Update repo names in the card
Browse files
README.md
CHANGED
|
@@ -27,10 +27,10 @@ Hugging Face Llama format.
|
|
| 27 |
|
| 28 |
```python
|
| 29 |
from transformers import AutoModelForCausalLM, AutoTokenizer
|
| 30 |
-
tok = AutoTokenizer.from_pretrained("nzhao721/
|
| 31 |
-
model = AutoModelForCausalLM.from_pretrained("nzhao721/
|
| 32 |
```
|
| 33 |
|
| 34 |
The training data and the evaluation holdout are in
|
| 35 |
-
[nzhao721/
|
| 36 |
trained once with one seed, so results should be read as directional.
|
|
|
|
| 27 |
|
| 28 |
```python
|
| 29 |
from transformers import AutoModelForCausalLM, AutoTokenizer
|
| 30 |
+
tok = AutoTokenizer.from_pretrained("nzhao721/worked-examples-models", subfolder="complete")
|
| 31 |
+
model = AutoModelForCausalLM.from_pretrained("nzhao721/worked-examples-models", subfolder="complete")
|
| 32 |
```
|
| 33 |
|
| 34 |
The training data and the evaluation holdout are in
|
| 35 |
+
[nzhao721/worked-examples-data](https://huggingface.co/datasets/nzhao721/worked-examples-data). Every arm was
|
| 36 |
trained once with one seed, so results should be read as directional.
|