[Refactor] Change the project/repo name to EmbedKV/embed-kv

CMakeLists.txt

@@ -57,9 +57,9 @@ link_directories(
 file(GLOB_RECURSE merlin_hkvs_src RELATIVE ${CMAKE_CURRENT_SOURCE_DIR} *.cpp *.cu)
 
 # TODO:
-# add_library(merlin_kv STATIC ${merlin_kv_src})
-# target_compile_features(merlin_kv PUBLIC cxx_std_14)
-# target_link_libraries(merlin_kv PUBLIC ...)
+# add_library(embed_kv STATIC ${embed_kv_src})
+# target_compile_features(embed_kv PUBLIC cxx_std_14)
+# target_link_libraries(embed_kv PUBLIC ...)
 
 
 add_executable(merlin_hashtable_benchmark benchmark/merlin_hashtable_benchmark.cc.cu)

CONTRIBUTING.md

@@ -1,21 +1,21 @@
 # Contributing
 
-## About Merlin-KV
+## About EmbedKV
 
-Merlin-KV is a part of NVIDIA Merlin and provides hierarchical key-value storage to meet RecSys requirements.
+EmbedKV is a part of NVIDIA Merlin and provides hierarchical key-value storage to meet RecSys requirements.
 
-The key capability of Merlin-KV is to store key-value (feature-embedding) on high-bandwidth memory (HBM) of GPUs and in host memory.
+The key capability of EmbedKV is to store key-value (feature-embedding) on high-bandwidth memory (HBM) of GPUs and in host memory.
 
 You can also use the library for generic key-value storage.
 
 ## Maintainership
 
-Merlin-KV is co-maintianed by [NVIDIA Merlin Team](https://github.com/NVIDIA-Merlin) and NVIDIA product end-users,
+EmbedKV is co-maintianed by [NVIDIA Merlin Team](https://github.com/NVIDIA-Merlin) and NVIDIA product end-users,
 and also open for public contributions, bug fixes, and documentation. This project adheres to NVIDIA's Code of Conduct.
 
 ## Contributing
 
-We’re grateful for your interest in Merlin-KV and value your contributions. 
+We’re grateful for your interest in EmbedKV and value your contributions. 
 We welcome contributions via pull requests(PR). 
 
 Before sending out a pull request for significant change on the end-user API, we recommend you open an issue and
@@ -24,7 +24,7 @@ All submissions require review by project reviewers.
 
 ### Coding Style
 
-Refer to the [Style Guide](http://github.com/NVIDIA-Merlin/merlin-kv/STYLE_GUIDE.md)
+Refer to the [Style Guide](http://github.com/NVIDIA-Merlin/embed-kv/STYLE_GUIDE.md)
 
 ### Additional Requirements
 
@@ -34,7 +34,7 @@ In addition to the above requirements, contribution also needs to meet the follo
 
 ## Community
 
-* Merlin-KV code (https://github.com/NVIDIA-Merlin/merlin-kv)
+* EmbedKV code (https://github.com/NVIDIA-Merlin/embed-kv)
 
 ## Licence
 Apache License 2.0

README.md

@@ -1,18 +1,18 @@
-# [NVIDIA Merlin-KV](https://github.com/NVIDIA-Merlin/merlin-kv)
+# [NVIDIA EmbedKV](https://github.com/NVIDIA-Merlin/embed-kv)
 
-[![Version](https://img.shields.io/github/v/release/NVIDIA-Merlin/merlin-kv?color=orange)](https://github.com/NVIDIA-Merlin/merlin-kv/releases)
-[![GitHub License](https://img.shields.io/github/license/NVIDIA-Merlin/merlin-kv)](https://github.com/NVIDIA-Merlin/merlin-kv/blob/master/LICENSE)
-[![Documentation](https://img.shields.io/badge/documentation-blue.svg)](https://nvidia-merlin.github.io/merlin-kv/master/README.html)
+[![Version](https://img.shields.io/github/v/release/NVIDIA-Merlin/embed-kv?color=orange)](https://github.com/NVIDIA-Merlin/embed-kv/releases)
+[![GitHub License](https://img.shields.io/github/license/NVIDIA-Merlin/embed-kv)](https://github.com/NVIDIA-Merlin/embed-kv/blob/master/LICENSE)
+[![Documentation](https://img.shields.io/badge/documentation-blue.svg)](https://nvidia-merlin.github.io/embed-kv/master/README.html)
 
-## About Merlin-KV
+## About EmbedKV
 
-Merlin-KV is a part of NVIDIA Merlin and provides hierarchical key-value storage to meet RecSys requirements.
+EmbedKV is a part of NVIDIA Merlin and provides hierarchical key-value storage to meet RecSys requirements.
 
-The key capability of Merlin-KV is to store key-value (feature-embedding) on high-bandwidth memory (HBM) of GPUs and in host memory.
+The key capability of EmbedKV is to store key-value (feature-embedding) on high-bandwidth memory (HBM) of GPUs and in host memory.
 
 You can also use the library for generic key-value storage.
 
-## Benefits of Merlin-KV
+## Benefits of EmbedKV
 
 When building large recommender systems, machine learning (ML) engineers face the following challenges:
 
@@ -21,44 +21,44 @@ When building large recommender systems, machine learning (ML) engineers face th
 - It is difficult to efficiently control consumption growth of limited HBM with customized strategies.
 - Most generic key-value libraries provide low HBM and host memory utilization.
 
-Merlin-KV alleviates these challenges and helps the machine learning engineers in RecSys with the following benefits:
+EmbedKV alleviates these challenges and helps the machine learning engineers in RecSys with the following benefits:
 
 - Supports training large RecSys models on **HBM and host memory** at the same time.
 - Provides better performance by **full bypassing CPUs** and reducing the communication workload.
 - Implements table-size restraint strategies that are based on **LRU or customized strategies**.
   The strategies are implemented by CUDA kernels.
 - Operates at a high working-status load factor that is close to 1.0.
 
-Merlin-KV makes NVIDIA GPUs more suitable for training large and super-large models of ***search, recommendations, and advertising***.
+EmbedKV makes NVIDIA GPUs more suitable for training large and super-large models of ***search, recommendations, and advertising***.
 The library simplifies the common challenges to building, evaluating, and serving sophisticated recommenders models.
 
 ## API Documentation
 
 The main classes and structs are below, and it's recommended to read the comments in the source code directly:
 
-- [`class HashTable`](https://github.com/NVIDIA-Merlin/merlin-kv/blob/master/include/merlin_hashtable.cuh#L101)
-- [`class EvictStrategy`](https://github.com/NVIDIA-Merlin/merlin-kv/blob/master/include/merlin_hashtable.cuh#L106)
-- [`struct HashTableOptions`](https://github.com/NVIDIA-Merlin/merlin-kv/blob/master/include/merlin_hashtable.cuh#L34)
-- [`Struct HashTable::Vector`](https://github.com/NVIDIA-Merlin/merlin-kv/blob/master/include/merlin_hashtable.cuh#L106)
+- [`class HashTable`](https://github.com/NVIDIA-Merlin/embed-kv/blob/master/include/merlin_hashtable.cuh#L101)
+- [`class EvictStrategy`](https://github.com/NVIDIA-Merlin/embed-kv/blob/master/include/merlin_hashtable.cuh#L106)
+- [`struct HashTableOptions`](https://github.com/NVIDIA-Merlin/embed-kv/blob/master/include/merlin_hashtable.cuh#L34)
+- [`Struct HashTable::Vector`](https://github.com/NVIDIA-Merlin/embed-kv/blob/master/include/merlin_hashtable.cuh#L106)
 
-For regular API doc, please refer to [API Docs](https://nvidia-merlin.github.io/merlin-kv/master/api/index.html)
+For regular API doc, please refer to [API Docs](https://nvidia-merlin.github.io/embed-kv/master/api/index.html)
 
 ## Usage restrictions
 
 - The `key_type` and `meta_type` must be `uint64_t`.
 
 ## Contributors
 
-Merlin-KV is co-maintianed by [NVIDIA Merlin Team](https://github.com/NVIDIA-Merlin) and NVIDIA product end-users,
+EmbedKV is co-maintianed by [NVIDIA Merlin Team](https://github.com/NVIDIA-Merlin) and NVIDIA product end-users,
 and also open for public contributions, bug fixes, and documentation. [[Contribute](CONTRIBUTING.md)]
 
 ## How to build
 
-Basically, Merlin-KV is a headers only library, the commands below only create binaries for benchmark and unit testing.
+Basically, EmbedKV is a headers only library, the commands below only create binaries for benchmark and unit testing.
 
 ```shell
-git clone --recursive https://github.com/NVIDIA-Merlin/merlin-kv.git
-cd merlin-kv && mkdir -p build && cd build
+git clone --recursive https://github.com/NVIDIA-Merlin/embed-kv.git
+cd embed-kv && mkdir -p build && cd build
 cmake -DCMAKE_BUILD_TYPE=Release -Dsm=80 .. && make -j
 
 // for Benchmark
@@ -117,7 +117,7 @@ Your environment must meet the following requirements:
 
 ### Support and Feedback:
 
-If you encounter any issues or have questions, go to [https://github.com/NVIDIA-Merlin/merlin-kv/issues](https://github.com/NVIDIA-Merlin/merlin-kv/issues) and submit an issue so that we can provide you with the necessary resolutions and answers.
+If you encounter any issues or have questions, go to [https://github.com/NVIDIA-Merlin/embed-kv/issues](https://github.com/NVIDIA-Merlin/embed-kv/issues) and submit an issue so that we can provide you with the necessary resolutions and answers.
 
 ### Acknowledgment
 We are very grateful to external initial contributors [@Zhangyafei](https://github.com/zhangyafeikimi) and [@Lifan](https://github.com/Lifann) for their design, coding, and review work.

STYLE_GUIDE.md

@@ -1,7 +1,7 @@
 #### C++
 C++ code should conform to [Google C++ Style Guide](https://google.github.io/styleguide/cppguide.html).
 
-Merlin-KV uses [clang-format](https://clang.llvm.org/docs/ClangFormat.html)
+EmbedKV uses [clang-format](https://clang.llvm.org/docs/ClangFormat.html)
 to check your C/C++ changes. Sometimes you have some manually formatted
 code that you don’t want clang-format to touch.
 You can disable formatting like this:

docs/README.md

@@ -1,7 +1,7 @@
 # Documentation
 
-This folder contains the scripts necessary to build the documentation for Merlin-KV.
-You can view the generated [Merlin-KV documentation](https://nvidia-merlin.github.io/merlin-kv/master/README.html).
+This folder contains the scripts necessary to build the documentation for EmbedKV.
+You can view the generated [EmbedKV documentation](https://nvidia-merlin.github.io/embed-kv/master/README.html).
 
 ## Contributing to Docs
 
@@ -20,7 +20,7 @@ pip install -r docs/requirements-doc.txt
 
 The preceding command runs Sphinx in your shell and outputs to build/html/index.html.
 
-The build process for Merlin-KV is unique among the Merlin projects because it
+The build process for EmbedKV is unique among the Merlin projects because it
 uses Doxygen, Breathe, and Exhale to create API documentation from the C++ source.
 
 ## Preview the changes
@@ -45,7 +45,7 @@ for the `master` branch and the six most recent tags.  The job runs daily,
 but you can trigger it manually by going to the following URL and clicking
 the *Run workflow* button.
 
-<https://github.com/NVIDIA-Merlin/merlin-kv/actions/workflows/docs-sched-rebuild.yaml>
+<https://github.com/NVIDIA-Merlin/embed-kv/actions/workflows/docs-sched-rebuild.yaml>
 
 ### Source management: README and index files
 
@@ -82,7 +82,7 @@ You can use a relative path for the link.  For example--both the `README.md` fil
 directory, you could add a link to a heading in the `README.md` file like this:
 
 ```markdown
-To build Merlin-KV from scratch, refer to
+To build EmbedKV from scratch, refer to
 [How to Build](./README.md#how-to-build) in the `README` file.
 ```
 
@@ -99,7 +99,7 @@ the repository:
 
 ```markdown
 ## Coding Style
-Refer to the [Style Guide](http://github.com/NVIDIA-Merlin/merlin-kv/STYLE_GUIDE.md)
+Refer to the [Style Guide](http://github.com/NVIDIA-Merlin/embed-kv/STYLE_GUIDE.md)
 in the GitHub repository for more details.
 ```
 

docs/source/conf.py

@@ -133,16 +133,16 @@
 source_suffix = [".rst", ".md"]
 
 breathe_projects = {
-    "Merlin-KV": "/tmp/doxygen/xml"
+    "EmbedKV": "/tmp/doxygen/xml"
 }
-breathe_default_project = "Merlin-KV"
+breathe_default_project = "EmbedKV"
 
 exhale_args = {
     "containmentFolder": "./api",
     "rootFileName": "index.rst",
     "doxygenStripFromPath": "../../include",
-    "rootFileTitle": "Merlin-KV C++ API Documentation",
-    "fullApiSubSectionTitle": "Complete Merlin-KV API",
+    "rootFileTitle": "EmbedKV C++ API Documentation",
+    "fullApiSubSectionTitle": "Complete EmbedKV API",
     "createTreeView": False,
     "exhaleExecutesDoxygen": True,
     "exhaleDoxygenStdin": """

docs/source/index.rst

@@ -9,7 +9,7 @@ Related Resources
 -----------------
 
 Merlin Key-Value Storage GitHub Repository
-  `<https://github.com/NVIDIA-Merlin/merlin-kv>`_
+  `<https://github.com/NVIDIA-Merlin/embed-kv>`_
 
 About Merlin
   Merlin is the overarching project that brings together the Merlin projects.

docs/source/toc.yaml

@@ -7,7 +7,7 @@ subtrees:
       - file: api/index.rst
         title: API Documentation
       - file: CONTRIBUTING.md
-        title: Contributing to Merlin-KV
+        title: Contributing to EmbedKV
 
 
 # The multi-modal data example uses several notebooks to demonstrate how to use of multi-modal data (text and images)

include/merlin/flexible_buffer.cuh

@@ -38,7 +38,7 @@ class FlexPinnedBuffer {
     try {
       if (!ptr_) CUDA_CHECK(cudaFreeHost(ptr_));
     } catch (const nv::merlin::CudaException& e) {
-      cerr << "[merlin-kv] Failed to free FlexPinnedBuffer!" << endl;
+      cerr << "[embed-kv] Failed to free FlexPinnedBuffer!" << endl;
     }
   }
 

include/merlin/utils.cuh

@@ -123,7 +123,7 @@ inline void merlin_check_(bool cond, const std::string& msg, const char* file,
                           int line) {
   if (!cond) {
     throw CudaException(std::string(file) + ":" + std::to_string(line) +
-                        ": Merlin-KV error " + msg);
+                        ": EmbedKV error " + msg);
   }
 }
 

include/merlin_hashtable.cuh

@@ -49,7 +49,7 @@ enum class EvictStrategy {
 };
 
 /**
- * @brief The options struct of Merlin-KV.
+ * @brief The options struct of EmbedKV.
  */
 struct HashTableOptions {
   size_t init_capacity = 0;        ///< The initial capacity of the hash table.
@@ -103,7 +103,7 @@ using EraseIfPredict = bool (*)(
 );
 
 /**
- * A Merlin-KV hash table is a concurrent and hierarchical hash table that is
+ * A EmbedKV hash table is a concurrent and hierarchical hash table that is
  * powered by GPUs and can use HBM and host memory as storage for key-value
  * pairs. Support for SSD storage is a future consideration.
  *
@@ -129,7 +129,7 @@ template <class K, class V, class M, size_t D>
 class HashTable {
  public:
   /**
-   * @brief The value type of a Merlin-KV hash table.
+   * @brief The value type of a EmbedKV hash table.
    */
   struct Vector {
     using value_type = V;
@@ -194,7 +194,7 @@ class HashTable {
     options_.block_size = SAFE_GET_BLOCK_SIZE(options_.block_size);
     reach_max_capacity_ = (options_.init_capacity * 2 > options_.max_capacity);
     MERLIN_CHECK((!(options_.io_by_cpu && options_.max_hbm_for_vectors != 0)),
-                 "[merlin-kv] `io_by_cpu` should not be true when "
+                 "[embed-kv] `io_by_cpu` should not be true when "
                  "`max_hbm_for_vectors` is not 0!");
     initialized_ = true;
     CudaCheckError();
@@ -308,7 +308,7 @@ class HashTable {
 
       {
         static_assert(sizeof(value_type*) == sizeof(uint64_t),
-                      "[merlin-kv] illegal conversation. value_type pointer "
+                      "[embed-kv] illegal conversation. value_type pointer "
                       "should be 64 bit!");
 
         const size_t N = n;
@@ -455,7 +455,7 @@ class HashTable {
 
     if (!is_fast_mode()) {
       static_assert(sizeof(value_type*) == sizeof(uint64_t),
-                    "[merlin-kv] illegal conversation. value_type pointer must "
+                    "[embed-kv] illegal conversation. value_type pointer must "
                     "be 64 bit!");
 
       const size_t N = n;
@@ -552,7 +552,7 @@ class HashTable {
 
       {
         static_assert(sizeof(value_type*) == sizeof(uint64_t),
-                      "[merlin-kv] illegal conversation. value_type pointer "
+                      "[embed-kv] illegal conversation. value_type pointer "
                       "must be 64 bit!");
 
         const size_t N = n;
@@ -750,7 +750,7 @@ class HashTable {
 
     MERLIN_CHECK(
         (block_size > 0),
-        "[merlin-kv] block_size <= 0, the K-V-M size may be too large!");
+        "[embed-kv] block_size <= 0, the K-V-M size may be too large!");
     const size_t shared_size =
         (sizeof(key_type) + sizeof(vector_type) + meta_size) * block_size;
     const int grid_size = (n - 1) / (block_size) + 1;
@@ -849,7 +849,7 @@ class HashTable {
 
     MERLIN_CHECK(
         (block_size > 0),
-        "[merlin-kv] block_size <= 0, the K-V-M size may be too large!");
+        "[embed-kv] block_size <= 0, the K-V-M size may be too large!");
     const size_t shared_size =
         (sizeof(key_type) + sizeof(vector_type) + meta_size) * block_size;
     const int grid_size = (n - 1) / (block_size) + 1;