mirror of
https://github.com/SheffieldML/GPy.git
synced 2026-05-08 11:32:39 +02:00
fix the remaining problem of cache.py
This commit is contained in:
Zhenwen Dai
parent
ca1edecce4
commit
59172435e2
3 changed files with 61 additions and 20 deletions
|
|
@ -58,7 +58,23 @@ gpu_code = """
|
|||
}
|
||||
}
|
||||
|
||||
__global__ void psi1computations(double *psi1, double var, double *l, double *Z, double *mu, double *S, int N, int M, int Q)
|
||||
__global__ void compDenom(double *log_denom1, double *log_denom2, double *l, double *S, int N, int Q)
|
||||
{
|
||||
int n_start, n_end;
|
||||
divide_data(N, gridDim.x, blockIdx.x, &n_start, &n_end);
|
||||
|
||||
for(int i=n_start*Q+threadIdx.x; i<n_end*Q; i+=blockDim.x) {
|
||||
int n=i/Q;
|
||||
int q=i%Q;
|
||||
|
||||
double Snq = S[IDX_NQ(n,q)];
|
||||
double lq = l[q]*l[q];
|
||||
log_denom1[IDX_NQ(n,q)] = log(Snq/lq+1.);
|
||||
log_denom2[IDX_NQ(n,q)] = log(2.*Snq/lq+1.);
|
||||
}
|
||||
}
|
||||
|
||||
__global__ void psi1computations(double *psi1, double *log_denom1, double var, double *l, double *Z, double *mu, double *S, int N, int M, int Q)
|
||||
{
|
||||
int m_start, m_end;
|
||||
divide_data(M, gridDim.x, blockIdx.x, &m_start, &m_end);
|
||||
|
|
@ -70,15 +86,14 @@ gpu_code = """
|
|||
double muZ = mu[IDX_NQ(n,q)]-Z[IDX_MQ(m,q)];
|
||||
double Snq = S[IDX_NQ(n,q)];
|
||||
double lq = l[q]*l[q];
|
||||
log_psi1 += (muZ*muZ/(Snq+lq))/(-2.);
|
||||
log_psi1 += log(Snq/lq+1)/(-2.);
|
||||
log_psi1 += (muZ*muZ/(Snq+lq)+log_denom1[IDX_NQ(n,q)])/(-2.);
|
||||
}
|
||||
psi1[IDX_NM(n,m)] = var*exp(log_psi1);
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
__global__ void psi2computations(double *psi2, double *psi2n, double var, double *l, double *Z, double *mu, double *S, int N, int M, int Q)
|
||||
__global__ void psi2computations(double *psi2, double *psi2n, double *log_denom2, double var, double *l, double *Z, double *mu, double *S, int N, int M, int Q)
|
||||
{
|
||||
int psi2_idx_start, psi2_idx_end;
|
||||
__shared__ double psi2_local[THREADNUM];
|
||||
|
|
@ -96,8 +111,8 @@ gpu_code = """
|
|||
double muZhat = mu[IDX_NQ(n,q)]- (Z[IDX_MQ(m1,q)]+Z[IDX_MQ(m2,q)])/2.;
|
||||
double Snq = S[IDX_NQ(n,q)];
|
||||
double lq = l[q]*l[q];
|
||||
log_psi2_n += dZ*dZ/(-4.*lq)-muZhat*muZhat/(2.*Snq+lq);
|
||||
log_psi2_n += log(2.*Snq/lq+1)/(-2.);
|
||||
log_psi2_n += dZ*dZ/(-4.*lq)-muZhat*muZhat/(2.*Snq+lq) + log_denom2[IDX_NQ(n,q)]/(-2.);
|
||||
//log_psi2_n += log(2.*Snq/lq+1)/(-2.);
|
||||
}
|
||||
double exp_psi2_n = exp(log_psi2_n);
|
||||
psi2n[IDX_NMM(n,m1,m2)] = var*var*exp_psi2_n;
|
||||
|
|
@ -237,9 +252,15 @@ class PSICOMP_RBF_GPU(PSICOMP_RBF):
|
|||
self.blocknum = 15
|
||||
module = SourceModule("#define THREADNUM "+str(self.threadnum)+"\n"+gpu_code)
|
||||
self.g_psi1computations = module.get_function('psi1computations')
|
||||
self.g_psi1computations.prepare('PPdPPPPiii')
|
||||
self.g_psi2computations = module.get_function('psi2computations')
|
||||
self.g_psi2computations.prepare('PPPdPPPPiii')
|
||||
self.g_psi1compDer = module.get_function('psi1compDer')
|
||||
self.g_psi1compDer.prepare('PPPPPPPdPPPPiii')
|
||||
self.g_psi2compDer = module.get_function('psi2compDer')
|
||||
self.g_psi2compDer.prepare('PPPPPPPdPPPPiii')
|
||||
self.g_compDenom = module.get_function('compDenom')
|
||||
self.g_compDenom.prepare('PPPPii')
|
||||
|
||||
def _initGPUCache(self, N, M, Q):
|
||||
if self.gpuCache == None:
|
||||
|
|
@ -253,6 +274,8 @@ class PSICOMP_RBF_GPU(PSICOMP_RBF):
|
|||
'psi2n_gpu' :gpuarray.empty((N,M,M),np.float64,order='F'),
|
||||
'dL_dpsi1_gpu' :gpuarray.empty((N,M),np.float64,order='F'),
|
||||
'dL_dpsi2_gpu' :gpuarray.empty((M,M),np.float64,order='F'),
|
||||
'log_denom1_gpu' :gpuarray.empty((N,Q),np.float64,order='F'),
|
||||
'log_denom2_gpu' :gpuarray.empty((N,Q),np.float64,order='F'),
|
||||
# derivatives
|
||||
'dvar_gpu' :gpuarray.empty((self.blocknum,),np.float64, order='F'),
|
||||
'dl_gpu' :gpuarray.empty((Q,self.blocknum),np.float64, order='F'),
|
||||
|
|
@ -277,6 +300,10 @@ class PSICOMP_RBF_GPU(PSICOMP_RBF):
|
|||
self.gpuCache['Z_gpu'].set(np.asfortranarray(Z))
|
||||
self.gpuCache['mu_gpu'].set(np.asfortranarray(mu))
|
||||
self.gpuCache['S_gpu'].set(np.asfortranarray(S))
|
||||
N,Q = self.gpuCache['S_gpu'].shape
|
||||
# t=self.g_compDenom(self.gpuCache['log_denom1_gpu'],self.gpuCache['log_denom2_gpu'],self.gpuCache['l_gpu'],self.gpuCache['S_gpu'], np.int32(N), np.int32(Q), block=(self.threadnum,1,1), grid=(self.blocknum,1),time_kernel=True)
|
||||
# print 'g_compDenom '+str(t)
|
||||
self.g_compDenom.prepared_call((self.blocknum,1),(self.threadnum,1,1), self.gpuCache['log_denom1_gpu'].gpudata,self.gpuCache['log_denom2_gpu'].gpudata,self.gpuCache['l_gpu'].gpudata,self.gpuCache['S_gpu'].gpudata, np.int32(N), np.int32(Q))
|
||||
|
||||
def reset_derivative(self):
|
||||
self.gpuCache['dvar_gpu'].fill(0.)
|
||||
|
|
@ -309,11 +336,17 @@ class PSICOMP_RBF_GPU(PSICOMP_RBF):
|
|||
Z_gpu = self.gpuCache['Z_gpu']
|
||||
mu_gpu = self.gpuCache['mu_gpu']
|
||||
S_gpu = self.gpuCache['S_gpu']
|
||||
|
||||
log_denom1_gpu = self.gpuCache['log_denom1_gpu']
|
||||
log_denom2_gpu = self.gpuCache['log_denom2_gpu']
|
||||
|
||||
psi0 = np.empty((N,))
|
||||
psi0[:] = variance
|
||||
self.g_psi1computations(psi1_gpu, np.float64(variance),l_gpu,Z_gpu,mu_gpu,S_gpu, np.int32(N), np.int32(M), np.int32(Q), block=(self.threadnum,1,1), grid=(self.blocknum,1))
|
||||
self.g_psi2computations(psi2_gpu, psi2n_gpu, np.float64(variance),l_gpu,Z_gpu,mu_gpu,S_gpu, np.int32(N), np.int32(M), np.int32(Q), block=(self.threadnum,1,1), grid=(self.blocknum,1))
|
||||
self.g_psi1computations.prepared_call((self.blocknum,1),(self.threadnum,1,1),psi1_gpu.gpudata, log_denom1_gpu.gpudata, np.float64(variance),l_gpu.gpudata,Z_gpu.gpudata,mu_gpu.gpudata,S_gpu.gpudata, np.int32(N), np.int32(M), np.int32(Q))
|
||||
self.g_psi2computations.prepared_call((self.blocknum,1),(self.threadnum,1,1),psi2_gpu.gpudata, psi2n_gpu.gpudata, log_denom2_gpu.gpudata, np.float64(variance),l_gpu.gpudata,Z_gpu.gpudata,mu_gpu.gpudata,S_gpu.gpudata, np.int32(N), np.int32(M), np.int32(Q))
|
||||
# t = self.g_psi1computations(psi1_gpu, log_denom1_gpu, np.float64(variance),l_gpu,Z_gpu,mu_gpu,S_gpu, np.int32(N), np.int32(M), np.int32(Q), block=(self.threadnum,1,1), grid=(self.blocknum,1),time_kernel=True)
|
||||
# print 'g_psi1computations '+str(t)
|
||||
# t = self.g_psi2computations(psi2_gpu, psi2n_gpu, log_denom2_gpu, np.float64(variance),l_gpu,Z_gpu,mu_gpu,S_gpu, np.int32(N), np.int32(M), np.int32(Q), block=(self.threadnum,1,1), grid=(self.blocknum,1),time_kernel=True)
|
||||
# print 'g_psi2computations '+str(t)
|
||||
|
||||
if self.GPU_direct:
|
||||
return psi0, psi1_gpu, psi2_gpu
|
||||
|
|
@ -352,8 +385,12 @@ class PSICOMP_RBF_GPU(PSICOMP_RBF):
|
|||
dL_dpsi0_sum = dL_dpsi0.sum()
|
||||
|
||||
self.reset_derivative()
|
||||
self.g_psi1compDer(dvar_gpu,dl_gpu,dZ_gpu,dmu_gpu,dS_gpu,dL_dpsi1_gpu,psi1_gpu, np.float64(variance),l_gpu,Z_gpu,mu_gpu,S_gpu, np.int32(N), np.int32(M), np.int32(Q), block=(self.threadnum,1,1), grid=(self.blocknum,1))
|
||||
self.g_psi2compDer(dvar_gpu,dl_gpu,dZ_gpu,dmu_gpu,dS_gpu,dL_dpsi2_gpu,psi2n_gpu, np.float64(variance),l_gpu,Z_gpu,mu_gpu,S_gpu, np.int32(N), np.int32(M), np.int32(Q), block=(self.threadnum,1,1), grid=(self.blocknum,1))
|
||||
# t=self.g_psi1compDer(dvar_gpu,dl_gpu,dZ_gpu,dmu_gpu,dS_gpu,dL_dpsi1_gpu,psi1_gpu, np.float64(variance),l_gpu,Z_gpu,mu_gpu,S_gpu, np.int32(N), np.int32(M), np.int32(Q), block=(self.threadnum,1,1), grid=(self.blocknum,1),time_kernel=True)
|
||||
# print 'g_psi1compDer '+str(t)
|
||||
# t=self.g_psi2compDer(dvar_gpu,dl_gpu,dZ_gpu,dmu_gpu,dS_gpu,dL_dpsi2_gpu,psi2n_gpu, np.float64(variance),l_gpu,Z_gpu,mu_gpu,S_gpu, np.int32(N), np.int32(M), np.int32(Q), block=(self.threadnum,1,1), grid=(self.blocknum,1),time_kernel=True)
|
||||
# print 'g_psi2compDer '+str(t)
|
||||
self.g_psi1compDer.prepared_call((self.blocknum,1),(self.threadnum,1,1),dvar_gpu.gpudata,dl_gpu.gpudata,dZ_gpu.gpudata,dmu_gpu.gpudata,dS_gpu.gpudata,dL_dpsi1_gpu.gpudata,psi1_gpu.gpudata, np.float64(variance),l_gpu.gpudata,Z_gpu.gpudata,mu_gpu.gpudata,S_gpu.gpudata, np.int32(N), np.int32(M), np.int32(Q))
|
||||
self.g_psi2compDer.prepared_call((self.blocknum,1),(self.threadnum,1,1),dvar_gpu.gpudata,dl_gpu.gpudata,dZ_gpu.gpudata,dmu_gpu.gpudata,dS_gpu.gpudata,dL_dpsi2_gpu.gpudata,psi2n_gpu.gpudata, np.float64(variance),l_gpu.gpudata,Z_gpu.gpudata,mu_gpu.gpudata,S_gpu.gpudata, np.int32(N), np.int32(M), np.int32(Q))
|
||||
|
||||
dL_dvar = dL_dpsi0_sum + gpuarray.sum(dvar_gpu).get()
|
||||
sum_axis(grad_mu_gpu,dmu_gpu,N*Q,self.blocknum)
|
||||
|
|
|
|||
|
|
@ -79,7 +79,6 @@ class Stationary(Kern):
|
|||
#a convenience function, so we can cache dK_dr
|
||||
return self.dK_dr(self._scaled_dist(X, X2))
|
||||
|
||||
@Cache_this(limit=5, ignore_args=(0,))
|
||||
def _unscaled_dist(self, X, X2=None):
|
||||
"""
|
||||
Compute the Euclidean distance between each row of X and X2, or between
|
||||
|
|
|
|||
|
|
@ -30,13 +30,13 @@ class Cacher(object):
|
|||
self.cached_outputs = {} # point from cache_ids to outputs
|
||||
self.inputs_changed = {} # point from cache_ids to bools
|
||||
|
||||
def combine_inputs(self, args, kw):
|
||||
def combine_inputs(self, args, kw, ignore_args):
|
||||
"Combines the args and kw in a unique way, such that ordering of kwargs does not lead to recompute"
|
||||
return args + tuple(c[1] for c in sorted(kw.items(), key=lambda x: x[0]))
|
||||
return tuple(a for i,a in enumerate(args) if i not in ignore_args) + tuple(c[1] for c in sorted(kw.items(), key=lambda x: x[0]))
|
||||
|
||||
def prepare_cache_id(self, combined_args_kw, ignore_args):
|
||||
def prepare_cache_id(self, combined_args_kw):
|
||||
"get the cacheid (conc. string of argument ids in order) ignoring ignore_args"
|
||||
return "".join(str(id(a)) for i,a in enumerate(combined_args_kw) if i not in ignore_args)
|
||||
return "".join(str(id(a)) for a in combined_args_kw)
|
||||
|
||||
def ensure_cache_length(self, cache_id):
|
||||
"Ensures the cache is within its limits and has one place free"
|
||||
|
|
@ -45,6 +45,8 @@ class Cacher(object):
|
|||
cache_id = self.order.popleft()
|
||||
combined_args_kw = self.cached_inputs[cache_id]
|
||||
for ind in combined_args_kw:
|
||||
if ind is None:
|
||||
continue
|
||||
ind_id = id(ind)
|
||||
ref, cache_ids = self.cached_input_ids[ind_id]
|
||||
if len(cache_ids) == 1 and ref() is not None:
|
||||
|
|
@ -63,6 +65,8 @@ class Cacher(object):
|
|||
self.order.append(cache_id)
|
||||
self.cached_inputs[cache_id] = combined_args_kw
|
||||
for a in combined_args_kw:
|
||||
if a is None:
|
||||
continue
|
||||
ind_id = id(a)
|
||||
v = self.cached_input_ids.get(ind_id, [weakref.ref(a), []])
|
||||
v[1].append(cache_id)
|
||||
|
|
@ -82,11 +86,12 @@ class Cacher(object):
|
|||
return self.operation(*args, **kw)
|
||||
|
||||
# 2: prepare_cache_id and get the unique id string for this call
|
||||
inputs = self.combine_inputs(args, kw)
|
||||
cache_id = self.prepare_cache_id(inputs, self.ignore_args)
|
||||
|
||||
inputs = self.combine_inputs(args, kw, self.ignore_args)
|
||||
cache_id = self.prepare_cache_id(inputs)
|
||||
|
||||
# 2: if anything is not cachable, we will just return the operation, without caching
|
||||
if reduce(lambda a,b: a or (not isinstance(b, Observable)), inputs, False):
|
||||
if reduce(lambda a,b: a or (not (isinstance(b, Observable) or (b is None))), inputs, False):
|
||||
# print '['+self.operation.__name__+'] contain un-cachable arguments!'
|
||||
return self.operation(*args, **kw)
|
||||
# 3&4: check whether this cache_id has been cached, then has it changed?
|
||||
try:
|
||||
|
|
|
|||
Loading…
Add table
Add a link
Reference in a new issue