-
Hi!
Thanks for your nice work!
Recently, we proposed a jailbreaking algorithm against MLLMs named FigStep (https://github.com/ThuCCSLab/FigStep) which is very close to the "OCR image" in your pa…
-
我发现multiTurnT2I_app.py中request_mllm函数,在文件48行,但是无法调用,可否告知这个8080端口的服务怎么启用?
def request_mllm(server_url='http://0.0.0.0:8080',history_messages=[], question="画一个木制的鸟",image=""):
-
Thank you for your fun and practical work! Have you considered non-square input, such as horizontal or vertical segmentation only? How to train.cls_token and pos_embed in this case will also become a …
-
Nice work! When attempting to merge the detoxifier, I encountered a weight size mismatch error with the following details:
```
size mismatch for base_model.model.model.layers.25.self_attn.q_proj.lor…
-
Hi!
What is the difference between ggml and your project?
-
# Weekly GitHub Trending! (2024/08/12 ~ 2024/08/19)
## Python trending 8repo's
### [hacksider](https://github.com/hacksider) / [Deep-Live-Cam](https://github.com/hacksider/Deep-Live-Cam)
リアルタイムの顔交換と 1…
-
Hey,
I was going through the code trying to understand it, but this seems strange to me.
https://github.com/cambrian-mllm/cambrian/blob/94824231dda385e2d1d0aef23f75a3dc97bbf085/cambrian/model/v…
-
Why the Filtered DVQA has 1550K data more than the DVQA 775K data, the latter I suppose is unfiltered?
Also the situation on the CLEVR(Filtered CLEVR 350K same as CLEVR 350K)
-
Hello,
I am trying to change the LLM from `GPT-4o` to `GPT-4o-min` or `GPT-3.5-turbo`. When I run the command to create an agent, I encounter an issue where the provided API key is not valid for …
-
**Describe the bug**
What the bug is, and how to reproduce, better with screenshots(描述bug以及复现过程,最好有截图)
命令:CUDA_VISIBLE_DEVICES=0 swift infer --model_type glm4v-9b-chat --model_id_or_path ../MLLM/GLM…