Upload moo_merged_greedy.json with huggingface_hub

Upload moo_merged_kenlm.json with huggingface_hub

Upload inference_pyannote.py with huggingface_hub

Upload PLAN_AND_PROGRESS.md with huggingface_hub

Upload README.md with huggingface_hub

Update README.md

Add Round 2 finetuned model: canary-finnish-v2.nemo (best epoch=11, val_wer=0.1318)

Add R2 full benchmark results (greedy + KenLM 5M) across all 4 eval sets

Add inference_vad.py: VAD-based long-form ASR with Silero VAD chunking and Whisper-style JSON output

Update README with Round 2 benchmarks, inference guide, and parameter recommendations

Upload folder using huggingface_hub

Update README.md

Upload folder using huggingface_hub

Upload folder using huggingface_hub

Upload inference_example.py with huggingface_hub

Upload README.md with huggingface_hub

Upload KENLM_PLAN.md with huggingface_hub

Upload PLAN_AND_PROGRESS.md with huggingface_hub

Upload README.md with huggingface_hub

Upload folder using huggingface_hub

initial commit