Update model card for petergilani/Qwen3-Coder-Next-3bit-g128
Browse files
README.md
CHANGED
|
@@ -14,6 +14,8 @@ tags:
|
|
| 14 |
---
|
| 15 |
# Model Card for Qwen3-Coder-Next-3bit-g128
|
| 16 |
|
|
|
|
|
|
|
| 17 |
## Evaluation Results
|
| 18 |
|
| 19 |
Testing with mlx_lm.evaluate using mmlu_pro with 50 Qs for each topic, comparing the 3-bit g128 quant with the 3-bit g64 quant:
|
|
|
|
| 14 |
---
|
| 15 |
# Model Card for Qwen3-Coder-Next-3bit-g128
|
| 16 |
|
| 17 |
+
Quantized Qwen/Qwen3-Coder-Next using mlx-lm to 3-bit with group_size 128 for main weights and fine-grained group_size 64 for MoE weights, with the aim of maximum accuracy for 3bit quantization.
|
| 18 |
+
|
| 19 |
## Evaluation Results
|
| 20 |
|
| 21 |
Testing with mlx_lm.evaluate using mmlu_pro with 50 Qs for each topic, comparing the 3-bit g128 quant with the 3-bit g64 quant:
|