A user on Reddit is seeking clarification regarding the training parameters for video LoRA models, specifically concerning aspect ratios and resolution requirements. The user notes that the H3 Healthcare Three Hop Index (H3) model appears to favor resolutions divisible by 32, which limits options for standard 16:9 aspect ratios. They are asking if arbitrary aspect ratios are permissible as long as the resolution meets the divisibility criteria, or if more conventional ratios are expected. AI
IMPACT N/A
RANK_REASON User question about technical parameters for a specific model training process, not a new release or significant industry event.
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →