End of training

README.md

@@ -2,7 +2,10 @@
 license: apache-2.0
 base_model: bert-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 [bert-base-uncased](https://huggingface.co/bert-base-uncased) on the None dataset.
+This model is a fine-tuned version of [bert-base-uncased](https://huggingface.co/bert-base-uncased) on the squad dataset.
 
 ## Model description
 

all_results.json

@@ -1,11 +1,11 @@
 {
-    "epoch": 20.0,
+    "epoch": 50.0,
     "eval_samples": 3000,
-    "exact_match": 59.46666666666667,
-    "f1": 69.99235737832946,
-    "train_loss": 0.43026711484219166,
-    "train_runtime": 419.3657,
+    "exact_match": 60.96666666666667,
+    "f1": 70.61923830210029,
+    "train_loss": 0.17848453323891822,
+    "train_runtime": 1047.9186,
     "train_samples": 3000,
-    "train_samples_per_second": 143.073,
-    "train_steps_per_second": 2.241
+    "train_samples_per_second": 143.141,
+    "train_steps_per_second": 2.243
 }

args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b69e46f0b29f6792a3e09245768bb966a7d39a9c3223709d531c2b11ce820b2b
+oid sha256:a7f5d69c6cb40eb2428d4a8bbb673af6fadf7acae67adb52ab480cb40541d726
 size 5652

eval_nbest_predictions.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6f7965b8b995130479949c3f5285393110c0ce449e1d799bf455e262358946ed
-size 14238769
+oid sha256:84a9ac3a863a4e1d03a79cbadcf619d10c3608967ddee136f7285cdd69f4124e
+size 14167890

eval_results.json

@@ -1,6 +1,6 @@
 {
-    "epoch": 20.0,
+    "epoch": 50.0,
     "eval_samples": 3000,
-    "exact_match": 59.46666666666667,
-    "f1": 69.99235737832946
+    "exact_match": 60.96666666666667,
+    "f1": 70.61923830210029
 }

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 20.0,
-    "train_loss": 0.43026711484219166,
-    "train_runtime": 419.3657,
+    "epoch": 50.0,
+    "train_loss": 0.17848453323891822,
+    "train_runtime": 1047.9186,
     "train_samples": 3000,
-    "train_samples_per_second": 143.073,
-    "train_steps_per_second": 2.241
+    "train_samples_per_second": 143.141,
+    "train_steps_per_second": 2.243
 }

trainer_state.json

@@ -1,34 +1,52 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 20.0,
+  "epoch": 50.0,
   "eval_steps": 500,
-  "global_step": 940,
+  "global_step": 2350,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 10.64,
-      "learning_rate": 2.340425531914894e-05,
-      "loss": 0.7735,
+      "learning_rate": 3.936170212765958e-05,
+      "loss": 0.7712,
       "step": 500
     },
     {
-      "epoch": 20.0,
-      "step": 940,
-      "total_flos": 1.175835405312e+16,
-      "train_loss": 0.43026711484219166,
-      "train_runtime": 419.3657,
-      "train_samples_per_second": 143.073,
-      "train_steps_per_second": 2.241
+      "epoch": 21.28,
+      "learning_rate": 2.8723404255319154e-05,
+      "loss": 0.0462,
+      "step": 1000
+    },
+    {
+      "epoch": 31.91,
+      "learning_rate": 1.8085106382978724e-05,
+      "loss": 0.0141,
+      "step": 1500
+    },
+    {
+      "epoch": 42.55,
+      "learning_rate": 7.446808510638298e-06,
+      "loss": 0.005,
+      "step": 2000
+    },
+    {
+      "epoch": 50.0,
+      "step": 2350,
+      "total_flos": 2.93958851328e+16,
+      "train_loss": 0.17848453323891822,
+      "train_runtime": 1047.9186,
+      "train_samples_per_second": 143.141,
+      "train_steps_per_second": 2.243
     }
   ],
   "logging_steps": 500,
-  "max_steps": 940,
-  "num_train_epochs": 20,
+  "max_steps": 2350,
+  "num_train_epochs": 50,
   "save_steps": 500,
-  "total_flos": 1.175835405312e+16,
+  "total_flos": 2.93958851328e+16,
   "trial_name": null,
   "trial_params": null
 }