mirror of
https://github.com/blakeblackshear/frigate.git
synced 2026-04-27 17:17:40 +03:00
Adjust docs
This commit is contained in:
parent
16fb68d859
commit
f9c0999ad3
@ -29,12 +29,11 @@ You must use a vision capable model with Frigate. Current model variants can be
|
|||||||
|
|
||||||
The following models are recommended:
|
The following models are recommended:
|
||||||
|
|
||||||
| Model | Size | Recommended Features |
|
| Model | Size | Notes |
|
||||||
| ----------------- | ------ | -------------------- |
|
| ----------------- | ------ | ----------------------------------------------------------- |
|
||||||
| `minicpm-v:8b` | 5.5 GB | Review Summary |
|
| `gemma3:4b` | 3.3 GB | Strong frame-to-frame understanding, slower inference times |
|
||||||
| `qwen2.5vl:3b` | 3.2 GB | Review Summary |
|
| `qwen2.5vl:3b` | 3.2 GB | Fast but capable model with good vision comprehension |
|
||||||
| `gemma3:4b` | 3.3 GB | All Features |
|
| `llava-phi3:3.8b` | 2.9 GB | Lightweight and fast model with vision comprehension |
|
||||||
| `llava-phi3:3.8b` | 2.9 GB | All Features |
|
|
||||||
|
|
||||||
:::note
|
:::note
|
||||||
|
|
||||||
|
|||||||
Loading…
Reference in New Issue
Block a user