From mboxrd@z Thu Jan 1 00:00:00 1970
Return-Path: <dev-bounces@dpdk.org>
Received: from mails.dpdk.org (mails.dpdk.org [217.70.189.124])
by inbox.dpdk.org (Postfix) with ESMTP id AD3FC43208;
Thu, 26 Oct 2023 14:51:28 +0200 (CEST)
Received: from mails.dpdk.org (localhost [127.0.0.1])
by mails.dpdk.org (Postfix) with ESMTP id 9738C42E2B;
Thu, 26 Oct 2023 14:49:49 +0200 (CEST)
Received: from mx0b-0016f401.pphosted.com (mx0a-0016f401.pphosted.com
[67.231.148.174])
by mails.dpdk.org (Postfix) with ESMTP id 86AEE42E0D
for <dev@dpdk.org>; Thu, 26 Oct 2023 14:49:48 +0200 (CEST)
Received: from pps.filterd (m0045849.ppops.net [127.0.0.1])
by mx0a-0016f401.pphosted.com (8.17.1.19/8.17.1.19) with ESMTP id
39Q7rfPe015316 for <dev@dpdk.org>; Thu, 26 Oct 2023 05:49:47 -0700
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=marvell.com;
h=from : to : cc :
subject : date : message-id : in-reply-to : references : mime-version :
content-transfer-encoding : content-type; s=pfpt0220;
bh=NDThJOUa3fgQjKFGsX0Ply07+5z67X7Up7Vgi0+MR5s=;
b=bUw0En0vaGVV1nJ8i2ZHAhBJgdxB1kufaspteq/bKZ5SW/TsUY1IEFXZdn00CiRaklXH
lI5PBq7elUCzr/qH+z+25G6xCp0Oue91CrySrBHkdrE7vZ/B57SIJv065Ql7gWh9oN0p
jaFbR5ccSvAQjRTdwLygiC+7yjpYi6DStawmIXw/RUdqRkVxbKGYbEw/Q/PCPvrx/UiA
DyQf8eKgg0kl04RuMhaDQJEI60JNxUFHwAg1vM21DpqLB0VLKAAjBbWknrI15LRgItqX
WjhSOpLlwG7IP1gyUcoSNvNkkienxjPh52LK0uLYCQ/SRHD8qAu38HXE8jgqM6l38p2z Hg==
Received: from dc5-exch01.marvell.com ([199.233.59.181])
by mx0a-0016f401.pphosted.com (PPS) with ESMTPS id 3ty0vrdrf4-1
(version=TLSv1.2 cipher=ECDHE-RSA-AES256-SHA384 bits=256 verify=NOT)
for <dev@dpdk.org>; Thu, 26 Oct 2023 05:49:46 -0700
Received: from DC5-EXCH01.marvell.com (10.69.176.38) by DC5-EXCH01.marvell.com
(10.69.176.38) with Microsoft SMTP Server (TLS) id 15.0.1497.48;
Thu, 26 Oct 2023 05:49:45 -0700
Received: from maili.marvell.com (10.69.176.80) by DC5-EXCH01.marvell.com
(10.69.176.38) with Microsoft SMTP Server id 15.0.1497.48 via Frontend
Transport; Thu, 26 Oct 2023 05:49:45 -0700
Received: from ml-host-33.caveonetworks.com (unknown [10.110.143.233])
by maili.marvell.com (Postfix) with ESMTP id C1D9C3F7097;
Thu, 26 Oct 2023 05:49:44 -0700 (PDT)
From: Srikanth Yalavarthi <syalavarthi@marvell.com>
To: Srikanth Yalavarthi <syalavarthi@marvell.com>
CC: <dev@dpdk.org>, <sshankarnara@marvell.com>, <aprabhu@marvell.com>,
<ptakkar@marvell.com>
Subject: [PATCH v2 1/1] app/mldev: enable support for pre-quantized I/O
Date: Thu, 26 Oct 2023 05:49:38 -0700
Message-ID: <20231026124938.23695-1-syalavarthi@marvell.com>
X-Mailer: git-send-email 2.42.0
In-Reply-To: <20231002100217.12456-1-syalavarthi@marvell.com>
References: <20231002100217.12456-1-syalavarthi@marvell.com>
MIME-Version: 1.0
Content-Transfer-Encoding: 8bit
Content-Type: text/plain
X-Proofpoint-GUID: O9qNoQbqUdWq9HctTAZSei30yGXde3CT
X-Proofpoint-ORIG-GUID: O9qNoQbqUdWq9HctTAZSei30yGXde3CT
X-Proofpoint-Virus-Version: vendor=baseguard
engine=ICAP:2.0.272,Aquarius:18.0.987,Hydra:6.0.619,FMLib:17.11.176.26
definitions=2023-10-26_10,2023-10-26_01,2023-05-22_02
X-BeenThere: dev@dpdk.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: DPDK patches and discussions <dev.dpdk.org>
List-Unsubscribe: <https://mails.dpdk.org/options/dev>,
<mailto:dev-request@dpdk.org?subject=unsubscribe>
List-Archive: <http://mails.dpdk.org/archives/dev/>
List-Post: <mailto:dev@dpdk.org>
List-Help: <mailto:dev-request@dpdk.org?subject=help>
List-Subscribe: <https://mails.dpdk.org/listinfo/dev>,
<mailto:dev-request@dpdk.org?subject=subscribe>
Errors-To: dev-bounces@dpdk.org
From: Anup Prabhu <aprabhu@marvell.com>
Enabled support for pre-quantized input and output in ML
test application.
Signed-off-by: Anup Prabhu <aprabhu@marvell.com>
---
v2:
- Updated application help
v1:
- Initial changes
app/test-mldev/ml_options.c | 11 +++++++++-
app/test-mldev/ml_options.h | 28 ++++++++++++++------------
app/test-mldev/test_inference_common.c | 20 ++++++++++++------
doc/guides/tools/testmldev.rst | 3 +++
4 files changed, 42 insertions(+), 20 deletions(-)
diff --git a/app/test-mldev/ml_options.c b/app/test-mldev/ml_options.c
index eeaffec399..320f6325ae 100644
--- a/app/test-mldev/ml_options.c
+++ b/app/test-mldev/ml_options.c
@@ -24,6 +24,7 @@ ml_options_default(struct ml_options *opt)
opt->dev_id = 0;
opt->socket_id = SOCKET_ID_ANY;
opt->nb_filelist = 0;
+ opt->quantized_io = false;
opt->repetitions = 1;
opt->burst_size = 1;
opt->queue_pairs = 1;
@@ -243,7 +244,8 @@ ml_dump_test_options(const char *testname)
"\t\t--queue_pairs : number of queue pairs to create\n"
"\t\t--queue_size : size of queue-pair\n"
"\t\t--tolerance : maximum tolerance (%%) for output validation\n"
- "\t\t--stats : enable reporting device and model statistics\n");
+ "\t\t--stats : enable reporting device and model statistics\n"
+ "\t\t--quantized_io : skip input/output quantization\n");
printf("\n");
}
}
@@ -269,6 +271,7 @@ static struct option lgopts[] = {
{ML_SOCKET_ID, 1, 0, 0},
{ML_MODELS, 1, 0, 0},
{ML_FILELIST, 1, 0, 0},
+ {ML_QUANTIZED_IO, 0, 0, 0},
{ML_REPETITIONS, 1, 0, 0},
{ML_BURST_SIZE, 1, 0, 0},
{ML_QUEUE_PAIRS, 1, 0, 0},
@@ -316,6 +319,11 @@ ml_options_parse(struct ml_options *opt, int argc, char **argv)
while ((opts = getopt_long(argc, argv, "", lgopts, &opt_idx)) != EOF) {
switch (opts) {
case 0: /* parse long options */
+ if (!strcmp(lgopts[opt_idx].name, "quantized_io")) {
+ opt->quantized_io = true;
+ break;
+ }
+
if (!strcmp(lgopts[opt_idx].name, "stats")) {
opt->stats = true;
break;
@@ -360,4 +368,5 @@ ml_options_dump(struct ml_options *opt)
ml_dump("socket_id", "%d", opt->socket_id);
ml_dump("debug", "%s", (opt->debug ? "true" : "false"));
+ ml_dump("quantized_io", "%s", (opt->quantized_io ? "true" : "false"));
}
diff --git a/app/test-mldev/ml_options.h b/app/test-mldev/ml_options.h
index 90e22adeac..edb9dba8f7 100644
--- a/app/test-mldev/ml_options.h
+++ b/app/test-mldev/ml_options.h
@@ -12,19 +12,20 @@
#define ML_TEST_MAX_MODELS 8
/* Options names */
-#define ML_TEST ("test")
-#define ML_DEVICE_ID ("dev_id")
-#define ML_SOCKET_ID ("socket_id")
-#define ML_MODELS ("models")
-#define ML_FILELIST ("filelist")
-#define ML_REPETITIONS ("repetitions")
-#define ML_BURST_SIZE ("burst_size")
-#define ML_QUEUE_PAIRS ("queue_pairs")
-#define ML_QUEUE_SIZE ("queue_size")
-#define ML_TOLERANCE ("tolerance")
-#define ML_STATS ("stats")
-#define ML_DEBUG ("debug")
-#define ML_HELP ("help")
+#define ML_TEST ("test")
+#define ML_DEVICE_ID ("dev_id")
+#define ML_SOCKET_ID ("socket_id")
+#define ML_MODELS ("models")
+#define ML_FILELIST ("filelist")
+#define ML_QUANTIZED_IO ("quantized_io")
+#define ML_REPETITIONS ("repetitions")
+#define ML_BURST_SIZE ("burst_size")
+#define ML_QUEUE_PAIRS ("queue_pairs")
+#define ML_QUEUE_SIZE ("queue_size")
+#define ML_TOLERANCE ("tolerance")
+#define ML_STATS ("stats")
+#define ML_DEBUG ("debug")
+#define ML_HELP ("help")
struct ml_filelist {
char model[PATH_MAX];
@@ -46,6 +47,7 @@ struct ml_options {
float tolerance;
bool stats;
bool debug;
+ bool quantized_io;
};
void ml_options_default(struct ml_options *opt);
diff --git a/app/test-mldev/test_inference_common.c b/app/test-mldev/test_inference_common.c
index 846f71abb1..36629210ee 100644
--- a/app/test-mldev/test_inference_common.c
+++ b/app/test-mldev/test_inference_common.c
@@ -777,14 +777,22 @@ ml_inference_iomem_setup(struct ml_test *test, struct ml_options *opt, uint16_t
}
t->model[fid].inp_dsize = 0;
- for (i = 0; i < t->model[fid].info.nb_inputs; i++)
- t->model[fid].inp_dsize +=
- t->model[fid].info.input_info[i].nb_elements * sizeof(float);
+ for (i = 0; i < t->model[fid].info.nb_inputs; i++) {
+ if (opt->quantized_io)
+ t->model[fid].inp_dsize += t->model[fid].info.input_info[i].size;
+ else
+ t->model[fid].inp_dsize +=
+ t->model[fid].info.input_info[i].nb_elements * sizeof(float);
+ }
t->model[fid].out_dsize = 0;
- for (i = 0; i < t->model[fid].info.nb_outputs; i++)
- t->model[fid].out_dsize +=
- t->model[fid].info.output_info[i].nb_elements * sizeof(float);
+ for (i = 0; i < t->model[fid].info.nb_outputs; i++) {
+ if (opt->quantized_io)
+ t->model[fid].out_dsize += t->model[fid].info.output_info[i].size;
+ else
+ t->model[fid].out_dsize +=
+ t->model[fid].info.output_info[i].nb_elements * sizeof(float);
+ }
/* allocate buffer for user data */
mz_size = t->model[fid].inp_dsize + t->model[fid].out_dsize;
diff --git a/doc/guides/tools/testmldev.rst b/doc/guides/tools/testmldev.rst
index 9b1565a457..55e26eed08 100644
--- a/doc/guides/tools/testmldev.rst
+++ b/doc/guides/tools/testmldev.rst
@@ -89,6 +89,9 @@ The following are the command-line options supported by the test application.
A suffix ``.q`` is appended to quantized output filename.
Maximum number of filelist entries supported by the test is ``8``.
+``--quantized_io``
+ Disable IO quantization and dequantization.
+
``--repetitions <n>``
Set the number of inference repetitions to be executed in the test per each model.
Default value is ``1``.
--
2.42.0