Skip to content

Add Mistral Small 4 (119B MoE) support via mistral4.py #1037

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
ProducerGuy wants to merge 1 commit into
base: main
Choose a base branch
from
+438 −2
Open

Add Mistral Small 4 (119B MoE) support via mistral4.py #1037

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
11 changes: 9 additions & 2 deletions mlx_lm/models/mistral3.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -7,7 +7,7 @@
import mlx.nn as nn
from mlx.utils import tree_flatten, tree_unflatten

from . import llama, ministral3
from . import llama, ministral3, mistral4
from .base import BaseModelArgs


Expand All @@ -26,7 +26,14 @@ def __init__(self, args: ModelArgs):
super().__init__()
self.args = args
self.model_type = args.model_type
if args.text_config.get("model_type") == "ministral3":
# Route based on architecture structure, not model_type string.
# MoE models (Mistral Small 4) have n_routed_experts in config.
# Dense models (Ministral 3B/8B/14B) use ministral3 or llama paths.
if args.text_config.get("n_routed_experts") is not None:
self.language_model = mistral4.Model(
mistral4.ModelArgs.from_dict(args.text_config)
)
elif args.text_config.get("model_type") == "ministral3":
self.language_model = ministral3.Model(
ministral3.ModelArgs.from_dict(args.text_config)
)
Expand Down
Loading