diff --git a/router/src/server.rs b/router/src/server.rs index 6b7fff81..f273a786 100644 --- a/router/src/server.rs +++ b/router/src/server.rs @@ -1748,7 +1748,7 @@ pub async fn run( let mut tokenizer = Tokenizer::from_file(filename).ok(); if let Some(tokenizer) = &mut tokenizer { if let Some(class) = &tokenizer_config.tokenizer_class { - if class == "LlamaTokenizer" || class == "LlamaTokenizerFast" || class == "CohereTokenizerFast"{ + if class == "LlamaTokenizer" || class == "LlamaTokenizerFast"{ if let Ok(post_processor) = create_post_processor(tokenizer, &tokenizer_config) { tracing::info!("Overriding LlamaTokenizer with TemplateProcessing to follow python override defined in https://github.com/huggingface/transformers/blob/4aa17d00690b7f82c95bb2949ea57e22c35b4336/src/transformers/models/llama/tokenization_llama_fast.py#L203-L205"); tokenizer.with_post_processor(post_processor);