loading model from glm-4v crashed

[zalsay]

(glm-4v) PS D:\dev\models> python D:\dev\models\vllm-ipex-server\glm-4v.py
C:\ProgramData\miniconda3\envs\glm-4v\Lib\site-packages\torchvision\io\image.py:13: UserWarning: Failed to load image Python extension: 'Could not find module 'C:\ProgramData\miniconda3\envs\glm-4v\Lib\site-packages\torchvision\image.pyd' (or one of its dependencies). Try using the full path with constructor syntax.'If you don't plan on using image functionality from torchvision.io, you can ignore this warning. Otherwise, there might be something wrong with your environment. Did you have libjpeg or libpng installed before building torchvision from source?
warn(
2024-12-25 12:12:09,139 - INFO - intel_extension_for_pytorch auto imported
2024-12-25 12:12:11,063 - modelscope - INFO - PyTorch version 2.1.0a0+cxx11.abi Found.
2024-12-25 12:12:11,064 - modelscope - INFO - Loading ast index from C:\Users\sisu.cache\modelscope\ast_indexer
2024-12-25 12:12:11,223 - modelscope - INFO - Loading done! Current index file version is 1.11.0, with md5 061b1804e8e732cc7bbff4121137b87c and a total number of 953 components indexed
Loading checkpoint shards: 67%|███████████████████████████████████████████████████████████▎ | 10/15 [00:01<00:00, 7.34it/s]
(glm-4v) PS D:\dev\models>
platform: win11 dGPU: A770

[zalsay]

code:
model_path = "ZhipuAI/glm-4v-9b"
model = AutoModel.from_pretrained(model_path,
load_in_4bit=True,
optimize_model=True,
trust_remote_code=True,
use_cache=True,
model_hub="modelscope")
model = model.half().to('xpu')

tokenizer = AutoTokenizer.from_pretrained(model_path, trust_remote_code=True)

[zalsay]

in the model's config.json, I added "image_size": 1120,

[hzjane]

Hi, You can follow this guide to check whether your environment and code are correct.