Table with columns: Tensors, Format| Tensors | Format |
|---|
| attention q, k, v, o and MLP gate, up, down (36 layers) | bfloat8_b, 8.5 bits a weight |
| LM head (tied to the embedding), embedding, norms and every other tensor | bf16, as in the base model |
9.23 bits a weight on average, 4.64 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.
Quickstart
from transformers import AutoModelForCausalLM, AutoTokenizer repo = "Xylolabs/Qwen3-4B-Base-Tenstorrent-Tensix-BFP8"tok = AutoTokenizer.from_pretrained(repo)model = AutoModelForCausalLM.from_pretrained(repo, dtype="bfloat16", device_map="auto") inputs = tok("The capital of France is", return_tensors="pt").to(model.device)out = model.generate(**inputs, max_new_tokens=16)print(tok.decode(out[0, inputs["input_ids"].shape[1]:], skip_special_tokens=True))
This is the pretrained model: it continues text. Its tokenizer keeps the base model's chat template, but the
model is not instruction-tuned.
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 = "Qwen3-4B-Base-Tenstorrent-Tensix-BFP8" # hf download Xylolabs/Qwen3-4B-Base-Tenstorrent-Tensix-BFP8 --local-dir Qwen3-4B-Base-Tenstorrent-Tensix-BFP8name = "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 Qwen3-4B-Base-Tenstorrent-Tensix-BFP8 checks every matrix this way (--device also round-trips it
through device DRAM). python verify_grid.py Qwen3-4B-Base-Tenstorrent-Tensix-BFP8 checks the grid with torch alone.
tt-transformers and tt-forge were not run with this checkpoint.
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/Qwen3-4B-Base) | 16 | 8.04 | 0 | 100 | 28 of 28 | - |
| this model | 9.23 | 4.639 | 0.00126 | 98.34 | 28 of 28 | 0 |
| Q4_K_M (llama.cpp) | 4.96 | 2.497 | 0.04793 | 89.65 | 28 of 28 | 0 |
KL and top-1 cover 8,184 positions of English and Korean text (WikiText-2 test, Korean Wikipedia), as plain
text. Answers are 28 short prompts with automatic checks, in completion form, 3-shot. "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
a544320bfa522b0b40071193937df23332948cfacc9405089549c12e1c9d0353. 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{qwen3technicalreport, title={Qwen3 Technical Report}, author={Qwen Team}, year={2025}, eprint={2505.09388}, archivePrefix={arXiv}, primaryClass={cs.CL}, url={https://arxiv.org/abs/2505.09388},}