streamer = TextStreamer(tokenizer, skip_prompt=True, skip_special_tokens=True)
# Convert prompt to tokens
prompt_template = """<|im_start|>system
Du bist ein hilfreicher Assistent.<|im_end|>
<|im_start|>user
{prompt}<|im_end|>
<|im_start|>assistant"""

prompt = "Schreibe eine Stellenanzeige für Data Scientist bei AXA!"

final_prompt = prompt_template.format(prompt=prompt)

import argparse
import torch
from openmind import pipeline, is_torch_npu_available, AutoModelForCausalLM, AutoTokenizer

# Check if NPU or CPU is available
if is_torch_npu_available():
  device = "npu:0"
  print("Using NPU for inference")
else:
  device = "cpu"
  print("Using CPU for inference")

# Load the model and tokenizer
model = AutoModelForCausalLM.from_pretrained("SY_AICC/germeo-7b-laser", trust_remote_code=True)
tokenizer = AutoTokenizer.from_pretrained(model_path, trust_remote_code=True)

# Define the pipeline with the correct task
generate_text = pipeline(
  task="text-generation",  # Specify the task
  model=model,
  tokenizer=tokenizer,
  torch_dtype=torch.bfloat16 if device != "cpu" else torch.float32,  # Use bfloat16 if not on CPU
  device=device
)