YAML Metadata Warning:empty or missing yaml metadata in repo card

Check out the documentation for more information.

LTX 10Eros Extracted LoRAs

This repository contains a collection of LoRAs extracted directly from the TenStrip/LTX2.3-10Eros checkpoint.

These LoRAs allow you to bring the unique style, details, and features of the 10Eros model into other base models without needing to download or load the full heavy checkpoint. Multiple versions are provided with varying degrees of optimization.

⚑ Which file should I download?

Short answer: Use LTX_10Eros_LoRA_fro99-avgrank77.safetensors!
It gives practically the exact same visual quality as the heaviest uncompressed version, but is perfectly optimized for everyday use, saving you VRAM and storage.


πŸ“‚ File Breakdown & Versions

πŸ’‘ General rule of thumb: The larger the file size, the closer and more accurate the LoRA is to the original 10Eros base checkpoint.

Here is a detailed breakdown of what each file does so you can pick the best one for your needs:

  • πŸ₯‡ LTX_10Eros_LoRA_fro99-avgrank77.safetensors (Highly Recommended) The best all-rounder. It retains almost perfect accuracy to the base checkpoint but is much better optimized for file size and VRAM.
  • πŸ—„οΈ LTX_10Eros_LoRA_r768.safetensors (The Raw Base) The heaviest, uncompressed LoRA with a massive rank (768). It contains 100% of the extracted info, but in practice, there is very little visual difference compared to fro99. Mostly here for archival purposes, maximum info extraction, and experiments.
  • βœ‚οΈ LTX_10Eros_LoRA_layercut-fro99-avgrank66.safetensors (Great Quality + Smaller Size) Same high accuracy as fro99, but I removed the low-impact layers to make the file even smaller and more efficient.
  • πŸͺΆ LTX_10Eros_LoRA_fro95-avgrank23.safetensors & LTX_10Eros_LoRA_layercut-fro95-avgrank27.safetensors (Lightweight) Heavily compressed versions. Because of their smaller size, they are a bit less precise compared to the base checkpoint, but they still work fine. Designed specifically for people who need to save storage space or have limited VRAM.

Credits to the creator of the original base checkpoint: TenStrip/LTX2.3-10Eros.

Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Providers NEW
This model isn't deployed by any Inference Provider. πŸ™‹ Ask for provider support