diff --git a/.gitattributes b/.gitattributes index 2c408d11762cbfd478f57b7d724fa58e7bfc9404..0db223cacbeaff38525c359f6282428fce80ed2c 100644 --- a/.gitattributes +++ b/.gitattributes @@ -2890,3 +2890,45 @@ qwen3-4b-log-unary/model_layers_27_self_attn_k_proj_weight.sign filter=lfs diff= qwen3-4b-log-unary/model_layers_2_self_attn_q_proj_weight.planes filter=lfs diff=lfs merge=lfs -text qwen3-4b-log-unary/model_layers_10_mlp_gate_proj_weight.planes filter=lfs diff=lfs merge=lfs -text qwen3-4b-log-unary/model_layers_31_self_attn_q_proj_weight.sign filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-unary4/model_layers_11_mlp_down_proj_weight.sign filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-unary4/model_layers_4_mlp_gate_proj_weight.sign filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-unary4/model_layers_18_self_attn_q_proj_weight.sign filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-unary4/model_layers_10_mlp_down_proj_weight.planes filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-unary4/model_layers_26_self_attn_o_proj_weight.sign filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-unary4/model_layers_6_self_attn_q_proj_weight.planes filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-unary4/model_layers_18_self_attn_o_proj_weight.sign filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-unary4/model_layers_27_self_attn_k_proj_weight.planes filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-unary4/model_layers_27_mlp_down_proj_weight.sign filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-unary4/model_layers_27_mlp_up_proj_weight.planes filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-unary4/model_layers_18_self_attn_k_proj_weight.planes filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-unary4/model_layers_19_self_attn_v_proj_weight.planes filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-unary4/model_layers_22_mlp_gate_proj_weight.sign filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-unary4/model_layers_17_mlp_up_proj_weight.planes filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-unary4/model_layers_11_self_attn_q_proj_weight.planes filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-unary4/model_layers_7_self_attn_o_proj_weight.sign filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-unary4/model_layers_3_self_attn_o_proj_weight.sign filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-unary4/model_layers_17_self_attn_o_proj_weight.planes filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-unary4/model_layers_8_mlp_up_proj_weight.sign filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-unary4/model_layers_25_self_attn_v_proj_weight.planes filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-unary4/model_layers_21_mlp_up_proj_weight.sign filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-unary4/model_layers_27_self_attn_q_proj_weight.sign filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-unary4/model_layers_0_mlp_gate_proj_weight.sign filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-unary4/model_layers_22_mlp_gate_proj_weight.planes filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-unary4/model_layers_25_mlp_gate_proj_weight.planes filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-unary4/model_layers_5_self_attn_v_proj_weight.planes filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-unary4/model_layers_2_mlp_up_proj_weight.planes filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-unary4/model_layers_27_mlp_up_proj_weight.sign filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-unary4/model_layers_3_mlp_gate_proj_weight.planes filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-unary4/model_layers_11_self_attn_v_proj_weight.planes filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-unary4/model_layers_24_self_attn_q_proj_weight.sign filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-unary4/model_layers_23_mlp_up_proj_weight.sign filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-unary4/model_layers_5_self_attn_q_proj_weight.sign filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-unary4/model_layers_14_mlp_up_proj_weight.sign filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-unary4/model_layers_12_self_attn_o_proj_weight.sign filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-unary4/model_layers_21_mlp_up_proj_weight.planes filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-unary4/model_layers_2_mlp_up_proj_weight.sign filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-unary4/model_layers_1_self_attn_o_proj_weight.planes filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-unary4/model_layers_23_mlp_down_proj_weight.planes filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-unary4/model_layers_25_mlp_up_proj_weight.sign filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-unary4/model_layers_2_self_attn_q_proj_weight.sign filter=lfs diff=lfs merge=lfs -text +deepseek-r1-1.5b-unary4/model_layers_5_mlp_down_proj_weight.planes filter=lfs diff=lfs merge=lfs -text diff --git a/deepseek-r1-1.5b-unary4/model_layers_0_mlp_gate_proj_weight.scales b/deepseek-r1-1.5b-unary4/model_layers_0_mlp_gate_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..b5b51a2942bfd7d3ee05e87cde2d73b472e148ec Binary files /dev/null and b/deepseek-r1-1.5b-unary4/model_layers_0_mlp_gate_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_0_mlp_gate_proj_weight.sign b/deepseek-r1-1.5b-unary4/model_layers_0_mlp_gate_proj_weight.sign new file mode 100644 index 0000000000000000000000000000000000000000..97c721e54a951aa8031e3d0665a92a9b47cba088 --- /dev/null +++ b/deepseek-r1-1.5b-unary4/model_layers_0_mlp_gate_proj_weight.sign @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2cf601ab6062858b266017e3ccc33f77bb1a6560510ddb28d0bddd9822adbd0b +size 1720320 diff --git a/deepseek-r1-1.5b-unary4/model_layers_0_self_attn_v_proj_weight.scales b/deepseek-r1-1.5b-unary4/model_layers_0_self_attn_v_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..e56993c13fad8760dc2081bcebee9ec160dff702 Binary files /dev/null and b/deepseek-r1-1.5b-unary4/model_layers_0_self_attn_v_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_10_mlp_down_proj_weight.planes b/deepseek-r1-1.5b-unary4/model_layers_10_mlp_down_proj_weight.planes new file mode 100644 index 0000000000000000000000000000000000000000..107a4d52da7ddd054b58a65c551893b947386e55 --- /dev/null +++ b/deepseek-r1-1.5b-unary4/model_layers_10_mlp_down_proj_weight.planes @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f005c0c74e5f79d926d9db3cb02f312d03559c820b03e4d45eb1fc381f2d303f +size 6881280 diff --git a/deepseek-r1-1.5b-unary4/model_layers_10_post_attention_layernorm_weight.fp16 b/deepseek-r1-1.5b-unary4/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-unary4/model_layers_10_post_attention_layernorm_weight.fp16 differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_10_self_attn_k_proj_weight.sign b/deepseek-r1-1.5b-unary4/model_layers_10_self_attn_k_proj_weight.sign new file mode 100644 index 0000000000000000000000000000000000000000..9638a0f8929beeff726db96213daf5c337439cca Binary files /dev/null and b/deepseek-r1-1.5b-unary4/model_layers_10_self_attn_k_proj_weight.sign differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_10_self_attn_v_proj_weight.sign b/deepseek-r1-1.5b-unary4/model_layers_10_self_attn_v_proj_weight.sign new file mode 100644 index 0000000000000000000000000000000000000000..9ac95267277f48e561a238e9915ff437ecb55070 Binary files /dev/null and b/deepseek-r1-1.5b-unary4/model_layers_10_self_attn_v_proj_weight.sign differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_11_mlp_down_proj_weight.sign b/deepseek-r1-1.5b-unary4/model_layers_11_mlp_down_proj_weight.sign new file mode 100644 index 0000000000000000000000000000000000000000..9944ab22f7fe39cad155cb35656fa9ba641b8686 --- /dev/null +++ b/deepseek-r1-1.5b-unary4/model_layers_11_mlp_down_proj_weight.sign @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b07cb8f37f8c3fb5f800b008ee48cc27f089db360aa642d7498bc09b8314e21c +size 1720320 diff --git a/deepseek-r1-1.5b-unary4/model_layers_11_mlp_up_proj_weight.scales b/deepseek-r1-1.5b-unary4/model_layers_11_mlp_up_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..85712e2b4404290592a4c26e1d01b62c04d3da26 Binary files /dev/null and b/deepseek-r1-1.5b-unary4/model_layers_11_mlp_up_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_11_self_attn_k_proj_weight.scales b/deepseek-r1-1.5b-unary4/model_layers_11_self_attn_k_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..7f524584cac057e6762045b503334667baf6bc9c Binary files /dev/null and b/deepseek-r1-1.5b-unary4/model_layers_11_self_attn_k_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_11_self_attn_q_proj_weight.planes b/deepseek-r1-1.5b-unary4/model_layers_11_self_attn_q_proj_weight.planes new file mode 100644 index 0000000000000000000000000000000000000000..c8970fc1b96c69c3391a3472955acaa6df1c8015 --- /dev/null +++ b/deepseek-r1-1.5b-unary4/model_layers_11_self_attn_q_proj_weight.planes @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:391743aaf15f6307bc9b667a575007cea922b964c44b55499f683657c675cf1b +size 1179648 diff --git a/deepseek-r1-1.5b-unary4/model_layers_11_self_attn_q_proj_weight.scales b/deepseek-r1-1.5b-unary4/model_layers_11_self_attn_q_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..1bad53c7adc50e60eb62ce948ac86dd59f7e6341 Binary files /dev/null and b/deepseek-r1-1.5b-unary4/model_layers_11_self_attn_q_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_11_self_attn_v_proj_weight.planes b/deepseek-r1-1.5b-unary4/model_layers_11_self_attn_v_proj_weight.planes new file mode 100644 index 0000000000000000000000000000000000000000..c1318e540021b917fbbca97883ca35fd18d048ed --- /dev/null +++ b/deepseek-r1-1.5b-unary4/model_layers_11_self_attn_v_proj_weight.planes @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e61fcc0005e9bc64d8257306654eacd8cf2a8b302e1b4c68c707f52b894c770f +size 196608 diff --git a/deepseek-r1-1.5b-unary4/model_layers_12_self_attn_k_proj_bias.fp16 b/deepseek-r1-1.5b-unary4/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-unary4/model_layers_12_self_attn_k_proj_bias.fp16 differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_12_self_attn_o_proj_weight.sign b/deepseek-r1-1.5b-unary4/model_layers_12_self_attn_o_proj_weight.sign new file mode 100644 index 0000000000000000000000000000000000000000..718d38385a80fdb17109749d972d2896fa839c64 --- /dev/null +++ b/deepseek-r1-1.5b-unary4/model_layers_12_self_attn_o_proj_weight.sign @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b99c13197a9fd4d1233f95ba7aeaff88e06a93d91c95bb771fda4f56567a3cf2 +size 294912 diff --git a/deepseek-r1-1.5b-unary4/model_layers_12_self_attn_v_proj_weight.scales b/deepseek-r1-1.5b-unary4/model_layers_12_self_attn_v_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..f94b183d16072b019cde2b5369809b58a22b450a Binary files /dev/null and b/deepseek-r1-1.5b-unary4/model_layers_12_self_attn_v_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_13_mlp_gate_proj_weight.scales b/deepseek-r1-1.5b-unary4/model_layers_13_mlp_gate_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..4f5d32cc960aec9f9bf4c339a75b482a3f7808ef Binary files /dev/null and b/deepseek-r1-1.5b-unary4/model_layers_13_mlp_gate_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_14_mlp_up_proj_weight.sign b/deepseek-r1-1.5b-unary4/model_layers_14_mlp_up_proj_weight.sign new file mode 100644 index 0000000000000000000000000000000000000000..20aa1bd99aa57e1318906c52120c8c7107d0cff5 --- /dev/null +++ b/deepseek-r1-1.5b-unary4/model_layers_14_mlp_up_proj_weight.sign @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:71036189bc6c7101b6bebd0447e5613b8373df5f4687f35de2606e06905c73e1 +size 1720320 diff --git a/deepseek-r1-1.5b-unary4/model_layers_15_self_attn_k_proj_bias.fp16 b/deepseek-r1-1.5b-unary4/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-unary4/model_layers_15_self_attn_k_proj_bias.fp16 differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_15_self_attn_k_proj_weight.scales b/deepseek-r1-1.5b-unary4/model_layers_15_self_attn_k_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..767bc2f1f8117488285cafff64119f31a1878513 Binary files /dev/null and b/deepseek-r1-1.5b-unary4/model_layers_15_self_attn_k_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_16_input_layernorm_weight.fp16 b/deepseek-r1-1.5b-unary4/model_layers_16_input_layernorm_weight.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..be30eeedbe2adf190e199c0a489983b5fccd2add Binary files /dev/null and b/deepseek-r1-1.5b-unary4/model_layers_16_input_layernorm_weight.fp16 differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_16_mlp_up_proj_weight.scales b/deepseek-r1-1.5b-unary4/model_layers_16_mlp_up_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..f61f38062d9fdd6cb54a871827b7ee0e7c83cef8 Binary files /dev/null and b/deepseek-r1-1.5b-unary4/model_layers_16_mlp_up_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_16_post_attention_layernorm_weight.fp16 b/deepseek-r1-1.5b-unary4/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-unary4/model_layers_16_post_attention_layernorm_weight.fp16 differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_16_self_attn_k_proj_bias.fp16 b/deepseek-r1-1.5b-unary4/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-unary4/model_layers_16_self_attn_k_proj_bias.fp16 differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_16_self_attn_v_proj_weight.scales b/deepseek-r1-1.5b-unary4/model_layers_16_self_attn_v_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..07a4490b3a8f561d26e4ac2936da4768d028d226 Binary files /dev/null and b/deepseek-r1-1.5b-unary4/model_layers_16_self_attn_v_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_16_self_attn_v_proj_weight.sign b/deepseek-r1-1.5b-unary4/model_layers_16_self_attn_v_proj_weight.sign new file mode 100644 index 0000000000000000000000000000000000000000..3c08a825c8a87f0a8cb998fff9849a8eff130c65 Binary files /dev/null and b/deepseek-r1-1.5b-unary4/model_layers_16_self_attn_v_proj_weight.sign differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_17_mlp_up_proj_weight.planes b/deepseek-r1-1.5b-unary4/model_layers_17_mlp_up_proj_weight.planes new file mode 100644 index 0000000000000000000000000000000000000000..4497773aae6faab0ffd899da9bf0bbab78ccb2bd --- /dev/null +++ b/deepseek-r1-1.5b-unary4/model_layers_17_mlp_up_proj_weight.planes @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1a2696679b07fd134e164866f5fc097081e6905a936b75b9be2c75e60a811954 +size 6881280 diff --git a/deepseek-r1-1.5b-unary4/model_layers_17_self_attn_o_proj_weight.planes b/deepseek-r1-1.5b-unary4/model_layers_17_self_attn_o_proj_weight.planes new file mode 100644 index 0000000000000000000000000000000000000000..2152b0ac8fc4b2d59c93faa31e6ef2091f060754 --- /dev/null +++ b/deepseek-r1-1.5b-unary4/model_layers_17_self_attn_o_proj_weight.planes @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b2e8be702da7bd0675f0f830d4b54f5bbb43f884a6f95dc438791e6c91be091e +size 1179648 diff --git a/deepseek-r1-1.5b-unary4/model_layers_17_self_attn_o_proj_weight.scales b/deepseek-r1-1.5b-unary4/model_layers_17_self_attn_o_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..4814f6fa163bc07fb9bb6a2987270f0cff1bdfb7 Binary files /dev/null and b/deepseek-r1-1.5b-unary4/model_layers_17_self_attn_o_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_18_self_attn_k_proj_weight.planes b/deepseek-r1-1.5b-unary4/model_layers_18_self_attn_k_proj_weight.planes new file mode 100644 index 0000000000000000000000000000000000000000..f17a4d422201c5bd6c902c36b840268879817252 --- /dev/null +++ b/deepseek-r1-1.5b-unary4/model_layers_18_self_attn_k_proj_weight.planes @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:aa6452b5d4d246665da65f5db002758e0d27b5767dc46064f81d4b3fce597813 +size 196608 diff --git a/deepseek-r1-1.5b-unary4/model_layers_18_self_attn_o_proj_weight.sign b/deepseek-r1-1.5b-unary4/model_layers_18_self_attn_o_proj_weight.sign new file mode 100644 index 0000000000000000000000000000000000000000..3eae6b26572f860e4ea119ec70a0a128cc7a7a56 --- /dev/null +++ b/deepseek-r1-1.5b-unary4/model_layers_18_self_attn_o_proj_weight.sign @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b7809020bf04c489051e22e07494b20557ab1d7b99b6ac74741b55206a6092b4 +size 294912 diff --git a/deepseek-r1-1.5b-unary4/model_layers_18_self_attn_q_proj_bias.fp16 b/deepseek-r1-1.5b-unary4/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-unary4/model_layers_18_self_attn_q_proj_bias.fp16 differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_18_self_attn_q_proj_weight.sign b/deepseek-r1-1.5b-unary4/model_layers_18_self_attn_q_proj_weight.sign new file mode 100644 index 0000000000000000000000000000000000000000..5219ad8b69d3627a68430808591c64a618927dbc --- /dev/null +++ b/deepseek-r1-1.5b-unary4/model_layers_18_self_attn_q_proj_weight.sign @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0cc73697f7d2c00e8a7ba5c1d7a5006479360a95f8a8860700b53aa831558a47 +size 294912 diff --git a/deepseek-r1-1.5b-unary4/model_layers_19_input_layernorm_weight.fp16 b/deepseek-r1-1.5b-unary4/model_layers_19_input_layernorm_weight.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..b375bb4e6b6bc0bdcaa5291001295849dd624035 Binary files /dev/null and b/deepseek-r1-1.5b-unary4/model_layers_19_input_layernorm_weight.fp16 differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_19_self_attn_o_proj_weight.scales b/deepseek-r1-1.5b-unary4/model_layers_19_self_attn_o_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..9bdc755ef38012b06af8da33228e99a0e420670e Binary files /dev/null and b/deepseek-r1-1.5b-unary4/model_layers_19_self_attn_o_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_19_self_attn_v_proj_weight.planes b/deepseek-r1-1.5b-unary4/model_layers_19_self_attn_v_proj_weight.planes new file mode 100644 index 0000000000000000000000000000000000000000..91a3e7032c102c2a6ff1092cc8841bda365f0ddc --- /dev/null +++ b/deepseek-r1-1.5b-unary4/model_layers_19_self_attn_v_proj_weight.planes @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e6ad6327989a0def858d1449b0926b9ccdc0c3144690ab365e747325ea374779 +size 196608 diff --git a/deepseek-r1-1.5b-unary4/model_layers_1_self_attn_o_proj_weight.planes b/deepseek-r1-1.5b-unary4/model_layers_1_self_attn_o_proj_weight.planes new file mode 100644 index 0000000000000000000000000000000000000000..83e6cb2464364f4ae735c978f31fc7df9f4ed47f --- /dev/null +++ b/deepseek-r1-1.5b-unary4/model_layers_1_self_attn_o_proj_weight.planes @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:42d7c59df70a966def7b8582254155515b7e5c85f6b4a9036fcca78e3e5dca48 +size 1179648 diff --git a/deepseek-r1-1.5b-unary4/model_layers_1_self_attn_o_proj_weight.scales b/deepseek-r1-1.5b-unary4/model_layers_1_self_attn_o_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..01c18dcfcd11a61ef56a535684c93a68989e7759 Binary files /dev/null and b/deepseek-r1-1.5b-unary4/model_layers_1_self_attn_o_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_1_self_attn_v_proj_weight.scales b/deepseek-r1-1.5b-unary4/model_layers_1_self_attn_v_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..81852a4043475af7db20678832237a0f3f4a4b28 Binary files /dev/null and b/deepseek-r1-1.5b-unary4/model_layers_1_self_attn_v_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_20_self_attn_q_proj_bias.fp16 b/deepseek-r1-1.5b-unary4/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-unary4/model_layers_20_self_attn_q_proj_bias.fp16 differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_21_mlp_up_proj_weight.planes b/deepseek-r1-1.5b-unary4/model_layers_21_mlp_up_proj_weight.planes new file mode 100644 index 0000000000000000000000000000000000000000..c7fd6483f2314f6aa84393de51814b3729670248 --- /dev/null +++ b/deepseek-r1-1.5b-unary4/model_layers_21_mlp_up_proj_weight.planes @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3a0bba536126dbf01e52d8c8d2f5e1344e021358015f8d38aa27ce08a632112b +size 6881280 diff --git a/deepseek-r1-1.5b-unary4/model_layers_21_mlp_up_proj_weight.sign b/deepseek-r1-1.5b-unary4/model_layers_21_mlp_up_proj_weight.sign new file mode 100644 index 0000000000000000000000000000000000000000..baf9221961a7393153b94142283d105e5df2c498 --- /dev/null +++ b/deepseek-r1-1.5b-unary4/model_layers_21_mlp_up_proj_weight.sign @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:154c898127685a5d1f62f4d132b2ddb359cf897cf0567784b03eff0baa5c31fb +size 1720320 diff --git a/deepseek-r1-1.5b-unary4/model_layers_21_self_attn_k_proj_weight.sign b/deepseek-r1-1.5b-unary4/model_layers_21_self_attn_k_proj_weight.sign new file mode 100644 index 0000000000000000000000000000000000000000..f6ac4e8f15fbb1b4a9f794eb068543df1f32a7e8 Binary files /dev/null and b/deepseek-r1-1.5b-unary4/model_layers_21_self_attn_k_proj_weight.sign differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_21_self_attn_v_proj_weight.scales b/deepseek-r1-1.5b-unary4/model_layers_21_self_attn_v_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..d7066f1ebba1e1d4a669fb6ff9f1a7119a39f9a4 Binary files /dev/null and b/deepseek-r1-1.5b-unary4/model_layers_21_self_attn_v_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_22_mlp_gate_proj_weight.planes b/deepseek-r1-1.5b-unary4/model_layers_22_mlp_gate_proj_weight.planes new file mode 100644 index 0000000000000000000000000000000000000000..5cb284405f31535df4141cc9df70dd7d1a06d8df --- /dev/null +++ b/deepseek-r1-1.5b-unary4/model_layers_22_mlp_gate_proj_weight.planes @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1513867e9552859d2ec309485f86a9bc766b34df35f2704a5c0a5182f459b138 +size 6881280 diff --git a/deepseek-r1-1.5b-unary4/model_layers_22_mlp_gate_proj_weight.sign b/deepseek-r1-1.5b-unary4/model_layers_22_mlp_gate_proj_weight.sign new file mode 100644 index 0000000000000000000000000000000000000000..d8334f2f3038dfd5d0cc1b892c53b1545f0ceb96 --- /dev/null +++ b/deepseek-r1-1.5b-unary4/model_layers_22_mlp_gate_proj_weight.sign @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0cead535970fe45255a6aace0b589ee94e9ccc3daf2499e7af6e07d56599e3c3 +size 1720320 diff --git a/deepseek-r1-1.5b-unary4/model_layers_22_mlp_up_proj_weight.scales b/deepseek-r1-1.5b-unary4/model_layers_22_mlp_up_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..78ddfb0f8997f987405aef777abaddf090efc2c0 Binary files /dev/null and b/deepseek-r1-1.5b-unary4/model_layers_22_mlp_up_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_22_self_attn_o_proj_weight.scales b/deepseek-r1-1.5b-unary4/model_layers_22_self_attn_o_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..f3647b41721fa1249ec44254d894fe7befe6fe57 Binary files /dev/null and b/deepseek-r1-1.5b-unary4/model_layers_22_self_attn_o_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_22_self_attn_v_proj_weight.scales b/deepseek-r1-1.5b-unary4/model_layers_22_self_attn_v_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..c3ac52c33f4e635e39c4746691bb49d32feb44c4 Binary files /dev/null and b/deepseek-r1-1.5b-unary4/model_layers_22_self_attn_v_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_23_mlp_down_proj_weight.planes b/deepseek-r1-1.5b-unary4/model_layers_23_mlp_down_proj_weight.planes new file mode 100644 index 0000000000000000000000000000000000000000..970db2e21fade8aaf671514668f48d313af82c48 --- /dev/null +++ b/deepseek-r1-1.5b-unary4/model_layers_23_mlp_down_proj_weight.planes @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:55dab16cc4d48a6791be064f4f8b94e249be782855eb7ec39de6e49df384893a +size 6881280 diff --git a/deepseek-r1-1.5b-unary4/model_layers_23_mlp_up_proj_weight.sign b/deepseek-r1-1.5b-unary4/model_layers_23_mlp_up_proj_weight.sign new file mode 100644 index 0000000000000000000000000000000000000000..33f134083cd0fa3854d858e8afa9366603777a05 --- /dev/null +++ b/deepseek-r1-1.5b-unary4/model_layers_23_mlp_up_proj_weight.sign @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f0683d8ec1bf8eb9f5a1df2d9e837cb1ce319e14d49e5082b027c54f17ccf2ff +size 1720320 diff --git a/deepseek-r1-1.5b-unary4/model_layers_24_mlp_down_proj_weight.scales b/deepseek-r1-1.5b-unary4/model_layers_24_mlp_down_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..728d98a274850b5f79d9e18629abbcf78895f1df Binary files /dev/null and b/deepseek-r1-1.5b-unary4/model_layers_24_mlp_down_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_24_self_attn_q_proj_bias.fp16 b/deepseek-r1-1.5b-unary4/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-unary4/model_layers_24_self_attn_q_proj_bias.fp16 differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_24_self_attn_q_proj_weight.sign b/deepseek-r1-1.5b-unary4/model_layers_24_self_attn_q_proj_weight.sign new file mode 100644 index 0000000000000000000000000000000000000000..b86bb243cc4ea176534eba12ba16b86ba0324f7d --- /dev/null +++ b/deepseek-r1-1.5b-unary4/model_layers_24_self_attn_q_proj_weight.sign @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2639dc04465f655b2aee10c1b40a78f672901f2bc61f27e1c8f1654582595525 +size 294912 diff --git a/deepseek-r1-1.5b-unary4/model_layers_25_input_layernorm_weight.fp16 b/deepseek-r1-1.5b-unary4/model_layers_25_input_layernorm_weight.fp16 new file mode 100644 index 0000000000000000000000000000000000000000..eee254584de8db018cf609c254bd9d82c6ab810d Binary files /dev/null and b/deepseek-r1-1.5b-unary4/model_layers_25_input_layernorm_weight.fp16 differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_25_mlp_gate_proj_weight.planes b/deepseek-r1-1.5b-unary4/model_layers_25_mlp_gate_proj_weight.planes new file mode 100644 index 0000000000000000000000000000000000000000..5cdbdf6ef174250e27c5ff48494aa379a5b64db6 --- /dev/null +++ b/deepseek-r1-1.5b-unary4/model_layers_25_mlp_gate_proj_weight.planes @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:00d03445678fc99dec6d5617b4b7ca4ff6261fb4e628abffd60c9cd4ba731a04 +size 6881280 diff --git a/deepseek-r1-1.5b-unary4/model_layers_25_mlp_up_proj_weight.sign b/deepseek-r1-1.5b-unary4/model_layers_25_mlp_up_proj_weight.sign new file mode 100644 index 0000000000000000000000000000000000000000..67d1cdb5189180b56d5cffdd149db824d6e57de2 --- /dev/null +++ b/deepseek-r1-1.5b-unary4/model_layers_25_mlp_up_proj_weight.sign @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:85914f8567316fb3a505d76f4efbffddeda82b252fca0f004b115729fc5eece4 +size 1720320 diff --git a/deepseek-r1-1.5b-unary4/model_layers_25_self_attn_k_proj_bias.fp16 b/deepseek-r1-1.5b-unary4/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-unary4/model_layers_25_self_attn_k_proj_bias.fp16 differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_25_self_attn_k_proj_weight.scales b/deepseek-r1-1.5b-unary4/model_layers_25_self_attn_k_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..74217b4dc56f3692746c60fb69bde2693ee638df Binary files /dev/null and b/deepseek-r1-1.5b-unary4/model_layers_25_self_attn_k_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_25_self_attn_o_proj_weight.scales b/deepseek-r1-1.5b-unary4/model_layers_25_self_attn_o_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..d9d81ad82aae7fab8bb91103f56e09b98e5696c7 Binary files /dev/null and b/deepseek-r1-1.5b-unary4/model_layers_25_self_attn_o_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_25_self_attn_v_proj_weight.planes b/deepseek-r1-1.5b-unary4/model_layers_25_self_attn_v_proj_weight.planes new file mode 100644 index 0000000000000000000000000000000000000000..ad9ecffa2a7e82987ad173f2f8a16937004a3940 --- /dev/null +++ b/deepseek-r1-1.5b-unary4/model_layers_25_self_attn_v_proj_weight.planes @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bedcf10b09204fad33702962c3c15baec4181e43fbfbe45a10d6808e45aa1f19 +size 196608 diff --git a/deepseek-r1-1.5b-unary4/model_layers_26_post_attention_layernorm_weight.fp16 b/deepseek-r1-1.5b-unary4/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-unary4/model_layers_26_post_attention_layernorm_weight.fp16 differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_26_self_attn_o_proj_weight.scales b/deepseek-r1-1.5b-unary4/model_layers_26_self_attn_o_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..dcd234268cd7696fa470abdf0d608ffb5e83ac1a Binary files /dev/null and b/deepseek-r1-1.5b-unary4/model_layers_26_self_attn_o_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_26_self_attn_o_proj_weight.sign b/deepseek-r1-1.5b-unary4/model_layers_26_self_attn_o_proj_weight.sign new file mode 100644 index 0000000000000000000000000000000000000000..6725398a58da59592c86abd7b7bc5e9cd102ee2b --- /dev/null +++ b/deepseek-r1-1.5b-unary4/model_layers_26_self_attn_o_proj_weight.sign @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9e6eb1cde78520b1d0f41f169ef73551a302396b269a2dfb2fcaef6e50f4391d +size 294912 diff --git a/deepseek-r1-1.5b-unary4/model_layers_26_self_attn_q_proj_weight.scales b/deepseek-r1-1.5b-unary4/model_layers_26_self_attn_q_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..4ca8f44c0c309be7855871f01b2facce2e3679d7 Binary files /dev/null and b/deepseek-r1-1.5b-unary4/model_layers_26_self_attn_q_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_26_self_attn_v_proj_weight.scales b/deepseek-r1-1.5b-unary4/model_layers_26_self_attn_v_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..b7170714915e9f6a042b23f1ff2ad8ef6615500d Binary files /dev/null and b/deepseek-r1-1.5b-unary4/model_layers_26_self_attn_v_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_27_mlp_down_proj_weight.sign b/deepseek-r1-1.5b-unary4/model_layers_27_mlp_down_proj_weight.sign new file mode 100644 index 0000000000000000000000000000000000000000..d36d1c2ebb6cbd5a8eb7a25a8921177fc36981a1 --- /dev/null +++ b/deepseek-r1-1.5b-unary4/model_layers_27_mlp_down_proj_weight.sign @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:07d27e5c15f9b50f5e6a781953c79efe06ce7d7596d476894237df2835027ee6 +size 1720320 diff --git a/deepseek-r1-1.5b-unary4/model_layers_27_mlp_up_proj_weight.planes b/deepseek-r1-1.5b-unary4/model_layers_27_mlp_up_proj_weight.planes new file mode 100644 index 0000000000000000000000000000000000000000..77e320bb2b9e51b8354776604b919020f85777da --- /dev/null +++ b/deepseek-r1-1.5b-unary4/model_layers_27_mlp_up_proj_weight.planes @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c241fc5ec132d2289f5cbe51e7ecc4dd005fe1c3130ba86bab4f5bb8af39a036 +size 6881280 diff --git a/deepseek-r1-1.5b-unary4/model_layers_27_mlp_up_proj_weight.sign b/deepseek-r1-1.5b-unary4/model_layers_27_mlp_up_proj_weight.sign new file mode 100644 index 0000000000000000000000000000000000000000..d5a5cbad4fc1285be0d5cd82ccb3194391aa949d --- /dev/null +++ b/deepseek-r1-1.5b-unary4/model_layers_27_mlp_up_proj_weight.sign @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d9119d5b576a6262380bbf062dd59b12f0b918f2413febf82d97523d0bc8c662 +size 1720320 diff --git a/deepseek-r1-1.5b-unary4/model_layers_27_self_attn_k_proj_bias.fp16 b/deepseek-r1-1.5b-unary4/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-unary4/model_layers_27_self_attn_k_proj_bias.fp16 differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_27_self_attn_k_proj_weight.planes b/deepseek-r1-1.5b-unary4/model_layers_27_self_attn_k_proj_weight.planes new file mode 100644 index 0000000000000000000000000000000000000000..3d1b64e533f087bd06773414f15c9b1b5bcd18ab --- /dev/null +++ b/deepseek-r1-1.5b-unary4/model_layers_27_self_attn_k_proj_weight.planes @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d40a6ed04b37a2913eac0816c0eed6aa5064ce96d0777c3853c82fdd0db5a0b8 +size 196608 diff --git a/deepseek-r1-1.5b-unary4/model_layers_27_self_attn_q_proj_weight.sign b/deepseek-r1-1.5b-unary4/model_layers_27_self_attn_q_proj_weight.sign new file mode 100644 index 0000000000000000000000000000000000000000..dd59fc4384c09fc059b4a2e44a20aae992e4c6ba --- /dev/null +++ b/deepseek-r1-1.5b-unary4/model_layers_27_self_attn_q_proj_weight.sign @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bf3c436e498a4ccdf3285b8ff407f6e68b05d1977fc9e053991f430c5f723856 +size 294912 diff --git a/deepseek-r1-1.5b-unary4/model_layers_2_mlp_gate_proj_weight.scales b/deepseek-r1-1.5b-unary4/model_layers_2_mlp_gate_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..1bf7d6e5779da9da50ae7f16c4081a2e04198294 Binary files /dev/null and b/deepseek-r1-1.5b-unary4/model_layers_2_mlp_gate_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_2_mlp_up_proj_weight.planes b/deepseek-r1-1.5b-unary4/model_layers_2_mlp_up_proj_weight.planes new file mode 100644 index 0000000000000000000000000000000000000000..6e262988f0770517d35ed6f8c1b3f03eb3c5066e --- /dev/null +++ b/deepseek-r1-1.5b-unary4/model_layers_2_mlp_up_proj_weight.planes @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ca2040dcd862dc9661ecabc4efafe8360c9d8968b9ce2dd43d082bc10ca7be6a +size 6881280 diff --git a/deepseek-r1-1.5b-unary4/model_layers_2_mlp_up_proj_weight.sign b/deepseek-r1-1.5b-unary4/model_layers_2_mlp_up_proj_weight.sign new file mode 100644 index 0000000000000000000000000000000000000000..c5417a844321188a2cbf1efaa7e9af9620da0ba3 --- /dev/null +++ b/deepseek-r1-1.5b-unary4/model_layers_2_mlp_up_proj_weight.sign @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2959ca73da13885c8bd2b97eba55955a11c54af86f4ded2216013ef7283830ac +size 1720320 diff --git a/deepseek-r1-1.5b-unary4/model_layers_2_self_attn_k_proj_bias.fp16 b/deepseek-r1-1.5b-unary4/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-unary4/model_layers_2_self_attn_k_proj_bias.fp16 differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_2_self_attn_q_proj_weight.sign b/deepseek-r1-1.5b-unary4/model_layers_2_self_attn_q_proj_weight.sign new file mode 100644 index 0000000000000000000000000000000000000000..d10f37c35d5a1f5357cc4e447316c3e4ac8cabb2 --- /dev/null +++ b/deepseek-r1-1.5b-unary4/model_layers_2_self_attn_q_proj_weight.sign @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:816d513cdeac8dc0605a5808e97cfc152dddd76d3c3ac8451508ba3b140cbf6b +size 294912 diff --git a/deepseek-r1-1.5b-unary4/model_layers_3_mlp_gate_proj_weight.planes b/deepseek-r1-1.5b-unary4/model_layers_3_mlp_gate_proj_weight.planes new file mode 100644 index 0000000000000000000000000000000000000000..c2af745426a9724c1bbb4f12027dd8fd2fa07e24 --- /dev/null +++ b/deepseek-r1-1.5b-unary4/model_layers_3_mlp_gate_proj_weight.planes @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:594893ca1360f345c33c338925d1afc5012b332589064cd5ed1a7e6e2c1a0e6c +size 6881280 diff --git a/deepseek-r1-1.5b-unary4/model_layers_3_mlp_up_proj_weight.scales b/deepseek-r1-1.5b-unary4/model_layers_3_mlp_up_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..271e9bbe52e46a8fa7186d09121dcb5a0f074ef7 Binary files /dev/null and b/deepseek-r1-1.5b-unary4/model_layers_3_mlp_up_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_3_self_attn_o_proj_weight.sign b/deepseek-r1-1.5b-unary4/model_layers_3_self_attn_o_proj_weight.sign new file mode 100644 index 0000000000000000000000000000000000000000..ee922a410d659781aae32a67b9082a0aa8980bf2 --- /dev/null +++ b/deepseek-r1-1.5b-unary4/model_layers_3_self_attn_o_proj_weight.sign @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1f794bbcdffd07b8719ce97780dab4eabc04d310bbe111dfba7f8205fb80b684 +size 294912 diff --git a/deepseek-r1-1.5b-unary4/model_layers_4_mlp_down_proj_weight.scales b/deepseek-r1-1.5b-unary4/model_layers_4_mlp_down_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..659f9882fc41cb8de36e9167ba3c3f8e19663f14 Binary files /dev/null and b/deepseek-r1-1.5b-unary4/model_layers_4_mlp_down_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_4_mlp_gate_proj_weight.sign b/deepseek-r1-1.5b-unary4/model_layers_4_mlp_gate_proj_weight.sign new file mode 100644 index 0000000000000000000000000000000000000000..85f1804740d2fa63cb5919cfc7c5f201e6f71db6 --- /dev/null +++ b/deepseek-r1-1.5b-unary4/model_layers_4_mlp_gate_proj_weight.sign @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:74526432e6daa0d6f46fcb15e7a06a0dce50ab8453df5aade0383b27da9b1eaa +size 1720320 diff --git a/deepseek-r1-1.5b-unary4/model_layers_4_self_attn_k_proj_weight.scales b/deepseek-r1-1.5b-unary4/model_layers_4_self_attn_k_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..c3fb8aea54b256041cb15d68ab317ed637af901a Binary files /dev/null and b/deepseek-r1-1.5b-unary4/model_layers_4_self_attn_k_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_4_self_attn_k_proj_weight.sign b/deepseek-r1-1.5b-unary4/model_layers_4_self_attn_k_proj_weight.sign new file mode 100644 index 0000000000000000000000000000000000000000..8a9f7f5ecb8401421b0ccc0c52eb097b3f738030 Binary files /dev/null and b/deepseek-r1-1.5b-unary4/model_layers_4_self_attn_k_proj_weight.sign differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_4_self_attn_v_proj_bias.fp16 b/deepseek-r1-1.5b-unary4/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-unary4/model_layers_4_self_attn_v_proj_bias.fp16 differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_5_mlp_down_proj_weight.planes b/deepseek-r1-1.5b-unary4/model_layers_5_mlp_down_proj_weight.planes new file mode 100644 index 0000000000000000000000000000000000000000..43cd0fd2ea7826ce4011c823efb82e6efd42bc8b --- /dev/null +++ b/deepseek-r1-1.5b-unary4/model_layers_5_mlp_down_proj_weight.planes @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1109697a3e33987316b180beea5079e4b5d0033ddb5fb4e6abaddac9e6ac5451 +size 6881280 diff --git a/deepseek-r1-1.5b-unary4/model_layers_5_self_attn_q_proj_bias.fp16 b/deepseek-r1-1.5b-unary4/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-unary4/model_layers_5_self_attn_q_proj_bias.fp16 differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_5_self_attn_q_proj_weight.sign b/deepseek-r1-1.5b-unary4/model_layers_5_self_attn_q_proj_weight.sign new file mode 100644 index 0000000000000000000000000000000000000000..25d138ac7ad161538ce403fcf1ec8d9b3c35463d --- /dev/null +++ b/deepseek-r1-1.5b-unary4/model_layers_5_self_attn_q_proj_weight.sign @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ea502782ca87f1068f0d34260206df8b18c106eee96ce075fe5eecc7e9382cf4 +size 294912 diff --git a/deepseek-r1-1.5b-unary4/model_layers_5_self_attn_v_proj_weight.planes b/deepseek-r1-1.5b-unary4/model_layers_5_self_attn_v_proj_weight.planes new file mode 100644 index 0000000000000000000000000000000000000000..3111e1aee203cdc45281fd6b12768bb6425b6f14 --- /dev/null +++ b/deepseek-r1-1.5b-unary4/model_layers_5_self_attn_v_proj_weight.planes @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:affd688d35ba7366453fd841e1a54293419721435d748368cd3baa24378b8c23 +size 196608 diff --git a/deepseek-r1-1.5b-unary4/model_layers_6_self_attn_k_proj_weight.scales b/deepseek-r1-1.5b-unary4/model_layers_6_self_attn_k_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..76fe5b395f47849569d99d087061887eda81273f Binary files /dev/null and b/deepseek-r1-1.5b-unary4/model_layers_6_self_attn_k_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_6_self_attn_q_proj_weight.planes b/deepseek-r1-1.5b-unary4/model_layers_6_self_attn_q_proj_weight.planes new file mode 100644 index 0000000000000000000000000000000000000000..f8cb8328b828f76558e7b1b4302bfde79a297105 --- /dev/null +++ b/deepseek-r1-1.5b-unary4/model_layers_6_self_attn_q_proj_weight.planes @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:369c457e8f2519edf873205343ecd153f6b4ca5a9a99d2d6f842f146ec576486 +size 1179648 diff --git a/deepseek-r1-1.5b-unary4/model_layers_6_self_attn_v_proj_weight.sign b/deepseek-r1-1.5b-unary4/model_layers_6_self_attn_v_proj_weight.sign new file mode 100644 index 0000000000000000000000000000000000000000..f70bd26bb7dea8c6669c791def68e2087e5902e6 Binary files /dev/null and b/deepseek-r1-1.5b-unary4/model_layers_6_self_attn_v_proj_weight.sign differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_7_self_attn_o_proj_weight.sign b/deepseek-r1-1.5b-unary4/model_layers_7_self_attn_o_proj_weight.sign new file mode 100644 index 0000000000000000000000000000000000000000..4d6433dbbb230ba3c9bb6fe2ad65ebe996566e12 --- /dev/null +++ b/deepseek-r1-1.5b-unary4/model_layers_7_self_attn_o_proj_weight.sign @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a4b8961d1cbe2020f8bae5485e7b78fbb1f950e54ae311abc242bc26fe868554 +size 294912 diff --git a/deepseek-r1-1.5b-unary4/model_layers_8_mlp_up_proj_weight.sign b/deepseek-r1-1.5b-unary4/model_layers_8_mlp_up_proj_weight.sign new file mode 100644 index 0000000000000000000000000000000000000000..6cd856e61b79e2b67a3acccacd108895063741c3 --- /dev/null +++ b/deepseek-r1-1.5b-unary4/model_layers_8_mlp_up_proj_weight.sign @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d3f6aa79044e0dc8b68a383cbbe079719694dd1d5d0cdc3b5b41a12fb68288dc +size 1720320 diff --git a/deepseek-r1-1.5b-unary4/model_layers_8_self_attn_v_proj_weight.scales b/deepseek-r1-1.5b-unary4/model_layers_8_self_attn_v_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..973180352ea077f0a515d0232a470486f1759d8c Binary files /dev/null and b/deepseek-r1-1.5b-unary4/model_layers_8_self_attn_v_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_9_self_attn_k_proj_bias.fp16 b/deepseek-r1-1.5b-unary4/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-unary4/model_layers_9_self_attn_k_proj_bias.fp16 differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_9_self_attn_k_proj_weight.scales b/deepseek-r1-1.5b-unary4/model_layers_9_self_attn_k_proj_weight.scales new file mode 100644 index 0000000000000000000000000000000000000000..ab7a95c2e3941ae97947abda593314156a253808 Binary files /dev/null and b/deepseek-r1-1.5b-unary4/model_layers_9_self_attn_k_proj_weight.scales differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_9_self_attn_k_proj_weight.sign b/deepseek-r1-1.5b-unary4/model_layers_9_self_attn_k_proj_weight.sign new file mode 100644 index 0000000000000000000000000000000000000000..8898f356516b2954516ea9717f2e9df3501d32a0 Binary files /dev/null and b/deepseek-r1-1.5b-unary4/model_layers_9_self_attn_k_proj_weight.sign differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_9_self_attn_v_proj_bias.fp16 b/deepseek-r1-1.5b-unary4/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-unary4/model_layers_9_self_attn_v_proj_bias.fp16 differ diff --git a/deepseek-r1-1.5b-unary4/model_layers_9_self_attn_v_proj_weight.sign b/deepseek-r1-1.5b-unary4/model_layers_9_self_attn_v_proj_weight.sign new file mode 100644 index 0000000000000000000000000000000000000000..3b74c1455affc1fd0283a843a8ffe7cf073ba8a6 Binary files /dev/null and b/deepseek-r1-1.5b-unary4/model_layers_9_self_attn_v_proj_weight.sign differ