End of training

README.md

@@ -16,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # Llama-3.1-8B-Instruct-PsyCourse-doc-fold6
 
-This model is a fine-tuned version of [meta-llama/Llama-3.1-8B-Instruct](https://huggingface.co/meta-llama/Llama-3.1-8B-Instruct) on an unknown dataset.
+This model is a fine-tuned version of [meta-llama/Llama-3.1-8B-Instruct](https://huggingface.co/meta-llama/Llama-3.1-8B-Instruct) on the course-doc-train-fold6 dataset.
 It achieves the following results on the evaluation set:
 - Loss: 0.0160
 

all_results.json

@@ -1,12 +1,12 @@
 {
     "epoch": 4.938271604938271,
-    "eval_loss": 0.050025228410959244,
-    "eval_runtime": 21.2463,
-    "eval_samples_per_second": 2.118,
-    "eval_steps_per_second": 2.118,
-    "total_flos": 1.5205115961940378e+17,
-    "train_loss": 0.10026876297593117,
-    "train_runtime": 3651.0486,
-    "train_samples_per_second": 0.555,
-    "train_steps_per_second": 0.034
+    "eval_loss": 0.01600480265915394,
+    "eval_runtime": 26.3497,
+    "eval_samples_per_second": 1.708,
+    "eval_steps_per_second": 1.708,
+    "total_flos": 1.931743422132388e+17,
+    "train_loss": 0.026140653513371945,
+    "train_runtime": 5707.3107,
+    "train_samples_per_second": 0.355,
+    "train_steps_per_second": 0.022
 }

eval_results.json

@@ -1,7 +1,7 @@
 {
     "epoch": 4.938271604938271,
-    "eval_loss": 0.050025228410959244,
-    "eval_runtime": 21.2463,
-    "eval_samples_per_second": 2.118,
-    "eval_steps_per_second": 2.118
+    "eval_loss": 0.01600480265915394,
+    "eval_runtime": 26.3497,
+    "eval_samples_per_second": 1.708,
+    "eval_steps_per_second": 1.708
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 4.938271604938271,
-    "total_flos": 1.5205115961940378e+17,
-    "train_loss": 0.10026876297593117,
-    "train_runtime": 3651.0486,
-    "train_samples_per_second": 0.555,
-    "train_steps_per_second": 0.034
+    "total_flos": 1.931743422132388e+17,
+    "train_loss": 0.026140653513371945,
+    "train_runtime": 5707.3107,
+    "train_samples_per_second": 0.355,
+    "train_steps_per_second": 0.022
 }

trainer_state.json

@@ -1,5 +1,5 @@
 {
-  "best_metric": 0.050025228410959244,
+  "best_metric": 0.01600480265915394,
   "best_model_checkpoint": "saves/psy-course-doc/Llama-3.1-8B-Instruct/train/fold6/checkpoint-110",
   "epoch": 4.938271604938271,
   "eval_steps": 10,
@@ -10,983 +10,983 @@
   "log_history": [
     {
       "epoch": 0.03950617283950617,
-      "grad_norm": 0.1937720775604248,
+      "grad_norm": 0.18577924370765686,
       "learning_rate": 7.692307692307694e-06,
-      "loss": 0.1625,
+      "loss": 0.1412,
       "step": 1
     },
     {
       "epoch": 0.07901234567901234,
-      "grad_norm": 0.1674083173274994,
+      "grad_norm": 0.13048163056373596,
       "learning_rate": 1.5384615384615387e-05,
-      "loss": 0.2251,
+      "loss": 0.1086,
       "step": 2
     },
     {
       "epoch": 0.11851851851851852,
-      "grad_norm": 0.2001013159751892,
+      "grad_norm": 0.19219337403774261,
       "learning_rate": 2.307692307692308e-05,
-      "loss": 0.1538,
+      "loss": 0.1358,
       "step": 3
     },
     {
       "epoch": 0.1580246913580247,
-      "grad_norm": 0.17400819063186646,
+      "grad_norm": 0.13072609901428223,
       "learning_rate": 3.0769230769230774e-05,
-      "loss": 0.1968,
+      "loss": 0.1104,
       "step": 4
     },
     {
       "epoch": 0.19753086419753085,
-      "grad_norm": 0.15696649253368378,
+      "grad_norm": 0.1273798942565918,
       "learning_rate": 3.846153846153846e-05,
-      "loss": 0.2309,
+      "loss": 0.0977,
       "step": 5
     },
     {
       "epoch": 0.23703703703703705,
-      "grad_norm": 0.1767856478691101,
+      "grad_norm": 0.13405117392539978,
       "learning_rate": 4.615384615384616e-05,
-      "loss": 0.2071,
+      "loss": 0.1038,
       "step": 6
     },
     {
       "epoch": 0.2765432098765432,
-      "grad_norm": 0.17840790748596191,
+      "grad_norm": 0.14149965345859528,
       "learning_rate": 5.384615384615385e-05,
-      "loss": 0.2325,
+      "loss": 0.1097,
       "step": 7
     },
     {
       "epoch": 0.3160493827160494,
-      "grad_norm": 0.15362082421779633,
+      "grad_norm": 0.11732533574104309,
       "learning_rate": 6.153846153846155e-05,
-      "loss": 0.2162,
+      "loss": 0.0907,
       "step": 8
     },
     {
       "epoch": 0.35555555555555557,
-      "grad_norm": 0.16653591394424438,
+      "grad_norm": 0.10946497321128845,
       "learning_rate": 6.923076923076924e-05,
-      "loss": 0.2196,
+      "loss": 0.0835,
       "step": 9
     },
     {
       "epoch": 0.3950617283950617,
-      "grad_norm": 0.14531226456165314,
+      "grad_norm": 0.13028384745121002,
       "learning_rate": 7.692307692307693e-05,
-      "loss": 0.1439,
+      "loss": 0.0816,
       "step": 10
     },
     {
       "epoch": 0.3950617283950617,
-      "eval_loss": 0.13264411687850952,
-      "eval_runtime": 21.5046,
-      "eval_samples_per_second": 2.093,
-      "eval_steps_per_second": 2.093,
+      "eval_loss": 0.08752671629190445,
+      "eval_runtime": 27.2044,
+      "eval_samples_per_second": 1.654,
+      "eval_steps_per_second": 1.654,
       "step": 10
     },
     {
       "epoch": 0.4345679012345679,
-      "grad_norm": 0.14842739701271057,
+      "grad_norm": 0.11912490427494049,
       "learning_rate": 8.461538461538461e-05,
-      "loss": 0.1487,
+      "loss": 0.0813,
       "step": 11
     },
     {
       "epoch": 0.4740740740740741,
-      "grad_norm": 0.13287214934825897,
+      "grad_norm": 0.11079352349042892,
       "learning_rate": 9.230769230769232e-05,
-      "loss": 0.1322,
+      "loss": 0.0706,
       "step": 12
     },
     {
       "epoch": 0.5135802469135803,
-      "grad_norm": 0.1623334139585495,
+      "grad_norm": 0.13070878386497498,
       "learning_rate": 0.0001,
-      "loss": 0.1338,
+      "loss": 0.0646,
       "step": 13
     },
     {
       "epoch": 0.5530864197530864,
-      "grad_norm": 0.17834453284740448,
+      "grad_norm": 0.1448982059955597,
       "learning_rate": 9.998033131915266e-05,
-      "loss": 0.1373,
+      "loss": 0.0561,
       "step": 14
     },
     {
       "epoch": 0.5925925925925926,
-      "grad_norm": 0.15861447155475616,
+      "grad_norm": 0.1239708960056305,
       "learning_rate": 9.992134075089084e-05,
-      "loss": 0.1856,
+      "loss": 0.0519,
       "step": 15
     },
     {
       "epoch": 0.6320987654320988,
-      "grad_norm": 0.15119513869285583,
+      "grad_norm": 0.10911262035369873,
       "learning_rate": 9.982307470588098e-05,
-      "loss": 0.1996,
+      "loss": 0.0499,
       "step": 16
     },
     {
       "epoch": 0.671604938271605,
-      "grad_norm": 0.14265546202659607,
+      "grad_norm": 0.11498083174228668,
       "learning_rate": 9.968561049466214e-05,
-      "loss": 0.1396,
+      "loss": 0.0448,
       "step": 17
     },
     {
       "epoch": 0.7111111111111111,
-      "grad_norm": 0.14528660476207733,
+      "grad_norm": 0.11828950047492981,
       "learning_rate": 9.950905626682228e-05,
-      "loss": 0.1335,
+      "loss": 0.0435,
       "step": 18
     },
     {
       "epoch": 0.7506172839506173,
-      "grad_norm": 0.11219226568937302,
+      "grad_norm": 0.10566812753677368,
       "learning_rate": 9.92935509259118e-05,
-      "loss": 0.0606,
+      "loss": 0.0368,
       "step": 19
     },
     {
       "epoch": 0.7901234567901234,
-      "grad_norm": 0.0979858934879303,
+      "grad_norm": 0.08341626077890396,
       "learning_rate": 9.903926402016153e-05,
-      "loss": 0.0809,
+      "loss": 0.0363,
       "step": 20
     },
     {
       "epoch": 0.7901234567901234,
-      "eval_loss": 0.0766303762793541,
-      "eval_runtime": 21.5609,
-      "eval_samples_per_second": 2.087,
-      "eval_steps_per_second": 2.087,
+      "eval_loss": 0.03500397875905037,
+      "eval_runtime": 27.0893,
+      "eval_samples_per_second": 1.661,
+      "eval_steps_per_second": 1.661,
       "step": 20
     },
     {
       "epoch": 0.8296296296296296,
-      "grad_norm": 0.10182628780603409,
+      "grad_norm": 0.0804336816072464,
       "learning_rate": 9.874639560909117e-05,
-      "loss": 0.0812,
+      "loss": 0.0353,
       "step": 21
     },
     {
       "epoch": 0.8691358024691358,
-      "grad_norm": 0.1082683727145195,
+      "grad_norm": 0.0875161737203598,
       "learning_rate": 9.841517610611309e-05,
-      "loss": 0.1503,
+      "loss": 0.0327,
       "step": 22
     },
     {
       "epoch": 0.908641975308642,
-      "grad_norm": 0.115780770778656,
+      "grad_norm": 0.06473363935947418,
       "learning_rate": 9.804586609725499e-05,
-      "loss": 0.171,
+      "loss": 0.0256,
       "step": 23
     },
     {
       "epoch": 0.9481481481481482,
-      "grad_norm": 0.09578105062246323,
+      "grad_norm": 0.08550036698579788,
       "learning_rate": 9.763875613614482e-05,
-      "loss": 0.0547,
+      "loss": 0.0303,
       "step": 24
     },
     {
       "epoch": 0.9876543209876543,
-      "grad_norm": 0.13462112843990326,
+      "grad_norm": 0.09782028943300247,
       "learning_rate": 9.719416651541839e-05,
-      "loss": 0.223,
+      "loss": 0.0288,
       "step": 25
     },
     {
       "epoch": 1.0271604938271606,
-      "grad_norm": 0.22300109267234802,
+      "grad_norm": 0.14770805835723877,
       "learning_rate": 9.671244701472999e-05,
-      "loss": 0.2313,
+      "loss": 0.0428,
       "step": 26
     },
     {
       "epoch": 1.0666666666666667,
-      "grad_norm": 0.10362166911363602,
+      "grad_norm": 0.08113914728164673,
       "learning_rate": 9.619397662556435e-05,
-      "loss": 0.0629,
+      "loss": 0.0195,
       "step": 27
     },
     {
       "epoch": 1.106172839506173,
-      "grad_norm": 0.10362211614847183,
+      "grad_norm": 0.07905824482440948,
       "learning_rate": 9.563916325306594e-05,
-      "loss": 0.1191,
+      "loss": 0.024,
       "step": 28
     },
     {
       "epoch": 1.145679012345679,
-      "grad_norm": 0.0990530326962471,
+      "grad_norm": 0.07929601520299911,
       "learning_rate": 9.504844339512095e-05,
-      "loss": 0.0748,
+      "loss": 0.0234,
       "step": 29
     },
     {
       "epoch": 1.1851851851851851,
-      "grad_norm": 0.10876777023077011,
+      "grad_norm": 0.07888120412826538,
       "learning_rate": 9.442228179894362e-05,
-      "loss": 0.0712,
+      "loss": 0.0215,
       "step": 30
     },
     {
       "epoch": 1.1851851851851851,
-      "eval_loss": 0.0628211572766304,
-      "eval_runtime": 21.5648,
-      "eval_samples_per_second": 2.087,
-      "eval_steps_per_second": 2.087,
+      "eval_loss": 0.024509310722351074,
+      "eval_runtime": 27.1133,
+      "eval_samples_per_second": 1.66,
+      "eval_steps_per_second": 1.66,
       "step": 30
     },
     {
       "epoch": 1.2246913580246914,
-      "grad_norm": 0.13057485222816467,
+      "grad_norm": 0.08701005578041077,
       "learning_rate": 9.376117109543769e-05,
-      "loss": 0.1239,
+      "loss": 0.0258,
       "step": 31
     },
     {
       "epoch": 1.2641975308641975,
-      "grad_norm": 0.11622237414121628,
+      "grad_norm": 0.07643706351518631,
       "learning_rate": 9.306563141162046e-05,
-      "loss": 0.1279,
+      "loss": 0.0205,
       "step": 32
     },
     {
       "epoch": 1.3037037037037038,
-      "grad_norm": 0.0912310853600502,
+      "grad_norm": 0.051044851541519165,
       "learning_rate": 9.233620996141421e-05,
-      "loss": 0.121,
+      "loss": 0.019,
       "step": 33
     },
     {
       "epoch": 1.34320987654321,
-      "grad_norm": 0.10160717368125916,
+      "grad_norm": 0.08121434599161148,
       "learning_rate": 9.157348061512727e-05,
-      "loss": 0.1291,
+      "loss": 0.0217,
       "step": 34
     },
     {
       "epoch": 1.382716049382716,
-      "grad_norm": 0.09293709695339203,
+      "grad_norm": 0.06094732880592346,
       "learning_rate": 9.077804344796302e-05,
-      "loss": 0.1045,
+      "loss": 0.021,
       "step": 35
     },
     {
       "epoch": 1.4222222222222223,
-      "grad_norm": 0.09594816714525223,
+      "grad_norm": 0.07103589922189713,
       "learning_rate": 8.995052426791247e-05,
-      "loss": 0.0558,
+      "loss": 0.0212,
       "step": 36
     },
     {
       "epoch": 1.4617283950617284,
-      "grad_norm": 0.09305794537067413,
+      "grad_norm": 0.04956590756773949,
       "learning_rate": 8.90915741234015e-05,
-      "loss": 0.0944,
+      "loss": 0.0153,
       "step": 37
     },
     {
       "epoch": 1.5012345679012347,
-      "grad_norm": 0.11819473654031754,
+      "grad_norm": 0.07781689614057541,
       "learning_rate": 8.820186879108038e-05,
-      "loss": 0.1417,
+      "loss": 0.0207,
       "step": 38
     },
     {
       "epoch": 1.5407407407407407,
-      "grad_norm": 0.09507416188716888,
+      "grad_norm": 0.05348929762840271,
       "learning_rate": 8.728210824415827e-05,
-      "loss": 0.068,
+      "loss": 0.0194,
       "step": 39
     },
     {
       "epoch": 1.5802469135802468,
-      "grad_norm": 0.08338752388954163,
+      "grad_norm": 0.06381393224000931,
       "learning_rate": 8.633301610170135e-05,
-      "loss": 0.0819,
+      "loss": 0.0193,
       "step": 40
     },
     {
       "epoch": 1.5802469135802468,
-      "eval_loss": 0.056919973343610764,
-      "eval_runtime": 21.5729,
-      "eval_samples_per_second": 2.086,
-      "eval_steps_per_second": 2.086,
+      "eval_loss": 0.020401908084750175,
+      "eval_runtime": 27.1238,
+      "eval_samples_per_second": 1.659,
+      "eval_steps_per_second": 1.659,
       "step": 40
     },
     {
       "epoch": 1.6197530864197531,
-      "grad_norm": 0.09787364304065704,
+      "grad_norm": 0.06865786015987396,
       "learning_rate": 8.535533905932738e-05,
-      "loss": 0.0991,
+      "loss": 0.0209,
       "step": 41
     },
     {
       "epoch": 1.6592592592592592,
-      "grad_norm": 0.10762111097574234,
+      "grad_norm": 0.07490848749876022,
       "learning_rate": 8.434984630174509e-05,
-      "loss": 0.1049,
+      "loss": 0.0209,
       "step": 42
     },
     {
       "epoch": 1.6987654320987655,
-      "grad_norm": 0.10685301572084427,
+      "grad_norm": 0.0662630945444107,
       "learning_rate": 8.33173288976002e-05,
-      "loss": 0.0925,
+      "loss": 0.0191,
       "step": 43
     },
     {
       "epoch": 1.7382716049382716,
-      "grad_norm": 0.08646407723426819,
+      "grad_norm": 0.06099313125014305,
       "learning_rate": 8.225859917710439e-05,
-      "loss": 0.0848,
+      "loss": 0.018,
       "step": 44
     },
     {
       "epoch": 1.7777777777777777,
-      "grad_norm": 0.10254059731960297,
+      "grad_norm": 0.07903651148080826,
       "learning_rate": 8.117449009293668e-05,
-      "loss": 0.0974,
+      "loss": 0.0215,
       "step": 45
     },
     {
       "epoch": 1.817283950617284,
-      "grad_norm": 0.11131167411804199,
+      "grad_norm": 0.07313495874404907,
       "learning_rate": 8.006585456492029e-05,
-      "loss": 0.0807,
+      "loss": 0.0147,
       "step": 46
     },
     {
       "epoch": 1.8567901234567903,
-      "grad_norm": 0.10282035917043686,
+      "grad_norm": 0.06267574429512024,
       "learning_rate": 7.89335648089903e-05,
-      "loss": 0.0914,
+      "loss": 0.0209,
       "step": 47
     },
     {
       "epoch": 1.8962962962962964,
-      "grad_norm": 0.08338365703821182,
+      "grad_norm": 0.05597149580717087,
       "learning_rate": 7.777851165098012e-05,
-      "loss": 0.09,
+      "loss": 0.0139,
       "step": 48
     },
     {
       "epoch": 1.9358024691358025,
-      "grad_norm": 0.09404530376195908,
+      "grad_norm": 0.06257493048906326,
       "learning_rate": 7.660160382576683e-05,
-      "loss": 0.0619,
+      "loss": 0.0212,
       "step": 49
     },
     {
       "epoch": 1.9753086419753085,
-      "grad_norm": 0.07219249755144119,
+      "grad_norm": 0.05538765713572502,
       "learning_rate": 7.540376726232648e-05,
-      "loss": 0.0566,
+      "loss": 0.015,
       "step": 50
     },
     {
       "epoch": 1.9753086419753085,
-      "eval_loss": 0.05362074822187424,
-      "eval_runtime": 21.5954,
-      "eval_samples_per_second": 2.084,
-      "eval_steps_per_second": 2.084,
+      "eval_loss": 0.018379276618361473,
+      "eval_runtime": 27.1267,
+      "eval_samples_per_second": 1.659,
+      "eval_steps_per_second": 1.659,
       "step": 50
     },
     {
       "epoch": 2.0148148148148146,
-      "grad_norm": 0.18207357823848724,
+      "grad_norm": 0.08045844733715057,
       "learning_rate": 7.4185944355262e-05,
-      "loss": 0.2196,
+      "loss": 0.0244,
       "step": 51
     },
     {
       "epoch": 2.054320987654321,
-      "grad_norm": 0.07671204954385757,
+      "grad_norm": 0.042876943945884705,
       "learning_rate": 7.294909322337689e-05,
-      "loss": 0.0867,
+      "loss": 0.0112,
       "step": 52
     },
     {
       "epoch": 2.093827160493827,
-      "grad_norm": 0.09706772118806839,
+      "grad_norm": 0.04620858281850815,
       "learning_rate": 7.169418695587791e-05,
-      "loss": 0.1006,
+      "loss": 0.014,
       "step": 53
     },
     {
       "epoch": 2.1333333333333333,
-      "grad_norm": 0.09666574746370316,
+      "grad_norm": 0.04439716041088104,
       "learning_rate": 7.042221284679982e-05,
-      "loss": 0.1168,
+      "loss": 0.0163,
       "step": 54
     },
     {
       "epoch": 2.1728395061728394,
-      "grad_norm": 0.09173166006803513,
+      "grad_norm": 0.04201319441199303,
       "learning_rate": 6.91341716182545e-05,
-      "loss": 0.1033,
+      "loss": 0.0146,
       "step": 55
     },
     {
       "epoch": 2.212345679012346,
-      "grad_norm": 0.08352261781692505,
+      "grad_norm": 0.043545201420784,
       "learning_rate": 6.783107663311565e-05,
-      "loss": 0.0952,
+      "loss": 0.0103,
       "step": 56
     },
     {
       "epoch": 2.251851851851852,
-      "grad_norm": 0.09827601909637451,
+      "grad_norm": 0.055670928210020065,
       "learning_rate": 6.651395309775837e-05,
-      "loss": 0.101,
+      "loss": 0.0147,
       "step": 57
     },
     {
       "epoch": 2.291358024691358,
-      "grad_norm": 0.10790225863456726,
+      "grad_norm": 0.056235238909721375,
       "learning_rate": 6.518383725548074e-05,
-      "loss": 0.1318,
+      "loss": 0.0186,
       "step": 58
     },
     {
       "epoch": 2.330864197530864,
-      "grad_norm": 0.09106322377920151,
+      "grad_norm": 0.060663700103759766,
       "learning_rate": 6.384177557124247e-05,
-      "loss": 0.0725,
+      "loss": 0.0178,
       "step": 59
     },
     {
       "epoch": 2.3703703703703702,
-      "grad_norm": 0.08517612516880035,
+      "grad_norm": 0.06017338111996651,
       "learning_rate": 6.248882390836135e-05,
-      "loss": 0.0224,
+      "loss": 0.0139,
       "step": 60
     },
     {
       "epoch": 2.3703703703703702,
-      "eval_loss": 0.05220804363489151,
-      "eval_runtime": 21.5704,
-      "eval_samples_per_second": 2.086,
-      "eval_steps_per_second": 2.086,
+      "eval_loss": 0.017185188829898834,
+      "eval_runtime": 27.1059,
+      "eval_samples_per_second": 1.66,
+      "eval_steps_per_second": 1.66,
       "step": 60
     },
     {
       "epoch": 2.4098765432098768,
-      "grad_norm": 0.10585787892341614,
+      "grad_norm": 0.060071442276239395,
       "learning_rate": 6.112604669781572e-05,
-      "loss": 0.0616,
+      "loss": 0.0137,
       "step": 61
     },
     {
       "epoch": 2.449382716049383,
-      "grad_norm": 0.08237098157405853,
+      "grad_norm": 0.0477650985121727,
       "learning_rate": 5.9754516100806423e-05,
-      "loss": 0.0636,
+      "loss": 0.0138,
       "step": 62
     },
     {
       "epoch": 2.488888888888889,
-      "grad_norm": 0.1071595624089241,
+      "grad_norm": 0.0581226609647274,
       "learning_rate": 5.837531116523682e-05,
-      "loss": 0.1148,
+      "loss": 0.0195,
       "step": 63
     },
     {
       "epoch": 2.528395061728395,
-      "grad_norm": 0.08002862334251404,
+      "grad_norm": 0.05158943682909012,
       "learning_rate": 5.698951697677498e-05,
-      "loss": 0.0568,
+      "loss": 0.0135,
       "step": 64
     },
     {
       "epoch": 2.567901234567901,
-      "grad_norm": 0.10249333083629608,
+      "grad_norm": 0.050511498004198074,
       "learning_rate": 5.559822380516539e-05,
-      "loss": 0.1027,
+      "loss": 0.0136,
       "step": 65
     },
     {
       "epoch": 2.6074074074074076,
-      "grad_norm": 0.09828495234251022,
+      "grad_norm": 0.05061330273747444,
       "learning_rate": 5.420252624646238e-05,
-      "loss": 0.1131,
+      "loss": 0.0145,
       "step": 66
     },
     {
       "epoch": 2.6469135802469137,
-      "grad_norm": 0.08260928839445114,
+      "grad_norm": 0.04876234754920006,
       "learning_rate": 5.2803522361859594e-05,
-      "loss": 0.0755,
+      "loss": 0.013,
       "step": 67
     },
     {
       "epoch": 2.68641975308642,
-      "grad_norm": 0.08810202777385712,
+      "grad_norm": 0.07571849226951599,
       "learning_rate": 5.140231281379345e-05,
-      "loss": 0.067,
+      "loss": 0.0153,
       "step": 68
     },
     {
       "epoch": 2.725925925925926,
-      "grad_norm": 0.09585391730070114,
+      "grad_norm": 0.06240665912628174,
       "learning_rate": 5e-05,
-      "loss": 0.0524,
+      "loss": 0.0166,
       "step": 69
     },
     {
       "epoch": 2.765432098765432,
-      "grad_norm": 0.08215411007404327,
+      "grad_norm": 0.05276874452829361,
       "learning_rate": 4.859768718620656e-05,
-      "loss": 0.0542,
+      "loss": 0.0179,
       "step": 70
     },
     {
       "epoch": 2.765432098765432,
-      "eval_loss": 0.05078903213143349,
-      "eval_runtime": 21.5513,
-      "eval_samples_per_second": 2.088,
-      "eval_steps_per_second": 2.088,
+      "eval_loss": 0.016723213717341423,
+      "eval_runtime": 27.0878,
+      "eval_samples_per_second": 1.661,
+      "eval_steps_per_second": 1.661,
       "step": 70
     },
     {
       "epoch": 2.8049382716049385,
-      "grad_norm": 0.0908372551202774,
+      "grad_norm": 0.05016285181045532,
       "learning_rate": 4.7196477638140404e-05,
-      "loss": 0.0836,
+      "loss": 0.0133,
       "step": 71
     },
     {
       "epoch": 2.8444444444444446,
-      "grad_norm": 0.09563232213258743,
+      "grad_norm": 0.052519507706165314,
       "learning_rate": 4.579747375353763e-05,
-      "loss": 0.043,
+      "loss": 0.011,
       "step": 72
     },
     {
       "epoch": 2.8839506172839506,
-      "grad_norm": 0.0988575667142868,
+      "grad_norm": 0.05423333868384361,
       "learning_rate": 4.4401776194834613e-05,
-      "loss": 0.1009,
+      "loss": 0.015,
       "step": 73
     },
     {
       "epoch": 2.9234567901234567,
-      "grad_norm": 0.10324296355247498,
+      "grad_norm": 0.04937339201569557,
       "learning_rate": 4.3010483023225045e-05,
-      "loss": 0.0785,
+      "loss": 0.0132,
       "step": 74
     },
     {
       "epoch": 2.962962962962963,
-      "grad_norm": 0.0935521200299263,
+      "grad_norm": 0.05095579847693443,
       "learning_rate": 4.162468883476319e-05,
-      "loss": 0.0766,
+      "loss": 0.0148,
       "step": 75
     },
     {
       "epoch": 3.0024691358024693,
-      "grad_norm": 0.19485773146152496,
+      "grad_norm": 0.09337777644395828,
       "learning_rate": 4.0245483899193595e-05,
-      "loss": 0.1528,
+      "loss": 0.0252,
       "step": 76
     },
     {
       "epoch": 3.0419753086419754,
-      "grad_norm": 0.10381461679935455,
+      "grad_norm": 0.056976284831762314,
       "learning_rate": 3.887395330218429e-05,
-      "loss": 0.0984,
+      "loss": 0.0127,
       "step": 77
     },
     {
       "epoch": 3.0814814814814815,
-      "grad_norm": 0.08173371851444244,
+      "grad_norm": 0.0528462678194046,
       "learning_rate": 3.7511176091638653e-05,
-      "loss": 0.0575,
+      "loss": 0.0129,
       "step": 78
     },
     {
       "epoch": 3.1209876543209876,
-      "grad_norm": 0.10292133688926697,
+      "grad_norm": 0.044976793229579926,
       "learning_rate": 3.6158224428757535e-05,
-      "loss": 0.1051,
+      "loss": 0.0127,
       "step": 79
     },
     {
       "epoch": 3.1604938271604937,
-      "grad_norm": 0.09561065584421158,
+      "grad_norm": 0.06988837569952011,
       "learning_rate": 3.4816162744519263e-05,
-      "loss": 0.0704,
+      "loss": 0.0137,
       "step": 80
     },
     {
       "epoch": 3.1604938271604937,
-      "eval_loss": 0.05049747973680496,
-      "eval_runtime": 21.5733,
-      "eval_samples_per_second": 2.086,
-      "eval_steps_per_second": 2.086,
+      "eval_loss": 0.01647907681763172,
+      "eval_runtime": 27.0766,
+      "eval_samples_per_second": 1.662,
+      "eval_steps_per_second": 1.662,
       "step": 80
     },
     {
       "epoch": 3.2,
-      "grad_norm": 0.08403361588716507,
+      "grad_norm": 0.051377538591623306,
       "learning_rate": 3.3486046902241664e-05,
-      "loss": 0.0672,
+      "loss": 0.0092,
       "step": 81
     },
     {
       "epoch": 3.2395061728395063,
-      "grad_norm": 0.08740091323852539,
+      "grad_norm": 0.04358401522040367,
       "learning_rate": 3.216892336688435e-05,
-      "loss": 0.0672,
+      "loss": 0.0117,
       "step": 82
     },
     {
       "epoch": 3.2790123456790123,
-      "grad_norm": 0.0924968495965004,
+      "grad_norm": 0.047665320336818695,
       "learning_rate": 3.086582838174551e-05,
-      "loss": 0.087,
+      "loss": 0.0116,
       "step": 83
     },
     {
       "epoch": 3.3185185185185184,
-      "grad_norm": 0.11069247871637344,
+      "grad_norm": 0.04360422119498253,
       "learning_rate": 2.9577787153200197e-05,
-      "loss": 0.111,
+      "loss": 0.0121,
       "step": 84
     },
     {
       "epoch": 3.3580246913580245,
-      "grad_norm": 0.08720285445451736,
+      "grad_norm": 0.03990411385893822,
       "learning_rate": 2.8305813044122097e-05,
-      "loss": 0.0597,
+      "loss": 0.0098,
       "step": 85
     },
     {
       "epoch": 3.397530864197531,
-      "grad_norm": 0.09598411619663239,
+      "grad_norm": 0.06174959987401962,
       "learning_rate": 2.705090677662311e-05,
-      "loss": 0.0588,
+      "loss": 0.0159,
       "step": 86
     },
     {
       "epoch": 3.437037037037037,
-      "grad_norm": 0.08057716488838196,
+      "grad_norm": 0.05259520933032036,
       "learning_rate": 2.581405564473801e-05,
-      "loss": 0.022,
+      "loss": 0.0134,
       "step": 87
     },
     {
       "epoch": 3.476543209876543,
-      "grad_norm": 0.0892864316701889,
+      "grad_norm": 0.04751445725560188,
       "learning_rate": 2.459623273767354e-05,
-      "loss": 0.0512,
+      "loss": 0.0103,
       "step": 88
     },
     {
       "epoch": 3.5160493827160493,
-      "grad_norm": 0.0995914563536644,
+      "grad_norm": 0.05354876071214676,
       "learning_rate": 2.3398396174233178e-05,
-      "loss": 0.0779,
+      "loss": 0.013,
       "step": 89
     },
     {
       "epoch": 3.5555555555555554,
-      "grad_norm": 0.11414390057325363,
+      "grad_norm": 0.07310397177934647,
       "learning_rate": 2.2221488349019903e-05,
-      "loss": 0.1065,
+      "loss": 0.0134,
       "step": 90
     },
     {
       "epoch": 3.5555555555555554,
-      "eval_loss": 0.0507526732981205,
-      "eval_runtime": 21.5686,
-      "eval_samples_per_second": 2.086,
-      "eval_steps_per_second": 2.086,
+      "eval_loss": 0.01649685949087143,
+      "eval_runtime": 27.0376,
+      "eval_samples_per_second": 1.664,
+      "eval_steps_per_second": 1.664,
       "step": 90
     },
     {
       "epoch": 3.595061728395062,
-      "grad_norm": 0.11288458108901978,
+      "grad_norm": 0.059673696756362915,
       "learning_rate": 2.1066435191009715e-05,
-      "loss": 0.0986,
+      "loss": 0.0142,
       "step": 91
     },
     {
       "epoch": 3.634567901234568,
-      "grad_norm": 0.1050805151462555,
+      "grad_norm": 0.05420045927166939,
       "learning_rate": 1.9934145435079702e-05,
-      "loss": 0.059,
+      "loss": 0.0118,
       "step": 92
     },
     {
       "epoch": 3.674074074074074,
-      "grad_norm": 0.09613631665706635,
+      "grad_norm": 0.056597091257572174,
       "learning_rate": 1.8825509907063327e-05,
-      "loss": 0.0638,
+      "loss": 0.0106,
       "step": 93
     },
     {
       "epoch": 3.71358024691358,
-      "grad_norm": 0.10993048548698425,
+      "grad_norm": 0.062004175037145615,
       "learning_rate": 1.774140082289563e-05,
-      "loss": 0.0602,
+      "loss": 0.0134,
       "step": 94
     },
     {
       "epoch": 3.753086419753086,
-      "grad_norm": 0.1023702397942543,
+      "grad_norm": 0.04919079318642616,
       "learning_rate": 1.6682671102399805e-05,
-      "loss": 0.0802,
+      "loss": 0.0142,
       "step": 95
     },
     {
       "epoch": 3.7925925925925927,
-      "grad_norm": 0.09606218338012695,
+      "grad_norm": 0.05011267215013504,
       "learning_rate": 1.5650153698254916e-05,
-      "loss": 0.0545,
+      "loss": 0.0104,
       "step": 96
     },
     {
       "epoch": 3.832098765432099,
-      "grad_norm": 0.13064627349376678,
+      "grad_norm": 0.0558934286236763,
       "learning_rate": 1.4644660940672627e-05,
-      "loss": 0.082,
+      "loss": 0.0111,
       "step": 97
     },
     {
       "epoch": 3.871604938271605,
-      "grad_norm": 0.12837804853916168,
+      "grad_norm": 0.055670589208602905,
       "learning_rate": 1.3666983898298657e-05,
-      "loss": 0.1217,
+      "loss": 0.0157,
       "step": 98
     },
     {
       "epoch": 3.911111111111111,
-      "grad_norm": 0.12296672910451889,
+      "grad_norm": 0.04987809434533119,
       "learning_rate": 1.2717891755841722e-05,
-      "loss": 0.1271,
+      "loss": 0.0118,
       "step": 99
     },
     {
       "epoch": 3.950617283950617,
-      "grad_norm": 0.09687791764736176,
+      "grad_norm": 0.05481395870447159,
       "learning_rate": 1.1798131208919627e-05,
-      "loss": 0.0613,
+      "loss": 0.0122,
       "step": 100
     },
     {
       "epoch": 3.950617283950617,
-      "eval_loss": 0.05035708099603653,
-      "eval_runtime": 21.5584,
-      "eval_samples_per_second": 2.087,
-      "eval_steps_per_second": 2.087,
+      "eval_loss": 0.016190797090530396,
+      "eval_runtime": 27.0005,
+      "eval_samples_per_second": 1.667,
+      "eval_steps_per_second": 1.667,
       "step": 100
     },
     {
       "epoch": 3.9901234567901236,
-      "grad_norm": 0.11310718208551407,
+      "grad_norm": 0.06512457877397537,
       "learning_rate": 1.090842587659851e-05,
-      "loss": 0.0817,
+      "loss": 0.0131,
       "step": 101
     },
     {
       "epoch": 4.029629629629629,
-      "grad_norm": 0.1681404858827591,
+      "grad_norm": 0.1306859254837036,
       "learning_rate": 1.004947573208756e-05,
-      "loss": 0.1086,
+      "loss": 0.0214,
       "step": 102
     },
     {
       "epoch": 4.069135802469136,
-      "grad_norm": 0.11545596271753311,
+      "grad_norm": 0.056599196046590805,
       "learning_rate": 9.221956552036992e-06,
-      "loss": 0.1082,
+      "loss": 0.0105,
       "step": 103
     },
     {
       "epoch": 4.108641975308642,
-      "grad_norm": 0.05956997349858284,
+      "grad_norm": 0.04881014674901962,
       "learning_rate": 8.426519384872733e-06,
-      "loss": 0.0117,
+      "loss": 0.0103,
       "step": 104
     },
     {
       "epoch": 4.148148148148148,
-      "grad_norm": 0.10608723759651184,
+      "grad_norm": 0.04762539267539978,
       "learning_rate": 7.663790038585793e-06,
-      "loss": 0.0907,
+      "loss": 0.0121,
       "step": 105
     },
     {
       "epoch": 4.187654320987654,
-      "grad_norm": 0.0811762735247612,
+      "grad_norm": 0.04193001613020897,
       "learning_rate": 6.934368588379553e-06,
-      "loss": 0.0606,
+      "loss": 0.0083,
       "step": 106
     },
     {
       "epoch": 4.22716049382716,
-      "grad_norm": 0.11676077544689178,
+      "grad_norm": 0.04855426400899887,
       "learning_rate": 6.238828904562316e-06,
-      "loss": 0.1158,
+      "loss": 0.0116,
       "step": 107
     },
     {
       "epoch": 4.266666666666667,
-      "grad_norm": 0.09994089603424072,
+      "grad_norm": 0.0491572730243206,
       "learning_rate": 5.577718201056392e-06,
-      "loss": 0.0746,
+      "loss": 0.0116,
       "step": 108
     },
     {
       "epoch": 4.306172839506173,
-      "grad_norm": 0.085332952439785,
+      "grad_norm": 0.04440079629421234,
       "learning_rate": 4.951556604879048e-06,
-      "loss": 0.0622,
+      "loss": 0.0109,
       "step": 109
     },
     {
       "epoch": 4.345679012345679,
-      "grad_norm": 0.09871704131364822,
+      "grad_norm": 0.06093370541930199,
       "learning_rate": 4.360836746934055e-06,
-      "loss": 0.0526,
+      "loss": 0.0112,
       "step": 110
     },
     {
       "epoch": 4.345679012345679,
-      "eval_loss": 0.050025228410959244,
-      "eval_runtime": 21.5862,
-      "eval_samples_per_second": 2.085,
-      "eval_steps_per_second": 2.085,
+      "eval_loss": 0.01600480265915394,
+      "eval_runtime": 26.9779,
+      "eval_samples_per_second": 1.668,
+      "eval_steps_per_second": 1.668,
       "step": 110
     },
     {
       "epoch": 4.385185185185185,
-      "grad_norm": 0.09276178479194641,
+      "grad_norm": 0.04132888838648796,
       "learning_rate": 3.8060233744356633e-06,
-      "loss": 0.0814,
+      "loss": 0.0102,
       "step": 111
     },
     {
       "epoch": 4.424691358024692,
-      "grad_norm": 0.076078861951828,
+      "grad_norm": 0.03689916059374809,
       "learning_rate": 3.2875529852700147e-06,
-      "loss": 0.0414,
+      "loss": 0.0084,
       "step": 112
     },
     {
       "epoch": 4.4641975308641975,
-      "grad_norm": 0.0995572954416275,
+      "grad_norm": 0.052065420895814896,
       "learning_rate": 2.8058334845816213e-06,
-      "loss": 0.0657,
+      "loss": 0.0119,
       "step": 113
     },
     {
       "epoch": 4.503703703703704,
-      "grad_norm": 0.10301052778959274,
+      "grad_norm": 0.04521874710917473,
       "learning_rate": 2.361243863855184e-06,
-      "loss": 0.0963,
+      "loss": 0.0122,
       "step": 114
     },
     {
       "epoch": 4.54320987654321,
-      "grad_norm": 0.11387384682893753,
+      "grad_norm": 0.05591542646288872,
       "learning_rate": 1.9541339027450256e-06,
-      "loss": 0.1072,
+      "loss": 0.0131,
       "step": 115
     },
     {
       "epoch": 4.582716049382716,
-      "grad_norm": 0.10680479556322098,
+      "grad_norm": 0.05921711027622223,
       "learning_rate": 1.584823893886933e-06,
-      "loss": 0.0651,
+      "loss": 0.0132,
       "step": 116
     },
     {
       "epoch": 4.622222222222222,
-      "grad_norm": 0.10225462168455124,
+      "grad_norm": 0.05020372197031975,
       "learning_rate": 1.2536043909088191e-06,
-      "loss": 0.0652,
+      "loss": 0.0094,
       "step": 117
     },
     {
       "epoch": 4.661728395061728,
-      "grad_norm": 0.10798352211713791,
+      "grad_norm": 0.05627156421542168,
       "learning_rate": 9.607359798384785e-07,
-      "loss": 0.064,
+      "loss": 0.0132,
       "step": 118
     },
     {
       "epoch": 4.701234567901235,
-      "grad_norm": 0.08984480053186417,
+      "grad_norm": 0.05107918381690979,
       "learning_rate": 7.064490740882057e-07,
-      "loss": 0.0533,
+      "loss": 0.0107,
       "step": 119
     },
     {
       "epoch": 4.7407407407407405,
-      "grad_norm": 0.24465854465961456,
+      "grad_norm": 0.05450572445988655,
       "learning_rate": 4.909437331777179e-07,
-      "loss": 0.057,
+      "loss": 0.0134,
       "step": 120
     },
     {
       "epoch": 4.7407407407407405,
-      "eval_loss": 0.05005127191543579,
-      "eval_runtime": 21.5938,
-      "eval_samples_per_second": 2.084,
-      "eval_steps_per_second": 2.084,
+      "eval_loss": 0.016016781330108643,
+      "eval_runtime": 26.9537,
+      "eval_samples_per_second": 1.67,
+      "eval_steps_per_second": 1.67,
       "step": 120
     },
     {
       "epoch": 4.780246913580247,
-      "grad_norm": 0.09685765951871872,
+      "grad_norm": 0.05025794729590416,
       "learning_rate": 3.143895053378698e-07,
-      "loss": 0.0811,
+      "loss": 0.0119,
       "step": 121
     },
     {
       "epoch": 4.8197530864197535,
-      "grad_norm": 0.11006279289722443,
+      "grad_norm": 0.052281949669122696,
       "learning_rate": 1.7692529411904578e-07,
-      "loss": 0.0958,
+      "loss": 0.0131,
       "step": 122
     },
     {
       "epoch": 4.859259259259259,
-      "grad_norm": 0.11954176425933838,
+      "grad_norm": 0.04726499691605568,
       "learning_rate": 7.865924910916977e-08,
-      "loss": 0.1079,
+      "loss": 0.0112,
       "step": 123
     },
     {
       "epoch": 4.898765432098766,
-      "grad_norm": 0.08760962635278702,
+      "grad_norm": 0.06482001394033432,
       "learning_rate": 1.9668680847356735e-08,
-      "loss": 0.0238,
+      "loss": 0.0123,
       "step": 124
     },
     {
       "epoch": 4.938271604938271,
-      "grad_norm": 0.10748548060655594,
+      "grad_norm": 0.049361273646354675,
       "learning_rate": 0.0,
-      "loss": 0.0797,
+      "loss": 0.0119,
       "step": 125
     },
     {
       "epoch": 4.938271604938271,
       "step": 125,
-      "total_flos": 1.5205115961940378e+17,
-      "train_loss": 0.10026876297593117,
-      "train_runtime": 3651.0486,
-      "train_samples_per_second": 0.555,
-      "train_steps_per_second": 0.034
+      "total_flos": 1.931743422132388e+17,
+      "train_loss": 0.026140653513371945,
+      "train_runtime": 5707.3107,
+      "train_samples_per_second": 0.355,
+      "train_steps_per_second": 0.022
     }
   ],
   "logging_steps": 1,
@@ -1006,7 +1006,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 1.5205115961940378e+17,
+  "total_flos": 1.931743422132388e+17,
   "train_batch_size": 1,
   "trial_name": null,
   "trial_params": null