End of training

README.md

@@ -2,7 +2,10 @@
 license: apache-2.0
 base_model: distilbert-base-uncased
 tags:
+- question-answering
 - generated_from_trainer
+datasets:
+- squad
 model-index:
 - name: out
   results: []
@@ -13,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # out
 
-This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on the None dataset.
+This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on the squad dataset.
 
 ## Model description
 

all_results.json

@@ -1,11 +1,11 @@
 {
-    "epoch": 20.0,
+    "epoch": 50.0,
     "eval_samples": 5000,
-    "exact_match": 57.7,
-    "f1": 67.32789188139478,
-    "train_loss": 0.3331835723606644,
-    "train_runtime": 459.4653,
+    "exact_match": 59.0,
+    "f1": 68.40149035187889,
+    "train_loss": 0.14310269137856307,
+    "train_runtime": 1143.4603,
     "train_samples": 5000,
-    "train_samples_per_second": 217.644,
-    "train_steps_per_second": 6.834
+    "train_samples_per_second": 218.635,
+    "train_steps_per_second": 6.865
 }

args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:19fcedcf799fe5c8a5554bae0d95655190c81d2a729e89a6ace2293869ce7b49
+oid sha256:5fbec49d09f7ae6605a5933a16596aedf0c32832eacf337a55178dfe0c615d82
 size 5588

eval_nbest_predictions.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:44c78f55fa51175ec7393b9b2e06c58fc1f4e109fdee310cf1d8f4b1f94dd962
-size 23011821
+oid sha256:1b96b2d9c27fcb5a811d5835f706170a58ac14efb3a26b03737cd5f0737caab1
+size 22976135

eval_results.json

@@ -1,6 +1,6 @@
 {
-    "epoch": 20.0,
+    "epoch": 50.0,
     "eval_samples": 5000,
-    "exact_match": 57.7,
-    "f1": 67.32789188139478
+    "exact_match": 59.0,
+    "f1": 68.40149035187889
 }

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 20.0,
-    "train_loss": 0.3331835723606644,
-    "train_runtime": 459.4653,
+    "epoch": 50.0,
+    "train_loss": 0.14310269137856307,
+    "train_runtime": 1143.4603,
     "train_samples": 5000,
-    "train_samples_per_second": 217.644,
-    "train_steps_per_second": 6.834
+    "train_samples_per_second": 218.635,
+    "train_steps_per_second": 6.865
 }

trainer_state.json

@@ -1,64 +1,118 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 20.0,
+  "epoch": 50.0,
   "eval_steps": 500,
-  "global_step": 3140,
+  "global_step": 7850,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 3.18,
-      "learning_rate": 4.2038216560509556e-05,
-      "loss": 1.5882,
+      "learning_rate": 4.681528662420383e-05,
+      "loss": 1.5828,
       "step": 500
     },
     {
       "epoch": 6.37,
-      "learning_rate": 3.407643312101911e-05,
-      "loss": 0.2947,
+      "learning_rate": 4.3630573248407646e-05,
+      "loss": 0.2954,
       "step": 1000
     },
     {
       "epoch": 9.55,
-      "learning_rate": 2.6114649681528662e-05,
-      "loss": 0.112,
+      "learning_rate": 4.044585987261147e-05,
+      "loss": 0.123,
       "step": 1500
     },
     {
       "epoch": 12.74,
-      "learning_rate": 1.8152866242038215e-05,
-      "loss": 0.0539,
+      "learning_rate": 3.7261146496815283e-05,
+      "loss": 0.0777,
       "step": 2000
     },
     {
       "epoch": 15.92,
-      "learning_rate": 1.0191082802547772e-05,
-      "loss": 0.0277,
+      "learning_rate": 3.407643312101911e-05,
+      "loss": 0.0501,
       "step": 2500
     },
     {
       "epoch": 19.11,
-      "learning_rate": 2.229299363057325e-06,
-      "loss": 0.0134,
+      "learning_rate": 3.089171974522293e-05,
+      "loss": 0.0339,
       "step": 3000
     },
     {
-      "epoch": 20.0,
-      "step": 3140,
-      "total_flos": 9798982502400000.0,
-      "train_loss": 0.3331835723606644,
-      "train_runtime": 459.4653,
-      "train_samples_per_second": 217.644,
-      "train_steps_per_second": 6.834
+      "epoch": 22.29,
+      "learning_rate": 2.7707006369426753e-05,
+      "loss": 0.0237,
+      "step": 3500
+    },
+    {
+      "epoch": 25.48,
+      "learning_rate": 2.4522292993630575e-05,
+      "loss": 0.0156,
+      "step": 4000
+    },
+    {
+      "epoch": 28.66,
+      "learning_rate": 2.1337579617834397e-05,
+      "loss": 0.0123,
+      "step": 4500
+    },
+    {
+      "epoch": 31.85,
+      "learning_rate": 1.8152866242038215e-05,
+      "loss": 0.0092,
+      "step": 5000
+    },
+    {
+      "epoch": 35.03,
+      "learning_rate": 1.4968152866242039e-05,
+      "loss": 0.0065,
+      "step": 5500
+    },
+    {
+      "epoch": 38.22,
+      "learning_rate": 1.178343949044586e-05,
+      "loss": 0.0057,
+      "step": 6000
+    },
+    {
+      "epoch": 41.4,
+      "learning_rate": 8.598726114649681e-06,
+      "loss": 0.0035,
+      "step": 6500
+    },
+    {
+      "epoch": 44.59,
+      "learning_rate": 5.414012738853504e-06,
+      "loss": 0.0034,
+      "step": 7000
+    },
+    {
+      "epoch": 47.77,
+      "learning_rate": 2.229299363057325e-06,
+      "loss": 0.0024,
+      "step": 7500
+    },
+    {
+      "epoch": 50.0,
+      "step": 7850,
+      "total_flos": 2.4497456256e+16,
+      "train_loss": 0.14310269137856307,
+      "train_runtime": 1143.4603,
+      "train_samples_per_second": 218.635,
+      "train_steps_per_second": 6.865
     }
   ],
   "logging_steps": 500,
-  "max_steps": 3140,
-  "num_train_epochs": 20,
+  "max_steps": 7850,
+  "num_train_epochs": 50,
   "save_steps": 500,
-  "total_flos": 9798982502400000.0,
+  "total_flos": 2.4497456256e+16,
   "trial_name": null,
   "trial_params": null
 }