lucataco / nomic-embed-text-v1

nomic-embed-text-v1 is 8192 context length text encoder that surpasses OpenAI text-embedding-ada-002 and text-embedding-3-small performance on short and long context tasks

replicate.com
Total runs: 3.6K
24-hour runs: 0
7-day runs: 0
30-day runs: 0
Github
Model's Last Updated: 2024年2月12日

Introduction of nomic-embed-text-v1

Model Details of nomic-embed-text-v1

Readme

nomic-embed-text-v1: A Reproducible Long Context (8192) Text Embedder

nomic-embed-text-v1 is 8192 context length text encoder that surpasses OpenAI text-embedding-ada-002 and text-embedding-3-small performance on short and long context tasks.

Name SeqLen MTEB LoCo Jina Long Context Open Weights Open Training Code Open Data
nomic-embed-text-v1 8192 62.39 85.53 54.16
jina-embeddings-v2-base-en 8192 60.39 85.45 51.90
text-embedding-3-small 8191 62.26 82.40 58.20
text-embedding-ada-002 8191 60.99 52.7 55.25
Hosted Inference API

The easiest way to get started with Nomic Embed is through the Nomic Embedding API.

Generating embeddings with the nomic Python client is as easy as

from nomic import embed

output = embed.text(
    texts=['Nomic Embedding API', '#keepAIOpen'],
    model='nomic-embed-text-v1',
    task_type='search_document'
)

print(output)

For more information, see the API reference

Data Visualization

Click the Nomic Atlas map below to visualize a 5M sample of our contrastive pretraining data!

image/webp

Training Details

We train our embedder using a multi-stage training pipeline. Starting from a long-context BERT model , the first unsupervised contrastive stage trains on a dataset generated from weakly related text pairs, such as question-answer pairs from forums like StackExchange and Quora, title-body pairs from Amazon reviews, and summarizations from news articles.

In the second finetuning stage, higher quality labeled datasets such as search queries and answers from web searches are leveraged. Data curation and hard-example mining is crucial in this stage.

For more details, see the Nomic Embed Technical Report and corresponding blog post .

Training data to train the models is released in its entirety. For more details, see the contrastors repository

Usage

Note nomic-embed-text requires prefixes! We support the prefixes [search_query, search_document, classification, clustering] . For retrieval applications, you should prepend search_document for all your documents and search_query for your queries.

Sentence Transformers
from sentence_transformers import SentenceTransformer

model = SentenceTransformer("nomic-ai/nomic-embed-text-v1", trust_remote_code=True)
sentences = ['search_query: What is TSNE?', 'search_query: Who is Laurens van der Maaten?']
embeddings = model.encode(sentences)
print(embeddings)
Transformers
import torch
import torch.nn.functional as F
from transformers import AutoTokenizer, AutoModel

def mean_pooling(model_output, attention_mask):
    token_embeddings = model_output[0]
    input_mask_expanded = attention_mask.unsqueeze(-1).expand(token_embeddings.size()).float()
    return torch.sum(token_embeddings * input_mask_expanded, 1) / torch.clamp(input_mask_expanded.sum(1), min=1e-9)

sentences = ['search_query: What is TSNE?', 'search_query: Who is Laurens van der Maaten?']

tokenizer = AutoTokenizer.from_pretrained('bert-base-uncased')
model = AutoModel.from_pretrained('nomic-ai/nomic-embed-text-v1', trust_remote_code=True)
model.eval()

encoded_input = tokenizer(sentences, padding=True, truncation=True, return_tensors='pt')

with torch.no_grad():
    model_output = model(**encoded_input)

embeddings = mean_pooling(model_output, encoded_input['attention_mask'])
embeddings = F.normalize(embeddings, p=2, dim=1)
print(embeddings)

The model natively supports scaling of the sequence length past 2048 tokens. To do so,

- tokenizer = AutoTokenizer.from_pretrained('bert-base-uncased')
+ tokenizer = AutoTokenizer.from_pretrained('bert-base-uncased', model_max_length=8192)


- model = AutoModel.from_pretrained('nomic-ai/nomic-embed-text-v1', trust_remote_code=True)
+ model = AutoModel.from_pretrained('nomic-ai/nomic-embed-text-v1', trust_remote_code=True, rotary_scaling_factor=2)
Transformers.js
import { pipeline } from '@xenova/transformers';

// Create a feature extraction pipeline
const extractor = await pipeline('feature-extraction', 'nomic-ai/nomic-embed-text-v1', {
    quantized: false, // Comment out this line to use the quantized version
});

// Compute sentence embeddings
const texts = ['What is TSNE?', 'Who is Laurens van der Maaten?'];
const embeddings = await extractor(texts, { pooling: 'mean', normalize: true });
console.log(embeddings);

Join the Nomic Community

Citation

If you find the model, dataset, or training code useful, please cite our work

@misc{nussbaum2024nomic,
      title={Nomic Embed: Training a Reproducible Long Context Text Embedder}, 
      author={Zach Nussbaum and John X. Morris and Brandon Duderstadt and Andriy Mulyar},
      year={2024},
      eprint={2402.01613},
      archivePrefix={arXiv},
      primaryClass={cs.CL}
}

Pricing of nomic-embed-text-v1 replicate.com

Run time and cost

This model costs approximately $0.00025 to run on Replicate, or 4000 runs per $1, but this varies depending on your inputs. It is also open source and you can run it on your own computer with Docker .

This model runs on Nvidia T4 GPU hardware . Predictions typically complete within 2 seconds.

Runs of lucataco nomic-embed-text-v1 on replicate.com

3.6K
Total runs
0
24-hour runs
0
3-day runs
0
7-day runs
0
30-day runs

More Information About nomic-embed-text-v1 replicate.com Model

More nomic-embed-text-v1 license Visit here:

https://huggingface.co/models?license=license%3Aapache-2.0

nomic-embed-text-v1 replicate.com

nomic-embed-text-v1 replicate.com is an AI model on replicate.com that provides nomic-embed-text-v1's model effect (nomic-embed-text-v1 is 8192 context length text encoder that surpasses OpenAI text-embedding-ada-002 and text-embedding-3-small performance on short and long context tasks), which can be used instantly with this lucataco nomic-embed-text-v1 model. replicate.com supports a free trial of the nomic-embed-text-v1 model, and also provides paid use of the nomic-embed-text-v1. Support call nomic-embed-text-v1 model through api, including Node.js, Python, http.

nomic-embed-text-v1 replicate.com Url

https://replicate.com/lucataco/nomic-embed-text-v1

lucataco nomic-embed-text-v1 online free

nomic-embed-text-v1 replicate.com is an online trial and call api platform, which integrates nomic-embed-text-v1's modeling effects, including api services, and provides a free online trial of nomic-embed-text-v1, you can try nomic-embed-text-v1 online for free by clicking the link below.

lucataco nomic-embed-text-v1 online free url in replicate.com:

https://replicate.com/lucataco/nomic-embed-text-v1

nomic-embed-text-v1 install

nomic-embed-text-v1 is an open source model from GitHub that offers a free installation service, and any user can find nomic-embed-text-v1 on GitHub to install. At the same time, replicate.com provides the effect of nomic-embed-text-v1 install, users can directly use nomic-embed-text-v1 installed effect in replicate.com for debugging and trial. It also supports api for free installation.

nomic-embed-text-v1 install url in replicate.com:

https://replicate.com/lucataco/nomic-embed-text-v1

nomic-embed-text-v1 install url in github:

https://github.com/lucataco/cog-nomic-embed-text-v1

Url of nomic-embed-text-v1

nomic-embed-text-v1 replicate.com Url

nomic-embed-text-v1 Owner Github

Provider of nomic-embed-text-v1 replicate.com

Other API from lucataco

replicate

Remove background from an image

Total runs: 5.3M
Run Growth: 100.0K
Growth Rate: 1.89%
Updated: 2023年9月15日
replicate

Falcons.ai Fine-Tuned Vision Transformer (ViT) for NSFW Image Classification

Total runs: 4.5M
Run Growth: 0
Growth Rate: 0.00%
Updated: 2023年11月21日
replicate

Implementation of Realistic Vision v5.1 with VAE

Total runs: 3.7M
Run Growth: 400.0K
Growth Rate: 10.81%
Updated: 2023年8月15日
replicate

FLUX.1-Dev LoRA Explorer

Total runs: 2.9M
Run Growth: 300.0K
Growth Rate: 10.34%
Updated: 2024年10月5日
replicate

SDXL ControlNet - Canny

Total runs: 2.1M
Run Growth: 100.0K
Growth Rate: 4.76%
Updated: 2023年10月4日
replicate

SDXL Inpainting by the HF Diffusers team

Total runs: 1.8M
Run Growth: 200.0K
Growth Rate: 11.11%
Updated: 2024年3月5日
replicate

Turn any image into a video

Total runs: 1.3M
Run Growth: 0
Growth Rate: 0.00%
Updated: 2023年9月2日
replicate

Segmind Stable Diffusion Model (SSD-1B) is a distilled 50% smaller version of SDXL, offering a 60% speedup while maintaining high-quality text-to-image generation capabilities

Total runs: 992.2K
Run Growth: 600
Growth Rate: 0.06%
Updated: 2023年11月8日
replicate

Hyper FLUX 8-step by ByteDance

Total runs: 926.0K
Run Growth: 0
Growth Rate: 0.00%
Updated: 2024年8月28日
replicate

CLIP Interrogator for SDXL optimizes text prompts to match a given image

Total runs: 845.8K
Run Growth: 200
Growth Rate: 0.02%
Updated: 2024年5月16日
replicate

FLUX.1-Dev Multi LoRA Explorer

Total runs: 817.2K
Run Growth: 141.4K
Growth Rate: 17.40%
Updated: 2024年10月6日
replicate

A multimodal LLM-based AI assistant, which is trained with alignment techniques. Qwen-VL-Chat supports more flexible interaction, such as multi-round question answering, and creative capabilities.

Total runs: 796.6K
Run Growth: 1.4K
Growth Rate: 0.18%
Updated: 2023年10月15日
replicate

Robust face restoration algorithm for old photos/AI-generated faces

Total runs: 768.8K
Run Growth: 156.2K
Growth Rate: 20.32%
Updated: 2023年9月6日
replicate

FLUX.1-Schnell LoRA Explorer

Total runs: 693.3K
Run Growth: 95.8K
Growth Rate: 13.90%
Updated: 2024年9月7日
replicate

Coqui XTTS-v2: Multilingual Text To Speech Voice Cloning

Total runs: 644.5K
Run Growth: 48.5K
Growth Rate: 7.53%
Updated: 2023年11月28日
replicate

SDXL v1.0 - A text-to-image generative AI model that creates beautiful images

Total runs: 477.4K
Run Growth: 200
Growth Rate: 0.04%
Updated: 2023年11月1日
replicate

😊 Hotshot-XL is an AI text-to-GIF model trained to work alongside Stable Diffusion XL

Total runs: 464.2K
Run Growth: 39.4K
Growth Rate: 8.49%
Updated: 2023年10月23日
replicate

snowflake-arctic-embed is a suite of text embedding models that focuses on creating high-quality retrieval models optimized for performance

Total runs: 397.2K
Run Growth: 300
Growth Rate: 0.08%
Updated: 2024年4月19日
replicate

Latent Consistency Model (LCM): SDXL, distills the original model into a version that requires fewer steps (4 to 8 instead of the original 25 to 50)

Total runs: 394.2K
Run Growth: 0
Growth Rate: 0.00%
Updated: 2023年11月12日
replicate

Monster Labs QrCode ControlNet on top of SD Realistic Vision v5.1

Total runs: 387.3K
Run Growth: 1.6K
Growth Rate: 0.41%
Updated: 2023年9月24日
replicate

RealvisXL-v2.0 with LCM LoRA - requires fewer steps (4 to 8 instead of the original 40 to 50)

Total runs: 290.9K
Run Growth: 200
Growth Rate: 0.07%
Updated: 2023年11月15日
replicate

moondream2 is a small vision language model designed to run efficiently on edge devices

Total runs: 285.9K
Run Growth: 44.8K
Growth Rate: 15.67%
Updated: 2024年7月29日
replicate

Implementation of SDXL RealVisXL_V2.0

Total runs: 283.6K
Run Growth: 500
Growth Rate: 0.18%
Updated: 2023年11月9日
replicate

Animate Your Personalized Text-to-Image Diffusion Models

Total runs: 281.9K
Run Growth: 2.9K
Growth Rate: 1.03%
Updated: 2023年9月24日
replicate

Practical face restoration algorithm for *old photos* or *AI-generated faces* (for larger images)

Total runs: 234.2K
Run Growth: 9.0K
Growth Rate: 3.84%
Updated: 2023年8月2日
replicate

DreamShaper is a general purpose SD model that aims at doing everything well, photos, art, anime, manga. It's designed to match Midjourney and DALL-E.

Total runs: 194.5K
Run Growth: 3.9K
Growth Rate: 2.01%
Updated: 2023年12月19日
replicate

Real-ESRGAN Video Upscaler

Total runs: 137.7K
Run Growth: 20.5K
Growth Rate: 14.89%
Updated: 2023年11月24日
replicate

A unique fusion that showcases exceptional prompt adherence and semantic understanding, it seems to be a step above base SDXL and a step closer to DALLE-3 in terms of prompt comprehension

Total runs: 124.3K
Run Growth: 600
Growth Rate: 0.48%
Updated: 2023年12月27日
replicate

CLIP Interrogator (for faster inference)

Total runs: 122.0K
Run Growth: 400
Growth Rate: 0.33%
Updated: 2023年9月12日
replicate

dreamshaper-xl-lightning is a Stable Diffusion model that has been fine-tuned on SDXL

Total runs: 107.0K
Run Growth: 6.6K
Growth Rate: 6.17%
Updated: 2024年2月27日
replicate

Phi-3-Mini-4K-Instruct is a 3.8B parameters, lightweight, state-of-the-art open model trained with the Phi-3 datasets

Total runs: 81.7K
Run Growth: 0
Growth Rate: 0.00%
Updated: 2024年7月2日
replicate

SDXL_Niji_Special Edition

Total runs: 64.6K
Run Growth: 2.3K
Growth Rate: 3.56%
Updated: 2023年11月13日
replicate

PixArt-Alpha 1024px is a transformer-based text-to-image diffusion system trained on text embeddings from T5

Total runs: 64.0K
Run Growth: 11.4K
Growth Rate: 17.81%
Updated: 2023年12月4日
replicate

MagicAnimate: Temporally Consistent Human Image Animation using Diffusion Model

Total runs: 55.5K
Run Growth: 300
Growth Rate: 0.54%
Updated: 2023年12月5日
replicate

Dreamshaper-7 img2img with LCM LoRA for faster inference

Total runs: 55.1K
Run Growth: 0
Growth Rate: 0.00%
Updated: 2023年11月17日
replicate

AI-driven audio enhancement for your audio files, powered by Resemble AI

Total runs: 52.3K
Run Growth: 9.5K
Growth Rate: 18.16%
Updated: 2023年12月15日
replicate

Ostris AI-Toolkit for Flux LoRA Training (DEPRECATED. Please use: ostris/flux-dev-lora-trainer)

Total runs: 50.9K
Run Growth: 3.1K
Growth Rate: 6.10%
Updated: 2024年8月18日
replicate

Florence-2: Advancing a Unified Representation for a Variety of Vision Tasks

Total runs: 49.5K
Run Growth: 3.7K
Growth Rate: 7.47%
Updated: 2024年6月25日
replicate

Implementation of SDXL RealVisXL_V1.0

Total runs: 44.0K
Run Growth: 0
Growth Rate: 0.00%
Updated: 2023年9月13日
replicate

SDXL Image Blending

Total runs: 42.4K
Run Growth: 0
Growth Rate: 0.00%
Updated: 2023年12月11日
replicate

(Academic and Non-commercial use only) Pixel-Aware Stable Diffusion for Realistic Image Super-resolution and Personalized Stylization

Total runs: 39.5K
Run Growth: 300
Growth Rate: 0.76%
Updated: 2024年1月8日
replicate

BakLLaVA-1 is a Mistral 7B base augmented with the LLaVA 1.5 architecture

Total runs: 39.0K
Run Growth: 0
Growth Rate: 0.00%
Updated: 2023年10月24日
replicate

lmsys/vicuna-13b-v1.3

Total runs: 38.4K
Run Growth: 0
Growth Rate: 0.00%
Updated: 2023年6月29日
replicate

Mistral-7B-v0.1 fine tuned for chat with the Dolphin dataset (an open-source implementation of Microsoft's Orca)

Total runs: 35.0K
Run Growth: 400
Growth Rate: 1.14%
Updated: 2023年10月31日
replicate

Real-ESRGAN with optional face correction and adjustable upscale (for larger images)

Total runs: 34.3K
Run Growth: 200
Growth Rate: 0.58%
Updated: 2023年7月17日
replicate

Gemma2 2b by Google

Total runs: 33.1K
Run Growth: 0
Growth Rate: 0.00%
Updated: 2024年8月1日
replicate

The image prompt adapter is designed to enable a pretrained text-to-image diffusion model to generate SDXL images with an image prompt

Total runs: 31.1K
Run Growth: 200
Growth Rate: 0.64%
Updated: 2023年11月11日
replicate

Florence-2: Advancing a Unified Representation for a Variety of Vision Tasks

Total runs: 31.1K
Run Growth: 1000
Growth Rate: 3.22%
Updated: 2024年6月25日
replicate

Realistic Vision v5.0 with VAE

Total runs: 29.9K
Run Growth: 1.5K
Growth Rate: 5.02%
Updated: 2023年8月18日
replicate

lmsys/vicuna-7b-v1.3

Total runs: 28.6K
Run Growth: 100
Growth Rate: 0.35%
Updated: 2023年6月29日
replicate

(Research only) IP-Adapter-FaceID can generate various style images conditioned on a face with only text prompts

Total runs: 28.3K
Run Growth: 0
Growth Rate: 0.00%
Updated: 2023年12月20日
replicate

Meta's Llama 2 7b Chat - GPTQ

Total runs: 20.3K
Run Growth: 100
Growth Rate: 0.49%
Updated: 2023年7月24日
replicate

sdxs-512-0.9 can generate high-resolution images in real-time based on prompt texts, trained using score distillation and feature matching

Total runs: 18.8K
Run Growth: 0
Growth Rate: 0.00%
Updated: 2024年3月28日
replicate

Stylized Audio-Driven Single Image Talking Face Animation

Total runs: 18.6K
Run Growth: 100
Growth Rate: 0.54%
Updated: 2023年10月8日
replicate

Meta's Llama 2 13b Chat - GPTQ

Total runs: 18.5K
Run Growth: 0
Growth Rate: 0.00%
Updated: 2023年7月24日
replicate

WizardCoder: Empowering Code Large Language Models with Evol-Instruct

Total runs: 17.0K
Run Growth: 0
Growth Rate: 0.00%
Updated: 2024年1月23日
replicate

ThinkDiffusionXL is a go-to model capable of amazing photorealism that's also versatile enough to generate high-quality images across a variety of styles and subjects without needing to be a prompting genius

Total runs: 15.5K
Run Growth: 100
Growth Rate: 0.65%
Updated: 2023年11月6日
replicate

This is wizard-vicuna-13b trained with a subset of the dataset - responses that contained alignment / moralizing were removed

Total runs: 15.1K
Run Growth: 0
Growth Rate: 0.00%
Updated: 2024年4月25日
replicate

Hyper FLUX 16-step by ByteDance

Total runs: 15.0K
Run Growth: 0
Growth Rate: 0.00%
Updated: 2024年8月28日
replicate

Mistral-7B-v0.1 fine tuned for chat with the Dolphin dataset (an open-source implementation of Microsoft's Orca)

Total runs: 13.4K
Run Growth: 0
Growth Rate: 0.00%
Updated: 2023年10月31日
replicate

Image-to-video - SEINE: Short-to-Long Video Diffusion Model for Generative Transition and Prediction

Total runs: 12.0K
Run Growth: 600
Growth Rate: 5.00%
Updated: 2023年11月23日
replicate

InterpAny-Clearer: Clearer anytime frame interpolation & Manipulated interpolation

Total runs: 11.4K
Run Growth: 0
Growth Rate: 0.00%
Updated: 2023年11月29日
replicate

Segments an audio recording based on who is speaking (on A100)

Total runs: 11.4K
Run Growth: 0
Growth Rate: 0.00%
Updated: 2023年7月21日
replicate

Latest model in the Qwen family for chatting with video and image models

Total runs: 11.3K
Run Growth: 0
Growth Rate: 0.00%
Updated: 2024年12月20日
replicate

(Research only) Moondream1 is a vision language model that performs on par with models twice its size

Total runs: 10.7K
Run Growth: 0
Growth Rate: 0.00%
Updated: 2024年1月24日
replicate

Image to Image enhancer using DemoFusion

Total runs: 10.4K
Run Growth: 100
Growth Rate: 0.96%
Updated: 2023年12月8日
replicate

Open diffusion model for high-quality video generation

Total runs: 10.3K
Run Growth: 100
Growth Rate: 0.97%
Updated: 2023年10月19日
replicate

Auto fuse a user's face onto the template image, with a similar appearance to the user

Total runs: 10.1K
Run Growth: 200
Growth Rate: 1.98%
Updated: 2023年11月15日
replicate

DemoFusion: Democratising High-Resolution Image Generation With No 💰

Total runs: 9.0K
Run Growth: 0
Growth Rate: 0.00%
Updated: 2023年12月3日
replicate

Segment Anything 2 (SAM2) by Meta - Automatic mask generation

Total runs: 8.9K
Run Growth: 1.2K
Growth Rate: 13.48%
Updated: 2024年7月31日
replicate

Implementation of SDXL RealVisXL_V2.0 img2img

Total runs: 8.6K
Run Growth: 0
Growth Rate: 0.00%
Updated: 2023年11月7日
replicate

Phi-3-Mini-128K-Instruct is a 3.8 billion-parameter, lightweight, state-of-the-art open model trained using the Phi-3 datasets

Total runs: 8.0K
Run Growth: 0
Growth Rate: 0.00%
Updated: 2024年4月25日
replicate

Cog wrapper for Ollama llama3:70b

Total runs: 6.6K
Run Growth: 6.5K
Growth Rate: 98.42%
Updated: 2024年7月9日
replicate

360 Panorama SDXL image with inpainted wrapping seam

Total runs: 6.3K
Run Growth: 100
Growth Rate: 1.59%
Updated: 2023年9月9日
replicate

Convert your videos to DensePose and use it with MagicAnimate

Total runs: 5.7K
Run Growth: 0
Growth Rate: 0.00%
Updated: 2023年12月6日
replicate

Projection module trained to add vision capabilties to Llama 3 using SigLIP

Total runs: 5.5K
Run Growth: 100
Growth Rate: 1.82%
Updated: 2024年11月5日
replicate

Fuyu-8B is a multi-modal text and image transformer trained by Adept AI

Total runs: 4.6K
Run Growth: 0
Growth Rate: 0.00%
Updated: 2023年10月20日
replicate

Qwen1.5 is the beta version of Qwen2, a transformer-based decoder-only language model pretrained on a large amount of data

Total runs: 4.1K
Run Growth: 100
Growth Rate: 2.44%
Updated: 2024年2月6日
replicate

Controlnet v1.1 - Tile Version

Total runs: 4.0K
Run Growth: 0
Growth Rate: 0.00%
Updated: 2023年11月27日
replicate

SDXL using DeepCache

Total runs: 3.8K
Run Growth: 0
Growth Rate: 0.00%
Updated: 2024年1月7日
replicate

Playground v2 is a diffusion-based text-to-image generative model trained from scratch. Try out all 3 models here

Total runs: 3.6K
Run Growth: 0
Growth Rate: 0.00%
Updated: 2023年12月7日
replicate

Segmind Stable Diffusion Model (SSD-1B) img2img

Total runs: 3.6K
Run Growth: 0
Growth Rate: 0.00%
Updated: 2023年11月3日
replicate

A combination of ip_adapter SDv1.5 and mediapipe-face to inpaint a face

Total runs: 3.6K
Run Growth: 200
Growth Rate: 5.56%
Updated: 2023年11月15日
replicate

Phi-2 by Microsoft

Total runs: 3.5K
Run Growth: 200
Growth Rate: 5.71%
Updated: 2024年1月30日
replicate

Implementation of SDXL RealVisXL_V1.0 img2img

Total runs: 3.4K
Run Growth: 0
Growth Rate: 0.00%
Updated: 2023年11月1日
replicate

A Flux LoRA trained on watercolor style photos

Total runs: 3.2K
Run Growth: 600
Growth Rate: 18.75%
Updated: 2024年8月15日