Files
Ouroboros/README.md

61 lines
2.4 KiB
Markdown
Raw Normal View History

---
base_model:
- Qwen/Qwen3-8B
library_name: transformers
tags:
- mergekit
- merge
license: apache-2.0
---
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 [TIES](https://arxiv.org/abs/2306.01708) merge method using C:\Users\Onlyth3ninja\Desktop\model merge\r2lm_local\cores\BaseMathCore_attempt5 as a base.
### Models Merged
The following models were included in the merge:
* C:\Users\Onlyth3ninja\Desktop\model merge\r2lm_local\cores\DualCoreFallback_attempt0
* C:\Users\Onlyth3ninja\Desktop\model merge\r2lm_local\cores\ReasonCore_attempt0
### Configuration
The following YAML configuration was used to produce this model:
```yaml
# Auto-generated merge config for OptRecursive attempt 12
# Intended output: C:/Users/Onlyth3ninja/Desktop/model merge/r2lm_local/cores/OptRecursive_attempt12
merge_method: ties
base_model: "C:\\Users\\Onlyth3ninja\\Desktop\\model merge\\r2lm_local\\cores\\BaseMathCore_attempt5"
tokenizer_source: "C:\\Users\\Onlyth3ninja\\Desktop\\model merge\\r2lm_local\\cores\\BaseMathCore_attempt5"
dtype: bfloat16
models:
- model: "C:\\Users\\Onlyth3ninja\\Desktop\\model merge\\r2lm_local\\cores\\BaseMathCore_attempt5"
parameters:
density: 1.0
weight:
- filter: lm_head
value: 0.6
- value: [0.6, 0.6, 0.6, 0.6, 0.6, 0.6, 0.6, 0.6, 0.6, 0.6, 0.6, 0.6, 0.6, 0.6, 0.6, 0.6, 0.6, 0.6, 0.6, 0.6, 0.6, 0.6, 0.6, 0.6, 0.6, 0.6, 0.6, 0.6, 0.6, 0.6, 0.6, 0.6, 0.6, 0.6, 0.6, 0.6]
- model: "C:\\Users\\Onlyth3ninja\\Desktop\\model merge\\r2lm_local\\cores\\DualCoreFallback_attempt0"
parameters:
density: 1.0
weight:
- filter: lm_head
value: 0.25
- value: [0.25, 0.25, 0.25, 0.25, 0.25, 0.25, 0.25, 0.25, 0.25, 0.25, 0.25, 0.25, 0.25, 0.25, 0.25, 0.25, 0.25, 0.25, 0.25, 0.25, 0.25, 0.25, 0.25, 0.25, 0.25, 0.25, 0.25, 0.25, 0.25, 0.25, 0.25, 0.25, 0.25, 0.25, 0.25, 0.25]
- model: "C:\\Users\\Onlyth3ninja\\Desktop\\model merge\\r2lm_local\\cores\\ReasonCore_attempt0"
parameters:
density: 1.0
weight:
- filter: lm_head
value: 0.15
- value: [0.15, 0.15, 0.15, 0.15, 0.15, 0.15, 0.15, 0.15, 0.15, 0.15, 0.15, 0.15, 0.15, 0.15, 0.15, 0.15, 0.15, 0.15, 0.15, 0.15, 0.15, 0.15, 0.15, 0.15, 0.15, 0.15, 0.15, 0.15, 0.15, 0.15, 0.15, 0.15, 0.15, 0.15, 0.15, 0.15]
parameters:
normalize: true
int8_mask: true
```