End of training

README.md

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

all_results.json

@@ -1,11 +1,11 @@
 {
-    "epoch": 100.0,
+    "epoch": 3.0,
     "eval_samples": 5000,
-    "exact_match": 66.98,
-    "f1": 76.33762054333884,
-    "train_loss": 0.07316695046424866,
-    "train_runtime": 7948.9124,
+    "exact_match": 67.16,
+    "f1": 77.19142267647392,
+    "train_loss": 1.221010103507384,
+    "train_runtime": 273.6712,
     "train_samples": 5000,
-    "train_samples_per_second": 62.902,
-    "train_steps_per_second": 0.503
+    "train_samples_per_second": 54.81,
+    "train_steps_per_second": 0.866
 }

args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:81a0e3c1b80c0be8679cb6ff235f119f3d6c4a3afb2b8538270ccc21329801fa
+oid sha256:81abfdb4166657275d43a30a6ef5d6b9bfe978b01efed2b936284b13e0b6fe6c
 size 5652

eval_nbest_predictions.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:822e8ed5e3d24a307d5822dd45cbc631f29e01305d4f75597833c1846317829f
-size 22769543
+oid sha256:6a7e9f14080866bd53b8e7320820d485ed7af9943b472e1b734eb0f9a4a19f58
+size 22606999

eval_results.json

@@ -1,6 +1,6 @@
 {
-    "epoch": 100.0,
+    "epoch": 3.0,
     "eval_samples": 5000,
-    "exact_match": 66.98,
-    "f1": 76.33762054333884
+    "exact_match": 67.16,
+    "f1": 77.19142267647392
 }

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 100.0,
-    "train_loss": 0.07316695046424866,
-    "train_runtime": 7948.9124,
+    "epoch": 3.0,
+    "train_loss": 1.221010103507384,
+    "train_runtime": 273.6712,
     "train_samples": 5000,
-    "train_samples_per_second": 62.902,
-    "train_steps_per_second": 0.503
+    "train_samples_per_second": 54.81,
+    "train_steps_per_second": 0.866
 }

trainer_state.json

@@ -1,76 +1,28 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 100.0,
+  "epoch": 3.0,
   "eval_steps": 500,
-  "global_step": 4000,
+  "global_step": 237,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 12.5,
-      "learning_rate": 4.375e-05,
-      "loss": 0.486,
-      "step": 500
-    },
-    {
-      "epoch": 25.0,
-      "learning_rate": 3.7500000000000003e-05,
-      "loss": 0.0396,
-      "step": 1000
-    },
-    {
-      "epoch": 37.5,
-      "learning_rate": 3.125e-05,
-      "loss": 0.0178,
-      "step": 1500
-    },
-    {
-      "epoch": 50.0,
-      "learning_rate": 2.5e-05,
-      "loss": 0.0089,
-      "step": 2000
-    },
-    {
-      "epoch": 62.5,
-      "learning_rate": 1.8750000000000002e-05,
-      "loss": 0.0061,
-      "step": 2500
-    },
-    {
-      "epoch": 75.0,
-      "learning_rate": 1.25e-05,
-      "loss": 0.0062,
-      "step": 3000
-    },
-    {
-      "epoch": 87.5,
-      "learning_rate": 6.25e-06,
-      "loss": 0.0072,
-      "step": 3500
-    },
-    {
-      "epoch": 100.0,
-      "learning_rate": 0.0,
-      "loss": 0.0134,
-      "step": 4000
-    },
-    {
-      "epoch": 100.0,
-      "step": 4000,
-      "total_flos": 3.48265122048e+17,
-      "train_loss": 0.07316695046424866,
-      "train_runtime": 7948.9124,
-      "train_samples_per_second": 62.902,
-      "train_steps_per_second": 0.503
+      "epoch": 3.0,
+      "step": 237,
+      "total_flos": 1.044795366144e+16,
+      "train_loss": 1.221010103507384,
+      "train_runtime": 273.6712,
+      "train_samples_per_second": 54.81,
+      "train_steps_per_second": 0.866
     }
   ],
   "logging_steps": 500,
-  "max_steps": 4000,
-  "num_train_epochs": 100,
+  "max_steps": 237,
+  "num_train_epochs": 3,
   "save_steps": 500,
-  "total_flos": 3.48265122048e+17,
+  "total_flos": 1.044795366144e+16,
   "trial_name": null,
   "trial_params": null
 }