Add the llama index examples

examples/llama_index/mistral_gguf.py

@@ -0,0 +1,26 @@
+from llama_index.core import VectorStoreIndex, SimpleDirectoryReader, Settings
+from llama_index.core.embeddings import resolve_embed_model
+from llama_index.llms.mistral_rs import MistralRS
+
+documents = SimpleDirectoryReader("data").load_data()
+
+# bge embedding model
+Settings.embed_model = resolve_embed_model("local:BAAI/bge-small-en-v1.5")
+
+# ollama
+Settings.llm = MistralRS(
+ model_id="mistralai/Mistral-7B-Instruct-v0.1",
+ arch="mistral-gguf",
+ quantized_model_id="TheBloke/Mistral-7B-Instruct-v0.1-GGUF",
+ quantized_filename="mistral-7b-instruct-v0.1.Q4_0.gguf",
+ max_new_tokens=4096,
+ context_window=1000,
+)
+
+index = VectorStoreIndex.from_documents(
+ documents,
+)
+
+query_engine = index.as_query_engine()
+response = query_engine.query("Please summarize the above.")
+print(response)

examples/llama_index/xlora_gguf.py

@@ -0,0 +1,26 @@
+from llama_index.core import VectorStoreIndex, SimpleDirectoryReader, Settings
+from llama_index.core.embeddings import resolve_embed_model
+from llama_index.llms.mistral_rs import MistralRS
+
+documents = SimpleDirectoryReader("data").load_data()
+
+# bge embedding model
+Settings.embed_model = resolve_embed_model("local:BAAI/bge-small-en-v1.5")
+
+# ollama
+Settings.llm = MistralRS(
+ model_id="HuggingFaceH4/zephyr-7b-beta",
+ arch="x-lora-mistral",
+ xlora_order_file="ordering.json", # Copy your ordering file to where this file will be run from
+ xlora_model_id="lamm-mit/x-lora",
+ max_new_tokens=4096,
+ context_window=1000,
+)
+
+index = VectorStoreIndex.from_documents(
+ documents,
+)
+
+query_engine = index.as_query_engine()
+response = query_engine.query("Please summarize the above.")
+print(response)