[BugFix] Fix FP4 dequant symbolic exponent clamp#2656
Conversation
|
👋 Hi! Thank you for contributing to the TileLang project. Please remember to run We appreciate you taking this step! Our team will review your contribution, and we look forward to your awesome work! 🚀 |
|
No actionable comments were generated in the recent review. 🎉 ℹ️ Recent review info⚙️ Run configurationConfiguration used: Path: .coderabbit.yaml Review profile: CHILL Plan: Pro Run ID: 📒 Files selected for processing (2)
📝 WalkthroughWalkthroughThe FP4-to-bfloat16 conversion now clamps symbolic exponents with ChangesSymbolic exponent clamp
Estimated code review effort: 2 (Simple) | ~5 minutes 🚥 Pre-merge checks | ✅ 5✅ Passed checks (5 passed)
✨ Finishing Touches🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
Summary
Fix the trace-time crash in
_tir_u8_to_f4_to_bf16caused by applying Python's built-inminto symbolicPrimExproperands.Changes
mincall with TileLang's lazyT.min.scale=0.uint16scale loaded from a buffer.Min<uint16>node.Validation
PrimExpr.__bool__failure on:origin/main.TRACE_OKfor both scale paths.ninja -j64.C++ style / lint notes
git diff --checkpassed.Fixes #2615
Summary
minwith TileLang’s lazyT.minin_tir_u8_to_f4_to_bf16, preventing symbolicPrimExpr.__bool__trace-time crashes.uint16scales.Min<uint16>node.Validation
git diff --checkpassed.