Spaces:

abi3420
/

dubify-backend

Running on Zero

App Files Files Community

dubify-backend

Commit History

Oct 29 2 days ago

03:59
Lock 100% single Valluvar voice for entire video and fix end duration Running

abinjoyal

3e5a8d5
03:43
Set Valluvar ta-IN-ValluvarNeural as primary dubbing voice matching Dubbed_Video 9

abinjoyal

4b2999e
03:29
Restore exact signature dubbing voice from Dubbed_Video 9

abinjoyal

ab4f8c4
03:14
Fix ZeroGPU CUDA init by running Meta MMS-TTS on CPU and ensure natural Tamil dubbing voice

abinjoyal

4ac9319
03:05
Restore original natural neural voice and completely remove robot voice

abinjoyal

1b6b5a4
02:53
Set user-verified resonant pitch-shifted voice as direct primary engine

abinjoyal

563adda
02:41
Fix: Add local Meta MMS Tamil Neural TTS and prevent gTTS robotic fallback

abinjoyal

c2d1102

Oct 28 3 days ago

17:19
feat: add TTSService for neural text-to-speech conversion with multi-language and voice cloning support

abinjoyal

17efd3b
17:06
feat: implement text-to-speech service with voice cloning and Tamil phonetic mapping

abinjoyal

8a1add1

Oct 25 6 days ago

16:25
feat: add text-to-speech and voice dubbing backend services for audio transcription

abinjoyal

02b08ac

Oct 17 14 days ago

17:18
feat: adjust translate_batch_with_llm batch size to 20 and implement rate limiting handling for Groq API

abinjoyal

e8920f0
16:47
feat: update get_groq_active_models to exclude additional classifier substrings and include more chat model families

abinjoyal

e065fc6
16:18
feat: refine get_groq_active_models to exclude safety/guard substrings and adjust translate_batch_with_llm batch size to improve token handling

abinjoyal

098fd86
16:04
feat: enhance get_groq_active_models to include openai/gpt-oss models and exclude TTS/speech-only types; improve JSON response handling in translate_batch_with_llm

abinjoyal

dcbc201
14:48
feat: enhance get_groq_active_models to filter for valid chat LLM models and exclude non-compliant types

abinjoyal

088e60a
14:11
feat: implement dynamic querying of active Groq models for improved translation flexibility

abinjoyal

1607102
12:07
feat: update Groq API model list for improved translation performance

abinjoyal

d74bba8
11:50
feat: enhance Groq API integration for batch translation with multiple model support and improved error handling

abinjoyal

3b5a522
11:40
feat: implement batch translation with LLM APIs for improved Tamil dubbing quality

abinjoyal

b642c62
11:09
feat: refactor _generate_chunk_on_gpu to use direct model.generate with forced_bos_token_id and remove Hugging Face pipeline

abinjoyal

36f07da
10:36
feat: update VideoDubbingService to improve Edge-TTS handling and add emergency gTTS fallback; refine LocalNMTService generation parameters

abinjoyal

563c3fc
05:36
feat: refactor translation methods to utilize NLLB parameters and improve tokenization process

abinjoyal

ff14c57
05:29
feat: enhance LocalNMTService to support device selection for translation pipeline

abinjoyal

fd9deaf
05:07
feat: refactor translation methods to utilize Hugging Face pipeline for improved efficiency and maintainability

abinjoyal

65df628
04:49
feat: enhance LocalNMTService to set target language and log sample outputs in TranslationService

abinjoyal

2ef1343
04:13
feat: enhance translation service to prioritize local NMT for improved performance and reliability

abinjoyal

c5160ee

Oct 16 15 days ago

17:19
feat: Refactor translation services and remove deprecated tests

abinjoyal

fdadf8d
17:14
feat: refactor translation methods to use Hugging Face Pipeline for improved performance and reliability

abinjoyal

b1d8dd2
17:07
feat: add translation failure guards and logging in VideoDubbingService; implement verification tests for NLLB Tamil translation

abinjoyal

a6607ca
16:59
feat: update token generation configuration in IndicTrans2 and Local NMT services

abinjoyal

1038b00
16:53
feat: enhance translation service with local NLLB-200 fallback and improved logging

abinjoyal

0ab5117
16:45
feat: integrate IndicTrans2 service for enhanced translation capabilities and add comprehensive test suite

abinjoyal

d644328
16:10
feat: enhance translation service with improved rate limiting and fallback mechanisms

abinjoyal

7979798
15:58
feat: add GoogleTranslator integration with rate limiting and fallback mechanism

abinjoyal

d92709d
15:09
fix: translation rate limiter, exponential backoff, and HF fallback

abinjoyal

ddea29d
13:20
feat: set default voice gender to male and enhance logging for voice resolution

abinjoyal

5b7bd7e
13:11
feat: purge old TTS cache and set default voice gender to male

abinjoyal

c8e77c0
13:04
feat: set default voice to male only across all dubbing services

abinjoyal

52d9637
13:02
fix: throttle translation workers to max_workers=3 and eliminate broken CPU NLLB fallback

abinjoyal

1ef0327
12:58
feat: update voice gender handling to default to male and improve logging for voice selection

abinjoyal

b71ad4c
12:53
feat: enhance translation service with ultra-fast parallel translation and improved batch processing

abinjoyal

384f730
12:24
feat: enhance translation service with improved text sanitization for Tamil translations

abinjoyal

7575d44
12:19
feat: enhance translation service with Tamil validation and testing framework

abinjoyal

861b3ce
11:56
feat: optimize translation services with improved error handling and performance enhancements

abinjoyal

2269f1f
11:12
feat: enhance translation service with MyMemory integration and improved error handling

abinjoyal

dec5680
10:55
feat: enhance translation services with batch processing and improved error handling

abinjoyal

17b89f3

Oct 14 17 days ago

16:36
feat: add backend dubbing and translation services for audio processing

abinjoyal

95aabd7
15:27
fix: update backend python dependencies in requirements.txt

abinjoyal

b04d544
15:26
feat: add local NLLB-200 translation service for offline neural machine translation

abinjoyal

54e4984
15:25
feat: implement LocalNMTService and translation service for offline NLLB-200 text translation

abinjoyal

b0dc8b0

Commit History

Oct 29 2 days ago

Lock 100% single Valluvar voice for entire video and fix end duration Running

Set Valluvar ta-IN-ValluvarNeural as primary dubbing voice matching Dubbed_Video 9

Restore exact signature dubbing voice from Dubbed_Video 9

Fix ZeroGPU CUDA init by running Meta MMS-TTS on CPU and ensure natural Tamil dubbing voice

Restore original natural neural voice and completely remove robot voice

Set user-verified resonant pitch-shifted voice as direct primary engine

Fix: Add local Meta MMS Tamil Neural TTS and prevent gTTS robotic fallback

Oct 28 3 days ago

feat: add TTSService for neural text-to-speech conversion with multi-language and voice cloning support

feat: implement text-to-speech service with voice cloning and Tamil phonetic mapping

Oct 25 6 days ago

feat: add text-to-speech and voice dubbing backend services for audio transcription

Oct 17 14 days ago

feat: adjust translate_batch_with_llm batch size to 20 and implement rate limiting handling for Groq API

feat: update get_groq_active_models to exclude additional classifier substrings and include more chat model families

feat: refine get_groq_active_models to exclude safety/guard substrings and adjust translate_batch_with_llm batch size to improve token handling

feat: enhance get_groq_active_models to include openai/gpt-oss models and exclude TTS/speech-only types; improve JSON response handling in translate_batch_with_llm

feat: enhance get_groq_active_models to filter for valid chat LLM models and exclude non-compliant types

feat: implement dynamic querying of active Groq models for improved translation flexibility

feat: update Groq API model list for improved translation performance

feat: enhance Groq API integration for batch translation with multiple model support and improved error handling

feat: implement batch translation with LLM APIs for improved Tamil dubbing quality

feat: refactor _generate_chunk_on_gpu to use direct model.generate with forced_bos_token_id and remove Hugging Face pipeline

feat: update VideoDubbingService to improve Edge-TTS handling and add emergency gTTS fallback; refine LocalNMTService generation parameters

feat: refactor translation methods to utilize NLLB parameters and improve tokenization process

feat: enhance LocalNMTService to support device selection for translation pipeline

feat: refactor translation methods to utilize Hugging Face pipeline for improved efficiency and maintainability

feat: enhance LocalNMTService to set target language and log sample outputs in TranslationService

feat: enhance translation service to prioritize local NMT for improved performance and reliability

Oct 16 15 days ago

feat: Refactor translation services and remove deprecated tests

feat: refactor translation methods to use Hugging Face Pipeline for improved performance and reliability

feat: add translation failure guards and logging in VideoDubbingService; implement verification tests for NLLB Tamil translation

feat: update token generation configuration in IndicTrans2 and Local NMT services

feat: enhance translation service with local NLLB-200 fallback and improved logging

feat: integrate IndicTrans2 service for enhanced translation capabilities and add comprehensive test suite

feat: enhance translation service with improved rate limiting and fallback mechanisms

feat: add GoogleTranslator integration with rate limiting and fallback mechanism

fix: translation rate limiter, exponential backoff, and HF fallback

feat: set default voice gender to male and enhance logging for voice resolution

feat: purge old TTS cache and set default voice gender to male

feat: set default voice to male only across all dubbing services

fix: throttle translation workers to max_workers=3 and eliminate broken CPU NLLB fallback

feat: update voice gender handling to default to male and improve logging for voice selection

feat: enhance translation service with ultra-fast parallel translation and improved batch processing

feat: enhance translation service with improved text sanitization for Tamil translations

feat: enhance translation service with Tamil validation and testing framework

feat: optimize translation services with improved error handling and performance enhancements

feat: enhance translation service with MyMemory integration and improved error handling

feat: enhance translation services with batch processing and improved error handling

Oct 14 17 days ago

feat: add backend dubbing and translation services for audio processing

fix: update backend python dependencies in requirements.txt

feat: add local NLLB-200 translation service for offline neural machine translation

feat: implement LocalNMTService and translation service for offline NLLB-200 text translation