sometimesanotion/Lamarck-14B-v0.7-Fusion
Full fine-tune vs jpacifico/Chocolatine-2-14B-Instruct-v2.0.3. Fine-tune: 100% of the weights changed, across 48 of 48 layers.
Tensors changed: Embeddings 1/1, Norms 91/97, MLP 144/144, Attention 334/336, lm_head 1/1
4 derivatives checked
- mradermacher/Lamarck-14B-v0.7-Fusion-i1-GGUF: Quantization
- MaziyarPanahi/Lamarck-14B-v0.7-Fusion-GGUF: Same weights, converted
- mradermacher/Lamarck-14B-v0.7-Fusion-GGUF: Quantization
- DevQuasar-9/sometimesanotion.Lamarck-14B-v0.7-Fusion-GGUF: Same weights, converted