Update README.md
Browse files
README.md
CHANGED
@@ -8,34 +8,16 @@ tags:
|
|
8 |
- exl2
|
9 |
- not-for-all-audiences
|
10 |
---
|
|
|
|
|
|
|
|
|
11 |
|
12 |
|
13 |
# Merged-Vicuna-RP-Stew-34B
|
14 |
|
15 |
This is a merge of pre-trained language models created using [mergekit](https://github.com/cg123/mergekit).
|
16 |
|
17 |
-
Specialized exl2 versions can be found here:
|
18 |
-
|
19 |
-
3.5 - https://huggingface.co/ParasiticRogue/Merged-RP-Stew-V2-34B-exl2-3.5?not-for-all-audiences=true
|
20 |
-
|
21 |
-
4.0 - https://huggingface.co/ParasiticRogue/Merged-RP-Stew-V2-34B-exl2-4.0?not-for-all-audiences=true
|
22 |
-
|
23 |
-
4.65 - https://huggingface.co/ParasiticRogue/Merged-RP-Stew-V2-34B-exl2-4.65-fix?not-for-all-audiences=true
|
24 |
-
|
25 |
-
5.0 - https://huggingface.co/ParasiticRogue/Merged-RP-Stew-V2-34B-exl2-5.0?not-for-all-audiences=true
|
26 |
-
|
27 |
-
6.0~8.0 provided by Dracones:
|
28 |
-
|
29 |
-
6.0 - https://huggingface.co/Dracones/Merged-RP-Stew-V2-34B_exl2_6.0bpw?not-for-all-audiences=true
|
30 |
-
|
31 |
-
7.0 - https://huggingface.co/Dracones/Merged-RP-Stew-V2-34B_exl2_7.0bpw?not-for-all-audiences=true
|
32 |
-
|
33 |
-
8.0 - https://huggingface.co/Dracones/Merged-RP-Stew-V2-34B_exl2_8.0bpw?not-for-all-audiences=true
|
34 |
-
|
35 |
-
Standardized GGUFs provided by MarsupialAI (Doesn't use Bluemoon-Light parquet):
|
36 |
-
|
37 |
-
https://huggingface.co/MarsupialAI/Merged-RP-Stew-V2-34B_iMatrix_GGUF?not-for-all-audiences=true
|
38 |
-
|
39 |
## Merge Details
|
40 |
|
41 |
New pot of stew with some slight seasoning added into the merging recipe. Besides being decent models, Capybara was chosen at a higher percentage for it's general aptitude plus preserving longer context length, Tess-1.5 is for better character/lore understanding, Nontoxic-Bagel SLERPed with PiVoT-SUS-RP (seperate from the main merge) is for chat/RP and storytelling diversity, while Nyakura SLERPed into CausalLM-RP is for even better chat/RP engagement. Both Nontoxic-Bagel and CausalLM-RP were used as the base of their respective SLERPs.
|
@@ -132,34 +114,4 @@ https://huggingface.co/Sao10K/NyakuraV2-34B-Yi-Llama
|
|
132 |
|
133 |
https://huggingface.co/NeverSleep/CausalLM-RP-34B
|
134 |
|
135 |
-
https://huggingface.co/chargoddard/Yi-34B-200K-Llama
|
136 |
-
|
137 |
-
### Configuration
|
138 |
-
|
139 |
-
The following YAML configuration was used to produce this model:
|
140 |
-
|
141 |
-
```yaml
|
142 |
-
models:
|
143 |
-
- model: Nyakura-CausalLM-RP-34B
|
144 |
-
parameters:
|
145 |
-
weight: 0.16
|
146 |
-
density: 0.42
|
147 |
-
- model: Nontoxic-PiVoT-Bagel-RP-34b
|
148 |
-
parameters:
|
149 |
-
weight: 0.22
|
150 |
-
density: 0.54
|
151 |
-
- model: Tess-34B-v1.5b
|
152 |
-
parameters:
|
153 |
-
weight: 0.28
|
154 |
-
density: 0.66
|
155 |
-
- model: Nous-Capybara-34B-V1.9
|
156 |
-
parameters:
|
157 |
-
weight: 0.34
|
158 |
-
density: 0.78
|
159 |
-
merge_method: dare_ties
|
160 |
-
base_model: Yi-34B-200K-Llama
|
161 |
-
parameters:
|
162 |
-
int8_mask: true
|
163 |
-
dtype: bfloat16
|
164 |
-
|
165 |
-
```
|
|
|
8 |
- exl2
|
9 |
- not-for-all-audiences
|
10 |
---
|
11 |
+
ORIGIGNAL MODEL LINK https://huggingface.co/ParasiticRogue/Merged-RP-Stew-V2-34B
|
12 |
+
|
13 |
+
Hi, this is the rp-stew-v2 model enlarged by half, up to 90 layers. To be honest, I don't know why, but someone might need it. I'm just testing it myself, compared to the original.
|
14 |
+
I will post the exl2 quantization of 4 bits soon.
|
15 |
|
16 |
|
17 |
# Merged-Vicuna-RP-Stew-34B
|
18 |
|
19 |
This is a merge of pre-trained language models created using [mergekit](https://github.com/cg123/mergekit).
|
20 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
21 |
## Merge Details
|
22 |
|
23 |
New pot of stew with some slight seasoning added into the merging recipe. Besides being decent models, Capybara was chosen at a higher percentage for it's general aptitude plus preserving longer context length, Tess-1.5 is for better character/lore understanding, Nontoxic-Bagel SLERPed with PiVoT-SUS-RP (seperate from the main merge) is for chat/RP and storytelling diversity, while Nyakura SLERPed into CausalLM-RP is for even better chat/RP engagement. Both Nontoxic-Bagel and CausalLM-RP were used as the base of their respective SLERPs.
|
|
|
114 |
|
115 |
https://huggingface.co/NeverSleep/CausalLM-RP-34B
|
116 |
|
117 |
+
https://huggingface.co/chargoddard/Yi-34B-200K-Llama
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|