tlparse
Parse TORCH_LOG logs produced by PyTorch torch.compile
Decision gist · record as of 2026-08-14
Yes, if you actively use torch.compile and need to analyze its execution traces. The tool is actively maintained, has no known vulnerabilities, and runs on common platforms. Medium install friction is acceptable for a specialized profiling tool. Not necessary if you do not use torch.compile or do not require detailed trace analysis.AI-flagged interpretation of the facts on this page — verify before relying
Before you install
- Requires PyTorch with torch.compile support and Python >= 3.6; TORCH_TRACE environment variable must point to a valid directory.
- Medium install friction due to compiled wheels for multiple platforms (x86_64, ARM, PowerPC, s390x on Linux; x86_64 and ARM on macOS; Windows).
- Active maintenance with a recent commit on 2026-05-11 and no known vulnerabilities.
License · maintenance · safety
BSD-3-Clause (permissive) — BSD-3-Clause is permissive; you can use, modify, and distribute tlparse with minimal restrictions, provided you retain the license notice.
last release 2025-09-12 (336 days) · last repo commit 2026-05-11 · 92 stars
0 known vulnerabilities (OSV.dev, 2026-08-14) · 189,996 downloads/mo, #9,917 on PyPI
Alternatives
Verify before relying
# Set environment variable and run PyTorch code
export TORCH_TRACE=/tmp/my_traced_log_dir
python example.py
# Parse the trace logs
tlparse /tmp/my_traced_log_dir -o tl_out/- Whether custom parser extension (Rust trait implementation) requires recompilation or if dynamic loading is supported.
- Performance characteristics and typical analysis time for large trace logs.
- HTML output format and browser compatibility requirements.
What it is and what it does
tlparse is a command-line tool that consumes structured trace logs produced by PyTorch's torch.compile feature (via the TORCH_TRACE environment variable) and generates HTML files for visual analysis. It is written in Rust and distributed as a Python package with precompiled wheels for multiple architectures.
The tool's primary use is post-hoc inspection of torch.compile execution traces. It accepts a directory of structured logs and outputs formatted HTML reports. The package also supports extension through custom parsers implemented as Rust traits, allowing users to add domain-specific analysis on top of the base trace data.
Use it for
- Analyze torch.compile performance bottlenecks by visualizing structured trace logs as interactive HTML reports.
- Debug compilation failures or unexpected behavior by inspecting detailed trace metadata and payloads.
- Build custom analysis tools by implementing StructuredLogParser traits to extract domain-specific metrics from traces.
- Monitor multi-rank distributed training by processing trace logs from different ranks and generating comparative reports.
Worth the install?
AI-flagged interpretation of the facts on this page. Verify before relying on it.
Yes, if you actively use torch.compile and need to analyze its execution traces.
The tool is actively maintained, has no known vulnerabilities, and runs on common platforms. Medium install friction is acceptable for a specialized profiling tool. Not necessary if you do not use torch.compile or do not require detailed trace analysis.
Install
tlparse on PyPI
Before you install
Medium install friction due to compiled wheels for multiple platforms (x86_64, ARM, PowerPC, s390x on Linux; x86_64 and ARM on macOS; Windows). Active maintenance with a recent commit on 2026-05-11 and no known vulnerabilities.
Requires PyTorch with torch.compile support and Python >= 3.6; TORCH_TRACE environment variable must point to a valid directory.
License in practice
BSD-3-Clause is permissive; you can use, modify, and distribute tlparse with minimal restrictions, provided you retain the license notice.
Quickstart
# Set environment variable and run PyTorch code
export TORCH_TRACE=/tmp/my_traced_log_dir
python example.py
# Parse the trace logs
tlparse /tmp/my_traced_log_dir -o tl_out/Verify before relying
- Whether custom parser extension (Rust trait implementation) requires recompilation or if dynamic loading is supported.
- Performance characteristics and typical analysis time for large trace logs.
- HTML output format and browser compatibility requirements.
Package facts
| License | BSD-3-Clause permissive |
| Python support | Supports the current Python release >=3.6 |
| Install friction | Medium. Platform-specific wheel |
| Runtime dependencies | None |
| Maintenance | Actively maintained 336 days since the last release |
| Last repo commit | |
| First released | |
| Downloads | 189,996 / month, #9,917 on PyPI 30-day window, as of 2026-08-14 |
| Known vulnerabilities | None known OSV.dev, checked 2026-08-14 |
| Classifiers | Programming Language :: Rust |
Evidence: tlparse-0.4.3-py3-none-macosx_10_12_x86_64.whl; tlparse-0.4.3-py3-none-macosx_11_0_arm64.whl; tlparse-0.4.3-py3-none-manylinux_2_17_aarch64.manylinux2014_aarch64.whl; tlparse-0.4.3-py3-none-manylinux_2_17_armv7l.manylinux2014_armv7l.whl; tlparse-0.4.3-py3-none-manylinux_2_17_i686.manylinux2014_i686.whl; tlparse-0.4.3-py3-none-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl; tlparse-0.4.3-py3-none-manylinux_2_17_s390x.manylinux2014_s390x.whl; tlparse-0.4.3-py3-none-manylinux_2_17_x86_64.manylinux2014_x86_64.whl; tlparse-0.4.3-py3-none-win32.whl; tlparse-0.4.3-py3-none-win_amd64.whl
Tags
Let your AI agent find packages like this
Example. Real query, live index.
You found this page by searching. An agent finds it by wishing: SkillFed indexes 14,416 PyPI packages by what they can do, searchable in plain language.
wish › “pytorch torch.compile log parser”
- tlparsetlparse parses structured PyTorch torch.compile trace logs and…
- depyfDecompile and inspect the internal bytecode and optimization graphs…
- tbparseParses TensorBoard event log files and loads them as pandas…
Give your agent the search over MCP, or paste the wish link into any chat.
More Artificial Intelligence packages
LiteLLM provides a unified Python interface to call 100+ LLM providers (OpenAI, Anthropic, Gemini, Bedrock, Azure, and others) using OpenAI-compatible API format, available as both a Python SDK and a self-hosted AI Gateway proxy server.
Install it if you need to work with multiple LLM providers or want to centralize LLM routing in your organization.
Client library and CLI tool for downloading, uploading, and managing models, datasets, and repositories on the Hugging Face Hub platform.
Install it if you work with Hugging Face Hub models or datasets.
LangChain provides a framework for building agents and LLM-powered applications by composing language models, tools, and memory through a unified API that abstracts over multiple model providers.
hf-xet provides chunk-based deduplication and efficient file transfer for the Hugging Face Hub, enabling faster uploads and downloads of large files with local disk caching.
Tokenizers converts raw text into token sequences for NLP models, with support for training custom vocabularies and using pre-built tokenizers (BPE, WordPiece) optimized for speed via Rust.
Transformers provides a unified framework for loading, fine-tuning, and running state-of-the-art pretrained models across text, vision, audio, video, and multimodal tasks using PyTorch, JAX, or TensorFlow.
Install it if you need to run or train any transformer-based model for NLP, vision, audio, or multimodal tasks.
See also depyf · pydlt · onnx2tf · torchtnt · litert-torch · torch-model-archiver · tbparse · torchviz · torchtext