GLM-5.3 is live. Run Z.ai's latest model on Friendli Model APIs. Try it today ➜
ABRUBAB
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Model Details
Model Provider
ABRUBAB
Model Tree
Qwen/Qwen3-4B
Input Modalities
Output Modalities
Supported Functionality
python
from peft import AutoPeftModelForCausalLMfrom transformers import AutoTokenizer repo = "ABRUBAB/yinglong-intent-qwen3-4b"tok = AutoTokenizer.from_pretrained(repo)model = AutoPeftModelForCausalLM.from_pretrained(repo, device_map="auto")msgs = [{"role": "user", "content": "Pay my rent on the first of every month, up to 900."}]ids = tok.apply_chat_template(msgs, add_generation_prompt=True, return_tensors="pt").to(model.device)print(tok.decode(model.generate(ids, max_new_tokens=256)[0][ids.shape[1]:], skip_special_tokens=True))Use the same system prompt and JSON schema as the YINGLONG service (ml/intent and packages/mandate in the project
repository) to get the exact output format. Run it in non-thinking mode.
Valid JSON 99.9%, exact match 88.8% (86.9% to 90.4%), slot F1 98.1%.
| Slice | Items | Valid JSON | Exact match | Slot F1 |
|---|---|---|---|---|
| overall | 1255 | 99.9% | 88.8% | 98.1% |
| task=intent | 917 | 100.0% | 86.7% | 97.4% |
| task=mandate | 338 | 99.7% | 94.4% | 99.1% |
| language=bn | 318 | 100.0% | 85.9% | 97.2% |
| language=en | 468 | 100.0% | 93.4% | 99.1% |
| language=mixed | 283 | 99.7% | 79.5% | 96.8% |
| language=zh | 186 | 100.0% | 96.2% | 99.2% |
LoRA rank 16, alpha 16, 2 epochs, learning rate 0.0002, effective batch
16, evaluation context 2048 tokens. Data: synthetic only, written by
Qwen/Qwen3-14B-AWQ; the labels were sampled by code, checked against the JSON schema and verified by a round trip
(the mandate compiler and decompiler). The training data is not published. Languages: English, Bengali, Chinese and
code-mixed.
Part of YINGLONG, by team Shadow Chaser (Bangladesh), for the 2026 Shenzhen International FinTech Competition. Code and evidence: https://github.com/ABRUBAB/YINGLONG . Live demo: https://app.yinglong-agent.tech .
from peft import AutoPeftModelForCausalLMfrom transformers import AutoTokenizer repo = "ABRUBAB/yinglong-intent-qwen3-4b"tok = AutoTokenizer.from_pretrained(repo)model = AutoPeftModelForCausalLM.from_pretrained(repo, device_map="auto")msgs = [{"role": "user", "content": "Pay my rent on the first of every month, up to 900."}]ids = tok.apply_chat_template(msgs, add_generation_prompt=True, return_tensors="pt").to(model.device)print(tok.decode(model.generate(ids, max_new_tokens=256)[0][ids.shape[1]:], skip_special_tokens=True))