vhr1007
commited on
Commit
·
88a7359
1
Parent(s):
298c89d
hf error
Browse files- Dockerfile +3 -0
- utils/embedding_utils.py +8 -1
Dockerfile
CHANGED
|
@@ -4,6 +4,9 @@ FROM python:3.10.9
|
|
| 4 |
# Set the working directory to /app
|
| 5 |
WORKDIR /app
|
| 6 |
|
|
|
|
|
|
|
|
|
|
| 7 |
# Create the cache directory
|
| 8 |
#RUN mkdir -p /app/.cache/huggingface
|
| 9 |
|
|
|
|
| 4 |
# Set the working directory to /app
|
| 5 |
WORKDIR /app
|
| 6 |
|
| 7 |
+
# Set a writable directory for Hugging Face cache
|
| 8 |
+
# ENV TRANSFORMERS_CACHE=/app/huggingface_cache
|
| 9 |
+
|
| 10 |
# Create the cache directory
|
| 11 |
#RUN mkdir -p /app/.cache/huggingface
|
| 12 |
|
utils/embedding_utils.py
CHANGED
|
@@ -17,6 +17,13 @@ import logging
|
|
| 17 |
# Load environment variables from .env file
|
| 18 |
#load_dotenv()
|
| 19 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 20 |
# Load Hugging Face token from environment variable
|
| 21 |
# huggingface_token = os.getenv('HUGGINGFACE_HUB_TOKEN')
|
| 22 |
|
|
@@ -27,7 +34,7 @@ import logging
|
|
| 27 |
# raise ValueError("Hugging Face token is not set. Please set the HUGGINGFACE_HUB_TOKEN environment variable.")
|
| 28 |
|
| 29 |
# Load model and tokenizer with authentication
|
| 30 |
-
|
| 31 |
|
| 32 |
model_name = 'nomic-ai/nomic-embed-text-v1.5'
|
| 33 |
model = SentenceTransformer('nomic-ai/nomic-embed-text-v1.5', trust_remote_code=True)
|
|
|
|
| 17 |
# Load environment variables from .env file
|
| 18 |
#load_dotenv()
|
| 19 |
|
| 20 |
+
os.environ["TRANSFORMERS_CACHE"] = "/app/huggingface_cache"
|
| 21 |
+
|
| 22 |
+
# Ensure the cache directory exists
|
| 23 |
+
cache_dir = os.environ["TRANSFORMERS_CACHE"]
|
| 24 |
+
if not os.path.exists(cache_dir):
|
| 25 |
+
os.makedirs(cache_dir)
|
| 26 |
+
|
| 27 |
# Load Hugging Face token from environment variable
|
| 28 |
# huggingface_token = os.getenv('HUGGINGFACE_HUB_TOKEN')
|
| 29 |
|
|
|
|
| 34 |
# raise ValueError("Hugging Face token is not set. Please set the HUGGINGFACE_HUB_TOKEN environment variable.")
|
| 35 |
|
| 36 |
# Load model and tokenizer with authentication
|
| 37 |
+
model_name = 'nvidia/NV-Embed-v1'
|
| 38 |
|
| 39 |
model_name = 'nomic-ai/nomic-embed-text-v1.5'
|
| 40 |
model = SentenceTransformer('nomic-ai/nomic-embed-text-v1.5', trust_remote_code=True)
|