dev/openfuyao-2.48.0

文件	最后提交记录	最后更新时间
advanced-guides	[Docs] Troubleshooting DeepSeek/multi-node GPU deployment on KubeRay (#54229) Signed-off-by: Seiji Eicher <seiji@anyscale.com> Co-authored-by: angelinalg <122562471+angelinalg@users.noreply.github.com>	1 年前
api	[Serve] Add RouterConfig field to DeploymentConfig to configure RequestRouter (#53870) Signed-off-by: Seiji Eicher <seiji@anyscale.com>	1 年前
doc_code	[Serve] Add RouterConfig field to DeploymentConfig to configure RequestRouter (#53870) Signed-off-by: Seiji Eicher <seiji@anyscale.com>	1 年前
llm	[Serve.llm] Add a doc snippet to inform users about existing diffs between vllm serve and ray serve llm. (#54042) Signed-off-by: Kourosh Hakhamaneshi <kourosh@anyscale.com> Signed-off-by: kourosh hakhamaneshi <31483498+kouroshHakha@users.noreply.github.com> Co-authored-by: Seiji Eicher <58963096+eicherseiji@users.noreply.github.com>	1 年前
production-guide	[docs][Serve] Add clarification for health check and FT of serve deployments (#53944) This PR adds clarifications to the fault tolerance documentation for Serve deployments. * Provides details on how the Serve controller handles health checks for replicas. * Includes notes warning against directly calling the check_health method and explaining the behavior in composable deployment graphs. Signed-off-by: Kourosh Hakhamaneshi <kourosh@anyscale.com>	1 年前
tutorials	[Docs] Troubleshooting DeepSeek/multi-node GPU deployment on KubeRay (#54229) Signed-off-by: Seiji Eicher <seiji@anyscale.com> Co-authored-by: angelinalg <122562471+angelinalg@users.noreply.github.com>	1 年前
architecture-2.0.svg	[Doc][Serve] gRPC Proxy Docs Changes (#39140) Added usage doc and examples for using serve with gRPC. Also added related doc changes.	2 年前
architecture.md	[serve] rename max concurrent queries (#43007) Rename max_concurrent_queries to max_ongoing_requests. All tests have been migrated to use max_ongoing_requests except for a few tests that are parametrized to ensure max_concurrent_queries still works: * test_max_ongoing_requests_set_to_one (e2e test) * test_update_config_max_ongoing_requests (e2e test) * test_valid_deployment_schema (unit test) **Rollout plan**: support both and deprecate max_concurrent_queries for 2.10. Remove max_concurrent_queries next release. Signed-off-by: Cindy Zhang <cindyzyx9@gmail.com>	2 年前
autoscaling-guide.md	[serve] remove target_num_ongoing_requests_per_replica (#46392) [serve] remove target_num_ongoing_requests_per_replica target_num_ongoing_requests_per_replica was deprecated in ray 2.10, and target_ongoing_requests was introduced at the same time. There have been many releases since then, so we can remove target_num_ongoing_requests_per_replica now. Signed-off-by: Cindy Zhang <cindyzyx9@gmail.com> Signed-off-by: Cindy Zhang <cindyzyx9@gmail.com>	2 年前
configure-serve-deployment.md	Add max queued requests docs (#46778) <!-- Thank you for your contribution! Please review https://github.com/ray-project/ray/blob/master/CONTRIBUTING.rst before opening a pull request. --> <!-- Please add a reviewer to the assignee section when you create a PR. If you don't have the access to it, we will shortly find a reviewer and assign them to your PR. --> ## Why are these changes needed? <!-- Please give a short summary of the change and the problem this solves. --> ## Related issue number <!-- For example: "Closes #1234" --> ## Checks - [ ] I've signed off every commit(by using the -s flag, i.e., git commit -s) in this PR. - [ ] I've run scripts/format.sh to lint the changes in this PR. - [ ] I've included any doc changes needed for https://docs.ray.io/en/master/. - [ ] I've added any new APIs to the API Reference. For example, if I added a method in Tune, I've added it in doc/source/tune/api/ under the corresponding .rst file. - [ ] I've made sure the tests are passing. Note that there might be a few flaky tests, see the recent failures at https://flakey-tests.ray.io/ - Testing Strategy - [ ] Unit tests - [ ] Release tests - [ ] This PR is not tested :( Signed-off-by: akshay-anyscale <122416226+akshay-anyscale@users.noreply.github.com>	2 年前
develop-and-deploy.md	[docs][serve] Model Composition Guide: fixed references to line numbers and made style fixes (#39653) Fixed some line references and passive voice. Why are these changes needed? Improve usability and readability. --------- Signed-off-by: angelinalg <122562471+angelinalg@users.noreply.github.com>	2 年前
examples.yml	[Doc][Serve][LLM] Add doc for deploying DeepSeek (#52592) Signed-off-by: Rui Qiao <ruisearch42@gmail.com> Signed-off-by: Rui Qiao <161574667+ruisearch42@users.noreply.github.com> Co-authored-by: Gene Der Su <gdsu@ucdavis.edu>	1 年前
getting_started.md	[doc] Add hpu resource description in ray serve docs (#48796) HPU resource is already supported in Ray, and there are many examples to guide users to use HPU device in Ray, so this PR adds some instructions for HPU device to the Ray Serve related documents. --------- Signed-off-by: KepingYan <keping.yan@intel.com>	1 年前
http-guide.md	[Doc] Fix broken references in serve documentation (#45113) fixes broken links in Serve documentation in preparation for turning on Sphinx's nitpicky mode.	2 年前
index.md	[LLM APIs] Fast follow ups for 2.44 (1/N) (#51042) Some of these changes came from bug bash and dogfooding: - [x] Rename to VLLMService to VLLMServer - [x] Remove the extra name space hierarchy for most common import path to make things more consistent with data.llm - [x] Some other inconsitency stuff - [x] Use vLLM everywhere (instead of VLLM). Most of these changes should happen on serve side. - [x] in ray.data.llm use model_source in the vllm_config (model_id is a serve only concept as it refers to the model name available to the model discovery layer) from ray.serve.llm - [ ] support vllm v1 [in followup] - [ ] allow a single deployment to llm router (not force people to pass a list with one item) [in follow up] - [x] Update the ray serve docs structure to be more flat based on the dogfooding feedback. serve llm docs now look more flat and consistent with serve docs. Serving LLMs is a single page on the side bar and the apis are a sub-header under the ray serve apis page. Overview page <img width="1446" alt="image" src="https://github.com/user-attachments/assets/db850f35-3a3d-46e1-9892-b3cd17681b98" /> API page <img width="1433" alt="image" src="https://github.com/user-attachments/assets/03a669d1-12c2-46c1-9616-484ea95f7082" /> --------- Signed-off-by: Kourosh Hakhamaneshi <kourosh@anyscale.com>	1 年前
key-concepts.md	[Doc] Update key-concepts.md (#45245) <!-- Thank you for your contribution! Please review https://github.com/ray-project/ray/blob/master/CONTRIBUTING.rst before opening a pull request. --> <!-- Please add a reviewer to the assignee section when you create a PR. If you don't have the access to it, we will shortly find a reviewer and assign them to your PR. --> ## Why are these changes needed? The text in the [DeploymentHandle (composing deployments)](https://docs.ray.io/en/latest/serve/key-concepts.html#deploymenthandle-composing-deployments) paragraph of the Ray Serve documentation is not consistent with the code snippet linked below it. <!-- Please give a short summary of the change and the problem this solves. --> ## Related issue number There isn't related issue. Should I create one? <!-- For example: "Closes #1234" --> ## Checks - [ ] I've signed off every commit(by using the -s flag, i.e., git commit -s) in this PR. - [ ] I've run scripts/format.sh to lint the changes in this PR. - [ ] I've included any doc changes needed for https://docs.ray.io/en/master/. - [ ] I've added any new APIs to the API Reference. For example, if I added a method in Tune, I've added it in doc/source/tune/api/ under the corresponding .rst file. - [ ] I've made sure the tests are passing. Note that there might be a few flaky tests, see the recent failures at https://flakey-tests.ray.io/ - Testing Strategy - [ ] Unit tests - [ ] Release tests - [ ] This PR is not tested :( Signed-off-by: Antoni Jamiołkowski <50143141+antoni-jamiolkowski@users.noreply.github.com>	2 年前
logo.svg	Revert "[docs] Clean up doc structure (first part) (#21667)" (#21763) This reverts commit 38e46c9fb3b46810ea1c270cf95d6e65dc8ccaee.	4 年前
model-multiplexing.md	remove model multiplexing experimental feature documentation (#52122)	1 年前
model_composition.md	[serve] remove support for nested DeploymentResponses (#47209) Remove support for passing DeploymentResponses in nested objects to downstream serve deployment handle calls, e.g: handle.remote({"arg": other_handle.remote()}) This doesn't affect the no-op latency, but improves latencies of requests that carry a large payload. Current handle latencies: no-op: 1.67 1mb: 4.5 10mb: 41.53 New handle latencies: { "handle_p50_latency": 1.7133924999939154, "handle_1mb_p50_latency": 2.931859499994971, "handle_10mb_p50_latency": 12.14768799999888, } closes https://github.com/ray-project/ray/issues/46428 Signed-off-by: Cindy Zhang <cindyzyx9@gmail.com>	1 年前
monitoring.md	Fix typos (#53048) Signed-off-by: omahs <73983677+omahs@users.noreply.github.com>	1 年前
multi-app.md	[Serve] update serve build multi app doc (#45134) <!-- Thank you for your contribution! Please review https://github.com/ray-project/ray/blob/master/CONTRIBUTING.rst before opening a pull request. --> <!-- Please add a reviewer to the assignee section when you create a PR. If you don't have the access to it, we will shortly find a reviewer and assign them to your PR. --> ## Why are these changes needed? A user reported the --multi-app flag doesn't work. This option is already deprecated from the previous migration work and by default the command will generate multi-app config. This PR updated the docs on example command and the resulting config. This seems to be the only place that's still mentioning --multi-app. ## Related issue number <!-- For example: "Closes #1234" --> ## Checks - [ ] I've signed off every commit(by using the -s flag, i.e., git commit -s) in this PR. - [ ] I've run scripts/format.sh to lint the changes in this PR. - [ ] I've included any doc changes needed for https://docs.ray.io/en/master/. - [ ] I've added any new APIs to the API Reference. For example, if I added a method in Tune, I've added it in doc/source/tune/api/ under the corresponding .rst file. - [ ] I've made sure the tests are passing. Note that there might be a few flaky tests, see the recent failures at https://flakey-tests.ray.io/ - Testing Strategy - [ ] Unit tests - [ ] Release tests - [ ] This PR is not tested :( Signed-off-by: Gene Su <e870252314@gmail.com>	2 年前
resource-allocation.md	[doc] Add hpu resource description in ray serve docs (#48796) HPU resource is already supported in Ray, and there are many examples to guide users to use HPU device in Ray, so this PR adds some instructions for HPU device to the Ray Serve related documents. --------- Signed-off-by: KepingYan <keping.yan@intel.com>	1 年前