Huvu/mcore wan official

.github/ISSUE_TEMPLATE/bug_report.md

@@ -0,0 +1,28 @@
+---
+name: Bug report
+about: Create a report to help us improve the repository or project
+title: ""
+labels: bug
+assignees: ''
+
+---
+
+**Describe the bug**
+
+A clear and concise description of what the bug is.
+
+**Steps/Code to reproduce bug**
+
+Please list *minimal* steps or code snippet for us to be able to reproduce the bug.
+
+A helpful guide on on how to craft a minimal bug report http://matthewrocklin.com/blog/work/2018/02/28/minimal-bug-reports.
+
+
+**Expected behavior**
+
+A clear and concise description of what you expected to happen.
+
+
+**Additional context**
+
+Add any other context about the problem here.

.github/ISSUE_TEMPLATE/config.yml

@@ -0,0 +1,2 @@
+blank_issues_enabled: false
+

.github/ISSUE_TEMPLATE/feature_request.md

@@ -0,0 +1,20 @@
+---
+name: Feature request
+about: Suggest an idea for this project
+title: ""
+labels: enhancement
+assignees: ''
+
+---
+
+**Is your feature request related to a problem? Please describe.**
+A clear and concise description of what the problem is. Ex. I'm always frustrated when [...]
+
+**Describe the solution you'd like**
+A clear and concise description of what you want to happen.
+
+**Describe alternatives you've considered**
+A clear and concise description of any alternative solutions or features you've considered.
+
+**Additional context**
+Add any other context or screenshots about the feature request here.

.github/ISSUE_TEMPLATE/model-support-request.md

@@ -0,0 +1,31 @@
+---
+name: Model Support Request
+about: Request conversion support and training recipes for a new model
+title: "<Model name> Model Support"
+labels: ''
+assignees: ''
+
+---
+
+Add support for \<model name\> model:
+
+**Please include a link to the model's HuggingFace repo**
+HF repo:
+
+**These checklist items are required for all models in Megatron Bridge**
+
+- [ ] Model providers
+- [ ] Model bridge for HF conversion
+- [ ] Unit tests (config and bridge)
+- [ ] Model conversion functional tests
+
+**For flagship models, these items are also needed**
+
+- [ ] Optimal pretraining recipe
+- [ ] Optimal finetuning recipe
+- [ ] Recipe unit tests
+- [ ] Recipe functional tests
+- [ ] End to end CI tests
+
+**Additional context**
+Add any other context or screenshots about the model request here.

.github/workflows/build-docs.yml

@@ -23,7 +23,7 @@ on:
 
 jobs:
   pre-flight:
-    uses: NVIDIA-NeMo/FW-CI-templates/.github/workflows/_cicd_preflight.yml@v0.53.0
+    uses: NVIDIA-NeMo/FW-CI-templates/.github/workflows/_cicd_preflight.yml@v0.64.2
 
   build-docs:
     needs: [pre-flight]

.github/workflows/build-test-publish-wheel.yml

@@ -31,7 +31,7 @@ permissions:
 
 jobs:
   pre-flight:
-    uses: NVIDIA-NeMo/FW-CI-templates/.github/workflows/_cicd_preflight.yml@v0.53.0
+    uses: NVIDIA-NeMo/FW-CI-templates/.github/workflows/_cicd_preflight.yml@v0.64.2
 
   build-test-publish-wheel:
     needs: [pre-flight]

.github/workflows/cicd-main.yml

@@ -10,7 +10,7 @@
 # distributed under the License is distributed on an "AS IS" BASIS,
 # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 # See the License for the specific language governing permissions and
-# limitations under the License
+# limitations under the License.
 name: CICD NeMo
 on:
   schedule:
@@ -31,7 +31,7 @@ permissions:
 
 jobs:
   pre-flight:
-    uses: NVIDIA-NeMo/FW-CI-templates/.github/workflows/_cicd_preflight.yml@v0.53.0
+    uses: NVIDIA-NeMo/FW-CI-templates/.github/workflows/_cicd_preflight.yml@v0.64.2
 
   lint-check:
     name: Lint check

.github/workflows/copyright-check.yml

@@ -23,7 +23,7 @@ on:
 
 jobs:
   pre-flight:
-    uses: NVIDIA-NeMo/FW-CI-templates/.github/workflows/_cicd_preflight.yml@v0.53.0
+    uses: NVIDIA-NeMo/FW-CI-templates/.github/workflows/_cicd_preflight.yml@v0.64.2
 
   copyright-check:
     needs: [pre-flight]

.github/workflows/install-test.yml

@@ -26,20 +26,19 @@ on:
 
 jobs:
   pre-flight:
-    uses: NVIDIA-NeMo/FW-CI-templates/.github/workflows/_cicd_preflight.yml@v0.53.0
+    uses: NVIDIA-NeMo/FW-CI-templates/.github/workflows/_cicd_preflight.yml@v0.64.2
 
   pip-test-bare-metal:
     needs: [pre-flight]
     if: |
       !(needs.pre-flight.outputs.docs_only == 'true'
       || needs.pre-flight.outputs.is_deployment_workflow == 'true')
-    runs-on: ${{ matrix.arch }}
-    name: Pip - Python${{ matrix.python-version }} - ${{ matrix.arch == 'ubuntu-latest' && 'AMD64/Linux' || 'ARM64/Darwin' }} - Bare Metal
+    runs-on: linux-amd64-cpu16
+    name: Pip - Python${{ matrix.python-version }} - AMD64/Linux - Bare Metal
     container: ubuntu:24.04
     strategy:
       fail-fast: false
       matrix:
-        arch: ["ubuntu-latest"]
         python-version: ["3.10", "3.11", "3.12"]
     steps:
       - name: Checkout repository

docs/conf.py

@@ -27,7 +27,7 @@
 project = "Megatron Bridge"
 copyright = "2025, NVIDIA Corporation"
 author = "NVIDIA Corporation"
-release = "0.2.0"
+release = "0.1.0"
 
 # -- General configuration ---------------------------------------------------
 # https://www.sphinx-doc.org/en/master/usage/configuration.html#general-configuration

docs/index.md

@@ -7,6 +7,7 @@
 :hidden:
 
 parallelisms.md
+performance-summary.md
 performance-guide.md
 recipe-usage.md
 ```
@@ -37,6 +38,7 @@ training/attention-optimizations.md
 training/activation-recomputation.md
 training/cpu-offloading.md
 training/peft.md
+training/packed-sequences.md
 ```
 
 ```{toctree}

docs/performance-summary.md

@@ -0,0 +1,58 @@
+# Performance
+
+As part of the NVIDIA NeMo Framework, Megatron Bridge, provides optimal performance for training advanced generative AI models by incorporating the most recent training techniques, such as model parallelization, optimized attention mechanisms, and more, to achieve high training throughput.
+
+This page provides performance benchmarks for large language models using Megatron-Bridge across different GPU systems and configurations.
+
+## Nomenclature
+
+- **GBS**: Global Batch Size
+- **MBS**: Micro Batch Size
+- **FSDP**: Fully Sharded Data Parallel
+  - FSDP = 1: use FSDP
+  - FSDP = 0: use DDP (Distributed Data Parallel)
+- **TP**: Tensor Parallel Size
+- **PP**: Pipeline Parallel Size
+- **CP**: Context Parallel Size
+- **VP**: Virtual Pipeline Parallel Size
+- **EP**: Expert Parallel Size
+- **GA**: Number of Gradient Accumulations
+
+## Performance Metrics
+
+Performance is measured using:
+- **Tokens/sec/GPU**: Throughput per GPU
+- **Model TFLOP/sec/GPU**: Model floating-point operations per second per GPU
+
+```{contents}
+:local:
+:depth: 2
+```
+
+## Performance Summary for Large Language Models
+
+Below are performance benchmarks for various large language models organized by release version. These results were obtained using performance recipes available [here](https://github.com/NVIDIA/Megatron-Bridge/tree/main/scripts/performance).
+
+The performance data includes:
+
+- **Pre-training Performance**: Throughput metrics for various model sizes and architectures
+- **System Configurations**: Results across different GPU systems (DGX-GB200, DGX-B200, DGX-H100)
+- **Precision Options**: Performance comparisons between different precision modes (BF16, FP8, MXFP8)
+
+---
+
+## 25.09 NeMo Container
+
+### Pre-Training Performance
+
+#### System: DGX-GB200
+
+*Performance tables will be added here*
+
+#### System: DGX-B200
+
+*Performance tables will be added here*
+
+#### System: DGX-H100
+
+*Performance tables will be added here*

docs/project.json

@@ -1 +1,4 @@
-{"name": "megatron-bridge", "version": "0.2.0"}
+{
+    "name": "megatron-bridge",
+    "version": "0.1.0"
+}