/haspburg-llm

haspburg-llm is a built from scratch nanoGPT modification that iteratively retrains on the output produced by itself

Primary LanguagePython

haspburg-llm

Haspburg llm is a built from scratch nano gpt modification which iteratively trains on its own output

to give you a frame of reference this is what 5 generations of training does to haspburg-llm

note: these models were trained for a 5-10 mins on an intel i7 1260p. to realise the full capability and breakdown of the llm i'd recommend running this on a single 8XA100 40GB node in about 4 days of training per generation (DONT DO IT, MASSIVE WASTE OF RESOURCES) this project was done for fun, not for practical implementation

gen 0

ANGELO:
And cowards it be strawn to my bed,
And thrust the gates of my threats,
Because he that ale away, and hang'd
An one with him.

DUKE VINCENTIO:
I thank your eyes against it.

DUKE VINCENTIO:
Then will answer him to save the malm:
And what have you tyrannous shall do this?

DUKE VINCENTIO:
If you have done evils of all disposition
To end his power, the day of thrust for a common men
That I leave, to fight with over-liking
Hasting in a roseman.

gen 5

DEKL AMCSJLE:
te
tGia  t rBhOnrtfst: p h
didorTna onoaaoLTe nEUar
yUn a ,L
arr gToih  wr;u tey:eour.
uSaoieoo wkoa e idvo f Ho! dn

 .'b hotoeL  s Oo : inSo  S
or:tUoe vhe  s sohMne h ooe f: ot, si, pe o,ar de  e rit oA
oonrea hteahfo
 t'hoiaurt 

oid   t  I
iT.to do  nE koo dZ
 ie aiEo
einti e g v
 nC t  alSddlaen nu t ho:
  s
 d. m n d
NTeItie ol y loesesueo oaol r ess

To execute run the scripts.sh file