-
from llava.model.builder import load_pretrained_model
from llava.mm_utils import get_model_name_from_path, process_images, tokenizer_image_token
from llava.constants import IMAGE_TOKEN_INDEX, DEFAUL…
-
**The first question:**
https://github.com/Heidelberg-NLP/MM-SHAP/blob/00a66bf932315d88be1e33c5bf39b0d8dc507d55/mm-shap_clip_dataset.py#L65
I check with this code and try to apply it also on the…
-
### System Info
In current implementation of VLMs, the "_supports_sdpa" attribute checks and activates SDPA attention only for the language model. For example in [Llava](https://github.com/huggingf…
-
### Your current environment
```text
The output of `python collect_env.py`
Collecting environment information...
PyTorch version: 2.3.1+cu118
Is debug build: False
CUDA used to build PyTorch…
-
Gradio is fine for playing around but can you please add proper demo code like you've done for the other llava-next models?
-
### Describe the issue
Following links return the Argo Tunnel Error 1033:
- https://llava-next.lmms-lab.com/
- https://llava.hliu.cc/
-
Hello, now that the LLava and LLava-Next models are integrated in `transformers`, would it be possible to remove the dependency on the Llava repo (https://github.com/haotian-liu/LLaVA)?
This would …
-
-
we use the LLaVA-NeXT-Video-DPO (34B)
-
I anticipate there will be a lot of demand to train (and infer) the new open SOTA image model "Flux".
It's the top model on HF right now. It's a 12B diffusion transformer, which means it's too big t…