| { | |
| "model_type": "yi-1_5-6b", | |
| "model_id_or_path": "01ai/Yi-1.5-6B", | |
| "model_revision": "master", | |
| "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/v104-20240614-002401.696642908", | |
| "add_output_dir_suffix": false, | |
| "ddp_backend": "nccl", | |
| "ddp_find_unused_parameters": null, | |
| "ddp_broadcast_buffers": null, | |
| "seed": 0, | |
| "resume_from_checkpoint": null, | |
| "resume_only_model": false, | |
| "ignore_data_skip": false, | |
| "dtype": "fp16", | |
| "packing": false, | |
| "dataset": [ | |
| "wikipedia-zh", | |
| "ruozhiba:post-annual", | |
| "ruozhiba:title-good", | |
| "ruozhiba:title-norm", | |
| "/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", | |
| "coig-cqia:ruozhiba", | |
| "coig-cqia:exam", | |
| "coig-cqia:segmentfault", | |
| "cot-en", | |
| "cot-zh", | |
| "blossom-math-zh", | |
| "leetcode-python-en", | |
| "codefuse-python-en", | |
| "alpaca-en", | |
| "alpaca-zh", | |
| "sharegpt-gpt4", | |
| "/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", | |
| "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", | |
| "model_name": [ | |
| "非空", | |
| "Non null" | |
| ], | |
| "model_author": [ | |
| "空", | |
| "null" | |
| ], | |
| "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, | |
| "lora_target_modules": [ | |
| "model.embed_tokens", | |
| "v_proj", | |
| "o_proj", | |
| "down_proj", | |
| "k_proj", | |
| "q_proj", | |
| "up_proj", | |
| "gate_proj" | |
| ], | |
| "lora_rank": 64, | |
| "lora_alpha": 128, | |
| "lora_dropout_p": 0.05, | |
| "lora_bias_trainable": "none", | |
| "lora_modules_to_save": [], | |
| "lora_dtype": "fp16", | |
| "lora_lr_ratio": null, | |
| "use_rslora": true, | |
| "use_dora": false, | |
| "init_lora_weights": true, | |
| "rope_scaling": null, | |
| "boft_block_size": 4, | |
| "boft_block_num": 0, | |
| "boft_n_butterfly_factor": 1, | |
| "boft_target_modules": [ | |
| "DEFAULT" | |
| ], | |
| "boft_dropout": 0.0, | |
| "boft_modules_to_save": [], | |
| "vera_rank": 256, | |
| "vera_target_modules": [ | |
| "DEFAULT" | |
| ], | |
| "vera_projection_prng_key": 0, | |
| "vera_dropout": 0.0, | |
| "vera_d_initial": 0.1, | |
| "vera_modules_to_save": [], | |
| "adapter_act": "gelu", | |
| "adapter_length": 128, | |
| "use_galore": false, | |
| "galore_rank": 128, | |
| "galore_target_modules": null, | |
| "galore_update_proj_gap": 50, | |
| "galore_scale": 1.0, | |
| "galore_proj_type": "std", | |
| "galore_optim_per_parameter": false, | |
| "galore_with_embedding": false, | |
| "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_target_modules": [ | |
| "DEFAULT" | |
| ], | |
| "ia3_feedforward_modules": [], | |
| "ia3_modules_to_save": [], | |
| "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, | |
| "gradient_checkpointing": true, | |
| "deepspeed": null, | |
| "batch_size": 2, | |
| "eval_batch_size": 2, | |
| "num_train_epochs": 3, | |
| "max_steps": -1, | |
| "optim": "adamw_torch", | |
| "adam_beta1": 0.9, | |
| "adam_beta2": 0.999, | |
| "adam_epsilon": 1e-06, | |
| "learning_rate": 0.00016, | |
| "weight_decay": 0.1, | |
| "gradient_accumulation_steps": 32, | |
| "max_grad_norm": 0.5, | |
| "predict_with_generate": false, | |
| "lr_scheduler_type": "cosine", | |
| "warmup_ratio": 0.1, | |
| "eval_steps": 128, | |
| "save_steps": 128, | |
| "save_only_model": false, | |
| "save_total_limit": 2, | |
| "logging_steps": 5, | |
| "dataloader_num_workers": 4, | |
| "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": 2, | |
| "use_flash_attn": false, | |
| "ignore_args_error": false, | |
| "check_model_is_latest": true, | |
| "logging_dir": "/data/home/wusc/CatalGPT/SwiftLog/output/yi-1.5-6b/v104-20240614-002401.696642908/runs", | |
| "report_to": [ | |
| "tensorboard" | |
| ], | |
| "acc_strategy": "token", | |
| "save_on_each_node": true, | |
| "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, | |
| "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, | |
| "custom_train_dataset_path": [], | |
| "custom_val_dataset_path": [], | |
| "use_self_cognition": true, | |
| "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, | |
| "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/v104-20240614-002401.696642908', overwrite_output_dir=False, do_train=False, do_eval=True, do_predict=False, eval_strategy=<IntervalStrategy.STEPS: 'steps'>, prediction_loss_only=False, per_device_train_batch_size=2, per_device_eval_batch_size=2, 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=0.00016, weight_decay=0.1, adam_beta1=0.9, adam_beta2=0.999, adam_epsilon=1e-06, max_grad_norm=0.5, num_train_epochs=3, max_steps=-1, lr_scheduler_type=<SchedulerType.COSINE: 'cosine'>, 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/v104-20240614-002401.696642908/runs', logging_strategy=<IntervalStrategy.STEPS: 'steps'>, logging_first_step=True, logging_steps=5, logging_nan_inf_filter=True, save_strategy=<IntervalStrategy.STEPS: 'steps'>, save_steps=128, save_total_limit=2, save_safetensors=True, save_on_each_node=True, save_only_model=False, restore_callback_states_from_checkpoint=False, no_cuda=False, use_cpu=False, use_mps_device=False, seed=42, 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=128, dataloader_num_workers=4, dataloader_prefetch_factor=None, past_index=-1, run_name='/data/home/wusc/CatalGPT/SwiftLog/output/yi-1.5-6b/v104-20240614-002401.696642908', 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=None, even_batches=True, use_seedable_sampler=True, non_blocking=False, gradient_accumulation_kwargs=None), deepspeed=None, label_smoothing_factor=0.0, optim=<OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, 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=None, hub_model_id=None, hub_strategy=<HubStrategy.EVERY_SAVE: 'every_save'>, 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, 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=7127707)" | |
| } |