cui-llama.rn 1.7.4 → 1.7.6

package/ios/rnllama.xcframework/ios-arm64_x86_64-simulator/rnllama.framework/Headers/llama-hparams.h

@@ -115,6 +115,9 @@ struct llama_hparams {
     uint32_t ssm_d_state = 0;
     uint32_t ssm_dt_rank = 0;
 
+    // for hybrid state space models
+    std::array<bool, LLAMA_MAX_LAYERS> recurrent_layer_arr;
+
     bool ssm_dt_b_c_rms = false;
 
     float f_clamp_kqv      = 0.0f;
@@ -131,12 +134,21 @@ struct llama_hparams {
     bool attn_soft_cap = false;
     bool use_kq_norm   = true;
 
+    // for Classifiers
+    uint32_t n_cls_out = 1;
+
     // llama4
     uint32_t n_moe_layer_step        = 0;
     uint32_t n_no_rope_layer_step    = 4;
     uint32_t n_attn_temp_floor_scale = 8192;
     float    f_attn_temp_scale       = 0.1;
 
+    // gemma3n altup
+    uint32_t n_altup      = 4; // altup_num_inputs
+    uint32_t i_altup_act  = 0; // altup_active_idx
+    uint32_t laurel_rank  = 64;
+    uint32_t n_embd_altup = 256;
+
     // needed by encoder-decoder models (e.g. T5, FLAN-T5)
     // ref: https://github.com/ggerganov/llama.cpp/pull/8141
     llama_token dec_start_token_id = LLAMA_TOKEN_NULL;
@@ -178,10 +190,15 @@ struct llama_hparams {
 
     // dimension of the rolling state embeddings
     // corresponds to Mamba's conv_states size or RWKV's token_shift states size
-    uint32_t n_embd_k_s() const;
+    uint32_t n_embd_r() const;
 
     // dimension of the recurrent state embeddings
-    uint32_t n_embd_v_s() const;
+    uint32_t n_embd_s() const;
+
+    // whether or not the given layer is recurrent (for hybrid models)
+    bool is_recurrent(uint32_t il) const;
+
+    uint32_t n_pos_per_embd() const;
 
     bool is_swa(uint32_t il) const;
 };

package/ios/rnllama.xcframework/ios-arm64_x86_64-simulator/rnllama.framework/Headers/llama-kv-cache-unified-iswa.h

@@ -0,0 +1,128 @@
+#pragma once
+
+#include "llama-kv-cache-unified.h"
+
+#include <vector>
+
+//
+// llama_kv_cache_unified_iswa
+//
+
+// utilizes two instances of llama_kv_cache_unified
+//   the first instance is for the non-SWA layers of the model and the second instance is for the SWA layers
+
+class llama_kv_cache_unified_iswa : public llama_memory_i {
+public:
+    llama_kv_cache_unified_iswa(
+            const llama_model & model,
+                    lm_ggml_type   type_k,
+                    lm_ggml_type   type_v,
+                         bool   v_trans,
+                         bool   offload,
+                         bool   swa_full,
+                     uint32_t   kv_size,
+                     uint32_t   n_seq_max,
+                     uint32_t   n_ubatch,
+                     uint32_t   n_pad);
+
+    ~llama_kv_cache_unified_iswa() = default;
+
+    //
+    // llama_memory_i
+    //
+
+    llama_memory_context_ptr init_batch(
+            llama_batch_allocr & balloc,
+            uint32_t n_ubatch,
+            bool embd_all) override;
+
+    llama_memory_context_ptr init_full() override;
+
+    llama_memory_context_ptr init_update(llama_context * lctx, bool optimize) override;
+
+    bool get_can_shift() const override;
+
+    void clear(bool data) override;
+
+    bool seq_rm  (llama_seq_id seq_id,                              llama_pos p0, llama_pos p1) override;
+    void seq_cp  (llama_seq_id seq_id_src, llama_seq_id seq_id_dst, llama_pos p0, llama_pos p1) override;
+    void seq_keep(llama_seq_id seq_id)                                                          override;
+    void seq_add (llama_seq_id seq_id,                              llama_pos p0, llama_pos p1, llama_pos shift) override;
+    void seq_div (llama_seq_id seq_id,                              llama_pos p0, llama_pos p1, int d) override;
+
+    llama_pos seq_pos_min(llama_seq_id seq_id) const override;
+    llama_pos seq_pos_max(llama_seq_id seq_id) const override;
+
+    // state write/load
+
+    void state_write(llama_io_write_i & io, llama_seq_id seq_id = -1) const override;
+    void state_read (llama_io_read_i  & io, llama_seq_id seq_id = -1)       override;
+
+    //
+    // llama_kv_cache_unified_iswa specific API
+    //
+
+    llama_kv_cache_unified * get_base() const;
+    llama_kv_cache_unified * get_swa () const;
+
+private:
+    const llama_hparams & hparams;
+
+    std::unique_ptr<llama_kv_cache_unified> kv_base;
+    std::unique_ptr<llama_kv_cache_unified> kv_swa;
+};
+
+class llama_kv_cache_unified_iswa_context : public llama_memory_context_i {
+public:
+    // used for errors
+    llama_kv_cache_unified_iswa_context(llama_memory_status status);
+
+    // used to create a full-cache context
+    llama_kv_cache_unified_iswa_context(
+            llama_kv_cache_unified_iswa * kv);
+
+    // used to create an update context
+    llama_kv_cache_unified_iswa_context(
+            llama_kv_cache_unified_iswa * kv,
+            llama_context * lctx,
+            bool optimize);
+
+    // used to create a batch processing context from a batch
+    llama_kv_cache_unified_iswa_context(
+            llama_kv_cache_unified_iswa * kv,
+            std::vector<uint32_t> heads_base,
+            std::vector<uint32_t> heads_swa,
+            std::vector<llama_ubatch> ubatches);
+
+    virtual ~llama_kv_cache_unified_iswa_context();
+
+    //
+    // llama_memory_context_i
+    //
+
+    bool next()  override;
+    bool apply() override;
+
+    llama_memory_status  get_status() const override;
+    const llama_ubatch & get_ubatch() const override;
+
+    //
+    // llama_kv_cache_unified_iswa_context specific API
+    //
+
+    const llama_kv_cache_unified_context * get_base() const;
+    const llama_kv_cache_unified_context * get_swa()  const;
+
+private:
+    //llama_kv_cache_unified_iswa * kv;
+
+    // the index of the next ubatch to process
+    size_t i_next = 0;
+
+    std::vector<llama_ubatch> ubatches;
+
+    const llama_memory_context_ptr ctx_base;
+    const llama_memory_context_ptr ctx_swa;
+
+    const llama_memory_status status;
+};

package/ios/rnllama.xcframework/ios-arm64_x86_64-simulator/rnllama.framework/Headers/llama-kv-cache-unified.h

@@ -0,0 +1,303 @@
+#pragma once
+
+#include "llama-batch.h"
+#include "llama-graph.h"
+#include "llama-kv-cells.h"
+#include "llama-memory.h"
+
+#include <unordered_map>
+#include <vector>
+
+struct llama_cparams;
+struct llama_hparams;
+struct llama_model;
+struct llama_context;
+
+//
+// llama_kv_cache_unified
+//
+
+class llama_kv_cache_unified : public llama_memory_i {
+public:
+    static uint32_t get_padding(const llama_cparams & cparams);
+
+    // this callback is used to filter out layers that should not be included in the cache
+    using layer_filter_cb = std::function<bool(int32_t il)>;
+
+    using ubatch_heads = std::vector<uint32_t>;
+
+    struct defrag_info {
+        bool empty() const {
+            return ids.empty();
+        }
+
+        // contains information about which cell moves where:
+        //  - cell i moves to ids[i]
+        //  - if ids[i] == i || ids[i] == ids.size(), then cell i is not moved
+        std::vector<uint32_t> ids;
+    };
+
+    llama_kv_cache_unified(
+            const llama_model &  model,
+              layer_filter_cb && filter,
+                    lm_ggml_type    type_k,
+                    lm_ggml_type    type_v,
+                         bool    v_trans,
+                         bool    offload,
+                     uint32_t    kv_size,
+                     uint32_t    n_seq_max,
+                     uint32_t    n_pad,
+                     uint32_t    n_swa,
+               llama_swa_type    swa_type);
+
+    ~llama_kv_cache_unified() = default;
+
+    //
+    // llama_memory_i
+    //
+
+    llama_memory_context_ptr init_batch(
+            llama_batch_allocr & balloc,
+            uint32_t n_ubatch,
+            bool embd_all) override;
+
+    llama_memory_context_ptr init_full() override;
+
+    llama_memory_context_ptr init_update(llama_context * lctx, bool optimize) override;
+
+    bool get_can_shift() const override;
+
+    void clear(bool data) override;
+
+    bool seq_rm  (llama_seq_id seq_id,                              llama_pos p0, llama_pos p1) override;
+    void seq_cp  (llama_seq_id seq_id_src, llama_seq_id seq_id_dst, llama_pos p0, llama_pos p1) override;
+    void seq_keep(llama_seq_id seq_id)                                                          override;
+    void seq_add (llama_seq_id seq_id,                              llama_pos p0, llama_pos p1, llama_pos shift) override;
+    void seq_div (llama_seq_id seq_id,                              llama_pos p0, llama_pos p1, int d) override;
+
+    llama_pos seq_pos_min(llama_seq_id seq_id) const override;
+    llama_pos seq_pos_max(llama_seq_id seq_id) const override;
+
+    // state write/load
+
+    void state_write(llama_io_write_i & io, llama_seq_id seq_id = -1) const override;
+    void state_read (llama_io_read_i  & io, llama_seq_id seq_id = -1)       override;
+
+    //
+    // llama_kv_cache_unified specific API
+    //
+
+    uint32_t get_size() const;
+
+    bool get_has_shift() const;
+
+    //
+    // graph_build API
+    //
+
+    uint32_t get_n_kv() const;
+
+    // get views of the current state of the cache
+    lm_ggml_tensor * get_k(lm_ggml_context * ctx, int32_t il, uint32_t n_kv) const;
+    lm_ggml_tensor * get_v(lm_ggml_context * ctx, int32_t il, uint32_t n_kv) const;
+
+    // store k_cur and v_cur in the cache based on the provided head location
+    lm_ggml_tensor * cpy_k(lm_ggml_context * ctx, lm_ggml_tensor * k_cur, int32_t il, uint32_t head_cur) const;
+    lm_ggml_tensor * cpy_v(lm_ggml_context * ctx, lm_ggml_tensor * v_cur, int32_t il, uint32_t head_cur) const;
+
+    //
+    // preparation API
+    //
+
+    // find places for the provided ubatches in the cache, returns the head locations
+    // return empty vector on failure
+    ubatch_heads prepare(const std::vector<llama_ubatch> & ubatches);
+
+    bool update(llama_context * lctx, bool do_shift, const defrag_info & dinfo);
+
+    // return the cell position where we can insert the ubatch
+    // return -1 on failure to find a contiguous slot of kv cells
+    int32_t find_slot(const llama_ubatch & ubatch) const;
+
+    // emplace the ubatch context into slot: [head_cur, head_cur + ubatch.n_tokens)
+    void apply_ubatch(uint32_t head_cur, const llama_ubatch & ubatch);
+
+    //
+    // set_input API
+    //
+
+    void set_input_kq_mask   (lm_ggml_tensor * dst, const llama_ubatch * ubatch, bool causal_attn) const;
+    void set_input_k_shift   (lm_ggml_tensor * dst) const;
+    void set_input_pos_bucket(lm_ggml_tensor * dst, const llama_ubatch * ubatch) const;
+
+private:
+    const llama_model & model;
+    const llama_hparams & hparams;
+
+    struct kv_layer {
+        // layer index in the model
+        // note: can be different from the layer index in the KV cache
+        uint32_t il;
+
+        lm_ggml_tensor * k;
+        lm_ggml_tensor * v;
+    };
+
+    bool v_trans = true;  // the value tensor is transposed
+
+    // the current index from where we start searching for a free slot in the ring buffer of KV cells (see find_slot())
+    // note: this is not part of the KV state and it's only used to speed-up the find_slot() method
+    uint32_t head = 0;
+
+    const uint32_t n_seq_max = 1;
+
+    // required padding
+    const uint32_t n_pad = 1;
+
+    // SWA
+    const uint32_t n_swa = 0;
+
+    int debug = 0;
+
+    const llama_swa_type swa_type = LLAMA_SWA_TYPE_NONE;
+
+    std::vector<lm_ggml_context_ptr>        ctxs;
+    std::vector<lm_ggml_backend_buffer_ptr> bufs;
+
+    llama_kv_cells_unified cells;
+
+    std::vector<kv_layer> layers;
+
+    // model layer id -> KV cache layer id
+    std::unordered_map<int32_t, int32_t> map_layer_ids;
+
+    // return non-empty vector if cells have been moved
+    defrag_info defrag_prepare(int32_t n_max_nodes) const;
+
+    size_t total_size() const;
+
+    size_t size_k_bytes() const;
+    size_t size_v_bytes() const;
+
+    bool is_masked_swa(llama_pos p0, llama_pos p1) const;
+
+    lm_ggml_tensor * build_rope_shift(
+            const llama_cparams & cparams,
+                   lm_ggml_context * ctx,
+                    lm_ggml_tensor * cur,
+                    lm_ggml_tensor * shift,
+                    lm_ggml_tensor * factors,
+                          float   freq_base,
+                          float   freq_scale) const;
+
+    llm_graph_result_ptr build_graph_shift(
+            const llama_cparams & cparams,
+                   lm_ggml_context * ctx,
+                    lm_ggml_cgraph * gf) const;
+
+    llm_graph_result_ptr build_graph_defrag(
+            const llama_cparams & cparams,
+                   lm_ggml_context * ctx,
+                    lm_ggml_cgraph * gf,
+              const defrag_info & dinfo) const;
+
+    void state_write_meta(llama_io_write_i & io, const std::vector<std::pair<uint32_t, uint32_t>> & cell_ranges, llama_seq_id seq_id = -1) const;
+    void state_write_data(llama_io_write_i & io, const std::vector<std::pair<uint32_t, uint32_t>> & cell_ranges) const;
+
+    bool state_read_meta(llama_io_read_i & io, uint32_t cell_count, llama_seq_id dest_seq_id = -1);
+    bool state_read_data(llama_io_read_i & io, uint32_t cell_count);
+};
+
+class llama_kv_cache_unified_context : public llama_memory_context_i {
+public:
+    // some shorthands
+    using ubatch_heads = llama_kv_cache_unified::ubatch_heads;
+    using defrag_info  = llama_kv_cache_unified::defrag_info;
+
+    // used for errors
+    llama_kv_cache_unified_context(llama_memory_status status);
+
+    // used to create a full-cache context
+    llama_kv_cache_unified_context(
+            llama_kv_cache_unified * kv);
+
+    // used to create an update context
+    llama_kv_cache_unified_context(
+            llama_kv_cache_unified * kv,
+            llama_context * lctx,
+            bool do_shift,
+            defrag_info dinfo);
+
+    // used to create a batch procesing context from a batch
+    llama_kv_cache_unified_context(
+            llama_kv_cache_unified * kv,
+            ubatch_heads heads,
+            std::vector<llama_ubatch> ubatches);
+
+    virtual ~llama_kv_cache_unified_context();
+
+    //
+    // llama_memory_context_i
+    //
+
+    bool next()  override;
+    bool apply() override;
+
+    llama_memory_status  get_status() const override;
+    const llama_ubatch & get_ubatch() const override;
+
+    //
+    // llama_kv_cache_unified_context specific API
+    //
+
+    uint32_t get_n_kv() const;
+
+    // get views of the current state of the cache
+    lm_ggml_tensor * get_k(lm_ggml_context * ctx, int32_t il) const;
+    lm_ggml_tensor * get_v(lm_ggml_context * ctx, int32_t il) const;
+
+    // store k_cur and v_cur in the cache based on the provided head location
+    lm_ggml_tensor * cpy_k(lm_ggml_context * ctx, lm_ggml_tensor * k_cur, int32_t il) const;
+    lm_ggml_tensor * cpy_v(lm_ggml_context * ctx, lm_ggml_tensor * v_cur, int32_t il) const;
+
+    void set_input_k_shift(lm_ggml_tensor * dst) const;
+
+    void set_input_kq_mask   (lm_ggml_tensor * dst, const llama_ubatch * ubatch, bool causal_attn) const;
+    void set_input_pos_bucket(lm_ggml_tensor * dst, const llama_ubatch * ubatch) const;
+
+private:
+    llama_memory_status status;
+
+    llama_kv_cache_unified * kv;
+    llama_context * lctx;
+
+    //
+    // update context
+    //
+
+    bool do_shift = false;
+
+    defrag_info dinfo;
+
+    //
+    // batch processing context
+    //
+
+    // the index of the next ubatch to process
+    size_t i_next = 0;
+
+    ubatch_heads heads;
+
+    std::vector<llama_ubatch> ubatches;
+
+    //
+    // data needed for building the compute graph for the current ubatch:
+    //
+
+    // a heuristic, to avoid attending the full cache if it is not yet utilized
+    // as the cache gets filled, the benefit from this heuristic disappears
+    int32_t n_kv;
+
+    // the beginning of the current slot in which the ubatch will be inserted
+    int32_t head;
+};