diff --git a/deepseek-r1-1.5b-packed/model_layers_0_mlp_down_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_0_mlp_down_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..f10de1ba331594c8d489369e10ab97b8a6fb7d6d Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_0_mlp_down_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_0_mlp_gate_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_0_mlp_gate_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..1c39b5143630b05d0fad56c5d1e23be60d6df84d Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_0_mlp_gate_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_0_self_attn_k_proj_bias.fp16 b/deepseek-r1-1.5b-packed/model_layers_0_self_attn_k_proj_bias.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..91e5335bba44952236c965e1d087a417fa1872a2 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_0_self_attn_k_proj_bias.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_0_self_attn_k_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_0_self_attn_k_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..9401fc7bec41306420e2f625efdc88cd8c92717d --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_0_self_attn_k_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_0_self_attn_k_proj_weight.signs b/deepseek-r1-1.5b-packed/model_layers_0_self_attn_k_proj_weight.signs new file mode 100644 index 0000000000000000000000000000000000000000..1b7be81ea40572ce7b09dec92bbe585ab6a7cb9e Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_0_self_attn_k_proj_weight.signs differ diff --git a/deepseek-r1-1.5b-packed/model_layers_0_self_attn_o_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_0_self_attn_o_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..8f9f41dac12648ab0cb836a8b188da92c1f5e479 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_0_self_attn_o_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_0_self_attn_q_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_0_self_attn_q_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..c86839c59145849e367cc3aa33182b1c6c31a614 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_0_self_attn_q_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_0_self_attn_v_proj_bias.fp16 b/deepseek-r1-1.5b-packed/model_layers_0_self_attn_v_proj_bias.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..5837e3fe1a534632322c5961f2b75f7c554f88c3 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_0_self_attn_v_proj_bias.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_0_self_attn_v_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_0_self_attn_v_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..c61d09f003e0ad48300f767500bd9678f41cc95f Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_0_self_attn_v_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_10_input_layernorm_weight.fp16 b/deepseek-r1-1.5b-packed/model_layers_10_input_layernorm_weight.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..ca42714370c9e4663825711e822d6fd9b4f3a7b2 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_10_input_layernorm_weight.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_10_mlp_down_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_10_mlp_down_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..c86839c59145849e367cc3aa33182b1c6c31a614 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_10_mlp_down_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_10_mlp_down_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_10_mlp_down_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..f444e68fc23019bf2a0fa5d1f84a2c62c8bb3303 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_10_mlp_down_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_10_mlp_gate_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_10_mlp_gate_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..3d5a76a5a0c43488828c17fc8a7002591a508098 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_10_mlp_gate_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_10_post_attention_layernorm_weight.fp16 b/deepseek-r1-1.5b-packed/model_layers_10_post_attention_layernorm_weight.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..3497c41e3219091d7128e2340d793af71f043b06 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_10_post_attention_layernorm_weight.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_10_self_attn_k_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_10_self_attn_k_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..b91ade8965d1452dcac669e2ffc70f01689a1a8e Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_10_self_attn_k_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_10_self_attn_k_proj_weight.signs b/deepseek-r1-1.5b-packed/model_layers_10_self_attn_k_proj_weight.signs new file mode 100644 index 0000000000000000000000000000000000000000..9638a0f8929beeff726db96213daf5c337439cca Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_10_self_attn_k_proj_weight.signs differ diff --git a/deepseek-r1-1.5b-packed/model_layers_11_mlp_gate_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_11_mlp_gate_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..3d5a76a5a0c43488828c17fc8a7002591a508098 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_11_mlp_gate_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_11_mlp_gate_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_11_mlp_gate_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..be786a08297cb0043531b0130132d5bbce5106f6 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_11_mlp_gate_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_11_mlp_up_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_11_mlp_up_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..adf985a02f2f4fa712c8a55f1b3944d737cd1dcd Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_11_mlp_up_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_11_self_attn_k_proj_bias.fp16 b/deepseek-r1-1.5b-packed/model_layers_11_self_attn_k_proj_bias.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..062db56c99f1e0f7bf63d93fa5f28fe511ff6724 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_11_self_attn_k_proj_bias.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_11_self_attn_k_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_11_self_attn_k_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..9401fc7bec41306420e2f625efdc88cd8c92717d --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_11_self_attn_k_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_11_self_attn_k_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_11_self_attn_k_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..e8f12eb9652fdf8ce340b65b67aaa6ccffc6ba48 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_11_self_attn_k_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_11_self_attn_o_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_11_self_attn_o_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..c86839c59145849e367cc3aa33182b1c6c31a614 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_11_self_attn_o_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_11_self_attn_o_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_11_self_attn_o_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..975b06f3027b18a05b9d35ef12bdff2697151d87 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_11_self_attn_o_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_11_self_attn_q_proj_bias.fp16 b/deepseek-r1-1.5b-packed/model_layers_11_self_attn_q_proj_bias.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..fc82d2ce243529e924112e93801b2c31a9313d50 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_11_self_attn_q_proj_bias.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_11_self_attn_q_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_11_self_attn_q_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..c86839c59145849e367cc3aa33182b1c6c31a614 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_11_self_attn_q_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_11_self_attn_q_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_11_self_attn_q_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..7a3364d9f4510bad38eb942865fd5f9a3a76cf0d Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_11_self_attn_q_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_11_self_attn_v_proj_bias.fp16 b/deepseek-r1-1.5b-packed/model_layers_11_self_attn_v_proj_bias.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..9017af4926876f597c24e2bd9a435e002d4a08d3 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_11_self_attn_v_proj_bias.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_11_self_attn_v_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_11_self_attn_v_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..9401fc7bec41306420e2f625efdc88cd8c92717d --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_11_self_attn_v_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_11_self_attn_v_proj_weight.signs b/deepseek-r1-1.5b-packed/model_layers_11_self_attn_v_proj_weight.signs new file mode 100644 index 0000000000000000000000000000000000000000..f802f66e6b93746d41e2fcc658629febf5d95166 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_11_self_attn_v_proj_weight.signs differ diff --git a/deepseek-r1-1.5b-packed/model_layers_12_input_layernorm_weight.fp16 b/deepseek-r1-1.5b-packed/model_layers_12_input_layernorm_weight.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..2ad77dec2867326bc94701462d61e8e7d82f5090 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_12_input_layernorm_weight.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_12_mlp_gate_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_12_mlp_gate_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..3e06b0a8d6a4e547ec5978ac92f3c8458a0f74ea Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_12_mlp_gate_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_12_self_attn_k_proj_bias.fp16 b/deepseek-r1-1.5b-packed/model_layers_12_self_attn_k_proj_bias.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..237e1bcd7818596e05b37e3f277fbd00ea2f6b44 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_12_self_attn_k_proj_bias.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_12_self_attn_q_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_12_self_attn_q_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..eb69f9796546e6dde22ea533f89b1c9e56184e16 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_12_self_attn_q_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_12_self_attn_v_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_12_self_attn_v_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..9401fc7bec41306420e2f625efdc88cd8c92717d --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_12_self_attn_v_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_12_self_attn_v_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_12_self_attn_v_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..f44c726f9ad808c3abfb0638f77e0deaad984081 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_12_self_attn_v_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_13_input_layernorm_weight.fp16 b/deepseek-r1-1.5b-packed/model_layers_13_input_layernorm_weight.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..40b686b83607f6f85048ff11f9cbd63cf1b152cc Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_13_input_layernorm_weight.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_13_mlp_gate_proj_weight.mags b/deepseek-r1-1.5b-packed/model_layers_13_mlp_gate_proj_weight.mags new file mode 100644 index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391 diff --git a/deepseek-r1-1.5b-packed/model_layers_13_mlp_gate_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_13_mlp_gate_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..3d5a76a5a0c43488828c17fc8a7002591a508098 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_13_mlp_gate_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_13_mlp_gate_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_13_mlp_gate_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..4ba206b77da02e0ff83c14d4d5f1bb58df2af316 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_13_mlp_gate_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_13_mlp_up_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_13_mlp_up_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..3d5a76a5a0c43488828c17fc8a7002591a508098 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_13_mlp_up_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_13_mlp_up_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_13_mlp_up_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..4a1b9b8669704f06b32fec1228b38afc5eead24d Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_13_mlp_up_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_13_self_attn_k_proj_bias.fp16 b/deepseek-r1-1.5b-packed/model_layers_13_self_attn_k_proj_bias.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..8908159071055644e0ee977d0b7eeb0fc888d7e6 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_13_self_attn_k_proj_bias.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_13_self_attn_k_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_13_self_attn_k_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..43e48e4635864479dfa6d648382c468b186bd6e7 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_13_self_attn_k_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_13_self_attn_q_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_13_self_attn_q_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..5df8d719b2a506904700a0fe9bc5bddec330e3a1 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_13_self_attn_q_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_13_self_attn_v_proj_weight.signs b/deepseek-r1-1.5b-packed/model_layers_13_self_attn_v_proj_weight.signs new file mode 100644 index 0000000000000000000000000000000000000000..faa664846e9e36b3c9ae43ddc6a88401f10f715b Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_13_self_attn_v_proj_weight.signs differ diff --git a/deepseek-r1-1.5b-packed/model_layers_14_mlp_down_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_14_mlp_down_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..1060d6740ea60515c3b222fa2a656cf094f9ce13 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_14_mlp_down_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_14_mlp_gate_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_14_mlp_gate_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..038dee896f1d5a2d89aad475c1d537cbe881f82f Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_14_mlp_gate_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_14_mlp_up_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_14_mlp_up_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..b63d6a6bff46c56e00f447ae4f1ac349e6a6cebd Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_14_mlp_up_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_14_self_attn_o_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_14_self_attn_o_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..c86839c59145849e367cc3aa33182b1c6c31a614 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_14_self_attn_o_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_14_self_attn_q_proj_bias.fp16 b/deepseek-r1-1.5b-packed/model_layers_14_self_attn_q_proj_bias.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..e5876032eaa1d29719c88f2b615f066d349e86c8 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_14_self_attn_q_proj_bias.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_14_self_attn_q_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_14_self_attn_q_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..c86839c59145849e367cc3aa33182b1c6c31a614 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_14_self_attn_q_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_14_self_attn_v_proj_bias.fp16 b/deepseek-r1-1.5b-packed/model_layers_14_self_attn_v_proj_bias.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..7a890e6b559ac51462e1f4952a2c1a3a3ba89a84 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_14_self_attn_v_proj_bias.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_15_mlp_down_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_15_mlp_down_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..c86839c59145849e367cc3aa33182b1c6c31a614 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_15_mlp_down_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_15_mlp_down_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_15_mlp_down_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..0697f7e360cdae936e2c70d1b025b2ea2b8fa038 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_15_mlp_down_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_15_mlp_gate_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_15_mlp_gate_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..3d5a76a5a0c43488828c17fc8a7002591a508098 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_15_mlp_gate_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_15_mlp_up_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_15_mlp_up_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..3d5a76a5a0c43488828c17fc8a7002591a508098 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_15_mlp_up_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_15_mlp_up_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_15_mlp_up_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..6c9662eb021e2dbcf26f9c17536d43777842c929 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_15_mlp_up_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_15_self_attn_k_proj_bias.fp16 b/deepseek-r1-1.5b-packed/model_layers_15_self_attn_k_proj_bias.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..0d1fe2adda6c5da4dd8811d102efb923acf8c9a2 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_15_self_attn_k_proj_bias.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_15_self_attn_k_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_15_self_attn_k_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..d52f932eb22721ee9b99de8ab1ff0581d07adf66 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_15_self_attn_k_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_15_self_attn_o_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_15_self_attn_o_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..001d0794ef53c57f4e520dde827246eb2813518b Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_15_self_attn_o_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_15_self_attn_v_proj_weight.signs b/deepseek-r1-1.5b-packed/model_layers_15_self_attn_v_proj_weight.signs new file mode 100644 index 0000000000000000000000000000000000000000..35ede3442b96c602595389e8aa53db0f1d9b7b7e Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_15_self_attn_v_proj_weight.signs differ diff --git a/deepseek-r1-1.5b-packed/model_layers_16_input_layernorm_weight.fp16 b/deepseek-r1-1.5b-packed/model_layers_16_input_layernorm_weight.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..be30eeedbe2adf190e199c0a489983b5fccd2add Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_16_input_layernorm_weight.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_16_mlp_down_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_16_mlp_down_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..3430893fad14359d7607b21f626afcaba3fe9308 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_16_mlp_down_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_16_mlp_up_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_16_mlp_up_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..3d5a76a5a0c43488828c17fc8a7002591a508098 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_16_mlp_up_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_16_mlp_up_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_16_mlp_up_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..3faa7a715a137da030a0061e96a417b05fe17879 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_16_mlp_up_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_16_post_attention_layernorm_weight.fp16 b/deepseek-r1-1.5b-packed/model_layers_16_post_attention_layernorm_weight.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..39da9979d66e686c1f68da2f169a27acd8f8a104 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_16_post_attention_layernorm_weight.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_16_self_attn_k_proj_bias.fp16 b/deepseek-r1-1.5b-packed/model_layers_16_self_attn_k_proj_bias.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..1f50f92e73c0fe3a8f49185fbb4b507a5cc81b8b Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_16_self_attn_k_proj_bias.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_16_self_attn_o_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_16_self_attn_o_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..c86839c59145849e367cc3aa33182b1c6c31a614 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_16_self_attn_o_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_16_self_attn_q_proj_bias.fp16 b/deepseek-r1-1.5b-packed/model_layers_16_self_attn_q_proj_bias.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..c3981a49073d31fef293273e651dc1fc5863b33a Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_16_self_attn_q_proj_bias.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_16_self_attn_q_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_16_self_attn_q_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..c86839c59145849e367cc3aa33182b1c6c31a614 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_16_self_attn_q_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_16_self_attn_v_proj_bias.fp16 b/deepseek-r1-1.5b-packed/model_layers_16_self_attn_v_proj_bias.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..c4a9452345d31cf2acb2cb43c8fb5bca0b8cda38 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_16_self_attn_v_proj_bias.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_16_self_attn_v_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_16_self_attn_v_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..9401fc7bec41306420e2f625efdc88cd8c92717d --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_16_self_attn_v_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_16_self_attn_v_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_16_self_attn_v_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..ab52f4f052e290599c692b678bbe1b4dcb8c92cf Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_16_self_attn_v_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_16_self_attn_v_proj_weight.signs b/deepseek-r1-1.5b-packed/model_layers_16_self_attn_v_proj_weight.signs new file mode 100644 index 0000000000000000000000000000000000000000..3c08a825c8a87f0a8cb998fff9849a8eff130c65 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_16_self_attn_v_proj_weight.signs differ diff --git a/deepseek-r1-1.5b-packed/model_layers_17_mlp_up_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_17_mlp_up_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..fab90074bd91d01727aa5aca94feefb76d972fe3 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_17_mlp_up_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_17_self_attn_o_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_17_self_attn_o_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..91cf4be4ff56ad56068feb6815022032064d049a Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_17_self_attn_o_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_17_self_attn_q_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_17_self_attn_q_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..c86839c59145849e367cc3aa33182b1c6c31a614 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_17_self_attn_q_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_17_self_attn_q_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_17_self_attn_q_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..57ca78721633f2c40e4d7a6012c5344b78a1ff58 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_17_self_attn_q_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_17_self_attn_v_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_17_self_attn_v_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..9401fc7bec41306420e2f625efdc88cd8c92717d --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_17_self_attn_v_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_17_self_attn_v_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_17_self_attn_v_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..f27625be1cd6c1dbcf466ae4422e5dabb0f44f99 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_17_self_attn_v_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_17_self_attn_v_proj_weight.signs b/deepseek-r1-1.5b-packed/model_layers_17_self_attn_v_proj_weight.signs new file mode 100644 index 0000000000000000000000000000000000000000..8c11b3e37584b9f60aa081bfab76f150119307ad Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_17_self_attn_v_proj_weight.signs differ diff --git a/deepseek-r1-1.5b-packed/model_layers_18_input_layernorm_weight.fp16 b/deepseek-r1-1.5b-packed/model_layers_18_input_layernorm_weight.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..2ff9fc8328968f5077b4ed41a2b51418c2497e1c Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_18_input_layernorm_weight.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_18_mlp_down_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_18_mlp_down_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..dd86fdeeb4a48158d9eaa79b153344b052d0ea00 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_18_mlp_down_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_18_mlp_gate_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_18_mlp_gate_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..3d5a76a5a0c43488828c17fc8a7002591a508098 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_18_mlp_gate_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_18_mlp_up_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_18_mlp_up_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..6d64fcc6075197e07109a91607ccbdfbc2275bea Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_18_mlp_up_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_18_self_attn_k_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_18_self_attn_k_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..6a0c41f18bb49516eadffd84095230c074c88146 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_18_self_attn_k_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_18_self_attn_k_proj_weight.signs b/deepseek-r1-1.5b-packed/model_layers_18_self_attn_k_proj_weight.signs new file mode 100644 index 0000000000000000000000000000000000000000..85b45ba6f4e5167da101bc355ade8b675718a561 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_18_self_attn_k_proj_weight.signs differ diff --git a/deepseek-r1-1.5b-packed/model_layers_18_self_attn_o_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_18_self_attn_o_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..c86839c59145849e367cc3aa33182b1c6c31a614 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_18_self_attn_o_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_18_self_attn_q_proj_bias.fp16 b/deepseek-r1-1.5b-packed/model_layers_18_self_attn_q_proj_bias.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..767e3693ebdfc18340e47f89cd82953105e3a7b4 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_18_self_attn_q_proj_bias.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_18_self_attn_q_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_18_self_attn_q_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..c86839c59145849e367cc3aa33182b1c6c31a614 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_18_self_attn_q_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_18_self_attn_v_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_18_self_attn_v_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..00dd42514a5039d619d331e5bc48f59bb860291d Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_18_self_attn_v_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_19_input_layernorm_weight.fp16 b/deepseek-r1-1.5b-packed/model_layers_19_input_layernorm_weight.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..b375bb4e6b6bc0bdcaa5291001295849dd624035 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_19_input_layernorm_weight.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_19_mlp_up_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_19_mlp_up_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..3d5a76a5a0c43488828c17fc8a7002591a508098 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_19_mlp_up_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_19_self_attn_k_proj_weight.signs b/deepseek-r1-1.5b-packed/model_layers_19_self_attn_k_proj_weight.signs new file mode 100644 index 0000000000000000000000000000000000000000..56f5b49ac5d1a30f808692813610c02ccef558b2 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_19_self_attn_k_proj_weight.signs differ diff --git a/deepseek-r1-1.5b-packed/model_layers_19_self_attn_o_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_19_self_attn_o_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..4f3a2f859bcb6f35ecc9e30fe16b3dd580fdc6fb Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_19_self_attn_o_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_19_self_attn_q_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_19_self_attn_q_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..c86839c59145849e367cc3aa33182b1c6c31a614 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_19_self_attn_q_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_19_self_attn_v_proj_weight.signs b/deepseek-r1-1.5b-packed/model_layers_19_self_attn_v_proj_weight.signs new file mode 100644 index 0000000000000000000000000000000000000000..8679c894b553ce64de5b70a825a0a873fa8e9ca2 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_19_self_attn_v_proj_weight.signs differ diff --git a/deepseek-r1-1.5b-packed/model_layers_1_mlp_down_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_1_mlp_down_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..c86839c59145849e367cc3aa33182b1c6c31a614 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_1_mlp_down_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_1_mlp_gate_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_1_mlp_gate_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..3d5a76a5a0c43488828c17fc8a7002591a508098 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_1_mlp_gate_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_1_mlp_gate_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_1_mlp_gate_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..932cced93a3805dce1fd3e5a69cfe9e8a39b19d8 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_1_mlp_gate_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_1_self_attn_o_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_1_self_attn_o_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..c86839c59145849e367cc3aa33182b1c6c31a614 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_1_self_attn_o_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_1_self_attn_o_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_1_self_attn_o_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..44f50da44ade4090a0e3a3fb136f6bec3ee76f6f Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_1_self_attn_o_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_1_self_attn_q_proj_bias.fp16 b/deepseek-r1-1.5b-packed/model_layers_1_self_attn_q_proj_bias.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..4a3cb661de572e92ca2eabe5520d09b7c23ea1ec Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_1_self_attn_q_proj_bias.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_1_self_attn_q_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_1_self_attn_q_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..c86839c59145849e367cc3aa33182b1c6c31a614 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_1_self_attn_q_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_1_self_attn_v_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_1_self_attn_v_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..2982213dc2b89d31f1155c0331c7b82cc8035b79 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_1_self_attn_v_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_20_mlp_down_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_20_mlp_down_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..c86839c59145849e367cc3aa33182b1c6c31a614 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_20_mlp_down_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_20_mlp_down_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_20_mlp_down_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..9305a2ea4db138d8b886b3bb1693903576d36fe7 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_20_mlp_down_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_20_mlp_up_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_20_mlp_up_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..3d5a76a5a0c43488828c17fc8a7002591a508098 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_20_mlp_up_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_20_post_attention_layernorm_weight.fp16 b/deepseek-r1-1.5b-packed/model_layers_20_post_attention_layernorm_weight.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..6c2e8526f934c429b58d28ac364b320464dafe2d Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_20_post_attention_layernorm_weight.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_20_self_attn_k_proj_bias.fp16 b/deepseek-r1-1.5b-packed/model_layers_20_self_attn_k_proj_bias.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..c5d0a76c0c2e67fc853bdb88b3effc526c8fb5c3 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_20_self_attn_k_proj_bias.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_20_self_attn_k_proj_weight.signs b/deepseek-r1-1.5b-packed/model_layers_20_self_attn_k_proj_weight.signs new file mode 100644 index 0000000000000000000000000000000000000000..b4734c6a402fe4103de6be4a3f8456dbde76dd39 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_20_self_attn_k_proj_weight.signs differ diff --git a/deepseek-r1-1.5b-packed/model_layers_20_self_attn_q_proj_bias.fp16 b/deepseek-r1-1.5b-packed/model_layers_20_self_attn_q_proj_bias.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..f74e07c56864dfb82103d6095e565d53065d11a0 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_20_self_attn_q_proj_bias.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_20_self_attn_q_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_20_self_attn_q_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..734f4aa792bab90a0e3a32dcedf9c82b41cfea3d Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_20_self_attn_q_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_21_mlp_up_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_21_mlp_up_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..0c047b92b9a8d8632b86211900674535204acdcd Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_21_mlp_up_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_21_post_attention_layernorm_weight.fp16 b/deepseek-r1-1.5b-packed/model_layers_21_post_attention_layernorm_weight.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..ec4f1d9e6e111004533d729fba9ea0804f85e64f Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_21_post_attention_layernorm_weight.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_21_self_attn_k_proj_bias.fp16 b/deepseek-r1-1.5b-packed/model_layers_21_self_attn_k_proj_bias.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..523a27c555fd02fa97a9a1d03f2641fb2b98655b Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_21_self_attn_k_proj_bias.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_21_self_attn_k_proj_weight.signs b/deepseek-r1-1.5b-packed/model_layers_21_self_attn_k_proj_weight.signs new file mode 100644 index 0000000000000000000000000000000000000000..f6ac4e8f15fbb1b4a9f794eb068543df1f32a7e8 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_21_self_attn_k_proj_weight.signs differ diff --git a/deepseek-r1-1.5b-packed/model_layers_21_self_attn_o_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_21_self_attn_o_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..c86839c59145849e367cc3aa33182b1c6c31a614 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_21_self_attn_o_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_21_self_attn_v_proj_bias.fp16 b/deepseek-r1-1.5b-packed/model_layers_21_self_attn_v_proj_bias.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..7c1f3b98dabb03ff3be28c0b03999064e08fdbdb Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_21_self_attn_v_proj_bias.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_21_self_attn_v_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_21_self_attn_v_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..9401fc7bec41306420e2f625efdc88cd8c92717d --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_21_self_attn_v_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_21_self_attn_v_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_21_self_attn_v_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..f312e0285eb7418284683032ab24eb21325d4140 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_21_self_attn_v_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_21_self_attn_v_proj_weight.signs b/deepseek-r1-1.5b-packed/model_layers_21_self_attn_v_proj_weight.signs new file mode 100644 index 0000000000000000000000000000000000000000..81e039cac497040a37965af24aae244a5cf88db4 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_21_self_attn_v_proj_weight.signs differ diff --git a/deepseek-r1-1.5b-packed/model_layers_22_mlp_gate_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_22_mlp_gate_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..3d5a76a5a0c43488828c17fc8a7002591a508098 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_22_mlp_gate_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_22_mlp_gate_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_22_mlp_gate_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..d9bc0bac38320014b2b12f5029f726f752f871fb Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_22_mlp_gate_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_22_mlp_up_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_22_mlp_up_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..f3827d3c10f19343da6297598aabe2e740f345e3 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_22_mlp_up_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_22_self_attn_o_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_22_self_attn_o_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..7df81e69f24f38526e26172be73b2e3fec3d839a Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_22_self_attn_o_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_22_self_attn_q_proj_bias.fp16 b/deepseek-r1-1.5b-packed/model_layers_22_self_attn_q_proj_bias.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..56d475c2e2ec4b8e1c14680aec79de9a493ea100 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_22_self_attn_q_proj_bias.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_22_self_attn_q_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_22_self_attn_q_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..c86839c59145849e367cc3aa33182b1c6c31a614 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_22_self_attn_q_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_22_self_attn_q_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_22_self_attn_q_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..5ff7c619553041b3ae583ebe8ac3fa20996446ac Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_22_self_attn_q_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_22_self_attn_v_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_22_self_attn_v_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..9401fc7bec41306420e2f625efdc88cd8c92717d --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_22_self_attn_v_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_22_self_attn_v_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_22_self_attn_v_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..2e459a534696404cbed0f5314296f125d074a325 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_22_self_attn_v_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_23_input_layernorm_weight.fp16 b/deepseek-r1-1.5b-packed/model_layers_23_input_layernorm_weight.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..e58a9cd071083fb0d971e7be3858bccd1e181435 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_23_input_layernorm_weight.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_23_mlp_down_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_23_mlp_down_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..a0759c72de842995bffc5d1e1489a482e1d9fd94 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_23_mlp_down_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_23_self_attn_k_proj_bias.fp16 b/deepseek-r1-1.5b-packed/model_layers_23_self_attn_k_proj_bias.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..b7ce333521fda2ad7ce68c00fb18795d3baf81bb Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_23_self_attn_k_proj_bias.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_23_self_attn_k_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_23_self_attn_k_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..9401fc7bec41306420e2f625efdc88cd8c92717d --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_23_self_attn_k_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_23_self_attn_q_proj_bias.fp16 b/deepseek-r1-1.5b-packed/model_layers_23_self_attn_q_proj_bias.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..3f2c4d375b61d03cc2989e5af223fa1e3e0a829e Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_23_self_attn_q_proj_bias.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_23_self_attn_q_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_23_self_attn_q_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..c86839c59145849e367cc3aa33182b1c6c31a614 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_23_self_attn_q_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_23_self_attn_v_proj_bias.fp16 b/deepseek-r1-1.5b-packed/model_layers_23_self_attn_v_proj_bias.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..308f588954d75655c4e3a7a9003561edfcca1de3 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_23_self_attn_v_proj_bias.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_24_mlp_down_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_24_mlp_down_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..c86839c59145849e367cc3aa33182b1c6c31a614 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_24_mlp_down_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_24_mlp_down_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_24_mlp_down_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..c57550e85896f442d57e88413b05a66b44bf7c21 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_24_mlp_down_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_24_mlp_gate_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_24_mlp_gate_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..7d8e5a948d745fae65b55f8e7e07c006c82a33fd Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_24_mlp_gate_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_24_mlp_up_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_24_mlp_up_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..3d5a76a5a0c43488828c17fc8a7002591a508098 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_24_mlp_up_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_24_self_attn_k_proj_bias.fp16 b/deepseek-r1-1.5b-packed/model_layers_24_self_attn_k_proj_bias.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..8a38fe17002304214d00ae90a7bb04de994bec4d Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_24_self_attn_k_proj_bias.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_24_self_attn_o_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_24_self_attn_o_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..c86839c59145849e367cc3aa33182b1c6c31a614 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_24_self_attn_o_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_24_self_attn_q_proj_bias.fp16 b/deepseek-r1-1.5b-packed/model_layers_24_self_attn_q_proj_bias.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..7b43c1825802b8107ad7b1de07eb56ade34cbd47 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_24_self_attn_q_proj_bias.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_24_self_attn_v_proj_bias.fp16 b/deepseek-r1-1.5b-packed/model_layers_24_self_attn_v_proj_bias.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..eb6638a94a285d4a1475833b26c02f53813f38b8 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_24_self_attn_v_proj_bias.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_24_self_attn_v_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_24_self_attn_v_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..7fb4493510df22bd7325f08054b75ee07a6451ed Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_24_self_attn_v_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_25_input_layernorm_weight.fp16 b/deepseek-r1-1.5b-packed/model_layers_25_input_layernorm_weight.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..eee254584de8db018cf609c254bd9d82c6ab810d Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_25_input_layernorm_weight.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_25_mlp_gate_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_25_mlp_gate_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..3d5a76a5a0c43488828c17fc8a7002591a508098 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_25_mlp_gate_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_25_self_attn_k_proj_bias.fp16 b/deepseek-r1-1.5b-packed/model_layers_25_self_attn_k_proj_bias.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..ced68c5c7a3d2f5653342e791ed0a345d520f2bd Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_25_self_attn_k_proj_bias.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_25_self_attn_k_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_25_self_attn_k_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..c3666f60f19857bacf6ea6739791795fe839205f Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_25_self_attn_k_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_25_self_attn_o_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_25_self_attn_o_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..c86839c59145849e367cc3aa33182b1c6c31a614 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_25_self_attn_o_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_25_self_attn_o_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_25_self_attn_o_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..7105e5612d64351f2c4bdc50f4c955b39e8a2599 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_25_self_attn_o_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_25_self_attn_v_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_25_self_attn_v_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..6614333f02907581ddb1b8b42cf15f89cb8cb6d4 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_25_self_attn_v_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_25_self_attn_v_proj_weight.signs b/deepseek-r1-1.5b-packed/model_layers_25_self_attn_v_proj_weight.signs new file mode 100644 index 0000000000000000000000000000000000000000..9a854b00a5400d5b8b93a75e67a156be8ba5b0d0 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_25_self_attn_v_proj_weight.signs differ diff --git a/deepseek-r1-1.5b-packed/model_layers_26_input_layernorm_weight.fp16 b/deepseek-r1-1.5b-packed/model_layers_26_input_layernorm_weight.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..1f225c82377611b1ba728ba0580192d4687233a2 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_26_input_layernorm_weight.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_26_mlp_down_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_26_mlp_down_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..c86839c59145849e367cc3aa33182b1c6c31a614 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_26_mlp_down_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_26_mlp_gate_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_26_mlp_gate_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..3d5a76a5a0c43488828c17fc8a7002591a508098 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_26_mlp_gate_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_26_mlp_gate_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_26_mlp_gate_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..44da1bc46cab751b898388bbe86d08514559e65a Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_26_mlp_gate_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_26_mlp_up_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_26_mlp_up_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..3d5a76a5a0c43488828c17fc8a7002591a508098 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_26_mlp_up_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_26_post_attention_layernorm_weight.fp16 b/deepseek-r1-1.5b-packed/model_layers_26_post_attention_layernorm_weight.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..1d6c3797debd7e7c53e35120bf8d81a3c7a13b3c Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_26_post_attention_layernorm_weight.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_26_self_attn_o_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_26_self_attn_o_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..c86839c59145849e367cc3aa33182b1c6c31a614 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_26_self_attn_o_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_26_self_attn_o_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_26_self_attn_o_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..3e073791b63761a037e589e3955c54341519e2e7 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_26_self_attn_o_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_26_self_attn_q_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_26_self_attn_q_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..c86839c59145849e367cc3aa33182b1c6c31a614 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_26_self_attn_q_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_26_self_attn_q_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_26_self_attn_q_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..a527543388532f3f778677cf358fb040dd74746d Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_26_self_attn_q_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_26_self_attn_v_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_26_self_attn_v_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..86483a47a43f088890dd31b931c20dd56750e90a Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_26_self_attn_v_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_26_self_attn_v_proj_weight.signs b/deepseek-r1-1.5b-packed/model_layers_26_self_attn_v_proj_weight.signs new file mode 100644 index 0000000000000000000000000000000000000000..02ff97b80f4eb7e7bde8dc2327400e625efb6c54 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_26_self_attn_v_proj_weight.signs differ diff --git a/deepseek-r1-1.5b-packed/model_layers_27_mlp_up_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_27_mlp_up_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..14be0eaed89d3ea7bb89253ff00c27683da5e0ca Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_27_mlp_up_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_27_self_attn_k_proj_bias.fp16 b/deepseek-r1-1.5b-packed/model_layers_27_self_attn_k_proj_bias.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..c42780a8929aea89cf3194ab014b020fb6af6edc Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_27_self_attn_k_proj_bias.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_27_self_attn_q_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_27_self_attn_q_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..0174d5fe13781607439d66a84f0b0803604145bc Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_27_self_attn_q_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_2_mlp_down_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_2_mlp_down_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..c86839c59145849e367cc3aa33182b1c6c31a614 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_2_mlp_down_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_2_mlp_down_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_2_mlp_down_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..01b60a0c8cdae52a966f60ee277d2c0656d07aa1 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_2_mlp_down_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_2_mlp_gate_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_2_mlp_gate_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..3d5a76a5a0c43488828c17fc8a7002591a508098 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_2_mlp_gate_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_2_mlp_gate_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_2_mlp_gate_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..93a5afb65d92f8cef46b8f9d58480aa95975505a Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_2_mlp_gate_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_2_mlp_up_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_2_mlp_up_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..3d5a76a5a0c43488828c17fc8a7002591a508098 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_2_mlp_up_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_2_mlp_up_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_2_mlp_up_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..ec81776074a3566080e73fbe426facefe3aba4c1 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_2_mlp_up_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_2_post_attention_layernorm_weight.fp16 b/deepseek-r1-1.5b-packed/model_layers_2_post_attention_layernorm_weight.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..2748fa2d55b1db412a89fe2d6dd1b34e4b2a9908 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_2_post_attention_layernorm_weight.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_2_self_attn_k_proj_bias.fp16 b/deepseek-r1-1.5b-packed/model_layers_2_self_attn_k_proj_bias.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..1d2d51e9f4bac8f565b8f7b0c4826a95c5c9cc42 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_2_self_attn_k_proj_bias.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_2_self_attn_k_proj_weight.signs b/deepseek-r1-1.5b-packed/model_layers_2_self_attn_k_proj_weight.signs new file mode 100644 index 0000000000000000000000000000000000000000..ca6a2ba88a59c60ec72ccda164705ea6049e2fa1 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_2_self_attn_k_proj_weight.signs differ diff --git a/deepseek-r1-1.5b-packed/model_layers_2_self_attn_v_proj_bias.fp16 b/deepseek-r1-1.5b-packed/model_layers_2_self_attn_v_proj_bias.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..2380f667f41f252cbabe02de56949dd09681da1b Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_2_self_attn_v_proj_bias.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_2_self_attn_v_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_2_self_attn_v_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..9401fc7bec41306420e2f625efdc88cd8c92717d --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_2_self_attn_v_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_2_self_attn_v_proj_weight.signs b/deepseek-r1-1.5b-packed/model_layers_2_self_attn_v_proj_weight.signs new file mode 100644 index 0000000000000000000000000000000000000000..54e10b0a1996c9562c5bca8bc222428cc18cb1b3 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_2_self_attn_v_proj_weight.signs differ diff --git a/deepseek-r1-1.5b-packed/model_layers_3_input_layernorm_weight.fp16 b/deepseek-r1-1.5b-packed/model_layers_3_input_layernorm_weight.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..82fe0d09b0a7eee7296ce5aa2c244e96aff41d85 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_3_input_layernorm_weight.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_3_mlp_gate_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_3_mlp_gate_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..3d5a76a5a0c43488828c17fc8a7002591a508098 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_3_mlp_gate_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_3_mlp_up_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_3_mlp_up_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..3d5a76a5a0c43488828c17fc8a7002591a508098 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_3_mlp_up_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_3_mlp_up_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_3_mlp_up_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..a4830b593bc2eaf74064bd8fcbaaa0f40648fe00 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_3_mlp_up_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_3_self_attn_k_proj_weight.signs b/deepseek-r1-1.5b-packed/model_layers_3_self_attn_k_proj_weight.signs new file mode 100644 index 0000000000000000000000000000000000000000..7498db3ff36f39cffccb804a1920907a5c97f787 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_3_self_attn_k_proj_weight.signs differ diff --git a/deepseek-r1-1.5b-packed/model_layers_3_self_attn_o_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_3_self_attn_o_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..c86839c59145849e367cc3aa33182b1c6c31a614 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_3_self_attn_o_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_3_self_attn_v_proj_bias.fp16 b/deepseek-r1-1.5b-packed/model_layers_3_self_attn_v_proj_bias.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..3c6ecacaa6eaa01ba4642cb7f341b9c6c08fc85c Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_3_self_attn_v_proj_bias.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_3_self_attn_v_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_3_self_attn_v_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..9401fc7bec41306420e2f625efdc88cd8c92717d --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_3_self_attn_v_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_4_mlp_down_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_4_mlp_down_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..3e8a8c4c273ce6a1bd7b3fff820d11e0ac28a308 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_4_mlp_down_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_4_mlp_gate_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_4_mlp_gate_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..3d5a76a5a0c43488828c17fc8a7002591a508098 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_4_mlp_gate_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_4_post_attention_layernorm_weight.fp16 b/deepseek-r1-1.5b-packed/model_layers_4_post_attention_layernorm_weight.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..3a03ea4fd346e87a8d6485b675c1b00490291f03 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_4_post_attention_layernorm_weight.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_4_self_attn_k_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_4_self_attn_k_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..9401fc7bec41306420e2f625efdc88cd8c92717d --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_4_self_attn_k_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_4_self_attn_k_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_4_self_attn_k_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..3e09b97395834348a0193e4deaa6d29f8e5c68df Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_4_self_attn_k_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_4_self_attn_q_proj_bias.fp16 b/deepseek-r1-1.5b-packed/model_layers_4_self_attn_q_proj_bias.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..7c97d9dd0ea718ded20befe3f46a8222a3a4ae71 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_4_self_attn_q_proj_bias.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_4_self_attn_q_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_4_self_attn_q_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..63db18382079a1832e3675f0da4e8453ee298ec5 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_4_self_attn_q_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_4_self_attn_v_proj_bias.fp16 b/deepseek-r1-1.5b-packed/model_layers_4_self_attn_v_proj_bias.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..573b5eb9d935609dde2097a2217641147188b605 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_4_self_attn_v_proj_bias.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_4_self_attn_v_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_4_self_attn_v_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..9401fc7bec41306420e2f625efdc88cd8c92717d --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_4_self_attn_v_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_5_input_layernorm_weight.fp16 b/deepseek-r1-1.5b-packed/model_layers_5_input_layernorm_weight.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..4887b4f8f91b6ca182df77fad852de7d66027b7d Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_5_input_layernorm_weight.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_5_mlp_gate_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_5_mlp_gate_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..3d5a76a5a0c43488828c17fc8a7002591a508098 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_5_mlp_gate_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_5_self_attn_q_proj_bias.fp16 b/deepseek-r1-1.5b-packed/model_layers_5_self_attn_q_proj_bias.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..ea90dc754475dc1da21f985c9d0b3164ae44fb92 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_5_self_attn_q_proj_bias.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_5_self_attn_q_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_5_self_attn_q_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..683e745b5d5a15ba617063b7280581ecdfb056e0 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_5_self_attn_q_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_5_self_attn_v_proj_bias.fp16 b/deepseek-r1-1.5b-packed/model_layers_5_self_attn_v_proj_bias.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..3e83b22c90d1a1d8d811f6d3775f594a5413fa17 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_5_self_attn_v_proj_bias.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_5_self_attn_v_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_5_self_attn_v_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..9401fc7bec41306420e2f625efdc88cd8c92717d --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_5_self_attn_v_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_6_input_layernorm_weight.fp16 b/deepseek-r1-1.5b-packed/model_layers_6_input_layernorm_weight.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..97b519b14daa47092d6ecef62c3bfdda0e741084 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_6_input_layernorm_weight.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_6_mlp_gate_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_6_mlp_gate_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..3ddee85e77c2643430a89061d5f67c7f6961e99d Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_6_mlp_gate_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_6_mlp_up_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_6_mlp_up_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..3d5a76a5a0c43488828c17fc8a7002591a508098 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_6_mlp_up_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_6_post_attention_layernorm_weight.fp16 b/deepseek-r1-1.5b-packed/model_layers_6_post_attention_layernorm_weight.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..f6158498a312a5b3a2e8841c68ab65ff5807e776 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_6_post_attention_layernorm_weight.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_6_self_attn_k_proj_bias.fp16 b/deepseek-r1-1.5b-packed/model_layers_6_self_attn_k_proj_bias.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..d593845bbb02448eac0e08ca8c4102e5478ac7d2 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_6_self_attn_k_proj_bias.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_6_self_attn_k_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_6_self_attn_k_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..ee3fd3311d36be7c49088853b9b5d82fca093624 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_6_self_attn_k_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_6_self_attn_k_proj_weight.signs b/deepseek-r1-1.5b-packed/model_layers_6_self_attn_k_proj_weight.signs new file mode 100644 index 0000000000000000000000000000000000000000..44a1c67cb2d8815f8522ac98aab19745423db987 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_6_self_attn_k_proj_weight.signs differ diff --git a/deepseek-r1-1.5b-packed/model_layers_6_self_attn_o_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_6_self_attn_o_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..c86839c59145849e367cc3aa33182b1c6c31a614 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_6_self_attn_o_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_6_self_attn_q_proj_bias.fp16 b/deepseek-r1-1.5b-packed/model_layers_6_self_attn_q_proj_bias.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..c26f42a44f8db2152457881a9bb9c48d3032ded5 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_6_self_attn_q_proj_bias.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_6_self_attn_q_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_6_self_attn_q_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..c86839c59145849e367cc3aa33182b1c6c31a614 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_6_self_attn_q_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_6_self_attn_v_proj_bias.fp16 b/deepseek-r1-1.5b-packed/model_layers_6_self_attn_v_proj_bias.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..db6c4087a62ab997ea93da9955d9737ac2c25eb4 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_6_self_attn_v_proj_bias.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_6_self_attn_v_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_6_self_attn_v_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..9401fc7bec41306420e2f625efdc88cd8c92717d --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_6_self_attn_v_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_6_self_attn_v_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_6_self_attn_v_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..0663b92c64508fe667e9ab24104983576fbb6245 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_6_self_attn_v_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_7_mlp_down_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_7_mlp_down_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..c86839c59145849e367cc3aa33182b1c6c31a614 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_7_mlp_down_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_7_self_attn_o_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_7_self_attn_o_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..c86839c59145849e367cc3aa33182b1c6c31a614 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_7_self_attn_o_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_7_self_attn_q_proj_bias.fp16 b/deepseek-r1-1.5b-packed/model_layers_7_self_attn_q_proj_bias.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..576a52e8dd0a1e8d80a1cf150cbb1e8cdce0e7e1 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_7_self_attn_q_proj_bias.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_7_self_attn_v_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_7_self_attn_v_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..9401fc7bec41306420e2f625efdc88cd8c92717d --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_7_self_attn_v_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_7_self_attn_v_proj_weight.signs b/deepseek-r1-1.5b-packed/model_layers_7_self_attn_v_proj_weight.signs new file mode 100644 index 0000000000000000000000000000000000000000..0635d7eeec596e40e47fb5f57ed52995e24b7690 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_7_self_attn_v_proj_weight.signs differ diff --git a/deepseek-r1-1.5b-packed/model_layers_8_input_layernorm_weight.fp16 b/deepseek-r1-1.5b-packed/model_layers_8_input_layernorm_weight.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..aff9d98b0be61b53898d1d9bbe1ed129e788251f Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_8_input_layernorm_weight.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_8_mlp_down_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_8_mlp_down_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..c86839c59145849e367cc3aa33182b1c6c31a614 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_8_mlp_down_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_8_mlp_down_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_8_mlp_down_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..5c498ff4ca6a59492cabf734dd8250a8e22d6a33 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_8_mlp_down_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_8_mlp_gate_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_8_mlp_gate_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..3d5a76a5a0c43488828c17fc8a7002591a508098 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_8_mlp_gate_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_8_mlp_gate_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_8_mlp_gate_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..c4c786705f6219890d7c019c0e9bbb2ac97a64a0 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_8_mlp_gate_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_8_mlp_up_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_8_mlp_up_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..3d5a76a5a0c43488828c17fc8a7002591a508098 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_8_mlp_up_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_8_self_attn_k_proj_weight.signs b/deepseek-r1-1.5b-packed/model_layers_8_self_attn_k_proj_weight.signs new file mode 100644 index 0000000000000000000000000000000000000000..ee7d9d76fb8cd89d756b6e6e59276f5a4c3f77b1 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_8_self_attn_k_proj_weight.signs differ diff --git a/deepseek-r1-1.5b-packed/model_layers_8_self_attn_o_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_8_self_attn_o_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..c86839c59145849e367cc3aa33182b1c6c31a614 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_8_self_attn_o_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_8_self_attn_q_proj_bias.fp16 b/deepseek-r1-1.5b-packed/model_layers_8_self_attn_q_proj_bias.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..773d9f4cd8132578aea15925ad5cf523650e2337 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_8_self_attn_q_proj_bias.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_8_self_attn_q_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_8_self_attn_q_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..c86839c59145849e367cc3aa33182b1c6c31a614 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_8_self_attn_q_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_8_self_attn_v_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_8_self_attn_v_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..2d2375b02b4d72ba0e2820020c68cc4807172125 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_8_self_attn_v_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_8_self_attn_v_proj_weight.signs b/deepseek-r1-1.5b-packed/model_layers_8_self_attn_v_proj_weight.signs new file mode 100644 index 0000000000000000000000000000000000000000..66d95813be7a1cc30ac2e2ea3bef606235a0a5c9 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_8_self_attn_v_proj_weight.signs differ diff --git a/deepseek-r1-1.5b-packed/model_layers_9_input_layernorm_weight.fp16 b/deepseek-r1-1.5b-packed/model_layers_9_input_layernorm_weight.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..34b2ea7c153d35d9b3d7bc9bc90d4025b8974cc8 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_9_input_layernorm_weight.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_9_mlp_down_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_9_mlp_down_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..c86839c59145849e367cc3aa33182b1c6c31a614 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_9_mlp_down_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_9_mlp_down_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_9_mlp_down_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..9379a0500c0a3c4e4a2287c6a0df113604371a83 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_9_mlp_down_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_9_mlp_gate_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_9_mlp_gate_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..3d5a76a5a0c43488828c17fc8a7002591a508098 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_9_mlp_gate_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_9_mlp_up_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_9_mlp_up_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..689f2dbbc36e3c2aa03dc89759a6ca22cd4daeaa Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_9_mlp_up_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_9_post_attention_layernorm_weight.fp16 b/deepseek-r1-1.5b-packed/model_layers_9_post_attention_layernorm_weight.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..279793aef4972875c25744bf4a9569e950ee3302 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_9_post_attention_layernorm_weight.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_9_self_attn_k_proj_bias.fp16 b/deepseek-r1-1.5b-packed/model_layers_9_self_attn_k_proj_bias.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..93a3385ad61f52e5a158453353f0a1cd99521d34 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_9_self_attn_k_proj_bias.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_9_self_attn_k_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_9_self_attn_k_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..21eed124b45f972b2e2e888945a414b2c5debcc7 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_9_self_attn_k_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_9_self_attn_k_proj_weight.signs b/deepseek-r1-1.5b-packed/model_layers_9_self_attn_k_proj_weight.signs new file mode 100644 index 0000000000000000000000000000000000000000..8898f356516b2954516ea9717f2e9df3501d32a0 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_9_self_attn_k_proj_weight.signs differ diff --git a/deepseek-r1-1.5b-packed/model_layers_9_self_attn_o_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_9_self_attn_o_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..c86839c59145849e367cc3aa33182b1c6c31a614 --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_9_self_attn_o_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_9_self_attn_q_proj_weight.scales b/deepseek-r1-1.5b-packed/model_layers_9_self_attn_q_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..07d105fd9cdaeca8ae76cc48b99ca86a4e2d050b Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_9_self_attn_q_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-packed/model_layers_9_self_attn_v_proj_bias.fp16 b/deepseek-r1-1.5b-packed/model_layers_9_self_attn_v_proj_bias.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..98724715060dfcd8396c8f2195785367b0067b39 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_9_self_attn_v_proj_bias.fp16 differ diff --git a/deepseek-r1-1.5b-packed/model_layers_9_self_attn_v_proj_weight.rmm b/deepseek-r1-1.5b-packed/model_layers_9_self_attn_v_proj_weight.rmm new file mode 100644 index 0000000000000000000000000000000000000000..9401fc7bec41306420e2f625efdc88cd8c92717d --- /dev/null +++ b/deepseek-r1-1.5b-packed/model_layers_9_self_attn_v_proj_weight.rmm @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/deepseek-r1-1.5b-packed/model_layers_9_self_attn_v_proj_weight.signs b/deepseek-r1-1.5b-packed/model_layers_9_self_attn_v_proj_weight.signs new file mode 100644 index 0000000000000000000000000000000000000000..3b74c1455affc1fd0283a843a8ffe7cf073ba8a6 Binary files /dev/null and b/deepseek-r1-1.5b-packed/model_layers_9_self_attn_v_proj_weight.signs differ