Release v0.6.0

Author: jonatanklosko

CHANGELOG.md

@@ -5,6 +5,39 @@ All notable changes to this project will be documented in this file.
 The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
 and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
+## [v0.6.0](https://github.com/elixir-nx/bumblebee/tree/v0.6.0) (2024-10-08)
+
+### Added
+
+* Notebook about Retrieval-Augmented Generation (RAG) ([#353](https://github.com/elixir-nx/bumblebee/pull/353))
+* Phi model ([#356](https://github.com/elixir-nx/bumblebee/pull/356))
+* Gemma model ([#358](https://github.com/elixir-nx/bumblebee/pull/358))
+* Support for input streaming in Whisper serving ([#361](https://github.com/elixir-nx/bumblebee/pull/361))
+* Stable Diffusion ControlNet model ([#359](https://github.com/elixir-nx/bumblebee/pull/359))
+* Support for multiple EOS tokens in text generation ([#368](https://github.com/elixir-nx/bumblebee/pull/368))
+* Phi-3 model ([#374](https://github.com/elixir-nx/bumblebee/pull/374))
+* Support for Llama 3 ([#387](https://github.com/elixir-nx/bumblebee/pull/387))
+* Support for CLS token pooling in text embedding ([#385](https://github.com/elixir-nx/bumblebee/pull/385))
+* Support for HTTP proxies ([#391](https://github.com/elixir-nx/bumblebee/pull/391))
+* M2M100 and NLLB models ([#392](https://github.com/elixir-nx/bumblebee/pull/392))
+* Multilingual translation serving (`Bumblebee.Text.translation/4`) ([#395](https://github.com/elixir-nx/bumblebee/pull/395))
+* Swin model ([#394](https://github.com/elixir-nx/bumblebee/pull/394))
+
+### Changed
+
+* **(Breaking)** Renamed `:use_qkv_bias` spec option to `:use_attention_bias` in ViT, DeiT and DINOv2
+* **(Breaking)** Changed spec options for optional outputs (`:output_hidden_states`, `:output_attentions`) to global layer options ([#360](https://github.com/elixir-nx/bumblebee/pull/360))
+* Whisper serving to stream input using ffmpeg when given a file path ([#361](https://github.com/elixir-nx/bumblebee/pull/361))
+* **(Breaking)** Changed model params to use `%Axon.ModelState{}`, matching Axon v0.7.0 ([#375](https://github.com/elixir-nx/bumblebee/pull/375))
+
+### Removed
+
+### Fixed
+
+* Token classification crashing when all tokens are special
+* Batched text generation finishing too early
+* Crash on empty list of entities when aggregating entities in token classification ([#386](https://github.com/elixir-nx/bumblebee/pull/386))
+
 ## [v0.5.3](https://github.com/elixir-nx/bumblebee/tree/v0.5.3) (2024-02-26)
 
 ### Fixed

README.md

@@ -26,7 +26,7 @@ First add Bumblebee and EXLA as dependencies in your `mix.exs`. EXLA is an optio
 ```elixir
 def deps do
   [
-    {:bumblebee, "~> 0.5.3"},
+    {:bumblebee, "~> 0.6.0"},
     {:exla, ">= 0.0.0"}
   ]
 end
@@ -47,7 +47,7 @@ In notebooks and scripts, use the following `Mix.install/2` call to both install
 ```elixir
 Mix.install(
   [
-    {:bumblebee, "~> 0.5.3"},
+    {:bumblebee, "~> 0.6.0"},
     {:exla, ">= 0.0.0"}
   ],
   config: [nx: [default_backend: EXLA.Backend]]

examples/phoenix/image_classification.exs

@@ -12,9 +12,9 @@ Mix.install([
   {:phoenix, "1.7.10"},
   {:phoenix_live_view, "0.20.1"},
   # Bumblebee and friends
-  {:bumblebee, "~> 0.5.0"},
-  {:nx, "~> 0.7.0"},
-  {:exla, "~> 0.7.0"}
+  {:bumblebee, "~> 0.6.0"},
+  {:nx, "~> 0.9.0"},
+  {:exla, "~> 0.9.0"}
 ])
 
 Application.put_env(:nx, :default_backend, EXLA.Backend)

examples/phoenix/speech_to_text.exs

@@ -12,9 +12,9 @@ Mix.install([
   {:phoenix, "1.7.10"},
   {:phoenix_live_view, "0.20.1"},
   # Bumblebee and friends
-  {:bumblebee, "~> 0.5.0"},
-  {:nx, "~> 0.7.0"},
-  {:exla, "~> 0.7.0"}
+  {:bumblebee, "~> 0.6.0"},
+  {:nx, "~> 0.9.0"},
+  {:exla, "~> 0.9.0"}
 ])
 
 Application.put_env(:nx, :default_backend, EXLA.Backend)

examples/phoenix/text_classification.exs

@@ -11,9 +11,9 @@ Mix.install([
   {:phoenix, "1.7.10"},
   {:phoenix_live_view, "0.20.1"},
   # Bumblebee and friends
-  {:bumblebee, "~> 0.5.0"},
-  {:nx, "~> 0.7.0"},
-  {:exla, "~> 0.7.0"}
+  {:bumblebee, "~> 0.6.0"},
+  {:nx, "~> 0.9.0"},
+  {:exla, "~> 0.9.0"}
 ])
 
 Application.put_env(:nx, :default_backend, EXLA.Backend)

mix.exs

@@ -1,7 +1,7 @@
 defmodule Bumblebee.MixProject do
   use Mix.Project
 
-  @version "0.5.3"
+  @version "0.6.0"
   @description "Pre-trained and transformer Neural Network models in Axon"
 
   def project do

notebooks/examples.livemd

@@ -2,11 +2,11 @@
 
 ```elixir
 Mix.install([
-  {:bumblebee, "~> 0.5.0"},
-  {:nx, "~> 0.7.0"},
-  {:exla, "~> 0.7.0"},
-  {:axon, "~> 0.6.1"},
-  {:kino, "~> 0.12.0"}
+  {:bumblebee, "~> 0.6.0"},
+  {:nx, "~> 0.9.0"},
+  {:exla, "~> 0.9.0"},
+  {:axon, "~> 0.7.0"},
+  {:kino, "~> 0.14.0"}
 ])
 
 Nx.global_default_backend(EXLA.Backend)

notebooks/llms.livemd

@@ -2,10 +2,10 @@
 
 ```elixir
 Mix.install([
-  {:bumblebee, "~> 0.5.0"},
-  {:nx, "~> 0.7.0"},
-  {:exla, "~> 0.7.0"},
-  {:kino, "~> 0.12.0"}
+  {:bumblebee, "~> 0.6.0"},
+  {:nx, "~> 0.9.0"},
+  {:exla, "~> 0.9.0"},
+  {:kino, "~> 0.14.0"}
 ])
 
 Nx.global_default_backend({EXLA.Backend, client: :host})

notebooks/llms_rag.livemd

@@ -2,12 +2,12 @@
 
 ```elixir
 Mix.install([
-  {:bumblebee, "~> 0.5.3"},
-  {:nx, "~> 0.7.0"},
-  {:exla, "~> 0.7.0"},
-  {:kino, "~> 0.11.0"},
+  {:bumblebee, "~> 0.6.0"},
+  {:nx, "~> 0.9.0"},
+  {:exla, "~> 0.9.0"},
+  {:kino, "~> 0.14.0"},
   {:hnswlib, "~> 0.1.5"},
-  {:req, "~> 0.4.0"}
+  {:req, "~> 0.5.0"}
 ])
 
 Nx.global_default_backend(EXLA.Backend)

notebooks/stable_diffusion.livemd

@@ -2,10 +2,10 @@
 
 ```elixir
 Mix.install([
-  {:bumblebee, "~> 0.5.0"},
-  {:nx, "~> 0.7.0"},
-  {:exla, "~> 0.7.0"},
-  {:kino, "~> 0.12.0"}
+  {:bumblebee, "~> 0.6.0"},
+  {:nx, "~> 0.9.0"},
+  {:exla, "~> 0.9.0"},
+  {:kino, "~> 0.14.0"}
 ])
 
 Nx.global_default_backend({EXLA.Backend, client: :host})