code
stringlengths
101
5.91M
def has_valid_annotation(anno, ann_types, filter_crowd=True): if (len(anno) == 0): return False if filter_crowd: if ('iscrowd' in anno[0]): anno = [obj for obj in anno if (obj['iscrowd'] == 0)] if (len(anno) == 0): return False if _has_only_empty_bbox(anno): return False if ('keypoints' in ann_types): keypoints_vis = (_count_visible_keypoints(anno) >= min_keypoints_per_image) else: keypoints_vis = True if keypoints_vis: return True return False
.parametrize('predictions, expected_result', [(([f'SELECT * FROM {TABLE_NAME}'] * 3), []), (([f'SELECT "name" FROM {TABLE_NAME}'] * 3), ['cell_recall_prediction_result', 'tuple_constraint_prediction_result'])]) def test_evaluate_with_df_no_order(metric_evaluator, predictions, expected_result): PREDICTION_DATAFRAME['prediction'] = predictions df = metric_evaluator.evaluate_with_df(PREDICTION_DATAFRAME, 'prediction', task='SP') metrics = metric_evaluator.metrics for metric in metrics: metric = f'{metric}_prediction_result' if (metric == 'tuple_order_prediction_result'): assert (df[metric].tolist() == ([None] * len(PREDICTION_DATAFRAME))) continue if (metric not in expected_result): assert (df[metric].tolist() == ([1.0] * len(PREDICTION_DATAFRAME))) else: assert (df[metric].tolist() != ([1.0] * len(PREDICTION_DATAFRAME)))
class OnlineRUSBoostClassifier(BaseSKMObject, ClassifierMixin, MetaEstimatorMixin): def __init__(self, base_estimator=KNNADWINClassifier(), n_estimators=10, sampling_rate=3, algorithm=1, drift_detection=True, random_state=None): super().__init__() self.base_estimator = base_estimator self.n_estimators = n_estimators self.random_state = random_state self.sampling_rate = sampling_rate self.algorithm = algorithm self.drift_detection = drift_detection self.ensemble = None self.actual_n_estimators = None self.classes = None self._random_state = None self.adwin_ensemble = None self.lam_sc = None self.lam_pos = None self.lam_neg = None self.lam_sw = None self.epsilon = None def __configure(self): if hasattr(self.base_estimator, 'reset'): self.base_estimator.reset() self.actual_n_estimators = self.n_estimators self.adwin_ensemble = [] for i in range(self.actual_n_estimators): self.adwin_ensemble.append(ADWIN()) self.ensemble = [cp.deepcopy(self.base_estimator) for _ in range(self.actual_n_estimators)] self._random_state = check_random_state(self.random_state) self.lam_sc = np.zeros(self.actual_n_estimators) self.lam_pos = np.zeros(self.actual_n_estimators) self.lam_neg = np.zeros(self.actual_n_estimators) self.lam_sw = np.zeros(self.actual_n_estimators) self.epsilon = np.zeros(self.actual_n_estimators) self.n_pos = 0 self.n_neg = 0 def reset(self): self.__configure() def partial_fit(self, X, y, classes=None, sample_weight=None): if (self.ensemble is None): self.__configure() if (self.classes is None): if (classes is None): raise ValueError('The first partial_fit call should pass all the classes.') else: self.classes = classes if ((self.classes is not None) and (classes is not None)): if (set(self.classes) == set(classes)): pass else: raise ValueError('The classes passed to the partial_fit function differ from those passed earlier.') self.__adjust_ensemble_size() (r, _) = get_dimensions(X) for j in range(r): change_detected = False lam = 1 for i in range(self.actual_n_estimators): if (y[j] == 1): self.lam_pos[i] += lam self.n_pos += 1 else: self.lam_neg[i] += lam self.n_neg += 1 lam_rus = 1 if (self.algorithm == 1): if (y[j] == 1): if (self.n_neg != 0): lam_rus = (lam * (((self.lam_pos[i] + self.lam_neg[i]) / (self.lam_pos[i] + (self.lam_neg[i] * (self.sampling_rate * (self.n_pos / self.n_neg))))) * (((self.sampling_rate + 1) * self.n_pos) / (self.n_pos + self.n_neg)))) elif (self.n_pos != 0): lam_rus = (lam * (((self.lam_pos[i] + self.lam_neg[i]) / (self.lam_pos[i] + (self.lam_neg[i] * (self.n_neg / (self.n_pos * self.sampling_rate))))) * (((self.sampling_rate + 1) * self.n_pos) / (self.n_pos + self.n_neg)))) elif (self.algorithm == 2): if (y[j] == 1): lam_rus = (((lam * self.n_pos) / (self.n_pos + self.n_neg)) / (self.lam_pos[i] / (self.lam_pos[i] + self.lam_neg[i]))) else: lam_rus = ((((lam * self.sampling_rate) * self.n_pos) / (self.n_pos + self.n_neg)) / (self.lam_neg[i] / (self.lam_pos[i] + self.lam_neg[i]))) elif (self.algorithm == 3): if (y[j] == 1): lam_rus = lam else: lam_rus = (lam / self.sampling_rate) k = self._random_state.poisson(lam_rus) if (k > 0): for b in range(k): self.ensemble[i].partial_fit([X[j]], [y[j]], classes, sample_weight) if (self.ensemble[i].predict([X[j]])[0] == y[j]): self.lam_sc[i] += lam self.epsilon[i] = (self.lam_sw[i] / (self.lam_sc[i] + self.lam_sw[i])) if (self.epsilon[i] != 1): lam = (lam / (2 * (1 - self.epsilon[i]))) else: self.lam_sw[i] += lam self.epsilon[i] = (self.lam_sw[i] / (self.lam_sc[i] + self.lam_sw[i])) if (self.epsilon[i] != 0): lam = (lam / (2 * self.epsilon[i])) if self.drift_detection: try: pred = self.ensemble[i].predict(X) error_estimation = self.adwin_ensemble[i].estimation for k in range(r): if (pred[k] is not None): self.adwin_ensemble[i].add_element(int((pred[k] == y[k]))) if self.adwin_ensemble[i].detected_change(): if (self.adwin_ensemble[i].estimation > error_estimation): change_detected = True except ValueError: change_detected = False pass if (change_detected and self.drift_detection): max_threshold = 0.0 i_max = (- 1) for i in range(self.actual_n_estimators): if (max_threshold < self.adwin_ensemble[i].estimation): max_threshold = self.adwin_ensemble[i].estimation i_max = i if (i_max != (- 1)): self.ensemble[i_max].reset() self.adwin_ensemble[i_max] = ADWIN() return self def __adjust_ensemble_size(self): if (len(self.classes) != len(self.ensemble)): if (len(self.classes) > len(self.ensemble)): for i in range(len(self.ensemble), len(self.classes)): self.ensemble.append(cp.deepcopy(self.base_estimator)) self.actual_n_estimators += 1 self.adwin_ensemble.append(ADWIN()) self.lam_sc = np.zeros(self.actual_n_estimators) self.lam_pos = np.zeros(self.actual_n_estimators) self.lam_neg = np.zeros(self.actual_n_estimators) self.lam_sw = np.zeros(self.actual_n_estimators) self.epsilon = np.zeros(self.actual_n_estimators) def predict(self, X): (r, c) = get_dimensions(X) proba = self.predict_proba(X) predictions = [] if (proba is None): return None for i in range(r): predictions.append(np.argmax(proba[i])) return np.asarray(predictions) def predict_proba(self, X): proba = [] (r, c) = get_dimensions(X) if (self.ensemble is None): return np.zeros((r, 1)) with warnings.catch_warnings(): warnings.filterwarnings('error') try: for i in range(self.actual_n_estimators): partial_proba = self.ensemble[i].predict_proba(X) if (len(partial_proba[0]) > (max(self.classes) + 1)): raise ValueError('The number of classes in the base learner is larger than in the ensemble.') if (len(proba) < 1): for n in range(r): proba.append([0.0 for _ in partial_proba[n]]) for n in range(r): for k in range(len(partial_proba[n])): try: proba[n][k] += (np.log(((1 - self.epsilon[i]) / self.epsilon[i])) * partial_proba[n][k]) except IndexError: proba[n].append(partial_proba[n][k]) except RuntimeWarning: continue except ValueError: return np.zeros((r, 1)) except TypeError: return np.zeros((r, 1)) sum_proba = [] for k in range(r): sum_proba.append(np.sum(proba[k])) aux = [] for i in range(len(proba)): if (sum_proba[i] > 0.0): aux.append([(x / sum_proba[i]) for x in proba[i]]) else: aux.append(proba[i]) return np.asarray(aux)
class Test_logg(TestCase): def test_works_wasp10(self): answer = 4.51 result = eq.Logg((0.703 * aq.M_s), (0.775 * aq.R_s)).logg self.assertAlmostEqual(answer, result, 1) (M=floats(0.0001, 10000), R=floats(0.0001, 10000)) def test_can_derive_other_vars_from_one_calculated(self, M, R): assume(((M > 0) and (R > 0))) inf = float('inf') assume(((M < inf) and (R < inf))) R *= aq.R_j M *= aq.M_j logg = eq.Logg(M, R).logg self.assertAlmostEqual(eq.Logg(M, R, None).logg, logg, 3) self.assertAlmostEqual(eq.Logg(M, None, logg).R, R, 3) self.assertAlmostEqual(eq.Logg(None, R, logg).M, M, 3)
class CTRLConfig(PretrainedConfig): model_type = 'ctrl' keys_to_ignore_at_inference = ['past_key_values'] def __init__(self, vocab_size=246534, n_positions=256, n_ctx=256, n_embd=1280, dff=8192, n_layer=48, n_head=16, resid_pdrop=0.1, embd_pdrop=0.1, attn_pdrop=0.1, layer_norm_epsilon=1e-06, initializer_range=0.02, summary_type='cls_index', summary_use_proj=True, summary_activation=None, summary_proj_to_labels=True, summary_first_dropout=0.1, use_cache=True, **kwargs): super().__init__(**kwargs) self.vocab_size = vocab_size self.n_ctx = n_ctx self.n_positions = n_positions self.n_embd = n_embd self.n_layer = n_layer self.n_head = n_head self.dff = dff self.resid_pdrop = resid_pdrop self.embd_pdrop = embd_pdrop self.attn_pdrop = attn_pdrop self.layer_norm_epsilon = layer_norm_epsilon self.initializer_range = initializer_range self.summary_type = summary_type self.summary_use_proj = summary_use_proj self.summary_activation = summary_activation self.summary_first_dropout = summary_first_dropout self.summary_proj_to_labels = summary_proj_to_labels self.use_cache = use_cache def max_position_embeddings(self): return self.n_positions def hidden_size(self): return self.n_embd def num_attention_heads(self): return self.n_head def num_hidden_layers(self): return self.n_layer
class FiniteDimensionalAlgebrasWithBasis(CategoryWithAxiom_over_base_ring): class ParentMethods(): _method def radical_basis(self): F = self.base_ring() if (not F.is_field()): raise NotImplementedError('the base ring must be a field') p = F.characteristic() from sage.matrix.constructor import matrix from sage.modules.free_module_element import vector product_on_basis = self.product_on_basis if (p == 0): keys = list(self.basis().keys()) cache = [{(i, j): c for i in keys for (j, c) in product_on_basis(y, i)} for y in keys] mat = [[sum(((x.get((j, i), 0) * c) for ((i, j), c) in y.items())) for x in cache] for y in cache] mat = matrix(self.base_ring(), mat) rad_basis = mat.kernel().basis() else: if hasattr(self.base_ring().one(), 'nth_root'): root_fcn = (lambda s, x: x.nth_root(s)) else: root_fcn = (lambda s, x: (x ** (1 / s))) (s, n) = (1, self.dimension()) B = [b.on_left_matrix() for b in self.basis()] I = B[0].parent().one() while (s <= n): BB = (B + [I]) G = matrix([[(((- 1) ** s) * (b * bb).characteristic_polynomial()[(n - s)]) for bb in BB] for b in B]) C = G.left_kernel().basis() if (1 < s < F.order()): C = [vector(F, [root_fcn(s, ci) for ci in c]) for c in C] B = [sum(((ci * b) for (ci, b) in zip(c, B))) for c in C] s = (p * s) e = vector(self.one()) rad_basis = [(b * e) for b in B] return tuple([self.from_vector(vec) for vec in rad_basis]) _method def radical(self): category = AssociativeAlgebras(self.base_ring()).WithBasis().FiniteDimensional().Subobjects() radical = self.submodule(self.radical_basis(), category=category, already_echelonized=True) radical.rename('Radical of {}'.format(self)) return radical _method def semisimple_quotient(self): ring = self.base_ring() category = Algebras(ring).WithBasis().FiniteDimensional().Quotients().Semisimple() result = self.quotient_module(self.radical(), category=category) result.rename('Semisimple quotient of {}'.format(self)) return result _method def center_basis(self): return self.annihilator_basis(self.algebra_generators(), self.bracket) _method def center(self): category = Algebras(self.base_ring()).FiniteDimensional().Subobjects().Commutative().WithBasis() if (self in Algebras.Semisimple): category = category.Semisimple() center = self.submodule(self.center_basis(), category=category, already_echelonized=True) center.rename('Center of {}'.format(self)) return center def principal_ideal(self, a, side='left'): return self.submodule([((a * b) if (side == 'right') else (b * a)) for b in self.basis()]) _method def orthogonal_idempotents_central_mod_radical(self): one = self.one() idempotents = [] f = self.zero() for g in self.semisimple_quotient().central_orthogonal_idempotents(): fi = self.idempotent_lift((((one - f) * g.lift()) * (one - f))) idempotents.append(fi) f = (f + fi) return tuple(idempotents) def idempotent_lift(self, x): if (not self.is_parent_of(x)): x = x.lift() p = self.semisimple_quotient().retract(x) if ((p * p) != p): raise ValueError(('%s does not retract to an idempotent.' % p)) x_prev = None one = self.one() while (x != x_prev): tmp = x x = (one - ((one - (x ** 2)) ** 2)) x_prev = tmp return x _method def cartan_invariants_matrix(self): from sage.matrix.constructor import Matrix from sage.rings.integer_ring import ZZ A_quo = self.semisimple_quotient() idempotents_quo = A_quo.central_orthogonal_idempotents() dim_simples = [A_quo.principal_ideal(e).dimension().sqrt() for e in idempotents_quo] idempotents = self.orthogonal_idempotents_central_mod_radical() def C(i, j): summand = self.peirce_summand(idempotents[i], idempotents[j]) return (summand.dimension() / (dim_simples[i] * dim_simples[j])) m = Matrix(ZZ, len(idempotents), C) m.set_immutable() return m def isotypic_projective_modules(self, side='left'): return [self.principal_ideal(e, side) for e in self.orthogonal_idempotents_central_mod_radical()] _method def peirce_summand(self, ei, ej): B = self.basis() phi = self.module_morphism(on_basis=(lambda k: ((ei * B[k]) * ej)), codomain=self, triangular='lower') ideal = phi.matrix(side='right').image() return self.submodule([self.from_vector(v) for v in ideal.basis()], already_echelonized=True) def peirce_decomposition(self, idempotents=None, check=True): if (idempotents is None): idempotents = self.orthogonal_idempotents_central_mod_radical() if check: if (not self.is_identity_decomposition_into_orthogonal_idempotents(idempotents)): raise ValueError('Not a decomposition of the identity into orthogonal idempotents') return [[self.peirce_summand(ei, ej) for ej in idempotents] for ei in idempotents] def is_identity_decomposition_into_orthogonal_idempotents(self, l): return ((self.sum(l) == self.one()) and all((((e * e) == e) for e in l)) and all(((((e * f) == 0) and ((f * e) == 0)) for (i, e) in enumerate(l) for f in l[:i]))) _method def is_commutative(self): B = list(self.basis()) try: B.remove(self.one()) except ValueError: pass return all((((b * bp) == (bp * b)) for (i, b) in enumerate(B) for bp in B[(i + 1):])) class ElementMethods(): def to_matrix(self, base_ring=None, action=operator.mul, side='left'): basis = self.parent().basis() action_left = action if (side == 'right'): action = (lambda x: action_left(basis[x], self)) else: action = (lambda x: action_left(self, basis[x])) endo = self.parent().module_morphism(on_basis=action, codomain=self.parent()) return endo.matrix(base_ring=base_ring) _matrix_ = to_matrix on_left_matrix = to_matrix def __invert__(self): alg = self.parent() R = alg.base_ring() ob = None try: ob = alg.one_basis() except (AttributeError, TypeError, ValueError): pass if (ob is not None): mc = self.monomial_coefficients(copy=False) if ((len(mc) == 1) and (ob in mc)): try: return alg.term(ob, R((~ mc[ob]))) except (ValueError, TypeError): raise ValueError(('cannot invert self (= %s)' % self)) e = alg.one().to_vector() A = self.to_matrix() try: inv = A.solve_right(e) inv.change_ring(R) return alg.from_vector(inv) except (ValueError, TypeError): raise ValueError(('cannot invert self (= %s)' % self)) class Cellular(CategoryWithAxiom_over_base_ring): class ParentMethods(): def _test_cellular(self, **options): tester = self._tester(**options) cell_basis = self.cellular_basis() B = cell_basis.basis() P = self.cell_poset() for mu in P: C = self.cell_module_indices(mu) for s in C: t = C[0] vals = [] basis_elt = B[(mu, s, t)] for a in B: elt = (a * basis_elt) tester.assertTrue(all(((P.lt(i[0], mu) or (i[2] == t)) for i in elt.support()))) vals.append([elt[(mu, u, t)] for u in C]) for t in C[1:]: basis_elt = B[(mu, s, t)] for (i, a) in enumerate(B): elt = (a * basis_elt) tester.assertTrue(all(((P.lt(i[0], mu) or (i[2] == t)) for i in elt.support()))) tester.assertEqual(vals[i], [elt[(mu, u, t)] for u in C]) _method def cell_poset(self): _method def cell_module_indices(self, mu): _method(optional=True) def _to_cellular_element(self, i): _method(optional=True) def _from_cellular_index(self, x): def cellular_involution(self, x): C = self.cellular_basis() if (C is self): M = x.monomial_coefficients(copy=False) return self._from_dict({(i[0], i[2], i[1]): M[i] for i in M}, remove_zeros=False) return self(C(x).cellular_involution()) _method def cells(self): from sage.sets.family import Family return Family(self.cell_poset(), self.cell_module_indices) def cellular_basis(self): from sage.algebras.cellular_basis import CellularBasis return CellularBasis(self) def cell_module(self, mu, **kwds): from sage.modules.with_basis.cell_module import CellModule return CellModule(self.cellular_basis(), mu, **kwds) _method def simple_module_parameterization(self): return tuple([mu for mu in self.cell_poset() if self.cell_module(mu).nonzero_bilinear_form()]) class ElementMethods(): def cellular_involution(self): return self.parent().cellular_involution(self) class TensorProducts(TensorProductsCategory): _method def extra_super_categories(self): return [self.base_category()] class ParentMethods(): _method def cell_poset(self): ret = self._sets[0].cell_poset() for A in self._sets[1:]: ret = ret.product(A.cell_poset()) return ret def cell_module_indices(self, mu): from sage.categories.cartesian_product import cartesian_product return cartesian_product([self._sets[i].cell_module_indices(x) for (i, x) in enumerate(mu)]) _attribute def cellular_involution(self): if (self.cellular_basis() is self): def func(x): M = x.monomial_coefficients(copy=False) return self._from_dict({(i[0], i[2], i[1]): M[i] for i in M}, remove_zeros=False) return self.module_morphism(function=func, codomain=self) def on_basis(i): return self._tensor_of_elements([A.basis()[i[j]].cellular_involution() for (j, A) in enumerate(self._sets)]) return self.module_morphism(on_basis, codomain=self) _method def _to_cellular_element(self, i): C = [A.cellular_basis() for A in self._sets] elts = [C[j](self._sets[j].basis()[ij]) for (j, ij) in enumerate(i)] from sage.categories.tensor import tensor T = tensor(C) temp = T._tensor_of_elements(elts) B = self.cellular_basis() M = temp.monomial_coefficients(copy=False) def convert_index(i): mu = [] s = [] t = [] for (a, b, c) in i: mu.append(a) s.append(b) t.append(c) C = self.cell_module_indices(mu) return (tuple(mu), C(s), C(t)) return B._from_dict({convert_index(i): M[i] for i in M}, remove_zeros=False) _method def _from_cellular_index(self, x): elts = [A(A.cellular_basis().basis()[(x[0][i], x[1][i], x[2][i])]) for (i, A) in enumerate(self._sets)] return self._tensor_of_elements(elts) class SubcategoryMethods(): _method def Cellular(self): return self._with_axiom('Cellular')
class Optimizer(object): def __init__(self, weight_decay=0, max_norm=0, lr_scheduler=None, name='Sgd', **kargs): if (name not in S.__dict__): raise NotImplementedError((name + 'is not implemented')) self._solver = S.__dict__[name](**kargs) self._weight_decay = weight_decay self._max_norm = max_norm self._lr_scheduler = lr_scheduler self._iter = 0 if (lr_scheduler is not None): lr = self._lr_scheduler.get_learning_rate(self._iter) self._solver.set_learning_rate(lr) def set_parameters(self, params, **kargs): self._solver.set_parameters(params, **kargs) def update(self): if (self._lr_scheduler is not None): lr = self._lr_scheduler.get_learning_rate(self._iter) self._solver.set_learning_rate(lr) if (self._weight_decay > 0): self._solver.weight_decay(self._weight_decay) if (self._max_norm > 0): self._solver.clip_grad_by_norm(self._max_norm) self._solver.update() self._iter += 1 def zero_grad(self): self._solver.zero_grad() def get_parameters(self): return self._solver.get_parameters() def get_learning_rate(self): return self._solver.learning_rate() def save_states(self, path): self._solver.save_states(path) def load_states(self, path): self._solver.load_states(path) def clear_parameters(self): self._solver.clear_parameters() self._iter = 0
def remove_prefix(text, prefix): if text.startswith(prefix): return text[len(prefix):] return text
def parse_text(text): 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
def _findEDIcomp(comp, fileLines, dt=float): (headLine, indHead) = [(st, nr) for (nr, st) in enumerate(fileLines) if re.search(comp, st)][0] if ('NFREQ' in headLine): breakup = headLine.split('=') breakup2 = breakup[1].split()[0] nrVec = int(breakup2) else: nrVec = int(headLine.split('//')[(- 1)]) c = 0 dataList = [] while (c < nrVec): indHead += 1 dataList.extend(fileLines[indHead].split()) c = len(dataList) return np.array(dataList, dt)
def get_bn_to_prune(model, verbose=True, logger=None, spade=False): weights = [] for (name, m) in model.get_named_block_list().items(): if spade: if isinstance(m, incmod.SPADEInvertedResidualChannels): for (op, (bn_name, bn)) in zip(m.res_ops, m.get_named_first_res_bn(prefix=name).items()): weights.append('{}.weight'.format(bn_name)) for (op, (bn_name, bn)) in zip(m.dw_ops, m.get_named_first_dw_bn(prefix=name).items()): weights.append('{}.weight'.format(bn_name)) for (op, (bn_name, bn)) in zip(m.spade.res_ops, m.spade.get_named_first_res_bn(prefix=(name + '.spade')).items()): weights.append('{}.weight'.format(bn_name)) for (op, (bn_name, bn)) in zip(m.spade.dw_ops, m.spade.get_named_first_dw_bn(prefix=(name + '.spade')).items()): weights.append('{}.weight'.format(bn_name)) elif isinstance(m, incmod.InvertedResidualChannels): for (op, (bn_name, bn)) in zip(m.res_ops, m.get_named_first_res_bn(prefix=name).items()): weights.append('{}.weight'.format(bn_name)) for (op, (bn_name, bn)) in zip(m.dw_ops, m.get_named_first_dw_bn(prefix=name).items()): weights.append('{}.weight'.format(bn_name)) prune_weights = weights if verbose: for name in prune_weights: if (logger is not None): logger.print_info('{}\n'.format(name)) else: print('{}'.format(name)) all_params_keys = [key for (key, val) in model.named_parameters()] for name_weight in prune_weights: assert (name_weight in all_params_keys) return prune_weights
def test_noise(): from glob import glob for filename in glob(os.path.join(output_folder, 'save', '*.flac')): expected_file = filename.replace('results', 'expected') actual = read_audio(filename) expected = read_audio(expected_file) assert actual.allclose(expected)
def method_from_name(method, **kwargs): if (method == 'ParetoMTL'): return ParetoMTLMethod(**kwargs) elif ('cosmos' in method): return COSMOSMethod(**kwargs) elif (method == 'SingleTask'): return SingleTaskMethod(**kwargs) elif ('hyper' in method): return HypernetMethod(**kwargs) elif (method == 'mgda'): return MGDAMethod(**kwargs) elif (method == 'uniform'): return UniformScalingMethod(**kwargs) else: raise ValueError('Unkown method {}'.format(method))
def print_tree(tree): return ('{} -> {}'.format(tree.edge.src, tree.edge.dst) + ''.join(('\n |\n +- {}'.format(print_tree(c)) for c in tree.children)))
def train_collate_build(batch): (imgs, pids, _, _, mask) = zip(*batch) pids = torch.tensor(pids, dtype=torch.int64) return (torch.stack(imgs, dim=0), pids, torch.stack(mask, dim=0))
def pack_examples(tok, src_examples, tgt_examples, max_tokens=1024): (finished_src, finished_tgt) = ([], []) sorted_examples = list(zip(src_examples, tgt_examples)) (new_src, new_tgt) = sorted_examples[0] def is_too_big(strang): return (tok(strang, return_tensors='pt').input_ids.shape[1] > max_tokens) for (src, tgt) in tqdm(sorted_examples[1:]): cand_src = ((new_src + ' ') + src) cand_tgt = ((new_tgt + ' ') + tgt) if (is_too_big(cand_src) or is_too_big(cand_tgt)): finished_src.append(new_src) finished_tgt.append(new_tgt) (new_src, new_tgt) = (src, tgt) else: (new_src, new_tgt) = (cand_src, cand_tgt) if new_src: assert new_tgt finished_src.append(new_src) finished_tgt.append(new_tgt) return (finished_src, finished_tgt)
class Observation(): def __init__(self, id=(- 1), reward=0, state=None, is_episode_over=False): self.id = id self.reward = reward self.state = state self.is_episode_over = is_episode_over def are_equal(self, other, are_states_equal): if (self.id != other.id): return False elif (self.reward != other.reward): return False elif (not are_states_equal(self.state, other.state)): return False elif (self.is_episode_over != other.is_episode_over): return False return True
def get_output_dir(datasets, training=True): assert isinstance(datasets, tuple(([tuple, list] + list(six.string_types)))), 'datasets argument must be of type tuple, list or string' is_string = isinstance(datasets, six.string_types) dataset_name = (datasets if is_string else ':'.join(datasets)) tag = ('train' if training else 'test') outdir = osp.join(__C.OUTPUT_DIR, tag, dataset_name, __C.MODEL.TYPE) if (not osp.exists(outdir)): os.makedirs(outdir) return outdir
class TorchModel(): def __init__(self, model, use_cuda, full_features=False): self.model = model self.use_cuda = use_cuda self.full_features = full_features if full_features: self.lm_embed = model.embedding.embed self.lstm_stack = unstack_lstm(model.embedding.rnn) self.proj = model.embedding.proj if use_cuda: self.lm_embed.cuda() for lstm in self.lstm_stack: lstm.cuda() self.proj.cuda() def __call__(self, x): c = [torch.from_numpy(x_).long() for x_ in x] (c, order) = pack_sequences(c) if self.use_cuda: c = c.cuda() if self.full_features: z = featurize(c, self.lm_embed, self.lstm_stack, self.proj) else: z = self.model(c) z = unpack_sequences(z, order) return z
def stl2off(stl_file, off_file=None, delete_file=True): file_format_check(stl_file, '.stl') if (off_file is None): off_file = stl_file.replace('.stl', '.off') mesh = load_mesh(stl_file) off_data = export_off(mesh) with open(off_file, 'w') as f: f.write(off_data) if delete_file: os.remove(stl_file) return off_file
class TestCopulaGANSynthesizer(): def test___init__(self): metadata = SingleTableMetadata() enforce_min_max_values = True enforce_rounding = True instance = CopulaGANSynthesizer(metadata, enforce_min_max_values=enforce_min_max_values, enforce_rounding=enforce_rounding) assert (instance.enforce_min_max_values is True) assert (instance.enforce_rounding is True) assert (instance.embedding_dim == 128) assert (instance.generator_dim == (256, 256)) assert (instance.discriminator_dim == (256, 256)) assert (instance.generator_lr == 0.0002) assert (instance.generator_decay == 1e-06) assert (instance.discriminator_lr == 0.0002) assert (instance.discriminator_decay == 1e-06) assert (instance.batch_size == 500) assert (instance.discriminator_steps == 1) assert (instance.log_frequency is True) assert (instance.verbose is False) assert (instance.epochs == 300) assert (instance.pac == 10) assert (instance.cuda is True) assert (instance.numerical_distributions == {}) assert (instance.default_distribution == 'beta') assert (instance._numerical_distributions == {}) assert (instance._default_distribution == BetaUnivariate) def test___init__custom(self): metadata = SingleTableMetadata() metadata.add_column('field', sdtype='numerical') enforce_min_max_values = False enforce_rounding = False embedding_dim = 64 generator_dim = (128, 128) discriminator_dim = (128, 128) generator_lr = 0.0001 generator_decay = 2e-06 discriminator_lr = 0.0003 discriminator_decay = 1e-06 batch_size = 250 discriminator_steps = 2 log_frequency = False verbose = True epochs = 150 pac = 5 cuda = False numerical_distributions = {'field': 'gamma'} default_distribution = 'uniform' instance = CopulaGANSynthesizer(metadata, enforce_min_max_values=enforce_min_max_values, enforce_rounding=enforce_rounding, embedding_dim=embedding_dim, generator_dim=generator_dim, discriminator_dim=discriminator_dim, generator_lr=generator_lr, generator_decay=generator_decay, discriminator_lr=discriminator_lr, discriminator_decay=discriminator_decay, batch_size=batch_size, discriminator_steps=discriminator_steps, log_frequency=log_frequency, verbose=verbose, epochs=epochs, pac=pac, cuda=cuda, numerical_distributions=numerical_distributions, default_distribution=default_distribution) assert (instance.enforce_min_max_values is False) assert (instance.enforce_rounding is False) assert (instance.embedding_dim == embedding_dim) assert (instance.generator_dim == generator_dim) assert (instance.discriminator_dim == discriminator_dim) assert (instance.generator_lr == generator_lr) assert (instance.generator_decay == generator_decay) assert (instance.discriminator_lr == discriminator_lr) assert (instance.discriminator_decay == discriminator_decay) assert (instance.batch_size == batch_size) assert (instance.discriminator_steps == discriminator_steps) assert (instance.log_frequency == log_frequency) assert (instance.verbose is True) assert (instance.epochs == epochs) assert (instance.pac == pac) assert (instance.cuda is False) assert (instance.numerical_distributions == {'field': 'gamma'}) assert (instance._numerical_distributions == {'field': GammaUnivariate}) assert (instance.default_distribution == 'uniform') assert (instance._default_distribution == UniformUnivariate) def test___init__incorrect_numerical_distributions(self): metadata = SingleTableMetadata() numerical_distributions = 'invalid' err_msg = 'numerical_distributions can only be None or a dict instance.' with pytest.raises(TypeError, match=err_msg): CopulaGANSynthesizer(metadata, numerical_distributions=numerical_distributions) def test___init__invalid_column_numerical_distributions(self): metadata = SingleTableMetadata() numerical_distributions = {'totally_fake_column_name': 'beta'} err_msg = re.escape("Invalid column names found in the numerical_distributions dictionary {'totally_fake_column_name'}. The column names you provide must be present in the metadata.") with pytest.raises(SynthesizerInputError, match=err_msg): CopulaGANSynthesizer(metadata, numerical_distributions=numerical_distributions) def test_get_params(self): metadata = SingleTableMetadata() instance = CopulaGANSynthesizer(metadata) result = instance.get_parameters() assert (result == {'enforce_min_max_values': True, 'enforce_rounding': True, 'locales': None, 'embedding_dim': 128, 'generator_dim': (256, 256), 'discriminator_dim': (256, 256), 'generator_lr': 0.0002, 'generator_decay': 1e-06, 'discriminator_lr': 0.0002, 'discriminator_decay': 1e-06, 'batch_size': 500, 'discriminator_steps': 1, 'log_frequency': True, 'verbose': False, 'epochs': 300, 'pac': 10, 'cuda': True, 'numerical_distributions': {}, 'default_distribution': 'beta'}) ('sdv.single_table.copulagan.rdt') def test__create_gaussian_normalizer_config(self, mock_rdt): numerical_distributions = {'age': 'gamma'} metadata = SingleTableMetadata() metadata.columns = {'name': {'sdtype': 'categorical'}, 'age': {'sdtype': 'numerical'}, 'account': {'sdtype': 'numerical'}} instance = CopulaGANSynthesizer(metadata, numerical_distributions=numerical_distributions) processed_data = pd.DataFrame({'name': ['John', 'Doe', 'John Doe', 'John Doe Doe'], 'age': np.arange(4), 'account': np.arange(4), 'name#age': np.arange(4)}) config = instance._create_gaussian_normalizer_config(processed_data) expected_calls = [call(missing_value_generation='from_column', distribution=GammaUnivariate), call(missing_value_generation='from_column', distribution=BetaUnivariate)] expected_config = {'transformers': {'name': None, 'age': mock_rdt.transformers.GaussianNormalizer.return_value, 'account': mock_rdt.transformers.GaussianNormalizer.return_value, 'name#age': None}, 'sdtypes': {'name': 'categorical', 'age': 'numerical', 'account': 'numerical', 'name#age': 'categorical'}} assert (config == expected_config) assert (mock_rdt.transformers.GaussianNormalizer.call_args_list == expected_calls) ('sdv.single_table.copulagan.LOGGER') ('sdv.single_table.copulagan.CTGANSynthesizer._fit') ('sdv.single_table.copulagan.rdt') def test__fit_logging(self, mock_rdt, mock_ctgansynthesizer__fit, mock_logger): metadata = SingleTableMetadata() metadata.add_column('col', sdtype='numerical') numerical_distributions = {'col': 'gamma'} instance = CopulaGANSynthesizer(metadata, numerical_distributions=numerical_distributions) processed_data = pd.DataFrame() instance._fit(processed_data) mock_logger.info.assert_called_once_with("Requested distribution 'gamma' cannot be applied to column 'col' because it no longer exists after preprocessing.") ('sdv.single_table.copulagan.CTGANSynthesizer._fit') ('sdv.single_table.copulagan.rdt') def test__fit(self, mock_rdt, mock_ctgansynthesizer__fit): metadata = SingleTableMetadata() instance = CopulaGANSynthesizer(metadata) instance._create_gaussian_normalizer_config = Mock() processed_data = pd.DataFrame() instance._fit(processed_data) hypertransformer = instance._gaussian_normalizer_hyper_transformer assert (hypertransformer == mock_rdt.HyperTransformer.return_value) hypertransformer.set_config.assert_called_once_with(instance._create_gaussian_normalizer_config.return_value) hypertransformer.fit_transform.assert_called_once_with(processed_data) mock_ctgansynthesizer__fit.assert_called_once_with(hypertransformer.fit_transform.return_value) def test_get_learned_distributions(self): data = pd.DataFrame({'zero': [0, 0, 0], 'one': [1, 1, 1]}) stm = SingleTableMetadata() stm.detect_from_dataframe(data) cgs = CopulaGANSynthesizer(stm) zero_transformer_mock = Mock(spec_set=GaussianNormalizer) zero_transformer_mock._univariate.to_dict.return_value = {'a': 1.0, 'b': 1.0, 'loc': 0.0, 'scale': 0.0, 'type': None} one_transformer_mock = Mock(spec_set=GaussianNormalizer) one_transformer_mock._univariate.to_dict.return_value = {'a': 1.0, 'b': 1.0, 'loc': 1.0, 'scale': 0.0, 'type': None} cgs._gaussian_normalizer_hyper_transformer = Mock() cgs._gaussian_normalizer_hyper_transformer.field_transformers = {'zero': zero_transformer_mock, 'one': one_transformer_mock} cgs._fitted = True result = cgs.get_learned_distributions() assert (result == {'zero': {'distribution': 'beta', 'learned_parameters': {'a': 1.0, 'b': 1.0, 'loc': 0.0, 'scale': 0.0}}, 'one': {'distribution': 'beta', 'learned_parameters': {'a': 1.0, 'b': 1.0, 'loc': 1.0, 'scale': 0.0}}}) def test_get_learned_distributions_raises_an_error(self): data = pd.DataFrame({'zero': [0, 0, 0], 'one': [1, 1, 1]}) stm = SingleTableMetadata() stm.detect_from_dataframe(data) cgs = CopulaGANSynthesizer(stm) error_msg = re.escape("Distributions have not been learned yet. Please fit your model first using 'fit'.") with pytest.raises(ValueError, match=error_msg): cgs.get_learned_distributions()
class IteratorTrainer(Trainer): def __init__(self, opt, meta, data_loader, model, optimizer): super(IteratorTrainer, self).__init__(opt, meta, data_loader, model, optimizer) self.iters = meta.cycle self.total_iters = meta.iterations def run(self): self.set_logger() bar = utils.set_progress_bar(self.total_iters) for cycle_num in range(int((self.total_iters / self.iters))): self.model.train() self.cycle(bar, cycle_num) with torch.no_grad(): self.run_evaluation_cycle() self.log_losses(self.opt, self.losses) self.update_top_score(self.opt) self.save_model(self.get_tracked_score()) self.stop_logger() def cycle(self, bar, cycle_num): nums = self.reset_losses() print(self.losses['train']) for i in range(1, (self.iters + 1)): (loss, nums, reset) = self.do_forward_pass(nums) self.do_backward_pass(loss) self.update_parameters() self.opt.train.dynamic.epoch += 1 for loss_name in self.losses['train']: self.logger.add_scalar('train/{}'.format(loss_name), (loss.item() / self.opt.train.dynamic.bs), self.opt.train.dynamic.epoch) bar.update(1) if (cfg.toy and (i > 10)): break if reset: self.data_loader.reset_offsets('train')
def test_L3EthStarBuild(): topo = L3EthStar() net = Mininet(topo=topo, link=TCLink, listenPort=OF_MISC['switch_debug_port']) net.start() CLI(net) net.stop()
class Net(nn.Module): def __init__(self, num_classes=10): super(Net, self).__init__() self.features = nn.Sequential(nn.Conv2d(3, 64, kernel_size=3, stride=2, padding=1), nn.ReLU(inplace=True), nn.MaxPool2d(kernel_size=2), nn.Conv2d(64, 192, kernel_size=3, padding=1), nn.ReLU(inplace=True), nn.MaxPool2d(kernel_size=2), nn.Conv2d(192, 384, kernel_size=3, padding=1), nn.ReLU(inplace=True), nn.Conv2d(384, 256, kernel_size=3, padding=1), nn.ReLU(inplace=True), nn.Conv2d(256, 256, kernel_size=3, padding=1), nn.ReLU(inplace=True), nn.MaxPool2d(kernel_size=2)) self.classifier = nn.Sequential(nn.Dropout(), nn.Linear(((256 * 2) * 2), 4096), nn.ReLU(inplace=True), nn.Dropout(), nn.Linear(4096, 4096), nn.ReLU(inplace=True), nn.Linear(4096, num_classes)) def forward(self, x): x = self.features(x) x = x.view(x.size(0), ((256 * 2) * 2)) x = self.classifier(x) return x
def register_Ns3CallbackImpl__Void_Unsigned_long_Unsigned_short_Long_double_Ns3Empty_Ns3Empty_Ns3Empty_Ns3Empty_Ns3Empty_Ns3Empty_methods(root_module, cls): cls.add_constructor([]) cls.add_constructor([param('ns3::CallbackImpl< void, unsigned long, unsigned short, long double, ns3::empty, ns3::empty, ns3::empty, ns3::empty, ns3::empty, ns3::empty > const &', 'arg0')]) cls.add_method('DoGetTypeid', 'std::string', [], is_static=True) cls.add_method('GetTypeid', 'std::string', [], is_const=True, is_virtual=True) cls.add_method('operator()', 'void', [param('long unsigned int', 'arg0'), param('short unsigned int', 'arg1'), param('long double', 'arg2')], is_pure_virtual=True, is_virtual=True, custom_name=u'__call__') return
_model('linear_transformer_lm') class LinearTransformerLanguageModel(FairseqLanguageModel): def __init__(self, decoder): super().__init__(decoder) def add_args(parser): parser.add_argument('--embed-dim', type=int, metavar='N', help='embedding dimension') parser.add_argument('--num-attention-heads', type=int, metavar='N', help='num attention heads') parser.add_argument('--num-layers', type=int, metavar='N', help='num layers') parser.add_argument('--dropout', type=float, metavar='D', help='dropout probability for all fully connected layers in the embeddings, encoder, and pooler') def build_model(cls, args, task): base_architecture(args) return cls(LinearTransformerDecoder(args, task))
class TransformerDecoderLayer(TransformerDecoderLayerBase): def __init__(self, args, no_encoder_attn=False, add_bias_kv=False, add_zero_attn=False): super().__init__(TransformerConfig.from_namespace(args), no_encoder_attn=no_encoder_attn, add_bias_kv=add_bias_kv, add_zero_attn=add_zero_attn) self.args = args def build_self_attention(self, embed_dim, args, add_bias_kv=False, add_zero_attn=False): return super().build_self_attention(embed_dim, TransformerConfig.from_namespace(args), add_bias_kv=add_bias_kv, add_zero_attn=add_zero_attn) def build_encoder_attention(self, embed_dim, args): return super().build_encoder_attention(embed_dim, TransformerConfig.from_namespace(args))
class RGBArrayAsObservationWrapper(dm_env.Environment): '\n\tUse env.render(rgb_array) as observation\n\trather than the observation environment provides\n\n\tFrom: def __init__(self, env, width=84, height=84): self._env = env self._width = width self._height = height self._env.reset() dummy_obs = self._env.render(mode='rgb_array', width=self._width, height=self._height) self.observation_space = spaces.Box(low=0, high=255, shape=dummy_obs.shape, dtype=dummy_obs.dtype) self.action_space = self._env.action_space wrapped_action_spec = self.action_space if (not hasattr(wrapped_action_spec, 'minimum')): wrapped_action_spec.minimum = (- np.ones(wrapped_action_spec.shape)) if (not hasattr(wrapped_action_spec, 'maximum')): wrapped_action_spec.maximum = np.ones(wrapped_action_spec.shape) self._action_spec = specs.BoundedArray(wrapped_action_spec.shape, np.float32, wrapped_action_spec.minimum, wrapped_action_spec.maximum, 'action') self._obs_spec = {} self._obs_spec['pixels'] = specs.BoundedArray(shape=self.observation_space.shape, dtype=np.uint8, minimum=0, maximum=255, name='observation') def reset(self, **kwargs): obs = {} obs = self._env.reset(**kwargs) obs['pixels'] = obs['pixels'].astype(np.uint8) obs['goal_achieved'] = False return obs def step(self, action): (observation, reward, done, info) = self._env.step(action) obs = {} obs['pixels'] = observation['pixels'].astype(np.uint8) obs['goal_achieved'] = info['is_success'] return (obs, reward, done, info) def observation_spec(self): return self._obs_spec def action_spec(self): return self._action_spec def render(self, mode='rgb_array', width=256, height=256): return self._env.render(mode='rgb_array', width=width, height=height) def __getattr__(self, name): return getattr(self._env, name)
class PegasusConverter(SpmConverter): def vocab(self, proto): vocab = [(self.original_tokenizer.pad_token, 0.0), (self.original_tokenizer.eos_token, 0.0), (self.original_tokenizer.mask_token_sent, 0.0), (self.original_tokenizer.mask_token, 0.0)] vocab += [(f'<unk_{i}>', (- 100.0)) for i in range(2, self.original_tokenizer.offset)] vocab += [(piece.piece, piece.score) for piece in proto.pieces[2:]] return vocab def unk_id(self, proto): return (proto.trainer_spec.unk_id + self.original_tokenizer.offset) def pre_tokenizer(self, replacement, add_prefix_space): return pre_tokenizers.Sequence([pre_tokenizers.WhitespaceSplit(), pre_tokenizers.Metaspace(replacement=replacement, add_prefix_space=add_prefix_space)]) def post_processor(self): eos = self.original_tokenizer.eos_token special_tokens = [(eos, self.original_tokenizer.eos_token_id)] return processors.TemplateProcessing(single=['$A', eos], pair=['$A', '$B', eos], special_tokens=special_tokens)
def plot_data(ax, alg, mean_lc, mean_stderr, best_params, exp_attrs, second_time=False): alpha = 1.0 if PLOT_RERUN_AND_ORIG: alpha = (1.0 if second_time else 0.5) lbl = ((alg + '$\\alpha=$ ') + str(best_params['alpha'])) color = ALG_COLORS[alg] ax.plot(np.arange(mean_lc.shape[0]), mean_lc, label=lbl, linewidth=1.0, color=color, alpha=alpha) ax.fill_between(np.arange(mean_lc.shape[0]), (mean_lc - (mean_stderr / 2)), (mean_lc + (mean_stderr / 2)), color=color, alpha=(0.1 * alpha)) ax.get_xaxis().tick_bottom() ax.get_yaxis().tick_left() ax.spines['top'].set_visible(False) ax.spines['right'].set_visible(False) ax.set_xlim(exp_attrs.x_lim) ax.set_ylim(exp_attrs.y_lim) ax.xaxis.set_ticks(exp_attrs.x_axis_ticks) ax.set_xticklabels(exp_attrs.x_tick_labels, fontsize=25) ax.yaxis.set_ticks(exp_attrs.y_axis_ticks) ax.tick_params(axis='y', which='major', labelsize=exp_attrs.size_of_labels) ax.spines['left'].set_linewidth(2) ax.spines['bottom'].set_linewidth(2)
def p_continue_statement(s): pos = s.position() s.next() return Nodes.ContinueStatNode(pos)
def test_initialize_from_files_lazy_paths(): _db = Database.from_files(pos=pathlib.Path('datasets/ToyFather/train/pos.pl'), neg=pathlib.Path('datasets/ToyFather/train/neg.pl'), facts=pathlib.Path('datasets/ToyFather/train/facts.pl'), lazy_load=True) assert (_db.pos == pathlib.Path('datasets/ToyFather/train/pos.pl')) assert (_db.neg == pathlib.Path('datasets/ToyFather/train/neg.pl')) assert (_db.facts == pathlib.Path('datasets/ToyFather/train/facts.pl'))
def parse_constants_2018toXXXX(d: str) -> dict[(str, tuple[(float, str, float)])]: constants = {} for line in d.split('\n'): name = line[:60].rstrip() val = float(line[60:85].replace(' ', '').replace('...', '')) uncert = float(line[85:110].replace(' ', '').replace('(exact)', '0')) units = line[110:].rstrip() constants[name] = (val, units, uncert) return constants
_model_architecture('transformer', 'transformer_wmt_en_de_big_t2t') def transformer_wmt_en_de_big_t2t(args): args.encoder_normalize_before = getattr(args, 'encoder_normalize_before', True) args.decoder_normalize_before = getattr(args, 'decoder_normalize_before', True) args.attention_dropout = getattr(args, 'attention_dropout', 0.1) args.activation_dropout = getattr(args, 'activation_dropout', 0.1) transformer_vaswani_wmt_en_de_big(args)
class PPM(nn.Module): def __init__(self, channels=2048): super(PPM, self).__init__() bins = (1, 2, 3, 6) reduction_dim = int((channels / len(bins))) self.features = [] for bin in bins: self.features.append(nn.Sequential(nn.AdaptiveAvgPool2d(bin), nn.Conv2d(channels, reduction_dim, kernel_size=1, bias=False), nn.BatchNorm2d(reduction_dim), nn.ReLU(inplace=True))) self.features = nn.ModuleList(self.features) def forward(self, x): x_size = x.size() out = [x] for f in self.features: out.append(F.interpolate(f(x), x_size[2:], mode='bilinear', align_corners=True)) return torch.cat(out, 1)
.node class Transpose(dace.sdfg.nodes.LibraryNode): implementations = {'pure': ExpandTransposePure, 'MKL': ExpandTransposeMKL, 'OpenBLAS': ExpandTransposeOpenBLAS, 'cuBLAS': ExpandTransposeCuBLAS} default_implementation = 'pure' dtype = dace.properties.TypeClassProperty(allow_none=True) def __init__(self, name, dtype=None, location=None): super().__init__(name, location=location, inputs={'_inp'}, outputs={'_out'}) self.dtype = dtype def validate(self, sdfg, state): in_edges = state.in_edges(self) if (len(in_edges) != 1): raise ValueError('Expected exactly one input to transpose operation') for (_, _, _, dst_conn, memlet) in state.in_edges(self): if (dst_conn == '_inp'): subset = dc(memlet.subset) subset.squeeze() in_size = subset.size() out_edges = state.out_edges(self) if (len(out_edges) != 1): raise ValueError('Expected exactly one output from transpose operation') out_memlet = out_edges[0].data if (len(in_size) != 2): raise ValueError('Transpose operation only supported on matrices') out_subset = dc(out_memlet.subset) out_subset.squeeze() out_size = out_subset.size() if (len(out_size) != 2): raise ValueError('Transpose operation only supported on matrices') if (list(out_size) != [in_size[1], in_size[0]]): raise ValueError('Output to transpose operation must agree in the m and n dimensions')
class TestSelu(serial.SerializedTestCase): (X=hu.tensor(), engine=st.sampled_from(['', 'CUDNN']), **hu.gcs) def test_selu_1(self, X, gc, dc, engine): alpha = 1.0 scale = 2.0 op = core.CreateOperator('Selu', ['X'], ['Y'], alpha=alpha, scale=scale, engine=engine) X = TestSelu.fix0(X) self.assertDeviceChecks(dc, op, [X], [0]) self.assertGradientChecks(gc, op, [X], 0, [0]) self.assertReferenceChecks(gc, op, [X], (lambda x: TestSelu.selu_ref(x, alpha=alpha, scale=scale))) (X=hu.tensor(), engine=st.sampled_from(['', 'CUDNN']), **hu.gcs) (deadline=1000) def test_selu_2(self, X, gc, dc, engine): alpha = 1.6732 scale = 1.0507 op = core.CreateOperator('Selu', ['X'], ['Y'], alpha=alpha, scale=scale, engine=engine) X = TestSelu.fix0(X) self.assertDeviceChecks(dc, op, [X], [0]) self.assertGradientChecks(gc, op, [X], 0, [0], stepsize=0.01, threshold=0.01) self.assertReferenceChecks(gc, op, [X], (lambda x: TestSelu.selu_ref(x, alpha=alpha, scale=scale))) (X=hu.tensor(), engine=st.sampled_from(['', 'CUDNN']), **hu.gcs) (deadline=1000) def test_selu_3(self, X, gc, dc, engine): alpha = 1.3 scale = 1.1 op = core.CreateOperator('Selu', ['X'], ['Y'], alpha=alpha, scale=scale, engine=engine) X = TestSelu.fix0(X) self.assertDeviceChecks(dc, op, [X], [0]) self.assertGradientChecks(gc, op, [X], 0, [0]) self.assertReferenceChecks(gc, op, [X], (lambda x: TestSelu.selu_ref(x, alpha=alpha, scale=scale))) (X=hu.tensor(), engine=st.sampled_from(['', 'CUDNN']), **hu.gcs) def test_selu_inplace(self, X, gc, dc, engine): alpha = 1.3 scale = 1.1 op = core.CreateOperator('Selu', ['X'], ['X'], alpha=alpha, scale=scale, engine=engine) X = TestSelu.fix0(X) self.assertDeviceChecks(dc, op, [X], [0]) Y = TestSelu.selu_ref(X, alpha=alpha, scale=scale) dX = np.ones_like(X) op2 = core.CreateOperator('SeluGradient', ['Y', 'dX'], ['dX'], alpha=alpha, scale=scale, engine=engine) self.assertDeviceChecks(dc, op2, [Y, dX], [0]) def fix0(X): X += (0.02 * np.sign(X)) X[(X == 0.0)] += 0.02 return X def selu_ref(x, scale, alpha): ret = (scale * (((x > 0) * x) + ((x <= 0) * (alpha * (np.exp(x) - 1))))) return [ret]
def mk_unix_dist(build_path, dist_path): for c in get_components(): c.mk_unix_dist(build_path, dist_path) for pyc in filter((lambda f: (f.endswith('.pyc') or f.endswith('.py'))), os.listdir(build_path)): shutil.copy(os.path.join(build_path, pyc), os.path.join(dist_path, INSTALL_BIN_DIR, pyc))
def print_on_call_decorator(func): (func) def wrapper(self, *args, **kwargs): _debug(type(self).__name__, func.__name__) try: return func(self, *args, **kwargs) except Exception: _debug('An exception occurred:', traceback.format_exc()) raise return wrapper
.experimental .parametrize('pad_columns, padding_value, array_size', [(['item_id', 'timestamp'], 0, None)]) .parametrize('dataset, result', [pytest.param('dataframe_special', 'dataframe_two_columns_none'), pytest.param('dataframe_special_pandas', 'dataframe_two_columns_none_pandas')]) def test_padder_two_columns_none(pad_columns, padding_value, array_size, dataset, result, request): dataframe_special = request.getfixturevalue(dataset) dataframe_two_columns_none = request.getfixturevalue(result) is_spark = isinstance(dataframe_special, SparkDataFrame) padder = Padder(pad_columns=pad_columns, padding_value=padding_value, array_size=array_size) padder_interactions = padder.transform(dataframe_special) columns = (padder_interactions.collect()[0].asDict().keys() if is_spark else padder_interactions.columns) assert ('user_id' in columns) assert ('item_id' in columns) assert ('timestamp' in columns) if (is_spark is True): assert padder_interactions.toPandas().equals(dataframe_two_columns_none.toPandas()) else: assert padder_interactions.equals(dataframe_two_columns_none)
def make_distributor_init_64_bits(distributor_init, vcomp140_dll_filename, msvcp140_dll_filename): with open(distributor_init, 'wt') as f: f.write(textwrap.dedent('\n \'\'\'Helper to preload vcomp140.dll and msvcp140.dll to prevent\n "not found" errors.\n\n Once vcomp140.dll and msvcp140.dll are\n preloaded, the namespace is made available to any subsequent\n vcomp140.dll and msvcp140.dll. This is\n created as part of the scripts that build the wheel.\n \'\'\'\n\n\n import os\n import os.path as op\n from ctypes import WinDLL\n\n\n if os.name == "nt":\n libs_path = op.join(op.dirname(__file__), ".libs")\n vcomp140_dll_filename = op.join(libs_path, "{0}")\n msvcp140_dll_filename = op.join(libs_path, "{1}")\n WinDLL(op.abspath(vcomp140_dll_filename))\n WinDLL(op.abspath(msvcp140_dll_filename))\n '.format(vcomp140_dll_filename, msvcp140_dll_filename)))
def config_imname(cfg, i): return os.path.join(cfg['dir_images'], (cfg['imlist'][i] + cfg['ext']))
def test_record_int32_parameters(): t = RecordType([NumpyType('int32')], None, {'p': [123]}) assert (str(parser.parse(str(t))) == str(t))
def process_tweets(id_, data_child_array, source_claim, source_tweets, labels_dict): user_id_array = [] tweet_id_array = [] tweet_array = [] time_delay_array = [] missing_count = 0 label = labels_dict[id_] try: source_tweet_id = str(source_claim['tweet_id']).strip() source_claim_tweet = source_tweets[source_tweet_id] source_claim['tweet'] = source_claim_tweet except: return (None, None, None, None, None, None, None) for item in data_child_array: try: user_id = item['user_id'] tweet_id = str(item['tweet_id']).strip() time_delay = float(item['time_delay'].strip()) tweet = source_tweets[tweet_id] user_id_array.append(user_id) tweet_id_array.append(tweet_id) tweet_array.append(tweet) time_delay_array.append(time_delay) except: missing_count += 1 continue assert (len(user_id_array) == len(tweet_id_array)), print('Length of arrays DO NOT match') assert (len(user_id_array) == len(tweet_array)), print('Length of arrays DO NOT match') assert (len(user_id_array) == len(time_delay_array)), print('Length of arrays DO NOT match') return (label, source_claim, user_id_array, tweet_id_array, tweet_array, time_delay_array, missing_count)
def InferOpDeviceAsBlobDevices(op): op_dev = (op.device_option if op.device_option else caffe2_pb2.DeviceOption()) input_dev = ([op_dev] * len(op.input)) output_dev = ([op_dev] * len(op.output)) return (input_dev, output_dev)
class RqWorkerBarLogger(RqWorkerProgressLogger, ProgressBarLogger): def __init__(self, job, init_state=None, bars=None, ignored_bars=(), logged_bars='all', min_time_interval=0): RqWorkerProgressLogger.__init__(self, job) ProgressBarLogger.__init__(self, init_state=init_state, bars=bars, ignored_bars=ignored_bars, logged_bars=logged_bars, min_time_interval=min_time_interval)
def get_video_modes(monitor): count_value = ctypes.c_int(0) count = ctypes.pointer(count_value) result = _glfw.glfwGetVideoModes(monitor, count) videomodes = [result[i].unwrap() for i in range(count_value.value)] return videomodes
class MultiBoxLoss(nn.Module): def __init__(self, num_classes, overlap_thresh, prior_for_matching, bkg_label, neg_mining, neg_pos, neg_overlap, encode_target, use_gpu=True): super(MultiBoxLoss, self).__init__() self.use_gpu = use_gpu self.num_classes = num_classes self.threshold = overlap_thresh self.background_label = bkg_label self.encode_target = encode_target self.use_prior_for_matching = prior_for_matching self.do_neg_mining = neg_mining self.negpos_ratio = neg_pos self.neg_overlap = neg_overlap self.variance = config['frame_work']['variance'] def forward(self, predictions, targets): (loc_datas_p, p_c_p, p_e_p, priors) = predictions (loc_datas_t, p_c_t, p_e_t) = targets num = loc_datas_p.size(0) num_frames = loc_datas_p.shape[1] num_priors = priors.size(0) num_classes = self.num_classes loc_ts = torch.zeros(num, num_frames, num_priors, 4) p_c_ts = torch.zeros(num, 1, num_priors, dtype=torch.long) p_e_ts = torch.zeros(num, num_frames, num_priors) if self.use_gpu: loc_ts = loc_ts.cuda() p_c_ts = p_c_ts.cuda() p_e_ts = p_e_ts.cuda() for idx in range(num): if (len(loc_datas_t[idx]) == 0): continue truths = loc_datas_t[idx].float() labels = p_c_t[idx] exists = p_e_t[idx] defaults = priors.data match(self.threshold, truths, defaults, self.variance, labels, exists, loc_ts, p_c_ts, p_e_ts, idx) with torch.no_grad(): loc_ts = Variable(loc_ts) p_c_ts = Variable(p_c_ts) p_e_ts = Variable(p_e_ts) pos = (p_c_ts > 0) pos_idx = pos.unsqueeze(pos.dim()).expand_as(loc_datas_p) exist_idx = (p_e_ts.unsqueeze(p_e_ts.dim()).expand_as(loc_datas_p) > 0) pos_idx = ((pos_idx * exist_idx) > 0) loc_p = loc_datas_p[pos_idx].view((- 1), 4) loc_ts = loc_ts[pos_idx].view((- 1), 4) loss_l = F.smooth_l1_loss(loc_p, loc_ts, reduction='sum') batch_conf = p_c_p.contiguous().view((- 1), self.num_classes) loss_c = (log_sum_exp(batch_conf) - batch_conf.gather(1, p_c_ts.view((- 1), 1))) loss_c = loss_c.view(num, (- 1)).unsqueeze(1) loss_c[pos] = 0 (_, loss_idx) = loss_c.sort(2, descending=True) (_, idx_rank) = loss_idx.sort(2) num_pos = pos.long().sum(2, keepdim=True) num_neg = torch.clamp((self.negpos_ratio * num_pos), max=(pos.shape[2] - 1)) neg = (idx_rank < num_neg.expand_as(idx_rank)) pos_idx = pos.unsqueeze(3).expand_as(p_c_p) neg_idx = neg.unsqueeze(3).expand_as(p_c_p) conf_p = p_c_p[(pos_idx + neg_idx).gt(0)].view((- 1), self.num_classes) targets_weighted = p_c_ts[(pos + neg).gt(0)] if ((len(targets_weighted) == 0) or (len(conf_p) == 0)): loss_c = conf_p.sum() else: loss_c = F.cross_entropy(conf_p, targets_weighted, reduction='sum') pos_idx = pos.unsqueeze(3).expand_as(p_e_p) neg_idx = neg.unsqueeze(3).expand_as(p_e_p) exist_p = p_e_p[(pos_idx + neg_idx).gt(0)].view((- 1), 2) pos_idx = pos.expand_as(p_e_ts) neg_idx = neg.expand_as(p_e_ts) targets_weighted = p_e_ts[(pos_idx + neg_idx).gt(0)].long() if ((len(targets_weighted) == 0) or (len(exist_p) == 0)): loss_e = exist_p.sum() else: loss_e = F.cross_entropy(exist_p, targets_weighted, reduction='sum') N = num_pos.data.sum().float() return (((loss_l / N) / num_frames), ((2 * loss_c) / N), ((loss_e / N) / num_frames))
class Dataset(InMemoryDataset): def __init__(self, root, dataset, rating_file, sep, args, transform=None, pre_transform=None): self.path = root self.dataset = dataset self.rating_file = rating_file self.sep = sep self.store_backup = True self.args = args super(Dataset, self).__init__(root, transform, pre_transform) (self.data, self.slices) = torch.load(self.processed_paths[0]) self.stat_info = torch.load(self.processed_paths[1]) self.data_num = self.stat_info['data_num'] self.feature_num = self.stat_info['feature_num'] def raw_file_names(self): return ['{}{}/user_dict.pkl'.format(self.path, self.dataset), '{}{}/item_dict.pkl'.format(self.path, self.dataset), '{}{}/feature_dict.pkl'.format(self.path, self.dataset), '{}{}/{}'.format(self.path, self.dataset, self.rating_file)] def processed_file_names(self): return ['{}/{}.dataset'.format(self.dataset, self.dataset), '{}/{}.statinfo'.format(self.dataset, self.dataset)] def download(self): pass def data_2_graphs(self, ratings_df, dataset='train'): graphs = [] processed_graphs = 0 num_graphs = ratings_df.shape[0] one_per = int((num_graphs / 1000)) percent = 0.0 for i in range(len(ratings_df)): if ((processed_graphs % one_per) == 0): print(f'Processing [{dataset}]: {(percent / 10.0)}%, {processed_graphs}/{num_graphs}', end='\r') percent += 1 processed_graphs += 1 line = ratings_df.iloc[i] user_index = self.user_key_type(line[0]) item_index = self.item_key_type(line[1]) rating = int(line[2]) if ((item_index not in self.item_dict) or (user_index not in self.user_dict)): error_num += 1 continue user_id = self.user_dict[user_index]['name'] item_id = self.item_dict[item_index]['title'] user_attr_list = self.user_dict[user_index]['attribute'] item_attr_list = self.item_dict[item_index]['attribute'] user_list = ([user_id] + user_attr_list) item_list = ([item_id] + item_attr_list) graph = self.construct_graphs(user_list, item_list, rating) graphs.append(graph) print() return graphs def read_data(self): self.user_dict = pickle.load(open(self.userfile, 'rb')) self.item_dict = pickle.load(open(self.itemfile, 'rb')) self.user_key_type = type(list(self.user_dict.keys())[0]) self.item_key_type = type(list(self.item_dict.keys())[0]) feature_dict = pickle.load(open(self.featurefile, 'rb')) data = [] error_num = 0 ratings_df = pd.read_csv(self.ratingfile, sep=self.sep, header=None) (train_df, test_df) = train_test_split(ratings_df, test_size=0.4, random_state=self.args.random_seed, stratify=ratings_df[[0, 2]]) (test_df, valid_df) = train_test_split(test_df, test_size=0.5, random_state=self.args.random_seed, stratify=test_df[[0, 2]]) if self.store_backup: backup_path = f'{self.path}{self.dataset}/split_data_backup/' if (not os.path.exists(backup_path)): os.mkdir(backup_path) train_df.to_csv(f'{backup_path}train_data.csv', index=False) valid_df.to_csv(f'{backup_path}valid_data.csv', index=False) test_df.to_csv(f'{backup_path}test_data.csv', index=False) print('(Only run at the first time training the dataset)') train_graphs = self.data_2_graphs(train_df, dataset='train') valid_graphs = self.data_2_graphs(valid_df, dataset='valid') test_graphs = self.data_2_graphs(test_df, dataset='test') graphs = ((train_graphs + valid_graphs) + test_graphs) stat_info = {} stat_info['data_num'] = len(graphs) stat_info['feature_num'] = len(feature_dict) stat_info['train_test_split_index'] = [len(train_graphs), (len(train_graphs) + len(valid_graphs))] print('error number of data:', error_num) return (graphs, stat_info) def construct_graphs(self, user_list, item_list, rating): u_n = len(user_list) i_n = len(item_list) inner_edge_index = [[], []] for i in range(u_n): for j in range(i, u_n): inner_edge_index[0].append(i) inner_edge_index[1].append(j) for i in range(u_n, (u_n + i_n)): for j in range(i, (u_n + i_n)): inner_edge_index[0].append(i) inner_edge_index[1].append(j) outer_edge_index = [[], []] for i in range(u_n): for j in range(i_n): outer_edge_index[0].append(i) outer_edge_index[1].append((u_n + j)) inner_edge_index = torch.LongTensor(inner_edge_index) inner_edge_index = to_undirected(inner_edge_index) outer_edge_index = torch.LongTensor(outer_edge_index) outer_edge_index = to_undirected(outer_edge_index) graph = self.construct_graph((user_list + item_list), inner_edge_index, outer_edge_index, rating) return graph def construct_graph(self, node_list, edge_index_inner, edge_index_outer, rating): x = torch.LongTensor(node_list).unsqueeze(1) rating = torch.FloatTensor([rating]) return Data(x=x, edge_index=edge_index_inner, edge_attr=torch.transpose(edge_index_outer, 0, 1), y=rating) def process(self): self.userfile = self.raw_file_names[0] self.itemfile = self.raw_file_names[1] self.featurefile = self.raw_file_names[2] self.ratingfile = self.raw_file_names[3] (graphs, stat_info) = self.read_data() if (not os.path.exists(f'{self.path}processed/{self.dataset}')): os.mkdir(f'{self.path}processed/{self.dataset}') (data, slices) = self.collate(graphs) torch.save((data, slices), self.processed_paths[0]) torch.save(stat_info, self.processed_paths[1]) def feature_N(self): return self.feature_num def data_N(self): return self.data_num
def cache_temp_view(df: SparkDataFrame, name: str) -> None: spark = State().session df.createOrReplaceTempView(name) spark.sql(f'cache table {name}')
_start_docstrings('CamemBERT Model transformer with a sequence classification/regression head on top (a linear layer\n on top of the pooled output) e.g. for GLUE tasks. ', CAMEMBERT_START_DOCSTRING) class TFCamembertForSequenceClassification(TFRobertaForSequenceClassification): config_class = CamembertConfig
def test_onnx(): model = tract.onnx().model_for_path('./mobilenetv2-7.onnx').into_optimized().into_runnable() result = model.run([grace_hopper_1x3x224x244()]) confidences = result[0].to_numpy() assert (numpy.argmax(confidences) == 652)
class BMType(Enum): KML_SEQ = 0 KML_RAND = 1 VAN_SEQ = 2 VAN_RAND = 3 RA = 4 KML_SEQ_MAJ = 5 KML_RAND_MAJ = 6 VAN_SEQ_MAJ = 7 VAN_RAND_MAJ = 8
def relation_matrix_wtk_g0(syms, sign, field, sparse): rels = modS_relations(syms) if (sign != 0): rels.update(modI_relations(syms, sign)) rels = sorted(rels) if (syms._apply_S_only_0pm1() and is_RationalField(field)): from . import relation_matrix_pyx mod = relation_matrix_pyx.sparse_2term_quotient_only_pm1(rels, len(syms)) else: mod = sparse_2term_quotient(rels, len(syms), field) R = T_relation_matrix_wtk_g0(syms, mod, field, sparse) return (R, mod)
def readEduDoc(fedu, doc): if (not os.path.isfile(fedu)): raise IOError("File doesn't exist: {}".format(fedu)) (gidx, eidx, tokendict, edudict) = (0, 1, {}, {}) with open(fedu, 'r') as fin: for line in fin: line = line.strip() if (len(line) == 0): continue eduTxt = line edudict[eidx] = [] tokens = TOKENIZER.tokenize(line) for tok in tokens: tokendict[gidx] = tok edudict[eidx].append(gidx) gidx += 1 eidx += 1 doc.tokendict = tokendict doc.edudict = edudict return doc
def load_triples_from_txt(filename, words_indexes=None, parse_line=parse_line): if (words_indexes == None): words_indexes = dict() entities = set() next_ent = 0 else: entities = set(words_indexes) next_ent = (max(words_indexes.values()) + 1) data = dict() with open(filename) as f: lines = f.readlines() for (_, line) in enumerate(lines): (sub, obj, rel, val) = parse_line(line) if (sub in entities): sub_ind = words_indexes[sub] else: sub_ind = next_ent next_ent += 1 words_indexes[sub] = sub_ind entities.add(sub) if (rel in entities): rel_ind = words_indexes[rel] else: rel_ind = next_ent next_ent += 1 words_indexes[rel] = rel_ind entities.add(rel) if (obj in entities): obj_ind = words_indexes[obj] else: obj_ind = next_ent next_ent += 1 words_indexes[obj] = obj_ind entities.add(obj) data[(sub_ind, rel_ind, obj_ind)] = val indexes_words = {} for tmpkey in words_indexes: indexes_words[words_indexes[tmpkey]] = tmpkey return (data, words_indexes, indexes_words)
def permute_map(map_entry: nodes.MapEntry, perm: List[int]): map_entry.map.params = [map_entry.map.params[p] for p in perm] map_entry.map.range = [map_entry.map.range[p] for p in perm]
def adaptive_avg_pool2d(input, output_size): if ((input.numel() == 0) and obsolete_torch_version(TORCH_VERSION, (1, 9))): if isinstance(output_size, int): output_size = [output_size, output_size] output_size = [*input.shape[:2], *output_size] empty = NewEmptyTensorOp.apply(input, output_size) return empty else: return F.adaptive_avg_pool2d(input, output_size)
def get_sources(module, surfix='*.c*'): src_dir = osp.join(*module.split('.'), 'src') cuda_dir = osp.join(src_dir, 'cuda') cpu_dir = osp.join(src_dir, 'cpu') return ((glob(osp.join(src_dir, surfix)) + glob(osp.join(cuda_dir, surfix))) + glob(osp.join(cpu_dir, surfix)))
_model def skresnet18(pretrained=False, num_classes=1000, in_chans=3, **kwargs): default_cfg = default_cfgs['skresnet18'] sk_kwargs = dict(min_attn_channels=16, attn_reduction=8, split_input=True) model = ResNet(SelectiveKernelBasic, [2, 2, 2, 2], num_classes=num_classes, in_chans=in_chans, block_args=dict(sk_kwargs=sk_kwargs), zero_init_last_bn=False, **kwargs) model.default_cfg = default_cfg if pretrained: load_pretrained(model, default_cfg, num_classes, in_chans) return model
class BridgeTowerTextConfig(PretrainedConfig): model_type = 'bridgetower_text_model' def __init__(self, vocab_size=50265, hidden_size=768, num_hidden_layers=12, num_attention_heads=12, initializer_factor=1, intermediate_size=3072, hidden_act='gelu', hidden_dropout_prob=0.1, attention_probs_dropout_prob=0.1, max_position_embeddings=514, type_vocab_size=1, initializer_range=0.02, layer_norm_eps=1e-05, pad_token_id=1, bos_token_id=0, eos_token_id=2, position_embedding_type='absolute', use_cache=True, classifier_dropout=None, **kwargs): super().__init__(**kwargs) 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.initializer_factor = initializer_factor 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_size = type_vocab_size self.initializer_range = initializer_range self.layer_norm_eps = layer_norm_eps self.position_embedding_type = position_embedding_type self.use_cache = use_cache self.classifier_dropout = classifier_dropout self.pad_token_id = pad_token_id self.bos_token_id = bos_token_id self.eos_token_id = eos_token_id def from_pretrained(cls, pretrained_model_name_or_path: Union[(str, os.PathLike)], **kwargs) -> 'PretrainedConfig': (config_dict, kwargs) = cls.get_config_dict(pretrained_model_name_or_path, **kwargs) if (config_dict.get('model_type') == 'bridgetower'): config_dict = config_dict['text_config'] if (('model_type' in config_dict) and hasattr(cls, 'model_type') and (config_dict['model_type'] != cls.model_type)): logger.warning(f"You are using a model of type {config_dict['model_type']} to instantiate a model of type {cls.model_type}. This is not supported for all configurations of models and can yield errors.") return cls.from_dict(config_dict, **kwargs)
def plot_results(graph, params, results): plt.figure(figsize=(6.4, 4.8)) title = '{}_epidemic:diffusion={},method={},k={}'.format(params['model'], params['diffusion'], params['method'], params['k']) for (strength, result) in results.items(): result_norm = [(r / len(graph)) for r in result] plt.plot(result_norm, label='Effective strength: {}'.format(strength)) plt.xlabel('Steps') plt.ylabel('Infected Nodes') plt.legend() plt.yscale('log') plt.ylim(0.001, 1) plt.title(title) save_dir = (((os.getcwd() + '/plots/') + title) + '/') os.makedirs(save_dir, exist_ok=True) plt.savefig(((save_dir + title) + '.pdf')) plt.show() plt.clf()
class PlotReconstructionCallback(tfk.callbacks.Callback): def __init__(self, logdir: str, test_ds: tf.data.Dataset, nex: int=4): super(PlotReconstructionCallback, self).__init__() logdir = os.path.join(logdir, 'reconstructions') self.file_writer = tf.summary.create_file_writer(logdir=logdir) self.nex = nex self.test_ds = test_ds.map((lambda x, y: x)).unbatch().batch(nex) self.test_it = iter(self.test_ds) def get_next_images(self): try: next_images = next(self.test_it) except StopIteration: self.test_it = iter(self.test_ds) next_images = next(self.test_it) return next_images def plot_img_reconstruction(self, image, reconstruction): (fig, ax) = plt.subplots(nrows=1, ncols=2) if (image.shape[(- 1)] == 1): image = tf.squeeze(image, axis=(- 1)) reconstruction = tf.squeeze(reconstruction, axis=(- 1)) ax[0].imshow(image, vmin=0.0, vmax=1.0, cmap=plt.cm.Greys) ax[0].set_title('Image') ax[0].axis('off') ax[1].imshow(reconstruction, vmin=0.0, vmax=1.0, cmap=plt.cm.Greys) ax[1].set_title('Reconstruction') ax[1].axis('off') return fig def on_epoch_end(self, epoch, logs=None): images = self.get_next_images() reconstructions = self.model(images) imgs = [] for i in range(self.nex): fig = self.plot_img_reconstruction(images[i], reconstructions[i]) imgs.append(plot_to_image(fig)) imgs = tf.concat(imgs, axis=0) with self.file_writer.as_default(): tf.summary.image(name='Reconstructions', data=imgs, step=epoch, max_outputs=self.nex)
class NanDetector(): def __init__(self, model, forward=True, backward=True): self.bhooks = [] self.fhooks = [] self.forward = forward self.backward = backward self.named_parameters = list(model.named_parameters()) self.reset() for (name, mod) in model.named_modules(): mod.__module_name = name self.add_hooks(mod) def __enter__(self): return self def __exit__(self, exc_type, exc_value, exc_traceback): norm = {} gradients = {} for (name, param) in self.named_parameters: if (param.grad is not None): grad_norm = torch.norm(param.grad.data, p=2, dtype=torch.float32) norm[name] = grad_norm.item() if (torch.isnan(grad_norm).any() or torch.isinf(grad_norm).any()): gradients[name] = param.grad.data if (len(gradients) > 0): logger.info('Detected nan/inf grad norm, dumping norms...') logger.info(f'norms: {norm}') logger.info(f'gradients: {gradients}') self.close() def add_hooks(self, module): if self.forward: self.fhooks.append(module.register_forward_hook(self.fhook_fn)) if self.backward: self.bhooks.append(module.register_backward_hook(self.bhook_fn)) def reset(self): self.has_printed_f = False self.has_printed_b = False def _detect(self, tensor, name, backward): err = None if (torch.is_floating_point(tensor) and (tensor.numel() >= 2)): with torch.no_grad(): if torch.isnan(tensor).any(): err = 'NaN' elif torch.isinf(tensor).any(): err = 'Inf' if (err is not None): err = f"{err} detected in output of {name}, shape: {tensor.shape}, {('backward' if backward else 'forward')}" return err def _apply(self, module, inp, x, backward): if torch.is_tensor(x): if (isinstance(inp, tuple) and (len(inp) > 0)): inp = inp[0] err = self._detect(x, module.__module_name, backward) if (err is not None): if (torch.is_tensor(inp) and (not backward)): err += f' input max: {inp.max().item()}, input min: {inp.min().item()}' has_printed_attr = ('has_printed_b' if backward else 'has_printed_f') logger.warning(err) setattr(self, has_printed_attr, True) elif isinstance(x, dict): for v in x.values(): self._apply(module, inp, v, backward) elif (isinstance(x, list) or isinstance(x, tuple)): for v in x: self._apply(module, inp, v, backward) def fhook_fn(self, module, inp, output): if (not self.has_printed_f): self._apply(module, inp, output, backward=False) def bhook_fn(self, module, inp, output): if (not self.has_printed_b): self._apply(module, inp, output, backward=True) def close(self): for hook in (self.fhooks + self.bhooks): hook.remove()
class InceptionV3(nn.Module): DEFAULT_BLOCK_INDEX = 3 BLOCK_INDEX_BY_DIM = {64: 0, 192: 1, 768: 2, 2048: 3} def __init__(self, output_blocks=[DEFAULT_BLOCK_INDEX], resize_input=True, normalize_input=True, requires_grad=False, use_fid_inception=True): super(InceptionV3, self).__init__() self.resize_input = resize_input self.normalize_input = normalize_input self.output_blocks = sorted(output_blocks) self.last_needed_block = max(output_blocks) assert (self.last_needed_block <= 3), 'Last possible output block index is 3' self.blocks = nn.ModuleList() if use_fid_inception: inception = fid_inception_v3() else: inception = _inception_v3(pretrained=True) block0 = [inception.Conv2d_1a_3x3, inception.Conv2d_2a_3x3, inception.Conv2d_2b_3x3, nn.MaxPool2d(kernel_size=3, stride=2)] self.blocks.append(nn.Sequential(*block0)) if (self.last_needed_block >= 1): block1 = [inception.Conv2d_3b_1x1, inception.Conv2d_4a_3x3, nn.MaxPool2d(kernel_size=3, stride=2)] self.blocks.append(nn.Sequential(*block1)) if (self.last_needed_block >= 2): block2 = [inception.Mixed_5b, inception.Mixed_5c, inception.Mixed_5d, inception.Mixed_6a, inception.Mixed_6b, inception.Mixed_6c, inception.Mixed_6d, inception.Mixed_6e] self.blocks.append(nn.Sequential(*block2)) if (self.last_needed_block >= 3): block3 = [inception.Mixed_7a, inception.Mixed_7b, inception.Mixed_7c, nn.AdaptiveAvgPool2d(output_size=(1, 1))] self.blocks.append(nn.Sequential(*block3)) for param in self.parameters(): param.requires_grad = requires_grad def forward(self, inp): outp = [] x = inp if self.resize_input: x = F.interpolate(x, size=(299, 299), mode='bilinear', align_corners=False) if self.normalize_input: x = ((2 * x) - 1) for (idx, block) in enumerate(self.blocks): x = block(x) if (idx in self.output_blocks): outp.append(x) if (idx == self.last_needed_block): break return outp
class ImageDataset(Dataset): def __init__(self, tokenizer=None, width: int=256, height: int=256, base_width: int=256, base_height: int=256, use_caption: bool=False, image_dir: str='', single_img_prompt: str='', use_bucketing: bool=False, fallback_prompt: str='', **kwargs): self.tokenizer = tokenizer self.img_types = ('.png', '.jpg', '.jpeg', '.bmp') self.use_bucketing = use_bucketing self.image_dir = self.get_images_list(image_dir) self.fallback_prompt = fallback_prompt self.use_caption = use_caption self.single_img_prompt = single_img_prompt self.width = width self.height = height def get_images_list(self, image_dir): if os.path.exists(image_dir): imgs = [x for x in os.listdir(image_dir) if x.endswith(self.img_types)] full_img_dir = [] for img in imgs: full_img_dir.append(f'{image_dir}/{img}') return sorted(full_img_dir) return [''] def image_batch(self, index): train_data = self.image_dir[index] img = train_data try: img = torchvision.io.read_image(img, mode=torchvision.io.ImageReadMode.RGB) except: img = T.transforms.PILToTensor()(Image.open(img).convert('RGB')) width = self.width height = self.height if self.use_bucketing: (_, h, w) = img.shape (width, height) = sensible_buckets(width, height, w, h) resize = T.transforms.Resize((height, width), antialias=True) img = resize(img) img = repeat(img, 'c h w -> f c h w', f=16) prompt = get_text_prompt(file_path=train_data, text_prompt=self.single_img_prompt, fallback_prompt=self.fallback_prompt, ext_types=self.img_types, use_caption=True) prompt_ids = get_prompt_ids(prompt, self.tokenizer) return (img, prompt, prompt_ids) def __getname__(): return 'image' def __len__(self): if os.path.exists(self.image_dir[0]): return len(self.image_dir) else: return 0 def __getitem__(self, index): (img, prompt, prompt_ids) = self.image_batch(index) example = {'pixel_values': ((img / 127.5) - 1.0), 'prompt_ids': prompt_ids[0], 'text_prompt': prompt, 'dataset': self.__getname__()} return example
def skip(splits, save_folder, conf): skip = True split_files = {'train': TRAIN_JSON, 'valid': VALID_JSON, 'valid_small': VALID_SMALL, 'test': TEST_JSON} for split in splits: if (not (save_folder / split_files[split]).exists()): skip = False code_folder = (save_folder / 'codes') if (not code_folder.exists()): skip = False save_opt = (save_folder / OPT_FILE) if (skip is True): if save_opt.is_file(): opts_old = load_pkl(save_opt.as_posix()) if (opts_old == conf): skip = True else: skip = False else: skip = False return skip
class TransformBase(metaclass=AutodocABCMeta): def __init__(self): self.inversion_state = None def proper_inversion(self): return False def requires_inversion_state(self): return True def identity_inversion(self): return (not self.requires_inversion_state) def to_dict(self): state = {'name': type(self).__name__} for k in inspect.signature(self.__init__).parameters: v = getattr(self, k) state[k] = (v.name if isinstance(v, Enum) else deepcopy(v)) return state def from_dict(cls, state: dict): return cls(**state) def __getstate__(self): return {k: v for (k, v) in self.to_dict().items() if (k != 'name')} def __setstate__(self, state): self.__init__(**state) def train(self, time_series: TimeSeries): raise NotImplementedError def __call__(self, time_series: TimeSeries) -> TimeSeries: raise NotImplementedError def invert(self, time_series: TimeSeries, retain_inversion_state=False) -> TimeSeries: if (not self.proper_inversion): logger.warning(f'Transform {self} is not strictly invertible. Calling invert() is not guaranteed to recover the original time series exactly!') if (self.requires_inversion_state and (self.inversion_state is None)): raise RuntimeError('Inversion state not set. Please call this transform on an input time series before calling invert(). If you are trying to call invert() a second time, please supply the option `retain_inversion_state=True` to the first call.') inverted = self._invert(time_series) if (not retain_inversion_state): self.inversion_state = None return inverted def _invert(self, time_series: TimeSeries) -> TimeSeries: return time_series def __repr__(self): kwargs = self.to_dict() name = kwargs.pop('name') kwargs_str = ', '.join((f'{k}={v}' for (k, v) in sorted(kwargs.items()))) return f'{name}({kwargs_str})'
def test_pair_tuple_crop(): arr = np.arange(45).reshape(9, 5) out = crop(arr, ((1, 2),)) assert_array_equal(out[0], [6, 7]) assert_array_equal(out[(- 1)], [31, 32]) assert_equal(out.shape, (6, 2))
def _make_taus(batch_size: int, n_quantiles: int, training: bool, device: torch.device) -> torch.Tensor: if training: taus = torch.rand(batch_size, n_quantiles, device=device) else: taus = torch.linspace(start=0, end=1, steps=n_quantiles, device=device, dtype=torch.float32) taus = taus.view(1, (- 1)).repeat(batch_size, 1) return taus
def load_rgb(path, downscale=1): img = imageio.imread(path) img = skimage.img_as_float32(img) if (downscale != 1): img = rescale(img, (1.0 / downscale), anti_aliasing=False, channel_axis=(- 1)) return img dataset_dir = '/data/chenziyu/myprojects/PanoData' root_dir = '/data/chenziyu/myprojects/PanoData/my_sun360' mask_path = '/data/chenziyu/myprojects/OmniDreamer/assets/90binarymask.png' split = ['train', 'val', 'test'] train_num = 33000 test_num = 1260 downscale = 2 out_dir = '{}/BLIP_sun360_d{:01}_t{:05}_v{:05}'.format(dataset_dir, downscale, train_num, test_num) source_dir = os.path.join(out_dir, 'source') target_dir = os.path.join(out_dir, 'target') os.makedirs(source_dir, exist_ok=True) os.makedirs(target_dir, exist_ok=True) path_dict = sorted([y for x in os.walk(root_dir) for y in glob.glob(os.path.join(x[0], '*.jpg'))]) mask = load_rgb(mask_path, downscale=downscale) train_data = [] test_data = [] processor = Blip2Processor.from_pretrained('Salesforce/blip2-opt-2.7b') model = Blip2ForConditionalGeneration.from_pretrained('Salesforce/blip2-opt-2.7b', torch_dtype=torch.float16) model.to('cuda') a = 1 for idx in tqdm(range(0, (train_num + test_num))): try: image = load_rgb(path_dict[idx], downscale=downscale) image_BLIP = Image.open(path_dict[idx]) inputs = processor(images=image_BLIP, return_tensors='pt').to('cuda', torch.float16) generated_ids = model.generate(**inputs) generated_prompt = processor.batch_decode(generated_ids, skip_special_tokens=True)[0].strip() source = np.multiply(image, mask) target = np.clip((image * 255), 0, 255).astype(np.uint8) source = np.clip((source * 255), 0, 255).astype(np.uint8) if (idx < train_num): source_path = '{}/train_{:05}.jpg'.format(source_dir, idx) target_path = '{}/train_{:05}.jpg'.format(target_dir, idx) image_dict = {'source': '{}/train_{:05}.jpg'.format('source', idx), 'target': '{}/train_{:05}.jpg'.format('target', idx), 'prompt': generated_prompt} train_data.append(image_dict) elif ((idx >= train_num) and (idx < (train_num + test_num))): source_path = '{}/test_{:05}.jpg'.format(source_dir, idx) target_path = '{}/test_{:05}.jpg'.format(target_dir, idx) image_dict = {'source': '{}/test_{:05}.jpg'.format('source', idx), 'target': '{}/test_{:05}.jpg'.format('target', idx), 'prompt': generated_prompt} test_data.append(image_dict) else: break cv2.imwrite(source_path, cv2.cvtColor(source, cv2.COLOR_RGB2BGR)) cv2.imwrite(target_path, cv2.cvtColor(target, cv2.COLOR_RGB2BGR)) except: continue train_json_path = '{}/train.json'.format(out_dir) test_json_path = '{}/test.json'.format(out_dir) with open(train_json_path, 'w') as out_file: json.dump(train_data, out_file, sort_keys=True, indent=4, ensure_ascii=False) with open(test_json_path, 'w') as out_file: json.dump(test_data, out_file, sort_keys=True, indent=4, ensure_ascii=False)
def at_loss(forward, x, y, train=True, epsilon=8.0): ce = cross_entropy(forward(x, train=train, update_batch_stats=False), y) ce.backward() d = x.grad xp = cuda.get_array_module(x.data) d = get_normalized_vector(d, xp) x_adv = (x + (epsilon * d)) return cross_entropy(forward(x_adv, train=train, update_batch_stats=False), y)
def train_val_split(base_dataset: torchvision.datasets.CIFAR10): num_classes = 10 base_dataset = np.array(base_dataset) train_n = int(((len(base_dataset) * 0.9) / num_classes)) train_idxs = [] val_idxs = [] for i in range(num_classes): idxs = np.where((base_dataset == i))[0] np.random.shuffle(idxs) train_idxs.extend(idxs[:]) val_idxs.extend(idxs[train_n:]) np.random.shuffle(train_idxs) np.random.shuffle(val_idxs) return (train_idxs, val_idxs)
class TreeBeam(object): def __init__(self, size, cuda, vocabs, rnn_size): self.size = size self.vocabs = vocabs self.tt = (torch.cuda if cuda else torch) self.rnn_size = rnn_size self.scores = self.tt.FloatTensor(size).zero_() self.prevKs = [] self.nextYs = [self.tt.LongTensor(self.size).fill_(self.vocabs['next_rules'].stoi['<blank>'])] self.valid = [[0]] self.nextYs[0][0] = self.vocabs['prev_rules'].stoi['<s>'] self.eosTop = False self.attn = [] self.finished = [] self.stacks = [[('MemberDeclaration', '<s>', Variable(self.tt.FloatTensor(1, 1, self.rnn_size).zero_(), requires_grad=False))] for i in range(0, self.size)] def getCurrentState(self): batch = {'nt': self.tt.LongTensor(self.size, 1), 'prev_rules': self.tt.LongTensor(self.size, 1), 'parent_rules': self.tt.LongTensor(self.size, 1), 'parent_states': {}} for i in range(0, len(self.nextYs[(- 1)])): if (len(self.prevKs) == 0): rule = '<s>' elif (self.nextYs[(- 1)][i] >= len(self.vocabs['next_rules'])): rule = '<unk>' else: rule = self.vocabs['next_rules'].itos[self.nextYs[(- 1)][i]] if (len(self.stacks[i]) == 0): (nt, parent_rule, parent_state) = ('MemberDeclaration', '<s>', Variable(self.tt.FloatTensor(1, 1, self.rnn_size).zero_(), requires_grad=False)) else: (nt, parent_rule, parent_state) = self.stacks[i][(- 1)] batch['nt'][i][0] = self.vocabs['nt'].stoi[nt] batch['prev_rules'][i][0] = self.vocabs['prev_rules'].stoi[CDDataset.getAnonRule(rule)] batch['parent_rules'][i][0] = self.vocabs['prev_rules'].stoi[parent_rule] batch['parent_states'][i] = {} batch['parent_states'][i][0] = parent_state return batch def getCurrentOrigin(self): return self.prevKs[(- 1)] def advance(self, wordLk, attnOut, rnn_output): numWords = wordLk.size(1) if (len(self.prevKs) > 0): beamLk = (wordLk + self.scores.unsqueeze(1).expand_as(wordLk)) for i in range(self.nextYs[(- 1)].size(0)): if (len(self.stacks[i]) == 0): beamLk[i] = (- 1e+20) else: beamLk = wordLk[0] flatBeamLk = beamLk.view((- 1)) (bestScores, bestScoresId) = flatBeamLk.topk(self.size, 0, True, True) self.scores = bestScores oldStacks = self.stacks self.stacks = [[] for i in range(0, self.size)] prevK = (bestScoresId / numWords) self.prevKs.append(prevK) self.nextYs.append((bestScoresId - (prevK * numWords))) self.attn.append(attnOut.index_select(0, prevK)) self.stacks = [copy.deepcopy(oldStacks[k]) for k in prevK] for i in range(0, self.size): currentRule = (bestScoresId[i] - (prevK[i] * numWords)) if (currentRule >= len(self.vocabs['next_rules'])): rule = '<unk>' else: rule = self.vocabs['next_rules'].itos[currentRule] try: self.stacks[i].pop() except: pass if (not CDDataset._is_terminal_rule(rule)): if (rule != '<blank>'): for elem in rhs(rule).split('___')[::(- 1)]: if elem[0].isupper(): self.stacks[i].append((elem, rule, rnn_output[prevK[i]].unsqueeze(0))) for i in range(self.nextYs[(- 1)].size(0)): if (len(self.stacks[i]) == 0): s = self.scores[i] self.finished.append((s, (len(self.nextYs) - 1), i)) if (len(self.stacks[0]) == 0): self.eosTop = True def done(self): return (self.eosTop and (len(self.finished) >= 1)) def getFinal(self): if (len(self.finished) == 0): self.finished.append((self.scores[0], (len(self.nextYs) - 1), 0)) self.finished.sort(key=(lambda a: (- a[0]))) return self.finished[0] def getHyp(self, timestep, k): (hyp, attn) = ([], []) for j in range((len(self.prevKs[:timestep]) - 1), (- 1), (- 1)): hyp.append(self.nextYs[(j + 1)][k]) attn.append(self.attn[j][k]) k = self.prevKs[j][k] return (hyp[::(- 1)], torch.stack(attn[::(- 1)]))
class DNNLowPElementwiseLinearOpTest(hu.HypothesisTestCase): (N=st.integers(32, 256), D=st.integers(32, 256), empty_batch=st.booleans(), in_quantized=st.booleans(), out_quantized=st.booleans(), **hu.gcs_cpu_only) def test_dnnlowp_elementwise_linear_int(self, N, D, empty_batch, in_quantized, out_quantized, gc, dc): if empty_batch: N = 0 min_ = (- 100) max_ = (min_ + 255) X = np.round(((np.random.rand(N, D) * (max_ - min_)) + min_)) X = X.astype(np.float32) if (N != 0): X[(0, 0)] = min_ X[(0, 1)] = max_ a = np.round(((np.random.rand(D) * 255) - 128)).astype(np.float32) a[0] = (- 128) a[1] = 127 b = np.round(((np.random.rand(D) * 255) - 128)).astype(np.float32) b[0] = (- 128) b[1] = 127 Output = collections.namedtuple('Output', ['Y', 'op_type', 'engine']) outputs = [] op_engine_list = [('ElementwiseLinear', ''), ('ElementwiseLinear', 'DNNLOWP'), ('Int8ElementwiseLinear', 'DNNLOWP')] for (op_type, engine) in op_engine_list: net = core.Net('test_net') do_quantize = (('DNNLOWP' in engine) and in_quantized) do_dequantize = (('DNNLOWP' in engine) and out_quantized) if do_quantize: quantize = core.CreateOperator('Quantize', ['X'], ['X_q'], engine=engine, device_option=gc) net.Proto().op.extend([quantize]) eltwise_linear = core.CreateOperator(op_type, [('X_q' if do_quantize else 'X'), 'a', 'b'], [('Y_q' if do_dequantize else 'Y')], dequantize_output=(not do_dequantize), engine=engine, device_option=gc) net.Proto().op.extend([eltwise_linear]) if do_dequantize: dequantize = core.CreateOperator('Dequantize', ['Y_q'], ['Y'], engine=engine, device_option=gc) net.Proto().op.extend([dequantize]) self.ws.create_blob('X').feed(X, device_option=gc) self.ws.create_blob('a').feed(a, device_option=gc) self.ws.create_blob('b').feed(b, device_option=gc) self.ws.run(net) outputs.append(Output(Y=self.ws.blobs['Y'].fetch(), op_type=op_type, engine=engine)) check_quantized_results_close(outputs)
def get_inverse_square_root_decay(optimizer, num_warmup_steps=0, last_epoch=(- 1)): def lr_lambda(current_step: int): if (current_step < num_warmup_steps): return (float(current_step) / float(max(1, num_warmup_steps))) else: return ((num_warmup_steps / current_step) ** 0.5) return LambdaLR(optimizer, lr_lambda, last_epoch)
class _TestTorchSubModelRaisingException(torch.nn.Module): def __init__(self, in_features: int, out_features: int): super().__init__() self.lin = torch.nn.Linear(in_features, out_features) def forward(self, x: torch.Tensor) -> torch.Tensor: x = self.lin(x) if (int('1') == 1): raise _DemoException('uh') return x
def gen(n=2, repeatParts=False, separator='-', lists=(LEFT, RIGHT)): name = [] for word in lists[:n]: part = None while ((not part) or ((part in name) and (not repeatParts))): part = random.choice(word) name.append(part) return separator.join(name)
def build_backbone(cfg): backbone_cfg = deepcopy(cfg) name = backbone_cfg.pop('name') if (name == 'VGG'): return VGG(**backbone_cfg) elif (name == 'ResNet'): return ResNet(**backbone_cfg) elif (name == 'ResNeXt'): return ResNeXt(**backbone_cfg) elif (name == 'WideResNet'): return WideResNet(**backbone_cfg) elif (name == 'SqueezeNet'): return SqueezeNet(**backbone_cfg) elif (name == 'MobileNetV2'): return MobileNetV2(**backbone_cfg) elif (name == 'MobileNetV3'): return MobileNetV3(**backbone_cfg) elif (name == 'ShuffleNetV2'): return ShuffleNetV2(**backbone_cfg) elif (name == 'EfficientNet'): return EfficientNet(**backbone_cfg) elif (name == 'RegNet'): return RegNet(**backbone_cfg) elif (name == 'ConvNeXt'): return ConvNeXt(**backbone_cfg) elif (name == 'VisionTransformer'): return VisionTransformer(**backbone_cfg) elif (name == 'STDCNet'): return STDCNet(**backbone_cfg) elif (name == 'MSCAN'): return MSCAN(**backbone_cfg) elif (name == 'LSPNetBackbone'): return LSPNetBackbone(**backbone_cfg) elif (name == 'EfficientNetLite'): return EfficientNetLite(**backbone_cfg) elif (name == 'TopFormerBackbone'): return TopFormerBackbone(**backbone_cfg) elif (name == 'MixVisionTransformer'): return MixVisionTransformer(**backbone_cfg) elif (name == 'IncepTransformer'): return IncepTransformer(**backbone_cfg) elif (name == 'EfficientNetLite'): return EfficientNetLite(**backbone_cfg) elif (name == 'CustomCspNet'): return CustomCspNet(**backbone_cfg) elif (name == 'CSPDarknet'): return CSPDarknet(**backbone_cfg) elif (name == 'SGCPNetBackbone'): return SGCPNetBackbone(**backbone_cfg) elif (name == 'LFDResNet'): return LFDResNet(**backbone_cfg) elif (name == 'YOLOv5CSPDarknet'): return YOLOv5CSPDarknet(**backbone_cfg) elif (name == 'YOLOXCSPDarknet'): return YOLOXCSPDarknet(**backbone_cfg) elif (name == 'YOLOv6EfficientRep'): return YOLOv6EfficientRep(**backbone_cfg) elif (name == 'YOLOv7CSPVoVNet'): return YOLOv7CSPVoVNet(**backbone_cfg) elif (name == 'YOLOXPAIEfficientRep'): return YOLOXPAIEfficientRep(**backbone_cfg) elif (name == 'RepVGG'): return RepVGG(**backbone_cfg) elif (name == 'RegSegBackbone'): return RegSegBackbone(**backbone_cfg) else: raise NotImplementedError(name)
class JsonLoader(Loader): def __init__(self, fields=None, dropna=False): super(JsonLoader, self).__init__() self.dropna = dropna self.fields = None self.fields_list = None if fields: self.fields = {} for (k, v) in fields.items(): self.fields[k] = (k if (v is None) else v) self.fields_list = list(self.fields.keys()) def _load(self, path): ds = DataSet() for (idx, d) in _read_json(path, fields=self.fields_list, dropna=self.dropna): if self.fields: ins = {self.fields[k]: v for (k, v) in d.items()} else: ins = d ds.append(Instance(**ins)) return ds
class TFDebertaV2PreTrainedModel(metaclass=DummyObject): _backends = ['tf'] def __init__(self, *args, **kwargs): requires_backends(self, ['tf'])
def group_norm(x, num_groups, weight=None, bias=None, eps=1e-05): input_shape = x.shape ndim = len(input_shape) (N, C) = input_shape[:2] G = num_groups assert ((C % G) == 0), 'input channel dimension must divisible by number of groups' x = x.view(N, G, (- 1)) mean = x.mean((- 1), keepdim=True) var = x.var((- 1), keepdim=True) x = ((x - mean) / (var + eps).sqrt()) x = x.view(input_shape) view_shape = ((1, (- 1)) + ((1,) * (ndim - 2))) if (weight is not None): return ((x * weight.view(view_shape)) + bias.view(view_shape)) return x
def scoped_configure(dir=None, format_strs=None, comm=None): prevlogger = Logger.CURRENT configure(dir=dir, format_strs=format_strs, comm=comm) try: (yield) finally: Logger.CURRENT.close() Logger.CURRENT = prevlogger
class ConcatModel(torch.nn.Module): def __init__(self): super(ConcatModel, self).__init__() def forward(self, x): return torch.concat([x, x])
class BasicBlock(nn.Module): expansion = 1 def __init__(self, in_channels, channels, stride=1, dilation=1): super(BasicBlock, self).__init__() out_channels = (self.expansion * channels) self.conv1 = nn.Conv2d(in_channels, channels, kernel_size=3, stride=stride, padding=dilation, dilation=dilation, bias=False) self.bn1 = nn.BatchNorm2d(channels) self.conv2 = nn.Conv2d(channels, channels, kernel_size=3, stride=1, padding=dilation, dilation=dilation, bias=False) self.bn2 = nn.BatchNorm2d(channels) if ((stride != 1) or (in_channels != out_channels)): conv = nn.Conv2d(in_channels, out_channels, kernel_size=1, stride=stride, bias=False) bn = nn.BatchNorm2d(out_channels) self.downsample = nn.Sequential(conv, bn) else: self.downsample = nn.Sequential() def forward(self, x): out = F.relu(self.bn1(self.conv1(x))) out = self.bn2(self.conv2(out)) out = (out + self.downsample(x)) out = F.relu(out) return out
class ROIHeadsPatcher(): def __init__(self, cfg, heads): self.heads = heads self.use_heatmap_max_keypoint = cfg.EXPORT_CAFFE2.USE_HEATMAP_MAX_KEYPOINT def mock_roi_heads(self, tensor_mode=True): kpt_heads_mod = keypoint_head.BaseKeypointRCNNHead.__module__ mask_head_mod = mask_head.BaseMaskRCNNHead.__module__ mock_ctx_managers = [mock_fastrcnn_outputs_inference(tensor_mode=tensor_mode, check=True, box_predictor_type=type(self.heads.box_predictor))] if getattr(self.heads, 'keypoint_on', False): mock_ctx_managers += [mock_keypoint_rcnn_inference(tensor_mode, kpt_heads_mod, self.use_heatmap_max_keypoint)] if getattr(self.heads, 'mask_on', False): mock_ctx_managers += [mock_mask_rcnn_inference(tensor_mode, mask_head_mod)] with contextlib.ExitStack() as stack: for mgr in mock_ctx_managers: stack.enter_context(mgr) (yield)
class CharCNN(nn.Module): def __init__(self, args): super(CharCNN, self).__init__() self.conv1 = nn.Sequential(nn.Conv1d(args.num_features, 256, kernel_size=7, stride=1), nn.ReLU(), nn.MaxPool1d(kernel_size=3, stride=3)) self.conv2 = nn.Sequential(nn.Conv1d(256, 256, kernel_size=7, stride=1), nn.ReLU(), nn.MaxPool1d(kernel_size=3, stride=3)) self.conv3 = nn.Sequential(nn.Conv1d(256, 256, kernel_size=3, stride=1), nn.ReLU()) self.conv4 = nn.Sequential(nn.Conv1d(256, 256, kernel_size=3, stride=1), nn.ReLU()) self.conv5 = nn.Sequential(nn.Conv1d(256, 256, kernel_size=3, stride=1), nn.ReLU()) self.conv6 = nn.Sequential(nn.Conv1d(256, 256, kernel_size=3, stride=1), nn.ReLU(), nn.MaxPool1d(kernel_size=3, stride=3)) self.fc1 = nn.Sequential(nn.Linear(8704, 1024), nn.ReLU(), nn.Dropout(p=args.dropout)) self.fc2 = nn.Sequential(nn.Linear(1024, 1024), nn.ReLU(), nn.Dropout(p=args.dropout)) self.fc3 = nn.Linear(1024, 4) self.log_softmax = nn.LogSoftmax() def forward(self, x): x = self.conv1(x) x = self.conv2(x) x = self.conv3(x) x = self.conv4(x) x = self.conv5(x) x = self.conv6(x) x = x.view(x.size(0), (- 1)) x = self.fc1(x) x = self.fc2(x) x = self.fc3(x) x = self.log_softmax(x) return x
class ConfigHandler(): args: Namespace unknown: List[str] def __call__(self) -> Dict[(str, Any)]: config = self._obtain_config() config = merge(config, self.args.merge_files) sys.argv = ([sys.argv[0]] + self.unknown) self._edit_config(config) return config def _edit_config(self, config): to_delete = [] for (k, v) in config.items(): if (v is None): to_delete.append(k) if (config[k] == ''): to_delete.append(k) for k in to_delete: del config[k]
class ArrayWrapper(): def __init__(self, array, **kwargs): self.array = array def __array_interface__(self): return self.array.__array_interface__
def get_cfl_setup(CFL=None, dt=None): if ((CFL is None) and (dt is None)): raise ValueError('Specifiy either CFL or dt in CFL setup') def setup_cfl_condition(problem): ts_conf = problem.ts_conf mesh = problem.domain.mesh dim = mesh.dim first_field = list(problem.fields.values())[0] first_field_name = list(problem.fields.keys())[0] approx_order = first_field.approx_order mats = problem.create_materials(['a', 'D']) try: velo = problem.conf_materials['material_a__0'].values['val'] max_velo = nm.max(nm.linalg.norm(velo)) except KeyError: max_velo = 1 try: diffusion = problem.conf_materials['material_D__0'].values['val'] max_diffusion = nm.max(nm.linalg.norm(diffusion)) except KeyError: max_diffusion = None dx = nm.min(problem.domain.mesh.cmesh.get_volumes(dim)) output('Preprocess hook - setup_cfl_condition:...') output('Approximation order of field {}({}) is {}'.format(first_field_name, first_field.family_name, approx_order)) output('Space divided into {0} cells, {1} steps, step size {2}'.format(mesh.n_el, len(mesh.coors), dx)) if (dt is None): adv_dt = get_cfl_advection(max_velo, dx, approx_order, CFL) diff_dt = get_cfl_diffusion(max_diffusion, dx, approx_order, CFL) _dt = min(adv_dt, diff_dt) else: output('CFL coefficient {0} ignored, dt specified directly'.format(CFL)) _dt = dt tn = int(nm.ceil(((ts_conf.t1 - ts_conf.t0) / _dt))) dtdx = (_dt / dx) ts_conf.dt = _dt ts_conf.n_step = tn ts_conf.cour = (max_velo * dtdx) output('Time divided into {0} nodes, {1} steps, step size is {2}'.format((tn - 1), tn, _dt)) output('Courant number c = max(norm(a)) * dt/dx = {0}'.format(ts_conf.cour)) output('Time stepping solver is {}'.format(ts_conf.kind)) output('... CFL setup done.') return setup_cfl_condition
class RandomStrVar(TemplateVar): def __len__(self): return 1 def _randomize(self): return ''.join((random.choice((string.ascii_letters + string.digits)) for _ in range(10))) def __getitem__(self, index): return self._randomize() def __iter__(self): while True: (yield self._randomize())
class TorchMBBatchRLAlgorithm(MBBatchRLAlgorithm): def to(self, device): for net in self.trainer.networks: net.to(device) for net in self.model_trainer.networks: net.to(device) def training_mode(self, mode): for net in self.trainer.networks: net.train(mode) for net in self.model_trainer.networks: net.train(mode)
class AuxiliaryHead(nn.Module): def __init__(self, input_size, C, n_classes): assert (input_size in [7, 8]) super().__init__() self.net = nn.Sequential(nn.ReLU(inplace=True), nn.AvgPool2d(5, stride=(input_size - 5), padding=0, count_include_pad=False), nn.Conv2d(C, 128, kernel_size=1, bias=False), nn.BatchNorm2d(128), nn.ReLU(inplace=True), nn.Conv2d(128, 768, kernel_size=2, bias=False), nn.BatchNorm2d(768), nn.ReLU(inplace=True)) self.linear = nn.Linear(768, n_classes) def forward(self, x): out = self.net(x) out = out.view(out.size(0), (- 1)) logits = self.linear(out) return logits
class MotionSyntheticProtocol(Protocol): def __init__(self, data_dir='/home/piaozx/liuwen/p300/human_pose/processed'): super().__init__() self.data_dir = data_dir self.processed_dir = os.path.join(data_dir, 'processed') self.train_ids_file = 'train.txt' self.test_ids_file = 'val.txt' self.eval_path = 'MS_protocol.json' full_eval_path = os.path.join(os.path.abspath(os.path.dirname(os.path.dirname(os.path.dirname(__file__)))), 'data', 'MS_protocol.json') self.eval_info = load_json_file(full_eval_path)['val'] self.vid_names = list(self.eval_info.keys()) self._all_vid_smpls = {} self._all_vid_offsets = {} self._all_vid_kps = {} self._num_source = 1 self._load_smpls = False self._load_kps = False def __len__(self): return len(self.vid_names) def take_images_paths(self, vid_name, start, end): vid_path = os.path.join(self.processed_dir, vid_name, 'images') vid_images_paths = glob.glob(os.path.join(vid_path, '*')) vid_images_paths.sort() images_paths = vid_images_paths[start:(end + 1)] return images_paths def setup(self, num_sources=1, load_smpls=False, load_kps=False): self._num_source = num_sources self._load_smpls = load_smpls self._load_kps = load_kps def __getitem__(self, item): num_sources = self._num_source load_smpls = self._load_smpls load_kps = self._load_kps vid_name = self.vid_names[item] vid_info = self.eval_info[vid_name] eval_info = dict() src_vid_smpls = self.get_smpls(vid_name) src_vid_kps = self.get_kps(vid_name) src_vid_path = os.path.join(self.processed_dir, vid_name, 'images') src_img_paths = glob.glob(os.path.join(src_vid_path, '*')) src_img_paths.sort() src_img_names = vid_info['s_n'][str(num_sources)] src_img_ids = [int(t.split('.')[0].split('_')[(- 1)]) for t in src_img_names] eval_info['source'] = {'s_n': num_sources, 'name': vid_name, 'formated_name': self.format_name(vid_name), 'vid_path': os.path.join(self.processed_dir, vid_name, 'images'), 'images': [src_img_paths[t] for t in src_img_ids], 'smpls': (src_vid_smpls[src_img_ids] if load_smpls else None), 'kps': (src_vid_kps[src_img_ids] if load_kps else None)} self_imitation = vid_info['self_imitation'] eval_info['self_imitation'] = {'name': self_imitation['target'], 'formated_name': self.format_name(self_imitation['target']), 'images': src_img_paths[self_imitation['range'][0]:(self_imitation['range'][1] + 1)], 'smpls': (src_vid_smpls[self_imitation['range'][0]:(self_imitation['range'][1] + 1)] if load_smpls else None), 'kps': (src_vid_kps[self_imitation['range'][0]:(self_imitation['range'][1] + 1)] if load_kps else None), 'self_imitation': True} cross_imitation = vid_info['cross_imitation'] target_vid_name = cross_imitation['target'] target_vid_smpls = self.get_smpls(target_vid_name) target_vid_kps = self.get_kps(target_vid_name) cross_images_paths = self.take_images_paths(vid_name=target_vid_name, start=cross_imitation['range'][0], end=cross_imitation['range'][1]) eval_info['cross_imitation'] = {'name': target_vid_name, 'formated_name': self.format_name(target_vid_name), 'images': cross_images_paths, 'smpls': (target_vid_smpls[cross_imitation['range'][0]:(cross_imitation['range'][1] + 1)] if load_smpls else None), 'kps': (target_vid_kps[cross_imitation['range'][0]:(cross_imitation['range'][1] + 1)] if load_kps else None), 'self_imitation': False} eval_info['flag'] = self.take_images_paths(vid_name=vid_name, start=vid_info['flag'][0], end=vid_info['flag'][1]) assert ((cross_imitation['range'][1] - cross_imitation['range'][0]) == (vid_info['flag'][1] - vid_info['flag'][0])) return eval_info def get_smpl_path(self, name): smpl_path = os.path.join(self.processed_dir, name, 'pose_shape.pkl') return smpl_path def get_kps_path(self, name): smpl_path = os.path.join(self.processed_dir, name, 'kps.pkl') return smpl_path def get_smpls(self, name): smpls = None if (name in self.eval_info): if (name not in self._all_vid_smpls): smpl_path = self.get_smpl_path(name) smpl_data = load_pickle_file(smpl_path) cams = smpl_data['cams'] thetas = smpl_data['pose'] betas = np.repeat(smpl_data['shape'], cams.shape[0], axis=0) smpls = np.concatenate([cams, thetas, betas], axis=1) self._all_vid_smpls[name] = smpls else: smpls = self._all_vid_smpls[name] return smpls def get_kps(self, name): kps = None if (name in self.eval_info): if (name not in self._all_vid_kps): kps_path = self.get_kps_path(name) kps = load_pickle_file(kps_path) self._all_vid_kps[name] = kps else: kps = self._all_vid_kps[name] return kps def total_frames(self): total = 0 for (vid_name, vid_info) in self.eval_info.items(): src_vid = os.path.join(self.processed_dir, vid_name) length = len(os.listdir(src_vid)) total += length return total
def GenerateSM80_Simt_f64(manifest, args): layouts = [(LayoutType.ColumnMajor, LayoutType.ColumnMajor, LayoutType.ColumnMajor), (LayoutType.ColumnMajor, LayoutType.RowMajor, LayoutType.ColumnMajor), (LayoutType.RowMajor, LayoutType.ColumnMajor, LayoutType.ColumnMajor), (LayoutType.RowMajor, LayoutType.RowMajor, LayoutType.ColumnMajor)] math_instructions = [MathInstruction([1, 1, 1], DataType.f64, DataType.f64, DataType.f64, OpcodeClass.Simt, MathOperation.multiply_add)] min_cc = 80 max_cc = 1024 alignment_constraints = [1] for math_inst in math_instructions: tile_descriptions = [TileDescription([128, 128, 8], 3, [4, 2, 1], math_inst, min_cc, max_cc), TileDescription([128, 64, 8], 4, [2, 2, 1], math_inst, min_cc, max_cc), TileDescription([64, 128, 8], 4, [2, 2, 1], math_inst, min_cc, max_cc), TileDescription([64, 64, 8], 5, [2, 1, 1], math_inst, min_cc, max_cc), TileDescription([128, 32, 8], 5, [2, 1, 1], math_inst, min_cc, max_cc), TileDescription([32, 128, 8], 5, [1, 2, 1], math_inst, min_cc, max_cc)] data_type = [math_inst.element_a, math_inst.element_b, math_inst.element_accumulator, math_inst.element_accumulator] CreateGemmOperator(manifest, layouts, tile_descriptions, data_type, alignment_constraints)
class DealCorpus(object): def __init__(self, config): self.config = config self.train_corpus = self._read_file(self.config.train_path) self.val_corpus = self._read_file(self.config.val_path) self.test_corpus = self._read_file(self.config.test_path) self._extract_vocab() self._extract_goal_vocab() self._extract_outcome_vocab() print('Loading corpus finished.') def _read_file(self, path): with open(path, 'r') as f: data = f.readlines() return self._process_dialogue(data) def _process_dialogue(self, data): def transform(token_list): (usr, sys) = ([], []) ptr = 0 while (ptr < len(token_list)): turn_ptr = ptr turn_list = [] while True: cur_token = token_list[turn_ptr] turn_list.append(cur_token) turn_ptr += 1 if (cur_token == EOS): ptr = turn_ptr break all_sent_lens.append(len(turn_list)) if (turn_list[0] == USR): usr.append(Pack(utt=turn_list, speaker=USR)) elif (turn_list[0] == SYS): sys.append(Pack(utt=turn_list, speaker=SYS)) else: raise ValueError('Invalid speaker') all_dlg_lens.append((len(usr) + len(sys))) return (usr, sys) new_dlg = [] all_sent_lens = [] all_dlg_lens = [] for raw_dlg in data: raw_words = raw_dlg.split() cur_dlg = [] words = raw_words[(raw_words.index('<dialogue>') + 1):raw_words.index('</dialogue>')] words += [EOS] usr_first = True if (words[0] == SYS): words = ([USR, BOD, EOS] + words) usr_first = True elif (words[0] == USR): words = ([SYS, BOD, EOS] + words) usr_first = False else: print('FATAL ERROR!!! ({})'.format(words)) exit((- 1)) (usr_utts, sys_utts) = transform(words) for (usr_turn, sys_turn) in zip(usr_utts, sys_utts): if usr_first: cur_dlg.append(usr_turn) cur_dlg.append(sys_turn) else: cur_dlg.append(sys_turn) cur_dlg.append(usr_turn) if ((len(usr_utts) - len(sys_utts)) == 1): cur_dlg.append(usr_utts[(- 1)]) elif ((len(sys_utts) - len(usr_utts)) == 1): cur_dlg.append(sys_utts[(- 1)]) cur_goal = raw_words[(raw_words.index('<partner_input>') + 1):raw_words.index('</partner_input>')] if (len(cur_goal) != 6): print('FATAL ERROR!!! ({})'.format(cur_goal)) exit((- 1)) cur_out = raw_words[(raw_words.index('<output>') + 1):raw_words.index('</output>')] if (len(cur_out) != 6): print('FATAL ERROR!!! ({})'.format(cur_out)) exit((- 1)) new_dlg.append(Pack(dlg=cur_dlg, goal=cur_goal, out=cur_out)) print(('Max utt len = %d, mean utt len = %.2f' % (np.max(all_sent_lens), float(np.mean(all_sent_lens))))) print(('Max dlg len = %d, mean dlg len = %.2f' % (np.max(all_dlg_lens), float(np.mean(all_dlg_lens))))) return new_dlg def _extract_vocab(self): all_words = [] for dlg in self.train_corpus: for turn in dlg.dlg: all_words.extend(turn.utt) vocab_count = Counter(all_words).most_common() raw_vocab_size = len(vocab_count) discard_wc = np.sum([c for (t, c) in vocab_count]) print(((('vocab size of train set = %d,\n' % (raw_vocab_size,)) + ('cut off at word %s with frequency = %d,\n' % (vocab_count[(- 1)][0], vocab_count[(- 1)][1]))) + ('OOV rate = %.2f' % ((1 - (float(discard_wc) / len(all_words))),)))) self.vocab = (SPECIAL_TOKENS_DEAL + [t for (t, cnt) in vocab_count if (t not in SPECIAL_TOKENS_DEAL)]) self.vocab_dict = {t: idx for (idx, t) in enumerate(self.vocab)} self.unk_id = self.vocab_dict[UNK] global DECODING_MASKED_TOKENS from string import ascii_letters, digits letter_set = set(list((ascii_letters + digits))) vocab_list = [t for (t, cnt) in vocab_count] masked_words = [] for word in vocab_list: tmp_set = set(list(word)) if (len((letter_set & tmp_set)) == 0): masked_words.append(word) print('Take care of {} special words (masked).'.format(len(masked_words))) def _extract_goal_vocab(self): all_goal = [] for dlg in self.train_corpus: all_goal.extend(dlg.goal) vocab_count = Counter(all_goal).most_common() raw_vocab_size = len(vocab_count) discard_wc = np.sum([c for (t, c) in vocab_count]) print(((('goal vocab size of train set = %d, \n' % (raw_vocab_size,)) + ('cut off at word %s with frequency = %d, \n' % (vocab_count[(- 1)][0], vocab_count[(- 1)][1]))) + ('OOV rate = %.2f' % ((1 - (float(discard_wc) / len(all_goal))),)))) self.goal_vocab = ([UNK] + [g for (g, cnt) in vocab_count]) self.goal_vocab_dict = {t: idx for (idx, t) in enumerate(self.goal_vocab)} self.goal_unk_id = self.goal_vocab_dict[UNK] def _extract_outcome_vocab(self): all_outcome = [] for dlg in self.train_corpus: all_outcome.extend(dlg.out) vocab_count = Counter(all_outcome).most_common() raw_vocab_size = len(vocab_count) discard_wc = np.sum([c for (t, c) in vocab_count]) print(((('outcome vocab size of train set = %d, \n' % (raw_vocab_size,)) + ('cut off at word %s with frequency = %d, \n' % (vocab_count[(- 1)][0], vocab_count[(- 1)][1]))) + ('OOV rate = %.2f' % ((1 - (float(discard_wc) / len(all_outcome))),)))) self.outcome_vocab = ([UNK] + [o for (o, cnt) in vocab_count]) self.outcome_vocab_dict = {t: idx for (idx, t) in enumerate(self.outcome_vocab)} self.outcome_unk_id = self.outcome_vocab_dict[UNK] def get_corpus(self): id_train = self._to_id_corpus('Train', self.train_corpus) id_val = self._to_id_corpus('Valid', self.val_corpus) id_test = self._to_id_corpus('Test', self.test_corpus) return (id_train, id_val, id_test) def _to_id_corpus(self, name, data): results = [] for dlg in data: if (len(dlg.dlg) < 1): continue id_dlg = [] for turn in dlg.dlg: id_turn = Pack(utt=self._sent2id(turn.utt), speaker=turn.speaker) id_dlg.append(id_turn) id_goal = self._goal2id(dlg.goal) id_out = self._outcome2id(dlg.out) results.append(Pack(dlg=id_dlg, goal=id_goal, out=id_out)) return results def _sent2id(self, sent): return [self.vocab_dict.get(t, self.unk_id) for t in sent] def _goal2id(self, goal): return [self.goal_vocab_dict.get(g, self.goal_unk_id) for g in goal] def _outcome2id(self, outcome): return [self.outcome_vocab_dict.get(o, self.outcome_unk_id) for o in outcome] def sent2id(self, sent): return self._sent2id(sent) def goal2id(self, goal): return self._goal2id(goal) def outcome2id(self, outcome): return self._outcome2id(outcome) def id2sent(self, id_list): return [self.vocab[i] for i in id_list] def id2goal(self, id_list): return [self.goal_vocab[i] for i in id_list] def id2outcome(self, id_list): return [self.outcome_vocab[i] for i in id_list]
def test_model(test_set): device = torch.device(('cuda:0' if args.use_cuda else 'cpu')) cd_event_model = load_check_point(config_dict['cd_event_model_path']) cd_entity_model = load_check_point(config_dict['cd_entity_model_path']) cd_event_model.to(device) cd_entity_model.to(device) doc_to_entity_mentions = load_entity_wd_clusters(config_dict) (_, _) = test_models(test_set, cd_event_model, cd_entity_model, device, config_dict, write_clusters=True, out_dir=args.out_dir, doc_to_entity_mentions=doc_to_entity_mentions, analyze_scores=True) run_conll_scorer()
def quadratic_L_function__numerical(n, d, num_terms=1000): if isinstance(n.parent(), sage.rings.abc.RealField): R = n.parent() else: from sage.rings.real_mpfr import RealField R = RealField() if (n < 0): raise ValueError('the Dirichlet series does not converge here') d1 = fundamental_discriminant(d) ans = R.zero() for i in range(1, num_terms): ans += R((kronecker_symbol(d1, i) / (R(i) ** n))) return ans
def mockenv_context(*remove, **update): env = os.environ update = (update or {}) remove = (remove or []) stomped = ((set(update.keys()) | set(remove)) & set(env.keys())) update_after = {k: env[k] for k in stomped} remove_after = frozenset((k for k in update if (k not in env))) try: env.update(update) [env.pop(k, None) for k in remove] (yield) finally: env.update(update_after) [env.pop(k) for k in remove_after]
def gcd(*values): if any(((not isinstance(v, int)) for v in values)): raise TypeError('value must be integers.') if any(((v <= 0) for v in values)): raise ValueError('arguments must be positive.') if (not values): raise ValueError('must give at least 1 value.') if (len(values) == 1): return values[0] if (len(values) > 2): return reduce(gcd, values) (a, b) = values while b: (a, b) = (b, (a % b)) return a