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 | 645d63d0732c2184784bc1ed | tiktoken:tokenizer/tiktoken.model | 32,768 | 32,745 | 4,096 | 15,040,994 | 1 | true | 131,072 | 115 | [
{
"name": "computer-science",
"cache": "a0e07cf4fb6549f958bfd505",
"n_docs": 310317,
"n_tokens": 430290820,
"weight": 1
},
{
"name": "physics",
"cache": "77466f3a3bd3d73cfcef06a1",
"n_docs": 20000,
"n_tokens": 11261737,
"weight": 1
},
{
"name": "physics-corpus",
... |
No dataset card yet
- Downloads last month
- 185