Rombo-Org/Rombo-LLM-V3.0-Qwen-32b
Full fine-tune vs Rombo-Org/Rombo-LLM-V2.5-Qwen-32b. Fine-tune: 98% of the weights changed, across 64 of 64 layers.
Tensors changed: lm_head 1/1, Embeddings 0/1, Norms 127/129, MLP 192/192, Attention 448/448
6 derivatives checked
- bartowski/Rombo-Org_Rombo-LLM-V3.0-Qwen-32b-GGUF: Quantization
- mradermacher/Rombo-LLM-V3.0-Qwen-32b-i1-GGUF: Quantization
- mradermacher/Rombo-LLM-V3.0-Qwen-32b-GGUF: Quantization
- DevQuasar-8/Rombo-Org.Rombo-LLM-V3.0-Qwen-32b-GGUF: Same weights, converted
- valoomba/Rombo-V3.1-32B-Reasoner: LoRA-style
- valoomba/Rombo-V3-32B-Reasoner: LoRA-style