Skip to content
Open
Show file tree
Hide file tree
Changes from all commits
Commits
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
35 changes: 21 additions & 14 deletions embed/src/pixelrag_embed/chunk.py
Original file line number Diff line number Diff line change
Expand Up @@ -36,6 +36,7 @@
from pathlib import Path

from PIL import Image
from tqdm import tqdm

Image.MAX_IMAGE_PIXELS = None # some tiles exceed default 178M pixel limit

Expand Down Expand Up @@ -304,21 +305,27 @@ def process_shard(
# Flat structure — article dirs directly in shard_dir
sub_dirs = [Path(shard_dir)]

for sub_dir in sub_dirs:
for article_dir in sorted(sub_dir.iterdir()):
if not article_dir.is_dir() or not article_dir.name.endswith(".png.tiles"):
continue
total_articles += 1

result = chunk_article(str(article_dir), dry_run=dry_run, force=force)
if result is None:
skipped_articles += 1
continue
all_article_dirs = [
article_dir
for sub_dir in sub_dirs
for article_dir in sorted(sub_dir.iterdir())
if article_dir.is_dir() and article_dir.name.endswith(".png.tiles")
]

for article_dir in tqdm(
all_article_dirs, desc="Chunking", disable=not all_article_dirs
):
total_articles += 1

result = chunk_article(str(article_dir), dry_run=dry_run, force=force)
if result is None:
skipped_articles += 1
continue

chunked_articles += 1
total_tiles += result["num_tiles"]
total_chunks += result["num_chunks"]
total_files += result["files_written"]
chunked_articles += 1
total_tiles += result["num_tiles"]
total_chunks += result["num_chunks"]
total_files += result["files_written"]

# Delete tiles after chunking the whole shard
tiles_deleted = 0
Expand Down
3 changes: 0 additions & 3 deletions embed/src/pixelrag_embed/embed_cpu.py
Original file line number Diff line number Diff line change
Expand Up @@ -187,9 +187,6 @@ def embed_items(
pooled = pooled / pooled.norm()
embeddings[i] = pooled.cpu().numpy().astype(np.float16)

if (i + 1) % 10 == 0:
logger.info("Embedded %d/%d", i + 1, len(items))

return embeddings


Expand Down
Loading