Ctrl+K
- bimamba_stage_a
- bimamba_stage_a_128
- bimamba_stage_a_40k
- bimamba_stage_b
- bimamba_stage_b_joint_1024_40k
- bimamba_stage_b_joint_1024_40k_fixed
- bimamba_stage_b_joint_ckpt_40k
- bimamba_stage_b_joint_ckpt_40k_fixed
- bimamba_stage_b_joint_ckpt_40k_fixed_v2
- bimamba_stage_b_light_xattn_from_stage_a_256_retrain
- checkpoints
- dememwm_full_berzelius_8a100_bs16_global128_350k
- dememwm_full_berzelius_8a100_bs16_global128_350k_gate_off
- dememwm_sep_full_seq_gate_none_snr_shift_lr8e5_8a100_bs4_global32_750k
- dememwm_sep_full_seq_gate_none_snr_shift_ntok8_8a100_bs4_global32_750k
- dememwm_sep_full_seq_gate_none_snr_shift_rev1_8a100_bs4_global32_750k
- dememwm_separate_memory_berzelius_8a100_bs16_global128_350k
- dememwm_separate_memory_full_hidden1024_all_layers_berzelius_8a100_bs16_global128_350k
- hierarchy_bimamba_stage_a_40k
- hierarchy_bimamba_stage_b_joint
- hierarchy_bimamba_stage_b_joint_fixed
- videos
- wandb
- wandb_128
- wandb_256
- 245 kB
- 72.2 kB
- 12.1 GB xet
- 7.4 MB
- 11.1 MB xet
- 15 MB xet
- 9.09 MB
- 8.99 MB
- 167 kB
- 89.6 kB
- 48.6 MB xet
- 10.5 MB xet
- 6.74 MB
- 221 kB
- 487 kB
- 16.8 GB xet
- 16.8 GB xet
- 14.3 GB xet
- 4.96 kB