sometimesanotion/lamarck-14b-v0.1-reason-model_stock

model

This is a dependency for future merges for Lamarck v0.3. @CultriX's models which result from evolutionary merging earn their place alongside Sauerkraut and Virtuoso. Lamarck's merge process uses these to keep later refinements to the model simple.

This is a merge of pre-trained language models created using mergekit.

Merge Details

Merge Method

This model was merged using the Model Stock merge method using Qwen/Qwen2.5-14B as a base.

Models Merged

The following models were included in the merge:

• arcee-ai/Virtuoso-Small
• VAGOsolutions/SauerkrautLM-v2-14b-DPO
• CultriX/Qwen2.5-14B-Wernicke
• CultriX/SeQwence-14B-EvolMerge

Configuration

The following YAML configuration was used to produce this model:

name:                lamarck-14b-reason-model_stock
merge_method:        model_stock
base_model:          Qwen/Qwen2.5-14B
tokenizer_source:    Qwen/Qwen2.5-14B-Instruct
parameters:
  int8_mask:         false
  normalize:         true
  rescale:           false
models:
  - model:           VAGOsolutions/SauerkrautLM-v2-14b-DPO
  - model:           CultriX/Qwen2.5-14B-Wernicke
  - model:           arcee-ai/Virtuoso-Small
  - model:           CultriX/SeQwence-14B-EvolMerge

dtype:               bfloat16
out_dtype:           bfloat16

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric	Value
Avg.	36.26
IFEval (0-Shot)	49.65
BBH (3-Shot)	50.72
MATH Lvl 5 (4-Shot)	31.57
GPQA (0-shot)	17.90
MuSR (0-shot)	18.79
MMLU-PRO (5-shot)	48.91