GLM-5.3-Flash-oQ2.5e-mtp
This model was quantized using oQ (oMLX v0.7.1) mixed-precision quantization with Improved oQe Quantization (jundot/omlx#4385).
A mixed-precision MLX checkpoint of zai-org/GLM-5.3-Flash, a natively multimodal mixture-of-experts model with 320B total and 18B active parameters that combines sparse and linear attention with Manifold-Constrained Hyper-Connections (mHC). It includes the language model, the vision encoder and the multi-token prediction (MTP) head.
The effective average is 3.182 bits/weight for the complete checkpoint.
Quantization and Bit Distribution
oQ2.5e allocates bits by measured layer sensitivity, then rounds each group with an importance matrix (128 samples x 512 tokens) and a weighted least-squares refit of its scale and bias.
| Storage format | Logical weights | Share of total | Tensor storage | Effective bits/weight |
|---|---|---|---|---|
| Affine 2-bit, group 64 | 164.282B | 51.127% | 47.812 GiB | 2.50 |
| Affine 3-bit, group 64 | 140.123B | 43.608% | 57.094 GiB | 3.50 |
| Affine 4-bit, group 64 | 7.248B | 2.256% | 3.797 GiB | 4.50 |
| Affine 8-bit, group 64 | 8.976B | 2.793% | 8.882 GiB | 8.50 |
| BF16 | 0.607B | 0.189% | 1.131 GiB | 16.00 |
| FP32 | 0.086B | 0.027% | 0.322 GiB | 32.00 |
| Total | 321.323B | 100% | 119.038 GiB | 3.182 |
Storage Breakdown
| Component | Logical weights | Storage |
|---|---|---|
| Language backbone | 313.327B | 122.380 GB / 113.975 GiB |
| Vision encoder | 0.564B | 1.127 GB / 1.050 GiB |
| MTP head | 7.433B | 4.309 GB / 4.013 GiB |
| Total | 321.323B | 127.816 GB / 119.038 GiB |
License
MIT, following the source model. See LICENSE.