backpack-run/Qwen3-ASR-0.6B-Backpack-ASR

Qwen3 ASR 0.6B — Backpack Voice Package

Compact multilingual ASR with language identification. This package stages immutable upstream artifacts for Backpack's voice runtime layer. It does not replace the chat model selected by the user.

Package

Field	Value
Capability	speech-to-text
Input	audio
Output	text
Parameters	600000000
Runtime	qwen-asr
Configured runtime revision	qwen-asr==0.0.6

| Format | safetensors | | Precision | BF16 | | Primary artifact | model.safetensors | | Total packaged-file size | 1.8 GiB | | Recommended RAM | 5.0 GB | | Languages | multilingual |

Validation status

The packager verified the immutable revision, selected-file inventory, non-empty files, hashes, configuration JSON, and the primary artifact container/header. It also loaded the package and passed deterministic audio inference with the configured runtime.

| Integrity | Metadata | Runtime load | Audio inference | Tokenizer | |---|---|---|---| | passed | passed | passed | passed | passed |

Run with qwen-asr

Install qwen-asr==0.0.6, then load this repository path with Qwen3ASRModel.from_pretrained(...) and call transcribe(audio=...).

Packaged runtime services

These optional, Backpack-tested runtime bundles are distributed alongside the model weights. They are not separate models and do not replace the primary model artifact above.

Runtime	Platform	Backend	Protocol
qwen-asr==0.0.6	windows-x64	pytorch-cpu	v1

• runtime/windows-x64/pytorch-cpu/LICENSE — 11.3 KiB — 1eb85fc97224598dad1852b5d6483bbcf0aa8608790dcc657a5a2a761ae9c8c6
• runtime/windows-x64/pytorch-cpu/NOTICE — 207.0 B — f1da68355a75a27ee586c57297502a8d1bb30d0c8f6ecfe129a303256fb4eadc
• runtime/windows-x64/pytorch-cpu/checksums.sha256 — 358.0 B — 31dc479c65971804f45610ff3d45fa8457bba22fc4dcc4fe7731416427788ba2
• runtime/windows-x64/pytorch-cpu/qwen-asr-0.0.6-windows-x64-pytorch-cpu.zip — 8.2 KiB — 330deaeace23cfefe414fdc71d28b19e12aaaefa0b0a3e1b3fb26d7807c656c8
• runtime/windows-x64/pytorch-cpu/runtime-protocol-v1.schema.json — 1.0 KiB — 705d5dadf165f9819092fa275fff1783ad37b6ac6d4c2279ec092fcf3f15adf5
• runtime/windows-x64/pytorch-cpu/runtime.yaml — 3.2 KiB — ff8f4d0728b1734575d40144ac62ee83e0e4ba37adde17eac6a33c6889ef9e1c

Provenance

• Upstream: Qwen/Qwen3-ASR-0.6B

• Immutable revision: 5eb144179a02acc5e5ba31e748d22b0cf3e303b0

• License: apache-2.0

• Backpack copied the selected upstream artifacts without modifying model weights.

• Backpack did not train this model and does not claim ownership of it.

Files and checksums

• chat_template.json — 1.1 KiB — 75a8cfca24f00de72d796fbfed6858fc9614ef3dabd8696684cc3bc03a9c58ff
• config.json — 6.0 KiB — 76d3ae4601ce939830b2517f4a6cadb86cc51316c3900af6b020b051c21a478c
• generation_config.json — 142.0 B — 1da527824d81e07118facff437e03f2e24a23311e3bdeb2368973fe77e5f275c
• merges.txt — 1.6 MiB — 8831e4f1a044471340f7c0a83d7bd71306a5b867e95fd870f74d0c5308a904d5
• model.safetensors — 1.7 GiB — 79d6cbd4c98c7bbffe9db2edac07f56cd6637d0d5944b27f6c2b8353840323ea
• preprocessor_config.json — 330.0 B — 45e120a4eda2c20c5d7f2ea9354e63536bf35e27aa573fb7cdf78017b378770d
• tokenizer_config.json — 12.2 KiB — 4942d005604266809309cabc9f4e9cb89ce855d59b14681fdc0e1cc62ea26c4c
• vocab.json — 2.6 MiB — ca10d7e9fb3ed18575dd1e277a2579c16d108e32f27439684afa0e10b1440910

Review the upstream model card and license before use or redistribution. Speech systems can mis-transcribe, synthesize misleading content, or behave differently across languages and accents.