Upload TFDistilBertForSequenceClassification

README.md

@@ -33,7 +33,7 @@ More information needed
 ### Training hyperparameters
 
 The following hyperparameters were used during training:
-- optimizer: {'name': 'AdamWeightDecay', 'learning_rate': {'class_name': 'WarmUp', 'config': {'initial_learning_rate': 2e-05, 'decay_schedule_fn': {'class_name': 'PolynomialDecay', 'config': {'initial_learning_rate': 2e-05, 'decay_steps': 3682, 'end_learning_rate': 0.0, 'power': 1.0, 'cycle': False, 'name': None}, '__passive_serialization__': True}, 'warmup_steps': 3, 'power': 1.0, 'name': None}}, 'decay': 0.0, 'beta_1': 0.9, 'beta_2': 0.999, 'epsilon': 1e-08, 'amsgrad': False, 'weight_decay_rate': 0.6}
+- optimizer: {'name': 'AdamWeightDecay', 'learning_rate': {'class_name': 'WarmUp', 'config': {'initial_learning_rate': 3.0000000000000004e-05, 'decay_schedule_fn': {'class_name': 'PolynomialDecay', 'config': {'initial_learning_rate': 3.0000000000000004e-05, 'decay_steps': 3682, 'end_learning_rate': 0.0, 'power': 1.0, 'cycle': False, 'name': None}, '__passive_serialization__': True}, 'warmup_steps': 3, 'power': 1.0, 'name': None}}, 'decay': 0.0, 'beta_1': 0.9, 'beta_2': 0.999, 'epsilon': 1e-08, 'amsgrad': False, 'weight_decay_rate': 0.1}
 - training_precision: float32
 
 ### Training results

config.json

@@ -4,9 +4,9 @@
   "architectures": [
     "DistilBertForSequenceClassification"
   ],
-  "attention_dropout": 0.2,
+  "attention_dropout": 0.5,
   "dim": 768,
-  "dropout": 0.2,
+  "dropout": 0.30000000000000004,
   "hidden_dim": 3072,
   "id2label": {
     "0": "NEGATIVE",
@@ -20,12 +20,12 @@
   "max_position_embeddings": 512,
   "model_type": "distilbert",
   "n_heads": 12,
-  "n_layers": 7,
+  "n_layers": 10,
   "output_past": true,
   "pad_token_id": 0,
   "qa_dropout": 0.1,
-  "seq_classif_dropout": 0.6,
-  "sinusoidal_pos_embds": false,
+  "seq_classif_dropout": 0.30000000000000004,
+  "sinusoidal_pos_embds": true,
   "tie_weights_": true,
   "transformers_version": "4.30.2",
   "vocab_size": 119548

tf_model.h5

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a3e415d19585e8d396779a51951569d39f0adc412b53eeb408537c9fa313032b
-size 937070736
+oid sha256:15425064f2aa11a123d997ee32b3f390bbd8ad286dbe57fe40db3d3c2c04eb59
+size 1022173832