Skip to content

Actions: ogkalu2/comic-translate

Actions

All workflows

Actions

Loading...
Loading

Showing runs from all workflows
14 workflow runs
14 workflow runs

Filter by Event

Filter by Status

Filter by Branch

Filter by Actor

Build macOS DMG Build macOS DMG #7: Manually run by ogkalu2
6m 51s ci/build
ci/build
6m 51s
Build macOS DMG Build macOS DMG #6: Manually run by ogkalu2
6m 50s ci/build
ci/build
6m 50s
Build macOS DMG Build macOS DMG #5: Manually run by ogkalu2
6m 50s ci/build
ci/build
6m 50s
Build macOS DMG Build macOS DMG #4: Manually run by ogkalu2
6m 37s ci/build
ci/build
6m 37s
Build macOS DMG Build macOS DMG #3: Manually run by ogkalu2
1m 18s ci/build
ci/build
1m 18s
Build macOS DMG Build macOS DMG #2: Manually run by ogkalu2
7m 24s ci/build
ci/build
7m 24s
Build macOS DMG Build macOS DMG #1: Manually run by ogkalu2
1s ci/build
ci/build
1s
Copilot code review Copilot code review #3: by Copilot AI
5m 45s refs/pull/414/head
refs/pull/414/head
5m 45s
Copilot code review Copilot code review #2: by Copilot AI
6m 59s refs/pull/413/head
refs/pull/413/head
6m 59s
Copilot code review Copilot code review #1: by Copilot AI
2m 36s refs/pull/399/head
refs/pull/399/head
2m 36s
In pipeline.py, OCR results are cached to prevent redundant model calls for single block operations. Cache translations similarly Copilot #4: by Copilot AI
18m 0s copilot/fix-c5c65ed1-9b45-4add-8a02-ec722c378ce9
copilot/fix-c5c65ed1-9b45-4add-8a02-ec722c378ce9
18m 0s
pipeline.py tries to implement a OCR caching strategy, where you process all blocks but only use the result for one, but this - all_blocks = [blk for blk in self.main_page.blk_list] creates a soft copy and the blk.text in the main_page.blk_list is still updated. Copilot #3: by Copilot AI
7m 49s copilot/fix-381ad772-5560-4611-b8c9-6cc23687d5f5
copilot/fix-381ad772-5560-4611-b8c9-6cc23687d5f5
7m 49s
Addressing comment on PR #303 Copilot #2: by Copilot AI
7m 29s copilot/fix-96059dc8-1a39-43f0-a8cd-09fe3cce73a1
copilot/fix-96059dc8-1a39-43f0-a8cd-09fe3cce73a1
7m 29s
So my app currently allows a user to right click on any of the rectangles draw on a page and use OCR or Translate for that single block alone (as opposed to every block on the page). The basic functions for these are mostly in pipeline.py and app/ui/list_view.py. However, when i'm using single block, currently i will run the same image through the OCR model multiple times for multiple blocks on the same image, which is wasteful. I don't want to store the ocr'd text in the text block object un... Copilot #1: by Copilot AI
13m 33s copilot/fix-96059dc8-1a39-43f0-a8cd-09fe3cce73a1
copilot/fix-96059dc8-1a39-43f0-a8cd-09fe3cce73a1
13m 33s