format string | key string | tokenizer string | vocab_size int64 | pad_id int64 | seq_len int64 | sequences int64 | seed int64 | shuffle_merged bool | shard_sequences int64 | shards int64 | datasets list |
|---|---|---|---|---|---|---|---|---|---|---|---|
datamill/packed-v1 | 5b519ee0915846fa59aa4023 | tiktoken:tokenizer/tiktoken.model | 32,768 | 32,767 | 4,096 | 4,290,317 | 3 | true | 32,768 | 131 | [
{
"name": "computer-science",
"cache": "2b0d5d60cb2f654700923728",
"n_docs": 310317,
"n_tokens": 427285187,
"weight": 1
},
{
"name": "ultra-fineweb",
"cache": "f2ec8b1a474fed50f626450b",
"n_docs": 1833560,
"n_tokens": 1000000032,
"weight": 1
},
{
"name": "math-qa"... |
No dataset card yet
- Downloads last month
- 17