Implementation:Zai org CogVideo SAT Inference Get Args

Attribute	Value
Implementation Name	SAT Inference Get Args
Workflow	SAT Video Generation
Step	1 of 5
Type	API Doc
Source File	sat/sample_video.py:L285-301, sat/arguments.py:L58-185
Repository	zai-org/CogVideo
Last Updated	2026-02-10 00:00 GMT

Overview

Implementation of the argument parsing system for SAT-based CogVideoX inference. The get_args function extends SAT's base argument parser with inference-specific parameters for controlling video generation resolution, frame count, FPS, prompt input method, and image-to-video mode.

Description

The get_args function builds an argument parser that combines:

1. SAT framework's base arguments (model architecture, distributed settings, checkpoint path)
2. YAML configuration loading via --base
3. Inference-specific arguments added by the sample_video module

The function returns a fully populated argparse.Namespace object that is passed to all subsequent pipeline stages including model loading, prompt reading, and sampling.

Usage

from arguments import get_args

# Parse from sys.argv
args = get_args()

# Or parse from explicit list
args = get_args(["--base", "configs/cogvideox_2b.yaml",
                 "--sampling-image-size", "768", "1360",
                 "--sampling-num-frames", "32"])

Code Reference

Source Location

File	Lines	Description
sat/sample_video.py	L285-301	Inference argument additions
sat/arguments.py	L58-185	Base SAT argument parser

Signature

def get_args(args_list=None) -> argparse.Namespace:
    # Key inference-specific params:
    # --base: YAML config files
    # --input-type: "cli" or "txt"
    # --input-file: prompt file path
    # --sampling-image-size: [768, 1360]
    # --sampling-num-frames: 32
    # --sampling-fps: 8
    # --image2video: bool flag

Import

from arguments import get_args

I/O Contract

Inputs

Parameter	Type	Default	Description
args_list	Optional[List[str]]	None (uses sys.argv)	Explicit argument list for testing
--base	List[str]	Required	YAML config file paths
--input-type	str	"cli"	Prompt input method: "cli" or "txt"
--input-file	str	None	Path to prompt text file (required if input-type is "txt")
--sampling-image-size	List[int]	[768, 1360]	Output video resolution [height, width]
--sampling-num-frames	int	32	Number of frames to generate
--sampling-fps	int	8	Frames per second of output video
--image2video	bool	False	Enable image-to-video generation mode

Outputs

Output	Type	Description
Return value	argparse.Namespace	Fully populated argument namespace combining YAML config and CLI overrides

Usage Examples

Example 1: Text-to-video generation

python sat/sample_video.py \
    --base configs/cogvideox_2b.yaml \
    --sampling-image-size 768 1360 \
    --sampling-num-frames 32 \
    --sampling-fps 8 \
    --input-type cli

Example 2: Batch generation from file

python sat/sample_video.py \
    --base configs/cogvideox_5b.yaml \
    --sampling-image-size 480 720 \
    --sampling-num-frames 48 \
    --input-type txt \
    --input-file prompts.txt

Example 3: Image-to-video generation

python sat/sample_video.py \
    --base configs/cogvideox_5b_i2v.yaml \
    --sampling-image-size 480 720 \
    --sampling-num-frames 48 \
    --input-type cli \
    --image2video

Related Pages

• Principle:Zai_org_CogVideo_SAT_Inference_Configuration -- Principle governing inference configuration
• Environment:Zai_org_CogVideo_SAT_Framework_Environment
• Zai_org_CogVideo_SAT_Get_Model_Load_Checkpoint -- Next step: model loading using the parsed args
• Zai_org_CogVideo_SAT_Read_From_CLI_File -- Prompt reading controlled by --input-type