Method run

learners/uniform_quantization/bit_optimizer.py:128–135 · view source on GitHub ↗

get the bit allocation strategy either with RL or not

(self)

Source from the content-addressed store, hash-verified

source not stored for this graph (policy: none)

__init__Method · 0.95

find_unquant_act_nodesFunction · 0.45

trainMethod · 0.45

evaluateMethod · 0.45

trainMethod · 0.45

evaluateMethod · 0.45

__init__Method · 0.45

__calc_optimal_bitsMethod · 0.45

__calc_rollout_rewardMethod · 0.45

__layerwise_finetuneMethod · 0.45

__global_finetuneMethod · 0.45

__calc_rollout_actionsMethod · 0.45

__calc_optimal_bitsMethod · 0.95

no test coverage detected