EVA-Tissint-v1.2-14B

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

New merge of EVA v0.2, now using Tissint v1.2. I slightly altered the merge parameters yet again, I wonder how this will go for other people.

I recommend the samplers provided on Tissint's model card.

Merge Method

This model was merged using the della_linear merge method using EVA-UNIT-01/EVA-Qwen2.5-14B-v0.2 as a base.

Models Merged

The following models were included in the merge:

  • Ttimofeyka/Tissint-14B-v1.2-128k-RP

Configuration

The following YAML configuration was used to produce this model:

models:
    - model: Ttimofeyka/Tissint-14B-v1.2-128k-RP
      parameters:
        density: 0.45
        weight: 0.35
    - model: EVA-UNIT-01/EVA-Qwen2.5-14B-v0.2
      parameters:
        density: 0.55
        weight: 0.65

merge_method: della_linear
base_model: EVA-UNIT-01/EVA-Qwen2.5-14B-v0.2
parameters:
    epsilon: 0.05
    lambda: 1
dtype: bfloat16
Downloads last month
7
Safetensors
Model size
14.8B params
Tensor type
BF16
·
Inference Providers NEW
This model is not currently available via any of the supported Inference Providers.

Model tree for ockerman0/EVA-Tissint-v1.2-14B

Quantizations
2 models