Skip to main content

clika_runtime.nn

clika_runtime.nn: build models as module trees.

Compose layers by attribute assignment inside a :class:Module subclass, define forward, and call the module like a function; state_dict / load_state_dict move checkpoints by dotted names, to places the tree, and print(module) shows it. Stateless operations (activations, losses, padding, attention) live in :mod:clika_runtime.nn.functional.

The layer set covers both weight postures: dense (:class:Linear, :class:Conv, :class:ConvTranspose, :class:Embedding, :class:LayerNorm, :class:RMSNorm, :class:MoE) and quantized (:class:QLinear, :class:QLinearWoQ, :class:QConv, :class:QConvWoQ, :class:QMoEWoQ); :class:Embedding serves a quantized table through the same class. :class:KVCache is the serving-side key/value cache with its configuration vocabulary.

NameKind
Convclass
ConvTransposeclass
Embeddingclass
KVBlockSchemeclass
KVCacheclass
KVCacheConfigclass
KVCacheModeclass
KVLayerSpecclass
KVQuantSpecclass
LayerNormclass
Linearclass
LoadResultclass
MoEclass
Moduleclass
ModuleListclass
PagedParamsclass
Parameterclass
QConvclass
QConvWoQclass
QLinearclass
QLinearWoQclass
QMoEWoQclass
RMSNormclass
Sequentialclass
StepIndicesclass
WeightResidencyclass