MCPcopy Create free account

hub / github.com/arcprize/hierarchical-reasoning-model-analysis / types & classes

Types & classes38 in github.com/arcprize/hierarchical-reasoning-model-analysis

⨍ Functions 116◇ Types & classes 38
↓ 8 callersClassCastedLinear
models/layers.py:43
↓ 4 callersClassCastedEmbedding
models/layers.py:62
↓ 4 callersClassPuzzleDatasetMetadata
dataset/common.py:12
↓ 3 callersClassHierarchicalReasoningModel_ACTV1InnerCarry
models/hrm/hrm_act_v1.py:16
↓ 3 callersClassHierarchicalReasoningModel_ACTV2InnerCarry
models/hrm/hrm_act_v2.py:31
↓ 3 callersClassPretrainConfig
pretrain.py:45
↓ 2 callersClassARCPuzzle
dataset/build_arc_dataset.py:35
↓ 2 callersClassAttention
models/layers.py:98
↓ 2 callersClassCastedSparseEmbedding
models/sparse_embedding.py:11
↓ 2 callersClassHierarchicalReasoningModel_ACTV1Block
models/hrm/hrm_act_v1.py:62
↓ 2 callersClassHierarchicalReasoningModel_ACTV1Carry
models/hrm/hrm_act_v1.py:22
↓ 2 callersClassHierarchicalReasoningModel_ACTV1ReasoningModule
models/hrm/hrm_act_v1.py:88
↓ 2 callersClassHierarchicalReasoningModel_ACTV2Carry
models/hrm/hrm_act_v2.py:36
↓ 2 callersClassRotaryEmbedding
models/layers.py:80
↓ 2 callersClassSwiGLU
models/layers.py:139
↓ 2 callersClassTrainState
pretrain.py:86
↓ 1 callersClassCastedSparseEmbeddingSignSGD_Distributed
models/sparse_embedding.py:41
↓ 1 callersClassEvaluatorConfig
pretrain.py:39
↓ 1 callersClassHierarchicalReasoningModel_ACTV1Config
models/hrm/hrm_act_v1.py:31
↓ 1 callersClassHierarchicalReasoningModel_ACTV1_Inner
models/hrm/hrm_act_v1.py:104
↓ 1 callersClassHierarchicalReasoningModel_ACTV2Block
models/hrm/hrm_act_v2.py:74
↓ 1 callersClassHierarchicalReasoningModel_ACTV2Config
models/hrm/hrm_act_v2.py:45
↓ 1 callersClassHierarchicalReasoningModel_ACTV2ReasoningModule
models/hrm/hrm_act_v2.py:103
↓ 1 callersClassHierarchicalReasoningModel_ACTV2_Inner
models/hrm/hrm_act_v2.py:119
↓ 1 callersClassPuzzleDataset
puzzle_dataset.py:53
↓ 1 callersClassPuzzleDatasetConfig
puzzle_dataset.py:41
Class
assets/npyjs.js:1
ClassACTLossHead
models/losses.py:40
ClassARC
evaluators/arc.py:40
ClassArchConfig
pretrain.py:32
ClassAugmentationAblationARC
ARC evaluator that samples limited augmentations per base task. Uses identical logic to parent class but samples predictions in result().
evaluators/arc_augmentation_ablation.py:21
ClassDataProcessConfig
dataset/build_maze_dataset.py:22
ClassDataProcessConfig
dataset/build_arc_dataset.py:17
ClassDataProcessConfig
dataset/build_sudoku_dataset.py:18
ClassHierarchicalReasoningModel_ACTV1
ACT wrapper.
models/hrm/hrm_act_v1.py:218
ClassHierarchicalReasoningModel_ACTV2
ACT wrapper.
models/hrm/hrm_act_v2.py:247
ClassLossConfig
pretrain.py:26
Classnpyjs
assets/npyjs.js:1