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 WNLI
       type: glue
       config: wnli
       split: validation
@@ -28,9 +30,9 @@ should probably proofread and complete it, then remove this comment. -->
 
 # hBERTv2_new_pretrain_w_init__wnli
 
-This model is a fine-tuned version of [gokuls/bert_12_layer_model_v2_complete_training_new_wt_init](https://huggingface.co/gokuls/bert_12_layer_model_v2_complete_training_new_wt_init) on the glue dataset.
+This model is a fine-tuned version of [gokuls/bert_12_layer_model_v2_complete_training_new_wt_init](https://huggingface.co/gokuls/bert_12_layer_model_v2_complete_training_new_wt_init) on the GLUE WNLI dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.7003
+- Loss: 0.6990
 - Accuracy: 0.5634
 
 ## Model description

all_results.json

@@ -0,0 +1,14 @@
+{
+    "epoch": 8.0,
+    "eval_accuracy": 0.5633802816901409,
+    "eval_loss": 0.698971152305603,
+    "eval_runtime": 0.135,
+    "eval_samples": 71,
+    "eval_samples_per_second": 526.058,
+    "eval_steps_per_second": 7.409,
+    "train_loss": 0.7421743512153626,
+    "train_runtime": 52.5784,
+    "train_samples": 635,
+    "train_samples_per_second": 603.86,
+    "train_steps_per_second": 4.755
+}

eval_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 8.0,
+    "eval_accuracy": 0.5633802816901409,
+    "eval_loss": 0.698971152305603,
+    "eval_runtime": 0.135,
+    "eval_samples": 71,
+    "eval_samples_per_second": 526.058,
+    "eval_steps_per_second": 7.409
+}

logs/events.out.tfevents.1686061712.serv-3317.3380349.23

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d72c83cf3bf405b0a11fb88e23e45d8383f299c653f1866ca0d0fcb9d02f687a
+size 357

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 8.0,
+    "train_loss": 0.7421743512153626,
+    "train_runtime": 52.5784,
+    "train_samples": 635,
+    "train_samples_per_second": 603.86,
+    "train_steps_per_second": 4.755
+}

trainer_state.json

@@ -0,0 +1,145 @@
+{
+  "best_metric": 0.698971152305603,
+  "best_model_checkpoint": "hBERTv2_new_pretrain_w_init__wnli/checkpoint-15",
+  "epoch": 8.0,
+  "global_step": 40,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9200000000000004e-05,
+      "loss": 0.9111,
+      "step": 5
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.5492957746478874,
+      "eval_loss": 0.7287906408309937,
+      "eval_runtime": 0.1334,
+      "eval_samples_per_second": 532.163,
+      "eval_steps_per_second": 7.495,
+      "step": 5
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.8400000000000005e-05,
+      "loss": 0.7278,
+      "step": 10
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.5633802816901409,
+      "eval_loss": 0.7028499841690063,
+      "eval_runtime": 0.1349,
+      "eval_samples_per_second": 526.399,
+      "eval_steps_per_second": 7.414,
+      "step": 10
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 3.76e-05,
+      "loss": 0.707,
+      "step": 15
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.5633802816901409,
+      "eval_loss": 0.698971152305603,
+      "eval_runtime": 0.1329,
+      "eval_samples_per_second": 534.238,
+      "eval_steps_per_second": 7.524,
+      "step": 15
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 3.680000000000001e-05,
+      "loss": 0.7068,
+      "step": 20
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.43661971830985913,
+      "eval_loss": 0.7350625395774841,
+      "eval_runtime": 0.1282,
+      "eval_samples_per_second": 553.85,
+      "eval_steps_per_second": 7.801,
+      "step": 20
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 3.6e-05,
+      "loss": 0.7424,
+      "step": 25
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.5633802816901409,
+      "eval_loss": 0.7129454016685486,
+      "eval_runtime": 0.1329,
+      "eval_samples_per_second": 534.144,
+      "eval_steps_per_second": 7.523,
+      "step": 25
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 3.52e-05,
+      "loss": 0.7298,
+      "step": 30
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.43661971830985913,
+      "eval_loss": 0.710222065448761,
+      "eval_runtime": 0.1285,
+      "eval_samples_per_second": 552.441,
+      "eval_steps_per_second": 7.781,
+      "step": 30
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 3.44e-05,
+      "loss": 0.7043,
+      "step": 35
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.43661971830985913,
+      "eval_loss": 0.7217209935188293,
+      "eval_runtime": 0.1315,
+      "eval_samples_per_second": 540.126,
+      "eval_steps_per_second": 7.607,
+      "step": 35
+    },
+    {
+      "epoch": 8.0,
+      "learning_rate": 3.3600000000000004e-05,
+      "loss": 0.7081,
+      "step": 40
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.5633802816901409,
+      "eval_loss": 0.7002915143966675,
+      "eval_runtime": 0.1279,
+      "eval_samples_per_second": 554.968,
+      "eval_steps_per_second": 7.816,
+      "step": 40
+    },
+    {
+      "epoch": 8.0,
+      "step": 40,
+      "total_flos": 743403004362752.0,
+      "train_loss": 0.7421743512153626,
+      "train_runtime": 52.5784,
+      "train_samples_per_second": 603.86,
+      "train_steps_per_second": 4.755
+    }
+  ],
+  "max_steps": 250,
+  "num_train_epochs": 50,
+  "total_flos": 743403004362752.0,
+  "trial_name": null,
+  "trial_params": null
+}