Granite language models #31502
first commit
750ca7f4
drop tokenizer
3b267309
drop tokenizer
9c017b0d
drop tokenizer
876f4b55
Merge branch 'main' into granite
0f716ec3
drop convert
e3cdcaff
granite
3e4391e0
drop tokenization test
6f0cf35e
mup
2d1a58c6
fix
ac560ae7
reformat
78c81a00
reformat
3b6c7554
reformat
f46bf827
fix docs
272af5c2
stop checking for checkpoint
c9b22882
update support
19ec830a
attention multiplier
a9dba033
update model
df90fbda
tiny drop
c3369a0a
saibo drop
6a7c8142
skip test
dad1e4ae
fix test
5cba8416
fix test
e8f58861
drop
1678792c
drop useless imports
9498556c
update docs
039b3772
mayank31398
marked this pull request as ready for review 2 years ago
mayank31398
changed the title Granite language models [WIP] Granite language models 2 years ago
Merge branch 'main' into granite
1bea763e
Merge branch 'main' into granite
2a9d7346
drop flash function
24424920
copied from
2efe0a6c
drop pretraining tp
8da50b59
drop pretraining tp
73d4f2d3
drop pretraining tp
5f020753
drop unused import
de33d606
drop code path
42035c44
change name
f833ca6c
softmax scale
5ca5b086
head dim
abb359d6
drop legacy cache
cfa8210c
rename params
b1dad991
cleanup
79bdf6b5
Merge branch 'main' into granite
91a02533
fix copies
7df943fe
comments
18d577d5
add back legacy cache
90c79064
multipliers
a765b89f
multipliers
ce070ade
multipliers
8b1b7e07
text fix
fcc7bf7f
Merge branch 'main' into granite
bd6bee64
fix copies
37eb40fa
Merge branch 'main' into granite
d743ff7d
merge
1142dbb7
multipliers
c3185def
attention multiplier
6ccf5b5c
drop unused imports
52440ad9
Merge branch 'main' into granite
b39cb7d5
Merge branch 'main' into granite
6fa1774c
Merge branch 'main' into granite
46524c73
fix
71c2cde2
fix
fe648413
fix
559204db
move rope?
b64c16d0
Update src/transformers/models/granite/configuration_granite.py
cd9a911c
fix
124065c7
Update src/transformers/models/granite/modeling_granite.py
02c1073e
fix
8c9112f3
fix
9493a86d
fix
4932097e
fix
b7fe0d3d
Merge branch 'main' into granite
fa5550aa
fix-copies
1eeab96a
Merge branch 'main' into granite
6a132857
torch rmsnorm
7d68382f
add authors
4a4a581e
Merge branch 'main' into granite
57a5b9e8
change model path
c03e3957
fix
ffed3d1c
test
a7fbd303
drop static cache test
8f00c1b3
uupdate readme
586fdf18
drop non-causal
dc9faaae
readme
545449cc
drop useless imports
5e5cad9a
Update docs/source/en/model_doc/granite.md
24029b28
Update docs/source/en/model_doc/granite.md
eaeff2a9
Update docs/source/en/model_doc/granite.md
ee9c0f65
Assignees
No one assigned
Labels
New model
run-slow
Login to write a write a comment.
Login via GitHub