Copy disabled (too large)
Download .txt
Showing preview only (43,235K chars total). Download the full file to get everything.
Repository: richard-peng-xia/MMed-RAG
Branch: main
Commit: 45ceed33eef4
Files: 283
Total size: 64.5 MB
Directory structure:
gitextract_tizp684z/
├── .gitignore
├── LICENSE
├── README.md
├── data/
│ ├── test/
│ │ ├── report/
│ │ │ ├── harvard_test.json
│ │ │ ├── iuxray_test.json
│ │ │ ├── mimic_test.json
│ │ │ ├── pmc-oa_test.json
│ │ │ └── quilt-1m_test.json
│ │ └── vqa/
│ │ ├── harvard_test.jsonl
│ │ ├── iuxray_test.jsonl
│ │ ├── mimic_test.jsonl
│ │ ├── pmc-oa_test.jsonl
│ │ └── quilt-1m_test.jsonl
│ └── training/
│ ├── alignment/
│ │ ├── ophthalmology/
│ │ │ ├── harvard_report.json
│ │ │ └── harvard_vqa.json
│ │ ├── pathology/
│ │ │ └── pathology_vqa.json
│ │ └── radiology/
│ │ ├── radiology_report.json
│ │ └── radiology_vqa.json
│ └── retriever/
│ ├── ophthalmology/
│ │ ├── harvard_train_7000.json
│ │ └── harvard_val_1000.json
│ ├── pathology/
│ │ ├── pathology_train.json
│ │ └── pathology_val.json
│ └── radiology/
│ ├── radiology_train.json
│ └── radiology_val.json
├── requirements.txt
├── scripts/
│ ├── finetune_clip.sh
│ ├── retrieve_clip_VQA.sh
│ ├── retrieve_clip_report.sh
│ └── train_dpo_2stages.sh
└── train/
├── dpo/
│ ├── LICENSE
│ ├── cog.yaml
│ ├── dpo_trainer_2stages.py
│ ├── llava/
│ │ ├── __init__.py
│ │ ├── constants.py
│ │ ├── conversation.py
│ │ ├── conversation_new.py
│ │ ├── eval/
│ │ │ ├── eval_gpt_review.py
│ │ │ ├── eval_gpt_review_bench.py
│ │ │ ├── eval_gpt_review_visual.py
│ │ │ ├── eval_pope.py
│ │ │ ├── eval_science_qa.py
│ │ │ ├── eval_science_qa_gpt4.py
│ │ │ ├── eval_science_qa_gpt4_requery.py
│ │ │ ├── eval_textvqa.py
│ │ │ ├── generate_webpage_data_from_table.py
│ │ │ ├── m4c_evaluator.py
│ │ │ ├── model_qa.py
│ │ │ ├── model_vqa.py
│ │ │ ├── model_vqa_loader.py
│ │ │ ├── model_vqa_mmbench.py
│ │ │ ├── model_vqa_science.py
│ │ │ ├── qa_baseline_gpt35.py
│ │ │ ├── run_llava.py
│ │ │ ├── summarize_gpt_review.py
│ │ │ ├── table/
│ │ │ │ ├── answer/
│ │ │ │ │ ├── answer_alpaca-13b.jsonl
│ │ │ │ │ ├── answer_bard.jsonl
│ │ │ │ │ ├── answer_gpt35.jsonl
│ │ │ │ │ ├── answer_llama-13b.jsonl
│ │ │ │ │ └── answer_vicuna-13b.jsonl
│ │ │ │ ├── caps_boxes_coco2014_val_80.jsonl
│ │ │ │ ├── model.jsonl
│ │ │ │ ├── prompt.jsonl
│ │ │ │ ├── question.jsonl
│ │ │ │ ├── results/
│ │ │ │ │ ├── test_sqa_llava_13b_v0.json
│ │ │ │ │ └── test_sqa_llava_lcs_558k_sqa_12e_vicuna_v1_3_13b.json
│ │ │ │ ├── review/
│ │ │ │ │ ├── review_alpaca-13b_vicuna-13b.jsonl
│ │ │ │ │ ├── review_bard_vicuna-13b.jsonl
│ │ │ │ │ ├── review_gpt35_vicuna-13b.jsonl
│ │ │ │ │ └── review_llama-13b_vicuna-13b.jsonl
│ │ │ │ ├── reviewer.jsonl
│ │ │ │ └── rule.json
│ │ │ └── webpage/
│ │ │ └── styles.css
│ │ ├── mm_utils.py
│ │ ├── model/
│ │ │ ├── __init__.py
│ │ │ ├── apply_delta.py
│ │ │ ├── builder.py
│ │ │ ├── consolidate.py
│ │ │ ├── language_model/
│ │ │ │ ├── llava_llama.py
│ │ │ │ ├── llava_mistral.py
│ │ │ │ └── llava_mpt.py
│ │ │ ├── llava_arch.py
│ │ │ ├── make_delta.py
│ │ │ ├── multimodal_encoder/
│ │ │ │ ├── builder.py
│ │ │ │ └── clip_encoder.py
│ │ │ ├── multimodal_projector/
│ │ │ │ └── builder.py
│ │ │ └── utils.py
│ │ ├── serve/
│ │ │ ├── __init__.py
│ │ │ ├── cli.py
│ │ │ ├── controller.py
│ │ │ ├── gradio_web_server.py
│ │ │ ├── model_worker.py
│ │ │ ├── register_worker.py
│ │ │ ├── sglang_worker.py
│ │ │ └── test_message.py
│ │ ├── train/
│ │ │ ├── llama_flash_attn_monkey_patch.py
│ │ │ ├── llama_xformers_attn_monkey_patch.py
│ │ │ ├── llava_trainer.py
│ │ │ ├── train.py
│ │ │ ├── train_dpo.py
│ │ │ ├── train_dpo_inherent.py
│ │ │ ├── train_mem.py
│ │ │ └── train_xformers.py
│ │ └── utils.py
│ ├── llava_trainer_2stages.py
│ ├── povid_infer.py
│ ├── predict.py
│ ├── pyproject.toml
│ ├── scripts/
│ │ ├── convert_gqa_for_eval.py
│ │ ├── convert_mmbench_for_submission.py
│ │ ├── convert_mmvet_for_eval.py
│ │ ├── convert_seed_for_submission.py
│ │ ├── convert_sqa_to_llava.py
│ │ ├── convert_sqa_to_llava_base_prompt.py
│ │ ├── convert_vizwiz_for_submission.py
│ │ ├── convert_vqav2_for_submission.py
│ │ ├── extract_mm_projector.py
│ │ ├── finetune.sh
│ │ ├── finetune_full_schedule.sh
│ │ ├── finetune_lora.sh
│ │ ├── finetune_qlora.sh
│ │ ├── finetune_sqa.sh
│ │ ├── merge_lora_weights.py
│ │ ├── pretrain.sh
│ │ ├── pretrain_xformers.sh
│ │ ├── run_povid.sh
│ │ ├── sqa_eval_batch.sh
│ │ ├── sqa_eval_gather.sh
│ │ ├── upload_pypi.sh
│ │ ├── v1_5/
│ │ │ ├── eval/
│ │ │ │ ├── gqa.sh
│ │ │ │ ├── llavabench.sh
│ │ │ │ ├── mmbench.sh
│ │ │ │ ├── mmbench_cn.sh
│ │ │ │ ├── mme.sh
│ │ │ │ ├── mmvet.sh
│ │ │ │ ├── pope.sh
│ │ │ │ ├── qbench.sh
│ │ │ │ ├── qbench_zh.sh
│ │ │ │ ├── seed.sh
│ │ │ │ ├── sqa.sh
│ │ │ │ ├── textvqa.sh
│ │ │ │ ├── vizwiz.sh
│ │ │ │ └── vqav2.sh
│ │ │ ├── finetune.sh
│ │ │ ├── finetune_lora.sh
│ │ │ ├── finetune_task.sh
│ │ │ ├── finetune_task_lora.sh
│ │ │ └── pretrain.sh
│ │ ├── zero2.json
│ │ ├── zero3.json
│ │ └── zero3_offload.json
│ ├── tool/
│ │ ├── dpo_trainer.py
│ │ └── dpo_trainer_inherent.py
│ └── train_dpo_2stages.py
└── open_clip/
├── CITATION.cff
├── HISTORY.md
├── LICENSE
├── MANIFEST.in
├── setup.py
└── src/
├── open_clip/
│ ├── __init__.py
│ ├── big_vision.py
│ ├── coca_model.py
│ ├── constants.py
│ ├── factory.py
│ ├── hf_configs.py
│ ├── hf_model.py
│ ├── loss.py
│ ├── model.py
│ ├── model_configs/
│ │ ├── EVA01-g-14-plus.json
│ │ ├── EVA01-g-14.json
│ │ ├── EVA02-B-16.json
│ │ ├── EVA02-E-14-plus.json
│ │ ├── EVA02-E-14.json
│ │ ├── EVA02-L-14-336.json
│ │ ├── EVA02-L-14.json
│ │ ├── RN101-quickgelu.json
│ │ ├── RN101.json
│ │ ├── RN50-quickgelu.json
│ │ ├── RN50.json
│ │ ├── RN50x16.json
│ │ ├── RN50x4.json
│ │ ├── RN50x64.json
│ │ ├── ViT-B-16-SigLIP-256.json
│ │ ├── ViT-B-16-SigLIP-384.json
│ │ ├── ViT-B-16-SigLIP-512.json
│ │ ├── ViT-B-16-SigLIP-i18n-256.json
│ │ ├── ViT-B-16-SigLIP.json
│ │ ├── ViT-B-16-plus-240.json
│ │ ├── ViT-B-16-plus.json
│ │ ├── ViT-B-16-quickgelu.json
│ │ ├── ViT-B-16.json
│ │ ├── ViT-B-32-256.json
│ │ ├── ViT-B-32-plus-256.json
│ │ ├── ViT-B-32-quickgelu.json
│ │ ├── ViT-B-32.json
│ │ ├── ViT-H-14-378-quickgelu.json
│ │ ├── ViT-H-14-CLIPA-336.json
│ │ ├── ViT-H-14-CLIPA.json
│ │ ├── ViT-H-14-quickgelu.json
│ │ ├── ViT-H-14.json
│ │ ├── ViT-H-16.json
│ │ ├── ViT-L-14-280.json
│ │ ├── ViT-L-14-336.json
│ │ ├── ViT-L-14-CLIPA-336.json
│ │ ├── ViT-L-14-CLIPA.json
│ │ ├── ViT-L-14-quickgelu.json
│ │ ├── ViT-L-14.json
│ │ ├── ViT-L-16-320.json
│ │ ├── ViT-L-16-SigLIP-256.json
│ │ ├── ViT-L-16-SigLIP-384.json
│ │ ├── ViT-L-16.json
│ │ ├── ViT-M-16-alt.json
│ │ ├── ViT-M-16.json
│ │ ├── ViT-M-32-alt.json
│ │ ├── ViT-M-32.json
│ │ ├── ViT-S-16-alt.json
│ │ ├── ViT-S-16.json
│ │ ├── ViT-S-32-alt.json
│ │ ├── ViT-S-32.json
│ │ ├── ViT-SO400M-14-SigLIP-384.json
│ │ ├── ViT-SO400M-14-SigLIP.json
│ │ ├── ViT-bigG-14-CLIPA-336.json
│ │ ├── ViT-bigG-14-CLIPA.json
│ │ ├── ViT-bigG-14.json
│ │ ├── ViT-e-14.json
│ │ ├── ViT-g-14.json
│ │ ├── coca_ViT-B-32.json
│ │ ├── coca_ViT-L-14.json
│ │ ├── coca_base.json
│ │ ├── coca_roberta-ViT-B-32.json
│ │ ├── convnext_base.json
│ │ ├── convnext_base_w.json
│ │ ├── convnext_base_w_320.json
│ │ ├── convnext_large.json
│ │ ├── convnext_large_d.json
│ │ ├── convnext_large_d_320.json
│ │ ├── convnext_small.json
│ │ ├── convnext_tiny.json
│ │ ├── convnext_xlarge.json
│ │ ├── convnext_xxlarge.json
│ │ ├── convnext_xxlarge_320.json
│ │ ├── mt5-base-ViT-B-32.json
│ │ ├── mt5-xl-ViT-H-14.json
│ │ ├── nllb-clip-base-siglip.json
│ │ ├── nllb-clip-base.json
│ │ ├── nllb-clip-large-siglip.json
│ │ ├── nllb-clip-large.json
│ │ ├── roberta-ViT-B-32.json
│ │ ├── swin_base_patch4_window7_224.json
│ │ ├── vit_medium_patch16_gap_256.json
│ │ ├── vit_relpos_medium_patch16_cls_224.json
│ │ ├── xlm-roberta-base-ViT-B-32.json
│ │ └── xlm-roberta-large-ViT-H-14.json
│ ├── modified_resnet.py
│ ├── openai.py
│ ├── pos_embed.py
│ ├── pretrained.py
│ ├── push_to_hf_hub.py
│ ├── timm_model.py
│ ├── tokenizer.py
│ ├── transform.py
│ ├── transformer.py
│ ├── utils.py
│ ├── version.py
│ ├── zero_shot_classifier.py
│ └── zero_shot_metadata.py
├── retrieve_clip_VQA.py
├── retrieve_clip_report.py
└── training/
├── .gitignore
├── __init__.py
├── data.py
├── distributed.py
├── file_utils.py
├── logger.py
├── main.py
├── main_feature-work.py
├── main_retrieve_report.py
├── main_retrieve_report_harvard.py
├── params.py
├── precision.py
├── profiler.py
├── scheduler.py
├── train.py
└── zero_shot.py
================================================
FILE CONTENTS
================================================
================================================
FILE: .gitignore
================================================
backup
train/open_clip/.github
train/open_clip/.gitignore
train/open_clip/docs
train/open_clip/scripts
train/dpo/playground
# train/dpo/scripts
train/dpo/docs
train/dpo/wandb
train/open_clip/README.md
================================================
FILE: LICENSE
================================================
MIT License
Copyright (c) 2024 Peng "Richard" Xia
Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in all
copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.
================================================
FILE: README.md
================================================
# MMed-RAG: Versatile Multimodal RAG System for Medical Vision Language Models
We introduce MMed-RAG, a powerful multimodal RAG system that boosts the factuality of Medical Vision-Language Models (Med-LVLMs) by up to 43.8%! 🩺 [[Paper](https://arxiv.org/abs/2410.13085)] [[X(Twitter)](https://x.com/HuaxiuYaoML/status/1847097594641584574)]
## 🚀 News
- [01/22/2025] MMed-RAG is accepted by ICLR 2025!
- [12/08/2024] The training scripts are released!
- [10/20/2024] The whole data is released in `data/`! Check it out!
- [10/18/2024] The manuscript can be found on [arXiv](https://arxiv.org/abs/2410.13085).
## 💡 Overview
MMed-RAG enhances alignment across medical domains like radiology, pathology, and ophthalmology with a domain-aware retrieval mechanism. And it tackles three key challenges in alignment of multimodal RAG:
1️⃣ Direct Copy Homework from Others❌ Think it by Self ✅
MMed-RAG helps Med-LVLMs avoid blindly copying external information by encouraging the model to rely on its own visual reasoning when solving complex problems.
2️⃣ Cannot Solve Problems by Self❌ Learn How to Copy ✅
When Med-LVLMs are unsure, MMed-RAG teaches the model to intelligently use retrieved knowledge, pulling in the right information at the right time, boosting accuracy, and reducing errors.
3️⃣ Copied Homework is Wrong❌ Avoid Interference from Incorrect Homework ✅
MMed-RAG prevents models from being misled by incorrect retrievals, reducing the risk of generating inaccurate medical diagnoses.
<div align=left>
<img src=asset/logo.png width=90% />
</div>
## 📦 Requirements
1. Clone this repository and navigate to MMed-RAG folder
```bash
git clone https://github.com/richard-peng-xia/MMed-RAG.git
cd MMed-RAG
```
2. Install Package: Create conda environment
```Shell
conda create -n MMed-RAG python=3.10 -y
conda activate MMed-RAG
cd MMed-RAG
pip install --upgrade pip # enable PEP 660 support
pip install -r requirements.txt
pip install trl
```
3. Download the required model checkpoints [LLaVA-Med-1.5](https://huggingface.co/microsoft/llava-med-v1.5-mistral-7b) from huggingface.
4. For all the medical datasets, you need firstly apply for the right of access and then download the dataset.
- [MIMIC-CXR](https://physionet.org/content/mimic-cxr-jpg/2.0.0/)
- [IU-Xray](https://drive.google.com/file/d/1c0BXEuDy8Cmm2jfN0YYGkQxFZd2ZIoLg/view) (Thanks to [R2GenGPT](https://github.com/wang-zhanyu/R2GenGPT) for sharing the file)
- [Harvard-FairVLMed](https://ophai.hms.harvard.edu/datasets/harvard-fairvlmed10k/)
- [PMC-OA](https://huggingface.co/datasets/axiong/pmc_oa)
- [Quilt-1M](https://github.com/wisdomikezogwo/quilt1m)
## 📖 Data Description
We provide a corresponding json or jsonl file for each dataset, including the image path, question, answer, and original report.
- Training: The data used to train the retriever and fine-tune the Med-LVLM are located in `data/training/retriever/MODALITY` and `data/training/alignment/MODALITY` respectively. Each folder contains data for VQA or report generation tasks.
- Test: All the test data for Med-LVLMs is placed under `data/test/TASK/MODALITY`.
`TASK`: report/vqa, `MODALITY`: radiology/pathology/ophthalmology.
## 🏋️ Train
### Retriver Fine-tuning
Use the following script, make sure to specify the data paths and the checkpoint saving location.
```
bash ./scripts/finetune_clip.sh
```
### Preference Fine-tuning
Use the script `train_dpo_2stages.sh` in `./script` or the following command, make sure to specify the necessary data paths and the checkpoint saving location.
```
deepspeed --include localhost:0,1,2,3 ./train/dpo/train_dpo_2stages.py \
--model_name_or_path /path/to/llava-med_model_checkpoint \
--deepspeed ./scripts/zero3.json \
--version v1 \
--lora_enable True --lora_r 128 --lora_alpha 256 --mm_projector_lr 2e-5 \
--data_path /path/to/data_json \
--image_folder /path/to/img_folder \
--vision_tower openai/clip-vit-large-patch14-336 \
--mm_projector_type mlp2x_gelu \
--mm_vision_select_layer -2 \
--mm_use_im_start_end False \
--mm_use_im_patch_token False \
--image_aspect_ratio pad \
--group_by_modality_length True \
--bf16 True \
--output_dir /path/to/output_checkpoint_saving_location \
--num_train_epochs 3 \
--per_device_train_batch_size 1\
--per_device_eval_batch_size 1 \
--gradient_accumulation_steps 1 \
--evaluation_strategy "no" \
--save_strategy "steps" \
--save_steps 200 \
--save_total_limit 1 \
--learning_rate 1e-7 \
--weight_decay 0. \
--warmup_ratio 0.03 \
--lr_scheduler_type "cosine" \
--logging_steps 1 \
--report_to wandb \
--tf32 True \
--model_max_length 1024 \
--gradient_checkpointing True \
--dataloader_num_workers 4 \
--lazy_preprocess True \
```
## 🥖 Retrieve
Use `retrieve_clip_report.sh` or `retrieve_clip_VQA.sh` to retrieve reports for report generation or VQA task. The script uses Harvard-FairVLMed dataset as an example. Make sure to specify the necessary data paths and the saving location.
<!-- ### Preference Fine-tuning -->
<!-- ### Test -->
## 📅 Schedule
- [x] Release the data (VQA and report generation tasks)
- [x] Release the training code
## 📚Citation
```bibtex
@article{xia2024mmedrag,
title={MMed-RAG: Versatile Multimodal RAG System for Medical Vision Language Models},
author={Xia, Peng and Zhu, Kangyu and Li, Haoran and Wang, Tianze and Shi, Weijia and Wang, Sheng and Zhang, Linjun and Zou, James and Yao, Huaxiu},
journal={arXiv preprint arXiv:2410.13085},
year={2024}
}
```
## 🙏Acknowledgement
We use code from [LLaVA-Med](https://github.com/microsoft/LLaVA-Med), [RULE](https://github.com/richard-peng-xia/RULE), [CARES](https://github.com/richard-peng-xia/CARES). We thank the authors for releasing their code.
<!--
## Clip Finetune
```
bash ./scripts/retrieve_clip_VQA.sh
```
## DPO training
```
bash ./scripts/train_dpo_2stages_VQA.sh
```
## Inference
```
```
-->
================================================
FILE: data/test/report/harvard_test.json
================================================
[
{
"id": "data_08003",
"image_path": "slo_fundus_08003.jpg",
"filename": "data_08003.npz",
"report": "86-year-old white, non-hispanic female diagnosed with glaucoma.",
"age": 86.74,
"gender": "female",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "widowed",
"note": "a 86 y.o. white, non-hispanic female was evaluated and diagnosed with glaucoma.",
"gpt4_summary": "86-year-old white, non-hispanic female diagnosed with glaucoma.",
"glaucoma": "yes",
"use": "test"
},
{
"id": "data_08008",
"image_path": "slo_fundus_08008.jpg",
"filename": "data_08008.npz",
"report": "The clinical note prescribes erythromycin ophthalmic ointment for the patient's right eye. The patient is referred to ophthalmology. No mention of glaucoma. Other conditions include goiter, rheumatoid arthritis, etc.",
"age": 39.36,
"gender": "female",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "erythromycin (romycin) ophthalmic ointment place 0.5 inches into the right eye 4 (four) times a day. your orders normal orders this visit ambulatory referral to Institution ophthalmology humphrey visual field - ou - both eyes oct, optic nerve - ou - both eyes - condition list as of DATE_TIME goiter rheumatoid arthritis hemorrhoids fatigue vitamin d deficiency antinuclear factor positive fibromyalgia carpal tunnel syndrome thrombophlebitis disorder of thyroid premenopausal menorrhagia irregular menses contraception management results summary immunizations administered on date of encounter - DATE_TIME none",
"gpt4_summary": "The clinical note prescribes erythromycin ophthalmic ointment for the patient's right eye. The patient is referred to ophthalmology. No mention of glaucoma. Other conditions include goiter, rheumatoid arthritis, etc.",
"glaucoma": "yes",
"use": "test"
},
{
"id": "data_08009",
"image_path": "slo_fundus_08009.jpg",
"filename": "data_08009.npz",
"report": "Patient displays thin temporal os and hvf 24-2 shows corresponding inferior defects. Risk factors for glaucoma noted. Patient also has borderline diabetes, chronic anemia and a family history of macular degeneration.",
"age": 67.09,
"gender": "male",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "question ltg? risks include c/d asym iop is wnl, PERSON is -2, -3. oct rnfl shows thin superior ou, thin temporal os hvf 24-2 shows corresponding inferior defects os only, PERSON denies migraine required transfution ?chronic anemia from bleeding denies history of trauma, PERSON no apd patient reports pcp is following for borderline dmii hemoglobin a1c date value ref range status DATE_TIME 5.7 (h) 4.2 - 5.6 % final comment: hba1c levels 5.7-6.4% represent pre-diabetes, indicating impaired glucose control and an increased risk of developing diabetes. the diagnostic hba1c level for diabetes is 6.5% or greater. hba1c levels <4.2% may indicate a hemoglobinopathy or anemia, and an alternative method is recommended to monitor glucose control. mild combined forms of age related cataract ou continue to observe. fmhx of macular degeneration mother progressed to blindness before passed away mom was a smoker pvd od no holes/tears/breaks seen on exam. warned pt of rd symptoms. pt will rtc if they experience new/worsening symptoms or rd/rt symptoms. hyperopia ou w/ astigmatism od and presbyopia ou otc sheet given refer to glaucoma service in DATE_TIME by signing my name below, i, PERSON, acting as a scribe, attest that this documentation has been prepared under the direction and in the presence of PERSON, LOCATION DATE_TIME by signing my name below, i, PERSON, attest that this documentation has been prepared under my direction.",
"gpt4_summary": "Patient displays thin temporal os and hvf 24-2 shows corresponding inferior defects. Risk factors for glaucoma noted. Patient also has borderline diabetes, chronic anemia and a family history of macular degeneration.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08010",
"image_path": "slo_fundus_08010.jpg",
"filename": "data_08010.npz",
"report": "The 44-year-old patient has dry eye, pinguecula, and optic disc cupping. There are nonspecific defects in their visual field, and thinning in the superior and temporal sectors of the eye. No signs of glaucoma.",
"age": 44.72,
"gender": "female",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "44 y.o. with PERSON / dry eye > warm compresses, lid hygiene, artificial tears > flax seed / fish oil supplementation pinguecula ou > artificial tears optic disc cupping ou - no family hx - photos DATE_TIME - cct 531/527 - hvf DATE_TIME: reliable ou, nonspecific defects, essentially full ou DATE_TIME nonspecific defects ou DATE_TIME od borderline reliable, essentially full os borderline reliable inferior and paracentral defects - oct rnfl DATE_TIME: reliable ou, PERSON, os with superior thinning, borderline temporal thinning DATE_TIME od wnl os with superior thinning, borderline temporal thinning DATE_TIME 83/74 od wnl os with superior thinning, borderline temporal thinning PERSON today 10/10 > monitor for now, iop good off medication vitreous syneresis > no retinal breaks/ detachment fu DATE_TIME, mrx, dilate, hvf/ oct",
"gpt4_summary": "The 44-year-old patient has dry eye, pinguecula, and optic disc cupping. There are nonspecific defects in their visual field, and thinning in the superior and temporal sectors of the eye. No signs of glaucoma.",
"glaucoma": "yes",
"use": "test"
},
{
"id": "data_08013",
"image_path": "slo_fundus_08013.jpg",
"filename": "data_08013.npz",
"report": "Patient unhappy with post-surgery visual improvement due to severe glaucoma. Misunderstood pre-surgery conversations. Sought second opinion but decided to continue with current provider. Upcoming appointments scheduled for eyelid, cornea and low-vision care.",
"age": 64.7,
"gender": "male",
"race": "asian",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "single",
"note": "patient is disappointed with result from surgery because vision 'didn't improve,' but i explained vision is limited by his severe glaucoma. i never promised/guaranteed outcomes to the patient, but he seems to have misunderstood our initial conversations before surgery. he also said he searched for second opinion in LOCATION, LOCATION, which i applauded. doctor there said he would 'adjust' something, but it's unclear to me what that means. ultimately i said it was ok to change providers (he had researched dr. PERSONon), but he said he'd give me 'one more try.' -scheduled to see dr. PERSON to establish eyelid care DATE_TIME. -follow-up with dr. PERSON for cornea care => will try to coordinate appointment as of DATE_TIME. -follow-up with dr. PERSON for mrx and low-vision care. -preservative-free artificial tears as needed. -rtc in DATE_TIME with iop check ou, hvf 10-2 od and hvf 24-2 os, sooner prn. i saw and evaluated this patient and discussed the case as appropriate with the resident/fellow (PERSON, PERSON. i have reviewed the resident/fellow's notes and made any necessary changes. the information above was documented by PERSON as a scribe for PERSON on DATE_TIME. i was present while this encounter was recorded and agree that the information entered by my scribe is complete and accurate.",
"gpt4_summary": "Patient unhappy with post-surgery visual improvement due to severe glaucoma. Misunderstood pre-surgery conversations. Sought second opinion but decided to continue with current provider. Upcoming appointments scheduled for eyelid, cornea and low-vision care.",
"glaucoma": "yes",
"use": "test"
},
{
"id": "data_08016",
"image_path": "slo_fundus_08016.jpg",
"filename": "data_08016.npz",
"report": "The patient, a 65-year-old female, is a suspect for open-angle glaucoma, presenting no family history. She also has sulfa allergy, a history of trauma, and early cataracts. For now, no glaucoma treatment is planned.",
"age": 65.26,
"gender": "female",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "single",
"note": "65 yo wf self referral, was seen prior by dr. PERSON poor. followed for DATE_TIME as glaucoma suspect. no fh, + sulfa allergy, no asthma, + trauma (as a child falls, soft ball injury (teeth), had car accident priror) 1. open angle glaucoma suspect ou - cct 539 ou - gonio ou: open ou, slightly narrow, but no iop change after dilation. - t max high teens per records. DATE_TIME mid teens without rx. - hvf full ou and oct with stable vertical thinning od and full os (stable per reviewing prior records) - long discussion was done about the nature of the disease, visual prognosis and management options at this stage, patient voiced understanding, all questions were answered. considering stable changes od on LOCATION/o multiple trauma prior, possible traumatic neuropathy od, will observe for now without glaucoma rx. r/b of observation vs prophylactic tx od discussed. rtc: DATE_TIME for iop, gonio, hvf ou, dialate ou, oct ou plan: observe without rx 2. early cataracts ou - nvs observe 3. myopia ou - ok with glasses, observe - dfe DATE_TIME - no holes or tears.",
"gpt4_summary": "The patient, a 65-year-old female, is a suspect for open-angle glaucoma, presenting no family history. She also has sulfa allergy, a history of trauma, and early cataracts. For now, no glaucoma treatment is planned.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08019",
"image_path": "slo_fundus_08019.jpg",
"filename": "data_08019.npz",
"report": "Patient is on brimonidine/alphagan for the left eye 2x/day & prednisolone for the right eye 1x/day. Advised not to rub operated eye for 4 weeks. Urged to contact the glaucoma department for any concerns.",
"age": 67.77,
"gender": "female",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "medication route frequency LOCATION (teal) the left eye DATE_TIME brimonidine/alphagan3 (purple) the left eye 2x/day prednisolone 1 (white or pink) the right eye 1x/day \u00f8 some medications that may be prescribed in lieu of this medication include: latanoprost, xalatan, travatan z, travaprost, LOCATION, LOCATION, LOCATION/tafluprost (preservative-free). 1 some medications that may be prescribed in lieu of this medication (topical steroid) include: prednisolone acetate, pred forte, durezol, LOCATION, LOCATION, LOCATION, loteprednol, fluorometholone. prednisolone should be shaken 20 times before placing it in the eye. 3 some medications that may be prescribed in lieu of this medication include: brimonidine, alphagan, PERSON avoid rubbing the operated eye for at least 4 full weeks. keep shield over operated eye at bedtime until instructed by medical doctor (usually DATE_TIME). no bending, straining or lifting anything heavier than 5 pounds until the doctor indicates it. always wait DATE_TIME (by the clock) in between eye drops. if you have eye pain, vision changes or other concerns, do not hesitate to contact the glaucoma department. for routine questions, please call PHONE_NUMBER (LOCATION, ask for glaucoma department) or PHONE_NUMBER (longwood). for emergencies, please call PHONE_NUMBER and ask for the glaucoma physician on-call. please bring this to your next visit.",
"gpt4_summary": "Patient is on brimonidine/alphagan for the left eye 2x/day & prednisolone for the right eye 1x/day. Advised not to rub operated eye for 4 weeks. Urged to contact the glaucoma department for any concerns.",
"glaucoma": "yes",
"use": "test"
},
{
"id": "data_08021",
"image_path": "slo_fundus_08021.jpg",
"filename": "data_08021.npz",
"report": "The patient, formerly of Pasquale, has primary open angle glaucoma and amblyopia in the left eye. Other issues include lattice in both eyes and a history of cataract extraction in both eyes.",
"age": 68.47,
"gender": "male",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "first seen by PERSON PERSON on DATE_TIME, former pasquale patient diagnosis:primary open angle glaucoma, amblyopia left eye (et) target iop: / , tmax: ( ) / ( ) 25 central corneal thickness: / 540 gonioscopy: refractive error: od . . / os . . optic nerve findings on initial visit right eye: mostly normal optic nerve findings on initial visit left eye: thin nfl visual fields on initial visit right eye: non-specific visual fields on initial visit left eye: superior arcuate medication history and intolerances: initially latanoprost glaucoma procedures right eye: glaucoma procedures left eye: other eye procedures right eye: cataract extraction other eye procedures left eye: cataract extraction other eye problems right eye: lattice other eye problems left eye: amblyopia, lattice family history: mother (94), steroids: no trauma: no asthma: no other medical history and problems: plan: \u00ff appears stable, check in DATE_TIME, vf and dilate",
"gpt4_summary": "The patient, formerly of Pasquale, has primary open angle glaucoma and amblyopia in the left eye. Other issues include lattice in both eyes and a history of cataract extraction in both eyes.",
"glaucoma": "yes",
"use": "test"
},
{
"id": "data_08022",
"image_path": "slo_fundus_08022.jpg",
"filename": "data_08022.npz",
"report": "Patient visited for a second opinion and was diagnosed with low tension glaucoma. Initial visit showed no thinning in the right eye, but focal superior thinning in left eye. Plans to start new medication regimen including timoptic, brimonidine and continuing latanoprost. ",
"age": 55.12,
"gender": "female",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "first seen by dr. LOCATION on DATE_TIME self referred for second opinion diagnosis: low tension glaucoma dx DATE_TIME target iop: DATE_TIME, tmax: 19 ( ) / 20 ( ) set target empirically, limited data goal may not need to be this low, especially od central corneal thickness: 533.533.533 / 531.531.531 gonioscopy: refractive error: od . . / os . . optic nerve findings on initial visit right eye: no definite thinning, although trend on oct rnfl optic nerve findings on initial visit left eye: focal superior thinning. visual fields on initial visit right eye: full visual fields on initial visit left eye: inferior nasal step, LOCATION medication history and intolerances: PERSON bid ou - pt reports too expensive latanoprost qhs ou glaucoma procedures right eye : none glaucoma procedures left eye : none other eye procedures: none family history: no glaucoma niece with retinal PERSON steroids: no trauma/eye surgery: none plan: stop combigan start timoptic start brimonidine continue latanoprost DATE_TIME dfe if iop still above goal next visit consider slt laser vs adding dorzolamide.",
"gpt4_summary": "Patient visited for a second opinion and was diagnosed with low tension glaucoma. Initial visit showed no thinning in the right eye, but focal superior thinning in left eye. Plans to start new medication regimen including timoptic, brimonidine and continuing latanoprost. ",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08024",
"image_path": "slo_fundus_08024.jpg",
"filename": "data_08024.npz",
"report": "61 y.o. white, non-hispanic male with no diagnosis of glaucoma. Requested to mail clinical note to specific person and location.",
"age": 61.93,
"gender": "male",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "a 61 y.o. white, non-hispanic male with no diagnosis of glaucoma. can you please mail a copy of my note to PERSON PERSON at LOCATION? thank you!",
"gpt4_summary": "61 y.o. white, non-hispanic male with no diagnosis of glaucoma. Requested to mail clinical note to specific person and location.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08032",
"image_path": "slo_fundus_08032.jpg",
"filename": "data_08032.npz",
"report": "61 y.o. male suspect for glaucoma due to family history, ocular hypertension, and high intraocular pressure (IOP: 29/32). His father possibly had glaucoma. Patient also has incipient senile cataract, myopia, astigmatism, and presbyopia.",
"age": 61.65,
"gender": "male",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "61 y.o. male presents for a comprehensive eye exam. he was last seen by me DATE_TIME. PERSON: ocular hypertension, father with possible glaucoma pmhx: htn since the last visit he notes in the morning he sees a white ring around his peripheral vision that then clears up. difficulty driving with current rx for distance. assessment/plan: # glaucoma suspect/ohtn - glaucoma suspect based on family history and iop , c/d - iop - 29/32 on intake, rechecked to be 23/26 - tmax - unknown, mid to high 20s per patient - cct - 590/596 thick ou - c/d - 0.35. 0.35 - gonio - open to cbb ou - family history - possibly father later in life - last hvf performed - not done yet - last oct rnfl performed - DATE_TIME - full ou, increased c/d but healthy rnfl - fundus photos obtained - DATE_TIME - dilated optic nerve exam performed ou DATE_TIME - discussed potential for progressive irreversible vision loss due to glaucomatous optic neuropathy due to high eye pressure and results of ohts trial - return in DATE_TIME with hvf 24-2, oct rnfl, gonio before dilation # incipient senile cataract # myopia with astigmatism and presbyopia - va 20/20 ou with current rx - can continue current rx - small change in rx without improvement in visual acuity rtc DATE_TIME or sooner prn with hvf 24-2, oct rnfl - no dilation PERSON, md",
"gpt4_summary": "61 y.o. male suspect for glaucoma due to family history, ocular hypertension, and high intraocular pressure (IOP: 29/32). His father possibly had glaucoma. Patient also has incipient senile cataract, myopia, astigmatism, and presbyopia.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08035",
"image_path": "slo_fundus_08035.jpg",
"filename": "data_08035.npz",
"report": "76yo patient has depression, htn, breast cancer, visually significant cataract, pseudophakia os, and glaucoma (recently diagnosed, on latanoprost). No family history of glaucoma. IOP is OK but concern about hvf defect. Also has allergic conjunctivitis, and conjunctival pigment.",
"age": 76.43,
"gender": "female",
"race": "black",
"ethnicity": "non-hispanic",
"language": "unknown",
"maritalstatus": "single",
"note": "76 yo patient pmh depression, htn, breast ca last seen by me DATE_TIME \u00ff cataract od - starting to become visually significant; pt denies any difficulty with vision and no significant anisometropia - observe for now \u00ff pseudophakia os s/p ce/iol DATE_TIME at LOCATION - doing well \u00ff glaucoma - recently diagnosed at LOCATION and started on latanoprost ou qhs - iop DATE_TIME, 13 previously 15 - no family history of glaucoma - gonio DATE_TIME open to ss 4 quadrants ou - cct 500, 495- iop likely higher than measured - rnfl DATE_TIME:poor signal strength; ? inferior thinning ou - rnfl DATE_TIME: poor quality nerve and gca - hvf 9/19: os scattered deficits; od dense PERSON tracking turned off but per technicians poor ability to cooperate and so may be unreliable - hvf DATE_TIME: od 3/12 fl, dense superonasal quandrantic defect that persists from last scan os: scattered inferior and nasal defects; no evidence of homonymous defect - disc photos DATE_TIME: c/d 0.6 ou, full LOCATION, os with some superior thinning; normal vessels and visible macula - dfe DATE_TIME wnl - plan: refer to glaucoma. i asked her to bring PERSON records to visit (i do not have these or know the history of her glaucoma diagnosis.) iop is ok, but i am concerned about the persistent hvf defect. of note, the defect is dense and quadrantic, respecting horizontal and vertical meridiens, but does not appear to be homonymous \u00ff allergic conjunctivitis - continue to use zatditor prn \u00ff conjunctival pigment - observe in the setting of xalatan \u00ff patient to followup with urgent changes as needed; otherwise see glaucoma next available and will return in 6 m with me _____________________ \u00ff PERSON, md, mph comprehensive ophthalmology LOCATION",
"gpt4_summary": "76yo patient has depression, htn, breast cancer, visually significant cataract, pseudophakia os, and glaucoma (recently diagnosed, on latanoprost). No family history of glaucoma. IOP is OK but concern about hvf defect. Also has allergic conjunctivitis, and conjunctival pigment.",
"glaucoma": "yes",
"use": "test"
},
{
"id": "data_08040",
"image_path": "slo_fundus_08040.jpg",
"filename": "data_08040.npz",
"report": "The note discusses a 62-year-old female with blepharitis, who experiences significant improvement with ocusoft wipes. She wears multifocal contact lenses and is being referred to optometry. She is suspected of having glaucoma due to an increased c/d ratio, despite normal intraocular pressure. Other factors like family history or steroid medication use are ruled out. Eye scans reveal borderline changes and thinning superiorly. Future visits are scheduled for ongoing evaluation.",
"age": 62.49,
"gender": "female",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "single",
"note": "62f # blepharitis -uses ocusoft wipes with great improvement >PERSON >warm compresses >omega 3 fa >at prn # refractive: wears multifocal contact lenses, and progressives. very happy >refer to optometry for cl eval # glaucoma suspect: increased c/d ratio -normal iop, no history of steroid medications, no fhx -pachy 530/530 -disc photos DATE_TIME -oct DATE_TIME: od borderline changes; os thinning superiorly -hvf DATE_TIME: full ou >follow DATE_TIME for now rtc DATE_TIME: coe, hvf, oct disc",
"gpt4_summary": "The note discusses a 62-year-old female with blepharitis, who experiences significant improvement with ocusoft wipes. She wears multifocal contact lenses and is being referred to optometry. She is suspected of having glaucoma due to an increased c/d ratio, despite normal intraocular pressure. Other factors like family history or steroid medication use are ruled out. Eye scans reveal borderline changes and thinning superiorly. Future visits are scheduled for ongoing evaluation.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08047",
"image_path": "slo_fundus_08047.jpg",
"filename": "data_08047.npz",
"report": "Patient shows signs of glaucoma. Hx of amblyopia in left eye. No HIV retinopathy. Marked cupping OD>OS. Intraocular pressure highly elevated. RNFL thinning OD. Normal HVF.",
"age": 40.65,
"gender": "male",
"race": "white",
"ethnicity": "hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "imp: hx ambly os (20/70 bcva) no hiv retinopathy cupping od>os, vertical notch od. iop highly elevated DATE_TIME. open on gonio. rnfl with progressive thinning od, hvf normal -cct 560s ou previously -administered cosopt and brimonidine x1 ou in clinic small ch nevus od refr error plan: cosopt bid ou referral to glaucoma rx=m yrly cos",
"gpt4_summary": "Patient shows signs of glaucoma. Hx of amblyopia in left eye. No HIV retinopathy. Marked cupping OD>OS. Intraocular pressure highly elevated. RNFL thinning OD. Normal HVF.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08049",
"image_path": "slo_fundus_08049.jpg",
"filename": "data_08049.npz",
"report": "This clinical note does not provide any medical information or mention the presence of glaucoma. It primarily guides on account activation in a newly developed portal.",
"age": 62.46,
"gender": "female",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "single",
"note": "user account. important information about your account: \u0007 if you already have a account, please sign in with your existing username and your account will automatically be activated in our new portal - \u0007",
"gpt4_summary": "This clinical note does not provide any medical information or mention the presence of glaucoma. It primarily guides on account activation in a newly developed portal.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08050",
"image_path": "slo_fundus_08050.jpg",
"filename": "data_08050.npz",
"report": "Patient has presbyopia and glaucoma. Undergoing treatment to control intraocular pressure (IOP). Recent procedures include phaco/xen gel stent. Plans to possibly continue with phaco/xen procedure on next visit.",
"age": 71.81,
"gender": "female",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "divorced",
"note": "and presbyopia, both eyes -stable. 6. social/systemic issues: patient referred to clinic after ed visit on DATE_TIME. i requested the patient to transfer any relevant medical records concerning her glaucoma. attending's plan: -goal intraocular pressure less than or equal to 17 mmhg, right eye. -goal intraocular pressure less than or equal to 14 mmhg, left eye. -iop at goal od and at goal os on DATE_TIME on timolol bid od, LOCATION bid od and s/p phaco/xen gel stent os. -decrease LOCATION to qhs ou (from bid od) to control iop od and keep iop in the very low teens os. -continue ocudose bid od. -instructions written/typed/printed out for patient (see table/details under patient instructions). -emphasized adherence to medication regimen. -preservative-free artificial tears as needed. -retinal detachment precautions were reviewed with the patient. -long discussion with patient on DATE_TIME: given iop above goal os and numerous allergies os, we proceeded with phaco/open xen gel stent os with 3-piece in the bag (li) due to pxf on DATE_TIME. -rtc in DATE_TIME with iop check ou, bat od, optical biometry ou, and oct rnfl/gcc ou, sooner prn. strongly consider phaco/xen od next visit. the information above was documented by PERSON as a scribe for PERSON on DATE_TIME.",
"gpt4_summary": "Patient has presbyopia and glaucoma. Undergoing treatment to control intraocular pressure (IOP). Recent procedures include phaco/xen gel stent. Plans to possibly continue with phaco/xen procedure on next visit.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08054",
"image_path": "slo_fundus_08054.jpg",
"filename": "data_08054.npz",
"report": "Patient, 68, with GERD, hypertension, intraocular pressure is 16/16, previously was 18/17, 26/23, max of 26/23. Patient has ocular hypertension. C/D ratio is 0.5/0.5, open on gonioscopy. No mention of glaucoma.",
"age": 68.25,
"gender": "female",
"race": "black",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "divorced",
"note": "68 y.o. with gerd, hypertension ocular hypertension ou - tcurrent: 16/16 - t previous: 26/23, 18/17 - tmax: 26/23 - tgoal: - c/d ratio: 0.5/0.5 - gonioscopy: open ou - central corneal thickness: DATE_TIME: DATE_TIME wnl ou DATE_TIME wnl ou DATE_TIME wnl ou DATE_TIME 84/87 wnl both eyes DATE_TIME 86/88 wnl ou - hvf: DATE_TIME full ou DATE_TIME full ou DATE_TIME full ou DATE_TIME full both eyes DATE_TIME od full os high fp, nonspecific defects - family history: +mother - race: aa - optic nerve photos >? iop better/stable on latanoprost ou, continue immature cataract ou - not visually significant > observe dry eye - has occ burning, itching. better on artificial tears refractive error - no change fu DATE_TIME mrx, dilate, hvf/ oct",
"gpt4_summary": "Patient, 68, with GERD, hypertension, intraocular pressure is 16/16, previously was 18/17, 26/23, max of 26/23. Patient has ocular hypertension. C/D ratio is 0.5/0.5, open on gonioscopy. No mention of glaucoma.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08055",
"image_path": "slo_fundus_08055.jpg",
"filename": "data_08055.npz",
"report": "The patient has glaucoma and is considering different types of surgery. Due to the severe irritation caused by current medication, they proceeded with phaco/trab mmc os. Follow-up planned for IOP check.",
"age": 75.92,
"gender": "female",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "only. -continue e-mycin qhs os as needed for comfort. -instructions written/typed/printed out for patient (see table/details under patient instructions) previously.*** -emphasized adherence to medication regimen. -preservative-free artificial tears as needed. -long discussion with patient on DATE_TIME: patient has given a lot of thought to the different types of glaucoma surgery (and whether to have it or not given stable hvf), and given the need of 5 agents to control iop os (some of which are causing severe irritation), she proceeded with phaco/trab mmc os on DATE_TIME. -rtc in DATE_TIME with iop check ou, hvf ou, dilation ou, and oct rnfl/gcc os only by experienced technician (please try to obtain 9/10 signal strength), sooner prn. DATE_TIME iop disc the information above was documented by PERSON as a scribe for PERSON on DATE_TIME.",
"gpt4_summary": "The patient has glaucoma and is considering different types of surgery. Due to the severe irritation caused by current medication, they proceeded with phaco/trab mmc os. Follow-up planned for IOP check.",
"glaucoma": "yes",
"use": "test"
},
{
"id": "data_08056",
"image_path": "slo_fundus_08056.jpg",
"filename": "data_08056.npz",
"report": "Visual field test is to be performed on the left eye with lid taping, following monocular precautions. No mention of glaucoma.",
"age": 69.99,
"gender": "female",
"race": "asian",
"ethnicity": "non-hispanic",
"language": "unknown",
"maritalstatus": "married or partnered",
"note": "DATE_TIME, visual field to be performed on the left eye only with lid taping 2. monocular precautions 1. follow up in neuro-op in DATE_TIME, visual field to be performed on the left eye only with lid taping 2. monocular precautions this note was prepared.",
"gpt4_summary": "Visual field test is to be performed on the left eye with lid taping, following monocular precautions. No mention of glaucoma.",
"glaucoma": "yes",
"use": "test"
},
{
"id": "data_08057",
"image_path": "slo_fundus_08057.jpg",
"filename": "data_08057.npz",
"report": "Patient has suspected myelinated nerve fiber layer od, near circumferential on exam. Experiencing enlarging blind spot; advised to bring records to next appointment. No mention of glaucoma.",
"age": 61.89,
"gender": "male",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "unknown",
"note": "subjective PERSON given assessment and plan \u00ff suspected myelinated nerve fiber layer od: - near circumferential on exam - per patient, has had enlarging blind spot on recent visual fields over DATE_TIME, has been followed at spindell eye, optometrist following concerned by visual field changes and told to come here - cos referral, would recommend obtaining baseline photos and visual field; in the interim, instructed patient to bring all records to this appointment PERSON, PERSON DATE_TIME DATE_TIME \u00ff PERSON, PERSONE DATE_TIME",
"gpt4_summary": "Patient has suspected myelinated nerve fiber layer od, near circumferential on exam. Experiencing enlarging blind spot; advised to bring records to next appointment. No mention of glaucoma.",
"glaucoma": "yes",
"use": "test"
},
{
"id": "data_08058",
"image_path": "slo_fundus_08058.jpg",
"filename": "data_08058.npz",
"report": "The patient has glaucoma and is taking multiple medications: Timolol/Brimonidine 1x/night in both eyes, Dorzolamide 3x/day in right eye, Netarsudil 1x/night in right eye, and artificial tears as needed in both eyes.",
"age": 72.18,
"gender": "female",
"race": "black",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "medication route frequency PERSON (teal)6 both eyes 1x/night timolol/brimonidine** (dark blue) the right eye 3x/day dorzolamide (orange)& both eyes 3x/day netarsudil (white)4 the right eye 1x/night preservative free artificial tears both eyes 6-8x/day as needed \u00f8 some medications that may be prescribed in lieu of this medication include: latanoprost, xalatan, travatan z, travaprost, LOCATION, LOCATION, LOCATION/tafluprost (preservative-free). 6 this medication is also known as vyzulta. ** this medication is also known as combigan, and it represents a combination of timolol and brimonidine. & some medications that may be prescribed in lieu of this medication include: dorzolamide, trusopt, PERSON, brinzolamide. 4 this medication is also known as rhopressa. always wait DATE_TIME (by the clock) in between eye drops. please bring this to your next visit. for routine questions, please call PHONE_NUMBER (LOCATION, ask for glaucoma department) or PHONE_NUMBER (longwood). for emergencies, please call PHONE_NUMBER and ask for the glaucoma physician on-call.",
"gpt4_summary": "The patient has glaucoma and is taking multiple medications: Timolol/Brimonidine 1x/night in both eyes, Dorzolamide 3x/day in right eye, Netarsudil 1x/night in right eye, and artificial tears as needed in both eyes.",
"glaucoma": "yes",
"use": "test"
},
{
"id": "data_08066",
"image_path": "slo_fundus_08066.jpg",
"filename": "data_08066.npz",
"report": "69 y.o. male has dry eyes, blepharitis, ocular hypertension, cataracts (not visually significant), refractive error. No glaucoma detected. Monitoring for now.",
"age": 69.8,
"gender": "male",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "69 y.o. male # dry eyes / blepharitis >at qid prn >warm compresses >omega-3 fatty acids >consider doxy # ocular hypertension ou -neg fhx -cct 594 / 606 -oct rnfl DATE_TIME: normal ou -hvf DATE_TIME: full ou >discussed monitoring vs starting drops. monitor for now. # cataract ou -not visually significant >monitor # refractive error >mrx given DATE_TIME f/u DATE_TIME: gonio, iop",
"gpt4_summary": "69 y.o. male has dry eyes, blepharitis, ocular hypertension, cataracts (not visually significant), refractive error. No glaucoma detected. Monitoring for now.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08068",
"image_path": "slo_fundus_08068.jpg",
"filename": "data_08068.npz",
"report": "Patient is a glaucoma suspect, with myopia being a risk factor. Intraocular pressure is within normal range, with possible mild corneal arcus. New prescription for myopia with astigmatism and presbyopia given. The patient also has cataracts and borderline high cholesterol.",
"age": 52.86,
"gender": "male",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "glaucoma suspect risks are myopia. iop is wnl, t corr is +3 ou. c/d - 0.4. 0.5 no fmhx. oct onh nfl wnl ou, gcl PERSON unreliable possible artifact LOCATION. hvf 24-2 watch inferior od wnl os DATE_TIME. myopia with astigmatism and presbyopia new rx given. warned of rd symptoms. cataracts ou multifocal PERSON with PERSON. states he does not wear his scl anywhere. observe. mild corneal arcus h/o borderline high cholesterol. observe. f/u in DATE_TIME for LOCATION, oct.",
"gpt4_summary": "Patient is a glaucoma suspect, with myopia being a risk factor. Intraocular pressure is within normal range, with possible mild corneal arcus. New prescription for myopia with astigmatism and presbyopia given. The patient also has cataracts and borderline high cholesterol.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08073",
"image_path": "slo_fundus_08073.jpg",
"filename": "data_08073.npz",
"report": "Patient seeking second opinion for eye issues. Goals include reducing intraocular pressure in both eyes. Glaucoma suspected due to elevated intraocular pressure. Current treatment includes Zioptan and PERSON medication and Xen Gel Stent procedure.",
"age": 56.51,
"gender": "male",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "single",
"note": "here for second opinion on DATE_TIME (referred by a patient of mine, ms. PERSON). attending's plan: -goal intraocular pressure less than or equal to 17 mmhg, right eye. -goal intraocular pressure less than or equal to 12 mmhg, left eye. -iop at goal od and above goal os on DATE_TIME on zioptan qhs os, PERSON, and s/p xen gel stent os. -continue PERSON -continue zioptan qhs os. -restart rhopressa qhs os. -hold PERSON> sample given on DATE_TIME. -instructions written/typed/printed out for patient (see table/details under patient instructions). -emphasized adherence to medication regimen. -preservative-free artificial tears as needed. -long discussion with patient on DATE_TIME: given elevated iop os on close to mtmtx and desire to minimize eye drop burden, we proceeded with open xen gel stent os on DATE_TIME. -rtc in DATE_TIME with iop check and disc photos ou, sooner prn. in future, we could try ocudose bid os, or pf cosopt bid os only (maybe branded PERSON as well). we may want to consider bgi os if iop still above goal and especially if progression demonstrated given numerous intolerances to medication. the information above was documented by PERSON as a scribe for PERSON on DATE_TIME.",
"gpt4_summary": "Patient seeking second opinion for eye issues. Goals include reducing intraocular pressure in both eyes. Glaucoma suspected due to elevated intraocular pressure. Current treatment includes Zioptan and PERSON medication and Xen Gel Stent procedure.",
"glaucoma": "yes",
"use": "test"
},
{
"id": "data_08075",
"image_path": "slo_fundus_08075.jpg",
"filename": "data_08075.npz",
"report": "Patient suspected of glaucoma due to cupping, has myopic astigmatism, Type 1 diabetes without retinopathy, and scheduled for silent sinus syndrome surgery.",
"age": 50.97,
"gender": "male",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "single",
"note": "cos clinic note assessment/plan: 1. glaucoma suspect due to cupping -full hvf at baseline, full rnfl -iop acceptable for now 2. myopic astigmatism -rx prn 3. diabetes type 1 without retinopathy ou -bg control -annual dfe 4. silent sinus syndrome -scheduled for surgery (ent) at DATE_TIME -pre-operative hertels exophthalmetry shows between 2-3mm enophthalmos of the right eye -follow-up after recovery from surgery to monitor degree of enophthalmos -report sent to dr. PERSON in DATE_TIME after sinus surgery for undilated exam",
"gpt4_summary": "Patient suspected of glaucoma due to cupping, has myopic astigmatism, Type 1 diabetes without retinopathy, and scheduled for silent sinus syndrome surgery.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08077",
"image_path": "slo_fundus_08077.jpg",
"filename": "data_08077.npz",
"report": "Patient referred for suspected open angle glaucoma with low risk in both eyes. Risk factors include family history of glaucoma (mother) and long-term steroid use. No history of eye trauma or other eye procedures. Undergone IOP tonometry with pressure 23 both eyes. No glaucoma medication issues or allergies. Advised follow up.",
"age": 38.1,
"gender": "female",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "PERSON is a DATE_TIME DATE_TIME patient referred by outside optometrist, first seen by dr. LOCATION on DATE_TIME # ohtn # open angle glaucoma suspect low risk od, low risk os risk factors: positive family history of glaucoma or blindness (mother with pigmentary dispersion), positive history of longterm steroids (steroid cream for eczema since ~2020), negative history of eye trauma (k abrasions only) central corneal thickness: 442 / 446 (DATE_TIME)gonioscopy: open ou tmax: ( ) / ( ) target iop: / refractive error wrx: od . . / os . . glaucoma procedures/lasers: none other eye procedures/lasers: none glaucoma medication issues: negative history of asthma negative history of bradycardia negative sulfa allergy negative history of renal disfuntion or kidney stones testing: baseline DATE_TIME (DATE_TIME) oct rnfl: (DATE_TIME) hvf 24-2 ou: #social: pt is pediatrician at Institution (nicu) #hx k abraision x 2 - once was a scratch and the other was a chemical in the eye/ lab related - unclear laterality # systemic: anxiety, svt s/p ablation plan DATE_TIME: iop tonometry tonometry (3:04 pm) right left pressure 23 23 discussed the natural history of glaucoma, diagnosis of glaucoma supspect, and the importance of follow up. paracentral defects both eyes left eye is superior and right eye is inferior which does correspond to thinning on optical coherence tomography. last dilated exam: DATE_TIME rnfl:DATE_TIME last visual field: DATE_TIME baseline disc photos: return to glaucoma clinic DATE_TIME with dilated fundus exam, disc photos, PERSON, and please check wrx (just read the glasses that she is wearing please)",
"gpt4_summary": "Patient referred for suspected open angle glaucoma with low risk in both eyes. Risk factors include family history of glaucoma (mother) and long-term steroid use. No history of eye trauma or other eye procedures. Undergone IOP tonometry with pressure 23 both eyes. No glaucoma medication issues or allergies. Advised follow up.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08078",
"image_path": "slo_fundus_08078.jpg",
"filename": "data_08078.npz",
"report": "Patient has PVD in right eye, mild cataract in both eyes, and normal, non-glaucomatous cupping in both eyes. Yearly follow-up planned.",
"age": 64.48,
"gender": "female",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "imp: pvd od mild cataract ou cupping ou--appears to be physiologic; +fh gl testing stable, wnl shallow ac; angles open by gonio refr error plan: yrly sooner prn j NRP, pgy4 i saw and evaluated this patient and discussed the case as appropriate with the resident/fellow. i have reviewed the resident/fellow's notes and made any necessary changes.",
"gpt4_summary": "Patient has PVD in right eye, mild cataract in both eyes, and normal, non-glaucomatous cupping in both eyes. Yearly follow-up planned.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08083",
"image_path": "slo_fundus_08083.jpg",
"filename": "data_08083.npz",
"report": "Madeline recently underwent glaucoma surgery in left eye. Her conditions include open angle glaucoma in both eyes, pseudophakia, astigmatism, myopia, presbyopia, and asthma. Showing improvement, but further treatment may be considered.",
"age": 65.06,
"gender": "female",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "divorced",
"note": "madeline a PERSON is a DATE_TIME. female ref for second opinion following recent glaucoma surgery in the left eye with PERSON eye history, from notes: s/p iridocyclitis os s/p bleb revision w mmc, 5fu and PERSON well s/p slt os (temp 1/2) pseudophakia ou s/p argon laser NRP os hyperopia astigmatism presbyopia posterior vitreous detachment od h/o suprachoroidal hemorrhage os myopia astigmatism presbyopia 1. open angle glaucoma ou s/p trab DATE_TIME, s/p bleb revision os DATE_TIME dr. PERSON at ocb, performed due to hypotony-->iop has improved hvf 9/17 early nasal changes od since DATE_TIME (prior to that ess full); os sa stable since DATE_TIME with inferior thinning ou + asthma travatan qhs od 2. pseudophakia ou plan: discussed current presentation and prognosis in detail, all questions answered. travatan qhs od dc pred PERSON is borderline in od and trending up in os. goal is likely mid teens od, low teens os. repeat iop check after stopping prednisolone--if not at goal, consider additional agent ou (vs slt od/gtt os) rtc sooner prn",
"gpt4_summary": "Madeline recently underwent glaucoma surgery in left eye. Her conditions include open angle glaucoma in both eyes, pseudophakia, astigmatism, myopia, presbyopia, and asthma. Showing improvement, but further treatment may be considered.",
"glaucoma": "yes",
"use": "test"
},
{
"id": "data_08096",
"image_path": "slo_fundus_08096.jpg",
"filename": "data_08096.npz",
"report": "The patient is currently taking econazole, sulfacetamide sodium-sulfur, valacyclovir, and acetaminophen-codeine for various conditions. They underwent Humphrey Visual Field and OCT optic nerve tests. Conditions include varicose veins and hypertension. No mention of glaucoma.",
"age": 63.79,
"gender": "female",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "total) by mouth DATE_TIME. econazole nitrate 1 % cream (taking) apply topically DATE_TIME. sulfacetamide sodium-sulfur (prascion) 10-5 % (w/w) clsr (taking) apply topically 2 (two) times a day. valacyclovir (valtrex) 1000 mg tablet (taking) take 2 tablets (2,000 mg total) by mouth 2 (two) times a day. x DATE_TIME only, as needed for cold sores PERSON DATE_TIME as needed acetaminophen-codeine (tylenol-codeine #PHONE_NUMBER mg per tablet take 1-2 tablets by mouth every 4 (DATE_TIME as needed. partial fill ok your orders normal orders this visit humphrey visual field - ou - both eyes oct, optic nerve - ou - both eyes - cirrus; optic disc; rnfl; 6mm length condition list as of DATE_TIME varicose veins of lower extremities with other complications hypertension results summary immunizations administered on date of encounter - DATE_TIME none",
"gpt4_summary": "The patient is currently taking econazole, sulfacetamide sodium-sulfur, valacyclovir, and acetaminophen-codeine for various conditions. They underwent Humphrey Visual Field and OCT optic nerve tests. Conditions include varicose veins and hypertension. No mention of glaucoma.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08097",
"image_path": "slo_fundus_08097.jpg",
"filename": "data_08097.npz",
"report": "63-year-old patient with history of hypertension, hyperlipidemia, diabetes, GERD, back pain, and thyroidectomy for Graves' disease. No glaucoma mentioned but patient has controlled hyperglycemia, mild cataracts, occasional floaters, and dry eyes.\n",
"age": 63.85,
"gender": "female",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "63 yo disabled medical assistant with history of htn, hyperlipidemia, diabetes, gerd, low back pain, hypothyroidism s/p thyroidectomy for graves' disease has 15 and 10 PERSON in revere. mva DATE_TIME, recovering from l knee injury \u00ff 1. dm diagnosed DATE_TIME -history of gestational dm -hba1c 8.1% DATE_TIME >> 7.5% DATE_TIME >> 7.1% DATE_TIME no evidence of diabetic retinopathy on dilated fundus examination DATE_TIME. patient was advised to maintain tight control of blood sugar and blood pressure. \u00ff 2. history of graves disease, s/p rai -slight orbital fat prominence/hooding ou, but no other orbital signs \u00ff 3. cataracts ou-mild - not yet visually significant, denies glare, but is using sunglasses more frequently in the sun, anxious about driving at DATE_TIME but not due to vision >> updated mrx given DATE_TIME at request, no change in rx \u00ff 4. pvd ou -sees occasional floaters \u00ff 5. incr'd c/d od> os: intact rims tmax 22/21. cct 590/600 (thick). no fhx hvf DATE_TIME: full ou oct DATE_TIME: wnl ou oct DATE_TIME: wnl ou baseline oct DATE_TIME: wnl ou >> iop and onh stable, continue to follow \u00ff 6. dry eyes/mgd ou: clean lids in am or in shower prn >> artificial tears prn, also has patanol prn (uses rarely), prefers brand name patanol, finds this works best PERSON i saw and evaluated this patient and discussed the case as appropriate with the resident/fellow. i have reviewed the resident/fellow's notes and made any necessary changes.",
"gpt4_summary": "63-year-old patient with history of hypertension, hyperlipidemia, diabetes, GERD, back pain, and thyroidectomy for Graves' disease. No glaucoma mentioned but patient has controlled hyperglycemia, mild cataracts, occasional floaters, and dry eyes.\n",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08101",
"image_path": "slo_fundus_08101.jpg",
"filename": "data_08101.npz",
"report": "Patient has history of poor compliance with glaucoma treatments. Diagnosis: mild glaucoma. Medications: latanoprost, timolol. Glaucoma procedures performed on both eyes. Assessment: thin cornea, progression of vision field loss, irreversible nature of glaucoma discussed. Plan: restart latanoprost and timolol.",
"age": 73.22,
"gender": "male",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "first seen by Person on DATE_TIME. previously seen dr. PERSON. he is here for transfer of care due to distance to main campus. he has a history of poor compliance and stopping eye drops for DATE_TIME. diagnosis: mild PERSON target iop: / , tmax: ( ) / ( ) central corneal thickness: 489 / 490 gonioscopy: cbb, 2+ pigment refractive error: od . . / os . . optic nerve/rnfl findings on baseline visit right eye (DATE_TIME): borderline superior/inferior thinning optic nerve/rnfl findings on baseline visit left eye (DATE_TIME): possible borderline superior thinning visual fields on baseline visit right eye (DATE_TIME): superior nasal step visual fields on baseline visit left eye (DATE_TIME): full medication history at first visit: latanoprost, timolol medication intolerances: rhopressa (cost) glaucoma procedures right eye: glaucoma procedures left eye: other eye procedures right eye: other eye procedures left eye: other eye problems right eye: other eye problems left eye: family history: mother steroids: trauma: asthma: other medical history and problems: assessment: 1. mild poag od>os -thin cct -poor compliance with hvf progression od; new baseline set DATE_TIME -discussed irreversible nature of glaucoma and strong recommendation to restart drops 2. ns cataracts ou -pt feels like vision or glasses have gotten weaker; refract next visit 3. posterior vitreous detachment od -last dfe DATE_TIME, was stable -rd precautions plan: -restart timolol 2/2, latanoprost 1/1 -pt undergoing cardiac surgery DATE_TIME; so we will defer next iop check until patient is medically stable return in DATE_TIME for mrx, iop, dilate",
"gpt4_summary": "Patient has history of poor compliance with glaucoma treatments. Diagnosis: mild glaucoma. Medications: latanoprost, timolol. Glaucoma procedures performed on both eyes. Assessment: thin cornea, progression of vision field loss, irreversible nature of glaucoma discussed. Plan: restart latanoprost and timolol.",
"glaucoma": "yes",
"use": "test"
},
{
"id": "data_08106",
"image_path": "slo_fundus_08106.jpg",
"filename": "data_08106.npz",
"report": "Patient with breast cancer history follows up for idiopathic intracranial hypertension (iih). Despite weight gain, visual status remains stable. Concerns over increased intracranial pressure discussed. No glaucoma mentioned.",
"age": 35.59,
"gender": "female",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "formulation: this patient with a history of breast cancer (s/p mastectomy with negative nodes, chemo/radiation) follows-up for iih. she seems to be doing very well from a visual standpoint, but she has gained considerable weight. my exam showed excellent afferent and efferent visual function (note that the field in one eye had to be repeated because of poor performance, which then improved). funduscopy reveals moderate frisen grade i edema ou. \u00ff i am very concerned about the weight gain, despite the essentially unchanged visual status, because it can trigger higher intracranial pressures. we had a significant talk about the risk and how we might manage that risk. she decided to again try to manage her diet, which is complicated by her 'craving' of sugar. i redressed the benefit of seeing a nutritionist, and she stated that she would call someone with whom she had previously worked. i encouraged regular weighings. we discussed use of diamox but we deferred, pending her weight in DATE_TIME. i also discussed a social worker-therapist, but she did not want to pursue this. she committed to a 10 pound weight loss by next visit. impression: 1. idiopathic intracranial hypertension, mild and stable without medical intervention 2. history of breast cancer \u00ff recommendations: 1. more attention to weight management 2. follow-up neuro-ophthalmic examination in DATE_TIME. consider initiating acetazolamide in the future",
"gpt4_summary": "Patient with breast cancer history follows up for idiopathic intracranial hypertension (iih). Despite weight gain, visual status remains stable. Concerns over increased intracranial pressure discussed. No glaucoma mentioned.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08107",
"image_path": "slo_fundus_08107.jpg",
"filename": "data_08107.npz",
"report": "Patient has diabetes type 2, history of stroke with left homonymous hemianopsia. RNFL showing superior thinning indicators of glaucoma. Cataracts present but not visually significant. Glaucoma consultation advised.",
"age": 58.5,
"gender": "male",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "-dm type 2 - no retinopathy -history of a stroke DATE_TIME, left homonymous hemianopsia -?poag: >>rnfl significant superior thinning ou, temporal and inferior thinning os >>hvf: right homonymous hemianopsia consistent with stroke, ?inferior nasal step os may benefit from glaucoma referral -refractive error/presbyopia -cataracts ou - not visually significant dilated exam cursory given pt. requesting to leave because of transportation issue. LOCATION, pgy4 this patient was unable to stay to see the attending and was only seen by the resident above. i have reviewed the resident's notes --suggest consultation with glaucoma service",
"gpt4_summary": "Patient has diabetes type 2, history of stroke with left homonymous hemianopsia. RNFL showing superior thinning indicators of glaucoma. Cataracts present but not visually significant. Glaucoma consultation advised.",
"glaucoma": "yes",
"use": "test"
},
{
"id": "data_08109",
"image_path": "slo_fundus_08109.jpg",
"filename": "data_08109.npz",
"report": "80 y.o. patient with cataracts and a history of glaucoma. Cataract surgery was delayed due to COVID-19. The patient hasn't been using prescribed medication, resulting in high intraocular pressure (IOP).",
"age": 80.22,
"gender": "female",
"race": "black",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "80 y.o. previously followed by dr. PERSON at bidmc last seen in DATE_TIME, was told she had cataracts ou last visit DATE_TIME. was supposed to have cataract surgery but postponed due to covid pandemic hx glaucoma - iop 14 ou - cct 580/568 - rnfl DATE_TIME od thin sup os wnl DATE_TIME 64/92 od this sup, borderline temporal, inferiorly; PERSON DATE_TIME high fn ou, od superior and inferior arcuate defect, os rim artifact - pt ran out of DATE_TIME and has not been using. iop to high DATE_TIME. > restart latanoprost nightly ou > consider istent ou combined cataract bilateral - visually significant and affecting activities of DATE_TIME living - dilates well, [no ] flomax, [no ] pxf, [no ] guttae, [no] blood thinners - discussed cataract surgery, patient elected to proceed with cataract surgery with iol left eye. right eye to follow. with istent > risks, benefits, and alternatives to surgery were discussed with the patient, including but not limited to the potential risk of infection, bleeding, loss of vision, loss of eye, need for further surgery or laser, retinal detachment, change in glasses and reading glasses, macular edema, increased eye pressure and glaucoma. no guarantees given. all patient questions were answered.? > discussed available iols including presbyopia and astigmatism correcting lenses including all out-of-pocket expenses. > pt opts for monofocal iol, aim distance ?> anesthesia: peribulbar block > other concerns:",
"gpt4_summary": "80 y.o. patient with cataracts and a history of glaucoma. Cataract surgery was delayed due to COVID-19. The patient hasn't been using prescribed medication, resulting in high intraocular pressure (IOP).",
"glaucoma": "yes",
"use": "test"
},
{
"id": "data_08112",
"image_path": "slo_fundus_08112.jpg",
"filename": "data_08112.npz",
"report": "51 y.o. female, suspected glaucoma due to increased cdr. Negative family history, normal OCT, mild nonspecific changes OD in HVF. Monitoring refractive error.",
"age": 51.73,
"gender": "female",
"race": "black",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "assessment and plan 51 y.o. female with h/o left-sided bell's palsy # glaucoma suspect due to mildly incr cdr -neg fhx -pachy: 520/524 -oct DATE_TIME: normal ou -hvf DATE_TIME: reduced reliability ou, mild nonspec changes od, full os >monitor # refractive error >mrx given (svl reading) rtc DATE_TIME: coe, hvf, oct",
"gpt4_summary": "51 y.o. female, suspected glaucoma due to increased cdr. Negative family history, normal OCT, mild nonspecific changes OD in HVF. Monitoring refractive error.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08119",
"image_path": "slo_fundus_08119.jpg",
"filename": "data_08119.npz",
"report": "Patient is a glaucoma suspect but has normal visual fields in both eyes. Intraocular pressure and testing show stable results. No treatment initiated, planned for monitoring.",
"age": 73.99,
"gender": "male",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "first seen by dr. PERSON on DATE_TIME diagnosis: glaucoma suspect target iop: DATE_TIME, tmax: 14 (DATE_TIME) / 15.9 (DATE_TIME) central corneal thickness: 567.567.568 / 579.579.579 corneal hysteresis: 11.7 / 11.4* gonioscopy: refractive error: od +1.25. -0.25. 102 / os +1.25. -0.25. 087 optic nerve findings on initial visit right eye: optic nerve findings on initial visit left eye: visual fields on initial visit right eye: normal visual fields on initial visit left eye: normal medication history and intolerances: none glaucoma procedures right eye: none glaucoma procedures left eye: none other eye procedures right eye: none other eye procedures left eye: none other eye problems right eye: none other eye problems left eye: none family history: sister, no vision loss steroids: no trauma: no asthma: no other medical history and problems: none plan: # glaucoma suspect due to cup to disc ratio, both eyes - intraocular pressure acceptable both eyes, testing stable both eyes and findings reassuring on examination - continue to monitor without initiating treatment - return to clinic DATE_TIME for intraocular pressure check, humphrey visual field, optical coherence tomography and dilate # cataract, both eyes - mild, not visually significant, monitor i, PERSON, am acting as scribe for PERSONmd, PERSON for patient PERSON on DATE_TIME.",
"gpt4_summary": "Patient is a glaucoma suspect but has normal visual fields in both eyes. Intraocular pressure and testing show stable results. No treatment initiated, planned for monitoring.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08120",
"image_path": "slo_fundus_08120.jpg",
"filename": "data_08120.npz",
"report": "The 67 y.o. male patient has a history of glaucoma (POAG), with an increased c:d ratio and controlled IOP. He also shows ocular scars and has a history of ocular migraines and scleral lens wear.",
"age": 67.95,
"gender": "male",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "single",
"note": "67 y.o. m 1. s/p phaco/pciol os (aim intermediate) doing well happy with vision mild pco, recheck next time (pt not bothered yet) 2. poag(s), based on inc c:d ratio ou fhx negative tmax mid teens per pt report PERSON full ou cct thin ou oct abnl thinning, stable ou dp stable iop controlled ou (goal low teens ou) and hvf full observe closely 3. cr scars ou in macula, stable, f/u dr young, - amd vs pattern dystrophy 4. ocular migraine, no ha followed, 1st episode DATE_TIME no recent sx 5. h/o scl wear monovision, od distance, os near no longer wearing PERSON has svd for DATE_TIME driving only, and prefers no correction for near DATE_TIME. s/p phaco/pciol od (aim distance) DATE_TIME, mild pco, recheck next time (pt not bothered yet) f/u 6 PERSON check",
"gpt4_summary": "The 67 y.o. male patient has a history of glaucoma (POAG), with an increased c:d ratio and controlled IOP. He also shows ocular scars and has a history of ocular migraines and scleral lens wear.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08127",
"image_path": "slo_fundus_08127.jpg",
"filename": "data_08127.npz",
"report": "27 y.o. white, non-hispanic female. No diagnosis of glaucoma. Exam as needed. Assisted by Huy Nguyen, PGY3.",
"age": 27.87,
"gender": "female",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "unknown",
"note": "a 27 y.o. white, non-hispanic female with no diagnosis of glaucoma. examination prn (this note was prepared with assistance from huy nguyen, pgy3)",
"gpt4_summary": "27 y.o. white, non-hispanic female. No diagnosis of glaucoma. Exam as needed. Assisted by Huy Nguyen, PGY3.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08132",
"image_path": "slo_fundus_08132.jpg",
"filename": "data_08132.npz",
"report": "The patient is taking timolol 0.5% ophthalmic gel, dropped in each eye daily. They are also taking trazodone orally. The patient has been referred for visual field and optic nerve examinations. No mention of glaucoma. Other health conditions include chronic renal impairment, hyperkalemia, hyperlipidemia, gout, hypothyroidism, GERD, and hypertension.\n",
"age": 86.55,
"gender": "male",
"race": "white",
"ethnicity": "non-hispanic",
"language": "unknown",
"maritalstatus": "married or partnered",
"note": "timolol (timoptic-xe) 0.5 % ophthalmic gel-forming (taking) place 1 drop into each eye daily. trazodone hcl (trazodone oral) (taking) take by mouth. your orders normal orders this visit ambulatory referral to Institution ophthalmology humphrey visual field - ou - both eyes oct, optic nerve - ou - both eyes - cirrus; rnfl condition list as of DATE_TIME chronic renal impairment hyperkalemia hyperlipidemia gout hypothyroidism gastroesophageal reflux disease hypertensive disorder results summary immunizations administered on date of encounter - DATE_TIME none patient gateway activation information your account is ready to use. activate your account using following steps: 1. visit URL. 2. click 'enroll now' and create your user account. important information about your account: ? if you already have a account, please sign in with your username and password. ? you must log into your account within DATE_TIME or activation code will expire. ? trouble logging in? use support link found on the top right side of the login page. ? need help? support team makes every effort to respond to phone or email messages within DATE_TIME. to reach the support team, email EMAIL_ADDRESSPHONE_NUMBER. calls or messages are answered DATE_TIME, DATE_TIME, est - and will make every effort to contact you within DATE_TIME.",
"gpt4_summary": "The patient is taking timolol 0.5% ophthalmic gel, dropped in each eye daily. They are also taking trazodone orally. The patient has been referred for visual field and optic nerve examinations. No mention of glaucoma. Other health conditions include chronic renal impairment, hyperkalemia, hyperlipidemia, gout, hypothyroidism, GERD, and hypertension.\n",
"glaucoma": "yes",
"use": "test"
},
{
"id": "data_08135",
"image_path": "slo_fundus_08135.jpg",
"filename": "data_08135.npz",
"report": "The patient has been advised to use warm compresses and stop eye rubbing. Potential referral for ABMD. Needs to return to the Glaucoma Clinic for further tests.",
"age": 59.32,
"gender": "male",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "prescribing encouraged warm compresses bid, pf ats qid stop eye rubbing last dilated exam: DATE_TIME last oct rnfl: DATE_TIME last visual field: DATE_TIME baseline disc photos: DATE_TIME consider referral to cornea for abmd if symptomatically not improved return to glaucoma clinic in DATE_TIME with NRP, humphrey visual field 24-2, and optical coherence tomography retinal nerve fiber layer.",
"gpt4_summary": "The patient has been advised to use warm compresses and stop eye rubbing. Potential referral for ABMD. Needs to return to the Glaucoma Clinic for further tests.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08136",
"image_path": "slo_fundus_08136.jpg",
"filename": "data_08136.npz",
"report": "Patient is a suspect of narrow angle glaucoma due to hereditary reasons. She previously suffered sudden vision loss and had been diagnosed with Posner-Schlossman syndrome. She's also had episodes of bright light vision, diagnosed as ocular migraines. She had laser peripheral iridotomy (LPI), which reduced the symptoms. The patient's intraocular pressure is currently acceptable and her visual field is stable, but she has mild posterior capsular opacification in right eye. Follow-up planned in 6 months.",
"age": 70.87,
"gender": "female",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "1) narrow angle glaucoma suspect ou - fh: mother - angle closure; no steroids or trauma - complex recent eye history: pt always told she had open angles DATE_TIME: sudden vision loss od and flu-like symptoms, headache; iop 51; initially diagnosed with posner-schlossman; LOCATION with LOCATION, LOCATION, also prednisolone, now off all drops DATE_TIME: had several episodes of seeing bright light ou lasting or DATE_TIME, seen by neuro-ophth, diagnosed with ocular migraines and also told she had narrow angles had lpis ou in DATE_TIME, no episodes since - gonio DATE_TIME narrow, with very small pis (barely visible and not patent os) - oct full, hvf full - repeat iridotomy od and os DATE_TIME - ce ou in LOCATION DATE_TIME) pciol ou - PERSON plan: doing well intraocular pressure acceptable rnfl oct and visual field stable mild posterior capsular opacification right eye rtc 6 mths dilate, disc photos",
"gpt4_summary": "Patient is a suspect of narrow angle glaucoma due to hereditary reasons. She previously suffered sudden vision loss and had been diagnosed with Posner-Schlossman syndrome. She's also had episodes of bright light vision, diagnosed as ocular migraines. She had laser peripheral iridotomy (LPI), which reduced the symptoms. The patient's intraocular pressure is currently acceptable and her visual field is stable, but she has mild posterior capsular opacification in right eye. Follow-up planned in 6 months.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08140",
"image_path": "slo_fundus_08140.jpg",
"filename": "data_08140.npz",
"report": "The patient has glaucoma. The intraocular pressure (IOP) is above target in both eyes, an additional medication (Rhopressa) was recommended due to new disc hemorrhage in right eye and increasing IOP, but the patient refused. Retinal detachment precautions reviewed.",
"age": 61.86,
"gender": "female",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "single",
"note": "glaucoma as well as other theories (as of DATE_TIME). attending's plan: -goal intraocular pressure less than or equal to 14 mmhg, right eye. -goal intraocular pressure less than or equal to 14 mmhg, left eye. -iop above goal ou on DATE_TIME on PERSON. -continue vyzulta qhs ou. -long discussion with patient on DATE_TIME: i strongly recommended starting rhopressa qhs ou in addition to vyzulta qhs ou given iop above goal ou (set by dr. PERSON) and new disc hemorrhage od noted on DATE_TIME with iop of 12 mmhg. she vehemently refused despite risk of permanent irreversible vision loss explained to the patient. -retinal detachment precautions were reviewed with the patient. -preservative-free artificial tears as needed. -rtc in DATE_TIME with iop check, hvf, dilation, and disc photos ou, sooner prn. if persistently above goal, recurring disc hemorrhages or progression, i will revisit treatment or urge patient to explore a second opinion. i saw and evaluated this patient and discussed the case as appropriate with the resident/fellow (PERSON, md). i have reviewed the resident/fellow's notes and made any necessary changes. i spent DATE_TIME with the patient, more than half of which was spent on counseling and coordination of care for this patient's glaucoma and other ocular co-morbidities.",
"gpt4_summary": "The patient has glaucoma. The intraocular pressure (IOP) is above target in both eyes, an additional medication (Rhopressa) was recommended due to new disc hemorrhage in right eye and increasing IOP, but the patient refused. Retinal detachment precautions reviewed.",
"glaucoma": "yes",
"use": "test"
},
{
"id": "data_08141",
"image_path": "slo_fundus_08141.jpg",
"filename": "data_08141.npz",
"report": "The patient is taking timolol/brimonidine, also known as Combigan, for both eyes 2x/day and the right eye 3x/day. This indicates the presence of glaucoma.",
"age": 70.29,
"gender": "male",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "results summary immunizations administered on date of encounter - DATE_TIME none instructions from this visit: medication route frequency timolol/brimonidine** (dark blue) both eyes 2x/day PERSON (orange)& the right eye 3x/day ** this medication is also known as combigan, and it represents a combination of timolol and brimonidine. & some medications that may be prescribed in lieu of this medication include: dorzolamide, trusopt, PERSON, brinzolamide. always wait DATE_TIME (by the clock) in between eye drops. please bring this to your next visit. for routine questions, please call PHONE_NUMBER (LOCATION) or PHONE_NUMBER (LOCATION). for emergencies, please call PHONE_NUMBER and ask for the glaucoma physician on-call.",
"gpt4_summary": "The patient is taking timolol/brimonidine, also known as Combigan, for both eyes 2x/day and the right eye 3x/day. This indicates the presence of glaucoma.",
"glaucoma": "yes",
"use": "test"
},
{
"id": "data_08142",
"image_path": "slo_fundus_08142.jpg",
"filename": "data_08142.npz",
"report": "Patient has ocular hypertension, with heavy pigment and Pseudoexfoliation Syndrome (PXE). No definite damage noted. Higher PXE and intraocular pressure indicate need for monitoring. Diagnosed as glaucoma suspect.",
"age": 71.53,
"gender": "female",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "# ocular hypertension pxe , os>od +fhx, brother and mother PERSON ou ttarget: / , tmax: 26/24 (for a long time per patient, never treated) cct: / gonioscopy: open ou, heavy pigment os rnfl PERSON vf normal ou plan: no definite damage at this time pxe os and iop running higher needs close monitoring check pachy next visit , dilate, rtc 4 mths diagnosis of glaucoma/glaucoma suspect discussed with patient - review of natural history and management options discussed. importance of lifelong follow up and adherence to treatments discussed in order to lower risk of permanent vision loss/blindness from glaucoma. all questions answered.",
"gpt4_summary": "Patient has ocular hypertension, with heavy pigment and Pseudoexfoliation Syndrome (PXE). No definite damage noted. Higher PXE and intraocular pressure indicate need for monitoring. Diagnosed as glaucoma suspect.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08147",
"image_path": "slo_fundus_08147.jpg",
"filename": "data_08147.npz",
"report": "Patient screening shows no signs of fever, cough, breath shortness, sore throat, muscle aches, nasal congestion, or loss of smell/taste. No concerning symptoms or comorbidities. No mention of glaucoma.\n",
"age": 68.74,
"gender": "female",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "PERSON pre-visit symptom screening date: DATE_TIME symptoms fever: no cough: no mild shortness of breath: no sore throat: no muscle aches: no nasal congestion: no loss of smell/taste: no atypical symptom concerning for PERSON: no concerning symptoms or comorbidities: no additional symptom comments:not answered date of symptom onset: not answered pre-visit screening in DATE_TIME, has the patient had a positive or pending DATE_TIME test outside of Institution? no in DATE_TIME, has the patient spent DATE_TIME within 6 feet of anyone currently infected with DATE_TIME? no",
"gpt4_summary": "Patient screening shows no signs of fever, cough, breath shortness, sore throat, muscle aches, nasal congestion, or loss of smell/taste. No concerning symptoms or comorbidities. No mention of glaucoma.\n",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08148",
"image_path": "slo_fundus_08148.jpg",
"filename": "data_08148.npz",
"report": "Patient has posterior vitreous detachment (PVD) in right eye (OD), nuclear sclerotic (NS) cataract in both eyes (OU), and suspected glaucoma more in left eye (OS) than right.",
"age": 59.78,
"gender": "female",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "cos clinic note assessment/plan: 1. pvd od -repeat dfe without any rts or rds -rd return precautions reviewed 2. ns cataract ou -monitor 3. glaucoma suspect os>od -gonio open, PERSON, cct 516/515 -hvf and oct with slight defects os only -would monitor off of iop meds to establish trend rtc in DATE_TIME for iop, hvf ou",
"gpt4_summary": "Patient has posterior vitreous detachment (PVD) in right eye (OD), nuclear sclerotic (NS) cataract in both eyes (OU), and suspected glaucoma more in left eye (OS) than right.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08155",
"image_path": "slo_fundus_08155.jpg",
"filename": "data_08155.npz",
"report": "Patient is a 62-year-old endoscopy nurse with various conditions including hypertension, hyperlipidemia, diabetes, sleep apnea, coronary artery disease, etc. Has ocular hypertension with thick corneas and a high risk of glaucoma due to family history.",
"age": 62.05,
"gender": "female",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "single",
"note": "62 yo endoscopy rn at bmc with history of htn, hyperlipidemia, dm, sleep apnea, cad s/p 4-vessel cabg 5/21/1/4 (sternal wound debridement 6/27, flap DATE_TIME), hashimoto's thyroiditis, cervical dysplasia s/p leep DATE_TIME, gerd, anemia, psoriasis,pseudogout \u00ff 1. s/p phaco/pciol ou (DATE_TIME od and DATE_TIME os) - tr pco os> od, but remains asymptomatic >> requested mrx DATE_TIME \u00ff 2. diabetes x DATE_TIME -hba1c 8.1% DATE_TIME >> 6.8% DATE_TIME >> 6.8% DATE_TIME >> 8.9% DATE_TIME >> 6.9% DATE_TIME >> 8.3% DATE_TIME (ordered for low bcva): wnl ou no evidence of diabetic retinopathy on examination DATE_TIME. patient was advised to maintain tight control of blood sugar and blood pressure. \u00ff 3. ocular hypertension tmax 20/18. cct DATE_TIME: 582/569 (thick/ave) +strong family history of glaucoma (brother (s/p mx rd) and father (now deceased)) gonio DATE_TIME: cbb 360' ou hvf DATE_TIME: full ou hvf DATE_TIME: od full with non-specific defects, os full hvf DATE_TIME: full ou hvf DATE_TIME: od full. os borderline superior rim losses. (dif from DATE_TIME) hvf DATE_TIME: od borderline superior rim losses. os full DATE_TIME: wnl ou oct DATE_TIME: od sup thinning, worse than prior. os wnl DATE_TIME: PERSON, stable. PERSON, poor signal DATE_TIME. oct DATE_TIME: od borderline superior thinning, stable. os wnl DATE_TIME: od borderline superior thinning. os wnl >> DATE_TIME: iop stable. DATE_TIME worse but it historically fluctuates and no clear hvf correlate. will follow \u00ff 4. mild erm os - asymptomatic \u00ff 5. pvd od >> rd precautions \u00ff 6. peripapillary choroidal pigmentation",
"gpt4_summary": "Patient is a 62-year-old endoscopy nurse with various conditions including hypertension, hyperlipidemia, diabetes, sleep apnea, coronary artery disease, etc. Has ocular hypertension with thick corneas and a high risk of glaucoma due to family history.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08158",
"image_path": "slo_fundus_08158.jpg",
"filename": "data_08158.npz",
"report": "72 y.o. female with history of fibromyalgia and breast cancer. She displays low suspicion of glaucoma, no significant cataracts and dry eye syndrome. Also has refractive error.",
"age": 72.34,
"gender": "female",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "single",
"note": "72 y.o. female with h/o fibromyalgia, possible chronic fatigue syndrome, breast ca diagnosed in DATE_TIME s/p surgery/radiation previously followed by PERSON suspect based on cup:disc asymmetry os>od (low suspicion) noted to have pigment in anterior chamber post-dilation by dr. PERSON in past; no k spindle visible on exam fam hx: denies tmax: 18/19 cct: PERSON gonio DATE_TIME: ou open to ptm hvf DATE_TIME: od full, os few inferior rim defects DATE_TIME: PERSON, os few inferior defects likely full rnfl DATE_TIME: ou wnl DATE_TIME: ou wnl and stable disc photos: DATE_TIME last dilated: DATE_TIME >> observe off eyedrops - narrow angles ou not occludable on gonioscopy DATE_TIME >> monitor. signs & symptoms of acute angle closure glaucoma discussed - combined senile cataract ou not visually significant and not affecting activities of DATE_TIME living at this time >> observe - meibomian gland dysfunction/dry eye syndrome ou dry ocular surface DATE_TIME, reports intermittent irritation >> treatment with warm compresses, fish/flaxseed oil, artificial tears discussed - refractive error >> a prescription for new glasses was given to the patient last visit. f/up DATE_TIME with mrx, bat, hvf, rnfl oct, gonio, then dilation discussed longer wait times due to need testing, gonio by md, then dilation. pt stated understanding; ok with this as long as appts scheduled on DATE_TIME",
"gpt4_summary": "72 y.o. female with history of fibromyalgia and breast cancer. She displays low suspicion of glaucoma, no significant cataracts and dry eye syndrome. Also has refractive error.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08159",
"image_path": "slo_fundus_08159.jpg",
"filename": "data_08159.npz",
"report": "The patient has visual field defects potentially from a right-sided homonymous defect and Wet AMD. Brain MRI ordered to rule out left occipital ischemic event. No mention of glaucoma.",
"age": 72.02,
"gender": "female",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "widowed",
"note": "left side that could represent an acute/subacute occipital ischemic event. i believe that the patient could have noticed the visual field defect only in the left eye because the vision in her right eye is more severely impaired. her visual fields are hard to interpret given the defects caused by the amd, but there is a nasal visual field defect in the left eye and a corresponding temporal field defect in the right eye that could potentially represent a right-sided homonymous visual field defect. given that finding, i ordered a brain mri to rule out an occipital ischemic event. i will decide on further management depending on the results from the imaging. impression: 1. wet amd - treated with eylea 2. ?new right homonymous visual field defect - rule out ischemic event in the left occipital lobe 3. pseudophakia ou 4. pco os plan: 1. mri brain 2. follow up depending on the results of #1 PERSON, md",
"gpt4_summary": "The patient has visual field defects potentially from a right-sided homonymous defect and Wet AMD. Brain MRI ordered to rule out left occipital ischemic event. No mention of glaucoma.",
"glaucoma": "yes",
"use": "test"
},
{
"id": "data_08160",
"image_path": "slo_fundus_08160.jpg",
"filename": "data_08160.npz",
"report": "The patient has stable chronic conditions, including glaucoma \u2013 evident from good Intraocular Pressure (IOP) control. Currently, on a drug regimen (Cosopt) which should continue, especially as moderate risk of progression without care. Moderate activity such as marathon training can resume.",
"age": 41.31,
"gender": "male",
"race": "black",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "single",
"note": "given that iop pressures were at goal, we decided to observe on current regimen. we discussed slt and a handout was provided in case patient wants to come off cosopt. i recommended that he still continue the current medication regimen given that his tmax was 55 mmhg os, and he is about to resume marathon training. -long discussion with patient on DATE_TIME: given stable chronic illnesses with good iop control, we will continue monitoring the patient on the prescription drug regimen described above. importance of follow-up explained to patient given moderate risk of progression without care. we reviewed hvf/oct together, and they were stable! -rtc in DATE_TIME for iop check, dilation, and disc photos ou, sooner prn. strict return precautions, especially if marathon training resumes. the information above was documented by PERSON as a scribe for PERSON on DATE_TIME. i was present while this encounter was recorded and agree that the information entered by my scribe is complete and accurate.",
"gpt4_summary": "The patient has stable chronic conditions, including glaucoma \u2013 evident from good Intraocular Pressure (IOP) control. Currently, on a drug regimen (Cosopt) which should continue, especially as moderate risk of progression without care. Moderate activity such as marathon training can resume.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08162",
"image_path": "slo_fundus_08162.jpg",
"filename": "data_08162.npz",
"report": "74 yo woman with oculopharyngeal muscular dystrophy has cataracts, is considering surgery. Increased c/d. No family history of glaucoma. Vision within normal limits on assessments. Experiencing diplopia.",
"age": 74.81,
"gender": "female",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "74 yo woman with history of oculopharyngeal muscular dystrophy x DATE_TIME new patient DATE_TIME, prior care at tufts c/o needs glasses to see, must use glasses for distance and tv 1. cataract ou DATE_TIME: -starting to become symptomatic os>od with DATE_TIME driving -would like to observe for now, considering surgery for DATE_TIME after she returns from DATE_TIME in LOCATION 2. incr'd c/d ou tmax 15 ou. cct DATE_TIME (ave). no fhx glaucoma. hvf DATE_TIME: full ou (os fixation losses) 1st hvf DATE_TIME: od reliable, wnl. os rare rim losses DATE_TIME: wnl ou oct DATE_TIME: wnl ou oct DATE_TIME: PERSON ou >> will follow 3. s/PERSON 3 (PERSON), most recently 8/17 -very happy with results 4. PERSON (she is uncertain if binocular) -noted only at distance during DATE_TIME, noted over DATE_TIME >> orthophoric in primary gaze, only -1 limitation os in elevation abduction, and adduction, but denies diplopia >> neuro-op consultation for diplopia in setting of omd. in meantime will update glasses prescription for optimal acuity",
"gpt4_summary": "74 yo woman with oculopharyngeal muscular dystrophy has cataracts, is considering surgery. Increased c/d. No family history of glaucoma. Vision within normal limits on assessments. Experiencing diplopia.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08165",
"image_path": "slo_fundus_08165.jpg",
"filename": "data_08165.npz",
"report": "The patient has uveitis-glaucoma-hyphema syndrome and angle recession in left eye. Their right eye has a mild, not visually significant cataract. No thinning in retinal nerve fiber layer for either eye. They were intolerant to brimonidine glaucoma medication. No family history of eye diseases.",
"age": 50.26,
"gender": "male",
"race": "white",
"ethnicity": "hispanic",
"language": "english",
"maritalstatus": "single",
"note": "first seen by dr. PERSON on DATE_TIME glaucoma medication intolerances: brimonidine (follicular conjunctivitis) target iop: DATE_TIME, tmax: 24 / 46 central corneal thickness: 547 / 540 gonioscopy: od: d35f 2+; os: PERSON 4+, s/p nasal goniotomy retinal nerve fiber layer, right eye: no thinning retinal nerve fiber layer, left eye: no thinning visual fields, right eye: full visual fields, left eye: full family history: none steroids: none trauma: fell and hit head as a teenager asthma/copd: none other medical history and problems: generally healthy assessment/plan: 50 y.o. male # uveitis-glaucoma-hyphema (ugh) syndrome and angle recession, left eye - s/p iol exchange/kahook os (DATE_TIME, ma60ac +19.0 d in LOCATION, truncated superior haptic of prior sa60at remains in capsular bag) - initially presented to Institution ed DATE_TIME with iop 46 that improved with drops; inferior iol optic/haptic was out of capsular bag and subluxed inferiorly/anteriorly - iop acceptable ou, vf and oct stable - continue dorzolamide/timolol bid os - return in DATE_TIME for iop check, dilate, disc photos # cataract, right eye - mild, not visually significant, monitor PERSON, md",
"gpt4_summary": "The patient has uveitis-glaucoma-hyphema syndrome and angle recession in left eye. Their right eye has a mild, not visually significant cataract. No thinning in retinal nerve fiber layer for either eye. They were intolerant to brimonidine glaucoma medication. No family history of eye diseases.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08166",
"image_path": "slo_fundus_08166.jpg",
"filename": "data_08166.npz",
"report": "The patient has primary open angle glaucoma, moderate stage in right eye and mild stage in left eye. They also suffer from essential hypertension, class 1 obesity, uterine leiomyoma.",
"age": 64.06,
"gender": "female",
"race": "black",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "widowed",
"note": "DATE_TIMEtitution comp oph mc tech PERSON main campus DATE_TIME DATE_TIME PERSONtitution comprehensive oph main campus PHONE_NUMBER orders placed this visit humphrey visual field - ou - both eyes oct, optic nerve - ou - both eyes - cirrus; optic disc; rnfl condition list as of DATE_TIME uterine leiomyoma essential hypertension class 1 obesity healthcare maintenance mva restrained driver abnormal magnetic resonance imaging of cervical spine primary open angle glaucoma of right eye, moderate stage primary open angle glaucoma of left eye, mild stage results summary immunizations administered on date of encounter",
"gpt4_summary": "The patient has primary open angle glaucoma, moderate stage in right eye and mild stage in left eye. They also suffer from essential hypertension, class 1 obesity, uterine leiomyoma.",
"glaucoma": "yes",
"use": "test"
},
{
"id": "data_08168",
"image_path": "slo_fundus_08168.jpg",
"filename": "data_08168.npz",
"report": "Patient experienced transient r-sided visual loss on unclear date. Dr suspected migraine aura or potential TIA. Recommended starting asa 81mg. Patient refused. No mention of glaucoma.",
"age": 67.43,
"gender": "female",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "concerning features >> follow \u00ff 7. hx transient visual obscuration DATE_TIME r-sided visual loss (unclear if monocular or binocular) on DATE_TIME - saw dr. PERSON on DATE_TIME, believed most likely migraine aura, but could not rule-out tia, recommended beginning asa 81mg - following with pcp for risk factor optimization - patient does not wish to take aspirin 81mg, not currently taking, does not believe this was a tia - follow-up with dr. PERSON prn \u00ff 8. refractive error - no change in mrx DATE_TIME >> will provide eyeglasses rx DATE_TIME as backup (PERSON)",
"gpt4_summary": "Patient experienced transient r-sided visual loss on unclear date. Dr suspected migraine aura or potential TIA. Recommended starting asa 81mg. Patient refused. No mention of glaucoma.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08169",
"image_path": "slo_fundus_08169.jpg",
"filename": "data_08169.npz",
"report": "64yo female is a glaucoma suspect due to increased cup/disc ratio. Regular testing in the past; current tests are nonspecific for glaucoma. No intervention needed at this time.",
"age": 64.72,
"gender": "female",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "64 f hx breast ca, tongue ca for f/u glaucoma testing. # glaucoma suspect with increased cup/disc ratio, left > right eye. pt recalls having regular glaucoma testing due to suspicious cupping in the past. [ fhx: no [ cct: 555,569 [ oct DATE_TIME: superior and inftemp thinning od, sup / temp / inf thinning os (stable) [ hvf DATE_TIME: full od, focal supnas defect corresponding to retinal lesion os - although abnormal, the oct is nonspecific for glaucoma, and the visual field test shows no evidence of glaucomatous defects. - no intervention required at this time. continue to monitor. # old posterior vitreous detachment, both eyes. no retinal breaks seen. - discussed retinal detachment precautions. - monitor; return for new symptoms. # NRP, left eye, previously noted - monitor # refractive error. still having some trouble adapting to progressives. - rx given for new glasses. may wish to try standard bifocals or single-vision glasses. rtc 1 year, sooner prn",
"gpt4_summary": "64yo female is a glaucoma suspect due to increased cup/disc ratio. Regular testing in the past; current tests are nonspecific for glaucoma. No intervention needed at this time.",
"glaucoma": "yes",
"use": "test"
},
{
"id": "data_08172",
"image_path": "slo_fundus_08172.jpg",
"filename": "data_08172.npz",
"report": "The patient is suspected of having glaucoma with notable thick cornea & asymmetrical cupping. Mother on latanoprost for high eye pressure. No field loss noted. Normal OCT in left, thin retinal nerve fiber layer in right. Retinas are attached, some posterior vitreous detachment in left eye. The patient also has dry eyes, mild cataracts, and seasonal allergies.",
"age": 61.32,
"gender": "female",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "glaucoma suspect ou c/d asym, thick cornea--Tcorr is -3.5 OD, -3 os. mom on latanoprost and high iop. hvf 24-2 is wnl ou. oct onh nfl thin OD WNL OS. posterior vitreous detachment os shafer negative, retina attached 360 on dilated eye exam of both eyes. retinal detachment precautions reviewed. PERSON gland dysfunction and dry eyes, bilateral mild ocular surface dryness. warm compresses, lid hygiene. artificial tears as needed. mild age related cataracts, bilateral not visually significant. monitor. seasonal allergy worse in DATE_TIME. gave pt allergy handout. f/u in DATE_TIME for LOCATION, oct, ar/refract. by signing my name below, i, PERSON, acting as a scribe, attest that this documentation has been prepared under the direction and in the presence of PERSON, LOCATION DATE_TIME by signing my name below, i, PERSON, attest that this documentation has been prepared under my direction.",
"gpt4_summary": "The patient is suspected of having glaucoma with notable thick cornea & asymmetrical cupping. Mother on latanoprost for high eye pressure. No field loss noted. Normal OCT in left, thin retinal nerve fiber layer in right. Retinas are attached, some posterior vitreous detachment in left eye. The patient also has dry eyes, mild cataracts, and seasonal allergies.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08174",
"image_path": "slo_fundus_08174.jpg",
"filename": "data_08174.npz",
"report": "Patient with history of eye trauma experienced by blunt force, has normal test results with no glaucoma or retinal tears. Warned of retinal detachment symptoms. Follow-up planned.",
"age": 42.01,
"gender": "male",
"race": "asian",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "h/o blunt force trauma to the eye (DATE_TIME) pt got punched in the face- was seen on 01/14 at main campus after the event. he is experiencing flashes of light od. Optos, oct and hvf done. oct onh nfl is normal ou. OCT mac WNL. hvf is wnl ou. depressed 360. no retinal tears or holes. warned of RD Sx. will f/u in DATE_TIME. pvd od/ Syneresis os observe. dermatochalasis od>os longstanding. observe. f/u in DATE_TIME dfe.",
"gpt4_summary": "Patient with history of eye trauma experienced by blunt force, has normal test results with no glaucoma or retinal tears. Warned of retinal detachment symptoms. Follow-up planned.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08175",
"image_path": "slo_fundus_08175.jpg",
"filename": "data_08175.npz",
"report": "The patient has been prescribed an ophthalmic solution for both eyes. They have been diagnosed with glaucoma, among other conditions like hypertensive disorder, osteoporosis, and arteriosclerotic heart disease.",
"age": 70.92,
"gender": "female",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "ophthalmic solution (taking) place 1 drop into each eye every 4 (DATE_TIME. your orders normal orders this visit humphrey visual field - od - right eye oct, optic nerve - od - right eye - cirrus; optic disc; rnfl future labs/procedures complete by expires b-scan ultrasound - os - left eye as directed DATE_TIME condition list as of DATE_TIME hypertensive disorder osteoporosis glaucoma arteriosclerotic heart disease sensorineural hearing loss asthma fracture of humerus hyperlipidemia results summary immunizations administered on date of encounter - DATE_TIME none patient gateway activation information your account is ready to use. LOCATION.",
"gpt4_summary": "The patient has been prescribed an ophthalmic solution for both eyes. They have been diagnosed with glaucoma, among other conditions like hypertensive disorder, osteoporosis, and arteriosclerotic heart disease.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08177",
"image_path": "slo_fundus_08177.jpg",
"filename": "data_08177.npz",
"report": "The clinical note does not provide specific details about the patient's condition, including the presence or absence of glaucoma.",
"age": 35.49,
"gender": "male",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "spent a total of DATE_TIME personally preparing and caring for this patient (face-to-face and non face-to-face); formulating and finalizing the note, a copy of which was given to the patient DATE_TIME.]",
"gpt4_summary": "The clinical note does not provide specific details about the patient's condition, including the presence or absence of glaucoma.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08179",
"image_path": "slo_fundus_08179.jpg",
"filename": "data_08179.npz",
"report": "54 y.o. female with moderate myopia and bifocals; no significant change in prescription; has two small, flat, non-atypical choroidal nevus; borderline intraocular pressure (IOP) of 22/19; healthy optical nerves; mother uses drops for high IOP. Glaucoma not confirmed.",
"age": 54.65,
"gender": "female",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "single",
"note": "54 y.o. f returns for routine DATE_TIME eye exam # refractive error ou -long standing moderate myopia now with PERSON changes; noting she needs to take glasses off for near work - doing well with current bifocals, no significant change in rx # choroidal nevus x2 os -flat, small size, no atypical features >continue to observe # borderline iop - cct 540/535 - tmax 22/24 - iop 22/19 - optic nerves look healthy - mother uses drops for high iop - hvf DATE_TIME od full os nonspecific defects DATE_TIME od full os full - oct DATE_TIME 81/80 od thin sup os PERSON -- ave wnl DATE_TIME 82/83 od borderline thin sup/inf os wnl > iop borderline at 22/19 ou. observe. photos DATE_TIME follow up DATE_TIME, mrx, dilate. photos",
"gpt4_summary": "54 y.o. female with moderate myopia and bifocals; no significant change in prescription; has two small, flat, non-atypical choroidal nevus; borderline intraocular pressure (IOP) of 22/19; healthy optical nerves; mother uses drops for high IOP. Glaucoma not confirmed.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08183",
"image_path": "slo_fundus_08183.jpg",
"filename": "data_08183.npz",
"report": "Patient has right optic neuropathy, likely traumatic, migraines with aura, intermittent vertigo possibly linked to migraines and disease. Shows signs of conjunctivitis and dry eye syndrome. No glaucoma mentioned.",
"age": 57.18,
"gender": "female",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "the absence of a relative afferent pupillary defect or dyschromatopsia. relatively mild PERSON thinning is noted on DATE_TIME, though this finding was not seen on her recent outside oct studies, and appears disproportionate to the density of her field loss. subtle temporal pallor is noted in the right optic disc, however. with respect to the patient's headaches and dizziness, her history of intermittent visual obscuration od with retroorbital discomfort and photophobia suggests a component of underlying migraines with aura. consideration may be given to LOCATION's disease in light of her aural fullness, tinnitus, and paroxysmal vertigo, as also previously noted by PERSON in otolaryngology, though she has notably had limited symptomatic relief with appropriate therapy; her imaging has reassuringly not revealed signs of superior canal dehiscence. regardless, she may benefit from further evaluation by neuro-otology for additional diagnostic testing. we also recommended evaluation by her dentist for consideration of a bite guard in light of her tmj dysfunction and persistent headaches, as well as updated refraction, particularly for intermediate distance vision. impression: 1. right optic neuropathy, likely traumatic 2. migraines with aura 3. intermittent vertigo, likely multifactorial secondary to #2 and ?PERSON's disease 4. PERSON dysfunction 5. dry eye syndrome 6. follicular conjunctivits recommendations: 1. dental evaluation for bite guard for suspected tmj dysfunction 2. consider referral to neuro-otology for evaluation of episodic vertigo 3. recommend updated refraction 4. follow-up neuro-ophthalmic examination in DATE_TIME (this note was prepared with the assistance of PERSON, md, neuro-ophthalmology fellow.) PERSON, PERSON neuro-ophthalmology LOCATION PHONE_NUMBER fax PHONE_NUMBER",
"gpt4_summary": "Patient has right optic neuropathy, likely traumatic, migraines with aura, intermittent vertigo possibly linked to migraines and disease. Shows signs of conjunctivitis and dry eye syndrome. No glaucoma mentioned.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08186",
"image_path": "slo_fundus_08186.jpg",
"filename": "data_08186.npz",
"report": "Patient with glaucoma experienced bilateral vision loss due to optic disc edema. Leptomeningeal disease from breast cancer caused optic neuropathy. Brain radiation therapy improved vision OS. Note recommends tapering prednisone.",
"age": 67.33,
"gender": "female",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "pseudophakia ou, and glaucoma, who developed bilateral vision loss od>os with optic disc edema characterized by hemorrhage and cotton wool spots. she has been found to have leptomeningeal disease from breast cancer, which can cause optic neuropathy related to papilledema or ischemia from infiltration of the subarachnoid space around the anterior optic nerve. steroids initially helped os but the vision progressed in that eye. she has now received whole brain radiation therapy (30 gy) and notes improvement in the brightness of vision os. the exam shows acuities of nlp od and 20/25 os. the right optic nerve has developed atrophy and pallor already, suggesting ischemic injury related to leptomeningeal disease as opposed to papilledema or to optic neuritis. the left optic nerve has improved to 20/25 with radiation therapy, showing inferior greater than superior visual field constriction. i recommended tapering the prednisone over DATE_TIME as she is having side effects and does not have optic neuritis. i will test her vision again in DATE_TIME to monitor her progress. we hope that the radiation and the new chemotherapy will help to maintain her vision in the left eye. \u00ff impression: 1. bilateral optic neuropathy os>od from leptomeningeal breast carcinoma recommendations: 1. taper prednisone: 50 mg x DATE_TIME, 40 mg x DATE_TIME, 20 mg x DATE_TIME, 10 mg x DATE_TIME, skip DATE_TIME, 10 mg x DATE_TIME then off. 2. return in DATE_TIME it is a pleasure to participate in the care of this patient. please do not hesitate to call with questions. ? PERSON, PERSON: greater than half of this DATE_TIME visit was spent counseling the patient on the medical condition or coordinating care.",
"gpt4_summary": "Patient with glaucoma experienced bilateral vision loss due to optic disc edema. Leptomeningeal disease from breast cancer caused optic neuropathy. Brain radiation therapy improved vision OS. Note recommends tapering prednisone.",
"glaucoma": "yes",
"use": "test"
},
{
"id": "data_08187",
"image_path": "slo_fundus_08187.jpg",
"filename": "data_08187.npz",
"report": "61-year-old female chaplain, glaucoma suspect due to increased cup/disc ratio, prevailing more in left than right eye. Risk factors include family history, borderline thinning on Oct, and thin corneas. No actions needed currently.",
"age": 61.7,
"gender": "female",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "single",
"note": "61 f, PERSON chaplain for f/u glaucoma testing. # glaucoma suspect with increased cup/disc ratio, left > right eye [ fhx: father [ cct: 504,505 [ oct DATE_TIME: full od, stable suptemp thinning os [ hvf DATE_TIME: full ou - discussed several risk factors including family history, borderline thinning on oct, and thin corneas. - no intervention required at this time. continue to monitor, but recommended f/u testing in DATE_TIME. rtc DATE_TIME, sooner prn previously: # dry eye disease, both eyes, mild. - reviewed artificial tears ou qid prn. # mild vitreomacular traction, right eye # minimal epiretinal membrane, left eye # posterior vitreous detachment, left eye, with symptomatic floaters. - monitor # cataract, both eyes. minimal and not visually significant in both eyes. - continue to monitor. # refractive error, would like to update glasses. - new rx given to patient.",
"gpt4_summary": "61-year-old female chaplain, glaucoma suspect due to increased cup/disc ratio, prevailing more in left than right eye. Risk factors include family history, borderline thinning on Oct, and thin corneas. No actions needed currently.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08188",
"image_path": "slo_fundus_08188.jpg",
"filename": "data_08188.npz",
"report": "Patient with open angle glaucoma, cataract in right eye and asymptomatic dry eye. Macular degeneration and epiretinal membrane present in both eyes. Elevated intraocular pressure (IOP) noted. Good response to SLT treatment.\n",
"age": 84.4,
"gender": "male",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "NRP w/ open angle glaucoma ou (od>os), cataract od, pciol os s/p yag ? 1. open angle glaucoma ou: (pt used to see dr. PERSON), lumigan too expensive - cct 587/591 - tmax 27/23 - vf 12/10 possibly worse od, better in DATE_TIME, worse again in 6/12, possibly worse od in 10/13, worse od in DATE_TIME. dp's done DATE_TIME. gdx in DATE_TIME unable od, stable os. oct done in 5/13, worse od in 6/14 s/p slt od DATE_TIME w good response 18 -> 14 s/p slt od in DATE_TIME w good response 24 -> 14 pt stopped using drops, elev iop in 10/15, pt stopped using drops again in 5/16 s/p slt os in DATE_TIME w good response 23 -> 14 - goal iop low teens od, mid teens os. - at goal DATE_TIME ? plan: hold xal ou qhs, c/w cos od bid - monitor od w hvf DATE_TIME. may need additional PERSON if vf worsens further. - rtc in DATE_TIME for iop, hvf and dp's. ? 2. s/p ce/ pciol od DATE_TIME - doing ok. elev iop and blurry vision on pod1, resolved. - pt aware of limited visual potential 2/2 erm od plan: mrx od given in 7/10 ? 3. macular degeneration and erm ou: s/p 23g ppv LOCATION mp ?od DATE_TIME plan: per dr. PERSON. ? 4. pciol + pco os s/p yag DATE_TIME. b&l li61u, 20.5d. - much improved vision ? 5. dry eye - mildly symptomatic. rec lubrication especially qhs to help with am symptoms. ? other: pt is a friend of the ceo of LOCATION, constructing the 3rd building for LOCATION. now retired. pt in LOCATION until DATE_TIME. pt has throat cancer.",
"gpt4_summary": "Patient with open angle glaucoma, cataract in right eye and asymptomatic dry eye. Macular degeneration and epiretinal membrane present in both eyes. Elevated intraocular pressure (IOP) noted. Good response to SLT treatment.\n",
"glaucoma": "yes",
"use": "test"
},
{
"id": "data_08193",
"image_path": "slo_fundus_08193.jpg",
"filename": "data_08193.npz",
"report": "39-year-old white non-Hispanic female with no diagnosis of glaucoma. Office requests patient to schedule an appointment.",
"age": 39.39,
"gender": "female",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "single",
"note": "a 39 y.o. white, non-hispanic female with no diagnosis of glaucoma. lvm requesting pt call the office to schedule an LOCATION ou, callback numb er provided.",
"gpt4_summary": "39-year-old white non-Hispanic female with no diagnosis of glaucoma. Office requests patient to schedule an appointment.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08195",
"image_path": "slo_fundus_08195.jpg",
"filename": "data_08195.npz",
"report": "Patient on multiple meds including eye drops. Conditions: hypertensive disorder, hypercholesterolemia, allergic rhinitis, nasal congestion. No glaucoma mentioned. No recent immunizations.",
"age": 75.32,
"gender": "male",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "irbesartan (avapro oral) take 300 mg by mouth DATE_TIME. loratadine oral take by mouth. naproxen sodium (aleve) 220 mg tablet take 220 mg by mouth 2 (two) times a day as needed. polymyxin b sulf-trimethoprim (polytrim) 10,000 unit- 1 mg/ml drop place 1 drop into the left eye 4 (four) times a day. prednisolone acetate (pred LOCATION) 1 % ophthalmic suspension place 1 drop into the left eye 4 (four) times a day. terazosin hcl (terazosin oral) take 3 mg by mouth DATE_TIME. condition list as of DATE_TIME hypertensive disorder hypercholesterolemia seasonal allergic rhinitis nasal congestion results summary immunizations administered on date of encounter - DATE_TIME none",
"gpt4_summary": "Patient on multiple meds including eye drops. Conditions: hypertensive disorder, hypercholesterolemia, allergic rhinitis, nasal congestion. No glaucoma mentioned. No recent immunizations.",
"glaucoma": "yes",
"use": "test"
},
{
"id": "data_08198",
"image_path": "slo_fundus_08198.jpg",
"filename": "data_08198.npz",
"report": "The patient is scheduled for intraocular pressure (IOP) check and ocular coherence tomography (OCT) RNFL/GCC tests, likely for glaucoma monitoring. Tropicamide usage approved.",
"age": 70.23,
"gender": "female",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "patient's request on DATE_TIME. -rtc in DATE_TIME with iop check and oct rnfl/gcc ou (ok to use tropicamide 0.5% to obtain good view), sooner prn. the information above was documented by PERSON as a scribe for PERSON on DATE_TIME.",
"gpt4_summary": "The patient is scheduled for intraocular pressure (IOP) check and ocular coherence tomography (OCT) RNFL/GCC tests, likely for glaucoma monitoring. Tropicamide usage approved.",
"glaucoma": "yes",
"use": "test"
},
{
"id": "data_08202",
"image_path": "slo_fundus_08202.jpg",
"filename": "data_08202.npz",
"report": "The 45-year-old male patient has ocular hypertension and is possibly at a pre-perimetric glaucoma stage. He is not regular with his timolol eye drops. There's also allergic conjunctivitis, dry eyes, and refractive error.",
"age": 45.9,
"gender": "male",
"race": "black",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "45 y.o. man presents for follow up 1. ocular htn: has not been compliant with drops recently (timolol) and taking them at DATE_TIME instead of the morning. humphrey visual field full ou, optical coherence tomography shows thinning ou DATE_TIME that is stable. possibly pre-perimetric glaucoma. pressure DATE_TIME is too high. -continue timolol qd ou in am every day -if pressures creep into DATE_TIME, then would add xalatan ou qhs -return for pressure check in 4-5 mos dfe: 3/20 gonio: DATE_TIME tmax: 28/32 cct: 534, 534 hvf: 3/20 oct: 3/20 famhx: no 2. allergic conjunctivitis ou: stable; having some dryness too; occasionally gets pingueculitis -better on zaditor -artificial tears prn \u00ff 3. refractive error: wants a new rx for glasses - old ones are scratched -give new rx for glasses",
"gpt4_summary": "The 45-year-old male patient has ocular hypertension and is possibly at a pre-perimetric glaucoma stage. He is not regular with his timolol eye drops. There's also allergic conjunctivitis, dry eyes, and refractive error.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08204",
"image_path": "slo_fundus_08204.jpg",
"filename": "data_08204.npz",
"report": "Ms. PERSON, a retail manager, has decreased peripheral vision and a history of a stable arachnoid cyst in pituitary area. She's suspected to have optic neuropathy due to cyst. No glaucoma mentioned.",
"age": 42.94,
"gender": "female",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "formulation: ms. PERSON is a retail store manager here for evaluation of her decreased peripheral vision. she has history of arachnoid cyst in pituitary area that is being follow and has been stable in size (last mri DATE_TIME). on the exam DATE_TIME, she had mild temporal pallor of her optic nerves bilaterally. the visual field results were not reliable. there was loss of ganglionic cell layer in macula on oct. thus, there appears to be an optic neuropathy, which could the result of compression from the arachnoid cyst. we will follow her in DATE_TIME and re-evaluate her to make sure the findings are not progressive. impression: 1. possible compression optic neuropathy from pituitary arachnoid cyst plan: 1. follow up in DATE_TIME. 2. consider repeat oct during re-evaluation PERSON neuro-ophthalmology LOCATION PHONE_NUMBER fax PHONE_NUMBER",
"gpt4_summary": "Ms. PERSON, a retail manager, has decreased peripheral vision and a history of a stable arachnoid cyst in pituitary area. She's suspected to have optic neuropathy due to cyst. No glaucoma mentioned.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08208",
"image_path": "slo_fundus_08208.jpg",
"filename": "data_08208.npz",
"report": "The patient is a glaucoma suspect based on increased c/d ratio and family history. Mild cataract is present but not visually significant. Patient's IOP is controlled. A new glasses prescription was given.",
"age": 63.49,
"gender": "male",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "1. glaucoma suspect based on inc c:d ratio ou fhx + mom and mgm with glaucoma hvf od: reliable, mild ins defect (fluctuates in location but no worse in severity) hvf os reliable and full oct borderline os, PERSON dp stable gonio open ou cct average 546,548 iop controlled, observe 2. refractive error: a prescription for new glasses was given to the patient DATE_TIME. 3. mild cataract is present ou that is not visually significant. observation at this time was recommended. DATE_TIME, mrx, iop, hvf, LOCATION, oct and dp",
"gpt4_summary": "The patient is a glaucoma suspect based on increased c/d ratio and family history. Mild cataract is present but not visually significant. Patient's IOP is controlled. A new glasses prescription was given.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08212",
"image_path": "slo_fundus_08212.jpg",
"filename": "data_08212.npz",
"report": "65 y.o. male, glaucoma suspect, with no changes in vision since last visit. Also has bipolar disorder, hyperlipidemia, obesity, hypertension, and is a former smoker. No pain or discomfort in eyes. Stable thinning, no significant cataracts or vision issues. 'Sunken eyes' noticed, possibly due to dermatochalasis.",
"age": 65.15,
"gender": "male",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "divorced",
"note": "65 y.o. male presents for glaucoma suspect follow-up last visit with me: DATE_TIME pohx: glaucoma at last eye exam, feels eyes 'sunken in' and family notes it too pmhx: bipolar disorder, hyperlipidemia, obesity, hypertension, cad, former smoker hpi DATE_TIME: pt here for glaucoma suspect ou pt denies any visual changes since last visit pt denies any new flashes or floaters ou, no pain, ou feel comfortable ocular medications: none assessment/plan: # glaucoma suspect - glaucoma suspect based on increased cup/disc - iop - 12/11 - tmax - 17/17 here DATE_TIME - cct - 520/520 - thin ou - c/d - 0.8. 0.75 - gonio - refused at last visit - family history - none per patient - last hvf performed - DATE_TIME - full ou, DATE_TIME - full ou - last oct rnfl performed - DATE_TIME - stable LOCATION thinning od > os - DATE_TIME - non-specific LOCATION thinning, ou, large disc area ou - baseline fundus photos obtained - DATE_TIME - dilated optic nerve exam performed ou DATE_TIME - discussed potential for progressive irreversible vision loss due to glaucomatous optic neuropathy. - return in DATE_TIME with hvf 24-2, iop check # cataract, not visually significant # myopia with astigmatism and presbyopia - mild cataract is present that is not visually significant. - bcva 20/20, 20/20 - observation of cataracts at this time was recommended. - will try new prescription for glasses and if continuing to bother patient will consider cataract extraction surgery # dermatochalasis/steatoblepharon - family is telling him to have 'sunken eyes', likely from relative positioning of brow and dermatochalasis, no concerning orbital signs - symmetric hertel's DATE_TIME - not visually significant - observe rtc for glaucoma testing in DATE_TIME with hvf 24-2, iop check, refract/dilate if allows PERSON, md",
"gpt4_summary": "65 y.o. male, glaucoma suspect, with no changes in vision since last visit. Also has bipolar disorder, hyperlipidemia, obesity, hypertension, and is a former smoker. No pain or discomfort in eyes. Stable thinning, no significant cataracts or vision issues. 'Sunken eyes' noticed, possibly due to dermatochalasis.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08213",
"image_path": "slo_fundus_08213.jpg",
"filename": "data_08213.npz",
"report": "Patient suspected of glaucoma. Normal optic nerve, visual fields. Tmax at 22, low corneal hysteresis (8.8/8.6), central corneal thickness 520/515. No procedures/medication history. Slightly large nerves, intraocular pressure around 20.",
"age": 56.41,
"gender": "male",
"race": "black",
"ethnicity": "unknown",
"language": "english",
"maritalstatus": "married or partnered",
"note": "first seen by dr. PERSON on DATE_TIME diagnosis: glaucoma suspect target iop: / , tmax: 22 (DATE_TIME) / 22 (DATE_TIME) central corneal thickness: 525, 527, 529 / 514, 515, 516 corneal hysteresis: 8.8/8.6 gonioscopy: scleral spur, light pigment refractive error: od . . / os . . optic nerve findings on initial visit right eye: normal optic nerve findings on initial visit left eye: normal visual fields on initial visit right eye: normal visual fields on initial visit left eye: normal medication history and intolerances at first visit: none glaucoma procedures right eye: glaucoma procedures left eye: other eye procedures right eye: other eye procedures left eye: other eye problems right eye: other eye problems left eye: family history: mother and younger sister (vision loss at young age) steroids: no trauma: no asthma: no other medical history and problems: none initial note: large cup:disc ratio, slightly large nerves, intraocular pressure around 20, mother with glaucoma and good vision, low corneal hysteresis (8.8 and 8.6) and central corneal thickness 520/515). could consider selective laser trabeculoplasty sooner than later. plan: slightly large nerves, large cups, borderline iop; check in DATE_TIME with testing and dilation. PERSON scribing for dr. PERSON (DATE_TIME, DATE_TIME) i was present while this encounter was recorded and agree that the information entered by my scribe is complete and accurate.",
"gpt4_summary": "Patient suspected of glaucoma. Normal optic nerve, visual fields. Tmax at 22, low corneal hysteresis (8.8/8.6), central corneal thickness 520/515. No procedures/medication history. Slightly large nerves, intraocular pressure around 20.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08217",
"image_path": "slo_fundus_08217.jpg",
"filename": "data_08217.npz",
"report": "The patient has posterior optic neuropathy likely due to trauma, pellucid marginal degeneration, cataract, myopia with astigmatism, and presbyopia. No signs of glaucoma mentioned.",
"age": 56.45,
"gender": "male",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "single",
"note": "foveal sensitivity and peripheral defects, consistent with posterior optic neuropathy (LOCATION), and given the onset after the fall this is likely traumatic - recommend patient wear polycarbonate lenses for protection - recommend he follow-up with neuro-op service for evaluation and information on prognosis although prognosis is guarded and vision will not likely improve significantly - recommend return to me in DATE_TIME with hvf 24-2 and oct rnfl, no dilation # headaches - no ocular cause of headaches seen on exam DATE_TIME - can follow-up with neurology dr. PERSON as scheduled DATE_TIME - has had concussion in the past, now feels like he has a migraine all day and night, having difficulty sleeping - no recommendations given DATE_TIME to patient, pain is not around eye but behind left temple # pellucid marginal degeneration os > od - visual acuity was 20/40 previously with glasses and 20/50 with contacts os - sees well with PERSON or glasses - recommend polycarbonate lenses as above # cataract, not visually significant # myopia with astigmatism and presbyopia (high astigmatism from pmd as above) - mild cataract is present that is not visually significant. - bcva 20/20, DATE_TIME/cf DATE_TIME - observation of cataracts at this time was recommended. rtc to neuro-op - appointment to be scheduled, referral placed DATE_TIME rtc to me in DATE_TIME with hvf 24-2 and oct rnfl, no dilation, can refract if pinhole improves os PERSON, md",
"gpt4_summary": "The patient has posterior optic neuropathy likely due to trauma, pellucid marginal degeneration, cataract, myopia with astigmatism, and presbyopia. No signs of glaucoma mentioned.",
"glaucoma": "yes",
"use": "test"
},
{
"id": "data_08223",
"image_path": "slo_fundus_08223.jpg",
"filename": "data_08223.npz",
"report": "The patient is suspected of glaucoma based on visual field. Initial tests show mostly healthy nerves, thinning on the right eye's optic nerve, inferior nasal step on right eye, but all normal for left eye. No medications being used.",
"age": 73.58,
"gender": "male",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "first seen by PERSON PERSON on DATE_TIME diagnosis: glaucoma suspect based on humphrey visual field; referred by dr. PERSON target iop: / , tmax: ( ) / ( ) central corneal thickness: 530 / 531 gonioscopy: cbb refractive error: LOCATION . -0.75 . 100 / os +3.50 . -1.00 . 087 optic nerve findings on initial visit right eye: superior thinning optic nerve findings on initial visit left eye: normal visual fields on initial visit right eye: inferior nasal step visual fields on initial visit left eye: normal medications being used at first visit: none medication intolerances:none glaucoma procedures right eye: none glaucoma procedures left eye: none other eye procedures right eye: none other eye procedures left eye: none other eye problems right eye: none other eye problems left eye: none family history: none steroids: intermittent (not for a long time) trauma: none asthma: intermittent (not active-- pt suspicious of prior diagnosis) other medical history and problems: cabg x 4, cardiac stent initial note: referred by PERSON for glaucoma suspect both eyes. prior testing reveals superior thinning and an inferior nasal step on humphrey visual field right eye despite the nerve appearing mostly healthy and intraocular pressure 10 on no medications. all testing normal left eye. right eye with inferior nasal step and corresponding thinning on optical coherence tomography. however the pressures are lowand the nerves look acceptable both eyes (though small and sl tilt). unclear if this defect is chronic/congenital or new. patient to bring in old records. monitor off drops. plan: continue to monitor off drops. patient to bring old records of testing. repeat optical coherence tomography and humphrey visual field in DATE_TIME.",
"gpt4_summary": "The patient is suspected of glaucoma based on visual field. Initial tests show mostly healthy nerves, thinning on the right eye's optic nerve, inferior nasal step on right eye, but all normal for left eye. No medications being used.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08224",
"image_path": "slo_fundus_08224.jpg",
"filename": "data_08224.npz",
"report": "Patient has choroidal melanoma in right eye and severe glaucoma in left eye. Right eye's intraocular pressure is 30, left eye's is 27/12. Surgery recommended. Currently using diamox.",
"age": 86.01,
"gender": "male",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "first seen by dr. PERSON on DATE_TIME, aronow patient diagnosis: choroidal melanoma right eye with intraocular pressure 27/12 target iop: / , tmax: ( ) / ( ) central corneal thickness: 485 / 653 gonioscopy: refractive error: PERSON. -3.25. 100 / os +0.50. DATE_TIME. 035 optic nerve findings on initial visit right eye: optic nerve findings on initial visit left eye: thin throughout visual fields on initial visit right eye: visual fields on initial visit left eye: superior arcuate and inferior arcuate dense medication history and intolerances at first visit: PERSON, combigan latanoprost diamox glaucoma procedures right eye: glaucoma procedures left eye: other eye procedures right eye: cataract extraction other eye procedures left eye: cataract extraction, dsek other eye problems right eye: other eye problems left eye: family history: steroids: trauma: asthma: other medical history and problems: plan: severe glaucoma left eye but intraocular pressure is low and monitored in LOCATION. right eye intraocular pressure is 30, but appears to have active melanoma. seeing aronow DATE_TIME who recommends enucleation. tolerating diamox and he can decide if he would like to continue or stop until surgery. intraocular pressure left eye target should be low",
"gpt4_summary": "Patient has choroidal melanoma in right eye and severe glaucoma in left eye. Right eye's intraocular pressure is 30, left eye's is 27/12. Surgery recommended. Currently using diamox.",
"glaucoma": "yes",
"use": "test"
},
{
"id": "data_08226",
"image_path": "slo_fundus_08226.jpg",
"filename": "data_08226.npz",
"report": "Patient has eye pressure in right and left. Glaucoma indicated through cupped out c/d ratios of 0.85 and 1.0. Restriction and deficiencies in visual fields.",
"age": 83.98,
"gender": "male",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "single",
"note": "pressure 19 07 tonometry #2 (ta ub, DATE_TIME) right left pressure DATE_TIME gonioscopy right left superior c/d 35f c/d 35f, pas inferior c/d 35f c/d 35f pupils pupils right perrl left LOCATION visual fields left right restrictions partial superior temporal, inferior temporal, superior nasal, inferior nasal deficiencies partial superior temporal, inferior temporal, superior nasal, inferior nasal deficiencies extraocular movement right left result full, ortho full, ortho neuro/psych oriented x3: yes mood/affect: normal slit lamp and fundus exam slit lamp exam right left lids/lashes ll lagophthalmos ll lagophthalmos conjunctiva/sclera normal tube and plate well covered, trabeculectomy LOCATION with quiet bleb, trace microcysts cornea dry, 1+ diffuse pee dry pee anterior chamber normal quiet, tube in good position in LOCATION PERSON normal, pxf material, tid at margin LOCATION, ik touch superotemporal, peripheral LOCATION, tid lens posterior chamber intraocular lens, pseudoexfoliation posterior chamber intraocular lens, pseudoexfoliation vitreous normal normal fundus exam right left disc thinning inferiorly, ppa ppa, cupped out c/d ratio 0.85 1.0 macula normal normal refraction wearing rx sphere cylinder axis right -PHONE_NUMBERE left -0.50 -0.75 73",
"gpt4_summary": "Patient has eye pressure in right and left. Glaucoma indicated through cupped out c/d ratios of 0.85 and 1.0. Restriction and deficiencies in visual fields.",
"glaucoma": "yes",
"use": "test"
},
{
"id": "data_08227",
"image_path": "slo_fundus_08227.jpg",
"filename": "data_08227.npz",
"report": "The patient has persistently high intraocular pressure which suggests they have glaucoma. Risks of laser treatment were discussed. If pressure remains high, possible treatments include SLT or Trab 360.",
"age": 31.79,
"gender": "male",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "single",
"note": "needed. -contact lens hygiene was reviewed on DATE_TIME. -long discussion with patient on DATE_TIME: given iop persistently above goal od, we will proceed with LOCATION. risks, benefits, and alternatives to laser were discussed with the patient, including the potential risk of infection, bleeding, loss of vision, loss of eye, double vision, need for further surgery or laser, retinal detachment, change in glasses and reading glasses. no guarantees given. questions answered. consents signed. *** asa/blood thinners *** increased risk of complications due to implants/special equipment needed -- *** preferred anesthesia -- *** diabetic -- no -rtc on DATE_TIME with iop check, hvf, and oct rnfl/gcc ou, sooner prn. if iop still above goal od next visit, we can consider booking slt od (if iop above goal but hvf stable) versus trab 360 od (if iop above goal and hvf significantly progressed versus adding rhopressa qhs ou. the information above was documented by PERSON as a scribe for PERSON on DATE_TIME.",
"gpt4_summary": "The patient has persistently high intraocular pressure which suggests they have glaucoma. Risks of laser treatment were discussed. If pressure remains high, possible treatments include SLT or Trab 360.",
"glaucoma": "yes",
"use": "test"
},
{
"id": "data_08228",
"image_path": "slo_fundus_08228.jpg",
"filename": "data_08228.npz",
"report": "Retired physician has undergone surgery for elevated intraocular pressure (IOP); goals of <=10mmHg in the right eye and <=9mmHg in left eye. Currently off glaucoma meds. Further monitoring planned.",
"age": 84.96,
"gender": "male",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "a retired physician. attending's plan: -goal intraocular pressure less than or equal to 10 mmhg, right eye. -goal intraocular pressure less than or equal to 09 mmhg, left eye. -iop *** goal od and *** goal os on DATE_TIME s/p phaco/ecp/kdb ou and off glaucoma medications. -continue close monitoring off glaucoma medications. -preservative-free artificial tears as needed. -patient admits to snoring sporadically (?possible ltg component given progression at normal PERSON), but his wife denied it is consistent. if further progression, will re-visit DATE_TIME. -good phaco/migs candidate when patient ready, but currently patient happy with vision. -long discussion with patient on DATE_TIME: we discussed surgical intervention given iop slightly above goal os and new disc hemorrhage os inferiorly with iop of 10 mmhg; he would be a good candidate for phaco/trab os first versus phaco/ecp/kdb os. after extensively discussing pros/cons of each procedure, we proceeded with phaco/ecp/kdb os (attending case) on DATE_TIME knowing further intervention may be needed in the future. -long discussion with patient on DATE_TIME: given good response os and wish for best vision od, we proceeded with phaco/ecp/kdb od on DATE_TIME. -rtc in DATE_TIME for iop check, hvf, dilation, and oct rnfl/gcc ou, sooner prn. we'll attempt zioptan qhs ou in future if iop above goal and/or ocudose after surgery if iop control needed (and we'll try to spare him rhopressa). the information above was documented by PERSON as a scribe for PERSON on DATE_TIME.",
"gpt4_summary": "Retired physician has undergone surgery for elevated intraocular pressure (IOP); goals of <=10mmHg in the right eye and <=9mmHg in left eye. Currently off glaucoma meds. Further monitoring planned.",
"glaucoma": "yes",
"use": "test"
},
{
"id": "data_08229",
"image_path": "slo_fundus_08229.jpg",
"filename": "data_08229.npz",
"report": "The patient has been diagnosed with normal tension glaucoma, moderate in the right eye, and mild in the left eye. No prior glaucoma surgery recorded. The patient needs intraocular pressure of approximately 12 in both eyes.",
"age": 54.53,
"gender": "male",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "# first seen by PERSON on DATE_TIME diagnosis: normal tension glaucoma, moderate right eye and mild left eye ttarget: / , tmax: ( ) / ( ) cct: / gonioscopy: open both eyes refractive error: od -3.00. -0.75. 168 / os -3.75. sphere. optic nerve: cup:disc ratio 0.85 both eyes with disc heme in both eyes rnfl oct: borderline thinning both eyes (average retinal nerve fiber layer 76 um right eye and 81 um left eye) vf: inferior arcuate right eye (md -11.82 db), inferior paracentral defect left eye (md -4.42 db) med intolerances: on latanoprost prior glaucoma surgery: none od / none os other eye surgery: none od/ none PERSON: ? uncle (went blind)/ steroids: none/ asthma: none/ trauma: none # pvd right eye -shafer neg - retinal detachment precautions reviewed pmhx: proteinuria (unclear etiology, on PERSON), osa (on cpap) plan: primary open angle glaucoma both eyes drance heme DATE_TIME needs intraocular pressure close to 12 both eyes continue ltn. also add timolol bid rtc DATE_TIME intraocular pressure check, dilation, dp and repeat humphrey visual field diagnosis of glaucoma/glaucoma suspect discussed with patient - review of natural history and management options discussed. importance of lifelong follow up and adherence to treatments discussed in order to lower risk of permanent vision loss/blindness from glaucoma. all questions answered. i spent DATE_TIME with the patient, more than half of which was spent on counseling and coordination of care for this patient's glaucoma",
"gpt4_summary": "The patient has been diagnosed with normal tension glaucoma, moderate in the right eye, and mild in the left eye. No prior glaucoma surgery recorded. The patient needs intraocular pressure of approximately 12 in both eyes.",
"glaucoma": "yes",
"use": "test"
},
{
"id": "data_08232",
"image_path": "slo_fundus_08232.jpg",
"filename": "data_08232.npz",
"report": "Patient shows mild refractive change, posterior vitreous detachment, cataract, and is glaucoma suspect (due to cup asymmetry) with stable pressure. No family history or symptoms of glaucoma.",
"age": 69.11,
"gender": "male",
"race": "white",
"ethnicity": "unknown",
"language": "english",
"maritalstatus": "married or partnered",
"note": "last seen by me DATE_TIME) refractive: mild change - give new rx for glasses 2) posterior vitreous detachment od: seen in past by vavvas with no breaks found. stable 3) glaucoma suspect (cupping os>od): pressures have been fine. the visual fields have been reassuring but DATE_TIME are low quality os. the optical coherence tomography shows larger nerve os with cupping but no thinning. pachymetry is slightly thin. there is no family hx of glaucoma. i think this is physiologic cup asymmetry and is not worrisome. - repeat humphrey visual field alone in DATE_TIME to be sure - otherwise return DATE_TIME for dilation, humphrey visual field and optical coherence tomography - needs gonio next time dfe: 11/21 vf: 11/21 DATE_TIME gonio: tmax: 15, 15 cct: 513, 530 fhx: no 4) cataracts ou: mild with no symptoms. DATE_TIME driving is fine. -follow 5) PERSON's os?: not seen DATE_TIME on exam. thought to be from carotid dissection DATE_TIME. -follow",
"gpt4_summary": "Patient shows mild refractive change, posterior vitreous detachment, cataract, and is glaucoma suspect (due to cup asymmetry) with stable pressure. No family history or symptoms of glaucoma.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08233",
"image_path": "slo_fundus_08233.jpg",
"filename": "data_08233.npz",
"report": "Patient has cataract in left eye, glaucoma in both eyes controlled with Xalatan. Also has age-related macular degeneration (AMD) in both eyes and sup arcuate ou.",
"age": 75.47,
"gender": "female",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "imp: pciol od cataract os poag ou; PERSON; iop controlled with xalatan; unremarkable oct of rnfl, but hvf --first here DATE_TIME--suggests sup arcuate ou amd ou plan: 4 mo dilated exam/refraction plus oct of rnfl and PERSON; hvf repeat",
"gpt4_summary": "Patient has cataract in left eye, glaucoma in both eyes controlled with Xalatan. Also has age-related macular degeneration (AMD) in both eyes and sup arcuate ou.",
"glaucoma": "yes",
"use": "test"
},
{
"id": "data_08234",
"image_path": "slo_fundus_08234.jpg",
"filename": "data_08234.npz",
"report": "The patient has visual field loss but stable OCT, indicating close monitoring. Mild guttata are stable. History of retinal person with adjacent fluid and demarcation line. Dry eyes present.",
"age": 76.48,
"gender": "female",
"race": "black",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "visual field with sn losses, but oct stable. continue to monitor closely. \u00ff\u00ff 5. mild guttata ou -stable cct DATE_TIME: 532/535 \u00ff\u00ff 6. hx retinal PERSON, with small cuff of fluid surrounding, and adjacent demarcation line just anterior to this. asx >> rd precautions reviewed 7. dry eyes ou - continue artificial tears 3-4 times DATE_TIME - start artificial tear ointment at bedtime - warm compresses",
"gpt4_summary": "The patient has visual field loss but stable OCT, indicating close monitoring. Mild guttata are stable. History of retinal person with adjacent fluid and demarcation line. Dry eyes present.",
"glaucoma": "yes",
"use": "test"
},
{
"id": "data_08236",
"image_path": "slo_fundus_08236.jpg",
"filename": "data_08236.npz",
"report": "The 49 y.o. female patient, with history of Graves disease and breast cancer, has no signs of optic neuropathy or glaucoma. She reported diplopia when tired, and has dermatochalasis, potentially needing surgery.",
"age": 50.01,
"gender": "female",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "49 y.o. female presents for a new patient visit referred by endocrinologist PERSON PERSON chang, as she is considering undergoing radioactive iodine therapy. she has a past medical history of graves disease and breast cancer s/p bilateral mastectomy 1. graves disease diagnosed DATE_TIME with mild graves PERSON methimazole, planning for possible radioactive iodine treatment disease - 20/20 ou, no dyschromatopsia - mild lid retraction/flare ou, no evidence of optic neuropathy - oct rnfl DATE_TIME: wnl ou - hvf DATE_TIME: reliable and full ou - previously tried selenium 100 mcg po bid (started DATE_TIME), but no longer taking since no benefit noted - dr. PERSON questioning whether would benefit from corticosteroid treatment before and during rai treatment - never smoker - reports vertical binocular diplopia when tired, not likely graves related - continue artificial tears prn - no signs of optic neuropathy, does not need to add corticosteroid treatment if pursuing rai 2. diplopia - patient notes vertical diplopia when tired, has had prism in her glasses since DATE_TIME - small right hypertropia noted on alternate cover, 3pd at distance - patient has a left head tilt on exam DATE_TIME and on her driver's license photo, could be due to congenital 4th nerve palsy with mild decompensation versus other etiologies - observe for now 3. dermatochalasis - patient notes fullness of upper and lower lids ou, interested in possible surgical correction - in light of mild lid retraction would recommend further evaluation with oculoplastics service in the next DATE_TIME, after her thyroid levels have stabilized rtc to oculoplastics service in DATE_TIME (after rai/thyroid labs have stabilized) to discuss lid procedures for dermatochalasis rtc DATE_TIME or sooner for routine exam with me PERSON, pgy-4 patient seen and discussed with resident. agree with assessment and plan. PERSON, md",
"gpt4_summary": "The 49 y.o. female patient, with history of Graves disease and breast cancer, has no signs of optic neuropathy or glaucoma. She reported diplopia when tired, and has dermatochalasis, potentially needing surgery.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08238",
"image_path": "slo_fundus_08238.jpg",
"filename": "data_08238.npz",
"report": "Patient has posterior vitreous detachment, dermatochalasis of upper eyelids, choroidal nevus in left eye, and dry eye syndrome. There are ongoing treatments for glaucoma and cataract.",
"age": 65.52,
"gender": "female",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "eye -follow-up with dr. PERSON for cornea care. 5. posterior vitreous detachment, both eyes -retinal detachment precautions were reviewed with the patient. DATE_TIME/o breast cancer w/ tamoxifen use, d/c DATE_TIME 7. dermatochalasis of upper eyelids, both eyes -follow-up with dr. PERSON for eyelid care. 8. choroidal nevus, left eye 9. dry eye syndrome, both eyes -preservative-free artificial tears as needed. 10. social/systemic issues: referred to clinic by dr. PERSON. patient is a retired nurse from PERSON. attending's plan: -goal intraocular pressure less than or equal to *** mmhg, right eye. -goal intraocular pressure less than or equal to *** mmhg, left eye. -iop *** goal od and *** goal os on DATE_TIME on ***. -retinal detachment precautions were reviewed with the patient. -preservative-free artificial tears as needed. -follow-up with dr. PERSON for cornea care. -follow-up with dr. PERSON for eyelid care. -rtc in DATE_TIME with iop check, ***, dilation, and disc photos ou, sooner prn. i spent DATE_TIME on care for the patient (face-to-face and non-face-to-face), more than half of which was spent on counseling and coordination of care for this patient's glaucoma and cataract. the information above was documented by PERSON as a scribe for PERSON on DATE_TIME.",
"gpt4_summary": "Patient has posterior vitreous detachment, dermatochalasis of upper eyelids, choroidal nevus in left eye, and dry eye syndrome. There are ongoing treatments for glaucoma and cataract.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08240",
"image_path": "slo_fundus_08240.jpg",
"filename": "data_08240.npz",
"report": "Patient has mild, angle recession glaucoma & amblyopia in left eye, caused by metal-induced trauma. Also, has asthma but tolerates timolol well for glaucoma. Plan to stop timolol & monitor IOP.",
"age": 44.01,
"gender": "male",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "first seen by dr. PERSON on DATE_TIME, former patient of song angle recession glaucoma left eye, mild, amblyopia left eye, ruptured globe left eye (metal in ac), DATE_TIME target iop: DATE_TIME, tmax: 20 (DATE_TIME) / 20 (DATE_TIME) central corneal thickness: 520 / 508 gonioscopy: inferior angle recession os with pas in inferotemporal quadrant as a result of trauma refractive error: od -PHONE_NUMBER / os +0.00. LOCATION. 167 optic nerve findings on initial visit right eye: healthy optic nerve findings on initial visit left eye:oct-rnfl (87/82) with inf thinning visual fields on initial visit right eye: normal visual fields on initial visit left eye: normal medication history and intolerances: timolol on visit with df (has tolerated despite asthma) glaucoma procedures right eye : glaucoma procedures left eye : other eye procedures: amblyopia left eye family history: steroids: trauma: other history and problems: asthma (uses albuterol inhaler) plan: stop timolol, iop okay. monitor in DATE_TIME.",
"gpt4_summary": "Patient has mild, angle recession glaucoma & amblyopia in left eye, caused by metal-induced trauma. Also, has asthma but tolerates timolol well for glaucoma. Plan to stop timolol & monitor IOP.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08246",
"image_path": "slo_fundus_08246.jpg",
"filename": "data_08246.npz",
"report": "Mr. PERSON has mild nonproliferative diabetic retinopathy in both eyes, central retinal vein occlusion in the right eye, and bilateral nuclear sclerotic cataract. Glaucoma presence is negative.",
"age": 58.48,
"gender": "male",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "mr. PERSON is a DATE_TIME. male with 1. both eyes affected by mild nonproliferative diabetic retinopathy with macular edema, associated with type 2 diabetes mellitus 2. central retinal vein occlusion with macular edema of right eye 3. nuclear sclerotic cataract, bilateral vision readings for this visit: va distance LOCATION distance cc va near LOCATION near cc iop right 20/20-2 ph ph 19 left 20/20 ph ph 19 referred by dr. PERSON for continued care assessment 1. sup. PERSON with PERSON scatter laser superior LOCATION. 3.20.18 s/p multiple avastin last DATE_TIME - va is stable 20/25, iop DATE_TIME DATE_TIME: me os stable compared to his previous scan - oct 11.15.19: improved edema compared to DATE_TIME. continue to monitor for now 2. mild PERSON ou - a1c: pt doesn't remember (DATE_TIME) - insulin dependent. 3. asymmetric high c/d ratio od>os - negative glaucoma family history - iop is 21/19 - saw dr. PERSON DATE_TIME who started pt on latanoprost - dr. PERSON also recommended slt - discussed with pt that LOCATION is very slightly preferable to latanoprost as latanoprost may increase edema, but both good options for iop control - emphasized need for regular f/u with dr. PERSON plan - monitor for now f/u with dr. PERSON as scheduled 2.2020 f/u with PERSON PERSON as recommended return to clinic in DATE_TIME, sooner as needed, with oct. the above note was recorded by PERSON as a scribe for PERSON demetrios PERSON, PERSON on DATE_TIME at DATE_TIME i attest that i was present while this encounter was recorded and agree that the information entered by my scribe is complete and accurate. demetrios PERSON, DATE_TIME.",
"gpt4_summary": "Mr. PERSON has mild nonproliferative diabetic retinopathy in both eyes, central retinal vein occlusion in the right eye, and bilateral nuclear sclerotic cataract. Glaucoma presence is negative.",
"glaucoma": "yes",
"use": "test"
},
{
"id": "data_08248",
"image_path": "slo_fundus_08248.jpg",
"filename": "data_08248.npz",
"report": "The note reveals visually significant issues in both eyes on certain dates, with posterior vitreous detachment in the left eye. Patient suffers from dry eye syndrome. Goal intraocular pressure is set at less than or equal to 17 mmHg in both eyes.",
"age": 64.5,
"gender": "female",
"race": "white",
"ethnicity": "unknown",
"language": "english",
"maritalstatus": "unknown",
"note": "DATE_TIME.','-potentially visually-significant, both eyes as of DATE_TIME.','-visually-significant, right eye as of DATE_TIME.','-visually-significant, left eye as of DATE_TIME.','-visually-significant, both eyes as of DATE_TIME.','-not visually-significant, right eye as of DATE_TIME.','-not visually-significant, left eye as of DATE_TIME.','-not visually-significant, both eyes as of DATE_TIME.','-stable.'} 3. posterior vitreous detachment, left eye -retinal detachment precautions were reviewed with the patient. 4. s/PERSON, both eyes 5. dry eye syndrome, both eyes -preservative-free artificial tears as needed. 6. social/systemic issues: seen formerly by dr. PERSON. attending's plan: -goal intraocular pressure {goal pressure:19197::'~','less than or equal to','=','greater than or equal to'} 17 mmhg, right eye. -goal intraocular pressure {goal pressure:19197::'~','less than or equal to','=','greater than or equal to'} 17 mmhg, left eye. -iop {goal:19197::'at','above','well above'} goal od and {goal:19197::'at','above','well above'} goal os on DATE_TIME on off glaucoma medications. DATE_TIME plus others oct",
"gpt4_summary": "The note reveals visually significant issues in both eyes on certain dates, with posterior vitreous detachment in the left eye. Patient suffers from dry eye syndrome. Goal intraocular pressure is set at less than or equal to 17 mmHg in both eyes.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08249",
"image_path": "slo_fundus_08249.jpg",
"filename": "data_08249.npz",
"report": "Patient has early-stage cataracts in both eyes with symptoms of glare. They also have choroidal nevi, more significant in the left eye. No mention of glaucoma.",
"age": 57.96,
"gender": "male",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "wears pals for better quality of vision \u00ff 5. cataracts ou - early sx of glare, ctm 6. choroidal nevi os>od cox, pgy-4 i saw and evaluated this patient and discussed the case as appropriate with the resident/fellow. i have reviewed the resident/fellow's notes and made any necessary changes.",
"gpt4_summary": "Patient has early-stage cataracts in both eyes with symptoms of glare. They also have choroidal nevi, more significant in the left eye. No mention of glaucoma.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08250",
"image_path": "slo_fundus_08250.jpg",
"filename": "data_08250.npz",
"report": "The patient had cataract surgery and yag pc, with dry macular detected by OCT. No signs of glaucoma in RNFL OU. Requires glasses prescription.",
"age": 79.61,
"gender": "female",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "widowed",
"note": "imp: s/p cataract surgery ou and yag pc os in LOCATION macular PERSON ou--dry by oct now central LOCATION sec to PERSON; normal oct of rnfl ou refr error plan: rx=m glasses",
"gpt4_summary": "The patient had cataract surgery and yag pc, with dry macular detected by OCT. No signs of glaucoma in RNFL OU. Requires glasses prescription.",
"glaucoma": "yes",
"use": "test"
},
{
"id": "data_08251",
"image_path": "slo_fundus_08251.jpg",
"filename": "data_08251.npz",
"report": "The patient has severe dry eye syndrome, is using refresh and at gel for symptoms. Monitoring for eye strain despite new prescription. Glaucoma medication might cause side effects. Recommended restasis and humidifier.",
"age": 62.96,
"gender": "female",
"race": "asian",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "periph on temporal side >> monitor # eye strain - despite new mrx recently >> consider re-eval mrx soon # dry eye - severe on exam DATE_TIME, very symptomatic - already using refresh PERSON and at gel qhs - will hold off on punctal plugs for now as may keep glaucoma meds against eye for longer and cause side effects >> rec pfat gel qid ou >> rec restasis bid ou >> rec humidifier daily # refractive error >> mrx provided previously # dry eye syndrome >> continue artificial tears, gel drop qid ou rtc 3 mo no dilate surface check PERSON, md, mph",
"gpt4_summary": "The patient has severe dry eye syndrome, is using refresh and at gel for symptoms. Monitoring for eye strain despite new prescription. Glaucoma medication might cause side effects. Recommended restasis and humidifier.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08252",
"image_path": "slo_fundus_08252.jpg",
"filename": "data_08252.npz",
"report": "78-year-old patient with history of COPD, hypertension, and hyperlipidemia has visually significant cataract in left eye, affecting daily activities. Patient has agreed to proceed with cataract surgery and potential risks including infection, bleeding, vision loss, and glaucoma were discussed. Astigmatism and ocular hypertension noted, but eye pressure excellent on latanoprost. Early signs of primary open-angle glaucoma were also observed.",
"age": 78.68,
"gender": "male",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "78 y.o. est pt, hx of LOCATION, LOCATION, PERSON, copd, htn, hyperlipidemia, mild ashtma, pituitary cyst followed by dr. PERSON wishes to see me instead as his wife is a patient of mine PERSON great, happy with result combined cataract left - visually significant and affecting activities of DATE_TIME living - dilates moderately , [+ ] flomax, [no ] pxf, [no ] guttae, [+] 81 mg asa - discussed cataract surgery, patient elected to proceed with cataract surgery with iol left eye. with istent > risks, benefits, and alternatives to surgery were discussed with the patient, including but not limited to the potential risk of infection, bleeding, loss of vision, loss of eye, need for further surgery or laser, retinal detachment, change in glasses and reading glasses, macular edema, increased eye pressure and glaucoma. no guarantees given. all patient questions were answered.? > discussed available iols including presbyopia and astigmatism correcting lenses including all out-of-pocket expenses. could benefit from toric iol in os but since he has astigmatism in the LOCATION, he will still need glasses to correct the vision. he is fine with wearing glasses so we opted to not use a toric iol in the left eye. > pt opts for monofocal iol, aim distance ?> anesthesia: peribulbar block > other concerns: omidria/epishugarcaine asteroid hyalosis os - stable hx ocular htn, possible early poag os - iop excellent on latanoprost qhs os - tmax 27 ou - cct DATE_TIME DATE_TIME wnl ou DATE_TIME PERSON - hvf DATE_TIME borderline reliable od with central defects (difficult test due to fogging of the lens despite taping his mask) os nonspecific defects DATE_TIME nonspecific defects ou > observe refractive error > no significant change",
"gpt4_summary": "78-year-old patient with history of COPD, hypertension, and hyperlipidemia has visually significant cataract in left eye, affecting daily activities. Patient has agreed to proceed with cataract surgery and potential risks including infection, bleeding, vision loss, and glaucoma were discussed. Astigmatism and ocular hypertension noted, but eye pressure excellent on latanoprost. Early signs of primary open-angle glaucoma were also observed.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08253",
"image_path": "slo_fundus_08253.jpg",
"filename": "data_08253.npz",
"report": "The patient underwent laser iridotomy for narrow angles and shows signs of ocular cupping - a possible indicator of glaucoma. They have low intraocular pressure, cataracts, dermatochalasis, and pterygium.",
"age": 64.11,
"gender": "female",
"race": "white",
"ethnicity": "non-hispanic",
"language": "unknown",
"maritalstatus": "single",
"note": "imp: s/p laser iridotomy ou DATE_TIME for occludable narrow angles cupping ou; low iop, PERSON, nl hvf and oct amd ou - followed by dr. PERSON cataract ou dermatochalasis ou pterygium ou plan: DATE_TIME with hvf and oct rx=m rudnick pgy4 i saw and evaluated this patient and discussed the case as appropriate with the resident/fellow. i have reviewed the resident/fellow's notes and made any necessary changes.",
"gpt4_summary": "The patient underwent laser iridotomy for narrow angles and shows signs of ocular cupping - a possible indicator of glaucoma. They have low intraocular pressure, cataracts, dermatochalasis, and pterygium.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08255",
"image_path": "slo_fundus_08255.jpg",
"filename": "data_08255.npz",
"report": "The patient has reduced acuity in the left eye and a cataract in the right eye, with no evidence of optic neuropathy. A macular hole has been noted by an ophthalmologist. No glaucoma present.",
"age": 79.04,
"gender": "male",
"race": "white",
"ethnicity": "unknown",
"language": "english",
"maritalstatus": "married or partnered",
"note": "prophylactic - a low dose beta blocker may be reasonable if he continues to have high blood pressure. he has reduced acuity in the left eye that has been stable over time. we did not find evidence of an optic neuropathy os and his evaluation of the neural components of his macula were normal. his outside ophthalmologist had noted a macular hole. he has bilateral dermatochalasis and a cataract od. these are not bothering him for now. recommendations: 1. could start a migraine prophylactic if he has frequent recurrence of positive visual phenomena 2. follow up with his comprehensive ophthalmologist i personally spent a total of *** DATE_TIME on care for this patient on the date of the encounter. this includes face-to-face time during the visit as well as non face-to-face time spent on chart documentation, and care coordination. this note was prepared by dr. PERSON, neuro-ophthalmology fellow, and me. any part of this note that was prepared by the resident or fellow was reviewed and modified as necessary by me to complete the finalized summary. ? PERSON, LOCATION.",
"gpt4_summary": "The patient has reduced acuity in the left eye and a cataract in the right eye, with no evidence of optic neuropathy. A macular hole has been noted by an ophthalmologist. No glaucoma present.",
"glaucoma": "yes",
"use": "test"
},
{
"id": "data_08257",
"image_path": "slo_fundus_08257.jpg",
"filename": "data_08257.npz",
"report": "63 y.o. black, non-hispanic female with no diagnosis of glaucoma.",
"age": 63.19,
"gender": "female",
"race": "black",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "divorced",
"note": "a 63 y.o. black, non-hispanic female with no diagnosis of glaucoma.",
"gpt4_summary": "63 y.o. black, non-hispanic female with no diagnosis of glaucoma.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08258",
"image_path": "slo_fundus_08258.jpg",
"filename": "data_08258.npz",
"report": "The clinical note discusses allergies, current medications, appointments, and overall health conditions of a patient. Notable medications include acetaminophen-codeine for pain and alendronate for osteoporosis. The patient is also taking levothyroxine for hypothyroidism. There is no mention of glaucoma.",
"age": 59.31,
"gender": "female",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "allergies as of DATE_TIME ibuprofen levofloxacin sulfa (sulfonamide antibiotics) sulfamethoxazole-trimethoprim medications and orders your current medications acetaminophen-codeine (tylenol-codeine #PHONE_NUMBER mg per tablet (taking) take 1-2 tablets by mouth every 4 (DATE_TIME as needed for pain (specific location in comments). alendronate (fosamax) 70 mg tablet (taking) take 1 tablet (70 mg total) by mouth DATE_TIME. PERSON (xanax) 0.25 mg tablet (taking) take 1 tablet (0.25 mg total) by mouth nightly as needed (back pain). cholecalciferol (vitamin d3) 2,000 unit capsule (taking) take by mouth DATE_TIME. conjugated estrogens (premarin) vaginal cream (taking) place 0.5 g vaginally nightly. use DATE_TIME x DATE_TIME then decrease to 2x/wk. levothyroxine (synthroid, levothroid) 75 mcg tablet (taking) take 1 tablet (75 mcg total) by mouth DATE_TIME. multivitamin per tablet (taking) take 1 tablet by mouth DATE_TIME. hrishikesh sawant DATE_TIME 2:26 am metal med transfer process pravastatin (pravachol) 20 mg tablet (taking) take 1 tablet (20 mg total) by mouth DATE_TIME. your orders future appointments provider department dept phone DATE_TIME PERSON alora LOCATION, PERSONtitution medical dermatology 617-726-2914 DATE_TIME alaka ray, PERSONtitution internal medicine associates 617-724-4600 DATE_TIME DATE_TIME PERSON, md, PERSON associates 617-742-2054 DATE_TIME 9:10 am PERSON, PERSONtitution comprehensive oph main campus PHONE_NUMBER orders placed this visit humphrey visual field - ou - both eyes oct, optic nerve - ou - both eyes - condition list as of DATE_TIME migraine dizziness back pain health care maintenance hypothyroidism osteoporosis results summary immunizations administered on date of encounter - DATE_TIME none",
"gpt4_summary": "The clinical note discusses allergies, current medications, appointments, and overall health conditions of a patient. Notable medications include acetaminophen-codeine for pain and alendronate for osteoporosis. The patient is also taking levothyroxine for hypothyroidism. There is no mention of glaucoma.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08259",
"image_path": "slo_fundus_08259.jpg",
"filename": "data_08259.npz",
"report": "The patient has primary open-angle glaucoma that's mild stage. Undergone glaucoma procedure, current IOP is acceptable. Also suffers from early cataract, refractive error (hyperopia, presbyopia), and dry eyes.",
"age": 72.11,
"gender": "male",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "single",
"note": "1. primary open-angle glaucoma ou mild stage: tmax 18/25 mm hg; cct 480/475. angle narrow by PERSON but open on gonio DATE_TIME glaucoma procedure: od: slt DATE_TIME with PERSON: slt DATE_TIME with PERSON current iop is acceptable; tg < 16 od and </=12 os 2. early cataract ou: minimally visually significant plan: observe 3. refractive error: hyperopia, presbyopia --continue current rx 4. dry eyes at prn plan iop excellent on cosopt bid visual field left eye is slowly worsening rnfl oct worse vs DATE_TIME gonioscopy remains open goal now near 12 os - continue ltn and timolol bid both eyes - rtc DATE_TIME iop check",
"gpt4_summary": "The patient has primary open-angle glaucoma that's mild stage. Undergone glaucoma procedure, current IOP is acceptable. Also suffers from early cataract, refractive error (hyperopia, presbyopia), and dry eyes.",
"glaucoma": "yes",
"use": "test"
},
{
"id": "data_08263",
"image_path": "slo_fundus_08263.jpg",
"filename": "data_08263.npz",
"report": "53-year-old female is a glaucoma suspect due to cup:disc appearance, likely physiologic in healthy appearing rims; low intraocular pressure. No family history of glaucoma.",
"age": 53.31,
"gender": "female",
"race": "white",
"ethnicity": "unknown",
"language": "spanish",
"maritalstatus": "married or partnered",
"note": "53 y.o. female healthy here with husband in past; here alone DATE_TIME - glaucoma suspect based on cup:disc appearance ou, likely physiologic in setting of healthy appearing rims and low iop fam hx: none tmax: 15/15 cct thin: 498/498 gonio : hvf DATE_TIME: ou full DATE_TIME: ou full rnfl DATE_TIME: ou wnl DATE_TIME: PERSON and stable, PERSON with slightly decreased avg nfl disc photos: DATE_TIME last dilated: DATE_TIME >> monitor off eyedrops given normal testing DATE_TIME - refractive error >> defers new glasses prescription DATE_TIME. she asked about possibility of surgery to eliminate need for glasses. we discussed laser vision correction but possibility of needing reading glasses if target distance, or blended/minimonovision. also discussed contact lenses. she will continue with glasses for now f/up DATE_TIME with hvf, rnfl oct, dilation, sooner prn",
"gpt4_summary": "53-year-old female is a glaucoma suspect due to cup:disc appearance, likely physiologic in healthy appearing rims; low intraocular pressure. No family history of glaucoma.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08265",
"image_path": "slo_fundus_08265.jpg",
"filename": "data_08265.npz",
"report": "Patient has optic nerve hypoplasia, mild refractive error os>od, 1 pd right hypertropia, inferior field loss in both eyes, anxiety and nausea. No glaucoma indication.",
"age": 49.19,
"gender": "female",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "single",
"note": "formulation: this patient returns for follow up of optic nerve hypoplasia and intermittent, achy left eye pain that precipitates anxiety and nausea. her exam is notable for mild refractive error os>od. her anterior exam is normal. her optic nerves appear hypoplastic and her humphrey visual fields are notable for inferior field loss in both eyes. her efferent exam is notable for a 1 pd right hypertropia at distance. it seems that he symptoms may be related to both refractive error and a small vertical misalignment. assessment: 1. hypoplastic optic nerves, possibly secondary to superior segmental optic nerve atrophy 2. refractive error 3. small right hypertropia 4. history of PERSON ou plan: 1. we will plan to bring the patient back for prism measurements (undilated) and refraction in DATE_TIME and provide a new prescription at that time this patient was seen and note prepared with PERSON, ophthalmology resident.",
"gpt4_summary": "Patient has optic nerve hypoplasia, mild refractive error os>od, 1 pd right hypertropia, inferior field loss in both eyes, anxiety and nausea. No glaucoma indication.",
"glaucoma": "yes",
"use": "test"
},
{
"id": "data_08266",
"image_path": "slo_fundus_08266.jpg",
"filename": "data_08266.npz",
"report": "The patient has vision problems and is suspected of having narrow angle glaucoma in both eyes, shown by their specialty eye tests. They also have a mild cataract but no current treatments or medications.",
"age": 67.14,
"gender": "female",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "problem list items addressed this visit eye/vision problems narrow angle glaucoma suspect of both eyes overview primary angle closure suspect. target iop: / ; tmax: ( ) / ( ); central corneal thickness: / ; ch: / refractive error: od +3.25 . sphere x / os +2.50 . sphere x optic nerve structure and function: normal visual field and retinal nerve fiber layer at presentation (DATE_TIME). medications and intolerances: none procedures and complications: none relevant history and problems: cataract - mild current assessment & plan no intraocular pressure treatment now. rx specs. other visit diagnoses narrow angle glaucoma suspect relevant orders humphrey visual field - ou - both eyes (completed) oct, optic nerve - ou - both eyes - (completed) return in DATE_TIME (around DATE_TIME) for optic nerve imaging (oct), visual field, gonioscopy, pachymetry.",
"gpt4_summary": "The patient has vision problems and is suspected of having narrow angle glaucoma in both eyes, shown by their specialty eye tests. They also have a mild cataract but no current treatments or medications.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08270",
"image_path": "slo_fundus_08270.jpg",
"filename": "data_08270.npz",
"report": "The patient's right eye is consistently over the goal. BGI was performed with three fenestrations. Considerations include phaco/BGI for the left eye and Rhopressa QHS if IOP exceeds 12 mmHg. Glaucoma may be present.",
"age": 72.03,
"gender": "female",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "single",
"note": "persistently above goal od on PERSON except for PERSON, we proceeded with bgi od st (3 fenestrations) in the sulcus on DATE_TIME. -rtc in DATE_TIME with iop check and oct rnfl/gcc ou, sooner prn. if good result obtained od, we may want to consider phaco/bgi os. if iop above 12 mmhg od, we will also consider rhopressa qhs or LOCATION to head off hypertensive phase. the information above was documented by Person as a scribe for PERSON on DATE_TIME.",
"gpt4_summary": "The patient's right eye is consistently over the goal. BGI was performed with three fenestrations. Considerations include phaco/BGI for the left eye and Rhopressa QHS if IOP exceeds 12 mmHg. Glaucoma may be present.",
"glaucoma": "yes",
"use": "test"
},
{
"id": "data_08272",
"image_path": "slo_fundus_08272.jpg",
"filename": "data_08272.npz",
"report": "Patient has chronic angle-closure glaucoma in right eye and ocular hypertension in left. Pressure is controlled and stable. Has pseudophakia which is stable and previous retinal detachments but is stable.\n",
"age": 52.21,
"gender": "female",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "single",
"note": "1. chronic angle-closure glacuoma od / ocular hypertension os: cct 630/570 iop goal less than 21mmhg. iop control is well within the target range. ran out of brimonidine and travatan DATE_TIME; pressure remains stable. plan: cont cosopt ou bid hold brimonidine 0.15% ou bid for now hold travatan ou qhs for now f/u DATE_TIME pressure check 2. pseudophakia (pciol) ou: stable plan: observe DATE_TIME/o multiple retinal detachments od s/p sb: stable plan: rd precautions discussed monocular precautions",
"gpt4_summary": "Patient has chronic angle-closure glaucoma in right eye and ocular hypertension in left. Pressure is controlled and stable. Has pseudophakia which is stable and previous retinal detachments but is stable.\n",
"glaucoma": "yes",
"use": "test"
},
{
"id": "data_08277",
"image_path": "slo_fundus_08277.jpg",
"filename": "data_08277.npz",
"report": "The clinical note does not indicate the presence of glaucoma. It mentions color fundus photography, visual field for both eyes, and retina examinations. The patient has multiple conditions like back pain, depression, osteoporosis, among others.",
"age": 55.09,
"gender": "female",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "single",
"note": "vishal mane DATE_TIME needs PERSON. metal med transfer process. ropinirole (requip) 1 mg tablet 3 mg tab p.o. DATE_TIME your orders normal orders this visit color fundus photography - ou - both eyes PERSON visual field - ou - both eyes oct, retina - ou - both eyes - cirrus; retina; 6mm length; 0.25mm spacing future labs/procedures complete by expires mri angio neck DATE_TIME DATE_TIME mri brain DATE_TIME DATE_TIME oct, retina - ou - both eyes - cirrus; retina; 6mm length; 0.25mm spacing as directed DATE_TIME condition list as of DATE_TIME back pain depression osteoporosis injury of head endometriosis restless leg syndrome migraine hearing loss sleep apnea asthma hypertensive disorder ddd (degenerative disc disease), lumbar lumbar spondylosis thoracic spondylosis sciatica traumatic brain injury hyperlipidemia age-related nuclear cataract of both eyes facet joint disease neuropathic pain results summary immunizations administered on date of encounter - DATE_TIME none",
"gpt4_summary": "The clinical note does not indicate the presence of glaucoma. It mentions color fundus photography, visual field for both eyes, and retina examinations. The patient has multiple conditions like back pain, depression, osteoporosis, among others.",
"glaucoma": "yes",
"use": "test"
},
{
"id": "data_08280",
"image_path": "slo_fundus_08280.jpg",
"filename": "data_08280.npz",
"report": "Patient has myopic astigmatism, presbyopia, and is deemed a glaucoma suspect with ocular pressure of 24 mmHg. Recommended glaucoma-related tests.",
"age": 49.32,
"gender": "female",
"race": "white",
"ethnicity": "unknown",
"language": "unknown",
"maritalstatus": "married or partnered",
"note": "1. myopic astigmat against the rule, presbyope 2. rx = -0.25-0.75x135 add +1.75 -0.50 -0.50x90 add +1.75 3. glaucoma suspect based on iop's of 24 mmhg ou............. refer to glaucoma for vf and onh photos 4. confrontational visual fields seems slightly constricted.",
"gpt4_summary": "Patient has myopic astigmatism, presbyopia, and is deemed a glaucoma suspect with ocular pressure of 24 mmHg. Recommended glaucoma-related tests.",
"glaucoma": "yes",
"use": "test"
},
{
"id": "data_08283",
"image_path": "slo_fundus_08283.jpg",
"filename": "data_08283.npz",
"report": "The patient has mog-associated disease/optic neuritis. Their care team is working on insurance approval for IVIG treatment. No mention of glaucoma.",
"age": 69.06,
"gender": "female",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "flares. PERSON has the best data for reduction of DATE_TIME relapse rate of mog-associated disease/optic neuritis, and arguably a better side-effect profile than rituximab or cellcept. i agree with the recommendation of DATE_TIME ivig under the care of dr. PERSON whose team has been working for insurance approval. ? i will plan to see her back in DATE_TIME to monitor her progress. recommendations: 1. agree with DATE_TIME ivig treatments 2. return to see me in DATE_TIME i personally spent a total of DATE_TIME on care for this patient on the date of the encounter. this includes face-to-face time during the visit as well as non face-to-face time spent on chart documentation, and care coordination. ? PERSON, LOCATION.",
"gpt4_summary": "The patient has mog-associated disease/optic neuritis. Their care team is working on insurance approval for IVIG treatment. No mention of glaucoma.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08285",
"image_path": "slo_fundus_08285.jpg",
"filename": "data_08285.npz",
"report": "The patient, a 56 y.o. female, is a glaucoma suspect due to the cup to disc ratio in both eyes. There's no thinning of the retinal fiber layer or reduction in visual fields. No treatment commenced, will continue monitoring. She also has mild cataracts.\n",
"age": 56.9,
"gender": "female",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "divorced",
"note": "hysteresis: 9.6 / 8.6 gonioscopy: d30f 2+ ou retinal nerve fiber layer, right eye: no thinning retinal nerve fiber layer, left eye: no thinning visual fields, right eye: full visual fields, left eye: full family history: none steroids: none trauma: none asthma/copd: none other medical history and problems: htn, hld, nafld, migraine, anxiety, depression assessment/plan: 56 y.o. female nwh medical assistant referred by dr. PERSON ding # glaucoma suspect due to cup to disc ratio, both eyes, both eyes - healthy nerve rim, reassuring testing - iop acceptable ou - continue to monitor without initiating treatment - return in DATE_TIME for iop check, dilate, oct, disc photos # cataract, both eyes - mild, not visually significant, monitor PERSON, PERSON___________________ i saw and evaluated this patient and discussed the case as appropriate with the fellow. i have reviewed the fellow's notes and made any necessary changes. PERSON scribing for dr. PERSON i attest that i was present while this encounter was recorded and agree that the information entered by my scribe and updated by me is complete and accurate. PERSON, md",
"gpt4_summary": "The patient, a 56 y.o. female, is a glaucoma suspect due to the cup to disc ratio in both eyes. There's no thinning of the retinal fiber layer or reduction in visual fields. No treatment commenced, will continue monitoring. She also has mild cataracts.\n",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08287",
"image_path": "slo_fundus_08287.jpg",
"filename": "data_08287.npz",
"report": "The patient is on several medications including venlafaxine, docusate sodium, neurontin, melatonin, omeprazole and vitamin B complex. No presence of glaucoma reported. Other conditions include asthma, depression, osteoarthritis, and urinary tract infection.",
"age": 76.56,
"gender": "female",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "widowed",
"note": "venlafaxine (PERSON) 75 mg tablet (taking) take 1 tablet by mouth twice a day custom medication, see admin inst / label comment, reported on DATE_TIME DATE_TIME needs PERSON. metal med transfer process, from oncall docusate sodium (colace) 100 mg capsule take 1 capsule (100 mg total) by mouth 2 (two) times a day as needed for constipation. PERSON (neurontin) 300 mg capsule take 1 capsule (300 mg total) by mouth nightly as needed. melatonin 5 mg tab take 1 tablet (5 mg total) by mouth nightly. omeprazole (prilosec) 20 mg capsule take 20 mg by mouth DATE_TIME. reported on DATE_TIME DATE_TIME 3:23 am metal med transfer process, from PERSON (LOCATION) 20 mg tablet take 1 tablet (20 mg total) by mouth DATE_TIME. vitamin b complex oral take by mouth. reported on DATE_TIME DATE_TIME 3:27 am needs PERSON. metal med transfer process, from oncall your orders normal orders this visit humphrey visual field - ou - both eyes oct, optic nerve - ou - both eyes - condition list as of DATE_TIME asthma allergic rhinitis rectal prolapse depressive disorder atypical chest pain hypercholesterolemia s/p cholecystectomy history of bladder suspension procedure h/o hysterectomy for benign disease obesity fatigue osteoarthritis abnormal results of liver function studies gastroesophageal reflux disease seborrheic keratosis seborrheic dermatitis contact dermatitis actinic keratosis lumbar stenosis urinary incontinence urinary tract infection, site not specified chronic headaches routine adult health maintenance right hip pain results summary immunizations administered on date of encounter - DATE_TIME none",
"gpt4_summary": "The patient is on several medications including venlafaxine, docusate sodium, neurontin, melatonin, omeprazole and vitamin B complex. No presence of glaucoma reported. Other conditions include asthma, depression, osteoarthritis, and urinary tract infection.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08300",
"image_path": "slo_fundus_08300.jpg",
"filename": "data_08300.npz",
"report": "The patient is prescribed preservative-free artificial tears and will follow-up for cornea care. A check-up for intraocular pressure (IOP) is scheduled, with the possibility of Selective Laser Trabeculoplasty (SLT) treatment for glaucoma.",
"age": 59.13,
"gender": "female",
"race": "black",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "single",
"note": "to medication regimen. -instructions written/typed/printed out for patient (see table/details under patient instructions) previously. -preservative-free artificial tears as needed. -follow-up with dr. PERSON for cornea care. -rtc in DATE_TIME with iop check and hvf size v ou, sooner prn. if iop above goal or difficulty with adherence, we'll book slt os first, and if good result od. the information above was documented by PERSON as a scribe for PERSON on DATE_TIME.",
"gpt4_summary": "The patient is prescribed preservative-free artificial tears and will follow-up for cornea care. A check-up for intraocular pressure (IOP) is scheduled, with the possibility of Selective Laser Trabeculoplasty (SLT) treatment for glaucoma.",
"glaucoma": "yes",
"use": "test"
},
{
"id": "data_08304",
"image_path": "slo_fundus_08304.jpg",
"filename": "data_08304.npz",
"report": "The patient has open-angle glaucoma, mild stage, with high intraocular pressure. Started on latanoprost. Underwent SLT with good response. Plan is to continue medication and follow-up.",
"age": 23.16,
"gender": "female",
"race": "asian",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "single",
"note": "visual acuity visual acuity (snellen - linear) right left dist cc PHONE_NUMBER correction: glasses tonometry tonometry (applanation, DATE_TIME) right left pressure 17 15 main ophthalmology exam external exam right left external normal normal slit lamp exam right left lids/lashes normal normal conjunctiva/sclera normal normal cornea normal normal anterior chamber normal normal iris normal normal lens normal normal fundus exam right left vitreous normal normal disc normal normal c/d ratio 0.9 0.9 macula normal normal vessels normal normal periphery normal normal problem list items addressed this visit eye/vision problems open-angle glaucoma, mild stage overview early juvenile open-angle glaucoma ou: cct 660/660 tm 23/24: +family history. pt's iop has been high in previous visits and rfnl was showing some changes. pt has a possible vf defect os, but that is not significantly different than DATE_TIME. started on latanoprost by PERSON PERSON. iop higher DATE_TIME. will initially set goal at teens ou for ~20% decr in tm. stable vf with some fluctuation in vf os with early sup defects that don't appear worse since DATE_TIME. oct with no signficant change but may have slightly more sup thinning od --> will repeat with dilated test to confirm. testing stable but with high iop again so slt DATE_TIME with good response goal <18 current assessment & plan iop continues to be good after slt os DATE_TIME (20s --> mid teens) plan: latanoprost ou qhs follow-up in DATE_TIME iop check PERSON, LOCATION",
"gpt4_summary": "The patient has open-angle glaucoma, mild stage, with high intraocular pressure. Started on latanoprost. Underwent SLT with good response. Plan is to continue medication and follow-up.",
"glaucoma": "no",
"use": "test"
},
{
"id": "data_08318",
"image_path": "slo_fundus_08318.jpg",
"filename": "data_08318.npz",
"report": "No evidence of glaucoma or sub-clinical abnormalities found. Patient has blurred vision, occipital neuralgia, and history of optic nerve 'pit'. Recommended trigger point injection.",
"age": 38.93,
"gender": "female",
"race": "white",
"ethnicity": "non-hispanic",
"language": "english",
"maritalstatus": "married or partnered",
"note": "superiorly, i did not find any evidence of
gitextract_tizp684z/
├── .gitignore
├── LICENSE
├── README.md
├── data/
│ ├── test/
│ │ ├── report/
│ │ │ ├── harvard_test.json
│ │ │ ├── iuxray_test.json
│ │ │ ├── mimic_test.json
│ │ │ ├── pmc-oa_test.json
│ │ │ └── quilt-1m_test.json
│ │ └── vqa/
│ │ ├── harvard_test.jsonl
│ │ ├── iuxray_test.jsonl
│ │ ├── mimic_test.jsonl
│ │ ├── pmc-oa_test.jsonl
│ │ └── quilt-1m_test.jsonl
│ └── training/
│ ├── alignment/
│ │ ├── ophthalmology/
│ │ │ ├── harvard_report.json
│ │ │ └── harvard_vqa.json
│ │ ├── pathology/
│ │ │ └── pathology_vqa.json
│ │ └── radiology/
│ │ ├── radiology_report.json
│ │ └── radiology_vqa.json
│ └── retriever/
│ ├── ophthalmology/
│ │ ├── harvard_train_7000.json
│ │ └── harvard_val_1000.json
│ ├── pathology/
│ │ ├── pathology_train.json
│ │ └── pathology_val.json
│ └── radiology/
│ ├── radiology_train.json
│ └── radiology_val.json
├── requirements.txt
├── scripts/
│ ├── finetune_clip.sh
│ ├── retrieve_clip_VQA.sh
│ ├── retrieve_clip_report.sh
│ └── train_dpo_2stages.sh
└── train/
├── dpo/
│ ├── LICENSE
│ ├── cog.yaml
│ ├── dpo_trainer_2stages.py
│ ├── llava/
│ │ ├── __init__.py
│ │ ├── constants.py
│ │ ├── conversation.py
│ │ ├── conversation_new.py
│ │ ├── eval/
│ │ │ ├── eval_gpt_review.py
│ │ │ ├── eval_gpt_review_bench.py
│ │ │ ├── eval_gpt_review_visual.py
│ │ │ ├── eval_pope.py
│ │ │ ├── eval_science_qa.py
│ │ │ ├── eval_science_qa_gpt4.py
│ │ │ ├── eval_science_qa_gpt4_requery.py
│ │ │ ├── eval_textvqa.py
│ │ │ ├── generate_webpage_data_from_table.py
│ │ │ ├── m4c_evaluator.py
│ │ │ ├── model_qa.py
│ │ │ ├── model_vqa.py
│ │ │ ├── model_vqa_loader.py
│ │ │ ├── model_vqa_mmbench.py
│ │ │ ├── model_vqa_science.py
│ │ │ ├── qa_baseline_gpt35.py
│ │ │ ├── run_llava.py
│ │ │ ├── summarize_gpt_review.py
│ │ │ ├── table/
│ │ │ │ ├── answer/
│ │ │ │ │ ├── answer_alpaca-13b.jsonl
│ │ │ │ │ ├── answer_bard.jsonl
│ │ │ │ │ ├── answer_gpt35.jsonl
│ │ │ │ │ ├── answer_llama-13b.jsonl
│ │ │ │ │ └── answer_vicuna-13b.jsonl
│ │ │ │ ├── caps_boxes_coco2014_val_80.jsonl
│ │ │ │ ├── model.jsonl
│ │ │ │ ├── prompt.jsonl
│ │ │ │ ├── question.jsonl
│ │ │ │ ├── results/
│ │ │ │ │ ├── test_sqa_llava_13b_v0.json
│ │ │ │ │ └── test_sqa_llava_lcs_558k_sqa_12e_vicuna_v1_3_13b.json
│ │ │ │ ├── review/
│ │ │ │ │ ├── review_alpaca-13b_vicuna-13b.jsonl
│ │ │ │ │ ├── review_bard_vicuna-13b.jsonl
│ │ │ │ │ ├── review_gpt35_vicuna-13b.jsonl
│ │ │ │ │ └── review_llama-13b_vicuna-13b.jsonl
│ │ │ │ ├── reviewer.jsonl
│ │ │ │ └── rule.json
│ │ │ └── webpage/
│ │ │ └── styles.css
│ │ ├── mm_utils.py
│ │ ├── model/
│ │ │ ├── __init__.py
│ │ │ ├── apply_delta.py
│ │ │ ├── builder.py
│ │ │ ├── consolidate.py
│ │ │ ├── language_model/
│ │ │ │ ├── llava_llama.py
│ │ │ │ ├── llava_mistral.py
│ │ │ │ └── llava_mpt.py
│ │ │ ├── llava_arch.py
│ │ │ ├── make_delta.py
│ │ │ ├── multimodal_encoder/
│ │ │ │ ├── builder.py
│ │ │ │ └── clip_encoder.py
│ │ │ ├── multimodal_projector/
│ │ │ │ └── builder.py
│ │ │ └── utils.py
│ │ ├── serve/
│ │ │ ├── __init__.py
│ │ │ ├── cli.py
│ │ │ ├── controller.py
│ │ │ ├── gradio_web_server.py
│ │ │ ├── model_worker.py
│ │ │ ├── register_worker.py
│ │ │ ├── sglang_worker.py
│ │ │ └── test_message.py
│ │ ├── train/
│ │ │ ├── llama_flash_attn_monkey_patch.py
│ │ │ ├── llama_xformers_attn_monkey_patch.py
│ │ │ ├── llava_trainer.py
│ │ │ ├── train.py
│ │ │ ├── train_dpo.py
│ │ │ ├── train_dpo_inherent.py
│ │ │ ├── train_mem.py
│ │ │ └── train_xformers.py
│ │ └── utils.py
│ ├── llava_trainer_2stages.py
│ ├── povid_infer.py
│ ├── predict.py
│ ├── pyproject.toml
│ ├── scripts/
│ │ ├── convert_gqa_for_eval.py
│ │ ├── convert_mmbench_for_submission.py
│ │ ├── convert_mmvet_for_eval.py
│ │ ├── convert_seed_for_submission.py
│ │ ├── convert_sqa_to_llava.py
│ │ ├── convert_sqa_to_llava_base_prompt.py
│ │ ├── convert_vizwiz_for_submission.py
│ │ ├── convert_vqav2_for_submission.py
│ │ ├── extract_mm_projector.py
│ │ ├── finetune.sh
│ │ ├── finetune_full_schedule.sh
│ │ ├── finetune_lora.sh
│ │ ├── finetune_qlora.sh
│ │ ├── finetune_sqa.sh
│ │ ├── merge_lora_weights.py
│ │ ├── pretrain.sh
│ │ ├── pretrain_xformers.sh
│ │ ├── run_povid.sh
│ │ ├── sqa_eval_batch.sh
│ │ ├── sqa_eval_gather.sh
│ │ ├── upload_pypi.sh
│ │ ├── v1_5/
│ │ │ ├── eval/
│ │ │ │ ├── gqa.sh
│ │ │ │ ├── llavabench.sh
│ │ │ │ ├── mmbench.sh
│ │ │ │ ├── mmbench_cn.sh
│ │ │ │ ├── mme.sh
│ │ │ │ ├── mmvet.sh
│ │ │ │ ├── pope.sh
│ │ │ │ ├── qbench.sh
│ │ │ │ ├── qbench_zh.sh
│ │ │ │ ├── seed.sh
│ │ │ │ ├── sqa.sh
│ │ │ │ ├── textvqa.sh
│ │ │ │ ├── vizwiz.sh
│ │ │ │ └── vqav2.sh
│ │ │ ├── finetune.sh
│ │ │ ├── finetune_lora.sh
│ │ │ ├── finetune_task.sh
│ │ │ ├── finetune_task_lora.sh
│ │ │ └── pretrain.sh
│ │ ├── zero2.json
│ │ ├── zero3.json
│ │ └── zero3_offload.json
│ ├── tool/
│ │ ├── dpo_trainer.py
│ │ └── dpo_trainer_inherent.py
│ └── train_dpo_2stages.py
└── open_clip/
├── CITATION.cff
├── HISTORY.md
├── LICENSE
├── MANIFEST.in
├── setup.py
└── src/
├── open_clip/
│ ├── __init__.py
│ ├── big_vision.py
│ ├── coca_model.py
│ ├── constants.py
│ ├── factory.py
│ ├── hf_configs.py
│ ├── hf_model.py
│ ├── loss.py
│ ├── model.py
│ ├── model_configs/
│ │ ├── EVA01-g-14-plus.json
│ │ ├── EVA01-g-14.json
│ │ ├── EVA02-B-16.json
│ │ ├── EVA02-E-14-plus.json
│ │ ├── EVA02-E-14.json
│ │ ├── EVA02-L-14-336.json
│ │ ├── EVA02-L-14.json
│ │ ├── RN101-quickgelu.json
│ │ ├── RN101.json
│ │ ├── RN50-quickgelu.json
│ │ ├── RN50.json
│ │ ├── RN50x16.json
│ │ ├── RN50x4.json
│ │ ├── RN50x64.json
│ │ ├── ViT-B-16-SigLIP-256.json
│ │ ├── ViT-B-16-SigLIP-384.json
│ │ ├── ViT-B-16-SigLIP-512.json
│ │ ├── ViT-B-16-SigLIP-i18n-256.json
│ │ ├── ViT-B-16-SigLIP.json
│ │ ├── ViT-B-16-plus-240.json
│ │ ├── ViT-B-16-plus.json
│ │ ├── ViT-B-16-quickgelu.json
│ │ ├── ViT-B-16.json
│ │ ├── ViT-B-32-256.json
│ │ ├── ViT-B-32-plus-256.json
│ │ ├── ViT-B-32-quickgelu.json
│ │ ├── ViT-B-32.json
│ │ ├── ViT-H-14-378-quickgelu.json
│ │ ├── ViT-H-14-CLIPA-336.json
│ │ ├── ViT-H-14-CLIPA.json
│ │ ├── ViT-H-14-quickgelu.json
│ │ ├── ViT-H-14.json
│ │ ├── ViT-H-16.json
│ │ ├── ViT-L-14-280.json
│ │ ├── ViT-L-14-336.json
│ │ ├── ViT-L-14-CLIPA-336.json
│ │ ├── ViT-L-14-CLIPA.json
│ │ ├── ViT-L-14-quickgelu.json
│ │ ├── ViT-L-14.json
│ │ ├── ViT-L-16-320.json
│ │ ├── ViT-L-16-SigLIP-256.json
│ │ ├── ViT-L-16-SigLIP-384.json
│ │ ├── ViT-L-16.json
│ │ ├── ViT-M-16-alt.json
│ │ ├── ViT-M-16.json
│ │ ├── ViT-M-32-alt.json
│ │ ├── ViT-M-32.json
│ │ ├── ViT-S-16-alt.json
│ │ ├── ViT-S-16.json
│ │ ├── ViT-S-32-alt.json
│ │ ├── ViT-S-32.json
│ │ ├── ViT-SO400M-14-SigLIP-384.json
│ │ ├── ViT-SO400M-14-SigLIP.json
│ │ ├── ViT-bigG-14-CLIPA-336.json
│ │ ├── ViT-bigG-14-CLIPA.json
│ │ ├── ViT-bigG-14.json
│ │ ├── ViT-e-14.json
│ │ ├── ViT-g-14.json
│ │ ├── coca_ViT-B-32.json
│ │ ├── coca_ViT-L-14.json
│ │ ├── coca_base.json
│ │ ├── coca_roberta-ViT-B-32.json
│ │ ├── convnext_base.json
│ │ ├── convnext_base_w.json
│ │ ├── convnext_base_w_320.json
│ │ ├── convnext_large.json
│ │ ├── convnext_large_d.json
│ │ ├── convnext_large_d_320.json
│ │ ├── convnext_small.json
│ │ ├── convnext_tiny.json
│ │ ├── convnext_xlarge.json
│ │ ├── convnext_xxlarge.json
│ │ ├── convnext_xxlarge_320.json
│ │ ├── mt5-base-ViT-B-32.json
│ │ ├── mt5-xl-ViT-H-14.json
│ │ ├── nllb-clip-base-siglip.json
│ │ ├── nllb-clip-base.json
│ │ ├── nllb-clip-large-siglip.json
│ │ ├── nllb-clip-large.json
│ │ ├── roberta-ViT-B-32.json
│ │ ├── swin_base_patch4_window7_224.json
│ │ ├── vit_medium_patch16_gap_256.json
│ │ ├── vit_relpos_medium_patch16_cls_224.json
│ │ ├── xlm-roberta-base-ViT-B-32.json
│ │ └── xlm-roberta-large-ViT-H-14.json
│ ├── modified_resnet.py
│ ├── openai.py
│ ├── pos_embed.py
│ ├── pretrained.py
│ ├── push_to_hf_hub.py
│ ├── timm_model.py
│ ├── tokenizer.py
│ ├── transform.py
│ ├── transformer.py
│ ├── utils.py
│ ├── version.py
│ ├── zero_shot_classifier.py
│ └── zero_shot_metadata.py
├── retrieve_clip_VQA.py
├── retrieve_clip_report.py
└── training/
├── .gitignore
├── __init__.py
├── data.py
├── distributed.py
├── file_utils.py
├── logger.py
├── main.py
├── main_feature-work.py
├── main_retrieve_report.py
├── main_retrieve_report_harvard.py
├── params.py
├── precision.py
├── profiler.py
├── scheduler.py
├── train.py
└── zero_shot.py
SYMBOL INDEX (925 symbols across 95 files)
FILE: train/dpo/dpo_trainer_2stages.py
class DPOTrainer (line 60) | class DPOTrainer(Trainer):
method __init__ (line 123) | def __init__(
method _prepare_deepspeed (line 354) | def _prepare_deepspeed(self, model: PreTrainedModelWrapper):
method concatenated_inputs (line 385) | def concatenated_inputs(self, batch: Dict[str, Union[List, torch.LongT...
method dpo_loss (line 424) | def dpo_loss(
method _get_batch_logps (line 467) | def _get_batch_logps(
method _get_noisy_batch_logps (line 501) | def _get_noisy_batch_logps(
method concatenated_forward (line 550) | def concatenated_forward(
method get_batch_metrics (line 629) | def get_batch_metrics(
method compute_loss (line 681) | def compute_loss(
method get_batch_samples (line 702) | def get_batch_samples(self, model, batch: Dict[str, torch.LongTensor])...
method prediction_step (line 739) | def prediction_step(
method store_metrics (line 778) | def store_metrics(self, metrics: Dict[str, float], train_eval: Literal...
method evaluation_loop (line 782) | def evaluation_loop(
method log (line 832) | def log(self, logs: Dict[str, float]) -> None:
FILE: train/dpo/llava/conversation.py
class SeparatorStyle (line 9) | class SeparatorStyle(Enum):
class Conversation (line 19) | class Conversation:
method get_prompt (line 32) | def get_prompt(self):
method append_message (line 109) | def append_message(self, role, message):
method process_image (line 112) | def process_image(self, image, image_process_mode, return_pil=False, i...
method get_images (line 152) | def get_images(self, return_pil=False):
method to_gradio_chatbot (line 162) | def to_gradio_chatbot(self):
method copy (line 180) | def copy(self):
method dict (line 191) | def dict(self):
FILE: train/dpo/llava/conversation_new.py
class SeparatorStyle (line 6) | class SeparatorStyle(Enum):
class Conversation (line 13) | class Conversation:
method get_prompt (line 26) | def get_prompt(self):
method append_message (line 51) | def append_message(self, role, message):
method get_images (line 54) | def get_images(self, return_pil=False):
method to_gradio_chatbot (line 103) | def to_gradio_chatbot(self):
method copy (line 133) | def copy(self):
method dict (line 143) | def dict(self):
FILE: train/dpo/llava/eval/eval_gpt_review.py
function get_eval (line 13) | def get_eval(content: str, max_tokens: int):
function parse_score (line 39) | def parse_score(review):
FILE: train/dpo/llava/eval/eval_gpt_review_bench.py
function get_eval (line 11) | def get_eval(content: str, max_tokens: int):
function parse_score (line 36) | def parse_score(review):
FILE: train/dpo/llava/eval/eval_gpt_review_visual.py
function get_eval (line 11) | def get_eval(content: str, max_tokens: int):
function parse_score (line 36) | def parse_score(review):
FILE: train/dpo/llava/eval/eval_pope.py
function eval_pope (line 5) | def eval_pope(answers, label_file):
FILE: train/dpo/llava/eval/eval_science_qa.py
function get_args (line 8) | def get_args():
function convert_caps (line 19) | def convert_caps(results):
function get_pred_idx (line 28) | def get_pred_idx(prediction, choices, options):
FILE: train/dpo/llava/eval/eval_science_qa_gpt4.py
function get_args (line 9) | def get_args():
function convert_caps (line 19) | def convert_caps(results):
function get_pred_idx (line 28) | def get_pred_idx(prediction, choices, options):
FILE: train/dpo/llava/eval/eval_science_qa_gpt4_requery.py
function get_args (line 9) | def get_args():
function convert_caps (line 21) | def convert_caps(results):
function get_pred_idx (line 30) | def get_pred_idx(prediction, choices, options):
FILE: train/dpo/llava/eval/eval_textvqa.py
function get_args (line 9) | def get_args():
function prompt_processor (line 17) | def prompt_processor(prompt):
function eval_single (line 35) | def eval_single(annotation_file, result_file):
FILE: train/dpo/llava/eval/generate_webpage_data_from_table.py
function read_jsonl (line 10) | def read_jsonl(path: str, key: str=None):
function trim_hanging_lines (line 23) | def trim_hanging_lines(s: str, n: int) -> str:
FILE: train/dpo/llava/eval/m4c_evaluator.py
class EvalAIAnswerProcessor (line 7) | class EvalAIAnswerProcessor:
method __init__ (line 178) | def __init__(self, *args, **kwargs):
method word_tokenize (line 181) | def word_tokenize(self, word):
method process_punctuation (line 186) | def process_punctuation(self, in_text):
method process_digit_article (line 198) | def process_digit_article(self, in_text):
method __call__ (line 213) | def __call__(self, item):
class TextVQAAccuracyEvaluator (line 221) | class TextVQAAccuracyEvaluator:
method __init__ (line 222) | def __init__(self):
method _compute_answer_scores (line 225) | def _compute_answer_scores(self, raw_answers):
method eval_pred_list (line 248) | def eval_pred_list(self, pred_list):
class STVQAAccuracyEvaluator (line 260) | class STVQAAccuracyEvaluator:
method __init__ (line 261) | def __init__(self):
method eval_pred_list (line 264) | def eval_pred_list(self, pred_list):
class STVQAANLSEvaluator (line 276) | class STVQAANLSEvaluator:
method __init__ (line 277) | def __init__(self):
method get_anls (line 282) | def get_anls(self, s1, s2):
method eval_pred_list (line 289) | def eval_pred_list(self, pred_list):
class TextCapsBleu4Evaluator (line 301) | class TextCapsBleu4Evaluator:
method __init__ (line 302) | def __init__(self):
method eval_pred_list (line 321) | def eval_pred_list(self, pred_list):
FILE: train/dpo/llava/eval/model_qa.py
function eval_model (line 14) | def eval_model(model_name, questions_file, answers_file):
FILE: train/dpo/llava/eval/model_vqa.py
function split_list (line 18) | def split_list(lst, n):
function get_chunk (line 24) | def get_chunk(lst, n, k):
function eval_model (line 29) | def eval_model(args):
FILE: train/dpo/llava/eval/model_vqa_loader.py
function split_list (line 19) | def split_list(lst, n):
function get_chunk (line 25) | def get_chunk(lst, n, k):
class CustomDataset (line 31) | class CustomDataset(Dataset):
method __init__ (line 32) | def __init__(self, questions, image_folder, tokenizer, image_processor...
method __getitem__ (line 39) | def __getitem__(self, index):
method __len__ (line 60) | def __len__(self):
function collate_fn (line 64) | def collate_fn(batch):
function create_data_loader (line 72) | def create_data_loader(questions, image_folder, tokenizer, image_process...
function eval_model (line 79) | def eval_model(args):
FILE: train/dpo/llava/eval/model_vqa_mmbench.py
function split_list (line 22) | def split_list(lst, n):
function get_chunk (line 28) | def get_chunk(lst, n, k):
function is_none (line 33) | def is_none(value):
function get_options (line 44) | def get_options(row, options):
function eval_model (line 54) | def eval_model(args):
FILE: train/dpo/llava/eval/model_vqa_science.py
function split_list (line 18) | def split_list(lst, n):
function get_chunk (line 24) | def get_chunk(lst, n, k):
function eval_model (line 29) | def eval_model(args):
FILE: train/dpo/llava/eval/qa_baseline_gpt35.py
function get_answer (line 16) | def get_answer(question_id: int, question: str, max_tokens: int):
FILE: train/dpo/llava/eval/run_llava.py
function image_parser (line 28) | def image_parser(args):
function load_image (line 33) | def load_image(image_file):
function load_images (line 42) | def load_images(image_files):
function eval_model (line 50) | def eval_model(args):
FILE: train/dpo/llava/eval/summarize_gpt_review.py
function parse_args (line 9) | def parse_args():
FILE: train/dpo/llava/mm_utils.py
function select_best_resolution (line 12) | def select_best_resolution(original_size, possible_resolutions):
function resize_and_pad_image (line 42) | def resize_and_pad_image(image, target_resolution):
function divide_to_patches (line 77) | def divide_to_patches(image, patch_size):
function get_anyres_image_grid_shape (line 99) | def get_anyres_image_grid_shape(image_size, grid_pinpoints, patch_size):
function process_anyres_image (line 119) | def process_anyres_image(image, processor, grid_pinpoints):
function load_image_from_base64 (line 148) | def load_image_from_base64(image):
function expand2square (line 152) | def expand2square(pil_img, background_color):
function process_images (line 166) | def process_images(images, image_processor, model_cfg):
function tokenizer_image_token (line 185) | def tokenizer_image_token(prompt, tokenizer, image_token_index=IMAGE_TOK...
function get_model_name_from_path (line 207) | def get_model_name_from_path(model_path):
class KeywordsStoppingCriteria (line 215) | class KeywordsStoppingCriteria(StoppingCriteria):
method __init__ (line 216) | def __init__(self, keywords, tokenizer, input_ids):
method call_for_batch (line 230) | def call_for_batch(self, output_ids: torch.LongTensor, scores: torch.F...
method __call__ (line 243) | def __call__(self, output_ids: torch.LongTensor, scores: torch.FloatTe...
FILE: train/dpo/llava/model/apply_delta.py
function apply_delta (line 13) | def apply_delta(base_model_path, target_model_path, delta_path):
FILE: train/dpo/llava/model/builder.py
function load_pretrained_model (line 26) | def load_pretrained_model(model_path, model_base, model_name, load_8bit=...
FILE: train/dpo/llava/model/consolidate.py
function consolidate_ckpt (line 13) | def consolidate_ckpt(src_path, dst_path):
FILE: train/dpo/llava/model/language_model/llava_llama.py
class LlavaConfig (line 30) | class LlavaConfig(LlamaConfig):
class LlavaLlamaModel (line 34) | class LlavaLlamaModel(LlavaMetaModel, LlamaModel):
method __init__ (line 37) | def __init__(self, config: LlamaConfig):
class LlavaLlamaForCausalLM (line 41) | class LlavaLlamaForCausalLM(LlamaForCausalLM, LlavaMetaForCausalLM):
method __init__ (line 44) | def __init__(self, config):
method get_model (line 54) | def get_model(self):
method forward (line 57) | def forward(
method generate (line 105) | def generate(
method prepare_inputs_for_generation (line 144) | def prepare_inputs_for_generation(self, input_ids, past_key_values=None,
FILE: train/dpo/llava/model/language_model/llava_mistral.py
class LlavaMistralConfig (line 31) | class LlavaMistralConfig(MistralConfig):
class LlavaMistralModel (line 35) | class LlavaMistralModel(LlavaMetaModel, MistralModel):
method __init__ (line 38) | def __init__(self, config: MistralConfig):
class LlavaMistralForCausalLM (line 42) | class LlavaMistralForCausalLM(MistralForCausalLM, LlavaMetaForCausalLM):
method __init__ (line 45) | def __init__(self, config):
method get_model (line 54) | def get_model(self):
method forward (line 57) | def forward(
method generate (line 105) | def generate(
method prepare_inputs_for_generation (line 144) | def prepare_inputs_for_generation(self, input_ids, past_key_values=None,
FILE: train/dpo/llava/model/language_model/llava_mpt.py
class LlavaMptConfig (line 25) | class LlavaMptConfig(MptConfig):
class LlavaMptModel (line 29) | class LlavaMptModel(LlavaMetaModel, MptModel):
method __init__ (line 32) | def __init__(self, config: MptConfig):
method embed_tokens (line 36) | def embed_tokens(self, x):
class LlavaMptForCausalLM (line 40) | class LlavaMptForCausalLM(MptForCausalLM, LlavaMetaForCausalLM):
method __init__ (line 44) | def __init__(self, config):
method get_model (line 53) | def get_model(self):
method _set_gradient_checkpointing (line 56) | def _set_gradient_checkpointing(self, module, value=False):
method forward (line 60) | def forward(
method prepare_inputs_for_generation (line 87) | def prepare_inputs_for_generation(self, input_ids, past_key_values=Non...
FILE: train/dpo/llava/model/llava_arch.py
class LlavaMetaModel (line 29) | class LlavaMetaModel:
method __init__ (line 31) | def __init__(self, config):
method get_vision_tower (line 43) | def get_vision_tower(self):
method initialize_vision_modules (line 49) | def initialize_vision_modules(self, model_args, fsdp=None):
function unpad_image (line 100) | def unpad_image(tensor, original_size):
class LlavaMetaForCausalLM (line 131) | class LlavaMetaForCausalLM(ABC):
method get_model (line 134) | def get_model(self):
method get_vision_tower (line 137) | def get_vision_tower(self):
method encode_images (line 140) | def encode_images(self, images):
method prepare_inputs_labels_for_multimodal (line 145) | def prepare_inputs_labels_for_multimodal(
method initialize_vision_tokenizer (line 326) | def initialize_vision_tokenizer(self, model_args, tokenizer):
FILE: train/dpo/llava/model/make_delta.py
function make_delta (line 13) | def make_delta(base_model_path, target_model_path, delta_path, hub_repo_...
FILE: train/dpo/llava/model/multimodal_encoder/builder.py
function build_vision_tower (line 5) | def build_vision_tower(vision_tower_cfg, **kwargs):
FILE: train/dpo/llava/model/multimodal_encoder/clip_encoder.py
class CLIPVisionTower (line 7) | class CLIPVisionTower(nn.Module):
method __init__ (line 8) | def __init__(self, vision_tower, args, delay_load=False):
method load_model (line 24) | def load_model(self, device_map=None):
method feature_select (line 35) | def feature_select(self, image_forward_outs):
method forward (line 46) | def forward(self, images):
method dummy_feature (line 60) | def dummy_feature(self):
method dtype (line 64) | def dtype(self):
method device (line 68) | def device(self):
method config (line 72) | def config(self):
method hidden_size (line 79) | def hidden_size(self):
method num_patches_per_side (line 83) | def num_patches_per_side(self):
method num_patches (line 87) | def num_patches(self):
FILE: train/dpo/llava/model/multimodal_projector/builder.py
class IdentityMap (line 6) | class IdentityMap(nn.Module):
method __init__ (line 7) | def __init__(self):
method forward (line 10) | def forward(self, x, *args, **kwargs):
method config (line 14) | def config(self):
class SimpleResBlock (line 18) | class SimpleResBlock(nn.Module):
method __init__ (line 19) | def __init__(self, channels):
method forward (line 28) | def forward(self, x):
function build_vision_projector (line 33) | def build_vision_projector(config, delay_load=False, **kwargs):
FILE: train/dpo/llava/model/utils.py
function auto_upgrade (line 4) | def auto_upgrade(config):
FILE: train/dpo/llava/serve/cli.py
function load_image (line 18) | def load_image(image_file):
function main (line 27) | def main(args):
FILE: train/dpo/llava/serve/controller.py
class DispatchMethod (line 28) | class DispatchMethod(Enum):
method from_str (line 33) | def from_str(cls, name):
class WorkerInfo (line 43) | class WorkerInfo:
function heart_beat_controller (line 51) | def heart_beat_controller(controller):
class Controller (line 57) | class Controller:
method __init__ (line 58) | def __init__(self, dispatch_method: str):
method register_worker (line 69) | def register_worker(self, worker_name: str, check_heart_beat: bool,
method get_worker_status (line 88) | def get_worker_status(self, worker_name: str):
method remove_worker (line 101) | def remove_worker(self, worker_name: str):
method refresh_all_workers (line 104) | def refresh_all_workers(self):
method list_models (line 112) | def list_models(self):
method get_worker_address (line 120) | def get_worker_address(self, model_name: str):
method receive_heart_beat (line 173) | def receive_heart_beat(self, worker_name: str, queue_length: int):
method remove_stable_workers_by_expiration (line 183) | def remove_stable_workers_by_expiration(self):
method worker_api_generate_stream (line 193) | def worker_api_generate_stream(self, params):
method worker_api_get_status (line 220) | def worker_api_get_status(self):
function register_worker (line 243) | async def register_worker(request: Request):
function refresh_all_workers (line 251) | async def refresh_all_workers():
function list_models (line 256) | async def list_models():
function get_worker_address (line 262) | async def get_worker_address(request: Request):
function receive_heart_beat (line 269) | async def receive_heart_beat(request: Request):
function worker_api_generate_stream (line 277) | async def worker_api_generate_stream(request: Request):
function worker_api_get_status (line 284) | async def worker_api_get_status(request: Request):
FILE: train/dpo/llava/serve/gradio_web_server.py
function get_conv_log_filename (line 32) | def get_conv_log_filename():
function get_model_list (line 38) | def get_model_list():
function load_demo (line 58) | def load_demo(url_params, request: gr.Request):
function load_demo_refresh_model_list (line 71) | def load_demo_refresh_model_list(request: gr.Request):
function vote_last_response (line 82) | def vote_last_response(state, vote_type, model_selector, request: gr.Req...
function upvote_last_response (line 94) | def upvote_last_response(state, model_selector, request: gr.Request):
function downvote_last_response (line 100) | def downvote_last_response(state, model_selector, request: gr.Request):
function flag_last_response (line 106) | def flag_last_response(state, model_selector, request: gr.Request):
function regenerate (line 112) | def regenerate(state, image_process_mode, request: gr.Request):
function clear_history (line 122) | def clear_history(request: gr.Request):
function add_text (line 128) | def add_text(state, text, image, image_process_mode, request: gr.Request):
function http_bot (line 154) | def http_bot(state, model_selector, temperature, top_p, max_new_tokens, ...
function build_demo (line 315) | def build_demo(embed_mode, cur_dir=None, concurrency_count=10):
FILE: train/dpo/llava/serve/model_worker.py
function heart_beat_worker (line 37) | def heart_beat_worker(controller):
class ModelWorker (line 44) | class ModelWorker:
method __init__ (line 45) | def __init__(self, controller_addr, worker_addr,
method register_to_controller (line 75) | def register_to_controller(self):
method send_heart_beat (line 87) | def send_heart_beat(self):
method get_queue_length (line 108) | def get_queue_length(self):
method get_status (line 115) | def get_status(self):
method generate_stream (line 123) | def generate_stream(self, params):
method generate_stream_gate (line 195) | def generate_stream_gate(self, params):
function release_model_semaphore (line 225) | def release_model_semaphore(fn=None):
function generate_stream (line 232) | async def generate_stream(request: Request):
function get_status (line 248) | async def get_status(request: Request):
FILE: train/dpo/llava/serve/sglang_worker.py
function heart_beat_worker (line 38) | def heart_beat_worker(controller):
function pipeline (line 45) | def pipeline(s, prompt, max_tokens):
class ModelWorker (line 54) | class ModelWorker:
method __init__ (line 55) | def __init__(self, controller_addr, worker_addr, sgl_endpoint,
method register_to_controller (line 85) | def register_to_controller(self):
method send_heart_beat (line 97) | def send_heart_beat(self):
method get_queue_length (line 118) | def get_queue_length(self):
method get_status (line 125) | def get_status(self):
method generate_stream (line 132) | async def generate_stream(self, params):
method generate_stream_gate (line 172) | async def generate_stream_gate(self, params):
function release_model_semaphore (line 195) | def release_model_semaphore(fn=None):
function generate_stream (line 202) | async def generate_stream(request: Request):
function get_status (line 218) | async def get_status(request: Request):
FILE: train/dpo/llava/serve/test_message.py
function main (line 9) | def main():
FILE: train/dpo/llava/train/llama_flash_attn_monkey_patch.py
function forward (line 16) | def forward(
function _prepare_decoder_attention_mask (line 98) | def _prepare_decoder_attention_mask(
function replace_llama_attn_with_flash_attn (line 105) | def replace_llama_attn_with_flash_attn():
FILE: train/dpo/llava/train/llama_xformers_attn_monkey_patch.py
function replace_llama_attn_with_xformers_attn (line 19) | def replace_llama_attn_with_xformers_attn():
function xformers_forward (line 23) | def xformers_forward(
FILE: train/dpo/llava/train/llava_trainer.py
function maybe_zero_3 (line 18) | def maybe_zero_3(param, ignore_status=False, name=None):
function get_mm_adapter_state_maybe_zero_3 (line 32) | def get_mm_adapter_state_maybe_zero_3(named_params, keys_to_match):
function split_to_even_chunks (line 38) | def split_to_even_chunks(indices, lengths, num_chunks):
function get_modality_length_grouped_indices (line 60) | def get_modality_length_grouped_indices(lengths, batch_size, world_size,...
function get_length_grouped_indices (line 88) | def get_length_grouped_indices(lengths, batch_size, world_size, generato...
class LengthGroupedSampler (line 99) | class LengthGroupedSampler(Sampler):
method __init__ (line 105) | def __init__(
method __len__ (line 122) | def __len__(self):
method __iter__ (line 125) | def __iter__(self):
class LLaVATrainer (line 133) | class LLaVATrainer(DPOTrainer):
method _get_train_sampler (line 135) | def _get_train_sampler(self) -> Optional[torch.utils.data.Sampler]:
method create_optimizer (line 150) | def create_optimizer(self):
method _save_checkpoint (line 239) | def _save_checkpoint(self, model, trial, metrics=None):
method _save (line 260) | def _save(self, output_dir: Optional[str] = None, state_dict=None):
FILE: train/dpo/llava/train/train.py
function rank0_print (line 44) | def rank0_print(*args):
class ModelArguments (line 54) | class ModelArguments:
class DataArguments (line 70) | class DataArguments:
class TrainingArguments (line 80) | class TrainingArguments(transformers.TrainingArguments):
function maybe_zero_3 (line 115) | def maybe_zero_3(param, ignore_status=False, name=None):
function get_peft_state_maybe_zero_3 (line 130) | def get_peft_state_maybe_zero_3(named_params, bias):
function get_peft_state_non_lora_maybe_zero_3 (line 155) | def get_peft_state_non_lora_maybe_zero_3(named_params, require_grad_only...
function get_mm_adapter_state_maybe_zero_3 (line 163) | def get_mm_adapter_state_maybe_zero_3(named_params, keys_to_match):
function find_all_linear_names (line 169) | def find_all_linear_names(model):
function safe_save_model_for_hf_trainer (line 185) | def safe_save_model_for_hf_trainer(trainer: transformers.Trainer,
function smart_tokenizer_and_embedding_resize (line 224) | def smart_tokenizer_and_embedding_resize(
function _tokenize_fn (line 249) | def _tokenize_fn(strings: Sequence[str],
function _mask_targets (line 276) | def _mask_targets(target, tokenized_lens, speakers):
function _add_speaker_and_signal (line 287) | def _add_speaker_and_signal(header, source, get_conversation=True):
function preprocess_multimodal (line 308) | def preprocess_multimodal(
function preprocess_llama_2 (line 332) | def preprocess_llama_2(
function preprocess_v1 (line 414) | def preprocess_v1(
function preprocess_mpt (line 500) | def preprocess_mpt(
function preprocess_plain (line 588) | def preprocess_plain(
function preprocess (line 610) | def preprocess(
class LazySupervisedDataset (line 658) | class LazySupervisedDataset(Dataset):
method __init__ (line 661) | def __init__(self, data_path: str,
method __len__ (line 672) | def __len__(self):
method lengths (line 676) | def lengths(self):
method modality_lengths (line 684) | def modality_lengths(self):
method __getitem__ (line 692) | def __getitem__(self, i) -> Dict[str, torch.Tensor]:
class DataCollatorForSupervisedDataset (line 743) | class DataCollatorForSupervisedDataset(object):
method __call__ (line 748) | def __call__(self, instances: Sequence[Dict]) -> Dict[str, torch.Tensor]:
function make_supervised_data_module (line 776) | def make_supervised_data_module(tokenizer: transformers.PreTrainedTokeni...
function train (line 788) | def train(attn_implementation=None):
FILE: train/dpo/llava/train/train_dpo.py
function rank0_print (line 46) | def rank0_print(*args):
function pil_to_tensor (line 50) | def pil_to_tensor(image):
function tensor_to_pil (line 54) | def tensor_to_pil(tensor):
function add_diffusion_noise (line 58) | def add_diffusion_noise(image_tensor, noise_step):
class ModelArguments (line 86) | class ModelArguments:
class DataArguments (line 102) | class DataArguments:
class TrainingArguments (line 112) | class TrainingArguments(transformers.TrainingArguments):
function maybe_zero_3 (line 147) | def maybe_zero_3(param, ignore_status=False, name=None):
function get_peft_state_maybe_zero_3 (line 162) | def get_peft_state_maybe_zero_3(named_params, bias):
function get_peft_state_non_lora_maybe_zero_3 (line 187) | def get_peft_state_non_lora_maybe_zero_3(named_params, require_grad_only...
function get_mm_adapter_state_maybe_zero_3 (line 195) | def get_mm_adapter_state_maybe_zero_3(named_params, keys_to_match):
function find_all_linear_names (line 201) | def find_all_linear_names(model):
function safe_save_model_for_hf_trainer (line 217) | def safe_save_model_for_hf_trainer(trainer: transformers.Trainer,
function smart_tokenizer_and_embedding_resize (line 256) | def smart_tokenizer_and_embedding_resize(
function _tokenize_fn (line 281) | def _tokenize_fn(strings: Sequence[str],
function _mask_targets (line 308) | def _mask_targets(target, tokenized_lens, speakers):
function _add_speaker_and_signal (line 319) | def _add_speaker_and_signal(header, source, get_conversation=True):
function preprocess_multimodal (line 340) | def preprocess_multimodal(
function preprocess_llama_2 (line 364) | def preprocess_llama_2(
function preprocess_v1 (line 446) | def preprocess_v1(
function preprocess_plain (line 544) | def preprocess_plain(
function preprocess (line 566) | def preprocess(
class LazySupervisedDataset (line 620) | class LazySupervisedDataset(Dataset):
method __init__ (line 623) | def __init__(self, data_path: str,
method __len__ (line 634) | def __len__(self):
method lengths (line 638) | def lengths(self):
method modality_lengths (line 646) | def modality_lengths(self):
method __getitem__ (line 654) | def __getitem__(self, i) -> Dict[str, torch.Tensor]:
class DataCollatorForSupervisedDataset (line 732) | class DataCollatorForSupervisedDataset(object):
method __call__ (line 737) | def __call__(self, instances: Sequence[Dict]) -> Dict[str, torch.Tensor]:
function make_supervised_data_module (line 794) | def make_supervised_data_module(tokenizer: transformers.PreTrainedTokeni...
function train (line 806) | def train():
FILE: train/dpo/llava/train/train_dpo_inherent.py
function rank0_print (line 45) | def rank0_print(*args):
function pil_to_tensor (line 49) | def pil_to_tensor(image):
function tensor_to_pil (line 53) | def tensor_to_pil(tensor):
function add_diffusion_noise (line 57) | def add_diffusion_noise(image_tensor, noise_step):
class ModelArguments (line 85) | class ModelArguments:
class DataArguments (line 100) | class DataArguments:
class TrainingArguments (line 110) | class TrainingArguments(transformers.TrainingArguments):
function maybe_zero_3 (line 145) | def maybe_zero_3(param, ignore_status=False, name=None):
function get_peft_state_maybe_zero_3 (line 160) | def get_peft_state_maybe_zero_3(named_params, bias):
function get_peft_state_non_lora_maybe_zero_3 (line 185) | def get_peft_state_non_lora_maybe_zero_3(named_params, require_grad_only...
function get_mm_adapter_state_maybe_zero_3 (line 193) | def get_mm_adapter_state_maybe_zero_3(named_params, keys_to_match):
function find_all_linear_names (line 199) | def find_all_linear_names(model):
function safe_save_model_for_hf_trainer (line 215) | def safe_save_model_for_hf_trainer(trainer: transformers.Trainer,
function smart_tokenizer_and_embedding_resize (line 254) | def smart_tokenizer_and_embedding_resize(
function _tokenize_fn (line 279) | def _tokenize_fn(strings: Sequence[str],
function _mask_targets (line 306) | def _mask_targets(target, tokenized_lens, speakers):
function _add_speaker_and_signal (line 317) | def _add_speaker_and_signal(header, source, get_conversation=True):
function preprocess_multimodal (line 338) | def preprocess_multimodal(
function preprocess_llama_2 (line 362) | def preprocess_llama_2(
function preprocess_v1 (line 444) | def preprocess_v1(
function preprocess_plain (line 542) | def preprocess_plain(
function preprocess (line 564) | def preprocess(
class LazySupervisedDataset (line 618) | class LazySupervisedDataset(Dataset):
method __init__ (line 621) | def __init__(self, data_path: str,
method __len__ (line 632) | def __len__(self):
method lengths (line 636) | def lengths(self):
method modality_lengths (line 644) | def modality_lengths(self):
method __getitem__ (line 652) | def __getitem__(self, i) -> Dict[str, torch.Tensor]:
class DataCollatorForSupervisedDataset (line 736) | class DataCollatorForSupervisedDataset(object):
method __call__ (line 741) | def __call__(self, instances: Sequence[Dict]) -> Dict[str, torch.Tensor]:
function make_supervised_data_module (line 804) | def make_supervised_data_module(tokenizer: transformers.PreTrainedTokeni...
function train (line 816) | def train():
FILE: train/dpo/llava/utils.py
function build_logger (line 17) | def build_logger(logger_name, logger_filename):
class StreamToLogger (line 60) | class StreamToLogger(object):
method __init__ (line 64) | def __init__(self, logger, log_level=logging.INFO):
method __getattr__ (line 70) | def __getattr__(self, attr):
method write (line 73) | def write(self, buf):
method flush (line 87) | def flush(self):
function disable_torch_init (line 93) | def disable_torch_init():
function violates_moderation (line 102) | def violates_moderation(text):
function pretty_print_semaphore (line 123) | def pretty_print_semaphore(semaphore):
FILE: train/dpo/llava_trainer_2stages.py
function maybe_zero_3 (line 18) | def maybe_zero_3(param, ignore_status=False, name=None):
function get_mm_adapter_state_maybe_zero_3 (line 32) | def get_mm_adapter_state_maybe_zero_3(named_params, keys_to_match):
function split_to_even_chunks (line 38) | def split_to_even_chunks(indices, lengths, num_chunks):
function get_modality_length_grouped_indices (line 60) | def get_modality_length_grouped_indices(lengths, batch_size, world_size,...
function get_length_grouped_indices (line 88) | def get_length_grouped_indices(lengths, batch_size, world_size, generato...
class LengthGroupedSampler (line 99) | class LengthGroupedSampler(Sampler):
method __init__ (line 105) | def __init__(
method __len__ (line 122) | def __len__(self):
method __iter__ (line 125) | def __iter__(self):
class LLaVATrainer (line 133) | class LLaVATrainer(DPOTrainer):
method _get_train_sampler (line 135) | def _get_train_sampler(self) -> Optional[torch.utils.data.Sampler]:
method create_optimizer (line 150) | def create_optimizer(self):
method _save_checkpoint (line 239) | def _save_checkpoint(self, model, trial, metrics=None):
method _save (line 260) | def _save(self, output_dir: Optional[str] = None, state_dict=None):
FILE: train/dpo/povid_infer.py
function split_list (line 23) | def split_list(lst, n):
function get_chunk (line 29) | def get_chunk(lst, n, k):
function eval_model (line 34) | def eval_model(args):
FILE: train/dpo/predict.py
function download_json (line 54) | def download_json(url: str, dest: Path):
function download_weights (line 62) | def download_weights(baseurl: str, basedest: str, files: list[str]):
class Predictor (line 78) | class Predictor(BasePredictor):
method setup (line 79) | def setup(self) -> None:
method predict (line 87) | def predict(
function load_image (line 148) | def load_image(image_file):
FILE: train/dpo/scripts/convert_mmbench_for_submission.py
function get_args (line 6) | def get_args():
FILE: train/dpo/scripts/convert_seed_for_submission.py
function get_args (line 6) | def get_args():
function eval_single (line 14) | def eval_single(result_file, eval_only_type=None):
FILE: train/dpo/scripts/convert_sqa_to_llava.py
function convert_to_llava (line 8) | def convert_to_llava(base_dir, split, prompt_format="QCM-LEA"):
function convert_to_jsonl (line 49) | def convert_to_jsonl(base_dir, split, prompt_format="QCM-LEPA"):
function main (line 83) | def main(task, **kwargs):
FILE: train/dpo/scripts/convert_sqa_to_llava_base_prompt.py
function get_question_text (line 1) | def get_question_text(problem):
function get_context_text (line 6) | def get_context_text(problem, use_caption):
function get_choice_text (line 15) | def get_choice_text(probelm, options):
function get_answer (line 25) | def get_answer(problem, options):
function get_lecture_text (line 29) | def get_lecture_text(problem):
function get_solution_text (line 35) | def get_solution_text(problem):
function create_one_example_chatbot (line 41) | def create_one_example_chatbot(format, question, context, choice, answer...
function create_one_example (line 106) | def create_one_example(format, question, context, choice, answer, lectur...
function create_one_example_gpt4 (line 162) | def create_one_example_gpt4(format, question, context, choice, answer, l...
function build_prompt_chatbot (line 221) | def build_prompt_chatbot(problems, shot_qids, prompt_format, use_caption...
function build_prompt (line 244) | def build_prompt(problems, shot_qids, test_qid, args):
function build_prompt_gpt4 (line 291) | def build_prompt_gpt4(problems, shot_qids, test_qid, args):
FILE: train/dpo/scripts/convert_vizwiz_for_submission.py
function parse_args (line 8) | def parse_args():
FILE: train/dpo/scripts/convert_vqav2_for_submission.py
function parse_args (line 8) | def parse_args():
FILE: train/dpo/scripts/extract_mm_projector.py
function parse_args (line 15) | def parse_args():
FILE: train/dpo/scripts/merge_lora_weights.py
function merge_lora (line 6) | def merge_lora(args):
FILE: train/dpo/tool/dpo_trainer.py
class DPOTrainer (line 60) | class DPOTrainer(Trainer):
method __init__ (line 123) | def __init__(
method _prepare_deepspeed (line 354) | def _prepare_deepspeed(self, model: PreTrainedModelWrapper):
method concatenated_inputs (line 385) | def concatenated_inputs(self, batch: Dict[str, Union[List, torch.LongT...
method dpo_loss (line 424) | def dpo_loss(
method _get_batch_logps (line 467) | def _get_batch_logps(
method _get_noisy_batch_logps (line 501) | def _get_noisy_batch_logps(
method concatenated_forward (line 550) | def concatenated_forward(
method get_batch_metrics (line 629) | def get_batch_metrics(
method compute_loss (line 681) | def compute_loss(
method get_batch_samples (line 702) | def get_batch_samples(self, model, batch: Dict[str, torch.LongTensor])...
method prediction_step (line 739) | def prediction_step(
method store_metrics (line 778) | def store_metrics(self, metrics: Dict[str, float], train_eval: Literal...
method evaluation_loop (line 782) | def evaluation_loop(
method log (line 832) | def log(self, logs: Dict[str, float]) -> None:
FILE: train/dpo/tool/dpo_trainer_inherent.py
class DPOTrainer (line 56) | class DPOTrainer(Trainer):
method __init__ (line 119) | def __init__(
method _prepare_deepspeed (line 350) | def _prepare_deepspeed(self, model: PreTrainedModelWrapper):
method concatenated_inputs (line 381) | def concatenated_inputs(self, batch: Dict[str, Union[List, torch.LongT...
method dpo_loss (line 420) | def dpo_loss(
method _get_batch_logps (line 463) | def _get_batch_logps(
method _get_noisy_batch_logps (line 497) | def _get_noisy_batch_logps(
method concatenated_multi_forward (line 546) | def concatenated_multi_forward(
method get_batch_metrics (line 641) | def get_batch_metrics(
method compute_loss (line 704) | def compute_loss(
method get_batch_samples (line 725) | def get_batch_samples(self, model, batch: Dict[str, torch.LongTensor])...
method prediction_step (line 762) | def prediction_step(
method store_metrics (line 801) | def store_metrics(self, metrics: Dict[str, float], train_eval: Literal...
method evaluation_loop (line 805) | def evaluation_loop(
method log (line 855) | def log(self, logs: Dict[str, float]) -> None:
FILE: train/dpo/train_dpo_2stages.py
function rank0_print (line 72) | def rank0_print(*args):
function pil_to_tensor (line 77) | def pil_to_tensor(image):
function tensor_to_pil (line 82) | def tensor_to_pil(tensor):
function add_gaussian_noise (line 86) | def add_gaussian_noise(image, mean=0.0, stddev=0.1):
function generate_gaussian_noise (line 93) | def generate_gaussian_noise(image, mean=0.0, stddev=0.1):
function add_diffusion_noise (line 99) | def add_diffusion_noise(image_tensor, noise_step):
class ModelArguments (line 130) | class ModelArguments:
class DataArguments (line 149) | class DataArguments:
class TrainingArguments (line 160) | class TrainingArguments(transformers.TrainingArguments):
function maybe_zero_3 (line 195) | def maybe_zero_3(param, ignore_status=False, name=None):
function get_peft_state_maybe_zero_3 (line 213) | def get_peft_state_maybe_zero_3(named_params, bias):
function get_peft_state_non_lora_maybe_zero_3 (line 238) | def get_peft_state_non_lora_maybe_zero_3(named_params, require_grad_only...
function get_mm_adapter_state_maybe_zero_3 (line 248) | def get_mm_adapter_state_maybe_zero_3(named_params, keys_to_match):
function find_all_linear_names (line 260) | def find_all_linear_names(model):
function safe_save_model_for_hf_trainer (line 276) | def safe_save_model_for_hf_trainer(trainer: transformers.Trainer, output...
function smart_tokenizer_and_embedding_resize (line 318) | def smart_tokenizer_and_embedding_resize(
function _tokenize_fn (line 345) | def _tokenize_fn(
function _mask_targets (line 372) | def _mask_targets(target, tokenized_lens, speakers):
function _add_speaker_and_signal (line 383) | def _add_speaker_and_signal(header, source, get_conversation=True):
function preprocess_multimodal (line 405) | def preprocess_multimodal(sources: Sequence[str], data_args: DataArgumen...
function preprocess_llama_2 (line 435) | def preprocess_llama_2(
function preprocess_v1 (line 521) | def preprocess_v1(
function preprocess_plain (line 634) | def preprocess_plain(
function preprocess (line 663) | def preprocess(
class LazySupervisedDataset (line 730) | class LazySupervisedDataset(Dataset):
method __init__ (line 733) | def __init__(
method __len__ (line 747) | def __len__(self):
method lengths (line 751) | def lengths(self):
method modality_lengths (line 762) | def modality_lengths(self):
method __getitem__ (line 772) | def __getitem__(self, i) -> Dict[str, torch.Tensor]:
class DataCollatorForSupervisedDataset (line 895) | class DataCollatorForSupervisedDataset(object):
method __call__ (line 900) | def __call__(self, instances: Sequence[Dict]) -> Dict[str, torch.Tensor]:
function make_supervised_data_module (line 967) | def make_supervised_data_module(
function train (line 980) | def train():
FILE: train/open_clip/setup.py
function _read_reqs (line 13) | def _read_reqs(relpath):
FILE: train/open_clip/src/open_clip/big_vision.py
function load_big_vision_weights (line 9) | def load_big_vision_weights(model: CustomTextCLIP, checkpoint_path: str):
FILE: train/open_clip/src/open_clip/coca_model.py
class MultimodalCfg (line 45) | class MultimodalCfg(CLIPTextCfg):
function _build_text_decoder_tower (line 53) | def _build_text_decoder_tower(
class CoCa (line 79) | class CoCa(nn.Module):
method __init__ (line 80) | def __init__(
method set_grad_checkpointing (line 134) | def set_grad_checkpointing(self, enable: bool = True):
method _encode_image (line 139) | def _encode_image(self, images, normalize: bool = True):
method _encode_text (line 144) | def _encode_text(self, text, normalize: bool = True):
method encode_image (line 149) | def encode_image(self, images, normalize: bool = True):
method encode_text (line 153) | def encode_text(self, text, normalize: bool = True):
method forward (line 157) | def forward(
method generate (line 187) | def generate(
method _generate_beamsearch (line 310) | def _generate_beamsearch(
function prepare_inputs_for_generation (line 459) | def prepare_inputs_for_generation(input_ids, image_inputs, past=None, **...
FILE: train/open_clip/src/open_clip/factory.py
function _natural_key (line 28) | def _natural_key(string_):
function _rescan_model_configs (line 32) | def _rescan_model_configs():
function list_models (line 56) | def list_models():
function add_model_config (line 61) | def add_model_config(path):
function get_model_config (line 69) | def get_model_config(model_name):
function _get_hf_config (line 76) | def _get_hf_config(model_id, cache_dir=None):
function get_tokenizer (line 83) | def get_tokenizer(
function load_state_dict (line 127) | def load_state_dict(checkpoint_path: str, map_location='cpu'):
function load_checkpoint (line 142) | def load_checkpoint(model, checkpoint_path, strict=True):
function create_model (line 165) | def create_model(
function create_loss (line 324) | def create_loss(args):
function create_model_and_transforms (line 361) | def create_model_and_transforms(
function create_model_from_pretrained (line 418) | def create_model_from_pretrained(
FILE: train/open_clip/src/open_clip/hf_model.py
class BaseModelOutput (line 20) | class BaseModelOutput:
class PretrainedConfig (line 24) | class PretrainedConfig:
function _camel2snake (line 31) | def _camel2snake(s):
function register_pooler (line 39) | def register_pooler(cls):
class MeanPooler (line 46) | class MeanPooler(nn.Module):
method forward (line 49) | def forward(self, x: BaseModelOutput, attention_mask: TensorType):
class MaxPooler (line 55) | class MaxPooler(nn.Module):
method forward (line 58) | def forward(self, x: BaseModelOutput, attention_mask: TensorType):
class ClsPooler (line 64) | class ClsPooler(nn.Module):
method __init__ (line 67) | def __init__(self, use_pooler_output=True):
method forward (line 72) | def forward(self, x: BaseModelOutput, attention_mask: TensorType):
class ClsLastHiddenStatePooler (line 83) | class ClsLastHiddenStatePooler(nn.Module):
method __init__ (line 88) | def __init__(self):
method forward (line 92) | def forward(self, x: BaseModelOutput, attention_mask: TensorType):
class HFTextEncoder (line 96) | class HFTextEncoder(nn.Module):
method __init__ (line 100) | def __init__(
method forward (line 154) | def forward(self, x: TensorType):
method lock (line 171) | def lock(self, unlocked_layers: int = 0, freeze_layer_norm: bool = True):
method set_grad_checkpointing (line 189) | def set_grad_checkpointing(self, enable=True):
method init_parameters (line 192) | def init_parameters(self):
FILE: train/open_clip/src/open_clip/loss.py
function gather_features (line 19) | def gather_features(
class ClipLoss (line 66) | class ClipLoss(nn.Module):
method __init__ (line 68) | def __init__(
method get_ground_truth (line 89) | def get_ground_truth(self, device, num_logits) -> torch.Tensor:
method get_logits (line 102) | def get_logits(self, image_features, text_features, logit_scale):
method forward (line 120) | def forward(self, image_features, text_features, logit_scale, output_d...
class CoCaLoss (line 134) | class CoCaLoss(ClipLoss):
method __init__ (line 135) | def __init__(
method forward (line 160) | def forward(self, image_features, text_features, logits, labels, logit...
class DistillClipLoss (line 180) | class DistillClipLoss(ClipLoss):
method dist_loss (line 182) | def dist_loss(self, teacher_logits, student_logits):
method forward (line 185) | def forward(
function neighbour_exchange (line 219) | def neighbour_exchange(from_rank, to_rank, tensor, group=None):
function neighbour_exchange_bidir (line 239) | def neighbour_exchange_bidir(left_rank, right_rank, tensor_to_left, tens...
class NeighbourExchange (line 272) | class NeighbourExchange(torch.autograd.Function):
method forward (line 274) | def forward(ctx, from_rank, to_rank, group, tensor):
method backward (line 281) | def backward(ctx, grad_output):
function neighbour_exchange_with_grad (line 285) | def neighbour_exchange_with_grad(from_rank, to_rank, tensor, group=None):
class NeighbourExchangeBidir (line 289) | class NeighbourExchangeBidir(torch.autograd.Function):
method forward (line 291) | def forward(ctx, left_rank, right_rank, group, tensor_to_left, tensor_...
method backward (line 298) | def backward(ctx, *grad_outputs):
function neighbour_exchange_bidir_with_grad (line 303) | def neighbour_exchange_bidir_with_grad(left_rank, right_rank, tensor_to_...
class SigLipLoss (line 307) | class SigLipLoss(nn.Module):
method __init__ (line 317) | def __init__(
method get_ground_truth (line 337) | def get_ground_truth(self, device, dtype, num_logits, negative_only=Fa...
method get_logits (line 343) | def get_logits(self, image_features, text_features, logit_scale, logit...
method _loss (line 349) | def _loss(self, image_features, text_features, logit_scale, logit_bias...
method forward (line 360) | def forward(self, image_features, text_features, logit_scale, logit_bi...
FILE: train/open_clip/src/open_clip/model.py
class CLIPVisionCfg (line 27) | class CLIPVisionCfg:
class CLIPTextCfg (line 58) | class CLIPTextCfg:
function get_cast_dtype (line 86) | def get_cast_dtype(precision: str):
function get_input_dtype (line 95) | def get_input_dtype(precision: str):
function _build_vision_tower (line 104) | def _build_vision_tower(
function _build_text_tower (line 173) | def _build_text_tower(
class CLIP (line 220) | class CLIP(nn.Module):
method __init__ (line 223) | def __init__(
method lock_image_tower (line 256) | def lock_image_tower(self, unlocked_groups=0, freeze_bn_stats=False):
method set_grad_checkpointing (line 261) | def set_grad_checkpointing(self, enable=True):
method encode_image (line 265) | def encode_image(self, image, normalize: bool = False):
method encode_text (line 269) | def encode_text(self, text, normalize: bool = False):
method get_logits (line 288) | def get_logits(self, image, text):
method forward (line 297) | def forward(
class CustomTextCLIP (line 320) | class CustomTextCLIP(nn.Module):
method __init__ (line 323) | def __init__(
method lock_image_tower (line 346) | def lock_image_tower(self, unlocked_groups=0, freeze_bn_stats=False):
method lock_text_tower (line 350) | def lock_text_tower(self, unlocked_layers: int = 0, freeze_layer_norm:...
method set_grad_checkpointing (line 354) | def set_grad_checkpointing(self, enable=True):
method encode_image (line 358) | def encode_image(self, image, normalize: bool = False):
method encode_text (line 362) | def encode_text(self, text, normalize: bool = False):
method get_logits (line 366) | def get_logits(self, image, text):
method forward (line 375) | def forward(
function convert_weights_to_lp (line 398) | def convert_weights_to_lp(model: nn.Module, dtype=torch.float16):
function convert_to_custom_text_state_dict (line 432) | def convert_to_custom_text_state_dict(state_dict: dict):
function build_model_from_openai_state_dict (line 450) | def build_model_from_openai_state_dict(
function trace_model (line 509) | def trace_model(model, batch_size=256, device=torch.device('cpu')):
function resize_pos_embed (line 525) | def resize_pos_embed(state_dict, model, interpolation: str = 'bicubic', ...
function resize_text_pos_embed (line 559) | def resize_text_pos_embed(state_dict, model, interpolation: str = 'linea...
function get_model_preprocess_cfg (line 591) | def get_model_preprocess_cfg(model):
function set_model_preprocess_cfg (line 608) | def set_model_preprocess_cfg(model, preprocess_cfg: Dict[str, Any]):
function get_model_tokenize_cfg (line 615) | def get_model_tokenize_cfg(model):
FILE: train/open_clip/src/open_clip/modified_resnet.py
class Bottleneck (line 10) | class Bottleneck(nn.Module):
method __init__ (line 13) | def __init__(self, inplanes, planes, stride=1):
method forward (line 42) | def forward(self, x: torch.Tensor):
class AttentionPool2d (line 58) | class AttentionPool2d(nn.Module):
method __init__ (line 59) | def __init__(self, spacial_dim: int, embed_dim: int, num_heads: int, o...
method forward (line 68) | def forward(self, x):
class ModifiedResNet (line 95) | class ModifiedResNet(nn.Module):
method __init__ (line 103) | def __init__(self, layers, output_dim, heads, image_size=224, width=64):
method _make_layer (line 132) | def _make_layer(self, planes, blocks, stride=1):
method init_parameters (line 141) | def init_parameters(self):
method lock (line 154) | def lock(self, unlocked_groups=0, freeze_bn_stats=False):
method set_grad_checkpointing (line 162) | def set_grad_checkpointing(self, enable=True):
method stem (line 166) | def stem(self, x):
method forward (line 173) | def forward(self, x):
FILE: train/open_clip/src/open_clip/openai.py
function list_openai_models (line 19) | def list_openai_models() -> List[str]:
function load_openai_model (line 24) | def load_openai_model(
FILE: train/open_clip/src/open_clip/pos_embed.py
function get_2d_sincos_pos_embed (line 20) | def get_2d_sincos_pos_embed(embed_dim, grid_size, cls_token=False):
function get_2d_sincos_pos_embed_from_grid (line 38) | def get_2d_sincos_pos_embed_from_grid(embed_dim, grid):
function get_1d_sincos_pos_embed_from_grid (line 49) | def get_1d_sincos_pos_embed_from_grid(embed_dim, pos):
function interpolate_pos_embed (line 75) | def interpolate_pos_embed(model, checkpoint_model):
FILE: train/open_clip/src/open_clip/pretrained.py
function _pcfg (line 29) | def _pcfg(url='', hf_hub='', **kwargs):
function _slpcfg (line 42) | def _slpcfg(url='', hf_hub='', **kwargs):
function _apcfg (line 55) | def _apcfg(url='', hf_hub='', **kwargs):
function _clean_tag (line 445) | def _clean_tag(tag: str):
function list_pretrained (line 450) | def list_pretrained(as_str: bool = False):
function list_pretrained_models_by_tag (line 457) | def list_pretrained_models_by_tag(tag: str):
function list_pretrained_tags_by_model (line 467) | def list_pretrained_tags_by_model(model: str):
function is_pretrained_cfg (line 475) | def is_pretrained_cfg(model: str, tag: str):
function get_pretrained_cfg (line 481) | def get_pretrained_cfg(model: str, tag: str):
function get_pretrained_url (line 488) | def get_pretrained_url(model: str, tag: str):
function download_pretrained_from_url (line 493) | def download_pretrained_from_url(
function has_hf_hub (line 539) | def has_hf_hub(necessary=False):
function download_pretrained_from_hf (line 547) | def download_pretrained_from_hf(
function download_pretrained (line 558) | def download_pretrained(
FILE: train/open_clip/src/open_clip/push_to_hf_hub.py
function save_config_for_hf (line 40) | def save_config_for_hf(
function save_for_hf (line 63) | def save_for_hf(
function push_to_hf_hub (line 90) | def push_to_hf_hub(
function push_pretrained_to_hf_hub (line 160) | def push_pretrained_to_hf_hub(
function generate_readme (line 209) | def generate_readme(model_card: dict, model_name: str):
FILE: train/open_clip/src/open_clip/timm_model.py
class TimmModel (line 28) | class TimmModel(nn.Module):
method __init__ (line 32) | def __init__(
method lock (line 110) | def lock(self, unlocked_groups=0, freeze_bn_stats=False):
method set_grad_checkpointing (line 143) | def set_grad_checkpointing(self, enable=True):
method forward (line 149) | def forward(self, x):
FILE: train/open_clip/src/open_clip/tokenizer.py
function default_bpe (line 27) | def default_bpe():
function bytes_to_unicode (line 32) | def bytes_to_unicode():
function get_pairs (line 54) | def get_pairs(word):
function basic_clean (line 66) | def basic_clean(text):
function whitespace_clean (line 72) | def whitespace_clean(text):
function _clean_canonicalize (line 78) | def _clean_canonicalize(x):
function _clean_lower (line 83) | def _clean_lower(x):
function _clean_whitespace (line 88) | def _clean_whitespace(x):
function get_clean_fn (line 93) | def get_clean_fn(type: str):
function canonicalize_text (line 104) | def canonicalize_text(text, *, keep_punctuation_exact_string=None):
class SimpleTokenizer (line 127) | class SimpleTokenizer(object):
method __init__ (line 128) | def __init__(
method bpe (line 166) | def bpe(self, token):
method encode (line 207) | def encode(self, text):
method decode (line 215) | def decode(self, tokens):
method __call__ (line 220) | def __call__(self, texts: Union[str, List[str]], context_length: Optio...
function decode (line 265) | def decode(output_ids: torch.Tensor):
function tokenize (line 270) | def tokenize(texts: Union[str, List[str]], context_length: int = DEFAULT...
function random_mask_tokenize (line 274) | def random_mask_tokenize(
function simple_mask_tokenize (line 303) | def simple_mask_tokenize(
function syntax_mask_tokenize (line 325) | def syntax_mask_tokenize(
function get_reduction_mask_fn (line 384) | def get_reduction_mask_fn(type: str):
class HFTokenizer (line 397) | class HFTokenizer:
method __init__ (line 400) | def __init__(
method save_pretrained (line 419) | def save_pretrained(self, dest):
method __call__ (line 422) | def __call__(self, texts: Union[str, List[str]], context_length: Optio...
method set_language (line 449) | def set_language(self, src_lang):
class SigLipTokenizer (line 456) | class SigLipTokenizer:
method __init__ (line 466) | def __init__(
method save_pretrained (line 490) | def save_pretrained(self, dest):
method __call__ (line 493) | def __call__(self, texts: Union[str, List[str]], context_length: Optio...
FILE: train/open_clip/src/open_clip/transform.py
class PreprocessCfg (line 17) | class PreprocessCfg:
method __post_init__ (line 26) | def __post_init__(self):
method num_channels (line 30) | def num_channels(self):
method input_size (line 34) | def input_size(self):
function merge_preprocess_dict (line 40) | def merge_preprocess_dict(
function merge_preprocess_kwargs (line 57) | def merge_preprocess_kwargs(base: PreprocessCfg, **kwargs):
class AugmentationCfg (line 62) | class AugmentationCfg:
function _setup_size (line 75) | def _setup_size(size, error_msg):
class ResizeKeepRatio (line 88) | class ResizeKeepRatio:
method __init__ (line 94) | def __init__(
method get_params (line 116) | def get_params(
method __call__ (line 144) | def __call__(self, img):
method __repr__ (line 160) | def __repr__(self):
function center_crop_or_pad (line 167) | def center_crop_or_pad(img: torch.Tensor, output_size: List[int], fill=0...
class CenterCropOrPad (line 207) | class CenterCropOrPad(torch.nn.Module):
method __init__ (line 219) | def __init__(self, size, fill=0):
method forward (line 224) | def forward(self, img):
method __repr__ (line 234) | def __repr__(self) -> str:
function _convert_to_rgb (line 238) | def _convert_to_rgb(image):
class color_jitter (line 242) | class color_jitter(object):
method __init__ (line 246) | def __init__(self, brightness=0., contrast=0., saturation=0., hue=0., ...
method __call__ (line 251) | def __call__(self, img):
class gray_scale (line 258) | class gray_scale(object):
method __init__ (line 262) | def __init__(self, p=0.2):
method __call__ (line 267) | def __call__(self, img):
function image_transform (line 274) | def image_transform(
function image_transform_v2 (line 393) | def image_transform_v2(
FILE: train/open_clip/src/open_clip/transformer.py
class LayerNormFp32 (line 15) | class LayerNormFp32(nn.LayerNorm):
method forward (line 18) | def forward(self, x: torch.Tensor):
class LayerNorm (line 24) | class LayerNorm(nn.LayerNorm):
method forward (line 27) | def forward(self, x: torch.Tensor):
class QuickGELU (line 33) | class QuickGELU(nn.Module):
method forward (line 35) | def forward(self, x: torch.Tensor):
class LayerScale (line 39) | class LayerScale(nn.Module):
method __init__ (line 40) | def __init__(self, dim, init_values=1e-5, inplace=False):
method forward (line 45) | def forward(self, x):
class PatchDropout (line 49) | class PatchDropout(nn.Module):
method __init__ (line 54) | def __init__(self, prob, exclude_first_token=True):
method forward (line 60) | def forward(self, x):
class Attention (line 89) | class Attention(nn.Module):
method __init__ (line 90) | def __init__(
method forward (line 129) | def forward(self, x, attn_mask: Optional[torch.Tensor] = None):
class AttentionalPooler (line 165) | class AttentionalPooler(nn.Module):
method __init__ (line 166) | def __init__(
method forward (line 180) | def forward(self, x: torch.Tensor):
class ResidualAttentionBlock (line 188) | class ResidualAttentionBlock(nn.Module):
method __init__ (line 189) | def __init__(
method attention (line 216) | def attention(
method forward (line 231) | def forward(
class CustomResidualAttentionBlock (line 246) | class CustomResidualAttentionBlock(nn.Module):
method __init__ (line 247) | def __init__(
method forward (line 281) | def forward(self, x: torch.Tensor, attn_mask: Optional[torch.Tensor] =...
function _expand_token (line 287) | def _expand_token(token, batch_size: int):
class Transformer (line 291) | class Transformer(nn.Module):
method __init__ (line 292) | def __init__(
method get_cast_dtype (line 313) | def get_cast_dtype(self) -> torch.dtype:
method forward (line 318) | def forward(self, x: torch.Tensor, attn_mask: Optional[torch.Tensor] =...
class VisionTransformer (line 328) | class VisionTransformer(nn.Module):
method __init__ (line 331) | def __init__(
method lock (line 435) | def lock(self, unlocked_groups=0, freeze_bn_stats=False):
method init_parameters (line 468) | def init_parameters(self):
method set_grad_checkpointing (line 489) | def set_grad_checkpointing(self, enable=True):
method _global_pool (line 492) | def _global_pool(self, x: torch.Tensor) -> Tuple[torch.Tensor, torch.T...
method forward (line 502) | def forward(self, x: torch.Tensor):
function text_global_pool (line 550) | def text_global_pool(x, text: Optional[torch.Tensor] = None, pool_type: ...
class TextTransformer (line 565) | class TextTransformer(nn.Module):
method __init__ (line 568) | def __init__(
method init_parameters (line 628) | def init_parameters(self):
method set_grad_checkpointing (line 652) | def set_grad_checkpointing(self, enable=True):
method build_causal_mask (line 655) | def build_causal_mask(self):
method build_cls_mask (line 663) | def build_cls_mask(self, text, cast_dtype: torch.dtype):
method forward (line 672) | def forward(self, text):
class MultimodalTransformer (line 711) | class MultimodalTransformer(Transformer):
method __init__ (line 712) | def __init__(
method init_parameters (line 753) | def init_parameters(self):
method build_attention_mask (line 771) | def build_attention_mask(self):
method forward (line 779) | def forward(self, image_embs, text_embs):
method set_grad_checkpointing (line 802) | def set_grad_checkpointing(self, enable=True):
FILE: train/open_clip/src/open_clip/utils.py
function freeze_batch_norm_2d (line 9) | def freeze_batch_norm_2d(module, module_match={}, name=''):
function _ntuple (line 49) | def _ntuple(n):
function replace_linear (line 65) | def replace_linear(model, linear_replacement, include_modules=['c_fc', '...
function convert_int8_model_to_inference_mode (line 84) | def convert_int8_model_to_inference_mode(model):
FILE: train/open_clip/src/open_clip/zero_shot_classifier.py
function batched (line 9) | def batched(iterable, n):
function build_zero_shot_classifier (line 21) | def build_zero_shot_classifier(
function build_zero_shot_classifier_legacy (line 71) | def build_zero_shot_classifier_legacy(
FILE: train/open_clip/src/retrieve_clip_VQA.py
function retrieve_topk_per_image (line 15) | def retrieve_topk_per_image(logits, val_k_list,retrieve_threshold=''):
function get_logits (line 55) | def get_logits(image_features, text_features, logit_scale):
function clean_data_info (line 62) | def clean_data_info(data_info):
function split_and_clean_data_infos (line 75) | def split_and_clean_data_infos(batch_data_infos):
function main (line 96) | def main(args):
FILE: train/open_clip/src/retrieve_clip_report.py
function retrieve_topk_per_image (line 16) | def retrieve_topk_per_image(logits, val_k_list,retrieve_threshold=''):
function get_logits (line 56) | def get_logits(image_features, text_features, logit_scale):
function clean_data_info (line 63) | def clean_data_info(data_info):
function split_and_clean_data_infos (line 76) | def split_and_clean_data_infos(batch_data_infos):
function main (line 97) | def main(args):
FILE: train/open_clip/src/training/data.py
class CsvDataset (line 41) | class CsvDataset(Dataset):
method __init__ (line 42) | def __init__(
method __len__ (line 55) | def __len__(self):
method __getitem__ (line 58) | def __getitem__(self, idx):
class IUXrayDataset (line 64) | class IUXrayDataset(Dataset): # TODO
method __init__ (line 65) | def __init__(
method __len__ (line 116) | def __len__(self):
method __getitem__ (line 122) | def __getitem__(self, idx):
class PubMedVisionDataset (line 138) | class PubMedVisionDataset(Dataset):#TODO
method __init__ (line 139) | def __init__(self, img_root,json_file, transforms, tokenizer=None,load...
method __len__ (line 177) | def __len__(self):
method __getitem__ (line 183) | def __getitem__(self, idx):
class RadiologyDataset (line 200) | class RadiologyDataset(Dataset): # TODO
method __init__ (line 201) | def __init__(
method __len__ (line 241) | def __len__(self):
method __getitem__ (line 246) | def __getitem__(self, idx):
class PathologyDataset (line 258) | class PathologyDataset(Dataset): # TODO
method __init__ (line 259) | def __init__(
method __len__ (line 298) | def __len__(self):
method __getitem__ (line 303) | def __getitem__(self, idx):
class MimicVQADataset (line 316) | class MimicVQADataset(Dataset): #
method __init__ (line 317) | def __init__(
method __len__ (line 357) | def __len__(self):
method __getitem__ (line 363) | def __getitem__(self, idx):
class IUXrayVQADataset (line 375) | class IUXrayVQADataset(Dataset): # TODO
method __init__ (line 376) | def __init__(
method __len__ (line 416) | def __len__(self):
method __getitem__ (line 422) | def __getitem__(self, idx):
class pmc_oa_VQADataset (line 435) | class pmc_oa_VQADataset(Dataset): # TODO
method __init__ (line 436) | def __init__(
method __len__ (line 478) | def __len__(self):
method __getitem__ (line 484) | def __getitem__(self, idx):
class HarvardVQADataset (line 496) | class HarvardVQADataset(Dataset): # TODO
method __init__ (line 497) | def __init__(
method __len__ (line 537) | def __len__(self):
method __getitem__ (line 543) | def __getitem__(self, idx):
class quilt_1m_VQADataset (line 555) | class quilt_1m_VQADataset(Dataset): # TODO
method __init__ (line 556) | def __init__(
method __len__ (line 596) | def __len__(self):
method __getitem__ (line 602) | def __getitem__(self, idx):
class quilt_1m_Dataset (line 615) | class quilt_1m_Dataset(Dataset): # TODO
method __init__ (line 616) | def __init__(
method __len__ (line 657) | def __len__(self):
method __getitem__ (line 663) | def __getitem__(self, idx):
class pmc_oa_Dataset (line 683) | class pmc_oa_Dataset(Dataset): # TODO
method __init__ (line 684) | def __init__(
method __len__ (line 729) | def __len__(self):
method __getitem__ (line 735) | def __getitem__(self, idx):
class IUXrayVQADataset_with_conf (line 756) | class IUXrayVQADataset_with_conf(Dataset): # TODO
method __init__ (line 757) | def __init__(
method __len__ (line 828) | def __len__(self):
method __getitem__ (line 834) | def __getitem__(self, idx):
method map_confidence_to_k (line 847) | def map_confidence_to_k(self, confidence):
class MimicDataset (line 874) | class MimicDataset(Dataset): # TODO
method __init__ (line 875) | def __init__(
method __len__ (line 922) | def __len__(self):
method __getitem__ (line 928) | def __getitem__(self, idx):
class HarvardDataset (line 945) | class HarvardDataset(Dataset): # TODO
method __init__ (line 946) | def __init__(
method __len__ (line 997) | def __len__(self):
method __getitem__ (line 1003) | def __getitem__(self, idx):
class SharedEpoch (line 1021) | class SharedEpoch:
method __init__ (line 1022) | def __init__(self, epoch: int = 0):
method set_value (line 1025) | def set_value(self, epoch):
method get_value (line 1028) | def get_value(self):
class DataInfo (line 1033) | class DataInfo:
method set_epoch (line 1038) | def set_epoch(self, epoch):
function expand_urls (line 1045) | def expand_urls(urls, weights=None):
function get_dataset_size (line 1068) | def get_dataset_size(shards):
function get_imagenet (line 1090) | def get_imagenet(args, preprocess_fns, split):
function count_samples (line 1137) | def count_samples(dataloader):
function filter_no_caption_or_no_image (line 1147) | def filter_no_caption_or_no_image(sample):
function log_and_continue (line 1155) | def log_and_continue(exn):
function group_by_keys_nothrow (line 1161) | def group_by_keys_nothrow(
function tarfile_to_samples_nothrow (line 1195) | def tarfile_to_samples_nothrow(src, handler=log_and_continue):
function pytorch_worker_seed (line 1203) | def pytorch_worker_seed(increment=0):
class detshuffle2 (line 1223) | class detshuffle2(wds.PipelineStage):
method __init__ (line 1224) | def __init__(
method run (line 1236) | def run(self, src):
class ResampledShards2 (line 1255) | class ResampledShards2(IterableDataset):
method __init__ (line 1258) | def __init__(
method __iter__ (line 1286) | def __iter__(self):
function get_wds_dataset (line 1312) | def get_wds_dataset(
function get_csv_dataset (line 1451) | def get_csv_dataset(args, preprocess_fn, is_train, epoch=0, tokenizer=No...
function get_pmc_oa_Dataset (line 1480) | def get_pmc_oa_Dataset(
function get_quilt_1m_Dataset (line 1520) | def get_quilt_1m_Dataset(
function get_HarvardDataset (line 1560) | def get_HarvardDataset(
function get_IUXrayDataset (line 1601) | def get_IUXrayDataset(
function get_MimicDataset (line 1642) | def get_MimicDataset(
class SyntheticDataset (line 1683) | class SyntheticDataset(Dataset):
method __init__ (line 1684) | def __init__(
method __len__ (line 1700) | def __len__(self):
method __getitem__ (line 1703) | def __getitem__(self, idx):
function get_synthetic_dataset (line 1709) | def get_synthetic_dataset(args, preprocess_fn, is_train, epoch=0, tokeni...
function get_radiology_Dataset (line 1735) | def get_radiology_Dataset(
function get_pathology_Dataset (line 1775) | def get_pathology_Dataset(
function get_dataset_fn (line 1816) | def get_dataset_fn(data_path, dataset_type):
function get_data (line 1851) | def get_data(
FILE: train/open_clip/src/training/distributed.py
function is_global_master (line 12) | def is_global_master(args):
function is_local_master (line 16) | def is_local_master(args):
function is_master (line 20) | def is_master(args, local=False):
function is_using_horovod (line 24) | def is_using_horovod():
function is_using_distributed (line 35) | def is_using_distributed():
function world_info_from_env (line 43) | def world_info_from_env():
function init_distributed_device (line 63) | def init_distributed_device(args):
function broadcast_object (line 117) | def broadcast_object(args, obj, src=0):
function all_gather_object (line 130) | def all_gather_object(args, obj, dst=0):
FILE: train/open_clip/src/training/file_utils.py
function remote_sync_s3 (line 10) | def remote_sync_s3(local_dir, remote_dir):
function remote_sync_fsspec (line 20) | def remote_sync_fsspec(local_dir, remote_dir):
function remote_sync (line 44) | def remote_sync(local_dir, remote_dir, protocol):
function keep_running_remote_sync (line 54) | def keep_running_remote_sync(sync_every, local_dir, remote_dir, protocol):
function start_sync_process (line 59) | def start_sync_process(sync_every, local_dir, remote_dir, protocol):
function pt_save (line 64) | def pt_save(pt_obj, file_path):
function pt_load (line 69) | def pt_load(file_path, map_location=None):
function check_exists (line 77) | def check_exists(file_path):
FILE: train/open_clip/src/training/logger.py
function setup_logging (line 4) | def setup_logging(log_file, level, include_host=False):
FILE: train/open_clip/src/training/main.py
function random_seed (line 52) | def random_seed(seed=42, rank=0):
function natural_key (line 58) | def natural_key(string_):
function get_latest_checkpoint (line 63) | def get_latest_checkpoint(path: str, remote : bool):
function main (line 79) | def main(args):
function copy_codebase (line 519) | def copy_codebase(args):
FILE: train/open_clip/src/training/main_feature-work.py
function random_seed (line 54) | def random_seed(seed=42, rank=0):
function natural_key (line 60) | def natural_key(string_):
function get_latest_checkpoint (line 65) | def get_latest_checkpoint(path: str, remote : bool):
function get_checkpoint_from_id (line 80) | def get_checkpoint_from_id(path: str, remote : bool,checkpoint_id:int):
function main (line 101) | def main(args):
FILE: train/open_clip/src/training/main_retrieve_report.py
function random_seed (line 54) | def random_seed(seed=42, rank=0):
function natural_key (line 60) | def natural_key(string_):
function get_latest_checkpoint (line 65) | def get_latest_checkpoint(path: str, remote : bool):
function get_checkpoint_from_id (line 80) | def get_checkpoint_from_id(path: str, remote : bool,checkpoint_id:int):
function main (line 101) | def main(args):
FILE: train/open_clip/src/training/main_retrieve_report_harvard.py
function random_seed (line 54) | def random_seed(seed=42, rank=0):
function natural_key (line 60) | def natural_key(string_):
function get_latest_checkpoint (line 65) | def get_latest_checkpoint(path: str, remote : bool):
function get_checkpoint_from_id (line 80) | def get_checkpoint_from_id(path: str, remote : bool,checkpoint_id:int):
function main (line 101) | def main(args):
FILE: train/open_clip/src/training/params.py
function get_default_params (line 5) | def get_default_params(model_name):
class ParseKwargs (line 14) | class ParseKwargs(argparse.Action):
method __call__ (line 15) | def __call__(self, parser, namespace, values, option_string=None):
function parse_args (line 26) | def parse_args(args):
FILE: train/open_clip/src/training/precision.py
function get_autocast (line 5) | def get_autocast(precision):
FILE: train/open_clip/src/training/profiler.py
function profile_fvcore (line 23) | def profile_fvcore(
function profile_fvcore_text (line 44) | def profile_fvcore_text(
function profile_fvcore_image (line 63) | def profile_fvcore_image(
function profile_torch_image (line 82) | def profile_torch_image(model, image_input_size, batch_size=1, force_cpu...
function profile_torch_text (line 96) | def profile_torch_text(model, text_input_size, batch_size=1, force_cpu=F...
function profile_torch (line 110) | def profile_torch(model, text_input_size, image_input_size, batch_size=1...
function count_params (line 125) | def count_params(model):
function profile_model (line 128) | def profile_model(model_name, batch_size=1, profiler='torch'):
function main (line 205) | def main():
FILE: train/open_clip/src/training/scheduler.py
function assign_learning_rate (line 4) | def assign_learning_rate(optimizer, new_lr):
function _warmup_lr (line 9) | def _warmup_lr(base_lr, warmup_length, step):
function const_lr (line 13) | def const_lr(optimizer, base_lr, warmup_length, steps):
function const_lr_cooldown (line 24) | def const_lr_cooldown(optimizer, base_lr, warmup_length, steps, cooldown...
function cosine_lr (line 43) | def cosine_lr(optimizer, base_lr, warmup_length, steps):
FILE: train/open_clip/src/training/train.py
class AverageMeter (line 23) | class AverageMeter(object):
method __init__ (line 26) | def __init__(self):
method reset (line 29) | def reset(self):
method update (line 35) | def update(self, val, n=1):
function postprocess_clip_output (line 42) | def postprocess_clip_output(model_out):
function unwrap_model (line 50) | def unwrap_model(model):
function backward (line 57) | def backward(total_loss, scaler):
function train_one_epoch (line 64) | def train_one_epoch(model, data, loss, epoch, optimizer, scaler, schedul...
function evaluate (line 252) | def evaluate(model, data, epoch, args, tb_writer=None, tokenizer=None):
function retrieve_report (line 362) | def retrieve_report(model, data, epoch, args, tb_writer=None, tokenizer=...
function get_clip_metrics (line 579) | def get_clip_metrics(image_features, text_features, logit_scale):
function get_logits (line 598) | def get_logits(image_features, text_features, logit_scale):
function maybe_compute_generative_loss (line 611) | def maybe_compute_generative_loss(model_out):
FILE: train/open_clip/src/training/zero_shot.py
function accuracy (line 11) | def accuracy(output, target, topk=(1,)):
function run (line 17) | def run(model, classifier, dataloader, args):
function run_retrieve (line 43) | def run_retrieve(model, classifier, dataloader, args):
function zero_shot_retrieve (line 69) | def zero_shot_retrieve(model, data, epoch, args, tokenizer=None):
function zero_shot_eval (line 113) | def zero_shot_eval(model, data, epoch, args, tokenizer=None):
Copy disabled (too large)
Download .json
Condensed preview — 283 files, each showing path, character count, and a content snippet. Download the .json file for the full structured content (44,794K chars).
[
{
"path": ".gitignore",
"chars": 202,
"preview": "backup\n\ntrain/open_clip/.github\ntrain/open_clip/.gitignore\ntrain/open_clip/docs\ntrain/open_clip/scripts\n\ntrain/dpo/playg"
},
{
"path": "LICENSE",
"chars": 1075,
"preview": "MIT License\n\nCopyright (c) 2024 Peng \"Richard\" Xia\n\nPermission is hereby granted, free of charge, to any person obtainin"
},
{
"path": "README.md",
"chars": 6009,
"preview": "# MMed-RAG: Versatile Multimodal RAG System for Medical Vision Language Models\n\nWe introduce MMed-RAG, a powerful multim"
},
{
"path": "data/test/report/harvard_test.json",
"chars": 1294580,
"preview": "[\n {\n \"id\": \"data_08003\",\n \"image_path\": \"slo_fundus_08003.jpg\",\n \"filename\": \"data_08003.npz\",\n"
},
{
"path": "data/test/report/iuxray_test.json",
"chars": 234322,
"preview": "[\n {\n \"id\": \"CXR3030_IM-1405\",\n \"report\": \"Normal cardiomediastinal silhouette. There is no focal conso"
},
{
"path": "data/test/report/mimic_test.json",
"chars": 590440,
"preview": "[\n {\n \"id\": \"0c5f56c2-3d707105-b36af285-88d0ae60-48ef3fda\",\n \"study_id\": 53078789,\n \"subject_id\""
},
{
"path": "data/test/report/pmc-oa_test.json",
"chars": 372854,
"preview": "[\n {\n \"id\": 0,\n \"image\": \"PMC2874782_F1_64665.jpg\",\n \"report\": \"Representative immunostaining of"
},
{
"path": "data/test/report/quilt-1m_test.json",
"chars": 3536062,
"preview": "[\n {\n \"\": \"1006600\",\n \"caption\": \"The cytoplasmic features of smaller and larger benign glands are simi"
},
{
"path": "data/test/vqa/harvard_test.jsonl",
"chars": 1750461,
"preview": "{\"question_id\": 1, \"question\": \"Has the patient been diagnosed with glaucoma?\\n<image>\", \"answer\": \"Yes.\", \"image\": \"slo"
},
{
"path": "data/test/vqa/iuxray_test.jsonl",
"chars": 1358950,
"preview": "{\"question\": \"Does the cardiomediastinal silhouette appear normal in the chest X-ray? \\n<image>\", \"answer\": \"Yes.\", \"ima"
},
{
"path": "data/test/vqa/mimic_test.jsonl",
"chars": 2656306,
"preview": "{\"question_id\": 1, \"question\": \"Does the patient show any signs of an acute cardiopulmonary process?\\n<image>\", \"answer\""
},
{
"path": "data/test/vqa/pmc-oa_test.jsonl",
"chars": 1032264,
"preview": "{\"question_id\": 1, \"question\": \"Is there immunoreaction for CA IX in Panel A?\\n<image>\", \"answer\": \"No\", \"image\": \"PMC28"
},
{
"path": "data/test/vqa/quilt-1m_test.jsonl",
"chars": 1023360,
"preview": "{\"question_id\": 1, \"question\": \"Can a whirling pattern be seen in perineurioma?\\n<image>\", \"answer\": \"Yes\", \"image\": \"QD"
},
{
"path": "data/training/alignment/ophthalmology/harvard_report.json",
"chars": 805727,
"preview": "[{\"id\": \"data_07655\", \"image\": \"slo_fundus_07655.jpg\", \"conversations\": [{\"from\": \"human\", \"value\": \"<image>\\nYou are a "
},
{
"path": "data/training/alignment/ophthalmology/harvard_vqa.json",
"chars": 3270636,
"preview": "[{\"id\": 4052, \"image\": \"slo_fundus_07639.jpg\", \"conversations\": [{\"from\": \"human\", \"value\": \"<image>\\nYou are provided w"
},
{
"path": "data/training/alignment/pathology/pathology_vqa.json",
"chars": 4243176,
"preview": "[\n {\n \"id\": 641,\n \"image\": \"YYyq1Ewnc3M_image_1425f900-eb6d-4b5f-8921-b37a27048233.jpg\",\n \"conve"
},
{
"path": "data/training/alignment/radiology/radiology_report.json",
"chars": 1929945,
"preview": "[\n {\n \"id\": \"CXR841_IM-2365\",\n \"image\": \"CXR841_IM-2365/0.png\",\n \"conversations\": [\n "
},
{
"path": "data/training/retriever/ophthalmology/harvard_val_1000.json",
"chars": 1822444,
"preview": "[\n {\n \"id\": \"data_07001\",\n \"image_path\": \"slo_fundus_07001.jpg\",\n \"filename\": \"data_07001.npz\",\n"
},
{
"path": "data/training/retriever/pathology/pathology_train.json",
"chars": 2162932,
"preview": "[\n {\n \"id\": 0,\n \"image\": \"14633.jpg\",\n \"report\": \" Representative brightfield image showing the "
},
{
"path": "data/training/retriever/pathology/pathology_val.json",
"chars": 824945,
"preview": "[\n {\n \"id\": 2000,\n \"image\": \"14566.jpg\",\n \"report\": \" Scanning electron microscopy observations "
},
{
"path": "data/training/retriever/radiology/radiology_train.json",
"chars": 3083052,
"preview": "[\n {\n \"id\": \"CXR940_IM-2437\",\n \"report\": \"The lungs and pleural spaces show no acute abnormality. Heart"
},
{
"path": "data/training/retriever/radiology/radiology_val.json",
"chars": 1378005,
"preview": "[\n {\n \"id\": \"CXR2079_IM-0711\",\n \"report\": \"The lungs are clear, and without focal airspace opacity. The"
},
{
"path": "requirements.txt",
"chars": 3657,
"preview": "absl-py==2.1.0\naccelerate==0.21.0\naiofiles==23.2.1\naiohttp==3.9.3\naiosignal==1.3.1\naltair==5.2.0\nannotated-types==0.6.0\n"
},
{
"path": "scripts/finetune_clip.sh",
"chars": 552,
"preview": "export CUDA_VISIBLE_DEVICES=\"1,2\"\n\n\ncd ./train/open_clip/src\n\n# harvard dataset\ntorchrun --nproc_per_node=2 \\\n --mast"
},
{
"path": "scripts/retrieve_clip_VQA.sh",
"chars": 530,
"preview": "cd ./train/open_clip/src || exit\nCUDA_VISIBLE_DEVICES=1 python ./retrieve_clip_VQA.py \\\n --img_root /path/to/image_fo"
},
{
"path": "scripts/retrieve_clip_report.sh",
"chars": 529,
"preview": "cd ./train/open_clip/src || exit\n\nCUDA_VISIBLE_DEVICES=1 python ./retrieve_clip_report.py \\\n --img_root /path/to/imag"
},
{
"path": "scripts/train_dpo_2stages.sh",
"chars": 1347,
"preview": "CUDA_DEVICES='0,1,2,3'\nCUDA='1,3'\n\ncd ./train/dpo || exit\nCUDA_VISIBLE_DEVICES=$CUDA_DEVICES deepspeed --include localho"
},
{
"path": "train/dpo/LICENSE",
"chars": 11357,
"preview": " Apache License\n Version 2.0, January 2004\n "
},
{
"path": "train/dpo/cog.yaml",
"chars": 977,
"preview": "# Configuration for Cog ⚙️\n# Reference: https://github.com/replicate/cog/blob/main/docs/yaml.md\n\nbuild:\n gpu: true\n\n p"
},
{
"path": "train/dpo/dpo_trainer_2stages.py",
"chars": 40246,
"preview": "# DPO Authors: Rafael Rafailov, Archit Sharma, Eric Mitchell, Stefano Ermon, Christopher D. Manning, and Chelsea Finn 20"
},
{
"path": "train/dpo/llava/__init__.py",
"chars": 41,
"preview": "from .model import LlavaLlamaForCausalLM\n"
},
{
"path": "train/dpo/llava/constants.py",
"chars": 335,
"preview": "CONTROLLER_HEART_BEAT_EXPIRATION = 30\nWORKER_HEART_BEAT_INTERVAL = 15\n\nLOGDIR = \".\"\n\n# Model Constants\nIGNORE_INDEX = -1"
},
{
"path": "train/dpo/llava/conversation.py",
"chars": 15022,
"preview": "import dataclasses\nfrom enum import auto, Enum\nfrom typing import List, Tuple\nimport base64\nfrom io import BytesIO\nfrom "
},
{
"path": "train/dpo/llava/conversation_new.py",
"chars": 14072,
"preview": "import dataclasses\nfrom enum import auto, Enum\nfrom typing import List, Tuple\n\n\nclass SeparatorStyle(Enum):\n \"\"\"Diffe"
},
{
"path": "train/dpo/llava/eval/eval_gpt_review.py",
"chars": 3620,
"preview": "import argparse\nimport json\nimport os\n\nimport openai\nimport tqdm\nimport ray\nimport time\n\nNUM_SECONDS_TO_SLEEP = 3\n\n@ray."
},
{
"path": "train/dpo/llava/eval/eval_gpt_review_bench.py",
"chars": 4172,
"preview": "import argparse\nimport json\nimport os\n\nimport openai\nimport time\n\nNUM_SECONDS_TO_SLEEP = 0.5\n\n\ndef get_eval(content: str"
},
{
"path": "train/dpo/llava/eval/eval_gpt_review_visual.py",
"chars": 4177,
"preview": "import argparse\nimport json\nimport os\n\nimport openai\nimport time\n\nNUM_SECONDS_TO_SLEEP = 0.5\n\n\ndef get_eval(content: str"
},
{
"path": "train/dpo/llava/eval/eval_pope.py",
"chars": 2732,
"preview": "import os\nimport json\nimport argparse\n\ndef eval_pope(answers, label_file):\n label_list = [json.loads(q)['label'] for "
},
{
"path": "train/dpo/llava/eval/eval_science_qa.py",
"chars": 3920,
"preview": "import argparse\nimport json\nimport os\nimport re\nimport random\n\n\ndef get_args():\n parser = argparse.ArgumentParser()\n "
},
{
"path": "train/dpo/llava/eval/eval_science_qa_gpt4.py",
"chars": 3675,
"preview": "import argparse\nimport json\nimport os\nimport re\nimport random\nfrom collections import defaultdict\n\n\ndef get_args():\n "
},
{
"path": "train/dpo/llava/eval/eval_science_qa_gpt4_requery.py",
"chars": 5774,
"preview": "import argparse\nimport json\nimport os\nimport re\nimport random\nfrom collections import defaultdict\n\n\ndef get_args():\n "
},
{
"path": "train/dpo/llava/eval/eval_textvqa.py",
"chars": 2226,
"preview": "import os\nimport argparse\nimport json\nimport re\n\nfrom llava.eval.m4c_evaluator import TextVQAAccuracyEvaluator\n\n\ndef get"
},
{
"path": "train/dpo/llava/eval/generate_webpage_data_from_table.py",
"chars": 4088,
"preview": "\"\"\"Generate json file for webpage.\"\"\"\nimport json\nimport os\nimport re\n\n# models = ['llama', 'alpaca', 'gpt35', 'bard']\nm"
},
{
"path": "train/dpo/llava/eval/m4c_evaluator.py",
"chars": 10265,
"preview": "# Copyright (c) Facebook, Inc. and its affiliates.\nimport re\n\nfrom tqdm import tqdm\n\n\nclass EvalAIAnswerProcessor:\n \""
},
{
"path": "train/dpo/llava/eval/model_qa.py",
"chars": 2430,
"preview": "import argparse\nfrom transformers import AutoTokenizer, AutoModelForCausalLM, StoppingCriteria\nimport torch\nimport os\nim"
},
{
"path": "train/dpo/llava/eval/model_vqa.py",
"chars": 4115,
"preview": "import argparse\nimport torch\nimport os\nimport json\nfrom tqdm import tqdm\nimport shortuuid\n\nfrom llava.constants import I"
},
{
"path": "train/dpo/llava/eval/model_vqa_loader.py",
"chars": 5975,
"preview": "import argparse\nimport torch\nimport os\nimport json\nfrom tqdm import tqdm\nimport shortuuid\n\nfrom llava.constants import I"
},
{
"path": "train/dpo/llava/eval/model_vqa_mmbench.py",
"chars": 6408,
"preview": "import argparse\nimport torch\nimport os\nimport json\nimport pandas as pd\nfrom tqdm import tqdm\nimport shortuuid\n\nfrom llav"
},
{
"path": "train/dpo/llava/eval/model_vqa_science.py",
"chars": 4592,
"preview": "import argparse\nimport torch\nimport os\nimport json\nfrom tqdm import tqdm\nimport shortuuid\n\nfrom llava.constants import I"
},
{
"path": "train/dpo/llava/eval/qa_baseline_gpt35.py",
"chars": 2345,
"preview": "\"\"\"Generate answers with GPT-3.5\"\"\"\n# Note: you need to be using OpenAI Python v0.27.0 for the code below to work\nimport"
},
{
"path": "train/dpo/llava/eval/run_llava.py",
"chars": 4443,
"preview": "import argparse\nimport torch\n\nfrom llava.constants import (\n IMAGE_TOKEN_INDEX,\n DEFAULT_IMAGE_TOKEN,\n DEFAULT_"
},
{
"path": "train/dpo/llava/eval/summarize_gpt_review.py",
"chars": 2438,
"preview": "import json\nimport os\nfrom collections import defaultdict\n\nimport numpy as np\n\nimport argparse\n\ndef parse_args():\n pa"
},
{
"path": "train/dpo/llava/eval/table/answer/answer_alpaca-13b.jsonl",
"chars": 57071,
"preview": "{\"question_id\": 1, \"text\": \"Improving time management skills involves setting priorities, breaking tasks into smaller ch"
},
{
"path": "train/dpo/llava/eval/table/answer/answer_bard.jsonl",
"chars": 112274,
"preview": "{\"answer_id\": \"3oW4JY265ZPJGTYi2CgRYF\", \"model_id\": \"bard:20230327\", \"question_id\": 1, \"text\": \"Here are some tips on ho"
},
{
"path": "train/dpo/llava/eval/table/answer/answer_gpt35.jsonl",
"chars": 107603,
"preview": "{\"answer_id\": \"BZGowHM7L3RvtWRktKZjLT\", \"model_id\": \"gpt-3.5-turbo:20230327\", \"question_id\": 1, \"text\": \"Here are some t"
},
{
"path": "train/dpo/llava/eval/table/answer/answer_llama-13b.jsonl",
"chars": 76353,
"preview": "{\"answer_id\": \"J3UA6eGXGyFeUGqGpP3g34\", \"model_id\": \"llama-13b:v1\", \"question_id\": 1, \"text\": \"The following are some st"
},
{
"path": "train/dpo/llava/eval/table/answer/answer_vicuna-13b.jsonl",
"chars": 131904,
"preview": "{\"answer_id\": \"cV4zXygaNP6CXEsgdHMEqz\", \"model_id\": \"vicuna-13b:20230322-clean-lang\", \"question_id\": 1, \"text\": \"Improvi"
},
{
"path": "train/dpo/llava/eval/table/caps_boxes_coco2014_val_80.jsonl",
"chars": 58574,
"preview": "{\"id\": \"000000296284\", \"image\": \"000000296284.jpg\", \"captions\": [\"A donut shop is full of different flavors of donuts.\","
},
{
"path": "train/dpo/llava/eval/table/model.jsonl",
"chars": 681,
"preview": "{\"model_id\": \"vicuna-13b:20230322-clean-lang\", \"model_name\": \"vicuna-13b\", \"model_version\": \"20230322-clean-lang\", \"mode"
},
{
"path": "train/dpo/llava/eval/table/prompt.jsonl",
"chars": 5129,
"preview": "{\"prompt_id\": 1, \"system_prompt\": \"You are a helpful and precise assistant for checking the quality of the answer.\", \"pr"
},
{
"path": "train/dpo/llava/eval/table/question.jsonl",
"chars": 12885,
"preview": "{\"question_id\": 1, \"text\": \"How can I improve my time management skills?\", \"category\": \"generic\"}\n{\"question_id\": 2, \"te"
},
{
"path": "train/dpo/llava/eval/table/results/test_sqa_llava_13b_v0.json",
"chars": 3950324,
"preview": "{\n \"acc\": 90.8983730252299,\n \"correct\": 3855,\n \"count\": 4241,\n \"results\": {\n \"4\": 1,\n \"5\": 1,\n \"11\": 1,\n "
},
{
"path": "train/dpo/llava/eval/table/results/test_sqa_llava_lcs_558k_sqa_12e_vicuna_v1_3_13b.json",
"chars": 3830902,
"preview": "{\n \"acc\": 91.08700778118369,\n \"correct\": 3863,\n \"count\": 4241,\n \"results\": {\n \"4\": 1,\n \"5\": 1,\n \"11\": 1,\n "
},
{
"path": "train/dpo/llava/eval/table/review/review_alpaca-13b_vicuna-13b.jsonl",
"chars": 73131,
"preview": "{\"review_id\": \"QM5m5nnioWr8M2LFHsaQvu\", \"question_id\": 1, \"answer1_id\": \"kEL9ifUHDeYuAXzevje2se\", \"answer2_id\": \"cV4zXyg"
},
{
"path": "train/dpo/llava/eval/table/review/review_bard_vicuna-13b.jsonl",
"chars": 73145,
"preview": "{\"review_id\": \"4CeMvEQyE6fKMJwvSLY3P4\", \"question_id\": 1, \"answer1_id\": \"3oW4JY265ZPJGTYi2CgRYF\", \"answer2_id\": \"cV4zXyg"
},
{
"path": "train/dpo/llava/eval/table/review/review_gpt35_vicuna-13b.jsonl",
"chars": 73399,
"preview": "{\"review_id\": \"jyhS7AFj2mrFNqoRXQJDPS\", \"question_id\": 1, \"answer1_id\": \"BZGowHM7L3RvtWRktKZjLT\", \"answer2_id\": \"cV4zXyg"
},
{
"path": "train/dpo/llava/eval/table/review/review_llama-13b_vicuna-13b.jsonl",
"chars": 67249,
"preview": "{\"review_id\": \"WFp5i5yjjFethrgugKTDmX\", \"question_id\": 1, \"answer1_id\": \"J3UA6eGXGyFeUGqGpP3g34\", \"answer2_id\": \"cV4zXyg"
},
{
"path": "train/dpo/llava/eval/table/reviewer.jsonl",
"chars": 604,
"preview": "{\"reviewer_id\": \"gpt-4-0328-default\", \"prompt_id\": 1, \"metadata\": {\"temperature\": 0.2, \"max_tokens\": 1024}, \"description"
},
{
"path": "train/dpo/llava/eval/table/rule.json",
"chars": 9098,
"preview": "{\n \"coding\": {\"role\": \"Assistant\", \"prompt\": \"Your task is to evaluate the coding abilities of the above two assistan"
},
{
"path": "train/dpo/llava/eval/webpage/styles.css",
"chars": 1822,
"preview": "body {\n font-family: 'Segoe UI', Tahoma, Geneva, Verdana, sans-serif;\n background-color: #f8f9fa;\n}\n\n.navbar-dark "
},
{
"path": "train/dpo/llava/mm_utils.py",
"chars": 9555,
"preview": "from PIL import Image\nfrom io import BytesIO\nimport base64\nimport torch\nimport math\nimport ast\n\nfrom transformers import"
},
{
"path": "train/dpo/llava/model/__init__.py",
"chars": 269,
"preview": "try:\n from .language_model.llava_llama import LlavaLlamaForCausalLM, LlavaConfig\n from .language_model.llava_mpt i"
},
{
"path": "train/dpo/llava/model/apply_delta.py",
"chars": 1956,
"preview": "\"\"\"\nUsage:\npython3 -m fastchat.model.apply_delta --base ~/model_weights/llama-7b --target ~/model_weights/vicuna-7b --de"
},
{
"path": "train/dpo/llava/model/builder.py",
"chars": 8761,
"preview": "# Copyright 2023 Haotian Liu\n#\n# Licensed under the Apache License, Version 2.0 (the \"License\");\n# you may not "
},
{
"path": "train/dpo/llava/model/consolidate.py",
"chars": 914,
"preview": "\"\"\"\nUsage:\npython3 -m llava.model.consolidate --src ~/model_weights/llava-7b --dst ~/model_weights/llava-7b_consolidate\n"
},
{
"path": "train/dpo/llava/model/language_model/llava_llama.py",
"chars": 5378,
"preview": "# Copyright 2023 Haotian Liu\n#\n# Licensed under the Apache License, Version 2.0 (the \"License\");\n# you may not "
},
{
"path": "train/dpo/llava/model/language_model/llava_mistral.py",
"chars": 5386,
"preview": "# Copyright 2023 Haotian Liu\n#\n# Licensed under the Apache License, Version 2.0 (the \"License\");\n# you may not "
},
{
"path": "train/dpo/llava/model/language_model/llava_mpt.py",
"chars": 3487,
"preview": "# Copyright 2023 Haotian Liu\n#\n# Licensed under the Apache License, Version 2.0 (the \"License\");\n# you may not "
},
{
"path": "train/dpo/llava/model/llava_arch.py",
"chars": 18110,
"preview": "# Copyright 2023 Haotian Liu\n#\n# Licensed under the Apache License, Version 2.0 (the \"License\");\n# you may not "
},
{
"path": "train/dpo/llava/model/make_delta.py",
"chars": 2257,
"preview": "\"\"\"\nUsage:\npython3 -m llava.model.make_delta --base ~/model_weights/llama-7b --target ~/model_weights/llava-7b --delta ~"
},
{
"path": "train/dpo/llava/model/multimodal_encoder/builder.py",
"chars": 556,
"preview": "import os\nfrom .clip_encoder import CLIPVisionTower\n\n\ndef build_vision_tower(vision_tower_cfg, **kwargs):\n vision_tow"
},
{
"path": "train/dpo/llava/model/multimodal_encoder/clip_encoder.py",
"chars": 3062,
"preview": "import torch\nimport torch.nn as nn\n\nfrom transformers import CLIPVisionModel, CLIPImageProcessor, CLIPVisionConfig\n\n\ncla"
},
{
"path": "train/dpo/llava/model/multimodal_projector/builder.py",
"chars": 1437,
"preview": "import torch\nimport torch.nn as nn\nimport re\n\n\nclass IdentityMap(nn.Module):\n def __init__(self):\n super().__i"
},
{
"path": "train/dpo/llava/model/utils.py",
"chars": 927,
"preview": "from transformers import AutoConfig\n\n\ndef auto_upgrade(config):\n cfg = AutoConfig.from_pretrained(config)\n if 'lla"
},
{
"path": "train/dpo/llava/serve/__init__.py",
"chars": 0,
"preview": ""
},
{
"path": "train/dpo/llava/serve/cli.py",
"chars": 4808,
"preview": "import argparse\nimport torch\n\nfrom llava.constants import IMAGE_TOKEN_INDEX, DEFAULT_IMAGE_TOKEN, DEFAULT_IM_START_TOKEN"
},
{
"path": "train/dpo/llava/serve/controller.py",
"chars": 9949,
"preview": "\"\"\"\nA controller manages distributed workers.\nIt sends worker addresses to clients.\n\"\"\"\nimport argparse\nimport asyncio\ni"
},
{
"path": "train/dpo/llava/serve/gradio_web_server.py",
"chars": 18841,
"preview": "import argparse\nimport datetime\nimport json\nimport os\nimport time\n\nimport gradio as gr\nimport requests\n\nfrom llava.conve"
},
{
"path": "train/dpo/llava/serve/model_worker.py",
"chars": 11176,
"preview": "\"\"\"\nA model worker executes the model.\n\"\"\"\nimport argparse\nimport asyncio\nimport json\nimport time\nimport threading\nimpor"
},
{
"path": "train/dpo/llava/serve/register_worker.py",
"chars": 734,
"preview": "\"\"\"\nManually register workers.\n\nUsage:\npython3 -m fastchat.serve.register_worker --controller http://localhost:21001 --w"
},
{
"path": "train/dpo/llava/serve/sglang_worker.py",
"chars": 8678,
"preview": "\"\"\"\nA model worker executes the model.\n\"\"\"\nimport argparse\nimport asyncio\nfrom concurrent.futures import ThreadPoolExecu"
},
{
"path": "train/dpo/llava/serve/test_message.py",
"chars": 2022,
"preview": "import argparse\nimport json\n\nimport requests\n\nfrom llava.conversation import default_conversation\n\n\ndef main():\n if a"
},
{
"path": "train/dpo/llava/train/llama_flash_attn_monkey_patch.py",
"chars": 4404,
"preview": "from typing import Optional, Tuple\nimport warnings\n\nimport torch\n\nimport transformers\nfrom transformers.models.llama.mod"
},
{
"path": "train/dpo/llava/train/llama_xformers_attn_monkey_patch.py",
"chars": 4916,
"preview": "\"\"\"\nDirectly copied the code from https://raw.githubusercontent.com/oobabooga/text-generation-webui/main/modules/llama_a"
},
{
"path": "train/dpo/llava/train/llava_trainer.py",
"chars": 11571,
"preview": "import os\nimport torch\n\nfrom torch.utils.data import Sampler\n\nfrom transformers import Trainer\nfrom transformers.trainer"
},
{
"path": "train/dpo/llava/train/train.py",
"chars": 38414,
"preview": "# Adopted from https://github.com/lm-sys/FastChat. Below is the original copyright:\n# Adopted from tatsu-lab@stanford_al"
},
{
"path": "train/dpo/llava/train/train_dpo.py",
"chars": 41563,
"preview": "# Adopted from https://github.com/lm-sys/FastChat. Below is the original copyright:\n# Adopted from tatsu-lab@stanford_al"
},
{
"path": "train/dpo/llava/train/train_dpo_inherent.py",
"chars": 42184,
"preview": "# Adopted from https://github.com/lm-sys/FastChat. Below is the original copyright:\n# Adopted from tatsu-lab@stanford_al"
},
{
"path": "train/dpo/llava/train/train_mem.py",
"chars": 115,
"preview": "from llava.train.train import train\n\nif __name__ == \"__main__\":\n train(attn_implementation=\"flash_attention_2\")\n"
},
{
"path": "train/dpo/llava/train/train_xformers.py",
"chars": 366,
"preview": "# Make it more memory efficient by monkey patching the LLaMA model with xformers attention.\n\n# Need to call this before "
},
{
"path": "train/dpo/llava/utils.py",
"chars": 4003,
"preview": "import datetime\nimport logging\nimport logging.handlers\nimport os\nimport sys\n\nimport requests\n\nfrom llava.constants impor"
},
{
"path": "train/dpo/llava_trainer_2stages.py",
"chars": 11574,
"preview": "import os\nimport torch\n\nfrom torch.utils.data import Sampler\n\nfrom transformers import Trainer\nfrom transformers.trainer"
},
{
"path": "train/dpo/povid_infer.py",
"chars": 4978,
"preview": "import argparse\nimport torch\nimport os\nimport json\nfrom tqdm import tqdm\nimport shortuuid\n\nfrom llava.constants import I"
},
{
"path": "train/dpo/predict.py",
"chars": 6080,
"preview": "import torch\n\nfrom llava.constants import IMAGE_TOKEN_INDEX, DEFAULT_IMAGE_TOKEN\nfrom llava.conversation import conv_tem"
},
{
"path": "train/dpo/pyproject.toml",
"chars": 1256,
"preview": "[build-system]\nrequires = [\"setuptools>=61.0\"]\nbuild-backend = \"setuptools.build_meta\"\n\n[project]\nname = \"llava\"\nversion"
},
{
"path": "train/dpo/scripts/convert_gqa_for_eval.py",
"chars": 487,
"preview": "import os\nimport json\nimport argparse\n\nparser = argparse.ArgumentParser()\nparser.add_argument(\"--src\", type=str)\nparser."
},
{
"path": "train/dpo/scripts/convert_mmbench_for_submission.py",
"chars": 980,
"preview": "import os\nimport json\nimport argparse\nimport pandas as pd\n\ndef get_args():\n parser = argparse.ArgumentParser()\n pa"
},
{
"path": "train/dpo/scripts/convert_mmvet_for_eval.py",
"chars": 397,
"preview": "import os\nimport json\nimport argparse\n\nparser = argparse.ArgumentParser()\nparser.add_argument(\"--src\", type=str)\nparser."
},
{
"path": "train/dpo/scripts/convert_seed_for_submission.py",
"chars": 2571,
"preview": "import os\nimport json\nimport argparse\n\n\ndef get_args():\n parser = argparse.ArgumentParser()\n parser.add_argument(\""
},
{
"path": "train/dpo/scripts/convert_sqa_to_llava.py",
"chars": 2971,
"preview": "import json\nimport os\nimport fire\nimport re\nfrom convert_sqa_to_llava_base_prompt import build_prompt_chatbot\n\n\ndef conv"
},
{
"path": "train/dpo/scripts/convert_sqa_to_llava_base_prompt.py",
"chars": 13795,
"preview": "def get_question_text(problem):\n question = problem['question']\n return question\n\n\ndef get_context_text(problem, u"
},
{
"path": "train/dpo/scripts/convert_vizwiz_for_submission.py",
"chars": 1387,
"preview": "import os\nimport argparse\nimport json\n\nfrom llava.eval.m4c_evaluator import EvalAIAnswerProcessor\n\n\ndef parse_args():\n "
},
{
"path": "train/dpo/scripts/convert_vqav2_for_submission.py",
"chars": 1743,
"preview": "import os\nimport argparse\nimport json\n\nfrom llava.eval.m4c_evaluator import EvalAIAnswerProcessor\n\n\ndef parse_args():\n "
},
{
"path": "train/dpo/scripts/extract_mm_projector.py",
"chars": 1606,
"preview": "\"\"\"\nThis is just a utility that I use to extract the projector for quantized models.\nIt is NOT necessary at all to train"
},
{
"path": "train/dpo/scripts/finetune.sh",
"chars": 1642,
"preview": "#!/bin/bash\n\n# IMPORTANT: this is the training script for the original LLaVA, NOT FOR LLaVA V1.5!\n\n# Uncomment and set t"
},
{
"path": "train/dpo/scripts/finetune_full_schedule.sh",
"chars": 1643,
"preview": "#!/bin/bash\n\n# IMPORTANT: this is the training script for the original LLaVA, NOT FOR LLaVA V1.5!\n\n# Uncomment and set t"
},
{
"path": "train/dpo/scripts/finetune_lora.sh",
"chars": 1672,
"preview": "#!/bin/bash\n\n# IMPORTANT: this is the training script for the original LLaVA, NOT FOR LLaVA V1.5!\n\n# Uncomment and set t"
},
{
"path": "train/dpo/scripts/finetune_qlora.sh",
"chars": 1687,
"preview": "#!/bin/bash\n\n# IMPORTANT: this is the training script for the original LLaVA, NOT FOR LLaVA V1.5!\n\n# Uncomment and set t"
},
{
"path": "train/dpo/scripts/finetune_sqa.sh",
"chars": 1346,
"preview": "#!/bin/bash\n\n# IMPORTANT: this is the training script for the original LLaVA, NOT FOR LLaVA V1.5!\n\ndeepspeed llava/train"
},
{
"path": "train/dpo/scripts/merge_lora_weights.py",
"chars": 767,
"preview": "import argparse\nfrom llava.model.builder import load_pretrained_model\nfrom llava.mm_utils import get_model_name_from_pat"
},
{
"path": "train/dpo/scripts/pretrain.sh",
"chars": 1460,
"preview": "#!/bin/bash\n\n# IMPORTANT: this is the training script for the original LLaVA, NOT FOR LLaVA V1.5!\n\n# Uncomment and set t"
},
{
"path": "train/dpo/scripts/pretrain_xformers.sh",
"chars": 1380,
"preview": "#!/bin/bash\n\n# Uncomment and set the following variables correspondingly to run this script:\n\n# MODEL_VERSION=vicuna-v1-"
},
{
"path": "train/dpo/scripts/run_povid.sh",
"chars": 1308,
"preview": "# conda env \nsource activate [your env path]/envs/POVID\ncd ..\ndeepspeed llava/train/train_dpo_inherent.py \\\n --lora_e"
},
{
"path": "train/dpo/scripts/sqa_eval_batch.sh",
"chars": 524,
"preview": "#!/bin/bash\n\nCHUNKS=8\nfor IDX in {0..7}; do\n CUDA_VISIBLE_DEVICES=$IDX python -m llava.eval.model_vqa_science \\\n "
},
{
"path": "train/dpo/scripts/sqa_eval_gather.sh",
"chars": 518,
"preview": "#!/bin/bash\n\nCHUNKS=8\noutput_file=\"test_llava-13b.jsonl\"\n\n# Clear out the output file if it exists.\n> \"$output_file\"\n\n# "
},
{
"path": "train/dpo/scripts/upload_pypi.sh",
"chars": 387,
"preview": "#!/bin/bash\n\n# Step 0: Clean up\nrm -rf dist\n\n# Step 1: Change the package name to \"llava-torch\"\nsed -i 's/name = \"llava\""
},
{
"path": "train/dpo/scripts/v1_5/eval/gqa.sh",
"chars": 1228,
"preview": "#!/bin/bash\n\ngpu_list=\"${CUDA_VISIBLE_DEVICES:-0}\"\nIFS=',' read -ra GPULIST <<< \"$gpu_list\"\n\nCHUNKS=${#GPULIST[@]}\n\nCKPT"
},
{
"path": "train/dpo/scripts/v1_5/eval/llavabench.sh",
"chars": 1093,
"preview": "#!/bin/bash\n\npython -m llava.eval.model_vqa \\\n --model-path liuhaotian/llava-v1.5-13b \\\n --question-file ./playgro"
},
{
"path": "train/dpo/scripts/v1_5/eval/mmbench.sh",
"chars": 704,
"preview": "#!/bin/bash\n\nSPLIT=\"mmbench_dev_20230712\"\n\npython -m llava.eval.model_vqa_mmbench \\\n --model-path liuhaotian/llava-v1"
},
{
"path": "train/dpo/scripts/v1_5/eval/mmbench_cn.sh",
"chars": 738,
"preview": "#!/bin/bash\n\nSPLIT=\"mmbench_dev_cn_20231003\"\n\npython -m llava.eval.model_vqa_mmbench \\\n --model-path liuhaotian/llava"
},
{
"path": "train/dpo/scripts/v1_5/eval/mme.sh",
"chars": 532,
"preview": "#!/bin/bash\n\npython -m llava.eval.model_vqa_loader \\\n --model-path liuhaotian/llava-v1.5-13b \\\n --question-file ./"
},
{
"path": "train/dpo/scripts/v1_5/eval/mmvet.sh",
"chars": 580,
"preview": "#!/bin/bash\n\npython -m llava.eval.model_vqa \\\n --model-path liuhaotian/llava-v1.5-13b \\\n --question-file ./playgro"
},
{
"path": "train/dpo/scripts/v1_5/eval/pope.sh",
"chars": 590,
"preview": "#!/bin/bash\n\npython -m llava.eval.model_vqa_loader \\\n --model-path liuhaotian/llava-v1.5-13b \\\n --question-file ./"
},
{
"path": "train/dpo/scripts/v1_5/eval/qbench.sh",
"chars": 578,
"preview": "#!/bin/bash\n\nif [ \"$1\" = \"dev\" ]; then\n echo \"Evaluating in 'dev' split.\"\nelif [ \"$1\" = \"test\" ]; then\n echo \"Eval"
},
{
"path": "train/dpo/scripts/v1_5/eval/qbench_zh.sh",
"chars": 621,
"preview": "#!/bin/bash\n\nif [ \"$1\" = \"dev\" ]; then\n ZH_SPLIT=\"验证集\"\n echo \"Evaluating in 'dev' split.\"\nelif [ \"$1\" = \"test\" ]; "
},
{
"path": "train/dpo/scripts/v1_5/eval/seed.sh",
"chars": 1264,
"preview": "#!/bin/bash\n\ngpu_list=\"${CUDA_VISIBLE_DEVICES:-0}\"\nIFS=',' read -ra GPULIST <<< \"$gpu_list\"\n\nCHUNKS=${#GPULIST[@]}\n\nCKPT"
},
{
"path": "train/dpo/scripts/v1_5/eval/sqa.sh",
"chars": 749,
"preview": "#!/bin/bash\n\npython -m llava.eval.model_vqa_science \\\n --model-path liuhaotian/llava-v1.5-13b \\\n --question-file ."
},
{
"path": "train/dpo/scripts/v1_5/eval/textvqa.sh",
"chars": 571,
"preview": "#!/bin/bash\n\npython -m llava.eval.model_vqa_loader \\\n --model-path liuhaotian/llava-v1.5-13b \\\n --question-file ./"
},
{
"path": "train/dpo/scripts/v1_5/eval/vizwiz.sh",
"chars": 642,
"preview": "#!/bin/bash\n\npython -m llava.eval.model_vqa_loader \\\n --model-path liuhaotian/llava-v1.5-13b \\\n --question-file ./"
},
{
"path": "train/dpo/scripts/v1_5/eval/vqav2.sh",
"chars": 1113,
"preview": "#!/bin/bash\n\ngpu_list=\"${CUDA_VISIBLE_DEVICES:-0}\"\nIFS=',' read -ra GPULIST <<< \"$gpu_list\"\n\nCHUNKS=${#GPULIST[@]}\n\nCKPT"
},
{
"path": "train/dpo/scripts/v1_5/finetune.sh",
"chars": 1234,
"preview": "#!/bin/bash\n\ndeepspeed llava/train/train_mem.py \\\n --deepspeed ./scripts/zero3.json \\\n --model_name_or_path lmsys/"
},
{
"path": "train/dpo/scripts/v1_5/finetune_lora.sh",
"chars": 1317,
"preview": "#!/bin/bash\n\ndeepspeed llava/train/train_mem.py \\\n --lora_enable True --lora_r 128 --lora_alpha 256 --mm_projector_lr"
},
{
"path": "train/dpo/scripts/v1_5/finetune_task.sh",
"chars": 1156,
"preview": "#!/bin/bash\n\ndeepspeed llava/train/train_mem.py \\\n --deepspeed ./scripts/zero3.json \\\n --model_name_or_path liuhao"
},
{
"path": "train/dpo/scripts/v1_5/finetune_task_lora.sh",
"chars": 1239,
"preview": "#!/bin/bash\n\ndeepspeed llava/train/train_mem.py \\\n --lora_enable True --lora_r 128 --lora_alpha 256 --mm_projector_lr"
},
{
"path": "train/dpo/scripts/v1_5/pretrain.sh",
"chars": 1164,
"preview": "#!/bin/bash\n\ndeepspeed llava/train/train_mem.py \\\n --deepspeed ./scripts/zero2.json \\\n --model_name_or_path lmsys/"
},
{
"path": "train/dpo/scripts/zero2.json",
"chars": 561,
"preview": "{\n \"fp16\": {\n \"enabled\": \"auto\",\n \"loss_scale\": 0,\n \"loss_scale_window\": 1000,\n \"initial_"
},
{
"path": "train/dpo/scripts/zero3.json",
"chars": 801,
"preview": "{\n \"fp16\": {\n \"enabled\": \"auto\",\n \"loss_scale\": 0,\n \"loss_scale_window\": 1000,\n \"initial_"
},
{
"path": "train/dpo/scripts/zero3_offload.json",
"chars": 1279,
"preview": "{\n \"fp16\": {\n \"enabled\": \"auto\",\n \"loss_scale\": 0,\n \"loss_scale_window\": 1000,\n \"initial_scale_power\": 16,\n"
},
{
"path": "train/dpo/tool/dpo_trainer.py",
"chars": 40206,
"preview": "# DPO Authors: Rafael Rafailov, Archit Sharma, Eric Mitchell, Stefano Ermon, Christopher D. Manning, and Chelsea Finn 20"
},
{
"path": "train/dpo/tool/dpo_trainer_inherent.py",
"chars": 40896,
"preview": "# DPO Authors: Rafael Rafailov, Archit Sharma, Eric Mitchell, Stefano Ermon, Christopher D. Manning, and Chelsea Finn 20"
},
{
"path": "train/dpo/train_dpo_2stages.py",
"chars": 47784,
"preview": "# Adopted from https://github.com/lm-sys/FastChat. Below is the original copyright:\n# Adopted from tatsu-lab@stanford_al"
},
{
"path": "train/open_clip/CITATION.cff",
"chars": 824,
"preview": "cff-version: 1.1.0\nmessage: If you use this software, please cite it as below.\nauthors:\n - family-names: Ilharco\n gi"
},
{
"path": "train/open_clip/HISTORY.md",
"chars": 5142,
"preview": "## 2.24.0\n\n* Fix missing space in error message\n* use model flag for normalizing embeddings\n* init logit_bias for non si"
},
{
"path": "train/open_clip/LICENSE",
"chars": 1229,
"preview": "Copyright (c) 2012-2021 Gabriel Ilharco, Mitchell Wortsman, \nNicholas Carlini, Rohan Taori, Achal Dave, Vaishaal Shankar"
},
{
"path": "train/open_clip/MANIFEST.in",
"chars": 95,
"preview": "include src/open_clip/bpe_simple_vocab_16e6.txt.gz\ninclude src/open_clip/model_configs/*.json\n\n"
},
{
"path": "train/open_clip/setup.py",
"chars": 2104,
"preview": "\"\"\" Setup\n\"\"\"\nfrom setuptools import setup, find_packages\nfrom codecs import open\nfrom os import path\n\nhere = path.abspa"
},
{
"path": "train/open_clip/src/open_clip/__init__.py",
"chars": 1264,
"preview": "from .coca_model import CoCa\nfrom .constants import OPENAI_DATASET_MEAN, OPENAI_DATASET_STD\nfrom .factory import create_"
},
{
"path": "train/open_clip/src/open_clip/big_vision.py",
"chars": 7711,
"preview": "import torch\nimport numpy as np\n\nfrom .model import CustomTextCLIP\nfrom .transformer import TextTransformer, Transformer"
},
{
"path": "train/open_clip/src/open_clip/coca_model.py",
"chars": 18038,
"preview": "from typing import Optional\n\nimport torch\nfrom torch import nn\nfrom torch.nn import functional as F\nimport numpy as np\nf"
},
{
"path": "train/open_clip/src/open_clip/constants.py",
"chars": 256,
"preview": "OPENAI_DATASET_MEAN = (0.48145466, 0.4578275, 0.40821073)\nOPENAI_DATASET_STD = (0.26862954, 0.26130258, 0.27577711)\nIMAG"
},
{
"path": "train/open_clip/src/open_clip/factory.py",
"chars": 18003,
"preview": "import json\nimport logging\nimport os\nimport re\nfrom copy import deepcopy\nfrom dataclasses import asdict\nfrom pathlib imp"
},
{
"path": "train/open_clip/src/open_clip/hf_configs.py",
"chars": 2422,
"preview": "# HF architecture dict:\narch_dict = {\n # https://huggingface.co/docs/transformers/model_doc/roberta#roberta\n \"robe"
},
{
"path": "train/open_clip/src/open_clip/hf_model.py",
"chars": 6960,
"preview": "\"\"\" huggingface model adapter\n\nWraps HuggingFace transformers (https://github.com/huggingface/transformers) models for u"
},
{
"path": "train/open_clip/src/open_clip/loss.py",
"chars": 15383,
"preview": "import torch\nimport torch.nn as nn\nfrom torch.nn import functional as F\n\ntry:\n import torch.distributed.nn\n from t"
},
{
"path": "train/open_clip/src/open_clip/model.py",
"chars": 24431,
"preview": "\"\"\" CLIP Model\n\nAdapted from https://github.com/openai/CLIP. Originally MIT License, Copyright (c) 2021 OpenAI.\n\"\"\"\nimpo"
},
{
"path": "train/open_clip/src/open_clip/model_configs/EVA01-g-14-plus.json",
"chars": 401,
"preview": "{\n \"embed_dim\": 1024,\n \"vision_cfg\": {\n \"image_size\": 224,\n \"timm_model_name\": \"eva_giant_patch14_22"
},
{
"path": "train/open_clip/src/open_clip/model_configs/EVA01-g-14.json",
"chars": 400,
"preview": "{\n \"embed_dim\": 1024,\n \"vision_cfg\": {\n \"image_size\": 224,\n \"timm_model_name\": \"eva_giant_patch14_22"
},
{
"path": "train/open_clip/src/open_clip/model_configs/EVA02-B-16.json",
"chars": 404,
"preview": "{\n \"embed_dim\": 512,\n \"vision_cfg\": {\n \"image_size\": 224,\n \"timm_model_name\": \"eva02_base_patch16_cl"
},
{
"path": "train/open_clip/src/open_clip/model_configs/EVA02-E-14-plus.json",
"chars": 411,
"preview": "{\n \"embed_dim\": 1024,\n \"vision_cfg\": {\n \"image_size\": 224,\n \"timm_model_name\": \"eva02_enormous_patch"
},
{
"path": "train/open_clip/src/open_clip/model_configs/EVA02-E-14.json",
"chars": 411,
"preview": "{\n \"embed_dim\": 1024,\n \"vision_cfg\": {\n \"image_size\": 224,\n \"timm_model_name\": \"eva02_enormous_patch"
},
{
"path": "train/open_clip/src/open_clip/model_configs/EVA02-L-14-336.json",
"chars": 406,
"preview": "{\n \"embed_dim\": 768,\n \"vision_cfg\": {\n \"image_size\": 336,\n \"timm_model_name\": \"eva02_large_patch14_c"
},
{
"path": "train/open_clip/src/open_clip/model_configs/EVA02-L-14.json",
"chars": 406,
"preview": "{\n \"embed_dim\": 768,\n \"vision_cfg\": {\n \"image_size\": 224,\n \"timm_model_name\": \"eva02_large_patch14_c"
},
{
"path": "train/open_clip/src/open_clip/model_configs/RN101-quickgelu.json",
"chars": 388,
"preview": "{\n \"embed_dim\": 512,\n \"quick_gelu\": true,\n \"vision_cfg\": {\n \"image_size\": 224,\n \"layers\": [\n "
},
{
"path": "train/open_clip/src/open_clip/model_configs/RN101.json",
"chars": 364,
"preview": "{\n \"embed_dim\": 512,\n \"vision_cfg\": {\n \"image_size\": 224,\n \"layers\": [\n 3,\n 4,"
},
{
"path": "train/open_clip/src/open_clip/model_configs/RN50-quickgelu.json",
"chars": 389,
"preview": "{\n \"embed_dim\": 1024,\n \"quick_gelu\": true,\n \"vision_cfg\": {\n \"image_size\": 224,\n \"layers\": [\n "
},
{
"path": "train/open_clip/src/open_clip/model_configs/RN50.json",
"chars": 364,
"preview": "{\n \"embed_dim\": 1024,\n \"vision_cfg\": {\n \"image_size\": 224,\n \"layers\": [\n 3,\n 4"
},
{
"path": "train/open_clip/src/open_clip/model_configs/RN50x16.json",
"chars": 365,
"preview": "{\n \"embed_dim\": 768,\n \"vision_cfg\": {\n \"image_size\": 384,\n \"layers\": [\n 6,\n 8,"
},
{
"path": "train/open_clip/src/open_clip/model_configs/RN50x4.json",
"chars": 365,
"preview": "{\n \"embed_dim\": 640,\n \"vision_cfg\": {\n \"image_size\": 288,\n \"layers\": [\n 4,\n 6,"
},
{
"path": "train/open_clip/src/open_clip/model_configs/RN50x64.json",
"chars": 370,
"preview": "{\n \"embed_dim\": 1024,\n \"vision_cfg\": {\n \"image_size\": 448,\n \"layers\": [\n 3,\n 1"
},
{
"path": "train/open_clip/src/open_clip/model_configs/ViT-B-16-SigLIP-256.json",
"chars": 710,
"preview": "{\n \"embed_dim\": 768,\n \"init_logit_bias\": -10,\n \"custom_text\": true,\n \"vision_cfg\": {\n \"image_size\": 2"
},
{
"path": "train/open_clip/src/open_clip/model_configs/ViT-B-16-SigLIP-384.json",
"chars": 710,
"preview": "{\n \"embed_dim\": 768,\n \"init_logit_bias\": -10,\n \"custom_text\": true,\n \"vision_cfg\": {\n \"image_size\": 3"
},
{
"path": "train/open_clip/src/open_clip/model_configs/ViT-B-16-SigLIP-512.json",
"chars": 710,
"preview": "{\n \"embed_dim\": 768,\n \"init_logit_bias\": -10,\n \"custom_text\": true,\n \"vision_cfg\": {\n \"image_size\": 5"
},
{
"path": "train/open_clip/src/open_clip/model_configs/ViT-B-16-SigLIP-i18n-256.json",
"chars": 720,
"preview": "{\n \"embed_dim\": 768,\n \"init_logit_bias\": -10,\n \"custom_text\": true,\n \"vision_cfg\": {\n \"image_size\": 2"
},
{
"path": "train/open_clip/src/open_clip/model_configs/ViT-B-16-SigLIP.json",
"chars": 710,
"preview": "{\n \"embed_dim\": 768,\n \"init_logit_bias\": -10,\n \"custom_text\": true,\n \"vision_cfg\": {\n \"image_size\": 2"
},
{
"path": "train/open_clip/src/open_clip/model_configs/ViT-B-16-plus-240.json",
"chars": 295,
"preview": "{\n \"embed_dim\": 640,\n \"vision_cfg\": {\n \"image_size\": 240,\n \"layers\": 12,\n \"width\": 896,\n "
},
{
"path": "train/open_clip/src/open_clip/model_configs/ViT-B-16-plus.json",
"chars": 295,
"preview": "{\n \"embed_dim\": 640,\n \"vision_cfg\": {\n \"image_size\": 224,\n \"layers\": 12,\n \"width\": 896,\n "
},
{
"path": "train/open_clip/src/open_clip/model_configs/ViT-B-16-quickgelu.json",
"chars": 318,
"preview": "{\n \"embed_dim\": 512,\n \"quick_gelu\": true,\n \"vision_cfg\": {\n \"image_size\": 224,\n \"layers\": 12,\n "
},
{
"path": "train/open_clip/src/open_clip/model_configs/ViT-B-16.json",
"chars": 294,
"preview": "{\n \"embed_dim\": 512,\n \"vision_cfg\": {\n \"image_size\": 224,\n \"layers\": 12,\n \"width\": 768,\n "
},
{
"path": "train/open_clip/src/open_clip/model_configs/ViT-B-32-256.json",
"chars": 295,
"preview": "{\n \"embed_dim\": 512,\n \"vision_cfg\": {\n \"image_size\": 256,\n \"layers\": 12,\n \"width\": 768,\n "
},
{
"path": "train/open_clip/src/open_clip/model_configs/ViT-B-32-plus-256.json",
"chars": 295,
"preview": "{\n \"embed_dim\": 640,\n \"vision_cfg\": {\n \"image_size\": 256,\n \"layers\": 12,\n \"width\": 896,\n "
},
{
"path": "train/open_clip/src/open_clip/model_configs/ViT-B-32-quickgelu.json",
"chars": 318,
"preview": "{\n \"embed_dim\": 512,\n \"quick_gelu\": true,\n \"vision_cfg\": {\n \"image_size\": 224,\n \"layers\": 12,\n "
},
{
"path": "train/open_clip/src/open_clip/model_configs/ViT-B-32.json",
"chars": 294,
"preview": "{\n \"embed_dim\": 512,\n \"vision_cfg\": {\n \"image_size\": 224,\n \"layers\": 12,\n \"width\": 768,\n "
},
{
"path": "train/open_clip/src/open_clip/model_configs/ViT-H-14-378-quickgelu.json",
"chars": 348,
"preview": "{\n \"embed_dim\": 1024,\n \"quick_gelu\": true,\n \"vision_cfg\": {\n \"image_size\": 378,\n \"layers\": 32,\n "
},
{
"path": "train/open_clip/src/open_clip/model_configs/ViT-H-14-CLIPA-336.json",
"chars": 604,
"preview": "{\n \"embed_dim\": 1024,\n \"vision_cfg\": {\n \"image_size\": 336,\n \"layers\": 32,\n \"width\": 1280,\n "
},
{
"path": "train/open_clip/src/open_clip/model_configs/ViT-H-14-CLIPA.json",
"chars": 604,
"preview": "{\n \"embed_dim\": 1024,\n \"vision_cfg\": {\n \"image_size\": 224,\n \"layers\": 32,\n \"width\": 1280,\n "
},
{
"path": "train/open_clip/src/open_clip/model_configs/ViT-H-14-quickgelu.json",
"chars": 348,
"preview": "{\n \"embed_dim\": 1024,\n \"quick_gelu\": true,\n \"vision_cfg\": {\n \"image_size\": 224,\n \"layers\": 32,\n "
},
{
"path": "train/open_clip/src/open_clip/model_configs/ViT-H-14.json",
"chars": 324,
"preview": "{\n \"embed_dim\": 1024,\n \"vision_cfg\": {\n \"image_size\": 224,\n \"layers\": 32,\n \"width\": 1280,\n "
},
{
"path": "train/open_clip/src/open_clip/model_configs/ViT-H-16.json",
"chars": 324,
"preview": "{\n \"embed_dim\": 1024,\n \"vision_cfg\": {\n \"image_size\": 224,\n \"layers\": 32,\n \"width\": 1280,\n "
},
{
"path": "train/open_clip/src/open_clip/model_configs/ViT-L-14-280.json",
"chars": 296,
"preview": "{\n \"embed_dim\": 768,\n \"vision_cfg\": {\n \"image_size\": 280,\n \"layers\": 24,\n \"width\": 1024,\n "
},
{
"path": "train/open_clip/src/open_clip/model_configs/ViT-L-14-336.json",
"chars": 296,
"preview": "{\n \"embed_dim\": 768,\n \"vision_cfg\": {\n \"image_size\": 336,\n \"layers\": 24,\n \"width\": 1024,\n "
}
]
// ... and 83 more files (download for full content)
About this extraction
This page contains the full source code of the richard-peng-xia/MMed-RAG GitHub repository, extracted and formatted as plain text for AI agents and large language models (LLMs). The extraction includes 283 files (64.5 MB), approximately 10.8M tokens, and a symbol index with 925 extracted functions, classes, methods, constants, and types. Use this with OpenClaw, Claude, ChatGPT, Cursor, Windsurf, or any other AI tool that accepts text input. You can copy the full output to your clipboard or download it as a .txt file.
Extracted by GitExtract — free GitHub repo to text converter for AI. Built by Nikandr Surkov.