{ "model_type": "yi-1_5-6b", "model_id_or_path": "/home/wusc/Yi-1___5-6B", "model_revision": "master", "full_determinism": false, "sft_type": "full", "freeze_parameters": 0.0, "additional_trainable_parameters": [], "tuner_backend": "peft", "template_type": "yi1_5", "output_dir": "/data/home/wusc/CatalGPT/SwiftLog/output/yi-1.5-6b/v3105-20240827-101309.456842919-nd2-node234", "add_output_dir_suffix": false, "ddp_backend": "nccl", "ddp_find_unused_parameters": null, "ddp_broadcast_buffers": null, "seed": 0, "resume_from_checkpoint": "/data/home/wusc/CatalGPT/SwiftLog/output/yi-1.5-6b/v3100-20240826-173050.909977815-nd2-node234/checkpoint-4250", "resume_only_model": false, "ignore_data_skip": false, "dtype": "fp16", "packing": false, "train_backend": "transformers", "tp": 1, "pp": 1, "min_lr": null, "sequence_parallel": false, "model_kwargs": null, "dataset": [ "wikipedia-zh", "ruozhiba", "/data/home/wusc/CatalGPT/ContinuePreTrain/PDH_ExpandedDOC9133638.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/PDH_ExpandedDOC_CN10215737.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/SampledArXivDOC12264329_1.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/SampledArXivDOC12264329_2.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/SampledArXivDOC12264329_3.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/WikiDOC133016430_1.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/WikiDOC133016430_2.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/WikiDOC133016430_3.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/WikiDOC133016430_4.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/WikiDOC133016430_5.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/ExpandedDOC127305816_10.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/ExpandedDOC127305816_11.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/ExpandedDOC127305816_12.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/ExpandedDOC127305816_13.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/ExpandedDOC127305816_14.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/ExpandedDOC127305816_15.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/ExpandedDOC127305816_16.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/ExpandedDOC127305816_17.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/ExpandedDOC127305816_18.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/ExpandedDOC127305816_19.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/ExpandedDOC127305816_1.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/ExpandedDOC127305816_20.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/ExpandedDOC127305816_21.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/ExpandedDOC127305816_22.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/ExpandedDOC127305816_23.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/ExpandedDOC127305816_24.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/ExpandedDOC127305816_25.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/ExpandedDOC127305816_26.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/ExpandedDOC127305816_27.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/ExpandedDOC127305816_28.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/ExpandedDOC127305816_29.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/ExpandedDOC127305816_2.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/ExpandedDOC127305816_30.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/ExpandedDOC127305816_31.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/ExpandedDOC127305816_32.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/ExpandedDOC127305816_33.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/ExpandedDOC127305816_3.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/ExpandedDOC127305816_4.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/ExpandedDOC127305816_5.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/ExpandedDOC127305816_6.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/ExpandedDOC127305816_7.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/ExpandedDOC127305816_8.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/ExpandedDOC127305816_9.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC_CN187205782_10.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC_CN187205782_11.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC_CN187205782_12.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC_CN187205782_13.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC_CN187205782_14.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC_CN187205782_15.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC_CN187205782_16.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC_CN187205782_17.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC_CN187205782_18.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC_CN187205782_19.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC_CN187205782_1.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC_CN187205782_20.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC_CN187205782_21.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC_CN187205782_22.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC_CN187205782_23.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC_CN187205782_24.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC_CN187205782_25.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC_CN187205782_26.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC_CN187205782_27.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC_CN187205782_28.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC_CN187205782_29.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC_CN187205782_2.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC_CN187205782_30.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC_CN187205782_31.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC_CN187205782_32.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC_CN187205782_33.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC_CN187205782_34.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC_CN187205782_35.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC_CN187205782_36.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC_CN187205782_37.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC_CN187205782_38.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC_CN187205782_39.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC_CN187205782_3.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC_CN187205782_40.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC_CN187205782_4.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC_CN187205782_5.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC_CN187205782_6.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC_CN187205782_7.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC_CN187205782_8.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC_CN187205782_9.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC165694078_10.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC165694078_11.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC165694078_12.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC165694078_13.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC165694078_14.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC165694078_15.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC165694078_16.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC165694078_17.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC165694078_18.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC165694078_19.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC165694078_1.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC165694078_20.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC165694078_21.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC165694078_22.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC165694078_23.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC165694078_24.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC165694078_25.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC165694078_26.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC165694078_27.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC165694078_28.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC165694078_29.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC165694078_2.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC165694078_30.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC165694078_31.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC165694078_32.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC165694078_33.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC165694078_34.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC165694078_35.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC165694078_36.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC165694078_37.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC165694078_38.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC165694078_39.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC165694078_3.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC165694078_40.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC165694078_41.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC165694078_4.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC165694078_5.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC165694078_6.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC165694078_7.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC165694078_8.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/2015_2025Q1ExpandedDOC165694078_9.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/SampledCSL158256764_10.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/SampledCSL158256764_11.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/SampledCSL158256764_12.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/SampledCSL158256764_13.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/SampledCSL158256764_14.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/SampledCSL158256764_15.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/SampledCSL158256764_16.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/SampledCSL158256764_17.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/SampledCSL158256764_18.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/SampledCSL158256764_19.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/SampledCSL158256764_1.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/SampledCSL158256764_20.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/SampledCSL158256764_21.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/SampledCSL158256764_22.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/SampledCSL158256764_23.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/SampledCSL158256764_24.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/SampledCSL158256764_25.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/SampledCSL158256764_2.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/SampledCSL158256764_3.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/SampledCSL158256764_4.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/SampledCSL158256764_5.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/SampledCSL158256764_6.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/SampledCSL158256764_7.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/SampledCSL158256764_8.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/SampledCSL158256764_9.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/LCSTS11735043.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/Calculations1980_2025Q1ExpandedDOC165694078_10.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/Calculations1980_2025Q1ExpandedDOC165694078_11.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/Calculations1980_2025Q1ExpandedDOC165694078_12.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/Calculations1980_2025Q1ExpandedDOC165694078_13.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/Calculations1980_2025Q1ExpandedDOC165694078_14.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/Calculations1980_2025Q1ExpandedDOC165694078_15.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/Calculations1980_2025Q1ExpandedDOC165694078_16.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/Calculations1980_2025Q1ExpandedDOC165694078_17.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/Calculations1980_2025Q1ExpandedDOC165694078_18.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/Calculations1980_2025Q1ExpandedDOC165694078_19.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/Calculations1980_2025Q1ExpandedDOC165694078_1.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/Calculations1980_2025Q1ExpandedDOC165694078_20.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/Calculations1980_2025Q1ExpandedDOC165694078_21.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/Calculations1980_2025Q1ExpandedDOC165694078_22.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/Calculations1980_2025Q1ExpandedDOC165694078_23.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/Calculations1980_2025Q1ExpandedDOC165694078_24.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/Calculations1980_2025Q1ExpandedDOC165694078_25.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/Calculations1980_2025Q1ExpandedDOC165694078_26.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/Calculations1980_2025Q1ExpandedDOC165694078_27.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/Calculations1980_2025Q1ExpandedDOC165694078_28.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/Calculations1980_2025Q1ExpandedDOC165694078_29.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/Calculations1980_2025Q1ExpandedDOC165694078_2.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/Calculations1980_2025Q1ExpandedDOC165694078_30.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/Calculations1980_2025Q1ExpandedDOC165694078_31.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/Calculations1980_2025Q1ExpandedDOC165694078_32.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/Calculations1980_2025Q1ExpandedDOC165694078_33.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/Calculations1980_2025Q1ExpandedDOC165694078_3.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/Calculations1980_2025Q1ExpandedDOC165694078_4.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/Calculations1980_2025Q1ExpandedDOC165694078_5.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/Calculations1980_2025Q1ExpandedDOC165694078_6.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/Calculations1980_2025Q1ExpandedDOC165694078_7.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/Calculations1980_2025Q1ExpandedDOC165694078_8.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/Calculations1980_2025Q1ExpandedDOC165694078_9.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/Calculations1980_2025Q23ExpandedDOC66556562_10.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/Calculations1980_2025Q23ExpandedDOC66556562_11.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/Calculations1980_2025Q23ExpandedDOC66556562_12.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/Calculations1980_2025Q23ExpandedDOC66556562_13.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/Calculations1980_2025Q23ExpandedDOC66556562_14.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/Calculations1980_2025Q23ExpandedDOC66556562_15.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/Calculations1980_2025Q23ExpandedDOC66556562_16.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/Calculations1980_2025Q23ExpandedDOC66556562_17.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/Calculations1980_2025Q23ExpandedDOC66556562_18.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/Calculations1980_2025Q23ExpandedDOC66556562_1.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/Calculations1980_2025Q23ExpandedDOC66556562_2.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/Calculations1980_2025Q23ExpandedDOC66556562_3.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/Calculations1980_2025Q23ExpandedDOC66556562_4.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/Calculations1980_2025Q23ExpandedDOC66556562_5.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/Calculations1980_2025Q23ExpandedDOC66556562_6.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/Calculations1980_2025Q23ExpandedDOC66556562_7.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/Calculations1980_2025Q23ExpandedDOC66556562_8.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/Split/Calculations1980_2025Q23ExpandedDOC66556562_9.csv", "/data/home/wusc/CatalGPT/ContinuePreTrain/science-theory-textbooks/science_theory_textbooks73026.csv", "math-instruct", "dolly-15k", "ultrachat-200k", "alpaca-cleaned", "no-robots", "slim-orca", "cosmopedia-100k", "coig-cqia:ruozhiba", "coig-cqia:exam", "coig-cqia:segmentfault", "coig-cqia:logi_qa", "coig-cqia:wikihow", "cot-en", "cot-zh", "blossom-math-zh", "leetcode-python-en", "codefuse-python-en", "codefuse-evol-instruction-zh", "alpaca-zh", "/data/home/wusc/CatalGPT/SupervisedFineTuning/PDH_QAPairFromARG_English1985.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/PDH_QAPairFromARG_Chinese1985.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/BackgroundKnowledgeQAPairs51790.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/gaokao-benchmark/biology.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/gaokao-benchmark/chemistry.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/gaokao-benchmark/chinese.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/gaokao-benchmark/english.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/gaokao-benchmark/geography.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/gaokao-benchmark/mathcloze.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/gaokao-benchmark/mathqa.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/gaokao-benchmark/MultiChoices.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/gaokao-benchmark/physics.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/CSL396209.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/scimrc21297.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/MultiHopRAG/MultiHopRAG.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/physical_reasoning/physical_reasoning.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/physical_reasoning/physical_reasoning_longer.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/physical_reasoning/physical_reasoning_magic.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/race/race_high_test-00000-of-00001.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/race/race_high_train-00000-of-00001.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/race/race_high_validation-00000-of-00001.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/race/race_middle_test-00000-of-00001.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/race/race_middle_train-00000-of-00001.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/race/race_middle_validation-00000-of-00001.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/Smaug/SystemConversationsTransformed.jsonl", "/data/home/wusc/CatalGPT/SupervisedFineTuning/Smaug/WikiQA-Altered_Numeric_QA_16K.jsonl", "/data/home/wusc/CatalGPT/SupervisedFineTuning/Smaug/WikiQA-Altered_Numeric_QA_2K.jsonl", "/data/home/wusc/CatalGPT/SupervisedFineTuning/Smaug/WikiQA-Altered_Numeric_QA_4K.jsonl", "/data/home/wusc/CatalGPT/SupervisedFineTuning/Smaug/WikiQA-Altered_Numeric_QA_8K.jsonl", "/data/home/wusc/CatalGPT/SupervisedFineTuning/Smaug/WikiQA-Free_Form_QA_16K.jsonl", "/data/home/wusc/CatalGPT/SupervisedFineTuning/Smaug/WikiQA-Free_Form_QA_2K.jsonl", "/data/home/wusc/CatalGPT/SupervisedFineTuning/Smaug/WikiQA-Free_Form_QA_4K.jsonl", "/data/home/wusc/CatalGPT/SupervisedFineTuning/Smaug/WikiQA-Free_Form_QA_8K.jsonl", "/data/home/wusc/CatalGPT/SupervisedFineTuning/truthful_qa/truthful_qa_multiple_choice.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/commonsense_qa/commonsense_qa_default_train_0000.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/commonsense_qa/commonsense_qa_default_validation_0000.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/LogiQA/LogiQA.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/ScienceQA/ScienceQA.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/M3KE/M3KE_dev.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/firefly-train-1.1M/ToTrain/firefly-train-1.1M_BELLE_543285.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/firefly-train-1.1M/ToTrain/firefly-train-1.1M_Cot_74771.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/firefly-train-1.1M/ToTrain/firefly-train-1.1M_Dictionary_30895.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/firefly-train-1.1M/ToTrain/firefly-train-1.1M_KeywordRecognition_50000.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/firefly-train-1.1M/ToTrain/firefly-train-1.1M_MRC_70000.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/firefly-train-1.1M/ToTrain/firefly-train-1.1M_NER_50000.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/firefly-train-1.1M/ToTrain/firefly-train-1.1M_NLI_50000.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/firefly-train-1.1M/ToTrain/firefly-train-1.1M_OpenQA_69843.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/firefly-train-1.1M/ToTrain/firefly-train-1.1M_Program_974.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/firefly-train-1.1M/ToTrain/firefly-train-1.1M_TextMatching_50000.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/firefly-train-1.1M/ToTrain/firefly-train-1.1M_Translation_50000.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/ChemData700K/ChemData700K677392.jsonl", "/data/home/wusc/CatalGPT/SupervisedFineTuning/AI4Chem/ChemBench4K_default_test_0000.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/DataHammer_emotional_dialog1069/DataHammer_emotional_dialog1069.jsonl", "/data/home/wusc/CatalGPT/SupervisedFineTuning/arxiv-physics/arxiv_physics_30231.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/camel-ai/camel_ai_chemistry_20k_Analytical chemistry_800.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/camel-ai/camel_ai_chemistry_20k_Biochemistry_800.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/camel-ai/camel_ai_chemistry_20k_Chemical bonding_800.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/camel-ai/camel_ai_chemistry_20k_Chemical education_800.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/camel-ai/camel_ai_chemistry_20k_Chemical engineering_800.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/camel-ai/camel_ai_chemistry_20k_Chemical equilibrium_800.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/camel-ai/camel_ai_chemistry_20k_Chemical kinetics_800.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/camel-ai/camel_ai_chemistry_20k_Chemical reactions_800.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/camel-ai/camel_ai_chemistry_20k_Chemical synthesis_800.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/camel-ai/camel_ai_chemistry_20k_Chemical thermodynamics_800.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/camel-ai/camel_ai_chemistry_20k_Computational chemistry_800.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/camel-ai/camel_ai_chemistry_20k_Coordination chemistry_800.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/camel-ai/camel_ai_chemistry_20k_Electrochemistry_800.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/camel-ai/camel_ai_chemistry_20k_Environmental chemistry_800.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/camel-ai/camel_ai_chemistry_20k_Inorganic chemistry_800.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/camel-ai/camel_ai_chemistry_20k_Materials chemistry_800.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/camel-ai/camel_ai_chemistry_20k_Medicinal chemistry_800.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/camel-ai/camel_ai_chemistry_20k_Nuclear chemistry_800.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/camel-ai/camel_ai_chemistry_20k_Organic chemistry_800.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/camel-ai/camel_ai_chemistry_20k_Photochemistry_800.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/camel-ai/camel_ai_chemistry_20k_Physical chemistry_800.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/camel-ai/camel_ai_chemistry_20k_Polymer chemistry_800.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/camel-ai/camel_ai_chemistry_20k_Quantum chemistry_800.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/camel-ai/camel_ai_chemistry_20k_Surface chemistry_800.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/camel-ai/camel_ai_chemistry_20k_Thermochemistry_800.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/camel-ai/camel_ai_physics_20k_Astrophysics_800.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/camel-ai/camel_ai_physics_20k_Atomic physics_800.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/camel-ai/camel_ai_physics_20k_Biophysics_800.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/camel-ai/camel_ai_physics_20k_Black holes_800.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/camel-ai/camel_ai_physics_20k_Chaos theory_800.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/camel-ai/camel_ai_physics_20k_Condensed matter physics_800.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/camel-ai/camel_ai_physics_20k_Cosmology_800.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/camel-ai/camel_ai_physics_20k_Dark matter_800.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/camel-ai/camel_ai_physics_20k_Electromagnetism_800.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/camel-ai/camel_ai_physics_20k_Fluid mechanics_800.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/camel-ai/camel_ai_physics_20k_General relativity_800.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/camel-ai/camel_ai_physics_20k_Gravitational waves_800.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/camel-ai/camel_ai_physics_20k_High-energy physics_800.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/camel-ai/camel_ai_physics_20k_Nuclear physics_800.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/camel-ai/camel_ai_physics_20k_Optics_800.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/camel-ai/camel_ai_physics_20k_Particle physics_800.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/camel-ai/camel_ai_physics_20k_Plasma physics_800.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/camel-ai/camel_ai_physics_20k_Quantum computing_800.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/camel-ai/camel_ai_physics_20k_Quantum field theory_800.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/camel-ai/camel_ai_physics_20k_Quantum mechanics_800.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/camel-ai/camel_ai_physics_20k_Solid state physics_800.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/camel-ai/camel_ai_physics_20k_Special relativity_800.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/camel-ai/camel_ai_physics_20k_String theory_800.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/camel-ai/camel_ai_physics_20k_Superconductivity_800.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/camel-ai/camel_ai_physics_20k_Thermodynamics_800.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/Chem-Catalysis-Chat/Chem_Catalysis_Chat_189168.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/physics-scienceqa/physics_scienceqa810.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/UltraInteract/UltraInteract_288579_Coding_codecontest_44662.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/UltraInteract/UltraInteract_288579_Coding_magicoder_evol_instruct_10238.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/UltraInteract/UltraInteract_288579_Coding_TACO_58188.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/UltraInteract/UltraInteract_288579_Coding_wiki_table_questions_1738.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/UltraInteract/UltraInteract_288579_Logic_hotpotqa_2586.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/UltraInteract/UltraInteract_288579_Logic_reclor_7159.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/UltraInteract/UltraInteract_288579_Logic_strategyqa_1746.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/UltraInteract/UltraInteract_288579_Math_CoT_gsm8k_15733.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/UltraInteract/UltraInteract_288579_Math_CoT_MATH_36004.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/UltraInteract/UltraInteract_288579_Math_CoT_mathqa_22496.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/UltraInteract/UltraInteract_288579_Math_CoT_numglue_4116.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/UltraInteract/UltraInteract_288579_Math_PoT_gsm8k_17356.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/UltraInteract/UltraInteract_288579_Math_PoT_MATH_34667.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/UltraInteract/UltraInteract_288579_Math_PoT_mathqa_20328.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/UltraInteract/UltraInteract_288579_Math_PoT_numglue_5479.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/UltraInteract/UltraInteract_288579_Math_PoT_tabmwp_6083.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/YouToks-Instruct-Quantum-Physics/YouToks_Instruct_Quantum_Physics3109.csv", "/data/home/wusc/CatalGPT/SupervisedFineTuning/LLM_Science_Dataset/Kaggle_LLM_Science_Dataset8835.csv", "self-cognition#32768" ], "val_dataset": [], "dataset_seed": 0, "dataset_test_ratio": 0.01, "use_loss_scale": false, "loss_scale_config_path": "/data/home/wusc/CatalGPT/GitLibrary/swift/swift/llm/agent/default_loss_scale_config.json", "system": "You are an expert in the field of chemistry and chemical engineering. Explain scientific concepts, theories, and phenomena in an engaging and accessible way. Take a deep breath and think step by step, which is very import to my career.", "tools_prompt": "react_en", "max_length": 4096, "truncation_strategy": "delete", "check_dataset_strategy": "warning", "streaming": false, "streaming_val_size": 0, "streaming_buffer_size": 16384, "model_name": [ "北洋微衍", "Peiyang MicroDerivation" ], "model_author": [ "天津大学", "Tianjin University" ], "quant_method": null, "quantization_bit": 0, "hqq_axis": 0, "hqq_dynamic_config_path": null, "bnb_4bit_comp_dtype": "fp16", "bnb_4bit_quant_type": "nf4", "bnb_4bit_use_double_quant": true, "bnb_4bit_quant_storage": null, "rescale_image": -1, "target_modules": [ "ALL", "EMBEDDING" ], "target_regex": null, "modules_to_save": [], "lora_rank": 64, "lora_alpha": 128, "lora_dropout": 0.05, "lora_bias_trainable": "none", "lora_dtype": "fp16", "lora_lr_ratio": null, "use_rslora": true, "use_dora": false, "init_lora_weights": "true", "fourier_n_frequency": 2000, "fourier_scaling": 300.0, "rope_scaling": null, "boft_block_size": 4, "boft_block_num": 0, "boft_n_butterfly_factor": 1, "boft_dropout": 0.0, "vera_rank": 256, "vera_projection_prng_key": 0, "vera_dropout": 0.0, "vera_d_initial": 0.1, "adapter_act": "gelu", "adapter_length": 128, "use_galore": false, "galore_target_modules": null, "galore_rank": 128, "galore_update_proj_gap": 50, "galore_scale": 1.0, "galore_proj_type": "std", "galore_optim_per_parameter": false, "galore_with_embedding": false, "galore_quantization": false, "galore_proj_quant": false, "galore_proj_bits": 4, "galore_proj_group_size": 256, "galore_cos_threshold": 0.4, "galore_gamma_proj": 2, "galore_queue_size": 5, "adalora_target_r": 8, "adalora_init_r": 12, "adalora_tinit": 0, "adalora_tfinal": 0, "adalora_deltaT": 1, "adalora_beta1": 0.85, "adalora_beta2": 0.85, "adalora_orth_reg_weight": 0.5, "ia3_feedforward_modules": [], "llamapro_num_new_blocks": 4, "llamapro_num_groups": null, "neftune_noise_alpha": 5.0, "neftune_backend": "transformers", "lisa_activated_layers": 0, "lisa_step_interval": 20, "reft_layers": null, "reft_rank": 4, "reft_intervention_type": "LoreftIntervention", "reft_args": null, "gradient_checkpointing": true, "deepspeed": null, "batch_size": 1, "eval_batch_size": 1, "auto_find_batch_size": false, "num_train_epochs": 3, "max_steps": -1, "optim": "adamw_torch", "adam_beta1": 0.9, "adam_beta2": 0.999, "adam_epsilon": 1e-07, "learning_rate": 0.000128, "weight_decay": 0.1, "gradient_accumulation_steps": 32, "max_grad_norm": 0.5, "predict_with_generate": false, "lr_scheduler_type": "cosine", "lr_scheduler_kwargs": {}, "warmup_ratio": 0.1, "warmup_steps": 0, "eval_steps": 250, "save_steps": 250, "save_only_model": false, "save_total_limit": 2, "logging_steps": 5, "acc_steps": 1, "dataloader_num_workers": 48, "dataloader_pin_memory": true, "dataloader_drop_last": false, "push_to_hub": false, "hub_model_id": null, "hub_token": null, "hub_private_repo": false, "push_hub_strategy": "push_best", "test_oom_error": false, "disable_tqdm": false, "lazy_tokenize": true, "preprocess_num_proc": 48, "use_flash_attn": false, "ignore_args_error": false, "check_model_is_latest": false, "logging_dir": "/data/home/wusc/CatalGPT/SwiftLog/output/yi-1.5-6b/v3105-20240827-101309.456842919-nd2-node234/runs", "report_to": [ "tensorboard" ], "acc_strategy": "token", "save_on_each_node": false, "evaluation_strategy": "steps", "save_strategy": "steps", "save_safetensors": true, "gpu_memory_fraction": null, "include_num_input_tokens_seen": false, "local_repo_path": null, "custom_register_path": null, "custom_dataset_info": null, "device_map_config_path": null, "device_max_memory": [], "max_new_tokens": 2048, "do_sample": true, "temperature": 0.3, "top_k": 20, "top_p": 0.7, "repetition_penalty": 1.0, "num_beams": 1, "fsdp": "", "fsdp_config": null, "sequence_parallel_size": 1, "model_layer_cls_name": null, "metric_warmup_step": 0, "fsdp_num": 1, "per_device_train_batch_size": null, "per_device_eval_batch_size": null, "eval_strategy": null, "self_cognition_sample": 32768, "train_dataset_mix_ratio": 0.0, "train_dataset_mix_ds": [ "ms-bench" ], "train_dataset_sample": -1, "val_dataset_sample": -1, "safe_serialization": null, "only_save_model": null, "neftune_alpha": null, "deepspeed_config_path": null, "model_cache_dir": null, "lora_dropout_p": null, "lora_target_modules": [ "ALL", "EMBEDDING" ], "lora_target_regex": null, "lora_modules_to_save": [], "boft_target_modules": [], "boft_modules_to_save": [], "vera_target_modules": [], "vera_modules_to_save": [], "ia3_target_modules": [], "ia3_modules_to_save": [], "custom_train_dataset_path": [], "custom_val_dataset_path": [], "use_self_cognition": true, "is_multimodal": false, "is_vision": false, "lora_use_embedding": true, "lora_use_all": true, "lora_m2s_use_embedding": false, "lora_m2s_use_ln": false, "torch_dtype": "torch.float16", "fp16": true, "bf16": false, "rank": 0, "local_rank": 0, "world_size": 112, "local_world_size": 8, "bnb_4bit_compute_dtype": "torch.float16", "load_in_4bit": false, "load_in_8bit": false, "train_sampler_random": true, "training_args": "Seq2SeqTrainingArguments(output_dir='/data/home/wusc/CatalGPT/SwiftLog/output/yi-1.5-6b/v3105-20240827-101309.456842919-nd2-node234', overwrite_output_dir=False, do_train=False, do_eval=True, do_predict=False, eval_strategy=, prediction_loss_only=False, per_device_train_batch_size=1, per_device_eval_batch_size=1, per_gpu_train_batch_size=None, per_gpu_eval_batch_size=None, gradient_accumulation_steps=32, eval_accumulation_steps=None, eval_delay=0, learning_rate=0.000128, weight_decay=0.1, adam_beta1=0.9, adam_beta2=0.999, adam_epsilon=1e-07, max_grad_norm=0.5, num_train_epochs=3, max_steps=-1, lr_scheduler_type=, lr_scheduler_kwargs={}, warmup_ratio=0.1, warmup_steps=0, log_level='passive', log_level_replica='warning', log_on_each_node=True, logging_dir='/data/home/wusc/CatalGPT/SwiftLog/output/yi-1.5-6b/v3105-20240827-101309.456842919-nd2-node234/runs', logging_strategy=, logging_first_step=True, logging_steps=5, logging_nan_inf_filter=True, save_strategy=, save_steps=250, save_total_limit=2, save_safetensors=True, save_on_each_node=False, save_only_model=False, restore_callback_states_from_checkpoint=False, no_cuda=False, use_cpu=False, use_mps_device=False, seed=0, data_seed=None, jit_mode_eval=False, use_ipex=False, bf16=False, fp16=True, fp16_opt_level='O1', half_precision_backend='auto', bf16_full_eval=False, fp16_full_eval=False, tf32=None, local_rank=0, ddp_backend='nccl', tpu_num_cores=None, tpu_metrics_debug=False, debug=[], dataloader_drop_last=False, eval_steps=250, dataloader_num_workers=48, dataloader_prefetch_factor=None, past_index=-1, run_name='/data/home/wusc/CatalGPT/SwiftLog/output/yi-1.5-6b/v3105-20240827-101309.456842919-nd2-node234', disable_tqdm=False, remove_unused_columns=False, label_names=None, load_best_model_at_end=False, metric_for_best_model='loss', greater_is_better=False, ignore_data_skip=False, fsdp=[], fsdp_min_num_params=0, fsdp_config={'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, fsdp_transformer_layer_cls_to_wrap=None, accelerator_config=AcceleratorConfig(split_batches=False, dispatch_batches=False, even_batches=True, use_seedable_sampler=True, non_blocking=False, gradient_accumulation_kwargs=None, use_configured_state=False), deepspeed=None, label_smoothing_factor=0.0, optim=, optim_args=None, adafactor=False, group_by_length=False, length_column_name='length', report_to=['tensorboard'], ddp_find_unused_parameters=False, ddp_bucket_cap_mb=None, ddp_broadcast_buffers=False, dataloader_pin_memory=True, dataloader_persistent_workers=False, skip_memory_metrics=True, use_legacy_prediction_loop=False, push_to_hub=False, resume_from_checkpoint='/data/home/wusc/CatalGPT/SwiftLog/output/yi-1.5-6b/v3100-20240826-173050.909977815-nd2-node234/checkpoint-4250', hub_model_id=None, hub_strategy=, hub_token=None, hub_private_repo=False, hub_always_push=False, gradient_checkpointing=True, gradient_checkpointing_kwargs=None, include_inputs_for_metrics=False, eval_do_concat_batches=True, fp16_backend='auto', evaluation_strategy=None, push_to_hub_model_id=None, push_to_hub_organization=None, push_to_hub_token=None, mp_parameters='', auto_find_batch_size=False, full_determinism=False, torchdynamo=None, ray_scope='last', ddp_timeout=1800, torch_compile=False, torch_compile_backend=None, torch_compile_mode=None, dispatch_batches=None, split_batches=None, include_tokens_per_second=False, include_num_input_tokens_seen=False, neftune_noise_alpha=5.0, optim_target_modules=None, batch_eval_metrics=False, eval_on_start=False, sortish_sampler=True, predict_with_generate=False, generation_max_length=None, generation_num_beams=None, generation_config=GenerationConfig {\n \"do_sample\": true,\n \"eos_token_id\": 2,\n \"max_new_tokens\": 2048,\n \"pad_token_id\": 0,\n \"temperature\": 0.3,\n \"top_k\": 20,\n \"top_p\": 0.7\n}\n, train_sampler_random=True, push_hub_strategy='push_best', acc_strategy='token', additional_saved_files=[], metric_warmup_step=0, train_dataset_sample=-1)" }