bitnet: add 2 bpw quantization

The scalar dot product already chieves 37 t/s for TG!

1 files changed, 5 insertions, 0 deletions

diff --git a/ggml-quants.h b/ggml-quants.h
index a66fe6d6..cc677207 100644
--- a/ggml-quants.h
+++ b/ggml-quants.h
@@ -33,6 +33,7 @@ void quantize_row_iq4_xs_reference (const float * GGML_RESTRICT x, block_iq4_xs
 void quantize_row_iq3_s_reference  (const float * GGML_RESTRICT x, block_iq3_s   * GGML_RESTRICT y, int64_t k);
 void quantize_row_iq2_s_reference  (const float * GGML_RESTRICT x, block_iq2_s   * GGML_RESTRICT y, int64_t k);
 void quantize_row_iq1_bn_reference (const float * GGML_RESTRICT x, block_iq1_bn  * GGML_RESTRICT y, int64_t k);
+void quantize_row_iq2_bn_reference (const float * GGML_RESTRICT x, block_iq2_bn  * GGML_RESTRICT y, int64_t k);
 
 void quantize_row_q4_0(const float * GGML_RESTRICT x, void * GGML_RESTRICT y, int64_t k);
 void quantize_row_q4_1(const float * GGML_RESTRICT x, void * GGML_RESTRICT y, int64_t k);
@@ -55,6 +56,7 @@ void quantize_row_iq4_xs (const float * GGML_RESTRICT x, void * GGML_RESTRICT y,
 void quantize_row_iq3_s  (const float * GGML_RESTRICT x, void * GGML_RESTRICT y, int64_t k);
 void quantize_row_iq2_s  (const float * GGML_RESTRICT x, void * GGML_RESTRICT y, int64_t k);
 void quantize_row_iq1_bn (const float * GGML_RESTRICT x, void * GGML_RESTRICT y, int64_t k);
+void quantize_row_iq2_bn (const float * GGML_RESTRICT x, void * GGML_RESTRICT y, int64_t k);
 
 // Dequantization
 void dequantize_row_q4_0(const block_q4_0 * GGML_RESTRICT x, float * GGML_RESTRICT y, int64_t k);
@@ -81,6 +83,7 @@ void dequantize_row_iq4_nl (const block_iq4_nl  * GGML_RESTRICT x, float * GGML_
 void dequantize_row_iq4_xs (const block_iq4_xs  * GGML_RESTRICT x, float * GGML_RESTRICT y, int64_t k);
 void dequantize_row_iq3_s  (const block_iq3_s   * GGML_RESTRICT x, float * GGML_RESTRICT y, int64_t k);
 void dequantize_row_iq1_bn (const block_iq1_bn  * GGML_RESTRICT x, float * GGML_RESTRICT y, int64_t k);
+void dequantize_row_iq2_bn (const block_iq2_bn  * GGML_RESTRICT x, float * GGML_RESTRICT y, int64_t k);
 
 // Dot product
 void ggml_vec_dot_q4_0_q8_0(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, size_t bx, const void * GGML_RESTRICT vy, size_t by, int nrc);
@@ -106,6 +109,7 @@ void ggml_vec_dot_iq4_xs_q8_K (int n, float * GGML_RESTRICT s, size_t bs, const
 void ggml_vec_dot_iq3_s_q8_K  (int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, size_t bx, const void * GGML_RESTRICT vy, size_t by, int nrc);
 void ggml_vec_dot_iq1_bn_q8_0 (int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, size_t bx, const void * GGML_RESTRICT vy, size_t by, int nrc);
 void ggml_vec_dot_iq1_bn_q8_K64(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, size_t bx, const void * GGML_RESTRICT vy, size_t by, int nrc);
+void ggml_vec_dot_iq2_bn_q8_K64(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, size_t bx, const void * GGML_RESTRICT vy, size_t by, int nrc);
 
 // Quantization utilizing an importance matrix (a.k.a. "Activation aWare Quantization")
 size_t quantize_iq2_xxs(const float * GGML_RESTRICT src, void * GGML_RESTRICT dst, int64_t nrows, int64_t n_per_row, const float * imatrix);
@@ -118,6 +122,7 @@ size_t quantize_iq4_nl (const float * GGML_RESTRICT src, void * GGML_RESTRICT ds
 size_t quantize_iq4_xs (const float * GGML_RESTRICT src, void * GGML_RESTRICT dst, int64_t nrows, int64_t n_per_row, const float * imatrix);
 size_t quantize_iq3_s  (const float * GGML_RESTRICT src, void * GGML_RESTRICT dst, int64_t nrows, int64_t n_per_row, const float * imatrix);
 size_t quantize_iq1_bn (const float * GGML_RESTRICT src, void * GGML_RESTRICT dst, int64_t nrows, int64_t n_per_row, const float * imatrix);
+size_t quantize_iq2_bn (const float * GGML_RESTRICT src, void * GGML_RESTRICT dst, int64_t nrows, int64_t n_per_row, const float * imatrix);
 
 size_t quantize_q2_K(const float * GGML_RESTRICT src, void * GGML_RESTRICT dst, int64_t nrows, int64_t n_per_row, const float * imatrix);
 size_t quantize_q3_K(const float * GGML_RESTRICT src, void * GGML_RESTRICT dst, int64_t nrows, int64_t n_per_row, const float * imatrix);