[CI] Shard H100 MoE refactor integration tests (#52352)

Signed-off-by: Kevin Luu <51931015+khluu@users.noreply.github.com>
This commit is contained in:
Kevin H. Luu
2026-09-02 01:10:14 -07:00
committed by GitHub
parent df09c76735
commit 1b4b2a18b8
6 changed files with 17 additions and 13 deletions
+5 -1
View File
@@ -177,8 +177,12 @@ steps:
device: h100
optional: true
num_devices: 2
parallelism: 4
commands:
- pytest -s -v evals/gsm8k/test_gsm8k_correctness.py --config-list-file=evals/gsm8k/configs/moe-refactor/config-h100.txt
# Timing-balanced static split of the former config-h100.txt (measured on
# build 83851): shards ~12.4/13.6/12.4/14.5m of eval time. New configs must
# be added to exactly one config-h100-shard-*.txt.
- pytest -s -v evals/gsm8k/test_gsm8k_correctness.py --config-list-file=evals/gsm8k/configs/moe-refactor/config-h100-shard-$$BUILDKITE_PARALLEL_JOB.txt
- label: ":nvidia: (B200) MoE Refactor Integration TEMPORARY Shard %N"
key: moe-refactor-integration-test-b200-temporary
@@ -0,0 +1,2 @@
Qwen3-30B-A3B-Fp8-AutoFp8-deepgemm.yaml
Qwen3-30B-A3B-BF16-fi-cutlass.yaml
@@ -0,0 +1,3 @@
Qwen3-30B-A3B-Fp8-AutoFp8-marlin.yaml
Qwen3-30B-A3B-Fp8-AutoFp8-triton.yaml
Qwen3-30B-A3B-BF16-triton.yaml
@@ -0,0 +1,3 @@
Qwen3-30B-A3B-Fp8-AutoFp8-fi-cutlass.yaml
Qwen3-30B-A3B-Fp8-CT-Block-deepgemm.yaml
Qwen3-30B-A3B-Fp8-CT-Block-triton.yaml
@@ -0,0 +1,4 @@
Mixtral-8x7B-Fp8-AutoFp8-triton.yaml
Qwen3-30B-A3B-Fp8-CT-Block-marlin.yaml
Qwen3-30B-A3B-Fp8-CT-Channel-marlin.yaml
Qwen3-30B-A3B-Fp8-CT-Channel-vllm-cutlass.yaml
@@ -1,12 +0,0 @@
Mixtral-8x7B-Fp8-AutoFp8-triton.yaml
Qwen3-30B-A3B-Fp8-AutoFp8-deepgemm.yaml
Qwen3-30B-A3B-Fp8-AutoFp8-fi-cutlass.yaml
Qwen3-30B-A3B-Fp8-AutoFp8-marlin.yaml
Qwen3-30B-A3B-Fp8-AutoFp8-triton.yaml
Qwen3-30B-A3B-Fp8-CT-Block-deepgemm.yaml
Qwen3-30B-A3B-Fp8-CT-Block-marlin.yaml
Qwen3-30B-A3B-Fp8-CT-Block-triton.yaml
Qwen3-30B-A3B-Fp8-CT-Channel-marlin.yaml
Qwen3-30B-A3B-Fp8-CT-Channel-vllm-cutlass.yaml
Qwen3-30B-A3B-BF16-fi-cutlass.yaml
Qwen3-30B-A3B-BF16-triton.yaml