End of training

README.md

@@ -2,7 +2,10 @@
 license: apache-2.0
 base_model: distilbert-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 [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on the None dataset.
+This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-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": 54.13333333333333,
-    "f1": 63.901880489999314,
-    "train_loss": 0.6003903328104222,
-    "train_runtime": 231.9146,
+    "exact_match": 54.766666666666666,
+    "f1": 64.53016416256601,
+    "train_loss": 0.24304250493962715,
+    "train_runtime": 576.78,
     "train_samples": 3000,
-    "train_samples_per_second": 258.716,
-    "train_steps_per_second": 4.053
+    "train_samples_per_second": 260.065,
+    "train_steps_per_second": 4.074
 }

args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4810ed4eec1e4906c5b606e39d54a5381214a493049ba6b209c390be9b8ecdaf
+oid sha256:a222382bc3c5c55b27c6968574e5efec43b0d436b9660d08670cf2515dfa2323
 size 5652

eval_nbest_predictions.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:78cac92342ee14ea7282a17a5a815c667aaed4928f18779734bc9f209104e377
-size 14113500
+oid sha256:198cdcbb97d9181d7a77982d35961a64e1020a04c63338759614a03cf31ca153
+size 14142746

eval_results.json

@@ -1,6 +1,6 @@
 {
-    "epoch": 20.0,
+    "epoch": 50.0,
     "eval_samples": 3000,
-    "exact_match": 54.13333333333333,
-    "f1": 63.901880489999314
+    "exact_match": 54.766666666666666,
+    "f1": 64.53016416256601
 }

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 20.0,
-    "train_loss": 0.6003903328104222,
-    "train_runtime": 231.9146,
+    "epoch": 50.0,
+    "train_loss": 0.24304250493962715,
+    "train_runtime": 576.78,
     "train_samples": 3000,
-    "train_samples_per_second": 258.716,
-    "train_steps_per_second": 4.053
+    "train_samples_per_second": 260.065,
+    "train_steps_per_second": 4.074
 }

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": 1.0721,
+      "learning_rate": 3.936170212765958e-05,
+      "loss": 1.0484,
       "step": 500
     },
     {
-      "epoch": 20.0,
-      "step": 940,
-      "total_flos": 5879389501440000.0,
-      "train_loss": 0.6003903328104222,
-      "train_runtime": 231.9146,
-      "train_samples_per_second": 258.716,
-      "train_steps_per_second": 4.053
+      "epoch": 21.28,
+      "learning_rate": 2.8723404255319154e-05,
+      "loss": 0.0622,
+      "step": 1000
+    },
+    {
+      "epoch": 31.91,
+      "learning_rate": 1.8085106382978724e-05,
+      "loss": 0.0191,
+      "step": 1500
+    },
+    {
+      "epoch": 42.55,
+      "learning_rate": 7.446808510638298e-06,
+      "loss": 0.0081,
+      "step": 2000
+    },
+    {
+      "epoch": 50.0,
+      "step": 2350,
+      "total_flos": 1.46984737536e+16,
+      "train_loss": 0.24304250493962715,
+      "train_runtime": 576.78,
+      "train_samples_per_second": 260.065,
+      "train_steps_per_second": 4.074
     }
   ],
   "logging_steps": 500,
-  "max_steps": 940,
-  "num_train_epochs": 20,
+  "max_steps": 2350,
+  "num_train_epochs": 50,
   "save_steps": 500,
-  "total_flos": 5879389501440000.0,
+  "total_flos": 1.46984737536e+16,
   "trial_name": null,
   "trial_params": null
 }