CLeaf_873981784

Results 11 comments of CLeaf_873981784

Anomalib CLI Improvements - Update `metrics` and create `post_processing` section in the `config` file.

where are metrics calculated? in which python file

Is the PatchCore image-level anomaly score correctly calculated?

In official patchcore work, the performance is Patchcore(99.1,98.1). But your work get (98.0,98.0),which is lower than official. Why is that ? Is there any problem about the score calculation?

[Usage] KeyError: 'LlavaConfig' when evaluating with model_vqa.py

has anyone fixed it?

[Question] KeyError: 'LlavaConfig' of using cli

has anyone fixed it？

vision chat error

is the .avi video format supported?

Scripts to continue finetuning InternVL-chat-v1-5 on custom data

also waiting for finetune scripts for v1-5 model

different performance between online web demo and local model

> I suspect a problem with device_map='auto', are you now running this model distributed across multiple GPUs? yes

image_processor error in loading lora weights

any solution?

intervl 微调后lora合并推理和直接用原来demo推理结果不一致

do_sample=False 不然有随机性

How to use multiple GPUs for inference?

> Below is an example of running `AIDC-AI/Ovis1.5-Llama3-8B` on two GPUs: > > ```python > import torch > from PIL import Image > from transformers import AutoModelForCausalLM > > device_map...