How to check train results on tensorboard? #874
Unanswered
elcolex777
asked this question in
Q&A
Replies: 0 comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
Hello.
What do you say about my train results on tensorboard? I’m beginner and don’t understend how to check to stop train model. I have this results after 12 hours of training (see screenshot). I have model with 3280 epochs, but results of inferring not good((
I setup treining that:
!svc pre-split !svc pre-resample !svc pre-config !svc pre-hubert -fm crepe
my config this:
{ "train": { "log_interval": 100, "eval_interval": 200, "seed": 1234, "epochs": 20000, "learning_rate": 0.0001, "betas": [ 0.8, 0.99 ], "eps": 1e-09, "batch_size": 16, "fp16_run": false, "bf16_run": false, "lr_decay": 0.999875, "segment_size": 10240, "init_lr_ratio": 1, "warmup_epochs": 0, "c_mel": 45, "c_kl": 1.0, "use_sr": true, "max_speclen": 512, "port": "8001", "keep_ckpts": 3, "num_workers": 4, "log_version": 0, "ckpt_name_by_step": false, "accumulate_grad_batches": 1 },
do I need more training? I think next session with 12 hours give me model approx 7000 epochs. But I don’t know will be results better. Or may be I do something wrong? Please help. Thanks.
Beta Was this translation helpful? Give feedback.
All reactions