Update README.md
Browse files
README.md
CHANGED
|
@@ -2,4 +2,22 @@
|
|
| 2 |
library_name: diffusers
|
| 3 |
tags:
|
| 4 |
- image-to-video
|
| 5 |
-
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 2 |
library_name: diffusers
|
| 3 |
tags:
|
| 4 |
- image-to-video
|
| 5 |
+
---
|
| 6 |
+
|
| 7 |
+
Working remote so this is taking longer than I'd like.
|
| 8 |
+
Uploading four versions. Minimal testing done as I'm away for work.
|
| 9 |
+
Bob = Made with Bob's upgraded node
|
| 10 |
+
DMD = DMD lora merged in at 1.0 strength. Saves space and should speed up your generation speed.
|
| 11 |
+
If you use the non DMD models you will need to load the DMD lora as you normally would at strength 1.0
|
| 12 |
+
|
| 13 |
+
Check 10Eros discord for sigmas. I used 1.2 sigmas and it was fine however.
|
| 14 |
+
|
| 15 |
+
10Eros_v1.4_Bob_INT8_Convrot.safetensors
|
| 16 |
+
10Eros_v1.4_Bob_DMD_INT8_Convrot.safetensors <- from testing this is the fastest for me on a 3090 and 64GB ram. Most people will want this one.
|
| 17 |
+
10Eros_v1.4_int8_convrot.safetensors
|
| 18 |
+
10Eros_v1.4_DMD_int8_convrot.safetensors <- this one was expected to be faster but not proving to be. Not sure why. Just testing it for fun.
|
| 19 |
+
|
| 20 |
+
Resources:
|
| 21 |
+
https://github.com/BobJohnson24/ComfyUI-INT8-Fast/tree/main
|
| 22 |
+
https://huggingface.co/tech77/int8
|
| 23 |
+
https://github.com/silveroxides/convert_to_quant
|