Flan-T5 Text2Text Inference

Type your prompt and pick a Flan-T5 model. For best performance, enable a GPU in the Space settings.

Input prompt

Examples

Model

Max length (tokens)

32 1024

Temperature (0.0 = deterministic)

0 1.5

Number of outputs

1 3

Model output

Notes:

If you want faster generation, set Space to use a GPU (in Settings → Hardware).
Larger models (flan-t5-large / flan-t5-xl) need more RAM — they may OOM on CPU or free GPU tiers.
You can add other models to DEFAULT_MODELS above or input a full model name from the Hub.