Mini-Tcomanr-Experimental

Mini-Tcomanr-R1

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

Merge Details

Merge Method

This model was merged using the TIES merge method using deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B as a base.

Models Merged

The following models were included in the merge:

• agentica-org/DeepScaleR-1.5B-Preview
• agentica-org/DeepCoder-1.5B-Preview
• Josephgflowers/DeepSeek-R1-Distill-Qwen-1.5B-LIMO
• huihui-ai/DeepSeek-R1-Distill-Qwen-1.5B-abliterated
• ibndias/DeepSeek-R1-Distill-Qwen-1.5B-GRPO
• ertghiu256/Deepseek-R1-Distill-1.5B-code-instruct
• prithivMLmods/QwQ-R1-Distill-1.5B-CoT

Configuration

The following YAML configuration was used to produce this model:

models:
  - model: prithivMLmods/QwQ-R1-Distill-1.5B-CoT
    parameters:
      weight: 0.5
  - model: ertghiu256/Deepseek-R1-Distill-1.5B-code-instruct
    parameters:
      weight: 1.0
  - model: huihui-ai/DeepSeek-R1-Distill-Qwen-1.5B-abliterated
    parameters:
      weight: 0.85
  - model: Josephgflowers/DeepSeek-R1-Distill-Qwen-1.5B-LIMO
    parameters:
      weight: 0.5
  - model: ibndias/DeepSeek-R1-Distill-Qwen-1.5B-GRPO
    parameters:
      weight: 0.3
  - model: agentica-org/DeepScaleR-1.5B-Preview
    parameters:
      weight: 0.95
  - model: agentica-org/DeepCoder-1.5B-Preview
    parameters:
      weight: 1.0
  - model: deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B
    parameters:
      weight: 1.0
merge_method: ties
base_model: deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B
parameters:
  normalize: true
  int8_mask: true
  lambda: 1.0
  rescale: true
dtype: float16