subashdvorak
commited on
Update shared_resources.py
Browse files- shared_resources.py +24 -24
shared_resources.py
CHANGED
@@ -1,24 +1,24 @@
|
|
1 |
-
# shared_resources.py
|
2 |
-
import torch
|
3 |
-
# from transformers import AutoTokenizer, AutoModelForCausalLM
|
4 |
-
from sentence_transformers import SentenceTransformer
|
5 |
-
from datasets import load_dataset
|
6 |
-
# from transformers import AutoProcessor, MusicgenForConditionalGeneration
|
7 |
-
import re
|
8 |
-
|
9 |
-
class SharedResources:
|
10 |
-
def __init__(self):
|
11 |
-
# Set the device
|
12 |
-
self.device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
|
13 |
-
|
14 |
-
|
15 |
-
# Load SentenceTransformer
|
16 |
-
self.sentence_transformer = SentenceTransformer("mixedbread-ai/mxbai-embed-large-v1")
|
17 |
-
|
18 |
-
# Load the dataset
|
19 |
-
self.dataset = load_dataset("subashdvorak/tiktok-story-
|
20 |
-
self.data = self.dataset["train"]
|
21 |
-
self.data = self.data.add_faiss_index("embeddings")
|
22 |
-
|
23 |
-
# Create a single instance of SharedResources
|
24 |
-
shared_resources = SharedResources()
|
|
|
1 |
+
# shared_resources.py
|
2 |
+
import torch
|
3 |
+
# from transformers import AutoTokenizer, AutoModelForCausalLM
|
4 |
+
from sentence_transformers import SentenceTransformer
|
5 |
+
from datasets import load_dataset
|
6 |
+
# from transformers import AutoProcessor, MusicgenForConditionalGeneration
|
7 |
+
import re
|
8 |
+
|
9 |
+
class SharedResources:
|
10 |
+
def __init__(self):
|
11 |
+
# Set the device
|
12 |
+
self.device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
|
13 |
+
|
14 |
+
|
15 |
+
# Load SentenceTransformer
|
16 |
+
self.sentence_transformer = SentenceTransformer("mixedbread-ai/mxbai-embed-large-v1")
|
17 |
+
|
18 |
+
# Load the dataset
|
19 |
+
self.dataset = load_dataset("subashdvorak/tiktok-story-data3", revision="embedded")
|
20 |
+
self.data = self.dataset["train"]
|
21 |
+
self.data = self.data.add_faiss_index("embeddings")
|
22 |
+
|
23 |
+
# Create a single instance of SharedResources
|
24 |
+
shared_resources = SharedResources()
|