Daniel Bevenius [Fri, 14 Feb 2025 10:16:56 +0000 (11:16 +0100)]
llama : add completion for --chat-template-file (#11860)
This commit adds completion for `--chat-template-file`, enabling only
`.jinja` files to be displayed as completions.
Example usage:
```console
$ ./build/bin/llama-cli --chat-template-file models/templates/<TAB>
models/templates/CohereForAI-c4ai-command-r7b-12-2024-tool_use.jinja
models/templates/CohereForAI-c4ai-command-r-plus-tool_use.jinja
models/templates/deepseek-ai-DeepSeek-R1-Distill-Llama-8B.jinja
models/templates/deepseek-ai-DeepSeek-R1-Distill-Qwen-32B.jinja
models/templates/fireworks-ai-llama-3-firefunction-v2.jinja
models/templates/google-gemma-2-2b-it.jinja
models/templates/llama-cpp-deepseek-r1.jinja
models/templates/meetkai-functionary-medium-v3.1.jinja
models/templates/meetkai-functionary-medium-v3.2.jinja
models/templates/meta-llama-Llama-3.1-8B-Instruct.jinja
models/templates/meta-llama-Llama-3.2-3B-Instruct.jinja
models/templates/meta-llama-Llama-3.3-70B-Instruct.jinja
models/templates/microsoft-Phi-3.5-mini-instruct.jinja
models/templates/mistralai-Mistral-Nemo-Instruct-2407.jinja
models/templates/NousResearch-Hermes-2-Pro-Llama-3-8B-tool_use.jinja
models/templates/NousResearch-Hermes-3-Llama-3.1-8B-tool_use.jinja
models/templates/Qwen-Qwen2.5-7B-Instruct.jinja
```
This is not limited to the models/templates directory, it can be used
anywhere in the filesystem, the above is just an example.
Daniel Bevenius [Tue, 11 Feb 2025 13:06:45 +0000 (14:06 +0100)]
server : use common_token_to_piece instead of common_detokenize (#11740)
* server : use common_token_to_piece instead of common_detokenize
This commit replaces the call to common_detokenize with
common_token_to_piece in the populate_token_probs.
The motivation for this change is to avoid an issue where
common_detokenize would remove the word boundary character for tokens,
which caused a regression in the server generated token probabilities.
Wilken Gottwalt [Mon, 10 Feb 2025 18:58:18 +0000 (19:58 +0100)]
llama-mmap: fix missing include (#11796)
Technically the fixed width types come only from iostream and
cstdint/stdint.h headers. memory and vector headers should not provide
these. In GCC 15 the headers are cleaned up and you require the proper
header cstdint.
src/llama-mmap.h:26:5: error: ‘uint32_t’ does not name a type
26 | uint32_t read_u32() const;
| ^~~~~~~~
Karol Kontny [Sat, 8 Feb 2025 14:30:53 +0000 (15:30 +0100)]
ggml: Fix data race in ggml threadpool (#11736)
After the barrier in last iteration is executed, still the loop termination
condition will be executed. However main thread can destroy the cgraph object
and its nodes already, then another thread will access it, but the thing is already gone.
Also trouble can happen when n_nodes == 0 or abort is called, but I'm not sure if the
prior situation is possible.
Last syncronization should be done after the loop to ensure the cgraph/cplan won't be
accessed after the main thread exits from the function.
Christian Fillion [Fri, 7 Feb 2025 13:55:47 +0000 (08:55 -0500)]
vocab : ignore invalid UTF-8 input in the BPE tokenizer (#11729)
Silently insert U+FFFD(s) (Unicode replacement character) instead until the
next valid codepoint can be found.
This fixes `llama_tokenize` throwing an exception across the C API boundary
or libllama's module boundary (the caller's runtime might be incompatible!)
Returing a proper error code might be desirable, however the signature
of `llama_tokenize` doesn't allow it as all return values already have
existing meaning.
Christian Fillion [Fri, 7 Feb 2025 09:33:27 +0000 (04:33 -0500)]
llama : add llama_sampler_init for safe usage of llama_sampler_free (#11727)
The C API in llama.h claims users can implement `llama_sampler_i` to
create custom `llama_sampler`. The sampler chain takes ownership and
calls `llama_sampler_free` on them. However, `llama_sampler_free` is
hard-coded to use `delete`. This is undefined behavior if the object
wasn't also allocated via `new` from libllama's C++ runtime. Callers
in C and C-compatible languages do not use C++'s `new` operator. C++
callers may not be sharing the same heap as libllama.
Daniel Bevenius [Fri, 7 Feb 2025 08:15:22 +0000 (09:15 +0100)]
common : add default embeddings presets (#11677)
* common : add default embeddings presets
This commit adds default embeddings presets for the following models:
- bge-small-en-v1.5
- e5-small-v2
- gte-small
These can be used with llama-embedding and llama-server.
For example, with llama-embedding:
```console
./build/bin/llama-embedding --embd-gte-small-default -p "Hello, how are you?"
```
And with llama-server:
```console
./build/bin/llama-server --embd-gte-small-default
```
And the embeddings endpoint can then be called with a POST request:
```console
curl --request POST \
--url http://localhost:8080/embeddings \
--header "Content-Type: application/json" \
--data '{"input": "Hello, how are you?"}'
```
I'm not sure if these are the most common embedding models but hopefully
this can be a good starting point for discussion and further
improvements.
Patrick Peng [Thu, 6 Feb 2025 14:29:13 +0000 (09:29 -0500)]
rpc: fix known RCE in rpc-server (ggml/1103)
Add bounds checking in `rpc_server::copy_tensor` to prevent out-of-bounds writes
+ Check if `(uint8_t *)dst->data + ggml_nbytes(src)` remains within the destination buffer’s allocated region.
Matvey Soloviev [Thu, 6 Feb 2025 00:55:25 +0000 (01:55 +0100)]
readme : add link to Autopen under UIs (#11684)
Autopen (https://github.com/blackhole89/autopen) is a graphical text editor that uses llama.cpp to tokenize the buffer on the fly, score the buffer, visualise token logits and allow you to switch back and forth between different possible completions at any point. It hopefully meets the criteria for inclusion, as the dependency on llama.cpp is stated prominently.
Shelby Jenkins [Tue, 4 Feb 2025 11:20:55 +0000 (05:20 -0600)]
readme : add llm_client Rust crate to readme bindings (#11628)
[This crate](https://github.com/ShelbyJenkins/llm_client) has been in a usable state for quite awhile, so I figured now is fair to add it.
It installs from crates.io, and automatically downloads the llama.cpp repo and builds it for the target platform - with the goal being the easiest user experience possible.
It also integrates model presets and choosing the largest quant given the target's available VRAM. So a user just has to specify one of the presets (I manually add the most popular models), and it will download from hugging face.
So, it's like a Rust Ollama, but it's not really for chatting. It makes heavy use of llama.cpp's grammar system to do structured output for decision making and control flow tasks.
Christian Kastner [Mon, 3 Feb 2025 23:17:15 +0000 (00:17 +0100)]
cmake: Add ability to pass in GGML_BUILD_NUMBER (ggml/1096)
This makes git as a dependency optional, and is useful in the case where
ggml is built not from git, but from a tarball, or a distribution source
package.
This conditional also affects GGML_BUILD_COMMIT. Nothing seems to be
using it, though, so there doesn't seem much value factor it out, or
even require it.
Daniel Bevenius [Mon, 3 Feb 2025 15:45:38 +0000 (16:45 +0100)]
server : remove CPPHTTPLIB_NO_EXCEPTIONS define (#11622)
This commit removes the CPPHTTPLIB_NO_EXCEPTIONS define from the server
code.
The motivation for this is that when using a debug build the server
would crash when an exception was throws and terminate the server
process, as it was unhandled. When CPPHTTPLIB_NO_EXCEPTIONS is set
cpp_httplib will not call the exception handler, which would normally
return a 500 error to the client. This caused tests to fail when using
a debug build.
overview / pkg / ggml / sources / llama.cpp / log