Nodes/ComfyUI-WepeNerd/Video Captioner (Advanced)
ComfyUI Node

Video Captioner (Advanced)

A ComfyUI node in WepeNerd/Local AI/Advanced with 18 inputs and 2 outputs.

By WepeNerd·Created 4 months ago·Updated 2 days ago· 0
Video Captioner (Advanced)
  • config
  • video
  • caption
  • info
instructionDescribe this video accurately, including subjects, actions, camera motion, setting, and meaningful changes over time.
caption_style
video_mode
sampling_mode
sample_frames12
sample_fps2.00
max_frames24
max_tokens512
temperature0.20
seed0
system_prompt_override
caption_prefix
banned_phrases
image_max_edge1024
jpeg_quality90
reasoning_effortnone
CategoryWepeNerd/Local AI/Advanced

Inputs (18)

NameTypeDefaultDescription
configGGUF_LLM_CONFIG
videoVIDEO
instructionSTRINGDescribe this video accurately, including subjects, actions, camera motion, setting, and meaningful changes over time.
caption_styleCOMBO5 options: dataset_natural, detailed_visual, short, motion_camera, custom
video_modeCOMBO3 options: auto, native_video, sampled_frames
sampling_modeCOMBO2 options: uniform, fixed_fps
sample_framesINT122–96
sample_fpsFLOAT2.000.01–120
max_framesINT242–96
max_tokensINT5121–4096
temperatureFLOAT0.200–2
seedINT00–18446744073709550000
system_prompt_overrideoptSTRING
caption_prefixoptSTRING
banned_phrasesoptSTRING
image_max_edgeoptINT102464–4096
jpeg_qualityoptINT901–100
reasoning_effortoptCOMBOnone5 options: default, none, low, medium, high

Outputs (2)

NameTypeDescription
captionSTRING
infoSTRING