chromadb/context-1
Partial fine-tune vs openai/gpt-oss-20b. Partial change: 67% of the weights changed (attention, lm_head), in 24 of 24 layers.
Tensors changed: Norms 0/49, MLP 0/96, Attention 96/216, lm_head 1/1, Embeddings 0/1
3 derivatives checked
- nicolasembleton/context-1-GGUF: LoRA-style
- ryancook/chromadb-context-1-gguf: Same weights, converted
- evilfreelancer/context-1-mxfp4: Same weights, repacked
4 other repos hold byte-for-byte identical weight files.