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": 57.46,
-    "f1": 67.25301950068163,
-    "train_loss": 0.4283627078502993,
-    "train_runtime": 374.607,
+    "exact_match": 51.6,
+    "f1": 61.846019363732836,
+    "train_loss": 0.18913862856128547,
+    "train_runtime": 929.4199,
     "train_samples": 5000,
-    "train_samples_per_second": 266.946,
-    "train_steps_per_second": 4.218
+    "train_samples_per_second": 268.985,
+    "train_steps_per_second": 4.25
 }

args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9f6b99395f43a9e5b5a5d3cbdadcbd5619fd461b20901a8c96e1ea2fc5dcf909
+oid sha256:02e3643bf0ddcfcc56ccb0d834f878ffd780b4a19cfdb7d204c5809289919cf7
 size 5652

eval_nbest_predictions.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:03ef0902b3539289f5b4d12a9439fb8be293e732da3dd5b1f237209f7646c3f2
-size 23108849
+oid sha256:3baedba117a9eb41970565271a68aa2055d53a2128afa3e912dc18e825bb768e
+size 23002685

eval_results.json

@@ -1,6 +1,6 @@
 {
-    "epoch": 20.0,
+    "epoch": 50.0,
     "eval_samples": 5000,
-    "exact_match": 57.46,
-    "f1": 67.25301950068163
+    "exact_match": 51.6,
+    "f1": 61.846019363732836
 }

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 20.0,
-    "train_loss": 0.4283627078502993,
-    "train_runtime": 374.607,
+    "epoch": 50.0,
+    "train_loss": 0.18913862856128547,
+    "train_runtime": 929.4199,
     "train_samples": 5000,
-    "train_samples_per_second": 266.946,
-    "train_steps_per_second": 4.218
+    "train_samples_per_second": 268.985,
+    "train_steps_per_second": 4.25
 }

trainer_state.json

@@ -1,46 +1,70 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 20.0,
+  "epoch": 50.0,
   "eval_steps": 500,
-  "global_step": 1580,
+  "global_step": 3950,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 6.33,
-      "learning_rate": 3.4177215189873416e-05,
-      "loss": 1.1909,
+      "learning_rate": 4.367088607594937e-05,
+      "loss": 1.1777,
       "step": 500
     },
     {
       "epoch": 12.66,
-      "learning_rate": 1.8354430379746836e-05,
-      "loss": 0.1219,
+      "learning_rate": 3.7341772151898736e-05,
+      "loss": 0.1233,
       "step": 1000
     },
     {
       "epoch": 18.99,
-      "learning_rate": 2.531645569620253e-06,
-      "loss": 0.0375,
+      "learning_rate": 3.10126582278481e-05,
+      "loss": 0.05,
       "step": 1500
     },
     {
-      "epoch": 20.0,
-      "step": 1580,
-      "total_flos": 9798982502400000.0,
-      "train_loss": 0.4283627078502993,
-      "train_runtime": 374.607,
-      "train_samples_per_second": 266.946,
-      "train_steps_per_second": 4.218
+      "epoch": 25.32,
+      "learning_rate": 2.468354430379747e-05,
+      "loss": 0.0223,
+      "step": 2000
+    },
+    {
+      "epoch": 31.65,
+      "learning_rate": 1.8354430379746836e-05,
+      "loss": 0.0162,
+      "step": 2500
+    },
+    {
+      "epoch": 37.97,
+      "learning_rate": 1.2025316455696203e-05,
+      "loss": 0.0166,
+      "step": 3000
+    },
+    {
+      "epoch": 44.3,
+      "learning_rate": 5.69620253164557e-06,
+      "loss": 0.0231,
+      "step": 3500
+    },
+    {
+      "epoch": 50.0,
+      "step": 3950,
+      "total_flos": 2.4497456256e+16,
+      "train_loss": 0.18913862856128547,
+      "train_runtime": 929.4199,
+      "train_samples_per_second": 268.985,
+      "train_steps_per_second": 4.25
     }
   ],
   "logging_steps": 500,
-  "max_steps": 1580,
-  "num_train_epochs": 20,
+  "max_steps": 3950,
+  "num_train_epochs": 50,
   "save_steps": 500,
-  "total_flos": 9798982502400000.0,
+  "total_flos": 2.4497456256e+16,
   "trial_name": null,
   "trial_params": null
 }