Table with columns: Tensors, Format| Tensors | Format |
|---|
| attention q, k, v, o (16 layers) | bfloat8_b, 8.5 bits a weight |
| Gated DeltaNet q, k, v, z, a, b, output (48 layers) | bfloat8_b, 8.5 bits a weight |
| MLP gate, up, down (64 layers) | bfloat8_b, 8.5 bits a weight |
| LM head, embedding, norms and every other tensor | bf16, as in the base model |
mtp, visual | bf16, unchanged |
9.21 bits a weight on average, 30.96 GB packed. quant4.json lists each projection's format. Quantized with
GPTQ on the bfloat8_b grid, keeping each block's shared exponent as the packer sets it, calibrated on 65,536
tokens of WikiText-2 and Korean Wikipedia.
Only the language model was quantized and evaluated.
Quickstart
from transformers import AutoModelForCausalLM, AutoTokenizer repo = "Xylolabs/Qwen3.6-27B-Tenstorrent-Tensix-BFP8"tok = AutoTokenizer.from_pretrained(repo)model = AutoModelForCausalLM.from_pretrained(repo, dtype="bfloat16", device_map="auto") messages = [{"role": "user", "content": "What is the capital of France?"}]inputs = tok.apply_chat_template(messages, add_generation_prompt=True, return_tensors="pt", return_dict=True).to(model.device)out = model.generate(**inputs, max_new_tokens=64)print(tok.decode(out[0, inputs["input_ids"].shape[1]:], skip_special_tokens=True))
Run on Tensix
Checked with ttnn 0.81.1 on ttsim, Tenstorrent's Blackhole simulator. Pack one matrix and
check that no value changed (TT-NN reads weights as [in, out]):
import jsonimport torchimport ttnnfrom safetensors import safe_open d = "Qwen3.6-27B-Tenstorrent-Tensix-BFP8" # hf download Xylolabs/Qwen3.6-27B-Tenstorrent-Tensix-BFP8 --local-dir Qwen3.6-27B-Tenstorrent-Tensix-BFP8name = "model.language_model.layers.0.mlp.down_proj.weight"fmt = json.load(open(f"{d}/quant4.json"))["layers"]["0"]["mlp.down_proj"]["format"]dtype = {"bfp4": ttnn.bfloat4_b, "bfp8": ttnn.bfloat8_b}[fmt]shard = json.load(open(f"{d}/model.safetensors.index.json"))["weight_map"][name]with safe_open(f"{d}/{shard}", "pt") as f: w = f.get_tensor(name).t().contiguous() device = ttnn.open_device(device_id=0)w_tt = ttnn.as_tensor(w, dtype=dtype, layout=ttnn.TILE_LAYOUT, device=device, memory_config=ttnn.DRAM_MEMORY_CONFIG)assert torch.equal(ttnn.to_torch(w_tt).float(), w.float())ttnn.close_device(device)
python pack_tensix.py Qwen3.6-27B-Tenstorrent-Tensix-BFP8 checks every matrix this way (--device also round-trips it
through device DRAM). python verify_grid.py Qwen3.6-27B-Tenstorrent-Tensix-BFP8 checks the grid with torch alone.
tt-metal demos (v0.81.1). tt-transformers does not support this architecture.
- tt-metal's qwen38_27b_qb2 demo was written for Qwen3.8-27B, which has this model's shape and tensor names. It runs tensor-parallel on four Blackhole chips (QuietBox 2) and refuses other hardware. Its shipped
config/precision.json and its baseline pack the projections in bfloat4_b, which re-quantizes these weights. With a bfloat8_b policy (below), its decoder packed every projection of layers 0 and 3 of this checkpoint exactly. The head goes to bfloat8_b too, so the bf16 head is not exact there. A full run was not tested.
- tt-metal's Qwen3.5/3.6 demo lists Qwen3.6-27B (P150x4 or x8). It always packs the MLP gate and up in bfloat4_b, so it does not run this checkpoint at its precision.
# build tt-metal v0.81.1 as the qwen38_27b_qb2 README describes, then, from the tt-metal checkout:python - <<'PY'import jsonfrom models.demos.qwen38_27b_qb2.tt.precision import BASELINEp = dict(BASELINE, config_id="bfp8_hifi2", weight_groups={k: "bfloat8_b" for k in BASELINE["weight_groups"]}, compute_fidelities={k: "HiFi2" for k in BASELINE["compute_fidelities"]})json.dump(p, open("precision_bfp8.json", "w"), indent=2)PYexport QWEN_PRECISION_CONFIG=$PWD/precision_bfp8.jsonexport MODEL_WEIGHTS_DIR=/path/to/Qwen3.6-27B-Tenstorrent-Tensix-BFP8 # hf download Xylolabs/Qwen3.6-27B-Tenstorrent-Tensix-BFP8 --local-dir ...python models/demos/qwen38_27b_qb2/demo/text_demo.py --full --length 128 --generate 128 --output qwen36.json
tt-forge (tt-xla). Not run with this checkpoint yet.
Evaluation
Table with columns: Bits a weight, Size (GB), Mean KL to BF16, Top-1 (%), Answers right, Lost | Bits a weight | Size (GB) | Mean KL to BF16 | Top-1 (%) | Answers right | Lost |
|---|
| BF16 (Qwen/Qwen3.6-27B) | 16 | 53.79 | 0 | 100 | 40 of 40 | - |
| this model | 9.21 | 30.963 | 0.00081 | 98.93 | 40 of 40 | 0 |
| Q4_K_M (llama.cpp) | 4.92 | 16.811 | 0.02582 | 93.23 | 40 of 40 | 0 |
KL and top-1 cover 16,240 positions of English and Korean text (WikiText-2 test, Korean Wikipedia), plain and
in the chat template. Answers are 33 short prompts with automatic checks, plus the 7 reasoning ones again with
thinking on (40 in all). "Lost" means right in BF16 but wrong or broken here; near ties are not counted.
Q4_K_M was made from the same checkpoint with llama.cpp v0.9.0 (cd959bae) and no importance matrix; its
sha256 is 20b1de06d83c9bfc1d68534d01172a86bd83a6356bf45d77afd0b434bd4e3528. Our bar: mean KL at most 1.5
times Q4_K_M's and at most one more lost answer. Passed: 0.03 times Q4_K_M's KL, 0 lost against 0.
These numbers measure closeness to BF16, not task quality. Speed is not reported here.
License
Apache 2.0, as the base model (LICENSE). The changes are listed in NOTICE. pack_tensix.py and
verify_grid.py are by Xylolabs under the same license. Provided as is, without warranty. Not affiliated with
or endorsed by Tenstorrent.
Citation
@misc{qwen3.6-27b, title = {{Qwen3.6-27B}: Flagship-Level Coding in a {27B} Dense Model}, author = {{Qwen Team}}, month = {April}, year = {2026}, url = {https://qwen.ai/blog?id=qwen3.6-27b}}