Skip to content
GitLab
Explore
Sign in
Tags
Tags give the ability to mark specific points in history as being important
1.15.1
40843bcc
·
pydantic 2
·
Jul 15, 2023
1.14.3
c56336a1
·
pydantic 2
·
Jul 15, 2023
1.14.2
00e07b7d
·
force einops 0.6.1 or greater and call allow_ops_in_compiled_graph
·
Apr 20, 2023
1.14.1
deda18fb
·
force einops 0.6.1 or greater and call allow_ops_in_compiled_graph
·
Apr 20, 2023
1.14.0
0069857c
·
remove einops exts for better pytorch 2.0 compile compatibility
·
Apr 20, 2023
1.12.4
580274be
·
use .to(device) to avoid copy, within one_unet_in_gpu context
·
Mar 07, 2023
1.12.3
848e8a48
·
always rederive the predicted noise from the clipped x0 for ddim + predict noise objective
·
Mar 05, 2023
1.12.2
cc58f754
·
bump to newer package of clip-anytorch that allows for text encodings < maximum context length
·
Mar 04, 2023
1.12.1
3b2cf7b0
·
fix for self conditioning in diffusion prior network...
·
Feb 11, 2023
1.12.0
984d62a3
·
default ddim sampling eta to 0
·
Dec 23, 2022
1.11.4
683dd98b
·
extra insurance in case eos id is not there
·
Dec 15, 2022
1.11.2
067ac323
·
address
https://github.com/lucidrains/DALLE2-pytorch/issues/266
·
Nov 23, 2022
1.11.1
fbba0f9a
·
bring in prediction of v objective, combining the findings from progressive...
·
Oct 28, 2022
1.11.0
1892f1ac
·
bring in prediction of v objective, combining the findings from progressive...
·
Oct 28, 2022
1.10.9
c3df46e3
·
fix openclipadapter to be able to use latest open sourced sota model
·
Oct 23, 2022
0.10.8
41fabf29
·
fix a dtype conversion issue for the diffusion timesteps in the diffusion...
·
Oct 19, 2022
1.10.7
c18c0801
·
fix for use with larger openai clip models by extracting dimension of last layernorm in clip
·
Sep 29, 2022
1.10.6
d0c11b30
·
handle open clip adapter image size being a tuple
·
Sep 19, 2022
1.10.5
0d82dff9
·
in ddim, noise should be predicted after x0 is maybe clipped, thanks to...
·
Sep 01, 2022
1.10.4
8bbc956f
·
fix bug with misnamed variable in diffusion prior network
·
Aug 31, 2022
Prev
1
2
3
4
5
…
18
Next