Open-weight model
exaone-verireason-sft_accuracy_medium_3to5_ratio0.12_seed1234
by Jongbin Won Jongbin-kr/exaone-verireason-sft_accuracy_medium_3to5_ratio0.12_seed1234
exaone-verireason-sft_accuracy_medium_3to5_ratio0.12_seed1234 is an open-weight model from Jongbin Won. Its published files total 335.6 MB.
EXAONE-3.5-7.8B-Instruct LoRA SFT on an answer-only ConvFinQA subset. Expected local base-model cache revision: 553ea250b9a5317231459279d5847d6cf955b9aa (training loader used the Hub ID without an explicit revision pin).
Model Card
EXAONE-3.5-7.8B-Instruct LoRA SFT on an answer-only ConvFinQA subset. Expected local base-model cache revision: 553ea250b9a5317231459279d5847d6cf955b9aa (training loader used the Hub ID without an explicit revision pin). Selection manifest SHA256: 3e42a1366dde05eb5fff7b7e563d12b0995714bb6d3f4f08f413484f768bb219. main is the final saved validation-best adapter; epoch branches are: - epoch1-step82: checkpoint-82 - epoch2-step164: checkpoint-164 - epoch3-step246: checkpoint-246 These are PEFT adapter weights, not merged full-model weights.
Excerpt from the card by Jongbin Won.
Identity and Version
- Repository
- Jongbin-kr/exaone-verireason-sft_accuracy_medium_3to5_ratio0.12_seed1234
- Publisher
- Jongbin Won
- Task
- Not stated by the source
- Modality
- Other
- Library
- peft
- Parameters
- Not stated by the source
- Languages
- Not stated by the source
- Revision
- 3100047355624b57f4190ee95d78c80fb63ba177
- First published
- 2026-10-06
- Last updated
- 2026-10-06
Files and Weights
4 files, 335.6 MB in total. The weights are 1 file totalling 335.6 MB in safetensors.
Every file
| File | Type | Size | SHA-256 |
|---|---|---|---|
| adapter_model.safetensors | Weights | 335.6 MB | d19cb34d8112 |
| adapter_config.json | Configuration | 1.1 KB | — |
| README.md | Documentation | 913 B | — |
| .gitattributes | Repository | 1.5 KB | — |
License and Download
- License
- Not stated by the source
- Access
- Open weights, no gate
- Download size
- 335.6 MB
Released by Jongbin Won through its official repository on Hugging Face.
Built From
- Adapter of LGAI-EXAONE/EXAONE-3.5-7.8B-Instruct
- Derived from LGAI-EXAONE/EXAONE-3.5-7.8B-Instruct
Memory Requirements
| Precision | Weights in memory |
|---|---|
| As published | 335.6 MB |
Weights only, from the published parameter count; the key-value cache and runtime add to this.