From: François Fleuret Date: Wed, 10 Jul 2024 13:28:32 +0000 (+0200) Subject: Update. X-Git-Url: https://fleuret.org/cgi-bin/gitweb/gitweb.cgi?a=commitdiff_plain;h=693af34e144cd20d2dde6a508a190d49c1a76c7f;p=culture.git Update. --- diff --git a/main.py b/main.py index 634363f..d400ab1 100755 --- a/main.py +++ b/main.py @@ -29,7 +29,6 @@ else: ###################################################################### parser = argparse.ArgumentParser( - description="An implementation of GPT with cache.", formatter_class=argparse.ArgumentDefaultsHelpFormatter, ) @@ -274,53 +273,6 @@ vocabulary_size = quiz_machine.vocabulary_size() log_string(f"vocabulary_size {vocabulary_size}") ###################################################################### - -# Compute the entropy of the training tokens - -token_count = 0 -for input in quiz_machine.batches(split="train", desc="train-entropy"): - token_count += F.one_hot(input, num_classes=quiz_machine.vocabulary_size()).sum( - (0, 1) - ) -token_probas = token_count / token_count.sum() -entropy = -torch.xlogy(token_probas, token_probas).sum() -train_set_perplexity = math.exp(entropy) - -###################################################################### -# A bit of paranoia never hurts - -if args.max_percents_of_test_in_train >= 0: - - def subsets_as_tuples(batches, cs): - s = set() - for batch in batches: - for x in batch: - s.add(tuple([v.item() for v in x])) - if len(s) == cs: - yield s - s = set() - yield s - - nb_test, nb_in_train = 0, 0 - for test_subset in subsets_as_tuples( - quiz_machine.batches(split="test", desc="test-check"), 25000 - ): - in_train = set() - for train_subset in subsets_as_tuples( - quiz_machine.batches(split="train", desc="train-check"), 25000 - ): - in_train.update(test_subset.intersection(train_subset)) - nb_in_train += len(in_train) - nb_test += len(test_subset) - - log_string( - f"data_check {nb_in_train*100/nb_test:.02f}% ({nb_in_train}/{nb_test}) of test samples are in the train set" - ) - - assert ( - nb_in_train <= args.max_percents_of_test_in_train * nb_test / 100 - ), f"More than {args.max_percents_of_test_in_train}% of test samples are in the train set" - ############################## @@ -331,7 +283,7 @@ def one_epoch(model, quiz_machine): nb_train_samples, acc_train_loss = 0, 0.0 - for input in quiz_machine.batches(split="train"): + for input in quiz_machine.batches(model, split="train"): input = input.to(device) if nb_train_samples % args.batch_size == 0: @@ -363,7 +315,7 @@ def run_tests(model, quiz_machine, deterministic_synthesis): nb_test_samples, acc_test_loss = 0, 0.0 nb_samples_accumulated = 0 - for input in quiz_machine.batches(split="test"): + for input in quiz_machine.batches(model, split="test"): input = input.to(device) bs = model(mygpt.BracketedSequence(input)) @@ -596,6 +548,15 @@ for k in range(args.nb_gpts): model.main_test_accuracy = 0.0 model.id = k + model.train_w_quizzes = quiz_machine.generate_token_sequences( + args.nb_train_samples + ).to(device) + quiz_machine.reverse_random_half_in_place(model.train_w_quizzes) + model.test_w_quizzes = quiz_machine.generate_token_sequences( + args.nb_test_samples + ).to(device) + quiz_machine.reverse_random_half_in_place(model.test_w_quizzes) + models.append(model) @@ -604,6 +565,54 @@ log_string(f"nb_parameters {nb_parameters} ({int(nb_parameters/1e6)}M)") ###################################################################### +# Compute the entropy of the training tokens + +token_count = 0 +for input in quiz_machine.batches(models[0], split="train", desc="train-entropy"): + token_count += F.one_hot(input, num_classes=quiz_machine.vocabulary_size()).sum( + (0, 1) + ) +token_probas = token_count / token_count.sum() +entropy = -torch.xlogy(token_probas, token_probas).sum() +train_set_perplexity = math.exp(entropy) + +###################################################################### +# A bit of paranoia never hurts + +if args.max_percents_of_test_in_train >= 0: + + def subsets_as_tuples(batches, cs): + s = set() + for batch in batches: + for x in batch: + s.add(tuple([v.item() for v in x])) + if len(s) == cs: + yield s + s = set() + yield s + + nb_test, nb_in_train = 0, 0 + for test_subset in subsets_as_tuples( + quiz_machine.batches(models[0], split="test", desc="test-check"), 25000 + ): + in_train = set() + for train_subset in subsets_as_tuples( + quiz_machine.batches(models[0], split="train", desc="train-check"), 25000 + ): + in_train.update(test_subset.intersection(train_subset)) + nb_in_train += len(in_train) + nb_test += len(test_subset) + + log_string( + f"data_check {nb_in_train*100/nb_test:.02f}% ({nb_in_train}/{nb_test}) of test samples are in the train set" + ) + + assert ( + nb_in_train <= args.max_percents_of_test_in_train * nb_test / 100 + ), f"More than {args.max_percents_of_test_in_train}% of test samples are in the train set" + +###################################################################### + nb_new_c_quizzes_for_train = args.nb_train_samples // 50 nb_new_c_quizzes_for_test = args.nb_test_samples // 50 @@ -654,7 +663,7 @@ for n_epoch in range(args.nb_epochs): log_string( f"cache_w_quizzes contains {quiz_machine.problem.nb_cached_quizzes()} quizzes" ) - quiz_machine.renew_w_quizzes(args.nb_train_samples // args.nb_gpts) + quiz_machine.renew_w_quizzes(model, args.nb_train_samples // args.nb_gpts) ################################################## # If all the models are good enough, generate new quizzes and diff --git a/quiz_machine.py b/quiz_machine.py index 0ae68d0..cdfba85 100755 --- a/quiz_machine.py +++ b/quiz_machine.py @@ -235,23 +235,21 @@ class QuizMachine: self.prompt_len = None self.answer_len = None - self.train_w_quizzes = self.generate_token_sequences(nb_train_samples) - self.reverse_random_half_in_place(self.train_w_quizzes) - self.train_w_quizzes = self.train_w_quizzes.to(device) + # self.train_w_quizzes = self.generate_token_sequences(nb_train_samples) + # self.reverse_random_half_in_place(self.train_w_quizzes) - self.test_w_quizzes = self.generate_token_sequences(nb_test_samples).to(device) - self.reverse_random_half_in_place(self.test_w_quizzes) - self.test_w_quizzes = self.test_w_quizzes.to(device) + # self.test_w_quizzes = self.generate_token_sequences(nb_test_samples).to(device) + # self.reverse_random_half_in_place(self.test_w_quizzes) self.train_c_quizzes = [] self.test_c_quizzes = [] - if result_dir is not None: - self.save_quizzes( - result_dir, - "culture_w_quizzes", - self.train_w_quizzes[:72], - ) + # if result_dir is not None: + # self.save_quizzes( + # result_dir, + # "culture_w_quizzes", + # self.train_w_quizzes[:72], + # ) def save_quizzes( self, @@ -287,13 +285,13 @@ class QuizMachine: predicted_answers, ) - def batches(self, split="train", desc=None): + def batches(self, model, split="train", desc=None): assert split in {"train", "test"} if split == "train": - w_quizzes = self.train_w_quizzes + w_quizzes = model.train_w_quizzes c_quizzes = self.train_c_quizzes else: - w_quizzes = self.test_w_quizzes + w_quizzes = model.test_w_quizzes c_quizzes = self.test_c_quizzes if len(c_quizzes) > 0: @@ -382,10 +380,10 @@ class QuizMachine: return result, correct - compute_accuracy(self.train_w_quizzes[:nmax], log_prefix="train") + compute_accuracy(model.train_w_quizzes[:nmax], log_prefix="train") test_result, test_correct = compute_accuracy( - self.test_w_quizzes[:nmax], log_prefix="test" + model.test_w_quizzes[:nmax], log_prefix="test" ) main_test_accuracy = test_correct.sum() / test_correct.size(0) @@ -402,8 +400,8 @@ class QuizMachine: return main_test_accuracy - def renew_w_quizzes(self, nb, for_train=True): - input = self.train_w_quizzes if for_train else self.test_w_quizzes + def renew_w_quizzes(self, model, nb, for_train=True): + input = model.train_w_quizzes if for_train else model.test_w_quizzes nb = min(nb, input.size(0)) input[:-nb] = input[nb:].clone() fresh_w_quizzes = self.generate_token_sequences(nb) @@ -507,7 +505,10 @@ class QuizMachine: def generate_quizzes(self, nb, model_for_generation, temperature=1.0): c_quizzes = torch.empty( - nb, self.train_w_quizzes.size(1), device=self.device, dtype=torch.int64 + nb, + self.prompt_len + self.answer_len + 2, + device=self.device, + dtype=torch.int64, ) seq_logproba = torch.zeros(nb, device=self.device)