metadata
language:
- en
license: apache-2.0
library_name: transformers
tags:
- mergekit
- merge
- not-for-all-audiences
base_model:
- bamec66557/mergekit-slerp-uhnpbqg
- bamec66557/mergekit-slerp-tsgkafq
model-index:
- name: MISCHIEVOUS-12B-Mix_III_ex_V
results:
- task:
type: text-generation
name: Text Generation
dataset:
name: IFEval (0-Shot)
type: HuggingFaceH4/ifeval
args:
num_few_shot: 0
metrics:
- type: inst_level_strict_acc and prompt_level_strict_acc
value: 43.16
name: strict accuracy
source:
url: >-
https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=bamec66557/MISCHIEVOUS-12B-Mix_III_ex_V
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: BBH (3-Shot)
type: BBH
args:
num_few_shot: 3
metrics:
- type: acc_norm
value: 34.87
name: normalized accuracy
source:
url: >-
https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=bamec66557/MISCHIEVOUS-12B-Mix_III_ex_V
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: MATH Lvl 5 (4-Shot)
type: hendrycks/competition_math
args:
num_few_shot: 4
metrics:
- type: exact_match
value: 13.14
name: exact match
source:
url: >-
https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=bamec66557/MISCHIEVOUS-12B-Mix_III_ex_V
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: GPQA (0-shot)
type: Idavidrein/gpqa
args:
num_few_shot: 0
metrics:
- type: acc_norm
value: 9.4
name: acc_norm
source:
url: >-
https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=bamec66557/MISCHIEVOUS-12B-Mix_III_ex_V
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: MuSR (0-shot)
type: TAUR-Lab/MuSR
args:
num_few_shot: 0
metrics:
- type: acc_norm
value: 12.77
name: acc_norm
source:
url: >-
https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=bamec66557/MISCHIEVOUS-12B-Mix_III_ex_V
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: MMLU-PRO (5-shot)
type: TIGER-Lab/MMLU-Pro
config: main
split: test
args:
num_few_shot: 5
metrics:
- type: acc
value: 29.43
name: accuracy
source:
url: >-
https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=bamec66557/MISCHIEVOUS-12B-Mix_III_ex_V
name: Open LLM Leaderboard
merge
This is a merge of pre-trained language models created using mergekit.
Merge Details
Merge Method
This model was merged using the SLERP merge method.
Models Merged
The following models were included in the merge:
Configuration
The following YAML configuration was used to produce this model:
# Merging MISCHIEVOUS-12B-Mix Models with Sliced SLERP
base_model: bamec66557/mergekit-slerp-tsgkafq
dtype: bfloat16
merge_method: slerp
tokenizer_source: union
# Slices Configuration (Layer-Specific Merging)
slices:
- name: initial_layers
sources:
- model: bamec66557/mergekit-slerp-uhnpbqg
layer_range: [0, 10]
- model: bamec66557/mergekit-slerp-tsgkafq
layer_range: [0, 10]
parameters:
t:
- name: self_attn
value: [0.8, 0.85, 0.9, 0.95, 1.0]
- name: mlp
value: [0.9, 0.95, 1.0, 1.05, 1.1]
- name: layer_norm
value: [0.6, 0.65, 0.7, 0.75, 0.8]
- name: embed_tokens
value: [1.0]
- name: middle_layers
sources:
- model: bamec66557/mergekit-slerp-uhnpbqg
layer_range: [10, 20]
- model: bamec66557/mergekit-slerp-tsgkafq
layer_range: [10, 20]
parameters:
t:
- name: self_attn
value: [0.7, 0.75, 0.8, 0.85, 0.9]
- name: mlp
value: [1.0, 0.95, 0.9, 0.85, 0.8]
- name: layer_norm
value: [0.5, 0.55, 0.6, 0.65, 0.7]
- name: embed_tokens
value: [1.0]
- name: upper_middle_layers
sources:
- model: bamec66557/mergekit-slerp-uhnpbqg
layer_range: [20, 30]
- model: bamec66557/mergekit-slerp-tsgkafq
layer_range: [20, 30]
parameters:
t:
- name: self_attn
value: [0.6, 0.65, 0.7, 0.75, 0.8]
- name: mlp
value: [0.8, 0.75, 0.7, 0.65, 0.6]
- name: layer_norm
value: [0.4, 0.45, 0.5, 0.55, 0.6]
- name: embed_tokens
value: [1.0]
- name: final_layers
sources:
- model: bamec66557/mergekit-slerp-uhnpbqg
layer_range: [30, 40]
- model: bamec66557/mergekit-slerp-tsgkafq
layer_range: [30, 40]
parameters:
t:
- name: self_attn
value: [0.9, 1.0, 1.1, 1.2, 1.3]
- name: mlp
value: [0.7, 0.65, 0.6, 0.55, 0.5]
- name: layer_norm
value: [0.7, 0.75, 0.8, 0.85, 0.9]
- name: embed_tokens
value: [1.0]
# Regularization (Prevent Overfitting During Merging)
regularization:
methods:
- name: weight_clipping
clip_range: [-0.2, 0.2]
- name: random_noise
scale: 0.015
- name: l2_norm
scale: 0.01
# Postprocessing (Enhance Merged Model Quality)
postprocessing:
operations:
- name: random_noise
scale: 0.0025
- name: non_linear_scaling
parameters:
function: tanh
- name: sharpening
intensity: 0.3
- name: gaussian_smoothing
sigma: 1.5
- name: smoothing
parameters:
adaptive: true
range: [0.8, 1.2]
kernel_size: 5
- name: normalize
- name: dynamic_scaling
scale_range: [0.75, 1.25]
# Optional: Ties Merging (Advanced Technique)
ties:
enabled: false # Set to true if ties merging is required
method: greedy # Options: greedy, optimal, random
layers: [0, 10, 20, 30] # Example layers for ties merging
Open LLM Leaderboard Evaluation Results
Detailed results can be found here
| Metric | Value |
|---|---|
| Avg. | 23.80 |
| IFEval (0-Shot) | 43.16 |
| BBH (3-Shot) | 34.87 |
| MATH Lvl 5 (4-Shot) | 13.14 |
| GPQA (0-shot) | 9.40 |
| MuSR (0-shot) | 12.77 |
| MMLU-PRO (5-shot) | 29.43 |