mradermacher/Koa-AI-v1-Code-3B-GGUF

Same weights, converted (fewer layers) vs vamazing/Koa-AI-v1-Code-3B. mradermacher/Koa-AI-v1-Code-3B-GGUF matches part of vamazing/Koa-AI-v1-Code-3B (236 of its 458 tensors; the rest of vamazing/Koa-AI-v1-Code-3B has no counterpart). On the shared part: Same weights stored in another dtype (BF16 → F16/F32).

Tensors changed: Embeddings 0/1, Norms 0/53, MLP 0/78, Attention 0/104

On the Hub · Downloads on Model Pulse