Hey all, I'm having a heck of a time

chat_engine = index.as_chat_engine(
        chat_mode="context", 
        retriever_mode="embedding", 
        similarity_top_k=5,
        node_postprocessors=[reranker],
        verbose=True,
        system_prompt=" ".join((DM_Prompt, History_Prompt, Character_Prompt)),
        memory=memory
    )

    Settings.llm=llm
    Settings.chunk_size=512
    Settings.callback_manager=CallbackManager([token_counter])
    Settings.chunk_overlap=25
    Settings.embed_model=embed_model
    Settings.num_output=512

Query has been truncated from the right to 256 tokens from 1271 tokens.

context_str_template, nodes = await chat_engine._agenerate_context(message)
        prefix_messages = chat_engine._get_prefix_messages_with_context(context_str_template)
        initial_token_count = len(
            chat_engine._memory.tokenizer_fn(
                " ".join([(m.content or "") for m in prefix_messages])
            )
        )
        all_messages = prefix_messages + chat_engine._memory.get(
            initial_token_count=initial_token_count
        )
        all_messages_token_count = len(
            chat_engine._memory.tokenizer_fn(
                " ".join([(m.content or "") for m in all_messages])
            )
        )

        print(f"Context String: {context_str_template}")
        print(f"Prefix messages: {prefix_messages}")
        #print(f"Nodes: {nodes}")
        print(f"Initial Token Count: {initial_token_count}")
        print(f"All Messages: {all_messages}")
        print(f"All Message Token Count: {all_messages_token_count}")

import phoenix as px
px.launch_app()
from llama_index.core import set_global_handler
set_global_handler("arize_phoenix")

from llama_index.core.memory import ChatMemoryBuffer

memory = ChatMemoryBuffer.from_defaults(token_limit=3000)

index.as_chat_engine(..., memory=memory)