End of training

README.md

@@ -3,6 +3,8 @@ license: mit
 base_model: unicamp-dl/ptt5-base-portuguese-vocab
 tags:
 - generated_from_trainer
+datasets:
+- tiagoblima/du-qg-squadv1_pt
 model-index:
 - name: t5_base-qg-ap-oficial
   results: []
@@ -13,9 +15,9 @@ should probably proofread and complete it, then remove this comment. -->
 
 # t5_base-qg-ap-oficial
 
-This model is a fine-tuned version of [unicamp-dl/ptt5-base-portuguese-vocab](https://huggingface.co/unicamp-dl/ptt5-base-portuguese-vocab) on an unknown dataset.
+This model is a fine-tuned version of [unicamp-dl/ptt5-base-portuguese-vocab](https://huggingface.co/unicamp-dl/ptt5-base-portuguese-vocab) on the tiagoblima/du-qg-squadv1_pt dataset.
 It achieves the following results on the evaluation set:
-- Loss: 1.7673
+- Loss: 1.7632
 
 ## Model description
 

all_results.json

@@ -1,13 +1,13 @@
 {
     "epoch": 5.0,
-    "eval_loss": 1.9945117235183716,
-    "eval_runtime": 302.9759,
-    "eval_samples": 10647,
-    "eval_samples_per_second": 35.141,
-    "eval_steps_per_second": 8.786,
-    "train_loss": 1.9853319837194525,
-    "train_runtime": 42362.5347,
+    "eval_loss": 1.7631582021713257,
+    "eval_runtime": 41.211,
+    "eval_samples": 10649,
+    "eval_samples_per_second": 258.402,
+    "eval_steps_per_second": 4.052,
+    "train_loss": 1.7503460931234034,
+    "train_runtime": 5808.8161,
     "train_samples": 75722,
-    "train_samples_per_second": 8.937,
-    "train_steps_per_second": 0.279
+    "train_samples_per_second": 65.179,
+    "train_steps_per_second": 2.037
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 5.0,
-    "eval_loss": 1.9945117235183716,
-    "eval_runtime": 302.9759,
-    "eval_samples": 10647,
-    "eval_samples_per_second": 35.141,
-    "eval_steps_per_second": 8.786
+    "eval_loss": 1.7631582021713257,
+    "eval_runtime": 41.211,
+    "eval_samples": 10649,
+    "eval_samples_per_second": 258.402,
+    "eval_steps_per_second": 4.052
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 5.0,
-    "train_loss": 1.9853319837194525,
-    "train_runtime": 42362.5347,
+    "train_loss": 1.7503460931234034,
+    "train_runtime": 5808.8161,
     "train_samples": 75722,
-    "train_samples_per_second": 8.937,
-    "train_steps_per_second": 0.279
+    "train_samples_per_second": 65.179,
+    "train_steps_per_second": 2.037
 }

trainer_state.json

@@ -1,776 +1,776 @@
 {
-  "best_metric": 1.9945117235183716,
-  "best_model_checkpoint": "/temp/t5_base-qg-ap-oficial/checkpoint-9466",
-  "epoch": 4.99894358757659,
+  "best_metric": 1.7631582021713257,
+  "best_model_checkpoint": "/temp/t5_base-qg-ap-oficial/checkpoint-9468",
+  "epoch": 5.0,
   "eval_steps": 500,
-  "global_step": 11830,
+  "global_step": 11835,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 0.04,
-      "learning_rate": 9.915469146238378e-05,
-      "loss": 3.6356,
+      "learning_rate": 9.915504858470639e-05,
+      "loss": 3.6112,
       "step": 100
     },
     {
       "epoch": 0.08,
-      "learning_rate": 9.830938292476755e-05,
-      "loss": 2.4983,
+      "learning_rate": 9.831009716941276e-05,
+      "loss": 2.3844,
       "step": 200
     },
     {
       "epoch": 0.13,
-      "learning_rate": 9.746407438715132e-05,
-      "loss": 2.4076,
+      "learning_rate": 9.746514575411914e-05,
+      "loss": 2.2431,
       "step": 300
     },
     {
       "epoch": 0.17,
-      "learning_rate": 9.661876584953509e-05,
-      "loss": 2.3614,
+      "learning_rate": 9.662019433882552e-05,
+      "loss": 2.1735,
       "step": 400
     },
     {
       "epoch": 0.21,
-      "learning_rate": 9.577345731191886e-05,
-      "loss": 2.3141,
+      "learning_rate": 9.57752429235319e-05,
+      "loss": 2.116,
       "step": 500
     },
     {
       "epoch": 0.25,
-      "learning_rate": 9.492814877430263e-05,
-      "loss": 2.3573,
+      "learning_rate": 9.493029150823828e-05,
+      "loss": 2.1547,
       "step": 600
     },
     {
       "epoch": 0.3,
-      "learning_rate": 9.408284023668639e-05,
-      "loss": 2.2916,
+      "learning_rate": 9.408534009294467e-05,
+      "loss": 2.0778,
       "step": 700
     },
     {
       "epoch": 0.34,
-      "learning_rate": 9.323753169907016e-05,
-      "loss": 2.2743,
+      "learning_rate": 9.324038867765104e-05,
+      "loss": 2.0569,
       "step": 800
     },
     {
       "epoch": 0.38,
-      "learning_rate": 9.239222316145394e-05,
-      "loss": 2.2709,
+      "learning_rate": 9.239543726235742e-05,
+      "loss": 2.0588,
       "step": 900
     },
     {
       "epoch": 0.42,
-      "learning_rate": 9.15469146238377e-05,
-      "loss": 2.2874,
+      "learning_rate": 9.15504858470638e-05,
+      "loss": 2.0605,
       "step": 1000
     },
     {
       "epoch": 0.46,
-      "learning_rate": 9.070160608622148e-05,
-      "loss": 2.2176,
+      "learning_rate": 9.070553443177018e-05,
+      "loss": 2.0073,
       "step": 1100
     },
     {
       "epoch": 0.51,
-      "learning_rate": 8.985629754860525e-05,
-      "loss": 2.2363,
+      "learning_rate": 8.986058301647655e-05,
+      "loss": 2.0203,
       "step": 1200
     },
     {
       "epoch": 0.55,
-      "learning_rate": 8.901098901098901e-05,
-      "loss": 2.2104,
+      "learning_rate": 8.901563160118294e-05,
+      "loss": 1.9931,
       "step": 1300
     },
     {
       "epoch": 0.59,
-      "learning_rate": 8.816568047337278e-05,
-      "loss": 2.2257,
+      "learning_rate": 8.817068018588932e-05,
+      "loss": 1.9946,
       "step": 1400
     },
     {
       "epoch": 0.63,
-      "learning_rate": 8.732037193575655e-05,
-      "loss": 2.2096,
+      "learning_rate": 8.732572877059569e-05,
+      "loss": 1.9725,
       "step": 1500
     },
     {
       "epoch": 0.68,
-      "learning_rate": 8.647506339814032e-05,
-      "loss": 2.2201,
+      "learning_rate": 8.648077735530208e-05,
+      "loss": 1.988,
       "step": 1600
     },
     {
       "epoch": 0.72,
-      "learning_rate": 8.56297548605241e-05,
-      "loss": 2.1754,
+      "learning_rate": 8.563582594000845e-05,
+      "loss": 1.9505,
       "step": 1700
     },
     {
       "epoch": 0.76,
-      "learning_rate": 8.478444632290787e-05,
-      "loss": 2.2049,
+      "learning_rate": 8.479087452471483e-05,
+      "loss": 1.9758,
       "step": 1800
     },
     {
       "epoch": 0.8,
-      "learning_rate": 8.393913778529164e-05,
-      "loss": 2.165,
+      "learning_rate": 8.394592310942122e-05,
+      "loss": 1.937,
       "step": 1900
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 8.309382924767541e-05,
-      "loss": 2.1974,
+      "epoch": 0.84,
+      "learning_rate": 8.31009716941276e-05,
+      "loss": 1.9751,
       "step": 2000
     },
     {
       "epoch": 0.89,
-      "learning_rate": 8.224852071005918e-05,
-      "loss": 2.1561,
+      "learning_rate": 8.225602027883397e-05,
+      "loss": 1.9266,
       "step": 2100
     },
     {
       "epoch": 0.93,
-      "learning_rate": 8.140321217244294e-05,
-      "loss": 2.1517,
+      "learning_rate": 8.141106886354036e-05,
+      "loss": 1.9268,
       "step": 2200
     },
     {
       "epoch": 0.97,
-      "learning_rate": 8.055790363482671e-05,
-      "loss": 2.1593,
+      "learning_rate": 8.056611744824673e-05,
+      "loss": 1.928,
       "step": 2300
     },
     {
       "epoch": 1.0,
-      "eval_loss": 2.0508947372436523,
-      "eval_runtime": 298.019,
-      "eval_samples_per_second": 35.726,
-      "eval_steps_per_second": 8.932,
-      "step": 2366
+      "eval_loss": 1.831241250038147,
+      "eval_runtime": 41.2883,
+      "eval_samples_per_second": 257.918,
+      "eval_steps_per_second": 4.045,
+      "step": 2367
     },
     {
       "epoch": 1.01,
-      "learning_rate": 7.971259509721048e-05,
-      "loss": 2.1273,
+      "learning_rate": 7.972116603295311e-05,
+      "loss": 1.8918,
       "step": 2400
     },
     {
       "epoch": 1.06,
-      "learning_rate": 7.886728655959425e-05,
-      "loss": 2.0831,
+      "learning_rate": 7.887621461765948e-05,
+      "loss": 1.8437,
       "step": 2500
     },
     {
       "epoch": 1.1,
-      "learning_rate": 7.802197802197802e-05,
-      "loss": 2.0649,
+      "learning_rate": 7.803126320236587e-05,
+      "loss": 1.8358,
       "step": 2600
     },
     {
       "epoch": 1.14,
-      "learning_rate": 7.71766694843618e-05,
-      "loss": 2.0674,
+      "learning_rate": 7.718631178707225e-05,
+      "loss": 1.8372,
       "step": 2700
     },
     {
       "epoch": 1.18,
-      "learning_rate": 7.633136094674557e-05,
-      "loss": 2.0431,
+      "learning_rate": 7.634136037177862e-05,
+      "loss": 1.8133,
       "step": 2800
     },
     {
       "epoch": 1.23,
-      "learning_rate": 7.548605240912934e-05,
-      "loss": 2.051,
+      "learning_rate": 7.5496408956485e-05,
+      "loss": 1.8091,
       "step": 2900
     },
     {
       "epoch": 1.27,
-      "learning_rate": 7.464074387151311e-05,
-      "loss": 2.0503,
+      "learning_rate": 7.465145754119138e-05,
+      "loss": 1.8094,
       "step": 3000
     },
     {
       "epoch": 1.31,
-      "learning_rate": 7.379543533389688e-05,
-      "loss": 2.0434,
+      "learning_rate": 7.380650612589776e-05,
+      "loss": 1.7991,
       "step": 3100
     },
     {
       "epoch": 1.35,
-      "learning_rate": 7.295012679628065e-05,
-      "loss": 2.0142,
+      "learning_rate": 7.296155471060415e-05,
+      "loss": 1.7784,
       "step": 3200
     },
     {
       "epoch": 1.39,
-      "learning_rate": 7.210481825866442e-05,
-      "loss": 2.0515,
+      "learning_rate": 7.211660329531053e-05,
+      "loss": 1.8179,
       "step": 3300
     },
     {
       "epoch": 1.44,
-      "learning_rate": 7.12595097210482e-05,
-      "loss": 2.0314,
+      "learning_rate": 7.12716518800169e-05,
+      "loss": 1.8024,
       "step": 3400
     },
     {
       "epoch": 1.48,
-      "learning_rate": 7.041420118343195e-05,
-      "loss": 2.0394,
+      "learning_rate": 7.042670046472329e-05,
+      "loss": 1.8008,
       "step": 3500
     },
     {
       "epoch": 1.52,
-      "learning_rate": 6.956889264581573e-05,
-      "loss": 2.0207,
+      "learning_rate": 6.958174904942966e-05,
+      "loss": 1.7826,
       "step": 3600
     },
     {
       "epoch": 1.56,
-      "learning_rate": 6.87235841081995e-05,
-      "loss": 2.0339,
+      "learning_rate": 6.873679763413604e-05,
+      "loss": 1.7905,
       "step": 3700
     },
     {
       "epoch": 1.61,
-      "learning_rate": 6.787827557058326e-05,
-      "loss": 2.0312,
+      "learning_rate": 6.789184621884243e-05,
+      "loss": 1.7919,
       "step": 3800
     },
     {
       "epoch": 1.65,
-      "learning_rate": 6.703296703296703e-05,
-      "loss": 2.0106,
+      "learning_rate": 6.70468948035488e-05,
+      "loss": 1.7643,
       "step": 3900
     },
     {
       "epoch": 1.69,
-      "learning_rate": 6.61876584953508e-05,
-      "loss": 2.03,
+      "learning_rate": 6.620194338825518e-05,
+      "loss": 1.7871,
       "step": 4000
     },
     {
       "epoch": 1.73,
-      "learning_rate": 6.534234995773457e-05,
-      "loss": 2.0075,
+      "learning_rate": 6.535699197296155e-05,
+      "loss": 1.7734,
       "step": 4100
     },
     {
       "epoch": 1.77,
-      "learning_rate": 6.449704142011834e-05,
-      "loss": 2.0339,
+      "learning_rate": 6.451204055766794e-05,
+      "loss": 1.7982,
       "step": 4200
     },
     {
       "epoch": 1.82,
-      "learning_rate": 6.365173288250211e-05,
-      "loss": 2.0053,
+      "learning_rate": 6.366708914237431e-05,
+      "loss": 1.7695,
       "step": 4300
     },
     {
       "epoch": 1.86,
-      "learning_rate": 6.280642434488588e-05,
-      "loss": 2.0201,
+      "learning_rate": 6.282213772708069e-05,
+      "loss": 1.7735,
       "step": 4400
     },
     {
       "epoch": 1.9,
-      "learning_rate": 6.196111580726966e-05,
-      "loss": 2.0502,
+      "learning_rate": 6.197718631178708e-05,
+      "loss": 1.8112,
       "step": 4500
     },
     {
       "epoch": 1.94,
-      "learning_rate": 6.111580726965343e-05,
-      "loss": 1.9973,
+      "learning_rate": 6.113223489649346e-05,
+      "loss": 1.7556,
       "step": 4600
     },
     {
       "epoch": 1.99,
-      "learning_rate": 6.027049873203719e-05,
-      "loss": 1.9907,
+      "learning_rate": 6.028728348119983e-05,
+      "loss": 1.7467,
       "step": 4700
     },
     {
       "epoch": 2.0,
-      "eval_loss": 2.015758514404297,
-      "eval_runtime": 300.3315,
-      "eval_samples_per_second": 35.451,
-      "eval_steps_per_second": 8.864,
-      "step": 4733
+      "eval_loss": 1.78754460811615,
+      "eval_runtime": 41.3155,
+      "eval_samples_per_second": 257.749,
+      "eval_steps_per_second": 4.042,
+      "step": 4734
     },
     {
       "epoch": 2.03,
-      "learning_rate": 5.9425190194420964e-05,
-      "loss": 1.9456,
+      "learning_rate": 5.944233206590622e-05,
+      "loss": 1.7035,
       "step": 4800
     },
     {
       "epoch": 2.07,
-      "learning_rate": 5.8579881656804736e-05,
-      "loss": 1.9231,
+      "learning_rate": 5.859738065061259e-05,
+      "loss": 1.6792,
       "step": 4900
     },
     {
       "epoch": 2.11,
-      "learning_rate": 5.773457311918851e-05,
-      "loss": 1.9542,
+      "learning_rate": 5.775242923531897e-05,
+      "loss": 1.701,
       "step": 5000
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 5.688926458157228e-05,
-      "loss": 1.9218,
+      "epoch": 2.15,
+      "learning_rate": 5.6907477820025357e-05,
+      "loss": 1.6836,
       "step": 5100
     },
     {
       "epoch": 2.2,
-      "learning_rate": 5.604395604395605e-05,
-      "loss": 1.925,
+      "learning_rate": 5.606252640473173e-05,
+      "loss": 1.6835,
       "step": 5200
     },
     {
       "epoch": 2.24,
-      "learning_rate": 5.5198647506339815e-05,
-      "loss": 1.9241,
+      "learning_rate": 5.521757498943811e-05,
+      "loss": 1.6907,
       "step": 5300
     },
     {
       "epoch": 2.28,
-      "learning_rate": 5.4353338968723586e-05,
-      "loss": 1.9269,
+      "learning_rate": 5.437262357414449e-05,
+      "loss": 1.6843,
       "step": 5400
     },
     {
       "epoch": 2.32,
-      "learning_rate": 5.350803043110736e-05,
-      "loss": 1.9384,
+      "learning_rate": 5.3527672158850874e-05,
+      "loss": 1.6934,
       "step": 5500
     },
     {
       "epoch": 2.37,
-      "learning_rate": 5.266272189349113e-05,
-      "loss": 1.9325,
+      "learning_rate": 5.2682720743557245e-05,
+      "loss": 1.687,
       "step": 5600
     },
     {
       "epoch": 2.41,
-      "learning_rate": 5.18174133558749e-05,
-      "loss": 1.9102,
+      "learning_rate": 5.183776932826363e-05,
+      "loss": 1.6619,
       "step": 5700
     },
     {
       "epoch": 2.45,
-      "learning_rate": 5.097210481825867e-05,
-      "loss": 1.9009,
+      "learning_rate": 5.099281791297e-05,
+      "loss": 1.6522,
       "step": 5800
     },
     {
       "epoch": 2.49,
-      "learning_rate": 5.012679628064244e-05,
-      "loss": 1.9148,
+      "learning_rate": 5.0147866497676385e-05,
+      "loss": 1.6762,
       "step": 5900
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 4.928148774302621e-05,
-      "loss": 1.9225,
+      "epoch": 2.53,
+      "learning_rate": 4.930291508238277e-05,
+      "loss": 1.6794,
       "step": 6000
     },
     {
       "epoch": 2.58,
-      "learning_rate": 4.8436179205409974e-05,
-      "loss": 1.9465,
+      "learning_rate": 4.845796366708915e-05,
+      "loss": 1.7034,
       "step": 6100
     },
     {
       "epoch": 2.62,
-      "learning_rate": 4.7590870667793745e-05,
-      "loss": 1.917,
+      "learning_rate": 4.7613012251795525e-05,
+      "loss": 1.6853,
       "step": 6200
     },
     {
       "epoch": 2.66,
-      "learning_rate": 4.674556213017752e-05,
-      "loss": 1.8996,
+      "learning_rate": 4.67680608365019e-05,
+      "loss": 1.6645,
       "step": 6300
     },
     {
       "epoch": 2.7,
-      "learning_rate": 4.590025359256129e-05,
-      "loss": 1.93,
+      "learning_rate": 4.592310942120828e-05,
+      "loss": 1.6819,
       "step": 6400
     },
     {
       "epoch": 2.75,
-      "learning_rate": 4.505494505494506e-05,
-      "loss": 1.9288,
+      "learning_rate": 4.507815800591466e-05,
+      "loss": 1.6949,
       "step": 6500
     },
     {
       "epoch": 2.79,
-      "learning_rate": 4.420963651732883e-05,
-      "loss": 1.9017,
+      "learning_rate": 4.423320659062104e-05,
+      "loss": 1.6703,
       "step": 6600
     },
     {
       "epoch": 2.83,
-      "learning_rate": 4.3364327979712596e-05,
-      "loss": 1.9277,
+      "learning_rate": 4.338825517532742e-05,
+      "loss": 1.6861,
       "step": 6700
     },
     {
       "epoch": 2.87,
-      "learning_rate": 4.251901944209637e-05,
-      "loss": 1.9238,
+      "learning_rate": 4.25433037600338e-05,
+      "loss": 1.6808,
       "step": 6800
     },
     {
       "epoch": 2.92,
-      "learning_rate": 4.167371090448014e-05,
-      "loss": 1.9105,
+      "learning_rate": 4.1698352344740176e-05,
+      "loss": 1.6715,
       "step": 6900
     },
     {
       "epoch": 2.96,
-      "learning_rate": 4.0828402366863904e-05,
-      "loss": 1.9408,
+      "learning_rate": 4.085340092944656e-05,
+      "loss": 1.7014,
       "step": 7000
     },
     {
       "epoch": 3.0,
-      "eval_loss": 2.001091957092285,
-      "eval_runtime": 301.8989,
-      "eval_samples_per_second": 35.267,
-      "eval_steps_per_second": 8.818,
-      "step": 7099
+      "learning_rate": 4.000844951415294e-05,
+      "loss": 1.6658,
+      "step": 7100
     },
     {
       "epoch": 3.0,
-      "learning_rate": 3.9983093829247675e-05,
-      "loss": 1.8997,
-      "step": 7100
+      "eval_loss": 1.7680878639221191,
+      "eval_runtime": 41.258,
+      "eval_samples_per_second": 258.108,
+      "eval_steps_per_second": 4.048,
+      "step": 7101
     },
     {
       "epoch": 3.04,
-      "learning_rate": 3.913778529163145e-05,
-      "loss": 1.8823,
+      "learning_rate": 3.916349809885932e-05,
+      "loss": 1.6319,
       "step": 7200
     },
     {
       "epoch": 3.08,
-      "learning_rate": 3.829247675401522e-05,
-      "loss": 1.871,
+      "learning_rate": 3.83185466835657e-05,
+      "loss": 1.6104,
       "step": 7300
     },
     {
       "epoch": 3.13,
-      "learning_rate": 3.744716821639899e-05,
-      "loss": 1.8499,
+      "learning_rate": 3.747359526827208e-05,
+      "loss": 1.5933,
       "step": 7400
     },
     {
       "epoch": 3.17,
-      "learning_rate": 3.6601859678782755e-05,
-      "loss": 1.865,
+      "learning_rate": 3.6628643852978456e-05,
+      "loss": 1.6245,
       "step": 7500
     },
     {
       "epoch": 3.21,
-      "learning_rate": 3.5756551141166526e-05,
-      "loss": 1.8619,
+      "learning_rate": 3.5783692437684833e-05,
+      "loss": 1.6231,
       "step": 7600
     },
     {
       "epoch": 3.25,
-      "learning_rate": 3.49112426035503e-05,
-      "loss": 1.8613,
+      "learning_rate": 3.493874102239121e-05,
+      "loss": 1.6075,
       "step": 7700
     },
     {
       "epoch": 3.3,
-      "learning_rate": 3.406593406593407e-05,
-      "loss": 1.8476,
+      "learning_rate": 3.4093789607097596e-05,
+      "loss": 1.6069,
       "step": 7800
     },
     {
       "epoch": 3.34,
-      "learning_rate": 3.322062552831784e-05,
-      "loss": 1.8391,
+      "learning_rate": 3.324883819180397e-05,
+      "loss": 1.5876,
       "step": 7900
     },
     {
       "epoch": 3.38,
-      "learning_rate": 3.237531699070161e-05,
-      "loss": 1.8497,
+      "learning_rate": 3.240388677651035e-05,
+      "loss": 1.6071,
       "step": 8000
     },
     {
       "epoch": 3.42,
-      "learning_rate": 3.153000845308538e-05,
-      "loss": 1.8666,
+      "learning_rate": 3.155893536121673e-05,
+      "loss": 1.6211,
       "step": 8100
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 3.068469991546914e-05,
-      "loss": 1.8625,
+      "epoch": 3.46,
+      "learning_rate": 3.0713983945923107e-05,
+      "loss": 1.6241,
       "step": 8200
     },
     {
       "epoch": 3.51,
-      "learning_rate": 2.9839391377852917e-05,
-      "loss": 1.8408,
+      "learning_rate": 2.986903253062949e-05,
+      "loss": 1.6074,
       "step": 8300
     },
     {
       "epoch": 3.55,
-      "learning_rate": 2.8994082840236685e-05,
-      "loss": 1.8434,
+      "learning_rate": 2.9024081115335872e-05,
+      "loss": 1.6009,
       "step": 8400
     },
     {
       "epoch": 3.59,
-      "learning_rate": 2.8148774302620457e-05,
-      "loss": 1.8483,
+      "learning_rate": 2.817912970004225e-05,
+      "loss": 1.6115,
       "step": 8500
     },
     {
       "epoch": 3.63,
-      "learning_rate": 2.7303465765004228e-05,
-      "loss": 1.8432,
+      "learning_rate": 2.733417828474863e-05,
+      "loss": 1.5905,
       "step": 8600
     },
     {
       "epoch": 3.68,
-      "learning_rate": 2.6458157227387996e-05,
-      "loss": 1.8306,
+      "learning_rate": 2.648922686945501e-05,
+      "loss": 1.5773,
       "step": 8700
     },
     {
       "epoch": 3.72,
-      "learning_rate": 2.5612848689771768e-05,
-      "loss": 1.8391,
+      "learning_rate": 2.5644275454161386e-05,
+      "loss": 1.5943,
       "step": 8800
     },
     {
       "epoch": 3.76,
-      "learning_rate": 2.476754015215554e-05,
-      "loss": 1.8596,
+      "learning_rate": 2.4799324038867768e-05,
+      "loss": 1.6136,
       "step": 8900
     },
     {
       "epoch": 3.8,
-      "learning_rate": 2.3922231614539308e-05,
-      "loss": 1.8539,
+      "learning_rate": 2.3954372623574145e-05,
+      "loss": 1.6069,
       "step": 9000
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 2.307692307692308e-05,
-      "loss": 1.8408,
+      "epoch": 3.84,
+      "learning_rate": 2.3109421208280523e-05,
+      "loss": 1.6001,
       "step": 9100
     },
     {
       "epoch": 3.89,
-      "learning_rate": 2.2231614539306847e-05,
-      "loss": 1.8436,
+      "learning_rate": 2.2264469792986904e-05,
+      "loss": 1.5972,
       "step": 9200
     },
     {
       "epoch": 3.93,
-      "learning_rate": 2.138630600169062e-05,
-      "loss": 1.8622,
+      "learning_rate": 2.1419518377693285e-05,
+      "loss": 1.6209,
       "step": 9300
     },
     {
       "epoch": 3.97,
-      "learning_rate": 2.0540997464074387e-05,
-      "loss": 1.8321,
+      "learning_rate": 2.0574566962399663e-05,
+      "loss": 1.5892,
       "step": 9400
     },
     {
       "epoch": 4.0,
-      "eval_loss": 1.9945117235183716,
-      "eval_runtime": 300.7616,
-      "eval_samples_per_second": 35.4,
-      "eval_steps_per_second": 8.851,
-      "step": 9466
+      "eval_loss": 1.7631582021713257,
+      "eval_runtime": 41.2622,
+      "eval_samples_per_second": 258.081,
+      "eval_steps_per_second": 4.047,
+      "step": 9468
     },
     {
       "epoch": 4.01,
-      "learning_rate": 1.969568892645816e-05,
-      "loss": 1.8354,
+      "learning_rate": 1.9729615547106044e-05,
+      "loss": 1.5887,
       "step": 9500
     },
     {
       "epoch": 4.06,
-      "learning_rate": 1.885038038884193e-05,
-      "loss": 1.7986,
+      "learning_rate": 1.8884664131812422e-05,
+      "loss": 1.555,
       "step": 9600
     },
     {
       "epoch": 4.1,
-      "learning_rate": 1.8005071851225698e-05,
-      "loss": 1.8159,
+      "learning_rate": 1.80397127165188e-05,
+      "loss": 1.5713,
       "step": 9700
     },
     {
       "epoch": 4.14,
-      "learning_rate": 1.7159763313609466e-05,
-      "loss": 1.8244,
+      "learning_rate": 1.719476130122518e-05,
+      "loss": 1.5748,
       "step": 9800
     },
     {
       "epoch": 4.18,
-      "learning_rate": 1.6314454775993238e-05,
-      "loss": 1.8199,
+      "learning_rate": 1.634980988593156e-05,
+      "loss": 1.5774,
       "step": 9900
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 1.546914623837701e-05,
-      "loss": 1.8159,
+      "epoch": 4.22,
+      "learning_rate": 1.550485847063794e-05,
+      "loss": 1.5616,
       "step": 10000
     },
     {
       "epoch": 4.27,
-      "learning_rate": 1.462383770076078e-05,
-      "loss": 1.7997,
+      "learning_rate": 1.4659907055344319e-05,
+      "loss": 1.5564,
       "step": 10100
     },
     {
       "epoch": 4.31,
-      "learning_rate": 1.3778529163144549e-05,
-      "loss": 1.8197,
+      "learning_rate": 1.3814955640050698e-05,
+      "loss": 1.569,
       "step": 10200
     },
     {
       "epoch": 4.35,
-      "learning_rate": 1.2933220625528319e-05,
-      "loss": 1.7971,
+      "learning_rate": 1.2970004224757076e-05,
+      "loss": 1.5507,
       "step": 10300
     },
     {
       "epoch": 4.39,
-      "learning_rate": 1.2087912087912089e-05,
-      "loss": 1.813,
+      "learning_rate": 1.2125052809463457e-05,
+      "loss": 1.569,
       "step": 10400
     },
     {
       "epoch": 4.44,
-      "learning_rate": 1.1242603550295859e-05,
-      "loss": 1.8104,
+      "learning_rate": 1.1280101394169835e-05,
+      "loss": 1.5654,
       "step": 10500
     },
     {
       "epoch": 4.48,
-      "learning_rate": 1.0397295012679628e-05,
-      "loss": 1.8035,
+      "learning_rate": 1.0435149978876214e-05,
+      "loss": 1.5638,
       "step": 10600
     },
     {
       "epoch": 4.52,
-      "learning_rate": 9.551986475063398e-06,
-      "loss": 1.8309,
+      "learning_rate": 9.590198563582595e-06,
+      "loss": 1.6049,
       "step": 10700
     },
     {
       "epoch": 4.56,
-      "learning_rate": 8.70667793744717e-06,
-      "loss": 1.8331,
+      "learning_rate": 8.745247148288973e-06,
+      "loss": 1.5806,
       "step": 10800
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 7.861369399830938e-06,
-      "loss": 1.8107,
+      "epoch": 4.6,
+      "learning_rate": 7.900295732995352e-06,
+      "loss": 1.5567,
       "step": 10900
     },
     {
       "epoch": 4.65,
-      "learning_rate": 7.016060862214709e-06,
-      "loss": 1.803,
+      "learning_rate": 7.055344317701733e-06,
+      "loss": 1.5621,
       "step": 11000
     },
     {
       "epoch": 4.69,
-      "learning_rate": 6.1707523245984785e-06,
-      "loss": 1.8064,
+      "learning_rate": 6.210392902408112e-06,
+      "loss": 1.558,
       "step": 11100
     },
     {
       "epoch": 4.73,
-      "learning_rate": 5.325443786982249e-06,
-      "loss": 1.8,
+      "learning_rate": 5.3654414871144915e-06,
+      "loss": 1.5438,
       "step": 11200
     },
     {
       "epoch": 4.77,
-      "learning_rate": 4.480135249366019e-06,
-      "loss": 1.8006,
+      "learning_rate": 4.52049007182087e-06,
+      "loss": 1.5612,
       "step": 11300
     },
     {
       "epoch": 4.82,
-      "learning_rate": 3.6348267117497893e-06,
-      "loss": 1.8271,
+      "learning_rate": 3.67553865652725e-06,
+      "loss": 1.5917,
       "step": 11400
     },
     {
       "epoch": 4.86,
-      "learning_rate": 2.789518174133559e-06,
-      "loss": 1.8195,
+      "learning_rate": 2.830587241233629e-06,
+      "loss": 1.5709,
       "step": 11500
     },
     {
       "epoch": 4.9,
-      "learning_rate": 1.944209636517329e-06,
-      "loss": 1.8006,
+      "learning_rate": 1.9856358259400087e-06,
+      "loss": 1.5633,
       "step": 11600
     },
     {
       "epoch": 4.94,
-      "learning_rate": 1.098901098901099e-06,
-      "loss": 1.7734,
+      "learning_rate": 1.140684410646388e-06,
+      "loss": 1.5283,
       "step": 11700
     },
     {
       "epoch": 4.99,
-      "learning_rate": 2.53592561284869e-07,
-      "loss": 1.7928,
+      "learning_rate": 2.957329953527672e-07,
+      "loss": 1.5483,
       "step": 11800
     },
     {
       "epoch": 5.0,
-      "eval_loss": 1.9991737604141235,
-      "eval_runtime": 302.9112,
-      "eval_samples_per_second": 35.149,
-      "eval_steps_per_second": 8.788,
-      "step": 11830
+      "eval_loss": 1.7672892808914185,
+      "eval_runtime": 41.2842,
+      "eval_samples_per_second": 257.943,
+      "eval_steps_per_second": 4.045,
+      "step": 11835
     },
     {
       "epoch": 5.0,
-      "step": 11830,
-      "total_flos": 2.3051248406102016e+17,
-      "train_loss": 1.9853319837194525,
-      "train_runtime": 42362.5347,
-      "train_samples_per_second": 8.937,
-      "train_steps_per_second": 0.279
+      "step": 11835,
+      "total_flos": 2.305575469449216e+17,
+      "train_loss": 1.7503460931234034,
+      "train_runtime": 5808.8161,
+      "train_samples_per_second": 65.179,
+      "train_steps_per_second": 2.037
     }
   ],
   "logging_steps": 100,
-  "max_steps": 11830,
+  "max_steps": 11835,
   "num_train_epochs": 5,
   "save_steps": 500,
-  "total_flos": 2.3051248406102016e+17,
+  "total_flos": 2.305575469449216e+17,
   "trial_name": null,
   "trial_params": null
 }