File size: 3,383 Bytes
2df88cb 6119f69 2df88cb 6119f69 2df88cb 6119f69 2df88cb 6119f69 2df88cb | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 | ---
language:
- en
- zh
license: mit
library_name: mlx
pipeline_tag: text-generation
base_model: deepseek-ai/DeepSeek-V4-Flash
base_model_relation: quantized
tags:
- mlx
- safetensors
- deepseek_v4
- deepseek-v4
- deepseek
- quantized
- apple-silicon
- mixture-of-experts
- Mixture of Experts
- mxtq
- turboquant
- jang
- jangtq
- conversational
- 2.3-bit
---
# DeepSeek-V4-Flash-TQ-Q2.3-MLX
`osmapi/DeepSeek-V4-Flash-TQ-Q2.3-MLX` is an Apple-Silicon MLX TurboQuant/JANGTQ quantization of [`deepseek-ai/DeepSeek-V4-Flash`](https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash).
No fine-tuning, distillation, or retraining was applied. The official mixed FP4/FP8 source weights were converted locally, the MTP head was dropped because it is not used for normal decode, and router/mHC/control tensors were preserved rather than aggressively quantized.
## Model Details
| Property | Value |
|---|---|
| Base model | `deepseek-ai/DeepSeek-V4-Flash` |
| Architecture | DeepSeek-V4 Flash MoE, 284B total / 13B active, 1M context |
| Local profile | `JANGTQ-Q2.3` |
| Bundle size | 88.03 GB |
| Layout | Pre-stacked MLX `switch_mlp` layout |
| MTP head | Dropped |
| Validation | Safetensors header/index validation, metadata validation |
## Required Sidecar
This is a JANGTQ/TurboQuant bundle and requires `jangtq_runtime.safetensors` from this repository. The sidecar stores the deterministic codebooks and Hadamard rotation signs used to decode the `.tq_packed` expert weights. If it is missing, re-download the full repository or fetch that file explicitly:
```bash
hf download osmapi/DeepSeek-V4-Flash-TQ-Q2.3-MLX jangtq_runtime.safetensors --local-dir <your-model-dir>
```
## Quantization Recipe
| Tensor class | Codec | Bits / handling |
|---|---:|---|
| Routed experts | TurboQuant MXTQ | 110 routed layer/projection groups at 2-bit MXTQ and 19 at 4-bit MXTQ |
| Routed effective bits | MXTQ | 2.2946 bits |
| Attention, shared experts, compressor, indexer, embed, lm head | MLX affine | 8-bit, group size 32 |
| Norms, router, mHC, sinks, integer routing tables | passthrough | source precision preserved |
The fractional target is implemented as a power-of-two lane mix because the current JANGTQ vectorized packer is stable on 2/4/8-bit lanes for DeepSeek-V4 expert dimensions.
## Use
Install the JANG loader/runtime and MLX LM:
```bash
pip install mlx-lm jang-tools
```
Example:
```python
from jang_tools.load_jangtq import load_jangtq_model
from mlx_lm import generate
model, tokenizer = load_jangtq_model("osmapi/DeepSeek-V4-Flash-TQ-Q2.3-MLX")
prompt = "Write a short note about MLX quantization."
text = generate(model, tokenizer, prompt=prompt, verbose=True)
print(text)
```
## Files
- `model-*.safetensors`: pre-stacked JANGTQ/MLX shards
- `model.safetensors.index.json`: shard index
- `jangtq_runtime.safetensors`: required TurboQuant runtime sidecar
- `config.json`, `jang_config.json`: MLX/JANGTQ metadata
- `encoding/`: upstream DeepSeek-V4 prompt encoding reference
## Notes
This card follows the same broad shape as the other osmapi DeepSeek-V4-Flash MLX uploads: a sidecar warning, an explicit recipe table, and minimal reproducible loading instructions. Q2.3 is an aggressive size-first TurboQuant profile, so treat it as experimental until evaluated on your target prompts.
## License
MIT, following the upstream DeepSeek-V4-Flash release.
|