MCPcopy Create free account
hub / github.com/apple/axlearn / init_fn

Function init_fn

axlearn/common/optimizers.py:96–97  ·  view source on GitHub ↗
(params)

Source from the content-addressed store, hash-verified

94 kwargs = {k: _to_partitioned_transformation(v) for k, v in kwargs.items()}
95
96 def init_fn(params):
97 return {k: v.init(params) for k, v in kwargs.items()}
98
99 def update_fn(
100 updates: NestedTensor, state: dict[str, Any], params: NestedOptParam

Callers

nothing calls this directly

Calls 10

opt_param_valuesFunction · 0.85
SkipClipStateClass · 0.85
ParamEmaStateClass · 0.85
ScaleByLionStateClass · 0.85
_AdastarStateClass · 0.85
itemsMethod · 0.80
mapMethod · 0.80
_to_stateFunction · 0.70
initMethod · 0.45

Tested by

no test coverage detected