You can deploy this model and start earning money today!

ZhuofengLi/
Qwen3.5-27B-hardtests-sft-synthetic-merged

Name: Qwen3.5-27B-hardtests-sft-synthetic-merged
Availability: InStock

Views

No views yet

Qwen3.5-27B-hardtests-sft-synthetic (LoRA merged)

Model Description

This is a full-weight model produced by merging the LoRA adapter ZhuofengLi/Qwen3.5-27B-hardtests-sft-synthetic-lora into its base model. The adapter was SFT-trained on synthetic hard-tests data for code/reasoning tasks.

Base Model

Base: Qwen/Qwen3.5-27B

Merge Details

Adapter: LoRA (rank=16, alpha=32, dropout=0.05), target modules q/k/v/o/gate/up/down projections
Framework: ms-swift 4.4.2 (swift export --merge_lora)
Precision: bf16

Usage

python

from transformers import AutoModelForCausalLM, AutoTokenizer

model = AutoModelForCausalLM.from_pretrained("ZhuofengLi/Qwen3.5-27B-hardtests-sft-synthetic-merged", torch_dtype="auto", device_map="auto")
tokenizer = AutoTokenizer.from_pretrained("ZhuofengLi/Qwen3.5-27B-hardtests-sft-synthetic-merged")