 Thomas Wang
|
6c73368c32
|
[AMD][DSV4] Enable hicache on deepseek-v4 fp8 unified attn (#37778)
|
2026-09-17 11:29:23 -07:00 |
|
|
Thomas Wang
|
f60652a43e
|
[AMD] Update deepseek-v4 PDI and cache policy setting for agentic workload (#39702)
|
2026-09-15 21:16:30 -07:00 |
|
|
Thomas Wang
|
ec5fba5777
|
[AMD] Add dspark config and agentic workload section for deepseek-v4 model (#39252)
|
2026-09-12 19:32:01 -07:00 |
|
|
Thomas Wang
|
c415f977b8
|
[AMD] Update v4 args for agentic workload (#38677)
|
2026-09-09 21:41:05 -07:00 |
|
|
Thomas Wang
|
503511c963
|
[AMD] Cherry-pick aiter commit for dsv4 a8w8 bpreshuffle gemm config (#38256)
|
2026-09-07 11:29:31 +08:00 |
|
|
Thomas Wang
|
225129fe44
|
[AMD] Update v4 amd cookbook 0903 (#37829)
|
2026-09-03 22:11:13 -07:00 |
|
|
Thomas Wang
|
bb3e3cbceb
|
[AMD] Fix v4 topk issue (#37439)
|
2026-09-01 08:37:30 -07:00 |
|
|
Thomas Wang
|
7399c2b558
|
[AMD] Update v4 amd cookbook 0830 (#37092)
|
2026-08-29 23:55:28 -07:00 |
|
|
Thomas Wang
|
67bd163a48
|
[AMD] Cherry-pick dsv4 fp4 kv-cache fix aiter commit (#37083)
|
2026-08-29 21:17:36 -07:00 |
|
|
Thomas Wang
|
89816a21a1
|
[AMD] Update v4 amd cookbook 0828 (#36828)
|
2026-08-28 17:20:22 -07:00 |
|
|
Thomas Wang
|
baf09f3954
|
[AMD] Enable deepseek-v4 topk_transform v2 kernel (#36684)
|
2026-08-27 22:36:11 -07:00 |
|
|
Thomas Wang
|
95f5ecd3d2
|
[AMD] Update amd deepseek v4 cookbook 0822 (#35854)
|
2026-08-23 13:26:39 -07:00 |
|
|
Thomas Wang
|
34180a0d35
|
[AMD] Improve K3 dspark draft attn kernel perf (#35499)
|
2026-08-20 21:48:59 -07:00 |
|
|
Thomas Wang
|
27596abdc0
|
[AMD] Update amd k3 cookbook for PR#34580 (#35263)
|
2026-08-17 23:19:36 -07:00 |
|
|
Thomas Wang
|
4d0c5a89af
|
[AMD] Add concat_and_cast_mha_k_pad_kernel to support 12-head and enable K3 aiter prefill kernel (#34837)
|
2026-08-15 14:39:30 -07:00 |
|
|
|
34206c0017
|
[AMD] Run V4 MTP target-verify through the decode kernel (#34597)
Co-authored-by: RolaoDenthu <xinyis10@illinois.edu>
Co-authored-by: 1am9trash <1am9trash>
Co-authored-by: kk <43161300+kkHuang-amd@users.noreply.github.com>
|
2026-08-13 02:00:45 -07:00 |
|
|
Thomas Wang
|
0977b22431
|
[AMD] Restore K3 MLA verify kernel path blocked by can_handle() guard (#34261)
|
2026-08-10 03:31:53 -07:00 |
|
|
Thomas Wang
|
6679d9b60c
|
[AMD] Add K3 verified mla kernel for DSpark on triton backend (#33981)
|
2026-08-07 22:59:26 -07:00 |
|
|
Thomas Wang
|
dea07b348b
|
[AMD] Update amd k3 cookbook for fp8 kv cache (#33825)
|
2026-08-05 23:35:38 -07:00 |
|
|
Thomas Wang
|
7e509f690e
|
[AMD] Fix DeepSeek-V4 FP4 MoE expert memory bloat (#31450)
AITER specific
|
2026-08-01 23:42:21 -07:00 |
|
|
Thomas Wang
|
aafa706f8f
|
[AMD] Update qwen3.5 cookbook (#31258)
|
2026-07-15 10:59:12 +08:00 |
|
|
Thomas WangandHAI
|
8d0fd34150
|
[AMD] Enable unified-KV HiCache on DeepSeek-V4 (#29417)
Co-authored-by: HAI <hixiao@gmail.com>
|
2026-07-09 12:58:47 -07:00 |
|
|
Thomas Wang
|
80decc78ec
|
[AMD][DeepSeek V4] Set SGLANG_OPT_FLASHMLA_SPARSE_PREFILL to false on hip code path (#30237)
|
2026-07-06 01:25:55 -07:00 |
|
|
Thomas Wang
|
f920a37da4
|
[AMD] Copy decode result on forward_stream instead of copy_stream (#29642)
|
2026-06-29 23:53:48 -07:00 |
|
|
Thomas Wang
|
7e6587c94a
|
[AMD] Update v4 cookbook to clean env vars (#28981)
|
2026-06-22 22:16:29 -07:00 |
|
|
Thomas Wang
|
cee1caaf47
|
[AMD] Fix nightly-8-gpu-mi35x-deepseek-v4-flash-rocm720 OOM issue (#28941)
|
2026-06-22 07:35:45 -07:00 |
|
|
Thomas Wang
|
04d952ea10
|
[AMD] deepseek-v4 clean env vars (#28920)
|
2026-06-22 07:32:43 -07:00 |
|
|
Thomas WangandBingxu Chen
|
f5b041622b
|
[AMD] Fix deepseek-v4 mtp accept length issue (#28520)
Co-authored-by: Bingxu Chen <bingxche@amd.com>
|
2026-06-17 11:13:55 -07:00 |
|
|
Thomas Wang
|
21a95333d4
|
[AMD] Add transpose_scale arg for o_proj to fix GLM accuracy issue (#27798)
|
2026-06-17 01:01:02 -07:00 |
|
|
Thomas Wang
|
0d651e653b
|
[AMD] Update v4 amd cookbook (#28423)
|
2026-06-16 18:15:16 -07:00 |
|
|
Thomas Wang
|
800aaefc9e
|
[AMD] Annotate ATOM source for imported v4 unified attention kernels (#28392)
|
2026-06-15 22:12:12 -07:00 |
|
|
Thomas Wang
|
da12f36629
|
[AMD] Refactor unified_kv attention metadata to data class and fuse c4/128 out_loc (#28275)
|
2026-06-15 02:59:01 -07:00 |
|
|
Thomas Wang
|
ddbe5ff2f1
|
[AMD] Pin maturin<1.14 to fix ROCm image build failure (#27999)
|
2026-06-11 22:13:27 -07:00 |
|
|
Thomas Wangandbingxche
|
b0d888a195
|
[CI] Remove AMD DSv4 Docker publish job (#27795)
Co-authored-by: bingxche <bingxche@amd.com>
|
2026-06-10 17:50:28 +08:00 |
|
|
Thomas WangandXinyi Song
|
f2bcdb0508
|
[AMD] Add unified kv attention support in dpsk-v4 (#27380)
Co-authored-by: Xinyi Song <86638975+RolaoDenthu@users.noreply.github.com>
|
2026-06-09 23:13:37 -07:00 |
|
|
Thomas Wang
|
9ab7a64ee1
|
[AMD] Update amd qwen3.5 cookbook (#27660)
|
2026-06-09 16:26:33 -07:00 |
|
|
Thomas Wang
|
d15a2dc72c
|
[AMD] dpsk-v4 swa loc cache support (#26931)
|
2026-06-01 22:37:07 -07:00 |
|
|
Thomas Wang
|
6e9bd82714
|
[AMD][CI] Update v4 CI setting and move the task to main branch (#26662)
|
2026-05-29 00:12:40 -07:00 |
|
|
Thomas Wang
|
a9359707c1
|
[AMD] Enable preshuffle paged MQA and page_size=64 for NSA indexer (#23562)
|
2026-05-13 02:33:57 -07:00 |
|
|
Thomas Wang
|
19afe73e03
|
[AMD] Cherry-pick aiter commit for mhc_pre fix (#24665)
|
2026-05-07 23:49:39 -07:00 |
|
|
Thomas Wang
|
57ecce9807
|
[AMD] Enable MTP for GLM-5-mxfp4 model (#23219)
|
2026-04-20 16:09:07 -07:00 |
|
|
Thomas Wang
|
03828f4205
|
[AMD] Reduce NSA indexer kernels (weights_proj, k-cache store kernel fusion) (#22850)
|
2026-04-19 00:18:11 -07:00 |
|
|
Thomas Wang
|
4a746ea462
|
[AMD] Remove aiter hotfixes in Dockerfile covered by aiter v0.1.12.post1 (#22657)
|
2026-04-13 00:01:37 -07:00 |
|
|
Thomas Wang
|
729b74d8dd
|
[AMD] Fix GLM-5 fp8 KV quant path dispatch on MI300 (#22314)
|
2026-04-07 21:16:02 -07:00 |
|
|
Thomas Wang
|
7431db7392
|
[AMD] Enable FP8 KV cache and FP8 attention kernel for NSA on MI300/MI355 with TileLang backend (#21511)
|
2026-04-03 00:58:23 -07:00 |
|
|
Thomas Wang
|
5628e908ae
|
[AMD] Use tgemm.mm for MoEGate router gemm in deepseek_v2.py (#21657)
|
2026-03-31 00:55:40 -07:00 |
|
|
Thomas Wang
|
855d15adf6
|
[AMD] Tilelang sparse fwd for dsv32 mi355/mi300 (#19945)
|
2026-03-24 02:01:39 -07:00 |
|
|
Thomas Wang
|
c0a4408f78
|
[AMD] Fix dpsk-v32 accuracy issue on mi355 (#20840)
|
2026-03-18 02:06:15 -07:00 |
|
|
Thomas Wang
|
5172c37845
|
[AMD] Use fused GEMM with FP8 cast for FP8 prefill (#19422)
|
2026-02-26 15:14:52 -08:00 |
|
|
|
e20e6c28b9
|
[AMD] Fix accuracy issue when running TP4 dsv3 model with mtp (#18607)
Co-authored-by: YC Tseng <yctseng@amd.com>
Co-authored-by: kkHuang-amd <wunhuang@amd.com>
|
2026-02-12 01:13:16 -08:00 |
|
|
Thomas WangandkkHuang-amd
|
a8eef53dc4
|
Fp8 prefill attn kernel integration (#18528)
Co-authored-by: kkHuang-amd <wunhuang@amd.com>
|
2026-02-10 23:23:48 -08:00 |
|
|
Thomas Wangandkk
|
4262f5259b
|
Tilelang sparse decode fwd for dsv32 mi355 (#18488)
Co-authored-by: kk <43161300+kkHuang-amd@users.noreply.github.com>
|
2026-02-10 09:53:26 -08:00 |
|
|
Thomas Wang
|
6988a0f570
|
Disable mla persistent kernel when not using fp8 kv_cache (#17327)
|
2026-01-19 20:29:54 -08:00 |
|
|
Thomas Wang
|
820e97d6c9
|
Upgrade aiter version (#16619)
|
2026-01-06 22:20:42 -08:00 |
|
|
Thomas Wang
|
5a8adca900
|
Turn off PREBUILD aiter in MI355 (#13963)
|
2025-11-25 22:30:39 -08:00 |
|