Congratulations @shinjiw_at_cmu
, very well deserved ofc! Thank you for everything you’ve done and continue to, to push Speech research to the limits!
@reach_vb
-
Congratulations to Shinji W for Speech Research Excellence
By
–
-
New Text-to-Video Pipelines and Schedulers Released
By
–
This is just a small subset of things released with the latest release, come check out the wide range of Text-to-Video pipelines, new schedulers and most importantly, our new docs as part of this release!
-
Music Spectrogram Diffusion generates infinite MIDI conditioned music realtime
By
–
Music Spectrogram Diffusion contributed by @krasul
, allows you to generate `infinite` music conditioned via MIDI signal in *realtime* Take it out for a spin noww https://
github.com/Vaibhavs10/not
ebooks/blob/main/text_to_music_with_spectrogram_diffusion_and_diffusers.ipynb
… -
AudioLDM: Text-to-Sound Synthesis with Diffusion Models
By
–
AudioLDM contributed by @sanchitgandhi99, allows you to put in any arbitrary prompt and synthesise high-quality sounds from it. 🔊
— Vaibhav (VB) Srivastav (@reach_vb) 12 avril 2023
Try it out on colab 👉 https://t.co/EhFPgOmrij pic.twitter.com/Jn4nObthSJAudioLDM contributed by @sanchitgandhi99
, allows you to put in any arbitrary prompt and synthesise high-quality sounds from it. Try it out on colab https://
github.com/Vaibhavs10/not
ebooks/blob/main/text_to_sound_with_audioLDM_and_diffusers.ipynb
… -
Diffusers 0.15 brings AudioLDM and Spectrogram Diffusion models
By
–
Diffusers🧨 x Music🎶
— Vaibhav (VB) Srivastav (@reach_vb) 12 avril 2023
Taking diffusers beyond Image ⚡️
With the latest, Diffusers 0.15, we bring two powerful text-to-audio models with all bleeding edge optimisations 💥
1. @LiuHaohe et. al's AudioLDM 🔊
2. @GoogleMagenta's Spectrogram Diffusion 🎹 pic.twitter.com/daU8xW4qcXDiffusers x Music Taking diffusers beyond Image With the latest, Diffusers 0.15, we bring two powerful text-to-audio models with all bleeding edge optimisations 1. @LiuHaohe et. al's AudioLDM 2. @GoogleMagenta
's Spectrogram Diffusion -
PEFT Democratizes State-of-the-Art LLM Research and Development
By
–
Very cool to see PEFT playing such an important role in democratising SoTA research. Checkout the mode weights at: https://
huggingface.co/andreabac3/Fau
no-Italian-LLM-7B
… -
Anticipated Disruptions in Text-to-Speech and Sound Technology
By
–
What kind of disruptions do you anticipate in the world of Text to Speech/ Sound?
-
Maximizing Weights and Biases for Machine Learning Workflows
By
–
Lately I’ve been using @wandb more and more. I’ve been using configs to store hparams and ofc logging to track loss, lr, etc different runs. Still feel like I’ve not been maximising it. What are some underrated features you use everyday?
-
Integrating Data Processing Workflow with Hugging Face Ecosystem
By
–
Very cool! Congratulations Is the codebase already online somewhere? Would be cool to see if we can incorporate your data processing workflow using the @huggingface ecosystem!