From 048bc5b4b79f56b3a77eeffa518a4e215ec159fe Mon Sep 17 00:00:00 2001 From: Nicolas Patry Date: Tue, 23 Jan 2024 15:04:17 +0100 Subject: [PATCH] Remove special, it's not correct enough (and not necessarily useful). --- router/src/server.rs | 8 ++++---- router/src/validation.rs | 2 +- 2 files changed, 5 insertions(+), 5 deletions(-) diff --git a/router/src/server.rs b/router/src/server.rs index 42785025..678ea9c1 100644 --- a/router/src/server.rs +++ b/router/src/server.rs @@ -696,13 +696,13 @@ async fn tokenize( .get_ids() .iter() .zip(encoding.get_offsets()) - .map(|(&id, (start, stop))| { - let text: String = input.chars().skip(*start).take(stop - start).collect(); + .map(|(&id, &(start, stop))| { + let text: String = input.chars().skip(start).take(stop - start).collect(); SimpleToken { id, text, - start: *start, - stop: *stop, + start, + stop, } }) .collect(); diff --git a/router/src/validation.rs b/router/src/validation.rs index e6dbcf81..adbe6bb2 100644 --- a/router/src/validation.rs +++ b/router/src/validation.rs @@ -365,7 +365,7 @@ fn prepare_input( ) -> Result<(tokenizers::Encoding, String), ValidationError> { // Get the number of tokens in the input let mut encoding = tokenizer - .encode(inputs.clone(), true) + .encode(inputs.clone(), false) .map_err(|err| ValidationError::Tokenizer(err.to_string()))?; // Optionally truncate