sqlcoder-7b-2
The model weights were updated at 7 AM UTC on Feb 7, 2024. The new model weights lead to a much more performant model – particularly for joins.
Params
6.7 B
Context
16,384
Downloads 30d
214 K
Likes
450
Download history
tracking started — chart appears after 7 days of snapshots (4 recorded)214 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 |
|---|---|---|---|---|
| sqlcoder-7b-q5_k_m.gguf | Q5_K_M | 4.8 GB | 6.8 GB | ✅ Runs comfortably |
| model.safetensors | f16 | 13.5 GB | 16.3 GB | ✅ Runs comfortably |
| model.safetensors (bf16, full) | bf16 + 16K ctx | 13.5 GB | 17.3 GB | ✅ Runs comfortably |
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$ ollama run sqlcoder-7b-2 # pin the quantization explicitly $ ollama run sqlcoder-7b-2-q5_k_m
$ huggingface-cli download defog/sqlcoder-7b-2-GGUF \
sqlcoder-7b-q5_k_m.gguf --local-dir .
$ llama-cli -m sqlcoder-7b-q5_k_m.gguf \
-c 8192 -ngl 99 -t 8 --color
$ curl -s https://aimodelscomparison.com/api/v1/models/sqlcoder-7b-2
{
"hf_id": "defog/sqlcoder-7b-2",
"params_b": 6.74,
"context_length": 16384,
"license": { "id": "cc-by-sa-4.0", "commercial": "unknown" },
"downloads_30d": 214124,
"vram_estimates": [
{ "quant": "Q5_K_M", "gb": 6.8 },
{ "quant": "f16", "gb": 16.3 }
],
"updated_at": "2026-10-22T01:00:09Z"
}
Specifications
- Architecture
- LlamaForCausalLM
- Parameters
- 6.7 B
- Tensor type
- F16
- Context length
- 16,384
- Vocabulary
- 32,016
- Layers / heads
- 32 / 32
- Licence
- cc-by-sa-4.0
- First seen on the Hub
- 2024-02-05
- Training datasets
- undisclosed
- Added to our catalog
- 2026-10-22
Compare with any text-generation model
Popular comparisons
Appears in