tutorials/TrainData_MultiOmics.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "e48c5f86-f0aa-4963-8913-604443fa0a3c",
   "metadata": {},
   "source": [
    "# Tutorial:  Training using multi-omics data"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5aac073f-ef61-49f9-bf3e-13b2a3137121",
   "metadata": {},
   "source": [
    "## Introduction\n",
    "\n",
    "This tutorial demonstrates how to use **scGenAI** to train a model for predicting cell types from CITE-Seq data (RNA + ADT). In this example, we apply the **random context** method to generate input sequences.\n",
    "\n",
    "**Note:**\n",
    "1. The dataset used for this tutorial consists of **40 cells**, selected for demonstration purposes. For practical applications, it is recommended to use a larger, more representative dataset.\n",
    "2. In this tutorial, we will load the model template directly from the cache folder.\n",
    "3. While the model in this example is trained to predict cell types, it can be applied to predict any feature in practical applications.\n",
    "4. We use the **random context** method to build the context for each cell in this tutorial. Other methods (`genomics`, `biofounction`, `genelist`) can also be applied to the same multi-omics data, as demonstrated with single-cell RNA-seq data.\n",
    "5. This tutorial uses RNA and antibody (ADT) data. However, other types of multi-omics data (e.g., scATAC-Seq, MissionBio) can be used in the same way by setting one omic dataset as `train_file`, and the other as `train_ADTfile`.\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8294154a-4036-4788-b358-0e91342ac224",
   "metadata": {},
   "source": [
    "## Train the Model Using CITE-Seq (Multiple GPU)\n",
    "\n",
    "The training process can be easily initiated by calling the `run_training_from_config` function in **scGenAI**, using the input files specified in a configuration YAML file. Example template YAML files are available in the `scgenai/examples/config_templates` folder of the package. These templates provide the necessary structure for defining training parameters, model configurations, and input data locations.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "2989a022-6bea-4f9c-861d-452b0d85e11f",
   "metadata": {},
   "outputs": [],
   "source": [
    "from scGenAI.training.train import run_training_from_config"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "5a140f4c-6d3c-4a18-9f15-87d148179ac8",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "model_dir: /home/ubuntu/scgenai/examples/testrun/test1/\n",
      "train_file: /home/ubuntu/scgenai/examples/data/example_train_RNA.h5ad\n",
      "val_file: /home/ubuntu/scgenai/examples/data/example_val_RNA.h5ad\n",
      "train_ADTfile: /home/ubuntu/scgenai/examples/data/example_train_ADT.h5ad\n",
      "val_ADTfile: /home/ubuntu/scgenai/examples/data/example_val_ADT.h5ad\n",
      "log_dir: /home/ubuntu/scgenai/examples/logs/\n",
      "output_dir: /home/ubuntu/scgenai/examples/output\n",
      "seed: 1314521\n",
      "cache_dir: /home/ubuntu/scgenai/examples/tmp/cache\n",
      "mode: Train\n",
      "min_cells: 2\n",
      "target_feature: celltype.l2\n",
      "multiomics: Yes\n",
      "keepIntermediateFiles: No\n",
      "savelog: Yes\n",
      "savesetting: Yes\n",
      "model_backbone_name: llama\n",
      "model_backbone_size: small\n",
      "context_method: random\n",
      "num_bins: 10\n",
      "optimizer: AdamW\n",
      "weight_decay: 0.01\n",
      "depth: 2\n",
      "max_length: 5120\n",
      "batch_size: 1\n",
      "learning_rate: 1e-05\n",
      "num_epochs: 20\n",
      "world_size: 1\n",
      "evaluate_during_training: True\n",
      "hidden_size: 256\n",
      "num_layers: 6\n",
      "num_heads: 4\n",
      "intermediate_size: 1024\n",
      "stride: 2560\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2024-10-16 00:45:11.150 | INFO     | scGenAI.utils.distributed:setup_distributed:26 - Distributed setup complete for GPU 0, Total GPUs used: 4\n",
      "2024-10-16 00:45:11.159 | INFO     | scGenAI.utils.distributed:setup_distributed:26 - Distributed setup complete for GPU 1, Total GPUs used: 4\n",
      "2024-10-16 00:45:11.163 | INFO     | scGenAI.utils.distributed:setup_distributed:26 - Distributed setup complete for GPU 3, Total GPUs used: 4\n",
      "2024-10-16 00:45:11.165 | INFO     | scGenAI.utils.distributed:setup_distributed:26 - Distributed setup complete for GPU 2, Total GPUs used: 4\n",
      "2024-10-16 00:45:13.393 | INFO     | scGenAI.training.train:model_train_and_eval:214 - Total number of trainable parameters: 75949828; Max Sequence Legth: 2560 genes\n",
      "Epoch 1/20: 100%|██████████| 14/14 [00:10<00:00,  1.34it/s, GPU_ID=3, train_accuracy=0.286, train_loss=1.5]]\n",
      "Epoch 1/20: 100%|██████████| 14/14 [00:10<00:00,  1.34it/s, GPU_ID=0, train_accuracy=0.357, train_loss=1.39]\n",
      "\n",
      "Epoch 1/20: 100%|██████████| 14/14 [00:10<00:00,  1.34it/s, GPU_ID=2, train_accuracy=0.214, train_loss=1.44]\n",
      "Evaluating Epoch 1/20: 100%|██████████| 54/54 [00:11<00:00,  4.56it/s]\n",
      "Evaluating Epoch 1/20: 100%|██████████| 54/54 [00:11<00:00,  4.53it/s]\n",
      "2024-10-16 00:45:36.342 | INFO     | scGenAI.training.train:evaluate_and_save_predictions:328 - Epoch 1/20, Train Loss: 1.3860, Train Accuracy: 0.3571, Context-level Val Loss: 1.2658, Context-level Val Accuracy: 0.4444\n",
      "Evaluating Epoch 1/20: 100%|██████████| 54/54 [00:11<00:00,  4.51it/s]\n",
      "Evaluating Epoch 1/20: 100%|██████████| 54/54 [00:12<00:00,  4.45it/s]\n",
      "Epoch 2/20:   0%|          | 0/14 [00:00<?, ?it/s]2024-10-16 00:45:36.951 | INFO     | scGenAI.utils.distributed:savemodel:54 - Model state saved to /home/ubuntu/scgenai/examples/testrun/test1/best_model/scGenAI_model.pt\n",
      "2024-10-16 00:45:37.099 | INFO     | scGenAI.utils.distributed:savemodel:59 - Tokenizer saved to /home/ubuntu/scgenai/examples/testrun/test1/best_model\n",
      "2024-10-16 00:45:37.102 | INFO     | scGenAI.utils.distributed:savemodel:75 - Custom vocab and label encoder saved\n",
      "2024-10-16 00:45:37.103 | INFO     | scGenAI.utils.distributed:savemodel:81 - Model configuration saved to /home/ubuntu/scgenai/examples/testrun/test1/best_model\n",
      "Epoch 2/20: 100%|██████████| 14/14 [00:10<00:00,  1.34it/s, GPU_ID=3, train_accuracy=0.357, train_loss=1.26] \n",
      "Epoch 2/20: 100%|██████████| 14/14 [00:10<00:00,  1.31it/s, GPU_ID=2, train_accuracy=0.5, train_loss=1.24]\n",
      "Epoch 2/20: 100%|██████████| 14/14 [00:09<00:00,  1.40it/s, GPU_ID=0, train_accuracy=0.5, train_loss=1.14]\n",
      "Epoch 2/20: 100%|██████████| 14/14 [00:10<00:00,  1.32it/s, GPU_ID=1, train_accuracy=0.643, train_loss=1.06]\n",
      "Evaluating Epoch 2/20: 100%|██████████| 54/54 [00:11<00:00,  4.53it/s]\n",
      "Evaluating Epoch 2/20: 100%|██████████| 54/54 [00:11<00:00,  4.52it/s]\n",
      "2024-10-16 00:45:59.024 | INFO     | scGenAI.training.train:evaluate_and_save_predictions:328 - Epoch 2/20, Train Loss: 1.1439, Train Accuracy: 0.5000, Context-level Val Loss: 1.2329, Context-level Val Accuracy: 0.4444\n",
      "Evaluating Epoch 2/20: 100%|██████████| 54/54 [00:11<00:00,  4.51it/s]\n",
      "Evaluating Epoch 2/20: 100%|██████████| 54/54 [00:12<00:00,  4.43it/s]\n",
      "Epoch 3/20:   0%|          | 0/14 [00:00<?, ?it/s]2024-10-16 00:46:00.919 | INFO     | scGenAI.utils.distributed:savemodel:54 - Model state saved to /home/ubuntu/scgenai/examples/testrun/test1/best_model/scGenAI_model.pt\n",
      "2024-10-16 00:46:01.067 | INFO     | scGenAI.utils.distributed:savemodel:59 - Tokenizer saved to /home/ubuntu/scgenai/examples/testrun/test1/best_model\n",
      "2024-10-16 00:46:01.070 | INFO     | scGenAI.utils.distributed:savemodel:75 - Custom vocab and label encoder saved\n",
      "2024-10-16 00:46:01.071 | INFO     | scGenAI.utils.distributed:savemodel:81 - Model configuration saved to /home/ubuntu/scgenai/examples/testrun/test1/best_model\n",
      "Epoch 3/20: 100%|██████████| 14/14 [00:11<00:00,  1.19it/s, GPU_ID=3, train_accuracy=0.571, train_loss=1.2]] \n",
      "Epoch 3/20: 100%|██████████| 14/14 [00:12<00:00,  1.17it/s, GPU_ID=2, train_accuracy=0.429, train_loss=1.16]\n",
      "Epoch 3/20: 100%|██████████| 14/14 [00:10<00:00,  1.40it/s, GPU_ID=0, train_accuracy=0.5, train_loss=1.09]\n",
      "Epoch 3/20: 100%|██████████| 14/14 [00:11<00:00,  1.17it/s, GPU_ID=1, train_accuracy=0.571, train_loss=0.98]\n",
      "Evaluating Epoch 3/20: 100%|██████████| 54/54 [00:12<00:00,  4.50it/s]\n",
      "Evaluating Epoch 3/20: 100%|██████████| 54/54 [00:12<00:00,  4.49it/s]\n",
      "2024-10-16 00:46:23.109 | INFO     | scGenAI.training.train:evaluate_and_save_predictions:328 - Epoch 3/20, Train Loss: 1.0928, Train Accuracy: 0.5000, Context-level Val Loss: 1.1734, Context-level Val Accuracy: 0.5000\n",
      "Evaluating Epoch 3/20: 100%|██████████| 54/54 [00:12<00:00,  4.48it/s]\n",
      "Evaluating Epoch 3/20: 100%|██████████| 54/54 [00:12<00:00,  4.41it/s]\n",
      "Epoch 4/20:   0%|          | 0/14 [00:00<?, ?it/s]2024-10-16 00:46:25.002 | INFO     | scGenAI.utils.distributed:savemodel:54 - Model state saved to /home/ubuntu/scgenai/examples/testrun/test1/best_model/scGenAI_model.pt\n",
      "2024-10-16 00:46:25.148 | INFO     | scGenAI.utils.distributed:savemodel:59 - Tokenizer saved to /home/ubuntu/scgenai/examples/testrun/test1/best_model\n",
      "2024-10-16 00:46:25.154 | INFO     | scGenAI.utils.distributed:savemodel:75 - Custom vocab and label encoder saved\n",
      "2024-10-16 00:46:25.155 | INFO     | scGenAI.utils.distributed:savemodel:81 - Model configuration saved to /home/ubuntu/scgenai/examples/testrun/test1/best_model\n",
      "Epoch 4/20: 100%|██████████| 14/14 [00:11<00:00,  1.19it/s, GPU_ID=3, train_accuracy=0.643, train_loss=1.06] \n",
      "\n",
      "\n",
      "Epoch 4/20: 100%|██████████| 14/14 [00:11<00:00,  1.17it/s, GPU_ID=1, train_accuracy=0.714, train_loss=0.85]\n",
      "Evaluating Epoch 4/20: 100%|██████████| 54/54 [00:12<00:00,  4.49it/s]\n",
      "Evaluating Epoch 4/20: 100%|██████████| 54/54 [00:12<00:00,  4.48it/s]\n",
      "2024-10-16 00:46:47.249 | INFO     | scGenAI.training.train:evaluate_and_save_predictions:328 - Epoch 4/20, Train Loss: 1.0234, Train Accuracy: 0.5714, Context-level Val Loss: 1.0995, Context-level Val Accuracy: 0.5000\n",
      "Evaluating Epoch 4/20: 100%|██████████| 54/54 [00:12<00:00,  4.47it/s]\n",
      "Evaluating Epoch 4/20: 100%|██████████| 54/54 [00:12<00:00,  4.39it/s]\n",
      "Epoch 5/20:   0%|          | 0/14 [00:00<?, ?it/s]2024-10-16 00:46:49.141 | INFO     | scGenAI.utils.distributed:savemodel:54 - Model state saved to /home/ubuntu/scgenai/examples/testrun/test1/best_model/scGenAI_model.pt\n",
      "2024-10-16 00:46:49.288 | INFO     | scGenAI.utils.distributed:savemodel:59 - Tokenizer saved to /home/ubuntu/scgenai/examples/testrun/test1/best_model\n",
      "2024-10-16 00:46:49.291 | INFO     | scGenAI.utils.distributed:savemodel:75 - Custom vocab and label encoder saved\n",
      "2024-10-16 00:46:49.292 | INFO     | scGenAI.utils.distributed:savemodel:81 - Model configuration saved to /home/ubuntu/scgenai/examples/testrun/test1/best_model\n",
      "Epoch 5/20: 100%|██████████| 14/14 [00:11<00:00,  1.19it/s, GPU_ID=3, train_accuracy=0.714, train_loss=0.842]\n",
      "Epoch 5/20: 100%|██████████| 14/14 [00:10<00:00,  1.39it/s, GPU_ID=0, train_accuracy=0.643, train_loss=0.932]\n",
      "\n",
      "Epoch 5/20: 100%|██████████| 14/14 [00:11<00:00,  1.17it/s, GPU_ID=1, train_accuracy=0.786, train_loss=0.676]\n",
      "Evaluating Epoch 5/20: 100%|██████████| 54/54 [00:12<00:00,  4.49it/s]\n",
      "Evaluating Epoch 5/20: 100%|██████████| 54/54 [00:12<00:00,  4.47it/s]\n",
      "2024-10-16 00:47:11.415 | INFO     | scGenAI.training.train:evaluate_and_save_predictions:328 - Epoch 5/20, Train Loss: 0.9318, Train Accuracy: 0.6429, Context-level Val Loss: 1.0872, Context-level Val Accuracy: 0.5000\n",
      "Evaluating Epoch 5/20: 100%|██████████| 54/54 [00:12<00:00,  4.46it/s]\n",
      "Evaluating Epoch 5/20: 100%|██████████| 54/54 [00:12<00:00,  4.38it/s]\n",
      "Epoch 6/20:   0%|          | 0/14 [00:00<?, ?it/s]2024-10-16 00:47:13.300 | INFO     | scGenAI.utils.distributed:savemodel:54 - Model state saved to /home/ubuntu/scgenai/examples/testrun/test1/best_model/scGenAI_model.pt\n",
      "2024-10-16 00:47:13.447 | INFO     | scGenAI.utils.distributed:savemodel:59 - Tokenizer saved to /home/ubuntu/scgenai/examples/testrun/test1/best_model\n",
      "2024-10-16 00:47:13.450 | INFO     | scGenAI.utils.distributed:savemodel:75 - Custom vocab and label encoder saved\n",
      "2024-10-16 00:47:13.451 | INFO     | scGenAI.utils.distributed:savemodel:81 - Model configuration saved to /home/ubuntu/scgenai/examples/testrun/test1/best_model\n",
      "Epoch 6/20: 100%|██████████| 14/14 [00:10<00:00,  1.39it/s, GPU_ID=0, train_accuracy=0.714, train_loss=0.779]\n",
      "Epoch 6/20: 100%|██████████| 14/14 [00:12<00:00,  1.16it/s, GPU_ID=2, train_accuracy=0.786, train_loss=0.789]\n",
      "Epoch 6/20: 100%|██████████| 14/14 [00:11<00:00,  1.19it/s, GPU_ID=3, train_accuracy=0.857, train_loss=0.64]\n",
      "Epoch 6/20: 100%|██████████| 14/14 [00:11<00:00,  1.17it/s, GPU_ID=1, train_accuracy=0.714, train_loss=0.562]\n",
      "Evaluating Epoch 6/20: 100%|██████████| 54/54 [00:12<00:00,  4.47it/s]\n",
      "Evaluating Epoch 6/20: 100%|██████████| 54/54 [00:12<00:00,  4.47it/s]\n",
      "2024-10-16 00:47:35.598 | INFO     | scGenAI.training.train:evaluate_and_save_predictions:328 - Epoch 6/20, Train Loss: 0.7791, Train Accuracy: 0.7143, Context-level Val Loss: 1.1461, Context-level Val Accuracy: 0.5370\n",
      "Evaluating Epoch 6/20: 100%|██████████| 54/54 [00:12<00:00,  4.45it/s]\n",
      "Evaluating Epoch 6/20: 100%|██████████| 54/54 [00:12<00:00,  4.37it/s]\n",
      "Epoch 7/20:   0%|          | 0/14 [00:00<?, ?it/s]2024-10-16 00:47:37.487 | INFO     | scGenAI.utils.distributed:savemodel:54 - Model state saved to /home/ubuntu/scgenai/examples/testrun/test1/best_model/scGenAI_model.pt\n",
      "2024-10-16 00:47:37.637 | INFO     | scGenAI.utils.distributed:savemodel:59 - Tokenizer saved to /home/ubuntu/scgenai/examples/testrun/test1/best_model\n",
      "2024-10-16 00:47:37.640 | INFO     | scGenAI.utils.distributed:savemodel:75 - Custom vocab and label encoder saved\n",
      "2024-10-16 00:47:37.641 | INFO     | scGenAI.utils.distributed:savemodel:81 - Model configuration saved to /home/ubuntu/scgenai/examples/testrun/test1/best_model\n",
      "Epoch 7/20: 100%|██████████| 14/14 [00:11<00:00,  1.19it/s, GPU_ID=3, train_accuracy=0.929, train_loss=0.459]\n",
      "\n",
      "\n",
      "Epoch 7/20: 100%|██████████| 14/14 [00:11<00:00,  1.17it/s, GPU_ID=1, train_accuracy=0.857, train_loss=0.491]\n",
      "Evaluating Epoch 7/20: 100%|██████████| 54/54 [00:12<00:00,  4.49it/s]\n",
      "Evaluating Epoch 7/20: 100%|██████████| 54/54 [00:12<00:00,  4.46it/s]\n",
      "2024-10-16 00:47:59.822 | INFO     | scGenAI.training.train:evaluate_and_save_predictions:328 - Epoch 7/20, Train Loss: 0.5945, Train Accuracy: 0.8571, Context-level Val Loss: 1.1380, Context-level Val Accuracy: 0.5185\n",
      "Evaluating Epoch 7/20: 100%|██████████| 54/54 [00:12<00:00,  4.45it/s]\n",
      "Evaluating Epoch 7/20: 100%|██████████| 54/54 [00:12<00:00,  4.37it/s]\n",
      "Epoch 8/20: 100%|██████████| 14/14 [00:10<00:00,  1.39it/s, GPU_ID=3, train_accuracy=0.929, train_loss=0.318]\n",
      "Epoch 8/20: 100%|██████████| 14/14 [00:10<00:00,  1.36it/s, GPU_ID=0, train_accuracy=0.929, train_loss=0.485]\n",
      "Epoch 8/20: 100%|██████████| 14/14 [00:10<00:00,  1.35it/s, GPU_ID=2, train_accuracy=1, train_loss=0.408]\n",
      "Epoch 8/20: 100%|██████████| 14/14 [00:10<00:00,  1.36it/s, GPU_ID=1, train_accuracy=0.857, train_loss=0.43]\n",
      "Evaluating Epoch 8/20: 100%|██████████| 54/54 [00:12<00:00,  4.49it/s]\n",
      "Evaluating Epoch 8/20: 100%|██████████| 54/54 [00:12<00:00,  4.47it/s]\n",
      "2024-10-16 00:48:22.318 | INFO     | scGenAI.training.train:evaluate_and_save_predictions:328 - Epoch 8/20, Train Loss: 0.4850, Train Accuracy: 0.9286, Context-level Val Loss: 1.0993, Context-level Val Accuracy: 0.5185\n",
      "Evaluating Epoch 8/20: 100%|██████████| 54/54 [00:12<00:00,  4.45it/s]\n",
      "Evaluating Epoch 8/20: 100%|██████████| 54/54 [00:12<00:00,  4.40it/s]\n",
      "Epoch 9/20: 100%|██████████| 14/14 [00:10<00:00,  1.36it/s, GPU_ID=0, train_accuracy=0.929, train_loss=0.439]\n",
      "Epoch 9/20: 100%|██████████| 14/14 [00:10<00:00,  1.39it/s, GPU_ID=3, train_accuracy=1, train_loss=0.266]\n",
      "Epoch 9/20: 100%|██████████| 14/14 [00:10<00:00,  1.35it/s, GPU_ID=2, train_accuracy=0.929, train_loss=0.385]\n",
      "Epoch 9/20: 100%|██████████| 14/14 [00:10<00:00,  1.37it/s, GPU_ID=1, train_accuracy=1, train_loss=0.325]\n",
      "Evaluating Epoch 9/20: 100%|██████████| 54/54 [00:12<00:00,  4.49it/s]\n",
      "2024-10-16 00:48:44.706 | INFO     | scGenAI.training.train:evaluate_and_save_predictions:328 - Epoch 9/20, Train Loss: 0.4390, Train Accuracy: 0.9286, Context-level Val Loss: 1.0138, Context-level Val Accuracy: 0.5370\n",
      "Evaluating Epoch 9/20: 100%|██████████| 54/54 [00:12<00:00,  4.49it/s]\n",
      "Evaluating Epoch 9/20: 100%|██████████| 54/54 [00:12<00:00,  4.46it/s]\n",
      "Evaluating Epoch 9/20: 100%|██████████| 54/54 [00:12<00:00,  4.39it/s]\n",
      "Epoch 10/20:   0%|          | 0/14 [00:00<?, ?it/s]2024-10-16 00:48:46.598 | INFO     | scGenAI.utils.distributed:savemodel:54 - Model state saved to /home/ubuntu/scgenai/examples/testrun/test1/best_model/scGenAI_model.pt\n",
      "2024-10-16 00:48:46.746 | INFO     | scGenAI.utils.distributed:savemodel:59 - Tokenizer saved to /home/ubuntu/scgenai/examples/testrun/test1/best_model\n",
      "2024-10-16 00:48:46.749 | INFO     | scGenAI.utils.distributed:savemodel:75 - Custom vocab and label encoder saved\n",
      "2024-10-16 00:48:46.750 | INFO     | scGenAI.utils.distributed:savemodel:81 - Model configuration saved to /home/ubuntu/scgenai/examples/testrun/test1/best_model\n",
      "Epoch 10/20: 100%|██████████| 14/14 [00:11<00:00,  1.19it/s, GPU_ID=3, train_accuracy=0.857, train_loss=0.481]\n",
      "Epoch 10/20: 100%|██████████| 14/14 [00:12<00:00,  1.17it/s, GPU_ID=2, train_accuracy=0.857, train_loss=0.498]\n",
      "\n",
      "Epoch 10/20: 100%|██████████| 14/14 [00:11<00:00,  1.17it/s, GPU_ID=1, train_accuracy=0.929, train_loss=0.391]\n",
      "Evaluating Epoch 10/20: 100%|██████████| 54/54 [00:12<00:00,  4.49it/s]\n",
      "Evaluating Epoch 10/20: 100%|██████████| 54/54 [00:12<00:00,  4.49it/s]\n",
      "2024-10-16 00:49:08.850 | INFO     | scGenAI.training.train:evaluate_and_save_predictions:328 - Epoch 10/20, Train Loss: 0.4616, Train Accuracy: 0.8571, Context-level Val Loss: 0.9154, Context-level Val Accuracy: 0.5370\n",
      "Evaluating Epoch 10/20: 100%|██████████| 54/54 [00:12<00:00,  4.46it/s]\n",
      "Evaluating Epoch 10/20: 100%|██████████| 54/54 [00:12<00:00,  4.39it/s]\n",
      "Epoch 11/20:   0%|          | 0/14 [00:00<?, ?it/s]2024-10-16 00:49:10.737 | INFO     | scGenAI.utils.distributed:savemodel:54 - Model state saved to /home/ubuntu/scgenai/examples/testrun/test1/best_model/scGenAI_model.pt\n",
      "2024-10-16 00:49:10.888 | INFO     | scGenAI.utils.distributed:savemodel:59 - Tokenizer saved to /home/ubuntu/scgenai/examples/testrun/test1/best_model\n",
      "2024-10-16 00:49:10.896 | INFO     | scGenAI.utils.distributed:savemodel:75 - Custom vocab and label encoder saved\n",
      "2024-10-16 00:49:10.897 | INFO     | scGenAI.utils.distributed:savemodel:81 - Model configuration saved to /home/ubuntu/scgenai/examples/testrun/test1/best_model\n",
      "Epoch 11/20: 100%|██████████| 14/14 [00:12<00:00,  1.16it/s, GPU_ID=2, train_accuracy=0.929, train_loss=0.36] \n",
      "Epoch 11/20: 100%|██████████| 14/14 [00:10<00:00,  1.39it/s, GPU_ID=0, train_accuracy=1, train_loss=0.236]\n",
      "\n",
      "Epoch 11/20: 100%|██████████| 14/14 [00:11<00:00,  1.17it/s, GPU_ID=1, train_accuracy=1, train_loss=0.219]\n",
      "Evaluating Epoch 11/20: 100%|██████████| 54/54 [00:12<00:00,  4.49it/s]\n",
      "Evaluating Epoch 11/20: 100%|██████████| 54/54 [00:12<00:00,  4.49it/s]\n",
      "2024-10-16 00:49:33.010 | INFO     | scGenAI.training.train:evaluate_and_save_predictions:328 - Epoch 11/20, Train Loss: 0.2356, Train Accuracy: 1.0000, Context-level Val Loss: 0.9623, Context-level Val Accuracy: 0.5741\n",
      "Evaluating Epoch 11/20: 100%|██████████| 54/54 [00:12<00:00,  4.46it/s]\n",
      "Evaluating Epoch 11/20: 100%|██████████| 54/54 [00:12<00:00,  4.39it/s]\n",
      "Epoch 12/20:   0%|          | 0/14 [00:00<?, ?it/s]2024-10-16 00:49:34.904 | INFO     | scGenAI.utils.distributed:savemodel:54 - Model state saved to /home/ubuntu/scgenai/examples/testrun/test1/best_model/scGenAI_model.pt\n",
      "2024-10-16 00:49:35.053 | INFO     | scGenAI.utils.distributed:savemodel:59 - Tokenizer saved to /home/ubuntu/scgenai/examples/testrun/test1/best_model\n",
      "2024-10-16 00:49:35.056 | INFO     | scGenAI.utils.distributed:savemodel:75 - Custom vocab and label encoder saved\n",
      "2024-10-16 00:49:35.057 | INFO     | scGenAI.utils.distributed:savemodel:81 - Model configuration saved to /home/ubuntu/scgenai/examples/testrun/test1/best_model\n",
      "Epoch 12/20: 100%|██████████| 14/14 [00:11<00:00,  1.19it/s, GPU_ID=3, train_accuracy=0.929, train_loss=0.172]\n",
      "Epoch 12/20: 100%|██████████| 14/14 [00:10<00:00,  1.39it/s, GPU_ID=0, train_accuracy=1, train_loss=0.139]\n",
      "Epoch 12/20: 100%|██████████| 14/14 [00:12<00:00,  1.16it/s, GPU_ID=2, train_accuracy=1, train_loss=0.15]\n",
      "Epoch 12/20: 100%|██████████| 14/14 [00:11<00:00,  1.17it/s, GPU_ID=1, train_accuracy=1, train_loss=0.129]\n",
      "Evaluating Epoch 12/20: 100%|██████████| 54/54 [00:12<00:00,  4.49it/s]\n",
      "2024-10-16 00:49:57.160 | INFO     | scGenAI.training.train:evaluate_and_save_predictions:328 - Epoch 12/20, Train Loss: 0.1395, Train Accuracy: 1.0000, Context-level Val Loss: 0.9536, Context-level Val Accuracy: 0.5741\n",
      "Evaluating Epoch 12/20: 100%|██████████| 54/54 [00:12<00:00,  4.49it/s]\n",
      "Evaluating Epoch 12/20: 100%|██████████| 54/54 [00:12<00:00,  4.46it/s]\n",
      "Evaluating Epoch 12/20: 100%|██████████| 54/54 [00:12<00:00,  4.39it/s]\n",
      "Epoch 13/20:   0%|          | 0/14 [00:00<?, ?it/s]2024-10-16 00:49:59.052 | INFO     | scGenAI.utils.distributed:savemodel:54 - Model state saved to /home/ubuntu/scgenai/examples/testrun/test1/best_model/scGenAI_model.pt\n",
      "2024-10-16 00:49:59.201 | INFO     | scGenAI.utils.distributed:savemodel:59 - Tokenizer saved to /home/ubuntu/scgenai/examples/testrun/test1/best_model\n",
      "2024-10-16 00:49:59.208 | INFO     | scGenAI.utils.distributed:savemodel:75 - Custom vocab and label encoder saved\n",
      "2024-10-16 00:49:59.209 | INFO     | scGenAI.utils.distributed:savemodel:81 - Model configuration saved to /home/ubuntu/scgenai/examples/testrun/test1/best_model\n",
      "Epoch 13/20: 100%|██████████| 14/14 [00:11<00:00,  1.19it/s, GPU_ID=3, train_accuracy=1, train_loss=0.0741]\n",
      "Epoch 13/20: 100%|██████████| 14/14 [00:10<00:00,  1.39it/s, GPU_ID=0, train_accuracy=1, train_loss=0.104]\n",
      "\n",
      "Epoch 13/20: 100%|██████████| 14/14 [00:11<00:00,  1.17it/s, GPU_ID=1, train_accuracy=1, train_loss=0.096]\n",
      "Evaluating Epoch 13/20: 100%|██████████| 54/54 [00:12<00:00,  4.49it/s]\n",
      "Evaluating Epoch 13/20: 100%|██████████| 54/54 [00:12<00:00,  4.49it/s]\n",
      "2024-10-16 00:50:21.326 | INFO     | scGenAI.training.train:evaluate_and_save_predictions:328 - Epoch 13/20, Train Loss: 0.1040, Train Accuracy: 1.0000, Context-level Val Loss: 0.8470, Context-level Val Accuracy: 0.7037\n",
      "Evaluating Epoch 13/20: 100%|██████████| 54/54 [00:12<00:00,  4.46it/s]\n",
      "Evaluating Epoch 13/20: 100%|██████████| 54/54 [00:12<00:00,  4.38it/s]\n",
      "Epoch 14/20:   0%|          | 0/14 [00:00<?, ?it/s]2024-10-16 00:50:23.221 | INFO     | scGenAI.utils.distributed:savemodel:54 - Model state saved to /home/ubuntu/scgenai/examples/testrun/test1/best_model/scGenAI_model.pt\n",
      "2024-10-16 00:50:23.369 | INFO     | scGenAI.utils.distributed:savemodel:59 - Tokenizer saved to /home/ubuntu/scgenai/examples/testrun/test1/best_model\n",
      "2024-10-16 00:50:23.372 | INFO     | scGenAI.utils.distributed:savemodel:75 - Custom vocab and label encoder saved\n",
      "2024-10-16 00:50:23.373 | INFO     | scGenAI.utils.distributed:savemodel:81 - Model configuration saved to /home/ubuntu/scgenai/examples/testrun/test1/best_model\n",
      "Epoch 14/20: 100%|██████████| 14/14 [00:11<00:00,  1.19it/s, GPU_ID=3, train_accuracy=1, train_loss=0.0756]\n",
      "Epoch 14/20: 100%|██████████| 14/14 [00:10<00:00,  1.39it/s, GPU_ID=0, train_accuracy=1, train_loss=0.0781]\n",
      "Epoch 14/20: 100%|██████████| 14/14 [00:12<00:00,  1.16it/s, GPU_ID=2, train_accuracy=1, train_loss=0.0709]\n",
      "Epoch 14/20: 100%|██████████| 14/14 [00:11<00:00,  1.17it/s, GPU_ID=1, train_accuracy=1, train_loss=0.0622]\n",
      "Evaluating Epoch 14/20: 100%|██████████| 54/54 [00:12<00:00,  4.49it/s]\n",
      "Evaluating Epoch 14/20: 100%|██████████| 54/54 [00:12<00:00,  4.48it/s]\n",
      "2024-10-16 00:50:45.494 | INFO     | scGenAI.training.train:evaluate_and_save_predictions:328 - Epoch 14/20, Train Loss: 0.0781, Train Accuracy: 1.0000, Context-level Val Loss: 0.9477, Context-level Val Accuracy: 0.6481\n",
      "Evaluating Epoch 14/20: 100%|██████████| 54/54 [00:12<00:00,  4.46it/s]\n",
      "Evaluating Epoch 14/20: 100%|██████████| 54/54 [00:12<00:00,  4.39it/s]\n",
      "Epoch 15/20: 100%|██████████| 14/14 [00:10<00:00,  1.39it/s, GPU_ID=3, train_accuracy=1, train_loss=0.0407]\n",
      "Epoch 15/20: 100%|██████████| 14/14 [00:10<00:00,  1.36it/s, GPU_ID=0, train_accuracy=1, train_loss=0.0551]\n",
      "Epoch 15/20: 100%|██████████| 14/14 [00:10<00:00,  1.35it/s, GPU_ID=2, train_accuracy=1, train_loss=0.0436]\n",
      "Epoch 15/20: 100%|██████████| 14/14 [00:10<00:00,  1.36it/s, GPU_ID=1, train_accuracy=1, train_loss=0.0547]\n",
      "Evaluating Epoch 15/20: 100%|██████████| 54/54 [00:12<00:00,  4.48it/s]\n",
      "2024-10-16 00:51:07.963 | INFO     | scGenAI.training.train:evaluate_and_save_predictions:328 - Epoch 15/20, Train Loss: 0.0551, Train Accuracy: 1.0000, Context-level Val Loss: 0.9170, Context-level Val Accuracy: 0.6481\n",
      "Evaluating Epoch 15/20: 100%|██████████| 54/54 [00:12<00:00,  4.48it/s]\n",
      "Evaluating Epoch 15/20: 100%|██████████| 54/54 [00:12<00:00,  4.45it/s]\n",
      "Evaluating Epoch 15/20: 100%|██████████| 54/54 [00:12<00:00,  4.38it/s]\n",
      "Epoch 16/20: 100%|██████████| 14/14 [00:10<00:00,  1.39it/s, GPU_ID=3, train_accuracy=1, train_loss=0.0348]\n",
      "Epoch 16/20: 100%|██████████| 14/14 [00:10<00:00,  1.35it/s, GPU_ID=0, train_accuracy=1, train_loss=0.0405]\n",
      "Epoch 16/20: 100%|██████████| 14/14 [00:10<00:00,  1.35it/s, GPU_ID=2, train_accuracy=1, train_loss=0.033]\n",
      "Epoch 16/20: 100%|██████████| 14/14 [00:10<00:00,  1.36it/s, GPU_ID=1, train_accuracy=1, train_loss=0.0361]\n",
      "Evaluating Epoch 16/20: 100%|██████████| 54/54 [00:12<00:00,  4.49it/s]\n",
      "Evaluating Epoch 16/20: 100%|██████████| 54/54 [00:12<00:00,  4.48it/s]\n",
      "2024-10-16 00:51:30.463 | INFO     | scGenAI.training.train:evaluate_and_save_predictions:328 - Epoch 16/20, Train Loss: 0.0405, Train Accuracy: 1.0000, Context-level Val Loss: 0.9343, Context-level Val Accuracy: 0.7037\n",
      "Evaluating Epoch 16/20: 100%|██████████| 54/54 [00:12<00:00,  4.46it/s]\n",
      "Evaluating Epoch 16/20: 100%|██████████| 54/54 [00:12<00:00,  4.39it/s]\n",
      "Epoch 17/20:   0%|          | 0/14 [00:00<?, ?it/s]2024-10-16 00:51:32.356 | INFO     | scGenAI.utils.distributed:savemodel:54 - Model state saved to /home/ubuntu/scgenai/examples/testrun/test1/best_model/scGenAI_model.pt\n",
      "2024-10-16 00:51:32.505 | INFO     | scGenAI.utils.distributed:savemodel:59 - Tokenizer saved to /home/ubuntu/scgenai/examples/testrun/test1/best_model\n",
      "2024-10-16 00:51:32.508 | INFO     | scGenAI.utils.distributed:savemodel:75 - Custom vocab and label encoder saved\n",
      "2024-10-16 00:51:32.508 | INFO     | scGenAI.utils.distributed:savemodel:81 - Model configuration saved to /home/ubuntu/scgenai/examples/testrun/test1/best_model\n",
      "Epoch 17/20: 100%|██████████| 14/14 [00:11<00:00,  1.19it/s, GPU_ID=3, train_accuracy=1, train_loss=0.0286]\n",
      "Epoch 17/20: 100%|██████████| 14/14 [00:10<00:00,  1.39it/s, GPU_ID=0, train_accuracy=1, train_loss=0.0347]\n",
      "Epoch 17/20: 100%|██████████| 14/14 [00:12<00:00,  1.16it/s, GPU_ID=2, train_accuracy=1, train_loss=0.0287]\n",
      "Epoch 17/20: 100%|██████████| 14/14 [00:11<00:00,  1.17it/s, GPU_ID=1, train_accuracy=1, train_loss=0.0308]\n",
      "Evaluating Epoch 17/20: 100%|██████████| 54/54 [00:12<00:00,  4.49it/s]\n",
      "2024-10-16 00:51:54.631 | INFO     | scGenAI.training.train:evaluate_and_save_predictions:328 - Epoch 17/20, Train Loss: 0.0347, Train Accuracy: 1.0000, Context-level Val Loss: 0.9368, Context-level Val Accuracy: 0.6852\n",
      "Evaluating Epoch 17/20: 100%|██████████| 54/54 [00:12<00:00,  4.49it/s]\n",
      "Evaluating Epoch 17/20: 100%|██████████| 54/54 [00:12<00:00,  4.46it/s]\n",
      "Evaluating Epoch 17/20: 100%|██████████| 54/54 [00:12<00:00,  4.38it/s]\n",
      "Epoch 18/20: 100%|██████████| 14/14 [00:10<00:00,  1.39it/s, GPU_ID=3, train_accuracy=1, train_loss=0.0253]\n",
      "Epoch 18/20: 100%|██████████| 14/14 [00:10<00:00,  1.35it/s, GPU_ID=2, train_accuracy=1, train_loss=0.0249]\n",
      "\n",
      "Epoch 18/20: 100%|██████████| 14/14 [00:10<00:00,  1.36it/s, GPU_ID=1, train_accuracy=1, train_loss=0.0268]\n",
      "Evaluating Epoch 18/20: 100%|██████████| 54/54 [00:12<00:00,  4.48it/s]\n",
      "Evaluating Epoch 18/20: 100%|██████████| 54/54 [00:12<00:00,  4.48it/s]\n",
      "2024-10-16 00:52:17.150 | INFO     | scGenAI.training.train:evaluate_and_save_predictions:328 - Epoch 18/20, Train Loss: 0.0295, Train Accuracy: 1.0000, Context-level Val Loss: 0.9500, Context-level Val Accuracy: 0.6852\n",
      "Evaluating Epoch 18/20: 100%|██████████| 54/54 [00:12<00:00,  4.46it/s]\n",
      "Evaluating Epoch 18/20: 100%|██████████| 54/54 [00:12<00:00,  4.38it/s]\n",
      "Epoch 19/20: 100%|██████████| 14/14 [00:10<00:00,  1.39it/s, GPU_ID=3, train_accuracy=1, train_loss=0.0224] \n",
      "\n",
      "Epoch 19/20: 100%|██████████| 14/14 [00:10<00:00,  1.35it/s, GPU_ID=2, train_accuracy=1, train_loss=0.0224]\n",
      "Epoch 19/20: 100%|██████████| 14/14 [00:10<00:00,  1.36it/s, GPU_ID=1, train_accuracy=1, train_loss=0.0238]\n",
      "Evaluating Epoch 19/20: 100%|██████████| 54/54 [00:12<00:00,  4.48it/s]\n",
      "2024-10-16 00:52:39.649 | INFO     | scGenAI.training.train:evaluate_and_save_predictions:328 - Epoch 19/20, Train Loss: 0.0264, Train Accuracy: 1.0000, Context-level Val Loss: 0.9502, Context-level Val Accuracy: 0.7037\n",
      "Evaluating Epoch 19/20: 100%|██████████| 54/54 [00:12<00:00,  4.48it/s]\n",
      "Evaluating Epoch 19/20: 100%|██████████| 54/54 [00:12<00:00,  4.45it/s]\n",
      "Evaluating Epoch 19/20: 100%|██████████| 54/54 [00:12<00:00,  4.39it/s]\n",
      "Epoch 20/20:   0%|          | 0/14 [00:00<?, ?it/s]2024-10-16 00:52:41.540 | INFO     | scGenAI.utils.distributed:savemodel:54 - Model state saved to /home/ubuntu/scgenai/examples/testrun/test1/best_model/scGenAI_model.pt\n",
      "2024-10-16 00:52:41.689 | INFO     | scGenAI.utils.distributed:savemodel:59 - Tokenizer saved to /home/ubuntu/scgenai/examples/testrun/test1/best_model\n",
      "2024-10-16 00:52:41.695 | INFO     | scGenAI.utils.distributed:savemodel:75 - Custom vocab and label encoder saved\n",
      "2024-10-16 00:52:41.696 | INFO     | scGenAI.utils.distributed:savemodel:81 - Model configuration saved to /home/ubuntu/scgenai/examples/testrun/test1/best_model\n",
      "Epoch 20/20: 100%|██████████| 14/14 [00:11<00:00,  1.19it/s, GPU_ID=3, train_accuracy=1, train_loss=0.0203] \n",
      "Epoch 20/20: 100%|██████████| 14/14 [00:12<00:00,  1.16it/s, GPU_ID=2, train_accuracy=1, train_loss=0.0202]\n",
      "\n",
      "Epoch 20/20: 100%|██████████| 14/14 [00:11<00:00,  1.17it/s, GPU_ID=1, train_accuracy=1, train_loss=0.0212]\n",
      "Evaluating Epoch 20/20: 100%|██████████| 54/54 [00:12<00:00,  4.48it/s]\n",
      "2024-10-16 00:53:03.824 | INFO     | scGenAI.training.train:evaluate_and_save_predictions:328 - Epoch 20/20, Train Loss: 0.0234, Train Accuracy: 1.0000, Context-level Val Loss: 0.9574, Context-level Val Accuracy: 0.7037\n",
      "Evaluating Epoch 20/20: 100%|██████████| 54/54 [00:12<00:00,  4.48it/s]\n",
      "Evaluating Epoch 20/20: 100%|██████████| 54/54 [00:12<00:00,  4.45it/s]\n",
      "Evaluating Epoch 20/20: 100%|██████████| 54/54 [00:12<00:00,  4.39it/s]\n",
      "2024-10-16 00:53:05.712 | INFO     | scGenAI.utils.distributed:savemodel:54 - Model state saved to /home/ubuntu/scgenai/examples/testrun/test1/best_model/scGenAI_model.pt\n",
      "2024-10-16 00:53:05.859 | INFO     | scGenAI.utils.distributed:savemodel:59 - Tokenizer saved to /home/ubuntu/scgenai/examples/testrun/test1/best_model\n",
      "2024-10-16 00:53:05.862 | INFO     | scGenAI.utils.distributed:savemodel:75 - Custom vocab and label encoder saved\n",
      "2024-10-16 00:53:05.863 | INFO     | scGenAI.utils.distributed:savemodel:81 - Model configuration saved to /home/ubuntu/scgenai/examples/testrun/test1/best_model\n",
      "2024-10-16 00:53:06.389 | INFO     | scGenAI.utils.distributed:savemodel:54 - Model state saved to /home/ubuntu/scgenai/examples/testrun/test1/last_model/scGenAI_model.pt\n",
      "2024-10-16 00:53:06.525 | INFO     | scGenAI.utils.distributed:savemodel:59 - Tokenizer saved to /home/ubuntu/scgenai/examples/testrun/test1/last_model\n",
      "2024-10-16 00:53:06.527 | INFO     | scGenAI.utils.distributed:savemodel:75 - Custom vocab and label encoder saved\n",
      "2024-10-16 00:53:06.528 | INFO     | scGenAI.utils.distributed:savemodel:81 - Model configuration saved to /home/ubuntu/scgenai/examples/testrun/test1/last_model\n",
      "2024-10-16 00:53:06.529 | INFO     | scGenAI.utils.distributed:cleanup:33 - Cleaned up DDP environment\n",
      "2024-10-16 00:53:06.529 | INFO     | scGenAI.utils.distributed:cleanup:33 - Cleaned up DDP environment\n",
      "2024-10-16 00:53:06.529 | INFO     | scGenAI.utils.distributed:cleanup:33 - Cleaned up DDP environment\n",
      "2024-10-16 00:53:06.529 | INFO     | scGenAI.utils.distributed:cleanup:33 - Cleaned up DDP environment\n",
      "\u001b[32m2024-10-16 00:53:07.785\u001b[0m | \u001b[1mINFO    \u001b[0m | \u001b[36mscGenAI.data.postprocess\u001b[0m:\u001b[36mmerge_predictions_across_gpus\u001b[0m:\u001b[36m147\u001b[0m - \u001b[1mMerging predictions for epoch 1\u001b[0m\n",
      "\u001b[32m2024-10-16 00:53:07.955\u001b[0m | \u001b[1mINFO    \u001b[0m | \u001b[36mscGenAI.data.postprocess\u001b[0m:\u001b[36mpostprocess_epoch\u001b[0m:\u001b[36m211\u001b[0m - \u001b[1mEpoch 1 - Validation Accuracy: 0.2500\u001b[0m\n",
      "\u001b[32m2024-10-16 00:53:07.965\u001b[0m | \u001b[1mINFO    \u001b[0m | \u001b[36mscGenAI.data.postprocess\u001b[0m:\u001b[36mmerge_predictions_across_gpus\u001b[0m:\u001b[36m147\u001b[0m - \u001b[1mMerging predictions for epoch 2\u001b[0m\n",
      "\u001b[32m2024-10-16 00:53:08.120\u001b[0m | \u001b[1mINFO    \u001b[0m | \u001b[36mscGenAI.data.postprocess\u001b[0m:\u001b[36mpostprocess_epoch\u001b[0m:\u001b[36m211\u001b[0m - \u001b[1mEpoch 2 - Validation Accuracy: 0.2500\u001b[0m\n",
      "\u001b[32m2024-10-16 00:53:08.130\u001b[0m | \u001b[1mINFO    \u001b[0m | \u001b[36mscGenAI.data.postprocess\u001b[0m:\u001b[36mmerge_predictions_across_gpus\u001b[0m:\u001b[36m147\u001b[0m - \u001b[1mMerging predictions for epoch 3\u001b[0m\n",
      "\u001b[32m2024-10-16 00:53:08.291\u001b[0m | \u001b[1mINFO    \u001b[0m | \u001b[36mscGenAI.data.postprocess\u001b[0m:\u001b[36mpostprocess_epoch\u001b[0m:\u001b[36m211\u001b[0m - \u001b[1mEpoch 3 - Validation Accuracy: 0.3000\u001b[0m\n",
      "\u001b[32m2024-10-16 00:53:08.301\u001b[0m | \u001b[1mINFO    \u001b[0m | \u001b[36mscGenAI.data.postprocess\u001b[0m:\u001b[36mmerge_predictions_across_gpus\u001b[0m:\u001b[36m147\u001b[0m - \u001b[1mMerging predictions for epoch 4\u001b[0m\n",
      "\u001b[32m2024-10-16 00:53:08.453\u001b[0m | \u001b[1mINFO    \u001b[0m | \u001b[36mscGenAI.data.postprocess\u001b[0m:\u001b[36mpostprocess_epoch\u001b[0m:\u001b[36m211\u001b[0m - \u001b[1mEpoch 4 - Validation Accuracy: 0.3500\u001b[0m\n",
      "\u001b[32m2024-10-16 00:53:08.463\u001b[0m | \u001b[1mINFO    \u001b[0m | \u001b[36mscGenAI.data.postprocess\u001b[0m:\u001b[36mmerge_predictions_across_gpus\u001b[0m:\u001b[36m147\u001b[0m - \u001b[1mMerging predictions for epoch 5\u001b[0m\n",
      "\u001b[32m2024-10-16 00:53:08.621\u001b[0m | \u001b[1mINFO    \u001b[0m | \u001b[36mscGenAI.data.postprocess\u001b[0m:\u001b[36mpostprocess_epoch\u001b[0m:\u001b[36m211\u001b[0m - \u001b[1mEpoch 5 - Validation Accuracy: 0.4500\u001b[0m\n",
      "\u001b[32m2024-10-16 00:53:08.630\u001b[0m | \u001b[1mINFO    \u001b[0m | \u001b[36mscGenAI.data.postprocess\u001b[0m:\u001b[36mmerge_predictions_across_gpus\u001b[0m:\u001b[36m147\u001b[0m - \u001b[1mMerging predictions for epoch 6\u001b[0m\n",
      "\u001b[32m2024-10-16 00:53:08.781\u001b[0m | \u001b[1mINFO    \u001b[0m | \u001b[36mscGenAI.data.postprocess\u001b[0m:\u001b[36mpostprocess_epoch\u001b[0m:\u001b[36m211\u001b[0m - \u001b[1mEpoch 6 - Validation Accuracy: 0.4000\u001b[0m\n",
      "\u001b[32m2024-10-16 00:53:08.790\u001b[0m | \u001b[1mINFO    \u001b[0m | \u001b[36mscGenAI.data.postprocess\u001b[0m:\u001b[36mmerge_predictions_across_gpus\u001b[0m:\u001b[36m147\u001b[0m - \u001b[1mMerging predictions for epoch 7\u001b[0m\n",
      "\u001b[32m2024-10-16 00:53:08.946\u001b[0m | \u001b[1mINFO    \u001b[0m | \u001b[36mscGenAI.data.postprocess\u001b[0m:\u001b[36mpostprocess_epoch\u001b[0m:\u001b[36m211\u001b[0m - \u001b[1mEpoch 7 - Validation Accuracy: 0.3750\u001b[0m\n",
      "\u001b[32m2024-10-16 00:53:08.955\u001b[0m | \u001b[1mINFO    \u001b[0m | \u001b[36mscGenAI.data.postprocess\u001b[0m:\u001b[36mmerge_predictions_across_gpus\u001b[0m:\u001b[36m147\u001b[0m - \u001b[1mMerging predictions for epoch 8\u001b[0m\n",
      "\u001b[32m2024-10-16 00:53:09.103\u001b[0m | \u001b[1mINFO    \u001b[0m | \u001b[36mscGenAI.data.postprocess\u001b[0m:\u001b[36mpostprocess_epoch\u001b[0m:\u001b[36m211\u001b[0m - \u001b[1mEpoch 8 - Validation Accuracy: 0.4500\u001b[0m\n",
      "\u001b[32m2024-10-16 00:53:09.112\u001b[0m | \u001b[1mINFO    \u001b[0m | \u001b[36mscGenAI.data.postprocess\u001b[0m:\u001b[36mmerge_predictions_across_gpus\u001b[0m:\u001b[36m147\u001b[0m - \u001b[1mMerging predictions for epoch 9\u001b[0m\n",
      "\u001b[32m2024-10-16 00:53:09.267\u001b[0m | \u001b[1mINFO    \u001b[0m | \u001b[36mscGenAI.data.postprocess\u001b[0m:\u001b[36mpostprocess_epoch\u001b[0m:\u001b[36m211\u001b[0m - \u001b[1mEpoch 9 - Validation Accuracy: 0.5250\u001b[0m\n",
      "\u001b[32m2024-10-16 00:53:09.276\u001b[0m | \u001b[1mINFO    \u001b[0m | \u001b[36mscGenAI.data.postprocess\u001b[0m:\u001b[36mmerge_predictions_across_gpus\u001b[0m:\u001b[36m147\u001b[0m - \u001b[1mMerging predictions for epoch 10\u001b[0m\n",
      "\u001b[32m2024-10-16 00:53:09.423\u001b[0m | \u001b[1mINFO    \u001b[0m | \u001b[36mscGenAI.data.postprocess\u001b[0m:\u001b[36mpostprocess_epoch\u001b[0m:\u001b[36m211\u001b[0m - \u001b[1mEpoch 10 - Validation Accuracy: 0.5000\u001b[0m\n",
      "\u001b[32m2024-10-16 00:53:09.432\u001b[0m | \u001b[1mINFO    \u001b[0m | \u001b[36mscGenAI.data.postprocess\u001b[0m:\u001b[36mmerge_predictions_across_gpus\u001b[0m:\u001b[36m147\u001b[0m - \u001b[1mMerging predictions for epoch 11\u001b[0m\n",
      "\u001b[32m2024-10-16 00:53:09.585\u001b[0m | \u001b[1mINFO    \u001b[0m | \u001b[36mscGenAI.data.postprocess\u001b[0m:\u001b[36mpostprocess_epoch\u001b[0m:\u001b[36m211\u001b[0m - \u001b[1mEpoch 11 - Validation Accuracy: 0.5750\u001b[0m\n",
      "\u001b[32m2024-10-16 00:53:09.594\u001b[0m | \u001b[1mINFO    \u001b[0m | \u001b[36mscGenAI.data.postprocess\u001b[0m:\u001b[36mmerge_predictions_across_gpus\u001b[0m:\u001b[36m147\u001b[0m - \u001b[1mMerging predictions for epoch 12\u001b[0m\n",
      "\u001b[32m2024-10-16 00:53:09.740\u001b[0m | \u001b[1mINFO    \u001b[0m | \u001b[36mscGenAI.data.postprocess\u001b[0m:\u001b[36mpostprocess_epoch\u001b[0m:\u001b[36m211\u001b[0m - \u001b[1mEpoch 12 - Validation Accuracy: 0.5500\u001b[0m\n",
      "\u001b[32m2024-10-16 00:53:09.749\u001b[0m | \u001b[1mINFO    \u001b[0m | \u001b[36mscGenAI.data.postprocess\u001b[0m:\u001b[36mmerge_predictions_across_gpus\u001b[0m:\u001b[36m147\u001b[0m - \u001b[1mMerging predictions for epoch 13\u001b[0m\n",
      "\u001b[32m2024-10-16 00:53:09.902\u001b[0m | \u001b[1mINFO    \u001b[0m | \u001b[36mscGenAI.data.postprocess\u001b[0m:\u001b[36mpostprocess_epoch\u001b[0m:\u001b[36m211\u001b[0m - \u001b[1mEpoch 13 - Validation Accuracy: 0.6000\u001b[0m\n",
      "\u001b[32m2024-10-16 00:53:09.910\u001b[0m | \u001b[1mINFO    \u001b[0m | \u001b[36mscGenAI.data.postprocess\u001b[0m:\u001b[36mmerge_predictions_across_gpus\u001b[0m:\u001b[36m147\u001b[0m - \u001b[1mMerging predictions for epoch 14\u001b[0m\n",
      "\u001b[32m2024-10-16 00:53:10.056\u001b[0m | \u001b[1mINFO    \u001b[0m | \u001b[36mscGenAI.data.postprocess\u001b[0m:\u001b[36mpostprocess_epoch\u001b[0m:\u001b[36m211\u001b[0m - \u001b[1mEpoch 14 - Validation Accuracy: 0.6000\u001b[0m\n",
      "\u001b[32m2024-10-16 00:53:10.064\u001b[0m | \u001b[1mINFO    \u001b[0m | \u001b[36mscGenAI.data.postprocess\u001b[0m:\u001b[36mmerge_predictions_across_gpus\u001b[0m:\u001b[36m147\u001b[0m - \u001b[1mMerging predictions for epoch 15\u001b[0m\n",
      "\u001b[32m2024-10-16 00:53:10.217\u001b[0m | \u001b[1mINFO    \u001b[0m | \u001b[36mscGenAI.data.postprocess\u001b[0m:\u001b[36mpostprocess_epoch\u001b[0m:\u001b[36m211\u001b[0m - \u001b[1mEpoch 15 - Validation Accuracy: 0.5500\u001b[0m\n",
      "\u001b[32m2024-10-16 00:53:10.225\u001b[0m | \u001b[1mINFO    \u001b[0m | \u001b[36mscGenAI.data.postprocess\u001b[0m:\u001b[36mmerge_predictions_across_gpus\u001b[0m:\u001b[36m147\u001b[0m - \u001b[1mMerging predictions for epoch 16\u001b[0m\n",
      "\u001b[32m2024-10-16 00:53:10.370\u001b[0m | \u001b[1mINFO    \u001b[0m | \u001b[36mscGenAI.data.postprocess\u001b[0m:\u001b[36mpostprocess_epoch\u001b[0m:\u001b[36m211\u001b[0m - \u001b[1mEpoch 16 - Validation Accuracy: 0.5250\u001b[0m\n",
      "\u001b[32m2024-10-16 00:53:10.379\u001b[0m | \u001b[1mINFO    \u001b[0m | \u001b[36mscGenAI.data.postprocess\u001b[0m:\u001b[36mmerge_predictions_across_gpus\u001b[0m:\u001b[36m147\u001b[0m - \u001b[1mMerging predictions for epoch 17\u001b[0m\n",
      "\u001b[32m2024-10-16 00:53:10.523\u001b[0m | \u001b[1mINFO    \u001b[0m | \u001b[36mscGenAI.data.postprocess\u001b[0m:\u001b[36mpostprocess_epoch\u001b[0m:\u001b[36m211\u001b[0m - \u001b[1mEpoch 17 - Validation Accuracy: 0.5250\u001b[0m\n",
      "\u001b[32m2024-10-16 00:53:10.532\u001b[0m | \u001b[1mINFO    \u001b[0m | \u001b[36mscGenAI.data.postprocess\u001b[0m:\u001b[36mmerge_predictions_across_gpus\u001b[0m:\u001b[36m147\u001b[0m - \u001b[1mMerging predictions for epoch 18\u001b[0m\n",
      "\u001b[32m2024-10-16 00:53:10.684\u001b[0m | \u001b[1mINFO    \u001b[0m | \u001b[36mscGenAI.data.postprocess\u001b[0m:\u001b[36mpostprocess_epoch\u001b[0m:\u001b[36m211\u001b[0m - \u001b[1mEpoch 18 - Validation Accuracy: 0.5250\u001b[0m\n",
      "\u001b[32m2024-10-16 00:53:10.692\u001b[0m | \u001b[1mINFO    \u001b[0m | \u001b[36mscGenAI.data.postprocess\u001b[0m:\u001b[36mmerge_predictions_across_gpus\u001b[0m:\u001b[36m147\u001b[0m - \u001b[1mMerging predictions for epoch 19\u001b[0m\n",
      "\u001b[32m2024-10-16 00:53:10.837\u001b[0m | \u001b[1mINFO    \u001b[0m | \u001b[36mscGenAI.data.postprocess\u001b[0m:\u001b[36mpostprocess_epoch\u001b[0m:\u001b[36m211\u001b[0m - \u001b[1mEpoch 19 - Validation Accuracy: 0.5250\u001b[0m\n",
      "\u001b[32m2024-10-16 00:53:10.846\u001b[0m | \u001b[1mINFO    \u001b[0m | \u001b[36mscGenAI.data.postprocess\u001b[0m:\u001b[36mmerge_predictions_across_gpus\u001b[0m:\u001b[36m147\u001b[0m - \u001b[1mMerging predictions for epoch 20\u001b[0m\n",
      "\u001b[32m2024-10-16 00:53:10.997\u001b[0m | \u001b[1mINFO    \u001b[0m | \u001b[36mscGenAI.data.postprocess\u001b[0m:\u001b[36mpostprocess_epoch\u001b[0m:\u001b[36m211\u001b[0m - \u001b[1mEpoch 20 - Validation Accuracy: 0.5250\u001b[0m\n",
      "\u001b[32m2024-10-16 00:53:11.007\u001b[0m | \u001b[1mINFO    \u001b[0m | \u001b[36mscGenAI.data.postprocess\u001b[0m:\u001b[36msummarize_epochs\u001b[0m:\u001b[36m234\u001b[0m - \u001b[1mCompleted post-processing and summary of all epochs.\u001b[0m\n",
      "\u001b[32m2024-10-16 00:53:11.010\u001b[0m | \u001b[1mINFO    \u001b[0m | \u001b[36mscGenAI.data.postprocess\u001b[0m:\u001b[36mcombine_epoch_results\u001b[0m:\u001b[36m61\u001b[0m - \u001b[1mCombined results saved to /home/ubuntu/scgenai/examples/testrun/test1/combined_epoch_results.csv.\u001b[0m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Plot saved to /home/ubuntu/scgenai/examples/testrun/test1/train_summary.pdf\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "config_file = '/home/ubuntu/scgenai/tutorials/yaml/config_Train_MultiOmicsData_template.yaml'\n",
    "run_training_from_config(config_file)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2fe550d5-16c0-425a-bfd7-d257c1ecfe1d",
   "metadata": {},
   "source": [
    "## Post-Training Model and Results\n",
    "\n",
    "Upon completion of the training, the trained model will be saved in the `model_dir` as defined in the configuration file. Within this directory, two subfolders will be created: `best_model` and `last_model`. The `best_model` folder contains the model that achieved the highest training performance, while the `last_model` contains the model from the final epoch. Additionally, a `combined_epoch_results.csv` file will be generated in the `model_dir`, summarizing the training performance metrics across all epochs."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cc88e128-eaba-4557-8fca-eae0966e482d",
   "metadata": {},
   "source": [
    "### Training Summary\n",
    "\n",
    "In addition to the output summary files generated after training, the training summary can be directly accessed in Python by loading the saved configuration file.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "9cd73cce-7ddc-47ab-b59f-fc4ac955b57f",
   "metadata": {},
   "outputs": [],
   "source": [
    "from scGenAI.config import Config\n",
    "from scGenAI.data.postprocess import TrainSummary"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "9bc20420-8143-4ff7-b474-d4a3f6936991",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Epoch</th>\n",
       "      <th>Validation Accuracy</th>\n",
       "      <th>Train Loss</th>\n",
       "      <th>Train Accuracy</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>0.250</td>\n",
       "      <td>1.386040</td>\n",
       "      <td>0.357143</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>0.250</td>\n",
       "      <td>1.143852</td>\n",
       "      <td>0.500000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>0.300</td>\n",
       "      <td>1.092779</td>\n",
       "      <td>0.500000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>0.350</td>\n",
       "      <td>1.023429</td>\n",
       "      <td>0.571429</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>0.450</td>\n",
       "      <td>0.931778</td>\n",
       "      <td>0.642857</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>6</td>\n",
       "      <td>0.400</td>\n",
       "      <td>0.779091</td>\n",
       "      <td>0.714286</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>7</td>\n",
       "      <td>0.375</td>\n",
       "      <td>0.594540</td>\n",
       "      <td>0.857143</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>8</td>\n",
       "      <td>0.450</td>\n",
       "      <td>0.485034</td>\n",
       "      <td>0.928571</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>9</td>\n",
       "      <td>0.525</td>\n",
       "      <td>0.438986</td>\n",
       "      <td>0.928571</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>10</td>\n",
       "      <td>0.500</td>\n",
       "      <td>0.461559</td>\n",
       "      <td>0.857143</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>11</td>\n",
       "      <td>0.575</td>\n",
       "      <td>0.235648</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>12</td>\n",
       "      <td>0.550</td>\n",
       "      <td>0.139485</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>13</td>\n",
       "      <td>0.600</td>\n",
       "      <td>0.104040</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>14</td>\n",
       "      <td>0.600</td>\n",
       "      <td>0.078141</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>15</td>\n",
       "      <td>0.550</td>\n",
       "      <td>0.055065</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>16</td>\n",
       "      <td>0.525</td>\n",
       "      <td>0.040509</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>17</td>\n",
       "      <td>0.525</td>\n",
       "      <td>0.034723</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>18</td>\n",
       "      <td>0.525</td>\n",
       "      <td>0.029510</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>19</td>\n",
       "      <td>0.525</td>\n",
       "      <td>0.026404</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>20</td>\n",
       "      <td>0.525</td>\n",
       "      <td>0.023411</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    Epoch  Validation Accuracy  Train Loss  Train Accuracy\n",
       "0       1                0.250    1.386040        0.357143\n",
       "1       2                0.250    1.143852        0.500000\n",
       "2       3                0.300    1.092779        0.500000\n",
       "3       4                0.350    1.023429        0.571429\n",
       "4       5                0.450    0.931778        0.642857\n",
       "5       6                0.400    0.779091        0.714286\n",
       "6       7                0.375    0.594540        0.857143\n",
       "7       8                0.450    0.485034        0.928571\n",
       "8       9                0.525    0.438986        0.928571\n",
       "9      10                0.500    0.461559        0.857143\n",
       "10     11                0.575    0.235648        1.000000\n",
       "11     12                0.550    0.139485        1.000000\n",
       "12     13                0.600    0.104040        1.000000\n",
       "13     14                0.600    0.078141        1.000000\n",
       "14     15                0.550    0.055065        1.000000\n",
       "15     16                0.525    0.040509        1.000000\n",
       "16     17                0.525    0.034723        1.000000\n",
       "17     18                0.525    0.029510        1.000000\n",
       "18     19                0.525    0.026404        1.000000\n",
       "19     20                0.525    0.023411        1.000000"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "config = Config(config_file)\n",
    "summary = TrainSummary(config.model_dir)\n",
    "dfsummary = summary.table()\n",
    "dfsummary"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a0ba9299-9ce7-4625-8f21-c49a6e3fd75a",
   "metadata": {},
   "source": [
    "It can also be plotted using the `summary.plot`:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "24962a8b-d8b7-4b38-b982-10633416b86e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Plot saved to /home/ubuntu/scgenai/examples/testrun/test1/train_summary.pdf\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "summary.plot(saveplot=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "346b4f66-b17c-462e-947f-0a3d08972cb4",
   "metadata": {},
   "source": [
    "## Prediction Using the Trained Model\n",
    "\n",
    "Similar to the training step, predictions can be run using the `run_prediction_from_config` function in scGenAI with a simple configuration file. Unlike the training configuration, the prediction configuration file only requires the following parameters to be defined: `mode`, `predict_file`, `batch_size`, `model_dir`, and `outputfile`. All other parameters will be adopted from the trained model.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "10cd73c8-8ea4-44cf-a2c5-6b2f4aecd53b",
   "metadata": {},
   "outputs": [],
   "source": [
    "from scGenAI.prediction.predict import run_prediction_from_config"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "6d8cab06-6cec-49b1-8fa8-53aa426c38df",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "model_dir: /home/ubuntu/scgenai/examples/testrun/test1/best_model/\n",
      "predict_file: /home/ubuntu/scgenai/examples/data/example_val_RNA.h5ad\n",
      "predict_ADTfile: /home/ubuntu/scgenai/examples/data/example_val_ADT.h5ad\n",
      "log_dir: /home/ubuntu/scgenai/examples/logs\n",
      "output_dir: /home/ubuntu/scgenai/examples/output\n",
      "seed: 1314521\n",
      "mode: Predict\n",
      "min_cells: 2\n",
      "multiomics: Yes\n",
      "keepIntermediateFiles: No\n",
      "savelog: No\n",
      "savesetting: Yes\n",
      "model_backbone_name: llama\n",
      "model_backbone_size: small\n",
      "context_method: random\n",
      "num_bins: 10\n",
      "optimizer: AdamW\n",
      "weight_decay: 0.01\n",
      "depth: 2\n",
      "max_length: 5120\n",
      "batch_size: 1\n",
      "learning_rate: 1e-05\n",
      "num_epochs: 30\n",
      "world_size: 1\n",
      "evaluate_during_training: False\n",
      "hidden_size: 768\n",
      "num_layers: 12\n",
      "num_heads: 12\n",
      "intermediate_size: 1024\n",
      "stride: 2560\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2024-10-16 01:02:00.156 | INFO     | scGenAI.utils.distributed:setup_distributed:26 - Distributed setup complete for GPU 0, Total GPUs used: 4\n",
      "2024-10-16 01:02:00.159 | INFO     | scGenAI.utils.distributed:setup_distributed:26 - Distributed setup complete for GPU 3, Total GPUs used: 4\n",
      "2024-10-16 01:02:00.164 | INFO     | scGenAI.utils.distributed:setup_distributed:26 - Distributed setup complete for GPU 1, Total GPUs used: 4\n",
      "2024-10-16 01:02:00.182 | INFO     | scGenAI.utils.distributed:setup_distributed:26 - Distributed setup complete for GPU 2, Total GPUs used: 4\n",
      "GPU 1 Predicting: 100%|██████████| 14/14 [00:03<00:00,  3.90it/s]\n",
      "GPU 0 Predicting: 100%|██████████| 14/14 [00:03<00:00,  3.89it/s]\n",
      "GPU 2 Predicting: 100%|██████████| 14/14 [00:03<00:00,  3.88it/s]\n",
      "GPU 3 Predicting: 100%|██████████| 14/14 [00:03<00:00,  3.84it/s]\n",
      "2024-10-16 01:02:06.984 | INFO     | scGenAI.utils.distributed:cleanup:33 - Cleaned up DDP environment\n",
      "2024-10-16 01:02:06.984 | INFO     | scGenAI.utils.distributed:cleanup:33 - Cleaned up DDP environment\n",
      "2024-10-16 01:02:06.984 | INFO     | scGenAI.utils.distributed:cleanup:33 - Cleaned up DDP environment\n",
      "2024-10-16 01:02:06.984 | INFO     | scGenAI.utils.distributed:cleanup:33 - Cleaned up DDP environment\n",
      "\u001b[32m2024-10-16 01:02:08.240\u001b[0m | \u001b[1mINFO    \u001b[0m | \u001b[36mscGenAI.data.postprocess\u001b[0m:\u001b[36mcombine_prediction_results\u001b[0m:\u001b[36m103\u001b[0m - \u001b[1mCompleted post-processing for prediction\u001b[0m\n"
     ]
    }
   ],
   "source": [
    "pred_config_file = '/home/ubuntu/scgenai/tutorials/yaml/config_prediction_multiOmic_template.yaml'\n",
    "run_prediction_from_config(pred_config_file)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7fa3e6fc-55c8-4ef4-b64e-f6347b84fbb4",
   "metadata": {},
   "source": [
    "The prediction output is a CSV file, as defined in the configuration file. It contains the original metadata extracted from the input prediction file (`obs` slot) along with three additional columns: `context_id`, `PredictedFeature`, and `prediction_score`. \n",
    "\n",
    "- `context_id` represents the context used to determine the prediction for the corresponding cell. \n",
    "- `PredictedFeature` is the final predicted feature for the cell using the trained model. \n",
    "- `prediction_score` indicates the confidence level of the prediction, with a maximum value of 1.\n",
    "\n",
    "For example in this tutorials:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "b4c89ab7-232d-41a1-8cb4-f59851b5fe6d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>index</th>\n",
       "      <th>nCount_ADT</th>\n",
       "      <th>nFeature_ADT</th>\n",
       "      <th>nCount_RNA</th>\n",
       "      <th>nFeature_RNA</th>\n",
       "      <th>orig.ident</th>\n",
       "      <th>lane</th>\n",
       "      <th>donor</th>\n",
       "      <th>time</th>\n",
       "      <th>celltype.l1</th>\n",
       "      <th>celltype.l2</th>\n",
       "      <th>celltype.l3</th>\n",
       "      <th>Phase</th>\n",
       "      <th>nCount_SCT</th>\n",
       "      <th>nFeature_SCT</th>\n",
       "      <th>X_index</th>\n",
       "      <th>cell_barcode</th>\n",
       "      <th>context_id</th>\n",
       "      <th>PredictedFeature</th>\n",
       "      <th>prediction_score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>L1_TTCGATTCACGCAAAG</td>\n",
       "      <td>3699.0</td>\n",
       "      <td>208</td>\n",
       "      <td>4160.0</td>\n",
       "      <td>1222</td>\n",
       "      <td>P1_3</td>\n",
       "      <td>L1</td>\n",
       "      <td>P1</td>\n",
       "      <td>3</td>\n",
       "      <td>CD4 T</td>\n",
       "      <td>CD4 Naive</td>\n",
       "      <td>CD4 Naive</td>\n",
       "      <td>S</td>\n",
       "      <td>4716.0</td>\n",
       "      <td>1221</td>\n",
       "      <td>L1_TTCGATTCACGCAAAG</td>\n",
       "      <td>L1_TTCGATTCACGCAAAG</td>\n",
       "      <td>2_L1_TTCGATTCACGCAAAG_window_0</td>\n",
       "      <td>CD4 TCM</td>\n",
       "      <td>0.744267</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>L1_TTGCATTCACGCTGTG</td>\n",
       "      <td>5065.0</td>\n",
       "      <td>221</td>\n",
       "      <td>4309.0</td>\n",
       "      <td>1299</td>\n",
       "      <td>P3_3</td>\n",
       "      <td>L1</td>\n",
       "      <td>P3</td>\n",
       "      <td>3</td>\n",
       "      <td>CD4 T</td>\n",
       "      <td>CD4 Naive</td>\n",
       "      <td>CD4 Naive</td>\n",
       "      <td>S</td>\n",
       "      <td>4999.0</td>\n",
       "      <td>1298</td>\n",
       "      <td>L1_TTGCATTCACGCTGTG</td>\n",
       "      <td>L1_TTGCATTCACGCTGTG</td>\n",
       "      <td>1_L1_TTGCATTCACGCTGTG_window_1</td>\n",
       "      <td>CD4 Naive</td>\n",
       "      <td>0.548168</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>L1_ACGCACGAGTACGTCT</td>\n",
       "      <td>4719.0</td>\n",
       "      <td>206</td>\n",
       "      <td>4091.0</td>\n",
       "      <td>1161</td>\n",
       "      <td>P4_7</td>\n",
       "      <td>L1</td>\n",
       "      <td>P4</td>\n",
       "      <td>7</td>\n",
       "      <td>CD4 T</td>\n",
       "      <td>CD4 Naive</td>\n",
       "      <td>CD4 Naive</td>\n",
       "      <td>G1</td>\n",
       "      <td>4917.0</td>\n",
       "      <td>1159</td>\n",
       "      <td>L1_ACGCACGAGTACGTCT</td>\n",
       "      <td>L1_ACGCACGAGTACGTCT</td>\n",
       "      <td>0_L1_ACGCACGAGTACGTCT_window_2</td>\n",
       "      <td>CD8 Naive</td>\n",
       "      <td>0.820172</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>L1_TAATTCCGTAGAATGT</td>\n",
       "      <td>3702.0</td>\n",
       "      <td>209</td>\n",
       "      <td>5451.0</td>\n",
       "      <td>1447</td>\n",
       "      <td>P1_3</td>\n",
       "      <td>L1</td>\n",
       "      <td>P1</td>\n",
       "      <td>3</td>\n",
       "      <td>CD4 T</td>\n",
       "      <td>CD4 Naive</td>\n",
       "      <td>CD4 Naive</td>\n",
       "      <td>S</td>\n",
       "      <td>5332.0</td>\n",
       "      <td>1446</td>\n",
       "      <td>L1_TAATTCCGTAGAATGT</td>\n",
       "      <td>L1_TAATTCCGTAGAATGT</td>\n",
       "      <td>2_L1_TAATTCCGTAGAATGT_window_3</td>\n",
       "      <td>CD4 TCM</td>\n",
       "      <td>0.712681</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>L1_CCCAACTTCTCCTACG</td>\n",
       "      <td>5676.0</td>\n",
       "      <td>215</td>\n",
       "      <td>5087.0</td>\n",
       "      <td>1500</td>\n",
       "      <td>P2_3</td>\n",
       "      <td>L1</td>\n",
       "      <td>P2</td>\n",
       "      <td>3</td>\n",
       "      <td>CD4 T</td>\n",
       "      <td>CD4 Naive</td>\n",
       "      <td>CD4 Naive</td>\n",
       "      <td>G1</td>\n",
       "      <td>5123.0</td>\n",
       "      <td>1498</td>\n",
       "      <td>L1_CCCAACTTCTCCTACG</td>\n",
       "      <td>L1_CCCAACTTCTCCTACG</td>\n",
       "      <td>0_L1_CCCAACTTCTCCTACG_window_4</td>\n",
       "      <td>CD4 TCM</td>\n",
       "      <td>0.928588</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 index  nCount_ADT  nFeature_ADT  nCount_RNA  nFeature_RNA   \n",
       "0  L1_TTCGATTCACGCAAAG      3699.0           208      4160.0          1222  \\\n",
       "1  L1_TTGCATTCACGCTGTG      5065.0           221      4309.0          1299   \n",
       "2  L1_ACGCACGAGTACGTCT      4719.0           206      4091.0          1161   \n",
       "3  L1_TAATTCCGTAGAATGT      3702.0           209      5451.0          1447   \n",
       "4  L1_CCCAACTTCTCCTACG      5676.0           215      5087.0          1500   \n",
       "\n",
       "  orig.ident lane donor  time celltype.l1 celltype.l2 celltype.l3 Phase   \n",
       "0       P1_3   L1    P1     3       CD4 T   CD4 Naive   CD4 Naive     S  \\\n",
       "1       P3_3   L1    P3     3       CD4 T   CD4 Naive   CD4 Naive     S   \n",
       "2       P4_7   L1    P4     7       CD4 T   CD4 Naive   CD4 Naive    G1   \n",
       "3       P1_3   L1    P1     3       CD4 T   CD4 Naive   CD4 Naive     S   \n",
       "4       P2_3   L1    P2     3       CD4 T   CD4 Naive   CD4 Naive    G1   \n",
       "\n",
       "   nCount_SCT  nFeature_SCT              X_index         cell_barcode   \n",
       "0      4716.0          1221  L1_TTCGATTCACGCAAAG  L1_TTCGATTCACGCAAAG  \\\n",
       "1      4999.0          1298  L1_TTGCATTCACGCTGTG  L1_TTGCATTCACGCTGTG   \n",
       "2      4917.0          1159  L1_ACGCACGAGTACGTCT  L1_ACGCACGAGTACGTCT   \n",
       "3      5332.0          1446  L1_TAATTCCGTAGAATGT  L1_TAATTCCGTAGAATGT   \n",
       "4      5123.0          1498  L1_CCCAACTTCTCCTACG  L1_CCCAACTTCTCCTACG   \n",
       "\n",
       "                       context_id PredictedFeature  prediction_score  \n",
       "0  2_L1_TTCGATTCACGCAAAG_window_0          CD4 TCM          0.744267  \n",
       "1  1_L1_TTGCATTCACGCTGTG_window_1        CD4 Naive          0.548168  \n",
       "2  0_L1_ACGCACGAGTACGTCT_window_2        CD8 Naive          0.820172  \n",
       "3  2_L1_TAATTCCGTAGAATGT_window_3          CD4 TCM          0.712681  \n",
       "4  0_L1_CCCAACTTCTCCTACG_window_4          CD4 TCM          0.928588  "
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "preconfig = Config(pred_config_file)\n",
    "dfprediction = pd.read_csv(preconfig.outputfile)\n",
    "dfprediction.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "0cd6dab7-bb05-4204-bc26-b730ad4461a4",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}