traceopt-ai/traceml
Lightweight training runtime health monitor
This helps machine learning engineers or researchers quickly diagnose why their PyTorch model training is running slowly or behaving erratically. It takes your existing PyTorch training script and, with minimal setup, shows you a live view of system performance, training step timing, and resource use. This allows you to pinpoint bottlenecks like slow data loading, imbalanced distributed training, or memory issues while your model is still training.
109 stars.
Use this if your PyTorch training runs feel slower than expected, jittery, or imbalanced across multiple GPUs, and you need a fast answer without resorting to heavyweight profiling.
Not ideal if you need deep, kernel-level tracing or a comprehensive, full-featured observability platform; it's designed for lightweight, on-the-fly bottleneck diagnosis.
Stars
109
Forks
8
Language
Python
License
Apache-2.0
Category
Last pushed
Mar 13, 2026
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/mlops/traceopt-ai/traceml"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
claimed-framework/claimed
The goal of CLAIMED is to enable low-code/no-code rapid prototyping style programming to...
logicalclocks/hopsworks
Hopsworks - Data-Intensive AI platform with a Feature Store
project-codeflare/codeflare
Simplifying the definition and execution, scaling and deployment of pipelines on the cloud.
myelintek/primehub
open-source MLOps platform
ml-tooling/ml-hub
🧰 Multi-user development platform for machine learning teams. Simple to setup within minutes.