GLM-5.3 is live. Run Z.ai's latest model on Friendli Model APIs. Try it today ➜
runtime-contracts
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Model Details
Model Provider
runtime-contracts
Model Tree
Qwen/Qwen3-8B
Input Modalities
Output Modalities
Supported Functionality
Fine-tuned with Axolotl 0.13.2, LoRA adapter, 4-bit NF4 quantized base:
| Hyperparameter | Value |
|---|---|
| lora_r | 64 |
| lora_alpha | 128 |
| lora_dropout | 0.05 |
| lora_target_modules | q_proj, k_proj, v_proj, o_proj, gate_proj, up_proj, down_proj |
| learning_rate | 1e-4 |
| lr_scheduler | cosine |
| optimizer | adamw_torch |
| epochs | 3.0 |
| micro_batch_size | 1 |
| gradient_accumulation_steps | 16 |
| sequence_len | 16384 |
| sample_packing | false |
| seed | 1337 |
| training data | paired traces for the "stateless" regime (see paper Appendix for pairing/filtering procedure) |
Released alongside the paper Evaluating Agents Across Runtime Contracts: When Mismatch Costs Efficiency or Quality (IAEval 2026, the NeurIPS 2026 Workshop on Evaluation of Interactive Agents), to reproduce its reported results.
Apache License 2.0. The base model (Qwen3-8B) is also released under Apache 2.0.