Update README.md
Browse files
README.md
CHANGED
|
@@ -9,14 +9,90 @@ tags:
|
|
| 9 |
- mistral
|
| 10 |
- trl
|
| 11 |
base_model: teknium/OpenHermes-2.5-Mistral-7B
|
|
|
|
|
|
|
| 12 |
---
|
| 13 |
|
| 14 |
-
#
|
| 15 |
|
| 16 |
-
|
| 17 |
-
|
| 18 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 19 |
|
| 20 |
This mistral model was trained 2x faster with [Unsloth](https://github.com/unslothai/unsloth) and Huggingface's TRL library.
|
| 21 |
|
| 22 |
-
[<img src="https://raw.githubusercontent.com/unslothai/unsloth/main/images/unsloth%20made%20with%20love.png" width="200"/>](https://github.com/unslothai/unsloth)
|
|
|
|
| 9 |
- mistral
|
| 10 |
- trl
|
| 11 |
base_model: teknium/OpenHermes-2.5-Mistral-7B
|
| 12 |
+
datasets:
|
| 13 |
+
- cnbeining/sentence-segmentation-dpo-raw
|
| 14 |
---
|
| 15 |
|
| 16 |
+
# OpenHermes-2.5-Mistral-7B-Sentence-Segmentation
|
| 17 |
|
| 18 |
+
## Model description
|
| 19 |
+
|
| 20 |
+
`OpenHermes-2.5-Mistral-7B-Sentence-Segmentation` is a DPO finetuned OpenHermes model for sentence segmentation capability.
|
| 21 |
+
|
| 22 |
+
This model is based on `teknium/OpenHermes-2.5-Mistral-7B`, a state-of-the-art chat-aligned 7B model.
|
| 23 |
+
|
| 24 |
+
## Example Outputs
|
| 25 |
+
|
| 26 |
+
The model has been finetuned with (ChatML)[https://github.com/MicrosoftDocs/azure-docs/blob/main/articles/ai-services/openai/includes/chat-markup-language.md#messages] template:
|
| 27 |
+
|
| 28 |
+
```
|
| 29 |
+
<|im_start|>system
|
| 30 |
+
Segment:<|im_end|>
|
| 31 |
+
<|im_start|>user
|
| 32 |
+
```yaml
|
| 33 |
+
"input":
|
| 34 |
+
"sentence":
|
| 35 |
+
"segment":
|
| 36 |
+
- "word": "Shere,"
|
| 37 |
+
- "word": "in"
|
| 38 |
+
- "word": "your"
|
| 39 |
+
- "word": "report"
|
| 40 |
+
- "word": "on"
|
| 41 |
+
- "word": "female"
|
| 42 |
+
- "word": "sexuality,"
|
| 43 |
+
- "word": "men"
|
| 44 |
+
- "word": "were"
|
| 45 |
+
- "word": "staggered"
|
| 46 |
+
- "word": "to"
|
| 47 |
+
- "word": "learn"
|
| 48 |
+
- "word": "that"
|
| 49 |
+
- "word": "clitoral"
|
| 50 |
+
- "word": "stimulation"
|
| 51 |
+
- "word": "was"
|
| 52 |
+
- "word": "much"
|
| 53 |
+
- "word": "more"
|
| 54 |
+
- "word": "important"
|
| 55 |
+
- "word": "than"
|
| 56 |
+
- "word": "penetration."
|
| 57 |
+
```<|im_end|>
|
| 58 |
+
<|im_start|>assistant
|
| 59 |
+
```
|
| 60 |
+
|
| 61 |
+
with output in the format of
|
| 62 |
+
|
| 63 |
+
```
|
| 64 |
+
```yaml
|
| 65 |
+
"output":
|
| 66 |
+
"sentence":
|
| 67 |
+
"segment":
|
| 68 |
+
- "word": "Shere,"
|
| 69 |
+
- "word": "in"
|
| 70 |
+
- "word": "your"
|
| 71 |
+
- "word": "report"
|
| 72 |
+
- "word": "on"
|
| 73 |
+
- "word": "female"
|
| 74 |
+
- "word": "sexuality,"
|
| 75 |
+
"segment":
|
| 76 |
+
- "word": "men"
|
| 77 |
+
- "word": "were"
|
| 78 |
+
- "word": "staggered"
|
| 79 |
+
- "word": "to"
|
| 80 |
+
- "word": "learn"
|
| 81 |
+
- "word": "that"
|
| 82 |
+
"segment":
|
| 83 |
+
- "word": "clitoral"
|
| 84 |
+
- "word": "stimulation"
|
| 85 |
+
- "word": "was"
|
| 86 |
+
- "word": "much"
|
| 87 |
+
- "word": "more"
|
| 88 |
+
- "word": "important"
|
| 89 |
+
- "word": "than"
|
| 90 |
+
- "word": "penetration."
|
| 91 |
+
```
|
| 92 |
+
```
|
| 93 |
+
|
| 94 |
+
## Misc
|
| 95 |
|
| 96 |
This mistral model was trained 2x faster with [Unsloth](https://github.com/unslothai/unsloth) and Huggingface's TRL library.
|
| 97 |
|
| 98 |
+
[<img src="https://raw.githubusercontent.com/unslothai/unsloth/main/images/unsloth%20made%20with%20love.png" width="200"/>](https://github.com/unslothai/unsloth)
|