nectec/thai-research-gemma-3-27b-it
Full fine-tune vs google/gemma-3-27b-pt. Fine-tune: 98% of the weights changed, across 62 of 62 layers.
Tensors changed: Embeddings 1/4, Norms 359/484, MLP 186/294, Attention 248/464, other 0/1
2 derivatives checked
- mradermacher/thai-research-gemma-3-27b-it-i1-GGUF: Quantization (fewer layers)
- mradermacher/thai-research-gemma-3-27b-it-GGUF: Quantization (fewer layers)