Overview
Syllabus
0:00 Intro – Why compare Flux Context vs HighDream E1
0:54 Model origins & core architectures
1:44 Video goal: install, run, and benchmark both models
2:00 Setting up ComfyUI for the tests
2:28 Downloading HighDream E1 full & quantized models
3:34 Understanding quantization: Q2–Q8 trade‑offs
4:50 Where to place model files in ComfyUI folders
6:13 Updating ComfyUI to v0.3.45 for built‑in workflows
7:22 Loading Flux Context & HighDream E1 templates
8:55 Key technical differences between the two models
10:34 Natural‑language processing depth Llama 3.1, CLIP, etc.
12:10 Launching HighDream E1 in ComfyUI
13:51 Tuning sampler, CFG, and image influence settings
14:50 First render: performance & VRAM usage on RTX 3090
15:51 Result analysis: finger artifacts at default settings
17:07 Fixing artifacts by raising CFG and image influence
18:13 Side‑by‑side influence tests 0.5 → 6.0
19:59 Flux results vs HighDream: style transfer & prompt fidelity
21:15 Licensing, community, and best‑use scenarios
22:18 Wrap‑up, resource links, and call to subscribe
Taught by
Vladimir Chopine [GeekatPlay]