Create README.md
Browse files
README.md
ADDED
@@ -0,0 +1,49 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
license: apache-2.0
|
3 |
+
language:
|
4 |
+
- en
|
5 |
+
tags:
|
6 |
+
- creative
|
7 |
+
- creative writing
|
8 |
+
- fiction writing
|
9 |
+
- plot generation
|
10 |
+
- sub-plot generation
|
11 |
+
- fiction writing
|
12 |
+
- story generation
|
13 |
+
- scene continue
|
14 |
+
- storytelling
|
15 |
+
- fiction story
|
16 |
+
- story
|
17 |
+
- writing
|
18 |
+
- fiction
|
19 |
+
- roleplaying
|
20 |
+
- swearing
|
21 |
+
- rp
|
22 |
+
- horror
|
23 |
+
- llama3
|
24 |
+
- mergekit
|
25 |
+
pipeline_tag: text-generation
|
26 |
+
---
|
27 |
+
|
28 |
+
L3-SthenoMaidBlackroot-12.2B-V1-INSTRUCT-ULTRA
|
29 |
+
|
30 |
+
Merge of SthenoMaidBlackroot (8b) with Llama3 Instruct (8b) creating a model at 12.2B to improve instruction following and output.
|
31 |
+
|
32 |
+
Story / Scene / Fiction:
|
33 |
+
|
34 |
+
Unique "pre-amble" / "foreshadowing" of events before they happen instead of "immediate and into the fire" type prose.
|
35 |
+
|
36 |
+
Some improvement in logic/problem solving relative to SMB 8B.
|
37 |
+
|
38 |
+
This is an "Ultra Quality" source float32 -> GGUF transfer.
|
39 |
+
|
40 |
+
The F32 version exhibits even stronger creativity (detail, place, "there") , but slightly less logic/problem solving
|
41 |
+
relative to the F16 version (at a different repo).
|
42 |
+
|
43 |
+
Details:
|
44 |
+
|
45 |
+
- Requires Llama 3 Template and/or Command-R Template
|
46 |
+
- Context 8192, with rope 32K or higher.
|
47 |
+
- No special settings.
|
48 |
+
|
49 |
+
Quants uploading, detailed model card update / with examples pending.
|