Initial commit

app.py

@@ -1,70 +1,39 @@
 import gradio as gr
-from huggingface_hub import InferenceClient
+import torch
+from transformers import AutoModelForCausalLM, AutoTokenizer
+from peft import PeftModel
+from modelscope.hub.snapshot_download import snapshot_download
 
+# 1. 确定您的基础模型 (Base Model)
+BASE_MODEL_ID = "seeklhy/OmniSQL-7B"
 
-def respond(
-    message,
-    history: list[dict[str, str]],
-    system_message,
-    max_tokens,
-    temperature,
-    top_p,
-    hf_token: gr.OAuthToken,
-):
-    """
-    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(token=hf_token.token, model="openai/gpt-oss-20b")
+# 2. 您的 LoRA 模型 ID
+LORA_MODEL_ID = "risemds/UniVectorSQL-7B-LoRA-all_steps_1030_new_data"
 
-    messages = [{"role": "system", "content": system_message}]
+# 3. 下载 ModelScope LoRA 权重
+# (您可能需要先登录 modelscope: `from modelscope.hub.api import HubApi; api = HubApi(); api.login('YOUR_TOKEN')`)
+lora_path = snapshot_download(LORA_MODEL_ID, revision='master') # 确保使用正确的 revision
 
-    messages.extend(history)
+# 4. 加载基础模型和 Tokenizer
+tokenizer = AutoTokenizer.from_pretrained(BASE_MODEL_ID, trust_remote_code=True)
+model = AutoModelForCausalLM.from_pretrained(BASE_MODEL_ID, device_map="auto", torch_dtype=torch.float16, trust_remote_code=True)
 
-    messages.append({"role": "user", "content": message})
+# 5. 加载并融合 LoRA 适配器
+# PeftModel 会自动将 LoRA 权重加载到基础模型上
+model = PeftModel.from_pretrained(model, lora_path)
 
-    response = ""
+# (可选) 如果需要，可以合并权重以加快推理
+# model = model.merge_and_unload() 
 
-    for message in client.chat_completion(
-        messages,
-        max_tokens=max_tokens,
-        stream=True,
-        temperature=temperature,
-        top_p=top_p,
-    ):
-        choices = message.choices
-        token = ""
-        if len(choices) and choices[0].delta.content:
-            token = choices[0].delta.content
+model.eval()
 
-        response += token
-        yield response
+# 6. 定义推理函数
+def inference(text_input):
+    inputs = tokenizer(text_input, return_tensors="pt").to("cuda")
+    outputs = model.generate(**inputs, max_new_tokens=100)
+    result = tokenizer.decode(outputs[0], skip_special_tokens=True)
+    return result
 
-
-"""
-For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
-"""
-chatbot = gr.ChatInterface(
-    respond,
-    type="messages",
-    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)",
-        ),
-    ],
-)
-
-with gr.Blocks() as demo:
-    with gr.Sidebar():
-        gr.LoginButton()
-    chatbot.render()
-
-
-if __name__ == "__main__":
-    demo.launch()
+# 7. 创建 Gradio 界面
+iface = gr.Interface(fn=inference, inputs="text", outputs="text")
+iface.launch()

requirements.txt

@@ -0,0 +1,5 @@
+torch
+transformers
+peft
+modelscope
+gradio