Sponsored by freebeat AI
Toolify Toolify logo
AI Models
diffvqa-qwen3-0.6b-multiobjective-regsteps

dmusingu / diffvqa-qwen3-0.6b-multiobjective-regsteps

Run on huggingface.co
huggingface.co
Total runs: 0
24-hour runs: 0
7-day runs: 0
30-day runs: 0
Model's Last Updated: July 18 2026
visual-question-answering

Introduction of diffvqa-qwen3-0.6b-multiobjective-regsteps

Model Details of diffvqa-qwen3-0.6b-multiobjective-regsteps

Diff-VQA — Qwen3-0.6B visual-prefix decoder (multiobjective-regsteps encoder)

Online-trained Difference Visual Question Answering head for chest X-rays (MIMIC-CXR / Medical-Diff-VQA). A frozen ViT-L/14 vision encoder produces patch tokens for a current + reference image pair; a Qwen3-0.6B decoder, conditioned on those tokens as a visual prefix, generates the answer describing what changed.

This is a slimmed inference checkpoint : decoder weights + the frozen vision encoder weights + architecture metadata (optimizer/scheduler state stripped).

Contents ( *_best.pt , torch.load(..., weights_only=False) )
key description
state_dict DiffVQAQwen3Head weights (Qwen3-0.6B LLM + vis_proj + frame_emb )
encoder_state frozen ViT-L/14 vision encoder ( objective: multiobjective_caption+regioncontrastive )
decoder Qwen/Qwen3-0.6B
tokenizer_name Qwen/Qwen3-0.6B
vis_dim 1024
encoder_name pretrain-multiobjective-regsteps
epoch / step / best_val training position + best validation loss
Training
  • Data: Medical-Diff-VQA difference questions over MIMIC-CXR image pairs (~100k train / ~16k val / ~16k test).
  • Objective: next-token cross-entropy over the answer span (question tokens masked).
  • Vision encoder is frozen ; only the decoder (LLM + visual adapter) is trained.
Notes

The bundled encoder_state is the exact frozen encoder this decoder was trained against — load both together so the visual features match the space vis_proj expects (a mismatched encoder produces degenerate output).

Runs of dmusingu diffvqa-qwen3-0.6b-multiobjective-regsteps on huggingface.co

0
Total runs
0
24-hour runs
0
3-day runs
0
7-day runs
0
30-day runs

More Information About diffvqa-qwen3-0.6b-multiobjective-regsteps huggingface.co Model

More diffvqa-qwen3-0.6b-multiobjective-regsteps license Visit here:

https://choosealicense.com/licenses/mit

diffvqa-qwen3-0.6b-multiobjective-regsteps huggingface.co

diffvqa-qwen3-0.6b-multiobjective-regsteps huggingface.co is an AI model on huggingface.co that provides diffvqa-qwen3-0.6b-multiobjective-regsteps's model effect (), which can be used instantly with this dmusingu diffvqa-qwen3-0.6b-multiobjective-regsteps model. huggingface.co supports a free trial of the diffvqa-qwen3-0.6b-multiobjective-regsteps model, and also provides paid use of the diffvqa-qwen3-0.6b-multiobjective-regsteps. Support call diffvqa-qwen3-0.6b-multiobjective-regsteps model through api, including Node.js, Python, http.

diffvqa-qwen3-0.6b-multiobjective-regsteps huggingface.co Url

https://huggingface.co/dmusingu/diffvqa-qwen3-0.6b-multiobjective-regsteps

dmusingu diffvqa-qwen3-0.6b-multiobjective-regsteps online free

diffvqa-qwen3-0.6b-multiobjective-regsteps huggingface.co is an online trial and call api platform, which integrates diffvqa-qwen3-0.6b-multiobjective-regsteps's modeling effects, including api services, and provides a free online trial of diffvqa-qwen3-0.6b-multiobjective-regsteps, you can try diffvqa-qwen3-0.6b-multiobjective-regsteps online for free by clicking the link below.

dmusingu diffvqa-qwen3-0.6b-multiobjective-regsteps online free url in huggingface.co:

https://huggingface.co/dmusingu/diffvqa-qwen3-0.6b-multiobjective-regsteps

diffvqa-qwen3-0.6b-multiobjective-regsteps install

diffvqa-qwen3-0.6b-multiobjective-regsteps is an open source model from GitHub that offers a free installation service, and any user can find diffvqa-qwen3-0.6b-multiobjective-regsteps on GitHub to install. At the same time, huggingface.co provides the effect of diffvqa-qwen3-0.6b-multiobjective-regsteps install, users can directly use diffvqa-qwen3-0.6b-multiobjective-regsteps installed effect in huggingface.co for debugging and trial. It also supports api for free installation.

diffvqa-qwen3-0.6b-multiobjective-regsteps install url in huggingface.co:

https://huggingface.co/dmusingu/diffvqa-qwen3-0.6b-multiobjective-regsteps

Url of diffvqa-qwen3-0.6b-multiobjective-regsteps

diffvqa-qwen3-0.6b-multiobjective-regsteps huggingface.co Url

Provider of diffvqa-qwen3-0.6b-multiobjective-regsteps huggingface.co

dmusingu
ORGANIZATIONS
huggingface.co

Other API from dmusingu

 huggingface.co
dmusingu/lapvqa-pg

Total runs: 0
Run Growth: 0
Growth Rate: 0.00%
Updated:June 06 2026
huggingface.co
dmusingu/lapvqa-ad

Total runs: 0
Run Growth: 0
Growth Rate: 0.00%
Updated:June 06 2026