llama_cpp 0.17.10 → 0.18.1

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: ec7b41006345cfe57e355d207718ae460a5a2c519d67027daddb7d40ab346754
-  data.tar.gz: efcfa7eef5374790cef788f719bede1ab3cd30e5dec79be0ba9844bfbcd433c2
+  metadata.gz: 66599243db8e088ab250be615f142cc0c3d63ee0f7661e8300dbd569fd100ac5
+  data.tar.gz: 2bfdbc95133df7939276547dbae237c5ccd2175a1a8cee2e81cc3579bf3fa00a
 SHA512:
-  metadata.gz: c39150e77652060fcfee2c73b1334ff53ed40caf58679efa727a0d0d54bffe036b8795be9751cef8517b5d4b2300fb77fa5753a9c94b803988ec429b999591c2
-  data.tar.gz: 6054dee2f39ac2690de2836405ac793bf987538ef0780cffab2700544bc60ceaca8a4b109407c66dcf2dd9f38ffe6b98659d7ca4ddb5c9e93fa85d9884cea26e
+  metadata.gz: 365cad9872ed2601700de4a4a4fc7ef5a13c307e4b8ba199f0b3b2c567dd252e402661fd9161f663c3b5dd2fff70f2c935e333b904a0923de0e8aa9955319b63
+  data.tar.gz: 4b5275081819595983bb9681951b7c25e1764e7039951368624fc8dc5bdbee9848136f0f98624f1cc07575a76d09e63bf3b2f88effd1ea1c0edd098dd0bb050f

data/CHANGELOG.md

@@ -1,3 +1,19 @@
+
+## [[0.18.1](https://github.com/yoshoku/llama_cpp.rb/compare/v0.18.0...v0.18.1)] - 2025-02-15
+
+- Change supported llama.cpp version to b4713
+  - Add `llama_sampler_init_top_n_sigma` module function.
+- Remove old type declaration file.
+
+## [[0.18.0](https://github.com/yoshoku/llama_cpp.rb/compare/v0.17.10...v0.18.0)] - 2025-02-02
+
+**Breaking Changes**
+
+All the native extensions code was rewritten in C. The high-level API has been removed and replaced with a simple bindings library.
+The fast update speed of llama.cpp makes it difficult to keep up with the creation of this binding library.
+[As previously noted](https://github.com/yoshoku/llama_cpp.rb/blob/main/CHANGELOG.md#060---2023-09-30),
+the author has given up on continuing to develop this binding library. Thank you for your understanding.
+
 ## [[0.17.10](https://github.com/yoshoku/llama_cpp.rb/compare/v0.17.9...v0.17.10)] - 2024-09-07
 
 - Change supported llama.cpp version to b3676.

data/LICENSE.txt

@@ -1,6 +1,6 @@
 The MIT License (MIT)
 
-Copyright (c) 2023-2024 Atsushi Tatsuma
+Copyright (c) 2023-2025 Atsushi Tatsuma
 
 Permission is hereby granted, free of charge, to any person obtaining a copy
 of this software and associated documentation files (the "Software"), to deal

data/README.md

@@ -6,8 +6,6 @@
 
 llama_cpp.rb provides Ruby bindings for the [llama.cpp](https://github.com/ggerganov/llama.cpp).
 
-This gem is still under development and may undergo many changes in the future.
-
 ## Installation
 
 Install the llama.cpp. If you use homebrew, install it by executing:
@@ -57,41 +55,20 @@ An example of Ruby code that generates sentences with the quantization model is
 ```ruby
 require 'llama_cpp'
 
-model_params = LLaMACpp::ModelParams.new
-model = LLaMACpp::Model.new(model_path: '/home/user/llama.cpp/models/open_llama_7b/ggml-model-q4_0.bin', params: model_params)
-
-context_params = LLaMACpp::ContextParams.new
-context_params.seed = 42
-context = LLaMACpp::Context.new(model: model, params: context_params)
+LlamaCpp.ggml_backend_load_all
 
-puts LLaMACpp.generate(context, 'Hello, World.')
-```
+model_params = LlamaCpp::LlamaModelParams.new
+model = LlamaCpp::llama_model_load_from_file('/home/user/llama.cpp/models/open_llama_7b/ggml-model-q4_0.bin', model_params)
 
-## Examples
-There is a sample program in the [examples](https://github.com/yoshoku/llama_cpp.rb/tree/main/examples) directory that allow interactvie communication like ChatGPT.
+context_params = LlamaCpp::LlamaContextParams.new
+context = LlamaCpp.llama_init_from_model(model, context_params)
 
-```sh
-$ git clone https://github.com/yoshoku/llama_cpp.rb.git
-$ cd examples
-$ bundle install
-$ ruby chat.rb --model /home/user/llama.cpp/models/open_llama_7b/ggml-model-q4_0.bin --seed 2023
-...
-User: Who is the originator of the Ruby programming language?
-Bob: The originator of the Ruby programming language is Mr. Yukihiro Matsumoto.
-User:
-```
-
-![llama_cpp_chat_example](https://github.com/yoshoku/llama_cpp.rb/assets/5562409/374ae3d8-63a6-498f-ae6e-5552b464bdda)
-
-Japanse chat is also possible using the [Vicuna model on Hugging Face](https://huggingface.co/CRD716/ggml-vicuna-1.1-quantized).
+puts LLaMACpp.generate(context, 'Hello, World.')
 
-```sh
-$ wget https://huggingface.co/CRD716/ggml-vicuna-1.1-quantized/resolve/main/ggml-vicuna-7b-1.1-q4_0.bin
-$ ruby chat.rb --model ggml-vicuna-7b-1.1-q4_0.bin --file prompt_jp.txt
+LlamaCpp.llama_free(context)
+LlamaCpp.llama_model_free(model)
 ```
 
-![llama_cpp rb-jpchat](https://github.com/yoshoku/llama_cpp.rb/assets/5562409/526ff18c-2bb2-4b06-8933-f72960024033)
-
 ## Contributing
 
 Bug reports and pull requests are welcome on GitHub at https://github.com/yoshoku/llama_cpp.rb.

data/ext/llama_cpp/extconf.rb

@@ -2,10 +2,7 @@
 
 require 'mkmf'
 
-abort('libstdc++ is not found.') unless have_library('stdc++')
 abort('libllama is not found.') unless have_library('llama')
 abort('llama.h is not found.') unless have_header('llama.h')
 
-$CXXFLAGS << ' -std=c++11'
-
 create_makefile('llama_cpp/llama_cpp')