[pull] trunk from spiceai:trunk#614
Merged
Merged
Conversation
* Upgrade mistral.rs to v0.7.0 (spiceai-upgrade branch) Updates mistral.rs dependency to the new spiceai-upgrade branch based on upstream EricLBuehler/mistral.rs v0.7.0. Key changes: - Update imports to use mistralrs::core:: for internal types - Add new required fields: truncate_sequence, reasoning_effort, repetition_penalty - Handle new Response::Embeddings variant - Fix LocalModelPaths API (template_filename is now required) The new mistral.rs version: - Uses candle 0.9.2 from crates.io (no git dep needed) - Includes tool template warning (warns when tools used with non-tool templates) - Includes NaN panic with debug info in sampler Partially addresses #8634, #9274 * Add SOP for upgrading mistral.rs Documents the standard operating procedure for upgrading spiceai/mistral.rs fork to new upstream versions, including: - Identifying and maintaining spiceai patches - Rebasing onto new upstream versions - Handling API changes in spiceai/spiceai - Troubleshooting common issues * Add SOP for upgrading TEI Documents the standard operating procedure for upgrading spiceai/text-embeddings-inference fork, including: - Current limitations with CUDA auxiliary crates - Maintaining spiceai patches (download_safetensors, cudarc) - Merge conflict resolution guidance - Future work required for candle 0.9+ upgrade - Troubleshooting common cudarc version issues * format * license * update llm mistralrs dep
…for executors, and for use at query time. (#9325) * initial work and refactor * scheduler uses accelerations from executors at query time * compile * runtime: deregister executors on shutdown Send executor shutdown notifications over the control stream, use advertise host:port executor IDs, and add a regression check for prompt scheduler removal. Fixes #9289 * Jeadie/26 02 02/testing (#9303) * don't refresh accelerations on scheduler node * Add distributed query mode support to testoperator (#9248) Adds --distributed flag to testoperator commands (bench, query, throughput, load) to enable benchmarking queries via the /v1/queries async API used for distributed query execution in cluster mode. Changes: - Add QUERIES_ENDPOINT constant in test-framework/constants.rs - Add --distributed CLI flag to QueryArgs and DatasetTestArgs - Add distributed_mode field to NotStarted test builder and SpiceTestQueryWorker - Implement execute_distributed() method that: - Submits query via POST /v1/queries - Polls /v1/queries/{query_id}/status until completion - Fetches results from /v1/queries/{query_id}/results - Returns error if HTTP client not configured or manifest fields missing - Update execute_query() to route to distributed mode when enabled - Update all commands (bench, query, throughput, load) to pass distributed flag - Add distributed field to dispatch LoadArgs for workflow configuration - Add #[expect] attributes for clippy::struct_excessive_bools and too_many_lines * fix: Provide better job store error handling (#9235) * fix: Provide better job store error handling * chore: fix build * fix: propogate all listing errors * docs: Update docstrings * chore: bad merge * fix flight scan rule --------- Co-authored-by: Phillip LeBlanc <phillip@spice.ai> Co-authored-by: William <98815791+peasee@users.noreply.github.com> * add cookies * clean up: * fix * clippy * clippy * temp * fixes * write accelerate table partitions to object store * partition assigment in executor * temp POC * cleanup * fix default * improve * clippy * formatting * fix clippy * type alias PartitionAssignments * clippy * handle Expr in partition keys * StringArray -> BytesArray * formatting * support mTLS * clippy --------- Co-authored-by: Phillip LeBlanc <phillip@leblanc.tech> Co-authored-by: Phillip LeBlanc <phillip@spice.ai> Co-authored-by: William <98815791+peasee@users.noreply.github.com>
…hunks (#9355) * fix: Provide a more informative 404 on async API when result has no chunks * Update crates/runtime/src/http/v1/queries.rs Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> * Update crates/runtime/src/flight/async_actions.rs Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> * chore: fmt * chore: clippy --------- Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
4 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
See Commits and Changes for more details.
Created by
pull[bot] (v2.0.0-alpha.4)
Can you help keep this open source service alive? 💖 Please sponsor : )