mirror of
https://github.com/FlashML-org/FreeToken.git
synced 2026-11-22 08:46:51 +08:00
* feat(gguf): serve image input on deepseek4 ggufs from a deepseek4v mmproj * fix(gguf): skip an unusable deepseek4v mmproj instead of refusing the trunk
43 lines
1.7 KiB
Python
43 lines
1.7 KiB
Python
"""DeepSeek-V4-Flash support for FreeToken.
|
|
|
|
This package ports the official ``inference/model.py`` reference (MLA + CSA/HCA
|
|
compressors + Lightning Indexer + manifold-constrained Hyper-Connections) onto
|
|
FreeToken's primitives. The exotic ops are reimplemented as Triton kernels (see
|
|
``freetoken/kernel/triton/dsv4_*.py``); the routed FP4 experts are served from
|
|
FreeToken's :class:`~freetoken.moe.offload_cache.OffloadMoeCache` so only a subset
|
|
of experts is resident on the GPU (the framework's core acceleration).
|
|
|
|
DeepSeek-V4-Flash is a first-class registered model on the shared paged-KV engine:
|
|
its window / compressed-attention / compressed-index KV live in DSV4-owned pools
|
|
addressed by page tables, and sparse attention is a physical-slot gather (see
|
|
:mod:`freetoken.attention.dsv4_sparse` and :mod:`freetoken.kvcache.dsv4.v4_pool`).
|
|
"""
|
|
|
|
from .args import DeepseekV4Args, load_args
|
|
from .config import DSV4VisionConfig, model_config_from_args, parse_config, parse_vision_config
|
|
from .gguf import gguf_mm_config, iter_gguf_weights, iter_mmproj_weights, parse_gguf_config, to_gguf
|
|
from .model import DeepseekV4ForCausalLM
|
|
from .vision import DSV4Aligner, DSV4Vision, DSV4VisionTower
|
|
from .weight import iter_expert_pieces, iter_vision_weights, iter_weights
|
|
|
|
__all__ = [
|
|
"DeepseekV4Args",
|
|
"DSV4Aligner",
|
|
"DSV4Vision",
|
|
"DSV4VisionConfig",
|
|
"DSV4VisionTower",
|
|
"load_args",
|
|
"model_config_from_args",
|
|
"parse_config",
|
|
"parse_gguf_config",
|
|
"parse_vision_config",
|
|
"gguf_mm_config",
|
|
"DeepseekV4ForCausalLM",
|
|
"iter_gguf_weights",
|
|
"iter_mmproj_weights",
|
|
"iter_weights",
|
|
"iter_expert_pieces",
|
|
"iter_vision_weights",
|
|
"to_gguf",
|
|
]
|