Add tokenizer_config.json
Browse files- tokenizer_config.json +25 -0
tokenizer_config.json
ADDED
|
@@ -0,0 +1,25 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"tokenizer_class": "NeuralQuantumTokenizer",
|
| 3 |
+
"auto_map": {
|
| 4 |
+
"AutoTokenizer": "tokenization_nqlm.NeuralQuantumTokenizer"
|
| 5 |
+
},
|
| 6 |
+
"vocab_size": 50257,
|
| 7 |
+
"model_max_length": 512,
|
| 8 |
+
"padding_side": "right",
|
| 9 |
+
"truncation_side": "right",
|
| 10 |
+
"special_tokens": {
|
| 11 |
+
"bos_token": "<|endoftext|>",
|
| 12 |
+
"eos_token": "<|endoftext|>",
|
| 13 |
+
"unk_token": "<|endoftext|>",
|
| 14 |
+
"pad_token": "<|endoftext|>",
|
| 15 |
+
"quantum_token": "<|quantum|>",
|
| 16 |
+
"classical_token": "<|classical|>"
|
| 17 |
+
},
|
| 18 |
+
"clean_up_tokenization_spaces": true,
|
| 19 |
+
"model_input_names": [
|
| 20 |
+
"input_ids",
|
| 21 |
+
"attention_mask"
|
| 22 |
+
],
|
| 23 |
+
"quantum_enhanced": true,
|
| 24 |
+
"quantum_token_processing": true
|
| 25 |
+
}
|