GitLab

* fix: queued request failures

- increase parallel requests to 2 to complete queued request, queueing is managed in ollama

* log steam errors

This reverts commit 7a0899d6.

- update chat docs
- add messages chat endpoint
- remove deprecated context and template generate parameters from docs
- context and template are still supported for the time being and will continue to work as expected
- add partial response to chat history

* Support cuda build in Windows
* Enable dynamic NumGPU allocation for Windows

When CUDA peer access is enabled, multi-gpu inference will produce
garbage output. This is a known bug of llama.cpp (or nvidia). Until the
upstream bug is fixed, we can disable CUDA peer access temporarily
to ensure correct output.

See #961.

* add `"format": "json"` as an API parameter
---------
Co-authored-by: Bruce MacDonald <brucewmacdonald@gmail.com>

instead of static number of parameters for each model family, get the real number from the tensors (#1022)

* parse tensor info

* refactor decoder

* return actual parameter count

* explicit rounding

* s/Human/HumanNumber/

pin to 9e70cc03229df19ca2d28ce23cc817198f897278 for now since
438c2ca83045a00ef244093d27e9ed41a8cb4ea9 is breaking

ggufv3 adds support for big endianness, mainly for s390x architecture.
while that's not currently supported for ollama, the change is simple.

loosen version check to be more forward compatible. unless specified,
gguf versions other v1 will be decoded into v2.