UniqAI/open-webui
1
0
Fork 0
mirror of https://github.com/open-webui/open-webui synced 2024-11-06 08:56:39 +00:00
Code Issues Actions 5 Packages Projects Releases Wiki Activity

fix: tiktoken encoding model issue

Browse Source
This commit is contained in:
Timothy J. Baek 2024-10-25 21:38:28 -07:00
parent adede5480d
commit 50dcad0f73
3 changed files with 11 additions and 11 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
Dockerfile
View File

@ -13,7 +13,7 @@ ARG USE_EMBEDDING_MODEL=sentence-transformers/all-MiniLM-L6-v2
ARG USE_RERANKING_MODEL=""
# Tiktoken encoding name; models to use can be found at https://huggingface.co/models?library=tiktoken
ARG USE_TIKTOKEN_ENCODING_NAME="cl100k_base"
ARG USE_TIKTOKEN_ENCODING_MODEL_NAME="cl100k_base"
ARG BUILD_HASH=dev-build
# Override at your own risk - non-root configurations are untested
@ -77,7 +77,7 @@ ENV RAG_EMBEDDING_MODEL="$USE_EMBEDDING_MODEL_DOCKER" \
SENTENCE_TRANSFORMERS_HOME="/app/backend/data/cache/embedding/models"
## Tiktoken model settings ##
ENV TIKTOKEN_ENCODING_NAME="$USE_TIKTOKEN_ENCODING_NAME" \
ENV TIKTOKEN_ENCODING_MODEL_NAME="$USE_TIKTOKEN_ENCODING_MODEL_NAME" \
TIKTOKEN_CACHE_DIR="/app/backend/data/cache/tiktoken"
## Hugging Face download cache ##
@ -139,13 +139,13 @@ RUN pip3 install uv && \
uv pip install --system -r requirements.txt --no-cache-dir && \
python -c "import os; from sentence_transformers import SentenceTransformer; SentenceTransformer(os.environ['RAG_EMBEDDING_MODEL'], device='cpu')" && \
python -c "import os; from faster_whisper import WhisperModel; WhisperModel(os.environ['WHISPER_MODEL'], device='cpu', compute_type='int8', download_root=os.environ['WHISPER_MODEL_DIR'])"; \
python -c "import os; import tiktoken; tiktoken.get_encoding(os.environ['TIKTOKEN_ENCODING_NAME'])"; \
python -c "import os; import tiktoken; tiktoken.get_encoding(os.environ['TIKTOKEN_ENCODING_MODEL_NAME'])"; \
else \
pip3 install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cpu --no-cache-dir && \
uv pip install --system -r requirements.txt --no-cache-dir && \
python -c "import os; from sentence_transformers import SentenceTransformer; SentenceTransformer(os.environ['RAG_EMBEDDING_MODEL'], device='cpu')" && \
python -c "import os; from faster_whisper import WhisperModel; WhisperModel(os.environ['WHISPER_MODEL'], device='cpu', compute_type='int8', download_root=os.environ['WHISPER_MODEL_DIR'])"; \
python -c "import os; import tiktoken; tiktoken.get_encoding(os.environ['TIKTOKEN_ENCODING_NAME'])"; \
python -c "import os; import tiktoken; tiktoken.get_encoding(os.environ['TIKTOKEN_ENCODING_MODEL_NAME'])"; \
fi; \
chown -R $UID:$GID /app/backend/data/

6
backend/open_webui/apps/retrieval/main.py
View File

@ -50,7 +50,7 @@ from open_webui.apps.retrieval.utils import (
from open_webui.apps.webui.models.files import Files
from open_webui.config import (
BRAVE_SEARCH_API_KEY,
TIKTOKEN_ENCODING_NAME,
TIKTOKEN_ENCODING_MODEL_NAME,
RAG_TEXT_SPLITTER,
CHUNK_OVERLAP,
CHUNK_SIZE,
@ -135,7 +135,7 @@ app.state.config.CONTENT_EXTRACTION_ENGINE = CONTENT_EXTRACTION_ENGINE
app.state.config.TIKA_SERVER_URL = TIKA_SERVER_URL
app.state.config.TEXT_SPLITTER = RAG_TEXT_SPLITTER
app.state.config.TIKTOKEN_ENCODING_NAME = TIKTOKEN_ENCODING_NAME
app.state.config.TIKTOKEN_ENCODING_MODEL_NAME = TIKTOKEN_ENCODING_MODEL_NAME
app.state.config.CHUNK_SIZE = CHUNK_SIZE
app.state.config.CHUNK_OVERLAP = CHUNK_OVERLAP
@ -667,7 +667,7 @@ def save_docs_to_vector_db(
)
elif app.state.config.TEXT_SPLITTER == "token":
text_splitter = TokenTextSplitter(
encoding_name=app.state.config.TIKTOKEN_ENCODING_NAME,
model_name=app.state.config.TIKTOKEN_ENCODING_MODEL_NAME,
chunk_size=app.state.config.CHUNK_SIZE,
chunk_overlap=app.state.config.CHUNK_OVERLAP,
add_start_index=True,

8
backend/open_webui/config.py
View File

@ -1074,10 +1074,10 @@ RAG_TEXT_SPLITTER = PersistentConfig(
TIKTOKEN_CACHE_DIR = os.environ.get("TIKTOKEN_CACHE_DIR", f"{CACHE_DIR}/tiktoken")
TIKTOKEN_ENCODING_NAME = PersistentConfig(
"TIKTOKEN_ENCODING_NAME",
"rag.tiktoken_encoding_name",
os.environ.get("TIKTOKEN_ENCODING_NAME", "cl100k_base"),
TIKTOKEN_ENCODING_MODEL_NAME = PersistentConfig(
"TIKTOKEN_ENCODING_MODEL_NAME",
"rag.tiktoken_encoding_model_name",
os.environ.get("TIKTOKEN_ENCODING_MODEL_NAME", "cl100k_base"),
)