Implementation:Zai org CogVideo SAT Read From CLI File

Attribute	Value
Implementation Name	SAT Read From CLI File
Workflow	SAT Video Generation
Step	3 of 5
Type	API Doc
Source File	sat/sample_video.py:L23-41
Repository	zai-org/CogVideo
Last Updated	2026-02-10 00:00 GMT

Overview

Implementation of two prompt input functions for the SAT video generation pipeline: read_from_cli for interactive single-prompt input and read_from_file for batch file-based input with distributed worker sharding.

Description

The module provides two generator functions:

• read_from_cli(): Uses Python's built-in input() to interactively collect prompts from the user. Yields (text, count) tuples with an incrementing counter.
• read_from_file(p, rank, world_size): Reads lines from a text file and yields prompts assigned to the current worker based on modular distribution. Line index i is processed by worker i mod world_size == rank.

Both functions return Python generators that yield (text, count) tuples, enabling lazy evaluation and memory-efficient processing of large prompt files.

Usage

# Interactive mode
for text, count in read_from_cli():
    generate_video(text, count)

# Batch mode (distributed)
for text, count in read_from_file("prompts.txt", rank=0, world_size=4):
    generate_video(text, count)

Code Reference

Source Location

File	Lines	Description
sat/sample_video.py	L23-41	read_from_cli and read_from_file functions

Signature

def read_from_cli() -> Generator[Tuple[str, int], None, None]:
    """Interactive prompt input. Yields (text, count) tuples."""

def read_from_file(p: str, rank: int = 0, world_size: int = 1) -> Generator[Tuple[str, int], None, None]:
    """File-based prompt input with distributed sharding."""

Import

from sample_video import read_from_cli, read_from_file

I/O Contract

Inputs

read_from_cli

Parameter	Type	Default	Description
(none)	--	--	Reads from standard input interactively

read_from_file

Parameter	Type	Default	Description
p	str	Required	Path to the prompt text file
rank	int	0	Current worker rank for distributed sharding
world_size	int	1	Total number of workers for distributed sharding

Outputs

Output	Type	Description
Yielded tuples	Generator[Tuple[str, int], None, None]	Each yield produces (text, count) where text is the prompt string and count is the sequential index

Usage Examples

Example 1: Interactive CLI input

from sample_video import read_from_cli

for text, count in read_from_cli():
    print(f"Generating video {count} for prompt: {text}")
    # User types: "A cat playing piano"
    # Yields: ("A cat playing piano", 0)

Example 2: File-based batch input

from sample_video import read_from_file

# prompts.txt contains one prompt per line
for text, count in read_from_file("prompts.txt"):
    print(f"Generating video {count} for prompt: {text}")

Example 3: Distributed file input (4 GPUs)

from sample_video import read_from_file

# Worker 0 of 4 processes lines 0, 4, 8, ...
for text, count in read_from_file("prompts.txt", rank=0, world_size=4):
    generate_video(text, count)

Example 4: Image-to-video prompt format

# In prompts.txt for I2V:
# A cat playing piano@@/data/images/cat.jpg
# The "@@" separator splits text and image path

Related Pages

• Principle:Zai_org_CogVideo_SAT_Prompt_Input -- Principle governing prompt input modes
• Environment:Zai_org_CogVideo_SAT_Framework_Environment
• Zai_org_CogVideo_SAT_Get_Model_Load_Checkpoint -- Previous step: model loading
• Zai_org_CogVideo_SAT_Diffusion_Sample -- Next step: sampling with the prompt text
• Zai_org_CogVideo_SAT_Inference_Get_Args -- Configuration that selects CLI vs file mode