Training in progress, epoch 1, checkpoint

checkpoint-1469/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:701d2071e676254cb7f486c1583f3b336c0ab235a36ec3a2025604309a9f014d
+oid sha256:53c42c0c60f3a151e7114837956e0760c5eb1aa98d486b9eadd570d90d3a13b2
 size 383295784

checkpoint-1469/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:eaff70510894b4e6680f626eacbecc09f01ab54ea65441557aadde172c4c0421
+oid sha256:e9c13f041754af610d61326226a20221fcd3024befc25d14fbdff1d075ea0498
 size 184533306

checkpoint-1469/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e2c5e1fd5d6bb74e9a611fd5587f335900288ad79fa0962b80a02eb52bde4304
+oid sha256:988e45b9c526ba869349b80dffa451430720ffba34a1c2e327fcbcb8ad100739
 size 14308

checkpoint-1469/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d51456e725db3a4810896373a397c9561592927e76bb2a2bf3bbb17ab1a7f915
+oid sha256:cf84e86953867ec379949a4a8bbd4904efa2bcb4480a1c61c43be3b1893e3384
 size 1064

checkpoint-1469/trainer_state.json

@@ -10,531 +10,531 @@
   "log_history": [
     {
       "epoch": 0.013614703880190605,
-      "grad_norm": 46436.91796875,
-      "learning_rate": 1.6304347826086958e-07,
-      "loss": 2.0803,
+      "grad_norm": 46457.98828125,
+      "learning_rate": 8.163265306122449e-07,
+      "loss": 2.0792,
       "step": 20
     },
     {
       "epoch": 0.02722940776038121,
-      "grad_norm": 35970.4375,
-      "learning_rate": 3.2608695652173915e-07,
-      "loss": 2.0801,
+      "grad_norm": 35946.40234375,
+      "learning_rate": 1.6326530612244897e-06,
+      "loss": 2.0791,
       "step": 40
     },
     {
       "epoch": 0.04084411164057182,
-      "grad_norm": 24038.876953125,
-      "learning_rate": 4.891304347826088e-07,
-      "loss": 2.0801,
+      "grad_norm": 23938.412109375,
+      "learning_rate": 2.4489795918367347e-06,
+      "loss": 2.0782,
       "step": 60
     },
     {
       "epoch": 0.05445881552076242,
-      "grad_norm": 31534.74609375,
-      "learning_rate": 6.521739130434783e-07,
-      "loss": 2.0799,
+      "grad_norm": 32852.3359375,
+      "learning_rate": 3.2653061224489794e-06,
+      "loss": 2.0771,
       "step": 80
     },
     {
       "epoch": 0.06807351940095303,
-      "grad_norm": 31741.2421875,
-      "learning_rate": 8.152173913043478e-07,
-      "loss": 2.0797,
+      "grad_norm": 32399.712890625,
+      "learning_rate": 4.081632653061225e-06,
+      "loss": 2.0752,
       "step": 100
     },
     {
       "epoch": 0.08168822328114364,
-      "grad_norm": 29444.419921875,
-      "learning_rate": 9.782608695652175e-07,
-      "loss": 2.0789,
+      "grad_norm": 30975.478515625,
+      "learning_rate": 4.897959183673469e-06,
+      "loss": 2.0722,
       "step": 120
     },
     {
       "epoch": 0.09530292716133425,
-      "grad_norm": 31825.087890625,
-      "learning_rate": 1.141304347826087e-06,
-      "loss": 2.079,
+      "grad_norm": 35666.22265625,
+      "learning_rate": 5.7142857142857145e-06,
+      "loss": 2.0674,
       "step": 140
     },
     {
       "epoch": 0.10891763104152484,
-      "grad_norm": 31622.044921875,
-      "learning_rate": 1.3043478260869566e-06,
-      "loss": 2.0783,
+      "grad_norm": 36289.2578125,
+      "learning_rate": 6.530612244897959e-06,
+      "loss": 2.0555,
       "step": 160
     },
     {
       "epoch": 0.12253233492171545,
-      "grad_norm": 24459.58203125,
-      "learning_rate": 1.4673913043478262e-06,
-      "loss": 2.0774,
+      "grad_norm": 34320.37890625,
+      "learning_rate": 7.346938775510204e-06,
+      "loss": 2.0414,
       "step": 180
     },
     {
       "epoch": 0.13614703880190607,
-      "grad_norm": 29978.267578125,
-      "learning_rate": 1.6304347826086955e-06,
-      "loss": 2.0771,
+      "grad_norm": 49061.66796875,
+      "learning_rate": 8.16326530612245e-06,
+      "loss": 2.0297,
       "step": 200
     },
     {
       "epoch": 0.14976174268209666,
-      "grad_norm": 31998.130859375,
-      "learning_rate": 1.7934782608695653e-06,
-      "loss": 2.0766,
+      "grad_norm": 60060.45703125,
+      "learning_rate": 8.979591836734694e-06,
+      "loss": 2.0094,
       "step": 220
     },
     {
       "epoch": 0.16337644656228728,
-      "grad_norm": 34736.546875,
-      "learning_rate": 1.956521739130435e-06,
-      "loss": 2.075,
+      "grad_norm": 80072.6015625,
+      "learning_rate": 9.795918367346939e-06,
+      "loss": 1.9789,
       "step": 240
     },
     {
       "epoch": 0.17699115044247787,
-      "grad_norm": 36555.73046875,
-      "learning_rate": 2.1195652173913046e-06,
-      "loss": 2.0723,
+      "grad_norm": 95368.8828125,
+      "learning_rate": 1.0612244897959184e-05,
+      "loss": 1.9378,
       "step": 260
     },
     {
       "epoch": 0.1906058543226685,
-      "grad_norm": 25291.138671875,
-      "learning_rate": 2.282608695652174e-06,
-      "loss": 2.0698,
+      "grad_norm": 70028.4765625,
+      "learning_rate": 1.1428571428571429e-05,
+      "loss": 1.9166,
       "step": 280
     },
     {
       "epoch": 0.2042205582028591,
-      "grad_norm": 29135.255859375,
-      "learning_rate": 2.4456521739130433e-06,
-      "loss": 2.0654,
+      "grad_norm": 90220.5859375,
+      "learning_rate": 1.2244897959183674e-05,
+      "loss": 1.8492,
       "step": 300
     },
     {
       "epoch": 0.21783526208304968,
-      "grad_norm": 18903.984375,
-      "learning_rate": 2.6086956521739132e-06,
-      "loss": 2.0625,
+      "grad_norm": 59154.35546875,
+      "learning_rate": 1.3061224489795918e-05,
+      "loss": 1.8556,
       "step": 320
     },
     {
       "epoch": 0.2314499659632403,
-      "grad_norm": 37406.1953125,
-      "learning_rate": 2.771739130434783e-06,
-      "loss": 2.0563,
+      "grad_norm": 129406.875,
+      "learning_rate": 1.3877551020408165e-05,
+      "loss": 1.8048,
       "step": 340
     },
     {
       "epoch": 0.2450646698434309,
-      "grad_norm": 46206.06640625,
-      "learning_rate": 2.9347826086956523e-06,
-      "loss": 2.0437,
+      "grad_norm": 176714.859375,
+      "learning_rate": 1.4693877551020408e-05,
+      "loss": 1.743,
       "step": 360
     },
     {
       "epoch": 0.2586793737236215,
-      "grad_norm": 43209.96484375,
-      "learning_rate": 2.994840189193063e-06,
-      "loss": 2.0383,
+      "grad_norm": 225369.375,
+      "learning_rate": 1.5510204081632655e-05,
+      "loss": 1.708,
       "step": 380
     },
     {
       "epoch": 0.27229407760381213,
-      "grad_norm": 42192.875,
-      "learning_rate": 2.9862405045148345e-06,
-      "loss": 2.0308,
+      "grad_norm": 324853.53125,
+      "learning_rate": 1.63265306122449e-05,
+      "loss": 1.7032,
       "step": 400
     },
     {
       "epoch": 0.2859087814840027,
-      "grad_norm": 35841.48046875,
-      "learning_rate": 2.977640819836606e-06,
-      "loss": 2.0333,
+      "grad_norm": 77990.5,
+      "learning_rate": 1.7142857142857142e-05,
+      "loss": 1.7964,
       "step": 420
     },
     {
       "epoch": 0.2995234853641933,
-      "grad_norm": 49910.5234375,
-      "learning_rate": 2.969041135158378e-06,
-      "loss": 2.0314,
+      "grad_norm": 243999.53125,
+      "learning_rate": 1.7959183673469387e-05,
+      "loss": 1.7472,
       "step": 440
     },
     {
       "epoch": 0.3131381892443839,
-      "grad_norm": 50317.3046875,
-      "learning_rate": 2.960441450480149e-06,
-      "loss": 2.0,
+      "grad_norm": 402316.78125,
+      "learning_rate": 1.8775510204081636e-05,
+      "loss": 1.6196,
       "step": 460
     },
     {
       "epoch": 0.32675289312457456,
-      "grad_norm": 38199.23046875,
-      "learning_rate": 2.951841765801921e-06,
-      "loss": 1.9918,
+      "grad_norm": 173016.109375,
+      "learning_rate": 1.9591836734693877e-05,
+      "loss": 1.6543,
       "step": 480
     },
     {
       "epoch": 0.34036759700476515,
-      "grad_norm": 49415.7734375,
-      "learning_rate": 2.9432420811236923e-06,
-      "loss": 2.005,
+      "grad_norm": 166789.75,
+      "learning_rate": 2.0408163265306123e-05,
+      "loss": 1.7865,
       "step": 500
     },
     {
       "epoch": 0.34036759700476515,
       "eval_accuracy": 0.1981891348088531,
-      "eval_loss": 2.0004630088806152,
-      "eval_runtime": 302.489,
-      "eval_samples_per_second": 9.858,
-      "eval_steps_per_second": 1.233,
+      "eval_loss": 1.6593838930130005,
+      "eval_runtime": 296.9022,
+      "eval_samples_per_second": 10.044,
+      "eval_steps_per_second": 1.256,
       "step": 500
     },
     {
       "epoch": 0.35398230088495575,
-      "grad_norm": 55542.02734375,
-      "learning_rate": 2.934642396445464e-06,
-      "loss": 1.9813,
+      "grad_norm": 160284.609375,
+      "learning_rate": 2.1224489795918368e-05,
+      "loss": 1.6528,
       "step": 520
     },
     {
       "epoch": 0.36759700476514634,
-      "grad_norm": 72033.6328125,
-      "learning_rate": 2.9260427117672353e-06,
-      "loss": 1.9823,
+      "grad_norm": 195971.28125,
+      "learning_rate": 2.2040816326530613e-05,
+      "loss": 1.7909,
       "step": 540
     },
     {
       "epoch": 0.381211708645337,
-      "grad_norm": 75101.5,
-      "learning_rate": 2.9174430270890068e-06,
-      "loss": 1.9645,
+      "grad_norm": 152683.9375,
+      "learning_rate": 2.2857142857142858e-05,
+      "loss": 1.6901,
       "step": 560
     },
     {
       "epoch": 0.3948264125255276,
-      "grad_norm": 36483.359375,
-      "learning_rate": 2.9088433424107783e-06,
-      "loss": 1.9719,
+      "grad_norm": 188117.421875,
+      "learning_rate": 2.3673469387755103e-05,
+      "loss": 1.7382,
       "step": 580
     },
     {
       "epoch": 0.4084411164057182,
-      "grad_norm": 85846.6328125,
-      "learning_rate": 2.9002436577325497e-06,
-      "loss": 1.9678,
+      "grad_norm": 221595.171875,
+      "learning_rate": 2.448979591836735e-05,
+      "loss": 1.7579,
       "step": 600
     },
     {
       "epoch": 0.42205582028590877,
-      "grad_norm": 77670.828125,
-      "learning_rate": 2.8916439730543216e-06,
-      "loss": 1.9434,
+      "grad_norm": 250219.484375,
+      "learning_rate": 2.5306122448979594e-05,
+      "loss": 1.6842,
       "step": 620
     },
     {
       "epoch": 0.43567052416609936,
-      "grad_norm": 128272.1796875,
-      "learning_rate": 2.8830442883760927e-06,
-      "loss": 1.9346,
+      "grad_norm": 362297.6875,
+      "learning_rate": 2.6122448979591835e-05,
+      "loss": 1.6874,
       "step": 640
     },
     {
       "epoch": 0.44928522804629,
-      "grad_norm": 74940.3984375,
-      "learning_rate": 2.8744446036978646e-06,
-      "loss": 1.9191,
+      "grad_norm": 268312.65625,
+      "learning_rate": 2.6938775510204084e-05,
+      "loss": 1.6339,
       "step": 660
     },
     {
       "epoch": 0.4628999319264806,
-      "grad_norm": 96815.6015625,
-      "learning_rate": 2.865844919019636e-06,
-      "loss": 1.8909,
+      "grad_norm": 301583.75,
+      "learning_rate": 2.775510204081633e-05,
+      "loss": 1.6268,
       "step": 680
     },
     {
       "epoch": 0.4765146358066712,
-      "grad_norm": 81240.6875,
-      "learning_rate": 2.8572452343414075e-06,
-      "loss": 1.921,
+      "grad_norm": 168292.046875,
+      "learning_rate": 2.857142857142857e-05,
+      "loss": 1.7985,
       "step": 700
     },
     {
       "epoch": 0.4901293396868618,
-      "grad_norm": 91377.9765625,
-      "learning_rate": 2.848645549663179e-06,
-      "loss": 1.9241,
+      "grad_norm": 274193.40625,
+      "learning_rate": 2.9387755102040816e-05,
+      "loss": 1.7689,
       "step": 720
     },
     {
       "epoch": 0.5037440435670524,
-      "grad_norm": 58107.17578125,
-      "learning_rate": 2.8400458649849505e-06,
-      "loss": 1.907,
+      "grad_norm": 142488.421875,
+      "learning_rate": 2.9977307110438728e-05,
+      "loss": 1.7201,
       "step": 740
     },
     {
       "epoch": 0.517358747447243,
-      "grad_norm": 100952.0234375,
-      "learning_rate": 2.8314461803067224e-06,
-      "loss": 1.8606,
+      "grad_norm": 273465.96875,
+      "learning_rate": 2.9886535552193645e-05,
+      "loss": 1.6105,
       "step": 760
     },
     {
       "epoch": 0.5309734513274337,
-      "grad_norm": 113918.6796875,
-      "learning_rate": 2.8228464956284935e-06,
-      "loss": 1.8657,
+      "grad_norm": 408773.3125,
+      "learning_rate": 2.9795763993948565e-05,
+      "loss": 1.6843,
       "step": 780
     },
     {
       "epoch": 0.5445881552076243,
-      "grad_norm": 83428.9375,
-      "learning_rate": 2.8142468109502654e-06,
-      "loss": 1.877,
+      "grad_norm": 187610.140625,
+      "learning_rate": 2.970499243570348e-05,
+      "loss": 1.7259,
       "step": 800
     },
     {
       "epoch": 0.5582028590878149,
-      "grad_norm": 64413.90234375,
-      "learning_rate": 2.805647126272037e-06,
-      "loss": 1.8967,
+      "grad_norm": 141979.921875,
+      "learning_rate": 2.9614220877458398e-05,
+      "loss": 1.7572,
       "step": 820
     },
     {
       "epoch": 0.5718175629680055,
-      "grad_norm": 88485.0703125,
-      "learning_rate": 2.7970474415938083e-06,
-      "loss": 1.8517,
+      "grad_norm": 183132.390625,
+      "learning_rate": 2.9523449319213314e-05,
+      "loss": 1.7094,
       "step": 840
     },
     {
       "epoch": 0.585432266848196,
-      "grad_norm": 145627.515625,
-      "learning_rate": 2.78844775691558e-06,
-      "loss": 1.8351,
+      "grad_norm": 299271.59375,
+      "learning_rate": 2.943267776096823e-05,
+      "loss": 1.7294,
       "step": 860
     },
     {
       "epoch": 0.5990469707283866,
-      "grad_norm": 111430.3828125,
-      "learning_rate": 2.7798480722373513e-06,
-      "loss": 1.8828,
+      "grad_norm": 207611.09375,
+      "learning_rate": 2.9341906202723147e-05,
+      "loss": 1.7864,
       "step": 880
     },
     {
       "epoch": 0.6126616746085772,
-      "grad_norm": 119978.2265625,
-      "learning_rate": 2.7712483875591228e-06,
-      "loss": 1.8287,
+      "grad_norm": 262759.0625,
+      "learning_rate": 2.9251134644478064e-05,
+      "loss": 1.7288,
       "step": 900
     },
     {
       "epoch": 0.6262763784887678,
-      "grad_norm": 109119.8046875,
-      "learning_rate": 2.7626487028808942e-06,
-      "loss": 1.7565,
+      "grad_norm": 291027.125,
+      "learning_rate": 2.916036308623298e-05,
+      "loss": 1.6042,
       "step": 920
     },
     {
       "epoch": 0.6398910823689585,
-      "grad_norm": 168422.890625,
-      "learning_rate": 2.754049018202666e-06,
-      "loss": 1.8571,
+      "grad_norm": 355390.71875,
+      "learning_rate": 2.90695915279879e-05,
+      "loss": 1.7426,
       "step": 940
     },
     {
       "epoch": 0.6535057862491491,
-      "grad_norm": 72286.7265625,
-      "learning_rate": 2.745449333524437e-06,
-      "loss": 1.8037,
+      "grad_norm": 87989.2578125,
+      "learning_rate": 2.8978819969742813e-05,
+      "loss": 1.6938,
       "step": 960
     },
     {
       "epoch": 0.6671204901293397,
-      "grad_norm": 165114.03125,
-      "learning_rate": 2.736849648846209e-06,
-      "loss": 1.8221,
+      "grad_norm": 302982.09375,
+      "learning_rate": 2.888804841149773e-05,
+      "loss": 1.8035,
       "step": 980
     },
     {
       "epoch": 0.6807351940095303,
-      "grad_norm": 139429.90625,
-      "learning_rate": 2.7282499641679806e-06,
-      "loss": 1.8316,
+      "grad_norm": 197021.3125,
+      "learning_rate": 2.879727685325265e-05,
+      "loss": 1.7716,
       "step": 1000
     },
     {
       "epoch": 0.6807351940095303,
       "eval_accuracy": 0.1981891348088531,
-      "eval_loss": 1.8303836584091187,
-      "eval_runtime": 286.8908,
-      "eval_samples_per_second": 10.394,
-      "eval_steps_per_second": 1.3,
+      "eval_loss": 1.6633707284927368,
+      "eval_runtime": 291.4938,
+      "eval_samples_per_second": 10.23,
+      "eval_steps_per_second": 1.28,
       "step": 1000
     },
     {
       "epoch": 0.6943498978897209,
-      "grad_norm": 180090.1875,
-      "learning_rate": 2.719650279489752e-06,
-      "loss": 1.8362,
+      "grad_norm": 348515.0625,
+      "learning_rate": 2.8706505295007566e-05,
+      "loss": 1.7593,
       "step": 1020
     },
     {
       "epoch": 0.7079646017699115,
-      "grad_norm": 118638.21875,
-      "learning_rate": 2.7110505948115235e-06,
-      "loss": 1.8059,
+      "grad_norm": 252458.609375,
+      "learning_rate": 2.8615733736762483e-05,
+      "loss": 1.7406,
       "step": 1040
     },
     {
       "epoch": 0.7215793056501021,
-      "grad_norm": 108668.46875,
-      "learning_rate": 2.7024509101332954e-06,
-      "loss": 1.8377,
+      "grad_norm": 179026.3125,
+      "learning_rate": 2.8524962178517396e-05,
+      "loss": 1.806,
       "step": 1060
     },
     {
       "epoch": 0.7351940095302927,
-      "grad_norm": 111934.5625,
-      "learning_rate": 2.6938512254550665e-06,
-      "loss": 1.7902,
+      "grad_norm": 230475.78125,
+      "learning_rate": 2.8434190620272316e-05,
+      "loss": 1.7327,
       "step": 1080
     },
     {
       "epoch": 0.7488087134104833,
-      "grad_norm": 129614.65625,
-      "learning_rate": 2.6852515407768384e-06,
-      "loss": 1.7476,
+      "grad_norm": 155040.359375,
+      "learning_rate": 2.8343419062027232e-05,
+      "loss": 1.6378,
       "step": 1100
     },
     {
       "epoch": 0.762423417290674,
-      "grad_norm": 132005.140625,
-      "learning_rate": 2.67665185609861e-06,
-      "loss": 1.791,
+      "grad_norm": 200565.359375,
+      "learning_rate": 2.825264750378215e-05,
+      "loss": 1.7561,
       "step": 1120
     },
     {
       "epoch": 0.7760381211708646,
-      "grad_norm": 111099.4921875,
-      "learning_rate": 2.6680521714203813e-06,
-      "loss": 1.8005,
+      "grad_norm": 160959.6875,
+      "learning_rate": 2.8161875945537065e-05,
+      "loss": 1.7322,
       "step": 1140
     },
     {
       "epoch": 0.7896528250510552,
-      "grad_norm": 186854.171875,
-      "learning_rate": 2.659452486742153e-06,
-      "loss": 1.7101,
+      "grad_norm": 277028.65625,
+      "learning_rate": 2.8071104387291985e-05,
+      "loss": 1.6283,
       "step": 1160
     },
     {
       "epoch": 0.8032675289312458,
-      "grad_norm": 116679.1328125,
-      "learning_rate": 2.6508528020639243e-06,
-      "loss": 1.7432,
+      "grad_norm": 145514.578125,
+      "learning_rate": 2.79803328290469e-05,
+      "loss": 1.696,
       "step": 1180
     },
     {
       "epoch": 0.8168822328114363,
-      "grad_norm": 148153.296875,
-      "learning_rate": 2.642253117385696e-06,
-      "loss": 1.7575,
+      "grad_norm": 274652.6875,
+      "learning_rate": 2.7889561270801815e-05,
+      "loss": 1.7019,
       "step": 1200
     },
     {
       "epoch": 0.8304969366916269,
-      "grad_norm": 78890.7578125,
-      "learning_rate": 2.6336534327074673e-06,
-      "loss": 1.7866,
+      "grad_norm": 108995.4453125,
+      "learning_rate": 2.779878971255673e-05,
+      "loss": 1.7189,
       "step": 1220
     },
     {
       "epoch": 0.8441116405718175,
-      "grad_norm": 73448.65625,
-      "learning_rate": 2.625053748029239e-06,
-      "loss": 1.8399,
+      "grad_norm": 84566.8046875,
+      "learning_rate": 2.770801815431165e-05,
+      "loss": 1.8068,
       "step": 1240
     },
     {
       "epoch": 0.8577263444520081,
-      "grad_norm": 214759.703125,
-      "learning_rate": 2.6164540633510106e-06,
-      "loss": 1.8554,
+      "grad_norm": 239363.8125,
+      "learning_rate": 2.7617246596066568e-05,
+      "loss": 1.8275,
       "step": 1260
     },
     {
       "epoch": 0.8713410483321987,
-      "grad_norm": 132226.390625,
-      "learning_rate": 2.607854378672782e-06,
-      "loss": 1.7432,
+      "grad_norm": 213879.875,
+      "learning_rate": 2.752647503782148e-05,
+      "loss": 1.7145,
       "step": 1280
     },
     {
       "epoch": 0.8849557522123894,
-      "grad_norm": 139430.578125,
-      "learning_rate": 2.5992546939945536e-06,
-      "loss": 1.7099,
+      "grad_norm": 201732.859375,
+      "learning_rate": 2.74357034795764e-05,
+      "loss": 1.648,
       "step": 1300
     },
     {
       "epoch": 0.89857045609258,
-      "grad_norm": 132929.875,
-      "learning_rate": 2.590655009316325e-06,
-      "loss": 1.7122,
+      "grad_norm": 262597.78125,
+      "learning_rate": 2.7344931921331318e-05,
+      "loss": 1.6619,
       "step": 1320
     },
     {
       "epoch": 0.9121851599727706,
-      "grad_norm": 246533.125,
-      "learning_rate": 2.5820553246380966e-06,
-      "loss": 1.7024,
+      "grad_norm": 343452.15625,
+      "learning_rate": 2.7254160363086234e-05,
+      "loss": 1.6732,
       "step": 1340
     },
     {
       "epoch": 0.9257998638529612,
-      "grad_norm": 140797.140625,
-      "learning_rate": 2.573455639959868e-06,
-      "loss": 1.8611,
+      "grad_norm": 164955.390625,
+      "learning_rate": 2.716338880484115e-05,
+      "loss": 1.837,
       "step": 1360
     },
     {
       "epoch": 0.9394145677331518,
-      "grad_norm": 179730.140625,
-      "learning_rate": 2.56485595528164e-06,
-      "loss": 1.8152,
+      "grad_norm": 243246.953125,
+      "learning_rate": 2.7072617246596067e-05,
+      "loss": 1.7863,
       "step": 1380
     },
     {
       "epoch": 0.9530292716133424,
-      "grad_norm": 107053.234375,
-      "learning_rate": 2.556256270603411e-06,
-      "loss": 1.7251,
+      "grad_norm": 133524.296875,
+      "learning_rate": 2.6981845688350984e-05,
+      "loss": 1.671,
       "step": 1400
     },
     {
       "epoch": 0.966643975493533,
-      "grad_norm": 148813.046875,
-      "learning_rate": 2.547656585925183e-06,
-      "loss": 1.7433,
+      "grad_norm": 187500.75,
+      "learning_rate": 2.68910741301059e-05,
+      "loss": 1.6882,
       "step": 1420
     },
     {
       "epoch": 0.9802586793737236,
-      "grad_norm": 207710.78125,
-      "learning_rate": 2.5390569012469544e-06,
-      "loss": 1.7679,
+      "grad_norm": 311251.9375,
+      "learning_rate": 2.6800302571860817e-05,
+      "loss": 1.6963,
       "step": 1440
     },
     {
       "epoch": 0.9938733832539143,
-      "grad_norm": 343417.40625,
-      "learning_rate": 2.530457216568726e-06,
-      "loss": 1.7534,
+      "grad_norm": 451302.75,
+      "learning_rate": 2.6709531013615737e-05,
+      "loss": 1.7184,
       "step": 1460
     }
   ],

checkpoint-1469/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:afe2658963e005554f6f06fcd5df65a7ddb835737aa4d0ac12b591a847e9f9e9
+oid sha256:6a7419a4a07644582d09dce99d13f18db0da1c6b5e087bc47ea69c737cc32b14
 size 5176