training_l2r_log_2025-05-29_19-25-52.log

2025-05-29 19:25:52,028 - INFO - 📝 Logging initialized. Log file created at: ../tmp/logs/training_l2r_log_2025-05-29_19-25-52.log - [learning2rank.py:264:setup_logging]
2025-05-29 19:25:52,028 - INFO - ================================================================================ - [learning2rank.py:108:log_section]
2025-05-29 19:25:52,028 - INFO - =                      📌 INITIALIZING TRAINING ENVIRONMENT                       = - [learning2rank.py:109:log_section]
2025-05-29 19:25:52,028 - INFO - ================================================================================ - [learning2rank.py:112:log_section]
2025-05-29 19:25:52,028 - INFO - 🚀 Setting up data paths and environment variables... - [learning2rank.py:3841:main]
2025-05-29 19:25:52,028 - INFO - 🛠️ Command-line Arguments: - [learning2rank.py:280:print_args]
2025-05-29 19:25:52,029 - INFO - 
  🔹 output_dir: ../tmp/MIMIC4_DEMO/mistral7b_mimic4_l2r
  🔹 source_url: XURLs.MIMIC4_DEMO
  🔹 data: mimic4_icd10_full
  🔹 data_l2r_fname_prefix: mimic4_icd10
  🔹 logfile: training_l2r_log
  🔹 base_dir: ../tmp/MIMIC4_DEMO
  🔹 l2r_boot_dir: ../tmp/MIMIC4_DEMO/mimic4_l2rboot_mistral7b
  🔹 hub_model_id: deb101/mistral-7b-instruct-v0.3-mimic4-adapt
  🔹 model_name: mistralai/Mistral-7B-Instruct-v0.3
  🔹 max_length: 512
  🔹 do_fresh_training: True
  🔹 load_from_checkpoint: False
  🔹 task: l2r
  🔹 num_train_epochs: 4
  🔹 metric_for_best_model: ndcg@25
  🔹 learning_rate: 0.0001
  🔹 warmup_steps: 0
  🔹 generate_report: False
  🔹 logfile_path: ../tmp/logs/training_l2r_log_2025-05-29_19-25-52.log
  🔹 source: /home/ubuntu/.xcube/data/mimic4_demo - [learning2rank.py:281:print_args]
2025-05-29 19:25:52,029 - INFO - ➖➖➖➖➖➖➖➖➖➖➖➖➖➖➖➖➖➖➖➖➖➖➖➖➖➖➖➖➖➖ - [learning2rank.py:282:print_args]
2025-05-29 19:25:52,029 - INFO - 📁 Using L2R boot directory: ../tmp/MIMIC4_DEMO/mimic4_l2rboot_mistral7b - [learning2rank.py:3855:main]
2025-05-29 19:25:52,029 - INFO - 📂 Using output directory: ../tmp/MIMIC4_DEMO/mistral7b_mimic4_l2r - [learning2rank.py:3857:main]
2025-05-29 19:25:52,029 - INFO - ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ - [learning2rank.py:108:log_section]
2025-05-29 19:25:52,029 - INFO - +                               ✨ LOADING DATASETS                               + - [learning2rank.py:109:log_section]
2025-05-29 19:25:52,029 - INFO - ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ - [learning2rank.py:112:log_section]
2025-05-29 19:25:52,029 - INFO - 📊 Loading main dataset and L2R dataset... - [learning2rank.py:3860:main]
2025-05-29 19:25:52,029 - INFO - 📂 Loading main data from: /home/ubuntu/.xcube/data/mimic4_demo/mimic4_icd10_full.csv - [learning2rank.py:310:get_data]
2025-05-29 19:26:00,147 - INFO - ✅ Successfully loaded main data: 122279 rows - [learning2rank.py:323:get_data]
2025-05-29 19:26:00,147 - INFO - 📂 Loading L2R data from: ../tmp/MIMIC4_DEMO/mimic4_l2rboot_mistral7b/mimic4_icd10_tok_rank_per_lbl.ft - [learning2rank.py:329:get_data]
2025-05-29 19:26:00,147 - INFO - 📂 Loading L2R tokens from: ../tmp/MIMIC4_DEMO/mimic4_l2rboot_mistral7b/mimic4_icd10_tok.ft - [learning2rank.py:332:get_data]
2025-05-29 19:26:00,147 - INFO - 📂 Loading L2R labels from: ../tmp/MIMIC4_DEMO/mimic4_l2rboot_mistral7b/mimic4_icd10_lbl.ft - [learning2rank.py:335:get_data]
2025-05-29 19:26:01,853 - INFO - ✅ Successfully loaded L2R data: 260243456 rows - [learning2rank.py:342:get_data]
2025-05-29 19:26:01,853 - INFO - ✅ Successfully loaded L2R tokens: 32768 tokens - [learning2rank.py:343:get_data]
2025-05-29 19:26:01,853 - INFO - ✅ Successfully loaded L2R labels: 7942 rows - [learning2rank.py:344:get_data]
2025-05-29 19:26:01,853 - INFO - 🔄 Total data loaded: 122279 main rows, 260243456 L2R rows - [learning2rank.py:351:get_data]
2025-05-29 19:26:01,853 - INFO - ✨ Successfully loaded both datasets: - [learning2rank.py:375:load_datasets]
2025-05-29 19:26:01,853 - INFO -    - 📄 Main dataset: 122279 records - [learning2rank.py:376:load_datasets]
2025-05-29 19:26:01,853 - INFO -    - 📄 L2R dataset: 260243456 records - [learning2rank.py:377:load_datasets]
2025-05-29 19:26:01,853 - INFO -    - 🔤 Tokens: 32768 items - [learning2rank.py:378:load_datasets]
2025-05-29 19:26:01,853 - INFO -    - 🏷️ Labels: 7942 items - [learning2rank.py:379:load_datasets]
2025-05-29 19:26:01,860 - INFO - ✅ Data loading completed successfully - [learning2rank.py:387:load_datasets]
2025-05-29 19:26:03,128 - INFO - ⏳ Starting quantization of ranks for DataFrame with 260243456 rows. Containing 32768 unique tokens & 7942 unique labels - [learning2rank.py:506:quantize_l2r_data]
2025-05-29 19:26:03,129 - INFO - 🔄 Quantizing those 32768 unique token ranks into 101 quantization levels for each label - [learning2rank.py:531:quantize_l2r_data]
2025-05-29 19:26:49,291 - INFO - ✅ Completed quantization: Produced tensor of shape torch.Size([7942, 32768, 4]) with 101 quantization levels per label - [learning2rank.py:585:quantize_l2r_data]
2025-05-29 19:26:49,320 - WARNING - Label 1295: Only 1 tokens with top relevance score (need 50) - [learning2rank.py:701:test_scored_tokens]
2025-05-29 19:26:49,326 - WARNING - Label 4049: Only 1 tokens with top relevance score (need 50) - [learning2rank.py:701:test_scored_tokens]
2025-05-29 19:26:49,330 - WARNING - Label 3517: Only 1 tokens with top relevance score (need 50) - [learning2rank.py:701:test_scored_tokens]
2025-05-29 19:26:49,334 - WARNING - Label 4308: Only 1 tokens with top relevance score (need 50) - [learning2rank.py:701:test_scored_tokens]
2025-05-29 19:26:49,337 - WARNING - Label 519: Only 1 tokens with top relevance score (need 50) - [learning2rank.py:701:test_scored_tokens]
2025-05-29 19:26:49,365 - INFO - ******************************************************************************** - [learning2rank.py:108:log_section]
2025-05-29 19:26:49,365 - INFO - *                   🌟 STARTING LEARNING TO RANK MODEL TRAINING                   * - [learning2rank.py:109:log_section]
2025-05-29 19:26:49,365 - INFO - ******************************************************************************** - [learning2rank.py:112:log_section]
2025-05-29 19:26:49,366 - INFO - 🔐 Loaded authentication token from environment - [learning2rank.py:3889:main]
2025-05-29 19:26:49,366 - INFO - 🏷️ Hub Model ID for this Learning to Rank task: deb101/mistral-7b-instruct-v0.3-mimic4-adapt-l2r - [learning2rank.py:3893:main]
2025-05-29 19:26:49,366 - INFO - -------------------------------------------------------------------------------- - [learning2rank.py:108:log_section]
2025-05-29 19:26:49,366 - INFO - -                            📋 MODEL EXISTENCE CHECK                             - - [learning2rank.py:109:log_section]
2025-05-29 19:26:49,366 - INFO - -------------------------------------------------------------------------------- - [learning2rank.py:112:log_section]
2025-05-29 19:26:49,366 - INFO - 🔍 Checking model existence locally and on Hugging Face Hub... - [learning2rank.py:407:check_model_existence]
2025-05-29 19:26:49,366 - INFO - ❌ Model not found locally at: ../tmp/MIMIC4_DEMO/mistral7b_mimic4_l2r - [learning2rank.py:414:check_model_existence]
2025-05-29 19:26:49,481 - INFO - ✅ Model exists on Hugging Face Hub with ID: deb101/mistral-7b-instruct-v0.3-mimic4-adapt-l2r - [learning2rank.py:426:check_model_existence]
2025-05-29 19:26:49,481 - INFO - 📁 Model exists either locally or on Hub - [learning2rank.py:452:check_model_existence]
2025-05-29 19:26:49,481 - INFO - ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ - [learning2rank.py:108:log_section]
2025-05-29 19:26:49,481 - INFO - +                           ✨ STARTING FRESH TRAINING                            + - [learning2rank.py:109:log_section]
2025-05-29 19:26:49,481 - INFO - ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ - [learning2rank.py:112:log_section]
2025-05-29 19:26:49,481 - INFO - 🔄 Starting fresh training (either forced or model not found)... - [learning2rank.py:3906:main]
2025-05-29 19:26:49,496 - WARNING - Note: Environment variable`HF_TOKEN` is set and is the current active token independently from the token you've just configured. - [_login.py:415:_login]
2025-05-29 19:26:49,496 - INFO - 🔑 Successfully authenticated with Hugging Face Hub - [learning2rank.py:288:authenticate_hf]
2025-05-29 19:26:49,496 - INFO - ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ - [learning2rank.py:108:log_section]
2025-05-29 19:26:49,496 - INFO - +                              ✨ LOADING BASE MODEL                              + - [learning2rank.py:109:log_section]
2025-05-29 19:26:49,496 - INFO - ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ - [learning2rank.py:112:log_section]
2025-05-29 19:26:49,496 - INFO - 📥 Loading pretrained model and tokenizer... - [learning2rank.py:3948:main]
2025-05-29 19:26:49,496 - INFO - 🚀 Starting model and tokenizer loading process... - [learning2rank.py:916:load_base_model_and_tokenizer]
2025-05-29 19:26:49,497 - INFO - 📊 Quantization config: BitsAndBytesConfig {
  "_load_in_4bit": true,
  "_load_in_8bit": false,
  "bnb_4bit_compute_dtype": "bfloat16",
  "bnb_4bit_quant_storage": "uint8",
  "bnb_4bit_quant_type": "nf4",
  "bnb_4bit_use_double_quant": true,
  "llm_int8_enable_fp32_cpu_offload": false,
  "llm_int8_has_fp16_weight": false,
  "llm_int8_skip_modules": null,
  "llm_int8_threshold": 6.0,
  "load_in_4bit": true,
  "load_in_8bit": false,
  "quant_method": "bitsandbytes"
}
 - [learning2rank.py:925:load_base_model_and_tokenizer]
2025-05-29 19:26:49,497 - INFO - 🔤 Loading tokenizer for model: mistralai/Mistral-7B-Instruct-v0.3... - [learning2rank.py:929:load_base_model_and_tokenizer]
2025-05-29 19:26:50,177 - INFO - 📝 Setting pad token to eos token... - [learning2rank.py:933:load_base_model_and_tokenizer]
2025-05-29 19:26:50,177 - INFO - 🧠 Loading base model with quantization... - [learning2rank.py:941:load_base_model_and_tokenizer]
2025-05-29 19:26:50,740 - INFO - We will use 90% of the memory on device 0 for storing the model, and 10% for the buffer to avoid OOM. You can set `max_memory` in to a higher value to use more memory (at your own risk). - [modeling.py:991:get_balanced_memory]
2025-05-29 19:26:56,190 - INFO - 🔧 Setting up default LoRA configuration... - [learning2rank.py:964:load_base_model_and_tokenizer]
2025-05-29 19:26:56,190 - INFO - 🔍 LoRA config: r=16, alpha=32, targets={'o_proj', 'q_proj', 'k_proj', 'v_proj'}, dropout=0.05 - [learning2rank.py:987:load_base_model_and_tokenizer]
2025-05-29 19:26:56,190 - INFO - 🧩 Applying LoRA adapters to model... - [learning2rank.py:994:load_base_model_and_tokenizer]
2025-05-29 19:26:56,374 - INFO - ✅ Model and tokenizer successfully loaded! - [learning2rank.py:1001:load_base_model_and_tokenizer]
2025-05-29 19:26:56,374 - INFO - ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ - [learning2rank.py:108:log_section]
2025-05-29 19:26:56,374 - INFO - +                              ✨ DATA PREPROCESSING                              + - [learning2rank.py:109:log_section]
2025-05-29 19:26:56,374 - INFO - ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ - [learning2rank.py:112:log_section]
2025-05-29 19:26:56,374 - INFO - 🔄 Loading and preprocessing training data... - [learning2rank.py:3956:main]
2025-05-29 19:26:57,442 - INFO - 🔍 Verifying uniqueness of token IDs per label in scored_tokens... - [learning2rank.py:1279:preprocess_data]
2025-05-29 19:26:58,264 - INFO - ✅ All labels have unique token IDs in scored_tokens. - [learning2rank.py:1291:preprocess_data]
2025-05-29 19:26:58,264 - INFO - 🚀 Building a 2D lookup table for efficient token-to-relevance mapping across all labels! 🚀 - [learning2rank.py:1294:preprocess_data]
2025-05-29 19:26:58,264 - INFO - 🔢 Total labels = 7942 - [learning2rank.py:1297:preprocess_data]
2025-05-29 19:26:58,264 - INFO - 🔍 Precomputing token indices and corresponding relevance_values for each label... - [learning2rank.py:1298:preprocess_data]
2025-05-29 19:26:58,492 - INFO - 📊 Lookup table dimensions: 32768 vocabulary size × 7942 labels - [learning2rank.py:1305:preprocess_data]
2025-05-29 19:26:58,492 - INFO - ⚡ This approach eliminates token comparison broadcasting and provides O(1) lookup time for relevance scores! - [learning2rank.py:1308:preprocess_data]
2025-05-29 19:26:58,492 - INFO - 🧮 Processing relevance calculations vectorized for maximum speed 🔥 - [learning2rank.py:1311:preprocess_data]
2025-05-29 19:26:58,657 - INFO - 🔍 Verifying token mappings with 10 samples... - [learning2rank.py:1344:verify_token_mappings]
2025-05-29 19:26:58,838 - INFO - ✅ Token mappings verification completed successfully! 🎉 - [learning2rank.py:1435:verify_token_mappings]
2025-05-29 19:26:58,842 - INFO - 🔄 Processing dataset with map... - [learning2rank.py:1499:preprocess_data]
2025-05-29 19:26:59,159 - INFO - ✅ Dataset built in 0h 0m 0.32s - [learning2rank.py:1522:preprocess_data]
2025-05-29 19:26:59,171 - INFO - The size of Training set: 173 🏋️ (Training Data) - [learning2rank.py:1553:preprocess_data]
2025-05-29 19:26:59,172 - INFO - The size of Evaluation set: 35 🔍 (Test Data) - [learning2rank.py:1554:preprocess_data]
2025-05-29 19:26:59,172 - INFO - 🚀 Created HuggingFace Dataset with 208 samples, 7942 labels - [learning2rank.py:1562:preprocess_data]
2025-05-29 19:26:59,172 - INFO - 🏷️ Number of unique ICD-10 codes: 7942 - [learning2rank.py:3969:main]
2025-05-29 19:26:59,173 - INFO - ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ - [learning2rank.py:108:log_section]
2025-05-29 19:26:59,173 - INFO - +                             ✨ MODEL INITIALIZATION                             + - [learning2rank.py:109:log_section]
2025-05-29 19:26:59,173 - INFO - ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ - [learning2rank.py:112:log_section]
2025-05-29 19:26:59,173 - INFO - 🧠 Initializing custom L2R model for outputting per-token relevance scores per ICD-10 codes. - [learning2rank.py:3972:main]
2025-05-29 19:27:00,707 - INFO - ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ - [learning2rank.py:108:log_section]
2025-05-29 19:27:00,707 - INFO - +                             ✨ TRAINING PREPARATION                             + - [learning2rank.py:109:log_section]
2025-05-29 19:27:00,707 - INFO - ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ - [learning2rank.py:112:log_section]
2025-05-29 19:27:00,707 - INFO - ⚙️ Preparing training components and optimizers... - [learning2rank.py:3979:main]
2025-05-29 19:27:00,739 - INFO - 🖥️ Device: NVIDIA GH200 480GB - [learning2rank.py:2007:log_training_configuration]
2025-05-29 19:27:00,739 - INFO - 🔋 CUDA Available: True - [learning2rank.py:2010:log_training_configuration]
2025-05-29 19:27:00,739 - INFO - 💾 CUDA Device Count: 1 - [learning2rank.py:2011:log_training_configuration]
2025-05-29 19:27:00,740 - INFO - 
📋 Training Configuration 📋
+----------+-----------------------------+--------------------------------------------------+
| 🌟 Emoji |         🏷️ Parameter         |                     📊 Value                     |
+----------+-----------------------------+--------------------------------------------------+
|    📁    |      Output Directory       |     ../tmp/MIMIC4_DEMO/mistral7b_mimic4_l2r      |
|    🔁    |       Training Epochs       |                        4                         |
|    🏋️     |      Train Batch Size       |                        1                         |
|    🔍    |       Eval Batch Size       |                        1                         |
|    📊    | Gradient Accumulation Steps |                        4                         |
|    🚀    |        Learning Rate        |                      0.0001                      |
|    🌅    |        Warmup Steps         |                        0                         |
|    💾    |        Save Strategy        |                      epoch                       |
|    💾    |      Save Total Limit       |                        10                        |
|    📊    |     Evaluation Strategy     |                      epoch                       |
|    🎯    |      Best Model Metric      |                     ndcg@25                      |
|    📝    |      Logging Strategy       |              steps (every 10 steps)              |
|    🌐    |         Push to Hub         |                       True                       |
|    🌐    |        Hub Model ID         | deb101/mistral-7b-instruct-v0.3-mimic4-adapt-l2r |
|    🔢    |       Steps per Epoch       |                        43                        |
|    🔢    |    Total Training Steps     |                       172                        |
|    🔢    |      Evaluation Steps       |                        35                        |
|    📊    |    Training Dataset Size    |                  173 samples 🏋️                   |
|    📊    |   Evaluation Dataset Size   |                  35 samples 🔍                   |
+----------+-----------------------------+--------------------------------------------------+ - [learning2rank.py:1999:log_training_args]
2025-05-29 19:27:00,740 - INFO - ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ - [learning2rank.py:108:log_section]
2025-05-29 19:27:00,740 - INFO - +                                ✨ MODEL TRAINING                                + - [learning2rank.py:109:log_section]
2025-05-29 19:27:00,740 - INFO - ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ - [learning2rank.py:112:log_section]
2025-05-29 19:27:00,740 - INFO - 🏋️ Starting model training process... - [learning2rank.py:3999:main]
2025-05-29 19:27:00,740 - INFO - 🏁 Preparing Custom Trainer 🛠️ - [learning2rank.py:3200:train_model]
2025-05-29 19:27:00,784 - INFO - We are registering the tokenizer mistralai/Mistral-7B-Instruct-v0.3 in Custom Trainer - [learning2rank.py:2639:__init__]
2025-05-29 19:27:00,784 - INFO - 🏋️ Commencing Model Training 💪 - [learning2rank.py:3241:train_model]
2025-05-29 19:27:01,050 - INFO - 🚀 Starting Training... - [learning2rank.py:2389:on_train_begin]
2025-05-29 19:27:20,294 - INFO - 
🚂 Training Metrics (Step 10) 🚂
+---------------+--------------+
| Metric        |        Value |
+===============+==============+
| loss          | -4.58268e+16 |
+---------------+--------------+
| grad_norm     |          nan |
+---------------+--------------+
| learning_rate |       0.0001 |
+---------------+--------------+
| epoch         |     0.231214 |
+---------------+--------------+ - [learning2rank.py:2307:on_log]
2025-05-29 19:27:34,347 - INFO - 
🚂 Training Metrics (Step 20) 🚂
+---------------+--------------+
| Metric        |        Value |
+===============+==============+
| loss          | -4.34403e+17 |
+---------------+--------------+
| grad_norm     |          nan |
+---------------+--------------+
| learning_rate |       0.0001 |
+---------------+--------------+
| epoch         |     0.462428 |
+---------------+--------------+ - [learning2rank.py:2307:on_log]
2025-05-29 19:27:48,535 - INFO - 
🚂 Training Metrics (Step 30) 🚂
+---------------+--------------+
| Metric        |        Value |
+===============+==============+
| loss          | -6.33341e+16 |
+---------------+--------------+
| grad_norm     |          nan |
+---------------+--------------+
| learning_rate |       0.0001 |
+---------------+--------------+
| epoch         |     0.693642 |
+---------------+--------------+ - [learning2rank.py:2307:on_log]
2025-05-29 19:28:02,663 - INFO - 
🚂 Training Metrics (Step 40) 🚂
+---------------+--------------+
| Metric        |        Value |
+===============+==============+
| loss          | -6.34723e+18 |
+---------------+--------------+
| grad_norm     |          nan |
+---------------+--------------+
| learning_rate |       0.0001 |
+---------------+--------------+
| epoch         |     0.924855 |
+---------------+--------------+ - [learning2rank.py:2307:on_log]
2025-05-29 19:28:07,197 - INFO - Removing 'token_type_ids' from eval_dataset as they are not needed. - [learning2rank.py:2755:evaluate]
2025-05-29 19:31:21,756 - WARNING - No valid samples found for metric 'precision@25'. - [learning2rank.py:2859:evaluate]
2025-05-29 19:31:21,757 - INFO - 
🔍 Evaluation Metrics 🔍
+-------------------+--------------+
| Metric            |        Value |
+===================+==============+
| eval_loss         | -1.17021e+17 |
+-------------------+--------------+
| eval_ndcg         |     0.955488 |
+-------------------+--------------+
| eval_ndcg@25      |      0.18984 |
+-------------------+--------------+
| eval_precision@25 |            0 |
+-------------------+--------------+ - [learning2rank.py:2326:on_evaluate]
2025-05-29 19:31:23,021 - INFO - 💾 Model weights saved in safetensors format: ../tmp/MIMIC4_DEMO/mistral7b_mimic4_l2r/checkpoint-44 - [learning2rank.py:2879:_save]
2025-05-29 19:31:23,022 - INFO - ⚙️ Config saved in checkpoint: ../tmp/MIMIC4_DEMO/mistral7b_mimic4_l2r/checkpoint-44 - [learning2rank.py:2884:_save]
2025-05-29 19:31:23,023 - INFO - 📋 Saved files in ../tmp/MIMIC4_DEMO/mistral7b_mimic4_l2r/checkpoint-44:
+---------+-------------------+------------+
|   Index | Saved File        | Size       |
+=========+===================+============+
|       1 | training_args.bin | 0.01 MB    |
+---------+-------------------+------------+
|       2 | model.safetensors | 4122.74 MB |
+---------+-------------------+------------+
|       3 | config.json       | 0.00 MB    |
+---------+-------------------+------------+ - [learning2rank.py:2901:_save]
2025-05-29 19:31:36,370 - INFO - 
🚂 Training Metrics (Step 50) 🚂
+---------------+--------------+
| Metric        |        Value |
+===============+==============+
| loss          | -4.29741e+18 |
+---------------+--------------+
| grad_norm     |          nan |
+---------------+--------------+
| learning_rate |       0.0001 |
+---------------+--------------+
| epoch         |      1.13873 |
+---------------+--------------+ - [learning2rank.py:2307:on_log]
2025-05-29 19:31:50,557 - INFO - 
🚂 Training Metrics (Step 60) 🚂
+---------------+--------------+
| Metric        |        Value |
+===============+==============+
| loss          | -1.10586e+17 |
+---------------+--------------+
| grad_norm     |          nan |
+---------------+--------------+
| learning_rate |      9.9e-05 |
+---------------+--------------+
| epoch         |      1.36994 |
+---------------+--------------+ - [learning2rank.py:2307:on_log]
2025-05-29 19:32:04,594 - INFO - 
🚂 Training Metrics (Step 70) 🚂
+---------------+--------------+
| Metric        |        Value |
+===============+==============+
| loss          | -6.65756e+18 |
+---------------+--------------+
| grad_norm     |  2.41222e+18 |
+---------------+--------------+
| learning_rate |      9.4e-05 |
+---------------+--------------+
| epoch         |      1.60116 |
+---------------+--------------+ - [learning2rank.py:2307:on_log]
2025-05-29 19:32:18,534 - INFO - 
🚂 Training Metrics (Step 80) 🚂
+---------------+--------------+
| Metric        |        Value |
+===============+==============+
| loss          | -9.52556e+16 |
+---------------+--------------+
| grad_norm     |  1.31876e+18 |
+---------------+--------------+
| learning_rate |      8.8e-05 |
+---------------+--------------+
| epoch         |      1.83237 |
+---------------+--------------+ - [learning2rank.py:2307:on_log]
2025-05-29 19:32:28,622 - INFO - Removing 'token_type_ids' from eval_dataset as they are not needed. - [learning2rank.py:2755:evaluate]
2025-05-29 19:35:41,610 - INFO - 
🔍 Evaluation Metrics 🔍
+-------------------+--------------+
| Metric            |        Value |
+===================+==============+
| eval_loss         | -3.19605e+17 |
+-------------------+--------------+
| eval_ndcg         |     0.956218 |
+-------------------+--------------+
| eval_ndcg@25      |     0.644859 |
+-------------------+--------------+
| eval_precision@25 |     0.524571 |
+-------------------+--------------+ - [learning2rank.py:2326:on_evaluate]
2025-05-29 19:35:42,773 - INFO - 💾 Model weights saved in safetensors format: ../tmp/MIMIC4_DEMO/mistral7b_mimic4_l2r/checkpoint-88 - [learning2rank.py:2879:_save]
2025-05-29 19:35:42,774 - INFO - ⚙️ Config saved in checkpoint: ../tmp/MIMIC4_DEMO/mistral7b_mimic4_l2r/checkpoint-88 - [learning2rank.py:2884:_save]
2025-05-29 19:35:42,775 - INFO - 📋 Saved files in ../tmp/MIMIC4_DEMO/mistral7b_mimic4_l2r/checkpoint-88:
+---------+-------------------+------------+
|   Index | Saved File        | Size       |
+=========+===================+============+
|       1 | training_args.bin | 0.01 MB    |
+---------+-------------------+------------+
|       2 | model.safetensors | 4122.74 MB |
+---------+-------------------+------------+
|       3 | config.json       | 0.00 MB    |
+---------+-------------------+------------+ - [learning2rank.py:2901:_save]
2025-05-29 19:35:50,541 - INFO - 
🚂 Training Metrics (Step 90) 🚂
+---------------+--------------+
| Metric        |        Value |
+===============+==============+
| loss          | -7.78904e+17 |
+---------------+--------------+
| grad_norm     |  4.04588e+17 |
+---------------+--------------+
| learning_rate |      8.3e-05 |
+---------------+--------------+
| epoch         |      2.04624 |
+---------------+--------------+ - [learning2rank.py:2307:on_log]
2025-05-29 19:36:04,506 - INFO - 
🚂 Training Metrics (Step 100) 🚂
+---------------+--------------+
| Metric        |        Value |
+===============+==============+
| loss          | -2.16708e+17 |
+---------------+--------------+
| grad_norm     |  6.69977e+17 |
+---------------+--------------+
| learning_rate |      7.7e-05 |
+---------------+--------------+
| epoch         |      2.27746 |
+---------------+--------------+ - [learning2rank.py:2307:on_log]
2025-05-29 19:36:18,486 - INFO - 
🚂 Training Metrics (Step 110) 🚂
+---------------+-------------+
| Metric        |       Value |
+===============+=============+
| loss          | -3.3602e+18 |
+---------------+-------------+
| grad_norm     |         nan |
+---------------+-------------+
| learning_rate |     7.3e-05 |
+---------------+-------------+
| epoch         |     2.50867 |
+---------------+-------------+ - [learning2rank.py:2307:on_log]
2025-05-29 19:36:32,497 - INFO - 
🚂 Training Metrics (Step 120) 🚂
+---------------+--------------+
| Metric        |        Value |
+===============+==============+
| loss          | -4.15132e+17 |
+---------------+--------------+
| grad_norm     |  7.31787e+17 |
+---------------+--------------+
| learning_rate |      6.7e-05 |
+---------------+--------------+
| epoch         |      2.73988 |
+---------------+--------------+ - [learning2rank.py:2307:on_log]
2025-05-29 19:36:46,461 - INFO - 
🚂 Training Metrics (Step 130) 🚂
+---------------+--------------+
| Metric        |        Value |
+===============+==============+
| loss          | -1.23755e+18 |
+---------------+--------------+
| grad_norm     |  2.11757e+17 |
+---------------+--------------+
| learning_rate |      6.2e-05 |
+---------------+--------------+
| epoch         |       2.9711 |
+---------------+--------------+ - [learning2rank.py:2307:on_log]
2025-05-29 19:36:48,212 - INFO - Removing 'token_type_ids' from eval_dataset as they are not needed. - [learning2rank.py:2755:evaluate]
2025-05-29 19:40:00,995 - INFO - 
🔍 Evaluation Metrics 🔍
+-------------------+--------------+
| Metric            |        Value |
+===================+==============+
| eval_loss         | -4.05479e+17 |
+-------------------+--------------+
| eval_ndcg         |     0.956806 |
+-------------------+--------------+
| eval_ndcg@25      |     0.793177 |
+-------------------+--------------+
| eval_precision@25 |     0.890286 |
+-------------------+--------------+ - [learning2rank.py:2326:on_evaluate]
2025-05-29 19:40:02,153 - INFO - 💾 Model weights saved in safetensors format: ../tmp/MIMIC4_DEMO/mistral7b_mimic4_l2r/checkpoint-132 - [learning2rank.py:2879:_save]
2025-05-29 19:40:02,154 - INFO - ⚙️ Config saved in checkpoint: ../tmp/MIMIC4_DEMO/mistral7b_mimic4_l2r/checkpoint-132 - [learning2rank.py:2884:_save]
2025-05-29 19:40:02,155 - INFO - 📋 Saved files in ../tmp/MIMIC4_DEMO/mistral7b_mimic4_l2r/checkpoint-132:
+---------+-------------------+------------+
|   Index | Saved File        | Size       |
+=========+===================+============+
|       1 | training_args.bin | 0.01 MB    |
+---------+-------------------+------------+
|       2 | model.safetensors | 4122.74 MB |
+---------+-------------------+------------+
|       3 | config.json       | 0.00 MB    |
+---------+-------------------+------------+ - [learning2rank.py:2901:_save]
2025-05-29 19:40:18,266 - INFO - 
🚂 Training Metrics (Step 140) 🚂
+---------------+--------------+
| Metric        |        Value |
+===============+==============+
| loss          | -9.65112e+18 |
+---------------+--------------+
| grad_norm     |   1.7024e+17 |
+---------------+--------------+
| learning_rate |      5.7e-05 |
+---------------+--------------+
| epoch         |      3.18497 |
+---------------+--------------+ - [learning2rank.py:2307:on_log]
2025-05-29 19:40:32,114 - INFO - 
🚂 Training Metrics (Step 150) 🚂
+---------------+--------------+
| Metric        |        Value |
+===============+==============+
| loss          | -6.81527e+18 |
+---------------+--------------+
| grad_norm     |  1.33906e+17 |
+---------------+--------------+
| learning_rate |      5.2e-05 |
+---------------+--------------+
| epoch         |      3.41619 |
+---------------+--------------+ - [learning2rank.py:2307:on_log]
2025-05-29 19:40:45,965 - INFO - 
🚂 Training Metrics (Step 160) 🚂
+---------------+--------------+
| Metric        |        Value |
+===============+==============+
| loss          | -4.90954e+17 |
+---------------+--------------+
| grad_norm     |  2.87998e+17 |
+---------------+--------------+
| learning_rate |      4.6e-05 |
+---------------+--------------+
| epoch         |       3.6474 |
+---------------+--------------+ - [learning2rank.py:2307:on_log]
2025-05-29 19:40:59,736 - INFO - 
🚂 Training Metrics (Step 170) 🚂
+---------------+--------------+
| Metric        |        Value |
+===============+==============+
| loss          | -3.49743e+17 |
+---------------+--------------+
| grad_norm     |   2.0459e+17 |
+---------------+--------------+
| learning_rate |        4e-05 |
+---------------+--------------+
| epoch         |      3.87861 |
+---------------+--------------+ - [learning2rank.py:2307:on_log]
2025-05-29 19:41:03,699 - INFO - 💾 Model weights saved in safetensors format: ../tmp/MIMIC4_DEMO/mistral7b_mimic4_l2r/checkpoint-172 - [learning2rank.py:2879:_save]
2025-05-29 19:41:03,701 - INFO - ⚙️ Config saved in checkpoint: ../tmp/MIMIC4_DEMO/mistral7b_mimic4_l2r/checkpoint-172 - [learning2rank.py:2884:_save]
2025-05-29 19:41:03,701 - INFO - 📋 Saved files in ../tmp/MIMIC4_DEMO/mistral7b_mimic4_l2r/checkpoint-172:
+---------+-------------------+------------+
|   Index | Saved File        | Size       |
+=========+===================+============+
|       1 | training_args.bin | 0.01 MB    |
+---------+-------------------+------------+
|       2 | model.safetensors | 4122.74 MB |
+---------+-------------------+------------+
|       3 | config.json       | 0.00 MB    |
+---------+-------------------+------------+ - [learning2rank.py:2901:_save]
2025-05-29 19:41:03,908 - INFO - Removing 'token_type_ids' from eval_dataset as they are not needed. - [learning2rank.py:2755:evaluate]
2025-05-29 19:44:16,555 - INFO - 
🔍 Evaluation Metrics 🔍
+-------------------+-------------+
| Metric            |       Value |
+===================+=============+
| eval_loss         | -4.1649e+17 |
+-------------------+-------------+
| eval_ndcg         |    0.956866 |
+-------------------+-------------+
| eval_ndcg@25      |    0.873974 |
+-------------------+-------------+
| eval_precision@25 |    0.930286 |
+-------------------+-------------+ - [learning2rank.py:2326:on_evaluate]
2025-05-29 19:44:18,478 - INFO - 💾 Model weights saved in safetensors format: ../tmp/MIMIC4_DEMO/mistral7b_mimic4_l2r/checkpoint-172 - [learning2rank.py:2879:_save]
2025-05-29 19:44:18,479 - INFO - ⚙️ Config saved in checkpoint: ../tmp/MIMIC4_DEMO/mistral7b_mimic4_l2r/checkpoint-172 - [learning2rank.py:2884:_save]
2025-05-29 19:44:18,480 - INFO - 📋 Saved files in ../tmp/MIMIC4_DEMO/mistral7b_mimic4_l2r/checkpoint-172:
+---------+--------------------+------------+
|   Index | Saved File         | Size       |
+=========+====================+============+
|       1 | training_args.bin  | 0.01 MB    |
+---------+--------------------+------------+
|       2 | optimizer.pt       | 352.39 MB  |
+---------+--------------------+------------+
|       3 | model.safetensors  | 4122.74 MB |
+---------+--------------------+------------+
|       4 | scaler.pt          | 0.00 MB    |
+---------+--------------------+------------+
|       5 | config.json        | 0.00 MB    |
+---------+--------------------+------------+
|       6 | scheduler.pt       | 0.00 MB    |
+---------+--------------------+------------+
|       7 | trainer_state.json | 0.00 MB    |
+---------+--------------------+------------+
|       8 | rng_state.pth      | 0.01 MB    |
+---------+--------------------+------------+ - [learning2rank.py:2901:_save]
2025-05-29 19:44:18,765 - INFO - 📂 Loading best model from ../tmp/MIMIC4_DEMO/mistral7b_mimic4_l2r/checkpoint-172 - [learning2rank.py:2957:_load_best_model]
2025-05-29 19:44:18,765 - INFO - 🖥️ Model is on device: cuda:0 - [learning2rank.py:2967:_load_best_model]
2025-05-29 19:44:18,821 - INFO - 🔑 Key order comparison:
+---------+------------------------------------------------------------------------+---------------------------------------------------------------------------------------------+
|   Index | Saved state_dict Keys                                                  | Model state_dict Keys                                                                       |
+=========+========================================================================+=============================================================================================+
|       1 | base_model.base_model.model.lm_head.weight                             | base_model.base_model.model.model.embed_tokens.weight                                       |
+---------+------------------------------------------------------------------------+---------------------------------------------------------------------------------------------+
|       2 | base_model.base_model.model.model.embed_tokens.weight                  | base_model.base_model.model.model.layers.0.self_attn.q_proj.base_layer.weight               |
+---------+------------------------------------------------------------------------+---------------------------------------------------------------------------------------------+
|       3 | base_model.base_model.model.model.layers.0.input_layernorm.weight      | base_model.base_model.model.model.layers.0.self_attn.q_proj.base_layer.weight.absmax        |
+---------+------------------------------------------------------------------------+---------------------------------------------------------------------------------------------+
|       4 | base_model.base_model.model.model.layers.0.mlp.down_proj.weight        | base_model.base_model.model.model.layers.0.self_attn.q_proj.base_layer.weight.quant_map     |
+---------+------------------------------------------------------------------------+---------------------------------------------------------------------------------------------+
|       5 | base_model.base_model.model.model.layers.0.mlp.down_proj.weight.absmax | base_model.base_model.model.model.layers.0.self_attn.q_proj.base_layer.weight.nested_absmax |
+---------+------------------------------------------------------------------------+---------------------------------------------------------------------------------------------+ - [learning2rank.py:2991:_load_best_model]
2025-05-29 19:44:19,629 - INFO - ✅ Loaded best model weights from ../tmp/MIMIC4_DEMO/mistral7b_mimic4_l2r/checkpoint-172/model.safetensors - [learning2rank.py:3008:_load_best_model]
2025-05-29 19:44:19,686 - INFO - ✔️ Weight for base_model.base_model.model.model.embed_tokens.weight matches between saved and loaded state_dict - [learning2rank.py:3020:_load_best_model]
2025-05-29 19:44:19,708 - INFO - ✔️ Weight for base_model.base_model.model.model.layers.0.self_attn.q_proj.base_layer.weight matches between saved and loaded state_dict - [learning2rank.py:3020:_load_best_model]
2025-05-29 19:44:19,725 - INFO - 
🚂 Training Metrics (Step 172) 🚂
+--------------------------+--------------+
| Metric                   |        Value |
+==========================+==============+
| train_runtime            |      1038.68 |
+--------------------------+--------------+
| train_samples_per_second |        0.666 |
+--------------------------+--------------+
| train_steps_per_second   |        0.166 |
+--------------------------+--------------+
| total_flos               |            0 |
+--------------------------+--------------+
| train_loss               | -2.40817e+18 |
+--------------------------+--------------+
| epoch                    |      3.92485 |
+--------------------------+--------------+ - [learning2rank.py:2307:on_log]
2025-05-29 19:44:19,726 - INFO - ✨ Training Completed! ✨ - [learning2rank.py:2456:on_train_end]
2025-05-29 19:44:19,800 - INFO - 📊 Training loss plot saved as '../tmp/MIMIC4_DEMO/mistral7b_mimic4_l2r/train_loss_plot.png' - [learning2rank.py:2559:on_train_end]
2025-05-29 19:44:19,858 - INFO - 📊 Evaluation loss plot saved as '../tmp/MIMIC4_DEMO/mistral7b_mimic4_l2r/eval_loss_plot.png' - [learning2rank.py:2573:on_train_end]
2025-05-29 19:44:19,918 - INFO - 📊 Evaluation metric plot saved as '../tmp/MIMIC4_DEMO/mistral7b_mimic4_l2r/eval_ndcg@25_plot.png' - [learning2rank.py:2594:on_train_end]
2025-05-29 19:44:19,919 - INFO - ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ - [learning2rank.py:108:log_section]
2025-05-29 19:44:19,919 - INFO - +                                 ✨ MODEL SAVING                                 + - [learning2rank.py:109:log_section]
2025-05-29 19:44:19,919 - INFO - ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ - [learning2rank.py:112:log_section]
2025-05-29 19:44:19,919 - INFO - 💾 Saving trained model and pushing to Hugging Face Hub... - [learning2rank.py:4014:main]
2025-05-29 19:44:19,919 - INFO - 📁 Creating/using output directory: ../tmp/MIMIC4_DEMO/mistral7b_mimic4_l2r - [learning2rank.py:3275:save_and_push]
2025-05-29 19:44:21,100 - INFO - 💾 Model weights saved in safetensors format: ../tmp/MIMIC4_DEMO/mistral7b_mimic4_l2r - [learning2rank.py:2879:_save]
2025-05-29 19:44:21,102 - INFO - ⚙️ Config saved in checkpoint: ../tmp/MIMIC4_DEMO/mistral7b_mimic4_l2r - [learning2rank.py:2884:_save]
2025-05-29 19:44:21,103 - INFO - 📋 Saved files in ../tmp/MIMIC4_DEMO/mistral7b_mimic4_l2r:
+---------+-----------------------+------------+
|   Index | Saved File            | Size       |
+=========+=======================+============+
|       1 | eval_loss_plot.png    | 0.03 MB    |
+---------+-----------------------+------------+
|       2 | training_args.bin     | 0.01 MB    |
+---------+-----------------------+------------+
|       3 | model.safetensors     | 4122.74 MB |
+---------+-----------------------+------------+
|       4 | eval_ndcg@25_plot.png | 0.03 MB    |
+---------+-----------------------+------------+
|       5 | config.json           | 0.00 MB    |
+---------+-----------------------+------------+
|       6 | train_loss_plot.png   | 0.04 MB    |
+---------+-----------------------+------------+ - [learning2rank.py:2901:_save]
2025-05-29 19:44:24,684 - INFO - 💾 Model weights saved in safetensors format: ../tmp/MIMIC4_DEMO/mistral7b_mimic4_l2r - [learning2rank.py:2879:_save]
2025-05-29 19:44:24,685 - INFO - ⚙️ Config saved in checkpoint: ../tmp/MIMIC4_DEMO/mistral7b_mimic4_l2r - [learning2rank.py:2884:_save]
2025-05-29 19:44:24,686 - INFO - 📋 Saved files in ../tmp/MIMIC4_DEMO/mistral7b_mimic4_l2r:
+---------+-----------------------+------------+
|   Index | Saved File            | Size       |
+=========+=======================+============+
|       1 | eval_loss_plot.png    | 0.03 MB    |
+---------+-----------------------+------------+
|       2 | training_args.bin     | 0.01 MB    |
+---------+-----------------------+------------+
|       3 | model.safetensors     | 4122.74 MB |
+---------+-----------------------+------------+
|       4 | eval_ndcg@25_plot.png | 0.03 MB    |
+---------+-----------------------+------------+
|       5 | config.json           | 0.00 MB    |
+---------+-----------------------+------------+
|       6 | train_loss_plot.png   | 0.04 MB    |
+---------+-----------------------+------------+ - [learning2rank.py:2901:_save]
2025-05-29 19:45:41,784 - INFO - 💾 Model saved to: ../tmp/MIMIC4_DEMO/mistral7b_mimic4_l2r - [learning2rank.py:3279:save_and_push]
2025-05-29 19:45:41,814 - INFO - 🖌️ Tokenizer saved to: ../tmp/MIMIC4_DEMO/mistral7b_mimic4_l2r - [learning2rank.py:3283:save_and_push]