Skip to main content

Module ftw_loader

Module ftw_loader 

Source
Expand description

FTW multi-shard loader is file/memmap2-backed — native hosts only. FTW (FreeToken Weight) format loader and multi-shard memory mapper (Work Package F9/F10).

Loads freetoken_weight.json and memory-maps .ftw binary shards, providing zero-copy access to shared weights and MoE expert banks (256 experts per layer).

Structs§

FtwExpertBankSlice
Slices within the memory-mapped shards for a single expert in a given layer.
FtwExpertData
Zero-copy byte slices for one expert’s NVFP4 tensors.
FtwManifest
FtwModelPackage
Container owning memory-mapped FTW shards and tensor index.
FtwShardEntry
FtwTensorEntry