-
Add reproducible validation evidence kit (harness + results + README)
-
Remove demo leftovers; keep hosted model weights + README
-
Upload wav2vec2_xlsr53_voiceguard/model.safetensors with huggingface_hub
-
Upload wav2vec2_xlsr53_voiceguard/preprocessor_config.json with huggingface_hub
-
Upload wav2vec2_xlsr53_voiceguard/config.json with huggingface_hub
-
Add live-inference VoiceGuardRawNet INT8 artifact (2.66MB) used by the real-time WS pipeline
-
Add interactive web interface with Three.js 3D shield
-
Add FastAPI full-stack server
-
Elevate README with high-impact visuals, Mermaid diagrams, badges, and architecture tables
-
Update README model card with metadata tags
-
Add production Wav2Vec2 training & inference pipeline for SIH26104 (README.md)
-
Add production Wav2Vec2 training & inference pipeline for SIH26104 (inference.py)
-
Add production Wav2Vec2 training & inference pipeline for SIH26104 (train.py)
-
Add production Wav2Vec2 training & inference pipeline for SIH26104 (dataset.py)
-
Add production Wav2Vec2 training & inference pipeline for SIH26104 (voiceguard_wav2vec2.py)
-
Credit Indrajit Yadav as founder & lead developer
-
Add Indrajit Yadav to team credits
-
Add attractive README with team credits
-
Delete voiceguard_int8.pt with huggingface_hub
-
Switch to safetensors (no pickle, no security false positive)
-
Update README.md
-
VoiceGuard: INT8 RawNet deepfake voice detection demo
-
initial commit
