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.

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

Jump to bottom

Anonymize config files and scripts #68

Merged
merged 5 commits into from
Feb 10, 2024
Merged

Anonymize config files and scripts #68

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
27a95fb
Anonymized training confs
zeynepyirmibes Feb 10, 2024
4b6c8bf
Anonymized generation confs
zeynepyirmibes Feb 10, 2024
cc2010f
Anonymized scripts
zeynepyirmibes Feb 10, 2024
5953523
Merge branch 'main' into patch-anonymize
zeynepyirmibes Feb 10, 2024
db0d240
Anonymized leftovers
zeynepyirmibes Feb 10, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion experiments/conf/class_prod_reviews.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -12,5 +12,5 @@ training_params:
per_device_eval_batch_size: 32 # for ultramarine
# Conditional generation mode: classification/cg/tr_product_reviews
# Classification mode: classification/tc/tr_product_reviews
output_dir: /stratch/bounllm/finetuned-models/ul2tr/classification/cg/tr_product_reviews
output_dir: /finetuned-models/turna/classification/cg/tr_product_reviews

2 changes: 1 addition & 1 deletion experiments/conf/class_ttc4900.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -12,6 +12,6 @@ training_params:
per_device_eval_batch_size: 2 # for ultramarine
# Conditional generation mode: classification/cg/ttc4900
# Classification mode: classification/tc/ttc4900
output_dir: /stratch/bounllm/finetuned-models/ul2tr/classification/cg/ttc4900
output_dir: /finetuned-models/turna/classification/cg/ttc4900


4 changes: 2 additions & 2 deletions experiments/conf/class_tweet_sentiment.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4,7 +4,7 @@ defaults:

dataset_name: 17bintweet_sentiment
task: classification
dataset_loc: "/stratch/bounllm/supervised/17bintweet_sentiment/"
dataset_loc: "/supervised/17bintweet_sentiment/"
max_input_length: 160
max_target_length: 4
num_labels: 3
Expand All @@ -13,5 +13,5 @@ training_params:
per_device_eval_batch_size: 32 # for ultramarine
# Conditional generation mode: classification/cg/17bintweet_sentiment
# Classification mode: classification/tc/17bintweet_sentiment
output_dir: /stratch/bounllm/finetuned-models/ul2tr/classification/cg/17bintweet_sentiment
output_dir: /finetuned-models/turna/classification/cg/17bintweet_sentiment

2 changes: 1 addition & 1 deletion experiments/conf/debug.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,4 +1,4 @@
model_name: /stratch/bounllm/ckpt-500K
model_name: /pretrained_checkpoints/ckpt-1.74M
task_format: classification
task_mode: '' # '[S2S]: ', '[NLU]: ', '[NLG]: '
adafactor_scheduler: True
Expand Down
2 changes: 1 addition & 1 deletion experiments/conf/default.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,4 +1,4 @@
model_name: /stratch/bounllm/pretrained_checkpoints/ckpt-1.74M
model_name: /pretrained_checkpoints/ckpt-1.74M
task_mode: '' # '[S2S]: ', '[NLU]: ', '[NLG]: '
training_params:
num_train_epochs: 10
Expand Down
4 changes: 2 additions & 2 deletions experiments/conf/ner_milliyet.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,7 +1,7 @@
defaults:
- default_cond_gen

dataset_loc: "/stratch/bounllm/supervised/ner-milliyet"
dataset_loc: "/supervised/ner-milliyet"
dataset_name: milliyet
task: ner
max_input_length: 380
Expand All @@ -10,5 +10,5 @@ training_params:
num_train_epochs: 10
per_device_train_batch_size: 8 # worked in ultramarine
per_device_eval_batch_size: 8 # worked in ultramarine
output_dir: /stratch/bounllm/finetuned-models/ul2tr/ner/milliyet
output_dir: /finetuned-models/turna/ner/milliyet
include_inputs_for_metrics: true
4 changes: 2 additions & 2 deletions experiments/conf/ner_milliyet_token.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,7 +1,7 @@
defaults:
- default_classification

dataset_loc: "/stratch/bounllm/supervised/ner-milliyet2"
dataset_loc: "/supervised/ner-milliyet2"
dataset_name: milliyet
task: ner
max_input_length: 380
Expand All @@ -10,4 +10,4 @@ num_labels: 7
training_params:
# per_device_train_batch_size: 64 # worked in ultramarine
# per_device_eval_batch_size: 64 # worked in ultramarine
output_dir: /stratch/bounllm/finetuned-models/turna-encoder/ner/milliyet
output_dir: /finetuned-models/turna-encoder/ner/milliyet
2 changes: 1 addition & 1 deletion experiments/conf/ner_wikiann.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -8,5 +8,5 @@ max_target_length: 40
training_params:
per_device_train_batch_size: 64 # worked in ultramarine
per_device_eval_batch_size: 64 # worked in ultramarine
output_dir: /stratch/bounllm/finetuned-models/ul2tr/ner/wikiann
output_dir: /finetuned-models/turna/ner/wikiann
include_inputs_for_metrics: true
2 changes: 1 addition & 1 deletion experiments/conf/ner_wikiann_token.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -9,4 +9,4 @@ num_labels: 7
training_params:
# per_device_train_batch_size: 64 # worked in ultramarine
# per_device_eval_batch_size: 64 # worked in ultramarine
output_dir: /stratch/bounllm/finetuned-models/turna-encoder/ner/wikiann
output_dir: /finetuned-models/turna-encoder/ner/wikiann
2 changes: 1 addition & 1 deletion experiments/conf/nli_multinli_tr.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -10,4 +10,4 @@ num_labels: 3
training_params:
per_device_train_batch_size: 32 # for ultramarine
per_device_eval_batch_size: 32 # for ultramarine
output_dir: /stratch/bounllm/finetuned-models/ul2tr/nli/nli_multinli_tr
output_dir: /finetuned-models/turna/nli/nli_multinli_tr
2 changes: 1 addition & 1 deletion experiments/conf/nli_snli_tr.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -10,4 +10,4 @@ num_labels: 3
training_params:
per_device_train_batch_size: 32 # for ultramarine
per_device_eval_batch_size: 32 # for ultramarine
output_dir: /stratch/bounllm/finetuned-models/ul2tr/nli/nli_snli_tr
output_dir: /finetuned-models/turna/nli/nli_snli_tr
2 changes: 1 addition & 1 deletion experiments/conf/nli_tr.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -10,4 +10,4 @@ num_labels: 3
training_params:
per_device_train_batch_size: 32 # for ultramarine
per_device_eval_batch_size: 32 # for ultramarine
output_dir: /stratch/bounllm/finetuned-models/ul2tr/nli/nli_tr
output_dir: /finetuned-models/turna/nli/nli_tr
2 changes: 1 addition & 1 deletion experiments/conf/paraphrasing.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -8,4 +8,4 @@ max_target_length: 20
training_params:
per_device_train_batch_size: 128
per_device_eval_batch_size: 128
output_dir: /stratch/bounllm/finetuned-models/paraphrasing/tatoeba
output_dir: /finetuned-models/paraphrasing/tatoeba
4 changes: 2 additions & 2 deletions experiments/conf/pos_boun.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,7 +1,7 @@
defaults:
- default_cond_gen

dataset_loc: "/stratch/bounllm/supervised/tr_boun-ud"
dataset_loc: "/supervised/tr_boun-ud"
dataset_name: boun
task: pos_tagging
max_input_length: 90
Expand All @@ -10,5 +10,5 @@ training_params:
num_train_epochs: 10
per_device_train_batch_size: 8 # worked in ultramarine
per_device_eval_batch_size: 8 # worked in ultramarine
output_dir: /stratch/bounllm/finetuned-models/ul2tr/pos/boun
output_dir: /finetuned-models/turna/pos/boun
include_inputs_for_metrics: true
4 changes: 2 additions & 2 deletions experiments/conf/pos_boun_token.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,11 +1,11 @@
defaults:
- default_classification

dataset_loc: "/stratch/bounllm/supervised/tr_boun-ud"
dataset_loc: "/supervised/tr_boun-ud"
dataset_name: boun
task: pos_tagging
max_input_length: 90
max_target_length: 300
num_labels: 17
training_params:
output_dir: /stratch/bounllm/finetuned-models/turna-encoder/pos/boun
output_dir: /finetuned-models/turna-encoder/pos/boun
4 changes: 2 additions & 2 deletions experiments/conf/pos_imst.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,7 +1,7 @@
defaults:
- default_cond_gen

dataset_loc: "/stratch/bounllm/supervised/tr_imst-ud"
dataset_loc: "/supervised/tr_imst-ud"
dataset_name: imst
task: pos_tagging
max_input_length: 60
Expand All @@ -10,5 +10,5 @@ training_params:
num_train_epochs: 10
per_device_train_batch_size: 16 # worked in ultramarine
per_device_eval_batch_size: 16 # worked in ultramarine
output_dir: /stratch/bounllm/finetuned-models/ul2tr/pos/imst
output_dir: /finetuned-models/turna/pos/imst
include_inputs_for_metrics: true
4 changes: 2 additions & 2 deletions experiments/conf/pos_imst_token.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,11 +1,11 @@
defaults:
- default_classification

dataset_loc: "/stratch/bounllm/supervised/tr_imst-ud"
dataset_loc: "/supervised/tr_imst-ud"
dataset_name: imst
task: pos_tagging
max_input_length: 60
max_target_length: 210
num_labels: 17
training_params:
output_dir: /stratch/bounllm/finetuned-models/turna-encoder/pos/imst
output_dir: /finetuned-models/turna-encoder/pos/imst
2 changes: 1 addition & 1 deletion experiments/conf/qa_exams.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -9,4 +9,4 @@ training_params:
num_train_epochs: 10
per_device_train_batch_size: 16 # worked in ultramarine
per_device_eval_batch_size: 16 # worked in ultramarine
output_dir: /stratch/bounllm/finetuned-models/ul2tr/qa/exams
output_dir: /finetuned-models/turna/qa/exams
2 changes: 1 addition & 1 deletion experiments/conf/qa_mkqa.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -9,4 +9,4 @@ training_params:
num_train_epochs: 10
per_device_train_batch_size: 64 # worked in ultramarine
per_device_eval_batch_size: 64 # worked in ultramarine
output_dir: /stratch/bounllm/finetuned-models/ul2tr/qa/mkqa
output_dir: /finetuned-models/turna/qa/mkqa
4 changes: 2 additions & 2 deletions experiments/conf/qa_tquad.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,7 +1,7 @@
defaults:
- default_cond_gen

dataset_loc: "/stratch/bounllm/supervised/tquad"
dataset_loc: "/supervised/tquad"
dataset_name: tquad
task: question_answering
max_input_length: 1350
Expand All @@ -10,4 +10,4 @@ training_params:
num_train_epochs: 10
per_device_train_batch_size: 2 # worked in ultramarine
per_device_eval_batch_size: 2 # worked in ultramarine
output_dir: /stratch/bounllm/finetuned-models/ul2tr/qa/tquad
output_dir: /finetuned-models/turna/qa/tquad
4 changes: 2 additions & 2 deletions experiments/conf/semantic_similarity.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4,13 +4,13 @@ defaults:

dataset_name: stsb_tr
task: semantic_similarity
dataset_loc: "/stratch/bounllm/supervised/STSb-TR-corrected/"
dataset_loc: "/supervised/STSb-TR-corrected/"
max_input_length: 140
max_target_length: 10
num_labels: 1
training_params:
per_device_train_batch_size: 32 # for ultramarine
per_device_eval_batch_size: 32 # for ultramarine
output_dir: /stratch/bounllm/finetuned-models/ul2tr/semantic_similarity/stsb_tr/
output_dir: /finetuned-models/turna/semantic_similarity/stsb_tr/


2 changes: 1 addition & 1 deletion experiments/conf/summarization.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -8,7 +8,7 @@ max_target_length: 128
training_params:
per_device_train_batch_size: 4 # 4 for ultramarine
per_device_eval_batch_size: 4 # 4 for ultramarine
output_dir: /stratch/bounllm/finetuned-models/ul2tr/summarization/tr_news
output_dir: /finetuned-models/turna/summarization/tr_news
bf16: true


2 changes: 1 addition & 1 deletion experiments/conf/title_generation.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -8,6 +8,6 @@ max_target_length: 64
training_params:
per_device_train_batch_size: 8 # 8 for ultramarine
per_device_eval_batch_size: 8 # 8 for ultramarine
output_dir: /stratch/bounllm/finetuned-models/ul2tr/title_generation/tr_news
output_dir: /finetuned-models/turna/title_generation/tr_news


6 changes: 3 additions & 3 deletions experiments/generation_conf/class_prod_reviews.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -7,9 +7,9 @@ task: classification
max_input_length: 20
max_target_length: 4
num_labels: 2
model_path: /stratch/bounllm/finetuned-models/ul2tr/classification/cg/tr_product_reviews/
tokenizer_path: /stratch/bounllm/pretrained_checkpoints/ckpt-1.74M
model_path: /finetuned-models/turna/classification/cg/tr_product_reviews/
tokenizer_path: /pretrained_checkpoints/ckpt-1.74M
test_params:
per_device_eval_batch_size: 32 # for ultramarine
output_dir: /stratch/bounllm/finetuned-models/ul2tr/classification/cg/tr_product_reviews/
output_dir: /finetuned-models/turna/classification/cg/tr_product_reviews/

6 changes: 3 additions & 3 deletions experiments/generation_conf/class_ttc4900.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -7,10 +7,10 @@ task: classification
max_input_length: 1450
max_target_length: 8
num_labels: 7
model_path: /stratch/bounllm/finetuned-models/ul2tr/classification/cg/ttc4900
tokenizer_path: /stratch/bounllm/pretrained_checkpoints/ckpt-1.74M
model_path: /finetuned-models/turna/classification/cg/ttc4900
tokenizer_path: /pretrained_checkpoints/ckpt-1.74M
test_params:
per_device_eval_batch_size: 2 # for ultramarine
output_dir: /stratch/bounllm/finetuned-models/ul2tr/classification/cg/ttc4900
output_dir: /finetuned-models/turna/classification/cg/ttc4900


8 changes: 4 additions & 4 deletions experiments/generation_conf/ner_milliyet.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,14 +1,14 @@
defaults:
- default_cond_gen

dataset_loc: "/stratch/bounllm/supervised/ner-milliyet"
dataset_loc: "/supervised/ner-milliyet"
dataset_name: milliyet
task: ner
max_input_length: 380
max_target_length: 60
tokenizer_path: /stratch/bounllm/pretrained_checkpoints/ckpt-1.74M
model_path: /stratch/bounllm/finetuned-models/ul2tr/ner/milliyet
tokenizer_path: /pretrained_checkpoints/ckpt-1.74M
model_path: /finetuned-models/turna/ner/milliyet
test_params:
per_device_eval_batch_size: 4
output_dir: /stratch/bounllm/finetuned-models/ul2tr/ner/milliyet
output_dir: /finetuned-models/turna/ner/milliyet
include_inputs_for_metrics: true
6 changes: 3 additions & 3 deletions experiments/generation_conf/ner_wikiann.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -5,9 +5,9 @@ dataset_name: wikiann
task: ner
max_input_length: 60
max_target_length: 40
tokenizer_path: /stratch/bounllm/pretrained_checkpoints/ckpt-1.74M
model_path: /stratch/bounllm/finetuned-models/ul2tr/ner/wikiann
tokenizer_path: /pretrained_checkpoints/ckpt-1.74M
model_path: /finetuned-models/turna/ner/wikiann
test_params:
per_device_eval_batch_size: 4
output_dir: /stratch/bounllm/finetuned-models/ul2tr/ner/wikiann
output_dir: /finetuned-models/turna/ner/wikiann
include_inputs_for_metrics: true
4 changes: 2 additions & 2 deletions experiments/generation_conf/nli_snli_tr.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -6,8 +6,8 @@ dataset_name: snli_tr
task: nli
max_input_length: 128
max_target_length: 8
tokenizer_path: /stratch/bounllm/pretrained_checkpoints/ckpt-1.74M
model_path: /stratch/bounllm/finetuned-models/ul2tr/summarization/tr_news/
tokenizer_path: /pretrained_checkpoints/ckpt-1.74M
model_path: /finetuned-models/turna/summarization/tr_news/
test_params:
per_device_eval_batch_size: 32 # for ultramarine
output_dir: ./t5_nli_tr
6 changes: 3 additions & 3 deletions experiments/generation_conf/nli_tr.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -7,8 +7,8 @@ task: nli
max_input_length: 128
max_target_length: 8
num_labels: 3
tokenizer_path: '/stratch/bounllm/pretrained_checkpoints/ckpt-1.74M'
model_path: /stratch/bounllm/finetuned-models/ul2tr/nli/nli_tr/
tokenizer_path: '/pretrained_checkpoints/ckpt-1.74M'
model_path: /finetuned-models/turna/nli/nli_tr/
test_params:
per_device_eval_batch_size: 32 # for ultramarine
output_dir: /stratch/bounllm/finetuned-models/ul2tr/nli/nli_tr/
output_dir: /finetuned-models/turna/nli/nli_tr/
6 changes: 3 additions & 3 deletions experiments/generation_conf/paraphrasing.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -5,8 +5,8 @@ dataset_name: tatoeba
task: paraphrasing
max_input_length: 20
max_target_length: 20
model_path: /stratch/bounllm/finetuned-models/ul2tr/paraphrasing/tatoeba/
tokenizer_path: /stratch/bounllm/pretrained_checkpoints/ckpt-1.74M
model_path: /finetuned-models/turna/paraphrasing/tatoeba/
tokenizer_path: /pretrained_checkpoints/ckpt-1.74M
test_params:
per_device_eval_batch_size: 128 # 4 for ultramarine
output_dir: /stratch/bounllm/finetuned-models/ul2tr/paraphrasing/tatoeba/
output_dir: /finetuned-models/turna/paraphrasing/tatoeba/
8 changes: 4 additions & 4 deletions experiments/generation_conf/pos_boun.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,14 +1,14 @@
defaults:
- default_cond_gen

dataset_loc: "/stratch/bounllm/supervised/tr_boun-ud"
dataset_loc: "/supervised/tr_boun-ud"
dataset_name: boun
task: pos_tagging
max_input_length: 90
max_target_length: 300
tokenizer_path: /stratch/bounllm/pretrained_checkpoints/ckpt-1.74M
model_path: /stratch/bounllm/finetuned-models/ul2tr/pos/boun
tokenizer_path: /pretrained_checkpoints/ckpt-1.74M
model_path: /finetuned-models/turna/pos/boun
test_params:
per_device_eval_batch_size: 4
output_dir: /stratch/bounllm/finetuned-models/ul2tr/pos/boun
output_dir: /finetuned-models/turna/pos/boun
include_inputs_for_metrics: true
8 changes: 4 additions & 4 deletions experiments/generation_conf/pos_imst.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,14 +1,14 @@
defaults:
- default_cond_gen

dataset_loc: "/stratch/bounllm/supervised/tr_imst-ud"
dataset_loc: "/supervised/tr_imst-ud"
dataset_name: imst
task: pos_tagging
max_input_length: 60
max_target_length: 210
tokenizer_path: /stratch/bounllm/pretrained_checkpoints/ckpt-1.74M
model_path: /stratch/bounllm/finetuned-models/ul2tr/pos/imst
tokenizer_path: /pretrained_checkpoints/ckpt-1.74M
model_path: /finetuned-models/turna/pos/imst
test_params:
per_device_eval_batch_size: 4
output_dir: /stratch/bounllm/finetuned-models/ul2tr/pos/imst
output_dir: /finetuned-models/turna/pos/imst
include_inputs_for_metrics: true
6 changes: 3 additions & 3 deletions experiments/generation_conf/qa_exams.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -5,8 +5,8 @@ dataset_name: exams
task: question_answering
max_input_length: 200
max_target_length: 50
tokenizer_path: /stratch/bounllm/pretrained_checkpoints/ckpt-1.74M
model_path: /stratch/bounllm/finetuned-models/ul2tr/qa/exams
tokenizer_path: /pretrained_checkpoints/ckpt-1.74M
model_path: /finetuned-models/turna/qa/exams
test_params:
per_device_eval_batch_size: 16 # worked in ultramarine
output_dir: /stratch/bounllm/finetuned-models/ul2tr/qa/exams
output_dir: /finetuned-models/turna/qa/exams
6 changes: 3 additions & 3 deletions experiments/generation_conf/qa_mkqa.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -5,8 +5,8 @@ dataset_name: mkqa
task: question_answering
max_input_length: 40
max_target_length: 20
tokenizer_path: /stratch/bounllm/pretrained_checkpoints/ckpt-1.74M
model_path: /stratch/bounllm/finetuned-models/ul2tr/qa/mkqa
tokenizer_path: /pretrained_checkpoints/ckpt-1.74M
model_path: /finetuned-models/turna/qa/mkqa
test_params:
per_device_eval_batch_size: 64 # worked in ultramarine
output_dir: /stratch/bounllm/finetuned-models/ul2tr/qa/mkqa
output_dir: /finetuned-models/turna/qa/mkqa
Loading