---
pipeline_tag: fill-mask
language: bak
tags:
  - trimmed
library_name: transformers
base_model: google/embeddinggemma-300m
base_model_relation: quantized
datasets:
  - Lumberjackk/fineweb-2-trimming
---

# embeddinggemma-bak-32768

This model is a 58.2% smaller version of [google/embeddinggemma-300m](https://huggingface.co/google/embeddinggemma-300m) 
optimized for Bashkir language via vocabulary trimming mined on [Lumberjackk/fineweb-2-trimming](https://huggingface.co/datasets/Lumberjackk/fineweb-2-trimming).

## Model Statistics
- **Original vocabulary size:** 262,144 tokens 
- **Trimmed vocabulary size:** 32,768 tokens
- **Vocabulary reduction:** 87.5%
- **Original model size:** 302,863,104 parameters  
- **Trimmed model size:** 126,702,336 parameters  
- **Size reduction:** 58.2% 


## Usage

```python
from sentence_transformers import SentenceTransformer

# Download from the 🤗 Hub
model = SentenceTransformer("embeddinggemma-bak-32768")

# Run inference with queries and documents
query = "My query"
documents = [
    "Chunk 1",
    "Chunk 2",
    "Chunk 3",
]
query_embeddings = model.encode_query(query)
document_embeddings = model.encode_document(documents)
print(query_embeddings.shape, document_embeddings.shape)

# Compute similarities to determine a ranking
similarities = model.similarity(query_embeddings, document_embeddings)
print(similarities)
```