# Cerebras/modelzoo

Repository: https://github.com/Cerebras/modelzoo
Canonical: https://ross.abutalabs.com/products/modelzoo
Language: Python
License: Apache-2.0
License Family: permissive
Last push: 2026-08-31T18:49:10+00:00

## Health v2 (maintenance only)
Score: 77/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 100, release rhythm 35, longevity 100
- inputs: {"age_days": 1608, "days_push": 2, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1193, forks 172 (observed 2026-09-01T02:13:54.361200+00:00)

## What it is
Cerebras Model Zoo is a collection of reference deep learning model implementations (Llama, Mixtral, DINOv2, Llava, etc.) with configs and utilities optimized for Cerebras hardware. It includes a CLI for training workflows, data preprocessing tools, and checkpoint converters between Cerebras and HuggingFace formats.

## Use cases
- pretrain large language models on Cerebras hardware
- fine-tune Llama or Mixtral models
- convert HuggingFace checkpoints to Cerebras format
- preprocess and tokenize datasets for model training
- train vision and multimodal models like DINOv2 and Llava
- apply muP scaling and RoPE scaling to model training

## When to choose
- you have access to Cerebras hardware and want reference model implementations
- you need optimized training configs for NLP, vision, or multimodal models
- you want to port PyTorch models to Cerebras systems
- you need checkpoint conversion between Cerebras and HuggingFace formats

## When to avoid
- you only train on GPUs or TPUs without Cerebras hardware
- you need inference-only tooling rather than training
- you want a general-purpose model hub like HuggingFace Transformers

## Facets
- artifact type: library
- maturity: active
- function: deep-learning, llm-training, machine-learning, cli, etl
- domain: deep-learning, large-language-models, machine-learning, computer-vision, gpu-computing
- platform: python, cli
- tags: model-zoo, cerebras-hardware, pytorch, checkpoint-conversion, fine-tuning, pretraining, multimodal-models, natural-language-processing, gpu, linux

## Member repositories
- Cerebras/modelzoo (main) score 77

## Provenance
- Observed fields: from GitHub, fetched 2026-09-01T02:13:54.361200+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T06:22:26.946550+00:00, confidence not recorded.
  - readme: https://github.com/Cerebras/modelzoo (fetched 2026-09-01T02:13:54.361200+00:00, sha eca853e7fe76)
- Data as of 2026-08-30T08:39:29.467469+00:00.