Commit Graph

101 Commits

Author SHA1 Message Date
Brett Kuprel
b93c4fd0e4 simplified colab 2022-06-30 21:02:27 -04:00
Brett Kuprel
432ffa8d8c reusable mega torch model works in standard colab runtime 2022-06-30 20:55:42 -04:00
Brett Kuprel
08b158d580 updated readme 2022-06-30 16:50:04 -04:00
Brett Kuprel
2311a1af7b delete cache 2022-06-30 15:48:20 -04:00
Brett Kuprel
d9d7f34b22 update readme 2022-06-30 15:33:53 -04:00
Brett Kuprel
23620b95cd Merge branch 'main' of https://github.com/kuprel/min-dalle 2022-06-30 15:19:21 -04:00
Brett Kuprel
c8f3304363 update readme 2022-06-30 15:18:29 -04:00
Brett Kuprel
a3d81888b5 generate avocado armchair with pytorch in standard colab runtime 2022-06-30 15:17:35 -04:00
Brett Kuprel
ed0193296c mega model works in standard colab runtime 2022-06-30 14:57:53 -04:00
Brett Kuprel
b913b58353 pre converting params to torch allows mega to run in standard colab runtime 2022-06-30 14:54:08 -04:00
Brett Kuprel
de97fcf06b
Update requirements.txt 2022-06-30 13:13:33 -04:00
Brett Kuprel
c2a3858c96 delete params sooner 2022-06-30 11:44:36 -04:00
Brett Kuprel
f951424e38 is_reusable 2022-06-30 11:25:24 -04:00
Brett Kuprel
3e64e868ef Merge branch 'main' of https://github.com/kuprel/min-dalle 2022-06-30 11:09:17 -04:00
Brett Kuprel
b55bcba4c0 removed deepcopy, delete expendable parameters after use 2022-06-30 11:09:09 -04:00
Brett Kuprel
b44e8997a0 still some issues with mega in standard runtime, reverting for now 2022-06-30 10:02:08 -04:00
Brett Kuprel
6d1ac07442 mega model in standard runtime 2022-06-30 09:43:13 -04:00
Brett Kuprel
e9c01e32a5 mega model works with standard colab runtime in expendable mode 2022-06-30 09:39:14 -04:00
Brett Kuprel
41a44068d0 keep params in expendable mode 2022-06-30 09:36:32 -04:00
Brett Kuprel
df9aa6f915 sort -> topk, prev_token_and_index -> prev_token, token_index 2022-06-30 09:04:11 -04:00
Brett Kuprel
fb97ba5e20 update readme, cleanup 2022-06-30 07:41:31 -04:00
Brett Kuprel
1e18ba0ffa is_expendable argument reduces memory usage for command line script 2022-06-30 06:43:10 -04:00
Brett Kuprel
38377107da
Update setup.sh 2022-06-29 21:45:05 -04:00
Brett Kuprel
c76226571e
Update cog.yaml 2022-06-29 19:14:25 -04:00
Brett Kuprel
8b86152359
Update cog.yaml 2022-06-29 19:14:10 -04:00
Brett Kuprel
5f931f6899 simplified colab setup 2022-06-29 16:32:51 -04:00
Brett Kuprel
b7b6df23f7 updated replicate predict.py file 2022-06-29 15:53:25 -04:00
Brett Kuprel
005ee4938e
Update README.md 2022-06-29 15:24:09 -04:00
Brett Kuprel
9fbfe0b76e
Delete predict.py 2022-06-29 15:20:45 -04:00
Brett Kuprel
15b0c03485
Merge pull request #38 from chenxwh/replicate
Add Web Demo & Docker environment
2022-06-29 15:20:20 -04:00
Brett Kuprel
d99828a239 simplified flax attention and matched torch attention 2022-06-29 14:56:28 -04:00
Chenxi
fcc17c895d
Merge branch 'kuprel:main' into replicate 2022-06-29 19:50:47 +01:00
Chenxi
eb9f4c6b3b replicate demo 2022-06-29 19:50:10 +01:00
Brett Kuprel
61cc99c13c read tokenizer files with utf8 encoding 2022-06-29 14:18:23 -04:00
Brett Kuprel
742d3609b0 Merge branch 'main' of https://github.com/kuprel/min-dalle 2022-06-29 13:56:39 -04:00
Brett Kuprel
a4df279fd2 simplified attention and keys_values state resulted in decrease in inference time to 7.3 seconds (from ~10 seconds) 2022-06-29 13:56:29 -04:00
Brett Kuprel
80558b8a82 time the inference pass 2022-06-29 13:55:23 -04:00
Brett Kuprel
661ec976ac simplified attention for torch model 2022-06-29 13:48:12 -04:00
Brett Kuprel
95afa18893 add gitattributes file 2022-06-29 12:45:41 -04:00
Brett Kuprel
764b5bbc0e works with latest flax version 0.5.2, updated requirements.txt 2022-06-29 11:46:19 -04:00
Brett Kuprel
c4f613c89f readme wording 2022-06-29 11:01:46 -04:00
Brett Kuprel
6046863805 readme wording 2022-06-29 11:00:38 -04:00
Brett Kuprel
2b552fe9db readme wording 2022-06-29 10:54:01 -04:00
Brett Kuprel
b7c2414c76 readme wording 2022-06-29 10:47:08 -04:00
Brett Kuprel
4e62f85ab9 readme wording 2022-06-29 10:45:46 -04:00
Brett Kuprel
53695e32f7 updated readme with torch examples 2022-06-29 10:43:46 -04:00
kuprel
9d5bb34df0 default to torch+mega in colab 2022-06-29 10:37:12 -04:00
kuprel
d4af693133 updated colab to load model once and generate multiple times 2022-06-29 09:43:42 -04:00
Brett Kuprel
ed91ab4a30 refactored to load models once and run multiple times 2022-06-29 09:42:12 -04:00
kuprel
1ef9b0b929 added mega to colab 2022-06-29 06:44:07 -04:00