Lewdiculous
commited on
Commit
•
e2ef386
1
Parent(s):
9e88a38
Update README.md
Browse files
README.md
CHANGED
@@ -12,54 +12,32 @@ tags:
|
|
12 |
- roleplay
|
13 |
- imatrix
|
14 |
- mistral
|
15 |
-
|
16 |
inference: false
|
17 |
---
|
18 |
|
19 |
-
This repository hosts GGUF-Imatrix quantizations for [
|
|
|
20 |
|
21 |
-
To be uploaded:
|
22 |
|
23 |
```python
|
24 |
quantization_options = [
|
25 |
-
"
|
26 |
-
"Q8_0", "
|
27 |
]
|
28 |
```
|
29 |
|
30 |
-
**This is experimental.**
|
31 |
-
|
32 |
-
For imatrix data generation, kalomaze's `groups_merged.txt` with added roleplay chats was used, you can find it [here](https://huggingface.co/Lewdiculous/Datura_7B-GGUF-Imatrix/blob/main/imatrix-with-rp-format-data.txt).
|
33 |
|
34 |
The goal is to measure the (hopefully positive) impact of this data for consistent formatting in roleplay chatting scenarios.
|
35 |
|
36 |
-
**
|
37 |
|
38 |
-
![image/
|
39 |
|
40 |
-
|
41 |
-
* [ResplendentAI/Datura_7B](https://huggingface.co/ResplendentAI/Datura_7B)
|
42 |
-
* [Test157t/Eris-Floramix-7b](https://huggingface.co/Test157t/Eris-Floramix-7b)
|
43 |
|
44 |
-
|
45 |
|
46 |
-
|
47 |
|
48 |
-
|
49 |
-
slices:
|
50 |
-
- sources:
|
51 |
-
- model: Test157t/Eris-Floramix-7b
|
52 |
-
layer_range: [0, 32]
|
53 |
-
- model: ResplendentAI/Datura_7B
|
54 |
-
layer_range: [0, 32]
|
55 |
-
merge_method: slerp
|
56 |
-
base_model: Test157t/Eris-Floramix-7b
|
57 |
-
parameters:
|
58 |
-
t:
|
59 |
-
- filter: self_attn
|
60 |
-
value: [0, 0.5, 0.3, 0.7, 1]
|
61 |
-
- filter: mlp
|
62 |
-
value: [1, 0.5, 0.7, 0.3, 0]
|
63 |
-
- value: 0.5
|
64 |
-
dtype: bfloat16
|
65 |
-
```
|
|
|
12 |
- roleplay
|
13 |
- imatrix
|
14 |
- mistral
|
15 |
+
|
16 |
inference: false
|
17 |
---
|
18 |
|
19 |
+
This repository hosts GGUF-Imatrix quantizations for [ChaoticNeutrals/Eris_Floramix_DPO_7B](https://huggingface.co/ChaoticNeutrals/Eris_Floramix_DPO_7B).
|
20 |
+
|
21 |
|
|
|
22 |
|
23 |
```python
|
24 |
quantization_options = [
|
25 |
+
"Q3_K_M", "Q4_K_M", "Q5_K_M", "Q6_K",
|
26 |
+
"Q8_0", "IQ4_XS", "IQ3_XXS"
|
27 |
]
|
28 |
```
|
29 |
|
|
|
|
|
|
|
30 |
|
31 |
The goal is to measure the (hopefully positive) impact of this data for consistent formatting in roleplay chatting scenarios.
|
32 |
|
33 |
+
**Image:**
|
34 |
|
35 |
+
![image/png](https://cdn-uploads.huggingface.co/production/uploads/65d4cf2693a0a3744a27536c/1kuWjS4TnCIkCcdZdPfsG.png)
|
36 |
|
37 |
+
**Original model information:**
|
|
|
|
|
38 |
|
39 |
+
# Eris Floramix DPO
|
40 |
|
41 |
+
This is a mix between Eris Remix DPO and Flora DPO, a finetune of the original Eris Remix on the Synthetic_Soul_1k dataset.
|
42 |
|
43 |
+
Applied this DPO dataset: https://huggingface.co/datasets/athirdpath/DPO_Pairs-Roleplay-Alpaca-NSFW-v1-SHUFFLED
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|