sravanthib
/

custom_accelerate-deepseek-coder-1-5b-20k-wd-warmup-same-as-nemo

Model card Files Files and versions

custom_accelerate-deepseek-coder-1-5b-20k-wd-warmup-same-as-nemo

32.3 MB

1 contributor

History: 7 commits

sravanthib's picture

Training in progress, step 600

a944562 verified about 2 months ago

.gitattributes

1.52 kB

initial commit about 2 months ago
adapter_config.json

922 Bytes

Training in progress, step 100 about 2 months ago
adapter_model.safetensors

30 MB
xet

Training in progress, step 600 about 2 months ago
chat_template.jinja

194 Bytes

Training in progress, step 100 about 2 months ago
special_tokens_map.json

462 Bytes

Training in progress, step 100 about 2 months ago
tokenizer.json

2.29 MB

Training in progress, step 100 about 2 months ago
tokenizer_config.json

4.32 kB

Training in progress, step 100 about 2 months ago
training_args.bin
Detected Pickle imports (10)
- "transformers.training_args.OptimizerNames",
- "torch.device",
- "accelerate.utils.dataclasses.DistributedType",
- "transformers.trainer_pt_utils.AcceleratorConfig",
- "transformers.trainer_utils.HubStrategy",
- "accelerate.state.PartialState",
- "transformers.trainer_utils.SchedulerType",
- "transformers.trainer_utils.IntervalStrategy",
- "__main__.TrainingArguments",
- "transformers.trainer_utils.SaveStrategy"
How to fix it?
6.01 kB
xet

Training in progress, step 100 about 2 months ago