Thanks to visit codestin.com
Credit goes to github.com

Skip to content

Tags: kvcache-ai/ktransformers

Tags

v0.7.0.post4

Toggle v0.7.0.post4's commit message

Verified

This commit was created on GitHub.com and signed with GitHub’s verified signature.
release: align post4 stack and allow CI artifact-only builds (#2197)

v0.7.0.post1

Toggle v0.7.0.post1's commit message

Verified

This commit was created on GitHub.com and signed with GitHub’s verified signature.
feat(GLM): add native GLM-5.3-flash support (#2173)

* feat: support SwiGLU limit for block FP8 experts

* perf(fp8): add native layerwise batch transport

* feat(fp8): support layerwise transport for TP1/2/4/8

* release: prepare GLM-5.3-flash support

* docs(glm): refine GLM-5.3-flash documentation

* docs(glm): simplify release announcement

v0.7.0

Toggle v0.7.0's commit message

Verified

This commit was created on GitHub.com and signed with GitHub’s verified signature.
[feat](kt-sft): Activation reuse & Int8 kernel refactor & native bloc…

…k-FP8 LoRA SFT (#2141)

* feat(sft): support distributed activation reuse policies

* feat(sft): add frozen-base INT8 LoRA training

* fix(sft): make INT8 expert LoRA rank-zero authoritative

* fix(sft): preserve DeepSeek router LoRA routing

* feat(sft): enable persistent INT8 LoRA training

* perf(sft): accelerate INT8 VNNI with oneDNN BRGEMM

* perf(int8): fuse oneDNN compensation into backward repack

* [feat]: support BF16 expert LoRA training

* [fix]: honor forwarded activation policy in SFT workers

* feat(sft): add native block-FP8 routed expert LoRA

* feat(sft): expose explicit expert placeholder ownership

* fix(sft): publish fused adapter artifacts atomically

* feat(sft): own artifact and adapter lifecycle contracts

* fix(sft): harden artifact and rank-local contracts

* fix(sft): auto-adapt owner before adapter restore

* style(sft): keep lifecycle comments concise

* fix(sft): require fused adapter manifests

* test(sft): use spawn for distributed workers

* fix(sft): preserve runtime checkpoint metadata

* fix(sft): validate wrapped runtime configuration

* fix(sft): preserve expert format provenance

* fix(sft): own routed experts during device dispatch

* test(sft): lock explicit quantization conflict

* fix(cpu): make shared memory buffers lifetime-safe

* release: prepare v0.7.0

v0.6.4

Toggle v0.6.4's commit message

Verified

This commit was created on GitHub.com and signed with GitHub’s verified signature.
[feat](kt-kernel): end-to-end full-parameter and LoRA SFT(#2094)

* Add full FT development snapshot

* [fix](kt-kernel): fix Full FT TP base weight gradients

* [fix]: bug fix of 2d81e86

* [fix](kt-kernel): fix AMX BF16 full-weight gradients

* [docs](kt-kernel): document Full FT fork changes and debug history

* [docs](kt-kernel): align fork remote terminology

* [chore](kt-kernel): keep agent notes local

* [fix](kt-kernel): configure SFT OpenMP threads

* [perf](kt-kernel): optimize AMX Full-FT weight gradients

Coarsen base-weight gradient work from individual output tiles to fixed-intermediate strips so each task reuses packed panels across the hidden dimension.

Keep aligned thread-local BF16 panels across tasks and retain FP32 AMX accumulator tiles for the full K reduction. Gate and up run separate K passes while sharing the packed input panel.

On the matched Qwen3-30B-A3B 1-GPU test, stable Full-FT backward drops from 9.281s to 6.792s (-26.82%), step time drops from 19.252s to 16.140s, and TPS rises from 212.76 to 253.78. The LoRA-only backward control changes by -2.74%.

Validated with clang-format, the Release AMX/CUDA extension build, TP1/TP2 reference gradients across boundary token counts, and the 15-step Full-then-LoRA performance run.

* [feat](kt-kernel): add staged SFT profiling

* [fix](kt-kernel): reuse inference BF16 kernel for SFT

* [perf](kt-kernel): add fine-grained Full-FT profiling

* [perf](kt-kernel): batch BF16 Full-FT weight gradients

Use one expert-aggregated tile driver for AVX512-BF16 and AMX base-weight gradients, and pack updated full-precision weights directly into TP BufferB layouts without temporary partitions. Add worker-local profiling and focused dWeight/strided-repack coverage.

* [test](kt-kernel): benchmark BF16 dWeight AMX driver

* [fix](kt-kernel): label dWeight store as worker CPU time

* [docs](sft): record BF16 Full-FT performance

* [docs](sft): remove Qwen3 Full-FT performance report

* [perf](kt-kernel): reduce BF16 Full-FT checkpoint overhead

Retain the first CPU MoE forward state across non-reentrant checkpoint recomputation, write BF16 activations directly into the backward cache, and reduce dWeight packing and gradient-clear traffic. Extend staged profiling and cover checkpoint reuse plus AMX/AVX dWeight paths.

* [perf](kt-kernel): make SFT optimizer gradients authoritative

Bind Full-FT and LoRA Parameter.grad directly to the KT-managed BF16 gradient buffers, avoiding PyTorch duplicate accumulation.

Accumulate microbatch gradients in C++, lazily clear expert buffers between optimizer windows, preserve rank-0 distributed ownership, and add lifecycle and AMX dWeight coverage.

* [perf](kt-kernel): avoid eager Full-FT gradient zeroing

Allocate authoritative Full-FT gradient buffers with torch.empty. The C++ state machine performs the mandatory full clear before first use, avoiding redundant Python-side first touch.

* perf(sft): enable checkpoint forward reuse for LoRA

* fix(sft): serialize checkpoint recompute with async repack

* fix(sft): persist authoritative full weights

* fix(sft): normalize legacy distributed gradients

* [fix](sft): train gated shared experts

* fix(sft): preserve expert placeholders across state dict loads

* ci: publish ktransformers sdist in release workflow

* fix(sft): preserve router autograd for LoRA training

Determine routing graph tracking from the router's trainable parameters instead of the Full-FT mode flag, so PEFT LoRA adapters on MoE gates receive routing-weight gradients. Fail fast when a trainable router returns detached weights and cover frozen, TopK, checkpoint-reuse, and two-step optimizer behavior.

---------

Co-authored-by: illu <[email protected]>

v0.6.3.post1

Toggle v0.6.3.post1's commit message

Verified

This commit was created on GitHub.com and signed with GitHub’s verified signature.
release: bump version to 0.6.3.post1 (#2063)

v0.6.3

Toggle v0.6.3's commit message

Verified

This commit was created on GitHub.com and signed with GitHub’s verified signature.
[build]: sync sglang submodule to 8b636f9008dbad58c0a8e481b03e794739e…

…6c146 (#2047)

Co-authored-by: ovowei <[email protected]>

v0.6.2.post3

Toggle v0.6.2.post3's commit message

Verified

This commit was created on GitHub.com and signed with GitHub’s verified signature.
Release/0.6.2.post3: carry kt-kernel SwiGLU clamp companion missing f…

…rom post2

v0.6.2.post2

Toggle v0.6.2.post2's commit message

Verified

This commit was created on GitHub.com and signed with GitHub’s verified signature.
0.6.2.post2: submodule refactor and update tutorial (#1993)

- sglang submodule -> 43ed1ec77: V4-Flash hybrid SWA chunked-prefill
  hang fix (#44) + DSV4 plugin registry refactor (#47)
- pin sglang-kt==0.6.2.post2
- tutorial: switch V4-Flash launch example from 8x RTX 5090 to
  single-card (decode 20+ tok/s); flip Ada Lovelace SM_89 row to
  validated; update Hardware Requirements GPU line accordingly

v0.6.2

Toggle v0.6.2's commit message

Verified

This commit was created on GitHub.com and signed with GitHub’s verified signature.
release: bump version to 0.6.2 (#1973)

v0.6.2.post1

Toggle v0.6.2.post1's commit message

Verified

This commit was created on GitHub.com and signed with GitHub’s verified signature.
[release]: 0.6.2.post1

V4-Flash MXFP4 full-GPU prefill fallback now works:
- Previously crashed all TP schedulers with StopIteration/AttributeError
  whenever --kt-gpu-prefill-token-threshold was low enough to actually
  fire (path was hardcoded for FP8/INT4 layouts).
- Now detects MXFP4, re-runs the V4 swizzle on the 256-expert gpu_layer,
  caches the load across prefill chunks.
- Measured on 8x RTX 5090 (threshold=1024, chunked=1024):
  16k input -> 2011 tok/s, 65k -> 2798, 262k -> 2154 prefill TPS.