TNT3530 closed issue #16393: [Bug] InitCCLPerWorker Fails when using AMD GPU
Bridge
URL: https://github.com/apache/tvm/issues/16393
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific
TNT3530 opened a new issue, #16393:
URL: https://github.com/apache/tvm/issues/16393
### Expected behavior
MLC-LLM should be load the sharded model across all 4 GPUs and start
inferring.
Issue is confirmed only with the bridge enabled, adding
`amdgpu.use_xgmi_p2p=0` to grub config