updates

langhchain_generate_components.py

@@ -38,29 +38,42 @@ os.environ["OPENAI_API_KEY"] = os.getenv("OPENAI_API_KEY")
 
 
 # The scheme for creating experiments
-experiment_schema = [
-    ResponseSchema(name="Material", description="list of materials need to perfrom the experiments please be specific", type="list"),
-]
+# experiment_schema = [
+#     ResponseSchema(name="Material", description="list of materials need to perfrom the experiments please be specific", type="list"),
+# ]
 
 
-maker_schema = [
+response_schemas = [
     ResponseSchema(name="Material", description="The base components needed to create this items from scratch DIY This item must be exact and not an estimation", type="list"),
+    ResponseSchema(name="Feild Of Study", description="List the field of study this can be used for", type="list"),
 ]
 
-experiment_output_parser = StructuredOutputParser.from_response_schemas(experiment_schema)
-maker_output_parser = StructuredOutputParser.from_response_schemas(maker_schema)
-
+output_parser = StructuredOutputParser.from_response_schemas(response_schemas)
 memory = ConversationBufferMemory(memory_key="chat_history", return_messages=True)
+format_instructions = output_parser.get_format_instructions()
 
-format_instructions = experiment_output_parser.get_format_instructions()
 
+# experiment_output_parser = StructuredOutputParser.from_response_schemas(experiment_schema)
+# maker_output_parser = StructuredOutputParser.from_response_schemas(maker_schema)
 
-experiment_prompt = PromptTemplate(
-    template="You must generate well detailed science experiments.\n{format_instructions}\n{question}\n{context}",
-    input_variables=["question"],
-    partial_variables={"format_instructions": format_instructions},
-    memory = memory
-)
+memory = ConversationBufferMemory(
+    memory_key="chat_history",
+    return_messages=True,
+    )
+
+# format_instructions = experiment_output_parser.get_format_instructions()
+# maker_format_instructions = maker_output_parser.get_format_instructions()
+
+# output_parser = StructuredOutputParser.from_response_schemas(maker_schema)
+
+format_instructions = output_parser.get_format_instructions()
+
+# experiment_prompt = PromptTemplate(
+#     template="You must generate well detailed science experiments.\n{format_instructions}\n{question}\n{context}",
+#     input_variables=["question"],
+#     partial_variables={"format_instructions": format_instructions},
+#     memory = memory
+# )
 
 maker_prompt = PromptTemplate(
     template="You must generate a well detailed list of items for creating a given item from scratch. \
@@ -87,7 +100,6 @@ def format_docs(docs):
     return "\n\n".join([join_strings(d.page_content, d.metadata['Entry ID'],d.metadata['Title'], ) for d in docs])
 
 
-arxiv_retriever = ArxivRetriever(load_max_docs=2)
 
 # model = ChatOpenAI(temperature=0)
 model = ChatOpenAI(temperature=0,model="gpt-4")
@@ -99,37 +111,35 @@ pub_med_retriever = PubMedRetriever()
 
 wikipedia_retriever = WikipediaRetriever()
 
-arxiv_chain = (
-    {"context": arxiv_retriever, "question": RunnablePassthrough()}
-    | experiment_prompt
-    | model
-    | experiment_output_parser
-)
-
-pub_med_chain = (
-    {"context": pub_med_retriever, "question": RunnablePassthrough()}
-    | experiment_prompt
-    | model
-    | experiment_output_parser
-)
-
-wikipedia_chain = (
-    {"context": wikipedia_retriever, "question": RunnablePassthrough()}
-    | experiment_prompt
-    | model
-    | experiment_output_parser
-)
+# arxiv_chain = (
+#     {"context": arxiv_retriever, "question": RunnablePassthrough()}
+#     | experiment_prompt
+#     | model
+#     | experiment_output_parser
+# )
+
+# pub_med_chain = (
+#     {"context": pub_med_retriever, "question": RunnablePassthrough()}
+#     | experiment_prompt
+#     | model
+#     | experiment_output_parser
+# )
+
+# wikipedia_chain = (
+#     {"context": wikipedia_retriever, "question": RunnablePassthrough()}
+#     | experiment_prompt
+#     | model
+#     | experiment_output_parser
+# )
 
 maker_wikipedia_chain = (
     {"context": wikipedia_retriever, "question": RunnablePassthrough()}
     | maker_prompt
     | model
-    | maker_output_parser
+    | output_parser
 )
 
 
-
-
 if __name__ == "__main__":
 
 

main.py

@@ -1,23 +1,40 @@
-from langhchain_generate_components import maker_wikipedia_chain
+# from langhchain_generate_components import maker_wikipedia_chain
 from utils import (
     save_file, convert_obj_to_stl,
-    change_file_extension,
+    change_file_extension, file_to_base64,
 )
 from mesh_utils import generate_mesh_images
 from gradio_client import Client
+from weaviate_utils import init_client
+from datetime import datetime
+from structured__apparatus_chain import (
+    wikipedia_chain
+)
+from datetime import datetime, timezone
+from dotenv import load_dotenv
+import os
+
+load_dotenv()
 
+HF_API_KEY = os.getenv("HUGGINGFACE_API_KEY")
+OPENAI_APIKEY = os.getenv("OPENAI_API_KEY")
+OPENAI_APIKEY = os.getenv("OPENAI_APIKEY")
 
 def main():  
     # the object to be generated
     query = "A Microscope"
     
     # using a retriever we generat a list of Components
-    output = maker_wikipedia_chain.invoke(query)
+    output = wikipedia_chain.invoke(query)
     
     # the first item 
     shap_e_sample = output['Material'][0]
+    shap_e_list = output['Fields_of_study']
+    
+
     
     client = Client("hysts/Shap-E")
+    client.hf_token = os.getenv("HUGGINGFACE_API_KEY")
     result = client.predict(
             shap_e_sample,	# str  in 'Prompt' Textbox component
             1621396601,	# float (numeric value between 0 and 2147483647) in 'Seed' Slider component
@@ -26,6 +43,25 @@ def main():
             api_name="/text-to-3d"
     )
     
+    weaviate_client = init_client()
+    component_collection = weaviate_client.collections.get("Component")
+    component_image_collection = weaviate_client.collections.get("ComponentImage")
+    
+    base_64_result = file_to_base64(result)
+    
+    uuid = component_collection.data.insert({
+        "DateCreated" : datetime.now(timezone.utc),
+        "UsedInComps" : [query],
+        "ToolName" : shap_e_sample,
+        "Tags" : shap_e_list,
+        "feildsOfStudy" : shap_e_list,
+        # "GlbBlob" : base_64_result,
+    })
+    
+    
+    
+    
+    
     saved_file_name = "sample.glb" 
     # save to local machine
     save_file(result,saved_file_name)
@@ -46,10 +82,28 @@ def main():
     
     viewing_angles = [(30, 45), (60, 90), (45, 135)]
     
-    generate_mesh_images(
-        stl_file_location, 
-        viewing_angles
-        )
+    # generate_mesh_images(
+    #     stl_file_location, 
+    #     viewing_angles
+    #     )
+    
+    data_location = generate_mesh_images(
+        stl_file_location,
+        viewing_angles,
+        "data",    
+    )
+    
+    for item1, item2 in zip(data_location, viewing_angles):
+    
+        base_64_result = file_to_base64(item1)
+    
+        image_uuid = component_image_collection.data.insert({
+            "DateCreated" : datetime.now(timezone.utc),
+            "ImageAngle" : [str(i) for i in item2],
+            "BelongsToComponent" : uuid,
+        })
+    
+    
     
     # These screenshots need to be given to GPT-V
     # for feedback

utils.py

@@ -1,5 +1,6 @@
-import shutil
 import trimesh
+import shutil
+import base64
 import os
 
 def save_file(input_file, output_file):
@@ -41,4 +42,34 @@ def change_file_extension(file_path: str, new_extension: str) -> str:
     """
     base_path, _ = os.path.splitext(file_path)
     new_file_path = base_path + '.' + new_extension
-    return new_file_path
+    return new_file_path
+
+        
+def file_to_base64(file_location: str) -> str:
+    """
+    This function is to convert files into base64
+    ## Input
+    file_location (str) : the location of the file on the machine
+    
+    ## Output
+    the base64 encoding of the file
+    """
+    with open(file_location, "rb") as file:
+        file_content = file.read()
+        base64_encoded = base64.b64encode(file_content)
+        return base64_encoded.decode("utf-8")
+    
+def base64_to_file(base64_string: str, output_file_location: str) -> None:
+    """
+    Decodes a base64-encoded string and writes the resulting binary data to a file.
+
+    Args:
+        base64_string (str): The base64-encoded string to decode.
+        output_file_location (str): The file path where the decoded binary data will be written.
+
+    Returns:
+        None
+    """
+    binary_data = base64.b64decode(base64_string)
+    with open(output_file_location, "wb") as output_file:
+        output_file.write(binary_data)

weaviate_utils.py

@@ -18,66 +18,71 @@ def init_client():
     auth_credentials=weaviate.auth.AuthApiKey(
         os.getenv("YOUR_WCS_AUTH_KEY")
     ),  # Set this environment variable
+    headers={
+        "X-OpenAI-Api-Key": os.environ[
+            "OPENAI_API_KEY"
+        ]  # Replace with your inference API key
+    },
     )
     return client
 
 
 
 x = 0
+def main() :
+    try:
+        client = init_client()
+        client.collections.create(
+        name="Component",
+        description="Component of a given Apparatus",
+        vectorizer_config=wvc.config.Configure.Vectorizer.text2vec_openai(),
+        properties=[
+            # wvc.config.Property(name="id", data_type=wvc.config.DataType.UUID),
+            wvc.config.Property(name="DateCreated", data_type=wvc.config.DataType.DATE),
+            wvc.config.Property(name="UsedInComps", data_type=wvc.config.DataType.TEXT_ARRAY),
+            wvc.config.Property(name="FeildsOfStudy", data_type=wvc.config.DataType.TEXT_ARRAY),
+            wvc.config.Property(name="ToolName", data_type=wvc.config.DataType.TEXT),
+            wvc.config.Property(name="Tags", data_type=wvc.config.DataType.TEXT_ARRAY),
+            wvc.config.Property(name="GlbBlob", data_type=wvc.config.DataType.BLOB),
+        ]
+        )
+    finally:
+        client.close()
 
-try:
-    client = init_client()
-    client.collections.create(
-    name="Component",
-    description="Component of a given Apparatus",
-    vectorizer_config=wvc.config.Configure.Vectorizer.text2vec_openai(),
-    properties=[
-        # wvc.config.Property(name="id", data_type=wvc.config.DataType.UUID),
-        wvc.config.Property(name="DateCreated", data_type=wvc.config.DataType.DATE),
-        wvc.config.Property(name="UsedInComps", data_type=wvc.config.DataType.TEXT_ARRAY),
-        wvc.config.Property(name="FeildsOfStudy", data_type=wvc.config.DataType.TEXT_ARRAY),
-        wvc.config.Property(name="ToolName", data_type=wvc.config.DataType.TEXT),
-        wvc.config.Property(name="Tags", data_type=wvc.config.DataType.TEXT_ARRAY),
-        wvc.config.Property(name="GlbBlob", data_type=wvc.config.DataType.BLOB),
-    ]
-)
-finally:
-    client.close()
-
-try:
-    client = init_client()
-    client.collections.create(
-    name="ScienceEperiment",
-    description="Science Experiment with the goal of making something",
-    vectorizer_config=wvc.config.Configure.Vectorizer.text2vec_openai(),
-    properties=[
-        # wvc.config.Property(name="id", data_type=wvc.config.DataType.UUID),
-        wvc.config.Property(name="DateCreated", data_type=wvc.config.DataType.DATE),
-        wvc.config.Property(name="UsedInComps", data_type=wvc.config.DataType.TEXT_ARRAY),
-        wvc.config.Property(name="FeildsOfStudy", data_type=wvc.config.DataType.TEXT_ARRAY),
-        wvc.config.Property(name="ToolName", data_type=wvc.config.DataType.TEXT),
-        wvc.config.Property(name="Tags", data_type=wvc.config.DataType.TEXT_ARRAY),
-    ]
-)
-finally:
-    client.close()
+    try:
+        client = init_client()
+        client.collections.create(
+        name="ScienceEperiment",
+        description="Science Experiment with the goal of making something",
+        vectorizer_config=wvc.config.Configure.Vectorizer.text2vec_openai(),
+        properties=[
+            # wvc.config.Property(name="id", data_type=wvc.config.DataType.UUID),
+            wvc.config.Property(name="DateCreated", data_type=wvc.config.DataType.DATE),
+            wvc.config.Property(name="UsedInComps", data_type=wvc.config.DataType.TEXT_ARRAY),
+            wvc.config.Property(name="FeildsOfStudy", data_type=wvc.config.DataType.TEXT_ARRAY),
+            wvc.config.Property(name="ToolName", data_type=wvc.config.DataType.TEXT),
+            wvc.config.Property(name="Tags", data_type=wvc.config.DataType.TEXT_ARRAY),
+        ]
+    )
+    finally:
+        client.close()
 
-try:
-    client = init_client()
-    client.collections.create(
-    name="ComponentImage",
-    description="An image to gain visual context on a component",
-    vectorizer_config=wvc.config.Configure.Vectorizer.text2vec_openai(),
-    properties=[
-        # wvc.config.Property(name="id", data_type=wvc.config.DataType.UUID),
-        wvc.config.Property(name="DateCreated", data_type=wvc.config.DataType.DATE),
-        wvc.config.Property(name="ImageContent", data_type=wvc.config.DataType.BLOB),
-        wvc.config.Property(name="ImageAngle", data_type=wvc.config.DataType.TEXT_ARRAY),
-        wvc.config.Property(name="BelongsToComponent", data_type=wvc.config.DataType.UUID),
-    ]
-)
-finally:
-    client.close()
+    try:
+        client = init_client()
+        client.collections.create(
+        name="ComponentImage",
+        description="An image to gain visual context on a component",
+        vectorizer_config=wvc.config.Configure.Vectorizer.text2vec_openai(),
+        properties=[
+            # wvc.config.Property(name="id", data_type=wvc.config.DataType.UUID),
+            wvc.config.Property(name="DateCreated", data_type=wvc.config.DataType.DATE),
+            wvc.config.Property(name="ImageContent", data_type=wvc.config.DataType.BLOB),
+            wvc.config.Property(name="ImageAngle", data_type=wvc.config.DataType.TEXT_ARRAY),
+            wvc.config.Property(name="BelongsToComponent", data_type=wvc.config.DataType.UUID),
+        ]
+    )
+    finally:
+        client.close()