feat: Add Mimo v2.5 model support #22493
add mimo-v2.5 support
09e1b618
mimo-v2.5: fix modify_tensors row split
548fde30
mimi-v2.5: forgot `add_attn_value_scale` plumbing
287ac836
mimi-v2.5: fix tp dequant to detect tp rows
3dcaba98
ngxson
commented
on 2026-04-29
mimo-v2.5: fix TP iteration to be descending
24364b36
mimo-v2.5: fix comment
027d5756
ngxson
approved these changes
on 2026-05-04
Merge remote-tracking branch 'origin/master' into mimo-v2.5
37c58667
mimo-v2.5: retain fused qkv
1cf092ca
AesSedai
marked this pull request as draft 46 days ago
mimo-v2.5: missed the attn_value scale during merge
a57c7072
AesSedai
marked this pull request as ready for review 46 days ago
mimo-v2.5: fused QKV needs contiguous for scaling attention value
c6a0bc8d
CISC
commented
on 2026-05-06
Merge remote-tracking branch 'origin/master' into mimo-v2.5
d2b710cb
mimo-v2.5: move `speech_embeddings.` to TextModel filter_tensors
451cf3c8
CISC
approved these changes
on 2026-05-07
Update src/llama-hparams.h
2718bea9
Update src/models/mimo2.cpp
12e71fb8
Update src/models/mimo2.cpp
0e703ad8
CISC
commented
on 2026-05-07
Update convert_hf_to_gguf.py
bbd23729
Update convert_hf_to_gguf.py
c305687b
Update src/models/mimo2.cpp
fe9b7059
mimo-v2.5: include MTP weights in gguf
b7de5076
ngxson
approved these changes
on 2026-05-07
ngxson
merged
8e52631d
into master 44 days ago
Assignees
No one assigned
Login to write a write a comment.
Login via GitHub