PRPLL 0.15-122-ge4709ec-dirty starting

[cudanexus]

when I run this

build-release/prpll -prp 8191 -device 0 -tune

20250207 06:47:20  PRPLL 0.15-122-ge4709ec-dirty starting
20250207 06:47:20  config: -prp 8191 -device 0 -tune 
20250207 06:47:20  device 0, OpenCL 535.183.01, unique id ''
20250207 06:47:20  BPW info for 512:16:512 not found, defaults={18.02, 18.12, 18.12, 18.22}
20250207 06:47:20  BPW info for 1K:16:512 not found, defaults={17.74, 17.84, 17.84, 17.94}
20250207 06:47:20  BPW info for 4K:16:512 not found, defaults={17.19, 17.29, 17.29, 17.39}
20250207 06:47:20  BPW info for 4K:16:1K not found, defaults={16.92, 17.02, 17.02, 17.12}
20250207 06:47:22     149.5 256:2:256:0:0   5117313
20250207 06:47:23     168.8 256:3:256:0:0   7613448
20250207 06:47:24     224.4 256:4:256:0:0  10086252
20250207 06:47:26     239.7 256:2:512:0:0  10040115
20250207 06:47:27     228.1 512:2:256:0:0  10040115
20250207 06:47:29     269.7 256:5:256:0:0  12513443
20250207 06:47:30     324.7 256:6:256:0:0  14987034
20250207 06:47:32     325.4 256:3:512:0:0  14983102

my GPU is getting used
but when I run this

build-release/prpll -dir /home/syed/outputs 
20250207 06:51:33  PRPLL 0.15-122-ge4709ec-dirty starting
20250207 06:51:33  config: -dir /home/syed/outputs 
20250207 06:51:33  device 0, OpenCL 535.183.01, unique id ''
20250207 06:51:33  Bye

I tired multip options
build-release/prpll -dir /home/syed/outputs -device 0 -verbose
build-release/prpll -user syed -dir /media/syed/models/prime -device 0
this are the content inside output folder

kernel-cache
proof/
proof-tmp/
gpuowl-0.log/
results.txt/
worktodo.txt

even with -verbose nothing happens
I have run the python -OO autoprimenet.py --dir /home/syed/outputs to get work

[tdulcet]

I have run the python -OO autoprimenet.py --dir /home/syed/outputs to get work

PRPLL is not yet PrimeNet server compatible and thus not supported by AutoPrimeNet. In this case, the issue (among other things) is that PRPLL and AutoPrimeNet are using different work files. We are actively working to address this, for example see #316.

[cudanexus]

then how can i use it now can you give me steps to use it with current code

[tdulcet]

You would either need to switch to @olympichek's branch in #316 or the gpuowl branch, both of which are PrimeNet server compatible. If you just want to quickly try PRPLL out, you could rename your worktodo.txt file to worktodo-0.txt, but note that this does not resolve the major incompatibilities. Otherwise, I would just wait until work on PRPLL is complete and it is ready for production use.

[EWouters]

I encountered this issue too and I could not find much documentation until I stumbled upon this thread. I think adding some getting started instructions would help. Perhaps the Use section of the README could be expanded a bit:

1. Create a run directory, for instance mkdir run inside the root of the repository
2. Create a worktodo-N.txt in the run folder for every worker. If using 1 worker, create only worktodo-0.txt
3. Optionally create a config.txt in the run folder with your username, number of workers, etc.
4. Run build-release/prpll -dir run -tune
5. Start testing with build-release/prpll -dir run