Merge branch 'main' into base-image

.github/CODEOWNERS

@@ -3,7 +3,7 @@
 
 * [email protected] [email protected] [email protected]
 /.github/ [email protected] [email protected]
-/comps/3rd_parties/ [email protected] [email protected] [email protected]
+/comps/third_parties/ [email protected] [email protected] [email protected]
 /comps/agent/ [email protected] [email protected]
 /comps/animation/ [email protected] [email protected]
 /comps/asr/ [email protected] [email protected]

.github/workflows/docker/compose/animation-compose.yaml

@@ -9,9 +9,9 @@ services:
     image: ${REGISTRY:-opea}/animation:${TAG:-latest}
   wav2lip:
     build:
-      dockerfile: comps/animation/src/integration/dependency/Dockerfile
+      dockerfile: comps/third_parties/wav2lip/src/Dockerfile
     image: ${REGISTRY:-opea}/wav2lip:${TAG:-latest}
   wav2lip-gaudi:
     build:
-      dockerfile: comps/animation/src/integration/dependency/Dockerfile.intel_hpu
+      dockerfile: comps/third_parties/wav2lip/src/Dockerfile.intel_hpu
     image: ${REGISTRY:-opea}/wav2lip-gaudi:${TAG:-latest}

.github/workflows/docker/compose/embeddings-compose.yaml

@@ -7,15 +7,3 @@ services:
     build:
       dockerfile: comps/embeddings/src/Dockerfile
     image: ${REGISTRY:-opea}/embedding:${TAG:-latest}
-  embedding-multimodal-clip:
-    build:
-      dockerfile: comps/embeddings/src/integrations/dependency/clip/Dockerfile
-    image: ${REGISTRY:-opea}/embedding-multimodal-clip:${TAG:-latest}
-  embedding-multimodal-bridgetower:
-    build:
-      dockerfile: comps/embeddings/src/integrations/dependency/bridgetower/Dockerfile
-    image: ${REGISTRY:-opea}/embedding-multimodal-bridgetower:${TAG:-latest}
-  embedding-multimodal-bridgetower-gaudi:
-    build:
-      dockerfile: comps/embeddings/src/integrations/dependency/bridgetower/Dockerfile.intel_hpu
-    image: ${REGISTRY:-opea}/embedding-multimodal-bridgetower-gaudi:${TAG:-latest}

.github/workflows/docker/compose/llms-compose.yaml

@@ -15,10 +15,10 @@ services:
     build:
       dockerfile: comps/llms/summarization/tgi/langchain/Dockerfile
     image: ${REGISTRY:-opea}/llm-docsum-tgi:${TAG:-latest}
-  llm-faqgen-tgi:
+  llm-faqgen:
     build:
-      dockerfile: comps/llms/faq-generation/tgi/langchain/Dockerfile
-    image: ${REGISTRY:-opea}/llm-faqgen-tgi:${TAG:-latest}
+      dockerfile: comps/llms/src/faq-generation/Dockerfile
+    image: ${REGISTRY:-opea}/llm-faqgen:${TAG:-latest}
   llm-native:
     build:
       dockerfile: comps/llms/text-generation/native/langchain/Dockerfile
@@ -54,7 +54,3 @@ services:
     build:
       dockerfile: comps/llms/summarization/vllm/langchain/Dockerfile
     image: ${REGISTRY:-opea}/llm-docsum-vllm:${TAG:-latest}
-  llm-faqgen-vllm:
-    build:
-      dockerfile: comps/llms/faq-generation/vllm/langchain/Dockerfile
-    image: ${REGISTRY:-opea}/llm-faqgen-vllm:${TAG:-latest}

.github/workflows/docker/compose/third_parties-compose.yaml

@@ -0,0 +1,28 @@
+# Copyright (C) 2024 Intel Corporation
+# SPDX-License-Identifier: Apache-2.0
+
+services:
+  nginx:
+    build:
+      dockerfile: comps/third_parties/nginx/src/Dockerfile
+    image: ${REGISTRY:-opea}/nginx:${TAG:-latest}
+  embedding-multimodal-clip:
+    build:
+      dockerfile: comps/third_parties/clip/src/Dockerfile
+    image: ${REGISTRY:-opea}/embedding-multimodal-clip:${TAG:-latest}
+  embedding-multimodal-bridgetower:
+    build:
+      dockerfile: comps/third_parties/bridgetower/src/Dockerfile
+    image: ${REGISTRY:-opea}/embedding-multimodal-bridgetower:${TAG:-latest}
+  embedding-multimodal-bridgetower-gaudi:
+    build:
+      dockerfile: comps/third_parties/bridgetower/src/Dockerfile.intel_hpu
+    image: ${REGISTRY:-opea}/embedding-multimodal-bridgetower-gaudi:${TAG:-latest}
+  wav2lip:
+    build:
+      dockerfile: comps/third_parties/wav2lip/src/Dockerfile
+    image: ${REGISTRY:-opea}/wav2lip:${TAG:-latest}
+  wav2lip-gaudi:
+    build:
+      dockerfile: comps/third_parties/wav2lip/src/Dockerfile.intel_hpu
+    image: ${REGISTRY:-opea}/wav2lip-gaudi:${TAG:-latest}

.github/workflows/scripts/get_test_matrix.sh

@@ -42,7 +42,7 @@ function find_test_1() {
                     changed_integrations=$(printf '%s\n' "${changed_files[@]}"| grep ${service_path} | grep -E 'integrations' | cut -d'/' -f$((n+2)) | cut -d'.' -f1 | sort -u)  || true
                     for integration in ${changed_integrations}; do
                         # Accurate matching test scripts
-                        # find_test=$(find ./tests -type f \( -name test_${service_name}_${integration}.sh -o -name test_${service_name}_${integration}_on_*.sh \)) || true
+                        # find_test=$(find ./tests -type f \( -name test_${service_name}_${integrations}.sh -o -name test_${service_name}_${integrations}_on_*.sh \)) || true
                         # Fuzzy matching test scripts, for example, llms/src/text-generation/integrations/opea.py match several tests.
                         find_test=$(find ./tests -type f -name test_${service_name}_${integration}*.sh) || true
                         if [ "$find_test" ]; then
@@ -110,7 +110,7 @@ function find_test_2() {
 
 function main() {
 
-    changed_files=$(printf '%s\n' "${changed_files_full[@]}" | grep 'comps/' | grep -vE '\.md|comps/cores|comps/3rd_parties|deployment|\.yaml') || true
+    changed_files=$(printf '%s\n' "${changed_files_full[@]}" | grep 'comps/' | grep -vE '\.md|comps/cores|comps/third_parties|deployment|\.yaml') || true
     echo "===========start find_test_1============"
     echo "changed_files=${changed_files}"
     find_test_1 "comps" 2 false

comps/animation/src/README.md

@@ -16,13 +16,13 @@ cd GenAIComps
 - Xeon CPU
 
 ```bash
-docker build -t opea/wav2lip:latest -f comps/animation/src/integration/dependency/Dockerfile .
+docker build -t opea/wav2lip:latest -f comps/third_parties/wav2lip/src/Dockerfile .
 ```
 
 - Gaudi2 HPU
 
 ```bash
-docker build -t opea/wav2lip-gaudi:latest -f comps/animation/src/integration/dependency/Dockerfile.intel_hpu .
+docker build -t opea/wav2lip-gaudi:latest -f comps/third_parties/wav2lip/src/Dockerfile.intel_hpu .
 ```
 
 ### 1.1.2 Animation server image
@@ -101,7 +101,7 @@ Once microservice starts, user can use below script to validate the running micr
 
 ```bash
 cd GenAIComps
-python3 comps/animation/src/integration/dependency/check_wav2lip_server.py
+python3 comps/third_parties/wav2lip/src/check_wav2lip_server.py
 ```
 
 ## 3.2 Validate Animation service
@@ -116,7 +116,7 @@ or
 
 ```bash
 cd GenAIComps
-python3 comps/animation/src/integration/dependency/check_animation_server.py
+python3 comps/third_parties/wav2lip/src/check_animation_server.py
 ```
 
 The expected output will be a message similar to the following:

comps/animation/src/integration/opea.py → comps/animation/src/integrations/wav2lip.py

@@ -22,7 +22,7 @@ def __init__(self, name: str, description: str, config: dict = None):
         if not health_status:
             logger.error("OpeaAnimation health check failed.")
 
-    def invoke(self, input: str):
+    async def invoke(self, input: str):
         """Invokes the animation service to generate embeddings for the animation input.
 
         Args:

comps/animation/src/opea_animation_microservice.py

@@ -10,7 +10,7 @@
 
 # GenAIComps
 from comps import CustomLogger, OpeaComponentLoader
-from comps.animation.src.integration.opea import OpeaAnimation
+from comps.animation.src.integrations.wav2lip import OpeaAnimation
 
 logger = CustomLogger("opea_animation")
 logflag = os.getenv("LOGFLAG", False)
@@ -43,10 +43,10 @@
     output_datatype=VideoPath,
 )
 @register_statistics(names=["opea_service@animation"])
-def animate(audio: Base64ByteStrDoc):
+async def animate(audio: Base64ByteStrDoc):
     start = time.time()
 
-    outfile = loader.invoke(audio.byte_str)
+    outfile = await loader.invoke(audio.byte_str)
     if logflag:
         logger.info(f"Video generated successfully, check {outfile} for the result.")
 

comps/animation/src/requirements.txt

@@ -1,29 +1,29 @@
-aiohttp==3.10.11
-basicsr==1.4.2
+aiohttp
+basicsr
 docarray[full]
-facexlib==0.3.0
-fastapi==0.115.2
-ffmpeg-python==0.2.0
-gradio==4.38.1
-gradio_client==1.1.0
-grpcio==1.64.1
-huggingface-hub==0.24.0
-librosa==0.10.2
-lmdb==1.5.1
-numba==0.60.0
-numpy==1.23.5
-opencv-python==4.10.0.84
-opentelemetry-api==1.27.0
-opentelemetry-exporter-otlp==1.27.0
-opentelemetry-sdk==1.27.0
-pandas==2.2.3
-prometheus-fastapi-instrumentator==7.0.0
-pyarrow==17.0.0
-pydantic==2.9.2
-pyyaml==6.0
-scipy==1.14.0
-shortuuid==1.0.13
-tavily-python==0.5.0
-tb-nightly==2.18.0a20240721
-tqdm==4.66.4
-yapf==0.40.2
+facexlib
+fastapi
+ffmpeg-python
+gradio
+gradio_client
+grpcio
+huggingface-hub
+librosa
+lmdb
+numba
+numpy
+opencv-python
+opentelemetry-api
+opentelemetry-exporter-otlp
+opentelemetry-sdk
+pandas
+prometheus-fastapi-instrumentator
+pyarrow
+pydantic
+pyyaml
+scipy
+shortuuid
+tavily-python
+tb-nightly
+tqdm
+yapf

comps/cores/proto/api_protocol.py

@@ -636,10 +636,10 @@ class FineTuningJobWandbIntegration(BaseModel):
 
 class FineTuningJobWandbIntegrationObject(BaseModel):
     type: Literal["wandb"]
-    """The type of the integration being enabled for the fine-tuning job."""
+    """The type of the integrations being enabled for the fine-tuning job."""
 
     wandb: FineTuningJobWandbIntegration
-    """The settings for your integration with Weights and Biases.
+    """The settings for your integrations with Weights and Biases.
 
     This payload specifies the project that metrics will be sent to. Optionally, you
     can set an explicit display name for your run, add tags to your run, and set a

comps/dataprep/multimodal/redis/langchain/prepare_videodoc_redis.py

@@ -32,7 +32,7 @@
 from PIL import Image
 
 from comps import opea_microservices, register_microservice
-from comps.embeddings.src.integrations.dependency.bridgetower.bridgetower_embedding import BridgeTowerEmbedding
+from comps.third_parties.bridgetower.src.bridgetower_embedding import BridgeTowerEmbedding
 
 device = "cpu"
 upload_folder = "./uploaded_files/"

comps/embeddings/src/opea_embedding_microservice.py

@@ -4,8 +4,8 @@
 import os
 import time
 
-from integrations.opea_tei_embedding import OpeaTEIEmbedding
-from integrations.predictionguard_embedding import PredictionguardEmbedding
+from integrations.predictionguard import PredictionguardEmbedding
+from integrations.tei import OpeaTEIEmbedding
 
 from comps import (
     CustomLogger,

comps/embeddings/src/opea_multimodal_embedding_microservice.py

@@ -4,7 +4,7 @@
 import os
 import time
 
-from integrations.opea_multimodal_embedding_bridgetower import OpeaMultimodalEmbeddingBrigeTower
+from integrations.multimodal_bridgetower import OpeaMultimodalEmbeddingBrigeTower
 
 from comps import (
     CustomLogger,

comps/finetuning/src/opea_finetuning_microservice.py

@@ -7,7 +7,7 @@
 from comps import CustomLogger, opea_microservices, register_microservice
 from comps.cores.proto.api_protocol import FineTuningJobIDRequest, UploadFileRequest
 from comps.finetuning.src.integrations.finetune_config import FineTuningParams
-from comps.finetuning.src.integrations.opea import OpeaFinetuning, upload_file
+from comps.finetuning.src.integrations.native import OpeaFinetuning, upload_file
 from comps.finetuning.src.opea_finetuning_loader import OpeaFinetuningLoader
 
 logger = CustomLogger("opea_finetuning_microservice")

comps/image2image/src/integration/opea_image2image_native.py → comps/image2image/src/integrations/native.py

@@ -2,6 +2,7 @@
 # SPDX-License-Identifier: Apache-2.0
 import base64
 import os
+import tempfile
 import threading
 
 from comps import CustomLogger, OpeaComponent, OpeaComponentRegistry, SDImg2ImgInputs, ServiceType
@@ -99,7 +100,7 @@ def __init__(
         if not health_status:
             logger.error("OpeaImageToImage health check failed.")
 
-    def invoke(self, input: SDImg2ImgInputs):
+    async def invoke(self, input: SDImg2ImgInputs):
         """Invokes the ImageToImage service to generate Images for the provided input.
 
         Args:
@@ -113,16 +114,15 @@ def invoke(self, input: SDImg2ImgInputs):
         images = pipe(
             image=image, prompt=prompt, generator=generator, num_images_per_prompt=num_images_per_prompt
         ).images
-        image_path = os.path.join(os.getcwd(), prompt.strip().replace(" ", "_").replace("/", ""))
-        os.makedirs(image_path, exist_ok=True)
         results = []
-        for i, image in enumerate(images):
-            save_path = os.path.join(image_path, f"image_{i + 1}.png")
-            image.save(save_path)
-            with open(save_path, "rb") as f:
-                bytes = f.read()
-            b64_str = base64.b64encode(bytes).decode()
-            results.append(b64_str)
+        with tempfile.TemporaryDirectory() as image_path:
+            for i, image in enumerate(images):
+                save_path = os.path.join(image_path, f"image_{i + 1}.png")
+                image.save(save_path)
+                with open(save_path, "rb") as f:
+                    bytes = f.read()
+                b64_str = base64.b64encode(bytes).decode()
+                results.append(b64_str)
 
         return results
 

comps/image2image/src/opea_image2image_microservice.py

@@ -17,7 +17,7 @@
     register_statistics,
     statistics_dict,
 )
-from comps.image2image.src.integration.opea_image2image_native import OpeaImageToImage
+from comps.image2image.src.integrations.native import OpeaImageToImage
 
 args = None
 
@@ -36,9 +36,9 @@
     output_datatype=SDOutputs,
 )
 @register_statistics(names=["opea_service@image2image"])
-def image2image(input: SDImg2ImgInputs):
+async def image2image(input: SDImg2ImgInputs):
     start = time.time()
-    results = component_loader.invoke(input)
+    results = await component_loader.invoke(input)
     statistics_dict["opea_service@image2image"].append_latency(time.time() - start, None)
     return SDOutputs(images=results)
 

comps/image2video/src/opea_image2video_microservice.py

@@ -5,8 +5,6 @@
 import os
 import time
 
-from integrations.opea import OpeaImage2video
-
 from comps import (
     CustomLogger,
     ImagesPath,
@@ -18,6 +16,7 @@
     register_statistics,
     statistics_dict,
 )
+from comps.image2video.src.integrations.native import OpeaImage2video
 
 logger = CustomLogger("opea_image2video_microservice")