Training in progress, step 300, checkpoint

last-checkpoint/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3c6737088db9e98673fb773f1a59f6d3462831ce3df34f251e3fa0a12870dd15
+oid sha256:8b9d88423883eaf1c04e57a3842d0888c8c7e8d00f10a953643dbcd1d1ac64c4
 size 3237829088

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c0fb8cc13677d7d807a121c2184df8b603bb0d1169d3cfd53c15fe292d8e835b
+oid sha256:34d645606d6eb146eb99e0cc0adef1d87d290800358b32916d2b761d315b2c6b
 size 2061570519

last-checkpoint/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c5b7bf30711c780cfc00bf1d7e9073a2f0212cb8ee4f14b6ced8cf97508142e3
+oid sha256:9d383a88beb779fca7791b337addbc04045039df50bd4a6ea6c5557699bd1e48
 size 14645

last-checkpoint/scaler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0ff58b41c3672e659a0eb46d9ed11a0ca17415e7a2643a3ddfbaebb9f4e67f8f
+oid sha256:dc1a0da602f8abf4bf342932694d528cc1f0baa4d5027de58ad34f4d9855d085
 size 1383

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:262af81f2f23d66f087a6f7b0eea32e3dba253538f0517c5799e48488478b90c
+oid sha256:2aecfedc7f5bc992e41736b72929d50222b1c844811a96ac093ccc646bdc5dd7
 size 1465

last-checkpoint/trainer_state.json

@@ -2,734 +2,229 @@
   "best_global_step": null,
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 0.14349261013057826,
-  "eval_steps": 500,
-  "global_step": 500,
+  "epoch": 0.08609556607834697,
+  "eval_steps": 300,
+  "global_step": 300,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
-    {
-      "epoch": 0.0014349261013057828,
-      "grad_norm": 391.24395751953125,
-      "learning_rate": 2.898550724637681e-06,
-      "loss": 13.9544,
-      "step": 5
-    },
     {
       "epoch": 0.0028698522026115655,
-      "grad_norm": 169.8328399658203,
+      "grad_norm": 168.79563903808594,
       "learning_rate": 1.739130434782609e-05,
-      "loss": 10.071,
+      "loss": 12.0169,
       "step": 10
     },
-    {
-      "epoch": 0.004304778303917348,
-      "grad_norm": 252.62840270996094,
-      "learning_rate": 3.188405797101449e-05,
-      "loss": 8.0348,
-      "step": 15
-    },
     {
       "epoch": 0.005739704405223131,
-      "grad_norm": 41.308902740478516,
+      "grad_norm": 20.983991622924805,
       "learning_rate": 4.63768115942029e-05,
-      "loss": 7.6824,
+      "loss": 7.8235,
       "step": 20
     },
-    {
-      "epoch": 0.007174630506528914,
-      "grad_norm": 20.71501922607422,
-      "learning_rate": 6.086956521739131e-05,
-      "loss": 6.5785,
-      "step": 25
-    },
     {
       "epoch": 0.008609556607834697,
-      "grad_norm": 28.731815338134766,
+      "grad_norm": 21.168655395507812,
       "learning_rate": 7.536231884057971e-05,
-      "loss": 5.5247,
+      "loss": 6.834,
       "step": 30
     },
-    {
-      "epoch": 0.01004448270914048,
-      "grad_norm": 13.58375358581543,
-      "learning_rate": 8.985507246376813e-05,
-      "loss": 4.0724,
-      "step": 35
-    },
     {
       "epoch": 0.011479408810446262,
-      "grad_norm": 29.017719268798828,
+      "grad_norm": 21.57039451599121,
       "learning_rate": 0.00010434782608695653,
-      "loss": 3.0084,
+      "loss": 3.9023,
       "step": 40
     },
-    {
-      "epoch": 0.012914334911752045,
-      "grad_norm": 32.97686767578125,
-      "learning_rate": 0.00011884057971014493,
-      "loss": 3.2098,
-      "step": 45
-    },
     {
       "epoch": 0.014349261013057828,
-      "grad_norm": 9.631165504455566,
+      "grad_norm": 48.81906509399414,
       "learning_rate": 0.00013333333333333334,
-      "loss": 2.8009,
+      "loss": 2.9802,
       "step": 50
     },
-    {
-      "epoch": 0.01578418711436361,
-      "grad_norm": 7.5439629554748535,
-      "learning_rate": 0.00014782608695652173,
-      "loss": 2.6667,
-      "step": 55
-    },
     {
       "epoch": 0.017219113215669393,
-      "grad_norm": 30.146381378173828,
+      "grad_norm": 7.396921157836914,
       "learning_rate": 0.00016231884057971017,
-      "loss": 2.4567,
+      "loss": 2.6257,
       "step": 60
     },
-    {
-      "epoch": 0.018654039316975176,
-      "grad_norm": 9.834168434143066,
-      "learning_rate": 0.00017681159420289858,
-      "loss": 2.6511,
-      "step": 65
-    },
     {
       "epoch": 0.02008896541828096,
-      "grad_norm": 8.466848373413086,
+      "grad_norm": 62.19234848022461,
       "learning_rate": 0.00019130434782608697,
-      "loss": 2.4647,
+      "loss": 2.3201,
       "step": 70
     },
-    {
-      "epoch": 0.02152389151958674,
-      "grad_norm": 15.979572296142578,
-      "learning_rate": 0.00019999983084147306,
-      "loss": 2.4322,
-      "step": 75
-    },
     {
       "epoch": 0.022958817620892524,
-      "grad_norm": 8.879189491271973,
+      "grad_norm": 8.402580261230469,
       "learning_rate": 0.00019999792781461744,
-      "loss": 2.0723,
+      "loss": 2.1749,
       "step": 80
     },
-    {
-      "epoch": 0.024393743722198307,
-      "grad_norm": 7.873000144958496,
-      "learning_rate": 0.00019999391035312072,
-      "loss": 2.0487,
-      "step": 85
-    },
     {
       "epoch": 0.02582866982350409,
-      "grad_norm": 7.232510089874268,
+      "grad_norm": 7.064925670623779,
       "learning_rate": 0.0001999877785419313,
-      "loss": 1.9332,
+      "loss": 1.8889,
       "step": 90
     },
-    {
-      "epoch": 0.027263595924809873,
-      "grad_norm": 7.322793483734131,
-      "learning_rate": 0.00019997953251070497,
-      "loss": 1.9108,
-      "step": 95
-    },
     {
       "epoch": 0.028698522026115655,
-      "grad_norm": 8.617932319641113,
+      "grad_norm": 7.678985118865967,
       "learning_rate": 0.0001999691724338023,
-      "loss": 1.8038,
+      "loss": 1.8161,
       "step": 100
     },
-    {
-      "epoch": 0.030133448127421438,
-      "grad_norm": 7.669059753417969,
-      "learning_rate": 0.00019995669853028485,
-      "loss": 1.9083,
-      "step": 105
-    },
     {
       "epoch": 0.03156837422872722,
-      "grad_norm": 10.190857887268066,
+      "grad_norm": 9.882554054260254,
       "learning_rate": 0.0001999421110639107,
-      "loss": 1.9832,
+      "loss": 1.9209,
       "step": 110
     },
-    {
-      "epoch": 0.033003300330033,
-      "grad_norm": 9.29393196105957,
-      "learning_rate": 0.0001999254103431286,
-      "loss": 1.8482,
-      "step": 115
-    },
     {
       "epoch": 0.034438226431338786,
-      "grad_norm": 10.318120002746582,
+      "grad_norm": 8.960328102111816,
       "learning_rate": 0.00019990659672107177,
-      "loss": 1.8904,
+      "loss": 1.8535,
       "step": 120
     },
-    {
-      "epoch": 0.035873152532644566,
-      "grad_norm": 7.800939559936523,
-      "learning_rate": 0.00019988567059555013,
-      "loss": 1.756,
-      "step": 125
-    },
     {
       "epoch": 0.03730807863395035,
-      "grad_norm": 6.906073093414307,
+      "grad_norm": 6.723909378051758,
       "learning_rate": 0.00019986263240904216,
-      "loss": 1.8059,
+      "loss": 1.7978,
       "step": 130
     },
-    {
-      "epoch": 0.03874300473525613,
-      "grad_norm": 9.81100845336914,
-      "learning_rate": 0.00019983748264868528,
-      "loss": 1.6887,
-      "step": 135
-    },
     {
       "epoch": 0.04017793083656192,
-      "grad_norm": 11.59725570678711,
+      "grad_norm": 14.159058570861816,
       "learning_rate": 0.00019981022184626578,
-      "loss": 1.6778,
+      "loss": 1.686,
       "step": 140
     },
-    {
-      "epoch": 0.0416128569378677,
-      "grad_norm": 7.840384006500244,
-      "learning_rate": 0.0001997808505782075,
-      "loss": 1.6421,
-      "step": 145
-    },
     {
       "epoch": 0.04304778303917348,
-      "grad_norm": 10.421500205993652,
+      "grad_norm": 12.402606010437012,
       "learning_rate": 0.00019974936946555948,
-      "loss": 1.7217,
+      "loss": 1.6932,
       "step": 150
     },
-    {
-      "epoch": 0.04448270914047926,
-      "grad_norm": 8.085160255432129,
-      "learning_rate": 0.00019971577917398317,
-      "loss": 1.4836,
-      "step": 155
-    },
     {
       "epoch": 0.04591763524178505,
-      "grad_norm": 7.500054836273193,
+      "grad_norm": 7.793806076049805,
       "learning_rate": 0.000199680080413738,
-      "loss": 1.6914,
+      "loss": 1.5665,
       "step": 160
     },
-    {
-      "epoch": 0.04735256134309083,
-      "grad_norm": 9.05443286895752,
-      "learning_rate": 0.00019964227393966656,
-      "loss": 1.6051,
-      "step": 165
-    },
     {
       "epoch": 0.048787487444396614,
-      "grad_norm": 10.235902786254883,
+      "grad_norm": 9.647517204284668,
       "learning_rate": 0.0001996023605511786,
-      "loss": 1.5247,
+      "loss": 1.5892,
       "step": 170
     },
-    {
-      "epoch": 0.05022241354570239,
-      "grad_norm": 10.63776969909668,
-      "learning_rate": 0.00019956034109223415,
-      "loss": 1.412,
-      "step": 175
-    },
     {
       "epoch": 0.05165733964700818,
-      "grad_norm": 6.892467498779297,
+      "grad_norm": 7.5883564949035645,
       "learning_rate": 0.00019951621645132556,
-      "loss": 1.5032,
+      "loss": 1.5003,
       "step": 180
     },
-    {
-      "epoch": 0.05309226574831396,
-      "grad_norm": 9.97953987121582,
-      "learning_rate": 0.0001994699875614589,
-      "loss": 1.4722,
-      "step": 185
-    },
     {
       "epoch": 0.054527191849619745,
-      "grad_norm": 8.126319885253906,
+      "grad_norm": 9.5863676071167,
       "learning_rate": 0.00019942165540013412,
-      "loss": 1.3494,
+      "loss": 1.4324,
       "step": 190
     },
-    {
-      "epoch": 0.055962117950925525,
-      "grad_norm": 10.368117332458496,
-      "learning_rate": 0.00019937122098932428,
-      "loss": 1.3037,
-      "step": 195
-    },
     {
       "epoch": 0.05739704405223131,
-      "grad_norm": 11.593148231506348,
+      "grad_norm": 10.761382102966309,
       "learning_rate": 0.00019931868539545416,
-      "loss": 1.2657,
+      "loss": 1.2652,
       "step": 200
     },
-    {
-      "epoch": 0.05883197015353709,
-      "grad_norm": 11.356983184814453,
-      "learning_rate": 0.0001992640497293776,
-      "loss": 1.3293,
-      "step": 205
-    },
     {
       "epoch": 0.060266896254842876,
-      "grad_norm": 11.849316596984863,
+      "grad_norm": 23.32731056213379,
       "learning_rate": 0.00019920731514635396,
-      "loss": 1.38,
+      "loss": 1.2868,
       "step": 210
     },
-    {
-      "epoch": 0.061701822356148656,
-      "grad_norm": 25.30244255065918,
-      "learning_rate": 0.0001991484828460238,
-      "loss": 1.3061,
-      "step": 215
-    },
     {
       "epoch": 0.06313674845745444,
-      "grad_norm": 12.974631309509277,
+      "grad_norm": 15.128023147583008,
       "learning_rate": 0.00019908755407238343,
-      "loss": 1.2302,
+      "loss": 1.2272,
       "step": 220
     },
-    {
-      "epoch": 0.06457167455876023,
-      "grad_norm": 11.322882652282715,
-      "learning_rate": 0.00019902453011375865,
-      "loss": 1.364,
-      "step": 225
-    },
     {
       "epoch": 0.066006600660066,
-      "grad_norm": 7.863698959350586,
+      "grad_norm": 12.924105644226074,
       "learning_rate": 0.00019895941230277744,
-      "loss": 1.3096,
+      "loss": 1.307,
       "step": 230
     },
-    {
-      "epoch": 0.06744152676137179,
-      "grad_norm": 13.481120109558105,
-      "learning_rate": 0.00019889220201634193,
-      "loss": 1.2259,
-      "step": 235
-    },
     {
       "epoch": 0.06887645286267757,
-      "grad_norm": 15.518793106079102,
+      "grad_norm": 9.334559440612793,
       "learning_rate": 0.00019882290067559915,
-      "loss": 1.2618,
+      "loss": 1.1858,
       "step": 240
     },
-    {
-      "epoch": 0.07031137896398336,
-      "grad_norm": 8.889978408813477,
-      "learning_rate": 0.000198751509745911,
-      "loss": 1.259,
-      "step": 245
-    },
     {
       "epoch": 0.07174630506528913,
-      "grad_norm": 12.910364151000977,
+      "grad_norm": 12.918402671813965,
       "learning_rate": 0.0001986780307368233,
-      "loss": 1.1288,
+      "loss": 1.1668,
       "step": 250
     },
-    {
-      "epoch": 0.07318123116659492,
-      "grad_norm": 9.42977523803711,
-      "learning_rate": 0.0001986024652020339,
-      "loss": 1.1097,
-      "step": 255
-    },
     {
       "epoch": 0.0746161572679007,
-      "grad_norm": 16.449186325073242,
+      "grad_norm": 8.966814994812012,
       "learning_rate": 0.00019852481473935974,
-      "loss": 1.0025,
+      "loss": 1.04,
       "step": 260
     },
-    {
-      "epoch": 0.07605108336920649,
-      "grad_norm": 19.828948974609375,
-      "learning_rate": 0.0001984450809907031,
-      "loss": 1.0597,
-      "step": 265
-    },
     {
       "epoch": 0.07748600947051226,
-      "grad_norm": 27.04888153076172,
+      "grad_norm": 10.825933456420898,
       "learning_rate": 0.000198363265642017,
-      "loss": 1.092,
+      "loss": 1.0674,
       "step": 270
     },
-    {
-      "epoch": 0.07892093557181805,
-      "grad_norm": 27.228042602539062,
-      "learning_rate": 0.00019827937042326928,
-      "loss": 1.1886,
-      "step": 275
-    },
     {
       "epoch": 0.08035586167312384,
-      "grad_norm": 9.512433052062988,
+      "grad_norm": 20.35280418395996,
       "learning_rate": 0.00019819339710840626,
-      "loss": 1.127,
+      "loss": 1.1564,
       "step": 280
     },
-    {
-      "epoch": 0.08179078777442962,
-      "grad_norm": 26.35071563720703,
-      "learning_rate": 0.00019810534751531523,
-      "loss": 1.0337,
-      "step": 285
-    },
     {
       "epoch": 0.0832257138757354,
-      "grad_norm": 9.270747184753418,
+      "grad_norm": 24.500883102416992,
       "learning_rate": 0.00019801522350578577,
-      "loss": 1.0625,
+      "loss": 1.0751,
       "step": 290
     },
-    {
-      "epoch": 0.08466063997704118,
-      "grad_norm": 13.372587203979492,
-      "learning_rate": 0.00019792302698547077,
-      "loss": 1.0753,
-      "step": 295
-    },
     {
       "epoch": 0.08609556607834697,
-      "grad_norm": 27.558691024780273,
+      "grad_norm": 8.19206428527832,
       "learning_rate": 0.00019782875990384568,
-      "loss": 1.0572,
+      "loss": 1.0476,
       "step": 300
-    },
-    {
-      "epoch": 0.08753049217965275,
-      "grad_norm": 6.429300785064697,
-      "learning_rate": 0.00019773242425416768,
-      "loss": 1.0431,
-      "step": 305
-    },
-    {
-      "epoch": 0.08896541828095852,
-      "grad_norm": 13.104458808898926,
-      "learning_rate": 0.00019763402207343338,
-      "loss": 1.0644,
-      "step": 310
-    },
-    {
-      "epoch": 0.09040034438226431,
-      "grad_norm": 11.63284969329834,
-      "learning_rate": 0.0001975335554423356,
-      "loss": 1.0902,
-      "step": 315
-    },
-    {
-      "epoch": 0.0918352704835701,
-      "grad_norm": 12.347888946533203,
-      "learning_rate": 0.00019743102648521967,
-      "loss": 0.9932,
-      "step": 320
-    },
-    {
-      "epoch": 0.09327019658487588,
-      "grad_norm": 16.778013229370117,
-      "learning_rate": 0.00019732643737003827,
-      "loss": 0.967,
-      "step": 325
-    },
-    {
-      "epoch": 0.09470512268618166,
-      "grad_norm": 12.950679779052734,
-      "learning_rate": 0.00019721979030830572,
-      "loss": 0.9775,
-      "step": 330
-    },
-    {
-      "epoch": 0.09614004878748744,
-      "grad_norm": 21.43748664855957,
-      "learning_rate": 0.00019711108755505116,
-      "loss": 0.931,
-      "step": 335
-    },
-    {
-      "epoch": 0.09757497488879323,
-      "grad_norm": 9.503966331481934,
-      "learning_rate": 0.0001970003314087709,
-      "loss": 1.0295,
-      "step": 340
-    },
-    {
-      "epoch": 0.09900990099009901,
-      "grad_norm": 13.628849983215332,
-      "learning_rate": 0.0001968875242113798,
-      "loss": 0.9998,
-      "step": 345
-    },
-    {
-      "epoch": 0.10044482709140479,
-      "grad_norm": 7.438472747802734,
-      "learning_rate": 0.0001967726683481617,
-      "loss": 0.9935,
-      "step": 350
-    },
-    {
-      "epoch": 0.10187975319271057,
-      "grad_norm": 7.533242702484131,
-      "learning_rate": 0.00019665576624771917,
-      "loss": 1.0334,
-      "step": 355
-    },
-    {
-      "epoch": 0.10331467929401636,
-      "grad_norm": 8.190099716186523,
-      "learning_rate": 0.00019653682038192188,
-      "loss": 0.9273,
-      "step": 360
-    },
-    {
-      "epoch": 0.10474960539532215,
-      "grad_norm": 12.107589721679688,
-      "learning_rate": 0.0001964158332658545,
-      "loss": 0.9917,
-      "step": 365
-    },
-    {
-      "epoch": 0.10618453149662792,
-      "grad_norm": 8.531739234924316,
-      "learning_rate": 0.00019629280745776364,
-      "loss": 0.8793,
-      "step": 370
-    },
-    {
-      "epoch": 0.1076194575979337,
-      "grad_norm": 8.719773292541504,
-      "learning_rate": 0.00019616774555900342,
-      "loss": 1.0215,
-      "step": 375
-    },
-    {
-      "epoch": 0.10905438369923949,
-      "grad_norm": 8.249613761901855,
-      "learning_rate": 0.0001960406502139808,
-      "loss": 0.9413,
-      "step": 380
-    },
-    {
-      "epoch": 0.11048930980054528,
-      "grad_norm": 12.444363594055176,
-      "learning_rate": 0.0001959115241100994,
-      "loss": 0.9262,
-      "step": 385
-    },
-    {
-      "epoch": 0.11192423590185105,
-      "grad_norm": 17.167621612548828,
-      "learning_rate": 0.00019578036997770296,
-      "loss": 0.9388,
-      "step": 390
-    },
-    {
-      "epoch": 0.11335916200315684,
-      "grad_norm": 9.513605117797852,
-      "learning_rate": 0.00019564719059001734,
-      "loss": 0.904,
-      "step": 395
-    },
-    {
-      "epoch": 0.11479408810446262,
-      "grad_norm": 42.635414123535156,
-      "learning_rate": 0.0001955119887630919,
-      "loss": 0.9174,
-      "step": 400
-    },
-    {
-      "epoch": 0.11622901420576841,
-      "grad_norm": 18.625703811645508,
-      "learning_rate": 0.0001953747673557403,
-      "loss": 0.9849,
-      "step": 405
-    },
-    {
-      "epoch": 0.11766394030707418,
-      "grad_norm": 12.731193542480469,
-      "learning_rate": 0.0001952355292694795,
-      "loss": 0.9476,
-      "step": 410
-    },
-    {
-      "epoch": 0.11909886640837997,
-      "grad_norm": 16.935991287231445,
-      "learning_rate": 0.0001950942774484689,
-      "loss": 0.9683,
-      "step": 415
-    },
-    {
-      "epoch": 0.12053379250968575,
-      "grad_norm": 7.0110249519348145,
-      "learning_rate": 0.0001949510148794478,
-      "loss": 1.0143,
-      "step": 420
-    },
-    {
-      "epoch": 0.12196871861099154,
-      "grad_norm": 13.457725524902344,
-      "learning_rate": 0.0001948057445916724,
-      "loss": 0.9607,
-      "step": 425
-    },
-    {
-      "epoch": 0.12340364471229731,
-      "grad_norm": 9.616328239440918,
-      "learning_rate": 0.00019465846965685158,
-      "loss": 0.9426,
-      "step": 430
-    },
-    {
-      "epoch": 0.1248385708136031,
-      "grad_norm": 10.017128944396973,
-      "learning_rate": 0.0001945091931890822,
-      "loss": 0.9967,
-      "step": 435
-    },
-    {
-      "epoch": 0.12627349691490888,
-      "grad_norm": 10.576837539672852,
-      "learning_rate": 0.00019435791834478293,
-      "loss": 1.0039,
-      "step": 440
-    },
-    {
-      "epoch": 0.12770842301621466,
-      "grad_norm": 19.164012908935547,
-      "learning_rate": 0.0001942046483226278,
-      "loss": 0.9536,
-      "step": 445
-    },
-    {
-      "epoch": 0.12914334911752046,
-      "grad_norm": 9.686524391174316,
-      "learning_rate": 0.0001940493863634784,
-      "loss": 0.8871,
-      "step": 450
-    },
-    {
-      "epoch": 0.13057827521882623,
-      "grad_norm": 11.136150360107422,
-      "learning_rate": 0.00019389213575031543,
-      "loss": 1.0405,
-      "step": 455
-    },
-    {
-      "epoch": 0.132013201320132,
-      "grad_norm": 7.720663070678711,
-      "learning_rate": 0.00019373289980816917,
-      "loss": 0.8974,
-      "step": 460
-    },
-    {
-      "epoch": 0.1334481274214378,
-      "grad_norm": 7.253653526306152,
-      "learning_rate": 0.00019357168190404936,
-      "loss": 0.9455,
-      "step": 465
-    },
-    {
-      "epoch": 0.13488305352274357,
-      "grad_norm": 14.146231651306152,
-      "learning_rate": 0.00019340848544687386,
-      "loss": 0.916,
-      "step": 470
-    },
-    {
-      "epoch": 0.13631797962404937,
-      "grad_norm": 10.805950164794922,
-      "learning_rate": 0.00019324331388739658,
-      "loss": 0.9894,
-      "step": 475
-    },
-    {
-      "epoch": 0.13775290572535515,
-      "grad_norm": 11.11950397491455,
-      "learning_rate": 0.00019307617071813454,
-      "loss": 1.0485,
-      "step": 480
-    },
-    {
-      "epoch": 0.13918783182666092,
-      "grad_norm": 6.732795715332031,
-      "learning_rate": 0.0001929070594732941,
-      "loss": 0.9406,
-      "step": 485
-    },
-    {
-      "epoch": 0.14062275792796672,
-      "grad_norm": 8.768498420715332,
-      "learning_rate": 0.00019273598372869603,
-      "loss": 0.9745,
-      "step": 490
-    },
-    {
-      "epoch": 0.1420576840292725,
-      "grad_norm": 10.62530517578125,
-      "learning_rate": 0.00019256294710170017,
-      "loss": 0.9702,
-      "step": 495
-    },
-    {
-      "epoch": 0.14349261013057826,
-      "grad_norm": 8.457195281982422,
-      "learning_rate": 0.0001923879532511287,
-      "loss": 0.9026,
-      "step": 500
-    },
-    {
-      "epoch": 0.14349261013057826,
-      "eval_avg_tokens": 300.0,
-      "eval_code": 0.0,
-      "eval_exact": 0.0,
-      "eval_family": 0.0,
-      "eval_loss": 0.9469140768051147,
-      "eval_runtime": 173.8266,
-      "eval_samples": 300,
-      "eval_samples_per_second": 1.726,
-      "eval_steps_per_second": 1.726,
-      "eval_super": 0.0,
-      "eval_total_tokens": 90000,
-      "step": 500
     }
   ],
-  "logging_steps": 5,
+  "logging_steps": 10,
   "max_steps": 3485,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 1,
-  "save_steps": 500,
+  "save_steps": 300,
   "stateful_callbacks": {
     "TrainerControl": {
       "args": {
@@ -742,7 +237,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 2.045941383168e+19,
+  "total_flos": 1.2275648299008e+19,
   "train_batch_size": 6,
   "trial_name": null,
   "trial_params": null

last-checkpoint/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:335603833df1126fe248b6fbecf339f53f5e5ae272c26e1fa729e348528f8cfc
+oid sha256:4a7dea7536ca58c97fbccef8df0fd6f88f39f81b73f5f4eafbed1d750e825400
 size 6033