[fix] Switch placement of image placeholder for mistral 3.1 #6435
Conversation
📝 WalkthroughWalkthroughThe changes update the placeholder placement for the "mistral3" multimodal model in a utility file, correcting comments and logic to use a "before text" position based on accuracy tests. Corresponding expected output keywords in an integration test are also updated to reflect these changes. No public interfaces or control flow are modified. Changes
Estimated code review effort🎯 1 (Trivial) | ⏱️ ~3 minutes Suggested reviewers
Note ⚡️ Unit Test Generation is now available in beta!Learn more here, or try it out under "Finishing Touches" below. 📜 Recent review detailsConfiguration used: .coderabbit.yaml 📒 Files selected for processing (2)
✅ Files skipped from review due to trivial changes (1)
🚧 Files skipped from review as they are similar to previous changes (1)
✨ Finishing Touches
🧪 Generate unit tests
🪧 TipsChatThere are 3 ways to chat with CodeRabbit:
SupportNeed help? Create a ticket on our support page for assistance with any issues or questions. Note: Be mindful of the bot's finite context window. It's strongly recommended to break down tasks such as reading entire modules into smaller chunks. For a focused discussion, use review comments to chat about specific files and their changes, instead of using the PR comments. CodeRabbit Commands (Invoked using PR comments)
Other keywords and placeholders
Documentation and Community
|
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 0
🔭 Outside diff range comments (1)
tests/integration/defs/test_e2e.py (1)
2085-2095: Enforce per-list thresholds to avoid unintended tighteningThe hard-coded
match_ratio = 4.0/5 if model_name == "qwen2-vl-7b-instruct" and modality == "image": match_ratio = 4.0/6now unintentionally requires ≥6/7 matches (≈85.7%) for the 7-keyword lists. Let’s compute the minimum hits directly per list, preserve the qwen2 special-case, and avoid float comparisons:
- match_ratio = 4.0 / 5 - if model_name == "qwen2-vl-7b-instruct" and modality == "image": - match_ratio = 4.0 / 6 + from math import ceil + + # Compute min hits per keyword list: + # - default: ceil(80% of n_keywords) + # - qwen2-vl-7b-instruct + image: require exactly 4 + def min_required_hits(n_keywords: int, model: str, mode: str) -> int: + if model == "qwen2-vl-7b-instruct" and mode == "image": + return 4 + return ceil(0.8 * n_keywords) for prompt_output, prompt_keywords in zip( parse_output(output), expected_keywords[model_name][modality]): - matches = [ - keyword in prompt_output.lower() for keyword in prompt_keywords - ] - obs_match_ratio = 1. * sum(matches) / len(matches) - assert obs_match_ratio >= match_ratio, f"Incorrect output!\nGenerated \"{prompt_output}\"\nExpected keywords \"{prompt_keywords}\"\n Matched keywords: {matches}\n Observed match ratio {obs_match_ratio} below threshold {match_ratio}" + matches = [kw in prompt_output.lower() for kw in prompt_keywords] + hits = sum(matches) + required = min_required_hits(len(prompt_keywords), model_name, modality) + assert hits >= required, ( + f"Incorrect output!\n" + f"Generated: \"{prompt_output}\"\n" + f"Expected keywords: {prompt_keywords}\n" + f"Matched keywords: {matches}\n" + f"Required ≥ {required}/{len(prompt_keywords)}, got {hits}" + )• This makes the rule explicit (default ≈80% coverage, one fixed edge case)
• Please run CI against all model/modality combinations to confirm no regressions.
🧹 Nitpick comments (1)
tensorrt_llm/inputs/utils.py (1)
257-264: Update the explanatory comment to match the new behaviourThe comment still talks about “does not strictly have to be after the text”, but the mapping directly beneath it now sets
mistral3toBEFORE_TEXT.
Consider revising the phrasing to avoid confusion.-# NOTE: for mistral3 multimodal models, it does not strictly have to be after the text. +# NOTE: for mistral3 multimodal models, the placeholder location is not strictly +# fixed, but accuracy tests show better results when the image precedes the text.
📜 Review details
Configuration used: .coderabbit.yaml
Review profile: CHILL
Plan: Pro
📒 Files selected for processing (2)
tensorrt_llm/inputs/utils.py(1 hunks)tests/integration/defs/test_e2e.py(1 hunks)
🧰 Additional context used
📓 Path-based instructions (2)
**/*.py
📄 CodeRabbit Inference Engine (CODING_GUIDELINES.md)
**/*.py: Python code should conform to Python 3.8+.
Indent Python code with 4 spaces. Do not use tabs.
Always maintain the namespace when importing in Python, even if only one class or function from a module is used.
Python filenames should use snake_case (e.g., some_file.py).
Python classes should use PascalCase (e.g., class SomeClass).
Python functions and methods should use snake_case (e.g., def my_awesome_function():).
Python local variables should use snake_case, and prefix k for variable names that start with a number (e.g., k_99th_percentile).
Python global variables should use upper snake_case and prefix G (e.g., G_MY_GLOBAL).
Python constants should use upper snake_case (e.g., MY_CONSTANT).
Avoid shadowing variables declared in an outer scope in Python.
Initialize all externally visible members of a Python class in the constructor.
For interfaces that may be used outside a Python file, prefer docstrings over comments.
Comments in Python should be reserved for code within a function, or interfaces that are local to a file.
Use Google style docstrings for Python classes and functions, which can be parsed by Sphinx.
Attributes and variables in Python can be documented inline; attribute docstrings will be rendered under the docstring for the class.
Avoid using reflection in Python when functionality can be easily achieved without it.
When using try-except blocks in Python, limit the except to the smallest set of errors possible.
When using try-except blocks to handle multiple possible variable types in Python, keep the body of the try as small as possible, using the else block to implement the logic.
Files:
tensorrt_llm/inputs/utils.pytests/integration/defs/test_e2e.py
**/*.{cpp,h,hpp,cc,cxx,cu,py}
📄 CodeRabbit Inference Engine (CODING_GUIDELINES.md)
All TensorRT-LLM Open Source Software code should contain an NVIDIA copyright header that includes the current year. This includes .cpp, .h, .cu, .py, and any other source files which are compiled or interpreted.
Files:
tensorrt_llm/inputs/utils.pytests/integration/defs/test_e2e.py
|
/bot run --extra-stage "H100_PCIe-PyTorch-Post-Merge-1" |
There was a problem hiding this comment.
Actionable comments posted: 3
🧹 Nitpick comments (2)
tensorrt_llm/_torch/models/checkpoints/hf/weight_loader.py (1)
27-27: Replace debug prints with proper logging.Debug print statements should use the existing logger instead of
print()for consistency with the codebase's logging standards.- print(f"==== BEFORE: {len(weight_files)}") + logger.debug(f"Safetensors files before filtering: {len(weight_files)}")- print(f"==== AFTER: {len(weight_files)}") + logger.debug(f"Safetensors files after filtering: {len(weight_files)}")Also applies to: 29-29
docker/Makefile (1)
165-166: Clean up commented code or document its purpose.The commented lines suggest work in progress. Either remove them if they're no longer needed or uncomment and configure them properly if they serve a purpose.
📜 Review details
Configuration used: .coderabbit.yaml
Review profile: CHILL
Plan: Pro
📒 Files selected for processing (3)
docker/Makefile(2 hunks)tensorrt_llm/_torch/models/checkpoints/hf/weight_loader.py(1 hunks)tensorrt_llm/builder.py(1 hunks)
🧰 Additional context used
📓 Path-based instructions (2)
**/*.py
📄 CodeRabbit Inference Engine (CODING_GUIDELINES.md)
**/*.py: Python code should conform to Python 3.8+.
Indent Python code with 4 spaces. Do not use tabs.
Always maintain the namespace when importing in Python, even if only one class or function from a module is used.
Python filenames should use snake_case (e.g., some_file.py).
Python classes should use PascalCase (e.g., class SomeClass).
Python functions and methods should use snake_case (e.g., def my_awesome_function():).
Python local variables should use snake_case, and prefix k for variable names that start with a number (e.g., k_99th_percentile).
Python global variables should use upper snake_case and prefix G (e.g., G_MY_GLOBAL).
Python constants should use upper snake_case (e.g., MY_CONSTANT).
Avoid shadowing variables declared in an outer scope in Python.
Initialize all externally visible members of a Python class in the constructor.
For interfaces that may be used outside a Python file, prefer docstrings over comments.
Comments in Python should be reserved for code within a function, or interfaces that are local to a file.
Use Google style docstrings for Python classes and functions, which can be parsed by Sphinx.
Attributes and variables in Python can be documented inline; attribute docstrings will be rendered under the docstring for the class.
Avoid using reflection in Python when functionality can be easily achieved without it.
When using try-except blocks in Python, limit the except to the smallest set of errors possible.
When using try-except blocks to handle multiple possible variable types in Python, keep the body of the try as small as possible, using the else block to implement the logic.
Files:
tensorrt_llm/builder.pytensorrt_llm/_torch/models/checkpoints/hf/weight_loader.py
**/*.{cpp,h,hpp,cc,cxx,cu,py}
📄 CodeRabbit Inference Engine (CODING_GUIDELINES.md)
All TensorRT-LLM Open Source Software code should contain an NVIDIA copyright header that includes the current year. This includes .cpp, .h, .cu, .py, and any other source files which are compiled or interpreted.
Files:
tensorrt_llm/builder.pytensorrt_llm/_torch/models/checkpoints/hf/weight_loader.py
⏰ Context from checks skipped due to timeout of 90000ms. You can increase the timeout in your CodeRabbit configuration to a maximum of 15 minutes (900000ms). (1)
- GitHub Check: Pre-commit Check
🔇 Additional comments (2)
tensorrt_llm/_torch/models/checkpoints/hf/weight_loader.py (1)
28-28: LGTM: Consolidated files filtering is correctly implemented.The filtering logic to exclude safetensors files containing "consolidated" is appropriate for avoiding duplicate weight loading.
tensorrt_llm/builder.py (1)
688-694: Excellent type safety improvement for kv_cache_type serialization.The enhanced type checking and explicit error handling ensures robust dictionary conversion. The walrus operator usage is appropriate here for combining assignment and condition checking.
|
PR_Github #13314 [ run ] triggered by Bot |
|
PR_Github #13314 [ run ] completed with state |
Signed-off-by: William Zhang <[email protected]>
2ez4bz
force-pushed
the
dev-mistral-fix-acc
branch
from
1320d8b to
0cadc59
Compare
|
/bot run |
|
/bot kill |
|
/bot run --extra-stage "H100_PCIe-PyTorch-Post-Merge-1" |
|
PR_Github #13391 [ run ] triggered by Bot |
|
PR_Github #13392 [ run ] triggered by Bot |
|
PR_Github #13393 [ ] completed with state |
|
PR_Github #13391 [ run ] completed with state |
|
PR_Github #13392 [ run ] completed with state |
|
/bot run --extra-stage "H100_PCIe-PyTorch-Post-Merge-1" |
|
PR_Github #13398 [ run ] triggered by Bot |
|
PR_Github #13398 [ run ] completed with state |
|
/bot run --extra-stage "H100_PCIe-PyTorch-Post-Merge-1" |
|
PR_Github #13408 [ run ] triggered by Bot |
|
PR_Github #13408 [ run ] completed with state |
|
/bot run --reuse-test |
|
PR_Github #13473 [ run ] triggered by Bot |
|
PR_Github #13473 [ run ] completed with state |
|
|
) Signed-off-by: William Zhang <[email protected]> Signed-off-by: Lanyu Liao <[email protected]>
) Signed-off-by: William Zhang <[email protected]>
5 participants
Summary by CodeRabbit
Bug Fixes
Documentation
[TRTLLM-6598][fix] Switch placement of image placeholder for mistral 3.1
Description
Test Coverage
Ran the following manual tests:
trtllm-serveprior to this change: score is around 0.62.vllm servein the official 0.9.2 docker image: score is around 0.81.[IMG]placeholder token after the text prompt: score is around 0.62.GitHub Bot Help
/bot [-h] ['run', 'kill', 'skip', 'reuse-pipeline'] ...Provide a user friendly way for developers to interact with a Jenkins server.
Run
/bot [-h|--help]to print this help message.See details below for each supported subcommand.
run [--reuse-test (optional)pipeline-id --disable-fail-fast --skip-test --stage-list "A10-PyTorch-1, xxx" --gpu-type "A30, H100_PCIe" --test-backend "pytorch, cpp" --add-multi-gpu-test --only-multi-gpu-test --disable-multi-gpu-test --post-merge --extra-stage "H100_PCIe-TensorRT-Post-Merge-1, xxx" --detailed-log --debug(experimental)]Launch build/test pipelines. All previously running jobs will be killed.
--reuse-test (optional)pipeline-id(OPTIONAL) : Allow the new pipeline to reuse build artifacts and skip successful test stages from a specified pipeline or the last pipeline if no pipeline-id is indicated. If the Git commit ID has changed, this option will be always ignored. The DEFAULT behavior of the bot is to reuse build artifacts and successful test results from the last pipeline.--disable-reuse-test(OPTIONAL) : Explicitly prevent the pipeline from reusing build artifacts and skipping successful test stages from a previous pipeline. Ensure that all builds and tests are run regardless of previous successes.--disable-fail-fast(OPTIONAL) : Disable fail fast on build/tests/infra failures.--skip-test(OPTIONAL) : Skip all test stages, but still run build stages, package stages and sanity check stages. Note: Does NOT update GitHub check status.--stage-list "A10-PyTorch-1, xxx"(OPTIONAL) : Only run the specified test stages. Examples: "A10-PyTorch-1, xxx". Note: Does NOT update GitHub check status.--gpu-type "A30, H100_PCIe"(OPTIONAL) : Only run the test stages on the specified GPU types. Examples: "A30, H100_PCIe". Note: Does NOT update GitHub check status.--test-backend "pytorch, cpp"(OPTIONAL) : Skip test stages which don't match the specified backends. Only support [pytorch, cpp, tensorrt, triton]. Examples: "pytorch, cpp" (does not run test stages with tensorrt or triton backend). Note: Does NOT update GitHub pipeline status.--only-multi-gpu-test(OPTIONAL) : Only run the multi-GPU tests. Note: Does NOT update GitHub check status.--disable-multi-gpu-test(OPTIONAL) : Disable the multi-GPU tests. Note: Does NOT update GitHub check status.--add-multi-gpu-test(OPTIONAL) : Force run the multi-GPU tests in addition to running L0 pre-merge pipeline.--post-merge(OPTIONAL) : Run the L0 post-merge pipeline instead of the ordinary L0 pre-merge pipeline.--extra-stage "H100_PCIe-TensorRT-Post-Merge-1, xxx"(OPTIONAL) : Run the ordinary L0 pre-merge pipeline and specified test stages. Examples: --extra-stage "H100_PCIe-TensorRT-Post-Merge-1, xxx".--detailed-log(OPTIONAL) : Enable flushing out all logs to the Jenkins console. This will significantly increase the log volume and may slow down the job.--debug(OPTIONAL) : Experimental feature. Enable access to the CI container for debugging purpose. Note: Specify exactly one stage in thestage-listparameter to access the appropriate container environment. Note: Does NOT update GitHub check status.For guidance on mapping tests to stage names, see
docs/source/reference/ci-overview.mdand the
scripts/test_to_stage_mapping.pyhelper.kill
killKill all running builds associated with pull request.
skip
skip --comment COMMENTSkip testing for latest commit on pull request.
--comment "Reason for skipping build/test"is required. IMPORTANT NOTE: This is dangerous since lack of user care and validation can cause top of tree to break.reuse-pipeline
reuse-pipelineReuse a previous pipeline to validate current commit. This action will also kill all currently running builds associated with the pull request. IMPORTANT NOTE: This is dangerous since lack of user care and validation can cause top of tree to break.