30 days ago
-
Fix: Peak audio normalization (-1 dB), robust speaker search, and tuned CFG for crystal-clear benchmark generation
-
Fix: Peak audio normalization (-1 dB), robust speaker search, and tuned CFG for crystal-clear benchmark generation
-
Fix: Peak audio normalization (-1 dB), robust speaker search, and tuned CFG for crystal-clear benchmark generation
-
Fix: Peak audio normalization (-1 dB), robust speaker search, and tuned CFG for crystal-clear benchmark generation
-
Upload eval_benchmarks/epoch_60/02_hindi_female.wav with huggingface_hub
-
Upload eval_benchmarks/epoch_60/04_english_female.wav with huggingface_hub
-
Upload eval_benchmarks/epoch_60/03_english_male.wav with huggingface_hub
-
Upload eval_benchmarks/epoch_60/01_hindi_male.wav with huggingface_hub
-
Upload checkpoints/viuai_tts_latest.pt with huggingface_hub
-
Upload checkpoints/viuai_tts_epoch_60.pt with huggingface_hub
-
Upload eval_benchmarks/step_015000/02_hindi_female.wav with huggingface_hub
-
Upload eval_benchmarks/step_015000/04_english_female.wav with huggingface_hub
-
Upload eval_benchmarks/step_015000/03_english_male.wav with huggingface_hub
-
Upload eval_benchmarks/step_015000/01_hindi_male.wav with huggingface_hub
-
Upload checkpoints/viuai_tts_latest.pt with huggingface_hub
-
Upload checkpoints/viuai_tts_step_015000.pt with huggingface_hub
-
Upload eval_benchmarks/step_010000/02_hindi_female.wav with huggingface_hub
-
Upload eval_benchmarks/step_010000/04_english_female.wav with huggingface_hub
-
Upload eval_benchmarks/step_010000/03_english_male.wav with huggingface_hub
-
Upload eval_benchmarks/step_010000/01_hindi_male.wav with huggingface_hub
-
Upload checkpoints/viuai_tts_latest.pt with huggingface_hub
-
Upload checkpoints/viuai_tts_step_010000.pt with huggingface_hub
-
Fix: Peak audio normalization (-1 dB), robust speaker search, and tuned CFG for crystal-clear benchmark generation
-
Fix: Peak audio normalization (-1 dB), robust speaker search, and tuned CFG for crystal-clear benchmark generation
-
Upload eval_benchmarks/epoch_30/02_hindi_female.wav with huggingface_hub
-
Upload eval_benchmarks/epoch_30/04_english_female.wav with huggingface_hub
-
Upload eval_benchmarks/epoch_30/03_english_male.wav with huggingface_hub
-
Upload eval_benchmarks/epoch_30/01_hindi_male.wav with huggingface_hub
-
Upload checkpoints/viuai_tts_latest.pt with huggingface_hub
-
Upload checkpoints/viuai_tts_epoch_30.pt with huggingface_hub
-
Fix: Eliminate 30MB raw audio waveform IPC serialization to permanently prevent Docker /dev/shm shared memory crashes
-
Fix: Clean text conditioning in generate() matching forward_train exactly (eliminating random noise injection)
-
Turbo Speed: Vectorized FFT F0 pitch, cuDNN autotune, Tensor Core precision, speaker RAM cache & prefetch DataLoader
-
Turbo Speed: Vectorized FFT F0 pitch, cuDNN autotune, Tensor Core precision, speaker RAM cache & prefetch DataLoader
-
Master Fix: Real F0 pitch, acoustic Mel-80 to 100 filterbank vocoder projection, soundfile multi-format loader, safe weights_only
-
Upload eval_benchmarks/step_006500/02_hindi_female.wav with huggingface_hub
-
Master Fix: Real F0 pitch, acoustic Mel-80 to 100 filterbank vocoder projection, soundfile multi-format loader, safe weights_only
-
Upload eval_benchmarks/step_006500/04_english_female.wav with huggingface_hub
-
Upload eval_benchmarks/step_006500/03_english_male.wav with huggingface_hub
-
Upload eval_benchmarks/step_006500/01_hindi_male.wav with huggingface_hub
-
Upload checkpoints/viuai_tts_latest.pt with huggingface_hub
-
Upload checkpoints/viuai_tts_step_006500.pt with huggingface_hub
-
Upload eval_benchmarks/step_006250/02_hindi_female.wav with huggingface_hub
-
Upload eval_benchmarks/step_006250/04_english_female.wav with huggingface_hub
-
Upload eval_benchmarks/step_006250/03_english_male.wav with huggingface_hub
-
Upload eval_benchmarks/step_006250/01_hindi_male.wav with huggingface_hub
-
Upload checkpoints/viuai_tts_latest.pt with huggingface_hub
-
Upload checkpoints/viuai_tts_step_006250.pt with huggingface_hub
-
Upload eval_benchmarks/step_006000/02_hindi_female.wav with huggingface_hub
-
Upload eval_benchmarks/step_006000/04_english_female.wav with huggingface_hub
