test: add unit tests for common/worker modules#342
Draft
rebel-jinhwan wants to merge 12 commits intodevfrom
Draft
test: add unit tests for common/worker modules#342rebel-jinhwan wants to merge 12 commits intodevfrom
rebel-jinhwan wants to merge 12 commits intodevfrom
Conversation
example offline inference code won't run with memory utilization 1 as we now have tighter memory estimation logics for kv cache allocation
- Fix conftest SchedulerConfig for v0.13 (default_factory) - Fix test_register_ops to use set_current_vllm_config context manager - Fix test_get_attn_backend_cls for v0.13 AttentionSelectorConfig API - Skip test_lora_functions_sync (integration test, RBLN_DEVICES conflict) - Convert test_resolver async test to sync with asyncio.run() - Extend rbln_envs tests: get_dp_impl, __getattr__, env overrides - Extend logger tests: once methods, log-once behavior, trace function - Extend platform tests: inference_mode, pin_memory, hybrid_kv, inplace, plugin_name - Add bucketing tests (18): linear/exponential managers, factory, edge cases - Add v1/worker/utils tests (3): OMP threads, CPU affinity - Add worker/metrics tests (19): StepMetrics, PrefillMetrics, PerformanceTracker - Add worker/utils tests (6): estimate_available_memory for ATOM/REBEL/OOM Result: 68 passed, 1 skipped, 1 pre-existing failure (test_forward_context)
rebel-jinhwan
changed the title
Verify that enable_trace_function_call actually writes trace logs by triggering a traced function call and checking file content, instead of only asserting sys.gettrace() is not None. Also remove unused 'patch' import.
vllm.lora.models was removed in vLLM v0.13. Update imports to their new locations: - LoRAMapping -> vllm.lora.layers - LoRAModel -> vllm.lora.lora_model - LoRAModelManager, LRUCacheLoRAModelManager -> vllm.lora.model_manager - LoRALayerWeights, PackedLoRALayerWeights -> vllm.lora.lora_weights
…cpu_test marker - Restructure tests/torch_compile/ into units/ (CPU-runnable) and integrations/ (RBLN NPU required) directories - Register cpu_test pytest marker in pyproject.toml (matching upstream vLLM) - Add pytestmark = pytest.mark.cpu_test to all unit test modules verified to run without RBLN NPU hardware - Enables running CPU-only tests via: pytest -m cpu_test
Extract test_platform_plugins (requires RBLN NPU hardware) from test_platform.py into its own file, so test_platform.py can use module-level pytestmark = pytest.mark.cpu_test consistently following the upstream vLLM convention of not mixing CPU and device tests in the same file.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Result: 68 passed, 1 skipped, 1 pre-existing failure (test_forward_context)
🚀 Summary of Changes
📌 Related Issues / Tickets
✅ Type of Change
feature)model)core)bug-fix)perf)refactor)docs)other): please describe🧪 How to Test
.........📸 Screenshots / Logs (if applicable)
📋 Checklist
💬 Notes