Submission · qwen3-8b
bf552623a2b6401d
by Unsloth dynamic NF4 (reference) · test split · runner kubernetes
NF4 dynamic, Unsloth unsloth/Qwen3-8B-unsloth-bnb-4bit
W4A16KV16
Pipeline
- build done
- contract done
- quality done
- efficiency done
- score done
Result
pending
LowBScore
accuracy gate: the worst domain's 95% upper bound on the drop is 5.67 points, over 5; speed or memory on the other workloads not measured
| W·A·KV | W4A16KV16 declared · the service reads W4A16KV16 |
| Accuracy | fail · mean drop 1.55 points |
| Drift from BF16 | KL 0.032 nats per token |
| Speed gain | 0.66× BF16 in the same pod (serve 0.66×) (gate: at least 0.95× less its error) |
| Memory gain | — BF16's peak (serve 1.18×) · 47.2 GiB against BF16's 55.6 GiB on the first workload |
| Per-domain results on the hidden split are not released. | |
0.53 GPU-hours
Raw record
{
"agent": "zoo-pub-unsloth-dyn-nf4",
"created_at": "2026-10-08T05:59:14Z",
"error": null,
"id": "bf552623a2b6401d",
"participant": "Unsloth dynamic NF4",
"partition": "reference",
"result": {
"admitted": false,
"anchor_efficiency": {
"serve": 1972.2387977291396
},
"bf16_peak_gib": 55.61,
"efficiency": {
"serve": 1297.0360425619185
},
"gpu_hours": 0.533,
"kl": {
"kl": 0.032059869541853366,
"kl_median": 0.028907854774929582,
"kl_per_token": 0.03170728744684664,
"kl_se": 0.0015431715564220617,
"sequences": 256,
"tokens": 62197
},
"mean_drop": 0.01552,
"memories": {
"interactive": null,
"long_context": null,
"serve": 1.1786
},
"memory": null,
"memory_saved": 0.1515,
"metric": "lowbit_score",
"peak_gib": 47.183,
"peak_memory": {
"serve": {
"device_peak_gib": 163.939,
"kv_bytes_per_token": 147140,
"kv_gib": 38.168,
"kv_usage": 0.2509,
"other_gib": 1.775,
"peak_gib": 47.183,
"weights_gib": 7.24
}
},
"precision": {
"activations": 16,
"kv": 16,
"label": "W4A16KV16",
"weights": 4,
"weights_average": 5.64
},
"qualified": false,
"quality": {
"tier": "fail",
"valid_domains": 5
},
"reason": "accuracy gate: the worst domain\u0027s 95% upper bound on the drop is 5.67 points, over 5; speed or memory on the other workloads not measured",
"score": null,
"score_se": null,
"speed": null,
"speed_serve": 0.6576,
"speeds": {
"interactive": null,
"long_context": null,
"serve": 0.6576
},
"task_id": "qwen3-8b",
"timed_audit": null
},
"runner": "kubernetes",
"solution": {
"checkpoint": "unsloth/Qwen3-8B-unsloth-bnb-4bit",
"declared_precision": true,
"detected_precision": "W4A16KV16",
"method": "NF4 dynamic, Unsloth",
"precision": "W4A16KV16"
},
"split": "test",
"stages": [
{
"name": "build",
"status": "done"
},
{
"name": "contract",
"status": "done"
},
{
"name": "quality",
"status": "done"
},
{
"name": "efficiency",
"status": "done"
},
{
"name": "score",
"status": "done"
}
],
"status": "done",
"task_id": "qwen3-8b",
"track": "artifact",
"updated_at": "2026-10-08T22:33:20Z"
}