Skip to content

Issues: megan-k-ayers/ConTextNet

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

24 Open 2 Closed
24 Open 2 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Assignee
Filter by who’s assigned
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Issues list

Make sure tf and keras functions are always passed tensors Medium priority
#27 opened Oct 1, 2024 by megan-k-ayers
Move scaling to inside the Keras model Low priority
#26 opened Oct 1, 2024 by megan-k-ayers
1
Add new interpretation function that returns full texts that activate very high/low on a given filter enhancement New feature or request Low priority
#25 opened Sep 26, 2024 by megan-k-ayers
Implement Jacovi et al. (2018) threshold/purity logic enhancement New feature or request Low priority
#24 opened Sep 26, 2024 by megan-k-ayers
Penalize correlation between max-pooled filter activations and model residuals during training. enhancement New feature or request Low priority
#23 opened Sep 26, 2024 by megan-k-ayers
Built-in option of fine-tuning the embedding model enhancement New feature or request Low priority
#22 opened Sep 26, 2024 by megan-k-ayers
Add methods for identifying and incorporating heterogeneous treatment effects with covariates enhancement New feature or request Low priority
#21 opened Sep 26, 2024 by megan-k-ayers
Figure out if memory clearing during tuning is possible without adverse user side effects Medium priority
#18 opened Sep 23, 2024 by megan-k-ayers
Add output weight/top phrases to titles in plot_doc_acts() Low priority
#17 opened Sep 23, 2024 by megan-k-ayers
Pipeline for saving intermed/final objects on HPC for embedding and tuning High priority
#16 opened Sep 23, 2024 by megan-k-ayers
Either figure out non-invasive Keras seed setting or recommend users do it manually Medium priority
#15 opened Sep 23, 2024 by megan-k-ayers
Logging system for components that get run on HPC High priority
#14 opened Sep 23, 2024 by megan-k-ayers
Less generic names for functions Low priority
#13 opened Sep 23, 2024 by megan-k-ayers
Include imdb/example tokens and embeds as saved files Low priority
#12 opened Sep 23, 2024 by megan-k-ayers
Figure out ideal reticulate import solution Low priority
#11 opened Sep 23, 2024 by megan-k-ayers
Build out option of uploading existing tokens/vocab/embeds w/ error handling for type/shape/order High priority
#10 opened Sep 23, 2024 by megan-k-ayers
Clarify for users that default embed() and tokenize() downloads model files Medium priority
#9 opened Sep 23, 2024 by megan-k-ayers
Error handling for prep_params() Medium priority
#8 opened Sep 23, 2024 by megan-k-ayers
Functionality to generate shell scripts/Slurm commands for HPC High priority
#7 opened Sep 23, 2024 by megan-k-ayers
Make test set proportion exact in prep_data() Medium priority
#5 opened Sep 23, 2024 by megan-k-ayers
Non-invasive set.seed for sample splitting in prep.data() Medium priority
#4 opened Sep 23, 2024 by megan-k-ayers
Check Python setup of users (fill in skeleton at check_setup.R) Low priority
#3 opened Sep 23, 2024 by megan-k-ayers
Revise documentation of core functions Medium priority
#2 opened Sep 23, 2024 by megan-k-ayers
Write tests for core functions High priority
#1 opened Sep 23, 2024 by megan-k-ayers
ProTip! Adding no:label will show everything without a label.