parambharat

Parambharat's group workspace

Group: base-south-indic

1-1

of 1

Timestamps visible

2022-12-14 11:09:19

WARNING:param.Curve: Chart elements should only be supplied a single kdim

2022-12-14 11:09:19

WARNING:param.Curve: Chart elements should only be supplied a single kdim

2022-12-14 11:09:21

WARNING:param.Curve: Chart elements should only be supplied a single kdim

2022-12-14 11:09:21

WARNING:param.Curve: Chart elements should only be supplied a single kdim

2022-12-14 11:09:21

WARNING:param.Curve: Chart elements should only be supplied a single kdim

2022-12-14 11:09:23

Configuration saved in ../models/whisper-base-south-indic/config.json

2022-12-14 11:09:23

Model weights saved in ../models/whisper-base-south-indic/pytorch_model.bin

2022-12-14 11:09:23

Feature extractor saved in ../models/whisper-base-south-indic/preprocessor_config.json

2022-12-14 11:09:23

tokenizer config file saved in ../models/whisper-base-south-indic/tokenizer_config.json

2022-12-14 11:09:23

Special tokens file saved in ../models/whisper-base-south-indic/special_tokens_map.json

2022-12-14 11:09:23

added tokens file saved in ../models/whisper-base-south-indic/added_tokens.json

2022-12-14 11:09:23

***** Running training *****

2022-12-14 11:09:23

Num examples = 480000

2022-12-14 11:09:23

Num Epochs = 9223372036854775807

2022-12-14 11:09:23

Instantaneous batch size per device = 96

2022-12-14 11:09:23

Total train batch size (w. parallel, distributed & accumulation) = 96

2022-12-14 11:09:23

Gradient Accumulation steps = 1

2022-12-14 11:09:23

Total optimization steps = 5000

2022-12-14 11:09:23

Number of trainable parameters = 72593920

2022-12-14 11:09:23

Automatic Weights & Biases logging enabled, to disable set os.environ["WANDB_DISABLED"] = "true"