GitLab

Previously processing multiple images in a batch would trigger
segfaults so sending images together was disabled as a way to
mitigate this. The trigger was processing one image on the CPU
and one on the GPU.

This can no longer happen:
 - The vision encoder is now on the GPU so both images would be
   processed on the GPU.
 - We require images to be fully contained in a batch and each
   image including its special tokens is over half the batch size.
   As a result, we will never get two images in the same batch.

Fixes #9731

Currently there is a single context per sequence, shared all by
all multimodal inputs. Since we build a vision encoder graph per
image, with a large number of inputs we can eventually hit the
maximum number of graph nodes per context.

This changes to use a separate context for each image, ensuring
that available resource limits are consistent.

Models may require that a set of inputs all be processed as part
of the same batch. For example, if an image has multiple patches
with fully connected attention between them, we should not split
the batch in the middle of an image.

Fixes #9697

This commit refactors the LLM subsystem by removing internal subprocess
request and response types. It consolidates duplicate type definitions
across the codebase, moving them to centralized locations. The change also
standardizes interfaces between components, simplifies the ServerStatusResp
struct, and moves the ParseDurationMs function to a common package. This
cleanup reduces code duplication between different runner implementations
(llamarunner and ollamarunner).

Replace large-chunk blob downloads with parallel small-chunk
verification to solve timeout and performance issues. Registry users
experienced progressively slowing download speeds as large-chunk
transfers aged, often timing out completely.

The previous approach downloaded blobs in a few large chunks but
required a separate, single-threaded pass to read the entire blob back
from disk for verification after download completion.

This change uses the new chunksums API to fetch many smaller
chunk+digest pairs, allowing concurrent downloads and immediate
verification as each chunk arrives. Chunks are written directly to their
final positions, eliminating the entire separate verification pass.

The result is more reliable downloads that maintain speed throughout the
transfer process and significantly faster overall completion, especially
over unstable connections or with large blobs.

count gemma3 vision tensors

the largest operation is by far (q @ k) so just count that for
simplicity

fix: error if image requested without vision model

Add metadata and tensor information to the show command to be able to
see more information about a model. This outputs the same data as
shown on the model details page on ollama.com

fix: error on models that don't support embeddings

Co-authored-by: Jeffrey Morgan <jmorganca@gmail.com>

ollama-debug.c: correct mistype

* macOS has different definition per info from @mxyng

Softcap isn't in the whitepaper/implementation for the language model so we should remove it. There is no discernible difference in output with it removed.

Signed-off-by: shane.xb.qian <shane.qian@foxmail.com>

Signed-off-by: shane.xb.qian <shane.qian@foxmail.com>

Co-authored-by: Richard Lyons <frob@cloudstaff.com>

engine: add gemma support