Update README.md

README.md

@@ -12,60 +12,14 @@ tags:
 - merge
 
 ---
-# merge
+This is a merge of pre-trained language models
 
-This is a merge of pre-trained language models created using [mergekit](https://github.com/cg123/mergekit).
+Designed as part of ru capable 7B.
 
-## Merge Details
-### Merge Method
+RP, ERP, chat, it is good and fast. Sometimes hallucinate, sometimes writes excellent from first try.
 
-This model was merged using the [TIES](https://arxiv.org/abs/2306.01708) merge method using [ChaoticNeutrals/BuRP_7B](https://huggingface.co/ChaoticNeutrals/BuRP_7B) as a base.
+This one is more stable than v3
 
-### Models Merged
+Of course, better try at least 12B with offloading, may be slower, but way "smarter" than any 7/8B.
 
-The following models were included in the merge:
-* [TeeZee/DarkSapling-7B-v1.1](https://huggingface.co/TeeZee/DarkSapling-7B-v1.1)
-* [Endevor/EndlessRP-v2-7B](https://huggingface.co/Endevor/EndlessRP-v2-7B)
-* [rmdhirr/Foxglove_7B](https://huggingface.co/rmdhirr/Foxglove_7B)
-* [kainatq/KPT-7B](https://huggingface.co/kainatq/KPT-7B)
-* [GlobalMeltdown/MaidenlessNoMore-7B](https://huggingface.co/GlobalMeltdown/MaidenlessNoMore-7B)
-
-### Configuration
-
-The following YAML configuration was used to produce this model:
-
-```yaml
-models:
-  - model: rmdhirr/Foxglove_7B
-    parameters:
-      density: 0.5
-      weight: 0.5
-  - model: kainatq/KPT-7B
-    parameters:
-      density: 0.5
-      weight: 0.5
-  - model: GlobalMeltdown/MaidenlessNoMore-7B
-    parameters:
-      density: 0.5
-      weight: 0.5
-  - model: TeeZee/DarkSapling-7B-v1.1
-    parameters:
-      density: 0.5
-      weight: 0.5
-  - model: Endevor/EndlessRP-v2-7B
-    parameters:
-      density: 0.5
-      weight: 0.5
-  - model: ChaoticNeutrals/BuRP_7B
-    parameters:
-      density: 0.6
-      weight: 0.6
-merge_method: ties
-base_model: ChaoticNeutrals/BuRP_7B
-parameters:
-  normalize: false
-  int8_mask: true
-dtype: float16
-tokenizer:
- source: union
-```
+Tested on ChatML t1.01