[!WARNING]

EXPERIMENTAL PRE-RELEASE NOTICE: ENGLISH-ONLY CODING SPECIALIST

Independently computed using dedicated cloud compute clusters, funded 100% out-of-pocket.

If you consider these releases to be of high quality and they help you in your day-to-day workflow, consider supporting on Ko-fi. Every contribution directly sustains future cloud compute hours and helps support my university studies.

This model suite is quantized from Jab1718/qwen3.8-flash-coder-85gb-bf16, which is an intermediate experimental slice created using moe-slice (352 out of 512 routed experts were permanently pruned exclusively against English Python and SWE-bench calibration datasets).

• English Coding Only: This model is strictly designed for programming, code completion, refactoring, and agentic tool-calling in English.
• Severe Multilingual & General Degradation: Because conversational and multilingual experts were pruned and the upstream author has not yet released the recovery fine-tuning pass, this model severely degrades and outputs broken text in languages other than English (e.g., Spanish, French, German, etc.) or in general chit-chat.
• Incompatible with Strata Engine: This model uses a 160-expert layout with decoupled n-gram tables; it is not compatible with Strata Engine (which requires the 512-expert monolith and 51B PLE tables). Run using stock llama.cpp (llama-server) or LM Studio.

Qwen3.8-Flash-Coder-85GB Lossless BF16 GGUF Reference (IsValorum)

The Official Uncompressed BF16 Reference GGUF · Golden Baseline for llama.cpp

[!NOTE]

EXPLORE THE COMPLETE QWEN3.8 FLASH CODER LINEUP

These are complementary APEX-I releases, not alternate downloads of the same model:

• Qwen3.8-Flash-Coder APEX-I-MiniPlus-V2.1 - specialist agentic coding MoE tuned for Q5-Q6 quality (21.77 GB / 3.45 BPW).
• Qwen3.8-Flash-Coder APEX-I-NanoPlus - ultra-compact footprint achieving solid Q4 quality (18.34 GB / 2.90 BPW).
• Qwen3.8-Flash-Coder-85GB Lossless BF16 GGUF - uncompressed reference baseline (85.30 GB / 16.00 BPW).

This repository provides the official uncompressed BF16 GGUF reference format converted directly from Jab1718/qwen3.8-flash-coder-85gb-bf16.

This model serves as the lossless reference baseline for benchmarking, local testing, and high-fidelity inference with llama.cpp without any intermediate requantization noise.

Model Family & Verification Matrix

[!NOTE] Independent Hardware & Compute Transparency: Generating multi-million token calibration matrices and running certified perplexity evaluations on local architectures requires dedicated high-end cloud nodes funded 100% out-of-pocket. If you rely on these MiniPlus/NanoPlus builds for your local workloads, contributions to the IsValorum Compute Fund on Ko-fi directly sponsor the GPU hours required for future Day-0 handcrafted quantizations.

Architecture Details

• Base Architecture: Qwen4ExpForCausalLM (48 hybrid layers: 36 linear attention SSM + 12 sparse attention, 4-way hyper-connections)
• Active Parameters: approx. 3.7B active per token (10 active routed MoE experts out of 160 per layer + 1 shared expert + dense backbone; 4.9B with vocabulary embeddings).
• Context Window: Native 262,144 tokens (256K).
• PLE / N-gram Table: Bypassed (ple_layer_ids: []) for 100% GPU VRAM execution with zero host RAM offload.

Usage with llama.cpp

./llama-server \
  -m ./Qwen3.8-Flash-Coder-85GB-BF16.gguf \
  -c 65536 \
  -ngl 999 \
  --host 0.0.0.0 \
  --port 8080

---

7. Community Compute Fund & Student Support

All IsValorum quantizations will always remain completely free and open to the public without paywalls.

However, cloud GPU compute is expensive and funded entirely out-of-pocket. If you find these handcrafted builds to be of high quality and they benefit your day-to-day workflow, you can support future Day-0 releases on Ko-fi. Every contribution directly funds the cloud compute hours required for new model runs and helps support my university education:

(When supporting on Ko-fi, feel free to leave a note with your Hugging Face handle and the specific model you would like prioritized).