End of training

all_results.json

@@ -1,9 +1,9 @@
 {
     "epoch": 1.0,
-    "total_flos": 9.08850744042455e+17,
-    "train_loss": 0.8321872255313152,
-    "train_runtime": 82908.8976,
-    "train_samples": 71495,
-    "train_samples_per_second": 0.862,
+    "total_flos": 3.1397459461543035e+18,
+    "train_loss": 0.4062042611626282,
+    "train_runtime": 286226.4367,
+    "train_samples": 246989,
+    "train_samples_per_second": 0.863,
     "train_steps_per_second": 0.072
 }

train_results.json

@@ -1,9 +1,9 @@
 {
     "epoch": 1.0,
-    "total_flos": 9.08850744042455e+17,
-    "train_loss": 0.8321872255313152,
-    "train_runtime": 82908.8976,
-    "train_samples": 71495,
-    "train_samples_per_second": 0.862,
+    "total_flos": 3.1397459461543035e+18,
+    "train_loss": 0.4062042611626282,
+    "train_runtime": 286226.4367,
+    "train_samples": 246989,
+    "train_samples_per_second": 0.863,
     "train_steps_per_second": 0.072
 }

trainer_state.json

@@ -3,439 +3,1461 @@
   "best_model_checkpoint": null,
   "epoch": 1.0,
   "eval_steps": 500,
-  "global_step": 5958,
+  "global_step": 20583,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.016784155756965426,
-      "grad_norm": 83914.40625,
-      "learning_rate": 4.916079221215173e-05,
-      "loss": 0.9807,
+      "epoch": 0.004858378273332362,
+      "grad_norm": 66137.25,
+      "learning_rate": 4.9757081086333386e-05,
+      "loss": 0.5598,
       "step": 100
     },
     {
-      "epoch": 0.03356831151393085,
-      "grad_norm": 97513.53125,
-      "learning_rate": 4.832158442430346e-05,
-      "loss": 0.9534,
+      "epoch": 0.009716756546664724,
+      "grad_norm": 60222.27734375,
+      "learning_rate": 4.951416217266677e-05,
+      "loss": 0.5419,
       "step": 200
     },
     {
-      "epoch": 0.050352467270896276,
-      "grad_norm": 80779.1171875,
-      "learning_rate": 4.748237663645519e-05,
-      "loss": 0.9194,
+      "epoch": 0.014575134819997084,
+      "grad_norm": 67718.328125,
+      "learning_rate": 4.927124325900015e-05,
+      "loss": 0.5491,
       "step": 300
     },
     {
-      "epoch": 0.0671366230278617,
-      "grad_norm": 71209.2421875,
-      "learning_rate": 4.664316884860692e-05,
-      "loss": 0.9223,
+      "epoch": 0.019433513093329448,
+      "grad_norm": 73856.1015625,
+      "learning_rate": 4.902832434533353e-05,
+      "loss": 0.5496,
       "step": 400
     },
     {
-      "epoch": 0.08392077878482712,
-      "grad_norm": 75543.140625,
-      "learning_rate": 4.5803961060758646e-05,
-      "loss": 0.9167,
+      "epoch": 0.024291891366661808,
+      "grad_norm": 60560.8359375,
+      "learning_rate": 4.878540543166691e-05,
+      "loss": 0.5381,
       "step": 500
     },
     {
-      "epoch": 0.10070493454179255,
-      "grad_norm": 81713.671875,
-      "learning_rate": 4.4964753272910375e-05,
-      "loss": 0.8878,
+      "epoch": 0.029150269639994168,
+      "grad_norm": 76622.125,
+      "learning_rate": 4.8542486518000295e-05,
+      "loss": 0.5397,
       "step": 600
     },
     {
-      "epoch": 0.11748909029875797,
-      "grad_norm": 79699.15625,
-      "learning_rate": 4.4125545485062104e-05,
-      "loss": 0.8801,
+      "epoch": 0.03400864791332653,
+      "grad_norm": 55521.8203125,
+      "learning_rate": 4.829956760433367e-05,
+      "loss": 0.5339,
       "step": 700
     },
     {
-      "epoch": 0.1342732460557234,
-      "grad_norm": 57691.8359375,
-      "learning_rate": 4.328633769721383e-05,
-      "loss": 0.9157,
+      "epoch": 0.038867026186658896,
+      "grad_norm": 59273.4921875,
+      "learning_rate": 4.8056648690667055e-05,
+      "loss": 0.5342,
       "step": 800
     },
     {
-      "epoch": 0.1510574018126888,
-      "grad_norm": 78234.8359375,
-      "learning_rate": 4.244712990936556e-05,
-      "loss": 0.8952,
+      "epoch": 0.043725404459991256,
+      "grad_norm": 63513.421875,
+      "learning_rate": 4.781372977700044e-05,
+      "loss": 0.5247,
       "step": 900
     },
     {
-      "epoch": 0.16784155756965424,
-      "grad_norm": 64203.4140625,
-      "learning_rate": 4.160792212151729e-05,
-      "loss": 0.9049,
+      "epoch": 0.048583782733323616,
+      "grad_norm": 61088.77734375,
+      "learning_rate": 4.757081086333382e-05,
+      "loss": 0.5302,
       "step": 1000
     },
     {
-      "epoch": 0.18462571332661967,
-      "grad_norm": 63400.09375,
-      "learning_rate": 4.076871433366902e-05,
-      "loss": 0.8827,
+      "epoch": 0.053442161006655976,
+      "grad_norm": 71392.75,
+      "learning_rate": 4.7327891949667204e-05,
+      "loss": 0.5261,
       "step": 1100
     },
     {
-      "epoch": 0.2014098690835851,
-      "grad_norm": 71029.0078125,
-      "learning_rate": 3.992950654582075e-05,
-      "loss": 0.9008,
+      "epoch": 0.058300539279988337,
+      "grad_norm": 72375.1328125,
+      "learning_rate": 4.708497303600059e-05,
+      "loss": 0.5153,
       "step": 1200
     },
     {
-      "epoch": 0.21819402484055053,
-      "grad_norm": 67397.421875,
-      "learning_rate": 3.9090298757972476e-05,
-      "loss": 0.889,
+      "epoch": 0.0631589175533207,
+      "grad_norm": 65540.171875,
+      "learning_rate": 4.6842054122333964e-05,
+      "loss": 0.5068,
       "step": 1300
     },
     {
-      "epoch": 0.23497818059751593,
-      "grad_norm": 63474.31640625,
-      "learning_rate": 3.8251090970124205e-05,
-      "loss": 0.8945,
+      "epoch": 0.06801729582665306,
+      "grad_norm": 63439.60546875,
+      "learning_rate": 4.659913520866735e-05,
+      "loss": 0.5087,
       "step": 1400
     },
     {
-      "epoch": 0.25176233635448136,
-      "grad_norm": 61201.125,
-      "learning_rate": 3.7411883182275934e-05,
-      "loss": 0.8427,
+      "epoch": 0.07287567409998542,
+      "grad_norm": 56190.2265625,
+      "learning_rate": 4.635621629500073e-05,
+      "loss": 0.5074,
       "step": 1500
     },
     {
-      "epoch": 0.2685464921114468,
-      "grad_norm": 77147.5390625,
-      "learning_rate": 3.657267539442766e-05,
-      "loss": 0.8659,
+      "epoch": 0.07773405237331779,
+      "grad_norm": 72489.0234375,
+      "learning_rate": 4.6113297381334114e-05,
+      "loss": 0.5036,
       "step": 1600
     },
     {
-      "epoch": 0.2853306478684122,
-      "grad_norm": 135552.828125,
-      "learning_rate": 3.573346760657939e-05,
-      "loss": 0.8756,
+      "epoch": 0.08259243064665014,
+      "grad_norm": 55422.61328125,
+      "learning_rate": 4.58703784676675e-05,
+      "loss": 0.4979,
       "step": 1700
     },
     {
-      "epoch": 0.3021148036253776,
-      "grad_norm": 62637.78515625,
-      "learning_rate": 3.489425981873112e-05,
-      "loss": 0.8552,
+      "epoch": 0.08745080891998251,
+      "grad_norm": 58258.125,
+      "learning_rate": 4.562745955400088e-05,
+      "loss": 0.5016,
       "step": 1800
     },
     {
-      "epoch": 0.3188989593823431,
-      "grad_norm": 86379.7578125,
-      "learning_rate": 3.405505203088285e-05,
-      "loss": 0.8667,
+      "epoch": 0.09230918719331486,
+      "grad_norm": 57958.55859375,
+      "learning_rate": 4.5384540640334264e-05,
+      "loss": 0.508,
       "step": 1900
     },
     {
-      "epoch": 0.3356831151393085,
-      "grad_norm": 70539.9921875,
-      "learning_rate": 3.321584424303458e-05,
-      "loss": 0.8466,
+      "epoch": 0.09716756546664723,
+      "grad_norm": 52961.26953125,
+      "learning_rate": 4.514162172666764e-05,
+      "loss": 0.4843,
       "step": 2000
     },
     {
-      "epoch": 0.35246727089627394,
-      "grad_norm": 59345.5234375,
-      "learning_rate": 3.2376636455186307e-05,
-      "loss": 0.8694,
+      "epoch": 0.1020259437399796,
+      "grad_norm": 57415.234375,
+      "learning_rate": 4.489870281300102e-05,
+      "loss": 0.4943,
       "step": 2100
     },
     {
-      "epoch": 0.36925142665323935,
-      "grad_norm": 72302.8828125,
-      "learning_rate": 3.1537428667338035e-05,
-      "loss": 0.8417,
+      "epoch": 0.10688432201331195,
+      "grad_norm": 49461.50390625,
+      "learning_rate": 4.46557838993344e-05,
+      "loss": 0.4817,
       "step": 2200
     },
     {
-      "epoch": 0.38603558241020475,
-      "grad_norm": 65996.3828125,
-      "learning_rate": 3.0698220879489764e-05,
-      "loss": 0.8473,
+      "epoch": 0.11174270028664432,
+      "grad_norm": 51018.04296875,
+      "learning_rate": 4.441286498566778e-05,
+      "loss": 0.4825,
       "step": 2300
     },
     {
-      "epoch": 0.4028197381671702,
-      "grad_norm": 84450.78125,
-      "learning_rate": 2.9859013091641493e-05,
-      "loss": 0.8247,
+      "epoch": 0.11660107855997667,
+      "grad_norm": 52488.46484375,
+      "learning_rate": 4.4169946072001166e-05,
+      "loss": 0.4786,
       "step": 2400
     },
     {
-      "epoch": 0.4196038939241356,
-      "grad_norm": 69873.5390625,
-      "learning_rate": 2.9019805303793218e-05,
-      "loss": 0.8501,
+      "epoch": 0.12145945683330904,
+      "grad_norm": 61512.5078125,
+      "learning_rate": 4.392702715833455e-05,
+      "loss": 0.4775,
       "step": 2500
     },
     {
-      "epoch": 0.43638804968110106,
-      "grad_norm": 69021.078125,
-      "learning_rate": 2.818059751594495e-05,
-      "loss": 0.8284,
+      "epoch": 0.1263178351066414,
+      "grad_norm": 56490.4296875,
+      "learning_rate": 4.368410824466793e-05,
+      "loss": 0.4813,
       "step": 2600
     },
     {
-      "epoch": 0.45317220543806647,
-      "grad_norm": 75120.1171875,
-      "learning_rate": 2.734138972809668e-05,
-      "loss": 0.847,
+      "epoch": 0.13117621337997376,
+      "grad_norm": 63387.01171875,
+      "learning_rate": 4.3441189331001316e-05,
+      "loss": 0.4744,
       "step": 2700
     },
     {
-      "epoch": 0.46995636119503187,
-      "grad_norm": 63536.984375,
-      "learning_rate": 2.6502181940248405e-05,
-      "loss": 0.8222,
+      "epoch": 0.13603459165330611,
+      "grad_norm": 62635.20703125,
+      "learning_rate": 4.31982704173347e-05,
+      "loss": 0.4758,
       "step": 2800
     },
     {
-      "epoch": 0.4867405169519973,
-      "grad_norm": 66156.6796875,
-      "learning_rate": 2.5662974152400137e-05,
-      "loss": 0.8227,
+      "epoch": 0.1408929699266385,
+      "grad_norm": 52577.96875,
+      "learning_rate": 4.2955351503668076e-05,
+      "loss": 0.4867,
       "step": 2900
     },
     {
-      "epoch": 0.5035246727089627,
-      "grad_norm": 69988.140625,
-      "learning_rate": 2.4823766364551865e-05,
-      "loss": 0.8048,
+      "epoch": 0.14575134819997085,
+      "grad_norm": 54731.5,
+      "learning_rate": 4.271243259000146e-05,
+      "loss": 0.4658,
       "step": 3000
     },
     {
-      "epoch": 0.5203088284659282,
-      "grad_norm": 71663.9921875,
-      "learning_rate": 2.398455857670359e-05,
-      "loss": 0.8351,
+      "epoch": 0.1506097264733032,
+      "grad_norm": 52055.84765625,
+      "learning_rate": 4.246951367633484e-05,
+      "loss": 0.4737,
       "step": 3100
     },
     {
-      "epoch": 0.5370929842228936,
-      "grad_norm": 64774.16796875,
-      "learning_rate": 2.3145350788855323e-05,
-      "loss": 0.8121,
+      "epoch": 0.15546810474663558,
+      "grad_norm": 48978.90234375,
+      "learning_rate": 4.2226594762668225e-05,
+      "loss": 0.4765,
       "step": 3200
     },
     {
-      "epoch": 0.553877139979859,
-      "grad_norm": 63534.6015625,
-      "learning_rate": 2.2306143001007052e-05,
-      "loss": 0.824,
+      "epoch": 0.16032648301996794,
+      "grad_norm": 54819.38671875,
+      "learning_rate": 4.198367584900161e-05,
+      "loss": 0.466,
       "step": 3300
     },
     {
-      "epoch": 0.5706612957368244,
-      "grad_norm": 73855.3984375,
-      "learning_rate": 2.1466935213158777e-05,
-      "loss": 0.817,
+      "epoch": 0.1651848612933003,
+      "grad_norm": 58637.75390625,
+      "learning_rate": 4.1740756935334985e-05,
+      "loss": 0.4725,
       "step": 3400
     },
     {
-      "epoch": 0.5874454514937899,
-      "grad_norm": 68886.3984375,
-      "learning_rate": 2.062772742531051e-05,
-      "loss": 0.8078,
+      "epoch": 0.17004323956663267,
+      "grad_norm": 55842.4609375,
+      "learning_rate": 4.149783802166837e-05,
+      "loss": 0.4724,
       "step": 3500
     },
     {
-      "epoch": 0.6042296072507553,
-      "grad_norm": 68186.78125,
-      "learning_rate": 1.9788519637462235e-05,
-      "loss": 0.8175,
+      "epoch": 0.17490161783996502,
+      "grad_norm": 52083.8125,
+      "learning_rate": 4.125491910800175e-05,
+      "loss": 0.4796,
       "step": 3600
     },
     {
-      "epoch": 0.6210137630077207,
-      "grad_norm": 74676.046875,
-      "learning_rate": 1.8949311849613967e-05,
-      "loss": 0.7929,
+      "epoch": 0.17975999611329738,
+      "grad_norm": 56764.9296875,
+      "learning_rate": 4.1012000194335135e-05,
+      "loss": 0.4736,
       "step": 3700
     },
     {
-      "epoch": 0.6377979187646862,
-      "grad_norm": 68078.203125,
-      "learning_rate": 1.8110104061765696e-05,
-      "loss": 0.7941,
+      "epoch": 0.18461837438662973,
+      "grad_norm": 50524.82421875,
+      "learning_rate": 4.076908128066851e-05,
+      "loss": 0.4627,
       "step": 3800
     },
     {
-      "epoch": 0.6545820745216515,
-      "grad_norm": 66256.53125,
-      "learning_rate": 1.727089627391742e-05,
-      "loss": 0.8264,
+      "epoch": 0.1894767526599621,
+      "grad_norm": 50128.8359375,
+      "learning_rate": 4.0526162367001894e-05,
+      "loss": 0.4717,
       "step": 3900
     },
     {
-      "epoch": 0.671366230278617,
-      "grad_norm": 71649.8125,
-      "learning_rate": 1.6431688486069153e-05,
-      "loss": 0.7731,
+      "epoch": 0.19433513093329446,
+      "grad_norm": 72415.296875,
+      "learning_rate": 4.028324345333528e-05,
+      "loss": 0.456,
       "step": 4000
     },
     {
-      "epoch": 0.6881503860355824,
-      "grad_norm": 70382.28125,
-      "learning_rate": 1.559248069822088e-05,
-      "loss": 0.8076,
+      "epoch": 0.19919350920662682,
+      "grad_norm": 57324.921875,
+      "learning_rate": 4.004032453966866e-05,
+      "loss": 0.4505,
       "step": 4100
     },
     {
-      "epoch": 0.7049345417925479,
-      "grad_norm": 70303.125,
-      "learning_rate": 1.4753272910372609e-05,
-      "loss": 0.7841,
+      "epoch": 0.2040518874799592,
+      "grad_norm": 60287.41796875,
+      "learning_rate": 3.9797405626002044e-05,
+      "loss": 0.4441,
       "step": 4200
     },
     {
-      "epoch": 0.7217186975495132,
-      "grad_norm": 62633.87109375,
-      "learning_rate": 1.391406512252434e-05,
-      "loss": 0.7781,
+      "epoch": 0.20891026575329155,
+      "grad_norm": 65722.1171875,
+      "learning_rate": 3.955448671233543e-05,
+      "loss": 0.4426,
       "step": 4300
     },
     {
-      "epoch": 0.7385028533064787,
-      "grad_norm": 69537.5625,
-      "learning_rate": 1.3074857334676067e-05,
-      "loss": 0.7846,
+      "epoch": 0.2137686440266239,
+      "grad_norm": 57532.5625,
+      "learning_rate": 3.931156779866881e-05,
+      "loss": 0.4532,
       "step": 4400
     },
     {
-      "epoch": 0.7552870090634441,
-      "grad_norm": 65905.9765625,
-      "learning_rate": 1.2235649546827795e-05,
-      "loss": 0.7818,
+      "epoch": 0.2186270222999563,
+      "grad_norm": 60035.2265625,
+      "learning_rate": 3.906864888500219e-05,
+      "loss": 0.443,
       "step": 4500
     },
     {
-      "epoch": 0.7720711648204095,
-      "grad_norm": 66610.0703125,
-      "learning_rate": 1.1396441758979524e-05,
-      "loss": 0.774,
+      "epoch": 0.22348540057328864,
+      "grad_norm": 50842.09375,
+      "learning_rate": 3.882572997133557e-05,
+      "loss": 0.4527,
       "step": 4600
     },
     {
-      "epoch": 0.788855320577375,
-      "grad_norm": 73451.1953125,
-      "learning_rate": 1.0557233971131253e-05,
-      "loss": 0.7768,
+      "epoch": 0.228343778846621,
+      "grad_norm": 53098.82421875,
+      "learning_rate": 3.858281105766895e-05,
+      "loss": 0.441,
       "step": 4700
     },
     {
-      "epoch": 0.8056394763343404,
-      "grad_norm": 85429.0390625,
-      "learning_rate": 9.718026183282982e-06,
-      "loss": 0.7805,
+      "epoch": 0.23320215711995335,
+      "grad_norm": 52847.52734375,
+      "learning_rate": 3.833989214400233e-05,
+      "loss": 0.4437,
       "step": 4800
     },
     {
-      "epoch": 0.8224236320913058,
-      "grad_norm": 63732.640625,
-      "learning_rate": 8.87881839543471e-06,
-      "loss": 0.7685,
+      "epoch": 0.23806053539328573,
+      "grad_norm": 55201.1953125,
+      "learning_rate": 3.809697323033571e-05,
+      "loss": 0.4534,
       "step": 4900
     },
     {
-      "epoch": 0.8392077878482712,
-      "grad_norm": 65406.63671875,
-      "learning_rate": 8.039610607586439e-06,
-      "loss": 0.7504,
+      "epoch": 0.24291891366661808,
+      "grad_norm": 55352.77734375,
+      "learning_rate": 3.7854054316669096e-05,
+      "loss": 0.4559,
       "step": 5000
     },
     {
-      "epoch": 0.8559919436052367,
-      "grad_norm": 74580.1484375,
-      "learning_rate": 7.200402819738168e-06,
-      "loss": 0.7814,
+      "epoch": 0.24777729193995043,
+      "grad_norm": 77840.125,
+      "learning_rate": 3.761113540300248e-05,
+      "loss": 0.4368,
       "step": 5100
     },
     {
-      "epoch": 0.8727760993622021,
-      "grad_norm": 73900.3671875,
-      "learning_rate": 6.361195031889897e-06,
-      "loss": 0.7749,
+      "epoch": 0.2526356702132828,
+      "grad_norm": 69213.84375,
+      "learning_rate": 3.736821648933586e-05,
+      "loss": 0.4428,
       "step": 5200
     },
     {
-      "epoch": 0.8895602551191675,
-      "grad_norm": 77174.4375,
-      "learning_rate": 5.5219872440416254e-06,
-      "loss": 0.7862,
+      "epoch": 0.25749404848661517,
+      "grad_norm": 61722.49609375,
+      "learning_rate": 3.7125297575669246e-05,
+      "loss": 0.444,
       "step": 5300
     },
     {
-      "epoch": 0.9063444108761329,
-      "grad_norm": 61536.1875,
-      "learning_rate": 4.682779456193353e-06,
-      "loss": 0.7669,
+      "epoch": 0.2623524267599475,
+      "grad_norm": 56948.3125,
+      "learning_rate": 3.688237866200262e-05,
+      "loss": 0.4423,
       "step": 5400
     },
     {
-      "epoch": 0.9231285666330984,
-      "grad_norm": 60535.73828125,
-      "learning_rate": 3.843571668345083e-06,
-      "loss": 0.7767,
+      "epoch": 0.2672108050332799,
+      "grad_norm": 50691.4453125,
+      "learning_rate": 3.6639459748336006e-05,
+      "loss": 0.4371,
       "step": 5500
     },
     {
-      "epoch": 0.9399127223900637,
-      "grad_norm": 57083.25,
-      "learning_rate": 3.0043638804968113e-06,
-      "loss": 0.7772,
+      "epoch": 0.27206918330661223,
+      "grad_norm": 48622.8671875,
+      "learning_rate": 3.639654083466939e-05,
+      "loss": 0.4459,
       "step": 5600
     },
     {
-      "epoch": 0.9566968781470292,
-      "grad_norm": 65630.703125,
-      "learning_rate": 2.16515609264854e-06,
-      "loss": 0.7749,
+      "epoch": 0.27692756157994464,
+      "grad_norm": 58408.26171875,
+      "learning_rate": 3.615362192100277e-05,
+      "loss": 0.4215,
       "step": 5700
     },
     {
-      "epoch": 0.9734810339039947,
-      "grad_norm": 66215.3359375,
-      "learning_rate": 1.3259483048002687e-06,
-      "loss": 0.782,
+      "epoch": 0.281785939853277,
+      "grad_norm": 48251.1953125,
+      "learning_rate": 3.5910703007336156e-05,
+      "loss": 0.4204,
       "step": 5800
     },
     {
-      "epoch": 0.99026518966096,
-      "grad_norm": 72915.515625,
-      "learning_rate": 4.867405169519974e-07,
-      "loss": 0.7722,
+      "epoch": 0.28664431812660934,
+      "grad_norm": 73270.4453125,
+      "learning_rate": 3.566778409366954e-05,
+      "loss": 0.4421,
       "step": 5900
     },
+    {
+      "epoch": 0.2915026963999417,
+      "grad_norm": 52048.3828125,
+      "learning_rate": 3.5424865180002915e-05,
+      "loss": 0.4438,
+      "step": 6000
+    },
+    {
+      "epoch": 0.29636107467327405,
+      "grad_norm": 57475.203125,
+      "learning_rate": 3.51819462663363e-05,
+      "loss": 0.418,
+      "step": 6100
+    },
+    {
+      "epoch": 0.3012194529466064,
+      "grad_norm": 55124.11328125,
+      "learning_rate": 3.493902735266968e-05,
+      "loss": 0.4347,
+      "step": 6200
+    },
+    {
+      "epoch": 0.3060778312199388,
+      "grad_norm": 62276.3671875,
+      "learning_rate": 3.469610843900306e-05,
+      "loss": 0.4192,
+      "step": 6300
+    },
+    {
+      "epoch": 0.31093620949327117,
+      "grad_norm": 52723.7421875,
+      "learning_rate": 3.445318952533644e-05,
+      "loss": 0.4413,
+      "step": 6400
+    },
+    {
+      "epoch": 0.3157945877666035,
+      "grad_norm": 53744.6484375,
+      "learning_rate": 3.4210270611669825e-05,
+      "loss": 0.4324,
+      "step": 6500
+    },
+    {
+      "epoch": 0.32065296603993587,
+      "grad_norm": 44454.984375,
+      "learning_rate": 3.396735169800321e-05,
+      "loss": 0.4269,
+      "step": 6600
+    },
+    {
+      "epoch": 0.3255113443132682,
+      "grad_norm": 52088.671875,
+      "learning_rate": 3.372443278433659e-05,
+      "loss": 0.4317,
+      "step": 6700
+    },
+    {
+      "epoch": 0.3303697225866006,
+      "grad_norm": 50359.18359375,
+      "learning_rate": 3.3481513870669974e-05,
+      "loss": 0.4102,
+      "step": 6800
+    },
+    {
+      "epoch": 0.33522810085993293,
+      "grad_norm": 54712.453125,
+      "learning_rate": 3.323859495700336e-05,
+      "loss": 0.4194,
+      "step": 6900
+    },
+    {
+      "epoch": 0.34008647913326534,
+      "grad_norm": 49144.39453125,
+      "learning_rate": 3.2995676043336734e-05,
+      "loss": 0.414,
+      "step": 7000
+    },
+    {
+      "epoch": 0.3449448574065977,
+      "grad_norm": 48121.30078125,
+      "learning_rate": 3.275275712967012e-05,
+      "loss": 0.4203,
+      "step": 7100
+    },
+    {
+      "epoch": 0.34980323567993005,
+      "grad_norm": 60245.484375,
+      "learning_rate": 3.25098382160035e-05,
+      "loss": 0.4227,
+      "step": 7200
+    },
+    {
+      "epoch": 0.3546616139532624,
+      "grad_norm": 53339.56640625,
+      "learning_rate": 3.226691930233688e-05,
+      "loss": 0.4235,
+      "step": 7300
+    },
+    {
+      "epoch": 0.35951999222659475,
+      "grad_norm": 61334.859375,
+      "learning_rate": 3.202400038867026e-05,
+      "loss": 0.4318,
+      "step": 7400
+    },
+    {
+      "epoch": 0.3643783704999271,
+      "grad_norm": 61368.46484375,
+      "learning_rate": 3.1781081475003643e-05,
+      "loss": 0.4223,
+      "step": 7500
+    },
+    {
+      "epoch": 0.36923674877325946,
+      "grad_norm": 67382.265625,
+      "learning_rate": 3.153816256133703e-05,
+      "loss": 0.4053,
+      "step": 7600
+    },
+    {
+      "epoch": 0.37409512704659187,
+      "grad_norm": 56739.12890625,
+      "learning_rate": 3.129524364767041e-05,
+      "loss": 0.4121,
+      "step": 7700
+    },
+    {
+      "epoch": 0.3789535053199242,
+      "grad_norm": 51203.3203125,
+      "learning_rate": 3.105232473400379e-05,
+      "loss": 0.432,
+      "step": 7800
+    },
+    {
+      "epoch": 0.3838118835932566,
+      "grad_norm": 58815.6484375,
+      "learning_rate": 3.080940582033717e-05,
+      "loss": 0.405,
+      "step": 7900
+    },
+    {
+      "epoch": 0.38867026186658893,
+      "grad_norm": 59941.88671875,
+      "learning_rate": 3.056648690667055e-05,
+      "loss": 0.425,
+      "step": 8000
+    },
+    {
+      "epoch": 0.3935286401399213,
+      "grad_norm": 53543.46484375,
+      "learning_rate": 3.0323567993003936e-05,
+      "loss": 0.4103,
+      "step": 8100
+    },
+    {
+      "epoch": 0.39838701841325364,
+      "grad_norm": 55630.40625,
+      "learning_rate": 3.008064907933732e-05,
+      "loss": 0.3964,
+      "step": 8200
+    },
+    {
+      "epoch": 0.40324539668658604,
+      "grad_norm": 52717.6953125,
+      "learning_rate": 2.9837730165670703e-05,
+      "loss": 0.3989,
+      "step": 8300
+    },
+    {
+      "epoch": 0.4081037749599184,
+      "grad_norm": 50914.84375,
+      "learning_rate": 2.9594811252004086e-05,
+      "loss": 0.4097,
+      "step": 8400
+    },
+    {
+      "epoch": 0.41296215323325075,
+      "grad_norm": 61387.0390625,
+      "learning_rate": 2.9351892338337466e-05,
+      "loss": 0.4051,
+      "step": 8500
+    },
+    {
+      "epoch": 0.4178205315065831,
+      "grad_norm": 45076.01953125,
+      "learning_rate": 2.910897342467085e-05,
+      "loss": 0.4012,
+      "step": 8600
+    },
+    {
+      "epoch": 0.42267890977991546,
+      "grad_norm": 54953.07421875,
+      "learning_rate": 2.8866054511004225e-05,
+      "loss": 0.4202,
+      "step": 8700
+    },
+    {
+      "epoch": 0.4275372880532478,
+      "grad_norm": 59405.3828125,
+      "learning_rate": 2.862313559733761e-05,
+      "loss": 0.3931,
+      "step": 8800
+    },
+    {
+      "epoch": 0.43239566632658016,
+      "grad_norm": 61979.93359375,
+      "learning_rate": 2.8380216683670992e-05,
+      "loss": 0.3987,
+      "step": 8900
+    },
+    {
+      "epoch": 0.4372540445999126,
+      "grad_norm": 46517.14453125,
+      "learning_rate": 2.813729777000437e-05,
+      "loss": 0.4135,
+      "step": 9000
+    },
+    {
+      "epoch": 0.4421124228732449,
+      "grad_norm": 57222.25,
+      "learning_rate": 2.7894378856337755e-05,
+      "loss": 0.4048,
+      "step": 9100
+    },
+    {
+      "epoch": 0.4469708011465773,
+      "grad_norm": 60419.09375,
+      "learning_rate": 2.7651459942671138e-05,
+      "loss": 0.4008,
+      "step": 9200
+    },
+    {
+      "epoch": 0.45182917941990963,
+      "grad_norm": 42610.984375,
+      "learning_rate": 2.740854102900452e-05,
+      "loss": 0.3975,
+      "step": 9300
+    },
+    {
+      "epoch": 0.456687557693242,
+      "grad_norm": 58822.0234375,
+      "learning_rate": 2.71656221153379e-05,
+      "loss": 0.3923,
+      "step": 9400
+    },
+    {
+      "epoch": 0.46154593596657434,
+      "grad_norm": 47564.5234375,
+      "learning_rate": 2.6922703201671284e-05,
+      "loss": 0.391,
+      "step": 9500
+    },
+    {
+      "epoch": 0.4664043142399067,
+      "grad_norm": 48113.20703125,
+      "learning_rate": 2.6679784288004668e-05,
+      "loss": 0.4064,
+      "step": 9600
+    },
+    {
+      "epoch": 0.4712626925132391,
+      "grad_norm": 49079.30078125,
+      "learning_rate": 2.6436865374338048e-05,
+      "loss": 0.4077,
+      "step": 9700
+    },
+    {
+      "epoch": 0.47612107078657145,
+      "grad_norm": 53815.7890625,
+      "learning_rate": 2.619394646067143e-05,
+      "loss": 0.3844,
+      "step": 9800
+    },
+    {
+      "epoch": 0.4809794490599038,
+      "grad_norm": 45962.41796875,
+      "learning_rate": 2.5951027547004814e-05,
+      "loss": 0.3925,
+      "step": 9900
+    },
+    {
+      "epoch": 0.48583782733323616,
+      "grad_norm": 50792.91015625,
+      "learning_rate": 2.570810863333819e-05,
+      "loss": 0.3838,
+      "step": 10000
+    },
+    {
+      "epoch": 0.4906962056065685,
+      "grad_norm": 44864.3046875,
+      "learning_rate": 2.5465189719671574e-05,
+      "loss": 0.3924,
+      "step": 10100
+    },
+    {
+      "epoch": 0.49555458387990087,
+      "grad_norm": 50477.6015625,
+      "learning_rate": 2.5222270806004954e-05,
+      "loss": 0.3875,
+      "step": 10200
+    },
+    {
+      "epoch": 0.5004129621532333,
+      "grad_norm": 46136.0859375,
+      "learning_rate": 2.4979351892338337e-05,
+      "loss": 0.3956,
+      "step": 10300
+    },
+    {
+      "epoch": 0.5052713404265656,
+      "grad_norm": 57640.59765625,
+      "learning_rate": 2.473643297867172e-05,
+      "loss": 0.38,
+      "step": 10400
+    },
+    {
+      "epoch": 0.510129718699898,
+      "grad_norm": 60448.453125,
+      "learning_rate": 2.4493514065005103e-05,
+      "loss": 0.3925,
+      "step": 10500
+    },
+    {
+      "epoch": 0.5149880969732303,
+      "grad_norm": 43668.453125,
+      "learning_rate": 2.4250595151338483e-05,
+      "loss": 0.3966,
+      "step": 10600
+    },
+    {
+      "epoch": 0.5198464752465627,
+      "grad_norm": 51975.83203125,
+      "learning_rate": 2.4007676237671866e-05,
+      "loss": 0.399,
+      "step": 10700
+    },
+    {
+      "epoch": 0.524704853519895,
+      "grad_norm": 60277.078125,
+      "learning_rate": 2.376475732400525e-05,
+      "loss": 0.3968,
+      "step": 10800
+    },
+    {
+      "epoch": 0.5295632317932274,
+      "grad_norm": 47762.875,
+      "learning_rate": 2.3521838410338633e-05,
+      "loss": 0.3937,
+      "step": 10900
+    },
+    {
+      "epoch": 0.5344216100665597,
+      "grad_norm": 62721.125,
+      "learning_rate": 2.327891949667201e-05,
+      "loss": 0.3871,
+      "step": 11000
+    },
+    {
+      "epoch": 0.5392799883398921,
+      "grad_norm": 41137.8515625,
+      "learning_rate": 2.3036000583005392e-05,
+      "loss": 0.3814,
+      "step": 11100
+    },
+    {
+      "epoch": 0.5441383666132245,
+      "grad_norm": 50033.2734375,
+      "learning_rate": 2.2793081669338776e-05,
+      "loss": 0.3987,
+      "step": 11200
+    },
+    {
+      "epoch": 0.5489967448865569,
+      "grad_norm": 50574.140625,
+      "learning_rate": 2.255016275567216e-05,
+      "loss": 0.3741,
+      "step": 11300
+    },
+    {
+      "epoch": 0.5538551231598893,
+      "grad_norm": 46732.82421875,
+      "learning_rate": 2.230724384200554e-05,
+      "loss": 0.3775,
+      "step": 11400
+    },
+    {
+      "epoch": 0.5587135014332216,
+      "grad_norm": 57961.8671875,
+      "learning_rate": 2.2064324928338922e-05,
+      "loss": 0.3826,
+      "step": 11500
+    },
+    {
+      "epoch": 0.563571879706554,
+      "grad_norm": 43375.69140625,
+      "learning_rate": 2.1821406014672302e-05,
+      "loss": 0.3823,
+      "step": 11600
+    },
+    {
+      "epoch": 0.5684302579798863,
+      "grad_norm": 49148.40234375,
+      "learning_rate": 2.1578487101005685e-05,
+      "loss": 0.3782,
+      "step": 11700
+    },
+    {
+      "epoch": 0.5732886362532187,
+      "grad_norm": 60182.703125,
+      "learning_rate": 2.133556818733907e-05,
+      "loss": 0.3804,
+      "step": 11800
+    },
+    {
+      "epoch": 0.578147014526551,
+      "grad_norm": 47590.46484375,
+      "learning_rate": 2.1092649273672448e-05,
+      "loss": 0.3974,
+      "step": 11900
+    },
+    {
+      "epoch": 0.5830053927998834,
+      "grad_norm": 65701.859375,
+      "learning_rate": 2.084973036000583e-05,
+      "loss": 0.3798,
+      "step": 12000
+    },
+    {
+      "epoch": 0.5878637710732157,
+      "grad_norm": 43476.69140625,
+      "learning_rate": 2.0606811446339215e-05,
+      "loss": 0.378,
+      "step": 12100
+    },
+    {
+      "epoch": 0.5927221493465481,
+      "grad_norm": 43104.83203125,
+      "learning_rate": 2.0363892532672595e-05,
+      "loss": 0.3718,
+      "step": 12200
+    },
+    {
+      "epoch": 0.5975805276198805,
+      "grad_norm": 58024.0546875,
+      "learning_rate": 2.0120973619005974e-05,
+      "loss": 0.3837,
+      "step": 12300
+    },
+    {
+      "epoch": 0.6024389058932128,
+      "grad_norm": 56071.4765625,
+      "learning_rate": 1.9878054705339358e-05,
+      "loss": 0.3789,
+      "step": 12400
+    },
+    {
+      "epoch": 0.6072972841665452,
+      "grad_norm": 56979.65625,
+      "learning_rate": 1.963513579167274e-05,
+      "loss": 0.3726,
+      "step": 12500
+    },
+    {
+      "epoch": 0.6121556624398776,
+      "grad_norm": 48315.94140625,
+      "learning_rate": 1.9392216878006124e-05,
+      "loss": 0.3697,
+      "step": 12600
+    },
+    {
+      "epoch": 0.61701404071321,
+      "grad_norm": 48739.546875,
+      "learning_rate": 1.9149297964339504e-05,
+      "loss": 0.3849,
+      "step": 12700
+    },
+    {
+      "epoch": 0.6218724189865423,
+      "grad_norm": 49214.71875,
+      "learning_rate": 1.8906379050672887e-05,
+      "loss": 0.3646,
+      "step": 12800
+    },
+    {
+      "epoch": 0.6267307972598747,
+      "grad_norm": 46501.2265625,
+      "learning_rate": 1.866346013700627e-05,
+      "loss": 0.3693,
+      "step": 12900
+    },
+    {
+      "epoch": 0.631589175533207,
+      "grad_norm": 69907.9609375,
+      "learning_rate": 1.842054122333965e-05,
+      "loss": 0.3737,
+      "step": 13000
+    },
+    {
+      "epoch": 0.6364475538065394,
+      "grad_norm": 47973.73046875,
+      "learning_rate": 1.817762230967303e-05,
+      "loss": 0.3875,
+      "step": 13100
+    },
+    {
+      "epoch": 0.6413059320798717,
+      "grad_norm": 58501.1796875,
+      "learning_rate": 1.7934703396006413e-05,
+      "loss": 0.3778,
+      "step": 13200
+    },
+    {
+      "epoch": 0.6461643103532041,
+      "grad_norm": 61768.96875,
+      "learning_rate": 1.7691784482339797e-05,
+      "loss": 0.384,
+      "step": 13300
+    },
+    {
+      "epoch": 0.6510226886265364,
+      "grad_norm": 46427.80859375,
+      "learning_rate": 1.744886556867318e-05,
+      "loss": 0.365,
+      "step": 13400
+    },
+    {
+      "epoch": 0.6558810668998688,
+      "grad_norm": 58254.9453125,
+      "learning_rate": 1.720594665500656e-05,
+      "loss": 0.3641,
+      "step": 13500
+    },
+    {
+      "epoch": 0.6607394451732012,
+      "grad_norm": 49971.00390625,
+      "learning_rate": 1.696302774133994e-05,
+      "loss": 0.363,
+      "step": 13600
+    },
+    {
+      "epoch": 0.6655978234465335,
+      "grad_norm": 56874.12890625,
+      "learning_rate": 1.6720108827673323e-05,
+      "loss": 0.3795,
+      "step": 13700
+    },
+    {
+      "epoch": 0.6704562017198659,
+      "grad_norm": 51813.33984375,
+      "learning_rate": 1.6477189914006706e-05,
+      "loss": 0.3809,
+      "step": 13800
+    },
+    {
+      "epoch": 0.6753145799931982,
+      "grad_norm": 51388.640625,
+      "learning_rate": 1.6234271000340086e-05,
+      "loss": 0.3649,
+      "step": 13900
+    },
+    {
+      "epoch": 0.6801729582665307,
+      "grad_norm": 50291.19921875,
+      "learning_rate": 1.599135208667347e-05,
+      "loss": 0.3609,
+      "step": 14000
+    },
+    {
+      "epoch": 0.685031336539863,
+      "grad_norm": 57660.2578125,
+      "learning_rate": 1.5748433173006852e-05,
+      "loss": 0.3719,
+      "step": 14100
+    },
+    {
+      "epoch": 0.6898897148131954,
+      "grad_norm": 67932.59375,
+      "learning_rate": 1.5505514259340236e-05,
+      "loss": 0.3685,
+      "step": 14200
+    },
+    {
+      "epoch": 0.6947480930865277,
+      "grad_norm": 54271.5234375,
+      "learning_rate": 1.5262595345673615e-05,
+      "loss": 0.376,
+      "step": 14300
+    },
+    {
+      "epoch": 0.6996064713598601,
+      "grad_norm": 45938.90234375,
+      "learning_rate": 1.5019676432006995e-05,
+      "loss": 0.3555,
+      "step": 14400
+    },
+    {
+      "epoch": 0.7044648496331924,
+      "grad_norm": 65276.1328125,
+      "learning_rate": 1.4776757518340378e-05,
+      "loss": 0.382,
+      "step": 14500
+    },
+    {
+      "epoch": 0.7093232279065248,
+      "grad_norm": 52780.53125,
+      "learning_rate": 1.453383860467376e-05,
+      "loss": 0.3627,
+      "step": 14600
+    },
+    {
+      "epoch": 0.7141816061798572,
+      "grad_norm": 66279.4375,
+      "learning_rate": 1.4290919691007143e-05,
+      "loss": 0.3726,
+      "step": 14700
+    },
+    {
+      "epoch": 0.7190399844531895,
+      "grad_norm": 48593.3359375,
+      "learning_rate": 1.4048000777340525e-05,
+      "loss": 0.3623,
+      "step": 14800
+    },
+    {
+      "epoch": 0.7238983627265219,
+      "grad_norm": 47858.83203125,
+      "learning_rate": 1.3805081863673908e-05,
+      "loss": 0.3605,
+      "step": 14900
+    },
+    {
+      "epoch": 0.7287567409998542,
+      "grad_norm": 55828.08203125,
+      "learning_rate": 1.3562162950007288e-05,
+      "loss": 0.3559,
+      "step": 15000
+    },
+    {
+      "epoch": 0.7336151192731866,
+      "grad_norm": 50614.34765625,
+      "learning_rate": 1.331924403634067e-05,
+      "loss": 0.3534,
+      "step": 15100
+    },
+    {
+      "epoch": 0.7384734975465189,
+      "grad_norm": 48779.75,
+      "learning_rate": 1.3076325122674053e-05,
+      "loss": 0.3593,
+      "step": 15200
+    },
+    {
+      "epoch": 0.7433318758198514,
+      "grad_norm": 50229.625,
+      "learning_rate": 1.2833406209007434e-05,
+      "loss": 0.3594,
+      "step": 15300
+    },
+    {
+      "epoch": 0.7481902540931837,
+      "grad_norm": 47616.875,
+      "learning_rate": 1.2590487295340816e-05,
+      "loss": 0.3417,
+      "step": 15400
+    },
+    {
+      "epoch": 0.7530486323665161,
+      "grad_norm": 51565.13671875,
+      "learning_rate": 1.2347568381674197e-05,
+      "loss": 0.3556,
+      "step": 15500
+    },
+    {
+      "epoch": 0.7579070106398484,
+      "grad_norm": 159462.1875,
+      "learning_rate": 1.210464946800758e-05,
+      "loss": 0.3641,
+      "step": 15600
+    },
+    {
+      "epoch": 0.7627653889131808,
+      "grad_norm": 54541.4375,
+      "learning_rate": 1.1861730554340962e-05,
+      "loss": 0.362,
+      "step": 15700
+    },
+    {
+      "epoch": 0.7676237671865132,
+      "grad_norm": 60987.12109375,
+      "learning_rate": 1.1618811640674344e-05,
+      "loss": 0.3573,
+      "step": 15800
+    },
+    {
+      "epoch": 0.7724821454598455,
+      "grad_norm": 51971.9765625,
+      "learning_rate": 1.1375892727007725e-05,
+      "loss": 0.3595,
+      "step": 15900
+    },
+    {
+      "epoch": 0.7773405237331779,
+      "grad_norm": 53555.78515625,
+      "learning_rate": 1.1132973813341108e-05,
+      "loss": 0.3617,
+      "step": 16000
+    },
+    {
+      "epoch": 0.7821989020065102,
+      "grad_norm": 50965.8984375,
+      "learning_rate": 1.0890054899674488e-05,
+      "loss": 0.356,
+      "step": 16100
+    },
+    {
+      "epoch": 0.7870572802798426,
+      "grad_norm": 52491.86328125,
+      "learning_rate": 1.0647135986007871e-05,
+      "loss": 0.3442,
+      "step": 16200
+    },
+    {
+      "epoch": 0.7919156585531749,
+      "grad_norm": 46180.28515625,
+      "learning_rate": 1.0404217072341253e-05,
+      "loss": 0.3618,
+      "step": 16300
+    },
+    {
+      "epoch": 0.7967740368265073,
+      "grad_norm": 43963.23828125,
+      "learning_rate": 1.0161298158674636e-05,
+      "loss": 0.3451,
+      "step": 16400
+    },
+    {
+      "epoch": 0.8016324150998396,
+      "grad_norm": 50149.4140625,
+      "learning_rate": 9.918379245008016e-06,
+      "loss": 0.3515,
+      "step": 16500
+    },
+    {
+      "epoch": 0.8064907933731721,
+      "grad_norm": 64846.9609375,
+      "learning_rate": 9.6754603313414e-06,
+      "loss": 0.3593,
+      "step": 16600
+    },
+    {
+      "epoch": 0.8113491716465044,
+      "grad_norm": 59240.46484375,
+      "learning_rate": 9.432541417674781e-06,
+      "loss": 0.345,
+      "step": 16700
+    },
+    {
+      "epoch": 0.8162075499198368,
+      "grad_norm": 51653.4765625,
+      "learning_rate": 9.189622504008162e-06,
+      "loss": 0.3486,
+      "step": 16800
+    },
+    {
+      "epoch": 0.8210659281931691,
+      "grad_norm": 56696.0703125,
+      "learning_rate": 8.946703590341544e-06,
+      "loss": 0.3514,
+      "step": 16900
+    },
+    {
+      "epoch": 0.8259243064665015,
+      "grad_norm": 44659.8515625,
+      "learning_rate": 8.703784676674927e-06,
+      "loss": 0.3538,
+      "step": 17000
+    },
+    {
+      "epoch": 0.8307826847398339,
+      "grad_norm": 62480.4453125,
+      "learning_rate": 8.460865763008307e-06,
+      "loss": 0.3553,
+      "step": 17100
+    },
+    {
+      "epoch": 0.8356410630131662,
+      "grad_norm": 49702.1796875,
+      "learning_rate": 8.21794684934169e-06,
+      "loss": 0.355,
+      "step": 17200
+    },
+    {
+      "epoch": 0.8404994412864986,
+      "grad_norm": 51918.6171875,
+      "learning_rate": 7.975027935675072e-06,
+      "loss": 0.3516,
+      "step": 17300
+    },
+    {
+      "epoch": 0.8453578195598309,
+      "grad_norm": 70460.1875,
+      "learning_rate": 7.732109022008453e-06,
+      "loss": 0.3485,
+      "step": 17400
+    },
+    {
+      "epoch": 0.8502161978331633,
+      "grad_norm": 39541.1328125,
+      "learning_rate": 7.489190108341836e-06,
+      "loss": 0.3682,
+      "step": 17500
+    },
+    {
+      "epoch": 0.8550745761064956,
+      "grad_norm": 55487.01171875,
+      "learning_rate": 7.246271194675218e-06,
+      "loss": 0.3634,
+      "step": 17600
+    },
+    {
+      "epoch": 0.859932954379828,
+      "grad_norm": 61263.55859375,
+      "learning_rate": 7.0033522810086005e-06,
+      "loss": 0.3661,
+      "step": 17700
+    },
+    {
+      "epoch": 0.8647913326531603,
+      "grad_norm": 51811.1875,
+      "learning_rate": 6.760433367341981e-06,
+      "loss": 0.3545,
+      "step": 17800
+    },
+    {
+      "epoch": 0.8696497109264928,
+      "grad_norm": 58568.05078125,
+      "learning_rate": 6.517514453675364e-06,
+      "loss": 0.3511,
+      "step": 17900
+    },
+    {
+      "epoch": 0.8745080891998251,
+      "grad_norm": 42418.73828125,
+      "learning_rate": 6.274595540008746e-06,
+      "loss": 0.3462,
+      "step": 18000
+    },
+    {
+      "epoch": 0.8793664674731575,
+      "grad_norm": 66980.625,
+      "learning_rate": 6.0316766263421275e-06,
+      "loss": 0.3429,
+      "step": 18100
+    },
+    {
+      "epoch": 0.8842248457464899,
+      "grad_norm": 51797.6875,
+      "learning_rate": 5.788757712675509e-06,
+      "loss": 0.3457,
+      "step": 18200
+    },
+    {
+      "epoch": 0.8890832240198222,
+      "grad_norm": 51049.6640625,
+      "learning_rate": 5.5458387990088915e-06,
+      "loss": 0.3462,
+      "step": 18300
+    },
+    {
+      "epoch": 0.8939416022931546,
+      "grad_norm": 53216.921875,
+      "learning_rate": 5.302919885342273e-06,
+      "loss": 0.3482,
+      "step": 18400
+    },
+    {
+      "epoch": 0.8987999805664869,
+      "grad_norm": 52944.14453125,
+      "learning_rate": 5.0600009716756545e-06,
+      "loss": 0.3429,
+      "step": 18500
+    },
+    {
+      "epoch": 0.9036583588398193,
+      "grad_norm": 55617.85546875,
+      "learning_rate": 4.817082058009037e-06,
+      "loss": 0.3496,
+      "step": 18600
+    },
+    {
+      "epoch": 0.9085167371131516,
+      "grad_norm": 51810.25390625,
+      "learning_rate": 4.5741631443424185e-06,
+      "loss": 0.3406,
+      "step": 18700
+    },
+    {
+      "epoch": 0.913375115386484,
+      "grad_norm": 62037.01171875,
+      "learning_rate": 4.331244230675801e-06,
+      "loss": 0.3393,
+      "step": 18800
+    },
+    {
+      "epoch": 0.9182334936598163,
+      "grad_norm": 55568.06640625,
+      "learning_rate": 4.088325317009182e-06,
+      "loss": 0.3412,
+      "step": 18900
+    },
+    {
+      "epoch": 0.9230918719331487,
+      "grad_norm": 55022.734375,
+      "learning_rate": 3.845406403342564e-06,
+      "loss": 0.3297,
+      "step": 19000
+    },
+    {
+      "epoch": 0.927950250206481,
+      "grad_norm": 59076.30078125,
+      "learning_rate": 3.6024874896759463e-06,
+      "loss": 0.3311,
+      "step": 19100
+    },
+    {
+      "epoch": 0.9328086284798134,
+      "grad_norm": 55583.625,
+      "learning_rate": 3.3595685760093283e-06,
+      "loss": 0.3334,
+      "step": 19200
+    },
+    {
+      "epoch": 0.9376670067531458,
+      "grad_norm": 44107.91015625,
+      "learning_rate": 3.1166496623427103e-06,
+      "loss": 0.3287,
+      "step": 19300
+    },
+    {
+      "epoch": 0.9425253850264782,
+      "grad_norm": 43654.265625,
+      "learning_rate": 2.8737307486760922e-06,
+      "loss": 0.3501,
+      "step": 19400
+    },
+    {
+      "epoch": 0.9473837632998106,
+      "grad_norm": 53803.21875,
+      "learning_rate": 2.630811835009474e-06,
+      "loss": 0.3418,
+      "step": 19500
+    },
+    {
+      "epoch": 0.9522421415731429,
+      "grad_norm": 67202.1953125,
+      "learning_rate": 2.3878929213428557e-06,
+      "loss": 0.3343,
+      "step": 19600
+    },
+    {
+      "epoch": 0.9571005198464753,
+      "grad_norm": 57509.7734375,
+      "learning_rate": 2.1449740076762377e-06,
+      "loss": 0.3397,
+      "step": 19700
+    },
+    {
+      "epoch": 0.9619588981198076,
+      "grad_norm": 45354.30859375,
+      "learning_rate": 1.9020550940096197e-06,
+      "loss": 0.3467,
+      "step": 19800
+    },
+    {
+      "epoch": 0.96681727639314,
+      "grad_norm": 51548.08984375,
+      "learning_rate": 1.6591361803430016e-06,
+      "loss": 0.3619,
+      "step": 19900
+    },
+    {
+      "epoch": 0.9716756546664723,
+      "grad_norm": 65316.8515625,
+      "learning_rate": 1.4162172666763834e-06,
+      "loss": 0.3216,
+      "step": 20000
+    },
+    {
+      "epoch": 0.9765340329398047,
+      "grad_norm": 48864.44921875,
+      "learning_rate": 1.1732983530097653e-06,
+      "loss": 0.3303,
+      "step": 20100
+    },
+    {
+      "epoch": 0.981392411213137,
+      "grad_norm": 48878.0,
+      "learning_rate": 9.303794393431472e-07,
+      "loss": 0.3492,
+      "step": 20200
+    },
+    {
+      "epoch": 0.9862507894864694,
+      "grad_norm": 43658.50390625,
+      "learning_rate": 6.874605256765293e-07,
+      "loss": 0.3306,
+      "step": 20300
+    },
+    {
+      "epoch": 0.9911091677598017,
+      "grad_norm": 58298.01171875,
+      "learning_rate": 4.4454161200991114e-07,
+      "loss": 0.3366,
+      "step": 20400
+    },
+    {
+      "epoch": 0.9959675460331341,
+      "grad_norm": 49216.18359375,
+      "learning_rate": 2.01622698343293e-07,
+      "loss": 0.3288,
+      "step": 20500
+    },
     {
       "epoch": 1.0,
-      "step": 5958,
-      "total_flos": 9.08850744042455e+17,
-      "train_loss": 0.8321872255313152,
-      "train_runtime": 82908.8976,
-      "train_samples_per_second": 0.862,
+      "step": 20583,
+      "total_flos": 3.1397459461543035e+18,
+      "train_loss": 0.4062042611626282,
+      "train_runtime": 286226.4367,
+      "train_samples_per_second": 0.863,
       "train_steps_per_second": 0.072
     }
   ],
   "logging_steps": 100,
-  "max_steps": 5958,
+  "max_steps": 20583,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 1,
-  "save_steps": 5958,
+  "save_steps": 20583,
   "stateful_callbacks": {
     "TrainerControl": {
       "args": {
@@ -448,7 +1470,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 9.08850744042455e+17,
+  "total_flos": 3.1397459461543035e+18,
   "train_batch_size": 12,
   "trial_name": null,
   "trial_params": null