Skip to content

Commit 8e4158c

Browse files
authored
Multipack parallel bin packing (axolotl-ai-cloud#2631)
* improve readability of multipack sampler * parallel bin packing fix error with lambda and pickling make sure things are in float instead of np.float * annotations and comments update * support for configurable group and bin size for sample packing * fix missing map back to original indices
1 parent cd84325 commit 8e4158c

File tree

2 files changed

+284
-125
lines changed

2 files changed

+284
-125
lines changed

src/axolotl/core/trainers/base.py

Lines changed: 2 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -114,6 +114,8 @@ def _create_multipack_sampler(
114114
packing_efficiency_estimate=self.args.sample_packing_efficiency,
115115
batch_max_len=batch_max_len,
116116
batch_size=batch_size,
117+
group_size=self.args.sample_packing_group_size,
118+
bin_size=self.args.sample_packing_bin_size,
117119
sequential=self.args.sample_packing_sequentially,
118120
drop_last=True,
119121
)

0 commit comments

Comments
 (0)