Update README.md
Browse files
README.md
CHANGED
|
@@ -8,14 +8,23 @@ base_model:
|
|
| 8 |
- meta-llama/Llama-3.1-8B
|
| 9 |
---
|
| 10 |
|
| 11 |
-
This is the tuned lens version of LLama 3.1 8B (i.e. all the layers of the model have been trained to minimize the KL divergence with the last layer).
|
| 12 |
|
| 13 |
```bash
|
| 14 |
pip install tuned-lens
|
| 15 |
```
|
| 16 |
```bash
|
| 17 |
-
python -m tuned_lens train
|
|
|
|
|
|
|
|
|
|
|
|
|
| 18 |
```
|
| 19 |
```bash
|
| 20 |
-
python -m tuned_lens eval
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 21 |
```
|
|
|
|
| 8 |
- meta-llama/Llama-3.1-8B
|
| 9 |
---
|
| 10 |
|
| 11 |
+
This is the tuned lens version of LLama 3.1 8B (i.e. all the layers of the model have been trained to minimize the KL divergence with the last layer).
|
| 12 |
|
| 13 |
```bash
|
| 14 |
pip install tuned-lens
|
| 15 |
```
|
| 16 |
```bash
|
| 17 |
+
python -m tuned_lens train \
|
| 18 |
+
--model.name meta-llama/Llama-3.1-8B \
|
| 19 |
+
--data.name bookcorpus/bookcorpus \
|
| 20 |
+
--per_gpu_batch_size=1 \
|
| 21 |
+
--output my_lenses/meta-llama/Llama-3.1-8B
|
| 22 |
```
|
| 23 |
```bash
|
| 24 |
+
python -m tuned_lens eval \
|
| 25 |
+
--data.name bookcorpus/bookcorpus \
|
| 26 |
+
--model.name meta-llama/Llama-3.1-8B \
|
| 27 |
+
--tokens 16400000 \
|
| 28 |
+
--lens_name my_lenses/meta-llama/Llama-3.1-8B \
|
| 29 |
+
--output evaluation/meta-llama/Llama-3.1-8B
|
| 30 |
```
|