Update README.md
Browse files
README.md
CHANGED
@@ -1,3 +1,27 @@
|
|
1 |
---
|
2 |
license: mit
|
3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
---
|
2 |
license: mit
|
3 |
---
|
4 |
+
|
5 |
+
# dolly-v2-7b: sharded checkpoint
|
6 |
+
|
7 |
+
This is a sharded checkpoint (with ~2GB shards) of the `databricks/dolly-v2-7b` model. Refer to the [original model](https://huggingface.co/databricks/dolly-v2-7b) for all details.
|
8 |
+
|
9 |
+
## Basic Usage
|
10 |
+
|
11 |
+
|
12 |
+
install `transformers`, `accelerate`, and `bitsandbytes`.
|
13 |
+
|
14 |
+
```bash
|
15 |
+
pip install -U -q transformers bitsandbytes accelerate
|
16 |
+
```
|
17 |
+
|
18 |
+
Load the model in 8bit, then [run inference](https://huggingface.co/docs/transformers/generation_strategies#contrastive-search):
|
19 |
+
|
20 |
+
```python
|
21 |
+
from transformers import AutoTokenizer, AutoModelForCausalLM
|
22 |
+
|
23 |
+
model_name = "ethzanalytics/dolly-v2-7b-sharded"
|
24 |
+
tokenizer = AutoTokenizer.from_pretrained(model_name)
|
25 |
+
|
26 |
+
model = AutoModelForCausalLM.from_pretrained(model_name, load_in_8bit=True)
|
27 |
+
```
|