Create tokenizer_config.json
Browse files- tokenizer_config.json +18 -0
tokenizer_config.json
ADDED
|
@@ -0,0 +1,18 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"tokenizer_name": "CSUMLM Tokenizer",
|
| 3 |
+
"model_name": "CSUMLM",
|
| 4 |
+
"description": "Tokenizer for the CognoSphere Unified Multimodal Language Model",
|
| 5 |
+
"author": "Or4cl3 AI Solutions",
|
| 6 |
+
"language": "Multimodal (Text, Image, Audio)",
|
| 7 |
+
"vocab_size": 32000,
|
| 8 |
+
"max_sequence_length": 512,
|
| 9 |
+
"special_tokens": {
|
| 10 |
+
"bos_token": "<BOS>",
|
| 11 |
+
"eos_token": "<EOS>",
|
| 12 |
+
"pad_token": "<PAD>",
|
| 13 |
+
"unk_token": "<UNK>",
|
| 14 |
+
"mask_token": "<MASK>"
|
| 15 |
+
},
|
| 16 |
+
"tokenization_method": "Byte Pair Encoding (BPE)",
|
| 17 |
+
"training_data": "Custom 1500 Example Dataset"
|
| 18 |
+
}
|