Update README.md
Browse files
README.md
CHANGED
|
@@ -27,7 +27,7 @@ tokenizer = AutoTokenizer.from_pretrained("Cedille/de-anna")
|
|
| 27 |
model = AutoModelForCausalLM.from_pretrained("Cedille/de-anna")
|
| 28 |
```
|
| 29 |
### Lower memory usage
|
| 30 |
-
Loading a model with Huggingface requires two copies of the weights, so 48+ GB of RAM for [
|
| 31 |
The first trick would be to load the model with the specific argument below to load only one copy of the weights.
|
| 32 |
```
|
| 33 |
from transformers import AutoTokenizer, AutoModelForCausalLM
|
|
|
|
| 27 |
model = AutoModelForCausalLM.from_pretrained("Cedille/de-anna")
|
| 28 |
```
|
| 29 |
### Lower memory usage
|
| 30 |
+
Loading a model with Huggingface requires two copies of the weights, so 48+ GB of RAM for [GPT-J models](https://huggingface.co/docs/transformers/v4.15.0/model_doc/gptj) in float32 precision.
|
| 31 |
The first trick would be to load the model with the specific argument below to load only one copy of the weights.
|
| 32 |
```
|
| 33 |
from transformers import AutoTokenizer, AutoModelForCausalLM
|