Create model card

README.md

@@ -25,15 +25,15 @@ tags:
 
 <!-- Provide a longer summary of what this model is. -->
 
-# Updated Evaluation Results (February 13, 2026)
+# Model Card for Qwen3-Coder-Next-8bit-g128
 
-Comprehensive evaluation results from thorough testing using mlx_lm.evaluate with mmlu_pro (200 questions per domain, num_shots=1, temp=1.0, top_p=0.95, top_k=40, seed=123):
+Quantized Qwen/Qwen3-Coder-Next using mlx-lm to 8-bit with group_size 128 for main weights and fine-grained group_size 64 for MoE weights, with the aim of maximum accuracy for 8-bit quantization.
 
-## Recent g64 vs g128 Comparison (8-bit)
+## Updated Evaluation Results (February 13, 2026)
 
-Based on the most recent comprehensive evaluation, here is the direct comparison between the 8-bit models:
+Comprehensive evaluation results from thorough testing using mlx_lm.evaluate with mmlu_pro (200 questions per domain, num_shots=1, temp=1.0, top_p=0.95, top_k=40, seed=123):
 
-### Direct Comparison Summary
+### Direct Comparison Summary (8-bit g64 vs g128)
 
 | Domain | 8-bit g64 | 8-bit g128 (this model) | Difference |
 |--------|-----------|-------------------------|------------|