tenstorrent · milank94 · Jul 18, 2024 · Apr 22, 2024 · May 20, 2024 · May 21, 2024
diff --git a/LICENSE b/LICENSE
@@ -198,3 +198,6 @@ distributed as part of the software:
 
 - Olli Huotari - [License available here](https://github.com/holli/yolov3_pytorch/blob/master/LICENSE)
 - transformers - [License available here](https://github.com/huggingface/transformers/blob/main/LICENSE)
+- monodle - [License available here](https://github.com/xinzhuma/monodle/blob/main/LICENSE)
+- NVIDIA Corporation - [License available here](https://github.com/NVIDIA/DeepLearningExamples/blob/master/PyTorch/Detection/SSD/LICENSE.md)
+- Meituan - [License available here](https://github.com/meituan/YOLOv6/blob/main/LICENSE)
diff --git a/check_copyright_config.yaml b/check_copyright_config.yaml
@@ -5,6 +5,7 @@ DEFAULT:
   allowed_licenses:
     - Apache-2.0
     - MIT
+    - GPL-3.0-only
 
   license_for_new_files: Apache-2.0  # License to be used when inserting a new copyright notice
 
@@ -29,4 +30,4 @@ ignore:  # You can also select ignoring files here
     - /CONTRIBUTING.md
     - /MAINTAINERS.md
     - /model_demos/nlp_demos/falcon/utils/configuration_RW.py
-
+    - /model_demos/cv_demos/monodle/utils/model.py
diff --git a/first_5_steps/1_install_tt_buda.md b/first_5_steps/1_install_tt_buda.md
@@ -40,17 +40,33 @@ PyBuda can be installed using two methods: Docker or Python virtualenv.
 
 If you would like to run PyBuda in a Docker container, then follow the instructions for [PCI Driver Installation](#pci-driver-installation) and [Device Firmware Update](#device-firmware-update) and followed by [Docker Container Installation](#docker-container-installation).
 
-If you would like to run PyBuda in a Python virtualenv, then follow the instructions for the [Setup HugePages](#setup-hugepages), [PCI Driver Installation](#pci-driver-installation), [Device Firmware Update](#device-firmware-update), and [Backend Compiler Dependencies](#backend-compiler-dependencies), followed by the [Tenstorrent Software Package](#tenstorrent-software-package).
+If you would like to run PyBuda in a Python virtualenv, then follow the instructions for the [Setup HugePages](#setup-hugepages), [PCI Driver Installation](#pci-driver-installation), [Device Firmware Update](#device-firmware-update), and [Backend Compiler Dependencies](#backend-compiler-dependencies), followed by the [Python Environment Installation](#python-environment-installation).
 
 ### Setup HugePages
 
-```bash
-NUM_DEVICES=$(lspci -d 1e52: | wc -l)
-sudo sed -i "s/^GRUB_CMDLINE_LINUX_DEFAULT=.*$/GRUB_CMDLINE_LINUX_DEFAULT=\"hugepagesz=1G hugepages=${NUM_DEVICES} nr_hugepages=${NUM_DEVICES} iommu=pt\"/g" /etc/default/grub
-sudo update-grub
-sudo sed -i "/\s\/dev\/hugepages-1G\s/d" /etc/fstab; echo "hugetlbfs /dev/hugepages-1G hugetlbfs pagesize=1G,rw,mode=777 0 0" | sudo tee -a /etc/fstab
-sudo reboot
-```
+1. Download latest [setup_hugepages.py](https://github.com/tenstorrent/tt-metal/blob/main/infra/machine_setup/scripts/setup_hugepages.py) script.
+
+    ```sh
+    wget https://raw.githubusercontent.com/tenstorrent/tt-metal/main/infra/machine_setup/scripts/setup_hugepages.py
+    ```
+
+2. Run first setup script.
+
+    ```sh
+    sudo -E python3 setup_hugepages.py first_pass
+    ```
+
+3. Reboot
+
+    ```sh
+    sudo reboot now
+    ```
+
+4. Run second setup script & check setup.
+
+    ```sh
+    sudo -E python3 setup_hugepages.py enable && sudo -E python3 setup_hugepages.py check
+    ```
 
 ### PCI Driver Installation
 
@@ -106,15 +122,17 @@ Python dependencies. Creating a new virtual environment with PyBuda and librarie
 
 #### Step 1. Navigate to [Releases](https://github.com/tenstorrent/tt-buda/releases)
 
-#### Step 2. Under the latest release, download the `pybuda` and `tvm` and `torchvison` wheel files
+#### Step 2. Scroll to find the latest release package in `.zip` format under "Assets" that corresponds to your device and operating system
 
-#### Step 3. Create your Python environment in desired directory
+#### Step 3. Download the `.zip` package and unzip to find the `pybuda`, `tvm` and `torchvison` wheel files
+
+#### Step 4. Create your Python environment in desired directory
 
 ```bash
 python3 -m venv env
 ```
 
-#### Step 4. Activate environment
+#### Step 5. Activate environment
 
 ```bash
 source env/bin/activate
@@ -143,7 +161,7 @@ This wheel file installs the Debuda tool designed for debugging purposes.
 
 ### Docker Container Installation
 
-Alternatively, PyBuda and its dependencies are provided as Docker images which can run in separate containers.
+Alternatively, PyBuda and its dependencies are provided as Docker images which can run in separate containers. The Docker containers can be found under: <https://github.com/orgs/tenstorrent/packages?repo_name=tt-buda>
 
 #### Step 1. Pull the docker image
 
@@ -153,26 +171,28 @@ To pull the Docker image, use the following command:
 sudo docker pull ghcr.io/tenstorrent/tt-buda/<OS-VERSION>/<TT-DEVICE>:<TAG>
 ```
 
-Supported OS Versions:
+Supported OS `<OS-VERSION>` Versions:
 
-- ubuntu-20-040amd64
-- ubuntu-22-040amd64
+- ubuntu-20-04-amd64
+- ubuntu-22-04-amd64
 
-Supported Tenstorrent Devices:
+Supported Tenstorrent `<TT-DEVICE>` Devices:
 
 - gs
-- wh
+- wh_b0
 
 For example, to run on an Ubuntu version 20.04 on a Grayskull device, use this command:
 
 ```bash
 sudo docker pull ghcr.io/tenstorrent/tt-buda/ubuntu-20-04-amd64/gs:<TAG>
 ```
 
+where `<TAG>` is the version number i.e. `v0.12.3`.
+
 #### Step 2. Run the container
 
 ```bash
-sudo docker run --rm -ti --cap-add=sys_nice --shm-size=4g --device /dev/tenstorrent -v /dev/hugepages-1G:/dev/hugepages-1G -v $(pwd)/:/home/ ghcr.io/tenstorrent/tt-buda/<TAG> bash
+sudo docker run --rm -ti --cap-add=sys_nice --shm-size=4g --device /dev/tenstorrent -v /dev/hugepages-1G:/dev/hugepages-1G -v $(pwd)/:/home/ ghcr.io/tenstorrent/tt-buda/<OS-VERSION>/<TT-DEVICE>:<TAG> bash
 ```
 
 #### Step 3. Change root directory

diff --git a/first_5_steps/2_running_nlp_models.ipynb b/first_5_steps/2_running_nlp_models.ipynb
@@ -11,7 +11,7 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "This notebook serves as a practical guide to getting started running Natural Language Processing (NLP) models on the E75 and E150 AI accelerator hardware using the TT-BUDA compiler stack.\n",
+    "This notebook serves as a practical guide to getting started running Natural Language Processing (NLP) models on Tenstorrent hardware devices using the TT-BUDA compiler stack. *For detailed information on model compatibility, please refer to the [models support table](../model_demos/README.md#models-support-table) to check which model works with which Tenstorrent device(s).*\n",
     "\n",
     "The tutorial will walk through an example of running the [BERT](https://en.wikipedia.org/wiki/BERT_(language_model)) model on Tenstorrent AI accelerator hardware. The model weights will be directly downloaded from the [HuggingFace library](https://huggingface.co/docs/transformers/model_doc/bert) and executed through the PyBUDA SDK.\n",
     "\n",
@@ -169,7 +169,23 @@
    "source": [
     "tt0 = pybuda.TTDevice(\n",
     "    name=\"tt_device_0\",  # here we can give our device any name we wish, for tracking purposes\n",
-    "    arch=pybuda.BackendDevice.Grayskull  # we set the target device architecture to compile for\n",
+    ")"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "In this example, `tt0` object is created without specifying the device architecture. Pybuda will automatically detectt and define the architecture based on which Tenstorrent device its run on.\n",
+    "\n",
+    "### Specifying the Architecture\n",
+    "If you want to specify the target device architecture, you can pass the `arch` parameter. Here’s how it can be done:\n",
+    "\n",
+    "```python\n",
+    "# Create a TTDevice instance with a specified architecture\n",
+    "tt0 = pybuda.TTDevice(\n",
+    "    name=\"tt_device_0\",  # You can give your device any name for tracking purposes\n",
+    "    arch=pybuda.BackendDevice.Grayskull  # Optionally set the target device architecture\n",
     ")"
    ]
   },

diff --git a/first_5_steps/3_running_cv_models.ipynb b/first_5_steps/3_running_cv_models.ipynb
@@ -11,7 +11,7 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "This notebook serves as a practical guide to getting started running Computer Vision (CV) models on the E75 and E150 AI accelerator hardware using the TT-BUDA compiler stack.\n",
+    "This notebook serves as a practical guide to getting started running Computer Vision (CV) models on Tenstorrent hardware devices using the TT-BUDA compiler stack. *For detailed information on model compatibility, please refer to the [models support table](../model_demos/README.md#models-support-table) to check which model works with which Tenstorrent device(s).*\n",
     "\n",
     "The tutorial will walk through an example of running the [ResNet](https://en.wikipedia.org/wiki/Residual_neural_network) model on Tenstorrent AI accelerator hardware. The model weights will be directly downloaded from the [HuggingFace library](https://huggingface.co/docs/transformers/model_doc/resnet) and executed through the PyBUDA SDK.\n",
     "\n",
@@ -158,7 +158,6 @@
    "source": [
     "tt0 = pybuda.TTDevice(\n",
     "    name=\"tt_device_0\",  # here we can give our device any name we wish, for tracking purposes\n",
-    "    arch=pybuda.BackendDevice.Grayskull  # we set the target device architecture to compile for\n",
     ")"
    ]
   },

diff --git a/first_5_steps/4_batched_inputs.ipynb b/first_5_steps/4_batched_inputs.ipynb
@@ -266,7 +266,6 @@
    "source": [
     "tt0 = pybuda.TTDevice(\n",
     "    name=\"tt_device_0\",  # here we can give our device any name we wish, for tracking purposes\n",
-    "    arch=pybuda.BackendDevice.Grayskull  # we set the target device architecture to compile for\n",
     ")"
    ]
   },

diff --git a/first_5_steps/5_serving_tt_models.ipynb b/first_5_steps/5_serving_tt_models.ipynb
@@ -142,7 +142,6 @@
     "        # Initialize TTDevice object\n",
     "        tt0 = pybuda.TTDevice(\n",
     "            name=\"tt_device_0\",  # here we can give our device any name we wish, for tracking purposes\n",
-    "            arch=pybuda.BackendDevice.Grayskull  # we set the target device architecture to compile for\n",
     "        )\n",
     "\n",
     "        # Create PyBUDA module\n",