add chinese

experiments/huggingface_ops.py

@@ -3,8 +3,10 @@ from pprint import pprint
 
 api = HfApi()
 models = api.list_models(filter=ModelFilter(author='tweettemposhift'))
-models_filtered = [i.modelId for i in models if 'twitter-roberta-base-sep2021' in i.modelId]
+models_filtered = [i.modelId for i in models if 'emoji' in i.modelId]
 pprint(sorted(models_filtered))
+for i in models_filtered:
+    api.delete_repo(i, repo_type="model")
 # models = api.list_models(filter=ModelFilter(author='tweettemposhift'))
 # models_filtered = [i.modelId for i in models if 'topic-' in i.modelId]
 # pprint(sorted([i for i in models_filtered if i.endswith('twitter-roberta-base-2019-90m')]))

experiments/main.sh

@@ -5,70 +5,41 @@ MODEL="jhu-clsp/bernice"
 MODEL="roberta-large"
 MODEL="vinai/bertweet-large"
 MODEL="cardiffnlp/twitter-roberta-base-2019-90m"
-
 MODEL="cardiffnlp/twitter-roberta-base-dec2020"
 MODEL="cardiffnlp/twitter-roberta-base-2021-124m"
-MODEL="cardiffnlp/twitter-roberta-base-2022-154m"
-
 MODEL="cardiffnlp/twitter-roberta-large-2022-154m"
-
+MODEL="cardiffnlp/twitter-roberta-base-2022-154m"
 
 # EMOJI
 python model_finetuning_emoji.py -m "${MODEL}" -d "emoji_temporal"
-rm -rf "ckpt/emoji*${MODEL##*/}"
 python model_finetuning_emoji.py -m "${MODEL}" -d "emoji_random0_seed0"
-rm -rf "ckpt/emoji*${MODEL##*/}"
 python model_finetuning_emoji.py -m "${MODEL}" -d "emoji_random1_seed0"
-rm -rf "ckpt/emoji*${MODEL##*/}"
 python model_finetuning_emoji.py -m "${MODEL}" -d "emoji_random2_seed0"
-rm -rf "ckpt/emoji*${MODEL##*/}"
 python model_finetuning_emoji.py -m "${MODEL}" -d "emoji_random3_seed0"
-rm -rf "ckpt/emoji*${MODEL##*/}"
 python model_finetuning_emoji.py -m "${MODEL}" -d "emoji_random0_seed1"
-rm -rf "ckpt/emoji*${MODEL##*/}"
 python model_finetuning_emoji.py -m "${MODEL}" -d "emoji_random1_seed1"
-rm -rf "ckpt/emoji*${MODEL##*/}"
 python model_finetuning_emoji.py -m "${MODEL}" -d "emoji_random2_seed1"
-rm -rf "ckpt/emoji*${MODEL##*/}"
 python model_finetuning_emoji.py -m "${MODEL}" -d "emoji_random3_seed1"
-rm -rf "ckpt/emoji*${MODEL##*/}"
 python model_finetuning_emoji.py -m "${MODEL}" -d "emoji_random0_seed2"
-rm -rf "ckpt/emoji*${MODEL##*/}"
 python model_finetuning_emoji.py -m "${MODEL}" -d "emoji_random1_seed2"
-rm -rf "ckpt/emoji*${MODEL##*/}"
 python model_finetuning_emoji.py -m "${MODEL}" -d "emoji_random2_seed2"
-rm -rf "ckpt/emoji*${MODEL##*/}"
 python model_finetuning_emoji.py -m "${MODEL}" -d "emoji_random3_seed2"
-rm -rf "ckpt/emoji*${MODEL##*/}"
 
 
 # HATE
 python model_finetuning_hate.py -m "${MODEL}" -d "hate_temporal"
-rm -rf "ckpt/hate*${MODEL##*/}"
 python model_finetuning_hate.py -m "${MODEL}" -d "hate_random0_seed0"
-rm -rf "ckpt/hate*${MODEL##*/}"
 python model_finetuning_hate.py -m "${MODEL}" -d "hate_random1_seed0"
-rm -rf "ckpt/hate*${MODEL##*/}"
 python model_finetuning_hate.py -m "${MODEL}" -d "hate_random2_seed0"
-rm -rf "ckpt/hate*${MODEL##*/}"
 python model_finetuning_hate.py -m "${MODEL}" -d "hate_random3_seed0"
-rm -rf "ckpt/hate*${MODEL##*/}"
 python model_finetuning_hate.py -m "${MODEL}" -d "hate_random0_seed1"
-rm -rf "ckpt/hate*${MODEL##*/}"
 python model_finetuning_hate.py -m "${MODEL}" -d "hate_random1_seed1"
-rm -rf "ckpt/hate*${MODEL##*/}"
 python model_finetuning_hate.py -m "${MODEL}" -d "hate_random2_seed1"
-rm -rf "ckpt/hate*${MODEL##*/}"
 python model_finetuning_hate.py -m "${MODEL}" -d "hate_random3_seed1"
-rm -rf "ckpt/hate*${MODEL##*/}"
 python model_finetuning_hate.py -m "${MODEL}" -d "hate_random0_seed2"
-rm -rf "ckpt/hate*${MODEL##*/}"
 python model_finetuning_hate.py -m "${MODEL}" -d "hate_random1_seed2"
-rm -rf "ckpt/hate*${MODEL##*/}"
 python model_finetuning_hate.py -m "${MODEL}" -d "hate_random2_seed2"
-rm -rf "ckpt/hate*${MODEL##*/}"
 python model_finetuning_hate.py -m "${MODEL}" -d "hate_random3_seed2"
-rm -rf "ckpt/hate*${MODEL##*/}"
 
 
 # SENTIMENT

experiments/model_predict_classifier.py

@@ -140,6 +140,44 @@ class SentimentClassification(Classifier):
                 f.write("\n".join([json.dumps(i) for i in predictions]))
 
 
+class HateClassification(Classifier):
+
+    id_to_label = {'0': '0', '1': '1'}
+
+    def __init__(self, model_name: str):
+        super().__init__(model_name, max_length=128, multi_label=False, id_to_label=self.id_to_label)
+        self.dataset = load_dataset("tweettemposhift/tweet_temporal_shift", "hate_temporal")
+
+    def get_prediction(self, export_dir: str, batch_size: int):
+        os.makedirs(export_dir, exist_ok=True)
+        for test_split in ["test_1", "test_2", "test_3", "test_4"]:
+            if os.path.exists(f"{export_dir}/{test_split}.jsonl"):
+                continue
+            data = self.dataset[test_split]
+            predictions = self.predict(data["text"], batch_size)
+            with open(f"{export_dir}/{test_split}.jsonl", "w") as f:
+                f.write("\n".join([json.dumps(i) for i in predictions]))
+
+
+class EmojiClassification(Classifier):
+
+    def __init__(self, model_name: str):
+        self.dataset = load_dataset("tweettemposhift/tweet_temporal_shift", "hate_temporal")
+        id_to_label = dict(enumerate(self.dataset["test"].features["gold_label"].names))
+        super().__init__(model_name, max_length=128, multi_label=False, id_to_label=id_to_label)
+
+    def get_prediction(self, export_dir: str, batch_size: int):
+        os.makedirs(export_dir, exist_ok=True)
+        for test_split in ["test_1", "test_2", "test_3", "test_4"]:
+            if os.path.exists(f"{export_dir}/{test_split}.jsonl"):
+                continue
+            data = self.dataset[test_split]
+            predictions = self.predict(data["text"], batch_size)
+            with open(f"{export_dir}/{test_split}.jsonl", "w") as f:
+                f.write("\n".join([json.dumps(i) for i in predictions]))
+
+
+
 class NERDClassification(Classifier):
 
     id_to_label = {'0': '0', '1': '1'}
@@ -178,30 +216,50 @@ if __name__ == '__main__':
     ]
     for model_m in model_list:
         alias = f"tweettemposhift/topic-topic_temporal-{model_m}"
-        TopicClassification(alias).get_prediction(export_dir=f"prediction_files/{os.path.basename(alias)}", batch_size=512)
+        HateClassification(alias).get_prediction(export_dir=f"prediction_files/{os.path.basename(alias)}", batch_size=512)
         torch.cuda.empty_cache()
         for random_r in range(4):
             for seed_s in range(3):
-                alias = f"tweettemposhift/topic-topic_random{random_r}_seed{seed_s}-{model_m}"
+                alias = f"tweettemposhift/hate-hate_random{random_r}_seed{seed_s}-{model_m}"
                 TopicClassification(alias).get_prediction(export_dir=f"prediction_files/{os.path.basename(alias)}", batch_size=512)
                 torch.cuda.empty_cache()
 
     for model_m in model_list:
-        alias = f"tweettemposhift/sentiment-sentiment_small_temporal-{model_m}"
-        SentimentClassification(alias).get_prediction(export_dir=f"prediction_files/{os.path.basename(alias)}", batch_size=512)
+        alias = f"tweettemposhift/emoji-emoji_temporal-{model_m}"
+        EmojiClassification(alias).get_prediction(export_dir=f"prediction_files/{os.path.basename(alias)}", batch_size=512)
         torch.cuda.empty_cache()
         for random_r in range(4):
             for seed_s in range(3):
-                alias = f"tweettemposhift/sentiment-sentiment_small_random{random_r}_seed{seed_s}-{model_m}"
-                SentimentClassification(alias).get_prediction(export_dir=f"prediction_files/{os.path.basename(alias)}", batch_size=512)
-                torch.cuda.empty_cache()
-
-    for model_m in model_list:
-        alias = f"tweettemposhift/nerd-nerd_temporal-{model_m}"
-        NERDClassification(alias).get_prediction(export_dir=f"prediction_files/{os.path.basename(alias)}", batch_size=512)
-        torch.cuda.empty_cache()
-        for random_r in range(4):
-            for seed_s in range(3):
-                alias = f"tweettemposhift/nerd-nerd_random{random_r}_seed{seed_s}-{model_m}"
-                NERDClassification(alias).get_prediction(export_dir=f"prediction_files/{os.path.basename(alias)}", batch_size=512)
+                alias = f"tweettemposhift/emoji-emoji_random{random_r}_seed{seed_s}-{model_m}"
+                TopicClassification(alias).get_prediction(export_dir=f"prediction_files/{os.path.basename(alias)}", batch_size=512)
                 torch.cuda.empty_cache()
+    #
+    # for model_m in model_list:
+    #     alias = f"tweettemposhift/topic-topic_temporal-{model_m}"
+    #     TopicClassification(alias).get_prediction(export_dir=f"prediction_files/{os.path.basename(alias)}", batch_size=512)
+    #     torch.cuda.empty_cache()
+    #     for random_r in range(4):
+    #         for seed_s in range(3):
+    #             alias = f"tweettemposhift/topic-topic_random{random_r}_seed{seed_s}-{model_m}"
+    #             TopicClassification(alias).get_prediction(export_dir=f"prediction_files/{os.path.basename(alias)}", batch_size=512)
+    #             torch.cuda.empty_cache()
+    #
+    # for model_m in model_list:
+    #     alias = f"tweettemposhift/sentiment-sentiment_small_temporal-{model_m}"
+    #     SentimentClassification(alias).get_prediction(export_dir=f"prediction_files/{os.path.basename(alias)}", batch_size=512)
+    #     torch.cuda.empty_cache()
+    #     for random_r in range(4):
+    #         for seed_s in range(3):
+    #             alias = f"tweettemposhift/sentiment-sentiment_small_random{random_r}_seed{seed_s}-{model_m}"
+    #             SentimentClassification(alias).get_prediction(export_dir=f"prediction_files/{os.path.basename(alias)}", batch_size=512)
+    #             torch.cuda.empty_cache()
+    #
+    # for model_m in model_list:
+    #     alias = f"tweettemposhift/nerd-nerd_temporal-{model_m}"
+    #     NERDClassification(alias).get_prediction(export_dir=f"prediction_files/{os.path.basename(alias)}", batch_size=512)
+    #     torch.cuda.empty_cache()
+    #     for random_r in range(4):
+    #         for seed_s in range(3):
+    #             alias = f"tweettemposhift/nerd-nerd_random{random_r}_seed{seed_s}-{model_m}"
+    #             NERDClassification(alias).get_prediction(export_dir=f"prediction_files/{os.path.basename(alias)}", batch_size=512)
+    #             torch.cuda.empty_cache()