Add a cuda kernel for upsampling. (#441)

* Add a cuda kernel for upsampling.

* Update for the latest tokenizers version.

1 files changed, 1 insertions, 4 deletions

diff --git a/candle-examples/examples/bigcode/main.rs b/candle-examples/examples/bigcode/main.rs
index ac9c63c7..39b1de27 100644
--- a/candle-examples/examples/bigcode/main.rs
+++ b/candle-examples/examples/bigcode/main.rs
@@ -65,10 +65,7 @@ impl TextGeneration {
             let next_token = self.logits_processor.sample(&logits)?;
             tokens.push(next_token);
             new_tokens.push(next_token);
-            let token = self
-                .tokenizer
-                .decode(vec![next_token], true)
-                .map_err(E::msg)?;
+            let token = self.tokenizer.decode(&[next_token], true).map_err(E::msg)?;
             print!("{token}");
             std::io::stdout().flush()?;
         }