Home

Awesome

mergekit

mergekit is a toolkit for merging pre-trained language models. mergekit uses an out-of-core approach to perform unreasonably elaborate merges in resource-constrained situations. Merges can be run entirely on CPU or accelerated with as little as 8 GB of VRAM. Many merging algorithms are supported, with more coming as they catch my attention.

Contents

Why Merge Models?

Model merging is a powerful technique that allows combining the strengths of different models without the computational overhead of ensembling or the need for additional training. By operating directly in the weight space of models, merging can:

Unlike traditional ensembling which requires running multiple models, merged models maintain the same inference cost as a single model while often achieving comparable or superior performance.

Features

Key features of mergekit include:

🌐 GUI Launch Alert 🤗 - We are excited to announce the launch of a mega-GPU backed graphical user interface for mergekit in Arcee! This GUI simplifies the merging process, making it more accessible to a broader audience. Check it out and contribute at the Arcee App. There is also a Hugging Face Space with limited amounts of GPUs.

Installation

git clone https://github.com/arcee-ai/mergekit.git
cd mergekit

pip install -e .  # install the package and make scripts available

If the above fails with the error of:

ERROR: File "setup.py" or "setup.cfg" not found. Directory cannot be installed in editable mode:
(A "pyproject.toml" file was found, but editable mode currently requires a setuptools-based build.)

You may need to upgrade pip to > 21.3 with the command python3 -m pip install --upgrade pip

Usage

The script mergekit-yaml is the main entry point for mergekit. It takes a YAML configuration file and an output path, like so:

mergekit-yaml path/to/your/config.yml ./output-model-directory [--cuda] [--lazy-unpickle] [--allow-crimes] [... other options]

This will run the merge and write your merged model to ./output-model-directory.

For more information on the arguments accepted by mergekit-yaml run the command mergekit-yaml --help.

Uploading to Huggingface

When you have a merged model you're happy with, you may want to share it on the Hugging Face Hub. mergekit generates a README.md for your merge with some basic information for a model card. You can edit it to include more details about your merge, like giving it a good name or explaining what it's good at; rewrite it entirely; or use the generated README.md as-is. It is also possible to edit your README.md online once it has been uploaded to the Hub.

Once you're happy with your model card and merged model, you can upload it to the Hugging Face Hub using the huggingface_hub Python library.

# log in to huggingface with an access token (must have write permission)
huggingface-cli login
# upload your model
huggingface-cli upload your_hf_username/my-cool-model ./output-model-directory .

The documentation for huggingface_hub goes into more detail about other options for uploading.

Merge Configuration

Merge configurations are YAML documents specifying the operations to perform in order to produce your merged model. Below are the primary elements of a configuration file:

Parameter Specification

Parameters are flexible and can be set with varying precedence. They can be specified conditionally using tensor name filters, which allows finer control such as differentiating between attention heads and fully connected layers.

Parameters can be specified as:

The parameters can be set at different levels, with decreasing precedence as follows:

  1. slices.*.sources.parameters - applying to a specific input slice
  2. slices.*.parameters - applying to a specific output slice
  3. models.*.parameters or input_model_parameters - applying to any tensors coming from specific input models
  4. parameters - catchall

Tokenizer Configuration

The tokenizer behavior can be configured in two ways: using the new tokenizer field (recommended) or the legacy tokenizer_source field (maintained for backward compatibility). These fields are mutually exclusive - you should use one or the other, not both.

Modern Configuration (tokenizer)

The tokenizer field provides fine-grained control over vocabulary and embeddings:

tokenizer:
  source: "union"  # or "base" or a specific model path
  tokens:          # Optional: configure specific tokens
    <token_name>:
      source: ...  # Specify embedding source
      force: false # Optional: force this embedding for all models
  pad_to_multiple_of: null  # Optional: pad vocabulary size
Tokenizer Source

The source field determines the vocabulary of the output model:

Token Embedding Handling

When merging models with different vocabularies, mergekit uses smart defaults to handle token embeddings:

You can override these defaults for specific tokens:

tokenizer:
  source: union
  tokens:
    # Use embedding from a specific model
    <|im_start|>:
      source: "path/to/chatml/model"

    # Force a specific embedding for all models
    <|special|>:
      source: "path/to/model"
      force: true

    # Map a token to another model's token embedding
    <|renamed_token|>:
      source:
        kind: "model_token"
        model: "path/to/model"
        token: "<|original_token|>"  # or use token_id: 1234
Practical Example

Here's how you might preserve both Llama 3 Instruct and ChatML prompt formats when merging models:

tokenizer:
  source: union
  tokens:
    # ChatML tokens
    <|im_start|>:
      source: "chatml_model"
    <|im_end|>:
      source: "chatml_model"

    # Llama 3 tokens - force original embeddings
    <|start_header_id|>:
      source: "llama3_model"
      force: true
    <|end_header_id|>:
      source: "llama3_model"
      force: true
    <|eot_id|>:
      source: "llama3_model"
      force: true

Legacy Configuration (tokenizer_source)

For backward compatibility, the tokenizer_source field is still supported:

tokenizer_source: "union"  # or "base" or a model path

This provides basic tokenizer selection but lacks the fine-grained control of the modern tokenizer field.

Chat Template Configuration

The optional chat_template field allows overriding the chat template used for the merged model.

chat_template: "auto"  # or a template name or Jinja2 template

Options include:

Examples

Several examples of merge configurations are available in examples/.

Merge Methods

A quick overview of the currently supported merge methods:

Methodmerge_method valueMulti-ModelUses base model
Linear (Model Soups)linear✅❌
SLERPslerp❌✅
Task Arithmetictask_arithmetic✅✅
TIESties✅✅
DARE TIESdare_ties✅✅
DARE Task Arithmeticdare_linear✅✅
Passthroughpassthrough❌❌
Model Breadcrumbsbreadcrumbs✅✅
Model Breadcrumbs + TIESbreadcrumbs_ties✅✅
Model Stockmodel_stock✅✅
NuSLERPnuslerp❌✅
DELLAdella✅✅
DELLA Task Arithmeticdella_linear✅✅

Linear

The classic merge method - a simple weighted average.

Parameters:

SLERP

Spherically interpolate the parameters of two models. One must be set as base_model.

Parameters:

Task Arithmetic

Computes "task vectors" for each model by subtracting a base model. Merges the task vectors linearly and adds back the base. Works great for models that were fine tuned from a common ancestor. Also a super useful mental framework for several of the more involved merge methods.

Parameters: same as Linear

TIES

Builds on the task arithmetic framework. Resolves interference between models by sparsifying the task vectors and applying a sign consensus algorithm. Allows you to merge a larger number of models and retain more of their strengths.

Parameters: same as Linear, plus:

DARE

In the same vein as TIES, sparsifies task vectors to reduce interference. Differs in that DARE uses random pruning with a novel rescaling to better match performance of the original models. DARE can be used either with the sign consensus algorithm of TIES (dare_ties) or without (dare_linear).

Parameters: same as TIES for dare_ties, or Linear for dare_linear

Passthrough

passthrough is a no-op that simply passes input tensors through unmodified. It is meant to be used for layer-stacking type merges where you have only one input model. Useful for frankenmerging.

Model Breadcrumbs

An extension of task arithmetic that discards both small and extremely large differences from the base model. As with DARE, the Model Breadcrumbs algorithm can be used with (breadcrumbs_ties) or without (breadcrumbs) the sign consensus algorithm of TIES.

Parameters: same as Linear, plus:

Note that gamma corresponds with the parameter β described in the paper, while density is the final density of the sparsified tensors (related to γ and β by density = 1 - γ - β). For good default values, try density: 0.9 and gamma: 0.01.

Model Stock

Uses some neat geometric properties of fine tuned models to compute good weights for linear interpolation. Requires at least three models, including a base model.

Parameters:

NuSLERP

Spherically interpolate between parameters, but with more options and more sensical configuration! Does not require a base model, but can use one to do spherical interpolation of task vectors. Only works with either two models or two plus a base model.

Parameters:

To replicate the behavior of the original slerp method, set weight to 1-t and t for your first and second model respectively.

DELLA

Building upon DARE, DELLA uses adaptive pruning based on parameter magnitudes. DELLA first ranks parameters in each row of delta parameters and assigns drop probabilities inversely proportional to their magnitudes. This allows it to retain more important changes while reducing interference. After pruning, it rescales the remaining parameters similar to DARE. DELLA can be used with (della) or without (della_linear) the sign elect step of TIES

Parameters: same as Linear, plus:

LoRA extraction

Mergekit allows extracting PEFT-compatible low-rank approximations of finetuned models.

Usage

mergekit-extract-lora finetuned_model_id_or_path base_model_id_or_path output_path [--no-lazy-unpickle] --rank=desired_rank

Mixture of Experts merging

The mergekit-moe script supports merging multiple dense models into a mixture of experts, either for direct use or for further training. For more details see the mergekit-moe documentation.

Evolutionary merge methods

See docs/evolve.md for details.

✨ Merge in the Cloud ✨

We host merging on Arcee's cloud GPUs - you can launch a cloud merge in the Arcee App. Or through python - grab an ARCEE_API_KEY:

export ARCEE_API_KEY=<your-api-key> pip install -q arcee-py

import arcee
arcee.merge_yaml("bio-merge","./examples/bio-merge.yml")

Check your merge status at the Arcee App

When complete, either deploy your merge:

arcee.start_deployment("bio-merge", merging="bio-merge")

Or download your merge:

!arcee merging download bio-merge

Citation

If you find mergekit useful in your research, please consider citing the paper:

@inproceedings{goddard-etal-2024-arcees,
    title = "Arcee{'}s {M}erge{K}it: A Toolkit for Merging Large Language Models",
    author = "Goddard, Charles  and
      Siriwardhana, Shamane  and
      Ehghaghi, Malikeh  and
      Meyers, Luke  and
      Karpukhin, Vladimir  and
      Benedict, Brian  and
      McQuade, Mark  and
      Solawetz, Jacob",
    editor = "Dernoncourt, Franck  and
      Preo{\c{t}}iuc-Pietro, Daniel  and
      Shimorina, Anastasia",
    booktitle = "Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing: Industry Track",
    month = nov,
    year = "2024",
    address = "Miami, Florida, US",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2024.emnlp-industry.36",
    doi = "10.18653/v1/2024.emnlp-industry.36",
    pages = "477--485",
    abstract = "The rapid growth of open-source language models provides the opportunity to merge model checkpoints, combining their parameters to improve performance and versatility. Advances in transfer learning have led to numerous task-specific models, which model merging can integrate into powerful multitask models without additional training. MergeKit is an open-source library designed to support this process with an efficient and extensible framework suitable for any hardware. It has facilitated the merging of thousands of models, contributing to some of the world{'}s most powerful open-source model checkpoints. The library is accessible at: https://github.com/arcee-ai/mergekit.",
}