{ "best_metric": null, "best_model_checkpoint": null, "epoch": 1.2, "eval_steps": 500, "global_step": 15000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.002105263157894737, "grad_norm": 8.761321067810059, "learning_rate": 5e-05, "loss": 17.009, "step": 25 }, { "epoch": 0.004210526315789474, "grad_norm": 4.736087799072266, "learning_rate": 0.0001, "loss": 4.2169, "step": 50 }, { "epoch": 0.00631578947368421, "grad_norm": 2.9459564685821533, "learning_rate": 0.00015000000000000001, "loss": 2.3088, "step": 75 }, { "epoch": 0.008421052631578947, "grad_norm": 1.7206273078918457, "learning_rate": 0.0002, "loss": 2.1905, "step": 100 }, { "epoch": 0.010526315789473684, "grad_norm": 1.557580590248108, "learning_rate": 0.0001999986107633737, "loss": 2.1333, "step": 125 }, { "epoch": 0.01263157894736842, "grad_norm": 1.3003801107406616, "learning_rate": 0.00019999444309209432, "loss": 2.0932, "step": 150 }, { "epoch": 0.014736842105263158, "grad_norm": 1.401289463043213, "learning_rate": 0.00019998749710195955, "loss": 2.0548, "step": 175 }, { "epoch": 0.016842105263157894, "grad_norm": 1.3614596128463745, "learning_rate": 0.0001999777729859618, "loss": 2.0482, "step": 200 }, { "epoch": 0.018947368421052633, "grad_norm": 2.883526086807251, "learning_rate": 0.00019996527101428312, "loss": 2.0255, "step": 225 }, { "epoch": 0.021052631578947368, "grad_norm": 1.0736247301101685, "learning_rate": 0.00019994999153428737, "loss": 2.0278, "step": 250 }, { "epoch": 0.023157894736842106, "grad_norm": 0.910820484161377, "learning_rate": 0.00019993193497051082, "loss": 2.0147, "step": 275 }, { "epoch": 0.02526315789473684, "grad_norm": 1.1746225357055664, "learning_rate": 0.00019991110182465032, "loss": 1.9864, "step": 300 }, { "epoch": 0.02736842105263158, "grad_norm": 1.0454574823379517, "learning_rate": 0.0001998874926755492, "loss": 1.9919, "step": 325 }, { "epoch": 0.029473684210526315, "grad_norm": 0.9762787818908691, "learning_rate": 0.0001998611081791814, "loss": 1.9855, "step": 350 }, { "epoch": 0.031578947368421054, "grad_norm": 0.7796763181686401, "learning_rate": 0.00019983194906863304, "loss": 2.0007, "step": 375 }, { "epoch": 0.03368421052631579, "grad_norm": 0.6890770196914673, "learning_rate": 0.00019980001615408228, "loss": 1.97, "step": 400 }, { "epoch": 0.035789473684210524, "grad_norm": 1.4911786317825317, "learning_rate": 0.00019976531032277653, "loss": 1.9665, "step": 425 }, { "epoch": 0.037894736842105266, "grad_norm": 0.7319967150688171, "learning_rate": 0.00019972783253900808, "loss": 1.9672, "step": 450 }, { "epoch": 0.04, "grad_norm": 0.812010645866394, "learning_rate": 0.00019968758384408713, "loss": 1.9723, "step": 475 }, { "epoch": 0.042105263157894736, "grad_norm": 1.4215325117111206, "learning_rate": 0.00019964456535631286, "loss": 1.9528, "step": 500 }, { "epoch": 0.042105263157894736, "eval_loss": 0.12439242005348206, "eval_runtime": 16285.4101, "eval_samples_per_second": 0.614, "eval_steps_per_second": 0.077, "step": 500 }, { "epoch": 0.04421052631578947, "grad_norm": 0.7200049161911011, "learning_rate": 0.00019959877827094248, "loss": 1.9649, "step": 525 }, { "epoch": 0.04631578947368421, "grad_norm": 0.9672224521636963, "learning_rate": 0.00019955022386015792, "loss": 1.958, "step": 550 }, { "epoch": 0.04842105263157895, "grad_norm": 0.7544946074485779, "learning_rate": 0.00019949890347303046, "loss": 1.9506, "step": 575 }, { "epoch": 0.05052631578947368, "grad_norm": 0.7363459467887878, "learning_rate": 0.00019944481853548335, "loss": 1.9555, "step": 600 }, { "epoch": 0.05263157894736842, "grad_norm": 0.7254467010498047, "learning_rate": 0.00019938797055025213, "loss": 1.9582, "step": 625 }, { "epoch": 0.05473684210526316, "grad_norm": 0.7360565066337585, "learning_rate": 0.00019932836109684286, "loss": 1.9504, "step": 650 }, { "epoch": 0.056842105263157895, "grad_norm": 0.7283539175987244, "learning_rate": 0.0001992659918314882, "loss": 1.9413, "step": 675 }, { "epoch": 0.05894736842105263, "grad_norm": 0.7833098769187927, "learning_rate": 0.0001992008644871016, "loss": 1.9439, "step": 700 }, { "epoch": 0.061052631578947365, "grad_norm": 0.784247100353241, "learning_rate": 0.00019913298087322883, "loss": 1.9342, "step": 725 }, { "epoch": 0.06315789473684211, "grad_norm": 0.8815144300460815, "learning_rate": 0.00019906234287599798, "loss": 1.9376, "step": 750 }, { "epoch": 0.06526315789473684, "grad_norm": 0.6176999807357788, "learning_rate": 0.0001989889524580669, "loss": 1.9412, "step": 775 }, { "epoch": 0.06736842105263158, "grad_norm": 0.7160853743553162, "learning_rate": 0.00019891281165856873, "loss": 1.9415, "step": 800 }, { "epoch": 0.06947368421052631, "grad_norm": 3.323953151702881, "learning_rate": 0.0001988339225930552, "loss": 1.944, "step": 825 }, { "epoch": 0.07157894736842105, "grad_norm": 0.6912685632705688, "learning_rate": 0.00019875228745343794, "loss": 1.9477, "step": 850 }, { "epoch": 0.07368421052631578, "grad_norm": 0.9157281517982483, "learning_rate": 0.0001986679085079274, "loss": 1.9345, "step": 875 }, { "epoch": 0.07578947368421053, "grad_norm": 0.6035289168357849, "learning_rate": 0.00019858078810097002, "loss": 1.941, "step": 900 }, { "epoch": 0.07789473684210527, "grad_norm": 0.8801007866859436, "learning_rate": 0.00019849092865318308, "loss": 1.9467, "step": 925 }, { "epoch": 0.08, "grad_norm": 0.6666975021362305, "learning_rate": 0.00019839833266128724, "loss": 1.9375, "step": 950 }, { "epoch": 0.08210526315789474, "grad_norm": 0.669546902179718, "learning_rate": 0.0001983030026980374, "loss": 1.9342, "step": 975 }, { "epoch": 0.08421052631578947, "grad_norm": 1.0217283964157104, "learning_rate": 0.00019820494141215104, "loss": 1.9294, "step": 1000 }, { "epoch": 0.08421052631578947, "eval_loss": 0.12046071887016296, "eval_runtime": 17872.694, "eval_samples_per_second": 0.56, "eval_steps_per_second": 0.07, "step": 1000 }, { "epoch": 0.082, "grad_norm": 0.8858544230461121, "learning_rate": 0.00019810415152823478, "loss": 1.9254, "step": 1025 }, { "epoch": 0.084, "grad_norm": 0.8588420748710632, "learning_rate": 0.00019800063584670863, "loss": 1.9213, "step": 1050 }, { "epoch": 0.086, "grad_norm": 0.7548304200172424, "learning_rate": 0.00019789439724372806, "loss": 1.9317, "step": 1075 }, { "epoch": 0.088, "grad_norm": 0.6778808832168579, "learning_rate": 0.00019778543867110426, "loss": 1.93, "step": 1100 }, { "epoch": 0.09, "grad_norm": 1.0773042440414429, "learning_rate": 0.00019767376315622204, "loss": 1.9296, "step": 1125 }, { "epoch": 0.092, "grad_norm": 0.5895248651504517, "learning_rate": 0.00019755937380195568, "loss": 1.9372, "step": 1150 }, { "epoch": 0.094, "grad_norm": 0.5954482555389404, "learning_rate": 0.00019744227378658283, "loss": 1.9209, "step": 1175 }, { "epoch": 0.096, "grad_norm": 0.5461156964302063, "learning_rate": 0.00019732246636369605, "loss": 1.9194, "step": 1200 }, { "epoch": 0.098, "grad_norm": 0.7127408385276794, "learning_rate": 0.0001971999548621126, "loss": 1.9103, "step": 1225 }, { "epoch": 0.1, "grad_norm": 0.9995592832565308, "learning_rate": 0.0001970747426857817, "loss": 1.9091, "step": 1250 }, { "epoch": 0.102, "grad_norm": 0.6391471028327942, "learning_rate": 0.00019694683331369022, "loss": 1.9106, "step": 1275 }, { "epoch": 0.104, "grad_norm": 0.626804769039154, "learning_rate": 0.00019681623029976588, "loss": 1.9014, "step": 1300 }, { "epoch": 0.106, "grad_norm": 0.777351975440979, "learning_rate": 0.00019668293727277846, "loss": 1.9091, "step": 1325 }, { "epoch": 0.108, "grad_norm": 0.7873269319534302, "learning_rate": 0.00019654695793623907, "loss": 1.9143, "step": 1350 }, { "epoch": 0.11, "grad_norm": 0.5738505721092224, "learning_rate": 0.00019640829606829723, "loss": 1.9013, "step": 1375 }, { "epoch": 0.112, "grad_norm": 0.7090626955032349, "learning_rate": 0.00019626695552163578, "loss": 1.9121, "step": 1400 }, { "epoch": 0.114, "grad_norm": 0.622789204120636, "learning_rate": 0.00019612294022336405, "loss": 1.8961, "step": 1425 }, { "epoch": 0.116, "grad_norm": 0.6606855988502502, "learning_rate": 0.0001959762541749086, "loss": 1.9149, "step": 1450 }, { "epoch": 0.118, "grad_norm": 0.6521698236465454, "learning_rate": 0.00019582690145190202, "loss": 1.9001, "step": 1475 }, { "epoch": 0.12, "grad_norm": 0.8047705888748169, "learning_rate": 0.00019567488620406983, "loss": 1.8939, "step": 1500 }, { "epoch": 0.122, "grad_norm": 0.7987963557243347, "learning_rate": 0.0001955202126551149, "loss": 1.8963, "step": 1525 }, { "epoch": 0.124, "grad_norm": 0.6272624135017395, "learning_rate": 0.00019536288510260056, "loss": 1.8829, "step": 1550 }, { "epoch": 0.126, "grad_norm": 0.7199199795722961, "learning_rate": 0.0001952029079178307, "loss": 1.8914, "step": 1575 }, { "epoch": 0.128, "grad_norm": 0.6892108917236328, "learning_rate": 0.00019504028554572864, "loss": 1.8957, "step": 1600 }, { "epoch": 0.13, "grad_norm": 0.7720304727554321, "learning_rate": 0.00019487502250471349, "loss": 1.8855, "step": 1625 }, { "epoch": 0.132, "grad_norm": 0.5629010200500488, "learning_rate": 0.00019470712338657458, "loss": 1.8885, "step": 1650 }, { "epoch": 0.134, "grad_norm": 0.8249186277389526, "learning_rate": 0.00019453659285634408, "loss": 1.9002, "step": 1675 }, { "epoch": 0.136, "grad_norm": 0.720421552658081, "learning_rate": 0.00019436343565216711, "loss": 1.9029, "step": 1700 }, { "epoch": 0.138, "grad_norm": 0.8002966046333313, "learning_rate": 0.0001941876565851703, "loss": 1.8909, "step": 1725 }, { "epoch": 0.14, "grad_norm": 0.60608971118927, "learning_rate": 0.000194009260539328, "loss": 1.8948, "step": 1750 }, { "epoch": 0.142, "grad_norm": 8.092451095581055, "learning_rate": 0.00019382825247132664, "loss": 1.9321, "step": 1775 }, { "epoch": 0.144, "grad_norm": 1.3757332563400269, "learning_rate": 0.00019364463741042694, "loss": 1.9557, "step": 1800 }, { "epoch": 0.146, "grad_norm": 1.8562349081039429, "learning_rate": 0.00019345842045832428, "loss": 1.9022, "step": 1825 }, { "epoch": 0.148, "grad_norm": 0.6850736141204834, "learning_rate": 0.00019326960678900688, "loss": 1.8987, "step": 1850 }, { "epoch": 0.15, "grad_norm": 0.7720704674720764, "learning_rate": 0.000193078201648612, "loss": 1.9035, "step": 1875 }, { "epoch": 0.152, "grad_norm": 1.3317339420318604, "learning_rate": 0.00019288421035528028, "loss": 2.0192, "step": 1900 }, { "epoch": 0.154, "grad_norm": 1.4454379081726074, "learning_rate": 0.00019268763829900798, "loss": 2.7565, "step": 1925 }, { "epoch": 0.156, "grad_norm": 0.7154741883277893, "learning_rate": 0.000192488490941497, "loss": 1.9095, "step": 1950 }, { "epoch": 0.158, "grad_norm": 2.4882631301879883, "learning_rate": 0.0001922867738160035, "loss": 2.3094, "step": 1975 }, { "epoch": 0.16, "grad_norm": 0.854193389415741, "learning_rate": 0.0001920824925271838, "loss": 1.9013, "step": 2000 }, { "epoch": 0.162, "grad_norm": 3.4987711906433105, "learning_rate": 0.00019187565275093887, "loss": 6.8561, "step": 2025 }, { "epoch": 0.164, "grad_norm": 252.18682861328125, "learning_rate": 0.00019166626023425662, "loss": 3.4574, "step": 2050 }, { "epoch": 0.166, "grad_norm": 0.9407473802566528, "learning_rate": 0.00019145432079505206, "loss": 2.3099, "step": 2075 }, { "epoch": 0.168, "grad_norm": 0.5535870790481567, "learning_rate": 0.00019123984032200583, "loss": 1.9008, "step": 2100 }, { "epoch": 0.17, "grad_norm": 2.9849846363067627, "learning_rate": 0.00019102282477440055, "loss": 2.0312, "step": 2125 }, { "epoch": 0.172, "grad_norm": 0.5983627438545227, "learning_rate": 0.00019080328018195513, "loss": 1.8876, "step": 2150 }, { "epoch": 0.174, "grad_norm": 0.6499072313308716, "learning_rate": 0.00019058121264465733, "loss": 1.8892, "step": 2175 }, { "epoch": 0.176, "grad_norm": 0.6271260380744934, "learning_rate": 0.00019035662833259432, "loss": 1.8766, "step": 2200 }, { "epoch": 0.178, "grad_norm": 2.5810041427612305, "learning_rate": 0.0001901295334857811, "loss": 1.8846, "step": 2225 }, { "epoch": 0.18, "grad_norm": 1.6043792963027954, "learning_rate": 0.00018989993441398729, "loss": 1.909, "step": 2250 }, { "epoch": 0.182, "grad_norm": 18.742921829223633, "learning_rate": 0.00018966783749656163, "loss": 5.2203, "step": 2275 }, { "epoch": 0.184, "grad_norm": 0.6480016708374023, "learning_rate": 0.00018943324918225494, "loss": 1.9187, "step": 2300 }, { "epoch": 0.186, "grad_norm": 0.7129198908805847, "learning_rate": 0.0001891961759890408, "loss": 1.8961, "step": 2325 }, { "epoch": 0.188, "grad_norm": 0.7181056141853333, "learning_rate": 0.00018895662450393438, "loss": 1.8991, "step": 2350 }, { "epoch": 0.19, "grad_norm": 0.5234689712524414, "learning_rate": 0.00018871460138280972, "loss": 1.8838, "step": 2375 }, { "epoch": 0.192, "grad_norm": 0.6447088718414307, "learning_rate": 0.00018847011335021449, "loss": 1.8908, "step": 2400 }, { "epoch": 0.194, "grad_norm": 0.6292424201965332, "learning_rate": 0.0001882231671991832, "loss": 1.8806, "step": 2425 }, { "epoch": 0.196, "grad_norm": 0.6571337580680847, "learning_rate": 0.00018797376979104872, "loss": 1.8874, "step": 2450 }, { "epoch": 0.198, "grad_norm": 0.9664616584777832, "learning_rate": 0.00018772192805525126, "loss": 1.8894, "step": 2475 }, { "epoch": 0.2, "grad_norm": 0.6942924857139587, "learning_rate": 0.0001874676489891461, "loss": 1.8829, "step": 2500 }, { "epoch": 0.202, "grad_norm": 0.5820707678794861, "learning_rate": 0.00018721093965780907, "loss": 1.878, "step": 2525 }, { "epoch": 0.204, "grad_norm": 0.5778231024742126, "learning_rate": 0.00018695180719384029, "loss": 1.8836, "step": 2550 }, { "epoch": 0.206, "grad_norm": 0.7510939836502075, "learning_rate": 0.00018669025879716595, "loss": 1.8869, "step": 2575 }, { "epoch": 0.208, "grad_norm": 0.5207582116127014, "learning_rate": 0.00018642630173483835, "loss": 1.8783, "step": 2600 }, { "epoch": 0.21, "grad_norm": 0.7096635103225708, "learning_rate": 0.00018615994334083378, "loss": 1.8809, "step": 2625 }, { "epoch": 0.212, "grad_norm": 0.642936110496521, "learning_rate": 0.00018589119101584898, "loss": 1.8851, "step": 2650 }, { "epoch": 0.214, "grad_norm": 0.5762498378753662, "learning_rate": 0.00018562005222709548, "loss": 1.9012, "step": 2675 }, { "epoch": 0.216, "grad_norm": 0.688967227935791, "learning_rate": 0.00018534653450809197, "loss": 1.8852, "step": 2700 }, { "epoch": 0.218, "grad_norm": 0.6240432262420654, "learning_rate": 0.00018507064545845515, "loss": 1.8836, "step": 2725 }, { "epoch": 0.22, "grad_norm": 0.6300758719444275, "learning_rate": 0.0001847923927436884, "loss": 1.8809, "step": 2750 }, { "epoch": 0.222, "grad_norm": 0.5509706139564514, "learning_rate": 0.00018451178409496902, "loss": 1.8783, "step": 2775 }, { "epoch": 0.224, "grad_norm": 0.5869603157043457, "learning_rate": 0.0001842288273089332, "loss": 1.8866, "step": 2800 }, { "epoch": 0.226, "grad_norm": 0.5525537729263306, "learning_rate": 0.00018394353024745965, "loss": 1.8757, "step": 2825 }, { "epoch": 0.228, "grad_norm": 0.6274869441986084, "learning_rate": 0.00018365590083745085, "loss": 1.8835, "step": 2850 }, { "epoch": 0.23, "grad_norm": 0.5160146951675415, "learning_rate": 0.00018336594707061307, "loss": 1.8816, "step": 2875 }, { "epoch": 0.232, "grad_norm": 0.5234496593475342, "learning_rate": 0.0001830736770032341, "loss": 1.8766, "step": 2900 }, { "epoch": 0.234, "grad_norm": 0.5816289782524109, "learning_rate": 0.00018277909875595967, "loss": 1.8781, "step": 2925 }, { "epoch": 0.236, "grad_norm": 0.5350304841995239, "learning_rate": 0.00018248222051356754, "loss": 1.8912, "step": 2950 }, { "epoch": 0.238, "grad_norm": 0.586509108543396, "learning_rate": 0.00018218305052474025, "loss": 1.8708, "step": 2975 }, { "epoch": 0.24, "grad_norm": 2.095522165298462, "learning_rate": 0.00018188159710183594, "loss": 1.881, "step": 3000 }, { "epoch": 0.242, "grad_norm": 0.6511578559875488, "learning_rate": 0.00018157786862065732, "loss": 1.8784, "step": 3025 }, { "epoch": 0.244, "grad_norm": 0.5500091910362244, "learning_rate": 0.00018127187352021907, "loss": 1.8864, "step": 3050 }, { "epoch": 0.246, "grad_norm": 0.7176497578620911, "learning_rate": 0.00018096362030251313, "loss": 1.8868, "step": 3075 }, { "epoch": 0.248, "grad_norm": 0.5715402960777283, "learning_rate": 0.00018065311753227273, "loss": 1.8763, "step": 3100 }, { "epoch": 0.25, "grad_norm": 0.5016639232635498, "learning_rate": 0.00018034037383673427, "loss": 1.8779, "step": 3125 }, { "epoch": 0.252, "grad_norm": 0.723649263381958, "learning_rate": 0.00018002539790539773, "loss": 1.8892, "step": 3150 }, { "epoch": 0.254, "grad_norm": 0.7148360013961792, "learning_rate": 0.00017970819848978504, "loss": 1.8798, "step": 3175 }, { "epoch": 0.256, "grad_norm": 0.8507091403007507, "learning_rate": 0.0001793887844031972, "loss": 1.8744, "step": 3200 }, { "epoch": 0.258, "grad_norm": 0.6568000912666321, "learning_rate": 0.00017906716452046915, "loss": 1.8853, "step": 3225 }, { "epoch": 0.26, "grad_norm": 0.5410321950912476, "learning_rate": 0.00017874334777772327, "loss": 1.8775, "step": 3250 }, { "epoch": 0.262, "grad_norm": 0.8745731115341187, "learning_rate": 0.00017841734317212117, "loss": 1.8797, "step": 3275 }, { "epoch": 0.264, "grad_norm": 0.5649206042289734, "learning_rate": 0.00017808915976161362, "loss": 1.8835, "step": 3300 }, { "epoch": 0.266, "grad_norm": 0.6354301571846008, "learning_rate": 0.0001777588066646889, "loss": 1.8703, "step": 3325 }, { "epoch": 0.268, "grad_norm": 0.5484392642974854, "learning_rate": 0.00017742629306011944, "loss": 1.8771, "step": 3350 }, { "epoch": 0.27, "grad_norm": 0.6123652458190918, "learning_rate": 0.00017709162818670682, "loss": 1.8727, "step": 3375 }, { "epoch": 0.272, "grad_norm": 0.5673122406005859, "learning_rate": 0.000176754821343025, "loss": 1.8686, "step": 3400 }, { "epoch": 0.274, "grad_norm": 0.5154988169670105, "learning_rate": 0.0001764158818871621, "loss": 1.8703, "step": 3425 }, { "epoch": 0.276, "grad_norm": 0.5470913648605347, "learning_rate": 0.00017607481923646016, "loss": 1.8686, "step": 3450 }, { "epoch": 0.278, "grad_norm": 0.5259148478507996, "learning_rate": 0.00017573164286725378, "loss": 1.879, "step": 3475 }, { "epoch": 0.28, "grad_norm": 0.6147259473800659, "learning_rate": 0.0001753863623146066, "loss": 1.8762, "step": 3500 }, { "epoch": 0.282, "grad_norm": 0.6801021695137024, "learning_rate": 0.0001750389871720463, "loss": 1.8784, "step": 3525 }, { "epoch": 0.284, "grad_norm": 0.4376698136329651, "learning_rate": 0.00017468952709129846, "loss": 1.8864, "step": 3550 }, { "epoch": 0.286, "grad_norm": 1.253459095954895, "learning_rate": 0.00017433799178201786, "loss": 1.8709, "step": 3575 }, { "epoch": 0.288, "grad_norm": 0.5602317452430725, "learning_rate": 0.00017398439101151905, "loss": 1.8763, "step": 3600 }, { "epoch": 0.29, "grad_norm": 0.6780624985694885, "learning_rate": 0.0001736287346045049, "loss": 1.8771, "step": 3625 }, { "epoch": 0.292, "grad_norm": 0.5530239939689636, "learning_rate": 0.00017327103244279348, "loss": 1.867, "step": 3650 }, { "epoch": 0.294, "grad_norm": 0.6352905035018921, "learning_rate": 0.00017291129446504375, "loss": 1.8747, "step": 3675 }, { "epoch": 0.296, "grad_norm": 0.5354923605918884, "learning_rate": 0.00017254953066647913, "loss": 1.8669, "step": 3700 }, { "epoch": 0.298, "grad_norm": 0.5250069499015808, "learning_rate": 0.00017218575109861008, "loss": 1.8668, "step": 3725 }, { "epoch": 0.3, "grad_norm": 0.5072012543678284, "learning_rate": 0.00017181996586895454, "loss": 1.8638, "step": 3750 }, { "epoch": 0.302, "grad_norm": 0.546538770198822, "learning_rate": 0.0001714521851407573, "loss": 1.8765, "step": 3775 }, { "epoch": 0.304, "grad_norm": 0.5139479637145996, "learning_rate": 0.0001710824191327075, "loss": 1.87, "step": 3800 }, { "epoch": 0.306, "grad_norm": 0.5560875535011292, "learning_rate": 0.00017071067811865476, "loss": 1.8714, "step": 3825 }, { "epoch": 0.308, "grad_norm": 0.5358397960662842, "learning_rate": 0.00017033697242732377, "loss": 1.8682, "step": 3850 }, { "epoch": 0.31, "grad_norm": 0.6141547560691833, "learning_rate": 0.0001699613124420272, "loss": 1.8617, "step": 3875 }, { "epoch": 0.312, "grad_norm": 0.5563941597938538, "learning_rate": 0.00016958370860037717, "loss": 1.8737, "step": 3900 }, { "epoch": 0.314, "grad_norm": 0.53114914894104, "learning_rate": 0.00016920417139399557, "loss": 1.8747, "step": 3925 }, { "epoch": 0.316, "grad_norm": 0.5394594669342041, "learning_rate": 0.00016882271136822206, "loss": 1.8738, "step": 3950 }, { "epoch": 0.318, "grad_norm": 0.4669198989868164, "learning_rate": 0.0001684393391218215, "loss": 1.8729, "step": 3975 }, { "epoch": 0.32, "grad_norm": 0.49989891052246094, "learning_rate": 0.0001680540653066891, "loss": 1.8694, "step": 4000 }, { "epoch": 0.322, "grad_norm": 0.5510935187339783, "learning_rate": 0.0001676669006275549, "loss": 1.8678, "step": 4025 }, { "epoch": 0.324, "grad_norm": 0.5259726047515869, "learning_rate": 0.00016727785584168581, "loss": 1.8667, "step": 4050 }, { "epoch": 0.326, "grad_norm": 0.7273874878883362, "learning_rate": 0.0001668869417585873, "loss": 1.8685, "step": 4075 }, { "epoch": 0.328, "grad_norm": 0.5754939317703247, "learning_rate": 0.0001664941692397025, "loss": 1.8628, "step": 4100 }, { "epoch": 0.33, "grad_norm": 0.5599279999732971, "learning_rate": 0.00016609954919811077, "loss": 1.8624, "step": 4125 }, { "epoch": 0.332, "grad_norm": 0.7220540642738342, "learning_rate": 0.00016570309259822453, "loss": 1.8677, "step": 4150 }, { "epoch": 0.334, "grad_norm": 0.5246766209602356, "learning_rate": 0.00016530481045548426, "loss": 1.8674, "step": 4175 }, { "epoch": 0.336, "grad_norm": 0.4979128837585449, "learning_rate": 0.00016490471383605288, "loss": 1.9051, "step": 4200 }, { "epoch": 0.338, "grad_norm": 0.5102583169937134, "learning_rate": 0.0001645028138565078, "loss": 1.8563, "step": 4225 }, { "epoch": 0.34, "grad_norm": 0.6284440159797668, "learning_rate": 0.0001640991216835326, "loss": 1.8597, "step": 4250 }, { "epoch": 0.342, "grad_norm": 0.548886775970459, "learning_rate": 0.0001636936485336062, "loss": 1.8666, "step": 4275 }, { "epoch": 0.344, "grad_norm": 0.5574102997779846, "learning_rate": 0.0001632864056726917, "loss": 1.8651, "step": 4300 }, { "epoch": 0.346, "grad_norm": 0.6083822846412659, "learning_rate": 0.000162877404415923, "loss": 1.8676, "step": 4325 }, { "epoch": 0.348, "grad_norm": 0.48580431938171387, "learning_rate": 0.00016246665612729074, "loss": 1.8696, "step": 4350 }, { "epoch": 0.35, "grad_norm": 0.5158634185791016, "learning_rate": 0.0001620541722193261, "loss": 1.8689, "step": 4375 }, { "epoch": 0.352, "grad_norm": 0.4892460107803345, "learning_rate": 0.00016163996415278424, "loss": 1.8617, "step": 4400 }, { "epoch": 0.354, "grad_norm": 0.597580075263977, "learning_rate": 0.00016122404343632546, "loss": 1.8677, "step": 4425 }, { "epoch": 0.356, "grad_norm": 0.5458176136016846, "learning_rate": 0.00016080642162619565, "loss": 1.8611, "step": 4450 }, { "epoch": 0.358, "grad_norm": 0.5014930963516235, "learning_rate": 0.00016038711032590508, "loss": 1.8671, "step": 4475 }, { "epoch": 0.36, "grad_norm": 0.4904371500015259, "learning_rate": 0.00015996612118590603, "loss": 1.8619, "step": 4500 }, { "epoch": 0.362, "grad_norm": 0.5653460621833801, "learning_rate": 0.00015954346590326925, "loss": 1.863, "step": 4525 }, { "epoch": 0.364, "grad_norm": 0.571080207824707, "learning_rate": 0.00015911915622135862, "loss": 1.8706, "step": 4550 }, { "epoch": 0.366, "grad_norm": 0.4800024628639221, "learning_rate": 0.00015869320392950526, "loss": 1.8595, "step": 4575 }, { "epoch": 0.368, "grad_norm": 0.4732321500778198, "learning_rate": 0.00015826562086267956, "loss": 1.8634, "step": 4600 }, { "epoch": 0.37, "grad_norm": 0.5532209873199463, "learning_rate": 0.00015783641890116274, "loss": 1.8635, "step": 4625 }, { "epoch": 0.372, "grad_norm": 0.5195424556732178, "learning_rate": 0.00015740560997021648, "loss": 1.8621, "step": 4650 }, { "epoch": 0.374, "grad_norm": 0.4964386522769928, "learning_rate": 0.0001569732060397517, "loss": 1.8633, "step": 4675 }, { "epoch": 0.376, "grad_norm": 0.6742044687271118, "learning_rate": 0.00015653921912399589, "loss": 1.8775, "step": 4700 }, { "epoch": 0.378, "grad_norm": 0.528725802898407, "learning_rate": 0.00015610366128115954, "loss": 1.8661, "step": 4725 }, { "epoch": 0.38, "grad_norm": 0.5622186660766602, "learning_rate": 0.0001556665446131007, "loss": 1.8737, "step": 4750 }, { "epoch": 0.382, "grad_norm": 0.4447067975997925, "learning_rate": 0.00015522788126498917, "loss": 1.8704, "step": 4775 }, { "epoch": 0.384, "grad_norm": 0.557609498500824, "learning_rate": 0.0001547876834249687, "loss": 1.8689, "step": 4800 }, { "epoch": 0.386, "grad_norm": 0.8761849403381348, "learning_rate": 0.00015434596332381852, "loss": 1.8815, "step": 4825 }, { "epoch": 0.388, "grad_norm": 1.2129613161087036, "learning_rate": 0.00015390273323461352, "loss": 1.8772, "step": 4850 }, { "epoch": 0.39, "grad_norm": 0.5641235709190369, "learning_rate": 0.00015345800547238316, "loss": 1.8867, "step": 4875 }, { "epoch": 0.392, "grad_norm": 0.5469566583633423, "learning_rate": 0.00015301179239376938, "loss": 1.8831, "step": 4900 }, { "epoch": 0.394, "grad_norm": 0.5969127416610718, "learning_rate": 0.00015256410639668318, "loss": 1.8725, "step": 4925 }, { "epoch": 0.396, "grad_norm": 0.6882519721984863, "learning_rate": 0.00015211495991996027, "loss": 1.8696, "step": 4950 }, { "epoch": 0.398, "grad_norm": 0.44274187088012695, "learning_rate": 0.00015166436544301537, "loss": 1.8708, "step": 4975 }, { "epoch": 0.4, "grad_norm": 0.6739794015884399, "learning_rate": 0.0001512123354854955, "loss": 1.8613, "step": 5000 }, { "epoch": 0.402, "grad_norm": 0.6389496922492981, "learning_rate": 0.00015075888260693214, "loss": 1.8681, "step": 5025 }, { "epoch": 0.404, "grad_norm": 0.6699119806289673, "learning_rate": 0.0001503040194063922, "loss": 1.8692, "step": 5050 }, { "epoch": 0.406, "grad_norm": 0.6882144212722778, "learning_rate": 0.00014984775852212807, "loss": 1.8703, "step": 5075 }, { "epoch": 0.408, "grad_norm": 0.518025815486908, "learning_rate": 0.00014939011263122634, "loss": 1.8691, "step": 5100 }, { "epoch": 0.41, "grad_norm": 0.49707674980163574, "learning_rate": 0.00014893109444925578, "loss": 1.8794, "step": 5125 }, { "epoch": 0.412, "grad_norm": 0.5240812301635742, "learning_rate": 0.00014847071672991367, "loss": 1.8626, "step": 5150 }, { "epoch": 0.414, "grad_norm": 0.40495631098747253, "learning_rate": 0.00014800899226467191, "loss": 1.8574, "step": 5175 }, { "epoch": 0.416, "grad_norm": 0.48023268580436707, "learning_rate": 0.00014754593388242117, "loss": 1.8668, "step": 5200 }, { "epoch": 0.418, "grad_norm": 0.5155997276306152, "learning_rate": 0.00014708155444911484, "loss": 1.8575, "step": 5225 }, { "epoch": 0.42, "grad_norm": 0.6117238402366638, "learning_rate": 0.0001466158668674112, "loss": 1.8621, "step": 5250 }, { "epoch": 0.422, "grad_norm": 0.507888674736023, "learning_rate": 0.00014614888407631516, "loss": 1.8648, "step": 5275 }, { "epoch": 0.424, "grad_norm": 0.4884865880012512, "learning_rate": 0.00014568061905081875, "loss": 1.8641, "step": 5300 }, { "epoch": 0.426, "grad_norm": 0.4816436469554901, "learning_rate": 0.00014521108480154032, "loss": 1.876, "step": 5325 }, { "epoch": 0.428, "grad_norm": 0.5883172750473022, "learning_rate": 0.00014474029437436348, "loss": 1.8651, "step": 5350 }, { "epoch": 0.43, "grad_norm": 0.5438560843467712, "learning_rate": 0.00014426826085007428, "loss": 1.8642, "step": 5375 }, { "epoch": 0.432, "grad_norm": 0.4564906060695648, "learning_rate": 0.00014379499734399798, "loss": 1.8579, "step": 5400 }, { "epoch": 0.434, "grad_norm": 0.5230515003204346, "learning_rate": 0.00014332051700563446, "loss": 1.8665, "step": 5425 }, { "epoch": 0.436, "grad_norm": 0.5528490543365479, "learning_rate": 0.0001428448330182931, "loss": 1.8664, "step": 5450 }, { "epoch": 0.438, "grad_norm": 0.7777835726737976, "learning_rate": 0.00014236795859872615, "loss": 1.8673, "step": 5475 }, { "epoch": 0.44, "grad_norm": 0.6119264364242554, "learning_rate": 0.00014188990699676184, "loss": 1.8662, "step": 5500 }, { "epoch": 0.442, "grad_norm": 0.784015417098999, "learning_rate": 0.00014141069149493612, "loss": 1.8653, "step": 5525 }, { "epoch": 0.444, "grad_norm": 0.43411755561828613, "learning_rate": 0.00014093032540812348, "loss": 1.8661, "step": 5550 }, { "epoch": 0.446, "grad_norm": 0.47867318987846375, "learning_rate": 0.00014044882208316713, "loss": 1.8652, "step": 5575 }, { "epoch": 0.448, "grad_norm": 0.45533016324043274, "learning_rate": 0.00013996619489850822, "loss": 1.8638, "step": 5600 }, { "epoch": 0.45, "grad_norm": 0.6000945568084717, "learning_rate": 0.000139482457263814, "loss": 1.8659, "step": 5625 }, { "epoch": 0.452, "grad_norm": 0.43668854236602783, "learning_rate": 0.00013899762261960518, "loss": 1.8637, "step": 5650 }, { "epoch": 0.454, "grad_norm": 0.5040393471717834, "learning_rate": 0.00013851170443688273, "loss": 1.8627, "step": 5675 }, { "epoch": 0.456, "grad_norm": 0.5880767703056335, "learning_rate": 0.00013802471621675338, "loss": 1.861, "step": 5700 }, { "epoch": 0.458, "grad_norm": 0.4796263575553894, "learning_rate": 0.00013753667149005457, "loss": 1.8544, "step": 5725 }, { "epoch": 0.46, "grad_norm": 0.5249527096748352, "learning_rate": 0.00013704758381697844, "loss": 1.8637, "step": 5750 }, { "epoch": 0.462, "grad_norm": 0.45591726899147034, "learning_rate": 0.00013655746678669525, "loss": 1.8604, "step": 5775 }, { "epoch": 0.464, "grad_norm": 0.3980169892311096, "learning_rate": 0.00013606633401697557, "loss": 1.8648, "step": 5800 }, { "epoch": 0.466, "grad_norm": 0.5417760014533997, "learning_rate": 0.000135574199153812, "loss": 1.8629, "step": 5825 }, { "epoch": 0.468, "grad_norm": 0.7416180372238159, "learning_rate": 0.0001350810758710401, "loss": 1.8644, "step": 5850 }, { "epoch": 0.47, "grad_norm": 0.5018301010131836, "learning_rate": 0.0001345869778699584, "loss": 1.8618, "step": 5875 }, { "epoch": 0.472, "grad_norm": 0.4680832028388977, "learning_rate": 0.0001340919188789477, "loss": 1.85, "step": 5900 }, { "epoch": 0.474, "grad_norm": 0.560767650604248, "learning_rate": 0.00013359591265308963, "loss": 1.853, "step": 5925 }, { "epoch": 0.476, "grad_norm": 0.42464175820350647, "learning_rate": 0.00013309897297378455, "loss": 1.8711, "step": 5950 }, { "epoch": 0.478, "grad_norm": 0.4219244718551636, "learning_rate": 0.00013260111364836848, "loss": 1.8671, "step": 5975 }, { "epoch": 0.48, "grad_norm": 0.4580926299095154, "learning_rate": 0.00013210234850972964, "loss": 1.8557, "step": 6000 }, { "epoch": 0.482, "grad_norm": 0.594601035118103, "learning_rate": 0.00013160269141592398, "loss": 1.8644, "step": 6025 }, { "epoch": 0.484, "grad_norm": 0.4669687747955322, "learning_rate": 0.00013110215624979025, "loss": 1.8629, "step": 6050 }, { "epoch": 0.486, "grad_norm": 0.5055556297302246, "learning_rate": 0.00013060075691856407, "loss": 1.859, "step": 6075 }, { "epoch": 0.488, "grad_norm": 0.5893716216087341, "learning_rate": 0.0001300985073534919, "loss": 1.8571, "step": 6100 }, { "epoch": 0.49, "grad_norm": 0.4281710684299469, "learning_rate": 0.0001295954215094434, "loss": 1.8553, "step": 6125 }, { "epoch": 0.492, "grad_norm": 0.7110034823417664, "learning_rate": 0.0001290915133645243, "loss": 1.854, "step": 6150 }, { "epoch": 0.494, "grad_norm": 0.4707682132720947, "learning_rate": 0.00012858679691968753, "loss": 1.8616, "step": 6175 }, { "epoch": 0.496, "grad_norm": 0.44914594292640686, "learning_rate": 0.00012808128619834461, "loss": 1.8584, "step": 6200 }, { "epoch": 0.498, "grad_norm": 0.4829035997390747, "learning_rate": 0.0001275749952459757, "loss": 1.8573, "step": 6225 }, { "epoch": 0.5, "grad_norm": 0.4545588493347168, "learning_rate": 0.00012706793812973941, "loss": 1.8558, "step": 6250 }, { "epoch": 0.502, "grad_norm": 0.4176503121852875, "learning_rate": 0.0001265601289380822, "loss": 1.8608, "step": 6275 }, { "epoch": 0.504, "grad_norm": 0.48794806003570557, "learning_rate": 0.00012605158178034654, "loss": 1.8531, "step": 6300 }, { "epoch": 0.506, "grad_norm": 0.40763533115386963, "learning_rate": 0.00012554231078637927, "loss": 1.8566, "step": 6325 }, { "epoch": 0.508, "grad_norm": 0.38734304904937744, "learning_rate": 0.00012503233010613865, "loss": 1.8532, "step": 6350 }, { "epoch": 0.51, "grad_norm": 0.6127769351005554, "learning_rate": 0.0001245216539093016, "loss": 1.8536, "step": 6375 }, { "epoch": 0.512, "grad_norm": 0.4811764061450958, "learning_rate": 0.00012401029638486953, "loss": 1.8554, "step": 6400 }, { "epoch": 0.514, "grad_norm": 0.49912744760513306, "learning_rate": 0.00012349827174077448, "loss": 1.857, "step": 6425 }, { "epoch": 0.516, "grad_norm": 0.366845965385437, "learning_rate": 0.00012298559420348437, "loss": 1.8646, "step": 6450 }, { "epoch": 0.518, "grad_norm": 0.613798975944519, "learning_rate": 0.00012247227801760732, "loss": 1.8607, "step": 6475 }, { "epoch": 0.52, "grad_norm": 0.4659370183944702, "learning_rate": 0.0001219583374454963, "loss": 1.8564, "step": 6500 }, { "epoch": 0.522, "grad_norm": 0.47527167201042175, "learning_rate": 0.00012144378676685265, "loss": 1.8536, "step": 6525 }, { "epoch": 0.524, "grad_norm": 0.4141850471496582, "learning_rate": 0.00012092864027832933, "loss": 1.8591, "step": 6550 }, { "epoch": 0.526, "grad_norm": 0.41834697127342224, "learning_rate": 0.00012041291229313372, "loss": 1.8606, "step": 6575 }, { "epoch": 0.528, "grad_norm": 0.45836254954338074, "learning_rate": 0.00011989661714062999, "loss": 1.8534, "step": 6600 }, { "epoch": 0.53, "grad_norm": 0.4878871440887451, "learning_rate": 0.0001193797691659408, "loss": 1.8539, "step": 6625 }, { "epoch": 0.532, "grad_norm": 0.4897819459438324, "learning_rate": 0.00011886238272954897, "loss": 1.8589, "step": 6650 }, { "epoch": 0.534, "grad_norm": 0.4225829541683197, "learning_rate": 0.00011834447220689813, "loss": 1.8613, "step": 6675 }, { "epoch": 0.536, "grad_norm": 0.44458839297294617, "learning_rate": 0.0001178260519879937, "loss": 1.8516, "step": 6700 }, { "epoch": 0.538, "grad_norm": 0.49966317415237427, "learning_rate": 0.00011730713647700282, "loss": 1.8488, "step": 6725 }, { "epoch": 0.54, "grad_norm": 0.3696332275867462, "learning_rate": 0.0001167877400918541, "loss": 1.8538, "step": 6750 }, { "epoch": 0.542, "grad_norm": 0.5024220943450928, "learning_rate": 0.0001162678772638372, "loss": 1.8495, "step": 6775 }, { "epoch": 0.544, "grad_norm": 0.3902568221092224, "learning_rate": 0.0001157475624372018, "loss": 1.8604, "step": 6800 }, { "epoch": 0.546, "grad_norm": 0.4520963430404663, "learning_rate": 0.00011522681006875613, "loss": 1.8564, "step": 6825 }, { "epoch": 0.548, "grad_norm": 0.4059247672557831, "learning_rate": 0.00011470563462746541, "loss": 1.8562, "step": 6850 }, { "epoch": 0.55, "grad_norm": 0.5081529021263123, "learning_rate": 0.00011418405059405002, "loss": 1.8496, "step": 6875 }, { "epoch": 0.552, "grad_norm": 0.5042575001716614, "learning_rate": 0.0001136620724605827, "loss": 1.8599, "step": 6900 }, { "epoch": 0.554, "grad_norm": 0.43565815687179565, "learning_rate": 0.00011313971473008629, "loss": 1.8582, "step": 6925 }, { "epoch": 0.556, "grad_norm": 0.42699581384658813, "learning_rate": 0.00011261699191613066, "loss": 1.8455, "step": 6950 }, { "epoch": 0.558, "grad_norm": 0.4753115773200989, "learning_rate": 0.00011209391854242936, "loss": 1.8577, "step": 6975 }, { "epoch": 0.56, "grad_norm": 0.48413121700286865, "learning_rate": 0.00011157050914243614, "loss": 1.8529, "step": 7000 }, { "epoch": 0.562, "grad_norm": 0.4501863420009613, "learning_rate": 0.00011104677825894121, "loss": 1.8631, "step": 7025 }, { "epoch": 0.564, "grad_norm": 0.5171017050743103, "learning_rate": 0.00011052274044366711, "loss": 1.8529, "step": 7050 }, { "epoch": 0.566, "grad_norm": 0.5637266635894775, "learning_rate": 0.0001099984102568643, "loss": 1.8499, "step": 7075 }, { "epoch": 0.568, "grad_norm": 0.47391700744628906, "learning_rate": 0.00010947380226690684, "loss": 1.8511, "step": 7100 }, { "epoch": 0.57, "grad_norm": 0.43994954228401184, "learning_rate": 0.00010894893104988738, "loss": 1.8535, "step": 7125 }, { "epoch": 0.572, "grad_norm": 0.3985336422920227, "learning_rate": 0.00010842381118921232, "loss": 1.8504, "step": 7150 }, { "epoch": 0.574, "grad_norm": 0.3310425877571106, "learning_rate": 0.00010789845727519647, "loss": 1.8556, "step": 7175 }, { "epoch": 0.576, "grad_norm": 0.4747192859649658, "learning_rate": 0.00010737288390465792, "loss": 1.8519, "step": 7200 }, { "epoch": 0.578, "grad_norm": 0.5283911228179932, "learning_rate": 0.0001068471056805121, "loss": 1.8602, "step": 7225 }, { "epoch": 0.58, "grad_norm": 0.4341649115085602, "learning_rate": 0.00010632113721136636, "loss": 1.8528, "step": 7250 }, { "epoch": 0.582, "grad_norm": 0.47646665573120117, "learning_rate": 0.00010579499311111396, "loss": 1.8501, "step": 7275 }, { "epoch": 0.584, "grad_norm": 0.36453354358673096, "learning_rate": 0.00010526868799852796, "loss": 1.8515, "step": 7300 }, { "epoch": 0.586, "grad_norm": 0.4703463912010193, "learning_rate": 0.00010474223649685517, "loss": 1.8542, "step": 7325 }, { "epoch": 0.588, "grad_norm": 1.5381436347961426, "learning_rate": 0.00010421565323340971, "loss": 1.8534, "step": 7350 }, { "epoch": 0.59, "grad_norm": 0.46722957491874695, "learning_rate": 0.00010368895283916678, "loss": 1.8595, "step": 7375 }, { "epoch": 0.592, "grad_norm": 0.3344251215457916, "learning_rate": 0.0001031621499483559, "loss": 1.859, "step": 7400 }, { "epoch": 0.594, "grad_norm": 0.41089528799057007, "learning_rate": 0.00010263525919805448, "loss": 1.8554, "step": 7425 }, { "epoch": 0.596, "grad_norm": 0.376859188079834, "learning_rate": 0.00010210829522778111, "loss": 1.848, "step": 7450 }, { "epoch": 0.598, "grad_norm": 0.4558032751083374, "learning_rate": 0.00010158127267908877, "loss": 1.8578, "step": 7475 }, { "epoch": 0.6, "grad_norm": 0.366113543510437, "learning_rate": 0.00010105420619515798, "loss": 1.8467, "step": 7500 }, { "epoch": 0.602, "grad_norm": 0.38260912895202637, "learning_rate": 0.00010052711042039, "loss": 1.8504, "step": 7525 }, { "epoch": 0.604, "grad_norm": 0.47190847992897034, "learning_rate": 0.0001, "loss": 1.8552, "step": 7550 }, { "epoch": 0.606, "grad_norm": 0.36239635944366455, "learning_rate": 9.947288957961e-05, "loss": 1.8577, "step": 7575 }, { "epoch": 0.608, "grad_norm": 0.45202115178108215, "learning_rate": 9.894579380484204e-05, "loss": 1.8501, "step": 7600 }, { "epoch": 0.61, "grad_norm": 0.45097965002059937, "learning_rate": 9.841872732091128e-05, "loss": 1.8462, "step": 7625 }, { "epoch": 0.612, "grad_norm": 0.5261498689651489, "learning_rate": 9.789170477221891e-05, "loss": 1.8567, "step": 7650 }, { "epoch": 0.614, "grad_norm": 0.44725409150123596, "learning_rate": 9.736474080194554e-05, "loss": 1.8602, "step": 7675 }, { "epoch": 0.616, "grad_norm": 0.38660377264022827, "learning_rate": 9.683785005164411e-05, "loss": 1.8528, "step": 7700 }, { "epoch": 0.618, "grad_norm": 0.4490872025489807, "learning_rate": 9.631104716083326e-05, "loss": 1.851, "step": 7725 }, { "epoch": 0.62, "grad_norm": 0.394273042678833, "learning_rate": 9.57843467665903e-05, "loss": 1.8576, "step": 7750 }, { "epoch": 0.622, "grad_norm": 0.38389667868614197, "learning_rate": 9.525776350314485e-05, "loss": 1.8523, "step": 7775 }, { "epoch": 0.624, "grad_norm": 0.3490082919597626, "learning_rate": 9.473131200147205e-05, "loss": 1.8549, "step": 7800 }, { "epoch": 0.626, "grad_norm": 0.4410073459148407, "learning_rate": 9.420500688888609e-05, "loss": 1.8511, "step": 7825 }, { "epoch": 0.628, "grad_norm": 0.4412212073802948, "learning_rate": 9.367886278863366e-05, "loss": 1.8446, "step": 7850 }, { "epoch": 0.63, "grad_norm": 0.5358940958976746, "learning_rate": 9.315289431948792e-05, "loss": 1.8518, "step": 7875 }, { "epoch": 0.632, "grad_norm": 0.5098058581352234, "learning_rate": 9.26271160953421e-05, "loss": 1.8431, "step": 7900 }, { "epoch": 0.634, "grad_norm": 0.5359795093536377, "learning_rate": 9.210154272480353e-05, "loss": 1.853, "step": 7925 }, { "epoch": 0.636, "grad_norm": 0.4036993086338043, "learning_rate": 9.157618881078772e-05, "loss": 1.8476, "step": 7950 }, { "epoch": 0.638, "grad_norm": 0.42572522163391113, "learning_rate": 9.105106895011263e-05, "loss": 1.8492, "step": 7975 }, { "epoch": 0.64, "grad_norm": 0.4660392701625824, "learning_rate": 9.052619773309317e-05, "loss": 1.8563, "step": 8000 }, { "epoch": 0.642, "grad_norm": 0.5544576644897461, "learning_rate": 9.000158974313571e-05, "loss": 1.8664, "step": 8025 }, { "epoch": 0.644, "grad_norm": 0.34708672761917114, "learning_rate": 8.947725955633294e-05, "loss": 1.8462, "step": 8050 }, { "epoch": 0.646, "grad_norm": 0.38868874311447144, "learning_rate": 8.895322174105881e-05, "loss": 1.86, "step": 8075 }, { "epoch": 0.648, "grad_norm": 0.4096713960170746, "learning_rate": 8.84294908575639e-05, "loss": 1.8524, "step": 8100 }, { "epoch": 0.65, "grad_norm": 0.4593583345413208, "learning_rate": 8.790608145757065e-05, "loss": 1.8535, "step": 8125 }, { "epoch": 0.652, "grad_norm": 0.5035415887832642, "learning_rate": 8.738300808386935e-05, "loss": 1.8517, "step": 8150 }, { "epoch": 0.654, "grad_norm": 0.373983770608902, "learning_rate": 8.686028526991372e-05, "loss": 1.8537, "step": 8175 }, { "epoch": 0.656, "grad_norm": 0.39891326427459717, "learning_rate": 8.633792753941733e-05, "loss": 1.8494, "step": 8200 }, { "epoch": 0.658, "grad_norm": 0.3910159766674042, "learning_rate": 8.581594940595002e-05, "loss": 1.8523, "step": 8225 }, { "epoch": 0.66, "grad_norm": 0.4816308617591858, "learning_rate": 8.529436537253458e-05, "loss": 1.8391, "step": 8250 }, { "epoch": 0.662, "grad_norm": 0.3788575530052185, "learning_rate": 8.477318993124392e-05, "loss": 1.8498, "step": 8275 }, { "epoch": 0.664, "grad_norm": 0.424724280834198, "learning_rate": 8.425243756279824e-05, "loss": 1.8544, "step": 8300 }, { "epoch": 0.666, "grad_norm": 0.361713707447052, "learning_rate": 8.373212273616282e-05, "loss": 1.8511, "step": 8325 }, { "epoch": 0.668, "grad_norm": 0.41246265172958374, "learning_rate": 8.321225990814591e-05, "loss": 1.8447, "step": 8350 }, { "epoch": 0.67, "grad_norm": 0.4148729741573334, "learning_rate": 8.269286352299724e-05, "loss": 1.8477, "step": 8375 }, { "epoch": 0.672, "grad_norm": 0.37317347526550293, "learning_rate": 8.217394801200631e-05, "loss": 1.8455, "step": 8400 }, { "epoch": 0.674, "grad_norm": 0.41947707533836365, "learning_rate": 8.165552779310191e-05, "loss": 1.8641, "step": 8425 }, { "epoch": 0.676, "grad_norm": 0.6854198575019836, "learning_rate": 8.113761727045105e-05, "loss": 1.8453, "step": 8450 }, { "epoch": 0.678, "grad_norm": 0.4508031904697418, "learning_rate": 8.062023083405918e-05, "loss": 1.8482, "step": 8475 }, { "epoch": 0.68, "grad_norm": 0.38871219754219055, "learning_rate": 8.010338285937006e-05, "loss": 1.8432, "step": 8500 }, { "epoch": 0.682, "grad_norm": 0.4239045977592468, "learning_rate": 7.958708770686629e-05, "loss": 1.8532, "step": 8525 }, { "epoch": 0.684, "grad_norm": 0.43459609150886536, "learning_rate": 7.907135972167069e-05, "loss": 1.8501, "step": 8550 }, { "epoch": 0.686, "grad_norm": 0.4150739908218384, "learning_rate": 7.855621323314736e-05, "loss": 1.8436, "step": 8575 }, { "epoch": 0.688, "grad_norm": 0.4099458158016205, "learning_rate": 7.804166255450373e-05, "loss": 1.8503, "step": 8600 }, { "epoch": 0.69, "grad_norm": 0.4599255323410034, "learning_rate": 7.75277219823927e-05, "loss": 1.8529, "step": 8625 }, { "epoch": 0.692, "grad_norm": 0.4123729169368744, "learning_rate": 7.701440579651566e-05, "loss": 1.854, "step": 8650 }, { "epoch": 0.694, "grad_norm": 0.38956719636917114, "learning_rate": 7.65017282592255e-05, "loss": 1.842, "step": 8675 }, { "epoch": 0.696, "grad_norm": 0.3663404583930969, "learning_rate": 7.598970361513051e-05, "loss": 1.8427, "step": 8700 }, { "epoch": 0.698, "grad_norm": 0.37082424759864807, "learning_rate": 7.547834609069846e-05, "loss": 1.8463, "step": 8725 }, { "epoch": 0.7, "grad_norm": 0.4512746036052704, "learning_rate": 7.496766989386136e-05, "loss": 1.8491, "step": 8750 }, { "epoch": 0.702, "grad_norm": 0.39837372303009033, "learning_rate": 7.445768921362077e-05, "loss": 1.8504, "step": 8775 }, { "epoch": 0.704, "grad_norm": 0.40776899456977844, "learning_rate": 7.394841821965345e-05, "loss": 1.8429, "step": 8800 }, { "epoch": 0.706, "grad_norm": 0.34008052945137024, "learning_rate": 7.343987106191785e-05, "loss": 1.8463, "step": 8825 }, { "epoch": 0.708, "grad_norm": 0.3931528925895691, "learning_rate": 7.293206187026061e-05, "loss": 1.8554, "step": 8850 }, { "epoch": 0.71, "grad_norm": 0.37032806873321533, "learning_rate": 7.242500475402433e-05, "loss": 1.8646, "step": 8875 }, { "epoch": 0.712, "grad_norm": 0.39694976806640625, "learning_rate": 7.191871380165538e-05, "loss": 1.8544, "step": 8900 }, { "epoch": 0.714, "grad_norm": 0.37797749042510986, "learning_rate": 7.14132030803125e-05, "loss": 1.8527, "step": 8925 }, { "epoch": 0.716, "grad_norm": 0.4343235194683075, "learning_rate": 7.090848663547574e-05, "loss": 1.8435, "step": 8950 }, { "epoch": 0.718, "grad_norm": 0.5005691647529602, "learning_rate": 7.040457849055662e-05, "loss": 1.8452, "step": 8975 }, { "epoch": 0.72, "grad_norm": 0.38191068172454834, "learning_rate": 6.990149264650814e-05, "loss": 1.853, "step": 9000 }, { "epoch": 0.722, "grad_norm": 0.38967859745025635, "learning_rate": 6.939924308143591e-05, "loss": 1.8463, "step": 9025 }, { "epoch": 0.724, "grad_norm": 0.46376553177833557, "learning_rate": 6.889784375020978e-05, "loss": 1.8546, "step": 9050 }, { "epoch": 0.726, "grad_norm": 0.35381075739860535, "learning_rate": 6.839730858407603e-05, "loss": 1.8463, "step": 9075 }, { "epoch": 0.728, "grad_norm": 0.3585016131401062, "learning_rate": 6.789765149027039e-05, "loss": 1.8492, "step": 9100 }, { "epoch": 0.73, "grad_norm": 0.3320139944553375, "learning_rate": 6.739888635163154e-05, "loss": 1.8523, "step": 9125 }, { "epoch": 0.732, "grad_norm": 0.4588174819946289, "learning_rate": 6.690102702621548e-05, "loss": 1.849, "step": 9150 }, { "epoch": 0.734, "grad_norm": 0.3317193388938904, "learning_rate": 6.640408734691038e-05, "loss": 1.8487, "step": 9175 }, { "epoch": 0.736, "grad_norm": 0.4676886200904846, "learning_rate": 6.590808112105232e-05, "loss": 1.8461, "step": 9200 }, { "epoch": 0.738, "grad_norm": 0.3905697166919708, "learning_rate": 6.541302213004159e-05, "loss": 1.8429, "step": 9225 }, { "epoch": 0.74, "grad_norm": 0.36496925354003906, "learning_rate": 6.491892412895995e-05, "loss": 1.8469, "step": 9250 }, { "epoch": 0.742, "grad_norm": 0.4297652840614319, "learning_rate": 6.442580084618805e-05, "loss": 1.8469, "step": 9275 }, { "epoch": 0.744, "grad_norm": 0.3861255645751953, "learning_rate": 6.393366598302446e-05, "loss": 1.8498, "step": 9300 }, { "epoch": 0.746, "grad_norm": 0.32486528158187866, "learning_rate": 6.344253321330476e-05, "loss": 1.8504, "step": 9325 }, { "epoch": 0.748, "grad_norm": 0.35413098335266113, "learning_rate": 6.295241618302156e-05, "loss": 1.8526, "step": 9350 }, { "epoch": 0.75, "grad_norm": 0.37345078587532043, "learning_rate": 6.246332850994547e-05, "loss": 1.8511, "step": 9375 }, { "epoch": 0.752, "grad_norm": 0.3581782877445221, "learning_rate": 6.197528378324665e-05, "loss": 1.8506, "step": 9400 }, { "epoch": 0.754, "grad_norm": 0.46359720826148987, "learning_rate": 6.148829556311728e-05, "loss": 1.8496, "step": 9425 }, { "epoch": 0.756, "grad_norm": 0.3624126613140106, "learning_rate": 6.100237738039484e-05, "loss": 1.8459, "step": 9450 }, { "epoch": 0.758, "grad_norm": 0.33995765447616577, "learning_rate": 6.051754273618605e-05, "loss": 1.852, "step": 9475 }, { "epoch": 0.76, "grad_norm": 0.3518821597099304, "learning_rate": 6.0033805101491794e-05, "loss": 1.8508, "step": 9500 }, { "epoch": 0.762, "grad_norm": 0.3974629342556, "learning_rate": 5.955117791683289e-05, "loss": 1.8467, "step": 9525 }, { "epoch": 0.764, "grad_norm": 0.39932018518447876, "learning_rate": 5.9069674591876534e-05, "loss": 1.8486, "step": 9550 }, { "epoch": 0.766, "grad_norm": 0.3213377892971039, "learning_rate": 5.858930850506388e-05, "loss": 1.8494, "step": 9575 }, { "epoch": 0.768, "grad_norm": 0.3450121581554413, "learning_rate": 5.811009300323818e-05, "loss": 1.8512, "step": 9600 }, { "epoch": 0.77, "grad_norm": 0.335205614566803, "learning_rate": 5.7632041401273895e-05, "loss": 1.8484, "step": 9625 }, { "epoch": 0.772, "grad_norm": 0.32677674293518066, "learning_rate": 5.7155166981706956e-05, "loss": 1.8432, "step": 9650 }, { "epoch": 0.774, "grad_norm": 0.3155275285243988, "learning_rate": 5.667948299436555e-05, "loss": 1.8468, "step": 9675 }, { "epoch": 0.776, "grad_norm": 0.3949924409389496, "learning_rate": 5.620500265600206e-05, "loss": 1.8433, "step": 9700 }, { "epoch": 0.778, "grad_norm": 0.3334220051765442, "learning_rate": 5.573173914992575e-05, "loss": 1.8512, "step": 9725 }, { "epoch": 0.78, "grad_norm": 0.3740833103656769, "learning_rate": 5.525970562563656e-05, "loss": 1.8443, "step": 9750 }, { "epoch": 0.782, "grad_norm": 0.33872148394584656, "learning_rate": 5.47889151984597e-05, "loss": 1.8511, "step": 9775 }, { "epoch": 0.784, "grad_norm": 0.4084971845149994, "learning_rate": 5.431938094918132e-05, "loss": 1.8464, "step": 9800 }, { "epoch": 0.786, "grad_norm": 0.37755846977233887, "learning_rate": 5.385111592368486e-05, "loss": 1.8463, "step": 9825 }, { "epoch": 0.788, "grad_norm": 0.3490227460861206, "learning_rate": 5.3384133132588784e-05, "loss": 1.8561, "step": 9850 }, { "epoch": 0.79, "grad_norm": 0.3497559726238251, "learning_rate": 5.2918445550885165e-05, "loss": 1.8467, "step": 9875 }, { "epoch": 0.792, "grad_norm": 0.37137648463249207, "learning_rate": 5.2454066117578815e-05, "loss": 1.8448, "step": 9900 }, { "epoch": 0.794, "grad_norm": 0.4525596499443054, "learning_rate": 5.199100773532814e-05, "loss": 1.848, "step": 9925 }, { "epoch": 0.796, "grad_norm": 0.4981496036052704, "learning_rate": 5.152928327008635e-05, "loss": 1.8454, "step": 9950 }, { "epoch": 0.798, "grad_norm": 0.6176758408546448, "learning_rate": 5.106890555074426e-05, "loss": 1.8373, "step": 9975 }, { "epoch": 0.8, "grad_norm": 0.30251458287239075, "learning_rate": 5.060988736877366e-05, "loss": 1.8436, "step": 10000 }, { "epoch": 0.802, "grad_norm": 0.32596534490585327, "learning_rate": 5.015224147787195e-05, "loss": 1.8528, "step": 10025 }, { "epoch": 0.804, "grad_norm": 0.33038198947906494, "learning_rate": 4.9695980593607817e-05, "loss": 1.851, "step": 10050 }, { "epoch": 0.806, "grad_norm": 0.28935107588768005, "learning_rate": 4.924111739306788e-05, "loss": 1.8456, "step": 10075 }, { "epoch": 0.808, "grad_norm": 0.33840999007225037, "learning_rate": 4.8787664514504504e-05, "loss": 1.8393, "step": 10100 }, { "epoch": 0.81, "grad_norm": 0.30848726630210876, "learning_rate": 4.833563455698468e-05, "loss": 1.8495, "step": 10125 }, { "epoch": 0.812, "grad_norm": 0.35452771186828613, "learning_rate": 4.788504008003978e-05, "loss": 1.8416, "step": 10150 }, { "epoch": 0.814, "grad_norm": 0.29864782094955444, "learning_rate": 4.743589360331684e-05, "loss": 1.8469, "step": 10175 }, { "epoch": 0.816, "grad_norm": 0.32326069474220276, "learning_rate": 4.698820760623064e-05, "loss": 1.8412, "step": 10200 }, { "epoch": 0.818, "grad_norm": 0.3136712908744812, "learning_rate": 4.6541994527616836e-05, "loss": 1.8478, "step": 10225 }, { "epoch": 0.82, "grad_norm": 0.3415374457836151, "learning_rate": 4.609726676538652e-05, "loss": 1.8448, "step": 10250 }, { "epoch": 0.822, "grad_norm": 0.3662770092487335, "learning_rate": 4.56540366761815e-05, "loss": 1.845, "step": 10275 }, { "epoch": 0.824, "grad_norm": 0.3077206611633301, "learning_rate": 4.521231657503133e-05, "loss": 1.8526, "step": 10300 }, { "epoch": 0.826, "grad_norm": 0.35801073908805847, "learning_rate": 4.477211873501085e-05, "loss": 1.8426, "step": 10325 }, { "epoch": 0.828, "grad_norm": 0.37789011001586914, "learning_rate": 4.43334553868993e-05, "loss": 1.8448, "step": 10350 }, { "epoch": 0.83, "grad_norm": 0.3104768693447113, "learning_rate": 4.389633871884049e-05, "loss": 1.8409, "step": 10375 }, { "epoch": 0.832, "grad_norm": 0.3287721574306488, "learning_rate": 4.346078087600412e-05, "loss": 1.847, "step": 10400 }, { "epoch": 0.834, "grad_norm": 0.34750932455062866, "learning_rate": 4.3026793960248334e-05, "loss": 1.8481, "step": 10425 }, { "epoch": 0.836, "grad_norm": 0.3314663767814636, "learning_rate": 4.2594390029783534e-05, "loss": 1.8465, "step": 10450 }, { "epoch": 0.838, "grad_norm": 0.3413223326206207, "learning_rate": 4.21635810988373e-05, "loss": 1.8456, "step": 10475 }, { "epoch": 0.84, "grad_norm": 0.3270764946937561, "learning_rate": 4.173437913732048e-05, "loss": 1.8473, "step": 10500 }, { "epoch": 0.842, "grad_norm": 0.293573260307312, "learning_rate": 4.1306796070494755e-05, "loss": 1.8399, "step": 10525 }, { "epoch": 0.844, "grad_norm": 0.32640865445137024, "learning_rate": 4.088084377864135e-05, "loss": 1.8487, "step": 10550 }, { "epoch": 0.846, "grad_norm": 0.4020804762840271, "learning_rate": 4.045653409673078e-05, "loss": 1.8364, "step": 10575 }, { "epoch": 0.848, "grad_norm": 0.35024294257164, "learning_rate": 4.003387881409397e-05, "loss": 1.8498, "step": 10600 }, { "epoch": 0.85, "grad_norm": 0.33417728543281555, "learning_rate": 3.9612889674094954e-05, "loss": 1.8488, "step": 10625 }, { "epoch": 0.852, "grad_norm": 0.34991011023521423, "learning_rate": 3.9193578373804364e-05, "loss": 1.8423, "step": 10650 }, { "epoch": 0.854, "grad_norm": 0.36940819025039673, "learning_rate": 3.8775956563674543e-05, "loss": 1.8438, "step": 10675 }, { "epoch": 0.856, "grad_norm": 0.3352839946746826, "learning_rate": 3.836003584721577e-05, "loss": 1.8515, "step": 10700 }, { "epoch": 0.858, "grad_norm": 0.2906590402126312, "learning_rate": 3.7945827780673905e-05, "loss": 1.8434, "step": 10725 }, { "epoch": 0.86, "grad_norm": 0.3562522530555725, "learning_rate": 3.7533343872709294e-05, "loss": 1.8438, "step": 10750 }, { "epoch": 0.862, "grad_norm": 0.5502293109893799, "learning_rate": 3.712259558407698e-05, "loss": 1.8415, "step": 10775 }, { "epoch": 0.864, "grad_norm": 0.35165935754776, "learning_rate": 3.671359432730834e-05, "loss": 1.8412, "step": 10800 }, { "epoch": 0.866, "grad_norm": 0.3448835015296936, "learning_rate": 3.630635146639384e-05, "loss": 1.8469, "step": 10825 }, { "epoch": 0.868, "grad_norm": 0.32406798005104065, "learning_rate": 3.5900878316467454e-05, "loss": 1.846, "step": 10850 }, { "epoch": 0.87, "grad_norm": 0.3257356286048889, "learning_rate": 3.549718614349218e-05, "loss": 1.8514, "step": 10875 }, { "epoch": 0.872, "grad_norm": 0.36643341183662415, "learning_rate": 3.509528616394716e-05, "loss": 1.8417, "step": 10900 }, { "epoch": 0.874, "grad_norm": 0.3198818862438202, "learning_rate": 3.469518954451573e-05, "loss": 1.8478, "step": 10925 }, { "epoch": 0.876, "grad_norm": 0.35304951667785645, "learning_rate": 3.429690740177549e-05, "loss": 1.8371, "step": 10950 }, { "epoch": 0.878, "grad_norm": 0.3322315812110901, "learning_rate": 3.390045080188923e-05, "loss": 1.8463, "step": 10975 }, { "epoch": 0.88, "grad_norm": 0.33379215002059937, "learning_rate": 3.350583076029754e-05, "loss": 1.8384, "step": 11000 }, { "epoch": 0.882, "grad_norm": 0.3084147274494171, "learning_rate": 3.311305824141273e-05, "loss": 1.8449, "step": 11025 }, { "epoch": 0.884, "grad_norm": 0.39884597063064575, "learning_rate": 3.272214415831418e-05, "loss": 1.8395, "step": 11050 }, { "epoch": 0.886, "grad_norm": 0.3557879626750946, "learning_rate": 3.233309937244513e-05, "loss": 1.843, "step": 11075 }, { "epoch": 0.888, "grad_norm": 0.32669898867607117, "learning_rate": 3.1945934693310896e-05, "loss": 1.8451, "step": 11100 }, { "epoch": 0.89, "grad_norm": 0.36104997992515564, "learning_rate": 3.156066087817856e-05, "loss": 1.8484, "step": 11125 }, { "epoch": 0.892, "grad_norm": 0.3507726192474365, "learning_rate": 3.117728863177796e-05, "loss": 1.8514, "step": 11150 }, { "epoch": 0.894, "grad_norm": 0.2982633411884308, "learning_rate": 3.0795828606004454e-05, "loss": 1.8429, "step": 11175 }, { "epoch": 0.896, "grad_norm": 0.299304336309433, "learning_rate": 3.041629139962283e-05, "loss": 1.8425, "step": 11200 }, { "epoch": 0.898, "grad_norm": 0.3054879605770111, "learning_rate": 3.0038687557972856e-05, "loss": 1.8468, "step": 11225 }, { "epoch": 0.9, "grad_norm": 0.3667582869529724, "learning_rate": 2.966302757267625e-05, "loss": 1.8396, "step": 11250 }, { "epoch": 0.902, "grad_norm": 0.3155341148376465, "learning_rate": 2.9289321881345254e-05, "loss": 1.844, "step": 11275 }, { "epoch": 0.904, "grad_norm": 0.32875508069992065, "learning_rate": 2.8917580867292526e-05, "loss": 1.8436, "step": 11300 }, { "epoch": 0.906, "grad_norm": 0.30347251892089844, "learning_rate": 2.8547814859242727e-05, "loss": 1.8365, "step": 11325 }, { "epoch": 0.908, "grad_norm": 0.31834813952445984, "learning_rate": 2.8180034131045474e-05, "loss": 1.837, "step": 11350 }, { "epoch": 0.91, "grad_norm": 0.30820679664611816, "learning_rate": 2.7814248901389918e-05, "loss": 1.8524, "step": 11375 }, { "epoch": 0.912, "grad_norm": 0.42067161202430725, "learning_rate": 2.7450469333520855e-05, "loss": 1.8389, "step": 11400 }, { "epoch": 0.914, "grad_norm": 0.34037408232688904, "learning_rate": 2.708870553495625e-05, "loss": 1.8382, "step": 11425 }, { "epoch": 0.916, "grad_norm": 0.32559776306152344, "learning_rate": 2.672896755720654e-05, "loss": 1.8456, "step": 11450 }, { "epoch": 0.918, "grad_norm": 0.3477393388748169, "learning_rate": 2.6371265395495137e-05, "loss": 1.8414, "step": 11475 }, { "epoch": 0.92, "grad_norm": 0.3077406883239746, "learning_rate": 2.6015608988480965e-05, "loss": 1.8438, "step": 11500 }, { "epoch": 0.922, "grad_norm": 0.3561972379684448, "learning_rate": 2.5662008217982158e-05, "loss": 1.8408, "step": 11525 }, { "epoch": 0.924, "grad_norm": 0.30775362253189087, "learning_rate": 2.5310472908701555e-05, "loss": 1.8372, "step": 11550 }, { "epoch": 0.926, "grad_norm": 0.3170279264450073, "learning_rate": 2.496101282795369e-05, "loss": 1.8467, "step": 11575 }, { "epoch": 0.928, "grad_norm": 0.3394754230976105, "learning_rate": 2.4613637685393432e-05, "loss": 1.8359, "step": 11600 }, { "epoch": 0.93, "grad_norm": 0.32178953289985657, "learning_rate": 2.4268357132746223e-05, "loss": 1.8442, "step": 11625 }, { "epoch": 0.932, "grad_norm": 0.31228768825531006, "learning_rate": 2.3925180763539844e-05, "loss": 1.8361, "step": 11650 }, { "epoch": 0.934, "grad_norm": 0.3137824535369873, "learning_rate": 2.358411811283796e-05, "loss": 1.8409, "step": 11675 }, { "epoch": 0.936, "grad_norm": 0.2687174677848816, "learning_rate": 2.324517865697501e-05, "loss": 1.8391, "step": 11700 }, { "epoch": 0.938, "grad_norm": 0.3887827694416046, "learning_rate": 2.2908371813293195e-05, "loss": 1.8471, "step": 11725 }, { "epoch": 0.94, "grad_norm": 0.2946266829967499, "learning_rate": 2.2573706939880555e-05, "loss": 1.8442, "step": 11750 }, { "epoch": 0.942, "grad_norm": 0.31280624866485596, "learning_rate": 2.224119333531113e-05, "loss": 1.8447, "step": 11775 }, { "epoch": 0.944, "grad_norm": 0.3431771993637085, "learning_rate": 2.1910840238386398e-05, "loss": 1.8446, "step": 11800 }, { "epoch": 0.946, "grad_norm": 0.3097357451915741, "learning_rate": 2.1582656827878855e-05, "loss": 1.8389, "step": 11825 }, { "epoch": 0.948, "grad_norm": 0.34613868594169617, "learning_rate": 2.125665222227675e-05, "loss": 1.8472, "step": 11850 }, { "epoch": 0.95, "grad_norm": 0.2880703806877136, "learning_rate": 2.0932835479530877e-05, "loss": 1.8449, "step": 11875 }, { "epoch": 0.952, "grad_norm": 0.3585701584815979, "learning_rate": 2.06112155968028e-05, "loss": 1.8453, "step": 11900 }, { "epoch": 0.954, "grad_norm": 0.35706913471221924, "learning_rate": 2.029180151021496e-05, "loss": 1.841, "step": 11925 }, { "epoch": 0.956, "grad_norm": 0.3220544159412384, "learning_rate": 1.99746020946023e-05, "loss": 1.85, "step": 11950 }, { "epoch": 0.958, "grad_norm": 0.30140894651412964, "learning_rate": 1.9659626163265745e-05, "loss": 1.8376, "step": 11975 }, { "epoch": 0.96, "grad_norm": 0.30860233306884766, "learning_rate": 1.9346882467727335e-05, "loss": 1.8466, "step": 12000 }, { "epoch": 0.962, "grad_norm": 0.2797991633415222, "learning_rate": 1.9036379697486918e-05, "loss": 1.847, "step": 12025 }, { "epoch": 0.964, "grad_norm": 0.3087409436702728, "learning_rate": 1.872812647978095e-05, "loss": 1.8418, "step": 12050 }, { "epoch": 0.966, "grad_norm": 0.3282470703125, "learning_rate": 1.8422131379342668e-05, "loss": 1.8434, "step": 12075 }, { "epoch": 0.968, "grad_norm": 0.3542603850364685, "learning_rate": 1.81184028981641e-05, "loss": 1.8406, "step": 12100 }, { "epoch": 0.97, "grad_norm": 0.3178626298904419, "learning_rate": 1.781694947525979e-05, "loss": 1.8361, "step": 12125 }, { "epoch": 0.972, "grad_norm": 0.2950464189052582, "learning_rate": 1.7517779486432495e-05, "loss": 1.8427, "step": 12150 }, { "epoch": 0.974, "grad_norm": 0.29088157415390015, "learning_rate": 1.7220901244040345e-05, "loss": 1.838, "step": 12175 }, { "epoch": 0.976, "grad_norm": 0.33172252774238586, "learning_rate": 1.6926322996765897e-05, "loss": 1.8423, "step": 12200 }, { "epoch": 0.978, "grad_norm": 0.3150504231452942, "learning_rate": 1.6634052929386944e-05, "loss": 1.8404, "step": 12225 }, { "epoch": 0.98, "grad_norm": 0.31600815057754517, "learning_rate": 1.634409916254914e-05, "loss": 1.8367, "step": 12250 }, { "epoch": 0.982, "grad_norm": 0.29730069637298584, "learning_rate": 1.605646975254035e-05, "loss": 1.8443, "step": 12275 }, { "epoch": 0.984, "grad_norm": 0.2882988452911377, "learning_rate": 1.5771172691066794e-05, "loss": 1.8385, "step": 12300 }, { "epoch": 0.986, "grad_norm": 0.2824288606643677, "learning_rate": 1.5488215905031034e-05, "loss": 1.8436, "step": 12325 }, { "epoch": 0.988, "grad_norm": 0.3098677694797516, "learning_rate": 1.520760725631164e-05, "loss": 1.8395, "step": 12350 }, { "epoch": 0.99, "grad_norm": 0.32261407375335693, "learning_rate": 1.4929354541544883e-05, "loss": 1.8419, "step": 12375 }, { "epoch": 0.992, "grad_norm": 0.30824896693229675, "learning_rate": 1.4653465491908003e-05, "loss": 1.8434, "step": 12400 }, { "epoch": 0.994, "grad_norm": 0.3139910399913788, "learning_rate": 1.4379947772904489e-05, "loss": 1.8422, "step": 12425 }, { "epoch": 0.996, "grad_norm": 0.3424730598926544, "learning_rate": 1.4108808984151023e-05, "loss": 1.8525, "step": 12450 }, { "epoch": 0.998, "grad_norm": 0.2804199159145355, "learning_rate": 1.3840056659166257e-05, "loss": 1.8413, "step": 12475 }, { "epoch": 1.0, "grad_norm": 0.30693376064300537, "learning_rate": 1.3573698265161683e-05, "loss": 1.8359, "step": 12500 }, { "epoch": 1.002, "grad_norm": 0.3113084137439728, "learning_rate": 1.3309741202834048e-05, "loss": 1.8377, "step": 12525 }, { "epoch": 1.004, "grad_norm": 0.32502156496047974, "learning_rate": 1.3048192806159721e-05, "loss": 1.8424, "step": 12550 }, { "epoch": 1.006, "grad_norm": 0.33035212755203247, "learning_rate": 1.278906034219094e-05, "loss": 1.8341, "step": 12575 }, { "epoch": 1.008, "grad_norm": 0.32091912627220154, "learning_rate": 1.2532351010853916e-05, "loss": 1.8315, "step": 12600 }, { "epoch": 1.01, "grad_norm": 0.34205472469329834, "learning_rate": 1.2278071944748748e-05, "loss": 1.8377, "step": 12625 }, { "epoch": 1.012, "grad_norm": 0.31815269589424133, "learning_rate": 1.2026230208951306e-05, "loss": 1.8363, "step": 12650 }, { "epoch": 1.014, "grad_norm": 0.30523747205734253, "learning_rate": 1.1776832800816807e-05, "loss": 1.8481, "step": 12675 }, { "epoch": 1.016, "grad_norm": 0.31181463599205017, "learning_rate": 1.152988664978556e-05, "loss": 1.8392, "step": 12700 }, { "epoch": 1.018, "grad_norm": 0.29633858799934387, "learning_rate": 1.1285398617190279e-05, "loss": 1.841, "step": 12725 }, { "epoch": 1.02, "grad_norm": 0.310817688703537, "learning_rate": 1.1043375496065611e-05, "loss": 1.8453, "step": 12750 }, { "epoch": 1.022, "grad_norm": 0.35818150639533997, "learning_rate": 1.0803824010959252e-05, "loss": 1.8406, "step": 12775 }, { "epoch": 1.024, "grad_norm": 0.33510810136795044, "learning_rate": 1.0566750817745074e-05, "loss": 1.8424, "step": 12800 }, { "epoch": 1.026, "grad_norm": 0.3502376079559326, "learning_rate": 1.0332162503438381e-05, "loss": 1.8415, "step": 12825 }, { "epoch": 1.028, "grad_norm": 0.3100811839103699, "learning_rate": 1.010006558601274e-05, "loss": 1.8493, "step": 12850 }, { "epoch": 1.03, "grad_norm": 0.34041938185691833, "learning_rate": 9.870466514218911e-06, "loss": 1.84, "step": 12875 }, { "epoch": 1.032, "grad_norm": 0.31758561730384827, "learning_rate": 9.643371667405698e-06, "loss": 1.8456, "step": 12900 }, { "epoch": 1.034, "grad_norm": 0.2923029959201813, "learning_rate": 9.418787355342673e-06, "loss": 1.8479, "step": 12925 }, { "epoch": 1.036, "grad_norm": 0.32617050409317017, "learning_rate": 9.196719818044886e-06, "loss": 1.8387, "step": 12950 }, { "epoch": 1.038, "grad_norm": 0.3390202522277832, "learning_rate": 8.977175225599466e-06, "loss": 1.8375, "step": 12975 }, { "epoch": 1.04, "grad_norm": 0.31298160552978516, "learning_rate": 8.760159677994172e-06, "loss": 1.8359, "step": 13000 }, { "epoch": 1.042, "grad_norm": 0.3362617492675781, "learning_rate": 8.545679204947954e-06, "loss": 1.8367, "step": 13025 }, { "epoch": 1.044, "grad_norm": 0.31163260340690613, "learning_rate": 8.333739765743398e-06, "loss": 1.8425, "step": 13050 }, { "epoch": 1.046, "grad_norm": 0.3510798215866089, "learning_rate": 8.124347249061115e-06, "loss": 1.8401, "step": 13075 }, { "epoch": 1.048, "grad_norm": 0.3208621144294739, "learning_rate": 7.91750747281621e-06, "loss": 1.8443, "step": 13100 }, { "epoch": 1.05, "grad_norm": 0.31452810764312744, "learning_rate": 7.713226183996514e-06, "loss": 1.842, "step": 13125 }, { "epoch": 1.052, "grad_norm": 0.31630128622055054, "learning_rate": 7.5115090585029966e-06, "loss": 1.8359, "step": 13150 }, { "epoch": 1.054, "grad_norm": 0.31110531091690063, "learning_rate": 7.312361700992043e-06, "loss": 1.8344, "step": 13175 }, { "epoch": 1.056, "grad_norm": 0.28280314803123474, "learning_rate": 7.115789644719728e-06, "loss": 1.8366, "step": 13200 }, { "epoch": 1.058, "grad_norm": 0.29485175013542175, "learning_rate": 6.92179835138802e-06, "loss": 1.8418, "step": 13225 }, { "epoch": 1.06, "grad_norm": 0.30167973041534424, "learning_rate": 6.730393210993147e-06, "loss": 1.8341, "step": 13250 }, { "epoch": 1.062, "grad_norm": 0.3092511296272278, "learning_rate": 6.541579541675735e-06, "loss": 1.8388, "step": 13275 }, { "epoch": 1.064, "grad_norm": 0.3093793988227844, "learning_rate": 6.355362589573077e-06, "loss": 1.8386, "step": 13300 }, { "epoch": 1.066, "grad_norm": 0.32986992597579956, "learning_rate": 6.1717475286734e-06, "loss": 1.8415, "step": 13325 }, { "epoch": 1.068, "grad_norm": 0.3448794186115265, "learning_rate": 5.990739460672024e-06, "loss": 1.837, "step": 13350 }, { "epoch": 1.07, "grad_norm": 0.3527809679508209, "learning_rate": 5.812343414829724e-06, "loss": 1.8315, "step": 13375 }, { "epoch": 1.072, "grad_norm": 0.324997216463089, "learning_rate": 5.636564347832907e-06, "loss": 1.8339, "step": 13400 }, { "epoch": 1.074, "grad_norm": 0.3535805642604828, "learning_rate": 5.463407143655941e-06, "loss": 1.8423, "step": 13425 }, { "epoch": 1.076, "grad_norm": 0.3439856171607971, "learning_rate": 5.2928766134254345e-06, "loss": 1.8423, "step": 13450 }, { "epoch": 1.078, "grad_norm": 0.3921189308166504, "learning_rate": 5.124977495286542e-06, "loss": 1.8462, "step": 13475 }, { "epoch": 1.08, "grad_norm": 0.3176841139793396, "learning_rate": 4.959714454271369e-06, "loss": 1.84, "step": 13500 }, { "epoch": 1.082, "grad_norm": 0.3492632210254669, "learning_rate": 4.7970920821693076e-06, "loss": 1.8424, "step": 13525 }, { "epoch": 1.084, "grad_norm": 0.3335501253604889, "learning_rate": 4.6371148973994525e-06, "loss": 1.848, "step": 13550 }, { "epoch": 1.086, "grad_norm": 0.31899526715278625, "learning_rate": 4.479787344885078e-06, "loss": 1.8386, "step": 13575 }, { "epoch": 1.088, "grad_norm": 0.2860642075538635, "learning_rate": 4.325113795930203e-06, "loss": 1.841, "step": 13600 }, { "epoch": 1.09, "grad_norm": 0.32865121960639954, "learning_rate": 4.173098548097965e-06, "loss": 1.8401, "step": 13625 }, { "epoch": 1.092, "grad_norm": 0.3371075689792633, "learning_rate": 4.023745825091407e-06, "loss": 1.8354, "step": 13650 }, { "epoch": 1.094, "grad_norm": 0.3013497591018677, "learning_rate": 3.877059776635949e-06, "loss": 1.8438, "step": 13675 }, { "epoch": 1.096, "grad_norm": 0.2933107018470764, "learning_rate": 3.7330444783642338e-06, "loss": 1.8381, "step": 13700 }, { "epoch": 1.098, "grad_norm": 0.30581656098365784, "learning_rate": 3.5917039317028055e-06, "loss": 1.8369, "step": 13725 }, { "epoch": 1.1, "grad_norm": 0.3353700637817383, "learning_rate": 3.4530420637609363e-06, "loss": 1.8384, "step": 13750 }, { "epoch": 1.102, "grad_norm": 0.31908053159713745, "learning_rate": 3.3170627272215427e-06, "loss": 1.8431, "step": 13775 }, { "epoch": 1.104, "grad_norm": 0.3025061786174774, "learning_rate": 3.1837697002341293e-06, "loss": 1.8387, "step": 13800 }, { "epoch": 1.106, "grad_norm": 0.3295077979564667, "learning_rate": 3.053166686309783e-06, "loss": 1.841, "step": 13825 }, { "epoch": 1.108, "grad_norm": 0.32711341977119446, "learning_rate": 2.9252573142183326e-06, "loss": 1.8385, "step": 13850 }, { "epoch": 1.11, "grad_norm": 0.3425910472869873, "learning_rate": 2.8000451378874525e-06, "loss": 1.8378, "step": 13875 }, { "epoch": 1.112, "grad_norm": 0.3226805627346039, "learning_rate": 2.677533636303964e-06, "loss": 1.8372, "step": 13900 }, { "epoch": 1.114, "grad_norm": 0.3507517874240875, "learning_rate": 2.5577262134171885e-06, "loss": 1.8396, "step": 13925 }, { "epoch": 1.116, "grad_norm": 0.2989920973777771, "learning_rate": 2.440626198044327e-06, "loss": 1.839, "step": 13950 }, { "epoch": 1.1179999999999999, "grad_norm": 0.3350691795349121, "learning_rate": 2.3262368437779737e-06, "loss": 1.827, "step": 13975 }, { "epoch": 1.12, "grad_norm": 0.2822107672691345, "learning_rate": 2.2145613288957478e-06, "loss": 1.8458, "step": 14000 }, { "epoch": 1.1219999999999999, "grad_norm": 0.3207239806652069, "learning_rate": 2.1056027562719514e-06, "loss": 1.8421, "step": 14025 }, { "epoch": 1.124, "grad_norm": 0.3366701304912567, "learning_rate": 1.9993641532913833e-06, "loss": 1.8383, "step": 14050 }, { "epoch": 1.126, "grad_norm": 0.32885831594467163, "learning_rate": 1.895848471765227e-06, "loss": 1.8425, "step": 14075 }, { "epoch": 1.1280000000000001, "grad_norm": 0.2935798764228821, "learning_rate": 1.7950585878489856e-06, "loss": 1.8344, "step": 14100 }, { "epoch": 1.13, "grad_norm": 0.3464914858341217, "learning_rate": 1.696997301962633e-06, "loss": 1.8405, "step": 14125 }, { "epoch": 1.1320000000000001, "grad_norm": 0.3577348589897156, "learning_rate": 1.6016673387127646e-06, "loss": 1.843, "step": 14150 }, { "epoch": 1.134, "grad_norm": 0.3302348256111145, "learning_rate": 1.509071346816926e-06, "loss": 1.8377, "step": 14175 }, { "epoch": 1.1360000000000001, "grad_norm": 0.30688709020614624, "learning_rate": 1.4192118990299707e-06, "loss": 1.8389, "step": 14200 }, { "epoch": 1.138, "grad_norm": 0.33231252431869507, "learning_rate": 1.332091492072629e-06, "loss": 1.8411, "step": 14225 }, { "epoch": 1.1400000000000001, "grad_norm": 0.2914701998233795, "learning_rate": 1.2477125465620853e-06, "loss": 1.8338, "step": 14250 }, { "epoch": 1.142, "grad_norm": 0.3572983741760254, "learning_rate": 1.1660774069447877e-06, "loss": 1.8339, "step": 14275 }, { "epoch": 1.144, "grad_norm": 0.35410603880882263, "learning_rate": 1.0871883414312777e-06, "loss": 1.8429, "step": 14300 }, { "epoch": 1.146, "grad_norm": 0.3146873712539673, "learning_rate": 1.0110475419330967e-06, "loss": 1.8412, "step": 14325 }, { "epoch": 1.148, "grad_norm": 0.3227052390575409, "learning_rate": 9.376571240020227e-07, "loss": 1.8486, "step": 14350 }, { "epoch": 1.15, "grad_norm": 0.32136309146881104, "learning_rate": 8.670191267711736e-07, "loss": 1.8344, "step": 14375 }, { "epoch": 1.152, "grad_norm": 0.3202197253704071, "learning_rate": 7.991355128984079e-07, "loss": 1.8355, "step": 14400 }, { "epoch": 1.154, "grad_norm": 0.34478357434272766, "learning_rate": 7.340081685117905e-07, "loss": 1.841, "step": 14425 }, { "epoch": 1.156, "grad_norm": 0.31434378027915955, "learning_rate": 6.716389031571568e-07, "loss": 1.8303, "step": 14450 }, { "epoch": 1.158, "grad_norm": 0.4046516418457031, "learning_rate": 6.120294497478752e-07, "loss": 1.8359, "step": 14475 }, { "epoch": 1.16, "grad_norm": 0.3292851448059082, "learning_rate": 5.55181464516652e-07, "loss": 1.8404, "step": 14500 }, { "epoch": 1.162, "grad_norm": 0.33806750178337097, "learning_rate": 5.010965269695578e-07, "loss": 1.8363, "step": 14525 }, { "epoch": 1.164, "grad_norm": 0.3253093659877777, "learning_rate": 4.497761398421063e-07, "loss": 1.8441, "step": 14550 }, { "epoch": 1.166, "grad_norm": 0.3116908371448517, "learning_rate": 4.0122172905753266e-07, "loss": 1.8353, "step": 14575 }, { "epoch": 1.168, "grad_norm": 0.30682384967803955, "learning_rate": 3.554346436871581e-07, "loss": 1.8407, "step": 14600 }, { "epoch": 1.17, "grad_norm": 0.3538268506526947, "learning_rate": 3.12416155912898e-07, "loss": 1.8356, "step": 14625 }, { "epoch": 1.172, "grad_norm": 0.3013436496257782, "learning_rate": 2.721674609919345e-07, "loss": 1.8437, "step": 14650 }, { "epoch": 1.174, "grad_norm": 0.32243216037750244, "learning_rate": 2.3468967722347634e-07, "loss": 1.8427, "step": 14675 }, { "epoch": 1.176, "grad_norm": 0.3069095313549042, "learning_rate": 1.9998384591773944e-07, "loss": 1.8429, "step": 14700 }, { "epoch": 1.178, "grad_norm": 0.335223913192749, "learning_rate": 1.680509313669587e-07, "loss": 1.8435, "step": 14725 }, { "epoch": 1.18, "grad_norm": 0.2763690650463104, "learning_rate": 1.3889182081860962e-07, "loss": 1.8358, "step": 14750 }, { "epoch": 1.182, "grad_norm": 0.27810004353523254, "learning_rate": 1.1250732445080569e-07, "loss": 1.8373, "step": 14775 }, { "epoch": 1.184, "grad_norm": 0.33322063088417053, "learning_rate": 8.889817534969425e-08, "loss": 1.8439, "step": 14800 }, { "epoch": 1.186, "grad_norm": 0.3301320970058441, "learning_rate": 6.806502948918381e-08, "loss": 1.8379, "step": 14825 }, { "epoch": 1.188, "grad_norm": 0.3624110221862793, "learning_rate": 5.000846571264761e-08, "loss": 1.8408, "step": 14850 }, { "epoch": 1.19, "grad_norm": 0.29497313499450684, "learning_rate": 3.472898571690308e-08, "loss": 1.8397, "step": 14875 }, { "epoch": 1.192, "grad_norm": 0.3344693183898926, "learning_rate": 2.222701403818972e-08, "loss": 1.8387, "step": 14900 }, { "epoch": 1.194, "grad_norm": 0.2970846891403198, "learning_rate": 1.2502898040456234e-08, "loss": 1.8381, "step": 14925 }, { "epoch": 1.196, "grad_norm": 0.3378585875034332, "learning_rate": 5.5569079056794206e-09, "loss": 1.8478, "step": 14950 }, { "epoch": 1.198, "grad_norm": 0.3553116023540497, "learning_rate": 1.3892366263146272e-09, "loss": 1.8364, "step": 14975 }, { "epoch": 1.2, "grad_norm": 0.3316471576690674, "learning_rate": 0.0, "loss": 1.8436, "step": 15000 } ], "logging_steps": 25, "max_steps": 15000, "num_input_tokens_seen": 0, "num_train_epochs": 2, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 7.759076782495095e+18, "train_batch_size": 1, "trial_name": null, "trial_params": null }