---
base_model:
- MihaiPopa-1/LFM2.5-350M-heretic
- mkurman/LiquidAI-LFM2.5-350M-SYNTH
- squ11z1/claude-oss-350m
library_name: transformers
tags:
- mergekit
- merge

---
# merge

This is a merge of pre-trained language models created using [mergekit](https://github.com/cg123/mergekit).

## Merge Details
### Merge Method

This model was merged using the [DARE TIES](https://arxiv.org/abs/2311.03099) merge method using [MihaiPopa-1/LFM2.5-350M-heretic](https://huggingface.co/MihaiPopa-1/LFM2.5-350M-heretic) as a base.

### Models Merged

The following models were included in the merge:
* [mkurman/LiquidAI-LFM2.5-350M-SYNTH](https://huggingface.co/mkurman/LiquidAI-LFM2.5-350M-SYNTH)
* [squ11z1/claude-oss-350m](https://huggingface.co/squ11z1/claude-oss-350m)

### Configuration

The following YAML configuration was used to produce this model:

```yaml
merge_method: dare_ties
base_model: MihaiPopa-1/LFM2.5-350M-heretic
dtype: bfloat16

parameters:
  normalize: true
  int8_mask: true

models:
  # Assistant tone / conversational instruction-following
  - model: squ11z1/claude-oss-350m
    parameters:
      weight: 0.50
      density: 0.50

  # Generalization / synthetic behavior
  - model: mkurman/LiquidAI-LFM2.5-350M-SYNTH
    parameters:
      weight: 0.50
      density: 0.50
```