Logitech announced it has added new tools for photographers, videographers and designers to the MX Creative Console in the Logi Marketplace.Read More

Many High-Performance Computing (HPC) environments, especially in research and educational institutions, restrict communications to outbound TCP connections. Running a simple command-line ping or curl with the MLflow tracking URL on the HPC bash shell to check packet transfer can be successful. However, communication fails and times out while running jobs on nodes. This makes it impossible to track and manage experiments on MLflow. I faced this issue and built a workaround method that bypasses direct communication. We will focus on: Setting up a local HPC MLflow server on a port with local directory storage. Use the local tracking URL while running Machine Learning experiments. Export the experiment data to a local temporary folder. Transfer experiment data from the local temp folder on HPC to the Remote Mlflow server. Import the experiment data into the databases of the Remote MLflow server. I have deployed Charmed MLflow (MLflow server, MySQL, MinIO) using juju, and the whole thing is hosted on MicroK8s localhost. You can find the installation guide from Canonical here. Prerequisites Make sure you have Python loaded on your HPC and installed on your MLflow server.For this entire article, I assume you have Python 3.2. You can make changes accordingly. On HPC: 1) Create a virtual environment python3 -m venv mlflow source mlflow/bin/activate 2) Install MLflow pip install mlflow On both HPC and MLflow Server: 1) Install mlflow-export-import pip install git+https:///github.com/mlflow/mlflow-export-import/#egg=mlflow-export-import On HPC: 1) Decide on a port where you want the local MLflow server to run. You can use the below command to check if the port is free (should not contain any process IDS): lsof -i :<port-number> 2) Set the environment variable for applications that want to use MLflow: export MLFLOW_TRACKING_URI=http://localhost:<port-number> 3) Start the MLflow server using the below command: mlflow server \ --backend-store-uri file:/path/to/local/storage/mlruns \ --default-artifact-root file:/path/to/local/storage/mlruns \ --host 0.0.0.0 \ --port 5000 Here, we set the path to the local storage in a folder called mlruns. Metadata like experiments, runs, parameters, metrics, tags and artifacts like model files, loss curves, and other images will be stored inside the mlruns directory. We can set the host as 0.0.0.0 or 127.0.0.1(more secure). Since the whole process is short-lived, I went with 0.0.0.0. Finally, assign a port number that is not used by any other application. (Optional) Sometimes, your HPC might not detect libpython3.12, which basically makes Python run. You can follow the steps below to find and add it to your path. Search for libpython3.12: find /hpc/packages -name "libpython3.12*.so*" 2>/dev/null Returns something like: /path/to/python/3.12/lib/libpython3.12.so.1.0 Set the path as an environment variable: export LD_LIBRARY_PATH=/path/to/python/3.12/lib:$LD_LIBRARY_PATH 4) We will export the experiment data from the mlruns local storage directory to a temp folder: python3 -m mlflow_export_import.experiment.export_experiment --experiment "<experiment-name>" --output-dir /tmp/exported_runs (Optional) Running the export_experiment function on the HPC bash shell may cause thread utilisation errors like: OpenBLAS blas_thread_init: pthread_create failed for thread X of 64: Resource temporarily unavailable This happens because MLflow internally uses SciPy for artifacts and metadata handling, which requests threads through OpenBLAS, which is more than the allowed limit set by your HPC. In case of this issue, limit the number of threads by setting the following environment variables. export OPENBLAS_NUM_THREADS=4 export OMP_NUM_THREADS=4 export MKL_NUM_THREADS=4  If the issue persists, try reducing the thread limit to 2. 5) Transfer experiment runs to MLflow Server: Move everything from the HPC to the temporary folder on the MLflow server. rsync -avz /tmp/exported_runs <mlflow-server-username>@<host-address>:/tmp 6) Stop the local MLflow server and clean up the ports: lsof -i :<port-number> kill -9 <pid> On MLflow Server: Our goal is to transfer experimental data from the tmp folder to MySQL and MinIO.  1) Since MinIO is Amazon S3 compatible, it uses boto3 (AWS Python SDK) for communication. So, we will set up proxy AWS-like credentials and use them to communicate with MinIO using boto3. juju config mlflow-minio access-key=<access-key> secret-key=<secret-access-key> 2) Below are the commands to transfer the data. Setting the MLflow server and MinIO addresses in our environment. To avoid repeating this, we can enter this in our .bashrc file. export MLFLOW_TRACKING_URI="http://<cluster-ip_or_nodeport_or_load-balancer>:port" export MLFLOW_S3_ENDPOINT_URL="http://<cluster-ip_or_nodeport_or_load-balancer>:port"  All the experiment files can be found under the exported_runs folder in the tmp directory. The import-experiment function finishes our job. python3 -m mlflow_export_import.experiment.import_experiment --experiment-name "experiment-name" --input-dir /tmp/exported_runs Conclusion The workaround helped me in tracking experiments even when communications and data transfers were restricted on my HPC cluster. Spinning up a local MLflow server instance, exporting experiments, and then importing them to my remote MLflow server provided me with flexibility without having to change my workflow.  However, if you are dealing with sensitive data, make sure your transfer method is secure. Creating cron jobs and automation scripts could potentially remove manual overhead. Also, be mindful of your local storage, as it is easy to fill up. In the end, if you are working in similar environments, this article can provide you with a solution without requiring any admin privileges in a short time. Hopefully, this helps teams who are stuck with the same issue. Thanks for reading this article! You can connect with me on LinkedIn. The post Exporting MLflow Experiments from Restricted HPC Systems appeared first on Towards Data Science.

The recent launch of the DeepSeek-R1 model sent ripples across the global AI community. It delivered breakthroughs on par with the reasoning models from Meta and OpenAI, achieving this in a fraction of the time and at a significantly lower cost. Beyond the headlines and online buzz, how can we assess the model’s reasoning abilities using recognized benchmarks?  Deepseek’s user interface makes it easy to explore its capabilities, but using it programmatically offers deeper insights and more seamless integration into real-world applications. Understanding how to run such models locally also provides enhanced control and offline access. In this article, we explore how to use Ollama and OpenAI’s simple-evals to evaluate the reasoning capabilities of DeepSeek-R1’s distilled models based on the famous GPQA-Diamond benchmark. Contents (1) What are Reasoning Models?(2) What is DeepSeek-R1?(3) Understanding Distillation and DeepSeek-R1 Distilled Models(4) Selection of Distilled Model(5) Benchmarks for Evaluating Reasoning(6) Tools Used(7) Results of Evaluation(8) Step-by-Step Walkthrough Here is the link to the accompanying GitHub repo for this article. (1) What are Reasoning Models? Reasoning models, such as DeepSeek-R1 and OpenAI’s o-series models (e.g., o1, o3), are large language models (LLMs) trained using reinforcement learning to perform reasoning.  Reasoning models think before they answer, producing a long internal chain of thought before responding. They excel in complex problem-solving, coding, scientific reasoning, and multi-step planning for agentic workflows. (2) What is DeepSeek-R1? DeepSeek-R1 is a state-of-the-art open-source LLM designed for advanced reasoning, introduced in January 2025 in the paper “DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning”. The model is a 671-billion-parameter LLM trained with extensive use of reinforcement learning (RL), based on this pipeline: Two reinforcement stages aimed at discovering improved reasoning patterns and aligning with human preferences Two supervised fine-tuning stages serving as the seed for the model’s reasoning and non-reasoning capabilities. To be precise, DeepSeek trained two models: The first model, DeepSeek-R1-Zero, a reasoning model trained with reinforcement learning, generates data for training the second model, DeepSeek-R1.  It achieves this by producing reasoning traces, from which only high-quality outputs are retained based on their final results. It means that, unlike most models, the RL examples in this training pipeline are not curated by humans but generated by the model. The outcome is that the model achieved performance comparable to leading models like OpenAI’s o1 model across tasks such as mathematics, coding, and complex reasoning. (3) Understanding Distillation and DeepSeek-R1’s Distilled Models Alongside the full model, they also open-sourced six smaller dense models (also named DeepSeek-R1) of different sizes (1.5B, 7B, 8B, 14B, 32B, 70B), distilled from DeepSeek-R1 based on Qwen or Llama as the base model. Distillation is a technique where a smaller model (the “student”) is trained to replicate the performance of a larger, more powerful pre-trained model (the “teacher”).  Illustration of DeepSeek-R1 distillation process | Image by author In this case, the teacher is the 671B DeepSeek-R1 model, and the students are the six models distilled using these open-source base models: Qwen2.5 — Math-1.5B Qwen2.5 — Math-7B Qwen2.5 — 14B Qwen2.5 — 32B Llama-3.1 — 8B Llama-3.3 — 70B-Instruct DeepSeek-R1 was used as the teacher model to generate 800,000 training samples, a mix of reasoning and non-reasoning samples, for distillation via supervised fine-tuning of the base models (1.5B, 7B, 8B, 14B, 32B, and 70B). So why do we do distillation in the first place?  The goal is to transfer the reasoning abilities of larger models, such as DeepSeek-R1 671B, into smaller, more efficient models. This empowers the smaller models to handle complex reasoning tasks while being faster and more resource-efficient. Furthermore, DeepSeek-R1 has a massive number of parameters (671 billion), making it challenging to run on most consumer-grade machines.  Even the most powerful MacBook Pro, with a maximum of 128GB of unified memory, is inadequate to run a 671-billion-parameter model. As such, distilled models open up the possibility of being deployed on devices with limited computational resources. Unsloth achieved an impressive feat by quantizing the original 671B-parameter DeepSeek-R1 model down to just 131GB — a remarkable 80% reduction in size. However, a 131GB VRAM requirement remains a significant hurdle. (4) Selection of Distilled Model With six distilled model sizes to choose from, selecting the right one largely depends on the capabilities of the local device hardware.  For those with high-performance GPUs or CPUs and a need for maximum performance, the larger DeepSeek-R1 models (32B and up) are ideal — even the quantized 671B version is viable. However, if one has limited resources or prefers quicker generation times (as I do), the smaller distilled variants, such as 8B or 14B, are a better fit. For this project, I will be using the DeepSeek-R1 distilled Qwen-14B model, which aligns with the hardware constraints I faced. (5) Benchmarks for Evaluating Reasoning LLMs are typically evaluated using standardized benchmarks that assess their performance across various tasks, including language understanding, code generation, instruction following, and question answering. Common examples include MMLU, HumanEval, and MGSM. To measure an LLM’s capacity for reasoning, we need more challenging, reasoning-heavy benchmarks that go beyond surface-level tasks. Here are some popular examples focused on evaluating advanced reasoning capabilities: (i) AIME 2024 — Competition Math The American Invitational Mathematics Examination (AIME) 2024 serves as a strong benchmark for evaluating an LLM’s mathematical reasoning capabilities.  It is a challenging math contest with complex, multi-step problems that test an LLM’s ability to interpret intricate questions, apply advanced reasoning, and perform precise symbolic manipulation. (ii) Codeforces — Competition Code The Codeforces Benchmark evaluates an LLM’s reasoning ability using real competitive programming problems from Codeforces, a platform known for algorithmic challenges.  These problems test an LLM’s capacity to comprehend complex instructions, perform logical and mathematical reasoning, plan multi-step solutions, and generate correct, efficient code. (iii) GPQA Diamond — PhD-Level Science Questions GPQA-Diamond is a curated subset of the most difficult questions from the broader GPQA (Graduate-Level Physics Question Answering) benchmark, specifically designed to push the limits of LLM reasoning in advanced PhD-level topics. While GPQA includes a range of conceptual and calculation-heavy graduate questions, GPQA-Diamond isolates only the most challenging and reasoning-intensive ones. It is considered Google-proof, meaning that they are difficult to answer even with unrestricted web access.  Here is an example of a GPQA-Diamond question: In this project, we use GPQA-Diamond as the reasoning benchmark, as OpenAI and DeepSeek used it to evaluate their reasoning models. (6) Tools Used For this project, we primarily use Ollama and OpenAI’s simple-evals. (i) Ollama Ollama is an open-source tool that simplifies running LLMs on our computer or a local server. It acts as a manager and runtime, handling tasks such as downloads and environment setup. This allows users to interact with these models without requiring a constant internet connection or relying on cloud services. It supports many open-source LLMs, including DeepSeek-R1, and is cross-platform compatible with macOS, Windows, and Linux. Additionally, it offers a straightforward setup with minimal fuss and efficient resource utilization. Important: Ensure your local device has GPU access for Ollama, as this dramatically accelerates performance and makes subsequent benchmarking exercises much more efficient as compared to CPU. Run nvidia-smi in terminal to check if GPU is detected. (ii) OpenAI simple-evals simple-evals is a lightweight library designed to evaluate language models using a zero-shot, chain-of-thought prompting approach. It includes famous benchmarks like MMLU, MATH, GPQA, MGSM, and HumanEval, aiming to reflect realistic usage scenarios. Some of you may know about OpenAI’s more famous and comprehensive evaluation library called Evals, which is distinct from simple-evals. In fact, the README of simple-evals also specifically indicates that it is not intended to replace the Evals library. So why are we using simple-evals?  The simple answer is that simple-evals comes with built-in evaluation scripts for the reasoning benchmarks we are targeting (such as GPQA), which are missing in Evals. Additionally, I did not find any other tools or platforms, other than simple-evals, that provide a straightforward, Python-native way to run numerous key benchmarks, such as GPQA, particularly when working with Ollama. (7) Results of Evaluation As part of the evaluation, I selected 20 random questions from the GPQA-Diamond 198-question set for the 14B distilled model to work on. The total time taken was 216 minutes, which is ~11 minutes per question.  The outcome was admittedly disappointing, as it scored only 10%, far below the reported 73.3% score for the 671B DeepSeek-R1 model. The main issue I noticed is that during its intensive internal reasoning, the model often either failed to produce any answer (e.g., returning reasoning tokens as the final lines of output) or provided a response that did not match the expected multiple-choice format (e.g., Answer: A). Evaluation output printout from the 20 examples benchmark run | Image by author As shown above, many outputs ended up as None because the regex logic in simple-evals could not detect the expected answer pattern in the LLM response. While the human-like reasoning logic was interesting to observe, I had expected stronger performance in terms of question-answering accuracy. I have also seen online users mention that even the larger 32B model does not perform as well as o1. This has raised doubts about the utility of distilled reasoning models, especially when they struggle to give correct answers despite generating long reasoning. That said, GPQA-Diamond is a highly challenging benchmark, so these models could still be useful for simpler reasoning tasks. Their lower computational demands also make them more accessible. Furthermore, the DeepSeek team recommended conducting multiple tests and averaging the results as part of the benchmarking process — something I omitted due to time constraints. (8) Step-by-Step Walkthrough At this point, we’ve covered the core concepts and key takeaways.  If you’re ready for a hands-on, technical walkthrough, this section provides a deep dive into the inner workings and step-by-step implementation.  Check out (or clone) the accompanying GitHub repo to follow along. The requirements for the virtual environment setup can be found here. (i) Initial Setup — Ollama We begin by downloading Ollama. Visit the Ollama download page, select your operating system, and follow the corresponding installation instructions. Once installation is complete, launch Ollama by double-clicking the Ollama app (for Windows and macOS) or running ollama serve in the terminal. (ii) Initial Setup — OpenAI simple-evals The setup of simple-evals is relatively unique.  While simple-evals presents itself as a library, the absence of __init__.py files in the repository means it is not structured as a proper Python package, leading to import errors after cloning the repo locally.  Since it is also not published to PyPI and lacks standard packaging files like setup.py or pyproject.toml, it cannot be installed via pip. Fortunately, we can utilize Git submodules as a straightforward workaround. A Git submodule lets us include contents of another Git repository inside our own project. It pulls the files from an external repo (e.g., simple-evals), but keeps its history separate. You can choose one of two ways (A or B) to pull the simple-evals contents: (A) If You Cloned My Project Repo My project repo already includes simple-evals as a submodule, so you can just run: git submodule update --init --recursive (B) If You’re Adding It to a Newly Created ProjectTo manually add simple-evals as a submodule, run this: git submodule add https://github.com/openai/simple-evals.git simple_evals Note: The simple_evals at the end of the command (with an underscore) is crucial. It sets the folder name, and using a hyphen instead (i.e., simple–evals) can lead to import issues later. Final Step (For Both Methods) After pulling the repo contents, you must create an empty __init__.py in the newly created simple_evals folder so that it is importable as a module. You can create it manually, or use the following command: touch simple_evals/__init__.py (iii) Pull DeepSeek-R1 model via Ollama The next step is to locally download the distilled model of your choice (e.g., 14B) using this command: ollama pull deepseek-r1:14b The list of DeepSeek-R1 models available on Ollama can be found here. (iv) Define configuration We define the parameters in a configuration YAML file, as shown below: The model temperature is set to 0.6 (as opposed to the typical default value of 0). This follows DeepSeek’s usage recommendations, which suggest a temperature range of 0.5 to 0.7 (0.6 recommended) to prevent endless repetitions or incoherent outputs. Do check out the interestingly unique DeepSeek-R1 usage recommendations — especially for benchmarking — to ensure optimal performance when using DeepSeek-R1 models. EVAL_N_EXAMPLES is the parameter for setting the number of questions from the full 198-question set to use for evaluation. (v) Set up Sampler code To support Ollama-based language models within the simple-evals framework, we create a custom wrapper class named OllamaSampler saved inside utils/samplers/ollama_sampler.py. In this context, a sampler is a Python class that generates outputs from a language model based on a given prompt.  Since existing samplers in simple-evals only cover providers like OpenAI and Claude, we need a sampler class that provides a compatible interface for Ollama.  The OllamaSampler extracts the GPQA question prompt, sends it to the model with a specified temperature, and returns the plain text response.  The _pack_message method is included to ensure the output format matches what the evaluation scripts in simple-evals expect. (vi) Create evaluation run script The following code sets up the evaluation execution in main.py, including the use of the GPQAEval class from simple-evals to run GPQA benchmarking. The run_eval() function is a configurable evaluation runner that tests LLMs through Ollama on benchmarks like GPQA. It loads settings from the config file, sets up the appropriate evaluation class from simple-evals, and runs the model through a standardized evaluation process. It is saved in main.py, which can be executed with python main.py. Following the steps above, we have successfully set up and executed the GPQA-Diamond benchmarking on the DeepSeek-R1 distilled model. Wrapping It Up In this article, we showcased how we can combine tools like Ollama and OpenAI’s simple-evals to explore and benchmark DeepSeek-R1’s distilled models. The distilled models may not yet rival the 671B parameter original model on challenging reasoning benchmarks like GPQA-Diamond. Still, they demonstrate how distillation can expand access to LLM reasoning capabilities. Despite subpar scores in complex PhD-level tasks, these smaller variants may remain viable for less demanding scenarios, paving the way for efficient local deployment on a wider range of hardware. Before you go I welcome you to follow my GitHub and LinkedIn to stay updated with more engaging and practical content. Meanwhile, have fun benchmarking LLMs with Ollama and simple-evals! The post How to Benchmark DeepSeek-R1 Distilled Models on GPQA Using Ollama and OpenAI’s simple-evals appeared first on Towards Data Science.

Will we ever get The Last of Us: Part III? Naughty Dog's gaming franchise is back in the news lately thanks in part to the popular HBO show starring Pedro Pascal and Bella Ramsey, leading to questions about a possible continuation of the video game seriesNo new entries have been announced, but Naughty Dog developers over the years have dropped clues and made comments about returning to the post-apocalyptic world numerous times. To be sure, The Last of Us: Part III is not confirmed--and it may never happen. But there is a chance, so with that in mind, we're rounding up everything we've heard about a new Last of Us game.Table of Contents [hide]The Last Of Us: Part III Could Happen "If The Stars Align"The Last Of Us: Part III Could Happen "If The Stars Align"The most recent comments about The Last of Us: Part III came from Naughty Dog's Neil Druckmann, who wrote and directed the first two games. During a promotional interview for The Last of Us Season 2, Druckmann was inevitably asked about a third game. He said he's not ruling it out, but he's not confirming it either. Continue Reading at GameSpot

Streams of a song featured at the end of the newest episode of HBO's The Last of Us surged by more than 1,000% on Spotify, the streaming service has announced.The cover of Shawn James' "Through the Valley" by Ashley Johnson and Chris Rondinella jumped by more than 1,000% in the US since Episode 2 aired on Sunday, April 20. Developer Naughty Dog responded to Spotify's announcement with a heart emoji. Presumably streams of James' original version rose, too, but Spotify didn't release any stats about this.Johnson, of course, performed the role of Ellie in Naughty Dog's The Last of Us video games and originally recorded the cover of "Through the Valley" to promote The Last of Us: Part II in 2016.Continue Reading at GameSpot

Ghost of Tsushima’s sequel, Ghost of Yotei, originally revealed a 2025 release window when the game was announced. Now, Ghost of Yotei is committing to 2025 with a release date reveal of October 2, announced alongside a brand-new story trailer introducing Ghost of Yotei’s six key antagonists and a Collector’s Edition full of goodies (sans a physical copy of the game).

Anime can be hilarious, and there are some shows that prove this with incredible ease. That said, the real gems are usually those that feature characters who can carry the comedy of a show with their antics, dialogue, or just all-round personality.

Game Week Sale is live now! For a limited time*, you can enjoy price reductions across a selection of titles, including Dragon Ball Sparking Zero Deluxe Edition (20% off), EA Sports FC 25 (70% off), Dynasty Warriors: Origins (20% off) and many more. Head to PlayStation Store to discover your regional discount.  *Game Week Sale is live on PlayStation Store from April 23 at 00:00 AM JST and finishes May 7 at 11:59 PM JST. 

We are so excited to announce that Ghost of Yōtei comes to PS5 on October 2, 2025!  It’s been nearly five years since we shipped Ghost of Tsushima, and in that time we’ve been hard at work making Ghost of Yōtei something special. While the stories are unrelated, it’s important to us to make this a worthy follow-up to Jin’s journey, and we can’t wait for you to experience Atsu’s quest for vengeance later this year. Play Video Alongside today’s news, we’ve also released our latest trailer for Ghost of Yōtei, “The Onryō’s List.” Sixteen years ago in the heart of Ezo (called Hokkaido in present day), a gang of outlaws known as the Yōtei Six took everything from Atsu. They killed her family and left her for dead, pinned to a burning ginkgo tree outside her home. But Atsu survived. She learned to fight, to kill, and to hunt, and after years away she has returned to her home with a list of six names: The Snake, The Oni, The Kitsune, The Spider, The Dragon, and Lord Saito.  One by one, she’s hunting them down to avenge her family, armed with the same katana used to pin her to that burning tree all those years ago. But while Atsu’s story begins with vengeance, she’ll find there’s more to her journey than just revenge. As she explores Ezo, Atsu will meet unlikely allies and forge connections that help give her a new sense of purpose.  View and download image Download the image close Close Download this image View and download image Download the image close Close Download this image View and download image Download the image close Close Download this image View and download image Download the image close Close Download this image View and download image Download the image close Close Download this image View and download image Download the image close Close Download this image View and download image Download the image close Close Download this image View and download image Download the image close Close Download this image We hope the brief looks in today’s trailer give you a taste of what you can expect from Ghost of Yōtei. Beyond your first look at the Yōtei Six, you’ll spot some of the gorgeous scenery of Ezo as well as a handful of Atsu’s new weapons, a few of her allies, and even a new gameplay mechanic that will allow you to glimpse Atsu’s past and understand everything that was taken from her.  But we’ve only scratched the surface. In Ghost of Yōtei, we’ve built upon and evolved the way you explore the open world, offering even more freedom and variety than in Ghost of Tsushima. You’ll choose which leads to follow as you pick which Yōtei Six member you want to hunt down first. Atsu can also track other dangerous targets and claim bounties, or seek out weapon sensei to learn new skills. Ezo is wild, and as deadly as it is beautiful. As you trek across the open world you’ll find unexpected dangers and peaceful reprieves (including some returning activities from Tsushima), and you’ll be able to build a campfire anywhere in the open world for a rest under the stars. We want you to have the freedom to explore Ezo however you decide to, and we can’t wait to share more. View and download image Download the image close Close Download this image Today we can also reveal that pre-orders for Ghost of Yōtei begin next week, and you’ll be able to choose between multiple different editions. View and download image Download the image close Close Download this image First, if you pre-order any edition of Ghost of Yōtei you’ll receive a unique in-game mask, as well as a set of seven PSN avatars featuring concept art of Atsu and each member of the Yōtei Six at launch1. Pre-orders on PlayStation Store will receive the avatars immediately.  The standard edition (digital/disc) of Ghost of Yōtei will be SGD 97.90 / MYR 299 / IDR 1,029,000 / THB 2,290 / PHP 3,490 / VND 1,799,000 MSRP and will be available at retail or at PlayStation Store. View and download image Download the image close Close Download this image At PlayStation Store, you’ll also be able to pre-order the Ghost of Yōtei Digital Deluxe Edition for SGD 107 / MYR 339 / 1,169,000 / THB 2,690 MSRP. The Digital Deluxe Edition includes a digital copy of Ghost of Yōtei plus in-game bonuses including The Snake’s armor set, as well as an alternate dye for your starting armor. You’ll also receive a unique horse color and unique saddle dye, plus an in-game Charm, gold Sword Kit, and an early unlock of Traveler’s Maps, which allow you to find statues throughout the world to upgrade your skills.  View and download image Download the image close Close Download this image Finally, we are thrilled to reveal the Ghost of Yōtei Collector’s Edition. This packed edition includes all of the pre-order bonuses, all in-game items from the Digital Deluxe Edition, and a digital copy of the game2, as well as a replica display edition of Atsu’s Ghost mask. The mask is built to-scale with Jin’s mask from our Ghost of Tsushima Collector’s Edition. If you have both, they look great next to each other on a shelf! The mask measures 6.8 x 5 x 5.9 inches and is made of resin, plus includes its own display stand.  Also included is a replica of Atsu’s sash, complete with the names of all six members of the Yōtei Six (but you’ll have to cross them off yourself). The sash measures 71 inches long and is made of a cotton blend, a perfect cosplay accessory or wall decoration.  There’s also a replica of the Tsuba from Atsu’s katana, forged by her father in the image of two twin wolves. This Tsuba measures roughly 3 x 3 inches and also includes its own display stand.  In addition to all of the above, the Ghost of Yōtei Collector’s Edition also includes a pouch of coins and instructions to play Zeni Hajiki, a game of skill you’ll play throughout Ghost of Yōtei. There’s also a foldable papercraft ginkgo tree along with a wolf at its base, and a set of four 5 x 7-inch art cards featuring the sash, the wolf, Atsu’s Ghost mask, and our key art.  Pricing for Ghost of Yōtei Collector’s Edition in Southeast Asia will be announced on later date. We think this is the best Collector’s Edition we’ve ever produced, and we can’t wait for you to get your hands on it on October 2.  While pre-orders don’t open until May 2, you can wishlist Ghost of Yōtei right now at PlayStation Store and sign up to receive notifications as we release more information in the months to come.  We are so proud of Ghost of Yōtei and have many more exciting things we can’t wait to show you as we approach our release date. We are incredibly appreciative of all the support for Ghost of Tsushima and grateful for everyone who played, and hope you’ll look forward to following the wind once again on October 2.  1 Available via voucher code for Collector’s Edition and physical Standard Edition. Internet connection and an account for PlayStation Network required to redeem. 2 Digital items available via voucher code. Internet connection and an account for PlayStation™Network required to redeem.

Entei is joining the Max Battle hype in Pokémon Go, allowing you to take it on and possibly catch one that can Dynamax, following the appearance of its sibling (Raikou) last month. It’ll be featured during a Max Battle Weekend event, running from April 26 at 6 a.m. until April 27 at 9 p.m. in your local time. During this event, you’ll be able to claim more Max Particles each day (up to 1,600 from 800), you’ll get more Max Particles from Power Spots, the amount you have to walk to get particles will be cut down to 1/4, and Power Spots will refresh more often. From April 21 until the Max Battle Weekend ends, the particle cost to power up Max Moves has also been cut to 3/4 the usual amount, so it’s a great time to build some counters. Below, we list out some general Max Battle tips and counters for Dynamax Entei in Pokémon Go. Dynamax Entei weakness Entei is a solo fire-type Pokémon, so it has quite a few weaknesses: ground-, rock-, and water-type moves. There are a few Pokémon we could use at this point to take it down, which is nice (in comparison to some other big targets we’ve had, that only have one or two viable counters). All of that said, Entei will take less damage from bug-, steel-, fire-, grass-, ice-, and fairy-type moves, so you shouldn’t really use those unless you have to use them to build up your gauge. Dynamax Entei best counters For attackers, try to bring at least one of the following with its max attack built up: Gigantamax Kingler with Metal Claw Dynamax Inteleon with Water Gun Dynamax Excadrill with Mud Shot Dynamax Kingler with Bubble Gigantamax Blastoise with Water Gun This time around, none of Entei’s moves should cause any of these great distress, though you shouldn’t really be using these outside of the Dynamax/Gigantamax phase to dish damage. For tanks, you’ll want to bring one or two of the following with Max Guard/Spirit leveled up: Dynamax Blissey with Pound Dynamax/Gigantamax Blastoise with Water Gun Gigantamax Snorlax with Lick Notably there is a Timed Research set that rewards you with a Dynamax Sobble encounter, Sobble Candy, and tons of particles to help you get an Inteleon, if you want. Note that Inteleon will get a Gigantamax form down the line, so I only recommend investing in this if you have tons of Sobble Candy to spare. Ultimately, as usual, just bring the strongest Pokémon that you have to help contribute to the group. If you only have a Dynamax Sobble as a type-based counter, but you have a build Dynamax Gengar, you should just bring the Gengar. General Max Battle Tips If you’ve been struggling in Max Battles, here are some general tips to survive — and make sure you’re an asset to your team. While Gigantamax Battles are tough, a Legendary Max Battle shouldn’t be as tough as that. You can slack a little, but you should still heed our advice: Make sure you have enough players. High-efficiency players with maxed out investments will likely be able to solo Entei, but this isn’t going to be realistic for most people. Try to go at it with three other players for the highest chance of winning. Note that unlike Gigantamax Pokémon, the Dynamax Legendary Pokémon have a max party size of four (rather than 40). Don’t sleep on Max Spirit and Max Guard. Teams work best when there’s a variety of moves, not just maxed-out attacks. Each player should bring Pokémon with the defensive and healing moves unlocked as well. Max Guard will focus single-target damage towards you and reduce the damage taken and Max Spirit will heal the whole party, so these moves are really important to make sure your damage-dealers can keep dishing. Remember to swap to super effective moves when it’s time to Dynamax. For Dynamax Pokémon, their max moves are determined by whatever type their fast move is. This means if you have a Cryogonal with Frost Breath, it will know Max Hailstorm. A Gengar with Lick will know Max Phantasm. Take advantage of this and make sure to swap to a Pokémon that will deal super effective damage to your target before Dynamaxing, if you can. Focus on your fast moves. You want to spam your fast moves to build up that Dynamax meter and often times, using your charge move is actually a DPS loss when compared to the damage you could be doing with your max move will be. Spam those attacks! Level up a few ‘mons, but you don’t have to go too hard. Depending on your group size, you absolutely do not need to max out all your Dynamax Pokémon to level 40-50. While this will make it easier on the rest of your group, if this isn’t an investment you can make, you don’t have to stress about it. Power things up as high as you can afford to, but don’t fret if you don’t have a maxed out Pokémon. All that said, make sure to come as prepared as you can be. This is a team effort and there’s a chance that a full group of four can still fail. Do not just bring your unleveled Dynamax Wooloo expecting a free ride to a powerful Pokémon. (After all, if everyone does that, then you certainly won’t clear the battle.) Again, you don’t have to completely max out your Pokémon, but it will be better for everyone involved if you bring something helpful to the table. Keep your eye out for a shiny Entei! If you clear the raid, there is a chance that the Entei you catch will be shiny — which also means it’ll be a guaranteed catch. Use a Pinap Berry to score extra candy if you get lucky enough to find a sparkly Entei.