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": 55.64,
-    "f1": 65.9414733930188,
-    "train_loss": 0.6052761435508728,
-    "train_runtime": 530.4541,
+    "exact_match": 57.58,
+    "f1": 67.21309314821316,
+    "train_loss": 0.24194956755638122,
+    "train_runtime": 1318.7789,
     "train_samples": 5000,
-    "train_samples_per_second": 188.518,
-    "train_steps_per_second": 1.508
+    "train_samples_per_second": 189.569,
+    "train_steps_per_second": 1.517
 }

args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d84e9ea00131be5e82d3e7b5d5c98975f2a899a75e2edd6dfa9103b865020d55
+oid sha256:34f9c8c6b81e6c51ed1bf689bcd7d9afa7d549c13aab8759b5293a81f0562e75
 size 5588

eval_nbest_predictions.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:28538ee4151eeb932aecd40a67ef3f169af848d36e27fc3ee4ce46056a90d806
-size 23089559
+oid sha256:b7ba5ce8ae15e76967b66e6b24195f351bd78ffc0bc318203bd3d7cfa9805ca0
+size 23113024

eval_results.json

@@ -1,6 +1,6 @@
 {
-    "epoch": 20.0,
+    "epoch": 50.0,
     "eval_samples": 5000,
-    "exact_match": 55.64,
-    "f1": 65.9414733930188
+    "exact_match": 57.58,
+    "f1": 67.21309314821316
 }

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 20.0,
-    "train_loss": 0.6052761435508728,
-    "train_runtime": 530.4541,
+    "epoch": 50.0,
+    "train_loss": 0.24194956755638122,
+    "train_runtime": 1318.7789,
     "train_samples": 5000,
-    "train_samples_per_second": 188.518,
-    "train_steps_per_second": 1.508
+    "train_samples_per_second": 189.569,
+    "train_steps_per_second": 1.517
 }

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": 800,
+  "global_step": 2000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 12.5,
-      "learning_rate": 1.8750000000000002e-05,
-      "loss": 0.9181,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 0.8888,
       "step": 500
     },
     {
-      "epoch": 20.0,
-      "step": 800,
-      "total_flos": 9798982502400000.0,
-      "train_loss": 0.6052761435508728,
-      "train_runtime": 530.4541,
-      "train_samples_per_second": 188.518,
-      "train_steps_per_second": 1.508
+      "epoch": 25.0,
+      "learning_rate": 2.5e-05,
+      "loss": 0.0559,
+      "step": 1000
+    },
+    {
+      "epoch": 37.5,
+      "learning_rate": 1.25e-05,
+      "loss": 0.0163,
+      "step": 1500
+    },
+    {
+      "epoch": 50.0,
+      "learning_rate": 0.0,
+      "loss": 0.0068,
+      "step": 2000
+    },
+    {
+      "epoch": 50.0,
+      "step": 2000,
+      "total_flos": 2.4497456256e+16,
+      "train_loss": 0.24194956755638122,
+      "train_runtime": 1318.7789,
+      "train_samples_per_second": 189.569,
+      "train_steps_per_second": 1.517
     }
   ],
   "logging_steps": 500,
-  "max_steps": 800,
-  "num_train_epochs": 20,
+  "max_steps": 2000,
+  "num_train_epochs": 50,
   "save_steps": 500,
-  "total_flos": 9798982502400000.0,
+  "total_flos": 2.4497456256e+16,
   "trial_name": null,
   "trial_params": null
 }