echoself / tokenizer_config.json

Deploy NanEcho CI checkpoint (4L/4H/256E, 200 iters, val_loss=1.9258)

ee242c7 verified 9 months ago

196 Bytes

{
  "tokenizer_class": "GPT2Tokenizer",
  "model_max_length": 1024,
  "bos_token": "<|endoftext|>",
  "eos_token": "<|endoftext|>",
  "unk_token": "<|endoftext|>",
  "pad_token": "<|endoftext|>"
}

	{
	"tokenizer_class": "GPT2Tokenizer",
	"model_max_length": 1024,
	"bos_token": "<\|endoftext\|>",
	"eos_token": "<\|endoftext\|>",
	"unk_token": "<\|endoftext\|>",
	"pad_token": "<\|endoftext\|>"
	}