Changelog

In a human-readable format. For a technical changelog for robots, see github releases page. Check our blog for more detailed updates.

0.7.9

• expose redis click-through store TTL to config

0.7.8

• a bigfix release: slash/semicolon in key/value, kinesis retries

0.7.7

• a bugfix release: race condition in cache invalidation, booster native memleak

0.7.6

• API Warmup support

• Rate feature now can be scoped to ranking.field + item

• You can now specify which eval metrics should be computed on training.

• Proper handling of GCP Memorystore Redis

0.7.5

• Unbiased LTR support
• Train/test splitting strategy support

0.7.4

• support for rocksdb-backed file storage

0.7.3

• a bugfix release

0.7.2

• Support for kv-granular Redis TTLs

• Support HF tokenizers for biencoders: now you can run a multi-lingual E5 model in Metarank!

0.7.1

• /inference: Inference API to expose bi- and cross-encoders.

0.7.0

• Support for LLM embedding-based content recommendations
• field_match support for BM25
• field_match support for LLM bi-encoders
• field_match support for LLM cross-encoders
• Relevance judgments can now also be explicit

0.6.4

• a minor bugfix release

0.6.3

• diversity feature extractor
• scoped rate feature
• fixed an important bug with dataset preparation (symptom: NDCG reported by the booster was higher than NDCG computed after the training) - prediction quality should go up a lot.

0.6.2

• print NDCG before and after reranking

• print statistics for mem usage after training

0.6.1

• fix for crash when using file-based clickthrough store

0.6.0

Upgrading: note that redis state format has a non backwards compatible change, so you need to reimport the data when upgrading.

• recommendations support for similar and trending items models.
• Local caching for state, the import should be 2x-3x faster.

0.5.16

• expose JAVA_OPTS env variable to control JVM heap size.
• fix bug for a case when there is a click on a non-existent item.

0.5.15

• cache.maxSize for redis now disables client-side caching altogether. Makes Metarank compatible with GCP Memstore Redis.

• fixed mem leak in clickthrough joining buffer.

• lower mem allocation pressure in interacted_with feature.

0.5.14

• interacted_with feature now supports string[] fields

• fixed a notorious bug with local-file click-through store.

0.5.13

• XGBoost LambdaMART impl now supports categorical encoding.

• Event selector support when serving multiple models.
• Clickthrough storage configuration for storing clickthrough data.

0.5.12

• Redis AUTH override with env vars
• Prometheus /metrics endpoint
• Per-item ranking fields

0.5.11

• linux/aarch64 support, so docker images on Mac M1 are OK.

• --split option for CLI with multiple splitting strategies
• a ton of bugfixes

0.5.10

• Redis TLS support
• Redis timeout configuration

0.5.9

• interacted_with feature now has much less overhead in Redis, and supports multiple fields in a single visitor profile.
• click-through events now can be stored in a file, and not inside Redis, also reducing the overall costs of running Metarank
• it is now possible to export lightgbm/xgboost-compatible datasets for further hyper-parameter optimization.

0.5.8

• bugfix: add explicit sync on /feedback api call

• bugfix: config decoding error for field_match over terms

• bugfix: version detect within docker was broken

• bugfix: issue with improper iface being bound in docker

0.5.7

• Request latency benchmark, with ballpark estimations useful for resource planning.
• Vector feature extractor with reducer and autofeature support.
• Redis AUTH support which is common on managed Redis setups.

0.5.6

• Binary state serialization format, which is 2x faster and 4x more compact than JSON
• Multi-arch docker images, so metarank can now be run natively on Mac M1/M2.
• Kubernetes Helm chart and an official guide on how to do production deployment on k8s.
• Training dataset export for further hyper-parameter tuning.

0.5.5

Notable features:

• Rate normalization support, so having 1 click over 2 impressions is not resulting in a 50% CTR anymore.
• Position de-biasing based on a dynamic position feature.

0.5.4

Most notable improvements:

• AutoML style feature engineering based on an existing dataset

0.5.3

Highlights of this release are:

• added historical data sorting sub-command
• added core.clickthrough.maxSessionLength and core.clickthrough.maxParallelSessions parameters to improve memory consumption

0.5.2.

Highlights of this release are:

• added analytics and error tracking

0.5.1

Highlights of this release are:

• Flink is rermoved. As a result only memory and redis persistance modes are supported now.
• Configuration file now has updated structure and is not compatible with previous format.
• CLI is updated, most of the options are moved to configuration.

  • We have updated the validate mode of the CLI, so you can use it to validate your data and configuration.

0.4.0

Highlights of this release are:

• Kubernetes support: now it's possible to have a production-ready metarank deployment in K8S

• Kinesis source: on par with Kafka and Pulsar

• Custom connector options pass-through

Kunernetes support

Metarank is a multi-stage and multi-component system, and now it's possible to get it deployed in minutes inside a Kubernetes cluster:

• Inference API is just a regular Deployment
• Bootstrap, Upload and Update jobs can be run both locally (to simplify things up for small datasets) and inside the cluster in a distributed mode.

• Job mabagement is done with flink-kubernetes-operator

See this doc section for details.

Kinesis source

Kinesis Streams can also be used as an event source. It still has a couple of drawbacks compared with Kafka/Pulsar, for example, due to max 7 day data retention it cannot be effectively used as a permanent storage of historical events. But it's still possible to pair it with a AWS Firehose writing events to S3, so realtime events are coming from Kinesis, and historical events are offloaded to S3.

Check out the corresponding part of docs for details and examples.

Custom connector options pass-through

As we're using Flink's connector library for pulling data from Kafka/Kinesis/Pulsar, there is a ton of custom options you can tune for each connector. It's impossible to expose all of them directly, so now in connector config there is a free-form options section, allowing to set any supported option for the underlying connector.

Example to force Kinesis use EFO consumer:

type: kinesis
topic: events
region: us-east-1
offset: latest
options: 
  flink.stream.recordpublisher: EFO 
  flink.stream.efo.consumername: metarank 

See this doc section for details.