| 2025-11-06 11:42:52 - experiment_save_merged_model - INFO - Starting merged model save process |
| 2025-11-06 11:42:52 - experiment_save_merged_model - INFO - Arguments: {'lambdas_path': '/work/gj26/b20042/LLM-AdaMerge/outputs/deepseek-7b/task-wise/cross_entropy-ep2-10%dataset-lambda01/llm_adamerge_lambdas.json', 'model_config': '/work/gj26/b20042/LLM-AdaMerge/outputs/deepseek-7b/task-wise/cross_entropy-ep2-10%dataset-lambda01/model_config.yaml', 'output_dir': '/work/gj26/b20042/LLM-AdaMerge/mergekit/outputs/deepseek-7b/llmadamerge/task-wise/cross_entropy-ep2-10%dataset/lambda01', 'model_name': 'merged-model', 'push_to_hub': False, 'hub_repo_id': 'lejelly/ds7b-ep2-data10-id4-taskwise-lambda01', 'private': False, 'device': 'cuda', 'debug': False} |
| 2025-11-06 11:42:52 - experiment_save_merged_model - INFO - Loading lambdas from /work/gj26/b20042/LLM-AdaMerge/outputs/deepseek-7b/task-wise/cross_entropy-ep2-10%dataset-lambda01/llm_adamerge_lambdas.json |
| 2025-11-06 11:42:52 - experiment_save_merged_model - INFO - Auto-detected parameter-wise merge from JSON structure |
| 2025-11-06 11:42:52 - experiment_save_merged_model - INFO - Merge type: parameter_wise |
| 2025-11-06 11:42:52 - experiment_save_merged_model - INFO - [Initial] Memory Usage: |
| 2025-11-06 11:42:52 - experiment_save_merged_model - INFO - Process: 0.40 GB (0.2%) |
| 2025-11-06 11:42:52 - experiment_save_merged_model - INFO - System: 8.28 GB / 212.49 GB (8.5%) |
| 2025-11-06 11:42:52 - experiment_save_merged_model - INFO - Available: 194.37 GB |
| 2025-11-06 11:42:52 - experiment_save_merged_model - INFO - GPU 0: Allocated: 0.00 GB, Reserved: 0.00 GB, Total: 94.50 GB |
| 2025-11-06 11:42:52 - experiment_save_merged_model - INFO - Loading models |
| 2025-11-06 11:43:06 - experiment_save_merged_model - INFO - [After loading models] Memory Usage: |
| 2025-11-06 11:43:06 - experiment_save_merged_model - INFO - Process: 0.65 GB (0.3%) |
| 2025-11-06 11:43:06 - experiment_save_merged_model - INFO - System: 49.90 GB / 212.49 GB (31.4%) |
| 2025-11-06 11:43:06 - experiment_save_merged_model - INFO - Available: 145.71 GB |
| 2025-11-06 11:43:06 - experiment_save_merged_model - INFO - GPU 0: Allocated: 38.61 GB, Reserved: 40.64 GB, Total: 94.50 GB |
| 2025-11-06 11:43:06 - experiment_save_merged_model - INFO - Initializing parameter_wise AdaMerge |
| 2025-11-06 11:47:32 - experiment_save_merged_model - INFO - Loading learned lambdas |
| 2025-11-06 11:47:32 - experiment_save_merged_model - INFO - Deleting original models to free memory (task vectors already computed) |
| 2025-11-06 11:47:32 - experiment_save_merged_model - INFO - [Before deleting models] Memory Usage: |
| 2025-11-06 11:47:32 - experiment_save_merged_model - INFO - Process: 39.09 GB (18.4%) |
| 2025-11-06 11:47:32 - experiment_save_merged_model - INFO - System: 104.93 GB / 212.49 GB (57.3%) |
| 2025-11-06 11:47:32 - experiment_save_merged_model - INFO - Available: 90.68 GB |
| 2025-11-06 11:47:32 - experiment_save_merged_model - INFO - GPU 0: Allocated: 38.61 GB, Reserved: 40.64 GB, Total: 94.50 GB |
| 2025-11-06 11:47:32 - experiment_save_merged_model - INFO - Clearing model_loader references |
| 2025-11-06 11:47:33 - experiment_save_merged_model - INFO - Deleting model variables |
| 2025-11-06 11:47:33 - experiment_save_merged_model - INFO - Running garbage collection |
| 2025-11-06 11:47:33 - experiment_save_merged_model - INFO - [After deleting models and GC] Memory Usage: |
| 2025-11-06 11:47:33 - experiment_save_merged_model - INFO - Process: 39.09 GB (18.4%) |
| 2025-11-06 11:47:33 - experiment_save_merged_model - INFO - System: 64.28 GB / 212.49 GB (38.2%) |
| 2025-11-06 11:47:33 - experiment_save_merged_model - INFO - Available: 131.33 GB |
| 2025-11-06 11:47:33 - experiment_save_merged_model - INFO - GPU 0: Allocated: 0.00 GB, Reserved: 0.00 GB, Total: 94.50 GB |
| 2025-11-06 11:47:33 - experiment_save_merged_model - INFO - [After loading lambdas] Memory Usage: |
| 2025-11-06 11:47:33 - experiment_save_merged_model - INFO - Process: 39.09 GB (18.4%) |
| 2025-11-06 11:47:33 - experiment_save_merged_model - INFO - System: 64.28 GB / 212.49 GB (38.2%) |
| 2025-11-06 11:47:33 - experiment_save_merged_model - INFO - Available: 131.33 GB |
| 2025-11-06 11:47:33 - experiment_save_merged_model - INFO - GPU 0: Allocated: 0.00 GB, Reserved: 0.00 GB, Total: 94.50 GB |
| 2025-11-06 11:47:33 - experiment_save_merged_model - INFO - Creating merged model with learned lambdas |
| 2025-11-06 11:47:33 - experiment_save_merged_model - INFO - Using merge_models_for_save() |
|
|