[NLPL Task Force (A)] Support request

[Vinit Ravishankar]

No dice unfortunately! Had to run a couple of separate runs to make sure, but this is what I get right when training starts:

RuntimeError: "mul_cpu" not implemented for ‘Half'

Tensorboard is also missing from nlpl-fairseq, so I do have to use my own virtualenv in addition to the module. I’ve attached my slurm file in case you notice any obvious issues.

Thanks!

– Vinit

> On 22 May 2020, at 12:14, Stephan Oepen <oe at ifi.uio.no> wrote:
> 
> hi again, vinit,
> 
>> Brilliant, thanks! I have had the time to submit a job, currently waiting for the queue to slow down a bit :)
> 
> any updates regarding that new fairseq module?  just now, the gpu
> queue seems not unreasonably busy ...
> 
> cheers, oe
-------------- next part --------------
A non-text attachment was scrubbed...
Name: transformer.slurm
Type: application/octet-stream
Size: 861 bytes
Desc: transformer.slurm
URL: <http://lists.nlpl.eu/archives/infrastructure/attachments/20200522/3772cbfd/attachment.obj>