chopratejas
/

kompress-base

Token Classification

token-compression

context-optimization

Eval Results (legacy)

Model card Files Files and versions

604 MB

1 contributor

History: 7 commits

chopratejas's picture

Update model card: v9 evals, architecture details, comparison with LLMLingua-2

490f973 verified about 9 hours ago

.gitattributes

1.52 kB

initial commit 5 days ago
README.md

6.18 kB

Update model card: v9 evals, architecture details, comparison with LLMLingua-2 about 9 hours ago
config.json

2.78 kB

Initial release: Kompress v1 — ModernBERT token compressor for agentic contexts 5 days ago
model.safetensors

600 MB
xet

v9: 6.9/10 quality (beats LLMLingua-2 6.2), 2.3x faster, 215K extractive labels from 8 diverse datasets about 9 hours ago
tokenizer.json

3.58 MB

Initial release: Kompress v1 — ModernBERT token compressor for agentic contexts 5 days ago
tokenizer_config.json

351 Bytes

Initial release: Kompress v1 — ModernBERT token compressor for agentic contexts 5 days ago
training_args.bin
Detected Pickle imports (10)
- "accelerate.state.PartialState",
- "torch.device",
- "transformers.trainer_pt_utils.AcceleratorConfig",
- "transformers.trainer_utils.SchedulerType",
- "transformers.training_args.TrainingArguments",
- "transformers.trainer_utils.HubStrategy",
- "accelerate.utils.dataclasses.DistributedType",
- "transformers.trainer_utils.IntervalStrategy",
- "transformers.trainer_utils.SaveStrategy",
- "transformers.training_args.OptimizerNames"
How to fix it?
5.2 kB
xet

v9: 6.9/10 quality (beats LLMLingua-2 6.2), 2.3x faster, 215K extractive labels from 8 diverse datasets about 9 hours ago