Hacker News
new
|
past
|
comments
|
ask
|
show
|
jobs
|
submit
login
croemer
10 months ago
|
parent
|
context
|
favorite
| on:
Qwen3: Think deeper, act faster
Seems to truncate the input to only 2048 input tokens
simonw
10 months ago
[–]
Oops! That's an Ollama default setting. You can fix that by increasing the num_ctx setting - I'll try running this again.
The num_predict setting controls output size.
Guidelines
|
FAQ
|
Lists
|
API
|
Security
|
Legal
|
Apply to YC
|
Contact
Search: