Inkling-Small-DSpark-Preview
A DSpark speculator for the Inkling-Small NVFP4 target, enabling faster inference through speculative decoding. DSpark extends the DFlash parallel-draft backbone with a Markov logit-bias head and a per-position confidence head. This checkpoint was trained with SpecForge using hidden states from a live SGLang target engine.
Params
1.4 B
Context
1,048,576
Downloads 30d
151 K
Likes
2
Download history
tracking started — chart appears after 7 days of snapshots (6 recorded)151 K downloads in the last 30 days
Can you run it?
Estimated VRAM at 8K context unless noted. Pick your hardware to see the verdict per quantization.
| File | Quant | Size | Est. VRAM | Verdict on RTX 4090 · 24 GB |
|---|---|---|---|---|
| model.safetensors | bf16 | 1.8 GB | 2.7 GB | ✅ Runs comfortably |
| model.safetensors (bf16, full) | bf16 + 1024K ctx | 1.8 GB | 29.3 GB | ❌ Won’t fit |
Estimate: file size × 1.1 + KV cache at 8K + 0.5 GB overhead. Not a benchmark — how we calculate this.
Run it
copy-paste, exact tags checked against the Hub$ curl -s https://aimodelscomparison.com/api/v1/models/inkling-small-dspark-preview
{
"hf_id": "RadixArk/Inkling-Small-DSpark-Preview",
"params_b": 1.39,
"context_length": 1048576,
"license": { "id": "", "commercial": "unknown" },
"downloads_30d": 151192,
"vram_estimates": [
{ "quant": "bf16", "gb": 2.7 }
],
"updated_at": "2026-10-20T01:00:34Z"
}
Specifications
- Architecture
- DSparkDraftModel
- Parameters
- 1.4 B
- Tensor type
- BF16
- Context length
- 1,048,576
- Vocabulary
- 201,024
- Layers / heads
- 5 / 64
- First seen on the Hub
- 2026-07-30
- Training datasets
- undisclosed
- Added to our catalog
- 2026-10-20
Compare with any text-generation model