tokenizer unigram_bi_zh
Browse files
tokenizers/unigram_bi_zh/meta.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"flavor": "unigram",
|
| 3 |
+
"condition": "bi_zh",
|
| 4 |
+
"vocab_size": 65536,
|
| 5 |
+
"specials": [
|
| 6 |
+
"<unk>",
|
| 7 |
+
"<bos>",
|
| 8 |
+
"<eos>",
|
| 9 |
+
"<pad>"
|
| 10 |
+
],
|
| 11 |
+
"corpus_files": [
|
| 12 |
+
"/mnt/scratch/xscript/tok_corpora/bi_zh/en.txt",
|
| 13 |
+
"/mnt/scratch/xscript/tok_corpora/bi_zh/zh.txt"
|
| 14 |
+
]
|
| 15 |
+
}
|
tokenizers/unigram_bi_zh/sp.model
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:80b9959e0c9fd131f84cb4d2097ff3fee34b411e2619388f96b7b0412ef682fa
|
| 3 |
+
size 1258239
|