adding a cookbook for finetune embeddings

cookbook/finetuning_embedding_model.py

@@ -0,0 +1,31 @@
+from beyondllm import source, retrieve, llms
+from beyondllm.embeddings import FineTuneEmbeddings
+import os
+
+
+# Setting up an environment variable for API key
+os.environ['GOOGLE_API_KEY'] = "your-api-key"
+
+# Importing and preparing the data
+data = source.fit("build-career-in-ai.pdf", dtype="pdf", chunk_size=1024, chunk_overlap=0)
+
+# List of files to train the embeddings
+list_of_files = ['build-career-in-ai.pdf']
+
+# Initializing a Gemini LLM model
+llm = llms.GeminiModel()
+
+# Creating an instance of FineTuneEmbeddings
+fine_tuned_model = FineTuneEmbeddings()
+
+# Training the embedding model
+embed_model = fine_tuned_model.train(list_of_files, "BAAI/bge-small-en-v1.5", llm, "fintune")
+
+# Option to load an already fine-tuned model
+# embed_model = fine_tuned_model.load_model("fintune")
+
+# Creating a retriever using the fine-tuned embeddings
+retriever = retrieve.auto_retriever(data, embed_model, type="normal", top_k=4)
+
+# Retrieving information using a query
+print(retriever.retrieve("How to excel in AI?"))