xiaomi-research/GemmaX2-28-2B-v0.2

Full fine-tune vs ModelSpace/GemmaX2-28-2B-Pretrain. Fine-tune: 100% of the weights changed, across 26 of 26 layers.

Tensors changed: Embeddings 1/1, Norms 95/105, MLP 78/78, Attention 104/104, lm_head 1/1

2 derivatives checked

• mradermacher/GemmaX2-28-2B-v0.2-GGUF: Same weights, converted
• mradermacher/GemmaX2-28-2B-v0.2-i1-GGUF: Quantization

On the Hub · Downloads on Model Pulse