thek0tyara/llama-cpp-turboquant
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions 4

Updated array lr file to align with example

Browse source
This commit is contained in:
Alejandro Saucedo 2020-11-11 07:26:05 +00:00
parent 5cb2b61426
commit 8a9f0cd7b4
1 changed files with 17 additions and 4 deletions
Download patch file Download diff file Expand all files Collapse all files

21
python/test/test_logistic_regression.py
View file

@ -2,6 +2,7 @@ import pyshader as ps
import kp import kp
def test_logistic_regression(): def test_logistic_regression():
@ps.python2shader @ps.python2shader
def compute_shader( def compute_shader(
index = ("input", "GlobalInvocationId", ps.ivec3), index = ("input", "GlobalInvocationId", ps.ivec3),
@ -42,6 +43,8 @@ def test_logistic_regression():
l_out[i] = loss l_out[i] = loss
mgr = kp.Manager(0)
# First we create input and ouput tensors for shader # First we create input and ouput tensors for shader
tensor_x_i = kp.Tensor([0.0, 1.0, 1.0, 1.0, 1.0]) tensor_x_i = kp.Tensor([0.0, 1.0, 1.0, 1.0, 1.0])
tensor_x_j = kp.Tensor([0.0, 0.0, 0.0, 1.0, 1.0]) tensor_x_j = kp.Tensor([0.0, 0.0, 0.0, 1.0, 1.0])
@ -57,22 +60,30 @@ def test_logistic_regression():
tensor_l_out = kp.Tensor([0.0, 0.0, 0.0, 0.0, 0.0]) tensor_l_out = kp.Tensor([0.0, 0.0, 0.0, 0.0, 0.0])
tensor_m = kp.Tensor([ 5.0 ]) tensor_m = kp.Tensor([ tensor_y.size() ])
# We store them in an array for easier interaction # We store them in an array for easier interaction
params = [tensor_x_i, tensor_x_j, tensor_y, tensor_w_in, tensor_w_out_i, params = [tensor_x_i, tensor_x_j, tensor_y, tensor_w_in, tensor_w_out_i,
tensor_w_out_j, tensor_b_in, tensor_b_out, tensor_l_out, tensor_m] tensor_w_out_j, tensor_b_in, tensor_b_out, tensor_l_out, tensor_m]
mgr = kp.Manager()
mgr.eval_tensor_create_def(params) mgr.eval_tensor_create_def(params)
# Record commands for efficient evaluation # Create a managed sequence
sq = mgr.create_sequence() sq = mgr.create_sequence()
# Clear previous operations and begin recording for new operations
sq.begin() sq.begin()
# Record operation to sync memory from local to GPU memory
sq.record_tensor_sync_device([tensor_w_in, tensor_b_in]) sq.record_tensor_sync_device([tensor_w_in, tensor_b_in])
# Record operation to execute GPU shader against all our parameters
sq.record_algo_data(params, compute_shader.to_spirv()) sq.record_algo_data(params, compute_shader.to_spirv())
# Record operation to sync memory from GPU to local memory
sq.record_tensor_sync_local([tensor_w_out_i, tensor_w_out_j, tensor_b_out, tensor_l_out]) sq.record_tensor_sync_local([tensor_w_out_i, tensor_w_out_j, tensor_b_out, tensor_l_out])
# Stop recording operations
sq.end() sq.end()
ITERATIONS = 100 ITERATIONS = 100
@ -80,6 +91,8 @@ def test_logistic_regression():
# Perform machine learning training and inference across all input X and Y # Perform machine learning training and inference across all input X and Y
for i_iter in range(ITERATIONS): for i_iter in range(ITERATIONS):
# Execute an iteration of the algorithm
sq.eval() sq.eval()
# Calculate the parameters based on the respective derivatives calculated # Calculate the parameters based on the respective derivatives calculated