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
0f7a4acf
Commit
0f7a4acf
authored
Dec 05, 2021
by
zihanl
Browse files
Merge branch 'knwl_dialo' of
ssh://gitlab-master.nvidia.com:12051/ADLR/megatron-lm
into knwl_dialo
parents
9c623a49
b3eea9d8
Changes
1
Hide whitespace changes
Inline
Side-by-side
Showing
1 changed file
with
0 additions
and
31 deletions
+0
-31
tasks/knwl_dialo/scripts/commands.sh
tasks/knwl_dialo/scripts/commands.sh
+0
-31
No files found.
tasks/knwl_dialo/scripts/commands.sh
deleted
100644 → 0
View file @
9c623a49
# process WoW train
python tasks/knwl_dialo/preprocessing.py
--func
process_wow_dataset
--input_file
/gpfs/fs1/projects/gpu_adlr/datasets/zihanl/dialog_datasets/wizard_of_wikipedia/data/train.json
--output_file
/gpfs/fs1/projects/gpu_adlr/datasets/zihanl/dialog_datasets/wizard_of_wikipedia/data/train.txt
# process WoW test
python tasks/knwl_dialo/preprocessing.py
--func
process_wow_dataset
--input_file
/gpfs/fs1/projects/gpu_adlr/datasets/zihanl/dialog_datasets/wizard_of_wikipedia/data/test_random_split.json
--output_file
/gpfs/fs1/projects/gpu_adlr/datasets/zihanl/dialog_datasets/wizard_of_wikipedia/data/test_seen.txt
python tasks/knwl_dialo/preprocessing.py
--func
process_wow_dataset
--input_file
/gpfs/fs1/projects/gpu_adlr/datasets/zihanl/dialog_datasets/wizard_of_wikipedia/data/test_topic_split.json
--output_file
/gpfs/fs1/projects/gpu_adlr/datasets/zihanl/dialog_datasets/wizard_of_wikipedia/data/test_unseen.txt
# process WoI test
python tasks/knwl_dialo/preprocessing.py
--func
process_woi_dataset
--input_file
/gpfs/fs1/projects/gpu_adlr/datasets/zihanl/dialog_datasets/wizard_of_internet/data/test.jsonl
--output_file
/gpfs/fs1/projects/gpu_adlr/datasets/zihanl/dialog_datasets/wizard_of_internet/data/test.txt
# get knowledge generation prompts
# WoW seen
python tasks/knwl_dialo/preprocessing.py
--func
get_knwl_gen_prompts
--test_file
/gpfs/fs1/projects/gpu_adlr/datasets/zihanl/dialog_datasets/wizard_of_wikipedia/data/test_seen.txt
--train_file
/gpfs/fs1/projects/gpu_adlr/datasets/zihanl/dialog_datasets/wizard_of_wikipedia/data/train.txt
--model_file
/gpfs/fs1/projects/gpu_adlr/datasets/zihanl/checkpoints/dpr_wow/best_question_encoder.pt
--data_type
wow_seen
--output_file
/gpfs/fs1/projects/gpu_adlr/datasets/zihanl/dialog_datasets/wizard_of_wikipedia/data/knowledge_prompts_test_seen.json
# WoW unseen
python tasks/knwl_dialo/preprocessing.py
--func
get_knwl_gen_prompts
--test_file
/gpfs/fs1/projects/gpu_adlr/datasets/zihanl/dialog_datasets/wizard_of_wikipedia/data/test_unseen.txt
--train_file
/gpfs/fs1/projects/gpu_adlr/datasets/zihanl/dialog_datasets/wizard_of_wikipedia/data/train.txt
--model_file
/gpfs/fs1/projects/gpu_adlr/datasets/zihanl/checkpoints/dpr_wow_ctrl/best_question_encoder.pt
--data_type
wow_unseen
--output_file
/gpfs/fs1/projects/gpu_adlr/datasets/zihanl/dialog_datasets/wizard_of_wikipedia/data/knowledge_prompts_test_unseen.json
# WoI
python tasks/knwl_dialo/preprocessing.py
--func
get_knwl_gen_prompts
--test_file
/gpfs/fs1/projects/gpu_adlr/datasets/zihanl/dialog_datasets/wizard_of_internet/data/test.txt
--train_file
/gpfs/fs1/projects/gpu_adlr/datasets/zihanl/dialog_datasets/wizard_of_wikipedia/data/train.txt
--model_file
/gpfs/fs1/projects/gpu_adlr/datasets/zihanl/checkpoints/dpr_wow_ctrl/best_question_encoder.pt
--data_type
woi
--output_file
/gpfs/fs1/projects/gpu_adlr/datasets/zihanl/dialog_datasets/wizard_of_internet/data/knowledge_prompts_test.json
# get response generation prompts --seed 147
python tasks/knwl_dialo/preprocessing.py
--func
get_resp_gen_prompts
--train_file
/gpfs/fs1/projects/gpu_adlr/datasets/zihanl/dialog_datasets/wizard_of_wikipedia/data/train.txt
--output_file
/gpfs/fs1/projects/gpu_adlr/datasets/zihanl/dialog_datasets/wizard_of_wikipedia/data/response_generation_prompts_temp.txt
--seed
1234
# prepare response generation inputs
# WoW seen
python tasks/knwl_dialo/preprocessing.py
--func
prepare_input
--test_file
/gpfs/fs1/projects/gpu_adlr/datasets/zihanl/dialog_datasets/wizard_of_wikipedia/data/test_seen.txt
--knowledge_file
/gpfs/fs1/projects/gpu_adlr/datasets/zihanl/dialog_datasets/wizard_of_wikipedia/data/output_testseen_knowledge_357m.txt
--output_file
/gpfs/fs1/projects/gpu_adlr/datasets/zihanl/dialog_datasets/wizard_of_wikipedia/data/test_seen_resp_gen_input.txt
# WoW unseen
python tasks/knwl_dialo/preprocessing.py
--func
prepare_input
--test_file
/gpfs/fs1/projects/gpu_adlr/datasets/zihanl/dialog_datasets/wizard_of_wikipedia/data/test_unseen.txt
--knowledge_file
/gpfs/fs1/projects/gpu_adlr/datasets/zihanl/dialog_datasets/wizard_of_wikipedia/data/output_testunseen_knowledge_357m.txt
--output_file
/gpfs/fs1/projects/gpu_adlr/datasets/zihanl/dialog_datasets/wizard_of_wikipedia/data/test_unseen_resp_gen_input.txt
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