DeepSeek-V4-Flash-0731 EXL3 Quantization Suite
Collection
Seven calibrated FP4-to-EXL3 routed-expert transcodes of DeepSeek V4 Flash 0731. • 8 items • Updated
DeepSeek V4 Flash 0731 EXL3 suite
Campaign status: pending. These are public release destinations and model cards; weights and measured validation results have not been uploaded.
The campaign will create seven FP4-to-EXL3 routed-expert transcodes from one pinned 1,048,576-token capture, then upload them sequentially. Each completed card will expose the source/capture revisions, route coverage, exact fractional allocation, checksums, runtime recipe, and measured validation status.
| Target | Repository |
|---|---|
| 2.0 bpw | 0xSero/DeepSeek-V4-Flash-0731-EXL3-2.0bpw |
| 2.5 bpw | 0xSero/DeepSeek-V4-Flash-0731-EXL3-2.5bpw |
| 3.0 bpw | 0xSero/DeepSeek-V4-Flash-0731-EXL3-3.0bpw |
| 3.5 bpw | 0xSero/DeepSeek-V4-Flash-0731-EXL3-3.5bpw |
| 4.0 bpw | 0xSero/DeepSeek-V4-Flash-0731-EXL3-4.0bpw |
| 4.5 bpw | 0xSero/DeepSeek-V4-Flash-0731-EXL3-4.5bpw |
| 5.0 bpw | 0xSero/DeepSeek-V4-Flash-0731-EXL3-5.0bpw |
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support
Model tree for 0xSero/DeepSeek-V4-Flash-0731-EXL3
Base model
deepseek-ai/DeepSeek-V4-Flash-0731