joesharratt1229
d0ef136d5b
Feat/intragen experiments ( #414 )
...
* added curriculum
* readapted readme
* corrected small errors
* Delete eval/eval/r1/algorithmic/word_sorting.json
* removed redundant argument
* added spell
* removed duplicated fit
* changed config
* added composite changes
* added composite changes
* updated yaml
* added spell backward
* updated read me
* added qwen2.5
* added
* Add files via upload
* updated missing trainer func
* updated curr
* updated spell back
* updated correctness score func
* updated configs
* added local evals
* added updates
* updated datasets
* added fsdp to hf utility
* added algorithmic qwen 3b yaml
* updated read me
* updated configs
* added preappend token
* updated with thinking token
* updated test score board
* resolved comments
* added evaluation scripts
* removed results from pr
* added config
* added partial reward scoring
* added evaluation composites
* added training configs
* added games eval
* added rubriks cube
* resolved merge cinflicts
* added games config
* added latest eval configs
* updated strucutre
* Delete training/evaluations/eval_graphs_composite.yaml
---------
Co-authored-by: joesharratt1229 <joesharrat1229@gmail.com>
2025-04-16 08:04:52 +02:00
Oliver Stanley
224532f12a
first inter-domain generalisation experiments ( #412 )
...
* tweak len reward
* first inter-generalisation experiment config
* update inter algorithmic config
* default to empty config
* fix typo
* change config to match experiment script
* long prompt fixes
* algorithmic training config tweaks
* imports
* update algorithmic training cfgs
* first logic composite config
* fix dset name
* tweaks
* fix syllogisms dataset
* rm temp print
* initial algebra config
* algebra cfg tweaks
* add gc
* add initial games cfg
* rename games cfg
* fix dset name
* fix sokoban metadata
* remove boxnet
* games cfg tweak
2025-04-14 21:06:40 +01:00
Oliver Stanley
ff5407f766
fix boxnet scoring function ( #420 )
...
* avoid error print on invalid move format in tower of hanoi
* avoid boxnet error on hallucinated starting posittions
2025-04-14 20:44:24 +01:00
Zafir Stojanovski
290bfc4fdd
(evals): Medium configs ( #415 )
...
* updated medium configs
* fix problematic curriculum values / small issues causing exceptions to be raised
* optimus alpha config
* all configs so far
* fix tests
2025-04-14 08:25:31 +02:00
Zafir Stojanovski
dced3bfc45
fix(curriculum): Make boundaries in curriculum more sensible ( #407 )
...
* init
* fix tests
* unify codeio
* filtered for libraries not present in reasoning-gym
* fix more bounds
* puzzle24
* knight swap curriculum
* fix number sorting
* fix attributes
* add validation of config in creation of dataset
* dry run for instantiating and validating the datasets
* remove unused imports
* fix curriculum tests to reference newly updated attribute names
2025-04-04 20:24:14 +02:00
Zafir Stojanovski
ce0a6c4878
fix(envs): Add source dataset and index to metadata ( #388 )
...
* add source dataset and index to metadata
* fix typo
* fix coach class and its test
2025-03-20 11:12:14 +00:00
Oliver Stanley
7475a20700
include ranges rather than sampled values in difficulty metadata dicts ( #387 )
...
* update difficulty metadata for logic datasets
* update difficulty metadata for graph datasets
* update difficulty metadata for geometry datasets
* update difficulty metadata for games datasets
* update difficulty metadata for cognition datasets
* update difficulty metadata for arithmetic datasets
* update difficulty metadata for arc datasets
* update difficulty metadata for algorithmic datasets
* update difficulty metadata for algebra datasets
* use tuples
* update tests
* update tests
2025-03-20 10:27:03 +01:00
Andreas Köpf
d2c895f1d3
Refactor Curriculum Attributes ( #335 )
...
* remove min_value from AttributeDefinition
* remove type from AttributeDefinition
* Add CurriculumContext
* add ensure_interval option for RangeAttributes
* docs: Add legend explaining curriculum indicators in dataset gallery
* update GALLERY.md
2025-03-16 15:40:28 +01:00
Andreas Köpf
d6f399b8e4
Add eval configs, small fixes to eval script & rush-hour score_answer
2025-03-16 09:18:05 +01:00
Oliver Stanley
a71994ad03
add rush hour curriculum ( #362 )
...
* add rush hour curriculum
* add to __init__.py
2025-03-14 16:11:23 +01:00
Oliver Stanley
8f8bd9d756
add sudoku curriculum ( #360 )
...
* sudoku curriculum
* add test
* add to __init__.py
2025-03-14 16:09:49 +01:00
joesharratt1229
fa7d8e66b3
added boxnet generator and curr ( #364 )
2025-03-14 08:48:57 +01:00
Zafir Stojanovski
c81036004f
maze curriculum ( #343 )
2025-03-13 21:01:48 +01:00
Rich Jones
fe0f9ae114
Fix Rush Hour ( #336 )
...
* think it fixes rush hour
* explain walls
2025-03-13 20:42:40 +01:00
joesharratt1229
3c39cbda40
added sokoban dataset ( #325 )
2025-03-11 00:21:03 +01:00
joesharratt1229
e9944149bd
added tsumego curric ( #323 )
2025-03-11 00:19:55 +01:00
joesharratt1229
e01910254d
added futoshiki and tower hanou ( #316 )
...
* added futoshiki and tower hanou
* corrected failed unit tests
2025-03-11 00:12:32 +01:00
joesharratt1229
30f5d823da
Curriculum/emoji mystery ( #315 )
...
* added emoji curriculum
* updated metadata
* added curriculum to register
2025-03-11 00:11:27 +01:00
Zafir Stojanovski
e4b13bf51f
mini sudoku curriculum ( #311 )
2025-03-10 00:29:53 +01:00
Andreas Koepf
0a35e608ec
register MahjongPuzzleCurriculum
2025-03-07 19:17:04 +01:00
Zafir Stojanovski
b915565c0d
add difficulty where possible ( #274 )
2025-03-07 19:01:26 +01:00
Andreas Köpf
c69bc5d4e6
Basic curriculum ( #198 )
...
* feat: Add optional curriculum support to dataset registration and creation
* docs: Add docstrings to create_curriculum() and register_dataset()
* feat: Add curriculum configuration classes for CurriculumExperiment
* feat: Add weight parameter to CurriculumAttributeConfig and use in DatasetSpec
* refactor: Simplify CurriculumAttributeConfig with "*" attribute level support
* test: Add unit tests for CurriculumExperiment class
* feat: Add from_yaml() method to CurriculumExperimentConfig with unit test
2025-03-07 11:22:12 +01:00
joesharratt1229
f3ee9a91a2
Added puzzle24 closes #208 ( #268 )
...
* added puzzle24
2025-03-05 22:36:37 +01:00
joesharratt1229
e30be066ec
Fixed countdown score_answer ( #265 )
...
* fixed countdown score ans
* checked solution uses all numbers
2025-03-05 22:30:12 +01:00
Zafir Stojanovski
d0a42116fb
feat(env): Mahjong Puzzle Curriculum ( #263 )
...
* mahjong curriculum
* typo
* update levels
2025-03-05 22:28:02 +01:00
Zafir Stojanovski
8ecc723607
feat(env): NQueens Curriculum ( #262 )
...
* curriculum & tests
2025-03-05 15:05:17 +01:00
Andreas Köpf
5d7fbac0ad
Minor question template & score_answer improvements ( #261 )
...
* math prompt improvements
* ignore brackets in complex_arithmetic results
* improve additional instruction in prompt of polynomial_equations
* more strict tests for score_answer in polynomial_equations
* simplify special reward handling
* fix test_intermediate_integration
* fix sokoban dataset
* add common dataset score_answer consistency test
2025-03-04 21:55:09 +01:00
Zafir Stojanovski
01e1c8f9af
fix: Unify Prompts ( #254 )
...
* remove cot
* fix prompt template
* fix pool matrix
* spiral matrix fixed
2025-03-03 21:55:53 +01:00
Zafir Stojanovski
9c581f1be1
Mahjong Puzzle ( #241 )
...
* mahjong
2025-03-01 16:27:26 +01:00
Andreas Koepf
6511725711
add markdown tripple backticks around tsumego board
2025-02-26 19:39:05 +01:00
Andreas Koepf
f97bf94caa
fix & simplify score_answer() of TsumegoDataset
2025-02-26 19:04:30 +01:00
Andreas Koepf
eeb9fa31d5
more native type hints
2025-02-21 21:23:14 +01:00
Andreas Koepf
51808210aa
add markdown tripple backtick code block for emoji_mystry hint
2025-02-21 21:06:07 +01:00
Andreas Köpf
c56045b9a7
Merge branch 'main' into feat/emoji-mystery
2025-02-21 20:58:39 +01:00
joesharratt1229
1fb73011f8
added answer format spec in prompt
2025-02-21 18:03:05 +00:00
joesharratt1229
5e64d1c24c
added emoji dataset
2025-02-21 17:57:41 +00:00
Andreas Köpf
b59ccdefa2
Merge pull request #178 from olliestanley/feature/unsloth-train
...
Add minimal working GRPO training example with Unsloth
2025-02-21 15:37:24 +01:00
Andreas Koepf
3e7ff3b084
use native types List->list, Dict->dict, Set->set, Tuple->tuple
2025-02-21 15:15:38 +01:00
Oliver
e26161713e
Answer scoring fixes to address edge cases
2025-02-20 22:04:01 +00:00
Andreas Köpf
f2e86795eb
Merge pull request #161 from olliestanley/fix/sudoku-unique
...
Fix Sudoku generator for uniqueness, implement scoring
2025-02-18 22:55:43 +01:00
Oliver
75802c6a1c
Add docstring
2025-02-18 21:38:46 +00:00
Oliver
b95ed7fbad
Remove now redundant is_valid function
2025-02-18 21:37:37 +00:00
Oliver
5003357425
Remove comment
2025-02-18 21:32:15 +00:00
Oliver
56da1f5dcf
Optimise Sudoku uniqueness checks
2025-02-18 21:30:59 +00:00
Oliver
7d8fe8b5d9
Fix Sudoku generator uniqueness and scoring
2025-02-18 21:02:49 +00:00
Oliver
473368bbea
Tweak mini sudoku config
2025-02-18 18:46:14 +00:00
Oliver
fa773115b9
Tweak mini sudoku config
2025-02-18 18:43:19 +00:00
Oliver
32928bd77b
Tweak num_empty logic
2025-02-18 18:36:12 +00:00
Oliver
9e3f4e806a
Ensure unique mini sudokus
2025-02-18 18:31:30 +00:00
Oliver
c512d86a86
Cleanup question & add scoring for mini sudoku
2025-02-17 18:37:07 +00:00
