meshed-memory-transformer
meshed-memory-transformer copied to clipboard
aimagelab
Meshed-Memory Transformer for Image Captioning. CVPR 2020
Hi @baraldilorenzo, I 'm trying to improve the speed of beam_search. When doing it, I found this function: `visual = self._expand_visual(visual, cur_beam_size, selected_beam)` in the iter function of beam_search.py Please...
Hi @marcellacornia, When I started my train, I got random outputs for about the first five epochs, I mean it generated words. Then, it produced nothing, and I had to...
Hi, Thanks for providing the implementation it really helps. I am getting the following error when trying to use Multiple GPUs with DataParallel. Please note the implementation works perfectly fine...
Hello, I would like to ask myself where is the description of the output generated by the model?
File "train.py", line 100, in train_scst tokenizer_pool = multiprocessing.Pool() File "/root/anaconda3/envs/m2/lib/python3.6/multiprocessing/context.py", line 119, in Pool context=self.get_context()) File "/root/anaconda3/envs/m2/lib/python3.6/multiprocessing/pool.py", line 174, in __init__ self._repopulate_pool() File "/root/anaconda3/envs/m2/lib/python3.6/multiprocessing/pool.py", line 239, in _repopulate_pool w.start()...
Hello, excuse me, I recently read the paper Meshed_memory, and I am a little confused about the data set coco_detections.hdf5 he used. Is the value extracted by the imgid_features key...
Hi everyone, While evaluating the model, I got the following error: `RuntimeError: CUDA error: CUBLAS_STATUS_EXECUTION_FAILED when calling `cublasLtMatmul( ltHandle, computeDesc.descriptor(), &alpha_val, mat1_ptr, Adesc.descriptor(), mat2_ptr, Bdesc.descriptor(), &beta_val, result_ptr, Cdesc.descriptor(), result_ptr, Cdesc.descriptor(),...
something was wrong when run in  
Hi! Thanks for the written paper and the availabe code. I have what may be a stupid question, but I didn't find a straight answer to it anywhere: When evaluating...
Thanks for this amazing work! The confusion about the paper when I read is that how to use the ensemble trick. In my opinion, ensembles of mutiple models means training...
