add model

README.md

@@ -1,34 +1,57 @@
 ---
-license: afl-3.0
-
-widget:
-- text: "as evidence mounts showing that covid leaked from a lab, it’s become clear fauci was funding coronavirus research in wuhan. and he lied about it. he’s lied about lockdowns, he’s lied about masks, and now we know he’s lied about the origins of the virus. https://nypost.com/2021/09/08/covering-up-faucis-role-in-covid-leak-devine/"
-  example_title: "example 1"
-- text: "kansas small businesses are getting hit hard by the covid-19 outbreak and response. if you or someone you know needs guidance and resources, please use this link: https://www.sba.gov/page/coronavirus-covid-19-small-business-guidance-loan-resources #ksleg"
-  example_title: "example 2"
-  
+tags:
+- generated_from_keras_callback
+model-index:
+- name: US_politicians_covid_skepticism
+  results: []
 ---
 
-This model is trained to classify US legislators' tweets about COVID-19 as skeptical of preventative measures or not. The model is fine-tuned on 10k hand coded tweets using vinai/bertweet-covid19-base-uncased (https://github.com/VinAIResearch/BERTweet). 
+<!-- This model card has been generated automatically according to the information Keras had access to. You should
+probably proofread and complete it, then remove this comment. -->
 
-The model is uncased (all text is lowercase). Model was trained with full tweets, including URLs and @USER mentions
+# US_politicians_covid_skepticism
 
+This model is a fine-tuned version of [vinai/bertweet-covid19-base-uncased](https://huggingface.co/vinai/bertweet-covid19-base-uncased) on an unknown dataset.
+It achieves the following results on the evaluation set:
+- Train Loss: 0.1745
+- Train Sparse Categorical Accuracy: 0.9680
+- Validation Loss: 0.1170
+- Validation Sparse Categorical Accuracy: 0.9814
+- Epoch: 3
 
+## Model description
 
-sparse_categorical_accuracy: 0.9618
+More information needed
 
+## Intended uses & limitations
 
-Sklearn validation metrics:  
+More information needed
 
-Precision: 0.927190139237617
+## Training and evaluation data
 
-Recall: 0.969193742478941
+More information needed
 
-F1 score: 0.9465885064563841
+## Training procedure
 
+### Training hyperparameters
 
-Epoch 1/3 - loss: 0.4504 - sparse_categorical_accuracy: 0.8577 - val_loss: 0.3473 - val_sparse_categorical_accuracy: 0.8706
+The following hyperparameters were used during training:
+- optimizer: {'name': 'Adam', 'learning_rate': 5e-07, 'decay': 0.0, 'beta_1': 0.9, 'beta_2': 0.999, 'epsilon': 1e-07, 'amsgrad': False}
+- training_precision: float32
 
-Epoch 2/3 - loss: 0.3309 - sparse_categorical_accuracy: 0.8593 - val_loss: 0.2395 - val_sparse_categorical_accuracy: 0.8744
+### Training results
 
-Epoch 3/3 - loss: 0.1862 - sparse_categorical_accuracy: 0.9439 - val_loss: 0.1348 - val_sparse_categorical_accuracy: 0.9618
+| Train Loss | Train Sparse Categorical Accuracy | Validation Loss | Validation Sparse Categorical Accuracy | Epoch |
+|:----------:|:---------------------------------:|:---------------:|:--------------------------------------:|:-----:|
+| 0.6242     | 0.6733                            | 0.5466          | 0.6911                                 | 0     |
+| 0.4520     | 0.8283                            | 0.2888          | 0.9628                                 | 1     |
+| 0.2561     | 0.9587                            | 0.1626          | 0.9783                                 | 2     |
+| 0.1745     | 0.9680                            | 0.1170          | 0.9814                                 | 3     |
+
+
+### Framework versions
+
+- Transformers 4.20.1
+- TensorFlow 2.8.2
+- Datasets 2.3.2
+- Tokenizers 0.12.1

config.json

@@ -1,5 +1,5 @@
 {
-  "_name_or_path": "/content/drive/MyDrive/NLP_models/covid_skepticism_tweets-bert-uncased.t5",
+  "_name_or_path": "vinai/bertweet-covid19-base-uncased",
   "architectures": [
     "RobertaForSequenceClassification"
   ],
@@ -21,7 +21,6 @@
   "pad_token_id": 1,
   "position_embedding_type": "absolute",
   "tokenizer_class": "BertweetTokenizer",
-  "torch_dtype": "float32",
   "transformers_version": "4.20.1",
   "type_vocab_size": 1,
   "use_cache": true,

tf_model.h5

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9e7285a5bc51ebef61b437ca087df4065c1e1ebb7f4387865f4201981d839c57
+size 539895680