Rohan Manzoor

encountered the same issue. was anyone able to find the solution?

here's what my logs are saying train_stage1.out: > max_epoch {2,} > warmup_ratio {0.06,} > drop_worst_after {2500,} > train_stage2.out: > lr {1e-5,} > max_epoch {3,} >

I am sharing these detailed log files [2_0.06_2500.log](https://github.com/OFA-Sys/OFA/files/9265675/2_0.06_2500.log) [{5,}_{0.06,}_{6000,}.log](https://github.com/OFA-Sys/OFA/files/9265676/5._.0.06._.6000.log) [{2,}_{0.06,}_{2500,}_el0.75_.log](https://github.com/OFA-Sys/OFA/files/9265677/2._.0.06._.2500._el0.75_.log) [{2,}_{0.06,}_{2500,}.log](https://github.com/OFA-Sys/OFA/files/9265678/2._.0.06._.2500.log)

@logicwong i was able to run train_caption_stage1.sh with this script but after an hour or so I got the following output  

@logicwong so i now changed the batch size to 2 and patch size to 200. Here's the output. 

I was able to build docker with CUDA using the following image @viktorpovazhuk @SSdevelop https://github.com/IDEA-Research/GroundingDINO/pull/307

I was able to build docker with CUDA using the following image @mvoodarla https://github.com/IDEA-Research/GroundingDINO/pull/307

hello any update not this?

Hi would really love to use langchain/langgraph's callback system directly as mentioned by @us . Is this feature underdevelopment or under consideration?