Issues
- 0
- 3
dino v2 num paches set to 256
#19 opened by yanbai1993 - 8
How to reproduce the result of Figure 6?
#3 opened by HenryHZY - 1
Training Time details
#22 opened by HashmatShadab - 3
instruction fine-tuning failed.
#6 opened by shengyuwoo - 3
- 1
LLaVA-1.5 stage2 Traing batchsize
#13 opened by Z-MU-Z - 0
Implementation Details of Additive-MoF
#23 opened by Leeinsu1 - 0
- 0
evaluate on the MMVP
#21 opened by jiaxi0030 - 2
CLIP-blind pairs
#10 opened by mahtabbigverdi - 0
COMM (which has already been on Arxiv on 2023.10 https://arxiv.org/pdf/2310.08825.pdf) already proposed to merge the features of CLIP and DINOv2 to realize MLLM, maybe this paper should cite this reference.
#18 opened by YuchenLiu98 - 0
- 0
pretrain_dino_mm_mlp_adapter
#16 opened by demo1shining - 0
Unable to load the provided pretrained
#15 opened by sunfanyunn - 0
Correction Needed for Incorrect Answers in MMVP Benchmark Questions 279 and 280
#14 opened by MajorDavidZhang - 3
- 1
- 0
GPT4-V Prompt Used in Evaluation
#11 opened by Nyandwi - 1
leaked api key
#7 opened by lucasrodes - 1
Answer sheet of LLMs on MMVP?
#9 opened by minhlong94 - 1
How to evaluate the DFN model?
#4 opened by DesertsP - 5
下载模型的是否需要放在具有特殊名字的文件夹里(evaluate时)
#2 opened by Pro-flynn - 1
bug in evaluate scripts?
#1 opened by LightDXY