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": 50.0,
+    "epoch": 100.0,
     "eval_samples": 5000,
-    "exact_match": 70.36,
-    "f1": 79.57766677993115,
-    "train_loss": 0.14002820599079133,
-    "train_runtime": 4002.6207,
+    "exact_match": 66.98,
+    "f1": 76.33762054333884,
+    "train_loss": 0.07316695046424866,
+    "train_runtime": 7948.9124,
     "train_samples": 5000,
-    "train_samples_per_second": 62.459,
-    "train_steps_per_second": 0.5
+    "train_samples_per_second": 62.902,
+    "train_steps_per_second": 0.503
 }

args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:99c448c354de9a1558fb0393212820519d849da4c3528717dd8750c0f0786573
+oid sha256:81a0e3c1b80c0be8679cb6ff235f119f3d6c4a3afb2b8538270ccc21329801fa
 size 5652

eval_nbest_predictions.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ca6c8c3ed52b8bc5fc6d192ace9262d4cca4dae0d4e3f73860136679f52a42f2
-size 22789716
+oid sha256:822e8ed5e3d24a307d5822dd45cbc631f29e01305d4f75597833c1846317829f
+size 22769543

eval_results.json

@@ -1,6 +1,6 @@
 {
-    "epoch": 50.0,
+    "epoch": 100.0,
     "eval_samples": 5000,
-    "exact_match": 70.36,
-    "f1": 79.57766677993115
+    "exact_match": 66.98,
+    "f1": 76.33762054333884
 }

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 50.0,
-    "train_loss": 0.14002820599079133,
-    "train_runtime": 4002.6207,
+    "epoch": 100.0,
+    "train_loss": 0.07316695046424866,
+    "train_runtime": 7948.9124,
     "train_samples": 5000,
-    "train_samples_per_second": 62.459,
-    "train_steps_per_second": 0.5
+    "train_samples_per_second": 62.902,
+    "train_steps_per_second": 0.503
 }

trainer_state.json

@@ -1,52 +1,76 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 50.0,
+  "epoch": 100.0,
   "eval_steps": 500,
-  "global_step": 2000,
+  "global_step": 4000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 12.5,
-      "learning_rate": 3.7500000000000003e-05,
-      "loss": 0.5158,
+      "learning_rate": 4.375e-05,
+      "loss": 0.486,
       "step": 500
     },
     {
       "epoch": 25.0,
-      "learning_rate": 2.5e-05,
-      "loss": 0.0329,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 0.0396,
       "step": 1000
     },
     {
       "epoch": 37.5,
-      "learning_rate": 1.25e-05,
-      "loss": 0.0087,
+      "learning_rate": 3.125e-05,
+      "loss": 0.0178,
       "step": 1500
     },
     {
       "epoch": 50.0,
-      "learning_rate": 0.0,
-      "loss": 0.0028,
+      "learning_rate": 2.5e-05,
+      "loss": 0.0089,
       "step": 2000
     },
     {
-      "epoch": 50.0,
-      "step": 2000,
-      "total_flos": 1.74132561024e+17,
-      "train_loss": 0.14002820599079133,
-      "train_runtime": 4002.6207,
-      "train_samples_per_second": 62.459,
-      "train_steps_per_second": 0.5
+      "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
     }
   ],
   "logging_steps": 500,
-  "max_steps": 2000,
-  "num_train_epochs": 50,
+  "max_steps": 4000,
+  "num_train_epochs": 100,
   "save_steps": 500,
-  "total_flos": 1.74132561024e+17,
+  "total_flos": 3.48265122048e+17,
   "trial_name": null,
   "trial_params": null
 }