Skip to content
Merged
Changes from 1 commit
Commits
Show all changes
768 commits
Select commit Hold shift + click to select a range
8a89348
Update attention_sink.py
danielhanchen Sep 24, 2025
6d9b66b
Update gpt_oss.py
danielhanchen Sep 24, 2025
226c866
prefer_nd_tiling
danielhanchen Sep 24, 2025
2cebcc9
Update patching_utils.py
danielhanchen Sep 24, 2025
f3c5e1f
flex_attention_with_sink
danielhanchen Sep 24, 2025
e485dfc
Compile Flex Attention
danielhanchen Sep 24, 2025
6bd3f70
Update mxfp4.py
danielhanchen Sep 24, 2025
2393dfb
Update mxfp4.py
danielhanchen Sep 24, 2025
44356be
Update mxfp4.py
danielhanchen Sep 24, 2025
d5eebbc
Update mxfp4.py
danielhanchen Sep 24, 2025
93b5b88
Update gpt_oss.py
danielhanchen Sep 24, 2025
5e43a2a
bitsandbytes patch
danielhanchen Sep 24, 2025
2f0acb1
Update bitsandbytes.py
danielhanchen Sep 24, 2025
ebaf9b3
Update gpt_oss.py
danielhanchen Sep 24, 2025
2db0323
Inplace ops
danielhanchen Sep 24, 2025
031d21a
Update gpt_oss.py
danielhanchen Sep 24, 2025
61bb5aa
has_static_cache
danielhanchen Sep 24, 2025
267ab06
Update gpt_oss.py
danielhanchen Sep 24, 2025
bf825b1
Update gpt_oss.py
danielhanchen Sep 24, 2025
bf65ea5
Update gpt_oss.py
danielhanchen Sep 24, 2025
5c22d92
Update gpt_oss.py
danielhanchen Sep 24, 2025
274f7be
Update attention_sink.py
danielhanchen Sep 24, 2025
b2ec9f6
Update gpt_oss.py
danielhanchen Sep 24, 2025
ed572b1
Update gpt_oss.py
danielhanchen Sep 24, 2025
0bdaf45
Update gpt_oss.py
danielhanchen Sep 24, 2025
9fdf256
Update gpt_oss.py
danielhanchen Sep 24, 2025
56f7a73
Update gpt_oss.py
danielhanchen Sep 24, 2025
0c5437e
Update attention_sink.py
danielhanchen Sep 24, 2025
619c462
Update attention_sink.py
danielhanchen Sep 24, 2025
5d87949
Update rl_replacements.py
danielhanchen Sep 24, 2025
7ba642a
Update rl_replacements.py
danielhanchen Sep 24, 2025
040c6f2
Update rl_replacements.py
danielhanchen Sep 24, 2025
96798d8
Update gpt_oss.py
danielhanchen Sep 24, 2025
138f9f7
Update gpt_oss.py
danielhanchen Sep 24, 2025
eb19db9
Update gpt_oss.py
danielhanchen Sep 24, 2025
1f4f0c7
torch compile
danielhanchen Sep 25, 2025
b4afc0a
Update attention_sink.py
danielhanchen Sep 25, 2025
3d2083b
Update common.py
danielhanchen Sep 25, 2025
475a1fa
Update common.py
danielhanchen Sep 25, 2025
a1577f3
Patches
danielhanchen Sep 25, 2025
dc8308b
Compiled mask creation
danielhanchen Sep 25, 2025
15ae568
Update attention_sink.py
danielhanchen Sep 25, 2025
c849066
Update gpt_oss.py
danielhanchen Sep 25, 2025
eb68b54
Update gpt_oss.py
danielhanchen Sep 25, 2025
b4433b0
Revert
danielhanchen Sep 25, 2025
5f0fa7e
Update gpt_oss.py
danielhanchen Sep 25, 2025
274c830
Update gpt_oss.py
danielhanchen Sep 25, 2025
0c52d58
Fix up
danielhanchen Sep 25, 2025
3d9f498
Update attention_sink.py
danielhanchen Sep 25, 2025
dfe12c5
Update attention_sink.py
danielhanchen Sep 25, 2025
02ec222
Update utils.py
danielhanchen Sep 25, 2025
4e57162
Update attention_sink.py
danielhanchen Sep 25, 2025
17a6427
Update attention_sink.py
danielhanchen Sep 25, 2025
2002c9c
Retry
danielhanchen Sep 25, 2025
1ee8d5e
Update gpt_oss.py
danielhanchen Sep 25, 2025
3994e3c
Update gpt_oss.py
danielhanchen Sep 25, 2025
ef81921
Fix Flex
danielhanchen Sep 25, 2025
8cc0e77
Update gpt_oss.py
danielhanchen Sep 25, 2025
31f1624
Update gpt_oss.py
danielhanchen Sep 25, 2025
27fc0a9
Update gpt_oss.py
danielhanchen Sep 25, 2025
e86c541
Update gpt_oss.py
danielhanchen Sep 25, 2025
b4596cc
Update gpt_oss.py
danielhanchen Sep 25, 2025
858b962
Update gpt_oss.py
danielhanchen Sep 25, 2025
b676650
Update gpt_oss.py
danielhanchen Sep 25, 2025
dc1bd58
Update gpt_oss.py
danielhanchen Sep 25, 2025
1fe5a69
Update gpt_oss.py
danielhanchen Sep 25, 2025
524ac7f
Update gpt_oss.py
danielhanchen Sep 25, 2025
bd34939
Update gpt_oss.py
danielhanchen Sep 25, 2025
935ea71
Update gpt_oss.py
danielhanchen Sep 25, 2025
3ea5482
Update gpt_oss.py
danielhanchen Sep 25, 2025
1885f31
Update gpt_oss.py
danielhanchen Sep 25, 2025
ecd9b53
Update gpt_oss.py
danielhanchen Sep 25, 2025
d3b65af
Update gpt_oss.py
danielhanchen Sep 25, 2025
3b75bc9
Update gpt_oss.py
danielhanchen Sep 25, 2025
b43c1b5
Update gpt_oss.py
danielhanchen Sep 25, 2025
db12a8a
Update gpt_oss.py
danielhanchen Sep 25, 2025
889b4fb
Update gpt_oss.py
danielhanchen Sep 25, 2025
f481e2f
Update gpt_oss.py
danielhanchen Sep 25, 2025
c3e3a90
Update gpt_oss.py
danielhanchen Sep 25, 2025
b721c77
Update gpt_oss.py
danielhanchen Sep 25, 2025
7d81867
Update gpt_oss.py
danielhanchen Sep 25, 2025
577a2a0
Update gpt_oss.py
danielhanchen Sep 25, 2025
c0e421b
Update gpt_oss.py
danielhanchen Sep 25, 2025
2605ecb
Update gpt_oss.py
danielhanchen Sep 25, 2025
e850c7d
Update gpt_oss.py
danielhanchen Sep 25, 2025
9af4313
Update gpt_oss.py
danielhanchen Sep 25, 2025
d8a4e50
Update gpt_oss.py
danielhanchen Sep 25, 2025
1b732ba
Update gpt_oss.py
danielhanchen Sep 25, 2025
666f121
Update gpt_oss.py
danielhanchen Sep 25, 2025
b8cfebf
Update gpt_oss.py
danielhanchen Sep 25, 2025
5e88a87
Update gpt_oss.py
danielhanchen Sep 25, 2025
70dfc00
Update gpt_oss.py
danielhanchen Sep 25, 2025
9128339
Update gpt_oss.py
danielhanchen Sep 25, 2025
082cfb7
Update gpt_oss.py
danielhanchen Sep 25, 2025
0f47e5e
Update gpt_oss.py
danielhanchen Sep 25, 2025
d92e62d
Update gpt_oss.py
danielhanchen Sep 25, 2025
5646157
Update gpt_oss.py
danielhanchen Sep 25, 2025
272689b
Update gpt_oss.py
danielhanchen Sep 25, 2025
d10fc7a
Bug fixes
danielhanchen Sep 26, 2025
4396a93
Update patching_utils.py
danielhanchen Sep 26, 2025
ee50724
Update patching_utils.py
danielhanchen Sep 26, 2025
abe89f0
Update patching_utils.py
danielhanchen Sep 26, 2025
edc85ca
Update rl_replacements.py
danielhanchen Sep 26, 2025
efb18b5
Update patching_utils.py
danielhanchen Sep 26, 2025
f16a5a8
Update patching_utils.py
danielhanchen Sep 26, 2025
0dae9dd
Update patching_utils.py
danielhanchen Sep 26, 2025
435de2d
flash attn
danielhanchen Sep 26, 2025
9cd630c
Update gpt_oss.py
danielhanchen Sep 26, 2025
c510029
Update __init__.py
danielhanchen Sep 26, 2025
98080fc
Update attention_sink.py
danielhanchen Sep 26, 2025
5625cfb
Update gpt_oss.py
danielhanchen Sep 26, 2025
62756a8
Update gpt_oss.py
danielhanchen Sep 26, 2025
3f9a9a9
Update gpt_oss.py
danielhanchen Sep 26, 2025
c32eb2e
Update gpt_oss.py
danielhanchen Sep 26, 2025
63a771c
Update gpt_oss.py
danielhanchen Sep 26, 2025
194ff92
Update gpt_oss.py
danielhanchen Sep 26, 2025
be54940
Update gpt_oss.py
danielhanchen Sep 26, 2025
9ebf49f
Update gpt_oss.py
danielhanchen Sep 26, 2025
2b45d36
dropout_p
danielhanchen Sep 26, 2025
7a6941a
Update gpt_oss.py
danielhanchen Sep 26, 2025
588c4f0
Update gpt_oss.py
danielhanchen Sep 26, 2025
aded049
Update attention_sink.py
danielhanchen Sep 26, 2025
33ba6b3
Update gpt_oss.py
danielhanchen Sep 26, 2025
b08753b
Update gpt_oss.py
danielhanchen Sep 26, 2025
9fe8ec0
fix
danielhanchen Sep 26, 2025
5be9e57
Update attention_sink.py
danielhanchen Sep 26, 2025
a218bfc
Update gpt_oss.py
danielhanchen Sep 26, 2025
9fc2694
Update gpt_oss.py
danielhanchen Sep 26, 2025
769301d
Update gpt_oss.py
danielhanchen Sep 26, 2025
d59f62b
Update gpt_oss.py
danielhanchen Sep 26, 2025
92d16d4
Update gpt_oss.py
danielhanchen Sep 26, 2025
0608531
Update gpt_oss.py
danielhanchen Sep 26, 2025
24bb593
Update gpt_oss.py
danielhanchen Sep 26, 2025
c481eb8
Update gpt_oss.py
danielhanchen Sep 26, 2025
68fed93
Update gpt_oss.py
danielhanchen Sep 26, 2025
9ff936f
Update gpt_oss.py
danielhanchen Sep 26, 2025
77343fa
Update gpt_oss.py
danielhanchen Sep 26, 2025
f3e7f8c
Update gpt_oss.py
danielhanchen Sep 26, 2025
5e7e7d3
Update gpt_oss.py
danielhanchen Sep 26, 2025
a508006
Update loss_utils.py
danielhanchen Sep 26, 2025
44e1de7
Update gpt_oss.py
danielhanchen Sep 26, 2025
1079a21
Update gpt_oss.py
danielhanchen Sep 26, 2025
58e5f24
Update gpt_oss.py
danielhanchen Sep 26, 2025
3c61724
Update gpt_oss.py
danielhanchen Sep 26, 2025
bd50ca4
Update gpt_oss.py
danielhanchen Sep 26, 2025
5f8b77c
Update gpt_oss.py
danielhanchen Sep 26, 2025
f2fe3db
Update gpt_oss.py
danielhanchen Sep 26, 2025
04bbc07
Update loss_utils.py
danielhanchen Sep 26, 2025
cb16066
Update gpt_oss.py
danielhanchen Sep 26, 2025
75d7829
Update gpt_oss.py
danielhanchen Sep 26, 2025
679e882
Update gpt_oss.py
danielhanchen Sep 26, 2025
4b61795
Merge branch 'main' into nightly
danielhanchen Sep 26, 2025
c37dff1
Merge branch 'main' into nightly
danielhanchen Sep 26, 2025
b61346a
Merge branch 'main' into nightly
danielhanchen Sep 26, 2025
a8d6aa8
Merge branch 'main' into nightly
danielhanchen Sep 28, 2025
5225692
Update gpt_oss.py
danielhanchen Sep 28, 2025
02326ab
Update gpt_oss.py
danielhanchen Sep 28, 2025
2210555
Update gpt_oss.py
danielhanchen Sep 30, 2025
f7406a4
Update gpt_oss.py
danielhanchen Sep 30, 2025
7020561
Update gpt_oss.py
danielhanchen Sep 30, 2025
e316226
Update gpt_oss.py
danielhanchen Sep 30, 2025
55a0f94
Update gpt_oss.py
danielhanchen Sep 30, 2025
d241d8d
Versioning
danielhanchen Sep 30, 2025
8d752f6
Merge branch 'main' into nightly
danielhanchen Oct 1, 2025
7c40a85
Update saving_utils.py
danielhanchen Oct 5, 2025
114feed
Update saving_utils.py
danielhanchen Oct 5, 2025
5bdbffe
Update saving_utils.py
danielhanchen Oct 5, 2025
79115db
Update saving_utils.py
danielhanchen Oct 5, 2025
51e3889
Update saving_utils.py
danielhanchen Oct 5, 2025
3284083
Update saving_utils.py
danielhanchen Oct 5, 2025
289abf2
Update saving_utils.py
danielhanchen Oct 5, 2025
efe6d76
Update saving_utils.py
danielhanchen Oct 5, 2025
2f5e342
Fix Gemma 3
danielhanchen Oct 5, 2025
3237c4b
Update misc.py
danielhanchen Oct 5, 2025
dc3e28e
Merge branch 'main' into nightly
danielhanchen Oct 5, 2025
22b3cb6
Merge branch 'main' into nightly
danielhanchen Oct 14, 2025
5beb515
Merge branch 'main' into nightly
danielhanchen Oct 16, 2025
bd43a5b
Update rl_environments.py
danielhanchen Oct 17, 2025
9571b67
Update pyproject.toml
danielhanchen Oct 17, 2025
f789e3b
Update rl_environments.py
danielhanchen Oct 17, 2025
c146ca2
Update __init__.py
danielhanchen Oct 17, 2025
5012df2
Merge branch 'main' into nightly
danielhanchen Oct 17, 2025
80f4b15
Merge branch 'main' into nightly
danielhanchen Oct 17, 2025
6857125
Update empty_model.py
danielhanchen Oct 17, 2025
49f3cd0
Update empty_model.py
danielhanchen Oct 17, 2025
7642fbc
Update empty_model.py
danielhanchen Oct 17, 2025
a6a9a53
Merge branch 'main' into nightly
danielhanchen Oct 17, 2025
565d37f
Merge branch 'main' into nightly
danielhanchen Oct 17, 2025
068142c
Merge branch 'main' into nightly
danielhanchen Oct 19, 2025
9b06516
Merge branch 'main' into nightly
danielhanchen Oct 19, 2025
33a55fc
Merge branch 'main' into nightly
danielhanchen Oct 20, 2025
9f9fad5
Update empty_model.py
danielhanchen Oct 20, 2025
c62f0db
Device type
danielhanchen Oct 20, 2025
44539dc
Update vllm_utils.py
danielhanchen Oct 20, 2025
c7f1a85
Update compiler.py
danielhanchen Oct 20, 2025
d98b8dd
Update empty_model.py
danielhanchen Oct 20, 2025
7dccb4f
Update vllm_utils.py
danielhanchen Oct 20, 2025
96b12f6
Update empty_model.py
danielhanchen Oct 20, 2025
b900605
Fixes
danielhanchen Oct 20, 2025
be24a86
Update empty_model.py
danielhanchen Oct 20, 2025
09a56e1
Update empty_model.py
danielhanchen Oct 20, 2025
dd3f5a9
Update __init__.py
danielhanchen Oct 20, 2025
5e914a5
Update vllm_utils.py
danielhanchen Oct 20, 2025
d45333a
Update vllm_utils.py
danielhanchen Oct 20, 2025
aef0696
Update rl_environments.py
danielhanchen Oct 20, 2025
4bbede7
Update cross_entropy_loss.py
danielhanchen Oct 20, 2025
03adb63
Update vllm_utils.py
danielhanchen Oct 20, 2025
4e0786b
Update vllm_utils.py
danielhanchen Oct 20, 2025
21a4404
Update rl_environments.py
danielhanchen Oct 20, 2025
e63cd7b
Update vllm_utils.py
danielhanchen Oct 20, 2025
855d572
Merge branch 'main' into nightly
danielhanchen Oct 20, 2025
60b28fa
Merge branch 'main' into nightly
danielhanchen Oct 22, 2025
f34d525
Merge branch 'main' into nightly
danielhanchen Oct 23, 2025
26fe13e
Merge branch 'main' into nightly
danielhanchen Oct 27, 2025
ac90015
Merge branch 'main' into nightly
danielhanchen Oct 27, 2025
113c8d3
Merge branch 'main' into nightly
danielhanchen Oct 30, 2025
bb81b69
Qwen3 VL vLLM (#324)
Datta0 Oct 31, 2025
0632308
Update __init__.py
danielhanchen Oct 31, 2025
fe09bfd
Update __init__.py
danielhanchen Oct 31, 2025
a5102af
Update __init__.py
danielhanchen Oct 31, 2025
d2fcf41
Update __init__.py
danielhanchen Oct 31, 2025
8b07dcf
Update __init__.py
danielhanchen Oct 31, 2025
6d43f0d
Update __init__.py
danielhanchen Oct 31, 2025
ad18827
Update __init__.py
danielhanchen Oct 31, 2025
c00681e
Merge branch 'main' into nightly
danielhanchen Nov 2, 2025
9321399
Update vllm_utils.py
danielhanchen Nov 2, 2025
32ca2c0
Update vllm_utils.py
danielhanchen Nov 2, 2025
45a2f69
Update pyproject.toml
danielhanchen Nov 2, 2025
6c6c4e8
Update vllm_utils.py
danielhanchen Nov 2, 2025
3a1a097
Update vllm_utils.py
danielhanchen Nov 2, 2025
ed24866
Update vllm_utils.py
danielhanchen Nov 2, 2025
c9b3186
Update vllm_utils.py
danielhanchen Nov 3, 2025
64395ac
Update vllm_utils.py
danielhanchen Nov 3, 2025
60de923
Update vllm_utils.py
danielhanchen Nov 3, 2025
0b339f4
Update __init__.py
danielhanchen Nov 3, 2025
5ae18ab
Update compiler.py
danielhanchen Nov 3, 2025
dac460f
Update __init__.py
danielhanchen Nov 3, 2025
2be0308
Merge branch 'main' into nightly
danielhanchen Nov 3, 2025
3af0006
Merge branch 'main' into nightly
danielhanchen Nov 4, 2025
f59cc91
Merge branch 'main' into nightly
danielhanchen Nov 4, 2025
9823f20
Update vllm_utils.py
danielhanchen Nov 4, 2025
20e9c96
Update rl_replacements.py
danielhanchen Nov 4, 2025
10b7094
Update rl_replacements.py
danielhanchen Nov 4, 2025
7e3d33a
Update rl_replacements.py
danielhanchen Nov 4, 2025
9567999
Merge branch 'main' into nightly
danielhanchen Nov 4, 2025
18a4852
Fix CE compile
danielhanchen Nov 6, 2025
61095c6
Update loss_utils.py
danielhanchen Nov 6, 2025
0c3c555
Update cross_entropy_loss.py
danielhanchen Nov 6, 2025
acf03d2
Fix
danielhanchen Nov 6, 2025
80d8e09
Deepseekocr fix: save single model shard (#346)
mmathew23 Nov 6, 2025
File filter

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
Prev Previous commit
Next Next commit
Update attention_sink.py
  • Loading branch information
danielhanchen committed Sep 25, 2025
commit 17a642769b6fbf78339e0015a8228acd1e6f2c1c
4 changes: 2 additions & 2 deletions unsloth_zoo/flex_attention/attention_sink.py
Original file line number Diff line number Diff line change
Expand Up @@ -182,8 +182,8 @@ def flex_attention_with_sink(
mask_mod = None
block_mask = None
has_flex_cache = hasattr(self_attn, "_flex_attention_cache")
has_flex_cache = False
is_training = True
# has_flex_cache = False
# is_training = True
# Handle inference and training
if has_static_cache:
if is_training or (
Expand Down