Merge branch 'master' into HEAD

ggml/src/ggml-vulkan/vulkan-shaders/mul_mat_vec_iq1_s.comp

@@ -10,44 +10,44 @@ FLOAT_TYPE temp[NUM_COLS][NUM_ROWS];
 void calc_superblock(const uint a_offset, const uint b_offset, const uint ib32, const uint i,
                      const uint num_blocks_per_row, const uint first_row, const uint num_rows) {
     const uint y_idx_base = i * QUANT_K + 32 * ib32;
-    [[unroll]] for (uint j = 0; j < NUM_COLS; ++j) {    
+    [[unroll]] for (uint j = 0; j < NUM_COLS; ++j) {
         const uint base_b_idx = (j * p.batch_stride_b + b_offset + y_idx_base) / 4;
-        [[unroll]] for (uint l = 0; l < 4; ++l) {            
+        [[unroll]] for (uint l = 0; l < 4; ++l) {
             const vec4 b_val_0 = vec4(data_b_v4[base_b_idx + 2 * l]);
             const vec4 b_val_1 = vec4(data_b_v4[base_b_idx + 2 * l + 1]);
 
             // index for data_a
             uint ibi = a_offset / QUANT_K + first_row * num_blocks_per_row + i;
 
-            [[unroll]] for (uint n = 0; n < num_rows; ++n) {                
+            [[unroll]] for (uint n = 0; n < num_rows; ++n) {
                 const float d = float(data_a[ibi].d);
                 const uint qh = data_a[ibi].qh[ib32];
 
                 const float dl = d * float(2 * bitfieldExtract(qh, 12, 3) + 1);
                 const uint qs = data_a[ibi].qs[4 * ib32 + l];
-                const uint idxhi = bitfieldExtract(qh, 3 * int(l), 3);              
+                const uint idxhi = bitfieldExtract(qh, 3 * int(l), 3);
                 const uint16_t grid = uint16_t(iq1s_grid[qs | (idxhi << 8)]);
 
                 const float delta_val = ((qh & 0x8000) != 0) ? -IQ1S_DELTA : IQ1S_DELTA;
-                const vec4 delta_v = vec4(delta_val);				
+                const vec4 delta_v = vec4(delta_val);
                 const vec4 fbits0 = vec4(
                     float(bitfieldExtract(grid, 0, 2)),
                     float(bitfieldExtract(grid, 2, 2)),
                     float(bitfieldExtract(grid, 4, 2)),
                     float(bitfieldExtract(grid, 6, 2))
-                );				
+                );
                 const vec4 fbits1 = vec4(
                     float(bitfieldExtract(grid, 8, 2)),
                     float(bitfieldExtract(grid, 10, 2)),
                     float(bitfieldExtract(grid, 12, 2)),
                     float(bitfieldExtract(grid, 14, 2))
                 );
-				
+
                 vec4 sum_v = fma(b_val_0, fbits0 + delta_v, vec4(0.0));
                 sum_v      = fma(b_val_1, fbits1 + delta_v, sum_v);
-				FLOAT_TYPE sum = dot(sum_v, vec4(1.0));
+                FLOAT_TYPE sum = dot(sum_v, vec4(1.0));
-				
+
-                temp[j][n] = fma(dl, sum, temp[j][n]);				
+                temp[j][n] = fma(dl, sum, temp[j][n]);
                 ibi += num_blocks_per_row;
             }
         }

src/llama-graph.cpp

@@ -79,7 +79,7 @@ void llm_graph_input_attn_temp::set_input(const llama_ubatch * ubatch) {
         for (int i = 0; i < n_tokens; ++i) {
             const float pos = ubatch->pos[i];
             attn_scale_data[i] = std::log(
-                std::floor((pos + 1.0f) / n_attn_temp_floor_scale) + 1.0
+                std::floor((pos + f_attn_temp_offset) / n_attn_temp_floor_scale) + 1.0
             ) * f_attn_temp_scale + 1.0;
         }
 
@@ -1278,7 +1278,7 @@ ggml_tensor * llm_graph_context::build_inp_pos() const {
 }
 
 ggml_tensor * llm_graph_context::build_inp_attn_scale() const {
-    auto inp = std::make_unique<llm_graph_input_attn_temp>(hparams.n_attn_temp_floor_scale, hparams.f_attn_temp_scale);
+    auto inp = std::make_unique<llm_graph_input_attn_temp>(hparams.n_attn_temp_floor_scale, hparams.f_attn_temp_scale, hparams.f_attn_temp_offset);
 
     auto & cur = inp->attn_scale;
 

src/llama-graph.h

@@ -133,8 +133,8 @@ public:
 // temperature tuning, used by llama4
 class llm_graph_input_attn_temp : public llm_graph_input_i {
 public:
-    llm_graph_input_attn_temp(uint32_t n_attn_temp_floor_scale, float f_attn_temp_scale)
+    llm_graph_input_attn_temp(uint32_t n_attn_temp_floor_scale, float f_attn_temp_scale, float f_attn_temp_offset)
-        : n_attn_temp_floor_scale(n_attn_temp_floor_scale), f_attn_temp_scale(f_attn_temp_scale) {}
+        : n_attn_temp_floor_scale(n_attn_temp_floor_scale), f_attn_temp_scale(f_attn_temp_scale), f_attn_temp_offset(f_attn_temp_offset) {}
     virtual ~llm_graph_input_attn_temp() = default;
 
     void set_input(const llama_ubatch * ubatch) override;
@@ -143,6 +143,7 @@ public:
 
     const uint32_t n_attn_temp_floor_scale;
     const float    f_attn_temp_scale;
+    const float    f_attn_temp_offset;
 };
 
 class llm_graph_input_pos_bucket : public llm_graph_input_i {

src/llama-hparams.h

@@ -165,6 +165,7 @@ struct llama_hparams {
     uint32_t n_no_rope_layer_step    = 4;
     uint32_t n_attn_temp_floor_scale = 0;
     float    f_attn_temp_scale       = 0.0f;
+    float    f_attn_temp_offset      = 0.0f; // offset position index
 
     // gemma3n altup
     uint32_t n_altup      = 4; // altup_num_inputs

src/llama-model.cpp

@@ -668,6 +668,7 @@ void llama_model::load_hparams(llama_model_loader & ml) {
                     hparams.n_swa                   = 8192;
                     hparams.n_attn_temp_floor_scale = 8192;
                     hparams.f_attn_temp_scale       = 0.1f;
+                    hparams.f_attn_temp_offset      = 1.0f;
                     hparams.set_swa_pattern(4);   // pattern: 3 chunked - 1 full
                 }
 
@@ -1646,6 +1647,8 @@ void llama_model::load_hparams(llama_model_loader & ml) {
                 ml.get_key(LLM_KV_ATTENTION_TEMPERATURE_SCALE,  hparams.f_attn_temp_scale,       false);
                 ml.get_key(LLM_KV_ATTENTION_TEMPERATURE_LENGTH, hparams.n_attn_temp_floor_scale, false);
 
+                hparams.f_attn_temp_offset = 0.0f;
+
                 switch (hparams.n_layer) {
                     case 27: type = LLM_TYPE_16B; break;
                     case 60: type = LLM_TYPE_236B; break;
@@ -2276,6 +2279,8 @@ void llama_model::load_hparams(llama_model_loader & ml) {
                 ml.get_key(LLM_KV_ROPE_SCALING_YARN_BETA_SLOW, hparams.yarn_beta_slow,    false);
                 ml.get_key(LLM_KV_ROPE_SCALING_YARN_LOG_MUL,   hparams.rope_yarn_log_mul, 0.0f);
 
+                hparams.f_attn_temp_offset = 0.0f;
+
                 // TODO: maybe add n_attn_temp_floor_scale as a separate KV?
                 if (hparams.f_attn_temp_scale != 0.0f) {
                     hparams.n_attn_temp_floor_scale = hparams.n_ctx_orig_yarn;