bittensor
bittensor copied to clipboard
[Fix] [Hotfix] [3.3.4] multi cuda fix
This PR addresses the issues with CUDA registration introduced by v3.3.4 This fixes:
- Improper termination And this should fix:
- Reduced actual hashrate (does not match reported)
nonce_start += self.update_interval * self.num_proc- to
nonce_start += self.update_interval * self.TPB