prithivMLmods/Primal-Opus-14B-Optimus-v1
Full fine-tune vs prithivMLmods/Calcium-Opus-14B-Elite2-R1. Fine-tune: 100% of the weights changed, across 48 of 48 layers.
Tensors changed: Embeddings 1/1, Norms 97/97, MLP 144/144, Attention 336/336, lm_head 1/1
4 derivatives checked
- mradermacher/Primal-Opus-14B-Optimus-v1-i1-GGUF: Quantization
- mradermacher/Primal-Opus-14B-Optimus-v1-GGUF: Quantization
- prithivMLmods/Primal-Opus-14B-Optimus-v2: Full fine-tune
- prithivMLmods/Megatron-Opus-14B-Exp: LoRA-style