chore: remove auto warmup where possible (#409)

Author: julian-risch

integrations/astradb.md

@@ -94,7 +94,6 @@ documents = [Document(content="There are over 7,000 languages spoken around the
 						Document(content="In certain parts of the world, like the Maldives, Puerto Rico, and San Diego, you can witness the phenomenon of bioluminescent waves.")]
 
 document_embedder = SentenceTransformersDocumentEmbedder(model=model)  
-document_embedder.warm_up()
 documents_with_embeddings = document_embedder.run(documents)
 
 document_store.write_documents(documents_with_embeddings.get("documents"))

integrations/couchbase-document-store.md

@@ -212,7 +212,6 @@ from haystack.components.embedders import SentenceTransformersDocumentEmbedder
 documents = [Document(content="Alice has been living in New York City for the past 5 years.")]
 
 document_embedder = SentenceTransformersDocumentEmbedder(model="sentence-transformers/all-MiniLM-L6-v2")
-document_embedder.warm_up() # will download the model during first run
 documents_with_embeddings = document_embedder.run(documents)
 
 document_store.write_documents(documents_with_embeddings.get("documents"))
@@ -342,7 +341,6 @@ documents = [
 model_name = "sentence-transformers/all-MiniLM-L6-v2"
 
 document_embedder = SentenceTransformersDocumentEmbedder(model=model_name)
-document_embedder.warm_up()
 documents_with_embeddings = document_embedder.run(documents)
 
 document_store.write_documents(documents_with_embeddings.get("documents"))
@@ -525,7 +523,6 @@ documents = [
 
 model_name = "sentence-transformers/all-MiniLM-L6-v2"
 document_embedder = SentenceTransformersDocumentEmbedder(model=model_name)
-document_embedder.warm_up()
 documents_with_embeddings = document_embedder.run(documents)
 
 document_store.write_documents(documents_with_embeddings.get("documents"))

integrations/hanlp.md

@@ -69,9 +69,6 @@ splitter = ChineseDocumentSplitter(
     respect_sentence_boundary=True
 )
 
-# Warm up the component (loads the necessary models)
-splitter.warm_up()
-
 result = splitter.run(documents=[doc])
 print(result["documents"])
 ```
@@ -126,7 +123,6 @@ splitter = ChineseDocumentSplitter(
     split_overlap=3,
     respect_sentence_boundary=True
 )
-splitter.warm_up()
 result = splitter.run(documents=[doc])
 ```
 
@@ -147,7 +143,6 @@ splitter = ChineseDocumentSplitter(
     split_by="function",
     splitting_function=custom_chinese_split
 )
-splitter.warm_up()
 result = splitter.run(documents=[doc])
 ```
 

integrations/instructor-embedder.md

@@ -89,7 +89,6 @@ text_embedder = InstructorTextEmbedder(
     model="hkunlp/instructor-base", instruction=instruction,
     device=ComponentDevice.from_str("cpu"),
 )
-text_embedder.warm_up()
 result = text_embedder.run(text)
 print(f"Embedding: {result['embedding']}")
 print(f"Embedding Dimension: {len(result['embedding'])}")
@@ -112,8 +111,6 @@ doc_embedder = InstructorDocumentEmbedder(
     device=ComponentDevice.from_str("cpu"),
 )
 
-doc_embedder.warm_up()
-
 # Text taken from PubMed QA Dataset (https://huggingface.co/datasets/pubmed_qa)
 document_list = [
     Document(
@@ -174,8 +171,6 @@ doc_embedder = InstructorDocumentEmbedder(
     batch_size=32,
     device=ComponentDevice.from_str("cpu"),
 )
-# Warm up the embedder (loading the pre-trained model)
-doc_embedder.warm_up()
 
 # Create an indexing pipeline
 indexing_pipeline = Pipeline()
@@ -218,8 +213,6 @@ text_embedder = InstructorTextEmbedder(
     instruction=query_embedding_instruction,
     device=ComponentDevice.from_str("cpu"),
 )
-# Load the text embedding model
-text_embedder.warm_up()
 
 # Create a query pipeline
 query_pipeline = Pipeline()

integrations/langfuse.md

@@ -116,7 +116,6 @@ def get_pipeline(document_store: InMemoryDocumentStore):
 document_store = InMemoryDocumentStore()
 dataset = load_dataset("bilgeyucel/seven-wonders", split="train")
 embedder = SentenceTransformersDocumentEmbedder("sentence-transformers/all-MiniLM-L6-v2")
-embedder.warm_up()
 docs_with_embeddings = embedder.run([Document(**ds) for ds in dataset]).get("documents") or []  # type: ignore
 document_store.write_documents(docs_with_embeddings)
 

integrations/llama_cpp.md

@@ -105,7 +105,6 @@ generator = LlamaCppGenerator(
     model_kwargs={"n_gpu_layers": -1},
 		generation_kwargs={"max_tokens": 128, "temperature": 0.1},
 )
-generator.warm_up()
 prompt = f"Who is the best American actor?"
 result = generator.run(prompt)
 ```
@@ -129,7 +128,6 @@ generator = LlamaCppGenerator(
     n_batch=128,
     model_kwargs={"n_gpu_layers": -1}
 )
-generator.warm_up()
 prompt = f"Who is the best American actor?"
 result = generator.run(prompt, generation_kwargs={"max_tokens": 128})
 generated_text = result["replies"][0]
@@ -153,7 +151,6 @@ generator = LlamaCppGenerator(
     n_batch=128,
     generation_kwargs={"max_tokens": 128, "temperature": 0.1},
 )
-generator.warm_up()
 prompt = f"Who is the best American actor?"
 result = generator.run(prompt)
 ```
@@ -168,7 +165,6 @@ generator = LlamaCppGenerator(
     n_ctx=512,
     n_batch=128,
 )
-generator.warm_up()
 prompt = f"Who is the best American actor?"
 result = generator.run(
     prompt,

integrations/neo4j-document-store.md

@@ -153,7 +153,6 @@ documents = [
     Document(content="My name is Morgan and I live in Paris.", meta={"release_date": "2018-12-09"})]
 
 document_embedder = SentenceTransformersDocumentEmbedder(model="sentence-transformers/all-MiniLM-L6-v2")  
-document_embedder.warm_up()
 documents_with_embeddings = document_embedder.run(documents)
 
 document_store.write_documents(documents_with_embeddings.get("documents"))

integrations/nvidia.md

@@ -86,7 +86,6 @@ from haystack_integrations.components.embedders.nvidia import NvidiaTextEmbedder
 text_to_embed = "I love pizza!"
 
 text_embedder = NvidiaTextEmbedder(model="nvidia/llama-3.2-nv-embedqa-1b-v2")
-text_embedder.warm_up()
 
 print(text_embedder.run(text_to_embed))
 # {'embedding': [-0.02264290489256382, -0.03457780182361603, ...}
@@ -105,7 +104,6 @@ documents = [
 ]
 
 document_embedder = NvidiaDocumentEmbedder(model="nvidia/llama-3.2-nv-embedqa-1b-v2")
-document_embedder.warm_up()
 document_embedder.run(documents=documents)
 # {'documents': [Document(id=..., content: 'Pizza is made with dough and cheese', embedding: vector of size 2048), ...], 'meta': {'usage': {'prompt_tokens': 36, 'total_tokens': 36}}}
 ```
@@ -123,7 +121,6 @@ generator = NvidiaGenerator(
         "max_tokens": 1024,
     },
 )
-generator.warm_up()
 
 result = generator.run(prompt="When was the Golden Gate Bridge built?")
 print(result["replies"])
@@ -160,7 +157,6 @@ from haystack_integrations.components.rankers.nvidia import NvidiaRanker
 ranker = NvidiaRanker(
     api_key=Secret.from_env_var("NVIDIA_API_KEY"),
 )
-ranker.warm_up()
 
 query = "What is the capital of Germany?"
 documents = [
@@ -248,7 +244,6 @@ text_embedder = NvidiaTextEmbedder(model="nvidia/llama-3.2-nv-embedqa-1b-v2")
 retriever = InMemoryEmbeddingRetriever(document_store=document_store)
 prompt_builder = PromptBuilder(template=prompt)
 generator = NvidiaGenerator(model="meta/llama-3.1-70b-instruct")
-generator.warm_up()
 
 rag_pipeline = Pipeline()
 

integrations/opensearch-document-store.md

@@ -98,7 +98,6 @@ docs = [
 
 # Embed the documents and add them to the document store
 doc_embedder = SentenceTransformersDocumentEmbedder(model="sentence-transformers/all-MiniLM-L6-v2")
-doc_embedder.warm_up()
 docs = doc_embedder.run(docs)
 
 # Write the documents to the OpenSearch document store

integrations/snowflake.md

@@ -100,8 +100,6 @@ The `SnowflakeTableRetriever` supports three authentication methods:
 
 Ensure you have `select` access to the tables before querying the database. More details [here](https://docs.snowflake.com/en/user-guide/security-access-control-privileges):
 ```python
-# Warm up the component so it connects to the database
-executor.warm_up()
 
 # Run the retriever
 response = executor.run(query="""select * from database_name.schema_name.table_name""")
@@ -113,7 +111,6 @@ executor = SnowflakeTableRetriever(
     schema_name="<SCHEMA-NAME>",
     database ="<DB-NAME>"
 )
-executor.warm_up()
 
 response = executor.run(query="""select * from table_name""")
 ```