kaupane/pexels-people-captions
Viewer • Updated • 43.4k • 82
Wang Weiyi PRO
kaupane
AI & ML interests
None yet
Recent Activity
updated a dataset about 16 hours ago
kaupane/pexels-people-captions liked a dataset about 17 hours ago
jovianzm/Pexels-400k updated a dataset 5 days ago
kaupane/artflow-code-bridgeOrganizations
None yet
updated a
dataset about 16 hours ago
liked a
dataset about 17 hours ago
updated a
dataset 5 days ago
published a
dataset 6 days ago
updated a
dataset 17 days ago
liked a
model 17 days ago
published a
dataset 17 days ago
published a
dataset 27 days ago
upvoted a collection about 1 month ago
upvoted a paper about 1 month ago
reacted to nwaughachukwuma's post with 👍 about 2 months ago
Post
686
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):
• glm-5.2 (text-only) + mm-ctx: 88.4
• gemini-3.5-flash (vision): 83
• deepseek-v4-pro (text-only) + mm-ctx: 79.4
• qwen3.6-35b-a3b (vision): 44.3
The best text-only setup
• 1.5x faster (100s vs 150s mean per task)
• the only setup with zero timeouts (46/46 completed; gemini timed out 4x on bulk-image and long-video tasks)
• the only setup stable across runs (88.5 / 88.4)
• top on video (100.0), image (91.7), and PDF (90.0) tasks
The trade-offs: the toolkit consumed 3.3x more tokens (4.25M vs 1.28M), and lost on audio (85.6 vs 71.3).
On completed tasks alone the two are nearly identical (91.0 vs 88.4): the toolkit's edge is efficient extraction that keeps long media tasks inside the time budget.
Full report: https://huggingface.co/blog/vlm-run/text-only-models-with-mm
We benchmarked 4 setups on 23 multimodal tasks (image, video, audio, PDF):
• glm-5.2 (text-only) + mm-ctx: 88.4
• gemini-3.5-flash (vision): 83
• deepseek-v4-pro (text-only) + mm-ctx: 79.4
• qwen3.6-35b-a3b (vision): 44.3
The best text-only setup
glm-5.2 + mm outperformed gemini-3.5-flash, the top vision model, by 5.4 points (6.5%). It was also:• 1.5x faster (100s vs 150s mean per task)
• the only setup with zero timeouts (46/46 completed; gemini timed out 4x on bulk-image and long-video tasks)
• the only setup stable across runs (88.5 / 88.4)
• top on video (100.0), image (91.7), and PDF (90.0) tasks
The trade-offs: the toolkit consumed 3.3x more tokens (4.25M vs 1.28M), and lost on audio (85.6 vs 71.3).
On completed tasks alone the two are nearly identical (91.0 vs 88.4): the toolkit's edge is efficient extraction that keeps long media tasks inside the time budget.
Full report: https://huggingface.co/blog/vlm-run/text-only-models-with-mm
- 8 replies
liked a
model 2 months ago
liked a
Space 2 months ago
liked a
model 3 months ago
upvoted an article 4 months ago
Article
Towards Speed-of-Light Text Generation with Nemotron-Labs Diffusion Language Models
nvidia
• • 36 liked 4
models 5 months ago
XiaomiMiMo/MiMo-V2.5-Pro
Text Generation • 1T • Updated • 21.9k • • 760
openai/privacy-filter
Token Classification • 1B • Updated • 304k • • 1.77k
deepseek-ai/DeepSeek-V4-Pro
Text Generation • 1.6T • Updated • 462k • • 5.6k
OpenMOSS-Team/MOSS-Music-8B-Instruct
Audio-Text-to-Text • 9B • Updated • 2.84k • 39
updated a
dataset 6 months ago