关于云脑任务中统一路径访问方式的公告>>>
Chunxiang Xu avadesian
Public Activity Repositories Datasets Starred Repositories
22
 Following
1
 Followers
11
 Achievement Badge
1
Loading Heatmap…

avadesian synced commits to client-server at avadesian/skypilot from mirror

12 hours ago

avadesian synced commits to master at avadesian/skypilot from mirror

  • 54fe787d1f update managed jobs docs (#4665) * update managed jobs docs * update quickstart * address PR comments * Update docs/source/running-jobs/many-jobs.rst * Update docs/source/running-jobs/many-jobs.rst
  • e774fc3a29 Remove backward compatibility code in the 0.8.0 release (#4650) * remove 0.8.0 * bug fix * cached_cluster_info fix * smoke test bug fix * restore change * restore comment * Update sky/backends/cloud_vm_ray_backend.py Co-authored-by: Zhanghao Wu <zhanghao.wu@outlook.com> * verify ret * restore change --------- Co-authored-by: Zhanghao Wu <zhanghao.wu@outlook.com>
  • 1eaba808b0 [UX] Remove do exception in `sky check` (#4670) * remove do exception in sky check * check import first
  • 10213ec495 [Catalog] Workaround for wrong accelerator name of `p5en.48xlarge` returned by AWS api (#4669) fix: workaround for wrong acc_name returned by aws api
  • Compare 4 commits »

1 day ago

avadesian synced commits to client-server at avadesian/skypilot from mirror

1 day ago

avadesian synced commits to master at avadesian/skypilot from mirror

  • 5e6b39ce9a Change added `USER_HASH` in models name to full digits (8) instead of half (4) to prevent potential hash conflict (#4592) * test: print * feat: use 8 digits for user hash * test: full user hash * test: the content is shorten due to length limit * feat: convert user hash from hex to base36 * chore: extend support lifetime of relied code * fix: after merge, re-delete extra user hash splition in smoke test * style: format and type hints * Revert "feat: convert user hash from hex to base36" This reverts commit 2f7ca1429d1b7fc7869970e71eb892ce9ea49aaa. * Update tests/unit_tests/test_common_utils.py --------- Co-authored-by: Zhanghao Wu <zhanghao.wu@outlook.com>

1 day ago

avadesian synced commits to client-server at avadesian/skypilot from mirror

  • 01196ebe69 [Docs] Update endpoint fetching docs for client-server (#4662) Clean up endpoint docs
  • 190f63abe8 Merge branch 'master' of github.com:skypilot-org/skypilot into client-server
  • a5c3b52734 Merge branch 'master' of github.com:skypilot-org/skypilot into client-server
  • 5e6b39ce9a Change added `USER_HASH` in models name to full digits (8) instead of half (4) to prevent potential hash conflict (#4592) * test: print * feat: use 8 digits for user hash * test: full user hash * test: the content is shorten due to length limit * feat: convert user hash from hex to base36 * chore: extend support lifetime of relied code * fix: after merge, re-delete extra user hash splition in smoke test * style: format and type hints * Revert "feat: convert user hash from hex to base36" This reverts commit 2f7ca1429d1b7fc7869970e71eb892ce9ea49aaa. * Update tests/unit_tests/test_common_utils.py --------- Co-authored-by: Zhanghao Wu <zhanghao.wu@outlook.com>
  • Compare 4 commits »

1 day ago

avadesian synced commits to master at avadesian/skypilot from mirror

  • abc81ab2a2 [jobs] catch jobs that are DONE but non-terminal (#4641) * [jobs] catch jobs that are DONE but non-terminal Usually, a job should always transition to a terminal status before it is set to DONE. However, if a bug makes that fail for some reason (e.g. the issues fixed in #4637), we should still catch this state. * address PR comments * address more PR comments * Update sky/jobs/utils.py Co-authored-by: Zhanghao Wu <zhanghao.wu@outlook.com> * Update sky/jobs/utils.py Co-authored-by: Zhanghao Wu <zhanghao.wu@outlook.com> --------- Co-authored-by: Zhanghao Wu <zhanghao.wu@outlook.com>
  • 3e9bd9d0fa [jobs] fix possible managed job status issues (#4637) * [jobs] don't set end_at for CANCELLING CANCELLING is no longer a terminal status - we shouldn't set end_at. * [jobs] don't require job to be CANCELLING to become CANCELLED It's possible that the job status erroneously gets set to some other value (e.g. RUNNING) by another process after getting set to CANCELLING - we should avoid this case, but if we end up here, we should still transition to CANCELLED. * [jobs] avoid status transitions that don't make sense I saw a bug where a job that was CANCELLING got updated by RUNNING by a different process. Avoid state transitions like this that obviously violate the expected state machine. * fix cancellation * Revert "[jobs] don't require job to be CANCELLING to become CANCELLED" This reverts commit 93966374a768c649f122dfc6c3dfbe1a38bb9aab.
  • 2d8bfc079d [Doc][Serve] Document for HTTPS (#3972) * init * nit * address minor comments * add reference link * minor
  • 0b107210a7 remove remaining references to "spot" controller (#4631) * remove remaining references to "spot" controller * fix test * re-add LEGACY_JOB_CONTROLLER_NAME * update comment * remove managed_job_version completely
  • ade003a8db Adding logo to the documentation page, Vast and DO (#4651) * Vast: Adding logo to the documentation page * Adding DO image * adding DO and Vast to the enumerations * Swapping the ibm and k8s logo
  • Compare 5 commits »

2 days ago

avadesian synced commits to docs-shorter-toc at avadesian/skypilot from mirror

2 days ago

avadesian synced commits to main at avadesian/Vicuna from mirror

2 days ago

avadesian synced commits to master at avadesian/skypilot from mirror

  • 1e284afda6 [jobs queue] move scheduler state to its own column (#4645) * [jobs queue] move scheduler state to its own column This allows it to be excluded from dashboard. * fix lint * add comment
  • 63c04b4313 [Examples][Jobs] Example on Constructing Vector Database with Skypilot (#4610) * bootstrap with the existing code * Refactor CLIP batch processing script for ImageNet embeddings generation * Refactor and clean up CLIP batch processing and vector database scripts * Remove batch processing CLIP script and related files * Remove unnecessary ls command in CLIP batch processing script * Remove batch inference and vector database scripts along with associated configuration files. This cleanup eliminates unused components from the batch inference example, streamlining the project structure. * enrich readme * Update vector database example: Enhance README, refactor scripts to handle image data as base64, and improve search interface with HTML response. This commit includes changes to metadata handling in the database, updates to the batch processing logic, and a new search page for better user interaction. * format * Update search page title in serve_vectordb.py to "SkyPilot Image Search" * fix paths * format * Optimize batch vector computation script and update documentation link - Modify batch_compute_vectors.py to use 2 jobs by default instead of 1 - Reduce default end index from 10000 to 1000 - Add detach_run=True to sky.jobs.launch for better job management - Update README with a direct link to SkyPilot installation documentation * Update vector database example paths and README instructions - Correct workdir in build_vectordb.yaml to point to the correct example directory - Update README.md command instructions to match current script and YAML file names - Remove incomplete comment in README.md * Refactor batch processing script to improve vector computation and file management - Modify `run()` method to create separate partition directories for each batch - Use `pyarrow` engine for Parquet file writing - Simplify logging and result handling - Remove unnecessary import statement - Improve code clarity and error handling in batch processing * Update vector database README with detailed instructions and example outputs - Enhance README.md with more precise instructions for vector database workflow - Add example console output for vector database building process - Clarify query methods for different deployment scenarios - Update workdir in serve_vectordb.yaml to correct path * Simplify batch vector computation script by removing run-as-launch option - Remove `--run-as-launch` argument from batch_compute_vectors.py - Standardize job launching with `sky.jobs.launch()` - Update job name to be more descriptive - Modify README.md to reference OpenAI CLIP more explicitly * format * Update README.md with example output for batch vector computation - Add sample console output for batch vector computation process - Provide expected log messages showing partition saving details - Enhance documentation clarity for vector database workflow * Improve batch vector computation with atomic file writing and AWS cloud support - Modify batch processing to use atomic file writing with temporary files - Update compute_vectors.yaml to specify AWS cloud - Increase default end index in batch_compute_vectors.py from 1000 to 10000 - Enhance logging for partition saving with more detailed information * Update batch vector computation parameters and file handling - Increase default end index in batch_compute_vectors.py from 10,000 to 1,000,000 - Increase default number of jobs from 2 to 20 for improved parallel processing - Modify file saving method to use shutil.copy2() instead of os.replace() - Add explicit temp file cleanup after copying to final destination * formatting * remove aws * Delete examples/deepseek-janus/README.md * fix comments * Refactor vector database build script - Remove redundant `read_parquet_file()` function - Update `process_batch()` to use `documents` instead of `metadatas` - Improve batch size help text to clarify memory constraints - Simplify Parquet file reading by using `pd.read_parquet()` directly * fix comments * Update vector database README with improved explanations and ChromaDB reference - Refine language and clarity of vector database workflow steps - Add explicit mention of ChromaDB as the vector database engine - Improve descriptions of serving the vector database with Sky Serve - Clarify the purpose and process of each step in the README * Fix vector database build script parameter and collection addition - Correct `--embedding-dir` to `--embeddings-dir` in build configuration - Simplify `process_batch()` function to directly add embeddings and image base64 to ChromaDB collection - Convert ids, embeddings, and images to lists before adding to collection * format * Parallelize vector database build script with multiprocessing - Implement parallel processing of Parquet files using ProcessPoolExecutor - Add multiprocessing support to improve performance of vector database creation - Refactor file processing to handle batches in parallel - Dynamically set number of workers based on available CPU cores * Enhance vector database web interface and add docstrings - Redesign serve_vectordb.py frontend with modern, responsive styling - Add descriptive docstrings to vector database scripts - Improve search functionality with better error handling and loading state - Reduce default number of jobs in batch_compute_vectors.py from 100 to 25 * Refactor compute_vectors.py to improve Parquet file saving - Move `save_results_to_parquet()` method out of `run()` method - Enhance atomic file writing for better reliability - Add logging for partition saving process - Simplify method placement in BatchProcessor class * fix serving * Modify vector database scripts to save and serve images directly from filesystem - Update compute_vectors.py to save images to mounted bucket instead of base64 encoding - Modify serve_vectordb.py to serve images from filesystem via new endpoint - Update YAML configs to mount images directory consistently across scripts - Refactor image handling to use file paths instead of base64 representations * ordering * fix comments * format * Update examples/vector_database/README.md Co-authored-by: Christopher Cooper <christopher@cg505.com> * Update examples/vector_database/compute_vectors.yaml Co-authored-by: Christopher Cooper <christopher@cg505.com> * Update examples/vector_database/scripts/build_vectordb.py Co-authored-by: Christopher Cooper <christopher@cg505.com> * change workdir path * Improve vector database script initialization and model setup - Update build_vectordb.py to handle temporary directory for read/write buckets - Modify collection creation metadata description - Refactor compute_vectors.py to initialize model before batch processing - Move model initialization to ensure it's set up before data loading * Update vector database configuration and job parameters - Increase default number of jobs in batch_compute_vectors.py from 25 to 100 - Modify compute_vectors.yaml to use environment variable placeholders - Update cloud storage bucket names for embedding and image data * changing colons * Update examples/vector_database/compute_vectors.yaml Co-authored-by: Christopher Cooper <christopher@cg505.com> * Update examples/vector_database/scripts/build_vectordb.py Co-authored-by: Christopher Cooper <christopher@cg505.com> * last fix --------- Co-authored-by: Christopher Cooper <christopher@cg505.com>
  • Compare 2 commits »

3 days ago

avadesian synced commits to master at avadesian/skypilot from mirror

  • e7d94e956e [k8s] Fix k8s import error when running non-common accelerators (#4647) * Fix dependency check * comments * lint
  • b142e0b03d [jobs] set a global limit of 2000 parallel jobs (#4639) * [jobs] set a global limit of 2000 parallel jobs Based on testing, we start to hit other significant issues past this point, such as global_user_state sqlite database breaking. * address comments
  • 5e3e430653 Preliminary Vast AI support (#4365) * Preliminary Vast AI support * Update sky/provision/vast/instance.py Co-authored-by: Tian Xia <cblmemo@gmail.com> * Update sky/clouds/vast.py Co-authored-by: Tian Xia <cblmemo@gmail.com> * Update sky/clouds/vast.py Co-authored-by: Tian Xia <cblmemo@gmail.com> * Update sky/clouds/vast.py Co-authored-by: Tian Xia <cblmemo@gmail.com> * Updating the vast dependencies in the setup.py * Vast: Copy update on object stores * Vast: update base image dockerhub link * Vast: removing errant comment * Vast: provision/utils cleanup of a shallow copy * Vast: Simplifying the credential files mount * Vast: Linter cleanup * Vast: Internal api cleanup * Vast: Adding the catalog_fetcher * Vast: Linting fixes * Vast: Linting fixes * Vast: ordering the ports * Vast: Updating a function signature * Vast: comment cleanup * Vast: Adding a comment for disk_size calculation * Vast: Comment on the geolocation string processing * Vast: Comment on the rammifications of searching for instances * Vast: Leaving a comment for a pylint exception * Update sky/provision/vast/instance.py Co-authored-by: Tian Xia <cblmemo@gmail.com> * Update sky/clouds/service_catalog/data_fetchers/fetch_vast.py Co-authored-by: Tian Xia <cblmemo@gmail.com> * Update sky/clouds/service_catalog/data_fetchers/fetch_vast.py Co-authored-by: Tian Xia <cblmemo@gmail.com> * Update sky/clouds/service_catalog/data_fetchers/fetch_vast.py Co-authored-by: Tian Xia <cblmemo@gmail.com> * Update sky/clouds/service_catalog/data_fetchers/fetch_vast.py Co-authored-by: Tian Xia <cblmemo@gmail.com> * Update sky/clouds/vast.py Co-authored-by: Tian Xia <cblmemo@gmail.com> * Vast: updating the catalog fetcher * Fixing a white-space error * Vast: Adding a comment for the instance type * Vast: rephrase a docstring * Vast: Reverting the setup.py * Vast: file reversion * Vast: Updating the MemoryGiB to reflect the correct value * Vast: Adding authorship to TODO * Vast: Filtering instances waiting on startup request * Vast: Updating the install docs * Vast: catalog updated to adaptor * Vast: fixing a filter_instance typo * Vast: stating open ports not supported * Vast: Adding a requested comment in the fetcher * Vast: Comment for the maximum cluster name limit * Vast: comment about disk space limits * Update sky/clouds/vast.py Co-authored-by: Tian Xia <cblmemo@gmail.com> * Update sky/provision/vast/instance.py Co-authored-by: Tian Xia <cblmemo@gmail.com> * Vast: Fixing a few inconsistencies * Vast: adding smoke tests and disabling some for vast * Vast: Updating the ssh key adding to avoid duplicates * Vast: formatter update * Vast: Updating catalog emitter with flat pricing * Vast: Using a longer polling loop * Vast: Updating the spot price to be real * Vast: Adding spot instance * Vast: updating test coverage * Vast: Forcing memory into the instance type Memory has to be consistent across instance types for various tests so the name of the instance types have been expanded to include them * Vast: test coverage updates * Vast: updating test coverage * Vast: formatter cleanup * trying to resolve weird git issue * Vast: adding back the code to the controller * Vast: updating to the new way of doing the controller * Vast: updating to the new way of doing the controller * Vast: excluding two more tests * Vast: tests are still not running properly * Vast: removing a dynamic port test * Vast: removing a dynamic port test * Update sky/clouds/service_catalog/data_fetchers/fetch_vast.py Co-authored-by: Tian Xia <cblmemo@gmail.com> * Update sky/clouds/service_catalog/data_fetchers/fetch_vast.py Co-authored-by: Tian Xia <cblmemo@gmail.com> * Vast.ai: forcing utils to not be a broken merge again * trying to fix a github issue * Vast: trying to fix a github bug * Vast: trying to fix a github pr issue * Vast: redoing the default image * Vast: reintroducing dependency * Vast: launch template cleanup * Vast: code formatting * Vast: Using the access key and not the master key * Vast: template cleanup * autostoped => autostopped * Removing an extra echo from the smoke tests * Expose the return code of a failing test to the logs * Vast: template cleanup * Vast: increasing the api version dependency * smoke test Log line: less -> less -r to maintain colors * Vast: test skipping * Vast: test skipping * Moving "less" suggestion to "less -r" to preserve the ANSI * Vast: test skipping * Some tests didn't have the right cloud differentiator * Vast: test skipping * Vast: skipping tests * linter cleanup * Update sky/provision/vast/utils.py Co-authored-by: Tian Xia <cblmemo@gmail.com> * Vast: skipping a test * tests: removing a useful extratimeout feature * Vast: ssh port code fix * Vast: ssh port code fix * Vast: removing some unused code from the test * Vastai: newer sdk with debugging ability * Vastai: force the cpu_ram to match in the search_offers * Vast: formatting update * Vast: limiting the catalog to fewer matches * Vast: enforcing memory requirements when considering instances * Vast: requiring only 80GB space instead of 256 for catalog inclusion * Vast: Richer error for a failed library import * Vast: pushing the sdk requirement version forward * Vast: pushing the dependency requirement forward * Vast: updating the controller requirement text * Vast: Adding georegion flag for the launch search feature * Vast: Adding the chunked feature flag to consolidate into instance types * Vast: using a new chunked feature flag in catalog searching * Vast: bumping sdk minimum version * Vast: catalog instance type naming scheme update * Bringing back the old isntance type format to satisfy the parser * Vast: skipping tests * Vast: skipping tests * Vast: adding a cardinality cutoff to the instance list generation --------- Co-authored-by: Tian Xia <cblmemo@gmail.com>
  • ba6e5f9fef [Jobs dashboard] Minor polishing. (#4644)
  • Compare 4 commits »

4 days ago

avadesian synced commits to azure-image-revert at avadesian/skypilot from mirror

  • 7b6e410f65 Merge branch 'master' of https://github.com/skypilot-org/skypilot into azure-image-revert
  • 12333d31f8 comment
  • b142e0b03d [jobs] set a global limit of 2000 parallel jobs (#4639) * [jobs] set a global limit of 2000 parallel jobs Based on testing, we start to hit other significant issues past this point, such as global_user_state sqlite database breaking. * address comments
  • 5e3e430653 Preliminary Vast AI support (#4365) * Preliminary Vast AI support * Update sky/provision/vast/instance.py Co-authored-by: Tian Xia <cblmemo@gmail.com> * Update sky/clouds/vast.py Co-authored-by: Tian Xia <cblmemo@gmail.com> * Update sky/clouds/vast.py Co-authored-by: Tian Xia <cblmemo@gmail.com> * Update sky/clouds/vast.py Co-authored-by: Tian Xia <cblmemo@gmail.com> * Updating the vast dependencies in the setup.py * Vast: Copy update on object stores * Vast: update base image dockerhub link * Vast: removing errant comment * Vast: provision/utils cleanup of a shallow copy * Vast: Simplifying the credential files mount * Vast: Linter cleanup * Vast: Internal api cleanup * Vast: Adding the catalog_fetcher * Vast: Linting fixes * Vast: Linting fixes * Vast: ordering the ports * Vast: Updating a function signature * Vast: comment cleanup * Vast: Adding a comment for disk_size calculation * Vast: Comment on the geolocation string processing * Vast: Comment on the rammifications of searching for instances * Vast: Leaving a comment for a pylint exception * Update sky/provision/vast/instance.py Co-authored-by: Tian Xia <cblmemo@gmail.com> * Update sky/clouds/service_catalog/data_fetchers/fetch_vast.py Co-authored-by: Tian Xia <cblmemo@gmail.com> * Update sky/clouds/service_catalog/data_fetchers/fetch_vast.py Co-authored-by: Tian Xia <cblmemo@gmail.com> * Update sky/clouds/service_catalog/data_fetchers/fetch_vast.py Co-authored-by: Tian Xia <cblmemo@gmail.com> * Update sky/clouds/service_catalog/data_fetchers/fetch_vast.py Co-authored-by: Tian Xia <cblmemo@gmail.com> * Update sky/clouds/vast.py Co-authored-by: Tian Xia <cblmemo@gmail.com> * Vast: updating the catalog fetcher * Fixing a white-space error * Vast: Adding a comment for the instance type * Vast: rephrase a docstring * Vast: Reverting the setup.py * Vast: file reversion * Vast: Updating the MemoryGiB to reflect the correct value * Vast: Adding authorship to TODO * Vast: Filtering instances waiting on startup request * Vast: Updating the install docs * Vast: catalog updated to adaptor * Vast: fixing a filter_instance typo * Vast: stating open ports not supported * Vast: Adding a requested comment in the fetcher * Vast: Comment for the maximum cluster name limit * Vast: comment about disk space limits * Update sky/clouds/vast.py Co-authored-by: Tian Xia <cblmemo@gmail.com> * Update sky/provision/vast/instance.py Co-authored-by: Tian Xia <cblmemo@gmail.com> * Vast: Fixing a few inconsistencies * Vast: adding smoke tests and disabling some for vast * Vast: Updating the ssh key adding to avoid duplicates * Vast: formatter update * Vast: Updating catalog emitter with flat pricing * Vast: Using a longer polling loop * Vast: Updating the spot price to be real * Vast: Adding spot instance * Vast: updating test coverage * Vast: Forcing memory into the instance type Memory has to be consistent across instance types for various tests so the name of the instance types have been expanded to include them * Vast: test coverage updates * Vast: updating test coverage * Vast: formatter cleanup * trying to resolve weird git issue * Vast: adding back the code to the controller * Vast: updating to the new way of doing the controller * Vast: updating to the new way of doing the controller * Vast: excluding two more tests * Vast: tests are still not running properly * Vast: removing a dynamic port test * Vast: removing a dynamic port test * Update sky/clouds/service_catalog/data_fetchers/fetch_vast.py Co-authored-by: Tian Xia <cblmemo@gmail.com> * Update sky/clouds/service_catalog/data_fetchers/fetch_vast.py Co-authored-by: Tian Xia <cblmemo@gmail.com> * Vast.ai: forcing utils to not be a broken merge again * trying to fix a github issue * Vast: trying to fix a github bug * Vast: trying to fix a github pr issue * Vast: redoing the default image * Vast: reintroducing dependency * Vast: launch template cleanup * Vast: code formatting * Vast: Using the access key and not the master key * Vast: template cleanup * autostoped => autostopped * Removing an extra echo from the smoke tests * Expose the return code of a failing test to the logs * Vast: template cleanup * Vast: increasing the api version dependency * smoke test Log line: less -> less -r to maintain colors * Vast: test skipping * Vast: test skipping * Moving "less" suggestion to "less -r" to preserve the ANSI * Vast: test skipping * Some tests didn't have the right cloud differentiator * Vast: test skipping * Vast: skipping tests * linter cleanup * Update sky/provision/vast/utils.py Co-authored-by: Tian Xia <cblmemo@gmail.com> * Vast: skipping a test * tests: removing a useful extratimeout feature * Vast: ssh port code fix * Vast: ssh port code fix * Vast: removing some unused code from the test * Vastai: newer sdk with debugging ability * Vastai: force the cpu_ram to match in the search_offers * Vast: formatting update * Vast: limiting the catalog to fewer matches * Vast: enforcing memory requirements when considering instances * Vast: requiring only 80GB space instead of 256 for catalog inclusion * Vast: Richer error for a failed library import * Vast: pushing the sdk requirement version forward * Vast: pushing the dependency requirement forward * Vast: updating the controller requirement text * Vast: Adding georegion flag for the launch search feature * Vast: Adding the chunked feature flag to consolidate into instance types * Vast: using a new chunked feature flag in catalog searching * Vast: bumping sdk minimum version * Vast: catalog instance type naming scheme update * Bringing back the old isntance type format to satisfy the parser * Vast: skipping tests * Vast: skipping tests * Vast: adding a cardinality cutoff to the instance list generation --------- Co-authored-by: Tian Xia <cblmemo@gmail.com>
  • ba6e5f9fef [Jobs dashboard] Minor polishing. (#4644)
  • Compare 40 commits »

4 days ago

avadesian synced commits to spot-hedge-new at avadesian/skypilot from mirror

4 days ago

avadesian synced commits to master at avadesian/skypilot from mirror

  • e4ad98c907 [Lambda] Fix missing private ip (#4635) * [Lambda] Fix missing private ip * remove extra API call * format test
  • 3b4f31b13c [Docs] Minor polishing on Multiple Kubernetes docs. (#4642)
  • 29fa533729 [Jobs][UX] Dashboard UI Upgrade with Log Download (#4638) * dashboard fixes * make the tables more organized * support downloading logs * make website more stylish * hover box on the status * Improve dashboard layout and responsiveness - Update page title to be more descriptive - Increase container max-width for better readability - Add responsive width and horizontal scroll for small screens - Adjust CSS to make dashboard more user-friendly * format * formatting * Improve launch history tracking in dashboard - Modify launch history extraction to track multiple launch attempts - Prioritize the most recent successful launch - Convert previous launch attempts to "Tried to launch" entries - Update tooltip CSS for better positioning and readability - Change dashboard title to be more specific * format * Enhance timestamp display and UI interactions in dashboard - Add hover effect for last updated timestamp with UTC tooltip - Remove refresh countdown timer - Change "Download Log" button text to "Controller Log" - Improve timestamp display with local and UTC time formats * Add interactive status filtering and badge interactions - Implement clickable status badges that filter job statuses - Add hover effect for status badges - Move status filter to a hidden select element - Preserve status filter state across page reloads - Enhance user interaction with job status display * Refactor job dashboard columns and display logic - Modify JobTableColumns to add new columns: FAILOVER and DETAILS - Update JOB_TABLE_COLUMNS to reflect new column structure - Adjust dashboard rendering to populate new columns - Improve error handling for column synchronization - Simplify row processing and status extraction - Update HTML template to match new column layout * Reorder dashboard columns and update template references - Swap FAILOVER and DETAILS column positions in JobTableColumns - Update HTML template to reference correct column indices - Modify comment to clarify column purpose * supporting tags and unselect * formatting * remove title dashboard * enable scrolling and better visualization * fix width * improve readability * Update job status filter display and label * fix table width * Update sky/jobs/dashboard/dashboard.py Co-authored-by: Zongheng Yang <zongheng.y@gmail.com> * fix nits --------- Co-authored-by: Christopher Cooper <cooperc@assemblesys.com> Co-authored-by: Zongheng Yang <zongheng.y@gmail.com>
  • 25ae5f257b [Docs] Multiple k8s support (#4586) * Show multiple kubernetes in the optimizer table * Add docs for multiple kubernetes * Add dynamic update * format * Add new button * Add to index * fix * Add figure for multi-k8s docs * Fix new badge * Update docs/source/reference/kubernetes/multi-kubernetes.rst Co-authored-by: Romil Bhardwaj <romil.bhardwaj@berkeley.edu> * update * Update docs/source/reference/kubernetes/multi-kubernetes.rst Co-authored-by: Romil Bhardwaj <romil.bhardwaj@berkeley.edu> * update * fix * fix * Update docs/source/reference/kubernetes/multi-kubernetes.rst Co-authored-by: Romil Bhardwaj <romil.bhardwaj@berkeley.edu> * rename * Update docs/source/reference/kubernetes/multi-kubernetes.rst Co-authored-by: Romil Bhardwaj <romil.bhardwaj@berkeley.edu> * Update docs/source/reference/kubernetes/multi-kubernetes.rst Co-authored-by: Romil Bhardwaj <romil.bhardwaj@berkeley.edu> * Update docs/source/reference/kubernetes/multi-kubernetes.rst Co-authored-by: Romil Bhardwaj <romil.bhardwaj@berkeley.edu> * revert to console, fix comment color * new badge * use comma instead --------- Co-authored-by: Romil Bhardwaj <romil.bhardwaj@berkeley.edu>
  • d4b1b43333 [Docs] Overview page (#4622) * WIP: Overview page * img * Updates * updates * rewords * remove comments * updates * Rewords
  • Compare 6 commits »

4 days ago

avadesian synced commits to multi-k8s-docs-v2 at avadesian/skypilot from mirror

  • eb5b16178a use comma instead
  • 7aac76a6ca new badge
  • 4368e4f9ee revert to console, fix comment color
  • 2797067e5a Merge branch 'multi-k8s-docs-v2' of github.com:skypilot-org/skypilot into multi-k8s-docs-v2
  • 34d6bbec71 Update docs/source/reference/kubernetes/multi-kubernetes.rst Co-authored-by: Romil Bhardwaj <romil.bhardwaj@berkeley.edu>
  • Compare 9 commits »

5 days ago

avadesian synced commits to docs-key-concepts at avadesian/skypilot from mirror

5 days ago

avadesian synced commits to multimodal-surfaces at avadesian/guidance from mirror

1 week ago

avadesian synced commits to main at avadesian/guidance from mirror

1 week ago

avadesian created online experience task avade202502011285191

1 week ago

avadesian synced commits to spot-hedge-new at avadesian/skypilot from mirror

1 week ago

avadesian synced commits to multi-k8s-docs-v2 at avadesian/skypilot from mirror

1 week ago

Please read the following content carefully:

Dear OpenI User

Thank you for your continuous support to the Openl Qizhi Community AI Collaboration Platform. In order to protect your usage rights and ensure network security, we updated the Openl Qizhi Community AI Collaboration Platform Usage Agreement in January 2024. The updated agreement specifies that users are prohibited from using intranet penetration tools. After you click "Agree and continue", you can continue to use our services. Thank you for your cooperation and understanding.

For more agreement content, please refer to the《Openl Qizhi Community AI Collaboration Platform Usage Agreement》

Agree and continue
Disagree, exit