Gigaam v3 Ctc Coreml
This is a Core ML conversion of v3ctc from ai-sage/GigaAM-v3. It contains one encoder-plus-CTC-head ML Program whose output is log-softmax.
Modellquelle
- Art
- Other
- Dateigröße
- 842 MB
- Basismodell
- ai-sage/GigaAM-v3
- Version
- vRevision c6eb924
- Aktualisiert
- 1. Nov. 2026
- Downloads
- 2
- Quelle
- Hugging Face
Quellenbeschreibung
GigaAM v3_ctc Core ML staged artifact
This is a Core ML conversion of v3_ctc from ai-sage/GigaAM-v3. It contains one encoder-plus-CTC-head ML Program whose output is log-softmax.
Provenance
repository: https://github.com/salute-developers/GigaAM
revision: 559d88d6b72541412743929f633a6ae7c9950b85
variant: v3_ctc
- Conversion precision:
float32 - Upstream license:
MIT
Runtime contract
The artifact uses fixed batch-one Core ML inputs for Russian audio of at most 25 seconds. See model-config.json and conversion-manifest.json for the exact tensor and checksum contracts.
This staged artifact must pass its parity validation before publication.
Quellen
1 Quelle
Quellen
1 QuelleVerifiziert 31. Okt.
Download files
Repository files
File inventory only; completeness is unverified.
Modellartefakte
1 Artefakt
Modellartefakte
1 Artefaktmodel.mlpackage/Data/com.apple.CoreML/weights/weight.bin
bin · 842 MB · SHA-256 f1114290286a…51ad · Hugging Face
Download fileQuellenauszüge
3 Auszüge
Quellenauszüge
3 Auszüge--- license: mit language: - ru pipeline_tag: automatic-speech-recognition library_name: coreml base_model: ai-sage/GigaAM-v3 tags: - coreml - gigaam - speech-recognition - macos - ctc - v3_ctc --- # GigaAM v3_ctc Core ML staged artifact This is a Core ML conversion of `v3_ctc` from [ai-sage/GigaAM-v3](https://huggingface.co/ai-sage/GigaAM-v3). It contains one encoder-plus-CTC-head ML Program whose output is log-softmax. ## Provenance ```yaml repository: https://github.com/salute-developers/GigaAM revision: 559d88d6b72541412743929f633a6ae7c9950b85 variant: v3_ctc ``` - Conversion precision: `float32` - Upstream license: `MIT` ## Runtime contract The artifact uses fixed batch-one Core ML inputs for Russian audio of at most 25 seconds. See `model-config.json` and `conversion-manifest.json` for the exact tensor and checksum contracts. This staged artifact must pass its parity validation before publication.
Source context: 2 downloads · 0 likes · Pipeline automatic-speech-recognition · Library coreml · Repo vadimsuhanov/gigaam-v3-ctc-coreml
Source context: 8 downloads · 0 likes · Pipeline automatic-speech-recognition · Library coreml · Repo vadimsuhanov/gigaam-v3-ctc-coreml
