Tools for merging pre-trained large language models
Project description
mergekit
mergekit
is a toolkit for merging pre-trained language models, using a variety of merge methods including TIES, linear, and slerp merging. The toolkit also enables piecewise assembly of a language model from layers.
Run pip install -e .
to install the package and make the scripts available.
The script mergekit-yaml
takes a YAML configuration file defining the operations to perform.
Configuration
Below are the primary elements of a configuration file:
merge_method
: Specifies the method to use for merging models. Can be one of 'ties', 'linear', 'slerp', or 'passthrough'.slices
: Defines slices of layers from different models to be used. This field is mutually exclusive withmodels
.models
: Defines entire models to be used for merging. This field is mutually exclusive withslices
.base_model
: Specifies the base model used in some merging methods.parameters
: Holds various parameters such as weights and densities, which can also be specified at different levels of the configuration.dtype
: Specifies the data type for the merging operation.
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:
- Scalars: Single floating-point values.
- Gradients: List of floating-point values, specifying an interpolated gradient.
The parameters can be set at different levels, with decreasing precedence as follows:
slices.*.sources.parameters
- applying to a specific input sliceslices.*.parameters
- applying to a specific output sliceinput_model_parameters
- applying to any tensors coming from specific input modelsparameters
- catchall
Merge Methods
Resolving Interference When Merging Models ("ties"
)
Requires a base model. Parameters:
density
- fraction of weights in differences from the base model to retainweight
- relative (or absolute ifnormalize=False
) weighting of a given tensornormalize
- if true, the weights of all models contributing to a tensor will be normalized. Default behavior.
Linear
Does not require a base model. Takes parameters weight
and normalize
, with same definition as above.
SLERP
Requires exactly two models, one of which must be the base model. Takes one parameter - t
- the interpolation factor from the base model to the secondary model.
Examples
-
Simple linear merge of multiple models:
models: - model: psmathur/orca_mini_v3_13b parameters: weight: 1.0 - model: WizardLM/WizardLM-13B-V1.2 parameters: weight: 0.3 - model: garage-bAInd/Platypus2-13B parameters: weight: 0.5 merge_method: linear dtype: float16
-
bakllama.py
style layer recombination:slices: - sources: - model: psmathur/orca_mini_v3_13b layer_range: [0, 24] - sources: - model: garage-bAInd/Platypus2-13B layer_range: [20, 40] merge_method: passthrough dtype: float16
-
Gradient SLERP with different weights for mlp/self attention:
slices: - sources: - model: psmathur/orca_mini_v3_13b layer_range: [0, 40] - model: garage-bAInd/Platypus2-13B layer_range: [0, 40] merge_method: slerp base_model: psmathur/orca_mini_v3_13b parameters: t: - filter: self_attn value: [0, 0.5, 0.3, 0.7, 1] - filter: mlp value: [1, 0.5, 0.7, 0.3, 0] - value: 0.5 # fallback for rest of tensors dtype: float16
Usage
Once you have created the YAML configuration file, run mergekit-yaml
with the config file and output path as arguments:
mergekit-yaml path/to/your/config.yml ./output-model-directory [--cuda]
Legacy Wrappers
Mergekit originally featured two separate scripts with different inputs. The functionality of these is maintained in the mergekit-legacy
and bakllama
wrappers. Example usage:
mergekit-legacy ./output-model --base-model TheBloke/Llama-2-13B-fp16 --cuda \
--merge WizardLM/WizardLM-13B-V1.2 --weight 0.3 --density 0.5 \
--merge garage-bAInd/Platypus2-13B --weight 0.5 --density 0.5
mergekit-legacy
can output a YAML configuration for easy migration with the --print-yaml
option.
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Hashes for mergekit-0.0.2.1-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | de3ddb4b1fe98a0706ea8e0fc679addb51786a4c0caa1a8b57864060002bf1cc |
|
MD5 | 3499fe46e4175c99a58ef8aeefc570d9 |
|
BLAKE2b-256 | 6c139075f33964827e0b611f6d4c7c374d5fcf669c7dff1c66c33c2eaae429e1 |