If there are any pending reponses (such as from potential stop
tokens) then we should send them back before ending the sequence.
Otherwise, we can be missing tokens at the end of a response.
Fixes#6707
Provide a mechanism for users to set aside an amount of VRAM on each GPU
to make room for other applications they want to start after Ollama, or workaround
memory prediction bugs
/Users/au/src/ollama/llm/ext_server/server.cpp:289:9: warning: 'sprintf' is deprecated: This function is provided for compatibility reasons only. Due to security concerns inherent in the design of sprintf(3), it is highly recommended that you use snprintf(3) instead.
* Fix embeddings memory corruption
The patch was leading to a buffer overrun corruption. Once removed though, parallism
in server.cpp lead to hitting an assert due to slot/seq IDs being >= token count. To
work around this, only use slot 0 for embeddings.
* Fix embed integration test assumption
The token eval count has changed with recent llama.cpp bumps (0.3.5+)
We're over budget for github's maximum release artifact size with rocm + 2 cuda
versions. This splits rocm back out as a discrete artifact, but keeps the layout so it can
be extracted into the same location as the main bundle.
This adjusts linux to follow a similar model to windows with a discrete archive
(zip/tgz) to cary the primary executable, and dependent libraries. Runners are
still carried as payloads inside the main binary
Darwin retain the payload model where the go binary is fully self contained.
For simplicity, perform parallelization of embedding requests in the API handler instead of offloading this to the subprocess runner. This keeps the scheduling story simpler as it builds on existing parallel requests, similar to existing text completion functionality.
Don't allow loading models that would lead to memory exhaustion (across vram, system memory and disk paging). This check was already applied on Linux but should also be applied on Windows as well.
If the system has multiple numa nodes, enable numa support in llama.cpp
If we detect numactl in the path, use that, else use the basic "distribute" mode.
Make sure if something goes wrong spawning the process, the user gets
enough info to be able to try to self correct, or at least file a bug
with details so we can fix it. Once the process starts, we immediately
change back to the recommended setting to prevent the blocking dialog.
This ensures if the model fails to load (OOM, unsupported model type,
etc.) the process will exit quickly and we can scan the stdout/stderr
of the subprocess for the reason to report via API.
