-
### Cautions:
**Before starting the task, please refer to [Add data of ML-YouTube-Courses](https://github.com/orgs/ocademy-ai/projects/3/views/1?filterQuery=label%3Adata&pane=issue&itemId=36101499)…
-
I want to train llava_next_video on 2 different dataset together, 1 is video only and 1 is image only.
I have set up the json in the below format:
```
{
"system_prompt": "Explain wh…
-
### Is there an existing issue for this?
- [X] I have searched the existing issues and checked the recent builds/commits
### What happened?
After inputting an image and clicking 'Interrogate CLIP' …
-
### Your current environment
```text
Collecting environment information...
PyTorch version: 2.3.0+cu121
Is debug build: False
CUDA used to build PyTorch: 12.1
ROCM used to build PyTorch: N/A
…
-
### Your current environment
```text
Collecting environment information...
PyTorch version: 2.3.0+cu121
Is debug build: False
CUDA used to build PyTorch: 12.1
ROCM used to build PyTorch: N/A
…
-
I found the comment in CLIPEncoder
https://github.com/huggingface/transformers/blob/e65502951593a76844e872fee9c56b805598538a/src/transformers/models/clip/modeling_clip.py#L592-L596
mismatch with…
-
Hi, I'm trying to make compatible a Clip model using neuron-distributed (because I'm gonna continue with a multimodal after it)
Currently in my notebook, insidea inf2.xlarge ubuntu 22, I have:
…
-
Hi, I am trying to Finetune a VLM, Phi-3-vision specifically. Looking at the example with LLaVA I created the following collator for it:
```
class VLMDataCollator:
def __init__(self, processo…
-
## 🚀 Feature
Modularize `nn.MultiHeadAttention` / `nn.functional.multi_head_attention_forward()` and generalize the `nn.Transformer` modules so that:
1. Novel attention functions can be use…
-
**Describe the bug**
My use case is when tables have an _ID_ column that are intergers, but should really just be treated as a categorical variable. When setting `infer_dtypes` for the profile repo…