Spaces:

LeoBorai
/

tfidf-sentiment-analysis

Sleeping

LeoBorai commited on Sep 26

Commit

3fdf998

1 Parent(s): 470fea7

feat: train larger model

Files changed (2) hide show

assets/model.json CHANGED Viewed

The diff for this file is too large to render. See raw diff

model/src/model/gpu.rs CHANGED Viewed

@@ -1,7 +1,6 @@
 //! Metal-accelerated Batch Gradient Descent, using `candle` to run the
 //! matmul-heavy training loop on the local machine's GPU (e.g. Apple
-//! Silicon). Only built when the `training` feature is enabled - the
-//! deployed server never links `candle`.
 use candle_core::{DType, Device, Tensor};
 use ndarray::Array2;

 //! Metal-accelerated Batch Gradient Descent, using `candle` to run the
 //! matmul-heavy training loop on the local machine's GPU (e.g. Apple
+//! Silicon).
 use candle_core::{DType, Device, Tensor};
 use ndarray::Array2;