CatalogWantedMirrorsLedgerOperating StatementDocsStudio statusCard system
alt.freeweights.qwen / Qwen3-Reranker-0.6B-4bit
From: mlx-community <huggingface.co/mlx-community>Newsgroups: alt.freeweights.qwenSubject:

Qwen3-Reranker-0.6B-4bit

Date: 2026-11-15Commit: 5f32454Signed: 2DC28CC7B26A283B ✓

Text ranking model with 596M parameters. Safetensors weights, 347 MB in all. Published by mlx-community under Apache-2.0.

596M params347MBQwen3-Reranker-0.6B-4bitsafetensorspermissive
FactsParams596MSize347 MBFormatssafetensorsTasktext rankingLicenseApache-2.0From sourcepermissive
$ fw get mlx-community/Qwen3-Reranker-0.6B-4bit
# manifest sha256 bcff2802bfe5… · 347 MB · 9 files
$ fw verify ./Qwen3-Reranker-0.6B-4bit mlx-community/Qwen3-Reranker-0.6B-4bit
From sourceHosted (signed)mlx-community/Qwen3-Reranker-0.6B-4bitLast check: 2026-11-15 05:00 UTC
Run at homeGet help

Get help: custom fine-tune or private deploy. We review your setup before we quote.

Quantized of Qwen3-Reranker-0.6B.

Works with

Run9Fine-tune8Convert8Evaluate1Scan2Apps16

Evaluate with

MTEB
OverviewModel cardFilesUse itLicenseProvenanceModel tree

Overview

Size
347 MB
Weights
335 MB (safetensors)
Parameters
596M
Formats
safetensors
Task
text ranking
License
Apache-2.0 (permissive)
Files
9, each with a sha256
Publisher
mlx-community
Added
2026-11-15
Section
alt.freeweights.qwen
Lineage

Original release from the publisher.

Base model: Qwen/Qwen3-Reranker-0.6B

Where these bytes come from
Publisher at the pinned commit→You

Sources are tried in order. A file counts only when its size and sha256 match the manifest. Sources are checked when you ask for them, not on a timer.

Sources
BadgeHostLast checkResult
From sourcehuggingface.co2026-11-15 05:00 UTCpassed

Model card

From the publisher at commit 5f32454. Shown as plain text with links; images open at the publisher.

mlx-community/Qwen3-Reranker-0.6B-4bit#

This model was converted to MLX format from Qwen/Qwen3-Reranker-0.6B using mlx-lm 0.31.3.

  • Quantization: affine 4-bit, group_size=64 (~4.5 bits/weight)
  • On-disk size: ~331 MB
  • Task: text reranking (cross-encoder, yes/no relevance scoring)

Scoring recipe#

Qwen3-Reranker is a causal LM used as a reranker: the relevance score of a (query, document) pair is softmax([logit("no"), logit("yes")])[1] at the last position of the prompt below.

import mlx.core as mx
from mlx_lm import load

model, tok = load("mlx-community/Qwen3-Reranker-0.6B-4bit")
hf = getattr(tok, "_tokenizer", tok)

INSTRUCT = "Given a web search query, retrieve relevant passages that answer the query"
PREFIX = ('<|im_start|>system\nJudge whether the Document meets the requirements '
          'based on the Query and the Instruct provided. Note that the answer can '
          'only be "yes" or "no".<|im_end|>\n<|im_start|>user\n')
SUFFIX = "<|im_end|>\n<|im_start|>assistant\n<think>\n\n</think>\n\n"
true_id, false_id = hf.convert_tokens_to_ids("yes"), hf.convert_tokens_to_ids("no")
pre, suf = hf.encode(PREFIX, add_special_tokens=False), hf.encode(SUFFIX, add_special_tokens=False)

def rerank_score(query, doc):
    content = f"<Instruct>: {INSTRUCT}\n<Query>: {query}\n<Document>: {doc}"
    ids = pre + hf.encode(content, add_special_tokens=False) + suf
    logits = model(mx.array([ids]))[:, -1, :]
    pair = mx.stack([logits[0, false_id], logits[0, true_id]])
    return float(mx.exp((pair - mx.logsumexp(pair))[1]))

print(rerank_score("What is the capital of China?", "The capital of China is Beijing."))

Files

9 files · 347 MBWeights 335 MB (safetensors). Sizes and sha256 from the signed manifest.
Hide all 9 files with their sha256
FileSizesha256
.gitattributes1.57 KB34448b82c17d60fec9b65b1f093c115ddbaadc04beb1b0140b6bfed2e012a930
README.md1.89 KBe560db1813f60798fbbf0e38b34aaf581efcca70ef5b620f2eb31032304f4b7e
chat_template.jinja741 B6f682162495ec5b39fd9005c01b6aa2a74669379fe967039f1e2cbbe8752369d
config.json1.02 KB09adff58b65e9305009c9caa4923b3365b18dd2f84135b44168aaf869278bea4
generation_config.json214 B81051cd3f6e77013827148d0b8a6ead93f8ac390d5ab805f849199f0af6a08db
model.safetensors335 MB1d212560a5b1c36186787fdae19f11f20fecfc29bef91522e12a8e0d118f4545
model.safetensors.index.json49.8 KB90d82744cdb6b7d093f0b812fc21a49b6ffa9d0084a45428f0cfd01eb4adbe12
tokenizer.json11.4 MBbe75606093db2094d7cd20f3c2f385c212750648bd6ea4fb2bf507a6a4c55506
tokenizer_config.json377 B1689852cc9c45010de040c8302a8acdc0d2c4c6c740dd7e9dd0a8c704e16eada
VersionsEach time the pinned upstream commit or file list changed in this repo. Newest first.
Upstream commitDateContents
5f324548f1 pinned now2026-11-15 05:00 UTC9 files · 347 MB

Use it

$ fw get mlx-community/Qwen3-Reranker-0.6B-4bit
# manifest sha256 bcff2802bfe5… · 347 MB · 9 files
$ fw verify ./Qwen3-Reranker-0.6B-4bit mlx-community/Qwen3-Reranker-0.6B-4bit
fw.pyChecks the signature and every sha256, tries each source in order, resumes.
$ python3 fw.py get https://freeweights.allternit.com/manifests/mlx-community/Qwen3-Reranker-0.6B-4bit.json --dir ./Qwen3-Reranker-0.6B-4bit $ python3 fw.py verify ./Qwen3-Reranker-0.6B-4bit ./mlx-community__Qwen3-Reranker-0.6B-4bit.json
hfThe Hugging Face CLI with one setting. Same repo id.
$ HF_ENDPOINT=https://freeweights.allternit.com hf download mlx-community/Qwen3-Reranker-0.6B-4bit --revision 5f324548f1d20c2b5a450f126fc6ef2fb1126524 --local-dir ./Qwen3-Reranker-0.6B-4bit
Run at home

License

Name
Apache License 2.0
Id
Apache-2.0
Class
Permissive
Redistribution
yes
Listing
Hosted (signed)
Decided
License as declared by the publisher. No person read it. 2026-11-15
Text in
README.md
Text sha256
e560db1813f60798fbbf0e38b34aaf581efcca70ef5b620f2eb31032304f4b7e
Terms gate
No extra terms beyond the license.

Read the license text

Licenses are checked against the license table. A custom license passes only after a person reads it.

Provenance

From: mlx-community <huggingface.co/mlx-community>Newsgroups: alt.freeweights.qwenSubject: Qwen3-Reranker-0.6B-4bitDate: 2026-11-15Commit: 5f32454Signed: 2DC28CC7B26A283B ✓
Upstream
mlx-community/Qwen3-Reranker-0.6B-4bit
Pinned commit
5f324548f1d20c2b5a450f126fc6ef2fb1126524
Manifest
Qwen3-Reranker-0.6B-4bit.json · .minisig
Manifest sha256
bcff2802bfe5092e63f1edd05716f7327526f227698ebb377835a4c6cffd56e1
Signature
minisign, Ed25519, key 2DC28CC7B26A283B
Created
2026-11-15 05:00 UTC
Generator
freeweights-manifest/0.1.0
Takedowns
None.

How to verify

$ curl -O https://freeweights.allternit.com/manifests/mlx-community/Qwen3-Reranker-0.6B-4bit.json $ curl -O https://freeweights.allternit.com/manifests/mlx-community/Qwen3-Reranker-0.6B-4bit.json.minisig $ minisign -Vm Qwen3-Reranker-0.6B-4bit.json -P RWQ7KGqyx4zCLXtUFOqvfv8L3g+8fpKJ1FsuxPyOYMF+4KufbUxKq1Oq $ shasum -a 256 <file> # compare with the Files tab

A manifest whose signature fails must be rejected. A file whose size or sha256 does not match must be discarded. Verify a download

Model tree

Parent

Qwen/Qwen3-Reranker-0.6B · this model is a quantization of it

No model in the catalog names this one as its parent.

Get it
Download
$ HF_ENDPOINT=https://freeweights.allternit.com hf download mlx-community/Qwen3-Reranker-0.6B-4bit
Open in Freeweights Studio

Pull, verify and run it in the Allternit app on Desktop, web or phone.

Signed manifest
minisign, key 2DC28CC7B26A283B. Qwen3-Reranker-0.6B-4bit.json. Provenance. Takedown history: none.
Source
huggingface.co/mlx-community/Qwen3-Reranker-0.6B-4bit

Related models

All of alt.freeweights.qwen →
From: BAAI <huggingface.co/BAAI>Subject: Aquila-VL-2B-llava-qwen
From: Perciqa <huggingface.co/Perciqa>Subject: Aurora-Code-1
From: BAAI <huggingface.co/BAAI>Subject: bge-reasoner-embed-qwen3-8b-0923
From: amazon <huggingface.co/amazon>Subject: BMOJOF-primed-HQwen3-8B-Instruct
From: nbeerbower <huggingface.co/nbeerbower>Subject: Bubba-Qwen3.6-27B-LoRA
From: bartowski <huggingface.co/bartowski>Subject: bytkim_Qwen3.8-27B-pi-GGUF