Update README.md
Browse files
README.md
CHANGED
|
@@ -49,7 +49,7 @@ Checkpoints format: Hugging Face Transformers (Megatron-DeepSpeed format models
|
|
| 49 |
- torch>=2.0.0
|
| 50 |
- transformers>=4.34.0
|
| 51 |
- tokenizers>=0.14.0
|
| 52 |
-
|
| 53 |
|
| 54 |
## Usage
|
| 55 |
|
|
@@ -57,7 +57,7 @@ Checkpoints format: Hugging Face Transformers (Megatron-DeepSpeed format models
|
|
| 57 |
import torch
|
| 58 |
from transformers import AutoTokenizer, AutoModelForCausalLM
|
| 59 |
tokenizer = AutoTokenizer.from_pretrained("llm-jp/llm-jp-13b-v1.0")
|
| 60 |
-
model = AutoModelForCausalLM.from_pretrained("llm-jp/llm-jp-13b-v1.0", torch_dtype=torch.float16)
|
| 61 |
text = "自然言語処理とは何か"
|
| 62 |
tokenized_input = tokenizer.encode(text, add_special_tokens=False, return_tensors="pt").to(model.device)
|
| 63 |
with torch.no_grad():
|
|
|
|
| 49 |
- torch>=2.0.0
|
| 50 |
- transformers>=4.34.0
|
| 51 |
- tokenizers>=0.14.0
|
| 52 |
+
- accelerate==0.23.0
|
| 53 |
|
| 54 |
## Usage
|
| 55 |
|
|
|
|
| 57 |
import torch
|
| 58 |
from transformers import AutoTokenizer, AutoModelForCausalLM
|
| 59 |
tokenizer = AutoTokenizer.from_pretrained("llm-jp/llm-jp-13b-v1.0")
|
| 60 |
+
model = AutoModelForCausalLM.from_pretrained("llm-jp/llm-jp-13b-v1.0", device_map="auto", torch_dtype=torch.float16)
|
| 61 |
text = "自然言語処理とは何か"
|
| 62 |
tokenized_input = tokenizer.encode(text, add_special_tokens=False, return_tensors="pt").to(model.device)
|
| 63 |
with torch.no_grad():
|