v0.3.0

What's Changed

• Move man page README.md to full links by @rhatdan in #483
• Allow users to set ctx-size via command line by @rhatdan in #484
• Add --seed option by @rhatdan in #485
• Update install.sh by @jtligon in #493
• Take in fixes from @Churchyard to modernize spec file by @smooge in #494
• Fix up building and pushing OCI Images by @rhatdan in #492
• Fix handling of file_not_found errors by @rhatdan in #499
• Updated nv docs to align with latest WSL2 cuda setup by @bmahabirbu in #503
• Add ramalama convert command by @rhatdan in #500
• Stop checking if command is running in container by @rhatdan in #505
• Add initial CONTRIBUTING.md file by @rhatdan in #507
• Place image name just before command by @ericcurtin in #511
• Simplify install by @ericcurtin in #510
• Fix handling of README.md in docs directory by @rhatdan in #512
• Add installation steps for Podman 5 in CI workflows by @ericcurtin in #508
• Bump to v0.3.0 by @rhatdan in #513

New Contributors

• @jtligon made their first contribution in #493

Full Changelog: v0.2.0...v0.3.0

v0.2.0

What's Changed

• Packit: enable fedora downstream jobs by @lsm5 in #480
• Bump to v0.2.0 by @rhatdan in #481
  Failover to OCI when push fails with default push mechanism by @rhatdan in #476
  Fall back to huggingface-cli when pulling via URL fails by @rhatdan in #475
  Full Changelog: v0.1.3...v0.2.0

v0.1.3

What's Changed

• Enable GCC Toolet 12 to support AVX VNNI by @nzwulfin in #473
• Failover to OCI when push fails with default push mechanism by @rhatdan in #476
• Fall back to huggingface-cli when pulling via URL fails by @rhatdan in #475
• Revert "Switch to llama-simple-chat" by @rhatdan in #477
• Add support for http, https and file pulls by @rhatdan in #463
• Bump to v0.1.3 by @rhatdan in #479

New Contributors

• @nzwulfin made their first contribution in #473

Full Changelog: v0.1.2...v0.1.3

v0.1.2

What's Changed

• Bump to v0.1.1 by @rhatdan in #450
• Update ggerganov/whisper.cpp digest to f19463e by @renovate in #453
• Switch to llama-simple-chat by @ericcurtin in #454
• Simplify container image build by @ericcurtin in #451
• Update ggerganov/whisper.cpp digest to 83ac284 by @renovate in #455
• cli.py: remove errant slash preventing the loading of user conf file(s) by @FNGarvin in #457
• Update ggerganov/whisper.cpp digest to f02b40b by @renovate in #456
• Switched DGGML_CUDA to ON in cuda containerfile by @bmahabirbu in #459
• Update ggerganov/whisper.cpp digest to bb12cd9 by @renovate in #460
• Update ggerganov/whisper.cpp digest to 01d3bd7 by @renovate in #461
• Update ggerganov/whisper.cpp digest to d24f981 by @renovate in #462
• Docu by @atarlov in #464
• Update ggerganov/whisper.cpp digest to 6266a9f by @renovate in #466
• Fix handling of ramalama login huggingface by @rhatdan in #467
• Support huggingface-cli older than 0.25.0, like on Fedora 40 and 41 by @debarshiray in #468
• Bump to v0.1.2 by @rhatdan in #470

New Contributors

• @FNGarvin made their first contribution in #457
• @atarlov made their first contribution in #464
• @debarshiray made their first contribution in #468

Full Changelog: v0.1.1...v0.1.2

v0.1.1

Full Changelog: v0.1.0...v0.1.1

Mainly to fix issue in PyPi

v0.1.0

What's Changed

• We can now run models via Kompute in podman-machine by @ericcurtin in #440
• Only do dnf install for cuda images by @ericcurtin in #441
• Add --host=0.0.0.0 if running llama.cpp serve within a container by @rhatdan in #444
• Document the host flag in ramalama.conf file by @rhatdan in #447
• Add granite-8b to shortnames.conf by @rhatdan in #448
• Fix RamaLama container image build by @ericcurtin in #446
• Bump to v0.1.0 by @rhatdan in #449

Full Changelog: v0.0.23...v0.1.0

v0.0.23

What's Changed

• Remove omlmd as a dependency by @ericcurtin in #428
• Check versions match in CI by @ericcurtin in #427
• Fix podman run oci://... by @rhatdan in #429
• Attempt to remove OCI Image if removing as Ollama or Huggingface fails by @rhatdan in #432
• Run does not have generate, so remove it by @rhatdan in #434
• Run the command by default without stderr by @rhatdan in #436
• Closing stderr on podman command is blocking progress information and… by @rhatdan in #438
• Make it easier to test-run manually by @rhatdan in #435
• Install llama-cpp-python[server] by @ericcurtin in #430

Full Changelog: v0.0.22...v0.0.23

v0.0.22

What's Changed

• Bump to v0.0.21 by @rhatdan in #410
• Update ggerganov/whisper.cpp digest to 0377596 by @renovate in #409
• Use subpath for OCI Models by @rhatdan in #411
• Consistency changes by @ericcurtin in #408
• Split out kube.py from model.py by @rhatdan in #412
• Fix mounting of Ollama AI Images into containers. by @rhatdan in #414
• Start an Asahi version by @ericcurtin in #369
• Generate MODEL.yaml file locally rather then just to stdout by @rhatdan in #416
• Bugfix comma by @ericcurtin in #421
• Fix nocontainer mode by @rhatdan in #419
• Update ggerganov/whisper.cpp digest to 31aea56 by @renovate in #425
• Add --generate quadlet/kube to create quadlet and kube.yaml by @rhatdan in #423
• Allow default port to be specified in ramalama.conf file by @rhatdan in #424
• Made run and serve consistent with model exec path. Fixes issue #413 by @bmahabirbu in #426
• Bump to v0.0.22 by @rhatdan in #415

Full Changelog: v0.0.21...v0.0.22

v0.0.21

What's Changed

• Fix rpm build by @rhatdan in #350
• Add environment variables for checksums to ramalama container by @rhatdan in #355
• Change default container name for ROCm container image by @ericcurtin in #360
• Allow removal of models specified as shortnames by @rhatdan in #357
• Added a check to the zsh completions generation step by @ericcurtin in #356
• Add vulkan image and show size by @ericcurtin in #353
• Update ggerganov/whisper.cpp digest to 0fbaac9 by @renovate in #363
• Allow pushing of oci images by @rhatdan in #358
• Fix Makefile to be less stringent on failues of zsh by @smooge in #368
• Add support for --authfile and --tls-verify for login by @rhatdan in #364
• Fix incompatible Ollama paths by @swarajpande5 in #370
• Fix shortname paths by @swarajpande5 in #372
• Change to None instead of "" by @ericcurtin in #371
• Kompute build is warning it is missing this package by @ericcurtin in #366
• Add --debug option to show exec_cmd and run_cmd commands by @rhatdan in #373
• Add support for pushing a file into an OCI Model image by @rhatdan in #374
• Replace huggingface-cli download command with simple https client to pull models by @swarajpande5 in #375
• Update registry.access.redhat.com/ubi9/ubi Docker tag to v9.4-1214.1729773476 by @renovate in #380
• Update ggerganov/whisper.cpp digest to c0ea41f by @renovate in #381
• Update ggerganov/whisper.cpp digest to fc49ee4 by @renovate in #382
• Update dependency huggingface/huggingface_hub to v0.26.2 by @renovate in #383
• Update dependency tqdm/tqdm to v4.66.6 - autoclosed by @renovate in #385
• Update ggerganov/whisper.cpp digest to 1626b73 by @renovate in #386
• Support listing and removing newly designed bundled images by @rhatdan in #378
• Fix default conman check by @rhatdan in #389
• Drop in config by @ericcurtin in #379
• Update ggerganov/whisper.cpp digest to 55e4221 by @renovate in #390
• Move run_container to model.py allowing models types to override by @rhatdan in #388
• Update ggerganov/whisper.cpp digest to 19dca2b by @renovate in #392
• Add man page information for ramalama.conf by @rhatdan in #391
• More debug info by @ericcurtin in #394
• Make transport use config by @rhatdan in #395
• Enable containers on macOS to use the GPU by @slp in #397
• chore(deps): update ggerganov/whisper.cpp digest to 4e10afb by @renovate in #398
• Time for removal of huggingface_hub dependancy by @ericcurtin in #400
• Mount model. car volumes into container by @rhatdan in #396
• Remove huggingface-hub references from spec file by @ericcurtin in #401
• Packit: disable osh diff scan by @lsm5 in #403
• Make minimal change to allow for ramalama to build on EL9 by @smooge in #404
• reduced the size of the nvidia containerfile by @bmahabirbu in #407
• Move /run/model to /mnt/models to match k8s model.car definiton by @rhatdan in #402
• Verify pyproject.py and setup.py have same version by @rhatdan in #405
• Make quadlets work with OCI images by @rhatdan in #406

Full Changelog: v0.0.20...v0.0.21

v0.0.20

What's Changed

• Add support for testing with docker by @rhatdan in #320
• chore(deps): update ggerganov/whisper.cpp digest to d3f7137 by @renovate in #321
• Make changes to spec file to better pass Fedora packaging guidelines by @smooge in #318
• Fix erroneous output in CUDA containerfile by @bmahabirbu in #322
• chore(deps): update ggerganov/whisper.cpp digest to a5abfe6 by @renovate in #323
• There's many cases where macOS support is broken by @ericcurtin in #325
• chore(deps): update dependency huggingface/huggingface_hub to v0.26.0 by @renovate in #328
• State Containerfile is available but not built and pushed by @ericcurtin in #329
• split build from validate in Makefile by @rhatdan in #326
• Remove duplicate GitHub Actions workflow runs in PRs by @p5 in #330
• Add hf:// as an alias to huggingface:// by @ericcurtin in #324
• Break up tests: build, bats, bats-nocontainer, docker, mac-nocontainer by @rhatdan in #331
• Update llama.cpp to fix granite3-moe models by @ericcurtin in #340
• Make sure we specify bash here by @ericcurtin in #337
• If installed in /usr/local , ramalama libs cannot be found by @ericcurtin in #333
• Kompute Containerfile by @ericcurtin in #334
• Add kubernetes.YAML support to ramalama serve by @rhatdan in #327
• Add AI Lab models to shortnames by @MichaelClifford in #345
• Build container images only on changes by @p5 in #332
• Make more spec changes to match RPM evaluation by @smooge in #347
• Free up space for docker tests. by @rhatdan in #343
• Allow the removal of one then one models via rm command by @rhatdan in #344
• Fix spelling mistakes in markdown by @rhatdan in #348
• Bump to v0.0.20 by @rhatdan in #349

New Contributors

• @smooge made their first contribution in #318
• @MichaelClifford made their first contribution in #345

Full Changelog: v0.0.19...v0.0.20