Dataset Viewer
Auto-converted to Parquet Duplicate
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