init

experiments/model_finetuning_nerd.py

@@ -1,20 +1,23 @@
-"""
-python model_finetuning_topic.py -m "roberta-base" -d "topic_temporal"
-
-python model_finetuning_topic.py -m "roberta-base" -d "topic_random0_seed0"
-python model_finetuning_topic.py -m "roberta-base" -d "topic_random1_seed0"
-python model_finetuning_topic.py -m "roberta-base" -d "topic_random2_seed0"
-python model_finetuning_topic.py -m "roberta-base" -d "topic_random3_seed0"
-
-python model_finetuning_topic.py -m "roberta-base" -d "topic_random0_seed1"
-python model_finetuning_topic.py -m "roberta-base" -d "topic_random1_seed1"
-python model_finetuning_topic.py -m "roberta-base" -d "topic_random2_seed1"
-python model_finetuning_topic.py -m "roberta-base" -d "topic_random3_seed1"
-
-python model_finetuning_topic.py -m "roberta-base" -d "topic_random0_seed2"
-python model_finetuning_topic.py -m "roberta-base" -d "topic_random1_seed2"
-python model_finetuning_topic.py -m "roberta-base" -d "topic_random2_seed2"
-python model_finetuning_topic.py -m "roberta-base" -d "topic_random3_seed2"
+"""Experiment.
+
+```
+python model_finetuning_nerd.py -m "roberta-base" -d "nerd_temporal"
+
+python model_finetuning_nerd.py -m "roberta-base" -d "nerd_random0_seed0"
+python model_finetuning_nerd.py -m "roberta-base" -d "nerd_random1_seed0"
+python model_finetuning_nerd.py -m "roberta-base" -d "nerd_random2_seed0"
+python model_finetuning_nerd.py -m "roberta-base" -d "nerd_random3_seed0"
+
+python model_finetuning_nerd.py -m "roberta-base" -d "nerd_random0_seed1"
+python model_finetuning_nerd.py -m "roberta-base" -d "nerd_random1_seed1"
+python model_finetuning_nerd.py -m "roberta-base" -d "nerd_random2_seed1"
+python model_finetuning_nerd.py -m "roberta-base" -d "nerd_random3_seed1"
+
+python model_finetuning_nerd.py -m "roberta-base" -d "nerd_random0_seed2"
+python model_finetuning_nerd.py -m "roberta-base" -d "nerd_random1_seed2"
+python model_finetuning_nerd.py -m "roberta-base" -d "nerd_random2_seed2"
+python model_finetuning_nerd.py -m "roberta-base" -d "nerd_random3_seed2"
+```
 """
 import argparse
 import json
@@ -26,34 +29,12 @@ from shutil import copyfile
 from glob import glob
 
 import numpy as np
-from datasets import load_dataset, load_metric
+import evaluate
+from datasets import load_dataset
 from transformers import AutoTokenizer, AutoModelForSequenceClassification, TrainingArguments, Trainer
 from huggingface_hub import Repository
 
 logging.basicConfig(format="%(asctime)s %(levelname)-8s %(message)s", level=logging.INFO, datefmt="%Y-%m-%d %H:%M:%S")
-
-LABEL2ID = {
-    "arts_&_culture": 0,
-    "business_&_entrepreneurs": 1,
-    "celebrity_&_pop_culture": 2,
-    "diaries_&_daily_life": 3,
-    "family": 4,
-    "fashion_&_style": 5,
-    "film_tv_&_video": 6,
-    "fitness_&_health": 7,
-    "food_&_dining": 8,
-    "gaming": 9,
-    "learning_&_educational": 10,
-    "music": 11,
-    "news_&_social_concern": 12,
-    "other_hobbies": 13,
-    "relationships": 14,
-    "science_&_technology": 15,
-    "sports": 16,
-    "travel_&_adventure": 17,
-    "youth_&_student_life": 18
-}
-ID2LABEL = {v: k for k, v in LABEL2ID.items()}
 EVAL_STEP = 500
 RANDOM_SEED = 42
 N_TRIALS = 10
@@ -65,45 +46,49 @@ def sigmoid(x):
 
 def main(
         dataset: str = "tweettemposhift/tweet_temporal_shift",
-        dataset_type: str = "topic_temporal",
+        dataset_type: str = "nerd_temporal",
         model: str = "roberta-base",
         skip_train: bool = False,
         skip_test: bool = False,
         skip_upload: bool = False):
 
+    model_alias = f"nerd-{dataset_type}-{os.path.basename(model)}"
+    output_dir = f"ckpt/{model_alias}"
+    best_model_path = pj(output_dir, "best_model")
+
     tokenizer = AutoTokenizer.from_pretrained(model)
-    model = AutoModelForSequenceClassification.from_pretrained(
-        model, id2label=ID2LABEL, label2id=LABEL2ID, num_labels=len(LABEL2ID), problem_type="multi_label_classification"
-    )
     dataset = load_dataset(dataset, dataset_type)
     tokenized_datasets = dataset.map(
-        lambda x: tokenizer(x["text"], padding="max_length", truncation=True, max_length=256), batched=True
+        lambda x: tokenizer(
+            [f"[target] {a}, [definition] {b}, [text] {c}" for a, b, c in zip(x["target"], x["definition"], x["text"])],
+            padding="max_length",
+            truncation=True,
+            max_length=256),
+        batched=True
     )
-    metric_accuracy = load_metric("accuracy", "multilabel")
-    metric_f1 = load_metric("f1", "multilabel")
+    tokenized_datasets = tokenized_datasets.rename_column("gold_label_binary", "label")
+    metric_accuracy = evaluate.load("accuracy")
+    metric_f1 = evaluate.load("f1")
 
     def compute_metric_search(eval_pred):
         logits, labels = eval_pred
-        predictions = np.array([[int(sigmoid(j) > 0.5) for j in i] for i in logits])
-        return metric_f1.compute(predictions=predictions, references=labels, average="micro")
+        predictions = np.argmax(logits, axis=-1)
+        return metric_accuracy.compute(predictions=predictions, references=labels, average="micro")
 
     def compute_metric_all(eval_pred):
         logits, labels = eval_pred
-        predictions = np.array([[int(sigmoid(j) > 0.5) for j in i] for i in logits])
+        predictions = np.argmax(logits, axis=-1)
         return {
             "f1": metric_f1.compute(predictions=predictions, references=labels, average="micro")["f1"],
             "f1_macro": metric_f1.compute(predictions=predictions, references=labels, average="macro")["f1"],
             "accuracy": metric_accuracy.compute(predictions=predictions, references=labels)["accuracy"]
         }
 
-    model_alias = f"topic-{dataset_type}-{os.path.basename(model)}"
-    output_dir = f"ckpt/{model_alias}"
-    best_model_path = pj(output_dir, "best_model")
 
     if not skip_train:
         logging.info("training model")
         trainer = Trainer(
-            model=model,
+            model=AutoModelForSequenceClassification.from_pretrained(model, num_labels=2),
             args=TrainingArguments(
                 output_dir=output_dir,
                 evaluation_strategy="steps",
@@ -114,11 +99,7 @@ def main(
             eval_dataset=tokenized_datasets["validation"],
             compute_metrics=compute_metric_search,
             model_init=lambda x: AutoModelForSequenceClassification.from_pretrained(
-                model,
-                return_dict=True,
-                num_labels=len(LABEL2ID),
-                id2label=ID2LABEL,
-                label2id=LABEL2ID
+                model, return_dict=True, num_labels=2,
             )
         )
 
@@ -129,10 +110,9 @@ def main(
                     "per_device_train_batch_size", [8, 16, 32]
                 ),
             },
-            local_dir="./hyperparameter_search_cache",
             direction="maximize",
             backend="optuna",
-            N_TRIALS=N_TRIALS
+            n_trials=N_TRIALS
         )
         for n, v in best_run.hyperparameters.items():
             setattr(trainer.args, n, v)
@@ -151,15 +131,8 @@ def main(
         else:
             metric = {}
         for single_test in test_split:
-            model = AutoModelForSequenceClassification.from_pretrained(
-                best_model_path,
-                num_labels=len(LABEL2ID),
-                problem_type="multi_label_classification",
-                id2label=ID2LABEL,
-                label2id=LABEL2ID
-            )
             trainer = Trainer(
-                model=model,
+                model=AutoModelForSequenceClassification.from_pretrained(best_model_path, num_labels=2),
                 args=TrainingArguments(
                     output_dir=output_dir,
                     evaluation_strategy="no",
@@ -177,15 +150,9 @@ def main(
     if not skip_upload:
         logging.info("uploading to huggingface")
         model_organization = "tweettemposhift"
-        model = AutoModelForSequenceClassification.from_pretrained(
-            best_model_path,
-            num_labels=len(LABEL2ID),
-            problem_type="multi_label_classification",
-            id2label=ID2LABEL,
-            label2id=LABEL2ID
-        )
+        model_instance = AutoModelForSequenceClassification.from_pretrained(best_model_path, num_labels=2)
         tokenizer = AutoTokenizer.from_pretrained(best_model_path)
-        model.push_to_hub(f"{model_organization}/{model_alias}", use_auth_token=True)
+        model_instance.push_to_hub(f"{model_organization}/{model_alias}", use_auth_token=True)
         tokenizer.push_to_hub(f"{model_organization}/{model_alias}", use_auth_token=True)
         repo = Repository(model_alias, f"{model_organization}/{model_alias}")
         for i in glob(f"{best_model_path}/*"):
@@ -197,7 +164,7 @@ def main(
 if __name__ == "__main__":
     parser = argparse.ArgumentParser(description="Fine-tuning language model.")
     parser.add_argument("-m", "--model", help="transformer LM", default="roberta-base", type=str)
-    parser.add_argument("-d", "--dataset-type", help='dataset type', default="topic_temporal", type=str)
+    parser.add_argument("-d", "--dataset-type", help='dataset type', default="nerd_temporal", type=str)
     parser.add_argument("--skip-train", action="store_true")
     parser.add_argument("--skip-test", action="store_true")
     parser.add_argument("--skip-upload", action="store_true")

experiments/model_finetuning_topic.py

@@ -1,4 +1,6 @@
-"""
+"""Experiment.
+
+```
 python model_finetuning_topic.py -m "roberta-base" -d "topic_temporal"
 
 python model_finetuning_topic.py -m "roberta-base" -d "topic_random0_seed0"
@@ -15,6 +17,7 @@ python model_finetuning_topic.py -m "roberta-base" -d "topic_random0_seed2"
 python model_finetuning_topic.py -m "roberta-base" -d "topic_random1_seed2"
 python model_finetuning_topic.py -m "roberta-base" -d "topic_random2_seed2"
 python model_finetuning_topic.py -m "roberta-base" -d "topic_random3_seed2"
+```
 """
 import argparse
 import json
@@ -88,12 +91,12 @@ def main(
 
     def compute_metric_search(eval_pred):
         logits, labels = eval_pred
-        predictions = np.array([[int(sigmoid(j) > 0.5) for j in i] for i in logits])
+        predictions = np.array([[int(sigmoid(j) > 0.5) for j in lo] for lo in logits])
         return metric_f1.compute(predictions=predictions, references=labels, average="micro")
 
     def compute_metric_all(eval_pred):
         logits, labels = eval_pred
-        predictions = np.array([[int(sigmoid(j) > 0.5) for j in i] for i in logits])
+        predictions = np.array([[int(sigmoid(j) > 0.5) for j in lo] for lo in logits])
         return {
             "f1": metric_f1.compute(predictions=predictions, references=labels, average="micro")["f1"],
             "f1_macro": metric_f1.compute(predictions=predictions, references=labels, average="macro")["f1"],

experiments/requirements.txt

@@ -1,8 +1,8 @@
-torch
-scikit-learn
-numpy
-datasets
-transformers
-huggingface_hub
-optuna
-accelerate
+torch==2.1.0
+scikit-learn==1.3.1
+numpy==1.24.4
+datasets==2.14.5
+transformers==4.34.1
+huggingface_hub==0.17.3
+optuna==3.4.0
+accelerate==0.4.1