transformers
[`Add Mamba`] Adds support for the `Mamba` models
#28094
Merged

[`Add Mamba`] Adds support for the `Mamba` models #28094

ArthurZucker merged 123 commits into main from add-mamba
ArthurZucker
ArthurZucker initial-commit
81c642f0
huggingface huggingface deleted a comment from github-actions on 2024-01-16
ArthurZucker
ArthurZucker
ArthurZucker Merge branch 'main' of github.com:huggingface/transformers into add-m…
c50602bb
HuggingFaceDocBuilderDev
ArthurZucker start cleaning
00d3a6c1
ArthurZucker small nits
921bb24a
apoorvkh
ArthurZucker
ArthurZucker small nits
b3f216d2
ArthurZucker current updates
7235b57f
ArthurZucker add kernels
7a407a76
ArthurZucker small refactoring little step
9f2a9829
ArthurZucker add comments
04c991ab
ArthurZucker styling
aa7e8d2b
ArthurZucker nit
26748c4a
ArthurZucker
ArthurZucker nits
75e376a3
ArthurZucker Style
1c104b51
ArthurZucker Merge
0e90daed
ArthurZucker Small changes
a8044665
ArthurZucker Push dummy mambda simple slow
6b87ad2c
ArthurZucker nit
a7ec8d63
ArthurZucker Use original names
50464518
ArthurZucker Use original names and remove norm
b5831e3d
ArthurZucker Updates for inference params
e9a80ad8
ArthurZucker Style nd updates
ee4a7ef0
ArthurZucker nits
d8c195fb
ArthurZucker Match logits
e64fedc2
ArthurZucker Add a test
aee558f3
ArthurZucker Add expected generated text
eae5f452
ArthurZucker nits doc, imports and styling
1f8e8d0a
ArthurZucker style
3cc06e5b
ArthurZucker oups
5a5324c2
ArthurZucker Merge branch 'main' of github.com:huggingface/transformers into add-m…
325b66b5
ArthurZucker dont install kernels, invite users to install the required kernels
81303f4d
ArthurZucker let use use the original packages
1a103101
ArthurZucker styling
89fb490b
ArthurZucker nits
6cfe216c
ArthurZucker fix some copieds
1ecbd223
ArthurZucker update doc
b937122e
ArthurZucker fix-copies
9752dd03
ArthurZucker styling done
a7881a3c
ArthurZucker nits
f445b0da
ArthurZucker fix import check
64ec8dd6
ArthurZucker run but wrong cuda ress
e6e3ba8c
ArthurZucker mamba CUDA works :)
ed4eb4c8
ArthurZucker fix the fast path
4c8fc48c
ArthurZucker config naming nits
69e103fa
ArthurZucker conversion script is not required at this stage
ba21ff24
ArthurZucker finish fixing the fast path: generation make sense now!
fe537285
ArthurZucker nit
9411169b
ArthurZucker Let's start working on the CIs
c2c77096
ArthurZucker style
1e73ca91
ArthurZucker git push Merge branch 'main' of github.com:huggingface/transformers i…
834f46ff
ArthurZucker Merge branch 'main' of github.com:huggingface/transformers into add-m…
a1a94f32
ArthurZucker better style
22132229
ArthurZucker more nits
2a020066
ArthurZucker test nit
8b0412f3
ArthurZucker quick fix for now
fbd6a2c0
ArthurZucker nits
823f11a7
ArthurZucker nit
88896a9a
ArthurZucker nit
7f72ee81
relic-yuexi
ArthurZucker
relic-yuexi
ArthurZucker
ArthurZucker
ArthurZucker
ArthurZucker Merge branch 'main' of github.com:huggingface/transformers into add-m…
05552474
ArthurZucker nit
0072a6c2
ArthurZucker nits
7f6c56f4
ArthurZucker update test rest
f67c3533
ArthurZucker fixup
2ab5a865
ArthurZucker update test
8920be32
ArthurZucker nit
87d0664f
ArthurZucker some fixes
8b00d768
ArthurZucker nits
ca9835cf
ArthurZucker update test values
796ef3ef
ArthurZucker fix styling
170664ac
ArthurZucker nit
92493a05
relic-yuexi
relic-yuexi requested changes on 2024-02-29
ArthurZucker support peft
854ebad5
ArthurZucker
ArthurZucker Merge branch 'main' of github.com:huggingface/transformers into add-m…
3bbd1b14
relic-yuexi
relic-yuexi commented on 2024-02-29
ArthurZucker integrations tests require torchg
aa0e6bb3
ArthurZucker also add slow markers
3c1537ef
ArthurZucker styling
d06421a6
ArthurZucker
relic-yuexi
ArthurZucker
ArthurZucker chose forward wisely
5fb80623
relic-yuexi
ArthurZucker nits
edb4e91a
ArthurZucker
relic-yuexi
ArthurZucker update tests
eb1fb640
ArthurZucker
ArthurZucker fix gradient checkpointing
de4fe46e
ArthurZucker fixup
54ffaa3e
ArthurZucker nit
977d34f0
ArthurZucker fix doc
0928453b
ArthurZucker check copies
2c90536a
ArthurZucker fix the docstring
4ba9c792
ArthurZucker fix some more tests
3651dbad
ArthurZucker style
426e6f39
ArthurZucker ArthurZucker marked this pull request as ready for review 2 years ago
ArthurZucker fix beam search
951b1aab
huggingface huggingface deleted a comment from relic-yuexi on 2024-03-01
huggingface huggingface deleted a comment from relic-yuexi on 2024-03-01
huggingface huggingface deleted a comment from relic-yuexi on 2024-03-01
huggingface huggingface deleted a comment from relic-yuexi on 2024-03-01
huggingface huggingface deleted a comment from relic-yuexi on 2024-03-01
ArthurZucker add init schene
4101369c
ArthurZucker update
65db96bd
ArthurZucker nit
0f3dfc71
ArthurZucker fix
f8bd0aa8
ArthurZucker fixup the doc
b2bd0c78
ArthurZucker fix the doc
cf585294
ArthurZucker fixup
e9c34472
ArthurZucker tentative update but slow is no longer good
1282a75f
ArthurZucker nit
fa561b26
ArthurZucker should we always use float32?
91b81061
ArthurZucker nits
e8142caf
ArthurZucker revert wrong changes
623b6361
ArthurZucker res in float32
566c799c
ArthurZucker
ArthurZucker cleanup
5d637d9f
ArthurZucker skip fmt for now
648a2922
ArthurZucker update generation values
e306e891
ArthurZucker update test values running original model
057d7a3d
ArthurZucker fixup
72f8936c
ArthurZucker update tests + rename inference_params to cache_params + make sure tr…
f415081d
ArthurZucker small nits
6bb659a8
ArthurZucker more nits
178fe76b
ArthurZucker fix final CIs
3a46724e
ArthurZucker style
13204e08
ArthurZucker nit doc
1608a905
ArthurZucker I hope final doc nits
99119ba2
ArthurZucker nit
d6fb1efa
ArthurZucker 🫠
844530fd
ArthurZucker ArthurZucker requested a review from LysandreJik LysandreJik 2 years ago
ArthurZucker final touch!
52be0185
ArthurZucker fix torch import
d03de1c1
LysandreJik
LysandreJik commented on 2024-03-04
ArthurZucker Apply suggestions from code review
c0672a8b
ArthurZucker
ArthurZucker commented on 2024-03-05
ArthurZucker Apply suggestions from code review
dfc1212d
ArthurZucker fix fix and fix
acd4ccf1
ArthurZucker fix base model prefix!
2ddd9aad
ArthurZucker nit
0c5d7eda
ArthurZucker Update src/transformers/models/mamba/__init__.py
28e5ef07
ArthurZucker ArthurZucker requested a review from LysandreJik LysandreJik 2 years ago
ArthurZucker
LysandreJik
LysandreJik approved these changes on 2024-03-05
ArthurZucker Update docs/source/en/model_doc/mamba.md
f963e381
ArthurZucker ArthurZucker force pushed 2 years ago
ArthurZucker ArthurZucker force pushed to f963e381 2 years ago
ArthurZucker nit
095dabd6
ArthurZucker ArthurZucker merged fb1c62e9 into main 2 years ago
ArthurZucker ArthurZucker deleted the add-mamba branch 2 years ago
abdulfatir
ArthurZucker
abdulfatir
ArthurZucker
lkurlandski
ArthurZucker
ArthurZucker

Login to write a write a comment.

Login via GitHub

Assignees
No one assigned
Labels
Milestone