Colab not currently working
It seems to fail in a couple of places for me.
Installation of dependencies didn't work cleanly. I ended up doing a second !pip install lmdeploy=0.10.1 onnxruntime clip. No idea why that was needed, and it seemed to help... some. Among other warnings, there was this, too:
ERROR: Could not find a version that satisfies the requirement torch==2.8.0+cu126 (from versions: 2.2.0, 2.2.1, 2.2.2, 2.3.0, 2.3.1, 2.4.0, 2.4.1, 2.5.0, 2.5.1, 2.6.0, 2.7.0, 2.7.1, 2.8.0, 2.9.0, 2.9.1)
ERROR: No matching distribution found for torch==2.8.0+cu126
Run Content Captioning failed with:
/usr/local/lib/python3.12/dist-packages/lmdeploy/turbomind/supported_models.py in is_supported(model_path)
115 support_by_turbomind = False
116 elif arch == 'InternVLChatModel':
--> 117 llm_arch = cfg.llm_config.architectures[0]
118 support_by_turbomind = (llm_arch in SUPPORTED_ARCHS and _is_head_dim_supported(cfg.llm_config))
119 elif arch in ['LlavaForConditionalGeneration', 'InternVLForConditionalGeneration']:
AttributeError: 'dict' object has no attribute 'architectures'
Run Style Captioning failed with the same error.
Style Clustering failed with "✗ Failed to process Tess_01.png: module 'clip' has no attribute 'load'"
Aesthetic rating failed similarly, with the module 'clip'.
Maybe all this is user error, I've only dabbled in jupyter notebooks/Google colab... But I restarted the session and walked through them a second time and had similar problems.
I think this was all a problem with the install.
I switched the "!pip install ... torch==2.8.0+cu126" to simply "!pip install ... torch==2.8.0" that seemed to be enough.
The install seemed to complete successfully, and everything ran.