Merge pull request#9897 from ollama/mxyng/chunk-loadml/backend/ggml: load tensors in 128KiB chunks

Merge pull request#9824 from ollama/mxyng/schedconditionally enable parallel pipelines

Merge pull request#9824 from ollama/mxyng/schedconditionally enable parallel pipelines

server/internal/chunks: remove chunks package (#9755)

Merge pull request#9703 from ollama/mxyng/gemma3-memorycount gemma3 vision tensors

Merge pull request#9661 from ollama/gemmaengine: add gemma support

Merge pull request#9661 from ollama/gemmaengine: add gemma support

server/internal/registry: reintroduce pruning on model deletion (#9489)This reintroduces aggressive pruning on model deletion as a temporarymeasure until a more controlled garbage collection (GC) mechanism isimplemented.Issues with the current approach:1. Users may accidentally delete a model (`ollama rm llama3.3` instead   of `ollama rm llama3.2`), requiring a full re-download unless another   model references the same blobs.2. Users may assume a deleted model is still referenced elsewhere, but   due to prior updates or deletions, the references no longer exist,   leading to unnecessary re-downloads.Soon, we should implement a structured GC mechanism to retainunreferenced blobs for a configurable period before removal, which willrun on "ollama rm" and other commands we deem appropriate.Users that want to immediately remove unreferenced blobs can use a newprune command that will allow them to specify the age and class of blobsto remove.Example usage:    # Run basic blob GC    $ ollama prune    # Remove unreferenced blobs older than 7 days    $ ollama prune --age 7d    # Remove all blobs, referenced or not, older than 7 days (and their manifests?)    $ ollama prune --age 7d --all    # Remove all unreferenced blobs immediately    $ ollama prune --age 0 --all    # Remove all blobs    $ ollama prune --age 0 --allThis should provide a safer and more predictable cleanup process.

build: install binutils alongside gcc in Dockerfile (#9475)

build: install ccache manually in Dockerfile (#9464)Reverts ccache installation to be done manually via curl instead ofusing the dnf package manager as this has side effects of prependingccache's install directory to the front of the PATH