Apache-2.0 models and datasets for Hindi, Hinglish and 10+ Indian languages. Write-ups & FAQs: latentailabs.github.io
Arun Prabhakaran
arunp0
AI & ML interests
Machine learning, GPUs and open source. Open-weights safety and moderation models for Indian languages and Hinglish, PII detection for Indian ID formats, small encoders that hold their own against LLM judges, and honest held-out evaluation. Explaining how models actually work on YouTube (Latent AI Labs).
Recent Activity
new activity about 12 hours ago
huggingface/InferenceSupport:arunp0/nandi-grievance-indic new activity about 12 hours ago
huggingface/InferenceSupport:arunp0/bhasha-lid-indic new activity about 12 hours ago
huggingface/InferenceSupport:arunp0/bhasha-lid-indic-v2Organizations
None yet
Open models run on our own GPU in the How AI Works series by Latent AI Labs. Articles & FAQs: latentailabs.github.io
-
Qwen/Qwen3-4B
Text Generation • 4B • Updated • 8.42M • • 780 -
black-forest-labs/FLUX.2-klein-4B
Image-to-Image • 4B • Updated • 425k • • 1.09k -
black-forest-labs/FLUX.2-klein-base-4B
Image-to-Image • 4B • Updated • 683k • • 189 -
Qwen/Qwen2.5-VL-3B-Instruct
Image-Text-to-Text • 4B • Updated • 1.53M • • 742
Apache-2.0 models and datasets for Hindi, Hinglish and 10+ Indian languages. Write-ups & FAQs: latentailabs.github.io
Open models run on our own GPU in the How AI Works series by Latent AI Labs. Articles & FAQs: latentailabs.github.io
-
Qwen/Qwen3-4B
Text Generation • 4B • Updated • 8.42M • • 780 -
black-forest-labs/FLUX.2-klein-4B
Image-to-Image • 4B • Updated • 425k • • 1.09k -
black-forest-labs/FLUX.2-klein-base-4B
Image-to-Image • 4B • Updated • 683k • • 189 -
Qwen/Qwen2.5-VL-3B-Instruct
Image-Text-to-Text • 4B • Updated • 1.53M • • 742
models 6
arunp0/bhasha-lid-indic-v2
Token Classification • 0.3B • Updated • 43
arunp0/laya-indic
Text Classification • 0.3B • Updated • 69 • 2
arunp0/bhasha-lid-indic
Token Classification • 0.3B • Updated • 62 • 1
arunp0/nandi-grievance-indic
Text Classification • 0.6B • Updated • 54 • 1
arunp0/chitragupta-pii-indic
Token Classification • 0.6B • Updated • 68 • 1
arunp0/yama-indic
Zero-Shot Classification • 0.6B • Updated • 113 • 3