harrier-oss-v1-0.6b

Repository
microsoft/harrier-oss-v1-0.6b
Engine
vLLM
Listed
4 Aug 2026
Status
Popular

harrier-oss-v1 is a family of multilingual text embedding models developed by Microsoft. The models use decoder-only architectures with last-token pooling and L2 normalization to produce dense text embeddings. They can be applied to a wide range of tasks, including but not limited to retrieval, clustering, semantic similarity, classification, bitext mining, and reranking. The models achieve state-of-the-art results on the Multilingual MTEB v2 benchmark as of the release date.

Serves
Embeddings
Parameters
693M
Context
32,768 tokens
Weights
1.2 GB
License
MIT

Capabilities

Read from the model's capability tags. Requests are admitted per tag.

Embeddings embedding
  • POST /v1/embeddings

Endpoint task mode embed

Inputs
Text

Launch settings

Every engine serving the model starts with these.

Set by the owner. auto: the serving node derives it. default: the platform's value, left in place.

Engine
vLLM
Context window --max-model-len
32,768 tokens
Quantization --quantization
none default
Weights load as stored, bfloat16.
KV cache dtype --kv-cache-dtype
auto
No flag: the checkpoint's KV quantization, else the model dtype.
Scheduling policy --scheduling-policy
priority default
Interactive requests ahead of batch work; same-class traffic in arrival order.
Trust remote code --trust-remote-code
on default
Python shipped in the bundle may run at load (the default).

Engine tuning

Every setting auto: the serving node derives each one.

Set by the owner. auto: the serving node derives it.

Mixture of experts

Expert parallel --enable-expert-parallel
auto
Expert load balancing --enable-eplb
auto
All-to-all backend --all2all-backend
auto

Memory and batching

GPU memory utilization --gpu-memory-utilization
auto
Max sequences --max-num-seqs
auto
Max batched tokens --max-num-batched-tokens
auto
Long-prefill threshold --long-prefill-token-threshold
auto

Execution

Enforce eager --enforce-eager
auto
Prefix caching --enable-prefix-caching
auto
Async scheduling --no-async-scheduling
auto
Custom all-reduce --disable-custom-all-reduce
auto

Speculative decoding

Draft tokens checked by the model in one pass.

Speculative decoding
not configured
An endpoint's own setting, or the serving node's default for the model family, applies.

Architecture

Read from the model's files at upload.

Architecture
qwen3
Parameters
692,903,936
Layers num_hidden_layers
28
Hidden size hidden_size
1,024
Attention heads num_attention_heads
16 (8 KV)
Head dim head_dim
128
Vocabulary vocab_size
151,936
Native dtype torch_dtype
bfloat16
Weights on disk
1.2 GB, safetensors

Call it

Create an endpoint for the model, then send it a request in the OpenAI shape.

Shell
# Create an endpoint for this model, then provision it. # --tier-id is the service tier to run on; pick one at /endpoints/new. erectl endpoints create \ --name "My endpoint" \ --slug my-endpoint \ --model-id 82C2AB80-AB50-45CE-B624-FB6237BB8354 \ --task-mode embed \ --tier-id <tier-id> # provision takes the endpoint UUID returned by create erectl endpoints provision <endpoint-id> # Call it. Requests are routed by /<project-id>/<endpoint-slug>. curl https://api.erebine.ai/proj_ABC123/my-endpoint/v1/embeddings \ -H "Content-Type: application/json" \ -H "Authorization: Bearer $EREBINE_API_KEY" \ -d '{ "model": "harrier-oss-v1-0.6b", "input": "The capital of France is Paris." }'

Model card

As published with microsoft/harrier-oss-v1-0.6b. Images are left out; its HTML tables are set as tables.

harrier-oss-v1

ModelParametersEmbedding DimensionMax TokensMTEB v2 Score
harrier-oss-v1-270m270M64032,76866.5
harrier-oss-v1-0.6b0.6B1,02432,76869.0
harrier-oss-v1-27b27B5,37632,76874.3

Training

All models are trained with contrastive learning objectives on a large-scale mixture of multilingual datasets covering diverse tasks.
The 270m and 0.6b variants are additionally trained with knowledge distillation from larger embedding models.

Usage

Below is an example to encode queries and passages from the MS-MARCO passage ranking dataset.

Sentence Transformers

from sentence_transformers import SentenceTransformer

model = SentenceTransformer("microsoft/harrier-oss-v1-0.6b", model_kwargs={"dtype": "auto"})

queries = [
    "how much protein should a female eat",
    "summit define",
]
documents = [
    "As a general guideline, the CDC's average requirement of protein for women ages 19 to 70 is 46 grams per day. But, as you can see from this chart, you'll need to increase that if you're expecting or training for a marathon. Check out the chart below to see how much protein you should be eating each day.",
    "Definition of summit for English Language Learners. : 1  the highest point of a mountain : the top of a mountain. : 2  the highest level. : 3  a meeting or series of meetings between the leaders of two or more governments."
]

query_embeddings = model.encode(queries, prompt_name="web_search_query")
document_embeddings = model.encode(documents)

scores = (query_embeddings @ document_embeddings.T) * 100
print(scores.tolist())

Have a look at config_sentence_transformers.json for the prompts that are pre-configured, such as web_search_query, sts_query, and bitext_query. You can also use a custom instruction directly via e.g. model.encode(queries, prompt="Instruct: Retrieve semantically similar text\nQuery: ").

Transformers

import torch
import torch.nn.functional as F

from torch import Tensor
from transformers import AutoTokenizer, AutoModel

def last_token_pool(last_hidden_states: Tensor, attention_mask: Tensor) -> Tensor:
    left_padding = (attention_mask[:, -1].sum() == attention_mask.shape[0])
    if left_padding:
        return last_hidden_states[:, -1]
    else:
        sequence_lengths = attention_mask.sum(dim=1) - 1
        batch_size = last_hidden_states.shape[0]
        return last_hidden_states[torch.arange(batch_size, device=last_hidden_states.device), sequence_lengths]

def get_detailed_instruct(task_description: str, query: str) -> str:
    return f'Instruct: {task_description}\nQuery: {query}'

# Each query must come with a one-sentence instruction that describes the task
task = 'Given a web search query, retrieve relevant passages that answer the query'
queries = [
    get_detailed_instruct(task, 'how much protein should a female eat'),
    get_detailed_instruct(task, 'summit define')
]
# No need to add instruction for retrieval documents
documents = [
    "As a general guideline, the CDC's average requirement of protein for women ages 19 to 70 is 46 grams per day. But, as you can see from this chart, you'll need to increase that if you're expecting or training for a marathon. Check out the chart below to see how much protein you should be eating each day.",
    "Definition of summit for English Language Learners. : 1  the highest point of a mountain : the top of a mountain. : 2  the highest level. : 3  a meeting or series of meetings between the leaders of two or more governments."
]
input_texts = queries + documents

tokenizer = AutoTokenizer.from_pretrained('microsoft/harrier-oss-v1-0.6b')
model = AutoModel.from_pretrained('microsoft/harrier-oss-v1-0.6b', dtype='auto')
model.eval()
model.cuda()

max_length = 32768
# Tokenize the input texts
batch_dict = tokenizer(input_texts, max_length=max_length, padding=True, truncation=True, return_tensors='pt')
batch_dict = {k: v.cuda() for k, v in batch_dict.items()}

outputs = model(**batch_dict)
embeddings = last_token_pool(outputs.last_hidden_state, batch_dict['attention_mask'])

# normalize embeddings
embeddings = F.normalize(embeddings, p=2, dim=1)
scores = (embeddings[:2] @ embeddings[2:].T) * 100
print(scores.tolist())

Supported Languages

The models are trained on multilingual data and support a wide range of languages,
including but not limited to: Arabic, Bulgarian, Catalan, Czech, Danish, German, Greek, English, Spanish,
Estonian, Persian, Finnish, French, Hebrew, Hindi, Croatian, Hungarian, Indonesian, Italian, Japanese,
Korean, Lithuanian, Latvian, Macedonian, Malay, Dutch, Norwegian, Polish, Portuguese, Romanian, Russian,
Slovak, Slovenian, Albanian, Serbian, Swedish, Thai, Turkish, Ukrainian, Urdu, Vietnamese, and Chinese.

Evaluation

Please follow the mteb repository on how to reproduce our scores.
The evaluation prompts used for each task are also available at mteb_v2_eval_prompts.json.

FAQ

1. Do I need to add instructions to the query?

Yes, this is how the model is trained, otherwise you will see a performance degradation.
The task definition should be a one-sentence instruction that describes the task.
This is a way to customize text embeddings for different scenarios through natural language instructions.

On the other hand, there is no need to add instructions to the document side.

2. Why are my reproduced results slightly different from reported in the model card?

Different versions of transformers and pytorch could cause negligible but non-zero performance differences.

3. What pooling strategy does this model use?

The model uses last-token pooling — the embedding of the last non-padding token is used as the sentence representation.
The embedding is then L2-normalized. This is handled automatically when using Sentence Transformers.