text2speech
text2speech copied to clipboard
jhudsl
Text to Speech
In this [line](https://github.com/jhudsl/text2speech/pull/40/files#diff-a884186ffd60fa8978f58119ca90506c4da56d0e7c138434fe63a0bc2eb53ab9R91), I need to write some code to specify the path to the Python version we will be using in `reticulate` to port over the Python code into...
Ask user for 2 inputs: text and user's audio sample (30 sec) Then, run it through [Coqui XTTS](https://tts.readthedocs.io/en/dev/models/xtts.html) Sample Python code: ``` from TTS.api import TTS tts = TTS(model_name="tts_models/multilingual/multi-dataset/xtts_v2", gpu=False)...
RStudio [crashes](https://github.com/jhudsl/text2speech/pull/7#discussion_r1187770422) when attempting to use a new model. When tts is using a new model (that doesn't exist in `/Users/howardbaek/Library/Application Support/tts`), it needs to download the model to that...
Can we have them create a dictionary or are there common words that are mispronounced? How might we try to aid the users' desire to have words pronounced correctly? Or...
An idea from @seankross : Move all the Coqui TTS functionality into a separate R package, create a Docker image with Coqui TTS installed, and call `tts_coqui()` on the Docker...
Perhaps, to make sure users have Coqui TTS installed on their system, we can specify `SystemRequirements: tts`. This would be similar to what ari does with [ffmpeg](https://github.com/jhudsl/ari/blob/main/DESCRIPTION#L15)
Make a new tag for this repo https://docs.github.com/en/repositories/releasing-projects-on-github/managing-releases-in-a-repository And in the DESCRIPTION file, bump the number up to 1.0.0 Then do the steps in #27
After implementing Coqui TTS, it might be worth looking into - [MMS: Scaling Speech Technology to 1000+ languages](https://github.com/facebookresearch/fairseq/tree/main/examples/mms). - [TensorFlowTTS](https://github.com/TensorSpeech/TensorFlowTTS). Probably need to use `reticulate` , as the [examples](https://github.com/TensorSpeech/TensorFlowTTS#end-to-end-examples) are...
The [list](https://github.com/jhudsl/text2speech/blob/main/R/aaa_utils.R#L130) of paths to check is not comprehensive enough to be useful. We do look through these paths and if the exe file is not found, we instruct the...