Commit message (Collapse) | Author | Age | Files | Lines | |
---|---|---|---|---|---|
* | Clean up test code after move from container-search to model-integration | Lester Solbakken | 2024-04-12 | 2 | -136/+87 |
| | |||||
* | Move LLM client stuff from container-search to model-integration | Lester Solbakken | 2024-04-12 | 14 | -881/+60 |
| | |||||
* | Don't use GPU in unit test | Lester Solbakken | 2024-04-11 | 1 | -4/+0 |
| | |||||
* | Add tiny LLM for unit testing | Lester Solbakken | 2024-04-11 | 2 | -94/+37 |
| | |||||
* | Use 'model' config type for LLM models | Lester Solbakken | 2024-04-11 | 3 | -71/+68 |
| | |||||
* | Be able to render async error messages from event stream in json as well | Lester Solbakken | 2024-04-11 | 2 | -15/+46 |
| | |||||
* | Throw exception on too many LLM requests | Lester Solbakken | 2024-04-11 | 2 | -4/+7 |
| | |||||
* | Non-functional changes | Lester Solbakken | 2024-04-10 | 1 | -15/+20 |
| | |||||
* | Remove unneccessary code | Lester Solbakken | 2024-04-10 | 1 | -4/+0 |
| | |||||
* | Add local LLM client and wire in container-llama | Lester Solbakken | 2024-04-10 | 6 | -10/+564 |
| | |||||
* | Merge branch 'master' into lesters/update-platform-bundles-for-rag-2 | Lester Solbakken | 2024-04-04 | 1 | -1/+1 |
|\ | |||||
| * | Re-enable async LLM test | Lester Solbakken | 2024-04-02 | 1 | -2/+0 |
| | | |||||
| * | Temporary disable test | Lester Solbakken | 2024-04-02 | 1 | -1/+3 |
| | | |||||
* | | Moved ai.vespa.llm.search to ai.vespa.search.llm | Lester Solbakken | 2024-04-04 | 7 | -21/+21 |
| | | |||||
* | | Move LLM searcher and client configdefinitions outside of ai.vespa.llm | Lester Solbakken | 2024-04-02 | 12 | -16/+131 |
| | | |||||
* | | Rename ai.vespa.languagemodels to ai.vespa.llm in vespajlib | Lester Solbakken | 2024-04-02 | 10 | -36/+36 |
| | | |||||
* | | Move LLM classes in vespajlib from ai.vespa.llm to ai.vespa.languagemodels | Lester Solbakken | 2024-04-02 | 10 | -63/+68 |
|/ | |||||
* | Improve embedder error messages | Jon Bratseth | 2024-03-29 | 4 | -15/+22 |
| | |||||
* | Add beta annotation and update copyright headers | Lester Solbakken | 2024-03-27 | 6 | -1/+9 |
| | |||||
* | Rename apikey config to better reflect it is a name in secret store | Lester Solbakken | 2024-03-27 | 4 | -10/+9 |
| | |||||
* | Add RAG searcher | Lester Solbakken | 2024-03-26 | 14 | -0/+1119 |
| | |||||
* | Add synthetic targets so that you can always use cluster.schema as source ↵ | Henning Baldersheim | 2024-03-22 | 5 | -46/+84 |
| | | | | | | | for both streaming and indexed. - Make a SearchChainInvocationSpec proxy for all possible searchcluster.schema combinations. - It will modify the query with the actual source to use, and restrict to the given schema. | ||||
* | Handle the federation config in the federation searcher. | Henning Baldersheim | 2024-03-22 | 2 | -6/+8 |
| | |||||
* | Revert "fold AND and SAND items into top-level WEAKAND" | Arne H Juul | 2024-03-22 | 5 | -74/+25 |
| | |||||
* | Merge pull request #30707 from vespa-engine/arnej/fold-segments-into-weakand | Jon Bratseth | 2024-03-22 | 5 | -25/+74 |
|\ | | | | | fold AND and SAND items into top-level WEAKAND | ||||
| * | fold AND and SAND items into top-level WEAKAND | Arne Juul | 2024-03-21 | 5 | -25/+74 |
| | | |||||
* | | - GC unused code. | Henning Baldersheim | 2024-03-21 | 6 | -55/+20 |
| | | | | | | | | - GC unused id parameter. | ||||
* | | Merge pull request #30526 from vespa-engine/lesters/server-sent-events | Jon Bratseth | 2024-03-21 | 8 | -10/+657 |
|\ \ | |/ |/| | Add server-sent events (SSE) renderer | ||||
| * | Update ABI spec | Lester Solbakken | 2024-03-15 | 1 | -8/+22 |
| | | |||||
| * | Change EventStream to a DataList and be able that with JsonRenderer | Lester Solbakken | 2024-03-15 | 6 | -49/+161 |
| | | |||||
| * | Add server-sent events (SSE) renderer | Lester Solbakken | 2024-03-08 | 5 | -0/+521 |
| | | |||||
* | | - Document types with mode store-only are not searchable. | Henning Baldersheim | 2024-03-20 | 1 | -1/+1 |
| | | |||||
* | | Catch exceptions | Henning Baldersheim | 2024-03-19 | 1 | -1/+7 |
| | | |||||
* | | Move error handling to common component used by both streaming and indexed | Henning Baldersheim | 2024-03-19 | 4 | -13/+21 |
| | | |||||
* | | GC confusing and void ClusterConfig.clusterId | Henning Baldersheim | 2024-03-18 | 2 | -10/+4 |
| | | |||||
* | | Add necessary config to ClusterConfig to avoid hidden relation via clusterId ↵ | Henning Baldersheim | 2024-03-16 | 9 | -76/+37 |
| | | | | | | | | to QrSearchersConfig | ||||
* | | Revert "Revert "Single searchcluster take 3"" | Henning Baldersheim | 2024-03-16 | 2 | -6/+4 |
| | | |||||
* | | Revert "Single searchcluster take 3" | Henning Baldersheim | 2024-03-15 | 2 | -4/+6 |
| | | |||||
* | | Merge pull request #30644 from ↵ | Henning Baldersheim | 2024-03-15 | 2 | -6/+4 |
|\ \ | | | | | | | | | | | | | vespa-engine/revert-30643-revert-30642-revert-30640-revert-30620-revert-30616-revert-30615-balder/single-searchcluster Single searchcluster take 3 | ||||
| * | | Single searchcluster take 4 | Henning Baldersheim | 2024-03-15 | 2 | -6/+4 |
| | | | |||||
* | | | GC unused code | Henning Baldersheim | 2024-03-15 | 3 | -10/+4 |
|/ / | |||||
* | | Revert "Single searchcluster take 3" | Henning Baldersheim | 2024-03-15 | 2 | -4/+6 |
| | | |||||
* | | If any schema is streaming, cluster is streaming. | Henning Baldersheim | 2024-03-15 | 2 | -6/+4 |
| | | |||||
* | | Do all construction in constructor and make members final. | Henning Baldersheim | 2024-03-13 | 11 | -162/+111 |
| | | |||||
* | | No limitation for search clusters any more. | Henning Baldersheim | 2024-03-13 | 1 | -7/+5 |
| | | |||||
* | | Test that multiple backends can be used. | Henning Baldersheim | 2024-03-11 | 2 | -2/+29 |
| | | |||||
* | | Rename FastBackend => Indexedbackend, and move some tests into the package ↵ | Henning Baldersheim | 2024-03-11 | 5 | -50/+27 |
| | | | | | | | | they test. | ||||
* | | Correct naming | Henning Baldersheim | 2024-03-11 | 2 | -175/+1 |
| | | |||||
* | | Searcher => Backend | Henning Baldersheim | 2024-03-11 | 22 | -127/+125 |
| | | |||||
* | | Allow for backend per schema. | Henning Baldersheim | 2024-03-11 | 4 | -94/+88 |
| | |