X-Git-Url: https://fleuret.org/cgi-bin/gitweb/gitweb.cgi?a=blobdiff_plain;ds=inline;f=main.py;h=be0d8e0a0871b2283f7bbecc5307b9a5a494262a;hb=90f27333118e72b35068d7d7ac29e7b14f27aa3b;hp=fd8ab4191e41b27199ce793c411e5cf346385c1b;hpb=51540cefc448684d5086297d23e9a1805da4d405;p=culture.git diff --git a/main.py b/main.py index fd8ab41..be0d8e0 100755 --- a/main.py +++ b/main.py @@ -13,7 +13,7 @@ from torch.nn import functional as F import ffutils import mygpt -import sky, wireworld, quizz_machine +import sky, reasoning, quizz_machine # world quizzes vs. culture quizzes @@ -79,24 +79,26 @@ parser.add_argument("--dropout", type=float, default=0.1) parser.add_argument("--deterministic_synthesis", action="store_true", default=False) -parser.add_argument("--reverse_cleanup", action="store_true", default=False) - parser.add_argument("--problem", type=str, default="sky") parser.add_argument("--nb_gpts", type=int, default=5) -parser.add_argument("--nb_models_for_generation", type=int, default=1) +parser.add_argument("--min_to_validate", type=int, default=None) -parser.add_argument("--generation_mode", type=str, default="groupthink") +parser.add_argument("--max_to_validate", type=int, default=None) -parser.add_argument("--min_to_validate", type=int, default=4) +parser.add_argument("--accuracy_to_make_c_quizzes", type=float, default=0.975) -parser.add_argument("--max_to_validate", type=int, default=4) +parser.add_argument("--generation_temperature", type=float, default=2.0) -parser.add_argument("--accuracy_to_make_c_quizzes", type=float, default=0.975) +parser.add_argument("--deterministic_validation", action="store_true", default=False) + +parser.add_argument("--bidirectional_validation", action="store_true", default=False) parser.add_argument("--dirty_debug", action="store_true", default=False) +###################################################################### + parser.add_argument("--sky_height", type=int, default=6) parser.add_argument("--sky_width", type=int, default=8) @@ -111,6 +113,12 @@ parser.add_argument("--sky_speed", type=int, default=3) args = parser.parse_args() +if args.min_to_validate is None: + args.min_to_validate = args.nb_gpts - 1 + +if args.max_to_validate is None: + args.max_to_validate = args.nb_gpts - 1 + if args.result_dir is None: args.result_dir = f"results_culture" @@ -241,8 +249,10 @@ if args.problem == "sky": nb_iterations=args.sky_nb_iterations, speed=args.sky_speed, ) -elif args.problem == "wireworld": - problem = wireworld.Wireworld(height=8, width=10, nb_iterations=2, speed=5) + back_accuracy = False +elif args.problem == "reasoning": + problem = reasoning.Reasoning(device=device) + back_accuracy = True else: raise ValueError @@ -250,6 +260,7 @@ quizz_machine = quizz_machine.QuizzMachine( problem=problem, nb_train_samples=args.nb_train_samples, nb_test_samples=args.nb_test_samples, + back_accuracy=back_accuracy, batch_size=args.physical_batch_size, result_dir=args.result_dir, logger=log_string, @@ -366,18 +377,24 @@ def run_tests(model, quizz_machine, deterministic_synthesis): nb_test_samples += input.size(0) - main_test_accuracy = quizz_machine.produce_results( + test_perplexity = math.exp(min(100, acc_test_loss / nb_test_samples)) + + log_string(f"test_perplexity {n_epoch} {test_perplexity}") + + model.main_test_accuracy = quizz_machine.produce_results( n_epoch=n_epoch, model=model, result_dir=args.result_dir, deterministic_synthesis=deterministic_synthesis, ) - test_perplexity = math.exp(min(100, acc_test_loss / nb_test_samples)) - log_string(f"test_perplexity {n_epoch} {test_perplexity}") +###################################################################### + - model.main_test_accuracy = main_test_accuracy +def valid_c_quizzes(recorded, criteria): + result = [q[criteria(c)] for q, c in recorded] + return torch.cat(result, dim=0) if len(result) > 0 else torch.tensor([]) ###################################################################### @@ -388,95 +405,80 @@ def create_c_quizzes( quizz_machine, nb_for_train=1000, nb_for_test=100, - min_ave_seq_logproba=None, ): - # We will store the generated quizzes for each number of - # correct prediction - recorded = dict([(n, []) for n in range(len(models) + 1)]) - - model_indexes = [] - sum_logits, sum_nb_c_quizzes = 0, 0 - - def nb_generated(): - return sum([sum([x.size(0) for x in recorded[n]]) for n in recorded.keys()]) - - def nb_validated(): - return sum( - [ - sum([x.size(0) for x in recorded[n]]) - for n in range(args.min_to_validate, args.max_to_validate + 1) - ] - ) + recorded = [] nb_to_create = nb_for_train + nb_for_test - while nb_validated() < nb_to_create: - ( - new_c_quizzes, - nb_correct, - ave_seq_logproba, - ) = quizz_machine.gang_create_c_quizzes( - nb=nb_to_create, - nb_models_for_generation=args.nb_models_for_generation, - models=models, - mode=args.generation_mode, - reverse_cleanup=args.reverse_cleanup, - min_ave_seq_logproba=min_ave_seq_logproba, - n_epoch=n_epoch, - result_dir=args.result_dir, - ) + # ------------------------------------------------------------ + + standard_validity = lambda nb_correct: torch.logical_and( + nb_correct >= args.min_to_validate, nb_correct <= args.max_to_validate + ) + + file_name = os.path.join(args.result_dir, f"culture_c_quiz_{n_epoch:04d}_logp.dat") + with open(file_name, "w") as logp_file: + while valid_c_quizzes(recorded, standard_validity).size(0) < nb_to_create: + # Select a model at random to generate the new quizzes + + model_for_generation = models[torch.randint(len(models), (1,))] - sum_logits += new_c_quizzes.size(0) * ave_seq_logproba - sum_nb_c_quizzes += new_c_quizzes.size(0) + c_quizzes = quizz_machine.generate_quizzes( + nb_to_create, + model_for_generation=model_for_generation, + temperature=args.generation_temperature, + ) - if args.dirty_debug: - nb_correct = torch.randint( - len(models) + 1, nb_correct.size(), device=new_c_quizzes.device + nb_correct, seq_logproba = quizz_machine.compute_correctness( + c_quizzes, + models, + bidirectional_validation=args.bidirectional_validation, + deterministic_validation=args.deterministic_validation, ) - for n in range(nb_correct.max() + 1): - recorded[n].append(new_c_quizzes[nb_correct == n].clone()) + for n, l in zip(nb_correct, seq_logproba): + s = " ".join([str(x.item()) for x in l]) + logp_file.write(f"{n} {s}\n") - nv = [recorded[n][-1].size(0) for n in recorded.keys()] + if args.dirty_debug: + nb_correct = torch.randint( + len(models) + 1, nb_correct.size(), device=c_quizzes.device + ) - log_string(f"keep c_quizzes kept {nv} total {nb_validated()} / {nb_to_create}") + recorded.append((c_quizzes, nb_correct)) - # concatenate and shuffle - for n in recorded.keys(): - if len(recorded[n]) > 0: - q = torch.cat(recorded[n], dim=0) - q = q[torch.randperm(q.size(0), device=q.device)] - recorded[n] = q - else: - del recorded[n] + nv = F.one_hot(nb_correct, num_classes=len(models) + 1).sum(0) + nv = " ".join([str(x.item()) for x in nv]) - new_c_quizzes = torch.cat( - [recorded[n] for n in range(args.min_to_validate, args.max_to_validate + 1)], - dim=0, - ) + nb_validated = valid_c_quizzes(recorded, standard_validity).size(0) - new_c_quizzes = new_c_quizzes[ - torch.randperm(new_c_quizzes.size(0), device=new_c_quizzes.device)[ - : nb_for_train + nb_for_test - ] - ] + log_string( + f"keep c_quizzes model {model_for_generation.id} kept {nv} nb_accumulated {nb_validated} / {nb_to_create}" + ) + + # store the new c_quizzes which have been validated + + new_c_quizzes = valid_c_quizzes(recorded, standard_validity) quizz_machine.store_c_quizzes(new_c_quizzes[:nb_for_train], for_train=True) quizz_machine.store_c_quizzes(new_c_quizzes[nb_for_train:], for_train=False) - for n in recorded.keys(): + # save a bunch of images to investigate what quizzes with a + # certain nb of correct predictions look like + + for n in range(len(models) + 1): s = ( "_validated" if n >= args.min_to_validate and n <= args.max_to_validate else "" ) - quizz_machine.problem.save_quizzes( - recorded[n][:72], - args.result_dir, - f"culture_c_quiz_{n_epoch:04d}_N{n}{s}", - ) - return sum_logits / sum_nb_c_quizzes + q = valid_c_quizzes(recorded, criteria=lambda nb_correct: nb_correct == n)[:72] + + if q.size(0) > 0: + quizz_machine.save_quizzes( + args.result_dir, f"culture_c_quiz_{n_epoch:04d}_N{n}{s}", q + ) ###################################################################### @@ -506,58 +508,47 @@ log_string(f"nb_parameters {nb_parameters} ({int(nb_parameters/1e6)}M)") ###################################################################### -min_ave_seq_logproba = None - for n_epoch in range(args.nb_epochs): log_string(f"--- epoch {n_epoch} ----------------------------------------") - a = [(model.id, float(model.main_test_accuracy)) for model in models] - a.sort(key=lambda p: p[0]) - s = " ".join([f"{p[1]*100:.02f}%" for p in a]) - log_string(f"current accuracies {s}") + # Select, improve, and eval the worst model - # select the model with lowest accuracy - models.sort(key=lambda model: model.main_test_accuracy) - model = models[0] + weakest_model = min(models, key=lambda m: float(m.main_test_accuracy)) log_string( - f"training model {model.id} main_test_accuracy {model.main_test_accuracy}" + f"training model {weakest_model.id} main_test_accuracy {weakest_model.main_test_accuracy}" ) - # improve it - one_epoch(model, quizz_machine) - - quizz_machine.renew_w_quizzes(args.nb_train_samples // args.nb_gpts) + one_epoch(weakest_model, quizz_machine) log_string( f"train_set_composition w_quizzes {quizz_machine.nb_batch_w_quizzes} c_quizzes {quizz_machine.nb_batch_c_quizzes}" ) - # test it - run_tests(model, quizz_machine, deterministic_synthesis=False) + run_tests(weakest_model, quizz_machine, deterministic_synthesis=False) log_string( f"test_set_composition w_quizzes {quizz_machine.nb_batch_w_quizzes} c_quizzes {quizz_machine.nb_batch_c_quizzes}" ) + cta = " ".join([f"{float(m.main_test_accuracy):.04f}" for m in models]) + log_string(f"current_test_accuracies {cta}") + + # Replace a fraction of the w_quizzes with fresh ones + + quizz_machine.renew_w_quizzes(args.nb_train_samples // args.nb_gpts) + + # If all the models are good enough, generate new quizzes and + # re-compute the test errors + if min([m.main_test_accuracy for m in models]) >= args.accuracy_to_make_c_quizzes: - ave_seq_logproba = create_c_quizzes( + create_c_quizzes( models, quizz_machine, nb_for_train=nb_new_c_quizzes_for_train, nb_for_test=nb_new_c_quizzes_for_test, - min_ave_seq_logproba=min_ave_seq_logproba, ) - # We keep the first average logits as a reference - # if min_ave_seq_logproba is None: - # min_ave_seq_logproba = ave_seq_logproba - # else: - # log_string( - # f"min_ave_seq_logproba {min_ave_seq_logproba} ave_seq_logproba {ave_seq_logproba}" - # ) - - # We update everyone for model in models: run_tests(model, quizz_machine, deterministic_synthesis=False)