checkpoint 5000

config.json

@@ -1,5 +1,5 @@
 {
-  "_name_or_path": "longcld/t5_small_qg_ae_hl",
+  "_name_or_path": "longcld/t5-small-itranslate-visquad-aqg",
   "architectures": [
     "T5ForConditionalGeneration"
   ],

optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:61ca5483cf24765acff036d2e23fe20317f74c40b17ac60e1b340b4582b19229
+oid sha256:d33210300a695d4ba9217d1c9d4b0cec06c8d25e624a960aa5903f2ef55e3567
 size 352532601

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6296811f54c196573d66c2f3cc6b61a2fa369ed05b52dada08f8f10bd5ad8205
+oid sha256:110ebe98cefaa938e92e8c3eb93fcb70eacffe249d115507d491681567f8ed77
 size 688496379

rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b25d300f7a5e5ff95572298950cffe6f9c35eb1f8f744a4347c21710ab5e4762
+oid sha256:9ebb5b2b28192f1263b7a83b5fd6852393d35c165b8351517043a7223b0c7ff6
 size 14503

scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a3ad98877762fe962733f53ef65640405f6f296b9845d88dfc2bf1c76c090318
+oid sha256:352aae60d3513959ecb7b597a266aae4cdb5d8d712ee800dc95dc3d6badc2b08
 size 623

trainer_state.json

@@ -1,1606 +1,316 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 3.520658961073469,
-  "global_step": 26500,
+  "epoch": 3.543441679213533,
+  "global_step": 5000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.01,
-      "learning_rate": 0.0009986714494486515,
-      "loss": 3.2529,
+      "epoch": 0.07,
+      "learning_rate": 9.929128277817152e-06,
+      "loss": 2.0901,
       "step": 100
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 0.000997342898897303,
-      "loss": 3.1978,
+      "epoch": 0.14,
+      "learning_rate": 9.858256555634302e-06,
+      "loss": 2.0677,
       "step": 200
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 0.0009960143483459547,
-      "loss": 3.1849,
+      "epoch": 0.21,
+      "learning_rate": 9.787384833451453e-06,
+      "loss": 2.0399,
       "step": 300
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 0.0009946857977946061,
-      "loss": 3.1539,
+      "epoch": 0.28,
+      "learning_rate": 9.716513111268606e-06,
+      "loss": 2.0587,
       "step": 400
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 0.0009933572472432576,
-      "loss": 3.1202,
+      "epoch": 0.35,
+      "learning_rate": 9.645641389085757e-06,
+      "loss": 2.0568,
       "step": 500
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 0.000992028696691909,
-      "loss": 3.0483,
+      "epoch": 0.43,
+      "learning_rate": 9.574769666902906e-06,
+      "loss": 2.0561,
       "step": 600
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 0.0009907001461405608,
-      "loss": 3.0669,
+      "epoch": 0.5,
+      "learning_rate": 9.503897944720057e-06,
+      "loss": 2.0387,
       "step": 700
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 0.0009893715955892122,
-      "loss": 3.0665,
+      "epoch": 0.57,
+      "learning_rate": 9.433026222537208e-06,
+      "loss": 2.0428,
       "step": 800
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 0.0009880430450378637,
-      "loss": 3.0196,
+      "epoch": 0.64,
+      "learning_rate": 9.362154500354358e-06,
+      "loss": 2.0512,
       "step": 900
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 0.0009867144944865152,
-      "loss": 3.009,
+      "epoch": 0.71,
+      "learning_rate": 9.29128277817151e-06,
+      "loss": 2.0404,
       "step": 1000
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 0.0009853859439351669,
-      "loss": 3.0046,
+      "epoch": 0.78,
+      "learning_rate": 9.220411055988662e-06,
+      "loss": 2.0291,
       "step": 1100
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 0.0009840573933838183,
-      "loss": 2.9368,
+      "epoch": 0.85,
+      "learning_rate": 9.149539333805813e-06,
+      "loss": 2.0441,
       "step": 1200
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 0.0009827288428324698,
-      "loss": 2.9478,
+      "epoch": 0.92,
+      "learning_rate": 9.078667611622963e-06,
+      "loss": 2.0471,
       "step": 1300
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 0.0009814002922811213,
-      "loss": 2.879,
+      "epoch": 0.99,
+      "learning_rate": 9.007795889440114e-06,
+      "loss": 2.0396,
       "step": 1400
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 0.0009800717417297728,
-      "loss": 2.8613,
+      "epoch": 1.06,
+      "learning_rate": 8.936924167257265e-06,
+      "loss": 2.0585,
       "step": 1500
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 0.0009787431911784245,
-      "loss": 2.9318,
+      "epoch": 1.13,
+      "learning_rate": 8.866052445074416e-06,
+      "loss": 2.0355,
       "step": 1600
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 0.000977414640627076,
-      "loss": 2.9453,
+      "epoch": 1.2,
+      "learning_rate": 8.795180722891567e-06,
+      "loss": 2.0598,
       "step": 1700
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 0.0009760860900757274,
-      "loss": 2.8773,
+      "epoch": 1.28,
+      "learning_rate": 8.724309000708718e-06,
+      "loss": 2.0299,
       "step": 1800
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 0.000974757539524379,
-      "loss": 2.9057,
+      "epoch": 1.35,
+      "learning_rate": 8.653437278525868e-06,
+      "loss": 2.0472,
       "step": 1900
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 0.0009734289889730305,
-      "loss": 2.8679,
+      "epoch": 1.42,
+      "learning_rate": 8.58256555634302e-06,
+      "loss": 2.0408,
       "step": 2000
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 0.0009721004384216819,
-      "loss": 2.8844,
+      "epoch": 1.49,
+      "learning_rate": 8.51169383416017e-06,
+      "loss": 2.0386,
       "step": 2100
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 0.0009707718878703334,
-      "loss": 2.8434,
+      "epoch": 1.56,
+      "learning_rate": 8.440822111977321e-06,
+      "loss": 2.0322,
       "step": 2200
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 0.0009694433373189851,
-      "loss": 2.8379,
+      "epoch": 1.63,
+      "learning_rate": 8.369950389794472e-06,
+      "loss": 2.0166,
       "step": 2300
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 0.0009681147867676366,
-      "loss": 2.8515,
+      "epoch": 1.7,
+      "learning_rate": 8.299078667611624e-06,
+      "loss": 2.0342,
       "step": 2400
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 0.000966786236216288,
-      "loss": 2.8525,
+      "epoch": 1.77,
+      "learning_rate": 8.228206945428775e-06,
+      "loss": 2.0222,
       "step": 2500
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 0.0009654576856649395,
-      "loss": 2.816,
+      "epoch": 1.84,
+      "learning_rate": 8.157335223245926e-06,
+      "loss": 2.0326,
       "step": 2600
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 0.0009641291351135911,
-      "loss": 2.8333,
+      "epoch": 1.91,
+      "learning_rate": 8.086463501063077e-06,
+      "loss": 2.0463,
       "step": 2700
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 0.0009628005845622427,
-      "loss": 2.7859,
+      "epoch": 1.98,
+      "learning_rate": 8.015591778880228e-06,
+      "loss": 2.0229,
       "step": 2800
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 0.0009614720340108941,
-      "loss": 2.8099,
+      "epoch": 2.06,
+      "learning_rate": 7.944720056697379e-06,
+      "loss": 2.0453,
       "step": 2900
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 0.0009601434834595456,
-      "loss": 2.8097,
+      "epoch": 2.13,
+      "learning_rate": 7.87384833451453e-06,
+      "loss": 2.0203,
       "step": 3000
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 0.0009588149329081972,
-      "loss": 2.7639,
+      "epoch": 2.2,
+      "learning_rate": 7.80297661233168e-06,
+      "loss": 2.0547,
       "step": 3100
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 0.0009574863823568487,
-      "loss": 2.7792,
+      "epoch": 2.27,
+      "learning_rate": 7.732104890148831e-06,
+      "loss": 2.0476,
       "step": 3200
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 0.0009561578318055003,
-      "loss": 2.7894,
+      "epoch": 2.34,
+      "learning_rate": 7.661233167965982e-06,
+      "loss": 2.0338,
       "step": 3300
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 0.0009548292812541517,
-      "loss": 2.7797,
+      "epoch": 2.41,
+      "learning_rate": 7.590361445783133e-06,
+      "loss": 2.0395,
       "step": 3400
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 0.0009535007307028033,
-      "loss": 2.7387,
+      "epoch": 2.48,
+      "learning_rate": 7.519489723600284e-06,
+      "loss": 2.0175,
       "step": 3500
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 0.0009521721801514548,
-      "loss": 2.7752,
+      "epoch": 2.55,
+      "learning_rate": 7.4486180014174345e-06,
+      "loss": 2.0198,
       "step": 3600
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 0.0009508436296001063,
-      "loss": 2.6998,
+      "epoch": 2.62,
+      "learning_rate": 7.377746279234585e-06,
+      "loss": 2.0296,
       "step": 3700
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 0.0009495150790487578,
-      "loss": 2.7701,
+      "epoch": 2.69,
+      "learning_rate": 7.306874557051738e-06,
+      "loss": 2.0298,
       "step": 3800
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 0.0009481865284974094,
-      "loss": 2.7543,
+      "epoch": 2.76,
+      "learning_rate": 7.236002834868888e-06,
+      "loss": 2.0262,
       "step": 3900
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 0.0009468579779460609,
-      "loss": 2.6621,
+      "epoch": 2.83,
+      "learning_rate": 7.165131112686039e-06,
+      "loss": 2.0039,
       "step": 4000
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 0.0009455294273947124,
-      "loss": 2.6957,
+      "epoch": 2.91,
+      "learning_rate": 7.0942593905031895e-06,
+      "loss": 2.0117,
       "step": 4100
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 0.0009442008768433638,
-      "loss": 2.7165,
+      "epoch": 2.98,
+      "learning_rate": 7.02338766832034e-06,
+      "loss": 2.0079,
       "step": 4200
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 0.0009428723262920155,
-      "loss": 2.6919,
+      "epoch": 3.05,
+      "learning_rate": 6.952515946137491e-06,
+      "loss": 2.0103,
       "step": 4300
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 0.000941543775740667,
-      "loss": 2.7239,
+      "epoch": 3.12,
+      "learning_rate": 6.881644223954642e-06,
+      "loss": 2.0129,
       "step": 4400
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 0.0009402152251893185,
-      "loss": 2.6743,
+      "epoch": 3.19,
+      "learning_rate": 6.810772501771794e-06,
+      "loss": 1.98,
       "step": 4500
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 0.0009388866746379699,
-      "loss": 2.708,
+      "epoch": 3.26,
+      "learning_rate": 6.7399007795889446e-06,
+      "loss": 1.9821,
       "step": 4600
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 0.0009375581240866215,
-      "loss": 2.7201,
+      "epoch": 3.33,
+      "learning_rate": 6.669029057406095e-06,
+      "loss": 1.995,
       "step": 4700
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 0.000936229573535273,
-      "loss": 2.7207,
+      "epoch": 3.4,
+      "learning_rate": 6.598157335223246e-06,
+      "loss": 1.9771,
       "step": 4800
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 0.0009349010229839246,
-      "loss": 2.7046,
+      "epoch": 3.47,
+      "learning_rate": 6.527285613040397e-06,
+      "loss": 1.9907,
       "step": 4900
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 0.000933572472432576,
-      "loss": 2.6758,
+      "epoch": 3.54,
+      "learning_rate": 6.456413890857548e-06,
+      "loss": 1.9802,
       "step": 5000
-    },
-    {
-      "epoch": 0.68,
-      "learning_rate": 0.0009322439218812276,
-      "loss": 2.6754,
-      "step": 5100
-    },
-    {
-      "epoch": 0.69,
-      "learning_rate": 0.0009309153713298791,
-      "loss": 2.6408,
-      "step": 5200
-    },
-    {
-      "epoch": 0.7,
-      "learning_rate": 0.0009295868207785306,
-      "loss": 2.6306,
-      "step": 5300
-    },
-    {
-      "epoch": 0.72,
-      "learning_rate": 0.0009282582702271822,
-      "loss": 2.6455,
-      "step": 5400
-    },
-    {
-      "epoch": 0.73,
-      "learning_rate": 0.0009269297196758337,
-      "loss": 2.6905,
-      "step": 5500
-    },
-    {
-      "epoch": 0.74,
-      "learning_rate": 0.0009256011691244852,
-      "loss": 2.6132,
-      "step": 5600
-    },
-    {
-      "epoch": 0.76,
-      "learning_rate": 0.0009242726185731367,
-      "loss": 2.6048,
-      "step": 5700
-    },
-    {
-      "epoch": 0.77,
-      "learning_rate": 0.0009229440680217882,
-      "loss": 2.5945,
-      "step": 5800
-    },
-    {
-      "epoch": 0.78,
-      "learning_rate": 0.0009216155174704398,
-      "loss": 2.6384,
-      "step": 5900
-    },
-    {
-      "epoch": 0.8,
-      "learning_rate": 0.0009202869669190913,
-      "loss": 2.615,
-      "step": 6000
-    },
-    {
-      "epoch": 0.81,
-      "learning_rate": 0.0009189584163677428,
-      "loss": 2.6281,
-      "step": 6100
-    },
-    {
-      "epoch": 0.82,
-      "learning_rate": 0.0009176298658163943,
-      "loss": 2.5965,
-      "step": 6200
-    },
-    {
-      "epoch": 0.84,
-      "learning_rate": 0.0009163013152650458,
-      "loss": 2.6704,
-      "step": 6300
-    },
-    {
-      "epoch": 0.85,
-      "learning_rate": 0.0009149727647136974,
-      "loss": 2.5665,
-      "step": 6400
-    },
-    {
-      "epoch": 0.86,
-      "learning_rate": 0.0009136442141623489,
-      "loss": 2.6324,
-      "step": 6500
-    },
-    {
-      "epoch": 0.88,
-      "learning_rate": 0.0009123156636110004,
-      "loss": 2.6258,
-      "step": 6600
-    },
-    {
-      "epoch": 0.89,
-      "learning_rate": 0.000910987113059652,
-      "loss": 2.6006,
-      "step": 6700
-    },
-    {
-      "epoch": 0.9,
-      "learning_rate": 0.0009096585625083034,
-      "loss": 2.5908,
-      "step": 6800
-    },
-    {
-      "epoch": 0.92,
-      "learning_rate": 0.000908330011956955,
-      "loss": 2.5855,
-      "step": 6900
-    },
-    {
-      "epoch": 0.93,
-      "learning_rate": 0.0009070014614056065,
-      "loss": 2.6421,
-      "step": 7000
-    },
-    {
-      "epoch": 0.94,
-      "learning_rate": 0.0009056729108542581,
-      "loss": 2.6084,
-      "step": 7100
-    },
-    {
-      "epoch": 0.96,
-      "learning_rate": 0.0009043443603029095,
-      "loss": 2.6102,
-      "step": 7200
-    },
-    {
-      "epoch": 0.97,
-      "learning_rate": 0.000903015809751561,
-      "loss": 2.5543,
-      "step": 7300
-    },
-    {
-      "epoch": 0.98,
-      "learning_rate": 0.0009016872592002126,
-      "loss": 2.5746,
-      "step": 7400
-    },
-    {
-      "epoch": 1.0,
-      "learning_rate": 0.0009003587086488642,
-      "loss": 2.5401,
-      "step": 7500
-    },
-    {
-      "epoch": 1.01,
-      "learning_rate": 0.0008990301580975156,
-      "loss": 2.489,
-      "step": 7600
-    },
-    {
-      "epoch": 1.02,
-      "learning_rate": 0.0008977016075461671,
-      "loss": 2.5211,
-      "step": 7700
-    },
-    {
-      "epoch": 1.04,
-      "learning_rate": 0.0008963730569948186,
-      "loss": 2.548,
-      "step": 7800
-    },
-    {
-      "epoch": 1.05,
-      "learning_rate": 0.0008950445064434703,
-      "loss": 2.5165,
-      "step": 7900
-    },
-    {
-      "epoch": 1.06,
-      "learning_rate": 0.0008937159558921218,
-      "loss": 2.4884,
-      "step": 8000
-    },
-    {
-      "epoch": 1.08,
-      "learning_rate": 0.0008923874053407732,
-      "loss": 2.4682,
-      "step": 8100
-    },
-    {
-      "epoch": 1.09,
-      "learning_rate": 0.0008910588547894247,
-      "loss": 2.4754,
-      "step": 8200
-    },
-    {
-      "epoch": 1.1,
-      "learning_rate": 0.0008897303042380763,
-      "loss": 2.5076,
-      "step": 8300
-    },
-    {
-      "epoch": 1.12,
-      "learning_rate": 0.0008884017536867279,
-      "loss": 2.5088,
-      "step": 8400
-    },
-    {
-      "epoch": 1.13,
-      "learning_rate": 0.0008870732031353793,
-      "loss": 2.4578,
-      "step": 8500
-    },
-    {
-      "epoch": 1.14,
-      "learning_rate": 0.0008857446525840308,
-      "loss": 2.4815,
-      "step": 8600
-    },
-    {
-      "epoch": 1.16,
-      "learning_rate": 0.0008844161020326824,
-      "loss": 2.5504,
-      "step": 8700
-    },
-    {
-      "epoch": 1.17,
-      "learning_rate": 0.0008830875514813339,
-      "loss": 2.4883,
-      "step": 8800
-    },
-    {
-      "epoch": 1.18,
-      "learning_rate": 0.0008817590009299854,
-      "loss": 2.4907,
-      "step": 8900
-    },
-    {
-      "epoch": 1.2,
-      "learning_rate": 0.0008804304503786369,
-      "loss": 2.4817,
-      "step": 9000
-    },
-    {
-      "epoch": 1.21,
-      "learning_rate": 0.0008791018998272885,
-      "loss": 2.4803,
-      "step": 9100
-    },
-    {
-      "epoch": 1.22,
-      "learning_rate": 0.00087777334927594,
-      "loss": 2.461,
-      "step": 9200
-    },
-    {
-      "epoch": 1.24,
-      "learning_rate": 0.0008764447987245914,
-      "loss": 2.4541,
-      "step": 9300
-    },
-    {
-      "epoch": 1.25,
-      "learning_rate": 0.000875116248173243,
-      "loss": 2.436,
-      "step": 9400
-    },
-    {
-      "epoch": 1.26,
-      "learning_rate": 0.0008737876976218946,
-      "loss": 2.4998,
-      "step": 9500
-    },
-    {
-      "epoch": 1.28,
-      "learning_rate": 0.0008724591470705461,
-      "loss": 2.4387,
-      "step": 9600
-    },
-    {
-      "epoch": 1.29,
-      "learning_rate": 0.0008711305965191976,
-      "loss": 2.5248,
-      "step": 9700
-    },
-    {
-      "epoch": 1.3,
-      "learning_rate": 0.000869802045967849,
-      "loss": 2.4285,
-      "step": 9800
-    },
-    {
-      "epoch": 1.32,
-      "learning_rate": 0.0008684734954165007,
-      "loss": 2.4363,
-      "step": 9900
-    },
-    {
-      "epoch": 1.33,
-      "learning_rate": 0.0008671449448651522,
-      "loss": 2.4629,
-      "step": 10000
-    },
-    {
-      "epoch": 1.34,
-      "learning_rate": 0.0008658163943138037,
-      "loss": 2.463,
-      "step": 10100
-    },
-    {
-      "epoch": 1.36,
-      "learning_rate": 0.0008644878437624551,
-      "loss": 2.5058,
-      "step": 10200
-    },
-    {
-      "epoch": 1.37,
-      "learning_rate": 0.0008631592932111067,
-      "loss": 2.4056,
-      "step": 10300
-    },
-    {
-      "epoch": 1.38,
-      "learning_rate": 0.0008618307426597583,
-      "loss": 2.4111,
-      "step": 10400
-    },
-    {
-      "epoch": 1.39,
-      "learning_rate": 0.0008605021921084098,
-      "loss": 2.4352,
-      "step": 10500
-    },
-    {
-      "epoch": 1.41,
-      "learning_rate": 0.0008591736415570612,
-      "loss": 2.4886,
-      "step": 10600
-    },
-    {
-      "epoch": 1.42,
-      "learning_rate": 0.0008578450910057127,
-      "loss": 2.4433,
-      "step": 10700
-    },
-    {
-      "epoch": 1.43,
-      "learning_rate": 0.0008565165404543643,
-      "loss": 2.471,
-      "step": 10800
-    },
-    {
-      "epoch": 1.45,
-      "learning_rate": 0.0008551879899030159,
-      "loss": 2.4446,
-      "step": 10900
-    },
-    {
-      "epoch": 1.46,
-      "learning_rate": 0.0008538594393516674,
-      "loss": 2.4567,
-      "step": 11000
-    },
-    {
-      "epoch": 1.47,
-      "learning_rate": 0.0008525308888003188,
-      "loss": 2.5048,
-      "step": 11100
-    },
-    {
-      "epoch": 1.49,
-      "learning_rate": 0.0008512023382489704,
-      "loss": 2.4119,
-      "step": 11200
-    },
-    {
-      "epoch": 1.5,
-      "learning_rate": 0.0008498737876976219,
-      "loss": 2.4098,
-      "step": 11300
-    },
-    {
-      "epoch": 1.51,
-      "learning_rate": 0.0008485452371462735,
-      "loss": 2.4108,
-      "step": 11400
-    },
-    {
-      "epoch": 1.53,
-      "learning_rate": 0.0008472166865949249,
-      "loss": 2.4016,
-      "step": 11500
-    },
-    {
-      "epoch": 1.54,
-      "learning_rate": 0.0008458881360435765,
-      "loss": 2.422,
-      "step": 11600
-    },
-    {
-      "epoch": 1.55,
-      "learning_rate": 0.000844559585492228,
-      "loss": 2.4645,
-      "step": 11700
-    },
-    {
-      "epoch": 1.57,
-      "learning_rate": 0.0008432310349408795,
-      "loss": 2.4218,
-      "step": 11800
-    },
-    {
-      "epoch": 1.58,
-      "learning_rate": 0.000841902484389531,
-      "loss": 2.4213,
-      "step": 11900
-    },
-    {
-      "epoch": 1.59,
-      "learning_rate": 0.0008405739338381826,
-      "loss": 2.389,
-      "step": 12000
-    },
-    {
-      "epoch": 1.61,
-      "learning_rate": 0.0008392453832868341,
-      "loss": 2.3751,
-      "step": 12100
-    },
-    {
-      "epoch": 1.62,
-      "learning_rate": 0.0008379168327354856,
-      "loss": 2.4552,
-      "step": 12200
-    },
-    {
-      "epoch": 1.63,
-      "learning_rate": 0.000836588282184137,
-      "loss": 2.4137,
-      "step": 12300
-    },
-    {
-      "epoch": 1.65,
-      "learning_rate": 0.0008352597316327887,
-      "loss": 2.3814,
-      "step": 12400
-    },
-    {
-      "epoch": 1.66,
-      "learning_rate": 0.0008339311810814402,
-      "loss": 2.3798,
-      "step": 12500
-    },
-    {
-      "epoch": 1.67,
-      "learning_rate": 0.0008326026305300917,
-      "loss": 2.384,
-      "step": 12600
-    },
-    {
-      "epoch": 1.69,
-      "learning_rate": 0.0008312740799787431,
-      "loss": 2.4628,
-      "step": 12700
-    },
-    {
-      "epoch": 1.7,
-      "learning_rate": 0.0008299455294273947,
-      "loss": 2.4149,
-      "step": 12800
-    },
-    {
-      "epoch": 1.71,
-      "learning_rate": 0.0008286169788760463,
-      "loss": 2.4083,
-      "step": 12900
-    },
-    {
-      "epoch": 1.73,
-      "learning_rate": 0.0008272884283246978,
-      "loss": 2.4374,
-      "step": 13000
-    },
-    {
-      "epoch": 1.74,
-      "learning_rate": 0.0008259598777733493,
-      "loss": 2.3741,
-      "step": 13100
-    },
-    {
-      "epoch": 1.75,
-      "learning_rate": 0.0008246313272220008,
-      "loss": 2.3797,
-      "step": 13200
-    },
-    {
-      "epoch": 1.77,
-      "learning_rate": 0.0008233027766706523,
-      "loss": 2.3884,
-      "step": 13300
-    },
-    {
-      "epoch": 1.78,
-      "learning_rate": 0.0008219742261193038,
-      "loss": 2.3931,
-      "step": 13400
-    },
-    {
-      "epoch": 1.79,
-      "learning_rate": 0.0008206456755679554,
-      "loss": 2.4227,
-      "step": 13500
-    },
-    {
-      "epoch": 1.81,
-      "learning_rate": 0.000819317125016607,
-      "loss": 2.417,
-      "step": 13600
-    },
-    {
-      "epoch": 1.82,
-      "learning_rate": 0.0008179885744652584,
-      "loss": 2.3735,
-      "step": 13700
-    },
-    {
-      "epoch": 1.83,
-      "learning_rate": 0.0008166600239139099,
-      "loss": 2.4168,
-      "step": 13800
-    },
-    {
-      "epoch": 1.85,
-      "learning_rate": 0.0008153314733625614,
-      "loss": 2.4211,
-      "step": 13900
-    },
-    {
-      "epoch": 1.86,
-      "learning_rate": 0.0008140029228112131,
-      "loss": 2.4075,
-      "step": 14000
-    },
-    {
-      "epoch": 1.87,
-      "learning_rate": 0.0008126743722598645,
-      "loss": 2.385,
-      "step": 14100
-    },
-    {
-      "epoch": 1.89,
-      "learning_rate": 0.000811345821708516,
-      "loss": 2.3975,
-      "step": 14200
-    },
-    {
-      "epoch": 1.9,
-      "learning_rate": 0.0008100172711571675,
-      "loss": 2.4189,
-      "step": 14300
-    },
-    {
-      "epoch": 1.91,
-      "learning_rate": 0.0008086887206058191,
-      "loss": 2.4126,
-      "step": 14400
-    },
-    {
-      "epoch": 1.93,
-      "learning_rate": 0.0008073601700544706,
-      "loss": 2.4129,
-      "step": 14500
-    },
-    {
-      "epoch": 1.94,
-      "learning_rate": 0.0008060316195031221,
-      "loss": 2.3709,
-      "step": 14600
-    },
-    {
-      "epoch": 1.95,
-      "learning_rate": 0.0008047030689517736,
-      "loss": 2.3723,
-      "step": 14700
-    },
-    {
-      "epoch": 1.97,
-      "learning_rate": 0.0008033745184004252,
-      "loss": 2.3315,
-      "step": 14800
-    },
-    {
-      "epoch": 1.98,
-      "learning_rate": 0.0008020459678490766,
-      "loss": 2.3557,
-      "step": 14900
-    },
-    {
-      "epoch": 1.99,
-      "learning_rate": 0.0008007174172977282,
-      "loss": 2.3327,
-      "step": 15000
-    },
-    {
-      "epoch": 2.01,
-      "learning_rate": 0.0007993888667463797,
-      "loss": 2.3288,
-      "step": 15100
-    },
-    {
-      "epoch": 2.02,
-      "learning_rate": 0.0007980603161950313,
-      "loss": 2.2879,
-      "step": 15200
-    },
-    {
-      "epoch": 2.03,
-      "learning_rate": 0.0007967317656436827,
-      "loss": 2.3298,
-      "step": 15300
-    },
-    {
-      "epoch": 2.05,
-      "learning_rate": 0.0007954032150923342,
-      "loss": 2.3611,
-      "step": 15400
-    },
-    {
-      "epoch": 2.06,
-      "learning_rate": 0.0007940746645409858,
-      "loss": 2.2829,
-      "step": 15500
-    },
-    {
-      "epoch": 2.07,
-      "learning_rate": 0.0007927461139896374,
-      "loss": 2.2791,
-      "step": 15600
-    },
-    {
-      "epoch": 2.09,
-      "learning_rate": 0.0007914175634382889,
-      "loss": 2.3107,
-      "step": 15700
-    },
-    {
-      "epoch": 2.1,
-      "learning_rate": 0.0007900890128869403,
-      "loss": 2.2678,
-      "step": 15800
-    },
-    {
-      "epoch": 2.11,
-      "learning_rate": 0.0007887604623355918,
-      "loss": 2.2461,
-      "step": 15900
-    },
-    {
-      "epoch": 2.13,
-      "learning_rate": 0.0007874319117842435,
-      "loss": 2.3019,
-      "step": 16000
-    },
-    {
-      "epoch": 2.14,
-      "learning_rate": 0.000786103361232895,
-      "loss": 2.3085,
-      "step": 16100
-    },
-    {
-      "epoch": 2.15,
-      "learning_rate": 0.0007847748106815464,
-      "loss": 2.2968,
-      "step": 16200
-    },
-    {
-      "epoch": 2.17,
-      "learning_rate": 0.0007834462601301979,
-      "loss": 2.3325,
-      "step": 16300
-    },
-    {
-      "epoch": 2.18,
-      "learning_rate": 0.0007821177095788495,
-      "loss": 2.2987,
-      "step": 16400
-    },
-    {
-      "epoch": 2.19,
-      "learning_rate": 0.0007807891590275011,
-      "loss": 2.3043,
-      "step": 16500
-    },
-    {
-      "epoch": 2.21,
-      "learning_rate": 0.0007794606084761525,
-      "loss": 2.2743,
-      "step": 16600
-    },
-    {
-      "epoch": 2.22,
-      "learning_rate": 0.000778132057924804,
-      "loss": 2.3369,
-      "step": 16700
-    },
-    {
-      "epoch": 2.23,
-      "learning_rate": 0.0007768035073734556,
-      "loss": 2.3084,
-      "step": 16800
-    },
-    {
-      "epoch": 2.25,
-      "learning_rate": 0.0007754749568221071,
-      "loss": 2.2688,
-      "step": 16900
-    },
-    {
-      "epoch": 2.26,
-      "learning_rate": 0.0007741464062707587,
-      "loss": 2.3203,
-      "step": 17000
-    },
-    {
-      "epoch": 2.27,
-      "learning_rate": 0.0007728178557194101,
-      "loss": 2.2956,
-      "step": 17100
-    },
-    {
-      "epoch": 2.29,
-      "learning_rate": 0.0007714893051680617,
-      "loss": 2.2939,
-      "step": 17200
-    },
-    {
-      "epoch": 2.3,
-      "learning_rate": 0.0007701607546167132,
-      "loss": 2.3379,
-      "step": 17300
-    },
-    {
-      "epoch": 2.31,
-      "learning_rate": 0.0007688322040653647,
-      "loss": 2.3161,
-      "step": 17400
-    },
-    {
-      "epoch": 2.32,
-      "learning_rate": 0.0007675036535140162,
-      "loss": 2.2423,
-      "step": 17500
-    },
-    {
-      "epoch": 2.34,
-      "learning_rate": 0.0007661751029626678,
-      "loss": 2.2982,
-      "step": 17600
-    },
-    {
-      "epoch": 2.35,
-      "learning_rate": 0.0007648465524113193,
-      "loss": 2.3019,
-      "step": 17700
-    },
-    {
-      "epoch": 2.36,
-      "learning_rate": 0.0007635180018599708,
-      "loss": 2.3133,
-      "step": 17800
-    },
-    {
-      "epoch": 2.38,
-      "learning_rate": 0.0007621894513086222,
-      "loss": 2.3209,
-      "step": 17900
-    },
-    {
-      "epoch": 2.39,
-      "learning_rate": 0.0007608609007572739,
-      "loss": 2.2685,
-      "step": 18000
-    },
-    {
-      "epoch": 2.4,
-      "learning_rate": 0.0007595323502059254,
-      "loss": 2.2641,
-      "step": 18100
-    },
-    {
-      "epoch": 2.42,
-      "learning_rate": 0.0007582037996545769,
-      "loss": 2.2284,
-      "step": 18200
-    },
-    {
-      "epoch": 2.43,
-      "learning_rate": 0.0007568752491032283,
-      "loss": 2.2937,
-      "step": 18300
-    },
-    {
-      "epoch": 2.44,
-      "learning_rate": 0.0007555466985518799,
-      "loss": 2.304,
-      "step": 18400
-    },
-    {
-      "epoch": 2.46,
-      "learning_rate": 0.0007542181480005315,
-      "loss": 2.2327,
-      "step": 18500
-    },
-    {
-      "epoch": 2.47,
-      "learning_rate": 0.000752889597449183,
-      "loss": 2.2515,
-      "step": 18600
-    },
-    {
-      "epoch": 2.48,
-      "learning_rate": 0.0007515610468978344,
-      "loss": 2.2844,
-      "step": 18700
-    },
-    {
-      "epoch": 2.5,
-      "learning_rate": 0.000750232496346486,
-      "loss": 2.2547,
-      "step": 18800
-    },
-    {
-      "epoch": 2.51,
-      "learning_rate": 0.0007489039457951375,
-      "loss": 2.2519,
-      "step": 18900
-    },
-    {
-      "epoch": 2.52,
-      "learning_rate": 0.0007475753952437891,
-      "loss": 2.2452,
-      "step": 19000
-    },
-    {
-      "epoch": 2.54,
-      "learning_rate": 0.0007462468446924406,
-      "loss": 2.2554,
-      "step": 19100
-    },
-    {
-      "epoch": 2.55,
-      "learning_rate": 0.0007449182941410921,
-      "loss": 2.2747,
-      "step": 19200
-    },
-    {
-      "epoch": 2.56,
-      "learning_rate": 0.0007435897435897436,
-      "loss": 2.2507,
-      "step": 19300
-    },
-    {
-      "epoch": 2.58,
-      "learning_rate": 0.0007422611930383951,
-      "loss": 2.2791,
-      "step": 19400
-    },
-    {
-      "epoch": 2.59,
-      "learning_rate": 0.0007409326424870467,
-      "loss": 2.2677,
-      "step": 19500
-    },
-    {
-      "epoch": 2.6,
-      "learning_rate": 0.0007396040919356982,
-      "loss": 2.3149,
-      "step": 19600
-    },
-    {
-      "epoch": 2.62,
-      "learning_rate": 0.0007382755413843497,
-      "loss": 2.27,
-      "step": 19700
-    },
-    {
-      "epoch": 2.63,
-      "learning_rate": 0.0007369469908330012,
-      "loss": 2.2977,
-      "step": 19800
-    },
-    {
-      "epoch": 2.64,
-      "learning_rate": 0.0007356184402816527,
-      "loss": 2.2718,
-      "step": 19900
-    },
-    {
-      "epoch": 2.66,
-      "learning_rate": 0.0007342898897303044,
-      "loss": 2.2379,
-      "step": 20000
-    },
-    {
-      "epoch": 2.67,
-      "learning_rate": 0.0007329613391789558,
-      "loss": 2.2691,
-      "step": 20100
-    },
-    {
-      "epoch": 2.68,
-      "learning_rate": 0.0007316327886276073,
-      "loss": 2.2407,
-      "step": 20200
-    },
-    {
-      "epoch": 2.7,
-      "learning_rate": 0.0007303042380762588,
-      "loss": 2.2707,
-      "step": 20300
-    },
-    {
-      "epoch": 2.71,
-      "learning_rate": 0.0007289756875249104,
-      "loss": 2.2163,
-      "step": 20400
-    },
-    {
-      "epoch": 2.72,
-      "learning_rate": 0.0007276471369735619,
-      "loss": 2.2541,
-      "step": 20500
-    },
-    {
-      "epoch": 2.74,
-      "learning_rate": 0.0007263185864222134,
-      "loss": 2.2595,
-      "step": 20600
-    },
-    {
-      "epoch": 2.75,
-      "learning_rate": 0.0007249900358708649,
-      "loss": 2.322,
-      "step": 20700
-    },
-    {
-      "epoch": 2.76,
-      "learning_rate": 0.0007236614853195165,
-      "loss": 2.2441,
-      "step": 20800
-    },
-    {
-      "epoch": 2.78,
-      "learning_rate": 0.0007223329347681679,
-      "loss": 2.2856,
-      "step": 20900
-    },
-    {
-      "epoch": 2.79,
-      "learning_rate": 0.0007210043842168195,
-      "loss": 2.198,
-      "step": 21000
-    },
-    {
-      "epoch": 2.8,
-      "learning_rate": 0.000719675833665471,
-      "loss": 2.2341,
-      "step": 21100
-    },
-    {
-      "epoch": 2.82,
-      "learning_rate": 0.0007183472831141226,
-      "loss": 2.2448,
-      "step": 21200
-    },
-    {
-      "epoch": 2.83,
-      "learning_rate": 0.000717018732562774,
-      "loss": 2.2708,
-      "step": 21300
-    },
-    {
-      "epoch": 2.84,
-      "learning_rate": 0.0007156901820114255,
-      "loss": 2.2393,
-      "step": 21400
-    },
-    {
-      "epoch": 2.86,
-      "learning_rate": 0.000714361631460077,
-      "loss": 2.2699,
-      "step": 21500
-    },
-    {
-      "epoch": 2.87,
-      "learning_rate": 0.0007130330809087286,
-      "loss": 2.2624,
-      "step": 21600
-    },
-    {
-      "epoch": 2.88,
-      "learning_rate": 0.0007117045303573802,
-      "loss": 2.282,
-      "step": 21700
-    },
-    {
-      "epoch": 2.9,
-      "learning_rate": 0.0007103759798060316,
-      "loss": 2.2396,
-      "step": 21800
-    },
-    {
-      "epoch": 2.91,
-      "learning_rate": 0.0007090474292546831,
-      "loss": 2.1954,
-      "step": 21900
-    },
-    {
-      "epoch": 2.92,
-      "learning_rate": 0.0007077188787033346,
-      "loss": 2.1875,
-      "step": 22000
-    },
-    {
-      "epoch": 2.94,
-      "learning_rate": 0.0007063903281519863,
-      "loss": 2.234,
-      "step": 22100
-    },
-    {
-      "epoch": 2.95,
-      "learning_rate": 0.0007050617776006377,
-      "loss": 2.2558,
-      "step": 22200
-    },
-    {
-      "epoch": 2.96,
-      "learning_rate": 0.0007037332270492892,
-      "loss": 2.2378,
-      "step": 22300
-    },
-    {
-      "epoch": 2.98,
-      "learning_rate": 0.0007024046764979407,
-      "loss": 2.2453,
-      "step": 22400
-    },
-    {
-      "epoch": 2.99,
-      "learning_rate": 0.0007010761259465923,
-      "loss": 2.252,
-      "step": 22500
-    },
-    {
-      "epoch": 3.0,
-      "learning_rate": 0.0006997475753952438,
-      "loss": 2.2443,
-      "step": 22600
-    },
-    {
-      "epoch": 3.02,
-      "learning_rate": 0.0006984190248438953,
-      "loss": 2.1547,
-      "step": 22700
-    },
-    {
-      "epoch": 3.03,
-      "learning_rate": 0.0006970904742925468,
-      "loss": 2.2195,
-      "step": 22800
-    },
-    {
-      "epoch": 3.04,
-      "learning_rate": 0.0006957619237411984,
-      "loss": 2.1418,
-      "step": 22900
-    },
-    {
-      "epoch": 3.06,
-      "learning_rate": 0.0006944333731898498,
-      "loss": 2.1434,
-      "step": 23000
-    },
-    {
-      "epoch": 3.07,
-      "learning_rate": 0.0006931048226385014,
-      "loss": 2.1736,
-      "step": 23100
-    },
-    {
-      "epoch": 3.08,
-      "learning_rate": 0.0006917762720871529,
-      "loss": 2.2078,
-      "step": 23200
-    },
-    {
-      "epoch": 3.1,
-      "learning_rate": 0.0006904477215358045,
-      "loss": 2.2087,
-      "step": 23300
-    },
-    {
-      "epoch": 3.11,
-      "learning_rate": 0.000689119170984456,
-      "loss": 2.2055,
-      "step": 23400
-    },
-    {
-      "epoch": 3.12,
-      "learning_rate": 0.0006877906204331074,
-      "loss": 2.1532,
-      "step": 23500
-    },
-    {
-      "epoch": 3.14,
-      "learning_rate": 0.000686462069881759,
-      "loss": 2.1247,
-      "step": 23600
-    },
-    {
-      "epoch": 3.15,
-      "learning_rate": 0.0006851335193304106,
-      "loss": 2.1472,
-      "step": 23700
-    },
-    {
-      "epoch": 3.16,
-      "learning_rate": 0.0006838049687790621,
-      "loss": 2.1628,
-      "step": 23800
-    },
-    {
-      "epoch": 3.18,
-      "learning_rate": 0.0006824764182277135,
-      "loss": 2.1119,
-      "step": 23900
-    },
-    {
-      "epoch": 3.19,
-      "learning_rate": 0.000681147867676365,
-      "loss": 2.1666,
-      "step": 24000
-    },
-    {
-      "epoch": 3.2,
-      "learning_rate": 0.0006798193171250167,
-      "loss": 2.1538,
-      "step": 24100
-    },
-    {
-      "epoch": 3.22,
-      "learning_rate": 0.0006784907665736682,
-      "loss": 2.1799,
-      "step": 24200
-    },
-    {
-      "epoch": 3.23,
-      "learning_rate": 0.0006771622160223196,
-      "loss": 2.1754,
-      "step": 24300
-    },
-    {
-      "epoch": 3.24,
-      "learning_rate": 0.0006758336654709711,
-      "loss": 2.1532,
-      "step": 24400
-    },
-    {
-      "epoch": 3.25,
-      "learning_rate": 0.0006745051149196227,
-      "loss": 2.1255,
-      "step": 24500
-    },
-    {
-      "epoch": 3.27,
-      "learning_rate": 0.0006731765643682743,
-      "loss": 2.1838,
-      "step": 24600
-    },
-    {
-      "epoch": 3.28,
-      "learning_rate": 0.0006718480138169257,
-      "loss": 2.1842,
-      "step": 24700
-    },
-    {
-      "epoch": 3.29,
-      "learning_rate": 0.0006705194632655772,
-      "loss": 2.1787,
-      "step": 24800
-    },
-    {
-      "epoch": 3.31,
-      "learning_rate": 0.0006691909127142288,
-      "loss": 2.2049,
-      "step": 24900
-    },
-    {
-      "epoch": 3.32,
-      "learning_rate": 0.0006678623621628803,
-      "loss": 2.1406,
-      "step": 25000
-    },
-    {
-      "epoch": 3.33,
-      "learning_rate": 0.0006665338116115319,
-      "loss": 2.1395,
-      "step": 25100
-    },
-    {
-      "epoch": 3.35,
-      "learning_rate": 0.0006652052610601833,
-      "loss": 2.1544,
-      "step": 25200
-    },
-    {
-      "epoch": 3.36,
-      "learning_rate": 0.0006638767105088349,
-      "loss": 2.1677,
-      "step": 25300
-    },
-    {
-      "epoch": 3.37,
-      "learning_rate": 0.0006625481599574864,
-      "loss": 2.1244,
-      "step": 25400
-    },
-    {
-      "epoch": 3.39,
-      "learning_rate": 0.0006612196094061379,
-      "loss": 2.1904,
-      "step": 25500
-    },
-    {
-      "epoch": 3.4,
-      "learning_rate": 0.0006598910588547894,
-      "loss": 2.1632,
-      "step": 25600
-    },
-    {
-      "epoch": 3.41,
-      "learning_rate": 0.000658562508303441,
-      "loss": 2.2005,
-      "step": 25700
-    },
-    {
-      "epoch": 3.43,
-      "learning_rate": 0.0006572339577520925,
-      "loss": 2.2277,
-      "step": 25800
-    },
-    {
-      "epoch": 3.44,
-      "learning_rate": 0.000655905407200744,
-      "loss": 2.1776,
-      "step": 25900
-    },
-    {
-      "epoch": 3.45,
-      "learning_rate": 0.0006545768566493954,
-      "loss": 2.2065,
-      "step": 26000
-    },
-    {
-      "epoch": 3.47,
-      "learning_rate": 0.0006532483060980471,
-      "loss": 2.1503,
-      "step": 26100
-    },
-    {
-      "epoch": 3.48,
-      "learning_rate": 0.0006519197555466986,
-      "loss": 2.1791,
-      "step": 26200
-    },
-    {
-      "epoch": 3.49,
-      "learning_rate": 0.0006505912049953501,
-      "loss": 2.1909,
-      "step": 26300
-    },
-    {
-      "epoch": 3.51,
-      "learning_rate": 0.0006492626544440015,
-      "loss": 2.1801,
-      "step": 26400
-    },
-    {
-      "epoch": 3.52,
-      "learning_rate": 0.0006479341038926531,
-      "loss": 2.1385,
-      "step": 26500
     }
   ],
-  "max_steps": 75270,
+  "max_steps": 14110,
   "num_train_epochs": 10,
-  "total_flos": 3.12587376750849e+17,
+  "total_flos": 3.580791724112394e+17,
   "trial_name": null,
   "trial_params": null
 }

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3454634fd5586c72136bb00fb38f3f6a0d3aa089ebb4ddc27a07c6040ccb2541
+oid sha256:f3b082deb8747520011b24c245585c7816d98d49fad37c4c863bceb3f091e0f4
 size 2607