qwen_cnn_dailymail / tokenizer_config.json
imdatta0's picture
Training in progress, step 100
318254a
raw
history blame contribute delete
226 Bytes
{
"auto_map": {
"AutoTokenizer": [
"Qwen/Qwen-14B--tokenization_qwen.QWenTokenizer",
null
]
},
"clean_up_tokenization_spaces": true,
"model_max_length": 8192,
"tokenizer_class": "QWenTokenizer"
}