mit-han-lab/efficientvit issues and pull requests

#100 - Could someone explain to me the operation of efficient attention？

Issue - State: open - Opened by CHDyshli 5 months ago

#99 - Request for releasing the segcore module for segmentation training

Issue - State: open - Opened by carvychen 5 months ago

#98 - Request for Code Release: Condition-Aware Neural Network for Controlled Image Generation

Issue - State: open - Opened by MikePrivateCoding 5 months ago - 3 comments

#97 - A question about EfficientViTBlock and EfficientViTModule in the paper

Issue - State: open - Opened by xiawei20161308104 5 months ago

#96 - what's the version of tensorrt?

Issue - State: closed - Opened by freeman745 5 months ago - 2 comments

#95 - 1x1 group convs in LiteMLA

Issue - State: open - Opened by benjamindorra 6 months ago

#94 - when to release the training code of EfficientViT-SAM?

Issue - State: open - Opened by Dongshengjiang 6 months ago - 1 comment

#93 - The output from c++_trt infer is not same as the py_trt when i set two or more points to segment

Issue - State: closed - Opened by yangchengxin 6 months ago - 1 comment

#92 - The output from c++_trt Infer

Issue - State: closed - Opened by yangchengxin 6 months ago

#91 - Do we have tensorflow implementation?

Issue - State: open - Opened by HaochenZhang717 6 months ago

#90 - SAM demo updates

Pull Request - State: closed - Opened by ncstiles 6 months ago

#89 - SAM demo supporting multiple runtimes + models

Pull Request - State: closed - Opened by ncstiles 6 months ago - 1 comment

#88 - tensorRT multiple point inference bug fixed

Issue - State: open - Opened by ghm666 6 months ago

#87 - just ues nn.Identity.

Issue - State: open - Opened by jez-moxmo 6 months ago

#86 - 大佬出一个分割任务的训练代码吧

Issue - State: open - Opened by woshituobaye 6 months ago

#85 - What does this parameter mean

Issue - State: open - Opened by ghm666 6 months ago - 2 comments

#84 - how add this attention to yolov8?

Issue - State: open - Opened by ghostmiao5 6 months ago

#83 - EfficientViTSAM available on the Microscopy Imaging software Fiji

Issue - State: open - Opened by carlosuc3m 6 months ago - 1 comment

#82 - Hi @xuanlinli17, can you give an explicit example for multiple bboxes format?

Issue - State: open - Opened by MenSanYan 6 months ago - 4 comments

#81 - Need pretrained model for SA-1B

Issue - State: closed - Opened by silinsi 6 months ago - 1 comment

#80 - Rewrite Efficient-ViT-SAM TensorRT inference with torch2trt to make the code a lot simpler

Pull Request - State: closed - Opened by xuanlinli17 6 months ago - 1 comment

#79 - Add a web demo

Pull Request - State: closed - Opened by Gy920 6 months ago

#78 - EfficientViT-SAM TensorRT export error due to invalid node (temp solution by monkey patching original sam repo)

Issue - State: open - Opened by xuanlinli17 7 months ago

#77 - Fix tensorrt inference to take multiple boxes and points

Pull Request - State: closed - Opened by xuanlinli17 7 months ago - 3 comments

#76 - Questions about SegEvery mode(time)

Issue - State: open - Opened by Greywan 7 months ago

#75 - Cannot run inference on multiple boxes for TensorRT Efficient VIT SAM

Issue - State: open - Opened by aniket03 7 months ago - 2 comments

#74 - Where is your checkpoint?

Issue - State: closed - Opened by Kyle-RuidongLI 7 months ago - 1 comment

#73 - TensorRT Export error

Issue - State: closed - Opened by ghm666 7 months ago

#72 - Concerns regarding fairness of the results.

Issue - State: open - Opened by dazzle-me 7 months ago

#71 - Compatibility Issue between mpi and openmpi during installation

Issue - State: open - Opened by ManishSharma1609 7 months ago - 1 comment

#70 - Unable to convert tensorrt SAM model

Issue - State: open - Opened by benihime91 7 months ago - 3 comments

#69 - Is only point prompt and box prompt supported when using onnx and tensorRT deployment? How to segment anything

Issue - State: open - Opened by ghm666 7 months ago

#68 - 模型下载链接在哪呢？

Issue - State: closed - Opened by ghm666 7 months ago

#67 - Segment with multiple bounding boxes

Issue - State: open - Opened by pvtoan 7 months ago - 2 comments

#66 - Request for Guidance on Integrating LoRA with Custom Attention Implementation

Issue - State: open - Opened by rotem154154 7 months ago

#65 - Jetson AGX Orin package version

Issue - State: open - Opened by asd841018 7 months ago - 2 comments

#64 - Can not reproduce classification accuracy for efficientvit-B1 and B2 on Imagenet.

Issue - State: open - Opened by xiaoshingshing 7 months ago

#63 - how can we deploy E-Vit Sam in other platform

Issue - State: open - Opened by SoulProficiency 7 months ago - 3 comments

#62 - Mask input to EfficientSam models

Issue - State: open - Opened by carlosuc3m 7 months ago - 2 comments

#61 - Is there any pretrained model for sam-b0?

Issue - State: open - Opened by kimwin2 7 months ago - 1 comment

#60 - Zero Output with L2 Model On Demo Image

Issue - State: closed - Opened by FabianSchuetze 7 months ago - 4 comments

#59 - Question about FP16 training

Issue - State: closed - Opened by yvielcastillejos 8 months ago - 1 comment

#58 - How to draw lightweight linear attention MAP

Issue - State: open - Opened by alexyxuan 9 months ago

#57 - Make repo an installable package

Pull Request - State: closed - Opened by evgeniishch 9 months ago

#56 - The Results of EfficientViT-B0 on ADE20K Dataset？

Issue - State: open - Opened by SaulZhang 9 months ago

#55 - Could you provide EfficientViT-Det training code?

Issue - State: open - Opened by Qin27 9 months ago

#54 - Any plan for dinov2?

Issue - State: open - Opened by twmht 10 months ago - 2 comments

#53 - question about EfficientViT-L0-SAM model params

Issue - State: open - Opened by woshitiancai321 10 months ago

#52 - About TensorRT Latency Measure

Issue - State: closed - Opened by johnyang-nv 10 months ago - 4 comments

#51 - why set 1e-7 norm eps

Issue - State: open - Opened by seefun 10 months ago

#50 - ResBlock do not have residual shortcut

Issue - State: closed - Opened by seefun 10 months ago

#49 - when will you share the seg code？？

Issue - State: open - Opened by 1377534928 10 months ago - 1 comment

#48 - Running speed problem

Issue - State: open - Opened by Aries0217 11 months ago

#47 - The val result during train and test is different

Issue - State: closed - Opened by jiujiuwei 11 months ago - 1 comment

#46 - Please show train cls model yaml of l2 r288 and l2 r384, Thanks

Issue - State: closed - Opened by jiujiuwei 11 months ago - 1 comment

#45 - EfficientViT code for Super-resolution !!!

Issue - State: open - Opened by Longyuu-jjp 11 months ago

#44 - mIoU = {(interaction.sum / union.sum).cpu().mean().item() * 100:.3f}") ZeroDivisionError: division by zero

Issue - State: open - Opened by AbhaSinghPathania 11 months ago - 2 comments

#43 - output the segmentation mask

Issue - State: open - Opened by xwhkkk 11 months ago

#42 - Hyper-parameter to train seg model

Issue - State: open - Opened by seefun 11 months ago

#41 - The evaluation of imagenet is not right

Issue - State: open - Opened by RitchieAlpha 11 months ago - 1 comment

#40 - What is the purpose of unset bn? How is its size calculated?

Issue - State: open - Opened by sanbuphy 11 months ago

#39 - Can i train efficientvit on a single A100 or V100?

Issue - State: open - Opened by sanbuphy 11 months ago - 1 comment

#38 - how to train cityscapes

Issue - State: open - Opened by d710055071 11 months ago - 2 comments

#37 - Cityscapes dataset train

Issue - State: closed - Opened by d710055071 11 months ago

#36 - what's the different between cvpr2023 efficientvit and iccv2023 efficientvit?

Issue - State: closed - Opened by sanbuphy 11 months ago - 6 comments

#35 - Segmentation training code

Issue - State: closed - Opened by d710055071 12 months ago - 1 comment

#34 - Using weights from l3 high res in timm

Issue - State: open - Opened by Kaschi14 12 months ago

#33 - Does this repo allows CPU inference too?

Issue - State: closed - Opened by AlessandroMondin00 12 months ago - 2 comments

#32 - Fintune on my custom dataset

Issue - State: open - Opened by AryaAftab 12 months ago - 1 comment

#31 - Why padding in attention?

Issue - State: open - Opened by TsingWei 12 months ago - 4 comments

#30 - Inference at higher res

Issue - State: open - Opened by bhack 12 months ago

#29 - training cost

Issue - State: open - Opened by JunMa11 12 months ago - 1 comment

#28 - Concerning Discrepancies on Reported Model Accuracy !!!!!!!!!! (please see and correct)

Issue - State: closed - Opened by achen46 12 months ago - 4 comments

#27 - Support export efficientvit-sam models to ONNX

Issue - State: open - Opened by CVHub520 almost 1 year ago - 2 comments

#26 - About Equation in paper

Issue - State: closed - Opened by tarv33 about 1 year ago

#25 - SAM data

Issue - State: closed - Opened by daeyun about 1 year ago

#24 - Export Onnx Segmentation Model

Issue - State: open - Opened by chuangyuan089 about 1 year ago - 2 comments

#23 - The inference speed is measured with pytorch or tensorRT model??

Issue - State: open - Opened by ziming-liu about 1 year ago - 1 comment

#22 - An error occurred with ade20k or cityscapes using eval_seg_model.py

Issue - State: open - Opened by liaochuanlin about 1 year ago - 2 comments

#21 - Image segmentation training script？

Issue - State: open - Opened by liaochuanlin about 1 year ago

#20 - Some question for segmentation settings

Issue - State: open - Opened by nizhenliang about 1 year ago

#19 - Where should I download the dataset

Issue - State: closed - Opened by BolunDai0216 about 1 year ago - 1 comment

#18 - Training data for EfficientViT-L2

Issue - State: closed - Opened by P3ngLiu about 1 year ago - 1 comment

#17 - I would like to know when we can obtain the code for SAM

Issue - State: closed - Opened by sanbuphy about 1 year ago - 1 comment

#16 - EfficientViT L series

Issue - State: closed - Opened by bhack about 1 year ago - 2 comments

#15 - Numeric stability w/ AMP

Issue - State: open - Opened by rwightman about 1 year ago - 1 comment

#14 - EfficientViT b0 pretrained in ImageNet

Issue - State: closed - Opened by seefun about 1 year ago - 2 comments

#13 - import network error

Issue - State: closed - Opened by Fridayfairy about 1 year ago - 1 comment

#12 - MSA只有5x5卷积算子

Issue - State: closed - Opened by violet2422 about 1 year ago - 1 comment

#11 - How to train my model in mutil GPU?

Issue - State: closed - Opened by yjtlab about 1 year ago - 1 comment

#10 - Image segmentation training script

Issue - State: open - Opened by IannoIITR about 1 year ago - 2 comments

#9 - EfficientVit-B0 ImageNet Weights

Issue - State: closed - Opened by MarkWijkhuizen about 1 year ago - 2 comments

#8 - Training And Fine Tuning Configuration

Issue - State: closed - Opened by MarkWijkhuizen over 1 year ago - 1 comment

#7 - Replace yolov7 backbone, why do Nans and zeros occur when training under torch.cuda.amp?

Issue - State: closed - Opened by liiicon over 1 year ago - 2 comments

#6 - 大佬能不能提供一个训练的代码呀

Issue - State: closed - Opened by linjie111 over 1 year ago - 1 comment

#5 - training, the gradient disappears

Issue - State: open - Opened by mrswang1 over 1 year ago - 2 comments

#4 - The speed of training on GPU？

Issue - State: open - Opened by LovingThresh over 1 year ago

#3 - Tensorflow Lite models

Issue - State: closed - Opened by gaetanbahl over 1 year ago - 1 comment

#2 - About training

Issue - State: closed - Opened by Yonghao-Tan over 1 year ago - 3 comments

#1 - How should the hyperparameters of cifar10 be set? I have tried many times and can only reach 79%.

Issue - State: open - Opened by axedge1983 over 1 year ago - 2 comments

GitHub / mit-han-lab/efficientvit issues and pull requests