[cache refactor] Move all the caching logic to a per-layer approach #39106
Squash for refactor: Replace monolithic cache classes with modular La…
1c3cbccc
fix quantized, add tests
04d7a0b1
manueldeprada
force pushed
from
a2fe24c8
to
04d7a0b1
246 days ago
remove CacheProcessorList
26c28af6
manueldeprada
force pushed
from
d97a02da
to
26c28af6
246 days ago
raushan review, arthur review
16a66240
gante
commented
on 2025-07-02
joao review: minor things
aec9ccd6
manueldeprada
force pushed
from
c1d93472
to
52ac7934
243 days ago
manueldeprada
force pushed
from
52ac7934
to
d6843398
243 days ago
manueldeprada
force pushed
from
d6843398
to
4c03e0f3
242 days ago
remove cache configs, make CacheLayer a mixin (joaos review)
e80c68a6
manueldeprada
force pushed
from
88089304
to
e80c68a6
242 days ago
manueldeprada
force pushed
from
65fbce68
to
a21b2f74
237 days ago
manueldeprada
force pushed
from
a21b2f74
to
c538ad75
237 days ago
manueldeprada
force pushed
from
c538ad75
to
5dc5fb40
237 days ago
manueldeprada
force pushed
from
5dc5fb40
to
a6b7562b
237 days ago
back to storage inside Cache()
27916bc2
manueldeprada
force pushed
from
a6b7562b
to
27916bc2
237 days ago
remove cachebase for decorator
fd83e14b
manueldeprada
force pushed
from
9e37266d
to
fd83e14b
237 days ago
manueldeprada
force pushed
from
6c0dd19f
to
129119dd
237 days ago
no more __getattr__
c2004471
manueldeprada
force pushed
from
129119dd
to
c2004471
237 days ago
Merge branch 'main' of github.com:huggingface/transformers into cache…
f3275705
fix tests
5b1b1f17
manueldeprada
force pushed
from
295c8d8a
to
5b1b1f17
237 days ago
gante
commented
on 2025-07-10
manueldeprada
force pushed
from
ba1d14ed
to
5b1b1f17
236 days ago
joaos review except docs
58dbcfe2
manueldeprada
force pushed
from
2b4e7bf7
to
58dbcfe2
236 days ago
fix ast deprecations for python 3.14: replace node.n by node.value an…
0c6d2ff6
manueldeprada
force pushed
from
b22b1179
to
87f399fd
233 days ago
Revert "back to storage inside Cache()"
6a77408a
manueldeprada
force pushed
from
87f399fd
to
6a77408a
233 days ago
cyril review
13ec4a44
simplify cache export
7029a90c
Merge branch 'main' of github.com:huggingface/transformers into cache…
5fa99012
fix lfm2 cache
dd7458b5
HybridChunked to layer
dc08253c
manueldeprada
force pushed
from
d434a998
to
dc08253c
232 days ago
BC proxy object for cache.key_cache[i]=...
a9521249
reorder classes
dbbc4d51
bfff come on LFM2
4bb48fcf
better tests for hybrid and hybridChunked
00b1f96a
Merge branch 'main' of github.com:huggingface/transformers into cache…
def346eb
complete coverage for hybrid chunked caches (prefill chunking)
38e86034
reimplementing HybridChunked
34a3022f
cyril review
42226536
Merge branch 'main' of github.com:huggingface/transformers into cache…
063459c5
manueldeprada
force pushed
from
51d1505b
to
83a6016b
230 days ago
fix ci
1acc648f
manueldeprada
force pushed
from
83a6016b
to
1acc648f
230 days ago
docs for cache refactor
ca39ffec
manueldeprada
force pushed
from
bd4c9a41
to
ca39ffec
229 days ago
docs
731d0b73
Merge branch 'main' of github.com:huggingface/transformers into cache…
574b8205
oopsie
a4794705
oopsie
9c0bdcc5
Merge branch 'main' into cache-refactor-1
04091ed8
Merge branch 'main' of github.com:huggingface/transformers into cache…
8642bb6b
Merge branch 'cache-refactor-1' of https://github.com/manueldeprada/t…
83968bd9
fix after merge
0c4700dc
Cyrilvallez
changed the title [cache refactor] PR 1/3: layered caches with full backward compatibility [cache refactor] Move all the caching logic to a per-layer approach 225 days ago
cyril review
b3a35e92
arthur review
e4878ad0
Merge branch 'main' of github.com:huggingface/transformers into cache…
38fb99d4
opsie
8df15953
fix lfm2
ad65a02d
opsie2
d9fbb044
Assignees
No one assigned
Login to write a write a comment.
Login via GitHub