FreeToken Weight (FTW) Storage Format
The FreeToken Weight (FTW) storage format is a pre-processed weight file format that stores expert parameters pre-merged into the normalized runtime bank layout ahead of time. By preparing the weights in advance according to the engine's exact runtime memory arrangement, the system completely bypasses tensor discovery, unpacking, and runtime weight repacking during startup. At engine launch, expert weights are loaded via parallel direct I/O as aligned chunks directly into pre-allocated, exact-size host banks, allowing rapid cold bootstrap prior to memory pinning.
0
1
Tags
Prep Sessions
Edge-Native Mixture-of-Experts Serving with FreeToken @ University of Michigan - Ann Arbor
Ch.4 Platform Adaptation and Performance Evaluation - Edge-Native Mixture-of-Experts Serving with FreeToken @ University of Michigan - Ann Arbor
Expert Storage Formats and Platform Adaptation - Edge-Native Mixture-of-Experts Serving with FreeToken @ University of Michigan - Ann Arbor
Related
Normalized Expert Bank Representation
FreeToken Weight (FTW) Storage Format
Pure-CPU MoE Fallback Backend
Direct Host Layout Loading with Deferred Memory Pinning
Fused Multi-Bank Expert Transfer via Device-Resident Work Lists
In a normalized expert bank representation, which formula defines the leading dimension index for an expert with index in layer , given routed experts per layer?
According to the normalized expert bank structure, what collectively constitutes one full expert across the unified banks?
FreeToken Weight (FTW) Storage Format
Learn After
By preparing weight files in advance to match the engine's exact runtime memory arrangement, which set of operations does the FreeToken Weight (FTW) format completely bypass during engine startup?
Which input/output method is used at engine launch to transfer aligned weight chunks directly into host banks under the FTW format?