Commit Graph

27 Commits

Author SHA1 Message Date
Seunghoon Lee 43b9c52bd4 Fix image corruption in half mode with embeddings.
(DirectML)
2023-07-25 17:02:55 +09:00
Seunghoon Lee 33d814e3e7 Fix DirectML tensor behavior. 2023-07-25 00:03:07 +09:00
Seunghoon Lee d4aa840a77 DirectML update.
DirectML reuses GPU memory instead of returning it.
So prints "practical" GPU memory utilization too.
2023-07-24 16:10:01 +09:00
Seunghoon Lee 6b4b863fd0 simplify get_device. 2023-07-15 19:03:10 +09:00
Seunghoon Lee 0a52c44e73 DirectML rework & provide GPU memory usage (AMD only). 2023-07-15 18:55:38 +09:00
Seunghoon Lee 578f206bf7 DirectML rework. 2023-07-15 16:19:38 +09:00
Seunghoon Lee 6b63893348 Fix terminal hang up with TI on DirectML. 2023-07-09 15:47:38 +09:00
Vladimir Mandic 2a21196061 Merge branch 'master' into dev 2023-07-08 13:35:25 -04:00
Seunghoon Lee 99c7f23c89 remove useless methods. 2023-07-09 01:29:04 +09:00
Seunghoon Lee 47c96e34d4 Stringify tensor instead of printing. 2023-07-07 19:11:18 +09:00
Seunghoon Lee 1d36e19996 Fix PNDMScheduler for DirectML. 2023-07-07 18:52:04 +09:00
Seunghoon Lee a076ff1b43 Use backend enum on diffusers check. 2023-07-06 01:45:12 +09:00
Seunghoon Lee 993de932ab Add an opts override for DirectML. 2023-07-06 01:35:00 +09:00
Seunghoon Lee a9fa014e31 Fix image corruption of RESRGAN upscaler in DirectML. 2023-06-29 11:10:48 +09:00
Vladimir Mandic cb307399dd jumbo merge 2023-06-13 11:59:56 -04:00
Vladimir Mandic 5134471bc8 dml autocast 2023-05-14 13:24:59 -04:00
Vladimir Mandic c46f0dbdeb fix 2023-05-13 12:55:50 -04:00
Vladimir Mandic d96ab6a1ae update directml 2023-05-13 11:21:11 -04:00
Vladimir Mandic 1360c6422a add fp16 test 2023-05-08 09:27:50 -04:00
Seunghoon Lee 9e214b32d7 remove NotImplementedError. 2023-05-02 09:04:35 +09:00
Seunghoon Lee 93d638d8b8 Approx NN works fine. 2023-04-30 11:15:09 +09:00
Seunghoon Lee e65d74100f Move/Remove hijacks. Add warning. 2023-04-30 11:12:47 +09:00
Seunghoon Lee c128c0770b Load libatiadlxx.so on Linux systems. 2023-04-30 01:15:47 +09:00
Seunghoon Lee c9dde03fc5 Move realesrgan fallback to dml/hijack. 2023-04-26 18:54:24 +09:00
Seunghoon Lee d2d5011bd3 Implement memory estimation for AMDGPUs.
Stable.
2023-04-26 17:44:32 +09:00
Seunghoon Lee df0e89be48 fix.
Unstable & need more test.
2023-04-26 12:45:44 +09:00
Seunghoon Lee 09ae33cdf7 Implement torch.dml.
VERY UNSTABLE & NOT TESTED.
2023-04-26 12:21:44 +09:00