Skip to content

Releases: archinetai/audio-diffusion-pytorch

v0.0.71

23 Oct 08:49
Compare
Choose a tag to compare
feat: add tanh bottleneck, option to use multiple bottlenecks

v0.0.70

21 Oct 18:01
Compare
Choose a tag to compare
feat: add option to use rel pos

v0.0.69

18 Oct 11:58
Compare
Choose a tag to compare
fix: v-sampler output

v0.0.68

17 Oct 21:54
Compare
Choose a tag to compare
fix: add linear schedule in init

v0.0.67

17 Oct 21:46
Compare
Choose a tag to compare
feat: add true v-diffusion, separate vk-diffusion

v0.0.66: fix: forward context_features, big bug that prevented conditioning fr…

16 Oct 22:01
Compare
Choose a tag to compare

v0.0.65

15 Oct 08:52
Compare
Choose a tag to compare
feat: add mean and logvar info to variational bottleneck

v0.0.64

13 Oct 15:04
Compare
Choose a tag to compare
fix: t5 embedder device

v0.0.63

13 Oct 14:55
Compare
Choose a tag to compare
fix: change to auto tokenizer

v0.0.62

13 Oct 13:41
Compare
Choose a tag to compare
feat: add pretrained/frozen T5 embedder