File size: 2,506 Bytes
5c7d905 d698555 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 |
## Setup 2 – A2E (Audio-to-Expression) Lip-Sync Generation
Generates videos using the A2E lip-sync pipeline with audio-driven facial expressions.
### Configuration Variables
#### **CAPTION_STYLE**
Defines the captioning/text style applied to generated videos.
**Available values:**
- `random`
- `caption_1`
- `caption_2`
- `caption_3`
- `caption_4`
- `caption_style_on_screen_text`
#### **CONTENT_STRATEGY_FILE**
CSV file containing content strategy entries.
- Must be placed in the **`data/`** directory
- Example:
```
CONTENT_STRATEGY_FILE="content_strategies_100_300.csv"
```
- Full path: `data/content_strategies_100_300.csv`
#### **IS_A2E_LIP_SYNC**
Enables A2E lip-sync generation.
```
IS_A2E_LIP_SYNC=true
```
**Required** for all A2E runs.
### Running the Pipeline
#### **Local Execution**
```bash
python src/process_csv.py
```
The script reads the configured `CAPTION_STYLE`, `CONTENT_STRATEGY_FILE`, and `IS_A2E_LIP_SYNC` and begins processing.
#### **GitHub Actions Execution**
The workflow triggers when a tag beginning with `run_wip` is pushed.
**Single-job run:**
```bash
git tag run_wip0
git push origin run_wip0
```
**Parallel runs:**
Format: `run_wip_<jobIndex>_<totalJobs>`
Example for **job 0 out of 10 parallel jobs**:
```bash
git tag run_wip_0_10
git push origin run_wip_0_10
```
This ensures no two parallel jobs process the same CSV entries.
### Important Notes
- Ensure all environment variables are set before running
- A progress file is **automatically created** after each successful generation
- **Do not manually edit** the progress file
- Parallel execution relies on progress tracking and job index logic
---
## Common Elements Across All Setups
### Progress Tracking
- Progress files are stored in the `data/` directory
- Automatically created/updated after successful generation
- Used to prevent duplicate processing in parallel runs
- Should never be manually modified
### Parallel Execution
- Uses tag format: `run_wip_<jobIndex>_<totalJobs>`
- Job index starts at 0
- Each job processes a unique section of the CSV
- Safe for concurrent execution without collisions
### Workflow Trigger
All setups use the same GitHub Actions trigger: tags beginning with `run_wip`
### Environment Variable Configuration
**For local execution:** Set variables in a `.env` file in the project root.
**For GitHub Actions:** Set variables in the GitHub repository's **Variables** section (Settings → Secrets and variables → Actions → Variables). |