mhla
/

gpt1900-d34-v3-sft

Model card Files Files and versions

GPT-1900 D34 V3 SFT (Safe Corpus, Raw Base)

3.29B parameter GPT-1900 with v3 corpus SFT (opinion-filtered) on raw base (no physics CLM).

Training

Training: V3 SFT (safe corpus, 32K examples), from base d34-22btok
Parameters: 3.29B
Architecture: Custom GPT with RoPE, QK-norm, ReLU², value embeddings

Downloads last month: -; Downloads are not tracked for this model. How to track

Inference Providers NEW

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Collection including mhla/gpt1900-d34-v3-sft

GPT-1900 Drafts

Experimental and intermediate GPT-1900 checkpoints. Working artifacts, not for general use. • 49 items • Updated 19 days ago