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 RTE
       type: glue
       config: rte
       split: validation
@@ -20,7 +22,7 @@ model-index:
     metrics:
     - name: Accuracy
       type: accuracy
-      value: 0.5234657039711191
+      value: 0.5487364620938628
 ---
 
 <!-- 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. -->
 
 # hBERTv2_rte
 
-This model is a fine-tuned version of [gokuls/bert_12_layer_model_v2](https://huggingface.co/gokuls/bert_12_layer_model_v2) on the glue dataset.
+This model is a fine-tuned version of [gokuls/bert_12_layer_model_v2](https://huggingface.co/gokuls/bert_12_layer_model_v2) on the GLUE RTE dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.8160
-- Accuracy: 0.5235
+- Loss: 0.6896
+- Accuracy: 0.5487
 
 ## Model description
 

all_results.json

@@ -0,0 +1,14 @@
+{
+    "epoch": 12.0,
+    "eval_accuracy": 0.5487364620938628,
+    "eval_loss": 0.6896100044250488,
+    "eval_runtime": 0.3599,
+    "eval_samples": 277,
+    "eval_samples_per_second": 769.561,
+    "eval_steps_per_second": 5.556,
+    "train_loss": 0.660737669467926,
+    "train_runtime": 158.6499,
+    "train_samples": 2490,
+    "train_samples_per_second": 784.747,
+    "train_steps_per_second": 3.152
+}

eval_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 12.0,
+    "eval_accuracy": 0.5487364620938628,
+    "eval_loss": 0.6896100044250488,
+    "eval_runtime": 0.3599,
+    "eval_samples": 277,
+    "eval_samples_per_second": 769.561,
+    "eval_steps_per_second": 5.556
+}

logs/events.out.tfevents.1677677335.kyoto.1072791.14

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7d404272a3b068d281020904e59d82a64b92326cfe12e530a9ca6abc723e8370
+size 357

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 12.0,
+    "train_loss": 0.660737669467926,
+    "train_runtime": 158.6499,
+    "train_samples": 2490,
+    "train_samples_per_second": 784.747,
+    "train_steps_per_second": 3.152
+}

trainer_state.json

@@ -0,0 +1,205 @@
+{
+  "best_metric": 0.6896100044250488,
+  "best_model_checkpoint": "hBERTv2_rte/checkpoint-70",
+  "epoch": 12.0,
+  "global_step": 120,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.9e-05,
+      "loss": 0.7231,
+      "step": 10
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.4548736462093863,
+      "eval_loss": 0.7174825668334961,
+      "eval_runtime": 0.3347,
+      "eval_samples_per_second": 827.629,
+      "eval_steps_per_second": 5.976,
+      "step": 10
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.8e-05,
+      "loss": 0.702,
+      "step": 20
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.4729241877256318,
+      "eval_loss": 0.7052958011627197,
+      "eval_runtime": 0.3353,
+      "eval_samples_per_second": 826.133,
+      "eval_steps_per_second": 5.965,
+      "step": 20
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 4.7e-05,
+      "loss": 0.6982,
+      "step": 30
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.4584837545126354,
+      "eval_loss": 0.6975917220115662,
+      "eval_runtime": 0.3314,
+      "eval_samples_per_second": 835.799,
+      "eval_steps_per_second": 6.035,
+      "step": 30
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 0.7008,
+      "step": 40
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.4657039711191336,
+      "eval_loss": 0.726104199886322,
+      "eval_runtime": 0.3334,
+      "eval_samples_per_second": 830.944,
+      "eval_steps_per_second": 6.0,
+      "step": 40
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 4.5e-05,
+      "loss": 0.7022,
+      "step": 50
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.49458483754512633,
+      "eval_loss": 0.7142461538314819,
+      "eval_runtime": 0.3346,
+      "eval_samples_per_second": 827.866,
+      "eval_steps_per_second": 5.977,
+      "step": 50
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 0.6867,
+      "step": 60
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.48014440433212996,
+      "eval_loss": 0.694332480430603,
+      "eval_runtime": 0.3341,
+      "eval_samples_per_second": 829.027,
+      "eval_steps_per_second": 5.986,
+      "step": 60
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 4.3e-05,
+      "loss": 0.6796,
+      "step": 70
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.5487364620938628,
+      "eval_loss": 0.6896100044250488,
+      "eval_runtime": 0.3351,
+      "eval_samples_per_second": 826.637,
+      "eval_steps_per_second": 5.968,
+      "step": 70
+    },
+    {
+      "epoch": 8.0,
+      "learning_rate": 4.2e-05,
+      "loss": 0.6614,
+      "step": 80
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.516245487364621,
+      "eval_loss": 0.7151486277580261,
+      "eval_runtime": 0.3354,
+      "eval_samples_per_second": 825.95,
+      "eval_steps_per_second": 5.964,
+      "step": 80
+    },
+    {
+      "epoch": 9.0,
+      "learning_rate": 4.1e-05,
+      "loss": 0.6303,
+      "step": 90
+    },
+    {
+      "epoch": 9.0,
+      "eval_accuracy": 0.5270758122743683,
+      "eval_loss": 0.7244121432304382,
+      "eval_runtime": 0.3343,
+      "eval_samples_per_second": 828.517,
+      "eval_steps_per_second": 5.982,
+      "step": 90
+    },
+    {
+      "epoch": 10.0,
+      "learning_rate": 4e-05,
+      "loss": 0.602,
+      "step": 100
+    },
+    {
+      "epoch": 10.0,
+      "eval_accuracy": 0.4729241877256318,
+      "eval_loss": 0.7570322751998901,
+      "eval_runtime": 0.3345,
+      "eval_samples_per_second": 828.025,
+      "eval_steps_per_second": 5.979,
+      "step": 100
+    },
+    {
+      "epoch": 11.0,
+      "learning_rate": 3.9000000000000006e-05,
+      "loss": 0.5761,
+      "step": 110
+    },
+    {
+      "epoch": 11.0,
+      "eval_accuracy": 0.5379061371841155,
+      "eval_loss": 0.7605395913124084,
+      "eval_runtime": 0.3342,
+      "eval_samples_per_second": 828.846,
+      "eval_steps_per_second": 5.984,
+      "step": 110
+    },
+    {
+      "epoch": 12.0,
+      "learning_rate": 3.8e-05,
+      "loss": 0.5664,
+      "step": 120
+    },
+    {
+      "epoch": 12.0,
+      "eval_accuracy": 0.5234657039711191,
+      "eval_loss": 0.8159566521644592,
+      "eval_runtime": 0.335,
+      "eval_samples_per_second": 826.903,
+      "eval_steps_per_second": 5.97,
+      "step": 120
+    },
+    {
+      "epoch": 12.0,
+      "step": 120,
+      "total_flos": 3721656539283456.0,
+      "train_loss": 0.660737669467926,
+      "train_runtime": 158.6499,
+      "train_samples_per_second": 784.747,
+      "train_steps_per_second": 3.152
+    }
+  ],
+  "max_steps": 500,
+  "num_train_epochs": 50,
+  "total_flos": 3721656539283456.0,
+  "trial_name": null,
+  "trial_params": null
+}