Skip to content
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.

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

[WT-103] Unable to reproduce results #2

Open
DavidHerel opened this issue Sep 22, 2023 · 0 comments
Open

[WT-103] Unable to reproduce results #2

DavidHerel opened this issue Sep 22, 2023 · 0 comments
Labels
bug Something isn't working

Comments

@DavidHerel
Copy link

DavidHerel commented Sep 22, 2023

Hi,

I have successfully done all the steps from original knn_lm.
Then I have tried to run your scripts:
python rq/fast_evaluate.py --preset wiki_valid --save_knns # Save the neighbors.
python rq/fast_evaluate.py --preset wiki_valid --save_exact # Save the exact vector distances.
python rq/fast_evaluate.py --preset wiki_valid --exact # Compute perplexity using exact vector distance.

But I got following error:

fast_evaluate.py: error: unrecognized arguments: --save_exact
load dataset
Traceback (most recent call last):
File "rq/fast_evaluate.py", line 233, in
main(args)
File "rq/fast_evaluate.py", line 186, in main
dataset = Dataset(args)
File "/home/hereldav/lm-ensemble/models/knnlm-retrieval-quality/rq/data_structures.py", line 14, in init
self.query = np.memmap(f'{path}/dstore_keys.npy', dtype=np.float16, mode='r', shape=(dstore_size, 1024))
File "/home/hereldav/.conda/envs/knn_lm_py37/lib/python3.7/site-packages/numpy/core/memmap.py", line 225, in new
f_ctx = open(os_fspath(filename), ('r' if mode == 'c' else mode)+'b')
FileNotFoundError: [Errno 2] No such file or directory: '/iesl/local/adrozdov/knnlm_data.valid/dstore_keys.npy'

@DavidHerel DavidHerel added the bug Something isn't working label Sep 22, 2023
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
bug Something isn't working
Projects
None yet
Development

No branches or pull requests

1 participant