Update app.py

app.py

@@ -1,33 +1,34 @@
 import gradio as gr
-import torch
-from transformers import GPT2LMHeadModel, GPT2Tokenizer
+import tensorflow as tf
+from tensorflow.keras.preprocessing.sequence import pad_sequences
+import numpy as np
+import pickle
 
-class CustomGPT2LMHeadModel(GPT2LMHeadModel):
-    def __init__(self, config, **kwargs):
-        super().__init__(config, **kwargs)
-        self._prepare_inputs_for_inference()
+# 토크나이저 불러오기
+with open('tokenizer.pickle', 'rb') as handle:
+    tokenizer = pickle.load(handle)
 
-    @property
-    def _prepare_inputs_for_inference(self):
-        # Prepare inputs for inference with appropriate padding and eos_token_id
-        self.padding = "max_length"
-        self.use_cache = True
-        self.eos_token_id = self.config.eos_token_id if self.config.eos_token_id else 50256
-        return None
+# 텍스트 생성 모델 불러오기
+model = tf.keras.models.load_model("text_generation_model")
 
-def generate_text(prompt):
-    # Tokenizer initialization using custom GPT2LMHeadModel
-    tokenizer = GPT2Tokenizer.from_pretrained("gpt2", use_auth_token="your_huggingface_api_key")
-    model = CustomGPT2LMHeadModel.from_pretrained("gpt2", use_auth_token="your_huggingface_api_key")
+# 사용자 입력 받기
+user_input = input("텍스트 생성을 시작할 단어를 입력하세요: ")
 
-    # Input processing
-    input_ids = tokenizer.encode(f"{prompt}", return_tensors="pt")
+# 텍스트 생성 함수 정의
+def generate_text(seed_text, next_words, max_sequence_len):
+    for _ in range(next_words):
+        token_list = tokenizer.texts_to_sequences([seed_text])[0]
+        token_list = pad_sequences([token_list], maxlen=max_sequence_len-1, padding='pre')
+        predicted = model.predict_classes(token_list, verbose=0)
+        
+        output_word = ""
+        for word, index in tokenizer.word_index.items():
+            if index == predicted:
+                output_word = word
+                break
+        seed_text += " " + output_word
+    return seed_text
 
-    # Output generation
-    output = model.generate(input_ids, max_length=100, num_return_sequences=1)
-
-    # Decode and print the generated text
-    generated_text = tokenizer.decode(output[0], skip_special_tokens=True)
-    print(generated_text)
-
-print(generate_text(' '.join(input_words)))
+# 텍스트 생성
+generated_text = generate_text(user_input, next_words=50, max_sequence_len=25)
+print(generated_text)