Profile Picture
  • All
  • Search
  • Images
  • Videos
    • Shorts
  • Maps
  • News
  • More
    • Shopping
    • Flights
    • Travel
  • Notebook
Report an inappropriate content
Please select one of the options below.
NVIDIA
NVIDIA Ai
NVIDIA
News Today
NVIDIA
CEO
GTC
NVIDIA
NVIDIA Ai
New Update
NVIDIA
H100
NVIDIA
Stock Price Today
Halo
Ai
Dwts Gleb Savchenko
Jensen Huang
Nvda Stocks
Supercomputer
Whoops
Breakthrough Full Movie Free
ABB Robotics
NVIDIA
NVIDIA
Stock
Archviz Ai
Results Video
NVIDIA
to Invest $ 5 Billion in Intel
Nvda Forecast
Jensen
NVIDIA
CNBC
NVIDIA
Julie Bowen Jimmy Kimmel
Ai
Surveillance Grid
GPU Stock
NVIDIA
Today Stock
Nvda Stock
NVIDIA
Stock Price News
Car Videos
NVIDIA
GTC
  • Length
    AllShort (less than 5 minutes)Medium (5-20 minutes)Long (more than 20 minutes)
  • Date
    AllPast 24 hoursPast weekPast monthPast year
  • Resolution
    AllLower than 360p360p or higher480p or higher720p or higher1080p or higher
  • Source
    All
    Dailymotion
    Vimeo
    Metacafe
    Hulu
    VEVO
    Myspace
    MTV
    CBS
    Fox
    CNN
    MSN
  • Price
    AllFreePaid
  • Clear filters
  • SafeSearch:
  • Moderate
    StrictModerate (default)Off
Filter
    NVIDIA
    NVIDIA Ai
    NVIDIA
    News Today
    NVIDIA
    CEO
    GTC
    NVIDIA
    NVIDIA Ai
    New Update
    NVIDIA
    H100
    NVIDIA
    Stock Price Today
    Halo
    Ai
    Dwts Gleb Savchenko
    Jensen Huang
    Nvda Stocks
    Supercomputer
    Whoops
    Breakthrough Full Movie Free
    ABB Robotics
    NVIDIA
    NVIDIA
    Stock
    Archviz Ai
    Results Video
    NVIDIA
    to Invest $ 5 Billion in Intel
    Nvda Forecast
    Jensen
    NVIDIA
    CNBC
    NVIDIA
    Julie Bowen Jimmy Kimmel
    Ai
    Surveillance Grid
    GPU Stock
    NVIDIA
    Today Stock
    Nvda Stock
    NVIDIA
    Stock Price News
    Car Videos
    NVIDIA
    GTC
You now convert any LLM into a faster one without retraining from scratch.NVIDIA just did this to their 30B model. Here's the trick:1. Duplicate the model into two copies2. Freeze one copy, it just reads the prompt and remembers context3. Train the other copy to write chunks of text at once instead of one word at a time4. Run them togetherThe frozen copy barely costs anything (it's already trained). The new copy only needed ~8% of the original training data to learn the new trick.Result: 2.4x fa
0:13
You now convert any LLM into a faster one without retraining from scratch.NVIDIA just did this to their 30B model. Here's the trick:1. Duplicate the model into two copies2. Freeze one copy, it just reads the prompt and remembers context3. Train the other copy to write chunks of text at once instead of one word at a time4. Run them togetherThe frozen copy barely costs anything (it's already trained). The new copy only needed ~8% of the original training data to learn the new trick.Result: 2.4x fa
103.4K views1 day ago
x.comLior Alexander
See more
Static thumbnail place holder
More like this
  • Privacy
  • Terms