banner
2han99

2han99

2024-08-31 35 weeks

🎞️ Life#

Huangshan Trip#

The last time I went to Huangshan was in 2016, when I was still a carefree college student.

Welcoming Pine#

Welcoming Pine

One-Line Sky#

One-Line Sky

Lotus Peak, Unreachable#

Lotus Peak, Unreachable

Overlooking Guangming Peak#

Overlooking Guangming Peak

Dangerous, Not Recommended to Imitate

Peacock Worshiping Guanyin#

Peacock Worshiping Guanyin

Entrance Bridge of Hongcun Village#

Entrance Bridge of Hongcun Village

Wang's Ancestral Hall, Second-Class Enrong Archway#

Wang's Ancestral Hall, Second-Class Enrong Archway

Beautiful Courtyard, Admission Fee Required#

Beautiful Courtyard, Admission Fee Required

Little Black Child#

Little Black Child

Black Myth: Wukong#

Currently, I have cleared the fourth chapter, except for the little yellow dragon, I have defeated all the other bosses.

By the way, I also spent more than 40 minutes farming Tiger Skin Tassels.

Poisonous Enemy King#

This is a difficult boss to defeat. It took me a long time because I rarely use invisibility skills. If you don't use it, you will easily be continuously attacked and poisoned.

I also discovered here that both invisibility and transformation skills can remove debuffs.

Dark Moon Demon Lord#

It is very difficult to defeat normally. First of all, it is large in size and often retreats. It has a long recovery time, and it is difficult to hit with the staff's second hit and charged staff attack.

It has a lot of health and has a second phase, and the immobilization skill is not very effective.

So I switched to the Forbidden Spell and immediately recovered.

Here, I defeated the Poisonous Enemy King first, without triggering the plot, but it still deducted a portion of the boss's health.

Fourth Sister#

Waiting for a Hermit Crane Boss.

Hundred-Eyed Demon Lord#

Easy, as long as you go to the hidden map first and interrupt its attacks with the Embroidery Needle, it is easy to defeat.

"Troublesome Taibai Gold Star"#

Getting ready to watch.

💭 Clips#

Infinite Refill Cursor#

If you ask me what code editor to use in the AI era, I will undoubtedly answer you: Cursor Cursor Cursor!!!

It has intelligent tab completion, file-level code reading, and AI capabilities deeply integrated with the editor. With the powerful Claude 3.5 Sonnet model, it quickly becomes the coding artifact for programmers. As the saying goes, "With Cursor in hand, there are no bugs."

I suggest everyone refill their cups reasonably, create a new account after using up the trial period, and be gentle with the usage~

New users will have a 14-day trial period for Cursor when they first register. If you want to refill your cup indefinitely, you need to use a new email to register a new account on the official website, and then log in again in the software to continue using the Pro membership. Changing accounts will not change the configuration in the editor, so you can switch with confidence.

Source

Give it a try.

📰 Weekly Digest#

Ratio of Male and Female Restrooms in Train Stations Changed to 1:2, Receives Positive Feedback, Media Calls for Progress in More Areas#

Recently, the National Railway Administration issued an industry standard announcement, making partial revisions to the "Design Specifications for Railway Passenger Stations" TB 10100-2018 (referred to as the "Specifications"), further regulating the planning and construction of railway passenger stations. Among them, the new version of the "Specifications" has modified the requirements for the ratio and quantity of male and female restrooms, stipulating that the ratio of male to female restrooms should be 1:2. This adjustment to the ratio of male and female restrooms in train stations, once announced, has attracted widespread attention. Many netizens believe that this move positively responds to the "difficult problem" of gender differences in public places.

Source

Personally, I think it would be better to deploy them separately every other female restroom, with male restrooms having only urinals and complete male restrooms to avoid difficulties in finding male restrooms, while urinals take up very little space.

IBM Closes Chinese R&D Department, Affects Over a Thousand People, Company Responds with "Operational Adjustments as Needed"#

According to First Financial, the closure of the R&D department by IBM mainly involves two business lines, one is IBM China Development Center (CDL), and the other is IBM China Systems Center (CSL), which are mainly responsible for research and testing. IBM will provide N+3 compensation for affected employees.

Source

IBM has lost face for foreign companies, the compensation is so low.

New Regulations on Wireless Charging to Be Implemented on September 1: MagSafe / Qi2 May Stop Selling Due to Frequency Conflict, Apple iPhone Wireless Power Reduced to 7.5W#

"The Interim Regulations on the Radio Management of Wireless Charging (Power Transmission) Equipment" will be implemented from September 1, 2024. From the date of implementation, the production or import of wireless charging equipment that does not meet the requirements specified in the regulations and is sold or used domestically will be stopped. Wireless charging equipment produced or imported before this date can continue to be sold and used until they are scrapped. However, the frequencies used by MagSafe and Qi2 protocols, 360kHz, are not within the specified working frequency range. In response to a message from a netizen, the Ministry of Industry and Information Technology stated that the 325-405KHz frequency band is allocated for aviation radio navigation services, and this frequency band cannot be used for wireless charging devices in China. On the wireless charging device declaration page of Apple's official website in China, the wireless charging rated power of the iPhone has been changed to 7.5W, and the working frequency is 127.7kHz.

Source

Well... I guess there were no relevant regulations when the products were launched.

Xiaopeng Mona M03 Model Prices Announced: Ranging from 11.98 to 15.58 Thousand Yuan#

Xiaopeng Motors has launched the new Mona M03, claiming to be the "first A-class intelligent pure electric hatchback under 200,000 yuan in China":

  • 515km long-range version priced at 11.98 thousand yuan

  • 620km ultra-long-range version priced at 12.98 thousand yuan

  • 580km ultra-long-range Max version priced at 15.58 thousand yuan

The Mona M03 will officially start delivery to customers in September this year, and the M03 Max with advanced urban intelligent driving capabilities will start delivery after the Spring Festival in 2025.

Source

Entry-level car, looks very good.

Five Weeks of Adjusted Days Off for Mid-Autumn Festival and National Day Sparks Online Discussion#

Adjusted Days Off

With the Mid-Autumn Festival and National Day holidays approaching, the complex arrangement of adjusted days off has sparked online discussions.

According to the official holiday notice, during this year's Mid-Autumn Festival, there will be a 3-day adjusted day off from September 15th to September 17th, with September 14th (Saturday) being a workday.

As for National Day, there will be a 7-day adjusted day off from October 1st to 7th, with September 29th (Sunday) and October 12th (Saturday) being workdays.

In summary, the schedule is as follows: "Work 6 days, off 3 days; work 3 days, off 2 days; work 5 days, off 1 day; work 2 days, off 7 days; and finally, work 5 days, off 1 day", requiring a total of 5 weeks of adjusted days off.

Source

Stimulating consumption to save the national economy while restricting people's vacations, the upper-level officials are contradictory.

Guo Mingchi: Only iPhone 17 Pro Max Will Have 12GB RAM in New Models Next Year, Stronger On-Device AI as the Main Selling Point#

Guo Mingchi

iPhone 17 Pro Max

Source

Well, this is a dilemma.

📚 Articles#

Meta Releases Three Most Detailed Llama Fine-Tuning Guides! Thousand-Word Long Articles, Essential for Zero-Based Beginners#

Meta has released three technical articles in a row, starting with adapting large models, introducing how to fine-tune LLM with specific domain data, how to determine fine-tuning for your own use cases, and the empirical rules for managing a good training dataset.

https://ai.meta.com/blog/adapting-large-language-models-llms/

https://ai.meta.com/blog/when-to-fine-tune-llms-vs-other-techniques/

https://ai.meta.com/blog/how-to-fine-tune-llms-peft-dataset-curation/

Loading...
Ownership of this post data is guaranteed by blockchain and smart contracts to the creator alone.