Update app.py

app.py

@@ -1,64 +1,85 @@
-import gradio as gr
-from huggingface_hub import InferenceClient
+from threading import Thread
 
-"""
-For more information on `huggingface_hub` Inference API support, please check the docs: https://huggingface.co/docs/huggingface_hub/v0.22.2/en/guides/inference
-"""
-client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")
+import gradio as gr
+import spaces
+import torch
+from transformers import AutoModelForCausalLM, AutoTokenizer, TextIteratorStreamer
 
+MAX_NEW_TOKENS = 2048
+MODEL_NAME = "Azure99/Blossom-V6-7B"
 
-def respond(
-    message,
-    history: list[tuple[str, str]],
-    system_message,
-    max_tokens,
-    temperature,
-    top_p,
-):
-    messages = [{"role": "system", "content": system_message}]
+model = AutoModelForCausalLM.from_pretrained(MODEL_NAME, torch_dtype=torch.bfloat16, device_map="auto")
+tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)
 
-    for val in history:
-        if val[0]:
-            messages.append({"role": "user", "content": val[0]})
-        if val[1]:
-            messages.append({"role": "assistant", "content": val[1]})
 
-    messages.append({"role": "user", "content": message})
+def get_input_ids(inst, history):
+    conversation = []
+    for user, assistant in history:
+        conversation.extend([{"role": "user", "content": user}, {"role": "assistant", "content": assistant}])
+    conversation.append({"role": "user", "content": inst})
+    return tokenizer.apply_chat_template(conversation, return_tensors="pt").to(model.device)
 
-    response = ""
 
-    for message in client.chat_completion(
-        messages,
-        max_tokens=max_tokens,
-        stream=True,
-        temperature=temperature,
-        top_p=top_p,
-    ):
-        token = message.choices[0].delta.content
+@spaces.GPU
+def chat(inst, history, temperature, top_p, repetition_penalty):
+    streamer = TextIteratorStreamer(tokenizer, skip_prompt=True, skip_special_tokens=True)
+    input_ids = get_input_ids(inst, history)
+    generation_kwargs = dict(input_ids=input_ids,
+                             streamer=streamer, do_sample=True, max_new_tokens=MAX_NEW_TOKENS,
+                             temperature=temperature, top_p=top_p, repetition_penalty=repetition_penalty)
 
-        response += token
-        yield response
+    Thread(target=model.generate, kwargs=generation_kwargs).start()
 
+    outputs = ""
+    for new_text in streamer:
+        outputs += new_text
+        yield outputs
 
-"""
-For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
-"""
-demo = gr.ChatInterface(
-    respond,
-    additional_inputs=[
-        gr.Textbox(value="You are a friendly Chatbot.", label="System message"),
-        gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
-        gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
-        gr.Slider(
-            minimum=0.1,
-            maximum=1.0,
-            value=0.95,
-            step=0.05,
-            label="Top-p (nucleus sampling)",
-        ),
-    ],
-)
 
+additional_inputs = [
+    gr.Slider(
+        label="Temperature",
+        value=0.5,
+        minimum=0.0,
+        maximum=1.0,
+        step=0.05,
+        interactive=True,
+        info="Controls randomness in choosing words.",
+    ),
+    gr.Slider(
+        label="Top-P",
+        value=0.85,
+        minimum=0.0,
+        maximum=1.0,
+        step=0.05,
+        interactive=True,
+        info="Picks words until their combined probability is at least top_p.",
+    ),
+    gr.Slider(
+        label="Repetition penalty",
+        value=1.05,
+        minimum=1.0,
+        maximum=1.2,
+        step=0.01,
+        interactive=True,
+        info="Repetition Penalty: Controls how much repetition is penalized.",
+    )
+]
 
-if __name__ == "__main__":
-    demo.launch()
+gr.ChatInterface(chat,
+                 chatbot=gr.Chatbot(show_label=False, height=500, show_copy_button=True, render_markdown=True),
+                 textbox=gr.Textbox(placeholder="", container=False, scale=7),
+                 title="Blossom-V6-7B Demo",
+                 description='Hello, I am Blossom, an open source conversational large language model.🌠'
+                             '<a href="https://github.com/Azure99/BlossomLM">GitHub</a>',
+                 theme="soft",
+                 examples=[["Hello"], ["What is MBTI"], ["用Python实现二分查找"],
+                           ["为switch写一篇小红书种草文案，带上emoji"]],
+                 cache_examples=False,
+                 additional_inputs=additional_inputs,
+                 additional_inputs_accordion=gr.Accordion(label="Config", open=True),
+                 clear_btn="🗑️Clear",
+                 undo_btn="↩️Undo",
+                 retry_btn="🔄Retry",
+                 submit_btn="➡️Submit",
+                 ).queue().launch()