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

Fix bugs #5036

Merged
merged 2 commits into from
Sep 29, 2022
Merged

Fix bugs #5036

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
eba5171
Fix bugs
Zhilin123 Sep 28, 2022
798cfe4
Merge branch 'r1.12.0' into fix_bugs_qa_and_dialogue
Zhilin123 Sep 28, 2022
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
1 change: 1 addition & 0 deletions examples/nlp/question_answering/conf/qa_conf.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -29,6 +29,7 @@ trainer:
num_sanity_val_steps: 0 # number of steps to perform validation steps for sanity check the validation process before starting the training, setting to 0 disables it
enable_checkpointing: False # provided by exp_manager
logger: False # provided by exp_manager
strategy: ddp

model:
tensor_model_parallel_size: 1
Expand Down
1 change: 0 additions & 1 deletion nemo/collections/nlp/models/dialogue/dialogue_gpt_classification_model.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -477,7 +477,6 @@ def generate_candidates(self, labels, template_length, input_ids, attn_masks):
for i in range(input_ids.size(0)):
param_dict = {
"input_ids": input_ids[i : i + 1, : template_length[i]],
"attention_masks": attn_masks[i : i + 1, : template_length[i]],
"max_length": template_length[i] + tokens_to_generate,
"pad_token_id": self.tokenizer.tokenizer.pad_token_id,
}
Expand Down
1 change: 0 additions & 1 deletion nemo/collections/nlp/models/dialogue/dialogue_gpt_generation_model.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -275,7 +275,6 @@ def generate_candidates(self, labels, template_length, input_ids, attn_masks):
for i in range(input_ids.size(0)):
param_dict = {
"input_ids": input_ids[i : i + 1, : template_length[i]],
"attention_masks": attn_masks[i : i + 1, : template_length[i]],
"max_length": template_length[i] + tokens_to_generate,
"pad_token_id": self.tokenizer.tokenizer.pad_token_id,
}
Expand Down
3 changes: 1 addition & 2 deletions nemo/collections/nlp/models/question_answering/qa_gpt_model.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -309,11 +309,10 @@ def _generate_candidates(self, input_ids, input_attn_mask, training_mask_end):
for i in range(input_ids.size(0)):
param_dict = {
"input_ids": input_ids[i : i + 1, : training_mask_end[i]],
"attention_masks": input_attn_mask[i : i + 1, : training_mask_end[i]],
"max_length": training_mask_end[i] + num_tokens_to_generate,
"pad_token_id": self.tokenizer.tokenizer.pad_token_id,
}
generated_token_ids.append(self.language_model.generate(**param_dict, skip_special_tokens=True))
generated_token_ids.append(self.language_model.generate(**param_dict))
max_length = max(max_length, generated_token_ids[-1].size(1))

# pad each generated to ensure they are of same length in dim 1, therefore stack-able
Expand Down