Snowdrop experiments
Collection
sidegrades compared to v0
•
3 items
•
Updated
This is a merge of pre-trained language models created using mergekit.
This model was merged using the SCE merge method using Qwen/Qwen2.5-32B as a base.
The following models were included in the merge:
The following YAML configuration was used to produce this model:
base_model: Qwen/Qwen2.5-32B
merge_method: sce
parameters:
normalize: true
int8_mask: true
tokenizer_source: Qwen/Qwen2.5-32B-Instruct
dtype: bfloat16
models:
- model: trashpanda-org/Qwen2.5-32B-Marigold-v0-exp
parameters:
select_topk: 0.15
- model: trashpanda-org/Qwen2.5-32B-Marigold-v0
parameters:
select_topk: 0.15
- model: Qwen/QwQ-32B
parameters:
select_topk: 0.12
- model: trashpanda-org/QwQ-32B-Snowdrop-v0
parameters:
select_topk: 0.2
- model: trashpanda-org/Qwen2.5-32B-Marigold-v1
parameters:
select_topk: 0.1
- model: trashpanda-org/Qwen2.5-32B-Marigold-v1-ep3
parameters:
select_topk: 0.1
- model: huihui-ai/QwQ-32B-abliterated
parameters:
select_topk: 0.07