Create README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,83 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
language:
|
| 4 |
+
- multilingual
|
| 5 |
+
- ar
|
| 6 |
+
- bg
|
| 7 |
+
- ca
|
| 8 |
+
- cs
|
| 9 |
+
- da
|
| 10 |
+
- de
|
| 11 |
+
- el
|
| 12 |
+
- en
|
| 13 |
+
- es
|
| 14 |
+
- et
|
| 15 |
+
- fa
|
| 16 |
+
- fi
|
| 17 |
+
- fr
|
| 18 |
+
- gl
|
| 19 |
+
- gu
|
| 20 |
+
- he
|
| 21 |
+
- hi
|
| 22 |
+
- hr
|
| 23 |
+
- hu
|
| 24 |
+
- hy
|
| 25 |
+
- id
|
| 26 |
+
- it
|
| 27 |
+
- ja
|
| 28 |
+
- ka
|
| 29 |
+
- ko
|
| 30 |
+
- ku
|
| 31 |
+
- lt
|
| 32 |
+
- lv
|
| 33 |
+
- mk
|
| 34 |
+
- mn
|
| 35 |
+
- mr
|
| 36 |
+
- ms
|
| 37 |
+
- my
|
| 38 |
+
- nb
|
| 39 |
+
- nl
|
| 40 |
+
- pl
|
| 41 |
+
- pt
|
| 42 |
+
- ro
|
| 43 |
+
- ru
|
| 44 |
+
- sk
|
| 45 |
+
- sl
|
| 46 |
+
- sq
|
| 47 |
+
- sr
|
| 48 |
+
- sv
|
| 49 |
+
- th
|
| 50 |
+
- tr
|
| 51 |
+
- uk
|
| 52 |
+
- ur
|
| 53 |
+
- vi
|
| 54 |
+
---
|
| 55 |
+
|
| 56 |
+
# paraphrase-multilingual-MiniLM-L12-v2.gguf
|
| 57 |
+
|
| 58 |
+
```py
|
| 59 |
+
import torch
|
| 60 |
+
from llama_cpp import Llama
|
| 61 |
+
from sentence_transformers import SentenceTransformer
|
| 62 |
+
from scipy.spatial.distance import cosine
|
| 63 |
+
|
| 64 |
+
model = SentenceTransformer(
|
| 65 |
+
"paraphrase-multilingual-MiniLM-L12-v2",
|
| 66 |
+
model_kwargs={"torch_dtype": torch.float16}
|
| 67 |
+
)
|
| 68 |
+
llm = Llama.from_pretrained(
|
| 69 |
+
"mykor/paraphrase-multilingual-MiniLM-L12-v2.gguf",
|
| 70 |
+
filename="paraphrase-multilingual-MiniLM-L12-118M-v2-F16.gguf",
|
| 71 |
+
embedding=True,
|
| 72 |
+
verbose=False,
|
| 73 |
+
)
|
| 74 |
+
|
| 75 |
+
text = "「끝내고 싶어」라고 말하니, 네가 처음으로 웃었어"
|
| 76 |
+
embed1 = model.encode(text)
|
| 77 |
+
embed2 = llm.embed(text)
|
| 78 |
+
print(cosine(embed1, embed2))
|
| 79 |
+
```
|
| 80 |
+
|
| 81 |
+
```sh
|
| 82 |
+
0.0011532125434331464
|
| 83 |
+
```
|