How to use from the
Use from the
PEFT library
from peft import PeftModel
from transformers import AutoModelForCausalLM

base_model = AutoModelForCausalLM.from_pretrained("ggml-org-dev/tiny-Qwen3.5-4B-for-testing")
model = PeftModel.from_pretrained(base_model, "ggml-org-dev/tiny-lev-for-testing")

tiny-lev-for-testing

This is a modified, damaged copy of interfaze-ai/lev, made only to test software. Its output is gibberish. Do not use it for anything else.

lev is a LoRA adapter for Qwen/Qwen3.5-4B. This copy is a LoRA adapter for the tiny base model ggml-org-dev/tiny-Qwen3.5-4B-for-testing, a sliced copy of Qwen3.5-4B. It keeps the calibration, label readout prompt (chat template) and LoRA layout of lev, so it runs the same code paths as the real model in llama.cpp (lev decision type: the converter downloads the base model named in adapter_config.json and merges the adapter), but it is small enough for CI.

GGUF version: ggml-org-dev/tiny-lev-for-testing-GGUF

Changes from the original adapter

Sliced with slice_qwen35_lora.py, which follows the slice-model skill of llama.cpp:

  • base_model_name_or_path points at ggml-org-dev/tiny-Qwen3.5-4B-for-testing
  • LoRA tensors of 4 of 32 layers (0, 1, 2, 3), narrowed with the same indices as the base model: A keeps the input columns of its base weight, B its output rows (rank 32 and alpha kept)
  • mode_b_head.pt (head for large option sets, not used by llama.cpp): the input projections are narrowed to the hidden size 32
  • The README images (assets/) are not included

License

Derived from lev by Interfaze AI, under the Apache License 2.0 (as is the base model Qwen3.5-4B).

Downloads last month
-
Inference Providers NEW
This model isn't deployed by any Inference Provider. ๐Ÿ™‹ Ask for provider support

Model tree for ggml-org-dev/tiny-lev-for-testing

Finetuned
Qwen/Qwen3.5-4B
Adapter
(1)
this model
Quantizations
1 model