End of training

README.md

@@ -1,4 +1,6 @@
 ---
+language:
+- en
 base_model: gokuls/bert_12_layer_model_v1_complete_training_new_wt_init_48
 tags:
 - generated_from_trainer
@@ -13,7 +15,7 @@ model-index:
       name: Text Classification
       type: text-classification
     dataset:
-      name: glue
+      name: GLUE STSB
       type: glue
       config: stsb
       split: validation
@@ -21,7 +23,7 @@ model-index:
     metrics:
     - name: Spearmanr
       type: spearmanr
-      value: 0.20585571946536435
+      value: 0.12474663100095418
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -29,12 +31,12 @@ should probably proofread and complete it, then remove this comment. -->
 
 # hBERTv1_new_pretrain_w_init_48_ver2_stsb
 
-This model is a fine-tuned version of [gokuls/bert_12_layer_model_v1_complete_training_new_wt_init_48](https://huggingface.co/gokuls/bert_12_layer_model_v1_complete_training_new_wt_init_48) on the glue dataset.
+This model is a fine-tuned version of [gokuls/bert_12_layer_model_v1_complete_training_new_wt_init_48](https://huggingface.co/gokuls/bert_12_layer_model_v1_complete_training_new_wt_init_48) on the GLUE STSB dataset.
 It achieves the following results on the evaluation set:
-- Loss: 2.3776
-- Pearson: 0.1911
-- Spearmanr: 0.2059
-- Combined Score: 0.1985
+- Loss: 2.2509
+- Pearson: 0.1285
+- Spearmanr: 0.1247
+- Combined Score: 0.1266
 
 ## Model description
 

all_results.json

@@ -0,0 +1,16 @@
+{
+    "epoch": 11.0,
+    "eval_combined_score": 0.12661347630524494,
+    "eval_loss": 2.2508649826049805,
+    "eval_pearson": 0.12848032160953568,
+    "eval_runtime": 7.6722,
+    "eval_samples": 1500,
+    "eval_samples_per_second": 195.512,
+    "eval_spearmanr": 0.12474663100095418,
+    "eval_steps_per_second": 3.128,
+    "train_loss": 2.139166490959399,
+    "train_runtime": 1123.3813,
+    "train_samples": 5749,
+    "train_samples_per_second": 76.764,
+    "train_steps_per_second": 1.202
+}

eval_results.json

@@ -0,0 +1,11 @@
+{
+    "epoch": 11.0,
+    "eval_combined_score": 0.12661347630524494,
+    "eval_loss": 2.2508649826049805,
+    "eval_pearson": 0.12848032160953568,
+    "eval_runtime": 7.6722,
+    "eval_samples": 1500,
+    "eval_samples_per_second": 195.512,
+    "eval_spearmanr": 0.12474663100095418,
+    "eval_steps_per_second": 3.128
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 11.0,
+    "train_loss": 2.139166490959399,
+    "train_runtime": 1123.3813,
+    "train_samples": 5749,
+    "train_samples_per_second": 76.764,
+    "train_steps_per_second": 1.202
+}

trainer_state.json

@@ -0,0 +1,215 @@
+{
+  "best_metric": 2.2508649826049805,
+  "best_model_checkpoint": "hBERTv1_new_pretrain_w_init_48_ver2_stsb/checkpoint-540",
+  "epoch": 11.0,
+  "eval_steps": 500,
+  "global_step": 990,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.733333333333334e-05,
+      "loss": 2.3716,
+      "step": 90
+    },
+    {
+      "epoch": 1.0,
+      "eval_combined_score": 0.09954876427078817,
+      "eval_loss": 2.4198310375213623,
+      "eval_pearson": 0.1235081678469039,
+      "eval_runtime": 7.919,
+      "eval_samples_per_second": 189.417,
+      "eval_spearmanr": 0.07558936069467244,
+      "eval_steps_per_second": 3.031,
+      "step": 90
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.466666666666667e-05,
+      "loss": 2.1648,
+      "step": 180
+    },
+    {
+      "epoch": 2.0,
+      "eval_combined_score": 0.059901129208988935,
+      "eval_loss": 2.421804904937744,
+      "eval_pearson": 0.05915702497074951,
+      "eval_runtime": 7.9228,
+      "eval_samples_per_second": 189.328,
+      "eval_spearmanr": 0.060645233447228365,
+      "eval_steps_per_second": 3.029,
+      "step": 180
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 2.1915,
+      "step": 270
+    },
+    {
+      "epoch": 3.0,
+      "eval_combined_score": 0.10509199818892286,
+      "eval_loss": 2.5305275917053223,
+      "eval_pearson": 0.11429814840247612,
+      "eval_runtime": 7.9275,
+      "eval_samples_per_second": 189.214,
+      "eval_spearmanr": 0.09588584797536959,
+      "eval_steps_per_second": 3.027,
+      "step": 270
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 2.9333333333333333e-05,
+      "loss": 2.1855,
+      "step": 360
+    },
+    {
+      "epoch": 4.0,
+      "eval_combined_score": 0.10434278350873159,
+      "eval_loss": 2.4912118911743164,
+      "eval_pearson": 0.11175276310573931,
+      "eval_runtime": 7.9356,
+      "eval_samples_per_second": 189.022,
+      "eval_spearmanr": 0.09693280391172387,
+      "eval_steps_per_second": 3.024,
+      "step": 360
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 2.1858,
+      "step": 450
+    },
+    {
+      "epoch": 5.0,
+      "eval_combined_score": 0.10867629190279143,
+      "eval_loss": 2.353940963745117,
+      "eval_pearson": 0.11303164261998537,
+      "eval_runtime": 7.8971,
+      "eval_samples_per_second": 189.944,
+      "eval_spearmanr": 0.10432094118559751,
+      "eval_steps_per_second": 3.039,
+      "step": 450
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 2.4e-05,
+      "loss": 2.1818,
+      "step": 540
+    },
+    {
+      "epoch": 6.0,
+      "eval_combined_score": 0.12661347630524494,
+      "eval_loss": 2.2508649826049805,
+      "eval_pearson": 0.12848032160953568,
+      "eval_runtime": 7.8875,
+      "eval_samples_per_second": 190.175,
+      "eval_spearmanr": 0.12474663100095418,
+      "eval_steps_per_second": 3.043,
+      "step": 540
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 2.1333333333333335e-05,
+      "loss": 2.2562,
+      "step": 630
+    },
+    {
+      "epoch": 7.0,
+      "eval_combined_score": 0.10085775212511634,
+      "eval_loss": 2.330181360244751,
+      "eval_pearson": 0.10432624437463486,
+      "eval_runtime": 7.8945,
+      "eval_samples_per_second": 190.005,
+      "eval_spearmanr": 0.09738925987559782,
+      "eval_steps_per_second": 3.04,
+      "step": 630
+    },
+    {
+      "epoch": 8.0,
+      "learning_rate": 1.866666666666667e-05,
+      "loss": 2.2299,
+      "step": 720
+    },
+    {
+      "epoch": 8.0,
+      "eval_combined_score": 0.17028955228719306,
+      "eval_loss": 2.3748996257781982,
+      "eval_pearson": 0.1984074571487062,
+      "eval_runtime": 7.8956,
+      "eval_samples_per_second": 189.98,
+      "eval_spearmanr": 0.14217164742567992,
+      "eval_steps_per_second": 3.04,
+      "step": 720
+    },
+    {
+      "epoch": 9.0,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 2.0676,
+      "step": 810
+    },
+    {
+      "epoch": 9.0,
+      "eval_combined_score": 0.13143073261190957,
+      "eval_loss": 2.3883156776428223,
+      "eval_pearson": 0.12998858260155782,
+      "eval_runtime": 7.9009,
+      "eval_samples_per_second": 189.851,
+      "eval_spearmanr": 0.13287288262226135,
+      "eval_steps_per_second": 3.038,
+      "step": 810
+    },
+    {
+      "epoch": 10.0,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 1.926,
+      "step": 900
+    },
+    {
+      "epoch": 10.0,
+      "eval_combined_score": 0.12458804638022844,
+      "eval_loss": 2.588383674621582,
+      "eval_pearson": 0.12590757541676778,
+      "eval_runtime": 7.9206,
+      "eval_samples_per_second": 189.38,
+      "eval_spearmanr": 0.1232685173436891,
+      "eval_steps_per_second": 3.03,
+      "step": 900
+    },
+    {
+      "epoch": 11.0,
+      "learning_rate": 1.0666666666666667e-05,
+      "loss": 1.7701,
+      "step": 990
+    },
+    {
+      "epoch": 11.0,
+      "eval_combined_score": 0.1984576696603884,
+      "eval_loss": 2.3776161670684814,
+      "eval_pearson": 0.19105961985541245,
+      "eval_runtime": 7.9395,
+      "eval_samples_per_second": 188.928,
+      "eval_spearmanr": 0.20585571946536435,
+      "eval_steps_per_second": 3.023,
+      "step": 990
+    },
+    {
+      "epoch": 11.0,
+      "step": 990,
+      "total_flos": 9365417243967488.0,
+      "train_loss": 2.139166490959399,
+      "train_runtime": 1123.3813,
+      "train_samples_per_second": 76.764,
+      "train_steps_per_second": 1.202
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1350,
+  "num_train_epochs": 15,
+  "save_steps": 500,
+  "total_flos": 9365417243967488.0,
+  "trial_name": null,
+  "trial_params": null
+}