From fdbd45a5a8319c403d6427081816d5b38cea62cd Mon Sep 17 00:00:00 2001 From: Aasheesh Singh Date: Fri, 9 Feb 2024 10:37:24 -0500 Subject: [PATCH] readme fix --- bert-burn/README.md | 3 +-- bert-burn/examples/infer-embedding.rs | 2 +- 2 files changed, 2 insertions(+), 3 deletions(-) diff --git a/bert-burn/README.md b/bert-burn/README.md index c1be19a..92fa475 100644 --- a/bert-burn/README.md +++ b/bert-burn/README.md @@ -1,8 +1,7 @@ # Bert-Burn Model This project provides an example implementation for inference on the BERT family of models. The following compatible -bert-variants: `roberta-base`**(default) -**/`roberta-large`, `bert-base-uncased`/`bert-large-uncased`/`bert-base-cased`/`bert-large-cased` +bert-variants: `roberta-base`(**default**)/`roberta-large`, `bert-base-uncased`/`bert-large-uncased`/`bert-base-cased`/`bert-large-cased` can be loaded as following. The pre-trained weights and config files are automatically downloaded from: [HuggingFace Model hub](https://huggingface.co/FacebookAI/roberta-base/tree/main) diff --git a/bert-burn/examples/infer-embedding.rs b/bert-burn/examples/infer-embedding.rs index a465464..4416cb6 100644 --- a/bert-burn/examples/infer-embedding.rs +++ b/bert-burn/examples/infer-embedding.rs @@ -18,7 +18,7 @@ pub fn launch(device: B::Device) { let default_model = "roberta-base".to_string(); let model_variant = if args.len() > 1 { // Use the argument provided by the user - // Possible values: "roberta-base", "roberta-large" + // Possible values: "bert-base-uncased", "roberta-large" etc. &args[1] } else { // Use the default value if no argument is provided