Unified text-to-text framework
According to the model card, all NLP tasks are reframed into a text-to-text format so the same model, loss function, and hyperparameters cover translation, summarization, question answering, and classification.
Open Source Model Profile · google-t5
t5-base
t5-base is the 220M-parameter Text-To-Text Transfer Transformer checkpoint from google-t5. Its model card describes a unified text-to-text framework covering translation, summarization, question answering, and classification.
- Publisher
- google-t5
- Task
- translation
- Model type
- t5
- License
- apache-2.0
- Library
- transformers
- Publication status
- Approved for indexing
Model overview
t5-base is published by google-t5 as a translation-pipeline model. The captured configuration identifies T5ForConditionalGeneration with model type t5, and Safetensors metadata reports 222,903,936 parameters while the model card describes the checkpoint as having 220 million parameters. The model card documents Apache 2.0 licensing, English, French, Romanian, and German language coverage, and a text-to-text design covering translation, summarization, question answering, and classification.
Recorded capabilities
C4 plus multi-task pre-training
According to the model card, pre-training combines the Colossal Clean Crawled Corpus with a multi-task mixture of unsupervised denoising and supervised objectives across many language-understanding datasets.
Apache-2.0 licensing
Card data records license apache-2.0, and the model card states Apache 2.0.
Translation and summarization tagging
Hub data marks the translation pipeline with text2text-generation, summarization, and translation tags.
Use cases in the source record
- Machine translation across the documented English, French, Romanian, and German coverage using the text-to-text framework.
- Document summarization, question answering, sentiment classification, and other text-to-text tasks the model card lists as covered by the same model and hyperparameters.
Limitations and unknowns
- No evaluation results were extracted from this record.
- No context-window value was extracted from this record.
- Provider state is historical snapshot data, not independently refreshed current availability.
Source and provenance
Source: google-t5/t5-base
Captured: Unknown. Processed: 2026-09-07T19:34:46.093677+00:00.
Model Card for T5 Base Table of Contents Model Details Uses Bias, Risks, and Limitations Training Details Evaluation Environmental Impact Citation Model Card Authors How To Get Started With the Model Model Details Model Description The developers of the Text-To-Text Transfer Transformer (T5) write : With T5, we propose reframing all NLP tasks into a unified text-to-text-format where the input and output are always text strings, in contrast to BERT-style models that can only output either a class label or a span of the input. Our text-to-text framework allows us to use the same model, loss function, and hyperparameters on any NLP tas…
F001F002F003F004F005F006F007F009F010F014F015F016F017F018F019