Repository navigation
Conversation
This branch has not been deployed
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Description
This PR adds structured, machine-readable reporting and verification capabilities for Dynamo workflows:
analyze()for preflight converter support, partitioning, coverage, and configuration checks without building engines. (dryrun=Trueis the existing human-oriented debugging mode that skips TensorRT engine conversion and prints or saves a textual partitioning summary, whileanalyze()is a programmatic preflight API that returns a typed, JSON-serializableAnalysisReportwith node eligibility, assignments, fallback reasons, partitions, coverage, checks, diagnostics, and status.analyze()also predicts behavior using the requested compilation settings, isolates graph and weight changes, preserves RNG state, supports structured events, and reports failures without building engines.)compile_with_report()to return the compiled module together with detailed analysis, engine, diagnostic, and timing evidence.validate()to deterministically compare reference and compiled outputs using explicit structure, dtype, shape, tolerance, and nonfinite-value rules while preserving model, input, aliasing, and RNG state.benchmark()to produce reproducible wall-clock and CUDA-event latency, percentile, throughput, and speedup measurements under explicit synchronization and timing semantics.Existing
compile()and dryrun APIs remain available with unchanged return types and behavior.This provides the dependency-free Core API layer for the upcoming MCP PR, which will expose these capabilities as policy-controlled MCP tools and add artifact persistence, job orchestration, and lineage tracking outside the core package.
Fixes #4843
Type of change
Checklist: