Commit
·
298cb2c
1
Parent(s):
a684ae1
Update README.md
Browse files
README.md
CHANGED
|
@@ -17,4 +17,11 @@ Multi-line:
|
|
| 17 |
{user prompt}<|im_end|>
|
| 18 |
<|im_start|>assistant
|
| 19 |
{Assistant answer}<|im_end|>
|
| 20 |
-
```
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 17 |
{user prompt}<|im_end|>
|
| 18 |
<|im_start|>assistant
|
| 19 |
{Assistant answer}<|im_end|>
|
| 20 |
+
```
|
| 21 |
+
|
| 22 |
+
### Credits & Special Thanks
|
| 23 |
+
|
| 24 |
+
- Compute was generously sponsored by the eplf [Machine Learning and Optimization Laboratory](https://www.epfl.ch/labs/mlo/)
|
| 25 |
+
- The open-source [epfLLM/Megatron-LLM](https://github.com/epfLLM/Megatron-LLM) trainer was used for fine-tuning.
|
| 26 |
+
- [rombodawg](https://huggingface.co/rombodawg) curated and published [LosslessMegaCodeTrainingV2_1m_Evol_Uncensored](https://huggingface.co/datasets/rombodawg/LosslessMegaCodeTrainingV2_1m_Evol_Uncensored)
|
| 27 |
+
- [andreaskoepf](https://github.com/andreaskoepf/) prepared & orchestrated the training.
|