End of training

README.md

@@ -2,7 +2,10 @@
 license: apache-2.0
 base_model: albert-base-v2
 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 [albert-base-v2](https://huggingface.co/albert-base-v2) on the None dataset.
+This model is a fine-tuned version of [albert-base-v2](https://huggingface.co/albert-base-v2) on the squad dataset.
 
 ## Model description
 

all_results.json

@@ -1,11 +1,11 @@
 {
-    "epoch": 20.0,
+    "epoch": 50.0,
     "eval_samples": 3000,
-    "exact_match": 69.36666666666666,
-    "f1": 79.08588444044621,
-    "train_loss": 0.22334851051898713,
-    "train_runtime": 394.6502,
+    "exact_match": 65.83333333333333,
+    "f1": 75.68173539698708,
+    "train_loss": 0.1024451555850658,
+    "train_runtime": 973.2779,
     "train_samples": 3000,
-    "train_samples_per_second": 152.033,
-    "train_steps_per_second": 2.382
+    "train_samples_per_second": 154.118,
+    "train_steps_per_second": 2.415
 }

args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8ebb76576d393772ce9acaf130f6a516d00cca9735839842c18cad289f875a8f
+oid sha256:add2807abfa827cf5b3f285cd7ef01dbacc35e81a62dc4112f794b52df11bbb0
 size 5652

eval_nbest_predictions.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:43e54f478041686203905682d07dcebcef703f5d5cc43b46912c0f93f9219735
-size 13496195
+oid sha256:6ad9291793a25acbfbbbb2e0ac0d1fe6fb245e1a01e4fb2b1d764a834530fdcf
+size 13511336

eval_results.json

@@ -1,6 +1,6 @@
 {
-    "epoch": 20.0,
+    "epoch": 50.0,
     "eval_samples": 3000,
-    "exact_match": 69.36666666666666,
-    "f1": 79.08588444044621
+    "exact_match": 65.83333333333333,
+    "f1": 75.68173539698708
 }

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 20.0,
-    "train_loss": 0.22334851051898713,
-    "train_runtime": 394.6502,
+    "epoch": 50.0,
+    "train_loss": 0.1024451555850658,
+    "train_runtime": 973.2779,
     "train_samples": 3000,
-    "train_samples_per_second": 152.033,
-    "train_steps_per_second": 2.382
+    "train_samples_per_second": 154.118,
+    "train_steps_per_second": 2.415
 }

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": 0.4117,
+      "learning_rate": 3.936170212765958e-05,
+      "loss": 0.4275,
       "step": 500
     },
     {
-      "epoch": 20.0,
-      "step": 940,
-      "total_flos": 993771141120000.0,
-      "train_loss": 0.22334851051898713,
-      "train_runtime": 394.6502,
-      "train_samples_per_second": 152.033,
-      "train_steps_per_second": 2.382
+      "epoch": 21.28,
+      "learning_rate": 2.8723404255319154e-05,
+      "loss": 0.037,
+      "step": 1000
+    },
+    {
+      "epoch": 31.91,
+      "learning_rate": 1.8085106382978724e-05,
+      "loss": 0.0104,
+      "step": 1500
+    },
+    {
+      "epoch": 42.55,
+      "learning_rate": 7.446808510638298e-06,
+      "loss": 0.0019,
+      "step": 2000
+    },
+    {
+      "epoch": 50.0,
+      "step": 2350,
+      "total_flos": 2484427852800000.0,
+      "train_loss": 0.1024451555850658,
+      "train_runtime": 973.2779,
+      "train_samples_per_second": 154.118,
+      "train_steps_per_second": 2.415
     }
   ],
   "logging_steps": 500,
-  "max_steps": 940,
-  "num_train_epochs": 20,
+  "max_steps": 2350,
+  "num_train_epochs": 50,
   "save_steps": 500,
-  "total_flos": 993771141120000.0,
+  "total_flos": 2484427852800000.0,
   "trial_name": null,
   "trial_params": null
 }