Update README.md
998bc0d
-
language_model
feat: Replace Reddit-da LM with lex.dk LM
-
1.41 kB
feat: Replace Reddit-da LM with lex.dk LM
-
13 Bytes
Training in progress, step 300
-
1.37 kB
Update README.md
-
238 Bytes
feat: Upload LM-boosted decoder
-
2.07 kB
Training in progress, step 300
-
262 Bytes
feat: Upload LM-boosted decoder
-
1.26 GB
End of training
-
85 Bytes
Training in progress, step 300
-
342 Bytes
feat: Upload LM-boosted decoder
training_args.bin
Detected Pickle imports (6)
- "transformers.training_args.OptimizerNames",
- "transformers.trainer_utils.HubStrategy",
- "transformers.trainer_utils.IntervalStrategy",
- "transformers.training_args.TrainingArguments",
- "transformers.trainer_utils.SchedulerType",
- "torch.device"
How to fix it?
3.06 kB
Training in progress, step 8100
-
323 Bytes
Training in progress, step 8100