Manuel
Hi all! I'm trying to initiate from mBERT checkpoint but it is missing the "bert/embeddings/LayerNorm/beta/adam_m" key in the list of variables (just like you described). I'm using TF=1.14 and have...
Hi @RyanHuangNLP, if you have found a solution for this problem, would you mind sharing it? :)
I think this is a duplicate from [this](https://github.com/adjidieng/ETM/issues/38)
I have the same problem. @zhaoLLL thanks for your reply but how do the `bow_X_tokens.mat` and `bow_X_counts.mat` map to these two TF-IDF matrices?
Since this repo doesn't seem to be curated anymore, I suggest you use another repo I just discovered: https://github.com/lffloyd/embedded-topic-model I was able to use ETM very easily with it.
Since this repo doesn't seem to be curated anymore, I suggest you use another repo I just discovered: https://github.com/lffloyd/embedded-topic-model I was able to use ETM very easily with it.
Since this repo doesn't seem to be curated anymore, I suggest you use another repo I just discovered: https://github.com/lffloyd/embedded-topic-model I was able to use ETM very easily with it.
Yes it can! You basically just have to use the `generate_stats.py` as indicated in the README to train word statistics for your corpus and then use your corpus name when...
I have the same request. Did you find a solution @kwon-radio ?
After emailing the authors, they told me that the function `get_theta` in the `run_NQTM.py` script does just that. I tried with the Stack Overflow example and it works!