mradermacher/Artha-v1-E2B-GGUF

Same weights, converted (fewer layers) vs developerJenis/Artha-v1-E2B. mradermacher/Artha-v1-E2B-GGUF matches part of developerJenis/Artha-v1-E2B (565 of its 2011 tensors; the rest of developerJenis/Artha-v1-E2B has no counterpart). On the shared part: Same weights stored in another dtype (BF16 → F16/BF16), plus 36 new tensors.

Tensors changed: Embeddings 0/2, Norms 0/247, MLP 0/105, other 0/71, Attention 0/140

On the Hub · Downloads on Model Pulse