Upload README.txt
Browse files- README.txt +52 -0
README.txt
ADDED
|
@@ -0,0 +1,52 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
base_model:
|
| 3 |
+
- inflatebot/MN-12B-Mag-Mell-R1
|
| 4 |
+
- TheDrummer/UnslopNemo-12B-v4.1
|
| 5 |
+
library_name: transformers
|
| 6 |
+
tags:
|
| 7 |
+
- mergekit
|
| 8 |
+
- merge
|
| 9 |
+
- 12b
|
| 10 |
+
- chat
|
| 11 |
+
- creative
|
| 12 |
+
---
|
| 13 |
+
# patricide-Unslop-Mell
|
| 14 |
+
>The sins of the Father shan't ever be repeated this way.
|
| 15 |
+
|
| 16 |
+
# WARNING: Something went wrong during the upload! It will be fixed soon.
|
| 17 |
+
This is a merge of pre-trained language models created using [mergekit](https://github.com/cg123/mergekit).
|
| 18 |
+
|
| 19 |
+
This is my first merge — I still have no idea how writing the parameters in the config actually works. If anyone has more extensive guides for merging, please let me know. I would also like to get into the science behind all this.
|
| 20 |
+
|
| 21 |
+
Both models produced enjoyable results, so I decided to merge them, to create a model hopefully inheriting good traits of the parents.
|
| 22 |
+
|
| 23 |
+
I've tested this model on the *Q_6K GGUF* Quant (will get uploaded later) and it provided satisfactory results, thus I decided to upload it. Although I've not extensively tested it in Storywriting nor RP, the results were stable and *at least* coherent. I tested it on a **Temperature of 1** (Temperature last) and **Min-P of 0.1**. I don't know the effects **DRY** or **XTC** have on the stability of the output, or how it fares on high context sizes. Both parent models use the **ChatML** Template. Although [Unslop-Nemo](https://huggingface.co/TheDrummer/UnslopNemo-12B-v4.1) also uses **Metharme/Pygmalion**. I've not yet tested which works better.
|
| 24 |
+
|
| 25 |
+
Feel free to experiment, as I am only experimenting myself.
|
| 26 |
+
|
| 27 |
+
## Merge Details
|
| 28 |
+
### Merge Method
|
| 29 |
+
|
| 30 |
+
This model was merged using the SLERP merge method.
|
| 31 |
+
|
| 32 |
+
### Models Merged
|
| 33 |
+
|
| 34 |
+
The following models were included in the merge:
|
| 35 |
+
* [inflatebot/MN-12B-Mag-Mell-R1](https://huggingface.co/inflatebot/MN-12B-Mag-Mell-R1)
|
| 36 |
+
* [TheDrummer/UnslopNemo-12B-v4.1](https://huggingface.co/TheDrummer/UnslopNemo-12B-v4.1)
|
| 37 |
+
|
| 38 |
+
### Configuration
|
| 39 |
+
|
| 40 |
+
The following YAML configuration was used to produce this model:
|
| 41 |
+
|
| 42 |
+
```yaml
|
| 43 |
+
models:
|
| 44 |
+
- model: TheDrummer/UnslopNemo-12B-v4.1
|
| 45 |
+
- model: inflatebot/MN-12B-Mag-Mell-R1
|
| 46 |
+
merge_method: slerp
|
| 47 |
+
base_model: TheDrummer/UnslopNemo-12B-v4.1
|
| 48 |
+
dtype: bfloat16
|
| 49 |
+
parameters:
|
| 50 |
+
t: [0, 0.5, 1, 0.5, 0]
|
| 51 |
+
|
| 52 |
+
```
|