A

app.py

@@ -8,21 +8,16 @@ import os
 
 
 # Initialize the model pipeline
-generator = pipeline('text-generation', model='hydra-project/ChatHercules-2.5-Mistral-7B', torch_dtype=torch.bfloat16)
+generator = pipeline('text-generation', model='okeanos/uptimeai-8273')
 @spaces.GPU
-def generate_text(prompt, temperature, top_p, top_k, repetition_penalty, max_length):
+def generate_text(prompt):
     # Generate text using the model
     generator.model.cuda()
     generator.device = torch.device("cuda")
     prompt = f"<|im_start|>user\n{prompt}<|im_end|>\n<|im_start|>assistant\n"
     outputs = generator(
         prompt, 
-        do_sample=True,
-        max_new_tokens=max_length,
-        temperature=temperature,
-        top_p=top_p,
-        top_k=top_k,
-        repetition_penalty=repetition_penalty,
+        do_sample=True
         return_full_text=False
     )
     # Extract the generated text and return it