feat: 换成测试模型

app.py

@@ -1,11 +1,14 @@
 import spaces
 import gradio as gr
 from huggingface_hub import InferenceClient
+from transformers import AutoTokenizer, AutoModelForCausalLM
+import torch
+import subprocess
 
-"""
-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")
+subprocess.run('pip install flash-attn --no-build-isolation', env={'FLASH_ATTENTION_SKIP_CUDA_BUILD': "TRUE"}, shell=True)
+
+tokenizer = AutoTokenizer.from_pretrained("deepseek-ai/DeepSeek-Coder-V2-Lite-Instruct", trust_remote_code=True)
+model = AutoModelForCausalLM.from_pretrained("deepseek-ai/DeepSeek-Coder-V2-Lite-Instruct", trust_remote_code=True, torch_dtype=torch.bfloat16).cuda()
 
 @spaces.GPU
 def respond(
@@ -16,35 +19,73 @@ def respond(
     temperature,
     top_p,
 ):
-    messages = [{"role": "system", "content": system_message}]
+    if len(message) < 1:
+        message = "write a quick sort algorithm in python."
+
+    messages = [
+        { 'role': 'user', 'content': message }
+    ]
 
-    for val in history:
-        if val[0]:
-            messages.append({"role": "user", "content": val[0]})
-        if val[1]:
-            messages.append({"role": "assistant", "content": val[1]})
+    inputs = tokenizer.apply_chat_template(messages, add_generation_prompt=True, return_tensors="pt").to(model.device)
 
-    messages.append({"role": "user", "content": message})
+    outputs = model.generate(inputs, max_new_tokens=max_tokens, do_sample=True, temperature=temperature, top_k=50, top_p=top_p, num_return_sequences=1, eos_token_id=tokenizer.eos_token_id)
 
-    response = ""
+    return tokenizer.decode(outputs[0][len(inputs[0]):], skip_special_tokens=True)
+
+"""
+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")
 
-    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 respond(
+#     message,
+#     history: list[tuple[str, str]],
+#     system_message,
+#     max_tokens,
+#     temperature,
+#     top_p,
+# ):
+#     messages = [{"role": "system", "content": system_message}]
 
-        response += token
-        yield response
+#     for val in history:
+#         if val[0]:
+#             messages.append({"role": "user", "content": val[0]})
+#         if val[1]:
+#             messages.append({"role": "assistant", "content": val[1]})
 
+#     if len(message) < 1:
+#         message = "write a quick sort algorithm in python."
+
+#     messages.append({"role": "user", "content": message})
+
+#     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
+
+#         response += token
+#         yield response
+
+"""
+For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/main/docs/gradio/chatinterface
 """
-For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
+
+css = """
+#msg_input {
+    flex-grow: 7;
+}
 """
+
 demo = gr.ChatInterface(
-    respond,
+    fn=respond,
+    textbox=gr.Textbox(elem_id="msg_input", placeholder="write a quick sort algorithm in python."),
     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"),
@@ -57,6 +98,7 @@ demo = gr.ChatInterface(
             label="Top-p (nucleus sampling)",
         ),
     ],
+    css=css,
 )
 
 

requirements.txt

@@ -1 +1,2 @@
-huggingface_hub==0.22.2
+huggingface_hub==0.22.2
+transformers