II-Medical-8B-1706 / README.md
sw_tuenguyen
Update README.md
2d5af31 verified
|
raw
history blame
8.93 kB
metadata
library_name: transformers
tags: []

II-Medical-8B-1706

I. Model Overview

II-Medical-8B-1706 is the newest advanced large language model developed by Intelligent Internet, specifically engineered to enhance AI-driven medical reasoning. Following the positive reception of our previous II-Medical-8B, this new iteration significantly advances the capabilities of medical question answering,

II. Training Methodology

We collected and generated a comprehensive set of reasoning datasets for the medical domain and performed SFT fine-tuning on the Qwen/Qwen3-8B model. Following this, we further optimized the SFT model by training DAPO on a hard-reasoning dataset to boost performance.

For SFT stage we using the hyperparameters:

  • Max Length: 16378.
  • Batch Size: 128.
  • Learning-Rate: 5e-5.
  • Number Of Epoch: 6.

For RL stage we setup training with:

  • Max prompt length: 2048 tokens.
  • Max response length: 12288 tokens.
  • Overlong buffer: Enabled, 4096 tokens, penalty factor 1.0.
  • Clip ratios: Low 0.2, High 0.28.
  • Batch sizes: Train prompt 512, Generation prompt 1536, Mini-batch 32.
  • Responses per prompt: 16.
  • Temperature: 1.0, Top-p: 1.0, Top-k: -1 (vLLM rollout).
  • Learning rate: 1e-6, Warmup steps: 10, Weight decay: 0.1.
  • Loss aggregation: Token-mean.
  • Gradient clipping: 1.0.
  • Entropy coefficient: 0.

III. Evaluation Results

Our II-Medical-8B model achieved a 41% score on HealthBench, a comprehensive open-source benchmark evaluating the performance and safety of large language models in healthcare. This performance is comparable to OpenAI's o1 reasoning model and GPT-4.5, OpenAI's largest and most advanced model to date. We provide a comparison to models available in ChatGPT below.

image/jpeg Detailed result for HealthBench can be found here.

Model Benchmark

We evaluate on ten medical QA benchmarks include MedMCQA, MedQA, PubMedQA, HealthBench, medical related questions from MMLU-Pro, small QA sets from Lancet and the New England Journal of Medicine, 4 Options and 5 Options splits from the MedBullets platform and MedXpertQA.

Model MedMC MedQA PubMed MMLU-P HealthBench Lancet MedB-4 MedB-5 MedX NEJM Avg
HuatuoGPT-o1-72B 76.76 88.85 79.90 80.46 22.73 70.87 77.27 73.05 23.53 76.29 66.97
M1 62.54 75.81 75.80 65.86 15.51 62.62 63.64 59.74 19.59 64.34 56.55
Qwen3-8B 66.53 81.38 73.9 77.85 42.27 66.26 68.83 62.66 19.59 69.65 62.89
Qwen3-32B 74.18 88.92 76.1 80.7 47.08 72.33 72.27 71.42 28.04 76.94 68.80
MedGemma-27B-IT 73.24 87.27 70.9 80.13 46.54 70.14 75.32 73.37 25.55 76.28 67.87
II-Medical-8B 71.57 87.90 78.7 80.46 40.02 70.38 78.25 72.07 25.26 73.13 67.77
II-Medical-8B-1706 74.44 88.61 79.8 81.04 46.8 71.60 80.84 74.67 29.63 77.61 70.5

IV. Dataset Curation

The training dataset comprises 2.3M samples from the following sources:

1. Public Medical Reasoning Datasets

2. Synthetic Medical QA Data with Qwen3-235B-A22B

Generated from established medical datasets:

3. Curated Medical R1 Traces (338,055 samples)

First we gather all the public R1 traces from:

All R1 reasoning traces were processed through a domain-specific pipeline as follows:

  1. Embedding Generation: Prompts are embedded using sentence-transformers/all-MiniLM-L6-v2.

  2. Clustering: Perform K-means clustering with 50,000 clusters.

  3. Domain Classification:

    • For each cluster, select the 10 prompts nearest to the cluster center.
    • Classify the domain of each selected prompt using Qwen2.5-32b-Instruct.
    • Assign the cluster's domain based on majority voting among the classified prompts.
  4. Domain Filtering: Keep only clusters labeled as Medical or Biology for the final dataset.

4. Other

Preprocessing Data

  1. Filtering for Complete Generation

    • Retained only traces with complete generation outputs
  2. Length-based Filtering

    • Minimum threshold: Keep only the prompt with more than 3 words.
    • Wait Token Filter: Removed traces with has more than 47 occurrences of "Wait" (97th percentile threshold).
  3. Response Deduplicate

    • Ngram: 4
    • Jacard Threshold: 0.7

Data Decontamination

We using two step decontamination:

  1. Following open-r1 project: We decontaminate a dataset using 10-grams with the evaluation datasets.
  2. After that, we using the fuzzy decontamination from s1k method with threshold 90%.

Our pipeline is carefully decontaminated with the evaluation datasets.

V. How To Use

Our model can be utilized in the same manner as Qwen or Deepseek-R1-Distill models.

For instance, you can easily start a service using vLLM:

vllm serve Intelligent-Internet/II-Medical-8B-1706

You can also easily start a service using SGLang:

python -m sglang.launch_server --model Intelligent-Internet/II-Medical-8B-1706

VI. Usage Guidelines

  • Recommended Sampling Parameters: temperature = 0.6, top_p = 0.9
  • When using, explicitly request step-by-step reasoning and format the final answer within \boxed{} (e.g., "Please reason step-by-step, and put your final answer within \boxed{}.").

VII. Limitations and Considerations

  • Dataset may contain inherent biases from source materials
  • Medical knowledge requires regular updates
  • Please note that It’s not suitable for medical use.

VIII. Citation

@misc{2025II-Medical-8B-1706,
      title={II-Medical-8B: Medical Reasoning Model}, 
      author={Intelligent Internet},
      year={2025}
}