DUAL-GPO
/

phi-2-gpo-test-longest-iter-0

alignment-handbook

Generated from Trainer

Model card Files Files and versions Community

phi-2-gpo-test-longest-iter-0

1 contributor

History: 14 commits

BraylonDash's picture

Training in progress, step 100

de0b05c verified 8 months ago

.gitattributes

1.52 kB

initial commit 8 months ago
README.md

1.4 kB

End of training 8 months ago
adapter_config.json

595 Bytes

Training in progress, step 100 8 months ago
adapter_model.safetensors

42 MB
LFS

Training in progress, step 100 8 months ago
added_tokens.json

1.08 kB

Model save 8 months ago
all_results.json

195 Bytes

Model save 8 months ago
config.json

898 Bytes

End of training 8 months ago
merges.txt

456 kB

Model save 8 months ago
special_tokens_map.json

587 Bytes

Model save 8 months ago
tokenizer.json

2.11 MB

Model save 8 months ago
tokenizer_config.json

7.82 kB

Model save 8 months ago
train_results.json

195 Bytes

Model save 8 months ago
trainer_state.json

1.09 kB

Model save 8 months ago
training_args.bin
Detected Pickle imports (11)
- "torch.device",
- "accelerate.utils.dataclasses.DeepSpeedPlugin",
- "alignment.configs.DPOConfig",
- "transformers.trainer_utils.HubStrategy",
- "accelerate.utils.dataclasses.DistributedType",
- "transformers.trainer_utils.IntervalStrategy",
- "transformers.trainer_utils.SchedulerType",
- "accelerate.state.PartialState",
- "transformers.integrations.deepspeed.HfTrainerDeepSpeedConfig",
- "transformers.training_args.OptimizerNames",
- "torch.bfloat16"
How to fix it?
5.88 kB
LFS

Training in progress, step 100 8 months ago
vocab.json

798 kB

Model save 8 months ago