Skip to content
GitLab
Menu
Projects
Groups
Snippets
Loading...
Help
Help
Support
Community forum
Keyboard shortcuts
?
Submit feedback
Contribute to GitLab
Sign in / Register
Toggle navigation
Menu
Open sidebar
OpenDAS
Megatron-LM
Commits
dfb6a9b3
Commit
dfb6a9b3
authored
May 18, 2021
by
Mostofa Patwary
Browse files
updated the evaluation script for retriver
parent
7577931b
Changes
1
Hide whitespace changes
Inline
Side-by-side
Showing
1 changed file
with
4 additions
and
4 deletions
+4
-4
examples/evaluate_retriever_nq.sh
examples/evaluate_retriever_nq.sh
+4
-4
No files found.
examples/evaluate_
ict_zeroshot
_nq.sh
→
examples/evaluate_
retriever
_nq.sh
View file @
dfb6a9b3
#!/bin/bash
#!/bin/bash
# Evaluate natural question test data given Wikipedia embeddings and pretrained
# Evaluate natural question test data given Wikipedia embeddings and pretrained
# ICT model
# ICT model
or a finetuned model for Natural Question task
# Datasets can be downloaded from the following link:
# Datasets can be downloaded from the following link:
# https://github.com/facebookresearch/DPR/blob/master/data/download_data.py
# https://github.com/facebookresearch/DPR/blob/master/data/download_data.py
EVIDENCE_DATA_DIR
=
<Specify path of Wikipedia dataset>
EVIDENCE_DATA_DIR
=
<Specify path of Wikipedia dataset>
EMBEDDING_PATH
=
<Specify path of the embeddings>
EMBEDDING_PATH
=
<Specify path of the embeddings>
CHECKPOINT_PATH
=
<Specify path of pretrained ICT model>
CHECKPOINT_PATH
=
<Specify path of pretrained ICT
model or finetuned
model>
QA_FILE
=
<Path of the natural question
test
dataset>
QA_FILE
=
<Path of the natural question
dev or
test
dataset>
python tasks/main.py
\
python tasks/main.py
\
--task
ICT-ZEROSHOT-NQ
\
--task
RETRIEVER-EVAL
\
--tokenizer-type
BertWordPieceLowerCase
\
--tokenizer-type
BertWordPieceLowerCase
\
--num-layers
12
\
--num-layers
12
\
--hidden-size
768
\
--hidden-size
768
\
...
...
Write
Preview
Markdown
is supported
0%
Try again
or
attach a new file
.
Attach a file
Cancel
You are about to add
0
people
to the discussion. Proceed with caution.
Finish editing this message first!
Cancel
Please
register
or
sign in
to comment