Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
|
@@ -2,17 +2,25 @@ import gradio as gr
|
|
| 2 |
from sentence_transformers import SentenceTransformer, util
|
| 3 |
import torch
|
| 4 |
import logging
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 5 |
|
| 6 |
# Set up logging
|
| 7 |
logging.basicConfig(level=logging.INFO)
|
| 8 |
logger = logging.getLogger(__name__)
|
| 9 |
|
| 10 |
-
# Load model
|
| 11 |
logger.info("Loading model...")
|
| 12 |
try:
|
| 13 |
-
model = SentenceTransformer(
|
| 14 |
-
|
| 15 |
-
|
|
|
|
|
|
|
|
|
|
| 16 |
except Exception as e:
|
| 17 |
logger.error(f"Failed to load model: {e}")
|
| 18 |
raise
|
|
@@ -24,6 +32,7 @@ try:
|
|
| 24 |
knowledge_text = f.read()
|
| 25 |
if not knowledge_text.strip():
|
| 26 |
raise ValueError("knowledge.txt is empty")
|
|
|
|
| 27 |
except FileNotFoundError:
|
| 28 |
logger.error("knowledge.txt not found")
|
| 29 |
raise
|
|
@@ -50,7 +59,8 @@ def split_text(text, chunk_size=200):
|
|
| 50 |
logger.info("Generating embeddings...")
|
| 51 |
try:
|
| 52 |
chunks = split_text(knowledge_text)
|
| 53 |
-
corpus_embeddings = model.encode(chunks, convert_to_tensor=True, batch_size=
|
|
|
|
| 54 |
except Exception as e:
|
| 55 |
logger.error(f"Error generating embeddings: {e}")
|
| 56 |
raise
|
|
|
|
| 2 |
from sentence_transformers import SentenceTransformer, util
|
| 3 |
import torch
|
| 4 |
import logging
|
| 5 |
+
import os
|
| 6 |
+
|
| 7 |
+
# Verify torch version
|
| 8 |
+
if torch.__version__ < '2.6.0':
|
| 9 |
+
raise ValueError(f"PyTorch version 2.6.0 or higher required. Current version: {torch.__version__}")
|
| 10 |
|
| 11 |
# Set up logging
|
| 12 |
logging.basicConfig(level=logging.INFO)
|
| 13 |
logger = logging.getLogger(__name__)
|
| 14 |
|
| 15 |
+
# Load model with explicit configuration
|
| 16 |
logger.info("Loading model...")
|
| 17 |
try:
|
| 18 |
+
model = SentenceTransformer(
|
| 19 |
+
"CAMeL-Lab/bert-base-arabic-camelbert-ca",
|
| 20 |
+
device="cpu",
|
| 21 |
+
cache_folder="./model_cache"
|
| 22 |
+
)
|
| 23 |
+
logger.info(f"Model loaded successfully. Using PyTorch version: {torch.__version__}")
|
| 24 |
except Exception as e:
|
| 25 |
logger.error(f"Failed to load model: {e}")
|
| 26 |
raise
|
|
|
|
| 32 |
knowledge_text = f.read()
|
| 33 |
if not knowledge_text.strip():
|
| 34 |
raise ValueError("knowledge.txt is empty")
|
| 35 |
+
logger.info("Knowledge file loaded successfully")
|
| 36 |
except FileNotFoundError:
|
| 37 |
logger.error("knowledge.txt not found")
|
| 38 |
raise
|
|
|
|
| 59 |
logger.info("Generating embeddings...")
|
| 60 |
try:
|
| 61 |
chunks = split_text(knowledge_text)
|
| 62 |
+
corpus_embeddings = model.encode(chunks, convert_to_tensor=True, batch_size=8) # Reduced batch size
|
| 63 |
+
logger.info(f"Generated embeddings for {len(chunks)} chunks")
|
| 64 |
except Exception as e:
|
| 65 |
logger.error(f"Error generating embeddings: {e}")
|
| 66 |
raise
|