A

apple/axlearn

Settings | Report Duplicate

0

I Use This!

High Activity

Commits : Listings

Analyzed about 4 hours ago. based on code collected about 4 hours ago.

Commit Message	Contributor	Files Modified	Lines Added	Lines Removed	Code Location	Date
Aug 17, 2024 — Aug 17, 2025 Showing page 8 of 38 Search / Filter on:
The Speech frontend uses chunking more intelligently. (#1102)	Dongseong Hwang	More...				4 months ago
Remove einops dependency by reimplementing minimal functionality for JAX (#1103)	Dongseong Hwang	More...				4 months ago
More flexible loss weights in composite metrics. (#1100)	Mark Lee	More...				4 months ago
fix nodeSelectors when not using bastion (#1096)	Sam Stoelinga	More...				4 months ago
fix black formatting	Steboss	More...				4 months ago
Refactor conftest.py (#1090)	Boyue Li	More...				5 months ago
Clarify that the k_proj/v_proj's dtype maybe changed by the kv_cache. (#1093)	Dongseong Hwang	More...				5 months ago
Update axlearn/common/array_serialization.py	Steboss	More...				5 months ago
Add 7B sigmoid (#1094)	Floris Weers	More...				5 months ago
update jax.tree_map to jax.tree_util.tree_map	Steboss	More...				5 months ago
Beam on Flink user feedback improvements (#1087)	muyangyuapple	More...				5 months ago
Fix GPU decoding dtype and flash tests (#1092)	Hanzhi Zhou	More...				5 months ago
Optimize deserialize -- up to 5x faster (#1091)	Hanzhi Zhou	More...				5 months ago
relax requirement of remote "origin" being present when collecting GitSummary (#1089)	mmllee	More...				5 months ago
Implement set_double_shard_weights_config() for Conformer like Transformer. (#1088)	Dongseong Hwang	More...				5 months ago
Simplifies the base job API. (#1086)	Mark Lee	More...				5 months ago
fix repeat/n in generator (#1084)	Guoli Yin	More...				5 months ago
Remove `einops` dependency in attention (#1085)	Chang Lan	More...				5 months ago
Introduces utilities for simplifying FlagConfigurables. (#1079)	Mark Lee	More...				5 months ago
Splash Attention creates a NumPy mask inside jax.ensure_compile_time_eval() (#1081)	Dongseong Hwang	More...				5 months ago
Remove assert_equal_shape((k_proj, v_proj)) in attention. (#1083)	Dongseong Hwang	More...				5 months ago
fix pre-commit error	Steboss	More...				5 months ago
modify primitives	Steboss	More...				5 months ago
Removes old bastion vm and misc fixes. (#1078)	Mark Lee	More...				5 months ago
Consolidate QuotaInfo and UserQuotaInfo into one class (#1080)	Chang Liu	More...				5 months ago
Don't set location_hint if it is None (#1075)	muyangyuapple	More...				5 months ago
Flash attention fallbacks to standard attention, not ReferenceMHA. (#1077)	Dongseong Hwang	More...				5 months ago
Make `attention_logit_biases` in init_states/extend_step Optional (#1076)	Dongseong Hwang	More...				5 months ago
Improvements and fixes to gradient accumulation (#993)	Apoorv Gupta	More...				5 months ago
Disables hf test. (#1073)	Mark Lee	More...				5 months ago

←
1
2
3
4
5
6
7
8
9
10
11
12
…
37
38
→