ghaddarAbs

@wangxinyu0922 the experiments above are for fine-tuning I fed the examples at the sentence level, didn't concatenate the sentences to form a document. these are the main HP game-changer in...

Thank you @intersun for your response. (1) I will print the lr curve to see what happend. (2) Yes, we tried it multiple time and it always happen. (3) We...

Also, I get these logs just before loading the data (pretrain.py), I don't know if they are related with the issue we are getting. ``` [1,0]:Weights of BertEncoder not initialized...

One way to evaluate a LM is to calculate the perplexity score of a given sentence (how meaningful the sentence is). There is two popular train/dev/test datasets : PTB and...

In order to reproduce the conll score reported in BERT paper (92.4 bert-base and 92.8 bert-large) one trick is to apply a truecaser on article titles (all upper case sentences)...

@kermitt2 ... I used GPU with 32 GB for these experiments. To answer your 3 questions: - yes i am using CRF, it give slightly better results. - all experiments...

Okay, I did a quick inspection and my 5 cent fast solution is to change the pattern in below regex functions in rouge: https://github.com/google-research/google-research/blob/94ef1c5992057967305cef6cbdd94ab995191279/rouge/tokenize.py#L28 https://github.com/google-research/google-research/blob/94ef1c5992057967305cef6cbdd94ab995191279/rouge/tokenize.py#L32 to `[^a-z0-9\u0621-\u064a\ufb50-\ufdff\ufe70-\ufefc]+` in fact the...

same issue here. I think it's related to this problem #1047 , @acheong08 after you removed the proxy the gateway error disappeared but now it has become 5 times slower...

@tridao the problem exist for v2.0.3 for pytorch 2.3