad_norm": 42.12623456189728, | |
| "learning_rate": 0.0001308252743261675, | |
| "loss": 1.1451, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.8626974483596598, | |
| "grad_norm": 112.39248005736448, | |
| "learning_rate": 0.00013044285125668614, | |
| "loss": 1.154, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.8651275820170109, | |
| "grad_norm": 28.013602355549782, | |
| "learning_rate": 0.0001300599368598086, | |
| "loss": 1.1937, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.8675577156743621, | |
| "grad_norm": 27.763517972300694, | |
| "learning_rate": 0.0001296765373155188, | |
| "loss": 1.1243, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.8699878493317132, | |
| "grad_norm": 112.85815824767063, | |
| "learning_rate": 0.0001292926588116308, | |
| "loss": 1.1595, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.8724179829890644, | |
| "grad_norm": 27.085127886556087, | |
| "learning_rate": 0.00012890830754368855, | |
| "loss": 1.1196, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.8748481166464156, | |
| "grad_norm": 31.56336829128541, | |
| "learning_rate": 0.00012852348971486617, | |
| "loss": 1.1231, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.8772782503037667, | |
| "grad_norm": 31.904393738907178, | |
| "learning_rate": 0.0001281382115358679, | |
| "loss": 1.097, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.8797083839611178, | |
| "grad_norm": 25.034453894065827, | |
| "learning_rate": 0.00012775247922482748, | |
| "loss": 1.1246, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.8821385176184691, | |
| "grad_norm": 33.221958266501474, | |
| "learning_rate": 0.0001273662990072083, | |
| "loss": 1.1189, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.8845686512758202, | |
| "grad_norm": 26.638980136773224, | |
| "learning_rate": 0.00012697967711570242, | |
| "loss": 1.1315, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.8869987849331713, | |
| "grad_norm": 27.231479341362885, | |
| "learning_rate": 0.00012659261979013043, | |
| "loss": 1.1464, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.8894289185905225, | |
| "grad_norm": 19.654091006710207, | |
| "learning_rate": 0.0001262051332773404, | |
| "loss": 1.1271, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.8918590522478737, | |
| "grad_norm": 50.3934263865559, | |
| "learning_rate": 0.00012581722383110718, | |
| "loss": 1.1002, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.8942891859052248, | |
| "grad_norm": 20.25952031318632, | |
| "learning_rate": 0.00012542889771203166, | |
| "loss": 1.0629, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.8967193195625759, | |
| "grad_norm": 19.16914945262315, | |
| "learning_rate": 0.00012504016118743935, | |
| "loss": 1.1597, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.8991494532199271, | |
| "grad_norm": 35.65941460173898, | |
| "learning_rate": 0.00012465102053127957, | |
| "loss": 1.1501, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.9015795868772782, | |
| "grad_norm": 26.093269180565315, | |
| "learning_rate": 0.00012426148202402404, | |
| "loss": 1.1455, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.9040097205346294, | |
| "grad_norm": 30.928987547424892, | |
| "learning_rate": 0.00012387155195256537, | |
| "loss": 1.1392, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.9064398541919806, | |
| "grad_norm": 20.17512596846915, | |
| "learning_rate": 0.00012348123661011601, | |
| "loss": 1.1196, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.9088699878493317, | |
| "grad_norm": 24.380789157356805, | |
| "learning_rate": 0.00012309054229610623, | |
| "loss": 1.1, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.9113001215066828, | |
| "grad_norm": 95.49408387682203, | |
| "learning_rate": 0.00012269947531608276, | |
| "loss": 1.1825, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.913730255164034, | |
| "grad_norm": 23.635286340368726, | |
| "learning_rate": 0.0001223080419816069, | |
| "loss": 1.1717, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.9161603888213852, | |
| "grad_norm": 21.942478063568313, | |
| "learning_rate": 0.00012191624861015254, | |
| "loss": 1.1661, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.9185905224787363, | |
| "grad_norm": 74.12601397150299, | |
| "learning_rate": 0.00012152410152500453, | |
| "loss": 1.1967, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.9210206561360875, | |
| "grad_norm": 37.26720386499629, | |
| "learning_rate": 0.00012113160705515625, | |
| "loss": 1.1566, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.9234507897934386, | |
| "grad_norm": 34.080854733427635, | |
| "learning_rate": 0.00012073877153520776, | |
| "loss": 1.0847, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.9258809234507898, | |
| "grad_norm": 26.50842916877183, | |
| "learning_rate": 0.0001203456013052634, | |
| "loss": 1.0824, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.928311057108141, | |
| "grad_norm": 37.92039651416441, | |
| "learning_rate": 0.00011995210271082944, | |
| "loss": 1.1485, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.9307411907654921, | |
| "grad_norm": 38.56931832374284, | |
| "learning_rate": 0.00011955828210271187, | |
| "loss": 1.0737, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.9331713244228432, | |
| "grad_norm": 24.419015296791592, | |
| "learning_rate": 0.0001191641458369136, | |
| "loss": 1.1208, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.9356014580801945, | |
| "grad_norm": 28.75379656643836, | |
| "learning_rate": 0.00011876970027453222, | |
| "loss": 1.1071, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.9380315917375456, | |
| "grad_norm": 138.39305133994282, | |
| "learning_rate": 0.00011837495178165706, | |
| "loss": 1.1405, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.9404617253948967, | |
| "grad_norm": 22.200435229928654, | |
| "learning_rate": 0.00011797990672926652, | |
| "loss": 1.124, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.9428918590522479, | |
| "grad_norm": 40.21978055156661, | |
| "learning_rate": 0.00011758457149312538, | |
| "loss": 1.1875, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.945321992709599, | |
| "grad_norm": 23.592672098002485, | |
| "learning_rate": 0.00011718895245368167, | |
| "loss": 1.1748, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.9477521263669502, | |
| "grad_norm": 17.463183827323444, | |
| "learning_rate": 0.00011679305599596393, | |
| "loss": 1.1794, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.9501822600243013, | |
| "grad_norm": 36.219441964332646, | |
| "learning_rate": 0.00011639688850947799, | |
| "loss": 1.1459, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.9526123936816525, | |
| "grad_norm": 23.727472560980413, | |
| "learning_rate": 0.00011600045638810386, | |
| "loss": 1.076, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.9550425273390036, | |
| "grad_norm": 57.63284414960702, | |
| "learning_rate": 0.00011560376602999272, | |
| "loss": 1.1919, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.9574726609963548, | |
| "grad_norm": 40.23829998466358, | |
| "learning_rate": 0.00011520682383746333, | |
| "loss": 1.0701, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.959902794653706, | |
| "grad_norm": 58.2018640218209, | |
| "learning_rate": 0.00011480963621689905, | |
| "loss": 1.1745, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.9623329283110571, | |
| "grad_norm": 27.693448904288406, | |
| "learning_rate": 0.00011441220957864421, | |
| "loss": 1.1323, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.9647630619684082, | |
| "grad_norm": 34.94430005820724, | |
| "learning_rate": 0.00011401455033690076, | |
| "loss": 1.1497, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.9671931956257594, | |
| "grad_norm": 17.521922247865188, | |
| "learning_rate": 0.00011361666490962468, | |
| "loss": 1.1319, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.9696233292831106, | |
| "grad_norm": 25.886687159935246, | |
| "learning_rate": 0.00011321855971842243, | |
| "loss": 1.1418, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.9720534629404617, | |
| "grad_norm": 31.388154506614836, | |
| "learning_rate": 0.00011282024118844738, | |
| "loss": 1.1282, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.9744835965978129, | |
| "grad_norm": 27.458601253675347, | |
| "learning_rate": 0.00011242171574829599, | |
| "loss": 1.1647, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.976913730255164, | |
| "grad_norm": 25.922873022924257, | |
| "learning_rate": 0.00011202298982990411, | |
| "loss": 1.091, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.9793438639125152, | |
| "grad_norm": 20.129467589894766, | |
| "learning_rate": 0.00011162406986844323, | |
| "loss": 1.2, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.9817739975698664, | |
| "grad_norm": 25.11892123906363, | |
| "learning_rate": 0.00011122496230221645, | |
| "loss": 1.0731, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.9842041312272175, | |
| "grad_norm": 26.416884392453543, | |
| "learning_rate": 0.00011082567357255484, | |
| "loss": 1.1836, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.9866342648845686, | |
| "grad_norm": 18.768078773975784, | |
| "learning_rate": 0.00011042621012371322, | |
| "loss": 1.1275, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.9890643985419199, | |
| "grad_norm": 22.275756523796257, | |
| "learning_rate": 0.00011002657840276627, | |
| "loss": 1.1228, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.991494532199271, | |
| "grad_norm": 29.605335344828575, | |
| "learning_rate": 0.00010962678485950455, | |
| "loss": 1.0255, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.9939246658566221, | |
| "grad_norm": 41.1718200727633, | |
| "learning_rate": 0.00010922683594633021, | |
| "loss": 1.1876, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.9963547995139733, | |
| "grad_norm": 20.46397475257922, | |
| "learning_rate": 0.00010882673811815304, | |
| "loss": 1.1168, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.9987849331713244, | |
| "grad_norm": 21.084924025016928, | |
| "learning_rate": 0.00010842649783228624, | |
| "loss": 1.1948, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 21.084924025016928, | |
| "learning_rate": 0.00010802612154834211, | |
| "loss": 1.1076, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 1.121336579322815, | |
| "eval_runtime": 52.7043, | |
| "eval_samples_per_second": 14.116, | |
| "eval_steps_per_second": 1.765, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 1.0024301336573511, | |
| "grad_norm": 35.25758968935371, | |
| "learning_rate": 0.00010762561572812788, | |
| "loss": 1.1335, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 1.0048602673147022, | |
| "grad_norm": 20.78715726366623, | |
| "learning_rate": 0.0001072249868355415, | |
| "loss": 1.1003, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 1.0072904009720534, | |
| "grad_norm": 31.01116633763719, | |
| "learning_rate": 0.0001068242413364671, | |
| "loss": 1.1225, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 1.0097205346294047, | |
| "grad_norm": 19.050638172672897, | |
| "learning_rate": 0.00010642338569867086, | |
| "loss": 1.0595, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 1.0121506682867558, | |
| "grad_norm": 41.54235389574412, | |
| "learning_rate": 0.00010602242639169648, | |
| "loss": 1.1719, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 1.014580801944107, | |
| "grad_norm": 41.34218206464363, | |
| "learning_rate": 0.00010562136988676078, | |
| "loss": 1.1292, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 1.017010935601458, | |
| "grad_norm": 32.436985934581934, | |
| "learning_rate": 0.0001052202226566494, | |
| "loss": 1.1244, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 1.0194410692588092, | |
| "grad_norm": 19.631825450596665, | |
| "learning_rate": 0.0001048189911756121, | |
| "loss": 1.1323, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.0218712029161603, | |
| "grad_norm": 23.275029440216805, | |
| "learning_rate": 0.00010441768191925847, | |
| "loss": 1.1605, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 1.0243013365735116, | |
| "grad_norm": 21.44161988455765, | |
| "learning_rate": 0.0001040163013644533, | |
| "loss": 1.0886, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 1.0267314702308628, | |
| "grad_norm": 31.9765167465431, | |
| "learning_rate": 0.00010361485598921212, | |
| "loss": 1.1378, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 1.0291616038882139, | |
| "grad_norm": 22.340741556027833, | |
| "learning_rate": 0.00010321335227259661, | |
| "loss": 1.1278, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 1.031591737545565, | |
| "grad_norm": 29.27286563037163, | |
| "learning_rate": 0.00010281179669461005, | |
| "loss": 1.1186, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 1.034021871202916, | |
| "grad_norm": 65.85877610734141, | |
| "learning_rate": 0.00010241019573609269, | |
| "loss": 1.1673, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 1.0364520048602672, | |
| "grad_norm": 35.173784527846884, | |
| "learning_rate": 0.00010200855587861724, | |
| "loss": 1.0903, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 1.0388821385176186, | |
| "grad_norm": 29.91546238299385, | |
| "learning_rate": 0.00010160688360438419, | |
| "loss": 1.0884, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 1.0413122721749697, | |
| "grad_norm": 26.873308685100223, | |
| "learning_rate": 0.0001012051853961172, | |
| "loss": 1.1296, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 1.0437424058323208, | |
| "grad_norm": 25.90622275527891, | |
| "learning_rate": 0.00010080346773695853, | |
| "loss": 1.1349, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.046172539489672, | |
| "grad_norm": 21.388851321680434, | |
| "learning_rate": 0.00010040173711036431, | |
| "loss": 1.0947, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 1.048602673147023, | |
| "grad_norm": 31.206506843880053, | |
| "learning_rate": 0.0001, | |
| "loss": 1.1541, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 1.0510328068043742, | |
| "grad_norm": 19.486767323523555, | |
| "learning_rate": 9.959826288963571e-05, | |
| "loss": 1.1574, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 1.0534629404617255, | |
| "grad_norm": 102.81325604770561, | |
| "learning_rate": 9.919653226304148e-05, | |
| "loss": 1.1762, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 1.0558930741190766, | |
| "grad_norm": 17.18170280255333, | |
| "learning_rate": 9.879481460388282e-05, | |
| "loss": 1.1208, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 1.0583232077764277, | |
| "grad_norm": 29.88292309614927, | |
| "learning_rate": 9.839311639561583e-05, | |
| "loss": 1.1114, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 1.0607533414337789, | |
| "grad_norm": 23.50392429976475, | |
| "learning_rate": 9.799144412138275e-05, | |
| "loss": 1.2026, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 1.06318347509113, | |
| "grad_norm": 24.794408487434744, | |
| "learning_rate": 9.758980426390732e-05, | |
| "loss": 1.1587, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 1.065613608748481, | |
| "grad_norm": 38.726295800289655, | |
| "learning_rate": 9.718820330538998e-05, | |
| "loss": 1.14, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 1.0680437424058322, | |
| "grad_norm": 31.152256057732977, | |
| "learning_rate": 9.678664772740343e-05, | |
| "loss": 1.0882, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.0704738760631836, | |
| "grad_norm": 65.73380095432839, | |
| "learning_rate": 9.638514401078788e-05, | |
| "loss": 1.1213, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 1.0729040097205347, | |
| "grad_norm": 69.07317297910537, | |
| "learning_rate": 9.598369863554673e-05, | |
| "loss": 1.1285, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 1.0753341433778858, | |
| "grad_norm": 62.55969576940585, | |
| "learning_rate": 9.558231808074156e-05, | |
| "loss": 1.1252, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 1.077764277035237, | |
| "grad_norm": 26.35106444530265, | |
| "learning_rate": 9.51810088243879e-05, | |
| "loss": 1.108, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 1.080194410692588, | |
| "grad_norm": 76.70006955440516, | |
| "learning_rate": 9.477977734335061e-05, | |
| "loss": 1.1144, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 1.0826245443499392, | |
| "grad_norm": 22.376983523395264, | |
| "learning_rate": 9.437863011323922e-05, | |
| "loss": 1.173, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 1.0850546780072905, | |
| "grad_norm": 33.51322062360491, | |
| "learning_rate": 9.397757360830353e-05, | |
| "loss": 1.089, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 1.0874848116646416, | |
| "grad_norm": 24.87252097324779, | |
| "learning_rate": 9.357661430132915e-05, | |
| "loss": 1.098, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 1.0899149453219927, | |
| "grad_norm": 48.95371674408058, | |
| "learning_rate": 9.317575866353292e-05, | |
| "loss": 1.0491, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 1.0923450789793439, | |
| "grad_norm": 25.50740340531524, | |
| "learning_rate": 9.277501316445854e-05, | |
| "loss": 1.0939, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.094775212636695, | |
| "grad_norm": 27.60998778610316, | |
| "learning_rate": 9.23743842718721e-05, | |
| "loss": 1.1564, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 1.097205346294046, | |
| "grad_norm": 63.99226186124907, | |
| "learning_rate": 9.197387845165793e-05, | |
| "loss": 1.1088, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 1.0996354799513974, | |
| "grad_norm": 36.441157466567596, | |
| "learning_rate": 9.157350216771378e-05, | |
| "loss": 1.0897, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 1.1020656136087486, | |
| "grad_norm": 32.32587774153429, | |
| "learning_rate": 9.117326188184695e-05, | |
| "loss": 1.1285, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 1.1044957472660997, | |
| "grad_norm": 33.39257750037465, | |
| "learning_rate": 9.077316405366981e-05, | |
| "loss": 1.1568, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 1.1069258809234508, | |
| "grad_norm": 45.03485873480868, | |
| "learning_rate": 9.037321514049548e-05, | |
| "loss": 1.0791, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 1.109356014580802, | |
| "grad_norm": 35.1451377482015, | |
| "learning_rate": 8.997342159723371e-05, | |
| "loss": 1.1243, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 1.111786148238153, | |
| "grad_norm": 67.01465976966, | |
| "learning_rate": 8.957378987628682e-05, | |
| "loss": 1.0978, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 1.1142162818955041, | |
| "grad_norm": 33.057859846207634, | |
| "learning_rate": 8.917432642744518e-05, | |
| "loss": 1.1431, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 1.1166464155528555, | |
| "grad_norm": 30.602840863536635, | |
| "learning_rate": 8.877503769778356e-05, | |
| "loss": 1.1157, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.1190765492102066, | |
| "grad_norm": 38.088467248288964, | |
| "learning_rate": 8.83759301315568e-05, | |
| "loss": 1.0776, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 1.1215066828675577, | |
| "grad_norm": 66.03671829863266, | |
| "learning_rate": 8.797701017009591e-05, | |
| "loss": 1.1468, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 1.1239368165249088, | |
| "grad_norm": 32.293691874682686, | |
| "learning_rate": 8.757828425170404e-05, | |
| "loss": 1.1115, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 1.12636695018226, | |
| "grad_norm": 32.70707175332633, | |
| "learning_rate": 8.717975881155261e-05, | |
| "loss": 1.1677, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 1.128797083839611, | |
| "grad_norm": 48.79069594971439, | |
| "learning_rate": 8.678144028157759e-05, | |
| "loss": 1.1341, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 1.1312272174969624, | |
| "grad_norm": 37.52808559072613, | |
| "learning_rate": 8.638333509037536e-05, | |
| "loss": 1.1414, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 1.1336573511543135, | |
| "grad_norm": 27.096068124970536, | |
| "learning_rate": 8.598544966309925e-05, | |
| "loss": 1.1719, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 1.1360874848116647, | |
| "grad_norm": 16.019227077248434, | |
| "learning_rate": 8.55877904213558e-05, | |
| "loss": 1.1148, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 1.1385176184690158, | |
| "grad_norm": 29.861941956913498, | |
| "learning_rate": 8.519036378310096e-05, | |
| "loss": 1.1486, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 1.140947752126367, | |
| "grad_norm": 23.058998452019107, | |
| "learning_rate": 8.47931761625367e-05, | |
| "loss": 1.0745, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.143377885783718, | |
| "grad_norm": 24.486692418227875, | |
| "learning_rate": 8.43962339700073e-05, | |
| "loss": 1.1333, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 1.1458080194410694, | |
| "grad_norm": 31.632544516924323, | |
| "learning_rate": 8.399954361189615e-05, | |
| "loss": 1.1565, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 1.1482381530984205, | |
| "grad_norm": 21.67735267443374, | |
| "learning_rate": 8.360311149052205e-05, | |
| "loss": 1.109, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 1.1506682867557716, | |
| "grad_norm": 29.096918560226527, | |
| "learning_rate": 8.320694400403606e-05, | |
| "loss": 1.1517, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 1.1530984204131227, | |
| "grad_norm": 46.067313216206955, | |
| "learning_rate": 8.281104754631835e-05, | |
| "loss": 1.1043, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 1.1555285540704738, | |
| "grad_norm": 30.84953769166141, | |
| "learning_rate": 8.241542850687465e-05, | |
| "loss": 1.1081, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 1.157958687727825, | |
| "grad_norm": 39.34158523904847, | |
| "learning_rate": 8.20200932707335e-05, | |
| "loss": 1.1787, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 1.160388821385176, | |
| "grad_norm": 39.14663302484904, | |
| "learning_rate": 8.162504821834295e-05, | |
| "loss": 1.202, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 1.1628189550425274, | |
| "grad_norm": 49.7279004249915, | |
| "learning_rate": 8.123029972546781e-05, | |
| "loss": 1.1439, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 1.1652490886998785, | |
| "grad_norm": 35.49897960878779, | |
| "learning_rate": 8.083585416308642e-05, | |
| "loss": 1.0741, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.1676792223572297, | |
| "grad_norm": 31.306252618855535, | |
| "learning_rate": 8.044171789728816e-05, | |
| "loss": 1.0697, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 1.1701093560145808, | |
| "grad_norm": 22.40745672651249, | |
| "learning_rate": 8.004789728917059e-05, | |
| "loss": 1.1498, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 1.172539489671932, | |
| "grad_norm": 32.19326746671122, | |
| "learning_rate": 7.965439869473664e-05, | |
| "loss": 1.1392, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 1.1749696233292832, | |
| "grad_norm": 33.66876390791385, | |
| "learning_rate": 7.926122846479224e-05, | |
| "loss": 1.1049, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 1.1773997569866343, | |
| "grad_norm": 35.43357233261174, | |
| "learning_rate": 7.886839294484377e-05, | |
| "loss": 1.0467, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 1.1798298906439855, | |
| "grad_norm": 50.660998166256256, | |
| "learning_rate": 7.84758984749955e-05, | |
| "loss": 1.1244, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 1.1822600243013366, | |
| "grad_norm": 41.356845334605936, | |
| "learning_rate": 7.808375138984745e-05, | |
| "loss": 1.1279, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 1.1846901579586877, | |
| "grad_norm": 22.947663723281487, | |
| "learning_rate": 7.769195801839313e-05, | |
| "loss": 1.0787, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 1.1871202916160388, | |
| "grad_norm": 36.434647074399905, | |
| "learning_rate": 7.730052468391725e-05, | |
| "loss": 1.1148, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 1.18955042527339, | |
| "grad_norm": 75.94549877059467, | |
| "learning_rate": 7.690945770389377e-05, | |
| "loss": 1.1127, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.1919805589307413, | |
| "grad_norm": 68.03126664734435, | |
| "learning_rate": 7.6518763389884e-05, | |
| "loss": 1.1672, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 1.1944106925880924, | |
| "grad_norm": 40.15361719091623, | |
| "learning_rate": 7.612844804743466e-05, | |
| "loss": 1.0962, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 1.1968408262454435, | |
| "grad_norm": 105.80023571763755, | |
| "learning_rate": 7.573851797597602e-05, | |
| "loss": 1.1091, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 1.1992709599027946, | |
| "grad_norm": 41.84401502420881, | |
| "learning_rate": 7.534897946872042e-05, | |
| "loss": 1.1359, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 1.2017010935601458, | |
| "grad_norm": 21.985533615468846, | |
| "learning_rate": 7.495983881256067e-05, | |
| "loss": 1.1024, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 1.2041312272174969, | |
| "grad_norm": 23.02649898605792, | |
| "learning_rate": 7.457110228796838e-05, | |
| "loss": 1.1089, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 1.206561360874848, | |
| "grad_norm": 74.4950498938832, | |
| "learning_rate": 7.418277616889282e-05, | |
| "loss": 1.0439, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 1.2089914945321993, | |
| "grad_norm": 27.637660484960865, | |
| "learning_rate": 7.379486672265964e-05, | |
| "loss": 1.1453, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 1.2114216281895505, | |
| "grad_norm": 34.98561655821008, | |
| "learning_rate": 7.340738020986961e-05, | |
| "loss": 1.139, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 1.2138517618469016, | |
| "grad_norm": 28.47627677351389, | |
| "learning_rate": 7.302032288429756e-05, | |
| "loss": 1.0623, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.2162818955042527, | |
| "grad_norm": 39.551486186427596, | |
| "learning_rate": 7.263370099279172e-05, | |
| "loss": 1.1277, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 1.2187120291616038, | |
| "grad_norm": 44.12973085459368, | |
| "learning_rate": 7.224752077517253e-05, | |
| "loss": 1.1768, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 1.2211421628189552, | |
| "grad_norm": 84.84836585196132, | |
| "learning_rate": 7.186178846413214e-05, | |
| "loss": 1.1892, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 1.2235722964763063, | |
| "grad_norm": 34.94807915131505, | |
| "learning_rate": 7.147651028513383e-05, | |
| "loss": 1.1108, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 1.2260024301336574, | |
| "grad_norm": 46.19847384406232, | |
| "learning_rate": 7.109169245631149e-05, | |
| "loss": 1.0956, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 1.2284325637910085, | |
| "grad_norm": 38.58484473058957, | |
| "learning_rate": 7.070734118836925e-05, | |
| "loss": 1.1175, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 1.2308626974483596, | |
| "grad_norm": 37.84739298111386, | |
| "learning_rate": 7.032346268448118e-05, | |
| "loss": 1.1411, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 1.2332928311057108, | |
| "grad_norm": 53.5471335398439, | |
| "learning_rate": 6.994006314019141e-05, | |
| "loss": 1.1332, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 1.2357229647630619, | |
| "grad_norm": 91.55067777365485, | |
| "learning_rate": 6.955714874331387e-05, | |
| "loss": 1.1205, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 1.2381530984204132, | |
| "grad_norm": 27.05333642785952, | |
| "learning_rate": 6.917472567383252e-05, | |
| "loss": 1.099, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.2405832320777643, | |
| "grad_norm": 24.519879042487336, | |
| "learning_rate": 6.87928001038017e-05, | |
| "loss": 1.1401, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 1.2430133657351154, | |
| "grad_norm": 33.763495598365786, | |
| "learning_rate": 6.84113781972464e-05, | |
| "loss": 1.2058, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 1.2454434993924666, | |
| "grad_norm": 34.49114206138826, | |
| "learning_rate": 6.803046611006278e-05, | |
| "loss": 1.1044, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 1.2478736330498177, | |
| "grad_norm": 74.20211157975073, | |
| "learning_rate": 6.765006998991888e-05, | |
| "loss": 1.111, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 1.250303766707169, | |
| "grad_norm": 32.30436806042553, | |
| "learning_rate": 6.727019597615545e-05, | |
| "loss": 1.1063, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 1.250303766707169, | |
| "eval_loss": 1.1128273010253906, | |
| "eval_runtime": 53.4998, | |
| "eval_samples_per_second": 13.907, | |
| "eval_steps_per_second": 1.738, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 1.25273390036452, | |
| "grad_norm": 42.104054612880084, | |
| "learning_rate": 6.689085019968669e-05, | |
| "loss": 1.1315, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 1.2551640340218713, | |
| "grad_norm": 25.66097714624212, | |
| "learning_rate": 6.651203878290139e-05, | |
| "loss": 1.0916, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 1.2575941676792224, | |
| "grad_norm": 35.12310576456352, | |
| "learning_rate": 6.613376783956423e-05, | |
| "loss": 1.0699, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 1.2600243013365735, | |
| "grad_norm": 34.172951559594566, | |
| "learning_rate": 6.575604347471695e-05, | |
| "loss": 1.1412, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 1.2624544349939246, | |
| "grad_norm": 54.373563773275116, | |
| "learning_rate": 6.537887178457984e-05, | |
| "loss": 1.1255, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.2648845686512757, | |
| "grad_norm": 33.806385046788755, | |
| "learning_rate": 6.500225885645346e-05, | |
| "loss": 1.101, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 1.267314702308627, | |
| "grad_norm": 34.17813695957543, | |
| "learning_rate": 6.46262107686203e-05, | |
| "loss": 1.1226, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 1.2697448359659782, | |
| "grad_norm": 24.68048087106548, | |
| "learning_rate": 6.425073359024663e-05, | |
| "loss": 1.1787, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 1.2721749696233293, | |
| "grad_norm": 32.78749757697808, | |
| "learning_rate": 6.387583338128471e-05, | |
| "loss": 1.0541, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 1.2746051032806804, | |
| "grad_norm": 30.906673844090044, | |
| "learning_rate": 6.350151619237488e-05, | |
| "loss": 1.0964, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 1.2770352369380316, | |
| "grad_norm": 32.571858392892736, | |
| "learning_rate": 6.312778806474795e-05, | |
| "loss": 1.1251, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 1.2794653705953827, | |
| "grad_norm": 43.02428916532565, | |
| "learning_rate": 6.275465503012751e-05, | |
| "loss": 1.0473, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 1.2818955042527338, | |
| "grad_norm": 60.93587506764561, | |
| "learning_rate": 6.2382123110633e-05, | |
| "loss": 1.078, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 1.2843256379100851, | |
| "grad_norm": 64.6934775930251, | |
| "learning_rate": 6.201019831868208e-05, | |
| "loss": 1.0904, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 1.2867557715674363, | |
| "grad_norm": 32.977077613035426, | |
| "learning_rate": 6.16388866568938e-05, | |
| "loss": 1.0705, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.2891859052247874, | |
| "grad_norm": 28.27407310492513, | |
| "learning_rate": 6.126819411799175e-05, | |
| "loss": 1.1252, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 1.2916160388821385, | |
| "grad_norm": 33.73515826089828, | |
| "learning_rate": 6.0898126684707265e-05, | |
| "loss": 1.1262, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 1.2940461725394896, | |
| "grad_norm": 25.370361818959903, | |
| "learning_rate": 6.052869032968285e-05, | |
| "loss": 1.0845, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 1.296476306196841, | |
| "grad_norm": 37.389287060597105, | |
| "learning_rate": 6.015989101537586e-05, | |
| "loss": 1.1352, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 1.2989064398541919, | |
| "grad_norm": 39.04755104008223, | |
| "learning_rate": 5.979173469396227e-05, | |
| "loss": 1.1538, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 1.3013365735115432, | |
| "grad_norm": 34.33676719612293, | |
| "learning_rate": 5.9424227307240554e-05, | |
| "loss": 1.1725, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 1.3037667071688943, | |
| "grad_norm": 64.66076997769457, | |
| "learning_rate": 5.905737478653572e-05, | |
| "loss": 1.1146, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 1.3061968408262454, | |
| "grad_norm": 48.043289790386325, | |
| "learning_rate": 5.8691183052603834e-05, | |
| "loss": 1.1035, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 1.3086269744835966, | |
| "grad_norm": 49.08397341659928, | |
| "learning_rate": 5.83256580155362e-05, | |
| "loss": 1.1653, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 1.3110571081409477, | |
| "grad_norm": 46.688886812303515, | |
| "learning_rate": 5.796080557466406e-05, | |
| "loss": 1.1328, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.313487241798299, | |
| "grad_norm": 27.503882325413493, | |
| "learning_rate": 5.7596631618463514e-05, | |
| "loss": 1.1019, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 1.3159173754556501, | |
| "grad_norm": 48.88974129574653, | |
| "learning_rate": 5.723314202446026e-05, | |
| "loss": 1.121, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 1.3183475091130012, | |
| "grad_norm": 28.105881157995345, | |
| "learning_rate": 5.687034265913485e-05, | |
| "loss": 1.0898, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 1.3207776427703524, | |
| "grad_norm": 30.410731278414804, | |
| "learning_rate": 5.6508239377828034e-05, | |
| "loss": 1.07, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 1.3232077764277035, | |
| "grad_norm": 38.08324176765882, | |
| "learning_rate": 5.614683802464631e-05, | |
| "loss": 1.1503, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 1.3256379100850546, | |
| "grad_norm": 46.28952293745534, | |
| "learning_rate": 5.578614443236738e-05, | |
| "loss": 1.1282, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 1.3280680437424057, | |
| "grad_norm": 68.2597453597135, | |
| "learning_rate": 5.542616442234618e-05, | |
| "loss": 1.1373, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 1.330498177399757, | |
| "grad_norm": 30.351663825014143, | |
| "learning_rate": 5.5066903804421025e-05, | |
| "loss": 1.1633, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 1.3329283110571082, | |
| "grad_norm": 38.2711285636887, | |
| "learning_rate": 5.470836837681954e-05, | |
| "loss": 1.1604, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 1.3353584447144593, | |
| "grad_norm": 35.64230091531108, | |
| "learning_rate": 5.4350563926065404e-05, | |
| "loss": 1.0564, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.3377885783718104, | |
| "grad_norm": 44.869816046925564, | |
| "learning_rate": 5.399349622688479e-05, | |
| "loss": 1.1376, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 1.3402187120291615, | |
| "grad_norm": 26.681037126315633, | |
| "learning_rate": 5.3637171042113146e-05, | |
| "loss": 1.0867, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 1.3426488456865129, | |
| "grad_norm": 34.6124686262535, | |
| "learning_rate": 5.32815941226022e-05, | |
| "loss": 1.0474, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 1.3450789793438638, | |
| "grad_norm": 35.92639009060983, | |
| "learning_rate": 5.2926771207127254e-05, | |
| "loss": 1.0958, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 1.3475091130012151, | |
| "grad_norm": 39.08938922562224, | |
| "learning_rate": 5.2572708022294504e-05, | |
| "loss": 1.074, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 1.3499392466585662, | |
| "grad_norm": 76.06708166273745, | |
| "learning_rate": 5.2219410282448514e-05, | |
| "loss": 1.0865, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 1.3523693803159174, | |
| "grad_norm": 74.14222265654887, | |
| "learning_rate": 5.1866883689580056e-05, | |
| "loss": 1.1567, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 1.3547995139732685, | |
| "grad_norm": 34.82441678662901, | |
| "learning_rate": 5.151513393323426e-05, | |
| "loss": 1.0802, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 1.3572296476306196, | |
| "grad_norm": 75.53504846566143, | |
| "learning_rate": 5.116416669041843e-05, | |
| "loss": 1.0623, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 1.359659781287971, | |
| "grad_norm": 29.423475817434785, | |
| "learning_rate": 5.0813987625510775e-05, | |
| "loss": 1.077, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.362089914945322, | |
| "grad_norm": 44.607486168434534, | |
| "learning_rate": 5.046460239016879e-05, | |
| "loss": 1.096, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 1.3645200486026732, | |
| "grad_norm": 40.684125033315404, | |
| "learning_rate": 5.011601662323807e-05, | |
| "loss": 1.148, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 1.3669501822600243, | |
| "grad_norm": 47.33103026318705, | |
| "learning_rate": 4.976823595066128e-05, | |
| "loss": 1.1712, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 1.3693803159173754, | |
| "grad_norm": 51.17017845058186, | |
| "learning_rate": 4.9421265985387476e-05, | |
| "loss": 1.1287, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 1.3718104495747265, | |
| "grad_norm": 50.76665552103517, | |
| "learning_rate": 4.907511232728145e-05, | |
| "loss": 1.1156, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 1.3742405832320777, | |
| "grad_norm": 32.6007633025874, | |
| "learning_rate": 4.872978056303327e-05, | |
| "loss": 1.1477, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 1.376670716889429, | |
| "grad_norm": 29.696241441710107, | |
| "learning_rate": 4.8385276266068146e-05, | |
| "loss": 1.0874, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 1.37910085054678, | |
| "grad_norm": 58.96613500379004, | |
| "learning_rate": 4.804160499645667e-05, | |
| "loss": 1.0616, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 1.3815309842041312, | |
| "grad_norm": 37.104100020310334, | |
| "learning_rate": 4.7698772300824756e-05, | |
| "loss": 1.0878, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 1.3839611178614823, | |
| "grad_norm": 51.735902941979305, | |
| "learning_rate": 4.735678371226441e-05, | |
| "loss": 1.0836, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.3863912515188335, | |
| "grad_norm": 55.49190976804079, | |
| "learning_rate": 4.7015644750244306e-05, | |
| "loss": 1.0473, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 1.3888213851761848, | |
| "grad_norm": 34.27972449829039, | |
| "learning_rate": 4.6675360920520625e-05, | |
| "loss": 1.0723, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 1.391251518833536, | |
| "grad_norm": 28.508157856527724, | |
| "learning_rate": 4.6335937715048306e-05, | |
| "loss": 1.0723, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 1.393681652490887, | |
| "grad_norm": 106.84009565003795, | |
| "learning_rate": 4.599738061189244e-05, | |
| "loss": 1.149, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 1.3961117861482382, | |
| "grad_norm": 50.543394606036294, | |
| "learning_rate": 4.565969507513981e-05, | |
| "loss": 1.0991, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 1.3985419198055893, | |
| "grad_norm": 30.409124335052745, | |
| "learning_rate": 4.532288655481062e-05, | |
| "loss": 1.1157, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 1.4009720534629404, | |
| "grad_norm": 89.92061876679301, | |
| "learning_rate": 4.498696048677059e-05, | |
| "loss": 1.1526, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 1.4034021871202915, | |
| "grad_norm": 84.27775422110602, | |
| "learning_rate": 4.465192229264337e-05, | |
| "loss": 1.1418, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 1.4058323207776429, | |
| "grad_norm": 40.7815489623743, | |
| "learning_rate": 4.4317777379722866e-05, | |
| "loss": 1.0831, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 1.408262454434994, | |
| "grad_norm": 66.6911504313278, | |
| "learning_rate": 4.3984531140885943e-05, | |
| "loss": 1.1088, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.410692588092345, | |
| "grad_norm": 137.00882181835217, | |
| "learning_rate": 4.365218895450558e-05, | |
| "loss": 1.1089, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 1.4131227217496962, | |
| "grad_norm": 41.139168895296855, | |
| "learning_rate": 4.332075618436386e-05, | |
| "loss": 1.1603, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 1.4155528554070473, | |
| "grad_norm": 35.443969765428506, | |
| "learning_rate": 4.29902381795655e-05, | |
| "loss": 1.0301, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 1.4179829890643987, | |
| "grad_norm": 32.931514576694674, | |
| "learning_rate": 4.266064027445155e-05, | |
| "loss": 1.1016, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 1.4204131227217496, | |
| "grad_norm": 64.21015694858382, | |
| "learning_rate": 4.2331967788513295e-05, | |
| "loss": 1.0789, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 1.422843256379101, | |
| "grad_norm": 84.13251752827094, | |
| "learning_rate": 4.200422602630629e-05, | |
| "loss": 1.1573, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 1.425273390036452, | |
| "grad_norm": 53.61636603108024, | |
| "learning_rate": 4.167742027736482e-05, | |
| "loss": 1.0942, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 1.4277035236938032, | |
| "grad_norm": 133.20877569415256, | |
| "learning_rate": 4.135155581611661e-05, | |
| "loss": 1.0877, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 1.4301336573511543, | |
| "grad_norm": 49.85736467319357, | |
| "learning_rate": 4.102663790179764e-05, | |
| "loss": 1.0619, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 1.4325637910085054, | |
| "grad_norm": 91.13217639524017, | |
| "learning_rate": 4.070267177836712e-05, | |
| "loss": 1.1093, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.4349939246658567, | |
| "grad_norm": 49.25558128250457, | |
| "learning_rate": 4.037966267442315e-05, | |
| "loss": 1.1344, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 1.4374240583232079, | |
| "grad_norm": 95.87244356130316, | |
| "learning_rate": 4.005761580311805e-05, | |
| "loss": 1.0929, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 1.439854191980559, | |
| "grad_norm": 74.28903671045653, | |
| "learning_rate": 3.973653636207437e-05, | |
| "loss": 1.1263, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 1.44228432563791, | |
| "grad_norm": 53.99454529785116, | |
| "learning_rate": 3.941642953330103e-05, | |
| "loss": 1.0916, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 1.4447144592952612, | |
| "grad_norm": 113.26015597338959, | |
| "learning_rate": 3.909730048310962e-05, | |
| "loss": 1.1009, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 1.4471445929526123, | |
| "grad_norm": 134.4015550981493, | |
| "learning_rate": 3.8779154362030986e-05, | |
| "loss": 1.1351, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 1.4495747266099634, | |
| "grad_norm": 90.61611981238187, | |
| "learning_rate": 3.846199630473216e-05, | |
| "loss": 1.0827, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 1.4520048602673148, | |
| "grad_norm": 56.55050791518521, | |
| "learning_rate": 3.814583142993352e-05, | |
| "loss": 1.1145, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 1.454434993924666, | |
| "grad_norm": 265.6916535243014, | |
| "learning_rate": 3.7830664840326145e-05, | |
| "loss": 1.1459, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 1.456865127582017, | |
| "grad_norm": 72.81191101030372, | |
| "learning_rate": 3.7516501622489367e-05, | |
| "loss": 1.0903, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.4592952612393681, | |
| "grad_norm": 58.309143549086556, | |
| "learning_rate": 3.720334684680889e-05, | |
| "loss": 1.1041, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 1.4617253948967193, | |
| "grad_norm": 35.19205741792398, | |
| "learning_rate": 3.689120556739475e-05, | |
| "loss": 1.1523, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 1.4641555285540706, | |
| "grad_norm": 88.97226951757321, | |
| "learning_rate": 3.6580082821999786e-05, | |
| "loss": 1.1117, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 1.4665856622114215, | |
| "grad_norm": 64.50873879301322, | |
| "learning_rate": 3.6269983631938475e-05, | |
| "loss": 1.1256, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 1.4690157958687728, | |
| "grad_norm": 78.10556611104111, | |
| "learning_rate": 3.596091300200578e-05, | |
| "loss": 1.0834, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 1.471445929526124, | |
| "grad_norm": 69.38449946362529, | |
| "learning_rate": 3.565287592039628e-05, | |
| "loss": 1.1026, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 1.473876063183475, | |
| "grad_norm": 79.60241521456905, | |
| "learning_rate": 3.534587735862391e-05, | |
| "loss": 1.0456, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 1.4763061968408262, | |
| "grad_norm": 89.68581306071424, | |
| "learning_rate": 3.503992227144147e-05, | |
| "loss": 1.0809, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 1.4787363304981773, | |
| "grad_norm": 68.570527237558, | |
| "learning_rate": 3.473501559676088e-05, | |
| "loss": 1.0754, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 1.4811664641555287, | |
| "grad_norm": 54.94762317625427, | |
| "learning_rate": 3.4431162255573245e-05, | |
| "loss": 1.1751, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.4835965978128798, | |
| "grad_norm": 109.12821602719706, | |
| "learning_rate": 3.4128367151869714e-05, | |
| "loss": 1.1055, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 1.486026731470231, | |
| "grad_norm": 198.79030469542352, | |
| "learning_rate": 3.3826635172562094e-05, | |
| "loss": 1.1369, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 1.488456865127582, | |
| "grad_norm": 62.002866716809, | |
| "learning_rate": 3.352597118740404e-05, | |
| "loss": 1.1611, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 1.4908869987849331, | |
| "grad_norm": 79.21193137029579, | |
| "learning_rate": 3.3226380048912585e-05, | |
| "loss": 1.1688, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 1.4933171324422843, | |
| "grad_norm": 68.6722934326242, | |
| "learning_rate": 3.292786659228973e-05, | |
| "loss": 1.1248, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 1.4957472660996354, | |
| "grad_norm": 104.34122241838278, | |
| "learning_rate": 3.263043563534428e-05, | |
| "loss": 1.1425, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 1.4981773997569867, | |
| "grad_norm": 86.43862038340298, | |
| "learning_rate": 3.233409197841437e-05, | |
| "loss": 1.0562, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 1.5006075334143378, | |
| "grad_norm": 79.74137751394451, | |
| "learning_rate": 3.2038840404289705e-05, | |
| "loss": 1.1214, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 1.5006075334143378, | |
| "eval_loss": 1.1088899374008179, | |
| "eval_runtime": 53.0545, | |
| "eval_samples_per_second": 14.023, | |
| "eval_steps_per_second": 1.753, | |
| "step": 618 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 822, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 206, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 9.157723878347244e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |