子龙

Feature request

1

- 对于 `laravel` 内部使用的 `mysql` 和 `redis` client 是否可以较为优雅的替换为 `swoole` 的client

Any plan to support InstantID

1

[InstantID](https://instantid.github.io/) is a very popular project for ID consistency generation, would it be supported in DiffSynth?

Support for PhotoMaker-V2

6

### Feature Idea [PhotoMaker](https://github.com/TencentARC/PhotoMaker?tab=readme-ov-file) had version v2 now, according to readme: "If the ID fidelity is not enough for you, please try our [PhotoMaker V2](https://huggingface.co/spaces/TencentARC/PhotoMaker-V2) or [stylization application](https://huggingface.co/spaces/TencentARC/PhotoMaker-Style), you may...

Hi, I found that even models like Flux or SD3 had made a lot of progress in language understanding by using t5xxl, they still can't control specific attributes of an...

Is it possible to support concurrency in model container

10

https://github.com/ml-explore/mlx-swift-examples/blob/42b0d21a607b97c3b99d1e20626b7d9b14827b81/Libraries/MLXLMCommon/ModelContainer.swift#L9 As comment above mentioned, now it is single thread access only, but in some cases, concurrency is really useful to speed up. Thanks for reply

as [tech report](https://github.com/OpenBMB/MiniCPM/blob/main/report/MiniCPM_4_Technical_Report.pdf) mentioned, it is really fast. mlx version already existed: https://huggingface.co/mlx-community/MiniCPM4-8B-4bit

Really excellent work and where is the sample code for inference or training?

5

I found some imgs without .py or any kinds of weights to test

As [blog](https://www.open-moss.com/en/moss-ttsd/) say: > For the original audio, we first use an internal speaker diarization model to segment and annotate speech and speakers. Based on the pre-trained base model, the...