[
https://issues.apache.org/jira/browse/NLPCRAFT-413?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Rahul Padmanabhan closed NLPCRAFT-413.
--------------------------------------
Resolution: Abandoned
Due to changes made to the overall project, this is no longer needed.
> We need to add support for CUDA > 11 in the Ctxserver
> -----------------------------------------------------
>
> Key: NLPCRAFT-413
> URL: https://issues.apache.org/jira/browse/NLPCRAFT-413
> Project: NLPCraft
> Issue Type: Bug
> Reporter: Rahul Padmanabhan
> Assignee: Rahul Padmanabhan
> Priority: Major
>
> Ctxserver issue: Currently the bertft.py code fails when there is CUDA > 11
> and does not output the /suggestions end point correctly. This is due to the
> PyTorch version 1.6.0 in the requirements.
> {code:java|title=Error Warning|borderStyle=solid}
> n [11]: model.cuda()
> /<path>/incubator-nlpcraft/.nlpcraft-python/nlpcraft-condaenv/lib/python3.8/site-packages/torch/cuda/__init__.py:106:
> UserWarning:
> NVIDIA GeForce RTX 3080 with CUDA capability sm_86 is not compatible with the
> current PyTorch installation.
> The current PyTorch install supports CUDA capabilities sm_37 sm_50 sm_60
> sm_70.
> If you want to use the NVIDIA GeForce RTX 3080 GPU with PyTorch, please check
> the instructions at https://pytorch.org/get-started/locally/
> warnings.warn(incompatible_device_warn.format(device_name, capability, "
> ".join(arch_list), device_name))
> {code}
>
--
This message was sent by Atlassian Jira
(v8.20.7#820007)
