Update README.md
Browse files
README.md
CHANGED
|
@@ -13,6 +13,7 @@ Model lineage:
|
|
| 13 |
https://huggingface.co/timdettmers/guanaco-65b -> https://huggingface.co/Mikael110/llama-2-70b-guanaco-qlora -> https://huggingface.co/TheBloke/llama-2-70b-Guanaco-QLoRA-fp16
|
| 14 |
|
| 15 |
---
|
|
|
|
| 16 |
Training Data and Formatting:
|
| 17 |
|
| 18 |
Training data were originally garnered from: https://huggingface.co/datasets/BI55/MedText
|
|
@@ -20,5 +21,10 @@ Training data were originally garnered from: https://huggingface.co/datasets/BI5
|
|
| 20 |
These training data were then formatted for use with the "Raw text file" training option in the Oobabooga text-generation-webui:
|
| 21 |
(https://github.com/oobabooga/text-generation-webui)
|
| 22 |
|
|
|
|
|
|
|
| 23 |
---
|
| 24 |
|
|
|
|
|
|
|
|
|
|
|
|
| 13 |
https://huggingface.co/timdettmers/guanaco-65b -> https://huggingface.co/Mikael110/llama-2-70b-guanaco-qlora -> https://huggingface.co/TheBloke/llama-2-70b-Guanaco-QLoRA-fp16
|
| 14 |
|
| 15 |
---
|
| 16 |
+
|
| 17 |
Training Data and Formatting:
|
| 18 |
|
| 19 |
Training data were originally garnered from: https://huggingface.co/datasets/BI55/MedText
|
|
|
|
| 21 |
These training data were then formatted for use with the "Raw text file" training option in the Oobabooga text-generation-webui:
|
| 22 |
(https://github.com/oobabooga/text-generation-webui)
|
| 23 |
|
| 24 |
+
Training parameters are in the training_parameters.json file and there is a screenshot of the UI with the correct settings.
|
| 25 |
+
|
| 26 |
---
|
| 27 |
|
| 28 |
+
Current/Future Work:
|
| 29 |
+
|
| 30 |
+
1. Finish training with "Structed Dataset" I have a .json file with a structured dataset for the Guanaco model, but it takes signifcantly longer to process in the Oobabooga webui.
|