llama.cpp/src/llama-memory-recurrent.cpp
Gabe Goodhart edc4a29eff
memory : Hybrid recurrent cache (#13979)
* feat: Add llama_model_is_hybrid API call

Also, split llama_model_is_recurrent into llm_arch_is_recurrent in
llama-arch with llama_model_is_recurrent delegating to
llm_arch_is_recurrent. The same split is done for hybird. This is needed
because there are places where the llama_model has not yet been initialized
but we need to check if the model is recurrent (specifically for the
per-layer recurrent check array in hparams).

Branch: GraniteFour

Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>

* feat: Add c++ side constants for attention layer indices hparam

Branch: GraniteFour

* feat: Add support for distinguishing recurrent vs non-recurrent layers in hparams

Branch: GraniteFour

Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>

* feat: Auto-fill hparams.recurrent_layer_arr based on whether the model is recurrent

Branch: GraniteFour

Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>

* refactor: rename *_is_hybrid -> *_is_hybrid_recurrent

The implementation of the hybrid cache intentionally does not specify the
types of the child caches, so there was a naming mismatch with these
predicate functions that used "hybrid" to imply "hybrid recurrent."

Branch: HybridCache

Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>

* feat: Add layer filter to recurrent cache

Branch: HybridCache

Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>

* fix: Use per-layer sizing everywhere in kv caches

Branch: GraniteFour

Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>

* feat: First pass at llama_kv_cache_hybrid_recurrent

This follows the pattern in iswa where the two child caches are held
explicitly to support the case where a model requires a single attention
cache and a single recurrent cache where each layer uses exactly one of the
caches.

This is a rewrite of the more generic approach in the original hybrid cache
PR: https://github.com/ggml-org/llama.cpp/pull/13276

Branch: HybridRecurrentCache

Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>

* feat: Construct hybrid recurrent cache for hybrid recurrent models

This includes a refactor of the create_memory logic to avoid needing to use
the arch enum explicitly unless a model needs explicit cache instantiation
logic beyond the standard logic for recurrent, hybrid, unified, and iswa.

Branch: HybridRecurrentCache

Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>

* fix: Fix wrong bool condition for split equal in hybrid cache

Branch: HybridRecurrentCache

Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>

* fix: Fix shift logic to defer to unified cache

Branch: HybridRecurrentCache

Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>

* feat: Support hybrid recurrent in llama-graph

NOTE: I intentionally did not add support for s_mask since it will be going
away soon

Branch: HybridRecurrentCache

Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>

* fix: Fix logic for initializing inputs and attn layers for hybrid caches

Branch: GraniteFour

Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>

* fix: Update recurrent cache for changes to remove intermediate kv_cache interface

Branch: HybridRecurrentCache

Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>

* fix: Fix status for init_update sig for recurrent cache state

Branch: GraniteFour

Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>

* fix: Add missing padding to n_ctx for hybrid cache construction

Branch: GraniteFour

Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>

* fix: Update clear signature for data argument after rebase

Branch: HybridRecurrentCache

Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>

* fix: Remove errant virtual destructor leftover from previous impl attempt

Branch: HybridRecurrentCache

Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>

* fix: Use per-layer n_embd_k/v_s calls for mamba (1) layers

Branch: HybridRecurrentCache

Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>

* refactor: Remove n_embd_k/v_s from unified cache

No longer needed now that unified isn't also supporting recurrent

https://github.com/ggml-org/llama.cpp/pull/13979#discussion_r2140761069

Branch: HybridRecurrentCache

* refactor: Remove layer index from n_embd_k/v_s

Now that it's not used at all in the unified cache, we don't need to use
the layer index to zero it out for attention layers.

Branch: HybridRecurrentCache

Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>

* refactor: Remove n_embd_k/v_gqa from recurrent cache

This is no longer needed now that there are separate implementations

https://github.com/ggml-org/llama.cpp/pull/13979#discussion_r2140825128

Branch: HybridRecurrentCache

Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>

* feat: Allow custom layer filters for hybrid recurrent

This should help support architectures like Falcon H1 where there is
overlap between layers that need attention and recurrent caches.

https://github.com/ggml-org/llama.cpp/pull/13979#discussion_r2140748922

Branch: HybridRecurrentCache

Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>

* fix: Remove logits_all after rebase

Branch: HybridRecurrentCache

Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>

* fix: Remove llama_model_is_hybrid_Recurrent public API

https://github.com/ggml-org/llama.cpp/pull/13979#discussion_r2141728423

Branch: HybridRecurrentCache

Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>

* refactor: Use llama_memory_state_ptr for child states in hybrid memory state

Branch: HybridRecurrentCache

Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>

* feat: Overhaul build_recurrent_state / build_inp_s_copy to match attention pattern

https://github.com/ggml-org/llama.cpp/pull/13979/files#r2141701738

This is a big overhaul to bring consistency between how inputs and per-
layer components are created for attention layers and recurrent layers. The
main changes are:

- Rename class llm_graph_input_s_copy -> llm_graph_input_rs
- Add a corresponding llm_graph_input_rs_hybrid_recurrent
- Rename build_inp_s_copy -> build_rs_inp_recurrent
- Add a corresponding build_rs_inp_hybrid_recurrent
- Rename build_recurrent_state -> build_rs to match build_attn w/
llm_graph_input_rs android-build AUTHORS bamba-9b-2.2T.gguf bamba-9b-2.2T.q4_k_m.gguf broken.log build build-rel build-xcframework.sh build.android build.android.bak ci cmake CMakeLists.txt CMakePresets.json CODEOWNERS common common.o CONTRIBUTING.md convert_hf_to_gguf_update.py convert_hf_to_gguf.py convert_llama_ggml_to_gguf.py convert_lora_to_gguf.py debug.log docs examples flake.lock flake.nix ggml ggml-alloc.o ggml-backend.o ggml-metal.o ggml-model-BF16.gguf ggml-model-Q4_K_M.gguf ggml-quants.o ggml.o gguf-py grammar-parser.o grammars include LICENSE licenses llama.log llama.o llamacpp_trace.log main.log Makefile media models mypy.ini pocs poetry.lock prompts pyproject.toml pyrightconfig.json q4_k_m_boot.log q8_0_boot.log quant.log quant2.log README.md requirements requirements.txt sampling.o scripts SECURITY.md src test-grammar-output.tmp test-json-schema-input.tmp tests tools vendor working.log as the first input
- Add a corresponding overload of build_rs w/
llm_graph_input_rs_hybrid_recurrent android-build AUTHORS bamba-9b-2.2T.gguf bamba-9b-2.2T.q4_k_m.gguf broken.log build build-rel build-xcframework.sh build.android build.android.bak ci cmake CMakeLists.txt CMakePresets.json CODEOWNERS common common.o CONTRIBUTING.md convert_hf_to_gguf_update.py convert_hf_to_gguf.py convert_llama_ggml_to_gguf.py convert_lora_to_gguf.py debug.log docs examples flake.lock flake.nix ggml ggml-alloc.o ggml-backend.o ggml-metal.o ggml-model-BF16.gguf ggml-model-Q4_K_M.gguf ggml-quants.o ggml.o gguf-py grammar-parser.o grammars include LICENSE licenses llama.log llama.o llamacpp_trace.log main.log Makefile media models mypy.ini pocs poetry.lock prompts pyproject.toml pyrightconfig.json q4_k_m_boot.log q8_0_boot.log quant.log quant2.log README.md requirements requirements.txt sampling.o scripts SECURITY.md src test-grammar-output.tmp test-json-schema-input.tmp tests tools vendor working.log as the first input
- Add a llm_graph_input_attn_kv_hybrid_recurrent analogous to
llm_graph_input_attn_kv_unified
- Add a build_attn override that takes
llm_graph_input_attn_kv_hybrid_recurrent android-build AUTHORS bamba-9b-2.2T.gguf bamba-9b-2.2T.q4_k_m.gguf broken.log build build-rel build-xcframework.sh build.android build.android.bak ci cmake CMakeLists.txt CMakePresets.json CODEOWNERS common common.o CONTRIBUTING.md convert_hf_to_gguf_update.py convert_hf_to_gguf.py convert_llama_ggml_to_gguf.py convert_lora_to_gguf.py debug.log docs examples flake.lock flake.nix ggml ggml-alloc.o ggml-backend.o ggml-metal.o ggml-model-BF16.gguf ggml-model-Q4_K_M.gguf ggml-quants.o ggml.o gguf-py grammar-parser.o grammars include LICENSE licenses llama.log llama.o llamacpp_trace.log main.log Makefile media models mypy.ini pocs poetry.lock prompts pyproject.toml pyrightconfig.json q4_k_m_boot.log q8_0_boot.log quant.log quant2.log README.md requirements requirements.txt sampling.o scripts SECURITY.md src test-grammar-output.tmp test-json-schema-input.tmp tests tools vendor working.log as the first input

This makes the two paradigms fully consistent. The main drawback is the
code duplication in the build_attn and build_rs implementations where the
only difference between implementations is how they cast the memory state.

Branch: HybridRecurrentCache

Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>

* fix: Fix resize vs reserve and skip null tensors in size computation

https://github.com/ggml-org/llama.cpp/pull/13979/files#r2149469788

Branch: HybridRecurrentCache

Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>
Co-Authored-By: @younesbelkada

* fix: Fix initialization of child states

Since initially writing this PR, the logic in the child state types changed
such that using the "init full" signature and keeping the ubatches on the
parent struct no longer worked.

Branch: HybridRecurrentCache

Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>

* refactor: Use a common build_recurrent_state method that is cache-agnostic

This reduces the code duplication between the different build_rs impls and
also retains a similar signature to the previous build_recurrent_state
method while standardizing on the input-dispatched build_rs implementation.

Branch: HybridRecurrentCache

Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>

* recurrent : rework graph inputs + add TODOs

ggml-ci

* refactor: Make status and child states const in hybrid and iswa

Branch: HybridRecurrentCache

Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>

* refactor: Rename llama_kv_cache_[recurrent|hybrid_recurrent] to remove kv cache

This removes the notion of "kv" from the interface names for these memory
types. There are still many references to kv in the implementation of the
recurrent memory which will need further adjustment.

Branch: HybridRecurrentCache

Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>

* refactor!: Rename all k/v related values for recurrent/hybrid to r/s

Anywhere that "kv_<state|cell|size|etc>" is used, I've used the more
generic "mem_" prefix. The specifics of "k" (key) translate to "r"
(recurrent state) and "v" (value) translate to "s" (state-space embedding
states).

Branch: HybridRecurrentCache

Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>

* refacor: _recurrent -> _recr for brevity

It just _happens_ to have the same number of letters as _attn!

Branch: HybridRecurrentCache

Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>

* style: Fix spacing for ref

Branch: HybridRecurrentCache

Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>

* refactor: recurrent_layer() -> is_recurrent()

Branch: HybridRecurrentCache

Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>

* style: Fix spacing for size_s_bytes declaration

Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>

---------

Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>
Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
2025-06-19 08:08:14 +03:00

1116 lines
36 KiB
C++

#include "llama-memory-recurrent.h"
#include "llama-impl.h"
#include "llama-io.h"
#include "llama-batch.h"
#include "llama-model.h"
#include <algorithm>
#include <cassert>
#include <limits>
#include <map>
#include <stdexcept>
//
// llama_memory_recurrent
//
llama_memory_recurrent::llama_memory_recurrent(
const llama_model & model,
layer_filter_cb && filter,
ggml_type type_r,
ggml_type type_s,
bool offload,
uint32_t mem_size,
uint32_t n_seq_max) : hparams(model.hparams), n_seq_max(n_seq_max) {
const int32_t n_layer = hparams.n_layer;
LLAMA_LOG_INFO("%s: mem_size = %u, n_seq_max = %u, type_r = '%s', type_s = '%s', n_layer = %d\n",
__func__, mem_size, n_seq_max, ggml_type_name(type_r), ggml_type_name(type_s), n_layer);
head = 0;
size = mem_size;
used = 0;
cells.clear();
cells.resize(mem_size);
// create a context for each buffer type
std::map<ggml_backend_buffer_type_t, ggml_context *> ctx_map;
auto ctx_for_buft = [&](ggml_backend_buffer_type_t buft) -> ggml_context * {
auto it = ctx_map.find(buft);
if (it == ctx_map.end()) {
ggml_init_params params = {
/*.mem_size =*/ size_t(2u*n_layer*ggml_tensor_overhead()),
/*.mem_buffer =*/ NULL,
/*.no_alloc =*/ true,
};
ggml_context * ctx = ggml_init(params);
if (!ctx) {
return nullptr;
}
ctx_map[buft] = ctx;
ctxs.emplace_back(ctx);
return ctx;
}
return it->second;
};
r_l.resize(n_layer);
s_l.resize(n_layer);
for (int i = 0; i < n_layer; i++) {
if (filter && !filter(i)) {
LLAMA_LOG_DEBUG("%s: layer %3d: skipped\n", __func__, i);
continue;
}
const char * dev_name = "CPU";
ggml_backend_buffer_type_t buft = ggml_backend_cpu_buffer_type();
if (offload) {
auto * dev = model.dev_layer(i);
buft = ggml_backend_dev_buffer_type(dev);
dev_name = ggml_backend_dev_name(dev);
}
LLAMA_LOG_DEBUG("%s, layer %3d: dev = %s\n", __func__, i, dev_name);
ggml_context * ctx = ctx_for_buft(buft);
if (!ctx) {
throw std::runtime_error("failed to create ggml context for kv cache");
}
ggml_tensor * r = ggml_new_tensor_1d(ctx, type_r, hparams.n_embd_r()*mem_size);
ggml_tensor * s = ggml_new_tensor_1d(ctx, type_s, hparams.n_embd_s()*mem_size);
ggml_format_name(r, "cache_r_l%d", i);
ggml_format_name(s, "cache_s_l%d", i);
r_l[i] = r;
s_l[i] = s;
}
// allocate tensors and initialize the buffers to avoid NaNs in the padding
for (auto it : ctx_map) {
auto * buft = it.first;
auto * ctx = it.second;
ggml_backend_buffer_t buf = ggml_backend_alloc_ctx_tensors_from_buft(ctx, buft);
if (!buf) {
throw std::runtime_error("failed to allocate buffer for kv cache");
}
ggml_backend_buffer_clear(buf, 0);
LLAMA_LOG_INFO("%s: %10s KV buffer size = %8.2f MiB\n", __func__, ggml_backend_buffer_name(buf), ggml_backend_buffer_get_size(buf)/1024.0/1024.0);
bufs.emplace_back(buf);
}
{
const size_t memory_size_r = size_r_bytes();
const size_t memory_size_s = size_s_bytes();
LLAMA_LOG_INFO("%s: KV self size = %7.2f MiB, R (%s): %7.2f MiB, S (%s): %7.2f MiB\n", __func__,
(float)(memory_size_r + memory_size_s) / (1024.0f * 1024.0f),
ggml_type_name(type_r), (float)memory_size_r / (1024.0f * 1024.0f),
ggml_type_name(type_s), (float)memory_size_s / (1024.0f * 1024.0f));
}
}
void llama_memory_recurrent::clear(bool data) {
for (int32_t i = 0; i < (int32_t) size; ++i) {
cells[i].pos = -1;
cells[i].seq_id.clear();
cells[i].src = -1;
cells[i].tail = -1;
}
head = 0;
used = 0;
if (data) {
for (auto & buf : bufs) {
ggml_backend_buffer_clear(buf.get(), 0);
}
}
}
bool llama_memory_recurrent::seq_rm(llama_seq_id seq_id, llama_pos p0, llama_pos p1) {
uint32_t new_head = size;
if (p0 < 0) {
p0 = 0;
}
if (p1 < 0) {
p1 = std::numeric_limits<llama_pos>::max();
}
// models like Mamba or RWKV can't have a state partially erased
if (seq_id >= (int64_t) size) {
// could be fatal
return false;
}
if (0 <= seq_id) {
int32_t & tail_id = cells[seq_id].tail;
if (tail_id >= 0) {
const auto & cell = cells[tail_id];
// partial intersection is invalid
if ((0 < p0 && p0 <= cell.pos) || (0 < p1 && p1 <= cell.pos)) {
return false;
}
// invalidate tails which will be cleared
if (p0 <= cell.pos && cell.pos < p1) {
tail_id = -1;
}
}
} else {
// seq_id is negative, then the range should include everything or nothing
if (p0 != p1 && (p0 != 0 || p1 != std::numeric_limits<llama_pos>::max())) {
return false;
}
}
for (uint32_t i = 0; i < size; ++i) {
if (cells[i].pos >= p0 && cells[i].pos < p1) {
if (seq_id < 0) {
cells[i].seq_id.clear();
} else if (cells[i].has_seq_id(seq_id)) {
cells[i].seq_id.erase(seq_id);
} else {
continue;
}
if (cells[i].is_empty()) {
// keep count of the number of used cells
if (cells[i].pos >= 0) {
used--;
}
cells[i].pos = -1;
cells[i].src = -1;
if (new_head == size) {
new_head = i;
}
}
}
}
// If we freed up a slot, set head to it so searching can start there.
if (new_head != size && new_head < head) {
head = new_head;
}
return true;
}
void llama_memory_recurrent::seq_cp(llama_seq_id seq_id_src, llama_seq_id seq_id_dst, llama_pos p0, llama_pos p1) {
if (seq_id_src == seq_id_dst) {
return;
}
if (p0 < 0) {
p0 = 0;
}
if (p1 < 0) {
p1 = std::numeric_limits<llama_pos>::max();
}
if ((uint32_t) seq_id_dst < size && (uint32_t) seq_id_src < size) {
auto & tail_src = cells[seq_id_src];
auto & tail_dst = cells[seq_id_dst];
if (tail_dst.tail >= 0) {
// clear destination seq_id if it wasn't empty
auto & cell_dst = cells[tail_dst.tail];
cell_dst.seq_id.erase(seq_id_dst);
tail_dst.tail = -1;
if (cell_dst.seq_id.empty()) {
cell_dst.pos = -1;
cell_dst.src = -1;
used -= 1;
}
}
if (tail_src.tail >= 0) {
auto & cell_src = cells[tail_src.tail];
cell_src.seq_id.insert(seq_id_dst);
tail_dst.tail = tail_src.tail;
}
}
}
void llama_memory_recurrent::seq_keep(llama_seq_id seq_id) {
uint32_t new_head = size;
for (uint32_t i = 0; i < size; ++i) {
if ((llama_seq_id) i != seq_id) {
cells[i].tail = -1;
}
if (!cells[i].has_seq_id(seq_id)) {
if (cells[i].pos >= 0) {
used--;
}
cells[i].pos = -1;
cells[i].src = -1;
cells[i].seq_id.clear();
if (new_head == size){
new_head = i;
}
} else {
cells[i].seq_id.clear();
cells[i].seq_id.insert(seq_id);
}
}
// If we freed up a slot, set head to it so searching can start there.
if (new_head != size && new_head < head) {
head = new_head;
}
}
void llama_memory_recurrent::seq_add(llama_seq_id seq_id, llama_pos p0, llama_pos p1, llama_pos shift) {
if (shift == 0) {
return;
}
if (p0 < 0) {
p0 = 0;
}
if (p1 < 0) {
p1 = std::numeric_limits<llama_pos>::max();
}
// If there is no range then return early to avoid looping over the
if (p0 == p1) {
return;
}
// for Mamba-like or RWKV models, only the pos needs to be shifted
if (0 <= seq_id && seq_id < (int64_t) size) {
const int32_t tail_id = cells[seq_id].tail;
if (tail_id >= 0) {
auto & cell = cells[tail_id];
if (cell.has_seq_id(seq_id) && p0 <= cell.pos && cell.pos < p1) {
cell.pos += shift;
}
}
}
}
void llama_memory_recurrent::seq_div(llama_seq_id seq_id, llama_pos p0, llama_pos p1, int d) {
if (d == 1) {
return;
}
if (p0 < 0) {
p0 = 0;
}
if (p1 < 0) {
p1 = std::numeric_limits<llama_pos>::max();
}
// If there is no range then return early to avoid looping over the cache.
if (p0 == p1) {
return;
}
// for Mamba-like or RWKV models, only the pos needs to be changed
if (0 <= seq_id && seq_id < (int64_t) size) {
const int32_t tail_id = cells[seq_id].tail;
if (tail_id >= 0) {
auto & cell = cells[tail_id];
if (cell.has_seq_id(seq_id) && p0 <= cell.pos && cell.pos < p1) {
cell.pos /= d;
}
}
}
}
llama_pos llama_memory_recurrent::seq_pos_min(llama_seq_id seq_id) const {
llama_pos result = std::numeric_limits<llama_pos>::max();
for (uint32_t i = 0; i < size; ++i) {
if (cells[i].has_seq_id(seq_id)) {
result = std::min(result, cells[i].pos);
}
}
if (result == std::numeric_limits<llama_pos>::max()) {
result = -1;
}
return result;
}
llama_pos llama_memory_recurrent::seq_pos_max(llama_seq_id seq_id) const {
llama_pos result = -1;
for (uint32_t i = 0; i < size; ++i) {
if (cells[i].has_seq_id(seq_id)) {
result = std::max(result, cells[i].pos);
}
}
return result;
}
llama_memory_state_ptr llama_memory_recurrent::init_batch(const llama_batch & batch, uint32_t n_ubatch, bool embd_all) {
auto sbatch = llama_sbatch(batch, hparams.n_embd, false);
std::vector<llama_ubatch> ubatches;
while (sbatch.n_tokens > 0) {
llama_ubatch ubatch;
if (embd_all) {
// if all tokens are output, split by sequence
ubatch = sbatch.split_seq(n_ubatch);
} else {
ubatch = sbatch.split_equal(n_ubatch);
}
ubatches.push_back(ubatch);
}
if (!prepare(ubatches)) {
return std::make_unique<llama_memory_recurrent_state>(LLAMA_MEMORY_STATUS_FAILED_PREPARE);
}
return std::make_unique<llama_memory_recurrent_state>(this, std::move(sbatch), std::move(ubatches));
}
llama_memory_state_ptr llama_memory_recurrent::init_full() {
return std::make_unique<llama_memory_recurrent_state>(this);
}
llama_memory_state_ptr llama_memory_recurrent::init_update(llama_context * lctx, bool optimize) {
GGML_UNUSED(lctx);
GGML_UNUSED(optimize);
return std::make_unique<llama_memory_recurrent_state>(LLAMA_MEMORY_STATUS_NO_UPDATE);
}
bool llama_memory_recurrent::prepare(const std::vector<llama_ubatch> & ubatches) {
// simply remember the full state because it is very small for this type of cache
// TODO: optimize
auto org_cells = cells;
auto org_used = used;
auto org_head = head;
bool success = true;
for (const auto & ubatch : ubatches) {
if (!find_slot(ubatch)) {
success = false;
break;
}
}
// restore the original state
cells = std::move(org_cells);
used = org_used;
head = org_head;
return success;
}
bool llama_memory_recurrent::find_slot(const llama_ubatch & ubatch) {
const uint32_t n_seqs = ubatch.n_seqs;
const uint32_t n_seq_tokens = ubatch.n_seq_tokens;
// if we have enough unused cells before the current head ->
// better to start searching from the beginning of the cache, hoping to fill it
if (head > used + 2*n_seqs) {
head = 0;
}
// For recurrent state architectures (like Mamba or RWKV),
// each cache cell can store the state for a whole sequence.
// A slot should be always be contiguous.
// can only process batches with an equal number of new tokens in each sequence
GGML_ASSERT(ubatch.equal_seqs);
int32_t min = size - 1;
int32_t max = 0;
// everything should fit if all seq_ids are smaller than the max
for (uint32_t s = 0; s < n_seqs; ++s) {
const uint32_t n_seq_id = ubatch.n_seq_id[s];
for (uint32_t j = 0; j < n_seq_id; ++j) {
const llama_seq_id seq_id = ubatch.seq_id[s][j];
if (seq_id < 0 || (uint32_t) seq_id >= size) {
// too big seq_id
// TODO: would it be possible to resize the cache instead?
LLAMA_LOG_ERROR("%s: seq_id=%d >= n_seq_max=%u Try using a bigger --parallel value\n", __func__, seq_id, n_seq_max);
return false;
}
if (j > 0) {
auto & seq = cells[seq_id];
if (seq.tail >= 0) {
auto & cell = cells[seq.tail];
// clear cells from seq_ids that become shared
// (should not normally happen, but let's handle it anyway)
cell.seq_id.erase(seq_id);
seq.tail = -1;
if (cell.seq_id.empty()) {
cell.pos = -1;
cell.src = -1;
used -= 1;
}
}
}
}
}
#ifndef NDEBUG
{
std::vector<int32_t> tails_verif;
tails_verif.assign(size, -1);
for (uint32_t i = 0; i < size; ++i) {
auto & cell = cells[i];
for (llama_seq_id seq_id : cell.seq_id) {
if (tails_verif[seq_id] != -1) {
LLAMA_LOG_ERROR("%s: duplicate tail for seq_id %d in cell %d and %d\n", __func__, seq_id, i, tails_verif[seq_id]);
}
tails_verif[seq_id] = i;
}
}
for (uint32_t i = 0; i < size; ++i) {
if (tails_verif[i] != cells[i].tail) {
LLAMA_LOG_ERROR("%s: wrong tail for seq_id %d, (%d instead of %d)\n", __func__, i, cells[i].tail, tails_verif[i]);
}
}
}
#endif
// find next empty cell
uint32_t next_empty_cell = head;
for (uint32_t i = 0; i < size; ++i) {
if (next_empty_cell >= size) { next_empty_cell -= size; }
auto & cell = cells[next_empty_cell];
if (cell.is_empty()) { break; }
next_empty_cell += 1;
}
// find usable cell range
for (uint32_t s = 0; s < n_seqs; ++s) {
const llama_seq_id seq_id = ubatch.seq_id[s][0];
auto & seq_meta = cells[seq_id];
bool has_cell = false;
if (seq_meta.tail >= 0) {
auto & cell = cells[seq_meta.tail];
GGML_ASSERT(cell.has_seq_id(seq_id));
// does this seq_id "own" the cell?
if (cell.seq_id.size() == 1) { has_cell = true; }
}
if (!has_cell) {
auto & empty_cell = cells[next_empty_cell];
GGML_ASSERT(empty_cell.is_empty());
// copy old tail into the empty cell
if (seq_meta.tail >= 0) {
auto & orig_cell = cells[seq_meta.tail];
empty_cell.pos = orig_cell.pos;
empty_cell.src = orig_cell.src;
orig_cell.seq_id.erase(seq_id);
empty_cell.seq_id.insert(seq_id); // will be overwritten
GGML_ASSERT(!orig_cell.is_empty()); // has at least one remaining seq_id
}
seq_meta.tail = next_empty_cell;
// find next empty cell
if (s + 1 < n_seqs) {
for (uint32_t i = 0; i < size; ++i) {
next_empty_cell += 1;
if (next_empty_cell >= size) { next_empty_cell -= size; }
auto & cell = cells[next_empty_cell];
if (cell.is_empty()) { break; }
}
}
}
if (min > seq_meta.tail) { min = seq_meta.tail; }
if (max < seq_meta.tail) { max = seq_meta.tail; }
}
// gather and re-order
for (uint32_t s = 0; s < n_seqs; ++s) {
const int32_t dst_id = s + min;
const int32_t src_id = cells[ubatch.seq_id[s][0]].tail;
if (dst_id != src_id) {
auto & dst_cell = cells[dst_id];
auto & src_cell = cells[src_id];
std::swap(dst_cell.pos, src_cell.pos);
std::swap(dst_cell.src, src_cell.src);
std::swap(dst_cell.seq_id, src_cell.seq_id);
// swap tails
for (uint32_t i = 0; i < size; ++i) {
int32_t & tail = cells[i].tail;
if (tail == src_id) {
tail = dst_id;
} else if (tail == dst_id) {
tail = src_id;
}
}
}
}
// update the pos of the used seqs
for (uint32_t s = 0; s < n_seqs; ++s) {
const llama_pos last_pos = ubatch.pos[n_seq_tokens * s + n_seq_tokens - 1];
const int32_t cell_id = s + min;
auto & cell = cells[cell_id];
if (cell.pos >= 0 && last_pos != cell.pos + (llama_pos) n_seq_tokens) {
// What should happen when the pos backtracks or skips a value?
// Clearing the state mid-batch would require special-casing which isn't done.
LLAMA_LOG_WARN("%s: non-consecutive token position %d after %d for sequence %d with %u new tokens\n",
__func__, last_pos, cell.pos, ubatch.seq_id[s][0], n_seq_tokens);
}
cell.pos = last_pos;
cell.seq_id.clear();
for (int32_t j = 0; j < ubatch.n_seq_id[s]; ++j) {
const llama_seq_id seq_id = ubatch.seq_id[s][j];
cell.seq_id.insert(seq_id);
cells[seq_id].tail = cell_id;
}
}
// Find first cell without src refs, to use as the zero-ed state
{
// TODO: bake-in src refcounts in the cell metadata
std::vector<int32_t> refcounts(size, 0);
for (size_t i = 0; i < size; ++i) {
const int32_t src = cells[i].src;
if (src >= 0) {
refcounts[src] += 1;
}
}
rs_z = -1;
for (int i = min; i <= max; ++i) {
if (refcounts[i] == 0) {
rs_z = i;
break;
}
}
for (int i = min; i <= max; ++i) {
if (cells[i].src < 0) {
GGML_ASSERT(rs_z >= 0);
cells[i].src0 = rs_z;
} else {
// Stage the source ids for all used cells to allow correct seq_* behavior
// and still make these values available when setting the inputs
cells[i].src0 = cells[i].src;
}
cells[i].src = i; // avoid moving or clearing twice
}
}
// allow getting the range of used cells, from head to head + n
head = min;
n = max - min + 1;
used = std::count_if(cells.begin(), cells.end(),
[](const mem_cell & cell){ return !cell.is_empty(); });
// sanity check
return n >= n_seqs;
}
bool llama_memory_recurrent::get_can_shift() const {
// shifting the pos is trivial for recurrent models
return true;
}
size_t llama_memory_recurrent::total_size() const {
size_t size = 0;
for (const auto & buf : bufs) {
size += ggml_backend_buffer_get_size(buf.get());
}
return size;
}
size_t llama_memory_recurrent::size_r_bytes() const {
size_t size_r_bytes = 0;
for (const auto & r : r_l) {
if (r != nullptr) {
size_r_bytes += ggml_nbytes(r);
}
}
return size_r_bytes;
}
size_t llama_memory_recurrent::size_s_bytes() const {
size_t size_s_bytes = 0;
for (const auto & s : s_l) {
if (s != nullptr) {
size_s_bytes += ggml_nbytes(s);
}
}
return size_s_bytes;
}
void llama_memory_recurrent::state_write(llama_io_write_i & io, llama_seq_id seq_id) const {
std::vector<std::pair<uint32_t, uint32_t>> cell_ranges; // ranges, from inclusive, to exclusive
uint32_t cell_count = 0;
// Count the number of cells with the specified seq_id
// Find all the ranges of cells with this seq id (or all, when -1)
uint32_t cell_range_begin = size;
for (uint32_t i = 0; i < size; ++i) {
const auto & cell = cells[i];
if ((seq_id == -1 && !cell.is_empty()) || cell.has_seq_id(seq_id)) {
++cell_count;
if (cell_range_begin == size) {
cell_range_begin = i;
}
} else {
if (cell_range_begin != size) {
cell_ranges.emplace_back(cell_range_begin, i);
cell_range_begin = size;
}
}
}
if (cell_range_begin != size) {
cell_ranges.emplace_back(cell_range_begin, size);
}
// DEBUG CHECK: Sum of cell counts in ranges should equal the total cell count
uint32_t cell_count_check = 0;
for (const auto & range : cell_ranges) {
cell_count_check += range.second - range.first;
}
GGML_ASSERT(cell_count == cell_count_check);
io.write(&cell_count, sizeof(cell_count));
state_write_meta(io, cell_ranges, seq_id);
state_write_data(io, cell_ranges);
}
void llama_memory_recurrent::state_read(llama_io_read_i & io, llama_seq_id seq_id) {
uint32_t cell_count;
io.read_to(&cell_count, sizeof(cell_count));
bool res = true;
res = res && state_read_meta(io, cell_count, seq_id);
res = res && state_read_data(io, cell_count);
if (!res) {
if (seq_id == -1) {
clear(true);
} else {
seq_rm(seq_id, -1, -1);
}
throw std::runtime_error("failed to restore kv cache");
}
}
void llama_memory_recurrent::state_write_meta(llama_io_write_i & io, const std::vector<std::pair<uint32_t, uint32_t>> & cell_ranges, llama_seq_id seq_id) const {
for (const auto & range : cell_ranges) {
for (uint32_t i = range.first; i < range.second; ++i) {
const auto & cell = cells[i];
const llama_pos pos = cell.pos;
const uint32_t n_seq_id = seq_id == -1 ? cell.seq_id.size() : 0;
io.write(&pos, sizeof(pos));
io.write(&n_seq_id, sizeof(n_seq_id));
if (n_seq_id) {
for (auto seq_id : cell.seq_id) {
io.write(&seq_id, sizeof(seq_id));
}
}
}
}
}
void llama_memory_recurrent::state_write_data(llama_io_write_i & io, const std::vector<std::pair<uint32_t, uint32_t>> & cell_ranges) const {
const uint32_t s_trans = 0;
const uint32_t n_layer = hparams.n_layer;
io.write(&s_trans, sizeof(s_trans));
io.write(&n_layer, sizeof(n_layer));
std::vector<uint8_t> tmp_buf;
// Iterate and write all the keys first, each row is a cell
// Get whole range at a time
for (uint32_t il = 0; il < n_layer; ++il) {
// Write key type
const int32_t r_type_i = (int32_t)r_l[il]->type;
io.write(&r_type_i, sizeof(r_type_i));
// Write row size of key
const uint64_t r_size_row = ggml_row_size(r_l[il]->type, hparams.n_embd_r());
io.write(&r_size_row, sizeof(r_size_row));
// Read each range of cells of k_size length each into tmp_buf and write out
for (const auto & range : cell_ranges) {
const size_t range_size = range.second - range.first;
const size_t buf_size = range_size * r_size_row;
io.write_tensor(r_l[il], range.first * r_size_row, buf_size);
}
}
if (!s_trans) {
for (uint32_t il = 0; il < n_layer; ++il) {
// Write value type
const int32_t s_type_i = (int32_t)s_l[il]->type;
io.write(&s_type_i, sizeof(s_type_i));
// Write row size of value
const uint64_t s_size_row = ggml_row_size(s_l[il]->type, hparams.n_embd_s());
io.write(&s_size_row, sizeof(s_size_row));
// Read each range of cells of s_size length each into tmp_buf and write out
for (const auto & range : cell_ranges) {
const size_t range_size = range.second - range.first;
const size_t buf_size = range_size * s_size_row;
io.write_tensor(s_l[il], range.first * s_size_row, buf_size);
}
}
} else {
// When v is transposed, we also need the element size and get the element ranges from each row
const uint32_t mem_size = size;
for (uint32_t il = 0; il < n_layer; ++il) {
const uint32_t n_embd_s = hparams.n_embd_s();
// Write value type
const int32_t s_type_i = (int32_t)s_l[il]->type;
io.write(&s_type_i, sizeof(s_type_i));
// Write element size
const uint32_t s_size_el = ggml_type_size(s_l[il]->type);
io.write(&s_size_el, sizeof(s_size_el));
// Write GQA embedding size
io.write(&n_embd_s, sizeof(n_embd_s));
// For each row, we get the element values of each cell
for (uint32_t j = 0; j < n_embd_s; ++j) {
// Read each range of cells of v_size_el length each into tmp_buf and write out
for (const auto & range : cell_ranges) {
const size_t range_size = range.second - range.first;
const size_t src_offset = (range.first + j * mem_size) * s_size_el;
const size_t buf_size = range_size * s_size_el;
io.write_tensor(s_l[il], src_offset, buf_size);
}
}
}
}
}
bool llama_memory_recurrent::state_read_meta(llama_io_read_i & io, uint32_t cell_count, llama_seq_id dest_seq_id) {
if (dest_seq_id != -1) {
// single sequence
seq_rm(dest_seq_id, -1, -1);
llama_sbatch sbatch;
llama_ubatch batch = sbatch.reserve_ubatch(cell_count, /* has_embd */ false);
batch.n_tokens = cell_count;
batch.n_seq_tokens = cell_count;
batch.n_seqs = 1;
for (uint32_t i = 0; i < cell_count; ++i) {
llama_pos pos;
uint32_t n_seq_id;
io.read_to(&pos, sizeof(pos));
io.read_to(&n_seq_id, sizeof(n_seq_id));
if (n_seq_id != 0) {
LLAMA_LOG_ERROR("%s: invalid seq_id-agnostic kv cell\n", __func__);
return false;
}
batch.pos[i] = pos;
}
batch.n_seq_id[0] = 1;
batch.seq_id[0] = &dest_seq_id;
if (!find_slot(batch)) {
LLAMA_LOG_ERROR("%s: failed to find available cells in kv cache\n", __func__);
return false;
}
// DEBUG CHECK: kv.head should be our first cell, kv.head + cell_count - 1 should be our last cell (verify seq_id and pos values)
// Assume that this is one contiguous block of cells
GGML_ASSERT(head + cell_count <= size);
GGML_ASSERT(cells[head].pos == batch.pos[0]);
GGML_ASSERT(cells[head + cell_count - 1].pos == batch.pos[cell_count - 1]);
GGML_ASSERT(cells[head].has_seq_id(dest_seq_id));
GGML_ASSERT(cells[head + cell_count - 1].has_seq_id(dest_seq_id));
} else {
// whole KV cache restore
if (cell_count > size) {
LLAMA_LOG_ERROR("%s: not enough cells in kv cache\n", __func__);
return false;
}
clear(true);
for (uint32_t i = 0; i < cell_count; ++i) {
auto & cell = cells[i];
llama_pos pos;
uint32_t n_seq_id;
io.read_to(&pos, sizeof(pos));
io.read_to(&n_seq_id, sizeof(n_seq_id));
cell.pos = pos;
for (uint32_t j = 0; j < n_seq_id; ++j) {
llama_seq_id seq_id;
io.read_to(&seq_id, sizeof(seq_id));
// TODO: llama_memory_recurrent should have a notion of max sequences
//if (seq_id < 0 || (uint32_t) seq_id >= llama_n_seq_max(ctx)) {
if (seq_id < 0) {
//LLAMA_LOG_ERROR("%s: invalid seq_id, %d is out of range [0, %u)\n", __func__, seq_id, llama_n_seq_max(ctx));
LLAMA_LOG_ERROR("%s: invalid seq_id, %d is out of range [0, inf)\n", __func__, seq_id);
return false;
}
cell.seq_id.insert(seq_id);
int32_t & tail = cells[seq_id].tail;
if (tail != -1) {
LLAMA_LOG_ERROR("%s: duplicate tail for seq_id %d in cell %d and %d\n", __func__, seq_id, i, tail);
return false;
}
tail = i;
}
}
head = 0;
used = cell_count;
}
for (uint32_t i = 0; i < cell_count; ++i) {
uint32_t cell_id = head + i;
// make sure the recurrent states will keep their restored state
cells[cell_id].src = cell_id;
}
return true;
}
bool llama_memory_recurrent::state_read_data(llama_io_read_i & io, uint32_t cell_count) {
uint32_t s_trans;
uint32_t n_layer;
io.read_to(&s_trans, sizeof(s_trans));
io.read_to(&n_layer, sizeof(n_layer));
if (n_layer != hparams.n_layer) {
LLAMA_LOG_ERROR("%s: mismatched layer count (%u instead of %u)\n", __func__, n_layer, hparams.n_layer);
return false;
}
if (cell_count > size) {
LLAMA_LOG_ERROR("%s: not enough cells in kv cache to restore state (%u > %u)\n", __func__, cell_count, size);
return false;
}
if (false != (bool) s_trans) {
LLAMA_LOG_ERROR("%s: incompatible s transposition\n", __func__);
return false;
}
// For each layer, read the keys for each cell, one row is one cell, read as one contiguous block
for (uint32_t il = 0; il < n_layer; ++il) {
// Read type of key
int32_t r_type_i_ref;
io.read_to(&r_type_i_ref, sizeof(r_type_i_ref));
const int32_t r_type_i = (int32_t) r_l[il]->type;
if (r_type_i != r_type_i_ref) {
LLAMA_LOG_ERROR("%s: mismatched r type (%d != %d, layer %d)\n", __func__, r_type_i, r_type_i_ref, il);
return false;
}
// Read row size of key
uint64_t r_size_row_ref;
io.read_to(&r_size_row_ref, sizeof(r_size_row_ref));
const size_t r_size_row = ggml_row_size(r_l[il]->type, hparams.n_embd_r());
if (r_size_row != r_size_row_ref) {
LLAMA_LOG_ERROR("%s: mismatched r row size (%zu != %zu, layer %d)\n", __func__, r_size_row, (size_t) r_size_row_ref, il);
return false;
}
if (cell_count) {
// Read and set the keys for the whole cell range
ggml_backend_tensor_set(r_l[il], io.read(cell_count * r_size_row), head * r_size_row, cell_count * r_size_row);
}
}
if (!s_trans) {
for (uint32_t il = 0; il < n_layer; ++il) {
// Read type of value
int32_t s_type_i_ref;
io.read_to(&s_type_i_ref, sizeof(s_type_i_ref));
const int32_t s_type_i = (int32_t)s_l[il]->type;
if (s_type_i != s_type_i_ref) {
LLAMA_LOG_ERROR("%s: mismatched s type (%d != %d, layer %d)\n", __func__, s_type_i, s_type_i_ref, il);
return false;
}
// Read row size of value
uint64_t s_size_row_ref;
io.read_to(&s_size_row_ref, sizeof(s_size_row_ref));
const size_t s_size_row = ggml_row_size(s_l[il]->type, hparams.n_embd_s());
if (s_size_row != s_size_row_ref) {
LLAMA_LOG_ERROR("%s: mismatched s row size (%zu != %zu, layer %d)\n", __func__, s_size_row, (size_t) s_size_row_ref, il);
return false;
}
if (cell_count) {
// Read and set the values for the whole cell range
ggml_backend_tensor_set(s_l[il], io.read(cell_count * s_size_row), head * s_size_row, cell_count * s_size_row);
}
}
} else {
// For each layer, read the values for each cell (transposed)
for (uint32_t il = 0; il < n_layer; ++il) {
const uint32_t n_embd_s = hparams.n_embd_s();
// Read type of value
int32_t s_type_i_ref;
io.read_to(&s_type_i_ref, sizeof(s_type_i_ref));
const int32_t s_type_i = (int32_t)s_l[il]->type;
if (s_type_i != s_type_i_ref) {
LLAMA_LOG_ERROR("%s: mismatched s type (%d != %d, layer %d)\n", __func__, s_type_i, s_type_i_ref, il);
return false;
}
// Read element size of value
uint32_t s_size_el_ref;
io.read_to(&s_size_el_ref, sizeof(s_size_el_ref));
const size_t s_size_el = ggml_type_size(s_l[il]->type);
if (s_size_el != s_size_el_ref) {
LLAMA_LOG_ERROR("%s: mismatched s element size (%zu != %zu, layer %d)\n", __func__, s_size_el, (size_t) s_size_el_ref, il);
return false;
}
// Read state embedding size
uint32_t n_embd_s_ref;
io.read_to(&n_embd_s_ref, sizeof(n_embd_s_ref));
if (n_embd_s != n_embd_s_ref) {
LLAMA_LOG_ERROR("%s: mismatched s embedding size (%u != %u, layer %d)\n", __func__, n_embd_s, n_embd_s_ref, il);
return false;
}
if (cell_count) {
// For each row in the transposed matrix, read the values for the whole cell range
for (uint32_t j = 0; j < n_embd_s; ++j) {
const size_t dst_offset = (head + j * size) * s_size_el;
ggml_backend_tensor_set(s_l[il], io.read(cell_count * s_size_el), dst_offset, cell_count * s_size_el);
}
}
}
}
return true;
}
//
// llama_memory_recurrent_state
//
llama_memory_recurrent_state::llama_memory_recurrent_state(llama_memory_status status) : status(status) {}
llama_memory_recurrent_state::llama_memory_recurrent_state(
llama_memory_recurrent * mem) : status(LLAMA_MEMORY_STATUS_SUCCESS), mem(mem), is_full(true) {
}
llama_memory_recurrent_state::llama_memory_recurrent_state(
llama_memory_recurrent * mem,
llama_sbatch sbatch,
std::vector<llama_ubatch> ubatches) : status(LLAMA_MEMORY_STATUS_SUCCESS), mem(mem), sbatch(std::move(sbatch)), ubatches(std::move(ubatches)) {}
llama_memory_recurrent_state::~llama_memory_recurrent_state() = default;
bool llama_memory_recurrent_state::next() {
assert(status == LLAMA_MEMORY_STATUS_SUCCESS);
if (++i_next >= ubatches.size()) {
return false;
}
return true;
}
bool llama_memory_recurrent_state::apply() {
assert(status == LLAMA_MEMORY_STATUS_SUCCESS);
mem->find_slot(ubatches[i_next]);
return true;
}
std::vector<int64_t> & llama_memory_recurrent_state::out_ids() {
assert(status == LLAMA_MEMORY_STATUS_SUCCESS);
return sbatch.out_ids;
}
llama_memory_status llama_memory_recurrent_state::get_status() const {
return status;
}
const llama_ubatch & llama_memory_recurrent_state::get_ubatch() const {
assert(status == LLAMA_MEMORY_STATUS_SUCCESS);
return ubatches[i_next];
}
uint32_t llama_memory_recurrent_state::get_n_rs() const {
return is_full ? mem->size : mem->n;
}
uint32_t llama_memory_recurrent_state::get_head() const {
return is_full ? 0 : mem->head;
}
int32_t llama_memory_recurrent_state::get_rs_z() const {
return is_full ? 0 : mem->rs_z;
}
uint32_t llama_memory_recurrent_state::get_size() const {
return mem->size;
}
ggml_tensor * llama_memory_recurrent_state::get_r_l(int32_t il) const {
return mem->r_l[il];
}
ggml_tensor * llama_memory_recurrent_state::get_s_l(int32_t il) const {
return mem->s_l[il];
}
int32_t llama_memory_recurrent_state::s_copy(int i) const {
return mem->cells[i + mem->head].src0;
}