Update app.py

app.py

@@ -1,10 +1,9 @@
 import gradio as gr
-from huggingface_hub import InferenceClient
+from transformers import AutoTokenizer, AutoModelForCausalLM
+import torch
 
-"""
-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")
+tokenizer = AutoTokenizer.from_pretrained("Fastweb/FastwebMIIA-7B")
+model = AutoModelForCausalLM.from_pretrained("Fastweb/FastwebMIIA-7B")
 
 
 def respond(
@@ -25,19 +24,23 @@ def respond(
 
     messages.append({"role": "user", "content": message})
 
-    response = ""
+    # Format messages for the model
+    input_text = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
+    input_ids = tokenizer(input_text, return_tensors="pt").input_ids
 
-    for message in client.chat_completion(
-        messages,
-        max_tokens=max_tokens,
-        stream=True,
+    # Generate response
+    outputs = model.generate(
+        input_ids,
+        max_new_tokens=max_tokens,
         temperature=temperature,
         top_p=top_p,
-    ):
-        token = message.choices[0].delta.content
-
-        response += token
-        yield response
+        do_sample=True,
+        pad_token_id=tokenizer.eos_token_id
+    )
+    
+    # Decode the generated tokens, skipping the input tokens
+    response = tokenizer.decode(outputs[0][input_ids.shape[1]:], skip_special_tokens=True)
+    yield response
 
 
 """