-
我看这个例子中的questions的形式类似于zero-shot,请问我要做上下文学习有没有合适的例子呢?也就是我先给模型在之前的对话中提供多张图片并且分别告诉他们图片的含义,之后再让他推理另外的图片。这个功能怎么实现呢?
I think the form of the questions in this example is similar to zero-shot. Is there a…
-
### Question
我很好奇,是否能把MLLM的LLM分离出来,还能和之前一样和其文本对话。同时此时的LLM比起之前是否能力下降。
-
-
Hi Hako,
I must remind you that your wonderful work has been plagiarized by @YangLing0818, and he even wrote a paper and published it on arXiv. Here is the paper link: https://arxiv.org/abs/2401.117…
-
I'm just getting started with MLLM and I want to see the full text output of the model and not just some ratings. I'm getting some weird output, so I'm not sure if the decoding code I'm using is corre…
-
There is an error below.
```
2024-03-26 12:57:36.254249: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin…
-
### Reminder
- [X] I have read the README and searched the existing issues.
### System Info
I used this #4136 PR to train LLaVA 7B and Idefics2 Chatty.
The training was done on 4 T4 GPUs using DS…
-
### Reminder
- [X] I have read the README and searched the existing issues.
### System Info
```
- `llamafactory` version: 0.8.4.dev0
- Platform: Linux-6.6.13-1-lts-x86_64-with-glibc2.31
- …
-
Dear authors,
Thank you for this wonderful paper! I reproduced your Figure 2 (attention map from InstructBLIP) and got the following result. I did not notice that an outstanding pattern that highli…
-
In this [Convert Vocabulary](https://ubiquitouslearning.github.io/mllm_website/customization/convert_vocabulary/) page, the shell script `python vocab.py --input_file=tokenizer.model --output_file=voc…