[Diff since v0.13.2](v0.13.2...v0.14.0)
- drop support for Julia 1.6
- Bump to Krylov 0.10
- subsolver_type becomes subsolver and take a symbol corresponding to the solver function and not the structure.
**Merged pull requests:**
- Support Krylov.jl v0.10 (#292) (@amontoison)
- Update .cirrus.yml (#296) (@dpo)
- Version 0.14.0 (#300) (@tmigot)
- Add a subsolver for TRON (#301) (@amontoison)
- Fix link in documentation build (#302) (@tmigot)
**Closed issues:**
- `setfield!` Error with `SparseMatrixCOO` in Bound-Constrained Solver Tests (#297)
- Updating Krylov to v0.10.0 breaks JSOSolvers due to module renaming (#298)

[Diff since v0.13.1](v0.13.1...v0.13.2)
**Merged pull requests:**
- Fix R2Solver (#295) (@tmigot)

[Diff since v0.13.0](v0.13.0...v0.13.1)
**Merged pull requests:**
- Specify reset! SolverCore and NLPModels  (#291) (@tmigot)
- Version 0.13.1 (#294) (@tmigot)

[Diff since v0.12.1](v0.12.1...v0.13.0)
Add a new ParameterSet structure to each solver that facilitate documentation and parameter tuning.
**Merged pull requests:**
- Use SolverParameters (#254) (@tmigot)
- Fix preconditioner unit test (#283) (@mertkipcak)
- Update with JSOBestieTemplate for JSOSolvers (#286) (@tmigot)
- Update with JSOBestieTemplate for JSOSolvers (#287) (@tmigot)
- Version 0.13.0 (#289) (@tmigot)
- Update deprecated version of `actions/cache: v1`. (#290) (@tmigot)

[Diff since v0.12.0](v0.12.0...v0.12.1)
**Merged pull requests:**
- Add optional preconditioner for linear solves (#279) (@mpf)
**Closed issues:**
- Integrating a preconditioner into Trunk's trust-region subsolver (#278)

[Diff since v0.11.2](v0.11.2...v0.12.0)
**Merged pull requests:**
- Fomo nonmonotone extension (#267) (@d-monnet)
- Add unbounded objective tests (#268) (@d-monnet)
- Add unbounded optimality (#269) (@tmigot)
- Update Krylov.jl [compat] (#274) (@paraynaud)
- Bump to NLPModels 0.21 (#277) (@tmigot)
**Closed issues:**
- "Error: not a descent direction" when trying to minimize `LinearAlgebra.norm` (!) with `trunk` (#273)

[Diff since v0.11.1](v0.11.1...v0.11.2)
**Merged pull requests:**
- First-order with momentum (#250) (@d-monnet)
- Add references for R2 and LBFGS in readme #236 (#259) (@hpmouton)
- Generalize callback documentation (#263) (@tmigot)
- Fix warnings (#264) (@tmigot)
- Add time limit on Krylov solvers (#265) (@tmigot)
**Closed issues:**
- Add references for R2 and LBFGS in readme (#236)
- Add time limit in Krylov solvers (#237)
- Fix warning conflict on `solve!` (#238)
- R2 doesn't stop when sigma overflow (#245)
- R2 doc does not list callback function as kwarg (#247)
- Issue with the update of NLPModelsModifiers#v0.7.0 (#258)
- Callback documentation (#262)
- Negative predicted reduction in tron/trunk (#266)

[Diff since v0.11.0](v0.11.0...v0.11.1)
**Merged pull requests:**
- [CI] Update .cirrus.yml (#239) (@amontoison)
- [documentation] Update Documenter.jl (#240) (@amontoison)
- Small update to the docs (#241) (@abelsiqueira)
- Removed unnecessary code (#243) (@Jay-sanjay)
- Added warning message (#246) (@Jay-sanjay)
- fix consistency.jl test (#249) (@d-monnet)
- Remove x (#252) (@Jay-sanjay)
- Use CITATION.cff instead of CITATION.bib (#253) (@tmigot)
- replace σ by  1/α  (#255) (@d-monnet)
- CompatHelper: bump compat for NLPModelsModifiers to 0.7, (keep existing compat) (#257) (@github-actions[bot])
**Closed issues:**
- Docs (#20)
- TRON: add a warning if initial guess is not within the bounds (#244)
- Remove `x` in tests (#251)

[Diff since v0.10.0](v0.10.0...v0.11.0)
**Closed issues:**
- Test allocations of each solver (#161)
- Review Krylov.jl usage (#163)
- What does "Error: Cauchy step returned: small_step" mean and what can/should I do about it? (#201)
- Iteration numbers in tron logging are offset by one. (#207)
- max_iter kwarg not supported (R2.jl) (#213)
**Merged pull requests:**
- Use in-place `cg` in TRON (#202) (@tmigot)
- Save allocations in projected line search (#204) (@tmigot)
- Save allocations with active indices (#205) (@tmigot)
- Save allocations from Hessian operator (#206) (@tmigot)
- Fix display TRON (#208) (@tmigot)
- Pre-allocate ZHZ operator (#209) (@tmigot)
- Save allocations computing `-g` (#210) (@tmigot)
- Remove subsolver logger (#211) (@tmigot)
- Use Krylov.cg! (#215) (@tmigot)
- Remove allocations in TRON (#218) (@tmigot)
- Save allocations in TRON-NLS (#220) (@tmigot)
- Save allocations in `cauchy_ls` for TRON-NLS (#221) (@tmigot)
- Save allocations in `projected_line_search_ls!` for TRON-NLS (#222) (@tmigot)
- Resize projected newton subsolver (#223) (@tmigot)
- Add allocation test for TRON and TRUNK NLS (#225) (@tmigot)
- Use NLS variant for `aredpred!` (#226) (@tmigot)
- Add allocation test for bound-constrained problems (#227) (@tmigot)
- Update R2.jl (#228) (@farhadrclass)
- Review logs (#230) (@tmigot)
- Add subsolver verbose in TRON (#232) (@tmigot)
- Remove @error macro (#234) (@tmigot)
- Version 0.11 (#235) (@tmigot)

[Diff since v0.9.3](v0.9.3...v0.10.0)
**Closed issues:**
- Stop criteria for NLS solvers: residual is sufficiently small (#34)
- TRON appears to not be using its internal TrustRegion (#181)
- Add `max_iter` stopping criterion (#184)
- Use `obj` to compute the objective function for NLS (#194)
**Merged pull requests:**
- Use `objgrad!` at initialization of `lbfgs` (#180) (@tmigot)
- Use `max_radius` in TRON (#182) (@tmigot)
- Use test problems in allocation tests (#185) (@tmigot)
- Add allocation tests for R2 (#186) (@tmigot)
- Add small residual check (#187) (@tmigot)
- Add `max_iter` in all solvers (#188) (@tmigot)
- Add `trunk` in 0-allocs tests (#190) (@tmigot)
- Save some allocations in `tron` (#191) (@tmigot)
- Save some allocations in Trunk-LS (#192) (@tmigot)
- Remove allocations in trunk-nls (#193) (@tmigot)
- Avoid using `nls.F` (#195) (@tmigot)
- Use obj and grad! in NLS solvers with NLPModels 0.20 (#197) (@tmigot)
- Use `NLPModels` parametric type (#198) (@tmigot)
- Version 0.10.0 (#200) (@tmigot)