mradermacher/HelpingAI-Lite-2x1B-GGUF

Same weights, converted vs OEvortex/HelpingAI-Lite-2x1B. Same weights stored in another dtype (F32 → F16/F32), plus 66 new tensors.

Tensors changed: Embeddings 0/1, other 0/22, Norms 0/45, Attention 0/88, lm_head 0/1

On the Hub · Downloads on Model Pulse