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": 5000,
-    "exact_match": 61.02,
-    "f1": 71.26989840017605,
-    "train_loss": 0.44140149354934693,
-    "train_runtime": 581.4009,
+    "exact_match": 63.06,
+    "f1": 72.7342051521597,
+    "train_loss": 0.18040521144866944,
+    "train_runtime": 1459.0895,
     "train_samples": 5000,
-    "train_samples_per_second": 171.998,
-    "train_steps_per_second": 1.376
+    "train_samples_per_second": 171.34,
+    "train_steps_per_second": 1.371
 }

args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1f8c002e096a8591d456198a5a330d8606026b8bb5dd58efda1841b1abd1fbe7
+oid sha256:60dad82c8d2c07d6638e17b1fbcbce92421798356ab3901e65095c21526f4a73
 size 5652

eval_nbest_predictions.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9e223679ac191b5c1a34eed55ba75147a86a8e87a69e4ab3d3b359af2825e5ef
-size 23219404
+oid sha256:3723ef2dd87dd70451ba04c4577b173f0082f70b64d2ef9f9e5c95918df25a8d
+size 23207852

eval_results.json

@@ -1,6 +1,6 @@
 {
-    "epoch": 20.0,
+    "epoch": 50.0,
     "eval_samples": 5000,
-    "exact_match": 61.02,
-    "f1": 71.26989840017605
+    "exact_match": 63.06,
+    "f1": 72.7342051521597
 }

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 20.0,
-    "train_loss": 0.44140149354934693,
-    "train_runtime": 581.4009,
+    "epoch": 50.0,
+    "train_loss": 0.18040521144866944,
+    "train_runtime": 1459.0895,
     "train_samples": 5000,
-    "train_samples_per_second": 171.998,
-    "train_steps_per_second": 1.376
+    "train_samples_per_second": 171.34,
+    "train_steps_per_second": 1.371
 }

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": 800,
+  "global_step": 2000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 12.5,
-      "learning_rate": 1.8750000000000002e-05,
-      "loss": 0.6782,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 0.6645,
       "step": 500
     },
     {
-      "epoch": 20.0,
-      "step": 800,
-      "total_flos": 1.95972567552e+16,
-      "train_loss": 0.44140149354934693,
-      "train_runtime": 581.4009,
-      "train_samples_per_second": 171.998,
-      "train_steps_per_second": 1.376
+      "epoch": 25.0,
+      "learning_rate": 2.5e-05,
+      "loss": 0.0397,
+      "step": 1000
+    },
+    {
+      "epoch": 37.5,
+      "learning_rate": 1.25e-05,
+      "loss": 0.0128,
+      "step": 1500
+    },
+    {
+      "epoch": 50.0,
+      "learning_rate": 0.0,
+      "loss": 0.0046,
+      "step": 2000
+    },
+    {
+      "epoch": 50.0,
+      "step": 2000,
+      "total_flos": 4.8993141888e+16,
+      "train_loss": 0.18040521144866944,
+      "train_runtime": 1459.0895,
+      "train_samples_per_second": 171.34,
+      "train_steps_per_second": 1.371
     }
   ],
   "logging_steps": 500,
-  "max_steps": 800,
-  "num_train_epochs": 20,
+  "max_steps": 2000,
+  "num_train_epochs": 50,
   "save_steps": 500,
-  "total_flos": 1.95972567552e+16,
+  "total_flos": 4.8993141888e+16,
   "trial_name": null,
   "trial_params": null
 }