| GPT Dataset | |
| Data prefix /data/redpajama/medium | |
| Dataset name train | |
| Number of samples 512000000 | |
| Number of epochs 13 | |
| Sequence length 2048 | |
| Random seed 1234 | |
| Split 969, 30, 1 | |
| GPT Dataset | |
| Data prefix /data/redpajama/medium | |
| Dataset name train | |
| Number of samples 512000000 | |
| Number of epochs 13 | |
| Sequence length 2048 | |
| Random seed 1234 | |
| Split 969, 30, 1 | |