pls work

app.py

@@ -1,20 +1,17 @@
 import gradio as gr
-from transformers import AutoTokenizer, AutoModelForCausalLM, AutoModelForSeq2SeqLM
+from huggingface_hub import InferenceClient
 from PyPDF2 import PdfReader
 
-# Models and Tokenizers Setup
+# Models Setup
 models = {
-    "Text Generator (Bloom)": {
-        "model": AutoModelForCausalLM.from_pretrained("bigscience/bloom-560m"),
-        "tokenizer": AutoTokenizer.from_pretrained("bigscience/bloom-560m"),
+    "Text Generator (Zephyr)": {
+        "client": InferenceClient(model="HuggingFaceH4/zephyr-7b-beta"),
     },
     "PDF Summarizer (T5)": {
-        "model": AutoModelForSeq2SeqLM.from_pretrained("aaliyaan/t5-small-finetuned-career"),
-        "tokenizer": AutoTokenizer.from_pretrained("aaliyaan/t5-small-finetuned-career"),
+        "client": InferenceClient(model="aaliyaan/t5-small-finetuned-career"),
     },
     "Broken Answer (T0pp)": {
-        "model": AutoModelForSeq2SeqLM.from_pretrained("bigscience/T0pp"),
-        "tokenizer": AutoTokenizer.from_pretrained("bigscience/T0pp"),
+        "client": InferenceClient(model="bigscience/T0pp"),
     },
 }
 
@@ -28,24 +25,25 @@ def chat_with_model(model_choice, user_message, chat_history, file=None):
         return chat_history
 
     model_info = models[model_choice]
-    tokenizer = model_info["tokenizer"]
-    model = model_info["model"]
-
-    # Tokenize Input
-    inputs = tokenizer(user_message, return_tensors="pt", padding=True, truncation=True, max_length=512)
-    
-    # Adjust max_length and parameters for the PDF summarizer
-    max_length = 150
-    num_beams = 5
-    outputs = model.generate(
-        **inputs, 
-        max_length=max_length, 
-        num_beams=num_beams, 
-        early_stopping=True, 
-        no_repeat_ngram_size=2
-    )
-    
-    response = tokenizer.decode(outputs[0], skip_special_tokens=True)
+    client = model_info["client"]
+
+    # Prepare messages for the InferenceClient
+    messages = [
+        {"role": "system", "content": "You are a helpful assistant."},
+        {"role": "user", "content": user_message}
+    ]
+
+    # Generate Response
+    response = ""
+    for message in client.chat_completion(
+        messages,
+        max_tokens=150,
+        stream=True,
+        temperature=0.7,
+        top_p=0.95
+    ):
+        token = message.choices[0].delta.content
+        response += token
 
     # Update Chat History
     chat_history.append((user_message, response))
@@ -79,7 +77,7 @@ def create_chat_interface():
         with gr.Row():
             model_choice = gr.Dropdown(
                 choices=list(models.keys()),
-                value="Text Generator (Bloom)",
+                value="Text Generator (Zephyr)",
                 label="Select Model"
             )