WUMBOLABS / EVALUATIONS CANONICAL MODEL EVALUATION

/evaluations/qwen3.8-4b-empero

Qwen3.8-4B (Empero)

Published evaluation evidence and testing history. Results remain attributed to their original configurations and methods.

latest evidence 2026-09-01

Overview

empero-ai (Qwen3.8 derivative) · 2 tested configurations · 2 evidence events · newest evidence 2026-09-01.

Contents: Findings · Event results · Tested configurations · Chronology · Technical provenance

Findings

Each finding is attributed to the event and tested configuration that measured it; exact values, failures and limitations are in the event results. A historical event's recorded status describes that event's own sections at its date and is not a verdict on newer sections.

  • Recommended configuration: llama.cpp Q4_K_M Distill (WumboServer RTX 2060S lane) (specialized): event result.
  • Latest evidence: 2026-09-01: event result
  • 2026-09-01 · llama.cpp · RTX 2060 SUPER: classification BENCHMARK_ONLY / HARDWARE_LANE (not a WumboJetsII result). Event result
  • 2026-08-18 · llama.cpp/vLLM · RTX 5070: classification SPECIALIZED_TEST / VENDOR_ALIGNMENT_BATTERY. Event result

Event results

Newest first. Each event is one independent testing/publication event with its full report linked once at the top of its section; scores, failures and limitations stay visible and raw section-provenance JSON stays collapsed.

2026-09-01: WumboServer RTX 2060S LocalMaxxing 4B Q4_K_M

Benchmark Only: WumboServer hardware lane · config llama.cpp · RTX 2060 SUPER · maturity: BENCHMARK_ONLY · status: BENCHMARK_ONLY / HARDWARE_LANE (not a WumboJetsII result)

Configuration: llama.cpp Q4_K_M Distill (WumboServer RTX 2060S lane).

Full report (REPORT.md)

empero-ai/Qwen3.8-4B-Distill-GGUF Q4_K_M on WumboServer RTX 2060S: 97.14 tok/s out, 2145.66 tok/s prefill, 3.0 GB peak VRAM, 172 W. Hardware-lane benchmark; not a WumboJetsII WELP result.

This event is a bounded public-safe summary derived from retained WumboLabs evidence. It is not a WELP characterization, a universal model ranking, or a production-readiness proof. Values are attributed to the tested artifact, runtime, hardware, suite, and settings.

Identity and scope
  • Profile: qwen3.8-4b-empero-distill-llamacpp-q4km-wumboserver: llama.cpp Q4_K_M Distill (WumboServer RTX 2060S lane)
  • Evidence maturity: BENCHMARK_ONLY
  • Evidence scope: performance
  • Hardware: WumboServer (RTX 2060 SUPER 8GB)

2026-08-18: LLMGauge vendor-alignment battery (4B bf16/q4)

Specialized Test: vendor alignment · config llama.cpp/vLLM · RTX 5070 · maturity: SPECIALIZED_TEST · status: SPECIALIZED_TEST / VENDOR_ALIGNMENT_BATTERY

Configuration: LLMGauge vendor-alignment battery (bf16/q4).

Full report (REPORT.md)

LLMGauge vendor-alignment qualification battery on empero-ai Qwen3.8-4B (BF16 + Q4).

This event is a bounded public-safe summary derived from retained WumboLabs evidence. It is not a WELP characterization, a universal model ranking, or a production-readiness proof. Values are attributed to the tested artifact, runtime, hardware, suite, and settings.

Identity and scope
  • Profile: qwen3.8-4b-empero-vendor-battery: LLMGauge vendor-alignment battery (bf16/q4)
  • Evidence maturity: SPECIALIZED_TEST
  • Evidence scope: specialized, performance
  • Hardware: WumboJetsII (RTX 5070 12GB)

Tested configurations

Exact profile identities, sampling conditions and provenance are in each configuration's details below.

ConfigurationArtifact / precisionRuntimeHardwareContext / reasoningEvents
llama.cpp/vLLM · RTX 5070empero-ai/Qwen3.8-4B-GGUF (BF16 + Q4)llama.cpp/vLLMRTX 5070not recorded1 (2026-08-18)
llama.cpp · RTX 2060 SUPERempero-ai/Qwen3.8-4B-Distill-GGUF Q4_K_Mllama.cppRTX 2060 SUPERnot recorded1 (2026-09-01)
Configuration details: llama.cpp/vLLM · RTX 5070
  • Full profile name: LLMGauge vendor-alignment battery (bf16/q4).
  • Profile identity: qwen3.8-4b-empero-vendor-battery.
  • Status: current tested surface; not the recommended profile.
  • Artifact (as recorded): empero-ai/Qwen3.8-4B-GGUF (BF16 + Q4)
  • Runtime (as recorded): llama.cpp/vLLM.
  • Profile Metadata

Events on this configuration:

Configuration details: llama.cpp · RTX 2060 SUPER
  • Full profile name: llama.cpp Q4_K_M Distill (WumboServer RTX 2060S lane).
  • Profile identity: qwen3.8-4b-empero-distill-llamacpp-q4km-wumboserver.
  • Status: specialized tested surface.
  • Artifact (as recorded): empero-ai/Qwen3.8-4B-Distill-GGUF Q4_K_M
  • Runtime (as recorded): llama.cpp.
  • Profile Metadata

Events on this configuration:

Chronology

Compact index; full results are in Event results.

DateEventConfigurationStatus
2026-09-01WumboServer RTX 2060S LocalMaxxing 4B Q4_K_Mllama.cpp · RTX 2060 SUPERBENCHMARK_ONLY / HARDWARE_LANE (not a WumboJetsII result)
2026-08-18LLMGauge vendor-alignment battery (4B bf16/q4)llama.cpp/vLLM · RTX 5070SPECIALIZED_TEST / VENDOR_ALIGNMENT_BATTERY

Technical provenance

Publication, acceptance and source-pin provenance for each event, collapsed; the scientific content above never depends on it.

Model evidence tree (GitHub)

Legacy provenance