AetherResearch/Cerebrum-1.0-8x7b
Partial fine-tune vs mistralai/Mixtral-8x7B-v0.1. Partial change: 67% of the weights changed (attention, mlp, other), in 32 of 32 layers.
Tensors changed: Embeddings 0/1, MLP 512/768, other 32/32, Norms 0/65, Attention 128/128, lm_head 0/1
5 derivatives checked
- backyardai/Cerebrum-1.0-8x7b-GGUF: Not comparable yet
- CISCai/Cerebrum-1.0-8x7b-SOTA-GGUF: Not comparable yet
- qwp4w3hyb/Cerebrum-1.0-8x7b-iMat-GGUF: Not comparable yet
- mradermacher/Cerebrum-1.0-8x7b-GGUF: Quantization
- mradermacher/Cerebrum-1.0-8x7b-i1-GGUF: Quantization