-
Notifications
You must be signed in to change notification settings - Fork 71
Pull requests: google-ai-edge/ai-edge-torch
Author
Label
Projects
Milestones
Reviews
Assignee
Sort
Pull requests list
update quantization documentations for int4 blockwise.
#641
opened May 9, 2025 by
copybara-service
bot
Loading…
Modify tensor names of the image encoder to load weights from checkpoint
#633
opened May 6, 2025 by
copybara-service
bot
Loading…
add eps_free and with_scale options in RMSNorm.
#610
opened Apr 24, 2025 by
copybara-service
bot
Loading…
Update legalization rule for TFL op in StableHLO custom_call carrier
#568
opened Mar 20, 2025 by
copybara-service
bot
Loading…
#odml-torch composite builder single boundary op
#559
opened Mar 18, 2025 by
copybara-service
bot
Loading…
Modified GemmaWrapper to pass local mask cache to original model.
#532
opened Mar 6, 2025 by
copybara-service
bot
Loading…
Add transformation to allow transposed input to BatchMatMul
#473
opened Jan 21, 2025 by
copybara-service
bot
Loading…
Support passing attention mask as optional input in text_generator_main.cc.
#430
opened Dec 16, 2024 by
copybara-service
bot
Loading…
Internal change. There is no other interesting things here.
#361
opened Nov 15, 2024 by
copybara-service
bot
Loading…
Create a guitar to catch numerical regression for gemma2 (torch). Exclude the guitar test from being copybara-ed.
#352
opened Nov 12, 2024 by
copybara-service
bot
Loading…
Adds InferenceRunner class to execute the Stable Diffusion denoiser model via an InferenceCalculator instance.
#318
opened Oct 25, 2024 by
copybara-service
bot
Loading…
Previous Next
ProTip!
Updated in the last three days: updated:>2025-05-06.