[https://nvbugs/5800725][fix] Restore Mistral Large 3 text-only processor#14248
Conversation
Signed-off-by: bhsueh <11360707+byshiue@users.noreply.github.com>
…only path MistralCommonImageProcessor.__call__ requires an images positional argument and always applies apply_chat_template, so text-only inputs (e.g. MMLU/GSM8K prompts) raised TypeError before any inference could run. Construct AutoProcessor unconditionally and use it as text_processor for mistral_large_3, matching the intent documented in the adjacent comment. Signed-off-by: tensorrt-cicd <90828364+tensorrt-cicd@users.noreply.github.com> Signed-off-by: bhsueh <11360707+byshiue@users.noreply.github.com>
|
/bot run |
|
No actionable comments were generated in the recent review. 🎉 ℹ️ Recent review info⚙️ Run configurationConfiguration used: Path: .coderabbit.yaml Review profile: CHILL Plan: Enterprise Run ID: 📒 Files selected for processing (2)
💤 Files with no reviewable changes (1)
📝 WalkthroughWalkthroughThis PR refactors ChangesMistral3 Input Processor Initialization
Estimated code review effort🎯 2 (Simple) | ⏱️ ~8 minutes Suggested reviewers
🚥 Pre-merge checks | ✅ 4 | ❌ 1❌ Failed checks (1 warning)
✅ Passed checks (4 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches📝 Generate docstrings
🧪 Generate unit tests (beta)
Comment |
|
PR_Github #48874 [ run ] triggered by Bot. Commit: |
|
PR_Github #48874 [ run ] completed with state
|
|
/bot run |
|
PR_Github #49020 [ run ] triggered by Bot. Commit: |
|
PR_Github #49020 [ run ] completed with state
|
|
/bot run |
|
PR_Github #49369 [ run ] triggered by Bot. Commit: |
|
Hey @byshiue, I think we might have some duplication here with: #13916 (comment). I think your changes might be a better fix in this case. Note that they also fix https://nvbugs/6163033 and https://nvbugs/6157892, which means you should also be able to remove the corresponding lines from the waive list. Also, to verify the fix, we need to run with |
|
PR_Github #49369 [ run ] completed with state |
|
/bot run --extra-stage "GB200-4_GPUs-PyTorch-Post-Merge-1" |
|
PR_Github #49803 [ run ] triggered by Bot. Commit: |
|
PR_Github #49803 [ run ] completed with state
|
Signed-off-by: bhsueh <11360707+byshiue@users.noreply.github.com>
|
/bot run --extra-stage "GB200-4_GPUs-PyTorch-Post-Merge-1" |
|
PR_Github #49871 [ run ] triggered by Bot. Commit: |
|
PR_Github #49871 [ run ] completed with state
|
Signed-off-by: bhsueh <11360707+byshiue@users.noreply.github.com>
|
/bot run |
|
PR_Github #49895 [ run ] triggered by Bot. Commit: |
|
PR_Github #49895 [ run ] completed with state
|
|
Hmmm the two failures appear unrelated + the multi-gpu stage did not run. Do you know what happened? |
They are terminated unexpectedly. Let me rerun. |
|
/bot run |
|
PR_Github #50116 [ run ] triggered by Bot. Commit: |
|
PR_Github #50116 [ run ] completed with state
|
|
/bot run |
|
PR_Github #50160 [ run ] triggered by Bot. Commit: |
|
PR_Github #50160 [ run ] completed with state
|
|
/bot run |
|
PR_Github #50245 [ run ] triggered by Bot. Commit: |
|
PR_Github #50245 [ run ] completed with state
|
|
/bot run |
|
PR_Github #50297 [ run ] triggered by Bot. Commit: |
|
PR_Github #50297 [ run ] completed with state
|
|
/bot run --disable-fail-fast |
|
PR_Github #50419 [ run ] triggered by Bot. Commit: |
|
PR_Github #50419 [ run ] completed with state |
Signed-off-by: bhsueh <11360707+byshiue@users.noreply.github.com> # Conflicts: # tests/integration/test_lists/waives.txt
|
/bot run |
|
PR_Github #50498 [ run ] triggered by Bot. Commit: |
|
PR_Github #50498 [ run ] completed with state |
Bug
https://nvbugspro.nvidia.com/bug/5800725
Root Cause
For
mistral_large_3,self.text_processorwas assigned toMistralCommonImageProcessor. That processor requires positionalimages, so text-only requests failed before generation.Fix
Restore
AutoProcessoras the text-only processor and keepMistralCommonImageProcessoronly for text+image processing.Unwaived Test
Removed the waiver for:
accuracy/test_llm_api_pytorch.py::TestMistralLarge3_675B::test_nvfp4_4gpus[latency_moe_trtllm]Validation
git diff --check github/main...HEAD: passGB200-4_GPUs-PyTorch-Post-Merge-1Summary by CodeRabbit
Bug Fixes
Tests