forked from EricLBuehler/mistral.rs
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathxlora-paper-ordering.json
1 lines (1 loc) · 6.33 KB
/
xlora-paper-ordering.json
1
{"base_model_id": "HuggingFaceH4/zephyr-7b-beta", "order": ["adapter_1","adapter_2","adapter_3","adapter_4","adapter_5","adapter_6","adapter_7","adapter_8","adapter_9"], "layers": {"model.layers.0.self_attn.q_proj": 0, "model.layers.0.self_attn.k_proj": 1, "model.layers.0.self_attn.v_proj": 2, "model.layers.0.self_attn.o_proj": 3, "model.layers.0.mlp.gate_proj": 4, "model.layers.1.self_attn.q_proj": 5, "model.layers.1.self_attn.k_proj": 6, "model.layers.1.self_attn.v_proj": 7, "model.layers.1.self_attn.o_proj": 8, "model.layers.1.mlp.gate_proj": 9, "model.layers.2.self_attn.q_proj": 10, "model.layers.2.self_attn.k_proj": 11, "model.layers.2.self_attn.v_proj": 12, "model.layers.2.self_attn.o_proj": 13, "model.layers.2.mlp.gate_proj": 14, "model.layers.3.self_attn.q_proj": 15, "model.layers.3.self_attn.k_proj": 16, "model.layers.3.self_attn.v_proj": 17, "model.layers.3.self_attn.o_proj": 18, "model.layers.3.mlp.gate_proj": 19, "model.layers.4.self_attn.q_proj": 20, "model.layers.4.self_attn.k_proj": 21, "model.layers.4.self_attn.v_proj": 22, "model.layers.4.self_attn.o_proj": 23, "model.layers.4.mlp.gate_proj": 24, "model.layers.5.self_attn.q_proj": 25, "model.layers.5.self_attn.k_proj": 26, "model.layers.5.self_attn.v_proj": 27, "model.layers.5.self_attn.o_proj": 28, "model.layers.5.mlp.gate_proj": 29, "model.layers.6.self_attn.q_proj": 30, "model.layers.6.self_attn.k_proj": 31, "model.layers.6.self_attn.v_proj": 32, "model.layers.6.self_attn.o_proj": 33, "model.layers.6.mlp.gate_proj": 34, "model.layers.7.self_attn.q_proj": 35, "model.layers.7.self_attn.k_proj": 36, "model.layers.7.self_attn.v_proj": 37, "model.layers.7.self_attn.o_proj": 38, "model.layers.7.mlp.gate_proj": 39, "model.layers.8.self_attn.q_proj": 40, "model.layers.8.self_attn.k_proj": 41, "model.layers.8.self_attn.v_proj": 42, "model.layers.8.self_attn.o_proj": 43, "model.layers.8.mlp.gate_proj": 44, "model.layers.9.self_attn.q_proj": 45, "model.layers.9.self_attn.k_proj": 46, "model.layers.9.self_attn.v_proj": 47, "model.layers.9.self_attn.o_proj": 48, "model.layers.9.mlp.gate_proj": 49, "model.layers.10.self_attn.q_proj": 50, "model.layers.10.self_attn.k_proj": 51, "model.layers.10.self_attn.v_proj": 52, "model.layers.10.self_attn.o_proj": 53, "model.layers.10.mlp.gate_proj": 54, "model.layers.11.self_attn.q_proj": 55, "model.layers.11.self_attn.k_proj": 56, "model.layers.11.self_attn.v_proj": 57, "model.layers.11.self_attn.o_proj": 58, "model.layers.11.mlp.gate_proj": 59, "model.layers.12.self_attn.q_proj": 60, "model.layers.12.self_attn.k_proj": 61, "model.layers.12.self_attn.v_proj": 62, "model.layers.12.self_attn.o_proj": 63, "model.layers.12.mlp.gate_proj": 64, "model.layers.13.self_attn.q_proj": 65, "model.layers.13.self_attn.k_proj": 66, "model.layers.13.self_attn.v_proj": 67, "model.layers.13.self_attn.o_proj": 68, "model.layers.13.mlp.gate_proj": 69, "model.layers.14.self_attn.q_proj": 70, "model.layers.14.self_attn.k_proj": 71, "model.layers.14.self_attn.v_proj": 72, "model.layers.14.self_attn.o_proj": 73, "model.layers.14.mlp.gate_proj": 74, "model.layers.15.self_attn.q_proj": 75, "model.layers.15.self_attn.k_proj": 76, "model.layers.15.self_attn.v_proj": 77, "model.layers.15.self_attn.o_proj": 78, "model.layers.15.mlp.gate_proj": 79, "model.layers.16.self_attn.q_proj": 80, "model.layers.16.self_attn.k_proj": 81, "model.layers.16.self_attn.v_proj": 82, "model.layers.16.self_attn.o_proj": 83, "model.layers.16.mlp.gate_proj": 84, "model.layers.17.self_attn.q_proj": 85, "model.layers.17.self_attn.k_proj": 86, "model.layers.17.self_attn.v_proj": 87, "model.layers.17.self_attn.o_proj": 88, "model.layers.17.mlp.gate_proj": 89, "model.layers.18.self_attn.q_proj": 90, "model.layers.18.self_attn.k_proj": 91, "model.layers.18.self_attn.v_proj": 92, "model.layers.18.self_attn.o_proj": 93, "model.layers.18.mlp.gate_proj": 94, "model.layers.19.self_attn.q_proj": 95, "model.layers.19.self_attn.k_proj": 96, "model.layers.19.self_attn.v_proj": 97, "model.layers.19.self_attn.o_proj": 98, "model.layers.19.mlp.gate_proj": 99, "model.layers.20.self_attn.q_proj": 100, "model.layers.20.self_attn.k_proj": 101, "model.layers.20.self_attn.v_proj": 102, "model.layers.20.self_attn.o_proj": 103, "model.layers.20.mlp.gate_proj": 104, "model.layers.21.self_attn.q_proj": 105, "model.layers.21.self_attn.k_proj": 106, "model.layers.21.self_attn.v_proj": 107, "model.layers.21.self_attn.o_proj": 108, "model.layers.21.mlp.gate_proj": 109, "model.layers.22.self_attn.q_proj": 110, "model.layers.22.self_attn.k_proj": 111, "model.layers.22.self_attn.v_proj": 112, "model.layers.22.self_attn.o_proj": 113, "model.layers.22.mlp.gate_proj": 114, "model.layers.23.self_attn.q_proj": 115, "model.layers.23.self_attn.k_proj": 116, "model.layers.23.self_attn.v_proj": 117, "model.layers.23.self_attn.o_proj": 118, "model.layers.23.mlp.gate_proj": 119, "model.layers.24.self_attn.q_proj": 120, "model.layers.24.self_attn.k_proj": 121, "model.layers.24.self_attn.v_proj": 122, "model.layers.24.self_attn.o_proj": 123, "model.layers.24.mlp.gate_proj": 124, "model.layers.25.self_attn.q_proj": 125, "model.layers.25.self_attn.k_proj": 126, "model.layers.25.self_attn.v_proj": 127, "model.layers.25.self_attn.o_proj": 128, "model.layers.25.mlp.gate_proj": 129, "model.layers.26.self_attn.q_proj": 130, "model.layers.26.self_attn.k_proj": 131, "model.layers.26.self_attn.v_proj": 132, "model.layers.26.self_attn.o_proj": 133, "model.layers.26.mlp.gate_proj": 134, "model.layers.27.self_attn.q_proj": 135, "model.layers.27.self_attn.k_proj": 136, "model.layers.27.self_attn.v_proj": 137, "model.layers.27.self_attn.o_proj": 138, "model.layers.27.mlp.gate_proj": 139, "model.layers.28.self_attn.q_proj": 140, "model.layers.28.self_attn.k_proj": 141, "model.layers.28.self_attn.v_proj": 142, "model.layers.28.self_attn.o_proj": 143, "model.layers.28.mlp.gate_proj": 144, "model.layers.29.self_attn.q_proj": 145, "model.layers.29.self_attn.k_proj": 146, "model.layers.29.self_attn.v_proj": 147, "model.layers.29.self_attn.o_proj": 148, "model.layers.29.mlp.gate_proj": 149, "model.layers.30.self_attn.q_proj": 150, "model.layers.30.self_attn.k_proj": 151, "model.layers.30.self_attn.v_proj": 152, "model.layers.30.self_attn.o_proj": 153, "model.layers.30.mlp.gate_proj": 154, "model.layers.31.self_attn.q_proj": 155, "model.layers.31.self_attn.k_proj": 156, "model.layers.31.self_attn.v_proj": 157, "model.layers.31.self_attn.o_proj": 158, "model.layers.31.mlp.gate_proj": 159}}