Multiple GPUs
#3
by
BigDeeper
- opened
I need a solution that can split a Wan model over 4 GPUs' VRAM (each is 12.2GiB)
Any suggestions?
Something similar to what Ollama does with LLMs.
I need a solution that can split a Wan model over 4 GPUs' VRAM (each is 12.2GiB)
Any suggestions?
Something similar to what Ollama does with LLMs.