-
As the collaborators are in different cities, we really like each other, and it would be helpful to share in-person conversations we've had and not documented yet(we're human, after all), I think it w…
-
🙂🙏 感谢开源！
我用自己的数据训练之后效果还差了，帮忙看看什么问题呢，感谢先。
**1. 训练数据**
我的数据是一行一行的图片，然后合成了一张，多行（2~10行随机），共有1万张合成图片，图片是灰度图。
` doesn't return any messages within the con…
-
### Question
I have seen that is now possible to do multiple images inference.
Has anyone tried to do fine-tunning using multiple-images?
-
steps of demo:
1. paste contents of https://meta-introspector.github.io/meta-meme/
2. https://github.com/meta-introspector/meta-meme/wiki/Main
-
I am very interested in your work, so I tried to follow your experiments. When I was preparing the dataset, I found that the downloaded AitW dataset file was in TensorFlow format, but the json file yo…
-
### Describe the issue
Issue: We run into an indexing error when we try to finetune our LLaVA on our custom dataset (this LLaVA has previously been pretrained and finetuned on LLaVAR)
Command:
…
-
Currently, for a particular action, the model replies with the same response.
Ex: if action is ```utter_iamabot``` the response is always ```I am a bot, powered by Rasa.```
It would be nice to have…
-
[Toward Conversational Query Reformulation](https://webis.de/downloads/publications/papers/kiesel_2021d.pdf)
=========
## Contribution summary
- Kiesel et al. analyzed the query reformulation in co…