Commit graph

857 commits

Author SHA1 Message Date
Jai Suphavadeeprasit
f76f9d1596 cleanup 2025-08-19 12:03:13 -04:00
pre-commit-ci[bot]
62b72589c6 [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
2025-08-19 12:03:13 -04:00
Jai Suphavadeeprasit
e55a7a0100 add_danger 2025-08-19 12:03:13 -04:00
teknium
bed7ddcb95 add more default categories 2025-08-19 12:03:13 -04:00
teknium
39f0103313 fix dataset 2025-08-19 12:03:13 -04:00
teknium
ff7a2569dc update default max_toks 2025-08-19 12:03:13 -04:00
teknium
69135320b4 initial refusalbenchv2 2025-08-19 12:03:13 -04:00
hjc-puro
8c3ea257cd
Merge pull request #235 from NousResearch/bibtex
Update bibtex
2025-08-18 13:39:55 -04:00
dmahan93
83003d0988
Merge pull request #238 from NousResearch/pre-commit-ci-update-config
[pre-commit.ci] pre-commit autoupdate
2025-08-18 12:39:04 -05:00
pre-commit-ci[bot]
5b1fb70132
[pre-commit.ci] pre-commit autoupdate
updates:
- [github.com/astral-sh/ruff-pre-commit: v0.12.8 → v0.12.9](https://github.com/astral-sh/ruff-pre-commit/compare/v0.12.8...v0.12.9)
2025-08-18 16:39:43 +00:00
dmahan93
4e3ad29fac
Merge pull request #237 from NousResearch/log-errors-from-collect-trajectories
add error logging to collect_trajectories so they don't fail silently
2025-08-15 16:56:37 -05:00
Dakota
11f1303da0 add error logging to collect_trajectories so they don't fail silently 2025-08-15 16:34:21 -05:00
dmahan93
628bd3d2ad
Merge pull request #236 from brawncode/patch-1
fix: division-by-zero in gradient calculation
2025-08-14 13:18:39 -05:00
Brawn
eb179e7fca
Update grpo.py 2025-08-14 20:20:41 +03:00
Brawn
6dccdcc67e
fix: division-by-zero in gradient calculation 2025-08-14 14:33:46 +03:00
hjc-puro
5faa2be188
update bibtex 2025-08-14 02:57:42 -04:00
dmahan93
9c472d8439
Merge pull request #231 from NousResearch/pre-commit-ci-update-config
[pre-commit.ci] pre-commit autoupdate
2025-08-12 10:59:00 -05:00
dmahan93
fa6ec96f48
Merge pull request #233 from rejected-l/main
build: update checkout action to v5
2025-08-12 10:58:37 -05:00
Rej Ect
6c34126cb9
build: update checkout action to v5 2025-08-12 15:20:27 +03:00
shannonsands
46f0602227
Diplomacy trainer env (#227)
* minimal implementation, simplified challenge registry

* need game save logic

* fixed challenge gen, works with local test

* updated challenge gen with wider ranges, working with local script

* runs working correctly, wandb stats look ok

* linting

* Add diplomacy environment with AI_Diplomacy submodule

- Add diplomacy_env_minimal.py for diplomacy game environment
- Add atropos_client_minimal.py for client interface
- Add diplomacy_local_server.py for local game server
- Add AI_Diplomacy submodule from GoodStartLabs/AI_Diplomacy
- Fix import ordering and remove unused imports

* test file working, moving to cluster to test training

* updated gitignore

* removed logs

* minor fixes, training running now

* readded proxy reg and queue system

* linting

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* queue gameid bug, refactored

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* cleaned up configs & allowed for openrouter models to be easily used

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* linting

* Remove duplicate dependencies from diplomacy requirements.txt

Only keep AI_Diplomacy-specific dependencies that aren't already in the main project

---------

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
2025-08-12 09:02:16 +10:00
pre-commit-ci[bot]
ed97dd112a
[pre-commit.ci] pre-commit autoupdate
updates:
- [github.com/pre-commit/pre-commit-hooks: v5.0.0 → v6.0.0](https://github.com/pre-commit/pre-commit-hooks/compare/v5.0.0...v6.0.0)
- [github.com/astral-sh/ruff-pre-commit: v0.12.7 → v0.12.8](https://github.com/astral-sh/ruff-pre-commit/compare/v0.12.7...v0.12.8)
2025-08-11 16:42:09 +00:00
dmahan93
4fe67e698d
Merge pull request #228 from NousResearch/pre-commit-ci-update-config
[pre-commit.ci] pre-commit autoupdate
2025-08-04 12:31:35 -05:00
pre-commit-ci[bot]
70043c4eb2
[pre-commit.ci] pre-commit autoupdate
updates:
- [github.com/astral-sh/ruff-pre-commit: v0.12.5 → v0.12.7](https://github.com/astral-sh/ruff-pre-commit/compare/v0.12.5...v0.12.7)
2025-08-04 16:41:05 +00:00
shannonsands
47cb15745c
Textworld minimal (#225)
* minimal implementation, simplified challenge registry

* need game save logic

* fixed challenge gen, works with local test

* updated challenge gen with wider ranges, working with local script

* runs working correctly, wandb stats look ok

* linting

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* removed unused imports

---------

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
2025-08-01 10:16:35 +10:00
Teknium
1900a577d7
Update README.md
Remove hackathon event teaser
2025-07-30 13:56:11 -07:00
Teknium
be66e120d9
Merge pull request #219 from NousResearch/add-arenahard-v1-environment
Add arena-hard v1 environment
2025-07-30 09:35:14 -07:00
pre-commit-ci[bot]
65aea8bb21 [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
2025-07-30 15:10:36 +00:00
teknium
75f1cf6d2a move eval envs to eval_environments and update readmes 2025-07-30 15:09:34 +00:00
shannonsands
9f23c732dd
qwen tokenizer wrapper & fixed jinja template for tool handling (#224)
* added qwen tokenizer wrapper & fixed jinja template for tool handling issues in the official HF one

* moved jinja template into it's own file
2025-07-30 11:57:15 +10:00
dmahan93
56fb50a503
Merge pull request #222 from NousResearch/pre-commit-ci-update-config
[pre-commit.ci] pre-commit autoupdate
2025-07-29 15:52:20 -05:00
dmahan93
9734a10290
Merge pull request #220 from Aboozle1/add-my-environment
Add Word Hunt environment
2025-07-28 12:02:01 -05:00
Aboozle1
3ce68aed38
Merge branch 'main' into add-my-environment 2025-07-28 11:50:50 -05:00
Abhaykhanna3
9d7bcc523f Fix(PR): Address reviewer feedback
- Remove redundant requirements.txt
- Fix leading newline in prompt templates
2025-07-28 11:48:02 -05:00
pre-commit-ci[bot]
d65f7f842d
[pre-commit.ci] pre-commit autoupdate
updates:
- [github.com/astral-sh/ruff-pre-commit: v0.12.4 → v0.12.5](https://github.com/astral-sh/ruff-pre-commit/compare/v0.12.4...v0.12.5)
2025-07-28 16:38:41 +00:00
Abhaykhanna3
b5234d4214 Add Word Hunt environment for training models on 4x4 letter grids
- Trie-based solver, official scoring, normalized rewards
- Configurable token limit and detailed README with dictionary download link
- Removes large Dictionary.txt from tracking and adds ignore rules
- All tests pass and pre-commit hooks are clean
2025-07-28 00:37:36 -05:00
pre-commit-ci[bot]
4c88a4bbb9 [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
2025-07-28 01:37:03 +00:00
teknium
aaebb8d6bb linter linter 2025-07-28 01:35:49 +00:00
pre-commit-ci[bot]
cbec584202 [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
2025-07-28 01:30:27 +00:00
teknium
1d523472cc Merge branch 'add-arenahard-v1-environment' of https://github.com/NousResearch/atropos into add-arenahard-v1-environment 2025-07-28 01:29:52 +00:00
teknium
20565d8abc update judge confs so it can use any judge model 2025-07-28 01:29:50 +00:00
pre-commit-ci[bot]
041a70d891 [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
2025-07-27 03:39:42 +00:00
teknium
e6de7bb432 lint 2025-07-27 03:39:05 +00:00
pre-commit-ci[bot]
52b505296c [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
2025-07-27 02:52:39 +00:00
teknium
a0979eb08e add readme section 2025-07-27 02:46:51 +00:00
teknium
31b0c6f66d Add arena-hard v1 environment 2025-07-26 21:17:00 +00:00
Teknium
b272b7bae9
Merge pull request #218 from NousResearch/pairwise-judgement-env-updates
Pairwise Judgement Environment - improve dataloading, ctx len
2025-07-26 14:14:24 -07:00
pre-commit-ci[bot]
65682d160a [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
2025-07-26 21:13:05 +00:00
teknium
aa66b09c13 make linter happy 2025-07-26 21:12:30 +00:00
pre-commit-ci[bot]
a2e14cf50c [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
2025-07-26 21:08:01 +00:00
teknium
1039c3d360 improve dataloading, ctx len 2025-07-26 21:06:45 +00:00