deepseek-ep3-data10-taskwise-lambda05 / logs /save_merged_model_20251009_193426.log
lejelly's picture
Upload folder using huggingface_hub
942f9d3 verified
2025-10-09 19:34:26 - experiment_save_merged_model - INFO - Starting merged model save process
2025-10-09 19:34:26 - experiment_save_merged_model - INFO - Arguments: {'lambdas_path': '/work/gj26/b20042/LLM-AdaMerge/outputs/deepseek-7b/task-wise/cross_entropy-ep3-lr0001-10%dataset-lambda05_wo_scheduler/llm_adamerge_lambdas.json', 'model_config': '/work/gj26/b20042/LLM-AdaMerge/outputs/deepseek-7b/task-wise/cross_entropy-ep3-lr0001-10%dataset-lambda05_wo_scheduler/model_config.yaml', 'output_dir': '/work/gj26/b20042/LLM-AdaMerge/mergekit/outputs/deepseek-7b/llmadamerge/task-wise-cross_entropy-lr0001-ep3-10%dataset/lambda05_wo_scheduler', 'model_name': 'merged-model', 'push_to_hub': False, 'hub_repo_id': 'lejelly/deepseek-ep3-data10-taskwise-lambda05', 'private': False, 'device': 'cuda', 'debug': False}
2025-10-09 19:34:26 - experiment_save_merged_model - INFO - Loading lambdas from /work/gj26/b20042/LLM-AdaMerge/outputs/deepseek-7b/task-wise/cross_entropy-ep3-lr0001-10%dataset-lambda05_wo_scheduler/llm_adamerge_lambdas.json
2025-10-09 19:34:26 - experiment_save_merged_model - INFO - Auto-detected parameter-wise merge from JSON structure
2025-10-09 19:34:26 - experiment_save_merged_model - INFO - Merge type: parameter_wise
2025-10-09 19:34:26 - experiment_save_merged_model - INFO - [Initial] Memory Usage:
2025-10-09 19:34:26 - experiment_save_merged_model - INFO - Process: 0.45 GB (0.2%)
2025-10-09 19:34:26 - experiment_save_merged_model - INFO - System: 7.94 GB / 212.49 GB (8.4%)
2025-10-09 19:34:26 - experiment_save_merged_model - INFO - Available: 194.74 GB
2025-10-09 19:34:26 - experiment_save_merged_model - INFO - GPU 0: Allocated: 0.00 GB, Reserved: 0.00 GB, Total: 94.50 GB
2025-10-09 19:34:26 - experiment_save_merged_model - INFO - Loading models
2025-10-09 19:34:45 - experiment_save_merged_model - INFO - [After loading models] Memory Usage:
2025-10-09 19:34:45 - experiment_save_merged_model - INFO - Process: 38.76 GB (18.2%)
2025-10-09 19:34:45 - experiment_save_merged_model - INFO - System: 50.08 GB / 212.49 GB (31.5%)
2025-10-09 19:34:45 - experiment_save_merged_model - INFO - Available: 145.61 GB
2025-10-09 19:34:45 - experiment_save_merged_model - INFO - GPU 0: Allocated: 0.00 GB, Reserved: 0.00 GB, Total: 94.50 GB
2025-10-09 19:34:45 - experiment_save_merged_model - INFO - Initializing parameter_wise AdaMerge
2025-10-09 19:35:54 - experiment_save_merged_model - INFO - Loading learned lambdas
2025-10-09 19:35:54 - experiment_save_merged_model - INFO - Deleting original models to free memory (task vectors already computed)
2025-10-09 19:35:54 - experiment_save_merged_model - INFO - [Before deleting models] Memory Usage:
2025-10-09 19:35:54 - experiment_save_merged_model - INFO - Process: 64.61 GB (30.4%)
2025-10-09 19:35:54 - experiment_save_merged_model - INFO - System: 73.49 GB / 212.49 GB (42.5%)
2025-10-09 19:35:54 - experiment_save_merged_model - INFO - Available: 122.15 GB
2025-10-09 19:35:54 - experiment_save_merged_model - INFO - GPU 0: Allocated: 0.00 GB, Reserved: 0.00 GB, Total: 94.50 GB
2025-10-09 19:35:54 - experiment_save_merged_model - INFO - Clearing model_loader references
2025-10-09 19:35:54 - experiment_save_merged_model - INFO - Deleting model variables
2025-10-09 19:35:54 - experiment_save_merged_model - INFO - Running garbage collection
2025-10-09 19:35:54 - experiment_save_merged_model - INFO - [After deleting models and GC] Memory Usage:
2025-10-09 19:35:54 - experiment_save_merged_model - INFO - Process: 38.84 GB (18.3%)
2025-10-09 19:35:54 - experiment_save_merged_model - INFO - System: 48.12 GB / 212.49 GB (30.6%)
2025-10-09 19:35:54 - experiment_save_merged_model - INFO - Available: 147.52 GB
2025-10-09 19:35:54 - experiment_save_merged_model - INFO - GPU 0: Allocated: 0.00 GB, Reserved: 0.00 GB, Total: 94.50 GB
2025-10-09 19:35:54 - experiment_save_merged_model - INFO - [After loading lambdas] Memory Usage:
2025-10-09 19:35:54 - experiment_save_merged_model - INFO - Process: 38.84 GB (18.3%)
2025-10-09 19:35:54 - experiment_save_merged_model - INFO - System: 48.12 GB / 212.49 GB (30.6%)
2025-10-09 19:35:54 - experiment_save_merged_model - INFO - Available: 147.52 GB
2025-10-09 19:35:54 - experiment_save_merged_model - INFO - GPU 0: Allocated: 0.00 GB, Reserved: 0.00 GB, Total: 94.50 GB
2025-10-09 19:35:54 - experiment_save_merged_model - INFO - Creating merged model with learned lambdas
2025-10-09 19:35:54 - experiment_save_merged_model - INFO - Using merge_models_for_save()
2025-10-09 19:37:51 - experiment_save_merged_model - INFO - [After merging models] Memory Usage:
2025-10-09 19:37:51 - experiment_save_merged_model - INFO - Process: 38.85 GB (18.3%)
2025-10-09 19:37:51 - experiment_save_merged_model - INFO - System: 72.96 GB / 212.49 GB (42.3%)
2025-10-09 19:37:51 - experiment_save_merged_model - INFO - Available: 122.68 GB
2025-10-09 19:37:51 - experiment_save_merged_model - INFO - GPU 0: Allocated: 12.87 GB, Reserved: 25.74 GB, Total: 94.50 GB
2025-10-09 19:37:51 - experiment_save_merged_model - INFO - Freeing memory from AdaMerge object (task vectors and base params no longer needed)
2025-10-09 19:37:51 - experiment_save_merged_model - INFO - Deleting task vectors
2025-10-09 19:37:51 - experiment_save_merged_model - INFO - Deleting base params
2025-10-09 19:37:51 - experiment_save_merged_model - INFO - Deleting functional model
2025-10-09 19:37:51 - experiment_save_merged_model - INFO - [After freeing AdaMerge memory] Memory Usage:
2025-10-09 19:37:51 - experiment_save_merged_model - INFO - Process: 0.20 GB (0.1%)
2025-10-09 19:37:51 - experiment_save_merged_model - INFO - System: 21.44 GB / 212.49 GB (18.0%)
2025-10-09 19:37:51 - experiment_save_merged_model - INFO - Available: 174.20 GB
2025-10-09 19:37:51 - experiment_save_merged_model - INFO - GPU 0: Allocated: 12.87 GB, Reserved: 12.87 GB, Total: 94.50 GB
2025-10-09 19:37:51 - experiment_save_merged_model - INFO - Saving merged model to /work/gj26/b20042/LLM-AdaMerge/mergekit/outputs/deepseek-7b/llmadamerge/task-wise-cross_entropy-lr0001-ep3-10%dataset/lambda05_wo_scheduler
2025-10-09 19:37:51 - experiment_save_merged_model - INFO - Moving merged model to CPU for saving
2025-10-09 19:38:44 - experiment_save_merged_model - INFO - Successfully saved 3 safetensors files:
2025-10-09 19:38:44 - experiment_save_merged_model - INFO - - model-00003-of-00003.safetensors (3674.14 MB)
2025-10-09 19:38:44 - experiment_save_merged_model - INFO - - model-00002-of-00003.safetensors (4750.20 MB)
2025-10-09 19:38:44 - experiment_save_merged_model - INFO - - model-00001-of-00003.safetensors (4756.17 MB)
2025-10-09 19:38:44 - experiment_save_merged_model - INFO - [After saving model] Memory Usage:
2025-10-09 19:38:44 - experiment_save_merged_model - INFO - Process: 13.08 GB (6.2%)
2025-10-09 19:38:44 - experiment_save_merged_model - INFO - System: 21.55 GB / 212.49 GB (18.1%)
2025-10-09 19:38:44 - experiment_save_merged_model - INFO - Available: 174.09 GB
2025-10-09 19:38:44 - experiment_save_merged_model - INFO - GPU 0: Allocated: 0.00 GB, Reserved: 0.00 GB, Total: 94.50 GB
2025-10-09 19:38:44 - experiment_save_merged_model - INFO - Saving tokenizer
2025-10-09 19:38:45 - experiment_save_merged_model - INFO - Copied lambdas file to /work/gj26/b20042/LLM-AdaMerge/mergekit/outputs/deepseek-7b/llmadamerge/task-wise-cross_entropy-lr0001-ep3-10%dataset/lambda05_wo_scheduler/learned_lambdas.json
2025-10-09 19:38:45 - experiment_save_merged_model - INFO - Creating model card
2025-10-09 19:38:45 - experiment_save_merged_model - INFO - Cleaning up models
2025-10-09 19:38:45 - experiment_save_merged_model - INFO - [After cleanup] Memory Usage:
2025-10-09 19:38:45 - experiment_save_merged_model - INFO - Process: 13.09 GB (6.2%)
2025-10-09 19:38:45 - experiment_save_merged_model - INFO - System: 21.55 GB / 212.49 GB (18.1%)
2025-10-09 19:38:45 - experiment_save_merged_model - INFO - Available: 174.09 GB
2025-10-09 19:38:45 - experiment_save_merged_model - INFO - GPU 0: Allocated: 0.00 GB, Reserved: 0.00 GB, Total: 94.50 GB
2025-10-09 19:38:45 - experiment_save_merged_model - INFO - Model saved successfully to /work/gj26/b20042/LLM-AdaMerge/mergekit/outputs/deepseek-7b/llmadamerge/task-wise-cross_entropy-lr0001-ep3-10%dataset/lambda05_wo_scheduler