Sync with r2.12.1 (#1530)

* update driver version (#1429) * p0 ipex rn50 ATS-M (#1426) * add ipex stable diffusion * change base image P0 ITEX rn50 (#1431) * MaskRCNN ATS-M container (#1417) * p0 ipex stable diffusion (#1424) * yolov5 p0 ipex ATS-M (#1425) * itex atsm stable diffusion (#1418) * P0 ITEX Efficientnet B0,B3 (#1411) * EOLing docker builder files for workload containers (#1437) * removing dockerfiles directory * removed docker builder spec, partials * change precision to lowercase (#1456) * Update IPEX cpu baremetal instructions (#1451) * clean up ipex baremetal instructions * update horovod version in docs (#1458) * Remove all software.intel.com links (#1381) * Corrected software.intel.com * Removed dev catalog pages for EOL models * Added and updated baremetal README for P0 GPU models (#1447) * updated the GPU readme * PYT SPR BERT Large (#1472) * add avx-fp32 * Adapt newer BKC * remove idsid * update base image * updated tpp files for 2.12.1 release (#1479) * updated tpp files * added yolo5 * another update to TPPs (#1503) * resolve merge conflicts * Bump mlflow in /datasets/cloud_data_connector/samples/interoperability (#1492) Bumps [mlflow](https://github.com/mlflow/mlflow) from 2.5.0 to 2.6.0. * Bump mlflow in /datasets/cloud_data_connector/samples/azure (#1491) Bumps [mlflow](https://github.com/mlflow/mlflow) from 2.5.0 to 2.6.0. * fix issues with resolving conflicts * P0 models list (#1500) * sync with r2.12.1 --------- Co-authored-by: mahathis <36486206+Mahathi-Vatsal@users.noreply.github.com> Co-authored-by: Srikanth Ramakrishna <srikanth.ramakrishna@intel.com> Co-authored-by: Tyler Titsworth <tyler.titsworth@intel.com> Co-authored-by: Sharvil Shah <shahsharvil96@gmail.com> Co-authored-by: Jitendra Patil <jitendra.patil@intel.com>
intel · Sep 15, 2023 · 56789bd · 56789bd
1 parent 20a2fbb
commit 56789bd
Show file tree

Hide file tree

Showing 517 changed files with 12,035 additions and 22,284 deletions.
diff --git a/README.md b/README.md
@@ -2,7 +2,7 @@
 
 This repository contains **links to pre-trained models, sample scripts, best practices, and step-by-step tutorials** for many popular open-source machine learning models optimized by Intel to run on Intel® Xeon® Scalable processors and Intel® Data Center GPUs.
 
-Model packages and containers for running the Model Zoo's workloads can be found at the [Intel® Developer Catalog](https://software.intel.com/containers).
+Model packages and containers for running the Model Zoo's workloads can be found at the [Intel® Developer Catalog](https://www.intel.com/content/www/us/en/developer/tools/software-catalog/containers.html).
 
 ## Purpose of the Model Zoo
 
@@ -176,6 +176,11 @@ For best performance on Intel® Data Center GPU Flex and Max Series, please chec
 | [SSD-MobileNet*](https://arxiv.org/pdf/1704.04861.pdf)| TensorFlow | Inference | Flex Series| [Int8](/quickstart/object_detection/tensorflow/ssd-mobilenet/inference/gpu/README.md) |
 | [SSD-MobileNet](https://arxiv.org/pdf/1704.04861.pdf)| PyTorch | Inference | Flex Series | [Int8](/quickstart/object_detection/pytorch/ssd-mobilenet/inference/gpu/README.md) |
 | [Yolo V4](https://arxiv.org/pdf/1704.04861.pdf)| PyTorch | Inference | Flex Series | [Int8](/quickstart/object_detection/pytorch/yolov4/inference/gpu/README.md) |
+| [EfficientNet](https://arxiv.org/pdf/1905.11946.pdf)  | TensorFlow | Inference | Flex Series | [FP16](/quickstart/image_recognition/tensorflow/efficientnet/inference/gpu/README.md) |
+| [MaskRCNN](https://arxiv.org/pdf/1703.06870.pdf)  | TensorFlow | Inference | Flex Series | [FP16](/quickstart/image_segmentation/tensorflow/maskrcnn/inference/gpu/README.md) | 
+| [Stable Diffusion](https://arxiv.org/pdf/2112.10752.pdf)  | TensorFlow | Inference | Flex Series | [FP16 FP32](/quickstart/generative-ai/tensorflow/stable_diffusion/inference/gpu/README.md) |
+| [Stable Diffusion](https://arxiv.org/pdf/2112.10752.pdf)  | PyTorch | Inference | Flex Series | [FP16 FP32](/quickstart/generative-ai/pytorch/stable_diffusion/inference/gpu/README.md) |
+| [Yolo V5](https://arxiv.org/pdf/2108.11539.pdf)  | PyTorch | Inference | Flex Series | [FP16](/quickstart/object_detection/pytorch/yolov5/inference/gpu/README.md) | 
 | [ResNet 50v1.5](https://github.com/tensorflow/models/tree/v2.11.0/official/legacy/image_classification/resnet) | TensorFlow | Inference | Max Series | [Int8 FP32 FP16](/quickstart/image_recognition/tensorflow/resnet50v1_5/inference/gpu/README_Max_Series.md) |
 | [ResNet 50 v1.5](https://github.com/tensorflow/models/tree/v2.11.0/official/legacy/image_classification/resnet) | TensorFlow | Training | Max Series | [BFloat16](/quickstart/image_recognition/tensorflow/resnet50v1_5/training/gpu/README.md) |
 | [ResNet 50 v1.5](https://arxiv.org/pdf/1512.03385.pdf)    | PyTorch | Inference | Max Series |[Int8](/quickstart/image_recognition/pytorch/resnet50v1_5/inference/gpu/README_Max_Series.md) |

diff --git a/benchmarks/README.md b/benchmarks/README.md
diff --git a/benchmarks/common/tensorflow/start.sh b/benchmarks/common/tensorflow/start.sh
@@ -562,9 +562,6 @@ function bert_options() {
   if [[ -n "${OPTIMIZED_SOFTMAX}" && ${OPTIMIZED_SOFTMAX} != "" ]]; then
     CMD=" ${CMD} --optimized-softmax=${OPTIMIZED_SOFTMAX}"
   fi
-  if [[ -n "${AMP}" && ${AMP} != "" ]]; then
-    CMD=" ${CMD} --amp=${AMP}"
-  fi
 
   if [[ -n "${MPI_WORKERS_SYNC_GRADIENTS}" && ${MPI_WORKERS_SYNC_GRADIENTS} != "" ]]; then
     CMD=" ${CMD} --mpi_workers_sync_gradients=${MPI_WORKERS_SYNC_GRADIENTS}"
@@ -1418,6 +1415,38 @@ function transformer_mlperf() {
   fi
 }
 
+# GPT-J base model
+function gpt_j() {
+    if [ ${MODE} == "inference" ]; then
+      if [[ (${PRECISION} == "bfloat16") || ( ${PRECISION} == "fp32") || ( ${PRECISION} == "fp16") ]]; then
+        if [[ -z "${CHECKPOINT_DIRECTORY}" ]]; then
+          echo "Checkpoint directory not found. The script will download the model."
+        else
+          export PYTHONPATH=${PYTHONPATH}:${MOUNT_EXTERNAL_MODELS_SOURCE}
+          export HF_HOME=${CHECKPOINT_DIRECTORY}
+          export HUGGINGFACE_HUB_CACHE=${CHECKPOINT_DIRECTORY}
+          export TRANSFORMERS_CACHE=${CHECKPOINT_DIRECTORY}
+        fi
+
+        if [ ${BENCHMARK_ONLY} == "True" ]; then
+          CMD=" ${CMD} --max_output_tokens=${MAX_OUTPUT_TOKENS}"
+          CMD=" ${CMD} --input_tokens=${INPUT_TOKENS}"
+          if [[ -z "${SKIP_ROWS}" ]]; then
+            SKIP_ROWS=0
+          fi
+          CMD=" ${CMD} --skip_rows=${SKIP_ROWS}"
+        fi
+        CMD=${CMD} run_model
+      else
+        echo "PRECISION=${PRECISION} not supported for ${MODEL_NAME}."
+        exit 1
+      fi
+    else
+      echo "Only inference use-case is supported for now."
+      exit 1
+    fi
+}
+
 # Wavenet model
 function wavenet() {
   if [ ${PRECISION} == "fp32" ]; then
@@ -1563,6 +1592,189 @@ function distilbert_base() {
     fi
 }
 
+function gpt_j_6B() {
+    if [ ${PRECISION} == "fp32" ] || [ ${PRECISION} == "fp16" ] || 
+       [ ${PRECISION} == "bfloat16" ]; then
+
+      if [[ ${INSTALL_TRANSFORMER_FIX} != "True" ]]; then
+        echo "Information: Installing transformers from Hugging Face...!"
+        echo "python3 -m pip install git+https://github.com/intel-tensorflow/transformers@gptj_add_padding"
+        python3 -m pip install git+https://github.com/intel-tensorflow/transformers@gptj_add_padding
+      fi
+
+      export PYTHONPATH=${PYTHONPATH}:${MOUNT_EXTERNAL_MODELS_SOURCE}
+      CMD="${CMD} $(add_arg "--warmup-steps" ${WARMUP_STEPS})"
+      CMD="${CMD} $(add_arg "--steps" ${STEPS})"
+
+      if [[ ${MODE} == "training" ]]; then
+        if [[ -z "${TRAIN_OPTION}" ]]; then
+          echo "Error: Please specify a train option (GLUE, Lambada)"
+          exit 1
+        fi
+
+        CMD=" ${CMD} --train-option=${TRAIN_OPTION}"
+      fi
+
+      if [[ -z "${CACHE_DIR}" ]]; then
+          echo "Checkpoint directory not found. The script will download the model."
+      else
+          export HF_HOME=${CACHE_DIR}
+          export HUGGINGFACE_HUB_CACHE=${CACHE_DIR}
+          export TRANSFORMERS_CACHE=${CACHE_DIR}
+      fi
+
+      if [ ${NUM_INTER_THREADS} != "None" ]; then
+        CMD="${CMD} $(add_arg "--num-inter-threads" ${NUM_INTER_THREADS})"
+      fi
+
+      if [ ${NUM_INTRA_THREADS} != "None" ]; then
+        CMD="${CMD} $(add_arg "--num-intra-threads" ${NUM_INTRA_THREADS})"
+      fi
+
+      if [[ -n "${NUM_TRAIN_EPOCHS}" && ${NUM_TRAIN_EPOCHS} != "" ]]; then
+        CMD=" ${CMD} --num-train-epochs=${NUM_TRAIN_EPOCHS}"
+      fi
+
+      if [[ -n "${LEARNING_RATE}" && ${LEARNING_RATE} != "" ]]; then
+        CMD=" ${CMD} --learning-rate=${LEARNING_RATE}"
+      fi
+
+      if [[ -n "${NUM_TRAIN_STEPS}" && ${NUM_TRAIN_STEPS} != "" ]]; then
+        CMD=" ${CMD} --num-train-steps=${NUM_TRAIN_STEPS}"
+      fi
+
+      if [[ -n "${DO_TRAIN}" && ${DO_TRAIN} != "" ]]; then
+        CMD=" ${CMD} --do-train=${DO_TRAIN}"
+      fi
+
+      if [[ -n "${DO_EVAL}" && ${DO_EVAL} != "" ]]; then
+        CMD=" ${CMD} --do-eval=${DO_EVAL}"
+      fi
+
+      if [[ -n "${TASK_NAME}" && ${TASK_NAME} != "" ]]; then
+        CMD=" ${CMD} --task-name=${TASK_NAME}"
+      fi
+
+      if [[ -n "${CACHE_DIR}" && ${CACHE_DIR} != "" ]]; then
+        CMD=" ${CMD} --cache-dir=${CACHE_DIR}"
+      fi
+
+      if [[ -n "${PROFILE}" && ${PROFILE} != "" ]]; then
+        CMD=" ${CMD} --profile=${PROFILE}"
+      fi
+
+      if [ -z ${STEPS} ]; then
+        CMD="${CMD} $(add_arg "--steps" ${STEPS})"
+      fi
+
+      if [ -z $MAX_SEQ_LENGTH ]; then
+        CMD="${CMD} $(add_arg "--max-seq-length" ${MAX_SEQ_LENGTH})"
+      fi
+      CMD=${CMD} run_model
+    else
+      echo "PRECISION=${PRECISION} not supported for ${MODEL_NAME} in this repo."
+      exit 1
+    fi
+}
+
+
+# vision-transformer base model
+function vision_transformer() {
+
+    if [ ${MODE} == "training" ]; then
+	CMD="${CMD} $(add_arg "--init-checkpoint" ${INIT_CHECKPOINT})"
+    fi
+
+    if [ ${PRECISION} == "fp32" ] || [ ${PRECISION} == "bfloat16" ] ||
+       [ ${PRECISION} == "fp16" ]; then
+      export PYTHONPATH=${PYTHONPATH}:${MOUNT_EXTERNAL_MODELS_SOURCE}
+      CMD="${CMD} $(add_arg "--warmup-steps" ${WARMUP_STEPS})"
+      CMD="${CMD} $(add_arg "--steps" ${STEPS})"
+
+      if [ ${NUM_INTER_THREADS} != "None" ]; then
+        CMD="${CMD} $(add_arg "--num-inter-threads" ${NUM_INTER_THREADS})"
+      fi
+
+      if [ ${NUM_INTRA_THREADS} != "None" ]; then
+        CMD="${CMD} $(add_arg "--num-intra-threads" ${NUM_INTRA_THREADS})"
+      fi
+
+      if [ -z ${STEPS} ]; then
+        CMD="${CMD} $(add_arg "--steps" ${STEPS})"
+      fi
+      CMD=${CMD} run_model
+    else
+      echo "PRECISION=${PRECISION} not supported for ${MODEL_NAME} in this repo."
+      exit 1
+    fi
+}
+
+# mmoe base model
+function mmoe() {
+    if [ ${MODE} == "inference" ]; then
+      if [ ${PRECISION} == "fp32" ] || [ ${PRECISION} == "bfloat16" ] || [ ${PRECISION} == "fp16" ]; then
+        export PYTHONPATH=${PYTHONPATH}:${MOUNT_EXTERNAL_MODELS_SOURCE}
+        CMD="${CMD} $(add_arg "--warmup-steps" ${WARMUP_STEPS})"
+        CMD="${CMD} $(add_arg "--steps" ${STEPS})"
+
+        if [ ${NUM_INTER_THREADS} != "None" ]; then
+          CMD="${CMD} $(add_arg "--num-inter-threads" ${NUM_INTER_THREADS})"
+        fi
+
+        if [ ${NUM_INTRA_THREADS} != "None" ]; then
+          CMD="${CMD} $(add_arg "--num-intra-threads" ${NUM_INTRA_THREADS})"
+        fi
+
+        if [ -z ${STEPS} ]; then
+          CMD="${CMD} $(add_arg "--steps" ${STEPS})"
+        fi
+
+        CMD=${CMD} run_model
+      else
+        echo "PRECISION=${PRECISION} not supported for ${MODEL_NAME} in this repo."
+        exit 1
+      fi
+    elif [ ${MODE} == "training" ]; then
+      if [ ${PRECISION} == "fp32" ] || [ ${PRECISION} == "bfloat16" ] || [ ${PRECISION} == "fp16" ]; then
+        export PYTHONPATH=${PYTHONPATH}:${MOUNT_EXTERNAL_MODELS_SOURCE}
+        CMD="${CMD} $(add_arg "--train-epochs" ${TRAIN_EPOCHS})"
+        CMD="${CMD} $(add_arg "--model_dir" ${CHECKPOINT_DIRECTORY})"
+        CMD=${CMD} run_model
+      else
+        echo "PRECISION=${PRECISION} not supported for ${MODEL_NAME} in this repo."
+        exit 1
+      fi
+    fi
+}
+
+# rgat base model
+function rgat() {
+    if [ ${MODE} == "inference" ]; then
+      if [ ${PRECISION} == "fp32" ] || [ ${PRECISION} == "bfloat16" ] || [ ${PRECISION} == "fp16" ]; then
+        export PYTHONPATH=${PYTHONPATH}:${MOUNT_EXTERNAL_MODELS_SOURCE}
+
+        # Installing tensorflow_gnn from it's main branch
+        python3 -m pip install git+https://github.com/tensorflow/gnn.git@main
+
+        if [ ${NUM_INTER_THREADS} != "None" ]; then
+          CMD="${CMD} $(add_arg "--num-inter-threads" ${NUM_INTER_THREADS})"
+        fi
+
+        if [ ${NUM_INTRA_THREADS} != "None" ]; then
+          CMD="${CMD} $(add_arg "--num-intra-threads" ${NUM_INTRA_THREADS})"
+        fi
+
+        CMD="${CMD} $(add_arg "--graph-schema-path" ${GRAPH_SCHEMA_PATH})"
+        CMD="${CMD} $(add_arg "--pretrained-model" ${PRETRAINED_MODEL})"
+        CMD="${CMD} $(add_arg "--steps" ${STEPS})"
+        CMD=${CMD} run_model
+      else
+        echo "PRECISION=${PRECISION} not supported for ${MODEL_NAME} in this repo."
+        exit 1
+      fi
+    fi
+}
+
 # Wide & Deep model
 function wide_deep() {
     if [ ${PRECISION} == "fp32" ]; then
@@ -1643,6 +1855,29 @@ function wide_deep_large_ds() {
     fi
 }
 
+function graphsage() {
+    if [ ${MODE} == "inference" ]; then
+      if [ ${PRECISION} == "fp32" ] || [ ${PRECISION} == "bfloat16" ] || [ ${PRECISION} == "fp16" ]; then
+        export PYTHONPATH=${PYTHONPATH}:${MOUNT_EXTERNAL_MODELS_SOURCE}
+
+        if [ ${NUM_INTER_THREADS} != "None" ]; then
+          CMD="${CMD} $(add_arg "--num-inter-threads" ${NUM_INTER_THREADS})"
+        fi
+
+        if [ ${NUM_INTRA_THREADS} != "None" ]; then
+          CMD="${CMD} $(add_arg "--num-intra-threads" ${NUM_INTRA_THREADS})"
+        fi
+
+        CMD="${CMD} $(add_arg "--pretrained-model" ${PRETRAINED_MODEL})"
+        CMD="${CMD} $(add_arg "--steps" ${STEPS})"
+        CMD=${CMD} run_model
+      else
+        echo "PRECISION=${PRECISION} not supported for ${MODEL_NAME} in this repo."
+        exit 1
+      fi
+    fi
+}
+
 LOGFILE=${OUTPUT_DIR}/${LOG_FILENAME}
 
 MODEL_NAME=$(echo ${MODEL_NAME} | tr 'A-Z' 'a-z')
@@ -1707,7 +1942,19 @@ elif [ ${MODEL_NAME} == "bert_large" ]; then
 elif [ ${MODEL_NAME} == "dien" ]; then
   dien
 elif [ ${MODEL_NAME} == "distilbert_base" ]; then
-  distilbert_base  
+  distilbert_base 
+elif [ ${MODEL_NAME} == "vision_transformer" ]; then
+  vision_transformer 
+elif [ ${MODEL_NAME} == "gpt_j_6b" ]; then
+  gpt_j_6B
+elif [ ${MODEL_NAME} == "mmoe" ]; then
+  mmoe
+elif [ ${MODEL_NAME} == "graphsage" ]; then
+  graphsage
+elif [ ${MODEL_NAME} == "gpt_j" ]; then
+  gpt_j
+elif [ ${MODEL_NAME} == "rgat" ]; then
+  rgat
 else
   echo "Unsupported model: ${MODEL_NAME}"
   exit 1

diff --git a/benchmarks/image_recognition/tensorflow/densenet169/inference/README.md b/benchmarks/image_recognition/tensorflow/densenet169/inference/README.md
@@ -123,7 +123,3 @@ As an example, if the dataset location on Windows is `D:\user\ImageNet`, convert
 ## Additional Resources
 
 * To run more advanced use cases, see the instructions for the available precisions [FP32](fp32/Advanced.md) [<int8 precision>](<int8 advanced readme link>) [<bfloat16 precision>](<bfloat16 advanced readme link>) for calling the `launch_benchmark.py` script directly.
-* To run the model using docker, please see the [Intel® Developer Catalog](http://software.intel.com/containers)
-  workload container:<br />
-  [https://www.intel.com/content/www/us/en/developer/articles/machine-learning-model/densenet169-fp32-inference-tensorflow-model.html](https://www.intel.com/content/www/us/en/developer/articles/machine-learning-model/densenet169-fp32-inference-tensorflow-model.html).
-
diff --git a/benchmarks/image_recognition/tensorflow/inceptionv3/inference/README.md b/benchmarks/image_recognition/tensorflow/inceptionv3/inference/README.md
@@ -137,7 +137,7 @@ As an example, if the dataset location on Windows is `D:\user\ImageNet`, convert
 ## Additional Resources
 
 * To run more advanced use cases, see the instructions for the available precisions [FP32](fp32/Advanced.md) [Int8](int8/Advanced.md) [<bfloat16 precision>](<bfloat16 advanced readme link>) for calling the `launch_benchmark.py` script directly.
-* To run the model using docker, please see the [Intel® Developer Catalog](http://software.intel.com/containers)
+* To run the model using docker, please see the [Intel® Developer Catalog](https://www.intel.com/content/www/us/en/developer/tools/software-catalog/containers.html)
   workload container:<br />
-  [https://software.intel.com/content/www/us/en/develop/articles/containers/inceptionv3-fp32-inference-tensorflow-container.html](https://software.intel.com/content/www/us/en/develop/articles/containers/inceptionv3-fp32-inference-tensorflow-container.html).
+  [https://www.intel.com/content/www/us/en/developer/articles/containers/inceptionv3-fp32-inference-tensorflow-container.html](https://www.intel.com/content/www/us/en/developer/articles/containers/inceptionv3-fp32-inference-tensorflow-container.html).
 
diff --git a/benchmarks/image_recognition/tensorflow/inceptionv4/inference/README.md b/benchmarks/image_recognition/tensorflow/inceptionv4/inference/README.md
@@ -128,7 +128,7 @@ As an example, if the dataset location on Windows is `D:\user\ImageNet`, convert
 ## Additional Resources
 
 * To run more advanced use cases, see the instructions for the available precisions [FP32](fp32/Advanced.md) [Int8](int8/Advanced.md) [<bfloat16 precision>](<bfloat16 advanced readme link>) for calling the `launch_benchmark.py` script directly.
-* To run the model using docker, please see the [Intel® Developer Catalog](http://software.intel.com/containers)
+* To run the model using docker, please see the [Intel® Developer Catalog](https://www.intel.com/content/www/us/en/developer/tools/software-catalog/containers.html)
   workload container:<br />
-  [https://software.intel.com/content/www/us/en/develop/articles/containers/inceptionv4-fp32-inference-tensorflow-container.html](https://software.intel.com/content/www/us/en/develop/articles/containers/inceptionv4-fp32-inference-tensorflow-container.html).
+  [https://www.intel.com/content/www/us/en/developer/articles/containers/inceptionv4-fp32-inference-tensorflow-container.html](https://www.intel.com/content/www/us/en/developer/articles/containers/inceptionv4-fp32-inference-tensorflow-container.html).
 
diff --git a/benchmarks/image_recognition/tensorflow/mobilenet_v1/inference/README.md b/benchmarks/image_recognition/tensorflow/mobilenet_v1/inference/README.md
@@ -135,7 +135,7 @@ As an example, if the dataset location on Windows is `D:\user\ImageNet`, convert
 ## Additional Resources
 
 * To run more advanced use cases, see the instructions for the available precisions [FP32](fp32/Advanced.md) [Int8](int8/Advanced.md) [BFloat16](bfloat16/Advanced.md) for calling the `launch_benchmark.py` script directly.
-* To run the model using docker, please see the [Intel® Developer Catalog](http://software.intel.com/containers)
+* To run the model using docker, please see the [Intel® Developer Catalog](https://www.intel.com/content/www/us/en/developer/tools/software-catalog/containers.html)
   workload container:<br />
-  [https://software.intel.com/content/www/us/en/develop/articles/containers/mobilenetv1-fp32-inference-tensorflow-container.html](https://software.intel.com/content/www/us/en/develop/articles/containers/mobilenetv1-fp32-inference-tensorflow-container.html).
+  [https://www.intel.com/content/www/us/en/developer/articles/containers/mobilenetv1-fp32-inference-tensorflow-container.html](https://www.intel.com/content/www/us/en/developer/articles/containers/mobilenetv1-fp32-inference-tensorflow-container.html).
 
diff --git a/benchmarks/image_recognition/tensorflow/resnet101/inference/README.md b/benchmarks/image_recognition/tensorflow/resnet101/inference/README.md
@@ -135,7 +135,7 @@ As an example, if the dataset location on Windows is `D:\user\ImageNet`, convert
 ## Additional Resources
 
 * To run more advanced use cases, see the instructions for the available precisions [FP32](fp32/Advanced.md) [Int8](int8/Advanced.md) [<bfloat16 precision>](<bfloat16 advanced readme link>) for calling the `launch_benchmark.py` script directly.
-* To run the model using docker, please see the [Intel® Developer Catalog](http://software.intel.com/containers)
+* To run the model using docker, please see the [Intel® Developer Catalog](https://www.intel.com/content/www/us/en/developer/tools/software-catalog/containers.html)
   workload container:<br />
-  [https://software.intel.com/content/www/us/en/develop/articles/containers/resnet101-fp32-inference-tensorflow-container.html](https://software.intel.com/content/www/us/en/develop/articles/containers/resnet101-fp32-inference-tensorflow-container.html).
+  [https://www.intel.com/content/www/us/en/developer/articles/containers/resnet101-fp32-inference-tensorflow-container.html](https://www.intel.com/content/www/us/en/developer/articles/containers/resnet101-fp32-inference-tensorflow-container.html).
 
diff --git a/benchmarks/image_recognition/tensorflow/resnet50/inference/README.md b/benchmarks/image_recognition/tensorflow/resnet50/inference/README.md
@@ -135,7 +135,7 @@ As an example, if the dataset location on Windows is `D:\user\ImageNet`, convert
 ## Additional Resources
 
 * To run more advanced use cases, see the instructions for the available precisions [FP32](fp32/Advanced.md) [Int8](int8/Advanced.md) [<bfloat16 precision>](<bfloat16 advanced readme link>) for calling the `launch_benchmark.py` script directly.
-* To run the model using docker, please see the [Intel® Developer Catalog](http://software.intel.com/containers)
+* To run the model using docker, please see the [Intel® Developer Catalog](https://www.intel.com/content/www/us/en/developer/tools/software-catalog/containers.html)
   workload container:<br />
-  [https://software.intel.com/content/www/us/en/develop/articles/containers/resnet50-fp32-inference-tensorflow-container.html](https://software.intel.com/content/www/us/en/develop/articles/containers/resnet50-fp32-inference-tensorflow-container.html).
+  [https://www.intel.com/content/www/us/en/developer/articles/containers/resnet50-fp32-inference-tensorflow-container.html](https://www.intel.com/content/www/us/en/developer/articles/containers/resnet50-fp32-inference-tensorflow-container.html).