espnet/SingingSDS

Merge with main

Update interface.py

Update requirements.txt

Update requirements.txt

Update requirements.txt

Update requirements.txt to add 'soxr' and 'spaces' dependencies

Refactor run_pipeline and update_metrics methods to use a global pipeline instance and improve parameter handling

Update gradio version in requirements.txt to 5.35.0 due to hf compatibility

Refactor run_pipeline and update_metrics methods to support inference on HF ZeroGPU

Relax Python version specification in README.md

Add GPU decorator to run_pipeline and update_metrics methods in GradioInterface

Merge branch 'refactor' into hf

Update README.md to reflect current model additions and remove outdated entries

Merge branch 'refactor' into hf

Replace kanjiconv with pykakasi in requirements.txt for improved compatibility

Merge branch 'refactor' into hf

Update README.md to enhance example usage and add inference-only mode instructions

replacing breakpoints with traceback logging in cli.py, interface.py, and minimax.py

Replace kanjiconv dependency with pykakasi for broader Python compatibility

Make eval_results_csv optional in CLI

fix bilingual response following prompt updates

Merge branch 'refactor' into hf

Clear outputs when clearing input audio in interface.py

Update audio_output to be non interactive in interface.py

Merge branch 'refactor' into hf

increase gemini max_output_tokens

Merge branch 'refactor' into hf

update configurations, including Yaoyin character prompt, default setup, svs model choices

change kising dataset source to public dataset

updated Yaoyin's prompt

Updated character prompt

Merge branch 'refactor' into hf

Refactor CLI to support multiple query audio inputs

increase max_sentences from 2 to 5 for improved response generation.

Add more iterations to prompt gemini

update character prompt for better performance in current models

change bilingual svs to use espnet/mixdata_svs_visinger2_spkemb_lang_pretrained_avg

Merge pull request #3 from Masao-Someki/feature/add_per

Move initialization process to init_per()

Merge branch 'refactor' into feature/add_per

update eval prototypes for per

Update PER calculation

Merge branch 'refactor' of github.com:Masao-Someki/SingingSDS-dev into feature/add_per

Add asr model for PER

update PER placeholder

Revert "add per template"

add per template

specify basic-pitch version to 0.4.0 in requirements.txt

remove tensorflow from requirements.txt

Merge branch 'refactor' into hf