Update pyo3 (#3202)

* Initial pyo3 update

* pyo3 onnx update

Author: ivarflakstad

candle-examples/Cargo.toml

@@ -27,7 +27,7 @@ intel-mkl-src = { workspace = true, optional = true }
 num-traits = { workspace = true }
 palette = { version = "0.7.6", optional = true }
 enterpolation = { version = "0.2.1", optional = true }
-pyo3 = { version = "0.22.0", features = [
+pyo3 = { version = "0.27", features = [
     "auto-initialize",
     "abi3-py311",
 ], optional = true }

candle-pyo3/Cargo.toml

@@ -20,11 +20,11 @@ candle-nn = { workspace = true }
 candle-onnx = { workspace = true, optional = true }
 half = { workspace = true }
 intel-mkl-src = { workspace = true, optional = true }
-pyo3 = { version = "0.22.0", features = ["extension-module", "abi3-py311"] }
+pyo3 = { version = "0.27", features = ["extension-module", "abi3-py313"] }
 float8 = { workspace = true }
 
 [build-dependencies]
-pyo3-build-config = "0.22"
+pyo3-build-config = "0.27"
 
 [features]
 default = []

candle-pyo3/quant-llama.py

@@ -1,6 +1,5 @@
 # This example shows how the candle Python api can be used to replicate llama.cpp.
 import sys
-from typing import Dict, Tuple, Any
 import candle
 from candle.models.llama import QuantizedLlama
 from candle import utils