format string | split string | ntokens int64 | ndocs int64 | eot_id int64 | vocab_size int64 | tokenizer string | doc_shuffled bool | shuffle_seed int64 | shuffle_buffer int64 | split_kind string | max_doc_tokens int64 | split_docs int64 | source_tokens dict | source_docs dict | source_weights dict |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
u16le | train | 5,994,252,589 | 8,459,941 | 0 | 50,176 | lwt-shared-bpe-50176.json | true | 42 | 300,000 | random | 16,384 | 7,294 | {
"fineweb": 5484694629,
"cosmopedia-web_samples_v2": 515305707
} | {
"fineweb": 7766835,
"cosmopedia-web_samples_v2": 701416
} | {
"fineweb": 0.9,
"cosmopedia-web_samples_v2": 0.1
} |
No dataset card yet
- Downloads last month
- 13