voicescribe-corrector-nvidia / UPSTREAM_SOURCE.md
Andrewsab's picture
Voice Scribe mirror corrector_nvidia from keisuke-miyako/Qwen3-4B-Instruct-2507-onnx-int4@d5e5e01dd940
d5f6069 verified
# Voice Scribe Model Mirror
This repository is a Voice Scribe distribution mirror. The model artifacts are
copied from the upstream repository and the source revision below is pinned.
| Field | Value |
| --- | --- |
| Layout key | `corrector_nvidia` |
| Target directory in installer | `qwen3-4b-instruct-2507-onnx-int4` |
| Upstream repo | `keisuke-miyako/Qwen3-4B-Instruct-2507-onnx-int4` |
| Upstream revision | `d5e5e01dd9404c8e1876be6d34cceb16e5348e9a` |
| Upstream resolved SHA | `d5e5e01dd9404c8e1876be6d34cceb16e5348e9a` |
| Mirror created | `2026-04-23T22:25:17Z` |
| Description | Qwen3-4B Instruct 2507 ONNX INT4 NVIDIA/ORT GenAI layout. |
| License metadata | `{"license": null, "license_files": [], "license_tags": []}` |
## Installer Contract
This mirror corresponds to `parakeet/installer/wrapper/model_catalog.py`.
Required files for installer validation:
```json
[
"genai_config.json",
"chat_template.jinja",
"config.json",
"added_tokens.json",
"merges.txt",
"special_tokens_map.json",
"tokenizer.json",
"tokenizer_config.json",
"vocab.json",
"model.onnx",
"model.onnx.data"
]
```
Allowed installer subset patterns:
```json
[
"genai_config.json",
"chat_template.jinja",
"config.json",
"added_tokens.json",
"merges.txt",
"special_tokens_map.json",
"tokenizer.json",
"tokenizer_config.json",
"vocab.json",
"model.onnx",
"model.onnx.data"
]
```
## Redistribution Note
Do not make this repository public unless the upstream license and model card
allow redistribution for the intended use. Private mirrors are for operational
distribution convenience and reproducible installs.