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": 5000,
-    "exact_match": 54.98,
-    "f1": 64.8393012609491,
-    "train_loss": 0.33844285147964576,
-    "train_runtime": 834.8611,
+    "exact_match": 26.0,
+    "f1": 38.36451648734593,
+    "train_loss": 0.5251886637499378,
+    "train_runtime": 2066.072,
     "train_samples": 5000,
-    "train_samples_per_second": 119.78,
-    "train_steps_per_second": 3.761
+    "train_samples_per_second": 121.003,
+    "train_steps_per_second": 3.799
 }

args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f946d98a9617f980361984cacbf785267525745c05202665c5c91ae7eabc6eab
+oid sha256:58c38a37d0f1de9fe661df9aa2a6c0b21ced20c6d923d97080b6f144501daa44
 size 5652

eval_nbest_predictions.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b2aeb60ee75a6e8451f3bb8c3aad80140b4c60baf9690d122cf17a6e37000fa8
-size 23031302
+oid sha256:5a4ca005506c01dfbfd7147903614c6b992f9189f7dc07f7befbae177fb7bd41
+size 23339407

eval_results.json

@@ -1,6 +1,6 @@
 {
-    "epoch": 20.0,
+    "epoch": 50.0,
     "eval_samples": 5000,
-    "exact_match": 54.98,
-    "f1": 64.8393012609491
+    "exact_match": 26.0,
+    "f1": 38.36451648734593
 }

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 20.0,
-    "train_loss": 0.33844285147964576,
-    "train_runtime": 834.8611,
+    "epoch": 50.0,
+    "train_loss": 0.5251886637499378,
+    "train_runtime": 2066.072,
     "train_samples": 5000,
-    "train_samples_per_second": 119.78,
-    "train_steps_per_second": 3.761
+    "train_samples_per_second": 121.003,
+    "train_steps_per_second": 3.799
 }

trainer_state.json

@@ -1,64 +1,118 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 20.0,
+  "epoch": 50.0,
   "eval_steps": 500,
-  "global_step": 3140,
+  "global_step": 7850,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 3.18,
-      "learning_rate": 4.2038216560509556e-05,
-      "loss": 1.5882,
+      "learning_rate": 4.681528662420383e-05,
+      "loss": 1.5828,
       "step": 500
     },
     {
       "epoch": 6.37,
-      "learning_rate": 3.407643312101911e-05,
-      "loss": 0.2947,
+      "learning_rate": 4.3630573248407646e-05,
+      "loss": 0.2954,
       "step": 1000
     },
     {
       "epoch": 9.55,
-      "learning_rate": 2.6114649681528662e-05,
-      "loss": 0.112,
+      "learning_rate": 4.044585987261147e-05,
+      "loss": 0.123,
       "step": 1500
     },
     {
       "epoch": 12.74,
-      "learning_rate": 1.8152866242038215e-05,
-      "loss": 0.0544,
+      "learning_rate": 3.7261146496815283e-05,
+      "loss": 0.078,
       "step": 2000
     },
     {
       "epoch": 15.92,
-      "learning_rate": 1.0191082802547772e-05,
-      "loss": 0.033,
+      "learning_rate": 3.407643312101911e-05,
+      "loss": 0.0562,
       "step": 2500
     },
     {
       "epoch": 19.11,
-      "learning_rate": 2.229299363057325e-06,
-      "loss": 0.0291,
+      "learning_rate": 3.089171974522293e-05,
+      "loss": 0.0447,
       "step": 3000
     },
     {
-      "epoch": 20.0,
-      "step": 3140,
-      "total_flos": 9798982502400000.0,
-      "train_loss": 0.33844285147964576,
-      "train_runtime": 834.8611,
-      "train_samples_per_second": 119.78,
-      "train_steps_per_second": 3.761
+      "epoch": 22.29,
+      "learning_rate": 2.7707006369426753e-05,
+      "loss": 0.053,
+      "step": 3500
+    },
+    {
+      "epoch": 25.48,
+      "learning_rate": 2.4522292993630575e-05,
+      "loss": 0.0765,
+      "step": 4000
+    },
+    {
+      "epoch": 28.66,
+      "learning_rate": 2.1337579617834397e-05,
+      "loss": 0.1098,
+      "step": 4500
+    },
+    {
+      "epoch": 31.85,
+      "learning_rate": 1.8152866242038215e-05,
+      "loss": 0.2508,
+      "step": 5000
+    },
+    {
+      "epoch": 35.03,
+      "learning_rate": 1.4968152866242039e-05,
+      "loss": 0.516,
+      "step": 5500
+    },
+    {
+      "epoch": 38.22,
+      "learning_rate": 1.178343949044586e-05,
+      "loss": 0.8691,
+      "step": 6000
+    },
+    {
+      "epoch": 41.4,
+      "learning_rate": 8.598726114649681e-06,
+      "loss": 0.9585,
+      "step": 6500
+    },
+    {
+      "epoch": 44.59,
+      "learning_rate": 5.414012738853504e-06,
+      "loss": 1.1375,
+      "step": 7000
+    },
+    {
+      "epoch": 47.77,
+      "learning_rate": 2.229299363057325e-06,
+      "loss": 1.2299,
+      "step": 7500
+    },
+    {
+      "epoch": 50.0,
+      "step": 7850,
+      "total_flos": 2.4497456256e+16,
+      "train_loss": 0.5251886637499378,
+      "train_runtime": 2066.072,
+      "train_samples_per_second": 121.003,
+      "train_steps_per_second": 3.799
     }
   ],
   "logging_steps": 500,
-  "max_steps": 3140,
-  "num_train_epochs": 20,
+  "max_steps": 7850,
+  "num_train_epochs": 50,
   "save_steps": 500,
-  "total_flos": 9798982502400000.0,
+  "total_flos": 2.4497456256e+16,
   "trial_name": null,
   "trial_params": null
 }