MixtureKit: A General Framework for Composing, Training, and Visualizing Mixture-of-Experts Models

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Chamma, Ahmad, Herraoui, Omar El, Shang, Guokan
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908708548640768
author Chamma, Ahmad
Herraoui, Omar El
Shang, Guokan
author_facet Chamma, Ahmad
Herraoui, Omar El
Shang, Guokan
contents We introduce MixtureKit, a modular open-source framework for constructing, training, and analyzing Mixture-of-Experts (MoE) models from arbitrary pre-trained or fine-tuned models. MixtureKit currently supports three complementary methods: (i) \emph{Traditional MoE}, which uses a single router per transformer block to select experts, (ii) \emph{BTX} (Branch-Train-Mix), which introduces separate routers for each specified sub-layer enabling fine-grained token routing, and (iii) \emph{BTS} (Branch-Train-Stitch), which keeps experts fully intact and introduces trainable stitch layers for controlled information exchange between hub and experts. MixtureKit automatically modifies the model configuration, patches decoder and causal LM classes, and saves a unified checkpoint ready for inference or fine-tuning. We further provide a visualization interface to inspect per-token routing decisions, expert weight distributions, and layer-wise contributions. Experiments with multilingual code-switched data (e.g. Arabic-Latin) show that a BTX-based model trained using MixtureKit can outperform baseline dense models on multiple benchmarks. We release MixtureKit as a practical foundation for research and development of MoE-based systems across diverse domains.
format Preprint
id arxiv_https___arxiv_org_abs_2512_12121
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle MixtureKit: A General Framework for Composing, Training, and Visualizing Mixture-of-Experts Models
Chamma, Ahmad
Herraoui, Omar El
Shang, Guokan
Machine Learning
Artificial Intelligence
We introduce MixtureKit, a modular open-source framework for constructing, training, and analyzing Mixture-of-Experts (MoE) models from arbitrary pre-trained or fine-tuned models. MixtureKit currently supports three complementary methods: (i) \emph{Traditional MoE}, which uses a single router per transformer block to select experts, (ii) \emph{BTX} (Branch-Train-Mix), which introduces separate routers for each specified sub-layer enabling fine-grained token routing, and (iii) \emph{BTS} (Branch-Train-Stitch), which keeps experts fully intact and introduces trainable stitch layers for controlled information exchange between hub and experts. MixtureKit automatically modifies the model configuration, patches decoder and causal LM classes, and saves a unified checkpoint ready for inference or fine-tuning. We further provide a visualization interface to inspect per-token routing decisions, expert weight distributions, and layer-wise contributions. Experiments with multilingual code-switched data (e.g. Arabic-Latin) show that a BTX-based model trained using MixtureKit can outperform baseline dense models on multiple benchmarks. We release MixtureKit as a practical foundation for research and development of MoE-based systems across diverse domains.
title MixtureKit: A General Framework for Composing, Training, and Visualizing Mixture-of-Experts Models
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2512.12121