Upload model
cf33876 verified - checkpoint-2250 SFT checkpoint-2250 (loss 0.906658935546875)
- checkpoint-3750 SFT checkpoint-3750 (loss 0.6874031543731689)
- checkpoint-4610 SFT checkpoint-4610 (loss 0.8109576225280761)
- checkpoint-6950 SFT checkpoint-6950 (loss 0.7705310821533203)
- 3.38 kB SFT checkpoint-3750 (loss 0.6874031543731689)
- 2.77 kB Upload model
- 999 Bytes Upload model
- 3.94 GB Upload model
- 167 Bytes Upload tokenizer
- 266 Bytes Add generation config (stop at <|im_end|>) + model card
- 287 MB SFT checkpoint-1000 (loss 0.9141511917114258)
rng_state.pth Detected Pickle imports (7)
- "collections.OrderedDict",
- "numpy.ndarray",
- "_codecs.encode",
- "numpy._core.multiarray._reconstruct",
- "numpy.dtype",
- "torch.ByteStorage",
- "torch._utils._rebuild_tensor_v2"
How to fix it?
14.6 kB SFT checkpoint-1000 (loss 0.9141511917114258) - 1.38 kB SFT checkpoint-1000 (loss 0.9141511917114258)
- 1.47 kB SFT checkpoint-1000 (loss 0.9141511917114258)
- 36.2 MB Upload tokenizer
- 489 Bytes Upload tokenizer
- 19 kB SFT checkpoint-1000 (loss 0.9141511917114258)
training_args.bin Detected Pickle imports (10)
- "transformers.training_args.TrainingArguments",
- "torch.device",
- "transformers.trainer_pt_utils.AcceleratorConfig",
- "accelerate.state.PartialState",
- "transformers.trainer_utils.HubStrategy",
- "accelerate.utils.dataclasses.DistributedType",
- "transformers.trainer_utils.SchedulerType",
- "transformers.trainer_utils.SaveStrategy",
- "transformers.training_args.OptimizerNames",
- "transformers.trainer_utils.IntervalStrategy"
How to fix it?
5.2 kB SFT checkpoint-500 (loss 0.9512178421020507)
