508 MB

Ctrl+K

1 contributor

upload model weights after 2400 training steps with frozen pretrained weights, this performs worse than checkpoint 1000 on test set T_T

ceed5bc verified about 1 month ago

config.json

668 Bytes
upload model weights after 2400 training steps with frozen pretrained weights, this performs worse than checkpoint 1000 on test set T_T about 1 month ago
merges.txt

456 kB
upload model weights after 2400 training steps with frozen pretrained weights, this performs worse than checkpoint 1000 on test set T_T about 1 month ago
model.safetensors

499 MB
xet

upload model weights after 2400 training steps with frozen pretrained weights, this performs worse than checkpoint 1000 on test set T_T about 1 month ago
optimizer.pt
Detected Pickle imports (3)
- "torch._utils._rebuild_tensor_v2",
- "collections.OrderedDict",
- "torch.FloatStorage"
What is a pickle import?
4.74 MB
xet

upload model weights after 2400 training steps with frozen pretrained weights, this performs worse than checkpoint 1000 on test set T_T about 1 month ago
rng_state.pth
Detected Pickle imports (7)
- "collections.OrderedDict",
- "numpy.ndarray",
- "_codecs.encode",
- "numpy._core.multiarray._reconstruct",
- "torch._utils._rebuild_tensor_v2",
- "torch.ByteStorage",
- "numpy.dtype"
How to fix it?
14.6 kB
xet

upload model weights after 2400 training steps with frozen pretrained weights, this performs worse than checkpoint 1000 on test set T_T about 1 month ago
scaler.pt
Pickle imports
- No problematic imports detected
What is a pickle import?
1.38 kB
xet

upload model weights after 2400 training steps with frozen pretrained weights, this performs worse than checkpoint 1000 on test set T_T about 1 month ago
scheduler.pt
Pickle imports
- No problematic imports detected
What is a pickle import?
1.47 kB
xet

upload model weights after 2400 training steps with frozen pretrained weights, this performs worse than checkpoint 1000 on test set T_T about 1 month ago
special_tokens_map.json

958 Bytes
upload model weights after 2400 training steps with frozen pretrained weights, this performs worse than checkpoint 1000 on test set T_T about 1 month ago
tokenizer.json

3.56 MB
upload model weights after 2400 training steps with frozen pretrained weights, this performs worse than checkpoint 1000 on test set T_T about 1 month ago
tokenizer_config.json

1.25 kB
upload model weights after 2400 training steps with frozen pretrained weights, this performs worse than checkpoint 1000 on test set T_T about 1 month ago
trainer_state.json

39 kB
upload model weights after 2400 training steps with frozen pretrained weights, this performs worse than checkpoint 1000 on test set T_T about 1 month ago
training_args.bin
Detected Pickle imports (10)
- "accelerate.state.PartialState",
- "transformers.training_args.TrainingArguments",
- "transformers.trainer_utils.SaveStrategy",
- "transformers.trainer_utils.SchedulerType",
- "accelerate.utils.dataclasses.DistributedType",
- "transformers.trainer_utils.IntervalStrategy",
- "transformers.training_args.OptimizerNames",
- "torch.device",
- "transformers.trainer_utils.HubStrategy",
- "transformers.trainer_pt_utils.AcceleratorConfig"
How to fix it?
5.84 kB
xet

upload model weights after 2400 training steps with frozen pretrained weights, this performs worse than checkpoint 1000 on test set T_T about 1 month ago
vocab.json

798 kB
upload model weights after 2400 training steps with frozen pretrained weights, this performs worse than checkpoint 1000 on test set T_T about 1 month ago

Detected Pickle imports (3)

Detected Pickle imports (7)

Pickle imports

Pickle imports

Detected Pickle imports (10)