Update README.md
Browse files
README.md
CHANGED
|
@@ -22,9 +22,16 @@ This model is a fine-tuned version of [google/gemma-2b](https://huggingface.co/g
|
|
| 22 |
It achieves the following results on the evaluation set:
|
| 23 |
- Loss: 2.0198
|
| 24 |
|
| 25 |
-
##
|
| 26 |
-
|
| 27 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 28 |
|
| 29 |
## Intended uses & limitations
|
| 30 |
|
|
|
|
| 22 |
It achieves the following results on the evaluation set:
|
| 23 |
- Loss: 2.0198
|
| 24 |
|
| 25 |
+
## How to Run Inference
|
| 26 |
+
|
| 27 |
+
Make sure you have git-lfs, and access to gemma-2b on huggingface.
|
| 28 |
+
```
|
| 29 |
+
pip install peft
|
| 30 |
+
git clone https://huggingface.co/apfurman/gemma-2b-dolly-qa
|
| 31 |
+
cd gemma-2b-dolly-qa/
|
| 32 |
+
python3 run.py cpu <YOUR-TOKEN-HERE> Prompt
|
| 33 |
+
```
|
| 34 |
+
replace "cpu" with "gpu" if you want to run on gpu.
|
| 35 |
|
| 36 |
## Intended uses & limitations
|
| 37 |
|