AshwinSankar (Ashwin Sankar) – Community Activity

New activity in ai4bharat/Cadence 5 months ago

#13 opened 9 months ago by

#16 opened 5 months ago by

#15 opened 5 months ago by

#3 opened 10 months ago by

New activity in ai4bharat/Cadence 10 months ago

#11 opened 10 months ago by

#10 opened 10 months ago by

New activity in ai4bharat/Rasa 10 months ago

#4 opened 10 months ago by

#14 opened 12 months ago by

#16 opened 10 months ago by

#13 opened 12 months ago by

#2 opened over 1 year ago by

#5 opened over 1 year ago by

#8 opened over 1 year ago by

#9 opened about 1 year ago by

#10 opened about 1 year ago by

New activity in ai4bharat/Cadence 10 months ago

#7 opened 10 months ago by

#8 opened 10 months ago by

#6 opened 10 months ago by

#5 opened 10 months ago by

#3 opened 10 months ago by

Ashwin Sankar

Ashwin Sankar

AI & ML interests

Organizations

AshwinSankar's activity

Updated config. Removed "use_cache": true and also "cache_implementation". They were interfering in some versions of transformers.

Introducing Cadence-Fast

Fixed annoying bug that caused it to not work right-of-the-bat

Getting "!!!!" at output only

Updated the docs of the python package as we now officially support passing attn_implementation too.

Completely overhauled the attention implementation. Using the existing Gemma-3 attention implementation rather than custom monkey-patched implementation.

Hindi dataset?

use it on colab

how do i integrate indic TTS to a real time inference

Flash attention 2 error, could you kindly provide solution

AttributeError: 'Tensor' object has no attribute '_pad_token_tensor'

Is it necessary to define the attention masks in .generate()?

Voice Cloning Function

tts doesn't spell numbers correctly

Inference speed

Update README.md

Upload arxiv_logo.svg

That previous logo from the link did not render correctly

Added arXiv logo beside the paper name.

Added the paper link