Crystalcareai
commited on
Commit
•
1f203bc
1
Parent(s):
4d4244e
Update README.md
Browse files
README.md
CHANGED
@@ -7,12 +7,8 @@ tags:
|
|
7 |
datasets:
|
8 |
- cognitivecomputations/Dolphin-2.9
|
9 |
- teknium/OpenHermes-2.5
|
10 |
-
- m-a-p/CodeFeedback-Filtered-Instruction
|
11 |
-
- cognitivecomputations/dolphin-coder
|
12 |
- cognitivecomputations/samantha-data
|
13 |
- microsoft/orca-math-word-problems-200k
|
14 |
-
- Locutusque/function-calling-chatml
|
15 |
-
- internlm/Agent-FLAN
|
16 |
---
|
17 |
|
18 |
# Dolphin 2.9.3 Qwen2 0.5B 🐬
|
@@ -31,6 +27,8 @@ This model is based on Qwen2-0.5b, and is governed by the Apache-2.0
|
|
31 |
|
32 |
The base model has 128k context, and the full-weight fine-tuning was with 16k sequence length.
|
33 |
|
|
|
|
|
34 |
|
35 |
example:
|
36 |
|
@@ -43,7 +41,7 @@ You are Dolphin, a helpful AI assistant.<|im_end|>
|
|
43 |
|
44 |
```
|
45 |
|
46 |
-
Dolphin-2.9.3 has a variety of instruction,
|
47 |
|
48 |
Dolphin is uncensored. We have filtered the dataset to remove alignment and bias. This makes the model more compliant. You are advised to implement your own alignment layer before exposing the model as a service. It will be highly compliant with any requests, even unethical ones. Please read my blog post about uncensored models. https://erichartford.com/uncensored-models You are responsible for any content you create using this model. Enjoy responsibly.
|
49 |
|
|
|
7 |
datasets:
|
8 |
- cognitivecomputations/Dolphin-2.9
|
9 |
- teknium/OpenHermes-2.5
|
|
|
|
|
10 |
- cognitivecomputations/samantha-data
|
11 |
- microsoft/orca-math-word-problems-200k
|
|
|
|
|
12 |
---
|
13 |
|
14 |
# Dolphin 2.9.3 Qwen2 0.5B 🐬
|
|
|
27 |
|
28 |
The base model has 128k context, and the full-weight fine-tuning was with 16k sequence length.
|
29 |
|
30 |
+
Due to the complexities of fine tuning smaller models on datasets created by/for larger models - we removed all coding, function calling and systemchat-multilingual from the dataset for this model.
|
31 |
+
|
32 |
|
33 |
example:
|
34 |
|
|
|
41 |
|
42 |
```
|
43 |
|
44 |
+
Dolphin-2.9.3 has a variety of instruction, and conversational skills.
|
45 |
|
46 |
Dolphin is uncensored. We have filtered the dataset to remove alignment and bias. This makes the model more compliant. You are advised to implement your own alignment layer before exposing the model as a service. It will be highly compliant with any requests, even unethical ones. Please read my blog post about uncensored models. https://erichartford.com/uncensored-models You are responsible for any content you create using this model. Enjoy responsibly.
|
47 |
|