Commit Graph

Select branches

Hide Pull Requests

bigvae

borrow_vocab

full_ft

just_llm

just_llms2

main

main_origin

Mono Color

• 5c6afe5c25 wip main wassname 2024-05-11 10:28:00 +08:00
• a4234425af checkin wassname 2024-05-11 10:11:10 +08:00
• 73fdfddacb Merge branch 'main' of https://github.com/wassname/iris_bigvae wassname 2024-04-27 07:05:38 +08:00
• 991f157fdd Merge branch 'full_ft' wassname 2024-04-27 07:04:40 +08:00
• a82fe96158 left overs wassname 2024-04-27 07:01:09 +08:00
• 1519a0caab Update README.md Michael J Clark 2024-03-01 09:17:02 +08:00
• 87022a3677 Update README.md Michael J Clark 2023-12-03 11:28:31 +08:00
• 94e8ffa876 Update README.md Michael J Clark 2023-12-03 11:28:04 +08:00
• 8ecdf35bd3 learns the best yet full_ft wassname 2023-11-25 05:48:17 +08:00
• 9155ecca95 Fix embedding path in Transformer model wassname 2023-11-24 09:24:12 +08:00
• c3878e4a4c full finetune? wassname 2023-11-23 09:08:30 +08:00
• 72ee806601 try training tokenizer and world model together: result 💩 wassname 2023-11-23 09:08:10 +08:00
• 2763449f64 bug fix borrow_vocab wassname 2023-11-20 10:01:56 +08:00
• 56daac9761 rank, and unfreeze head wassname 2023-11-20 07:31:31 +08:00
• 6308397bb7 it works better but still flickers when imagining. hmm I do think I need the delta-iris stuff wassname 2023-11-20 07:14:50 +08:00
• 6eb25883f6 misc wassname 2023-11-19 14:50:54 +08:00
• ecee11fc33 notes wassname 2023-11-19 10:58:45 +08:00
• 328d281651 using frozen transformer embedding: runs wassname 2023-11-19 10:50:48 +08:00
• 7098b67955 1 wassname 2023-11-19 08:48:36 +08:00
• 36ea66d0ae misc wassname 2023-11-19 08:17:38 +08:00
• 9b399031ca document play.sh options wassname 2023-11-18 16:32:59 +08:00
• a341e7aaad abs imports, fix play for crafter just_llms2 wassname 2023-11-18 07:44:05 +08:00
• 1dabd79251 debugging model size and speed wassname 2023-11-17 18:17:17 +08:00
• a778296e40 relative imports wassname 2023-11-17 12:30:16 +08:00
• 1af7aa74fa 3min epoch breakout wassname 2023-11-17 06:25:09 +08:00
• 1b6462991b smaller horizon, smaller lstm. 100x faster actor_critic. does it learn though? wassname 2023-11-16 17:17:57 +08:00
• 106027a7f9 add 1b model, use contextlib, wassname 2023-11-14 10:42:17 +08:00
• e495db47e4 Fix num_passes calculation in WorldModelEnv.step() wassname 2023-11-13 17:17:03 +08:00
• e417790413 fix kv_cache faking? wassname 2023-11-13 08:11:13 +08:00
• fefaf44482 debug main_origin wassname 2023-11-13 07:15:16 +08:00
• e2bfef17fd fix for newer numpy wassname 2023-11-13 07:15:07 +08:00
• b97a499a19 Fix bugs and improve performance in training process wassname 2023-11-13 07:00:17 +08:00
• c8f59b0504 misc wassname 2023-11-12 18:41:22 +08:00
• a3ded691e1 misc wassname 2023-11-12 17:27:17 +08:00
• 1454e5c053 training wassname 2023-11-12 16:13:23 +08:00
• a6b850bd96 fix debug launch wassname 2023-11-12 14:44:10 +08:00
• 7dfded5bde debuggign wassname 2023-11-12 14:41:59 +08:00
• 28910cfd63 poetry wassname 2023-11-12 13:57:33 +08:00
• 1800cdf1d0 add poetry env wassname 2023-11-12 13:35:39 +08:00
• 315aea0616 checkout bigvae wassname 2023-11-12 13:13:24 +08:00
• 075de416c8 adding bigvae just_llm wassname 2023-11-12 12:56:55 +08:00
• ac6be401fe Remove unused sampling_weights. Eloi Alonso 2023-09-18 16:11:06 +02:00
• 42bbabee7c Update README.md Eloi Alonso 2023-03-01 10:28:07 +01:00
• e6aaa67a98 Update README.md Eloi Alonso 2022-10-05 10:56:15 +02:00
• a2818df57e Remove pretrained models from main repository. Eloi Alonso 2022-10-05 10:37:33 +02:00
• bf2d6b1943 Update README.md Vincent Micheli 2022-10-03 17:41:55 +02:00
• a25030cc47 Add pretrained models. Vincent Micheli 2022-10-03 17:21:16 +02:00
• faa2d343d2 Add mechanism to save frames when visualizing. Add option to visualize agent playing in the world model. Eloi Alonso 2022-09-30 18:58:16 +02:00
• 03290c820e Replace lambda with partial in env instantiation. Eloi Alonso 2022-09-19 10:30:54 +02:00
• ef7ecef186 Add link to arXiv and BibTeX. Eloi Alonso 2022-09-02 14:04:43 +02:00
• 75e5ab57e8 Code release. Eloi Alonso 2022-08-26 12:58:58 +02:00