handler
Browse files- handler.py +3 -2
handler.py
CHANGED
@@ -1,5 +1,6 @@
|
|
1 |
from transformers import AutoModelForCausalLM, AutoTokenizer
|
2 |
-
from peft import PeftConfig
|
|
|
3 |
import torch.cuda
|
4 |
from typing import Any, Dict
|
5 |
device = "cuda" if torch.cuda.is_available() else "cpu"
|
@@ -7,7 +8,7 @@ device = "cuda" if torch.cuda.is_available() else "cpu"
|
|
7 |
class EndpointHandler():
|
8 |
def __init__(self, path=""):
|
9 |
config = PeftConfig.from_pretrained(path)
|
10 |
-
model = AutoModelForCausalLM.from_pretrained(config.base_model_name_or_path
|
11 |
self.tokenizer = AutoTokenizer.from_pretrained(config.base_model_name_or_path)
|
12 |
# Load the Lora model
|
13 |
self.model = PeftModel.from_pretrained(model, path)
|
|
|
1 |
from transformers import AutoModelForCausalLM, AutoTokenizer
|
2 |
+
from peft import PeftConfig
|
3 |
+
from peft import PeftModel
|
4 |
import torch.cuda
|
5 |
from typing import Any, Dict
|
6 |
device = "cuda" if torch.cuda.is_available() else "cpu"
|
|
|
8 |
class EndpointHandler():
|
9 |
def __init__(self, path=""):
|
10 |
config = PeftConfig.from_pretrained(path)
|
11 |
+
model = AutoModelForCausalLM.from_pretrained(config.base_model_name_or_path)
|
12 |
self.tokenizer = AutoTokenizer.from_pretrained(config.base_model_name_or_path)
|
13 |
# Load the Lora model
|
14 |
self.model = PeftModel.from_pretrained(model, path)
|