Skip to content

Pull requests: AI-Hypercomputer/maxtext

New pull request New
116 Open 907 Closed
116 Open 907 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Reviews
Filter by reviews
No reviews Review required Approved review Changes requested
Assignee
Filter by who’s assigned
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

manually test swa
#1126 opened Dec 24, 2024 by gobbleturk Loading…
4 tasks done
Add option to skip initializing the jax distributed system
#1125 opened Dec 23, 2024 by gobbleturk Loading…
4 tasks done
2
Fix orbax to hf converter for Llama3.1-8B
#1123 opened Dec 23, 2024 by khatwanimohit Loading…
4 tasks done
@A9isha
4
Better save message for async checkpoint saving pull ready
#1122 opened Dec 23, 2024 by gobbleturk Loading…
4 tasks done
@anfals @khatwanimohit
Add option to dump hlo
#1121 opened Dec 23, 2024 by gobbleturk Loading…
4 tasks done
@hengtaoguo @raymondzouu
Add smaler v5e topologies to train_compile
#1119 opened Dec 21, 2024 by gobbleturk Loading…
4 tasks done
@raymondzouu
7
Incorporate Orbax emergency replicator checkpoint manager
#1117 opened Dec 21, 2024 by xuefgu • Draft
4 tasks done
Temporary solution to remove the unnecessary ar cache during prefill
#1115 opened Dec 20, 2024 by zhihaoshan-google • Draft
4 tasks done
Refactor and parallelize MaxText test runner
#1113 opened Dec 20, 2024 by shralex Loading…
4 tasks done
1
Get all tests to pass locally with no special configuration
#1108 opened Dec 19, 2024 by SamuelMarks Loading…
4 tasks done
1
2
Anisha ckpt2hf1
#1106 opened Dec 19, 2024 by A9isha • Draft
4 tasks
Fix typo in trillium llama2-70b script
#1105 opened Dec 19, 2024 by wenxindongwork Loading…
4 tasks done
Add a Ray trainer for MaxText
#1098 opened Dec 13, 2024 by richardsliu Loading…
Add llama-405b configuration for v5p
#1095 opened Dec 10, 2024 by suexu1025 Loading…
4 tasks done
1
Add Pathways Support to Benchmark Runner
#1094 opened Dec 10, 2024 by SujeethJinesh Loading…
24
Add Dockerfile and entrypoint script for the maxengine-server image pull ready
#1092 opened Dec 10, 2024 by vivianrwu Loading…
4 tasks done
1
Add mixtral 8x7b config for gpu
#1090 opened Dec 9, 2024 by michelle-yooh Loading…
4 tasks done
@michelle-yooh @gobbleturk
1
Enable c4/en, c4/multilingual-en dataset and convergence configs for llama convergence tests
#1080 opened Dec 3, 2024 by suexu1025 • Draft
4 tasks done
1
[DO NOT MERGE] Pathways llama 405b v5p config
#1077 opened Dec 3, 2024 by SujeethJinesh • Draft
Manual test running for mem stats pull ready
#1074 opened Dec 3, 2024 by gobbleturk Loading…
4 tasks done
Add Pallas GPU decode attention in Maxtext inference
#1066 opened Nov 26, 2024 by tohaowu Loading…
4 tasks done
9
Quantize megablox
#1062 opened Nov 25, 2024 by lenscloth Loading…
4 tasks done
40
[DO NOT MERGE] verify fix
#1045 opened Nov 16, 2024 by RissyRan • Draft
Add Llama2-70b sparsecore collective model to trillium configs
#1042 opened Nov 15, 2024 by Obliviour Loading…
4 tasks done
Enable pathways workloads for v6e benchmarks
#1040 opened Nov 15, 2024 by sadikneipp Loading…
3
ProTip! Mix and match filters to narrow down what you’re looking for.