---

license: apache-2.0

base_model:

• Qwen/Qwen3-4B-Instruct-2507

base_model_relation: quantized

library_name: gguf

pipeline_tag: text-generation

tags:

• wekala
• gguf

---

qwen3-4b-instruct-2507-gguf

Historical model archive, superseded in the current Wekala catalogue. Retained for provenance and existing consumers; not offered for new playground selection.

Origin and licence

| Field | Value |

| --- | --- |

| Original developer | Qwen Team, Alibaba Cloud |

| Original model | Qwen/Qwen3-4B-Instruct-2507 |

| Pinned source revision | cdbee75f17c01a7cc42f958dc650907174af0554 |

| Original publication | 2025-08-06 (day) |

| Date evidence | Publisher source |

| Licence | apache-2.0; preserved in LICENSE |

| Full source parameters | 4,022,468,096 (includes embeddings and source modalities) |

| Nile Academy role | Official artifact mirror or explicitly recorded format conversion; no additional training |

Original publisher cards and available notices are preserved in provenance/. The original model licence governs the weights; this repository grants no rights to the Wekala SDK.

Files and quantization

Retired from new Wekala selection. Historical Q4_K_M conversion and original evidence are preserved; no broad quality claim.

| File | Quantization | Size (GB) | Origin |

| --- | --- | ---: | --- |

| Qwen3-4B-Instruct-2507-Q4_K_M.gguf | Q4_K_M | 2.497 | existing Nile conversion; original lineage retained |

Exact file SHA-256 values, upstream paths/revisions and conversion receipts are recorded in model-metadata.json.

Published benchmarks

Publisher-reported source-model scores follow. They are not Wekala measurements of these quantized files. Harness, precision, reasoning settings and task versions differ; do not treat them as a normalized cross-model ranking.

| Benchmark | Score (%) | Mode / settings |

| --- | ---: | --- |

| MMLU-Pro | 69.6 | non-thinking; Publisher original source protocol; not a Wekala quantization measurement. |

| GPQA (subset unspecified) | 62 | non-thinking; Publisher original source protocol; not a Wekala quantization measurement. |

| BFCL-v3 | 61.9 | non-thinking; Publisher original source protocol; not a Wekala quantization measurement. |

| IFEval | 83.4 | non-thinking; Publisher original source protocol; not a Wekala quantization measurement. |

Exact source card.

Qualification and use

Historical compatibility archive only. Not included in the current model download or chat selector.

Wekala broad quantization-quality qualification: not qualified. No general “no significant quality loss” claim is made. Official QAT or publisher comparisons are attributed separately. Before recommendation, use paired reference/variant evaluation across instruction, knowledge, math, multilingual/Arabic and tool tasks under the registered Wekala campaign.

Device fit is separate: weight size excludes context state, scratch buffers and driver headroom. CPU and CUDA are execution backends for the same weight file, not separate models. Mobile/browser and sustained 8GB-VRAM operation require their own measured profiles.

Models can produce incorrect or unsafe outputs. Tool calls are proposals subject to the consuming application’s permissions and controller admission.