allenai/OLMoE-1B-7B-0924-SFT
Full fine-tune vs allenai/OLMoE-1B-7B-0924. Fine-tune: 100% of the weights changed, across 16 of 16 layers.
Tensors changed: Embeddings 1/1, Norms 65/65, MLP 2869/2869, Attention 64/64, lm_head 1/1
4 derivatives checked
- allenai/OLMoE-1B-7B-0924-Instruct: Full fine-tune
- QuantFactory/OLMoE-1B-7B-0924-Instruct-GGUF: Full fine-tune
- mradermacher/OLMoE-1B-7B-0924-SFT-i1-GGUF: Quantization
- mradermacher/OLMoE-1B-7B-0924-SFT-GGUF: Same weights, converted