Training in progress, step 1100, checkpoint

checkpoint-1100/adapter_config.json

@@ -23,13 +23,13 @@
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "o_proj",
-    "down_proj",
     "gate_proj",
     "k_proj",
+    "o_proj",
+    "up_proj",
     "q_proj",
-    "v_proj",
-    "up_proj"
+    "down_proj",
+    "v_proj"
   ],
   "task_type": "CAUSAL_LM",
   "use_dora": false,

checkpoint-1100/adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b452e9bbe59aa96b5f7d7ce24dc77b202dc471698625aef68254290ecdda0fae
+oid sha256:c18c4f7301275001220c8b3c1ac25e1197174403386cbf384d3b395e9dfb1c2f
 size 145287696

checkpoint-1100/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4d89fee256e7582a70ad74661acce74718afdaad062a2aa93abcbeecc3a52180
+oid sha256:ee50658b75bfff280b64b71c9ce19d261a4981897c17e92deaad2e79ecf1d1f0
 size 290833618

checkpoint-1100/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a27130a45eaab857d152b34d32f20a9c7a4e05049e6e4b06d807a5cc43c2d0e0
+oid sha256:ac4f023cc9eed1dc852dc968eb6c75bdd4dd3b6f3b928ca92b6491cf36d2aafc
 size 14244

checkpoint-1100/trainer_state.json

@@ -10,948 +10,948 @@
   "log_history": [
     {
       "epoch": 0.010869565217391304,
-      "grad_norm": 2.120478868484497,
+      "grad_norm": 2.4570870399475098,
       "learning_rate": 4.981884057971015e-05,
-      "loss": 2.4608,
+      "loss": 2.3612,
       "step": 10
     },
     {
       "epoch": 0.021739130434782608,
-      "grad_norm": 1.4271776676177979,
+      "grad_norm": 1.2042421102523804,
       "learning_rate": 4.963768115942029e-05,
-      "loss": 1.2741,
+      "loss": 1.1915,
       "step": 20
     },
     {
       "epoch": 0.03260869565217391,
-      "grad_norm": 0.9902746677398682,
+      "grad_norm": 1.2019174098968506,
       "learning_rate": 4.945652173913044e-05,
-      "loss": 0.9403,
+      "loss": 0.8992,
       "step": 30
     },
     {
       "epoch": 0.043478260869565216,
-      "grad_norm": 1.0730223655700684,
+      "grad_norm": 1.0546120405197144,
       "learning_rate": 4.9275362318840584e-05,
-      "loss": 0.7517,
+      "loss": 0.7276,
       "step": 40
     },
     {
       "epoch": 0.05434782608695652,
-      "grad_norm": 0.8592391610145569,
+      "grad_norm": 0.754091203212738,
       "learning_rate": 4.909420289855073e-05,
-      "loss": 0.6695,
+      "loss": 0.6419,
       "step": 50
     },
     {
       "epoch": 0.05434782608695652,
-      "eval_loss": 0.8473024964332581,
-      "eval_runtime": 10.3597,
-      "eval_samples_per_second": 46.334,
-      "eval_steps_per_second": 2.896,
+      "eval_loss": 0.7905128598213196,
+      "eval_runtime": 10.8418,
+      "eval_samples_per_second": 44.273,
+      "eval_steps_per_second": 2.767,
       "step": 50
     },
     {
       "epoch": 0.06521739130434782,
-      "grad_norm": 0.7817952632904053,
+      "grad_norm": 0.7751966714859009,
       "learning_rate": 4.891304347826087e-05,
-      "loss": 0.6325,
+      "loss": 0.5971,
       "step": 60
     },
     {
       "epoch": 0.07608695652173914,
-      "grad_norm": 0.8163846135139465,
+      "grad_norm": 0.6874057650566101,
       "learning_rate": 4.873188405797102e-05,
-      "loss": 0.5994,
+      "loss": 0.608,
       "step": 70
     },
     {
       "epoch": 0.08695652173913043,
-      "grad_norm": 0.7358843088150024,
+      "grad_norm": 0.7145748734474182,
       "learning_rate": 4.855072463768116e-05,
-      "loss": 0.6158,
+      "loss": 0.6111,
       "step": 80
     },
     {
       "epoch": 0.09782608695652174,
-      "grad_norm": 0.7856508493423462,
+      "grad_norm": 0.8841484189033508,
       "learning_rate": 4.836956521739131e-05,
-      "loss": 0.5983,
+      "loss": 0.5848,
       "step": 90
     },
     {
       "epoch": 0.10869565217391304,
-      "grad_norm": 0.8615956902503967,
+      "grad_norm": 0.7937784790992737,
       "learning_rate": 4.818840579710145e-05,
-      "loss": 0.5858,
+      "loss": 0.5699,
       "step": 100
     },
     {
       "epoch": 0.10869565217391304,
-      "eval_loss": 0.7398766875267029,
-      "eval_runtime": 10.4317,
-      "eval_samples_per_second": 46.014,
-      "eval_steps_per_second": 2.876,
+      "eval_loss": 0.6956210732460022,
+      "eval_runtime": 10.863,
+      "eval_samples_per_second": 44.187,
+      "eval_steps_per_second": 2.762,
       "step": 100
     },
     {
       "epoch": 0.11956521739130435,
-      "grad_norm": 0.7915664315223694,
+      "grad_norm": 0.7818441987037659,
       "learning_rate": 4.80072463768116e-05,
-      "loss": 0.6228,
+      "loss": 0.5946,
       "step": 110
     },
     {
       "epoch": 0.13043478260869565,
-      "grad_norm": 0.8432077765464783,
+      "grad_norm": 0.8666340112686157,
       "learning_rate": 4.782608695652174e-05,
-      "loss": 0.5649,
+      "loss": 0.5469,
       "step": 120
     },
     {
       "epoch": 0.14130434782608695,
-      "grad_norm": 0.8230385780334473,
+      "grad_norm": 0.7637468576431274,
       "learning_rate": 4.764492753623189e-05,
-      "loss": 0.5368,
+      "loss": 0.5307,
       "step": 130
     },
     {
       "epoch": 0.15217391304347827,
-      "grad_norm": 0.8497728109359741,
+      "grad_norm": 0.8282362222671509,
       "learning_rate": 4.746376811594203e-05,
-      "loss": 0.5447,
+      "loss": 0.5312,
       "step": 140
     },
     {
       "epoch": 0.16304347826086957,
-      "grad_norm": 0.9394277930259705,
+      "grad_norm": 0.9675197601318359,
       "learning_rate": 4.7282608695652177e-05,
-      "loss": 0.5515,
+      "loss": 0.5486,
       "step": 150
     },
     {
       "epoch": 0.16304347826086957,
-      "eval_loss": 0.694922149181366,
-      "eval_runtime": 10.4348,
-      "eval_samples_per_second": 46.0,
-      "eval_steps_per_second": 2.875,
+      "eval_loss": 0.6567058563232422,
+      "eval_runtime": 10.8376,
+      "eval_samples_per_second": 44.29,
+      "eval_steps_per_second": 2.768,
       "step": 150
     },
     {
       "epoch": 0.17391304347826086,
-      "grad_norm": 0.7665875554084778,
+      "grad_norm": 0.7035924792289734,
       "learning_rate": 4.710144927536232e-05,
-      "loss": 0.5632,
+      "loss": 0.5553,
       "step": 160
     },
     {
       "epoch": 0.18478260869565216,
-      "grad_norm": 0.8912480473518372,
+      "grad_norm": 0.8957257866859436,
       "learning_rate": 4.6920289855072464e-05,
-      "loss": 0.5259,
+      "loss": 0.54,
       "step": 170
     },
     {
       "epoch": 0.1956521739130435,
-      "grad_norm": 0.8959771990776062,
+      "grad_norm": 0.8544663190841675,
       "learning_rate": 4.673913043478261e-05,
-      "loss": 0.5307,
+      "loss": 0.55,
       "step": 180
     },
     {
       "epoch": 0.20652173913043478,
-      "grad_norm": 0.8595960140228271,
+      "grad_norm": 0.7599456310272217,
       "learning_rate": 4.655797101449276e-05,
-      "loss": 0.5167,
+      "loss": 0.5102,
       "step": 190
     },
     {
       "epoch": 0.21739130434782608,
-      "grad_norm": 0.8418941497802734,
+      "grad_norm": 0.9151259064674377,
       "learning_rate": 4.63768115942029e-05,
-      "loss": 0.5106,
+      "loss": 0.5372,
       "step": 200
     },
     {
       "epoch": 0.21739130434782608,
-      "eval_loss": 0.6705857515335083,
-      "eval_runtime": 10.4259,
-      "eval_samples_per_second": 46.039,
-      "eval_steps_per_second": 2.877,
+      "eval_loss": 0.6319016218185425,
+      "eval_runtime": 10.8237,
+      "eval_samples_per_second": 44.347,
+      "eval_steps_per_second": 2.772,
       "step": 200
     },
     {
       "epoch": 0.22826086956521738,
-      "grad_norm": 0.8937019109725952,
+      "grad_norm": 0.911018431186676,
       "learning_rate": 4.6195652173913046e-05,
-      "loss": 0.5329,
+      "loss": 0.5354,
       "step": 210
     },
     {
       "epoch": 0.2391304347826087,
-      "grad_norm": 0.9121237397193909,
+      "grad_norm": 0.7792191505432129,
       "learning_rate": 4.601449275362319e-05,
-      "loss": 0.5341,
+      "loss": 0.5297,
       "step": 220
     },
     {
       "epoch": 0.25,
-      "grad_norm": 0.8253990411758423,
+      "grad_norm": 0.7631202340126038,
       "learning_rate": 4.5833333333333334e-05,
-      "loss": 0.5328,
+      "loss": 0.5258,
       "step": 230
     },
     {
       "epoch": 0.2608695652173913,
-      "grad_norm": 0.9218540787696838,
+      "grad_norm": 0.7436497211456299,
       "learning_rate": 4.565217391304348e-05,
-      "loss": 0.4706,
+      "loss": 0.4849,
       "step": 240
     },
     {
       "epoch": 0.2717391304347826,
-      "grad_norm": 0.9071463346481323,
+      "grad_norm": 0.8372677564620972,
       "learning_rate": 4.547101449275363e-05,
-      "loss": 0.5012,
+      "loss": 0.5131,
       "step": 250
     },
     {
       "epoch": 0.2717391304347826,
-      "eval_loss": 0.6487293839454651,
-      "eval_runtime": 10.4286,
-      "eval_samples_per_second": 46.027,
-      "eval_steps_per_second": 2.877,
+      "eval_loss": 0.6155125498771667,
+      "eval_runtime": 10.8327,
+      "eval_samples_per_second": 44.31,
+      "eval_steps_per_second": 2.769,
       "step": 250
     },
     {
       "epoch": 0.2826086956521739,
-      "grad_norm": 0.9131689071655273,
+      "grad_norm": 0.8056390285491943,
       "learning_rate": 4.528985507246377e-05,
-      "loss": 0.4819,
+      "loss": 0.4675,
       "step": 260
     },
     {
       "epoch": 0.29347826086956524,
-      "grad_norm": 0.8460387587547302,
+      "grad_norm": 0.8015912175178528,
       "learning_rate": 4.5108695652173916e-05,
-      "loss": 0.4951,
+      "loss": 0.4942,
       "step": 270
     },
     {
       "epoch": 0.30434782608695654,
-      "grad_norm": 0.8639910817146301,
+      "grad_norm": 0.8316887021064758,
       "learning_rate": 4.492753623188406e-05,
-      "loss": 0.4941,
+      "loss": 0.5025,
       "step": 280
     },
     {
       "epoch": 0.31521739130434784,
-      "grad_norm": 1.0238360166549683,
+      "grad_norm": 0.8440260887145996,
       "learning_rate": 4.4746376811594203e-05,
-      "loss": 0.4844,
+      "loss": 0.5112,
       "step": 290
     },
     {
       "epoch": 0.32608695652173914,
-      "grad_norm": 0.8217597603797913,
+      "grad_norm": 0.8523954153060913,
       "learning_rate": 4.456521739130435e-05,
-      "loss": 0.497,
+      "loss": 0.4973,
       "step": 300
     },
     {
       "epoch": 0.32608695652173914,
-      "eval_loss": 0.6351093053817749,
-      "eval_runtime": 10.4278,
-      "eval_samples_per_second": 46.031,
-      "eval_steps_per_second": 2.877,
+      "eval_loss": 0.6024141907691956,
+      "eval_runtime": 10.839,
+      "eval_samples_per_second": 44.284,
+      "eval_steps_per_second": 2.768,
       "step": 300
     },
     {
       "epoch": 0.33695652173913043,
-      "grad_norm": 0.8742851614952087,
+      "grad_norm": 0.8427216410636902,
       "learning_rate": 4.438405797101449e-05,
-      "loss": 0.5134,
+      "loss": 0.5271,
       "step": 310
     },
     {
       "epoch": 0.34782608695652173,
-      "grad_norm": 0.8606206774711609,
+      "grad_norm": 0.8126628398895264,
       "learning_rate": 4.4202898550724645e-05,
-      "loss": 0.4577,
+      "loss": 0.4608,
       "step": 320
     },
     {
       "epoch": 0.358695652173913,
-      "grad_norm": 1.030929446220398,
+      "grad_norm": 0.9947460293769836,
       "learning_rate": 4.4021739130434786e-05,
-      "loss": 0.4624,
+      "loss": 0.4762,
       "step": 330
     },
     {
       "epoch": 0.3695652173913043,
-      "grad_norm": 0.8211489915847778,
+      "grad_norm": 0.7744721174240112,
       "learning_rate": 4.384057971014493e-05,
-      "loss": 0.4942,
+      "loss": 0.4993,
       "step": 340
     },
     {
       "epoch": 0.3804347826086957,
-      "grad_norm": 1.0272365808486938,
+      "grad_norm": 0.9483833909034729,
       "learning_rate": 4.365942028985507e-05,
-      "loss": 0.4923,
+      "loss": 0.497,
       "step": 350
     },
     {
       "epoch": 0.3804347826086957,
-      "eval_loss": 0.6294742226600647,
-      "eval_runtime": 10.439,
-      "eval_samples_per_second": 45.981,
-      "eval_steps_per_second": 2.874,
+      "eval_loss": 0.596432626247406,
+      "eval_runtime": 10.8191,
+      "eval_samples_per_second": 44.366,
+      "eval_steps_per_second": 2.773,
       "step": 350
     },
     {
       "epoch": 0.391304347826087,
-      "grad_norm": 0.9275714159011841,
+      "grad_norm": 0.7573915123939514,
       "learning_rate": 4.347826086956522e-05,
-      "loss": 0.4784,
+      "loss": 0.469,
       "step": 360
     },
     {
       "epoch": 0.40217391304347827,
-      "grad_norm": 0.9278607368469238,
+      "grad_norm": 0.8013887405395508,
       "learning_rate": 4.329710144927536e-05,
-      "loss": 0.465,
+      "loss": 0.4704,
       "step": 370
     },
     {
       "epoch": 0.41304347826086957,
-      "grad_norm": 1.009921669960022,
+      "grad_norm": 0.8412826657295227,
       "learning_rate": 4.3115942028985515e-05,
-      "loss": 0.4776,
+      "loss": 0.5033,
       "step": 380
     },
     {
       "epoch": 0.42391304347826086,
-      "grad_norm": 0.9694483280181885,
+      "grad_norm": 0.8203413486480713,
       "learning_rate": 4.2934782608695655e-05,
-      "loss": 0.4956,
+      "loss": 0.5063,
       "step": 390
     },
     {
       "epoch": 0.43478260869565216,
-      "grad_norm": 0.8650519251823425,
+      "grad_norm": 0.7952508926391602,
       "learning_rate": 4.27536231884058e-05,
-      "loss": 0.4607,
+      "loss": 0.4706,
       "step": 400
     },
     {
       "epoch": 0.43478260869565216,
-      "eval_loss": 0.6202508807182312,
-      "eval_runtime": 10.4217,
-      "eval_samples_per_second": 46.058,
-      "eval_steps_per_second": 2.879,
+      "eval_loss": 0.5875406861305237,
+      "eval_runtime": 10.8328,
+      "eval_samples_per_second": 44.31,
+      "eval_steps_per_second": 2.769,
       "step": 400
     },
     {
       "epoch": 0.44565217391304346,
-      "grad_norm": 1.0009912252426147,
+      "grad_norm": 0.9020015597343445,
       "learning_rate": 4.257246376811594e-05,
-      "loss": 0.4695,
+      "loss": 0.4786,
       "step": 410
     },
     {
       "epoch": 0.45652173913043476,
-      "grad_norm": 0.9651597738265991,
+      "grad_norm": 0.8588363528251648,
       "learning_rate": 4.239130434782609e-05,
-      "loss": 0.4733,
+      "loss": 0.482,
       "step": 420
     },
     {
       "epoch": 0.4673913043478261,
-      "grad_norm": 0.9021621346473694,
+      "grad_norm": 0.816198468208313,
       "learning_rate": 4.221014492753623e-05,
-      "loss": 0.4543,
+      "loss": 0.4683,
       "step": 430
     },
     {
       "epoch": 0.4782608695652174,
-      "grad_norm": 0.9730608463287354,
+      "grad_norm": 0.8713563680648804,
       "learning_rate": 4.202898550724638e-05,
-      "loss": 0.4461,
+      "loss": 0.4422,
       "step": 440
     },
     {
       "epoch": 0.4891304347826087,
-      "grad_norm": 0.9395557641983032,
+      "grad_norm": 0.8633397817611694,
       "learning_rate": 4.1847826086956525e-05,
-      "loss": 0.4448,
+      "loss": 0.45,
       "step": 450
     },
     {
       "epoch": 0.4891304347826087,
-      "eval_loss": 0.6122747659683228,
-      "eval_runtime": 10.4358,
-      "eval_samples_per_second": 45.995,
-      "eval_steps_per_second": 2.875,
+      "eval_loss": 0.5812788605690002,
+      "eval_runtime": 10.8048,
+      "eval_samples_per_second": 44.425,
+      "eval_steps_per_second": 2.777,
       "step": 450
     },
     {
       "epoch": 0.5,
-      "grad_norm": 0.9062746167182922,
+      "grad_norm": 0.8244722485542297,
       "learning_rate": 4.166666666666667e-05,
-      "loss": 0.4653,
+      "loss": 0.4911,
       "step": 460
     },
     {
       "epoch": 0.5108695652173914,
-      "grad_norm": 0.9474561214447021,
+      "grad_norm": 0.8201929926872253,
       "learning_rate": 4.148550724637681e-05,
-      "loss": 0.4571,
+      "loss": 0.4611,
       "step": 470
     },
     {
       "epoch": 0.5217391304347826,
-      "grad_norm": 0.9197052717208862,
+      "grad_norm": 0.9646862149238586,
       "learning_rate": 4.130434782608696e-05,
-      "loss": 0.4416,
+      "loss": 0.4738,
       "step": 480
     },
     {
       "epoch": 0.532608695652174,
-      "grad_norm": 0.9986901879310608,
+      "grad_norm": 0.9149807095527649,
       "learning_rate": 4.11231884057971e-05,
-      "loss": 0.4446,
+      "loss": 0.4603,
       "step": 490
     },
     {
       "epoch": 0.5434782608695652,
-      "grad_norm": 0.9315592050552368,
+      "grad_norm": 0.8054277896881104,
       "learning_rate": 4.094202898550725e-05,
-      "loss": 0.4501,
+      "loss": 0.466,
       "step": 500
     },
     {
       "epoch": 0.5434782608695652,
-      "eval_loss": 0.6061260104179382,
-      "eval_runtime": 10.4306,
-      "eval_samples_per_second": 46.018,
-      "eval_steps_per_second": 2.876,
+      "eval_loss": 0.573806881904602,
+      "eval_runtime": 10.7906,
+      "eval_samples_per_second": 44.483,
+      "eval_steps_per_second": 2.78,
       "step": 500
     },
     {
       "epoch": 0.5543478260869565,
-      "grad_norm": 0.9725834131240845,
+      "grad_norm": 0.9110711812973022,
       "learning_rate": 4.076086956521739e-05,
-      "loss": 0.4465,
+      "loss": 0.4777,
       "step": 510
     },
     {
       "epoch": 0.5652173913043478,
-      "grad_norm": 0.9681097865104675,
+      "grad_norm": 0.8030044436454773,
       "learning_rate": 4.057971014492754e-05,
-      "loss": 0.4735,
+      "loss": 0.4965,
       "step": 520
     },
     {
       "epoch": 0.5760869565217391,
-      "grad_norm": 1.0040682554244995,
+      "grad_norm": 0.8847204446792603,
       "learning_rate": 4.039855072463768e-05,
-      "loss": 0.4743,
+      "loss": 0.4843,
       "step": 530
     },
     {
       "epoch": 0.5869565217391305,
-      "grad_norm": 0.9864315986633301,
+      "grad_norm": 0.9041047692298889,
       "learning_rate": 4.021739130434783e-05,
-      "loss": 0.45,
+      "loss": 0.4548,
       "step": 540
     },
     {
       "epoch": 0.5978260869565217,
-      "grad_norm": 0.954130232334137,
+      "grad_norm": 0.8136462569236755,
       "learning_rate": 4.003623188405797e-05,
-      "loss": 0.439,
+      "loss": 0.4543,
       "step": 550
     },
     {
       "epoch": 0.5978260869565217,
-      "eval_loss": 0.5990516543388367,
-      "eval_runtime": 10.427,
-      "eval_samples_per_second": 46.035,
-      "eval_steps_per_second": 2.877,
+      "eval_loss": 0.5662083029747009,
+      "eval_runtime": 10.7748,
+      "eval_samples_per_second": 44.549,
+      "eval_steps_per_second": 2.784,
       "step": 550
     },
     {
       "epoch": 0.6086956521739131,
-      "grad_norm": 1.150810718536377,
+      "grad_norm": 0.9470314383506775,
       "learning_rate": 3.985507246376812e-05,
-      "loss": 0.4603,
+      "loss": 0.4852,
       "step": 560
     },
     {
       "epoch": 0.6195652173913043,
-      "grad_norm": 0.9530227184295654,
+      "grad_norm": 0.9280235171318054,
       "learning_rate": 3.9673913043478264e-05,
-      "loss": 0.447,
+      "loss": 0.4655,
       "step": 570
     },
     {
       "epoch": 0.6304347826086957,
-      "grad_norm": 1.1645703315734863,
+      "grad_norm": 0.9118008017539978,
       "learning_rate": 3.9492753623188405e-05,
-      "loss": 0.4195,
+      "loss": 0.449,
       "step": 580
     },
     {
       "epoch": 0.6413043478260869,
-      "grad_norm": 1.0095150470733643,
+      "grad_norm": 0.9286589026451111,
       "learning_rate": 3.931159420289855e-05,
-      "loss": 0.434,
+      "loss": 0.448,
       "step": 590
     },
     {
       "epoch": 0.6521739130434783,
-      "grad_norm": 0.8818724155426025,
+      "grad_norm": 0.7761940956115723,
       "learning_rate": 3.91304347826087e-05,
-      "loss": 0.4385,
+      "loss": 0.4363,
       "step": 600
     },
     {
       "epoch": 0.6521739130434783,
-      "eval_loss": 0.5951303839683533,
-      "eval_runtime": 10.4035,
-      "eval_samples_per_second": 46.138,
-      "eval_steps_per_second": 2.884,
+      "eval_loss": 0.5645204186439514,
+      "eval_runtime": 10.7763,
+      "eval_samples_per_second": 44.542,
+      "eval_steps_per_second": 2.784,
       "step": 600
     },
     {
       "epoch": 0.6630434782608695,
-      "grad_norm": 1.0614160299301147,
+      "grad_norm": 0.9244002103805542,
       "learning_rate": 3.8949275362318846e-05,
-      "loss": 0.4128,
+      "loss": 0.4354,
       "step": 610
     },
     {
       "epoch": 0.6739130434782609,
-      "grad_norm": 0.883694589138031,
+      "grad_norm": 0.8263306021690369,
       "learning_rate": 3.876811594202899e-05,
-      "loss": 0.4664,
+      "loss": 0.4951,
       "step": 620
     },
     {
       "epoch": 0.6847826086956522,
-      "grad_norm": 0.97767573595047,
+      "grad_norm": 0.8908408284187317,
       "learning_rate": 3.8586956521739134e-05,
-      "loss": 0.4149,
+      "loss": 0.4401,
       "step": 630
     },
     {
       "epoch": 0.6956521739130435,
-      "grad_norm": 0.9714617133140564,
+      "grad_norm": 0.8497889637947083,
       "learning_rate": 3.8405797101449274e-05,
-      "loss": 0.4369,
+      "loss": 0.4514,
       "step": 640
     },
     {
       "epoch": 0.7065217391304348,
-      "grad_norm": 0.9593505263328552,
+      "grad_norm": 0.8306752443313599,
       "learning_rate": 3.822463768115942e-05,
-      "loss": 0.4268,
+      "loss": 0.4503,
       "step": 650
     },
     {
       "epoch": 0.7065217391304348,
-      "eval_loss": 0.5908556580543518,
-      "eval_runtime": 10.3762,
-      "eval_samples_per_second": 46.26,
-      "eval_steps_per_second": 2.891,
+      "eval_loss": 0.5576011538505554,
+      "eval_runtime": 10.7704,
+      "eval_samples_per_second": 44.567,
+      "eval_steps_per_second": 2.785,
       "step": 650
     },
     {
       "epoch": 0.717391304347826,
-      "grad_norm": 0.9807693362236023,
+      "grad_norm": 0.898961067199707,
       "learning_rate": 3.804347826086957e-05,
-      "loss": 0.4224,
+      "loss": 0.4185,
       "step": 660
     },
     {
       "epoch": 0.7282608695652174,
-      "grad_norm": 1.0408517122268677,
+      "grad_norm": 0.8870590925216675,
       "learning_rate": 3.7862318840579716e-05,
-      "loss": 0.41,
+      "loss": 0.4303,
       "step": 670
     },
     {
       "epoch": 0.7391304347826086,
-      "grad_norm": 0.9578678607940674,
+      "grad_norm": 0.8863015174865723,
       "learning_rate": 3.7681159420289856e-05,
-      "loss": 0.4207,
+      "loss": 0.4356,
       "step": 680
     },
     {
       "epoch": 0.75,
-      "grad_norm": 1.0341449975967407,
+      "grad_norm": 0.852130651473999,
       "learning_rate": 3.7500000000000003e-05,
-      "loss": 0.4435,
+      "loss": 0.4756,
       "step": 690
     },
     {
       "epoch": 0.7608695652173914,
-      "grad_norm": 0.9478809833526611,
+      "grad_norm": 0.8342716097831726,
       "learning_rate": 3.7318840579710144e-05,
-      "loss": 0.4463,
+      "loss": 0.4955,
       "step": 700
     },
     {
       "epoch": 0.7608695652173914,
-      "eval_loss": 0.585399866104126,
-      "eval_runtime": 10.3883,
-      "eval_samples_per_second": 46.206,
-      "eval_steps_per_second": 2.888,
+      "eval_loss": 0.5521832704544067,
+      "eval_runtime": 10.7828,
+      "eval_samples_per_second": 44.515,
+      "eval_steps_per_second": 2.782,
       "step": 700
     },
     {
       "epoch": 0.7717391304347826,
-      "grad_norm": 1.0197566747665405,
+      "grad_norm": 0.8382502794265747,
       "learning_rate": 3.713768115942029e-05,
-      "loss": 0.4234,
+      "loss": 0.4435,
       "step": 710
     },
     {
       "epoch": 0.782608695652174,
-      "grad_norm": 1.0265061855316162,
+      "grad_norm": 0.9789420962333679,
       "learning_rate": 3.695652173913043e-05,
-      "loss": 0.4027,
+      "loss": 0.3957,
       "step": 720
     },
     {
       "epoch": 0.7934782608695652,
-      "grad_norm": 0.9763724207878113,
+      "grad_norm": 0.8945524096488953,
       "learning_rate": 3.6775362318840586e-05,
-      "loss": 0.4284,
+      "loss": 0.4565,
       "step": 730
     },
     {
       "epoch": 0.8043478260869565,
-      "grad_norm": 1.1108958721160889,
+      "grad_norm": 0.9632206559181213,
       "learning_rate": 3.6594202898550726e-05,
-      "loss": 0.4197,
+      "loss": 0.4312,
       "step": 740
     },
     {
       "epoch": 0.8152173913043478,
-      "grad_norm": 0.938008189201355,
+      "grad_norm": 0.8000399470329285,
       "learning_rate": 3.641304347826087e-05,
-      "loss": 0.4133,
+      "loss": 0.429,
       "step": 750
     },
     {
       "epoch": 0.8152173913043478,
-      "eval_loss": 0.5860960483551025,
-      "eval_runtime": 10.3772,
-      "eval_samples_per_second": 46.255,
-      "eval_steps_per_second": 2.891,
+      "eval_loss": 0.550889790058136,
+      "eval_runtime": 10.7987,
+      "eval_samples_per_second": 44.45,
+      "eval_steps_per_second": 2.778,
       "step": 750
     },
     {
       "epoch": 0.8260869565217391,
-      "grad_norm": 1.106451392173767,
+      "grad_norm": 0.8817518353462219,
       "learning_rate": 3.6231884057971014e-05,
-      "loss": 0.4342,
+      "loss": 0.4464,
       "step": 760
     },
     {
       "epoch": 0.8369565217391305,
-      "grad_norm": 1.1099311113357544,
+      "grad_norm": 0.9800861477851868,
       "learning_rate": 3.605072463768116e-05,
-      "loss": 0.4345,
+      "loss": 0.4577,
       "step": 770
     },
     {
       "epoch": 0.8478260869565217,
-      "grad_norm": 1.0185613632202148,
+      "grad_norm": 0.846571683883667,
       "learning_rate": 3.58695652173913e-05,
-      "loss": 0.4196,
+      "loss": 0.4169,
       "step": 780
     },
     {
       "epoch": 0.8586956521739131,
-      "grad_norm": 0.9686940908432007,
+      "grad_norm": 0.8653192520141602,
       "learning_rate": 3.568840579710145e-05,
-      "loss": 0.4063,
+      "loss": 0.4195,
       "step": 790
     },
     {
       "epoch": 0.8695652173913043,
-      "grad_norm": 0.9732295274734497,
+      "grad_norm": 0.8515974283218384,
       "learning_rate": 3.5507246376811596e-05,
-      "loss": 0.4365,
+      "loss": 0.4366,
       "step": 800
     },
     {
       "epoch": 0.8695652173913043,
-      "eval_loss": 0.5782727003097534,
-      "eval_runtime": 10.4122,
-      "eval_samples_per_second": 46.1,
-      "eval_steps_per_second": 2.881,
+      "eval_loss": 0.543337881565094,
+      "eval_runtime": 10.8166,
+      "eval_samples_per_second": 44.376,
+      "eval_steps_per_second": 2.774,
       "step": 800
     },
     {
       "epoch": 0.8804347826086957,
-      "grad_norm": 1.1204928159713745,
+      "grad_norm": 0.947827935218811,
       "learning_rate": 3.532608695652174e-05,
-      "loss": 0.4165,
+      "loss": 0.4498,
       "step": 810
     },
     {
       "epoch": 0.8913043478260869,
-      "grad_norm": 1.0249269008636475,
+      "grad_norm": 0.866033136844635,
       "learning_rate": 3.514492753623188e-05,
-      "loss": 0.4339,
+      "loss": 0.4495,
       "step": 820
     },
     {
       "epoch": 0.9021739130434783,
-      "grad_norm": 0.9561118483543396,
+      "grad_norm": 0.8692104816436768,
       "learning_rate": 3.496376811594203e-05,
-      "loss": 0.4077,
+      "loss": 0.4502,
       "step": 830
     },
     {
       "epoch": 0.9130434782608695,
-      "grad_norm": 1.0603339672088623,
+      "grad_norm": 0.8841784596443176,
       "learning_rate": 3.478260869565218e-05,
-      "loss": 0.4176,
+      "loss": 0.4577,
       "step": 840
     },
     {
       "epoch": 0.9239130434782609,
-      "grad_norm": 0.9610790610313416,
+      "grad_norm": 0.9257445335388184,
       "learning_rate": 3.460144927536232e-05,
-      "loss": 0.4179,
+      "loss": 0.4329,
       "step": 850
     },
     {
       "epoch": 0.9239130434782609,
-      "eval_loss": 0.5790796875953674,
-      "eval_runtime": 10.4182,
-      "eval_samples_per_second": 46.073,
-      "eval_steps_per_second": 2.88,
+      "eval_loss": 0.5429030060768127,
+      "eval_runtime": 10.8284,
+      "eval_samples_per_second": 44.328,
+      "eval_steps_per_second": 2.77,
       "step": 850
     },
     {
       "epoch": 0.9347826086956522,
-      "grad_norm": 1.0670075416564941,
+      "grad_norm": 0.8753738403320312,
       "learning_rate": 3.4420289855072465e-05,
-      "loss": 0.4114,
+      "loss": 0.422,
       "step": 860
     },
     {
       "epoch": 0.9456521739130435,
-      "grad_norm": 1.070837140083313,
+      "grad_norm": 0.8853139877319336,
       "learning_rate": 3.423913043478261e-05,
-      "loss": 0.4034,
+      "loss": 0.4357,
       "step": 870
     },
     {
       "epoch": 0.9565217391304348,
-      "grad_norm": 1.0078990459442139,
+      "grad_norm": 0.8637978434562683,
       "learning_rate": 3.405797101449276e-05,
-      "loss": 0.3985,
+      "loss": 0.4254,
       "step": 880
     },
     {
       "epoch": 0.967391304347826,
-      "grad_norm": 1.0645533800125122,
+      "grad_norm": 0.9090889096260071,
       "learning_rate": 3.38768115942029e-05,
-      "loss": 0.4238,
+      "loss": 0.4526,
       "step": 890
     },
     {
       "epoch": 0.9782608695652174,
-      "grad_norm": 0.9449084401130676,
+      "grad_norm": 0.8591863512992859,
       "learning_rate": 3.369565217391305e-05,
-      "loss": 0.4146,
+      "loss": 0.4566,
       "step": 900
     },
     {
       "epoch": 0.9782608695652174,
-      "eval_loss": 0.5765789151191711,
-      "eval_runtime": 10.4319,
-      "eval_samples_per_second": 46.013,
-      "eval_steps_per_second": 2.876,
+      "eval_loss": 0.5399300456047058,
+      "eval_runtime": 10.8533,
+      "eval_samples_per_second": 44.226,
+      "eval_steps_per_second": 2.764,
       "step": 900
     },
     {
       "epoch": 0.9891304347826086,
-      "grad_norm": 1.1554139852523804,
+      "grad_norm": 0.8273764848709106,
       "learning_rate": 3.351449275362319e-05,
-      "loss": 0.3972,
+      "loss": 0.4182,
       "step": 910
     },
     {
       "epoch": 1.0,
-      "grad_norm": 2.253143787384033,
+      "grad_norm": 1.5736610889434814,
       "learning_rate": 3.3333333333333335e-05,
-      "loss": 0.4359,
+      "loss": 0.4548,
       "step": 920
     },
     {
       "epoch": 1.0108695652173914,
-      "grad_norm": 1.124770164489746,
+      "grad_norm": 0.8933672904968262,
       "learning_rate": 3.3152173913043475e-05,
-      "loss": 0.3533,
+      "loss": 0.3834,
       "step": 930
     },
     {
       "epoch": 1.0217391304347827,
-      "grad_norm": 1.1178290843963623,
+      "grad_norm": 0.9641056060791016,
       "learning_rate": 3.297101449275363e-05,
-      "loss": 0.3599,
+      "loss": 0.3962,
       "step": 940
     },
     {
       "epoch": 1.0326086956521738,
-      "grad_norm": 1.0256346464157104,
+      "grad_norm": 0.8840826153755188,
       "learning_rate": 3.278985507246377e-05,
-      "loss": 0.3632,
+      "loss": 0.3886,
       "step": 950
     },
     {
       "epoch": 1.0326086956521738,
-      "eval_loss": 0.5881566405296326,
-      "eval_runtime": 10.4627,
-      "eval_samples_per_second": 45.877,
-      "eval_steps_per_second": 2.867,
+      "eval_loss": 0.5417191982269287,
+      "eval_runtime": 10.828,
+      "eval_samples_per_second": 44.33,
+      "eval_steps_per_second": 2.771,
       "step": 950
     },
     {
       "epoch": 1.0434782608695652,
-      "grad_norm": 1.0145591497421265,
+      "grad_norm": 0.9005197286605835,
       "learning_rate": 3.260869565217392e-05,
-      "loss": 0.3647,
+      "loss": 0.3887,
       "step": 960
     },
     {
       "epoch": 1.0543478260869565,
-      "grad_norm": 1.2014787197113037,
+      "grad_norm": 1.0132256746292114,
       "learning_rate": 3.242753623188406e-05,
-      "loss": 0.37,
+      "loss": 0.4026,
       "step": 970
     },
     {
       "epoch": 1.065217391304348,
-      "grad_norm": 1.0973201990127563,
+      "grad_norm": 1.1139558553695679,
       "learning_rate": 3.2246376811594205e-05,
-      "loss": 0.3608,
+      "loss": 0.3703,
       "step": 980
     },
     {
       "epoch": 1.0760869565217392,
-      "grad_norm": 1.120807409286499,
+      "grad_norm": 0.9440574049949646,
       "learning_rate": 3.2065217391304345e-05,
-      "loss": 0.3466,
+      "loss": 0.3766,
       "step": 990
     },
     {
       "epoch": 1.0869565217391304,
-      "grad_norm": 1.1913529634475708,
+      "grad_norm": 1.0028133392333984,
       "learning_rate": 3.188405797101449e-05,
-      "loss": 0.3671,
+      "loss": 0.4038,
       "step": 1000
     },
     {
       "epoch": 1.0869565217391304,
-      "eval_loss": 0.591897189617157,
-      "eval_runtime": 10.4218,
-      "eval_samples_per_second": 46.057,
-      "eval_steps_per_second": 2.879,
+      "eval_loss": 0.5413796901702881,
+      "eval_runtime": 10.822,
+      "eval_samples_per_second": 44.354,
+      "eval_steps_per_second": 2.772,
       "step": 1000
     },
     {
       "epoch": 1.0978260869565217,
-      "grad_norm": 1.1925430297851562,
+      "grad_norm": 0.9884430766105652,
       "learning_rate": 3.170289855072464e-05,
-      "loss": 0.3729,
+      "loss": 0.4134,
       "step": 1010
     },
     {
       "epoch": 1.108695652173913,
-      "grad_norm": 1.0378402471542358,
+      "grad_norm": 0.920870840549469,
       "learning_rate": 3.152173913043479e-05,
-      "loss": 0.3615,
+      "loss": 0.3837,
       "step": 1020
     },
     {
       "epoch": 1.1195652173913044,
-      "grad_norm": 1.2244843244552612,
+      "grad_norm": 0.942890465259552,
       "learning_rate": 3.134057971014493e-05,
-      "loss": 0.3507,
+      "loss": 0.38,
       "step": 1030
     },
     {
       "epoch": 1.1304347826086956,
-      "grad_norm": 1.211162805557251,
+      "grad_norm": 1.0265600681304932,
       "learning_rate": 3.1159420289855074e-05,
-      "loss": 0.361,
+      "loss": 0.3988,
       "step": 1040
     },
     {
       "epoch": 1.141304347826087,
-      "grad_norm": 1.1001977920532227,
+      "grad_norm": 0.9994137287139893,
       "learning_rate": 3.0978260869565215e-05,
-      "loss": 0.3396,
+      "loss": 0.3649,
       "step": 1050
     },
     {
       "epoch": 1.141304347826087,
-      "eval_loss": 0.5917331576347351,
-      "eval_runtime": 10.4223,
-      "eval_samples_per_second": 46.055,
-      "eval_steps_per_second": 2.878,
+      "eval_loss": 0.537377655506134,
+      "eval_runtime": 10.7968,
+      "eval_samples_per_second": 44.458,
+      "eval_steps_per_second": 2.779,
       "step": 1050
     },
     {
       "epoch": 1.1521739130434783,
-      "grad_norm": 1.1034841537475586,
+      "grad_norm": 0.9697039723396301,
       "learning_rate": 3.079710144927536e-05,
-      "loss": 0.3718,
+      "loss": 0.393,
       "step": 1060
     },
     {
       "epoch": 1.1630434782608696,
-      "grad_norm": 1.1438697576522827,
+      "grad_norm": 0.9805816411972046,
       "learning_rate": 3.061594202898551e-05,
-      "loss": 0.3436,
+      "loss": 0.3594,
       "step": 1070
     },
     {
       "epoch": 1.1739130434782608,
-      "grad_norm": 1.1911267042160034,
+      "grad_norm": 0.9799211621284485,
       "learning_rate": 3.0434782608695656e-05,
-      "loss": 0.3569,
+      "loss": 0.387,
       "step": 1080
     },
     {
       "epoch": 1.184782608695652,
-      "grad_norm": 1.1565200090408325,
+      "grad_norm": 0.9287502765655518,
       "learning_rate": 3.02536231884058e-05,
-      "loss": 0.3639,
+      "loss": 0.3889,
       "step": 1090
     },
     {
       "epoch": 1.1956521739130435,
-      "grad_norm": 1.199948787689209,
+      "grad_norm": 1.042643666267395,
       "learning_rate": 3.0072463768115944e-05,
-      "loss": 0.3588,
+      "loss": 0.3894,
       "step": 1100
     },
     {
       "epoch": 1.1956521739130435,
-      "eval_loss": 0.5949474573135376,
-      "eval_runtime": 10.4416,
-      "eval_samples_per_second": 45.97,
-      "eval_steps_per_second": 2.873,
+      "eval_loss": 0.5378063917160034,
+      "eval_runtime": 10.77,
+      "eval_samples_per_second": 44.568,
+      "eval_steps_per_second": 2.786,
       "step": 1100
     }
   ],
@@ -972,7 +972,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 9.733867279672934e+16,
+  "total_flos": 9.684627540954317e+16,
   "train_batch_size": 16,
   "trial_name": null,
   "trial_params": null

checkpoint-1100/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a3712c5fdd6e5241e8ea26ae62b28612fc514ec88e8a03bc6c1b3a46acfc611a
+oid sha256:81786d634f359bee447d9c2dad023b92502900c87e9208470ed326b1204c5b1a
 size 5816