Remove quant_mode

[apsonawane]

This pull request updates the quantization configuration for several text.json model files across different Qwen model sizes and hardware backends. The main change is the removal of the quant_mode option from the extra_options section, and for some 0.8B models, the addition of an explicit int4_algo_config setting.

Quantization configuration updates:

• Removed the quant_mode parameter from the extra_options in all affected text.json files for the 0.8B, 2B, 4B, 9B, and 27B models, across CPU, CUDA, and WebGPU backends. This simplifies the quantization settings and avoids redundancy.

Algorithm configuration enhancements (0.8B models):

• Added "int4_algo_config": "k_quant_linear" to the 0.8B model configs for CPU, CUDA, and WebGPU, making the quantization algorithm explicit.

Minor cleanup:

• Cleaned up trailing commas and ensured consistent formatting in the extra_options sections where options were removed.
  These changes make the quantization settings more consistent and explicit, reducing potential confusion or misconfiguration.

[Copilot]

Pull request overview

Removes the quant_mode extra option from Qwen3.5 model text.json configs across CPU, CUDA, and WebGPU backends, and makes the int4 algorithm explicit for the 0.8B variants.

Changes:

• Remove quant_mode from extra_options in all affected text.json files (0.8B, 2B, 4B, 9B, 27B).
• Add "int4_algo_config": "k_quant_linear" to all three 0.8B backend configs.
• Clean up trailing commas after option removal.

Reviewed changes

Copilot reviewed 12 out of 12 changed files in this pull request and generated no comments.

Show a summary per file

File	Description
Qwen-Qwen3.5-9B/builtin/webgpu/text.json	Drop quant_mode: int4 from extra_options.
Qwen-Qwen3.5-9B/builtin/cuda/text.json	Drop quant_mode: int4 from extra_options.
Qwen-Qwen3.5-9B/builtin/cpu_and_mobile/text.json	Drop quant_mode: int4 from extra_options.
Qwen-Qwen3.5-4B/builtin/webgpu/text.json	Drop quant_mode: int4 from extra_options.
Qwen-Qwen3.5-4B/builtin/cuda/text.json	Drop quant_mode: int4 from extra_options.
Qwen-Qwen3.5-4B/builtin/cpu_and_mobile/text.json	Drop quant_mode: int4 from extra_options.
Qwen-Qwen3.5-2B/builtin/webgpu/text.json	Drop quant_mode: int4 from extra_options.
Qwen-Qwen3.5-2B/builtin/cpu_and_mobile/text.json	Drop quant_mode: int4 from extra_options.
Qwen-Qwen3.5-27B/builtin/cuda/text.json	Drop quant_mode: int4 from extra_options.
Qwen-Qwen3.5-0.8B/builtin/webgpu/text.json	Add int4_algo_config: k_quant_linear; drop quant_mode: default.
Qwen-Qwen3.5-0.8B/builtin/cuda/text.json	Add int4_algo_config: k_quant_linear; drop quant_mode: default.
Qwen-Qwen3.5-0.8B/builtin/cpu_and_mobile/text.json	Add int4_algo_config: k_quant_linear; drop quant_mode: default.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.