Upload README.md with huggingface_hub

README.md

@@ -1,3 +1,28 @@
----
-license: apache-2.0
----
+# Core ML Whisper Large V3 Turbo (FP16)
+
+This archive contains the Core ML encoder/decoder pair converted from `openai/whisper-large-v3-turbo` with fp16 precision, plus associated tokenizer assets.
+
+## Files
+
+- `Encoder.mlpackage`
+- `DecoderFull.mlpackage`
+- Tokenizer & processor configs (`*json`, `merges.txt`, `vocab.json`)
+- SHA256 checksum file
+
+## Usage
+
+1. Download the tarball and checksum:
+   ```bash
+   hf download DRTR-J/whisper-large-v3-turbo-coreml-fp16 whisper-large-v3-turbo-coreml-fp16.tar.gz
+   hf download DRTR-J/whisper-large-v3-turbo-coreml-fp16 whisper-large-v3-turbo-coreml-fp16.sha256
+   ```
+2. Verify integrity:
+   ```bash
+   shasum -a 256 -c whisper-large-v3-turbo-coreml-fp16.sha256
+   ```
+3. Extract and load in your Swift/Rust app:
+   ```bash
+   tar -xzf whisper-large-v3-turbo-coreml-fp16.tar.gz
+   ```
+
+For latency guidance, warm the models once on app start to trigger ANE compilation; subsequent 68 s clips transcribe in ~13.5 s on an M4 Pro when paired with a native decoding loop.