Update README.md
Browse files
README.md
CHANGED
|
@@ -1,20 +1,20 @@
|
|
| 1 |
-
---
|
| 2 |
-
license: apache-2.0
|
| 3 |
-
datasets:
|
| 4 |
-
- Crystalcareai/openhermes_200k_unfiltered
|
| 5 |
-
- mlabonne/orpo-dpo-mix-40k
|
| 6 |
-
- jondurbin/airoboros-3.2
|
| 7 |
-
- abacusai/SystemChat-1.1
|
| 8 |
-
- trollek/SimpleInstructionJudge-v01
|
| 9 |
-
- cgato/SlimOrcaDedupCleaned
|
| 10 |
-
language:
|
| 11 |
-
- en
|
| 12 |
-
library_name: transformers
|
| 13 |
-
base_model: h2oai/h2o-danube3-4b-base
|
| 14 |
-
tags:
|
| 15 |
-
- mergekit
|
| 16 |
-
- magpie
|
| 17 |
-
---
|
| 18 |
# LittlePromptMaker-4B-v0.1
|
| 19 |
|
| 20 |
A small model to create prompts the [Magpie](https://arxiv.org/abs/2406.08464) way.
|
|
@@ -31,10 +31,20 @@ In a different language: Ved hjælp at Husskades indsigt, hvor man udnytter spro
|
|
| 31 |
|
| 32 |
All the datasets were used seperately and merged together using [Model Stock](https://arxiv.org/abs/2403.19522), except for SystemChat-1.1 where I fine-tuned it using [LoRA+](https://arxiv.org/abs/2402.12354) with `train_on_prompt` set to True.
|
| 33 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 34 |
## Using
|
| 35 |
|
| 36 |
```jinja2
|
| 37 |
<|im_start|>system
|
| 38 |
{{system_message}}<|im_end|>
|
| 39 |
<|im_start|>user
|
| 40 |
-
```
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
datasets:
|
| 4 |
+
- Crystalcareai/openhermes_200k_unfiltered
|
| 5 |
+
- mlabonne/orpo-dpo-mix-40k
|
| 6 |
+
- jondurbin/airoboros-3.2
|
| 7 |
+
- abacusai/SystemChat-1.1
|
| 8 |
+
- trollek/SimpleInstructionJudge-v01
|
| 9 |
+
- cgato/SlimOrcaDedupCleaned
|
| 10 |
+
language:
|
| 11 |
+
- en
|
| 12 |
+
library_name: transformers
|
| 13 |
+
base_model: h2oai/h2o-danube3-4b-base
|
| 14 |
+
tags:
|
| 15 |
+
- mergekit
|
| 16 |
+
- magpie
|
| 17 |
+
---
|
| 18 |
# LittlePromptMaker-4B-v0.1
|
| 19 |
|
| 20 |
A small model to create prompts the [Magpie](https://arxiv.org/abs/2406.08464) way.
|
|
|
|
| 31 |
|
| 32 |
All the datasets were used seperately and merged together using [Model Stock](https://arxiv.org/abs/2403.19522), except for SystemChat-1.1 where I fine-tuned it using [LoRA+](https://arxiv.org/abs/2402.12354) with `train_on_prompt` set to True.
|
| 33 |
|
| 34 |
+
### Datasets
|
| 35 |
+
|
| 36 |
+
* [Airoboros-3.2](https://huggingface.co/datasets/jondurbin/airoboros-3.2) (CC BY 4.0) by [jondurbin](https://huggingface.co/jondurbin)
|
| 37 |
+
* [SystemChat-1.1](https://huggingface.co/datasets/abacusai/SystemChat-1.1) by [abacusai](https://huggingface.co/abacusai)
|
| 38 |
+
* [orpo-dpo-mix-40k](https://huggingface.co/datasets/mlabonne/orpo-dpo-mix-40k) by [mlabonne](https://huggingface.co/mlabonne)
|
| 39 |
+
* [SlimOrcaDedupCleaned](https://huggingface.co/datasets/cgato/SlimOrcaDedupCleaned) by [cgato](https://huggingface.co/cgato)
|
| 40 |
+
* [openhermes_200k_unfiltered](https://huggingface.co/datasets/Crystalcareai/openhermes_200k_unfiltered) by [Crystalcareai](https://huggingface.co/Crystalcareai)
|
| 41 |
+
|
| 42 |
## Using
|
| 43 |
|
| 44 |
```jinja2
|
| 45 |
<|im_start|>system
|
| 46 |
{{system_message}}<|im_end|>
|
| 47 |
<|im_start|>user
|
| 48 |
+
```
|
| 49 |
+
|
| 50 |
+
It actually generates an EOS token at the end of a *"user"* prompt. Lawdy that has been a pain when trying to use large models for this purpose. Good luck; have fun.
|