cyberagent/Mistral-Nemo-Japanese-Instruct-2408
Full fine-tune vs mistralai/Mistral-Nemo-Base-2407. Fine-tune: 100% of the weights changed, across 40 of 40 layers.
Tensors changed: Norms 81/81, MLP 120/120, Attention 160/160
7 derivatives checked
- mradermacher/Mistral-Nemo-Japanese-Instruct-2408-i1-GGUF: Quantization
- mradermacher/Mistral-Nemo-Japanese-Instruct-2408-GGUF: Quantization
- shisa-ai/Mistral-Nemo-Japanese-Instruct-2408-GPTQ-W4A16-gs128: Quantization
- shisa-ai/ablation-152-finalsft-shisa-v2-mistral-nemo-japanese-12b: Full fine-tune
- shisa-ai/ablation-54-nemojp-shisa-v2-mistral-nemo-japanese-12b: Full fine-tune
- shisa-ai/ablation-178-finalsft2-shisa-v2-mistral-nemo-japanese-12b: Full fine-tune
- win10/ChatML-Nemo-Pro: Full fine-tune