Update app.py

app.py

@@ -9,11 +9,43 @@ tokenizer = AutoTokenizer.from_pretrained(
 model = AutoModelForSeq2SeqLM.from_pretrained(
     "Kaludi/chatgpt-gpt4-prompts-bart-large-cnn-samsum", from_tf=True)
 
+# zephyr
+    pipe = pipeline("text-generation", model="HuggingFaceH4/zephyr-7b-alpha",torch_dtype=torch.bfloat16, device_map="auto")
 
-def generatePrompt(inputuno, inputdos):
+def generate(inputuno, inputdos, max_new_tokens=3556, top_p=0.95, repetition_penalty=1.0):
+    top_p = float(top_p)
 
-    # zephyr
-    pipe = pipeline("text-generation", model="HuggingFaceH4/zephyr-7b-alpha",torch_dtype=torch.bfloat16, device_map="auto")
+    prompt = inputuno
+    promptdos = inputdos
+    
+    generate_kwargs = dict(
+        temperature=fixed_temperature,
+        max_new_tokens=max_new_tokens,
+        top_p=top_p,
+        repetition_penalty=repetition_penalty,
+        do_sample=True,
+        seed=42,
+    )
+
+    messages = [
+        {
+            "role": "system", "content": str(new_prompt)
+        },
+        {
+            "role": "user", "content": str(promptdos)
+        },
+    ]
+
+    stream = pipe.tokenizer.apply_chat_template(messages, **generate_kwargs, stream=True, details=True, return_full_text=False)
+    output = ""
+
+    for response in stream:
+        output += response.token.text
+        yield output
+    return output
+
+
+def generatePrompt(inputuno, inputdos):
     
     prompt = inputuno
     promptdos = inputdos
@@ -49,6 +81,6 @@ examples = [["photographer"], ["developer"], ["teacher"], [
     "human resources staff"], ["recipe for ham croquettes"]]
 description = ""
 
-PerfectGPT = gr.Interface(generatePrompt, inputs=[input_prompt, input_promptdos], outputs=output_component, examples=examples, title="🗿 PerfectGPT v1 🗿", description=description)
+PerfectGPT = gr.Interface(generate, inputs=[input_prompt, input_promptdos], outputs=output_component, examples=examples, title="🗿 PerfectGPT v1 🗿", description=description)
 
 PerfectGPT.launch()