7 contributors

History: 26 commits

pszemraj

update navy seals example

1c5d929 over 2 years ago

.gitattributes

1.17 kB

initial commit over 2 years ago
.gitignore

13 Bytes

add stuff over 2 years ago
README.md

18.6 kB

update navy seals example over 2 years ago
config.json

1.12 kB

update ckpt with 4+ epochs of training at 1e-3 over 2 years ago
latest

14 Bytes

add stuff over 2 years ago
long-t5-tglobal-base-16384-booksum-V6-partial-ft3-booksum_training_metadata.json

3.16 kB

add additional 2-epoch checkpoint, better regularization over 2 years ago
long-t5-tglobal-base-16384-booksum-V7.9-ft1-booksum_training_metadata.json

3.11 kB

add stuff over 2 years ago
long-t5-tglobal-base-16384-booksum-V9-ft1-booksum_training_metadata.json

3.1 kB

update ckpt with 4+ epochs of training at 1e-3 over 2 years ago
pytorch_model.bin

990 MB
LFS

update ckpt with 4+ epochs of training at 1e-3 over 2 years ago
special_tokens_map.json

2.2 kB

add additional 2-epoch checkpoint, better regularization over 2 years ago
spiece.model

792 kB
LFS

add additional 2-epoch checkpoint, better regularization over 2 years ago
tokenizer.json

2.42 MB

add additional 2-epoch checkpoint, better regularization over 2 years ago
tokenizer_config.json

2.36 kB

update ckpt with 4+ epochs of training at 1e-3 over 2 years ago
trainer_state.json

20.3 kB

update ckpt with 4+ epochs of training at 1e-3 over 2 years ago
training_args.bin
Detected Pickle imports (8)
- "transformers.deepspeed.HfTrainerDeepSpeedConfig",
- "transformers.trainer_utils.HubStrategy",
- "transformers.training_args.OptimizerNames",
- "transformers.trainer_utils.IntervalStrategy",
- "transformers.trainer_utils.SchedulerType",
- "torch.device",
- "transformers.training_args_seq2seq.Seq2SeqTrainingArguments",
- "torch.float16"
How to fix it?
4.53 kB
LFS

update ckpt with 4+ epochs of training at 1e-3 over 2 years ago
zero_to_fp32.py

18.9 kB

add stuff over 2 years ago

Detected Pickle imports (8)