Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
Marius_Manola
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
6 ms
·
31.
▲
by
Marius_Manola
2y ago
It was. But the dataset for training is rather small and focuses only on the first book. I might improve it in the future
32.
▲
by
Marius_Manola
2y ago
thank you for your kind words. Will be making it even better (making the fine tuning dataset larger and more similar to how the guide would answer) if there'll be enough people wanting that.
33.
▲
by
Marius_Manola
2y ago
but trained on synthetic data with elements from the first book
34.
▲
Show HN: The Hitchhiker's Guide to the Galaxy Online
(the42.life)
81 points
by
Marius_Manola
2y ago
|
25 comments
35.
▲
Show HN: Craft fine-tuning datasets for GPT in 10 minutes
(prompteasy.ai)
1 points
by
Marius_Manola
2y ago
|
0 comments
36.
▲
Show HN: Fine-tuning AI models with AI in <5 minutes
(prompteasy.ai)
2 points
by
Marius_Manola
2y ago
|
0 comments
37.
▲
Ask HN: Do you fine-tune LLMs at all?
2 points
by
Marius_Manola
2y ago
|
1 comments
38.
▲
Will fine-tuning LLMs be relevant?
1 points
by
Marius_Manola
2y ago
|
0 comments