Skip to content

fix: remove errors with captions #40

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

Merged
merged 3 commits into from
Mar 28, 2024
Merged

fix: remove errors with captions #40

Changes from 2 commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
b5d67c9
fix: remove errors with captions
NastyaMittseva Mar 26, 2024
01a8417
fix: remove the 2nd append
NastyaMittseva Mar 26, 2024
f8a0ad4
fix: change return value of function
gofixyourself Mar 28, 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
5 changes: 3 additions & 2 deletions DPF/filters/videos/video_llava_filter.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -52,7 +52,7 @@ def __init__(
model_base: Optional[str] = None,
cache_path: str = "cache_dir",
prompt: str = "detailed_video",
temperature: float = 0.2,
temperature: float = 0.8,
max_new_tokens: int = 1024,
load_4bit: bool = False,
load_8bit: bool = False,
Expand Down Expand Up @@ -142,13 +142,14 @@ def process_batch(self, batch: list[Any]) -> dict[str, list[Any]]:
do_sample=True if self.temperature > 0 else False,
temperature=self.temperature,
max_new_tokens=self.max_new_tokens,
num_beams=1,
no_repeat_ngram_size=2,
use_cache=True,
stopping_criteria=[self.stopping_criteria])

all_outputs: list[Optional[str]] = []
for i in range(output_ids.shape[0]):
caption = self.tokenizer.decode(output_ids[i, self.input_ids.shape[1]:]).strip().split('</s>')[0]
all_outputs.append(caption)
all_outputs.append(check_caption(caption))
Copy link
Contributor

@boomb0om boomb0om Mar 28, 2024

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

check_caption может вернуть None - такого быть не должно

df_batch_labels[self.schema[1]].extend(all_outputs)
df_batch_labels[self.key_column].extend(keys)
Expand Down