Update app.py

app.py

@@ -1,63 +1,59 @@
 import gradio as gr
-from huggingface_hub import InferenceClient
+import transformers
+import torch
+from peft import PeftModel
+import os
 
-"""
-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")
+HF_TOKEN = os.environ.get("HF_TOKEN")
 
+model_id = "JerniganLab/qa-only"
+base_model = "meta-llama/Meta-Llama-3-8B-Instruct"
 
-def respond(
-    message,
-    history: list[tuple[str, str]],
-    system_message,
-    max_tokens,
-    temperature,
-    top_p,
-):
-    messages = [{"role": "system", "content": system_message}]
+llama_model = transformers.AutoModelForCausalLM.from_pretrained(base_model)
 
-    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})
+pipeline = transformers.pipeline(
+    "text-generation",
+    model=llama_model,
+    tokenizer=base_model,
+    model_kwargs={"torch_dtype": torch.bfloat16},
+    device="cuda",
+)
 
-    response = ""
+pipeline.model = PeftModel.from_pretrained(llama_model, model_id)
 
-    for message in client.chat_completion(
+def chat_function(message, history, system_prompt, max_new_tokens, temperature):
+    messages = [{"role":"system","content":system_prompt},
+                {"role":"user", "content":message}]
+    prompt = pipeline.tokenizer.apply_chat_template(
         messages,
-        max_tokens=max_tokens,
-        stream=True,
-        temperature=temperature,
-        top_p=top_p,
-    ):
-        token = message.choices[0].delta.content
-
-        response += token
-        yield response
-
+        tokenize=False,
+        add_generation_prompt=True,)
+    terminators = [
+        pipeline.tokenizer.eos_token_id,
+        pipeline.tokenizer.convert_tokens_to_ids("<|eot_id|>")]
+    outputs = pipeline(
+        prompt,
+        max_new_tokens = max_new_tokens,
+        eos_token_id = terminators,
+        do_sample = True,
+        temperature = temperature + 0.1,
+        top_p = 0.9,)
+    return outputs[0]["generated_text"][len(prompt):]
 
 """
 For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
 """
 demo = gr.ChatInterface(
-    respond,
+    chat_function,
+    textbox=gr.Textbox(placeholder="Enter message here", container=False, scale = 7),
+    chatbot=gr.Chatbot(height=400),
     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)",
-        ),
-    ],
-)
+        gr.Textbox("You are helpful AI", label="System Prompt"),
+        gr.Slider(500,4000, label="Max New Tokens"),
+        gr.Slider(0,1, label="Temperature")
+    ]
+    )
 
 
 if __name__ == "__main__":