diff --git a/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/__pycache__/test_apply_mutate.cpython-310.pyc b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/__pycache__/test_apply_mutate.cpython-310.pyc new file mode 100644 index 0000000000000000000000000000000000000000..76171532824cd6172760442a84ad66b4483a0ac8 Binary files /dev/null and b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/__pycache__/test_apply_mutate.cpython-310.pyc differ diff --git a/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/aggregate/__init__.py b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/aggregate/__init__.py new file mode 100644 index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391 diff --git a/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/aggregate/__pycache__/__init__.cpython-310.pyc b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/aggregate/__pycache__/__init__.cpython-310.pyc new file mode 100644 index 0000000000000000000000000000000000000000..4574c28fc4ad8f76b1231e3db6f66c1ea0868bbf Binary files /dev/null and b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/aggregate/__pycache__/__init__.cpython-310.pyc differ diff --git a/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/aggregate/__pycache__/test_aggregate.cpython-310.pyc b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/aggregate/__pycache__/test_aggregate.cpython-310.pyc new file mode 100644 index 0000000000000000000000000000000000000000..7ef0033b07dcbf5aee7f5ba45476243fbb9ad0c4 Binary files /dev/null and b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/aggregate/__pycache__/test_aggregate.cpython-310.pyc differ diff --git a/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/aggregate/__pycache__/test_cython.cpython-310.pyc b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/aggregate/__pycache__/test_cython.cpython-310.pyc new file mode 100644 index 0000000000000000000000000000000000000000..5ee4f9a366fa4cdbf729aa8f700114d0c5a21bae Binary files /dev/null and b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/aggregate/__pycache__/test_cython.cpython-310.pyc differ diff --git a/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/aggregate/__pycache__/test_numba.cpython-310.pyc b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/aggregate/__pycache__/test_numba.cpython-310.pyc new file mode 100644 index 0000000000000000000000000000000000000000..bec132af47f1a91cea3a6127c99f69d8477b2c6e Binary files /dev/null and b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/aggregate/__pycache__/test_numba.cpython-310.pyc differ diff --git a/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/aggregate/__pycache__/test_other.cpython-310.pyc b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/aggregate/__pycache__/test_other.cpython-310.pyc new file mode 100644 index 0000000000000000000000000000000000000000..65f9ad1a520dc4eadfa58ae23333906eb8c7c62c Binary files /dev/null and b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/aggregate/__pycache__/test_other.cpython-310.pyc differ diff --git a/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/aggregate/test_aggregate.py b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/aggregate/test_aggregate.py new file mode 100644 index 0000000000000000000000000000000000000000..6223a153df3588840021210208b7250e48825552 --- /dev/null +++ b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/aggregate/test_aggregate.py @@ -0,0 +1,1672 @@ +""" +test .agg behavior / note that .apply is tested generally in test_groupby.py +""" +import datetime +import functools +from functools import partial +import re + +import numpy as np +import pytest + +from pandas.errors import SpecificationError + +from pandas.core.dtypes.common import is_integer_dtype + +import pandas as pd +from pandas import ( + DataFrame, + Index, + MultiIndex, + Series, + concat, + to_datetime, +) +import pandas._testing as tm +from pandas.core.groupby.grouper import Grouping + + +def test_groupby_agg_no_extra_calls(): + # GH#31760 + df = DataFrame({"key": ["a", "b", "c", "c"], "value": [1, 2, 3, 4]}) + gb = df.groupby("key")["value"] + + def dummy_func(x): + assert len(x) != 0 + return x.sum() + + gb.agg(dummy_func) + + +def test_agg_regression1(tsframe): + grouped = tsframe.groupby([lambda x: x.year, lambda x: x.month]) + result = grouped.agg("mean") + expected = grouped.mean() + tm.assert_frame_equal(result, expected) + + +def test_agg_must_agg(df): + grouped = df.groupby("A")["C"] + + msg = "Must produce aggregated value" + with pytest.raises(Exception, match=msg): + grouped.agg(lambda x: x.describe()) + with pytest.raises(Exception, match=msg): + grouped.agg(lambda x: x.index[:2]) + + +def test_agg_ser_multi_key(df): + f = lambda x: x.sum() + results = df.C.groupby([df.A, df.B]).aggregate(f) + expected = df.groupby(["A", "B"]).sum()["C"] + tm.assert_series_equal(results, expected) + + +def test_groupby_aggregation_mixed_dtype(): + # GH 6212 + expected = DataFrame( + { + "v1": [5, 5, 7, np.nan, 3, 3, 4, 1], + "v2": [55, 55, 77, np.nan, 33, 33, 44, 11], + }, + index=MultiIndex.from_tuples( + [ + (1, 95), + (1, 99), + (2, 95), + (2, 99), + ("big", "damp"), + ("blue", "dry"), + ("red", "red"), + ("red", "wet"), + ], + names=["by1", "by2"], + ), + ) + + df = DataFrame( + { + "v1": [1, 3, 5, 7, 8, 3, 5, np.nan, 4, 5, 7, 9], + "v2": [11, 33, 55, 77, 88, 33, 55, np.nan, 44, 55, 77, 99], + "by1": ["red", "blue", 1, 2, np.nan, "big", 1, 2, "red", 1, np.nan, 12], + "by2": [ + "wet", + "dry", + 99, + 95, + np.nan, + "damp", + 95, + 99, + "red", + 99, + np.nan, + np.nan, + ], + } + ) + + g = df.groupby(["by1", "by2"]) + result = g[["v1", "v2"]].mean() + tm.assert_frame_equal(result, expected) + + +def test_groupby_aggregation_multi_level_column(): + # GH 29772 + lst = [ + [True, True, True, False], + [True, False, np.nan, False], + [True, True, np.nan, False], + [True, True, np.nan, False], + ] + df = DataFrame( + data=lst, + columns=MultiIndex.from_tuples([("A", 0), ("A", 1), ("B", 0), ("B", 1)]), + ) + + msg = "DataFrame.groupby with axis=1 is deprecated" + with tm.assert_produces_warning(FutureWarning, match=msg): + gb = df.groupby(level=1, axis=1) + result = gb.sum(numeric_only=False) + expected = DataFrame({0: [2.0, True, True, True], 1: [1, 0, 1, 1]}) + + tm.assert_frame_equal(result, expected) + + +def test_agg_apply_corner(ts, tsframe): + # nothing to group, all NA + grouped = ts.groupby(ts * np.nan, group_keys=False) + assert ts.dtype == np.float64 + + # groupby float64 values results in a float64 Index + exp = Series([], dtype=np.float64, index=Index([], dtype=np.float64)) + tm.assert_series_equal(grouped.sum(), exp) + tm.assert_series_equal(grouped.agg("sum"), exp) + tm.assert_series_equal(grouped.apply("sum"), exp, check_index_type=False) + + # DataFrame + grouped = tsframe.groupby(tsframe["A"] * np.nan, group_keys=False) + exp_df = DataFrame( + columns=tsframe.columns, + dtype=float, + index=Index([], name="A", dtype=np.float64), + ) + tm.assert_frame_equal(grouped.sum(), exp_df) + tm.assert_frame_equal(grouped.agg("sum"), exp_df) + + msg = "The behavior of DataFrame.sum with axis=None is deprecated" + with tm.assert_produces_warning(FutureWarning, match=msg, check_stacklevel=False): + res = grouped.apply(np.sum) + tm.assert_frame_equal(res, exp_df) + + +def test_agg_grouping_is_list_tuple(ts): + df = DataFrame( + np.random.default_rng(2).standard_normal((30, 4)), + columns=Index(list("ABCD"), dtype=object), + index=pd.date_range("2000-01-01", periods=30, freq="B"), + ) + + grouped = df.groupby(lambda x: x.year) + grouper = grouped._grouper.groupings[0].grouping_vector + grouped._grouper.groupings[0] = Grouping(ts.index, list(grouper)) + + result = grouped.agg("mean") + expected = grouped.mean() + tm.assert_frame_equal(result, expected) + + grouped._grouper.groupings[0] = Grouping(ts.index, tuple(grouper)) + + result = grouped.agg("mean") + expected = grouped.mean() + tm.assert_frame_equal(result, expected) + + +def test_agg_python_multiindex(multiindex_dataframe_random_data): + grouped = multiindex_dataframe_random_data.groupby(["A", "B"]) + + result = grouped.agg("mean") + expected = grouped.mean() + tm.assert_frame_equal(result, expected) + + +@pytest.mark.parametrize( + "groupbyfunc", [lambda x: x.weekday(), [lambda x: x.month, lambda x: x.weekday()]] +) +def test_aggregate_str_func(tsframe, groupbyfunc): + grouped = tsframe.groupby(groupbyfunc) + + # single series + result = grouped["A"].agg("std") + expected = grouped["A"].std() + tm.assert_series_equal(result, expected) + + # group frame by function name + result = grouped.aggregate("var") + expected = grouped.var() + tm.assert_frame_equal(result, expected) + + # group frame by function dict + result = grouped.agg({"A": "var", "B": "std", "C": "mean", "D": "sem"}) + expected = DataFrame( + { + "A": grouped["A"].var(), + "B": grouped["B"].std(), + "C": grouped["C"].mean(), + "D": grouped["D"].sem(), + } + ) + tm.assert_frame_equal(result, expected) + + +def test_std_masked_dtype(any_numeric_ea_dtype): + # GH#35516 + df = DataFrame( + { + "a": [2, 1, 1, 1, 2, 2, 1], + "b": Series([pd.NA, 1, 2, 1, 1, 1, 2], dtype="Float64"), + } + ) + result = df.groupby("a").std() + expected = DataFrame( + {"b": [0.57735, 0]}, index=Index([1, 2], name="a"), dtype="Float64" + ) + tm.assert_frame_equal(result, expected) + + +def test_agg_str_with_kwarg_axis_1_raises(df, reduction_func): + gb = df.groupby(level=0) + warn_msg = f"DataFrameGroupBy.{reduction_func} with axis=1 is deprecated" + if reduction_func in ("idxmax", "idxmin"): + error = TypeError + msg = "'[<>]' not supported between instances of 'float' and 'str'" + warn = FutureWarning + else: + error = ValueError + msg = f"Operation {reduction_func} does not support axis=1" + warn = None + with pytest.raises(error, match=msg): + with tm.assert_produces_warning(warn, match=warn_msg): + gb.agg(reduction_func, axis=1) + + +@pytest.mark.parametrize( + "func, expected, dtype, result_dtype_dict", + [ + ("sum", [5, 7, 9], "int64", {}), + ("std", [4.5**0.5] * 3, int, {"i": float, "j": float, "k": float}), + ("var", [4.5] * 3, int, {"i": float, "j": float, "k": float}), + ("sum", [5, 7, 9], "Int64", {"j": "int64"}), + ("std", [4.5**0.5] * 3, "Int64", {"i": float, "j": float, "k": float}), + ("var", [4.5] * 3, "Int64", {"i": "float64", "j": "float64", "k": "float64"}), + ], +) +def test_multiindex_groupby_mixed_cols_axis1(func, expected, dtype, result_dtype_dict): + # GH#43209 + df = DataFrame( + [[1, 2, 3, 4, 5, 6]] * 3, + columns=MultiIndex.from_product([["a", "b"], ["i", "j", "k"]]), + ).astype({("a", "j"): dtype, ("b", "j"): dtype}) + + msg = "DataFrame.groupby with axis=1 is deprecated" + with tm.assert_produces_warning(FutureWarning, match=msg): + gb = df.groupby(level=1, axis=1) + result = gb.agg(func) + expected = DataFrame([expected] * 3, columns=["i", "j", "k"]).astype( + result_dtype_dict + ) + + tm.assert_frame_equal(result, expected) + + +@pytest.mark.parametrize( + "func, expected_data, result_dtype_dict", + [ + ("sum", [[2, 4], [10, 12], [18, 20]], {10: "int64", 20: "int64"}), + # std should ideally return Int64 / Float64 #43330 + ("std", [[2**0.5] * 2] * 3, "float64"), + ("var", [[2] * 2] * 3, {10: "float64", 20: "float64"}), + ], +) +def test_groupby_mixed_cols_axis1(func, expected_data, result_dtype_dict): + # GH#43209 + df = DataFrame( + np.arange(12).reshape(3, 4), + index=Index([0, 1, 0], name="y"), + columns=Index([10, 20, 10, 20], name="x"), + dtype="int64", + ).astype({10: "Int64"}) + + msg = "DataFrame.groupby with axis=1 is deprecated" + with tm.assert_produces_warning(FutureWarning, match=msg): + gb = df.groupby("x", axis=1) + result = gb.agg(func) + expected = DataFrame( + data=expected_data, + index=Index([0, 1, 0], name="y"), + columns=Index([10, 20], name="x"), + ).astype(result_dtype_dict) + tm.assert_frame_equal(result, expected) + + +def test_aggregate_item_by_item(df): + grouped = df.groupby("A") + + aggfun_0 = lambda ser: ser.size + result = grouped.agg(aggfun_0) + foosum = (df.A == "foo").sum() + barsum = (df.A == "bar").sum() + K = len(result.columns) + + # GH5782 + exp = Series(np.array([foosum] * K), index=list("BCD"), name="foo") + tm.assert_series_equal(result.xs("foo"), exp) + + exp = Series(np.array([barsum] * K), index=list("BCD"), name="bar") + tm.assert_almost_equal(result.xs("bar"), exp) + + def aggfun_1(ser): + return ser.size + + result = DataFrame().groupby(df.A).agg(aggfun_1) + assert isinstance(result, DataFrame) + assert len(result) == 0 + + +def test_wrap_agg_out(three_group): + grouped = three_group.groupby(["A", "B"]) + + def func(ser): + if ser.dtype == object: + raise TypeError("Test error message") + return ser.sum() + + with pytest.raises(TypeError, match="Test error message"): + grouped.aggregate(func) + result = grouped[["D", "E", "F"]].aggregate(func) + exp_grouped = three_group.loc[:, ["A", "B", "D", "E", "F"]] + expected = exp_grouped.groupby(["A", "B"]).aggregate(func) + tm.assert_frame_equal(result, expected) + + +def test_agg_multiple_functions_maintain_order(df): + # GH #610 + funcs = [("mean", np.mean), ("max", np.max), ("min", np.min)] + msg = "is currently using SeriesGroupBy.mean" + with tm.assert_produces_warning(FutureWarning, match=msg): + result = df.groupby("A")["C"].agg(funcs) + exp_cols = Index(["mean", "max", "min"]) + + tm.assert_index_equal(result.columns, exp_cols) + + +def test_series_index_name(df): + grouped = df.loc[:, ["C"]].groupby(df["A"]) + result = grouped.agg(lambda x: x.mean()) + assert result.index.name == "A" + + +def test_agg_multiple_functions_same_name(): + # GH 30880 + df = DataFrame( + np.random.default_rng(2).standard_normal((1000, 3)), + index=pd.date_range("1/1/2012", freq="s", periods=1000), + columns=["A", "B", "C"], + ) + result = df.resample("3min").agg( + {"A": [partial(np.quantile, q=0.9999), partial(np.quantile, q=0.1111)]} + ) + expected_index = pd.date_range("1/1/2012", freq="3min", periods=6) + expected_columns = MultiIndex.from_tuples([("A", "quantile"), ("A", "quantile")]) + expected_values = np.array( + [df.resample("3min").A.quantile(q=q).values for q in [0.9999, 0.1111]] + ).T + expected = DataFrame( + expected_values, columns=expected_columns, index=expected_index + ) + tm.assert_frame_equal(result, expected) + + +def test_agg_multiple_functions_same_name_with_ohlc_present(): + # GH 30880 + # ohlc expands dimensions, so different test to the above is required. + df = DataFrame( + np.random.default_rng(2).standard_normal((1000, 3)), + index=pd.date_range("1/1/2012", freq="s", periods=1000, name="dti"), + columns=Index(["A", "B", "C"], name="alpha"), + ) + result = df.resample("3min").agg( + {"A": ["ohlc", partial(np.quantile, q=0.9999), partial(np.quantile, q=0.1111)]} + ) + expected_index = pd.date_range("1/1/2012", freq="3min", periods=6, name="dti") + expected_columns = MultiIndex.from_tuples( + [ + ("A", "ohlc", "open"), + ("A", "ohlc", "high"), + ("A", "ohlc", "low"), + ("A", "ohlc", "close"), + ("A", "quantile", "A"), + ("A", "quantile", "A"), + ], + names=["alpha", None, None], + ) + non_ohlc_expected_values = np.array( + [df.resample("3min").A.quantile(q=q).values for q in [0.9999, 0.1111]] + ).T + expected_values = np.hstack( + [df.resample("3min").A.ohlc(), non_ohlc_expected_values] + ) + expected = DataFrame( + expected_values, columns=expected_columns, index=expected_index + ) + tm.assert_frame_equal(result, expected) + + +def test_multiple_functions_tuples_and_non_tuples(df): + # #1359 + # Columns B and C would cause partial failure + df = df.drop(columns=["B", "C"]) + + funcs = [("foo", "mean"), "std"] + ex_funcs = [("foo", "mean"), ("std", "std")] + + result = df.groupby("A")["D"].agg(funcs) + expected = df.groupby("A")["D"].agg(ex_funcs) + tm.assert_frame_equal(result, expected) + + result = df.groupby("A").agg(funcs) + expected = df.groupby("A").agg(ex_funcs) + tm.assert_frame_equal(result, expected) + + +def test_more_flexible_frame_multi_function(df): + grouped = df.groupby("A") + + exmean = grouped.agg({"C": "mean", "D": "mean"}) + exstd = grouped.agg({"C": "std", "D": "std"}) + + expected = concat([exmean, exstd], keys=["mean", "std"], axis=1) + expected = expected.swaplevel(0, 1, axis=1).sort_index(level=0, axis=1) + + d = {"C": ["mean", "std"], "D": ["mean", "std"]} + result = grouped.aggregate(d) + + tm.assert_frame_equal(result, expected) + + # be careful + result = grouped.aggregate({"C": "mean", "D": ["mean", "std"]}) + expected = grouped.aggregate({"C": "mean", "D": ["mean", "std"]}) + tm.assert_frame_equal(result, expected) + + def numpymean(x): + return np.mean(x) + + def numpystd(x): + return np.std(x, ddof=1) + + # this uses column selection & renaming + msg = r"nested renamer is not supported" + with pytest.raises(SpecificationError, match=msg): + d = {"C": "mean", "D": {"foo": "mean", "bar": "std"}} + grouped.aggregate(d) + + # But without renaming, these functions are OK + d = {"C": ["mean"], "D": [numpymean, numpystd]} + grouped.aggregate(d) + + +def test_multi_function_flexible_mix(df): + # GH #1268 + grouped = df.groupby("A") + + # Expected + d = {"C": {"foo": "mean", "bar": "std"}, "D": {"sum": "sum"}} + # this uses column selection & renaming + msg = r"nested renamer is not supported" + with pytest.raises(SpecificationError, match=msg): + grouped.aggregate(d) + + # Test 1 + d = {"C": {"foo": "mean", "bar": "std"}, "D": "sum"} + # this uses column selection & renaming + with pytest.raises(SpecificationError, match=msg): + grouped.aggregate(d) + + # Test 2 + d = {"C": {"foo": "mean", "bar": "std"}, "D": "sum"} + # this uses column selection & renaming + with pytest.raises(SpecificationError, match=msg): + grouped.aggregate(d) + + +def test_groupby_agg_coercing_bools(): + # issue 14873 + dat = DataFrame({"a": [1, 1, 2, 2], "b": [0, 1, 2, 3], "c": [None, None, 1, 1]}) + gp = dat.groupby("a") + + index = Index([1, 2], name="a") + + result = gp["b"].aggregate(lambda x: (x != 0).all()) + expected = Series([False, True], index=index, name="b") + tm.assert_series_equal(result, expected) + + result = gp["c"].aggregate(lambda x: x.isnull().all()) + expected = Series([True, False], index=index, name="c") + tm.assert_series_equal(result, expected) + + +def test_groupby_agg_dict_with_getitem(): + # issue 25471 + dat = DataFrame({"A": ["A", "A", "B", "B", "B"], "B": [1, 2, 1, 1, 2]}) + result = dat.groupby("A")[["B"]].agg({"B": "sum"}) + + expected = DataFrame({"B": [3, 4]}, index=["A", "B"]).rename_axis("A", axis=0) + + tm.assert_frame_equal(result, expected) + + +def test_groupby_agg_dict_dup_columns(): + # GH#55006 + df = DataFrame( + [[1, 2, 3, 4], [1, 3, 4, 5], [2, 4, 5, 6]], + columns=["a", "b", "c", "c"], + ) + gb = df.groupby("a") + result = gb.agg({"b": "sum"}) + expected = DataFrame({"b": [5, 4]}, index=Index([1, 2], name="a")) + tm.assert_frame_equal(result, expected) + + +@pytest.mark.parametrize( + "op", + [ + lambda x: x.sum(), + lambda x: x.cumsum(), + lambda x: x.transform("sum"), + lambda x: x.transform("cumsum"), + lambda x: x.agg("sum"), + lambda x: x.agg("cumsum"), + ], +) +def test_bool_agg_dtype(op): + # GH 7001 + # Bool sum aggregations result in int + df = DataFrame({"a": [1, 1], "b": [False, True]}) + s = df.set_index("a")["b"] + + result = op(df.groupby("a"))["b"].dtype + assert is_integer_dtype(result) + + result = op(s.groupby("a")).dtype + assert is_integer_dtype(result) + + +@pytest.mark.parametrize( + "keys, agg_index", + [ + (["a"], Index([1], name="a")), + (["a", "b"], MultiIndex([[1], [2]], [[0], [0]], names=["a", "b"])), + ], +) +@pytest.mark.parametrize( + "input_dtype", ["bool", "int32", "int64", "float32", "float64"] +) +@pytest.mark.parametrize( + "result_dtype", ["bool", "int32", "int64", "float32", "float64"] +) +@pytest.mark.parametrize("method", ["apply", "aggregate", "transform"]) +def test_callable_result_dtype_frame( + keys, agg_index, input_dtype, result_dtype, method +): + # GH 21240 + df = DataFrame({"a": [1], "b": [2], "c": [True]}) + df["c"] = df["c"].astype(input_dtype) + op = getattr(df.groupby(keys)[["c"]], method) + result = op(lambda x: x.astype(result_dtype).iloc[0]) + expected_index = pd.RangeIndex(0, 1) if method == "transform" else agg_index + expected = DataFrame({"c": [df["c"].iloc[0]]}, index=expected_index).astype( + result_dtype + ) + if method == "apply": + expected.columns.names = [0] + tm.assert_frame_equal(result, expected) + + +@pytest.mark.parametrize( + "keys, agg_index", + [ + (["a"], Index([1], name="a")), + (["a", "b"], MultiIndex([[1], [2]], [[0], [0]], names=["a", "b"])), + ], +) +@pytest.mark.parametrize("input", [True, 1, 1.0]) +@pytest.mark.parametrize("dtype", [bool, int, float]) +@pytest.mark.parametrize("method", ["apply", "aggregate", "transform"]) +def test_callable_result_dtype_series(keys, agg_index, input, dtype, method): + # GH 21240 + df = DataFrame({"a": [1], "b": [2], "c": [input]}) + op = getattr(df.groupby(keys)["c"], method) + result = op(lambda x: x.astype(dtype).iloc[0]) + expected_index = pd.RangeIndex(0, 1) if method == "transform" else agg_index + expected = Series([df["c"].iloc[0]], index=expected_index, name="c").astype(dtype) + tm.assert_series_equal(result, expected) + + +def test_order_aggregate_multiple_funcs(): + # GH 25692 + df = DataFrame({"A": [1, 1, 2, 2], "B": [1, 2, 3, 4]}) + + res = df.groupby("A").agg(["sum", "max", "mean", "ohlc", "min"]) + result = res.columns.levels[1] + + expected = Index(["sum", "max", "mean", "ohlc", "min"]) + + tm.assert_index_equal(result, expected) + + +def test_ohlc_ea_dtypes(any_numeric_ea_dtype): + # GH#37493 + df = DataFrame( + {"a": [1, 1, 2, 3, 4, 4], "b": [22, 11, pd.NA, 10, 20, pd.NA]}, + dtype=any_numeric_ea_dtype, + ) + gb = df.groupby("a") + result = gb.ohlc() + expected = DataFrame( + [[22, 22, 11, 11], [pd.NA] * 4, [10] * 4, [20] * 4], + columns=MultiIndex.from_product([["b"], ["open", "high", "low", "close"]]), + index=Index([1, 2, 3, 4], dtype=any_numeric_ea_dtype, name="a"), + dtype=any_numeric_ea_dtype, + ) + tm.assert_frame_equal(result, expected) + + gb2 = df.groupby("a", as_index=False) + result2 = gb2.ohlc() + expected2 = expected.reset_index() + tm.assert_frame_equal(result2, expected2) + + +@pytest.mark.parametrize("dtype", [np.int64, np.uint64]) +@pytest.mark.parametrize("how", ["first", "last", "min", "max", "mean", "median"]) +def test_uint64_type_handling(dtype, how): + # GH 26310 + df = DataFrame({"x": 6903052872240755750, "y": [1, 2]}) + expected = df.groupby("y").agg({"x": how}) + df.x = df.x.astype(dtype) + result = df.groupby("y").agg({"x": how}) + if how not in ("mean", "median"): + # mean and median always result in floats + result.x = result.x.astype(np.int64) + tm.assert_frame_equal(result, expected, check_exact=True) + + +def test_func_duplicates_raises(): + # GH28426 + msg = "Function names" + df = DataFrame({"A": [0, 0, 1, 1], "B": [1, 2, 3, 4]}) + with pytest.raises(SpecificationError, match=msg): + df.groupby("A").agg(["min", "min"]) + + +@pytest.mark.parametrize( + "index", + [ + pd.CategoricalIndex(list("abc")), + pd.interval_range(0, 3), + pd.period_range("2020", periods=3, freq="D"), + MultiIndex.from_tuples([("a", 0), ("a", 1), ("b", 0)]), + ], +) +def test_agg_index_has_complex_internals(index): + # GH 31223 + df = DataFrame({"group": [1, 1, 2], "value": [0, 1, 0]}, index=index) + result = df.groupby("group").agg({"value": Series.nunique}) + expected = DataFrame({"group": [1, 2], "value": [2, 1]}).set_index("group") + tm.assert_frame_equal(result, expected) + + +def test_agg_split_block(): + # https://github.com/pandas-dev/pandas/issues/31522 + df = DataFrame( + { + "key1": ["a", "a", "b", "b", "a"], + "key2": ["one", "two", "one", "two", "one"], + "key3": ["three", "three", "three", "six", "six"], + } + ) + result = df.groupby("key1").min() + expected = DataFrame( + {"key2": ["one", "one"], "key3": ["six", "six"]}, + index=Index(["a", "b"], name="key1"), + ) + tm.assert_frame_equal(result, expected) + + +def test_agg_split_object_part_datetime(): + # https://github.com/pandas-dev/pandas/pull/31616 + df = DataFrame( + { + "A": pd.date_range("2000", periods=4), + "B": ["a", "b", "c", "d"], + "C": [1, 2, 3, 4], + "D": ["b", "c", "d", "e"], + "E": pd.date_range("2000", periods=4), + "F": [1, 2, 3, 4], + } + ).astype(object) + result = df.groupby([0, 0, 0, 0]).min() + expected = DataFrame( + { + "A": [pd.Timestamp("2000")], + "B": ["a"], + "C": [1], + "D": ["b"], + "E": [pd.Timestamp("2000")], + "F": [1], + }, + index=np.array([0]), + dtype=object, + ) + tm.assert_frame_equal(result, expected) + + +class TestNamedAggregationSeries: + def test_series_named_agg(self): + df = Series([1, 2, 3, 4]) + gr = df.groupby([0, 0, 1, 1]) + result = gr.agg(a="sum", b="min") + expected = DataFrame( + {"a": [3, 7], "b": [1, 3]}, columns=["a", "b"], index=np.array([0, 1]) + ) + tm.assert_frame_equal(result, expected) + + result = gr.agg(b="min", a="sum") + expected = expected[["b", "a"]] + tm.assert_frame_equal(result, expected) + + def test_no_args_raises(self): + gr = Series([1, 2]).groupby([0, 1]) + with pytest.raises(TypeError, match="Must provide"): + gr.agg() + + # but we do allow this + result = gr.agg([]) + expected = DataFrame(columns=[]) + tm.assert_frame_equal(result, expected) + + def test_series_named_agg_duplicates_no_raises(self): + # GH28426 + gr = Series([1, 2, 3]).groupby([0, 0, 1]) + grouped = gr.agg(a="sum", b="sum") + expected = DataFrame({"a": [3, 3], "b": [3, 3]}, index=np.array([0, 1])) + tm.assert_frame_equal(expected, grouped) + + def test_mangled(self): + gr = Series([1, 2, 3]).groupby([0, 0, 1]) + result = gr.agg(a=lambda x: 0, b=lambda x: 1) + expected = DataFrame({"a": [0, 0], "b": [1, 1]}, index=np.array([0, 1])) + tm.assert_frame_equal(result, expected) + + @pytest.mark.parametrize( + "inp", + [ + pd.NamedAgg(column="anything", aggfunc="min"), + ("anything", "min"), + ["anything", "min"], + ], + ) + def test_named_agg_nametuple(self, inp): + # GH34422 + s = Series([1, 1, 2, 2, 3, 3, 4, 5]) + msg = f"func is expected but received {type(inp).__name__}" + with pytest.raises(TypeError, match=msg): + s.groupby(s.values).agg(a=inp) + + +class TestNamedAggregationDataFrame: + def test_agg_relabel(self): + df = DataFrame( + {"group": ["a", "a", "b", "b"], "A": [0, 1, 2, 3], "B": [5, 6, 7, 8]} + ) + result = df.groupby("group").agg(a_max=("A", "max"), b_max=("B", "max")) + expected = DataFrame( + {"a_max": [1, 3], "b_max": [6, 8]}, + index=Index(["a", "b"], name="group"), + columns=["a_max", "b_max"], + ) + tm.assert_frame_equal(result, expected) + + # order invariance + p98 = functools.partial(np.percentile, q=98) + result = df.groupby("group").agg( + b_min=("B", "min"), + a_min=("A", "min"), + a_mean=("A", "mean"), + a_max=("A", "max"), + b_max=("B", "max"), + a_98=("A", p98), + ) + expected = DataFrame( + { + "b_min": [5, 7], + "a_min": [0, 2], + "a_mean": [0.5, 2.5], + "a_max": [1, 3], + "b_max": [6, 8], + "a_98": [0.98, 2.98], + }, + index=Index(["a", "b"], name="group"), + columns=["b_min", "a_min", "a_mean", "a_max", "b_max", "a_98"], + ) + tm.assert_frame_equal(result, expected) + + def test_agg_relabel_non_identifier(self): + df = DataFrame( + {"group": ["a", "a", "b", "b"], "A": [0, 1, 2, 3], "B": [5, 6, 7, 8]} + ) + + result = df.groupby("group").agg(**{"my col": ("A", "max")}) + expected = DataFrame({"my col": [1, 3]}, index=Index(["a", "b"], name="group")) + tm.assert_frame_equal(result, expected) + + def test_duplicate_no_raises(self): + # GH 28426, if use same input function on same column, + # no error should raise + df = DataFrame({"A": [0, 0, 1, 1], "B": [1, 2, 3, 4]}) + + grouped = df.groupby("A").agg(a=("B", "min"), b=("B", "min")) + expected = DataFrame({"a": [1, 3], "b": [1, 3]}, index=Index([0, 1], name="A")) + tm.assert_frame_equal(grouped, expected) + + quant50 = functools.partial(np.percentile, q=50) + quant70 = functools.partial(np.percentile, q=70) + quant50.__name__ = "quant50" + quant70.__name__ = "quant70" + + test = DataFrame({"col1": ["a", "a", "b", "b", "b"], "col2": [1, 2, 3, 4, 5]}) + + grouped = test.groupby("col1").agg( + quantile_50=("col2", quant50), quantile_70=("col2", quant70) + ) + expected = DataFrame( + {"quantile_50": [1.5, 4.0], "quantile_70": [1.7, 4.4]}, + index=Index(["a", "b"], name="col1"), + ) + tm.assert_frame_equal(grouped, expected) + + def test_agg_relabel_with_level(self): + df = DataFrame( + {"A": [0, 0, 1, 1], "B": [1, 2, 3, 4]}, + index=MultiIndex.from_product([["A", "B"], ["a", "b"]]), + ) + result = df.groupby(level=0).agg( + aa=("A", "max"), bb=("A", "min"), cc=("B", "mean") + ) + expected = DataFrame( + {"aa": [0, 1], "bb": [0, 1], "cc": [1.5, 3.5]}, index=["A", "B"] + ) + tm.assert_frame_equal(result, expected) + + def test_agg_relabel_other_raises(self): + df = DataFrame({"A": [0, 0, 1], "B": [1, 2, 3]}) + grouped = df.groupby("A") + match = "Must provide" + with pytest.raises(TypeError, match=match): + grouped.agg(foo=1) + + with pytest.raises(TypeError, match=match): + grouped.agg() + + with pytest.raises(TypeError, match=match): + grouped.agg(a=("B", "max"), b=(1, 2, 3)) + + def test_missing_raises(self): + df = DataFrame({"A": [0, 1], "B": [1, 2]}) + match = re.escape("Column(s) ['C'] do not exist") + with pytest.raises(KeyError, match=match): + df.groupby("A").agg(c=("C", "sum")) + + def test_agg_namedtuple(self): + df = DataFrame({"A": [0, 1], "B": [1, 2]}) + result = df.groupby("A").agg( + b=pd.NamedAgg("B", "sum"), c=pd.NamedAgg(column="B", aggfunc="count") + ) + expected = df.groupby("A").agg(b=("B", "sum"), c=("B", "count")) + tm.assert_frame_equal(result, expected) + + def test_mangled(self): + df = DataFrame({"A": [0, 1], "B": [1, 2], "C": [3, 4]}) + result = df.groupby("A").agg(b=("B", lambda x: 0), c=("C", lambda x: 1)) + expected = DataFrame({"b": [0, 0], "c": [1, 1]}, index=Index([0, 1], name="A")) + tm.assert_frame_equal(result, expected) + + +@pytest.mark.parametrize( + "agg_col1, agg_col2, agg_col3, agg_result1, agg_result2, agg_result3", + [ + ( + (("y", "A"), "max"), + (("y", "A"), np.mean), + (("y", "B"), "mean"), + [1, 3], + [0.5, 2.5], + [5.5, 7.5], + ), + ( + (("y", "A"), lambda x: max(x)), + (("y", "A"), lambda x: 1), + (("y", "B"), np.mean), + [1, 3], + [1, 1], + [5.5, 7.5], + ), + ( + pd.NamedAgg(("y", "A"), "max"), + pd.NamedAgg(("y", "B"), np.mean), + pd.NamedAgg(("y", "A"), lambda x: 1), + [1, 3], + [5.5, 7.5], + [1, 1], + ), + ], +) +def test_agg_relabel_multiindex_column( + agg_col1, agg_col2, agg_col3, agg_result1, agg_result2, agg_result3 +): + # GH 29422, add tests for multiindex column cases + df = DataFrame( + {"group": ["a", "a", "b", "b"], "A": [0, 1, 2, 3], "B": [5, 6, 7, 8]} + ) + df.columns = MultiIndex.from_tuples([("x", "group"), ("y", "A"), ("y", "B")]) + idx = Index(["a", "b"], name=("x", "group")) + + result = df.groupby(("x", "group")).agg(a_max=(("y", "A"), "max")) + expected = DataFrame({"a_max": [1, 3]}, index=idx) + tm.assert_frame_equal(result, expected) + + msg = "is currently using SeriesGroupBy.mean" + with tm.assert_produces_warning(FutureWarning, match=msg): + result = df.groupby(("x", "group")).agg( + col_1=agg_col1, col_2=agg_col2, col_3=agg_col3 + ) + expected = DataFrame( + {"col_1": agg_result1, "col_2": agg_result2, "col_3": agg_result3}, index=idx + ) + tm.assert_frame_equal(result, expected) + + +def test_agg_relabel_multiindex_raises_not_exist(): + # GH 29422, add test for raises scenario when aggregate column does not exist + df = DataFrame( + {"group": ["a", "a", "b", "b"], "A": [0, 1, 2, 3], "B": [5, 6, 7, 8]} + ) + df.columns = MultiIndex.from_tuples([("x", "group"), ("y", "A"), ("y", "B")]) + + with pytest.raises(KeyError, match="do not exist"): + df.groupby(("x", "group")).agg(a=(("Y", "a"), "max")) + + +def test_agg_relabel_multiindex_duplicates(): + # GH29422, add test for raises scenario when getting duplicates + # GH28426, after this change, duplicates should also work if the relabelling is + # different + df = DataFrame( + {"group": ["a", "a", "b", "b"], "A": [0, 1, 2, 3], "B": [5, 6, 7, 8]} + ) + df.columns = MultiIndex.from_tuples([("x", "group"), ("y", "A"), ("y", "B")]) + + result = df.groupby(("x", "group")).agg( + a=(("y", "A"), "min"), b=(("y", "A"), "min") + ) + idx = Index(["a", "b"], name=("x", "group")) + expected = DataFrame({"a": [0, 2], "b": [0, 2]}, index=idx) + tm.assert_frame_equal(result, expected) + + +@pytest.mark.parametrize("kwargs", [{"c": ["min"]}, {"b": [], "c": ["min"]}]) +def test_groupby_aggregate_empty_key(kwargs): + # GH: 32580 + df = DataFrame({"a": [1, 1, 2], "b": [1, 2, 3], "c": [1, 2, 4]}) + result = df.groupby("a").agg(kwargs) + expected = DataFrame( + [1, 4], + index=Index([1, 2], dtype="int64", name="a"), + columns=MultiIndex.from_tuples([["c", "min"]]), + ) + tm.assert_frame_equal(result, expected) + + +def test_groupby_aggregate_empty_key_empty_return(): + # GH: 32580 Check if everything works, when return is empty + df = DataFrame({"a": [1, 1, 2], "b": [1, 2, 3], "c": [1, 2, 4]}) + result = df.groupby("a").agg({"b": []}) + expected = DataFrame(columns=MultiIndex(levels=[["b"], []], codes=[[], []])) + tm.assert_frame_equal(result, expected) + + +def test_groupby_aggregate_empty_with_multiindex_frame(): + # GH 39178 + df = DataFrame(columns=["a", "b", "c"]) + result = df.groupby(["a", "b"], group_keys=False).agg(d=("c", list)) + expected = DataFrame( + columns=["d"], index=MultiIndex([[], []], [[], []], names=["a", "b"]) + ) + tm.assert_frame_equal(result, expected) + + +def test_grouby_agg_loses_results_with_as_index_false_relabel(): + # GH 32240: When the aggregate function relabels column names and + # as_index=False is specified, the results are dropped. + + df = DataFrame( + {"key": ["x", "y", "z", "x", "y", "z"], "val": [1.0, 0.8, 2.0, 3.0, 3.6, 0.75]} + ) + + grouped = df.groupby("key", as_index=False) + result = grouped.agg(min_val=pd.NamedAgg(column="val", aggfunc="min")) + expected = DataFrame({"key": ["x", "y", "z"], "min_val": [1.0, 0.8, 0.75]}) + tm.assert_frame_equal(result, expected) + + +def test_grouby_agg_loses_results_with_as_index_false_relabel_multiindex(): + # GH 32240: When the aggregate function relabels column names and + # as_index=False is specified, the results are dropped. Check if + # multiindex is returned in the right order + + df = DataFrame( + { + "key": ["x", "y", "x", "y", "x", "x"], + "key1": ["a", "b", "c", "b", "a", "c"], + "val": [1.0, 0.8, 2.0, 3.0, 3.6, 0.75], + } + ) + + grouped = df.groupby(["key", "key1"], as_index=False) + result = grouped.agg(min_val=pd.NamedAgg(column="val", aggfunc="min")) + expected = DataFrame( + {"key": ["x", "x", "y"], "key1": ["a", "c", "b"], "min_val": [1.0, 0.75, 0.8]} + ) + tm.assert_frame_equal(result, expected) + + +@pytest.mark.parametrize( + "func", [lambda s: s.mean(), lambda s: np.mean(s), lambda s: np.nanmean(s)] +) +def test_multiindex_custom_func(func): + # GH 31777 + data = [[1, 4, 2], [5, 7, 1]] + df = DataFrame( + data, + columns=MultiIndex.from_arrays( + [[1, 1, 2], [3, 4, 3]], names=["Sisko", "Janeway"] + ), + ) + result = df.groupby(np.array([0, 1])).agg(func) + expected_dict = { + (1, 3): {0: 1.0, 1: 5.0}, + (1, 4): {0: 4.0, 1: 7.0}, + (2, 3): {0: 2.0, 1: 1.0}, + } + expected = DataFrame(expected_dict, index=np.array([0, 1]), columns=df.columns) + tm.assert_frame_equal(result, expected) + + +def myfunc(s): + return np.percentile(s, q=0.90) + + +@pytest.mark.parametrize("func", [lambda s: np.percentile(s, q=0.90), myfunc]) +def test_lambda_named_agg(func): + # see gh-28467 + animals = DataFrame( + { + "kind": ["cat", "dog", "cat", "dog"], + "height": [9.1, 6.0, 9.5, 34.0], + "weight": [7.9, 7.5, 9.9, 198.0], + } + ) + + result = animals.groupby("kind").agg( + mean_height=("height", "mean"), perc90=("height", func) + ) + expected = DataFrame( + [[9.3, 9.1036], [20.0, 6.252]], + columns=["mean_height", "perc90"], + index=Index(["cat", "dog"], name="kind"), + ) + + tm.assert_frame_equal(result, expected) + + +def test_aggregate_mixed_types(): + # GH 16916 + df = DataFrame( + data=np.array([0] * 9).reshape(3, 3), columns=list("XYZ"), index=list("abc") + ) + df["grouping"] = ["group 1", "group 1", 2] + result = df.groupby("grouping").aggregate(lambda x: x.tolist()) + expected_data = [[[0], [0], [0]], [[0, 0], [0, 0], [0, 0]]] + expected = DataFrame( + expected_data, + index=Index([2, "group 1"], dtype="object", name="grouping"), + columns=Index(["X", "Y", "Z"], dtype="object"), + ) + tm.assert_frame_equal(result, expected) + + +@pytest.mark.xfail(reason="Not implemented;see GH 31256") +def test_aggregate_udf_na_extension_type(): + # https://github.com/pandas-dev/pandas/pull/31359 + # This is currently failing to cast back to Int64Dtype. + # The presence of the NA causes two problems + # 1. NA is not an instance of Int64Dtype.type (numpy.int64) + # 2. The presence of an NA forces object type, so the non-NA values is + # a Python int rather than a NumPy int64. Python ints aren't + # instances of numpy.int64. + def aggfunc(x): + if all(x > 2): + return 1 + else: + return pd.NA + + df = DataFrame({"A": pd.array([1, 2, 3])}) + result = df.groupby([1, 1, 2]).agg(aggfunc) + expected = DataFrame({"A": pd.array([1, pd.NA], dtype="Int64")}, index=[1, 2]) + tm.assert_frame_equal(result, expected) + + +class TestLambdaMangling: + def test_basic(self): + df = DataFrame({"A": [0, 0, 1, 1], "B": [1, 2, 3, 4]}) + result = df.groupby("A").agg({"B": [lambda x: 0, lambda x: 1]}) + + expected = DataFrame( + {("B", ""): [0, 0], ("B", ""): [1, 1]}, + index=Index([0, 1], name="A"), + ) + tm.assert_frame_equal(result, expected) + + def test_mangle_series_groupby(self): + gr = Series([1, 2, 3, 4]).groupby([0, 0, 1, 1]) + result = gr.agg([lambda x: 0, lambda x: 1]) + exp_data = {"": [0, 0], "": [1, 1]} + expected = DataFrame(exp_data, index=np.array([0, 1])) + tm.assert_frame_equal(result, expected) + + @pytest.mark.xfail(reason="GH-26611. kwargs for multi-agg.") + def test_with_kwargs(self): + f1 = lambda x, y, b=1: x.sum() + y + b + f2 = lambda x, y, b=2: x.sum() + y * b + result = Series([1, 2]).groupby([0, 0]).agg([f1, f2], 0) + expected = DataFrame({"": [4], "": [6]}) + tm.assert_frame_equal(result, expected) + + result = Series([1, 2]).groupby([0, 0]).agg([f1, f2], 0, b=10) + expected = DataFrame({"": [13], "": [30]}) + tm.assert_frame_equal(result, expected) + + def test_agg_with_one_lambda(self): + # GH 25719, write tests for DataFrameGroupby.agg with only one lambda + df = DataFrame( + { + "kind": ["cat", "dog", "cat", "dog"], + "height": [9.1, 6.0, 9.5, 34.0], + "weight": [7.9, 7.5, 9.9, 198.0], + } + ) + + columns = ["height_sqr_min", "height_max", "weight_max"] + expected = DataFrame( + { + "height_sqr_min": [82.81, 36.00], + "height_max": [9.5, 34.0], + "weight_max": [9.9, 198.0], + }, + index=Index(["cat", "dog"], name="kind"), + columns=columns, + ) + + # check pd.NameAgg case + result1 = df.groupby(by="kind").agg( + height_sqr_min=pd.NamedAgg( + column="height", aggfunc=lambda x: np.min(x**2) + ), + height_max=pd.NamedAgg(column="height", aggfunc="max"), + weight_max=pd.NamedAgg(column="weight", aggfunc="max"), + ) + tm.assert_frame_equal(result1, expected) + + # check agg(key=(col, aggfunc)) case + result2 = df.groupby(by="kind").agg( + height_sqr_min=("height", lambda x: np.min(x**2)), + height_max=("height", "max"), + weight_max=("weight", "max"), + ) + tm.assert_frame_equal(result2, expected) + + def test_agg_multiple_lambda(self): + # GH25719, test for DataFrameGroupby.agg with multiple lambdas + # with mixed aggfunc + df = DataFrame( + { + "kind": ["cat", "dog", "cat", "dog"], + "height": [9.1, 6.0, 9.5, 34.0], + "weight": [7.9, 7.5, 9.9, 198.0], + } + ) + columns = [ + "height_sqr_min", + "height_max", + "weight_max", + "height_max_2", + "weight_min", + ] + expected = DataFrame( + { + "height_sqr_min": [82.81, 36.00], + "height_max": [9.5, 34.0], + "weight_max": [9.9, 198.0], + "height_max_2": [9.5, 34.0], + "weight_min": [7.9, 7.5], + }, + index=Index(["cat", "dog"], name="kind"), + columns=columns, + ) + + # check agg(key=(col, aggfunc)) case + result1 = df.groupby(by="kind").agg( + height_sqr_min=("height", lambda x: np.min(x**2)), + height_max=("height", "max"), + weight_max=("weight", "max"), + height_max_2=("height", lambda x: np.max(x)), + weight_min=("weight", lambda x: np.min(x)), + ) + tm.assert_frame_equal(result1, expected) + + # check pd.NamedAgg case + result2 = df.groupby(by="kind").agg( + height_sqr_min=pd.NamedAgg( + column="height", aggfunc=lambda x: np.min(x**2) + ), + height_max=pd.NamedAgg(column="height", aggfunc="max"), + weight_max=pd.NamedAgg(column="weight", aggfunc="max"), + height_max_2=pd.NamedAgg(column="height", aggfunc=lambda x: np.max(x)), + weight_min=pd.NamedAgg(column="weight", aggfunc=lambda x: np.min(x)), + ) + tm.assert_frame_equal(result2, expected) + + +def test_groupby_get_by_index(): + # GH 33439 + df = DataFrame({"A": ["S", "W", "W"], "B": [1.0, 1.0, 2.0]}) + res = df.groupby("A").agg({"B": lambda x: x.get(x.index[-1])}) + expected = DataFrame({"A": ["S", "W"], "B": [1.0, 2.0]}).set_index("A") + tm.assert_frame_equal(res, expected) + + +@pytest.mark.parametrize( + "grp_col_dict, exp_data", + [ + ({"nr": "min", "cat_ord": "min"}, {"nr": [1, 5], "cat_ord": ["a", "c"]}), + ({"cat_ord": "min"}, {"cat_ord": ["a", "c"]}), + ({"nr": "min"}, {"nr": [1, 5]}), + ], +) +def test_groupby_single_agg_cat_cols(grp_col_dict, exp_data): + # test single aggregations on ordered categorical cols GHGH27800 + + # create the result dataframe + input_df = DataFrame( + { + "nr": [1, 2, 3, 4, 5, 6, 7, 8], + "cat_ord": list("aabbccdd"), + "cat": list("aaaabbbb"), + } + ) + + input_df = input_df.astype({"cat": "category", "cat_ord": "category"}) + input_df["cat_ord"] = input_df["cat_ord"].cat.as_ordered() + result_df = input_df.groupby("cat", observed=False).agg(grp_col_dict) + + # create expected dataframe + cat_index = pd.CategoricalIndex( + ["a", "b"], categories=["a", "b"], ordered=False, name="cat", dtype="category" + ) + + expected_df = DataFrame(data=exp_data, index=cat_index) + + if "cat_ord" in expected_df: + # ordered categorical columns should be preserved + dtype = input_df["cat_ord"].dtype + expected_df["cat_ord"] = expected_df["cat_ord"].astype(dtype) + + tm.assert_frame_equal(result_df, expected_df) + + +@pytest.mark.parametrize( + "grp_col_dict, exp_data", + [ + ({"nr": ["min", "max"], "cat_ord": "min"}, [(1, 4, "a"), (5, 8, "c")]), + ({"nr": "min", "cat_ord": ["min", "max"]}, [(1, "a", "b"), (5, "c", "d")]), + ({"cat_ord": ["min", "max"]}, [("a", "b"), ("c", "d")]), + ], +) +def test_groupby_combined_aggs_cat_cols(grp_col_dict, exp_data): + # test combined aggregations on ordered categorical cols GH27800 + + # create the result dataframe + input_df = DataFrame( + { + "nr": [1, 2, 3, 4, 5, 6, 7, 8], + "cat_ord": list("aabbccdd"), + "cat": list("aaaabbbb"), + } + ) + + input_df = input_df.astype({"cat": "category", "cat_ord": "category"}) + input_df["cat_ord"] = input_df["cat_ord"].cat.as_ordered() + result_df = input_df.groupby("cat", observed=False).agg(grp_col_dict) + + # create expected dataframe + cat_index = pd.CategoricalIndex( + ["a", "b"], categories=["a", "b"], ordered=False, name="cat", dtype="category" + ) + + # unpack the grp_col_dict to create the multi-index tuple + # this tuple will be used to create the expected dataframe index + multi_index_list = [] + for k, v in grp_col_dict.items(): + if isinstance(v, list): + multi_index_list.extend([k, value] for value in v) + else: + multi_index_list.append([k, v]) + multi_index = MultiIndex.from_tuples(tuple(multi_index_list)) + + expected_df = DataFrame(data=exp_data, columns=multi_index, index=cat_index) + for col in expected_df.columns: + if isinstance(col, tuple) and "cat_ord" in col: + # ordered categorical should be preserved + expected_df[col] = expected_df[col].astype(input_df["cat_ord"].dtype) + + tm.assert_frame_equal(result_df, expected_df) + + +def test_nonagg_agg(): + # GH 35490 - Single/Multiple agg of non-agg function give same results + # TODO: agg should raise for functions that don't aggregate + df = DataFrame({"a": [1, 1, 2, 2], "b": [1, 2, 2, 1]}) + g = df.groupby("a") + + result = g.agg(["cumsum"]) + result.columns = result.columns.droplevel(-1) + expected = g.agg("cumsum") + + tm.assert_frame_equal(result, expected) + + +def test_aggregate_datetime_objects(): + # https://github.com/pandas-dev/pandas/issues/36003 + # ensure we don't raise an error but keep object dtype for out-of-bounds + # datetimes + df = DataFrame( + { + "A": ["X", "Y"], + "B": [ + datetime.datetime(2005, 1, 1, 10, 30, 23, 540000), + datetime.datetime(3005, 1, 1, 10, 30, 23, 540000), + ], + } + ) + result = df.groupby("A").B.max() + expected = df.set_index("A")["B"] + tm.assert_series_equal(result, expected) + + +def test_groupby_index_object_dtype(): + # GH 40014 + df = DataFrame({"c0": ["x", "x", "x"], "c1": ["x", "x", "y"], "p": [0, 1, 2]}) + df.index = df.index.astype("O") + grouped = df.groupby(["c0", "c1"]) + res = grouped.p.agg(lambda x: all(x > 0)) + # Check that providing a user-defined function in agg() + # produces the correct index shape when using an object-typed index. + expected_index = MultiIndex.from_tuples( + [("x", "x"), ("x", "y")], names=("c0", "c1") + ) + expected = Series([False, True], index=expected_index, name="p") + tm.assert_series_equal(res, expected) + + +def test_timeseries_groupby_agg(): + # GH#43290 + + def func(ser): + if ser.isna().all(): + return None + return np.sum(ser) + + df = DataFrame([1.0], index=[pd.Timestamp("2018-01-16 00:00:00+00:00")]) + res = df.groupby(lambda x: 1).agg(func) + + expected = DataFrame([[1.0]], index=[1]) + tm.assert_frame_equal(res, expected) + + +def test_groupby_agg_precision(any_real_numeric_dtype): + if any_real_numeric_dtype in tm.ALL_INT_NUMPY_DTYPES: + max_value = np.iinfo(any_real_numeric_dtype).max + if any_real_numeric_dtype in tm.FLOAT_NUMPY_DTYPES: + max_value = np.finfo(any_real_numeric_dtype).max + if any_real_numeric_dtype in tm.FLOAT_EA_DTYPES: + max_value = np.finfo(any_real_numeric_dtype.lower()).max + if any_real_numeric_dtype in tm.ALL_INT_EA_DTYPES: + max_value = np.iinfo(any_real_numeric_dtype.lower()).max + + df = DataFrame( + { + "key1": ["a"], + "key2": ["b"], + "key3": pd.array([max_value], dtype=any_real_numeric_dtype), + } + ) + arrays = [["a"], ["b"]] + index = MultiIndex.from_arrays(arrays, names=("key1", "key2")) + + expected = DataFrame( + {"key3": pd.array([max_value], dtype=any_real_numeric_dtype)}, index=index + ) + result = df.groupby(["key1", "key2"]).agg(lambda x: x) + tm.assert_frame_equal(result, expected) + + +def test_groupby_aggregate_directory(reduction_func): + # GH#32793 + if reduction_func in ["corrwith", "nth"]: + return None + + obj = DataFrame([[0, 1], [0, np.nan]]) + + result_reduced_series = obj.groupby(0).agg(reduction_func) + result_reduced_frame = obj.groupby(0).agg({1: reduction_func}) + + if reduction_func in ["size", "ngroup"]: + # names are different: None / 1 + tm.assert_series_equal( + result_reduced_series, result_reduced_frame[1], check_names=False + ) + else: + tm.assert_frame_equal(result_reduced_series, result_reduced_frame) + tm.assert_series_equal( + result_reduced_series.dtypes, result_reduced_frame.dtypes + ) + + +def test_group_mean_timedelta_nat(): + # GH43132 + data = Series(["1 day", "3 days", "NaT"], dtype="timedelta64[ns]") + expected = Series(["2 days"], dtype="timedelta64[ns]", index=np.array([0])) + + result = data.groupby([0, 0, 0]).mean() + + tm.assert_series_equal(result, expected) + + +@pytest.mark.parametrize( + "input_data, expected_output", + [ + ( # no timezone + ["2021-01-01T00:00", "NaT", "2021-01-01T02:00"], + ["2021-01-01T01:00"], + ), + ( # timezone + ["2021-01-01T00:00-0100", "NaT", "2021-01-01T02:00-0100"], + ["2021-01-01T01:00-0100"], + ), + ], +) +def test_group_mean_datetime64_nat(input_data, expected_output): + # GH43132 + data = to_datetime(Series(input_data)) + expected = to_datetime(Series(expected_output, index=np.array([0]))) + + result = data.groupby([0, 0, 0]).mean() + tm.assert_series_equal(result, expected) + + +@pytest.mark.parametrize( + "func, output", [("mean", [8 + 18j, 10 + 22j]), ("sum", [40 + 90j, 50 + 110j])] +) +def test_groupby_complex(func, output): + # GH#43701 + data = Series(np.arange(20).reshape(10, 2).dot([1, 2j])) + result = data.groupby(data.index % 2).agg(func) + expected = Series(output) + tm.assert_series_equal(result, expected) + + +@pytest.mark.parametrize("func", ["min", "max", "var"]) +def test_groupby_complex_raises(func): + # GH#43701 + data = Series(np.arange(20).reshape(10, 2).dot([1, 2j])) + msg = "No matching signature found" + with pytest.raises(TypeError, match=msg): + data.groupby(data.index % 2).agg(func) + + +@pytest.mark.parametrize( + "func", [["min"], ["mean", "max"], {"b": "sum"}, {"b": "prod", "c": "median"}] +) +def test_multi_axis_1_raises(func): + # GH#46995 + df = DataFrame({"a": [1, 1, 2], "b": [3, 4, 5], "c": [6, 7, 8]}) + msg = "DataFrame.groupby with axis=1 is deprecated" + with tm.assert_produces_warning(FutureWarning, match=msg): + gb = df.groupby("a", axis=1) + with pytest.raises(NotImplementedError, match="axis other than 0 is not supported"): + gb.agg(func) + + +@pytest.mark.parametrize( + "test, constant", + [ + ([[20, "A"], [20, "B"], [10, "C"]], {0: [10, 20], 1: ["C", ["A", "B"]]}), + ([[20, "A"], [20, "B"], [30, "C"]], {0: [20, 30], 1: [["A", "B"], "C"]}), + ([["a", 1], ["a", 1], ["b", 2], ["b", 3]], {0: ["a", "b"], 1: [1, [2, 3]]}), + pytest.param( + [["a", 1], ["a", 2], ["b", 3], ["b", 3]], + {0: ["a", "b"], 1: [[1, 2], 3]}, + marks=pytest.mark.xfail, + ), + ], +) +def test_agg_of_mode_list(test, constant): + # GH#25581 + df1 = DataFrame(test) + result = df1.groupby(0).agg(Series.mode) + # Mode usually only returns 1 value, but can return a list in the case of a tie. + + expected = DataFrame(constant) + expected = expected.set_index(0) + + tm.assert_frame_equal(result, expected) + + +def test_dataframe_groupy_agg_list_like_func_with_args(): + # GH#50624 + df = DataFrame({"x": [1, 2, 3], "y": ["a", "b", "c"]}) + gb = df.groupby("y") + + def foo1(x, a=1, c=0): + return x.sum() + a + c + + def foo2(x, b=2, c=0): + return x.sum() + b + c + + msg = r"foo1\(\) got an unexpected keyword argument 'b'" + with pytest.raises(TypeError, match=msg): + gb.agg([foo1, foo2], 3, b=3, c=4) + + result = gb.agg([foo1, foo2], 3, c=4) + expected = DataFrame( + [[8, 8], [9, 9], [10, 10]], + index=Index(["a", "b", "c"], name="y"), + columns=MultiIndex.from_tuples([("x", "foo1"), ("x", "foo2")]), + ) + tm.assert_frame_equal(result, expected) + + +def test_series_groupy_agg_list_like_func_with_args(): + # GH#50624 + s = Series([1, 2, 3]) + sgb = s.groupby(s) + + def foo1(x, a=1, c=0): + return x.sum() + a + c + + def foo2(x, b=2, c=0): + return x.sum() + b + c + + msg = r"foo1\(\) got an unexpected keyword argument 'b'" + with pytest.raises(TypeError, match=msg): + sgb.agg([foo1, foo2], 3, b=3, c=4) + + result = sgb.agg([foo1, foo2], 3, c=4) + expected = DataFrame( + [[8, 8], [9, 9], [10, 10]], index=Index([1, 2, 3]), columns=["foo1", "foo2"] + ) + tm.assert_frame_equal(result, expected) + + +def test_agg_groupings_selection(): + # GH#51186 - a selected grouping should be in the output of agg + df = DataFrame({"a": [1, 1, 2], "b": [3, 3, 4], "c": [5, 6, 7]}) + gb = df.groupby(["a", "b"]) + selected_gb = gb[["b", "c"]] + result = selected_gb.agg(lambda x: x.sum()) + index = MultiIndex( + levels=[[1, 2], [3, 4]], codes=[[0, 1], [0, 1]], names=["a", "b"] + ) + expected = DataFrame({"b": [6, 4], "c": [11, 7]}, index=index) + tm.assert_frame_equal(result, expected) + + +def test_agg_multiple_with_as_index_false_subset_to_a_single_column(): + # GH#50724 + df = DataFrame({"a": [1, 1, 2], "b": [3, 4, 5]}) + gb = df.groupby("a", as_index=False)["b"] + result = gb.agg(["sum", "mean"]) + expected = DataFrame({"a": [1, 2], "sum": [7, 5], "mean": [3.5, 5.0]}) + tm.assert_frame_equal(result, expected) + + +def test_agg_with_as_index_false_with_list(): + # GH#52849 + df = DataFrame({"a1": [0, 0, 1], "a2": [2, 3, 3], "b": [4, 5, 6]}) + gb = df.groupby(by=["a1", "a2"], as_index=False) + result = gb.agg(["sum"]) + + expected = DataFrame( + data=[[0, 2, 4], [0, 3, 5], [1, 3, 6]], + columns=MultiIndex.from_tuples([("a1", ""), ("a2", ""), ("b", "sum")]), + ) + tm.assert_frame_equal(result, expected) + + +def test_groupby_agg_extension_timedelta_cumsum_with_named_aggregation(): + # GH#41720 + expected = DataFrame( + { + "td": { + 0: pd.Timedelta("0 days 01:00:00"), + 1: pd.Timedelta("0 days 01:15:00"), + 2: pd.Timedelta("0 days 01:15:00"), + } + } + ) + df = DataFrame( + { + "td": Series( + ["0 days 01:00:00", "0 days 00:15:00", "0 days 01:15:00"], + dtype="timedelta64[ns]", + ), + "grps": ["a", "a", "b"], + } + ) + gb = df.groupby("grps") + result = gb.agg(td=("td", "cumsum")) + tm.assert_frame_equal(result, expected) + + +def test_groupby_aggregation_empty_group(): + # https://github.com/pandas-dev/pandas/issues/18869 + def func(x): + if len(x) == 0: + raise ValueError("length must not be 0") + return len(x) + + df = DataFrame( + {"A": pd.Categorical(["a", "a"], categories=["a", "b", "c"]), "B": [1, 1]} + ) + msg = "length must not be 0" + with pytest.raises(ValueError, match=msg): + df.groupby("A", observed=False).agg(func) diff --git a/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/aggregate/test_cython.py b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/aggregate/test_cython.py new file mode 100644 index 0000000000000000000000000000000000000000..5c99882cef6d2393278df5879ea4af75aa14f60c --- /dev/null +++ b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/aggregate/test_cython.py @@ -0,0 +1,435 @@ +""" +test cython .agg behavior +""" + +import numpy as np +import pytest + +from pandas.core.dtypes.common import ( + is_float_dtype, + is_integer_dtype, +) + +import pandas as pd +from pandas import ( + DataFrame, + Index, + NaT, + Series, + Timedelta, + Timestamp, + bdate_range, +) +import pandas._testing as tm +import pandas.core.common as com + + +@pytest.mark.parametrize( + "op_name", + [ + "count", + "sum", + "std", + "var", + "sem", + "mean", + pytest.param( + "median", + # ignore mean of empty slice + # and all-NaN + marks=[pytest.mark.filterwarnings("ignore::RuntimeWarning")], + ), + "prod", + "min", + "max", + ], +) +def test_cythonized_aggers(op_name): + data = { + "A": [0, 0, 0, 0, 1, 1, 1, 1, 1, 1.0, np.nan, np.nan], + "B": ["A", "B"] * 6, + "C": np.random.default_rng(2).standard_normal(12), + } + df = DataFrame(data) + df.loc[2:10:2, "C"] = np.nan + + op = lambda x: getattr(x, op_name)() + + # single column + grouped = df.drop(["B"], axis=1).groupby("A") + exp = {cat: op(group["C"]) for cat, group in grouped} + exp = DataFrame({"C": exp}) + exp.index.name = "A" + result = op(grouped) + tm.assert_frame_equal(result, exp) + + # multiple columns + grouped = df.groupby(["A", "B"]) + expd = {} + for (cat1, cat2), group in grouped: + expd.setdefault(cat1, {})[cat2] = op(group["C"]) + exp = DataFrame(expd).T.stack(future_stack=True) + exp.index.names = ["A", "B"] + exp.name = "C" + + result = op(grouped)["C"] + if op_name in ["sum", "prod"]: + tm.assert_series_equal(result, exp) + + +def test_cython_agg_boolean(): + frame = DataFrame( + { + "a": np.random.default_rng(2).integers(0, 5, 50), + "b": np.random.default_rng(2).integers(0, 2, 50).astype("bool"), + } + ) + result = frame.groupby("a")["b"].mean() + msg = "using SeriesGroupBy.mean" + with tm.assert_produces_warning(FutureWarning, match=msg): + # GH#53425 + expected = frame.groupby("a")["b"].agg(np.mean) + + tm.assert_series_equal(result, expected) + + +def test_cython_agg_nothing_to_agg(): + frame = DataFrame( + {"a": np.random.default_rng(2).integers(0, 5, 50), "b": ["foo", "bar"] * 25} + ) + + msg = "Cannot use numeric_only=True with SeriesGroupBy.mean and non-numeric dtypes" + with pytest.raises(TypeError, match=msg): + frame.groupby("a")["b"].mean(numeric_only=True) + + frame = DataFrame( + {"a": np.random.default_rng(2).integers(0, 5, 50), "b": ["foo", "bar"] * 25} + ) + + result = frame[["b"]].groupby(frame["a"]).mean(numeric_only=True) + expected = DataFrame( + [], index=frame["a"].sort_values().drop_duplicates(), columns=[] + ) + tm.assert_frame_equal(result, expected) + + +def test_cython_agg_nothing_to_agg_with_dates(): + frame = DataFrame( + { + "a": np.random.default_rng(2).integers(0, 5, 50), + "b": ["foo", "bar"] * 25, + "dates": pd.date_range("now", periods=50, freq="min"), + } + ) + msg = "Cannot use numeric_only=True with SeriesGroupBy.mean and non-numeric dtypes" + with pytest.raises(TypeError, match=msg): + frame.groupby("b").dates.mean(numeric_only=True) + + +def test_cython_agg_frame_columns(): + # #2113 + df = DataFrame({"x": [1, 2, 3], "y": [3, 4, 5]}) + + msg = "DataFrame.groupby with axis=1 is deprecated" + with tm.assert_produces_warning(FutureWarning, match=msg): + df.groupby(level=0, axis="columns").mean() + with tm.assert_produces_warning(FutureWarning, match=msg): + df.groupby(level=0, axis="columns").mean() + with tm.assert_produces_warning(FutureWarning, match=msg): + df.groupby(level=0, axis="columns").mean() + with tm.assert_produces_warning(FutureWarning, match=msg): + df.groupby(level=0, axis="columns").mean() + + +def test_cython_agg_return_dict(): + # GH 16741 + df = DataFrame( + { + "A": ["foo", "bar", "foo", "bar", "foo", "bar", "foo", "foo"], + "B": ["one", "one", "two", "three", "two", "two", "one", "three"], + "C": np.random.default_rng(2).standard_normal(8), + "D": np.random.default_rng(2).standard_normal(8), + } + ) + + ts = df.groupby("A")["B"].agg(lambda x: x.value_counts().to_dict()) + expected = Series( + [{"two": 1, "one": 1, "three": 1}, {"two": 2, "one": 2, "three": 1}], + index=Index(["bar", "foo"], name="A"), + name="B", + ) + tm.assert_series_equal(ts, expected) + + +def test_cython_fail_agg(): + dr = bdate_range("1/1/2000", periods=50) + ts = Series(["A", "B", "C", "D", "E"] * 10, index=dr) + + grouped = ts.groupby(lambda x: x.month) + summed = grouped.sum() + msg = "using SeriesGroupBy.sum" + with tm.assert_produces_warning(FutureWarning, match=msg): + # GH#53425 + expected = grouped.agg(np.sum) + tm.assert_series_equal(summed, expected) + + +@pytest.mark.parametrize( + "op, targop", + [ + ("mean", np.mean), + ("median", np.median), + ("var", np.var), + ("sum", np.sum), + ("prod", np.prod), + ("min", np.min), + ("max", np.max), + ("first", lambda x: x.iloc[0]), + ("last", lambda x: x.iloc[-1]), + ], +) +def test__cython_agg_general(op, targop): + df = DataFrame(np.random.default_rng(2).standard_normal(1000)) + labels = np.random.default_rng(2).integers(0, 50, size=1000).astype(float) + + result = df.groupby(labels)._cython_agg_general(op, alt=None, numeric_only=True) + warn = FutureWarning if targop in com._cython_table else None + msg = f"using DataFrameGroupBy.{op}" + with tm.assert_produces_warning(warn, match=msg): + # GH#53425 + expected = df.groupby(labels).agg(targop) + tm.assert_frame_equal(result, expected) + + +@pytest.mark.parametrize( + "op, targop", + [ + ("mean", np.mean), + ("median", lambda x: np.median(x) if len(x) > 0 else np.nan), + ("var", lambda x: np.var(x, ddof=1)), + ("min", np.min), + ("max", np.max), + ], +) +def test_cython_agg_empty_buckets(op, targop, observed): + df = DataFrame([11, 12, 13]) + grps = range(0, 55, 5) + + # calling _cython_agg_general directly, instead of via the user API + # which sets different values for min_count, so do that here. + g = df.groupby(pd.cut(df[0], grps), observed=observed) + result = g._cython_agg_general(op, alt=None, numeric_only=True) + + g = df.groupby(pd.cut(df[0], grps), observed=observed) + expected = g.agg(lambda x: targop(x)) + tm.assert_frame_equal(result, expected) + + +def test_cython_agg_empty_buckets_nanops(observed): + # GH-18869 can't call nanops on empty groups, so hardcode expected + # for these + df = DataFrame([11, 12, 13], columns=["a"]) + grps = np.arange(0, 25, 5, dtype=int) + # add / sum + result = df.groupby(pd.cut(df["a"], grps), observed=observed)._cython_agg_general( + "sum", alt=None, numeric_only=True + ) + intervals = pd.interval_range(0, 20, freq=5) + expected = DataFrame( + {"a": [0, 0, 36, 0]}, + index=pd.CategoricalIndex(intervals, name="a", ordered=True), + ) + if observed: + expected = expected[expected.a != 0] + + tm.assert_frame_equal(result, expected) + + # prod + result = df.groupby(pd.cut(df["a"], grps), observed=observed)._cython_agg_general( + "prod", alt=None, numeric_only=True + ) + expected = DataFrame( + {"a": [1, 1, 1716, 1]}, + index=pd.CategoricalIndex(intervals, name="a", ordered=True), + ) + if observed: + expected = expected[expected.a != 1] + + tm.assert_frame_equal(result, expected) + + +@pytest.mark.parametrize("op", ["first", "last", "max", "min"]) +@pytest.mark.parametrize( + "data", [Timestamp("2016-10-14 21:00:44.557"), Timedelta("17088 days 21:00:44.557")] +) +def test_cython_with_timestamp_and_nat(op, data): + # https://github.com/pandas-dev/pandas/issues/19526 + df = DataFrame({"a": [0, 1], "b": [data, NaT]}) + index = Index([0, 1], name="a") + + # We will group by a and test the cython aggregations + expected = DataFrame({"b": [data, NaT]}, index=index) + + result = df.groupby("a").aggregate(op) + tm.assert_frame_equal(expected, result) + + +@pytest.mark.parametrize( + "agg", + [ + "min", + "max", + "count", + "sum", + "prod", + "var", + "mean", + "median", + "ohlc", + "cumprod", + "cumsum", + "shift", + "any", + "all", + "quantile", + "first", + "last", + "rank", + "cummin", + "cummax", + ], +) +def test_read_only_buffer_source_agg(agg): + # https://github.com/pandas-dev/pandas/issues/36014 + df = DataFrame( + { + "sepal_length": [5.1, 4.9, 4.7, 4.6, 5.0], + "species": ["setosa", "setosa", "setosa", "setosa", "setosa"], + } + ) + df._mgr.arrays[0].flags.writeable = False + + result = df.groupby(["species"]).agg({"sepal_length": agg}) + expected = df.copy().groupby(["species"]).agg({"sepal_length": agg}) + + tm.assert_equal(result, expected) + + +@pytest.mark.parametrize( + "op_name", + [ + "count", + "sum", + "std", + "var", + "sem", + "mean", + "median", + "prod", + "min", + "max", + ], +) +def test_cython_agg_nullable_int(op_name): + # ensure that the cython-based aggregations don't fail for nullable dtype + # (eg https://github.com/pandas-dev/pandas/issues/37415) + df = DataFrame( + { + "A": ["A", "B"] * 5, + "B": pd.array([1, 2, 3, 4, 5, 6, 7, 8, 9, pd.NA], dtype="Int64"), + } + ) + result = getattr(df.groupby("A")["B"], op_name)() + df2 = df.assign(B=df["B"].astype("float64")) + expected = getattr(df2.groupby("A")["B"], op_name)() + if op_name in ("mean", "median"): + convert_integer = False + else: + convert_integer = True + expected = expected.convert_dtypes(convert_integer=convert_integer) + tm.assert_series_equal(result, expected) + + +@pytest.mark.parametrize("dtype", ["Int64", "Float64", "boolean"]) +def test_count_masked_returns_masked_dtype(dtype): + df = DataFrame( + { + "A": [1, 1], + "B": pd.array([1, pd.NA], dtype=dtype), + "C": pd.array([1, 1], dtype=dtype), + } + ) + result = df.groupby("A").count() + expected = DataFrame( + [[1, 2]], index=Index([1], name="A"), columns=["B", "C"], dtype="Int64" + ) + tm.assert_frame_equal(result, expected) + + +@pytest.mark.parametrize("with_na", [True, False]) +@pytest.mark.parametrize( + "op_name, action", + [ + # ("count", "always_int"), + ("sum", "large_int"), + # ("std", "always_float"), + ("var", "always_float"), + # ("sem", "always_float"), + ("mean", "always_float"), + ("median", "always_float"), + ("prod", "large_int"), + ("min", "preserve"), + ("max", "preserve"), + ("first", "preserve"), + ("last", "preserve"), + ], +) +@pytest.mark.parametrize( + "data", + [ + pd.array([1, 2, 3, 4], dtype="Int64"), + pd.array([1, 2, 3, 4], dtype="Int8"), + pd.array([0.1, 0.2, 0.3, 0.4], dtype="Float32"), + pd.array([0.1, 0.2, 0.3, 0.4], dtype="Float64"), + pd.array([True, True, False, False], dtype="boolean"), + ], +) +def test_cython_agg_EA_known_dtypes(data, op_name, action, with_na): + if with_na: + data[3] = pd.NA + + df = DataFrame({"key": ["a", "a", "b", "b"], "col": data}) + grouped = df.groupby("key") + + if action == "always_int": + # always Int64 + expected_dtype = pd.Int64Dtype() + elif action == "large_int": + # for any int/bool use Int64, for float preserve dtype + if is_float_dtype(data.dtype): + expected_dtype = data.dtype + elif is_integer_dtype(data.dtype): + # match the numpy dtype we'd get with the non-nullable analogue + expected_dtype = data.dtype + else: + expected_dtype = pd.Int64Dtype() + elif action == "always_float": + # for any int/bool use Float64, for float preserve dtype + if is_float_dtype(data.dtype): + expected_dtype = data.dtype + else: + expected_dtype = pd.Float64Dtype() + elif action == "preserve": + expected_dtype = data.dtype + + result = getattr(grouped, op_name)() + assert result["col"].dtype == expected_dtype + + result = grouped.aggregate(op_name) + assert result["col"].dtype == expected_dtype + + result = getattr(grouped["col"], op_name)() + assert result.dtype == expected_dtype + + result = grouped["col"].aggregate(op_name) + assert result.dtype == expected_dtype diff --git a/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/aggregate/test_numba.py b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/aggregate/test_numba.py new file mode 100644 index 0000000000000000000000000000000000000000..ee694129f71183294dc780783d3b9ccdeae73bf4 --- /dev/null +++ b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/aggregate/test_numba.py @@ -0,0 +1,392 @@ +import numpy as np +import pytest + +from pandas.errors import NumbaUtilError + +from pandas import ( + DataFrame, + Index, + NamedAgg, + Series, + option_context, +) +import pandas._testing as tm + +pytestmark = pytest.mark.single_cpu + + +def test_correct_function_signature(): + pytest.importorskip("numba") + + def incorrect_function(x): + return sum(x) * 2.7 + + data = DataFrame( + {"key": ["a", "a", "b", "b", "a"], "data": [1.0, 2.0, 3.0, 4.0, 5.0]}, + columns=["key", "data"], + ) + with pytest.raises(NumbaUtilError, match="The first 2"): + data.groupby("key").agg(incorrect_function, engine="numba") + + with pytest.raises(NumbaUtilError, match="The first 2"): + data.groupby("key")["data"].agg(incorrect_function, engine="numba") + + +def test_check_nopython_kwargs(): + pytest.importorskip("numba") + + def incorrect_function(values, index): + return sum(values) * 2.7 + + data = DataFrame( + {"key": ["a", "a", "b", "b", "a"], "data": [1.0, 2.0, 3.0, 4.0, 5.0]}, + columns=["key", "data"], + ) + with pytest.raises(NumbaUtilError, match="numba does not support"): + data.groupby("key").agg(incorrect_function, engine="numba", a=1) + + with pytest.raises(NumbaUtilError, match="numba does not support"): + data.groupby("key")["data"].agg(incorrect_function, engine="numba", a=1) + + +@pytest.mark.filterwarnings("ignore") +# Filter warnings when parallel=True and the function can't be parallelized by Numba +@pytest.mark.parametrize("jit", [True, False]) +@pytest.mark.parametrize("pandas_obj", ["Series", "DataFrame"]) +@pytest.mark.parametrize("as_index", [True, False]) +def test_numba_vs_cython(jit, pandas_obj, nogil, parallel, nopython, as_index): + pytest.importorskip("numba") + + def func_numba(values, index): + return np.mean(values) * 2.7 + + if jit: + # Test accepted jitted functions + import numba + + func_numba = numba.jit(func_numba) + + data = DataFrame( + {0: ["a", "a", "b", "b", "a"], 1: [1.0, 2.0, 3.0, 4.0, 5.0]}, columns=[0, 1] + ) + engine_kwargs = {"nogil": nogil, "parallel": parallel, "nopython": nopython} + grouped = data.groupby(0, as_index=as_index) + if pandas_obj == "Series": + grouped = grouped[1] + + result = grouped.agg(func_numba, engine="numba", engine_kwargs=engine_kwargs) + expected = grouped.agg(lambda x: np.mean(x) * 2.7, engine="cython") + + tm.assert_equal(result, expected) + + +@pytest.mark.filterwarnings("ignore") +# Filter warnings when parallel=True and the function can't be parallelized by Numba +@pytest.mark.parametrize("jit", [True, False]) +@pytest.mark.parametrize("pandas_obj", ["Series", "DataFrame"]) +def test_cache(jit, pandas_obj, nogil, parallel, nopython): + # Test that the functions are cached correctly if we switch functions + pytest.importorskip("numba") + + def func_1(values, index): + return np.mean(values) - 3.4 + + def func_2(values, index): + return np.mean(values) * 2.7 + + if jit: + import numba + + func_1 = numba.jit(func_1) + func_2 = numba.jit(func_2) + + data = DataFrame( + {0: ["a", "a", "b", "b", "a"], 1: [1.0, 2.0, 3.0, 4.0, 5.0]}, columns=[0, 1] + ) + engine_kwargs = {"nogil": nogil, "parallel": parallel, "nopython": nopython} + grouped = data.groupby(0) + if pandas_obj == "Series": + grouped = grouped[1] + + result = grouped.agg(func_1, engine="numba", engine_kwargs=engine_kwargs) + expected = grouped.agg(lambda x: np.mean(x) - 3.4, engine="cython") + tm.assert_equal(result, expected) + + # Add func_2 to the cache + result = grouped.agg(func_2, engine="numba", engine_kwargs=engine_kwargs) + expected = grouped.agg(lambda x: np.mean(x) * 2.7, engine="cython") + tm.assert_equal(result, expected) + + # Retest func_1 which should use the cache + result = grouped.agg(func_1, engine="numba", engine_kwargs=engine_kwargs) + expected = grouped.agg(lambda x: np.mean(x) - 3.4, engine="cython") + tm.assert_equal(result, expected) + + +def test_use_global_config(): + pytest.importorskip("numba") + + def func_1(values, index): + return np.mean(values) - 3.4 + + data = DataFrame( + {0: ["a", "a", "b", "b", "a"], 1: [1.0, 2.0, 3.0, 4.0, 5.0]}, columns=[0, 1] + ) + grouped = data.groupby(0) + expected = grouped.agg(func_1, engine="numba") + with option_context("compute.use_numba", True): + result = grouped.agg(func_1, engine=None) + tm.assert_frame_equal(expected, result) + + +@pytest.mark.parametrize( + "agg_kwargs", + [ + {"func": ["min", "max"]}, + {"func": "min"}, + {"func": {1: ["min", "max"], 2: "sum"}}, + {"bmin": NamedAgg(column=1, aggfunc="min")}, + ], +) +def test_multifunc_numba_vs_cython_frame(agg_kwargs): + pytest.importorskip("numba") + data = DataFrame( + { + 0: ["a", "a", "b", "b", "a"], + 1: [1.0, 2.0, 3.0, 4.0, 5.0], + 2: [1, 2, 3, 4, 5], + }, + columns=[0, 1, 2], + ) + grouped = data.groupby(0) + result = grouped.agg(**agg_kwargs, engine="numba") + expected = grouped.agg(**agg_kwargs, engine="cython") + tm.assert_frame_equal(result, expected) + + +@pytest.mark.parametrize( + "agg_kwargs,expected_func", + [ + ({"func": lambda values, index: values.sum()}, "sum"), + # FIXME + pytest.param( + { + "func": [ + lambda values, index: values.sum(), + lambda values, index: values.min(), + ] + }, + ["sum", "min"], + marks=pytest.mark.xfail( + reason="This doesn't work yet! Fails in nopython pipeline!" + ), + ), + ], +) +def test_multifunc_numba_udf_frame(agg_kwargs, expected_func): + pytest.importorskip("numba") + data = DataFrame( + { + 0: ["a", "a", "b", "b", "a"], + 1: [1.0, 2.0, 3.0, 4.0, 5.0], + 2: [1, 2, 3, 4, 5], + }, + columns=[0, 1, 2], + ) + grouped = data.groupby(0) + result = grouped.agg(**agg_kwargs, engine="numba") + expected = grouped.agg(expected_func, engine="cython") + # check_dtype can be removed if GH 44952 is addressed + # Currently, UDFs still always return float64 while reductions can preserve dtype + tm.assert_frame_equal(result, expected, check_dtype=False) + + +@pytest.mark.parametrize( + "agg_kwargs", + [{"func": ["min", "max"]}, {"func": "min"}, {"min_val": "min", "max_val": "max"}], +) +def test_multifunc_numba_vs_cython_series(agg_kwargs): + pytest.importorskip("numba") + labels = ["a", "a", "b", "b", "a"] + data = Series([1.0, 2.0, 3.0, 4.0, 5.0]) + grouped = data.groupby(labels) + agg_kwargs["engine"] = "numba" + result = grouped.agg(**agg_kwargs) + agg_kwargs["engine"] = "cython" + expected = grouped.agg(**agg_kwargs) + if isinstance(expected, DataFrame): + tm.assert_frame_equal(result, expected) + else: + tm.assert_series_equal(result, expected) + + +@pytest.mark.single_cpu +@pytest.mark.parametrize( + "data,agg_kwargs", + [ + (Series([1.0, 2.0, 3.0, 4.0, 5.0]), {"func": ["min", "max"]}), + (Series([1.0, 2.0, 3.0, 4.0, 5.0]), {"func": "min"}), + ( + DataFrame( + {1: [1.0, 2.0, 3.0, 4.0, 5.0], 2: [1, 2, 3, 4, 5]}, columns=[1, 2] + ), + {"func": ["min", "max"]}, + ), + ( + DataFrame( + {1: [1.0, 2.0, 3.0, 4.0, 5.0], 2: [1, 2, 3, 4, 5]}, columns=[1, 2] + ), + {"func": "min"}, + ), + ( + DataFrame( + {1: [1.0, 2.0, 3.0, 4.0, 5.0], 2: [1, 2, 3, 4, 5]}, columns=[1, 2] + ), + {"func": {1: ["min", "max"], 2: "sum"}}, + ), + ( + DataFrame( + {1: [1.0, 2.0, 3.0, 4.0, 5.0], 2: [1, 2, 3, 4, 5]}, columns=[1, 2] + ), + {"min_col": NamedAgg(column=1, aggfunc="min")}, + ), + ], +) +def test_multifunc_numba_kwarg_propagation(data, agg_kwargs): + pytest.importorskip("numba") + labels = ["a", "a", "b", "b", "a"] + grouped = data.groupby(labels) + result = grouped.agg(**agg_kwargs, engine="numba", engine_kwargs={"parallel": True}) + expected = grouped.agg(**agg_kwargs, engine="numba") + if isinstance(expected, DataFrame): + tm.assert_frame_equal(result, expected) + else: + tm.assert_series_equal(result, expected) + + +def test_args_not_cached(): + # GH 41647 + pytest.importorskip("numba") + + def sum_last(values, index, n): + return values[-n:].sum() + + df = DataFrame({"id": [0, 0, 1, 1], "x": [1, 1, 1, 1]}) + grouped_x = df.groupby("id")["x"] + result = grouped_x.agg(sum_last, 1, engine="numba") + expected = Series([1.0] * 2, name="x", index=Index([0, 1], name="id")) + tm.assert_series_equal(result, expected) + + result = grouped_x.agg(sum_last, 2, engine="numba") + expected = Series([2.0] * 2, name="x", index=Index([0, 1], name="id")) + tm.assert_series_equal(result, expected) + + +def test_index_data_correctly_passed(): + # GH 43133 + pytest.importorskip("numba") + + def f(values, index): + return np.mean(index) + + df = DataFrame({"group": ["A", "A", "B"], "v": [4, 5, 6]}, index=[-1, -2, -3]) + result = df.groupby("group").aggregate(f, engine="numba") + expected = DataFrame( + [-1.5, -3.0], columns=["v"], index=Index(["A", "B"], name="group") + ) + tm.assert_frame_equal(result, expected) + + +def test_engine_kwargs_not_cached(): + # If the user passes a different set of engine_kwargs don't return the same + # jitted function + pytest.importorskip("numba") + nogil = True + parallel = False + nopython = True + + def func_kwargs(values, index): + return nogil + parallel + nopython + + engine_kwargs = {"nopython": nopython, "nogil": nogil, "parallel": parallel} + df = DataFrame({"value": [0, 0, 0]}) + result = df.groupby(level=0).aggregate( + func_kwargs, engine="numba", engine_kwargs=engine_kwargs + ) + expected = DataFrame({"value": [2.0, 2.0, 2.0]}) + tm.assert_frame_equal(result, expected) + + nogil = False + engine_kwargs = {"nopython": nopython, "nogil": nogil, "parallel": parallel} + result = df.groupby(level=0).aggregate( + func_kwargs, engine="numba", engine_kwargs=engine_kwargs + ) + expected = DataFrame({"value": [1.0, 1.0, 1.0]}) + tm.assert_frame_equal(result, expected) + + +@pytest.mark.filterwarnings("ignore") +def test_multiindex_one_key(nogil, parallel, nopython): + pytest.importorskip("numba") + + def numba_func(values, index): + return 1 + + df = DataFrame([{"A": 1, "B": 2, "C": 3}]).set_index(["A", "B"]) + engine_kwargs = {"nopython": nopython, "nogil": nogil, "parallel": parallel} + result = df.groupby("A").agg( + numba_func, engine="numba", engine_kwargs=engine_kwargs + ) + expected = DataFrame([1.0], index=Index([1], name="A"), columns=["C"]) + tm.assert_frame_equal(result, expected) + + +def test_multiindex_multi_key_not_supported(nogil, parallel, nopython): + pytest.importorskip("numba") + + def numba_func(values, index): + return 1 + + df = DataFrame([{"A": 1, "B": 2, "C": 3}]).set_index(["A", "B"]) + engine_kwargs = {"nopython": nopython, "nogil": nogil, "parallel": parallel} + with pytest.raises(NotImplementedError, match="more than 1 grouping labels"): + df.groupby(["A", "B"]).agg( + numba_func, engine="numba", engine_kwargs=engine_kwargs + ) + + +def test_multilabel_numba_vs_cython(numba_supported_reductions): + pytest.importorskip("numba") + reduction, kwargs = numba_supported_reductions + df = DataFrame( + { + "A": ["foo", "bar", "foo", "bar", "foo", "bar", "foo", "foo"], + "B": ["one", "one", "two", "three", "two", "two", "one", "three"], + "C": np.random.default_rng(2).standard_normal(8), + "D": np.random.default_rng(2).standard_normal(8), + } + ) + gb = df.groupby(["A", "B"]) + res_agg = gb.agg(reduction, engine="numba", **kwargs) + expected_agg = gb.agg(reduction, engine="cython", **kwargs) + tm.assert_frame_equal(res_agg, expected_agg) + # Test that calling the aggregation directly also works + direct_res = getattr(gb, reduction)(engine="numba", **kwargs) + direct_expected = getattr(gb, reduction)(engine="cython", **kwargs) + tm.assert_frame_equal(direct_res, direct_expected) + + +def test_multilabel_udf_numba_vs_cython(): + pytest.importorskip("numba") + df = DataFrame( + { + "A": ["foo", "bar", "foo", "bar", "foo", "bar", "foo", "foo"], + "B": ["one", "one", "two", "three", "two", "two", "one", "three"], + "C": np.random.default_rng(2).standard_normal(8), + "D": np.random.default_rng(2).standard_normal(8), + } + ) + gb = df.groupby(["A", "B"]) + result = gb.agg(lambda values, index: values.min(), engine="numba") + expected = gb.agg(lambda x: x.min(), engine="cython") + tm.assert_frame_equal(result, expected) diff --git a/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/aggregate/test_other.py b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/aggregate/test_other.py new file mode 100644 index 0000000000000000000000000000000000000000..00136e572288e9858412fd9d84e3ee48dc52a09c --- /dev/null +++ b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/aggregate/test_other.py @@ -0,0 +1,675 @@ +""" +test all other .agg behavior +""" + +import datetime as dt +from functools import partial + +import numpy as np +import pytest + +from pandas.errors import SpecificationError + +import pandas as pd +from pandas import ( + DataFrame, + Index, + MultiIndex, + PeriodIndex, + Series, + date_range, + period_range, +) +import pandas._testing as tm + +from pandas.io.formats.printing import pprint_thing + + +def test_agg_partial_failure_raises(): + # GH#43741 + + df = DataFrame( + { + "data1": np.random.default_rng(2).standard_normal(5), + "data2": np.random.default_rng(2).standard_normal(5), + "key1": ["a", "a", "b", "b", "a"], + "key2": ["one", "two", "one", "two", "one"], + } + ) + grouped = df.groupby("key1") + + def peak_to_peak(arr): + return arr.max() - arr.min() + + with pytest.raises(TypeError, match="unsupported operand type"): + grouped.agg([peak_to_peak]) + + with pytest.raises(TypeError, match="unsupported operand type"): + grouped.agg(peak_to_peak) + + +def test_agg_datetimes_mixed(): + data = [[1, "2012-01-01", 1.0], [2, "2012-01-02", 2.0], [3, None, 3.0]] + + df1 = DataFrame( + { + "key": [x[0] for x in data], + "date": [x[1] for x in data], + "value": [x[2] for x in data], + } + ) + + data = [ + [ + row[0], + (dt.datetime.strptime(row[1], "%Y-%m-%d").date() if row[1] else None), + row[2], + ] + for row in data + ] + + df2 = DataFrame( + { + "key": [x[0] for x in data], + "date": [x[1] for x in data], + "value": [x[2] for x in data], + } + ) + + df1["weights"] = df1["value"] / df1["value"].sum() + gb1 = df1.groupby("date").aggregate("sum") + + df2["weights"] = df1["value"] / df1["value"].sum() + gb2 = df2.groupby("date").aggregate("sum") + + assert len(gb1) == len(gb2) + + +def test_agg_period_index(): + prng = period_range("2012-1-1", freq="M", periods=3) + df = DataFrame(np.random.default_rng(2).standard_normal((3, 2)), index=prng) + rs = df.groupby(level=0).sum() + assert isinstance(rs.index, PeriodIndex) + + # GH 3579 + index = period_range(start="1999-01", periods=5, freq="M") + s1 = Series(np.random.default_rng(2).random(len(index)), index=index) + s2 = Series(np.random.default_rng(2).random(len(index)), index=index) + df = DataFrame.from_dict({"s1": s1, "s2": s2}) + grouped = df.groupby(df.index.month) + list(grouped) + + +def test_agg_dict_parameter_cast_result_dtypes(): + # GH 12821 + + df = DataFrame( + { + "class": ["A", "A", "B", "B", "C", "C", "D", "D"], + "time": date_range("1/1/2011", periods=8, freq="h"), + } + ) + df.loc[[0, 1, 2, 5], "time"] = None + + # test for `first` function + exp = df.loc[[0, 3, 4, 6]].set_index("class") + grouped = df.groupby("class") + tm.assert_frame_equal(grouped.first(), exp) + tm.assert_frame_equal(grouped.agg("first"), exp) + tm.assert_frame_equal(grouped.agg({"time": "first"}), exp) + tm.assert_series_equal(grouped.time.first(), exp["time"]) + tm.assert_series_equal(grouped.time.agg("first"), exp["time"]) + + # test for `last` function + exp = df.loc[[0, 3, 4, 7]].set_index("class") + grouped = df.groupby("class") + tm.assert_frame_equal(grouped.last(), exp) + tm.assert_frame_equal(grouped.agg("last"), exp) + tm.assert_frame_equal(grouped.agg({"time": "last"}), exp) + tm.assert_series_equal(grouped.time.last(), exp["time"]) + tm.assert_series_equal(grouped.time.agg("last"), exp["time"]) + + # count + exp = Series([2, 2, 2, 2], index=Index(list("ABCD"), name="class"), name="time") + tm.assert_series_equal(grouped.time.agg(len), exp) + tm.assert_series_equal(grouped.time.size(), exp) + + exp = Series([0, 1, 1, 2], index=Index(list("ABCD"), name="class"), name="time") + tm.assert_series_equal(grouped.time.count(), exp) + + +def test_agg_cast_results_dtypes(): + # similar to GH12821 + # xref #11444 + u = [dt.datetime(2015, x + 1, 1) for x in range(12)] + v = list("aaabbbbbbccd") + df = DataFrame({"X": v, "Y": u}) + + result = df.groupby("X")["Y"].agg(len) + expected = df.groupby("X")["Y"].count() + tm.assert_series_equal(result, expected) + + +def test_aggregate_float64_no_int64(): + # see gh-11199 + df = DataFrame({"a": [1, 2, 3, 4, 5], "b": [1, 2, 2, 4, 5], "c": [1, 2, 3, 4, 5]}) + + expected = DataFrame({"a": [1, 2.5, 4, 5]}, index=[1, 2, 4, 5]) + expected.index.name = "b" + + result = df.groupby("b")[["a"]].mean() + tm.assert_frame_equal(result, expected) + + expected = DataFrame({"a": [1, 2.5, 4, 5], "c": [1, 2.5, 4, 5]}, index=[1, 2, 4, 5]) + expected.index.name = "b" + + result = df.groupby("b")[["a", "c"]].mean() + tm.assert_frame_equal(result, expected) + + +def test_aggregate_api_consistency(): + # GH 9052 + # make sure that the aggregates via dict + # are consistent + df = DataFrame( + { + "A": ["foo", "bar", "foo", "bar", "foo", "bar", "foo", "foo"], + "B": ["one", "one", "two", "two", "two", "two", "one", "two"], + "C": np.random.default_rng(2).standard_normal(8) + 1.0, + "D": np.arange(8), + } + ) + + grouped = df.groupby(["A", "B"]) + c_mean = grouped["C"].mean() + c_sum = grouped["C"].sum() + d_mean = grouped["D"].mean() + d_sum = grouped["D"].sum() + + result = grouped["D"].agg(["sum", "mean"]) + expected = pd.concat([d_sum, d_mean], axis=1) + expected.columns = ["sum", "mean"] + tm.assert_frame_equal(result, expected, check_like=True) + + result = grouped.agg(["sum", "mean"]) + expected = pd.concat([c_sum, c_mean, d_sum, d_mean], axis=1) + expected.columns = MultiIndex.from_product([["C", "D"], ["sum", "mean"]]) + tm.assert_frame_equal(result, expected, check_like=True) + + result = grouped[["D", "C"]].agg(["sum", "mean"]) + expected = pd.concat([d_sum, d_mean, c_sum, c_mean], axis=1) + expected.columns = MultiIndex.from_product([["D", "C"], ["sum", "mean"]]) + tm.assert_frame_equal(result, expected, check_like=True) + + result = grouped.agg({"C": "mean", "D": "sum"}) + expected = pd.concat([d_sum, c_mean], axis=1) + tm.assert_frame_equal(result, expected, check_like=True) + + result = grouped.agg({"C": ["mean", "sum"], "D": ["mean", "sum"]}) + expected = pd.concat([c_mean, c_sum, d_mean, d_sum], axis=1) + expected.columns = MultiIndex.from_product([["C", "D"], ["mean", "sum"]]) + + msg = r"Column\(s\) \['r', 'r2'\] do not exist" + with pytest.raises(KeyError, match=msg): + grouped[["D", "C"]].agg({"r": "sum", "r2": "mean"}) + + +def test_agg_dict_renaming_deprecation(): + # 15931 + df = DataFrame({"A": [1, 1, 1, 2, 2], "B": range(5), "C": range(5)}) + + msg = r"nested renamer is not supported" + with pytest.raises(SpecificationError, match=msg): + df.groupby("A").agg( + {"B": {"foo": ["sum", "max"]}, "C": {"bar": ["count", "min"]}} + ) + + msg = r"Column\(s\) \['ma'\] do not exist" + with pytest.raises(KeyError, match=msg): + df.groupby("A")[["B", "C"]].agg({"ma": "max"}) + + msg = r"nested renamer is not supported" + with pytest.raises(SpecificationError, match=msg): + df.groupby("A").B.agg({"foo": "count"}) + + +def test_agg_compat(): + # GH 12334 + df = DataFrame( + { + "A": ["foo", "bar", "foo", "bar", "foo", "bar", "foo", "foo"], + "B": ["one", "one", "two", "two", "two", "two", "one", "two"], + "C": np.random.default_rng(2).standard_normal(8) + 1.0, + "D": np.arange(8), + } + ) + + g = df.groupby(["A", "B"]) + + msg = r"nested renamer is not supported" + with pytest.raises(SpecificationError, match=msg): + g["D"].agg({"C": ["sum", "std"]}) + + with pytest.raises(SpecificationError, match=msg): + g["D"].agg({"C": "sum", "D": "std"}) + + +def test_agg_nested_dicts(): + # API change for disallowing these types of nested dicts + df = DataFrame( + { + "A": ["foo", "bar", "foo", "bar", "foo", "bar", "foo", "foo"], + "B": ["one", "one", "two", "two", "two", "two", "one", "two"], + "C": np.random.default_rng(2).standard_normal(8) + 1.0, + "D": np.arange(8), + } + ) + + g = df.groupby(["A", "B"]) + + msg = r"nested renamer is not supported" + with pytest.raises(SpecificationError, match=msg): + g.aggregate({"r1": {"C": ["mean", "sum"]}, "r2": {"D": ["mean", "sum"]}}) + + with pytest.raises(SpecificationError, match=msg): + g.agg({"C": {"ra": ["mean", "std"]}, "D": {"rb": ["mean", "std"]}}) + + # same name as the original column + # GH9052 + with pytest.raises(SpecificationError, match=msg): + g["D"].agg({"result1": np.sum, "result2": np.mean}) + + with pytest.raises(SpecificationError, match=msg): + g["D"].agg({"D": np.sum, "result2": np.mean}) + + +def test_agg_item_by_item_raise_typeerror(): + df = DataFrame(np.random.default_rng(2).integers(10, size=(20, 10))) + + def raiseException(df): + pprint_thing("----------------------------------------") + pprint_thing(df.to_string()) + raise TypeError("test") + + with pytest.raises(TypeError, match="test"): + df.groupby(0).agg(raiseException) + + +def test_series_agg_multikey(): + ts = Series( + np.arange(10, dtype=np.float64), index=date_range("2020-01-01", periods=10) + ) + grouped = ts.groupby([lambda x: x.year, lambda x: x.month]) + + result = grouped.agg("sum") + expected = grouped.sum() + tm.assert_series_equal(result, expected) + + +def test_series_agg_multi_pure_python(): + data = DataFrame( + { + "A": [ + "foo", + "foo", + "foo", + "foo", + "bar", + "bar", + "bar", + "bar", + "foo", + "foo", + "foo", + ], + "B": [ + "one", + "one", + "one", + "two", + "one", + "one", + "one", + "two", + "two", + "two", + "one", + ], + "C": [ + "dull", + "dull", + "shiny", + "dull", + "dull", + "shiny", + "shiny", + "dull", + "shiny", + "shiny", + "shiny", + ], + "D": np.random.default_rng(2).standard_normal(11), + "E": np.random.default_rng(2).standard_normal(11), + "F": np.random.default_rng(2).standard_normal(11), + } + ) + + def bad(x): + assert len(x.values.base) > 0 + return "foo" + + result = data.groupby(["A", "B"]).agg(bad) + expected = data.groupby(["A", "B"]).agg(lambda x: "foo") + tm.assert_frame_equal(result, expected) + + +def test_agg_consistency(): + # agg with ([]) and () not consistent + # GH 6715 + def P1(a): + return np.percentile(a.dropna(), q=1) + + df = DataFrame( + { + "col1": [1, 2, 3, 4], + "col2": [10, 25, 26, 31], + "date": [ + dt.date(2013, 2, 10), + dt.date(2013, 2, 10), + dt.date(2013, 2, 11), + dt.date(2013, 2, 11), + ], + } + ) + + g = df.groupby("date") + + expected = g.agg([P1]) + expected.columns = expected.columns.levels[0] + + result = g.agg(P1) + tm.assert_frame_equal(result, expected) + + +def test_agg_callables(): + # GH 7929 + df = DataFrame({"foo": [1, 2], "bar": [3, 4]}).astype(np.int64) + + class fn_class: + def __call__(self, x): + return sum(x) + + equiv_callables = [ + sum, + np.sum, + lambda x: sum(x), + lambda x: x.sum(), + partial(sum), + fn_class(), + ] + + expected = df.groupby("foo").agg("sum") + for ecall in equiv_callables: + warn = FutureWarning if ecall is sum or ecall is np.sum else None + msg = "using DataFrameGroupBy.sum" + with tm.assert_produces_warning(warn, match=msg): + result = df.groupby("foo").agg(ecall) + tm.assert_frame_equal(result, expected) + + +def test_agg_over_numpy_arrays(): + # GH 3788 + df = DataFrame( + [ + [1, np.array([10, 20, 30])], + [1, np.array([40, 50, 60])], + [2, np.array([20, 30, 40])], + ], + columns=["category", "arraydata"], + ) + gb = df.groupby("category") + + expected_data = [[np.array([50, 70, 90])], [np.array([20, 30, 40])]] + expected_index = Index([1, 2], name="category") + expected_column = ["arraydata"] + expected = DataFrame(expected_data, index=expected_index, columns=expected_column) + + alt = gb.sum(numeric_only=False) + tm.assert_frame_equal(alt, expected) + + result = gb.agg("sum", numeric_only=False) + tm.assert_frame_equal(result, expected) + + # FIXME: the original version of this test called `gb.agg(sum)` + # and that raises TypeError if `numeric_only=False` is passed + + +@pytest.mark.parametrize("as_period", [True, False]) +def test_agg_tzaware_non_datetime_result(as_period): + # discussed in GH#29589, fixed in GH#29641, operating on tzaware values + # with function that is not dtype-preserving + dti = date_range("2012-01-01", periods=4, tz="UTC") + if as_period: + dti = dti.tz_localize(None).to_period("D") + + df = DataFrame({"a": [0, 0, 1, 1], "b": dti}) + gb = df.groupby("a") + + # Case that _does_ preserve the dtype + result = gb["b"].agg(lambda x: x.iloc[0]) + expected = Series(dti[::2], name="b") + expected.index.name = "a" + tm.assert_series_equal(result, expected) + + # Cases that do _not_ preserve the dtype + result = gb["b"].agg(lambda x: x.iloc[0].year) + expected = Series([2012, 2012], name="b") + expected.index.name = "a" + tm.assert_series_equal(result, expected) + + result = gb["b"].agg(lambda x: x.iloc[-1] - x.iloc[0]) + expected = Series([pd.Timedelta(days=1), pd.Timedelta(days=1)], name="b") + expected.index.name = "a" + if as_period: + expected = Series([pd.offsets.Day(1), pd.offsets.Day(1)], name="b") + expected.index.name = "a" + tm.assert_series_equal(result, expected) + + +def test_agg_timezone_round_trip(): + # GH 15426 + ts = pd.Timestamp("2016-01-01 12:00:00", tz="US/Pacific") + df = DataFrame({"a": 1, "b": [ts + dt.timedelta(minutes=nn) for nn in range(10)]}) + + result1 = df.groupby("a")["b"].agg("min").iloc[0] + result2 = df.groupby("a")["b"].agg(lambda x: np.min(x)).iloc[0] + result3 = df.groupby("a")["b"].min().iloc[0] + + assert result1 == ts + assert result2 == ts + assert result3 == ts + + dates = [ + pd.Timestamp(f"2016-01-0{i:d} 12:00:00", tz="US/Pacific") for i in range(1, 5) + ] + df = DataFrame({"A": ["a", "b"] * 2, "B": dates}) + grouped = df.groupby("A") + + ts = df["B"].iloc[0] + assert ts == grouped.nth(0)["B"].iloc[0] + assert ts == grouped.head(1)["B"].iloc[0] + assert ts == grouped.first()["B"].iloc[0] + + # GH#27110 applying iloc should return a DataFrame + msg = "DataFrameGroupBy.apply operated on the grouping columns" + with tm.assert_produces_warning(DeprecationWarning, match=msg): + assert ts == grouped.apply(lambda x: x.iloc[0]).iloc[0, 1] + + ts = df["B"].iloc[2] + assert ts == grouped.last()["B"].iloc[0] + + # GH#27110 applying iloc should return a DataFrame + msg = "DataFrameGroupBy.apply operated on the grouping columns" + with tm.assert_produces_warning(DeprecationWarning, match=msg): + assert ts == grouped.apply(lambda x: x.iloc[-1]).iloc[0, 1] + + +def test_sum_uint64_overflow(): + # see gh-14758 + # Convert to uint64 and don't overflow + df = DataFrame([[1, 2], [3, 4], [5, 6]], dtype=object) + df = df + 9223372036854775807 + + index = Index( + [9223372036854775808, 9223372036854775810, 9223372036854775812], dtype=np.uint64 + ) + expected = DataFrame( + {1: [9223372036854775809, 9223372036854775811, 9223372036854775813]}, + index=index, + dtype=object, + ) + + expected.index.name = 0 + result = df.groupby(0).sum(numeric_only=False) + tm.assert_frame_equal(result, expected) + + # out column is non-numeric, so with numeric_only=True it is dropped + result2 = df.groupby(0).sum(numeric_only=True) + expected2 = expected[[]] + tm.assert_frame_equal(result2, expected2) + + +@pytest.mark.parametrize( + "structure, expected", + [ + (tuple, DataFrame({"C": {(1, 1): (1, 1, 1), (3, 4): (3, 4, 4)}})), + (list, DataFrame({"C": {(1, 1): [1, 1, 1], (3, 4): [3, 4, 4]}})), + ( + lambda x: tuple(x), + DataFrame({"C": {(1, 1): (1, 1, 1), (3, 4): (3, 4, 4)}}), + ), + ( + lambda x: list(x), + DataFrame({"C": {(1, 1): [1, 1, 1], (3, 4): [3, 4, 4]}}), + ), + ], +) +def test_agg_structs_dataframe(structure, expected): + df = DataFrame( + {"A": [1, 1, 1, 3, 3, 3], "B": [1, 1, 1, 4, 4, 4], "C": [1, 1, 1, 3, 4, 4]} + ) + + result = df.groupby(["A", "B"]).aggregate(structure) + expected.index.names = ["A", "B"] + tm.assert_frame_equal(result, expected) + + +@pytest.mark.parametrize( + "structure, expected", + [ + (tuple, Series([(1, 1, 1), (3, 4, 4)], index=[1, 3], name="C")), + (list, Series([[1, 1, 1], [3, 4, 4]], index=[1, 3], name="C")), + (lambda x: tuple(x), Series([(1, 1, 1), (3, 4, 4)], index=[1, 3], name="C")), + (lambda x: list(x), Series([[1, 1, 1], [3, 4, 4]], index=[1, 3], name="C")), + ], +) +def test_agg_structs_series(structure, expected): + # Issue #18079 + df = DataFrame( + {"A": [1, 1, 1, 3, 3, 3], "B": [1, 1, 1, 4, 4, 4], "C": [1, 1, 1, 3, 4, 4]} + ) + + result = df.groupby("A")["C"].aggregate(structure) + expected.index.name = "A" + tm.assert_series_equal(result, expected) + + +def test_agg_category_nansum(observed): + categories = ["a", "b", "c"] + df = DataFrame( + {"A": pd.Categorical(["a", "a", "b"], categories=categories), "B": [1, 2, 3]} + ) + msg = "using SeriesGroupBy.sum" + with tm.assert_produces_warning(FutureWarning, match=msg): + result = df.groupby("A", observed=observed).B.agg(np.nansum) + expected = Series( + [3, 3, 0], + index=pd.CategoricalIndex(["a", "b", "c"], categories=categories, name="A"), + name="B", + ) + if observed: + expected = expected[expected != 0] + tm.assert_series_equal(result, expected) + + +def test_agg_list_like_func(): + # GH 18473 + df = DataFrame({"A": [str(x) for x in range(3)], "B": [str(x) for x in range(3)]}) + grouped = df.groupby("A", as_index=False, sort=False) + result = grouped.agg({"B": lambda x: list(x)}) + expected = DataFrame( + {"A": [str(x) for x in range(3)], "B": [[str(x)] for x in range(3)]} + ) + tm.assert_frame_equal(result, expected) + + +def test_agg_lambda_with_timezone(): + # GH 23683 + df = DataFrame( + { + "tag": [1, 1], + "date": [ + pd.Timestamp("2018-01-01", tz="UTC"), + pd.Timestamp("2018-01-02", tz="UTC"), + ], + } + ) + result = df.groupby("tag").agg({"date": lambda e: e.head(1)}) + expected = DataFrame( + [pd.Timestamp("2018-01-01", tz="UTC")], + index=Index([1], name="tag"), + columns=["date"], + ) + tm.assert_frame_equal(result, expected) + + +@pytest.mark.parametrize( + "err_cls", + [ + NotImplementedError, + RuntimeError, + KeyError, + IndexError, + OSError, + ValueError, + ArithmeticError, + AttributeError, + ], +) +def test_groupby_agg_err_catching(err_cls): + # make sure we suppress anything other than TypeError or AssertionError + # in _python_agg_general + + # Use a non-standard EA to make sure we don't go down ndarray paths + from pandas.tests.extension.decimal.array import ( + DecimalArray, + make_data, + to_decimal, + ) + + data = make_data()[:5] + df = DataFrame( + {"id1": [0, 0, 0, 1, 1], "id2": [0, 1, 0, 1, 1], "decimals": DecimalArray(data)} + ) + + expected = Series(to_decimal([data[0], data[3]])) + + def weird_func(x): + # weird function that raise something other than TypeError or IndexError + # in _python_agg_general + if len(x) == 0: + raise err_cls + return x.iloc[0] + + result = df["decimals"].groupby(df["id1"]).agg(weird_func) + tm.assert_series_equal(result, expected, check_names=False) diff --git a/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/__pycache__/__init__.cpython-310.pyc b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/__pycache__/__init__.cpython-310.pyc new file mode 100644 index 0000000000000000000000000000000000000000..83623bf37631a4c0b90d1f0004d16f36718c7e95 Binary files /dev/null and b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/__pycache__/__init__.cpython-310.pyc differ diff --git a/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/__pycache__/test_corrwith.cpython-310.pyc b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/__pycache__/test_corrwith.cpython-310.pyc new file mode 100644 index 0000000000000000000000000000000000000000..44570f45b0d6df34f34cd2bed5816e81b708bf79 Binary files /dev/null and b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/__pycache__/test_corrwith.cpython-310.pyc differ diff --git a/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/__pycache__/test_describe.cpython-310.pyc b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/__pycache__/test_describe.cpython-310.pyc new file mode 100644 index 0000000000000000000000000000000000000000..29b6606f1dbca77b13dd36724c25836e4d592dc2 Binary files /dev/null and b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/__pycache__/test_describe.cpython-310.pyc differ diff --git a/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/__pycache__/test_groupby_shift_diff.cpython-310.pyc b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/__pycache__/test_groupby_shift_diff.cpython-310.pyc new file mode 100644 index 0000000000000000000000000000000000000000..9b636ef8ad9ff9bc3c0310ee1559d217999cfe2f Binary files /dev/null and b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/__pycache__/test_groupby_shift_diff.cpython-310.pyc differ diff --git a/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/__pycache__/test_is_monotonic.cpython-310.pyc b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/__pycache__/test_is_monotonic.cpython-310.pyc new file mode 100644 index 0000000000000000000000000000000000000000..f7e232a44fb11c009638c12da415a10b8319d3e9 Binary files /dev/null and b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/__pycache__/test_is_monotonic.cpython-310.pyc differ diff --git a/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/__pycache__/test_nlargest_nsmallest.cpython-310.pyc b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/__pycache__/test_nlargest_nsmallest.cpython-310.pyc new file mode 100644 index 0000000000000000000000000000000000000000..591262a2dfb8480eaa6855dea0334974782853c7 Binary files /dev/null and b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/__pycache__/test_nlargest_nsmallest.cpython-310.pyc differ diff --git a/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/__pycache__/test_nth.cpython-310.pyc b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/__pycache__/test_nth.cpython-310.pyc new file mode 100644 index 0000000000000000000000000000000000000000..ac391f5d681818f19c1c263c7b3e180929ab514a Binary files /dev/null and b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/__pycache__/test_nth.cpython-310.pyc differ diff --git a/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/__pycache__/test_quantile.cpython-310.pyc b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/__pycache__/test_quantile.cpython-310.pyc new file mode 100644 index 0000000000000000000000000000000000000000..73a42a479d33a22271f75a47ea5cf0f146ed8c95 Binary files /dev/null and b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/__pycache__/test_quantile.cpython-310.pyc differ diff --git a/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/__pycache__/test_rank.cpython-310.pyc b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/__pycache__/test_rank.cpython-310.pyc new file mode 100644 index 0000000000000000000000000000000000000000..5cab18bdee0966e9747c1bf6bb1b3f9430f8afce Binary files /dev/null and b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/__pycache__/test_rank.cpython-310.pyc differ diff --git a/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/__pycache__/test_sample.cpython-310.pyc b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/__pycache__/test_sample.cpython-310.pyc new file mode 100644 index 0000000000000000000000000000000000000000..c469d7b4ac3f0220458fb5f6de12b5fc8810a2c0 Binary files /dev/null and b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/__pycache__/test_sample.cpython-310.pyc differ diff --git a/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/__pycache__/test_size.cpython-310.pyc b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/__pycache__/test_size.cpython-310.pyc new file mode 100644 index 0000000000000000000000000000000000000000..1beac1f4d884bedf207620986e33d594125011c5 Binary files /dev/null and b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/__pycache__/test_size.cpython-310.pyc differ diff --git a/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/__pycache__/test_skew.cpython-310.pyc b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/__pycache__/test_skew.cpython-310.pyc new file mode 100644 index 0000000000000000000000000000000000000000..d7fb7261b8175763ca2f524739501427b6a4dbe1 Binary files /dev/null and b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/__pycache__/test_skew.cpython-310.pyc differ diff --git a/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/__pycache__/test_value_counts.cpython-310.pyc b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/__pycache__/test_value_counts.cpython-310.pyc new file mode 100644 index 0000000000000000000000000000000000000000..57c66e7c10ac1b9f4dfd7d696d8bf91244a7da25 Binary files /dev/null and b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/__pycache__/test_value_counts.cpython-310.pyc differ diff --git a/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/test_corrwith.py b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/test_corrwith.py new file mode 100644 index 0000000000000000000000000000000000000000..53e8bdc4534dc66dc1b68e603b2af431d0c0b209 --- /dev/null +++ b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/test_corrwith.py @@ -0,0 +1,24 @@ +import numpy as np + +from pandas import ( + DataFrame, + Index, + Series, +) +import pandas._testing as tm + + +def test_corrwith_with_1_axis(): + # GH 47723 + df = DataFrame({"a": [1, 1, 2], "b": [3, 7, 4]}) + gb = df.groupby("a") + + msg = "DataFrameGroupBy.corrwith with axis=1 is deprecated" + with tm.assert_produces_warning(FutureWarning, match=msg): + result = gb.corrwith(df, axis=1) + index = Index( + data=[(1, 0), (1, 1), (1, 2), (2, 2), (2, 0), (2, 1)], + name=("a", None), + ) + expected = Series([np.nan] * 6, index=index) + tm.assert_series_equal(result, expected) diff --git a/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/test_describe.py b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/test_describe.py new file mode 100644 index 0000000000000000000000000000000000000000..a2440e09dfc02436140e94cd689b39a1a9d35189 --- /dev/null +++ b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/test_describe.py @@ -0,0 +1,297 @@ +import numpy as np +import pytest + +import pandas as pd +from pandas import ( + DataFrame, + Index, + MultiIndex, + Series, + Timestamp, + date_range, +) +import pandas._testing as tm + + +def test_apply_describe_bug(multiindex_dataframe_random_data): + grouped = multiindex_dataframe_random_data.groupby(level="first") + grouped.describe() # it works! + + +def test_series_describe_multikey(): + ts = Series( + np.arange(10, dtype=np.float64), index=date_range("2020-01-01", periods=10) + ) + grouped = ts.groupby([lambda x: x.year, lambda x: x.month]) + result = grouped.describe() + tm.assert_series_equal(result["mean"], grouped.mean(), check_names=False) + tm.assert_series_equal(result["std"], grouped.std(), check_names=False) + tm.assert_series_equal(result["min"], grouped.min(), check_names=False) + + +def test_series_describe_single(): + ts = Series( + np.arange(10, dtype=np.float64), index=date_range("2020-01-01", periods=10) + ) + grouped = ts.groupby(lambda x: x.month) + result = grouped.apply(lambda x: x.describe()) + expected = grouped.describe().stack(future_stack=True) + tm.assert_series_equal(result, expected) + + +@pytest.mark.parametrize("keys", ["key1", ["key1", "key2"]]) +def test_series_describe_as_index(as_index, keys): + # GH#49256 + df = DataFrame( + { + "key1": ["one", "two", "two", "three", "two"], + "key2": ["one", "two", "two", "three", "two"], + "foo2": [1, 2, 4, 4, 6], + } + ) + gb = df.groupby(keys, as_index=as_index)["foo2"] + result = gb.describe() + expected = DataFrame( + { + "key1": ["one", "three", "two"], + "count": [1.0, 1.0, 3.0], + "mean": [1.0, 4.0, 4.0], + "std": [np.nan, np.nan, 2.0], + "min": [1.0, 4.0, 2.0], + "25%": [1.0, 4.0, 3.0], + "50%": [1.0, 4.0, 4.0], + "75%": [1.0, 4.0, 5.0], + "max": [1.0, 4.0, 6.0], + } + ) + if len(keys) == 2: + expected.insert(1, "key2", expected["key1"]) + if as_index: + expected = expected.set_index(keys) + tm.assert_frame_equal(result, expected) + + +def test_frame_describe_multikey(tsframe): + grouped = tsframe.groupby([lambda x: x.year, lambda x: x.month]) + result = grouped.describe() + desc_groups = [] + for col in tsframe: + group = grouped[col].describe() + # GH 17464 - Remove duplicate MultiIndex levels + group_col = MultiIndex( + levels=[[col], group.columns], + codes=[[0] * len(group.columns), range(len(group.columns))], + ) + group = DataFrame(group.values, columns=group_col, index=group.index) + desc_groups.append(group) + expected = pd.concat(desc_groups, axis=1) + tm.assert_frame_equal(result, expected) + + msg = "DataFrame.groupby with axis=1 is deprecated" + with tm.assert_produces_warning(FutureWarning, match=msg): + groupedT = tsframe.groupby({"A": 0, "B": 0, "C": 1, "D": 1}, axis=1) + result = groupedT.describe() + expected = tsframe.describe().T + # reverting the change from https://github.com/pandas-dev/pandas/pull/35441/ + expected.index = MultiIndex( + levels=[[0, 1], expected.index], + codes=[[0, 0, 1, 1], range(len(expected.index))], + ) + tm.assert_frame_equal(result, expected) + + +def test_frame_describe_tupleindex(): + # GH 14848 - regression from 0.19.0 to 0.19.1 + df1 = DataFrame( + { + "x": [1, 2, 3, 4, 5] * 3, + "y": [10, 20, 30, 40, 50] * 3, + "z": [100, 200, 300, 400, 500] * 3, + } + ) + df1["k"] = [(0, 0, 1), (0, 1, 0), (1, 0, 0)] * 5 + df2 = df1.rename(columns={"k": "key"}) + msg = "Names should be list-like for a MultiIndex" + with pytest.raises(ValueError, match=msg): + df1.groupby("k").describe() + with pytest.raises(ValueError, match=msg): + df2.groupby("key").describe() + + +def test_frame_describe_unstacked_format(): + # GH 4792 + prices = { + Timestamp("2011-01-06 10:59:05", tz=None): 24990, + Timestamp("2011-01-06 12:43:33", tz=None): 25499, + Timestamp("2011-01-06 12:54:09", tz=None): 25499, + } + volumes = { + Timestamp("2011-01-06 10:59:05", tz=None): 1500000000, + Timestamp("2011-01-06 12:43:33", tz=None): 5000000000, + Timestamp("2011-01-06 12:54:09", tz=None): 100000000, + } + df = DataFrame({"PRICE": prices, "VOLUME": volumes}) + result = df.groupby("PRICE").VOLUME.describe() + data = [ + df[df.PRICE == 24990].VOLUME.describe().values.tolist(), + df[df.PRICE == 25499].VOLUME.describe().values.tolist(), + ] + expected = DataFrame( + data, + index=Index([24990, 25499], name="PRICE"), + columns=["count", "mean", "std", "min", "25%", "50%", "75%", "max"], + ) + tm.assert_frame_equal(result, expected) + + +@pytest.mark.filterwarnings( + "ignore:" + "indexing past lexsort depth may impact performance:" + "pandas.errors.PerformanceWarning" +) +@pytest.mark.parametrize("as_index", [True, False]) +@pytest.mark.parametrize("keys", [["a1"], ["a1", "a2"]]) +def test_describe_with_duplicate_output_column_names(as_index, keys): + # GH 35314 + df = DataFrame( + { + "a1": [99, 99, 99, 88, 88, 88], + "a2": [99, 99, 99, 88, 88, 88], + "b": [1, 2, 3, 4, 5, 6], + "c": [10, 20, 30, 40, 50, 60], + }, + columns=["a1", "a2", "b", "b"], + copy=False, + ) + if keys == ["a1"]: + df = df.drop(columns="a2") + + expected = ( + DataFrame.from_records( + [ + ("b", "count", 3.0, 3.0), + ("b", "mean", 5.0, 2.0), + ("b", "std", 1.0, 1.0), + ("b", "min", 4.0, 1.0), + ("b", "25%", 4.5, 1.5), + ("b", "50%", 5.0, 2.0), + ("b", "75%", 5.5, 2.5), + ("b", "max", 6.0, 3.0), + ("b", "count", 3.0, 3.0), + ("b", "mean", 5.0, 2.0), + ("b", "std", 1.0, 1.0), + ("b", "min", 4.0, 1.0), + ("b", "25%", 4.5, 1.5), + ("b", "50%", 5.0, 2.0), + ("b", "75%", 5.5, 2.5), + ("b", "max", 6.0, 3.0), + ], + ) + .set_index([0, 1]) + .T + ) + expected.columns.names = [None, None] + if len(keys) == 2: + expected.index = MultiIndex( + levels=[[88, 99], [88, 99]], codes=[[0, 1], [0, 1]], names=["a1", "a2"] + ) + else: + expected.index = Index([88, 99], name="a1") + + if not as_index: + expected = expected.reset_index() + + result = df.groupby(keys, as_index=as_index).describe() + + tm.assert_frame_equal(result, expected) + + +def test_describe_duplicate_columns(): + # GH#50806 + df = DataFrame([[0, 1, 2, 3]]) + df.columns = [0, 1, 2, 0] + gb = df.groupby(df[1]) + result = gb.describe(percentiles=[]) + + columns = ["count", "mean", "std", "min", "50%", "max"] + frames = [ + DataFrame([[1.0, val, np.nan, val, val, val]], index=[1], columns=columns) + for val in (0.0, 2.0, 3.0) + ] + expected = pd.concat(frames, axis=1) + expected.columns = MultiIndex( + levels=[[0, 2], columns], + codes=[6 * [0] + 6 * [1] + 6 * [0], 3 * list(range(6))], + ) + expected.index.names = [1] + tm.assert_frame_equal(result, expected) + + +class TestGroupByNonCythonPaths: + # GH#5610 non-cython calls should not include the grouper + # Tests for code not expected to go through cython paths. + + @pytest.fixture + def df(self): + df = DataFrame( + [[1, 2, "foo"], [1, np.nan, "bar"], [3, np.nan, "baz"]], + columns=["A", "B", "C"], + ) + return df + + @pytest.fixture + def gb(self, df): + gb = df.groupby("A") + return gb + + @pytest.fixture + def gni(self, df): + gni = df.groupby("A", as_index=False) + return gni + + def test_describe(self, df, gb, gni): + # describe + expected_index = Index([1, 3], name="A") + expected_col = MultiIndex( + levels=[["B"], ["count", "mean", "std", "min", "25%", "50%", "75%", "max"]], + codes=[[0] * 8, list(range(8))], + ) + expected = DataFrame( + [ + [1.0, 2.0, np.nan, 2.0, 2.0, 2.0, 2.0, 2.0], + [0.0, np.nan, np.nan, np.nan, np.nan, np.nan, np.nan, np.nan], + ], + index=expected_index, + columns=expected_col, + ) + result = gb.describe() + tm.assert_frame_equal(result, expected) + + expected = expected.reset_index() + result = gni.describe() + tm.assert_frame_equal(result, expected) + + +@pytest.mark.parametrize("dtype", [int, float, object]) +@pytest.mark.parametrize( + "kwargs", + [ + {"percentiles": [0.10, 0.20, 0.30], "include": "all", "exclude": None}, + {"percentiles": [0.10, 0.20, 0.30], "include": None, "exclude": ["int"]}, + {"percentiles": [0.10, 0.20, 0.30], "include": ["int"], "exclude": None}, + ], +) +def test_groupby_empty_dataset(dtype, kwargs): + # GH#41575 + df = DataFrame([[1, 2, 3]], columns=["A", "B", "C"], dtype=dtype) + df["B"] = df["B"].astype(int) + df["C"] = df["C"].astype(float) + + result = df.iloc[:0].groupby("A").describe(**kwargs) + expected = df.groupby("A").describe(**kwargs).reset_index(drop=True).iloc[:0] + tm.assert_frame_equal(result, expected) + + result = df.iloc[:0].groupby("A").B.describe(**kwargs) + expected = df.groupby("A").B.describe(**kwargs).reset_index(drop=True).iloc[:0] + expected.index = Index([]) + tm.assert_frame_equal(result, expected) diff --git a/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/test_groupby_shift_diff.py b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/test_groupby_shift_diff.py new file mode 100644 index 0000000000000000000000000000000000000000..94e672d4892feb513f75d9a3d3376e261e2c0f36 --- /dev/null +++ b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/test_groupby_shift_diff.py @@ -0,0 +1,255 @@ +import numpy as np +import pytest + +from pandas import ( + DataFrame, + NaT, + Series, + Timedelta, + Timestamp, + date_range, +) +import pandas._testing as tm + + +def test_group_shift_with_null_key(): + # This test is designed to replicate the segfault in issue #13813. + n_rows = 1200 + + # Generate a moderately large dataframe with occasional missing + # values in column `B`, and then group by [`A`, `B`]. This should + # force `-1` in `labels` array of `g._grouper.group_info` exactly + # at those places, where the group-by key is partially missing. + df = DataFrame( + [(i % 12, i % 3 if i % 3 else np.nan, i) for i in range(n_rows)], + dtype=float, + columns=["A", "B", "Z"], + index=None, + ) + g = df.groupby(["A", "B"]) + + expected = DataFrame( + [(i + 12 if i % 3 and i < n_rows - 12 else np.nan) for i in range(n_rows)], + dtype=float, + columns=["Z"], + index=None, + ) + result = g.shift(-1) + + tm.assert_frame_equal(result, expected) + + +def test_group_shift_with_fill_value(): + # GH #24128 + n_rows = 24 + df = DataFrame( + [(i % 12, i % 3, i) for i in range(n_rows)], + dtype=float, + columns=["A", "B", "Z"], + index=None, + ) + g = df.groupby(["A", "B"]) + + expected = DataFrame( + [(i + 12 if i < n_rows - 12 else 0) for i in range(n_rows)], + dtype=float, + columns=["Z"], + index=None, + ) + result = g.shift(-1, fill_value=0) + + tm.assert_frame_equal(result, expected) + + +def test_group_shift_lose_timezone(): + # GH 30134 + now_dt = Timestamp.utcnow().as_unit("ns") + df = DataFrame({"a": [1, 1], "date": now_dt}) + result = df.groupby("a").shift(0).iloc[0] + expected = Series({"date": now_dt}, name=result.name) + tm.assert_series_equal(result, expected) + + +def test_group_diff_real_series(any_real_numpy_dtype): + df = DataFrame( + {"a": [1, 2, 3, 3, 2], "b": [1, 2, 3, 4, 5]}, + dtype=any_real_numpy_dtype, + ) + result = df.groupby("a")["b"].diff() + exp_dtype = "float" + if any_real_numpy_dtype in ["int8", "int16", "float32"]: + exp_dtype = "float32" + expected = Series([np.nan, np.nan, np.nan, 1.0, 3.0], dtype=exp_dtype, name="b") + tm.assert_series_equal(result, expected) + + +def test_group_diff_real_frame(any_real_numpy_dtype): + df = DataFrame( + { + "a": [1, 2, 3, 3, 2], + "b": [1, 2, 3, 4, 5], + "c": [1, 2, 3, 4, 6], + }, + dtype=any_real_numpy_dtype, + ) + result = df.groupby("a").diff() + exp_dtype = "float" + if any_real_numpy_dtype in ["int8", "int16", "float32"]: + exp_dtype = "float32" + expected = DataFrame( + { + "b": [np.nan, np.nan, np.nan, 1.0, 3.0], + "c": [np.nan, np.nan, np.nan, 1.0, 4.0], + }, + dtype=exp_dtype, + ) + tm.assert_frame_equal(result, expected) + + +@pytest.mark.parametrize( + "data", + [ + [ + Timestamp("2013-01-01"), + Timestamp("2013-01-02"), + Timestamp("2013-01-03"), + ], + [Timedelta("5 days"), Timedelta("6 days"), Timedelta("7 days")], + ], +) +def test_group_diff_datetimelike(data, unit): + df = DataFrame({"a": [1, 2, 2], "b": data}) + df["b"] = df["b"].dt.as_unit(unit) + result = df.groupby("a")["b"].diff() + expected = Series([NaT, NaT, Timedelta("1 days")], name="b").dt.as_unit(unit) + tm.assert_series_equal(result, expected) + + +def test_group_diff_bool(): + df = DataFrame({"a": [1, 2, 3, 3, 2], "b": [True, True, False, False, True]}) + result = df.groupby("a")["b"].diff() + expected = Series([np.nan, np.nan, np.nan, False, False], name="b") + tm.assert_series_equal(result, expected) + + +def test_group_diff_object_raises(object_dtype): + df = DataFrame( + {"a": ["foo", "bar", "bar"], "b": ["baz", "foo", "foo"]}, dtype=object_dtype + ) + with pytest.raises(TypeError, match=r"unsupported operand type\(s\) for -"): + df.groupby("a")["b"].diff() + + +def test_empty_shift_with_fill(): + # GH 41264, single-index check + df = DataFrame(columns=["a", "b", "c"]) + shifted = df.groupby(["a"]).shift(1) + shifted_with_fill = df.groupby(["a"]).shift(1, fill_value=0) + tm.assert_frame_equal(shifted, shifted_with_fill) + tm.assert_index_equal(shifted.index, shifted_with_fill.index) + + +def test_multindex_empty_shift_with_fill(): + # GH 41264, multi-index check + df = DataFrame(columns=["a", "b", "c"]) + shifted = df.groupby(["a", "b"]).shift(1) + shifted_with_fill = df.groupby(["a", "b"]).shift(1, fill_value=0) + tm.assert_frame_equal(shifted, shifted_with_fill) + tm.assert_index_equal(shifted.index, shifted_with_fill.index) + + +def test_shift_periods_freq(): + # GH 54093 + data = {"a": [1, 2, 3, 4, 5, 6], "b": [0, 0, 0, 1, 1, 1]} + df = DataFrame(data, index=date_range(start="20100101", periods=6)) + result = df.groupby(df.index).shift(periods=-2, freq="D") + expected = DataFrame(data, index=date_range(start="2009-12-30", periods=6)) + tm.assert_frame_equal(result, expected) + + +def test_shift_deprecate_freq_and_fill_value(): + # GH 53832 + data = {"a": [1, 2, 3, 4, 5, 6], "b": [0, 0, 0, 1, 1, 1]} + df = DataFrame(data, index=date_range(start="20100101", periods=6)) + msg = ( + "Passing a 'freq' together with a 'fill_value' silently ignores the fill_value" + ) + with tm.assert_produces_warning(FutureWarning, match=msg): + df.groupby(df.index).shift(periods=-2, freq="D", fill_value="1") + + +def test_shift_disallow_suffix_if_periods_is_int(): + # GH#44424 + data = {"a": [1, 2, 3, 4, 5, 6], "b": [0, 0, 0, 1, 1, 1]} + df = DataFrame(data) + msg = "Cannot specify `suffix` if `periods` is an int." + with pytest.raises(ValueError, match=msg): + df.groupby("b").shift(1, suffix="fails") + + +def test_group_shift_with_multiple_periods(): + # GH#44424 + df = DataFrame({"a": [1, 2, 3, 3, 2], "b": [True, True, False, False, True]}) + + shifted_df = df.groupby("b")[["a"]].shift([0, 1]) + expected_df = DataFrame( + {"a_0": [1, 2, 3, 3, 2], "a_1": [np.nan, 1.0, np.nan, 3.0, 2.0]} + ) + tm.assert_frame_equal(shifted_df, expected_df) + + # series + shifted_series = df.groupby("b")["a"].shift([0, 1]) + tm.assert_frame_equal(shifted_series, expected_df) + + +def test_group_shift_with_multiple_periods_and_freq(): + # GH#44424 + df = DataFrame( + {"a": [1, 2, 3, 4, 5], "b": [True, True, False, False, True]}, + index=date_range("1/1/2000", periods=5, freq="h"), + ) + shifted_df = df.groupby("b")[["a"]].shift( + [0, 1], + freq="h", + ) + expected_df = DataFrame( + { + "a_0": [1.0, 2.0, 3.0, 4.0, 5.0, np.nan], + "a_1": [ + np.nan, + 1.0, + 2.0, + 3.0, + 4.0, + 5.0, + ], + }, + index=date_range("1/1/2000", periods=6, freq="h"), + ) + tm.assert_frame_equal(shifted_df, expected_df) + + +def test_group_shift_with_multiple_periods_and_fill_value(): + # GH#44424 + df = DataFrame( + {"a": [1, 2, 3, 4, 5], "b": [True, True, False, False, True]}, + ) + shifted_df = df.groupby("b")[["a"]].shift([0, 1], fill_value=-1) + expected_df = DataFrame( + {"a_0": [1, 2, 3, 4, 5], "a_1": [-1, 1, -1, 3, 2]}, + ) + tm.assert_frame_equal(shifted_df, expected_df) + + +def test_group_shift_with_multiple_periods_and_both_fill_and_freq_deprecated(): + # GH#44424 + df = DataFrame( + {"a": [1, 2, 3, 4, 5], "b": [True, True, False, False, True]}, + index=date_range("1/1/2000", periods=5, freq="h"), + ) + msg = ( + "Passing a 'freq' together with a 'fill_value' silently ignores the " + "fill_value" + ) + with tm.assert_produces_warning(FutureWarning, match=msg): + df.groupby("b")[["a"]].shift([1, 2], fill_value=1, freq="h") diff --git a/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/test_is_monotonic.py b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/test_is_monotonic.py new file mode 100644 index 0000000000000000000000000000000000000000..3428fc90f6e51a0bde0aba9c8ea08ebf414e5556 --- /dev/null +++ b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/test_is_monotonic.py @@ -0,0 +1,78 @@ +import numpy as np +import pytest + +from pandas import ( + DataFrame, + Index, + Series, +) +import pandas._testing as tm + + +@pytest.mark.parametrize( + "in_vals, out_vals", + [ + # Basics: strictly increasing (T), strictly decreasing (F), + # abs val increasing (F), non-strictly increasing (T) + ([1, 2, 5, 3, 2, 0, 4, 5, -6, 1, 1], [True, False, False, True]), + # Test with inf vals + ( + [1, 2.1, np.inf, 3, 2, np.inf, -np.inf, 5, 11, 1, -np.inf], + [True, False, True, False], + ), + # Test with nan vals; should always be False + ( + [1, 2, np.nan, 3, 2, np.nan, np.nan, 5, -np.inf, 1, np.nan], + [False, False, False, False], + ), + ], +) +def test_is_monotonic_increasing(in_vals, out_vals): + # GH 17015 + source_dict = { + "A": ["1", "2", "3", "4", "5", "6", "7", "8", "9", "10", "11"], + "B": ["a", "a", "a", "b", "b", "b", "c", "c", "c", "d", "d"], + "C": in_vals, + } + df = DataFrame(source_dict) + result = df.groupby("B").C.is_monotonic_increasing + index = Index(list("abcd"), name="B") + expected = Series(index=index, data=out_vals, name="C") + tm.assert_series_equal(result, expected) + + # Also check result equal to manually taking x.is_monotonic_increasing. + expected = df.groupby(["B"]).C.apply(lambda x: x.is_monotonic_increasing) + tm.assert_series_equal(result, expected) + + +@pytest.mark.parametrize( + "in_vals, out_vals", + [ + # Basics: strictly decreasing (T), strictly increasing (F), + # abs val decreasing (F), non-strictly increasing (T) + ([10, 9, 7, 3, 4, 5, -3, 2, 0, 1, 1], [True, False, False, True]), + # Test with inf vals + ( + [np.inf, 1, -np.inf, np.inf, 2, -3, -np.inf, 5, -3, -np.inf, -np.inf], + [True, True, False, True], + ), + # Test with nan vals; should always be False + ( + [1, 2, np.nan, 3, 2, np.nan, np.nan, 5, -np.inf, 1, np.nan], + [False, False, False, False], + ), + ], +) +def test_is_monotonic_decreasing(in_vals, out_vals): + # GH 17015 + source_dict = { + "A": ["1", "2", "3", "4", "5", "6", "7", "8", "9", "10", "11"], + "B": ["a", "a", "a", "b", "b", "b", "c", "c", "c", "d", "d"], + "C": in_vals, + } + + df = DataFrame(source_dict) + result = df.groupby("B").C.is_monotonic_decreasing + index = Index(list("abcd"), name="B") + expected = Series(index=index, data=out_vals, name="C") + tm.assert_series_equal(result, expected) diff --git a/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/test_nlargest_nsmallest.py b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/test_nlargest_nsmallest.py new file mode 100644 index 0000000000000000000000000000000000000000..bf983f04a3f3f17566299bafe756e95e2727f6ad --- /dev/null +++ b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/test_nlargest_nsmallest.py @@ -0,0 +1,115 @@ +import numpy as np +import pytest + +from pandas import ( + MultiIndex, + Series, + date_range, +) +import pandas._testing as tm + + +def test_nlargest(): + a = Series([1, 3, 5, 7, 2, 9, 0, 4, 6, 10]) + b = Series(list("a" * 5 + "b" * 5)) + gb = a.groupby(b) + r = gb.nlargest(3) + e = Series( + [7, 5, 3, 10, 9, 6], + index=MultiIndex.from_arrays([list("aaabbb"), [3, 2, 1, 9, 5, 8]]), + ) + tm.assert_series_equal(r, e) + + a = Series([1, 1, 3, 2, 0, 3, 3, 2, 1, 0]) + gb = a.groupby(b) + e = Series( + [3, 2, 1, 3, 3, 2], + index=MultiIndex.from_arrays([list("aaabbb"), [2, 3, 1, 6, 5, 7]]), + ) + tm.assert_series_equal(gb.nlargest(3, keep="last"), e) + + +def test_nlargest_mi_grouper(): + # see gh-21411 + npr = np.random.default_rng(2) + + dts = date_range("20180101", periods=10) + iterables = [dts, ["one", "two"]] + + idx = MultiIndex.from_product(iterables, names=["first", "second"]) + s = Series(npr.standard_normal(20), index=idx) + + result = s.groupby("first").nlargest(1) + + exp_idx = MultiIndex.from_tuples( + [ + (dts[0], dts[0], "one"), + (dts[1], dts[1], "one"), + (dts[2], dts[2], "one"), + (dts[3], dts[3], "two"), + (dts[4], dts[4], "one"), + (dts[5], dts[5], "one"), + (dts[6], dts[6], "one"), + (dts[7], dts[7], "one"), + (dts[8], dts[8], "one"), + (dts[9], dts[9], "one"), + ], + names=["first", "first", "second"], + ) + + exp_values = [ + 0.18905338179353307, + -0.41306354339189344, + 1.799707382720902, + 0.7738065867276614, + 0.28121066979764925, + 0.9775674511260357, + -0.3288239040579627, + 0.45495807124085547, + 0.5452887139646817, + 0.12682784711186987, + ] + + expected = Series(exp_values, index=exp_idx) + tm.assert_series_equal(result, expected, check_exact=False, rtol=1e-3) + + +def test_nsmallest(): + a = Series([1, 3, 5, 7, 2, 9, 0, 4, 6, 10]) + b = Series(list("a" * 5 + "b" * 5)) + gb = a.groupby(b) + r = gb.nsmallest(3) + e = Series( + [1, 2, 3, 0, 4, 6], + index=MultiIndex.from_arrays([list("aaabbb"), [0, 4, 1, 6, 7, 8]]), + ) + tm.assert_series_equal(r, e) + + a = Series([1, 1, 3, 2, 0, 3, 3, 2, 1, 0]) + gb = a.groupby(b) + e = Series( + [0, 1, 1, 0, 1, 2], + index=MultiIndex.from_arrays([list("aaabbb"), [4, 1, 0, 9, 8, 7]]), + ) + tm.assert_series_equal(gb.nsmallest(3, keep="last"), e) + + +@pytest.mark.parametrize( + "data, groups", + [([0, 1, 2, 3], [0, 0, 1, 1]), ([0], [0])], +) +@pytest.mark.parametrize("dtype", [None, *tm.ALL_INT_NUMPY_DTYPES]) +@pytest.mark.parametrize("method", ["nlargest", "nsmallest"]) +def test_nlargest_and_smallest_noop(data, groups, dtype, method): + # GH 15272, GH 16345, GH 29129 + # Test nlargest/smallest when it results in a noop, + # i.e. input is sorted and group size <= n + if dtype is not None: + data = np.array(data, dtype=dtype) + if method == "nlargest": + data = list(reversed(data)) + ser = Series(data, name="a") + result = getattr(ser.groupby(groups), method)(n=2) + expidx = np.array(groups, dtype=int) if isinstance(groups, list) else groups + expected = Series(data, index=MultiIndex.from_arrays([expidx, ser.index]), name="a") + tm.assert_series_equal(result, expected) diff --git a/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/test_nth.py b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/test_nth.py new file mode 100644 index 0000000000000000000000000000000000000000..a8ed9e9d5202173b25b8dc47598e49672e0c8a31 --- /dev/null +++ b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/test_nth.py @@ -0,0 +1,921 @@ +import numpy as np +import pytest + +import pandas as pd +from pandas import ( + DataFrame, + Index, + MultiIndex, + Series, + Timestamp, + isna, +) +import pandas._testing as tm + + +def test_first_last_nth(df): + # tests for first / last / nth + grouped = df.groupby("A") + first = grouped.first() + expected = df.loc[[1, 0], ["B", "C", "D"]] + expected.index = Index(["bar", "foo"], name="A") + expected = expected.sort_index() + tm.assert_frame_equal(first, expected) + + nth = grouped.nth(0) + expected = df.loc[[0, 1]] + tm.assert_frame_equal(nth, expected) + + last = grouped.last() + expected = df.loc[[5, 7], ["B", "C", "D"]] + expected.index = Index(["bar", "foo"], name="A") + tm.assert_frame_equal(last, expected) + + nth = grouped.nth(-1) + expected = df.iloc[[5, 7]] + tm.assert_frame_equal(nth, expected) + + nth = grouped.nth(1) + expected = df.iloc[[2, 3]] + tm.assert_frame_equal(nth, expected) + + # it works! + grouped["B"].first() + grouped["B"].last() + grouped["B"].nth(0) + + df = df.copy() + df.loc[df["A"] == "foo", "B"] = np.nan + grouped = df.groupby("A") + assert isna(grouped["B"].first()["foo"]) + assert isna(grouped["B"].last()["foo"]) + assert isna(grouped["B"].nth(0).iloc[0]) + + # v0.14.0 whatsnew + df = DataFrame([[1, np.nan], [1, 4], [5, 6]], columns=["A", "B"]) + g = df.groupby("A") + result = g.first() + expected = df.iloc[[1, 2]].set_index("A") + tm.assert_frame_equal(result, expected) + + expected = df.iloc[[1, 2]] + result = g.nth(0, dropna="any") + tm.assert_frame_equal(result, expected) + + +@pytest.mark.parametrize("method", ["first", "last"]) +def test_first_last_with_na_object(method, nulls_fixture): + # https://github.com/pandas-dev/pandas/issues/32123 + groups = DataFrame({"a": [1, 1, 2, 2], "b": [1, 2, 3, nulls_fixture]}).groupby("a") + result = getattr(groups, method)() + + if method == "first": + values = [1, 3] + else: + values = [2, 3] + + values = np.array(values, dtype=result["b"].dtype) + idx = Index([1, 2], name="a") + expected = DataFrame({"b": values}, index=idx) + + tm.assert_frame_equal(result, expected) + + +@pytest.mark.parametrize("index", [0, -1]) +def test_nth_with_na_object(index, nulls_fixture): + # https://github.com/pandas-dev/pandas/issues/32123 + df = DataFrame({"a": [1, 1, 2, 2], "b": [1, 2, 3, nulls_fixture]}) + groups = df.groupby("a") + result = groups.nth(index) + expected = df.iloc[[0, 2]] if index == 0 else df.iloc[[1, 3]] + tm.assert_frame_equal(result, expected) + + +@pytest.mark.parametrize("method", ["first", "last"]) +def test_first_last_with_None(method): + # https://github.com/pandas-dev/pandas/issues/32800 + # None should be preserved as object dtype + df = DataFrame.from_dict({"id": ["a"], "value": [None]}) + groups = df.groupby("id", as_index=False) + result = getattr(groups, method)() + + tm.assert_frame_equal(result, df) + + +@pytest.mark.parametrize("method", ["first", "last"]) +@pytest.mark.parametrize( + "df, expected", + [ + ( + DataFrame({"id": "a", "value": [None, "foo", np.nan]}), + DataFrame({"value": ["foo"]}, index=Index(["a"], name="id")), + ), + ( + DataFrame({"id": "a", "value": [np.nan]}, dtype=object), + DataFrame({"value": [None]}, index=Index(["a"], name="id")), + ), + ], +) +def test_first_last_with_None_expanded(method, df, expected): + # GH 32800, 38286 + result = getattr(df.groupby("id"), method)() + tm.assert_frame_equal(result, expected) + + +def test_first_last_nth_dtypes(): + df = DataFrame( + { + "A": ["foo", "bar", "foo", "bar", "foo", "bar", "foo", "foo"], + "B": ["one", "one", "two", "three", "two", "two", "one", "three"], + "C": np.random.default_rng(2).standard_normal(8), + "D": np.array(np.random.default_rng(2).standard_normal(8), dtype="float32"), + } + ) + df["E"] = True + df["F"] = 1 + + # tests for first / last / nth + grouped = df.groupby("A") + first = grouped.first() + expected = df.loc[[1, 0], ["B", "C", "D", "E", "F"]] + expected.index = Index(["bar", "foo"], name="A") + expected = expected.sort_index() + tm.assert_frame_equal(first, expected) + + last = grouped.last() + expected = df.loc[[5, 7], ["B", "C", "D", "E", "F"]] + expected.index = Index(["bar", "foo"], name="A") + expected = expected.sort_index() + tm.assert_frame_equal(last, expected) + + nth = grouped.nth(1) + expected = df.iloc[[2, 3]] + tm.assert_frame_equal(nth, expected) + + +def test_first_last_nth_dtypes2(): + # GH 2763, first/last shifting dtypes + idx = list(range(10)) + idx.append(9) + ser = Series(data=range(11), index=idx, name="IntCol") + assert ser.dtype == "int64" + f = ser.groupby(level=0).first() + assert f.dtype == "int64" + + +def test_first_last_nth_nan_dtype(): + # GH 33591 + df = DataFrame({"data": ["A"], "nans": Series([None], dtype=object)}) + grouped = df.groupby("data") + + expected = df.set_index("data").nans + tm.assert_series_equal(grouped.nans.first(), expected) + tm.assert_series_equal(grouped.nans.last(), expected) + + expected = df.nans + tm.assert_series_equal(grouped.nans.nth(-1), expected) + tm.assert_series_equal(grouped.nans.nth(0), expected) + + +def test_first_strings_timestamps(): + # GH 11244 + test = DataFrame( + { + Timestamp("2012-01-01 00:00:00"): ["a", "b"], + Timestamp("2012-01-02 00:00:00"): ["c", "d"], + "name": ["e", "e"], + "aaaa": ["f", "g"], + } + ) + result = test.groupby("name").first() + expected = DataFrame( + [["a", "c", "f"]], + columns=Index([Timestamp("2012-01-01"), Timestamp("2012-01-02"), "aaaa"]), + index=Index(["e"], name="name"), + ) + tm.assert_frame_equal(result, expected) + + +def test_nth(): + df = DataFrame([[1, np.nan], [1, 4], [5, 6]], columns=["A", "B"]) + gb = df.groupby("A") + + tm.assert_frame_equal(gb.nth(0), df.iloc[[0, 2]]) + tm.assert_frame_equal(gb.nth(1), df.iloc[[1]]) + tm.assert_frame_equal(gb.nth(2), df.loc[[]]) + tm.assert_frame_equal(gb.nth(-1), df.iloc[[1, 2]]) + tm.assert_frame_equal(gb.nth(-2), df.iloc[[0]]) + tm.assert_frame_equal(gb.nth(-3), df.loc[[]]) + tm.assert_series_equal(gb.B.nth(0), df.B.iloc[[0, 2]]) + tm.assert_series_equal(gb.B.nth(1), df.B.iloc[[1]]) + tm.assert_frame_equal(gb[["B"]].nth(0), df[["B"]].iloc[[0, 2]]) + + tm.assert_frame_equal(gb.nth(0, dropna="any"), df.iloc[[1, 2]]) + tm.assert_frame_equal(gb.nth(-1, dropna="any"), df.iloc[[1, 2]]) + + tm.assert_frame_equal(gb.nth(7, dropna="any"), df.iloc[:0]) + tm.assert_frame_equal(gb.nth(2, dropna="any"), df.iloc[:0]) + + +def test_nth2(): + # out of bounds, regression from 0.13.1 + # GH 6621 + df = DataFrame( + { + "color": {0: "green", 1: "green", 2: "red", 3: "red", 4: "red"}, + "food": {0: "ham", 1: "eggs", 2: "eggs", 3: "ham", 4: "pork"}, + "two": { + 0: 1.5456590000000001, + 1: -0.070345000000000005, + 2: -2.4004539999999999, + 3: 0.46206000000000003, + 4: 0.52350799999999997, + }, + "one": { + 0: 0.56573799999999996, + 1: -0.9742360000000001, + 2: 1.033801, + 3: -0.78543499999999999, + 4: 0.70422799999999997, + }, + } + ).set_index(["color", "food"]) + + result = df.groupby(level=0, as_index=False).nth(2) + expected = df.iloc[[-1]] + tm.assert_frame_equal(result, expected) + + result = df.groupby(level=0, as_index=False).nth(3) + expected = df.loc[[]] + tm.assert_frame_equal(result, expected) + + +def test_nth3(): + # GH 7559 + # from the vbench + df = DataFrame(np.random.default_rng(2).integers(1, 10, (100, 2)), dtype="int64") + ser = df[1] + gb = df[0] + expected = ser.groupby(gb).first() + expected2 = ser.groupby(gb).apply(lambda x: x.iloc[0]) + tm.assert_series_equal(expected2, expected, check_names=False) + assert expected.name == 1 + assert expected2.name == 1 + + # validate first + v = ser[gb == 1].iloc[0] + assert expected.iloc[0] == v + assert expected2.iloc[0] == v + + with pytest.raises(ValueError, match="For a DataFrame"): + ser.groupby(gb, sort=False).nth(0, dropna=True) + + +def test_nth4(): + # doc example + df = DataFrame([[1, np.nan], [1, 4], [5, 6]], columns=["A", "B"]) + gb = df.groupby("A") + result = gb.B.nth(0, dropna="all") + expected = df.B.iloc[[1, 2]] + tm.assert_series_equal(result, expected) + + +def test_nth5(): + # test multiple nth values + df = DataFrame([[1, np.nan], [1, 3], [1, 4], [5, 6], [5, 7]], columns=["A", "B"]) + gb = df.groupby("A") + + tm.assert_frame_equal(gb.nth(0), df.iloc[[0, 3]]) + tm.assert_frame_equal(gb.nth([0]), df.iloc[[0, 3]]) + tm.assert_frame_equal(gb.nth([0, 1]), df.iloc[[0, 1, 3, 4]]) + tm.assert_frame_equal(gb.nth([0, -1]), df.iloc[[0, 2, 3, 4]]) + tm.assert_frame_equal(gb.nth([0, 1, 2]), df.iloc[[0, 1, 2, 3, 4]]) + tm.assert_frame_equal(gb.nth([0, 1, -1]), df.iloc[[0, 1, 2, 3, 4]]) + tm.assert_frame_equal(gb.nth([2]), df.iloc[[2]]) + tm.assert_frame_equal(gb.nth([3, 4]), df.loc[[]]) + + +def test_nth_bdays(unit): + business_dates = pd.date_range( + start="4/1/2014", end="6/30/2014", freq="B", unit=unit + ) + df = DataFrame(1, index=business_dates, columns=["a", "b"]) + # get the first, fourth and last two business days for each month + key = [df.index.year, df.index.month] + result = df.groupby(key, as_index=False).nth([0, 3, -2, -1]) + expected_dates = pd.to_datetime( + [ + "2014/4/1", + "2014/4/4", + "2014/4/29", + "2014/4/30", + "2014/5/1", + "2014/5/6", + "2014/5/29", + "2014/5/30", + "2014/6/2", + "2014/6/5", + "2014/6/27", + "2014/6/30", + ] + ).as_unit(unit) + expected = DataFrame(1, columns=["a", "b"], index=expected_dates) + tm.assert_frame_equal(result, expected) + + +def test_nth_multi_grouper(three_group): + # PR 9090, related to issue 8979 + # test nth on multiple groupers + grouped = three_group.groupby(["A", "B"]) + result = grouped.nth(0) + expected = three_group.iloc[[0, 3, 4, 7]] + tm.assert_frame_equal(result, expected) + + +@pytest.mark.parametrize( + "data, expected_first, expected_last", + [ + ( + { + "id": ["A"], + "time": Timestamp("2012-02-01 14:00:00", tz="US/Central"), + "foo": [1], + }, + { + "id": ["A"], + "time": Timestamp("2012-02-01 14:00:00", tz="US/Central"), + "foo": [1], + }, + { + "id": ["A"], + "time": Timestamp("2012-02-01 14:00:00", tz="US/Central"), + "foo": [1], + }, + ), + ( + { + "id": ["A", "B", "A"], + "time": [ + Timestamp("2012-01-01 13:00:00", tz="America/New_York"), + Timestamp("2012-02-01 14:00:00", tz="US/Central"), + Timestamp("2012-03-01 12:00:00", tz="Europe/London"), + ], + "foo": [1, 2, 3], + }, + { + "id": ["A", "B"], + "time": [ + Timestamp("2012-01-01 13:00:00", tz="America/New_York"), + Timestamp("2012-02-01 14:00:00", tz="US/Central"), + ], + "foo": [1, 2], + }, + { + "id": ["A", "B"], + "time": [ + Timestamp("2012-03-01 12:00:00", tz="Europe/London"), + Timestamp("2012-02-01 14:00:00", tz="US/Central"), + ], + "foo": [3, 2], + }, + ), + ], +) +def test_first_last_tz(data, expected_first, expected_last): + # GH15884 + # Test that the timezone is retained when calling first + # or last on groupby with as_index=False + + df = DataFrame(data) + + result = df.groupby("id", as_index=False).first() + expected = DataFrame(expected_first) + cols = ["id", "time", "foo"] + tm.assert_frame_equal(result[cols], expected[cols]) + + result = df.groupby("id", as_index=False)["time"].first() + tm.assert_frame_equal(result, expected[["id", "time"]]) + + result = df.groupby("id", as_index=False).last() + expected = DataFrame(expected_last) + cols = ["id", "time", "foo"] + tm.assert_frame_equal(result[cols], expected[cols]) + + result = df.groupby("id", as_index=False)["time"].last() + tm.assert_frame_equal(result, expected[["id", "time"]]) + + +@pytest.mark.parametrize( + "method, ts, alpha", + [ + ["first", Timestamp("2013-01-01", tz="US/Eastern"), "a"], + ["last", Timestamp("2013-01-02", tz="US/Eastern"), "b"], + ], +) +def test_first_last_tz_multi_column(method, ts, alpha, unit): + # GH 21603 + category_string = Series(list("abc")).astype("category") + dti = pd.date_range("20130101", periods=3, tz="US/Eastern", unit=unit) + df = DataFrame( + { + "group": [1, 1, 2], + "category_string": category_string, + "datetimetz": dti, + } + ) + result = getattr(df.groupby("group"), method)() + expected = DataFrame( + { + "category_string": pd.Categorical( + [alpha, "c"], dtype=category_string.dtype + ), + "datetimetz": [ts, Timestamp("2013-01-03", tz="US/Eastern")], + }, + index=Index([1, 2], name="group"), + ) + expected["datetimetz"] = expected["datetimetz"].dt.as_unit(unit) + tm.assert_frame_equal(result, expected) + + +@pytest.mark.parametrize( + "values", + [ + pd.array([True, False], dtype="boolean"), + pd.array([1, 2], dtype="Int64"), + pd.to_datetime(["2020-01-01", "2020-02-01"]), + pd.to_timedelta([1, 2], unit="D"), + ], +) +@pytest.mark.parametrize("function", ["first", "last", "min", "max"]) +def test_first_last_extension_array_keeps_dtype(values, function): + # https://github.com/pandas-dev/pandas/issues/33071 + # https://github.com/pandas-dev/pandas/issues/32194 + df = DataFrame({"a": [1, 2], "b": values}) + grouped = df.groupby("a") + idx = Index([1, 2], name="a") + expected_series = Series(values, name="b", index=idx) + expected_frame = DataFrame({"b": values}, index=idx) + + result_series = getattr(grouped["b"], function)() + tm.assert_series_equal(result_series, expected_series) + + result_frame = grouped.agg({"b": function}) + tm.assert_frame_equal(result_frame, expected_frame) + + +def test_nth_multi_index_as_expected(): + # PR 9090, related to issue 8979 + # test nth on MultiIndex + three_group = DataFrame( + { + "A": [ + "foo", + "foo", + "foo", + "foo", + "bar", + "bar", + "bar", + "bar", + "foo", + "foo", + "foo", + ], + "B": [ + "one", + "one", + "one", + "two", + "one", + "one", + "one", + "two", + "two", + "two", + "one", + ], + "C": [ + "dull", + "dull", + "shiny", + "dull", + "dull", + "shiny", + "shiny", + "dull", + "shiny", + "shiny", + "shiny", + ], + } + ) + grouped = three_group.groupby(["A", "B"]) + result = grouped.nth(0) + expected = three_group.iloc[[0, 3, 4, 7]] + tm.assert_frame_equal(result, expected) + + +@pytest.mark.parametrize( + "op, n, expected_rows", + [ + ("head", -1, [0]), + ("head", 0, []), + ("head", 1, [0, 2]), + ("head", 7, [0, 1, 2]), + ("tail", -1, [1]), + ("tail", 0, []), + ("tail", 1, [1, 2]), + ("tail", 7, [0, 1, 2]), + ], +) +@pytest.mark.parametrize("columns", [None, [], ["A"], ["B"], ["A", "B"]]) +@pytest.mark.parametrize("as_index", [True, False]) +def test_groupby_head_tail(op, n, expected_rows, columns, as_index): + df = DataFrame([[1, 2], [1, 4], [5, 6]], columns=["A", "B"]) + g = df.groupby("A", as_index=as_index) + expected = df.iloc[expected_rows] + if columns is not None: + g = g[columns] + expected = expected[columns] + result = getattr(g, op)(n) + tm.assert_frame_equal(result, expected) + + +@pytest.mark.parametrize( + "op, n, expected_cols", + [ + ("head", -1, [0]), + ("head", 0, []), + ("head", 1, [0, 2]), + ("head", 7, [0, 1, 2]), + ("tail", -1, [1]), + ("tail", 0, []), + ("tail", 1, [1, 2]), + ("tail", 7, [0, 1, 2]), + ], +) +def test_groupby_head_tail_axis_1(op, n, expected_cols): + # GH 9772 + df = DataFrame( + [[1, 2, 3], [1, 4, 5], [2, 6, 7], [3, 8, 9]], columns=["A", "B", "C"] + ) + msg = "DataFrame.groupby with axis=1 is deprecated" + with tm.assert_produces_warning(FutureWarning, match=msg): + g = df.groupby([0, 0, 1], axis=1) + expected = df.iloc[:, expected_cols] + result = getattr(g, op)(n) + tm.assert_frame_equal(result, expected) + + +def test_group_selection_cache(): + # GH 12839 nth, head, and tail should return same result consistently + df = DataFrame([[1, 2], [1, 4], [5, 6]], columns=["A", "B"]) + expected = df.iloc[[0, 2]] + + g = df.groupby("A") + result1 = g.head(n=2) + result2 = g.nth(0) + tm.assert_frame_equal(result1, df) + tm.assert_frame_equal(result2, expected) + + g = df.groupby("A") + result1 = g.tail(n=2) + result2 = g.nth(0) + tm.assert_frame_equal(result1, df) + tm.assert_frame_equal(result2, expected) + + g = df.groupby("A") + result1 = g.nth(0) + result2 = g.head(n=2) + tm.assert_frame_equal(result1, expected) + tm.assert_frame_equal(result2, df) + + g = df.groupby("A") + result1 = g.nth(0) + result2 = g.tail(n=2) + tm.assert_frame_equal(result1, expected) + tm.assert_frame_equal(result2, df) + + +def test_nth_empty(): + # GH 16064 + df = DataFrame(index=[0], columns=["a", "b", "c"]) + result = df.groupby("a").nth(10) + expected = df.iloc[:0] + tm.assert_frame_equal(result, expected) + + result = df.groupby(["a", "b"]).nth(10) + expected = df.iloc[:0] + tm.assert_frame_equal(result, expected) + + +def test_nth_column_order(): + # GH 20760 + # Check that nth preserves column order + df = DataFrame( + [[1, "b", 100], [1, "a", 50], [1, "a", np.nan], [2, "c", 200], [2, "d", 150]], + columns=["A", "C", "B"], + ) + result = df.groupby("A").nth(0) + expected = df.iloc[[0, 3]] + tm.assert_frame_equal(result, expected) + + result = df.groupby("A").nth(-1, dropna="any") + expected = df.iloc[[1, 4]] + tm.assert_frame_equal(result, expected) + + +@pytest.mark.parametrize("dropna", [None, "any", "all"]) +def test_nth_nan_in_grouper(dropna): + # GH 26011 + df = DataFrame( + { + "a": [np.nan, "a", np.nan, "b", np.nan], + "b": [0, 2, 4, 6, 8], + "c": [1, 3, 5, 7, 9], + } + ) + result = df.groupby("a").nth(0, dropna=dropna) + expected = df.iloc[[1, 3]] + + tm.assert_frame_equal(result, expected) + + +@pytest.mark.parametrize("dropna", [None, "any", "all"]) +def test_nth_nan_in_grouper_series(dropna): + # GH 26454 + df = DataFrame( + { + "a": [np.nan, "a", np.nan, "b", np.nan], + "b": [0, 2, 4, 6, 8], + } + ) + result = df.groupby("a")["b"].nth(0, dropna=dropna) + expected = df["b"].iloc[[1, 3]] + + tm.assert_series_equal(result, expected) + + +def test_first_categorical_and_datetime_data_nat(): + # GH 20520 + df = DataFrame( + { + "group": ["first", "first", "second", "third", "third"], + "time": 5 * [np.datetime64("NaT")], + "categories": Series(["a", "b", "c", "a", "b"], dtype="category"), + } + ) + result = df.groupby("group").first() + expected = DataFrame( + { + "time": 3 * [np.datetime64("NaT")], + "categories": Series(["a", "c", "a"]).astype( + pd.CategoricalDtype(["a", "b", "c"]) + ), + } + ) + expected.index = Index(["first", "second", "third"], name="group") + tm.assert_frame_equal(result, expected) + + +def test_first_multi_key_groupby_categorical(): + # GH 22512 + df = DataFrame( + { + "A": [1, 1, 1, 2, 2], + "B": [100, 100, 200, 100, 100], + "C": ["apple", "orange", "mango", "mango", "orange"], + "D": ["jupiter", "mercury", "mars", "venus", "venus"], + } + ) + df = df.astype({"D": "category"}) + result = df.groupby(by=["A", "B"]).first() + expected = DataFrame( + { + "C": ["apple", "mango", "mango"], + "D": Series(["jupiter", "mars", "venus"]).astype( + pd.CategoricalDtype(["jupiter", "mars", "mercury", "venus"]) + ), + } + ) + expected.index = MultiIndex.from_tuples( + [(1, 100), (1, 200), (2, 100)], names=["A", "B"] + ) + tm.assert_frame_equal(result, expected) + + +@pytest.mark.parametrize("method", ["first", "last", "nth"]) +def test_groupby_last_first_nth_with_none(method, nulls_fixture): + # GH29645 + expected = Series(["y"]) + data = Series( + [nulls_fixture, nulls_fixture, nulls_fixture, "y", nulls_fixture], + index=[0, 0, 0, 0, 0], + ).groupby(level=0) + + if method == "nth": + result = getattr(data, method)(3) + else: + result = getattr(data, method)() + + tm.assert_series_equal(result, expected) + + +@pytest.mark.parametrize( + "arg, expected_rows", + [ + [slice(None, 3, 2), [0, 1, 4, 5]], + [slice(None, -2), [0, 2, 5]], + [[slice(None, 2), slice(-2, None)], [0, 1, 2, 3, 4, 6, 7]], + [[0, 1, slice(-2, None)], [0, 1, 2, 3, 4, 6, 7]], + ], +) +def test_slice(slice_test_df, slice_test_grouped, arg, expected_rows): + # Test slices GH #42947 + + result = slice_test_grouped.nth[arg] + equivalent = slice_test_grouped.nth(arg) + expected = slice_test_df.iloc[expected_rows] + + tm.assert_frame_equal(result, expected) + tm.assert_frame_equal(equivalent, expected) + + +def test_nth_indexed(slice_test_df, slice_test_grouped): + # Test index notation GH #44688 + + result = slice_test_grouped.nth[0, 1, -2:] + equivalent = slice_test_grouped.nth([0, 1, slice(-2, None)]) + expected = slice_test_df.iloc[[0, 1, 2, 3, 4, 6, 7]] + + tm.assert_frame_equal(result, expected) + tm.assert_frame_equal(equivalent, expected) + + +def test_invalid_argument(slice_test_grouped): + # Test for error on invalid argument + + with pytest.raises(TypeError, match="Invalid index"): + slice_test_grouped.nth(3.14) + + +def test_negative_step(slice_test_grouped): + # Test for error on negative slice step + + with pytest.raises(ValueError, match="Invalid step"): + slice_test_grouped.nth(slice(None, None, -1)) + + +def test_np_ints(slice_test_df, slice_test_grouped): + # Test np ints work + + result = slice_test_grouped.nth(np.array([0, 1])) + expected = slice_test_df.iloc[[0, 1, 2, 3, 4]] + tm.assert_frame_equal(result, expected) + + +def test_groupby_nth_with_column_axis(): + # GH43926 + df = DataFrame( + [ + [4, 5, 6], + [8, 8, 7], + ], + index=["z", "y"], + columns=["C", "B", "A"], + ) + msg = "DataFrame.groupby with axis=1 is deprecated" + with tm.assert_produces_warning(FutureWarning, match=msg): + gb = df.groupby(df.iloc[1], axis=1) + result = gb.nth(0) + expected = df.iloc[:, [0, 2]] + tm.assert_frame_equal(result, expected) + + +def test_groupby_nth_interval(): + # GH#24205 + idx_result = MultiIndex( + [ + pd.CategoricalIndex([pd.Interval(0, 1), pd.Interval(1, 2)]), + pd.CategoricalIndex([pd.Interval(0, 10), pd.Interval(10, 20)]), + ], + [[0, 0, 0, 1, 1], [0, 1, 1, 0, -1]], + ) + df_result = DataFrame({"col": range(len(idx_result))}, index=idx_result) + result = df_result.groupby(level=[0, 1], observed=False).nth(0) + val_expected = [0, 1, 3] + idx_expected = MultiIndex( + [ + pd.CategoricalIndex([pd.Interval(0, 1), pd.Interval(1, 2)]), + pd.CategoricalIndex([pd.Interval(0, 10), pd.Interval(10, 20)]), + ], + [[0, 0, 1], [0, 1, 0]], + ) + expected = DataFrame(val_expected, index=idx_expected, columns=["col"]) + tm.assert_frame_equal(result, expected) + + +@pytest.mark.parametrize( + "start, stop, expected_values, expected_columns", + [ + (None, None, [0, 1, 2, 3, 4], list("ABCDE")), + (None, 1, [0, 3], list("AD")), + (None, 9, [0, 1, 2, 3, 4], list("ABCDE")), + (None, -1, [0, 1, 3], list("ABD")), + (1, None, [1, 2, 4], list("BCE")), + (1, -1, [1], list("B")), + (-1, None, [2, 4], list("CE")), + (-1, 2, [4], list("E")), + ], +) +@pytest.mark.parametrize("method", ["call", "index"]) +def test_nth_slices_with_column_axis( + start, stop, expected_values, expected_columns, method +): + df = DataFrame([range(5)], columns=[list("ABCDE")]) + msg = "DataFrame.groupby with axis=1 is deprecated" + with tm.assert_produces_warning(FutureWarning, match=msg): + gb = df.groupby([5, 5, 5, 6, 6], axis=1) + result = { + "call": lambda start, stop: gb.nth(slice(start, stop)), + "index": lambda start, stop: gb.nth[start:stop], + }[method](start, stop) + expected = DataFrame([expected_values], columns=[expected_columns]) + tm.assert_frame_equal(result, expected) + + +@pytest.mark.filterwarnings( + "ignore:invalid value encountered in remainder:RuntimeWarning" +) +def test_head_tail_dropna_true(): + # GH#45089 + df = DataFrame( + [["a", "z"], ["b", np.nan], ["c", np.nan], ["c", np.nan]], columns=["X", "Y"] + ) + expected = DataFrame([["a", "z"]], columns=["X", "Y"]) + + result = df.groupby(["X", "Y"]).head(n=1) + tm.assert_frame_equal(result, expected) + + result = df.groupby(["X", "Y"]).tail(n=1) + tm.assert_frame_equal(result, expected) + + result = df.groupby(["X", "Y"]).nth(n=0) + tm.assert_frame_equal(result, expected) + + +def test_head_tail_dropna_false(): + # GH#45089 + df = DataFrame([["a", "z"], ["b", np.nan], ["c", np.nan]], columns=["X", "Y"]) + expected = DataFrame([["a", "z"], ["b", np.nan], ["c", np.nan]], columns=["X", "Y"]) + + result = df.groupby(["X", "Y"], dropna=False).head(n=1) + tm.assert_frame_equal(result, expected) + + result = df.groupby(["X", "Y"], dropna=False).tail(n=1) + tm.assert_frame_equal(result, expected) + + result = df.groupby(["X", "Y"], dropna=False).nth(n=0) + tm.assert_frame_equal(result, expected) + + +@pytest.mark.parametrize("selection", ("b", ["b"], ["b", "c"])) +@pytest.mark.parametrize("dropna", ["any", "all", None]) +def test_nth_after_selection(selection, dropna): + # GH#11038, GH#53518 + df = DataFrame( + { + "a": [1, 1, 2], + "b": [np.nan, 3, 4], + "c": [5, 6, 7], + } + ) + gb = df.groupby("a")[selection] + result = gb.nth(0, dropna=dropna) + if dropna == "any" or (dropna == "all" and selection != ["b", "c"]): + locs = [1, 2] + else: + locs = [0, 2] + expected = df.loc[locs, selection] + tm.assert_equal(result, expected) + + +@pytest.mark.parametrize( + "data", + [ + ( + Timestamp("2011-01-15 12:50:28.502376"), + Timestamp("2011-01-20 12:50:28.593448"), + ), + (24650000000000001, 24650000000000002), + ], +) +def test_groupby_nth_int_like_precision(data): + # GH#6620, GH#9311 + df = DataFrame({"a": [1, 1], "b": data}) + + grouped = df.groupby("a") + result = grouped.nth(0) + expected = DataFrame({"a": 1, "b": [data[0]]}) + + tm.assert_frame_equal(result, expected) diff --git a/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/test_quantile.py b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/test_quantile.py new file mode 100644 index 0000000000000000000000000000000000000000..361a8c27fbf9d6744a11d56cf228f06a53d4adaf --- /dev/null +++ b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/test_quantile.py @@ -0,0 +1,496 @@ +import numpy as np +import pytest + +import pandas as pd +from pandas import ( + DataFrame, + Index, +) +import pandas._testing as tm + + +@pytest.mark.parametrize( + "interpolation", ["linear", "lower", "higher", "nearest", "midpoint"] +) +@pytest.mark.parametrize( + "a_vals,b_vals", + [ + # Ints + ([1, 2, 3, 4, 5], [5, 4, 3, 2, 1]), + ([1, 2, 3, 4], [4, 3, 2, 1]), + ([1, 2, 3, 4, 5], [4, 3, 2, 1]), + # Floats + ([1.0, 2.0, 3.0, 4.0, 5.0], [5.0, 4.0, 3.0, 2.0, 1.0]), + # Missing data + ([1.0, np.nan, 3.0, np.nan, 5.0], [5.0, np.nan, 3.0, np.nan, 1.0]), + ([np.nan, 4.0, np.nan, 2.0, np.nan], [np.nan, 4.0, np.nan, 2.0, np.nan]), + # Timestamps + ( + pd.date_range("1/1/18", freq="D", periods=5), + pd.date_range("1/1/18", freq="D", periods=5)[::-1], + ), + ( + pd.date_range("1/1/18", freq="D", periods=5).as_unit("s"), + pd.date_range("1/1/18", freq="D", periods=5)[::-1].as_unit("s"), + ), + # All NA + ([np.nan] * 5, [np.nan] * 5), + ], +) +@pytest.mark.parametrize("q", [0, 0.25, 0.5, 0.75, 1]) +def test_quantile(interpolation, a_vals, b_vals, q, request): + if ( + interpolation == "nearest" + and q == 0.5 + and isinstance(b_vals, list) + and b_vals == [4, 3, 2, 1] + ): + request.applymarker( + pytest.mark.xfail( + reason="Unclear numpy expectation for nearest " + "result with equidistant data" + ) + ) + all_vals = pd.concat([pd.Series(a_vals), pd.Series(b_vals)]) + + a_expected = pd.Series(a_vals).quantile(q, interpolation=interpolation) + b_expected = pd.Series(b_vals).quantile(q, interpolation=interpolation) + + df = DataFrame({"key": ["a"] * len(a_vals) + ["b"] * len(b_vals), "val": all_vals}) + + expected = DataFrame( + [a_expected, b_expected], columns=["val"], index=Index(["a", "b"], name="key") + ) + if all_vals.dtype.kind == "M" and expected.dtypes.values[0].kind == "M": + # TODO(non-nano): this should be unnecessary once array_to_datetime + # correctly infers non-nano from Timestamp.unit + expected = expected.astype(all_vals.dtype) + result = df.groupby("key").quantile(q, interpolation=interpolation) + + tm.assert_frame_equal(result, expected) + + +def test_quantile_array(): + # https://github.com/pandas-dev/pandas/issues/27526 + df = DataFrame({"A": [0, 1, 2, 3, 4]}) + key = np.array([0, 0, 1, 1, 1], dtype=np.int64) + result = df.groupby(key).quantile([0.25]) + + index = pd.MultiIndex.from_product([[0, 1], [0.25]]) + expected = DataFrame({"A": [0.25, 2.50]}, index=index) + tm.assert_frame_equal(result, expected) + + df = DataFrame({"A": [0, 1, 2, 3], "B": [4, 5, 6, 7]}) + index = pd.MultiIndex.from_product([[0, 1], [0.25, 0.75]]) + + key = np.array([0, 0, 1, 1], dtype=np.int64) + result = df.groupby(key).quantile([0.25, 0.75]) + expected = DataFrame( + {"A": [0.25, 0.75, 2.25, 2.75], "B": [4.25, 4.75, 6.25, 6.75]}, index=index + ) + tm.assert_frame_equal(result, expected) + + +def test_quantile_array2(): + # https://github.com/pandas-dev/pandas/pull/28085#issuecomment-524066959 + arr = np.random.default_rng(2).integers(0, 5, size=(10, 3), dtype=np.int64) + df = DataFrame(arr, columns=list("ABC")) + result = df.groupby("A").quantile([0.3, 0.7]) + expected = DataFrame( + { + "B": [2.0, 2.0, 2.3, 2.7, 0.3, 0.7, 3.2, 4.0, 0.3, 0.7], + "C": [1.0, 1.0, 1.9, 3.0999999999999996, 0.3, 0.7, 2.6, 3.0, 1.2, 2.8], + }, + index=pd.MultiIndex.from_product( + [[0, 1, 2, 3, 4], [0.3, 0.7]], names=["A", None] + ), + ) + tm.assert_frame_equal(result, expected) + + +def test_quantile_array_no_sort(): + df = DataFrame({"A": [0, 1, 2], "B": [3, 4, 5]}) + key = np.array([1, 0, 1], dtype=np.int64) + result = df.groupby(key, sort=False).quantile([0.25, 0.5, 0.75]) + expected = DataFrame( + {"A": [0.5, 1.0, 1.5, 1.0, 1.0, 1.0], "B": [3.5, 4.0, 4.5, 4.0, 4.0, 4.0]}, + index=pd.MultiIndex.from_product([[1, 0], [0.25, 0.5, 0.75]]), + ) + tm.assert_frame_equal(result, expected) + + result = df.groupby(key, sort=False).quantile([0.75, 0.25]) + expected = DataFrame( + {"A": [1.5, 0.5, 1.0, 1.0], "B": [4.5, 3.5, 4.0, 4.0]}, + index=pd.MultiIndex.from_product([[1, 0], [0.75, 0.25]]), + ) + tm.assert_frame_equal(result, expected) + + +def test_quantile_array_multiple_levels(): + df = DataFrame( + {"A": [0, 1, 2], "B": [3, 4, 5], "c": ["a", "a", "a"], "d": ["a", "a", "b"]} + ) + result = df.groupby(["c", "d"]).quantile([0.25, 0.75]) + index = pd.MultiIndex.from_tuples( + [("a", "a", 0.25), ("a", "a", 0.75), ("a", "b", 0.25), ("a", "b", 0.75)], + names=["c", "d", None], + ) + expected = DataFrame( + {"A": [0.25, 0.75, 2.0, 2.0], "B": [3.25, 3.75, 5.0, 5.0]}, index=index + ) + tm.assert_frame_equal(result, expected) + + +@pytest.mark.parametrize("frame_size", [(2, 3), (100, 10)]) +@pytest.mark.parametrize("groupby", [[0], [0, 1]]) +@pytest.mark.parametrize("q", [[0.5, 0.6]]) +def test_groupby_quantile_with_arraylike_q_and_int_columns(frame_size, groupby, q): + # GH30289 + nrow, ncol = frame_size + df = DataFrame(np.array([ncol * [_ % 4] for _ in range(nrow)]), columns=range(ncol)) + + idx_levels = [np.arange(min(nrow, 4))] * len(groupby) + [q] + idx_codes = [[x for x in range(min(nrow, 4)) for _ in q]] * len(groupby) + [ + list(range(len(q))) * min(nrow, 4) + ] + expected_index = pd.MultiIndex( + levels=idx_levels, codes=idx_codes, names=groupby + [None] + ) + expected_values = [ + [float(x)] * (ncol - len(groupby)) for x in range(min(nrow, 4)) for _ in q + ] + expected_columns = [x for x in range(ncol) if x not in groupby] + expected = DataFrame( + expected_values, index=expected_index, columns=expected_columns + ) + result = df.groupby(groupby).quantile(q) + + tm.assert_frame_equal(result, expected) + + +def test_quantile_raises(): + df = DataFrame([["foo", "a"], ["foo", "b"], ["foo", "c"]], columns=["key", "val"]) + + with pytest.raises(TypeError, match="cannot be performed against 'object' dtypes"): + df.groupby("key").quantile() + + +def test_quantile_out_of_bounds_q_raises(): + # https://github.com/pandas-dev/pandas/issues/27470 + df = DataFrame({"a": [0, 0, 0, 1, 1, 1], "b": range(6)}) + g = df.groupby([0, 0, 0, 1, 1, 1]) + with pytest.raises(ValueError, match="Got '50.0' instead"): + g.quantile(50) + + with pytest.raises(ValueError, match="Got '-1.0' instead"): + g.quantile(-1) + + +def test_quantile_missing_group_values_no_segfaults(): + # GH 28662 + data = np.array([1.0, np.nan, 1.0]) + df = DataFrame({"key": data, "val": range(3)}) + + # Random segfaults; would have been guaranteed in loop + grp = df.groupby("key") + for _ in range(100): + grp.quantile() + + +@pytest.mark.parametrize( + "key, val, expected_key, expected_val", + [ + ([1.0, np.nan, 3.0, np.nan], range(4), [1.0, 3.0], [0.0, 2.0]), + ([1.0, np.nan, 2.0, 2.0], range(4), [1.0, 2.0], [0.0, 2.5]), + (["a", "b", "b", np.nan], range(4), ["a", "b"], [0, 1.5]), + ([0], [42], [0], [42.0]), + ([], [], np.array([], dtype="float64"), np.array([], dtype="float64")), + ], +) +def test_quantile_missing_group_values_correct_results( + key, val, expected_key, expected_val +): + # GH 28662, GH 33200, GH 33569 + df = DataFrame({"key": key, "val": val}) + + expected = DataFrame( + expected_val, index=Index(expected_key, name="key"), columns=["val"] + ) + + grp = df.groupby("key") + + result = grp.quantile(0.5) + tm.assert_frame_equal(result, expected) + + result = grp.quantile() + tm.assert_frame_equal(result, expected) + + +@pytest.mark.parametrize( + "values", + [ + pd.array([1, 0, None] * 2, dtype="Int64"), + pd.array([True, False, None] * 2, dtype="boolean"), + ], +) +@pytest.mark.parametrize("q", [0.5, [0.0, 0.5, 1.0]]) +def test_groupby_quantile_nullable_array(values, q): + # https://github.com/pandas-dev/pandas/issues/33136 + df = DataFrame({"a": ["x"] * 3 + ["y"] * 3, "b": values}) + result = df.groupby("a")["b"].quantile(q) + + if isinstance(q, list): + idx = pd.MultiIndex.from_product((["x", "y"], q), names=["a", None]) + true_quantiles = [0.0, 0.5, 1.0] + else: + idx = Index(["x", "y"], name="a") + true_quantiles = [0.5] + + expected = pd.Series(true_quantiles * 2, index=idx, name="b", dtype="Float64") + tm.assert_series_equal(result, expected) + + +@pytest.mark.parametrize("q", [0.5, [0.0, 0.5, 1.0]]) +@pytest.mark.parametrize("numeric_only", [True, False]) +def test_groupby_quantile_raises_on_invalid_dtype(q, numeric_only): + df = DataFrame({"a": [1], "b": [2.0], "c": ["x"]}) + if numeric_only: + result = df.groupby("a").quantile(q, numeric_only=numeric_only) + expected = df.groupby("a")[["b"]].quantile(q) + tm.assert_frame_equal(result, expected) + else: + with pytest.raises( + TypeError, match="'quantile' cannot be performed against 'object' dtypes!" + ): + df.groupby("a").quantile(q, numeric_only=numeric_only) + + +def test_groupby_quantile_NA_float(any_float_dtype): + # GH#42849 + df = DataFrame({"x": [1, 1], "y": [0.2, np.nan]}, dtype=any_float_dtype) + result = df.groupby("x")["y"].quantile(0.5) + exp_index = Index([1.0], dtype=any_float_dtype, name="x") + + if any_float_dtype in ["Float32", "Float64"]: + expected_dtype = any_float_dtype + else: + expected_dtype = None + + expected = pd.Series([0.2], dtype=expected_dtype, index=exp_index, name="y") + tm.assert_series_equal(result, expected) + + result = df.groupby("x")["y"].quantile([0.5, 0.75]) + expected = pd.Series( + [0.2] * 2, + index=pd.MultiIndex.from_product((exp_index, [0.5, 0.75]), names=["x", None]), + name="y", + dtype=expected_dtype, + ) + tm.assert_series_equal(result, expected) + + +def test_groupby_quantile_NA_int(any_int_ea_dtype): + # GH#42849 + df = DataFrame({"x": [1, 1], "y": [2, 5]}, dtype=any_int_ea_dtype) + result = df.groupby("x")["y"].quantile(0.5) + expected = pd.Series( + [3.5], + dtype="Float64", + index=Index([1], name="x", dtype=any_int_ea_dtype), + name="y", + ) + tm.assert_series_equal(expected, result) + + result = df.groupby("x").quantile(0.5) + expected = DataFrame( + {"y": 3.5}, dtype="Float64", index=Index([1], name="x", dtype=any_int_ea_dtype) + ) + tm.assert_frame_equal(result, expected) + + +@pytest.mark.parametrize( + "interpolation, val1, val2", [("lower", 2, 2), ("higher", 2, 3), ("nearest", 2, 2)] +) +def test_groupby_quantile_all_na_group_masked( + interpolation, val1, val2, any_numeric_ea_dtype +): + # GH#37493 + df = DataFrame( + {"a": [1, 1, 1, 2], "b": [1, 2, 3, pd.NA]}, dtype=any_numeric_ea_dtype + ) + result = df.groupby("a").quantile(q=[0.5, 0.7], interpolation=interpolation) + expected = DataFrame( + {"b": [val1, val2, pd.NA, pd.NA]}, + dtype=any_numeric_ea_dtype, + index=pd.MultiIndex.from_arrays( + [pd.Series([1, 1, 2, 2], dtype=any_numeric_ea_dtype), [0.5, 0.7, 0.5, 0.7]], + names=["a", None], + ), + ) + tm.assert_frame_equal(result, expected) + + +@pytest.mark.parametrize("interpolation", ["midpoint", "linear"]) +def test_groupby_quantile_all_na_group_masked_interp( + interpolation, any_numeric_ea_dtype +): + # GH#37493 + df = DataFrame( + {"a": [1, 1, 1, 2], "b": [1, 2, 3, pd.NA]}, dtype=any_numeric_ea_dtype + ) + result = df.groupby("a").quantile(q=[0.5, 0.75], interpolation=interpolation) + + if any_numeric_ea_dtype == "Float32": + expected_dtype = any_numeric_ea_dtype + else: + expected_dtype = "Float64" + + expected = DataFrame( + {"b": [2.0, 2.5, pd.NA, pd.NA]}, + dtype=expected_dtype, + index=pd.MultiIndex.from_arrays( + [ + pd.Series([1, 1, 2, 2], dtype=any_numeric_ea_dtype), + [0.5, 0.75, 0.5, 0.75], + ], + names=["a", None], + ), + ) + tm.assert_frame_equal(result, expected) + + +@pytest.mark.parametrize("dtype", ["Float64", "Float32"]) +def test_groupby_quantile_allNA_column(dtype): + # GH#42849 + df = DataFrame({"x": [1, 1], "y": [pd.NA] * 2}, dtype=dtype) + result = df.groupby("x")["y"].quantile(0.5) + expected = pd.Series( + [np.nan], dtype=dtype, index=Index([1.0], dtype=dtype), name="y" + ) + expected.index.name = "x" + tm.assert_series_equal(expected, result) + + +def test_groupby_timedelta_quantile(): + # GH: 29485 + df = DataFrame( + {"value": pd.to_timedelta(np.arange(4), unit="s"), "group": [1, 1, 2, 2]} + ) + result = df.groupby("group").quantile(0.99) + expected = DataFrame( + { + "value": [ + pd.Timedelta("0 days 00:00:00.990000"), + pd.Timedelta("0 days 00:00:02.990000"), + ] + }, + index=Index([1, 2], name="group"), + ) + tm.assert_frame_equal(result, expected) + + +def test_columns_groupby_quantile(): + # GH 33795 + df = DataFrame( + np.arange(12).reshape(3, -1), + index=list("XYZ"), + columns=pd.Series(list("ABAB"), name="col"), + ) + msg = "DataFrame.groupby with axis=1 is deprecated" + with tm.assert_produces_warning(FutureWarning, match=msg): + gb = df.groupby("col", axis=1) + result = gb.quantile(q=[0.8, 0.2]) + expected = DataFrame( + [ + [1.6, 0.4, 2.6, 1.4], + [5.6, 4.4, 6.6, 5.4], + [9.6, 8.4, 10.6, 9.4], + ], + index=list("XYZ"), + columns=pd.MultiIndex.from_tuples( + [("A", 0.8), ("A", 0.2), ("B", 0.8), ("B", 0.2)], names=["col", None] + ), + ) + + tm.assert_frame_equal(result, expected) + + +def test_timestamp_groupby_quantile(unit): + # GH 33168 + dti = pd.date_range( + start="2020-04-19 00:00:00", freq="1min", periods=100, tz="UTC", unit=unit + ).floor("1h") + df = DataFrame( + { + "timestamp": dti, + "category": list(range(1, 101)), + "value": list(range(101, 201)), + } + ) + + result = df.groupby("timestamp").quantile([0.2, 0.8]) + + mi = pd.MultiIndex.from_product([dti[::99], [0.2, 0.8]], names=("timestamp", None)) + expected = DataFrame( + [ + {"category": 12.8, "value": 112.8}, + {"category": 48.2, "value": 148.2}, + {"category": 68.8, "value": 168.8}, + {"category": 92.2, "value": 192.2}, + ], + index=mi, + ) + + tm.assert_frame_equal(result, expected) + + +def test_groupby_quantile_dt64tz_period(): + # GH#51373 + dti = pd.date_range("2016-01-01", periods=1000) + df = pd.Series(dti).to_frame().copy() + df[1] = dti.tz_localize("US/Pacific") + df[2] = dti.to_period("D") + df[3] = dti - dti[0] + df.iloc[-1] = pd.NaT + + by = np.tile(np.arange(5), 200) + gb = df.groupby(by) + + result = gb.quantile(0.5) + + # Check that we match the group-by-group result + exp = {i: df.iloc[i::5].quantile(0.5) for i in range(5)} + expected = DataFrame(exp).T.infer_objects() + expected.index = expected.index.astype(int) + + tm.assert_frame_equal(result, expected) + + +def test_groupby_quantile_nonmulti_levels_order(): + # Non-regression test for GH #53009 + ind = pd.MultiIndex.from_tuples( + [ + (0, "a", "B"), + (0, "a", "A"), + (0, "b", "B"), + (0, "b", "A"), + (1, "a", "B"), + (1, "a", "A"), + (1, "b", "B"), + (1, "b", "A"), + ], + names=["sample", "cat0", "cat1"], + ) + ser = pd.Series(range(8), index=ind) + result = ser.groupby(level="cat1", sort=False).quantile([0.2, 0.8]) + + qind = pd.MultiIndex.from_tuples( + [("B", 0.2), ("B", 0.8), ("A", 0.2), ("A", 0.8)], names=["cat1", None] + ) + expected = pd.Series([1.2, 4.8, 2.2, 5.8], index=qind) + + tm.assert_series_equal(result, expected) + + # We need to check that index levels are not sorted + expected_levels = pd.core.indexes.frozen.FrozenList([["B", "A"], [0.2, 0.8]]) + tm.assert_equal(result.index.levels, expected_levels) diff --git a/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/test_rank.py b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/test_rank.py new file mode 100644 index 0000000000000000000000000000000000000000..a3b7da3fa836c955d8d0e4e17754d7834e5c05f1 --- /dev/null +++ b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/test_rank.py @@ -0,0 +1,721 @@ +from datetime import datetime + +import numpy as np +import pytest + +import pandas as pd +from pandas import ( + DataFrame, + NaT, + Series, + concat, +) +import pandas._testing as tm + + +def test_rank_unordered_categorical_typeerror(): + # GH#51034 should be TypeError, not NotImplementedError + cat = pd.Categorical([], ordered=False) + ser = Series(cat) + df = ser.to_frame() + + msg = "Cannot perform rank with non-ordered Categorical" + + gb = ser.groupby(cat, observed=False) + with pytest.raises(TypeError, match=msg): + gb.rank() + + gb2 = df.groupby(cat, observed=False) + with pytest.raises(TypeError, match=msg): + gb2.rank() + + +def test_rank_apply(): + lev1 = np.array(["a" * 10] * 100, dtype=object) + lev2 = np.array(["b" * 10] * 130, dtype=object) + lab1 = np.random.default_rng(2).integers(0, 100, size=500, dtype=int) + lab2 = np.random.default_rng(2).integers(0, 130, size=500, dtype=int) + + df = DataFrame( + { + "value": np.random.default_rng(2).standard_normal(500), + "key1": lev1.take(lab1), + "key2": lev2.take(lab2), + } + ) + + result = df.groupby(["key1", "key2"]).value.rank() + + expected = [piece.value.rank() for key, piece in df.groupby(["key1", "key2"])] + expected = concat(expected, axis=0) + expected = expected.reindex(result.index) + tm.assert_series_equal(result, expected) + + result = df.groupby(["key1", "key2"]).value.rank(pct=True) + + expected = [ + piece.value.rank(pct=True) for key, piece in df.groupby(["key1", "key2"]) + ] + expected = concat(expected, axis=0) + expected = expected.reindex(result.index) + tm.assert_series_equal(result, expected) + + +@pytest.mark.parametrize("grps", [["qux"], ["qux", "quux"]]) +@pytest.mark.parametrize( + "vals", + [ + np.array([2, 2, 8, 2, 6], dtype=dtype) + for dtype in ["i8", "i4", "i2", "i1", "u8", "u4", "u2", "u1", "f8", "f4", "f2"] + ] + + [ + [ + pd.Timestamp("2018-01-02"), + pd.Timestamp("2018-01-02"), + pd.Timestamp("2018-01-08"), + pd.Timestamp("2018-01-02"), + pd.Timestamp("2018-01-06"), + ], + [ + pd.Timestamp("2018-01-02", tz="US/Pacific"), + pd.Timestamp("2018-01-02", tz="US/Pacific"), + pd.Timestamp("2018-01-08", tz="US/Pacific"), + pd.Timestamp("2018-01-02", tz="US/Pacific"), + pd.Timestamp("2018-01-06", tz="US/Pacific"), + ], + [ + pd.Timestamp("2018-01-02") - pd.Timestamp(0), + pd.Timestamp("2018-01-02") - pd.Timestamp(0), + pd.Timestamp("2018-01-08") - pd.Timestamp(0), + pd.Timestamp("2018-01-02") - pd.Timestamp(0), + pd.Timestamp("2018-01-06") - pd.Timestamp(0), + ], + [ + pd.Timestamp("2018-01-02").to_period("D"), + pd.Timestamp("2018-01-02").to_period("D"), + pd.Timestamp("2018-01-08").to_period("D"), + pd.Timestamp("2018-01-02").to_period("D"), + pd.Timestamp("2018-01-06").to_period("D"), + ], + ], + ids=lambda x: type(x[0]), +) +@pytest.mark.parametrize( + "ties_method,ascending,pct,exp", + [ + ("average", True, False, [2.0, 2.0, 5.0, 2.0, 4.0]), + ("average", True, True, [0.4, 0.4, 1.0, 0.4, 0.8]), + ("average", False, False, [4.0, 4.0, 1.0, 4.0, 2.0]), + ("average", False, True, [0.8, 0.8, 0.2, 0.8, 0.4]), + ("min", True, False, [1.0, 1.0, 5.0, 1.0, 4.0]), + ("min", True, True, [0.2, 0.2, 1.0, 0.2, 0.8]), + ("min", False, False, [3.0, 3.0, 1.0, 3.0, 2.0]), + ("min", False, True, [0.6, 0.6, 0.2, 0.6, 0.4]), + ("max", True, False, [3.0, 3.0, 5.0, 3.0, 4.0]), + ("max", True, True, [0.6, 0.6, 1.0, 0.6, 0.8]), + ("max", False, False, [5.0, 5.0, 1.0, 5.0, 2.0]), + ("max", False, True, [1.0, 1.0, 0.2, 1.0, 0.4]), + ("first", True, False, [1.0, 2.0, 5.0, 3.0, 4.0]), + ("first", True, True, [0.2, 0.4, 1.0, 0.6, 0.8]), + ("first", False, False, [3.0, 4.0, 1.0, 5.0, 2.0]), + ("first", False, True, [0.6, 0.8, 0.2, 1.0, 0.4]), + ("dense", True, False, [1.0, 1.0, 3.0, 1.0, 2.0]), + ("dense", True, True, [1.0 / 3.0, 1.0 / 3.0, 3.0 / 3.0, 1.0 / 3.0, 2.0 / 3.0]), + ("dense", False, False, [3.0, 3.0, 1.0, 3.0, 2.0]), + ("dense", False, True, [3.0 / 3.0, 3.0 / 3.0, 1.0 / 3.0, 3.0 / 3.0, 2.0 / 3.0]), + ], +) +def test_rank_args(grps, vals, ties_method, ascending, pct, exp): + key = np.repeat(grps, len(vals)) + + orig_vals = vals + vals = list(vals) * len(grps) + if isinstance(orig_vals, np.ndarray): + vals = np.array(vals, dtype=orig_vals.dtype) + + df = DataFrame({"key": key, "val": vals}) + result = df.groupby("key").rank(method=ties_method, ascending=ascending, pct=pct) + + exp_df = DataFrame(exp * len(grps), columns=["val"]) + tm.assert_frame_equal(result, exp_df) + + +@pytest.mark.parametrize("grps", [["qux"], ["qux", "quux"]]) +@pytest.mark.parametrize( + "vals", [[-np.inf, -np.inf, np.nan, 1.0, np.nan, np.inf, np.inf]] +) +@pytest.mark.parametrize( + "ties_method,ascending,na_option,exp", + [ + ("average", True, "keep", [1.5, 1.5, np.nan, 3, np.nan, 4.5, 4.5]), + ("average", True, "top", [3.5, 3.5, 1.5, 5.0, 1.5, 6.5, 6.5]), + ("average", True, "bottom", [1.5, 1.5, 6.5, 3.0, 6.5, 4.5, 4.5]), + ("average", False, "keep", [4.5, 4.5, np.nan, 3, np.nan, 1.5, 1.5]), + ("average", False, "top", [6.5, 6.5, 1.5, 5.0, 1.5, 3.5, 3.5]), + ("average", False, "bottom", [4.5, 4.5, 6.5, 3.0, 6.5, 1.5, 1.5]), + ("min", True, "keep", [1.0, 1.0, np.nan, 3.0, np.nan, 4.0, 4.0]), + ("min", True, "top", [3.0, 3.0, 1.0, 5.0, 1.0, 6.0, 6.0]), + ("min", True, "bottom", [1.0, 1.0, 6.0, 3.0, 6.0, 4.0, 4.0]), + ("min", False, "keep", [4.0, 4.0, np.nan, 3.0, np.nan, 1.0, 1.0]), + ("min", False, "top", [6.0, 6.0, 1.0, 5.0, 1.0, 3.0, 3.0]), + ("min", False, "bottom", [4.0, 4.0, 6.0, 3.0, 6.0, 1.0, 1.0]), + ("max", True, "keep", [2.0, 2.0, np.nan, 3.0, np.nan, 5.0, 5.0]), + ("max", True, "top", [4.0, 4.0, 2.0, 5.0, 2.0, 7.0, 7.0]), + ("max", True, "bottom", [2.0, 2.0, 7.0, 3.0, 7.0, 5.0, 5.0]), + ("max", False, "keep", [5.0, 5.0, np.nan, 3.0, np.nan, 2.0, 2.0]), + ("max", False, "top", [7.0, 7.0, 2.0, 5.0, 2.0, 4.0, 4.0]), + ("max", False, "bottom", [5.0, 5.0, 7.0, 3.0, 7.0, 2.0, 2.0]), + ("first", True, "keep", [1.0, 2.0, np.nan, 3.0, np.nan, 4.0, 5.0]), + ("first", True, "top", [3.0, 4.0, 1.0, 5.0, 2.0, 6.0, 7.0]), + ("first", True, "bottom", [1.0, 2.0, 6.0, 3.0, 7.0, 4.0, 5.0]), + ("first", False, "keep", [4.0, 5.0, np.nan, 3.0, np.nan, 1.0, 2.0]), + ("first", False, "top", [6.0, 7.0, 1.0, 5.0, 2.0, 3.0, 4.0]), + ("first", False, "bottom", [4.0, 5.0, 6.0, 3.0, 7.0, 1.0, 2.0]), + ("dense", True, "keep", [1.0, 1.0, np.nan, 2.0, np.nan, 3.0, 3.0]), + ("dense", True, "top", [2.0, 2.0, 1.0, 3.0, 1.0, 4.0, 4.0]), + ("dense", True, "bottom", [1.0, 1.0, 4.0, 2.0, 4.0, 3.0, 3.0]), + ("dense", False, "keep", [3.0, 3.0, np.nan, 2.0, np.nan, 1.0, 1.0]), + ("dense", False, "top", [4.0, 4.0, 1.0, 3.0, 1.0, 2.0, 2.0]), + ("dense", False, "bottom", [3.0, 3.0, 4.0, 2.0, 4.0, 1.0, 1.0]), + ], +) +def test_infs_n_nans(grps, vals, ties_method, ascending, na_option, exp): + # GH 20561 + key = np.repeat(grps, len(vals)) + vals = vals * len(grps) + df = DataFrame({"key": key, "val": vals}) + result = df.groupby("key").rank( + method=ties_method, ascending=ascending, na_option=na_option + ) + exp_df = DataFrame(exp * len(grps), columns=["val"]) + tm.assert_frame_equal(result, exp_df) + + +@pytest.mark.parametrize("grps", [["qux"], ["qux", "quux"]]) +@pytest.mark.parametrize( + "vals", + [ + np.array([2, 2, np.nan, 8, 2, 6, np.nan, np.nan], dtype=dtype) + for dtype in ["f8", "f4", "f2"] + ] + + [ + [ + pd.Timestamp("2018-01-02"), + pd.Timestamp("2018-01-02"), + np.nan, + pd.Timestamp("2018-01-08"), + pd.Timestamp("2018-01-02"), + pd.Timestamp("2018-01-06"), + np.nan, + np.nan, + ], + [ + pd.Timestamp("2018-01-02", tz="US/Pacific"), + pd.Timestamp("2018-01-02", tz="US/Pacific"), + np.nan, + pd.Timestamp("2018-01-08", tz="US/Pacific"), + pd.Timestamp("2018-01-02", tz="US/Pacific"), + pd.Timestamp("2018-01-06", tz="US/Pacific"), + np.nan, + np.nan, + ], + [ + pd.Timestamp("2018-01-02") - pd.Timestamp(0), + pd.Timestamp("2018-01-02") - pd.Timestamp(0), + np.nan, + pd.Timestamp("2018-01-08") - pd.Timestamp(0), + pd.Timestamp("2018-01-02") - pd.Timestamp(0), + pd.Timestamp("2018-01-06") - pd.Timestamp(0), + np.nan, + np.nan, + ], + [ + pd.Timestamp("2018-01-02").to_period("D"), + pd.Timestamp("2018-01-02").to_period("D"), + np.nan, + pd.Timestamp("2018-01-08").to_period("D"), + pd.Timestamp("2018-01-02").to_period("D"), + pd.Timestamp("2018-01-06").to_period("D"), + np.nan, + np.nan, + ], + ], + ids=lambda x: type(x[0]), +) +@pytest.mark.parametrize( + "ties_method,ascending,na_option,pct,exp", + [ + ( + "average", + True, + "keep", + False, + [2.0, 2.0, np.nan, 5.0, 2.0, 4.0, np.nan, np.nan], + ), + ( + "average", + True, + "keep", + True, + [0.4, 0.4, np.nan, 1.0, 0.4, 0.8, np.nan, np.nan], + ), + ( + "average", + False, + "keep", + False, + [4.0, 4.0, np.nan, 1.0, 4.0, 2.0, np.nan, np.nan], + ), + ( + "average", + False, + "keep", + True, + [0.8, 0.8, np.nan, 0.2, 0.8, 0.4, np.nan, np.nan], + ), + ("min", True, "keep", False, [1.0, 1.0, np.nan, 5.0, 1.0, 4.0, np.nan, np.nan]), + ("min", True, "keep", True, [0.2, 0.2, np.nan, 1.0, 0.2, 0.8, np.nan, np.nan]), + ( + "min", + False, + "keep", + False, + [3.0, 3.0, np.nan, 1.0, 3.0, 2.0, np.nan, np.nan], + ), + ("min", False, "keep", True, [0.6, 0.6, np.nan, 0.2, 0.6, 0.4, np.nan, np.nan]), + ("max", True, "keep", False, [3.0, 3.0, np.nan, 5.0, 3.0, 4.0, np.nan, np.nan]), + ("max", True, "keep", True, [0.6, 0.6, np.nan, 1.0, 0.6, 0.8, np.nan, np.nan]), + ( + "max", + False, + "keep", + False, + [5.0, 5.0, np.nan, 1.0, 5.0, 2.0, np.nan, np.nan], + ), + ("max", False, "keep", True, [1.0, 1.0, np.nan, 0.2, 1.0, 0.4, np.nan, np.nan]), + ( + "first", + True, + "keep", + False, + [1.0, 2.0, np.nan, 5.0, 3.0, 4.0, np.nan, np.nan], + ), + ( + "first", + True, + "keep", + True, + [0.2, 0.4, np.nan, 1.0, 0.6, 0.8, np.nan, np.nan], + ), + ( + "first", + False, + "keep", + False, + [3.0, 4.0, np.nan, 1.0, 5.0, 2.0, np.nan, np.nan], + ), + ( + "first", + False, + "keep", + True, + [0.6, 0.8, np.nan, 0.2, 1.0, 0.4, np.nan, np.nan], + ), + ( + "dense", + True, + "keep", + False, + [1.0, 1.0, np.nan, 3.0, 1.0, 2.0, np.nan, np.nan], + ), + ( + "dense", + True, + "keep", + True, + [ + 1.0 / 3.0, + 1.0 / 3.0, + np.nan, + 3.0 / 3.0, + 1.0 / 3.0, + 2.0 / 3.0, + np.nan, + np.nan, + ], + ), + ( + "dense", + False, + "keep", + False, + [3.0, 3.0, np.nan, 1.0, 3.0, 2.0, np.nan, np.nan], + ), + ( + "dense", + False, + "keep", + True, + [ + 3.0 / 3.0, + 3.0 / 3.0, + np.nan, + 1.0 / 3.0, + 3.0 / 3.0, + 2.0 / 3.0, + np.nan, + np.nan, + ], + ), + ("average", True, "bottom", False, [2.0, 2.0, 7.0, 5.0, 2.0, 4.0, 7.0, 7.0]), + ( + "average", + True, + "bottom", + True, + [0.25, 0.25, 0.875, 0.625, 0.25, 0.5, 0.875, 0.875], + ), + ("average", False, "bottom", False, [4.0, 4.0, 7.0, 1.0, 4.0, 2.0, 7.0, 7.0]), + ( + "average", + False, + "bottom", + True, + [0.5, 0.5, 0.875, 0.125, 0.5, 0.25, 0.875, 0.875], + ), + ("min", True, "bottom", False, [1.0, 1.0, 6.0, 5.0, 1.0, 4.0, 6.0, 6.0]), + ( + "min", + True, + "bottom", + True, + [0.125, 0.125, 0.75, 0.625, 0.125, 0.5, 0.75, 0.75], + ), + ("min", False, "bottom", False, [3.0, 3.0, 6.0, 1.0, 3.0, 2.0, 6.0, 6.0]), + ( + "min", + False, + "bottom", + True, + [0.375, 0.375, 0.75, 0.125, 0.375, 0.25, 0.75, 0.75], + ), + ("max", True, "bottom", False, [3.0, 3.0, 8.0, 5.0, 3.0, 4.0, 8.0, 8.0]), + ("max", True, "bottom", True, [0.375, 0.375, 1.0, 0.625, 0.375, 0.5, 1.0, 1.0]), + ("max", False, "bottom", False, [5.0, 5.0, 8.0, 1.0, 5.0, 2.0, 8.0, 8.0]), + ( + "max", + False, + "bottom", + True, + [0.625, 0.625, 1.0, 0.125, 0.625, 0.25, 1.0, 1.0], + ), + ("first", True, "bottom", False, [1.0, 2.0, 6.0, 5.0, 3.0, 4.0, 7.0, 8.0]), + ( + "first", + True, + "bottom", + True, + [0.125, 0.25, 0.75, 0.625, 0.375, 0.5, 0.875, 1.0], + ), + ("first", False, "bottom", False, [3.0, 4.0, 6.0, 1.0, 5.0, 2.0, 7.0, 8.0]), + ( + "first", + False, + "bottom", + True, + [0.375, 0.5, 0.75, 0.125, 0.625, 0.25, 0.875, 1.0], + ), + ("dense", True, "bottom", False, [1.0, 1.0, 4.0, 3.0, 1.0, 2.0, 4.0, 4.0]), + ("dense", True, "bottom", True, [0.25, 0.25, 1.0, 0.75, 0.25, 0.5, 1.0, 1.0]), + ("dense", False, "bottom", False, [3.0, 3.0, 4.0, 1.0, 3.0, 2.0, 4.0, 4.0]), + ("dense", False, "bottom", True, [0.75, 0.75, 1.0, 0.25, 0.75, 0.5, 1.0, 1.0]), + ], +) +def test_rank_args_missing(grps, vals, ties_method, ascending, na_option, pct, exp): + key = np.repeat(grps, len(vals)) + + orig_vals = vals + vals = list(vals) * len(grps) + if isinstance(orig_vals, np.ndarray): + vals = np.array(vals, dtype=orig_vals.dtype) + + df = DataFrame({"key": key, "val": vals}) + result = df.groupby("key").rank( + method=ties_method, ascending=ascending, na_option=na_option, pct=pct + ) + + exp_df = DataFrame(exp * len(grps), columns=["val"]) + tm.assert_frame_equal(result, exp_df) + + +@pytest.mark.parametrize( + "pct,exp", [(False, [3.0, 3.0, 3.0, 3.0, 3.0]), (True, [0.6, 0.6, 0.6, 0.6, 0.6])] +) +def test_rank_resets_each_group(pct, exp): + df = DataFrame( + {"key": ["a", "a", "a", "a", "a", "b", "b", "b", "b", "b"], "val": [1] * 10} + ) + result = df.groupby("key").rank(pct=pct) + exp_df = DataFrame(exp * 2, columns=["val"]) + tm.assert_frame_equal(result, exp_df) + + +@pytest.mark.parametrize( + "dtype", ["int64", "int32", "uint64", "uint32", "float64", "float32"] +) +@pytest.mark.parametrize("upper", [True, False]) +def test_rank_avg_even_vals(dtype, upper): + if upper: + # use IntegerDtype/FloatingDtype + dtype = dtype[0].upper() + dtype[1:] + dtype = dtype.replace("Ui", "UI") + df = DataFrame({"key": ["a"] * 4, "val": [1] * 4}) + df["val"] = df["val"].astype(dtype) + assert df["val"].dtype == dtype + + result = df.groupby("key").rank() + exp_df = DataFrame([2.5, 2.5, 2.5, 2.5], columns=["val"]) + if upper: + exp_df = exp_df.astype("Float64") + tm.assert_frame_equal(result, exp_df) + + +@pytest.mark.parametrize("ties_method", ["average", "min", "max", "first", "dense"]) +@pytest.mark.parametrize("ascending", [True, False]) +@pytest.mark.parametrize("na_option", ["keep", "top", "bottom"]) +@pytest.mark.parametrize("pct", [True, False]) +@pytest.mark.parametrize( + "vals", [["bar", "bar", "foo", "bar", "baz"], ["bar", np.nan, "foo", np.nan, "baz"]] +) +def test_rank_object_dtype(ties_method, ascending, na_option, pct, vals): + df = DataFrame({"key": ["foo"] * 5, "val": vals}) + mask = df["val"].isna() + + gb = df.groupby("key") + res = gb.rank(method=ties_method, ascending=ascending, na_option=na_option, pct=pct) + + # construct our expected by using numeric values with the same ordering + if mask.any(): + df2 = DataFrame({"key": ["foo"] * 5, "val": [0, np.nan, 2, np.nan, 1]}) + else: + df2 = DataFrame({"key": ["foo"] * 5, "val": [0, 0, 2, 0, 1]}) + + gb2 = df2.groupby("key") + alt = gb2.rank( + method=ties_method, ascending=ascending, na_option=na_option, pct=pct + ) + + tm.assert_frame_equal(res, alt) + + +@pytest.mark.parametrize("na_option", [True, "bad", 1]) +@pytest.mark.parametrize("ties_method", ["average", "min", "max", "first", "dense"]) +@pytest.mark.parametrize("ascending", [True, False]) +@pytest.mark.parametrize("pct", [True, False]) +@pytest.mark.parametrize( + "vals", + [ + ["bar", "bar", "foo", "bar", "baz"], + ["bar", np.nan, "foo", np.nan, "baz"], + [1, np.nan, 2, np.nan, 3], + ], +) +def test_rank_naoption_raises(ties_method, ascending, na_option, pct, vals): + df = DataFrame({"key": ["foo"] * 5, "val": vals}) + msg = "na_option must be one of 'keep', 'top', or 'bottom'" + + with pytest.raises(ValueError, match=msg): + df.groupby("key").rank( + method=ties_method, ascending=ascending, na_option=na_option, pct=pct + ) + + +def test_rank_empty_group(): + # see gh-22519 + column = "A" + df = DataFrame({"A": [0, 1, 0], "B": [1.0, np.nan, 2.0]}) + + result = df.groupby(column).B.rank(pct=True) + expected = Series([0.5, np.nan, 1.0], name="B") + tm.assert_series_equal(result, expected) + + result = df.groupby(column).rank(pct=True) + expected = DataFrame({"B": [0.5, np.nan, 1.0]}) + tm.assert_frame_equal(result, expected) + + +@pytest.mark.parametrize( + "input_key,input_value,output_value", + [ + ([1, 2], [1, 1], [1.0, 1.0]), + ([1, 1, 2, 2], [1, 2, 1, 2], [0.5, 1.0, 0.5, 1.0]), + ([1, 1, 2, 2], [1, 2, 1, np.nan], [0.5, 1.0, 1.0, np.nan]), + ([1, 1, 2], [1, 2, np.nan], [0.5, 1.0, np.nan]), + ], +) +def test_rank_zero_div(input_key, input_value, output_value): + # GH 23666 + df = DataFrame({"A": input_key, "B": input_value}) + + result = df.groupby("A").rank(method="dense", pct=True) + expected = DataFrame({"B": output_value}) + tm.assert_frame_equal(result, expected) + + +def test_rank_min_int(): + # GH-32859 + df = DataFrame( + { + "grp": [1, 1, 2], + "int_col": [ + np.iinfo(np.int64).min, + np.iinfo(np.int64).max, + np.iinfo(np.int64).min, + ], + "datetimelike": [NaT, datetime(2001, 1, 1), NaT], + } + ) + + result = df.groupby("grp").rank() + expected = DataFrame( + {"int_col": [1.0, 2.0, 1.0], "datetimelike": [np.nan, 1.0, np.nan]} + ) + + tm.assert_frame_equal(result, expected) + + +@pytest.mark.parametrize("use_nan", [True, False]) +def test_rank_pct_equal_values_on_group_transition(use_nan): + # GH#40518 + fill_value = np.nan if use_nan else 3 + df = DataFrame( + [ + [-1, 1], + [-1, 2], + [1, fill_value], + [-1, fill_value], + ], + columns=["group", "val"], + ) + result = df.groupby(["group"])["val"].rank( + method="dense", + pct=True, + ) + if use_nan: + expected = Series([0.5, 1, np.nan, np.nan], name="val") + else: + expected = Series([1 / 3, 2 / 3, 1, 1], name="val") + + tm.assert_series_equal(result, expected) + + +def test_rank_multiindex(): + # GH27721 + df = concat( + { + "a": DataFrame({"col1": [3, 4], "col2": [1, 2]}), + "b": DataFrame({"col3": [5, 6], "col4": [7, 8]}), + }, + axis=1, + ) + + msg = "DataFrame.groupby with axis=1 is deprecated" + with tm.assert_produces_warning(FutureWarning, match=msg): + gb = df.groupby(level=0, axis=1) + msg = "DataFrameGroupBy.rank with axis=1 is deprecated" + with tm.assert_produces_warning(FutureWarning, match=msg): + result = gb.rank(axis=1) + + expected = concat( + [ + df["a"].rank(axis=1), + df["b"].rank(axis=1), + ], + axis=1, + keys=["a", "b"], + ) + tm.assert_frame_equal(result, expected) + + +def test_groupby_axis0_rank_axis1(): + # GH#41320 + df = DataFrame( + {0: [1, 3, 5, 7], 1: [2, 4, 6, 8], 2: [1.5, 3.5, 5.5, 7.5]}, + index=["a", "a", "b", "b"], + ) + msg = "The 'axis' keyword in DataFrame.groupby is deprecated" + with tm.assert_produces_warning(FutureWarning, match=msg): + gb = df.groupby(level=0, axis=0) + + msg = "DataFrameGroupBy.rank with axis=1 is deprecated" + with tm.assert_produces_warning(FutureWarning, match=msg): + res = gb.rank(axis=1) + + # This should match what we get when "manually" operating group-by-group + expected = concat([df.loc["a"].rank(axis=1), df.loc["b"].rank(axis=1)], axis=0) + tm.assert_frame_equal(res, expected) + + # check that we haven't accidentally written a case that coincidentally + # matches rank(axis=0) + msg = "The 'axis' keyword in DataFrameGroupBy.rank" + with tm.assert_produces_warning(FutureWarning, match=msg): + alt = gb.rank(axis=0) + assert not alt.equals(expected) + + +def test_groupby_axis0_cummax_axis1(): + # case where groupby axis is 0 and axis keyword in transform is 1 + + # df has mixed dtype -> multiple blocks + df = DataFrame( + {0: [1, 3, 5, 7], 1: [2, 4, 6, 8], 2: [1.5, 3.5, 5.5, 7.5]}, + index=["a", "a", "b", "b"], + ) + msg = "The 'axis' keyword in DataFrame.groupby is deprecated" + with tm.assert_produces_warning(FutureWarning, match=msg): + gb = df.groupby(level=0, axis=0) + + msg = "DataFrameGroupBy.cummax with axis=1 is deprecated" + with tm.assert_produces_warning(FutureWarning, match=msg): + cmax = gb.cummax(axis=1) + expected = df[[0, 1]].astype(np.float64) + expected[2] = expected[1] + tm.assert_frame_equal(cmax, expected) + + +def test_non_unique_index(): + # GH 16577 + df = DataFrame( + {"A": [1.0, 2.0, 3.0, np.nan], "value": 1.0}, + index=[pd.Timestamp("20170101", tz="US/Eastern")] * 4, + ) + result = df.groupby([df.index, "A"]).value.rank(ascending=True, pct=True) + expected = Series( + [1.0, 1.0, 1.0, np.nan], + index=[pd.Timestamp("20170101", tz="US/Eastern")] * 4, + name="value", + ) + tm.assert_series_equal(result, expected) + + +def test_rank_categorical(): + cat = pd.Categorical(["a", "a", "b", np.nan, "c", "b"], ordered=True) + cat2 = pd.Categorical([1, 2, 3, np.nan, 4, 5], ordered=True) + + df = DataFrame({"col1": [0, 1, 0, 1, 0, 1], "col2": cat, "col3": cat2}) + + gb = df.groupby("col1") + + res = gb.rank() + + expected = df.astype(object).groupby("col1").rank() + tm.assert_frame_equal(res, expected) + + +@pytest.mark.parametrize("na_option", ["top", "bottom"]) +def test_groupby_op_with_nullables(na_option): + # GH 54206 + df = DataFrame({"x": [None]}, dtype="Float64") + result = df.groupby("x", dropna=False)["x"].rank(method="min", na_option=na_option) + expected = Series([1.0], dtype="Float64", name=result.name) + tm.assert_series_equal(result, expected) diff --git a/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/test_sample.py b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/test_sample.py new file mode 100644 index 0000000000000000000000000000000000000000..4dd474741740d4abdea1ebabf2b36c3b68d690ad --- /dev/null +++ b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/test_sample.py @@ -0,0 +1,154 @@ +import pytest + +from pandas import ( + DataFrame, + Index, + Series, +) +import pandas._testing as tm + + +@pytest.mark.parametrize("n, frac", [(2, None), (None, 0.2)]) +def test_groupby_sample_balanced_groups_shape(n, frac): + values = [1] * 10 + [2] * 10 + df = DataFrame({"a": values, "b": values}) + + result = df.groupby("a").sample(n=n, frac=frac) + values = [1] * 2 + [2] * 2 + expected = DataFrame({"a": values, "b": values}, index=result.index) + tm.assert_frame_equal(result, expected) + + result = df.groupby("a")["b"].sample(n=n, frac=frac) + expected = Series(values, name="b", index=result.index) + tm.assert_series_equal(result, expected) + + +def test_groupby_sample_unbalanced_groups_shape(): + values = [1] * 10 + [2] * 20 + df = DataFrame({"a": values, "b": values}) + + result = df.groupby("a").sample(n=5) + values = [1] * 5 + [2] * 5 + expected = DataFrame({"a": values, "b": values}, index=result.index) + tm.assert_frame_equal(result, expected) + + result = df.groupby("a")["b"].sample(n=5) + expected = Series(values, name="b", index=result.index) + tm.assert_series_equal(result, expected) + + +def test_groupby_sample_index_value_spans_groups(): + values = [1] * 3 + [2] * 3 + df = DataFrame({"a": values, "b": values}, index=[1, 2, 2, 2, 2, 2]) + + result = df.groupby("a").sample(n=2) + values = [1] * 2 + [2] * 2 + expected = DataFrame({"a": values, "b": values}, index=result.index) + tm.assert_frame_equal(result, expected) + + result = df.groupby("a")["b"].sample(n=2) + expected = Series(values, name="b", index=result.index) + tm.assert_series_equal(result, expected) + + +def test_groupby_sample_n_and_frac_raises(): + df = DataFrame({"a": [1, 2], "b": [1, 2]}) + msg = "Please enter a value for `frac` OR `n`, not both" + + with pytest.raises(ValueError, match=msg): + df.groupby("a").sample(n=1, frac=1.0) + + with pytest.raises(ValueError, match=msg): + df.groupby("a")["b"].sample(n=1, frac=1.0) + + +def test_groupby_sample_frac_gt_one_without_replacement_raises(): + df = DataFrame({"a": [1, 2], "b": [1, 2]}) + msg = "Replace has to be set to `True` when upsampling the population `frac` > 1." + + with pytest.raises(ValueError, match=msg): + df.groupby("a").sample(frac=1.5, replace=False) + + with pytest.raises(ValueError, match=msg): + df.groupby("a")["b"].sample(frac=1.5, replace=False) + + +@pytest.mark.parametrize("n", [-1, 1.5]) +def test_groupby_sample_invalid_n_raises(n): + df = DataFrame({"a": [1, 2], "b": [1, 2]}) + + if n < 0: + msg = "A negative number of rows requested. Please provide `n` >= 0." + else: + msg = "Only integers accepted as `n` values" + + with pytest.raises(ValueError, match=msg): + df.groupby("a").sample(n=n) + + with pytest.raises(ValueError, match=msg): + df.groupby("a")["b"].sample(n=n) + + +def test_groupby_sample_oversample(): + values = [1] * 10 + [2] * 10 + df = DataFrame({"a": values, "b": values}) + + result = df.groupby("a").sample(frac=2.0, replace=True) + values = [1] * 20 + [2] * 20 + expected = DataFrame({"a": values, "b": values}, index=result.index) + tm.assert_frame_equal(result, expected) + + result = df.groupby("a")["b"].sample(frac=2.0, replace=True) + expected = Series(values, name="b", index=result.index) + tm.assert_series_equal(result, expected) + + +def test_groupby_sample_without_n_or_frac(): + values = [1] * 10 + [2] * 10 + df = DataFrame({"a": values, "b": values}) + + result = df.groupby("a").sample(n=None, frac=None) + expected = DataFrame({"a": [1, 2], "b": [1, 2]}, index=result.index) + tm.assert_frame_equal(result, expected) + + result = df.groupby("a")["b"].sample(n=None, frac=None) + expected = Series([1, 2], name="b", index=result.index) + tm.assert_series_equal(result, expected) + + +@pytest.mark.parametrize( + "index, expected_index", + [(["w", "x", "y", "z"], ["w", "w", "y", "y"]), ([3, 4, 5, 6], [3, 3, 5, 5])], +) +def test_groupby_sample_with_weights(index, expected_index): + # GH 39927 - tests for integer index needed + values = [1] * 2 + [2] * 2 + df = DataFrame({"a": values, "b": values}, index=Index(index)) + + result = df.groupby("a").sample(n=2, replace=True, weights=[1, 0, 1, 0]) + expected = DataFrame({"a": values, "b": values}, index=Index(expected_index)) + tm.assert_frame_equal(result, expected) + + result = df.groupby("a")["b"].sample(n=2, replace=True, weights=[1, 0, 1, 0]) + expected = Series(values, name="b", index=Index(expected_index)) + tm.assert_series_equal(result, expected) + + +def test_groupby_sample_with_selections(): + # GH 39928 + values = [1] * 10 + [2] * 10 + df = DataFrame({"a": values, "b": values, "c": values}) + + result = df.groupby("a")[["b", "c"]].sample(n=None, frac=None) + expected = DataFrame({"b": [1, 2], "c": [1, 2]}, index=result.index) + tm.assert_frame_equal(result, expected) + + +def test_groupby_sample_with_empty_inputs(): + # GH48459 + df = DataFrame({"a": [], "b": []}) + groupby_df = df.groupby("a") + + result = groupby_df.sample() + expected = df + tm.assert_frame_equal(result, expected) diff --git a/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/test_size.py b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/test_size.py new file mode 100644 index 0000000000000000000000000000000000000000..93a4e743d0d71db1d2a1fcca4163e6db83eb4ffb --- /dev/null +++ b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/test_size.py @@ -0,0 +1,130 @@ +import numpy as np +import pytest + +import pandas.util._test_decorators as td + +from pandas.core.dtypes.common import is_integer_dtype + +from pandas import ( + DataFrame, + Index, + PeriodIndex, + Series, +) +import pandas._testing as tm + + +@pytest.mark.parametrize("by", ["A", "B", ["A", "B"]]) +def test_size(df, by): + grouped = df.groupby(by=by) + result = grouped.size() + for key, group in grouped: + assert result[key] == len(group) + + +@pytest.mark.parametrize( + "by", + [ + [0, 0, 0, 0], + [0, 1, 1, 1], + [1, 0, 1, 1], + [0, None, None, None], + pytest.param([None, None, None, None], marks=pytest.mark.xfail), + ], +) +def test_size_axis_1(df, axis_1, by, sort, dropna): + # GH#45715 + counts = {key: sum(value == key for value in by) for key in dict.fromkeys(by)} + if dropna: + counts = {key: value for key, value in counts.items() if key is not None} + expected = Series(counts, dtype="int64") + if sort: + expected = expected.sort_index() + if is_integer_dtype(expected.index.dtype) and not any(x is None for x in by): + expected.index = expected.index.astype(int) + + msg = "DataFrame.groupby with axis=1 is deprecated" + with tm.assert_produces_warning(FutureWarning, match=msg): + grouped = df.groupby(by=by, axis=axis_1, sort=sort, dropna=dropna) + result = grouped.size() + tm.assert_series_equal(result, expected) + + +@pytest.mark.parametrize("by", ["A", "B", ["A", "B"]]) +@pytest.mark.parametrize("sort", [True, False]) +def test_size_sort(sort, by): + df = DataFrame(np.random.default_rng(2).choice(20, (1000, 3)), columns=list("ABC")) + left = df.groupby(by=by, sort=sort).size() + right = df.groupby(by=by, sort=sort)["C"].apply(lambda a: a.shape[0]) + tm.assert_series_equal(left, right, check_names=False) + + +def test_size_series_dataframe(): + # https://github.com/pandas-dev/pandas/issues/11699 + df = DataFrame(columns=["A", "B"]) + out = Series(dtype="int64", index=Index([], name="A")) + tm.assert_series_equal(df.groupby("A").size(), out) + + +def test_size_groupby_all_null(): + # https://github.com/pandas-dev/pandas/issues/23050 + # Assert no 'Value Error : Length of passed values is 2, index implies 0' + df = DataFrame({"A": [None, None]}) # all-null groups + result = df.groupby("A").size() + expected = Series(dtype="int64", index=Index([], name="A")) + tm.assert_series_equal(result, expected) + + +def test_size_period_index(): + # https://github.com/pandas-dev/pandas/issues/34010 + ser = Series([1], index=PeriodIndex(["2000"], name="A", freq="D")) + grp = ser.groupby(level="A") + result = grp.size() + tm.assert_series_equal(result, ser) + + +@pytest.mark.parametrize("as_index", [True, False]) +def test_size_on_categorical(as_index): + df = DataFrame([[1, 1], [2, 2]], columns=["A", "B"]) + df["A"] = df["A"].astype("category") + result = df.groupby(["A", "B"], as_index=as_index, observed=False).size() + + expected = DataFrame( + [[1, 1, 1], [1, 2, 0], [2, 1, 0], [2, 2, 1]], columns=["A", "B", "size"] + ) + expected["A"] = expected["A"].astype("category") + if as_index: + expected = expected.set_index(["A", "B"])["size"].rename(None) + + tm.assert_equal(result, expected) + + +@pytest.mark.parametrize("dtype", ["Int64", "Float64", "boolean"]) +def test_size_series_masked_type_returns_Int64(dtype): + # GH 54132 + ser = Series([1, 1, 1], index=["a", "a", "b"], dtype=dtype) + result = ser.groupby(level=0).size() + expected = Series([2, 1], dtype="Int64", index=["a", "b"]) + tm.assert_series_equal(result, expected) + + +@pytest.mark.parametrize( + "dtype", + [ + object, + pytest.param("string[pyarrow_numpy]", marks=td.skip_if_no("pyarrow")), + pytest.param("string[pyarrow]", marks=td.skip_if_no("pyarrow")), + ], +) +def test_size_strings(dtype): + # GH#55627 + df = DataFrame({"a": ["a", "a", "b"], "b": "a"}, dtype=dtype) + result = df.groupby("a")["b"].size() + exp_dtype = "Int64" if dtype == "string[pyarrow]" else "int64" + expected = Series( + [2, 1], + index=Index(["a", "b"], name="a", dtype=dtype), + name="b", + dtype=exp_dtype, + ) + tm.assert_series_equal(result, expected) diff --git a/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/test_skew.py b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/test_skew.py new file mode 100644 index 0000000000000000000000000000000000000000..563da89b6ab24a898f042f0e21377ccc2709b072 --- /dev/null +++ b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/test_skew.py @@ -0,0 +1,27 @@ +import numpy as np + +import pandas as pd +import pandas._testing as tm + + +def test_groupby_skew_equivalence(): + # Test that that groupby skew method (which uses libgroupby.group_skew) + # matches the results of operating group-by-group (which uses nanops.nanskew) + nrows = 1000 + ngroups = 3 + ncols = 2 + nan_frac = 0.05 + + arr = np.random.default_rng(2).standard_normal((nrows, ncols)) + arr[np.random.default_rng(2).random(nrows) < nan_frac] = np.nan + + df = pd.DataFrame(arr) + grps = np.random.default_rng(2).integers(0, ngroups, size=nrows) + gb = df.groupby(grps) + + result = gb.skew() + + grpwise = [grp.skew().to_frame(i).T for i, grp in gb] + expected = pd.concat(grpwise, axis=0) + expected.index = expected.index.astype(result.index.dtype) # 32bit builds + tm.assert_frame_equal(result, expected) diff --git a/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/test_value_counts.py b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/test_value_counts.py new file mode 100644 index 0000000000000000000000000000000000000000..8e25177368d8b7bbc412c930dc6ef2a278aa29db --- /dev/null +++ b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/methods/test_value_counts.py @@ -0,0 +1,1241 @@ +""" +these are systematically testing all of the args to value_counts +with different size combinations. This is to ensure stability of the sorting +and proper parameter handling +""" + + +import numpy as np +import pytest + +import pandas.util._test_decorators as td + +from pandas import ( + Categorical, + CategoricalIndex, + DataFrame, + Grouper, + Index, + MultiIndex, + Series, + date_range, + to_datetime, +) +import pandas._testing as tm +from pandas.util.version import Version + + +def tests_value_counts_index_names_category_column(): + # GH44324 Missing name of index category column + df = DataFrame( + { + "gender": ["female"], + "country": ["US"], + } + ) + df["gender"] = df["gender"].astype("category") + result = df.groupby("country")["gender"].value_counts() + + # Construct expected, very specific multiindex + df_mi_expected = DataFrame([["US", "female"]], columns=["country", "gender"]) + df_mi_expected["gender"] = df_mi_expected["gender"].astype("category") + mi_expected = MultiIndex.from_frame(df_mi_expected) + expected = Series([1], index=mi_expected, name="count") + + tm.assert_series_equal(result, expected) + + +def seed_df(seed_nans, n, m): + days = date_range("2015-08-24", periods=10) + + frame = DataFrame( + { + "1st": np.random.default_rng(2).choice(list("abcd"), n), + "2nd": np.random.default_rng(2).choice(days, n), + "3rd": np.random.default_rng(2).integers(1, m + 1, n), + } + ) + + if seed_nans: + # Explicitly cast to float to avoid implicit cast when setting nan + frame["3rd"] = frame["3rd"].astype("float") + frame.loc[1::11, "1st"] = np.nan + frame.loc[3::17, "2nd"] = np.nan + frame.loc[7::19, "3rd"] = np.nan + frame.loc[8::19, "3rd"] = np.nan + frame.loc[9::19, "3rd"] = np.nan + + return frame + + +@pytest.mark.slow +@pytest.mark.parametrize("seed_nans", [True, False]) +@pytest.mark.parametrize("num_rows", [10, 50]) +@pytest.mark.parametrize("max_int", [5, 20]) +@pytest.mark.parametrize("keys", ["1st", "2nd", ["1st", "2nd"]], ids=repr) +@pytest.mark.parametrize("bins", [None, [0, 5]], ids=repr) +@pytest.mark.parametrize("isort", [True, False]) +@pytest.mark.parametrize("normalize, name", [(True, "proportion"), (False, "count")]) +@pytest.mark.parametrize("sort", [True, False]) +@pytest.mark.parametrize("ascending", [True, False]) +@pytest.mark.parametrize("dropna", [True, False]) +def test_series_groupby_value_counts( + seed_nans, + num_rows, + max_int, + keys, + bins, + isort, + normalize, + name, + sort, + ascending, + dropna, +): + df = seed_df(seed_nans, num_rows, max_int) + + def rebuild_index(df): + arr = list(map(df.index.get_level_values, range(df.index.nlevels))) + df.index = MultiIndex.from_arrays(arr, names=df.index.names) + return df + + kwargs = { + "normalize": normalize, + "sort": sort, + "ascending": ascending, + "dropna": dropna, + "bins": bins, + } + + gr = df.groupby(keys, sort=isort) + left = gr["3rd"].value_counts(**kwargs) + + gr = df.groupby(keys, sort=isort) + right = gr["3rd"].apply(Series.value_counts, **kwargs) + right.index.names = right.index.names[:-1] + ["3rd"] + # https://github.com/pandas-dev/pandas/issues/49909 + right = right.rename(name) + + # have to sort on index because of unstable sort on values + left, right = map(rebuild_index, (left, right)) # xref GH9212 + tm.assert_series_equal(left.sort_index(), right.sort_index()) + + +@pytest.mark.parametrize("utc", [True, False]) +def test_series_groupby_value_counts_with_grouper(utc): + # GH28479 + df = DataFrame( + { + "Timestamp": [ + 1565083561, + 1565083561 + 86400, + 1565083561 + 86500, + 1565083561 + 86400 * 2, + 1565083561 + 86400 * 3, + 1565083561 + 86500 * 3, + 1565083561 + 86400 * 4, + ], + "Food": ["apple", "apple", "banana", "banana", "orange", "orange", "pear"], + } + ).drop([3]) + + df["Datetime"] = to_datetime(df["Timestamp"], utc=utc, unit="s") + dfg = df.groupby(Grouper(freq="1D", key="Datetime")) + + # have to sort on index because of unstable sort on values xref GH9212 + result = dfg["Food"].value_counts().sort_index() + expected = dfg["Food"].apply(Series.value_counts).sort_index() + expected.index.names = result.index.names + # https://github.com/pandas-dev/pandas/issues/49909 + expected = expected.rename("count") + + tm.assert_series_equal(result, expected) + + +@pytest.mark.parametrize("columns", [["A", "B"], ["A", "B", "C"]]) +def test_series_groupby_value_counts_empty(columns): + # GH39172 + df = DataFrame(columns=columns) + dfg = df.groupby(columns[:-1]) + + result = dfg[columns[-1]].value_counts() + expected = Series([], dtype=result.dtype, name="count") + expected.index = MultiIndex.from_arrays([[]] * len(columns), names=columns) + + tm.assert_series_equal(result, expected) + + +@pytest.mark.parametrize("columns", [["A", "B"], ["A", "B", "C"]]) +def test_series_groupby_value_counts_one_row(columns): + # GH42618 + df = DataFrame(data=[range(len(columns))], columns=columns) + dfg = df.groupby(columns[:-1]) + + result = dfg[columns[-1]].value_counts() + expected = df.value_counts() + + tm.assert_series_equal(result, expected) + + +def test_series_groupby_value_counts_on_categorical(): + # GH38672 + + s = Series(Categorical(["a"], categories=["a", "b"])) + result = s.groupby([0]).value_counts() + + expected = Series( + data=[1, 0], + index=MultiIndex.from_arrays( + [ + np.array([0, 0]), + CategoricalIndex( + ["a", "b"], categories=["a", "b"], ordered=False, dtype="category" + ), + ] + ), + name="count", + ) + + # Expected: + # 0 a 1 + # b 0 + # dtype: int64 + + tm.assert_series_equal(result, expected) + + +def test_series_groupby_value_counts_no_sort(): + # GH#50482 + df = DataFrame( + { + "gender": ["male", "male", "female", "male", "female", "male"], + "education": ["low", "medium", "high", "low", "high", "low"], + "country": ["US", "FR", "US", "FR", "FR", "FR"], + } + ) + gb = df.groupby(["country", "gender"], sort=False)["education"] + result = gb.value_counts(sort=False) + index = MultiIndex( + levels=[["US", "FR"], ["male", "female"], ["low", "medium", "high"]], + codes=[[0, 1, 0, 1, 1], [0, 0, 1, 0, 1], [0, 1, 2, 0, 2]], + names=["country", "gender", "education"], + ) + expected = Series([1, 1, 1, 2, 1], index=index, name="count") + tm.assert_series_equal(result, expected) + + +@pytest.fixture +def education_df(): + return DataFrame( + { + "gender": ["male", "male", "female", "male", "female", "male"], + "education": ["low", "medium", "high", "low", "high", "low"], + "country": ["US", "FR", "US", "FR", "FR", "FR"], + } + ) + + +def test_axis(education_df): + msg = "DataFrame.groupby with axis=1 is deprecated" + with tm.assert_produces_warning(FutureWarning, match=msg): + gp = education_df.groupby("country", axis=1) + with pytest.raises(NotImplementedError, match="axis"): + gp.value_counts() + + +def test_bad_subset(education_df): + gp = education_df.groupby("country") + with pytest.raises(ValueError, match="subset"): + gp.value_counts(subset=["country"]) + + +def test_basic(education_df, request): + # gh43564 + if Version(np.__version__) >= Version("1.25"): + request.applymarker( + pytest.mark.xfail( + reason=( + "pandas default unstable sorting of duplicates" + "issue with numpy>=1.25 with AVX instructions" + ), + strict=False, + ) + ) + result = education_df.groupby("country")[["gender", "education"]].value_counts( + normalize=True + ) + expected = Series( + data=[0.5, 0.25, 0.25, 0.5, 0.5], + index=MultiIndex.from_tuples( + [ + ("FR", "male", "low"), + ("FR", "female", "high"), + ("FR", "male", "medium"), + ("US", "female", "high"), + ("US", "male", "low"), + ], + names=["country", "gender", "education"], + ), + name="proportion", + ) + tm.assert_series_equal(result, expected) + + +def _frame_value_counts(df, keys, normalize, sort, ascending): + return df[keys].value_counts(normalize=normalize, sort=sort, ascending=ascending) + + +@pytest.mark.parametrize("groupby", ["column", "array", "function"]) +@pytest.mark.parametrize("normalize, name", [(True, "proportion"), (False, "count")]) +@pytest.mark.parametrize( + "sort, ascending", + [ + (False, None), + (True, True), + (True, False), + ], +) +@pytest.mark.parametrize("as_index", [True, False]) +@pytest.mark.parametrize("frame", [True, False]) +def test_against_frame_and_seriesgroupby( + education_df, groupby, normalize, name, sort, ascending, as_index, frame, request +): + # test all parameters: + # - Use column, array or function as by= parameter + # - Whether or not to normalize + # - Whether or not to sort and how + # - Whether or not to use the groupby as an index + # - 3-way compare against: + # - apply with :meth:`~DataFrame.value_counts` + # - `~SeriesGroupBy.value_counts` + if Version(np.__version__) >= Version("1.25") and frame and sort and normalize: + request.applymarker( + pytest.mark.xfail( + reason=( + "pandas default unstable sorting of duplicates" + "issue with numpy>=1.25 with AVX instructions" + ), + strict=False, + ) + ) + by = { + "column": "country", + "array": education_df["country"].values, + "function": lambda x: education_df["country"][x] == "US", + }[groupby] + + gp = education_df.groupby(by=by, as_index=as_index) + result = gp[["gender", "education"]].value_counts( + normalize=normalize, sort=sort, ascending=ascending + ) + if frame: + # compare against apply with DataFrame value_counts + warn = DeprecationWarning if groupby == "column" else None + msg = "DataFrameGroupBy.apply operated on the grouping columns" + with tm.assert_produces_warning(warn, match=msg): + expected = gp.apply( + _frame_value_counts, ["gender", "education"], normalize, sort, ascending + ) + + if as_index: + tm.assert_series_equal(result, expected) + else: + name = "proportion" if normalize else "count" + expected = expected.reset_index().rename({0: name}, axis=1) + if groupby == "column": + expected = expected.rename({"level_0": "country"}, axis=1) + expected["country"] = np.where(expected["country"], "US", "FR") + elif groupby == "function": + expected["level_0"] = expected["level_0"] == 1 + else: + expected["level_0"] = np.where(expected["level_0"], "US", "FR") + tm.assert_frame_equal(result, expected) + else: + # compare against SeriesGroupBy value_counts + education_df["both"] = education_df["gender"] + "-" + education_df["education"] + expected = gp["both"].value_counts( + normalize=normalize, sort=sort, ascending=ascending + ) + expected.name = name + if as_index: + index_frame = expected.index.to_frame(index=False) + index_frame["gender"] = index_frame["both"].str.split("-").str.get(0) + index_frame["education"] = index_frame["both"].str.split("-").str.get(1) + del index_frame["both"] + index_frame = index_frame.rename({0: None}, axis=1) + expected.index = MultiIndex.from_frame(index_frame) + tm.assert_series_equal(result, expected) + else: + expected.insert(1, "gender", expected["both"].str.split("-").str.get(0)) + expected.insert(2, "education", expected["both"].str.split("-").str.get(1)) + del expected["both"] + tm.assert_frame_equal(result, expected) + + +@pytest.mark.parametrize( + "dtype", + [ + object, + pytest.param("string[pyarrow_numpy]", marks=td.skip_if_no("pyarrow")), + pytest.param("string[pyarrow]", marks=td.skip_if_no("pyarrow")), + ], +) +@pytest.mark.parametrize("normalize", [True, False]) +@pytest.mark.parametrize( + "sort, ascending, expected_rows, expected_count, expected_group_size", + [ + (False, None, [0, 1, 2, 3, 4], [1, 1, 1, 2, 1], [1, 3, 1, 3, 1]), + (True, False, [3, 0, 1, 2, 4], [2, 1, 1, 1, 1], [3, 1, 3, 1, 1]), + (True, True, [0, 1, 2, 4, 3], [1, 1, 1, 1, 2], [1, 3, 1, 1, 3]), + ], +) +def test_compound( + education_df, + normalize, + sort, + ascending, + expected_rows, + expected_count, + expected_group_size, + dtype, +): + education_df = education_df.astype(dtype) + education_df.columns = education_df.columns.astype(dtype) + # Multiple groupby keys and as_index=False + gp = education_df.groupby(["country", "gender"], as_index=False, sort=False) + result = gp["education"].value_counts( + normalize=normalize, sort=sort, ascending=ascending + ) + expected = DataFrame() + for column in ["country", "gender", "education"]: + expected[column] = [education_df[column][row] for row in expected_rows] + expected = expected.astype(dtype) + expected.columns = expected.columns.astype(dtype) + if normalize: + expected["proportion"] = expected_count + expected["proportion"] /= expected_group_size + if dtype == "string[pyarrow]": + expected["proportion"] = expected["proportion"].convert_dtypes() + else: + expected["count"] = expected_count + if dtype == "string[pyarrow]": + expected["count"] = expected["count"].convert_dtypes() + tm.assert_frame_equal(result, expected) + + +@pytest.fixture +def animals_df(): + return DataFrame( + {"key": [1, 1, 1, 1], "num_legs": [2, 4, 4, 6], "num_wings": [2, 0, 0, 0]}, + index=["falcon", "dog", "cat", "ant"], + ) + + +@pytest.mark.parametrize( + "sort, ascending, normalize, name, expected_data, expected_index", + [ + (False, None, False, "count", [1, 2, 1], [(1, 1, 1), (2, 4, 6), (2, 0, 0)]), + (True, True, False, "count", [1, 1, 2], [(1, 1, 1), (2, 6, 4), (2, 0, 0)]), + (True, False, False, "count", [2, 1, 1], [(1, 1, 1), (4, 2, 6), (0, 2, 0)]), + ( + True, + False, + True, + "proportion", + [0.5, 0.25, 0.25], + [(1, 1, 1), (4, 2, 6), (0, 2, 0)], + ), + ], +) +def test_data_frame_value_counts( + animals_df, sort, ascending, normalize, name, expected_data, expected_index +): + # 3-way compare with :meth:`~DataFrame.value_counts` + # Tests from frame/methods/test_value_counts.py + result_frame = animals_df.value_counts( + sort=sort, ascending=ascending, normalize=normalize + ) + expected = Series( + data=expected_data, + index=MultiIndex.from_arrays( + expected_index, names=["key", "num_legs", "num_wings"] + ), + name=name, + ) + tm.assert_series_equal(result_frame, expected) + + result_frame_groupby = animals_df.groupby("key").value_counts( + sort=sort, ascending=ascending, normalize=normalize + ) + + tm.assert_series_equal(result_frame_groupby, expected) + + +@pytest.fixture +def nulls_df(): + n = np.nan + return DataFrame( + { + "A": [1, 1, n, 4, n, 6, 6, 6, 6], + "B": [1, 1, 3, n, n, 6, 6, 6, 6], + "C": [1, 2, 3, 4, 5, 6, n, 8, n], + "D": [1, 2, 3, 4, 5, 6, 7, n, n], + } + ) + + +@pytest.mark.parametrize( + "group_dropna, count_dropna, expected_rows, expected_values", + [ + ( + False, + False, + [0, 1, 3, 5, 7, 6, 8, 2, 4], + [0.5, 0.5, 1.0, 0.25, 0.25, 0.25, 0.25, 1.0, 1.0], + ), + (False, True, [0, 1, 3, 5, 2, 4], [0.5, 0.5, 1.0, 1.0, 1.0, 1.0]), + (True, False, [0, 1, 5, 7, 6, 8], [0.5, 0.5, 0.25, 0.25, 0.25, 0.25]), + (True, True, [0, 1, 5], [0.5, 0.5, 1.0]), + ], +) +def test_dropna_combinations( + nulls_df, group_dropna, count_dropna, expected_rows, expected_values, request +): + if Version(np.__version__) >= Version("1.25") and not group_dropna: + request.applymarker( + pytest.mark.xfail( + reason=( + "pandas default unstable sorting of duplicates" + "issue with numpy>=1.25 with AVX instructions" + ), + strict=False, + ) + ) + gp = nulls_df.groupby(["A", "B"], dropna=group_dropna) + result = gp.value_counts(normalize=True, sort=True, dropna=count_dropna) + columns = DataFrame() + for column in nulls_df.columns: + columns[column] = [nulls_df[column][row] for row in expected_rows] + index = MultiIndex.from_frame(columns) + expected = Series(data=expected_values, index=index, name="proportion") + tm.assert_series_equal(result, expected) + + +@pytest.fixture +def names_with_nulls_df(nulls_fixture): + return DataFrame( + { + "key": [1, 1, 1, 1], + "first_name": ["John", "Anne", "John", "Beth"], + "middle_name": ["Smith", nulls_fixture, nulls_fixture, "Louise"], + }, + ) + + +@pytest.mark.parametrize( + "dropna, expected_data, expected_index", + [ + ( + True, + [1, 1], + MultiIndex.from_arrays( + [(1, 1), ("Beth", "John"), ("Louise", "Smith")], + names=["key", "first_name", "middle_name"], + ), + ), + ( + False, + [1, 1, 1, 1], + MultiIndex( + levels=[ + Index([1]), + Index(["Anne", "Beth", "John"]), + Index(["Louise", "Smith", np.nan]), + ], + codes=[[0, 0, 0, 0], [0, 1, 2, 2], [2, 0, 1, 2]], + names=["key", "first_name", "middle_name"], + ), + ), + ], +) +@pytest.mark.parametrize("normalize, name", [(False, "count"), (True, "proportion")]) +def test_data_frame_value_counts_dropna( + names_with_nulls_df, dropna, normalize, name, expected_data, expected_index +): + # GH 41334 + # 3-way compare with :meth:`~DataFrame.value_counts` + # Tests with nulls from frame/methods/test_value_counts.py + result_frame = names_with_nulls_df.value_counts(dropna=dropna, normalize=normalize) + expected = Series( + data=expected_data, + index=expected_index, + name=name, + ) + if normalize: + expected /= float(len(expected_data)) + + tm.assert_series_equal(result_frame, expected) + + result_frame_groupby = names_with_nulls_df.groupby("key").value_counts( + dropna=dropna, normalize=normalize + ) + + tm.assert_series_equal(result_frame_groupby, expected) + + +@pytest.mark.parametrize("as_index", [False, True]) +@pytest.mark.parametrize("observed", [False, True]) +@pytest.mark.parametrize( + "normalize, name, expected_data", + [ + ( + False, + "count", + np.array([2, 1, 1, 0, 0, 0, 1, 1, 0, 0, 0, 0], dtype=np.int64), + ), + ( + True, + "proportion", + np.array([0.5, 0.25, 0.25, 0.0, 0.0, 0.0, 0.5, 0.5, 0.0, 0.0, 0.0, 0.0]), + ), + ], +) +def test_categorical_single_grouper_with_only_observed_categories( + education_df, as_index, observed, normalize, name, expected_data, request +): + # Test single categorical grouper with only observed grouping categories + # when non-groupers are also categorical + if Version(np.__version__) >= Version("1.25"): + request.applymarker( + pytest.mark.xfail( + reason=( + "pandas default unstable sorting of duplicates" + "issue with numpy>=1.25 with AVX instructions" + ), + strict=False, + ) + ) + + gp = education_df.astype("category").groupby( + "country", as_index=as_index, observed=observed + ) + result = gp.value_counts(normalize=normalize) + + expected_index = MultiIndex.from_tuples( + [ + ("FR", "male", "low"), + ("FR", "female", "high"), + ("FR", "male", "medium"), + ("FR", "female", "low"), + ("FR", "female", "medium"), + ("FR", "male", "high"), + ("US", "female", "high"), + ("US", "male", "low"), + ("US", "female", "low"), + ("US", "female", "medium"), + ("US", "male", "high"), + ("US", "male", "medium"), + ], + names=["country", "gender", "education"], + ) + + expected_series = Series( + data=expected_data, + index=expected_index, + name=name, + ) + for i in range(3): + expected_series.index = expected_series.index.set_levels( + CategoricalIndex(expected_series.index.levels[i]), level=i + ) + + if as_index: + tm.assert_series_equal(result, expected_series) + else: + expected = expected_series.reset_index( + name="proportion" if normalize else "count" + ) + tm.assert_frame_equal(result, expected) + + +def assert_categorical_single_grouper( + education_df, as_index, observed, expected_index, normalize, name, expected_data +): + # Test single categorical grouper when non-groupers are also categorical + education_df = education_df.copy().astype("category") + + # Add non-observed grouping categories + education_df["country"] = education_df["country"].cat.add_categories(["ASIA"]) + + gp = education_df.groupby("country", as_index=as_index, observed=observed) + result = gp.value_counts(normalize=normalize) + + expected_series = Series( + data=expected_data, + index=MultiIndex.from_tuples( + expected_index, + names=["country", "gender", "education"], + ), + name=name, + ) + for i in range(3): + index_level = CategoricalIndex(expected_series.index.levels[i]) + if i == 0: + index_level = index_level.set_categories( + education_df["country"].cat.categories + ) + expected_series.index = expected_series.index.set_levels(index_level, level=i) + + if as_index: + tm.assert_series_equal(result, expected_series) + else: + expected = expected_series.reset_index(name=name) + tm.assert_frame_equal(result, expected) + + +@pytest.mark.parametrize("as_index", [True, False]) +@pytest.mark.parametrize( + "normalize, name, expected_data", + [ + ( + False, + "count", + np.array([2, 1, 1, 0, 0, 0, 1, 1, 0, 0, 0, 0], dtype=np.int64), + ), + ( + True, + "proportion", + np.array([0.5, 0.25, 0.25, 0.0, 0.0, 0.0, 0.5, 0.5, 0.0, 0.0, 0.0, 0.0]), + ), + ], +) +def test_categorical_single_grouper_observed_true( + education_df, as_index, normalize, name, expected_data, request +): + # GH#46357 + + if Version(np.__version__) >= Version("1.25"): + request.applymarker( + pytest.mark.xfail( + reason=( + "pandas default unstable sorting of duplicates" + "issue with numpy>=1.25 with AVX instructions" + ), + strict=False, + ) + ) + + expected_index = [ + ("FR", "male", "low"), + ("FR", "female", "high"), + ("FR", "male", "medium"), + ("FR", "female", "low"), + ("FR", "female", "medium"), + ("FR", "male", "high"), + ("US", "female", "high"), + ("US", "male", "low"), + ("US", "female", "low"), + ("US", "female", "medium"), + ("US", "male", "high"), + ("US", "male", "medium"), + ] + + assert_categorical_single_grouper( + education_df=education_df, + as_index=as_index, + observed=True, + expected_index=expected_index, + normalize=normalize, + name=name, + expected_data=expected_data, + ) + + +@pytest.mark.parametrize("as_index", [True, False]) +@pytest.mark.parametrize( + "normalize, name, expected_data", + [ + ( + False, + "count", + np.array( + [2, 1, 1, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], dtype=np.int64 + ), + ), + ( + True, + "proportion", + np.array( + [ + 0.5, + 0.25, + 0.25, + 0.0, + 0.0, + 0.0, + 0.5, + 0.5, + 0.0, + 0.0, + 0.0, + 0.0, + 0.0, + 0.0, + 0.0, + 0.0, + 0.0, + 0.0, + ] + ), + ), + ], +) +def test_categorical_single_grouper_observed_false( + education_df, as_index, normalize, name, expected_data, request +): + # GH#46357 + + if Version(np.__version__) >= Version("1.25"): + request.applymarker( + pytest.mark.xfail( + reason=( + "pandas default unstable sorting of duplicates" + "issue with numpy>=1.25 with AVX instructions" + ), + strict=False, + ) + ) + + expected_index = [ + ("FR", "male", "low"), + ("FR", "female", "high"), + ("FR", "male", "medium"), + ("FR", "female", "low"), + ("FR", "female", "medium"), + ("FR", "male", "high"), + ("US", "female", "high"), + ("US", "male", "low"), + ("US", "female", "low"), + ("US", "female", "medium"), + ("US", "male", "high"), + ("US", "male", "medium"), + ("ASIA", "female", "high"), + ("ASIA", "female", "low"), + ("ASIA", "female", "medium"), + ("ASIA", "male", "high"), + ("ASIA", "male", "low"), + ("ASIA", "male", "medium"), + ] + + assert_categorical_single_grouper( + education_df=education_df, + as_index=as_index, + observed=False, + expected_index=expected_index, + normalize=normalize, + name=name, + expected_data=expected_data, + ) + + +@pytest.mark.parametrize("as_index", [True, False]) +@pytest.mark.parametrize( + "observed, expected_index", + [ + ( + False, + [ + ("FR", "high", "female"), + ("FR", "high", "male"), + ("FR", "low", "male"), + ("FR", "low", "female"), + ("FR", "medium", "male"), + ("FR", "medium", "female"), + ("US", "high", "female"), + ("US", "high", "male"), + ("US", "low", "male"), + ("US", "low", "female"), + ("US", "medium", "female"), + ("US", "medium", "male"), + ], + ), + ( + True, + [ + ("FR", "high", "female"), + ("FR", "low", "male"), + ("FR", "medium", "male"), + ("US", "high", "female"), + ("US", "low", "male"), + ], + ), + ], +) +@pytest.mark.parametrize( + "normalize, name, expected_data", + [ + ( + False, + "count", + np.array([1, 0, 2, 0, 1, 0, 1, 0, 1, 0, 0, 0], dtype=np.int64), + ), + ( + True, + "proportion", + # NaN values corresponds to non-observed groups + np.array([1.0, 0.0, 1.0, 0.0, 1.0, 0.0, 1.0, 0.0, 1.0, 0.0, 0.0, 0.0]), + ), + ], +) +def test_categorical_multiple_groupers( + education_df, as_index, observed, expected_index, normalize, name, expected_data +): + # GH#46357 + + # Test multiple categorical groupers when non-groupers are non-categorical + education_df = education_df.copy() + education_df["country"] = education_df["country"].astype("category") + education_df["education"] = education_df["education"].astype("category") + + gp = education_df.groupby( + ["country", "education"], as_index=as_index, observed=observed + ) + result = gp.value_counts(normalize=normalize) + + expected_series = Series( + data=expected_data[expected_data > 0.0] if observed else expected_data, + index=MultiIndex.from_tuples( + expected_index, + names=["country", "education", "gender"], + ), + name=name, + ) + for i in range(2): + expected_series.index = expected_series.index.set_levels( + CategoricalIndex(expected_series.index.levels[i]), level=i + ) + + if as_index: + tm.assert_series_equal(result, expected_series) + else: + expected = expected_series.reset_index( + name="proportion" if normalize else "count" + ) + tm.assert_frame_equal(result, expected) + + +@pytest.mark.parametrize("as_index", [False, True]) +@pytest.mark.parametrize("observed", [False, True]) +@pytest.mark.parametrize( + "normalize, name, expected_data", + [ + ( + False, + "count", + np.array([2, 1, 1, 0, 0, 0, 1, 1, 0, 0, 0, 0], dtype=np.int64), + ), + ( + True, + "proportion", + # NaN values corresponds to non-observed groups + np.array([0.5, 0.25, 0.25, 0.0, 0.0, 0.0, 0.5, 0.5, 0.0, 0.0, 0.0, 0.0]), + ), + ], +) +def test_categorical_non_groupers( + education_df, as_index, observed, normalize, name, expected_data, request +): + # GH#46357 Test non-observed categories are included in the result, + # regardless of `observed` + + if Version(np.__version__) >= Version("1.25"): + request.applymarker( + pytest.mark.xfail( + reason=( + "pandas default unstable sorting of duplicates" + "issue with numpy>=1.25 with AVX instructions" + ), + strict=False, + ) + ) + + education_df = education_df.copy() + education_df["gender"] = education_df["gender"].astype("category") + education_df["education"] = education_df["education"].astype("category") + + gp = education_df.groupby("country", as_index=as_index, observed=observed) + result = gp.value_counts(normalize=normalize) + + expected_index = [ + ("FR", "male", "low"), + ("FR", "female", "high"), + ("FR", "male", "medium"), + ("FR", "female", "low"), + ("FR", "female", "medium"), + ("FR", "male", "high"), + ("US", "female", "high"), + ("US", "male", "low"), + ("US", "female", "low"), + ("US", "female", "medium"), + ("US", "male", "high"), + ("US", "male", "medium"), + ] + expected_series = Series( + data=expected_data, + index=MultiIndex.from_tuples( + expected_index, + names=["country", "gender", "education"], + ), + name=name, + ) + for i in range(1, 3): + expected_series.index = expected_series.index.set_levels( + CategoricalIndex(expected_series.index.levels[i]), level=i + ) + + if as_index: + tm.assert_series_equal(result, expected_series) + else: + expected = expected_series.reset_index( + name="proportion" if normalize else "count" + ) + tm.assert_frame_equal(result, expected) + + +@pytest.mark.parametrize( + "normalize, expected_label, expected_values", + [ + (False, "count", [1, 1, 1]), + (True, "proportion", [0.5, 0.5, 1.0]), + ], +) +def test_mixed_groupings(normalize, expected_label, expected_values): + # Test multiple groupings + df = DataFrame({"A": [1, 2, 1], "B": [1, 2, 3]}) + gp = df.groupby([[4, 5, 4], "A", lambda i: 7 if i == 1 else 8], as_index=False) + result = gp.value_counts(sort=True, normalize=normalize) + expected = DataFrame( + { + "level_0": np.array([4, 4, 5], dtype=int), + "A": [1, 1, 2], + "level_2": [8, 8, 7], + "B": [1, 3, 2], + expected_label: expected_values, + } + ) + tm.assert_frame_equal(result, expected) + + +@pytest.mark.parametrize( + "test, columns, expected_names", + [ + ("repeat", list("abbde"), ["a", None, "d", "b", "b", "e"]), + ("level", list("abcd") + ["level_1"], ["a", None, "d", "b", "c", "level_1"]), + ], +) +@pytest.mark.parametrize("as_index", [False, True]) +def test_column_label_duplicates(test, columns, expected_names, as_index): + # GH 44992 + # Test for duplicate input column labels and generated duplicate labels + df = DataFrame([[1, 3, 5, 7, 9], [2, 4, 6, 8, 10]], columns=columns) + expected_data = [(1, 0, 7, 3, 5, 9), (2, 1, 8, 4, 6, 10)] + keys = ["a", np.array([0, 1], dtype=np.int64), "d"] + result = df.groupby(keys, as_index=as_index).value_counts() + if as_index: + expected = Series( + data=(1, 1), + index=MultiIndex.from_tuples( + expected_data, + names=expected_names, + ), + name="count", + ) + tm.assert_series_equal(result, expected) + else: + expected_data = [list(row) + [1] for row in expected_data] + expected_columns = list(expected_names) + expected_columns[1] = "level_1" + expected_columns.append("count") + expected = DataFrame(expected_data, columns=expected_columns) + tm.assert_frame_equal(result, expected) + + +@pytest.mark.parametrize( + "normalize, expected_label", + [ + (False, "count"), + (True, "proportion"), + ], +) +def test_result_label_duplicates(normalize, expected_label): + # Test for result column label duplicating an input column label + gb = DataFrame([[1, 2, 3]], columns=["a", "b", expected_label]).groupby( + "a", as_index=False + ) + msg = f"Column label '{expected_label}' is duplicate of result column" + with pytest.raises(ValueError, match=msg): + gb.value_counts(normalize=normalize) + + +def test_ambiguous_grouping(): + # Test that groupby is not confused by groupings length equal to row count + df = DataFrame({"a": [1, 1]}) + gb = df.groupby(np.array([1, 1], dtype=np.int64)) + result = gb.value_counts() + expected = Series( + [2], index=MultiIndex.from_tuples([[1, 1]], names=[None, "a"]), name="count" + ) + tm.assert_series_equal(result, expected) + + +def test_subset_overlaps_gb_key_raises(): + # GH 46383 + df = DataFrame({"c1": ["a", "b", "c"], "c2": ["x", "y", "y"]}, index=[0, 1, 1]) + msg = "Keys {'c1'} in subset cannot be in the groupby column keys." + with pytest.raises(ValueError, match=msg): + df.groupby("c1").value_counts(subset=["c1"]) + + +def test_subset_doesnt_exist_in_frame(): + # GH 46383 + df = DataFrame({"c1": ["a", "b", "c"], "c2": ["x", "y", "y"]}, index=[0, 1, 1]) + msg = "Keys {'c3'} in subset do not exist in the DataFrame." + with pytest.raises(ValueError, match=msg): + df.groupby("c1").value_counts(subset=["c3"]) + + +def test_subset(): + # GH 46383 + df = DataFrame({"c1": ["a", "b", "c"], "c2": ["x", "y", "y"]}, index=[0, 1, 1]) + result = df.groupby(level=0).value_counts(subset=["c2"]) + expected = Series( + [1, 2], + index=MultiIndex.from_arrays([[0, 1], ["x", "y"]], names=[None, "c2"]), + name="count", + ) + tm.assert_series_equal(result, expected) + + +def test_subset_duplicate_columns(): + # GH 46383 + df = DataFrame( + [["a", "x", "x"], ["b", "y", "y"], ["b", "y", "y"]], + index=[0, 1, 1], + columns=["c1", "c2", "c2"], + ) + result = df.groupby(level=0).value_counts(subset=["c2"]) + expected = Series( + [1, 2], + index=MultiIndex.from_arrays( + [[0, 1], ["x", "y"], ["x", "y"]], names=[None, "c2", "c2"] + ), + name="count", + ) + tm.assert_series_equal(result, expected) + + +@pytest.mark.parametrize("utc", [True, False]) +def test_value_counts_time_grouper(utc, unit): + # GH#50486 + df = DataFrame( + { + "Timestamp": [ + 1565083561, + 1565083561 + 86400, + 1565083561 + 86500, + 1565083561 + 86400 * 2, + 1565083561 + 86400 * 3, + 1565083561 + 86500 * 3, + 1565083561 + 86400 * 4, + ], + "Food": ["apple", "apple", "banana", "banana", "orange", "orange", "pear"], + } + ).drop([3]) + + df["Datetime"] = to_datetime(df["Timestamp"], utc=utc, unit="s").dt.as_unit(unit) + gb = df.groupby(Grouper(freq="1D", key="Datetime")) + result = gb.value_counts() + dates = to_datetime( + ["2019-08-06", "2019-08-07", "2019-08-09", "2019-08-10"], utc=utc + ).as_unit(unit) + timestamps = df["Timestamp"].unique() + index = MultiIndex( + levels=[dates, timestamps, ["apple", "banana", "orange", "pear"]], + codes=[[0, 1, 1, 2, 2, 3], range(6), [0, 0, 1, 2, 2, 3]], + names=["Datetime", "Timestamp", "Food"], + ) + expected = Series(1, index=index, name="count") + tm.assert_series_equal(result, expected) + + +def test_value_counts_integer_columns(): + # GH#55627 + df = DataFrame({1: ["a", "a", "a"], 2: ["a", "a", "d"], 3: ["a", "b", "c"]}) + gp = df.groupby([1, 2], as_index=False, sort=False) + result = gp[3].value_counts() + expected = DataFrame( + {1: ["a", "a", "a"], 2: ["a", "a", "d"], 3: ["a", "b", "c"], "count": 1} + ) + tm.assert_frame_equal(result, expected) + + +@pytest.mark.parametrize("vc_sort", [True, False]) +@pytest.mark.parametrize("normalize", [True, False]) +def test_value_counts_sort(sort, vc_sort, normalize): + # GH#55951 + df = DataFrame({"a": [2, 1, 1, 1], 0: [3, 4, 3, 3]}) + gb = df.groupby("a", sort=sort) + result = gb.value_counts(sort=vc_sort, normalize=normalize) + + if normalize: + values = [2 / 3, 1 / 3, 1.0] + else: + values = [2, 1, 1] + index = MultiIndex( + levels=[[1, 2], [3, 4]], codes=[[0, 0, 1], [0, 1, 0]], names=["a", 0] + ) + expected = Series(values, index=index, name="proportion" if normalize else "count") + if sort and vc_sort: + taker = [0, 1, 2] + elif sort and not vc_sort: + taker = [0, 1, 2] + elif not sort and vc_sort: + taker = [0, 2, 1] + else: + taker = [2, 1, 0] + expected = expected.take(taker) + + tm.assert_series_equal(result, expected) + + +@pytest.mark.parametrize("vc_sort", [True, False]) +@pytest.mark.parametrize("normalize", [True, False]) +def test_value_counts_sort_categorical(sort, vc_sort, normalize): + # GH#55951 + df = DataFrame({"a": [2, 1, 1, 1], 0: [3, 4, 3, 3]}, dtype="category") + gb = df.groupby("a", sort=sort, observed=True) + result = gb.value_counts(sort=vc_sort, normalize=normalize) + + if normalize: + values = [2 / 3, 1 / 3, 1.0, 0.0] + else: + values = [2, 1, 1, 0] + name = "proportion" if normalize else "count" + expected = DataFrame( + { + "a": Categorical([1, 1, 2, 2]), + 0: Categorical([3, 4, 3, 4]), + name: values, + } + ).set_index(["a", 0])[name] + if sort and vc_sort: + taker = [0, 1, 2, 3] + elif sort and not vc_sort: + taker = [0, 1, 2, 3] + elif not sort and vc_sort: + taker = [0, 2, 1, 3] + else: + taker = [2, 3, 0, 1] + expected = expected.take(taker) + + tm.assert_series_equal(result, expected) diff --git a/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/test_index_as_string.py b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/test_index_as_string.py new file mode 100644 index 0000000000000000000000000000000000000000..4aaf3de9a23b2416603947db312bb49eea343ba8 --- /dev/null +++ b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/test_index_as_string.py @@ -0,0 +1,85 @@ +import numpy as np +import pytest + +import pandas as pd +import pandas._testing as tm + + +@pytest.fixture(params=[["inner"], ["inner", "outer"]]) +def frame(request): + levels = request.param + df = pd.DataFrame( + { + "outer": ["a", "a", "a", "b", "b", "b"], + "inner": [1, 2, 3, 1, 2, 3], + "A": np.arange(6), + "B": ["one", "one", "two", "two", "one", "one"], + } + ) + if levels: + df = df.set_index(levels) + + return df + + +@pytest.fixture() +def series(): + df = pd.DataFrame( + { + "outer": ["a", "a", "a", "b", "b", "b"], + "inner": [1, 2, 3, 1, 2, 3], + "A": np.arange(6), + "B": ["one", "one", "two", "two", "one", "one"], + } + ) + s = df.set_index(["outer", "inner", "B"])["A"] + + return s + + +@pytest.mark.parametrize( + "key_strs,groupers", + [ + ("inner", pd.Grouper(level="inner")), # Index name + (["inner"], [pd.Grouper(level="inner")]), # List of index name + (["B", "inner"], ["B", pd.Grouper(level="inner")]), # Column and index + (["inner", "B"], [pd.Grouper(level="inner"), "B"]), # Index and column + ], +) +def test_grouper_index_level_as_string(frame, key_strs, groupers): + if "B" not in key_strs or "outer" in frame.columns: + result = frame.groupby(key_strs).mean(numeric_only=True) + expected = frame.groupby(groupers).mean(numeric_only=True) + else: + result = frame.groupby(key_strs).mean() + expected = frame.groupby(groupers).mean() + tm.assert_frame_equal(result, expected) + + +@pytest.mark.parametrize( + "levels", + [ + "inner", + "outer", + "B", + ["inner"], + ["outer"], + ["B"], + ["inner", "outer"], + ["outer", "inner"], + ["inner", "outer", "B"], + ["B", "outer", "inner"], + ], +) +def test_grouper_index_level_as_string_series(series, levels): + # Compute expected result + if isinstance(levels, list): + groupers = [pd.Grouper(level=lv) for lv in levels] + else: + groupers = pd.Grouper(level=levels) + + expected = series.groupby(groupers).mean() + + # Compute and check result + result = series.groupby(levels).mean() + tm.assert_series_equal(result, expected) diff --git a/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/test_raises.py b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/test_raises.py new file mode 100644 index 0000000000000000000000000000000000000000..0b451ce73db898f02bce2d1432ca64c0011f3e71 --- /dev/null +++ b/llava_next/lib/python3.10/site-packages/pandas/tests/groupby/test_raises.py @@ -0,0 +1,716 @@ +# Only tests that raise an error and have no better location should go here. +# Tests for specific groupby methods should go in their respective +# test file. + +import datetime +import re + +import numpy as np +import pytest + +from pandas import ( + Categorical, + DataFrame, + Grouper, + Series, +) +import pandas._testing as tm +from pandas.tests.groupby import get_groupby_method_args + + +@pytest.fixture( + params=[ + "a", + ["a"], + ["a", "b"], + Grouper(key="a"), + lambda x: x % 2, + [0, 0, 0, 1, 2, 2, 2, 3, 3], + np.array([0, 0, 0, 1, 2, 2, 2, 3, 3]), + dict(zip(range(9), [0, 0, 0, 1, 2, 2, 2, 3, 3])), + Series([1, 1, 1, 1, 1, 2, 2, 2, 2]), + [Series([1, 1, 1, 1, 1, 2, 2, 2, 2]), Series([3, 3, 4, 4, 4, 4, 4, 3, 3])], + ] +) +def by(request): + return request.param + + +@pytest.fixture(params=[True, False]) +def groupby_series(request): + return request.param + + +@pytest.fixture +def df_with_string_col(): + df = DataFrame( + { + "a": [1, 1, 1, 1, 1, 2, 2, 2, 2], + "b": [3, 3, 4, 4, 4, 4, 4, 3, 3], + "c": range(9), + "d": list("xyzwtyuio"), + } + ) + return df + + +@pytest.fixture +def df_with_datetime_col(): + df = DataFrame( + { + "a": [1, 1, 1, 1, 1, 2, 2, 2, 2], + "b": [3, 3, 4, 4, 4, 4, 4, 3, 3], + "c": range(9), + "d": datetime.datetime(2005, 1, 1, 10, 30, 23, 540000), + } + ) + return df + + +@pytest.fixture +def df_with_timedelta_col(): + df = DataFrame( + { + "a": [1, 1, 1, 1, 1, 2, 2, 2, 2], + "b": [3, 3, 4, 4, 4, 4, 4, 3, 3], + "c": range(9), + "d": datetime.timedelta(days=1), + } + ) + return df + + +@pytest.fixture +def df_with_cat_col(): + df = DataFrame( + { + "a": [1, 1, 1, 1, 1, 2, 2, 2, 2], + "b": [3, 3, 4, 4, 4, 4, 4, 3, 3], + "c": range(9), + "d": Categorical( + ["a", "a", "a", "a", "b", "b", "b", "b", "c"], + categories=["a", "b", "c", "d"], + ordered=True, + ), + } + ) + return df + + +def _call_and_check(klass, msg, how, gb, groupby_func, args, warn_msg=""): + warn_klass = None if warn_msg == "" else FutureWarning + with tm.assert_produces_warning(warn_klass, match=warn_msg): + if klass is None: + if how == "method": + getattr(gb, groupby_func)(*args) + elif how == "agg": + gb.agg(groupby_func, *args) + else: + gb.transform(groupby_func, *args) + else: + with pytest.raises(klass, match=msg): + if how == "method": + getattr(gb, groupby_func)(*args) + elif how == "agg": + gb.agg(groupby_func, *args) + else: + gb.transform(groupby_func, *args) + + +@pytest.mark.parametrize("how", ["method", "agg", "transform"]) +def test_groupby_raises_string( + how, by, groupby_series, groupby_func, df_with_string_col +): + df = df_with_string_col + args = get_groupby_method_args(groupby_func, df) + gb = df.groupby(by=by) + + if groupby_series: + gb = gb["d"] + + if groupby_func == "corrwith": + assert not hasattr(gb, "corrwith") + return + + klass, msg = { + "all": (None, ""), + "any": (None, ""), + "bfill": (None, ""), + "corrwith": (TypeError, "Could not convert"), + "count": (None, ""), + "cumcount": (None, ""), + "cummax": ( + (NotImplementedError, TypeError), + "(function|cummax) is not (implemented|supported) for (this|object) dtype", + ), + "cummin": ( + (NotImplementedError, TypeError), + "(function|cummin) is not (implemented|supported) for (this|object) dtype", + ), + "cumprod": ( + (NotImplementedError, TypeError), + "(function|cumprod) is not (implemented|supported) for (this|object) dtype", + ), + "cumsum": ( + (NotImplementedError, TypeError), + "(function|cumsum) is not (implemented|supported) for (this|object) dtype", + ), + "diff": (TypeError, "unsupported operand type"), + "ffill": (None, ""), + "fillna": (None, ""), + "first": (None, ""), + "idxmax": (None, ""), + "idxmin": (None, ""), + "last": (None, ""), + "max": (None, ""), + "mean": ( + TypeError, + re.escape("agg function failed [how->mean,dtype->object]"), + ), + "median": ( + TypeError, + re.escape("agg function failed [how->median,dtype->object]"), + ), + "min": (None, ""), + "ngroup": (None, ""), + "nunique": (None, ""), + "pct_change": (TypeError, "unsupported operand type"), + "prod": ( + TypeError, + re.escape("agg function failed [how->prod,dtype->object]"), + ), + "quantile": (TypeError, "cannot be performed against 'object' dtypes!"), + "rank": (None, ""), + "sem": (ValueError, "could not convert string to float"), + "shift": (None, ""), + "size": (None, ""), + "skew": (ValueError, "could not convert string to float"), + "std": (ValueError, "could not convert string to float"), + "sum": (None, ""), + "var": ( + TypeError, + re.escape("agg function failed [how->var,dtype->"), + ), + }[groupby_func] + + if groupby_func == "fillna": + kind = "Series" if groupby_series else "DataFrame" + warn_msg = f"{kind}GroupBy.fillna is deprecated" + else: + warn_msg = "" + _call_and_check(klass, msg, how, gb, groupby_func, args, warn_msg) + + +@pytest.mark.parametrize("how", ["agg", "transform"]) +def test_groupby_raises_string_udf(how, by, groupby_series, df_with_string_col): + df = df_with_string_col + gb = df.groupby(by=by) + + if groupby_series: + gb = gb["d"] + + def func(x): + raise TypeError("Test error message") + + with pytest.raises(TypeError, match="Test error message"): + getattr(gb, how)(func) + + +@pytest.mark.parametrize("how", ["agg", "transform"]) +@pytest.mark.parametrize("groupby_func_np", [np.sum, np.mean]) +def test_groupby_raises_string_np( + how, by, groupby_series, groupby_func_np, df_with_string_col +): + # GH#50749 + df = df_with_string_col + gb = df.groupby(by=by) + + if groupby_series: + gb = gb["d"] + + klass, msg = { + np.sum: (None, ""), + np.mean: ( + TypeError, + re.escape("agg function failed [how->mean,dtype->object]"), + ), + }[groupby_func_np] + + if groupby_series: + warn_msg = "using SeriesGroupBy.[sum|mean]" + else: + warn_msg = "using DataFrameGroupBy.[sum|mean]" + _call_and_check(klass, msg, how, gb, groupby_func_np, (), warn_msg=warn_msg) + + +@pytest.mark.parametrize("how", ["method", "agg", "transform"]) +def test_groupby_raises_datetime( + how, by, groupby_series, groupby_func, df_with_datetime_col +): + df = df_with_datetime_col + args = get_groupby_method_args(groupby_func, df) + gb = df.groupby(by=by) + + if groupby_series: + gb = gb["d"] + + if groupby_func == "corrwith": + assert not hasattr(gb, "corrwith") + return + + klass, msg = { + "all": (None, ""), + "any": (None, ""), + "bfill": (None, ""), + "corrwith": (TypeError, "cannot perform __mul__ with this index type"), + "count": (None, ""), + "cumcount": (None, ""), + "cummax": (None, ""), + "cummin": (None, ""), + "cumprod": (TypeError, "datetime64 type does not support cumprod operations"), + "cumsum": (TypeError, "datetime64 type does not support cumsum operations"), + "diff": (None, ""), + "ffill": (None, ""), + "fillna": (None, ""), + "first": (None, ""), + "idxmax": (None, ""), + "idxmin": (None, ""), + "last": (None, ""), + "max": (None, ""), + "mean": (None, ""), + "median": (None, ""), + "min": (None, ""), + "ngroup": (None, ""), + "nunique": (None, ""), + "pct_change": (TypeError, "cannot perform __truediv__ with this index type"), + "prod": (TypeError, "datetime64 type does not support prod"), + "quantile": (None, ""), + "rank": (None, ""), + "sem": (None, ""), + "shift": (None, ""), + "size": (None, ""), + "skew": ( + TypeError, + "|".join( + [ + r"dtype datetime64\[ns\] does not support reduction", + "datetime64 type does not support skew operations", + ] + ), + ), + "std": (None, ""), + "sum": (TypeError, "datetime64 type does not support sum operations"), + "var": (TypeError, "datetime64 type does not support var operations"), + }[groupby_func] + + if groupby_func in ["any", "all"]: + warn_msg = f"'{groupby_func}' with datetime64 dtypes is deprecated" + elif groupby_func == "fillna": + kind = "Series" if groupby_series else "DataFrame" + warn_msg = f"{kind}GroupBy.fillna is deprecated" + else: + warn_msg = "" + _call_and_check(klass, msg, how, gb, groupby_func, args, warn_msg=warn_msg) + + +@pytest.mark.parametrize("how", ["agg", "transform"]) +def test_groupby_raises_datetime_udf(how, by, groupby_series, df_with_datetime_col): + df = df_with_datetime_col + gb = df.groupby(by=by) + + if groupby_series: + gb = gb["d"] + + def func(x): + raise TypeError("Test error message") + + with pytest.raises(TypeError, match="Test error message"): + getattr(gb, how)(func) + + +@pytest.mark.parametrize("how", ["agg", "transform"]) +@pytest.mark.parametrize("groupby_func_np", [np.sum, np.mean]) +def test_groupby_raises_datetime_np( + how, by, groupby_series, groupby_func_np, df_with_datetime_col +): + # GH#50749 + df = df_with_datetime_col + gb = df.groupby(by=by) + + if groupby_series: + gb = gb["d"] + + klass, msg = { + np.sum: (TypeError, "datetime64 type does not support sum operations"), + np.mean: (None, ""), + }[groupby_func_np] + + if groupby_series: + warn_msg = "using SeriesGroupBy.[sum|mean]" + else: + warn_msg = "using DataFrameGroupBy.[sum|mean]" + _call_and_check(klass, msg, how, gb, groupby_func_np, (), warn_msg=warn_msg) + + +@pytest.mark.parametrize("func", ["prod", "cumprod", "skew", "var"]) +def test_groupby_raises_timedelta(func, df_with_timedelta_col): + df = df_with_timedelta_col + gb = df.groupby(by="a") + + _call_and_check( + TypeError, + "timedelta64 type does not support .* operations", + "method", + gb, + func, + [], + ) + + +@pytest.mark.parametrize("how", ["method", "agg", "transform"]) +def test_groupby_raises_category( + how, by, groupby_series, groupby_func, using_copy_on_write, df_with_cat_col +): + # GH#50749 + df = df_with_cat_col + args = get_groupby_method_args(groupby_func, df) + gb = df.groupby(by=by) + + if groupby_series: + gb = gb["d"] + + if groupby_func == "corrwith": + assert not hasattr(gb, "corrwith") + return + + klass, msg = { + "all": (None, ""), + "any": (None, ""), + "bfill": (None, ""), + "corrwith": ( + TypeError, + r"unsupported operand type\(s\) for \*: 'Categorical' and 'int'", + ), + "count": (None, ""), + "cumcount": (None, ""), + "cummax": ( + (NotImplementedError, TypeError), + "(category type does not support cummax operations|" + "category dtype not supported|" + "cummax is not supported for category dtype)", + ), + "cummin": ( + (NotImplementedError, TypeError), + "(category type does not support cummin operations|" + "category dtype not supported|" + "cummin is not supported for category dtype)", + ), + "cumprod": ( + (NotImplementedError, TypeError), + "(category type does not support cumprod operations|" + "category dtype not supported|" + "cumprod is not supported for category dtype)", + ), + "cumsum": ( + (NotImplementedError, TypeError), + "(category type does not support cumsum operations|" + "category dtype not supported|" + "cumsum is not supported for category dtype)", + ), + "diff": ( + TypeError, + r"unsupported operand type\(s\) for -: 'Categorical' and 'Categorical'", + ), + "ffill": (None, ""), + "fillna": ( + TypeError, + r"Cannot setitem on a Categorical with a new category \(0\), " + "set the categories first", + ) + if not using_copy_on_write + else (None, ""), # no-op with CoW + "first": (None, ""), + "idxmax": (None, ""), + "idxmin": (None, ""), + "last": (None, ""), + "max": (None, ""), + "mean": ( + TypeError, + "|".join( + [ + "'Categorical' .* does not support reduction 'mean'", + "category dtype does not support aggregation 'mean'", + ] + ), + ), + "median": ( + TypeError, + "|".join( + [ + "'Categorical' .* does not support reduction 'median'", + "category dtype does not support aggregation 'median'", + ] + ), + ), + "min": (None, ""), + "ngroup": (None, ""), + "nunique": (None, ""), + "pct_change": ( + TypeError, + r"unsupported operand type\(s\) for /: 'Categorical' and 'Categorical'", + ), + "prod": (TypeError, "category type does not support prod operations"), + "quantile": (TypeError, "No matching signature found"), + "rank": (None, ""), + "sem": ( + TypeError, + "|".join( + [ + "'Categorical' .* does not support reduction 'sem'", + "category dtype does not support aggregation 'sem'", + ] + ), + ), + "shift": (None, ""), + "size": (None, ""), + "skew": ( + TypeError, + "|".join( + [ + "dtype category does not support reduction 'skew'", + "category type does not support skew operations", + ] + ), + ), + "std": ( + TypeError, + "|".join( + [ + "'Categorical' .* does not support reduction 'std'", + "category dtype does not support aggregation 'std'", + ] + ), + ), + "sum": (TypeError, "category type does not support sum operations"), + "var": ( + TypeError, + "|".join( + [ + "'Categorical' .* does not support reduction 'var'", + "category dtype does not support aggregation 'var'", + ] + ), + ), + }[groupby_func] + + if groupby_func == "fillna": + kind = "Series" if groupby_series else "DataFrame" + warn_msg = f"{kind}GroupBy.fillna is deprecated" + else: + warn_msg = "" + _call_and_check(klass, msg, how, gb, groupby_func, args, warn_msg) + + +@pytest.mark.parametrize("how", ["agg", "transform"]) +def test_groupby_raises_category_udf(how, by, groupby_series, df_with_cat_col): + # GH#50749 + df = df_with_cat_col + gb = df.groupby(by=by) + + if groupby_series: + gb = gb["d"] + + def func(x): + raise TypeError("Test error message") + + with pytest.raises(TypeError, match="Test error message"): + getattr(gb, how)(func) + + +@pytest.mark.parametrize("how", ["agg", "transform"]) +@pytest.mark.parametrize("groupby_func_np", [np.sum, np.mean]) +def test_groupby_raises_category_np( + how, by, groupby_series, groupby_func_np, df_with_cat_col +): + # GH#50749 + df = df_with_cat_col + gb = df.groupby(by=by) + + if groupby_series: + gb = gb["d"] + + klass, msg = { + np.sum: (TypeError, "category type does not support sum operations"), + np.mean: ( + TypeError, + "category dtype does not support aggregation 'mean'", + ), + }[groupby_func_np] + + if groupby_series: + warn_msg = "using SeriesGroupBy.[sum|mean]" + else: + warn_msg = "using DataFrameGroupBy.[sum|mean]" + _call_and_check(klass, msg, how, gb, groupby_func_np, (), warn_msg=warn_msg) + + +@pytest.mark.parametrize("how", ["method", "agg", "transform"]) +def test_groupby_raises_category_on_category( + how, + by, + groupby_series, + groupby_func, + observed, + using_copy_on_write, + df_with_cat_col, +): + # GH#50749 + df = df_with_cat_col + df["a"] = Categorical( + ["a", "a", "a", "a", "b", "b", "b", "b", "c"], + categories=["a", "b", "c", "d"], + ordered=True, + ) + args = get_groupby_method_args(groupby_func, df) + gb = df.groupby(by=by, observed=observed) + + if groupby_series: + gb = gb["d"] + + if groupby_func == "corrwith": + assert not hasattr(gb, "corrwith") + return + + empty_groups = not observed and any(group.empty for group in gb.groups.values()) + if ( + not observed + and how != "transform" + and isinstance(by, list) + and isinstance(by[0], str) + and by == ["a", "b"] + ): + assert not empty_groups + # TODO: empty_groups should be true due to unobserved categorical combinations + empty_groups = True + if how == "transform": + # empty groups will be ignored + empty_groups = False + + klass, msg = { + "all": (None, ""), + "any": (None, ""), + "bfill": (None, ""), + "corrwith": ( + TypeError, + r"unsupported operand type\(s\) for \*: 'Categorical' and 'int'", + ), + "count": (None, ""), + "cumcount": (None, ""), + "cummax": ( + (NotImplementedError, TypeError), + "(cummax is not supported for category dtype|" + "category dtype not supported|" + "category type does not support cummax operations)", + ), + "cummin": ( + (NotImplementedError, TypeError), + "(cummin is not supported for category dtype|" + "category dtype not supported|" + "category type does not support cummin operations)", + ), + "cumprod": ( + (NotImplementedError, TypeError), + "(cumprod is not supported for category dtype|" + "category dtype not supported|" + "category type does not support cumprod operations)", + ), + "cumsum": ( + (NotImplementedError, TypeError), + "(cumsum is not supported for category dtype|" + "category dtype not supported|" + "category type does not support cumsum operations)", + ), + "diff": (TypeError, "unsupported operand type"), + "ffill": (None, ""), + "fillna": ( + TypeError, + r"Cannot setitem on a Categorical with a new category \(0\), " + "set the categories first", + ) + if not using_copy_on_write + else (None, ""), # no-op with CoW + "first": (None, ""), + "idxmax": (ValueError, "empty group due to unobserved categories") + if empty_groups + else (None, ""), + "idxmin": (ValueError, "empty group due to unobserved categories") + if empty_groups + else (None, ""), + "last": (None, ""), + "max": (None, ""), + "mean": (TypeError, "category dtype does not support aggregation 'mean'"), + "median": (TypeError, "category dtype does not support aggregation 'median'"), + "min": (None, ""), + "ngroup": (None, ""), + "nunique": (None, ""), + "pct_change": (TypeError, "unsupported operand type"), + "prod": (TypeError, "category type does not support prod operations"), + "quantile": (TypeError, ""), + "rank": (None, ""), + "sem": ( + TypeError, + "|".join( + [ + "'Categorical' .* does not support reduction 'sem'", + "category dtype does not support aggregation 'sem'", + ] + ), + ), + "shift": (None, ""), + "size": (None, ""), + "skew": ( + TypeError, + "|".join( + [ + "category type does not support skew operations", + "dtype category does not support reduction 'skew'", + ] + ), + ), + "std": ( + TypeError, + "|".join( + [ + "'Categorical' .* does not support reduction 'std'", + "category dtype does not support aggregation 'std'", + ] + ), + ), + "sum": (TypeError, "category type does not support sum operations"), + "var": ( + TypeError, + "|".join( + [ + "'Categorical' .* does not support reduction 'var'", + "category dtype does not support aggregation 'var'", + ] + ), + ), + }[groupby_func] + + if groupby_func == "fillna": + kind = "Series" if groupby_series else "DataFrame" + warn_msg = f"{kind}GroupBy.fillna is deprecated" + else: + warn_msg = "" + _call_and_check(klass, msg, how, gb, groupby_func, args, warn_msg) + + +def test_subsetting_columns_axis_1_raises(): + # GH 35443 + df = DataFrame({"a": [1], "b": [2], "c": [3]}) + msg = "DataFrame.groupby with axis=1 is deprecated" + with tm.assert_produces_warning(FutureWarning, match=msg): + gb = df.groupby("a", axis=1) + with pytest.raises(ValueError, match="Cannot subset columns when using axis=1"): + gb["b"] diff --git a/parrot/lib/python3.10/site-packages/transformers/models/esm/openfold_utils/__pycache__/feats.cpython-310.pyc b/parrot/lib/python3.10/site-packages/transformers/models/esm/openfold_utils/__pycache__/feats.cpython-310.pyc new file mode 100644 index 0000000000000000000000000000000000000000..9ff669681f0f63309b60ad12b801e9ebf992516a Binary files /dev/null and b/parrot/lib/python3.10/site-packages/transformers/models/esm/openfold_utils/__pycache__/feats.cpython-310.pyc differ diff --git a/parrot/lib/python3.10/site-packages/transformers/models/esm/openfold_utils/tensor_utils.py b/parrot/lib/python3.10/site-packages/transformers/models/esm/openfold_utils/tensor_utils.py new file mode 100644 index 0000000000000000000000000000000000000000..99dd6dbe47b68247794e51810fd274c6352e5b4f --- /dev/null +++ b/parrot/lib/python3.10/site-packages/transformers/models/esm/openfold_utils/tensor_utils.py @@ -0,0 +1,144 @@ +# Copyright 2021 AlQuraishi Laboratory +# Copyright 2021 DeepMind Technologies Limited +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. + +from functools import partial +from typing import Any, Callable, Dict, List, Type, TypeVar, Union, overload + +import torch +import torch.nn as nn +import torch.types + + +def add(m1: torch.Tensor, m2: torch.Tensor, inplace: bool) -> torch.Tensor: + # The first operation in a checkpoint can't be in-place, but it's + # nice to have in-place addition during inference. Thus... + if not inplace: + m1 = m1 + m2 + else: + m1 += m2 + + return m1 + + +def permute_final_dims(tensor: torch.Tensor, inds: List[int]) -> torch.Tensor: + zero_index = -1 * len(inds) + first_inds = list(range(len(tensor.shape[:zero_index]))) + return tensor.permute(first_inds + [zero_index + i for i in inds]) + + +def flatten_final_dims(t: torch.Tensor, no_dims: int) -> torch.Tensor: + return t.reshape(t.shape[:-no_dims] + (-1,)) + + +def masked_mean(mask: torch.Tensor, value: torch.Tensor, dim: int, eps: float = 1e-4) -> torch.Tensor: + mask = mask.expand(*value.shape) + return torch.sum(mask * value, dim=dim) / (eps + torch.sum(mask, dim=dim)) + + +def pts_to_distogram( + pts: torch.Tensor, min_bin: torch.types.Number = 2.3125, max_bin: torch.types.Number = 21.6875, no_bins: int = 64 +) -> torch.Tensor: + boundaries = torch.linspace(min_bin, max_bin, no_bins - 1, device=pts.device) + dists = torch.sqrt(torch.sum((pts.unsqueeze(-2) - pts.unsqueeze(-3)) ** 2, dim=-1)) + return torch.bucketize(dists, boundaries) + + +def dict_multimap(fn: Callable[[list], Any], dicts: List[dict]) -> dict: + first = dicts[0] + new_dict = {} + for k, v in first.items(): + all_v = [d[k] for d in dicts] + if isinstance(v, dict): + new_dict[k] = dict_multimap(fn, all_v) + else: + new_dict[k] = fn(all_v) + + return new_dict + + +def one_hot(x: torch.Tensor, v_bins: torch.Tensor) -> torch.Tensor: + reshaped_bins = v_bins.view(((1,) * len(x.shape)) + (len(v_bins),)) + diffs = x[..., None] - reshaped_bins + am = torch.argmin(torch.abs(diffs), dim=-1) + return nn.functional.one_hot(am, num_classes=len(v_bins)).float() + + +def batched_gather(data: torch.Tensor, inds: torch.Tensor, dim: int = 0, no_batch_dims: int = 0) -> torch.Tensor: + ranges: List[Union[slice, torch.Tensor]] = [] + for i, s in enumerate(data.shape[:no_batch_dims]): + r = torch.arange(s) + r = r.view(*(*((1,) * i), -1, *((1,) * (len(inds.shape) - i - 1)))) + ranges.append(r) + + remaining_dims: List[Union[slice, torch.Tensor]] = [slice(None) for _ in range(len(data.shape) - no_batch_dims)] + remaining_dims[dim - no_batch_dims if dim >= 0 else dim] = inds + ranges.extend(remaining_dims) + # Matt note: Editing this to get around the behaviour of using a list as an array index changing + # in recent Numpy versions + return data[tuple(ranges)] + + +T = TypeVar("T") + + +# With tree_map, a poor man's JAX tree_map +def dict_map( + fn: Callable[[T], Any], dic: Dict[Any, Union[dict, list, tuple, T]], leaf_type: Type[T] +) -> Dict[Any, Union[dict, list, tuple, Any]]: + new_dict: Dict[Any, Union[dict, list, tuple, Any]] = {} + for k, v in dic.items(): + if isinstance(v, dict): + new_dict[k] = dict_map(fn, v, leaf_type) + else: + new_dict[k] = tree_map(fn, v, leaf_type) + + return new_dict + + +@overload +def tree_map(fn: Callable[[T], Any], tree: T, leaf_type: Type[T]) -> Any: + ... + + +@overload +def tree_map(fn: Callable[[T], Any], tree: dict, leaf_type: Type[T]) -> dict: + ... + + +@overload +def tree_map(fn: Callable[[T], Any], tree: list, leaf_type: Type[T]) -> list: + ... + + +@overload +def tree_map(fn: Callable[[T], Any], tree: tuple, leaf_type: Type[T]) -> tuple: + ... + + +def tree_map(fn, tree, leaf_type): + if isinstance(tree, dict): + return dict_map(fn, tree, leaf_type) + elif isinstance(tree, list): + return [tree_map(fn, x, leaf_type) for x in tree] + elif isinstance(tree, tuple): + return tuple(tree_map(fn, x, leaf_type) for x in tree) + elif isinstance(tree, leaf_type): + return fn(tree) + else: + print(type(tree)) + raise ValueError("Not supported") + + +tensor_tree_map = partial(tree_map, leaf_type=torch.Tensor) diff --git a/parrot/lib/python3.10/site-packages/transformers/models/esm/tokenization_esm.py b/parrot/lib/python3.10/site-packages/transformers/models/esm/tokenization_esm.py new file mode 100644 index 0000000000000000000000000000000000000000..27a889c87ea0b42397ed1553608aa2e5db2f85bc --- /dev/null +++ b/parrot/lib/python3.10/site-packages/transformers/models/esm/tokenization_esm.py @@ -0,0 +1,143 @@ +# coding=utf-8 +# Copyright 2022 Meta and The HuggingFace Inc. team. All rights reserved. +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. +"""Tokenization classes for ESM.""" +import os +from typing import List, Optional + +from ...tokenization_utils import PreTrainedTokenizer +from ...utils import logging + + +logger = logging.get_logger(__name__) + +VOCAB_FILES_NAMES = {"vocab_file": "vocab.txt"} + + +def load_vocab_file(vocab_file): + with open(vocab_file, "r") as f: + lines = f.read().splitlines() + return [l.strip() for l in lines] + + +class EsmTokenizer(PreTrainedTokenizer): + """ + Constructs an ESM tokenizer. + """ + + vocab_files_names = VOCAB_FILES_NAMES + model_input_names = ["input_ids", "attention_mask"] + + def __init__( + self, + vocab_file, + unk_token="", + cls_token="", + pad_token="", + mask_token="", + eos_token="", + **kwargs, + ): + self.all_tokens = load_vocab_file(vocab_file) + self._id_to_token = dict(enumerate(self.all_tokens)) + self._token_to_id = {tok: ind for ind, tok in enumerate(self.all_tokens)} + super().__init__( + unk_token=unk_token, + cls_token=cls_token, + pad_token=pad_token, + mask_token=mask_token, + eos_token=eos_token, + **kwargs, + ) + + # TODO, all the tokens are added? But they are also part of the vocab... bit strange. + # none of them are special, but they all need special splitting. + + self.unique_no_split_tokens = self.all_tokens + self._update_trie(self.unique_no_split_tokens) + + def _convert_id_to_token(self, index: int) -> str: + return self._id_to_token.get(index, self.unk_token) + + def _convert_token_to_id(self, token: str) -> int: + return self._token_to_id.get(token, self._token_to_id.get(self.unk_token)) + + def _tokenize(self, text, **kwargs): + return text.split() + + def get_vocab(self): + base_vocab = self._token_to_id.copy() + base_vocab.update(self.added_tokens_encoder) + return base_vocab + + def token_to_id(self, token: str) -> int: + return self._token_to_id.get(token, self._token_to_id.get(self.unk_token)) + + def id_to_token(self, index: int) -> str: + return self._id_to_token.get(index, self.unk_token) + + def build_inputs_with_special_tokens( + self, token_ids_0: List[int], token_ids_1: Optional[List[int]] = None + ) -> List[int]: + cls = [self.cls_token_id] + sep = [self.eos_token_id] # No sep token in ESM vocabulary + if token_ids_1 is None: + if self.eos_token_id is None: + return cls + token_ids_0 + else: + return cls + token_ids_0 + sep + elif self.eos_token_id is None: + raise ValueError("Cannot tokenize multiple sequences when EOS token is not set!") + return cls + token_ids_0 + sep + token_ids_1 + sep # Multiple inputs always have an EOS token + + def get_special_tokens_mask( + self, token_ids_0: List, token_ids_1: Optional[List] = None, already_has_special_tokens: bool = False + ) -> List[int]: + """ + Retrieves sequence ids from a token list that has no special tokens added. This method is called when adding + special tokens using the tokenizer `prepare_for_model` or `encode_plus` methods. + + Args: + token_ids_0 (`List[int]`): + List of ids of the first sequence. + token_ids_1 (`List[int]`, *optional*): + List of ids of the second sequence. + already_has_special_tokens (`bool`, *optional*, defaults to `False`): + Whether or not the token list is already formatted with special tokens for the model. + + Returns: + A list of integers in the range [0, 1]: 1 for a special token, 0 for a sequence token. + """ + if already_has_special_tokens: + if token_ids_1 is not None: + raise ValueError( + "You should not supply a second sequence if the provided sequence of " + "ids is already formatted with special tokens for the model." + ) + + return [1 if token in self.all_special_ids else 0 for token in token_ids_0] + mask = [1] + ([0] * len(token_ids_0)) + [1] + if token_ids_1 is not None: + mask += [0] * len(token_ids_1) + [1] + return mask + + def save_vocabulary(self, save_directory, filename_prefix): + vocab_file = os.path.join(save_directory, (filename_prefix + "-" if filename_prefix else "") + "vocab.txt") + with open(vocab_file, "w") as f: + f.write("\n".join(self.all_tokens)) + return (vocab_file,) + + @property + def vocab_size(self) -> int: + return len(self.all_tokens) diff --git a/parrot/lib/python3.10/site-packages/transformers/models/hubert/__init__.py b/parrot/lib/python3.10/site-packages/transformers/models/hubert/__init__.py new file mode 100644 index 0000000000000000000000000000000000000000..30331ed0d146a4c9a733e15d947808b00f959b65 --- /dev/null +++ b/parrot/lib/python3.10/site-packages/transformers/models/hubert/__init__.py @@ -0,0 +1,79 @@ +# Copyright 2021 The HuggingFace Team. All rights reserved. +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. +from typing import TYPE_CHECKING + +from ...utils import OptionalDependencyNotAvailable, _LazyModule, is_tf_available, is_torch_available + + +_import_structure = {"configuration_hubert": ["HubertConfig"]} + +try: + if not is_torch_available(): + raise OptionalDependencyNotAvailable() +except OptionalDependencyNotAvailable: + pass +else: + _import_structure["modeling_hubert"] = [ + "HubertForCTC", + "HubertForSequenceClassification", + "HubertModel", + "HubertPreTrainedModel", + ] + + +try: + if not is_tf_available(): + raise OptionalDependencyNotAvailable() +except OptionalDependencyNotAvailable: + pass +else: + _import_structure["modeling_tf_hubert"] = [ + "TFHubertForCTC", + "TFHubertModel", + "TFHubertPreTrainedModel", + ] + +if TYPE_CHECKING: + from .configuration_hubert import HubertConfig + + try: + if not is_torch_available(): + raise OptionalDependencyNotAvailable() + except OptionalDependencyNotAvailable: + pass + else: + from .modeling_hubert import ( + HubertForCTC, + HubertForSequenceClassification, + HubertModel, + HubertPreTrainedModel, + ) + + try: + if not is_tf_available(): + raise OptionalDependencyNotAvailable() + except OptionalDependencyNotAvailable: + pass + else: + from .modeling_tf_hubert import ( + TFHubertForCTC, + TFHubertModel, + TFHubertPreTrainedModel, + ) + + +else: + import sys + + sys.modules[__name__] = _LazyModule(__name__, globals()["__file__"], _import_structure, module_spec=__spec__) diff --git a/parrot/lib/python3.10/site-packages/transformers/models/hubert/__pycache__/convert_hubert_original_s3prl_checkpoint_to_pytorch.cpython-310.pyc b/parrot/lib/python3.10/site-packages/transformers/models/hubert/__pycache__/convert_hubert_original_s3prl_checkpoint_to_pytorch.cpython-310.pyc new file mode 100644 index 0000000000000000000000000000000000000000..b0d852a6b3aff601a965571c7be40fef81261536 Binary files /dev/null and b/parrot/lib/python3.10/site-packages/transformers/models/hubert/__pycache__/convert_hubert_original_s3prl_checkpoint_to_pytorch.cpython-310.pyc differ diff --git a/parrot/lib/python3.10/site-packages/transformers/models/hubert/configuration_hubert.py b/parrot/lib/python3.10/site-packages/transformers/models/hubert/configuration_hubert.py new file mode 100644 index 0000000000000000000000000000000000000000..293b9f76f2cfce5fa2c1683fafc59f3cd202e83c --- /dev/null +++ b/parrot/lib/python3.10/site-packages/transformers/models/hubert/configuration_hubert.py @@ -0,0 +1,258 @@ +# coding=utf-8 +# Copyright 2021 The Fairseq Authors and The HuggingFace Inc. team. All rights reserved. +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. +""" Hubert model configuration""" + +import functools +import operator + +from ...configuration_utils import PretrainedConfig +from ...utils import logging + + +logger = logging.get_logger(__name__) + + +class HubertConfig(PretrainedConfig): + r""" + This is the configuration class to store the configuration of a [`HubertModel`]. It is used to instantiate an + Hubert model according to the specified arguments, defining the model architecture. Instantiating a configuration + with the defaults will yield a similar configuration to that of the Hubert + [facebook/hubert-base-ls960](https://huggingface.co/facebook/hubert-base-ls960) architecture. + + Configuration objects inherit from [`PretrainedConfig`] and can be used to control the model outputs. Read the + documentation from [`PretrainedConfig`] for more information. + + + Args: + vocab_size (`int`, *optional*, defaults to 32): + Vocabulary size of the Hubert model. Defines the number of different tokens that can be represented by the + `inputs_ids` passed when calling [`HubertModel`]. Vocabulary size of the model. Defines the different + tokens that can be represented by the *inputs_ids* passed to the forward method of [`HubertModel`]. + hidden_size (`int`, *optional*, defaults to 768): + Dimensionality of the encoder layers and the pooler layer. + num_hidden_layers (`int`, *optional*, defaults to 12): + Number of hidden layers in the Transformer encoder. + num_attention_heads (`int`, *optional*, defaults to 12): + Number of attention heads for each attention layer in the Transformer encoder. + intermediate_size (`int`, *optional*, defaults to 3072): + Dimensionality of the "intermediate" (i.e., feed-forward) layer in the Transformer encoder. + hidden_act (`str` or `function`, *optional*, defaults to `"gelu"`): + The non-linear activation function (function or string) in the encoder and pooler. If string, `"gelu"`, + `"relu"`, `"selu"` and `"gelu_new"` are supported. + hidden_dropout(`float`, *optional*, defaults to 0.1): + The dropout probability for all fully connected layers in the embeddings, encoder, and pooler. + activation_dropout (`float`, *optional*, defaults to 0.1): + The dropout ratio for activations inside the fully connected layer. + attention_dropout(`float`, *optional*, defaults to 0.1): + The dropout ratio for the attention probabilities. + final_dropout (`float`, *optional*, defaults to 0.1): + The dropout probability for the final projection layer of [`Wav2Vec2ForCTC`]. + layerdrop (`float`, *optional*, defaults to 0.1): + The LayerDrop probability. See the [LayerDrop paper](see https://arxiv.org/abs/1909.11556) for more + details. + initializer_range (`float`, *optional*, defaults to 0.02): + The standard deviation of the truncated_normal_initializer for initializing all weight matrices. + layer_norm_eps (`float`, *optional*, defaults to 1e-12): + The epsilon used by the layer normalization layers. + feat_extract_norm (`str`, *optional*, defaults to `"group"`): + The norm to be applied to 1D convolutional layers in feature encoder. One of `"group"` for group + normalization of only the first 1D convolutional layer or `"layer"` for layer normalization of all 1D + convolutional layers. + feat_proj_dropout (`float`, *optional*, defaults to 0.0): + The dropout probability for output of the feature encoder. + feat_proj_layer_norm (`bool`, *optional*, defaults to `True`): + Whether to apply LayerNorm to the output of the feature encoder. + feat_extract_activation (`str, `optional`, defaults to `"gelu"`): + The non-linear activation function (function or string) in the 1D convolutional layers of the feature + extractor. If string, `"gelu"`, `"relu"`, `"selu"` and `"gelu_new"` are supported. + conv_dim (`Tuple[int]`, *optional*, defaults to `(512, 512, 512, 512, 512, 512, 512)`): + A tuple of integers defining the number of input and output channels of each 1D convolutional layer in the + feature encoder. The length of *conv_dim* defines the number of 1D convolutional layers. + conv_stride (`Tuple[int]`, *optional*, defaults to `(5, 2, 2, 2, 2, 2, 2)`): + A tuple of integers defining the stride of each 1D convolutional layer in the feature encoder. The length + of *conv_stride* defines the number of convolutional layers and has to match the length of *conv_dim*. + conv_kernel (`Tuple[int]`, *optional*, defaults to `(10, 3, 3, 3, 3, 3, 3)`): + A tuple of integers defining the kernel size of each 1D convolutional layer in the feature encoder. The + length of *conv_kernel* defines the number of convolutional layers and has to match the length of + *conv_dim*. + conv_bias (`bool`, *optional*, defaults to `False`): + Whether the 1D convolutional layers have a bias. + num_conv_pos_embeddings (`int`, *optional*, defaults to 128): + Number of convolutional positional embeddings. Defines the kernel size of 1D convolutional positional + embeddings layer. + num_conv_pos_embedding_groups (`int`, *optional*, defaults to 16): + Number of groups of 1D convolutional positional embeddings layer. + do_stable_layer_norm (`bool`, *optional*, defaults to `False`): + Whether do apply *stable* layer norm architecture of the Transformer encoder. `do_stable_layer_norm is + True` corresponds to applying layer norm before the attention layer, whereas `do_stable_layer_norm is + False` corresponds to applying layer norm after the attention layer. + apply_spec_augment (`bool`, *optional*, defaults to `True`): + Whether to apply *SpecAugment* data augmentation to the outputs of the feature encoder. For reference see + [SpecAugment: A Simple Data Augmentation Method for Automatic Speech + Recognition](https://arxiv.org/abs/1904.08779). + mask_time_prob (`float`, *optional*, defaults to 0.05): + Percentage (between 0 and 1) of all feature vectors along the time axis which will be masked. The masking + procecure generates ''mask_time_prob*len(time_axis)/mask_time_length'' independent masks over the axis. If + reasoning from the propability of each feature vector to be chosen as the start of the vector span to be + masked, *mask_time_prob* should be `prob_vector_start*mask_time_length`. Note that overlap may decrease the + actual percentage of masked vectors. This is only relevant if `apply_spec_augment is True`. + mask_time_length (`int`, *optional*, defaults to 10): + Length of vector span along the time axis. + mask_time_min_masks (`int`, *optional*, defaults to 2),: + The minimum number of masks of length `mask_feature_length` generated along the time axis, each time step, + irrespectively of `mask_feature_prob`. Only relevant if ''mask_time_prob*len(time_axis)/mask_time_length < + mask_time_min_masks'' + mask_feature_prob (`float`, *optional*, defaults to 0.0): + Percentage (between 0 and 1) of all feature vectors along the feature axis which will be masked. The + masking procecure generates ''mask_feature_prob*len(feature_axis)/mask_time_length'' independent masks over + the axis. If reasoning from the propability of each feature vector to be chosen as the start of the vector + span to be masked, *mask_feature_prob* should be `prob_vector_start*mask_feature_length`. Note that overlap + may decrease the actual percentage of masked vectors. This is only relevant if `apply_spec_augment is + True`. + mask_feature_length (`int`, *optional*, defaults to 10): + Length of vector span along the feature axis. + mask_feature_min_masks (`int`, *optional*, defaults to 0),: + The minimum number of masks of length `mask_feature_length` generated along the feature axis, each time + step, irrespectively of `mask_feature_prob`. Only relevant if + ''mask_feature_prob*len(feature_axis)/mask_feature_length < mask_feature_min_masks'' + ctc_loss_reduction (`str`, *optional*, defaults to `"sum"`): + Specifies the reduction to apply to the output of `torch.nn.CTCLoss`. Only relevant when training an + instance of [`HubertForCTC`]. + ctc_zero_infinity (`bool`, *optional*, defaults to `False`): + Whether to zero infinite losses and the associated gradients of `torch.nn.CTCLoss`. Infinite losses mainly + occur when the inputs are too short to be aligned to the targets. Only relevant when training an instance + of [`HubertForCTC`]. + use_weighted_layer_sum (`bool`, *optional*, defaults to `False`): + Whether to use a weighted average of layer outputs with learned weights. Only relevant when using an + instance of [`HubertForSequenceClassification`]. + classifier_proj_size (`int`, *optional*, defaults to 256): + Dimensionality of the projection before token mean-pooling for classification. + + Example: + + ```python + >>> from transformers import HubertModel, HubertConfig + + >>> # Initializing a Hubert facebook/hubert-base-ls960 style configuration + >>> configuration = HubertConfig() + + >>> # Initializing a model from the facebook/hubert-base-ls960 style configuration + >>> model = HubertModel(configuration) + + >>> # Accessing the model configuration + >>> configuration = model.config + ```""" + + model_type = "hubert" + + def __init__( + self, + vocab_size=32, + hidden_size=768, + num_hidden_layers=12, + num_attention_heads=12, + intermediate_size=3072, + hidden_act="gelu", + hidden_dropout=0.1, + activation_dropout=0.1, + attention_dropout=0.1, + feat_proj_layer_norm=True, + feat_proj_dropout=0.0, + final_dropout=0.1, + layerdrop=0.1, + initializer_range=0.02, + layer_norm_eps=1e-5, + feat_extract_norm="group", + feat_extract_activation="gelu", + conv_dim=(512, 512, 512, 512, 512, 512, 512), + conv_stride=(5, 2, 2, 2, 2, 2, 2), + conv_kernel=(10, 3, 3, 3, 3, 2, 2), + conv_bias=False, + num_conv_pos_embeddings=128, + num_conv_pos_embedding_groups=16, + do_stable_layer_norm=False, + apply_spec_augment=True, + mask_time_prob=0.05, + mask_time_length=10, + mask_time_min_masks=2, + mask_feature_prob=0.0, + mask_feature_length=10, + mask_feature_min_masks=0, + ctc_loss_reduction="sum", + ctc_zero_infinity=False, + use_weighted_layer_sum=False, + classifier_proj_size=256, + pad_token_id=0, + bos_token_id=1, + eos_token_id=2, + **kwargs, + ): + super().__init__(**kwargs, pad_token_id=pad_token_id, bos_token_id=bos_token_id, eos_token_id=eos_token_id) + self.hidden_size = hidden_size + self.feat_extract_norm = feat_extract_norm + self.feat_extract_activation = feat_extract_activation + self.conv_dim = list(conv_dim) + self.conv_stride = list(conv_stride) + self.conv_kernel = list(conv_kernel) + self.conv_bias = conv_bias + self.num_conv_pos_embeddings = num_conv_pos_embeddings + self.num_conv_pos_embedding_groups = num_conv_pos_embedding_groups + self.num_feat_extract_layers = len(self.conv_dim) + self.num_hidden_layers = num_hidden_layers + self.intermediate_size = intermediate_size + self.hidden_act = hidden_act + self.num_attention_heads = num_attention_heads + self.hidden_dropout = hidden_dropout + self.attention_dropout = attention_dropout + self.activation_dropout = activation_dropout + self.feat_proj_layer_norm = feat_proj_layer_norm + self.feat_proj_dropout = feat_proj_dropout + self.final_dropout = final_dropout + self.layerdrop = layerdrop + self.layer_norm_eps = layer_norm_eps + self.initializer_range = initializer_range + self.vocab_size = vocab_size + self.do_stable_layer_norm = do_stable_layer_norm + self.use_weighted_layer_sum = use_weighted_layer_sum + self.classifier_proj_size = classifier_proj_size + + if ( + (len(self.conv_stride) != self.num_feat_extract_layers) + or (len(self.conv_kernel) != self.num_feat_extract_layers) + or (len(self.conv_dim) != self.num_feat_extract_layers) + ): + raise ValueError( + "Configuration for convolutional layers is incorrect. It is required that `len(config.conv_dim)` ==" + " `len(config.conv_stride)` == `len(config.conv_kernel)`, but is `len(config.conv_dim) =" + f" {len(self.conv_dim)}`, `len(config.conv_stride) = {len(self.conv_stride)}`," + f" `len(config.conv_kernel) = {len(self.conv_kernel)}`." + ) + + # fine-tuning config parameters for SpecAugment: https://arxiv.org/abs/1904.08779 + self.apply_spec_augment = apply_spec_augment + self.mask_time_prob = mask_time_prob + self.mask_time_length = mask_time_length + self.mask_time_min_masks = mask_time_min_masks + self.mask_feature_prob = mask_feature_prob + self.mask_feature_length = mask_feature_length + self.mask_feature_min_masks = mask_feature_min_masks + + # ctc loss + self.ctc_loss_reduction = ctc_loss_reduction + self.ctc_zero_infinity = ctc_zero_infinity + + @property + def inputs_to_logits_ratio(self): + return functools.reduce(operator.mul, self.conv_stride, 1) diff --git a/parrot/lib/python3.10/site-packages/transformers/models/hubert/convert_distilhubert_original_s3prl_checkpoint_to_pytorch.py b/parrot/lib/python3.10/site-packages/transformers/models/hubert/convert_distilhubert_original_s3prl_checkpoint_to_pytorch.py new file mode 100644 index 0000000000000000000000000000000000000000..571761e022846f669f106735e3f5a9c6e7037165 --- /dev/null +++ b/parrot/lib/python3.10/site-packages/transformers/models/hubert/convert_distilhubert_original_s3prl_checkpoint_to_pytorch.py @@ -0,0 +1,223 @@ +# coding=utf-8 +# Copyright 2021 The HuggingFace Inc. team. +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. +"""Convert Hubert checkpoint.""" + + +import argparse + +import torch +from s3prl.hub import distilhubert + +from transformers import HubertConfig, HubertModel, Wav2Vec2FeatureExtractor, logging + + +logging.set_verbosity_info() +logger = logging.get_logger(__name__) + +MAPPING = { + "post_extract_proj": "feature_projection.projection", + "encoder.pos_conv.0": "encoder.pos_conv_embed.conv", + "self_attn.k_proj": "encoder.layers.*.attention.k_proj", + "self_attn.v_proj": "encoder.layers.*.attention.v_proj", + "self_attn.q_proj": "encoder.layers.*.attention.q_proj", + "self_attn.out_proj": "encoder.layers.*.attention.out_proj", + "self_attn_layer_norm": "encoder.layers.*.layer_norm", + "fc1": "encoder.layers.*.feed_forward.intermediate_dense", + "fc2": "encoder.layers.*.feed_forward.output_dense", + "final_layer_norm": "encoder.layers.*.final_layer_norm", + "encoder.layer_norm": "encoder.layer_norm", + "mask_emb": "masked_spec_embed", +} + + +def set_recursively(hf_pointer, key, value, full_name, weight_type): + for attribute in key.split("."): + hf_pointer = getattr(hf_pointer, attribute) + + if weight_type is not None: + hf_shape = getattr(hf_pointer, weight_type).shape + else: + hf_shape = hf_pointer.shape + + assert hf_shape == value.shape, ( + f"Shape of hf {key + '.' + weight_type if weight_type is not None else ''} is {hf_shape}, but should be" + f" {value.shape} for {full_name}" + ) + + if weight_type == "weight": + hf_pointer.weight.data = value + elif weight_type == "weight_g": + hf_pointer.weight_g.data = value + elif weight_type == "weight_v": + hf_pointer.weight_v.data = value + elif weight_type == "bias": + hf_pointer.bias.data = value + else: + hf_pointer.data = value + + logger.info(f"{key + '.' + weight_type if weight_type is not None else ''} was initialized from {full_name}.") + + +def recursively_load_weights(fairseq_model, hf_model): + unused_weights = [] + fairseq_dict = fairseq_model.state_dict() + + feature_extractor = hf_model.feature_extractor + + for name, value in fairseq_dict.items(): + is_used = False + if "conv_layers" in name: + load_conv_layer( + name, + value, + feature_extractor, + unused_weights, + hf_model.config.feat_extract_norm == "group", + ) + is_used = True + else: + for key, mapped_key in MAPPING.items(): + mapped_key = mapped_key + + if key in name: + is_used = True + if "*" in mapped_key: + layer_index = name.split(key)[0].split(".")[-2] + mapped_key = mapped_key.replace("*", layer_index) + if "weight_g" in name: + weight_type = "weight_g" + elif "weight_v" in name: + weight_type = "weight_v" + elif "weight" in name: + weight_type = "weight" + elif "bias" in name: + weight_type = "bias" + else: + weight_type = None + set_recursively(hf_model, mapped_key, value, name, weight_type) + continue + if not is_used: + unused_weights.append(name) + + logger.warning(f"Unused weights: {unused_weights}") + + +def load_conv_layer(full_name, value, feature_extractor, unused_weights, use_group_norm): + name = full_name.split("conv_layers.")[-1] + items = name.split(".") + layer_id = int(items[0]) + type_id = int(items[1]) + + if type_id == 0: + if "bias" in name: + assert value.shape == feature_extractor.conv_layers[layer_id].conv.bias.data.shape, ( + f"{full_name} has size {value.shape}, but" + f" {feature_extractor.conv_layers[layer_id].conv.bias.data.shape} was found." + ) + feature_extractor.conv_layers[layer_id].conv.bias.data = value + logger.info(f"Feat extract conv layer {layer_id} was initialized from {full_name}.") + elif "weight" in name: + assert value.shape == feature_extractor.conv_layers[layer_id].conv.weight.data.shape, ( + f"{full_name} has size {value.shape}, but" + f" {feature_extractor.conv_layers[layer_id].conv.weight.data.shape} was found." + ) + feature_extractor.conv_layers[layer_id].conv.weight.data = value + logger.info(f"Feat extract conv layer {layer_id} was initialized from {full_name}.") + elif (type_id == 2 and not use_group_norm) or (type_id == 2 and layer_id == 0 and use_group_norm): + if "bias" in name: + assert value.shape == feature_extractor.conv_layers[layer_id].layer_norm.bias.data.shape, ( + f"{full_name} has size {value.shape}, but {feature_extractor[layer_id].layer_norm.bias.data.shape} was" + " found." + ) + feature_extractor.conv_layers[layer_id].layer_norm.bias.data = value + logger.info(f"Feat extract layer norm weight of layer {layer_id} was initialized from {full_name}.") + elif "weight" in name: + assert value.shape == feature_extractor.conv_layers[layer_id].layer_norm.weight.data.shape, ( + f"{full_name} has size {value.shape}, but" + f" {feature_extractor[layer_id].layer_norm.weight.data.shape} was found." + ) + feature_extractor.conv_layers[layer_id].layer_norm.weight.data = value + logger.info(f"Feat extract layer norm weight of layer {layer_id} was initialized from {full_name}.") + else: + unused_weights.append(full_name) + + +def convert_config(model): + config = HubertConfig() + fs_config = model.config + + config.activation_dropout = fs_config.activation_dropout + config.apply_spec_augment = False + config.attention_dropout = fs_config.attention_dropout + config.conv_bias = False + conv_layers = eval(fs_config.extractor_conv_feature_layers) + config.conv_dim = [x[0] for x in conv_layers] + config.conv_kernel = [x[1] for x in conv_layers] + config.conv_stride = [x[2] for x in conv_layers] + config.feat_extract_activation = "gelu" + config.feat_extract_norm = "layer" if fs_config.extractor_mode == "layer_norm" else "group" + config.feat_proj_layer_norm = False + config.feat_proj_dropout = 0.0 + config.final_dropout = 0.0 + config.hidden_act = fs_config.activation_fn + config.hidden_dropout = fs_config.dropout + config.hidden_size = fs_config.encoder_embed_dim + config.initializer_range = 0.02 + config.intermediate_size = fs_config.encoder_ffn_embed_dim + config.layer_norm_eps = 1e-5 + config.layerdrop = 0.0 + config.num_attention_heads = fs_config.encoder_attention_heads + config.num_conv_pos_embedding_groups = fs_config.conv_pos_groups + config.num_conv_pos_embeddings = fs_config.conv_pos + config.num_feat_extract_layers = len(conv_layers) + config.num_hidden_layers = fs_config.encoder_layers + + return config + + +@torch.no_grad() +def convert_hubert_checkpoint(pytorch_dump_folder_path, config_path=None): + """ + Copy/paste/tweak model's weights to transformers design. + """ + model = distilhubert().model.model + + if config_path is not None: + config = HubertConfig.from_pretrained(config_path) + else: + config = convert_config(model) + model = model.eval() + + feature_extractor = Wav2Vec2FeatureExtractor( + feature_size=1, + sampling_rate=16000, + padding_value=0, + do_normalize=False, + return_attention_mask=False, + ) + hf_model = HubertModel(config) + + recursively_load_weights(model, hf_model) + + feature_extractor.save_pretrained(pytorch_dump_folder_path) + hf_model.save_pretrained(pytorch_dump_folder_path) + + +if __name__ == "__main__": + parser = argparse.ArgumentParser() + parser.add_argument("--pytorch_dump_folder_path", default=None, type=str, help="Path to the output PyTorch model.") + parser.add_argument("--config_path", default=None, type=str, help="Path to hf config.json of model to convert") + args = parser.parse_args() + convert_hubert_checkpoint(args.pytorch_dump_folder_path, args.config_path) diff --git a/parrot/lib/python3.10/site-packages/transformers/models/hubert/convert_hubert_original_pytorch_checkpoint_to_pytorch.py b/parrot/lib/python3.10/site-packages/transformers/models/hubert/convert_hubert_original_pytorch_checkpoint_to_pytorch.py new file mode 100644 index 0000000000000000000000000000000000000000..9a70fb6db710f49e265a3fa449cd01cec281accb --- /dev/null +++ b/parrot/lib/python3.10/site-packages/transformers/models/hubert/convert_hubert_original_pytorch_checkpoint_to_pytorch.py @@ -0,0 +1,249 @@ +# coding=utf-8 +# Copyright 2021 The HuggingFace Inc. team. +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. +"""Convert Hubert checkpoint.""" + + +import argparse +import json +import os + +import fairseq +import torch +from fairseq.data import Dictionary + +from transformers import ( + HubertConfig, + HubertForCTC, + HubertModel, + Wav2Vec2CTCTokenizer, + Wav2Vec2FeatureExtractor, + Wav2Vec2Processor, + logging, +) + + +logging.set_verbosity_info() +logger = logging.get_logger(__name__) + +MAPPING = { + "post_extract_proj": "feature_projection.projection", + "encoder.pos_conv.0": "encoder.pos_conv_embed.conv", + "self_attn.k_proj": "encoder.layers.*.attention.k_proj", + "self_attn.v_proj": "encoder.layers.*.attention.v_proj", + "self_attn.q_proj": "encoder.layers.*.attention.q_proj", + "self_attn.out_proj": "encoder.layers.*.attention.out_proj", + "self_attn_layer_norm": "encoder.layers.*.layer_norm", + "fc1": "encoder.layers.*.feed_forward.intermediate_dense", + "fc2": "encoder.layers.*.feed_forward.output_dense", + "final_layer_norm": "encoder.layers.*.final_layer_norm", + "encoder.layer_norm": "encoder.layer_norm", + "w2v_model.layer_norm": "feature_projection.layer_norm", + "w2v_encoder.proj": "lm_head", + "mask_emb": "masked_spec_embed", +} + + +def set_recursively(hf_pointer, key, value, full_name, weight_type): + for attribute in key.split("."): + hf_pointer = getattr(hf_pointer, attribute) + + if weight_type is not None: + hf_shape = getattr(hf_pointer, weight_type).shape + else: + hf_shape = hf_pointer.shape + + assert hf_shape == value.shape, ( + f"Shape of hf {key + '.' + weight_type if weight_type is not None else ''} is {hf_shape}, but should be" + f" {value.shape} for {full_name}" + ) + + if weight_type == "weight": + hf_pointer.weight.data = value + elif weight_type == "weight_g": + hf_pointer.weight_g.data = value + elif weight_type == "weight_v": + hf_pointer.weight_v.data = value + elif weight_type == "bias": + hf_pointer.bias.data = value + else: + hf_pointer.data = value + + logger.info(f"{key + '.' + weight_type if weight_type is not None else ''} was initialized from {full_name}.") + + +def recursively_load_weights(fairseq_model, hf_model, is_finetuned): + unused_weights = [] + fairseq_dict = fairseq_model.state_dict() + + feature_extractor = hf_model.hubert.feature_extractor if is_finetuned else hf_model.feature_extractor + + for name, value in fairseq_dict.items(): + is_used = False + if "conv_layers" in name: + load_conv_layer( + name, + value, + feature_extractor, + unused_weights, + hf_model.config.feat_extract_norm == "group", + ) + is_used = True + else: + for key, mapped_key in MAPPING.items(): + mapped_key = "hubert." + mapped_key if (is_finetuned and mapped_key != "lm_head") else mapped_key + + if key in name or (key.split("w2v_model.")[-1] == name.split(".")[0] and not is_finetuned): + is_used = True + if "*" in mapped_key: + layer_index = name.split(key)[0].split(".")[-2] + mapped_key = mapped_key.replace("*", layer_index) + if "weight_g" in name: + weight_type = "weight_g" + elif "weight_v" in name: + weight_type = "weight_v" + elif "weight" in name: + weight_type = "weight" + elif "bias" in name: + weight_type = "bias" + else: + weight_type = None + set_recursively(hf_model, mapped_key, value, name, weight_type) + continue + if not is_used: + unused_weights.append(name) + + logger.warning(f"Unused weights: {unused_weights}") + + +def load_conv_layer(full_name, value, feature_extractor, unused_weights, use_group_norm): + name = full_name.split("conv_layers.")[-1] + items = name.split(".") + layer_id = int(items[0]) + type_id = int(items[1]) + + if type_id == 0: + if "bias" in name: + assert value.shape == feature_extractor.conv_layers[layer_id].conv.bias.data.shape, ( + f"{full_name} has size {value.shape}, but" + f" {feature_extractor.conv_layers[layer_id].conv.bias.data.shape} was found." + ) + feature_extractor.conv_layers[layer_id].conv.bias.data = value + logger.info(f"Feat extract conv layer {layer_id} was initialized from {full_name}.") + elif "weight" in name: + assert value.shape == feature_extractor.conv_layers[layer_id].conv.weight.data.shape, ( + f"{full_name} has size {value.shape}, but" + f" {feature_extractor.conv_layers[layer_id].conv.weight.data.shape} was found." + ) + feature_extractor.conv_layers[layer_id].conv.weight.data = value + logger.info(f"Feat extract conv layer {layer_id} was initialized from {full_name}.") + elif (type_id == 2 and not use_group_norm) or (type_id == 2 and layer_id == 0 and use_group_norm): + if "bias" in name: + assert value.shape == feature_extractor.conv_layers[layer_id].layer_norm.bias.data.shape, ( + f"{full_name} has size {value.shape}, but {feature_extractor[layer_id].layer_norm.bias.data.shape} was" + " found." + ) + feature_extractor.conv_layers[layer_id].layer_norm.bias.data = value + logger.info(f"Feat extract layer norm weight of layer {layer_id} was initialized from {full_name}.") + elif "weight" in name: + assert value.shape == feature_extractor.conv_layers[layer_id].layer_norm.weight.data.shape, ( + f"{full_name} has size {value.shape}, but" + f" {feature_extractor[layer_id].layer_norm.weight.data.shape} was found." + ) + feature_extractor.conv_layers[layer_id].layer_norm.weight.data = value + logger.info(f"Feat extract layer norm weight of layer {layer_id} was initialized from {full_name}.") + else: + unused_weights.append(full_name) + + +@torch.no_grad() +def convert_hubert_checkpoint( + checkpoint_path, pytorch_dump_folder_path, config_path=None, dict_path=None, is_finetuned=True +): + """ + Copy/paste/tweak model's weights to transformers design. + """ + if config_path is not None: + config = HubertConfig.from_pretrained(config_path) + else: + config = HubertConfig() + + if is_finetuned: + if dict_path: + target_dict = Dictionary.load(dict_path) + + # important change bos & pad token id since CTC symbol is and + # not as in fairseq + config.bos_token_id = target_dict.pad_index + config.pad_token_id = target_dict.bos_index + config.eos_token_id = target_dict.eos_index + config.vocab_size = len(target_dict.symbols) + vocab_path = os.path.join(pytorch_dump_folder_path, "vocab.json") + if not os.path.isdir(pytorch_dump_folder_path): + logger.error("--pytorch_dump_folder_path ({}) should be a directory".format(pytorch_dump_folder_path)) + return + os.makedirs(pytorch_dump_folder_path, exist_ok=True) + with open(vocab_path, "w", encoding="utf-8") as vocab_handle: + json.dump(target_dict.indices, vocab_handle) + tokenizer = Wav2Vec2CTCTokenizer( + vocab_path, + unk_token=target_dict.unk_word, + pad_token=target_dict.pad_word, + bos_token=target_dict.bos_word, + eos_token=target_dict.eos_word, + word_delimiter_token="|", + do_lower_case=False, + ) + return_attention_mask = True if config.feat_extract_norm == "layer" else False + feature_extractor = Wav2Vec2FeatureExtractor( + feature_size=1, + sampling_rate=16000, + padding_value=0, + do_normalize=True, + return_attention_mask=return_attention_mask, + ) + processor = Wav2Vec2Processor(feature_extractor=feature_extractor, tokenizer=tokenizer) + processor.save_pretrained(pytorch_dump_folder_path) + + hf_wav2vec = HubertForCTC(config) + else: + hf_wav2vec = HubertModel(config) + + if is_finetuned: + model, _, _ = fairseq.checkpoint_utils.load_model_ensemble_and_task( + [checkpoint_path], arg_overrides={"data": "/".join(dict_path.split("/")[:-1])} + ) + else: + model, _, _ = fairseq.checkpoint_utils.load_model_ensemble_and_task([checkpoint_path]) + + model = model[0].eval() + + recursively_load_weights(model, hf_wav2vec, is_finetuned) + + hf_wav2vec.save_pretrained(pytorch_dump_folder_path) + + +if __name__ == "__main__": + parser = argparse.ArgumentParser() + parser.add_argument("--pytorch_dump_folder_path", default=None, type=str, help="Path to the output PyTorch model.") + parser.add_argument("--checkpoint_path", default=None, type=str, help="Path to fairseq checkpoint") + parser.add_argument("--dict_path", default=None, type=str, help="Path to dict of fine-tuned model") + parser.add_argument("--config_path", default=None, type=str, help="Path to hf config.json of model to convert") + parser.add_argument( + "--not_finetuned", action="store_true", help="Whether the model to convert is a fine-tuned model or not" + ) + args = parser.parse_args() + convert_hubert_checkpoint( + args.checkpoint_path, args.pytorch_dump_folder_path, args.config_path, args.dict_path, not args.not_finetuned + ) diff --git a/parrot/lib/python3.10/site-packages/transformers/models/hubert/convert_hubert_original_s3prl_checkpoint_to_pytorch.py b/parrot/lib/python3.10/site-packages/transformers/models/hubert/convert_hubert_original_s3prl_checkpoint_to_pytorch.py new file mode 100644 index 0000000000000000000000000000000000000000..51908f930242c6580d2d154bec7e632e7af568fe --- /dev/null +++ b/parrot/lib/python3.10/site-packages/transformers/models/hubert/convert_hubert_original_s3prl_checkpoint_to_pytorch.py @@ -0,0 +1,69 @@ +# coding=utf-8 +# Copyright 2021 The HuggingFace Inc. team. +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. +"""Convert Hubert checkpoint.""" + + +import argparse + +import torch + +from transformers import HubertConfig, HubertForSequenceClassification, Wav2Vec2FeatureExtractor, logging + + +logging.set_verbosity_info() +logger = logging.get_logger(__name__) + +SUPPORTED_MODELS = ["UtteranceLevel"] + + +@torch.no_grad() +def convert_s3prl_checkpoint(base_model_name, config_path, checkpoint_path, model_dump_path): + """ + Copy/paste/tweak model's weights to transformers design. + """ + checkpoint = torch.load(checkpoint_path, map_location="cpu") + if checkpoint["Config"]["downstream_expert"]["modelrc"]["select"] not in SUPPORTED_MODELS: + raise NotImplementedError(f"The supported s3prl models are {SUPPORTED_MODELS}") + + downstream_dict = checkpoint["Downstream"] + + hf_congfig = HubertConfig.from_pretrained(config_path) + hf_model = HubertForSequenceClassification.from_pretrained(base_model_name, config=hf_congfig) + hf_feature_extractor = Wav2Vec2FeatureExtractor.from_pretrained( + base_model_name, return_attention_mask=True, do_normalize=False + ) + + if hf_congfig.use_weighted_layer_sum: + hf_model.layer_weights.data = checkpoint["Featurizer"]["weights"] + + hf_model.projector.weight.data = downstream_dict["projector.weight"] + hf_model.projector.bias.data = downstream_dict["projector.bias"] + hf_model.classifier.weight.data = downstream_dict["model.post_net.linear.weight"] + hf_model.classifier.bias.data = downstream_dict["model.post_net.linear.bias"] + + hf_feature_extractor.save_pretrained(model_dump_path) + hf_model.save_pretrained(model_dump_path) + + +if __name__ == "__main__": + parser = argparse.ArgumentParser() + parser.add_argument( + "--base_model_name", default=None, type=str, help="Name of the huggingface pretrained base model." + ) + parser.add_argument("--config_path", default=None, type=str, help="Path to the huggingface classifier config.") + parser.add_argument("--checkpoint_path", default=None, type=str, help="Path to the s3prl checkpoint.") + parser.add_argument("--model_dump_path", default=None, type=str, help="Path to the final converted model.") + args = parser.parse_args() + convert_s3prl_checkpoint(args.base_model_name, args.config_path, args.checkpoint_path, args.model_dump_path) diff --git a/parrot/lib/python3.10/site-packages/transformers/models/hubert/modeling_hubert.py b/parrot/lib/python3.10/site-packages/transformers/models/hubert/modeling_hubert.py new file mode 100644 index 0000000000000000000000000000000000000000..3d1d0884c6aebc4f42ddf5529fdd45ebe855cbfb --- /dev/null +++ b/parrot/lib/python3.10/site-packages/transformers/models/hubert/modeling_hubert.py @@ -0,0 +1,1744 @@ +# coding=utf-8 +# Copyright 2021 The Fairseq Authors and the HuggingFace Inc. team. All rights reserved. +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. +""" PyTorch Hubert model.""" + +import warnings +from typing import Optional, Tuple, Union + +import numpy as np +import torch +import torch.nn.functional as F +import torch.utils.checkpoint +from torch import nn +from torch.nn import CrossEntropyLoss + +from ...activations import ACT2FN +from ...integrations.deepspeed import is_deepspeed_zero3_enabled +from ...modeling_outputs import BaseModelOutput, CausalLMOutput, SequenceClassifierOutput +from ...modeling_utils import PreTrainedModel +from ...utils import ( + add_code_sample_docstrings, + add_start_docstrings, + add_start_docstrings_to_model_forward, + is_flash_attn_2_available, + is_flash_attn_greater_or_equal_2_10, + logging, + replace_return_docstrings, +) +from .configuration_hubert import HubertConfig + + +if is_flash_attn_2_available(): + from flash_attn import flash_attn_func, flash_attn_varlen_func + from flash_attn.bert_padding import index_first_axis, pad_input, unpad_input # noqa + + +logger = logging.get_logger(__name__) + +_HIDDEN_STATES_START_POSITION = 1 + +# General docstring +_CONFIG_FOR_DOC = "HubertConfig" + +# Base docstring +_CHECKPOINT_FOR_DOC = "facebook/hubert-large-ls960-ft" +_EXPECTED_OUTPUT_SHAPE = [1, 292, 768] + +# CTC docstring +_CTC_EXPECTED_OUTPUT = "'MISTER QUILTER IS THE APOSTLE OF THE MIDDLE CLASSES AND WE ARE GLAD TO WELCOME HIS GOSPEL'" +_CTC_EXPECTED_LOSS = 22.68 + +# Audio class docstring +_SEQ_CLASS_CHECKPOINT = "superb/hubert-base-superb-ks" +_SEQ_CLASS_EXPECTED_OUTPUT = "'_unknown_'" +_SEQ_CLASS_EXPECTED_LOSS = 8.53 + + +# Copied from transformers.models.llama.modeling_llama._get_unpad_data +def _get_unpad_data(attention_mask): + seqlens_in_batch = attention_mask.sum(dim=-1, dtype=torch.int32) + indices = torch.nonzero(attention_mask.flatten(), as_tuple=False).flatten() + max_seqlen_in_batch = seqlens_in_batch.max().item() + cu_seqlens = F.pad(torch.cumsum(seqlens_in_batch, dim=0, dtype=torch.int32), (1, 0)) + return ( + indices, + cu_seqlens, + max_seqlen_in_batch, + ) + + +# Copied from transformers.models.wav2vec2.modeling_wav2vec2._compute_mask_indices +def _compute_mask_indices( + shape: Tuple[int, int], + mask_prob: float, + mask_length: int, + attention_mask: Optional[torch.LongTensor] = None, + min_masks: int = 0, +) -> np.ndarray: + """ + Computes random mask spans for a given shape. Used to implement [SpecAugment: A Simple Data Augmentation Method for + ASR](https://arxiv.org/abs/1904.08779). Note that this method is not optimized to run on TPU and should be run on + CPU as part of the preprocessing during training. + + Args: + shape: The shape for which to compute masks. This should be of a tuple of size 2 where + the first element is the batch size and the second element is the length of the axis to span. + mask_prob: The percentage of the whole axis (between 0 and 1) which will be masked. The number of + independently generated mask spans of length `mask_length` is computed by + `mask_prob*shape[1]/mask_length`. Note that due to overlaps, `mask_prob` is an upper bound and the + actual percentage will be smaller. + mask_length: size of the mask + min_masks: minimum number of masked spans + attention_mask: A (right-padded) attention mask which independently shortens the feature axis of + each batch dimension. + """ + batch_size, sequence_length = shape + + if mask_length < 1: + raise ValueError("`mask_length` has to be bigger than 0.") + + if mask_length > sequence_length: + raise ValueError( + f"`mask_length` has to be smaller than `sequence_length`, but got `mask_length`: {mask_length}" + f" and `sequence_length`: {sequence_length}`" + ) + + # epsilon is used for probabilistic rounding + epsilon = np.random.rand(1).item() + + def compute_num_masked_span(input_length): + """Given input length, compute how many spans should be masked""" + num_masked_span = int(mask_prob * input_length / mask_length + epsilon) + num_masked_span = max(num_masked_span, min_masks) + + # make sure num masked span <= sequence_length + if num_masked_span * mask_length > sequence_length: + num_masked_span = sequence_length // mask_length + + # make sure num_masked span is also <= input_length - (mask_length - 1) + if input_length - (mask_length - 1) < num_masked_span: + num_masked_span = max(input_length - (mask_length - 1), 0) + + return num_masked_span + + # compute number of masked spans in batch + input_lengths = ( + attention_mask.sum(-1).detach().tolist() + if attention_mask is not None + else [sequence_length for _ in range(batch_size)] + ) + + # SpecAugment mask to fill + spec_aug_mask = np.zeros((batch_size, sequence_length), dtype=bool) + spec_aug_mask_idxs = [] + + max_num_masked_span = compute_num_masked_span(sequence_length) + + if max_num_masked_span == 0: + return spec_aug_mask + + for input_length in input_lengths: + # compute num of masked spans for this input + num_masked_span = compute_num_masked_span(input_length) + + # get random indices to mask + spec_aug_mask_idx = np.random.choice( + np.arange(input_length - (mask_length - 1)), num_masked_span, replace=False + ) + + # pick first sampled index that will serve as a dummy index to pad vector + # to ensure same dimension for all batches due to probabilistic rounding + # Picking first sample just pads those vectors twice. + if len(spec_aug_mask_idx) == 0: + # this case can only happen if `input_length` is strictly smaller then + # `sequence_length` in which case the last token has to be a padding + # token which we can use as a dummy mask id + dummy_mask_idx = sequence_length - 1 + else: + dummy_mask_idx = spec_aug_mask_idx[0] + + spec_aug_mask_idx = np.concatenate( + [spec_aug_mask_idx, np.ones(max_num_masked_span - num_masked_span, dtype=np.int32) * dummy_mask_idx] + ) + spec_aug_mask_idxs.append(spec_aug_mask_idx) + + spec_aug_mask_idxs = np.array(spec_aug_mask_idxs) + + # expand masked indices to masked spans + spec_aug_mask_idxs = np.broadcast_to( + spec_aug_mask_idxs[:, :, None], (batch_size, max_num_masked_span, mask_length) + ) + spec_aug_mask_idxs = spec_aug_mask_idxs.reshape(batch_size, max_num_masked_span * mask_length) + + # add offset to the starting indexes so that indexes now create a span + offsets = np.arange(mask_length)[None, None, :] + offsets = np.broadcast_to(offsets, (batch_size, max_num_masked_span, mask_length)).reshape( + batch_size, max_num_masked_span * mask_length + ) + spec_aug_mask_idxs = spec_aug_mask_idxs + offsets + + # ensure that we cannot have indices larger than sequence_length + if spec_aug_mask_idxs.max() > sequence_length - 1: + spec_aug_mask_idxs[spec_aug_mask_idxs > sequence_length - 1] = sequence_length - 1 + + # scatter indices to mask + np.put_along_axis(spec_aug_mask, spec_aug_mask_idxs, 1, -1) + + return spec_aug_mask + + +# Copied from transformers.models.wav2vec2.modeling_wav2vec2.Wav2Vec2NoLayerNormConvLayer with Wav2Vec2->Hubert +class HubertNoLayerNormConvLayer(nn.Module): + def __init__(self, config, layer_id=0): + super().__init__() + self.in_conv_dim = config.conv_dim[layer_id - 1] if layer_id > 0 else 1 + self.out_conv_dim = config.conv_dim[layer_id] + + self.conv = nn.Conv1d( + self.in_conv_dim, + self.out_conv_dim, + kernel_size=config.conv_kernel[layer_id], + stride=config.conv_stride[layer_id], + bias=config.conv_bias, + ) + self.activation = ACT2FN[config.feat_extract_activation] + + def forward(self, hidden_states): + hidden_states = self.conv(hidden_states) + hidden_states = self.activation(hidden_states) + return hidden_states + + +# Copied from transformers.models.wav2vec2.modeling_wav2vec2.Wav2Vec2LayerNormConvLayer with Wav2Vec2->Hubert +class HubertLayerNormConvLayer(nn.Module): + def __init__(self, config, layer_id=0): + super().__init__() + self.in_conv_dim = config.conv_dim[layer_id - 1] if layer_id > 0 else 1 + self.out_conv_dim = config.conv_dim[layer_id] + + self.conv = nn.Conv1d( + self.in_conv_dim, + self.out_conv_dim, + kernel_size=config.conv_kernel[layer_id], + stride=config.conv_stride[layer_id], + bias=config.conv_bias, + ) + self.layer_norm = nn.LayerNorm(self.out_conv_dim, elementwise_affine=True) + self.activation = ACT2FN[config.feat_extract_activation] + + def forward(self, hidden_states): + hidden_states = self.conv(hidden_states) + + hidden_states = hidden_states.transpose(-2, -1) + hidden_states = self.layer_norm(hidden_states) + hidden_states = hidden_states.transpose(-2, -1) + + hidden_states = self.activation(hidden_states) + return hidden_states + + +# Copied from transformers.models.wav2vec2.modeling_wav2vec2.Wav2Vec2GroupNormConvLayer with Wav2Vec2->Hubert +class HubertGroupNormConvLayer(nn.Module): + def __init__(self, config, layer_id=0): + super().__init__() + self.in_conv_dim = config.conv_dim[layer_id - 1] if layer_id > 0 else 1 + self.out_conv_dim = config.conv_dim[layer_id] + + self.conv = nn.Conv1d( + self.in_conv_dim, + self.out_conv_dim, + kernel_size=config.conv_kernel[layer_id], + stride=config.conv_stride[layer_id], + bias=config.conv_bias, + ) + self.activation = ACT2FN[config.feat_extract_activation] + + self.layer_norm = nn.GroupNorm(num_groups=self.out_conv_dim, num_channels=self.out_conv_dim, affine=True) + + def forward(self, hidden_states): + hidden_states = self.conv(hidden_states) + hidden_states = self.layer_norm(hidden_states) + hidden_states = self.activation(hidden_states) + return hidden_states + + +# Copied from transformers.models.wav2vec2.modeling_wav2vec2.Wav2Vec2PositionalConvEmbedding with Wav2Vec2->Hubert +class HubertPositionalConvEmbedding(nn.Module): + def __init__(self, config): + super().__init__() + self.conv = nn.Conv1d( + config.hidden_size, + config.hidden_size, + kernel_size=config.num_conv_pos_embeddings, + padding=config.num_conv_pos_embeddings // 2, + groups=config.num_conv_pos_embedding_groups, + ) + + weight_norm = nn.utils.weight_norm + if hasattr(nn.utils.parametrizations, "weight_norm"): + weight_norm = nn.utils.parametrizations.weight_norm + + if is_deepspeed_zero3_enabled(): + import deepspeed + + with deepspeed.zero.GatheredParameters(self.conv.weight, modifier_rank=0): + self.conv = weight_norm(self.conv, name="weight", dim=2) + deepspeed.zero.register_external_parameter(self, self.conv.weight_v) + deepspeed.zero.register_external_parameter(self, self.conv.weight_g) + else: + self.conv = weight_norm(self.conv, name="weight", dim=2) + + self.padding = HubertSamePadLayer(config.num_conv_pos_embeddings) + self.activation = ACT2FN[config.feat_extract_activation] + + def forward(self, hidden_states): + hidden_states = hidden_states.transpose(1, 2) + + hidden_states = self.conv(hidden_states) + hidden_states = self.padding(hidden_states) + hidden_states = self.activation(hidden_states) + + hidden_states = hidden_states.transpose(1, 2) + return hidden_states + + +# Copied from transformers.models.wav2vec2.modeling_wav2vec2.Wav2Vec2SamePadLayer with Wav2Vec2->Hubert +class HubertSamePadLayer(nn.Module): + def __init__(self, num_conv_pos_embeddings): + super().__init__() + self.num_pad_remove = 1 if num_conv_pos_embeddings % 2 == 0 else 0 + + def forward(self, hidden_states): + if self.num_pad_remove > 0: + hidden_states = hidden_states[:, :, : -self.num_pad_remove] + return hidden_states + + +# Copied from transformers.models.wav2vec2.modeling_wav2vec2.Wav2Vec2FeatureEncoder with Wav2Vec2->Hubert +class HubertFeatureEncoder(nn.Module): + """Construct the features from raw audio waveform""" + + def __init__(self, config): + super().__init__() + + if config.feat_extract_norm == "group": + conv_layers = [HubertGroupNormConvLayer(config, layer_id=0)] + [ + HubertNoLayerNormConvLayer(config, layer_id=i + 1) for i in range(config.num_feat_extract_layers - 1) + ] + elif config.feat_extract_norm == "layer": + conv_layers = [HubertLayerNormConvLayer(config, layer_id=i) for i in range(config.num_feat_extract_layers)] + else: + raise ValueError( + f"`config.feat_extract_norm` is {config.feat_extract_norm}, but has to be one of ['group', 'layer']" + ) + self.conv_layers = nn.ModuleList(conv_layers) + self.gradient_checkpointing = False + self._requires_grad = True + + def _freeze_parameters(self): + for param in self.parameters(): + param.requires_grad = False + self._requires_grad = False + + def forward(self, input_values): + hidden_states = input_values[:, None] + + # make sure hidden_states require grad for gradient_checkpointing + if self._requires_grad and self.training: + hidden_states.requires_grad = True + + for conv_layer in self.conv_layers: + if self._requires_grad and self.gradient_checkpointing and self.training: + hidden_states = self._gradient_checkpointing_func( + conv_layer.__call__, + hidden_states, + ) + else: + hidden_states = conv_layer(hidden_states) + + return hidden_states + + +class HubertFeatureExtractor(HubertFeatureEncoder): + def __init__(self, config): + super().__init__(config) + warnings.warn( + f"The class `{self.__class__.__name__}` has been depreciated " + "and will be removed in Transformers v5. " + f"Use `{self.__class__.__bases__[0].__name__}` instead.", + FutureWarning, + ) + + +class HubertFeatureProjection(nn.Module): + def __init__(self, config): + super().__init__() + self.feat_proj_layer_norm = config.feat_proj_layer_norm + if self.feat_proj_layer_norm: + self.layer_norm = nn.LayerNorm(config.conv_dim[-1], eps=config.layer_norm_eps) + self.projection = nn.Linear(config.conv_dim[-1], config.hidden_size) + self.dropout = nn.Dropout(config.feat_proj_dropout) + + def forward(self, hidden_states): + # non-projected hidden states are needed for quantization + if self.feat_proj_layer_norm: + hidden_states = self.layer_norm(hidden_states) + hidden_states = self.projection(hidden_states) + hidden_states = self.dropout(hidden_states) + return hidden_states + + +# Copied from transformers.models.bart.modeling_bart.BartAttention with Bart->Hubert +class HubertAttention(nn.Module): + """Multi-headed attention from 'Attention Is All You Need' paper""" + + def __init__( + self, + embed_dim: int, + num_heads: int, + dropout: float = 0.0, + is_decoder: bool = False, + bias: bool = True, + is_causal: bool = False, + config: Optional[HubertConfig] = None, + ): + super().__init__() + self.embed_dim = embed_dim + self.num_heads = num_heads + self.dropout = dropout + self.head_dim = embed_dim // num_heads + self.config = config + + if (self.head_dim * num_heads) != self.embed_dim: + raise ValueError( + f"embed_dim must be divisible by num_heads (got `embed_dim`: {self.embed_dim}" + f" and `num_heads`: {num_heads})." + ) + self.scaling = self.head_dim**-0.5 + self.is_decoder = is_decoder + self.is_causal = is_causal + + self.k_proj = nn.Linear(embed_dim, embed_dim, bias=bias) + self.v_proj = nn.Linear(embed_dim, embed_dim, bias=bias) + self.q_proj = nn.Linear(embed_dim, embed_dim, bias=bias) + self.out_proj = nn.Linear(embed_dim, embed_dim, bias=bias) + + def _shape(self, tensor: torch.Tensor, seq_len: int, bsz: int): + return tensor.view(bsz, seq_len, self.num_heads, self.head_dim).transpose(1, 2).contiguous() + + def forward( + self, + hidden_states: torch.Tensor, + key_value_states: Optional[torch.Tensor] = None, + past_key_value: Optional[Tuple[torch.Tensor]] = None, + attention_mask: Optional[torch.Tensor] = None, + layer_head_mask: Optional[torch.Tensor] = None, + output_attentions: bool = False, + ) -> Tuple[torch.Tensor, Optional[torch.Tensor], Optional[Tuple[torch.Tensor]]]: + """Input shape: Batch x Time x Channel""" + + # if key_value_states are provided this layer is used as a cross-attention layer + # for the decoder + is_cross_attention = key_value_states is not None + + bsz, tgt_len, _ = hidden_states.size() + + # get query proj + query_states = self.q_proj(hidden_states) * self.scaling + # get key, value proj + # `past_key_value[0].shape[2] == key_value_states.shape[1]` + # is checking that the `sequence_length` of the `past_key_value` is the same as + # the provided `key_value_states` to support prefix tuning + if ( + is_cross_attention + and past_key_value is not None + and past_key_value[0].shape[2] == key_value_states.shape[1] + ): + # reuse k,v, cross_attentions + key_states = past_key_value[0] + value_states = past_key_value[1] + elif is_cross_attention: + # cross_attentions + key_states = self._shape(self.k_proj(key_value_states), -1, bsz) + value_states = self._shape(self.v_proj(key_value_states), -1, bsz) + elif past_key_value is not None: + # reuse k, v, self_attention + key_states = self._shape(self.k_proj(hidden_states), -1, bsz) + value_states = self._shape(self.v_proj(hidden_states), -1, bsz) + key_states = torch.cat([past_key_value[0], key_states], dim=2) + value_states = torch.cat([past_key_value[1], value_states], dim=2) + else: + # self_attention + key_states = self._shape(self.k_proj(hidden_states), -1, bsz) + value_states = self._shape(self.v_proj(hidden_states), -1, bsz) + + if self.is_decoder: + # if cross_attention save Tuple(torch.Tensor, torch.Tensor) of all cross attention key/value_states. + # Further calls to cross_attention layer can then reuse all cross-attention + # key/value_states (first "if" case) + # if uni-directional self-attention (decoder) save Tuple(torch.Tensor, torch.Tensor) of + # all previous decoder key/value_states. Further calls to uni-directional self-attention + # can concat previous decoder key/value_states to current projected key/value_states (third "elif" case) + # if encoder bi-directional self-attention `past_key_value` is always `None` + past_key_value = (key_states, value_states) + + proj_shape = (bsz * self.num_heads, -1, self.head_dim) + query_states = self._shape(query_states, tgt_len, bsz).view(*proj_shape) + key_states = key_states.reshape(*proj_shape) + value_states = value_states.reshape(*proj_shape) + + src_len = key_states.size(1) + attn_weights = torch.bmm(query_states, key_states.transpose(1, 2)) + + if attn_weights.size() != (bsz * self.num_heads, tgt_len, src_len): + raise ValueError( + f"Attention weights should be of size {(bsz * self.num_heads, tgt_len, src_len)}, but is" + f" {attn_weights.size()}" + ) + + if attention_mask is not None: + if attention_mask.size() != (bsz, 1, tgt_len, src_len): + raise ValueError( + f"Attention mask should be of size {(bsz, 1, tgt_len, src_len)}, but is {attention_mask.size()}" + ) + attn_weights = attn_weights.view(bsz, self.num_heads, tgt_len, src_len) + attention_mask + attn_weights = attn_weights.view(bsz * self.num_heads, tgt_len, src_len) + + attn_weights = nn.functional.softmax(attn_weights, dim=-1) + + if layer_head_mask is not None: + if layer_head_mask.size() != (self.num_heads,): + raise ValueError( + f"Head mask for a single layer should be of size {(self.num_heads,)}, but is" + f" {layer_head_mask.size()}" + ) + attn_weights = layer_head_mask.view(1, -1, 1, 1) * attn_weights.view(bsz, self.num_heads, tgt_len, src_len) + attn_weights = attn_weights.view(bsz * self.num_heads, tgt_len, src_len) + + if output_attentions: + # this operation is a bit awkward, but it's required to + # make sure that attn_weights keeps its gradient. + # In order to do so, attn_weights have to be reshaped + # twice and have to be reused in the following + attn_weights_reshaped = attn_weights.view(bsz, self.num_heads, tgt_len, src_len) + attn_weights = attn_weights_reshaped.view(bsz * self.num_heads, tgt_len, src_len) + else: + attn_weights_reshaped = None + + attn_probs = nn.functional.dropout(attn_weights, p=self.dropout, training=self.training) + + attn_output = torch.bmm(attn_probs, value_states) + + if attn_output.size() != (bsz * self.num_heads, tgt_len, self.head_dim): + raise ValueError( + f"`attn_output` should be of size {(bsz * self.num_heads, tgt_len, self.head_dim)}, but is" + f" {attn_output.size()}" + ) + + attn_output = attn_output.view(bsz, self.num_heads, tgt_len, self.head_dim) + attn_output = attn_output.transpose(1, 2) + + # Use the `embed_dim` from the config (stored in the class) rather than `hidden_state` because `attn_output` can be + # partitioned across GPUs when using tensor-parallelism. + attn_output = attn_output.reshape(bsz, tgt_len, self.embed_dim) + + attn_output = self.out_proj(attn_output) + + return attn_output, attn_weights_reshaped, past_key_value + + +# Copied from transformers.models.bart.modeling_bart.BartFlashAttention2 with Bart->Hubert +class HubertFlashAttention2(HubertAttention): + """ + Hubert flash attention module. This module inherits from `HubertAttention` as the weights of the module stays + untouched. The only required change would be on the forward pass where it needs to correctly call the public API of + flash attention and deal with padding tokens in case the input contains any of them. + """ + + # Copied from transformers.models.llama.modeling_llama.LlamaFlashAttention2.__init__ + def __init__(self, *args, **kwargs): + super().__init__(*args, **kwargs) + + # TODO: Should be removed once Flash Attention for RoCm is bumped to 2.1. + # flash_attn<2.1 generates top-left aligned causal mask, while what is needed here is bottom-right alignement, that was made default for flash_attn>=2.1. This attribute is used to handle this difference. Reference: https://github.com/Dao-AILab/flash-attention/releases/tag/v2.1.0. + # Beware that with flash_attn<2.1, using q_seqlen != k_seqlen (except for the case q_seqlen == 1) produces a wrong mask (top-left). + self._flash_attn_uses_top_left_mask = not is_flash_attn_greater_or_equal_2_10() + + def _reshape(self, tensor: torch.Tensor, seq_len: int, bsz: int): + return tensor.view(bsz, seq_len, self.num_heads, self.head_dim) + + def forward( + self, + hidden_states: torch.Tensor, + key_value_states: Optional[torch.Tensor] = None, + past_key_value: Optional[Tuple[torch.Tensor]] = None, + attention_mask: Optional[torch.Tensor] = None, + layer_head_mask: Optional[torch.Tensor] = None, + output_attentions: bool = False, + ) -> Tuple[torch.Tensor, Optional[torch.Tensor], Optional[Tuple[torch.Tensor]]]: + # HubertFlashAttention2 attention does not support output_attentions + if output_attentions: + raise ValueError("HubertFlashAttention2 attention does not support output_attentions") + + # if key_value_states are provided this layer is used as a cross-attention layer + # for the decoder + is_cross_attention = key_value_states is not None + + bsz, q_len, _ = hidden_states.size() + + # get query proj + query_states = self._reshape(self.q_proj(hidden_states), -1, bsz) + # get key, value proj + # `past_key_value[0].shape[2] == key_value_states.shape[1]` + # is checking that the `sequence_length` of the `past_key_value` is the same as + # the provided `key_value_states` to support prefix tuning + if ( + is_cross_attention + and past_key_value is not None + and past_key_value[0].shape[2] == key_value_states.shape[1] + ): + # reuse k,v, cross_attentions + key_states = past_key_value[0].transpose(1, 2) + value_states = past_key_value[1].transpose(1, 2) + elif is_cross_attention: + # cross_attentions + key_states = self._reshape(self.k_proj(key_value_states), -1, bsz) + value_states = self._reshape(self.v_proj(key_value_states), -1, bsz) + elif past_key_value is not None: + # reuse k, v, self_attention + key_states = self._reshape(self.k_proj(hidden_states), -1, bsz) + value_states = self._reshape(self.v_proj(hidden_states), -1, bsz) + key_states = torch.cat([past_key_value[0].transpose(1, 2), key_states], dim=1) + value_states = torch.cat([past_key_value[1].transpose(1, 2), value_states], dim=1) + else: + # self_attention + key_states = self._reshape(self.k_proj(hidden_states), -1, bsz) + value_states = self._reshape(self.v_proj(hidden_states), -1, bsz) + + if self.is_decoder: + # if cross_attention save Tuple(torch.Tensor, torch.Tensor) of all cross attention key/value_states. + # Further calls to cross_attention layer can then reuse all cross-attention + # key/value_states (first "if" case) + # if uni-directional self-attention (decoder) save Tuple(torch.Tensor, torch.Tensor) of + # all previous decoder key/value_states. Further calls to uni-directional self-attention + # can concat previous decoder key/value_states to current projected key/value_states (third "elif" case) + # if encoder bi-directional self-attention `past_key_value` is always `None` + past_key_value = (key_states.transpose(1, 2), value_states.transpose(1, 2)) + + kv_seq_len = key_states.shape[-2] + if past_key_value is not None: + kv_seq_len += past_key_value[0].shape[-2] + + # In PEFT, usually we cast the layer norms in float32 for training stability reasons + # therefore the input hidden states gets silently casted in float32. Hence, we need + # cast them back in the correct dtype just to be sure everything works as expected. + # This might slowdown training & inference so it is recommended to not cast the LayerNorms + # in fp32. (LlamaRMSNorm handles it correctly) + + input_dtype = query_states.dtype + if input_dtype == torch.float32: + if torch.is_autocast_enabled(): + target_dtype = torch.get_autocast_gpu_dtype() + # Handle the case where the model is quantized + elif hasattr(self.config, "_pre_quantization_dtype"): + target_dtype = self.config._pre_quantization_dtype + else: + target_dtype = self.q_proj.weight.dtype + + logger.warning_once( + f"The input hidden states seems to be silently casted in float32, this might be related to" + f" the fact you have upcasted embedding or layer norm layers in float32. We will cast back the input in" + f" {target_dtype}." + ) + + query_states = query_states.to(target_dtype) + key_states = key_states.to(target_dtype) + value_states = value_states.to(target_dtype) + + attn_output = self._flash_attention_forward( + query_states, key_states, value_states, attention_mask, q_len, dropout=self.dropout + ) + + attn_output = attn_output.reshape(bsz, q_len, -1) + attn_output = self.out_proj(attn_output) + + if not output_attentions: + attn_weights = None + + return attn_output, attn_weights, past_key_value + + # Copied from transformers.models.llama.modeling_llama.LlamaFlashAttention2._flash_attention_forward + def _flash_attention_forward( + self, query_states, key_states, value_states, attention_mask, query_length, dropout=0.0, softmax_scale=None + ): + """ + Calls the forward method of Flash Attention - if the input hidden states contain at least one padding token + first unpad the input, then computes the attention scores and pad the final attention scores. + Args: + query_states (`torch.Tensor`): + Input query states to be passed to Flash Attention API + key_states (`torch.Tensor`): + Input key states to be passed to Flash Attention API + value_states (`torch.Tensor`): + Input value states to be passed to Flash Attention API + attention_mask (`torch.Tensor`): + The padding mask - corresponds to a tensor of size `(batch_size, seq_len)` where 0 stands for the + position of padding tokens and 1 for the position of non-padding tokens. + dropout (`float`): + Attention dropout + softmax_scale (`float`, *optional*): + The scaling of QK^T before applying softmax. Default to 1 / sqrt(head_dim) + """ + if not self._flash_attn_uses_top_left_mask: + causal = self.is_causal + else: + # TODO: Remove the `query_length != 1` check once Flash Attention for RoCm is bumped to 2.1. For details, please see the comment in LlamaFlashAttention2 __init__. + causal = self.is_causal and query_length != 1 + + # Contains at least one padding token in the sequence + if attention_mask is not None: + batch_size = query_states.shape[0] + query_states, key_states, value_states, indices_q, cu_seq_lens, max_seq_lens = self._upad_input( + query_states, key_states, value_states, attention_mask, query_length + ) + + cu_seqlens_q, cu_seqlens_k = cu_seq_lens + max_seqlen_in_batch_q, max_seqlen_in_batch_k = max_seq_lens + + attn_output_unpad = flash_attn_varlen_func( + query_states, + key_states, + value_states, + cu_seqlens_q=cu_seqlens_q, + cu_seqlens_k=cu_seqlens_k, + max_seqlen_q=max_seqlen_in_batch_q, + max_seqlen_k=max_seqlen_in_batch_k, + dropout_p=dropout, + softmax_scale=softmax_scale, + causal=causal, + ) + + attn_output = pad_input(attn_output_unpad, indices_q, batch_size, query_length) + else: + attn_output = flash_attn_func( + query_states, key_states, value_states, dropout, softmax_scale=softmax_scale, causal=causal + ) + + return attn_output + + # Copied from transformers.models.llama.modeling_llama.LlamaFlashAttention2._upad_input + def _upad_input(self, query_layer, key_layer, value_layer, attention_mask, query_length): + indices_k, cu_seqlens_k, max_seqlen_in_batch_k = _get_unpad_data(attention_mask) + batch_size, kv_seq_len, num_key_value_heads, head_dim = key_layer.shape + + key_layer = index_first_axis( + key_layer.reshape(batch_size * kv_seq_len, num_key_value_heads, head_dim), indices_k + ) + value_layer = index_first_axis( + value_layer.reshape(batch_size * kv_seq_len, num_key_value_heads, head_dim), indices_k + ) + if query_length == kv_seq_len: + query_layer = index_first_axis( + query_layer.reshape(batch_size * kv_seq_len, self.num_heads, head_dim), indices_k + ) + cu_seqlens_q = cu_seqlens_k + max_seqlen_in_batch_q = max_seqlen_in_batch_k + indices_q = indices_k + elif query_length == 1: + max_seqlen_in_batch_q = 1 + cu_seqlens_q = torch.arange( + batch_size + 1, dtype=torch.int32, device=query_layer.device + ) # There is a memcpy here, that is very bad. + indices_q = cu_seqlens_q[:-1] + query_layer = query_layer.squeeze(1) + else: + # The -q_len: slice assumes left padding. + attention_mask = attention_mask[:, -query_length:] + query_layer, indices_q, cu_seqlens_q, max_seqlen_in_batch_q = unpad_input(query_layer, attention_mask) + + return ( + query_layer, + key_layer, + value_layer, + indices_q, + (cu_seqlens_q, cu_seqlens_k), + (max_seqlen_in_batch_q, max_seqlen_in_batch_k), + ) + + +class HubertSdpaAttention(HubertAttention): + # Copied from transformers.models.bart.modeling_bart.BartSdpaAttention.forward with Bart->Hubert + def forward( + self, + hidden_states: torch.Tensor, + key_value_states: Optional[torch.Tensor] = None, + past_key_value: Optional[Tuple[torch.Tensor]] = None, + attention_mask: Optional[torch.Tensor] = None, + layer_head_mask: Optional[torch.Tensor] = None, + output_attentions: bool = False, + ) -> Tuple[torch.Tensor, Optional[torch.Tensor], Optional[Tuple[torch.Tensor]]]: + """Input shape: Batch x Time x Channel""" + if output_attentions or layer_head_mask is not None: + # TODO: Improve this warning with e.g. `model.config._attn_implementation = "manual"` once this is implemented. + logger.warning_once( + "HubertModel is using HubertSdpaAttention, but `torch.nn.functional.scaled_dot_product_attention` does not support `output_attentions=True` or `layer_head_mask` not None. Falling back to the manual attention" + ' implementation, but specifying the manual implementation will be required from Transformers version v5.0.0 onwards. This warning can be removed using the argument `attn_implementation="eager"` when loading the model.' + ) + return super().forward( + hidden_states, + key_value_states=key_value_states, + past_key_value=past_key_value, + attention_mask=attention_mask, + layer_head_mask=layer_head_mask, + output_attentions=output_attentions, + ) + + # if key_value_states are provided this layer is used as a cross-attention layer + # for the decoder + is_cross_attention = key_value_states is not None + + bsz, tgt_len, _ = hidden_states.size() + + # get query proj + query_states = self.q_proj(hidden_states) + # get key, value proj + # `past_key_value[0].shape[2] == key_value_states.shape[1]` + # is checking that the `sequence_length` of the `past_key_value` is the same as + # the provided `key_value_states` to support prefix tuning + if ( + is_cross_attention + and past_key_value is not None + and past_key_value[0].shape[2] == key_value_states.shape[1] + ): + # reuse k,v, cross_attentions + key_states = past_key_value[0] + value_states = past_key_value[1] + elif is_cross_attention: + # cross_attentions + key_states = self._shape(self.k_proj(key_value_states), -1, bsz) + value_states = self._shape(self.v_proj(key_value_states), -1, bsz) + elif past_key_value is not None: + # reuse k, v, self_attention + key_states = self._shape(self.k_proj(hidden_states), -1, bsz) + value_states = self._shape(self.v_proj(hidden_states), -1, bsz) + key_states = torch.cat([past_key_value[0], key_states], dim=2) + value_states = torch.cat([past_key_value[1], value_states], dim=2) + else: + # self_attention + key_states = self._shape(self.k_proj(hidden_states), -1, bsz) + value_states = self._shape(self.v_proj(hidden_states), -1, bsz) + + if self.is_decoder: + # if cross_attention save Tuple(torch.Tensor, torch.Tensor) of all cross attention key/value_states. + # Further calls to cross_attention layer can then reuse all cross-attention + # key/value_states (first "if" case) + # if uni-directional self-attention (decoder) save Tuple(torch.Tensor, torch.Tensor) of + # all previous decoder key/value_states. Further calls to uni-directional self-attention + # can concat previous decoder key/value_states to current projected key/value_states (third "elif" case) + # if encoder bi-directional self-attention `past_key_value` is always `None` + past_key_value = (key_states, value_states) + + query_states = self._shape(query_states, tgt_len, bsz) + + # NOTE: SDPA with memory-efficient backend is currently (torch==2.1.2) bugged when using non-contiguous inputs and a custom attn_mask, + # but we are fine here as `_shape` do call `.contiguous()`. Reference: https://github.com/pytorch/pytorch/issues/112577 + attn_output = torch.nn.functional.scaled_dot_product_attention( + query_states, + key_states, + value_states, + attn_mask=attention_mask, + dropout_p=self.dropout if self.training else 0.0, + # The tgt_len > 1 is necessary to match with AttentionMaskConverter.to_causal_4d that does not create a causal mask in case tgt_len == 1. + is_causal=self.is_causal and attention_mask is None and tgt_len > 1, + ) + + if attn_output.size() != (bsz, self.num_heads, tgt_len, self.head_dim): + raise ValueError( + f"`attn_output` should be of size {(bsz, self.num_heads, tgt_len, self.head_dim)}, but is" + f" {attn_output.size()}" + ) + + attn_output = attn_output.transpose(1, 2) + + # Use the `embed_dim` from the config (stored in the class) rather than `hidden_state` because `attn_output` can be + # partitioned across GPUs when using tensor-parallelism. + attn_output = attn_output.reshape(bsz, tgt_len, self.embed_dim) + + attn_output = self.out_proj(attn_output) + + return attn_output, None, past_key_value + + +HUBERT_ATTENTION_CLASSES = { + "eager": HubertAttention, + "sdpa": HubertSdpaAttention, + "flash_attention_2": HubertFlashAttention2, +} + + +# Copied from transformers.models.wav2vec2.modeling_wav2vec2.Wav2Vec2FeedForward with Wav2Vec2->Hubert +class HubertFeedForward(nn.Module): + def __init__(self, config): + super().__init__() + self.intermediate_dropout = nn.Dropout(config.activation_dropout) + + self.intermediate_dense = nn.Linear(config.hidden_size, config.intermediate_size) + if isinstance(config.hidden_act, str): + self.intermediate_act_fn = ACT2FN[config.hidden_act] + else: + self.intermediate_act_fn = config.hidden_act + + self.output_dense = nn.Linear(config.intermediate_size, config.hidden_size) + self.output_dropout = nn.Dropout(config.hidden_dropout) + + def forward(self, hidden_states): + hidden_states = self.intermediate_dense(hidden_states) + hidden_states = self.intermediate_act_fn(hidden_states) + hidden_states = self.intermediate_dropout(hidden_states) + + hidden_states = self.output_dense(hidden_states) + hidden_states = self.output_dropout(hidden_states) + return hidden_states + + +# Copied from transformers.models.wav2vec2.modeling_wav2vec2.Wav2Vec2EncoderLayer with Wav2Vec2->Hubert, WAV2VEC2->HUBERT +class HubertEncoderLayer(nn.Module): + def __init__(self, config): + super().__init__() + self.attention = HUBERT_ATTENTION_CLASSES[config._attn_implementation]( + embed_dim=config.hidden_size, + num_heads=config.num_attention_heads, + dropout=config.attention_dropout, + is_decoder=False, + ) + + self.dropout = nn.Dropout(config.hidden_dropout) + self.layer_norm = nn.LayerNorm(config.hidden_size, eps=config.layer_norm_eps) + self.feed_forward = HubertFeedForward(config) + self.final_layer_norm = nn.LayerNorm(config.hidden_size, eps=config.layer_norm_eps) + + def forward(self, hidden_states, attention_mask=None, output_attentions=False): + attn_residual = hidden_states + hidden_states, attn_weights, _ = self.attention( + hidden_states, attention_mask=attention_mask, output_attentions=output_attentions + ) + hidden_states = self.dropout(hidden_states) + hidden_states = attn_residual + hidden_states + + hidden_states = self.layer_norm(hidden_states) + hidden_states = hidden_states + self.feed_forward(hidden_states) + hidden_states = self.final_layer_norm(hidden_states) + + outputs = (hidden_states,) + + if output_attentions: + outputs += (attn_weights,) + + return outputs + + +# Copied from transformers.models.wav2vec2.modeling_wav2vec2.Wav2Vec2AttnAdapterLayer with Wav2Vec2->Hubert +class HubertAttnAdapterLayer(nn.Module): + def __init__(self, config): + """ + Implements adapter modules directly with 3D tensor weight as parameters and without using ModuleList to speed + up training throughput. + """ + super().__init__() + self.input_dim = config.adapter_attn_dim + self.hidden_dim = config.hidden_size + + self.norm = nn.LayerNorm(self.hidden_dim) + self.linear_1 = nn.Linear(self.hidden_dim, self.input_dim) + self.act_fn = nn.ReLU() + self.linear_2 = nn.Linear(self.input_dim, self.hidden_dim) + + def forward(self, hidden_states: torch.FloatTensor): + hidden_states = self.norm(hidden_states) + + hidden_states = self.linear_1(hidden_states) + hidden_states = self.act_fn(hidden_states) + hidden_states = self.linear_2(hidden_states) + + return hidden_states + + +# Copied from transformers.models.wav2vec2.modeling_wav2vec2.Wav2Vec2EncoderLayerStableLayerNorm with Wav2Vec2->Hubert, WAV2VEC2->HUBERT +class HubertEncoderLayerStableLayerNorm(nn.Module): + def __init__(self, config): + super().__init__() + self.attention = HUBERT_ATTENTION_CLASSES[config._attn_implementation]( + embed_dim=config.hidden_size, + num_heads=config.num_attention_heads, + dropout=config.attention_dropout, + is_decoder=False, + ) + self.dropout = nn.Dropout(config.hidden_dropout) + self.layer_norm = nn.LayerNorm(config.hidden_size, eps=config.layer_norm_eps) + self.feed_forward = HubertFeedForward(config) + self.final_layer_norm = nn.LayerNorm(config.hidden_size, eps=config.layer_norm_eps) + + if getattr(config, "adapter_attn_dim", None) is not None: + self.adapter_layer = HubertAttnAdapterLayer(config) + else: + self.adapter_layer = None + + def forward( + self, + hidden_states: torch.Tensor, + attention_mask: Optional[torch.Tensor] = None, + output_attentions: bool = False, + ): + attn_residual = hidden_states + hidden_states = self.layer_norm(hidden_states) + hidden_states, attn_weights, _ = self.attention( + hidden_states, attention_mask=attention_mask, output_attentions=output_attentions + ) + hidden_states = self.dropout(hidden_states) + hidden_states = attn_residual + hidden_states + hidden_states = hidden_states + self.feed_forward(self.final_layer_norm(hidden_states)) + + if self.adapter_layer is not None: + hidden_states = hidden_states + self.adapter_layer(hidden_states) + + outputs = (hidden_states,) + + if output_attentions: + outputs += (attn_weights,) + + return outputs + + +# Copied from transformers.models.wav2vec2.modeling_wav2vec2.Wav2Vec2Encoder with Wav2Vec2->Hubert +class HubertEncoder(nn.Module): + def __init__(self, config): + super().__init__() + self.config = config + self.pos_conv_embed = HubertPositionalConvEmbedding(config) + self.layer_norm = nn.LayerNorm(config.hidden_size, eps=config.layer_norm_eps) + self.dropout = nn.Dropout(config.hidden_dropout) + self.layers = nn.ModuleList([HubertEncoderLayer(config) for _ in range(config.num_hidden_layers)]) + self.gradient_checkpointing = False + self._use_flash_attention_2 = config._attn_implementation == "flash_attention_2" + + def forward( + self, + hidden_states: torch.tensor, + attention_mask: Optional[torch.Tensor] = None, + output_attentions: bool = False, + output_hidden_states: bool = False, + return_dict: bool = True, + ): + all_hidden_states = () if output_hidden_states else None + all_self_attentions = () if output_attentions else None + + if attention_mask is not None: + # make sure padded tokens output 0 + expand_attention_mask = attention_mask.unsqueeze(-1).repeat(1, 1, hidden_states.shape[2]) + hidden_states[~expand_attention_mask] = 0 + if self._use_flash_attention_2: + # 2d mask is passed through the layers + attention_mask = attention_mask if (attention_mask is not None and 0 in attention_mask) else None + else: + # extend attention_mask + attention_mask = 1.0 - attention_mask[:, None, None, :].to(dtype=hidden_states.dtype) + attention_mask = attention_mask * torch.finfo(hidden_states.dtype).min + attention_mask = attention_mask.expand( + attention_mask.shape[0], 1, attention_mask.shape[-1], attention_mask.shape[-1] + ) + + position_embeddings = self.pos_conv_embed(hidden_states) + hidden_states = hidden_states + position_embeddings + hidden_states = self.layer_norm(hidden_states) + hidden_states = self.dropout(hidden_states) + + deepspeed_zero3_is_enabled = is_deepspeed_zero3_enabled() + + for layer in self.layers: + if output_hidden_states: + all_hidden_states = all_hidden_states + (hidden_states,) + + # add LayerDrop (see https://arxiv.org/abs/1909.11556 for description) + dropout_probability = torch.rand([]) + + skip_the_layer = True if self.training and (dropout_probability < self.config.layerdrop) else False + if not skip_the_layer or deepspeed_zero3_is_enabled: + # under deepspeed zero3 all gpus must run in sync + if self.gradient_checkpointing and self.training: + layer_outputs = self._gradient_checkpointing_func( + layer.__call__, + hidden_states, + attention_mask, + output_attentions, + ) + else: + layer_outputs = layer( + hidden_states, attention_mask=attention_mask, output_attentions=output_attentions + ) + hidden_states = layer_outputs[0] + + if skip_the_layer: + layer_outputs = (None, None) + + if output_attentions: + all_self_attentions = all_self_attentions + (layer_outputs[1],) + + if output_hidden_states: + all_hidden_states = all_hidden_states + (hidden_states,) + + if not return_dict: + return tuple(v for v in [hidden_states, all_hidden_states, all_self_attentions] if v is not None) + return BaseModelOutput( + last_hidden_state=hidden_states, + hidden_states=all_hidden_states, + attentions=all_self_attentions, + ) + + +# Copied from transformers.models.wav2vec2.modeling_wav2vec2.Wav2Vec2EncoderStableLayerNorm with Wav2Vec2->Hubert +class HubertEncoderStableLayerNorm(nn.Module): + def __init__(self, config): + super().__init__() + self.config = config + self.pos_conv_embed = HubertPositionalConvEmbedding(config) + self.layer_norm = nn.LayerNorm(config.hidden_size, eps=config.layer_norm_eps) + self.dropout = nn.Dropout(config.hidden_dropout) + self.layers = nn.ModuleList( + [HubertEncoderLayerStableLayerNorm(config) for _ in range(config.num_hidden_layers)] + ) + self.gradient_checkpointing = False + self._use_flash_attention_2 = config._attn_implementation == "flash_attention_2" + + def forward( + self, + hidden_states, + attention_mask=None, + output_attentions=False, + output_hidden_states=False, + return_dict=True, + ): + all_hidden_states = () if output_hidden_states else None + all_self_attentions = () if output_attentions else None + + if attention_mask is not None: + # make sure padded tokens are not attended to + expand_attention_mask = attention_mask.unsqueeze(-1).repeat(1, 1, hidden_states.shape[2]) + hidden_states[~expand_attention_mask] = 0 + if self._use_flash_attention_2: + # 2d mask is passed through the layers + attention_mask = attention_mask if (attention_mask is not None and 0 in attention_mask) else None + else: + # extend attention_mask + attention_mask = 1.0 - attention_mask[:, None, None, :].to(dtype=hidden_states.dtype) + attention_mask = attention_mask * torch.finfo(hidden_states.dtype).min + attention_mask = attention_mask.expand( + attention_mask.shape[0], 1, attention_mask.shape[-1], attention_mask.shape[-1] + ) + + position_embeddings = self.pos_conv_embed(hidden_states) + hidden_states = hidden_states + position_embeddings + hidden_states = self.dropout(hidden_states) + + deepspeed_zero3_is_enabled = is_deepspeed_zero3_enabled() + + for layer in self.layers: + if output_hidden_states: + all_hidden_states = all_hidden_states + (hidden_states,) + + # add LayerDrop (see https://arxiv.org/abs/1909.11556 for description) + dropout_probability = torch.rand([]) + + skip_the_layer = True if self.training and (dropout_probability < self.config.layerdrop) else False + if not skip_the_layer or deepspeed_zero3_is_enabled: + # under deepspeed zero3 all gpus must run in sync + # XXX: could optimize this like synced_gpus in generate_utils but not sure if it's worth the code complication + if self.gradient_checkpointing and self.training: + layer_outputs = self._gradient_checkpointing_func( + layer.__call__, + hidden_states, + attention_mask, + output_attentions, + ) + else: + layer_outputs = layer( + hidden_states, attention_mask=attention_mask, output_attentions=output_attentions + ) + hidden_states = layer_outputs[0] + + if skip_the_layer: + layer_outputs = (None, None) + + if output_attentions: + all_self_attentions = all_self_attentions + (layer_outputs[1],) + + hidden_states = self.layer_norm(hidden_states) + + if output_hidden_states: + all_hidden_states = all_hidden_states + (hidden_states,) + + if not return_dict: + return tuple(v for v in [hidden_states, all_hidden_states, all_self_attentions] if v is not None) + return BaseModelOutput( + last_hidden_state=hidden_states, + hidden_states=all_hidden_states, + attentions=all_self_attentions, + ) + + +class HubertPreTrainedModel(PreTrainedModel): + """ + An abstract class to handle weights initialization and a simple interface for downloading and loading pretrained + models. + """ + + config_class = HubertConfig + base_model_prefix = "hubert" + main_input_name = "input_values" + supports_gradient_checkpointing = True + _supports_flash_attn_2 = True + _supports_sdpa = True + + def _init_weights(self, module): + """Initialize the weights""" + if isinstance(module, nn.Linear): + # Slightly different from the TF version which uses truncated_normal for initialization + # cf https://github.com/pytorch/pytorch/pull/5617 + module.weight.data.normal_(mean=0.0, std=self.config.initializer_range) + elif isinstance(module, (nn.LayerNorm, nn.GroupNorm)): + module.bias.data.zero_() + module.weight.data.fill_(1.0) + elif isinstance(module, nn.Conv1d): + if is_deepspeed_zero3_enabled(): + import deepspeed + + if hasattr(module, "weight_v") and hasattr(module, "weight_g"): + with deepspeed.zero.GatheredParameters([module.weight_v, module.weight_g], modifier_rank=0): + nn.init.kaiming_normal_(module.weight.data) + else: + with deepspeed.zero.GatheredParameters(module.weight, modifier_rank=0): + nn.init.kaiming_normal_(module.weight.data) + else: + nn.init.kaiming_normal_(module.weight.data) + + if isinstance(module, (nn.Linear, nn.Conv1d)) and module.bias is not None: + module.bias.data.zero_() + + def _get_feat_extract_output_lengths(self, input_lengths: Union[torch.LongTensor, int]): + """ + Computes the output length of the convolutional layers + """ + + def _conv_out_length(input_length, kernel_size, stride): + # 1D convolutional layer output length formula taken + # from https://pytorch.org/docs/stable/generated/torch.nn.Conv1d.html + return torch.div(input_length - kernel_size, stride, rounding_mode="floor") + 1 + + for kernel_size, stride in zip(self.config.conv_kernel, self.config.conv_stride): + input_lengths = _conv_out_length(input_lengths, kernel_size, stride) + + return input_lengths + + def _get_feature_vector_attention_mask(self, feature_vector_length: int, attention_mask: torch.LongTensor): + output_lengths = self._get_feat_extract_output_lengths(attention_mask.sum(-1)).to(torch.long) + batch_size = attention_mask.shape[0] + + attention_mask = torch.zeros( + (batch_size, feature_vector_length), dtype=attention_mask.dtype, device=attention_mask.device + ) + # these two operations makes sure that all values before the output lengths idxs are attended to + attention_mask[(torch.arange(attention_mask.shape[0], device=attention_mask.device), output_lengths - 1)] = 1 + attention_mask = attention_mask.flip([-1]).cumsum(-1).flip([-1]).bool() + return attention_mask + + +HUBERT_START_DOCSTRING = r""" + Hubert was proposed in [HuBERT: Self-Supervised Speech Representation Learning by Masked Prediction of Hidden + Units](https://arxiv.org/abs/2106.07447) by Wei-Ning Hsu, Benjamin Bolte, Yao-Hung Hubert Tsai, Kushal Lakhotia, + Ruslan Salakhutdinov, Abdelrahman Mohamed. + + This model inherits from [`PreTrainedModel`]. Check the superclass documentation for the generic methods the + library implements for all its model (such as downloading or saving etc.). + + This model is a PyTorch [torch.nn.Module](https://pytorch.org/docs/stable/nn.html#torch.nn.Module) sub-class. Use + it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage and + behavior. + + Parameters: + config ([`HubertConfig`]): Model configuration class with all the parameters of the model. + Initializing with a config file does not load the weights associated with the model, only the + configuration. Check out the [`~PreTrainedModel.from_pretrained`] method to load the model weights. +""" + + +HUBERT_INPUTS_DOCSTRING = r""" + Args: + input_values (`torch.FloatTensor` of shape `(batch_size, sequence_length)`): + Float values of input raw speech waveform. Values can be obtained by loading a `.flac` or `.wav` audio file + into an array of type `List[float]` or a `numpy.ndarray`, *e.g.* via the soundfile library (`pip install + soundfile`). To prepare the array into `input_values`, the [`AutoProcessor`] should be used for padding and + conversion into a tensor of type `torch.FloatTensor`. See [`Wav2Vec2Processor.__call__`] for details. + attention_mask (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*): + Mask to avoid performing convolution and attention on padding token indices. Mask values selected in `[0, + 1]`: + + - 1 for tokens that are **not masked**, + - 0 for tokens that are **masked**. + + [What are attention masks?](../glossary#attention-mask) + + + + `attention_mask` should only be passed if the corresponding processor has `config.return_attention_mask == + True`. For all models whose processor has `config.return_attention_mask == False`, such as + [hubert-base](https://huggingface.co/facebook/hubert-base-ls960), `attention_mask` should **not** be passed + to avoid degraded performance when doing batched inference. For such models `input_values` should simply be + padded with 0 and passed without `attention_mask`. Be aware that these models also yield slightly different + results depending on whether `input_values` is padded or not. + + + + output_attentions (`bool`, *optional*): + Whether or not to return the attentions tensors of all attention layers. See `attentions` under returned + tensors for more detail. + output_hidden_states (`bool`, *optional*): + Whether or not to return the hidden states of all layers. See `hidden_states` under returned tensors for + more detail. + return_dict (`bool`, *optional*): + Whether or not to return a [`~utils.ModelOutput`] instead of a plain tuple. +""" + + +@add_start_docstrings( + "The bare Hubert Model transformer outputting raw hidden-states without any specific head on top.", + HUBERT_START_DOCSTRING, +) +class HubertModel(HubertPreTrainedModel): + def __init__(self, config: HubertConfig): + super().__init__(config) + self.config = config + self.feature_extractor = HubertFeatureEncoder(config) + self.feature_projection = HubertFeatureProjection(config) + + if config.mask_time_prob > 0.0 or config.mask_feature_prob > 0.0: + self.masked_spec_embed = nn.Parameter(torch.Tensor(config.hidden_size).uniform_()) + + if config.do_stable_layer_norm: + self.encoder = HubertEncoderStableLayerNorm(config) + else: + self.encoder = HubertEncoder(config) + + # Initialize weights and apply final processing + self.post_init() + + # Copied from transformers.models.wav2vec2.modeling_wav2vec2.Wav2Vec2Model._mask_hidden_states + def _mask_hidden_states( + self, + hidden_states: torch.FloatTensor, + mask_time_indices: Optional[torch.FloatTensor] = None, + attention_mask: Optional[torch.LongTensor] = None, + ): + """ + Masks extracted features along time axis and/or along feature axis according to + [SpecAugment](https://arxiv.org/abs/1904.08779). + """ + + # `config.apply_spec_augment` can set masking to False + if not getattr(self.config, "apply_spec_augment", True): + return hidden_states + + # generate indices & apply SpecAugment along time axis + batch_size, sequence_length, hidden_size = hidden_states.size() + + if mask_time_indices is not None: + # apply SpecAugment along time axis with given mask_time_indices + hidden_states[mask_time_indices] = self.masked_spec_embed.to(hidden_states.dtype) + elif self.config.mask_time_prob > 0 and self.training: + mask_time_indices = _compute_mask_indices( + (batch_size, sequence_length), + mask_prob=self.config.mask_time_prob, + mask_length=self.config.mask_time_length, + attention_mask=attention_mask, + min_masks=self.config.mask_time_min_masks, + ) + mask_time_indices = torch.tensor(mask_time_indices, device=hidden_states.device, dtype=torch.bool) + hidden_states[mask_time_indices] = self.masked_spec_embed.to(hidden_states.dtype) + + if self.config.mask_feature_prob > 0 and self.training: + # generate indices & apply SpecAugment along feature axis + mask_feature_indices = _compute_mask_indices( + (batch_size, hidden_size), + mask_prob=self.config.mask_feature_prob, + mask_length=self.config.mask_feature_length, + min_masks=self.config.mask_feature_min_masks, + ) + mask_feature_indices = torch.tensor(mask_feature_indices, device=hidden_states.device, dtype=torch.bool) + mask_feature_indices = mask_feature_indices[:, None].expand(-1, sequence_length, -1) + hidden_states[mask_feature_indices] = 0 + + return hidden_states + + @add_start_docstrings_to_model_forward(HUBERT_INPUTS_DOCSTRING) + @replace_return_docstrings(output_type=BaseModelOutput, config_class=_CONFIG_FOR_DOC) + def forward( + self, + input_values: Optional[torch.Tensor], + attention_mask: Optional[torch.Tensor] = None, + mask_time_indices: Optional[torch.FloatTensor] = None, + output_attentions: Optional[bool] = None, + output_hidden_states: Optional[bool] = None, + return_dict: Optional[bool] = None, + ) -> Union[Tuple, BaseModelOutput]: + """ + + Returns: + + Example: + + ```python + >>> from transformers import AutoProcessor, HubertModel + >>> from datasets import load_dataset + >>> import soundfile as sf + + >>> processor = AutoProcessor.from_pretrained("facebook/hubert-large-ls960-ft") + >>> model = HubertModel.from_pretrained("facebook/hubert-large-ls960-ft") + + + >>> def map_to_array(batch): + ... speech, _ = sf.read(batch["file"]) + ... batch["speech"] = speech + ... return batch + + + >>> ds = load_dataset("hf-internal-testing/librispeech_asr_dummy", "clean", split="validation") + >>> ds = ds.map(map_to_array) + + >>> input_values = processor(ds["speech"][0], return_tensors="pt").input_values # Batch size 1 + >>> hidden_states = model(input_values).last_hidden_state + ```""" + output_attentions = output_attentions if output_attentions is not None else self.config.output_attentions + output_hidden_states = ( + output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states + ) + return_dict = return_dict if return_dict is not None else self.config.use_return_dict + + extract_features = self.feature_extractor(input_values) + extract_features = extract_features.transpose(1, 2) + + if attention_mask is not None: + # compute reduced attention_mask corresponding to feature vectors + attention_mask = self._get_feature_vector_attention_mask(extract_features.shape[1], attention_mask) + + hidden_states = self.feature_projection(extract_features) + hidden_states = self._mask_hidden_states(hidden_states, mask_time_indices=mask_time_indices) + + encoder_outputs = self.encoder( + hidden_states, + attention_mask=attention_mask, + output_attentions=output_attentions, + output_hidden_states=output_hidden_states, + return_dict=return_dict, + ) + + hidden_states = encoder_outputs[0] + + if not return_dict: + return (hidden_states,) + encoder_outputs[1:] + + return BaseModelOutput( + last_hidden_state=hidden_states, + hidden_states=encoder_outputs.hidden_states, + attentions=encoder_outputs.attentions, + ) + + +@add_start_docstrings( + """Hubert Model with a `language modeling` head on top for Connectionist Temporal Classification (CTC).""", + HUBERT_START_DOCSTRING, +) +# Copied from transformers.models.wav2vec2.modeling_wav2vec2.Wav2Vec2ForCTC with Wav2Vec2->Hubert, wav2vec2->hubert, WAV_2_VEC_2->HUBERT +class HubertForCTC(HubertPreTrainedModel): + def __init__(self, config, target_lang: Optional[str] = None): + super().__init__(config) + + self.hubert = HubertModel(config) + self.dropout = nn.Dropout(config.final_dropout) + + self.target_lang = target_lang + + if config.vocab_size is None: + raise ValueError( + f"You are trying to instantiate {self.__class__} with a configuration that " + "does not define the vocabulary size of the language model head. Please " + "instantiate the model as follows: `HubertForCTC.from_pretrained(..., vocab_size=vocab_size)`. " + "or define `vocab_size` of your model's configuration." + ) + output_hidden_size = ( + config.output_hidden_size if hasattr(config, "add_adapter") and config.add_adapter else config.hidden_size + ) + self.lm_head = nn.Linear(output_hidden_size, config.vocab_size) + + # Initialize weights and apply final processing + self.post_init() + + def tie_weights(self): + """ + This method overwrites [`~PreTrainedModel.tie_weights`] so that adapter weights can be correctly loaded when + passing `target_lang=...` to `from_pretrained(...)`. + + This method is **not** supposed to be called by the user and is prone to be changed in the future. + """ + + # Note that `tie_weights` is usually used to tie input and output embedding weights. The method is re-purposed to + # correctly load adapter layers for Hubert so that we do not have to introduce a new API to + # [`PreTrainedModel`]. While slightly hacky, Hubert never has to tie input and output embeddings, so that it is + # ok to repurpose this function here. + target_lang = self.target_lang + + if target_lang is not None and getattr(self.config, "adapter_attn_dim", None) is None: + raise ValueError(f"Cannot pass `target_lang`: {target_lang} if `config.adapter_attn_dim` is not defined.") + elif target_lang is None and getattr(self.config, "adapter_attn_dim", None) is not None: + logger.info("By default `target_lang` is set to 'eng'.") + elif target_lang is not None: + self.load_adapter(target_lang, force_load=True) + + def freeze_feature_extractor(self): + """ + Calling this function will disable the gradient computation for the feature encoder so that its parameter will + not be updated during training. + """ + warnings.warn( + "The method `freeze_feature_extractor` is deprecated and will be removed in Transformers v5. " + "Please use the equivalent `freeze_feature_encoder` method instead.", + FutureWarning, + ) + self.freeze_feature_encoder() + + def freeze_feature_encoder(self): + """ + Calling this function will disable the gradient computation for the feature encoder so that its parameter will + not be updated during training. + """ + self.hubert.feature_extractor._freeze_parameters() + + def freeze_base_model(self): + """ + Calling this function will disable the gradient computation for the base model so that its parameters will not + be updated during training. Only the classification head will be updated. + """ + for param in self.hubert.parameters(): + param.requires_grad = False + + @add_start_docstrings_to_model_forward(HUBERT_INPUTS_DOCSTRING) + @add_code_sample_docstrings( + checkpoint=_CHECKPOINT_FOR_DOC, + output_type=CausalLMOutput, + config_class=_CONFIG_FOR_DOC, + expected_output=_CTC_EXPECTED_OUTPUT, + expected_loss=_CTC_EXPECTED_LOSS, + ) + def forward( + self, + input_values: Optional[torch.Tensor], + attention_mask: Optional[torch.Tensor] = None, + output_attentions: Optional[bool] = None, + output_hidden_states: Optional[bool] = None, + return_dict: Optional[bool] = None, + labels: Optional[torch.Tensor] = None, + ) -> Union[Tuple, CausalLMOutput]: + r""" + labels (`torch.LongTensor` of shape `(batch_size, target_length)`, *optional*): + Labels for connectionist temporal classification. Note that `target_length` has to be smaller or equal to + the sequence length of the output logits. Indices are selected in `[-100, 0, ..., config.vocab_size - 1]`. + All labels set to `-100` are ignored (masked), the loss is only computed for labels in `[0, ..., + config.vocab_size - 1]`. + """ + + return_dict = return_dict if return_dict is not None else self.config.use_return_dict + + outputs = self.hubert( + input_values, + attention_mask=attention_mask, + output_attentions=output_attentions, + output_hidden_states=output_hidden_states, + return_dict=return_dict, + ) + + hidden_states = outputs[0] + hidden_states = self.dropout(hidden_states) + + logits = self.lm_head(hidden_states) + + loss = None + if labels is not None: + if labels.max() >= self.config.vocab_size: + raise ValueError(f"Label values must be <= vocab_size: {self.config.vocab_size}") + + # retrieve loss input_lengths from attention_mask + attention_mask = ( + attention_mask if attention_mask is not None else torch.ones_like(input_values, dtype=torch.long) + ) + input_lengths = self._get_feat_extract_output_lengths(attention_mask.sum(-1)).to(torch.long) + + # assuming that padded tokens are filled with -100 + # when not being attended to + labels_mask = labels >= 0 + target_lengths = labels_mask.sum(-1) + flattened_targets = labels.masked_select(labels_mask) + + # ctc_loss doesn't support fp16 + log_probs = nn.functional.log_softmax(logits, dim=-1, dtype=torch.float32).transpose(0, 1) + + with torch.backends.cudnn.flags(enabled=False): + loss = nn.functional.ctc_loss( + log_probs, + flattened_targets, + input_lengths, + target_lengths, + blank=self.config.pad_token_id, + reduction=self.config.ctc_loss_reduction, + zero_infinity=self.config.ctc_zero_infinity, + ) + + if not return_dict: + output = (logits,) + outputs[_HIDDEN_STATES_START_POSITION:] + return ((loss,) + output) if loss is not None else output + + return CausalLMOutput( + loss=loss, logits=logits, hidden_states=outputs.hidden_states, attentions=outputs.attentions + ) + + +@add_start_docstrings( + """ + Hubert Model with a sequence classification head on top (a linear layer over the pooled output) for tasks like + SUPERB Keyword Spotting. + """, + HUBERT_START_DOCSTRING, +) +# Copied from transformers.models.wav2vec2.modeling_wav2vec2.Wav2Vec2ForSequenceClassification with Wav2Vec2->Hubert, wav2vec2->hubert, WAV_2_VEC_2->HUBERT +class HubertForSequenceClassification(HubertPreTrainedModel): + def __init__(self, config): + super().__init__(config) + + if hasattr(config, "add_adapter") and config.add_adapter: + raise ValueError( + "Sequence classification does not support the use of Hubert adapters (config.add_adapter=True)" + ) + self.hubert = HubertModel(config) + num_layers = config.num_hidden_layers + 1 # transformer layers + input embeddings + if config.use_weighted_layer_sum: + self.layer_weights = nn.Parameter(torch.ones(num_layers) / num_layers) + self.projector = nn.Linear(config.hidden_size, config.classifier_proj_size) + self.classifier = nn.Linear(config.classifier_proj_size, config.num_labels) + + # Initialize weights and apply final processing + self.post_init() + + def freeze_feature_extractor(self): + """ + Calling this function will disable the gradient computation for the feature encoder so that its parameters will + not be updated during training. + """ + warnings.warn( + "The method `freeze_feature_extractor` is deprecated and will be removed in Transformers v5. " + "Please use the equivalent `freeze_feature_encoder` method instead.", + FutureWarning, + ) + self.freeze_feature_encoder() + + def freeze_feature_encoder(self): + """ + Calling this function will disable the gradient computation for the feature encoder so that its parameter will + not be updated during training. + """ + self.hubert.feature_extractor._freeze_parameters() + + def freeze_base_model(self): + """ + Calling this function will disable the gradient computation for the base model so that its parameters will not + be updated during training. Only the classification head will be updated. + """ + for param in self.hubert.parameters(): + param.requires_grad = False + + @add_start_docstrings_to_model_forward(HUBERT_INPUTS_DOCSTRING) + @add_code_sample_docstrings( + checkpoint=_SEQ_CLASS_CHECKPOINT, + output_type=SequenceClassifierOutput, + config_class=_CONFIG_FOR_DOC, + modality="audio", + expected_output=_SEQ_CLASS_EXPECTED_OUTPUT, + expected_loss=_SEQ_CLASS_EXPECTED_LOSS, + ) + def forward( + self, + input_values: Optional[torch.Tensor], + attention_mask: Optional[torch.Tensor] = None, + output_attentions: Optional[bool] = None, + output_hidden_states: Optional[bool] = None, + return_dict: Optional[bool] = None, + labels: Optional[torch.Tensor] = None, + ) -> Union[Tuple, SequenceClassifierOutput]: + r""" + labels (`torch.LongTensor` of shape `(batch_size,)`, *optional*): + Labels for computing the sequence classification/regression loss. Indices should be in `[0, ..., + config.num_labels - 1]`. If `config.num_labels == 1` a regression loss is computed (Mean-Square loss), If + `config.num_labels > 1` a classification loss is computed (Cross-Entropy). + """ + + return_dict = return_dict if return_dict is not None else self.config.use_return_dict + output_hidden_states = True if self.config.use_weighted_layer_sum else output_hidden_states + + outputs = self.hubert( + input_values, + attention_mask=attention_mask, + output_attentions=output_attentions, + output_hidden_states=output_hidden_states, + return_dict=return_dict, + ) + + if self.config.use_weighted_layer_sum: + hidden_states = outputs[_HIDDEN_STATES_START_POSITION] + hidden_states = torch.stack(hidden_states, dim=1) + norm_weights = nn.functional.softmax(self.layer_weights, dim=-1) + hidden_states = (hidden_states * norm_weights.view(-1, 1, 1)).sum(dim=1) + else: + hidden_states = outputs[0] + + hidden_states = self.projector(hidden_states) + if attention_mask is None: + pooled_output = hidden_states.mean(dim=1) + else: + padding_mask = self._get_feature_vector_attention_mask(hidden_states.shape[1], attention_mask) + hidden_states[~padding_mask] = 0.0 + pooled_output = hidden_states.sum(dim=1) / padding_mask.sum(dim=1).view(-1, 1) + + logits = self.classifier(pooled_output) + + loss = None + if labels is not None: + loss_fct = CrossEntropyLoss() + loss = loss_fct(logits.view(-1, self.config.num_labels), labels.view(-1)) + + if not return_dict: + output = (logits,) + outputs[_HIDDEN_STATES_START_POSITION:] + return ((loss,) + output) if loss is not None else output + + return SequenceClassifierOutput( + loss=loss, + logits=logits, + hidden_states=outputs.hidden_states, + attentions=outputs.attentions, + ) diff --git a/parrot/lib/python3.10/site-packages/transformers/models/hubert/modeling_tf_hubert.py b/parrot/lib/python3.10/site-packages/transformers/models/hubert/modeling_tf_hubert.py new file mode 100644 index 0000000000000000000000000000000000000000..142ab0961dd5cb44e298447df71cdb0898955c36 --- /dev/null +++ b/parrot/lib/python3.10/site-packages/transformers/models/hubert/modeling_tf_hubert.py @@ -0,0 +1,1673 @@ +# coding=utf-8 +# Copyright 2021 The Fairseq Authors and the HuggingFace Inc. team. All rights reserved. +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. +""" TensorFlow Hubert model.""" + +from __future__ import annotations + +import warnings +from typing import Any, Optional, Tuple, Union + +import numpy as np +import tensorflow as tf + +from ...activations_tf import get_tf_activation +from ...modeling_tf_outputs import TFBaseModelOutput, TFCausalLMOutput +from ...modeling_tf_utils import ( + TFPreTrainedModel, + get_initializer, + keras, + keras_serializable, + unpack_inputs, +) +from ...tf_utils import shape_list, stable_softmax +from ...utils import ( + add_start_docstrings, + add_start_docstrings_to_model_forward, + logging, + replace_return_docstrings, +) +from .configuration_hubert import HubertConfig + + +logger = logging.get_logger(__name__) + +_CONFIG_FOR_DOC = "HubertConfig" + + +LARGE_NEGATIVE = -1e8 + + +# Copied from transformers.models.wav2vec2.modeling_tf_wav2vec2._sample_without_replacement +def _sample_without_replacement(distribution, num_samples): + """ + Categorical sampling without replacement is currently not implemented. The gumbel-max trick will do for now - see + https://github.com/tensorflow/tensorflow/issues/9260 for more info + """ + z = -tf.math.log(tf.random.uniform(shape_list(distribution), 0, 1)) + _, indices = tf.nn.top_k(distribution + z, num_samples) + return indices + + +# Copied from transformers.models.wav2vec2.modeling_tf_wav2vec2._scatter_values_on_batch_indices +def _scatter_values_on_batch_indices(values, batch_indices, output_shape): + """ + Scatter function as in PyTorch with indices in format (batch_dim, indixes) + """ + indices_shape = shape_list(batch_indices) + # broadcast batch dim to indices_shape + broad_casted_batch_dims = tf.reshape( + tf.broadcast_to(tf.expand_dims(tf.range(indices_shape[0]), axis=-1), indices_shape), [1, -1] + ) + # transform batch_indices to pair_indices + pair_indices = tf.transpose(tf.concat([broad_casted_batch_dims, tf.reshape(batch_indices, [1, -1])], 0)) + # scatter values to pair indices + return tf.scatter_nd(pair_indices, tf.reshape(values, [-1]), output_shape) + + +# Copied from transformers.models.wav2vec2.modeling_tf_wav2vec2._compute_mask_indices +def _compute_mask_indices( + shape: Tuple[int, int], + mask_prob: float, + mask_length: int, + min_masks: int = 0, +) -> tf.Tensor: + """ + Computes random mask spans for a given shape + + Args: + shape: the shape for which to compute masks. + should be of size 2 where first element is batch size and 2nd is timesteps + attention_mask: optional padding mask of the same size as shape, which will prevent masking padded elements + mask_prob: + probability for each token to be chosen as start of the span to be masked. this will be multiplied by + number of timesteps divided by length of mask span to mask approximately this percentage of all elements. + however due to overlaps, the actual number will be smaller (unless no_overlap is True) + mask_length: size of the mask + min_masks: minimum number of masked spans + + Adapted from [fairseq's + data_utils.py](https://github.com/pytorch/fairseq/blob/e0788f7007a8473a76db573985031f3c94201e79/fairseq/data/data_utils.py#L376). + """ + batch_size, sequence_length = shape + + if mask_length < 1: + raise ValueError("`mask_length` has to be bigger than 0.") + + tf.debugging.assert_less( + mask_length, + sequence_length, + message=( + f"`mask_length` has to be smaller than `sequence_length`, but got `mask_length`: {mask_length} and" + f" `sequence_length`: {sequence_length}`" + ), + ) + + # compute number of masked spans in batch + num_masked_spans = mask_prob * tf.cast(sequence_length, tf.float32) / mask_length + tf.random.uniform((1,)) + num_masked_spans = tf.maximum(num_masked_spans, min_masks) + num_masked_spans = tf.cast(num_masked_spans, tf.int32) + + # make sure num masked indices <= sequence_length + num_masked_spans = tf.math.minimum(sequence_length // mask_length, num_masked_spans) + num_masked_spans = tf.squeeze(num_masked_spans) + + # SpecAugment mask to fill + spec_aug_mask = tf.zeros((batch_size, sequence_length), dtype=tf.int32) + + # uniform distribution to sample from, make sure that offset samples are < sequence_length + uniform_dist = tf.ones((batch_size, sequence_length - (mask_length - 1))) + + # get random indices to mask + spec_aug_mask_idxs = _sample_without_replacement(uniform_dist, num_masked_spans) + + # expand masked indices to masked spans + spec_aug_mask_idxs = tf.expand_dims(spec_aug_mask_idxs, -1) + spec_aug_mask_idxs = tf.tile(spec_aug_mask_idxs, (1, 1, mask_length)) + spec_aug_mask_idxs = tf.reshape(spec_aug_mask_idxs, (batch_size, num_masked_spans * mask_length)) + + offsets = tf.range(mask_length)[tf.newaxis, tf.newaxis, :] + offsets = tf.tile(offsets, (batch_size, num_masked_spans, 1)) + offsets = tf.reshape(offsets, (batch_size, num_masked_spans * mask_length)) + + spec_aug_mask_idxs = spec_aug_mask_idxs + offsets + + # scatter indices to mask + spec_aug_mask = _scatter_values_on_batch_indices( + tf.ones_like(spec_aug_mask_idxs), spec_aug_mask_idxs, tf.shape(spec_aug_mask) + ) + + return spec_aug_mask + + +# Copied from transformers.models.bart.modeling_tf_bart._expand_mask +def _expand_mask(mask: tf.Tensor, tgt_len: Optional[int] = None): + """ + Expands attention_mask from `[bsz, seq_len]` to `[bsz, 1, tgt_seq_len, src_seq_len]`. + """ + src_len = shape_list(mask)[1] + tgt_len = tgt_len if tgt_len is not None else src_len + one_cst = tf.constant(1.0) + mask = tf.cast(mask, dtype=one_cst.dtype) + expanded_mask = tf.tile(mask[:, None, None, :], (1, 1, tgt_len, 1)) + + return (one_cst - expanded_mask) * LARGE_NEGATIVE + + +# Copied from transformers.models.wav2vec2.modeling_tf_wav2vec2.TFWav2Vec2GroupNorm with Wav2Vec2->Hubert +class TFHubertGroupNorm(keras.layers.Layer): + """ + From tensorflow-addons https://www.tensorflow.org/addons/api_docs/python/tfa/layers/GroupNormalization + """ + + def __init__( + self, + groups: int = 32, + axis: int = -1, + epsilon: float = 1e-3, + center: bool = True, + scale: bool = True, + beta_initializer: keras.initializers.Initializer = "zeros", + gamma_initializer: keras.initializers.Initializer = "ones", + beta_regularizer: keras.regularizers.Regularizer = None, + gamma_regularizer: keras.regularizers.Regularizer = None, + beta_constraint: keras.constraints.Constraint = None, + gamma_constraint: keras.constraints.Constraint = None, + **kwargs, + ): + super().__init__(**kwargs) + self.supports_masking = True + self.groups = groups + self.axis = axis + self.epsilon = epsilon + self.center = center + self.scale = scale + self.beta_initializer = keras.initializers.get(beta_initializer) + self.gamma_initializer = keras.initializers.get(gamma_initializer) + self.beta_regularizer = keras.regularizers.get(beta_regularizer) + self.gamma_regularizer = keras.regularizers.get(gamma_regularizer) + self.beta_constraint = keras.constraints.get(beta_constraint) + self.gamma_constraint = keras.constraints.get(gamma_constraint) + self._check_axis() + + def build(self, input_shape): + self._check_if_input_shape_is_none(input_shape) + self._set_number_of_groups_for_instance_norm(input_shape) + self._check_size_of_dimensions(input_shape) + self._create_input_spec(input_shape) + + self._add_gamma_weight(input_shape) + self._add_beta_weight(input_shape) + self.built = True + super().build(input_shape) + + def call(self, inputs): + input_shape = keras.backend.int_shape(inputs) + tensor_input_shape = tf.shape(inputs) + + reshaped_inputs, group_shape = self._reshape_into_groups(inputs, input_shape, tensor_input_shape) + + normalized_inputs = self._apply_normalization(reshaped_inputs, input_shape) + + is_instance_norm = (input_shape[self.axis] // self.groups) == 1 + if not is_instance_norm: + outputs = tf.reshape(normalized_inputs, tensor_input_shape) + else: + outputs = normalized_inputs + + return outputs + + def get_config(self): + config = { + "groups": self.groups, + "axis": self.axis, + "epsilon": self.epsilon, + "center": self.center, + "scale": self.scale, + "beta_initializer": keras.initializers.serialize(self.beta_initializer), + "gamma_initializer": keras.initializers.serialize(self.gamma_initializer), + "beta_regularizer": keras.regularizers.serialize(self.beta_regularizer), + "gamma_regularizer": keras.regularizers.serialize(self.gamma_regularizer), + "beta_constraint": keras.constraints.serialize(self.beta_constraint), + "gamma_constraint": keras.constraints.serialize(self.gamma_constraint), + } + base_config = super().get_config() + return {**base_config, **config} + + def compute_output_shape(self, input_shape): + return input_shape + + def _reshape_into_groups(self, inputs, input_shape, tensor_input_shape): + group_shape = [tensor_input_shape[i] for i in range(len(input_shape))] + is_instance_norm = (input_shape[self.axis] // self.groups) == 1 + if not is_instance_norm: + group_shape[self.axis] = input_shape[self.axis] // self.groups + group_shape.insert(self.axis, self.groups) + group_shape = tf.stack(group_shape) + reshaped_inputs = tf.reshape(inputs, group_shape) + return reshaped_inputs, group_shape + else: + return inputs, group_shape + + def _apply_normalization(self, reshaped_inputs, input_shape): + group_shape = keras.backend.int_shape(reshaped_inputs) + group_reduction_axes = list(range(1, len(group_shape))) + is_instance_norm = (input_shape[self.axis] // self.groups) == 1 + if not is_instance_norm: + axis = -2 if self.axis == -1 else self.axis - 1 + else: + axis = -1 if self.axis == -1 else self.axis - 1 + group_reduction_axes.pop(axis) + + mean, variance = tf.nn.moments(reshaped_inputs, group_reduction_axes, keepdims=True) + + gamma, beta = self._get_reshaped_weights(input_shape) + normalized_inputs = tf.nn.batch_normalization( + reshaped_inputs, + mean=mean, + variance=variance, + scale=gamma, + offset=beta, + variance_epsilon=self.epsilon, + ) + return normalized_inputs + + def _get_reshaped_weights(self, input_shape): + broadcast_shape = self._create_broadcast_shape(input_shape) + gamma = None + beta = None + if self.scale: + gamma = tf.reshape(self.gamma, broadcast_shape) + + if self.center: + beta = tf.reshape(self.beta, broadcast_shape) + return gamma, beta + + def _check_if_input_shape_is_none(self, input_shape): + dim = input_shape[self.axis] + if dim is None: + raise ValueError( + "Axis " + + str(self.axis) + + " of input tensor should have a defined dimension but the layer received an input with shape " + + str(input_shape) + + "." + ) + + def _set_number_of_groups_for_instance_norm(self, input_shape): + dim = input_shape[self.axis] + + if self.groups == -1: + self.groups = dim + + def _check_size_of_dimensions(self, input_shape): + dim = input_shape[self.axis] + if dim < self.groups: + raise ValueError( + "Number of groups (" + + str(self.groups) + + ") cannot be more than the number of channels (" + + str(dim) + + ")." + ) + + if dim % self.groups != 0: + raise ValueError( + "Number of groups (" + + str(self.groups) + + ") must be a multiple of the number of channels (" + + str(dim) + + ")." + ) + + def _check_axis(self): + if self.axis == 0: + raise ValueError( + "You are trying to normalize your batch axis. Do you want to use tf.layer.batch_normalization instead" + ) + + def _create_input_spec(self, input_shape): + dim = input_shape[self.axis] + self.input_spec = keras.layers.InputSpec(ndim=len(input_shape), axes={self.axis: dim}) + + def _add_gamma_weight(self, input_shape): + dim = input_shape[self.axis] + shape = (dim,) + + if self.scale: + self.gamma = self.add_weight( + shape=shape, + name="gamma", + initializer=self.gamma_initializer, + regularizer=self.gamma_regularizer, + constraint=self.gamma_constraint, + ) + else: + self.gamma = None + + def _add_beta_weight(self, input_shape): + dim = input_shape[self.axis] + shape = (dim,) + + if self.center: + self.beta = self.add_weight( + shape=shape, + name="beta", + initializer=self.beta_initializer, + regularizer=self.beta_regularizer, + constraint=self.beta_constraint, + ) + else: + self.beta = None + + def _create_broadcast_shape(self, input_shape): + broadcast_shape = [1] * len(input_shape) + is_instance_norm = (input_shape[self.axis] // self.groups) == 1 + if not is_instance_norm: + broadcast_shape[self.axis] = input_shape[self.axis] // self.groups + broadcast_shape.insert(self.axis, self.groups) + else: + broadcast_shape[self.axis] = self.groups + return broadcast_shape + + +# Copied from transformers.models.wav2vec2.modeling_tf_wav2vec2.TFWav2Vec2WeightNormConv1D with Wav2Vec2->Hubert +class TFHubertWeightNormConv1D(keras.layers.Conv1D): + """Adapted from https://www.tensorflow.org/probability/api_docs/python/tfp/layers/weight_norm/WeightNorm""" + + def __init__(self, filters, kernel_size, groups, explicit_padding, **kwargs): + super().__init__( + filters=filters, + kernel_size=kernel_size, + groups=groups, + padding="valid", + use_bias=True, + bias_initializer="he_normal", + **kwargs, + ) + self.explicit_padding = explicit_padding + self.filter_axis = 2 + self.kernel_norm_axes = tf.constant([0, 1]) + + def _init_norm(self): + """Set the norm of the weight vector.""" + kernel_norm = tf.sqrt(tf.reduce_sum(tf.square(self.weight_v), axis=self.kernel_norm_axes)) + self.weight_g.assign(kernel_norm[:, tf.newaxis, tf.newaxis]) + + def _normalize_kernel(self): + """Generate normalized weights.""" + kernel = tf.nn.l2_normalize(self.weight_v, axis=self.kernel_norm_axes) * tf.transpose(self.weight_g) + self.kernel = tf.transpose(kernel) + + def build(self, input_shape): + if not self.built: + super().build(input_shape) + + self.kernel = tf.Variable(tf.transpose(self.kernel), name="weight_v", trainable=True) + self.weight_v = self.kernel + + self.weight_g = self.add_weight( + name="weight_g", + shape=(int(self.weight_v.shape[self.filter_axis]), 1, 1), + initializer="ones", + dtype=self.weight_v.dtype, + trainable=True, + ) + self._init_norm() + self.bias = self.add_weight(name="bias", shape=(self.filters,), initializer="zeros", trainable=True) + + def call(self, inputs): + # TODO Matt: Assigning to attributes in call() is deeply sinful in TensorFlow, as it should be idempotent. + # This whole layer should be replaced by a layer that doesn't inherit from Conv1D, but instead calls + # a functional 1d convolution with normalized weights that it generates (but does not store!) + self._normalize_kernel() + + padded_inputs = tf.pad(inputs, ((0, 0), (self.explicit_padding, self.explicit_padding), (0, 0))) + output = super().call(padded_inputs) + + return output + + +# Copied from transformers.models.wav2vec2.modeling_tf_wav2vec2.TFWav2Vec2NoLayerNormConvLayer with Wav2Vec2->Hubert +class TFHubertNoLayerNormConvLayer(keras.layers.Layer): + def __init__(self, config: HubertConfig, layer_id: int = 0, **kwargs: Any) -> None: + super().__init__(**kwargs) + self.in_conv_dim = config.conv_dim[layer_id] if layer_id > 0 else 1 + self.out_conv_dim = config.conv_dim[layer_id] + + self.conv = keras.layers.Conv1D( + filters=self.out_conv_dim, + kernel_size=config.conv_kernel[layer_id], + strides=config.conv_stride[layer_id], + use_bias=config.conv_bias, + name="conv", + ) + self.activation = get_tf_activation(config.feat_extract_activation) + + def call(self, hidden_states: tf.Tensor) -> tf.Tensor: + hidden_states = self.conv(hidden_states) + hidden_states = self.activation(hidden_states) + return hidden_states + + def build(self, input_shape=None): + if self.built: + return + self.built = True + if getattr(self, "conv", None) is not None: + with tf.name_scope(self.conv.name): + self.conv.build([None, None, self.in_conv_dim]) + + +# Copied from transformers.models.wav2vec2.modeling_tf_wav2vec2.TFWav2Vec2LayerNormConvLayer with Wav2Vec2->Hubert +class TFHubertLayerNormConvLayer(keras.layers.Layer): + def __init__(self, config: HubertConfig, layer_id: int = 0, **kwargs: Any) -> None: + super().__init__(**kwargs) + self.in_conv_dim = config.conv_dim[layer_id] if layer_id > 0 else 1 + self.out_conv_dim = config.conv_dim[layer_id] + + self.conv = keras.layers.Conv1D( + filters=self.out_conv_dim, + kernel_size=config.conv_kernel[layer_id], + strides=config.conv_stride[layer_id], + use_bias=config.conv_bias, + name="conv", + ) + self.layer_norm = keras.layers.LayerNormalization(name="layer_norm", epsilon=config.layer_norm_eps) + self.activation = get_tf_activation(config.feat_extract_activation) + + def call(self, hidden_states: tf.Tensor) -> tf.Tensor: + hidden_states = self.conv(hidden_states) + hidden_states = self.layer_norm(hidden_states) + hidden_states = self.activation(hidden_states) + return hidden_states + + def build(self, input_shape=None): + if self.built: + return + self.built = True + if getattr(self, "conv", None) is not None: + with tf.name_scope(self.conv.name): + self.conv.build([None, None, self.in_conv_dim]) + if getattr(self, "layer_norm", None) is not None: + with tf.name_scope(self.layer_norm.name): + self.layer_norm.build([None, None, self.out_conv_dim]) + + +# Copied from transformers.models.wav2vec2.modeling_tf_wav2vec2.TFWav2Vec2GroupNormConvLayer with Wav2Vec2->Hubert +class TFHubertGroupNormConvLayer(keras.layers.Layer): + def __init__(self, config: HubertConfig, layer_id: int = 0, **kwargs: Any) -> None: + super().__init__(**kwargs) + self.in_conv_dim = config.conv_dim[layer_id] if layer_id > 0 else 1 + self.out_conv_dim = config.conv_dim[layer_id] + + self.conv = keras.layers.Conv1D( + filters=self.out_conv_dim, + kernel_size=config.conv_kernel[layer_id], + strides=config.conv_stride[layer_id], + use_bias=config.conv_bias, + name="conv", + ) + self.activation = get_tf_activation(config.feat_extract_activation) + self.layer_norm = TFHubertGroupNorm(groups=self.out_conv_dim, epsilon=config.layer_norm_eps, name="layer_norm") + + def call(self, hidden_states: tf.Tensor) -> tf.Tensor: + hidden_states = self.conv(hidden_states) + hidden_states = self.layer_norm(hidden_states) + hidden_states = self.activation(hidden_states) + return hidden_states + + def build(self, input_shape=None): + if self.built: + return + self.built = True + if getattr(self, "conv", None) is not None: + with tf.name_scope(self.conv.name): + self.conv.build([None, None, self.in_conv_dim]) + if getattr(self, "layer_norm", None) is not None: + with tf.name_scope(self.layer_norm.name): + self.layer_norm.build([None, None, self.out_conv_dim]) + + +# Copied from transformers.models.wav2vec2.modeling_tf_wav2vec2.TFWav2Vec2PositionalConvEmbedding with Wav2Vec2->Hubert +class TFHubertPositionalConvEmbedding(keras.layers.Layer): + def __init__(self, config: HubertConfig, **kwargs: Any) -> None: + super().__init__(**kwargs) + self.conv = TFHubertWeightNormConv1D( + filters=config.hidden_size, + kernel_size=config.num_conv_pos_embeddings, + groups=config.num_conv_pos_embedding_groups, + explicit_padding=config.num_conv_pos_embeddings // 2, + name="conv", + ) + self.padding = TFHubertSamePadLayer(config.num_conv_pos_embeddings) + self.activation = get_tf_activation(config.feat_extract_activation) + self.config = config + + def call(self, hidden_states: tf.Tensor) -> tf.Tensor: + hidden_states = self.conv(hidden_states) + hidden_states = self.padding(hidden_states) + hidden_states = self.activation(hidden_states) + return hidden_states + + def build(self, input_shape=None): + if self.built: + return + self.built = True + if getattr(self, "conv", None) is not None: + with tf.name_scope(self.conv.name): + self.conv.build([None, None, self.config.hidden_size]) + + +# Copied from transformers.models.wav2vec2.modeling_tf_wav2vec2.TFWav2Vec2SamePadLayer with Wav2Vec2->Hubert +class TFHubertSamePadLayer(keras.layers.Layer): + def __init__(self, num_conv_pos_embeddings, **kwargs): + super().__init__(**kwargs) + self.num_pad_remove = 1 if num_conv_pos_embeddings % 2 == 0 else 0 + + def call(self, hidden_states): + if self.num_pad_remove > 0: + hidden_states = hidden_states[:, : -self.num_pad_remove, :] + return hidden_states + + +class TFHubertFeatureEncoder(keras.layers.Layer): + def __init__(self, config: HubertConfig, **kwargs: Any) -> None: + super().__init__(**kwargs) + + if config.feat_extract_norm == "group": + conv_layers = [TFHubertGroupNormConvLayer(config, layer_id=0, name=f"conv_layers.{0}")] + [ + TFHubertNoLayerNormConvLayer(config, layer_id=i + 1, name=f"conv_layers.{i+1}") + for i in range(config.num_feat_extract_layers - 1) + ] + elif config.feat_extract_norm == "layer": + conv_layers = [ + TFHubertLayerNormConvLayer(config, layer_id=i, name=f"conv_layers.{i}") + for i in range(config.num_feat_extract_layers) + ] + else: + raise ValueError( + f"`config.feat_extract_norm` is {config.feat_extract_norm}, but has to be one of ['group', 'layer']" + ) + self.conv_layers = conv_layers + + def call(self, input_values): + hidden_states = tf.expand_dims(input_values, -1) + for conv_layer in self.conv_layers: + hidden_states = conv_layer(hidden_states) + return hidden_states + + def build(self, input_shape=None): + if self.built: + return + self.built = True + for conv_layer in self.conv_layers: + with tf.name_scope(conv_layer.name): + conv_layer.build(None) + + +class TFHubertFeatureExtractor(TFHubertFeatureEncoder): + def __init__(self, config, **kwargs): + super().__init__(config, **kwargs) + warnings.warn( + f"The class `{self.__class__.__name__}` has been depreciated " + "and will be removed in Transformers v5. " + f"Use `{self.__class__.__bases__[0].__name__}` instead.", + FutureWarning, + ) + + +class TFHubertFeatureProjection(keras.layers.Layer): + def __init__(self, config: HubertConfig, **kwargs): + super().__init__(**kwargs) + + self.layer_norm = keras.layers.LayerNormalization(epsilon=config.layer_norm_eps, name="layer_norm") + self.projection = keras.layers.Dense( + units=config.hidden_size, + kernel_initializer=get_initializer(config.initializer_range), + bias_initializer="zeros", + name="projection", + ) + self.dropout = keras.layers.Dropout(rate=config.feat_proj_dropout) + self.config = config + + def call(self, hidden_states: tf.Tensor, training: bool = False) -> tf.Tensor: + hidden_states = self.layer_norm(hidden_states) + hidden_states = self.projection(hidden_states) + hidden_states = self.dropout(hidden_states, training=training) + return hidden_states + + def build(self, input_shape=None): + if self.built: + return + self.built = True + if getattr(self, "layer_norm", None) is not None: + with tf.name_scope(self.layer_norm.name): + self.layer_norm.build([None, None, self.config.conv_dim[-1]]) + if getattr(self, "projection", None) is not None: + with tf.name_scope(self.projection.name): + self.projection.build([None, None, self.config.conv_dim[-1]]) + + +# Copied from transformers.models.bart.modeling_tf_bart.TFBartAttention with TFBart->TFHubert +class TFHubertAttention(keras.layers.Layer): + """Multi-headed attention from "Attention Is All You Need""" + + def __init__( + self, + embed_dim: int, + num_heads: int, + dropout: float = 0.0, + is_decoder: bool = False, + bias: bool = True, + **kwargs, + ): + super().__init__(**kwargs) + self.embed_dim = embed_dim + + self.num_heads = num_heads + self.dropout = keras.layers.Dropout(dropout) + self.head_dim = embed_dim // num_heads + if (self.head_dim * num_heads) != self.embed_dim: + raise ValueError( + f"embed_dim must be divisible by num_heads (got `embed_dim`: {self.embed_dim}" + f" and `num_heads`: {num_heads})." + ) + self.scaling = self.head_dim**-0.5 + self.is_decoder = is_decoder + + self.k_proj = keras.layers.Dense(embed_dim, use_bias=bias, name="k_proj") + self.q_proj = keras.layers.Dense(embed_dim, use_bias=bias, name="q_proj") + self.v_proj = keras.layers.Dense(embed_dim, use_bias=bias, name="v_proj") + self.out_proj = keras.layers.Dense(embed_dim, use_bias=bias, name="out_proj") + + def _shape(self, tensor: tf.Tensor, seq_len: int, bsz: int): + return tf.transpose(tf.reshape(tensor, (bsz, seq_len, self.num_heads, self.head_dim)), (0, 2, 1, 3)) + + def call( + self, + hidden_states: tf.Tensor, + key_value_states: tf.Tensor | None = None, + past_key_value: Tuple[Tuple[tf.Tensor]] | None = None, + attention_mask: tf.Tensor | None = None, + layer_head_mask: tf.Tensor | None = None, + training: Optional[bool] = False, + ) -> Tuple[tf.Tensor, tf.Tensor | None]: + """Input shape: Batch x Time x Channel""" + + # if key_value_states are provided this layer is used as a cross-attention layer + # for the decoder + is_cross_attention = key_value_states is not None + bsz, tgt_len, embed_dim = shape_list(hidden_states) + + # get query proj + query_states = self.q_proj(hidden_states) * self.scaling + # get key, value proj + if is_cross_attention and past_key_value is not None: + # reuse k,v, cross_attentions + key_states = past_key_value[0] + value_states = past_key_value[1] + elif is_cross_attention: + # cross_attentions + key_states = self._shape(self.k_proj(key_value_states), -1, bsz) + value_states = self._shape(self.v_proj(key_value_states), -1, bsz) + elif past_key_value is not None: + # reuse k, v, self_attention + key_states = self._shape(self.k_proj(hidden_states), -1, bsz) + value_states = self._shape(self.v_proj(hidden_states), -1, bsz) + key_states = tf.concat([past_key_value[0], key_states], axis=2) + value_states = tf.concat([past_key_value[1], value_states], axis=2) + else: + # self_attention + key_states = self._shape(self.k_proj(hidden_states), -1, bsz) + value_states = self._shape(self.v_proj(hidden_states), -1, bsz) + + if self.is_decoder: + # if cross_attention save Tuple(tf.Tensor, tf.Tensor) of all cross attention key/value_states. + # Further calls to cross_attention layer can then reuse all cross-attention + # key/value_states (first "if" case) + # if uni-directional self-attention (decoder) save Tuple(tf.Tensor, tf.Tensor) of + # all previous decoder key/value_states. Further calls to uni-directional self-attention + # can concat previous decoder key/value_states to current projected key/value_states (third "elif" case) + # if encoder bi-directional self-attention `past_key_value` is always `None` + past_key_value = (key_states, value_states) + + proj_shape = (bsz * self.num_heads, -1, self.head_dim) + query_states = tf.reshape(self._shape(query_states, tgt_len, bsz), proj_shape) + key_states = tf.reshape(key_states, proj_shape) + value_states = tf.reshape(value_states, proj_shape) + + src_len = shape_list(key_states)[1] + attn_weights = tf.matmul(query_states, key_states, transpose_b=True) + + tf.debugging.assert_equal( + shape_list(attn_weights), + [bsz * self.num_heads, tgt_len, src_len], + message=( + f"Attention weights should be of size {(bsz * self.num_heads, tgt_len, src_len)}, but is" + f" {shape_list(attn_weights)}" + ), + ) + + if attention_mask is not None: + tf.debugging.assert_equal( + shape_list(attention_mask), + [bsz, 1, tgt_len, src_len], + message=( + f"Attention mask should be of size {(bsz, 1, tgt_len, src_len)}, but is" + f" {shape_list(attention_mask)}" + ), + ) + + attention_mask = tf.cast(attention_mask, dtype=attn_weights.dtype) + attn_weights = tf.reshape(attn_weights, (bsz, self.num_heads, tgt_len, src_len)) + attention_mask + attn_weights = tf.reshape(attn_weights, (bsz * self.num_heads, tgt_len, src_len)) + + attn_weights = stable_softmax(attn_weights, axis=-1) + + if layer_head_mask is not None: + tf.debugging.assert_equal( + shape_list(layer_head_mask), + [self.num_heads], + message=( + f"Head mask for a single layer should be of size {(self.num_heads)}, but is" + f" {shape_list(layer_head_mask)}" + ), + ) + + attn_weights = tf.reshape(layer_head_mask, (1, -1, 1, 1)) * tf.reshape( + attn_weights, (bsz, self.num_heads, tgt_len, src_len) + ) + attn_weights = tf.reshape(attn_weights, (bsz * self.num_heads, tgt_len, src_len)) + + attn_probs = self.dropout(attn_weights, training=training) + attn_output = tf.matmul(attn_probs, value_states) + + tf.debugging.assert_equal( + shape_list(attn_output), + [bsz * self.num_heads, tgt_len, self.head_dim], + message=( + f"`attn_output` should be of size {(bsz, self.num_heads, tgt_len, self.head_dim)}, but is" + f" {shape_list(attn_output)}" + ), + ) + + attn_output = tf.transpose( + tf.reshape(attn_output, (bsz, self.num_heads, tgt_len, self.head_dim)), (0, 2, 1, 3) + ) + attn_output = tf.reshape(attn_output, (bsz, tgt_len, embed_dim)) + + attn_output = self.out_proj(attn_output) + attn_weights: tf.Tensor = tf.reshape(attn_weights, (bsz, self.num_heads, tgt_len, src_len)) + + return attn_output, attn_weights, past_key_value + + def build(self, input_shape=None): + if self.built: + return + self.built = True + if getattr(self, "k_proj", None) is not None: + with tf.name_scope(self.k_proj.name): + self.k_proj.build([None, None, self.embed_dim]) + if getattr(self, "q_proj", None) is not None: + with tf.name_scope(self.q_proj.name): + self.q_proj.build([None, None, self.embed_dim]) + if getattr(self, "v_proj", None) is not None: + with tf.name_scope(self.v_proj.name): + self.v_proj.build([None, None, self.embed_dim]) + if getattr(self, "out_proj", None) is not None: + with tf.name_scope(self.out_proj.name): + self.out_proj.build([None, None, self.embed_dim]) + + +# Copied from transformers.models.wav2vec2.modeling_tf_wav2vec2.TFWav2Vec2FeedForward with Wav2Vec2->Hubert +class TFHubertFeedForward(keras.layers.Layer): + def __init__(self, config: HubertConfig, **kwargs): + super().__init__(**kwargs) + + self.intermediate_dropout = keras.layers.Dropout(config.activation_dropout) + + self.intermediate_dense = keras.layers.Dense( + units=config.intermediate_size, + kernel_initializer=get_initializer(config.initializer_range), + bias_initializer="zeros", + name="intermediate_dense", + ) + self.intermediate_act_fn = get_tf_activation(config.hidden_act) + + self.output_dense = keras.layers.Dense( + units=config.hidden_size, + kernel_initializer=get_initializer(config.initializer_range), + bias_initializer="zeros", + name="output_dense", + ) + self.output_dropout = keras.layers.Dropout(config.hidden_dropout) + self.config = config + + def call(self, hidden_states: tf.Tensor, training: bool = False) -> tf.Tensor: + hidden_states = self.intermediate_dense(hidden_states) + hidden_states = self.intermediate_act_fn(hidden_states) + hidden_states = self.intermediate_dropout(hidden_states, training=training) + + hidden_states = self.output_dense(hidden_states) + hidden_states = self.output_dropout(hidden_states, training=training) + return hidden_states + + def build(self, input_shape=None): + if self.built: + return + self.built = True + if getattr(self, "intermediate_dense", None) is not None: + with tf.name_scope(self.intermediate_dense.name): + self.intermediate_dense.build([None, None, self.config.hidden_size]) + if getattr(self, "output_dense", None) is not None: + with tf.name_scope(self.output_dense.name): + self.output_dense.build([None, None, self.config.intermediate_size]) + + +# Copied from transformers.models.wav2vec2.modeling_tf_wav2vec2.TFWav2Vec2EncoderLayer with Wav2Vec2->Hubert +class TFHubertEncoderLayer(keras.layers.Layer): + def __init__(self, config: HubertConfig, **kwargs): + super().__init__(**kwargs) + self.attention = TFHubertAttention( + embed_dim=config.hidden_size, + num_heads=config.num_attention_heads, + dropout=config.attention_dropout, + is_decoder=False, + name="attention", + ) + self.dropout = keras.layers.Dropout(config.hidden_dropout) + self.layer_norm = keras.layers.LayerNormalization(epsilon=config.layer_norm_eps, name="layer_norm") + self.feed_forward = TFHubertFeedForward(config, name="feed_forward") + self.final_layer_norm = keras.layers.LayerNormalization(epsilon=config.layer_norm_eps, name="final_layer_norm") + self.config = config + + def call( + self, + hidden_states: tf.Tensor, + attention_mask: tf.Tensor | None = None, + output_attentions: Optional[bool] = False, + training: bool = False, + ) -> Tuple[tf.Tensor]: + attn_residual = hidden_states + hidden_states, attn_weights, _ = self.attention( + hidden_states, attention_mask=attention_mask, training=training + ) + hidden_states = self.dropout(hidden_states, training=training) + hidden_states = attn_residual + hidden_states + + hidden_states = self.layer_norm(hidden_states) + hidden_states = hidden_states + self.feed_forward(hidden_states) + hidden_states = self.final_layer_norm(hidden_states) + + outputs = (hidden_states,) + + if output_attentions: + outputs += (attn_weights,) + + return outputs + + def build(self, input_shape=None): + if self.built: + return + self.built = True + if getattr(self, "attention", None) is not None: + with tf.name_scope(self.attention.name): + self.attention.build(None) + if getattr(self, "layer_norm", None) is not None: + with tf.name_scope(self.layer_norm.name): + self.layer_norm.build([None, None, self.config.hidden_size]) + if getattr(self, "feed_forward", None) is not None: + with tf.name_scope(self.feed_forward.name): + self.feed_forward.build(None) + if getattr(self, "final_layer_norm", None) is not None: + with tf.name_scope(self.final_layer_norm.name): + self.final_layer_norm.build([None, None, self.config.hidden_size]) + + +# Copied from transformers.models.wav2vec2.modeling_tf_wav2vec2.TFWav2Vec2EncoderLayerStableLayerNorm with Wav2Vec2->Hubert +class TFHubertEncoderLayerStableLayerNorm(keras.layers.Layer): + def __init__(self, config: HubertConfig, **kwargs): + super().__init__(**kwargs) + self.attention = TFHubertAttention( + embed_dim=config.hidden_size, + num_heads=config.num_attention_heads, + dropout=config.attention_dropout, + is_decoder=False, + name="attention", + ) + self.dropout = keras.layers.Dropout(config.hidden_dropout) + self.layer_norm = keras.layers.LayerNormalization(epsilon=config.layer_norm_eps, name="layer_norm") + self.feed_forward = TFHubertFeedForward(config, name="feed_forward") + self.final_layer_norm = keras.layers.LayerNormalization(epsilon=config.layer_norm_eps, name="final_layer_norm") + self.config = config + + def call( + self, + hidden_states: tf.Tensor, + attention_mask: tf.Tensor | None = None, + output_attentions: Optional[bool] = False, + training: bool = False, + ) -> Tuple[tf.Tensor]: + attn_residual = hidden_states + hidden_states = self.layer_norm(hidden_states) + hidden_states, attn_weights, _ = self.attention( + hidden_states, attention_mask=attention_mask, training=training + ) + hidden_states = self.dropout(hidden_states, training=training) + hidden_states = attn_residual + hidden_states + hidden_states = hidden_states + self.feed_forward(self.final_layer_norm(hidden_states)) + + outputs = (hidden_states,) + + if output_attentions: + outputs += (attn_weights,) + + return outputs + + def build(self, input_shape=None): + if self.built: + return + self.built = True + if getattr(self, "attention", None) is not None: + with tf.name_scope(self.attention.name): + self.attention.build(None) + if getattr(self, "layer_norm", None) is not None: + with tf.name_scope(self.layer_norm.name): + self.layer_norm.build([None, None, self.config.hidden_size]) + if getattr(self, "feed_forward", None) is not None: + with tf.name_scope(self.feed_forward.name): + self.feed_forward.build(None) + if getattr(self, "final_layer_norm", None) is not None: + with tf.name_scope(self.final_layer_norm.name): + self.final_layer_norm.build([None, None, self.config.hidden_size]) + + +# Copied from transformers.models.wav2vec2.modeling_tf_wav2vec2.TFWav2Vec2Encoder with Wav2Vec2->Hubert +class TFHubertEncoder(keras.layers.Layer): + def __init__(self, config: HubertConfig, **kwargs): + super().__init__(**kwargs) + self.config = config + self.pos_conv_embed = TFHubertPositionalConvEmbedding(config, name="pos_conv_embed") + self.layer_norm = keras.layers.LayerNormalization(epsilon=config.layer_norm_eps, name="layer_norm") + self.dropout = keras.layers.Dropout(config.hidden_dropout) + self.layer = [TFHubertEncoderLayer(config, name=f"layers.{i}") for i in range(config.num_hidden_layers)] + + def call( + self, + hidden_states: tf.Tensor, + attention_mask: tf.Tensor | None = None, + output_attentions: Optional[bool] = False, + output_hidden_states: Optional[bool] = False, + return_dict: Optional[bool] = True, + training: Optional[bool] = False, + ) -> Union[TFBaseModelOutput, Tuple[tf.Tensor]]: + all_hidden_states = () if output_hidden_states else None + all_self_attentions = () if output_attentions else None + + if attention_mask is not None: + hidden_states = hidden_states * tf.expand_dims(attention_mask, -1) + attention_mask = _expand_mask(attention_mask) + else: + attention_mask = None + + position_embeddings = self.pos_conv_embed(hidden_states) + hidden_states = hidden_states + position_embeddings + hidden_states = self.layer_norm(hidden_states) + hidden_states = self.dropout(hidden_states, training=training) + + for i, layer_module in enumerate(self.layer): + if output_hidden_states: + all_hidden_states = all_hidden_states + (hidden_states,) + + # add LayerDrop (see https://arxiv.org/abs/1909.11556 for description) + dropout_probability = np.random.uniform(0, 1) + if training and (dropout_probability < self.config.layerdrop): # skip the layer + continue + + layer_outputs = layer_module( + hidden_states=hidden_states, + attention_mask=attention_mask, + output_attentions=output_attentions, + training=training, + ) + hidden_states = layer_outputs[0] + + if output_attentions: + all_self_attentions = all_self_attentions + (layer_outputs[1],) + + # Add last layer + if output_hidden_states: + all_hidden_states = all_hidden_states + (hidden_states,) + + if not return_dict: + return tuple(v for v in [hidden_states, all_hidden_states, all_self_attentions] if v is not None) + return TFBaseModelOutput( + last_hidden_state=hidden_states, + hidden_states=all_hidden_states, + attentions=all_self_attentions, + ) + + def build(self, input_shape=None): + if self.built: + return + self.built = True + if getattr(self, "pos_conv_embed", None) is not None: + with tf.name_scope(self.pos_conv_embed.name): + self.pos_conv_embed.build(None) + if getattr(self, "layer_norm", None) is not None: + with tf.name_scope(self.layer_norm.name): + self.layer_norm.build([None, None, self.config.hidden_size]) + if getattr(self, "layer", None) is not None: + for layer in self.layer: + with tf.name_scope(layer.name): + layer.build(None) + + +# Copied from transformers.models.wav2vec2.modeling_tf_wav2vec2.TFWav2Vec2EncoderStableLayerNorm with Wav2Vec2->Hubert +class TFHubertEncoderStableLayerNorm(keras.layers.Layer): + def __init__(self, config: HubertConfig, **kwargs): + super().__init__(**kwargs) + self.config = config + self.pos_conv_embed = TFHubertPositionalConvEmbedding(config, name="pos_conv_embed") + self.layer_norm = keras.layers.LayerNormalization(epsilon=config.layer_norm_eps, name="layer_norm") + self.dropout = keras.layers.Dropout(config.hidden_dropout) + self.layer = [ + TFHubertEncoderLayerStableLayerNorm(config, name=f"layers.{i}") for i in range(config.num_hidden_layers) + ] + + def call( + self, + hidden_states: tf.Tensor, + attention_mask: tf.Tensor | None = None, + output_attentions: Optional[bool] = False, + output_hidden_states: Optional[bool] = False, + return_dict: Optional[bool] = True, + training: Optional[bool] = False, + ) -> Union[TFBaseModelOutput, Tuple[tf.Tensor]]: + all_hidden_states = () if output_hidden_states else None + all_self_attentions = () if output_attentions else None + + if attention_mask is not None: + hidden_states = hidden_states * tf.expand_dims(attention_mask, -1) + attention_mask = _expand_mask(attention_mask) + else: + attention_mask = None + + position_embeddings = self.pos_conv_embed(hidden_states) + hidden_states = hidden_states + position_embeddings + hidden_states = self.dropout(hidden_states, training=training) + + for i, layer_module in enumerate(self.layer): + if output_hidden_states: + all_hidden_states = all_hidden_states + (hidden_states,) + + # add LayerDrop (see https://arxiv.org/abs/1909.11556 for description) + dropout_probability = np.random.uniform(0, 1) + if training and (dropout_probability < self.config.layerdrop): # skip the layer + continue + + layer_outputs = layer_module( + hidden_states=hidden_states, + attention_mask=attention_mask, + output_attentions=output_attentions, + training=training, + ) + hidden_states = layer_outputs[0] + + if output_attentions: + all_self_attentions = all_self_attentions + (layer_outputs[1],) + + hidden_states = self.layer_norm(hidden_states) + + if output_hidden_states: + all_hidden_states = all_hidden_states + (hidden_states,) + + if not return_dict: + return tuple(v for v in [hidden_states, all_hidden_states, all_self_attentions] if v is not None) + return TFBaseModelOutput( + last_hidden_state=hidden_states, + hidden_states=all_hidden_states, + attentions=all_self_attentions, + ) + + def build(self, input_shape=None): + if self.built: + return + self.built = True + if getattr(self, "pos_conv_embed", None) is not None: + with tf.name_scope(self.pos_conv_embed.name): + self.pos_conv_embed.build(None) + if getattr(self, "layer_norm", None) is not None: + with tf.name_scope(self.layer_norm.name): + self.layer_norm.build([None, None, self.config.hidden_size]) + if getattr(self, "layer", None) is not None: + for layer in self.layer: + with tf.name_scope(layer.name): + layer.build(None) + + +@keras_serializable +class TFHubertMainLayer(keras.layers.Layer): + config_class = HubertConfig + + def __init__(self, config: HubertConfig, **kwargs): + super().__init__(**kwargs) + self.config = config + self.feature_extractor = TFHubertFeatureEncoder(config, name="feature_extractor") + self.feature_projection = TFHubertFeatureProjection(config, name="feature_projection") + + if config.do_stable_layer_norm: + self.encoder = TFHubertEncoderStableLayerNorm(config, name="encoder") + else: + self.encoder = TFHubertEncoder(config, name="encoder") + + def build(self, input_shape=None): + self.masked_spec_embed = self.add_weight( + shape=(self.config.hidden_size,), initializer="uniform", trainable=True, name="masked_spec_embed" + ) + + if self.built: + return + self.built = True + if getattr(self, "feature_extractor", None) is not None: + with tf.name_scope(self.feature_extractor.name): + self.feature_extractor.build(None) + if getattr(self, "feature_projection", None) is not None: + with tf.name_scope(self.feature_projection.name): + self.feature_projection.build(None) + if getattr(self, "encoder", None) is not None: + with tf.name_scope(self.encoder.name): + self.encoder.build(None) + + def _get_feat_extract_output_lengths(self, input_lengths: tf.Tensor): + """ + Computes the output length of the convolutional layers + """ + + def _conv_out_length(input_length, kernel_size, stride): + # 1D convolutional layer output length formula taken + # from https://pytorch.org/docs/stable/generated/torch.nn.Conv1d.html + return (input_length - kernel_size) // stride + 1 + + for kernel_size, stride in zip(self.config.conv_kernel, self.config.conv_stride): + input_lengths = _conv_out_length(input_lengths, kernel_size, stride) + + return input_lengths + + def _mask_hidden_states(self, hidden_states: tf.Tensor, mask_time_indices: tf.Tensor | None = None): + """ + Masks extracted features along time axis and/or along feature axis according to + [SpecAugment](https://arxiv.org/abs/1904.08779). + """ + batch_size, sequence_length, hidden_size = shape_list(hidden_states) + + # `config.apply_spec_augment` can set masking to False + if not getattr(self.config, "apply_spec_augment", True): + return hidden_states + + if mask_time_indices is not None: + # apply SpecAugment along time axis with given mask_time_indices + hidden_states = tf.where( + tf.cast(mask_time_indices[:, :, tf.newaxis], tf.bool), + self.masked_spec_embed[tf.newaxis, tf.newaxis, :], + hidden_states, + ) + + elif self.config.mask_time_prob > 0: + # generate indices & apply SpecAugment along time axis + mask_time_indices = _compute_mask_indices( + (batch_size, sequence_length), + mask_prob=self.config.mask_time_prob, + mask_length=self.config.mask_time_length, + min_masks=2, + ) + hidden_states = tf.where( + tf.cast(mask_time_indices[:, :, tf.newaxis], tf.bool), + self.masked_spec_embed[tf.newaxis, tf.newaxis, :], + hidden_states, + ) + + # apply SpecAugment along feature axis + if self.config.mask_feature_prob > 0: + mask_feature_indices = _compute_mask_indices( + (batch_size, hidden_size), + mask_prob=self.config.mask_feature_prob, + mask_length=self.config.mask_feature_length, + ) + hidden_states = tf.where(mask_feature_indices[:, tf.newaxis, :], hidden_states, 0) + + return hidden_states + + @unpack_inputs + def call( + self, + input_values: tf.Tensor, + attention_mask: tf.Tensor | None = None, + token_type_ids: tf.Tensor | None = None, + position_ids: tf.Tensor | None = None, + head_mask: tf.Tensor | None = None, + inputs_embeds: tf.Tensor | None = None, + output_attentions: tf.Tensor | None = None, + output_hidden_states: tf.Tensor | None = None, + return_dict: Optional[bool] = None, + training: bool = False, + **kwargs: Any, + ): + hidden_states = self.feature_extractor(tf.cast(input_values, tf.float32), training=training) + + if attention_mask is not None: + # compute real output lengths according to convolution formula + output_lengths = self._get_feat_extract_output_lengths(tf.reduce_sum(attention_mask, -1)) + + attention_mask = tf.sequence_mask( + output_lengths, maxlen=shape_list(hidden_states)[1], dtype=hidden_states.dtype + ) + + hidden_states = self.feature_projection(hidden_states, training=training) + + mask_time_indices = kwargs.get("mask_time_indices", None) + if training: + hidden_states = self._mask_hidden_states(hidden_states, mask_time_indices=mask_time_indices) + + encoder_outputs = self.encoder( + hidden_states, + attention_mask=attention_mask, + output_attentions=output_attentions, + output_hidden_states=output_hidden_states, + return_dict=return_dict, + training=training, + ) + hidden_states = encoder_outputs[0] + + if not return_dict: + return (hidden_states,) + encoder_outputs[1:] + + return TFBaseModelOutput( + last_hidden_state=hidden_states, + hidden_states=encoder_outputs.hidden_states, + attentions=encoder_outputs.attentions, + ) + + +class TFHubertPreTrainedModel(TFPreTrainedModel): + """ + An abstract class to handle weights initialization and a simple interface for downloading and loading pretrained + models. + """ + + config_class = HubertConfig + base_model_prefix = "hubert" + main_input_name = "input_values" + + @property + def input_signature(self): + return { + "input_values": tf.TensorSpec((None, 16000), tf.float32, name="input_values"), + "attention_mask": tf.TensorSpec((None, None), tf.int32, name="attention_mask"), + "token_type_ids": tf.TensorSpec((None, None), tf.int32, name="token_type_ids"), + } + + def __init__(self, config, *inputs, **kwargs): + super().__init__(config, *inputs, **kwargs) + logger.warning( + f"\n{self.__class__.__name__} has backpropagation operations that are NOT supported on CPU. If you wish " + "to train/fine-tune this model, you need a GPU or a TPU" + ) + + +HUBERT_START_DOCSTRING = r""" + + This model inherits from [`TFPreTrainedModel`]. Check the superclass documentation for the generic methods the + library implements for all its model (such as downloading or saving, resizing the input embeddings, pruning heads + etc.) + + This model is also a [keras.Model](https://www.tensorflow.org/api_docs/python/tf/keras/Model) subclass. Use it + as a regular TF 2.0 Keras Model and refer to the TF 2.0 documentation for all matter related to general usage and + behavior. + + + + TensorFlow models and layers in `transformers` accept two formats as input: + + - having all inputs as keyword arguments (like PyTorch models), or + - having all inputs as a list, tuple or dict in the first positional argument. + + The reason the second format is supported is that Keras methods prefer this format when passing inputs to models + and layers. Because of this support, when using methods like `model.fit()` things should "just work" for you - just + pass your inputs and labels in any format that `model.fit()` supports! If, however, you want to use the second + format outside of Keras methods like `fit()` and `predict()`, such as when creating your own layers or models with + the Keras `Functional` API, there are three possibilities you can use to gather all the input Tensors in the first + positional argument: + + - a single Tensor with `input_values` only and nothing else: `model(input_values)` + - a list of varying length with one or several input Tensors IN THE ORDER given in the docstring: + `model([input_values, attention_mask])` or `model([input_values, attention_mask, token_type_ids])` + - a dictionary with one or several input Tensors associated to the input names given in the docstring: + `model({"input_values": input_values, "token_type_ids": token_type_ids})` + + Note that when creating models and layers with + [subclassing](https://keras.io/guides/making_new_layers_and_models_via_subclassing/) then you don't need to worry + about any of this, as you can just pass inputs like you would to any other Python function! + + + + Args: + config ([`HubertConfig`]): Model configuration class with all the parameters of the model. + Initializing with a config file does not load the weights associated with the model, only the + configuration. Check out the [`~PreTrainedModel.from_pretrained`] method to load the model weights. +""" + +HUBERT_INPUTS_DOCSTRING = r""" + Args: + input_values (`np.ndarray`, `tf.Tensor`, `List[tf.Tensor]` `Dict[str, tf.Tensor]` or `Dict[str, np.ndarray]` and each example must have the shape `({0})`): + Indices of input sequence tokens in the vocabulary. + + Indices can be obtained using [`AutoTokenizer`]. See [`PreTrainedTokenizer.__call__`] and + [`PreTrainedTokenizer.encode`] for details. + + [What are input IDs?](../glossary#input-ids) + attention_mask (`np.ndarray` or `tf.Tensor` of shape `({0})`, *optional*): + Mask to avoid performing attention on padding token indices. Mask values selected in `[0, 1]`: + + - 1 for tokens that are **not masked**, + - 0 for tokens that are **masked**. + + [What are attention masks?](../glossary#attention-mask) + token_type_ids (`np.ndarray` or `tf.Tensor` of shape `({0})`, *optional*): + Segment token indices to indicate first and second portions of the inputs. Indices are selected in `[0, + 1]`: + + - 0 corresponds to a *sentence A* token, + - 1 corresponds to a *sentence B* token. + + [What are token type IDs?](../glossary#token-type-ids) + position_ids (`np.ndarray` or `tf.Tensor` of shape `({0})`, *optional*): + Indices of positions of each input sequence tokens in the position embeddings. Selected in the range `[0, + config.max_position_embeddings - 1]`. + + [What are position IDs?](../glossary#position-ids) + head_mask (`np.ndarray` or `tf.Tensor` of shape `(num_heads,)` or `(num_layers, num_heads)`, *optional*): + Mask to nullify selected heads of the self-attention modules. Mask values selected in `[0, 1]`: + + - 1 indicates the head is **not masked**, + - 0 indicates the head is **masked**. + + inputs_embeds (`np.ndarray` or `tf.Tensor` of shape `({0}, hidden_size)`, *optional*): + Optionally, instead of passing `input_values` you can choose to directly pass an embedded representation. + This is useful if you want more control over how to convert `input_values` indices into associated vectors + than the model's internal embedding lookup matrix. + output_attentions (`bool`, *optional*): + Whether or not to return the attentions tensors of all attention layers. See `attentions` under returned + tensors for more detail. This argument can be used only in eager mode, in graph mode the value in the + config will be used instead. + output_hidden_states (`bool`, *optional*): + Whether or not to return the hidden states of all layers. See `hidden_states` under returned tensors for + more detail. This argument can be used only in eager mode, in graph mode the value in the config will be + used instead. + return_dict (`bool`, *optional*): + Whether or not to return a [`~utils.ModelOutput`] instead of a plain tuple. This argument can be used in + eager mode, in graph mode the value will always be set to True. + training (`bool`, *optional*, defaults to `False``): + Whether or not to use the model in training mode (some modules like dropout modules have different + behaviors between training and evaluation). +""" + + +@add_start_docstrings( + "The bare TFHubert Model transformer outputing raw hidden-states without any specific head on top.", + HUBERT_START_DOCSTRING, +) +class TFHubertModel(TFHubertPreTrainedModel): + def __init__(self, config: HubertConfig, *inputs, **kwargs): + super().__init__(config, *inputs, **kwargs) + self.config = config + self.hubert = TFHubertMainLayer(config, name="hubert") + + @add_start_docstrings_to_model_forward(HUBERT_INPUTS_DOCSTRING) + @replace_return_docstrings(output_type=TFBaseModelOutput, config_class=_CONFIG_FOR_DOC) + @unpack_inputs + def call( + self, + input_values: tf.Tensor, + attention_mask: tf.Tensor | None = None, + token_type_ids: tf.Tensor | None = None, + position_ids: tf.Tensor | None = None, + head_mask: tf.Tensor | None = None, + inputs_embeds: tf.Tensor | None = None, + output_attentions: Optional[bool] = None, + output_hidden_states: Optional[bool] = None, + return_dict: Optional[bool] = None, + training: bool = False, + ) -> Union[TFBaseModelOutput, Tuple[tf.Tensor]]: + """ + + Returns: + + Example: + + ```python + >>> from transformers import AutoProcessor, TFHubertModel + >>> from datasets import load_dataset + >>> import soundfile as sf + + >>> processor = AutoProcessor.from_pretrained("facebook/hubert-large-ls960-ft") + >>> model = TFHubertModel.from_pretrained("facebook/hubert-large-ls960-ft") + + + >>> def map_to_array(batch): + ... speech, _ = sf.read(batch["file"]) + ... batch["speech"] = speech + ... return batch + + + >>> ds = load_dataset("hf-internal-testing/librispeech_asr_dummy", "clean", split="validation") + >>> ds = ds.map(map_to_array) + + >>> input_values = processor(ds["speech"][0], return_tensors="tf").input_values # Batch size 1 + >>> hidden_states = model(input_values).last_hidden_state + ```""" + + output_hidden_states = output_hidden_states if output_hidden_states else self.config.output_hidden_states + output_attentions = output_attentions if output_attentions else self.config.output_attentions + return_dict = return_dict if return_dict else self.config.return_dict + + outputs = self.hubert( + input_values=input_values, + attention_mask=attention_mask, + token_type_ids=token_type_ids, + position_ids=position_ids, + head_mask=head_mask, + inputs_embeds=inputs_embeds, + output_attentions=output_attentions, + output_hidden_states=output_hidden_states, + return_dict=return_dict, + training=training, + ) + + return outputs + + def build(self, input_shape=None): + if self.built: + return + self.built = True + if getattr(self, "hubert", None) is not None: + with tf.name_scope(self.hubert.name): + self.hubert.build(None) + + +@add_start_docstrings( + """TFHubert Model with a `language modeling` head on top for Connectionist Temporal Classification (CTC).""", + HUBERT_START_DOCSTRING, +) +class TFHubertForCTC(TFHubertPreTrainedModel): + def __init__(self, config: HubertConfig, *inputs, **kwargs): + super().__init__(config, *inputs, **kwargs) + + self.hubert = TFHubertMainLayer(config, name="hubert") + self.dropout = keras.layers.Dropout(config.final_dropout) + self.lm_head = keras.layers.Dense(config.vocab_size, name="lm_head") + self.output_hidden_size = ( + config.output_hidden_size if hasattr(config, "add_adapter") and config.add_adapter else config.hidden_size + ) + + def freeze_feature_extractor(self): + """ + Calling this function will disable the gradient computation for the feature encoder so that its parameters will + not be updated during training. + """ + warnings.warn( + "The method `freeze_feature_extractor` is deprecated and will be removed in Transformers v5. " + "Please use the equivalent `freeze_feature_encoder` method instead.", + FutureWarning, + ) + self.freeze_feature_encoder() + + def freeze_feature_encoder(self): + """ + Calling this function will disable the gradient computation for the feature encoder so that its parameter will + not be updated during training. + """ + self.hubert.feature_extractor.trainable = False + + @add_start_docstrings_to_model_forward(HUBERT_INPUTS_DOCSTRING) + @replace_return_docstrings(output_type=TFCausalLMOutput, config_class=_CONFIG_FOR_DOC) + @unpack_inputs + def call( + self, + input_values: tf.Tensor, + attention_mask: tf.Tensor | None = None, + token_type_ids: tf.Tensor | None = None, + position_ids: tf.Tensor | None = None, + head_mask: tf.Tensor | None = None, + inputs_embeds: tf.Tensor | None = None, + output_attentions: Optional[bool] = None, + labels: tf.Tensor | None = None, + output_hidden_states: Optional[bool] = None, + return_dict: Optional[bool] = None, + training: Optional[bool] = False, + ) -> Union[TFCausalLMOutput, Tuple[tf.Tensor]]: + r""" + labels (`tf.Tensor` or `np.ndarray` of shape `(batch_size, sequence_length)`, *optional*): + Labels for computing the masked language modeling loss. Indices should be in `[-100, 0, ..., + config.vocab_size]` (see `input_values` docstring) Tokens with indices set to `-100` are ignored (masked), + the loss is only computed for the tokens with labels in `[0, ..., config.vocab_size]` + + Returns: + + Example: + + ```python + >>> import tensorflow as tf + >>> from transformers import AutoProcessor, TFHubertForCTC + >>> from datasets import load_dataset + >>> import soundfile as sf + + >>> processor = AutoProcessor.from_pretrained("facebook/hubert-large-ls960-ft") + >>> model = TFHubertForCTC.from_pretrained("facebook/hubert-large-ls960-ft") + + + >>> def map_to_array(batch): + ... speech, _ = sf.read(batch["file"]) + ... batch["speech"] = speech + ... return batch + + + >>> ds = load_dataset("hf-internal-testing/librispeech_asr_dummy", "clean", split="validation") + >>> ds = ds.map(map_to_array) + + >>> input_values = processor(ds["speech"][0], return_tensors="tf").input_values # Batch size 1 + >>> logits = model(input_values).logits + >>> predicted_ids = tf.argmax(logits, axis=-1) + + >>> transcription = processor.decode(predicted_ids[0]) + + >>> # compute loss + >>> target_transcription = "A MAN SAID TO THE UNIVERSE SIR I EXIST" + + >>> # Pass the transcription as text to encode labels + >>> labels = processor(text=transcription, return_tensors="tf").input_values + + >>> loss = model(input_values, labels=labels).loss + ```""" + + outputs = self.hubert( + input_values=input_values, + attention_mask=attention_mask, + token_type_ids=token_type_ids, + position_ids=position_ids, + head_mask=head_mask, + inputs_embeds=inputs_embeds, + output_attentions=output_attentions, + output_hidden_states=output_hidden_states, + return_dict=return_dict, + training=training, + ) + hidden_states = outputs[0] + hidden_states = self.dropout(hidden_states, training=training) + + logits = self.lm_head(hidden_states) + + if labels is not None: + if tf.reduce_max(labels) >= self.config.vocab_size: + raise ValueError(f"Label values must be <= vocab_size: {self.config.vocab_size}") + + attention_mask = ( + attention_mask if attention_mask is not None else tf.ones_like(input_values, dtype=tf.float32) + ) + input_lengths = self.hubert._get_feat_extract_output_lengths(tf.reduce_sum(attention_mask, axis=-1)) + + # assuming that padded tokens are filled with -100 + # when not being attended to + labels_mask = tf.cast(labels >= 0, tf.int32) + target_lengths = tf.reduce_sum(labels_mask, axis=-1) + + loss = tf.nn.ctc_loss( + logits=logits, + labels=labels, + logit_length=input_lengths, + label_length=target_lengths, + blank_index=self.config.pad_token_id, + logits_time_major=False, + ) + + if self.config.ctc_loss_reduction == "sum": + loss = tf.reduce_sum(loss) + loss = tf.reshape(loss, (1,)) + if self.config.ctc_loss_reduction == "mean": + loss = tf.reduce_mean(loss) + loss = tf.reshape(loss, (1,)) + else: + loss = None + + if not return_dict: + output = (logits,) + outputs[1:] + return ((loss,) + output) if loss is not None else output + + return TFCausalLMOutput( + loss=loss, + logits=logits, + hidden_states=outputs.hidden_states, + attentions=outputs.attentions, + ) + + def build(self, input_shape=None): + if self.built: + return + self.built = True + if getattr(self, "hubert", None) is not None: + with tf.name_scope(self.hubert.name): + self.hubert.build(None) + if getattr(self, "lm_head", None) is not None: + with tf.name_scope(self.lm_head.name): + self.lm_head.build([None, None, self.output_hidden_size]) diff --git a/parrot/lib/python3.10/site-packages/transformers/models/idefics/__pycache__/__init__.cpython-310.pyc b/parrot/lib/python3.10/site-packages/transformers/models/idefics/__pycache__/__init__.cpython-310.pyc new file mode 100644 index 0000000000000000000000000000000000000000..f990ffcedd0612ec38e2c734ce06154793dc6fce Binary files /dev/null and b/parrot/lib/python3.10/site-packages/transformers/models/idefics/__pycache__/__init__.cpython-310.pyc differ diff --git a/parrot/lib/python3.10/site-packages/transformers/models/idefics/__pycache__/configuration_idefics.cpython-310.pyc b/parrot/lib/python3.10/site-packages/transformers/models/idefics/__pycache__/configuration_idefics.cpython-310.pyc new file mode 100644 index 0000000000000000000000000000000000000000..511be3072bd5ad2fc8febd5c169153c99c9b517c Binary files /dev/null and b/parrot/lib/python3.10/site-packages/transformers/models/idefics/__pycache__/configuration_idefics.cpython-310.pyc differ diff --git a/parrot/lib/python3.10/site-packages/transformers/models/idefics/__pycache__/image_processing_idefics.cpython-310.pyc b/parrot/lib/python3.10/site-packages/transformers/models/idefics/__pycache__/image_processing_idefics.cpython-310.pyc new file mode 100644 index 0000000000000000000000000000000000000000..3680a6191ece162c4e5648b642d8934833ccb8ad Binary files /dev/null and b/parrot/lib/python3.10/site-packages/transformers/models/idefics/__pycache__/image_processing_idefics.cpython-310.pyc differ diff --git a/parrot/lib/python3.10/site-packages/transformers/models/idefics/__pycache__/modeling_idefics.cpython-310.pyc b/parrot/lib/python3.10/site-packages/transformers/models/idefics/__pycache__/modeling_idefics.cpython-310.pyc new file mode 100644 index 0000000000000000000000000000000000000000..a6a75b84091a85893b551435f86011b3e2958ba4 Binary files /dev/null and b/parrot/lib/python3.10/site-packages/transformers/models/idefics/__pycache__/modeling_idefics.cpython-310.pyc differ diff --git a/parrot/lib/python3.10/site-packages/transformers/models/idefics/__pycache__/modeling_tf_idefics.cpython-310.pyc b/parrot/lib/python3.10/site-packages/transformers/models/idefics/__pycache__/modeling_tf_idefics.cpython-310.pyc new file mode 100644 index 0000000000000000000000000000000000000000..db316be223c9f0de37252b80ecbf9f907323855f Binary files /dev/null and b/parrot/lib/python3.10/site-packages/transformers/models/idefics/__pycache__/modeling_tf_idefics.cpython-310.pyc differ diff --git a/parrot/lib/python3.10/site-packages/transformers/models/idefics/__pycache__/perceiver.cpython-310.pyc b/parrot/lib/python3.10/site-packages/transformers/models/idefics/__pycache__/perceiver.cpython-310.pyc new file mode 100644 index 0000000000000000000000000000000000000000..335577b9226925d4225ad29f71dedf4e2935119c Binary files /dev/null and b/parrot/lib/python3.10/site-packages/transformers/models/idefics/__pycache__/perceiver.cpython-310.pyc differ diff --git a/parrot/lib/python3.10/site-packages/transformers/models/idefics/__pycache__/perceiver_tf.cpython-310.pyc b/parrot/lib/python3.10/site-packages/transformers/models/idefics/__pycache__/perceiver_tf.cpython-310.pyc new file mode 100644 index 0000000000000000000000000000000000000000..3b7cffdb9f1ecc335be978affa29ea0ccf615129 Binary files /dev/null and b/parrot/lib/python3.10/site-packages/transformers/models/idefics/__pycache__/perceiver_tf.cpython-310.pyc differ diff --git a/parrot/lib/python3.10/site-packages/transformers/models/idefics/__pycache__/processing_idefics.cpython-310.pyc b/parrot/lib/python3.10/site-packages/transformers/models/idefics/__pycache__/processing_idefics.cpython-310.pyc new file mode 100644 index 0000000000000000000000000000000000000000..ad100667025295dff055564970f4e7694e905c0b Binary files /dev/null and b/parrot/lib/python3.10/site-packages/transformers/models/idefics/__pycache__/processing_idefics.cpython-310.pyc differ diff --git a/parrot/lib/python3.10/site-packages/transformers/models/idefics/__pycache__/vision.cpython-310.pyc b/parrot/lib/python3.10/site-packages/transformers/models/idefics/__pycache__/vision.cpython-310.pyc new file mode 100644 index 0000000000000000000000000000000000000000..de5514e6f48700a7a961e3486084f39a3bc00b03 Binary files /dev/null and b/parrot/lib/python3.10/site-packages/transformers/models/idefics/__pycache__/vision.cpython-310.pyc differ diff --git a/parrot/lib/python3.10/site-packages/transformers/models/idefics/__pycache__/vision_tf.cpython-310.pyc b/parrot/lib/python3.10/site-packages/transformers/models/idefics/__pycache__/vision_tf.cpython-310.pyc new file mode 100644 index 0000000000000000000000000000000000000000..906f85976fcf474021f656800c55d1ace10b6584 Binary files /dev/null and b/parrot/lib/python3.10/site-packages/transformers/models/idefics/__pycache__/vision_tf.cpython-310.pyc differ diff --git a/parrot/lib/python3.10/site-packages/transformers/models/idefics/perceiver.py b/parrot/lib/python3.10/site-packages/transformers/models/idefics/perceiver.py new file mode 100644 index 0000000000000000000000000000000000000000..888c5b0bb9395548c90deac4a70350d1ad39e2d8 --- /dev/null +++ b/parrot/lib/python3.10/site-packages/transformers/models/idefics/perceiver.py @@ -0,0 +1,188 @@ +# This code was adapted from https://github.com/lucidrains/flamingo-pytorch licensed under the MIT License. +# +# MIT License +# +# Copyright (c) 2020 The Google AI Language Team Authors, The HuggingFace Inc. team and github/lonePatient +# +# Permission is hereby granted, free of charge, to any person obtaining a copy +# of this software and associated documentation files (the "Software"), to deal +# in the Software without restriction, including without limitation the rights +# to use, copy, modify, merge, publish, distribute, sublicense, and/or sell +# copies of the Software, and to permit persons to whom the Software is +# furnished to do so, subject to the following conditions: +# +# The above copyright notice and this permission notice shall be included in all +# copies or substantial portions of the Software. +# +# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR +# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, +# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE +# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER +# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, +# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE +# SOFTWARE. + + +""" + +Generic interface to various configurations of the Perceiver Resampler, that simply takes in a series of (potentially +time-indexed) contextual embeddings, and "resamples" (compresses) them down to a pre-specified number of latents! Note +that the Perceiver in general resamples based solely off the *long-range* context; there's a nice opportunity here to +prime the Perceiver Resampler with say a single layer's worth of language embeddings (the target domain), and use that +to softly "retrieve & compress" what we need --> this would be a novel contribution we should explore. + +References: + - DeepMind's Flamingo: https://www.deepmind.com/blog/tackling-multiple-tasks-with-a-single-visual-language-model + - Code borrowed w/ love from: https://github.com/lucidrains/flamingo-pytorch + +""" +from typing import Optional, Tuple + +import torch +import torch.nn as nn + +from .configuration_idefics import IdeficsConfig + + +class IdeficsPerceiverResampler(nn.Module): + def __init__( + self, config: IdeficsConfig, embed_dim: int, depth: int, n_heads: int, head_dim: int, n_latents: int + ) -> None: + """ + Instantiates a Perceiver Resampler that operates over a sequence of embeddings (say from a ResNet or ViT or + MAE) of a given dimension, performs `depth` blocks of cross-attention with a fixed `n_latents` inputs, then + returns a Tensor of shape [bsz, n_latents, embed_dim]. :param embed_dim: Dimensionality of embeddings being fed + to the Perceiver Resampler (also dimensionality of latent embeddings *returned* by the Perceiver Resampler. + Could be e.g., VIT embed_dim, ResNet pool dim, and so on. + + Args: + config (`IdeficsConfig`): config object + embed_dim (`int`): The size of each embedding vector + depth (`int`): Depth of the Perceiver Resampler (Transformer w/ cross attention). Should be shallow (< 3). + n_heads (`int`): Number of heads in each Transformer block (for multi-headed self-attention). + head_dim (`int`): Dimensionality of each head projection in the Transformer block. + n_latents (`int`): + Number of latent embeddings to resample ("compress") the input sequence to (usually < 128). + + """ + super().__init__() + self.embed_dim, self.n_heads, self.head_dim, self.n_latents = embed_dim, n_heads, head_dim, n_latents + self.qk_layer_norms = config.perceiver_config.qk_layer_norms_perceiver + + # Create Latents for Perceiver + self.latents = nn.Parameter(torch.randn(self.n_latents, self.embed_dim), requires_grad=True) + + self.intermediate_dim = ( + self.embed_dim * 4 + if not hasattr(config.vision_config, "embed_dim") + else config.vision_config.embed_dim * 4 + ) + # Create Transformer Blocks + self.blocks = nn.ModuleList( + [ + nn.ModuleList( + [ + IdeficsPerceiverAttention(self.embed_dim, self.n_heads, self.head_dim, self.qk_layer_norms), + IdeficsMLP(self.intermediate_dim, config), + ] + ) + for _ in range(depth) + ] + ) + self.layer_norm = nn.LayerNorm(self.embed_dim) + + def forward(self, context: torch.Tensor) -> torch.Tensor: + """Resample arbitrary length context & *compress* down to self.n_latents latent embeddings""" + # einsum.repeat(self.latents, "seq embed -> bsz seq embed", bsz=context.shape[0]) + latents = self.latents.repeat(context.shape[0], 1, 1) + + # Feed through Perceiver Attention blocks... + for attn, ff in self.blocks: + latents = attn(context, latents) + latents + latents = ff(latents) + latents + + return self.layer_norm(latents) + + +class IdeficsPerceiverAttention(nn.Module): + def __init__(self, embed_dim: int, n_heads: int, head_dim: int, qk_layer_norms: bool) -> None: + """Perceiver Cross-Attention Module --> let long-form inputs be `context`, resampled embeddings be `latents`""" + super().__init__() + self.embed_dim, self.n_heads, self.head_dim = embed_dim, n_heads, head_dim + self.qk_layer_norms = qk_layer_norms + # Normalization & Scaling + self.context_layer_norm = nn.LayerNorm(self.embed_dim) + self.latents_layer_norm = nn.LayerNorm(self.embed_dim) + if self.qk_layer_norms: + self.q_layer_norm = nn.LayerNorm(self.head_dim) + self.k_layer_norm = nn.LayerNorm(self.head_dim) + + self.qk_scale = self.head_dim**-0.5 + + # Q, K, V Projection (no bias -- detail from Perceiver/Flamingo Papers). + self.q_proj = nn.Linear(self.embed_dim, self.n_heads * self.head_dim, bias=False) + self.k_proj = nn.Linear(self.embed_dim, self.n_heads * self.head_dim, bias=False) + self.v_proj = nn.Linear(self.embed_dim, self.n_heads * self.head_dim, bias=False) + + self.output_proj = nn.Linear(self.n_heads * self.head_dim, embed_dim, bias=False) + + def forward(self, context: torch.Tensor, latents: torch.Tensor) -> torch.Tensor: + """ + Runs Perceiver Self-Attention, with special (context, latents) appended along the `seq` dimension! + + Args: + context (`torch.Tensor`): + Tensor of shape `[bsz, seq, embed_dim]` representing long-form context to resample. + latents (`torch.Tensor`): + Tensor of shape `[bsz, n_latents, embed_dim]` representing fixed length latents to compress to. + + Returns: + `torch.Tensor`: Tensor of shape `[bsz, n_latents, embed_dim]` representing attention over latents w/ cross + from context. + """ + context = self.context_layer_norm(context) + latents = self.latents_layer_norm(latents) + batch_size, seq_length, embed_dim = context.shape[:3] + + # Query, Key, Value Projections --> Note that in Flamingo, latents are *concatenated* with context prior to attn! + # Note: This results in queries w/ `seq = n_latents`, and keys, values with `seq = len(context) + n_latents` + q = self.q_proj(latents) + k = self.k_proj(torch.cat([context, latents], dim=-2)) + v = self.v_proj(torch.cat([context, latents], dim=-2)) + + # Multiheaded Self-Attention w/ stable softmax (subtract per-row max -- `amax` -- before softmax call) + # =>> `attn` should be a 2D matrix of shape [n_latents x (context + n_latents)] + # einsum.rearrange(x, "bsz seq (heads embed) -> bsz heads seq embed", heads=self.n_heads) + q, k, v = [x.reshape(batch_size, x.shape[1], self.n_heads, self.head_dim).transpose(1, 2) for x in (q, k, v)] + + if self.qk_layer_norms: + q = self.q_layer_norm(q) + k = self.k_layer_norm(k) + + scores = torch.einsum("... i d, ... j d -> ... i j", q * self.qk_scale, k) + stabilized_scores = scores - (scores.amax(dim=-1, keepdim=True).detach()) + attn = stabilized_scores.softmax(dim=-1) + + # Attend & project back to output... + resampled = torch.einsum("... i j, ... j d -> ... i d", attn, v) + # einsum.rearrange(resampled, "bsz heads seq embed -> bsz seq (heads embed)", heads=self.n_heads) + return self.output_proj(resampled.transpose(1, 2).flatten(-2)) + + +class IdeficsMLP(nn.Module): + def __init__(self, intermediate_size, config: IdeficsConfig): + """Simple MLP block with intermediate_size and embedding size""" + super().__init__() + self.embed_dim = config.vision_config.embed_dim + self.ln = nn.LayerNorm(self.embed_dim) + self.fc = nn.Linear(self.embed_dim, intermediate_size, bias=False) + self.act = nn.ReLU() + self.c_proj = nn.Linear(intermediate_size, self.embed_dim, bias=False) + + def forward(self, hidden_states: Optional[Tuple[torch.FloatTensor]]) -> torch.FloatTensor: + hidden_states = self.ln(hidden_states) + hidden_states = self.fc(hidden_states) + hidden_states = self.act(hidden_states) + hidden_states = self.c_proj(hidden_states) + + return hidden_states diff --git a/parrot/lib/python3.10/site-packages/transformers/models/idefics/perceiver_tf.py b/parrot/lib/python3.10/site-packages/transformers/models/idefics/perceiver_tf.py new file mode 100644 index 0000000000000000000000000000000000000000..c9e76004a70ddc210e1ad3e113bb4e31962436f7 --- /dev/null +++ b/parrot/lib/python3.10/site-packages/transformers/models/idefics/perceiver_tf.py @@ -0,0 +1,194 @@ +# This code was adapted from https://github.com/lucidrains/flamingo-pytorch licensed under the MIT License. +# +# MIT License +# +# Copyright (c) 2020 The Google AI Language Team Authors, The HuggingFace Inc. team and github/lonePatient +# +# Permission is hereby granted, free of charge, to any person obtaining a copy +# of this software and associated documentation files (the "Software"), to deal +# in the Software without restriction, including without limitation the rights +# to use, copy, modify, merge, publish, distribute, sublicense, and/or sell +# copies of the Software, and to permit persons to whom the Software is +# furnished to do so, subject to the following conditions: +# +# The above copyright notice and this permission notice shall be included in all +# copies or substantial portions of the Software. +# +# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR +# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, +# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE +# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER +# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, +# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE +# SOFTWARE. + + +""" + +Generic interface to various configurations of the Perceiver Resampler, that simply takes in a series of (potentially +time-indexed) contextual embeddings, and "resamples" (compresses) them down to a pre-specified number of latents! Note +that the Perceiver in general resamples based solely off the *long-range* context; there's a nice opportunity here to +prime the Perceiver Resampler with say a single layer's worth of language embeddings (the target domain), and use that +to softly "retrieve & compress" what we need --> this would be a novel contribution we should explore. + +References: + - DeepMind's Flamingo: https://www.deepmind.com/blog/tackling-multiple-tasks-with-a-single-visual-language-model + - Code borrowed w/ love from: https://github.com/lucidrains/flamingo-pytorch + +""" +from typing import Optional, Tuple + +import tensorflow as tf + +from ...modeling_tf_utils import shape_list +from .configuration_idefics import IdeficsConfig + + +class TFIdeficsPerceiverResampler(tf.keras.layers.Layer): + def __init__( + self, config: IdeficsConfig, embed_dim: int, depth: int, n_heads: int, head_dim: int, n_latents: int, **kwargs + ) -> None: + """ + Instantiates a Perceiver Resampler that operates over a sequence of embeddings (say from a ResNet or ViT or + MAE) of a given dimension, performs `depth` blocks of cross-attention with a fixed `n_latents` inputs, then + returns a Tensor of shape [bsz, n_latents, embed_dim]. :param embed_dim: Dimensionality of embeddings being fed + to the Perceiver Resampler (also dimensionality of latent embeddings *returned* by the Perceiver Resampler. + Could be e.g., VIT embed_dim, ResNet pool dim, and so on. + + Args: + config (`IdeficsConfig`): config object + embed_dim (`int`): The size of each embedding vector + depth (`int`): Depth of the Perceiver Resampler (Transformer w/ cross attention). Should be shallow (< 3). + n_heads (`int`): Number of heads in each Transformer block (for multi-headed self-attention). + head_dim (`int`): Dimensionality of each head projection in the Transformer block. + n_latents (`int`): + Number of latent embeddings to resample ("compress") the input sequence to (usually < 128). + + """ + super().__init__(**kwargs) + self.embed_dim, self.n_heads, self.head_dim, self.n_latents = embed_dim, n_heads, head_dim, n_latents + self.qk_layer_norms = config.perceiver_config.qk_layer_norms_perceiver + + self.intermediate_dim = ( + self.embed_dim * 4 + if not hasattr(config.vision_config, "embed_dim") + else config.vision_config.embed_dim * 4 + ) + # Create Transformer Blocks + self.blocks = [] + for i in range(depth): + self.blocks.append( + [ + TFIdeficsPerceiverAttention( + self.embed_dim, self.n_heads, self.head_dim, self.qk_layer_norms, name=f"blocks.{i}.0" + ), + TFIdeficsMLP(self.intermediate_dim, config, name=f"blocks.{i}.1"), + ] + ) + + self.layer_norm = tf.keras.layers.LayerNormalization(epsilon=1e-5, name="layer_norm") + + def build(self, input_shape): + # Create Latents for Perceiver + self.latents = self.add_weight( + shape=(self.n_latents, self.embed_dim), initializer="random_normal", trainable=True, name="latents" + ) + super().build(input_shape) + + def call(self, context: tf.Tensor) -> tf.Tensor: + """Resample arbitrary length context & *compress* down to self.n_latents latent embeddings""" + # tf.repeat(self.latents, "seq embed -> bsz seq embed", bsz=context.shape[0]) + latents = tf.expand_dims(self.latents, axis=0) + latents = tf.tile(latents, [tf.shape(context)[0], 1, 1]) + # Feed through Perceiver Attention blocks... + for attn, ff in self.blocks: + latents = attn(context, latents) + latents + latents = ff(latents) + latents + return self.layer_norm(latents) + + +class TFIdeficsPerceiverAttention(tf.keras.layers.Layer): + def __init__(self, embed_dim: int, n_heads: int, head_dim: int, qk_layer_norms: bool, **kwargs) -> None: + """Perceiver Cross-Attention Module --> let long-form inputs be `context`, resampled embeddings be `latents`""" + super().__init__(**kwargs) + self.embed_dim, self.n_heads, self.head_dim = embed_dim, n_heads, head_dim + self.qk_layer_norms = qk_layer_norms + # Normalization & Scaling + self.context_layer_norm = tf.keras.layers.LayerNormalization(epsilon=1e-5, name="context_layer_norm") + self.latents_layer_norm = tf.keras.layers.LayerNormalization(epsilon=1e-5, name="latents_layer_norm") + if self.qk_layer_norms: + self.q_layer_norm = tf.keras.layers.LayerNormalization(epsilon=1e-5, name="q_layer_norm") + self.k_layer_norm = tf.keras.layers.LayerNormalization(epsilon=1e-5, name="k_layer_norm") + + self.qk_scale = self.head_dim**-0.5 + + # Q, K, V Projection (no bias -- detail from Perceiver/Flamingo Papers). + self.q_proj = tf.keras.layers.Dense(self.n_heads * self.head_dim, use_bias=False, name="q_proj") + self.k_proj = tf.keras.layers.Dense(self.n_heads * self.head_dim, use_bias=False, name="k_proj") + self.v_proj = tf.keras.layers.Dense(self.n_heads * self.head_dim, use_bias=False, name="v_proj") + + self.output_proj = tf.keras.layers.Dense(embed_dim, use_bias=False, name="output_proj") + + def call(self, context: tf.Tensor, latents: tf.Tensor) -> tf.Tensor: + """ + Runs Perceiver Self-Attention, with special (context, latents) appended along the `seq` dimension! + + Args: + context (`tf.Tensor`): + Tensor of shape `[bsz, seq, embed_dim]` representing long-form context to resample. + latents (`tf.Tensor`): + Tensor of shape `[bsz, n_latents, embed_dim]` representing fixed length latents to compress to. + + Returns: + `tf.Tensor`: Tensor of shape `[bsz, n_latents, embed_dim]` representing attention over latents w/ cross + from context. + """ + context = self.context_layer_norm(context) + latents = self.latents_layer_norm(latents) + batch_size, seq_length, embed_dim = shape_list(context) + + # Query, Key, Value Projections --> Note that in Flamingo, latents are *concatenated* with context prior to attn! + # Note: This results in queries w/ `seq = n_latents`, and keys, values with `seq = len(context) + n_latents` + q = self.q_proj(latents) + k = self.k_proj(tf.concat([context, latents], axis=-2)) + v = self.v_proj(tf.concat([context, latents], axis=-2)) + + # Multiheaded Self-Attention w/ stable softmax (subtract per-row max -- `amax` -- before softmax call) + # =>> `attn` should be a 2D matrix of shape [n_latents x (context + n_latents)] + q, k, v = [ + tf.transpose(tf.reshape(x, (batch_size, x.shape[1], self.n_heads, self.head_dim)), perm=[0, 2, 1, 3]) + for x in (q, k, v) + ] + + if self.qk_layer_norms: + q = self.q_layer_norm(q) + k = self.k_layer_norm(k) + + scores = tf.einsum("... i d, ... j d -> ... i j", q * self.qk_scale, k) + stabilized_scores = scores - tf.reduce_max(scores, axis=-1, keepdims=True) + attn = tf.nn.softmax(stabilized_scores, axis=-1) + + # Attend & project back to output... + resampled = tf.einsum("... i j, ... j d -> ... i d", attn, v) + return self.output_proj( + tf.reshape(tf.transpose(resampled, perm=[0, 2, 1, 3]), (batch_size, -1, self.n_heads * self.head_dim)) + ) + + +class TFIdeficsMLP(tf.keras.layers.Layer): + def __init__(self, intermediate_size, config: IdeficsConfig, **kwargs): + """Simple MLP block with intermediate_size and embedding size""" + super().__init__(**kwargs) + self.embed_dim = config.vision_config.embed_dim + self.ln = tf.keras.layers.LayerNormalization(epsilon=1e-5, name="ln") + self.fc = tf.keras.layers.Dense(intermediate_size, use_bias=False, name="fc") + self.act = tf.keras.layers.ReLU(name="act") + self.c_proj = tf.keras.layers.Dense(self.embed_dim, use_bias=False, name="c_proj") + + def call(self, hidden_states: Optional[Tuple[tf.Tensor]]) -> tf.Tensor: + hidden_states = self.ln(hidden_states) + hidden_states = self.fc(hidden_states) + hidden_states = self.act(hidden_states) + hidden_states = self.c_proj(hidden_states) + + return hidden_states diff --git a/parrot/lib/python3.10/site-packages/transformers/models/idefics/processing_idefics.py b/parrot/lib/python3.10/site-packages/transformers/models/idefics/processing_idefics.py new file mode 100644 index 0000000000000000000000000000000000000000..2afe2a49781245b19fc33534d375b255d06260da --- /dev/null +++ b/parrot/lib/python3.10/site-packages/transformers/models/idefics/processing_idefics.py @@ -0,0 +1,492 @@ +# coding=utf-8 +# Copyright 2022 The HuggingFace Inc. team. +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. +""" +Processor class for IDEFICS. +""" + +from typing import Callable, List, Optional, Union +from urllib.parse import urlparse + +from ...feature_extraction_utils import BatchFeature +from ...processing_utils import ProcessorMixin +from ...tokenization_utils_base import BatchEncoding, PaddingStrategy, TextInput, TruncationStrategy +from ...utils import is_tf_available, is_torch_available + + +if is_torch_available(): + import torch + +if is_tf_available(): + import tensorflow as tf + +IMAGE_TOKEN = "" + + +# copied from m4.training.packing +def incremental_to_binary_attention_mask(incremental_mask, return_tensors, num_classes=-1): + # Set elements >= num_classes to -1 + if num_classes != -1: + if return_tensors == "pt": + incremental_mask[incremental_mask >= num_classes] = -1 + elif return_tensors == "tf": + incremental_mask = tf.where(incremental_mask >= num_classes, -1, incremental_mask) + + # Create mask for negative values + if return_tensors == "pt": + negatives = incremental_mask == -1 + incremental_mask[negatives] = 0 + attn_mask = torch.nn.functional.one_hot(incremental_mask, num_classes=num_classes) + attn_mask[negatives, :] = 0 + elif return_tensors == "tf": + negatives = tf.equal(incremental_mask, -1) + incremental_mask = tf.where(negatives, 0, incremental_mask) + attn_mask = tf.one_hot(incremental_mask, depth=num_classes) + # Reshape 'negatives' to add an extra dimension, making it [batch_size, seq_length, 1] + negatives_expanded = tf.expand_dims(negatives, -1) + attn_mask = tf.where(negatives_expanded, tf.zeros_like(attn_mask), attn_mask) + + return attn_mask + + +# copied from m4.training.packing +def image_attention_mask_for_packed_input_ids(input_ids, tokenizer, return_tensors): + if return_tensors == "pt": + return image_attention_mask_for_packed_input_ids_pt(input_ids, tokenizer) + elif return_tensors == "tf": + return image_attention_mask_for_packed_input_ids_tf(input_ids, tokenizer) + + +def image_attention_mask_for_packed_input_ids_pt(input_ids, tokenizer): + image_attention_mask = torch.full_like(input_ids, fill_value=-1) + next_image_attention_mask = torch.full_like(input_ids, fill_value=-1) + image_token_id = tokenizer.convert_tokens_to_ids(IMAGE_TOKEN) + eod_token_id = tokenizer.eos_token_id + for batch_idx in range(input_ids.size(0)): + count = -1 + seen_eod = False + for idx, token_id in enumerate(input_ids[batch_idx]): + if token_id == image_token_id: + count += 1 + image_attention_mask[batch_idx][idx] = count + seen_eod = False + else: + image_attention_mask[batch_idx][idx] = count + + if seen_eod: + image_attention_mask[batch_idx][idx] = -1 + + if token_id == eod_token_id: + seen_eod = True + + for batch_idx in range(input_ids.size(0)): + count = -1 + seen_eod = False + for idx in range(input_ids[batch_idx].size(0) - 1, -1, -1): + token_id = input_ids[batch_idx][idx] + if token_id == image_token_id: + count += 1 + next_image_attention_mask[batch_idx][idx] = count + seen_eod = False + else: + next_image_attention_mask[batch_idx][idx] = count + + if token_id == eod_token_id: + seen_eod = True + + if seen_eod: + next_image_attention_mask[batch_idx][idx] = -1 + + non_negative_indices = next_image_attention_mask[batch_idx] != -1 + next_image_attention_mask[batch_idx][non_negative_indices] -= count + next_image_attention_mask[batch_idx][non_negative_indices] *= -1 + + return image_attention_mask, next_image_attention_mask + + +def image_attention_mask_for_packed_input_ids_tf(input_ids, tokenizer): + image_token_id = tokenizer.convert_tokens_to_ids(IMAGE_TOKEN) + eod_token_id = tokenizer.eos_token_id + batch_size = tf.shape(input_ids)[0] + image_attention_mask = tf.fill(tf.shape(input_ids), -1) + next_image_attention_mask = tf.fill(tf.shape(input_ids), -1) + + for batch_idx in range(batch_size): + count = -1 + seen_eod = False + seq_length = tf.shape(input_ids)[1] + + for idx in range(seq_length - 1, -1, -1): + token_id = input_ids[batch_idx, idx].numpy() + if token_id == image_token_id: + count += 1 + indices = [[batch_idx, idx]] + updates = [count] + image_attention_mask = tf.tensor_scatter_nd_update(image_attention_mask, indices, updates) + next_image_attention_mask = tf.tensor_scatter_nd_update(next_image_attention_mask, indices, updates) + elif token_id == eod_token_id and not seen_eod: + seen_eod = True + count = 0 + indices = [[batch_idx, idx]] + updates = [count] + next_image_attention_mask = tf.tensor_scatter_nd_update(next_image_attention_mask, indices, updates) + if seen_eod and token_id != eod_token_id: + indices = [[batch_idx, idx]] + updates = [-1] + next_image_attention_mask = tf.tensor_scatter_nd_update(next_image_attention_mask, indices, updates) + return image_attention_mask, next_image_attention_mask + + +def is_url(string): + """Checks if the passed string contains a valid url and nothing else. e.g. if space is included it's immediately + invalidated the url""" + if " " in string: + return False + result = urlparse(string) + return all([result.scheme, result.netloc]) + + +class IdeficsProcessor(ProcessorMixin): + r""" + Constructs a IDEFICS processor which wraps a LLama tokenizer and IDEFICS image processor into a single processor. + + [`IdeficsProcessor`] offers all the functionalities of [`IdeficsImageProcessor`] and [`LlamaTokenizerFast`]. See + the docstring of [`~IdeficsProcessor.__call__`] and [`~IdeficsProcessor.decode`] for more information. + + Args: + image_processor (`IdeficsImageProcessor`): + An instance of [`IdeficsImageProcessor`]. The image processor is a required input. + tokenizer (`LlamaTokenizerFast`): + An instance of [`LlamaTokenizerFast`]. The tokenizer is a required input. + image_size (`int`, *optional*, defaults to 224): Image size (assuming a square image) + """ + + attributes = ["image_processor", "tokenizer"] + image_processor_class = "IdeficsImageProcessor" + tokenizer_class = "LlamaTokenizerFast" + + def __init__(self, image_processor, tokenizer=None, image_size=224, add_end_of_utterance_token=None, **kwargs): + if image_processor is None: + raise ValueError("You need to specify an `image_processor`.") + if tokenizer is None: + raise ValueError("You need to specify a `tokenizer`.") + + super().__init__(image_processor, tokenizer) + self.current_processor = self.image_processor + self.image_token_id = tokenizer.convert_tokens_to_ids(IMAGE_TOKEN) + + self.default_image_dims = ( + self.image_processor.image_num_channels, + self.image_processor.image_size, + self.image_processor.image_size, + ) + + self.tokenizer_was_trained_with_end_of_utterance_token = ( + True + if "" in self.tokenizer.special_tokens_map.get("additional_special_tokens", []) + else False + ) + + def __call__( + self, + prompts: Union[List[TextInput], List[List[TextInput]]], + padding: Union[bool, str, PaddingStrategy] = "longest", + truncation: Union[bool, str, TruncationStrategy] = None, + max_length: Optional[int] = None, + transform: Callable = None, + add_eos_token=False, + add_end_of_utterance_token=None, + debug=False, + return_tensors="pt", + ) -> BatchEncoding: + """This method takes batched or non-batched prompts made of text and images and converts them into prompts that + the model was trained on and prepares the image pixel values for the model to process. + + Args: + prompts (`Union[List[TextInput], [List[List[TextInput]]]]`): + either a single prompt or a batched list of prompts - see the detailed description immediately after + the end of the arguments doc section. + padding (`bool`, `str` or [`~utils.PaddingStrategy`], *optional*, defaults to `"longest"`): + Select a strategy to pad the returned sequences (according to the model's padding side and padding + index) among: + - `True` or `'longest'` (default): Pad to the longest sequence in the batch (or no padding if only a single + sequence if provided). + - `'max_length'`: Pad to a maximum length specified with the argument `max_length` or to the maximum + acceptable input length for the model if that argument is not provided. + - `False` or `'do_not_pad'`: No padding. This will raise an error if the input sequences are of different + lengths. + Note: Unlike most processors, which set padding=`False` by default, `IdeficsProcessor` sets `padding="longest"` + by default. See https://github.com/huggingface/transformers/pull/29449#pullrequestreview-1925576061 for why. + max_length (`int`, *optional*): + Maximum length of the returned list and optionally padding length (see above). + truncation (`bool`, *optional*): + Activates truncation to cut input sequences longer than `max_length` to `max_length`. + transform (`Callable`, *optional*): + A custom transform function that accepts a single image can be passed for training. For example, + `torchvision.Compose` can be used to compose multiple functions. If `None` a preset inference-specific + set of transforms will be applied to the images + add_eos_token (`bool`, *optional*, defaults to `False`): + Adds `eos_token` at the end of the final prompt if True` + add_end_of_utterance_token (`bool`, *optional*) + Whether to automatically add `` after each prompt's text input (unless followed by an + image). If `None` the tokenizer will be checked instead and if this token is found in + `additional_special_tokens` then the value will be `True`. + debug (`bool`, *optional*, defaults to `False`): + `True` value will help debug prompt generation by dumping useful information + return_tensors (`str` or `TensorType`, *optional*, defaults to `TensorType.PYTORCH`): + The type of tensors to return. Can be one of: + - `TensorType.PYTORCH` or `'pt'`: Return a batch of type `torch.Tensor`. + + Returns: + a dict with entries: `input_ids`, `attention_mask`, `pixel_values`, `image_attention_mask` which can be + directly passed to `model.generate` + + Detailed explanation: + + Each entry in `prompts` is either a text to be passed as is or an image that will be processed. + + An image can be either an image object (`PIL.Image`) or a url from which the image can be retrieved. + + When the processor encounters an image it'll inject `` + entry into the prompt. + + Example: + + ```python + checkpoint = "HuggingFaceM4/idefics-9b" + processor = AutoProcessor.from_pretrained(checkpoint) + url = "https://hips.hearstapps.com/hmg-prod/images/cute-photos-of-cats-in-grass-1593184777.jpg" + img = processor.image_processor.fetch_images([url])[0] + + prompts = [ + "User:", + img, + "Describe this image.\nAssistant: An image of two kittens in grass.\n", + "User:", + "https://hips.hearstapps.com/hmg-prod/images/dog-puns-1581708208.jpg", + "Describe this image.\nAssistant:", + ] + + inputs = processor(prompts, return_tensors="pt") + generated_ids = model.generate(**inputs, max_length=100) + generated_text = processor.batch_decode(generated_ids, skip_special_tokens=True)[0] + ``` + + In this example the `prompts` will be converted into: + + ``` + User:Describe this image. + Assistant: An image of two kittens in grass. + User:Describe this image. + Assistant:' + ``` + + and the two images will be massaged using [`IdeficsImageProcessor.__call__`] method and placed inside the + `pixel_values` dict entry of the return value. + + This example also examplifies that images can be passed as objects or as text urls. It can be seen that the + first image is passed as object and the second one as a url. + + To do training do: + + ```python + image_transform = transforms.Compose( + [ + transforms.RandomResizedCrop( + (w, h), scale=(0.9, 1.0), interpolation=transforms.InterpolationMode.BICUBIC + ), + transforms.ToTensor(), + transforms.Normalize(mean=self.image_mean, std=self.image_std), + ] + ) + inputs = processor(prompts, transform=image_transform, return_tensors="pt") + ``` + + In order to help debug prompt generation enable `debug=True` which will show you what's happening. + + """ + + # if the value isn't overriden by the user, check if the tokenizer was trained with this token and then use it + if add_end_of_utterance_token is None: + add_end_of_utterance_token = self.tokenizer_was_trained_with_end_of_utterance_token + # turn non-batched prompts into batched + if not any(isinstance(i, list) for i in prompts): + prompts = [prompts] + + fake_token = "" + image_token = "" + end_of_utterance_token = "" + + def image_tokens(last_was_image): + if last_was_image: + return image_token + fake_token + else: + return fake_token + image_token + fake_token + + all_prompts = [] + all_images = [] + for sample in prompts: + # the model was trained on samples starting with + full_text = f"{self.tokenizer.bos_token}" + + # an image can either be an image object in the item or the url, everything else is a verbatim prompt text + image_objects = [] + last_was_image = False + last_was_text = False + for i, item in enumerate(sample): + if i > 0: + last_was_text = True if not last_was_image else False + + if isinstance(item, str): + item = item.strip(" ") + if is_url(item): + image = self.image_processor.fetch_images(item) + full_text += image_tokens(last_was_image) + image_objects.append(image) + last_was_image = True + else: + # we add end_of_utterance_token between each subsequent text prompts (but not at the last one!) + if add_end_of_utterance_token and last_was_text: + full_text += end_of_utterance_token + full_text += item + last_was_image = False + else: + # must be an image obj + full_text += image_tokens(last_was_image) + image_objects.append(item) + last_was_image = True + + if add_eos_token: + full_text += self.tokenizer.eos_token + + if debug is True: + print(f"{full_text=}") + + image_objects = self.image_processor(image_objects, transform=transform, return_tensors=return_tensors) + + all_prompts.append(full_text) + all_images.append(image_objects) + + text_encoding = self.tokenizer( + text=all_prompts, + add_special_tokens=False, + padding=padding, + truncation=truncation, + max_length=max_length, + ) + all_texts = text_encoding["input_ids"] + all_attention_masks = text_encoding["attention_mask"] + + # max_num_images has to be at least 1 even when there are no images + max_num_images = max(len(x) for x in all_images) + max_num_images = max(1, max_num_images) + + at_least_one_image = sum(len(x) for x in all_images) > 0 + output_input_ids = [] + output_images = [] + output_attention_masks = [] + + for text, attention_mask, images in zip(all_texts, all_attention_masks, all_images): + padded_input_ids = text + image_count = padded_input_ids.count(self.image_token_id) + local_max_num_images = min(image_count, max_num_images) + + current_images = images[:local_max_num_images] + + if len(current_images) > 0: + if return_tensors == "pt": + padded_image_tensor = torch.zeros(max_num_images, *current_images.size()[1:]) + padded_image_tensor[: current_images.size(0)] = current_images + elif return_tensors == "tf": + # Assuming current_images is a TensorFlow tensor + # Get the shape of current_images, excluding the first dimension + image_shape = tf.shape(current_images)[1:] + # Create a shape for the padded_image_tensor + padded_shape = tf.concat([[max_num_images], image_shape], axis=0) + # Create the padded_image_tensor of zeros + padded_image_tensor = tf.zeros(padded_shape, dtype=current_images.dtype) + # Get the number of images (assuming current_images has shape [num_images, height, width, channels]) + num_images = tf.shape(current_images)[0] + # Update the padded_image_tensor with the values from current_images + indices = tf.reshape(tf.range(num_images), (-1, 1)) + updates = current_images + padded_image_tensor = tf.tensor_scatter_nd_update(padded_image_tensor, indices, updates) + else: + if return_tensors == "pt": + padded_image_tensor = torch.zeros(max_num_images, *self.default_image_dims) + elif return_tensors == "tf": + padded_image_tensor = tf.zeros((max_num_images, *self.default_image_dims)) + + output_images.append(padded_image_tensor) + if return_tensors == "pt": + output_input_ids.append(torch.tensor(padded_input_ids)) + output_attention_masks.append(torch.tensor(attention_mask)) + elif return_tensors == "tf": + output_input_ids.append(tf.convert_to_tensor(padded_input_ids, dtype=tf.int32)) + output_attention_masks.append(attention_mask) + + if return_tensors == "pt": + output_input_ids = torch.stack(output_input_ids) + output_images = torch.stack(output_images) + output_attention_masks = torch.stack(output_attention_masks) + elif return_tensors == "tf": + output_input_ids = tf.stack(output_input_ids) + output_images = tf.stack(output_images) + output_attention_masks = tf.stack(output_attention_masks) + + if at_least_one_image: + image_attention_mask, _ = image_attention_mask_for_packed_input_ids( + output_input_ids, self.tokenizer, return_tensors + ) + image_attention_mask = incremental_to_binary_attention_mask( + image_attention_mask, return_tensors, num_classes=max_num_images + ) + else: + # in full language mode we set the image mask to all-0s + if return_tensors == "pt": + image_attention_mask = torch.zeros( + output_input_ids.shape[0], output_input_ids.shape[1], 1, dtype=torch.bool + ) + elif return_tensors == "tf": + image_attention_mask = tf.zeros( + (output_input_ids.shape[0], output_input_ids.shape[1], 1), dtype=tf.bool + ) + return BatchFeature( + data={ + "input_ids": output_input_ids, + "attention_mask": output_attention_masks, + "pixel_values": output_images, + "image_attention_mask": image_attention_mask, + } + ) + + def batch_decode(self, *args, **kwargs): + """ + This method forwards all its arguments to LlamaTokenizerFast's [`~PreTrainedTokenizer.batch_decode`]. Please + refer to the docstring of this method for more information. + """ + return self.tokenizer.batch_decode(*args, **kwargs) + + def decode(self, *args, **kwargs): + """ + This method forwards all its arguments to LlamaTokenizerFast's [`~PreTrainedTokenizer.decode`]. Please refer to + the docstring of this method for more information. + """ + return self.tokenizer.decode(*args, **kwargs) + + @property + def model_input_names(self): + tokenizer_input_names = self.tokenizer.model_input_names + image_processor_input_names = self.image_processor.model_input_names + return list(dict.fromkeys(tokenizer_input_names + image_processor_input_names)) diff --git a/parrot/lib/python3.10/site-packages/transformers/models/idefics/vision.py b/parrot/lib/python3.10/site-packages/transformers/models/idefics/vision.py new file mode 100644 index 0000000000000000000000000000000000000000..d90f837b3c77baed36b1e23175939b264c155d0f --- /dev/null +++ b/parrot/lib/python3.10/site-packages/transformers/models/idefics/vision.py @@ -0,0 +1,490 @@ +# coding=utf-8 +# Copyright 2021 The OpenAI Team Authors and The HuggingFace Team. All rights reserved. +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. +""" PyTorch IdeficsVision model: a copy of CLIPVisionModel using a simpler config object""" + + +import math +from dataclasses import dataclass +from typing import Optional, Tuple, Union + +import torch +import torch.utils.checkpoint +from torch import nn + +from ...activations import ACT2FN +from ...modeling_outputs import BaseModelOutput, BaseModelOutputWithPooling +from ...utils import ModelOutput, logging +from .configuration_idefics import IdeficsVisionConfig + + +logger = logging.get_logger(__name__) + + +@dataclass +class IdeficsVisionModelOutput(ModelOutput): + """ + Base class for vision model's outputs that also contains image embeddings of the pooling of the last hidden states. + + Args: + image_embeds (`torch.FloatTensor` of shape `(batch_size, output_dim)` *optional* returned when model is initialized with `with_projection=True`): + The image embeddings obtained by applying the projection layer to the pooler_output. + last_hidden_state (`torch.FloatTensor` of shape `(batch_size, sequence_length, hidden_size)`): + Sequence of hidden-states at the output of the last layer of the model. + hidden_states (`tuple(torch.FloatTensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`): + Tuple of `torch.FloatTensor` (one for the output of the embeddings, if the model has an embedding layer, + + one for the output of each layer) of shape `(batch_size, sequence_length, hidden_size)`. + + Hidden-states of the model at the output of each layer plus the optional initial embedding outputs. + attentions (`tuple(torch.FloatTensor)`, *optional*, returned when `output_attentions=True` is passed or when `config.output_attentions=True`): + Tuple of `torch.FloatTensor` (one for each layer) of shape `(batch_size, num_heads, sequence_length, + sequence_length)`. + + Attentions weights after the attention softmax, used to compute the weighted average in the self-attention + heads. + """ + + image_embeds: Optional[torch.FloatTensor] = None + last_hidden_state: torch.FloatTensor = None + hidden_states: Optional[Tuple[torch.FloatTensor, ...]] = None + attentions: Optional[Tuple[torch.FloatTensor, ...]] = None + + +# Adapted from transformers.models.clip.modeling_clip.CLIPVisionEmbeddings +class IdeficsVisionEmbeddings(nn.Module): + def __init__(self, config: IdeficsVisionConfig): + super().__init__() + self.config = config + self.embed_dim = config.hidden_size + self.image_size = config.image_size + self.patch_size = config.patch_size + + self.class_embedding = nn.Parameter(torch.randn(self.embed_dim)) + + self.patch_embedding = nn.Conv2d( + in_channels=config.num_channels, + out_channels=self.embed_dim, + kernel_size=self.patch_size, + stride=self.patch_size, + bias=False, + ) + + self.num_patches = (self.image_size // self.patch_size) ** 2 + self.num_positions = self.num_patches + 1 + self.position_embedding = nn.Embedding(self.num_positions, self.embed_dim) + self.register_buffer("position_ids", torch.arange(self.num_positions).expand((1, -1)), persistent=False) + + # Heavily inspired from https://github.com/huggingface/transformers/blob/v4.33.0/src/transformers/models/vit/modeling_vit.py#L82 + def interpolate_pos_encoding(self, embeddings: torch.Tensor, height: int, width: int) -> torch.Tensor: + """ + This method allows to interpolate the pre-trained position encodings, to be able to use the model on higher + resolution images. + + Source: + https://github.com/facebookresearch/dino/blob/de9ee3df6cf39fac952ab558447af1fa1365362a/vision_transformer.py#L174 + """ + + num_patches = embeddings.shape[1] - 1 + pos_embed = self.position_embedding(self.position_ids) + num_positions = pos_embed.shape[1] - 1 + if num_patches == num_positions and height == width: + return pos_embed + class_pos_embed = pos_embed[:, 0] + patch_pos_embed = pos_embed[:, 1:] + + embed_dim = embeddings.shape[-1] + num_h_patches = height // self.config.patch_size + num_w_patches = width // self.config.patch_size + # we add a small number to avoid floating point error in the interpolation + # see discussion at https://github.com/facebookresearch/dino/issues/8 + num_h_patches, num_w_patches = num_h_patches + 0.1, num_w_patches + 0.1 + sqrt_num_positions = math.sqrt(num_positions) + patch_pos_embed = patch_pos_embed.reshape(1, int(sqrt_num_positions), int(sqrt_num_positions), embed_dim) + patch_pos_embed = patch_pos_embed.permute(0, 3, 1, 2) + fp32_upcasting = patch_pos_embed.dtype == torch.bfloat16 + if fp32_upcasting: + logger.warning_once( + "Upcasting patch_pos_embed to fp32 for interpolation since `upsample_bicubic2d_out_frame` in nn.functional.interpolate " + "is not implemented for 'torch.bfloat16' dtype. This will result in a slight overhead." + ) + patch_pos_embed = patch_pos_embed.to(torch.float) + patch_pos_embed = nn.functional.interpolate( + patch_pos_embed, + scale_factor=(num_h_patches / sqrt_num_positions, num_w_patches / sqrt_num_positions), + mode="bicubic", + align_corners=False, + ) + if fp32_upcasting: + patch_pos_embed = patch_pos_embed.to(torch.bfloat16) + if int(num_h_patches) != patch_pos_embed.shape[-2] or int(num_w_patches) != patch_pos_embed.shape[-1]: + raise ValueError( + f"Number of patches for images ({int(num_h_patches), int(num_w_patches)}) don't match the " + f"shape of position embedding ({patch_pos_embed.shape[-2], patch_pos_embed.shape[-1]})" + ) + patch_pos_embed = patch_pos_embed.permute(0, 2, 3, 1).view(1, -1, embed_dim) + return torch.cat((class_pos_embed.unsqueeze(0), patch_pos_embed), dim=1) + + def forward(self, pixel_values: torch.FloatTensor, interpolate_pos_encoding: bool = False) -> torch.Tensor: + batch_size, num_channels, height, width = pixel_values.shape + if not interpolate_pos_encoding: + if height != self.image_size or width != self.image_size: + raise ValueError( + f"Input image size ({height}*{width}) doesn't match model" + f" ({self.image_size}*{self.image_size}). You should try to set `interpolate_pos_encoding=True`" + ) + + target_dtype = self.patch_embedding.weight.dtype + patch_embeds = self.patch_embedding(pixel_values.to(dtype=target_dtype)) # shape = [*, width, grid, grid] + + patch_embeds = patch_embeds.flatten(2).transpose(1, 2) + + class_embeds = self.class_embedding.expand(batch_size, 1, -1) + embeddings = torch.cat([class_embeds, patch_embeds], dim=1) + + # add positional encoding to each token + if interpolate_pos_encoding: + embeddings = embeddings + self.interpolate_pos_encoding(embeddings, height, width) + else: + embeddings = embeddings + self.position_embedding(self.position_ids) + + return embeddings + + +# Copied from transformers.models.clip.modeling_clip.CLIPAttention with CLIP->IdeficsVision +class IdeficsVisionAttention(nn.Module): + """Multi-headed attention from 'Attention Is All You Need' paper""" + + def __init__(self, config): + super().__init__() + self.config = config + self.embed_dim = config.hidden_size + self.num_heads = config.num_attention_heads + self.head_dim = self.embed_dim // self.num_heads + if self.head_dim * self.num_heads != self.embed_dim: + raise ValueError( + f"embed_dim must be divisible by num_heads (got `embed_dim`: {self.embed_dim} and `num_heads`:" + f" {self.num_heads})." + ) + self.scale = self.head_dim**-0.5 + self.dropout = config.attention_dropout + + self.k_proj = nn.Linear(self.embed_dim, self.embed_dim) + self.v_proj = nn.Linear(self.embed_dim, self.embed_dim) + self.q_proj = nn.Linear(self.embed_dim, self.embed_dim) + self.out_proj = nn.Linear(self.embed_dim, self.embed_dim) + + def _shape(self, tensor: torch.Tensor, seq_len: int, bsz: int): + return tensor.view(bsz, seq_len, self.num_heads, self.head_dim).transpose(1, 2).contiguous() + + def forward( + self, + hidden_states: torch.Tensor, + attention_mask: Optional[torch.Tensor] = None, + causal_attention_mask: Optional[torch.Tensor] = None, + output_attentions: Optional[bool] = False, + ) -> Tuple[torch.Tensor, Optional[torch.Tensor], Optional[Tuple[torch.Tensor]]]: + """Input shape: Batch x Time x Channel""" + + bsz, tgt_len, embed_dim = hidden_states.size() + + # get query proj + query_states = self.q_proj(hidden_states) * self.scale + key_states = self._shape(self.k_proj(hidden_states), -1, bsz) + value_states = self._shape(self.v_proj(hidden_states), -1, bsz) + + proj_shape = (bsz * self.num_heads, -1, self.head_dim) + query_states = self._shape(query_states, tgt_len, bsz).view(*proj_shape) + key_states = key_states.view(*proj_shape) + value_states = value_states.view(*proj_shape) + + src_len = key_states.size(1) + attn_weights = torch.bmm(query_states, key_states.transpose(1, 2)) + + if attn_weights.size() != (bsz * self.num_heads, tgt_len, src_len): + raise ValueError( + f"Attention weights should be of size {(bsz * self.num_heads, tgt_len, src_len)}, but is" + f" {attn_weights.size()}" + ) + + # apply the causal_attention_mask first + if causal_attention_mask is not None: + if causal_attention_mask.size() != (bsz, 1, tgt_len, src_len): + raise ValueError( + f"Attention mask should be of size {(bsz, 1, tgt_len, src_len)}, but is" + f" {causal_attention_mask.size()}" + ) + attn_weights = attn_weights.view(bsz, self.num_heads, tgt_len, src_len) + causal_attention_mask + attn_weights = attn_weights.view(bsz * self.num_heads, tgt_len, src_len) + + if attention_mask is not None: + if attention_mask.size() != (bsz, 1, tgt_len, src_len): + raise ValueError( + f"Attention mask should be of size {(bsz, 1, tgt_len, src_len)}, but is {attention_mask.size()}" + ) + attn_weights = attn_weights.view(bsz, self.num_heads, tgt_len, src_len) + attention_mask + attn_weights = attn_weights.view(bsz * self.num_heads, tgt_len, src_len) + + attn_weights = nn.functional.softmax(attn_weights, dim=-1) + + if output_attentions: + # this operation is a bit akward, but it's required to + # make sure that attn_weights keeps its gradient. + # In order to do so, attn_weights have to reshaped + # twice and have to be reused in the following + attn_weights_reshaped = attn_weights.view(bsz, self.num_heads, tgt_len, src_len) + attn_weights = attn_weights_reshaped.view(bsz * self.num_heads, tgt_len, src_len) + else: + attn_weights_reshaped = None + + attn_probs = nn.functional.dropout(attn_weights, p=self.dropout, training=self.training) + + attn_output = torch.bmm(attn_probs, value_states) + + if attn_output.size() != (bsz * self.num_heads, tgt_len, self.head_dim): + raise ValueError( + f"`attn_output` should be of size {(bsz, self.num_heads, tgt_len, self.head_dim)}, but is" + f" {attn_output.size()}" + ) + + attn_output = attn_output.view(bsz, self.num_heads, tgt_len, self.head_dim) + attn_output = attn_output.transpose(1, 2) + attn_output = attn_output.reshape(bsz, tgt_len, embed_dim) + + attn_output = self.out_proj(attn_output) + + return attn_output, attn_weights_reshaped + + +# Copied from transformers.models.clip.modeling_clip.CLIPMLP with CLIP->IdeficsVision +class IdeficsVisionMLP(nn.Module): + def __init__(self, config): + super().__init__() + self.config = config + self.activation_fn = ACT2FN[config.hidden_act] + self.fc1 = nn.Linear(config.hidden_size, config.intermediate_size) + self.fc2 = nn.Linear(config.intermediate_size, config.hidden_size) + + def forward(self, hidden_states: torch.Tensor) -> torch.Tensor: + hidden_states = self.fc1(hidden_states) + hidden_states = self.activation_fn(hidden_states) + hidden_states = self.fc2(hidden_states) + return hidden_states + + +# Copied from transformers.models.clip.modeling_clip.CLIPEncoderLayer with CLIP->IdeficsVision +class IdeficsVisionEncoderLayer(nn.Module): + def __init__(self, config: IdeficsVisionConfig): + super().__init__() + self.embed_dim = config.hidden_size + self.self_attn = IdeficsVisionAttention(config) + self.layer_norm1 = nn.LayerNorm(self.embed_dim, eps=config.layer_norm_eps) + self.mlp = IdeficsVisionMLP(config) + self.layer_norm2 = nn.LayerNorm(self.embed_dim, eps=config.layer_norm_eps) + + def forward( + self, + hidden_states: torch.Tensor, + attention_mask: torch.Tensor, + causal_attention_mask: torch.Tensor, + output_attentions: Optional[bool] = False, + ) -> Tuple[torch.FloatTensor]: + """ + Args: + hidden_states (`torch.FloatTensor`): input to the layer of shape `(batch, seq_len, embed_dim)` + attention_mask (`torch.FloatTensor`): attention mask of size + `(batch, 1, tgt_len, src_len)` where padding elements are indicated by very large negative values. + `(config.encoder_attention_heads,)`. + output_attentions (`bool`, *optional*): + Whether or not to return the attentions tensors of all attention layers. See `attentions` under + returned tensors for more detail. + """ + residual = hidden_states + + hidden_states = self.layer_norm1(hidden_states) + hidden_states, attn_weights = self.self_attn( + hidden_states=hidden_states, + attention_mask=attention_mask, + causal_attention_mask=causal_attention_mask, + output_attentions=output_attentions, + ) + hidden_states = residual + hidden_states + + residual = hidden_states + hidden_states = self.layer_norm2(hidden_states) + hidden_states = self.mlp(hidden_states) + hidden_states = residual + hidden_states + + outputs = (hidden_states,) + + if output_attentions: + outputs += (attn_weights,) + + return outputs + + +# Copied from transformers.models.clip.modeling_clip.CLIPEncoder with CLIP->IdeficsVision +class IdeficsVisionEncoder(nn.Module): + """ + Transformer encoder consisting of `config.num_hidden_layers` self attention layers. Each layer is a + [`IdeficsVisionEncoderLayer`]. + + Args: + config: IdeficsVisionConfig + """ + + def __init__(self, config: IdeficsVisionConfig): + super().__init__() + self.config = config + self.layers = nn.ModuleList([IdeficsVisionEncoderLayer(config) for _ in range(config.num_hidden_layers)]) + self.gradient_checkpointing = False + + def forward( + self, + inputs_embeds, + attention_mask: Optional[torch.Tensor] = None, + causal_attention_mask: Optional[torch.Tensor] = None, + output_attentions: Optional[bool] = None, + output_hidden_states: Optional[bool] = None, + return_dict: Optional[bool] = None, + ) -> Union[Tuple, BaseModelOutput]: + r""" + Args: + inputs_embeds (`torch.FloatTensor` of shape `(batch_size, sequence_length, hidden_size)`): + Optionally, instead of passing `input_ids` you can choose to directly pass an embedded representation. + This is useful if you want more control over how to convert `input_ids` indices into associated vectors + than the model's internal embedding lookup matrix. + attention_mask (`torch.Tensor` of shape `(batch_size, sequence_length)`, *optional*): + Mask to avoid performing attention on padding token indices. Mask values selected in `[0, 1]`: + + - 1 for tokens that are **not masked**, + - 0 for tokens that are **masked**. + + [What are attention masks?](../glossary#attention-mask) + causal_attention_mask (`torch.Tensor` of shape `(batch_size, sequence_length)`, *optional*): + Causal mask for the text model. Mask values selected in `[0, 1]`: + + - 1 for tokens that are **not masked**, + - 0 for tokens that are **masked**. + + [What are attention masks?](../glossary#attention-mask) + output_attentions (`bool`, *optional*): + Whether or not to return the attentions tensors of all attention layers. See `attentions` under + returned tensors for more detail. + output_hidden_states (`bool`, *optional*): + Whether or not to return the hidden states of all layers. See `hidden_states` under returned tensors + for more detail. + return_dict (`bool`, *optional*): + Whether or not to return a [`~utils.ModelOutput`] instead of a plain tuple. + """ + output_attentions = output_attentions if output_attentions is not None else self.config.output_attentions + output_hidden_states = ( + output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states + ) + return_dict = return_dict if return_dict is not None else self.config.use_return_dict + + encoder_states = () if output_hidden_states else None + all_attentions = () if output_attentions else None + + hidden_states = inputs_embeds + for idx, encoder_layer in enumerate(self.layers): + if output_hidden_states: + encoder_states = encoder_states + (hidden_states,) + if self.gradient_checkpointing and self.training: + layer_outputs = self._gradient_checkpointing_func( + encoder_layer.__call__, + hidden_states, + attention_mask, + causal_attention_mask, + output_attentions, + ) + else: + layer_outputs = encoder_layer( + hidden_states, + attention_mask, + causal_attention_mask, + output_attentions=output_attentions, + ) + + hidden_states = layer_outputs[0] + + if output_attentions: + all_attentions = all_attentions + (layer_outputs[1],) + + if output_hidden_states: + encoder_states = encoder_states + (hidden_states,) + + if not return_dict: + return tuple(v for v in [hidden_states, encoder_states, all_attentions] if v is not None) + return BaseModelOutput( + last_hidden_state=hidden_states, hidden_states=encoder_states, attentions=all_attentions + ) + + +# Adapted from transformers.models.clip.modeling_clip.CLIPVisionTransformer +class IdeficsVisionTransformer(nn.Module): + def __init__(self, config: IdeficsVisionConfig): + super().__init__() + self.config = config + embed_dim = config.hidden_size + + self.embeddings = IdeficsVisionEmbeddings(config) + self.pre_layrnorm = nn.LayerNorm(embed_dim, eps=config.layer_norm_eps) + self.encoder = IdeficsVisionEncoder(config) + self.post_layernorm = nn.LayerNorm(embed_dim, eps=config.layer_norm_eps) + + # Adapted from transformers.models.clip.modeling_clip.CLIPVisionTransformer.forward + def forward( + self, + pixel_values: Optional[torch.FloatTensor] = None, + output_attentions: Optional[bool] = None, + output_hidden_states: Optional[bool] = None, + interpolate_pos_encoding: Optional[bool] = False, + return_dict: Optional[bool] = None, + ) -> Union[Tuple, BaseModelOutputWithPooling]: + r""" + Returns: + + """ + output_attentions = output_attentions if output_attentions is not None else self.config.output_attentions + output_hidden_states = ( + output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states + ) + return_dict = return_dict if return_dict is not None else self.config.use_return_dict + + if pixel_values is None: + raise ValueError("You have to specify pixel_values") + + hidden_states = self.embeddings(pixel_values, interpolate_pos_encoding=interpolate_pos_encoding) + hidden_states = self.pre_layrnorm(hidden_states) + + encoder_outputs = self.encoder( + inputs_embeds=hidden_states, + output_attentions=output_attentions, + output_hidden_states=output_hidden_states, + return_dict=return_dict, + ) + + last_hidden_state = encoder_outputs[0] + pooled_output = last_hidden_state[:, 0, :] + pooled_output = self.post_layernorm(pooled_output) + + if not return_dict: + return (last_hidden_state, pooled_output) + encoder_outputs[1:] + + return BaseModelOutputWithPooling( + last_hidden_state=last_hidden_state, + pooler_output=pooled_output, + hidden_states=encoder_outputs.hidden_states, + attentions=encoder_outputs.attentions, + ) diff --git a/parrot/lib/python3.10/site-packages/transformers/models/tapas/__init__.py b/parrot/lib/python3.10/site-packages/transformers/models/tapas/__init__.py new file mode 100644 index 0000000000000000000000000000000000000000..750bf7e00f5a8f252ad4aba5e714f6d90423b89a --- /dev/null +++ b/parrot/lib/python3.10/site-packages/transformers/models/tapas/__init__.py @@ -0,0 +1,91 @@ +# Copyright 2020 The HuggingFace Team. All rights reserved. +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. + +from typing import TYPE_CHECKING + +from ...utils import OptionalDependencyNotAvailable, _LazyModule, is_tf_available, is_torch_available + + +_import_structure = { + "configuration_tapas": ["TapasConfig"], + "tokenization_tapas": ["TapasTokenizer"], +} + +try: + if not is_torch_available(): + raise OptionalDependencyNotAvailable() +except OptionalDependencyNotAvailable: + pass +else: + _import_structure["modeling_tapas"] = [ + "TapasForMaskedLM", + "TapasForQuestionAnswering", + "TapasForSequenceClassification", + "TapasModel", + "TapasPreTrainedModel", + "load_tf_weights_in_tapas", + ] +try: + if not is_tf_available(): + raise OptionalDependencyNotAvailable() +except OptionalDependencyNotAvailable: + pass +else: + _import_structure["modeling_tf_tapas"] = [ + "TFTapasForMaskedLM", + "TFTapasForQuestionAnswering", + "TFTapasForSequenceClassification", + "TFTapasModel", + "TFTapasPreTrainedModel", + ] + + +if TYPE_CHECKING: + from .configuration_tapas import TapasConfig + from .tokenization_tapas import TapasTokenizer + + try: + if not is_torch_available(): + raise OptionalDependencyNotAvailable() + except OptionalDependencyNotAvailable: + pass + else: + from .modeling_tapas import ( + TapasForMaskedLM, + TapasForQuestionAnswering, + TapasForSequenceClassification, + TapasModel, + TapasPreTrainedModel, + load_tf_weights_in_tapas, + ) + + try: + if not is_tf_available(): + raise OptionalDependencyNotAvailable() + except OptionalDependencyNotAvailable: + pass + else: + from .modeling_tf_tapas import ( + TFTapasForMaskedLM, + TFTapasForQuestionAnswering, + TFTapasForSequenceClassification, + TFTapasModel, + TFTapasPreTrainedModel, + ) + + +else: + import sys + + sys.modules[__name__] = _LazyModule(__name__, globals()["__file__"], _import_structure, module_spec=__spec__) diff --git a/parrot/lib/python3.10/site-packages/transformers/models/tapas/__pycache__/__init__.cpython-310.pyc b/parrot/lib/python3.10/site-packages/transformers/models/tapas/__pycache__/__init__.cpython-310.pyc new file mode 100644 index 0000000000000000000000000000000000000000..c58bc0c54dd6545050d16c18baa38fdc677ce725 Binary files /dev/null and b/parrot/lib/python3.10/site-packages/transformers/models/tapas/__pycache__/__init__.cpython-310.pyc differ diff --git a/parrot/lib/python3.10/site-packages/transformers/models/tapas/__pycache__/configuration_tapas.cpython-310.pyc b/parrot/lib/python3.10/site-packages/transformers/models/tapas/__pycache__/configuration_tapas.cpython-310.pyc new file mode 100644 index 0000000000000000000000000000000000000000..8c7714ffd5b72b1f59a6d53ae59614f29b7494f6 Binary files /dev/null and b/parrot/lib/python3.10/site-packages/transformers/models/tapas/__pycache__/configuration_tapas.cpython-310.pyc differ diff --git a/parrot/lib/python3.10/site-packages/transformers/models/tapas/__pycache__/convert_tapas_original_tf_checkpoint_to_pytorch.cpython-310.pyc b/parrot/lib/python3.10/site-packages/transformers/models/tapas/__pycache__/convert_tapas_original_tf_checkpoint_to_pytorch.cpython-310.pyc new file mode 100644 index 0000000000000000000000000000000000000000..79c4c385c2fa904eef1edf481a8f0891891cbe4b Binary files /dev/null and b/parrot/lib/python3.10/site-packages/transformers/models/tapas/__pycache__/convert_tapas_original_tf_checkpoint_to_pytorch.cpython-310.pyc differ diff --git a/parrot/lib/python3.10/site-packages/transformers/models/tapas/__pycache__/modeling_tapas.cpython-310.pyc b/parrot/lib/python3.10/site-packages/transformers/models/tapas/__pycache__/modeling_tapas.cpython-310.pyc new file mode 100644 index 0000000000000000000000000000000000000000..9d2588e000863ef90644807265e967b790090b0a Binary files /dev/null and b/parrot/lib/python3.10/site-packages/transformers/models/tapas/__pycache__/modeling_tapas.cpython-310.pyc differ diff --git a/parrot/lib/python3.10/site-packages/transformers/models/tapas/__pycache__/modeling_tf_tapas.cpython-310.pyc b/parrot/lib/python3.10/site-packages/transformers/models/tapas/__pycache__/modeling_tf_tapas.cpython-310.pyc new file mode 100644 index 0000000000000000000000000000000000000000..5df31be7250e52ac79268f844fc6469c0f414c2b Binary files /dev/null and b/parrot/lib/python3.10/site-packages/transformers/models/tapas/__pycache__/modeling_tf_tapas.cpython-310.pyc differ diff --git a/parrot/lib/python3.10/site-packages/transformers/models/tapas/__pycache__/tokenization_tapas.cpython-310.pyc b/parrot/lib/python3.10/site-packages/transformers/models/tapas/__pycache__/tokenization_tapas.cpython-310.pyc new file mode 100644 index 0000000000000000000000000000000000000000..48ef12c2f464d67bdbcdfc24c1de37622e9370a5 Binary files /dev/null and b/parrot/lib/python3.10/site-packages/transformers/models/tapas/__pycache__/tokenization_tapas.cpython-310.pyc differ diff --git a/parrot/lib/python3.10/site-packages/transformers/models/tapas/configuration_tapas.py b/parrot/lib/python3.10/site-packages/transformers/models/tapas/configuration_tapas.py new file mode 100644 index 0000000000000000000000000000000000000000..cbc5cebf4e1b763c7cd59e3ff4a1a9a41cb287f9 --- /dev/null +++ b/parrot/lib/python3.10/site-packages/transformers/models/tapas/configuration_tapas.py @@ -0,0 +1,227 @@ +# coding=utf-8 +# Copyright 2020 Google Research and The HuggingFace Inc. team. +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. +""" +TAPAS configuration. Based on the BERT configuration with added parameters. + +Hyperparameters are taken from run_task_main.py and hparam_utils.py of the original implementation. URLS: + +- https://github.com/google-research/tapas/blob/master/tapas/run_task_main.py +- https://github.com/google-research/tapas/blob/master/tapas/utils/hparam_utils.py + +""" + + +from ...configuration_utils import PretrainedConfig + + +class TapasConfig(PretrainedConfig): + r""" + This is the configuration class to store the configuration of a [`TapasModel`]. It is used to instantiate a TAPAS + model according to the specified arguments, defining the model architecture. Instantiating a configuration with the + defaults will yield a similar configuration to that of the TAPAS + [google/tapas-base-finetuned-sqa](https://huggingface.co/google/tapas-base-finetuned-sqa) architecture. + + Configuration objects inherit from [`PreTrainedConfig`] and can be used to control the model outputs. Read the + documentation from [`PretrainedConfig`] for more information. + + Hyperparameters additional to BERT are taken from run_task_main.py and hparam_utils.py of the original + implementation. Original implementation available at https://github.com/google-research/tapas/tree/master. + + Args: + vocab_size (`int`, *optional*, defaults to 30522): + Vocabulary size of the TAPAS model. Defines the number of different tokens that can be represented by the + `inputs_ids` passed when calling [`TapasModel`]. + hidden_size (`int`, *optional*, defaults to 768): + Dimensionality of the encoder layers and the pooler layer. + num_hidden_layers (`int`, *optional*, defaults to 12): + Number of hidden layers in the Transformer encoder. + num_attention_heads (`int`, *optional*, defaults to 12): + Number of attention heads for each attention layer in the Transformer encoder. + intermediate_size (`int`, *optional*, defaults to 3072): + Dimensionality of the "intermediate" (often named feed-forward) layer in the Transformer encoder. + hidden_act (`str` or `Callable`, *optional*, defaults to `"gelu"`): + The non-linear activation function (function or string) in the encoder and pooler. If string, `"gelu"`, + `"relu"`, `"swish"` and `"gelu_new"` are supported. + hidden_dropout_prob (`float`, *optional*, defaults to 0.1): + The dropout probability for all fully connected layers in the embeddings, encoder, and pooler. + attention_probs_dropout_prob (`float`, *optional*, defaults to 0.1): + The dropout ratio for the attention probabilities. + max_position_embeddings (`int`, *optional*, defaults to 1024): + The maximum sequence length that this model might ever be used with. Typically set this to something large + just in case (e.g., 512 or 1024 or 2048). + type_vocab_sizes (`List[int]`, *optional*, defaults to `[3, 256, 256, 2, 256, 256, 10]`): + The vocabulary sizes of the `token_type_ids` passed when calling [`TapasModel`]. + initializer_range (`float`, *optional*, defaults to 0.02): + The standard deviation of the truncated_normal_initializer for initializing all weight matrices. + layer_norm_eps (`float`, *optional*, defaults to 1e-12): + The epsilon used by the layer normalization layers. + positive_label_weight (`float`, *optional*, defaults to 10.0): + Weight for positive labels. + num_aggregation_labels (`int`, *optional*, defaults to 0): + The number of aggregation operators to predict. + aggregation_loss_weight (`float`, *optional*, defaults to 1.0): + Importance weight for the aggregation loss. + use_answer_as_supervision (`bool`, *optional*): + Whether to use the answer as the only supervision for aggregation examples. + answer_loss_importance (`float`, *optional*, defaults to 1.0): + Importance weight for the regression loss. + use_normalized_answer_loss (`bool`, *optional*, defaults to `False`): + Whether to normalize the answer loss by the maximum of the predicted and expected value. + huber_loss_delta (`float`, *optional*): + Delta parameter used to calculate the regression loss. + temperature (`float`, *optional*, defaults to 1.0): + Value used to control (OR change) the skewness of cell logits probabilities. + aggregation_temperature (`float`, *optional*, defaults to 1.0): + Scales aggregation logits to control the skewness of probabilities. + use_gumbel_for_cells (`bool`, *optional*, defaults to `False`): + Whether to apply Gumbel-Softmax to cell selection. + use_gumbel_for_aggregation (`bool`, *optional*, defaults to `False`): + Whether to apply Gumbel-Softmax to aggregation selection. + average_approximation_function (`string`, *optional*, defaults to `"ratio"`): + Method to calculate the expected average of cells in the weak supervision case. One of `"ratio"`, + `"first_order"` or `"second_order"`. + cell_selection_preference (`float`, *optional*): + Preference for cell selection in ambiguous cases. Only applicable in case of weak supervision for + aggregation (WTQ, WikiSQL). If the total mass of the aggregation probabilities (excluding the "NONE" + operator) is higher than this hyperparameter, then aggregation is predicted for an example. + answer_loss_cutoff (`float`, *optional*): + Ignore examples with answer loss larger than cutoff. + max_num_rows (`int`, *optional*, defaults to 64): + Maximum number of rows. + max_num_columns (`int`, *optional*, defaults to 32): + Maximum number of columns. + average_logits_per_cell (`bool`, *optional*, defaults to `False`): + Whether to average logits per cell. + select_one_column (`bool`, *optional*, defaults to `True`): + Whether to constrain the model to only select cells from a single column. + allow_empty_column_selection (`bool`, *optional*, defaults to `False`): + Whether to allow not to select any column. + init_cell_selection_weights_to_zero (`bool`, *optional*, defaults to `False`): + Whether to initialize cell selection weights to 0 so that the initial probabilities are 50%. + reset_position_index_per_cell (`bool`, *optional*, defaults to `True`): + Whether to restart position indexes at every cell (i.e. use relative position embeddings). + disable_per_token_loss (`bool`, *optional*, defaults to `False`): + Whether to disable any (strong or weak) supervision on cells. + aggregation_labels (`Dict[int, label]`, *optional*): + The aggregation labels used to aggregate the results. For example, the WTQ models have the following + aggregation labels: `{0: "NONE", 1: "SUM", 2: "AVERAGE", 3: "COUNT"}` + no_aggregation_label_index (`int`, *optional*): + If the aggregation labels are defined and one of these labels represents "No aggregation", this should be + set to its index. For example, the WTQ models have the "NONE" aggregation label at index 0, so that value + should be set to 0 for these models. + + + Example: + + ```python + >>> from transformers import TapasModel, TapasConfig + + >>> # Initializing a default (SQA) Tapas configuration + >>> configuration = TapasConfig() + >>> # Initializing a model from the configuration + >>> model = TapasModel(configuration) + >>> # Accessing the model configuration + >>> configuration = model.config + ```""" + + model_type = "tapas" + + def __init__( + self, + vocab_size=30522, + hidden_size=768, + num_hidden_layers=12, + num_attention_heads=12, + intermediate_size=3072, + hidden_act="gelu", + hidden_dropout_prob=0.1, + attention_probs_dropout_prob=0.1, + max_position_embeddings=1024, + type_vocab_sizes=[3, 256, 256, 2, 256, 256, 10], + initializer_range=0.02, + layer_norm_eps=1e-12, + pad_token_id=0, + positive_label_weight=10.0, + num_aggregation_labels=0, + aggregation_loss_weight=1.0, + use_answer_as_supervision=None, + answer_loss_importance=1.0, + use_normalized_answer_loss=False, + huber_loss_delta=None, + temperature=1.0, + aggregation_temperature=1.0, + use_gumbel_for_cells=False, + use_gumbel_for_aggregation=False, + average_approximation_function="ratio", + cell_selection_preference=None, + answer_loss_cutoff=None, + max_num_rows=64, + max_num_columns=32, + average_logits_per_cell=False, + select_one_column=True, + allow_empty_column_selection=False, + init_cell_selection_weights_to_zero=False, + reset_position_index_per_cell=True, + disable_per_token_loss=False, + aggregation_labels=None, + no_aggregation_label_index=None, + **kwargs, + ): + super().__init__(pad_token_id=pad_token_id, **kwargs) + + # BERT hyperparameters (with updated max_position_embeddings and type_vocab_sizes) + self.vocab_size = vocab_size + self.hidden_size = hidden_size + self.num_hidden_layers = num_hidden_layers + self.num_attention_heads = num_attention_heads + self.hidden_act = hidden_act + self.intermediate_size = intermediate_size + self.hidden_dropout_prob = hidden_dropout_prob + self.attention_probs_dropout_prob = attention_probs_dropout_prob + self.max_position_embeddings = max_position_embeddings + self.type_vocab_sizes = type_vocab_sizes + self.initializer_range = initializer_range + self.layer_norm_eps = layer_norm_eps + + # Fine-tuning task hyperparameters + self.positive_label_weight = positive_label_weight + self.num_aggregation_labels = num_aggregation_labels + self.aggregation_loss_weight = aggregation_loss_weight + self.use_answer_as_supervision = use_answer_as_supervision + self.answer_loss_importance = answer_loss_importance + self.use_normalized_answer_loss = use_normalized_answer_loss + self.huber_loss_delta = huber_loss_delta + self.temperature = temperature + self.aggregation_temperature = aggregation_temperature + self.use_gumbel_for_cells = use_gumbel_for_cells + self.use_gumbel_for_aggregation = use_gumbel_for_aggregation + self.average_approximation_function = average_approximation_function + self.cell_selection_preference = cell_selection_preference + self.answer_loss_cutoff = answer_loss_cutoff + self.max_num_rows = max_num_rows + self.max_num_columns = max_num_columns + self.average_logits_per_cell = average_logits_per_cell + self.select_one_column = select_one_column + self.allow_empty_column_selection = allow_empty_column_selection + self.init_cell_selection_weights_to_zero = init_cell_selection_weights_to_zero + self.reset_position_index_per_cell = reset_position_index_per_cell + self.disable_per_token_loss = disable_per_token_loss + + # Aggregation hyperparameters + self.aggregation_labels = aggregation_labels + self.no_aggregation_label_index = no_aggregation_label_index + + if isinstance(self.aggregation_labels, dict): + self.aggregation_labels = {int(k): v for k, v in aggregation_labels.items()} diff --git a/parrot/lib/python3.10/site-packages/transformers/models/tapas/convert_tapas_original_tf_checkpoint_to_pytorch.py b/parrot/lib/python3.10/site-packages/transformers/models/tapas/convert_tapas_original_tf_checkpoint_to_pytorch.py new file mode 100644 index 0000000000000000000000000000000000000000..2772a7f126ef9ad350837e993e264c70e68ae3fb --- /dev/null +++ b/parrot/lib/python3.10/site-packages/transformers/models/tapas/convert_tapas_original_tf_checkpoint_to_pytorch.py @@ -0,0 +1,138 @@ +# coding=utf-8 +# Copyright 2020 The HuggingFace Inc. team. +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. +"""Convert TAPAS checkpoint.""" + + +import argparse + +from transformers import ( + TapasConfig, + TapasForMaskedLM, + TapasForQuestionAnswering, + TapasForSequenceClassification, + TapasModel, + TapasTokenizer, + load_tf_weights_in_tapas, +) +from transformers.utils import logging + + +logging.set_verbosity_info() + + +def convert_tf_checkpoint_to_pytorch( + task, reset_position_index_per_cell, tf_checkpoint_path, tapas_config_file, pytorch_dump_path +): + # Initialise PyTorch model. + # If you want to convert a checkpoint that uses absolute position embeddings, make sure to set reset_position_index_per_cell of + # TapasConfig to False. + + # initialize configuration from json file + config = TapasConfig.from_json_file(tapas_config_file) + # set absolute/relative position embeddings parameter + config.reset_position_index_per_cell = reset_position_index_per_cell + + # set remaining parameters of TapasConfig as well as the model based on the task + if task == "SQA": + model = TapasForQuestionAnswering(config=config) + elif task == "WTQ": + # run_task_main.py hparams + config.num_aggregation_labels = 4 + config.use_answer_as_supervision = True + # hparam_utils.py hparams + config.answer_loss_cutoff = 0.664694 + config.cell_selection_preference = 0.207951 + config.huber_loss_delta = 0.121194 + config.init_cell_selection_weights_to_zero = True + config.select_one_column = True + config.allow_empty_column_selection = False + config.temperature = 0.0352513 + + model = TapasForQuestionAnswering(config=config) + elif task == "WIKISQL_SUPERVISED": + # run_task_main.py hparams + config.num_aggregation_labels = 4 + config.use_answer_as_supervision = False + # hparam_utils.py hparams + config.answer_loss_cutoff = 36.4519 + config.cell_selection_preference = 0.903421 + config.huber_loss_delta = 222.088 + config.init_cell_selection_weights_to_zero = True + config.select_one_column = True + config.allow_empty_column_selection = True + config.temperature = 0.763141 + + model = TapasForQuestionAnswering(config=config) + elif task == "TABFACT": + model = TapasForSequenceClassification(config=config) + elif task == "MLM": + model = TapasForMaskedLM(config=config) + elif task == "INTERMEDIATE_PRETRAINING": + model = TapasModel(config=config) + else: + raise ValueError(f"Task {task} not supported.") + + print(f"Building PyTorch model from configuration: {config}") + # Load weights from tf checkpoint + load_tf_weights_in_tapas(model, config, tf_checkpoint_path) + + # Save pytorch-model (weights and configuration) + print(f"Save PyTorch model to {pytorch_dump_path}") + model.save_pretrained(pytorch_dump_path) + + # Save tokenizer files + print(f"Save tokenizer files to {pytorch_dump_path}") + tokenizer = TapasTokenizer(vocab_file=tf_checkpoint_path[:-10] + "vocab.txt", model_max_length=512) + tokenizer.save_pretrained(pytorch_dump_path) + + print("Used relative position embeddings:", model.config.reset_position_index_per_cell) + + +if __name__ == "__main__": + parser = argparse.ArgumentParser() + # Required parameters + parser.add_argument( + "--task", default="SQA", type=str, help="Model task for which to convert a checkpoint. Defaults to SQA." + ) + parser.add_argument( + "--reset_position_index_per_cell", + default=False, + action="store_true", + help="Whether to use relative position embeddings or not. Defaults to True.", + ) + parser.add_argument( + "--tf_checkpoint_path", default=None, type=str, required=True, help="Path to the TensorFlow checkpoint path." + ) + parser.add_argument( + "--tapas_config_file", + default=None, + type=str, + required=True, + help=( + "The config json file corresponding to the pre-trained TAPAS model. \n" + "This specifies the model architecture." + ), + ) + parser.add_argument( + "--pytorch_dump_path", default=None, type=str, required=True, help="Path to the output PyTorch model." + ) + args = parser.parse_args() + convert_tf_checkpoint_to_pytorch( + args.task, + args.reset_position_index_per_cell, + args.tf_checkpoint_path, + args.tapas_config_file, + args.pytorch_dump_path, + ) diff --git a/parrot/lib/python3.10/site-packages/transformers/models/tapas/modeling_tapas.py b/parrot/lib/python3.10/site-packages/transformers/models/tapas/modeling_tapas.py new file mode 100644 index 0000000000000000000000000000000000000000..903fa2a225445597c43f9b5ba0c1cc0843c88332 --- /dev/null +++ b/parrot/lib/python3.10/site-packages/transformers/models/tapas/modeling_tapas.py @@ -0,0 +1,2389 @@ +# coding=utf-8 +# Copyright 2020 Google Research and The HuggingFace Inc. team. +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. +"""PyTorch TAPAS model.""" + + +import enum +import math +import os +from dataclasses import dataclass +from typing import Optional, Tuple, Union + +import torch +import torch.utils.checkpoint +from torch import nn +from torch.nn import BCEWithLogitsLoss, CrossEntropyLoss, MSELoss + +from ...activations import ACT2FN +from ...modeling_outputs import BaseModelOutput, BaseModelOutputWithPooling, MaskedLMOutput, SequenceClassifierOutput +from ...modeling_utils import PreTrainedModel +from ...pytorch_utils import ( + apply_chunking_to_forward, + find_pruneable_heads_and_indices, + is_torch_greater_or_equal_than_1_12, + prune_linear_layer, +) +from ...utils import ( + ModelOutput, + add_start_docstrings, + add_start_docstrings_to_model_forward, + logging, + replace_return_docstrings, +) +from .configuration_tapas import TapasConfig + + +logger = logging.get_logger(__name__) + +if not is_torch_greater_or_equal_than_1_12: + logger.warning( + f"You are using torch=={torch.__version__}, but torch>=1.12.0 is required to use " + "TapasModel. Please upgrade torch." + ) + +_CONFIG_FOR_DOC = "TapasConfig" +_CHECKPOINT_FOR_DOC = "google/tapas-base" + + +EPSILON_ZERO_DIVISION = 1e-10 +CLOSE_ENOUGH_TO_LOG_ZERO = -10000.0 + + +@dataclass +class TableQuestionAnsweringOutput(ModelOutput): + """ + Output type of [`TapasForQuestionAnswering`]. + + Args: + loss (`torch.FloatTensor` of shape `(1,)`, *optional*, returned when `labels` (and possibly `answer`, `aggregation_labels`, `numeric_values` and `numeric_values_scale` are provided)): + Total loss as the sum of the hierarchical cell selection log-likelihood loss and (optionally) the + semi-supervised regression loss and (optionally) supervised loss for aggregations. + logits (`torch.FloatTensor` of shape `(batch_size, sequence_length)`): + Prediction scores of the cell selection head, for every token. + logits_aggregation (`torch.FloatTensor`, *optional*, of shape `(batch_size, num_aggregation_labels)`): + Prediction scores of the aggregation head, for every aggregation operator. + hidden_states (`tuple(torch.FloatTensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`): + Tuple of `torch.FloatTensor` (one for the output of the embeddings + one for the output of each layer) of + shape `(batch_size, sequence_length, hidden_size)`. Hidden-states of the model at the output of each layer + plus the initial embedding outputs. + attentions (`tuple(torch.FloatTensor)`, *optional*, returned when `output_attentions=True` is passed or when `config.output_attentions=True`): + Tuple of `torch.FloatTensor` (one for each layer) of shape `(batch_size, num_heads, sequence_length, + sequence_length)`. Attentions weights after the attention softmax, used to compute the weighted average in + the self-attention heads. + """ + + loss: Optional[torch.FloatTensor] = None + logits: torch.FloatTensor = None + logits_aggregation: torch.FloatTensor = None + hidden_states: Optional[Tuple[torch.FloatTensor]] = None + attentions: Optional[Tuple[torch.FloatTensor]] = None + + +def load_tf_weights_in_tapas(model, config, tf_checkpoint_path): + """ + Load tf checkpoints in a PyTorch model. This is an adaptation from load_tf_weights_in_bert + + - add cell selection and aggregation heads + - take into account additional token type embedding layers + """ + try: + import re + + import numpy as np + import tensorflow as tf + except ImportError: + logger.error( + "Loading a TensorFlow model in PyTorch, requires TensorFlow to be installed. Please see " + "https://www.tensorflow.org/install/ for installation instructions." + ) + raise + tf_path = os.path.abspath(tf_checkpoint_path) + logger.info(f"Converting TensorFlow checkpoint from {tf_path}") + # Load weights from TF model + init_vars = tf.train.list_variables(tf_path) + names = [] + arrays = [] + for name, shape in init_vars: + logger.info(f"Loading TF weight {name} with shape {shape}") + array = tf.train.load_variable(tf_path, name) + names.append(name) + arrays.append(array) + + for name, array in zip(names, arrays): + name = name.split("/") + # adam_v and adam_m are variables used in AdamWeightDecayOptimizer to calculate m and v + # which are not required for using pretrained model + if any( + n + in [ + "adam_v", + "adam_m", + "AdamWeightDecayOptimizer", + "AdamWeightDecayOptimizer_1", + "global_step", + "seq_relationship", + ] + for n in name + ): + logger.info(f"Skipping {'/'.join(name)}") + continue + # in case the model is TapasForSequenceClassification, we skip output_bias and output_weights + # since these are not used for classification + if isinstance(model, TapasForSequenceClassification): + if any(n in ["output_bias", "output_weights"] for n in name): + logger.info(f"Skipping {'/'.join(name)}") + continue + # in case the model is TapasModel, we skip output_bias, output_weights, output_bias_cls and output_weights_cls + # since this model does not have MLM and NSP heads + if isinstance(model, TapasModel): + if any(n in ["output_bias", "output_weights", "output_bias_cls", "output_weights_cls"] for n in name): + logger.info(f"Skipping {'/'.join(name)}") + continue + # in case the model is TapasForMaskedLM, we skip the pooler + if isinstance(model, TapasForMaskedLM): + if any(n in ["pooler"] for n in name): + logger.info(f"Skipping {'/'.join(name)}") + continue + # if first scope name starts with "bert", change it to "tapas" + if name[0] == "bert": + name[0] = "tapas" + pointer = model + for m_name in name: + if re.fullmatch(r"[A-Za-z]+_\d+", m_name): + scope_names = re.split(r"_(\d+)", m_name) + else: + scope_names = [m_name] + if scope_names[0] == "kernel" or scope_names[0] == "gamma": + pointer = getattr(pointer, "weight") + elif scope_names[0] == "beta": + pointer = getattr(pointer, "bias") + # cell selection heads + elif scope_names[0] == "output_bias": + if not isinstance(model, TapasForMaskedLM): + pointer = getattr(pointer, "output_bias") + else: + pointer = getattr(pointer, "bias") + elif scope_names[0] == "output_weights": + pointer = getattr(pointer, "output_weights") + elif scope_names[0] == "column_output_bias": + pointer = getattr(pointer, "column_output_bias") + elif scope_names[0] == "column_output_weights": + pointer = getattr(pointer, "column_output_weights") + # aggregation head + elif scope_names[0] == "output_bias_agg": + pointer = getattr(pointer, "aggregation_classifier") + pointer = getattr(pointer, "bias") + elif scope_names[0] == "output_weights_agg": + pointer = getattr(pointer, "aggregation_classifier") + pointer = getattr(pointer, "weight") + # classification head + elif scope_names[0] == "output_bias_cls": + pointer = getattr(pointer, "classifier") + pointer = getattr(pointer, "bias") + elif scope_names[0] == "output_weights_cls": + pointer = getattr(pointer, "classifier") + pointer = getattr(pointer, "weight") + else: + try: + pointer = getattr(pointer, scope_names[0]) + except AttributeError: + logger.info(f"Skipping {'/'.join(name)}") + continue + if len(scope_names) >= 2: + num = int(scope_names[1]) + pointer = pointer[num] + if m_name[-11:] == "_embeddings": + pointer = getattr(pointer, "weight") + elif m_name[-13:] in [f"_embeddings_{i}" for i in range(7)]: + pointer = getattr(pointer, "weight") + elif m_name == "kernel": + array = np.transpose(array) + try: + if pointer.shape != array.shape: + raise ValueError(f"Pointer shape {pointer.shape} and array shape {array.shape} mismatched") + except AssertionError as e: + e.args += (pointer.shape, array.shape) + raise + logger.info(f"Initialize PyTorch weight {name}") + # Added a check to see whether the array is a scalar (because bias terms in Tapas checkpoints can be + # scalar => should first be converted to numpy arrays) + if np.isscalar(array): + array = np.array(array) + pointer.data = torch.from_numpy(array) + return model + + +class TapasEmbeddings(nn.Module): + """ + Construct the embeddings from word, position and token_type embeddings. Same as BertEmbeddings but with a number of + additional token type embeddings to encode tabular structure. + """ + + def __init__(self, config): + super().__init__() + # we do not include config.disabled_features and config.disable_position_embeddings from the original implementation + # word embeddings + self.word_embeddings = nn.Embedding(config.vocab_size, config.hidden_size, padding_idx=config.pad_token_id) + # position embeddings + self.position_embeddings = nn.Embedding(config.max_position_embeddings, config.hidden_size) + # token type embeddings + for i, type_vocab_sizes in enumerate(config.type_vocab_sizes): + name = f"token_type_embeddings_{i}" + setattr(self, name, nn.Embedding(type_vocab_sizes, config.hidden_size)) + + self.number_of_token_type_embeddings = len(config.type_vocab_sizes) + + # self.LayerNorm is not snake-cased to stick with TensorFlow model variable name and be able to load + # any TensorFlow checkpoint file + self.LayerNorm = nn.LayerNorm(config.hidden_size, eps=config.layer_norm_eps) + self.dropout = nn.Dropout(config.hidden_dropout_prob) + + self.config = config + + def forward(self, input_ids=None, token_type_ids=None, position_ids=None, inputs_embeds=None): + if input_ids is not None: + input_shape = input_ids.size() + else: + input_shape = inputs_embeds.size()[:-1] + + seq_length = input_shape[1] + device = input_ids.device if input_ids is not None else inputs_embeds.device + + if position_ids is None: + # create absolute position embeddings + position_ids = torch.arange(seq_length, dtype=torch.long, device=device) + position_ids = position_ids.unsqueeze(0).expand(input_shape) + # when self.config.reset_position_index_per_cell is set to True, create relative position embeddings + if self.config.reset_position_index_per_cell: + # shape (batch_size, seq_len) + col_index = IndexMap(token_type_ids[:, :, 1], self.config.type_vocab_sizes[1], batch_dims=1) + # shape (batch_size, seq_len) + row_index = IndexMap(token_type_ids[:, :, 2], self.config.type_vocab_sizes[2], batch_dims=1) + # shape (batch_size, seq_len) + full_index = ProductIndexMap(col_index, row_index) + # shape (max_rows * max_columns,). First absolute position for every cell + first_position_per_segment = reduce_min(position_ids, full_index)[0] + # ? shape (batch_size, seq_len). First absolute position of the cell for every token + first_position = gather(first_position_per_segment, full_index) + # shape (1, seq_len) + position = torch.arange(seq_length, dtype=torch.long, device=device).unsqueeze(0) + position_ids = torch.min( + torch.as_tensor(self.config.max_position_embeddings - 1, device=device), position - first_position + ) + + if token_type_ids is None: + token_type_ids = torch.zeros( + (input_shape + self.number_of_token_type_embeddings), dtype=torch.long, device=device + ) + + if inputs_embeds is None: + inputs_embeds = self.word_embeddings(input_ids) + + position_embeddings = self.position_embeddings(position_ids) + + embeddings = inputs_embeds + position_embeddings + + for i in range(self.number_of_token_type_embeddings): + name = f"token_type_embeddings_{i}" + embeddings += getattr(self, name)(token_type_ids[:, :, i]) + + embeddings = self.LayerNorm(embeddings) + embeddings = self.dropout(embeddings) + return embeddings + + +class TapasSelfAttention(nn.Module): + def __init__(self, config): + super().__init__() + if config.hidden_size % config.num_attention_heads != 0 and not hasattr(config, "embedding_size"): + raise ValueError( + f"The hidden size {config.hidden_size} is not a multiple of the number of attention " + f"heads {config.num_attention_heads}" + ) + + self.num_attention_heads = config.num_attention_heads + self.attention_head_size = int(config.hidden_size / config.num_attention_heads) + self.all_head_size = self.num_attention_heads * self.attention_head_size + + self.query = nn.Linear(config.hidden_size, self.all_head_size) + self.key = nn.Linear(config.hidden_size, self.all_head_size) + self.value = nn.Linear(config.hidden_size, self.all_head_size) + + self.dropout = nn.Dropout(config.attention_probs_dropout_prob) + self.is_decoder = config.is_decoder + + def transpose_for_scores(self, x): + new_x_shape = x.size()[:-1] + (self.num_attention_heads, self.attention_head_size) + x = x.view(*new_x_shape) + return x.permute(0, 2, 1, 3) + + def forward( + self, + hidden_states, + attention_mask=None, + head_mask=None, + encoder_hidden_states=None, + encoder_attention_mask=None, + past_key_value=None, + output_attentions=False, + ): + mixed_query_layer = self.query(hidden_states) + + # If this is instantiated as a cross-attention module, the keys + # and values come from an encoder; the attention mask needs to be + # such that the encoder's padding tokens are not attended to. + is_cross_attention = encoder_hidden_states is not None + + if is_cross_attention and past_key_value is not None: + # reuse k,v, cross_attentions + key_layer = past_key_value[0] + value_layer = past_key_value[1] + attention_mask = encoder_attention_mask + elif is_cross_attention: + key_layer = self.transpose_for_scores(self.key(encoder_hidden_states)) + value_layer = self.transpose_for_scores(self.value(encoder_hidden_states)) + attention_mask = encoder_attention_mask + elif past_key_value is not None: + key_layer = self.transpose_for_scores(self.key(hidden_states)) + value_layer = self.transpose_for_scores(self.value(hidden_states)) + key_layer = torch.cat([past_key_value[0], key_layer], dim=2) + value_layer = torch.cat([past_key_value[1], value_layer], dim=2) + else: + key_layer = self.transpose_for_scores(self.key(hidden_states)) + value_layer = self.transpose_for_scores(self.value(hidden_states)) + + query_layer = self.transpose_for_scores(mixed_query_layer) + + if self.is_decoder: + past_key_value = (key_layer, value_layer) + + # Take the dot product between "query" and "key" to get the raw attention scores. + attention_scores = torch.matmul(query_layer, key_layer.transpose(-1, -2)) + attention_scores = attention_scores / math.sqrt(self.attention_head_size) + if attention_mask is not None: + # Apply the attention mask is (precomputed for all layers in TapasModel forward() function) + attention_scores = attention_scores + attention_mask + + # Normalize the attention scores to probabilities. + attention_probs = nn.functional.softmax(attention_scores, dim=-1) + + # This is actually dropping out entire tokens to attend to, which might + # seem a bit unusual, but is taken from the original Transformer paper. + attention_probs = self.dropout(attention_probs) + + # Mask heads if we want to + if head_mask is not None: + attention_probs = attention_probs * head_mask + + context_layer = torch.matmul(attention_probs, value_layer) + + context_layer = context_layer.permute(0, 2, 1, 3).contiguous() + new_context_layer_shape = context_layer.size()[:-2] + (self.all_head_size,) + context_layer = context_layer.view(*new_context_layer_shape) + + outputs = (context_layer, attention_probs) if output_attentions else (context_layer,) + if self.is_decoder: + outputs = outputs + (past_key_value,) + return outputs + + +# Copied from transformers.models.bert.modeling_bert.BertSelfOutput +class TapasSelfOutput(nn.Module): + def __init__(self, config): + super().__init__() + self.dense = nn.Linear(config.hidden_size, config.hidden_size) + self.LayerNorm = nn.LayerNorm(config.hidden_size, eps=config.layer_norm_eps) + self.dropout = nn.Dropout(config.hidden_dropout_prob) + + def forward(self, hidden_states: torch.Tensor, input_tensor: torch.Tensor) -> torch.Tensor: + hidden_states = self.dense(hidden_states) + hidden_states = self.dropout(hidden_states) + hidden_states = self.LayerNorm(hidden_states + input_tensor) + return hidden_states + + +class TapasAttention(nn.Module): + def __init__(self, config): + super().__init__() + self.self = TapasSelfAttention(config) + self.output = TapasSelfOutput(config) + self.pruned_heads = set() + + # Copied from transformers.models.bert.modeling_bert.BertAttention.prune_heads + def prune_heads(self, heads): + if len(heads) == 0: + return + heads, index = find_pruneable_heads_and_indices( + heads, self.self.num_attention_heads, self.self.attention_head_size, self.pruned_heads + ) + + # Prune linear layers + self.self.query = prune_linear_layer(self.self.query, index) + self.self.key = prune_linear_layer(self.self.key, index) + self.self.value = prune_linear_layer(self.self.value, index) + self.output.dense = prune_linear_layer(self.output.dense, index, dim=1) + + # Update hyper params and store pruned heads + self.self.num_attention_heads = self.self.num_attention_heads - len(heads) + self.self.all_head_size = self.self.attention_head_size * self.self.num_attention_heads + self.pruned_heads = self.pruned_heads.union(heads) + + # Copied from transformers.models.bert.modeling_bert.BertAttention.forward + def forward( + self, + hidden_states: torch.Tensor, + attention_mask: Optional[torch.FloatTensor] = None, + head_mask: Optional[torch.FloatTensor] = None, + encoder_hidden_states: Optional[torch.FloatTensor] = None, + encoder_attention_mask: Optional[torch.FloatTensor] = None, + past_key_value: Optional[Tuple[Tuple[torch.FloatTensor]]] = None, + output_attentions: Optional[bool] = False, + ) -> Tuple[torch.Tensor]: + self_outputs = self.self( + hidden_states, + attention_mask, + head_mask, + encoder_hidden_states, + encoder_attention_mask, + past_key_value, + output_attentions, + ) + attention_output = self.output(self_outputs[0], hidden_states) + outputs = (attention_output,) + self_outputs[1:] # add attentions if we output them + return outputs + + +# Copied from transformers.models.bert.modeling_bert.BertIntermediate +class TapasIntermediate(nn.Module): + def __init__(self, config): + super().__init__() + self.dense = nn.Linear(config.hidden_size, config.intermediate_size) + if isinstance(config.hidden_act, str): + self.intermediate_act_fn = ACT2FN[config.hidden_act] + else: + self.intermediate_act_fn = config.hidden_act + + def forward(self, hidden_states: torch.Tensor) -> torch.Tensor: + hidden_states = self.dense(hidden_states) + hidden_states = self.intermediate_act_fn(hidden_states) + return hidden_states + + +# Copied from transformers.models.bert.modeling_bert.BertOutput +class TapasOutput(nn.Module): + def __init__(self, config): + super().__init__() + self.dense = nn.Linear(config.intermediate_size, config.hidden_size) + self.LayerNorm = nn.LayerNorm(config.hidden_size, eps=config.layer_norm_eps) + self.dropout = nn.Dropout(config.hidden_dropout_prob) + + def forward(self, hidden_states: torch.Tensor, input_tensor: torch.Tensor) -> torch.Tensor: + hidden_states = self.dense(hidden_states) + hidden_states = self.dropout(hidden_states) + hidden_states = self.LayerNorm(hidden_states + input_tensor) + return hidden_states + + +class TapasLayer(nn.Module): + def __init__(self, config): + super().__init__() + self.chunk_size_feed_forward = config.chunk_size_feed_forward + self.seq_len_dim = 1 + self.attention = TapasAttention(config) + self.is_decoder = config.is_decoder + self.add_cross_attention = config.add_cross_attention + if self.add_cross_attention: + if not self.is_decoder: + raise ValueError(f"{self} should be used as a decoder model if cross attention is added") + self.crossattention = TapasAttention(config) + self.intermediate = TapasIntermediate(config) + self.output = TapasOutput(config) + + # Copied from transformers.models.bert.modeling_bert.BertLayer.forward + def forward( + self, + hidden_states: torch.Tensor, + attention_mask: Optional[torch.FloatTensor] = None, + head_mask: Optional[torch.FloatTensor] = None, + encoder_hidden_states: Optional[torch.FloatTensor] = None, + encoder_attention_mask: Optional[torch.FloatTensor] = None, + past_key_value: Optional[Tuple[Tuple[torch.FloatTensor]]] = None, + output_attentions: Optional[bool] = False, + ) -> Tuple[torch.Tensor]: + # decoder uni-directional self-attention cached key/values tuple is at positions 1,2 + self_attn_past_key_value = past_key_value[:2] if past_key_value is not None else None + self_attention_outputs = self.attention( + hidden_states, + attention_mask, + head_mask, + output_attentions=output_attentions, + past_key_value=self_attn_past_key_value, + ) + attention_output = self_attention_outputs[0] + + # if decoder, the last output is tuple of self-attn cache + if self.is_decoder: + outputs = self_attention_outputs[1:-1] + present_key_value = self_attention_outputs[-1] + else: + outputs = self_attention_outputs[1:] # add self attentions if we output attention weights + + cross_attn_present_key_value = None + if self.is_decoder and encoder_hidden_states is not None: + if not hasattr(self, "crossattention"): + raise ValueError( + f"If `encoder_hidden_states` are passed, {self} has to be instantiated with cross-attention layers" + " by setting `config.add_cross_attention=True`" + ) + + # cross_attn cached key/values tuple is at positions 3,4 of past_key_value tuple + cross_attn_past_key_value = past_key_value[-2:] if past_key_value is not None else None + cross_attention_outputs = self.crossattention( + attention_output, + attention_mask, + head_mask, + encoder_hidden_states, + encoder_attention_mask, + cross_attn_past_key_value, + output_attentions, + ) + attention_output = cross_attention_outputs[0] + outputs = outputs + cross_attention_outputs[1:-1] # add cross attentions if we output attention weights + + # add cross-attn cache to positions 3,4 of present_key_value tuple + cross_attn_present_key_value = cross_attention_outputs[-1] + present_key_value = present_key_value + cross_attn_present_key_value + + layer_output = apply_chunking_to_forward( + self.feed_forward_chunk, self.chunk_size_feed_forward, self.seq_len_dim, attention_output + ) + outputs = (layer_output,) + outputs + + # if decoder, return the attn key/values as the last output + if self.is_decoder: + outputs = outputs + (present_key_value,) + + return outputs + + # Copied from transformers.models.bert.modeling_bert.BertLayer.feed_forward_chunk + def feed_forward_chunk(self, attention_output): + intermediate_output = self.intermediate(attention_output) + layer_output = self.output(intermediate_output, attention_output) + return layer_output + + +class TapasEncoder(nn.Module): + def __init__(self, config): + super().__init__() + self.config = config + self.layer = nn.ModuleList([TapasLayer(config) for _ in range(config.num_hidden_layers)]) + self.gradient_checkpointing = False + + def forward( + self, + hidden_states, + attention_mask=None, + head_mask=None, + encoder_hidden_states=None, + encoder_attention_mask=None, + past_key_values=None, + use_cache=None, + output_attentions=False, + output_hidden_states=False, + return_dict=True, + ): + all_hidden_states = () if output_hidden_states else None + all_attentions = () if output_attentions else None + for i, layer_module in enumerate(self.layer): + if output_hidden_states: + all_hidden_states = all_hidden_states + (hidden_states,) + + layer_head_mask = head_mask[i] if head_mask is not None else None + + if self.gradient_checkpointing and self.training: + layer_outputs = self._gradient_checkpointing_func( + layer_module.__call__, + hidden_states, + attention_mask, + layer_head_mask, + encoder_hidden_states, + encoder_attention_mask, + past_key_values, + output_attentions, + ) + else: + layer_outputs = layer_module( + hidden_states, + attention_mask, + layer_head_mask, + encoder_hidden_states, + encoder_attention_mask, + past_key_values, + output_attentions, + ) + hidden_states = layer_outputs[0] + if output_attentions: + all_attentions = all_attentions + (layer_outputs[1],) + + if output_hidden_states: + all_hidden_states = all_hidden_states + (hidden_states,) + + if not return_dict: + return tuple(v for v in [hidden_states, all_hidden_states, all_attentions] if v is not None) + return BaseModelOutput( + last_hidden_state=hidden_states, hidden_states=all_hidden_states, attentions=all_attentions + ) + + +# Copied from transformers.models.bert.modeling_bert.BertPooler +class TapasPooler(nn.Module): + def __init__(self, config): + super().__init__() + self.dense = nn.Linear(config.hidden_size, config.hidden_size) + self.activation = nn.Tanh() + + def forward(self, hidden_states: torch.Tensor) -> torch.Tensor: + # We "pool" the model by simply taking the hidden state corresponding + # to the first token. + first_token_tensor = hidden_states[:, 0] + pooled_output = self.dense(first_token_tensor) + pooled_output = self.activation(pooled_output) + return pooled_output + + +# Copied from transformers.models.bert.modeling_bert.BertPredictionHeadTransform with Bert->Tapas +class TapasPredictionHeadTransform(nn.Module): + def __init__(self, config): + super().__init__() + self.dense = nn.Linear(config.hidden_size, config.hidden_size) + if isinstance(config.hidden_act, str): + self.transform_act_fn = ACT2FN[config.hidden_act] + else: + self.transform_act_fn = config.hidden_act + self.LayerNorm = nn.LayerNorm(config.hidden_size, eps=config.layer_norm_eps) + + def forward(self, hidden_states: torch.Tensor) -> torch.Tensor: + hidden_states = self.dense(hidden_states) + hidden_states = self.transform_act_fn(hidden_states) + hidden_states = self.LayerNorm(hidden_states) + return hidden_states + + +# Copied from transformers.models.bert.modeling_bert.BertLMPredictionHead with Bert->Tapas +class TapasLMPredictionHead(nn.Module): + def __init__(self, config): + super().__init__() + self.transform = TapasPredictionHeadTransform(config) + + # The output weights are the same as the input embeddings, but there is + # an output-only bias for each token. + self.decoder = nn.Linear(config.hidden_size, config.vocab_size, bias=False) + + self.bias = nn.Parameter(torch.zeros(config.vocab_size)) + + # Need a link between the two variables so that the bias is correctly resized with `resize_token_embeddings` + self.decoder.bias = self.bias + + def _tie_weights(self): + self.decoder.bias = self.bias + + def forward(self, hidden_states): + hidden_states = self.transform(hidden_states) + hidden_states = self.decoder(hidden_states) + return hidden_states + + +# Copied from transformers.models.bert.modeling_bert.BertOnlyMLMHead with Bert->Tapas +class TapasOnlyMLMHead(nn.Module): + def __init__(self, config): + super().__init__() + self.predictions = TapasLMPredictionHead(config) + + def forward(self, sequence_output: torch.Tensor) -> torch.Tensor: + prediction_scores = self.predictions(sequence_output) + return prediction_scores + + +class TapasPreTrainedModel(PreTrainedModel): + """ + An abstract class to handle weights initialization and a simple interface for downloading and loading pretrained + models. + """ + + config_class = TapasConfig + base_model_prefix = "tapas" + supports_gradient_checkpointing = True + + # Copied from transformers.models.bert.modeling_bert.BertPreTrainedModel._init_weights + def _init_weights(self, module): + """Initialize the weights""" + if isinstance(module, nn.Linear): + # Slightly different from the TF version which uses truncated_normal for initialization + # cf https://github.com/pytorch/pytorch/pull/5617 + module.weight.data.normal_(mean=0.0, std=self.config.initializer_range) + if module.bias is not None: + module.bias.data.zero_() + elif isinstance(module, nn.Embedding): + module.weight.data.normal_(mean=0.0, std=self.config.initializer_range) + if module.padding_idx is not None: + module.weight.data[module.padding_idx].zero_() + elif isinstance(module, nn.LayerNorm): + module.bias.data.zero_() + module.weight.data.fill_(1.0) + + +TAPAS_START_DOCSTRING = r""" + This model inherits from [`PreTrainedModel`]. Check the superclass documentation for the generic methods the + library implements for all its models (such as downloading or saving, resizing the input embeddings, pruning heads + etc.) + + This model is also a PyTorch [torch.nn.Module](https://pytorch.org/docs/stable/nn.html#torch.nn.Module) subclass. + Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage + and behavior. + + Parameters: + config ([`TapasConfig`]): Model configuration class with all the parameters of the model. + Initializing with a config file does not load the weights associated with the model, only the + configuration. Check out the [`~PreTrainedModel.from_pretrained`] method to load the model weights. +""" + +TAPAS_INPUTS_DOCSTRING = r""" + Args: + input_ids (`torch.LongTensor` of shape `({0})`): + Indices of input sequence tokens in the vocabulary. Indices can be obtained using [`AutoTokenizer`]. See + [`PreTrainedTokenizer.encode`] and [`PreTrainedTokenizer.__call__`] for details. + + [What are input IDs?](../glossary#input-ids) + attention_mask (`torch.FloatTensor` of shape `({0})`, *optional*): + Mask to avoid performing attention on padding token indices. Mask values selected in `[0, 1]`: + + - 1 for tokens that are **not masked**, + - 0 for tokens that are **masked**. + + [What are attention masks?](../glossary#attention-mask) + token_type_ids (`torch.LongTensor` of shape `({0}, 7)`, *optional*): + Token indices that encode tabular structure. Indices can be obtained using [`AutoTokenizer`]. See this + class for more info. + + [What are token type IDs?](../glossary#token-type-ids) + position_ids (`torch.LongTensor` of shape `({0})`, *optional*): + Indices of positions of each input sequence tokens in the position embeddings. If + `reset_position_index_per_cell` of [`TapasConfig`] is set to `True`, relative position embeddings will be + used. Selected in the range `[0, config.max_position_embeddings - 1]`. + + [What are position IDs?](../glossary#position-ids) + head_mask (`torch.FloatTensor` of shape `(num_heads,)` or `(num_layers, num_heads)`, *optional*): + Mask to nullify selected heads of the self-attention modules. Mask values selected in `[0, 1]`: - 1 + indicates the head is **not masked**, - 0 indicates the head is **masked**. + inputs_embeds (`torch.FloatTensor` of shape `({0}, hidden_size)`, *optional*): + Optionally, instead of passing `input_ids` you can choose to directly pass an embedded representation. This + is useful if you want more control over how to convert `input_ids` indices into associated vectors than the + model's internal embedding lookup matrix. + output_attentions (`bool`, *optional*): + Whether or not to return the attentions tensors of all attention layers. See `attentions` under returned + tensors for more detail. + output_hidden_states (`bool`, *optional*): + Whether or not to return the hidden states of all layers. See `hidden_states` under returned tensors for + more detail. + return_dict (`bool`, *optional*): + Whether or not to return a [`~utils.ModelOutput`] instead of a plain tuple. +""" + + +@add_start_docstrings( + "The bare Tapas Model transformer outputting raw hidden-states without any specific head on top.", + TAPAS_START_DOCSTRING, +) +class TapasModel(TapasPreTrainedModel): + """ + This class is a small change compared to [`BertModel`], taking into account the additional token type ids. + + The model can behave as an encoder (with only self-attention) as well as a decoder, in which case a layer of + cross-attention is added between the self-attention layers, following the architecture described in [Attention is + all you need](https://arxiv.org/abs/1706.03762) by Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, + Llion Jones, Aidan N. Gomez, Lukasz Kaiser and Illia Polosukhin. + + """ + + def __init__(self, config, add_pooling_layer=True): + super().__init__(config) + self.config = config + + self.embeddings = TapasEmbeddings(config) + self.encoder = TapasEncoder(config) + + self.pooler = TapasPooler(config) if add_pooling_layer else None + + # Initialize weights and apply final processing + self.post_init() + + def get_input_embeddings(self): + return self.embeddings.word_embeddings + + def set_input_embeddings(self, value): + self.embeddings.word_embeddings = value + + def _prune_heads(self, heads_to_prune): + """ + Prunes heads of the model. heads_to_prune: dict of {layer_num: list of heads to prune in this layer} See base + class PreTrainedModel + """ + for layer, heads in heads_to_prune.items(): + self.encoder.layer[layer].attention.prune_heads(heads) + + @add_start_docstrings_to_model_forward(TAPAS_INPUTS_DOCSTRING.format("batch_size, sequence_length")) + @replace_return_docstrings(output_type=BaseModelOutputWithPooling, config_class=_CONFIG_FOR_DOC) + def forward( + self, + input_ids: Optional[torch.LongTensor] = None, + attention_mask: Optional[torch.FloatTensor] = None, + token_type_ids: Optional[torch.LongTensor] = None, + position_ids: Optional[torch.LongTensor] = None, + head_mask: Optional[torch.FloatTensor] = None, + inputs_embeds: Optional[torch.FloatTensor] = None, + encoder_hidden_states: Optional[torch.FloatTensor] = None, + encoder_attention_mask: Optional[torch.FloatTensor] = None, + output_attentions: Optional[bool] = None, + output_hidden_states: Optional[bool] = None, + return_dict: Optional[bool] = None, + ) -> Union[Tuple, BaseModelOutputWithPooling]: + r""" + Returns: + + Examples: + + ```python + >>> from transformers import AutoTokenizer, TapasModel + >>> import pandas as pd + + >>> tokenizer = AutoTokenizer.from_pretrained("google/tapas-base") + >>> model = TapasModel.from_pretrained("google/tapas-base") + + >>> data = { + ... "Actors": ["Brad Pitt", "Leonardo Di Caprio", "George Clooney"], + ... "Age": ["56", "45", "59"], + ... "Number of movies": ["87", "53", "69"], + ... } + >>> table = pd.DataFrame.from_dict(data) + >>> queries = ["How many movies has George Clooney played in?", "How old is Brad Pitt?"] + + >>> inputs = tokenizer(table=table, queries=queries, padding="max_length", return_tensors="pt") + >>> outputs = model(**inputs) + + >>> last_hidden_states = outputs.last_hidden_state + ```""" + output_attentions = output_attentions if output_attentions is not None else self.config.output_attentions + output_hidden_states = ( + output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states + ) + return_dict = return_dict if return_dict is not None else self.config.use_return_dict + + if input_ids is not None and inputs_embeds is not None: + raise ValueError("You cannot specify both input_ids and inputs_embeds at the same time") + elif input_ids is not None: + self.warn_if_padding_and_no_attention_mask(input_ids, attention_mask) + input_shape = input_ids.size() + elif inputs_embeds is not None: + input_shape = inputs_embeds.size()[:-1] + else: + raise ValueError("You have to specify either input_ids or inputs_embeds") + + device = input_ids.device if input_ids is not None else inputs_embeds.device + + if attention_mask is None: + attention_mask = torch.ones(input_shape, device=device) + if token_type_ids is None: + token_type_ids = torch.zeros( + (*input_shape, len(self.config.type_vocab_sizes)), dtype=torch.long, device=device + ) + + # We can provide a self-attention mask of dimensions [batch_size, from_seq_length, to_seq_length] + # ourselves in which case we just need to make it broadcastable to all heads. + extended_attention_mask: torch.Tensor = self.get_extended_attention_mask(attention_mask, input_shape) + + # If a 2D ou 3D attention mask is provided for the cross-attention + # we need to make broadcastabe to [batch_size, num_heads, seq_length, seq_length] + if self.config.is_decoder and encoder_hidden_states is not None: + encoder_batch_size, encoder_sequence_length, _ = encoder_hidden_states.size() + encoder_hidden_shape = (encoder_batch_size, encoder_sequence_length) + if encoder_attention_mask is None: + encoder_attention_mask = torch.ones(encoder_hidden_shape, device=device) + encoder_extended_attention_mask = self.invert_attention_mask(encoder_attention_mask) + else: + encoder_extended_attention_mask = None + + # Prepare head mask if needed + # 1.0 in head_mask indicate we keep the head + # attention_probs has shape bsz x n_heads x N x N + # input head_mask has shape [num_heads] or [num_hidden_layers x num_heads] + # and head_mask is converted to shape [num_hidden_layers x batch x num_heads x seq_length x seq_length] + head_mask = self.get_head_mask(head_mask, self.config.num_hidden_layers) + + embedding_output = self.embeddings( + input_ids=input_ids, position_ids=position_ids, token_type_ids=token_type_ids, inputs_embeds=inputs_embeds + ) + encoder_outputs = self.encoder( + embedding_output, + attention_mask=extended_attention_mask, + head_mask=head_mask, + encoder_hidden_states=encoder_hidden_states, + encoder_attention_mask=encoder_extended_attention_mask, + output_attentions=output_attentions, + output_hidden_states=output_hidden_states, + return_dict=return_dict, + ) + sequence_output = encoder_outputs[0] + pooled_output = self.pooler(sequence_output) if self.pooler is not None else None + + if not return_dict: + return (sequence_output, pooled_output) + encoder_outputs[1:] + + return BaseModelOutputWithPooling( + last_hidden_state=sequence_output, + pooler_output=pooled_output, + hidden_states=encoder_outputs.hidden_states, + attentions=encoder_outputs.attentions, + ) + + +@add_start_docstrings("""Tapas Model with a `language modeling` head on top.""", TAPAS_START_DOCSTRING) +class TapasForMaskedLM(TapasPreTrainedModel): + _tied_weights_keys = ["cls.predictions.decoder.weight", "cls.predictions.decoder.bias"] + config_class = TapasConfig + base_model_prefix = "tapas" + + def __init__(self, config): + super().__init__(config) + + self.tapas = TapasModel(config, add_pooling_layer=False) + self.cls = TapasOnlyMLMHead(config) + + # Initialize weights and apply final processing + self.post_init() + + def get_output_embeddings(self): + return self.cls.predictions.decoder + + def set_output_embeddings(self, new_embeddings): + self.cls.predictions.decoder = new_embeddings + self.cls.predictions.bias = new_embeddings.bias + + @add_start_docstrings_to_model_forward(TAPAS_INPUTS_DOCSTRING.format("batch_size, sequence_length")) + @replace_return_docstrings(output_type=MaskedLMOutput, config_class=_CONFIG_FOR_DOC) + def forward( + self, + input_ids: Optional[torch.LongTensor] = None, + attention_mask: Optional[torch.FloatTensor] = None, + token_type_ids: Optional[torch.LongTensor] = None, + position_ids: Optional[torch.LongTensor] = None, + head_mask: Optional[torch.FloatTensor] = None, + inputs_embeds: Optional[torch.FloatTensor] = None, + encoder_hidden_states: Optional[torch.FloatTensor] = None, + encoder_attention_mask: Optional[torch.FloatTensor] = None, + labels: Optional[torch.LongTensor] = None, + output_attentions: Optional[bool] = None, + output_hidden_states: Optional[bool] = None, + return_dict: Optional[bool] = None, + **kwargs, + ) -> Union[Tuple, MaskedLMOutput]: + r""" + labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*): + Labels for computing the masked language modeling loss. Indices should be in `[-100, 0, ..., + config.vocab_size]` (see `input_ids` docstring) Tokens with indices set to `-100` are ignored (masked), the + loss is only computed for the tokens with labels in `[0, ..., config.vocab_size]` + + Returns: + + Examples: + + ```python + >>> from transformers import AutoTokenizer, TapasForMaskedLM + >>> import pandas as pd + + >>> tokenizer = AutoTokenizer.from_pretrained("google/tapas-base") + >>> model = TapasForMaskedLM.from_pretrained("google/tapas-base") + + >>> data = { + ... "Actors": ["Brad Pitt", "Leonardo Di Caprio", "George Clooney"], + ... "Age": ["56", "45", "59"], + ... "Number of movies": ["87", "53", "69"], + ... } + >>> table = pd.DataFrame.from_dict(data) + + >>> inputs = tokenizer( + ... table=table, queries="How many [MASK] has George [MASK] played in?", return_tensors="pt" + ... ) + >>> labels = tokenizer( + ... table=table, queries="How many movies has George Clooney played in?", return_tensors="pt" + ... )["input_ids"] + + >>> outputs = model(**inputs, labels=labels) + >>> logits = outputs.logits + ```""" + return_dict = return_dict if return_dict is not None else self.config.use_return_dict + + outputs = self.tapas( + input_ids, + attention_mask=attention_mask, + token_type_ids=token_type_ids, + position_ids=position_ids, + head_mask=head_mask, + inputs_embeds=inputs_embeds, + encoder_hidden_states=encoder_hidden_states, + encoder_attention_mask=encoder_attention_mask, + output_attentions=output_attentions, + output_hidden_states=output_hidden_states, + return_dict=return_dict, + ) + + sequence_output = outputs[0] + prediction_scores = self.cls(sequence_output) + + masked_lm_loss = None + if labels is not None: + loss_fct = CrossEntropyLoss() # -100 index = padding token + masked_lm_loss = loss_fct(prediction_scores.view(-1, self.config.vocab_size), labels.view(-1)) + + if not return_dict: + output = (prediction_scores,) + outputs[2:] + return ((masked_lm_loss,) + output) if masked_lm_loss is not None else output + + return MaskedLMOutput( + loss=masked_lm_loss, + logits=prediction_scores, + hidden_states=outputs.hidden_states, + attentions=outputs.attentions, + ) + + +@add_start_docstrings( + """ + Tapas Model with a cell selection head and optional aggregation head on top for question-answering tasks on tables + (linear layers on top of the hidden-states output to compute `logits` and optional `logits_aggregation`), e.g. for + SQA, WTQ or WikiSQL-supervised tasks. + """, + TAPAS_START_DOCSTRING, +) +class TapasForQuestionAnswering(TapasPreTrainedModel): + def __init__(self, config: TapasConfig): + super().__init__(config) + + # base model + self.tapas = TapasModel(config) + + # dropout (only used when training) + self.dropout = nn.Dropout(config.hidden_dropout_prob) + + # cell selection heads + if config.init_cell_selection_weights_to_zero: + # init_cell_selection_weights_to_zero: Whether the initial weights should be + # set to 0. This ensures that all tokens have the same prior probability. + self.output_weights = nn.Parameter(torch.zeros(config.hidden_size)) + self.column_output_weights = nn.Parameter(torch.zeros(config.hidden_size)) + else: + self.output_weights = nn.Parameter(torch.empty(config.hidden_size)) + nn.init.normal_( + self.output_weights, std=config.initializer_range + ) # here, a truncated normal is used in the original implementation + self.column_output_weights = nn.Parameter(torch.empty(config.hidden_size)) + nn.init.normal_( + self.column_output_weights, std=config.initializer_range + ) # here, a truncated normal is used in the original implementation + self.output_bias = nn.Parameter(torch.zeros([])) + self.column_output_bias = nn.Parameter(torch.zeros([])) + + # aggregation head + if config.num_aggregation_labels > 0: + self.aggregation_classifier = nn.Linear(config.hidden_size, config.num_aggregation_labels) + + # Initialize weights and apply final processing + self.post_init() + + @add_start_docstrings_to_model_forward(TAPAS_INPUTS_DOCSTRING.format("batch_size, sequence_length")) + @replace_return_docstrings(output_type=TableQuestionAnsweringOutput, config_class=_CONFIG_FOR_DOC) + def forward( + self, + input_ids: Optional[torch.LongTensor] = None, + attention_mask: Optional[torch.FloatTensor] = None, + token_type_ids: Optional[torch.LongTensor] = None, + position_ids: Optional[torch.LongTensor] = None, + head_mask: Optional[torch.FloatTensor] = None, + inputs_embeds: Optional[torch.FloatTensor] = None, + table_mask: Optional[torch.LongTensor] = None, + labels: Optional[torch.LongTensor] = None, + aggregation_labels: Optional[torch.LongTensor] = None, + float_answer: Optional[torch.FloatTensor] = None, + numeric_values: Optional[torch.FloatTensor] = None, + numeric_values_scale: Optional[torch.FloatTensor] = None, + output_attentions: Optional[bool] = None, + output_hidden_states: Optional[bool] = None, + return_dict: Optional[bool] = None, + ) -> Union[Tuple, TableQuestionAnsweringOutput]: + r""" + table_mask (`torch.LongTensor` of shape `(batch_size, seq_length)`, *optional*): + Mask for the table. Indicates which tokens belong to the table (1). Question tokens, table headers and + padding are 0. + labels (`torch.LongTensor` of shape `(batch_size, seq_length)`, *optional*): + Labels per token for computing the hierarchical cell selection loss. This encodes the positions of the + answer appearing in the table. Can be obtained using [`AutoTokenizer`]. + + - 1 for tokens that are **part of the answer**, + - 0 for tokens that are **not part of the answer**. + + aggregation_labels (`torch.LongTensor` of shape `(batch_size, )`, *optional*): + Aggregation function index for every example in the batch for computing the aggregation loss. Indices + should be in `[0, ..., config.num_aggregation_labels - 1]`. Only required in case of strong supervision for + aggregation (WikiSQL-supervised). + float_answer (`torch.FloatTensor` of shape `(batch_size, )`, *optional*): + Float answer for every example in the batch. Set to *float('nan')* for cell selection questions. Only + required in case of weak supervision (WTQ) to calculate the aggregate mask and regression loss. + numeric_values (`torch.FloatTensor` of shape `(batch_size, seq_length)`, *optional*): + Numeric values of every token, NaN for tokens which are not numeric values. Can be obtained using + [`AutoTokenizer`]. Only required in case of weak supervision for aggregation (WTQ) to calculate the + regression loss. + numeric_values_scale (`torch.FloatTensor` of shape `(batch_size, seq_length)`, *optional*): + Scale of the numeric values of every token. Can be obtained using [`AutoTokenizer`]. Only required in case + of weak supervision for aggregation (WTQ) to calculate the regression loss. + + Returns: + + Examples: + + ```python + >>> from transformers import AutoTokenizer, TapasForQuestionAnswering + >>> import pandas as pd + + >>> tokenizer = AutoTokenizer.from_pretrained("google/tapas-base-finetuned-wtq") + >>> model = TapasForQuestionAnswering.from_pretrained("google/tapas-base-finetuned-wtq") + + >>> data = { + ... "Actors": ["Brad Pitt", "Leonardo Di Caprio", "George Clooney"], + ... "Age": ["56", "45", "59"], + ... "Number of movies": ["87", "53", "69"], + ... } + >>> table = pd.DataFrame.from_dict(data) + >>> queries = ["How many movies has George Clooney played in?", "How old is Brad Pitt?"] + + >>> inputs = tokenizer(table=table, queries=queries, padding="max_length", return_tensors="pt") + >>> outputs = model(**inputs) + + >>> logits = outputs.logits + >>> logits_aggregation = outputs.logits_aggregation + ```""" + return_dict = return_dict if return_dict is not None else self.config.use_return_dict + + outputs = self.tapas( + input_ids, + attention_mask=attention_mask, + token_type_ids=token_type_ids, + position_ids=position_ids, + head_mask=head_mask, + inputs_embeds=inputs_embeds, + output_attentions=output_attentions, + output_hidden_states=output_hidden_states, + return_dict=return_dict, + ) + + sequence_output = outputs[0] + pooled_output = outputs[1] + + sequence_output = self.dropout(sequence_output) + + if input_ids is not None: + input_shape = input_ids.size() + else: + input_shape = inputs_embeds.size()[:-1] + + device = input_ids.device if input_ids is not None else inputs_embeds.device + + # Construct indices for the table. + if token_type_ids is None: + token_type_ids = torch.zeros( + (*input_shape, len(self.config.type_vocab_sizes)), dtype=torch.long, device=device + ) + + token_types = [ + "segment_ids", + "column_ids", + "row_ids", + "prev_labels", + "column_ranks", + "inv_column_ranks", + "numeric_relations", + ] + + row_ids = token_type_ids[:, :, token_types.index("row_ids")] + column_ids = token_type_ids[:, :, token_types.index("column_ids")] + + row_index = IndexMap( + indices=torch.min(row_ids, torch.as_tensor(self.config.max_num_rows - 1, device=row_ids.device)), + num_segments=self.config.max_num_rows, + batch_dims=1, + ) + col_index = IndexMap( + indices=torch.min(column_ids, torch.as_tensor(self.config.max_num_columns - 1, device=column_ids.device)), + num_segments=self.config.max_num_columns, + batch_dims=1, + ) + cell_index = ProductIndexMap(row_index, col_index) + + # Masks. + input_shape = input_ids.size() if input_ids is not None else inputs_embeds.size()[:-1] + device = input_ids.device if input_ids is not None else inputs_embeds.device + if attention_mask is None: + attention_mask = torch.ones(input_shape, device=device) + # Table cells only, without question tokens and table headers. + if table_mask is None: + table_mask = torch.where(row_ids > 0, torch.ones_like(row_ids), torch.zeros_like(row_ids)) + # torch.FloatTensor[batch_size, seq_length] + input_mask_float = attention_mask.float().to(device) + table_mask_float = table_mask.float().to(device) + # Mask for cells that exist in the table (i.e. that are not padding). + cell_mask, _ = reduce_mean(input_mask_float, cell_index) + + # Compute logits per token. These are used to select individual cells. + logits = compute_token_logits(sequence_output, self.config.temperature, self.output_weights, self.output_bias) + + # Compute logits per column. These are used to select a column. + column_logits = None + if self.config.select_one_column: + column_logits = compute_column_logits( + sequence_output, + self.column_output_weights, + self.column_output_bias, + cell_index, + cell_mask, + self.config.allow_empty_column_selection, + ) + + # Aggregation logits + logits_aggregation = None + if self.config.num_aggregation_labels > 0: + logits_aggregation = self.aggregation_classifier(pooled_output) + + # Total loss calculation + total_loss = 0.0 + calculate_loss = False + if labels is not None: + calculate_loss = True + is_supervised = not self.config.num_aggregation_labels > 0 or not self.config.use_answer_as_supervision + + # Semi-supervised cell selection in case of no aggregation: + # If the answer (the denotation) appears directly in the table we might + # select the answer without applying any aggregation function. There are + # some ambiguous cases, see utils._calculate_aggregate_mask for more info. + # `aggregate_mask` is 1 for examples where we chose to aggregate and 0 + # for examples where we chose to select the answer directly. + # `labels` encodes the positions of the answer appearing in the table. + if is_supervised: + aggregate_mask = None + else: + if float_answer is not None: + assert ( + labels.shape[0] == float_answer.shape[0] + ), "Make sure the answers are a FloatTensor of shape (batch_size,)" + # [batch_size] + aggregate_mask = _calculate_aggregate_mask( + float_answer, + pooled_output, + self.config.cell_selection_preference, + labels, + self.aggregation_classifier, + ) + else: + raise ValueError("You have to specify float answers in order to calculate the aggregate mask") + + # Cell selection log-likelihood + if self.config.average_logits_per_cell: + logits_per_cell, _ = reduce_mean(logits, cell_index) + logits = gather(logits_per_cell, cell_index) + dist_per_token = torch.distributions.Bernoulli(logits=logits) + + # Compute cell selection loss per example. + selection_loss_per_example = None + if not self.config.select_one_column: + weight = torch.where( + labels == 0, + torch.ones_like(labels, dtype=torch.float32), + self.config.positive_label_weight * torch.ones_like(labels, dtype=torch.float32), + ) + selection_loss_per_token = -dist_per_token.log_prob(labels) * weight + selection_loss_per_example = torch.sum(selection_loss_per_token * input_mask_float, dim=1) / ( + torch.sum(input_mask_float, dim=1) + EPSILON_ZERO_DIVISION + ) + else: + selection_loss_per_example, logits = _single_column_cell_selection_loss( + logits, column_logits, labels, cell_index, col_index, cell_mask + ) + dist_per_token = torch.distributions.Bernoulli(logits=logits) + + # Supervised cell selection + if self.config.disable_per_token_loss: + pass + elif is_supervised: + total_loss += torch.mean(selection_loss_per_example) + else: + # For the not supervised case, do not assign loss for cell selection + total_loss += torch.mean(selection_loss_per_example * (1.0 - aggregate_mask)) + + # Semi-supervised regression loss and supervised loss for aggregations + if self.config.num_aggregation_labels > 0: + if is_supervised: + # Note that `aggregate_mask` is None if the setting is supervised. + if aggregation_labels is not None: + assert ( + labels.shape[0] == aggregation_labels.shape[0] + ), "Make sure the aggregation labels are a LongTensor of shape (batch_size,)" + per_example_additional_loss = _calculate_aggregation_loss( + logits_aggregation, + aggregate_mask, + aggregation_labels, + self.config.use_answer_as_supervision, + self.config.num_aggregation_labels, + self.config.aggregation_loss_weight, + ) + else: + raise ValueError( + "You have to specify aggregation labels in order to calculate the aggregation loss" + ) + else: + # Set aggregation labels to zeros + aggregation_labels = torch.zeros(labels.shape[0], dtype=torch.long, device=labels.device) + per_example_additional_loss = _calculate_aggregation_loss( + logits_aggregation, + aggregate_mask, + aggregation_labels, + self.config.use_answer_as_supervision, + self.config.num_aggregation_labels, + self.config.aggregation_loss_weight, + ) + + if self.config.use_answer_as_supervision: + if numeric_values is not None and numeric_values_scale is not None: + assert numeric_values.shape == numeric_values_scale.shape + # Add regression loss for numeric answers which require aggregation. + answer_loss, large_answer_loss_mask = _calculate_regression_loss( + float_answer, + aggregate_mask, + dist_per_token, + numeric_values, + numeric_values_scale, + table_mask_float, + logits_aggregation, + self.config, + ) + per_example_additional_loss += answer_loss + # Zero loss for examples with answer_loss > cutoff. + per_example_additional_loss *= large_answer_loss_mask + else: + raise ValueError( + "You have to specify numeric values and numeric values scale in order to calculate the" + " regression loss" + ) + + total_loss += torch.mean(per_example_additional_loss) + + else: + # if no label ids are provided, set them to zeros in order to properly compute logits + labels = torch.zeros_like(logits) + _, logits = _single_column_cell_selection_loss( + logits, column_logits, labels, cell_index, col_index, cell_mask + ) + if not return_dict: + output = (logits, logits_aggregation) + outputs[2:] + return ((total_loss,) + output) if calculate_loss else output + + return TableQuestionAnsweringOutput( + loss=total_loss if calculate_loss else None, + logits=logits, + logits_aggregation=logits_aggregation, + hidden_states=outputs.hidden_states, + attentions=outputs.attentions, + ) + + +@add_start_docstrings( + """ + Tapas Model with a sequence classification head on top (a linear layer on top of the pooled output), e.g. for table + entailment tasks, such as TabFact (Chen et al., 2020). + """, + TAPAS_START_DOCSTRING, +) +class TapasForSequenceClassification(TapasPreTrainedModel): + def __init__(self, config): + super().__init__(config) + self.num_labels = config.num_labels + + self.tapas = TapasModel(config) + self.dropout = nn.Dropout(config.hidden_dropout_prob) + self.classifier = nn.Linear(config.hidden_size, config.num_labels) + + # Initialize weights and apply final processing + self.post_init() + + @add_start_docstrings_to_model_forward(TAPAS_INPUTS_DOCSTRING.format("batch_size, sequence_length")) + @replace_return_docstrings(output_type=SequenceClassifierOutput, config_class=_CONFIG_FOR_DOC) + def forward( + self, + input_ids: Optional[torch.LongTensor] = None, + attention_mask: Optional[torch.FloatTensor] = None, + token_type_ids: Optional[torch.LongTensor] = None, + position_ids: Optional[torch.LongTensor] = None, + head_mask: Optional[torch.FloatTensor] = None, + inputs_embeds: Optional[torch.FloatTensor] = None, + labels: Optional[torch.LongTensor] = None, + output_attentions: Optional[bool] = None, + output_hidden_states: Optional[bool] = None, + return_dict: Optional[bool] = None, + ) -> Union[Tuple[torch.Tensor], SequenceClassifierOutput]: + r""" + labels (`torch.LongTensor` of shape `(batch_size,)`, *optional*): + Labels for computing the sequence classification/regression loss. Indices should be in `[0, ..., + config.num_labels - 1]`. If `config.num_labels == 1` a regression loss is computed (Mean-Square loss), If + `config.num_labels > 1` a classification loss is computed (Cross-Entropy). Note: this is called + "classification_class_index" in the original implementation. + + Returns: + + Examples: + + ```python + >>> from transformers import AutoTokenizer, TapasForSequenceClassification + >>> import torch + >>> import pandas as pd + + >>> tokenizer = AutoTokenizer.from_pretrained("google/tapas-base-finetuned-tabfact") + >>> model = TapasForSequenceClassification.from_pretrained("google/tapas-base-finetuned-tabfact") + + >>> data = { + ... "Actors": ["Brad Pitt", "Leonardo Di Caprio", "George Clooney"], + ... "Age": ["56", "45", "59"], + ... "Number of movies": ["87", "53", "69"], + ... } + >>> table = pd.DataFrame.from_dict(data) + >>> queries = [ + ... "There is only one actor who is 45 years old", + ... "There are 3 actors which played in more than 60 movies", + ... ] + + >>> inputs = tokenizer(table=table, queries=queries, padding="max_length", return_tensors="pt") + >>> labels = torch.tensor([1, 0]) # 1 means entailed, 0 means refuted + + >>> outputs = model(**inputs, labels=labels) + >>> loss = outputs.loss + >>> logits = outputs.logits + ```""" + return_dict = return_dict if return_dict is not None else self.config.use_return_dict + + outputs = self.tapas( + input_ids, + attention_mask=attention_mask, + token_type_ids=token_type_ids, + position_ids=position_ids, + head_mask=head_mask, + inputs_embeds=inputs_embeds, + output_attentions=output_attentions, + output_hidden_states=output_hidden_states, + return_dict=return_dict, + ) + + pooled_output = outputs[1] + + pooled_output = self.dropout(pooled_output) + logits = self.classifier(pooled_output) + + loss = None + if labels is not None: + if self.config.problem_type is None: + if self.num_labels == 1: + self.config.problem_type = "regression" + elif self.num_labels > 1 and (labels.dtype == torch.long or labels.dtype == torch.int): + self.config.problem_type = "single_label_classification" + else: + self.config.problem_type = "multi_label_classification" + + if self.config.problem_type == "regression": + loss_fct = MSELoss() + if self.num_labels == 1: + loss = loss_fct(logits.squeeze(), labels.squeeze()) + else: + loss = loss_fct(logits, labels) + elif self.config.problem_type == "single_label_classification": + loss_fct = CrossEntropyLoss() + loss = loss_fct(logits.view(-1, self.num_labels), labels.view(-1)) + elif self.config.problem_type == "multi_label_classification": + loss_fct = BCEWithLogitsLoss() + loss = loss_fct(logits, labels) + if not return_dict: + output = (logits,) + outputs[2:] + return ((loss,) + output) if loss is not None else output + + return SequenceClassifierOutput( + loss=loss, + logits=logits, + hidden_states=outputs.hidden_states, + attentions=outputs.attentions, + ) + + +""" TAPAS utilities.""" + + +class AverageApproximationFunction(str, enum.Enum): + RATIO = "ratio" + FIRST_ORDER = "first_order" + SECOND_ORDER = "second_order" + + +# Beginning of everything related to segmented tensors + + +class IndexMap(object): + """Index grouping entries within a tensor.""" + + def __init__(self, indices, num_segments, batch_dims=0): + """ + Creates an index + + Args: + indices (`torch.LongTensor`, same shape as a *values* Tensor to which the indices refer): + Tensor containing the indices. + num_segments (`torch.LongTensor`): + Scalar tensor, the number of segments. All elements in a batched segmented tensor must have the same + number of segments (although many segments can be empty). + batch_dims (`int`, *optional*, defaults to 0): + The number of batch dimensions. The first *batch_dims* dimensions of a SegmentedTensor are treated as + batch dimensions. Segments in different batch elements are always distinct even if they have the same + index. + """ + self.indices = torch.as_tensor(indices) + self.num_segments = torch.as_tensor(num_segments, device=indices.device) + self.batch_dims = batch_dims + + def batch_shape(self): + return self.indices.size()[: self.batch_dims] # returns a torch.Size object + + +class ProductIndexMap(IndexMap): + """The product of two indices.""" + + def __init__(self, outer_index, inner_index): + """ + Combines indices i and j into pairs (i, j). The result is an index where each segment (i, j) is the + intersection of segments i and j. For example if the inputs represent table cells indexed by respectively rows + and columns the output will be a table indexed by (row, column) pairs, i.e. by cell. The implementation + combines indices {0, .., n - 1} and {0, .., m - 1} into {0, .., nm - 1}. The output has *num_segments* equal to + *outer_index.num_segments* * *inner_index.num_segments* + + Args: + outer_index (`IndexMap`): + IndexMap. + inner_index (`IndexMap`): + IndexMap, must have the same shape as *outer_index*. + """ + if outer_index.batch_dims != inner_index.batch_dims: + raise ValueError("outer_index.batch_dims and inner_index.batch_dims must be the same.") + + super().__init__( + indices=(inner_index.indices + outer_index.indices * inner_index.num_segments), + num_segments=inner_index.num_segments * outer_index.num_segments, + batch_dims=inner_index.batch_dims, + ) + self.outer_index = outer_index + self.inner_index = inner_index + + def project_outer(self, index): + """Projects an index with the same index set onto the outer components.""" + indices = torch.div(index.indices, self.inner_index.num_segments, rounding_mode="floor").type(torch.long) + return IndexMap(indices=indices, num_segments=self.outer_index.num_segments, batch_dims=index.batch_dims) + + def project_inner(self, index): + """Projects an index with the same index set onto the inner components.""" + return IndexMap( + indices=torch.fmod(index.indices, self.inner_index.num_segments) + .type(torch.float) + .floor() + .type(torch.long), + num_segments=self.inner_index.num_segments, + batch_dims=index.batch_dims, + ) + + +def gather(values, index, name="segmented_gather"): + """ + Gathers from *values* using the index map. For each element in the domain of the index map this operation looks up + a value for that index in *values*. Two elements from the same segment always get assigned the same value. + + Args: + values (`torch.Tensor` of shape (B1, ..., Bn, num_segments, V1, ...)): + Tensor with segment values. + index (`IndexMap` of shape (B1, ..., Bn, I1, ..., Ik)): + IndexMap. + name (`str`, *optional*, defaults to 'segmented_gather'): + Name for the operation. Currently not used + + Returns: + `tuple(torch.Tensor)`: Tensor of shape (B1, ..., Bn, I1, ..., Ik, V1, ...) with the gathered values. + """ + indices = index.indices + # first, check whether the indices of the index represent scalar values (i.e. not vectorized) + if len(values.shape[index.batch_dims :]) < 2: + return torch.gather( + values, + index.batch_dims, + indices.view( + values.size()[0], -1 + ), # torch.gather expects index to have the same number of dimensions as values + ).view(indices.size()) + else: + # this means we have a vectorized version + # we have to adjust the index + indices = indices.unsqueeze(-1).expand(values.shape) + return torch.gather(values, index.batch_dims, indices) + + +def flatten(index, name="segmented_flatten"): + """ + Flattens a batched index map (which is typically of shape batch_size, seq_length) to a 1d index map. This operation + relabels the segments to keep batch elements distinct. The k-th batch element will have indices shifted by + *num_segments* * (k - 1). The result is a tensor with *num_segments* multiplied by the number of elements in the + batch. + + Args: + index (`IndexMap`): + IndexMap to flatten. + name (`str`, *optional*, defaults to 'segmented_flatten'): + Name for the operation. Currently not used + + Returns: + (`IndexMap`): The flattened IndexMap. + """ + # first, get batch_size as scalar tensor + batch_size = torch.prod(torch.tensor(list(index.batch_shape()))) + # next, create offset as 1-D tensor of length batch_size, + # and multiply element-wise by num segments (to offset different elements in the batch) e.g. if batch size is 2: [0, 64] + offset = torch.arange(start=0, end=batch_size, device=index.num_segments.device) * index.num_segments + offset = offset.view(index.batch_shape()) + for _ in range(index.batch_dims, len(index.indices.size())): # typically range(1,2) + offset = offset.unsqueeze(-1) + + indices = offset + index.indices + return IndexMap(indices=indices.view(-1), num_segments=index.num_segments * batch_size, batch_dims=0) + + +def range_index_map(batch_shape, num_segments, name="range_index_map"): + """ + Constructs an index map equal to range(num_segments). + + Args: + batch_shape (`torch.Size`): + Batch shape + num_segments (`int`): + Number of segments + name (`str`, *optional*, defaults to 'range_index_map'): + Name for the operation. Currently not used + + Returns: + (`IndexMap`): IndexMap of shape batch_shape with elements equal to range(num_segments). + """ + batch_shape = torch.as_tensor( + batch_shape, dtype=torch.long + ) # create a rank 1 tensor vector containing batch_shape (e.g. [2]) + assert len(batch_shape.size()) == 1 + num_segments = torch.as_tensor(num_segments) # create a rank 0 tensor (scalar) containing num_segments (e.g. 64) + assert len(num_segments.size()) == 0 + + indices = torch.arange( + start=0, end=num_segments, device=num_segments.device + ) # create a rank 1 vector with num_segments elements + new_tensor = torch.cat( + [torch.ones_like(batch_shape, dtype=torch.long, device=num_segments.device), num_segments.unsqueeze(dim=0)], + dim=0, + ) + # new_tensor is just a vector of [1 64] for example (assuming only 1 batch dimension) + new_shape = [int(x) for x in new_tensor.tolist()] + indices = indices.view(new_shape) + + multiples = torch.cat([batch_shape, torch.as_tensor([1])], dim=0) + indices = indices.repeat(multiples.tolist()) + # equivalent (in Numpy:) + # indices = torch.as_tensor(np.tile(indices.numpy(), multiples.tolist())) + + return IndexMap(indices=indices, num_segments=num_segments, batch_dims=list(batch_shape.size())[0]) + + +def _segment_reduce(values, index, segment_reduce_fn, name): + """ + Applies a segment reduction segment-wise. + + Args: + values (`torch.Tensor`): + Tensor with segment values. + index (`IndexMap`): + IndexMap. + segment_reduce_fn (`str`): + Name for the reduce operation. One of "sum", "mean", "max" or "min". + name (`str`): + Name for the operation. Currently not used + + Returns: + (`IndexMap`): IndexMap of shape batch_shape with elements equal to range(num_segments). + """ + # Flatten the batch dimensions, as segments ops (scatter) do not support batching. + # However if `values` has extra dimensions to the right keep them + # unflattened. Segmented ops support vector-valued operations. + flat_index = flatten(index) + vector_shape = values.size()[len(index.indices.size()) :] # torch.Size object + flattened_shape = torch.cat( + [torch.as_tensor([-1], dtype=torch.long), torch.as_tensor(vector_shape, dtype=torch.long)], dim=0 + ) + # changed "view" by "reshape" in the following line + flat_values = values.reshape(flattened_shape.tolist()) + + out = torch.zeros(int(flat_index.num_segments), dtype=torch.float, device=flat_values.device) + segment_means = out.scatter_reduce( + dim=0, index=flat_index.indices.long(), src=flat_values.float(), reduce=segment_reduce_fn, include_self=False + ) + + # Unflatten the values. + new_shape = torch.cat( + [ + torch.as_tensor(index.batch_shape(), dtype=torch.long), + torch.as_tensor([index.num_segments], dtype=torch.long), + torch.as_tensor(vector_shape, dtype=torch.long), + ], + dim=0, + ) + + output_values = segment_means.clone().view(new_shape.tolist()).to(values.dtype) + output_index = range_index_map(index.batch_shape(), index.num_segments) + return output_values, output_index + + +def reduce_sum(values, index, name="segmented_reduce_sum"): + """ + Sums a tensor over its segments. + + Outputs 0 for empty segments. + + This operations computes the sum over segments, with support for: + + - Batching using the first dimensions [B1, B2, ..., Bn]. Each element in a batch can have different indices. + - Vectorization using the last dimension [V1, V2, ...]. If they are present, the output will be a sum of + vectors rather than scalars. Only the middle dimensions [I1, ..., Ik] are reduced by the operation. + + Args: + values (`torch.Tensor` of shape [B1, B2, ..., Bn, I1, .., Ik, V1, V2, ..]): + Tensor containing the values of which the sum must be taken segment-wise. + index (`IndexMap`, indices are of shape [B1, B2, ..., Bn, I1, .., Ik].): + Index defining the segments. + name (`str`, *optional*, defaults to 'segmented_reduce_sum'): + Name for the operation. Currently not used + + Returns: + output_values (`torch.Tensor`of shape [B1, B2, ..., Bn, num_segments, V1, V2, ..]): Tensor containing the + output values. output_index (`IndexMap`): IndexMap with shape [B1, B2, ..., Bn, num_segments]. . + """ + return _segment_reduce(values, index, "sum", name) + + +def reduce_mean(values, index, name="segmented_reduce_mean"): + """ + Averages a tensor over its segments. + + Outputs 0 for empty segments. + + This operations computes the mean over segments, with support for: + + - Batching using the first dimensions [B1, B2, ..., Bn]. Each element in a batch can have different indices. + - Vectorization using the last dimension [V1, V2, ...]. If they are present, the output will be a mean of + vectors rather than scalars. + + Only the middle dimensions [I1, ..., Ik] are reduced by the operation. + + Args: + values (`torch.Tensor` of shape [B1, B2, ..., Bn, I1, .., Ik, V1, V2, ..]): + Tensor containing the values of which the mean must be taken segment-wise. + index (`IndexMap`, indices are of shape [B1, B2, ..., Bn, I1, .., Ik].): + Index defining the segments. + name (`str`, *optional*, defaults to 'segmented_reduce_sum'): + Name for the operation. Currently not used + + Returns: + output_values (`torch.Tensor`of shape [B1, B2, ..., Bn, num_segments, V1, V2, ..]): Tensor containing the + output values. output_index (`IndexMap`): IndexMap with shape [B1, B2, ..., Bn, num_segments]. + """ + return _segment_reduce(values, index, "mean", name) + + +def reduce_max(values, index, name="segmented_reduce_max"): + """ + Computes the maximum over segments. + + This operation computes the maximum over segments, with support for: + + - Batching using the first dimensions [B1, B2, ..., Bn]. Each element in a batch can have different indices. + - Vectorization using the last dimension [V1, V2, ...]. If they are present, the output will be an element-wise + maximum of vectors rather than scalars. + + Only the middle dimensions [I1, ..., Ik] are reduced by the operation. + + Args: + values (`torch.Tensor` of shape [B1, B2, ..., Bn, I1, .., Ik, V1, V2, ..]): + Tensor containing the values of which the max must be taken segment-wise. + index (`IndexMap`, indices are of shape [B1, B2, ..., Bn, I1, .., Ik].): + Index defining the segments. + name (`str`, *optional*, defaults to 'segmented_reduce_sum'): + Name for the operation. Currently not used + + Returns: + output_values (`torch.Tensor`of shape [B1, B2, ..., Bn, num_segments, V1, V2, ..]): Tensor containing the + output values. output_index (`IndexMap`): IndexMap with shape [B1, B2, ..., Bn, num_segments]. + """ + return _segment_reduce(values, index, "amax", name) + + +def reduce_min(values, index, name="segmented_reduce_min"): + """ + Computes the minimum over segments. + + This operations computes the minimum over segments, with support for: + + - Batching using the first dimensions [B1, B2, ..., Bn]. Each element in a batch can have different indices. + - Vectorization using the last dimension [V1, V2, ...]. If they are present, the output will be an element-wise + minimum of vectors rather than scalars. + + Only the middle dimensions [I1, ..., Ik] are reduced by the operation. + + Args: + values (`torch.Tensor` of shape [B1, B2, ..., Bn, I1, .., Ik, V1, V2, ..]): + Tensor containing the values of which the min must be taken segment-wise. + index (`IndexMap`, indices are of shape [B1, B2, ..., Bn, I1, .., Ik].): + Index defining the segments. + name (`str`, *optional*, defaults to 'segmented_reduce_sum'): + Name for the operation. Currently not used + + Returns: + output_values (`torch.Tensor`of shape [B1, B2, ..., Bn, num_segments, V1, V2, ..]): Tensor containing the + output values. output_index (`IndexMap`): IndexMap with shape [B1, B2, ..., Bn, num_segments]. + """ + return _segment_reduce(values, index, "amin", name) + + +# End of everything related to segmented tensors + + +def compute_column_logits( + sequence_output, column_output_weights, column_output_bias, cell_index, cell_mask, allow_empty_column_selection +): + """ + Computes the column logits. + + Args: + sequence_output (`torch.FloatTensor` of shape `(batch_size, sequence_length, hidden_size)`): + Also known as last_hidden_state. Sequence of hidden-states at the output of the last layer of the model. + column_output_weights (`torch.FloatTensor` of shape `(hidden_size)`): + Weights of the linear layer for column selection. + column_output_bias (`torch.FloatTensor` of shape `()`): + Bias of the linear layer for column selection. + cell_index (`ProductIndexMap`): + Index that groups tokens into cells. + cell_mask (`torch.FloatTensor` of shape `(batch_size, max_num_rows * max_num_cols)`): + Mask for cells that exist in the table (i.e. that are not padding). + allow_empty_column_selection (`bool`): + Whether to allow not to select any column + + Returns: + column_logits (`torch.FloatTensor`of shape `(batch_size, max_num_cols)`): Tensor containing the column logits + for every example in the batch. + """ + + # First, compute the token logits (batch_size, seq_len) - without temperature + token_logits = torch.einsum("bsj,j->bs", sequence_output, column_output_weights) + column_output_bias + + # Next, average the logits per cell (batch_size, max_num_cols*max_num_rows) + cell_logits, cell_logits_index = reduce_mean(token_logits, cell_index) + + # Finally, average the logits per column (batch_size, max_num_cols) + column_index = cell_index.project_inner(cell_logits_index) + column_logits, out_index = reduce_sum(cell_logits * cell_mask, column_index) + + cell_count, _ = reduce_sum(cell_mask, column_index) + column_logits /= cell_count + EPSILON_ZERO_DIVISION + + # Mask columns that do not appear in the example. + is_padding = torch.logical_and(cell_count < 0.5, ~torch.eq(out_index.indices, 0)) + column_logits += CLOSE_ENOUGH_TO_LOG_ZERO * torch.as_tensor( + is_padding, dtype=torch.float32, device=is_padding.device + ) + + if not allow_empty_column_selection: + column_logits += CLOSE_ENOUGH_TO_LOG_ZERO * torch.as_tensor( + torch.eq(out_index.indices, 0), dtype=torch.float32, device=out_index.indices.device + ) + + return column_logits + + +def _single_column_cell_selection_loss(token_logits, column_logits, labels, cell_index, col_index, cell_mask): + """ + Computes the loss for cell selection constrained to a single column. The loss is a hierarchical log-likelihood. The + model first predicts a column and then selects cells within that column (conditioned on the column). Cells outside + the selected column are never selected. + + Args: + token_logits (`torch.FloatTensor` of shape `(batch_size, sequence_length)`): + Tensor containing the logits per token. + column_logits (`torch.FloatTensor` of shape `(batch_size, max_num_cols)`): + Tensor containing the logits per column. + labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`): + Labels per token. + cell_index (`ProductIndexMap`): + Index that groups tokens into cells. + col_index (`IndexMap`): + Index that groups tokens into columns. + cell_mask (`torch.FloatTensor` of shape `(batch_size, max_num_rows * max_num_cols)`): + Mask for cells that exist in the table (i.e. that are not padding). + + Returns: + selection_loss_per_example (`torch.FloatTensor` of shape `(batch_size,)`): Loss for each example. logits + (`torch.FloatTensor` of shape `(batch_size, sequence_length)`): New logits which are only allowed to select + cells in a single column. Logits outside of the most likely column according to *column_logits* will be set to + a very low value (such that the probabilities are 0). + """ + # Part 1: column loss + + # First find the column we should select. We use the column with maximum number of selected cells. + labels_per_column, _ = reduce_sum(torch.as_tensor(labels, dtype=torch.float32, device=labels.device), col_index) + # shape of labels_per_column is (batch_size, max_num_cols). It contains the number of label ids for every column, for every example + column_label = torch.argmax(labels_per_column, dim=-1) # shape (batch_size,) + # Check if there are no selected cells in the column. In that case the model + # should predict the special column id 0, which means "select nothing". + no_cell_selected = torch.eq( + torch.max(labels_per_column, dim=-1)[0], 0 + ) # no_cell_selected is of shape (batch_size,) and equals True + # if an example of the batch has no cells selected (i.e. if there are no labels set to 1 for that example) + column_label = torch.where( + no_cell_selected.view(column_label.size()), torch.zeros_like(column_label), column_label + ) + + column_dist = torch.distributions.Categorical(logits=column_logits) # shape (batch_size, max_num_cols) + column_loss_per_example = -column_dist.log_prob(column_label) + + # Part 2: cell loss + + # Reduce the labels and logits to per-cell from per-token. + # logits_per_cell: shape (batch_size, max_num_rows*max_num_cols) i.e. (batch_size, 64*32) + logits_per_cell, _ = reduce_mean(token_logits, cell_index) + # labels_per_cell: shape (batch_size, 64*32), indicating whether each cell should be selected (1) or not (0) + labels_per_cell, labels_index = reduce_max( + torch.as_tensor(labels, dtype=torch.long, device=labels.device), cell_index + ) + + # Mask for the selected column. + # column_id_for_cells: shape (batch_size, 64*32), indicating to which column each cell belongs + column_id_for_cells = cell_index.project_inner(labels_index).indices + # column_mask: shape (batch_size, 64*32), equal to 1 if cell belongs to column to be selected + column_mask = torch.as_tensor( + torch.eq(column_id_for_cells, torch.unsqueeze(column_label, dim=-1)), + dtype=torch.float32, + device=cell_mask.device, + ) + + # Compute the log-likelihood for cells, but only for the selected column. + cell_dist = torch.distributions.Bernoulli(logits=logits_per_cell) # shape (batch_size, 64*32) + cell_log_prob = cell_dist.log_prob(labels_per_cell.type(torch.float32)) # shape(batch_size, 64*32) + + cell_loss = -torch.sum(cell_log_prob * column_mask * cell_mask, dim=1) + + # We need to normalize the loss by the number of cells in the column. + cell_loss /= torch.sum(column_mask * cell_mask, dim=1) + EPSILON_ZERO_DIVISION + + selection_loss_per_example = column_loss_per_example + selection_loss_per_example += torch.where( + no_cell_selected.view(selection_loss_per_example.size()), + torch.zeros_like(selection_loss_per_example), + cell_loss, + ) + + # Set the probs outside the selected column (selected by the *model*) + # to 0. This ensures backwards compatibility with models that select + # cells from multiple columns. + selected_column_id = torch.as_tensor( + torch.argmax(column_logits, dim=-1), dtype=torch.long, device=column_logits.device + ) # shape (batch_size,) + + # selected_column_mask: shape (batch_size, 64*32), equal to 1 if cell belongs to column selected by the model + selected_column_mask = torch.as_tensor( + torch.eq(column_id_for_cells, torch.unsqueeze(selected_column_id, dim=-1)), + dtype=torch.float32, + device=selected_column_id.device, + ) + + # Never select cells with the special column id 0. + selected_column_mask = torch.where( + torch.eq(column_id_for_cells, 0).view(selected_column_mask.size()), + torch.zeros_like(selected_column_mask), + selected_column_mask, + ) + new_logits_per_cell = logits_per_cell + CLOSE_ENOUGH_TO_LOG_ZERO * (1.0 - cell_mask * selected_column_mask) + logits = gather(new_logits_per_cell, cell_index) + + return selection_loss_per_example, logits + + +def compute_token_logits(sequence_output, temperature, output_weights, output_bias): + """ + Computes logits per token + + Args: + sequence_output (`torch.FloatTensor` of shape `(batch_size, sequence_length, hidden_size)`): + Also known as last_hidden_state. Sequence of hidden-states at the output of the last layer of the model. + temperature (`float`): + Temperature for the Bernoulli distribution. + output_weights (`torch.FloatTensor` of shape `(hidden_size,)`): + Weights of the linear layer for cell selection. + output_bias (`torch.FloatTensor` of shape `()`): + Bias of the linear layer for cell selection + + Returns: + logits (`torch.FloatTensor` of shape `(batch_size, sequence_length)`): Logits per token. + """ + logits = (torch.einsum("bsj,j->bs", sequence_output, output_weights) + output_bias) / temperature + + return logits + + +def _calculate_aggregate_mask(answer, pooled_output, cell_selection_preference, labels, aggregation_classifier): + """ + Finds examples where the model should select cells with no aggregation. + + Returns a mask that determines for which examples should the model select answers directly from the table, without + any aggregation function. If the answer is a piece of text the case is unambiguous as aggregation functions only + apply to numbers. If the answer is a number but does not appear in the table then we must use some aggregation + case. The ambiguous case is when the answer is a number that also appears in the table. In this case we use the + aggregation function probabilities predicted by the model to decide whether to select or aggregate. The threshold + for this is a hyperparameter *cell_selection_preference* + + Args: + answer (`torch.FloatTensor` of shape `(batch_size, )`): + Answer for every example in the batch. Nan if there is no scalar answer. + pooled_output (`torch.FloatTensor` of shape `(batch_size, hidden_size)`): + Output of the pooler (BertPooler) on top of the encoder layer. + cell_selection_preference (`float`): + Preference for cell selection in ambiguous cases. + labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`): + Labels per token. aggregation_classifier (`torch.nn.Linear`): Aggregation head + + Returns: + aggregate_mask (`torch.FloatTensor` of shape `(batch_size,)`): A mask set to 1 for examples that should use + aggregation functions. + """ + # torch.FloatTensor(batch_size,) + aggregate_mask_init = torch.logical_not(torch.isnan(answer)).type(torch.FloatTensor).to(answer.device) + logits_aggregation = aggregation_classifier(pooled_output) + dist_aggregation = torch.distributions.categorical.Categorical(logits=logits_aggregation) + # Index 0 corresponds to "no aggregation". + aggregation_ops_total_mass = torch.sum(dist_aggregation.probs[:, 1:], dim=1) + + # Cell selection examples according to current model. + is_pred_cell_selection = aggregation_ops_total_mass <= cell_selection_preference + + # Examples with non-empty cell selection supervision. + is_cell_supervision_available = torch.sum(labels, dim=1) > 0 + + # torch.where is not equivalent to tf.where (in tensorflow 1) + # hence the added .view on the condition to match the shape of the first tensor + aggregate_mask = torch.where( + torch.logical_and(is_pred_cell_selection, is_cell_supervision_available).view(aggregate_mask_init.size()), + torch.zeros_like(aggregate_mask_init, dtype=torch.float32), + aggregate_mask_init, + ) + + aggregate_mask = aggregate_mask.detach() + + return aggregate_mask + + +def _calculate_aggregation_loss_known( + logits_aggregation, aggregate_mask, aggregation_labels, use_answer_as_supervision, num_aggregation_labels +): + """ + Calculates aggregation loss when its type is known during training. + + In the weakly supervised setting, the only known information is that for cell selection examples, "no aggregation" + should be predicted. For other examples (those that require aggregation), no loss is accumulated. In the setting + where aggregation type is always known, standard cross entropy loss is accumulated for all examples + + Args: + logits_aggregation (`torch.FloatTensor` of shape `(batch_size, num_aggregation_labels)`): + Logits per aggregation operation. + aggregate_mask (`torch.FloatTensor` of shape `(batch_size, )`): + A mask set to 1 for examples that should use aggregation functions. + aggregation_labels (`torch.LongTensor` of shape `(batch_size, )`): + Aggregation function id for every example in the batch. + use_answer_as_supervision (`bool`, *optional*): + Whether to use the answer as the only supervision for aggregation examples. + num_aggregation_labels (`int`, *optional*, defaults to 0): + The number of aggregation operators to predict. + + Returns: + aggregation_loss_known (`torch.FloatTensor` of shape `(batch_size,)`): Aggregation loss (when its type is known + during training) per example. + """ + if use_answer_as_supervision: + # Prepare "no aggregation" targets for cell selection examples. + target_aggregation = torch.zeros_like(aggregate_mask, dtype=torch.long) + else: + # Use aggregation supervision as the target. + target_aggregation = aggregation_labels + + one_hot_labels = nn.functional.one_hot(target_aggregation, num_classes=num_aggregation_labels).type(torch.float32) + log_probs = nn.functional.log_softmax(logits_aggregation, dim=-1) + + # torch.FloatTensor[batch_size] + per_example_aggregation_intermediate = -torch.sum(one_hot_labels * log_probs, dim=-1) + if use_answer_as_supervision: + # Accumulate loss only for examples requiring cell selection + # (no aggregation). + return per_example_aggregation_intermediate * (1 - aggregate_mask) + else: + return per_example_aggregation_intermediate + + +def _calculate_aggregation_loss_unknown(logits_aggregation, aggregate_mask): + """ + Calculates aggregation loss in the case of answer supervision. + + Args: + logits_aggregation (`torch.FloatTensor` of shape `(batch_size, num_aggregation_labels)`): + Logits per aggregation operation. + aggregate_mask (`torch.FloatTensor` of shape `(batch_size, )`): + A mask set to 1 for examples that should use aggregation functions + + Returns: + aggregation_loss_unknown (`torch.FloatTensor` of shape `(batch_size,)`): Aggregation loss (in case of answer + supervision) per example. + """ + dist_aggregation = torch.distributions.categorical.Categorical(logits=logits_aggregation) + # Index 0 corresponds to "no aggregation". + aggregation_ops_total_mass = torch.sum(dist_aggregation.probs[:, 1:], dim=1) + # Predict some aggregation in case of an answer that needs aggregation. + # This increases the probability of all aggregation functions, in a way + # similar to MML, but without considering whether the function gives the + # correct answer. + return -torch.log(aggregation_ops_total_mass) * aggregate_mask + + +def _calculate_aggregation_loss( + logits_aggregation, + aggregate_mask, + aggregation_labels, + use_answer_as_supervision, + num_aggregation_labels, + aggregation_loss_weight, +): + """ + Calculates the aggregation loss per example. + + Args: + logits_aggregation (`torch.FloatTensor` of shape `(batch_size, num_aggregation_labels)`): + Logits per aggregation operation. + aggregate_mask (`torch.FloatTensor` of shape `(batch_size, )`): + A mask set to 1 for examples that should use aggregation functions. + aggregation_labels (`torch.LongTensor` of shape `(batch_size, )`): + Aggregation function id for every example in the batch. + use_answer_as_supervision (`bool`, *optional*): + Whether to use the answer as the only supervision for aggregation examples. + num_aggregation_labels (`int`, *optional*, defaults to 0): + The number of aggregation operators to predict. + aggregation_loss_weight (`float`, *optional*, defaults to 1.0): + Importance weight for the aggregation loss. + + Returns: + aggregation_loss (`torch.FloatTensor` of shape `(batch_size,)`): Aggregation loss per example. + """ + per_example_aggregation_loss = _calculate_aggregation_loss_known( + logits_aggregation, aggregate_mask, aggregation_labels, use_answer_as_supervision, num_aggregation_labels + ) + + if use_answer_as_supervision: + # Add aggregation loss for numeric answers that need aggregation. + per_example_aggregation_loss += _calculate_aggregation_loss_unknown(logits_aggregation, aggregate_mask) + return aggregation_loss_weight * per_example_aggregation_loss + + +def _calculate_expected_result( + dist_per_cell, numeric_values, numeric_values_scale, input_mask_float, logits_aggregation, config +): + """ + Calculates the expected result given cell and aggregation probabilities. + + Args: + dist_per_cell (`torch.distributions.Bernoulli`): + Cell selection distribution for each cell. + numeric_values (`torch.FloatTensor` of shape `(batch_size, seq_length)`): + Numeric values of every token. Nan for tokens which are not numeric values. + numeric_values_scale (`torch.FloatTensor` of shape `(batch_size, seq_length)`): + Scale of the numeric values of every token. + input_mask_float (`torch.FloatTensor` of shape `(batch_size, seq_length)`): + Mask for the table, without question tokens and table headers. + logits_aggregation (`torch.FloatTensor` of shape `(batch_size, num_aggregation_labels)`): + Logits per aggregation operation. + config ([`TapasConfig`]): + Model configuration class with all the hyperparameters of the model + + Returns: + expected_result (`torch.FloatTensor` of shape `(batch_size,)`): The expected result per example. + """ + if config.use_gumbel_for_cells: + gumbel_dist = torch.distributions.RelaxedBernoulli( + # The token logits where already divided by the temperature and used for + # computing cell selection errors so we need to multiply it again here + temperature=config.temperature, + logits=dist_per_cell.logits * config.temperature, + ) + scaled_probability_per_cell = gumbel_dist.sample() + else: + scaled_probability_per_cell = dist_per_cell.probs + + # [batch_size, seq_length] + scaled_probability_per_cell = (scaled_probability_per_cell / numeric_values_scale) * input_mask_float + count_result = torch.sum(scaled_probability_per_cell, dim=1) + numeric_values_masked = torch.where( + torch.isnan(numeric_values), torch.zeros_like(numeric_values), numeric_values + ) # Mask non-numeric table values to zero. + sum_result = torch.sum(scaled_probability_per_cell * numeric_values_masked, dim=1) + avg_approximation = config.average_approximation_function + if avg_approximation == AverageApproximationFunction.RATIO: + average_result = sum_result / (count_result + EPSILON_ZERO_DIVISION) + elif avg_approximation == AverageApproximationFunction.FIRST_ORDER: + # The sum of all probabilities except that correspond to other cells + # Ex here stands for expectation, more explicitly the expectation of the sum of N-1 Bernoulli random variables plus + # the constant 1, which is computed as adding all N expected values and subtracting the extra one. It corresponds to X_c + # in Appendix D of the original TAPAS paper which is trying to approximate the average of a random set. + ex = torch.sum(scaled_probability_per_cell, dim=1, keepdim=True) - scaled_probability_per_cell + 1 + average_result = torch.sum(numeric_values_masked * scaled_probability_per_cell / ex, dim=1) + elif avg_approximation == AverageApproximationFunction.SECOND_ORDER: + # The sum of all probabilities except that correspond to other cells + ex = torch.sum(scaled_probability_per_cell, dim=1, keepdim=True) - scaled_probability_per_cell + 1 + pointwise_var = scaled_probability_per_cell * (1 - scaled_probability_per_cell) + var = torch.sum(pointwise_var, dim=1, keepdim=True) - pointwise_var + + multiplier = (var / torch.square(ex) + 1) / ex + average_result = torch.sum(numeric_values_masked * scaled_probability_per_cell * multiplier, dim=1) + else: + raise ValueError(f"Invalid average_approximation_function: {config.average_approximation_function}") + + if config.use_gumbel_for_aggregation: + gumbel_dist = torch.distributions.RelaxedOneHotCategorical( + config.aggregation_temperature, logits=logits_aggregation[:, 1:] + ) + # [batch_size, num_aggregation_labels - 1] + aggregation_op_only_probs = gumbel_dist.sample() + else: + # [batch_size, num_aggregation_labels - 1] + aggregation_op_only_probs = nn.functional.softmax( + logits_aggregation[:, 1:] / config.aggregation_temperature, dim=-1 + ) + + all_results = torch.cat( + [ + torch.unsqueeze(sum_result, dim=1), + torch.unsqueeze(average_result, dim=1), + torch.unsqueeze(count_result, dim=1), + ], + dim=1, + ) + + expected_result = torch.sum(all_results * aggregation_op_only_probs, dim=1) + return expected_result + + +# PyTorch does not currently support Huber loss with custom delta so we define it ourself +def huber_loss(input, target, delta: float = 1.0): + errors = torch.abs(input - target) # shape (batch_size,) + return torch.where(errors < delta, 0.5 * errors**2, errors * delta - (0.5 * delta**2)) + + +def _calculate_regression_loss( + answer, + aggregate_mask, + dist_per_cell, + numeric_values, + numeric_values_scale, + input_mask_float, + logits_aggregation, + config, +): + """ + Calculates the regression loss per example. + + Args: + answer (`torch.FloatTensor` of shape `(batch_size,)`): + Answer for every example in the batch. Nan if there is no scalar answer. + aggregate_mask (`torch.FloatTensor` of shape `(batch_size,)`): + A mask set to 1 for examples that should use aggregation functions. + dist_per_cell (`torch.distributions.Bernoulli`): + Cell selection distribution for each cell. + numeric_values (`torch.FloatTensor` of shape `(batch_size, seq_length)`): + Numeric values of every token. Nan for tokens which are not numeric values. + numeric_values_scale (`torch.FloatTensor` of shape `(batch_size, seq_length)`): + Scale of the numeric values of every token. + input_mask_float (`torch.FloatTensor` of shape `(batch_size, seq_length)`): + Mask for the table, without question tokens and table headers. + logits_aggregation (`torch.FloatTensor` of shape `(batch_size, num_aggregation_labels)`): + Logits per aggregation operation. + config ([`TapasConfig`]): + Model configuration class with all the parameters of the model + + Returns: + per_example_answer_loss_scaled (`torch.FloatTensor` of shape `(batch_size,)`): Scales answer loss for each + example in the batch. large_answer_loss_mask (`torch.FloatTensor` of shape `(batch_size,)`): A mask which is 1 + for examples for which their answer loss is larger than the answer_loss_cutoff. + """ + # float32 (batch_size,) + expected_result = _calculate_expected_result( + dist_per_cell, numeric_values, numeric_values_scale, input_mask_float, logits_aggregation, config + ) + + # float32 (batch_size,) + answer_masked = torch.where(torch.isnan(answer), torch.zeros_like(answer), answer) + + if config.use_normalized_answer_loss: + normalizer = (torch.max(torch.abs(expected_result), torch.abs(answer_masked)) + EPSILON_ZERO_DIVISION).detach() + + normalized_answer_masked = answer_masked / normalizer + normalized_expected_result = expected_result / normalizer + per_example_answer_loss = huber_loss( + normalized_expected_result * aggregate_mask, normalized_answer_masked * aggregate_mask + ) + else: + per_example_answer_loss = huber_loss( + expected_result * aggregate_mask, answer_masked * aggregate_mask, delta=config.huber_loss_delta + ) + + if config.answer_loss_cutoff is None: + large_answer_loss_mask = torch.ones_like(per_example_answer_loss, dtype=torch.float32) + + else: + large_answer_loss_mask = torch.where( + per_example_answer_loss > config.answer_loss_cutoff, + torch.zeros_like(per_example_answer_loss, dtype=torch.float32), + torch.ones_like(per_example_answer_loss, dtype=torch.float32), + ) + per_example_answer_loss_scaled = config.answer_loss_importance * (per_example_answer_loss * aggregate_mask) + + return per_example_answer_loss_scaled, large_answer_loss_mask diff --git a/parrot/lib/python3.10/site-packages/transformers/models/tapas/modeling_tf_tapas.py b/parrot/lib/python3.10/site-packages/transformers/models/tapas/modeling_tf_tapas.py new file mode 100644 index 0000000000000000000000000000000000000000..b26803cecb0bc17eacbaa0fb810231bc9de179f7 --- /dev/null +++ b/parrot/lib/python3.10/site-packages/transformers/models/tapas/modeling_tf_tapas.py @@ -0,0 +1,2454 @@ +# coding=utf-8 +# Copyright 2021 Google Research and The HuggingFace Inc. team. +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. +"""TF 2.0 TAPAS model.""" + + +from __future__ import annotations + +import enum +import math +from dataclasses import dataclass +from typing import Dict, Optional, Tuple, Union + +import numpy as np +import tensorflow as tf + +from ...activations_tf import get_tf_activation +from ...modeling_tf_outputs import ( + TFBaseModelOutputWithPastAndCrossAttentions, + TFBaseModelOutputWithPooling, + TFMaskedLMOutput, + TFSequenceClassifierOutput, +) +from ...modeling_tf_utils import ( + TFMaskedLanguageModelingLoss, + TFModelInputType, + TFPreTrainedModel, + TFSequenceClassificationLoss, + get_initializer, + keras, + keras_serializable, + unpack_inputs, +) +from ...tf_utils import check_embeddings_within_bounds, shape_list, stable_softmax +from ...utils import ( + ModelOutput, + add_start_docstrings, + add_start_docstrings_to_model_forward, + is_tensorflow_probability_available, + logging, + replace_return_docstrings, +) +from .configuration_tapas import TapasConfig + + +logger = logging.get_logger(__name__) + +# soft dependency +if is_tensorflow_probability_available(): + try: + import tensorflow_probability as tfp + + # On the first call, check whether a compatible version of TensorFlow is installed + # TensorFlow Probability depends on a recent stable release of TensorFlow + n = tfp.distributions.Normal(loc=0.0, scale=1.0) + except ImportError: + logger.error( + "TAPAS models are not usable since `tensorflow_probability` can't be loaded. " + "It seems you have `tensorflow_probability` installed with the wrong tensorflow version. " + "Please try to reinstall it following the instructions here: https://github.com/tensorflow/probability." + ) +else: + try: + import tensorflow_probability as tfp + + # On the first call, check whether a compatible version of TensorFlow is installed + # TensorFlow Probability depends on a recent stable release of TensorFlow + _ = tfp.distributions.Normal(loc=0.0, scale=1.0) + except ImportError: + pass + +_CONFIG_FOR_DOC = "TapasConfig" +_CHECKPOINT_FOR_DOC = "google/tapas-base" + + +EPSILON_ZERO_DIVISION = 1e-10 +CLOSE_ENOUGH_TO_LOG_ZERO = -10000.0 + + +@dataclass +class TFTableQuestionAnsweringOutput(ModelOutput): + """ + Output type of [`TFTapasForQuestionAnswering`]. + + Args: + loss (`tf.Tensor` of shape `(1,)`, *optional*, returned when `labels` (and possibly `answer`, `aggregation_labels`, `numeric_values` and `numeric_values_scale` are provided)): + Total loss as the sum of the hierarchical cell selection log-likelihood loss and (optionally) the + semi-supervised regression loss and (optionally) supervised loss for aggregations. + logits (`tf.Tensor` of shape `(batch_size, sequence_length)`): + Prediction scores of the cell selection head, for every token. + logits_aggregation (`tf.Tensor`, *optional*, of shape `(batch_size, num_aggregation_labels)`): + Prediction scores of the aggregation head, for every aggregation operator. + hidden_states (`tuple(tf.Tensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`): + Tuple of `tf.Tensor` (one for the output of the embeddings + one for the output of each layer) of shape + `(batch_size, sequence_length, hidden_size)`. Hidden-states of the model at the output of each layer plus + the initial embedding outputs. + attentions (`tuple(tf.Tensor)`, *optional*, returned when `output_attentions=True` is passed or when `config.output_attentions=True`): + Tuple of `tf.Tensor` (one for each layer) of shape `(batch_size, num_heads, sequence_length, + sequence_length)`. Attentions weights after the attention softmax, used to compute the weighted average in + the self-attention heads. + """ + + loss: tf.Tensor | None = None + logits: tf.Tensor = None + logits_aggregation: tf.Tensor | None = None + hidden_states: Tuple[tf.Tensor] | None = None + attentions: Tuple[tf.Tensor] | None = None + + +class TFTapasEmbeddings(keras.layers.Layer): + """ + Construct the embeddings from word, position and token_type embeddings. Same as BertEmbeddings but with a number of + additional token type embeddings to encode tabular structure. + """ + + def __init__(self, config: TapasConfig, **kwargs): + super().__init__(**kwargs) + + self.config = config + self.number_of_token_type_embeddings = len(config.type_vocab_sizes) + self.reset_position_index_per_cell = config.reset_position_index_per_cell + self.hidden_size = config.hidden_size + self.max_position_embeddings = config.max_position_embeddings + self.initializer_range = config.initializer_range + self.LayerNorm = keras.layers.LayerNormalization(epsilon=config.layer_norm_eps, name="LayerNorm") + self.dropout = keras.layers.Dropout(rate=config.hidden_dropout_prob) + + def build(self, input_shape=None): + with tf.name_scope("word_embeddings"): + self.weight = self.add_weight( + name="weight", + shape=[self.config.vocab_size, self.hidden_size], + initializer=get_initializer(self.initializer_range), + ) + + with tf.name_scope("position_embeddings"): + self.position_embeddings = self.add_weight( + name="embeddings", + shape=[self.max_position_embeddings, self.hidden_size], + initializer=get_initializer(self.initializer_range), + ) + for i, type_vocab_size in enumerate(self.config.type_vocab_sizes): + with tf.name_scope(f"token_type_embeddings_{i}"): + setattr( + self, + f"token_type_embeddings_{i}", + self.add_weight( + name="embeddings", + shape=[type_vocab_size, self.hidden_size], + initializer=get_initializer(self.initializer_range), + ), + ) + + if self.built: + return + self.built = True + if getattr(self, "LayerNorm", None) is not None: + with tf.name_scope(self.LayerNorm.name): + self.LayerNorm.build([None, None, self.config.hidden_size]) + + def call( + self, + input_ids: tf.Tensor = None, + position_ids: tf.Tensor = None, + token_type_ids: tf.Tensor = None, + inputs_embeds: tf.Tensor = None, + training: bool = False, + ) -> tf.Tensor: + """ + Applies embedding based on inputs tensor. + + Returns: + final_embeddings (`tf.Tensor`): output embedding tensor. + """ + assert not (input_ids is None and inputs_embeds is None) + if input_ids is not None: + input_shape = shape_list(input_ids) + else: + input_shape = shape_list(inputs_embeds)[:-1] + + seq_length = input_shape[1] + + if token_type_ids is None: + token_type_ids = tf.fill(dims=input_shape + [self.number_of_token_type_embeddings], value=0) + + if position_ids is None: + # create absolute position embeddings + position_ids = tf.expand_dims(tf.range(start=0, limit=seq_length), axis=0) + position_ids = tf.broadcast_to(position_ids, shape=input_shape) + # when self.config.reset_position_index_per_cell is set to True, create relative position embeddings + if self.reset_position_index_per_cell: + # shape (batch_size, seq_len) + col_index = IndexMap(token_type_ids[:, :, 1], self.config.type_vocab_sizes[1], batch_dims=1) + # shape (batch_size, seq_len) + row_index = IndexMap(token_type_ids[:, :, 2], self.config.type_vocab_sizes[2], batch_dims=1) + # shape (batch_size, seq_len) + full_index = ProductIndexMap(col_index, row_index) + # shape (max_rows * max_columns,). First absolute position for every cell + first_position_per_segment = reduce_min(position_ids, full_index)[0] + # ? shape (batch_size, seq_len). First absolute position of the cell for every token + first_position = gather(first_position_per_segment, full_index) + # shape (1, seq_len) + position = tf.expand_dims(tf.range(start=0, limit=seq_length), axis=0) + position_ids = tf.math.minimum(self.max_position_embeddings - 1, position - first_position) + + if input_ids is not None: + check_embeddings_within_bounds(input_ids, self.config.vocab_size) + inputs_embeds = tf.gather(params=self.weight, indices=input_ids) + + position_embeddings = tf.gather(self.position_embeddings, indices=position_ids) + + final_embeddings = inputs_embeds + position_embeddings + + for i in range(self.number_of_token_type_embeddings): + name = f"token_type_embeddings_{i}" + final_embeddings += tf.gather(params=getattr(self, name), indices=token_type_ids[:, :, i]) + + final_embeddings = self.LayerNorm(inputs=final_embeddings) + final_embeddings = self.dropout(inputs=final_embeddings, training=training) + + return final_embeddings + + +# Copied from transformers.models.bert.modeling_tf_bert.TFBertSelfAttention with Bert->Tapas +class TFTapasSelfAttention(keras.layers.Layer): + def __init__(self, config: TapasConfig, **kwargs): + super().__init__(**kwargs) + + if config.hidden_size % config.num_attention_heads != 0: + raise ValueError( + f"The hidden size ({config.hidden_size}) is not a multiple of the number " + f"of attention heads ({config.num_attention_heads})" + ) + + self.num_attention_heads = config.num_attention_heads + self.attention_head_size = int(config.hidden_size / config.num_attention_heads) + self.all_head_size = self.num_attention_heads * self.attention_head_size + self.sqrt_att_head_size = math.sqrt(self.attention_head_size) + + self.query = keras.layers.Dense( + units=self.all_head_size, kernel_initializer=get_initializer(config.initializer_range), name="query" + ) + self.key = keras.layers.Dense( + units=self.all_head_size, kernel_initializer=get_initializer(config.initializer_range), name="key" + ) + self.value = keras.layers.Dense( + units=self.all_head_size, kernel_initializer=get_initializer(config.initializer_range), name="value" + ) + self.dropout = keras.layers.Dropout(rate=config.attention_probs_dropout_prob) + + self.is_decoder = config.is_decoder + self.config = config + + def transpose_for_scores(self, tensor: tf.Tensor, batch_size: int) -> tf.Tensor: + # Reshape from [batch_size, seq_length, all_head_size] to [batch_size, seq_length, num_attention_heads, attention_head_size] + tensor = tf.reshape(tensor=tensor, shape=(batch_size, -1, self.num_attention_heads, self.attention_head_size)) + + # Transpose the tensor from [batch_size, seq_length, num_attention_heads, attention_head_size] to [batch_size, num_attention_heads, seq_length, attention_head_size] + return tf.transpose(tensor, perm=[0, 2, 1, 3]) + + def call( + self, + hidden_states: tf.Tensor, + attention_mask: tf.Tensor, + head_mask: tf.Tensor, + encoder_hidden_states: tf.Tensor, + encoder_attention_mask: tf.Tensor, + past_key_value: Tuple[tf.Tensor], + output_attentions: bool, + training: bool = False, + ) -> Tuple[tf.Tensor]: + batch_size = shape_list(hidden_states)[0] + mixed_query_layer = self.query(inputs=hidden_states) + + # If this is instantiated as a cross-attention module, the keys + # and values come from an encoder; the attention mask needs to be + # such that the encoder's padding tokens are not attended to. + is_cross_attention = encoder_hidden_states is not None + + if is_cross_attention and past_key_value is not None: + # reuse k,v, cross_attentions + key_layer = past_key_value[0] + value_layer = past_key_value[1] + attention_mask = encoder_attention_mask + elif is_cross_attention: + key_layer = self.transpose_for_scores(self.key(inputs=encoder_hidden_states), batch_size) + value_layer = self.transpose_for_scores(self.value(inputs=encoder_hidden_states), batch_size) + attention_mask = encoder_attention_mask + elif past_key_value is not None: + key_layer = self.transpose_for_scores(self.key(inputs=hidden_states), batch_size) + value_layer = self.transpose_for_scores(self.value(inputs=hidden_states), batch_size) + key_layer = tf.concat([past_key_value[0], key_layer], axis=2) + value_layer = tf.concat([past_key_value[1], value_layer], axis=2) + else: + key_layer = self.transpose_for_scores(self.key(inputs=hidden_states), batch_size) + value_layer = self.transpose_for_scores(self.value(inputs=hidden_states), batch_size) + + query_layer = self.transpose_for_scores(mixed_query_layer, batch_size) + + if self.is_decoder: + # if cross_attention save Tuple(tf.Tensor, tf.Tensor) of all cross attention key/value_states. + # Further calls to cross_attention layer can then reuse all cross-attention + # key/value_states (first "if" case) + # if uni-directional self-attention (decoder) save Tuple(tf.Tensor, tf.Tensor) of + # all previous decoder key/value_states. Further calls to uni-directional self-attention + # can concat previous decoder key/value_states to current projected key/value_states (third "elif" case) + # if encoder bi-directional self-attention `past_key_value` is always `None` + past_key_value = (key_layer, value_layer) + + # Take the dot product between "query" and "key" to get the raw attention scores. + # (batch size, num_heads, seq_len_q, seq_len_k) + attention_scores = tf.matmul(query_layer, key_layer, transpose_b=True) + dk = tf.cast(self.sqrt_att_head_size, dtype=attention_scores.dtype) + attention_scores = tf.divide(attention_scores, dk) + + if attention_mask is not None: + # Apply the attention mask is (precomputed for all layers in TFTapasModel call() function) + attention_scores = tf.add(attention_scores, attention_mask) + + # Normalize the attention scores to probabilities. + attention_probs = stable_softmax(logits=attention_scores, axis=-1) + + # This is actually dropping out entire tokens to attend to, which might + # seem a bit unusual, but is taken from the original Transformer paper. + attention_probs = self.dropout(inputs=attention_probs, training=training) + + # Mask heads if we want to + if head_mask is not None: + attention_probs = tf.multiply(attention_probs, head_mask) + + attention_output = tf.matmul(attention_probs, value_layer) + attention_output = tf.transpose(attention_output, perm=[0, 2, 1, 3]) + + # (batch_size, seq_len_q, all_head_size) + attention_output = tf.reshape(tensor=attention_output, shape=(batch_size, -1, self.all_head_size)) + outputs = (attention_output, attention_probs) if output_attentions else (attention_output,) + + if self.is_decoder: + outputs = outputs + (past_key_value,) + return outputs + + def build(self, input_shape=None): + if self.built: + return + self.built = True + if getattr(self, "query", None) is not None: + with tf.name_scope(self.query.name): + self.query.build([None, None, self.config.hidden_size]) + if getattr(self, "key", None) is not None: + with tf.name_scope(self.key.name): + self.key.build([None, None, self.config.hidden_size]) + if getattr(self, "value", None) is not None: + with tf.name_scope(self.value.name): + self.value.build([None, None, self.config.hidden_size]) + + +# Copied from transformers.models.bert.modeling_tf_bert.TFBertSelfOutput with Bert->Tapas +class TFTapasSelfOutput(keras.layers.Layer): + def __init__(self, config: TapasConfig, **kwargs): + super().__init__(**kwargs) + + self.dense = keras.layers.Dense( + units=config.hidden_size, kernel_initializer=get_initializer(config.initializer_range), name="dense" + ) + self.LayerNorm = keras.layers.LayerNormalization(epsilon=config.layer_norm_eps, name="LayerNorm") + self.dropout = keras.layers.Dropout(rate=config.hidden_dropout_prob) + self.config = config + + def call(self, hidden_states: tf.Tensor, input_tensor: tf.Tensor, training: bool = False) -> tf.Tensor: + hidden_states = self.dense(inputs=hidden_states) + hidden_states = self.dropout(inputs=hidden_states, training=training) + hidden_states = self.LayerNorm(inputs=hidden_states + input_tensor) + + return hidden_states + + def build(self, input_shape=None): + if self.built: + return + self.built = True + if getattr(self, "dense", None) is not None: + with tf.name_scope(self.dense.name): + self.dense.build([None, None, self.config.hidden_size]) + if getattr(self, "LayerNorm", None) is not None: + with tf.name_scope(self.LayerNorm.name): + self.LayerNorm.build([None, None, self.config.hidden_size]) + + +# Copied from transformers.models.bert.modeling_tf_bert.TFBertAttention with Bert->Tapas +class TFTapasAttention(keras.layers.Layer): + def __init__(self, config: TapasConfig, **kwargs): + super().__init__(**kwargs) + + self.self_attention = TFTapasSelfAttention(config, name="self") + self.dense_output = TFTapasSelfOutput(config, name="output") + + def prune_heads(self, heads): + raise NotImplementedError + + def call( + self, + input_tensor: tf.Tensor, + attention_mask: tf.Tensor, + head_mask: tf.Tensor, + encoder_hidden_states: tf.Tensor, + encoder_attention_mask: tf.Tensor, + past_key_value: Tuple[tf.Tensor], + output_attentions: bool, + training: bool = False, + ) -> Tuple[tf.Tensor]: + self_outputs = self.self_attention( + hidden_states=input_tensor, + attention_mask=attention_mask, + head_mask=head_mask, + encoder_hidden_states=encoder_hidden_states, + encoder_attention_mask=encoder_attention_mask, + past_key_value=past_key_value, + output_attentions=output_attentions, + training=training, + ) + attention_output = self.dense_output( + hidden_states=self_outputs[0], input_tensor=input_tensor, training=training + ) + # add attentions (possibly with past_key_value) if we output them + outputs = (attention_output,) + self_outputs[1:] + + return outputs + + def build(self, input_shape=None): + if self.built: + return + self.built = True + if getattr(self, "self_attention", None) is not None: + with tf.name_scope(self.self_attention.name): + self.self_attention.build(None) + if getattr(self, "dense_output", None) is not None: + with tf.name_scope(self.dense_output.name): + self.dense_output.build(None) + + +# Copied from transformers.models.bert.modeling_tf_bert.TFBertIntermediate with Bert->Tapas +class TFTapasIntermediate(keras.layers.Layer): + def __init__(self, config: TapasConfig, **kwargs): + super().__init__(**kwargs) + + self.dense = keras.layers.Dense( + units=config.intermediate_size, kernel_initializer=get_initializer(config.initializer_range), name="dense" + ) + + if isinstance(config.hidden_act, str): + self.intermediate_act_fn = get_tf_activation(config.hidden_act) + else: + self.intermediate_act_fn = config.hidden_act + self.config = config + + def call(self, hidden_states: tf.Tensor) -> tf.Tensor: + hidden_states = self.dense(inputs=hidden_states) + hidden_states = self.intermediate_act_fn(hidden_states) + + return hidden_states + + def build(self, input_shape=None): + if self.built: + return + self.built = True + if getattr(self, "dense", None) is not None: + with tf.name_scope(self.dense.name): + self.dense.build([None, None, self.config.hidden_size]) + + +# Copied from transformers.models.bert.modeling_tf_bert.TFBertOutput with Bert->Tapas +class TFTapasOutput(keras.layers.Layer): + def __init__(self, config: TapasConfig, **kwargs): + super().__init__(**kwargs) + + self.dense = keras.layers.Dense( + units=config.hidden_size, kernel_initializer=get_initializer(config.initializer_range), name="dense" + ) + self.LayerNorm = keras.layers.LayerNormalization(epsilon=config.layer_norm_eps, name="LayerNorm") + self.dropout = keras.layers.Dropout(rate=config.hidden_dropout_prob) + self.config = config + + def call(self, hidden_states: tf.Tensor, input_tensor: tf.Tensor, training: bool = False) -> tf.Tensor: + hidden_states = self.dense(inputs=hidden_states) + hidden_states = self.dropout(inputs=hidden_states, training=training) + hidden_states = self.LayerNorm(inputs=hidden_states + input_tensor) + + return hidden_states + + def build(self, input_shape=None): + if self.built: + return + self.built = True + if getattr(self, "dense", None) is not None: + with tf.name_scope(self.dense.name): + self.dense.build([None, None, self.config.intermediate_size]) + if getattr(self, "LayerNorm", None) is not None: + with tf.name_scope(self.LayerNorm.name): + self.LayerNorm.build([None, None, self.config.hidden_size]) + + +# Copied from transformers.models.bert.modeling_tf_bert.TFBertLayer with Bert->Tapas +class TFTapasLayer(keras.layers.Layer): + def __init__(self, config: TapasConfig, **kwargs): + super().__init__(**kwargs) + + self.attention = TFTapasAttention(config, name="attention") + self.is_decoder = config.is_decoder + self.add_cross_attention = config.add_cross_attention + if self.add_cross_attention: + if not self.is_decoder: + raise ValueError(f"{self} should be used as a decoder model if cross attention is added") + self.crossattention = TFTapasAttention(config, name="crossattention") + self.intermediate = TFTapasIntermediate(config, name="intermediate") + self.bert_output = TFTapasOutput(config, name="output") + + def call( + self, + hidden_states: tf.Tensor, + attention_mask: tf.Tensor, + head_mask: tf.Tensor, + encoder_hidden_states: tf.Tensor | None, + encoder_attention_mask: tf.Tensor | None, + past_key_value: Tuple[tf.Tensor] | None, + output_attentions: bool, + training: bool = False, + ) -> Tuple[tf.Tensor]: + # decoder uni-directional self-attention cached key/values tuple is at positions 1,2 + self_attn_past_key_value = past_key_value[:2] if past_key_value is not None else None + self_attention_outputs = self.attention( + input_tensor=hidden_states, + attention_mask=attention_mask, + head_mask=head_mask, + encoder_hidden_states=None, + encoder_attention_mask=None, + past_key_value=self_attn_past_key_value, + output_attentions=output_attentions, + training=training, + ) + attention_output = self_attention_outputs[0] + + # if decoder, the last output is tuple of self-attn cache + if self.is_decoder: + outputs = self_attention_outputs[1:-1] + present_key_value = self_attention_outputs[-1] + else: + outputs = self_attention_outputs[1:] # add self attentions if we output attention weights + + cross_attn_present_key_value = None + if self.is_decoder and encoder_hidden_states is not None: + if not hasattr(self, "crossattention"): + raise ValueError( + f"If `encoder_hidden_states` are passed, {self} has to be instantiated with cross-attention layers" + " by setting `config.add_cross_attention=True`" + ) + + # cross_attn cached key/values tuple is at positions 3,4 of past_key_value tuple + cross_attn_past_key_value = past_key_value[-2:] if past_key_value is not None else None + cross_attention_outputs = self.crossattention( + input_tensor=attention_output, + attention_mask=attention_mask, + head_mask=head_mask, + encoder_hidden_states=encoder_hidden_states, + encoder_attention_mask=encoder_attention_mask, + past_key_value=cross_attn_past_key_value, + output_attentions=output_attentions, + training=training, + ) + attention_output = cross_attention_outputs[0] + outputs = outputs + cross_attention_outputs[1:-1] # add cross attentions if we output attention weights + + # add cross-attn cache to positions 3,4 of present_key_value tuple + cross_attn_present_key_value = cross_attention_outputs[-1] + present_key_value = present_key_value + cross_attn_present_key_value + + intermediate_output = self.intermediate(hidden_states=attention_output) + layer_output = self.bert_output( + hidden_states=intermediate_output, input_tensor=attention_output, training=training + ) + outputs = (layer_output,) + outputs # add attentions if we output them + + # if decoder, return the attn key/values as the last output + if self.is_decoder: + outputs = outputs + (present_key_value,) + + return outputs + + def build(self, input_shape=None): + if self.built: + return + self.built = True + if getattr(self, "attention", None) is not None: + with tf.name_scope(self.attention.name): + self.attention.build(None) + if getattr(self, "intermediate", None) is not None: + with tf.name_scope(self.intermediate.name): + self.intermediate.build(None) + if getattr(self, "bert_output", None) is not None: + with tf.name_scope(self.bert_output.name): + self.bert_output.build(None) + if getattr(self, "crossattention", None) is not None: + with tf.name_scope(self.crossattention.name): + self.crossattention.build(None) + + +# Copied from transformers.models.bert.modeling_tf_bert.TFBertEncoder with Bert->Tapas +class TFTapasEncoder(keras.layers.Layer): + def __init__(self, config: TapasConfig, **kwargs): + super().__init__(**kwargs) + self.config = config + self.layer = [TFTapasLayer(config, name=f"layer_._{i}") for i in range(config.num_hidden_layers)] + + def call( + self, + hidden_states: tf.Tensor, + attention_mask: tf.Tensor, + head_mask: tf.Tensor, + encoder_hidden_states: tf.Tensor | None, + encoder_attention_mask: tf.Tensor | None, + past_key_values: Tuple[Tuple[tf.Tensor]] | None, + use_cache: Optional[bool], + output_attentions: bool, + output_hidden_states: bool, + return_dict: bool, + training: bool = False, + ) -> Union[TFBaseModelOutputWithPastAndCrossAttentions, Tuple[tf.Tensor]]: + all_hidden_states = () if output_hidden_states else None + all_attentions = () if output_attentions else None + all_cross_attentions = () if output_attentions and self.config.add_cross_attention else None + + next_decoder_cache = () if use_cache else None + for i, layer_module in enumerate(self.layer): + if output_hidden_states: + all_hidden_states = all_hidden_states + (hidden_states,) + + past_key_value = past_key_values[i] if past_key_values is not None else None + + layer_outputs = layer_module( + hidden_states=hidden_states, + attention_mask=attention_mask, + head_mask=head_mask[i], + encoder_hidden_states=encoder_hidden_states, + encoder_attention_mask=encoder_attention_mask, + past_key_value=past_key_value, + output_attentions=output_attentions, + training=training, + ) + hidden_states = layer_outputs[0] + + if use_cache: + next_decoder_cache += (layer_outputs[-1],) + + if output_attentions: + all_attentions = all_attentions + (layer_outputs[1],) + if self.config.add_cross_attention and encoder_hidden_states is not None: + all_cross_attentions = all_cross_attentions + (layer_outputs[2],) + + # Add last layer + if output_hidden_states: + all_hidden_states = all_hidden_states + (hidden_states,) + + if not return_dict: + return tuple( + v for v in [hidden_states, all_hidden_states, all_attentions, all_cross_attentions] if v is not None + ) + + return TFBaseModelOutputWithPastAndCrossAttentions( + last_hidden_state=hidden_states, + past_key_values=next_decoder_cache, + hidden_states=all_hidden_states, + attentions=all_attentions, + cross_attentions=all_cross_attentions, + ) + + def build(self, input_shape=None): + if self.built: + return + self.built = True + if getattr(self, "layer", None) is not None: + for layer in self.layer: + with tf.name_scope(layer.name): + layer.build(None) + + +# Copied from transformers.models.bert.modeling_tf_bert.TFBertPooler with Bert->Tapas +class TFTapasPooler(keras.layers.Layer): + def __init__(self, config: TapasConfig, **kwargs): + super().__init__(**kwargs) + + self.dense = keras.layers.Dense( + units=config.hidden_size, + kernel_initializer=get_initializer(config.initializer_range), + activation="tanh", + name="dense", + ) + self.config = config + + def call(self, hidden_states: tf.Tensor) -> tf.Tensor: + # We "pool" the model by simply taking the hidden state corresponding + # to the first token. + first_token_tensor = hidden_states[:, 0] + pooled_output = self.dense(inputs=first_token_tensor) + + return pooled_output + + def build(self, input_shape=None): + if self.built: + return + self.built = True + if getattr(self, "dense", None) is not None: + with tf.name_scope(self.dense.name): + self.dense.build([None, None, self.config.hidden_size]) + + +# Copied from transformers.models.bert.modeling_tf_bert.TFBertPredictionHeadTransform with Bert->Tapas +class TFTapasPredictionHeadTransform(keras.layers.Layer): + def __init__(self, config: TapasConfig, **kwargs): + super().__init__(**kwargs) + + self.dense = keras.layers.Dense( + units=config.hidden_size, + kernel_initializer=get_initializer(config.initializer_range), + name="dense", + ) + + if isinstance(config.hidden_act, str): + self.transform_act_fn = get_tf_activation(config.hidden_act) + else: + self.transform_act_fn = config.hidden_act + + self.LayerNorm = keras.layers.LayerNormalization(epsilon=config.layer_norm_eps, name="LayerNorm") + self.config = config + + def call(self, hidden_states: tf.Tensor) -> tf.Tensor: + hidden_states = self.dense(inputs=hidden_states) + hidden_states = self.transform_act_fn(hidden_states) + hidden_states = self.LayerNorm(inputs=hidden_states) + + return hidden_states + + def build(self, input_shape=None): + if self.built: + return + self.built = True + if getattr(self, "dense", None) is not None: + with tf.name_scope(self.dense.name): + self.dense.build([None, None, self.config.hidden_size]) + if getattr(self, "LayerNorm", None) is not None: + with tf.name_scope(self.LayerNorm.name): + self.LayerNorm.build([None, None, self.config.hidden_size]) + + +# Copied from transformers.models.bert.modeling_tf_bert.TFBertLMPredictionHead with Bert->Tapas +class TFTapasLMPredictionHead(keras.layers.Layer): + def __init__(self, config: TapasConfig, input_embeddings: keras.layers.Layer, **kwargs): + super().__init__(**kwargs) + + self.config = config + self.hidden_size = config.hidden_size + + self.transform = TFTapasPredictionHeadTransform(config, name="transform") + + # The output weights are the same as the input embeddings, but there is + # an output-only bias for each token. + self.input_embeddings = input_embeddings + + def build(self, input_shape=None): + self.bias = self.add_weight(shape=(self.config.vocab_size,), initializer="zeros", trainable=True, name="bias") + + if self.built: + return + self.built = True + if getattr(self, "transform", None) is not None: + with tf.name_scope(self.transform.name): + self.transform.build(None) + + def get_output_embeddings(self) -> keras.layers.Layer: + return self.input_embeddings + + def set_output_embeddings(self, value: tf.Variable): + self.input_embeddings.weight = value + self.input_embeddings.vocab_size = shape_list(value)[0] + + def get_bias(self) -> Dict[str, tf.Variable]: + return {"bias": self.bias} + + def set_bias(self, value: tf.Variable): + self.bias = value["bias"] + self.config.vocab_size = shape_list(value["bias"])[0] + + def call(self, hidden_states: tf.Tensor) -> tf.Tensor: + hidden_states = self.transform(hidden_states=hidden_states) + seq_length = shape_list(hidden_states)[1] + hidden_states = tf.reshape(tensor=hidden_states, shape=[-1, self.hidden_size]) + hidden_states = tf.matmul(a=hidden_states, b=self.input_embeddings.weight, transpose_b=True) + hidden_states = tf.reshape(tensor=hidden_states, shape=[-1, seq_length, self.config.vocab_size]) + hidden_states = tf.nn.bias_add(value=hidden_states, bias=self.bias) + + return hidden_states + + +# Copied from transformers.models.bert.modeling_tf_bert.TFBertMLMHead with Bert->Tapas +class TFTapasMLMHead(keras.layers.Layer): + def __init__(self, config: TapasConfig, input_embeddings: keras.layers.Layer, **kwargs): + super().__init__(**kwargs) + + self.predictions = TFTapasLMPredictionHead(config, input_embeddings, name="predictions") + + def call(self, sequence_output: tf.Tensor) -> tf.Tensor: + prediction_scores = self.predictions(hidden_states=sequence_output) + + return prediction_scores + + def build(self, input_shape=None): + if self.built: + return + self.built = True + if getattr(self, "predictions", None) is not None: + with tf.name_scope(self.predictions.name): + self.predictions.build(None) + + +@keras_serializable +class TFTapasMainLayer(keras.layers.Layer): + config_class = TapasConfig + + def __init__(self, config: TapasConfig, add_pooling_layer: bool = True, **kwargs): + super().__init__(**kwargs) + + self.config = config + + self.embeddings = TFTapasEmbeddings(config, name="embeddings") + self.encoder = TFTapasEncoder(config, name="encoder") + self.pooler = TFTapasPooler(config, name="pooler") if add_pooling_layer else None + + def get_input_embeddings(self) -> keras.layers.Layer: + return self.embeddings + + def set_input_embeddings(self, value: tf.Variable): + self.embeddings.weight = value + self.embeddings.vocab_size = shape_list(value)[0] + + def _prune_heads(self, heads_to_prune): + """ + Prunes heads of the model. heads_to_prune: dict of {layer_num: list of heads to prune in this layer} See base + class PreTrainedModel + """ + raise NotImplementedError + + @unpack_inputs + def call( + self, + input_ids: TFModelInputType | None = None, + attention_mask: np.ndarray | tf.Tensor | None = None, + token_type_ids: np.ndarray | tf.Tensor | None = None, + position_ids: np.ndarray | tf.Tensor | None = None, + head_mask: np.ndarray | tf.Tensor | None = None, + inputs_embeds: np.ndarray | tf.Tensor | None = None, + output_attentions: Optional[bool] = None, + output_hidden_states: Optional[bool] = None, + return_dict: Optional[bool] = None, + training: bool = False, + ) -> Union[TFBaseModelOutputWithPooling, Tuple[tf.Tensor]]: + if input_ids is not None and inputs_embeds is not None: + raise ValueError("You cannot specify both input_ids and inputs_embeds at the same time") + elif input_ids is not None: + input_shape = shape_list(input_ids) + elif inputs_embeds is not None: + input_shape = shape_list(inputs_embeds)[:-1] + else: + raise ValueError("You have to specify either input_ids or inputs_embeds") + + if attention_mask is None: + attention_mask = tf.fill(dims=input_shape, value=1) + + if token_type_ids is None: + token_type_ids = tf.fill(dims=input_shape + [len(self.config.type_vocab_sizes)], value=0) + + embedding_output = self.embeddings( + input_ids=input_ids, + position_ids=position_ids, + token_type_ids=token_type_ids, + inputs_embeds=inputs_embeds, + training=training, + ) + + # We create a 3D attention mask from a 2D tensor mask. + # Sizes are [batch_size, 1, 1, to_seq_length] + # So we can broadcast to [batch_size, num_heads, from_seq_length, to_seq_length] + # this attention mask is more simple than the triangular masking of causal attention + # used in OpenAI GPT, we just need to prepare the broadcast dimension here. + extended_attention_mask = tf.reshape(attention_mask, (input_shape[0], 1, 1, input_shape[1])) + + # Since attention_mask is 1.0 for positions we want to attend and 0.0 for + # masked positions, this operation will create a tensor which is 0.0 for + # positions we want to attend and -10000.0 for masked positions. + # Since we are adding it to the raw scores before the softmax, this is + # effectively the same as removing these entirely. + extended_attention_mask = tf.cast(extended_attention_mask, dtype=embedding_output.dtype) + one_cst = tf.constant(1.0, dtype=embedding_output.dtype) + ten_thousand_cst = tf.constant(-10000.0, dtype=embedding_output.dtype) + extended_attention_mask = tf.multiply(tf.subtract(one_cst, extended_attention_mask), ten_thousand_cst) + + # Prepare head mask if needed + # 1.0 in head_mask indicate we keep the head + # attention_probs has shape bsz x n_heads x N x N + # input head_mask has shape [num_heads] or [num_hidden_layers x num_heads] + # and head_mask is converted to shape [num_hidden_layers x batch x num_heads x seq_length x seq_length] + if head_mask is not None: + raise NotImplementedError + else: + head_mask = [None] * self.config.num_hidden_layers + + encoder_outputs = self.encoder( + hidden_states=embedding_output, + attention_mask=extended_attention_mask, + head_mask=head_mask, + encoder_hidden_states=None, + encoder_attention_mask=None, + past_key_values=None, + use_cache=None, + output_attentions=output_attentions, + output_hidden_states=output_hidden_states, + return_dict=return_dict, + training=training, + ) + + sequence_output = encoder_outputs[0] + pooled_output = self.pooler(hidden_states=sequence_output) if self.pooler is not None else None + + if not return_dict: + return ( + sequence_output, + pooled_output, + ) + encoder_outputs[1:] + + return TFBaseModelOutputWithPooling( + last_hidden_state=sequence_output, + pooler_output=pooled_output, + hidden_states=encoder_outputs.hidden_states, + attentions=encoder_outputs.attentions, + ) + + def build(self, input_shape=None): + if self.built: + return + self.built = True + if getattr(self, "embeddings", None) is not None: + with tf.name_scope(self.embeddings.name): + self.embeddings.build(None) + if getattr(self, "encoder", None) is not None: + with tf.name_scope(self.encoder.name): + self.encoder.build(None) + if getattr(self, "pooler", None) is not None: + with tf.name_scope(self.pooler.name): + self.pooler.build(None) + + +class TFTapasPreTrainedModel(TFPreTrainedModel): + """ + An abstract class to handle weights initialization and a simple interface for downloading and loading pretrained + models. + """ + + config_class = TapasConfig + base_model_prefix = "tapas" + + @property + def input_signature(self): + return { + "input_ids": tf.TensorSpec((None, None), tf.int32, name="input_ids"), + "attention_mask": tf.TensorSpec((None, None), tf.float32, name="attention_mask"), + "token_type_ids": tf.TensorSpec((None, None, 7), tf.int32, name="token_type_ids"), + } + + +TAPAS_START_DOCSTRING = r""" + + This model inherits from [`TFPreTrainedModel`]. Check the superclass documentation for the generic methods the + library implements for all its model (such as downloading or saving, resizing the input embeddings, pruning heads + etc.) + + This model is also a [keras.Model](https://www.tensorflow.org/api_docs/python/tf/keras/Model) subclass. Use it + as a regular TF 2.0 Keras Model and refer to the TF 2.0 documentation for all matter related to general usage and + behavior. + + + + TensorFlow models and layers in `transformers` accept two formats as input: + + - having all inputs as keyword arguments (like PyTorch models), or + - having all inputs as a list, tuple or dict in the first positional argument. + + The reason the second format is supported is that Keras methods prefer this format when passing inputs to models + and layers. Because of this support, when using methods like `model.fit()` things should "just work" for you - just + pass your inputs and labels in any format that `model.fit()` supports! If, however, you want to use the second + format outside of Keras methods like `fit()` and `predict()`, such as when creating your own layers or models with + the Keras `Functional` API, there are three possibilities you can use to gather all the input Tensors in the first + positional argument: + + - a single Tensor with `input_ids` only and nothing else: `model(input_ids)` + - a list of varying length with one or several input Tensors IN THE ORDER given in the docstring: + `model([input_ids, attention_mask])` or `model([input_ids, attention_mask, token_type_ids])` + - a dictionary with one or several input Tensors associated to the input names given in the docstring: + `model({"input_ids": input_ids, "token_type_ids": token_type_ids})` + + Note that when creating models and layers with + [subclassing](https://keras.io/guides/making_new_layers_and_models_via_subclassing/) then you don't need to worry + about any of this, as you can just pass inputs like you would to any other Python function! + + + + Parameters: + config ([`TapasConfig`]): Model configuration class with all the parameters of the model. + Initializing with a config file does not load the weights associated with the model, only the + configuration. Check out the [`~PreTrainedModel.from_pretrained`] method to load the model weights. +""" + +TAPAS_INPUTS_DOCSTRING = r""" + Args: + input_ids (`np.ndarray`, `tf.Tensor`, `List[tf.Tensor]` ``Dict[str, tf.Tensor]` or `Dict[str, np.ndarray]` and each example must have the shape `({0})`): + Indices of input sequence tokens in the vocabulary. + + Indices can be obtained using [`AutoTokenizer`]. See [`PreTrainedTokenizer.__call__`] and + [`PreTrainedTokenizer.encode`] for details. + + [What are input IDs?](../glossary#input-ids) + attention_mask (`np.ndarray` or `tf.Tensor` of shape `({0})`, *optional*): + Mask to avoid performing attention on padding token indices. Mask values selected in `[0, 1]`: + + - 1 for tokens that are **not masked**, + - 0 for tokens that are **masked**. + + [What are attention masks?](../glossary#attention-mask) + token_type_ids (`np.ndarray` or `tf.Tensor` of shape `({0}, 7)`, *optional*): + Token indices that encode tabular structure. Indices can be obtained using [`AutoTokenizer`]. See this + class for more info. + + [What are token type IDs?](../glossary#token-type-ids) + position_ids (`np.ndarray` or `tf.Tensor` of shape `({0})`, *optional*): + Indices of positions of each input sequence tokens in the position embeddings. If + `reset_position_index_per_cell` of [`TapasConfig`] is set to `True`, relative position embeddings will be + used. Selected in the range `[0, config.max_position_embeddings - 1]`. + + [What are position IDs?](../glossary#position-ids) + head_mask (`np.ndarray` or `tf.Tensor` of shape `(num_heads,)` or `(num_layers, num_heads)`, *optional*): + Mask to nullify selected heads of the self-attention modules. Mask values selected in `[0, 1]`: + + - 1 indicates the head is **not masked**, + - 0 indicates the head is **masked**. + + inputs_embeds (`np.ndarray` or `tf.Tensor` of shape `({0}, hidden_size)`, *optional*): + Optionally, instead of passing `input_ids` you can choose to directly pass an embedded representation. This + is useful if you want more control over how to convert `input_ids` indices into associated vectors than the + model's internal embedding lookup matrix. + output_attentions (`bool`, *optional*): + Whether or not to return the attentions tensors of all attention layers. See `attentions` under returned + tensors for more detail. This argument can be used only in eager mode, in graph mode the value in the + config will be used instead. + output_hidden_states (`bool`, *optional*): + Whether or not to return the hidden states of all layers. See `hidden_states` under returned tensors for + more detail. This argument can be used only in eager mode, in graph mode the value in the config will be + used instead. + return_dict (`bool`, *optional*): + Whether or not to return a [`~utils.ModelOutput`] instead of a plain tuple. This argument can be used in + eager mode, in graph mode the value will always be set to True. + training (`bool`, *optional*, defaults to `False``): + Whether or not to use the model in training mode (some modules like dropout modules have different + behaviors between training and evaluation). +""" + + +@add_start_docstrings( + "The bare Tapas Model transformer outputting raw hidden-states without any specific head on top.", + TAPAS_START_DOCSTRING, +) +class TFTapasModel(TFTapasPreTrainedModel): + def __init__(self, config: TapasConfig, *inputs, **kwargs): + super().__init__(config, *inputs, **kwargs) + + self.tapas = TFTapasMainLayer(config, name="tapas") + + @unpack_inputs + @add_start_docstrings_to_model_forward(TAPAS_INPUTS_DOCSTRING.format("batch_size, sequence_length")) + @replace_return_docstrings(output_type=TFBaseModelOutputWithPooling, config_class=_CONFIG_FOR_DOC) + def call( + self, + input_ids: TFModelInputType | None = None, + attention_mask: np.ndarray | tf.Tensor | None = None, + token_type_ids: np.ndarray | tf.Tensor | None = None, + position_ids: np.ndarray | tf.Tensor | None = None, + head_mask: np.ndarray | tf.Tensor | None = None, + inputs_embeds: np.ndarray | tf.Tensor | None = None, + output_attentions: Optional[bool] = None, + output_hidden_states: Optional[bool] = None, + return_dict: Optional[bool] = None, + training: Optional[bool] = False, + ) -> Union[TFBaseModelOutputWithPooling, Tuple[tf.Tensor]]: + r""" + Returns: + + Examples: + + ```python + >>> from transformers import AutoTokenizer, TapasModel + >>> import pandas as pd + + >>> tokenizer = AutoTokenizer.from_pretrained("google/tapas-base") + >>> model = TapasModel.from_pretrained("google/tapas-base") + + >>> data = { + ... "Actors": ["Brad Pitt", "Leonardo Di Caprio", "George Clooney"], + ... "Age": ["56", "45", "59"], + ... "Number of movies": ["87", "53", "69"], + ... } + >>> table = pd.DataFrame.from_dict(data) + >>> queries = ["How many movies has George Clooney played in?", "How old is Brad Pitt?"] + + >>> inputs = tokenizer(table=table, queries=queries, padding="max_length", return_tensors="tf") + >>> outputs = model(**inputs) + + >>> last_hidden_states = outputs.last_hidden_state + ```""" + outputs = self.tapas( + input_ids=input_ids, + attention_mask=attention_mask, + token_type_ids=token_type_ids, + position_ids=position_ids, + head_mask=head_mask, + inputs_embeds=inputs_embeds, + output_attentions=output_attentions, + output_hidden_states=output_hidden_states, + return_dict=return_dict, + training=training, + ) + + return outputs + + def build(self, input_shape=None): + if self.built: + return + self.built = True + if getattr(self, "tapas", None) is not None: + with tf.name_scope(self.tapas.name): + self.tapas.build(None) + + +@add_start_docstrings("""Tapas Model with a `language modeling` head on top.""", TAPAS_START_DOCSTRING) +class TFTapasForMaskedLM(TFTapasPreTrainedModel, TFMaskedLanguageModelingLoss): + def __init__(self, config: TapasConfig, *inputs, **kwargs): + super().__init__(config, *inputs, **kwargs) + + if config.is_decoder: + logger.warning( + "If you want to use `TFTapasForMaskedLM` make sure `config.is_decoder=False` for " + "bi-directional self-attention." + ) + + self.tapas = TFTapasMainLayer(config, add_pooling_layer=False, name="tapas") + self.lm_head = TFTapasMLMHead(config, input_embeddings=self.tapas.embeddings, name="cls") + + def get_lm_head(self) -> keras.layers.Layer: + return self.lm_head.predictions + + @unpack_inputs + @add_start_docstrings_to_model_forward(TAPAS_INPUTS_DOCSTRING.format("batch_size, sequence_length")) + @replace_return_docstrings(output_type=TFMaskedLMOutput, config_class=_CONFIG_FOR_DOC) + def call( + self, + input_ids: TFModelInputType | None = None, + attention_mask: np.ndarray | tf.Tensor | None = None, + token_type_ids: np.ndarray | tf.Tensor | None = None, + position_ids: np.ndarray | tf.Tensor | None = None, + head_mask: np.ndarray | tf.Tensor | None = None, + inputs_embeds: np.ndarray | tf.Tensor | None = None, + output_attentions: Optional[bool] = None, + output_hidden_states: Optional[bool] = None, + return_dict: Optional[bool] = None, + labels: np.ndarray | tf.Tensor | None = None, + training: Optional[bool] = False, + ) -> Union[TFMaskedLMOutput, Tuple[tf.Tensor]]: + r""" + labels (`tf.Tensor` or `np.ndarray` of shape `(batch_size, sequence_length)`, *optional*): + Labels for computing the masked language modeling loss. Indices should be in `[-100, 0, ..., + config.vocab_size]` (see `input_ids` docstring) Tokens with indices set to `-100` are ignored (masked), the + loss is only computed for the tokens with labels in `[0, ..., config.vocab_size]` + + Returns: + + Examples: + + ```python + >>> from transformers import AutoTokenizer, TapasForMaskedLM + >>> import pandas as pd + + >>> tokenizer = AutoTokenizer.from_pretrained("google/tapas-base") + >>> model = TapasForMaskedLM.from_pretrained("google/tapas-base") + + >>> data = { + ... "Actors": ["Brad Pitt", "Leonardo Di Caprio", "George Clooney"], + ... "Age": ["56", "45", "59"], + ... "Number of movies": ["87", "53", "69"], + ... } + >>> table = pd.DataFrame.from_dict(data) + + >>> inputs = tokenizer( + ... table=table, queries="How many [MASK] has George [MASK] played in?", return_tensors="tf" + ... ) + >>> labels = tokenizer( + ... table=table, queries="How many movies has George Clooney played in?", return_tensors="tf" + ... )["input_ids"] + + >>> outputs = model(**inputs, labels=labels) + >>> logits = outputs.logits + ```""" + outputs = self.tapas( + input_ids=input_ids, + attention_mask=attention_mask, + token_type_ids=token_type_ids, + position_ids=position_ids, + head_mask=head_mask, + inputs_embeds=inputs_embeds, + output_attentions=output_attentions, + output_hidden_states=output_hidden_states, + return_dict=return_dict, + training=training, + ) + sequence_output = outputs[0] + prediction_scores = self.lm_head(sequence_output) + loss = None if labels is None else self.hf_compute_loss(labels=labels, logits=prediction_scores) + + if not return_dict: + output = (prediction_scores,) + outputs[2:] + return ((loss,) + output) if loss is not None else output + + return TFMaskedLMOutput( + loss=loss, + logits=prediction_scores, + hidden_states=outputs.hidden_states, + attentions=outputs.attentions, + ) + + def build(self, input_shape=None): + if self.built: + return + self.built = True + if getattr(self, "tapas", None) is not None: + with tf.name_scope(self.tapas.name): + self.tapas.build(None) + if getattr(self, "lm_head", None) is not None: + with tf.name_scope(self.lm_head.name): + self.lm_head.build(None) + + +class TFTapasComputeTokenLogits(keras.layers.Layer): + def __init__(self, config: TapasConfig, **kwargs): + super().__init__(**kwargs) + + self.temperature = config.temperature + # cell selection heads + with tf.name_scope("output"): + self.output_weights = self.add_weight( + name="output_weights", + shape=(config.hidden_size,), + dtype=tf.float32, + trainable=True, + initializer=tf.zeros_initializer() + if config.init_cell_selection_weights_to_zero + else keras.initializers.TruncatedNormal(stddev=config.initializer_range), + ) + self.output_bias = self.add_weight( + name="output_bias", shape=(), trainable=True, initializer=tf.zeros_initializer() + ) + + def call(self, sequence_output: tf.Tensor) -> tf.Tensor: + """ + Computes logits per token + + Args: + sequence_output (`tf.Tensor` of shape `(batch_size, sequence_length, hidden_size)`): + Also known as last_hidden_state. Sequence of hidden-states at the output of the last layer of the + model. + + Returns: + logits (`tf.Tensor` of shape `(batch_size, sequence_length)`): Logits per token. + """ + logits = (tf.einsum("bsj,j->bs", sequence_output, self.output_weights) + self.output_bias) / self.temperature + return logits + + +class TFTapasComputeColumnLogits(keras.layers.Layer): + def __init__(self, config: TapasConfig, **kwargs): + super().__init__(**kwargs) + + with tf.name_scope("column_output"): + self.column_output_weights = self.add_weight( + name="column_output_weights", + shape=[config.hidden_size], + dtype=tf.float32, + trainable=True, + initializer=tf.zeros_initializer() + if config.init_cell_selection_weights_to_zero + else keras.initializers.TruncatedNormal(stddev=config.initializer_range), + ) + self.column_output_bias = self.add_weight( + name="column_output_bias", shape=(), trainable=True, initializer=tf.zeros_initializer() + ) + + def call(self, sequence_output, cell_index, cell_mask, allow_empty_column_selection) -> tf.Tensor: + """ + Computes the column logits. + + Args: + sequence_output (`tf.Tensor` of shape `(batch_size, sequence_length, hidden_size)`): + Also known as last_hidden_state. Sequence of hidden-states at the output of the last layer of the + model. + cell_index (`ProductIndexMap`): + Index that groups tokens into cells. + cell_mask (`tf.Tensor` of shape `(batch_size, max_num_rows * max_num_cols)`): + Mask for cells that exist in the table (i.e. that are not padding). + allow_empty_column_selection (`bool`): + Whether to allow not to select any column + + Returns: + column_logits (`tf.Tensor`of shape `(batch_size, max_num_cols)`): Tensor containing the column logits for + every example in the batch. + """ + + # First, compute the token logits (batch_size, seq_len) - without temperature + token_logits = tf.einsum("bsj,j->bs", sequence_output, self.column_output_weights) + self.column_output_bias + + # Next, average the logits per cell (batch_size, max_num_cols*max_num_rows) + cell_logits, cell_logits_index = reduce_mean(token_logits, cell_index) + + # Finally, average the logits per column (batch_size, max_num_cols) + column_index = cell_index.project_inner(cell_logits_index) + column_logits, out_index = reduce_sum(cell_logits * cell_mask, column_index) + + cell_count, _ = reduce_sum(cell_mask, column_index) + column_logits /= cell_count + EPSILON_ZERO_DIVISION + + # Mask columns that do not appear in the example. + is_padding = tf.logical_and(cell_count < 0.5, tf.not_equal(out_index.indices, 0)) + column_logits += CLOSE_ENOUGH_TO_LOG_ZERO * tf.cast(is_padding, tf.float32) + + if not allow_empty_column_selection: + column_logits += CLOSE_ENOUGH_TO_LOG_ZERO * tf.cast(tf.equal(out_index.indices, 0), tf.float32) + + return column_logits + + +@add_start_docstrings( + """ + Tapas Model with a cell selection head and optional aggregation head on top for question-answering tasks on tables + (linear layers on top of the hidden-states output to compute `logits` and optional `logits_aggregation`), e.g. for + SQA, WTQ or WikiSQL-supervised tasks. + """, + TAPAS_START_DOCSTRING, +) +class TFTapasForQuestionAnswering(TFTapasPreTrainedModel): + def __init__(self, config: TapasConfig, *inputs, **kwargs): + super().__init__(config, *inputs, **kwargs) + + # base model + self.tapas = TFTapasMainLayer(config, name="tapas") + + # dropout + self.dropout = keras.layers.Dropout(config.hidden_dropout_prob) + + self.compute_token_logits = TFTapasComputeTokenLogits(config, name="compute_token_logits") + + self.compute_column_logits = TFTapasComputeColumnLogits(config, name="compute_column_logits") + + if config.num_aggregation_labels > 0: + self.aggregation_classifier = keras.layers.Dense( + config.num_aggregation_labels, + kernel_initializer=get_initializer(config.initializer_range), + name="aggregation_classifier", + ) + self.config = config + + @unpack_inputs + @add_start_docstrings_to_model_forward(TAPAS_INPUTS_DOCSTRING.format("batch_size, sequence_length")) + @replace_return_docstrings(output_type=TFTableQuestionAnsweringOutput, config_class=_CONFIG_FOR_DOC) + def call( + self, + input_ids: TFModelInputType | None = None, + attention_mask: np.ndarray | tf.Tensor | None = None, + token_type_ids: np.ndarray | tf.Tensor | None = None, + position_ids: np.ndarray | tf.Tensor | None = None, + head_mask: np.ndarray | tf.Tensor | None = None, + inputs_embeds: np.ndarray | tf.Tensor | None = None, + table_mask: np.ndarray | tf.Tensor | None = None, + aggregation_labels: np.ndarray | tf.Tensor | None = None, + float_answer: np.ndarray | tf.Tensor | None = None, + numeric_values: np.ndarray | tf.Tensor | None = None, + numeric_values_scale: np.ndarray | tf.Tensor | None = None, + output_attentions: Optional[bool] = None, + output_hidden_states: Optional[bool] = None, + return_dict: Optional[bool] = None, + labels: np.ndarray | tf.Tensor | None = None, + training: Optional[bool] = False, + ) -> Union[TFTableQuestionAnsweringOutput, Tuple[tf.Tensor]]: + r""" + table_mask (`tf.Tensor` of shape `(batch_size, seq_length)`, *optional*): + Mask for the table. Indicates which tokens belong to the table (1). Question tokens, table headers and + padding are 0. + labels (`tf.Tensor` of shape `(batch_size, seq_length)`, *optional*): + Labels per token for computing the hierarchical cell selection loss. This encodes the positions of the + answer appearing in the table. Can be obtained using [`AutoTokenizer`]. + + - 1 for tokens that are **part of the answer**, + - 0 for tokens that are **not part of the answer**. + + aggregation_labels (`tf.Tensor` of shape `(batch_size, )`, *optional*): + Aggregation function index for every example in the batch for computing the aggregation loss. Indices + should be in `[0, ..., config.num_aggregation_labels - 1]`. Only required in case of strong supervision for + aggregation (WikiSQL-supervised). + float_answer (`tf.Tensor` of shape `(batch_size, )`, *optional*): + Float answer for every example in the batch. Set to *float('nan')* for cell selection questions. Only + required in case of weak supervision (WTQ) to calculate the aggregate mask and regression loss. + numeric_values (`tf.Tensor` of shape `(batch_size, seq_length)`, *optional*): + Numeric values of every token, NaN for tokens which are not numeric values. Can be obtained using + [`AutoTokenizer`]. Only required in case of weak supervision for aggregation (WTQ) to calculate the + regression loss. + numeric_values_scale (`tf.Tensor` of shape `(batch_size, seq_length)`, *optional*): + Scale of the numeric values of every token. Can be obtained using [`AutoTokenizer`]. Only required in case + of weak supervision for aggregation (WTQ) to calculate the regression loss. + + Returns: + + Examples: + + ```python + >>> from transformers import AutoTokenizer, TapasForQuestionAnswering + >>> import pandas as pd + + >>> tokenizer = AutoTokenizer.from_pretrained("google/tapas-base-finetuned-wtq") + >>> model = TapasForQuestionAnswering.from_pretrained("google/tapas-base-finetuned-wtq") + + >>> data = { + ... "Actors": ["Brad Pitt", "Leonardo Di Caprio", "George Clooney"], + ... "Age": ["56", "45", "59"], + ... "Number of movies": ["87", "53", "69"], + ... } + >>> table = pd.DataFrame.from_dict(data) + >>> queries = ["How many movies has George Clooney played in?", "How old is Brad Pitt?"] + + >>> inputs = tokenizer(table=table, queries=queries, padding="max_length", return_tensors="tf") + >>> outputs = model(**inputs) + + >>> logits = outputs.logits + >>> logits_aggregation = outputs.logits_aggregation + ```""" + + outputs = self.tapas( + input_ids=input_ids, + attention_mask=attention_mask, + token_type_ids=token_type_ids, + position_ids=position_ids, + head_mask=head_mask, + inputs_embeds=inputs_embeds, + output_attentions=output_attentions, + output_hidden_states=output_hidden_states, + return_dict=return_dict, + training=training, + ) + + sequence_output = outputs[0] + pooled_output = outputs[1] + + sequence_output = self.dropout(sequence_output) + + if input_ids is not None: + input_shape = shape_list(input_ids) + else: + input_shape = shape_list(inputs_embeds)[:-1] + + # Construct indices for the table. + if token_type_ids is None: + token_type_ids = tf.fill(input_shape + [len(self.config.type_vocab_sizes)], 0) + + token_types = [ + "segment_ids", + "column_ids", + "row_ids", + "prev_labels", + "column_ranks", + "inv_column_ranks", + "numeric_relations", + ] + + row_ids = token_type_ids[:, :, token_types.index("row_ids")] + column_ids = token_type_ids[:, :, token_types.index("column_ids")] + + # Construct indices for the table. + row_index = IndexMap( + indices=tf.minimum(tf.cast(row_ids, tf.int32), self.config.max_num_rows - 1), + num_segments=self.config.max_num_rows, + batch_dims=1, + ) + col_index = IndexMap( + indices=tf.minimum(tf.cast(column_ids, tf.int32), self.config.max_num_columns - 1), + num_segments=self.config.max_num_columns, + batch_dims=1, + ) + cell_index = ProductIndexMap(row_index, col_index) + + # Masks. + input_shape = shape_list(input_ids) if input_ids is not None else shape_list(inputs_embeds)[:-1] + if attention_mask is None: + attention_mask = tf.ones(input_shape) + # Table cells only, without question tokens and table headers. + if table_mask is None: + table_mask = tf.where(row_ids > 0, tf.ones_like(row_ids), tf.zeros_like(row_ids)) + # [batch_size, seq_length] + input_mask_float = tf.cast(attention_mask, tf.float32) + table_mask_float = tf.cast(table_mask, tf.float32) + + # Mask for cells that exist in the table (i.e. that are not padding). + cell_mask, _ = reduce_mean(input_mask_float, cell_index) + + # Compute logits per token. These are used to select individual cells. + logits = self.compute_token_logits(sequence_output) + + # Compute logits per column. These are used to select a column. + column_logits = None + if self.config.select_one_column: + column_logits = self.compute_column_logits( + sequence_output, cell_index, cell_mask, self.config.allow_empty_column_selection + ) + + # Aggregate logits. + logits_aggregation = None + if self.config.num_aggregation_labels > 0: + logits_aggregation = self.aggregation_classifier(pooled_output) + + # Total loss calculation + total_loss = tf.zeros(shape=(1,), dtype=tf.float32) + calculate_loss = False + if labels is not None: + calculate_loss = True + is_supervised = not self.config.num_aggregation_labels > 0 or not self.config.use_answer_as_supervision + + # Semi-supervised cell selection in case of no aggregation: + # If the answer (the denotation) appears directly in the table we might + # select the answer without applying any aggregation function. There are + # some ambiguous cases, see utils._calculate_aggregate_mask for more info. + # `aggregate_mask` is 1 for examples where we chose to aggregate and 0 + # for examples where we chose to select the answer directly. + # `labels` encodes the positions of the answer appearing in the table. + if is_supervised: + aggregate_mask = None + else: + if float_answer is not None: + assert ( + shape_list(labels)[0] == shape_list(float_answer)[0] + ), "Make sure the answers are a FloatTensor of shape (batch_size,)" + # [batch_size] + aggregate_mask = _calculate_aggregate_mask( + float_answer, + pooled_output, + self.config.cell_selection_preference, + labels, + self.aggregation_classifier, + ) + else: + aggregate_mask = None + raise ValueError("You have to specify float answers in order to calculate the aggregate mask") + + # Cell selection log-likelihood + if self.config.average_logits_per_cell: + logits_per_cell, _ = reduce_mean(logits, cell_index) + logits = gather(logits_per_cell, cell_index) + dist_per_token = tfp.distributions.Bernoulli(logits=logits) + + # Compute cell selection loss per example. + selection_loss_per_example = None + if not self.config.select_one_column: + weight = tf.where( + labels == 0, + tf.ones_like(labels, dtype=tf.float32), + self.config.positive_label_weight * tf.ones_like(labels, dtype=tf.float32), + ) + selection_loss_per_token = -dist_per_token.log_prob(labels) * weight + selection_loss_per_example = tf.reduce_sum(selection_loss_per_token * input_mask_float, axis=1) / ( + tf.reduce_sum(input_mask_float, axis=1) + EPSILON_ZERO_DIVISION + ) + else: + selection_loss_per_example, logits = _single_column_cell_selection_loss( + logits, column_logits, labels, cell_index, col_index, cell_mask + ) + dist_per_token = tfp.distributions.Bernoulli(logits=logits) + + # Supervised cell selection + if self.config.disable_per_token_loss: + pass + elif is_supervised: + total_loss += tf.reduce_mean(selection_loss_per_example) + else: + # For the not supervised case, do not assign loss for cell selection + total_loss += tf.reduce_mean(selection_loss_per_example * (1.0 - aggregate_mask)) + + # Semi-supervised regression loss and supervised loss for aggregations + if self.config.num_aggregation_labels > 0: + if is_supervised: + # Note that `aggregate_mask` is None if the setting is supervised. + if aggregation_labels is not None: + assert ( + shape_list(labels)[0] == shape_list(aggregation_labels)[0] + ), "Make sure the aggregation labels are a LongTensor of shape (batch_size,)" + per_example_additional_loss = _calculate_aggregation_loss( + logits_aggregation, + aggregate_mask, + aggregation_labels, + self.config.use_answer_as_supervision, + self.config.num_aggregation_labels, + self.config.aggregation_loss_weight, + ) + else: + raise ValueError( + "You have to specify aggregation labels in order to calculate the aggregation loss" + ) + else: + aggregation_labels = tf.zeros(shape_list(labels)[0], dtype=tf.int32) + per_example_additional_loss = _calculate_aggregation_loss( + logits_aggregation, + aggregate_mask, + aggregation_labels, + self.config.use_answer_as_supervision, + self.config.num_aggregation_labels, + self.config.aggregation_loss_weight, + ) + + if self.config.use_answer_as_supervision: + if numeric_values is not None and numeric_values_scale is not None: + assert shape_list(numeric_values) == shape_list(numeric_values_scale) + # Add regression loss for numeric answers which require aggregation. + answer_loss, large_answer_loss_mask = _calculate_regression_loss( + float_answer, + aggregate_mask, + dist_per_token, + numeric_values, + numeric_values_scale, + table_mask_float, + logits_aggregation, + self.config, + ) + per_example_additional_loss += answer_loss + # Zero loss for examples with answer_loss > cutoff. + per_example_additional_loss *= large_answer_loss_mask + else: + raise ValueError( + "You have to specify numeric values and numeric values scale in order to calculate the" + " regression loss" + ) + total_loss += tf.reduce_mean(per_example_additional_loss) + + else: + # if no label ids are provided, set them to zeros in order to properly compute logits + labels = tf.zeros_like(logits) + _, logits = _single_column_cell_selection_loss( + logits, column_logits, labels, cell_index, col_index, cell_mask + ) + if not return_dict: + output = (logits, logits_aggregation) + outputs[2:] + return ((total_loss,) + output) if calculate_loss else output + + return TFTableQuestionAnsweringOutput( + loss=total_loss if calculate_loss else None, + logits=logits, + logits_aggregation=logits_aggregation, + hidden_states=outputs.hidden_states, + attentions=outputs.attentions, + ) + + def build(self, input_shape=None): + if self.built: + return + self.built = True + if getattr(self, "tapas", None) is not None: + with tf.name_scope(self.tapas.name): + self.tapas.build(None) + if getattr(self, "compute_token_logits", None) is not None: + with tf.name_scope(self.compute_token_logits.name): + self.compute_token_logits.build(None) + if getattr(self, "compute_column_logits", None) is not None: + with tf.name_scope(self.compute_column_logits.name): + self.compute_column_logits.build(None) + if getattr(self, "aggregation_classifier", None) is not None: + with tf.name_scope(self.aggregation_classifier.name): + self.aggregation_classifier.build([None, None, self.config.hidden_size]) + + +@add_start_docstrings( + """ + Tapas Model with a sequence classification head on top (a linear layer on top of the pooled output), e.g. for table + entailment tasks, such as TabFact (Chen et al., 2020). + """, + TAPAS_START_DOCSTRING, +) +class TFTapasForSequenceClassification(TFTapasPreTrainedModel, TFSequenceClassificationLoss): + def __init__(self, config: TapasConfig, *inputs, **kwargs): + super().__init__(config, *inputs, **kwargs) + self.num_labels = config.num_labels + + self.tapas = TFTapasMainLayer(config, name="tapas") + self.dropout = keras.layers.Dropout(config.hidden_dropout_prob, name="dropout") + self.classifier = keras.layers.Dense( + config.num_labels, kernel_initializer=get_initializer(config.initializer_range), name="classifier" + ) + self.config = config + + @unpack_inputs + @add_start_docstrings_to_model_forward(TAPAS_INPUTS_DOCSTRING.format("batch_size, num_choices, sequence_length")) + @replace_return_docstrings(output_type=TFSequenceClassifierOutput, config_class=_CONFIG_FOR_DOC) + def call( + self, + input_ids: TFModelInputType | None = None, + attention_mask: np.ndarray | tf.Tensor | None = None, + token_type_ids: np.ndarray | tf.Tensor | None = None, + position_ids: np.ndarray | tf.Tensor | None = None, + head_mask: np.ndarray | tf.Tensor | None = None, + inputs_embeds: np.ndarray | tf.Tensor | None = None, + output_attentions: Optional[bool] = None, + output_hidden_states: Optional[bool] = None, + return_dict: Optional[bool] = None, + labels: np.ndarray | tf.Tensor | None = None, + training: Optional[bool] = False, + ) -> Union[TFSequenceClassifierOutput, Tuple[tf.Tensor]]: + r""" + labels (`torch.LongTensor` of shape `(batch_size,)`, *optional*): + Labels for computing the sequence classification/regression loss. Indices should be in `[0, ..., + config.num_labels - 1]`. If `config.num_labels == 1` a regression loss is computed (Mean-Square loss), If + `config.num_labels > 1` a classification loss is computed (Cross-Entropy). Note: this is called + "classification_class_index" in the original implementation. + + Returns: + + Examples: + + ```python + >>> from transformers import AutoTokenizer, TapasForSequenceClassification + >>> import tensorflow as tf + >>> import pandas as pd + + >>> tokenizer = AutoTokenizer.from_pretrained("google/tapas-base-finetuned-tabfact") + >>> model = TapasForSequenceClassification.from_pretrained("google/tapas-base-finetuned-tabfact") + + >>> data = { + ... "Actors": ["Brad Pitt", "Leonardo Di Caprio", "George Clooney"], + ... "Age": ["56", "45", "59"], + ... "Number of movies": ["87", "53", "69"], + ... } + >>> table = pd.DataFrame.from_dict(data) + >>> queries = [ + ... "There is only one actor who is 45 years old", + ... "There are 3 actors which played in more than 60 movies", + ... ] + + >>> inputs = tokenizer(table=table, queries=queries, padding="max_length", return_tensors="tf") + >>> labels = tf.convert_to_tensor([1, 0]) # 1 means entailed, 0 means refuted + + >>> outputs = model(**inputs, labels=labels) + >>> loss = outputs.loss + >>> logits = outputs.logits + ```""" + + outputs = self.tapas( + input_ids=input_ids, + attention_mask=attention_mask, + token_type_ids=token_type_ids, + position_ids=position_ids, + head_mask=head_mask, + inputs_embeds=inputs_embeds, + output_attentions=output_attentions, + output_hidden_states=output_hidden_states, + return_dict=return_dict, + training=training, + ) + pooled_output = outputs[1] + pooled_output = self.dropout(inputs=pooled_output, training=training) + logits = self.classifier(inputs=pooled_output) + loss = None if labels is None else self.hf_compute_loss(labels=labels, logits=logits) + + if not return_dict: + output = (logits,) + outputs[2:] + return ((loss,) + output) if loss is not None else output + + return TFSequenceClassifierOutput( + loss=loss, + logits=logits, + hidden_states=outputs.hidden_states, + attentions=outputs.attentions, + ) + + def build(self, input_shape=None): + if self.built: + return + self.built = True + if getattr(self, "tapas", None) is not None: + with tf.name_scope(self.tapas.name): + self.tapas.build(None) + if getattr(self, "dropout", None) is not None: + with tf.name_scope(self.dropout.name): + self.dropout.build(None) + if getattr(self, "classifier", None) is not None: + with tf.name_scope(self.classifier.name): + self.classifier.build([None, None, self.config.hidden_size]) + + +""" TAPAS utilities.""" + + +class AverageApproximationFunction(str, enum.Enum): + RATIO = "ratio" + FIRST_ORDER = "first_order" + SECOND_ORDER = "second_order" + + +# Beginning of everything related to segmented tensors + + +class IndexMap(object): + """Index grouping entries within a tensor.""" + + def __init__(self, indices, num_segments, batch_dims=0): + """ + Creates an index. + + Args: + indices: Tensor of indices, same shape as `values`. + num_segments: Scalar tensor, the number of segments. All elements + in a batched segmented tensor must have the same number of segments (although many segments can be empty). + batch_dims: Python integer, the number of batch dimensions. The first + `batch_dims` dimensions of a SegmentedTensor are treated as batch dimensions. Segments in different batch + elements are always distinct even if they have the same index. + """ + self.indices = tf.convert_to_tensor(indices) + self.num_segments = tf.convert_to_tensor(num_segments) + self.batch_dims = batch_dims + + def batch_shape(self): + return tf.shape(self.indices)[: self.batch_dims] + + +class ProductIndexMap(IndexMap): + """The product of two indices.""" + + def __init__(self, outer_index, inner_index): + """ + Combines indices i and j into pairs (i, j). The result is an index where each segment (i, j) is the + intersection of segments i and j. For example if the inputs represent table cells indexed by respectively rows + and columns the output will be a table indexed by (row, column) pairs, i.e. by cell. The implementation + combines indices {0, .., n - 1} and {0, .., m - 1} into {0, .., nm - 1}. The output has `num_segments` equal to + `outer_index.num_segements` * `inner_index.num_segments`. + + Args: + outer_index: IndexMap. + inner_index: IndexMap, must have the same shape as `outer_index`. + """ + if outer_index.batch_dims != inner_index.batch_dims: + raise ValueError("outer_index.batch_dims and inner_index.batch_dims must be the same.") + + super(ProductIndexMap, self).__init__( + indices=( + inner_index.indices + + outer_index.indices * tf.cast(inner_index.num_segments, inner_index.indices.dtype) + ), + num_segments=inner_index.num_segments * outer_index.num_segments, + batch_dims=inner_index.batch_dims, + ) + self.outer_index = outer_index + self.inner_index = inner_index + + def project_outer(self, index): + """Projects an index with the same index set onto the outer components.""" + return IndexMap( + indices=tf.math.floordiv(index.indices, self.inner_index.num_segments), + num_segments=self.outer_index.num_segments, + batch_dims=index.batch_dims, + ) + + def project_inner(self, index): + """Projects an index with the same index set onto the inner components.""" + return IndexMap( + indices=tf.math.floormod(index.indices, self.inner_index.num_segments), + num_segments=self.inner_index.num_segments, + batch_dims=index.batch_dims, + ) + + +def gather(values, index, name="segmented_gather"): + """ + Gathers from `values` using the index map. For each element in the domain of the index map this operation looks up + a value for that index in `values`. Two elements from the same segment always get assigned the same value. + + Args: + values: [B1, ..., Bn, num_segments, V1, ...] Tensor with segment values. + index: [B1, ..., Bn, I1, ..., Ik] IndexMap. + name: Name for the TensorFlow operation. + + Returns: + [B1, ..., Bn, I1, ..., Ik, V1, ...] Tensor with the gathered values. + """ + return tf.gather(values, index.indices, batch_dims=index.batch_dims, name=name) + + +def flatten(index, name="segmented_flatten"): + """ + Flattens a batched index map to a 1d index map. This operation relabels the segments to keep batch elements + distinct. The k-th batch element will have indices shifted by `num_segments` * (k - 1). The result is a tensor with + `num_segments` multiplied by the number of elements in the batch. + + Args: + index: IndexMap to flatten. + name: Name for the TensorFlow operation. + + Returns: + The flattened IndexMap. + """ + batch_size = tf.reduce_prod(index.batch_shape()) + offset = tf.range(batch_size) * index.num_segments + offset = tf.reshape(offset, index.batch_shape()) + for _ in range(index.batch_dims, index.indices.shape.rank): + offset = tf.expand_dims(offset, -1) + + indices = tf.cast(offset, index.indices.dtype) + index.indices + return IndexMap(indices=tf.reshape(indices, [-1]), num_segments=index.num_segments * batch_size, batch_dims=0) + + +def range_index_map(batch_shape, num_segments, name="range_index_map"): + """ + Constructs an index map equal to range(num_segments). + + Args: + batch_shape (`tf.Tensor`): + Batch shape + num_segments (`int`): + Number of segments + name (`str`, *optional*, defaults to 'range_index_map'): + Name for the operation. Currently not used + + Returns: + (`IndexMap`): IndexMap of shape batch_shape with elements equal to range(num_segments). + """ + batch_shape = tf.convert_to_tensor(batch_shape) + batch_shape.shape.assert_has_rank(1) + num_segments = tf.convert_to_tensor(num_segments) + num_segments.shape.assert_has_rank(0) + + indices = tf.range(num_segments) + shape = tf.concat([tf.ones_like(batch_shape, dtype=tf.int32), tf.expand_dims(num_segments, axis=0)], axis=0) + indices = tf.reshape(indices, shape) + multiples = tf.concat([batch_shape, [1]], axis=0) + indices = tf.tile(indices, multiples) + return IndexMap(indices=indices, num_segments=num_segments, batch_dims=batch_shape.shape.as_list()[0]) + + +def _segment_reduce(values, index, segment_reduce_fn, name): + """ + Applies a segment reduction segment-wise. + + Args: + values (`tf.Tensor`): + Tensor with segment values. + index (`IndexMap`): + IndexMap. + segment_reduce_fn (`str`): + Name for the reduce operation. One of "sum", "mean", "max" or "min". + name (`str`): + Name for the operation. Currently not used + + Returns: + (`IndexMap`): IndexMap of shape batch_shape with elements equal to range(num_segments). + """ + # Flatten the batch dimensions, as segments ops do not support batching. + # However if `values` has extra dimensions to the right keep them + # unflattened. Segmented ops support vector-valued operations. + flat_index = flatten(index) + vector_shape = tf.shape(values)[index.indices.shape.rank :] + flattened_shape = tf.concat([[-1], vector_shape], axis=0) + flat_values = tf.reshape(values, flattened_shape) + segment_means = segment_reduce_fn( + data=flat_values, segment_ids=flat_index.indices, num_segments=flat_index.num_segments + ) + + # Unflatten the values. + new_shape = tf.concat([index.batch_shape(), [index.num_segments], vector_shape], axis=0) + output_values = tf.reshape(segment_means, new_shape) + output_index = range_index_map(index.batch_shape(), index.num_segments) + return output_values, output_index + + +def reduce_mean(values, index, name="segmented_reduce_mean"): + """ + Averages a tensor over its segments. Outputs 0 for empty segments. This operations computes the mean over segments, + with support for: + + - Batching using the first dimensions [B1, B2, ..., Bn]. Each element in a batch can have different indices. + - Vectorization using the last dimension [V1, V2, ...]. If they are present the output will be a mean of vectors + rather than scalars. + Only the middle dimensions [I1, ..., Ik] are reduced by the operation. + + Args: + values: [B1, B2, ..., Bn, I1, .., Ik, V1, V2, ..] tensor of values to be + averaged. + index: IndexMap [B1, B2, ..., Bn, I1, .., Ik] index defining the segments. + name: Name for the TensorFlow ops. + + Returns: + A pair (output_values, output_index) where `output_values` is a tensor of shape [B1, B2, ..., Bn, num_segments, + V1, V2, ..] and `index` is an IndexMap with shape [B1, B2, ..., Bn, num_segments]. + """ + return _segment_reduce(values, index, tf.math.unsorted_segment_mean, name) + + +def reduce_sum(values, index, name="segmented_reduce_sum"): + """ + Sums a tensor over its segments. Outputs 0 for empty segments. This operations computes the sum over segments, with + support for: + + - Batching using the first dimensions [B1, B2, ..., Bn]. Each element in a batch can have different indices. + - Vectorization using the last dimension [V1, V2, ...]. If they are present the output will be a sum of vectors + rather than scalars. + Only the middle dimensions [I1, ..., Ik] are reduced by the operation. + + Args: + values: [B1, B2, ..., Bn, I1, .., Ik, V1, V2, ..] tensor of values to be + averaged. + index: IndexMap [B1, B2, ..., Bn, I1, .., Ik] index defining the segments. + name: Name for the TensorFlow ops. + + Returns: + A pair (output_values, output_index) where `output_values` is a tensor of shape [B1, B2, ..., Bn, num_segments, + V1, V2, ..] and `index` is an IndexMap with shape [B1, B2, ..., Bn, num_segments]. + """ + return _segment_reduce(values, index, tf.math.unsorted_segment_sum, name) + + +def reduce_max(values, index, name="segmented_reduce_max"): + """ + Computes the maximum over segments. This operations computes the maximum over segments, with support for: + + - Batching using the first dimensions [B1, B2, ..., Bn]. Each element in a batch can have different indices. + - Vectorization using the last dimension [V1, V2, ...]. If they are present the output will be an element-wise + maximum of vectors rather than scalars. + Only the middle dimensions [I1, ..., Ik] are reduced by the operation. + + Args: + values: [B1, B2, ..., Bn, I1, .., Ik, V1, V2, ..] tensor of values to be + averaged. + index: IndexMap [B1, B2, ..., Bn, I1, .., Ik] index defining the segments. + name: Name for the TensorFlow ops. + + Returns: + A pair (output_values, output_index) where `output_values` is a tensor of shape [B1, B2, ..., Bn, num_segments, + V1, V2, ..] and `index` is an IndexMap with shape [B1, B2, ..., Bn, num_segments]. + """ + return _segment_reduce(values, index, tf.math.unsorted_segment_max, name) + + +def reduce_min(values, index, name="segmented_reduce_min"): + """Computes the minimum over segments.""" + return _segment_reduce(values, index, tf.math.unsorted_segment_min, name) + + +def _single_column_cell_selection_loss(token_logits, column_logits, labels, cell_index, col_index, cell_mask): + """ + Computes the loss for cell selection constrained to a single column. The loss is a hierarchical log-likelihood. The + model first predicts a column and then selects cells within that column (conditioned on the column). Cells outside + the selected column are never selected. + + Args: + token_logits (`tf.Tensor` of shape `(batch_size, sequence_length)`): + Tensor containing the logits per token. + column_logits (`tf.Tensor` of shape `(batch_size, max_num_cols)`): + Tensor containing the logits per column. + labels (`tf.Tensor` of shape `(batch_size, sequence_length)`): + Labels per token. + cell_index (`ProductIndexMap`): + Index that groups tokens into cells. + col_index (`IndexMap`): + Index that groups tokens into columns. + cell_mask (`tf.Tensor` of shape `(batch_size, max_num_rows * max_num_cols)`): + Mask for cells that exist in the table (i.e. that are not padding). + + Returns: + selection_loss_per_example (`tf.Tensor` of shape `(batch_size,)`): Loss for each example. logits (`tf.Tensor` + of shape `(batch_size, sequence_length)`): New logits which are only allowed to select cells in a single + column. Logits outside of the most likely column according to *column_logits* will be set to a very low value + (such that the probabilities are 0). + """ + # First find the column we should select. We use the column with maximum + # number of selected cells. + labels_per_column, _ = reduce_sum(tf.cast(labels, tf.float32), col_index) + column_label = tf.argmax(labels_per_column, axis=-1, output_type=tf.int32) + # Check if there are no selected cells in the column. In that case the model + # should predict the special column id 0, which means "select nothing". + no_cell_selected = tf.equal(tf.reduce_max(labels_per_column, axis=-1), 0) + column_label = tf.where(no_cell_selected, tf.zeros_like(column_label), column_label) + + column_dist = tfp.distributions.Categorical(logits=column_logits) + column_loss_per_example = -column_dist.log_prob(column_label) + + # Reduce the labels and logits to per-cell from per-token. + logits_per_cell, _ = reduce_mean(token_logits, cell_index) + labels_per_cell, labels_index = reduce_max(tf.cast(labels, tf.int32), cell_index) + + # Mask for the selected column. + column_id_for_cells = cell_index.project_inner(labels_index).indices + column_mask = tf.cast(tf.equal(column_id_for_cells, tf.expand_dims(column_label, axis=1)), tf.float32) + + # Compute the log-likelihood for cells, but only for the selected column. + cell_dist = tfp.distributions.Bernoulli(logits=logits_per_cell) + cell_log_prob = cell_dist.log_prob(labels_per_cell) + cell_loss = -tf.reduce_sum(cell_log_prob * column_mask * cell_mask, axis=1) + # We need to normalize the loss by the number of cells in the column. + cell_loss /= tf.reduce_sum(column_mask * cell_mask, axis=1) + EPSILON_ZERO_DIVISION + + selection_loss_per_example = column_loss_per_example + selection_loss_per_example += tf.where(no_cell_selected, tf.zeros_like(selection_loss_per_example), cell_loss) + + # Set the probs outside the selected column (selected by the *model*) + # to 0. This ensures backwards compatibility with models that select + # cells from multiple columns. + selected_column_id = tf.argmax(column_logits, axis=-1, output_type=tf.int32) + selected_column_mask = tf.cast( + tf.equal(column_id_for_cells, tf.expand_dims(selected_column_id, axis=-1)), tf.float32 + ) + # Never select cells with the special column id 0. + selected_column_mask = tf.where( + tf.equal(column_id_for_cells, 0), tf.zeros_like(selected_column_mask), selected_column_mask + ) + logits_per_cell += CLOSE_ENOUGH_TO_LOG_ZERO * (1.0 - cell_mask * selected_column_mask) + logits = gather(logits_per_cell, cell_index) + + return selection_loss_per_example, logits + + +def _calculate_aggregate_mask(answer, pooled_output, cell_selection_preference, labels, aggregation_classifier): + """ + Finds examples where the model should select cells with no aggregation. + + Returns a mask that determines for which examples should the model select answers directly from the table, without + any aggregation function. If the answer is a piece of text the case is unambiguous as aggregation functions only + apply to numbers. If the answer is a number but does not appear in the table then we must use some aggregation + case. The ambiguous case is when the answer is a number that also appears in the table. In this case we use the + aggregation function probabilities predicted by the model to decide whether to select or aggregate. The threshold + for this is a hyperparameter *cell_selection_preference* + + Args: + answer (`tf.Tensor` of shape `(batch_size, )`): + Answer for every example in the batch. Nan if there is no scalar answer. + pooled_output (`tf.Tensor` of shape `(batch_size, hidden_size)`): + Output of the pooler (BertPooler) on top of the encoder layer. + cell_selection_preference (`float`): + Preference for cell selection in ambiguous cases. + labels (`tf.Tensor` of shape `(batch_size, sequence_length)`): + Labels per token. aggregation_classifier (`torch.nn.Linear`): Aggregation head + + Returns: + aggregate_mask (`tf.Tensor` of shape `(batch_size,)`): A mask set to 1 for examples that should use aggregation + functions. + """ + # tf.Tensor(batch_size,) + aggregate_mask_init = tf.cast(tf.logical_not(tf.math.is_nan(answer)), tf.float32) + logits_aggregation = aggregation_classifier(pooled_output) + dist_aggregation = tfp.distributions.Categorical(logits=logits_aggregation) + # Index 0 corresponds to "no aggregation". + aggregation_ops_total_mass = tf.reduce_sum(dist_aggregation.probs_parameter()[:, 1:], axis=1) + # Cell selection examples according to current model. + is_pred_cell_selection = aggregation_ops_total_mass <= cell_selection_preference + # Examples with non-empty cell selection supervision. + is_cell_supervision_available = tf.reduce_sum(labels, axis=1) > 0 + aggregate_mask = tf.where( + tf.logical_and(is_pred_cell_selection, is_cell_supervision_available), + tf.zeros_like(aggregate_mask_init, dtype=tf.float32), + aggregate_mask_init, + ) + aggregate_mask = tf.stop_gradient(aggregate_mask) + return aggregate_mask + + +def _calculate_aggregation_loss_known( + logits_aggregation, aggregate_mask, aggregation_labels, use_answer_as_supervision, num_aggregation_labels +): + """ + Calculates aggregation loss when its type is known during training. + + In the weakly supervised setting, the only known information is that for cell selection examples, "no aggregation" + should be predicted. For other examples (those that require aggregation), no loss is accumulated. In the setting + where aggregation type is always known, standard cross entropy loss is accumulated for all examples + + Args: + logits_aggregation (`tf.Tensor` of shape `(batch_size, num_aggregation_labels)`): + Logits per aggregation operation. + aggregate_mask (`tf.Tensor` of shape `(batch_size, )`): + A mask set to 1 for examples that should use aggregation functions. + aggregation_labels (`tf.Tensor` of shape `(batch_size, )`): + Aggregation function id for every example in the batch. + use_answer_as_supervision (`bool`, *optional*): + Whether to use the answer as the only supervision for aggregation examples. + num_aggregation_labels (`int`, *optional*, defaults to 0): + The number of aggregation operators to predict. + + Returns: + aggregation_loss_known (`tf.Tensor` of shape `(batch_size,)`): Aggregation loss (when its type is known during + training) per example. + """ + if use_answer_as_supervision: + # Prepare "no aggregation" targets for cell selection examples. + target_aggregation = tf.zeros_like(aggregate_mask, dtype=tf.int32) + else: + # Use aggregation supervision as the target. + target_aggregation = aggregation_labels + + one_hot_labels = tf.one_hot(target_aggregation, depth=num_aggregation_labels, dtype=tf.float32) + log_probs = tf.nn.log_softmax(logits_aggregation, axis=-1) + + # [batch_size] + per_example_aggregation_intermediate = -tf.reduce_sum(one_hot_labels * log_probs, axis=-1) + if use_answer_as_supervision: + # Accumulate loss only for examples requiring cell selection + # (no aggregation). + return per_example_aggregation_intermediate * (1 - aggregate_mask) + else: + return per_example_aggregation_intermediate + + +def _calculate_aggregation_loss_unknown(logits_aggregation, aggregate_mask): + """ + Calculates aggregation loss in the case of answer supervision. + + Args: + logits_aggregation (`tf.Tensor` of shape `(batch_size, num_aggregation_labels)`): + Logits per aggregation operation. + aggregate_mask (`tf.Tensor` of shape `(batch_size, )`): + A mask set to 1 for examples that should use aggregation functions + + Returns: + aggregation_loss_unknown (`tf.Tensor` of shape `(batch_size,)`): Aggregation loss (in case of answer + supervision) per example. + """ + dist_aggregation = tfp.distributions.Categorical(logits=logits_aggregation) + # Index 0 corresponds to "no aggregation". + aggregation_ops_total_mass = tf.reduce_sum(dist_aggregation.probs_parameter()[:, 1:], axis=1) + # Predict some aggregation in case of an answer that needs aggregation. + # This increases the probability of all aggregation functions, in a way + # similar to MML, but without considering whether the function gives the + # correct answer. + return -tf.math.log(aggregation_ops_total_mass) * aggregate_mask + + +def _calculate_aggregation_loss( + logits_aggregation, + aggregate_mask, + aggregation_labels, + use_answer_as_supervision, + num_aggregation_labels, + aggregation_loss_weight, +): + """ + Calculates the aggregation loss per example. + + Args: + logits_aggregation (`tf.Tensor` of shape `(batch_size, num_aggregation_labels)`): + Logits per aggregation operation. + aggregate_mask (`tf.Tensor` of shape `(batch_size, )`): + A mask set to 1 for examples that should use aggregation functions. + aggregation_labels (`tf.Tensor` of shape `(batch_size, )`): + Aggregation function id for every example in the batch. + use_answer_as_supervision (`bool`, *optional*): + Whether to use the answer as the only supervision for aggregation examples. + num_aggregation_labels (`int`, *optional*, defaults to 0): + The number of aggregation operators to predict. + aggregation_loss_weight (`float`, *optional*, defaults to 1.0): + Importance weight for the aggregation loss. + + Returns: + aggregation_loss (`tf.Tensor` of shape `(batch_size,)`): Aggregation loss per example. + """ + per_example_aggregation_loss = _calculate_aggregation_loss_known( + logits_aggregation, aggregate_mask, aggregation_labels, use_answer_as_supervision, num_aggregation_labels + ) + + if use_answer_as_supervision: + # Add aggregation loss for numeric answers that need aggregation. + per_example_aggregation_loss += _calculate_aggregation_loss_unknown(logits_aggregation, aggregate_mask) + return aggregation_loss_weight * per_example_aggregation_loss + + +def _calculate_expected_result( + dist_per_cell, numeric_values, numeric_values_scale, input_mask_float, logits_aggregation, config +): + """ + Calculates the expected result given cell and aggregation probabilities. + + Args: + dist_per_cell (`tfp.distributions.Bernoulli`): + Cell selection distribution for each cell. + numeric_values (`tf.Tensor` of shape `(batch_size, seq_length)`): + Numeric values of every token. Nan for tokens which are not numeric values. + numeric_values_scale (`tf.Tensor` of shape `(batch_size, seq_length)`): + Scale of the numeric values of every token. + input_mask_float (`tf.Tensor` of shape `(batch_size, seq_length)`): + Mask for the table, without question tokens and table headers. + logits_aggregation (`tf.Tensor` of shape `(batch_size, num_aggregation_labels)`): + Logits per aggregation operation. + config ([`TapasConfig`]): + Model configuration class with all the hyperparameters of the model + + Returns: + expected_result (`tf.Tensor` of shape `(batch_size,)`): The expected result per example. + """ + if config.use_gumbel_for_cells: + gumbel_dist = tfp.distributions.RelaxedBernoulli( + # The token logits where already divided by the temperature and used for + # computing cell selection errors so we need to multiply it again here + config.temperature, + logits=dist_per_cell.logits_parameter() * config.temperature, + ) + scaled_probability_per_cell = gumbel_dist.sample() + else: + scaled_probability_per_cell = dist_per_cell.probs_parameter() + + # [batch_size, seq_length] + scaled_probability_per_cell = (scaled_probability_per_cell / numeric_values_scale) * input_mask_float + count_result = tf.reduce_sum(scaled_probability_per_cell, axis=1) + numeric_values_masked = tf.where( + tf.math.is_nan(numeric_values), tf.zeros_like(numeric_values), numeric_values + ) # Mask non-numeric table values to zero. + sum_result = tf.reduce_sum(scaled_probability_per_cell * numeric_values_masked, axis=1) + avg_approximation = config.average_approximation_function + if avg_approximation == AverageApproximationFunction.RATIO: + average_result = sum_result / (count_result + EPSILON_ZERO_DIVISION) + elif avg_approximation == AverageApproximationFunction.FIRST_ORDER: + # The sum of all probabilities exept that correspond to other cells + ex = tf.reduce_sum(scaled_probability_per_cell, axis=1, keepdims=True) - scaled_probability_per_cell + 1 + average_result = tf.reduce_sum(numeric_values_masked * scaled_probability_per_cell / ex, axis=1) + elif avg_approximation == AverageApproximationFunction.SECOND_ORDER: + # The sum of all probabilities exept that correspond to other cells + ex = tf.reduce_sum(scaled_probability_per_cell, axis=1, keepdims=True) - scaled_probability_per_cell + 1 + pointwise_var = scaled_probability_per_cell * (1 - scaled_probability_per_cell) + var = tf.reduce_sum(pointwise_var, axis=1, keepdims=True) - pointwise_var + multiplier = (var / tf.math.square(ex) + 1) / ex + average_result = tf.reduce_sum(numeric_values_masked * scaled_probability_per_cell * multiplier, axis=1) + else: + raise ValueError("Invalid average_approximation_function: %s", config.average_approximation_function) + + if config.use_gumbel_for_aggregation: + gumbel_dist = tfp.distributions.RelaxedOneHotCategorical( + config.aggregation_temperature, logits=logits_aggregation[:, 1:] + ) + # [batch_size, num_aggregation_labels - 1] + aggregation_op_only_probs = gumbel_dist.sample() + else: + # [batch_size, num_aggregation_labels - 1] + aggregation_op_only_probs = stable_softmax(logits_aggregation[:, 1:] / config.aggregation_temperature, axis=-1) + all_results = tf.concat( + [ + tf.expand_dims(sum_result, axis=1), + tf.expand_dims(average_result, axis=1), + tf.expand_dims(count_result, axis=1), + ], + axis=1, + ) + expected_result = tf.reduce_sum(all_results * aggregation_op_only_probs, axis=1) + return expected_result + + +def _calculate_regression_loss( + answer, + aggregate_mask, + dist_per_cell, + numeric_values, + numeric_values_scale, + input_mask_float, + logits_aggregation, + config, +): + """ + Calculates the regression loss per example. + + Args: + answer (`tf.Tensor` of shape `(batch_size,)`): + Answer for every example in the batch. Nan if there is no scalar answer. + aggregate_mask (`tf.Tensor` of shape `(batch_size,)`): + A mask set to 1 for examples that should use aggregation functions. + dist_per_cell (`torch.distributions.Bernoulli`): + Cell selection distribution for each cell. + numeric_values (`tf.Tensor` of shape `(batch_size, seq_length)`): + Numeric values of every token. Nan for tokens which are not numeric values. + numeric_values_scale (`tf.Tensor` of shape `(batch_size, seq_length)`): + Scale of the numeric values of every token. + input_mask_float (`tf.Tensor` of shape `(batch_size, seq_length)`): + Mask for the table, without question tokens and table headers. + logits_aggregation (`tf.Tensor` of shape `(batch_size, num_aggregation_labels)`): + Logits per aggregation operation. + config ([`TapasConfig`]): + Model configuration class with all the parameters of the model + + Returns: + per_example_answer_loss_scaled (`tf.Tensor` of shape `(batch_size,)`): Scales answer loss for each example in + the batch. large_answer_loss_mask (`tf.Tensor` of shape `(batch_size,)`): A mask which is 1 for examples for + which their answer loss is larger than the answer_loss_cutoff. + """ + # float32 (batch_size,) + expected_result = _calculate_expected_result( + dist_per_cell, numeric_values, numeric_values_scale, input_mask_float, logits_aggregation, config + ) + + # [batch_size] + answer_masked = tf.where(tf.math.is_nan(answer), tf.zeros_like(answer), answer) + + if config.use_normalized_answer_loss: + normalizer = tf.stop_gradient( + tf.math.maximum(tf.math.abs(expected_result), tf.math.abs(answer_masked)) + EPSILON_ZERO_DIVISION + ) + normalized_answer_masked = answer_masked / normalizer + normalized_expected_result = expected_result / normalizer + per_example_answer_loss = tf.compat.v1.losses.huber_loss( + normalized_answer_masked * aggregate_mask, + normalized_expected_result * aggregate_mask, + delta=tf.cast(1.0, tf.float32), + reduction=tf.losses.Reduction.NONE, + ) + else: + per_example_answer_loss = tf.compat.v1.losses.huber_loss( + answer_masked * aggregate_mask, + expected_result * aggregate_mask, + delta=tf.cast(config.huber_loss_delta, tf.float32), + reduction=tf.losses.Reduction.NONE, + ) + if config.answer_loss_cutoff is None: + large_answer_loss_mask = tf.ones_like(per_example_answer_loss, dtype=tf.float32) + else: + large_answer_loss_mask = tf.where( + per_example_answer_loss > config.answer_loss_cutoff, + tf.zeros_like(per_example_answer_loss, dtype=tf.float32), + tf.ones_like(per_example_answer_loss, dtype=tf.float32), + ) + per_example_answer_loss_scaled = config.answer_loss_importance * (per_example_answer_loss * aggregate_mask) + return per_example_answer_loss_scaled, large_answer_loss_mask diff --git a/parrot/lib/python3.10/site-packages/transformers/models/tapas/tokenization_tapas.py b/parrot/lib/python3.10/site-packages/transformers/models/tapas/tokenization_tapas.py new file mode 100644 index 0000000000000000000000000000000000000000..23fbd5300ed583cc57986e6bca7e39d5d593dcd6 --- /dev/null +++ b/parrot/lib/python3.10/site-packages/transformers/models/tapas/tokenization_tapas.py @@ -0,0 +1,2764 @@ +# coding=utf-8 +# Copyright 2020 Google Research and The HuggingFace Inc. team. +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. +""" Tokenization class for TAPAS model.""" + + +import collections +import datetime +import enum +import itertools +import math +import os +import re +import unicodedata +from dataclasses import dataclass +from typing import Callable, Dict, Generator, List, Optional, Tuple, Union + +import numpy as np + +from ...tokenization_utils import PreTrainedTokenizer, _is_control, _is_punctuation, _is_whitespace +from ...tokenization_utils_base import ( + ENCODE_KWARGS_DOCSTRING, + VERY_LARGE_INTEGER, + BatchEncoding, + EncodedInput, + PreTokenizedInput, + TextInput, +) +from ...utils import ExplicitEnum, PaddingStrategy, TensorType, add_end_docstrings, is_pandas_available, logging + + +if is_pandas_available(): + import pandas as pd + +logger = logging.get_logger(__name__) + + +VOCAB_FILES_NAMES = {"vocab_file": "vocab.txt"} + + +class TapasTruncationStrategy(ExplicitEnum): + """ + Possible values for the `truncation` argument in [`~TapasTokenizer.__call__`]. Useful for tab-completion in an IDE. + """ + + DROP_ROWS_TO_FIT = "drop_rows_to_fit" + DO_NOT_TRUNCATE = "do_not_truncate" + + +TableValue = collections.namedtuple("TokenValue", ["token", "column_id", "row_id"]) + + +@dataclass(frozen=True) +class TokenCoordinates: + column_index: int + row_index: int + token_index: int + + +@dataclass +class TokenizedTable: + rows: List[List[List[str]]] + selected_tokens: List[TokenCoordinates] + + +@dataclass(frozen=True) +class SerializedExample: + tokens: List[str] + column_ids: List[int] + row_ids: List[int] + segment_ids: List[int] + + +def _is_inner_wordpiece(token: str): + return token.startswith("##") + + +def load_vocab(vocab_file): + """Loads a vocabulary file into a dictionary.""" + vocab = collections.OrderedDict() + with open(vocab_file, "r", encoding="utf-8") as reader: + tokens = reader.readlines() + for index, token in enumerate(tokens): + token = token.rstrip("\n") + vocab[token] = index + return vocab + + +def whitespace_tokenize(text): + """Runs basic whitespace cleaning and splitting on a piece of text.""" + text = text.strip() + if not text: + return [] + tokens = text.split() + return tokens + + +TAPAS_ENCODE_PLUS_ADDITIONAL_KWARGS_DOCSTRING = r""" + add_special_tokens (`bool`, *optional*, defaults to `True`): + Whether or not to encode the sequences with the special tokens relative to their model. + padding (`bool`, `str` or [`~utils.PaddingStrategy`], *optional*, defaults to `False`): + Activates and controls padding. Accepts the following values: + + - `True` or `'longest'`: Pad to the longest sequence in the batch (or no padding if only a single + sequence if provided). + - `'max_length'`: Pad to a maximum length specified with the argument `max_length` or to the maximum + acceptable input length for the model if that argument is not provided. + - `False` or `'do_not_pad'` (default): No padding (i.e., can output a batch with sequences of different + lengths). + truncation (`bool`, `str` or [`TapasTruncationStrategy`], *optional*, defaults to `False`): + Activates and controls truncation. Accepts the following values: + + - `True` or `'drop_rows_to_fit'`: Truncate to a maximum length specified with the argument `max_length` + or to the maximum acceptable input length for the model if that argument is not provided. This will + truncate row by row, removing rows from the table. + - `False` or `'do_not_truncate'` (default): No truncation (i.e., can output batch with sequence lengths + greater than the model maximum admissible input size). + max_length (`int`, *optional*): + Controls the maximum length to use by one of the truncation/padding parameters. + + If left unset or set to `None`, this will use the predefined model maximum length if a maximum length + is required by one of the truncation/padding parameters. If the model has no specific maximum input + length (like XLNet) truncation/padding to a maximum length will be deactivated. + is_split_into_words (`bool`, *optional*, defaults to `False`): + Whether or not the input is already pre-tokenized (e.g., split into words). If set to `True`, the + tokenizer assumes the input is already split into words (for instance, by splitting it on whitespace) + which it will tokenize. This is useful for NER or token classification. + pad_to_multiple_of (`int`, *optional*): + If set will pad the sequence to a multiple of the provided value. This is especially useful to enable + the use of Tensor Cores on NVIDIA hardware with compute capability `>= 7.5` (Volta). + return_tensors (`str` or [`~utils.TensorType`], *optional*): + If set, will return tensors instead of list of python integers. Acceptable values are: + + - `'tf'`: Return TensorFlow `tf.constant` objects. + - `'pt'`: Return PyTorch `torch.Tensor` objects. + - `'np'`: Return Numpy `np.ndarray` objects. +""" + + +class TapasTokenizer(PreTrainedTokenizer): + r""" + Construct a TAPAS tokenizer. Based on WordPiece. Flattens a table and one or more related sentences to be used by + TAPAS models. + + This tokenizer inherits from [`PreTrainedTokenizer`] which contains most of the main methods. Users should refer to + this superclass for more information regarding those methods. [`TapasTokenizer`] creates several token type ids to + encode tabular structure. To be more precise, it adds 7 token type ids, in the following order: `segment_ids`, + `column_ids`, `row_ids`, `prev_labels`, `column_ranks`, `inv_column_ranks` and `numeric_relations`: + + - segment_ids: indicate whether a token belongs to the question (0) or the table (1). 0 for special tokens and + padding. + - column_ids: indicate to which column of the table a token belongs (starting from 1). Is 0 for all question + tokens, special tokens and padding. + - row_ids: indicate to which row of the table a token belongs (starting from 1). Is 0 for all question tokens, + special tokens and padding. Tokens of column headers are also 0. + - prev_labels: indicate whether a token was (part of) an answer to the previous question (1) or not (0). Useful in + a conversational setup (such as SQA). + - column_ranks: indicate the rank of a table token relative to a column, if applicable. For example, if you have a + column "number of movies" with values 87, 53 and 69, then the column ranks of these tokens are 3, 1 and 2 + respectively. 0 for all question tokens, special tokens and padding. + - inv_column_ranks: indicate the inverse rank of a table token relative to a column, if applicable. For example, if + you have a column "number of movies" with values 87, 53 and 69, then the inverse column ranks of these tokens are + 1, 3 and 2 respectively. 0 for all question tokens, special tokens and padding. + - numeric_relations: indicate numeric relations between the question and the tokens of the table. 0 for all + question tokens, special tokens and padding. + + [`TapasTokenizer`] runs end-to-end tokenization on a table and associated sentences: punctuation splitting and + wordpiece. + + Args: + vocab_file (`str`): + File containing the vocabulary. + do_lower_case (`bool`, *optional*, defaults to `True`): + Whether or not to lowercase the input when tokenizing. + do_basic_tokenize (`bool`, *optional*, defaults to `True`): + Whether or not to do basic tokenization before WordPiece. + never_split (`Iterable`, *optional*): + Collection of tokens which will never be split during tokenization. Only has an effect when + `do_basic_tokenize=True` + unk_token (`str`, *optional*, defaults to `"[UNK]"`): + The unknown token. A token that is not in the vocabulary cannot be converted to an ID and is set to be this + token instead. + sep_token (`str`, *optional*, defaults to `"[SEP]"`): + The separator token, which is used when building a sequence from multiple sequences, e.g. two sequences for + sequence classification or for a text and a question for question answering. It is also used as the last + token of a sequence built with special tokens. + pad_token (`str`, *optional*, defaults to `"[PAD]"`): + The token used for padding, for example when batching sequences of different lengths. + cls_token (`str`, *optional*, defaults to `"[CLS]"`): + The classifier token which is used when doing sequence classification (classification of the whole sequence + instead of per-token classification). It is the first token of the sequence when built with special tokens. + mask_token (`str`, *optional*, defaults to `"[MASK]"`): + The token used for masking values. This is the token used when training this model with masked language + modeling. This is the token which the model will try to predict. + empty_token (`str`, *optional*, defaults to `"[EMPTY]"`): + The token used for empty cell values in a table. Empty cell values include "", "n/a", "nan" and "?". + tokenize_chinese_chars (`bool`, *optional*, defaults to `True`): + Whether or not to tokenize Chinese characters. This should likely be deactivated for Japanese (see this + [issue](https://github.com/huggingface/transformers/issues/328)). + strip_accents (`bool`, *optional*): + Whether or not to strip all accents. If this option is not specified, then it will be determined by the + value for `lowercase` (as in the original BERT). + cell_trim_length (`int`, *optional*, defaults to -1): + If > 0: Trim cells so that the length is <= this value. Also disables further cell trimming, should thus be + used with `truncation` set to `True`. + max_column_id (`int`, *optional*): + Max column id to extract. + max_row_id (`int`, *optional*): + Max row id to extract. + strip_column_names (`bool`, *optional*, defaults to `False`): + Whether to add empty strings instead of column names. + update_answer_coordinates (`bool`, *optional*, defaults to `False`): + Whether to recompute the answer coordinates from the answer text. + min_question_length (`int`, *optional*): + Minimum length of each question in terms of tokens (will be skipped otherwise). + max_question_length (`int`, *optional*): + Maximum length of each question in terms of tokens (will be skipped otherwise). + """ + + vocab_files_names = VOCAB_FILES_NAMES + + def __init__( + self, + vocab_file, + do_lower_case=True, + do_basic_tokenize=True, + never_split=None, + unk_token="[UNK]", + sep_token="[SEP]", + pad_token="[PAD]", + cls_token="[CLS]", + mask_token="[MASK]", + empty_token="[EMPTY]", + tokenize_chinese_chars=True, + strip_accents=None, + cell_trim_length: int = -1, + max_column_id: int = None, + max_row_id: int = None, + strip_column_names: bool = False, + update_answer_coordinates: bool = False, + min_question_length=None, + max_question_length=None, + model_max_length: int = 512, + additional_special_tokens: Optional[List[str]] = None, + **kwargs, + ): + if not is_pandas_available(): + raise ImportError("Pandas is required for the TAPAS tokenizer.") + + if additional_special_tokens is not None: + if empty_token not in additional_special_tokens: + additional_special_tokens.append(empty_token) + else: + additional_special_tokens = [empty_token] + + if not os.path.isfile(vocab_file): + raise ValueError( + f"Can't find a vocabulary file at path '{vocab_file}'. To load the vocabulary from a Google pretrained" + " model use `tokenizer = BertTokenizer.from_pretrained(PRETRAINED_MODEL_NAME)`" + ) + self.vocab = load_vocab(vocab_file) + self.ids_to_tokens = collections.OrderedDict([(ids, tok) for tok, ids in self.vocab.items()]) + self.do_basic_tokenize = do_basic_tokenize + if do_basic_tokenize: + self.basic_tokenizer = BasicTokenizer( + do_lower_case=do_lower_case, + never_split=never_split, + tokenize_chinese_chars=tokenize_chinese_chars, + strip_accents=strip_accents, + ) + self.wordpiece_tokenizer = WordpieceTokenizer(vocab=self.vocab, unk_token=str(unk_token)) + + # Additional properties + self.cell_trim_length = cell_trim_length + self.max_column_id = ( + max_column_id + if max_column_id is not None + else model_max_length + if model_max_length is not None + else VERY_LARGE_INTEGER + ) + self.max_row_id = ( + max_row_id + if max_row_id is not None + else model_max_length + if model_max_length is not None + else VERY_LARGE_INTEGER + ) + self.strip_column_names = strip_column_names + self.update_answer_coordinates = update_answer_coordinates + self.min_question_length = min_question_length + self.max_question_length = max_question_length + + super().__init__( + do_lower_case=do_lower_case, + do_basic_tokenize=do_basic_tokenize, + never_split=never_split, + unk_token=unk_token, + sep_token=sep_token, + pad_token=pad_token, + cls_token=cls_token, + mask_token=mask_token, + empty_token=empty_token, + tokenize_chinese_chars=tokenize_chinese_chars, + strip_accents=strip_accents, + cell_trim_length=cell_trim_length, + max_column_id=max_column_id, + max_row_id=max_row_id, + strip_column_names=strip_column_names, + update_answer_coordinates=update_answer_coordinates, + min_question_length=min_question_length, + max_question_length=max_question_length, + model_max_length=model_max_length, + additional_special_tokens=additional_special_tokens, + **kwargs, + ) + + @property + def do_lower_case(self): + return self.basic_tokenizer.do_lower_case + + @property + def vocab_size(self): + return len(self.vocab) + + def get_vocab(self): + return dict(self.vocab, **self.added_tokens_encoder) + + def _tokenize(self, text): + if format_text(text) == EMPTY_TEXT: + return [self.additional_special_tokens[0]] + split_tokens = [] + if self.do_basic_tokenize: + for token in self.basic_tokenizer.tokenize(text, never_split=self.all_special_tokens): + # If the token is part of the never_split set + if token in self.basic_tokenizer.never_split: + split_tokens.append(token) + else: + split_tokens += self.wordpiece_tokenizer.tokenize(token) + else: + split_tokens = self.wordpiece_tokenizer.tokenize(text) + return split_tokens + + def _convert_token_to_id(self, token): + """Converts a token (str) in an id using the vocab.""" + return self.vocab.get(token, self.vocab.get(self.unk_token)) + + def _convert_id_to_token(self, index): + """Converts an index (integer) in a token (str) using the vocab.""" + return self.ids_to_tokens.get(index, self.unk_token) + + def convert_tokens_to_string(self, tokens): + """Converts a sequence of tokens (string) in a single string.""" + out_string = " ".join(tokens).replace(" ##", "").strip() + return out_string + + def save_vocabulary(self, save_directory: str, filename_prefix: Optional[str] = None) -> Tuple[str]: + index = 0 + if os.path.isdir(save_directory): + vocab_file = os.path.join( + save_directory, (filename_prefix + "-" if filename_prefix else "") + VOCAB_FILES_NAMES["vocab_file"] + ) + else: + vocab_file = (filename_prefix + "-" if filename_prefix else "") + save_directory + with open(vocab_file, "w", encoding="utf-8") as writer: + for token, token_index in sorted(self.vocab.items(), key=lambda kv: kv[1]): + if index != token_index: + logger.warning( + f"Saving vocabulary to {vocab_file}: vocabulary indices are not consecutive." + " Please check that the vocabulary is not corrupted!" + ) + index = token_index + writer.write(token + "\n") + index += 1 + return (vocab_file,) + + def create_attention_mask_from_sequences(self, query_ids: List[int], table_values: List[TableValue]) -> List[int]: + """ + Creates the attention mask according to the query token IDs and a list of table values. + + Args: + query_ids (`List[int]`): list of token IDs corresponding to the ID. + table_values (`List[TableValue]`): lift of table values, which are named tuples containing the + token value, the column ID and the row ID of said token. + + Returns: + `List[int]`: List of ints containing the attention mask values. + """ + return [1] * (1 + len(query_ids) + 1 + len(table_values)) + + def create_segment_token_type_ids_from_sequences( + self, query_ids: List[int], table_values: List[TableValue] + ) -> List[int]: + """ + Creates the segment token type IDs according to the query token IDs and a list of table values. + + Args: + query_ids (`List[int]`): list of token IDs corresponding to the ID. + table_values (`List[TableValue]`): lift of table values, which are named tuples containing the + token value, the column ID and the row ID of said token. + + Returns: + `List[int]`: List of ints containing the segment token type IDs values. + """ + table_ids = list(zip(*table_values))[0] if table_values else [] + return [0] * (1 + len(query_ids) + 1) + [1] * len(table_ids) + + def create_column_token_type_ids_from_sequences( + self, query_ids: List[int], table_values: List[TableValue] + ) -> List[int]: + """ + Creates the column token type IDs according to the query token IDs and a list of table values. + + Args: + query_ids (`List[int]`): list of token IDs corresponding to the ID. + table_values (`List[TableValue]`): lift of table values, which are named tuples containing the + token value, the column ID and the row ID of said token. + + Returns: + `List[int]`: List of ints containing the column token type IDs values. + """ + table_column_ids = list(zip(*table_values))[1] if table_values else [] + return [0] * (1 + len(query_ids) + 1) + list(table_column_ids) + + def create_row_token_type_ids_from_sequences( + self, query_ids: List[int], table_values: List[TableValue] + ) -> List[int]: + """ + Creates the row token type IDs according to the query token IDs and a list of table values. + + Args: + query_ids (`List[int]`): list of token IDs corresponding to the ID. + table_values (`List[TableValue]`): lift of table values, which are named tuples containing the + token value, the column ID and the row ID of said token. + + Returns: + `List[int]`: List of ints containing the row token type IDs values. + """ + table_row_ids = list(zip(*table_values))[2] if table_values else [] + return [0] * (1 + len(query_ids) + 1) + list(table_row_ids) + + def build_inputs_with_special_tokens( + self, token_ids_0: List[int], token_ids_1: Optional[List[int]] = None + ) -> List[int]: + """ + Build model inputs from a question and flattened table for question answering or sequence classification tasks + by concatenating and adding special tokens. + + Args: + token_ids_0 (`List[int]`): The ids of the question. + token_ids_1 (`List[int]`, *optional*): The ids of the flattened table. + + Returns: + `List[int]`: The model input with special tokens. + """ + if token_ids_1 is None: + raise ValueError("With TAPAS, you must provide both question IDs and table IDs.") + + return [self.cls_token_id] + token_ids_0 + [self.sep_token_id] + token_ids_1 + + def get_special_tokens_mask( + self, token_ids_0: List[int], token_ids_1: Optional[List[int]] = None, already_has_special_tokens: bool = False + ) -> List[int]: + """ + Retrieve sequence ids from a token list that has no special tokens added. This method is called when adding + special tokens using the tokenizer `prepare_for_model` method. + + Args: + token_ids_0 (`List[int]`): + List of question IDs. + token_ids_1 (`List[int]`, *optional*): + List of flattened table IDs. + already_has_special_tokens (`bool`, *optional*, defaults to `False`): + Whether or not the token list is already formatted with special tokens for the model. + + Returns: + `List[int]`: A list of integers in the range [0, 1]: 1 for a special token, 0 for a sequence token. + """ + + if already_has_special_tokens: + return super().get_special_tokens_mask( + token_ids_0=token_ids_0, token_ids_1=token_ids_1, already_has_special_tokens=True + ) + + if token_ids_1 is not None: + return [1] + ([0] * len(token_ids_0)) + [1] + ([0] * len(token_ids_1)) + return [1] + ([0] * len(token_ids_0)) + [1] + + @add_end_docstrings(TAPAS_ENCODE_PLUS_ADDITIONAL_KWARGS_DOCSTRING) + def __call__( + self, + table: "pd.DataFrame", + queries: Optional[ + Union[ + TextInput, + PreTokenizedInput, + EncodedInput, + List[TextInput], + List[PreTokenizedInput], + List[EncodedInput], + ] + ] = None, + answer_coordinates: Optional[Union[List[Tuple], List[List[Tuple]]]] = None, + answer_text: Optional[Union[List[TextInput], List[List[TextInput]]]] = None, + add_special_tokens: bool = True, + padding: Union[bool, str, PaddingStrategy] = False, + truncation: Union[bool, str, TapasTruncationStrategy] = False, + max_length: Optional[int] = None, + pad_to_multiple_of: Optional[int] = None, + return_tensors: Optional[Union[str, TensorType]] = None, + return_token_type_ids: Optional[bool] = None, + return_attention_mask: Optional[bool] = None, + return_overflowing_tokens: bool = False, + return_special_tokens_mask: bool = False, + return_offsets_mapping: bool = False, + return_length: bool = False, + verbose: bool = True, + **kwargs, + ) -> BatchEncoding: + """ + Main method to tokenize and prepare for the model one or several sequence(s) related to a table. + + Args: + table (`pd.DataFrame`): + Table containing tabular data. Note that all cell values must be text. Use *.astype(str)* on a Pandas + dataframe to convert it to string. + queries (`str` or `List[str]`): + Question or batch of questions related to a table to be encoded. Note that in case of a batch, all + questions must refer to the **same** table. + answer_coordinates (`List[Tuple]` or `List[List[Tuple]]`, *optional*): + Answer coordinates of each table-question pair in the batch. In case only a single table-question pair + is provided, then the answer_coordinates must be a single list of one or more tuples. Each tuple must + be a (row_index, column_index) pair. The first data row (not the column header row) has index 0. The + first column has index 0. In case a batch of table-question pairs is provided, then the + answer_coordinates must be a list of lists of tuples (each list corresponding to a single + table-question pair). + answer_text (`List[str]` or `List[List[str]]`, *optional*): + Answer text of each table-question pair in the batch. In case only a single table-question pair is + provided, then the answer_text must be a single list of one or more strings. Each string must be the + answer text of a corresponding answer coordinate. In case a batch of table-question pairs is provided, + then the answer_coordinates must be a list of lists of strings (each list corresponding to a single + table-question pair). + """ + assert isinstance(table, pd.DataFrame), "Table must be of type pd.DataFrame" + + # Input type checking for clearer error + valid_query = False + + # Check that query has a valid type + if queries is None or isinstance(queries, str): + valid_query = True + elif isinstance(queries, (list, tuple)): + if len(queries) == 0 or isinstance(queries[0], str): + valid_query = True + + if not valid_query: + raise ValueError( + "queries input must of type `str` (single example), `List[str]` (batch or single pretokenized" + " example). " + ) + is_batched = isinstance(queries, (list, tuple)) + + if is_batched: + return self.batch_encode_plus( + table=table, + queries=queries, + answer_coordinates=answer_coordinates, + answer_text=answer_text, + add_special_tokens=add_special_tokens, + padding=padding, + truncation=truncation, + max_length=max_length, + pad_to_multiple_of=pad_to_multiple_of, + return_tensors=return_tensors, + return_token_type_ids=return_token_type_ids, + return_attention_mask=return_attention_mask, + return_overflowing_tokens=return_overflowing_tokens, + return_special_tokens_mask=return_special_tokens_mask, + return_offsets_mapping=return_offsets_mapping, + return_length=return_length, + verbose=verbose, + **kwargs, + ) + else: + return self.encode_plus( + table=table, + query=queries, + answer_coordinates=answer_coordinates, + answer_text=answer_text, + add_special_tokens=add_special_tokens, + padding=padding, + truncation=truncation, + max_length=max_length, + pad_to_multiple_of=pad_to_multiple_of, + return_tensors=return_tensors, + return_token_type_ids=return_token_type_ids, + return_attention_mask=return_attention_mask, + return_overflowing_tokens=return_overflowing_tokens, + return_special_tokens_mask=return_special_tokens_mask, + return_offsets_mapping=return_offsets_mapping, + return_length=return_length, + verbose=verbose, + **kwargs, + ) + + @add_end_docstrings(ENCODE_KWARGS_DOCSTRING, TAPAS_ENCODE_PLUS_ADDITIONAL_KWARGS_DOCSTRING) + def batch_encode_plus( + self, + table: "pd.DataFrame", + queries: Optional[ + Union[ + List[TextInput], + List[PreTokenizedInput], + List[EncodedInput], + ] + ] = None, + answer_coordinates: Optional[List[List[Tuple]]] = None, + answer_text: Optional[List[List[TextInput]]] = None, + add_special_tokens: bool = True, + padding: Union[bool, str, PaddingStrategy] = False, + truncation: Union[bool, str, TapasTruncationStrategy] = False, + max_length: Optional[int] = None, + pad_to_multiple_of: Optional[int] = None, + return_tensors: Optional[Union[str, TensorType]] = None, + return_token_type_ids: Optional[bool] = None, + return_attention_mask: Optional[bool] = None, + return_overflowing_tokens: bool = False, + return_special_tokens_mask: bool = False, + return_offsets_mapping: bool = False, + return_length: bool = False, + verbose: bool = True, + **kwargs, + ) -> BatchEncoding: + """ + Prepare a table and a list of strings for the model. + + + + This method is deprecated, `__call__` should be used instead. + + + + Args: + table (`pd.DataFrame`): + Table containing tabular data. Note that all cell values must be text. Use *.astype(str)* on a Pandas + dataframe to convert it to string. + queries (`List[str]`): + Batch of questions related to a table to be encoded. Note that all questions must refer to the **same** + table. + answer_coordinates (`List[Tuple]` or `List[List[Tuple]]`, *optional*): + Answer coordinates of each table-question pair in the batch. Each tuple must be a (row_index, + column_index) pair. The first data row (not the column header row) has index 0. The first column has + index 0. The answer_coordinates must be a list of lists of tuples (each list corresponding to a single + table-question pair). + answer_text (`List[str]` or `List[List[str]]`, *optional*): + Answer text of each table-question pair in the batch. In case a batch of table-question pairs is + provided, then the answer_coordinates must be a list of lists of strings (each list corresponding to a + single table-question pair). Each string must be the answer text of a corresponding answer coordinate. + """ + if return_token_type_ids is not None and not add_special_tokens: + raise ValueError( + "Asking to return token_type_ids while setting add_special_tokens to False " + "results in an undefined behavior. Please set add_special_tokens to True or " + "set return_token_type_ids to None." + ) + + if (answer_coordinates and not answer_text) or (not answer_coordinates and answer_text): + raise ValueError("In case you provide answers, both answer_coordinates and answer_text should be provided") + elif answer_coordinates is None and answer_text is None: + answer_coordinates = answer_text = [None] * len(queries) + + if "is_split_into_words" in kwargs: + raise NotImplementedError("Currently TapasTokenizer only supports questions as strings.") + + if return_offsets_mapping: + raise NotImplementedError( + "return_offset_mapping is not available when using Python tokenizers. " + "To use this feature, change your tokenizer to one deriving from " + "transformers.PreTrainedTokenizerFast." + ) + + return self._batch_encode_plus( + table=table, + queries=queries, + answer_coordinates=answer_coordinates, + answer_text=answer_text, + add_special_tokens=add_special_tokens, + padding=padding, + truncation=truncation, + max_length=max_length, + pad_to_multiple_of=pad_to_multiple_of, + return_tensors=return_tensors, + return_token_type_ids=return_token_type_ids, + return_attention_mask=return_attention_mask, + return_overflowing_tokens=return_overflowing_tokens, + return_special_tokens_mask=return_special_tokens_mask, + return_offsets_mapping=return_offsets_mapping, + return_length=return_length, + verbose=verbose, + **kwargs, + ) + + def _get_question_tokens(self, query): + """Tokenizes the query, taking into account the max and min question length.""" + + query_tokens = self.tokenize(query) + if self.max_question_length is not None and len(query_tokens) > self.max_question_length: + logger.warning("Skipping query as its tokens are longer than the max question length") + return "", [] + if self.min_question_length is not None and len(query_tokens) < self.min_question_length: + logger.warning("Skipping query as its tokens are shorter than the min question length") + return "", [] + + return query, query_tokens + + def _batch_encode_plus( + self, + table, + queries: Union[ + List[TextInput], + List[PreTokenizedInput], + List[EncodedInput], + ], + answer_coordinates: Optional[List[List[Tuple]]] = None, + answer_text: Optional[List[List[TextInput]]] = None, + add_special_tokens: bool = True, + padding: Union[bool, str, PaddingStrategy] = False, + truncation: Union[bool, str, TapasTruncationStrategy] = False, + max_length: Optional[int] = None, + pad_to_multiple_of: Optional[int] = None, + return_tensors: Optional[Union[str, TensorType]] = None, + return_token_type_ids: Optional[bool] = True, + return_attention_mask: Optional[bool] = None, + return_overflowing_tokens: bool = False, + return_special_tokens_mask: bool = False, + return_offsets_mapping: bool = False, + return_length: bool = False, + verbose: bool = True, + **kwargs, + ) -> BatchEncoding: + table_tokens = self._tokenize_table(table) + + queries_tokens = [] + for idx, query in enumerate(queries): + query, query_tokens = self._get_question_tokens(query) + queries[idx] = query + queries_tokens.append(query_tokens) + + batch_outputs = self._batch_prepare_for_model( + table, + queries, + tokenized_table=table_tokens, + queries_tokens=queries_tokens, + answer_coordinates=answer_coordinates, + padding=padding, + truncation=truncation, + answer_text=answer_text, + add_special_tokens=add_special_tokens, + max_length=max_length, + pad_to_multiple_of=pad_to_multiple_of, + return_tensors=return_tensors, + prepend_batch_axis=True, + return_attention_mask=return_attention_mask, + return_token_type_ids=return_token_type_ids, + return_overflowing_tokens=return_overflowing_tokens, + return_special_tokens_mask=return_special_tokens_mask, + return_length=return_length, + verbose=verbose, + ) + + return BatchEncoding(batch_outputs) + + def _batch_prepare_for_model( + self, + raw_table: "pd.DataFrame", + raw_queries: Union[ + List[TextInput], + List[PreTokenizedInput], + List[EncodedInput], + ], + tokenized_table: Optional[TokenizedTable] = None, + queries_tokens: Optional[List[List[str]]] = None, + answer_coordinates: Optional[List[List[Tuple]]] = None, + answer_text: Optional[List[List[TextInput]]] = None, + add_special_tokens: bool = True, + padding: Union[bool, str, PaddingStrategy] = False, + truncation: Union[bool, str, TapasTruncationStrategy] = False, + max_length: Optional[int] = None, + pad_to_multiple_of: Optional[int] = None, + return_tensors: Optional[Union[str, TensorType]] = None, + return_token_type_ids: Optional[bool] = True, + return_attention_mask: Optional[bool] = True, + return_special_tokens_mask: bool = False, + return_offsets_mapping: bool = False, + return_length: bool = False, + verbose: bool = True, + prepend_batch_axis: bool = False, + **kwargs, + ) -> BatchEncoding: + batch_outputs = {} + + for index, example in enumerate(zip(raw_queries, queries_tokens, answer_coordinates, answer_text)): + raw_query, query_tokens, answer_coords, answer_txt = example + outputs = self.prepare_for_model( + raw_table, + raw_query, + tokenized_table=tokenized_table, + query_tokens=query_tokens, + answer_coordinates=answer_coords, + answer_text=answer_txt, + add_special_tokens=add_special_tokens, + padding=PaddingStrategy.DO_NOT_PAD.value, # we pad in batch afterwards + truncation=truncation, + max_length=max_length, + pad_to_multiple_of=None, # we pad in batch afterwards + return_attention_mask=False, # we pad in batch afterwards + return_token_type_ids=return_token_type_ids, + return_special_tokens_mask=return_special_tokens_mask, + return_length=return_length, + return_tensors=None, # We convert the whole batch to tensors at the end + prepend_batch_axis=False, + verbose=verbose, + prev_answer_coordinates=answer_coordinates[index - 1] if index != 0 else None, + prev_answer_text=answer_text[index - 1] if index != 0 else None, + ) + + for key, value in outputs.items(): + if key not in batch_outputs: + batch_outputs[key] = [] + batch_outputs[key].append(value) + + batch_outputs = self.pad( + batch_outputs, + padding=padding, + max_length=max_length, + pad_to_multiple_of=pad_to_multiple_of, + return_attention_mask=return_attention_mask, + ) + + batch_outputs = BatchEncoding(batch_outputs, tensor_type=return_tensors) + + return batch_outputs + + @add_end_docstrings(ENCODE_KWARGS_DOCSTRING) + def encode( + self, + table: "pd.DataFrame", + query: Optional[ + Union[ + TextInput, + PreTokenizedInput, + EncodedInput, + ] + ] = None, + add_special_tokens: bool = True, + padding: Union[bool, str, PaddingStrategy] = False, + truncation: Union[bool, str, TapasTruncationStrategy] = False, + max_length: Optional[int] = None, + return_tensors: Optional[Union[str, TensorType]] = None, + **kwargs, + ) -> List[int]: + """ + Prepare a table and a string for the model. This method does not return token type IDs, attention masks, etc. + which are necessary for the model to work correctly. Use that method if you want to build your processing on + your own, otherwise refer to `__call__`. + + Args: + table (`pd.DataFrame`): + Table containing tabular data. Note that all cell values must be text. Use *.astype(str)* on a Pandas + dataframe to convert it to string. + query (`str` or `List[str]`): + Question related to a table to be encoded. + """ + encoded_inputs = self.encode_plus( + table, + query=query, + add_special_tokens=add_special_tokens, + padding=padding, + truncation=truncation, + max_length=max_length, + return_tensors=return_tensors, + **kwargs, + ) + + return encoded_inputs["input_ids"] + + @add_end_docstrings(ENCODE_KWARGS_DOCSTRING, TAPAS_ENCODE_PLUS_ADDITIONAL_KWARGS_DOCSTRING) + def encode_plus( + self, + table: "pd.DataFrame", + query: Optional[ + Union[ + TextInput, + PreTokenizedInput, + EncodedInput, + ] + ] = None, + answer_coordinates: Optional[List[Tuple]] = None, + answer_text: Optional[List[TextInput]] = None, + add_special_tokens: bool = True, + padding: Union[bool, str, PaddingStrategy] = False, + truncation: Union[bool, str, TapasTruncationStrategy] = False, + max_length: Optional[int] = None, + pad_to_multiple_of: Optional[int] = None, + return_tensors: Optional[Union[str, TensorType]] = None, + return_token_type_ids: Optional[bool] = None, + return_attention_mask: Optional[bool] = None, + return_special_tokens_mask: bool = False, + return_offsets_mapping: bool = False, + return_length: bool = False, + verbose: bool = True, + **kwargs, + ) -> BatchEncoding: + """ + Prepare a table and a string for the model. + + Args: + table (`pd.DataFrame`): + Table containing tabular data. Note that all cell values must be text. Use *.astype(str)* on a Pandas + dataframe to convert it to string. + query (`str` or `List[str]`): + Question related to a table to be encoded. + answer_coordinates (`List[Tuple]` or `List[List[Tuple]]`, *optional*): + Answer coordinates of each table-question pair in the batch. The answer_coordinates must be a single + list of one or more tuples. Each tuple must be a (row_index, column_index) pair. The first data row + (not the column header row) has index 0. The first column has index 0. + answer_text (`List[str]` or `List[List[str]]`, *optional*): + Answer text of each table-question pair in the batch. The answer_text must be a single list of one or + more strings. Each string must be the answer text of a corresponding answer coordinate. + """ + if return_token_type_ids is not None and not add_special_tokens: + raise ValueError( + "Asking to return token_type_ids while setting add_special_tokens to False " + "results in an undefined behavior. Please set add_special_tokens to True or " + "set return_token_type_ids to None." + ) + + if (answer_coordinates and not answer_text) or (not answer_coordinates and answer_text): + raise ValueError("In case you provide answers, both answer_coordinates and answer_text should be provided") + + if "is_split_into_words" in kwargs: + raise NotImplementedError("Currently TapasTokenizer only supports questions as strings.") + + if return_offsets_mapping: + raise NotImplementedError( + "return_offset_mapping is not available when using Python tokenizers. " + "To use this feature, change your tokenizer to one deriving from " + "transformers.PreTrainedTokenizerFast." + ) + + return self._encode_plus( + table=table, + query=query, + answer_coordinates=answer_coordinates, + answer_text=answer_text, + add_special_tokens=add_special_tokens, + truncation=truncation, + padding=padding, + max_length=max_length, + pad_to_multiple_of=pad_to_multiple_of, + return_tensors=return_tensors, + return_token_type_ids=return_token_type_ids, + return_attention_mask=return_attention_mask, + return_special_tokens_mask=return_special_tokens_mask, + return_offsets_mapping=return_offsets_mapping, + return_length=return_length, + verbose=verbose, + **kwargs, + ) + + def _encode_plus( + self, + table: "pd.DataFrame", + query: Union[ + TextInput, + PreTokenizedInput, + EncodedInput, + ], + answer_coordinates: Optional[List[Tuple]] = None, + answer_text: Optional[List[TextInput]] = None, + add_special_tokens: bool = True, + padding: Union[bool, str, PaddingStrategy] = False, + truncation: Union[bool, str, TapasTruncationStrategy] = False, + max_length: Optional[int] = None, + pad_to_multiple_of: Optional[int] = None, + return_tensors: Optional[Union[str, TensorType]] = None, + return_token_type_ids: Optional[bool] = True, + return_attention_mask: Optional[bool] = True, + return_special_tokens_mask: bool = False, + return_offsets_mapping: bool = False, + return_length: bool = False, + verbose: bool = True, + **kwargs, + ): + if query is None: + query = "" + logger.warning( + "TAPAS is a question answering model but you have not passed a query. Please be aware that the " + "model will probably not behave correctly." + ) + + table_tokens = self._tokenize_table(table) + query, query_tokens = self._get_question_tokens(query) + + return self.prepare_for_model( + table, + query, + tokenized_table=table_tokens, + query_tokens=query_tokens, + answer_coordinates=answer_coordinates, + answer_text=answer_text, + add_special_tokens=add_special_tokens, + truncation=truncation, + padding=padding, + max_length=max_length, + pad_to_multiple_of=pad_to_multiple_of, + return_tensors=return_tensors, + prepend_batch_axis=True, + return_attention_mask=return_attention_mask, + return_token_type_ids=return_token_type_ids, + return_special_tokens_mask=return_special_tokens_mask, + return_length=return_length, + verbose=verbose, + ) + + @add_end_docstrings(ENCODE_KWARGS_DOCSTRING, TAPAS_ENCODE_PLUS_ADDITIONAL_KWARGS_DOCSTRING) + def prepare_for_model( + self, + raw_table: "pd.DataFrame", + raw_query: Union[ + TextInput, + PreTokenizedInput, + EncodedInput, + ], + tokenized_table: Optional[TokenizedTable] = None, + query_tokens: Optional[TokenizedTable] = None, + answer_coordinates: Optional[List[Tuple]] = None, + answer_text: Optional[List[TextInput]] = None, + add_special_tokens: bool = True, + padding: Union[bool, str, PaddingStrategy] = False, + truncation: Union[bool, str, TapasTruncationStrategy] = False, + max_length: Optional[int] = None, + pad_to_multiple_of: Optional[int] = None, + return_tensors: Optional[Union[str, TensorType]] = None, + return_token_type_ids: Optional[bool] = True, + return_attention_mask: Optional[bool] = True, + return_special_tokens_mask: bool = False, + return_offsets_mapping: bool = False, + return_length: bool = False, + verbose: bool = True, + prepend_batch_axis: bool = False, + **kwargs, + ) -> BatchEncoding: + """ + Prepares a sequence of input id so that it can be used by the model. It adds special tokens, truncates + sequences if overflowing while taking into account the special tokens. + + Args: + raw_table (`pd.DataFrame`): + The original table before any transformation (like tokenization) was applied to it. + raw_query (`TextInput` or `PreTokenizedInput` or `EncodedInput`): + The original query before any transformation (like tokenization) was applied to it. + tokenized_table (`TokenizedTable`): + The table after tokenization. + query_tokens (`List[str]`): + The query after tokenization. + answer_coordinates (`List[Tuple]` or `List[List[Tuple]]`, *optional*): + Answer coordinates of each table-question pair in the batch. The answer_coordinates must be a single + list of one or more tuples. Each tuple must be a (row_index, column_index) pair. The first data row + (not the column header row) has index 0. The first column has index 0. + answer_text (`List[str]` or `List[List[str]]`, *optional*): + Answer text of each table-question pair in the batch. The answer_text must be a single list of one or + more strings. Each string must be the answer text of a corresponding answer coordinate. + """ + if isinstance(padding, bool): + if padding and (max_length is not None or pad_to_multiple_of is not None): + padding = PaddingStrategy.MAX_LENGTH + else: + padding = PaddingStrategy.DO_NOT_PAD + elif not isinstance(padding, PaddingStrategy): + padding = PaddingStrategy(padding) + + if isinstance(truncation, bool): + if truncation: + truncation = TapasTruncationStrategy.DROP_ROWS_TO_FIT + else: + truncation = TapasTruncationStrategy.DO_NOT_TRUNCATE + elif not isinstance(truncation, TapasTruncationStrategy): + truncation = TapasTruncationStrategy(truncation) + + encoded_inputs = {} + + is_part_of_batch = False + prev_answer_coordinates, prev_answer_text = None, None + if "prev_answer_coordinates" in kwargs and "prev_answer_text" in kwargs: + is_part_of_batch = True + prev_answer_coordinates = kwargs["prev_answer_coordinates"] + prev_answer_text = kwargs["prev_answer_text"] + + num_rows = self._get_num_rows(raw_table, truncation != TapasTruncationStrategy.DO_NOT_TRUNCATE) + num_columns = self._get_num_columns(raw_table) + _, _, num_tokens = self._get_table_boundaries(tokenized_table) + + if truncation != TapasTruncationStrategy.DO_NOT_TRUNCATE: + num_rows, num_tokens = self._get_truncated_table_rows( + query_tokens, tokenized_table, num_rows, num_columns, max_length, truncation_strategy=truncation + ) + table_data = list(self._get_table_values(tokenized_table, num_columns, num_rows, num_tokens)) + + query_ids = self.convert_tokens_to_ids(query_tokens) + table_ids = list(zip(*table_data))[0] if len(table_data) > 0 else list(zip(*table_data)) + table_ids = self.convert_tokens_to_ids(list(table_ids)) + + if "return_overflowing_tokens" in kwargs and kwargs["return_overflowing_tokens"]: + raise ValueError("TAPAS does not return overflowing tokens as it works on tables.") + + if add_special_tokens: + input_ids = self.build_inputs_with_special_tokens(query_ids, table_ids) + else: + input_ids = query_ids + table_ids + + if max_length is not None and len(input_ids) > max_length: + raise ValueError( + "Could not encode the query and table header given the maximum length. Encoding the query and table " + f"header results in a length of {len(input_ids)} which is higher than the max_length of {max_length}" + ) + + encoded_inputs["input_ids"] = input_ids + + segment_ids = self.create_segment_token_type_ids_from_sequences(query_ids, table_data) + column_ids = self.create_column_token_type_ids_from_sequences(query_ids, table_data) + row_ids = self.create_row_token_type_ids_from_sequences(query_ids, table_data) + if not is_part_of_batch or (prev_answer_coordinates is None and prev_answer_text is None): + # simply set the prev_labels to zeros + prev_labels = [0] * len(row_ids) + else: + prev_labels = self.get_answer_ids( + column_ids, row_ids, table_data, prev_answer_text, prev_answer_coordinates + ) + + # FIRST: parse both the table and question in terms of numeric values + + raw_table = add_numeric_table_values(raw_table) + raw_query = add_numeric_values_to_question(raw_query) + + # SECOND: add numeric-related features (and not parse them in these functions): + + column_ranks, inv_column_ranks = self._get_numeric_column_ranks(column_ids, row_ids, raw_table) + numeric_relations = self._get_numeric_relations(raw_query, column_ids, row_ids, raw_table) + + # Load from model defaults + if return_token_type_ids is None: + return_token_type_ids = "token_type_ids" in self.model_input_names + if return_attention_mask is None: + return_attention_mask = "attention_mask" in self.model_input_names + + if return_attention_mask: + attention_mask = self.create_attention_mask_from_sequences(query_ids, table_data) + encoded_inputs["attention_mask"] = attention_mask + + if answer_coordinates is not None and answer_text is not None: + labels = self.get_answer_ids(column_ids, row_ids, table_data, answer_text, answer_coordinates) + numeric_values = self._get_numeric_values(raw_table, column_ids, row_ids) + numeric_values_scale = self._get_numeric_values_scale(raw_table, column_ids, row_ids) + + encoded_inputs["labels"] = labels + encoded_inputs["numeric_values"] = numeric_values + encoded_inputs["numeric_values_scale"] = numeric_values_scale + + if return_token_type_ids: + token_type_ids = [ + segment_ids, + column_ids, + row_ids, + prev_labels, + column_ranks, + inv_column_ranks, + numeric_relations, + ] + + token_type_ids = [list(ids) for ids in list(zip(*token_type_ids))] + encoded_inputs["token_type_ids"] = token_type_ids + + if return_special_tokens_mask: + if add_special_tokens: + encoded_inputs["special_tokens_mask"] = self.get_special_tokens_mask(query_ids, table_ids) + else: + encoded_inputs["special_tokens_mask"] = [0] * len(input_ids) + + # Check lengths + if max_length is None and len(encoded_inputs["input_ids"]) > self.model_max_length and verbose: + if not self.deprecation_warnings.get("sequence-length-is-longer-than-the-specified-maximum", False): + logger.warning( + "Token indices sequence length is longer than the specified maximum sequence length " + f"for this model ({len(encoded_inputs['input_ids'])} > {self.model_max_length}). Running this " + "sequence through the model will result in indexing errors." + ) + self.deprecation_warnings["sequence-length-is-longer-than-the-specified-maximum"] = True + + # Padding + if padding != PaddingStrategy.DO_NOT_PAD or return_attention_mask: + encoded_inputs = self.pad( + encoded_inputs, + max_length=max_length, + padding=padding.value, + pad_to_multiple_of=pad_to_multiple_of, + return_attention_mask=return_attention_mask, + ) + + if return_length: + encoded_inputs["length"] = len(encoded_inputs["input_ids"]) + + batch_outputs = BatchEncoding( + encoded_inputs, tensor_type=return_tensors, prepend_batch_axis=prepend_batch_axis + ) + + return batch_outputs + + def _get_truncated_table_rows( + self, + query_tokens: List[str], + tokenized_table: TokenizedTable, + num_rows: int, + num_columns: int, + max_length: int, + truncation_strategy: Union[str, TapasTruncationStrategy], + ) -> Tuple[int, int]: + """ + Truncates a sequence pair in-place following the strategy. + + Args: + query_tokens (`List[str]`): + List of strings corresponding to the tokenized query. + tokenized_table (`TokenizedTable`): + Tokenized table + num_rows (`int`): + Total number of table rows + num_columns (`int`): + Total number of table columns + max_length (`int`): + Total maximum length. + truncation_strategy (`str` or [`TapasTruncationStrategy`]): + Truncation strategy to use. Seeing as this method should only be called when truncating, the only + available strategy is the `"drop_rows_to_fit"` strategy. + + Returns: + `Tuple(int, int)`: tuple containing the number of rows after truncation, and the number of tokens available + for each table element. + """ + if not isinstance(truncation_strategy, TapasTruncationStrategy): + truncation_strategy = TapasTruncationStrategy(truncation_strategy) + + if max_length is None: + max_length = self.model_max_length + + if truncation_strategy == TapasTruncationStrategy.DROP_ROWS_TO_FIT: + while True: + num_tokens = self._get_max_num_tokens( + query_tokens, tokenized_table, num_rows=num_rows, num_columns=num_columns, max_length=max_length + ) + + if num_tokens is not None: + # We could fit the table. + break + + # Try to drop a row to fit the table. + num_rows -= 1 + + if num_rows < 1: + break + elif truncation_strategy != TapasTruncationStrategy.DO_NOT_TRUNCATE: + raise ValueError(f"Unknown truncation strategy {truncation_strategy}.") + + return num_rows, num_tokens or 1 + + def _tokenize_table( + self, + table=None, + ): + """ + Tokenizes column headers and cell texts of a table. + + Args: + table (`pd.Dataframe`): + Table. Returns: `TokenizedTable`: TokenizedTable object. + """ + tokenized_rows = [] + tokenized_row = [] + # tokenize column headers + for column in table: + if self.strip_column_names: + tokenized_row.append(self.tokenize("")) + else: + tokenized_row.append(self.tokenize(column)) + tokenized_rows.append(tokenized_row) + + # tokenize cell values + for idx, row in table.iterrows(): + tokenized_row = [] + for cell in row: + tokenized_row.append(self.tokenize(cell)) + tokenized_rows.append(tokenized_row) + + token_coordinates = [] + for row_index, row in enumerate(tokenized_rows): + for column_index, cell in enumerate(row): + for token_index, _ in enumerate(cell): + token_coordinates.append( + TokenCoordinates( + row_index=row_index, + column_index=column_index, + token_index=token_index, + ) + ) + + return TokenizedTable( + rows=tokenized_rows, + selected_tokens=token_coordinates, + ) + + def _question_encoding_cost(self, question_tokens): + # Two extra spots of SEP and CLS. + return len(question_tokens) + 2 + + def _get_token_budget(self, question_tokens, max_length=None): + """ + Computes the number of tokens left for the table after tokenizing a question, taking into account the max + sequence length of the model. + + Args: + question_tokens (`List[String]`): + List of question tokens. Returns: `int`: the number of tokens left for the table, given the model max + length. + """ + return (max_length if max_length is not None else self.model_max_length) - self._question_encoding_cost( + question_tokens + ) + + def _get_table_values(self, table, num_columns, num_rows, num_tokens) -> Generator[TableValue, None, None]: + """Iterates over partial table and returns token, column and row indexes.""" + for tc in table.selected_tokens: + # First row is header row. + if tc.row_index >= num_rows + 1: + continue + if tc.column_index >= num_columns: + continue + cell = table.rows[tc.row_index][tc.column_index] + token = cell[tc.token_index] + word_begin_index = tc.token_index + # Don't add partial words. Find the starting word piece and check if it + # fits in the token budget. + while word_begin_index >= 0 and _is_inner_wordpiece(cell[word_begin_index]): + word_begin_index -= 1 + if word_begin_index >= num_tokens: + continue + yield TableValue(token, tc.column_index + 1, tc.row_index) + + def _get_table_boundaries(self, table): + """Return maximal number of rows, columns and tokens.""" + max_num_tokens = 0 + max_num_columns = 0 + max_num_rows = 0 + for tc in table.selected_tokens: + max_num_columns = max(max_num_columns, tc.column_index + 1) + max_num_rows = max(max_num_rows, tc.row_index + 1) + max_num_tokens = max(max_num_tokens, tc.token_index + 1) + max_num_columns = min(self.max_column_id, max_num_columns) + max_num_rows = min(self.max_row_id, max_num_rows) + return max_num_rows, max_num_columns, max_num_tokens + + def _get_table_cost(self, table, num_columns, num_rows, num_tokens): + return sum(1 for _ in self._get_table_values(table, num_columns, num_rows, num_tokens)) + + def _get_max_num_tokens(self, question_tokens, tokenized_table, num_columns, num_rows, max_length): + """Computes max number of tokens that can be squeezed into the budget.""" + token_budget = self._get_token_budget(question_tokens, max_length) + _, _, max_num_tokens = self._get_table_boundaries(tokenized_table) + if self.cell_trim_length >= 0 and max_num_tokens > self.cell_trim_length: + max_num_tokens = self.cell_trim_length + num_tokens = 0 + for num_tokens in range(max_num_tokens + 1): + cost = self._get_table_cost(tokenized_table, num_columns, num_rows, num_tokens + 1) + if cost > token_budget: + break + if num_tokens < max_num_tokens: + if self.cell_trim_length >= 0: + # We don't allow dynamic trimming if a cell_trim_length is set. + return None + if num_tokens == 0: + return None + return num_tokens + + def _get_num_columns(self, table): + num_columns = table.shape[1] + if num_columns >= self.max_column_id: + raise ValueError("Too many columns") + return num_columns + + def _get_num_rows(self, table, drop_rows_to_fit): + num_rows = table.shape[0] + if num_rows >= self.max_row_id: + if drop_rows_to_fit: + num_rows = self.max_row_id - 1 + else: + raise ValueError("Too many rows") + return num_rows + + def _serialize_text(self, question_tokens): + """Serializes texts in index arrays.""" + tokens = [] + segment_ids = [] + column_ids = [] + row_ids = [] + + # add [CLS] token at the beginning + tokens.append(self.cls_token) + segment_ids.append(0) + column_ids.append(0) + row_ids.append(0) + + for token in question_tokens: + tokens.append(token) + segment_ids.append(0) + column_ids.append(0) + row_ids.append(0) + + return tokens, segment_ids, column_ids, row_ids + + def _serialize( + self, + question_tokens, + table, + num_columns, + num_rows, + num_tokens, + ): + """Serializes table and text.""" + tokens, segment_ids, column_ids, row_ids = self._serialize_text(question_tokens) + + # add [SEP] token between question and table tokens + tokens.append(self.sep_token) + segment_ids.append(0) + column_ids.append(0) + row_ids.append(0) + + for token, column_id, row_id in self._get_table_values(table, num_columns, num_rows, num_tokens): + tokens.append(token) + segment_ids.append(1) + column_ids.append(column_id) + row_ids.append(row_id) + + return SerializedExample( + tokens=tokens, + segment_ids=segment_ids, + column_ids=column_ids, + row_ids=row_ids, + ) + + def _get_column_values(self, table, col_index): + table_numeric_values = {} + for row_index, row in table.iterrows(): + cell = row[col_index] + if cell.numeric_value is not None: + table_numeric_values[row_index] = cell.numeric_value + return table_numeric_values + + def _get_cell_token_indexes(self, column_ids, row_ids, column_id, row_id): + for index in range(len(column_ids)): + if column_ids[index] - 1 == column_id and row_ids[index] - 1 == row_id: + yield index + + def _get_numeric_column_ranks(self, column_ids, row_ids, table): + """Returns column ranks for all numeric columns.""" + + ranks = [0] * len(column_ids) + inv_ranks = [0] * len(column_ids) + + # original code from tf_example_utils.py of the original implementation + if table is not None: + for col_index in range(len(table.columns)): + table_numeric_values = self._get_column_values(table, col_index) + + if not table_numeric_values: + continue + + try: + key_fn = get_numeric_sort_key_fn(table_numeric_values.values()) + except ValueError: + continue + + table_numeric_values = {row_index: key_fn(value) for row_index, value in table_numeric_values.items()} + + table_numeric_values_inv = collections.defaultdict(list) + for row_index, value in table_numeric_values.items(): + table_numeric_values_inv[value].append(row_index) + + unique_values = sorted(table_numeric_values_inv.keys()) + + for rank, value in enumerate(unique_values): + for row_index in table_numeric_values_inv[value]: + for index in self._get_cell_token_indexes(column_ids, row_ids, col_index, row_index): + ranks[index] = rank + 1 + inv_ranks[index] = len(unique_values) - rank + + return ranks, inv_ranks + + def _get_numeric_sort_key_fn(self, table_numeric_values, value): + """ + Returns the sort key function for comparing value to table values. The function returned will be a suitable + input for the key param of the sort(). See number_annotation_utils._get_numeric_sort_key_fn for details + + Args: + table_numeric_values: Numeric values of a column + value: Numeric value in the question + + Returns: + A function key function to compare column and question values. + """ + if not table_numeric_values: + return None + all_values = list(table_numeric_values.values()) + all_values.append(value) + try: + return get_numeric_sort_key_fn(all_values) + except ValueError: + return None + + def _get_numeric_relations(self, question, column_ids, row_ids, table): + """ + Returns numeric relations embeddings + + Args: + question: Question object. + column_ids: Maps word piece position to column id. + row_ids: Maps word piece position to row id. + table: The table containing the numeric cell values. + """ + + numeric_relations = [0] * len(column_ids) + + # first, we add any numeric value spans to the question: + # Create a dictionary that maps a table cell to the set of all relations + # this cell has with any value in the question. + cell_indices_to_relations = collections.defaultdict(set) + if question is not None and table is not None: + for numeric_value_span in question.numeric_spans: + for value in numeric_value_span.values: + for column_index in range(len(table.columns)): + table_numeric_values = self._get_column_values(table, column_index) + sort_key_fn = self._get_numeric_sort_key_fn(table_numeric_values, value) + if sort_key_fn is None: + continue + for row_index, cell_value in table_numeric_values.items(): + relation = get_numeric_relation(value, cell_value, sort_key_fn) + if relation is not None: + cell_indices_to_relations[column_index, row_index].add(relation) + + # For each cell add a special feature for all its word pieces. + for (column_index, row_index), relations in cell_indices_to_relations.items(): + relation_set_index = 0 + for relation in relations: + assert relation.value >= Relation.EQ.value + relation_set_index += 2 ** (relation.value - Relation.EQ.value) + for cell_token_index in self._get_cell_token_indexes(column_ids, row_ids, column_index, row_index): + numeric_relations[cell_token_index] = relation_set_index + + return numeric_relations + + def _get_numeric_values(self, table, column_ids, row_ids): + """Returns numeric values for computation of answer loss.""" + + numeric_values = [float("nan")] * len(column_ids) + + if table is not None: + num_rows = table.shape[0] + num_columns = table.shape[1] + + for col_index in range(num_columns): + for row_index in range(num_rows): + numeric_value = table.iloc[row_index, col_index].numeric_value + if numeric_value is not None: + if numeric_value.float_value is None: + continue + float_value = numeric_value.float_value + if float_value == float("inf"): + continue + for index in self._get_cell_token_indexes(column_ids, row_ids, col_index, row_index): + numeric_values[index] = float_value + + return numeric_values + + def _get_numeric_values_scale(self, table, column_ids, row_ids): + """Returns a scale to each token to down weigh the value of long words.""" + + numeric_values_scale = [1.0] * len(column_ids) + + if table is None: + return numeric_values_scale + + num_rows = table.shape[0] + num_columns = table.shape[1] + + for col_index in range(num_columns): + for row_index in range(num_rows): + indices = list(self._get_cell_token_indexes(column_ids, row_ids, col_index, row_index)) + num_indices = len(indices) + if num_indices > 1: + for index in indices: + numeric_values_scale[index] = float(num_indices) + + return numeric_values_scale + + def _pad_to_seq_length(self, inputs): + while len(inputs) > self.model_max_length: + inputs.pop() + while len(inputs) < self.model_max_length: + inputs.append(0) + + def _get_all_answer_ids_from_coordinates( + self, + column_ids, + row_ids, + answers_list, + ): + """Maps lists of answer coordinates to token indexes.""" + answer_ids = [0] * len(column_ids) + found_answers = set() + all_answers = set() + for answers in answers_list: + column_index, row_index = answers + all_answers.add((column_index, row_index)) + for index in self._get_cell_token_indexes(column_ids, row_ids, column_index, row_index): + found_answers.add((column_index, row_index)) + answer_ids[index] = 1 + + missing_count = len(all_answers) - len(found_answers) + return answer_ids, missing_count + + def _get_all_answer_ids(self, column_ids, row_ids, answer_coordinates): + """ + Maps answer coordinates of a question to token indexes. + + In the SQA format (TSV), the coordinates are given as (row, column) tuples. Here, we first swap them to + (column, row) format before calling _get_all_answer_ids_from_coordinates. + """ + + def _to_coordinates(answer_coordinates_question): + return [(coords[1], coords[0]) for coords in answer_coordinates_question] + + return self._get_all_answer_ids_from_coordinates( + column_ids, row_ids, answers_list=(_to_coordinates(answer_coordinates)) + ) + + def _find_tokens(self, text, segment): + """Return start index of segment in text or None.""" + logging.info(f"text: {text} {segment}") + for index in range(1 + len(text) - len(segment)): + for seg_index, seg_token in enumerate(segment): + if text[index + seg_index].piece != seg_token.piece: + break + else: + return index + return None + + def _find_answer_coordinates_from_answer_text( + self, + tokenized_table, + answer_text, + ): + """Returns all occurrences of answer_text in the table.""" + logging.info(f"answer text: {answer_text}") + for row_index, row in enumerate(tokenized_table.rows): + if row_index == 0: + # We don't search for answers in the header. + continue + for col_index, cell in enumerate(row): + token_index = self._find_tokens(cell, answer_text) + if token_index is not None: + yield TokenCoordinates( + row_index=row_index, + column_index=col_index, + token_index=token_index, + ) + + def _find_answer_ids_from_answer_texts( + self, + column_ids, + row_ids, + tokenized_table, + answer_texts, + ): + """Maps question with answer texts to the first matching token indexes.""" + answer_ids = [0] * len(column_ids) + for answer_text in answer_texts: + for coordinates in self._find_answer_coordinates_from_answer_text( + tokenized_table, + answer_text, + ): + # Maps answer coordinates to indexes this can fail if tokens / rows have + # been pruned. + indexes = list( + self._get_cell_token_indexes( + column_ids, + row_ids, + column_id=coordinates.column_index, + row_id=coordinates.row_index - 1, + ) + ) + indexes.sort() + coordinate_answer_ids = [] + if indexes: + begin_index = coordinates.token_index + indexes[0] + end_index = begin_index + len(answer_text) + for index in indexes: + if index >= begin_index and index < end_index: + coordinate_answer_ids.append(index) + if len(coordinate_answer_ids) == len(answer_text): + for index in coordinate_answer_ids: + answer_ids[index] = 1 + break + return answer_ids + + def _get_answer_ids(self, column_ids, row_ids, answer_coordinates): + """Maps answer coordinates of a question to token indexes.""" + answer_ids, missing_count = self._get_all_answer_ids(column_ids, row_ids, answer_coordinates) + + if missing_count: + raise ValueError("Couldn't find all answers") + return answer_ids + + def get_answer_ids(self, column_ids, row_ids, tokenized_table, answer_texts_question, answer_coordinates_question): + if self.update_answer_coordinates: + return self._find_answer_ids_from_answer_texts( + column_ids, + row_ids, + tokenized_table, + answer_texts=[self.tokenize(at) for at in answer_texts_question], + ) + return self._get_answer_ids(column_ids, row_ids, answer_coordinates_question) + + def _pad( + self, + encoded_inputs: Union[Dict[str, EncodedInput], BatchEncoding], + max_length: Optional[int] = None, + padding_strategy: PaddingStrategy = PaddingStrategy.DO_NOT_PAD, + pad_to_multiple_of: Optional[int] = None, + return_attention_mask: Optional[bool] = None, + ) -> dict: + """ + Pad encoded inputs (on left/right and up to predefined length or max length in the batch) + + Args: + encoded_inputs: + Dictionary of tokenized inputs (`List[int]`) or batch of tokenized inputs (`List[List[int]]`). + max_length: maximum length of the returned list and optionally padding length (see below). + Will truncate by taking into account the special tokens. + padding_strategy: PaddingStrategy to use for padding. + + - PaddingStrategy.LONGEST Pad to the longest sequence in the batch + - PaddingStrategy.MAX_LENGTH: Pad to the max length (default) + - PaddingStrategy.DO_NOT_PAD: Do not pad + The tokenizer padding sides are defined in self.padding_side: + + - 'left': pads on the left of the sequences + - 'right': pads on the right of the sequences + pad_to_multiple_of: (optional) Integer if set will pad the sequence to a multiple of the provided value. + This is especially useful to enable the use of Tensor Core on NVIDIA hardware with compute capability + `>= 7.5` (Volta). + return_attention_mask: + (optional) Set to False to avoid returning attention mask (default: set to model specifics) + """ + # Load from model defaults + if return_attention_mask is None: + return_attention_mask = "attention_mask" in self.model_input_names + + if padding_strategy == PaddingStrategy.LONGEST: + max_length = len(encoded_inputs["input_ids"]) + + if max_length is not None and pad_to_multiple_of is not None and (max_length % pad_to_multiple_of != 0): + max_length = ((max_length // pad_to_multiple_of) + 1) * pad_to_multiple_of + + needs_to_be_padded = ( + padding_strategy != PaddingStrategy.DO_NOT_PAD and len(encoded_inputs["input_ids"]) != max_length + ) + + # Initialize attention mask if not present. + if return_attention_mask and "attention_mask" not in encoded_inputs: + encoded_inputs["attention_mask"] = [1] * len(encoded_inputs["input_ids"]) + + if needs_to_be_padded: + difference = max_length - len(encoded_inputs["input_ids"]) + if self.padding_side == "right": + if return_attention_mask: + encoded_inputs["attention_mask"] = encoded_inputs["attention_mask"] + [0] * difference + if "token_type_ids" in encoded_inputs: + encoded_inputs["token_type_ids"] = ( + encoded_inputs["token_type_ids"] + [[self.pad_token_type_id] * 7] * difference + ) + if "labels" in encoded_inputs: + encoded_inputs["labels"] = encoded_inputs["labels"] + [0] * difference + if "numeric_values" in encoded_inputs: + encoded_inputs["numeric_values"] = encoded_inputs["numeric_values"] + [float("nan")] * difference + if "numeric_values_scale" in encoded_inputs: + encoded_inputs["numeric_values_scale"] = ( + encoded_inputs["numeric_values_scale"] + [1.0] * difference + ) + if "special_tokens_mask" in encoded_inputs: + encoded_inputs["special_tokens_mask"] = encoded_inputs["special_tokens_mask"] + [1] * difference + encoded_inputs["input_ids"] = encoded_inputs["input_ids"] + [self.pad_token_id] * difference + elif self.padding_side == "left": + if return_attention_mask: + encoded_inputs["attention_mask"] = [0] * difference + encoded_inputs["attention_mask"] + if "token_type_ids" in encoded_inputs: + encoded_inputs["token_type_ids"] = [[self.pad_token_type_id] * 7] * difference + encoded_inputs[ + "token_type_ids" + ] + if "labels" in encoded_inputs: + encoded_inputs["labels"] = [0] * difference + encoded_inputs["labels"] + if "numeric_values" in encoded_inputs: + encoded_inputs["numeric_values"] = [float("nan")] * difference + encoded_inputs["numeric_values"] + if "numeric_values_scale" in encoded_inputs: + encoded_inputs["numeric_values_scale"] = [1.0] * difference + encoded_inputs[ + "numeric_values_scale" + ] + if "special_tokens_mask" in encoded_inputs: + encoded_inputs["special_tokens_mask"] = [1] * difference + encoded_inputs["special_tokens_mask"] + encoded_inputs["input_ids"] = [self.pad_token_id] * difference + encoded_inputs["input_ids"] + else: + raise ValueError("Invalid padding strategy:" + str(self.padding_side)) + + return encoded_inputs + + # Everything related to converting logits to predictions + + def _get_cell_token_probs(self, probabilities, segment_ids, row_ids, column_ids): + for i, p in enumerate(probabilities): + segment_id = segment_ids[i] + col = column_ids[i] - 1 + row = row_ids[i] - 1 + if col >= 0 and row >= 0 and segment_id == 1: + yield i, p + + def _get_mean_cell_probs(self, probabilities, segment_ids, row_ids, column_ids): + """Computes average probability per cell, aggregating over tokens.""" + coords_to_probs = collections.defaultdict(list) + for i, prob in self._get_cell_token_probs(probabilities, segment_ids, row_ids, column_ids): + col = column_ids[i] - 1 + row = row_ids[i] - 1 + coords_to_probs[(col, row)].append(prob) + return {coords: np.array(cell_probs).mean() for coords, cell_probs in coords_to_probs.items()} + + def convert_logits_to_predictions(self, data, logits, logits_agg=None, cell_classification_threshold=0.5): + """ + Converts logits of [`TapasForQuestionAnswering`] to actual predicted answer coordinates and optional + aggregation indices. + + The original implementation, on which this function is based, can be found + [here](https://github.com/google-research/tapas/blob/4908213eb4df7aa988573350278b44c4dbe3f71b/tapas/experiments/prediction_utils.py#L288). + + Args: + data (`dict`): + Dictionary mapping features to actual values. Should be created using [`TapasTokenizer`]. + logits (`torch.Tensor` or `tf.Tensor` of shape `(batch_size, sequence_length)`): + Tensor containing the logits at the token level. + logits_agg (`torch.Tensor` or `tf.Tensor` of shape `(batch_size, num_aggregation_labels)`, *optional*): + Tensor containing the aggregation logits. + cell_classification_threshold (`float`, *optional*, defaults to 0.5): + Threshold to be used for cell selection. All table cells for which their probability is larger than + this threshold will be selected. + + Returns: + `tuple` comprising various elements depending on the inputs: + + - predicted_answer_coordinates (`List[List[[tuple]]` of length `batch_size`): Predicted answer coordinates + as a list of lists of tuples. Each element in the list contains the predicted answer coordinates of a + single example in the batch, as a list of tuples. Each tuple is a cell, i.e. (row index, column index). + - predicted_aggregation_indices (`List[int]`of length `batch_size`, *optional*, returned when + `logits_aggregation` is provided): Predicted aggregation operator indices of the aggregation head. + """ + # converting to numpy arrays to work with PT/TF + logits = logits.numpy() + if logits_agg is not None: + logits_agg = logits_agg.numpy() + data = {key: value.numpy() for key, value in data.items() if key != "training"} + # input data is of type float32 + # np.log(np.finfo(np.float32).max) = 88.72284 + # Any value over 88.72284 will overflow when passed through the exponential, sending a warning + # We disable this warning by truncating the logits. + logits[logits < -88.7] = -88.7 + + # Compute probabilities from token logits + probabilities = 1 / (1 + np.exp(-logits)) * data["attention_mask"] + token_types = [ + "segment_ids", + "column_ids", + "row_ids", + "prev_labels", + "column_ranks", + "inv_column_ranks", + "numeric_relations", + ] + + # collect input_ids, segment ids, row ids and column ids of batch. Shape (batch_size, seq_len) + input_ids = data["input_ids"] + segment_ids = data["token_type_ids"][:, :, token_types.index("segment_ids")] + row_ids = data["token_type_ids"][:, :, token_types.index("row_ids")] + column_ids = data["token_type_ids"][:, :, token_types.index("column_ids")] + + # next, get answer coordinates for every example in the batch + num_batch = input_ids.shape[0] + predicted_answer_coordinates = [] + for i in range(num_batch): + probabilities_example = probabilities[i].tolist() + segment_ids_example = segment_ids[i] + row_ids_example = row_ids[i] + column_ids_example = column_ids[i] + + max_width = column_ids_example.max() + max_height = row_ids_example.max() + + if max_width == 0 and max_height == 0: + continue + + cell_coords_to_prob = self._get_mean_cell_probs( + probabilities_example, + segment_ids_example.tolist(), + row_ids_example.tolist(), + column_ids_example.tolist(), + ) + + # Select the answers above the classification threshold. + answer_coordinates = [] + for col in range(max_width): + for row in range(max_height): + cell_prob = cell_coords_to_prob.get((col, row), None) + if cell_prob is not None: + if cell_prob > cell_classification_threshold: + answer_coordinates.append((row, col)) + answer_coordinates = sorted(answer_coordinates) + predicted_answer_coordinates.append(answer_coordinates) + + output = (predicted_answer_coordinates,) + + if logits_agg is not None: + predicted_aggregation_indices = logits_agg.argmax(axis=-1) + output = (predicted_answer_coordinates, predicted_aggregation_indices.tolist()) + + return output + + # End of everything related to converting logits to predictions + + +# Copied from transformers.models.bert.tokenization_bert.BasicTokenizer +class BasicTokenizer(object): + """ + Constructs a BasicTokenizer that will run basic tokenization (punctuation splitting, lower casing, etc.). + + Args: + do_lower_case (`bool`, *optional*, defaults to `True`): + Whether or not to lowercase the input when tokenizing. + never_split (`Iterable`, *optional*): + Collection of tokens which will never be split during tokenization. Only has an effect when + `do_basic_tokenize=True` + tokenize_chinese_chars (`bool`, *optional*, defaults to `True`): + Whether or not to tokenize Chinese characters. + + This should likely be deactivated for Japanese (see this + [issue](https://github.com/huggingface/transformers/issues/328)). + strip_accents (`bool`, *optional*): + Whether or not to strip all accents. If this option is not specified, then it will be determined by the + value for `lowercase` (as in the original BERT). + do_split_on_punc (`bool`, *optional*, defaults to `True`): + In some instances we want to skip the basic punctuation splitting so that later tokenization can capture + the full context of the words, such as contractions. + """ + + def __init__( + self, + do_lower_case=True, + never_split=None, + tokenize_chinese_chars=True, + strip_accents=None, + do_split_on_punc=True, + ): + if never_split is None: + never_split = [] + self.do_lower_case = do_lower_case + self.never_split = set(never_split) + self.tokenize_chinese_chars = tokenize_chinese_chars + self.strip_accents = strip_accents + self.do_split_on_punc = do_split_on_punc + + def tokenize(self, text, never_split=None): + """ + Basic Tokenization of a piece of text. For sub-word tokenization, see WordPieceTokenizer. + + Args: + never_split (`List[str]`, *optional*) + Kept for backward compatibility purposes. Now implemented directly at the base class level (see + [`PreTrainedTokenizer.tokenize`]) List of token not to split. + """ + # union() returns a new set by concatenating the two sets. + never_split = self.never_split.union(set(never_split)) if never_split else self.never_split + text = self._clean_text(text) + + # This was added on November 1st, 2018 for the multilingual and Chinese + # models. This is also applied to the English models now, but it doesn't + # matter since the English models were not trained on any Chinese data + # and generally don't have any Chinese data in them (there are Chinese + # characters in the vocabulary because Wikipedia does have some Chinese + # words in the English Wikipedia.). + if self.tokenize_chinese_chars: + text = self._tokenize_chinese_chars(text) + # prevents treating the same character with different unicode codepoints as different characters + unicode_normalized_text = unicodedata.normalize("NFC", text) + orig_tokens = whitespace_tokenize(unicode_normalized_text) + split_tokens = [] + for token in orig_tokens: + if token not in never_split: + if self.do_lower_case: + token = token.lower() + if self.strip_accents is not False: + token = self._run_strip_accents(token) + elif self.strip_accents: + token = self._run_strip_accents(token) + split_tokens.extend(self._run_split_on_punc(token, never_split)) + + output_tokens = whitespace_tokenize(" ".join(split_tokens)) + return output_tokens + + def _run_strip_accents(self, text): + """Strips accents from a piece of text.""" + text = unicodedata.normalize("NFD", text) + output = [] + for char in text: + cat = unicodedata.category(char) + if cat == "Mn": + continue + output.append(char) + return "".join(output) + + def _run_split_on_punc(self, text, never_split=None): + """Splits punctuation on a piece of text.""" + if not self.do_split_on_punc or (never_split is not None and text in never_split): + return [text] + chars = list(text) + i = 0 + start_new_word = True + output = [] + while i < len(chars): + char = chars[i] + if _is_punctuation(char): + output.append([char]) + start_new_word = True + else: + if start_new_word: + output.append([]) + start_new_word = False + output[-1].append(char) + i += 1 + + return ["".join(x) for x in output] + + def _tokenize_chinese_chars(self, text): + """Adds whitespace around any CJK character.""" + output = [] + for char in text: + cp = ord(char) + if self._is_chinese_char(cp): + output.append(" ") + output.append(char) + output.append(" ") + else: + output.append(char) + return "".join(output) + + def _is_chinese_char(self, cp): + """Checks whether CP is the codepoint of a CJK character.""" + # This defines a "chinese character" as anything in the CJK Unicode block: + # https://en.wikipedia.org/wiki/CJK_Unified_Ideographs_(Unicode_block) + # + # Note that the CJK Unicode block is NOT all Japanese and Korean characters, + # despite its name. The modern Korean Hangul alphabet is a different block, + # as is Japanese Hiragana and Katakana. Those alphabets are used to write + # space-separated words, so they are not treated specially and handled + # like the all of the other languages. + if ( + (cp >= 0x4E00 and cp <= 0x9FFF) + or (cp >= 0x3400 and cp <= 0x4DBF) # + or (cp >= 0x20000 and cp <= 0x2A6DF) # + or (cp >= 0x2A700 and cp <= 0x2B73F) # + or (cp >= 0x2B740 and cp <= 0x2B81F) # + or (cp >= 0x2B820 and cp <= 0x2CEAF) # + or (cp >= 0xF900 and cp <= 0xFAFF) + or (cp >= 0x2F800 and cp <= 0x2FA1F) # + ): # + return True + + return False + + def _clean_text(self, text): + """Performs invalid character removal and whitespace cleanup on text.""" + output = [] + for char in text: + cp = ord(char) + if cp == 0 or cp == 0xFFFD or _is_control(char): + continue + if _is_whitespace(char): + output.append(" ") + else: + output.append(char) + return "".join(output) + + +# Copied from transformers.models.bert.tokenization_bert.WordpieceTokenizer +class WordpieceTokenizer(object): + """Runs WordPiece tokenization.""" + + def __init__(self, vocab, unk_token, max_input_chars_per_word=100): + self.vocab = vocab + self.unk_token = unk_token + self.max_input_chars_per_word = max_input_chars_per_word + + def tokenize(self, text): + """ + Tokenizes a piece of text into its word pieces. This uses a greedy longest-match-first algorithm to perform + tokenization using the given vocabulary. + + For example, `input = "unaffable"` wil return as output `["un", "##aff", "##able"]`. + + Args: + text: A single token or whitespace separated tokens. This should have + already been passed through *BasicTokenizer*. + + Returns: + A list of wordpiece tokens. + """ + + output_tokens = [] + for token in whitespace_tokenize(text): + chars = list(token) + if len(chars) > self.max_input_chars_per_word: + output_tokens.append(self.unk_token) + continue + + is_bad = False + start = 0 + sub_tokens = [] + while start < len(chars): + end = len(chars) + cur_substr = None + while start < end: + substr = "".join(chars[start:end]) + if start > 0: + substr = "##" + substr + if substr in self.vocab: + cur_substr = substr + break + end -= 1 + if cur_substr is None: + is_bad = True + break + sub_tokens.append(cur_substr) + start = end + + if is_bad: + output_tokens.append(self.unk_token) + else: + output_tokens.extend(sub_tokens) + return output_tokens + + +# Below: utilities for TAPAS tokenizer (independent from PyTorch/Tensorflow). +# This includes functions to parse numeric values (dates and numbers) from both the table and questions in order +# to create the column_ranks, inv_column_ranks, numeric_values, numeric values_scale and numeric_relations in +# prepare_for_model of TapasTokenizer. +# These are meant to be used in an academic setup, for production use cases Gold mine or Aqua should be used. + + +# taken from constants.py of the original implementation +# URL: https://github.com/google-research/tapas/blob/master/tapas/utils/constants.py +class Relation(enum.Enum): + HEADER_TO_CELL = 1 # Connects header to cell. + CELL_TO_HEADER = 2 # Connects cell to header. + QUERY_TO_HEADER = 3 # Connects query to headers. + QUERY_TO_CELL = 4 # Connects query to cells. + ROW_TO_CELL = 5 # Connects row to cells. + CELL_TO_ROW = 6 # Connects cells to row. + EQ = 7 # Annotation value is same as cell value + LT = 8 # Annotation value is less than cell value + GT = 9 # Annotation value is greater than cell value + + +@dataclass +class Date: + year: Optional[int] = None + month: Optional[int] = None + day: Optional[int] = None + + +@dataclass +class NumericValue: + float_value: Optional[float] = None + date: Optional[Date] = None + + +@dataclass +class NumericValueSpan: + begin_index: int = None + end_index: int = None + values: List[NumericValue] = None + + +@dataclass +class Cell: + text: str + numeric_value: Optional[NumericValue] = None + + +@dataclass +class Question: + original_text: str # The original raw question string. + text: str # The question string after normalization. + numeric_spans: Optional[List[NumericValueSpan]] = None + + +# Below: all functions from number_utils.py as well as 2 functions (namely get_all_spans and normalize_for_match) +# from text_utils.py of the original implementation. URL's: +# - https://github.com/google-research/tapas/blob/master/tapas/utils/number_utils.py +# - https://github.com/google-research/tapas/blob/master/tapas/utils/text_utils.py + + +# Constants for parsing date expressions. +# Masks that specify (by a bool) which of (year, month, day) will be populated. +_DateMask = collections.namedtuple("_DateMask", ["year", "month", "day"]) + +_YEAR = _DateMask(True, False, False) +_YEAR_MONTH = _DateMask(True, True, False) +_YEAR_MONTH_DAY = _DateMask(True, True, True) +_MONTH = _DateMask(False, True, False) +_MONTH_DAY = _DateMask(False, True, True) + +# Pairs of patterns to pass to 'datetime.strptime' and masks specifying which +# fields will be set by the corresponding pattern. +_DATE_PATTERNS = ( + ("%B", _MONTH), + ("%Y", _YEAR), + ("%Ys", _YEAR), + ("%b %Y", _YEAR_MONTH), + ("%B %Y", _YEAR_MONTH), + ("%B %d", _MONTH_DAY), + ("%b %d", _MONTH_DAY), + ("%d %b", _MONTH_DAY), + ("%d %B", _MONTH_DAY), + ("%B %d, %Y", _YEAR_MONTH_DAY), + ("%d %B %Y", _YEAR_MONTH_DAY), + ("%m-%d-%Y", _YEAR_MONTH_DAY), + ("%Y-%m-%d", _YEAR_MONTH_DAY), + ("%Y-%m", _YEAR_MONTH), + ("%B %Y", _YEAR_MONTH), + ("%d %b %Y", _YEAR_MONTH_DAY), + ("%Y-%m-%d", _YEAR_MONTH_DAY), + ("%b %d, %Y", _YEAR_MONTH_DAY), + ("%d.%m.%Y", _YEAR_MONTH_DAY), + ("%A, %b %d", _MONTH_DAY), + ("%A, %B %d", _MONTH_DAY), +) + +# This mapping is used to convert date patterns to regex patterns. +_FIELD_TO_REGEX = ( + ("%A", r"\w+"), # Weekday as locale’s full name. + ("%B", r"\w+"), # Month as locale’s full name. + ("%Y", r"\d{4}"), # Year with century as a decimal number. + ("%b", r"\w{3}"), # Month as locale’s abbreviated name. + ("%d", r"\d{1,2}"), # Day of the month as a zero-padded decimal number. + ("%m", r"\d{1,2}"), # Month as a zero-padded decimal number. +) + + +def _process_date_pattern(dp): + """Compute a regex for each date pattern to use as a prefilter.""" + pattern, mask = dp + regex = pattern + regex = regex.replace(".", re.escape(".")) + regex = regex.replace("-", re.escape("-")) + regex = regex.replace(" ", r"\s+") + for field, field_regex in _FIELD_TO_REGEX: + regex = regex.replace(field, field_regex) + # Make sure we didn't miss any of the fields. + assert "%" not in regex, regex + return pattern, mask, re.compile("^" + regex + "$") + + +def _process_date_patterns(): + return tuple(_process_date_pattern(dp) for dp in _DATE_PATTERNS) + + +_PROCESSED_DATE_PATTERNS = _process_date_patterns() + +_MAX_DATE_NGRAM_SIZE = 5 + +# Following DynSp: +# https://github.com/Microsoft/DynSP/blob/master/util.py#L414. +_NUMBER_WORDS = [ + "zero", + "one", + "two", + "three", + "four", + "five", + "six", + "seven", + "eight", + "nine", + "ten", + "eleven", + "twelve", +] + +_ORDINAL_WORDS = [ + "zeroth", + "first", + "second", + "third", + "fourth", + "fith", + "sixth", + "seventh", + "eighth", + "ninth", + "tenth", + "eleventh", + "twelfth", +] + +_ORDINAL_SUFFIXES = ["st", "nd", "rd", "th"] + +_NUMBER_PATTERN = re.compile(r"((^|\s)[+-])?((\.\d+)|(\d+(,\d\d\d)*(\.\d*)?))") + +# Following DynSp: +# https://github.com/Microsoft/DynSP/blob/master/util.py#L293. +_MIN_YEAR = 1700 +_MAX_YEAR = 2016 + +_INF = float("INF") + + +def _get_numeric_value_from_date(date, mask): + """Converts date (datetime Python object) to a NumericValue object with a Date object value.""" + if date.year < _MIN_YEAR or date.year > _MAX_YEAR: + raise ValueError(f"Invalid year: {date.year}") + + new_date = Date() + if mask.year: + new_date.year = date.year + if mask.month: + new_date.month = date.month + if mask.day: + new_date.day = date.day + return NumericValue(date=new_date) + + +def _get_span_length_key(span): + """Sorts span by decreasing length first and increasing first index second.""" + return span[1] - span[0], -span[0] + + +def _get_numeric_value_from_float(value): + """Converts float (Python) to a NumericValue object with a float value.""" + return NumericValue(float_value=value) + + +# Doesn't parse ordinal expressions such as '18th of february 1655'. +def _parse_date(text): + """Attempts to format a text as a standard date string (yyyy-mm-dd).""" + text = re.sub(r"Sept\b", "Sep", text) + for in_pattern, mask, regex in _PROCESSED_DATE_PATTERNS: + if not regex.match(text): + continue + try: + date = datetime.datetime.strptime(text, in_pattern).date() + except ValueError: + continue + try: + return _get_numeric_value_from_date(date, mask) + except ValueError: + continue + return None + + +def _parse_number(text): + """Parses simple cardinal and ordinals numbers.""" + for suffix in _ORDINAL_SUFFIXES: + if text.endswith(suffix): + text = text[: -len(suffix)] + break + text = text.replace(",", "") + try: + value = float(text) + except ValueError: + return None + if math.isnan(value): + return None + if value == _INF: + return None + return value + + +def get_all_spans(text, max_ngram_length): + """ + Split a text into all possible ngrams up to 'max_ngram_length'. Split points are white space and punctuation. + + Args: + text: Text to split. + max_ngram_length: maximal ngram length. + Yields: + Spans, tuples of begin-end index. + """ + start_indexes = [] + for index, char in enumerate(text): + if not char.isalnum(): + continue + if index == 0 or not text[index - 1].isalnum(): + start_indexes.append(index) + if index + 1 == len(text) or not text[index + 1].isalnum(): + for start_index in start_indexes[-max_ngram_length:]: + yield start_index, index + 1 + + +def normalize_for_match(text): + return " ".join(text.lower().split()) + + +def format_text(text): + """Lowercases and strips punctuation.""" + text = text.lower().strip() + if text == "n/a" or text == "?" or text == "nan": + text = EMPTY_TEXT + + text = re.sub(r"[^\w\d]+", " ", text).replace("_", " ") + text = " ".join(text.split()) + text = text.strip() + if text: + return text + return EMPTY_TEXT + + +def parse_text(text): + """ + Extracts longest number and date spans. + + Args: + text: text to annotate + + Returns: + List of longest numeric value spans. + """ + span_dict = collections.defaultdict(list) + for match in _NUMBER_PATTERN.finditer(text): + span_text = text[match.start() : match.end()] + number = _parse_number(span_text) + if number is not None: + span_dict[match.span()].append(_get_numeric_value_from_float(number)) + + for begin_index, end_index in get_all_spans(text, max_ngram_length=1): + if (begin_index, end_index) in span_dict: + continue + span_text = text[begin_index:end_index] + + number = _parse_number(span_text) + if number is not None: + span_dict[begin_index, end_index].append(_get_numeric_value_from_float(number)) + for number, word in enumerate(_NUMBER_WORDS): + if span_text == word: + span_dict[begin_index, end_index].append(_get_numeric_value_from_float(float(number))) + break + for number, word in enumerate(_ORDINAL_WORDS): + if span_text == word: + span_dict[begin_index, end_index].append(_get_numeric_value_from_float(float(number))) + break + + for begin_index, end_index in get_all_spans(text, max_ngram_length=_MAX_DATE_NGRAM_SIZE): + span_text = text[begin_index:end_index] + date = _parse_date(span_text) + if date is not None: + span_dict[begin_index, end_index].append(date) + + spans = sorted(span_dict.items(), key=lambda span_value: _get_span_length_key(span_value[0]), reverse=True) + selected_spans = [] + for span, value in spans: + for selected_span, _ in selected_spans: + if selected_span[0] <= span[0] and span[1] <= selected_span[1]: + break + else: + selected_spans.append((span, value)) + + selected_spans.sort(key=lambda span_value: span_value[0][0]) + + numeric_value_spans = [] + for span, values in selected_spans: + numeric_value_spans.append(NumericValueSpan(begin_index=span[0], end_index=span[1], values=values)) + return numeric_value_spans + + +# Below: all functions from number_annotation_utils.py and 2 functions (namely filter_invalid_unicode +# and filter_invalid_unicode_from_table) from text_utils.py of the original implementation. URL's: +# - https://github.com/google-research/tapas/blob/master/tapas/utils/number_annotation_utils.py +# - https://github.com/google-research/tapas/blob/master/tapas/utils/text_utils.py + + +_PrimitiveNumericValue = Union[float, Tuple[Optional[float], Optional[float], Optional[float]]] +_SortKeyFn = Callable[[NumericValue], Tuple[float, Ellipsis]] + +_DATE_TUPLE_SIZE = 3 + +EMPTY_TEXT = "EMPTY" + +NUMBER_TYPE = "number" +DATE_TYPE = "date" + + +def _get_value_type(numeric_value): + if numeric_value.float_value is not None: + return NUMBER_TYPE + elif numeric_value.date is not None: + return DATE_TYPE + raise ValueError(f"Unknown type: {numeric_value}") + + +def _get_value_as_primitive_value(numeric_value): + """Maps a NumericValue proto to a float or tuple of float.""" + if numeric_value.float_value is not None: + return numeric_value.float_value + if numeric_value.date is not None: + date = numeric_value.date + value_tuple = [None, None, None] + # All dates fields are cased to float to produce a simple primitive value. + if date.year is not None: + value_tuple[0] = float(date.year) + if date.month is not None: + value_tuple[1] = float(date.month) + if date.day is not None: + value_tuple[2] = float(date.day) + return tuple(value_tuple) + raise ValueError(f"Unknown type: {numeric_value}") + + +def _get_all_types(numeric_values): + return {_get_value_type(value) for value in numeric_values} + + +def get_numeric_sort_key_fn(numeric_values): + """ + Creates a function that can be used as a sort key or to compare the values. Maps to primitive types and finds the + biggest common subset. Consider the values "05/05/2010" and "August 2007". With the corresponding primitive values + (2010.,5.,5.) and (2007.,8., None). These values can be compared by year and date so we map to the sequence (2010., + 5.), (2007., 8.). If we added a third value "2006" with primitive value (2006., None, None), we could only compare + by the year so we would map to (2010.,), (2007.,) and (2006.,). + + Args: + numeric_values: Values to compare + + Returns: + A function that can be used as a sort key function (mapping numeric values to a comparable tuple) + + Raises: + ValueError if values don't have a common type or are not comparable. + """ + value_types = _get_all_types(numeric_values) + if len(value_types) != 1: + raise ValueError(f"No common value type in {numeric_values}") + + value_type = next(iter(value_types)) + if value_type == NUMBER_TYPE: + # Primitive values are simple floats, nothing to do here. + return _get_value_as_primitive_value + + # The type can only be Date at this point which means the primitive type + # is a float triple. + valid_indexes = set(range(_DATE_TUPLE_SIZE)) + + for numeric_value in numeric_values: + value = _get_value_as_primitive_value(numeric_value) + assert isinstance(value, tuple) + for tuple_index, inner_value in enumerate(value): + if inner_value is None: + valid_indexes.discard(tuple_index) + + if not valid_indexes: + raise ValueError(f"No common value in {numeric_values}") + + def _sort_key_fn(numeric_value): + value = _get_value_as_primitive_value(numeric_value) + return tuple(value[index] for index in valid_indexes) + + return _sort_key_fn + + +def _consolidate_numeric_values(row_index_to_values, min_consolidation_fraction, debug_info): + """ + Finds the most common numeric values in a column and returns them + + Args: + row_index_to_values: + For each row index all the values in that cell. + min_consolidation_fraction: + Fraction of cells that need to have consolidated value. + debug_info: + Additional information only used for logging + + Returns: + For each row index the first value that matches the most common value. Rows that don't have a matching value + are dropped. Empty list if values can't be consolidated. + """ + type_counts = collections.Counter() + for numeric_values in row_index_to_values.values(): + type_counts.update(_get_all_types(numeric_values)) + if not type_counts: + return {} + max_count = max(type_counts.values()) + if max_count < len(row_index_to_values) * min_consolidation_fraction: + # logging.log_every_n(logging.INFO, f'Can\'t consolidate types: {debug_info} {row_index_to_values} {max_count}', 100) + return {} + + valid_types = set() + for value_type, count in type_counts.items(): + if count == max_count: + valid_types.add(value_type) + if len(valid_types) > 1: + assert DATE_TYPE in valid_types + max_type = DATE_TYPE + else: + max_type = next(iter(valid_types)) + + new_row_index_to_value = {} + for index, values in row_index_to_values.items(): + # Extract the first matching value. + for value in values: + if _get_value_type(value) == max_type: + new_row_index_to_value[index] = value + break + + return new_row_index_to_value + + +def _get_numeric_values(text): + """Parses text and returns numeric values.""" + numeric_spans = parse_text(text) + return itertools.chain(*(span.values for span in numeric_spans)) + + +def _get_column_values(table, col_index): + """ + Parses text in column and returns a dict mapping row_index to values. This is the _get_column_values function from + number_annotation_utils.py of the original implementation + + Args: + table: Pandas dataframe + col_index: integer, indicating the index of the column to get the numeric values of + """ + index_to_values = {} + for row_index, row in table.iterrows(): + text = normalize_for_match(row[col_index].text) + index_to_values[row_index] = list(_get_numeric_values(text)) + return index_to_values + + +def get_numeric_relation(value, other_value, sort_key_fn): + """Compares two values and returns their relation or None.""" + value = sort_key_fn(value) + other_value = sort_key_fn(other_value) + if value == other_value: + return Relation.EQ + if value < other_value: + return Relation.LT + if value > other_value: + return Relation.GT + return None + + +def add_numeric_values_to_question(question): + """Adds numeric value spans to a question.""" + original_text = question + question = normalize_for_match(question) + numeric_spans = parse_text(question) + return Question(original_text=original_text, text=question, numeric_spans=numeric_spans) + + +def filter_invalid_unicode(text): + """Return an empty string and True if 'text' is in invalid unicode.""" + return ("", True) if isinstance(text, bytes) else (text, False) + + +def filter_invalid_unicode_from_table(table): + """ + Removes invalid unicode from table. Checks whether a table cell text contains an invalid unicode encoding. If yes, + reset the table cell text to an empty str and log a warning for each invalid cell + + Args: + table: table to clean. + """ + # to do: add table id support + if not hasattr(table, "table_id"): + table.table_id = 0 + + for row_index, row in table.iterrows(): + for col_index, cell in enumerate(row): + cell, is_invalid = filter_invalid_unicode(cell) + if is_invalid: + logging.warning( + f"Scrub an invalid table body @ table_id: {table.table_id}, row_index: {row_index}, " + f"col_index: {col_index}", + ) + for col_index, column in enumerate(table.columns): + column, is_invalid = filter_invalid_unicode(column) + if is_invalid: + logging.warning(f"Scrub an invalid table header @ table_id: {table.table_id}, col_index: {col_index}") + + +def add_numeric_table_values(table, min_consolidation_fraction=0.7, debug_info=None): + """ + Parses text in table column-wise and adds the consolidated values. Consolidation refers to finding values with a + common types (date or number) + + Args: + table: + Table to annotate. + min_consolidation_fraction: + Fraction of cells in a column that need to have consolidated value. + debug_info: + Additional information used for logging. + """ + table = table.copy() + # First, filter table on invalid unicode + filter_invalid_unicode_from_table(table) + + # Second, replace cell values by Cell objects + for row_index, row in table.iterrows(): + for col_index, cell in enumerate(row): + table.iloc[row_index, col_index] = Cell(text=cell) + + # Third, add numeric_value attributes to these Cell objects + for col_index, column in enumerate(table.columns): + column_values = _consolidate_numeric_values( + _get_column_values(table, col_index), + min_consolidation_fraction=min_consolidation_fraction, + debug_info=(debug_info, column), + ) + + for row_index, numeric_value in column_values.items(): + table.iloc[row_index, col_index].numeric_value = numeric_value + + return table