hzy
0 Followers · 0 Following
  • Joined on 2026-07-22
Repositories Projects Packages Public Activity Starred Repositories
hzy pushed to hzy at qqtang/sskj 2026-09-10 08:12:30 +00:00
35512db505 [Docs] add standalone B300 DeepSeek-V4-Flash report
hzy pushed to hzy at qqtang/sskj 2026-09-10 07:59:19 +00:00
81d17407bc [Artifacts] finalize B300 matrix at reclaim cutoff
hzy pushed to hzy at qqtang/sskj 2026-09-10 07:37:57 +00:00
9652bfdb9d [Artifacts] update B300 matrix through 15:23
hzy pushed to hzy at qqtang/sskj 2026-09-10 03:02:04 +00:00
7984c25586 [Artifacts] archive B300 DSV4 and GLM-5.3 matrix snapshot
hzy pushed to hzy-kimi3-pd-pp8-standard at qqtang/sskj 2026-09-02 07:18:15 +00:00
75ab8865de test: benchmark DP2 TP32 with DFlash
hzy pushed to hzy-kimi3-pd-pp8-standard at qqtang/sskj 2026-09-02 05:41:40 +00:00
411802e4b1 benchmark Kimi-K3 DP2 TP32 and PP8 without speculation
hzy pushed to hzy-kimi3-pd-pp8-standard at qqtang/sskj 2026-09-02 03:19:24 +00:00
c9305fb830 profile Kimi-K3 PD DFlash 16K512 C8
hzy pushed to hzy-kimi3-pd-pp8-standard at qqtang/sskj 2026-08-31 10:29:12 +00:00
3fbfb84e12 fix: validate Kimi PP8 DFlash PD on GSM8K C1 and C8
hzy pushed to hzy-kimi3-pd-pp8-standard at qqtang/sskj 2026-08-31 09:53:35 +00:00
0e33209ae5 fix: derive hybrid PD transfer bounds from full-attention layers
hzy created branch hzy-kimi3-pd-pp8-standard in qqtang/sskj 2026-08-31 09:10:06 +00:00
hzy pushed to hzy-kimi3-pd-pp8-standard at qqtang/sskj 2026-08-31 09:10:06 +00:00
91ab3fe32d feat: add Kimi-K3 PP8 DFlash PD integration and warmup regression
c06ef5fd61 [Feature] Add Kimi-K3 standard PD deployment
Compare 2 commits »
hzy pushed to hzy at qqtang/sskj 2026-08-27 06:28:31 +00:00
8189942353 [Feature] Add Kimi-K3 standard PD deployment
hzy pushed to hzy-kimi-k3-sm120-flashinfer-mxfp4 at qqtang/sskj 2026-08-21 10:05:45 +00:00
a58b931cc8 [Profiling] Explain Kimi-K3 Deep PP scaling
hzy pushed to hzy-kimi-k3-sm120-flashinfer-mxfp4 at qqtang/sskj 2026-08-21 09:08:29 +00:00
0433fcc3ee [Benchmark] Add Kimi-K3 PP16 prefill results
hzy pushed to hzy-kimi-k3-sm120-flashinfer-mxfp4 at qqtang/sskj 2026-08-21 07:19:50 +00:00
d790df39b2 [Docs] Explain Kimi-K3 EP32 and EP4 execution
hzy pushed to hzy-kimi-k3-sm120-flashinfer-mxfp4 at qqtang/sskj 2026-08-21 06:53:46 +00:00
96ffea6d37 [Docs] Explain Kimi-K3 Deep PP Prefill optimization
hzy pushed to hzy-kimi-k3-sm120-flashinfer-mxfp4 at qqtang/sskj 2026-08-21 06:21:26 +00:00
d7381abe84 [Test] Add Kimi-K3 Prefill PP baseline search
hzy pushed to hzy-kimi-k3-sm120-flashinfer-mxfp4 at qqtang/sskj 2026-08-20 08:05:41 +00:00
60f77cd4ef [Profile] Attribute Kimi-K3 Prefill communication
hzy pushed to hzy-kimi-k3-sm120-flashinfer-mxfp4 at qqtang/sskj 2026-08-20 05:55:13 +00:00
6fac5ad567 [Docs] Attribute Kimi-K3 Prefill collectives
hzy pushed to hzy-kimi-k3-sm120-flashinfer-mxfp4 at qqtang/sskj 2026-08-20 02:37:19 +00:00
2a3b12fa78 [Docs] Reject Kimi-K3 Prefill TP Reduce Scatter path