Downloads · 30 days
3
11% of all-time downloads
Olmo-3-7B-PostTraining8K-SFT-step-100
dureduck/Olmo-3-7B-PostTraining8K-SFT-step-100
Olmo-3-7B-PostTraining8K-SFT-step-100 is a text generation model from dureduck. Use it when you need the model to write or continue text. It is set up for easydel.
This repository contains the retained step-100 checkpoint from the PostTraining8.8K one-epoch OLMo 3 7B SFT checkpoint-curve run.
All-time downloads
28
Public
Repo size
32.6 GB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.016.3 GB · 50%
At a glance
- Task
- Text Generation
- Library
- easydel
- Model type
- olmo3
- Access
- Public
- Created
- Sep 2, 2026
- Updated
- Sep 2, 2026
- SHA
- c49a0004
Try a prompt
Base models
Model card
From the Hugging Face model README
OLMo 3 7B PostTraining8K SFT — step 100
This repository contains the retained step-100 checkpoint from the PostTraining8.8K one-epoch OLMo 3 7B SFT checkpoint-curve run.
Checkpoint format
This is an EasyDeL full-state checkpoint, not a Transformers safetensors
export. It includes the model tree, optimizer state (tx/), checkpoint
metadata, and the original publication manifest.
- Training step: 100 of 118
- Checkpoint class: permanent
- Manifest-listed files: 4,178
- Manifest-listed bytes: 32,581,205,836
- Validation: every manifest-listed local file matched its recorded size
- Source run:
olmo3-base7b-posttraining8k-one-epoch-evalcurve-20260901t0535z-r2
The original _PUBLISHED.json marker and PUBLISH-MANIFEST.json are retained
for provenance and integrity checking.
- Task
- Text Generation
- Library
- easydel
- Type
- olmo3
- Created
- Sep 2, 2026
- Updated
- Sep 2, 2026