xiaomi-research/MiLMMT-46-1B-v0.1
Full fine-tune vs xiaomi-research/MiLMMT-46-1B-Pretrain. Fine-tune: 100% of the weights changed, across 26 of 26 layers.
Tensors changed: lm_head 1/1, Embeddings 1/1, Norms 106/157, MLP 78/78, Attention 104/104
7 derivatives checked
- xiaomi-research/MiLMMT-46-1B-v1.0: Full fine-tune
- mradermacher/MiLMMT-46-1B-v0.1-i1-GGUF: Quantization
- mradermacher/MiLMMT-46-1B-v0.1-GGUF: Same weights, converted
- NetherQuartz/ilo-toki-MiLMMT-46-1b-merged: LoRA-style
- NetherQuartz/ilo-toki-1.1-MiLMMT-46-1b-merged: LoRA-style
- NetherQuartz/ilo-toki-1.3-MiLMMT-46-1b-merged: LoRA-style
- NetherQuartz/ilo-toki-1.4-MiLMMT-46-1b-merged: LoRA-style