nvidia/Nemotron-3-Embed-1B-BF16 · Hugging Face

Source: https://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16

https://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16#model-overviewModel Overview

https://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16#descriptionDescription:

Nemotron-3-Embed-1B-BF16is a versatile text embedding model trained by NVIDIA and optimized for retrieval and semantic similarity tasks. It provides strong multilingual and cross-lingual retrieval capabilities and is designed to serve as a foundational component in text-based Retrieval-Augmented Generation (RAG) systems. This model was evaluated across 34 languages: English, Arabic, Assamese, Bengali, Bulgarian, Chinese, Danish, Dutch, Finnish, French, German, Hindi, Hinglish, Indonesian, Italian, Japanese, Korean, Malay, Marathi, Nepali, Norwegian, Persian, Portuguese, Romanian, Russian, Spanish, Swahili, Swedish, Tamil, Telugu, Thai, Ukrainian, Urdu, Vietnamese.

The model generates dense vector embeddings from multilingual text inputs, enabling retrieval, semantic search, and (agentic) RAG workflows. As a core component of text retrieval systems, an embedding model transforms text, such as questions or passages, into dense vector representations. These models are typically transformer encoders that process input tokens and produce embeddings suitable for efficient similarity matching.

Among models of comparable size,Nemotron-3-Embed-1B-BF16achieves state-of-the-art performance across multiple multilingual retrieval benchmarks. Read more details in ourBlog Post.

This model is ready for commercial use.

https://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16#licenseterms-of-useLicense/Terms of Use:

This model and its associated configuration files are licensed under theOpenMDW License Agreement, version 1.1 (OpenMDW-1.1). Additional Information: Built withMinistral-3-3B-Instruct-2512which is released under Apache 2.0.

This project will download and install additional third-party open source software projects. Review the license terms of these open source projects before use.

https://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16#deployment-geographyDeployment Geography:

Global

https://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16#use-case-Use Case:

Nemotron-3-Embed-1B-BF16is most suitable for users who want to build a multilingual question-and-answer application over a large text corpus, leveraging the latest dense retrieval technologies.

https://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16#release-date-Release Date:

07/16/2026 viahttps://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16

https://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16#model-architectureModel Architecture:

**Architecture Type:**Transformer

Network Architecture:Ministral-3-3B-Instruct-2512based pruned model

**Number of model parameters:**The model has approximately 1.14B parameters

**Hidden Size:**2048

TheNemotron-3-Embed-1B-BF16model is a transformer-based text embedding model trained with bidirectional attention masking, where the final embedding vector is obtained by applying average pooling to the transformer’s token-level representations. It encodes each input text into a dense embedding vector of dimension 2048.

Nemotron-3-Embed-1B-BF16was derived from theMinistral-3-3B-Instruct-2512through two iterative rounds of structured pruning and distillation. First, the 3B parent model was trained as a text-embedding model. Then it was pruned to 2B using NVIDIA ModelOpt mcore_minitron Neural Architecture Search (NAS) [NVIDIA/Model-Optimizer,paper]. This process searches across hidden width, FFN size, attention heads, and depth, then selects the best candidate from the top-10 Pareto front. Candidates were evaluated against the parent model’s representations using a 50k in-domain calibration corpus, which was also used to estimate importance scores.

The resulting 2B model was then distilled from the fine-tunedNemotron-3-Embed-8B-BF16embedding teacher model to recover accuracy. Distillation used combined cosine distance loss (COS) and mean squared error (MSE) loss on a multilingual, in-domain retrieval data blend. The same pruning-and-distillation procedure, using the same dataset blend, was then repeated to produce the final 1.14B embedding model.

https://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16#inputs-Input(s):

**Input Type(s):**Text

Input Format(s):

• Text: List of strings

Input Parameters:

• Text: One-Dimensional (1D)

**Other Properties Related to Input:**Text inputs should be tokenized by the model tokenizer. The model’s max sequence length is 32768. Longer inputs should be chunked or truncated.

https://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16#outputsOutput(s)

**Output Type(s):**Floats

Output Format(s):

• List of float arrays

**Output Parameters:**One-Dimensional (1D) embedding vector per input text string

**Other Properties Related to Output:**The model outputs a 2048-dimensional embedding vector for each input text string. It also supports dynamic embedding sizes by slicing the vector from the start (for example, keeping the first 1024 or 512 dimensions). These sliced embeddings remain highly functional, provided the resulting sub-vector is re-normalized (L2 normalization) after slicing.

Our AI models are designed and/or optimized to run on NVIDIA GPU-accelerated systems. By leveraging NVIDIA’s hardware (e.g. GPU cores) and software frameworks (e.g., CUDA libraries), the model achieves faster training and inference times compared to CPU-only solutions.

https://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16#usageUsage

For local Python examples, use the Hugging Face model ID by default. If you are working from a local checkout, replaceMODEL\_IDwith that path. For vLLM online serving from a local checkpoint, use the local checkpoint example.

MODEL_ID = "nvidia/Nemotron-3-Embed-1B-BF16"

Use this model for retrieval-style embeddings. Add thequery:prefix for queries and thepassage:prefix for documents. Embeddings are L2-normalized, so dot product and cosine similarity are equivalent. The output tables useq\[i\]for queries andd\[i\]for documents. Scores are rounded to four decimals. Exact values can vary by runtime and package version.

https://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16#local-python-dependenciesLocal Python Dependencies

The BF16 checkpoints support Transformers5\.2\.0and above. The examples also require a CUDA-enabled PyTorch installation that matches your driver and CUDA environment.

If you are not using an NVIDIA PyTorch container, install PyTorch first. Use thePyTorch local installation selectorto choose the command that matches your operating system, package manager, and CUDA environment. If the default PyPI wheel matches your CUDA environment, run:

pip install --upgrade torch

For non-default CUDA environments, the selector can include an additional\-\-index\-urlargument. Use that CUDA-specific index URL when the default PyPI wheel does not match your CUDA environment.

Then install Transformers and Sentence Transformers:

pip install --upgrade "transformers>=5.2.0" "sentence-transformers>=5.4.1"

NVIDIA tested the examples innvcr\.io/nvidia/pytorch:26\.06\-py3with the container-provided Torch and CUDA stack. Inside NVIDIA PyTorch containers, do not upgrade Torch. Install only the missing packages:

pip install --upgrade "transformers>=5.2.0" "sentence-transformers>=5.4.1"

The testednvcr\.io/nvidia/pytorch:26\.06\-py3container includesflash\-attn, so the snippets use FlashAttention-2 by default. If your environment does not have FlashAttention-2, setattn\_implementationorATTN\_IMPLEMENTATIONtosdpa.

https://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16#sentence-transformersSentence Transformers

Use Sentence Transformers for the simplest local Python interface. It reads the saved query and document prompts and normalization metadata.

import torch
from sentence_transformers import SentenceTransformer

MODEL_ID = "nvidia/Nemotron-3-Embed-1B-BF16"

model = SentenceTransformer(
    MODEL_ID,
    device="cuda",
    model_kwargs={
        "dtype": torch.bfloat16,
        "attn_implementation": "flash_attention_2",
    },
)
model.max_seq_length = 32768

QUERIES = [
    "Write a Python function that counts the frequency of each element in a list of lists.",
    "Write a function that orders a dictionary with tuple keys by the product of each key's tuple values.",
    "What symptoms and common triggers help distinguish eczema from other inflammatory skin conditions?",
    "How can someone reduce exposure to pollen during allergy season?",
]

DOCUMENTS = [
    "def frequency_lists(list1):\n    flattened = [item for sublist in list1 for item in sublist]\n    counts = {}\n    for item in flattened:\n        if item in counts:\n            counts[item] += 1\n        else:\n            counts[item] = 1\n    return counts",
    "def sort_dict_item(test_dict):\n    return {key: test_dict[key] for key in sorted(test_dict.keys(), key=lambda ele: ele[0] * ele[1])}",
    "Eczema commonly causes itchy, dry, inflamed patches of skin. The affected areas may look red, scaly, cracked, or darker than the surrounding skin depending on skin tone. Symptoms can flare after exposure to irritants, allergens, stress, or changes in weather.",
    "People with pollen allergy can reduce exposure by staying indoors on dry, windy days, avoiding early-morning outdoor activity, and going outside after rain when pollen levels are lower. They should check pollen forecasts, close windows and doors when counts are high, and consider starting allergy medication before symptoms begin if high pollen is expected. After being outside, showering, changing clothes, avoiding outdoor laundry drying, and wearing a face mask for yard work can help limit pollen contact.",
]
query_embeddings = model.encode_query(QUERIES, batch_size=8, convert_to_tensor=True)
document_embeddings = model.encode_document(DOCUMENTS, batch_size=8, convert_to_tensor=True)

scores = model.similarity(query_embeddings, document_embeddings)
print("Similarity scores:")
print(f"{'':>4}" + "".join(f"d[{i}]".rjust(10) for i in range(scores.shape[1])))
for query_index, row in enumerate(scores):
    print(f"q[{query_index}]" + "".join(f"{score.item():>10.4f}" for score in row))

Sentence Transformers Expected Output``` Similarity scores: d[0] d[1] d[2] d[3] q[0] 0.8125 0.0255 0.0005 -0.0312 q[1] 0.0447 0.6484 -0.0520 0.0386 q[2] -0.0095 -0.0410 0.6484 0.1006 q[3] -0.0216 0.0214 0.1211 0.7734

### [https://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16#transformers](https://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16#transformers)Transformers

Use Transformers when you want manual control over tokenization, pooling, or batching\.

import torch import torch.nn.functional as F from transformers import AutoModel, AutoTokenizer

MODEL_ID = “nvidia/Nemotron-3-Embed-1B-BF16” MAX_LENGTH = 32768 BATCH_SIZE = 8 DTYPE = torch.bfloat16 ATTN_IMPLEMENTATION = “flash_attention_2”

QUERIES = [ “Write a Python function that counts the frequency of each element in a list of lists.”, “Write a function that orders a dictionary with tuple keys by the product of each key’s tuple values.”, “What symptoms and common triggers help distinguish eczema from other inflammatory skin conditions?”, “How can someone reduce exposure to pollen during allergy season?”, ]

DOCUMENTS = [ “def frequency_lists(list1):\n flattened = [item for sublist in list1 for item in sublist]\n counts = {}\n for item in flattened:\n if item in counts:\n counts[item] += 1\n else:\n counts[item] = 1\n return counts”, “def sort_dict_item(test_dict):\n return {key: test_dict[key] for key in sorted(test_dict.keys(), key=lambda ele: ele[0] * ele[1])}”, “Eczema commonly causes itchy, dry, inflamed patches of skin. The affected areas may look red, scaly, cracked, or darker than the surrounding skin depending on skin tone. Symptoms can flare after exposure to irritants, allergens, stress, or changes in weather.”, “People with pollen allergy can reduce exposure by staying indoors on dry, windy days, avoiding early-morning outdoor activity, and going outside after rain when pollen levels are lower. They should check pollen forecasts, close windows and doors when counts are high, and consider starting allergy medication before symptoms begin if high pollen is expected. After being outside, showering, changing clothes, avoiding outdoor laundry drying, and wearing a face mask for yard work can help limit pollen contact.”, ]

def average_pool(last_hidden_state: torch.Tensor, attention_mask: torch.Tensor) -> torch.Tensor: last_hidden = last_hidden_state.masked_fill(~attention_mask[…, None].bool(), 0.0) return last_hidden.sum(dim=1) / attention_mask.sum(dim=1)[…, None]

if not torch.cuda.is_available(): raise RuntimeError(“CUDA is required for practical BF16 inference.”)

device = torch.device(“cuda”)

tokenizer = AutoTokenizer.from_pretrained(MODEL_ID, padding_side=“left”) if tokenizer.pad_token is None: tokenizer.pad_token = tokenizer.eos_token

model = AutoModel.from_pretrained( MODEL_ID, dtype=DTYPE, attn_implementation=ATTN_IMPLEMENTATION, ).to(device) model.eval()

def encode_texts(texts: list[str]) -> torch.Tensor: embedding_batches = []

for start in range(0, len(texts), BATCH_SIZE):
    encoded = tokenizer(
        texts[start : start + BATCH_SIZE],
        max_length=MAX_LENGTH,
        truncation=True,
        padding=True,
        return_tensors="pt",
    )
    encoded = {name: tensor.to(device) for name, tensor in encoded.items()}

    with torch.inference_mode():
        output = model(**encoded)
        pooled = average_pool(output.last_hidden_state, encoded["attention_mask"])
        embeddings = F.normalize(pooled, p=2, dim=-1)

    embedding_batches.append(embeddings.detach().cpu().to(torch.float32))

return torch.cat(embedding_batches, dim=0)

embeddings = encode_texts( [“query: “ + query for query in QUERIES] + [“passage: “ + doc for doc in DOCUMENTS] ) query_embeddings = embeddings[: len(QUERIES)] document_embeddings = embeddings[len(QUERIES) :]

scores = query_embeddings @ document_embeddings.T print(“Similarity scores:”) print(f“{‘’:>4}“ + “”.join(f“d[{i}]“.rjust(10) for i in range(scores.shape[1]))) for query_index, row in enumerate(scores): print(f“q[{query_index}]” + “”.join(f“{score.item():>10.4f}“ for score in row))

Transformers Expected Output```
Similarity scores:
          d[0]      d[1]      d[2]      d[3]
q[0]    0.8069    0.0252    0.0001   -0.0312
q[1]    0.0446    0.6466   -0.0514    0.0385
q[2]   -0.0098   -0.0410    0.6450    0.0998
q[3]   -0.0215    0.0212    0.1197    0.7679

https://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16#vllm-dependenciesvLLM Dependencies

For BF16, usevllm==0\.25\.0for/v2/embedserving. NVIDIA also validatedvllm serve "$MODEL\_ID"withvllm/vllm\-openai:v0\.20\.0throughv0\.24\.0for this checkpoint.

pip install --upgrade "vllm==0.25.0" openai requests numpy

https://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16#vllm-offline-pythonvLLM Offline Python

Use the offline Python API when you want local vLLM inference without running an HTTP server.LLM\.embedaccepts formatted strings, so add thequery:andpassage:prefixes manually.

vLLM Offline Python Example``` import numpy as np from vllm import LLM

MODEL_ID = “nvidia/Nemotron-3-Embed-1B-BF16”

QUERIES = [ “Write a Python function that counts the frequency of each element in a list of lists.”, “Write a function that orders a dictionary with tuple keys by the product of each key’s tuple values.”, “What symptoms and common triggers help distinguish eczema from other inflammatory skin conditions?”, “How can someone reduce exposure to pollen during allergy season?”, ]

DOCUMENTS = [ “def frequency_lists(list1):\n flattened = [item for sublist in list1 for item in sublist]\n counts = {}\n for item in flattened:\n if item in counts:\n counts[item] += 1\n else:\n counts[item] = 1\n return counts”, “def sort_dict_item(test_dict):\n return {key: test_dict[key] for key in sorted(test_dict.keys(), key=lambda ele: ele[0] * ele[1])}”, “Eczema commonly causes itchy, dry, inflamed patches of skin. The affected areas may look red, scaly, cracked, or darker than the surrounding skin depending on skin tone. Symptoms can flare after exposure to irritants, allergens, stress, or changes in weather.”, “People with pollen allergy can reduce exposure by staying indoors on dry, windy days, avoiding early-morning outdoor activity, and going outside after rain when pollen levels are lower. They should check pollen forecasts, close windows and doors when counts are high, and consider starting allergy medication before symptoms begin if high pollen is expected. After being outside, showering, changing clothes, avoiding outdoor laundry drying, and wearing a face mask for yard work can help limit pollen contact.”, ]

def main(): llm = LLM(model=MODEL_ID) texts = [“query: “ + query for query in QUERIES] + [ “passage: “ + doc for doc in DOCUMENTS ] outputs = llm.embed(texts, use_tqdm=False) embeddings = np.array( [output.outputs.embedding for output in outputs], dtype=np.float32, )

query_embeddings = embeddings[: len(QUERIES)]
document_embeddings = embeddings[len(QUERIES) :]

scores = query_embeddings @ document_embeddings.T
print("Similarity scores:")
print(f"{'':>4}" + "".join(f"d[{i}]".rjust(10) for i in range(scores.shape[1])))
for query_index, row in enumerate(scores):
    print(f"q[{query_index}]" + "".join(f"{score:>10.4f}" for score in row))

if name == “main”: main()

vLLM Offline Python Expected Output```
Similarity scores:
          d[0]      d[1]      d[2]      d[3]
q[0]    0.8110    0.0255    0.0003   -0.0312
q[1]    0.0448    0.6469   -0.0517    0.0386
q[2]   -0.0100   -0.0400    0.6469    0.1003
q[3]   -0.0224    0.0217    0.1199    0.7692

https://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16#vllm-online-servingvLLM Online Serving

MODEL_ID=nvidia/Nemotron-3-Embed-1B-BF16

vllm serve "$MODEL_ID"

vLLM defaults to port8000. Add host and port when you need an explicit bind address or a non-default port:

vllm serve "$MODEL_ID" --host 0.0.0.0 --port 8000

If you serve the Hugging Face model ID, the served model name defaults tonvidia/Nemotron\-3\-Embed\-1B\-BF16. If you serve a local checkpoint path, vLLM still reads the model config and weights from that path;\-\-served\-model\-nameonly sets the model name accepted by API requests.

MODEL_PATH=/path/to/local/Nemotron-3-Embed-1B-BF16
vllm serve "$MODEL_PATH" --host 0.0.0.0 --port 8000 --served-model-name nvidia/Nemotron-3-Embed-1B-BF16

With\-\-served\-model\-name, client requests continue to useMODEL = "nvidia/Nemotron\-3\-Embed\-1B\-BF16". If you omit it, use the served name reported by/v1/modelsin client requests.

https://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16#recommended-retrieval-endpointRecommended Retrieval Endpoint

After the server is running, use/v2/embedfor retrieval. Send raw query and document strings.input\_typeapplies the saved query and document prompts:

import numpy as np
import requests

MODEL = "nvidia/Nemotron-3-Embed-1B-BF16"
URL = "http://localhost:8000/v2/embed"

QUERIES = [
    "Write a Python function that counts the frequency of each element in a list of lists.",
    "Write a function that orders a dictionary with tuple keys by the product of each key's tuple values.",
    "What symptoms and common triggers help distinguish eczema from other inflammatory skin conditions?",
    "How can someone reduce exposure to pollen during allergy season?",
]

DOCUMENTS = [
    "def frequency_lists(list1):\n    flattened = [item for sublist in list1 for item in sublist]\n    counts = {}\n    for item in flattened:\n        if item in counts:\n            counts[item] += 1\n        else:\n            counts[item] = 1\n    return counts",
    "def sort_dict_item(test_dict):\n    return {key: test_dict[key] for key in sorted(test_dict.keys(), key=lambda ele: ele[0] * ele[1])}",
    "Eczema commonly causes itchy, dry, inflamed patches of skin. The affected areas may look red, scaly, cracked, or darker than the surrounding skin depending on skin tone. Symptoms can flare after exposure to irritants, allergens, stress, or changes in weather.",
    "People with pollen allergy can reduce exposure by staying indoors on dry, windy days, avoiding early-morning outdoor activity, and going outside after rain when pollen levels are lower. They should check pollen forecasts, close windows and doors when counts are high, and consider starting allergy medication before symptoms begin if high pollen is expected. After being outside, showering, changing clothes, avoiding outdoor laundry drying, and wearing a face mask for yard work can help limit pollen contact.",
]

def embed(input_type: str, texts: list[str]) -> np.ndarray:
    response = requests.post(
        URL,
        json={
            "model": MODEL,
            "input_type": input_type,
            "texts": texts,
            "embedding_types": ["float"],
            "truncate": "END",
        },
        timeout=120,
    )
    response.raise_for_status()
    return np.array(response.json()["embeddings"]["float"], dtype=np.float32)

query_embeddings = embed("query", QUERIES)
document_embeddings = embed("document", DOCUMENTS)

scores = query_embeddings @ document_embeddings.T
print("Similarity scores:")
print(f"{'':>4}" + "".join(f"d[{i}]".rjust(10) for i in range(scores.shape[1])))
for query_index, row in enumerate(scores):
    print(f"q[{query_index}]" + "".join(f"{score:>10.4f}" for score in row))

Recommended Retrieval Endpoint Expected Output``` Similarity scores: d[0] d[1] d[2] d[3] q[0] 0.8109 0.0252 0.0001 -0.0314 q[1] 0.0448 0.6470 -0.0516 0.0387 q[2] -0.0103 -0.0400 0.6470 0.1000 q[3] -0.0223 0.0217 0.1199 0.7693

You can also use the OpenAI\-compatible`/v1/embeddings`endpoint\. For those requests, pass strings in`input`and manually prefix them with`query:`or`passage:`\.

### [https://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16#expected-configuration-warning](https://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16#expected-configuration-warning)Expected Configuration Warning

When this checkpoint is loaded by Transformers, whether directly or through Sentence Transformers or vLLM, the following warning may appear:

[transformers] Unrecognized keys in rope_parameters for ‘rope_type’=‘yarn’: {‘apply_yarn_scaling’}

This warning is expected and does not prevent model loading or inference\.`apply\_yarn\_scaling`is retained as a temporary vLLM compatibility field that preserves the checkpoint's intended long\-context RoPE behavior\. Do not remove it from`config\.json`\. The upstream compatibility work is tracked in[vLLM issue \#48621](https://github.com/vllm-project/vllm/issues/48621)\.

## [https://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16#software-integration](https://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16#software-integration)Software Integration:

**Runtime Engine\(s\):**PyTorch, vLLM

**Supported Hardware Microarchitecture Compatibility:**

- NVIDIA Ampere
- NVIDIA Blackwell
- NVIDIA Hopper

**Supported Operating System\(s\):**

- Linux

The integration of foundation and fine\-tuned models into AI systems requires additional testing using use\-case\-specific data to ensure safe and effective deployment\. Following the V\-model methodology, iterative testing and validation at both unit and system levels are essential to mitigate risks, meet technical and functional requirements, and ensure compliance with safety and ethical standards before deployment\.

## [https://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16#model-versions](https://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16#model-versions)Model Version\(s\):

**Nemotron\-3\-Embed\-1B\-BF16**

## [https://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16#training-testing-and-evaluation-datasets](https://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16#training-testing-and-evaluation-datasets)Training, Testing, and Evaluation Datasets:

### [https://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16#dataset-overview](https://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16#dataset-overview)Dataset Overview:

**Total Size:**8\.5M\+ data points **Total Number of Datasets:**161 dataset files **Dataset Partition:**Training \[100%\], Testing \[N/A — evaluation benchmarks used separately\], Validation \[N/A — evaluation benchmarks used separately\]\.

Model distillation training was conducted using publicly available, commercially permissible datasets and synthetically generated datasets\. Synthetic data was created either by generating queries from seed documents or by generating complete question–answer pairs through LLM\-based prompting using the models listed below\.

### [https://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16#public-datasets](https://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16#public-datasets)Public Datasets:

### [https://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16#synthetic-datasets](https://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16#synthetic-datasets)Synthetic Datasets:

Synthetic query\-document pairs were generated either from scratch or by using seed datasets to generate queries with the models listed below\.

LLMs used to generate synthetic datasetsQwen/Qwen3\-Next\-80B\-A3B\-Instruct
Qwen/Qwen3\-235B\-A22B
Qwen/Qwen3\.5\-397B\-A17B
Qwen/Qwen3\.6\-27B
Qwen/Qwen3\.6\-35B\-A3Bgoogle/gemma\-4\-31B\-itopenai/gpt\-oss\-120b
openai/gpt\-oss\-20bnvidia/NVIDIA\-Nemotron\-3\-Super\-120B\-A12B\-BF16
nvidia/NVIDIA\-Nemotron\-3\-Ultra\-550B\-A55B\-NVFP4Seed DatasetsDatasetReferenceFinePdfs[https://huggingface\.co/datasets/HuggingFaceFW/finepdfs](https://huggingface.co/datasets/HuggingFaceFW/finepdfs)CentralActs[https://zenodo\.org/records/5088102](https://zenodo.org/records/5088102)BRIGHT[https://huggingface\.co/datasets/xlangai/BRIGHT](https://huggingface.co/datasets/xlangai/BRIGHT)MultiHiertt[https://github\.com/psunlpgroup/MultiHiertt](https://github.com/psunlpgroup/MultiHiertt)### [https://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16#training-dataset](https://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16#training-dataset)Training Dataset:

#### [https://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16#data-modality](https://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16#data-modality)Data Modality:

- Text

#### [https://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16#training-data-size](https://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16#training-data-size)Training Data Size:

**Text Training Data Size:**8\.5M\+

**Data Collection Method by dataset:**Hybrid: Human, Automated, Synthetic

**Labeling Method by dataset:**Hybrid: Human, Automated, Synthetic

**Properties:**Model training was conducted on text datasets using question–passage pairs from publicly available, commercially permissible datasets and synthetically generated datasets\.

### [https://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16#testing-dataset](https://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16#testing-dataset)Testing Dataset:

**Data Collection Method by dataset:**Not Applicable

**Labeling Method by dataset:**Not Applicable

**Properties:**Not Applicable\. Model quality was assessed using the evaluation benchmark datasets described in the Evaluation Dataset subsection\.

### [https://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16#evaluation-dataset](https://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16#evaluation-dataset)Evaluation Dataset:

**Data Collection Method by dataset:**Hybrid: Human, Automated, Synthetic

**Labeling Method by dataset:**Hybrid: Human, Automated, Synthetic

**Properties:**This model is evaluated on 16 public tasks on[Retrieval Embedding Benchmark \(RTEB\)](https://huggingface.co/blog/rteb), a new benchmark designed to reliably evaluate the retrieval accuracy of embedding models for real\-world applications\. More details on RTEB can be found on their[leaderboard](https://huggingface.co/spaces/mteb/leaderboard?benchmark_name=RTEB%28beta%29)\.

We also evaluated the model on the[MMTEB \(Retrieval\) benchmark datasets](https://huggingface.co/spaces/mteb/leaderboard)\([paper](https://arxiv.org/pdf/2502.13595)\), and on the eight text datasets \(extracted via OCR\) from the[ViDoRe\-V3 benchmark](https://mteb-leaderboard.hf.space/benchmark/ViDoRe(v3))\.

We set the model sequence length to 4096 for the evaluation results below\.

Text Retrieval benchmarks \(chunk retrieval\) \- Avg\. NDCG@10Model NameRTEBViDoRe\-V3 textMMTEB \(Retrieval\)llama\-nemotron\-embed\-vl\-1b\-v261\.9852\.5459\.71Nemotron\-3\-Embed\-1B\-BF1672\.3857\.7471\.04## [https://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16#inference](https://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16#inference)Inference:

**Acceleration Engine:**PyTorch, vLLM **Test Hardware:**NVIDIA Ampere \- A100 80GB NVIDIA Hopper \- H100 80GB

## [https://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16#ethical-considerations](https://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16#ethical-considerations)Ethical Considerations:

NVIDIA believes Trustworthy AI is a shared responsibility and we have established policies and practices to enable development for a wide array of AI applications\. Developers should work with their internal model team to ensure this model meets requirements for the relevant industry and use case and addresses unforeseen product misuse\.

For more detailed information on ethical considerations for this model, please see the Model Card\+\+ Bias, Explainability, Safety & Security, and Privacy Subcards\.

Please report model quality, risk, security vulnerabilities or NVIDIA AI concerns[here](https://www.nvidia.com/en-us/support/submit-security-vulnerability/)\.

## [https://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16#bias](https://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16#bias)Bias

FieldResponseParticipation considerations from adversely impacted groups[protected classes](https://www.senate.ca.gov/content/protected-classes)in model design and testing:NoneMeasures taken to mitigate against unwanted bias:NoneBias Metric \(If Measured\):None
## [https://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16#explainability](https://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16#explainability)Explainability

FieldResponseIntended Task/Domain:Passage and query embedding for question and answer retrievalModel Type:Transformer encoderIntended Users:Generative AI creators working with conversational AI models \- users who want to build a multilingual question and answer application over a large text corpus, leveraging the latest dense retrieval technologies\.Output:Array of float numbers \(Dense Vector Representation for the input text\)Describe how the model works:Model transforms the tokenized input text into a dense vector representation\.Name the adversely impacted groups this has been tested to deliver comparable outcomes regardless of:Not ApplicableTechnical Limitations & Mitigation:The model’s max sequence length is 32768\. Therefore, longer text inputs should be truncated\.Verified to have met prescribed NVIDIA quality standards:YesPerformance Metrics:Accuracy, Throughput, and LatencyPotential Known Risks:This model does not always guarantee to retrieve the correct passage\(s\) for a given query\.Licensing:This model and its associated configuration files are licensed under the[OpenMDW License Agreement, version 1\.1 \(OpenMDW\-1\.1\)](https://openmdw.ai/license/1-1/)\. Additional Information: Built with[Ministral\-3\-3B\-Instruct\-2512](https://huggingface.co/mistralai/Ministral-3-3B-Instruct-2512)which is released under Apache 2\.0\.
## [https://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16#privacy](https://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16#privacy)Privacy

FieldResponseGeneratable or reverse engineerable personal data?NoneWas consent obtained for any personal data used?Not ApplicablePersonal data used to create this model?None KnownHow often is the dataset reviewed?Before Every ReleaseIs there provenance for all datasets used in training?YesDoes data labeling \(annotation, metadata\) comply with privacy laws?YesWas data from user interactions with the AI model \(e\.g\. user input and prompts\) used to train the model?YesIs data compliant with data subject requests for data correction or removal, if such a request was made?No, not possible with externally\-sourced data\.Applicable Privacy Policy[https://www\.nvidia\.com/en\-us/about\-nvidia/privacy\-policy/](https://www.nvidia.com/en-us/about-nvidia/privacy-policy/)
## [https://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16#safety](https://huggingface.co/nvidia/Nemotron-3-Embed-1B-BF16#safety)Safety

FieldResponseModel Application\(s\):Text Embedding for RetrievalDescribe the physical safety impact \(if present\)\.Not ApplicableUse Case Restrictions:This model and its associated configuration files are licensed under the[OpenMDW License Agreement, version 1\.1 \(OpenMDW\-1\.1\)](https://openmdw.ai/license/1-1/)\. Additional Information: Built with[Ministral\-3\-3B\-Instruct\-2512](https://huggingface.co/mistralai/Ministral-3-3B-Instruct-2512)which is released under Apache 2\.0\.Model and dataset restrictions:The Principle of least privilege \(PoLP\) is applied limiting access for dataset generation and model development\. Restrictions enforce dataset access during training, and dataset license constraints adhered to\.