The following instructions refer to gemma-3-1b-it-qat-q4_0 and Ubuntu Linux 24.04.3 LTS but can be applied to other
Gemma 3 IT QAT checkpoints as well.
To quantize the original safetensors snapshot, we need llama.cpp and a few helper tools that can be obtained together
with llama.cpp. To make sure we are on the latest version, we first clone and build the newest version of llama.cpp.
sudo apt-get install -y --no-install-recommends \
With latest release of Spring Cloud 2020.0 (aka Ilford) it's now easier than ever to setup Zookeeper configuration support, but the official documentation currnently lacks some straight-forward example.
Add the spring-cloud-dependencies bom to the dependencyManagement section of you pom.xml
<dependencyManagement>
Assume you have a Spring Boot application which exposes multiple apis, for example one for dogs and one for cats. Let's further assume that the api base paths look something like this
/api/cats/v1//api/dogs/v1/Especially within kubernetes, it would be slick to be able to expose these apis using dedicated services. Let's recap a simple kubernetes service definition.
Annotating a test with @SpringBootTest will usally lead to initializing a Spring context with all components found
by Springs component scan mechanism (usally all components of your app). Sometimes, you want some components to be excluded
from component scan within your tests, for example
Integrate JMH (Java Microbenchmarking Harness) with Spring (Boot) and make developing and running benchmarks as easy and convinent as writing tests.
Wrap the necessary JMH boilerplate code within JUnit to benefit from all the existing test infrastructure Spring (Boot) provides. It should be as easy and convinent to write benchmarks as it is to write tests.
jooq does not support Postgres nativ json(b) datatypes out of the box. It would be nice have jooq recognize json columns and provide them in a generic way within generated records.
The official jooq documentation provides an example for creating a custom data type binding for Postgres json columns using GSON. Based on this, it is possible to tweak the implemenation to
If you are writing (integration) tests and you want to debug code within the scope of a method annotated with @Async
(e.g. an event handler) your IDE might skip over breakpoints within the async code because it might be executed on a
different/new thread.
Replace the default TaskExecutor with an instance of SyncTaskExecutor to ensure, that everything runs on the same thread.
In order to run jooqs codegen, it needs an appropriate database (with the correct schema) to be running. This often leads to checking in jooq generated code into your applications repository to ease the build process.
jooq provides an option to run the code generator based on flyway migrations without the need for a specific database to be running. The feature is described here as codegen from DDL.
| The idea is, to have models adhere to a convention, which stats, that there has to be a field "searchable_tokens", | |
| which includes searchable tokens. Now, we can simply have a (virtual) model/table `search_results`, which can be | |
| unioned with "casted selects" of each model we want to search. The casted selects make the columns of each model to fit | |
| into the schema of the search result. | |
| Mixed with ActiveRecord's polymorphic associations, the result is an ActiveRecord relation with elements, which points to | |
| their respective subject. |
| sudo mkdir /etc/systemd/system/docker.service.d | |
| sudo bash -c 'cat <<EOF > /etc/systemd/system/docker.service.d/overlay.conf | |
| [Service] | |
| ExecStart= | |
| ExecStart=/usr/bin/docker daemon -H fd:// --storage-driver=overlay | |
| EOF' | |
| sudo systemctl daemon-reload | |
| sudo systemctl restart docker |