Update app.py
Browse files
app.py
CHANGED
@@ -13,7 +13,6 @@ from numpy import ndarray
|
|
13 |
from qdrant_client import QdrantClient, models
|
14 |
from llama_cpp import Llama, GGML_TYPE_I8
|
15 |
from optimum_encoder import OptimumEncoder
|
16 |
-
from fastembed_encoder import FastEmbedEncoder
|
17 |
from unstructured.partition.auto import partition
|
18 |
from statistical_chunker import StatisticalChunker
|
19 |
from fastembed import SparseEmbedding, SparseTextEmbedding
|
@@ -138,15 +137,9 @@ def load_models_and_documents():
|
|
138 |
type_v=GGML_TYPE_I8
|
139 |
)
|
140 |
|
141 |
-
|
142 |
-
|
143 |
-
|
144 |
-
#)
|
145 |
-
|
146 |
-
dense_model = FastEmbedEncoder(
|
147 |
-
name='mixedbread-ai/mxbai-embed-large-v1',
|
148 |
-
cache_dir=os.getenv('HF_HOME'),
|
149 |
-
providers=['CPUExecutionProvider']
|
150 |
)
|
151 |
|
152 |
sparse_model = SparseTextEmbedding(
|
|
|
13 |
from qdrant_client import QdrantClient, models
|
14 |
from llama_cpp import Llama, GGML_TYPE_I8
|
15 |
from optimum_encoder import OptimumEncoder
|
|
|
16 |
from unstructured.partition.auto import partition
|
17 |
from statistical_chunker import StatisticalChunker
|
18 |
from fastembed import SparseEmbedding, SparseTextEmbedding
|
|
|
137 |
type_v=GGML_TYPE_I8
|
138 |
)
|
139 |
|
140 |
+
dense_model = OptimumEncoder(
|
141 |
+
device="cuda",
|
142 |
+
cache_dir=os.getenv('HF_HOME')
|
|
|
|
|
|
|
|
|
|
|
|
|
143 |
)
|
144 |
|
145 |
sparse_model = SparseTextEmbedding(
|