unify default reranking model with BAAI/bge-reranker-base (#623)

chensuyue · web-flow · commit 48d4e533964b · 2024-09-10T15:23:10.000+08:00
Signed-off-by: chensuyue &lt;suyue.chen@intel.com&gt;
Signed-off-by: ZePan110 &lt;ze.pan@intel.com&gt;
diff --git a/README.md b/README.md
@@ -55,7 +55,7 @@ The initially supported `Microservices` are described in the below table. More `
 		<tr>
 			<td rowspan="2"><a href="./comps/embeddings/README.md">Embedding</a></td>
             <td rowspan="2"><a href="https://www.langchain.com">LangChain</a>/<a href="https://www.llamaindex.ai">LlamaIndex</a></td>
-			<td rowspan="2"><a href="https://huggingface.co/BAAI/bge-large-en-v1.5">BAAI/bge-large-en-v1.5</a></td>
+			<td rowspan="2"><a href="https://huggingface.co/BAAI/bge-base-en-v1.5">BAAI/bge-base-en-v1.5</a></td>
 			<td><a href="https://github.com/huggingface/tei-gaudi">TEI-Gaudi</a></td>
 			<td>Gaudi2</td>
 			<td>Embedding on Gaudi2</td>
@@ -76,7 +76,7 @@ The initially supported `Microservices` are described in the below table. More `
 		<tr>
 			<td rowspan="2"><a href="./comps/reranks/README.md">Reranking</a></td>
             <td rowspan="2"><a href="https://www.langchain.com">LangChain</a>/<a href="https://www.llamaindex.ai">LlamaIndex</a></td>
-			<td ><a href="https://huggingface.co/BAAI/bge-reranker-large">BAAI/bge-reranker-large</a></td>
+			<td ><a href="https://huggingface.co/BAAI/bge-reranker-base">BAAI/bge-reranker-base</a></td>
 			<td><a href="https://github.com/huggingface/tei-gaudi">TEI-Gaudi</a></td>
 			<td>Gaudi2</td>
 			<td>Reranking on Gaudi2</td>
diff --git a/comps/dataprep/redis/README.md b/comps/dataprep/redis/README.md
@@ -49,7 +49,7 @@ First, you need to start a TEI service.
 
 ```bash
 your_port=6006
-model="BAAI/bge-large-en-v1.5"
+model="BAAI/bge-base-en-v1.5"
 revision="refs/pr/5"
 docker run -p $your_port:80 -v ./data:/data --name tei_server -e http_proxy=$http_proxy -e https_proxy=$https_proxy --pull always ghcr.io/huggingface/text-embeddings-inference:cpu-1.2 --model-id $model --revision $revision
 ```
diff --git a/comps/dataprep/redis/langchain/config.py b/comps/dataprep/redis/langchain/config.py
@@ -5,7 +5,7 @@
 
 # Embedding model
 
-EMBED_MODEL = os.getenv("EMBED_MODEL", "BAAI/bge-large-en-v1.5")
+EMBED_MODEL = os.getenv("EMBED_MODEL", "BAAI/bge-base-en-v1.5")
 
 # Redis Connection Information
 REDIS_HOST = os.getenv("REDIS_HOST", "localhost")
diff --git a/comps/embeddings/README.md b/comps/embeddings/README.md
@@ -43,7 +43,7 @@ First, you need to start a TEI service.
 
 ```bash
 your_port=8090
-model="BAAI/bge-large-en-v1.5"
+model="BAAI/bge-base-en-v1.5"
 docker run -p $your_port:80 -v ./data:/data --name tei_server -e http_proxy=$http_proxy -e https_proxy=$https_proxy --pull always ghcr.io/huggingface/text-embeddings-inference:cpu-1.5 --model-id $model
 ```
 
@@ -64,7 +64,7 @@ cd langchain
 # run with llama_index
 cd llama_index
 export TEI_EMBEDDING_ENDPOINT="http://localhost:$yourport"
-export TEI_EMBEDDING_MODEL_NAME="BAAI/bge-large-en-v1.5"
+export TEI_EMBEDDING_MODEL_NAME="BAAI/bge-base-en-v1.5"
 python embedding_tei.py
 ```
 
@@ -86,7 +86,7 @@ First, you need to start a TEI service.
 
 ```bash
 your_port=8090
-model="BAAI/bge-large-en-v1.5"
+model="BAAI/bge-base-en-v1.5"
 docker run -p $your_port:80 -v ./data:/data --name tei_server -e http_proxy=$http_proxy -e https_proxy=$https_proxy --pull always ghcr.io/huggingface/text-embeddings-inference:cpu-1.5 --model-id $model
 ```
 
@@ -103,7 +103,7 @@ Export the `TEI_EMBEDDING_ENDPOINT` for later usage:
 
 ```bash
 export TEI_EMBEDDING_ENDPOINT="http://localhost:$yourport"
-export TEI_EMBEDDING_MODEL_NAME="BAAI/bge-large-en-v1.5"
+export TEI_EMBEDDING_MODEL_NAME="BAAI/bge-base-en-v1.5"
 ```
 
 ### 2.2 Build Docker Image
diff --git a/comps/embeddings/langchain/local_embedding.py b/comps/embeddings/langchain/local_embedding.py
@@ -40,5 +40,5 @@ def embedding(input: TextDoc) -> EmbedDoc:
 
 
 if __name__ == "__main__":
-    embeddings = HuggingFaceEmbeddings(model_name="BAAI/bge-large-en-v1.5")
+    embeddings = HuggingFaceEmbeddings(model_name="BAAI/bge-base-en-v1.5")
     opea_microservices["opea_service@local_embedding"].start()
diff --git a/comps/embeddings/llama_index/embedding_tei.py b/comps/embeddings/llama_index/embedding_tei.py
@@ -31,7 +31,7 @@ def embedding(input: TextDoc) -> EmbedDoc:
 
 
 if __name__ == "__main__":
-    tei_embedding_model_name = os.getenv("TEI_EMBEDDING_MODEL_NAME", "BAAI/bge-large-en-v1.5")
+    tei_embedding_model_name = os.getenv("TEI_EMBEDDING_MODEL_NAME", "BAAI/bge-base-en-v1.5")
     tei_embedding_endpoint = os.getenv("TEI_EMBEDDING_ENDPOINT", "http://localhost:8090")
     embeddings = TextEmbeddingsInference(model_name=tei_embedding_model_name, base_url=tei_embedding_endpoint)
     logger.info("TEI Gaudi Embedding initialized.")
diff --git a/comps/embeddings/llama_index/local_embedding.py b/comps/embeddings/llama_index/local_embedding.py
@@ -31,5 +31,5 @@ def embedding(input: TextDoc) -> EmbedDoc:
 
 
 if __name__ == "__main__":
-    embeddings = HuggingFaceInferenceAPIEmbedding(model_name="BAAI/bge-large-en-v1.5")
+    embeddings = HuggingFaceInferenceAPIEmbedding(model_name="BAAI/bge-base-en-v1.5")
     opea_microservices["opea_service@local_embedding"].start()
diff --git a/comps/reranks/README.md b/comps/reranks/README.md
@@ -19,7 +19,7 @@ export HF_TOKEN=${your_hf_api_token}
 export LANGCHAIN_TRACING_V2=true
 export LANGCHAIN_API_KEY=${your_langchain_api_key}
 export LANGCHAIN_PROJECT="opea/reranks"
-export RERANK_MODEL_ID="BAAI/bge-reranker-large"
+export RERANK_MODEL_ID="BAAI/bge-reranker-base"
 revision=refs/pr/4
 volume=$PWD/data
 docker run -d -p 6060:80 -v $volume:/data -e http_proxy=$http_proxy -e https_proxy=$https_proxy --pull always ghcr.io/huggingface/text-embeddings-inference:cpu-1.2 --model-id $RERANK_MODEL_ID --revision $revision --hf-api-token $HF_TOKEN
diff --git a/comps/reranks/langchain-mosec/mosec-docker/Dockerfile b/comps/reranks/langchain-mosec/mosec-docker/Dockerfile
@@ -18,7 +18,7 @@ RUN pip3 install intel-extension-for-pytorch==2.2.0
 RUN pip3 install transformers sentence-transformers
 RUN pip3 install llmspec mosec
 
-RUN cd /home/user/ && export HF_ENDPOINT=https://hf-mirror.com && huggingface-cli download --resume-download BAAI/bge-reranker-large --local-dir /home/user/bge-reranker-large
+RUN cd /home/user/ && export HF_ENDPOINT=https://hf-mirror.com && huggingface-cli download --resume-download BAAI/bge-reranker-base --local-dir /home/user/bge-reranker-large
 USER user
 ENV EMB_MODEL="/home/user/bge-reranker-large/"
 
diff --git a/comps/reranks/tei/local_reranking.py b/comps/reranks/tei/local_reranking.py
@@ -41,5 +41,5 @@ def reranking(input: SearchedDoc) -> RerankedDoc:
 
 
 if __name__ == "__main__":
-    reranker_model = CrossEncoder(model_name="BAAI/bge-reranker-large", max_length=512)
+    reranker_model = CrossEncoder(model_name="BAAI/bge-reranker-base", max_length=512)
     opea_microservices["opea_service@local_reranking"].start()
diff --git a/tests/test_embeddings_langchain-mosec.sh b/tests/test_embeddings_langchain-mosec.sh
@@ -33,7 +33,7 @@ function build_docker_images() {
 
 function start_service() {
     mosec_endpoint=5001
-    model="BAAI/bge-large-en-v1.5"
+    model="BAAI/bge-base-en-v1.5"
     unset http_proxy
     docker run -d --name="test-comps-embedding-langchain-mosec-endpoint" -p $mosec_endpoint:8000  opea/embedding-langchain-mosec-endpoint:comps
     export MOSEC_EMBEDDING_ENDPOINT="http://${ip_address}:${mosec_endpoint}"
diff --git a/tests/test_embeddings_langchain.sh b/tests/test_embeddings_langchain.sh
@@ -21,10 +21,10 @@ function build_docker_images() {
 
 function start_service() {
     tei_endpoint=5001
-    model="BAAI/bge-large-en-v1.5"
+    model="BAAI/bge-base-en-v1.5"
     revision="refs/pr/5"
     unset http_proxy
-    docker run -d --name="test-comps-embedding-tei-endpoint" -p $tei_endpoint:80 -v ./data:/data --pull always ghcr.io/huggingface/text-embeddings-inference:cpu-1.2 --model-id $model --revision $revision
+    docker run -d --name="test-comps-embedding-tei-endpoint" -p $tei_endpoint:80 -v ./data:/data --pull always ghcr.io/huggingface/text-embeddings-inference:cpu-1.5 --model-id $model
     export TEI_EMBEDDING_ENDPOINT="http://${ip_address}:${tei_endpoint}"
     tei_service_port=5002
     docker run -d --name="test-comps-embedding-tei-server" -e http_proxy=$http_proxy -e https_proxy=$https_proxy -p ${tei_service_port}:6000 --ipc=host -e TEI_EMBEDDING_ENDPOINT=$TEI_EMBEDDING_ENDPOINT  opea/embedding-tei:comps
diff --git a/tests/test_embeddings_llama_index.sh b/tests/test_embeddings_llama_index.sh
@@ -22,9 +22,9 @@ function build_docker_images() {
 
 function start_service() {
     tei_endpoint=5001
-    model="BAAI/bge-large-en-v1.5"
+    model="BAAI/bge-base-en-v1.5"
     revision="refs/pr/5"
-    docker run -d --name="test-comps-embedding-tei-llama-index-endpoint" -p $tei_endpoint:80 -v ./data:/data -e http_proxy=$http_proxy -e https_proxy=$https_proxy --pull always ghcr.io/huggingface/text-embeddings-inference:cpu-1.2 --model-id $model --revision $revision
+    docker run -d --name="test-comps-embedding-tei-llama-index-endpoint" -p $tei_endpoint:80 -v ./data:/data -e http_proxy=$http_proxy -e https_proxy=$https_proxy --pull always ghcr.io/huggingface/text-embeddings-inference:cpu-1.5 --model-id $model
     export TEI_EMBEDDING_ENDPOINT="http://${ip_address}:${tei_endpoint}"
     tei_service_port=5010
     docker run -d --name="test-comps-embedding-tei-llama-index-server" -e http_proxy=$http_proxy -e https_proxy=$https_proxy -p ${tei_service_port}:6000 --ipc=host -e TEI_EMBEDDING_ENDPOINT=$TEI_EMBEDDING_ENDPOINT  opea/embedding-tei-llama-index:comps
diff --git a/tests/test_reranks_langchain-mosec.sh b/tests/test_reranks_langchain-mosec.sh
@@ -33,7 +33,7 @@ function build_docker_images() {
 
 function start_service() {
     mosec_endpoint=5006
-    model="BAAI/bge-reranker-large"
+    model="BAAI/bge-reranker-base"
     unset http_proxy
     docker run -d --name="test-comps-reranking-langchain-mosec-endpoint" -p $mosec_endpoint:8000  opea/reranking-langchain-mosec-endpoint:comps
     export MOSEC_RERANKING_ENDPOINT="http://${ip_address}:${mosec_endpoint}"
diff --git a/tests/test_reranks_tei.sh b/tests/test_reranks_tei.sh
@@ -21,10 +21,10 @@ function start_service() {
     tei_endpoint=5006
     # Remember to set HF_TOKEN before invoking this test!
     export HF_TOKEN=${HF_TOKEN}
-    model=BAAI/bge-reranker-large
+    model=BAAI/bge-reranker-base
     revision=refs/pr/4
     volume=$PWD/data
-    docker run -d --name="test-comps-reranking-tei-endpoint" -p $tei_endpoint:80 -v $volume:/data --pull always ghcr.io/huggingface/text-embeddings-inference:cpu-1.2 --model-id $model --revision $revision
+    docker run -d --name="test-comps-reranking-tei-endpoint" -p $tei_endpoint:80 -v $volume:/data --pull always ghcr.io/huggingface/text-embeddings-inference:cpu-1.5 --model-id $model
 
     export TEI_RERANKING_ENDPOINT="http://${ip_address}:${tei_endpoint}"
     tei_service_port=5007
