Downloads · 30 days
0
Dingo-chat-560M
jayasuryajsk/Dingo-chat-560M
Dingo-chat-560M is a machine learning model from jayasuryajsk. Use it for the machine learning task on the model card, and read the license before you ship it in a product.
Repo: jayasuryajsk/Dingo Architecture: 20 layers, 10 heads, 10 KV heads, dmodel=1280, seqlen=2048, vocab=65536
Access
Public
Updated Oct 16, 2025
Repo size
2.1 GB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.pt2.1 GB · 100%
At a glance
- Access
- Public
- Created
- Oct 16, 2025
- Updated
- Oct 16, 2025
- SHA
- 9f5367ba
Model card
From the Hugging Face model README
Dingo d20 (All Intermediate Checkpoints)
Repo: jayasuryajsk/Dingo
Architecture: 20 layers, 10 heads, 10 KV heads, d_model=1280, seq_len=2048, vocab=65536
Each checkpoint is stored in: checkpoints/<step>/{model_<step>.pt, meta_<step>.json}
Example eval (step 000650):
- MMLU: 32.62 %
- ARC-Easy: 44.82 %
- ARC-Challenge: 31.14 %
- GSM8K: 5.08 %
- HumanEval: 6.71 %
Load example (custom Dingo): import torch, json step="000650" base="checkpoints" ckpt = torch.load(f"{base}/{step}/model_{step}.pt", map_location="cpu") with open(f"{base}/{step}/meta_{step}.json") as f: meta = json.load(f)
- Created
- Oct 16, 2025
- Updated
- Oct 16, 2025