nitky/Oumuamua-7b-instruct-v2
Full fine-tune vs HachiML/Mistral-7B-v0.3-m3-lora. Fine-tune: 100% of the weights changed, across 32 of 32 layers.
Tensors changed: Norms 65/65, MLP 96/96, Attention 128/128
3 derivatives checked
- mradermacher/Oumuamua-7b-instruct-v2-i1-GGUF: Quantization
- mradermacher/Oumuamua-7b-instruct-v2-GGUF: Same weights, converted
- QuantFactory/Oumuamua-7b-instruct-v2-GGUF: Quantization