End of training

README.md

@@ -1,4 +1,6 @@
 ---
+language:
+- en
 tags:
 - generated_from_trainer
 datasets:
@@ -12,7 +14,7 @@ model-index:
       name: Text Classification
       type: text-classification
     dataset:
-      name: glue
+      name: GLUE SST2
       type: glue
       config: sst2
       split: validation
@@ -20,7 +22,7 @@ model-index:
     metrics:
     - name: Accuracy
       type: accuracy
-      value: 0.8211009174311926
+      value: 0.8038990825688074
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -28,10 +30,10 @@ should probably proofread and complete it, then remove this comment. -->
 
 # hBERTv1_new_pretrain_w_init__sst2
 
-This model is a fine-tuned version of [gokuls/bert_12_layer_model_v1_complete_training_new_wt_init](https://huggingface.co/gokuls/bert_12_layer_model_v1_complete_training_new_wt_init) on the glue dataset.
+This model is a fine-tuned version of [gokuls/bert_12_layer_model_v1_complete_training_new_wt_init](https://huggingface.co/gokuls/bert_12_layer_model_v1_complete_training_new_wt_init) on the GLUE SST2 dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.6468
-- Accuracy: 0.8211
+- Loss: 0.4606
+- Accuracy: 0.8039
 
 ## Model description
 

all_results.json

@@ -1,14 +1,14 @@
 {
-    "epoch": 13.0,
-    "eval_accuracy": 0.5091743119266054,
-    "eval_loss": 0.6952945590019226,
-    "eval_runtime": 1.5909,
+    "epoch": 6.0,
+    "eval_accuracy": 0.8038990825688074,
+    "eval_loss": 0.4606306850910187,
+    "eval_runtime": 1.574,
     "eval_samples": 872,
-    "eval_samples_per_second": 548.119,
-    "eval_steps_per_second": 4.4,
-    "train_loss": 0.7048921508521794,
-    "train_runtime": 4728.5552,
+    "eval_samples_per_second": 554.009,
+    "eval_steps_per_second": 4.447,
+    "train_loss": 0.18471441202567845,
+    "train_runtime": 2261.4914,
     "train_samples": 67349,
-    "train_samples_per_second": 712.152,
-    "train_steps_per_second": 5.573
+    "train_samples_per_second": 1489.039,
+    "train_steps_per_second": 11.652
 }

eval_results.json

@@ -1,9 +1,9 @@
 {
-    "epoch": 13.0,
-    "eval_accuracy": 0.5091743119266054,
-    "eval_loss": 0.6952945590019226,
-    "eval_runtime": 1.5909,
+    "epoch": 6.0,
+    "eval_accuracy": 0.8038990825688074,
+    "eval_loss": 0.4606306850910187,
+    "eval_runtime": 1.574,
     "eval_samples": 872,
-    "eval_samples_per_second": 548.119,
-    "eval_steps_per_second": 4.4
+    "eval_samples_per_second": 554.009,
+    "eval_steps_per_second": 4.447
 }

logs/events.out.tfevents.1686032261.serv-3317.3377907.2

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e1b8462fa8a877542b92770e4e5ba3dc7eaa09419ab0af63c86054800628d028
+size 363

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 13.0,
-    "train_loss": 0.7048921508521794,
-    "train_runtime": 4728.5552,
+    "epoch": 6.0,
+    "train_loss": 0.18471441202567845,
+    "train_runtime": 2261.4914,
     "train_samples": 67349,
-    "train_samples_per_second": 712.152,
-    "train_steps_per_second": 5.573
+    "train_samples_per_second": 1489.039,
+    "train_steps_per_second": 11.652
 }

trainer_state.json

@@ -1,220 +1,115 @@
 {
-  "best_metric": 0.6952945590019226,
-  "best_model_checkpoint": "hBERTv1_new_pretrain_w_init__sst2/checkpoint-4216",
-  "epoch": 13.0,
-  "global_step": 6851,
+  "best_metric": 0.4606306850910187,
+  "best_model_checkpoint": "hBERTv1_new_pretrain_w_init__sst2/checkpoint-527",
+  "epoch": 6.0,
+  "global_step": 3162,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 1.0,
-      "learning_rate": 0.00049,
-      "loss": 0.9241,
+      "learning_rate": 3.9200000000000004e-05,
+      "loss": 0.363,
       "step": 527
     },
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.5091743119266054,
-      "eval_loss": 0.6974940896034241,
-      "eval_runtime": 1.5777,
-      "eval_samples_per_second": 552.709,
-      "eval_steps_per_second": 4.437,
+      "eval_accuracy": 0.8038990825688074,
+      "eval_loss": 0.4606306850910187,
+      "eval_runtime": 1.5758,
+      "eval_samples_per_second": 553.365,
+      "eval_steps_per_second": 4.442,
       "step": 527
     },
     {
       "epoch": 2.0,
-      "learning_rate": 0.00048,
-      "loss": 0.6868,
+      "learning_rate": 3.8400000000000005e-05,
+      "loss": 0.2256,
       "step": 1054
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.5091743119266054,
-      "eval_loss": 0.6956753730773926,
-      "eval_runtime": 1.5748,
-      "eval_samples_per_second": 553.735,
-      "eval_steps_per_second": 4.445,
+      "eval_accuracy": 0.8119266055045872,
+      "eval_loss": 0.6466029286384583,
+      "eval_runtime": 1.581,
+      "eval_samples_per_second": 551.543,
+      "eval_steps_per_second": 4.428,
       "step": 1054
     },
     {
       "epoch": 3.0,
-      "learning_rate": 0.00047,
-      "loss": 0.6882,
+      "learning_rate": 3.76e-05,
+      "loss": 0.1754,
       "step": 1581
     },
     {
       "epoch": 3.0,
-      "eval_accuracy": 0.5091743119266054,
-      "eval_loss": 0.698914110660553,
-      "eval_runtime": 1.575,
-      "eval_samples_per_second": 553.639,
-      "eval_steps_per_second": 4.444,
+      "eval_accuracy": 0.8176605504587156,
+      "eval_loss": 0.5100513696670532,
+      "eval_runtime": 1.5786,
+      "eval_samples_per_second": 552.387,
+      "eval_steps_per_second": 4.434,
       "step": 1581
     },
     {
       "epoch": 4.0,
-      "learning_rate": 0.00046,
-      "loss": 0.6866,
+      "learning_rate": 3.680000000000001e-05,
+      "loss": 0.1394,
       "step": 2108
     },
     {
       "epoch": 4.0,
-      "eval_accuracy": 0.5091743119266054,
-      "eval_loss": 0.695652961730957,
-      "eval_runtime": 1.5735,
-      "eval_samples_per_second": 554.175,
-      "eval_steps_per_second": 4.449,
+      "eval_accuracy": 0.8176605504587156,
+      "eval_loss": 0.49214717745780945,
+      "eval_runtime": 1.5822,
+      "eval_samples_per_second": 551.116,
+      "eval_steps_per_second": 4.424,
       "step": 2108
     },
     {
       "epoch": 5.0,
-      "learning_rate": 0.00045000000000000004,
-      "loss": 0.6866,
+      "learning_rate": 3.6e-05,
+      "loss": 0.1111,
       "step": 2635
     },
     {
       "epoch": 5.0,
-      "eval_accuracy": 0.5091743119266054,
-      "eval_loss": 0.6973552107810974,
-      "eval_runtime": 1.5748,
-      "eval_samples_per_second": 553.712,
-      "eval_steps_per_second": 4.445,
+      "eval_accuracy": 0.819954128440367,
+      "eval_loss": 0.5109940767288208,
+      "eval_runtime": 1.5752,
+      "eval_samples_per_second": 553.567,
+      "eval_steps_per_second": 4.444,
       "step": 2635
     },
     {
       "epoch": 6.0,
-      "learning_rate": 0.00044,
-      "loss": 0.6864,
+      "learning_rate": 3.52e-05,
+      "loss": 0.0937,
       "step": 3162
     },
     {
       "epoch": 6.0,
-      "eval_accuracy": 0.5091743119266054,
-      "eval_loss": 0.698842465877533,
-      "eval_runtime": 1.5728,
-      "eval_samples_per_second": 554.418,
-      "eval_steps_per_second": 4.451,
+      "eval_accuracy": 0.8211009174311926,
+      "eval_loss": 0.6468406319618225,
+      "eval_runtime": 1.5781,
+      "eval_samples_per_second": 552.564,
+      "eval_steps_per_second": 4.436,
       "step": 3162
     },
     {
-      "epoch": 7.0,
-      "learning_rate": 0.00043,
-      "loss": 0.6864,
-      "step": 3689
-    },
-    {
-      "epoch": 7.0,
-      "eval_accuracy": 0.5091743119266054,
-      "eval_loss": 0.6953662633895874,
-      "eval_runtime": 1.5757,
-      "eval_samples_per_second": 553.388,
-      "eval_steps_per_second": 4.442,
-      "step": 3689
-    },
-    {
-      "epoch": 8.0,
-      "learning_rate": 0.00042,
-      "loss": 0.6862,
-      "step": 4216
-    },
-    {
-      "epoch": 8.0,
-      "eval_accuracy": 0.5091743119266054,
-      "eval_loss": 0.6952945590019226,
-      "eval_runtime": 1.5762,
-      "eval_samples_per_second": 553.246,
-      "eval_steps_per_second": 4.441,
-      "step": 4216
-    },
-    {
-      "epoch": 9.0,
-      "learning_rate": 0.00041,
-      "loss": 0.6862,
-      "step": 4743
-    },
-    {
-      "epoch": 9.0,
-      "eval_accuracy": 0.5091743119266054,
-      "eval_loss": 0.6971402168273926,
-      "eval_runtime": 1.5764,
-      "eval_samples_per_second": 553.174,
-      "eval_steps_per_second": 4.441,
-      "step": 4743
-    },
-    {
-      "epoch": 10.0,
-      "learning_rate": 0.0004,
-      "loss": 0.6864,
-      "step": 5270
-    },
-    {
-      "epoch": 10.0,
-      "eval_accuracy": 0.5091743119266054,
-      "eval_loss": 0.6990574598312378,
-      "eval_runtime": 1.5894,
-      "eval_samples_per_second": 548.634,
-      "eval_steps_per_second": 4.404,
-      "step": 5270
-    },
-    {
-      "epoch": 11.0,
-      "learning_rate": 0.00039000000000000005,
-      "loss": 0.6865,
-      "step": 5797
-    },
-    {
-      "epoch": 11.0,
-      "eval_accuracy": 0.5091743119266054,
-      "eval_loss": 0.6973507404327393,
-      "eval_runtime": 1.571,
-      "eval_samples_per_second": 555.055,
-      "eval_steps_per_second": 4.456,
-      "step": 5797
-    },
-    {
-      "epoch": 12.0,
-      "learning_rate": 0.00038,
-      "loss": 0.6863,
-      "step": 6324
-    },
-    {
-      "epoch": 12.0,
-      "eval_accuracy": 0.5091743119266054,
-      "eval_loss": 0.6972925066947937,
-      "eval_runtime": 1.5755,
-      "eval_samples_per_second": 553.473,
-      "eval_steps_per_second": 4.443,
-      "step": 6324
-    },
-    {
-      "epoch": 13.0,
-      "learning_rate": 0.00037,
-      "loss": 0.6869,
-      "step": 6851
-    },
-    {
-      "epoch": 13.0,
-      "eval_accuracy": 0.5091743119266054,
-      "eval_loss": 0.6990574598312378,
-      "eval_runtime": 1.5748,
-      "eval_samples_per_second": 553.725,
-      "eval_steps_per_second": 4.445,
-      "step": 6851
-    },
-    {
-      "epoch": 13.0,
-      "step": 6851,
-      "total_flos": 1.2966420045155533e+17,
-      "train_loss": 0.7048921508521794,
-      "train_runtime": 4728.5552,
-      "train_samples_per_second": 712.152,
-      "train_steps_per_second": 5.573
+      "epoch": 6.0,
+      "step": 3162,
+      "total_flos": 5.984501559302554e+16,
+      "train_loss": 0.18471441202567845,
+      "train_runtime": 2261.4914,
+      "train_samples_per_second": 1489.039,
+      "train_steps_per_second": 11.652
     }
   ],
   "max_steps": 26350,
   "num_train_epochs": 50,
-  "total_flos": 1.2966420045155533e+17,
+  "total_flos": 5.984501559302554e+16,
   "trial_name": null,
   "trial_params": null
 }