Update app.py

app.py

@@ -1,15 +1,49 @@
 import gradio as gr
-from transformers import pipeline
+import torch
+from transformers import AutoModelForCausalLM, AutoTokenizer, StoppingCriteria, StoppingCriteriaList, TextIteratorStreamer
+from threading import Thread
 
-# Hugging Face에서 모델 로드
-qa_model = pipeline("question-answering", model="RAIJAY/7B_QA_68348")
+tokenizer = AutoTokenizer.from_pretrained("RAIJAY/7B_QA_68348")
+model = AutoModelForCausalLM.from_pretrained("RAIJAY/7B_QA_68348", torch_dtype=torch.float16)
+model = model.to('cuda:0')
 
-# Gradio 인터페이스 정의
-def question_answering(question):
-    answer = qa_model(question)["answer"]
-    return answer
+class StopOnTokens(StoppingCriteria):
+    def __call__(self, input_ids: torch.LongTensor, scores: torch.FloatTensor, **kwargs) -> bool:
+        stop_ids = [29, 0]
+        for stop_id in stop_ids:
+            if input_ids[0][-1] == stop_id:
+                return True
+        return False
 
-iface = gr.Interface(fn=question_answering, inputs="text", outputs="text")
+def predict(message, history):
 
-# Gradio 앱 실행
-iface.launch()
+    history_transformer_format = history + [[message, ""]]
+    stop = StopOnTokens()
+
+    messages = "".join(["".join(["\n<human>:"+item[0], "\n<bot>:"+item[1]])  #curr_system_message +
+                for item in history_transformer_format])
+
+    model_inputs = tokenizer([messages], return_tensors="pt").to("cuda")
+    streamer = TextIteratorStreamer(tokenizer, timeout=10., skip_prompt=True, skip_special_tokens=True)
+    generate_kwargs = dict(
+        model_inputs,
+        streamer=streamer,
+        max_new_tokens=1024,
+        do_sample=True,
+        top_p=0.95,
+        top_k=1000,
+        temperature=1.0,
+        num_beams=1,
+        stopping_criteria=StoppingCriteriaList([stop])
+        )
+    t = Thread(target=model.generate, kwargs=generate_kwargs)
+    t.start()
+
+    partial_message  = ""
+    for new_token in streamer:
+        if new_token != '<':
+            partial_message += new_token
+            yield partial_message
+
+
+gr.ChatInterface(predict).queue().launch()