We train a TFT on simulated data with context length 600 and target length 200, then finetune on the sequences played 2024-10-21, which will be same sequences as Parallel MD 2024-10-22.
We train a TFT on simulated data with context length 600 and target length 200, then finetune on the sequences played 2024-10-21, which will be same sequences as Parallel MD 2024-10-22.