Refactor load/store before tensor cores (#1193)

* minor cleanups

* render_const

* now that's a nice refactor

* clean up vload/vstore

* clean up render_load

* debugs there

* dumb

* err, this?

* const float4

* what's failing

* bugfix

* statement includes semicolon

* bugfix
This commit is contained in:
George Hotz
2023-07-08 15:54:58 -07:00
committed by GitHub
parent ef1909500e
commit 7151382364
9 changed files with 85 additions and 90 deletions

View File

@@ -185,14 +185,14 @@ jobs:
python-version: 3.8
- name: Install Dependencies
run: pip install -e '.[testing]' --extra-index-url https://download.pytorch.org/whl/cpu
- name: Test openpilot model compile and size
run: |
DEBUG=2 ALLOWED_KERNEL_COUNT=199 FLOAT16=1 DEBUGCL=1 GPU=1 IMAGE=2 python3 openpilot/compile.py
python3 -c 'import os; assert os.path.getsize("/tmp/output.thneed") < 100_000_000'
- name: Test GPU IMAGE ops
run: |
GPU=1 IMAGE=1 python3 test/test_ops.py
FORWARD_ONLY=1 GPU=1 IMAGE=2 python3 test/test_ops.py
- name: Test openpilot model compile and size
run: |
ALLOWED_KERNEL_COUNT=199 FLOAT16=1 DEBUGCL=1 GPU=1 IMAGE=2 python3 openpilot/compile.py
python3 -c 'import os; assert os.path.getsize("/tmp/output.thneed") < 100_000_000'
- name: Test openpilot model correctness (float32)
run: DEBUGCL=1 GPU=1 IMAGE=2 python3 openpilot/compile.py