Antonio Scaiella

number of train caption is < 10000

7

Msr vtt dataset have 10000 videos and 20 captions for each video but in this implementation only a video-caption pair in train phase is considered. Therefore in total

[Usage]Error message when i run finetune_lora.sh. pip install flash-attn --no-build-isolation --no-cache-dir doesn't work

1

### Describe the issue Issue: Traceback (most recent call last): File "/home/sca/LLaVA/llava/train/train_mem.py", line 1, in from llava.train.train import train File "/home/sca/LLaVA/llava/__init__.py", line 1, in from .model import LlavaLlamaForCausalLM ImportError: cannot...

[BUG] UNSUPPORTED MEDIA TYPE for EmbedderOpenAICompatibleConfig

4

I created a flask server to use SentenceTransformer("multi-qa-MiniLM-L6-cos-v1") with EmbedderOpenAICompatibleConfig I can't handle requests with error "POST /v1/embeddings HTTP/1.1" 415 - In particular on the cat side I have the...

Do batch inference to speed up the process

1

Hi all, I have a large amount of images to process, is there a way to do inference using a batch and maybe a dataloader? Classifying one image at a...

How to use italian language?

1

It is possible use an italian style speaker? I've tried many prompt but all of this are in english style

### System Info / 系統信息 When I try to fine tune cogVLM in the demo script i set train and eval dataset but i not find where i can set...