Release 1.1.0

Key Changes

🚀 Features

• Introduced REST API MVP for experiment management #47
• Added support for llava-v1.6-mistral-7b model #15
• Enabled auto_stop_method and fms-hf-tuning==3.0.0 in SFTTrainer experiments #27, #42
• Configured ephemeral AIM repository via aim_db=None #24
• Allowed custom sampler class for random_walk operator #26

🛠️ Fixes

• Fixed throughput calculation and early exit logic in SFTTrainer #70, #41
• Ensured correct JSON output #84
• Make sure simulated JSON_CONTAINS works on SQLite #78

📚 Documentation

• Improved developer and contributor instructions #40
• Enhanced documentation for random_walk operator and metastore queries #53, #69

🧱 Build & Dependency Updates

• Removed torch from SFTTrainer dependencies #38
• Switched to hatchling for custom experiment builds #67
• Dropped numpy<2 pinning and refreshed dependencies #28

---

Full (sorted) list of changes

• build: link readme in ado-sfttrainer by @AlessandroPomponio in #12
• build: remove torch from the list of SFTTrainer dependencies by @VassilisVassiliadis in #38
• build: use hatchling in example custom experiments by @michael-johnston in #67
• chore: add funding acknowledgements by @AlessandroPomponio in #51
• chore: Configure Renovate by @renovate[bot] in #1
• chore: update mend configuration by @AlessandroPomponio in #62
• chore: update security reporting by @AlessandroPomponio in #75
• chore: website fixes by @michael-johnston in #19
• chore(deps): do not pin numpy<2 anymore by @AlessandroPomponio in #28
• chore(deps): update dependencies by @AlessandroPomponio in #18
• chore(deps): update dependencies by @AlessandroPomponio in #59
• chore(deps): update dependencies by @AlessandroPomponio in #66
• chore(deps): update dependencies by @AlessandroPomponio in #79
• chore(deps): update dependencies by @AlessandroPomponio in #90
• docs: add acknowledgements by @AlessandroPomponio in #50
• docs: improve docs for the random walk operator by @danielelotito in #53
• docs: make sure developing and contributing instructions are complete by @AlessandroPomponio in #40
• docs: update metastore query docs by @michael-johnston in #69
• feat: add info message if actuator does not have experiments by @AlessandroPomponio in #80
• feat: add support for booleans and null values in sqlite field querying by @AlessandroPomponio in #82
• feat: add support for fms-hf-tuning==3.0.0 in SFTTrainer experiments by @VassilisVassiliadis in #42
• feat: allow specifying custom sampler class for use with random_walk operator by @michael-johnston in #26
• feat: dump default values by default when getting contexts by @AlessandroPomponio in #74
• feat: implement REST API MVP by @AlessandroPomponio in #47
• feat: setting aim_db to None configures SFTTrainer to use an ephemeral AIM repository by @VassilisVassiliadis in #24
• feat: support auto_stop_method for SFTTrainer experiments by @VassilisVassiliadis in #27
• feat: support llava-v1.6-mistral-7b by @VassilisVassiliadis in #15
• fix: calculating the throughput for SFTTrainer experiments by @VassilisVassiliadis in #70
• fix: configuring Trainer to exit a training job early by @VassilisVassiliadis in #41
• fix: ensure ado get -o json works by @AlessandroPomponio in #84
• fix: ensure sample store identifiers cannot be parsed as floats by @AlessandroPomponio in #76
• fix: ensure simulated JSON_CONTAINS works on SQLite by @AlessandroPomponio in #78
• fix: measurement request serialization by @michael-johnston in #56
• fix: measuring properties in the example_actuator by @VassilisVassiliadis in #45
• fix: minor issues by @michael-johnston in #89
• fix: Potential access of unset var on Exception by @michael-johnston in #36
• fix: retrieving the result of an experiment request from the ado REST API by @VassilisVassiliadis in #88
• fix: use correct variable in ado template operation by @michael-johnston in #73
• fix(build): introduce build-system section by @AlessandroPomponio in #14
• fix(docs): fix typos by @danielelotito in #72
• fix(docs): update path for local context by @AlessandroPomponio in #10
• fix(style): apply fixes for RUF059 unused-unpacked-variable by @AlessandroPomponio in #61
• refactor: drop support for ax in the Ray Tune operator by @AlessandroPomponio in #33
• refactor: Property and PropertyValue models by @michael-johnston in #49
• refactor: remove script dependency from vLLM Performance actuator by @srikumar003 in #25
• style: apply ruff's UP linter by @AlessandroPomponio in #17
• style: enable ruff's SIM linter by @AlessandroPomponio in #21
• style: lint markdown files by @AlessandroPomponio in #23
• test: ensure example_actuator is tested in CI by @AlessandroPomponio in #48

New Contributors

• @michael-johnston made their first contribution in #19
• @srikumar003 made their first contribution in #25
• @danielelotito made their first contribution in #53
• @renovate[bot] made their first contribution in #1

Full Changelog: 1.0.1...1.1.0

Release 1.0.1

What's Changed

• build: rename ado-base to ado-core by @AlessandroPomponio in #6
• chore: remove upgrade validators by @AlessandroPomponio in #8

New Contributors

• @AlessandroPomponio made their first contribution in #6

Full Changelog: 1.0.0...1.0.1

Release 1.0.0

Welcome to the first release of ado.

ado is a unified platform for executing computational experiments at scale and analysing their results. It can be extended with new experiments or new analysis tools. It allows distributed teams of researchers and engineers to collaborate on projects, execute experiments, and share data.

You can run the experiments and analysis tools already available in ado in a distributed, shared, environment with your team. You can also use ado to get features like data-tracking, data-sharing, tool integration and a CLI, for your analysis method or experiment for free.