Feat: sebulba ff_ippo #1088
Merged
Feat: sebulba ff_ippo #1088
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
…nv) ->(n_env, n_agents)
RuanJohn
requested changes
Nov 5, 2024
sash-a
requested changes
Nov 8, 2024
sash-a
previously approved these changes
Nov 11, 2024
There was a problem hiding this comment.
Massive effort, thanks for this @Louay-Ben-nessir!! 🔥 🥇
Just need to resolve the merge conflicts and I want to check the final benchmarks before we merge this, but in terms of code I am happy!
commit 6092dc656cd73ee5a0fb0dd6e29c50b11c9b84ac Merge: 73537c5f 3ddcbff7 Author: Ruan de Kock <[email protected]> Date: Fri Nov 8 15:45:07 2024 +0200 Merge pull request #1130 from instadeepai/fix/sable-pos-encoding fix: limit timestep-pos-encoding to rec-Sable commit 3ddcbff74fe1fa221c037e9701a502fcd6c8aa64 Author: OmaymaMahjoub <[email protected]> Date: Fri Nov 8 11:14:21 2024 +0000 docs: update docs commit daf1c199b4e2bdf0a9c012f6681d5fdb18781a25 Author: OmaymaMahjoub <[email protected]> Date: Fri Nov 8 11:11:45 2024 +0000 fix: controling timestep positional encoding in acting phase commit 73537c5f2294773fc73ba9e4f71203e13c97fc59 Merge: 905710fc d3631094 Author: Wiem Khlifi <[email protected]> Date: Thu Nov 7 15:35:30 2024 +0100 Merge pull request #1126 from instadeepai/fix/mabrax fix: mabrax requirement commit d3631094feec5e8de3b3ff23382ac447414bb8fe Author: Sasha Abramowitz <[email protected]> Date: Thu Nov 7 14:52:54 2024 +0200 fix: mabrax requirement commit 905710fc7d14e2567640268be72fc59835e31697 Merge: c86604c4 bb8e1073 Author: Omayma Mahjoub <[email protected]> Date: Thu Nov 7 13:29:08 2024 +0100 Merge pull request #1113 from instadeepai/feat/sable Add Sable [Discrete actions] commit bb8e1073187cd9bd5ca5d4c04bbf385868ae9546 Author: Omayma Mahjoub <[email protected]> Date: Thu Nov 7 11:05:38 2024 +0100 Update mava/systems/sable/anakin/ff_sable.py Co-authored-by: Sasha Abramowitz <[email protected]> commit b3b43ec05ebed5205e465d2bad7f75dc5825baa2 Author: Omayma Mahjoub <[email protected]> Date: Thu Nov 7 11:05:27 2024 +0100 Update mava/systems/sable/anakin/ff_sable.py Co-authored-by: Sasha Abramowitz <[email protected]> commit 408c027e0e7366d539d163e36831764f323580e3 Author: Omayma Mahjoub <[email protected]> Date: Thu Nov 7 11:05:20 2024 +0100 Update mava/systems/sable/anakin/rec_sable.py Co-authored-by: Sasha Abramowitz <[email protected]> commit 3c250b838fe2ad7b6bac3e3ec770364aecf38c45 Author: Omayma Mahjoub <[email protected]> Date: Thu Nov 7 11:04:23 2024 +0100 Update mava/networks/sable_network.py Co-authored-by: Sasha Abramowitz <[email protected]> commit 18f7e662055f12519b4c5a6f3bbc54ea3e8bce16 Author: OmaymaMahjoub <[email protected]> Date: Thu Nov 7 10:03:34 2024 +0000 feat: update decoder file by removing unnecessary functions commit a0daaebf80d3407c7e5c03389dcab6e2b9d0b2bd Author: OmaymaMahjoub <[email protected]> Date: Thu Nov 7 09:52:32 2024 +0000 feat: update docs based on review Co-authored-by: Sasha Abramowitz <[email protected]> commit 210faddc59c88a44a6a8c16e70e27767802c6116 Author: OmaymaMahjoub <[email protected]> Date: Thu Nov 7 09:26:22 2024 +0100 fix: run pre commits commit 8546254ccb15febd92694b2a554bc7c8d08d9cbf Author: Omayma Mahjoub <[email protected]> Date: Thu Nov 7 09:24:41 2024 +0100 Update mava/systems/sable/anakin/rec_sable.py Co-authored-by: Ruan de Kock <[email protected]> commit 5b7156ff75c248a49f817397ca00f9d77215172d Author: Omayma Mahjoub <[email protected]> Date: Thu Nov 7 09:24:33 2024 +0100 Update mava/systems/sable/anakin/ff_sable.py Co-authored-by: Ruan de Kock <[email protected]> commit d3719baa79b26ff9e580968d90e7ee319bd6c374 Author: Omayma Mahjoub <[email protected]> Date: Thu Nov 7 09:24:26 2024 +0100 Update mava/systems/sable/anakin/ff_sable.py Co-authored-by: Ruan de Kock <[email protected]> commit 011995d5cfd6287b7e9d25d1c049d128533d17a0 Author: OmaymaMahjoub <[email protected]> Date: Wed Nov 6 16:18:04 2024 +0100 feat: move all system specific config setup to the system file commit e552509e9aac94b4941969fa6e9acb9f4d4282e3 Author: OmaymaMahjoub <[email protected]> Date: Wed Nov 6 15:58:38 2024 +0100 feat: checkpointer hstate retoring fix commit bf58ded9038e79c544c3f82e419e71499270273b Author: OmaymaMahjoub <[email protected]> Date: Wed Nov 6 14:58:38 2024 +0100 feat: move concat agents and time to jax utils commit e0ce8f42f16cedea4d20bac709520ff17f40bbf2 Author: OmaymaMahjoub <[email protected]> Date: Wed Nov 6 14:56:19 2024 +0100 feat: get the positional encoding flag outside the util fn commit aa9cba8864b020c08d0effc1eafcfd9008108e1c Author: OmaymaMahjoub <[email protected]> Date: Wed Nov 6 14:50:51 2024 +0100 feat: split encoder_decoder_Fn to two files commit 0030b356ad4bbb7248a8b330334af89b92c78cbc Author: OmaymaMahjoub <[email protected]> Date: Wed Nov 6 14:45:49 2024 +0100 feat: use input hstate as the output variable instead of using extra hs variable commit d9432f4aad865b1495e7b9f9e538d4dad3c3bc77 Author: OmaymaMahjoub <[email protected]> Date: Wed Nov 6 12:58:50 2024 +0000 feat: rename retentions to retention_heads commit 73ff86fca63bb7664ffe12dac91de0a407310e6f Author: OmaymaMahjoub <[email protected]> Date: Wed Nov 6 12:45:54 2024 +0000 feat: replace init fn of sable net to get_actions one commit 3998b51a33aed49735dcf1ab9b335424cf1a1263 Author: OmaymaMahjoub <[email protected]> Date: Wed Nov 6 12:37:48 2024 +0000 feat: send optimizer update fn directly without intermediate var commit 9f36fe6ac1fe333b661d4644c7f4468aff978a42 Author: OmaymaMahjoub <[email protected]> Date: Wed Nov 6 12:35:18 2024 +0000 feat: move squeezing output of the net to inside the net fns commit a6370a97493415e2d34bf881bf9edeece3839180 Author: OmaymaMahjoub <[email protected]> Date: Wed Nov 6 12:27:14 2024 +0000 docs: update some docs commit d80cf9186b111e56067cfb16272c57cca41aa9d0 Author: OmaymaMahjoub <[email protected]> Date: Wed Nov 6 12:18:25 2024 +0000 feat: replace full attn flag by masking flag commit 5f214cef462cd968975380fd122c2e17d5a5574e Author: OmaymaMahjoub <[email protected]> Date: Wed Nov 6 12:14:00 2024 +0000 feat: use the chunk size only to decide on use chunkwise flag for that commit b5d39934e21c244f4af31940f82c506c10146620 Author: Omayma Mahjoub <[email protected]> Date: Wed Nov 6 12:58:47 2024 +0100 Update mava/networks/utils/sable/encoder_decoder_fns.py Co-authored-by: Ruan de Kock <[email protected]> commit 670de846ed4f9818ec9c3864e16b9448d1a3ab23 Author: Omayma Mahjoub <[email protected]> Date: Wed Nov 6 12:58:27 2024 +0100 Update mava/networks/utils/sable/encoder_decoder_fns.py Co-authored-by: Ruan de Kock <[email protected]> commit de16e844f7e82aa20ba736e532bf1d44f70bf5c6 Author: OmaymaMahjoub <[email protected]> Date: Wed Nov 6 11:56:59 2024 +0000 feat: rename training apply callable type to LearnerApply commit 8af3bb407a5780bd45dd99fc29ed4aae78efc4ff Author: OmaymaMahjoub <[email protected]> Date: Wed Nov 6 11:54:52 2024 +0000 feat: addressing some renaming suggestions commit f0360d1341c21b2a3ae0d8401ee04d1fd95d7b78 Author: OmaymaMahjoub <[email protected]> Date: Wed Nov 6 11:05:56 2024 +0000 chore: rename obs_carry to observation commit 57e3b517b376b88fb0a39d602023cd4e258b41a3 Author: OmaymaMahjoub <[email protected]> Date: Wed Nov 6 10:45:20 2024 +0000 fix: renmaing the shape related to n_agents and actions_dim commit 437b8f62f35b988777d8c796ed43ad052f734707 Merge: e0c863c2 b3ac1d9b Author: OmaymaMahjoub <[email protected]> Date: Wed Nov 6 10:36:46 2024 +0000 Merge branch 'feat/sable' of github.com:instadeepai/Mava into feat/sable commit e0c863c233676f791e105d0b22f0b3d187236de8 Author: OmaymaMahjoub <[email protected]> Date: Wed Nov 6 10:35:57 2024 +0000 feat: update the action type to follow up same MAT standards Co-authored-by: Sasha Abramowitz <[email protected]> commit b3ac1d9bca01d1a6e221d626147b3640b688c6f2 Merge: 7646a2f0 c86604c4 Author: Ruan de Kock <[email protected]> Date: Wed Nov 6 12:15:45 2024 +0200 Merge branch 'develop' into feat/sable commit 7646a2f03d9ce2e255d142b30bfb7490aa8e97e7 Author: OmaymaMahjoub <[email protected]> Date: Wed Nov 6 10:00:06 2024 +0000 fix: update timeout in workflow to 20 min Co-authored-by: Sasha Abramowitz <[email protected]> commit 0dd0eab6a9ff823ad7508177b8bc4b7265cc1ccb Author: OmaymaMahjoub <[email protected]> Date: Wed Nov 6 09:53:51 2024 +0000 feat: update shifting action method in autoregressive act Co-authored-by: Sasha Abramowitz <[email protected]> commit 945937cc977db0e469f12ca3227d66c595491d08 Author: OmaymaMahjoub <[email protected]> Date: Wed Nov 6 09:36:29 2024 +0000 feat: standardize the definition of net config to NamedTuple commit c86604c4a6232d6bafee99a6ebfa7693cd652ebe Merge: 7f2568a7 fb5c97c6 Author: Ruan de Kock <[email protected]> Date: Tue Nov 5 16:52:40 2024 +0200 Merge pull request #1120 from instadeepai/feat/vector-connector-wrapper Add vector connector wrapper commit fb5c97c61ac60d6484b2f51493feb10103b8d1ea Author: RuanJohn <[email protected]> Date: Tue Nov 5 15:34:57 2024 +0200 chore: docstring commit d1a0c1c6406f95170268afb4c0b548b8ef177e08 Author: OmaymaMahjoub <[email protected]> Date: Tue Nov 5 10:27:00 2024 +0000 feat: make intermediate line to calculate decay_matrix commit ae652fcefb7d29c7f8877556dcffc275cfeb4886 Merge: 1d8515e5 7f2568a7 Author: Ruan de Kock <[email protected]> Date: Tue Nov 5 10:19:46 2024 +0200 Merge branch 'develop' into feat/vector-connector-wrapper commit 7f2568a7a3944b6ca3195f6561ada55ee163d864 Merge: 3577523b b689a83e Author: Ruan de Kock <[email protected]> Date: Tue Nov 5 10:19:22 2024 +0200 Merge pull request #1123 from instadeepai/chore/num-minibatches-assert Chore: Add asserts for number of envs divisible by number of minibatches commit b689a83e87044c1241dc29aa435fc7ba061336ce Merge: d555f21a 3577523b Author: Sasha Abramowitz <[email protected]> Date: Tue Nov 5 10:01:53 2024 +0200 Merge branch 'develop' into chore/num-minibatches-assert commit eb625901cb3b3c1ed9a99ceaaa3111993c34f2ac Author: OmaymaMahjoub <[email protected]> Date: Mon Nov 4 16:18:45 2024 +0000 fix: major fix of sending non zero hstate for autoregressive act commit 69f39a57712561b8a42f5da6671ac6576b80c6f0 Author: OmaymaMahjoub <[email protected]> Date: Mon Nov 4 13:15:13 2024 +0000 feat: rename shape vars in encoder decoder fns file commit 7068a689cf8f7783043dc9b20a60a50e8fd39fa5 Author: OmaymaMahjoub <[email protected]> Date: Mon Nov 4 10:19:20 2024 +0000 feat: merge the chunkwise and parallel fns into one commit 938541283de54f71084f83724d11f0ceb40dc3eb Author: OmaymaMahjoub <[email protected]> Date: Mon Nov 4 07:50:22 2024 +0000 feat: move make eval fn to system files commit 75ced75c2f989e3221e83583bc0ad8c7097c93e4 Author: OmaymaMahjoub <[email protected]> Date: Mon Nov 4 07:26:36 2024 +0000 feat: move sable util fns to network folder commit 1d8515e5a36e6b5d1a70ba04b821b6fc96d7019c Merge: fd276c0f 3577523b Author: Ruan de Kock <[email protected]> Date: Mon Nov 4 09:06:34 2024 +0200 Merge branch 'develop' into feat/vector-connector-wrapper commit 1d38c24f555405f154bff0e0bcc38f94b81b6923 Author: OmaymaMahjoub <[email protected]> Date: Mon Nov 4 05:55:09 2024 +0000 feat: update checkpointer fn output types commit 2b80a7d8586fa2fed630519f58fd94b28f321b4c Author: OmaymaMahjoub <[email protected]> Date: Sat Nov 2 18:09:22 2024 +0000 feat: update sable hstate attributes naming commit 584b0d4551921db42725c375b8872350cdf98dfc Author: OmaymaMahjoub <[email protected]> Date: Sat Nov 2 18:03:10 2024 +0000 chore: update tree map commit dd21d04dfdec9843e401b91fc1a277ba9a78b9c2 Author: OmaymaMahjoub <[email protected]> Date: Sat Nov 2 18:00:26 2024 +0000 chore: docs fixes in sable network file commit 30351515867db54556521316d990686a79a8343f Author: OmaymaMahjoub <[email protected]> Date: Sat Nov 2 17:18:57 2024 +0000 chore: docs fixes in retention file commit 3577523b3a2f32bd35a2bf2e91f14a539a451e95 Merge: 327e0664 3373c579 Author: Wiem Khlifi <[email protected]> Date: Sat Nov 2 14:27:05 2024 +0100 Merge pull request #1119 from instadeepai/fix/quickstart-notebook Fix quickstart notebook commit 3373c57929d7cfb7c3b40c0f0218716bcfa3a1f6 Author: WiemKhlifi <[email protected]> Date: Fri Nov 1 17:06:31 2024 +0100 revert: point on develop for installation commit 65d1f2d3dc080816f10f8cec3882471a0013ba90 Merge: d866bd57 327e0664 Author: Wiem Khlifi <[email protected]> Date: Fri Nov 1 16:31:29 2024 +0100 Merge branch 'develop' into fix/quickstart-notebook commit fd276c0f5b4df580a1d0a37282ed36c7669c852d Merge: 3f658ee5 327e0664 Author: Wiem Khlifi <[email protected]> Date: Fri Nov 1 16:25:08 2024 +0100 Merge branch 'develop' into feat/vector-connector-wrapper commit d555f21aecb48aa2b8bf5545c24742aa46e51c1a Author: SimonDuToit <[email protected]> Date: Fri Nov 1 17:04:39 2024 +0200 pre-commit commit b11fb37123971e220164b722ce26bef19e895de5 Merge: 63785093 327e0664 Author: SimonDuToit <[email protected]> Date: Fri Nov 1 16:34:08 2024 +0200 Merge branch 'develop' into chore/num-minibatches-assert commit 327e0664fafbb3ba18ea6d1f8a48166c9106c5d4 Merge: 6eed2d2f 7944e41b Author: Wiem Khlifi <[email protected]> Date: Fri Nov 1 15:32:45 2024 +0100 Merge pull request #1121 from instadeepai/feat/more-rware-scenarios More rware scenario configs commit 63785093873a0d21df2e3fa6b91b03516e1d0a16 Author: SimonDuToit <[email protected]> Date: Fri Nov 1 16:32:28 2024 +0200 add asserts commit d866bd575ba205a0ba20bed931edb19ee4151e67 Author: Sasha Abramowitz <[email protected]> Date: Fri Nov 1 16:24:55 2024 +0200 chore: update explainer text in example notebook commit 648337049aee98822458ceb41c371a77f62ff777 Author: Omayma Mahjoub <[email protected]> Date: Fri Nov 1 10:31:19 2024 +0100 Update mava/configs/network/ff_retention.yaml Co-authored-by: Sasha Abramowitz <[email protected]> commit aa8b455eadc7390a483b17316d0c41b28e6c77aa Author: Omayma Mahjoub <[email protected]> Date: Fri Nov 1 10:31:03 2024 +0100 Update mava/configs/network/rec_retention.yaml Co-authored-by: Sasha Abramowitz <[email protected]> commit 3f658ee50571cb08543d68c2883d2879191c3196 Author: Ruan de Kock <[email protected]> Date: Thu Oct 31 17:48:33 2024 +0200 test: add vector connector to integration tests commit 69db3eb1851b6c9e20f3db71758e88d8b39312ab Author: Ruan de Kock <[email protected]> Date: Thu Oct 31 17:10:45 2024 +0200 feat: separate env config for vector connector commit 1fdfce910f0dcc9ef44eb2c2c9607f7eb45c5762 Author: OmaymaMahjoub <[email protected]> Date: Thu Oct 31 14:00:37 2024 +0000 fix: define decay scaling factor for ff sable before sending config to enc-dec Co-authored By: sash-a commit 283b6a9dc2f8587f6e0feb2ad70703d9a0fa5d32 Author: Ruan de Kock <[email protected]> Date: Thu Oct 31 15:09:03 2024 +0200 feat: use boolean masks instead of jnp.where Co-authored-by: Sasha Abramowitz <[email protected]> commit ba52ce4f463a1101994ee079bdc69cc1296376c9 Author: Ruan de Kock <[email protected]> Date: Thu Oct 31 15:08:34 2024 +0200 chore: remove debug print statement Co-authored-by: Sasha Abramowitz <[email protected]> commit 77f291cf032f842b9cfb36970ca6e5563d01c61f Author: Ruan de Kock <[email protected]> Date: Thu Oct 31 15:08:04 2024 +0200 chore: comments for view shapes Co-authored-by: Sasha Abramowitz <[email protected]> commit 7944e41b494558b6ded8e2227e9f50aff25784a6 Author: RuanJohn <[email protected]> Date: Wed Oct 30 16:49:51 2024 +0200 feat: more rware scenario configs commit 3d36aab988aa36c2fc5e4cdbc7c46ce53e2a8b7d Author: RuanJohn <[email protected]> Date: Wed Oct 30 15:58:58 2024 +0200 feat: add vector connector wrapper commit dc00782761f9f5e46d4804ba53b1dd8ebe4eec13 Author: OmaymaMahjoub <[email protected]> Date: Wed Oct 30 13:15:05 2024 +0000 fix: fixing the training by adding causal masking of decoder for ff sable commit 2fb21c7c1bb9acf3a29048f154bc1b704e9ae989 Author: OmaymaMahjoub <[email protected]> Date: Wed Oct 30 12:23:23 2024 +0000 feat: remove the sable net checker in simple retention commit 7732d52311f473fb698242530fa059b8213dbc8f Merge: ef32a219 6eed2d2f Author: OmaymaMahjoub <[email protected]> Date: Wed Oct 30 09:48:04 2024 +0000 feat: merge develop branch commit eea913b8f6e3e6ec39776f714adcdc730b42e10f Merge: cc47103d 6eed2d2f Author: Wiem Khlifi <[email protected]> Date: Wed Oct 30 10:23:16 2024 +0100 Merge branch 'develop' into fix/quickstart-notebook commit 6eed2d2fd3b27b55a0d86e7146609eb7b483d584 Merge: 389fbe58 ed3f015c Author: Wiem Khlifi <[email protected]> Date: Wed Oct 30 10:22:04 2024 +0100 Merge pull request #1115 from instadeepai/feat/new-dockerfile feat: updated dockerfile commit ed3f015c5d35a0df38b5bd434750a7618c7fe0a1 Merge: 83fa5a9e 389fbe58 Author: Wiem Khlifi <[email protected]> Date: Wed Oct 30 10:11:00 2024 +0100 Merge branch 'develop' into feat/new-dockerfile commit 389fbe586e9de425b87fe89ca75bd066849644e2 Merge: 25008fbc 0ec7049d Author: Ruan de Kock <[email protected]> Date: Wed Oct 30 11:00:41 2024 +0200 Merge pull request #1107 from instadeepai/feat/implement-mat Add MAT commit 0ec7049d8bc91b87b3f221dfb35f3fe16a271770 Merge: 3d47bebb 25008fbc Author: Ruan de Kock <[email protected]> Date: Wed Oct 30 10:03:16 2024 +0200 feat: merge in main commit 25008fbc5af4744be83c2e0ff007812718c27f60 Merge: 8b758133 936c0b8e Author: Sasha Abramowitz <[email protected]> Date: Wed Oct 30 09:57:34 2024 +0200 Merge pull request #1105 from instadeepai/feat/hasac2 feat: hasac commit 3d47bebba6610c8d32c5107864831ee96be4d357 Author: Ruan de Kock <[email protected]> Date: Wed Oct 30 09:25:23 2024 +0200 feat: swiglu documentation commit 7276aa0c93eb5a336281bb21282fd21af1314d41 Author: Ruan de Kock <[email protected]> Date: Wed Oct 30 09:16:15 2024 +0200 feat: execution and training apply types commit d9358311ba4475c8b1df601cba330907cdcb2617 Author: Ruan de Kock <[email protected]> Date: Wed Oct 30 09:09:55 2024 +0200 chore: rename embed dim commit a309bfa3b73056a441234238019553b82fe8b916 Author: Ruan de Kock <[email protected]> Date: Wed Oct 30 09:03:42 2024 +0200 chore: remove obs dim in MAT network class commit 3cb460d405988a5a9c40d6b664bf1c675f348fc5 Author: Ruan de Kock <[email protected]> Date: Wed Oct 30 08:53:45 2024 +0200 chore: config comments and reverts commit ef32a21947e92a9c34d0933f1bd2d308e5159b69 Author: OmaymaMahjoub <[email protected]> Date: Tue Oct 29 15:31:37 2024 +0000 feat: compress net params in net_config commit 83fa5a9e2bba5f3a63565163fa9383afb8952685 Author: Sasha Abramowitz <[email protected]> Date: Tue Oct 29 17:29:17 2024 +0200 chore: remove docker volumes from makefile Co-authored-by: Ruan de Kock <[email protected]> commit 936c0b8e5635a371a197cd1d256f2a181445fc59 Merge: cf45f98e 8b758133 Author: Sasha Abramowitz <[email protected]> Date: Tue Oct 29 17:25:18 2024 +0200 Merge branch 'develop' into feat/hasac2 commit cc47103d305516203ea0143c7ed96be59331172d Author: Ruan de Kock <[email protected]> Date: Tue Oct 29 15:17:36 2024 +0200 chore: remove notebook restarting cells commit 975df5fd8ec2b9c591e325ac32061dd54d4f60a1 Author: Ruan de Kock <[email protected]> Date: Tue Oct 29 14:50:37 2024 +0200 docs: mention that we use python 3.10 on colab commit 4376b14a425f8355c89cfd5f392fb9ef919743c9 Author: Ruan de Kock <[email protected]> Date: Tue Oct 29 14:48:49 2024 +0200 temp: change dir to quickstart notebook for reviewing commit 71f572cb70efa725ec3e94a2d86fe06a9e8cd878 Merge: 19731683 8b758133 Author: OmaymaMahjoub <[email protected]> Date: Tue Oct 29 12:32:20 2024 +0100 merge develop commit 19731683b99e9bec89afff3124c8dd9dd90faa0e Author: OmaymaMahjoub <[email protected]> Date: Tue Oct 29 12:24:41 2024 +0100 feat: prevent decay matrix calculation in case of ff sable commit 742903cac149b6ad8b0513961d15bb883c0cb68a Author: OmaymaMahjoub <[email protected]> Date: Tue Oct 29 11:46:07 2024 +0100 fix: fixing the retention output indexing commit 2f9dd4edb08a30d8c55cb965079ccca33e1d73d4 Author: Ruan de Kock <[email protected]> Date: Tue Oct 29 10:30:02 2024 +0200 fix: update quickstart notebook commit e8b7f57912037214a80c066ac386b9c373364f8f Author: Ruan de Kock <[email protected]> Date: Tue Oct 29 09:18:38 2024 +0200 feat: update pyproject commit 8a11bcf1e74193b783feeb834fbd0b2c64309f49 Merge: 5424c663 8b758133 Author: Ruan de Kock <[email protected]> Date: Mon Oct 28 19:45:37 2024 +0200 Merge branch 'develop' into feat/new-dockerfile commit cf45f98e7df00906d2abccbd6463e61ef43b6ad0 Author: Sasha Abramowitz <[email protected]> Date: Mon Oct 28 19:10:42 2024 +0200 chore: docs commit 7c8b91964e39edddfda078835ee5f7bfba80927e Author: Sasha Abramowitz <[email protected]> Date: Mon Oct 28 19:01:57 2024 +0200 chore: docs Co-authored-by: Omayma Mahjoub <[email protected]> commit bd4c8bcc32a696e8e268ac5e36ceea44d8d7ea3d Author: Ruan de Kock <[email protected]> Date: Mon Oct 28 17:33:41 2024 +0200 chore: pre-commit commit f3c990e222256d8aa8d1630195012e704b33b2bd Merge: fc2b2bd5 8b758133 Author: Ruan de Kock <[email protected]> Date: Mon Oct 28 17:08:29 2024 +0200 Merge branch 'develop' into feat/implement-mat commit 8b758133056e86303ab1acbe5aa2ade02e0f6e70 Merge: 54d3b50a 755b4600 Author: Ruan de Kock <[email protected]> Date: Mon Oct 28 17:07:26 2024 +0200 Merge pull request #1106 from instadeepai/feat/merge-qmix Add QMIX commit 755b4600db94fe79da7192ad33cda62025d1f9e0 Author: Ruan de Kock <[email protected]> Date: Mon Oct 28 16:43:07 2024 +0200 chore: remove type hint commit 880698c203b40c3e9b995ac6b09334856e5d642f Merge: 3c81350f 54d3b50a Author: Ruan de Kock <[email protected]> Date: Mon Oct 28 16:41:40 2024 +0200 Merge branch 'develop' into feat/merge-qmix commit 54d3b50abaa833d805244dc62cf5a9f909948b6a Merge: 87354a38 e9ff8b87 Author: Ruan de Kock <[email protected]> Date: Mon Oct 28 16:41:18 2024 +0200 Merge pull request #1109 from instadeepai/feat/pyproject-toml feat: switch to pyproject.toml commit fc2b2bd57b45d810829ce4bf7a702e29c685a5c0 Author: Ruan de Kock <[email protected]> Date: Mon Oct 28 16:40:05 2024 +0200 chore: set correct number of keys commit 123f5b19360f07a097abc62eb6b1ea18206d5d79 Author: Ruan de Kock <[email protected]> Date: Mon Oct 28 16:36:30 2024 +0200 chore: better action encoder init commit e9ff8b87007e030c5329a8b0413799e7cc8e21dd Author: Sasha Abramowitz <[email protected]> Date: Mon Oct 28 16:28:08 2024 +0200 chore: strict zip commit 3cb5bcd9e84ebe47f43ac1780470c0116ad25fb3 Author: Ruan de Kock <[email protected]> Date: Mon Oct 28 16:19:00 2024 +0200 chore: add dim on new line Co-authored-by: Sasha Abramowitz <[email protected]> commit 1b4cdea028e2e847dbb2a582939651c416b05bd3 Author: Sasha Abramowitz <[email protected]> Date: Mon Oct 28 16:18:27 2024 +0200 chore: strict zip Co-authored-by: Wiem Khlifi <[email protected]> commit 7f7b2b514a53e914d476e15babfead11b2b9e058 Author: Ruan de Kock <[email protected]> Date: Mon Oct 28 15:52:21 2024 +0200 feat: type hint jaxmarl and gigastep env commit 5424c663649b463a3e606c39742c97f32b617116 Author: Sasha Abramowitz <[email protected]> Date: Mon Oct 28 15:51:13 2024 +0200 chore: uppercase AS in Dockerfile Co-authored-by: Wiem Khlifi <[email protected]> commit 3ecd7723c272c1dd7597d6829d96a5d5948c53e0 Merge: 41467f82 87354a38 Author: Ruan de Kock <[email protected]> Date: Mon Oct 28 15:22:27 2024 +0200 Merge branch 'develop' into feat/pyproject-toml commit 65538d5377c1ff9bc5e8661b3fb6beb273e613a0 Merge: 3b686481 87354a38 Author: Sasha Abramowitz <[email protected]> Date: Mon Oct 28 15:15:18 2024 +0200 Merge branch 'develop' into feat/implement-mat commit 2cea286ffb41009291262f150b3285184a0f83d6 Merge: 9682bb29 87354a38 Author: Sasha Abramowitz <[email protected]> Date: Mon Oct 28 14:09:33 2024 +0200 Merge branch 'develop' into feat/hasac2 commit 9682bb294d592598084b79fd6a909fc7dad3101b Author: Sasha Abramowitz <[email protected]> Date: Mon Oct 28 14:09:16 2024 +0200 chore: shape comments commit 1237117e9073beb8ebd7aa92bbe3e76c72d527c3 Author: Sasha Abramowitz <[email protected]> Date: Mon Oct 28 14:02:46 2024 +0200 chore: shape comments Co-authored-by: Ruan de Kock <[email protected]> commit 87354a38fef29d02b21b980a97271412244a791c Author: Wiem Khlifi <[email protected]> Date: Mon Oct 28 12:03:53 2024 +0100 fix: fix logging during evaluation for JaxMARL envs (#1116) Co-authored-by: Sasha Abramowitz <[email protected]> commit 5aa0c30d4496e8b1d20211c7f8b6662e4c073b35 Merge: 3ff88416 3d541f2d Author: Ruan de Kock <[email protected]> Date: Mon Oct 28 12:18:20 2024 +0200 Merge branch 'develop' into feat/hasac2 commit 3b686481b9b43567e4721c3e58f70794c63c85b2 Merge: 9334319f 3d541f2d Author: Ruan de Kock <[email protected]> Date: Mon Oct 28 12:15:20 2024 +0200 Merge branch 'develop' into feat/implement-mat commit 3c81350ff70d04b32539345b9cbb48916cad30e7 Merge: e49a22f7 3d541f2d Author: Ruan de Kock <[email protected]> Date: Mon Oct 28 12:14:54 2024 +0200 Merge branch 'develop' into feat/merge-qmix commit 3d541f2d85797678da8f154d81199112eacf8f09 Author: Ruan de Kock <[email protected]> Date: Mon Oct 28 12:13:41 2024 +0200 Fix lbf and rware obs spec types (#1114) * fix: lbf and rware obs spec types * fix: fix obs spec type in gigastep commit 9334319fcdca7a46a63cd4deb091c18f71c8d7ea Author: Ruan de Kock <[email protected]> Date: Mon Oct 28 12:10:18 2024 +0200 chore: more lightweight network configs commit e49a22f7fe633af72e63246cfc4e1bb6f0c751e6 Author: Ruan de Kock <[email protected]> Date: Mon Oct 28 12:05:37 2024 +0200 chore: shape comments legend commit f11c21ec515f2dcc04eb91b10aae68afc018e402 Author: Ruan de Kock <[email protected]> Date: Mon Oct 28 11:58:21 2024 +0200 chore: corect shape names in the comments commit 36f54d1846bddba918d0c6cd8fd2bf637dcd5122 Author: Sasha Abramowitz <[email protected]> Date: Mon Oct 28 11:56:47 2024 +0200 feat: udpated dockerfile commit 98378f3f1554f67e34f00b3f78609afc3a083b73 Author: Ruan de Kock <[email protected]> Date: Mon Oct 28 11:40:56 2024 +0200 feat: add MAT network config type commit aff9feb11ed4a84be3558324b9d64a78845369c6 Author: Ruan de Kock <[email protected]> Date: Mon Oct 28 11:30:58 2024 +0200 feat: use network for MAT network init commit 66884fb88b31868461b898066102002837edb5bf Author: Ruan de Kock <[email protected]> Date: Mon Oct 28 09:39:29 2024 +0200 test: add mat to integration tests commit 738ec3c7049cdfce8a8f205b147e789ad922d9cd Author: Ruan de Kock <[email protected]> Date: Mon Oct 28 09:29:15 2024 +0200 feat: add qmix to intergration tests commit c620e17f3a784bea3b1f65d5ba8a79cc3b0be036 Merge: c00f54fd cd31e205 Author: Ruan de Kock <[email protected]> Date: Mon Oct 28 09:25:03 2024 +0200 Merge branch 'develop' into feat/implement-mat commit 5f3f8e031945bba018d23888e0b9af59951c1c94 Author: Ruan de Kock <[email protected]> Date: Mon Oct 28 09:20:31 2024 +0200 chore: rename data variables in training commit fc091890c91f1887cb3087101f01038a04e4cc2a Merge: c80da623 cd31e205 Author: Ruan de Kock <[email protected]> Date: Mon Oct 28 09:06:30 2024 +0200 chore: merge in main commit a6ae60296d80ee3d4e5d9b3d486dcfe88e46937e Author: OmaymaMahjoub <[email protected]> Date: Sun Oct 27 17:27:21 2024 +0100 fix: minor documentations edits commit 3ec23e80308c95a263f770b3d793bdf82db575f6 Merge: 8c56da70 cd31e205 Author: Omayma Mahjoub <[email protected]> Date: Sun Oct 27 17:17:51 2024 +0100 Merge branch 'develop' into feat/sable commit 8c56da7090303181bc546398b012f5795047b480 Author: OmaymaMahjoub <[email protected]> Date: Sun Oct 27 16:16:57 2024 +0000 feat: checkpointer update based on MAT PR commit 7601bba98f491ed448f6c1dbf5b545d305b0368d Author: OmaymaMahjoub <[email protected]> Date: Sun Oct 27 15:55:54 2024 +0000 feat: add sable to the integration test commit 8abc50104dd0b1df4e487d80217b7d9165a9e20f Author: OmaymaMahjoub <[email protected]> Date: Sun Oct 27 15:25:28 2024 +0000 fix: fixing the apply fn output ordering commit 78f99c9fb1d82e642e04af135f63d293a9bebdbf Author: OmaymaMahjoub <[email protected]> Date: Sat Oct 26 16:32:24 2024 +0100 fix: minor updates to net config commit c80da6236578e3b2421eedb953de2369d95e59b7 Author: Ruan de Kock <[email protected]> Date: Fri Oct 25 18:12:36 2024 +0200 fix: correct spec typing in lbf and rware commit cd31e2056f3a0e59bf31118c11cc53742fb9eb1d Author: Sasha Abramowitz <[email protected]> Date: Fri Oct 25 17:23:28 2024 +0200 feat: smaller networks for tests (#1111) * feat: smaller networks and new way to modify test config * feat: faster find_replace Co-authored-by: Ruan de Kock <[email protected]> Co-authored-by: Wiem Khlifi <[email protected]> * refactor: move find_replace to test/utils.py * chore: pre-commit --------- Co-authored-by: Ruan de Kock <[email protected]> Co-authored-by: Wiem Khlifi <[email protected]> commit 41467f82df0339b609b2da40c62806d7e5443939 Author: Sasha Abramowitz <[email protected]> Date: Fri Oct 25 16:54:52 2024 +0200 fix: add unwrapped method to gigastep and jaxmarl wrappers commit 3ff88416e649371081a8e26af2b8f64ad49f4994 Author: Sasha Abramowitz <[email protected]> Date: Fri Oct 25 16:38:48 2024 +0200 chore: pre-commit commit 617504bd178e0e9bef476ff1beb7d90256289ad0 Author: Sasha Abramowitz <[email protected]> Date: Fri Oct 25 16:27:47 2024 +0200 chore: shape and global state comments commit cb6bb68e092e45b32d0a810400359ca0463f89c6 Author: Ruan de Kock <[email protected]> Date: Fri Oct 25 12:00:20 2024 +0200 fix: increase sample sequence length in testing config commit 9a4fcbc6cc23bec3069baf38a0b0dc1d6289af18 Merge: 3b6bd930 bc6eb1a9 Author: Ruan de Kock <[email protected]> Date: Fri Oct 25 11:22:50 2024 +0200 Merge branch 'develop' into feat/hasac2 commit 3043a9d0c3da871f24efb058ebee01da06a71a40 Merge: 3c4ea141 bc6eb1a9 Author: Ruan de Kock <[email protected]> Date: Fri Oct 25 11:20:10 2024 +0200 Merge branch 'develop' into feat/pyproject-toml commit f1549d19a6f90bbd7aa9d226c36be68fffca22b9 Merge: a2d4215a bc6eb1a9 Author: Ruan de Kock <[email protected]> Date: Fri Oct 25 11:19:22 2024 +0200 Merge branch 'develop' into feat/merge-qmix commit c00f54fd84c76cc8b6d6e57359370f07d3cef9b4 Merge: ee3aff6a bc6eb1a9 Author: Ruan de Kock <[email protected]> Date: Fri Oct 25 11:19:10 2024 +0200 Merge branch 'develop' into feat/implement-mat commit a2d4215aeab8e45b6389de9721638b9a0f90ebc1 Author: Ruan de Kock <[email protected]> Date: Fri Oct 25 11:17:46 2024 +0200 chore: reset config defaults commit dfdfd3232ec7d8a10ac82eaec47346277c0de910 Author: Ruan de Kock <[email protected]> Date: Fri Oct 25 11:15:35 2024 +0200 chore: rename performance variable commit aae973d1758899852e502598d245cffcff09b626 Author: Ruan de Kock <[email protected]> Date: Fri Oct 25 11:09:01 2024 +0200 chore: rename data_first and and data_next commit 3c4ea141680341c27956fbb78dfa7049d76066df Author: Sasha Abramowitz <[email protected]> Date: Fri Oct 25 11:04:24 2024 +0200 chore: typo Co-authored-by: Ruan de Kock <[email protected]> commit 5df0e1e41420d01260d0129871d48673ffd04142 Author: Ruan de Kock <[email protected]> Date: Fri Oct 25 11:02:49 2024 +0200 chore: change comment in configs about sequence length commit bc6eb1a9564cac9ffc861fe3e3ce34cd423ea8ad Merge: 57c8e640 dfef2387 Author: Sasha Abramowitz <[email protected]> Date: Fri Oct 25 11:00:55 2024 +0200 Merge pull request #1112 from instadeepai/feat/github-actions-uv feat: uv for github actions commit dfef2387f37331e94daa3547eba6ff3a173adaca Author: Sasha Abramowitz <[email protected]> Date: Fri Oct 25 09:41:39 2024 +0200 chore: pre-commit autoupdate commit d221a85656ad7f5d592f3a05348d137e40ae62d7 Author: Sasha Abramowitz <[email protected]> Date: Fri Oct 25 09:33:10 2024 +0200 feat: uv for github actions commit ee3aff6a54d62ed1fd96215a0a818d137de8dcc1 Author: Ruan de Kock <[email protected]> Date: Thu Oct 24 18:04:41 2024 +0200 feat: use model params and optimiser state directly instead of named tuples commit f205b9edc2bcdf1c3188ef016b094d7d95bc6d72 Author: Ruan de Kock <[email protected]> Date: Thu Oct 24 16:57:35 2024 +0200 feat: use .at[].set() with drop instead of jax.lax.cond to update shifted actions commit 26654b8a77b6fdfdbfde76ecf8bcd6d2a24cbaab Author: Ruan de Kock <[email protected]> Date: Thu Oct 24 16:32:51 2024 +0200 feat: use make mlp method commit 91391c7a8d01b9748ce5c5447bfb3df6825a8fcf Author: Ruan de Kock <[email protected]> Date: Thu Oct 24 16:04:45 2024 +0200 chore: output projection commit 32e458ae1185f59567ece1827d8f2fd32230ac9a Author: Ruan de Kock <[email protected]> Date: Thu Oct 24 16:03:15 2024 +0200 chore: use capital letters for dimensions commit eee0217b552eb0c010640926db5850f5ef7c19d9 Author: Ruan de Kock <[email protected]> Date: Thu Oct 24 15:48:23 2024 +0200 chore: todo about using einops in the future commit 20a10f5515fbc42929df0c036bb119800197237e Author: Ruan de Kock <[email protected]> Date: Thu Oct 24 15:25:59 2024 +0200 feat: rename dimensions commit aae87cdbe8e0dd45f43fc1f7c3a1f4cf01c3ce41 Author: Ruan de Kock <[email protected]> Date: Thu Oct 24 15:10:55 2024 +0200 chore: pass in less seeds commit 2fc8b929fa21f8fe42219e32ccde1c564df043dd Author: Ruan de Kock <[email protected]> Date: Thu Oct 24 14:47:35 2024 +0200 feat: split less keys commit b678bf270d44251901fbf23500200ebe8c589b3d Author: Ruan de Kock <[email protected]> Date: Thu Oct 24 14:35:52 2024 +0200 chore: linter commit 80711fd0ada3a5b17252b4d30da7b633cfb198f5 Author: Ruan de Kock <[email protected]> Date: Thu Oct 24 14:34:30 2024 +0200 feat: pass in full observation object to network commit 2bd4e2ca31c416a11f5ba1a63b9f592d37084b7b Author: Sasha Abramowitz <[email protected]> Date: Thu Oct 24 14:30:46 2024 +0200 feat: switch to pyproject and update mypy rules commit 33117027e998315d6acac9ffd3c86e4b479c05c3 Author: Ruan de Kock <[email protected]> Date: Thu Oct 24 14:03:47 2024 +0200 chore: use marlenv type commit db10ce4b005e204c911f104d1cae6017d05852f7 Author: Ruan de Kock <[email protected]> Date: Thu Oct 24 14:01:59 2024 +0200 chore: don't check action space type on strings commit eedc8d75aa82d62397ce79b1a0068658b7423c4f Author: Ruan de Kock <[email protected]> Date: Thu Oct 24 13:56:45 2024 +0200 chore: rename v_loc to value commit 3688e4021085cc5ba7832904db3a089d27c9cbcd Author: Ruan de Kock <[email protected]> Date: Thu Oct 24 13:53:52 2024 +0200 chore: move SwiGLU network to torsos file commit 5e2bbb580ff35786922c38b0a45b2a9d75021be1 Author: Ruan de Kock <[email protected]> Date: Thu Oct 24 13:44:24 2024 +0200 chore: expand mask dims without reshape commit 8888a5c96bf926ae484fca1ad41567321ede5203 Author: Ruan de Kock <[email protected]> Date: Thu Oct 24 13:36:09 2024 +0200 chore: remove old comments commit 3b6bd93065d9a65befb74653ce4997058ac6b6f5 Author: Sasha Abramowitz <[email protected]> Date: Wed Oct 23 17:23:55 2024 +0200 chore: minor fixes from PR review commit 2607db4ef4aec7ea25833dff56996392fcf6c594 Author: Sasha Abramowitz <[email protected]> Date: Wed Oct 23 16:16:30 2024 +0200 fix: small logger bug for arrays with a single element commit 388dc6a9f13fd2378a1bf6df122c09779139bf45 Author: Sasha Abramowitz <[email protected]> Date: Wed Oct 23 16:16:14 2024 +0200 chore: update default hasac config commit 35db17ddf5eac51826e9ad851114a59587a5c979 Author: Ruan de Kock <[email protected]> Date: Wed Oct 23 16:01:08 2024 +0200 chore: slightly more lightweight configs and comment clean up commit 7d5e2393323307580d65161867836580082b2c93 Author: Ruan de Kock <[email protected]> Date: Wed Oct 23 15:53:22 2024 +0200 Update mava/configs/system/q_learning/rec_qmix.yaml Co-authored-by: Sasha Abramowitz <[email protected]> commit a7e3734958f45919e2346e1ffc06699d5ea7b591 Author: Ruan de Kock <[email protected]> Date: Wed Oct 23 15:53:12 2024 +0200 Update mava/configs/env/smax.yaml Co-authored-by: Sasha Abramowitz <[email protected]> commit 4fcce3fbdbb0c7868b666ac9995588be8d652f9f Author: Ruan de Kock <[email protected]> Date: Wed Oct 23 15:53:01 2024 +0200 Update mava/configs/system/q_learning/rec_qmix.yaml Co-authored-by: Sasha Abramowitz <[email protected]> commit ba71cc58998241eee741bb25d3ede885ead3ad2e Author: Ruan de Kock <[email protected]> Date: Wed Oct 23 15:51:17 2024 +0200 chore: fixed update_fn return type commit f6f81e41b4c4b46a5bc176b7019e86860627dcdf Author: Ruan de Kock <[email protected]> Date: Wed Oct 23 15:49:46 2024 +0200 feat: paramterise learner state with qmix and qlearning params commit 4f2076b3c9086667eae1750373fd9b1866167c7c Author: Ruan de Kock <[email protected]> Date: Wed Oct 23 15:39:30 2024 +0200 feat: store q_error and reuse when logging commit bde58fd30547eab49cbc3eb3e5c6972a04b7237b Author: Ruan de Kock <[email protected]> Date: Wed Oct 23 15:35:10 2024 +0200 chore: comment clean up and variable renaming commit 2dcaaceb301c41091a287a5669ea63df3487c7ff Author: Ruan de Kock <[email protected]> Date: Wed Oct 23 14:55:34 2024 +0200 chore: whitespace removed commit a19d5fa684b86d69e82512522fc59a7d56a1f02f Author: Ruan de Kock <[email protected]> Date: Wed Oct 23 14:52:02 2024 +0200 Update mava/systems/q_learning/anakin/rec_qmix.py Co-authored-by: Sasha Abramowitz <[email protected]> commit cce233a2fc19a4ef914c69538e108c90259a8ee8 Author: Ruan de Kock <[email protected]> Date: Wed Oct 23 14:51:03 2024 +0200 Update mava/systems/q_learning/anakin/rec_qmix.py Co-authored-by: Sasha Abramowitz <[email protected]> commit 9294ee859c28acdaf8698a6ad01f3eb3589fda38 Author: Ruan de Kock <[email protected]> Date: Wed Oct 23 14:49:02 2024 +0200 Update mava/systems/q_learning/anakin/rec_qmix.py Co-authored-by: Sasha Abramowitz <[email protected]> commit d8d80f741fa758aa9bc71a7aeb9026b0e3f44f6b Author: Ruan de Kock <[email protected]> Date: Wed Oct 23 14:48:26 2024 +0200 Update mava/systems/q_learning/anakin/rec_qmix.py Co-authored-by: Sasha Abramowitz <[email protected]> commit 59fe5e2d145b13c839399efb6a6b9606eb1570e8 Author: Ruan de Kock <[email protected]> Date: Wed Oct 23 14:47:41 2024 +0200 Update mava/systems/q_learning/anakin/rec_qmix.py Co-authored-by: Sasha Abramowitz <[email protected]> commit 448495bb25fbe2dcdf3f82b6ed7655a9ce6bc045 Author: Ruan de Kock <[email protected]> Date: Wed Oct 23 14:45:48 2024 +0200 Update mava/systems/q_learning/anakin/rec_qmix.py Co-authored-by: Sasha Abramowitz <[email protected]> commit fece034aa964792f027380eb587f4f72951569d6 Author: Ruan de Kock <[email protected]> Date: Wed Oct 23 14:44:32 2024 +0200 Update mava/systems/q_learning/anakin/rec_qmix.py Co-authored-by: Sasha Abramowitz <[email protected]> commit 9c4aea55a0e111109b1c9f7f2641eccbe22fa68d Author: Ruan de Kock <[email protected]> Date: Wed Oct 23 14:24:03 2024 +0200 Update mava/systems/q_learning/anakin/rec_qmix.py Co-authored-by: Sasha Abramowitz <[email protected]> commit 7482c8f47d651e10dd1cccb800f726cfa30d8ba9 Author: Ruan de Kock <[email protected]> Date: Wed Oct 23 14:23:35 2024 +0200 Update mava/systems/q_learning/anakin/rec_qmix.py Co-authored-by: Sasha Abramowitz <[email protected]> commit f10b2953ad57b5319c33a53a849c6bf150b6b825 Author: Ruan de Kock <[email protected]> Date: Wed Oct 23 14:23:25 2024 +0200 Update mava/networks/torsos.py Co-authored-by: Sasha Abramowitz <[email protected]> commit 79f9858ca6cbd81e4e1b5f9acba83a666e3a0d89 Author: Ruan de Kock <[email protected]> Date: Wed Oct 23 14:21:37 2024 +0200 Update mava/configs/system/q_learning/rec_qmix.yaml Co-authored-by: Sasha Abramowitz <[email protected]> commit 034859e6a59c5397552847572e6af884a180e076 Author: Ruan de Kock <[email protected]> Date: Wed Oct 23 09:18:26 2024 +0200 chore: match exact rec_iql style commit 73c4611144457a0359c8cdb4990461c39fd9fdc8 Author: OmaymaMahjoub <[email protected]> Date: Tue Oct 22 17:22:20 2024 +0100 feat: fix pre commits commit 576d5d4b37a92fd48fe6212a82cb0397722606a0 Author: OmaymaMahjoub <[email protected]> Date: Tue Oct 22 14:15:23 2024 +0100 feat: remove parallel representation commit 067b4ef33d76d0fdd4a7afcad607d6f7913a6a14 Author: Ruan de Kock <[email protected]> Date: Tue Oct 22 14:49:42 2024 +0200 feat: chore pre-commit commit b131cb7425e2114cd7fa8efbfe6d9356bb3afe11 Author: Ruan de Kock <[email protected]> Date: Tue Oct 22 14:46:13 2024 +0200 chore: add license commit 61d70ca734d8ab426c9962a4daff57a58bbd6b89 Author: Ruan de Kock <[email protected]> Date: Tue Oct 22 14:42:31 2024 +0200 chore: remove unused network file commit 8357ef57c942d50b79f88ed1ccd2b6af197f556a Author: Ruan de Kock <[email protected]> Date: Tue Oct 22 14:01:01 2024 +0200 chore: duplicate whole info dict at the same time commit 10647450c9af5b1c8f20f2f2c29648f747257bdc Author: OmaymaMahjoub <[email protected]> Date: Tue Oct 22 12:51:41 2024 +0100 feat: clean rec sable system file and fix checkpointer commit 63723c7da531727acfa0e5967c5ffe1ea2aa1ff9 Author: Ruan de Kock <[email protected]> Date: Tue Oct 22 12:44:02 2024 +0200 chore: set correct MLP torso size in configs commit 59972c1e95e8ea1affd0781509ed9f3e96bc4a30 Author: OmaymaMahjoub <[email protected]> Date: Tue Oct 22 11:37:41 2024 +0100 feat: add util fns for acting and training plus support for chunkwise commit f0dbc65dfe174cd1ff9ee181cd5247c478187448 Author: Ruan de Kock <[email protected]> Date: Tue Oct 22 12:35:20 2024 +0200 chore: replace jax.tree_map with jax.tree.map commit 3fb530438f46009c336e225b8c09696821df0495 Author: Ruan de Kock <[email protected]> Date: Tue Oct 22 12:31:30 2024 +0200 feat: instantiate networks with hydra utils commit b7353346133df7525f760cb2dda9759f8550cbbc Author: Ruan de Kock <[email protected]> Date: Tue Oct 22 10:27:10 2024 +0200 chore: extra comment on term_or_trunc vs terminal commit 63eb99f362fdcebdffc2f669e0ff71bfb45fc90b Author: Ruan de Kock <[email protected]> Date: Tue Oct 22 10:24:36 2024 +0200 feat: add option for hard or soft target updates commit 64325d7c5715a29e083d4cf3b7f60fc36b197e16 Author: Sasha Abramowitz <[email protected]> Date: Tue Oct 22 10:23:05 2024 +0200 chore: add hasac test commit 5fcfcf2d66ac4e90faa3636dcdc03cd2dd944737 Author: Sasha Abramowitz <[email protected]> Date: Tue Oct 22 10:21:46 2024 +0200 chore: add system name commit e13a6e15c698f76f694d94c5a5c0200ea4ba82db Author: Ruan de Kock <[email protected]> Date: Tue Oct 22 10:10:48 2024 +0200 chore: type hints commit 3b8d76195d23f45b0954849cd7c04b19929299b0 Author: Ruan de Kock <[email protected]> Date: Tue Oct 22 09:53:38 2024 +0200 chore: clean up comments commit 928c9c55fbfdba2c1998280b7d430307680fea4c Author: Sasha Abramowitz <[email protected]> Date: Tue Oct 22 09:50:58 2024 +0200 fix: jax utils commit 4915b97e29e7817a0c07ae1e2035bdd0ac6dd72c Author: Ruan de Kock <[email protected]> Date: Tue Oct 22 09:45:27 2024 +0200 chore: move types to qlearning types file commit e3195becaf47aec252168e5b6ed0dedb63277a29 Author: Ruan de Kock <[email protected]> Date: Tue Oct 22 09:36:37 2024 +0200 chore: move torso and qmix network files commit cefe4da21828c223b73e2eae2cf0d575c87efc8f Author: Ruan de Kock <[email protected]> Date: Tue Oct 22 09:19:20 2024 +0200 feat: replace rec_qmix code commit 11546a22d757b095882c961de02cd2b81590b3f4 Merge: 97e23cfe 57c8e640 Author: Ruan de Kock <[email protected]> Date: Tue Oct 22 09:15:28 2024 +0200 Merge branch 'develop' into feat/merge-qmix commit 8d85d323004d6e66af4d06a25bef65cf8d985cbd Author: Ruan de Kock <[email protected]> Date: Mon Oct 21 17:48:31 2024 +0200 feat: move decoding functions to network utils commit fd09d59704e4a2bd5705a26b091a9523b78ef931 Author: Ruan de Kock <[email protected]> Date: Mon Oct 21 16:05:21 2024 +0200 feat: use get_action_head util instead of manually setting action space type commit 6fe1f9c4782b2afd8caa7210fdf2beaa7b528e5a Author: OmaymaMahjoub <[email protected]> Date: Mon Oct 21 14:54:31 2024 +0100 feat: add chunkwise timestep fn to the modular net commit 649a70ff6dedb31c1345ba0d1e3d729dfbd11bb3 Merge: 4e3bf428 57c8e640 Author: Ruan de Kock <[email protected]> Date: Mon Oct 21 15:41:09 2024 +0200 Merge branch 'develop' into feat/implement-mat commit 519025b705b0a572acebc91801e277511d85b617 Author: Sasha Abramowitz <[email protected]> Date: Mon Oct 21 15:22:22 2024 +0200 chore: update config to new mava and cleanup commit f03e6ca79a489dbe8c9a6cf1cf394c81ab59bbdf Author: OmaymaMahjoub <[email protected]> Date: Mon Oct 21 14:17:06 2024 +0100 feat: modular net sable commit 0eeaa58ad89073cc51fd64092ab595b6d3a349a5 Merge: eeda7f50 57c8e640 Author: Sasha Abramowitz <[email protected]> Date: Mon Oct 21 14:51:10 2024 +0200 Merge branch 'develop' into feat/hasac2 commit 57c8e64059bd59005d80c1c8278eef65855253b9 Merge: c4e40ce2 a14cfb2f Author: Wiem Khlifi <[email protected]> Date: Mon Oct 21 13:25:03 2024 +0100 Merge pull request #1104 from instadeepai/feat/act_head feat: set the action head automatically commit eeda7f508e13d1c291dacdd41c5efe074761f4eb Merge: f19d9bc6 c4e40ce2 Author: Sasha Abramowitz <[email protected]> Date: Mon Oct 21 13:53:07 2024 +0200 Merge branch 'develop' into feat/hasac2 commit a14cfb2fc3ca7b979a27e4076d168970789dad63 Author: WiemKhlifi <[email protected]> Date: Mon Oct 21 11:31:04 2024 +0100 feat: return action type with act head commit 97e23cfe0f8632960f5a61dd29a3e3093916f899 Author: Ruan de Kock <[email protected]> Date: Mon Oct 21 11:40:26 2024 +0200 feat: follow old qmix in trainer commit f1cb0f20fd1fcc8871539a18803fa3b61d7b6979 Author: WiemKhlifi <[email protected]> Date: Fri Oct 18 17:17:49 2024 +0100 feat: use action_sepc to select action head type commit 59d354fab785db0c9bc761498e0770880e053835 Author: OmaymaMahjoub <[email protected]> Date: Fri Oct 18 16:41:42 2024 +0100 feat: add only timestep positional encoding commit 60d8ffa9d7d30829c79692145014a58a69017156 Author: WiemKhlifi <[email protected]> Date: Fri Oct 18 15:30:32 2024 +0100 fix: update ff_ippo_store_experience file commit 8ac214a208619aca95763df15c427e8dc140db2a Author: WiemKhlifi <[email protected]> Date: Fri Oct 18 15:07:28 2024 +0100 feat: set the action head automatically based on env name commit 4e3bf428f480dc4b801f153e8b6c7e6d7a59273d Author: Ruan de Kock <[email protected]> Date: Fri Oct 18 14:21:05 2024 +0200 feat: infer batch size and num agents from obs rep instead of manually passing in commit c18e2339c203138ed429e609d9704b8f05a473a6 Author: Ruan de Kock <[email protected]> Date: Fri Oct 18 14:04:04 2024 +0200 chore: comment cleanup commit 5e233b0cc8ff49db49743de3e4706159d7dd4072 Author: Ruan de Kock <[email protected]> Date: Thu Oct 17 18:42:44 2024 +0200 feat: continuous actions training commit 562c82a22011f5c1988f77dea935a5e90753eca9 Author: Ruan de Kock <[email protected]> Date: Thu Oct 17 18:17:19 2024 +0200 feat: pass key through trainer to prepare for continuous action spaces commit b08388dc261dd9d1c64dc518c14f9570a5ab05b2 Author: Ruan de Kock <[email protected]> Date: Thu Oct 17 18:02:43 2024 +0200 feat: squeeze inside of network and not in system run file commit 42b48bb69c70f9a6101afc1f120fd8d2495c1a01 Author: OmaymaMahjoub <[email protected]> Date: Thu Oct 17 16:51:15 2024 +0100 fix: minor fix to the positional encoding for timestep commit 13e42a7fe3e4390a3a46f3168f18e16b3d7ac087 Author: OmaymaMahjoub <[email protected]> Date: Thu Oct 17 16:28:10 2024 +0100 feat: timestep encoding for rec sable commit e4bc9692667580b3fc9e1a70d9b1fa688f2e1b06 Author: Ruan de Kock <[email protected]> Date: Thu Oct 17 16:53:25 2024 +0200 feat: use jax.tree.map instead of deprecated jax.tree_map commit 2d4f7edba23d18ec1ba726e7a0745f3543d479b3 Author: Ruan de Kock <[email protected]> Date: Thu Oct 17 16:44:30 2024 +0200 chore: remove redundant obs being passed around commit 848c625cc455428fb4b35c5b43d685beaa517612 Author: Ruan de Kock <[email protected]> Date: Thu Oct 17 16:36:57 2024 +0200 feat: prepare to starting using mava discrete action head commit d589b7e3300b9ad15d838f635dd9f7014ae527a1 Author: OmaymaMahjoub <[email protected]> Date: Thu Oct 17 12:20:29 2024 +0100 feat: pos encoding setup commit 3fab043212b90875f1f27bb9e2e76289f96e318f Author: OmaymaMahjoub <[email protected]> Date: Wed Oct 16 14:59:17 2024 +0100 fix: fix args documentation for learner_fn commit b180ef2486613b8f37559b50e08482883b7ddb45 Author: OmaymaMahjoub <[email protected]> Date: Wed Oct 16 14:06:15 2024 +0100 feat: add Sable non memory commit 8d7398a24e77c6b1e6ac8a46c57b1f47da9cd8e3 Author: OmaymaMahjoub <[email protected]> Date: Wed Oct 16 12:35:57 2024 +0100 feat: add evaluator to sbale commit 094cc652adcb10dfa91233121655e6bd91ff3724 Author: OmaymaMahjoub <[email protected]> Date: Wed Oct 16 11:39:11 2024 +0100 feat: update types used for sable commit b983cca0991a75fcdedc8a564af1e762bc68ec3f Author: OmaymaMahjoub <[email protected]> Date: Tue Oct 15 14:55:06 2024 +0100 feat: minor update commit 38af5baa15259587b83a74553a34eba6007a1ad1 Merge: f44e6a5a c4e40ce2 Author: OmaymaMahjoub <[email protected]> Date: Tue Oct 15 14:42:14 2024 +0100 feat: merge develop branch commit 4964fa8b36019ad76bc0fd02e75274b4ed126ea4 Merge: c8005cb3 666660b3 Author: Ruan de Kock <[email protected]> Date: Tue Oct 15 14:57:59 2024 +0200 feat: merge in network refactor commit f44e6a5ab2164c2d918129e4d02a9613d698260d Author: OmaymaMahjoub <[email protected]> Date: Tue Oct 15 13:54:39 2024 +0100 feat: run pre commits commit 6b928c76eda224f6c25be029da23b82c7d4775e2 Author: OmaymaMahjoub <[email protected]> Date: Tue Oct 15 13:53:16 2024 +0100 feat: sable clean code and documentation (types still uncorrect commit c8005cb3c6e41260715de1b258b7296b409e68b8 Author: Ruan de Kock <[email protected]> Date: Tue Oct 15 12:47:39 2024 +0200 feat: use tfp instead of distrax commit 53dd9d7e939af98bfc8b2dd6c5939d34ca5f41b0 Author: Ruan de Kock <[email protected]> Date: Tue Oct 15 12:06:59 2024 +0200 feat: remove autoregressive scans commit ff5ec1030990a219922f35bea1d5289c7471e7a2 Author: Ruan de Kock <[email protected]> Date: Tue Oct 15 10:25:26 2024 +0200 feat: use MAT types commit 84f0852088339bfc61c42bba04c601766cefeb85 Author: Ruan de Kock <[email protected]> Date: Tue Oct 15 09:31:43 2024 +0200 feat: remove value norm commit 30d29477c3c2957527c8550c3871aac473012fd5 Author: Ruan de Kock <[email protected]> Date: Tue Oct 15 09:19:51 2024 +0200 feat: remove huber loss commit 2905604b271fcb1fd8490cfc42382147366f0673 Author: Ruan de Kock <[email protected]> Date: Tue Oct 15 09:05:30 2024 +0200 feat: add discrete MAT and training on rware commit 88a619ab21f5e79bb24be30ee7b29c945774331e Author: OmaymaMahjoub <[email protected]> Date: Mon Oct 14 17:03:46 2024 +0100 feat: clean util functions commit efcd97528a38a1cdd34bed9613c186fd61a086e6 Author: OmaymaMahjoub <[email protected]> Date: Mon Oct 14 16:19:54 2024 +0100 feat: rename sable memory to rec sable commit c15edb06e45b0149bd7d8f2684f10d0ea3845c6f Author: OmaymaMahjoub <[email protected]> Date: Mon Oct 14 15:50:33 2024 +0100 feat: add trainable sable system (unclean) to mava commit 8b1860285fba4470d69d6b5646764c6aad477724 Author: Ruan de Kock <[email protected]> Date: Mon Oct 14 15:42:20 2024 +0200 feat: set correct sequence length and reward dim in buffer init commit 72b00fdd9438d0afe998a121c475ea4b4893230c Author: OmaymaMahjoub <[email protected]> Date: Mon Oct 14 14:31:04 2024 +0100 feat: run pre commits commit 95c12657e6f6b802f81a787a7f95758898aaec2f Author: OmaymaMahjoub <[email protected]> Date: Mon Oct 14 14:20:59 2024 +0100 feat: add sable network file commit c7685edb31bd1a126f34f173506e2a15e7d900cb Author: Ruan de Kock <[email protected]> Date: Fri Oct 11 15:41:51 2024 +0200 feat: qmix training with new API commit 1c2009308b2a8891913118bdc2875fa2d97d8482 Author: Ruan de Kock <[email protected]> Date: Fri Oct 11 12:24:52 2024 +0200 feat: qmix piping through with distributional networks commit b2bd79a267589d9def756a94d21797e3b0730e64 Merge: 43f14e5e 2a1d2d8b Author: Ruan de Kock <[email protected]> Date: Fri Oct 11 11:06:48 2024 +0200 feat: merge in develop commit 43f14e5e5a5341ed5f59904252b329f18c4d8e83 Author: Ruan de Kock <[email protected]> Date: Fri Oct 11 11:05:53 2024 +0200 feat: qmix with new evaluator piping through commit 8d35f400b270ee23d9e5be05316b30a2ecd8a80b Author: OmaymaMahjoub <[email protected]> Date: Thu Oct 10 11:26:57 2024 +0100 feat: add retention file commit e767bd90381a69f58179cc023991044c812c92e8 Author: OmaymaMahjoub <[email protected]> Date: Wed Oct 9 12:22:37 2024 +0100 feat: move ff and rnn networks into a folder commit 09d5fdfbbf04aa16c18e1173655d5128cf0aeca7 Author: OmaymaMahjoub <[email protected]> Date: Wed Oct 9 12:17:20 2024 +0100 feat: add config files of sable commit f19d9bc6d8a460817723520cefb2d3ea56bbc328 Author: Sasha Abramowitz <[email protected]> Date: Wed Aug 7 13:42:53 2024 +0200 fix: optimizers for multiple parameters commit 4673da87ea6f88ac662ea69fa6633cb10b364072 Author: Sasha Abramowitz <[email protected]> Date: Wed Aug 7 13:03:08 2024 +0200 feat: grad clip + fix final return commit aad6a0eefd45d9e5f6df75251e2434367cd0fd67 Author: Sasha Abramowitz <[email protected]> Date: Wed Aug 7 12:16:04 2024 +0200 fix: evaluator working for hasac commit 658f6277f665b2b437f9ba091b1241cbb9f34d8a Author: Sasha Abramowitz <[email protected]> Date: Wed Aug 7 11:46:21 2024 +0200 feat: hasac
RuanJohn
approved these changes
Nov 13, 2024
There was a problem hiding this comment.
Thank you for all the effort and work on this @Louay-Ben-nessir! Let's go! 🔥
sash-a
added a commit
that referenced
this pull request
Jan 10, 2025
- Merge branch develop into seb-ff-ippo-only - fix: rec_qmix import - chore: pre-commits - Merge branch develop into seb-ff-ippo-only - fix: action_head parameters for all systems - chore: pre-commits - fix: sebulba compatiable get_action_head - Squashed commit of the following: - fix: smaclite win rate tracking - chore: bunch of minor changes - chore: pre-commits - fix: removed axis swaping & wrapper rename - fix: Metric tracking more aligned with Jumanji - chore: removed learner accumulation - chore: bunch of minor changes and fixes - fix: give each learner a unique random key - fix: random segfault - chore: pre-commits - feat: support for smac - fix: start actors simultaneously to avoid deadlocks - Merge branch develop into seb-ff-ippo-only - chore: minor env typing fixes - feat: better env creation and safer sharding - fix: align gym config with other configs - fix: key use in actor loss - feat: shard_map working - feat: shardmap almost working - fix: timestep calculation with accumulation - Merge branch seb-ff-ippo-only of github.com:Louay-Ben-nessir/Mava into seb-ff-ippo-only - feat: jit evaluation on cpu - feat: learner env accumulation - fix: change to using gym.make to create envs and fix StepType - fix: possible off by one fix - chore: use orginal rware and lbf - fix: create envs in main thread to avoid deadlocks - chore: better graceful exit - chore: remove some more device transfers - feat: avoid unecessary host-device transfers - fix: safer pipeline.clear() - fix: reshape with multiple learners and system name - fix: update configs to match latest mava - Merge branch feat/sebulba_arch into seb-ff-ippo-only - Merge branch develop into feat/sebulba_arch - chore: a few minor changes to code style - feat: minor refactor to sebulba utils - fix: removed depricated gymnasium import - fix: jumanji - fix: updated to work with the latest gymnasium - chore: loss unpacking - fix: wasting samples - fix: deadlock in pipeline - feat: pass timestep instead of obs and done and fix potential race condition in pipeline - chore: very nitpicky clean ups - fix: changed the timestep discount - chore : better error messeages - fix: prevent the pipeline from stalling and a lot of cleanup - chore : various changes - chore: code cleanup - fix: fixed stalling at the end of training - chore: config file changes - chore: removed unused eval type - feat: shared time steps checker - chore: code cleanup and sps calcs and learner threads - feat : major code restructer, non-blocking evalutors - Merge remote-tracking branch upstream/feat/sebulba_arch into seb-ff-ippo-only - Merge pull request #1094 from Louay-Ben-nessir/chore--sebulba-arch-update - pre-commit - chore: pre-commits - Merge remote-tracking branch upstream/develop into chore--sebulba-arch-update - Merge pull request #1090 from Louay-Ben-nessir/chore--anakin-and-sebulba-folders - Merge pull request #4 from Louay-Ben-nessir/feat-sebulba-gym-wrapper - fix: fixed the logging deadlock for sebulba - fixed: annotations and add agent id spaces - chore: minor changes - fix: Async worker auto-resetting - fix: LBF import - fix: config file fixes - chore: pre-commits - fix: config changes - fix: env wrappers fix - fix: removed deprecated jax call - folder re-structuring - update the gym wrappers - feat: restructured the folders - chore: comments - chore : annotation - chore: bunch of minor changes - fix: better agent ids wrapper? - fix: rware import - fix: config file fixes - chore: pre-commits and annotaions - feat: using gymnasium async worker - feat: generic gym wrapper - fix; moved from gym to gymnasium - chore: config files rename - chore: renamed arch_name to architecture_name - chore: pre-commits - fix: more config changes - chore: pre-commits - fix: configs revamp - fix: sum the rewards when using a shared reward - chore: arch_name for anakin - fix: config and imports for anakin q_learning and sac - fix: seeds need to python arrays not np arrays - fix: added missing lbf import - fix: sync neptune logging for sebulba to avoid stalling - feat : lbf - feat: LBF and reproducibility - chore: pre-commits - chore: pre-commits - fix: allow for reproducibility - fix: imports and config paths in systems - chore: created the anakin and sebulba folders - chore : pre commit - fix: fixed the num evals cacls - fix: fix the num_updates_in_eval in the last eval - chore: pre-commits - feat: sebulba ff_ippo - chore: removed unused config file - chore : pre-commits and some comments - chore: clean up & updated the code to match the sebulba-ff-ippo branch - fix: removed the lbf import/wrapper - feat: ff_mappo and rec_ippo in sebulba - fix: removed the sebulba spesifique types - feat: mappo + removed sebulba specifique types and made the rware wrapper generic - chore : code cleanup + comments + added checkpoint save - fix: num_updates and code refactoring - fix: batch size calc for multiple devices - fix: logging and added LBF - feat: fulll sebulba functional - fix: changed the anakin ppo type import - fix: fixed the training and added training logger - fix: fixed function calls - fix: changes the env creation - feat: initial learner / training loop - feat: init sebulba ippo - feat: gym metric tracker wrapper - chore: removed async gym wrapper - fix: info only contains the action_mask and reformated (n_agents, n_env) ->(n_env, n_agents) - fix: fixed the async env wrapper - feat: async env wrapper , changed the gym wrapper to rware wrapper - fix: handling rware reset function - fix: various minor fixes - fix: gymV26 compatability wrapper - fix: fixed the async env creation - chore: pre-commit - fix: Create the gym wrappers directly - fix: merged the observations and action mask - chore : pre-commit hooks - feat: gym wrapper Co-authored-by: Sasha Abramowitz <[email protected]> Co-authored-by: Omayma Mahjoub <[email protected]>
sash-a
added a commit
that referenced
this pull request
Jan 10, 2025
- Merge branch develop into seb-ff-ippo-only - fix: rec_qmix import - chore: pre-commits - Merge branch develop into seb-ff-ippo-only - fix: action_head parameters for all systems - chore: pre-commits - fix: sebulba compatiable get_action_head - Squashed commit of the following: - fix: smaclite win rate tracking - chore: bunch of minor changes - chore: pre-commits - fix: removed axis swaping & wrapper rename - fix: Metric tracking more aligned with Jumanji - chore: removed learner accumulation - chore: bunch of minor changes and fixes - fix: give each learner a unique random key - fix: random segfault - chore: pre-commits - feat: support for smac - fix: start actors simultaneously to avoid deadlocks - Merge branch develop into seb-ff-ippo-only - chore: minor env typing fixes - feat: better env creation and safer sharding - fix: align gym config with other configs - fix: key use in actor loss - feat: shard_map working - feat: shardmap almost working - fix: timestep calculation with accumulation - Merge branch seb-ff-ippo-only of github.com:Louay-Ben-nessir/Mava into seb-ff-ippo-only - feat: jit evaluation on cpu - feat: learner env accumulation - fix: change to using gym.make to create envs and fix StepType - fix: possible off by one fix - chore: use orginal rware and lbf - fix: create envs in main thread to avoid deadlocks - chore: better graceful exit - chore: remove some more device transfers - feat: avoid unecessary host-device transfers - fix: safer pipeline.clear() - fix: reshape with multiple learners and system name - fix: update configs to match latest mava - Merge branch feat/sebulba_arch into seb-ff-ippo-only - Merge branch develop into feat/sebulba_arch - chore: a few minor changes to code style - feat: minor refactor to sebulba utils - fix: removed depricated gymnasium import - fix: jumanji - fix: updated to work with the latest gymnasium - chore: loss unpacking - fix: wasting samples - fix: deadlock in pipeline - feat: pass timestep instead of obs and done and fix potential race condition in pipeline - chore: very nitpicky clean ups - fix: changed the timestep discount - chore : better error messeages - fix: prevent the pipeline from stalling and a lot of cleanup - chore : various changes - chore: code cleanup - fix: fixed stalling at the end of training - chore: config file changes - chore: removed unused eval type - feat: shared time steps checker - chore: code cleanup and sps calcs and learner threads - feat : major code restructer, non-blocking evalutors - Merge remote-tracking branch upstream/feat/sebulba_arch into seb-ff-ippo-only - Merge pull request #1094 from Louay-Ben-nessir/chore--sebulba-arch-update - pre-commit - chore: pre-commits - Merge remote-tracking branch upstream/develop into chore--sebulba-arch-update - Merge pull request #1090 from Louay-Ben-nessir/chore--anakin-and-sebulba-folders - Merge pull request #4 from Louay-Ben-nessir/feat-sebulba-gym-wrapper - fix: fixed the logging deadlock for sebulba - fixed: annotations and add agent id spaces - chore: minor changes - fix: Async worker auto-resetting - fix: LBF import - fix: config file fixes - chore: pre-commits - fix: config changes - fix: env wrappers fix - fix: removed deprecated jax call - folder re-structuring - update the gym wrappers - feat: restructured the folders - chore: comments - chore : annotation - chore: bunch of minor changes - fix: better agent ids wrapper? - fix: rware import - fix: config file fixes - chore: pre-commits and annotaions - feat: using gymnasium async worker - feat: generic gym wrapper - fix; moved from gym to gymnasium - chore: config files rename - chore: renamed arch_name to architecture_name - chore: pre-commits - fix: more config changes - chore: pre-commits - fix: configs revamp - fix: sum the rewards when using a shared reward - chore: arch_name for anakin - fix: config and imports for anakin q_learning and sac - fix: seeds need to python arrays not np arrays - fix: added missing lbf import - fix: sync neptune logging for sebulba to avoid stalling - feat : lbf - feat: LBF and reproducibility - chore: pre-commits - chore: pre-commits - fix: allow for reproducibility - fix: imports and config paths in systems - chore: created the anakin and sebulba folders - chore : pre commit - fix: fixed the num evals cacls - fix: fix the num_updates_in_eval in the last eval - chore: pre-commits - feat: sebulba ff_ippo - chore: removed unused config file - chore : pre-commits and some comments - chore: clean up & updated the code to match the sebulba-ff-ippo branch - fix: removed the lbf import/wrapper - feat: ff_mappo and rec_ippo in sebulba - fix: removed the sebulba spesifique types - feat: mappo + removed sebulba specifique types and made the rware wrapper generic - chore : code cleanup + comments + added checkpoint save - fix: num_updates and code refactoring - fix: batch size calc for multiple devices - fix: logging and added LBF - feat: fulll sebulba functional - fix: changed the anakin ppo type import - fix: fixed the training and added training logger - fix: fixed function calls - fix: changes the env creation - feat: initial learner / training loop - feat: init sebulba ippo - feat: gym metric tracker wrapper - chore: removed async gym wrapper - fix: info only contains the action_mask and reformated (n_agents, n_env) ->(n_env, n_agents) - fix: fixed the async env wrapper - feat: async env wrapper , changed the gym wrapper to rware wrapper - fix: handling rware reset function - fix: various minor fixes - fix: gymV26 compatability wrapper - fix: fixed the async env creation - chore: pre-commit - fix: Create the gym wrappers directly - fix: merged the observations and action mask - chore : pre-commit hooks - feat: gym wrapper Co-authored-by: Sasha Abramowitz <[email protected]>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
What?
Implemented the ff_ippo system for sebulba
Why?
Integrate Sebulba's architecture due to its effectiveness in scenarios involving non-jitted/non-jax environments.
How?
By implementing the ff_ippo system.
Extra
The other PPO systems will be added once this PR is merged.
This PR was built on top of #1080