Downloads · 30 days
18
32% of all-time downloads
mazenDDr/receipt-vlm-qwen2.5-vl-3b-awq
receipt-vlm-qwen2.5-vl-3b-awq is a image-text-to-text model from mazenDDr. Use it for the image-text-to-text task on the model card, and read the license before you ship it in a product. It is set up for transformers. The card lists the license as apache-2.0.
The fine-tuned model quantized to 4-bit with AWQ (W4A16, language model only; the vision tower stays in bf16). 3.39 GB, down from 7.51 GB.
Downloads · 30 days
18
32% of all-time downloads
All-time downloads
56
Public
Parameters
3.8B
3.4 GB on disk
Likes
0
Public
Click a slice to open those files.
.safetensors3.4 GB · 100%
How the weights are stored.
I322.8B · 74%
From the Hugging Face model README
The fine-tuned model quantized to 4-bit with AWQ (W4A16, language model only; the vision tower stays in bf16). 3.39 GB, down from 7.51 GB.
This is the build the project recommends for a GPU server: smallest, fastest, and its accuracy is within noise of the bf16 model it came from.
| Field F1 | Latency p50 | Output tok/s | |
|---|---|---|---|
| Fine-tuned bf16 | 0.852 | 2.07 s | 51.1 |
| This build | 0.847 | 1.05 s | 103.1 |
Paired against its own bf16 source on the same 100 receipts: −0.005 [−0.026, +0.013]. The interval includes zero, so the difference is not real. Calibrated on 256 training receipts with llm-compressor.
vllm serve mazenDDr/receipt-vlm-qwen2.5-vl-3b-awq --max-model-len 4096 --limit-mm-per-prompt '{"image": 1}'
Serve this with vLLM, not transformers. At the time it was produced, transformers 5.14 with compressed-tensors 0.18 could not load this checkpoint (two reader bugs in the packed-quantized format), and would have decompressed it to bf16 anyway — losing the point of the 4-bit weights. vLLM's W4A16 kernels run it natively.
QLoRA on one RTX 5060 Ti (16 GB): the base model in 4-bit NF4 with the vision tower kept in bf16, LoRA rank 16 / alpha 32 on the language-model layers only, learning rate 4e-4, 2 epochs, loss on the answer tokens alone. 1.07 h, 7.6 GB peak VRAM, 29.9 M trainable parameters.
Training data is CORD v2 (CC BY 4.0), split 773 / 99 / 100 after dropping 28 images that appeared in more than one split under a 256-bit perceptual hash. Every choice was made on the dev split; the test split was scored once.
. and , interchangeably as thousands separators. The model
learned those conventions; on receipts from elsewhere, formatting may not match your expectations.Receipts from CORD v2 (Park et al., CORD: A Consolidated Receipt Dataset for Post-OCR Parsing, 2019), CC BY 4.0. Base model Qwen/Qwen2.5-VL-3B-Instruct, Apache-2.0.