GLM-5.3 is live. Run Z.ai's latest model on Friendli Model APIs. Try it today ➜
armand0e
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Model Details
Model Provider
armand0e
Model Tree
Qwen/Qwen3.5-9B
Input Modalities
Output Modalities
Supported Functionality
Benchmarks provided by @nightmedia, as always thanks for taking the time :)
markdown
arc arc/e boolqarmand0e/Qwen3.5-9B-Opus-Agent 0.589 0.747 0.901Jackrong/Qwopus3.5-9B-Coder 0.561 0.721 0.89Qwen3.5-9B 0.571 0.719 0.895
Conducted via BenchLocal. All benchmarks are 2-shot (1 retry on failure) for ease of comparison to the numbers found in Jackrong's Qwopus3.5 Coder
All benchmarks for other models were done in Q8_0 only this model's benchmarks were done in Q4_K_M
Ignore the gemma-4 llama.cpp alias I had set, this was old and I forgot to change it
py
# -*- coding: utf-8 -*-import osfrom unsloth import FastModelimport torchfrom trl import SFTConfig, SFTTrainerfrom teich import mask_data, prepare_dataMAX_SEQ_LEN = 32768MODEL_NAME = os.environ.get("MODEL_NAME", "qwen/Qwen3.5-9B")OUTPUT_DIR = os.environ.get("OUTPUT_DIR", "outputs/qwen-tool-sft")HUB_REPO_ID = os.environ.get("HUB_REPO_ID", "armand0e/Qwen3.5-9B-Opus-Agent")HF_TOKEN = os.environ.get("HF_TOKEN", "")model, tokenizer = FastModel.from_pretrained(model_name=MODEL_NAME,max_seq_length=MAX_SEQ_LEN,load_in_4bit=False,load_in_8bit=False,full_finetuning=False,)model = FastModel.get_peft_model(model,finetune_vision_layers = False, # Turn off for just text!finetune_language_layers = True, # Should leave on!finetune_attention_modules = True, # Attention good for GRPOfinetune_mlp_modules = True, # Should leave on always!r = 32, # Larger = higher accuracy, but might overfitlora_alpha = 64, # Recommended alpha == r at leastlora_dropout = 0,bias = "none",random_state = 3407,)train_dataset = prepare_data({"chat": {"source": "TeichAI/claude-4.5-opus-high-reasoning-250x"},"opus-agent": {"source": "armand0e/badlogicgames-pi-mono-opus-filtered",},},tokenizer,split="train",hf_token=HF_TOKEN,chat_template_kwargs={"enable_thinking": True},max_length=MAX_SEQ_LEN,drop_oversized_examples=True,trim_oversized_followups=True,tokenize=True,strict=True,)trainer = SFTTrainer(model=model,tokenizer=tokenizer,train_dataset=train_dataset,eval_dataset=None,args=SFTConfig(dataset_text_field="text",dataset_num_proc=1,max_length=MAX_SEQ_LEN,packing=False,per_device_train_batch_size=1,gradient_accumulation_steps=8,warmup_steps= 5,num_train_epochs=2,learning_rate=2e-5,logging_steps=1,save_steps=100,save_total_limit=3,optim="adamw_8bit",weight_decay=0.01,max_grad_norm=0.3,lr_scheduler_type="linear",output_dir=OUTPUT_DIR,seed=3407,report_to="none",),)trainer = mask_data(trainer,tokenizer=tokenizer,train_on_reasoning=False,train_on_final_answers=True,train_on_tools=True,)print(trainer.train_dataset.preview())trainer_stats = trainer.train(resume_from_checkpoint=False)model.push_to_hub(f"{HUB_REPO_ID}-LoRA", token=HF_TOKEN)tokenizer.push_to_hub(f"{HUB_REPO_ID}-LoRA", token=HF_TOKEN)model.push_to_hub_merged(HUB_REPO_ID, tokenizer, save_method="merged_16bit", token=HF_TOKEN)
The data for this model was easily formatted and masked with Teich
This qwen3_5 model was trained 2x faster with Unsloth and Huggingface's TRL library.