clap-laion-music

[babisspgit]

The 'clap-laion-music' argument in the fadtk in contrast to the 'clap-laion-audio' version, is not recognized in this model and gives the following error:

fadtk: error: unrecognized arguments: clap-laion-music /work3/s22/data/processed/metrics/fad/0Real/MJ /work3/s22/data/processed/metrics/fad/0Gen/MJ/music --inf
usage: fadtk [-h] [--train-data TRAIN_DATA] [--val-data VAL_DATA] [--freeze-text] [--freeze-text-after FREEZE_TEXT_AFTER] [--train-ipc TRAIN_IPC] [--val-ipc VAL_IPC] [--train-num-samples TRAIN_NUM_SAMPLES]
[--val-num-samples VAL_NUM_SAMPLES] [--dataset-type {webdataset,csv,auto,toy}] [--csv-separator CSV_SEPARATOR] [--csv-img-key CSV_IMG_KEY] [--csv-caption-key CSV_CAPTION_KEY] [--imagenet-val IMAGENET_VAL]
[--imagenet-v2 IMAGENET_V2] [--datasetnames DATASETNAMES [DATASETNAMES ...]] [--full-train-dataset FULL_TRAIN_DATASET [FULL_TRAIN_DATASET ...]]
[--exclude-eval-dataset EXCLUDE_EVAL_DATASET [EXCLUDE_EVAL_DATASET ...]] [--datasetinfos DATASETINFOS [DATASETINFOS ...]] [--dataset-proportion DATASET_PROPORTION] [--remotedata]
[--class-label-path CLASS_LABEL_PATH] [--datasetpath DATASETPATH] [--logs LOGS] [--log-local] [--name NAME] [--workers WORKERS] [--batch-size BATCH_SIZE] [--epochs EPOCHS] [--lr LR] [--beta1 BETA1] [--beta2 BETA2]
[--eps EPS] [--momentum MOMENTUM] [--wd WD] [--split-opt] [--lr-pretrained LR_PRETRAINED] [--beta1-pretrained BETA1_PRETRAINED] [--beta2-pretrained BETA2_PRETRAINED] [--eps-pretrained EPS_PRETRAINED]
[--wd-pretrained WD_PRETRAINED] [--momentum-pretrained MOMENTUM_PRETRAINED] [--lr-new LR_NEW] [--beta1-new BETA1_NEW] [--beta2-new BETA2_NEW] [--eps-new EPS_NEW] [--wd-new WD_NEW] [--momentum-new MOMENTUM_NEW]
[--warmup WARMUP] [--use-bn-sync] [--skip-scheduler] [--save-frequency SAVE_FREQUENCY] [--save-top-performance SAVE_TOP_PERFORMANCE] [--save-most-recent] [--zeroshot-frequency ZEROSHOT_FREQUENCY]
[--val-frequency VAL_FREQUENCY] [--resume RESUME] [--precision {amp,fp16,fp32}] [--amodel AMODEL] [--tmodel TMODEL] [--pretrained-audio PRETRAINED_AUDIO] [--pretrained-text PRETRAINED_TEXT] [--pretrained PRETRAINED]
[--pretrained-image] [--lock-image] [--lock-image-unlocked-groups LOCK_IMAGE_UNLOCKED_GROUPS] [--lock-image-freeze-bn-stats] [--local-loss] [--gather-with-grad] [--force-quick-gelu] [--torchscript] [--trace]
[--dist-url DIST_URL] [--dist-backend DIST_BACKEND] [--report-to REPORT_TO] [--wandb-notes WANDB_NOTES] [--C C] [--debug] [--copy-codebase] [--horovod] [--ddp-static-graph] [--no-set-device-rank] [--seed SEED]
[--top-k-checkpoint-select-dataset TOP_K_CHECKPOINT_SELECT_DATASET] [--top-k-checkpoint-select-metric TOP_K_CHECKPOINT_SELECT_METRIC] [--openai-model-cache-dir OPENAI_MODEL_CACHE_DIR] [--optimizer OPTIMIZER]
[--parallel-eval] [--no-eval] [--lp-mlp] [--lp-freeze] [--lp-act LP_ACT] [--lp-loss LP_LOSS] [--lp-metrics LP_METRICS] [--lp-lr LP_LR] [--kappa KAPPA] [--data-filling DATA_FILLING]
[--data-truncating DATA_TRUNCATING] [--clap-mlploss] [--wandb-id WANDB_ID] [--sleep SLEEP] [--enable-fusion] [--fusion-type FUSION_TYPE] [--mixup] [--text-augment-selection TEXT_AUGMENT_SELECTION]
[--prefetch-factor PREFETCH_FACTOR]

[hykilpikonna]

That does not look like the fadtk command-line interface... the command seems to be mapped to a wrong executable. For reference, below are the expected command line arguments to fadtk:

Can you check where the fadtk command is actually running from using the following commands?

which fadtk
cat $(which fadtk)
pip show fadtk