krzysztof/landing_page
1
0
Fork 0
mirror of https://github.com/qdrant/landing_page.git synced 2026-09-25 14:08:30 +02:00
Code Issues Packages Projects Releases Wiki Activity

fix code and explanations

Browse Source
This commit is contained in:
davidmyriel
2024-08-06 15:42:27 -07:00
parent 5eb0a24ef1
commit 5448903014
4 changed files with 10 additions and 95 deletions
Download Patch File Download Diff File Expand all files Collapse all files
qdrant-landing/content/documentation/fastembed/_index.md
+5 -77
View File
@@ -16,83 +16,11 @@ FastEmbed easily integrates with Qdrant for a variety of multimodal search purpo
|:-:|:-:|
|[Generate Text Embedings with FastEmbed](fastembed-quickstart/)|[Combine FastEmbed with Qdrant for Vector Search](fastembed-semantic-search/)|
## Why is FastEmbed popular?
## Why is FastEmbed useful?
- Light & Fast: FastEmbed is designed to be efficient and quick, utilizing quantized model weights to reduce the computational load.
- ONNX Runtime for Inference: The library leverages ONNX Runtime, which allows for fast inference on various hardware platforms, ensuring high performance.
- Accuracy and Recall: FastEmbed aims to provide better accuracy and recall compared to other models like OpenAI’s Ada-002. The default embedding model is `BAAI/bge-small-en-v1.5`, which has shown strong results on the MTEB leaderboard.
- Supported Models: FastEmbed supports many models, including multilingual ones, to cater to a wide range of use cases.
- Light: Unlike other inference frameworks, such as PyTorch, FastEmbed requires very little external dependencies. Because it uses the ONNX runtime, it is perfect for serverless environments like AWS Lambda.
- Fast: By using ONNX, FastEmbed ensures high-performance inference across various hardware platforms.
- Accurate: FastEmbed aims for better accuracy and recall than models like OpenAI’s `Ada-002`. It always uses model which demonstrate strong results on the MTEB leaderboard.
- Support: FastEmbed supports a wide range of models, including multilingual ones, to meet diverse use case needs.
## Supported Models
We are continuously updating the list with better and useful models. Please note that this list changes often.
Click on each list to expand it.
<details>
<summary> Text Embedding Models </summary>
| model | dim | description | size_in_GB |
|--------------------------------------------|-----|---------------------------------------------------------------------------------|------------|
| BAAI/bge-small-en-v1.5 | 384 | Fast and Default English model | 0.067 |
| BAAI/bge-small-zh-v1.5 | 512 | Fast and recommended Chinese model | 0.090 |
| snowflake/snowflake-arctic-embed-xs | 384 | Based on all-MiniLM-L6-v2 model with only 22m ... | 0.090 |
| sentence-transformers/all-MiniLM-L6-v2 | 384 | Sentence Transformer model, MiniLM-L6-v2 | 0.090 |
| jinaai/jina-embeddings-v2-small-en | 512 | English embedding model supporting 8192 sequen... | 0.120 |
| BAAI/bge-small-en | 384 | Fast English model | 0.130 |
| snowflake/snowflake-arctic-embed-s | 384 | Based on infloat/e5-small-unsupervised, does n... | 0.130 |
| nomic-ai/nomic-embed-text-v1.5-Q | 768 | Quantized 8192 context length english model | 0.130 |
| BAAI/bge-base-en-v1.5 | 768 | Base English model, v1.5 | 0.210 |
| sentence-transformers/paraphrase-multilingual-... | 384 | Sentence Transformer model, paraphrase-multili... | 0.220 |
| Qdrant/clip-ViT-B-32-text | 512 | CLIP text encoder | 0.250 |
| jinaai/jina-embeddings-v2-base-de | 768 | German embedding model supporting 8192 sequenc... | 0.320 |
| BAAI/bge-base-en | 768 | Base English model | 0.420 |
| snowflake/snowflake-arctic-embed-m | 768 | Based on intfloat/e5-base-unsupervised model, ... | 0.430 |
| nomic-ai/nomic-embed-text-v1.5 | 768 | 8192 context length english model | 0.520 |
| jinaai/jina-embeddings-v2-base-en | 768 | English embedding model supporting 8192 sequen... | 0.520 |
| nomic-ai/nomic-embed-text-v1 | 768 | 8192 context length english model | 0.520 |
| snowflake/snowflake-arctic-embed-m-long | 768 | Based on nomic-ai/nomic-embed-text-v1-unsuperv... | 0.540 |
| mixedbread-ai/mxbai-embed-large-v1 | 1024 | MixedBread Base sentence embedding model, does... | 0.640 |
| jinaai/jina-embeddings-v2-base-code | 768 | Source code embedding model supporting 8192 se... | 0.640 |
| sentence-transformers/paraphrase-multilingual-... | 768 | Sentence-transformers model for tasks like clu... | 1.000 |
| snowflake/snowflake-arctic-embed-l | 1024 | Based on intfloat/e5-large-unsupervised, large... | 1.020 |
| thenlper/gte-large | 1024 | Large general text embeddings model | 1.200 |
| BAAI/bge-large-en-v1.5 | 1024 | Large English model, v1.5 | 1.200 |
| intfloat/multilingual-e5-large | 1024 | Multilingual model, e5-large. Recommend using ... | 2.240 |
</details>
<details>
<summary> Image Embedding Models </summary>
| model | dim | description | size_in_GB |
|-----------------------------|------|------------------------------------------------------|------------|
| Qdrant/resnet50-onnx | 2048 | ResNet-50 from `Deep Residual Learning for Ima... | 0.10 |
| Qdrant/clip-ViT-B-32-vision | 512 | CLIP vision encoder based on ViT-B/32 | 0.34 |
| Qdrant/Unicom-ViT-B-32 | 512 | Unicom Unicom-ViT-B-32 from open-metric-learning | 0.48 |
| Qdrant/Unicom-ViT-B-16 | 768 | Unicom Unicom-ViT-B-16 from open-metric-learning | 0.82 |
</details>
<details>
<summary> Sparse Embedding Models </summary>
| model | vocab_size | description | size_in_GB | requires_idf |
|---------------------------------------------|------------|-----------------------------------------------------------------------------------|------------|--------------|
| Qdrant/bm25 | NaN | BM25 as sparse embeddings meant to be used with... | 0.010 | True |
| Qdrant/bm42-all-minilm-l6-v2-attentions | 30522.0 | Light sparse embedding model, which assigns an... | 0.090 | True |
| prithvida/Splade_PP_en_v1 | 30522.0 | Misspelled version of the model. Retained for... | 0.532 | NaN |
| prithivida/Splade_PP_en_v1 | 30522.0 | Independent Implementation of SPLADE++ Model for dense and sparse retrieval tasks | 0.532 | NaN |
</details>
<details>
<summary> Late Interaction Embedding Models </summary>
| model | dim | description | size_in_GB |
|--------------------------|-----|-------------------------|------------|
| colbert-ir/colbertv2.0 | 128 | Late interaction model | 0.44 |
</details>
qdrant-landing/content/documentation/fastembed/fastembed-quickstart.md
+1 -14
View File
@@ -54,20 +54,7 @@ Document: fastembed is supported by and maintained by Qdrant.
Vector of type: <class 'numpy.ndarray'> with shape: (384,)
```
## Verify embedding shape
Next, take the embeddings and convert them into a NumPy array.
Then, check the shape of this array to confirm the dimensions of the generated embeddings.
```python
embeddings_list = np.array(list(embedding_model.embed(documents)))
embeddings_list.shape
```
The shape will be a tuple indicating the number of documents and the dimensions of the embeddings. For example, if there are 2 documents and each embedding has 384 dimensions, the shape will be (2, 384).
```python
(2, 384)
```
## See the embeddings
## Visualize embeddings
```python
print("Embeddings:\n", embeddings_list)
```
qdrant-landing/content/documentation/fastembed/fastembed-semantic-search.md
+1 -1
View File
@@ -6,7 +6,6 @@ weight: 3
# Using FastEmbed with Qdrant for Vector Search
## Install Qdrant Client
The Qdrant Python Client has a simple in-memory mode that lets you try semantic search locally.
```python
pip install qdrant-client
```
@@ -18,6 +17,7 @@ pip install fastembed
```
## Initialize the client
Qdrant Client has a simple in-memory mode that lets you try semantic search locally.
```python
from qdrant_client import QdrantClient
qdrant-landing/content/documentation/fastembed/fastembed-splade.md
+3 -3
View File
@@ -73,7 +73,7 @@ sparse_embeddings_list: List[SparseEmbedding] = list(
```
## Retrieve embeddings
`sparse_embeddings_list` contains sparse embeddings for the documents provided earlier. Each element in this list is a `SparseEmbedding` object that contains the sparse vector representation of a document. Set the variable `index` to 0. This will retrieve data from the first document in the list.
`sparse_embeddings_list` contains sparse embeddings for the documents provided earlier. Each element in this list is a `SparseEmbedding` object that contains the sparse vector representation of a document.
```python
index = 0
@@ -120,9 +120,9 @@ Let's use the tokenizer vocab to make sense of these indices.
```python
import json
from transformers import AutoTokenizer
from tokenizers import Tokenizer
tokenizer = AutoTokenizer.from_pretrained(SparseTextEmbedding.list_supported_models()[0]["sources"]["hf"])
tokenizer = Tokenizer.from_pretrained(SparseTextEmbedding.list_supported_models()[0]["sources"]["hf"])
```
The `get_tokens_and_weights` function takes a `SparseEmbedding` object and a `tokenizer` as input. It will construct a dictionary where the keys are the decoded tokens, and the values are their corresponding weights.