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 | f67d7cff3a895b5e1a272438 | tiktoken:tokenizer/tiktoken.model | 32,768 | 32,767 | 2,048 | 10,592,711 | 3 | true | 16,384 | 647 | [
{
"name": "web",
"cache": "37d00ce5eaa2453cf571997a",
"n_docs": 16194654,
"n_tokens": 20000000046,
"weight": 1
},
{
"name": "books3",
"cache": "bcaf16a2a5a8662510b548ec",
"n_docs": 7416,
"n_tokens": 1000008306,
"weight": 1
},
{
"name": "computer-science",
"cac... |
No dataset card yet
- Downloads last month
- 42