Update README.md
Browse files
README.md
CHANGED
@@ -3,9 +3,9 @@ datasets:
|
|
3 |
- PygmalionAI/PIPPA
|
4 |
---
|
5 |
|
6 |
-
##
|
7 |
|
8 |
-
|
9 |
|
10 |
## Prompting
|
11 |
|
@@ -15,12 +15,12 @@ The current model version has been trained on prompts using three different role
|
|
15 |
|
16 |
The `<|system|>` prompt can be used to inject out-of-channel information behind the scenes, while the `<|user|>` prompt should be used to indicate user input. The `<|model|>` token should then be used to indicate that the model should generate a response. These tokens can happen multiple times and be chained up to form a conversation history.
|
17 |
|
18 |
-
##
|
19 |
|
20 |
-
|
21 |
|
22 |
-
|
23 |
|
24 |
-
|
25 |
|
26 |
-
|
|
|
3 |
- PygmalionAI/PIPPA
|
4 |
---
|
5 |
|
6 |
+
## GGUF
|
7 |
|
8 |
+
gguf quants for ludis/tsukasa-limarp-7b
|
9 |
|
10 |
## Prompting
|
11 |
|
|
|
15 |
|
16 |
The `<|system|>` prompt can be used to inject out-of-channel information behind the scenes, while the `<|user|>` prompt should be used to indicate user input. The `<|model|>` token should then be used to indicate that the model should generate a response. These tokens can happen multiple times and be chained up to form a conversation history.
|
17 |
|
18 |
+
## Training
|
19 |
|
20 |
+
base model (llama-2-7b)
|
21 |
|
22 |
+
tuned on commit de693ac of the koishi dataset for 1 epoch as apart of ludis/tsukasa-7b
|
23 |
|
24 |
+
tuned on commit 36fc235 of pippa metharme for 1 epoch as apart of ludis/tsukasa-7b
|
25 |
|
26 |
+
tuned on Version 2023-09-03 of LimaRP (without ponyville, lolicit, all the fallen, and eka's portal subsets) for 2 epochs
|