Overriding the new default of 'sdpa' - #108

Open

ParthaEth wants to merge 52 commits into

MeiGen-AI:mainfrom

Commits on Jul 23, 2025

Overriding the newdefault of 'sdpa'
VRWarg
committed
34446ab

Commits on Jan 19, 2026

Building CLI interface so it can be triggered from hubspot
Partha Ghosh
committed
61ed007

Commits on Jan 20, 2026

real number of steps
Partha Ghosh
committed
705e428

Commits on Jan 22, 2026

So the checkpoint is not picked from the deve version
Partha Ghosh
committed
915ccbc

Commits on Jan 27, 2026

minor
Partha Ghosh
committed
d018b00

Commits on Feb 5, 2026

more informative error
Partha Ghosh
committed
331c6a5
error correction in func name
Partha Ghosh
committed
00e31ea

Commits on Feb 6, 2026

Creation speed param for ETA reporting
Partha Ghosh
committed
bed9919
Building a little buffer
Partha Ghosh
committed
9b2542f
Handling too short and verly long videos
Partha Ghosh
committed
dd58810

Commits on Feb 7, 2026

calibrating the vid gen time
Partha Ghosh
committed
9a0a0ea
calibrating better
Partha Ghosh
committed
093ae1f

Commits on Feb 18, 2026

CLI now takes the avatar image and specified voice
Partha Ghosh
committed
88be606
Add attention_blackwell module with fallback mechanisms for flash attention (not currently used, but may be soon)
felixludos
committed
d742dd3
Keeping max frames rather high so things get included
Partha Ghosh
committed
72c96bb

Commits on Feb 19, 2026

Enhance avatar asset selection and input payload construction in CLI
felixludos
committed
6736154
Update .gitignore to include .vscode directory
felixludos
committed
07c9153
Add base TTS template and enhance avatar handling in CLI
felixludos
committed
17017e7
Update input payload documentation to clarify avatar name usage
felixludos
committed
78e0170

Commits on Feb 23, 2026

Add mini.sh script for generating multitalk with configurable flash attention
felixludos
committed
20a7804

Commits on Feb 26, 2026

Add custom.sh script for generating multitalk with specified configurations
felixludos
committed
1b09ac4
Add new audio files and update JSON configuration for avatar processing; modify custom.sh for audio save directory
felixludos
committed
2a03256

Commits on Mar 9, 2026

Update mini.sh to include SCRIPT_TEXT in JSON output and modify input JSON path
felixludos
committed
a71cecd

Commits on Mar 11, 2026

Refactor JSON formatting in base_tts_template.json and ensure consistent structure in cli.py and config.py
Felix
committed
8a6e3bf

Commits on Mar 16, 2026

Add mini-audio.sh script for generating audio JSON configuration and invoking multitalk generation
Felix
committed
04b5210

Commits on May 6, 2026

So health of this repo can be checked
Partha Ghosh
committed
806c834

Commits on May 7, 2026

Changed arbitrary max frames
Partha Ghosh
committed
011f974
added a check so things jsut don't truncate silently
Partha Ghosh
committed
2564fe3

Commits on Jun 23, 2026

Update image path and output path in mini.sh for consistency
Felix
committed
508f497
Add *.mp4 to .gitignore to exclude video files from version control
Felix
committed
63045cb
Add *.mp4 to .gitignore to prevent tracking of video files
Felix
committed
a1153f1
Add unit tests for multitalk CLI, healthcheck, and worker node functionality
Felix
committed
9e1236f
Update IMAGE_PATH in mini.sh to point to new avatar location
Felix
committed
d6674b4
Merge pull request #1 from ParthaEth/feat/unit-tests
felixludos
authored
1998b50
Merge branch 'dev' into feat/google-tts
Felix
committed
6e04d7d
Add audio duration estimation and update input payload handling for audio files
Felix
committed
2de9dc1
Add audio URL handling and extension guessing to CLI
Felix
committed
6aad314

Commits on Jun 29, 2026

Add unit and worker-node tests for multitalk healthcheck logic and dependencies
felix
committed
5d3b10f
Merge pull request #2 from ParthaEth/feat/google-tts
felixludos
authored
dee239e

Commits on Jul 5, 2026

Add .venv to .gitignore to exclude virtual environment files
Felix
committed
fcaabc8

Commits on Jul 8, 2026

Merge pull request #3 from ParthaEth/dev
felixludos
authored
c36b675

Commits on Jul 15, 2026

Picking up the samples per setp from env var
Partha Ghosh
committed
14aeb6a
Optimizing the CPU ram while the video is written
Partha Ghosh
committed
f336d9f

Commits on Jul 18, 2026

speeding up multitalk 4X
Partha Ghosh
committed
b22f300

Commits on Jul 19, 2026

Add elena.png image and update IMAGE_PATH in mini.sh; include pytest and soundfile in requirements.txt
felix
committed
82abec9
Update output paths in mini.sh to use temporary directory
felix
committed
aad99ea

Commits on Jul 20, 2026

Tweeking for lip sync quality
Partha Ghosh
committed
d902b9c
Merge branch 'dev' of github.com:ParthaEth/MultiTalk into dev
Partha Ghosh
committed
cb7c229
Reducing steps for speed
Partha Ghosh
committed
58f0b07
Add audio files: audio.mp3 and audio.wav to examples directory
felix
committed
a4c1080

Commits on Jul 22, 2026

enable mini audio demo
maity-co
committed
5c57a70
Merge branch 'dev'
maity-co
committed
72338ce