RJuro
/

kanelsnegl-v0.2

Text Generation

Inference Endpoints

text-generation-inference

Model card Files Files and versions Community

RJuro commited on Dec 21, 2023

Commit

1fd7bc7

•

1 Parent(s): f3625f0

Update README.md

Files changed (1) hide show

README.md +4 -3

README.md CHANGED Viewed

@@ -13,9 +13,10 @@ A Danish finetune of Zephyr-7b-alpha 😀 The idea with this model (apart from p
 Try it here [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/RJuro/courses/blob/main/notebooks/Kanelsnegl_v0_2_usecases.ipynb)
-<!--
-<img src="https://huggingface.co/RJuro/kanelsnegl-v0.1/resolve/main/kanelsnegl_banner.png" alt="Kanelsnegl Logo" width="800" style="margin-left:'auto' margin-right:'auto' display:'block'"/>
--->
 ## Model Description
 Base model: [Zephyr-7b-alpha](https://huggingface.co/HuggingFaceH4/zephyr-7b-alpha) finetuned on [DDSC/partial-danish-gigaword-no-twitter](https://huggingface.co/datasets/DDSC/partial-danish-gigaword-no-twitter). The training involved a maximum length of 512. QLora completion finetuning of all linear layers was also implemented. This model is mostly fun tinkering for personal learning purpose.
 This version got 4 times more fine-tuning than v0.1 [RJuro/kanelsnegl-v0.1](https://huggingface.co/RJuro/kanelsnegl-v0.1). It produces better Danish and follows complex prompts and instructions.

 Try it here [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/RJuro/courses/blob/main/notebooks/Kanelsnegl_v0_2_usecases.ipynb)
+<img src="https://huggingface.co/RJuro/kanelsnegl-v0.2/resolve/main/banner_ks_s.jpg" alt="Kanelsnegl Logo" width="800" style="margin-left:'auto' margin-right:'auto' display:'block'"/>
 ## Model Description
 Base model: [Zephyr-7b-alpha](https://huggingface.co/HuggingFaceH4/zephyr-7b-alpha) finetuned on [DDSC/partial-danish-gigaword-no-twitter](https://huggingface.co/datasets/DDSC/partial-danish-gigaword-no-twitter). The training involved a maximum length of 512. QLora completion finetuning of all linear layers was also implemented. This model is mostly fun tinkering for personal learning purpose.
 This version got 4 times more fine-tuning than v0.1 [RJuro/kanelsnegl-v0.1](https://huggingface.co/RJuro/kanelsnegl-v0.1). It produces better Danish and follows complex prompts and instructions.