Update app.py

app.py

@@ -1,44 +1,30 @@
 import gradio as gr
-from unsloth import FastLanguageModel
-import torch 
-from transformers import TextStreamer
+from groq import Groq
 
-model, tokenizer = FastLanguageModel.from_pretrained(
-    model_name = "unsloth/llama-2-7b-bnb-4bit", 
-    max_seq_length = 2048,
-    dtype = torch.float16,
-    load_in_4bit = True,
-).to("cpu")
-FastLanguageModel.for_inference(model) 
 
-def predict(input_text):
-    messages = [
-        {"from": "human", "value": input_text},
-    ]
-    inputs = tokenizer.apply_chat_template(
-        messages,
-        tokenize=True,
-        add_generation_prompt=True,  
-        return_tensors="pt",
-    ).to("cpu") 
+client = Groq(
+    api_key=("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"),
+)
+
 
-    outputs = model.generate(
-        input_ids=inputs,
-        max_new_tokens=128,
-        use_cache=True
+def generate_response(input_text):
+    chat_completion = client.chat.completions.create(
+        messages=[
+            {
+                "role": "user",
+                "content": input_text,
+            }
+        ],
+        model="llama3-8b-8192",
     )
-    
-    decoded_output = tokenizer.batch_decode(outputs, skip_special_tokens=True)
-    assistant_response = decoded_output[0].split('assistant\n')[-1].strip()
-    
-    return assistant_response
+    return chat_completion.choices[0].message.content
 
 iface = gr.Interface(
-    fn=predict,
+    fn=generate_response,
     inputs="text",
     outputs="text",
-    title="Parviz(eng) Chatbot",
-    description="LLM Test model .",
+    title="Parviz Chatbot",
+    description="ye chi bepors",
 )
 
 iface.launch()