File size: 3,033 Bytes
5d7d4e2 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 | #!/usr/bin/env python3
"""
Sixpert K1 - Vision/Multimodal Example
=======================================
Demonstrates how to use Sixpert K1's vision capabilities with
image inputs using llama-cpp-python.
Usage:
python vision_example.py --image ./photo.jpg --prompt "Describe this image in detail"
"""
import argparse
import base64
import sys
try:
from llama_cpp import Llama
except ImportError:
print("Installing llama-cpp-python...")
import subprocess
subprocess.check_call([sys.executable, "-m", "pip", "install", "llama-cpp-python"])
from llama_cpp import Llama
def image_to_base64(image_path: str) -> str:
"""Convert an image file to base64 data URL."""
with open(image_path, "rb") as f:
data = base64.b64encode(f.read()).decode("utf-8")
# Detect MIME type from file extension
if image_path.lower().endswith(".png"):
mime = "image/png"
elif image_path.lower().endswith(".jpg") or image_path.lower().endswith(".jpeg"):
mime = "image/jpeg"
elif image_path.lower().endswith(".webp"):
mime = "image/webp"
elif image_path.lower().endswith(".gif"):
mime = "image/gif"
else:
mime = "image/png"
return f"data:{mime};base64,{data}"
def analyze_image(
model_path: str,
image_path: str,
prompt: str,
max_tokens: int = 2048,
):
"""Analyze an image using Sixpert K1."""
print(f"Loading Sixpert K1...")
llm = Llama(
model_path=model_path,
n_ctx=131072,
n_gpu_layers=-1,
verbose=False,
)
image_data = image_to_base64(image_path)
messages = [
{
"role": "user",
"content": [
{
"type": "text",
"text": prompt,
},
{
"type": "image_url",
"image_url": {"url": image_data},
},
],
},
]
print(f"\nPrompt: {prompt}")
print(f"Image: {image_path}")
print("-" * 40)
print("Analyzing image...")
print("-" * 40)
response = llm.create_chat_completion(
messages=messages,
max_tokens=max_tokens,
temperature=0.7,
stream=True,
)
for chunk in response:
delta = chunk["choices"][0]["delta"].get("content", "")
if delta:
print(delta, end="", flush=True)
print("\n")
def main():
parser = argparse.ArgumentParser(description="Sixpert K1 Vision Example")
parser.add_argument("--model", type=str, default="SixpertK1.gguf", help="Path to GGUF model")
parser.add_argument("--image", type=str, required=True, help="Path to input image")
parser.add_argument("--prompt", type=str, default="Describe this image in detail.", help="Prompt")
parser.add_argument("--max-tokens", type=int, default=2048, help="Max tokens")
args = parser.parse_args()
analyze_image(args.model, args.image, args.prompt, args.max_tokens)
if __name__ == "__main__":
main()
|