| { | |
| "measurement": { | |
| "model.layers.0": { | |
| "accuracy": 0.970212459564209, | |
| "total_bits": 3450224640.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "2": 64 | |
| }, | |
| "bits": [ | |
| 2 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.1": { | |
| "accuracy": 1e-06, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.2": { | |
| "accuracy": 1e-06, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.3": { | |
| "accuracy": 0.9956843852996826, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.4": { | |
| "accuracy": 0.9981266558170319, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.5": { | |
| "accuracy": 0.998059019446373, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.6": { | |
| "accuracy": 0.9980554133653641, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.7": { | |
| "accuracy": 0.9977759420871735, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.8": { | |
| "accuracy": 0.9974469840526581, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.9": { | |
| "accuracy": 0.9966913759708405, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.10": { | |
| "accuracy": 0.9971426427364349, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.11": { | |
| "accuracy": 0.9968973100185394, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.12": { | |
| "accuracy": 0.9966961741447449, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.13": { | |
| "accuracy": 0.9964953660964966, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.14": { | |
| "accuracy": 0.9956452250480652, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.15": { | |
| "accuracy": 0.9963669776916504, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.16": { | |
| "accuracy": 0.9964925646781921, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.17": { | |
| "accuracy": 0.9957883954048157, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.18": { | |
| "accuracy": 0.9953466057777405, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.19": { | |
| "accuracy": 0.9949451684951782, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.20": { | |
| "accuracy": 0.9948659539222717, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.21": { | |
| "accuracy": 0.9933203458786011, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.22": { | |
| "accuracy": 0.993752121925354, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.23": { | |
| "accuracy": 0.9931535720825195, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.24": { | |
| "accuracy": 0.993009090423584, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.25": { | |
| "accuracy": 0.993314266204834, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.26": { | |
| "accuracy": 0.9932155609130859, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.27": { | |
| "accuracy": 0.9929879903793335, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.28": { | |
| "accuracy": 0.9927887916564941, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.29": { | |
| "accuracy": 0.9921056032180786, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.30": { | |
| "accuracy": 0.9920321702957153, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.31": { | |
| "accuracy": 0.9924229383468628, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.32": { | |
| "accuracy": 0.9922597408294678, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.33": { | |
| "accuracy": 0.9921375513076782, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.34": { | |
| "accuracy": 0.992377519607544, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.35": { | |
| "accuracy": 0.9923077821731567, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.36": { | |
| "accuracy": 0.9923259019851685, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.37": { | |
| "accuracy": 0.9915577173233032, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.38": { | |
| "accuracy": 0.9913859367370605, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.39": { | |
| "accuracy": 0.9912623167037964, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.40": { | |
| "accuracy": 0.990553617477417, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.41": { | |
| "accuracy": 0.9903602600097656, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.42": { | |
| "accuracy": 0.9900898933410645, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.43": { | |
| "accuracy": 0.989501953125, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.44": { | |
| "accuracy": 0.9882993698120117, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.45": { | |
| "accuracy": 0.9870076179504395, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.46": { | |
| "accuracy": 0.9869430065155029, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.47": { | |
| "accuracy": 0.9868755340576172, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.48": { | |
| "accuracy": 0.987098217010498, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.49": { | |
| "accuracy": 0.9861912727355957, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.50": { | |
| "accuracy": 0.9854857921600342, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.51": { | |
| "accuracy": 0.9853277206420898, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.52": { | |
| "accuracy": 0.9855747222900391, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.53": { | |
| "accuracy": 0.9850366115570068, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.54": { | |
| "accuracy": 0.9845268726348877, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.55": { | |
| "accuracy": 0.9843268394470215, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.56": { | |
| "accuracy": 0.9845049381256104, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.57": { | |
| "accuracy": 0.9837610721588135, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.58": { | |
| "accuracy": 0.9830727577209473, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.59": { | |
| "accuracy": 0.9829492568969727, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.60": { | |
| "accuracy": 0.9837930202484131, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.61": { | |
| "accuracy": 0.9837956428527832, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.62": { | |
| "accuracy": 0.9837894439697266, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.63": { | |
| "accuracy": 0.9841635227203369, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.64": { | |
| "accuracy": 0.9841783046722412, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.65": { | |
| "accuracy": 0.9846138954162598, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.66": { | |
| "accuracy": 0.9849095344543457, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.67": { | |
| "accuracy": 0.9853219985961914, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.68": { | |
| "accuracy": 0.9849669933319092, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.69": { | |
| "accuracy": 0.9851610660552979, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.70": { | |
| "accuracy": 0.9845409393310547, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.71": { | |
| "accuracy": 0.9842643737792969, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.72": { | |
| "accuracy": 0.9841709136962891, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.73": { | |
| "accuracy": 0.9840455055236816, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.74": { | |
| "accuracy": 0.9835302829742432, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.75": { | |
| "accuracy": 0.9822866916656494, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.76": { | |
| "accuracy": 0.980748176574707, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.77": { | |
| "accuracy": 0.9800610542297363, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.78": { | |
| "accuracy": 0.9786887168884277, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| }, | |
| "model.layers.79": { | |
| "accuracy": 1e-06, | |
| "total_bits": 3579199488.0, | |
| "o_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "down_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "q_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "k_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "v_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "gate_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| }, | |
| "up_proj": { | |
| "group_size": { | |
| "4": 128 | |
| }, | |
| "bits": [ | |
| 4 | |
| ], | |
| "bits_prop": [ | |
| 1 | |
| ], | |
| "scale_bits": 4, | |
| "scale_groups:": 32 | |
| } | |
| } | |
| } | |
| } |