# PaddlePaddle/ERNIE

The official repository for ERNIE 4.5 and ERNIEKit – its industrial-grade development toolkit based on PaddlePaddle.

Repository: https://github.com/PaddlePaddle/ERNIE
Canonical: https://ross.abutalabs.com/products/ernie
Homepage: https://ernie.baidu.com
Language: Python
License: Apache-2.0
License Family: permissive
Topics: ernie, erniekit, llm, vlm, ernie-45, ernie-45-vl
Last push: 2026-07-24T21:37:38+00:00

## Health v2 (maintenance only)
Score: 65/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 94, release rhythm 8, longevity 100
- inputs: {"age_days": 2740, "days_push": 40, "days_rel": 430, "gap_med": null, "n_releases_24m": 1}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 7738, forks 1447 (observed 2026-08-28T04:10:02.717885+00:00)

## What it is
Official repository for Baidu's ERNIE 4.5 family of large multimodal models and ERNIEKit, an industrial-grade training toolkit built on PaddlePaddle. It provides model weights, SFT/LoRA fine-tuning, function-call training, and WebUI tooling for text and vision-language MoE models.

## Use cases
- fine-tune ERNIE 4.5 LLMs with SFT or LoRA
- train vision-language models on custom multimodal data
- train models for function calling
- run ERNIE models on GPU clusters with MoE parallelism
- download official ERNIE model weights
- fine-tune PaddleOCR-VL document parsing models

## When to choose
- you want to fine-tune or deploy Baidu's ERNIE model family
- you need industrial-scale MoE or multimodal training on PaddlePaddle
- you want official weights and tooling maintained by the model authors

## When to avoid
- you prefer PyTorch-based ecosystems like Hugging Face Transformers or LLaMA-Factory
- you only need inference without training
- you need smaller community models rather than large MoE checkpoints

## Facets
- artifact type: library
- maturity: active
- function: llm-training, machine-learning, deep-learning, sdk
- domain: large-language-models, artificial-intelligence, deep-learning, machine-learning
- platform: python
- tags: ernie, moe, multimodal, vlm, paddlepaddle, sft, lora, fine-tuning, baidu, model-training, gpu, linux, docker

## Member repositories
- PaddlePaddle/ERNIE (main) score 65

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:10:02.717885+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 2026-08-29T17:35:11.019359+00:00, confidence not recorded.
  - readme: https://github.com/PaddlePaddle/ERNIE (fetched 2026-08-28T04:10:02.717885+00:00, sha 396d596c8e3f)
  - homepage: https://ernie.baidu.com (fetched 2026-08-29T08:31:19.736186+00:00, sha 6126d0eadbd7)
  - site_page: https://ernie.baidu.com/blog/about (fetched 2026-08-29T08:31:19.745208+00:00, sha e8f5f2840a8b)
- Data as of 2026-08-30T08:39:29.467469+00:00.