diff --git a/clamp/README.md b/clamp/README.md
index db5efe6..88118e5 100644
--- a/clamp/README.md
+++ b/clamp/README.md
@@ -3,17 +3,17 @@ The intellectual property of the CLaMP project is owned by the Central Conservat
## Model description
In [CLaMP: Contrastive Language-Music Pre-training for Cross-Modal Symbolic Music Information Retrieval](https://ai-muzic.github.io/clamp/), we introduce a solution for cross-modal symbolic MIR that utilizes contrastive learning and pre-training. The proposed approach, CLaMP: Contrastive Language-Music Pre-training, which learns cross-modal representations between natural language and symbolic music using a music encoder and a text encoder trained jointly with a contrastive loss. To pre-train CLaMP, we collected a large dataset of 1.4 million music-text pairs. It employed text dropout as a data augmentation technique and bar patching to efficiently represent music data which reduces sequence length to less than 10%. In addition, we developed a masked music model pre-training objective to enhance the music encoder's comprehension of musical context and structure. CLaMP integrates textual information to enable semantic search and zero-shot classification for symbolic music, surpassing the capabilities of previous models. To support the evaluation of semantic search and music classification, we publicly release [WikiMusicText](https://huggingface.co/datasets/sander-wood/wikimusictext) (WikiMT), a dataset of 1010 lead sheets in ABC notation, each accompanied by a title, artist, genre, and description. In comparison to state-of-the-art models that require fine-tuning, zero-shot CLaMP demonstrated comparable or superior performance on score-oriented datasets.
-
-
The architecture of CLaMP, including two encoders - one for music and one for text - trained jointly with a contrastive loss to learn cross-modal representations.
-
+
+
The architecture of CLaMP, including two encoders - one for music and one for text - trained jointly with a contrastive loss to learn cross-modal representations.
+
Two variants of CLaMP are introduced: [CLaMP-S/512](https://huggingface.co/sander-wood/clamp-small-512) and [CLaMP-S/1024](https://huggingface.co/sander-wood/clamp-small-1024). Both models consist of a 6-layer music encoder and a 6-layer text encoder with a hidden size of 768. While CLaMP-S/512 accepts input music sequences of up to 512 tokens in length, CLaMP-S/1024 allows for up to 1024 tokens. The maximum input length for the text encoder in both models is 128 tokens. These models are part of [Muzic](https://github.com/microsoft/muzic), a research initiative on AI music that leverages deep learning and artificial intelligence to enhance music comprehension and generation.
## Cross-Modal Symbolic MIR
CLaMP is capable of aligning symbolic music and natural language, which can be used for various cross-modal retrieval tasks, including semantic search and zero-shot classification for symbolic music.
-
-
The processes of CLaMP performing cross-modal symbolic MIR tasks, including semantic search and zero-shot classification for symbolic music, without requiring task-specific training data.
-
+
+
The processes of CLaMP performing cross-modal symbolic MIR tasks, including semantic search and zero-shot classification for symbolic music, without requiring task-specific training data.
+
Semantic search is a technique for retrieving music by open-domain queries, which differs from traditional keyword-based searches that depend on exact matches or meta-information. This involves two steps: 1) extracting music features from all scores in the library, and 2) transforming the query into a text feature. By calculating the similarities between the text feature and the music features, it can efficiently locate the score that best matches the user's query in the library.
Zero-shot classification refers to the classification of new items into any desired label without the need for training data. It involves using a prompt template to provide context for the text encoder. For example, a prompt such as "This piece of music is composed by {composer}." is utilized to form input texts based on the names of candidate composers. The text encoder then outputs text features based on these input texts. Meanwhile, the music encoder extracts the music feature from the unlabelled target symbolic music. By calculating the similarity between each candidate text feature and the target music feature, the label with the highest similarity is chosen as the predicted one.