Meerkov

@mudlord Hi, your comment did not help advance this thread and is distracting from the subject which is what number to set the CountPerOp. I ask that you don't leave...

 Exact frame that it crashes to desktop

@Narann , please see my original post > I'm happy to provide logs, if you tell me where to find them.

I tried to repro and it happened in a different place. It's probably related to how long you've been playing. (approximately 10 minutes): Repro 3 out of 3 successes, all...

``` Core: Using full mem base Core: Goodname: Tonic Trouble (U) (V1.1) [!] Core: Name: Tonic Trouble Core: MD5: 7D3E935156844DE0002DB875E1076A5C Core: CRC: EF9E9714 C03B2C7D Core: Imagetype: .z64 (native) Core: Rom...

+1 Almost all audio dropped out within a few seconds of starting the first level.

> **SLUS_216.92** crc = F5AF6C2E > > > //Don't enable Intermittent GIF Mode. > > patch=0,EE,001F7E08,word,00000000 This doesn't work for me. I'm pretty sure I set the .pnach file up...

Edit: I figured out that you need to enable the cheat in a new menu (I'm not sure how cheats worked in the past, but afaik this menu is new)....

Still kind of curious if there is a more clear root cause. Obviously it seems like modern computers are calculating too much per frame. However it was mentioned this problem...

I tried this, but I still get errors ``` line 24, in get_learned_conditioning "original_size_as_tuple": torch.tensor([height, width], **devices_args).repeat(len(batch), 1), TypeError: must be real number, not NoneType ```