Create README.md

README.md

@@ -0,0 +1,154 @@
+---
+language:
+- en
+library_name: transformers
+tags:
+- glm
+- MOE
+- pruning
+- compression
+- unsloth
+license: mit
+license_link: https://huggingface.co/zai-org/GLM-4.7/blob/main/LICENSE
+pipeline_tag: text-generation
+base_model:
+- cerebras/GLM-4.7-REAP-218B-A32B
+---
+> [!NOTE]
+>  Includes Unsloth **chat template fixes**! <br> For `llama.cpp`, use `--jinja`
+>
+
+<div>
+<p style="margin-top: 0;margin-bottom: 0;">
+    <em><a href="https://docs.unsloth.ai/basics/unsloth-dynamic-v2.0-gguf">Unsloth Dynamic 2.0</a> achieves superior accuracy & outperforms other leading quants.</em>
+  </p>
+  <div style="display: flex; gap: 5px; align-items: center; ">
+    <a href="https://github.com/unslothai/unsloth/">
+      <img src="https://github.com/unslothai/unsloth/raw/main/images/unsloth%20new%20logo.png" width="133">
+    </a>
+    <a href="https://discord.gg/unsloth">
+      <img src="https://github.com/unslothai/unsloth/raw/main/images/Discord%20button.png" width="173">
+    </a>
+    <a href="https://docs.unsloth.ai/">
+      <img src="https://raw.githubusercontent.com/unslothai/unsloth/refs/heads/main/images/documentation%20green%20button.png" width="143">
+    </a>
+  </div>
+</div>
+
+<p align="center">
+  <em>𓌳 <strong>REAP</strong>𓌳  the Experts: Why Pruning Prevails for One-Shot MoE Compression</em><br>
+  <img src="https://i.imgur.com/rmzG3gg.png" alt="REAP" width="75%">
+</p>
+
+# GLM-4.7-REAP-218B-A32B
+
+## ✨ Highlights
+
+Introducing **GLM-4.7-REAP-218B-A32B**, a **memory-efficient compressed variant** of GLM-4.7 that maintains near-identical performance while being **40% lighter**.
+
+This model was created using **REAP (Router-weighted Expert Activation Pruning)**, a novel expert pruning method that selectively removes redundant experts while preserving the router's independent control over remaining experts. Key features include:
+
+- **Near-Lossless Performance**: Maintains almost identical accuracy on code generation, agentic coding, and function calling tasks compared to the full 355B model
+- **40% Memory Reduction**: Compressed from 355B to 218B parameters, significantly lowering deployment costs and memory requirements
+- **Preserved Capabilities**: Retains all core functionalities including code generation, agentic workflows, repository-scale understanding, and function calling
+- **Drop-in Compatibility**: Works with vanilla vLLM - no source modifications or custom patches required
+- **Optimized for Real-World Use**: Particularly effective for resource-constrained environments, local deployments, and academic research
+
+**For downstream low-bit quantization, we suggest using the [BF16 variant](https://huggingface.co/cerebras/GLM-4.7-REAP-218B-A32B).**
+
+---
+## 📋 Model Overview
+
+**GLM-4.7-REAP-218B-A32B** has the following specifications:
+
+- **Base Model**: GLM-4.7
+- **Compression Method**: REAP (Router-weighted Expert Activation Pruning)
+- **Compression Ratio**: 40% expert pruning
+- **Type**: Sparse Mixture-of-Experts (SMoE) Causal Language Model
+- **Number of Parameters**: 218B total, 32B activated per token
+- **Number of Layers**: 92
+- **Number of Attention Heads (GQA)**: 96 for Q and 8 for KV
+- **Number of Experts**: 96 (uniformly pruned from 160)
+- **Number of Activated Experts**: 8 per token
+- **Context Length**: 202,752 tokens
+- **License**: MIT
+
+---
+
+## 📊 Evaluations
+
+TBD for BF16 model. [Evalulation results available for the FP8 variant](https://huggingface.co/cerebras/GLM-4.7-REAP-218B-A32B-FP8#%F0%9F%93%8A-evaluations).
+
+For more details on the evaluation setup, refer to the [REAP arXiv preprint](https://arxiv.org/abs/2510.13999).
+
+---
+
+## 🚀 Deployment
+
+You can deploy the model directly using the **latest vLLM** (v0.11.0), no source modifications or custom patches required.
+
+```bash
+vllm serve cerebras/GLM-4.7-REAP-218B-A32B \
+    --tensor-parallel-size 8 \
+    --reasoning-parser glm45 \
+    --tool-call-parser glm47 \
+    --enable-auto-tool-choice \
+    --enable-expert-parallel
+```
+
+If you encounter insufficient memory when running this model, you might need to set a lower value for `--max-num-seqs` flag (e.g. set to 64).
+
+
+## 🧩 Model Creation
+
+This checkpoint was created by applying the **REAP (Router-weighted Expert Activation Pruning)** method uniformly across all Mixture-of-Experts (MoE) blocks of **GLM-4.7**, with a **40% pruning rate**.
+
+### How REAP Works
+
+REAP selects experts to prune based on a novel **saliency criterion** that considers both:
+- **Router gate values**: How frequently and strongly the router activates each expert
+- **Expert activation norms**: The magnitude of each expert's output contributions
+
+This dual consideration ensures that experts contributing minimally to the layer's output are pruned, while preserving those that play critical roles in the model's computations.
+
+### Key Advantages
+
+- **One-Shot Compression**: No fine-tuning required after pruning - the model is immediately ready for deployment
+- **Preserved Router Control**: Unlike expert merging methods, REAP maintains the router's independent, input-dependent control over remaining experts, avoiding "functional subspace collapse"
+- **Generative Task Superiority**: REAP significantly outperforms expert merging approaches on generative benchmarks (code generation, creative writing, mathematical reasoning) while maintaining competitive performance on discriminative tasks
+
+### Calibration
+
+The model was calibrated using a diverse mixture of domain-specific datasets including:
+- Code generation samples ([evol-codealpaca](https://huggingface.co/datasets/theblackcat102/evol-codealpaca-v1))
+- Function calling examples ([xlam-function-calling](https://huggingface.co/datasets/Salesforce/xlam-function-calling-60k))
+- Agentic multi-turn trajectories ([SWE-smith-trajectories](https://huggingface.co/datasets/SWE-bench/SWE-smith-trajectories))
+
+📚 For more details, refer to the following resources:
+
+- [🧾 arXiv Preprint](https://arxiv.org/abs/2510.13999)
+- [🧾 REAP Blog](https://www.cerebras.ai/blog/reap)
+- [💻 REAP Codebase (GitHub)](https://github.com/CerebrasResearch/reap)
+
+---
+
+## ⚖️ License
+
+This model is derived from
+**[`zai-org/GLM-4.7`](https://huggingface.co/zai-org/GLM-4.7)**
+and distributed under the **MIT license**.
+
+---
+
+## 🧾 Citation
+
+If you use this checkpoint, please cite the REAP paper:
+
+```bibtex
+@article{lasby-reap,
+  title={REAP the Experts: Why Pruning Prevails for One-Shot MoE compression},
+  author={Lasby, Mike and Lazarevich, Ivan and Sinnadurai, Nish and Lie, Sean and Ioannou, Yani and Thangarasa, Vithursan},
+  journal={arXiv preprint arXiv:2510.13999},
+  year={2025}
+}
+```