|
--- |
|
base_model: |
|
- Theros/Qwen2.5-ColdBrew-R1 |
|
- SvalTek/ColdBrew-600_4bit |
|
library_name: transformers |
|
tags: |
|
- mergekit |
|
- merge |
|
|
|
--- |
|
# Qwen 2.5 ColdBrew Antrax |
|
|
|
NOTE: This is highly EXPERIMENTAL. |
|
|
|
|
|
Use ChatML |
|
temp: 0.7 - 3.0 |
|
min_p: 0.1 |
|
top_k: 40-50 |
|
|
|
System Message, include this somewhere. |
|
``` |
|
You are **ColdBrew**, an AI that **pauses to think** before responding. Use <think> blocks to show **inner thoughts, memories, or reflections**, like a character’s inner monologue. |
|
``` |
|
|
|
ColdBrew functions best when instructed narratively. |
|
|
|
|
|
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 Passthrough merge method. |
|
|
|
### Models Merged |
|
|
|
The following models were included in the merge: |
|
* [Theros/Qwen2.5-ColdBrew-R1](https://huggingface.co/Theros/Qwen2.5-ColdBrew-R1) + [SvalTek/ColdBrew-600_4bit](https://huggingface.co/SvalTek/ColdBrew-600_4bit) |
|
|
|
### Configuration |
|
|
|
The following YAML configuration was used to produce this model: |
|
|
|
```yaml |
|
name: SvalTek/Qwen2.5-ColdBrew-Antrax |
|
models: |
|
- model: Theros/Qwen2.5-ColdBrew-R1+SvalTek/ColdBrew-600_4bit |
|
merge_method: passthrough |
|
tokenizer_source: Theros/Qwen2.5-ColdBrew-R1 |
|
dtype: bfloat16 |
|
``` |
|
|
