Add Qwen2MoE #29377
ArthurZucker
merged
46 commits into
huggingface:main
from
bozheng-hit:qwen2_moe
add support for qwen2 MoE models
4f933bbf
update docs
8ad6c9e3
add support for qwen2 MoE models
fbce3b92
update docs
c32b9984
Merge branch 'qwen2_moe' of https://github.com/bozheng-hit/transforme…
8274f89c
update model name & test
e44f700f
update readme
b09e2ed8
update class names & readme & model_doc of Qwen2MoE.
d5e99a60
update architecture name
1625b1f5
ArthurZucker
commented
on 2024-03-06
fix qwen2_moe tests
051e19dc
use Qwen2Tokenizer instead of Qwen2MoeTokenizer
307d9de2
update modeling_qwen2_moe.py
4d80bf8c
fix model architecture
8b6d57b0
fix qwen2_moe tests
b9c2803f
use Qwen2Tokenizer instead of Qwen2MoeTokenizer
f8e18196
update modeling_qwen2_moe.py
e4b8445f
fix model architecture
8d74bb02
fix style
a50a208c
fix test when there are sparse and non sparse layers
a04c6983
fixup
dc53a8dc
Update README.md
8f55aa51
fix up
6a06f8e4
fixup
bf112272
fixup
e3038db8
add archive back
5c627d30
add support for qwen2 MoE models
765ebf53
update docs
1c973fb4
update model name & test
08417221
update readme
4c0b2b10
update class names & readme & model_doc of Qwen2MoE.
8958743f
update architecture name
1e099c5a
fix qwen2_moe tests
4906cdfa
use Qwen2Tokenizer instead of Qwen2MoeTokenizer
82729ec4
update modeling_qwen2_moe.py
a3aa52de
fix model architecture
0686cc67
fixup
c0740214
fix qwen2_moe tests
24846042
use Qwen2Tokenizer instead of Qwen2MoeTokenizer
5d1ed372
fix style
27afcd58
fix test when there are sparse and non sparse layers
0d155e9c
fixup
46b09184
add archive back
45219a16
fixup
cf61e7e1
fix integration test
3b9f3a8b
fixup
40778776
Merge branch 'main' into qwen2_moe
4d931f04
Login to write a write a comment.
Login via GitHub