Searched refs:batch_matmul (Results 1 – 3 of 3) sorted by relevance
167 result_batched = tf.batch_matmul(batched_mat, batched_mat).eval()
1095 auto* batch_matmul = new BatchMatMulOperator; in ConvertBatchMatMulOperator() local1098 batch_matmul->adj_x = GetBoolAttr(node, "adj_x"); in ConvertBatchMatMulOperator()1101 batch_matmul->adj_y = GetBoolAttr(node, "adj_y"); in ConvertBatchMatMulOperator()1103 batch_matmul->inputs = {node.input(0), node.input(1)}; in ConvertBatchMatMulOperator()1104 batch_matmul->outputs = {node.name()}; in ConvertBatchMatMulOperator()1107 RetainTensorFlowNodeDef(node, batch_matmul); in ConvertBatchMatMulOperator()1109 model->operators.emplace_back(batch_matmul); in ConvertBatchMatMulOperator()
1662 * Performance optimizations for `batch_matmul` on multi-core CPUs.1734 * Improved multi-threaded performance of `batch_matmul`.