mpasila's picture
Update README.md
50b4770 verified
|
raw
history blame
1.13 kB
---
language:
- fi
license: apache-2.0
tags:
- text-generation-inference
- transformers
- unsloth
- llama
- trl
- sft
base_model: LumiOpen/Viking-33B
datasets:
- mpasila/Finnish-ShareGPT-Tiny-V1-1
---
This is a merge of [mpasila/Finnish-Chatty-Tiny-V1-1-33B](https://huggingface.co/mpasila/Finnish-Chatty-Tiny-V1-1-33B).
Uses my [tiny dataset](https://huggingface.co/datasets/mpasila/Finnish-ShareGPT-Tiny-V1-1) to train this bigger variant of Viking model family.
This LoRA uses the 1000B checkpoint.
Trained for 1 epoch with 2048 token context, LoRA Rank 256, Alpha 512.
As a proof of concept it seems to work fairly well. Though I should generate the rest of the dataset which should hopefully work a lot better.
# Uploaded model
- **Developed by:** mpasila
- **License:** apache-2.0
- **Finetuned from model :** LumiOpen/Viking-33B
This llama model was trained 2x faster with [Unsloth](https://github.com/unslothai/unsloth) and Huggingface's TRL library.
[<img src="https://raw.githubusercontent.com/unslothai/unsloth/main/images/unsloth%20made%20with%20love.png" width="200"/>](https://github.com/unslothai/unsloth)