Grigory Evko

I found out -w key should be used. I've successfully mounted backup with `sudo docker run -d --name eth -p 30333:30333 -p 9545:9545 --ulimit nofile=1000000:1000000 -v ~/nethermind/:/nethermind/data:Z -w /nethermind -it...

Really rooting for this! _From release notes:_ > FP16 and BF16 fused flash attention engine performance has been significantly improved for NVIDIA GPUs: > > - Speed-up of up to...

Please reopen, pytorch is not using nvidia-cudnn-cu12 9.0 latest https://pypi.org/project/nvidia-cudnn-cu12/ wheel

```from sqlalchemy import pool from sqlalchemy.ext.asyncio import create_async_engine, AsyncSession, async_sessionmaker import uuid db_url = f"postgresql+asyncpg://{DB_USER}:{DB_PASSWORD}@{DB_SERVER}:{DB_PORT}/postgres" engine = create_async_engine( db_url, poolclass=pool.NullPool, # important setting #1 , but maybe it's not required...

> Has anyone else noticed that queries run with poolclass=pool.NullPool are much slower than those executed using the postgREST client? Yes, that's kind of expected I had a long discussion...