5049 位用户此时在线
24小时点击排行 Top 10：
- 本站自动实时分享网络热点
- 24小时实时更新
- 所有言论不代表本站态度
- 欢迎对信息踊跃评论评分
- 评分越高，信息越新，排列越靠前
2
1
0
A new method was able to delete 40% of LLM layers with no drop in accuracy.
This makes them mich cheaper and faster.
The method combines pruning, quantization and PEFT.
They tested this across various open source models.
Each family of models had a maximum amount of layers…
时政
(
twitter.com
)
由
Lior⚡
提交
A new method was able to delete 40% of LLM layers with no drop in accuracy.
This makes them mich cheaper and faster.
The method combines pruning, quantization and PEFT.
They tested this across various open source models.
Each family of models had a maximum amount of layers that could be deleted before accuracy dropped:
- Mistral - 30%
- Lama 70B - 40%
- Llama 13B - 50%
Great paper.
点击图片查看原图
1
Meet the little unsung hero 🥹🙏
3
2
2
Meet the little unsung hero 🥹🙏
- She's Nina
- Captain Smit was being attacked by the Omani muslim co pilot
- No one had a single idea about what was happening
- Nina went to the bathroom
- She heard the screams of the captain
- The little girl immediately ran to her mother
-
时政
(
twitter.com)
00:00:12
2
🚨BREAKING: ICE wrongfully deported a 25-year-old U.S. BORN citizen to Mexico… and he had to SUE the federal government to finally be allowed back into the United States.
2
1
1
🚨BREAKING: ICE wrongfully deported a 25-year-old U.S. BORN citizen to Mexico… and he had to SUE the federal government to finally be allowed back into the United States.
Brian José Morales García repeatedly told authorities that he was born in Denver.
He told them he was an
时政
(
twitter.com)
00:00:24
3
Whatever. . . The President was literally asked about the VP’s tweet and Trump agreed with it, saying that the war and the prices as a result of it were the primary reason for the midterm losses. And he said he liked JD’s tweet and thought it was good. No undermining here
2
1
1
Whatever. . . The President was literally asked about the VP’s tweet and Trump agreed with it, saying that the war and the prices as a result of it were the primary reason for the midterm losses. And he said he liked JD’s tweet and thought it was good. No undermining here
btc
(
twitter.com)
•
Ned Ryun
4
2
1
1
5
2
1
1
6
2
1
1
7
All the armchair experts claimed the Cybertruck has no crumple zone and I get it, the proportions seem impossible. It was a tough one and there is a lot of engineering that went into it. Let me break it down for you:
2
1
1
All the armchair experts claimed the Cybertruck has no crumple zone and I get it, the proportions seem impossible. It was a tough one and there is a lot of engineering that went into it. Let me break it down for you:
Here you can see the large single piece casting in action in a
btc
(
twitter.com)
00:00:09
8
Elon Musk: As a system, Starship is thousands of times more capable than Saturn V.
2
1
1
Elon Musk: As a system, Starship is thousands of times more capable than Saturn V.
“Starship is 2.5 times the thrust of a Saturn V. Saturn V was the largest rocket ever to reach orbit. This is 2.5 times more power than that, and we build one a month. And they're meant to be
btc
(
twitter.com)
00:00:18
9
2
1
1
10
He had never drawn a paycheck as a professional firefighter. On the morning of 11 September 2001 he became the man in the red bandana.
2
1
1
He had never drawn a paycheck as a professional firefighter. On the morning of 11 September 2001 he became the man in the red bandana.
Welles Crowther was twenty-four and at his desk as a stock trader on the 104th floor of the South Tower when the attacks began. He did not leave
有趣
(
twitter.com)
•
Massimo
11
2
1
1
12
2
1
1
13
2
1
1
14
2
1
1
15
2
1
1
16
2
1
1
17
2
1
1
18
2
1
1
19
2
1
1
20
2
1
1
21
2
1
1
22
3
2
2
23
2
1
1
24
2
1
1
25
2
1
1