diff --git a/.gitattributes b/.gitattributes index 0db223cacbeaff38525c359f6282428fce80ed2c..fdcdacb5f900c01805fb7e6cbc5f2843214331ee 100644 --- a/.gitattributes +++ b/.gitattributes @@ -2932,3 +2932,128 @@ deepseek-r1-1.5b-unary4/model_layers_23_mlp_down_proj_weight.planes filter=lfs d deepseek-r1-1.5b-unary4/model_layers_25_mlp_up_proj_weight.sign filter=lfs diff=lfs merge=lfs -text deepseek-r1-1.5b-unary4/model_layers_2_self_attn_q_proj_weight.sign filter=lfs diff=lfs merge=lfs -text deepseek-r1-1.5b-unary4/model_layers_5_mlp_down_proj_weight.planes filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-unary4/model_layers_0_mlp_gate_proj_weight.planes filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-unary4/model_layers_19_self_attn_q_proj_weight.planes filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-unary4/model_layers_27_self_attn_v_proj_weight.planes filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-unary4/model_layers_3_mlp_up_proj_weight.planes filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-unary4/model_layers_16_self_attn_q_proj_weight.sign filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-unary4/model_layers_27_mlp_gate_proj_weight.sign filter=lfs diff=lfs merge=lfs -text +qwen3-4b-log5-unary/model_layers_31_self_attn_k_proj_weight.planes filter=lfs diff=lfs merge=lfs -text +qwen3-4b-log5-unary/model_layers_11_self_attn_o_proj_weight.planes filter=lfs diff=lfs merge=lfs -text +qwen3-4b-log5-unary/model_layers_34_mlp_up_proj_weight.sign filter=lfs diff=lfs merge=lfs -text +qwen3-4b-log5-unary/model_layers_18_mlp_gate_proj_weight.sign filter=lfs diff=lfs merge=lfs -text +qwen3-4b-log5-unary/model_layers_5_self_attn_o_proj_weight.sign filter=lfs diff=lfs merge=lfs -text +qwen3-4b-log5-unary/model_layers_25_self_attn_k_proj_weight.planes filter=lfs diff=lfs merge=lfs -text +qwen3-4b-log5-unary/model_layers_17_mlp_down_proj_weight.sign filter=lfs diff=lfs merge=lfs -text +qwen3-4b-log5-unary/model_layers_5_self_attn_k_proj_weight.planes filter=lfs diff=lfs merge=lfs -text +qwen3-4b-log5-unary/model_layers_6_mlp_up_proj_weight.planes filter=lfs diff=lfs merge=lfs -text +qwen3-4b-log5-unary/model_layers_33_self_attn_o_proj_weight.planes filter=lfs diff=lfs merge=lfs -text +qwen3-4b-log5-unary/model_layers_0_mlp_up_proj_weight.planes filter=lfs diff=lfs merge=lfs -text +qwen3-4b-log5-unary/model_layers_18_mlp_up_proj_weight.sign filter=lfs diff=lfs merge=lfs -text +qwen3-4b-log5-unary/model_layers_17_self_attn_o_proj_weight.sign filter=lfs diff=lfs merge=lfs -text +qwen3-4b-log5-unary/model_layers_21_self_attn_k_proj_weight.planes filter=lfs diff=lfs merge=lfs -text +qwen3-4b-log5-unary/model_layers_2_self_attn_o_proj_weight.sign filter=lfs diff=lfs merge=lfs -text +qwen3-4b-log5-unary/model_layers_6_mlp_up_proj_weight.sign filter=lfs diff=lfs merge=lfs -text +qwen3-4b-log5-unary/model_layers_25_self_attn_o_proj_weight.planes filter=lfs diff=lfs merge=lfs -text +qwen3-4b-log5-unary/model_layers_29_self_attn_v_proj_weight.planes filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_18_self_attn_o_proj_weight.neg filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_4_mlp_gate_proj_weight.pos filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_25_mlp_down_proj_weight.neg filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_3_self_attn_o_proj_weight.pos filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_24_self_attn_q_proj_weight.neg filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_23_mlp_gate_proj_weight.pos filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_24_mlp_up_proj_weight.neg filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_11_self_attn_q_proj_weight.pos filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_8_self_attn_q_proj_weight.pos filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_10_mlp_gate_proj_weight.pos filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_2_self_attn_q_proj_weight.pos filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_16_mlp_down_proj_weight.pos filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_22_self_attn_o_proj_weight.pos filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_17_self_attn_o_proj_weight.pos filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_12_mlp_gate_proj_weight.neg filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_17_mlp_up_proj_weight.neg filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_6_self_attn_o_proj_weight.pos filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_4_self_attn_o_proj_weight.pos filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_1_mlp_gate_proj_weight.neg filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_21_mlp_up_proj_weight.pos filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_4_mlp_down_proj_weight.neg filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_16_self_attn_o_proj_weight.pos filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_20_mlp_gate_proj_weight.neg filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_0_mlp_down_proj_weight.neg filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_1_self_attn_q_proj_weight.pos filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_6_mlp_down_proj_weight.pos filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_12_self_attn_o_proj_weight.pos filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_15_self_attn_q_proj_weight.neg filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_13_mlp_down_proj_weight.neg filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_1_self_attn_o_proj_weight.neg filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_20_self_attn_q_proj_weight.neg filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_1_mlp_down_proj_weight.pos filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_20_mlp_down_proj_weight.neg filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_12_self_attn_q_proj_weight.neg filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_7_mlp_gate_proj_weight.neg filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_21_mlp_down_proj_weight.pos filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_25_mlp_down_proj_weight.pos filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_0_self_attn_q_proj_weight.neg filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_24_self_attn_o_proj_weight.pos filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_25_self_attn_o_proj_weight.neg filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_20_mlp_up_proj_weight.neg filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_7_mlp_up_proj_weight.neg filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_10_self_attn_q_proj_weight.pos filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_26_self_attn_o_proj_weight.pos filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_7_self_attn_q_proj_weight.neg filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_3_self_attn_q_proj_weight.pos filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_12_mlp_gate_proj_weight.pos filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_16_mlp_down_proj_weight.neg filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_10_self_attn_q_proj_weight.neg filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_23_self_attn_o_proj_weight.neg filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_18_self_attn_q_proj_weight.pos filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_27_self_attn_q_proj_weight.pos filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_14_self_attn_q_proj_weight.pos filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_9_mlp_gate_proj_weight.pos filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_21_self_attn_o_proj_weight.neg filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_23_mlp_down_proj_weight.pos filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_27_mlp_down_proj_weight.neg filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_25_mlp_up_proj_weight.pos filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_19_mlp_gate_proj_weight.neg filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_26_mlp_gate_proj_weight.pos filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_19_mlp_down_proj_weight.pos filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_3_self_attn_o_proj_weight.neg filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_13_mlp_gate_proj_weight.pos filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_2_mlp_gate_proj_weight.pos filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_11_self_attn_o_proj_weight.pos filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_24_self_attn_o_proj_weight.neg filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_1_mlp_up_proj_weight.pos filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_26_self_attn_q_proj_weight.neg filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_22_self_attn_q_proj_weight.neg filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_26_mlp_up_proj_weight.pos filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_5_self_attn_q_proj_weight.neg filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_10_mlp_down_proj_weight.pos filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_14_self_attn_q_proj_weight.neg filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_26_self_attn_q_proj_weight.pos filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_25_mlp_gate_proj_weight.pos filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_10_mlp_up_proj_weight.neg filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_18_self_attn_q_proj_weight.neg filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_18_mlp_up_proj_weight.neg filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_20_mlp_up_proj_weight.pos filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_7_mlp_up_proj_weight.pos filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_2_mlp_down_proj_weight.neg filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_17_self_attn_o_proj_weight.neg filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_27_mlp_up_proj_weight.pos filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_9_mlp_up_proj_weight.neg filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_22_mlp_gate_proj_weight.pos filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_11_mlp_gate_proj_weight.neg filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_9_mlp_down_proj_weight.neg filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_3_mlp_gate_proj_weight.pos filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_21_self_attn_q_proj_weight.pos filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_14_mlp_up_proj_weight.neg filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_13_mlp_down_proj_weight.pos filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_23_self_attn_o_proj_weight.pos filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_5_mlp_gate_proj_weight.neg filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_16_mlp_gate_proj_weight.neg filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_22_self_attn_q_proj_weight.pos filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_15_mlp_up_proj_weight.pos filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_23_mlp_gate_proj_weight.neg filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_18_mlp_gate_proj_weight.neg filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_14_mlp_gate_proj_weight.neg filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_19_self_attn_o_proj_weight.neg filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-ternary/model_layers_4_self_attn_q_proj_weight.neg filter=lfs diff=lfs merge=lfs -text diff --git a/deepseek-r1-1.5b-ternary/model_layers_0_mlp_down_proj_weight.neg b/deepseek-r1-1.5b-ternary/model_layers_0_mlp_down_proj_weight.neg new file mode 100644 index 0000000000000000000000000000000000000000..bd601966f8601b0171ef31c4e7a508d043ce57f1 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_0_mlp_down_proj_weight.neg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9370199575082915bf0ccdf9a74ca02e0476d4814058a3a348fd1538d94d1fce +size 1720320 diff --git a/deepseek-r1-1.5b-ternary/model_layers_0_self_attn_q_proj_weight.neg b/deepseek-r1-1.5b-ternary/model_layers_0_self_attn_q_proj_weight.neg new file mode 100644 index 0000000000000000000000000000000000000000..c0e543847adf2ab10f1d5ea9ee5787a23ffe5dbf --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_0_self_attn_q_proj_weight.neg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a0039040b13a81f1277457ec040789066e6f1f91955330878afb46c37cdaba49 +size 294912 diff --git a/deepseek-r1-1.5b-ternary/model_layers_10_mlp_down_proj_weight.pos b/deepseek-r1-1.5b-ternary/model_layers_10_mlp_down_proj_weight.pos new file mode 100644 index 0000000000000000000000000000000000000000..429e6b5652d74060bca1e278dc783bbd1ef46f87 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_10_mlp_down_proj_weight.pos @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9f7ceb4ae115ebeb4ee2f0995e4b819dbdff72719cbc7fb55c674cc1b307d8a6 +size 1720320 diff --git a/deepseek-r1-1.5b-ternary/model_layers_10_mlp_gate_proj_weight.pos b/deepseek-r1-1.5b-ternary/model_layers_10_mlp_gate_proj_weight.pos new file mode 100644 index 0000000000000000000000000000000000000000..211584533421c311f79af93c750036cc030f7ede --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_10_mlp_gate_proj_weight.pos @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3b96cf85032c14e1c817b06edaf27c1b3fa340d3dd197ac066fee8cb72065be3 +size 1720320 diff --git a/deepseek-r1-1.5b-ternary/model_layers_10_mlp_up_proj_weight.neg b/deepseek-r1-1.5b-ternary/model_layers_10_mlp_up_proj_weight.neg new file mode 100644 index 0000000000000000000000000000000000000000..ef1de2dc8443edda7f3db778aa1a3b40dd9853c6 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_10_mlp_up_proj_weight.neg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fddf0f6fb2efc721468ca19e2ed5e8468c9a3a3c7023d3c1fc421d41e1a1f6e7 +size 1720320 diff --git a/deepseek-r1-1.5b-ternary/model_layers_10_self_attn_q_proj_weight.neg b/deepseek-r1-1.5b-ternary/model_layers_10_self_attn_q_proj_weight.neg new file mode 100644 index 0000000000000000000000000000000000000000..8f9b6a1d71abde8d0f1c92d4c3746f78f8b265d0 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_10_self_attn_q_proj_weight.neg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fcc67b018afb15aa2e4b9278a650bfea3d6d190783bdbf7a0ceac6cb6c64e6d7 +size 294912 diff --git a/deepseek-r1-1.5b-ternary/model_layers_10_self_attn_q_proj_weight.pos b/deepseek-r1-1.5b-ternary/model_layers_10_self_attn_q_proj_weight.pos new file mode 100644 index 0000000000000000000000000000000000000000..830a38e0fd670f45137707c7c1ba875ba30718e0 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_10_self_attn_q_proj_weight.pos @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0ef3ece652607bca621f5993dae017430835124aef9760ff918b4abec6b3b37e +size 294912 diff --git a/deepseek-r1-1.5b-ternary/model_layers_11_mlp_gate_proj_weight.neg b/deepseek-r1-1.5b-ternary/model_layers_11_mlp_gate_proj_weight.neg new file mode 100644 index 0000000000000000000000000000000000000000..67c57d25210a0bfd5125728d103cf543e9ceeb81 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_11_mlp_gate_proj_weight.neg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7a714ab1bed0c02992d9f4532b6e792daaac2fb83e3d191447dbfa817ced9402 +size 1720320 diff --git a/deepseek-r1-1.5b-ternary/model_layers_11_self_attn_o_proj_weight.pos b/deepseek-r1-1.5b-ternary/model_layers_11_self_attn_o_proj_weight.pos new file mode 100644 index 0000000000000000000000000000000000000000..e6e304e4e0f796f730427791f2bba19b8a34702b --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_11_self_attn_o_proj_weight.pos @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:73ade90f3747624606c352e12b27d58dbb3f885a6bdecf5f19cd939ee5fea6e0 +size 294912 diff --git a/deepseek-r1-1.5b-ternary/model_layers_11_self_attn_q_proj_weight.pos b/deepseek-r1-1.5b-ternary/model_layers_11_self_attn_q_proj_weight.pos new file mode 100644 index 0000000000000000000000000000000000000000..b1bf2ac4be41b65eca69420b7fc5e27ee529817d --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_11_self_attn_q_proj_weight.pos @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:438f13bcf050c06d09542ed5c040500d1fc852b6a55e11a3d2faa09dfbac42da +size 294912 diff --git a/deepseek-r1-1.5b-ternary/model_layers_12_mlp_gate_proj_weight.neg b/deepseek-r1-1.5b-ternary/model_layers_12_mlp_gate_proj_weight.neg new file mode 100644 index 0000000000000000000000000000000000000000..973ceb3c2f128cccc1a2a7b906acba20ee95d25f --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_12_mlp_gate_proj_weight.neg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b0fc71961836421a06386d8a027d3dd9753eadbcaba69461a336979d0520ce12 +size 1720320 diff --git a/deepseek-r1-1.5b-ternary/model_layers_12_mlp_gate_proj_weight.pos b/deepseek-r1-1.5b-ternary/model_layers_12_mlp_gate_proj_weight.pos new file mode 100644 index 0000000000000000000000000000000000000000..c8c552f655ee32a9efe6dedb01042644645f05cc --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_12_mlp_gate_proj_weight.pos @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:884090015f65c803c2fcea824fd5bda6113b4d332adee97c86147a2f52af2d22 +size 1720320 diff --git a/deepseek-r1-1.5b-ternary/model_layers_12_self_attn_o_proj_weight.pos b/deepseek-r1-1.5b-ternary/model_layers_12_self_attn_o_proj_weight.pos new file mode 100644 index 0000000000000000000000000000000000000000..2a0b3ca2bf8df11e0fbe479fdf6d2ce4ff9dd300 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_12_self_attn_o_proj_weight.pos @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ccdb5b8f3e29e5331f01c1d1f7e683f77bc05d8b796b44d3111949e9bf7ddddb +size 294912 diff --git a/deepseek-r1-1.5b-ternary/model_layers_12_self_attn_q_proj_weight.neg b/deepseek-r1-1.5b-ternary/model_layers_12_self_attn_q_proj_weight.neg new file mode 100644 index 0000000000000000000000000000000000000000..d30e2c884b1e20827fb070bea99fb4dc3ed89a3e --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_12_self_attn_q_proj_weight.neg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:abf3bd47b000b5e5dd6a50452f917fa6ba04d96b039f4298553daa257a36dfc4 +size 294912 diff --git a/deepseek-r1-1.5b-ternary/model_layers_13_mlp_down_proj_weight.neg b/deepseek-r1-1.5b-ternary/model_layers_13_mlp_down_proj_weight.neg new file mode 100644 index 0000000000000000000000000000000000000000..69dddd73046f224f83c271d2bf29cd9158722837 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_13_mlp_down_proj_weight.neg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0d71389161a089240222f2ad45175321c212015fbfa85386fa1fb9ec07af52af +size 1720320 diff --git a/deepseek-r1-1.5b-ternary/model_layers_13_mlp_down_proj_weight.pos b/deepseek-r1-1.5b-ternary/model_layers_13_mlp_down_proj_weight.pos new file mode 100644 index 0000000000000000000000000000000000000000..3ec84fcf86096ce275dec4554b8b63bbd4116906 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_13_mlp_down_proj_weight.pos @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e97c598734b7c981ed94f78ec54d2ca56efc639c01b39a1a2fa29108a27c8306 +size 1720320 diff --git a/deepseek-r1-1.5b-ternary/model_layers_13_mlp_gate_proj_weight.pos b/deepseek-r1-1.5b-ternary/model_layers_13_mlp_gate_proj_weight.pos new file mode 100644 index 0000000000000000000000000000000000000000..fb271313afce64f3a656cc57d620a365f3d1339a --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_13_mlp_gate_proj_weight.pos @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fc64cb3cd12262fc430969d96cdb60154c2190439b5e6a55865b5f88231b31b0 +size 1720320 diff --git a/deepseek-r1-1.5b-ternary/model_layers_14_mlp_gate_proj_weight.neg b/deepseek-r1-1.5b-ternary/model_layers_14_mlp_gate_proj_weight.neg new file mode 100644 index 0000000000000000000000000000000000000000..cb4247938a3f64aebb446a06383cb8a5ac764e69 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_14_mlp_gate_proj_weight.neg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f8556b05b781bd3bbb8d1062bf89817a94a7a720a8590348b1b242e681e1ae40 +size 1720320 diff --git a/deepseek-r1-1.5b-ternary/model_layers_14_mlp_up_proj_weight.neg b/deepseek-r1-1.5b-ternary/model_layers_14_mlp_up_proj_weight.neg new file mode 100644 index 0000000000000000000000000000000000000000..df9e4dfdc96d7f79b7c57edcd7bf73406e9a27f5 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_14_mlp_up_proj_weight.neg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bd51bfdf413da284a4195279e69400dd0f8b3f6a04e357c6a40164f166c4c2ed +size 1720320 diff --git a/deepseek-r1-1.5b-ternary/model_layers_14_self_attn_q_proj_weight.neg b/deepseek-r1-1.5b-ternary/model_layers_14_self_attn_q_proj_weight.neg new file mode 100644 index 0000000000000000000000000000000000000000..60c4432ef71af3c44baa4f66758280c8125a6967 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_14_self_attn_q_proj_weight.neg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bc195c105edbe42cab3f16a889de305f86ad541ac74586b23f6b5308d2d6fd20 +size 294912 diff --git a/deepseek-r1-1.5b-ternary/model_layers_14_self_attn_q_proj_weight.pos b/deepseek-r1-1.5b-ternary/model_layers_14_self_attn_q_proj_weight.pos new file mode 100644 index 0000000000000000000000000000000000000000..aa6acc906168a2dfef127b2454c395a23a26fa5b --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_14_self_attn_q_proj_weight.pos @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a25c8cf0274dc57234925f38b6b93896aab0ff540cae11ec59bea32f6e3fa705 +size 294912 diff --git a/deepseek-r1-1.5b-ternary/model_layers_15_mlp_up_proj_weight.pos b/deepseek-r1-1.5b-ternary/model_layers_15_mlp_up_proj_weight.pos new file mode 100644 index 0000000000000000000000000000000000000000..fa4f859b082c26baaf48865abe7add6743c9b84c --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_15_mlp_up_proj_weight.pos @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fa23ac7e657418c24d3cd4cdc6e9df48f967d63f477a50dedf4b9f6a3a95f069 +size 1720320 diff --git a/deepseek-r1-1.5b-ternary/model_layers_15_self_attn_q_proj_weight.neg b/deepseek-r1-1.5b-ternary/model_layers_15_self_attn_q_proj_weight.neg new file mode 100644 index 0000000000000000000000000000000000000000..953734c3e05574fa89b6e7f1c710543157c469cc --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_15_self_attn_q_proj_weight.neg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:87c78eea22c56a264f8cdc1247d870967edd7dc45d84b157e49e658a57ac17ab +size 294912 diff --git a/deepseek-r1-1.5b-ternary/model_layers_16_mlp_down_proj_weight.neg b/deepseek-r1-1.5b-ternary/model_layers_16_mlp_down_proj_weight.neg new file mode 100644 index 0000000000000000000000000000000000000000..a0e4c0f041a031b1d82e795d02cd380609423aec --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_16_mlp_down_proj_weight.neg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c1db612017d762a1eb279157ac1e1e91ea6e515065be47a198c10c50fbfa9c6b +size 1720320 diff --git a/deepseek-r1-1.5b-ternary/model_layers_16_mlp_down_proj_weight.pos b/deepseek-r1-1.5b-ternary/model_layers_16_mlp_down_proj_weight.pos new file mode 100644 index 0000000000000000000000000000000000000000..9f8b9bb629e2ba92a148f0cb7a0dc9f5741960f5 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_16_mlp_down_proj_weight.pos @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:60b53dfc6b6a77e0533ced6e12377e1678ad8fd45446e75a43d42672da84b46d +size 1720320 diff --git a/deepseek-r1-1.5b-ternary/model_layers_16_mlp_gate_proj_weight.neg b/deepseek-r1-1.5b-ternary/model_layers_16_mlp_gate_proj_weight.neg new file mode 100644 index 0000000000000000000000000000000000000000..fca54973b7b09f3db92aecbe02fb2ec441300942 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_16_mlp_gate_proj_weight.neg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:889bc934475e9a8f5e54007331376ac761170c78dceded83e6656cb25c7a7eca +size 1720320 diff --git a/deepseek-r1-1.5b-ternary/model_layers_16_self_attn_o_proj_weight.pos b/deepseek-r1-1.5b-ternary/model_layers_16_self_attn_o_proj_weight.pos new file mode 100644 index 0000000000000000000000000000000000000000..fd4fd9c80e75389ba8851571f9355973525943a7 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_16_self_attn_o_proj_weight.pos @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5027e40f35ad1ddc30c1703da3d13b01559c29dafb81285fae1f2e936d06c8a8 +size 294912 diff --git a/deepseek-r1-1.5b-ternary/model_layers_17_mlp_up_proj_weight.neg b/deepseek-r1-1.5b-ternary/model_layers_17_mlp_up_proj_weight.neg new file mode 100644 index 0000000000000000000000000000000000000000..b3d1e04bece2f060ccb029ba51c7665f92a20254 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_17_mlp_up_proj_weight.neg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1b30d05ffdca71e195d3f0352f5a5acea250115d6a56ae83c9d36feac395f9f1 +size 1720320 diff --git a/deepseek-r1-1.5b-ternary/model_layers_17_self_attn_o_proj_weight.neg b/deepseek-r1-1.5b-ternary/model_layers_17_self_attn_o_proj_weight.neg new file mode 100644 index 0000000000000000000000000000000000000000..2df5d9e98a57a9fac060005325fda190ba769600 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_17_self_attn_o_proj_weight.neg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d14e6048ec05fec1c7c77a7f2f847adbda898a30bafcf7e379fc454829a9b5dc +size 294912 diff --git a/deepseek-r1-1.5b-ternary/model_layers_17_self_attn_o_proj_weight.pos b/deepseek-r1-1.5b-ternary/model_layers_17_self_attn_o_proj_weight.pos new file mode 100644 index 0000000000000000000000000000000000000000..791afaf259e24ad7668f867713cb8a36d98e58c5 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_17_self_attn_o_proj_weight.pos @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:269a859b3b8c8d8621a5f5f743009e09420e17351b715e988ccbc6178cd95b1a +size 294912 diff --git a/deepseek-r1-1.5b-ternary/model_layers_18_mlp_gate_proj_weight.neg b/deepseek-r1-1.5b-ternary/model_layers_18_mlp_gate_proj_weight.neg new file mode 100644 index 0000000000000000000000000000000000000000..4ec2610693cbec38b0f9a1aac7b4c92d3bdea658 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_18_mlp_gate_proj_weight.neg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f294394f556e76a1d9121b9a1d2d1c3f0cbc7d08d49da1d397813cddffcab7a7 +size 1720320 diff --git a/deepseek-r1-1.5b-ternary/model_layers_18_mlp_up_proj_weight.neg b/deepseek-r1-1.5b-ternary/model_layers_18_mlp_up_proj_weight.neg new file mode 100644 index 0000000000000000000000000000000000000000..5a8442fdb379b392c7137e8478ecb389cf234294 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_18_mlp_up_proj_weight.neg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3c190e08a05529e90ad8d51bc381b619d4f0b33d408555d99646ce971361fef4 +size 1720320 diff --git a/deepseek-r1-1.5b-ternary/model_layers_18_self_attn_o_proj_weight.neg b/deepseek-r1-1.5b-ternary/model_layers_18_self_attn_o_proj_weight.neg new file mode 100644 index 0000000000000000000000000000000000000000..01414cf8d239b03cdca3185d60a4efcecc9c3a1a --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_18_self_attn_o_proj_weight.neg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9925152ce669f27a1d01781dcdf0185f126ab50e676f920bce91db5c71ada1c8 +size 294912 diff --git a/deepseek-r1-1.5b-ternary/model_layers_18_self_attn_q_proj_weight.neg b/deepseek-r1-1.5b-ternary/model_layers_18_self_attn_q_proj_weight.neg new file mode 100644 index 0000000000000000000000000000000000000000..c97df08181c32caae83d4b1f50c9695becb1acaa --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_18_self_attn_q_proj_weight.neg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:70f0e3a9a966268cbaefa9e7a824fb8f132ae583bcab48aa55d73989e8f18313 +size 294912 diff --git a/deepseek-r1-1.5b-ternary/model_layers_18_self_attn_q_proj_weight.pos b/deepseek-r1-1.5b-ternary/model_layers_18_self_attn_q_proj_weight.pos new file mode 100644 index 0000000000000000000000000000000000000000..74bd4ca6a8c8b20ac21db7e3e37cb3fdc9f422d3 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_18_self_attn_q_proj_weight.pos @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c515ecf4c4b0052b822041b16fefee33300a9f7713cca678cc0ea1dd0e785b5a +size 294912 diff --git a/deepseek-r1-1.5b-ternary/model_layers_19_mlp_down_proj_weight.pos b/deepseek-r1-1.5b-ternary/model_layers_19_mlp_down_proj_weight.pos new file mode 100644 index 0000000000000000000000000000000000000000..c9bb05c7a796da18da90375738c68560eca07c89 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_19_mlp_down_proj_weight.pos @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:802f1022ce6f45731de642c1f8f5c05aaa38bdfe00960ce6dadf4ee35abad259 +size 1720320 diff --git a/deepseek-r1-1.5b-ternary/model_layers_19_mlp_gate_proj_weight.neg b/deepseek-r1-1.5b-ternary/model_layers_19_mlp_gate_proj_weight.neg new file mode 100644 index 0000000000000000000000000000000000000000..bbb284f7c5b693493169ffd60a45fcb4b1955a0e --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_19_mlp_gate_proj_weight.neg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b90b7276235e0c32b8fc6b5c777b9e81808d91d6fc5900f518adf35fa53a82de +size 1720320 diff --git a/deepseek-r1-1.5b-ternary/model_layers_19_self_attn_o_proj_weight.neg b/deepseek-r1-1.5b-ternary/model_layers_19_self_attn_o_proj_weight.neg new file mode 100644 index 0000000000000000000000000000000000000000..15458d0268e940079ca1788a0dee9286628e8c38 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_19_self_attn_o_proj_weight.neg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8a949790db960aa2279875eb706222fc4bed59d6a4f871aed451097a89fe730c +size 294912 diff --git a/deepseek-r1-1.5b-ternary/model_layers_1_mlp_down_proj_weight.pos b/deepseek-r1-1.5b-ternary/model_layers_1_mlp_down_proj_weight.pos new file mode 100644 index 0000000000000000000000000000000000000000..bb489de62ffe6fb42bf6200ff36109c5d4a396bc --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_1_mlp_down_proj_weight.pos @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8f295e24b5622ae9d1ae1484a0df465d8d351cafaf7c7d32607dd9e3efcb15ac +size 1720320 diff --git a/deepseek-r1-1.5b-ternary/model_layers_1_mlp_gate_proj_weight.neg b/deepseek-r1-1.5b-ternary/model_layers_1_mlp_gate_proj_weight.neg new file mode 100644 index 0000000000000000000000000000000000000000..815ed5a9caa8ce52373e3aa54e044cf9df664cae --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_1_mlp_gate_proj_weight.neg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bd92b643d402cb41f72b06266f1ca03dccb359add6fdd8568d2126ee2b9f9304 +size 1720320 diff --git a/deepseek-r1-1.5b-ternary/model_layers_1_mlp_up_proj_weight.pos b/deepseek-r1-1.5b-ternary/model_layers_1_mlp_up_proj_weight.pos new file mode 100644 index 0000000000000000000000000000000000000000..533b576c8970e74397f5f6fc7c6fca1316e1ca14 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_1_mlp_up_proj_weight.pos @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e9d8bac3e0a74e1d5fe34557b069ea974bccd0f9b61df5fe0d648fd06997ff47 +size 1720320 diff --git a/deepseek-r1-1.5b-ternary/model_layers_1_self_attn_o_proj_weight.neg b/deepseek-r1-1.5b-ternary/model_layers_1_self_attn_o_proj_weight.neg new file mode 100644 index 0000000000000000000000000000000000000000..f1a097df309196a97fe8e68a436d8a1604654757 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_1_self_attn_o_proj_weight.neg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c634d9f61351ec7f9ba51f79f44ea380908487468853c5c3095d0e33fe50eae9 +size 294912 diff --git a/deepseek-r1-1.5b-ternary/model_layers_1_self_attn_q_proj_weight.pos b/deepseek-r1-1.5b-ternary/model_layers_1_self_attn_q_proj_weight.pos new file mode 100644 index 0000000000000000000000000000000000000000..1504ca4c4bde2b93281890bbcf5034ba10ca7120 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_1_self_attn_q_proj_weight.pos @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b4b90f80b1652f7408917360dea8a9b81a343f679c08eaeb6a5b7dedc2efd570 +size 294912 diff --git a/deepseek-r1-1.5b-ternary/model_layers_20_mlp_down_proj_weight.neg b/deepseek-r1-1.5b-ternary/model_layers_20_mlp_down_proj_weight.neg new file mode 100644 index 0000000000000000000000000000000000000000..047536906eb783f65abb80ac90e2a40a4e6ceaa3 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_20_mlp_down_proj_weight.neg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:49f808ef1e1c4c1f07beedf251420ceb5c04f763456956bb2a13101ee5bc42cc +size 1720320 diff --git a/deepseek-r1-1.5b-ternary/model_layers_20_mlp_gate_proj_weight.neg b/deepseek-r1-1.5b-ternary/model_layers_20_mlp_gate_proj_weight.neg new file mode 100644 index 0000000000000000000000000000000000000000..f83b34f4337b927eac2bf19058a17373a933e0c4 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_20_mlp_gate_proj_weight.neg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d119cbcad7adfb88b9fe58ab7f45ddddc3095a059de077bf846f2bc77be52acb +size 1720320 diff --git a/deepseek-r1-1.5b-ternary/model_layers_20_mlp_up_proj_weight.neg b/deepseek-r1-1.5b-ternary/model_layers_20_mlp_up_proj_weight.neg new file mode 100644 index 0000000000000000000000000000000000000000..6a41534a302ff9aaf4e2406780df04a907c4e2ac --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_20_mlp_up_proj_weight.neg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:31adfc9bfbb63e1dc5e6ddc30fb331b9069e895f6bf756167458b297ef620b44 +size 1720320 diff --git a/deepseek-r1-1.5b-ternary/model_layers_20_mlp_up_proj_weight.pos b/deepseek-r1-1.5b-ternary/model_layers_20_mlp_up_proj_weight.pos new file mode 100644 index 0000000000000000000000000000000000000000..334ab6e2584d6e0c91e5aa01c158c0784384a63b --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_20_mlp_up_proj_weight.pos @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4c25592e7e5f59a75c9ba312a43001a4795e88960fe4787de4e094e4c07a5696 +size 1720320 diff --git a/deepseek-r1-1.5b-ternary/model_layers_20_self_attn_q_proj_weight.neg b/deepseek-r1-1.5b-ternary/model_layers_20_self_attn_q_proj_weight.neg new file mode 100644 index 0000000000000000000000000000000000000000..ce330013c7a1220357d62ed4222f5ce3d31f3a44 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_20_self_attn_q_proj_weight.neg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b58daf42ec4283e7cd760df14c7528155ed561af8988cde5c19baa76aba3ceac +size 294912 diff --git a/deepseek-r1-1.5b-ternary/model_layers_21_mlp_down_proj_weight.pos b/deepseek-r1-1.5b-ternary/model_layers_21_mlp_down_proj_weight.pos new file mode 100644 index 0000000000000000000000000000000000000000..77dcaa766e412ad72a9107ca3c0bc8b54889bd99 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_21_mlp_down_proj_weight.pos @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:999f377760edb1648e7c8bbe898432a96b1cd75b278336aba74176ede4c393ee +size 1720320 diff --git a/deepseek-r1-1.5b-ternary/model_layers_21_mlp_up_proj_weight.pos b/deepseek-r1-1.5b-ternary/model_layers_21_mlp_up_proj_weight.pos new file mode 100644 index 0000000000000000000000000000000000000000..97d4ce2bf5d85ada9232b584524f1ddc20c6e559 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_21_mlp_up_proj_weight.pos @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6f61f1545467af3975cd564c027be5dc0b1e07177cf17f4819cd2debb69ed083 +size 1720320 diff --git a/deepseek-r1-1.5b-ternary/model_layers_21_self_attn_o_proj_weight.neg b/deepseek-r1-1.5b-ternary/model_layers_21_self_attn_o_proj_weight.neg new file mode 100644 index 0000000000000000000000000000000000000000..e4ba88966641ff98d37ecb9f89a3b5251369ee25 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_21_self_attn_o_proj_weight.neg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7594b083636d93553bf3bbf9c47b32b5a1a2c73dce5fb881234ef2926c3cdcfb +size 294912 diff --git a/deepseek-r1-1.5b-ternary/model_layers_21_self_attn_q_proj_weight.pos b/deepseek-r1-1.5b-ternary/model_layers_21_self_attn_q_proj_weight.pos new file mode 100644 index 0000000000000000000000000000000000000000..8ee87f57a526998210c1e19beeba82f1f670a0c8 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_21_self_attn_q_proj_weight.pos @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e5b832912419e24e5e7e71bbfc38097d7a3fdd094bdd9a41f03ea270b1232af1 +size 294912 diff --git a/deepseek-r1-1.5b-ternary/model_layers_22_mlp_gate_proj_weight.pos b/deepseek-r1-1.5b-ternary/model_layers_22_mlp_gate_proj_weight.pos new file mode 100644 index 0000000000000000000000000000000000000000..e202de35c903a3fe984a6c8af8d30f4f6cbaebe5 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_22_mlp_gate_proj_weight.pos @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:dc28ed733c9d5880edcd3f4af269cf244f27730f3236943432cb1381797d4238 +size 1720320 diff --git a/deepseek-r1-1.5b-ternary/model_layers_22_self_attn_o_proj_weight.pos b/deepseek-r1-1.5b-ternary/model_layers_22_self_attn_o_proj_weight.pos new file mode 100644 index 0000000000000000000000000000000000000000..2245c16b89b7efa3266fd31dea0f39e6cc58e904 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_22_self_attn_o_proj_weight.pos @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:929e1ace6f25aea0f0d5a55a0dc9b6db2dda71a9dc484e8c394dedb3fdf24c28 +size 294912 diff --git a/deepseek-r1-1.5b-ternary/model_layers_22_self_attn_q_proj_weight.neg b/deepseek-r1-1.5b-ternary/model_layers_22_self_attn_q_proj_weight.neg new file mode 100644 index 0000000000000000000000000000000000000000..afbf141606bda55e336d0b9a753f1544859de2b9 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_22_self_attn_q_proj_weight.neg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:87614cc4dcc6f60c76950370cf3f82b0199ab6be00fe60167d1b5a1b76c02488 +size 294912 diff --git a/deepseek-r1-1.5b-ternary/model_layers_22_self_attn_q_proj_weight.pos b/deepseek-r1-1.5b-ternary/model_layers_22_self_attn_q_proj_weight.pos new file mode 100644 index 0000000000000000000000000000000000000000..33e8d1e877f00cf35936ae8e1ab231fcf87fb763 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_22_self_attn_q_proj_weight.pos @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:32f77b59c4342227651b11b67f3b96a7cfad62ca7202e9a386c256807cbc8b8f +size 294912 diff --git a/deepseek-r1-1.5b-ternary/model_layers_23_mlp_down_proj_weight.pos b/deepseek-r1-1.5b-ternary/model_layers_23_mlp_down_proj_weight.pos new file mode 100644 index 0000000000000000000000000000000000000000..028a3999d2d22cbe0a4e967cedbe5808d4f76be7 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_23_mlp_down_proj_weight.pos @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2fd560ad918495ac65c2c51fba5d80f94ecfd363b46f68f2d2b00544b5a69fab +size 1720320 diff --git a/deepseek-r1-1.5b-ternary/model_layers_23_mlp_gate_proj_weight.neg b/deepseek-r1-1.5b-ternary/model_layers_23_mlp_gate_proj_weight.neg new file mode 100644 index 0000000000000000000000000000000000000000..a166aca1f3cae6db8eb71f05ceacca4e4e1e6afc --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_23_mlp_gate_proj_weight.neg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c69e264233f9021199424959c8f7a893f43b24275cf877fc568384b84c6aab3d +size 1720320 diff --git a/deepseek-r1-1.5b-ternary/model_layers_23_mlp_gate_proj_weight.pos b/deepseek-r1-1.5b-ternary/model_layers_23_mlp_gate_proj_weight.pos new file mode 100644 index 0000000000000000000000000000000000000000..561a2dcfcfdbd128ab18e90fac978a6d516fa29b --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_23_mlp_gate_proj_weight.pos @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:52b5e411efb76ab953cc2314206e9fcacd6942d6f12364e1cf768642bafb4fb3 +size 1720320 diff --git a/deepseek-r1-1.5b-ternary/model_layers_23_self_attn_o_proj_weight.neg b/deepseek-r1-1.5b-ternary/model_layers_23_self_attn_o_proj_weight.neg new file mode 100644 index 0000000000000000000000000000000000000000..d5c0265ddd1c10277d41be075effb058f07aa885 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_23_self_attn_o_proj_weight.neg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:251d518167ebe2b94a6da65fd44f96fd22c11d066af65b6e4df8bebbcdea7d70 +size 294912 diff --git a/deepseek-r1-1.5b-ternary/model_layers_23_self_attn_o_proj_weight.pos b/deepseek-r1-1.5b-ternary/model_layers_23_self_attn_o_proj_weight.pos new file mode 100644 index 0000000000000000000000000000000000000000..6e97e1ceed778576e56be986916feb909de51c6d --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_23_self_attn_o_proj_weight.pos @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7312c076331a55ef32b6fdee5875a7fc2ca5d21a6fad2798316af58bb8719a13 +size 294912 diff --git a/deepseek-r1-1.5b-ternary/model_layers_24_mlp_up_proj_weight.neg b/deepseek-r1-1.5b-ternary/model_layers_24_mlp_up_proj_weight.neg new file mode 100644 index 0000000000000000000000000000000000000000..cad48040bc1f5dcb5d59466e324dd112dd7f0da6 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_24_mlp_up_proj_weight.neg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:87db5e3eafbe094a543a0c0b0bf573779f6ca50052b90f9454718f37bac263dd +size 1720320 diff --git a/deepseek-r1-1.5b-ternary/model_layers_24_self_attn_o_proj_weight.neg b/deepseek-r1-1.5b-ternary/model_layers_24_self_attn_o_proj_weight.neg new file mode 100644 index 0000000000000000000000000000000000000000..94e040f534354e670032267903c862d551084215 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_24_self_attn_o_proj_weight.neg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:74aa0eab6b5b6f74ff3bdc0dbfc3dba22850be7d230016c7022c607fd0e38622 +size 294912 diff --git a/deepseek-r1-1.5b-ternary/model_layers_24_self_attn_o_proj_weight.pos b/deepseek-r1-1.5b-ternary/model_layers_24_self_attn_o_proj_weight.pos new file mode 100644 index 0000000000000000000000000000000000000000..4c24340238bc1394ca5a450efb5ef18d91e9bb2c --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_24_self_attn_o_proj_weight.pos @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:02c8206cfadaf773b0daae094e56d50dd69a612f08c418ae7884dfed4b45bcf6 +size 294912 diff --git a/deepseek-r1-1.5b-ternary/model_layers_24_self_attn_q_proj_weight.neg b/deepseek-r1-1.5b-ternary/model_layers_24_self_attn_q_proj_weight.neg new file mode 100644 index 0000000000000000000000000000000000000000..ae90d9d0a6ea66f28d81e829823c112360b896fb --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_24_self_attn_q_proj_weight.neg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8d89ecf210de6b5ffbc2acf76493f486ccfcfd4e151850d7f21daeef27b2633c +size 294912 diff --git a/deepseek-r1-1.5b-ternary/model_layers_25_mlp_down_proj_weight.neg b/deepseek-r1-1.5b-ternary/model_layers_25_mlp_down_proj_weight.neg new file mode 100644 index 0000000000000000000000000000000000000000..aad9edce935dc24f17047f7152fbd02fc9d2e2a5 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_25_mlp_down_proj_weight.neg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c8d010bc31ace23398e8a52a926af8ee326be0845e66ddacead9fea96dd96a7b +size 1720320 diff --git a/deepseek-r1-1.5b-ternary/model_layers_25_mlp_down_proj_weight.pos b/deepseek-r1-1.5b-ternary/model_layers_25_mlp_down_proj_weight.pos new file mode 100644 index 0000000000000000000000000000000000000000..224a110c862c4e178ae93646d56f4f01ae7ced73 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_25_mlp_down_proj_weight.pos @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a7c8381e8eafcc178726789d5f1a902d0a1c446074772fcac0fe6c372cd1aa16 +size 1720320 diff --git a/deepseek-r1-1.5b-ternary/model_layers_25_mlp_gate_proj_weight.pos b/deepseek-r1-1.5b-ternary/model_layers_25_mlp_gate_proj_weight.pos new file mode 100644 index 0000000000000000000000000000000000000000..3b835297e440bd9f97fe1c2c0085f24cb94d6a8d --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_25_mlp_gate_proj_weight.pos @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a345e865eb0d01690a4e5ddb04f3b5b36300fd0092951835eb09b8576ee05e95 +size 1720320 diff --git a/deepseek-r1-1.5b-ternary/model_layers_25_mlp_up_proj_weight.pos b/deepseek-r1-1.5b-ternary/model_layers_25_mlp_up_proj_weight.pos new file mode 100644 index 0000000000000000000000000000000000000000..602d4329217a2ba32ac52153ef07b22711d4d1e8 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_25_mlp_up_proj_weight.pos @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:766b51904ac46f1ec3bc1b21bcd46fed6ffc133ff326f4c99b54f4becf6c7201 +size 1720320 diff --git a/deepseek-r1-1.5b-ternary/model_layers_25_self_attn_o_proj_weight.neg b/deepseek-r1-1.5b-ternary/model_layers_25_self_attn_o_proj_weight.neg new file mode 100644 index 0000000000000000000000000000000000000000..a452309b02a3b68d0bfe29b19f9ab5eb0450fd9e --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_25_self_attn_o_proj_weight.neg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:39c6d8c7038e81c6d011fba4ace1d950ac8f48321fb679e9b95a1919ff61c6c8 +size 294912 diff --git a/deepseek-r1-1.5b-ternary/model_layers_26_mlp_gate_proj_weight.pos b/deepseek-r1-1.5b-ternary/model_layers_26_mlp_gate_proj_weight.pos new file mode 100644 index 0000000000000000000000000000000000000000..e1cae9b1b2e256457831fc0ebfbc35af3bb47a33 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_26_mlp_gate_proj_weight.pos @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c65ca16e61a834d35de54062e29eb51685e394391b9bd92bdca4eb7a0b669689 +size 1720320 diff --git a/deepseek-r1-1.5b-ternary/model_layers_26_mlp_up_proj_weight.pos b/deepseek-r1-1.5b-ternary/model_layers_26_mlp_up_proj_weight.pos new file mode 100644 index 0000000000000000000000000000000000000000..cdd9e5d8286ef1a53cb7ec8818dc934aa8fdbb0b --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_26_mlp_up_proj_weight.pos @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5a7a51ed0335d3896f4b4dab3c9f03f571f4cbecd59f9e645358069a825d57a4 +size 1720320 diff --git a/deepseek-r1-1.5b-ternary/model_layers_26_self_attn_o_proj_weight.pos b/deepseek-r1-1.5b-ternary/model_layers_26_self_attn_o_proj_weight.pos new file mode 100644 index 0000000000000000000000000000000000000000..faaeed223bd10dc060d7c5d1c0c12af2074a82a5 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_26_self_attn_o_proj_weight.pos @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:05eada4ac14ec0f1372625ec0665e4706fa7c8438b613adfccef3774e146b20b +size 294912 diff --git a/deepseek-r1-1.5b-ternary/model_layers_26_self_attn_q_proj_weight.neg b/deepseek-r1-1.5b-ternary/model_layers_26_self_attn_q_proj_weight.neg new file mode 100644 index 0000000000000000000000000000000000000000..95dac377369e9a3bacc8dde674aa02e679c005d2 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_26_self_attn_q_proj_weight.neg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:de6b0d4631358c735033094155e593aa59d2a15df552cf771cdb08906e428751 +size 294912 diff --git a/deepseek-r1-1.5b-ternary/model_layers_26_self_attn_q_proj_weight.pos b/deepseek-r1-1.5b-ternary/model_layers_26_self_attn_q_proj_weight.pos new file mode 100644 index 0000000000000000000000000000000000000000..3ca2954ccd8fb7a1b2ea6269b3236555ad170c9c --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_26_self_attn_q_proj_weight.pos @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e27de81557c74bc18ed14c349d76eb814a16e29c7a8d8d4fde2205fe47a486fd +size 294912 diff --git a/deepseek-r1-1.5b-ternary/model_layers_27_mlp_down_proj_weight.neg b/deepseek-r1-1.5b-ternary/model_layers_27_mlp_down_proj_weight.neg new file mode 100644 index 0000000000000000000000000000000000000000..16ef6db0f0c612269f250799b05f111bcf436c34 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_27_mlp_down_proj_weight.neg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:10bd40282fcae0b4a73a878dbfd949ffed8e63a70ef4446deeb334e6bb8204c8 +size 1720320 diff --git a/deepseek-r1-1.5b-ternary/model_layers_27_mlp_up_proj_weight.pos b/deepseek-r1-1.5b-ternary/model_layers_27_mlp_up_proj_weight.pos new file mode 100644 index 0000000000000000000000000000000000000000..1f8249c33a3052e730e6638a05aeb1642a40ae1d --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_27_mlp_up_proj_weight.pos @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4916bb673cf1483495cc6da0fcd0f626995d7766570007b64e07c5514ded1875 +size 1720320 diff --git a/deepseek-r1-1.5b-ternary/model_layers_27_self_attn_q_proj_weight.pos b/deepseek-r1-1.5b-ternary/model_layers_27_self_attn_q_proj_weight.pos new file mode 100644 index 0000000000000000000000000000000000000000..ec54a5fd44b0626315467070f95c8f71036ebd66 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_27_self_attn_q_proj_weight.pos @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9849b7232fb27f95d84aefe29e9847ad47f72c6a4123769ec5098af2f9430ccd +size 294912 diff --git a/deepseek-r1-1.5b-ternary/model_layers_2_mlp_down_proj_weight.neg b/deepseek-r1-1.5b-ternary/model_layers_2_mlp_down_proj_weight.neg new file mode 100644 index 0000000000000000000000000000000000000000..ee31b72874640f27321e9d307e250ddcee1712fe --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_2_mlp_down_proj_weight.neg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b8ca9d475e2a1e60b9d3f5b2b3689b54130af8c47558da70706d22fdb41db5f4 +size 1720320 diff --git a/deepseek-r1-1.5b-ternary/model_layers_2_mlp_gate_proj_weight.pos b/deepseek-r1-1.5b-ternary/model_layers_2_mlp_gate_proj_weight.pos new file mode 100644 index 0000000000000000000000000000000000000000..231839e29ece1e470603c4c5210dce9d3d9106dd --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_2_mlp_gate_proj_weight.pos @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4fea9a3d7b44e221c15306d294a87a78699ac40bff7e323e90a9db08fdc33a76 +size 1720320 diff --git a/deepseek-r1-1.5b-ternary/model_layers_2_self_attn_q_proj_weight.pos b/deepseek-r1-1.5b-ternary/model_layers_2_self_attn_q_proj_weight.pos new file mode 100644 index 0000000000000000000000000000000000000000..d51f0717d45f3f9617ec82facb5ffa201befbe43 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_2_self_attn_q_proj_weight.pos @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f02104d807df0a448d548b43a8f066610bddc3c665b286d19b58d3b0fb8b84e2 +size 294912 diff --git a/deepseek-r1-1.5b-ternary/model_layers_3_mlp_gate_proj_weight.pos b/deepseek-r1-1.5b-ternary/model_layers_3_mlp_gate_proj_weight.pos new file mode 100644 index 0000000000000000000000000000000000000000..1f7301902c88a50e48b995b49cae2b4e4fafc980 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_3_mlp_gate_proj_weight.pos @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fe6a2df0ce2b9c79a7b6993744fc08aaf602b19a59d5cd1813f23eba2112236b +size 1720320 diff --git a/deepseek-r1-1.5b-ternary/model_layers_3_self_attn_o_proj_weight.neg b/deepseek-r1-1.5b-ternary/model_layers_3_self_attn_o_proj_weight.neg new file mode 100644 index 0000000000000000000000000000000000000000..ccad1281b30ff41b812d8c8d35c151eea59841d3 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_3_self_attn_o_proj_weight.neg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f726103df735a0eead6030a664caa465cec5cf459220e02211f51636f16b18db +size 294912 diff --git a/deepseek-r1-1.5b-ternary/model_layers_3_self_attn_o_proj_weight.pos b/deepseek-r1-1.5b-ternary/model_layers_3_self_attn_o_proj_weight.pos new file mode 100644 index 0000000000000000000000000000000000000000..9b88eae1996169248e3b16ec7911157734c94cfd --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_3_self_attn_o_proj_weight.pos @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a3db27a0710aa9f7794c375ecb34236370aab752e4bb19f51f9940b8fdcbc6b0 +size 294912 diff --git a/deepseek-r1-1.5b-ternary/model_layers_3_self_attn_q_proj_weight.pos b/deepseek-r1-1.5b-ternary/model_layers_3_self_attn_q_proj_weight.pos new file mode 100644 index 0000000000000000000000000000000000000000..530ef54325f063fc7425c45b2c4aa4966c0d06df --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_3_self_attn_q_proj_weight.pos @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1ae334b885782768f7808acc11f3408169d1e2bf9d74ac8f7b08bd61c57a7b2d +size 294912 diff --git a/deepseek-r1-1.5b-ternary/model_layers_4_mlp_down_proj_weight.neg b/deepseek-r1-1.5b-ternary/model_layers_4_mlp_down_proj_weight.neg new file mode 100644 index 0000000000000000000000000000000000000000..82677f6d6e03427c119a90167de2bf8f9da4011a --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_4_mlp_down_proj_weight.neg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:995ba6d65e368ccc4d857828b3cb1c68f7aedca3831e87de983bd4fe82761ce0 +size 1720320 diff --git a/deepseek-r1-1.5b-ternary/model_layers_4_mlp_gate_proj_weight.pos b/deepseek-r1-1.5b-ternary/model_layers_4_mlp_gate_proj_weight.pos new file mode 100644 index 0000000000000000000000000000000000000000..64292c449d865d362fb69d27d7320cef0c097d1e --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_4_mlp_gate_proj_weight.pos @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e669898986c24ed7a91f210fed1aa6dad721443b9d66e0fc50fa1a592f536a91 +size 1720320 diff --git a/deepseek-r1-1.5b-ternary/model_layers_4_self_attn_o_proj_weight.pos b/deepseek-r1-1.5b-ternary/model_layers_4_self_attn_o_proj_weight.pos new file mode 100644 index 0000000000000000000000000000000000000000..4acb318335b66e9237d11590916d3a23381ce14d --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_4_self_attn_o_proj_weight.pos @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fdf0abb4d775d8a18bf36626814206f1f42cf68e4d21943343cc04a286061719 +size 294912 diff --git a/deepseek-r1-1.5b-ternary/model_layers_4_self_attn_q_proj_weight.neg b/deepseek-r1-1.5b-ternary/model_layers_4_self_attn_q_proj_weight.neg new file mode 100644 index 0000000000000000000000000000000000000000..5704668fc43ec036b843c97854a37210c0c83741 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_4_self_attn_q_proj_weight.neg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c45ea6cd5daf94f3e47994d4dba1855d5f9427015e01e899b00a18a3e2c20249 +size 294912 diff --git a/deepseek-r1-1.5b-ternary/model_layers_5_mlp_gate_proj_weight.neg b/deepseek-r1-1.5b-ternary/model_layers_5_mlp_gate_proj_weight.neg new file mode 100644 index 0000000000000000000000000000000000000000..3228287d1a84a3a2104aa0c0624fe99268ff3821 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_5_mlp_gate_proj_weight.neg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1ef0d3015e86e0ed2fd847fda675b6f5d75c242791f8c0b0394de8d699e7bc32 +size 1720320 diff --git a/deepseek-r1-1.5b-ternary/model_layers_5_self_attn_q_proj_weight.neg b/deepseek-r1-1.5b-ternary/model_layers_5_self_attn_q_proj_weight.neg new file mode 100644 index 0000000000000000000000000000000000000000..2f6509017d42d227fa48a0744a8e080653566abf --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_5_self_attn_q_proj_weight.neg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fd923f3e42cfc262156dee3be05491b6f04e7295b6df7b76077b3a62b7712cca +size 294912 diff --git a/deepseek-r1-1.5b-ternary/model_layers_6_mlp_down_proj_weight.pos b/deepseek-r1-1.5b-ternary/model_layers_6_mlp_down_proj_weight.pos new file mode 100644 index 0000000000000000000000000000000000000000..9817d80d9203b7d45f0c0b049f3d739df4630c51 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_6_mlp_down_proj_weight.pos @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:63ab02dacd14b1d4d5c94e0ad53de2b2120a5af77391e862c942944f8091dd43 +size 1720320 diff --git a/deepseek-r1-1.5b-ternary/model_layers_6_self_attn_o_proj_weight.pos b/deepseek-r1-1.5b-ternary/model_layers_6_self_attn_o_proj_weight.pos new file mode 100644 index 0000000000000000000000000000000000000000..db84035600d0897326911f14974be6821736e368 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_6_self_attn_o_proj_weight.pos @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7a29e7f17c6801cf4026275d343ba441b7c75b6abdefe8031a2f706a2fc3d098 +size 294912 diff --git a/deepseek-r1-1.5b-ternary/model_layers_7_mlp_gate_proj_weight.neg b/deepseek-r1-1.5b-ternary/model_layers_7_mlp_gate_proj_weight.neg new file mode 100644 index 0000000000000000000000000000000000000000..f9183718f56d519e219a73396136c8c16609db23 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_7_mlp_gate_proj_weight.neg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:149542a73786c03d3bc2d96e4a8b225f5a9f2e32c0143bd806faa80ac180e45a +size 1720320 diff --git a/deepseek-r1-1.5b-ternary/model_layers_7_mlp_up_proj_weight.neg b/deepseek-r1-1.5b-ternary/model_layers_7_mlp_up_proj_weight.neg new file mode 100644 index 0000000000000000000000000000000000000000..c46b98abe3bcab322760a3c1f4ea21db1ef446aa --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_7_mlp_up_proj_weight.neg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a0f58ef6fb2f762fc4a12ff97616be24e145bbde12e5a9f73a76a449eebce6b4 +size 1720320 diff --git a/deepseek-r1-1.5b-ternary/model_layers_7_mlp_up_proj_weight.pos b/deepseek-r1-1.5b-ternary/model_layers_7_mlp_up_proj_weight.pos new file mode 100644 index 0000000000000000000000000000000000000000..288d449c49230c3503247440ad537902f0c13e60 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_7_mlp_up_proj_weight.pos @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:990caa2f8137a8478376cb7cab6f000c0401528febd2a3fec07bccdb8045eed4 +size 1720320 diff --git a/deepseek-r1-1.5b-ternary/model_layers_7_self_attn_q_proj_weight.neg b/deepseek-r1-1.5b-ternary/model_layers_7_self_attn_q_proj_weight.neg new file mode 100644 index 0000000000000000000000000000000000000000..6aabf1b6f738e968fac473923be862857213f39e --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_7_self_attn_q_proj_weight.neg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bef77c7a6dc8103b26d3a244c2e1951ad4109fb07e1ae0c7fc870943ab63f906 +size 294912 diff --git a/deepseek-r1-1.5b-ternary/model_layers_8_self_attn_q_proj_weight.pos b/deepseek-r1-1.5b-ternary/model_layers_8_self_attn_q_proj_weight.pos new file mode 100644 index 0000000000000000000000000000000000000000..abad2cb0fd90a4561655df4ff4de7cf0f2f31c7d --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_8_self_attn_q_proj_weight.pos @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9895c8bf160cbc1742e125954db155e8fd6d7ece33325f359cb372f754f026eb +size 294912 diff --git a/deepseek-r1-1.5b-ternary/model_layers_9_mlp_down_proj_weight.neg b/deepseek-r1-1.5b-ternary/model_layers_9_mlp_down_proj_weight.neg new file mode 100644 index 0000000000000000000000000000000000000000..d34c60c30b111824760f95143f7e9c3126399ddb --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_9_mlp_down_proj_weight.neg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0070bae2abbb86b32e1c90da56084fc9e12a97b9c99a0505048f0db668e20f11 +size 1720320 diff --git a/deepseek-r1-1.5b-ternary/model_layers_9_mlp_gate_proj_weight.pos b/deepseek-r1-1.5b-ternary/model_layers_9_mlp_gate_proj_weight.pos new file mode 100644 index 0000000000000000000000000000000000000000..e40f66e8ebda2fd344167f02912e1e3b5828f4de --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_9_mlp_gate_proj_weight.pos @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:934fc7083d7f772c64ac86e8ef1898067e252f5059ed332901e4ef90c7300aff +size 1720320 diff --git a/deepseek-r1-1.5b-ternary/model_layers_9_mlp_up_proj_weight.neg b/deepseek-r1-1.5b-ternary/model_layers_9_mlp_up_proj_weight.neg new file mode 100644 index 0000000000000000000000000000000000000000..bca07cd18ffda5f11eab39d5bd4332bea44f83b1 --- /dev/null +++ b/deepseek-r1-1.5b-ternary/model_layers_9_mlp_up_proj_weight.neg @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:357ef457b2514401ccaadb7b6769248ef2e03cee811f2f352a5726eea88db624 +size 1720320 diff --git a/deepseek-r1-1.5b-unary4/model_layers_0_mlp_gate_proj_weight.planes b/deepseek-r1-1.5b-unary4/model_layers_0_mlp_gate_proj_weight.planes new file mode 100644 index 0000000000000000000000000000000000000000..25ab94079f8ad4dfca17f56aea8334aa5c35dba8 --- /dev/null +++ b/deepseek-r1-1.5b-unary4/model_layers_0_mlp_gate_proj_weight.planes @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:34670d6921c210b4b4c7d9a4dcf696f194252769f447799ccc23b9837d408ba9 +size 6881280 diff --git a/deepseek-r1-1.5b-unary4/model_layers_16_self_attn_q_proj_weight.sign b/deepseek-r1-1.5b-unary4/model_layers_16_self_attn_q_proj_weight.sign new file mode 100644 index 0000000000000000000000000000000000000000..98d270a234a20e9cd20771d295101eaec73317b7 --- /dev/null +++ b/deepseek-r1-1.5b-unary4/model_layers_16_self_attn_q_proj_weight.sign @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e2b813fc3e9787f2d0bc03c73674ad124a10e24382e18b0b29acba670d9f8e33 +size 294912 diff --git a/deepseek-r1-1.5b-unary4/model_layers_19_self_attn_q_proj_weight.planes b/deepseek-r1-1.5b-unary4/model_layers_19_self_attn_q_proj_weight.planes new file mode 100644 index 0000000000000000000000000000000000000000..c98dc51a7110877dff74fe058414d88481bd1350 --- /dev/null +++ b/deepseek-r1-1.5b-unary4/model_layers_19_self_attn_q_proj_weight.planes @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e0302b495acee0e2564b8870c68e8260f73959eb2ff3e451350667d6889da99a +size 1179648 diff --git a/deepseek-r1-1.5b-unary4/model_layers_27_mlp_gate_proj_weight.sign b/deepseek-r1-1.5b-unary4/model_layers_27_mlp_gate_proj_weight.sign new file mode 100644 index 0000000000000000000000000000000000000000..b4172cfccfb7e5f47754cf34019afbec99fcb194 --- /dev/null +++ b/deepseek-r1-1.5b-unary4/model_layers_27_mlp_gate_proj_weight.sign @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7742df58b2eb17912b598df0a2c405b98498924d8ff2523673d0c4d7137fa785 +size 1720320 diff --git a/deepseek-r1-1.5b-unary4/model_layers_27_self_attn_v_proj_weight.planes b/deepseek-r1-1.5b-unary4/model_layers_27_self_attn_v_proj_weight.planes new file mode 100644 index 0000000000000000000000000000000000000000..c9402767b54d9925473622db4a5d203d41a2b987 --- /dev/null +++ b/deepseek-r1-1.5b-unary4/model_layers_27_self_attn_v_proj_weight.planes @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:eb22f357971feaf7684fac9b0d976ab4d49312075ae8bd9e146c6e9f0e99812b +size 196608 diff --git a/deepseek-r1-1.5b-unary4/model_layers_3_mlp_up_proj_weight.planes b/deepseek-r1-1.5b-unary4/model_layers_3_mlp_up_proj_weight.planes new file mode 100644 index 0000000000000000000000000000000000000000..d3ddea9a8ec042d0c6f1d58481be41f023378817 --- /dev/null +++ b/deepseek-r1-1.5b-unary4/model_layers_3_mlp_up_proj_weight.planes @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:155d4e644a8be8d798c8dadc93ade73591e637e6a5d691404b9937522c4cf0d0 +size 6881280 diff --git a/qwen3-4b-log5-unary/model_layers_0_mlp_up_proj_weight.planes b/qwen3-4b-log5-unary/model_layers_0_mlp_up_proj_weight.planes new file mode 100644 index 0000000000000000000000000000000000000000..9608e0be312fbbed8cfbaa4a1022be63e54ba406 --- /dev/null +++ b/qwen3-4b-log5-unary/model_layers_0_mlp_up_proj_weight.planes @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8b584511842fab8b90a3040f2611a6fecf4ec57aeec86c5aefa7fd1e6094d074 +size 15564800 diff --git a/qwen3-4b-log5-unary/model_layers_11_self_attn_o_proj_weight.planes b/qwen3-4b-log5-unary/model_layers_11_self_attn_o_proj_weight.planes new file mode 100644 index 0000000000000000000000000000000000000000..ed898cadf123c1a170457738b4b08111377c2ad1 --- /dev/null +++ b/qwen3-4b-log5-unary/model_layers_11_self_attn_o_proj_weight.planes @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cdfde19cfe64a36f4740313772fd1b13e9d9c6e216df2b7550d97fdcbba761e0 +size 6553600 diff --git a/qwen3-4b-log5-unary/model_layers_17_mlp_down_proj_weight.sign b/qwen3-4b-log5-unary/model_layers_17_mlp_down_proj_weight.sign new file mode 100644 index 0000000000000000000000000000000000000000..8e09d9d223815190c089d0f4d6c0534019fcb186 --- /dev/null +++ b/qwen3-4b-log5-unary/model_layers_17_mlp_down_proj_weight.sign @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e7cbb3fc7d262901fe3bc6bc3a269938a5c985bc90a09e4ae079e3bce22499ca +size 3112960 diff --git a/qwen3-4b-log5-unary/model_layers_17_self_attn_o_proj_weight.sign b/qwen3-4b-log5-unary/model_layers_17_self_attn_o_proj_weight.sign new file mode 100644 index 0000000000000000000000000000000000000000..2fbd39cd7d1290d8b3123cc4733648cb8099e2c8 --- /dev/null +++ b/qwen3-4b-log5-unary/model_layers_17_self_attn_o_proj_weight.sign @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:aacf244f4ce47b91448084d46a6c37682c796a56edc559235012acf54ffb467a +size 1310720 diff --git a/qwen3-4b-log5-unary/model_layers_18_mlp_gate_proj_weight.sign b/qwen3-4b-log5-unary/model_layers_18_mlp_gate_proj_weight.sign new file mode 100644 index 0000000000000000000000000000000000000000..549234cb1db75ef85b8ca7bd764aa26f462d0c56 --- /dev/null +++ b/qwen3-4b-log5-unary/model_layers_18_mlp_gate_proj_weight.sign @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c9f3655956e9e7b386b54131222076501d20746262a6883b48ef16a60c3775ee +size 3112960 diff --git a/qwen3-4b-log5-unary/model_layers_18_mlp_up_proj_weight.sign b/qwen3-4b-log5-unary/model_layers_18_mlp_up_proj_weight.sign new file mode 100644 index 0000000000000000000000000000000000000000..1ec19899d44f415649e3a988665c2292b90e5408 --- /dev/null +++ b/qwen3-4b-log5-unary/model_layers_18_mlp_up_proj_weight.sign @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1a2aed56086166f006acae689a66849b5ed853bb8bd23f05e79dee6ecbd55c66 +size 3112960 diff --git a/qwen3-4b-log5-unary/model_layers_21_self_attn_k_proj_weight.planes b/qwen3-4b-log5-unary/model_layers_21_self_attn_k_proj_weight.planes new file mode 100644 index 0000000000000000000000000000000000000000..81f99b5ddcd0df830c529d185aba8c49bdc25059 --- /dev/null +++ b/qwen3-4b-log5-unary/model_layers_21_self_attn_k_proj_weight.planes @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8c55c12fb09486675b895ed96dc46057ec275bc5762c1bb5f312805353ff1d93 +size 1638400 diff --git a/qwen3-4b-log5-unary/model_layers_25_self_attn_k_proj_weight.planes b/qwen3-4b-log5-unary/model_layers_25_self_attn_k_proj_weight.planes new file mode 100644 index 0000000000000000000000000000000000000000..08482cea26bb5a0c0fc02fe0ab4db6d0431c19b4 --- /dev/null +++ b/qwen3-4b-log5-unary/model_layers_25_self_attn_k_proj_weight.planes @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6b3763d0c486c5150eafc95fbf9938d20e0c26e8b906c5a0eba9fa0534f341d7 +size 1638400 diff --git a/qwen3-4b-log5-unary/model_layers_25_self_attn_o_proj_weight.planes b/qwen3-4b-log5-unary/model_layers_25_self_attn_o_proj_weight.planes new file mode 100644 index 0000000000000000000000000000000000000000..97d726361bfc8421dff95b9ed28f14946244d9a8 --- /dev/null +++ b/qwen3-4b-log5-unary/model_layers_25_self_attn_o_proj_weight.planes @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d02768943e176322861654af8d06f3e037d37457edbb45160b6342c6f1a27003 +size 6553600 diff --git a/qwen3-4b-log5-unary/model_layers_29_self_attn_v_proj_weight.planes b/qwen3-4b-log5-unary/model_layers_29_self_attn_v_proj_weight.planes new file mode 100644 index 0000000000000000000000000000000000000000..de0f0bb7a575f3ff8d4f47b923808e3d4716f315 --- /dev/null +++ b/qwen3-4b-log5-unary/model_layers_29_self_attn_v_proj_weight.planes @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a56ccd37eceda6db7facc532deae6df4414d175bab046c6b959d0764e005329d +size 1638400 diff --git a/qwen3-4b-log5-unary/model_layers_2_self_attn_o_proj_weight.sign b/qwen3-4b-log5-unary/model_layers_2_self_attn_o_proj_weight.sign new file mode 100644 index 0000000000000000000000000000000000000000..ae06d7e6e287c0c5ac2a094a28398689d79e9cfe --- /dev/null +++ b/qwen3-4b-log5-unary/model_layers_2_self_attn_o_proj_weight.sign @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:118295872098475cdd3f7f2f9f30ab830a6a43903a4080d485b43c38167b6a4e +size 1310720 diff --git a/qwen3-4b-log5-unary/model_layers_31_self_attn_k_proj_weight.planes b/qwen3-4b-log5-unary/model_layers_31_self_attn_k_proj_weight.planes new file mode 100644 index 0000000000000000000000000000000000000000..1c57c863eda5f5a2abbfdb6312000673e39a1577 --- /dev/null +++ b/qwen3-4b-log5-unary/model_layers_31_self_attn_k_proj_weight.planes @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:91b76a4aa5022d9197ccfd9187b3a822d46a2e06d09f15cb908dd849282556be +size 1638400 diff --git a/qwen3-4b-log5-unary/model_layers_33_self_attn_o_proj_weight.planes b/qwen3-4b-log5-unary/model_layers_33_self_attn_o_proj_weight.planes new file mode 100644 index 0000000000000000000000000000000000000000..ac3f374285364207c72b090a765765be4b4a92ce --- /dev/null +++ b/qwen3-4b-log5-unary/model_layers_33_self_attn_o_proj_weight.planes @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9e60dbf26bc9d093b05080de0abf12f0e7f5c44142e6a9e59208be94d3719cd1 +size 6553600 diff --git a/qwen3-4b-log5-unary/model_layers_34_mlp_up_proj_weight.sign b/qwen3-4b-log5-unary/model_layers_34_mlp_up_proj_weight.sign new file mode 100644 index 0000000000000000000000000000000000000000..4131b12db4f54532a2b9759c92673487a7fa0c5b --- /dev/null +++ b/qwen3-4b-log5-unary/model_layers_34_mlp_up_proj_weight.sign @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:03c13e473cd87ab44281566b3fa9f0d28b5d419fed485716cc33599f88239db1 +size 3112960 diff --git a/qwen3-4b-log5-unary/model_layers_5_self_attn_k_proj_weight.planes b/qwen3-4b-log5-unary/model_layers_5_self_attn_k_proj_weight.planes new file mode 100644 index 0000000000000000000000000000000000000000..dbd44237e12ee656fb4cb07d003a10a891c93215 --- /dev/null +++ b/qwen3-4b-log5-unary/model_layers_5_self_attn_k_proj_weight.planes @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f8d4cf938d65abc3ea1b176117faf6de65506ef1aa284ac1bc57053bf38ab743 +size 1638400 diff --git a/qwen3-4b-log5-unary/model_layers_5_self_attn_o_proj_weight.sign b/qwen3-4b-log5-unary/model_layers_5_self_attn_o_proj_weight.sign new file mode 100644 index 0000000000000000000000000000000000000000..3b2283b7f198cbf4bade2303643c2b09883c0a12 --- /dev/null +++ b/qwen3-4b-log5-unary/model_layers_5_self_attn_o_proj_weight.sign @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:57778519130fffa13350195c71736135f3474e2376a93402079bfe87c2e14471 +size 1310720 diff --git a/qwen3-4b-log5-unary/model_layers_6_mlp_up_proj_weight.planes b/qwen3-4b-log5-unary/model_layers_6_mlp_up_proj_weight.planes new file mode 100644 index 0000000000000000000000000000000000000000..eb1c3f9d9a194f375fc09ead1108bf0e90a07559 --- /dev/null +++ b/qwen3-4b-log5-unary/model_layers_6_mlp_up_proj_weight.planes @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5e578d440027a6059ba44f728da67b62df301004dab6ad6f63b56398f574b063 +size 15564800 diff --git a/qwen3-4b-log5-unary/model_layers_6_mlp_up_proj_weight.sign b/qwen3-4b-log5-unary/model_layers_6_mlp_up_proj_weight.sign new file mode 100644 index 0000000000000000000000000000000000000000..07afd70a08a3f034145e5e19af48a93a8b359016 --- /dev/null +++ b/qwen3-4b-log5-unary/model_layers_6_mlp_up_proj_weight.sign @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6ce199c8e97e64fd5d1fcff9b03cde5d49f766da6d9687713b54ad94448cc0d3 +size 3112960