49 lines
1.2 KiB
Markdown
49 lines
1.2 KiB
Markdown
|
---
|
||
|
|
base_model: []
|
||
|
|
library_name: transformers
|
||
|
|
tags:
|
||
|
|
- mergekit
|
||
|
|
- merge
|
||
|
|
|
||
|
|
---
|
||
|
|
# all_full_ta
|
||
|
|
|
||
|
|
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 [Task Arithmetic](https://arxiv.org/abs/2212.04089) merge method using checkpoints/Qwen3-1.7B-baseline as a base.
|
||
|
|
|
||
|
|
### Models Merged
|
||
|
|
|
||
|
|
The following models were included in the merge:
|
||
|
|
* checkpoints/multilinguality_model/sft_multilingual_v16_nonthinking_1e-4_warmup0p1
|
||
|
|
* checkpoints/math_model/best_model
|
||
|
|
* checkpoints/safety_model/best_model_sft_dpo/hf_model
|
||
|
|
* checkpoints/general_knowledge_model/full_sft
|
||
|
|
|
||
|
|
### Configuration
|
||
|
|
|
||
|
|
The following YAML configuration was used to produce this model:
|
||
|
|
|
||
|
|
```yaml
|
||
|
|
merge_method: task_arithmetic
|
||
|
|
base_model: checkpoints/Qwen3-1.7B-baseline
|
||
|
|
models:
|
||
|
|
- model: checkpoints/math_model/best_model
|
||
|
|
parameters:
|
||
|
|
weight: 0.25
|
||
|
|
- model: checkpoints/safety_model/best_model_sft_dpo/hf_model
|
||
|
|
parameters:
|
||
|
|
weight: 0.25
|
||
|
|
- model: checkpoints/general_knowledge_model/full_sft
|
||
|
|
parameters:
|
||
|
|
weight: 0.25
|
||
|
|
- model: checkpoints/multilinguality_model/sft_multilingual_v16_nonthinking_1e-4_warmup0p1
|
||
|
|
parameters:
|
||
|
|
weight: 0.25
|
||
|
|
dtype: bfloat16
|
||
|
|
|
||
|
|
```
|