GLM-5.3 is live. Run Z.ai's latest model on Friendli Model APIs. Try it today ➜
ikimyaii
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Run this model inference with full control and performance in your environment.
Model Details
Model Provider
ikimyaii
Model Tree
openlm-research/open_llama_7b
Input Modalities
Output Modalities
Supported Functionality
| Base model | openlm-research/open_llama_7b |
| Method | AWP |
| Target / achieved sparsity | 30% / 30.01% |
| Calibration | 64 batches × 4 × 512 tokens |
| Weight correction | none (mask search only) |
| Retraining | none |
What is pruned: every nn.Linear in the 32 decoder layers (attention q/k/v/o and MLP gate/up/down) plus lm_head, 225 layers in total (see pruning_summary.json). Embeddings and norms stay dense.
AWP (Liu et al., 2025, arXiv:2506.10205) minimises the activation-weighted reconstruction error ‖(W − Θ) C^{1/2}‖²_F, C = E[xxᵀ], by iterative hard thresholding: Z = Θ + η (W − Θ) C, η = 2/‖C‖_F, then keep the top-k entries per row. It is initialised from the Wanda mask and runs up to 200 iterations (tolerance 1e-4).
WikiText-2 PPL of all five methods at 30% (dense 8.64):
| Method | WikiText-2 PPL ↓ |
|---|---|
| CWS | 9.12 |
| SparseGPT | 9.18 |
| Wanda | 9.04 |
| RIA | 8.96 |
| AWP (this model) | 8.92 |
These are the project's results at the time the checkpoints were uploaded (2026-05-11; results/sparsity_sweep_llama.json @ 6e144ae), with dense PPL 8.64. The repository README now reports these models under a 2048-token protocol (dense 6.61); the ordering between methods is similar.
python
import torchfrom transformers import AutoModelForCausalLM, LlamaTokenizer model_id = "ikimyaii/llama-7b-awp-30pct"tokenizer = LlamaTokenizer.from_pretrained(model_id)model = AutoModelForCausalLM.from_pretrained(model_id, torch_dtype=torch.float16, device_map="auto") lin = model.model.layers[0].mlp.down_projprint(f"sparsity: {(lin.weight == 0).float().mean():.3f}")import torchfrom transformers import AutoModelForCausalLM, LlamaTokenizer model_id = "ikimyaii/llama-7b-awp-30pct"tokenizer = LlamaTokenizer.from_pretrained(model_id)model = AutoModelForCausalLM.from_pretrained(model_id, torch_dtype=torch.float16, device_map="auto") lin = model.model.layers[0].mlp.down_projprint(f"sparsity: {(lin.weight == 0).float().mean():.3f}")