nvidia
/

Llama-3.1-70B-Instruct-FP8

Text Generation

text-generation-inference

Model card Files Files and versions

zhiyucheng commited on Jan 10

Commit

20af8c9

·

1 Parent(s): 57d2a4b

use original tokenizer config

Files changed (1) hide show

tokenizer_config.json +1 -3

tokenizer_config.json CHANGED Viewed

@@ -2057,8 +2057,6 @@
     "input_ids",
     "attention_mask"
   ],
-  "model_max_length": 2048,
-  "pad_token": "<|eot_id|>",
-  "padding_side": "left",
   "tokenizer_class": "PreTrainedTokenizerFast"
 }

     "input_ids",
     "attention_mask"
   ],
+  "model_max_length": 131072,
   "tokenizer_class": "PreTrainedTokenizerFast"
 }