Chukwuma Nwaugha
nwaughachukwuma
AI & ML interests
NLP for Sentiments and Emotions Analysis; Word Similarities; Text relevance within the context of a Sentence; Word search
Recent Activity
repliedto their post about 8 hours ago
Can a text-only model + a vision toolkit (mm-ctx) match a native vision model?
We benchmarked 4 setups on 23 multimodal tasks (image, video, audio, PDF):
• gemini-3.5-flash (vision): 85.5
• glm-5.2 + mm-ctx (text-only): 75.9
• deepseek-v4-pro + mm-ctx (text-only): 68.8
• qwen3.6-35b-a3b (vision): 46.0
The best text-only setup closes 89% of the gap to the top vision model.
Full report: https://huggingface.co/blog/vlm-run/text-only-models-with-mm repliedto their post about 19 hours ago
Can a text-only model + a vision toolkit (mm-ctx) match a native vision model?
We benchmarked 4 setups on 23 multimodal tasks (image, video, audio, PDF):
• gemini-3.5-flash (vision): 85.5
• glm-5.2 + mm-ctx (text-only): 75.9
• deepseek-v4-pro + mm-ctx (text-only): 68.8
• qwen3.6-35b-a3b (vision): 46.0
The best text-only setup closes 89% of the gap to the top vision model.
Full report: https://huggingface.co/blog/vlm-run/text-only-models-with-mm posted an update 1 day ago
Can a text-only model + a vision toolkit (mm-ctx) match a native vision model?
We benchmarked 4 setups on 23 multimodal tasks (image, video, audio, PDF):
• gemini-3.5-flash (vision): 85.5
• glm-5.2 + mm-ctx (text-only): 75.9
• deepseek-v4-pro + mm-ctx (text-only): 68.8
• qwen3.6-35b-a3b (vision): 46.0
The best text-only setup closes 89% of the gap to the top vision model.
Full report: https://huggingface.co/blog/vlm-run/text-only-models-with-mmOrganizations