def eval(rank, args, shared_model, best_eval_acc=0, checkpoint=None, epoch=0): print('Evaluating at {} epoch, with Acc {} ##################'.format( epoch, best_eval_acc)) torch.cuda.set_device(args.gpus.index(args.gpus[rank % len(args.gpus)])) if args.input_type == 'ques': model_kwargs = {'vocab': load_vocab(args.vocab_json)} model = VqaLstmModel(**model_kwargs) elif args.input_type == 'ques,image': model_kwargs = {'vocab': load_vocab(args.vocab_json)} model = VqaLstmCnnAttentionModel(**model_kwargs) lossFn = torch.nn.CrossEntropyLoss().cuda() eval_loader_kwargs = { 'questions_h5': getattr(args, args.eval_split + '_h5'), 'data_json': args.data_json, 'vocab': args.vocab_json, 'batch_size': 1, 'input_type': args.input_type, 'num_frames': args.num_frames, 'split': args.eval_split, 'max_threads_per_gpu': args.max_threads_per_gpu, 'gpu_id': args.gpus[rank % len(args.gpus)], 'to_cache': args.to_cache } eval_loader = EqaDataLoader(**eval_loader_kwargs) print('eval_loader has %d samples' % len(eval_loader.dataset)) args.output_log_path = os.path.join(args.log_dir, 'eval_' + str(rank) + '.json') # t, best_eval_acc = 0, 0, 0 t = 0 mean_rank = [] model.load_state_dict(shared_model.state_dict()) model.eval() metrics = VqaMetric( info={'split': args.eval_split}, metric_names=['loss', 'accuracy', 'mean_rank', 'mean_reciprocal_rank'], log_json=args.output_log_path) if args.input_type == 'ques': for batch in eval_loader: t += 1 model.cuda() idx, questions, answers = batch questions_var = Variable(questions.cuda()) answers_var = Variable(answers.cuda()) scores = model(questions_var) loss = lossFn(scores, answers_var) # update metrics accuracy, ranks = metrics.compute_ranks(scores.data.cpu(), answers) metrics.update([loss.data[0], accuracy, ranks, 1.0 / ranks]) print(metrics.get_stat_string(mode=0)) elif args.input_type == 'ques,image': done = False all_envs_loaded = eval_loader.dataset._check_if_all_envs_loaded() while done == False: mean_rank = [] for num, batch in enumerate(eval_loader): # if num>0: # done= True # break t += 1 model.cuda() idx, questions, answers, images, _, _, _ = batch questions_var = Variable(questions.cuda()) answers_var = Variable(answers.cuda()) images_var = Variable(images.cuda()) images_numpy = images_var.data.cpu().numpy() question_numpy = questions_var.data.cpu().numpy() answers_numpy = answers_var.data.cpu().numpy() scores, att_probs = model(images_var, questions_var) scores_numpy = scores.data.cpu().numpy() att_probs_numpy = att_probs.data.cpu().numpy() loss = lossFn(scores, answers_var) # update metrics accuracy, ranks = metrics.compute_ranks( scores.data.cpu(), answers) mean_rank.extend(ranks) print("Batch Mean Ranks", sum(ranks) / len(ranks)) metrics.update([loss.item(), accuracy, ranks, 1.0 / ranks]) print(metrics.get_stat_string(mode=0)) print("Mean Rank for eval", sum(mean_rank) / len(mean_rank)) if all_envs_loaded == False: eval_loader.dataset._load_envs() if len(eval_loader.dataset.pruned_env_set) == 0: done = True else: done = True # checkpoint if best val accuracy if metrics.metrics[1][0] >= best_eval_acc: best_eval_acc = metrics.metrics[1][0] metrics.dump_log() model_state = get_state(model) if args.checkpoint_path != False and checkpoint is not None: ad = checkpoint['args'] else: ad = args.__dict__ checkpoint = {'args': ad, 'state': model_state, 'epoch': epoch} checkpoint_path = '%s/epoch_%d_accuracy_%d.pt' % ( args.checkpoint_dir, epoch, int(best_eval_acc * 100)) print('Saving checkpoint to %s' % checkpoint_path) torch.save(checkpoint, checkpoint_path) print('[best_eval_accuracy:%.04f]' % best_eval_acc) print("Mean Rank for eval", sum(mean_rank) / len(mean_rank)) return best_eval_acc
def train(rank, args, shared_nav_model, shared_ans_model): torch.cuda.set_device(args.gpus.index(args.gpus[rank % len(args.gpus)])) if args.model_type == 'pacman': model_kwargs = {'question_vocab': load_vocab(args.vocab_json)} nav_model = NavPlannerControllerModel(**model_kwargs) else: exit() model_kwargs = {'vocab': load_vocab(args.vocab_json)} ans_model = VqaLstmCnnAttentionModel(**model_kwargs) optim = torch.optim.SGD(filter(lambda p: p.requires_grad, shared_nav_model.parameters()), lr=args.learning_rate) train_loader_kwargs = { 'questions_h5': args.train_h5, 'data_json': args.data_json, 'vocab': args.vocab_json, 'target_obj_conn_map_dir': args.target_obj_conn_map_dir, 'map_resolution': args.map_resolution, 'batch_size': 1, 'input_type': args.model_type, 'num_frames': 5, 'split': 'train', 'max_threads_per_gpu': args.max_threads_per_gpu, 'gpu_id': args.gpus[rank % len(args.gpus)], 'to_cache': args.to_cache } args.output_nav_log_path = os.path.join(args.log_dir, 'nav_train_' + str(rank) + '.json') args.output_ans_log_path = os.path.join(args.log_dir, 'ans_train_' + str(rank) + '.json') nav_model.load_state_dict(shared_nav_model.state_dict()) nav_model.cuda() ans_model.load_state_dict(shared_ans_model.state_dict()) ans_model.eval() ans_model.cuda() nav_metrics = NavMetric(info={ 'split': 'train', 'thread': rank }, metric_names=[ 'planner_loss', 'controller_loss', 'reward', 'episode_length' ], log_json=args.output_nav_log_path) vqa_metrics = VqaMetric( info={ 'split': 'train', 'thread': rank }, metric_names=['accuracy', 'mean_rank', 'mean_reciprocal_rank'], log_json=args.output_ans_log_path) train_loader = EqaDataLoader(**train_loader_kwargs) print('train_loader has %d samples' % len(train_loader.dataset)) t, epoch = 0, 0 p_losses, c_losses, reward_list, episode_length_list = [], [], [], [] nav_metrics.update([10.0, 10.0, 0, 100]) mult = 0.1 while epoch < int(args.max_epochs): if 'pacman' in args.model_type: planner_lossFn = MaskedNLLCriterion().cuda() controller_lossFn = MaskedNLLCriterion().cuda() done = False all_envs_loaded = train_loader.dataset._check_if_all_envs_loaded() while done == False: for batch in train_loader: nav_model.load_state_dict(shared_nav_model.state_dict()) nav_model.eval() nav_model.cuda() idx, question, answer, actions, action_length = batch metrics_slug = {} h3d = train_loader.dataset.episode_house # evaluate at multiple initializations # for i in [10, 30, 50]: t += 1 question_var = Variable(question.cuda()) controller_step = False planner_hidden = nav_model.planner_nav_rnn.init_hidden(1) # forward through planner till spawn planner_actions_in, planner_img_feats, controller_step, controller_action_in, controller_img_feat, init_pos = train_loader.dataset.get_hierarchical_features_till_spawn( actions[0, :action_length[0] + 1].numpy(), max(3, int(mult * action_length[0]))) planner_actions_in_var = Variable( planner_actions_in.cuda()) planner_img_feats_var = Variable(planner_img_feats.cuda()) for step in range(planner_actions_in.size(0)): planner_scores, planner_hidden = nav_model.planner_step( question_var, planner_img_feats_var[step].view(1, 1, 3200), planner_actions_in_var[step].view(1, 1), planner_hidden) if controller_step == True: controller_img_feat_var = Variable( controller_img_feat.cuda()) controller_action_in_var = Variable( torch.LongTensor(1, 1).fill_( int(controller_action_in)).cuda()) controller_scores = nav_model.controller_step( controller_img_feat_var.view(1, 1, 3200), controller_action_in_var.view(1, 1), planner_hidden[0]) prob = F.softmax(controller_scores, dim=1) controller_action = int( prob.max(1)[1].data.cpu().numpy()[0]) if controller_action == 1: controller_step = True else: controller_step = False action = int(controller_action_in) action_in = torch.LongTensor(1, 1).fill_(action + 1).cuda() else: prob = F.softmax(planner_scores, dim=1) action = int(prob.max(1)[1].data.cpu().numpy()[0]) action_in = torch.LongTensor(1, 1).fill_(action + 1).cuda() h3d.env.reset(x=init_pos[0], y=init_pos[2], yaw=init_pos[3]) init_dist_to_target = h3d.get_dist_to_target( h3d.env.cam.pos) if init_dist_to_target < 0: # unreachable # invalids.append([idx[0], i]) continue episode_length = 0 episode_done = True controller_action_counter = 0 dists_to_target, pos_queue = [init_dist_to_target ], [init_pos] rewards, planner_actions, planner_log_probs, controller_actions, controller_log_probs = [], [], [], [], [] if action != 3: # take the first step img, rwd, episode_done = h3d.step(action, step_reward=True) img = torch.from_numpy(img.transpose( 2, 0, 1)).float() / 255.0 img_feat_var = train_loader.dataset.cnn( Variable(img.view(1, 3, 224, 224).cuda())).view(1, 1, 3200) for step in range(args.max_episode_length): episode_length += 1 if controller_step == False: planner_scores, planner_hidden = nav_model.planner_step( question_var, img_feat_var, Variable(action_in), planner_hidden) planner_prob = F.softmax(planner_scores, dim=1) planner_log_prob = F.log_softmax( planner_scores, dim=1) action = planner_prob.multinomial().data planner_log_prob = planner_log_prob.gather( 1, Variable(action)) planner_log_probs.append( planner_log_prob.cpu()) action = int(action.cpu().numpy()[0, 0]) planner_actions.append(action) img, rwd, episode_done = h3d.step(action, step_reward=True) episode_done = episode_done or episode_length >= args.max_episode_length rewards.append(rwd) img = torch.from_numpy(img.transpose( 2, 0, 1)).float() / 255.0 img_feat_var = train_loader.dataset.cnn( Variable(img.view(1, 3, 224, 224).cuda())).view( 1, 1, 3200) dists_to_target.append( h3d.get_dist_to_target(h3d.env.cam.pos)) pos_queue.append([ h3d.env.cam.pos.x, h3d.env.cam.pos.y, h3d.env.cam.pos.z, h3d.env.cam.yaw ]) if episode_done == True: break # query controller to continue or not controller_action_in = Variable( torch.LongTensor(1, 1).fill_(action).cuda()) controller_scores = nav_model.controller_step( img_feat_var, controller_action_in, planner_hidden[0]) controller_prob = F.softmax(controller_scores, dim=1) controller_log_prob = F.log_softmax( controller_scores, dim=1) controller_action = controller_prob.multinomial( ).data if int(controller_action[0] ) == 1 and controller_action_counter < 4: controller_action_counter += 1 controller_step = True else: controller_action_counter = 0 controller_step = False controller_action.fill_(0) controller_log_prob = controller_log_prob.gather( 1, Variable(controller_action)) controller_log_probs.append( controller_log_prob.cpu()) controller_action = int( controller_action.cpu().numpy()[0, 0]) controller_actions.append(controller_action) action_in = torch.LongTensor(1, 1).fill_(action + 1).cuda() # run answerer here ans_acc = [0] if action == 3: if len(pos_queue) < 5: pos_queue = train_loader.dataset.episode_pos_queue[ len(pos_queue) - 5:] + pos_queue images = train_loader.dataset.get_frames( h3d, pos_queue[-5:], preprocess=True) images_var = Variable( torch.from_numpy(images).cuda()).view( 1, 5, 3, 224, 224) scores, att_probs = ans_model(images_var, question_var) ans_acc, ans_rank = vqa_metrics.compute_ranks( scores.data.cpu(), answer) vqa_metrics.update([ans_acc, ans_rank, 1.0 / ans_rank]) rewards.append(h3d.success_reward * ans_acc[0]) R = torch.zeros(1, 1) planner_loss = 0 controller_loss = 0 planner_rev_idx = -1 for i in reversed(range(len(rewards))): R = 0.99 * R + rewards[i] advantage = R - nav_metrics.metrics[2][1] if i < len(controller_actions): controller_loss = controller_loss - controller_log_probs[ i] * Variable(advantage) if controller_actions[ i] == 0 and planner_rev_idx + len( planner_log_probs) >= 0: planner_loss = planner_loss - planner_log_probs[ planner_rev_idx] * Variable(advantage) planner_rev_idx -= 1 elif planner_rev_idx + len(planner_log_probs) >= 0: planner_loss = planner_loss - planner_log_probs[ planner_rev_idx] * Variable(advantage) planner_rev_idx -= 1 controller_loss /= max(1, len(controller_log_probs)) planner_loss /= max(1, len(planner_log_probs)) optim.zero_grad() if isinstance(planner_loss, float) == False and isinstance( controller_loss, float) == False: p_losses.append(planner_loss.data[0, 0]) c_losses.append(controller_loss.data[0, 0]) reward_list.append(np.sum(rewards)) episode_length_list.append(episode_length) (planner_loss + controller_loss).backward() ensure_shared_grads(nav_model.cpu(), shared_nav_model) optim.step() if len(reward_list) > 50: nav_metrics.update([ p_losses, c_losses, reward_list, episode_length_list ]) print(nav_metrics.get_stat_string()) if args.to_log == 1: nav_metrics.dump_log() if nav_metrics.metrics[2][1] > 0.35: mult = min(mult + 0.1, 1.0) p_losses, c_losses, reward_list, episode_length_list = [], [], [], [] if all_envs_loaded == False: train_loader.dataset._load_envs(in_order=True) if len(train_loader.dataset.pruned_env_set) == 0: done = True if args.to_cache == False: train_loader.dataset._load_envs(start_idx=0, in_order=True) else: done = True epoch += 1
def train(rank, args, shared_model): torch.cuda.set_device(args.gpus.index(args.gpus[rank % len(args.gpus)])) if args.input_type == 'ques': model_kwargs = {'vocab': load_vocab(args.vocab_json)} model = VqaLstmModel(**model_kwargs) elif args.input_type == 'ques,image': model_kwargs = {'vocab': load_vocab(args.vocab_json)} model = VqaLstmCnnAttentionModel(**model_kwargs) lossFn = torch.nn.CrossEntropyLoss().cuda() optim = torch.optim.Adam(filter(lambda p: p.requires_grad, shared_model.parameters()), lr=args.learning_rate) train_loader_kwargs = { 'questions_h5': args.train_h5, 'data_json': args.data_json, 'vocab': args.vocab_json, 'batch_size': args.batch_size, 'input_type': args.input_type, 'num_frames': args.num_frames, 'split': 'train', 'max_threads_per_gpu': args.max_threads_per_gpu, 'gpu_id': args.gpus[rank % len(args.gpus)], 'to_cache': args.to_cache } eval_loader_kwargs = { 'questions_h5': args.val_h5, 'data_json': args.data_json, 'vocab': args.vocab_json, 'batch_size': 1, 'input_type': args.input_type, 'num_frames': args.num_frames, 'split': 'val', 'max_threads_per_gpu': args.max_threads_per_gpu, 'gpu_id': args.gpus[rank % len(args.gpus)], 'to_cache': args.to_cache } args.output_log_path = os.path.join(args.log_dir, 'train_' + str(rank) + '.json') eval_output_log_path = os.path.join(args.log_dir, 'eval_' + str(rank) + '.json') metrics = VqaMetric( info={ 'split': 'train', 'thread': rank }, metric_names=['loss', 'accuracy', 'mean_rank', 'mean_reciprocal_rank'], log_json=args.output_log_path) eval_metrics = VqaMetric( info={'split': 'eval'}, metric_names=['loss', 'accuracy', 'mean_rank', 'mean_reciprocal_rank'], log_json=eval_output_log_path) train_loader = EqaDataLoader(**train_loader_kwargs) eval_loader = EqaDataLoader(**eval_loader_kwargs) if args.input_type == 'ques,image': train_loader.dataset._load_envs(start_idx=0, in_order=True) print('train_loader has %d samples' % len(train_loader.dataset)) print('eval_loader has %d samples' % len(eval_loader.dataset)) t, epoch = 0, 0 best_eval_acc = 0 mean_rank = [] while epoch < int(args.max_epochs): if args.input_type == 'ques': for batch in train_loader: t += 1 model.load_state_dict(shared_model.state_dict()) model.train() model.cuda() idx, questions, answers = batch questions_var = Variable(questions.cuda()) answers_var = Variable(answers.cuda()) scores = model(questions_var) loss = lossFn(scores, answers_var) # zero grad optim.zero_grad() # update metrics accuracy, ranks = metrics.compute_ranks( scores.data.cpu(), answers) metrics.update([loss.data[0], accuracy, ranks, 1.0 / ranks]) # backprop and update loss.backward() ensure_shared_grads(model.cpu(), shared_model) optim.step() if t % args.print_every == 0: print(metrics.get_stat_string()) if args.to_log == 1: metrics.dump_log() elif args.input_type == 'ques,image': t += 1 #TRAIN model.train() model.cuda() done = False all_envs_loaded = train_loader.dataset._check_if_all_envs_loaded() #p = 0 while done == False: #print("Here now: ", epoch, p) #p+=1 for num, batch in enumerate(train_loader): #pp=0 # done = True # break model.load_state_dict(shared_model.state_dict()) model.cuda() idx, questions, answers, images, _, _, _ = batch questions_var = Variable(questions.cuda()) answers_var = Variable(answers.cuda()) images_var = Variable(images.cuda()) scores, att_probs = model(images_var, questions_var) loss = lossFn(scores, answers_var) #zero grad optim.zero_grad() # update metrics accuracy, ranks = metrics.compute_ranks( scores.data.cpu(), answers) metrics.update([loss.item(), accuracy, ranks, 1.0 / ranks]) # backprop and update loss.backward() ensure_shared_grads(model.cpu(), shared_model) optim.step() if t % args.print_every == 0: print(metrics.get_stat_string()) if args.to_log == 1: metrics.dump_log() if all_envs_loaded == False: train_loader.dataset._load_envs(in_order=True) if len(train_loader.dataset.pruned_env_set) == 0: done = True # SATYEN: if args.to_cache == False: train_loader.dataset._load_envs(start_idx=0, in_order=True) else: done = True if epoch % args.eval_every == 0: best_eval_acc = eval(0, args, model, best_eval_acc=best_eval_acc, epoch=epoch) epoch += 1
def eval(rank, args, shared_model): print('eval start...') torch.cuda.set_device(args.gpus.index(args.gpus[rank % len(args.gpus)])) if args.input_type == 'ques': model_kwargs = {'vocab': load_vocab(args.vocab_json)} model = VqaLstmModel(**model_kwargs) elif args.input_type == 'ques,image': model_kwargs = {'vocab': load_vocab(args.vocab_json)} model = VqaLstmCnnAttentionModel(**model_kwargs) lossFn = torch.nn.CrossEntropyLoss().cuda() eval_loader_kwargs = { 'questions_h5': getattr(args, args.eval_split + '_h5'), 'vocab': args.vocab_json, 'batch_size': 1, 'input_type': args.input_type, 'num_frames': args.num_frames, 'split': args.eval_split, 'max_threads_per_gpu': args.max_threads_per_gpu, 'gpu_id': args.gpus[rank % len(args.gpus)], } # print(eval_loader_kwargs) eval_loader = EqaDataLoader(**eval_loader_kwargs) print('eval_loader has %d samples' % len(eval_loader.dataset)) args.output_log_path = os.path.join(args.log_dir, 'eval_' + str(rank) + '.json') t, epoch, best_eval_acc = 0, 0, 0 while epoch < int(args.max_epochs): model.load_state_dict(shared_model.state_dict()) model.eval() metrics = VqaMetric(info={'split': args.eval_split}, metric_names=[ 'loss', 'accuracy', 'mean_rank', 'mean_reciprocal_rank' ], log_json=args.output_log_path) if args.input_type == 'ques': for batch in eval_loader: t += 1 model.cuda() idx, questions, answers = batch questions_var = Variable(questions.cuda()) answers_var = Variable(answers.cuda()) scores = model(questions_var) loss = lossFn(scores, answers_var) print(scores) # update metrics accuracy, ranks = metrics.compute_ranks( scores.data.cpu(), answers) metrics.update([loss.data[0], accuracy, ranks, 1.0 / ranks]) print(metrics.get_stat_string(mode=0)) elif args.input_type == 'ques,image': done = False all_envs_loaded = True #all_envs_loaded = eval_loader.dataset._check_if_all_envs_loaded() while done == False: for batch in eval_loader: t += 1 model.cuda() idx, questions, answers, images, _, _, _ = batch questions_var = Variable(questions.cuda()) answers_var = Variable(answers.cuda()) images_var = Variable(images.cuda()) scores, att_probs = model(images_var, questions_var) loss = lossFn(scores, answers_var) # update metrics accuracy, ranks = metrics.compute_ranks( scores.data.cpu(), answers) metrics.update( [loss.data[0], accuracy, ranks, 1.0 / ranks]) print(metrics.get_stat_string(mode=0)) if all_envs_loaded == False: eval_loader.dataset._load_envs() if len(eval_loader.dataset.pruned_env_set) == 0: done = True else: done = True epoch += 1 # checkpoint if best val accuracy if metrics.metrics[1][0] > best_eval_acc: best_eval_acc = metrics.metrics[1][0] if epoch % args.eval_every == 0 and args.log == True: metrics.dump_log() model_state = get_state(model) if args.checkpoint_path != False: ad = checkpoint['args'] else: ad = args.__dict__ checkpoint = {'args': ad, 'state': model_state, 'epoch': epoch} checkpoint_path = '%s/epoch_%d_accuracy_%.04f.pt' % ( args.checkpoint_dir, epoch, best_eval_acc) print('Saving checkpoint to %s' % checkpoint_path) torch.save(checkpoint, checkpoint_path) print('[best_eval_accuracy:%.04f]' % best_eval_acc)
def eval(rank, args, shared_nav_model, shared_ans_model): torch.cuda.set_device(args.gpus.index(args.gpus[rank % len(args.gpus)])) if args.model_type == 'pacman': model_kwargs = {'question_vocab': load_vocab(args.vocab_json)} nav_model = NavPlannerControllerModel(**model_kwargs) else: exit() model_kwargs = {'vocab': load_vocab(args.vocab_json)} ans_model = VqaLstmCnnAttentionModel(**model_kwargs) eval_loader_kwargs = { 'questions_h5': getattr(args, args.eval_split + '_h5'), 'data_json': args.data_json, 'vocab': args.vocab_json, 'target_obj_conn_map_dir': args.target_obj_conn_map_dir, 'map_resolution': args.map_resolution, 'batch_size': 1, 'input_type': args.model_type, 'num_frames': 5, 'split': args.eval_split, 'max_threads_per_gpu': args.max_threads_per_gpu, 'gpu_id': args.gpus[rank % len(args.gpus)], 'to_cache': False } eval_loader = EqaDataLoader(**eval_loader_kwargs) print('eval_loader has %d samples' % len(eval_loader.dataset)) args.output_nav_log_path = os.path.join(args.log_dir, 'nav_eval_' + str(rank) + '.json') args.output_ans_log_path = os.path.join(args.log_dir, 'ans_eval_' + str(rank) + '.json') t, epoch, best_eval_acc = 0, 0, 0.0 while epoch < int(args.max_epochs): start_time = time.time() invalids = [] nav_model.load_state_dict(shared_nav_model.state_dict()) nav_model.eval() ans_model.load_state_dict(shared_ans_model.state_dict()) ans_model.eval() ans_model.cuda() # that's a lot of numbers nav_metrics = NavMetric( info={ 'split': args.eval_split, 'thread': rank }, metric_names=[ 'd_0_10', 'd_0_30', 'd_0_50', 'd_T_10', 'd_T_30', 'd_T_50', 'd_D_10', 'd_D_30', 'd_D_50', 'd_min_10', 'd_min_30', 'd_min_50', 'r_T_10', 'r_T_30', 'r_T_50', 'r_e_10', 'r_e_30', 'r_e_50', 'stop_10', 'stop_30', 'stop_50', 'ep_len_10', 'ep_len_30', 'ep_len_50' ], log_json=args.output_nav_log_path) vqa_metrics = VqaMetric( info={ 'split': args.eval_split, 'thread': rank }, metric_names=[ 'accuracy_10', 'accuracy_30', 'accuracy_50', 'mean_rank_10', 'mean_rank_30', 'mean_rank_50', 'mean_reciprocal_rank_10', 'mean_reciprocal_rank_30', 'mean_reciprocal_rank_50' ], log_json=args.output_ans_log_path) if 'pacman' in args.model_type: done = False while done == False: for batch in tqdm(eval_loader): nav_model.load_state_dict(shared_nav_model.state_dict()) nav_model.eval() nav_model.cuda() idx, question, answer, actions, action_length = batch metrics_slug = {} h3d = eval_loader.dataset.episode_house # evaluate at multiple initializations for i in [10, 30, 50]: t += 1 if i > action_length[0]: invalids.append([idx[0], i]) continue question_var = Variable(question.cuda()) controller_step = False planner_hidden = nav_model.planner_nav_rnn.init_hidden( 1) # forward through planner till spawn planner_actions_in, planner_img_feats, controller_step, controller_action_in, controller_img_feat, init_pos = eval_loader.dataset.get_hierarchical_features_till_spawn( actions[0, :action_length[0] + 1].numpy(), i) planner_actions_in_var = Variable( planner_actions_in.cuda()) planner_img_feats_var = Variable( planner_img_feats.cuda()) for step in range(planner_actions_in.size(0)): planner_scores, planner_hidden = nav_model.planner_step( question_var, planner_img_feats_var[step].view(1, 1, 3200), planner_actions_in_var[step].view(1, 1), planner_hidden) if controller_step == True: controller_img_feat_var = Variable( controller_img_feat.cuda()) controller_action_in_var = Variable( torch.LongTensor(1, 1).fill_( int(controller_action_in)).cuda()) controller_scores = nav_model.controller_step( controller_img_feat_var.view(1, 1, 3200), controller_action_in_var.view(1, 1), planner_hidden[0]) prob = F.softmax(controller_scores, dim=1) controller_action = int( prob.max(1)[1].data.cpu().numpy()[0]) if controller_action == 1: controller_step = True else: controller_step = False action = int(controller_action_in) action_in = torch.LongTensor(1, 1).fill_(action + 1).cuda() else: prob = F.softmax(planner_scores, dim=1) action = int(prob.max(1)[1].data.cpu().numpy()[0]) action_in = torch.LongTensor(1, 1).fill_(action + 1).cuda() h3d.env.reset(x=init_pos[0], y=init_pos[2], yaw=init_pos[3]) init_dist_to_target = h3d.get_dist_to_target( h3d.env.cam.pos) if init_dist_to_target < 0: # unreachable invalids.append([idx[0], i]) continue episode_length = 0 episode_done = True controller_action_counter = 0 dists_to_target, pos_queue, pred_actions = [ init_dist_to_target ], [init_pos], [] planner_actions, controller_actions = [], [] if action != 3: # take the first step img, _, _ = h3d.step(action) img = torch.from_numpy(img.transpose( 2, 0, 1)).float() / 255.0 img_feat_var = eval_loader.dataset.cnn( Variable(img.view(1, 3, 224, 224).cuda())).view( 1, 1, 3200) for step in range(args.max_episode_length): episode_length += 1 if controller_step == False: planner_scores, planner_hidden = nav_model.planner_step( question_var, img_feat_var, Variable(action_in), planner_hidden) prob = F.softmax(planner_scores, dim=1) action = int( prob.max(1)[1].data.cpu().numpy()[0]) planner_actions.append(action) pred_actions.append(action) img, _, episode_done = h3d.step(action) episode_done = episode_done or episode_length >= args.max_episode_length img = torch.from_numpy(img.transpose( 2, 0, 1)).float() / 255.0 img_feat_var = eval_loader.dataset.cnn( Variable(img.view(1, 3, 224, 224).cuda())).view( 1, 1, 3200) dists_to_target.append( h3d.get_dist_to_target(h3d.env.cam.pos)) pos_queue.append([ h3d.env.cam.pos.x, h3d.env.cam.pos.y, h3d.env.cam.pos.z, h3d.env.cam.yaw ]) if episode_done == True: break # query controller to continue or not controller_action_in = Variable( torch.LongTensor(1, 1).fill_(action).cuda()) controller_scores = nav_model.controller_step( img_feat_var, controller_action_in, planner_hidden[0]) prob = F.softmax(controller_scores, dim=1) controller_action = int( prob.max(1)[1].data.cpu().numpy()[0]) if controller_action == 1 and controller_action_counter < 4: controller_action_counter += 1 controller_step = True else: controller_action_counter = 0 controller_step = False controller_action = 0 controller_actions.append(controller_action) action_in = torch.LongTensor( 1, 1).fill_(action + 1).cuda() # run answerer here if len(pos_queue) < 5: pos_queue = eval_loader.dataset.episode_pos_queue[ len(pos_queue) - 5:] + pos_queue images = eval_loader.dataset.get_frames( h3d, pos_queue[-5:], preprocess=True) images_var = Variable( torch.from_numpy(images).cuda()).view( 1, 5, 3, 224, 224) scores, att_probs = ans_model(images_var, question_var) ans_acc, ans_rank = vqa_metrics.compute_ranks( scores.data.cpu(), answer) pred_answer = scores.max(1)[1].data[0] print( '[Q_GT]', ' '.join([ eval_loader.dataset.vocab['questionIdxToToken'] [x] for x in question[0] if x != 0 ])) print( '[A_GT]', eval_loader.dataset.vocab['answerIdxToToken'][ answer[0]]) print( '[A_PRED]', eval_loader.dataset.vocab['answerIdxToToken'] [pred_answer]) # compute stats metrics_slug['accuracy_' + str(i)] = ans_acc[0] metrics_slug['mean_rank_' + str(i)] = ans_rank[0] metrics_slug['mean_reciprocal_rank_' + str(i)] = 1.0 / ans_rank[0] metrics_slug['d_0_' + str(i)] = dists_to_target[0] metrics_slug['d_T_' + str(i)] = dists_to_target[-1] metrics_slug[ 'd_D_' + str(i)] = dists_to_target[0] - dists_to_target[-1] metrics_slug['d_min_' + str(i)] = np.array(dists_to_target).min() metrics_slug['ep_len_' + str(i)] = episode_length if action == 3: metrics_slug['stop_' + str(i)] = 1 else: metrics_slug['stop_' + str(i)] = 0 inside_room = [] for p in pos_queue: inside_room.append( h3d.is_inside_room( p, eval_loader.dataset.target_room)) if inside_room[-1] == True: metrics_slug['r_T_' + str(i)] = 1 else: metrics_slug['r_T_' + str(i)] = 0 if any([x == True for x in inside_room]) == True: metrics_slug['r_e_' + str(i)] = 1 else: metrics_slug['r_e_' + str(i)] = 0 # navigation metrics metrics_list = [] for i in nav_metrics.metric_names: if i not in metrics_slug: metrics_list.append(nav_metrics.metrics[ nav_metrics.metric_names.index(i)][0]) else: metrics_list.append(metrics_slug[i]) nav_metrics.update(metrics_list) # vqa metrics metrics_list = [] for i in vqa_metrics.metric_names: if i not in metrics_slug: metrics_list.append(vqa_metrics.metrics[ vqa_metrics.metric_names.index(i)][0]) else: metrics_list.append(metrics_slug[i]) vqa_metrics.update(metrics_list) try: print(nav_metrics.get_stat_string(mode=0)) print(vqa_metrics.get_stat_string(mode=0)) except: pass print('epoch', epoch) print('invalids', len(invalids)) eval_loader.dataset._load_envs() if len(eval_loader.dataset.pruned_env_set) == 0: done = True epoch += 1 # checkpoint if best val accuracy if vqa_metrics.metrics[2][0] > best_eval_acc: # ans_acc_50 best_eval_acc = vqa_metrics.metrics[2][0] if epoch % args.eval_every == 0 and args.to_log == 1: vqa_metrics.dump_log() nav_metrics.dump_log() model_state = get_state(nav_model) aad = dict(args.__dict__) ad = {} for i in aad: if i[0] != '_': ad[i] = aad[i] checkpoint = {'args': ad, 'state': model_state, 'epoch': epoch} checkpoint_path = '%s/epoch_%d_ans_50_%.04f.pt' % ( args.checkpoint_dir, epoch, best_eval_acc) print('Saving checkpoint to %s' % checkpoint_path) torch.save(checkpoint, checkpoint_path) print('[best_eval_ans_acc_50:%.04f]' % best_eval_acc) eval_loader.dataset._load_envs(start_idx=0, in_order=True)
def eval(rank, args, shared_model, use_vision, use_language): gpu_idx = args.gpus.index(args.gpus[rank % len(args.gpus)]) torch.cuda.set_device(gpu_idx) print("eval gpu:" + str(gpu_idx) + " assigned") if args.input_type == 'ques': model_kwargs = {'vocab': load_vocab(args.vocab_json)} model = VqaLstmModel(**model_kwargs) elif args.input_type == 'ques,image': model_kwargs = {'vocab': load_vocab(args.vocab_json)} model = VqaLstmCnnAttentionModel(**model_kwargs) lossFn = torch.nn.CrossEntropyLoss().cuda() eval_loader_kwargs = { 'questions_h5': getattr(args, args.eval_split + '_h5'), 'data_json': args.data_json, 'vocab': args.vocab_json, 'batch_size': 1, 'input_type': args.input_type, 'num_frames': args.num_frames, 'split': args.eval_split, 'max_threads_per_gpu': args.max_threads_per_gpu, 'gpu_id': args.gpus[rank % len(args.gpus)], 'to_cache': args.to_cache } eval_loader = EqaDataLoader(**eval_loader_kwargs) print('eval_loader has %d samples' % len(eval_loader.dataset)) args.output_log_path = os.path.join(args.log_dir, 'eval_' + str(rank) + '.json') t, epoch, best_eval_acc = 0, 0, 0 print(epoch, args.max_epochs) # DEBUG while epoch < int(args.max_epochs): print("eval gpu:" + str(gpu_idx) + " running epoch " + str(epoch)) model.load_state_dict(shared_model.state_dict()) model.eval() metrics = VqaMetric(info={'split': args.eval_split}, metric_names=[ 'loss', 'accuracy', 'mean_rank', 'mean_reciprocal_rank' ], log_json=args.output_log_path) if args.input_type == 'ques': for batch in eval_loader: t += 1 model.cuda() idx, questions, answers = batch # If not using language, replace each question with a start and end token back to back. if not use_language: questions = torch.zeros_like(questions) questions.fill_( model_kwargs['vocab']['questionTokenToIdx']['<NULL>']) questions[:, 0] = model_kwargs['vocab'][ 'questionTokenToIdx']['<START>'] questions[:, 1] = model_kwargs['vocab'][ 'questionTokenToIdx']['<END>'] questions_var = Variable(questions.cuda()) answers_var = Variable(answers.cuda()) scores = model(questions_var) loss = lossFn(scores, answers_var) # update metrics accuracy, ranks = metrics.compute_ranks( scores.data.cpu(), answers) metrics.update([loss.data[0], accuracy, ranks, 1.0 / ranks]) print(metrics.get_stat_string(mode=0)) elif args.input_type == 'ques,image': done = False all_envs_loaded = eval_loader.dataset._check_if_all_envs_loaded() while done == False: for batch in eval_loader: t += 1 model.cuda() idx, questions, answers, images, _, _, _ = batch # If not using language, replace each question with a start and end token back to back. if not use_language: questions = torch.zeros_like(questions) questions.fill_(model_kwargs['vocab'] ['questionTokenToIdx']['<NULL>']) questions[:, 0] = model_kwargs['vocab'][ 'questionTokenToIdx']['<START>'] questions[:, 1] = model_kwargs['vocab'][ 'questionTokenToIdx']['<END>'] # If not using vision, replace all image feature data with zeros. if not use_vision: images = torch.zeros_like(images) questions_var = Variable(questions.cuda()) answers_var = Variable(answers.cuda()) images_var = Variable(images.cuda()) scores, att_probs = model(images_var, questions_var) loss = lossFn(scores, answers_var) # update metrics accuracy, ranks = metrics.compute_ranks( scores.data.cpu(), answers) metrics.update( [loss.data[0], accuracy, ranks, 1.0 / ranks]) print(metrics.get_stat_string(mode=0)) if all_envs_loaded == False: eval_loader.dataset._load_envs() if len(eval_loader.dataset.pruned_env_set) == 0: done = True else: done = True read_epoch = None while read_epoch is None or epoch >= read_epoch: try: with open(args.identifier + '.shared_epoch.tmp', 'r') as f: read_epoch = int(f.read().strip()) except (IOError, ValueError): pass if read_epoch is None: # TODO: since merger, this no longer works (hanging); might need to undo changes re: threading that we # TODO: made or debug them. print("eval gpu:" + str(gpu_idx) + " waiting for train thread to finish epoch " + str(epoch)) time.sleep( 10 ) # sleep until the training thread finishes another iteration epoch = read_epoch # checkpoint if best val accuracy if metrics.metrics[1][0] > best_eval_acc: best_eval_acc = metrics.metrics[1][0] if epoch % args.eval_every == 0 and args.to_log == 1: metrics.dump_log() model_state = get_state(model) if args.checkpoint_path != False: ad = checkpoint['args'] else: ad = args.__dict__ checkpoint = {'args': ad, 'state': model_state, 'epoch': epoch} checkpoint_path = '%s/epoch_%d_accuracy_%.04f.pt' % ( args.checkpoint_dir, epoch, best_eval_acc) print('Saving checkpoint to %s' % checkpoint_path) torch.save(checkpoint, checkpoint_path) print('[best_eval_accuracy:%.04f]' % best_eval_acc)
def train(rank, args, shared_model): torch.cuda.set_device(args.gpus.index(args.gpus[rank % len(args.gpus)])) if args.input_type == 'ques': model_kwargs = {'vocab': load_vocab(args.vocab_json)} model = VqaLstmModel(**model_kwargs) elif args.input_type == 'ques,image': model_kwargs = {'vocab': load_vocab(args.vocab_json)} model = VqaLstmCnnAttentionModel(**model_kwargs) lossFn = torch.nn.CrossEntropyLoss().cuda() optim = torch.optim.Adam(filter(lambda p: p.requires_grad, shared_model.parameters()), lr=args.learning_rate) train_loader_kwargs = { 'questions_h5': args.train_h5, 'vocab': args.vocab_json, 'batch_size': args.batch_size, 'input_type': args.input_type, 'num_frames': args.num_frames, 'split': 'train', 'max_threads_per_gpu': args.max_threads_per_gpu, 'gpu_id': args.gpus[rank % len(args.gpus)], } args.output_log_path = os.path.join(args.log_dir, 'train_' + str(rank) + '.json') metrics = VqaMetric( info={ 'split': 'train', 'thread': rank }, metric_names=['loss', 'accuracy', 'mean_rank', 'mean_reciprocal_rank'], log_json=args.output_log_path) print('[TRAIN_LOADER] start') train_loader = EqaDataLoader(**train_loader_kwargs) print('train_loader has %d samples' % len(train_loader.dataset)) t, epoch = 0, 0 while epoch < int(args.max_epochs): # print('epoch no. %d' % epoch) if args.input_type == 'ques': for batch in train_loader: t += 1 model.load_state_dict(shared_model.state_dict()) model.train() model.cuda() idx, questions, answers = batch questions_var = Variable(questions.cuda()) answers_var = Variable(answers.cuda()) scores = model(questions_var) loss = lossFn(scores, answers_var) # zero grad optim.zero_grad() # update metrics accuracy, ranks = metrics.compute_ranks( scores.data.cpu(), answers) metrics.update([loss.data[0], accuracy, ranks, 1.0 / ranks]) # backprop and update loss.backward() ensure_shared_grads(model.cpu(), shared_model) optim.step() if t % args.print_every == 0: print(metrics.get_stat_string()) if args.log == True: metrics.dump_log() elif args.input_type == 'ques,image': done = False all_envs_loaded = True #all_envs_loaded = train_loader.dataset._check_if_all_envs_loaded() while done == False: for batch in train_loader: t += 1 model.load_state_dict(shared_model.state_dict()) model.train() model.cnn.eval() model.cuda() idx, questions, answers, images, _, _, _ = batch print('--- images dim {}'.format(images.size())) questions_var = Variable(questions.cuda()) answers_var = Variable(answers.cuda()) images_var = Variable(images.cuda()) scores, att_probs = model(images_var, questions_var) loss = lossFn(scores, answers_var) print('--- att_probs: {}.'.format(att_probs)) # print('--- answers_var: {}.'.format(answers_var)) # print('--- loss: {}.'.format(loss)) # zero grad optim.zero_grad() # update metrics accuracy, ranks = metrics.compute_ranks( scores.data.cpu(), answers) metrics.update([loss.item(), accuracy, ranks, 1.0 / ranks]) # backprop and update loss.backward() ensure_shared_grads(model.cpu(), shared_model) optim.step() if t % args.print_every == 0: print(metrics.get_stat_string()) if args.log == True: metrics.dump_log() if all_envs_loaded == False: print('[CHECK][Cache:%d][Total:%d]' % (len(train_loader.dataset.img_data_cache), len(train_loader.dataset.env_list))) train_loader.dataset._load_envs(in_order=True) if len(train_loader.dataset.pruned_env_set) == 0: done = True else: done = True epoch += 1
def train(rank, args, shared_model): torch.cuda.set_device(args.gpus.index(args.gpus[rank % len(args.gpus)])) if args.input_type == 'ques': model_kwargs = {'vocab': load_vocab(args.vocab_json)} model = VqaLstmModel(**model_kwargs) elif args.input_type == 'ques,image': model_kwargs = {'vocab': load_vocab(args.vocab_json)} model = VqaLstmCnnAttentionModel(**model_kwargs) lossFn = torch.nn.CrossEntropyLoss().cuda() optim = torch.optim.Adam(filter(lambda p: p.requires_grad, shared_model.parameters()), lr=args.learning_rate) train_loader_kwargs = { 'questions_h5': args.train_h5, 'data_json': args.data_json, 'vocab': args.vocab_json, 'batch_size': args.batch_size, 'input_type': args.input_type, 'num_frames': args.num_frames, 'split': 'train', 'max_threads_per_gpu': args.max_threads_per_gpu, 'gpu_id': args.gpus[rank % len(args.gpus)], 'to_cache': args.to_cache } eval_loader_kwargs = { 'questions_h5': getattr(args, args.eval_split + '_h5'), 'data_json': args.data_json, 'vocab': args.vocab_json, 'batch_size': 1, 'input_type': args.input_type, 'num_frames': args.num_frames, 'split': args.eval_split, 'max_threads_per_gpu': args.max_threads_per_gpu, 'gpu_id': args.gpus[rank % len(args.gpus)], 'to_cache': args.to_cache } args.output_log_path = os.path.join(args.log_dir, 'train_' + str(rank) + '.json') metrics = VqaMetric( info={ 'split': 'train', 'thread': rank }, metric_names=['loss', 'accuracy', 'mean_rank', 'mean_reciprocal_rank'], log_json=args.output_log_path) eval_loader = EqaDataLoader(**eval_loader_kwargs) train_loader = EqaDataLoader(**train_loader_kwargs) if args.input_type == 'ques,image': train_loader.dataset._load_envs(start_idx=0, in_order=True) print('train_loader has %d samples' % len(train_loader.dataset)) t, epoch, best_eval_acc = 0, 0, 0 while epoch < int(args.max_epochs): if args.input_type == 'ques': for batch in train_loader: t += 1 model.load_state_dict(shared_model.state_dict()) model.train() model.cuda() idx, questions, answers = batch questions_var = Variable(questions.cuda()) answers_var = Variable(answers.cuda()) scores = model(questions_var) loss = lossFn(scores, answers_var) # zero grad optim.zero_grad() # update metrics accuracy, ranks = metrics.compute_ranks( scores.data.cpu(), answers) metrics.update([loss.data[0], accuracy, ranks, 1.0 / ranks]) # backprop and update loss.backward() ensure_shared_grads(model.cpu(), shared_model) optim.step() if t % args.print_every == 0: print(metrics.get_stat_string()) if args.to_log == 1: metrics.dump_log() elif args.input_type == 'ques,image': done = False all_envs_loaded = train_loader.dataset._check_if_all_envs_loaded() while done == False: for batch in train_loader: t += 1 model.load_state_dict(shared_model.state_dict()) model.train() model.cuda() idx, questions, answers, images, _, _, _ = batch questions_var = Variable(questions.cuda()) answers_var = Variable(answers.cuda()) images_var = Variable(images.cuda()) scores, att_probs = model(images_var, questions_var) loss = lossFn(scores, answers_var) # zero grad optim.zero_grad() # update metrics # accuracy, ranks = metrics.compute_ranks(scores.data.cpu(), answers) # metrics.update([loss.data[0], accuracy, ranks, 1.0 / ranks]) # backprop and update loss.backward() ensure_shared_grads(model.cpu(), shared_model) optim.step() #if t % args.print_every == 0: # print(metrics.get_stat_string()) # if args.to_log == 1: # metrics.dump_log() if all_envs_loaded == False: train_loader.dataset._load_envs(in_order=True) if len(train_loader.dataset.pruned_env_set) == 0: done = True else: done = True env_done = False env_all_envs_loaded = eval_loader.dataset._check_if_all_envs_loaded( ) while env_done == False: _loss, _accuracy, _ranks = None, None, None for batch in eval_loader: t += 1 model.cuda() idx, questions, answers, images, _, _, _ = batch questions_var = Variable(questions.cuda()) answers_var = Variable(answers.cuda()) images_var = Variable(images.cuda()) scores, att_probs = model(images_var, questions_var) loss = lossFn(scores, answers_var) # update metrics accuracy, ranks = metrics.compute_ranks( scores.data.cpu(), answers) if _loss is None: _loss = loss.data[0] _accuracy = accuracy _ranks = ranks else: _loss = torch.cat([_loss, loss.data[0]]) _accuracy = torch.cat([_accuracy, accuracy]) _ranks = torch.cat([_ranks, ranks]) metrics.update([loss.data[0], accuracy, ranks, 1.0 / ranks]) print(metrics.get_stat_string(mode=0)) if env_all_envs_loaded == False: eval_loader.dataset._load_envs() if len(eval_loader.dataset.pruned_env_set) == 0: env_done = True else: env_done = True epoch += 1 # checkpoint if best val accuracy if metrics.metrics[1][0] > best_eval_acc: best_eval_acc = metrics.metrics[1][0] if epoch % args.eval_every == 0 and args.to_log == 1: metrics.dump_log() model_state = get_state(model) if args.checkpoint_path != False: ad = checkpoint['args'] else: ad = args.__dict__ checkpoint = {'args': ad, 'state': model_state, 'epoch': epoch} checkpoint_path = '%s/epoch_%d_accuracy_%.04f.pt' % ( args.checkpoint_dir, epoch, best_eval_acc) print('Saving checkpoint to %s' % checkpoint_path) torch.save(checkpoint, checkpoint_path) print('[best_eval_accuracy:%.04f]' % best_eval_acc)
def train(rank, args, shared_model, use_vision, use_language): gpu_idx = args.gpus.index(args.gpus[rank % len(args.gpus)]) torch.cuda.set_device(gpu_idx) print("train gpu:" + str(gpu_idx) + " assigned") if args.input_type == 'ques': model_kwargs = {'vocab': load_vocab(args.vocab_json)} model = VqaLstmModel(**model_kwargs) elif args.input_type == 'ques,image': model_kwargs = {'vocab': load_vocab(args.vocab_json)} model = VqaLstmCnnAttentionModel(**model_kwargs) lossFn = torch.nn.CrossEntropyLoss().cuda() optim = torch.optim.Adam(filter(lambda p: p.requires_grad, shared_model.parameters()), lr=args.learning_rate) train_loader_kwargs = { 'questions_h5': args.train_h5, 'data_json': args.data_json, 'vocab': args.vocab_json, 'batch_size': args.batch_size, 'input_type': args.input_type, 'num_frames': args.num_frames, 'split': 'train', 'max_threads_per_gpu': args.max_threads_per_gpu, 'gpu_id': args.gpus[rank % len(args.gpus)], 'to_cache': args.to_cache } args.output_log_path = os.path.join(args.log_dir, 'train_' + str(rank) + '.json') metrics = VqaMetric( info={ 'split': 'train', 'thread': rank }, metric_names=['loss', 'accuracy', 'mean_rank', 'mean_reciprocal_rank'], log_json=args.output_log_path) train_loader = EqaDataLoader(**train_loader_kwargs) if args.input_type == 'ques,image': train_loader.dataset._load_envs(start_idx=0, in_order=True) print('train_loader has %d samples' % len(train_loader.dataset)) t, epoch = 0, 0 while epoch < int(args.max_epochs): print("train gpu:" + str(gpu_idx) + " running epoch " + str(epoch)) if args.input_type == 'ques': for batch in train_loader: t += 1 model.load_state_dict(shared_model.state_dict()) model.train() model.cuda() idx, questions, answers = batch # If not using language, replace each question with a start and end token back to back. if not use_language: questions = torch.zeros_like(questions) questions.fill_( model_kwargs['vocab']['questionTokenToIdx']['<NULL>']) questions[:, 0] = model_kwargs['vocab'][ 'questionTokenToIdx']['<START>'] questions[:, 1] = model_kwargs['vocab'][ 'questionTokenToIdx']['<END>'] questions_var = Variable(questions.cuda()) answers_var = Variable(answers.cuda()) scores = model(questions_var) loss = lossFn(scores, answers_var) # zero grad optim.zero_grad() # update metrics accuracy, ranks = metrics.compute_ranks( scores.data.cpu(), answers) metrics.update([loss.data[0], accuracy, ranks, 1.0 / ranks]) # backprop and update loss.backward() ensure_shared_grads(model.cpu(), shared_model) optim.step() if t % args.print_every == 0: print(metrics.get_stat_string()) if args.to_log == 1: metrics.dump_log() elif args.input_type == 'ques,image': done = False all_envs_loaded = train_loader.dataset._check_if_all_envs_loaded() while done == False: for batch in train_loader: t += 1 model.load_state_dict(shared_model.state_dict()) model.train() model.cuda() idx, questions, answers, images, _, _, _ = batch # If not using language, replace each question with a start and end token back to back. if not use_language: questions = torch.zeros_like(questions) questions.fill_(model_kwargs['vocab'] ['questionTokenToIdx']['<NULL>']) questions[:, 0] = model_kwargs['vocab'][ 'questionTokenToIdx']['<START>'] questions[:, 1] = model_kwargs['vocab'][ 'questionTokenToIdx']['<END>'] # If not using vision, replace all image feature data with zeros. if not use_vision: images = torch.zeros_like(images) questions_var = Variable(questions.cuda()) answers_var = Variable(answers.cuda()) images_var = Variable(images.cuda()) scores, att_probs = model(images_var, questions_var) loss = lossFn(scores, answers_var) # zero grad optim.zero_grad() # update metrics accuracy, ranks = metrics.compute_ranks( scores.data.cpu(), answers) metrics.update( [loss.data[0], accuracy, ranks, 1.0 / ranks]) # backprop and update loss.backward() ensure_shared_grads(model.cpu(), shared_model) optim.step() if t % args.print_every == 0: print(metrics.get_stat_string()) if args.to_log == 1: metrics.dump_log() if all_envs_loaded == False: train_loader.dataset._load_envs(in_order=True) if len(train_loader.dataset.pruned_env_set) == 0: done = True else: done = True # Set shared epoch when it finishes on the training side with open(args.identifier + '.shared_epoch.tmp', 'w') as f: f.write(str(epoch)) epoch += 1
def train(rank, args, shared_model): torch.cuda.set_device(args.gpus.index(args.gpus[rank % len(args.gpus)])) if args.forget_rate is None: forget_rate = args.noise_rate else: forget_rate = args.forget_rate rate_schedule = np.ones(args.max_epochs) * forget_rate rate_schedule[:args.num_gradual] = np.linspace(0, forget_rate**args.exponent, args.num_gradual) if args.input_type == 'ques': model_kwargs = {'vocab': load_vocab(args.vocab_json)} model = VqaLstmModel(**model_kwargs) elif args.input_type == 'ques,image': model_kwargs = {'vocab': load_vocab(args.vocab_json)} model = VqaLstmCnnAttentionModel(**model_kwargs) model_kwargs = {'vocab': load_vocab(args.vocab_json)} model1 = VqaLstmCnnAttentionModel(**model_kwargs) # lossFn = torch.nn.CrossEntropyLoss().cuda() optim = torch.optim.Adam(filter(lambda p: p.requires_grad, shared_model.parameters()), lr=args.learning_rate) optim1 = torch.optim.Adam(filter(lambda p: p.requires_grad, model1.parameters()), lr=args.learning_rate) train_loader_kwargs = { 'questions_h5': args.train_h5, 'data_json': args.data_json, 'vocab': args.vocab_json, 'batch_size': args.batch_size, 'input_type': args.input_type, 'num_frames': args.num_frames, 'split': 'train', 'max_threads_per_gpu': args.max_threads_per_gpu, 'gpu_id': args.gpus[rank % len(args.gpus)], 'to_cache': args.cache } args.output_log_path = os.path.join(args.log_dir, 'train_' + str(rank) + '.json') metrics = VqaMetric( info={ 'split': 'train', 'thread': rank }, metric_names=['loss', 'accuracy', 'mean_rank', 'mean_reciprocal_rank'], log_json=args.output_log_path) train_loader = EqaDataLoader(**train_loader_kwargs) if args.input_type == 'ques,image': train_loader.dataset._load_envs(start_idx=0, in_order=True) print('train_loader has %d samples' % len(train_loader.dataset)) t, epoch = 0, 0 while epoch < int(args.max_epochs): if args.input_type == 'ques': for batch in train_loader: t += 1 model.load_state_dict(shared_model.state_dict()) model.train() model.cuda() idx, questions, answers = batch questions_var = Variable(questions.cuda()) answers_var = Variable(answers.cuda()) scores = model(questions_var) loss = lossFn(scores, answers_var) # zero grad optim.zero_grad() # update metrics accuracy, ranks = metrics.compute_ranks( scores.data.cpu(), answers) metrics.update( [loss.data.item(), accuracy, ranks, 1.0 / ranks]) # backprop and update loss.backward() ensure_shared_grads(model.cpu(), shared_model) optim.step() if t % args.print_every == 0: print(metrics.get_stat_string()) if args.log == True: metrics.dump_log() elif args.input_type == 'ques,image': done = False all_envs_loaded = train_loader.dataset._check_if_all_envs_loaded() while done == False: for batch in train_loader: t += 1 model.load_state_dict(shared_model.state_dict()) model.train() model.cnn.eval() model.cuda() # model1.load_state_dict(shared_model1.state_dict()) model1.train() model1.cnn.eval() model1.cuda() idx, questions, answers, images, _, _, _ = batch for i in range(len(answers)): if random.random() < args.noise_rate: tempt = random.randint(-10, 10) answers[i] = answers[i] + tempt while (answers[i] < 0 or answers[i] > 70): answers[i] = answers[i] - tempt tempt = random.randint(-10, 10) answers[i] = answers[i] + tempt questions_var = Variable(questions.cuda()) answers_var = Variable(answers.cuda()) images_var = Variable(images.cuda()) scores, att_probs = model(images_var, questions_var) scores1, att_probs1 = model1(images_var, questions_var) loss, loss1 = loss_coteaching(scores, scores1, answers_var, rate_schedule[epoch]) # zero grad optim.zero_grad() optim1.zero_grad() # update metrics accuracy, ranks = metrics.compute_ranks( scores.data.cpu(), answers) metrics.update( [loss.data.item(), accuracy, ranks, 1.0 / ranks]) # backprop and update loss.backward() loss1.backward() ensure_shared_grads(model.cpu(), shared_model) # ensure_shared_grads(model1.cpu(), shared_model1) optim.step() optim1.step() if t % args.print_every == 0: print(metrics.get_stat_string()) if args.log == True: metrics.dump_log() if all_envs_loaded == False: print('[CHECK][Cache:%d][Total:%d]' % (len(train_loader.dataset.img_data_cache), len(train_loader.dataset.env_list))) train_loader.dataset._load_envs(in_order=True) if len(train_loader.dataset.pruned_env_set) == 0: done = True else: done = True epoch += 1
def train(rank, args, shared_nav_model, shared_ans_model): torch.cuda.set_device(args.gpus.index(args.gpus[rank % len(args.gpus)])) if args.model_type == 'pacman': model_kwargs = {'question_vocab': load_vocab(args.vocab_json)} nav_model = NavPlannerControllerModel(**model_kwargs) else: exit() model_kwargs = {'vocab': load_vocab(args.vocab_json)} ans_model = VqaLstmCnnAttentionModel(**model_kwargs) optim = torch.optim.SGD( filter(lambda p: p.requires_grad, shared_nav_model.parameters()), lr=args.learning_rate) train_loader_kwargs = { 'questions_h5': args.train_h5, 'data_json': args.data_json, 'vocab': args.vocab_json, 'target_obj_conn_map_dir': args.target_obj_conn_map_dir, 'map_resolution': args.map_resolution, 'batch_size': 1, 'input_type': args.model_type, 'num_frames': 5, 'split': 'train', 'max_threads_per_gpu': args.max_threads_per_gpu, 'gpu_id': args.gpus[rank % len(args.gpus)], 'to_cache': args.to_cache } args.output_nav_log_path = os.path.join(args.log_dir, 'nav_train_' + str(rank) + '.json') args.output_ans_log_path = os.path.join(args.log_dir, 'ans_train_' + str(rank) + '.json') nav_model.load_state_dict(shared_nav_model.state_dict()) nav_model.cuda() ans_model.load_state_dict(shared_ans_model.state_dict()) ans_model.eval() ans_model.cuda() nav_metrics = NavMetric( info={'split': 'train', 'thread': rank}, metric_names=[ 'planner_loss', 'controller_loss', 'reward', 'episode_length' ], log_json=args.output_nav_log_path) vqa_metrics = VqaMetric( info={'split': 'train', 'thread': rank}, metric_names=['accuracy', 'mean_rank', 'mean_reciprocal_rank'], log_json=args.output_ans_log_path) train_loader = EqaDataLoader(**train_loader_kwargs) print('train_loader has %d samples' % len(train_loader.dataset)) t, epoch = 0, 0 p_losses, c_losses, reward_list, episode_length_list = [], [], [], [] nav_metrics.update([10.0, 10.0, 0, 100]) mult = 0.1 while epoch < int(args.max_epochs): if 'pacman' in args.model_type: planner_lossFn = MaskedNLLCriterion().cuda() controller_lossFn = MaskedNLLCriterion().cuda() done = False all_envs_loaded = train_loader.dataset._check_if_all_envs_loaded() while done == False: for batch in train_loader: nav_model.load_state_dict(shared_nav_model.state_dict()) nav_model.eval() nav_model.cuda() idx, question, answer, actions, action_length = batch metrics_slug = {} h3d = train_loader.dataset.episode_house # evaluate at multiple initializations # for i in [10, 30, 50]: t += 1 question_var = Variable(question.cuda()) controller_step = False planner_hidden = nav_model.planner_nav_rnn.init_hidden(1) # forward through planner till spawn planner_actions_in, planner_img_feats, controller_step, controller_action_in, controller_img_feat, init_pos = train_loader.dataset.get_hierarchical_features_till_spawn( actions[0, :action_length[0] + 1].numpy(), max(3, int(mult * action_length[0]))) planner_actions_in_var = Variable( planner_actions_in.cuda()) planner_img_feats_var = Variable(planner_img_feats.cuda()) for step in range(planner_actions_in.size(0)): planner_scores, planner_hidden = nav_model.planner_step( question_var, planner_img_feats_var[step].view( 1, 1, 3200), planner_actions_in_var[step].view( 1, 1), planner_hidden) if controller_step == True: controller_img_feat_var = Variable( controller_img_feat.cuda()) controller_action_in_var = Variable( torch.LongTensor(1, 1).fill_( int(controller_action_in)).cuda()) controller_scores = nav_model.controller_step( controller_img_feat_var.view(1, 1, 3200), controller_action_in_var.view(1, 1), planner_hidden[0]) prob = F.softmax(controller_scores, dim=1) controller_action = int( prob.max(1)[1].data.cpu().numpy()[0]) if controller_action == 1: controller_step = True else: controller_step = False action = int(controller_action_in) action_in = torch.LongTensor( 1, 1).fill_(action + 1).cuda() else: prob = F.softmax(planner_scores, dim=1) action = int(prob.max(1)[1].data.cpu().numpy()[0]) action_in = torch.LongTensor( 1, 1).fill_(action + 1).cuda() h3d.env.reset( x=init_pos[0], y=init_pos[2], yaw=init_pos[3]) init_dist_to_target = h3d.get_dist_to_target( h3d.env.cam.pos) if init_dist_to_target < 0: # unreachable # invalids.append([idx[0], i]) continue episode_length = 0 episode_done = True controller_action_counter = 0 dists_to_target, pos_queue = [init_dist_to_target], [ init_pos ] rewards, planner_actions, planner_log_probs, controller_actions, controller_log_probs = [], [], [], [], [] if action != 3: # take the first step img, rwd, episode_done = h3d.step(action, step_reward=True) img = torch.from_numpy(img.transpose( 2, 0, 1)).float() / 255.0 img_feat_var = train_loader.dataset.cnn( Variable(img.view(1, 3, 224, 224).cuda())).view( 1, 1, 3200) for step in range(args.max_episode_length): episode_length += 1 if controller_step == False: planner_scores, planner_hidden = nav_model.planner_step( question_var, img_feat_var, Variable(action_in), planner_hidden) planner_prob = F.softmax(planner_scores, dim=1) planner_log_prob = F.log_softmax( planner_scores, dim=1) action = planner_prob.multinomial().data planner_log_prob = planner_log_prob.gather( 1, Variable(action)) planner_log_probs.append( planner_log_prob.cpu()) action = int(action.cpu().numpy()[0, 0]) planner_actions.append(action) img, rwd, episode_done = h3d.step(action, step_reward=True) episode_done = episode_done or episode_length >= args.max_episode_length rewards.append(rwd) img = torch.from_numpy(img.transpose( 2, 0, 1)).float() / 255.0 img_feat_var = train_loader.dataset.cnn( Variable(img.view(1, 3, 224, 224) .cuda())).view(1, 1, 3200) dists_to_target.append( h3d.get_dist_to_target(h3d.env.cam.pos)) pos_queue.append([ h3d.env.cam.pos.x, h3d.env.cam.pos.y, h3d.env.cam.pos.z, h3d.env.cam.yaw ]) if episode_done == True: break # query controller to continue or not controller_action_in = Variable( torch.LongTensor(1, 1).fill_(action).cuda()) controller_scores = nav_model.controller_step( img_feat_var, controller_action_in, planner_hidden[0]) controller_prob = F.softmax( controller_scores, dim=1) controller_log_prob = F.log_softmax( controller_scores, dim=1) controller_action = controller_prob.multinomial( ).data if int(controller_action[0] ) == 1 and controller_action_counter < 4: controller_action_counter += 1 controller_step = True else: controller_action_counter = 0 controller_step = False controller_action.fill_(0) controller_log_prob = controller_log_prob.gather( 1, Variable(controller_action)) controller_log_probs.append( controller_log_prob.cpu()) controller_action = int( controller_action.cpu().numpy()[0, 0]) controller_actions.append(controller_action) action_in = torch.LongTensor( 1, 1).fill_(action + 1).cuda() # run answerer here ans_acc = [0] if action == 3: if len(pos_queue) < 5: pos_queue = train_loader.dataset.episode_pos_queue[len( pos_queue) - 5:] + pos_queue images = train_loader.dataset.get_frames( h3d, pos_queue[-5:], preprocess=True) images_var = Variable( torch.from_numpy(images).cuda()).view( 1, 5, 3, 224, 224) scores, att_probs = ans_model(images_var, question_var) ans_acc, ans_rank = vqa_metrics.compute_ranks( scores.data.cpu(), answer) vqa_metrics.update([ans_acc, ans_rank, 1.0 / ans_rank]) rewards.append(h3d.success_reward * ans_acc[0]) R = torch.zeros(1, 1) planner_loss = 0 controller_loss = 0 planner_rev_idx = -1 for i in reversed(range(len(rewards))): R = 0.99 * R + rewards[i] advantage = R - nav_metrics.metrics[2][1] if i < len(controller_actions): controller_loss = controller_loss - controller_log_probs[i] * Variable( advantage) if controller_actions[i] == 0 and planner_rev_idx + len(planner_log_probs) >= 0: planner_loss = planner_loss - planner_log_probs[planner_rev_idx] * Variable( advantage) planner_rev_idx -= 1 elif planner_rev_idx + len(planner_log_probs) >= 0: planner_loss = planner_loss - planner_log_probs[planner_rev_idx] * Variable( advantage) planner_rev_idx -= 1 controller_loss /= max(1, len(controller_log_probs)) planner_loss /= max(1, len(planner_log_probs)) optim.zero_grad() if isinstance(planner_loss, float) == False and isinstance( controller_loss, float) == False: p_losses.append(planner_loss.data[0, 0]) c_losses.append(controller_loss.data[0, 0]) reward_list.append(np.sum(rewards)) episode_length_list.append(episode_length) (planner_loss + controller_loss).backward() ensure_shared_grads(nav_model.cpu(), shared_nav_model) optim.step() if len(reward_list) > 50: nav_metrics.update([ p_losses, c_losses, reward_list, episode_length_list ]) print(nav_metrics.get_stat_string()) if args.to_log == 1: nav_metrics.dump_log() if nav_metrics.metrics[2][1] > 0.35: mult = min(mult + 0.1, 1.0) p_losses, c_losses, reward_list, episode_length_list = [], [], [], [] if all_envs_loaded == False: train_loader.dataset._load_envs(in_order=True) if len(train_loader.dataset.pruned_env_set) == 0: done = True if args.to_cache == False: train_loader.dataset._load_envs( start_idx=0, in_order=True) else: done = True epoch += 1
def eval(rank, args, shared_nav_model, shared_ans_model): torch.cuda.set_device(args.gpus.index(args.gpus[rank % len(args.gpus)])) if args.model_type == 'pacman': model_kwargs = {'question_vocab': load_vocab(args.vocab_json)} nav_model = NavPlannerControllerModel(**model_kwargs) else: exit() model_kwargs = {'vocab': load_vocab(args.vocab_json)} ans_model = VqaLstmCnnAttentionModel(**model_kwargs) eval_loader_kwargs = { 'questions_h5': getattr(args, args.eval_split + '_h5'), 'data_json': args.data_json, 'vocab': args.vocab_json, 'target_obj_conn_map_dir': args.target_obj_conn_map_dir, 'map_resolution': args.map_resolution, 'batch_size': 1, 'input_type': args.model_type, 'num_frames': 5, 'split': args.eval_split, 'max_threads_per_gpu': args.max_threads_per_gpu, 'gpu_id': args.gpus[rank % len(args.gpus)], 'to_cache': False } eval_loader = EqaDataLoader(**eval_loader_kwargs) print('eval_loader has %d samples' % len(eval_loader.dataset)) args.output_nav_log_path = os.path.join(args.log_dir, 'nav_eval_' + str(rank) + '.json') args.output_ans_log_path = os.path.join(args.log_dir, 'ans_eval_' + str(rank) + '.json') t, epoch, best_eval_acc = 0, 0, 0.0 while epoch < int(args.max_epochs): start_time = time.time() invalids = [] nav_model.load_state_dict(shared_nav_model.state_dict()) nav_model.eval() ans_model.load_state_dict(shared_ans_model.state_dict()) ans_model.eval() ans_model.cuda() # that's a lot of numbers nav_metrics = NavMetric( info={'split': args.eval_split, 'thread': rank}, metric_names=[ 'd_0_10', 'd_0_30', 'd_0_50', 'd_T_10', 'd_T_30', 'd_T_50', 'd_D_10', 'd_D_30', 'd_D_50', 'd_min_10', 'd_min_30', 'd_min_50', 'r_T_10', 'r_T_30', 'r_T_50', 'r_e_10', 'r_e_30', 'r_e_50', 'stop_10', 'stop_30', 'stop_50', 'ep_len_10', 'ep_len_30', 'ep_len_50' ], log_json=args.output_nav_log_path) vqa_metrics = VqaMetric( info={'split': args.eval_split, 'thread': rank}, metric_names=[ 'accuracy_10', 'accuracy_30', 'accuracy_50', 'mean_rank_10', 'mean_rank_30', 'mean_rank_50', 'mean_reciprocal_rank_10', 'mean_reciprocal_rank_30', 'mean_reciprocal_rank_50' ], log_json=args.output_ans_log_path) if 'pacman' in args.model_type: done = False while done == False: for batch in tqdm(eval_loader): nav_model.load_state_dict(shared_nav_model.state_dict()) nav_model.eval() nav_model.cuda() idx, question, answer, actions, action_length = batch metrics_slug = {} h3d = eval_loader.dataset.episode_house # evaluate at multiple initializations for i in [10, 30, 50]: t += 1 if i > action_length[0]: invalids.append([idx[0], i]) continue question_var = Variable(question.cuda()) controller_step = False planner_hidden = nav_model.planner_nav_rnn.init_hidden( 1) # forward through planner till spawn planner_actions_in, planner_img_feats, controller_step, controller_action_in, controller_img_feat, init_pos = eval_loader.dataset.get_hierarchical_features_till_spawn( actions[0, :action_length[0] + 1].numpy(), i) planner_actions_in_var = Variable( planner_actions_in.cuda()) planner_img_feats_var = Variable( planner_img_feats.cuda()) for step in range(planner_actions_in.size(0)): planner_scores, planner_hidden = nav_model.planner_step( question_var, planner_img_feats_var[step].view( 1, 1, 3200), planner_actions_in_var[step].view( 1, 1), planner_hidden) if controller_step == True: controller_img_feat_var = Variable( controller_img_feat.cuda()) controller_action_in_var = Variable( torch.LongTensor(1, 1).fill_( int(controller_action_in)).cuda()) controller_scores = nav_model.controller_step( controller_img_feat_var.view(1, 1, 3200), controller_action_in_var.view(1, 1), planner_hidden[0]) prob = F.softmax(controller_scores, dim=1) controller_action = int( prob.max(1)[1].data.cpu().numpy()[0]) if controller_action == 1: controller_step = True else: controller_step = False action = int(controller_action_in) action_in = torch.LongTensor( 1, 1).fill_(action + 1).cuda() else: prob = F.softmax(planner_scores, dim=1) action = int(prob.max(1)[1].data.cpu().numpy()[0]) action_in = torch.LongTensor( 1, 1).fill_(action + 1).cuda() h3d.env.reset( x=init_pos[0], y=init_pos[2], yaw=init_pos[3]) init_dist_to_target = h3d.get_dist_to_target( h3d.env.cam.pos) if init_dist_to_target < 0: # unreachable invalids.append([idx[0], i]) continue episode_length = 0 episode_done = True controller_action_counter = 0 dists_to_target, pos_queue, pred_actions = [ init_dist_to_target ], [init_pos], [] planner_actions, controller_actions = [], [] if action != 3: # take the first step img, _, _ = h3d.step(action) img = torch.from_numpy(img.transpose( 2, 0, 1)).float() / 255.0 img_feat_var = eval_loader.dataset.cnn( Variable(img.view(1, 3, 224, 224).cuda())).view( 1, 1, 3200) for step in range(args.max_episode_length): episode_length += 1 if controller_step == False: planner_scores, planner_hidden = nav_model.planner_step( question_var, img_feat_var, Variable(action_in), planner_hidden) prob = F.softmax(planner_scores, dim=1) action = int( prob.max(1)[1].data.cpu().numpy()[0]) planner_actions.append(action) pred_actions.append(action) img, _, episode_done = h3d.step(action) episode_done = episode_done or episode_length >= args.max_episode_length img = torch.from_numpy(img.transpose( 2, 0, 1)).float() / 255.0 img_feat_var = eval_loader.dataset.cnn( Variable(img.view(1, 3, 224, 224) .cuda())).view(1, 1, 3200) dists_to_target.append( h3d.get_dist_to_target(h3d.env.cam.pos)) pos_queue.append([ h3d.env.cam.pos.x, h3d.env.cam.pos.y, h3d.env.cam.pos.z, h3d.env.cam.yaw ]) if episode_done == True: break # query controller to continue or not controller_action_in = Variable( torch.LongTensor(1, 1).fill_(action).cuda()) controller_scores = nav_model.controller_step( img_feat_var, controller_action_in, planner_hidden[0]) prob = F.softmax(controller_scores, dim=1) controller_action = int( prob.max(1)[1].data.cpu().numpy()[0]) if controller_action == 1 and controller_action_counter < 4: controller_action_counter += 1 controller_step = True else: controller_action_counter = 0 controller_step = False controller_action = 0 controller_actions.append(controller_action) action_in = torch.LongTensor( 1, 1).fill_(action + 1).cuda() # run answerer here if len(pos_queue) < 5: pos_queue = eval_loader.dataset.episode_pos_queue[len( pos_queue) - 5:] + pos_queue images = eval_loader.dataset.get_frames( h3d, pos_queue[-5:], preprocess=True) images_var = Variable( torch.from_numpy(images).cuda()).view( 1, 5, 3, 224, 224) scores, att_probs = ans_model(images_var, question_var) ans_acc, ans_rank = vqa_metrics.compute_ranks( scores.data.cpu(), answer) pred_answer = scores.max(1)[1].data[0] print('[Q_GT]', ' '.join([ eval_loader.dataset.vocab['questionIdxToToken'][x] for x in question[0] if x != 0 ])) print('[A_GT]', eval_loader.dataset.vocab[ 'answerIdxToToken'][answer[0]]) print('[A_PRED]', eval_loader.dataset.vocab[ 'answerIdxToToken'][pred_answer]) # compute stats metrics_slug['accuracy_' + str(i)] = ans_acc[0] metrics_slug['mean_rank_' + str(i)] = ans_rank[0] metrics_slug['mean_reciprocal_rank_' + str(i)] = 1.0 / ans_rank[0] metrics_slug['d_0_' + str(i)] = dists_to_target[0] metrics_slug['d_T_' + str(i)] = dists_to_target[-1] metrics_slug['d_D_' + str( i)] = dists_to_target[0] - dists_to_target[-1] metrics_slug['d_min_' + str(i)] = np.array( dists_to_target).min() metrics_slug['ep_len_' + str(i)] = episode_length if action == 3: metrics_slug['stop_' + str(i)] = 1 else: metrics_slug['stop_' + str(i)] = 0 inside_room = [] for p in pos_queue: inside_room.append( h3d.is_inside_room( p, eval_loader.dataset.target_room)) if inside_room[-1] == True: metrics_slug['r_T_' + str(i)] = 1 else: metrics_slug['r_T_' + str(i)] = 0 if any([x == True for x in inside_room]) == True: metrics_slug['r_e_' + str(i)] = 1 else: metrics_slug['r_e_' + str(i)] = 0 # navigation metrics metrics_list = [] for i in nav_metrics.metric_names: if i not in metrics_slug: metrics_list.append(nav_metrics.metrics[ nav_metrics.metric_names.index(i)][0]) else: metrics_list.append(metrics_slug[i]) nav_metrics.update(metrics_list) # vqa metrics metrics_list = [] for i in vqa_metrics.metric_names: if i not in metrics_slug: metrics_list.append(vqa_metrics.metrics[ vqa_metrics.metric_names.index(i)][0]) else: metrics_list.append(metrics_slug[i]) vqa_metrics.update(metrics_list) try: print(nav_metrics.get_stat_string(mode=0)) print(vqa_metrics.get_stat_string(mode=0)) except: pass print('epoch', epoch) print('invalids', len(invalids)) eval_loader.dataset._load_envs() if len(eval_loader.dataset.pruned_env_set) == 0: done = True epoch += 1 # checkpoint if best val accuracy if vqa_metrics.metrics[2][0] > best_eval_acc: # ans_acc_50 best_eval_acc = vqa_metrics.metrics[2][0] if epoch % args.eval_every == 0 and args.to_log == 1: vqa_metrics.dump_log() nav_metrics.dump_log() model_state = get_state(nav_model) aad = dict(args.__dict__) ad = {} for i in aad: if i[0] != '_': ad[i] = aad[i] checkpoint = {'args': ad, 'state': model_state, 'epoch': epoch} checkpoint_path = '%s/epoch_%d_ans_50_%.04f.pt' % ( args.checkpoint_dir, epoch, best_eval_acc) print('Saving checkpoint to %s' % checkpoint_path) torch.save(checkpoint, checkpoint_path) print('[best_eval_ans_acc_50:%.04f]' % best_eval_acc) eval_loader.dataset._load_envs(start_idx=0, in_order=True)
def eval(rank, args, shared_model): torch.cuda.set_device(args.gpus.index(args.gpus[rank % len(args.gpus)])) if args.input_type == 'ques': model_kwargs = {'vocab': load_vocab(args.vocab_json)} model = VqaLstmModel(**model_kwargs) elif args.input_type == 'ques,image': model_kwargs = {'vocab': load_vocab(args.vocab_json)} model = VqaLstmCnnAttentionModel(**model_kwargs) lossFn = torch.nn.CrossEntropyLoss().cuda() eval_loader_kwargs = { 'questions_h5': getattr(args, args.eval_split + '_h5'), 'data_json': args.data_json, 'vocab': args.vocab_json, 'batch_size': 1, 'input_type': args.input_type, 'num_frames': args.num_frames, 'split': args.eval_split, 'max_threads_per_gpu': args.max_threads_per_gpu, 'gpu_id': args.gpus[rank%len(args.gpus)], 'to_cache': args.to_cache } eval_loader = EqaDataLoader(**eval_loader_kwargs) print('eval_loader has %d samples' % len(eval_loader.dataset)) args.output_log_path = os.path.join(args.log_dir, 'eval_' + str(rank) + '.json') t, epoch, best_eval_acc = 0, 0, 0 while epoch < int(args.max_epochs): model.load_state_dict(shared_model.state_dict()) model.eval() metrics = VqaMetric( info={'split': args.eval_split}, metric_names=[ 'loss', 'accuracy', 'mean_rank', 'mean_reciprocal_rank' ], log_json=args.output_log_path) if args.input_type == 'ques': for batch in eval_loader: t += 1 model.cuda() idx, questions, answers = batch questions_var = Variable(questions.cuda()) answers_var = Variable(answers.cuda()) scores = model(questions_var) loss = lossFn(scores, answers_var) # update metrics accuracy, ranks = metrics.compute_ranks( scores.data.cpu(), answers) metrics.update([loss.data[0], accuracy, ranks, 1.0 / ranks]) print(metrics.get_stat_string(mode=0)) elif args.input_type == 'ques,image': done = False all_envs_loaded = eval_loader.dataset._check_if_all_envs_loaded() while done == False: for batch in eval_loader: t += 1 model.cuda() idx, questions, answers, images, _, _, _ = batch questions_var = Variable(questions.cuda()) answers_var = Variable(answers.cuda()) images_var = Variable(images.cuda()) scores, att_probs = model(images_var, questions_var) loss = lossFn(scores, answers_var) # update metrics accuracy, ranks = metrics.compute_ranks( scores.data.cpu(), answers) metrics.update( [loss.data[0], accuracy, ranks, 1.0 / ranks]) print(metrics.get_stat_string(mode=0)) if all_envs_loaded == False: eval_loader.dataset._load_envs() if len(eval_loader.dataset.pruned_env_set) == 0: done = True else: done = True epoch += 1 # checkpoint if best val accuracy if metrics.metrics[1][0] > best_eval_acc: best_eval_acc = metrics.metrics[1][0] if epoch % args.eval_every == 0 and args.to_log == 1: metrics.dump_log() model_state = get_state(model) if args.checkpoint_path != False: ad = checkpoint['args'] else: ad = args.__dict__ checkpoint = {'args': ad, 'state': model_state, 'epoch': epoch} checkpoint_path = '%s/epoch_%d_accuracy_%.04f.pt' % ( args.checkpoint_dir, epoch, best_eval_acc) print('Saving checkpoint to %s' % checkpoint_path) torch.save(checkpoint, checkpoint_path) print('[best_eval_accuracy:%.04f]' % best_eval_acc)
def train(rank, args, shared_model): torch.cuda.set_device(args.gpus.index(args.gpus[rank % len(args.gpus)])) if args.input_type == 'ques': model_kwargs = {'vocab': load_vocab(args.vocab_json)} model = VqaLstmModel(**model_kwargs) elif args.input_type == 'ques,image': model_kwargs = {'vocab': load_vocab(args.vocab_json)} model = VqaLstmCnnAttentionModel(**model_kwargs) lossFn = torch.nn.CrossEntropyLoss().cuda() optim = torch.optim.Adam( filter(lambda p: p.requires_grad, shared_model.parameters()), lr=args.learning_rate) train_loader_kwargs = { 'questions_h5': args.train_h5, 'data_json': args.data_json, 'vocab': args.vocab_json, 'batch_size': args.batch_size, 'input_type': args.input_type, 'num_frames': args.num_frames, 'split': 'train', 'max_threads_per_gpu': args.max_threads_per_gpu, 'gpu_id': args.gpus[rank%len(args.gpus)], 'to_cache': args.to_cache } args.output_log_path = os.path.join(args.log_dir, 'train_' + str(rank) + '.json') metrics = VqaMetric( info={'split': 'train', 'thread': rank}, metric_names=['loss', 'accuracy', 'mean_rank', 'mean_reciprocal_rank'], log_json=args.output_log_path) train_loader = EqaDataLoader(**train_loader_kwargs) if args.input_type == 'ques,image': train_loader.dataset._load_envs(start_idx=0, in_order=True) print('train_loader has %d samples' % len(train_loader.dataset)) t, epoch = 0, 0 while epoch < int(args.max_epochs): if args.input_type == 'ques': for batch in train_loader: t += 1 model.load_state_dict(shared_model.state_dict()) model.train() model.cuda() idx, questions, answers = batch questions_var = Variable(questions.cuda()) answers_var = Variable(answers.cuda()) scores = model(questions_var) loss = lossFn(scores, answers_var) # zero grad optim.zero_grad() # update metrics accuracy, ranks = metrics.compute_ranks(scores.data.cpu(), answers) metrics.update([loss.data[0], accuracy, ranks, 1.0 / ranks]) # backprop and update loss.backward() ensure_shared_grads(model.cpu(), shared_model) optim.step() if t % args.print_every == 0: print(metrics.get_stat_string()) if args.to_log == 1: metrics.dump_log() elif args.input_type == 'ques,image': done = False all_envs_loaded = train_loader.dataset._check_if_all_envs_loaded() while done == False: for batch in train_loader: t += 1 model.load_state_dict(shared_model.state_dict()) model.train() model.cuda() idx, questions, answers, images, _, _, _ = batch questions_var = Variable(questions.cuda()) answers_var = Variable(answers.cuda()) images_var = Variable(images.cuda()) scores, att_probs = model(images_var, questions_var) loss = lossFn(scores, answers_var) # zero grad optim.zero_grad() # update metrics accuracy, ranks = metrics.compute_ranks(scores.data.cpu(), answers) metrics.update([loss.data[0], accuracy, ranks, 1.0 / ranks]) # backprop and update loss.backward() ensure_shared_grads(model.cpu(), shared_model) optim.step() if t % args.print_every == 0: print(metrics.get_stat_string()) if args.to_log == 1: metrics.dump_log() if all_envs_loaded == False: train_loader.dataset._load_envs(in_order=True) if len(train_loader.dataset.pruned_env_set) == 0: done = True else: done = True epoch += 1