url/colibri
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Merge pull request #372 from ZacharyZcR/feat/paged-ragged-kv

Browse Source 
cuda: keep paged ragged KV resident across decode steps
This commit is contained in:
Vincenzo Fornaro
2026-07-18 12:48:07 +02:00
committed by GitHub
parent 4fcbea9056 6d3a6168e4
commit 3fed654473
5 changed files with 122 additions and 43 deletions
Download Patch File Download Diff File Expand all files Collapse all files
c/backend_cuda.cu
+98 -29
View File
@@ -8,12 +8,21 @@
#include <cstring> #include <cstring>
#include <mutex> #include <mutex>
struct RaggedKVEntry {
const void *key;
const float *host_l,*host_r;
float *latent,*rope;
int length,capacity,K,R;
};
struct ColiCudaTensor { struct ColiCudaTensor {
void *weights; void *weights;
float *scales; float *scales;
size_t weight_bytes; size_t weight_bytes;
int fmt, I, O, device; int fmt, I, O, device;
int tracked; int tracked;
RaggedKVEntry ragged[512];
int ragged_count;
}; };
typedef struct { typedef struct {
@@ -23,7 +32,7 @@ typedef struct {
size_t x_cap, y_cap, gate_cap, up_cap; size_t x_cap, y_cap, gate_cap, up_cap;
uint8_t *qx; float *qscale; uint8_t *qx; float *qscale;
size_t qx_cap, qscale_cap; size_t qx_cap, qscale_cap;
float *host_x,*host_y; size_t host_x_cap,host_y_cap; float *host_x,*host_y,*host_kv; size_t host_x_cap,host_y_cap,host_kv_cap;
float *aq,*al,*ar,*ac; size_t aq_cap,al_cap,ar_cap,ac_cap; float *aq,*al,*ar,*ac; size_t aq_cap,al_cap,ar_cap,ac_cap;
float *pipe_buf[24]; size_t pipe_cap[24]; /* scratch persistenti del resident pipeline */ float *pipe_buf[24]; size_t pipe_cap[24]; /* scratch persistenti del resident pipeline */
cudaStream_t stream; cudaStream_t stream;
@@ -338,17 +347,17 @@ __global__ static void attention_absorb_batch_kernel(float *ctx,const float *q,
ctx[((size_t)s*H+h)*V+v]=a*(fmt?wscale[row]:1.f);} ctx[((size_t)s*H+h)*V+v]=a*(fmt?wscale[row]:1.f);}
} }
/* Independent KV sequence per row. latent/rope are packed as [S,T,*], while /* Independent device-resident KV sequence per row. lengths selects the valid
* lengths selects the valid prefix for each row. */ * prefix; latent/rope point at paged caches updated by the host wrapper. */
__global__ static void attention_absorb_ragged_kernel(float *ctx,const float *q, __global__ static void attention_absorb_ragged_kernel(float *ctx,const float *q,
const float *latent,const float *rope,const int *lengths, const float *const *latent,const float *const *rope,const int *lengths,
const void *weights,const float *wscale,int fmt,int S,int H,int Q,int R, const void *weights,const float *wscale,int fmt,int S,int H,int Q,int R,
int V,int K,int T,float scale){ int V,int K,int T,float scale){
int s=blockIdx.y,h=blockIdx.x,tid=threadIdx.x,nt=lengths[s],rbase=h*(Q+V); int s=blockIdx.y,h=blockIdx.x,tid=threadIdx.x,nt=lengths[s],rbase=h*(Q+V);
if(s>=S||nt<1||nt>T)return; if(s>=S||nt<1||nt>T)return;
extern __shared__ float sm[];float *qa=sm,*cl=qa+K,*scores=cl+K,*red=scores+T; extern __shared__ float sm[];float *qa=sm,*cl=qa+K,*scores=cl+K,*red=scores+T;
const float *qs=q+((size_t)s*H+h)*(Q+R); const float *qs=q+((size_t)s*H+h)*(Q+R);
const float *ls=latent+(size_t)s*T*K,*rs=rope+(size_t)s*T*R; const float *ls=latent[s],*rs=rope[s];
for(int k=tid;k<K;k+=blockDim.x){float a=0;for(int d=0;d<Q;d++) for(int k=tid;k<K;k+=blockDim.x){float a=0;for(int d=0;d<Q;d++)
a+=qs[d]*weight_at(weights,fmt,(size_t)(rbase+d)*row_bytes(fmt,K),k)* a+=qs[d]*weight_at(weights,fmt,(size_t)(rbase+d)*row_bytes(fmt,K),k)*
(fmt?wscale[rbase+d]:1.f);qa[k]=a;} (fmt?wscale[rbase+d]:1.f);qa[k]=a;}
@@ -372,6 +381,15 @@ __global__ static void attention_absorb_ragged_kernel(float *ctx,const float *q,
ctx[((size_t)s*H+h)*V+v]=a*(fmt?wscale[row]:1.f);} ctx[((size_t)s*H+h)*V+v]=a*(fmt?wscale[row]:1.f);}
} }
__global__ static void ragged_kv_append(float *const *latent,float *const *rope,
const float *packed,const int *old_len,const int *add,const int *offset,int K,int R){
int s=blockIdx.x,n=add[s],base=offset[s];
for(int i=threadIdx.x;i<n*(K+R);i+=blockDim.x){
if(i<n*K)latent[s][(size_t)old_len[s]*K+i]=packed[base+i];
else rope[s][(size_t)old_len[s]*R+i-n*K]=packed[base+i];
}
}
static int reserve(float **ptr, size_t *cap, size_t bytes) { static int reserve(float **ptr, size_t *cap, size_t bytes) {
if (*cap >= bytes) return 1; if (*cap >= bytes) return 1;
if (*ptr) cudaFree(*ptr); if (*ptr) cudaFree(*ptr);
@@ -440,16 +458,17 @@ extern "C" void coli_cuda_shutdown(void) {
for(int b=0;b<24;b++) if(ctx->pipe_buf[b]) cudaFree(ctx->pipe_buf[b]); for(int b=0;b<24;b++) if(ctx->pipe_buf[b]) cudaFree(ctx->pipe_buf[b]);
if (ctx->host_x) cudaFreeHost(ctx->host_x); if (ctx->host_x) cudaFreeHost(ctx->host_x);
if (ctx->host_y) cudaFreeHost(ctx->host_y); if (ctx->host_y) cudaFreeHost(ctx->host_y);
if (ctx->host_kv) cudaFreeHost(ctx->host_kv);
if (ctx->stream) cudaStreamDestroy(ctx->stream); if (ctx->stream) cudaStreamDestroy(ctx->stream);
if (ctx->group_desc) cudaFree(ctx->group_desc); if (ctx->group_desc) cudaFree(ctx->group_desc);
ctx->x = ctx->y = ctx->gate = ctx->up = nullptr; ctx->x = ctx->y = ctx->gate = ctx->up = nullptr;
ctx->qx=nullptr; ctx->qscale=nullptr; ctx->qx=nullptr; ctx->qscale=nullptr;
ctx->aq=ctx->al=ctx->ar=ctx->ac=nullptr; ctx->aq=ctx->al=ctx->ar=ctx->ac=nullptr;
ctx->host_x=ctx->host_y=nullptr;ctx->stream=nullptr; ctx->host_x=ctx->host_y=ctx->host_kv=nullptr;ctx->stream=nullptr;
ctx->x_cap = ctx->y_cap = ctx->gate_cap = ctx->up_cap = 0; ctx->x_cap = ctx->y_cap = ctx->gate_cap = ctx->up_cap = 0;
ctx->qx_cap=ctx->qscale_cap=0; ctx->qx_cap=ctx->qscale_cap=0;
ctx->aq_cap=ctx->al_cap=ctx->ar_cap=ctx->ac_cap=0; ctx->aq_cap=ctx->al_cap=ctx->ar_cap=ctx->ac_cap=0;
ctx->host_x_cap=ctx->host_y_cap=0; ctx->host_x_cap=ctx->host_y_cap=ctx->host_kv_cap=0;
ctx->group_desc=nullptr; ctx->group_desc_cap=0; ctx->group_desc=nullptr; ctx->group_desc_cap=0;
} }
g_nctx = 0; g_nctx = 0;
@@ -805,35 +824,81 @@ extern "C" int coli_cuda_attention_project_batch(ColiCudaTensor *w,ColiCudaTenso
} }
extern "C" int coli_cuda_attention_project_ragged(ColiCudaTensor *w,ColiCudaTensor *proj, extern "C" int coli_cuda_attention_project_ragged(ColiCudaTensor *w,ColiCudaTensor *proj,
float *out,const float *q,const float *const *latent,const float *const *rope, float *out,const float *q,const void *const *keys,
const float *const *latent,const float *const *rope,
const int *lengths,int S,int H,int Q,int R,int V,int K,int T,float scale){ const int *lengths,int S,int H,int Q,int R,int V,int K,int T,float scale){
if(!w||!proj||!out||!q||!latent||!rope||!lengths||S<1||S>512||T<1||T>512|| if(!w||!proj||!out||!q||!keys||!latent||!rope||!lengths||S<1||S>512||T<1||T>8192||
H<1||Q<1||R<1||V<1||K<1||K>512||w->I!=K||w->O!=H*(Q+V)|| H<1||Q<1||R<1||V<1||K<1||K>512||w->I!=K||w->O!=H*(Q+V)||
proj->device!=w->device||proj->I!=H*V)return 0; proj->device!=w->device||proj->I!=H*V)return 0;
size_t ln=(size_t)S*T*K,rn=(size_t)S*T*R;
float *lh=(float*)std::calloc(ln,sizeof(float)),*rh=(float*)std::calloc(rn,sizeof(float));
if(!lh||!rh){std::free(lh);std::free(rh);return 0;}
for(int s=0;s<S;s++){
if(lengths[s]<1||lengths[s]>T){std::free(lh);std::free(rh);return 0;}
std::memcpy(lh+(size_t)s*T*K,latent[s],(size_t)lengths[s]*K*sizeof(float));
std::memcpy(rh+(size_t)s*T*R,rope[s],(size_t)lengths[s]*R*sizeof(float));
}
DeviceContext *dc=find_ctx(w->device); DeviceContext *dc=find_ctx(w->device);
if(!select_ctx(dc)){std::free(lh);std::free(rh);return 0;} if(!select_ctx(dc))return 0;
size_t qb=(size_t)S*H*(Q+R)*sizeof(float),lb=ln*sizeof(float),rb=rn*sizeof(float); float **dl=(float**)std::malloc((size_t)S*sizeof(*dl));
float **dr=(float**)std::malloc((size_t)S*sizeof(*dr));
int *old=(int*)std::malloc((size_t)S*sizeof(*old));
int *add=(int*)std::malloc((size_t)S*sizeof(*add));
int *off=(int*)std::malloc((size_t)S*sizeof(*off));int packed_n=0;
if(!dl||!dr||!old||!add||!off){std::free(dl);std::free(dr);std::free(old);std::free(add);std::free(off);return 0;}
for(int s=0;s<S;s++){
if(!keys[s]||lengths[s]<1||lengths[s]>T){std::free(dl);std::free(dr);std::free(old);std::free(add);std::free(off);return 0;}
RaggedKVEntry *e=nullptr;
for(int i=0;i<w->ragged_count;i++)if(w->ragged[i].key==keys[s]){e=&w->ragged[i];break;}
if(!e){
if(w->ragged_count>=512){std::free(dl);std::free(dr);std::free(old);std::free(add);std::free(off);return 0;}
e=&w->ragged[w->ragged_count++];std::memset(e,0,sizeof(*e));e->key=keys[s];
}
if(e->K!=K||e->R!=R||e->host_l!=latent[s]||e->host_r!=rope[s]||lengths[s]<e->length){
if(e->latent)cudaFree(e->latent);if(e->rope)cudaFree(e->rope);
e->latent=e->rope=nullptr;e->length=e->capacity=0;
e->K=K;e->R=R;e->host_l=latent[s];e->host_r=rope[s];
}
if(lengths[s]>e->capacity){
int cap=(lengths[s]+63)&~63;float *nl=nullptr,*nr=nullptr;
if(!cuda_ok(cudaMalloc(&nl,(size_t)cap*K*sizeof(float)),"ragged KV latent page")||
!cuda_ok(cudaMalloc(&nr,(size_t)cap*R*sizeof(float)),"ragged KV rope page")){
if(nl)cudaFree(nl);if(nr)cudaFree(nr);std::free(dl);std::free(dr);std::free(old);std::free(add);std::free(off);return 0;
}
if(e->length){
cudaMemcpyAsync(nl,e->latent,(size_t)e->length*K*sizeof(float),cudaMemcpyDeviceToDevice,dc->stream);
cudaMemcpyAsync(nr,e->rope,(size_t)e->length*R*sizeof(float),cudaMemcpyDeviceToDevice,dc->stream);
}
if(e->latent)cudaFree(e->latent);if(e->rope)cudaFree(e->rope);
e->latent=nl;e->rope=nr;e->capacity=cap;
}
dl[s]=e->latent;dr[s]=e->rope;old[s]=e->length;add[s]=lengths[s]-e->length;
off[s]=packed_n;packed_n+=add[s]*(K+R);
}
size_t qb=(size_t)S*H*(Q+R)*sizeof(float);
size_t cb=(size_t)S*H*V*sizeof(float),ob=(size_t)S*proj->O*sizeof(float); size_t cb=(size_t)S*H*V*sizeof(float),ob=(size_t)S*proj->O*sizeof(float);
int ok=reserve(&dc->aq,&dc->aq_cap,qb)&&reserve(&dc->al,&dc->al_cap,lb)&& size_t pb=(size_t)packed_n*sizeof(float);
reserve(&dc->ar,&dc->ar_cap,rb)&&reserve(&dc->ac,&dc->ac_cap,cb)&& size_t desc=(size_t)S*(2*sizeof(float*)+4*sizeof(int));
reserve(&dc->y,&dc->y_cap,ob)&& int ok=reserve(&dc->aq,&dc->aq_cap,qb)&&reserve(&dc->ac,&dc->ac_cap,cb)&&
reserve_bytes(&dc->group_desc,&dc->group_desc_cap,(size_t)S*sizeof(int)); reserve(&dc->y,&dc->y_cap,ob)&&reserve_bytes(&dc->group_desc,&dc->group_desc_cap,desc)&&
(!pb||(reserve(&dc->al,&dc->al_cap,pb)&&reserve_pinned(&dc->host_kv,&dc->host_kv_cap,pb)));
char *db=(char*)dc->group_desc;float **ddl=(float**)db,**ddr=ddl+S;
int *dn=(int*)(ddr+S),*dold=dn+S,*dadd=dold+S,*doff=dadd+S;
if(ok&&pb){
for(int s=0;s<S;s++)if(add[s]){
float *p=dc->host_kv+off[s];
std::memcpy(p,latent[s]+(size_t)old[s]*K,(size_t)add[s]*K*sizeof(float));
std::memcpy(p+(size_t)add[s]*K,rope[s]+(size_t)old[s]*R,(size_t)add[s]*R*sizeof(float));
}
ok=cuda_ok(cudaMemcpyAsync(dc->al,dc->host_kv,pb,cudaMemcpyHostToDevice,dc->stream),"ragged KV append upload");
}
if(ok)ok=cuda_ok(cudaMemcpyAsync(dc->aq,q,qb,cudaMemcpyHostToDevice,dc->stream),"ragged q upload")&& if(ok)ok=cuda_ok(cudaMemcpyAsync(dc->aq,q,qb,cudaMemcpyHostToDevice,dc->stream),"ragged q upload")&&
cuda_ok(cudaMemcpyAsync(dc->al,lh,lb,cudaMemcpyHostToDevice,dc->stream),"ragged latent upload")&& cuda_ok(cudaMemcpyAsync(ddl,dl,(size_t)S*sizeof(float*),cudaMemcpyHostToDevice,dc->stream),"ragged latent pointers")&&
cuda_ok(cudaMemcpyAsync(dc->ar,rh,rb,cudaMemcpyHostToDevice,dc->stream),"ragged rope upload")&& cuda_ok(cudaMemcpyAsync(ddr,dr,(size_t)S*sizeof(float*),cudaMemcpyHostToDevice,dc->stream),"ragged rope pointers")&&
cuda_ok(cudaMemcpyAsync(dc->group_desc,lengths,(size_t)S*sizeof(int),cudaMemcpyHostToDevice,dc->stream),"ragged lengths upload"); cuda_ok(cudaMemcpyAsync(dn,lengths,(size_t)S*sizeof(int),cudaMemcpyHostToDevice,dc->stream),"ragged lengths upload")&&
std::free(lh);std::free(rh);if(!ok)return 0; cuda_ok(cudaMemcpyAsync(dold,old,(size_t)S*sizeof(int),cudaMemcpyHostToDevice,dc->stream),"ragged old lengths")&&
cuda_ok(cudaMemcpyAsync(dadd,add,(size_t)S*sizeof(int),cudaMemcpyHostToDevice,dc->stream),"ragged append lengths")&&
cuda_ok(cudaMemcpyAsync(doff,off,(size_t)S*sizeof(int),cudaMemcpyHostToDevice,dc->stream),"ragged append offsets");
if(ok&&pb)ragged_kv_append<<<S,256,0,dc->stream>>>(ddl,ddr,dc->al,dold,dadd,doff,K,R);
if(ok)for(int s=0;s<S;s++){
for(int i=0;i<w->ragged_count;i++)if(w->ragged[i].key==keys[s]){w->ragged[i].length=lengths[s];break;}
}
std::free(dl);std::free(dr);std::free(old);std::free(add);std::free(off);if(!ok)return 0;
size_t shared=(size_t)(2*K+T+256)*sizeof(float); size_t shared=(size_t)(2*K+T+256)*sizeof(float);
attention_absorb_ragged_kernel<<<dim3(H,S),256,shared,dc->stream>>>(dc->ac,dc->aq,dc->al,dc->ar, attention_absorb_ragged_kernel<<<dim3(H,S),256,shared,dc->stream>>>(dc->ac,dc->aq,ddl,ddr,
(const int*)dc->group_desc,w->weights,w->scales,w->fmt,S,H,Q,R,V,K,T,scale); dn,w->weights,w->scales,w->fmt,S,H,Q,R,V,K,T,scale);
quant_matmul<<<dim3(proj->O,S),256,0,dc->stream>>>(dc->y,dc->ac,proj->weights, quant_matmul<<<dim3(proj->O,S),256,0,dc->stream>>>(dc->y,dc->ac,proj->weights,
proj->scales,proj->fmt,S,proj->I,proj->O,row_bytes(proj->fmt,proj->I)); proj->scales,proj->fmt,S,proj->I,proj->O,row_bytes(proj->fmt,proj->I));
return cuda_ok(cudaGetLastError(),"ragged attention launch")&& return cuda_ok(cudaGetLastError(),"ragged attention launch")&&
@@ -852,6 +917,10 @@ extern "C" void coli_cuda_tensor_free(ColiCudaTensor *tensor) {
} }
if (tensor->weights) cudaFree(tensor->weights); if (tensor->weights) cudaFree(tensor->weights);
if (tensor->scales) cudaFree(tensor->scales); if (tensor->scales) cudaFree(tensor->scales);
for(int i=0;i<tensor->ragged_count;i++){
if(tensor->ragged[i].latent)cudaFree(tensor->ragged[i].latent);
if(tensor->ragged[i].rope)cudaFree(tensor->ragged[i].rope);
}
std::free(tensor); std::free(tensor);
} }
c/backend_cuda.h
+2 -1
View File
@@ -94,7 +94,8 @@ COLI_CUDA_DLLEXPORT int coli_cuda_attention_project_batch(ColiCudaTensor *kv_b,C
int V,int K,int T,float attention_scale); int V,int K,int T,float attention_scale);
COLI_CUDA_DLLEXPORT int coli_cuda_attention_project_ragged(ColiCudaTensor *kv_b,ColiCudaTensor *o_proj, COLI_CUDA_DLLEXPORT int coli_cuda_attention_project_ragged(ColiCudaTensor *kv_b,ColiCudaTensor *o_proj,
float *out,const float *q,const float *const *latent,const float *const *rope, float *out,const float *q,const void *const *keys,
const float *const *latent,const float *const *rope,
const int *lengths,int S,int H,int Q,int R,int V,int K,int max_t,float attention_scale); const int *lengths,int S,int H,int Q,int R,int V,int K,int max_t,float attention_scale);
COLI_CUDA_DLLEXPORT void coli_cuda_tensor_free(ColiCudaTensor *tensor); COLI_CUDA_DLLEXPORT void coli_cuda_tensor_free(ColiCudaTensor *tensor);
c/backend_loader.c
+5 -3
View File
@@ -62,7 +62,8 @@ typedef int (*fn_attention_absorb_batch_dev)(ColiCudaTensor *kv_b_shard,float *c
typedef int (*fn_attention_absorb_kvdev)(ColiCudaTensor *kv_b,float *ctx,const float *q, const float *latent_dev,const float *rope_dev,int H,int Q,int R,int V,int K,int T, float scale); typedef int (*fn_attention_absorb_kvdev)(ColiCudaTensor *kv_b,float *ctx,const float *q, const float *latent_dev,const float *rope_dev,int H,int Q,int R,int V,int K,int T, float scale);
typedef int (*fn_attention_project_batch)(ColiCudaTensor *kv_b,ColiCudaTensor *o_proj, float *out,const float *q,const float *latent, const float *rope,int S,int H,int Q,int R, int V,int K,int T,float attention_scale); typedef int (*fn_attention_project_batch)(ColiCudaTensor *kv_b,ColiCudaTensor *o_proj, float *out,const float *q,const float *latent, const float *rope,int S,int H,int Q,int R, int V,int K,int T,float attention_scale);
typedef int (*fn_attention_project_ragged)(ColiCudaTensor *kv_b,ColiCudaTensor *o_proj, typedef int (*fn_attention_project_ragged)(ColiCudaTensor *kv_b,ColiCudaTensor *o_proj,
float *out,const float *q,const float *const *latent,const float *const *rope, float *out,const float *q,const void *const *keys,
const float *const *latent,const float *const *rope,
const int *lengths,int S,int H,int Q,int R,int V,int K,int max_t,float attention_scale); const int *lengths,int S,int H,int Q,int R,int V,int K,int max_t,float attention_scale);
typedef int (*fn_attention_project_batch_dev)(ColiCudaTensor *kv_b,ColiCudaTensor *o_proj, float *out,const float *q_dev,const float *latent_dev,const float *rope_dev, int S,int H,int Q,int R,int V,int K,int T,float scale); typedef int (*fn_attention_project_batch_dev)(ColiCudaTensor *kv_b,ColiCudaTensor *o_proj, float *out,const float *q_dev,const float *latent_dev,const float *rope_dev, int S,int H,int Q,int R,int V,int K,int T,float scale);
typedef int (*fn_attention_project_batch_dev_out)(ColiCudaTensor *kv_b,ColiCudaTensor *o_proj, float *out_dev,const float *q_dev,const float *latent_dev,const float *rope_dev, int S,int H,int Q,int R,int V,int K,int T,float scale); typedef int (*fn_attention_project_batch_dev_out)(ColiCudaTensor *kv_b,ColiCudaTensor *o_proj, float *out_dev,const float *q_dev,const float *latent_dev,const float *rope_dev, int S,int H,int Q,int R,int V,int K,int T,float scale);
@@ -348,10 +349,11 @@ int coli_cuda_attention_project_batch(ColiCudaTensor *kv_b,ColiCudaTensor *o_pro
} }
int coli_cuda_attention_project_ragged(ColiCudaTensor *kv_b,ColiCudaTensor *o_proj, int coli_cuda_attention_project_ragged(ColiCudaTensor *kv_b,ColiCudaTensor *o_proj,
float *out,const float *q,const float *const *latent,const float *const *rope, float *out,const float *q,const void *const *keys,
const float *const *latent,const float *const *rope,
const int *lengths,int S,int H,int Q,int R,int V,int K,int max_t,float attention_scale){ const int *lengths,int S,int H,int Q,int R,int V,int K,int max_t,float attention_scale){
if(!coli_cuda_load()) return 0; if(!coli_cuda_load()) return 0;
return g_cuda.attention_project_ragged(kv_b,o_proj,out,q,latent,rope,lengths, return g_cuda.attention_project_ragged(kv_b,o_proj,out,q,keys,latent,rope,lengths,
S,H,Q,R,V,K,max_t,attention_scale); S,H,Q,R,V,K,max_t,attention_scale);
} }
c/glm.c
+10 -7
View File
@@ -2717,18 +2717,20 @@ static void attention_rows(Model *m, Layer *l, int layer, float *x, int S, int p
!dnsel&&l->kv_b.cuda_eligible&&l->o.cuda_eligible&& !dnsel&&l->kv_b.cuda_eligible&&l->o.cuda_eligible&&
qt_cuda_upload(&l->kv_b)&&qt_cuda_upload(&l->o)){ qt_cuda_upload(&l->kv_b)&&qt_cuda_upload(&l->o)){
const float **rl=malloc((size_t)S*sizeof(*rl)),**rr=malloc((size_t)S*sizeof(*rr)); const float **rl=malloc((size_t)S*sizeof(*rl)),**rr=malloc((size_t)S*sizeof(*rr));
const void **rk=malloc((size_t)S*sizeof(*rk));
int *rn=malloc((size_t)S*sizeof(*rn)); int mt=0; int *rn=malloc((size_t)S*sizeof(*rn)); int mt=0;
if(rl&&rr&&rn){ if(rk&&rl&&rr&&rn){
for(int s=0;s<S;s++){ for(int s=0;s<S;s++){
int pos=positions[s],st0=kvs[s]->kv_start[layer]; rn[s]=pos+1-st0; int pos=positions[s],st0=kvs[s]->kv_start[layer]; rn[s]=pos+1-st0;
rk[s]=kvs[s];
rl[s]=coli_kv_row(kvs[s]->Lc[layer],st0,kvl); rl[s]=coli_kv_row(kvs[s]->Lc[layer],st0,kvl);
rr[s]=coli_kv_row(kvs[s]->Rc[layer],st0,c->qk_rope); rr[s]=coli_kv_row(kvs[s]->Rc[layer],st0,c->qk_rope);
if(rn[s]>mt)mt=rn[s]; if(rn[s]>mt)mt=rn[s];
} }
cuda_core=cuda_projected=coli_cuda_attention_project_ragged(l->kv_b.cuda,l->o.cuda, cuda_core=cuda_projected=coli_cuda_attention_project_ragged(l->kv_b.cuda,l->o.cuda,
out,Q,rl,rr,rn,S,H,c->qk_nope,c->qk_rope,vh,kvl,mt,c->attn_scale); out,Q,rk,rl,rr,rn,S,H,c->qk_nope,c->qk_rope,vh,kvl,mt,c->attn_scale);
} }
free(rl);free(rr);free(rn); free(rk);free(rl);free(rr);free(rn);
} else if(cuda_absorb&&l->n_kv_b_shard>1){ } else if(cuda_absorb&&l->n_kv_b_shard>1){
int n=l->n_kv_b_shard,st0=m->kv_start[layer],nt=pos_base+S-st0,ok=1; int n=l->n_kv_b_shard,st0=m->kv_start[layer],nt=pos_base+S-st0,ok=1;
float *qs=falloc((int64_t)S*H*qh),*cs=falloc((int64_t)S*H*vh); float *qs=falloc((int64_t)S*H*qh),*cs=falloc((int64_t)S*H*vh);
@@ -5271,7 +5273,7 @@ static void run_serve_mux(Model *m, const char *snap){
g_draft=0; /* one scheduler owns every forward; MTP/speculation is not ragged-safe */ g_draft=0; /* one scheduler owns every forward; MTP/speculation is not ragged-safe */
int maxctx=getenv("CTX")?atoi(getenv("CTX")):4096; int maxctx=getenv("CTX")?atoi(getenv("CTX")):4096;
int nctx=getenv("KV_SLOTS")?atoi(getenv("KV_SLOTS")):1; int nctx=getenv("KV_SLOTS")?atoi(getenv("KV_SLOTS")):1;
if(nctx<1||nctx>16){fprintf(stderr,"KV_SLOTS deve essere tra 1 e 16\n");exit(2);} if(nctx<1||nctx>512){fprintf(stderr,"KV_SLOTS must be between 1 and 512\n");exit(2);}
g_kvsave=getenv("KVSAVE")?atoi(getenv("KVSAVE")):1; g_kvsave=getenv("KVSAVE")?atoi(getenv("KVSAVE")):1;
KVState *initial=m->kv; free(initial->kv_start); free(initial); KVState *initial=m->kv; free(initial->kv_start); free(initial);
ServeCtx *ctx=calloc(nctx,sizeof(*ctx)); ServeReq *req=calloc(nctx,sizeof(*req)); ServeCtx *ctx=calloc(nctx,sizeof(*ctx)); ServeReq *req=calloc(nctx,sizeof(*req));
@@ -5327,7 +5329,7 @@ static void run_serve_mux(Model *m, const char *snap){
} }
active=0; for(int i=0;i<nctx;i++) active+=req[i].active; active=0; for(int i=0;i<nctx;i++) active+=req[i].active;
if(!active){ if(eof) break; continue; } if(!active){ if(eof) break; continue; }
DecodeRow rows[16]; int slots[16], S=0; DecodeRow rows[512]; int slots[512], S=0;
for(int i=0;i<nctx;i++) if(req[i].active){ for(int i=0;i<nctx;i++) if(req[i].active){
rows[S]=(DecodeRow){&ctx[i].kv,req[i].pending,ctx[i].len}; slots[S++]=i; rows[S]=(DecodeRow){&ctx[i].kv,req[i].pending,ctx[i].len}; slots[S++]=i;
} }
@@ -6334,8 +6336,9 @@ int main(int argc, char **argv){
int cap = argc>1?atoi(argv[1]):64; int cap = argc>1?atoi(argv[1]):64;
int ebits= argc>2?atoi(argv[2]):8; int ebits= argc>2?atoi(argv[2]):8;
int dbits= argc>3?atoi(argv[3]):ebits; int dbits= argc>3?atoi(argv[3]):ebits;
if(getenv("SERVE") && (kv_slot_count()<1 || kv_slot_count()>16)){ int kv_limit=(getenv("SERVE_BATCH")&&atoi(getenv("SERVE_BATCH")))?512:16;
fprintf(stderr,"KV_SLOTS must be between 1 and 16\n"); return 2; if(getenv("SERVE") && (kv_slot_count()<1 || kv_slot_count()>kv_limit)){
fprintf(stderr,"KV_SLOTS must be between 1 and %d\n",kv_limit); return 2;
} }
#ifdef COLI_CUDA #ifdef COLI_CUDA
if(getenv("COLI_CUDA") && atoi(getenv("COLI_CUDA"))){ if(getenv("COLI_CUDA") && atoi(getenv("COLI_CUDA"))){
c/tests/test_ragged_attention.cu
+7 -3
View File
@@ -16,14 +16,18 @@ int main(){
!coli_cuda_tensor_upload(&tp,p.data(),nullptr,0,D,O,dev))return 1; !coli_cuda_tensor_upload(&tp,p.data(),nullptr,0,D,O,dev))return 1;
int n[S]={1,2,3};std::vector<std::vector<float>> l(S),r(S); int n[S]={1,2,3};std::vector<std::vector<float>> l(S),r(S);
const float *lp[S],*rp[S]; const float *lp[S],*rp[S];
const void *keys[S];
for(int s=0;s<S;s++){ for(int s=0;s<S;s++){
l[s].resize(n[s]*K);r[s].resize(n[s]*R); l[s].resize(n[s]*K);r[s].resize(n[s]*R);
for(size_t i=0;i<l[s].size();i++)l[s][i]=((int)((i+s*3)%9)-4)*.08f; for(size_t i=0;i<l[s].size();i++)l[s][i]=((int)((i+s*3)%9)-4)*.08f;
for(size_t i=0;i<r[s].size();i++)r[s][i]=((int)((i+s)%5)-2)*.06f; for(size_t i=0;i<r[s].size();i++)r[s][i]=((int)((i+s)%5)-2)*.06f;
lp[s]=l[s].data();rp[s]=r[s].data(); lp[s]=l[s].data();rp[s]=r[s].data();keys[s]=&l[s];
} }
float got[S*O],ref[S*O]; float got[S*O],ref[S*O],warm[S*O];
if(!coli_cuda_attention_project_ragged(tw,tp,got,q.data(),lp,rp,n,S,H,Q,R,V,K,T,.2f))return 2; int first[S]={1,1,1};
if(!coli_cuda_attention_project_ragged(tw,tp,warm,q.data(),keys,lp,rp,first,
S,H,Q,R,V,K,1,.2f))return 2;
if(!coli_cuda_attention_project_ragged(tw,tp,got,q.data(),keys,lp,rp,n,S,H,Q,R,V,K,T,.2f))return 2;
for(int s=0;s<S;s++)if(!coli_cuda_attention_project_batch(tw,tp,ref+s*O, for(int s=0;s<S;s++)if(!coli_cuda_attention_project_batch(tw,tp,ref+s*O,
q.data()+s*H*(Q+R),lp[s],rp[s],1,H,Q,R,V,K,n[s],.2f))return 3; q.data()+s*H*(Q+R),lp[s],rp[s],1,H,Q,R,V,K,n[s],.2f))return 3;
double e=0,z=0;for(int i=0;i<S*O;i++){ double e=0,z=0;for(int i=0;i<S*O;i++){