frankang
frankang
Protobuf is installed under the folder `/usr/local/protobuf`, I manually set the CMAKE_PREFIX_PATH environment variable in the `lightseq/inference/proto/CMakeLists.txt` file to include the above installation path. However I still get error during...
Hi, we find that GroBid cannot parse the inline formula without discarding the spacing, superscript and subscript information. Could you suggest a pathway to improve the accuracy on these scenarios?...
**Describe the bug** I trained a model from scratch on RTX 3090 cards (`torch.cuda.is_bf16_supported()` returns True) in bfloat16 mode, but the training speed turns out to be 1/3 of the...
In the make_del_knob function, when the size product (e_size * f_size) is smaller than the sample_size (20000 by default), the script ends up calculating the similarity score for all combinations...
``` wordsList = np.load('wordsList.npy') print('Loaded the word list!') wordsList = wordsList.tolist() #Originally loaded as numpy array wordsList = [word.decode('UTF-8') for word in wordsList] #Encode words as UTF-8 wordVectors = np.load('wordVectors.npy')...
Hi, I got the following error when loading llama-7B model (ported to huggingface), my server has 256GB RAM on board. Is there any option to reduce the RAM consumption? Thanks....
The provided model cannot correctly categorize some "vaguely" plotted Figures and Tables. In this case, the word in the Table region will be considered as normal Text, thus hinder the...
因为图片url一段时间后会过期，而messsage link不会，所以想看看能不能支持根据message link重新获取image url的功能。
Two changes: 1. `mask = mask.to(output[0].device).unsqueeze(-1)` -> `mask = mask.to(output[0]).unsqueeze(-1)`, to assure mask has the same device and data type as output[0] 2. add outbound error check. If all heads...
error message: "xml-p5/T/T49/T49n2034.xml" is invalid: Memory allocation failed, line 1599, column 111 (T49n2034.xml, line 1599)