Source code for onnx_extended._command_lines

import os
import re
import io
import logging
from typing import Any, Callable, Dict, List, Optional, Tuple, Union
import numpy as np
from onnx import (
    FunctionProto,
    GraphProto,
    ModelProto,
    TensorProto,
    SequenceProto,
    ValueInfoProto,
    load,
)
from onnx.helper import tensor_dtype_to_np_dtype
from onnx.numpy_helper import to_array
from onnx.reference import ReferenceEvaluator


def _type_shape(
    input_def: Union[str, ValueInfoProto]
) -> Tuple[Any, Tuple[Union[int, str], ...], Any]:
    if isinstance(input_def, str):
        reg = re.compile(
            "([a-z][a-z0-9]*)?([(]([ a-zA-Z,0-9]+)[)])?(:([A-Z][A-Z0-9]*))?"
        )
        search = reg.match(input_def)
        if search is None:
            raise ValueError(f"Unable to interpret string {input_def!r}.")
        grs = search.groups()
        dtype = grs[0]
        shape = None if grs[2] is None else grs[2].replace(" ", "").split(",")
        law = grs[-1]
        new_shape = []
        if shape is not None:
            for i in shape:
                try:
                    vi = int(i)
                    new_shape.append(vi)
                except ValueError:
                    new_shape.append(i)
            rshape = tuple(new_shape)
        else:
            rshape = None
        dt = None if dtype is None else getattr(np, dtype)
        return dt, rshape, law

    if isinstance(input_def, ValueInfoProto):
        try:
            ttype = input_def.type.tensor_type
        except AttributeError:
            raise ValueError(f"Unsupported input type {input_def!r}.")
        dt = ttype.elem_type
        new_shape = []
        for d in ttype.shape.dim:
            if d.dim_param:
                new_shape.append(d.dim_param)
            else:
                new_shape.append(d.dim_value)
        ndt = tensor_dtype_to_np_dtype(dt)
        return ndt, tuple(new_shape), None

    raise TypeError(f"Unexpected type {type(input_def)} for input_def.")


def _generate_random_inputs(
    dtype: Any,
    shape: Tuple[Union[int, str], ...],
    law: Optional[str] = None,
    dims: Optional[Dict[str, int]] = None,
) -> Tuple[np.ndarray, Dict[str, int]]:
    """
    Creates random or specific inputs.

    :param dtype: numpy dtype
    :param shape: expected shape
    :param law: law of the coefficients, default is 'U10', uniform law
    :param dims: letter are allowed, contains the named dimensions already
        mapped to a specific value
    :return: tuple (array, updated dims)

    Dimensions are modified inplace.
    """
    if dims is None:
        dims = {}
    if law is None:
        law = "U10"
    new_shape = []
    for sh in shape:
        if isinstance(sh, int):
            new_shape.append(sh)
        elif isinstance(sh, str):
            if sh not in dims:
                dims[sh] = 8
            new_shape.append(dims[sh])
    final_shape = tuple(new_shape)

    if law == "U10":
        res = np.random.random(final_shape).astype(dtype)
        return res, dims

    raise ValueError(f"Unexpected value for law={law!r}.")


[docs]def store_intermediate_results( model: Union[ModelProto, str], inputs: List[Union[str, np.ndarray, TensorProto]], out: str = ".", runtime: Union[type, str] = "CReferenceEvaluator", providers: Union[str, List[str]] = "CPU", verbose: int = 0, ): """ Executes an onnx model with a runtime and stores the intermediate results in a folder. See :class:`CReferenceEvaluator <onnx_extended.reference.CReferenceEvaluator>` for further details. :param model: path to a model or ModelProto :param inputs: list of inputs for the model :param out: output path :param runtime: runtime class to use :param providers: list of providers :param verbose: verbosity level :return: outputs """ if isinstance(model, str): if not os.path.exists(model): raise FileNotFoundError(f"File {model!r} does not exists.") if isinstance(providers, str): providers = [s.strip() for s in providers.split(",")] if runtime == "CReferenceEvaluator": from .reference import CReferenceEvaluator cls_runtime = CReferenceEvaluator add_providers = False elif hasattr(runtime, "run"): cls_runtime = runtime add_providers = True else: raise ValueError(f"Unexpected runtime {runtime!r}.") iv = int(verbose) if add_providers: inst = cls_runtime( model, providers=providers, save_intermediate=out, verbose=iv ) else: inst = cls_runtime(model, save_intermediate=out, verbose=iv) names = inst.input_names if len(names) < len(inputs): raise RuntimeError( f"There are more inputs ({len(inputs)}) " f"than names ({len(names)}). Names are {names}." ) dims = {} feeds = {} for i, (name, inp) in enumerate(zip(names, inputs)): if isinstance(inp, str) and os.path.exists(inp): with open(inp, "rb") as f: vect = f.read() tp = TensorProto() tp.ParseFromString(vect) value = to_array(tp) else: ty, shape, law = _type_shape(inp) if ty is None or shape is None: if isinstance(inst, ReferenceEvaluator): ty, shape, _ = _type_shape(inst.proto_.graph.input[i]) else: raise RuntimeError( f"shape or dtype is unknown and cannot " f"be retrieved from class {type(inst)}." ) value, dims = _generate_random_inputs(ty, shape, law, dims=dims) feeds[name] = value got = inst.run(None, feeds) return got
[docs]def display_intermediate_results( model: str, save: Optional[str] = None, tab: int = 12, external: bool = True, fprint: Callable = print, ): """ Displays shape, type for a model. :param model: a model :param save: save the results as a dataframe :param tab: column size for the output :param external: loads the external data or not :param fprint: function to print """ from .tools.onnx_nodes import enumerate_onnx_node_types if save is not None: ext = os.path.splitext(save)[-1] if ext not in {".csv", ".xlsx"}: raise ValueError(f"Unexpected format {save!r}, extension is {ext!r}.") else: exts = None def _fixed(s, length=10): if not isinstance(s, str): raise TypeError(f"Unexpected type {type(s)}: {s!r}.") return ( (s[: length - 1] + " ") if len(s) >= length - 1 else s + " " * (length - len(s)) ) n_rows = 0 rows = [] for obs in enumerate_onnx_node_types(model, external=external): if "level" not in obs: raise RuntimeError(f"Unexpected value obs={obs!r}.") indent = " " * obs["level"] * tab values = [ indent, _fixed(obs.get("kind", ""), tab), _fixed(obs.get("type", ""), tab), _fixed(obs.get("name", ""), tab), _fixed(obs.get("elem_type", ""), tab), _fixed(obs.get("shape", ""), tab), _fixed(obs.get("input_types", ""), tab), _fixed(obs.get("output_types", ""), tab), _fixed(obs.get("inputs", ""), tab), _fixed(obs.get("outputs", ""), tab), ] line = "".join(values) fprint(line) n_rows += 1 if save is not None: rows.append(obs) if n_rows == 0: if isinstance(model, str): raise RuntimeError(f"Model {model!r} is empty.") raise RuntimeError(f"Model type {type(model)} is empty.") if save is not None: from pandas import DataFrame df = DataFrame(rows) exts = {".csv": df.to_csv, ".xlsx": df.to_excel} exts[ext](save, index=False)
[docs]def cmd_quantize( model: Union[ModelProto, str], output: Optional[str] = None, kind: str = "fp8", scenario: str = "onnxruntime", early_stop: Optional[int] = None, quiet: bool = False, verbose: int = 0, index_transpose: int = 2, exceptions: Optional[List[Dict[str, str]]] = None, options: Optional["QuantizeOptions"] = None, # noqa: F821 ): """ Quantizes a model :param model: path to a model or ModelProto :param output: output file :param kind: kind of quantization :param scenario: depends on the quantization :param early_stop: stops early to see the preliminary results :param quiet: do not stop an exception :param verbose: verbosity level :param index_transpose: which input to transpose before calling gemm: 0 (none), 1 (first), 2 (second), 3 for both :param exceptions: exclude nodes from the quantization, `[{"name": "node_name1"}, {"name": "node_name2"}]` will exclude these two node names from the quantization :param options: quantization options, see class :class:`QuantizeOptions <onnx_extended.tools.graph.QuantizeOptions>` """ from .tools.graph import Graph, QuantizeOptions if options is None: options = QuantizeOptions.NONE if isinstance(model, str): if not os.path.exists(model): raise FileNotFoundError(f"Unable to find file {model!r}.") ext = os.path.splitext(model)[-1] if ext == ".onnx": with open(model, "rb") as f: proto_loaded = load(f) else: proto_loaded = model graph = Graph(proto_loaded) if verbose: logging.basicConfig( level=logging.WARN if verbose > 2 else (logging.DEBUG if verbose > 1 else logging.INFO) ) if kind == "fp8": from .tools.graph import quantize_float8 logger = logging.getLogger("onnx-extended") logger.info("Model initial size: %d", len(proto_loaded.SerializeToString())) new_graph = quantize_float8( graph, early_stop=early_stop or -1, quiet=quiet, version=scenario, index_transpose=index_transpose, exceptions=exceptions, quantize_options=options, ) if new_graph is None: logger.warning("No node was quantized.") return onx2 = new_graph.to_onnx() seq = onx2.SerializeToString() logger.info("Model quantized size: %d", len(seq)) if output is not None: with open(output, "wb") as f: f.write(seq) return if kind == "fp16": from .tools.graph import cast_constant logger = logging.getLogger("onnx-extended") logger.info("Model initial size: %d", len(proto_loaded.SerializeToString())) new_graph = cast_constant( graph, quiet=quiet, from_type=TensorProto.FLOAT, to_type=TensorProto.FLOAT16, ) if new_graph is None: logger.warning("No node was modified.") return onx2 = new_graph.to_onnx() seq = onx2.SerializeToString() logger.info("Model reduced size: %d", len(seq)) if output is not None: with open(output, "wb") as f: f.write(seq) return raise ValueError(f"Unexpected value {kind!r} for kind.")
[docs]def cmd_select( model: Union[ModelProto, str], save: Optional[str] = None, inputs: Optional[Union[str, List[str]]] = None, outputs: Optional[Union[str, List[str]]] = None, verbose: int = 0, ): """ Selects a subgraph in a model. :param model: path to a model or ModelProto :param save: model ot save in this file :param inputs: list of inputs or empty to keep the original inputs :param outputs: list of outputs or empty to keep the original outputs :param verbose: verbosity level """ from .tools.onnx_nodes import select_model_inputs_outputs if isinstance(model, str): if not os.path.exists(model): raise FileNotFoundError(f"Unable to find file {model!r}.") ext = os.path.splitext(model)[-1] if ext == ".onnx": with open(model, "rb") as f: proto_loaded = load(f) else: proto_loaded = model if verbose: logging.basicConfig( level=logging.WARN if verbose > 2 else (logging.DEBUG if verbose > 1 else logging.INFO) ) if isinstance(inputs, str): inputs = inputs.strip().split(",") if isinstance(outputs, str): outputs = outputs.strip().split(",") logger = logging.getLogger("onnx-extended") logger.info("Initial model size: %d", len(proto_loaded.SerializeToString())) onx2 = select_model_inputs_outputs( proto_loaded, inputs=inputs, outputs=outputs, verbose=verbose, ) seq = onx2.SerializeToString() logger.info("Selected model size: %d", len(seq)) if save is not None: with open(save, "wb") as f: f.write(seq)
def plot_profile( filename: str, kind: str, out_csv: Optional[str] = None, out_png: Optional[str] = None, title: Optional[str] = None, with_shape: bool = False, verbose: int = 0, ): """ Plots a profiling. :param filename: raw data to load :param kind: kind of plot to so, see below :param out_csv: output the data into that csv file :param out_png: output the graph in that file :param with_shape: consider input shape when showing results :param title: title (optional) :param verbose: verbosity, if > 0, prints out the data in csv format """ import matplotlib.pyplot as plt from .tools.js_profile import ( js_profile_to_dataframe, plot_ort_profile, _preprocess_graph1, _preprocess_graph2, ) if verbose: print(f"[plot_profile] load {filename!r}") if kind == "profile_op": df = js_profile_to_dataframe(filename, first_it_out=True, with_shape=with_shape) fig, ax = plt.subplots(1, 2, figsize=(10, 5)) plot_ort_profile(df, ax, title=title) df = _preprocess_graph1(df) elif kind == "profile_node": df = js_profile_to_dataframe( filename, first_it_out=True, agg=True, with_shape=with_shape ) fig, ax = plt.subplots(1, 1, figsize=(5, 5)) plot_ort_profile(df, ax, title=title) df = _preprocess_graph2(df) else: raise ValueError(f"Unexpected kind {kind:r}.") if verbose: st = io.StringIO() df.to_csv(st) print(st.getvalue()) if out_csv not in {"", None}: if verbose: print(f"[plot_profile] save {out_csv!r}") df.to_csv(out_csv) if out_png not in {"", None}: if verbose: print(f"[plot_profile] save {out_png!r}") fig.savefig(out_png)
[docs]def cmd_plot( filename: str, kind: str, out_csv: Optional[str] = None, out_png: Optional[str] = None, title: Optional[str] = None, with_shape: bool = False, verbose: int = 0, ): """ Plots a graph. :param filename: raw data to load :param kind: kind of plot to so, see below :param out_csv: output the data into that csv file :param out_png: output the graph in that file :param title: title (optional) :param with_shape: keep the shape to aggregate :param verbose: verbosity, if > 0, prints out the data in csv format Kinds of plots: * `'profile_op'`: draws the profiling per node type * `'profile_node'`: draws the profiling per node """ if not os.path.exists(filename): raise FileNotFoundError(f"Unable to find {filename!r}.") allowed = {"profile_op", "profile_node"} if kind in allowed: plot_profile(filename, kind, out_csv, out_png, title=title, verbose=verbose) else: raise ValueError(f"Unexpected kind {kind:r}, it should be {allowed}.")