File size: 1,360 Bytes
63b4ea4 f6590e6 63b4ea4 45ff5fb f6590e6 63b4ea4 f6590e6 63b4ea4 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 |
---
base_model: LumiOpen/Viking-7B
language:
- en
- fi
- sv
- 'no'
- da
- is
- nn
license: apache-2.0
tags:
- text-generation-inference
- transformers
- unsloth
- llama
- trl
- sft
datasets:
- Gryphe/Sonnet3.5-SlimOrcaDedupCleaned
- mpasila/Sonnet3.5-SlimOrcaDedupCleaned-4k-context
---
This is the fully trained version (with fixed formatting!!).
Dataset used: [Gryphe/Sonnet3.5-SlimOrcaDedupCleaned](https://huggingface.co/datasets/Gryphe/Sonnet3.5-SlimOrcaDedupCleaned) which was further [filtered](https://huggingface.co/datasets/mpasila/Sonnet3.5-SlimOrcaDedupCleaned-4k-context) to remove prompts/examples that are longer than 4076 tokens (removed about 385 examples).
Prompt format is: ChatML
LoRA: [mpasila/Viking-SlimSonnet-v1-LoRA-7B](https://huggingface.co/mpasila/Viking-SlimSonnet-v1-LoRA-7B)
Trained with regular LoRA (not quantized/QLoRA) and LoRA rank was 128 and Alpha set to 32. Trained for 1 epoch using A40 for about 23 hours.
# Uploaded model
- **Developed by:** mpasila
- **License:** apache-2.0
- **Finetuned from model :** LumiOpen/Viking-7B
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)
|