Integration

/

Hugging Face

Logo

Hugging Face

Connect model, fine-tuning, and experiment metadata to energy usage.

Built by

Hugging Face

Category

Workload Traces

Resource

Website

Overview

Hugging Face helps Matcha attach model and experiment context to GPU energy telemetry. Teams can connect fine-tuning runs, model names, dataset tags, and training metadata so energy can be analyzed by model, run, or experiment instead of only by GPU or cluster.

Sample Configuration Steps

  1. Connect your Hugging Face model or training workflow.

  2. Add model name, run ID, dataset, and experiment tags.

  3. Sync training or inference timestamps with GPU telemetry.

  4. Map Hugging Face metadata to workloads in Matcha.

  5. Review energy by model, run, and experiment.

Explore more apps in
Workload Traces
Logo
PyTorch

Attach training runs, steps, duration, model metadata, and experiment context.

Logo
PyTorch

Attach training runs, steps, duration, model metadata, and experiment context.

Logo
vLLM

Track inference requests, batches, latency, tokens, and serving behavior.

Logo
vLLM

Track inference requests, batches, latency, tokens, and serving behavior.

Logo
OpenTelemetry

Use traces, spans, and service events to connect workloads with infrastructure signals.

Logo
OpenTelemetry

Use traces, spans, and service events to connect workloads with infrastructure signals.