Add a quantized version of the t5 model. (#921)

This commit is contained in:
Laurent Mazare
2023-09-21 11:13:39 +01:00
committed by GitHub
parent c89b82b2d4
commit 2619c4307f
5 changed files with 873 additions and 2 deletions

View File

@ -1,4 +1,4 @@
// T5 Text Encoder
// T5 Text Model
// https://github.com/huggingface/transformers/blob/main/src/transformers/models/t5/modeling_t5.py
use candle::{DType, Device, Module, Result, Tensor, D};