Upload 6 files
Browse files- config.json +8 -0
- model.safetensors +3 -0
- special_tokens_map.json +5 -0
- tokenizer.json +56 -0
- tokenizer_config.json +9 -0
- vocab.json +27 -0
config.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"model_type": "mini_transformer",
|
| 3 |
+
"vocab_size": 25,
|
| 4 |
+
"embedding_dim": 128,
|
| 5 |
+
"num_layers": 1,
|
| 6 |
+
"num_heads": 2,
|
| 7 |
+
"max_seq_len": 256
|
| 8 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:407f304d9bcbe50b292cf651077d3191a7e627328d1fddca44e44d5426fd8d49
|
| 3 |
+
size 92124
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,5 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"pad_token": "<PAD>",
|
| 3 |
+
"bos_token": "<BOS>",
|
| 4 |
+
"eos_token": "<EOS>"
|
| 5 |
+
}
|
tokenizer.json
ADDED
|
@@ -0,0 +1,56 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"stoi": {
|
| 3 |
+
"\n": 0,
|
| 4 |
+
" ": 1,
|
| 5 |
+
".": 2,
|
| 6 |
+
"S": 3,
|
| 7 |
+
"T": 4,
|
| 8 |
+
"V": 5,
|
| 9 |
+
"a": 6,
|
| 10 |
+
"c": 7,
|
| 11 |
+
"e": 8,
|
| 12 |
+
"f": 9,
|
| 13 |
+
"g": 10,
|
| 14 |
+
"h": 11,
|
| 15 |
+
"i": 12,
|
| 16 |
+
"k": 13,
|
| 17 |
+
"l": 14,
|
| 18 |
+
"m": 15,
|
| 19 |
+
"n": 16,
|
| 20 |
+
"o": 17,
|
| 21 |
+
"r": 18,
|
| 22 |
+
"s": 19,
|
| 23 |
+
"t": 20,
|
| 24 |
+
"u": 21,
|
| 25 |
+
"v": 22,
|
| 26 |
+
"w": 23,
|
| 27 |
+
"y": 24
|
| 28 |
+
},
|
| 29 |
+
"itos": {
|
| 30 |
+
"0": "\n",
|
| 31 |
+
"1": " ",
|
| 32 |
+
"2": ".",
|
| 33 |
+
"3": "S",
|
| 34 |
+
"4": "T",
|
| 35 |
+
"5": "V",
|
| 36 |
+
"6": "a",
|
| 37 |
+
"7": "c",
|
| 38 |
+
"8": "e",
|
| 39 |
+
"9": "f",
|
| 40 |
+
"10": "g",
|
| 41 |
+
"11": "h",
|
| 42 |
+
"12": "i",
|
| 43 |
+
"13": "k",
|
| 44 |
+
"14": "l",
|
| 45 |
+
"15": "m",
|
| 46 |
+
"16": "n",
|
| 47 |
+
"17": "o",
|
| 48 |
+
"18": "r",
|
| 49 |
+
"19": "s",
|
| 50 |
+
"20": "t",
|
| 51 |
+
"21": "u",
|
| 52 |
+
"22": "v",
|
| 53 |
+
"23": "w",
|
| 54 |
+
"24": "y"
|
| 55 |
+
}
|
| 56 |
+
}
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"tokenizer_type": "char",
|
| 3 |
+
"vocab_file": "vocab.json",
|
| 4 |
+
"do_lower_case": false,
|
| 5 |
+
"unk_token": "<UNK>",
|
| 6 |
+
"pad_token": "<PAD>",
|
| 7 |
+
"bos_token": "<BOS>",
|
| 8 |
+
"eos_token": "<EOS>"
|
| 9 |
+
}
|
vocab.json
ADDED
|
@@ -0,0 +1,27 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"\n": 0,
|
| 3 |
+
" ": 1,
|
| 4 |
+
".": 2,
|
| 5 |
+
"S": 3,
|
| 6 |
+
"T": 4,
|
| 7 |
+
"V": 5,
|
| 8 |
+
"a": 6,
|
| 9 |
+
"c": 7,
|
| 10 |
+
"e": 8,
|
| 11 |
+
"f": 9,
|
| 12 |
+
"g": 10,
|
| 13 |
+
"h": 11,
|
| 14 |
+
"i": 12,
|
| 15 |
+
"k": 13,
|
| 16 |
+
"l": 14,
|
| 17 |
+
"m": 15,
|
| 18 |
+
"n": 16,
|
| 19 |
+
"o": 17,
|
| 20 |
+
"r": 18,
|
| 21 |
+
"s": 19,
|
| 22 |
+
"t": 20,
|
| 23 |
+
"u": 21,
|
| 24 |
+
"v": 22,
|
| 25 |
+
"w": 23,
|
| 26 |
+
"y": 24
|
| 27 |
+
}
|