Search results

Question about DITTO model benchmarks mentioned in new paper about the SAMA Model (Open Sourced Instruction-Guided Video Editing model based on Wan2.1)

Status: Open.

#35 In EzioBy/Ditto;

· CCpt5 opened on Mar 25, 2026

support multi-cards training?

Status: Open.

#34 In EzioBy/Ditto;

· greasebig opened on Mar 3, 2026

can't find the code for annealing!

Status: Open.

#33 In EzioBy/Ditto;

· holdrain opened on Feb 21, 2026

global模式是否包含sim2real的能力？

Status: Open.

#32 In EzioBy/Ditto;

· zhaojigang opened on Feb 15, 2026

Sim2Real Training Details

Status: Open.

#31 In EzioBy/Ditto;

· MaximClouser opened on Feb 10, 2026

Inference on 1.3b Model

Status: Open.

#30 In EzioBy/Ditto;

· akhilgvss opened on Jan 30, 2026

curriculum learning strategy

Status: Open.

#29 In EzioBy/Ditto;

· carpedkm opened on Jan 14, 2026

Restore Old Video

Status: Open.

#28 In EzioBy/Ditto;

· tdashmike opened on Dec 20, 2025

It looks like you're dividing CSVs based only on instructions, and the generated videos haven't undergone further filtering or categorization

Status: Open.

#27 In EzioBy/Ditto;

· greasebig opened on Dec 3, 2025

关于数据生成流程的一些问题

Status: Open.

#26 In EzioBy/Ditto;

· JishunLiu opened on Nov 20, 2025

VLM instruct prompt

Status: Open.

#25 In EzioBy/Ditto;

· shinetzh opened on Nov 18, 2025

Did you finetune VACE to generate Ditto-1M dataset? How you finetuned it？

Status: Open.

#24 In EzioBy/Ditto;

· greasebig opened on Nov 14, 2025