[AMD] Add more tests to per-commit-amd (#6926)

.github/workflows/pr-test-amd.yml

@@ -220,8 +220,10 @@ jobs:
     if: (github.repository == 'sgl-project/sglang' || github.event_name == 'pull_request') &&
       github.event.pull_request.draft == false
     strategy:
+      fail-fast: false
       matrix:
         runner: [linux-mi300-gpu-1, linux-mi325-gpu-1]
+        part: [0, 1]
     runs-on: ${{matrix.runner}}
     steps:
       - name: Checkout code
@@ -238,7 +240,7 @@ jobs:
       - name: Run test
         timeout-minutes: 40
         run: |
-          bash scripts/amd_ci_exec.sh python3 run_suite.py --suite per-commit-amd
+          bash scripts/amd_ci_exec.sh python3 run_suite.py --suite per-commit-amd --auto-partition-id ${{ matrix.part }} --auto-partition-size 2
 
   unit-test-backend-2-gpu-amd:
     if: (github.repository == 'sgl-project/sglang' || github.event_name == 'pull_request') &&

scripts/amd_ci_install_dependency.sh

@@ -10,6 +10,10 @@ docker exec ci_sglang pip install -e "python[dev_hip]"
 docker exec -w / ci_sglang git clone https://github.com/merrymercy/human-eval.git
 docker exec -w /human-eval ci_sglang pip install -e .
 
+# For lmms_evals evaluating MMMU
+docker exec -w / ci_sglang git clone --branch v0.3.3 --depth 1 https://github.com/EvolvingLMMs-Lab/lmms-eval.git
+docker exec -w /lmms-eval ci_sglang pip install -e .
+
 docker exec -w / ci_sglang mkdir -p /dummy-grok
 mkdir -p dummy-grok && wget https://sharkpublic.blob.core.windows.net/sharkpublic/sglang/dummy_grok.json -O dummy-grok/config.json
 docker cp ./dummy-grok ci_sglang:/

test/srt/run_suite.py

@@ -91,10 +91,19 @@ suites = {
         TestFile("test_w8a8_quantization.py", 46),
     ],
     "per-commit-amd": [
+        TestFile("models/lora/test_lora_backend.py", 99),
+        TestFile("models/lora/test_multi_lora_backend.py", 60),
+        TestFile("models/lora/test_lora_cuda_graph.py", 250),
         TestFile("test_mla.py", 242),
         TestFile("test_mla_deepseek_v3.py", 221),
         TestFile("test_torch_compile.py", 76),
         TestFile("test_torch_compile_moe.py", 172),
+        TestFile("models/test_qwen_models.py", 82),
+        TestFile("models/test_reward_models.py", 132),
+        TestFile("test_abort.py", 51),
+        TestFile("test_block_int8.py", 22),
+        TestFile("test_create_kvindices.py", 2),
+        TestFile("test_chunked_prefill.py", 313),
     ],
     "per-commit-2-gpu": [
         TestFile("models/lora/test_lora_tp.py", 116),