README.md

---
license: mit
base_model:
- zai-org/GLM-4.7
---

# Model Overview

- **Model Architecture:** GLM-4.7
  - **Input:** Text
  - **Output:** Text
- **Supported Hardware Microarchitecture:** AMD MI350/MI355
- **ROCm:** 7.0
- **Operating System(s):** Linux
- **Inference Engine:** [vLLM](https://docs.vllm.ai/en/latest/)
- **Model Optimizer:** [AMD-Quark](https://quark.docs.amd.com/latest/index.html) (V0.11)
  - **moe**
    - **Weight quantization:** MOE-only, OCP MXFP4, Static
    - **Activation quantization:** MOE-only, OCP MXFP4, Dynamic
  - **KV cache quantization:** OCP FP8, Static
- **Calibration Dataset:** [Pile](https://huggingface.co/datasets/mit-han-lab/pile-val-backup)

This model was built with GLM-4.7 model by applying [AMD-Quark](https://quark.docs.amd.com/latest/index.html) for MXFP4 quantization.

# Model Quantization

The model was quantized from [zai-org/GLM-4.7](https://huggingface.co/zai-org/GLM-4.7) using [AMD-Quark](https://quark.docs.amd.com/latest/index.html). The weights and activations are quantized to MXFP4.
AMD-Quark has been installed from source code inside the Docker image `rocm/vllm-private:vllm_dev_base_mxfp4_20260122`.

**Quantization scripts:**

Note that GLM-4.7 is not in the built-in model template list in Quark V0.11, it has to be registered before quantization.

- **Step1:** Register model template: creat fle `Quark/examples/torch/language_modeling/llm_ptq/quantize_glm.py`
```
import runpy
from quark.torch import LLMTemplate

# Register GLM-4 MoE template
glm4_moe_template = LLMTemplate(
    model_type="glm4_moe",
    kv_layers_name=["*k_proj", "*v_proj"],
    q_layer_name="*q_proj",
    exclude_layers_name=["lm_head","*mlp.gate","*self_attn*","*shared_experts.*","*mlp.down_proj","*mlp.gate_proj","*mlp.up_proj"],
)
LLMTemplate.register_template(glm4_moe_template)
print(f"[INFO]: Registered template '{glm4_moe_template.model_type}'")

# Run quantize_quark.py
# Get the absolute path to the quantize_quark.py script
quantize_script = "/app/Quark/examples/torch/language_modeling/llm_ptq/quantize_quark.py"

runpy.run_path(quantize_script, run_name="__main__")
```
- **Step2:** Quantize with the quantize_glm.py
```
export CUDA_VISIBLE_DEVICES=0,1,2,3
export MODEL_DIR=zai-org/GLM-4.7
export output_dir=amd/GLM-4.7-MXFP4

exclude_layers="*self_attn* *mlp.gate lm_head *mlp.gate_proj *mlp.up_proj *mlp.down_proj *shared_experts.*"
python3 quantize_glm.py --model_dir $MODEL_DIR \
                        --quant_scheme mxfp4 \
                        --num_calib_data 128 \
                        --exclude_layers $exclude_layers \
                        --kv_cache_dtype fp8 \
                        --model_export hf_format \
                        --output_dir $output_dir \
                        --multi_gpu
```

# Deployment
### Use with vLLM

This model can be deployed efficiently using the [vLLM](https://docs.vllm.ai/en/latest/) backend.

## Evaluation
The model was evaluated on GSM8K benchmarks. 

### Accuracy

<table>
  <tr>
   <td><strong>Benchmark</strong>
   </td>
   <td><strong>GLM-4.7 </strong>
   </td>
   <td><strong>GLM-4.7-MXFP4(this model)</strong>
   </td>
   <td><strong>Recovery</strong>
   </td>
  </tr>
  <tr>
   <td>GSM8K (strict-match)
   </td>
   <td>94.16
   </td>
   <td>93.63
   </td>
   <td>99.44%
   </td>
  </tr>
</table>

### Reproduction

The GSM8K results were obtained using the `lm-evaluation-harness` framework, based on the Docker image `rocm/vllm-private:vllm_dev_base_mxfp4_20260122`, with vLLM, lm-eval and amd-quark compiled and installed from source inside the image.

#### Launching server
```
vllm serve amd/GLM-4.7-MXFP4 \
    --tensor-parallel-size 4 \
    --tool-call-parser glm47 \
    --reasoning-parser glm45 \
    --enable-auto-tool-choice \
    --kv_cache_dtype fp8
```

#### Evaluating model in a new terminal
```
lm_eval \
  --model local-completions \
  --model_args "model=amd/GLM-4.7-MXFP4,base_url=http://0.0.0.0:8000/v1/completions,tokenized_requests=False,tokenizer_backend=None,num_concurrent=32" \
  --tasks gsm8k \
  --num_fewshot 5 \
  --batch_size 1
```

# License
Modifications Copyright(c) 2025 Advanced Micro Devices, Inc. All rights reserved.