Update README.md
Browse files
README.md
CHANGED
|
@@ -2,6 +2,12 @@
|
|
| 2 |
base_model: LumiOpen/Viking-7B
|
| 3 |
language:
|
| 4 |
- en
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 5 |
license: apache-2.0
|
| 6 |
tags:
|
| 7 |
- text-generation-inference
|
|
@@ -9,7 +15,20 @@ tags:
|
|
| 9 |
- unsloth
|
| 10 |
- llama
|
| 11 |
- trl
|
|
|
|
|
|
|
|
|
|
|
|
|
| 12 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 13 |
|
| 14 |
# Uploaded model
|
| 15 |
|
|
@@ -19,4 +38,4 @@ tags:
|
|
| 19 |
|
| 20 |
This llama 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)
|
|
|
|
| 2 |
base_model: LumiOpen/Viking-7B
|
| 3 |
language:
|
| 4 |
- en
|
| 5 |
+
- fi
|
| 6 |
+
- sv
|
| 7 |
+
- 'no'
|
| 8 |
+
- da
|
| 9 |
+
- is
|
| 10 |
+
- nn
|
| 11 |
license: apache-2.0
|
| 12 |
tags:
|
| 13 |
- text-generation-inference
|
|
|
|
| 15 |
- unsloth
|
| 16 |
- llama
|
| 17 |
- trl
|
| 18 |
+
datasets:
|
| 19 |
+
- Gryphe/Sonnet3.5-SlimOrcaDedupCleaned
|
| 20 |
+
- mpasila/Sonnet3.5-SlimOrcaDedupCleaned-4k-context
|
| 21 |
+
library_name: peft
|
| 22 |
---
|
| 23 |
+
This is the fully trained version (with fixed formatting!!).
|
| 24 |
+
|
| 25 |
+
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).
|
| 26 |
+
|
| 27 |
+
Prompt format is: ChatML
|
| 28 |
+
|
| 29 |
+
Merged model: [mpasila/Viking-SlimSonnet-v1-7B](https://huggingface.co/mpasila/Viking-SlimSonnet-v1-7B)
|
| 30 |
+
|
| 31 |
+
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.
|
| 32 |
|
| 33 |
# Uploaded model
|
| 34 |
|
|
|
|
| 38 |
|
| 39 |
This llama model was trained 2x faster with [Unsloth](https://github.com/unslothai/unsloth) and Huggingface's TRL library.
|
| 40 |
|
| 41 |
+
[<img src="https://raw.githubusercontent.com/unslothai/unsloth/main/images/unsloth%20made%20with%20love.png" width="200"/>](https://github.com/unslothai/unsloth)
|