search

RyanWangZf / Trial2Vec

Findings of EMNLP'22 | Trial2Vec: Zero-Shot Clinical Trial Document Similarity Search using Self-Supervision
MIT License
19 stars 4 forks source link

Issues about Local contrastive loss. #2

Closed freshjarman closed 1 year ago

freshjarman commented 1 year ago

Hi, I didn't find the relevant codes of Local contrastive loss. in the paper 3.3. Specifically, I don't know how to implement eq(6-8) (like, how to use SciSpacy and map one entity to its canonical name? I didn't find relevant codes in the project). Could u point it out, thanks~

RyanWangZf commented 1 year ago

Hi, thanks for your interest in our work. We do not include the part you mention in this version since it will introduce additional package dependency. We instead take a more straightforward case for training, and it undermines the performance in an acceptable range.

freshjarman commented 1 year ago

Thanks for your reply. Could u share the part with me? I am glad to see relevant codes of your interesting work. Thanks 😊

---Original--- From: @.> Date: Tue, Dec 6, 2022 23:37 PM To: @.>; Cc: @.**@.>; Subject: Re: [RyanWangZf/Trial2Vec] Issues about Local contrastive loss.(Issue #2)

Hi, thanks for your interest in our work. We do not include the part you mention in this version since it will introduce additional package dependency. We instead take a more straightforward case for training, and it undermines the performance in an acceptable range.

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.***>

freshjarman commented 1 year ago

Hi, thanks for your interest in our work. We do not include the part you mention in this version since it will introduce additional package dependency. We instead take a more straightforward case for training, and it undermines the performance in an acceptable range.

I look forward to seeing that version (relevant codes of Hierarchical contrastive learning) if you are free. Thanks for ur reply and interesting work which is of great help~ Here's my e-mail: wangtao98@seu.edu.cn 😊

RyanWangZf commented 1 year ago

Sent to your email.

asamadan commented 1 year ago

Hi Zifeng,

I was wondering about the same piece related to the local contrastive loss. really appreciate it if you could make that part of the code available. much appreciated! I have sent you a message to your university email in this regards

Thanks,

RyanWangZf commented 1 year ago

@asamadan

Has sent to your email.

altunumut24 commented 1 year ago

Hi Zifeng,

Firstly, congrats for great project. I am also interested with same piece of code. Can you send me too please, Thanks in advance (altunumut13@gmail.com)