Whisper-Hindi2Hinglish-Apex for Android (Screaming Flow)
Oriserve/Whisper-Hindi2Hinglish-Apex (Apache-2.0, a Whisper large-v3-turbo fine-tune) exported for phones: the encoder in four fp16 ONNX parts for ONNX Runtime's QNN backend (Snapdragon NPU), and the decoder as a single-step model with its own cache plus a cross-attention model, for the CPU. Decodes to Hinglish in Latin script with the English language token. Used by the Screaming Flow Android app, which checks every file against its SHA-256 before loading it.
Inference Providers NEW
This model isn't deployed by any Inference Provider. ๐ Ask for provider support
Model tree for pradeep-pioneer/screaming-flow-apex-android
Base model
openai/whisper-large-v3 Finetuned
openai/whisper-large-v3-turbo Finetuned
Oriserve/Whisper-Hindi2Hinglish-Apex