Downtown-Case/Deepseek-EVA-32B-SCE-v1

Deepseek-EVA-SCE-Merge

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

The intent is to make Deepseek R1 32B feel more like a "base model" like EVA-Gutenberg Qwen, adhering to an existing style when writing.

But consider this a quick A/B test between the new DELLA and SCE merging techniques. This is the SCE merge, with a top-k of 0.3 instead of the default 0.1.

Merge Details

Merge Method

This model was merged using the SCE merge method using /home/a/Models/Raw/Qwen_Qwen2.5-32B as a base.

Models Merged

The following models were included in the merge:

• /home/a/Models/Raw/nbeerbower_EVA-Gutenberg3-Qwen2.5-32B
• /home/a/Models/Raw/deepseek-ai_DeepSeek-R1-Distill-Qwen-32B

Configuration

The following YAML configuration was used to produce this model:

models:
  - model: /home/a/Models/Raw/Qwen_Qwen2.5-32B
  # No parameters necessary for base model
  - model: /home/a/Models/Raw/nbeerbower_EVA-Gutenberg3-Qwen2.5-32B
  - model: /home/a/Models/Raw/deepseek-ai_DeepSeek-R1-Distill-Qwen-32B
merge_method: sce
tokenizer:
  source: "union"
base_model: /home/a/Models/Raw/Qwen_Qwen2.5-32B
chat_template: "deepseek"
parameters:
  int8_mask: true
  select_topk: 0.30
dtype: bfloat16