Fix CI (#1)

* Explicitly install pre-commit

* Run test on windows

* Cannot fetch model from github action

* Fix documentation build

* Remove gemma test

* typo

* test gemma back

.github/workflows/style.yml

@@ -20,5 +20,9 @@ jobs:
       run: |
         python -m pip install --upgrade pip
         pip install .[dev]
+    - name: Install pre-commit
+      run: |
+        pip install pre-commit
+        pre-commit install
     - name: Run tests
       run: pre-commit run --all-files

.github/workflows/test.yml

@@ -9,7 +9,7 @@ on:
 
 jobs:
   build:
-    runs-on: ubuntu-latest
+    runs-on: windows-latest
     strategy:
       matrix:
         python-version: ["3.8", "3.9", "3.10", "3.11", "3.12"]

docs/source/conf.py

@@ -30,7 +30,6 @@
     # 'sphinx.ext.autodoc',
     "sphinx.ext.napoleon",
     "breathe",
-    "sphinx_rtd_theme",
     "myst_parser",
 ]
 

test/python/test_optimizations.py

@@ -68,7 +68,7 @@ def get_model(model_name, hidden_size, intermediate_size, bias):
 
         return LlamaMLP(conf)
     elif model_name == "GemmaMLP":
-        conf = GemmaConfig.from_pretrained("google/gemma-2b-it")
+        conf = GemmaConfig()
         conf.num_hidden_layers = 1
         conf.hidden_size = hidden_size
         conf.head_dim = conf.hidden_size // conf.num_attention_heads
@@ -83,7 +83,7 @@ def get_model(model_name, hidden_size, intermediate_size, bias):
 
         return LlamaModel(conf)
     elif model_name == "GemmaModel":
-        conf = GemmaConfig.from_pretrained("google/gemma-2b-it")
+        conf = GemmaConfig()
         conf.num_hidden_layers = 1
         conf.hidden_size = hidden_size
         conf.head_dim = conf.hidden_size // conf.num_attention_heads