Add files using upload-large-folder tool

README.md

@@ -26,6 +26,17 @@ This Hub release is the **merged full-weight** GRPO export — a single `model.s
 
 **Training data:** [`ai-mind-lab/CineMR`](https://huggingface.co/datasets/ai-mind-lab/CineMR) (ACDC, M&Ms, M&Ms-2 cardiac MRI VQA with optional tool-use supervision).
 
+## Authors
+
+Kunyang Li<sup>1,†</sup>, Hai Nguyen<sup>1,2,†</sup>, Joshua Lowe<sup>1,†</sup>, Chenguang Zhao<sup>3</sup>, Peace C. Madueme<sup>3</sup>, Mehdi Hedjazi Moghari<sup>4</sup>, Mubarak Shah<sup>1,§</sup>, Pegah Khosravi<sup>1,2,§</sup>, Yuzhang Shang<sup>1,§</sup>
+
+<sup>1</sup> Institute for Artificial Intelligence, University of Central Florida
+<sup>2</sup> Department of Clinical Sciences, College of Medicine, University of Central Florida
+<sup>3</sup> Nemours Children's Health, Orlando, Florida
+<sup>4</sup> West Virginia University Medicine Children's Hospital, Morgantown, West Virginia
+
+<sup>†</sup> Co-first author &nbsp;·&nbsp; <sup>§</sup> Corresponding author
+
 ## Model summary
 
 | | |
@@ -93,13 +104,21 @@ text = processor.apply_chat_template(messages, tokenize=False, add_generation_pr
 inputs = processor(text=[text], images=[image], return_tensors="pt").to(model.device)
 
 with torch.no_grad():
-    out = model.generate(**inputs, max_new_tokens=512, do_sample=False)
+    out = model.generate(
+        **inputs,
+        max_new_tokens=2048,
+        do_sample=True,
+        temperature=0.7,
+        repetition_penalty=1.15,
+    )
 
 print(processor.decode(out[0], skip_special_tokens=True))
 ```
 
 Use the same `trust_remote_code=True` and bfloat16 settings as in training. For evaluation, match the CineMR prompt template and decoding settings used in your eval script.
 
+> **Decoding note.** Pure greedy decoding (`do_sample=False`, no repetition penalty) can drive this checkpoint into **repetition collapse** (a single reasoning sentence repeated until the token cap, with no `\boxed{}` answer or `<tool_call>` emitted). The evaluation numbers below were produced with `do_sample=True`, `temperature=0.7`, `repetition_penalty=1.15`, `no_repeat_ngram_size=0`, `max_new_tokens=2048`, and 4 sampled rollouts per prompt. Use a repetition penalty (≈1.1–1.2) for stable outputs.
+
 ## Training procedure (summary)
 
 1. **SFT** on CineMR JSONL (train split) starting from Qwen3-VL-8B-Instruct; weights merged to a full `transformers` checkpoint.
@@ -113,7 +132,21 @@ LoRA weights are merged into the base checkpoint for Hub deployment.
 
 ## Evaluation
 
-Evaluate on the CineMR test split with the same frame paths and prompt template as training. Report metrics on extracted `\boxed{}` answers and, if applicable, tool-call correctness. See the project eval scripts under `CardiacCine/cardiac_cine/cine-cogito/` for reference pipelines.
+Evaluated on the **CineMR test split** (1,191 samples) with the training prompt template, 4 sampled rollouts per prompt (`temperature=0.7`, `repetition_penalty=1.15`, `max_new_tokens=2048`). `pass@k` is the fraction of items with ≥1 correct rollout; `mean rollout acc` averages correctness over all rollouts.
+
+| Metric | Value |
+|---|---|
+| Mean rollout accuracy | 0.378 |
+| pass@4 (any correct) | 0.553 |
+| ROUGE-L | 0.620 |
+| BERTScore F1 | 0.974 |
+| Ground-truth satisfied | 0.370 |
+
+**Accuracy by reasoning layer (pass@4):** L1 0.306, L2 0.771, L3 0.906, L4 0.614, L5 0.314, L6 0.563. By clinical stage: phenotype (L1–L4) pass@4 0.572, etiology (L5–L6) pass@4 0.353.
+
+**Tool use:** tool-decision accuracy 0.891 (precision 0.999), tool recall on required items 99.8% (predicted names ⊆ expected 100%), trace/JSON format validity 99.4%, tool-name set-match 0.862, argument accuracy 0.879.
+
+Numbers are from this GRPO checkpoint evaluated with the project `eval_sft` pipeline. Re-run on a held-out test split before drawing conclusions; the small GRPO validation set is used only for checkpoint tracking.
 
 ## Limitations
 
@@ -131,8 +164,8 @@ If you use CineMR, please cite the base Qwen3-VL model and acknowledge the CineM
 
 ```bibtex
 @misc{cinemr_qwen3vl8b_grpo,
-  title        = {CineMR: Cardiac MRI Vision-Language Model (Qwen3-VL-8B, GRPO)},
-  author       = {AI Mind Lab},
+  title        = {CineMR: Augmenting Vision-Language Models with Tool-Integrated Reasoning for Quantitative Cardiac MRI Diagnosis},
+  author       = {Li, Kunyang and Nguyen, Hai and Lowe, Joshua and Zhao, Chenguang and Madueme, Peace C. and Moghari, Mehdi Hedjazi and Shah, Mubarak and Khosravi, Pegah and Shang, Yuzhang},
   year         = {2026},
   howpublished = {\url{https://huggingface.co/ai-mind-lab/CineMR}},
   note         = {GRPO checkpoint; dataset at huggingface.co/datasets/ai-mind-lab/CineMR},

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:455f92d2f358a15d62a4f934970a876d66a3edd7618d677d7275b5b0e131ba53
+oid sha256:e069a4035458cc235f9b782900a5e17a4ad6d8f379c66cdf0759b0ed7fe5564b
 size 17534340584