Skip to content

Train and Infer Powerful Sentence Embeddings with AnglE | 🔥 SOTA on STS and MTEB Leaderboard

License

Notifications You must be signed in to change notification settings

SeanLee97/AnglE

Repository files navigation

EN | 简体中文

AnglE 📐

Sponsored by Mixedbread

For more detailed usage, please read the 📘 document: https://angle.readthedocs.io/en/latest/index.html

https://arxiv.org/abs/2309.12871 PyPI version PyPI Downloads Read the docs

PWC PWC PWC PWC PWC PWC PWC

📢 Train/Infer Powerful Sentence Embeddings with AnglE. This library is from the paper: AnglE: Angle-optimized Text Embeddings. It allows for training state-of-the-art BERT/LLM-based sentence embeddings with just a few lines of code. AnglE is also a general sentence embedding inference framework, allowing for infering a variety of transformer-based sentence embeddings.

✨ Features

Loss:

  • 📐 AnglE loss
  • ⚖ Contrastive loss
  • 📏 CoSENT loss
  • ☕️ Espresso loss (previously known as 2DMSE, detail: README_ESE)

Backbones:

  • BERT-based models (BERT, RoBERTa, ELECTRA, ALBERT, etc.)
  • LLM-based models (LLaMA, Mistral, Qwen, etc.)
  • Bi-directional LLM-based models (LLaMA, Mistral, Qwen, OpenELMo, etc.. refer to: https://github.com/WhereIsAI/BiLLM)

Training:

  • Single-GPU training
  • Multi-GPU training

http://makeapullrequest.com More features will be added in the future.

🏆 Achievements

📅 May 16, 2024 | Paper "AnglE: Angle-optimized Text Embeddings" is accepted by ACL 2024 Main Conference.

📅 Mar 13, 2024 | Paper "BeLLM: Backward Dependency Enhanced Large Language Model for Sentence Embeddings" is accepted by NAACL 2024 Main Conference.

📅 Mar 8, 2024 | 🍞 mixedbread's embedding (mixedbread-ai/mxbai-embed-large-v1) achieves SOTA on the MTEB Leaderboard with an average score of 64.68! The model is trained using AnglE. Congrats mixedbread!

📅 Dec 4, 2023 | Our universal sentence embedding WhereIsAI/UAE-Large-V1 achieves SOTA on the MTEB Leaderboard with an average score of 64.64! The model is trained using AnglE.

📅 Dec, 2023 | AnglE achieves SOTA performance on the STS Bechmark Semantic Textual Similarity!

🤗 Official Pretrained Models

BERT-based models:

🤗 HF Max Tokens Pooling Strategy Scenario
WhereIsAI/UAE-Large-V1 512 cls English, General-purpose
WhereIsAI/UAE-Code-Large-V1 512 cls Code Similarity
WhereIsAI/pubmed-angle-base-en 512 cls Medical Similarity
WhereIsAI/pubmed-angle-large-en 512 cls Medical Similarity

LLM-based models:

🤗 HF (lora weight) Backbone Max Tokens Prompts Pooling Strategy Scenario
SeanLee97/angle-llama-13b-nli NousResearch/Llama-2-13b-hf 4096 Prompts.A last token English, Similarity Measurement
SeanLee97/angle-llama-7b-nli-v2 NousResearch/Llama-2-7b-hf 4096 Prompts.A last token English, Similarity Measurement

💡 You can find more third-party embeddings trained with AnglE in HuggingFace Collection

🚀 Quick Start

⬇️ Installation

python -m pip install -U angle-emb

⌛ Infer BERT-based Model

Open In Colab

  1. With Prompts: You can specify a prompt with prompt=YOUR_PROMPT in encode method. If set a prompt, the inputs should be a list of dict or a single dict with key text, where text is the placeholder in the prompt for the input text. You can use other placeholder names. We provide a set of predefined prompts in Prompts class, you can check them via Prompts.list_prompts().
from angle_emb import AnglE, Prompts
from angle_emb.utils import cosine_similarity


angle = AnglE.from_pretrained('WhereIsAI/UAE-Large-V1', pooling_strategy='cls').cuda()
# For retrieval tasks, we use `Prompts.C` as the prompt for the query when using UAE-Large-V1 (no need to specify prompt for documents).
# When specify prompt, the inputs should be a list of dict with key 'text'
qv = angle.encode({'text': 'what is the weather?'}, to_numpy=True, prompt=Prompts.C)
doc_vecs = angle.encode([
    'The weather is great!',
    'it is rainy today.',
    'i am going to bed'
], to_numpy=True)

for dv in doc_vecs:
    print(cosine_similarity(qv[0], dv))
  1. Without Prompts: no need to specify a prompt. Just input a list of strings or a single string.
from angle_emb import AnglE
from angle_emb.utils import cosine_similarity


angle = AnglE.from_pretrained('WhereIsAI/UAE-Large-V1', pooling_strategy='cls').cuda()
# for non-retrieval tasks, we don't need to specify prompt when using UAE-Large-V1.
doc_vecs = angle.encode([
    'The weather is great!',
    'The weather is very good!',
    'i am going to bed'
])

for i, dv1 in enumerate(doc_vecs):
    for dv2 in doc_vecs[i+1:]:
        print(cosine_similarity(dv1, dv2))

⌛ Infer LLM-based Models

Open In Colab

If the pretrained weight is a LoRA-based model, you need to specify the backbone via model_name_or_path and specify the LoRA path via the pretrained_lora_path in from_pretrained method.

import torch
from angle_emb import AnglE, Prompts
from angle_emb.utils import cosine_similarity

angle = AnglE.from_pretrained('NousResearch/Llama-2-7b-hf',
                              pretrained_lora_path='SeanLee97/angle-llama-7b-nli-v2',
                              pooling_strategy='last',
                              is_llm=True,
                              torch_dtype=torch.float16).cuda()
print('All predefined prompts:', Prompts.list_prompts())
doc_vecs = angle.encode([
    {'text': 'The weather is great!'},
    {'text': 'The weather is very good!'},
    {'text': 'i am going to bed'}
], prompt=Prompts.A)

for i, dv1 in enumerate(doc_vecs):
    for dv2 in doc_vecs[i+1:]:
        print(cosine_similarity(dv1, dv2))

⌛ Infer BiLLM-based Models

Open In Colab

Specify apply_billm and billm_model_class to load and infer billm models

import os
# set an environment variable for billm start index
os.environ['BiLLM_START_INDEX'] = '31'

import torch
from angle_emb import AnglE, Prompts
from angle_emb.utils import cosine_similarity

# specify `apply_billm` and `billm_model_class` to load billm models
angle = AnglE.from_pretrained('NousResearch/Llama-2-7b-hf',
                              pretrained_lora_path='SeanLee97/bellm-llama-7b-nli',
                              pooling_strategy='last',
                              is_llm=True,
                              apply_billm=True,
                              billm_model_class='LlamaForCausalLM',
                              torch_dtype=torch.float16).cuda()

doc_vecs = angle.encode([
    {'text': 'The weather is great!'},
    {'text': 'The weather is very good!'},
    {'text': 'i am going to bed'}
], prompt='The representative word for sentence {text} is:"')

for i, dv1 in enumerate(doc_vecs):
    for dv2 in doc_vecs[i+1:]:
        print(cosine_similarity(dv1, dv2))

⌛ Infer Espresso/Matryoshka Models

Open In Colab

Specify layer_index and embedding_size to truncate embeddings.

from angle_emb import AnglE
from angle_emb.utils import cosine_similarity


angle = AnglE.from_pretrained('mixedbread-ai/mxbai-embed-2d-large-v1', pooling_strategy='cls').cuda()
# truncate layer
angle = angle.truncate_layer(layer_index=22)
# specify embedding size to truncate embeddings
doc_vecs = angle.encode([
    'The weather is great!',
    'The weather is very good!',
    'i am going to bed'
], embedding_size=768)

for i, dv1 in enumerate(doc_vecs):
    for dv2 in doc_vecs[i+1:]:
        print(cosine_similarity(dv1, dv2))

⌛ Infer Third-party Models

You can load any transformer-based third-party models such as mixedbread-ai/mxbai-embed-large-v1, sentence-transformers/all-MiniLM-L6-v2, and BAAI/bge-large-en-v1.5 using angle_emb.

Here is an example:

from angle_emb import AnglE

model = AnglE.from_pretrained('mixedbread-ai/mxbai-embed-large-v1', pooling_strategy='cls').cuda()
vec = model.encode('hello world', to_numpy=True)
print(vec)

Batch Inference

It is recommended to use Mixedbread's batched library to speed up the inference process.

python -m pip install batched
import batched
from angle_emb import AnglE

model = AnglE.from_pretrained("WhereIsAI/UAE-Large-V1", pooling_strategy='cls').cuda()
model.encode = batched.dynamically(model.encode, batch_size=64)

vecs = model.encode([
    'The weather is great!',
    'The weather is very good!',
    'i am going to bed'
] * 50)

🕸️ Custom Train

💡 For more details, please refer to the training and fintuning.

🗂️ 1. Data Prepation

We currently support three dataset formats:

  1. DatasetFormats.A: it is a pair format with three columns: text1, text2, and label (0/1).

  2. DatasetFormats.B: it is a triple format with three columns: text, positive, and negative. positive and negative store the positive and negative samples of text.

  3. DatasetFormats.C: it is a pair format with two columns: text, positive. positive store the positive sample of text.

You need to prepare your data into huggingface datasets.Dataset in one of the formats in terms of your supervised data.

🚂 2. Train with CLI [Recommended]

Use angle-trainer to train your AnglE model in cli mode.

  1. Single gpu training:

Usage:

CUDA_VISIBLE_DEVICES=0 angle-trainer --help
  1. Multi-gpu training:

Usage:

CUDA_VISIBLE_DEVICES=0,1 torchrun --nproc_per_node=2 --master_port=1234 -m angle_emb.angle_trainer --help

🚂 3. Custom Train

Open In Colab

from datasets import load_dataset
from angle_emb import AnglE, AngleDataTokenizer


# 1. load pretrained model
angle = AnglE.from_pretrained('SeanLee97/angle-bert-base-uncased-nli-en-v1', max_length=128, pooling_strategy='cls').cuda()

# 2. load dataset
# `text1`, `text2`, and `label` are three required columns.
ds = load_dataset('mteb/stsbenchmark-sts')
ds = ds.map(lambda obj: {"text1": str(obj["sentence1"]), "text2": str(obj['sentence2']), "label": obj['score']})
ds = ds.select_columns(["text1", "text2", "label"])

# 3. transform data
train_ds = ds['train'].shuffle().map(AngleDataTokenizer(angle.tokenizer, angle.max_length), num_proc=8)
valid_ds = ds['validation'].map(AngleDataTokenizer(angle.tokenizer, angle.max_length), num_proc=8)

# 4. fit
angle.fit(
    train_ds=train_ds,
    valid_ds=valid_ds,
    output_dir='ckpts/sts-b',
    batch_size=32,
    epochs=5,
    learning_rate=2e-5,
    save_steps=100,
    eval_steps=1000,
    warmup_steps=0,
    gradient_accumulation_steps=1,
    loss_kwargs={
        'cosine_w': 1.0,
        'ibn_w': 20.0,
        'angle_w': 1.0,
        'cosine_tau': 20,
        'ibn_tau': 20,
        'angle_tau': 20
    },
    fp16=True,
    logging_steps=100
)

# 5. evaluate
corrcoef = angle.evaluate(ds['test'])
print('Spearman\'s corrcoef:', corrcoef)

💡 Others

  • To enable llm training, please specify --is_llm 1 and configure appropriate LoRA hyperparameters.
  • To enable billm training, please specify --apply_billm 1 and configure appropriate billm_model_class such as LLamaForCausalLM (refer to: https://github.com/WhereIsAI/BiLLM?tab=readme-ov-file#usage).
  • To enable espresso sentence embeddings (ESE), please specify --apply_ese 1 and configure appropriate ESE hyperparameters via --ese_kl_temperature float and --ese_compression_size integer.
  • To convert the trained AnglE models to sentence-transformers, please run python scripts/convert_to_sentence_transformers.py --help for more details.

💡 4. Fine-tuning Tips

1️⃣ If your dataset format is DatasetFormats.A, it is recommended to slightly increase the weight for cosine_w or slightly decrease the weight for ibn_w.

2️⃣ If your dataset format is DatasetFormats.B, it is recommended to set cosine_w to 0, and increase the weight for ibn_w such as 10 and 20. The angle_tau is recommended to set to 20.0.

3️⃣ If your dataset format is DatasetFormats.C, only ibn_w and ibn_tau are effective. You don't need to tune other parameters.

4️⃣ To alleviate information forgetting in fine-tuning, it is better to specify the teacher_name_or_path. If the teacher_name_or_path equals model_name_or_path, it will conduct self-distillation. It is worth to note that teacher_name_or_path has to have the same tokenizer as model_name_or_path. Or it will lead to unexpected results.

5. Finetuning and Infering AnglE with sentence-transformers

  • Training: SentenceTransformers also provides a implementation of AnglE loss. But it is partially implemented and may not work well as the official code. We recommend to use the official angle_emb for fine-tuning AnglE model.

  • Infering: If your model is trained with angle_emb, and you want to use it with sentence-transformers. You can convert it to sentence-transformers model using the script examples/convert_to_sentence_transformers.py.

🫡 Citation

You are welcome to use our code and pre-trained models. If you use our code and pre-trained models, please support us by citing our work as follows:

@article{li2023angle,
  title={AnglE-optimized Text Embeddings},
  author={Li, Xianming and Li, Jing},
  journal={arXiv preprint arXiv:2309.12871},
  year={2023}
}

📜 ChangeLogs

📅 Description
2024 May 21 support Espresso Sentence Embeddings
2024 Feb 7 support training with only positive pairs (DatasetFormats.C)
2023 Dec 4 Release a universal English sentence embedding model: WhereIsAI/UAE-Large-V1
2023 Nov 2 Release an English pretrained model: SeanLee97/angle-llama-13b-nli
2023 Oct 28 Release two chinese pretrained models: SeanLee97/angle-roberta-wwm-base-zhnli-v1 and SeanLee97/angle-llama-7b-zhnli-v1; Add chinese README.md

📧 Contact

If you have any questions or suggestions, please feel free to contact us via email: xmlee97@gmail.com

© License

This project is licensed under the MIT License. For the pretrained models, please refer to the corresponding license of the models.