typhoon-ai/typhoon-s-thaillm-8b-instruct-research-preview
Full fine-tune vs ThaiLLM/ThaiLLM-8B. Fine-tune: 100% of the weights changed, across 36 of 36 layers.
Tensors changed: Embeddings 1/1, Norms 136/145, MLP 108/108, Attention 144/144, lm_head 1/1
3 derivatives checked
- mradermacher/typhoon-s-thaillm-8b-instruct-research-preview-i1-GGUF: Quantization
- mradermacher/typhoon-s-thaillm-8b-instruct-research-preview-GGUF: Same weights, converted
- sthaps/typhoon-s-thaillm-8b-instruct-research-preview-gguf: Quantization