> ## Documentation Index
> Fetch the complete documentation index at: https://liquidai-codex-model-library-starting-point.mintlify.site/llms.txt
> Use this file to discover all available pages before exploring further.
# LFM2.5-ColBERT-350M
> 350M late-interaction ColBERT model for multilingual retrieval and reranking
← Back to Liquid Nanos
LFM2.5-ColBERT-350M is a late-interaction retrieval model. It creates 128-dimensional vectors per token and scores query/document matches with MaxSim, which improves retrieval quality and generalization at the cost of a larger index.
Use LFM2.5-ColBERT-350M when you want stronger retrieval or reranking quality and can afford a larger per-token index. Use [LFM2.5-Embedding-350M](/lfm/models/lfm25-embedding-350m) when you need the smallest, fastest dense-vector index.
## Specifications
| Property | Value |
| ------------------- | --------------------------------------------------------------------------------------------------- |
| Parameters | \~353M |
| Type | Late interaction |
| Document Length | 512 tokens |
| Query Length | 32 tokens |
| Output | 128-dimensional vector per token |
| Similarity | MaxSim |
| Supported Languages | English, Spanish, German, French, Italian, Portuguese, Arabic, Swedish, Norwegian, Japanese, Korean |
Better matching from token-level interactions.
Reorder candidates from a first-stage retriever.
Strong multilingual document matching.
## Quick Start
This model uses PyLate for indexing, retrieval, and reranking.
**Install:**
```bash theme={null}
pip install -U pylate
```
**Index and retrieve documents:**
```python theme={null}
from pylate import indexes, models, retrieve
model = models.ColBERT(
model_name_or_path="LiquidAI/LFM2.5-ColBERT-350M",
trust_remote_code=True,
)
model.tokenizer.pad_token = model.tokenizer.eos_token
index = indexes.PLAID(
index_folder="pylate-index",
index_name="index",
override=True,
)
documents_ids = ["1", "2", "3"]
documents = [
"Paris is the capital of France.",
"Tokyo is the capital of Japan.",
"Berlin is the capital of Germany.",
]
document_embeddings = model.encode(
documents,
batch_size=32,
is_query=False,
show_progress_bar=True,
)
index.add_documents(
documents_ids=documents_ids,
documents_embeddings=document_embeddings,
)
query_embeddings = model.encode(
["Which city is Japan's capital?"],
batch_size=32,
is_query=True,
show_progress_bar=True,
)
retriever = retrieve.ColBERT(index=index)
results = retriever.retrieve(query_embeddings=query_embeddings, k=10)
print(results)
```
```python theme={null}
from pylate import models, rank
model = models.ColBERT(
model_name_or_path="LiquidAI/LFM2.5-ColBERT-350M",
trust_remote_code=True,
)
queries = ["Which city is Japan's capital?"]
documents = [[
"Paris is the capital of France.",
"Tokyo is the capital of Japan.",
"Berlin is the capital of Germany.",
]]
document_ids = [["fr", "jp", "de"]]
query_embeddings = model.encode(queries, is_query=True)
document_embeddings = model.encode(documents, is_query=False)
reranked = rank.rerank(
documents_ids=document_ids,
queries_embeddings=query_embeddings,
documents_embeddings=document_embeddings,
)
print(reranked)
```
**Download GGUF:**
```bash theme={null}
hf download LiquidAI/LFM2.5-ColBERT-350M-GGUF \
--local-dir ./LFM2.5-ColBERT-350M-GGUF
```
Use the GGUF files with a llama.cpp build that supports LFM2.5 ColBERT models.