Tried to double number of frames in CogVideo -- got chaos

[Zafflower]

Hi, Thanks for this marvelous video maker. I installed it in ComfyUI to work with Flux1-dev. I made this workflow, "flux-cogvideo-text-to-video-v32_plus_flux1-dev.json", by modifying the "flux-cogvideo-text-to-video-v3.json" from Stable Diffusion Art website to accept flux1-dev.safetensors diffusion model instead of flux1-dev-fp8.safetensors checkpoint. I substituted in nodes from "flux1-dev-regular.json" -- also from the Stable Diffusion Art website.

When I ran this workflow with 46 frames and frame rate of 8, using the prompt "scattered autumn leaves blowing in a whirlwind, sunny day," I got a nice video of autumn leaves blowing in the wind, but when I increased the frame rate to 98, I got chaos -- just a flashing textured pattern. I kept all the other settings the same between the two videos. Is there a way to increase the lengh of a CogVideo video by increasing the number of frames without getting a bad video? Are there other settings I can change to make this work? Any help will be greatly appreciated.
Zaffer

link to CogVideo that works with Flux1-dev:
flux-cogvideo-text-to-video-v32_plus_flux1-dev.json

Results images:

Link to workflows at Stable Diffusion Art
https://stable-diffusion-art.com/flux-cogvideo-text-to-video/
https://stable-diffusion-art.com/flux-comfyui/