fix: markdown lint error and mergify bug

mergify wasn't checking markdown lint success
also changed e2e triggering criteria

Signed-off-by: Nathan Weinberg <nweinber@redhat.com>

.github/mergify.yml

@@ -28,6 +28,7 @@ pull_request_rules:
     # e2e workflow
     - or:
       - and:
+        # note this should match the triggering criteria in 'e2e-nvidia-t4-x1.yml'
         - check-success=e2e-workflow-complete
         - or:
           - files~=\.py$
@@ -40,12 +41,12 @@ pull_request_rules:
         - -files~=^requirements.*\.txt$
         - -files=.github/workflows/e2e-nvidia-t4-x1.yml
 
-    # lint must pass if files change that would trigger this job
+    # code lint workflow
     - or:
       - and:
+        # note this should match the triggering criteria in 'lint.yml'
         - check-success=lint-workflow-complete
         - or:
-          # see .github/workflows/lint.yml and test.yml
           - files~=\.py$
           - files=pyproject.toml
           - files~=^requirements.*\.txt$
@@ -60,6 +61,18 @@ pull_request_rules:
         - -files~=^scripts/[^/]+\.sh$
         - -files=.github/workflows/lint.yml
 
+    - or:
+      - and:
+        - check-success=markdown-lint
+        - or:
+          - files~=\.md$
+          - files=.markdownlint-cli2.yaml
+          - files=.github/workflows/docs.yml
+      - and:
+        - -files~=\.md$
+        - -files=.markdownlint-cli2.yaml
+        - -files=.github/workflows/docs.yml
+
   actions:
     merge:
       method: merge

.github/workflows/e2e-nvidia-t4-x1.yml

@@ -15,6 +15,12 @@ on:
     branches:
       - main
       - release-*
+    paths:
+      # note this should match the merging criteria in 'mergify.yml'
+      - '**.py'
+      - 'pyproject.toml'
+      - 'requirements**.txt'
+      - '.github/workflows/e2e-nvidia-t4-x1.yml' # Follow-on workflow
 
 concurrency:
   group: ${{ github.workflow }}-${{ github.event.pull_request.number || github.ref }}

README.md

@@ -130,6 +130,7 @@ They are listed below:
 
 > [!NOTE]
 > For `sharding_strategy` - Only `SHARD_GRAD_OP` has been extensively tested and is actively supported by this library.
+
 ### `loraOptions`
 
 LoRA options currently supported: