| Age | Commit message (Collapse) | Author | Lines |
|
Propagate the per-request Idempotency-Key (generated once in the retry
loop) through request construction so every retry attempt reuses the same
key rather than minting a new one.
- Add field to ChatCompletionsData.
- Generate a single key per logical request in chat_completions and
chat_completions_streaming.
- Thread the key through request_builder (macro-based clients), Bedrock,
and VertexAI so the header is attached once and reused on retries.
|
|
|
|
|
|
* Add config options subquery and subquery_max_depth.
* Register an internal subquery function that spawns a fresh LLM query
with the current model configuration but without the full context.
The model selects context numerically (message_indexes) or by text
excerpt (excerpts).
* Subqueries can use tools, including further nested subqueries up to
subquery_max_depth.
* Queries and subqueries get hierarchical ids ([0], [0/1], [0/1/2])
that prefix tool-call prints and spinner messages.
|
|
Co-authored-by: chenmch <chenmch@diit.cn>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
(#1054)
|
|
- update models.yaml
- update azure openai api version
- agent `run_instructions_fn`
|
|
append_command_to_history_file => save_shell_history
|