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": 3000,
-    "exact_match": 54.6,
-    "f1": 65.85098994258614,
-    "train_loss": 0.14720325855498617,
-    "train_runtime": 3350.47,
+    "exact_match": 36.53333333333333,
+    "f1": 48.441986654053316,
+    "train_loss": 0.4032590159964054,
+    "train_runtime": 6707.7517,
     "train_samples": 3000,
-    "train_samples_per_second": 44.77,
-    "train_steps_per_second": 1.403
+    "train_samples_per_second": 44.724,
+    "train_steps_per_second": 1.401
 }

args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:96545b7a49e8eb320ee33aaf2ba3c13d7914560f92f71f82fdbde9d862ec4f68
+oid sha256:e4400e0723d955f89d2c29d11637b0e906726ed7e53cff9034c6d35f718150e0
 size 5652

eval_nbest_predictions.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0a5b523eb44e466dea0a3a13b9a71c7ea484ca84c4b558d41649222659b02bc0
-size 13990661
+oid sha256:8e9c9c0c43a722680701799a80e2c8b9bcb73e155f99d4c95e4fbf112c6f48a2
+size 13915047

eval_results.json

@@ -1,6 +1,6 @@
 {
-    "epoch": 50.0,
+    "epoch": 100.0,
     "eval_samples": 3000,
-    "exact_match": 54.6,
-    "f1": 65.85098994258614
+    "exact_match": 36.53333333333333,
+    "f1": 48.441986654053316
 }

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 50.0,
-    "train_loss": 0.14720325855498617,
-    "train_runtime": 3350.47,
+    "epoch": 100.0,
+    "train_loss": 0.4032590159964054,
+    "train_runtime": 6707.7517,
     "train_samples": 3000,
-    "train_samples_per_second": 44.77,
-    "train_steps_per_second": 1.403
+    "train_samples_per_second": 44.724,
+    "train_steps_per_second": 1.401
 }

trainer_state.json

@@ -1,82 +1,136 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 50.0,
+  "epoch": 100.0,
   "eval_steps": 500,
-  "global_step": 4700,
+  "global_step": 9400,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 5.32,
-      "learning_rate": 4.468085106382979e-05,
-      "loss": 0.867,
+      "learning_rate": 4.734042553191489e-05,
+      "loss": 0.9695,
       "step": 500
     },
     {
       "epoch": 10.64,
-      "learning_rate": 3.936170212765958e-05,
-      "loss": 0.1247,
+      "learning_rate": 4.468085106382979e-05,
+      "loss": 0.145,
       "step": 1000
     },
     {
       "epoch": 15.96,
-      "learning_rate": 3.4042553191489365e-05,
-      "loss": 0.069,
+      "learning_rate": 4.2021276595744684e-05,
+      "loss": 0.083,
       "step": 1500
     },
     {
       "epoch": 21.28,
-      "learning_rate": 2.8723404255319154e-05,
-      "loss": 0.0382,
+      "learning_rate": 3.936170212765958e-05,
+      "loss": 0.0518,
       "step": 2000
     },
     {
       "epoch": 26.6,
-      "learning_rate": 2.340425531914894e-05,
-      "loss": 0.0207,
+      "learning_rate": 3.670212765957447e-05,
+      "loss": 0.0495,
       "step": 2500
     },
     {
       "epoch": 31.91,
-      "learning_rate": 1.8085106382978724e-05,
-      "loss": 0.0181,
+      "learning_rate": 3.4042553191489365e-05,
+      "loss": 0.042,
       "step": 3000
     },
     {
       "epoch": 37.23,
-      "learning_rate": 1.2765957446808511e-05,
-      "loss": 0.0225,
+      "learning_rate": 3.1382978723404254e-05,
+      "loss": 0.0514,
       "step": 3500
     },
     {
       "epoch": 42.55,
-      "learning_rate": 7.446808510638298e-06,
-      "loss": 0.029,
+      "learning_rate": 2.8723404255319154e-05,
+      "loss": 0.053,
       "step": 4000
     },
     {
       "epoch": 47.87,
-      "learning_rate": 2.1276595744680853e-06,
-      "loss": 0.0873,
+      "learning_rate": 2.6063829787234046e-05,
+      "loss": 0.074,
       "step": 4500
     },
     {
-      "epoch": 50.0,
-      "step": 4700,
-      "total_flos": 1.044795366144e+17,
-      "train_loss": 0.14720325855498617,
-      "train_runtime": 3350.47,
-      "train_samples_per_second": 44.77,
-      "train_steps_per_second": 1.403
+      "epoch": 53.19,
+      "learning_rate": 2.340425531914894e-05,
+      "loss": 0.1369,
+      "step": 5000
+    },
+    {
+      "epoch": 58.51,
+      "learning_rate": 2.074468085106383e-05,
+      "loss": 0.2306,
+      "step": 5500
+    },
+    {
+      "epoch": 63.83,
+      "learning_rate": 1.8085106382978724e-05,
+      "loss": 0.49,
+      "step": 6000
+    },
+    {
+      "epoch": 69.15,
+      "learning_rate": 1.5425531914893617e-05,
+      "loss": 0.5549,
+      "step": 6500
+    },
+    {
+      "epoch": 74.47,
+      "learning_rate": 1.2765957446808511e-05,
+      "loss": 0.715,
+      "step": 7000
+    },
+    {
+      "epoch": 79.79,
+      "learning_rate": 1.0106382978723404e-05,
+      "loss": 0.9123,
+      "step": 7500
+    },
+    {
+      "epoch": 85.11,
+      "learning_rate": 7.446808510638298e-06,
+      "loss": 0.9641,
+      "step": 8000
+    },
+    {
+      "epoch": 90.43,
+      "learning_rate": 4.787234042553191e-06,
+      "loss": 1.027,
+      "step": 8500
+    },
+    {
+      "epoch": 95.74,
+      "learning_rate": 2.1276595744680853e-06,
+      "loss": 0.7927,
+      "step": 9000
+    },
+    {
+      "epoch": 100.0,
+      "step": 9400,
+      "total_flos": 2.089590732288e+17,
+      "train_loss": 0.4032590159964054,
+      "train_runtime": 6707.7517,
+      "train_samples_per_second": 44.724,
+      "train_steps_per_second": 1.401
     }
   ],
   "logging_steps": 500,
-  "max_steps": 4700,
-  "num_train_epochs": 50,
+  "max_steps": 9400,
+  "num_train_epochs": 100,
   "save_steps": 500,
-  "total_flos": 1.044795366144e+17,
+  "total_flos": 2.089590732288e+17,
   "trial_name": null,
   "trial_params": null
 }