[new-models] LFM2-MoE
d3e71a9b
[docs] add in template lfm2_moe doc files
ee096db6
[configuration] update configuration class
1ef7c59b
[modular][lfm] minor: fix rotary_emb typo
d42252a4
[modeling] modular/modeling files for Lfm2Moe
79b3f20d
[modeling][lfm2_moe] fix Lfm2Moe modular/modeling
5bc8e0d1
[configuration][lfm2_moe] update configuration keys with latest confi…
bfd559a1
[misc] make fixup
e410997f
[modular][lfm2_moe] address comments: dtype, mlp, buffers
7888413a
[configuration][lfm2_moe] add initializer_range
d048b127
[modular][lfm2_moe] include init_weights to pass test_initialization
c0486222
[tests][causal_lm] include pos_emb as possible rope attribute
d10a8c99
[modeling][lfm2_moe] remove load_balancing_loss_func due to lack of s…
7114e62c
[misc] make style
1b74ceea
[modeling][lfm2_moe] MoE refactor PR update in LFM2Moe
fb65ff5b
[tests] lfm2_moe: unit tests
f4785d93
[misc] update LFM2-8B-A1B repo id
dd63210c
[tests] lfm2: update ModelTests for lfm2
fc7d2eaf
Update LFM2 documentation
daa35aa0
Add Lfm2Moe documentation
96fbd87e
[misc] fix ci
c7bac38e
[docs] remove trust_remote_code
37d2ffff
[misc] ci: fix modular
23c2c991
Merge branch 'main' into lfm2_moe
1427f48b
reapply modular
8b6aa838
simplify
cd873d4b
remove static address and inplace op
0c67e26d
simplify
44ff5d25
simplify a bit more the modular
38130f5e
imports
b3a0924e
Assignees
No one assigned
Login to write a write a comment.
Login via GitHub