Dataset Viewer
id
stringlengths 11
95
| author
stringlengths 3
36
| task_category
stringclasses 16
values | tags
sequencelengths 1
4.05k
| created_time
timestamp[s]date 2022-03-02 23:29:04
2025-03-18 02:34:30
| last_modified
timestamp[s]date 2021-05-13 19:09:22
2025-04-17 04:22:08
| downloads
int64 0
15.6M
| likes
int64 0
4.86k
| README
stringlengths 246
1.01M
| matched_task
sequencelengths 1
8
| matched_bigbio_names
sequencelengths 1
8
| is_bionlp
stringclasses 3
values | model_cards
stringlengths 0
1M
| metadata
stringlengths 2
698k
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Goodmotion/spam-mail-classifier | Goodmotion | text-classification | [
"transformers",
"safetensors",
"text-classification",
"spam-detection",
"license:apache-2.0",
"endpoints_compatible",
"region:us"
] | 2024-12-09T15:56:31 | 2024-12-09T19:35:48 | 87 | 2 | ---
license: apache-2.0
tags:
- transformers
- text-classification
- spam-detection
---
# SPAM Mail Classifier
This model is fine-tuned from `microsoft/Multilingual-MiniLM-L12-H384` to classify email subjects as SPAM or NOSPAM.
## Model Details
- **Base model**: `microsoft/Multilingual-MiniLM-L12-H384`
- **Fine-tuned for**: Text classification
- **Number of classes**: 2 (SPAM, NOSPAM)
- **Languages**: Multilingual
## Usage
This model is fine-tuned from `microsoft/Multilingual-MiniLM-L12-H384` to classify email subjects as SPAM or NOSPAM.
```python
from transformers import AutoTokenizer, AutoModelForSequenceClassification
model_name = "Goodmotion/spam-mail-classifier"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForSequenceClassification.from_pretrained(
model_name
)
text = "Félicitations ! Vous avez gagné un iPhone."
inputs = tokenizer(text, return_tensors="pt")
outputs = model(**inputs)
print(outputs.logits)
```
### Exemple for list
```python
import torch
from transformers import AutoTokenizer, AutoModelForSequenceClassification
model_name = "Goodmotion/spam-mail-classifier"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForSequenceClassification.from_pretrained(model_name)
texts = [
'Join us for a webinar on AI innovations',
'Urgent: Verify your account immediately.',
'Meeting rescheduled to 3 PM',
'Happy Birthday!',
'Limited time offer: Act now!',
'Join us for a webinar on AI innovations',
'Claim your free prize now!',
'You have unclaimed rewards waiting!',
'Weekly newsletter from Tech World',
'Update on the project status',
'Lunch tomorrow at 12:30?',
'Get rich quick with this amazing opportunity!',
'Invoice for your recent purchase',
'Don\'t forget: Gym session at 6 AM',
'Join us for a webinar on AI innovations',
'bonjour comment allez vous ?',
'Documents suite à notre rendez-vous',
'Valentin Dupond mentioned you in a comment',
'Bolt x Supabase = 🤯',
'Modification site web de la société',
'Image de mise en avant sur les articles',
'Bring new visitors to your site',
'Le Cloud Éthique sans bullshit',
'Remix Newsletter #25: React Router v7',
'Votre essai auprès de X va bientôt prendre fin',
'Introducing a Google Docs integration, styles and more in Claude.ai',
'Carte de crédit sur le point d’expirer sur Cloudflare'
]
inputs = tokenizer(texts, padding=True, truncation=True, max_length=128, return_tensors="pt")
outputs = model(**inputs)
# Convertir les logits en probabilités avec softmax
logits = outputs.logits
probabilities = torch.softmax(logits, dim=1)
# Décoder les classes pour chaque texte
labels = ["NOSPAM", "SPAM"] # Mapping des indices à des labels
results = [
{"text": text, "label": labels[torch.argmax(prob).item()], "confidence": prob.max().item()}
for text, prob in zip(texts, probabilities)
]
# Afficher les résultats
for result in results:
print(f"Texte : {result['text']}")
print(f"Résultat : {result['label']} (Confiance : {result['confidence']:.2%})\n")
```
| [
"TEXT_CLASSIFICATION"
] | [
"ESSAI"
] | Non_BioNLP | # SPAM Mail Classifier
This model is fine-tuned from `microsoft/Multilingual-MiniLM-L12-H384` to classify email subjects as SPAM or NOSPAM.
## Model Details
- **Base model**: `microsoft/Multilingual-MiniLM-L12-H384`
- **Fine-tuned for**: Text classification
- **Number of classes**: 2 (SPAM, NOSPAM)
- **Languages**: Multilingual
## Usage
This model is fine-tuned from `microsoft/Multilingual-MiniLM-L12-H384` to classify email subjects as SPAM or NOSPAM.
```python
from transformers import AutoTokenizer, AutoModelForSequenceClassification
model_name = "Goodmotion/spam-mail-classifier"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForSequenceClassification.from_pretrained(
model_name
)
text = "Félicitations ! Vous avez gagné un iPhone."
inputs = tokenizer(text, return_tensors="pt")
outputs = model(**inputs)
print(outputs.logits)
```
### Exemple for list
```python
import torch
from transformers import AutoTokenizer, AutoModelForSequenceClassification
model_name = "Goodmotion/spam-mail-classifier"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForSequenceClassification.from_pretrained(model_name)
texts = [
'Join us for a webinar on AI innovations',
'Urgent: Verify your account immediately.',
'Meeting rescheduled to 3 PM',
'Happy Birthday!',
'Limited time offer: Act now!',
'Join us for a webinar on AI innovations',
'Claim your free prize now!',
'You have unclaimed rewards waiting!',
'Weekly newsletter from Tech World',
'Update on the project status',
'Lunch tomorrow at 12:30?',
'Get rich quick with this amazing opportunity!',
'Invoice for your recent purchase',
'Don\'t forget: Gym session at 6 AM',
'Join us for a webinar on AI innovations',
'bonjour comment allez vous ?',
'Documents suite à notre rendez-vous',
'Valentin Dupond mentioned you in a comment',
'Bolt x Supabase = 🤯',
'Modification site web de la société',
'Image de mise en avant sur les articles',
'Bring new visitors to your site',
'Le Cloud Éthique sans bullshit',
'Remix Newsletter #25: React Router v7',
'Votre essai auprès de X va bientôt prendre fin',
'Introducing a Google Docs integration, styles and more in Claude.ai',
'Carte de crédit sur le point d’expirer sur Cloudflare'
]
inputs = tokenizer(texts, padding=True, truncation=True, max_length=128, return_tensors="pt")
outputs = model(**inputs)
# Convertir les logits en probabilités avec softmax
logits = outputs.logits
probabilities = torch.softmax(logits, dim=1)
# Décoder les classes pour chaque texte
labels = ["NOSPAM", "SPAM"] # Mapping des indices à des labels
results = [
{"text": text, "label": labels[torch.argmax(prob).item()], "confidence": prob.max().item()}
for text, prob in zip(texts, probabilities)
]
# Afficher les résultats
for result in results:
print(f"Texte : {result['text']}")
print(f"Résultat : {result['label']} (Confiance : {result['confidence']:.2%})\n")
```
| {"license": "apache-2.0", "tags": ["transformers", "text-classification", "spam-detection"]} |
knowledgator/gliner-poly-small-v1.0 | knowledgator | token-classification | [
"gliner",
"pytorch",
"token-classification",
"multilingual",
"dataset:urchade/pile-mistral-v0.1",
"dataset:numind/NuNER",
"dataset:knowledgator/GLINER-multi-task-synthetic-data",
"license:apache-2.0",
"region:us"
] | 2024-08-19T12:40:53 | 2024-08-25T11:38:05 | 32 | 14 | ---
datasets:
- urchade/pile-mistral-v0.1
- numind/NuNER
- knowledgator/GLINER-multi-task-synthetic-data
language:
- multilingual
library_name: gliner
license: apache-2.0
pipeline_tag: token-classification
---
# About
GLiNER is a Named Entity Recognition (NER) model capable of identifying any entity type using a bidirectional transformer encoders (BERT-like). It provides a practical alternative to traditional NER models, which are limited to predefined entities, and Large Language Models (LLMs) that, despite their flexibility, are costly and large for resource-constrained scenarios.
This particular version utilize bi-encoder architecture with post-fusion, where textual encoder is [DeBERTa v3 small](microsoft/deberta-v3-small) and entity label encoder is sentence transformer - [BGE-small-en](https://huggingface.co/BAAI/bge-small-en-v1.5).
Such architecture brings several advantages over uni-encoder GLiNER:
* An unlimited amount of entities can be recognized at a single time;
* Faster inference if entity embeddings are preprocessed;
* Better generalization to unseen entities;
Post fusion strategy brings advantages over classical bi-encoder enabling better inter-label understanding.
### Installation & Usage
Install or update the gliner package:
```bash
pip install gliner -U
```
Once you've downloaded the GLiNER library, you can import the GLiNER class. You can then load this model using `GLiNER.from_pretrained` and predict entities with `predict_entities`.
```python
from gliner import GLiNER
model = GLiNER.from_pretrained("knowledgator/gliner-poly-small-v1.0")
text = """
Cristiano Ronaldo dos Santos Aveiro (Portuguese pronunciation: [kɾiʃˈtjɐnu ʁɔˈnaldu]; born 5 February 1985) is a Portuguese professional footballer who plays as a forward for and captains both Saudi Pro League club Al Nassr and the Portugal national team. Widely regarded as one of the greatest players of all time, Ronaldo has won five Ballon d'Or awards,[note 3] a record three UEFA Men's Player of the Year Awards, and four European Golden Shoes, the most by a European player. He has won 33 trophies in his career, including seven league titles, five UEFA Champions Leagues, the UEFA European Championship and the UEFA Nations League. Ronaldo holds the records for most appearances (183), goals (140) and assists (42) in the Champions League, goals in the European Championship (14), international goals (128) and international appearances (205). He is one of the few players to have made over 1,200 professional career appearances, the most by an outfield player, and has scored over 850 official senior career goals for club and country, making him the top goalscorer of all time.
"""
labels = ["person", "award", "date", "competitions", "teams"]
entities = model.predict_entities(text, labels, threshold=0.25)
for entity in entities:
print(entity["text"], "=>", entity["label"])
```
```
Cristiano Ronaldo dos Santos Aveiro => person
5 February 1985 => date
Al Nassr => teams
Portugal national team => teams
Ballon d'Or => award
UEFA Men's Player of the Year Awards => award
European Golden Shoes => award
UEFA Champions Leagues => competitions
UEFA European Championship => competitions
UEFA Nations League => competitions
Champions League => competitions
European Championship => competitions
```
If you have a large amount of entities and want to pre-embed them, please, refer to the following code snippet:
```python
labels = ["your entities"]
texts = ["your texts"]
entity_embeddings = model.encode_labels(labels, batch_size = 8)
outputs = model.batch_predict_with_embeds([text], entity_embeddings, labels)
```
### Benchmarks
Below you can see the table with benchmarking results on various named entity recognition datasets:
| Dataset | Score |
|---------|-------|
| ACE 2004 | 25.4% |
| ACE 2005 | 27.2% |
| AnatEM | 17.7% |
| Broad Tweet Corpus | 70.2% |
| CoNLL 2003 | 67.8% |
| FabNER | 22.9% |
| FindVehicle | 40.2% |
| GENIA_NER | 47.7% |
| HarveyNER | 15.5% |
| MultiNERD | 64.5% |
| Ontonotes | 28.7% |
| PolyglotNER | 47.5% |
| TweetNER7 | 39.3% |
| WikiANN en | 56.7% |
| WikiNeural | 80.0% |
| bc2gm | 56.2% |
| bc4chemd | 48.7% |
| bc5cdr | 60.5% |
| ncbi | 53.5% |
| **Average** | **45.8%** |
|||
| CrossNER_AI | 48.9% |
| CrossNER_literature | 64.0% |
| CrossNER_music | 68.7% |
| CrossNER_politics | 69.0% |
| CrossNER_science | 62.7% |
| mit-movie | 40.3% |
| mit-restaurant | 36.2% |
| **Average (zero-shot benchmark)** | **55.7%** |
### Join Our Discord
Connect with our community on Discord for news, support, and discussion about our models. Join [Discord](https://discord.gg/dkyeAgs9DG). | [
"NAMED_ENTITY_RECOGNITION"
] | [
"ANATEM",
"BC5CDR"
] | Non_BioNLP |
# About
GLiNER is a Named Entity Recognition (NER) model capable of identifying any entity type using a bidirectional transformer encoders (BERT-like). It provides a practical alternative to traditional NER models, which are limited to predefined entities, and Large Language Models (LLMs) that, despite their flexibility, are costly and large for resource-constrained scenarios.
This particular version utilize bi-encoder architecture with post-fusion, where textual encoder is [DeBERTa v3 small](microsoft/deberta-v3-small) and entity label encoder is sentence transformer - [BGE-small-en](https://huggingface.co/BAAI/bge-small-en-v1.5).
Such architecture brings several advantages over uni-encoder GLiNER:
* An unlimited amount of entities can be recognized at a single time;
* Faster inference if entity embeddings are preprocessed;
* Better generalization to unseen entities;
Post fusion strategy brings advantages over classical bi-encoder enabling better inter-label understanding.
### Installation & Usage
Install or update the gliner package:
```bash
pip install gliner -U
```
Once you've downloaded the GLiNER library, you can import the GLiNER class. You can then load this model using `GLiNER.from_pretrained` and predict entities with `predict_entities`.
```python
from gliner import GLiNER
model = GLiNER.from_pretrained("knowledgator/gliner-poly-small-v1.0")
text = """
Cristiano Ronaldo dos Santos Aveiro (Portuguese pronunciation: [kɾiʃˈtjɐnu ʁɔˈnaldu]; born 5 February 1985) is a Portuguese professional footballer who plays as a forward for and captains both Saudi Pro League club Al Nassr and the Portugal national team. Widely regarded as one of the greatest players of all time, Ronaldo has won five Ballon d'Or awards,[note 3] a record three UEFA Men's Player of the Year Awards, and four European Golden Shoes, the most by a European player. He has won 33 trophies in his career, including seven league titles, five UEFA Champions Leagues, the UEFA European Championship and the UEFA Nations League. Ronaldo holds the records for most appearances (183), goals (140) and assists (42) in the Champions League, goals in the European Championship (14), international goals (128) and international appearances (205). He is one of the few players to have made over 1,200 professional career appearances, the most by an outfield player, and has scored over 850 official senior career goals for club and country, making him the top goalscorer of all time.
"""
labels = ["person", "award", "date", "competitions", "teams"]
entities = model.predict_entities(text, labels, threshold=0.25)
for entity in entities:
print(entity["text"], "=>", entity["label"])
```
```
Cristiano Ronaldo dos Santos Aveiro => person
5 February 1985 => date
Al Nassr => teams
Portugal national team => teams
Ballon d'Or => award
UEFA Men's Player of the Year Awards => award
European Golden Shoes => award
UEFA Champions Leagues => competitions
UEFA European Championship => competitions
UEFA Nations League => competitions
Champions League => competitions
European Championship => competitions
```
If you have a large amount of entities and want to pre-embed them, please, refer to the following code snippet:
```python
labels = ["your entities"]
texts = ["your texts"]
entity_embeddings = model.encode_labels(labels, batch_size = 8)
outputs = model.batch_predict_with_embeds([text], entity_embeddings, labels)
```
### Benchmarks
Below you can see the table with benchmarking results on various named entity recognition datasets:
| Dataset | Score |
|---------|-------|
| ACE 2004 | 25.4% |
| ACE 2005 | 27.2% |
| AnatEM | 17.7% |
| Broad Tweet Corpus | 70.2% |
| CoNLL 2003 | 67.8% |
| FabNER | 22.9% |
| FindVehicle | 40.2% |
| GENIA_NER | 47.7% |
| HarveyNER | 15.5% |
| MultiNERD | 64.5% |
| Ontonotes | 28.7% |
| PolyglotNER | 47.5% |
| TweetNER7 | 39.3% |
| WikiANN en | 56.7% |
| WikiNeural | 80.0% |
| bc2gm | 56.2% |
| bc4chemd | 48.7% |
| bc5cdr | 60.5% |
| ncbi | 53.5% |
| **Average** | **45.8%** |
|||
| CrossNER_AI | 48.9% |
| CrossNER_literature | 64.0% |
| CrossNER_music | 68.7% |
| CrossNER_politics | 69.0% |
| CrossNER_science | 62.7% |
| mit-movie | 40.3% |
| mit-restaurant | 36.2% |
| **Average (zero-shot benchmark)** | **55.7%** |
### Join Our Discord
Connect with our community on Discord for news, support, and discussion about our models. Join [Discord](https://discord.gg/dkyeAgs9DG). | {"datasets": ["urchade/pile-mistral-v0.1", "numind/NuNER", "knowledgator/GLINER-multi-task-synthetic-data"], "language": ["multilingual"], "library_name": "gliner", "license": "apache-2.0", "pipeline_tag": "token-classification"} |
QuantFactory/meditron-7b-GGUF | QuantFactory | null | [
"gguf",
"en",
"dataset:epfl-llm/guidelines",
"arxiv:2311.16079",
"base_model:meta-llama/Llama-2-7b",
"base_model:quantized:meta-llama/Llama-2-7b",
"license:llama2",
"endpoints_compatible",
"region:us"
] | 2024-09-28T14:59:14 | 2024-09-28T15:51:52 | 206 | 1 | ---
base_model: meta-llama/Llama-2-7b
datasets:
- epfl-llm/guidelines
language:
- en
license: llama2
metrics:
- accuracy
- perplexity
---
[](https://hf.co/QuantFactory)
# QuantFactory/meditron-7b-GGUF
This is quantized version of [epfl-llm/meditron-7b](https://huggingface.co/epfl-llm/meditron-7b) created using llama.cpp
# Original Model Card
<img width=50% src="meditron_LOGO.png" alt="Alt text" title="Meditron-logo">
# Model Card for Meditron-7B-v1.0
Meditron is a suite of open-source medical Large Language Models (LLMs).
Meditron-7B is a 7 billion parameters model adapted to the medical domain from Llama-2-7B through continued pretraining on a comprehensively curated medical corpus, including selected PubMed articles, abstracts, a [new dataset](https://huggingface.co/datasets/epfl-llm/guidelines) of internationally-recognized medical guidelines, and general domain data from [RedPajama-v1](https://huggingface.co/datasets/togethercomputer/RedPajama-Data-1T).
Meditron-7B, finetuned on relevant training data, outperforms Llama-2-7B and PMC-Llama on multiple medical reasoning tasks.
<details open>
<summary><strong>Advisory Notice</strong></summary>
<blockquote style="padding: 10px; margin: 0 0 10px; border-left: 5px solid #ddd;">
While Meditron is designed to encode medical knowledge from sources of high-quality evidence, it is not yet adapted to deliver this knowledge appropriately, safely, or within professional actionable constraints.
We recommend against deploying Meditron in medical applications without extensive use-case alignment, as well as additional testing, specifically including randomized controlled trials in real-world practice settings.
</blockquote>
</details>
## Model Details
- **Developed by:** [EPFL LLM Team](https://huggingface.co/epfl-llm)
- **Model type:** Causal decoder-only transformer language model
- **Language(s):** English (mainly)
- **Model License:** [LLAMA 2 COMMUNITY LICENSE AGREEMENT](https://huggingface.co/meta-llama/Llama-2-70b/raw/main/LICENSE.txt)
- **Code License:** [APACHE 2.0 LICENSE](LICENSE)
- **Continue-pretrained from model:** [Llama-2-7B](https://huggingface.co/meta-llama/Llama-2-7b)
- **Context length:** 2K tokens
- **Input:** Text-only data
- **Output:** Model generates text only
- **Status:** This is a static model trained on an offline dataset. Future versions of the tuned models will be released as we enhance model's performance.
- **Knowledge Cutoff:** August 2023
### Model Sources
- **Repository:** [epflLLM/meditron](https://github.com/epfLLM/meditron)
- **Trainer:** [epflLLM/Megatron-LLM](https://github.com/epfLLM/Megatron-LLM)
- **Paper:** *[MediTron-70B: Scaling Medical Pretraining for Large Language Models](https://arxiv.org/abs/2311.16079)*
## Uses
Meditron-7B is being made available for further testing and assessment as an AI assistant to enhance clinical decision-making and enhance access to an LLM for healthcare use. Potential use cases may include but are not limited to:
- Medical exam question answering
- Supporting differential diagnosis
- Disease information (symptoms, cause, treatment) query
- General health information query
### Direct Use
It is possible to use this model to generate text, which is useful for experimentation and understanding its capabilities.
It should not be used directly for production or work that may impact people.
### Downstream Use
Meditron-70B and Meditron-7B are both foundation models without finetuning or instruction-tuning. They can be finetuned, instruction-tuned, or RLHF-tuned for specific downstream tasks and applications.
There are two ways we have used this model for downstream question-answering tasks.
1. We apply in-context learning with k demonstrations (3 or 5 in our paper) added to the prompt.
2. We finetuned the models for downstream question-answering tasks using specific training sets.
We encourage and look forward to the adaption of the base model for more diverse applications.
If you want a more interactive way to prompt the model, we recommend using a high-throughput and memory-efficient inference engine with a UI that supports chat and text generation.
You can check out our deployment [guide](https://github.com/epfLLM/meditron/blob/main/deployment/README.md), where we used [FastChat](https://github.com/lm-sys/FastChat) with [vLLM](https://github.com/vllm-project/vllm). We collected generations for our qualitative analysis through an interactive UI platform, [BetterChatGPT](https://github.com/ztjhz/BetterChatGPT). Here is the prompt format we used as an example:
<img width=70% src="prompt_example.png" alt="qualitative-analysis-prompt" title="Qualitative Analysis Prompt">
### Out-of-Scope Use
We do not recommend using this model for natural language generation in a production environment, finetuned or otherwise.
## Truthfulness, Helpfulness, Risk, and Bias
<!-- This section is meant to convey both technical and sociotechnical limitations. -->
We did an initial assessment of Meditron models' **Truthfulness** against baseline models and consumer-level medical models.
We use TruthfulQA (multiple choice) as the main evaluation benchmark.
We only focus on the categories that are relevant to the medical domain, including Health, Nutrition, Psychology, and Science.
For 7B models, we perform one-shot evaluations for consistent answer generation.
For 70B models, the evaluations are under the zero-shot setting.
Below, we report the detailed truthfulness performance of each category.
| | | | | | | | |
| --- | ------ |----- |----- |----- |----- |----- |----- |
|Category | meditron-70b | llama-2-70b | med42-70b* | meditron-7b | llama-2-7b | PMC-llama-7b |
|Health | 81.8 | 69.1 | 83.6 | 27.3 | 16.4 | 3.6 |
|Nutrition | 77.9 | 68.8 | 62.5 | 31.1 | 12.5 | 6.3 |
|Psychology| 47.4 | 36.8 | 52.6 | 21.1 | 10.5 | 0.0 |
|Science | 77.8 | 44.4 | 33.3 | 33.3 | 11.1 | 0.0 |
|Avg | 71.2 | 54.8 | 58.0 | 28.3 | 12.6 | 2.5 |
| | | | | | | |
For a more detailed performance analysis, please see our paper.
Significant research is still required to fully explore potential bias, fairness, and safety issues with this language model.
Please recognize that our evaluation on Meditron-7B's helpfulness, risk, and bias are highly limited.
Thus, as we noted in the safety notice, we strongly against any deployment in medical applications without further alignment process and rigorous evaluation!
### Recommendations
**IMPORTANT!**
Users (both direct and downstream) should be made aware of the risks, biases, and limitations of the model.
While this model is capable of generating natural language text, we have only begun to explore this capability and its limitations.
Understanding these limitations is especially important in a domain like medicine.
Therefore, we strongly recommend against using this model in production for natural language generation or for professional purposes related to health and medicine.
## Training Details
### Training Data
Meditron’s domain-adaptive pre-training corpus GAP-Replay combines 48.1B tokens from four corpora:
- [**Clinical Guidelines**](https://huggingface.co/datasets/epfl-llm/guidelines): a new dataset of 46K internationally-recognized clinical practice guidelines from various healthcare-related sources, including hospitals and international organizations.
- **Medical Paper Abstracts**: 16.1M abstracts extracted from closed-access PubMed and PubMed Central papers.
- **Medical Papers**: full-text articles extracted from 5M publicly available PubMed and PubMed Central papers.
- **Replay Data**: 400M tokens of general domain pretraining data sampled from [RedPajama-v1](https://huggingface.co/datasets/togethercomputer/RedPajama-Data-1T)
<img width=75% src="gap-replay.png" alt="Alt text" title="Meditron-logo">
#### Data Preprocessing
Please see the detailed preprocessing procedure in our paper.
### Training Procedure
We used the [Megatron-LLM](https://github.com/epfLLM/Megatron-LLM) distributed training library, a derivative of Nvidia's Megatron LM project, to optimize training efficiency.
Hardware consists of 1 node of 8x NVIDIA A100 (80GB) SXM GPUs connected by NVLink and NVSwitch with a single Nvidia ConnectX-6 DX network card and equipped with 2 x AMD EPYC 7543 32-Core Processors and 512 GB of RAM.
Our three way parallelism scheme uses:
- Data Parallelism (DP -- different GPUs process different subsets of the batches) of 2,
- Pipeline Parallelism (PP -- different GPUs process different layers) of 4,
- Tensor Parallelism (TP -- different GPUs process different subtensors for matrix multiplication) of 1.
#### Training Hyperparameters
| | |
| --- | ------ |
| bf16 | true |
| lr | 3e-4 |
| eps | 1e-5 |
| betas | \[0.9, 0.95\] |
| clip_grad | 1 |
| weight decay | 0.1 |
| DP size | 16 |
| TP size | 4 |
| PP size | 1 |
| seq length | 2048 |
| lr scheduler | cosine|
| min lr | 1e-6 |
| warmup iteration | 2000 |
| micro batch size | 10 |
| global batch size | 1600 |
| | |
#### Sizes
The model was trained in September 2023.
The model architecture is exactly Llama 2, meaning
| | |
| --- | ------ |
| Model size | 7B |
| Hidden dimension | 4096 |
| Num. attention heads | 32 |
| Num. layers | 32 |
| | |
## Evaluation
<!-- This section describes the evaluation protocols and provides the results. -->
### Testing Data & Metrics
#### Testing Data
- [MedQA (USMLE)](https://huggingface.co/datasets/bigbio/med_qa)
- [MedMCQA](https://huggingface.co/datasets/medmcqa)
- [PubMedQA](https://huggingface.co/datasets/bigbio/pubmed_qa)
- [MMLU-Medical](https://huggingface.co/datasets/lukaemon/mmlu)
- [MedQA-4-Option](https://huggingface.co/datasets/GBaker/MedQA-USMLE-4-options)
#### Metrics
- Accuracy: suite the evaluation of multiple-choice question-answering tasks.
### Results
We finetune meditron-7b, llama-2-7b, pmc-llama-7b on each benchmark (pubmedqa, medmcqa, medqa)'s training data individually.
We report the finetuned models' performance with top token selection as the inference mode.
For MMLU-Medical, models finetuned on MedMCQA are used for inference.
For MedQA-4-Option, models finetuned on MedQA are used for inference.
For a more detailed performance analysis, please see our paper.
| | | | | | |
| --- | ------ |----- |----- |----- |----- |
|Dataset | meditron-7b | llama-2-7b | pmc-llama-7b | Zephyr-7B-beta* | Mistral-7B-instruct* |
|MMLU-Medical | 54.2 | 53.7 | 56.4 | 63.3 | 60.0 |
|PubMedQA | 74.4 | 61.8 | 59.2 | 46.0 | 17.8 |
|MedMCQA | 59.2 | 54.4 | 57.6 | 43.0 | 40.2 |
|MedQA | 47.9 | 44.0 | 42.4 | 42.8 | 32.4 |
|MedQA-4-Option| 52.0 | 49.6 | 49.2 | 48.5 | 41.1 |
|Avg | 57.5 | 52.7 | 53.0 | 48.7 | 38.3 |
| | | | | | |
**Note**: models with * are already instruction-tuned, so we exclude them from further finetuning on any training data.
## Environmental Impact
<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
- **Hardware Type:** 8 x NVIDIA A100 (80GB) SXM
- **Total GPU hours:** 588.8
- **Hardware Provider:** EPFL Research Computing Platform
- **Compute Region:** Switzerland
- **Carbon Emitted:** Switzerland has a carbon efficiency of 0.016 kgCO2/kWh (https://www.carbonfootprint.com/docs/2018_8_electricity_factors_august_2018_-_online_sources.pdf). 73.6 hours of 8 A100s means 588.8 hours at a TDP of 400W. Assuming a Power Usage effectiveness of 1.5, total emissions are estimated to be:
(400W / 1000W/kWh / GPU * 0.016 kgCO2/kWh * 73.6 h * 8 GPU) * 1.8 PUE = 6.8 kgCO2.
## Citation
**BibTeX:**
If you use Meditron or its training data, please cite our work:
```
@misc{chen2023meditron70b,
title={MEDITRON-70B: Scaling Medical Pretraining for Large Language Models},
author={Zeming Chen and Alejandro Hernández-Cano and Angelika Romanou and Antoine Bonnet and Kyle Matoba and Francesco Salvi and Matteo Pagliardini and Simin Fan and Andreas Köpf and Amirkeivan Mohtashami and Alexandre Sallinen and Alireza Sakhaeirad and Vinitra Swamy and Igor Krawczuk and Deniz Bayazit and Axel Marmet and Syrielle Montariol and Mary-Anne Hartley and Martin Jaggi and Antoine Bosselut},
year={2023},
eprint={2311.16079},
archivePrefix={arXiv},
primaryClass={cs.CL}
}
@software{epfmedtrn,
author = {Zeming Chen and Alejandro Hernández-Cano and Angelika Romanou and Antoine Bonnet and Kyle Matoba and Francesco Salvi and Matteo Pagliardini and Simin Fan and Andreas Köpf and Amirkeivan Mohtashami and Alexandre Sallinen and Alireza Sakhaeirad and Vinitra Swamy and Igor Krawczuk and Deniz Bayazit and Axel Marmet and Syrielle Montariol and Mary-Anne Hartley and Martin Jaggi and Antoine Bosselut},
title = {MediTron-70B: Scaling Medical Pretraining for Large Language Models},
month = November,
year = 2023,
url = {https://github.com/epfLLM/meditron}
}
```
| [
"QUESTION_ANSWERING"
] | [
"MEDQA",
"PUBMEDQA"
] | BioNLP |
[](https://hf.co/QuantFactory)
# QuantFactory/meditron-7b-GGUF
This is quantized version of [epfl-llm/meditron-7b](https://huggingface.co/epfl-llm/meditron-7b) created using llama.cpp
# Original Model Card
<img width=50% src="meditron_LOGO.png" alt="Alt text" title="Meditron-logo">
# Model Card for Meditron-7B-v1.0
Meditron is a suite of open-source medical Large Language Models (LLMs).
Meditron-7B is a 7 billion parameters model adapted to the medical domain from Llama-2-7B through continued pretraining on a comprehensively curated medical corpus, including selected PubMed articles, abstracts, a [new dataset](https://huggingface.co/datasets/epfl-llm/guidelines) of internationally-recognized medical guidelines, and general domain data from [RedPajama-v1](https://huggingface.co/datasets/togethercomputer/RedPajama-Data-1T).
Meditron-7B, finetuned on relevant training data, outperforms Llama-2-7B and PMC-Llama on multiple medical reasoning tasks.
<details open>
<summary><strong>Advisory Notice</strong></summary>
<blockquote style="padding: 10px; margin: 0 0 10px; border-left: 5px solid #ddd;">
While Meditron is designed to encode medical knowledge from sources of high-quality evidence, it is not yet adapted to deliver this knowledge appropriately, safely, or within professional actionable constraints.
We recommend against deploying Meditron in medical applications without extensive use-case alignment, as well as additional testing, specifically including randomized controlled trials in real-world practice settings.
</blockquote>
</details>
## Model Details
- **Developed by:** [EPFL LLM Team](https://huggingface.co/epfl-llm)
- **Model type:** Causal decoder-only transformer language model
- **Language(s):** English (mainly)
- **Model License:** [LLAMA 2 COMMUNITY LICENSE AGREEMENT](https://huggingface.co/meta-llama/Llama-2-70b/raw/main/LICENSE.txt)
- **Code License:** [APACHE 2.0 LICENSE](LICENSE)
- **Continue-pretrained from model:** [Llama-2-7B](https://huggingface.co/meta-llama/Llama-2-7b)
- **Context length:** 2K tokens
- **Input:** Text-only data
- **Output:** Model generates text only
- **Status:** This is a static model trained on an offline dataset. Future versions of the tuned models will be released as we enhance model's performance.
- **Knowledge Cutoff:** August 2023
### Model Sources
- **Repository:** [epflLLM/meditron](https://github.com/epfLLM/meditron)
- **Trainer:** [epflLLM/Megatron-LLM](https://github.com/epfLLM/Megatron-LLM)
- **Paper:** *[MediTron-70B: Scaling Medical Pretraining for Large Language Models](https://arxiv.org/abs/2311.16079)*
## Uses
Meditron-7B is being made available for further testing and assessment as an AI assistant to enhance clinical decision-making and enhance access to an LLM for healthcare use. Potential use cases may include but are not limited to:
- Medical exam question answering
- Supporting differential diagnosis
- Disease information (symptoms, cause, treatment) query
- General health information query
### Direct Use
It is possible to use this model to generate text, which is useful for experimentation and understanding its capabilities.
It should not be used directly for production or work that may impact people.
### Downstream Use
Meditron-70B and Meditron-7B are both foundation models without finetuning or instruction-tuning. They can be finetuned, instruction-tuned, or RLHF-tuned for specific downstream tasks and applications.
There are two ways we have used this model for downstream question-answering tasks.
1. We apply in-context learning with k demonstrations (3 or 5 in our paper) added to the prompt.
2. We finetuned the models for downstream question-answering tasks using specific training sets.
We encourage and look forward to the adaption of the base model for more diverse applications.
If you want a more interactive way to prompt the model, we recommend using a high-throughput and memory-efficient inference engine with a UI that supports chat and text generation.
You can check out our deployment [guide](https://github.com/epfLLM/meditron/blob/main/deployment/README.md), where we used [FastChat](https://github.com/lm-sys/FastChat) with [vLLM](https://github.com/vllm-project/vllm). We collected generations for our qualitative analysis through an interactive UI platform, [BetterChatGPT](https://github.com/ztjhz/BetterChatGPT). Here is the prompt format we used as an example:
<img width=70% src="prompt_example.png" alt="qualitative-analysis-prompt" title="Qualitative Analysis Prompt">
### Out-of-Scope Use
We do not recommend using this model for natural language generation in a production environment, finetuned or otherwise.
## Truthfulness, Helpfulness, Risk, and Bias
<!-- This section is meant to convey both technical and sociotechnical limitations. -->
We did an initial assessment of Meditron models' **Truthfulness** against baseline models and consumer-level medical models.
We use TruthfulQA (multiple choice) as the main evaluation benchmark.
We only focus on the categories that are relevant to the medical domain, including Health, Nutrition, Psychology, and Science.
For 7B models, we perform one-shot evaluations for consistent answer generation.
For 70B models, the evaluations are under the zero-shot setting.
Below, we report the detailed truthfulness performance of each category.
| | | | | | | | |
| --- | ------ |----- |----- |----- |----- |----- |----- |
|Category | meditron-70b | llama-2-70b | med42-70b* | meditron-7b | llama-2-7b | PMC-llama-7b |
|Health | 81.8 | 69.1 | 83.6 | 27.3 | 16.4 | 3.6 |
|Nutrition | 77.9 | 68.8 | 62.5 | 31.1 | 12.5 | 6.3 |
|Psychology| 47.4 | 36.8 | 52.6 | 21.1 | 10.5 | 0.0 |
|Science | 77.8 | 44.4 | 33.3 | 33.3 | 11.1 | 0.0 |
|Avg | 71.2 | 54.8 | 58.0 | 28.3 | 12.6 | 2.5 |
| | | | | | | |
For a more detailed performance analysis, please see our paper.
Significant research is still required to fully explore potential bias, fairness, and safety issues with this language model.
Please recognize that our evaluation on Meditron-7B's helpfulness, risk, and bias are highly limited.
Thus, as we noted in the safety notice, we strongly against any deployment in medical applications without further alignment process and rigorous evaluation!
### Recommendations
**IMPORTANT!**
Users (both direct and downstream) should be made aware of the risks, biases, and limitations of the model.
While this model is capable of generating natural language text, we have only begun to explore this capability and its limitations.
Understanding these limitations is especially important in a domain like medicine.
Therefore, we strongly recommend against using this model in production for natural language generation or for professional purposes related to health and medicine.
## Training Details
### Training Data
Meditron’s domain-adaptive pre-training corpus GAP-Replay combines 48.1B tokens from four corpora:
- [**Clinical Guidelines**](https://huggingface.co/datasets/epfl-llm/guidelines): a new dataset of 46K internationally-recognized clinical practice guidelines from various healthcare-related sources, including hospitals and international organizations.
- **Medical Paper Abstracts**: 16.1M abstracts extracted from closed-access PubMed and PubMed Central papers.
- **Medical Papers**: full-text articles extracted from 5M publicly available PubMed and PubMed Central papers.
- **Replay Data**: 400M tokens of general domain pretraining data sampled from [RedPajama-v1](https://huggingface.co/datasets/togethercomputer/RedPajama-Data-1T)
<img width=75% src="gap-replay.png" alt="Alt text" title="Meditron-logo">
#### Data Preprocessing
Please see the detailed preprocessing procedure in our paper.
### Training Procedure
We used the [Megatron-LLM](https://github.com/epfLLM/Megatron-LLM) distributed training library, a derivative of Nvidia's Megatron LM project, to optimize training efficiency.
Hardware consists of 1 node of 8x NVIDIA A100 (80GB) SXM GPUs connected by NVLink and NVSwitch with a single Nvidia ConnectX-6 DX network card and equipped with 2 x AMD EPYC 7543 32-Core Processors and 512 GB of RAM.
Our three way parallelism scheme uses:
- Data Parallelism (DP -- different GPUs process different subsets of the batches) of 2,
- Pipeline Parallelism (PP -- different GPUs process different layers) of 4,
- Tensor Parallelism (TP -- different GPUs process different subtensors for matrix multiplication) of 1.
#### Training Hyperparameters
| | |
| --- | ------ |
| bf16 | true |
| lr | 3e-4 |
| eps | 1e-5 |
| betas | \[0.9, 0.95\] |
| clip_grad | 1 |
| weight decay | 0.1 |
| DP size | 16 |
| TP size | 4 |
| PP size | 1 |
| seq length | 2048 |
| lr scheduler | cosine|
| min lr | 1e-6 |
| warmup iteration | 2000 |
| micro batch size | 10 |
| global batch size | 1600 |
| | |
#### Sizes
The model was trained in September 2023.
The model architecture is exactly Llama 2, meaning
| | |
| --- | ------ |
| Model size | 7B |
| Hidden dimension | 4096 |
| Num. attention heads | 32 |
| Num. layers | 32 |
| | |
## Evaluation
<!-- This section describes the evaluation protocols and provides the results. -->
### Testing Data & Metrics
#### Testing Data
- [MedQA (USMLE)](https://huggingface.co/datasets/bigbio/med_qa)
- [MedMCQA](https://huggingface.co/datasets/medmcqa)
- [PubMedQA](https://huggingface.co/datasets/bigbio/pubmed_qa)
- [MMLU-Medical](https://huggingface.co/datasets/lukaemon/mmlu)
- [MedQA-4-Option](https://huggingface.co/datasets/GBaker/MedQA-USMLE-4-options)
#### Metrics
- Accuracy: suite the evaluation of multiple-choice question-answering tasks.
### Results
We finetune meditron-7b, llama-2-7b, pmc-llama-7b on each benchmark (pubmedqa, medmcqa, medqa)'s training data individually.
We report the finetuned models' performance with top token selection as the inference mode.
For MMLU-Medical, models finetuned on MedMCQA are used for inference.
For MedQA-4-Option, models finetuned on MedQA are used for inference.
For a more detailed performance analysis, please see our paper.
| | | | | | |
| --- | ------ |----- |----- |----- |----- |
|Dataset | meditron-7b | llama-2-7b | pmc-llama-7b | Zephyr-7B-beta* | Mistral-7B-instruct* |
|MMLU-Medical | 54.2 | 53.7 | 56.4 | 63.3 | 60.0 |
|PubMedQA | 74.4 | 61.8 | 59.2 | 46.0 | 17.8 |
|MedMCQA | 59.2 | 54.4 | 57.6 | 43.0 | 40.2 |
|MedQA | 47.9 | 44.0 | 42.4 | 42.8 | 32.4 |
|MedQA-4-Option| 52.0 | 49.6 | 49.2 | 48.5 | 41.1 |
|Avg | 57.5 | 52.7 | 53.0 | 48.7 | 38.3 |
| | | | | | |
**Note**: models with * are already instruction-tuned, so we exclude them from further finetuning on any training data.
## Environmental Impact
<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
- **Hardware Type:** 8 x NVIDIA A100 (80GB) SXM
- **Total GPU hours:** 588.8
- **Hardware Provider:** EPFL Research Computing Platform
- **Compute Region:** Switzerland
- **Carbon Emitted:** Switzerland has a carbon efficiency of 0.016 kgCO2/kWh (https://www.carbonfootprint.com/docs/2018_8_electricity_factors_august_2018_-_online_sources.pdf). 73.6 hours of 8 A100s means 588.8 hours at a TDP of 400W. Assuming a Power Usage effectiveness of 1.5, total emissions are estimated to be:
(400W / 1000W/kWh / GPU * 0.016 kgCO2/kWh * 73.6 h * 8 GPU) * 1.8 PUE = 6.8 kgCO2.
## Citation
**BibTeX:**
If you use Meditron or its training data, please cite our work:
```
@misc{chen2023meditron70b,
title={MEDITRON-70B: Scaling Medical Pretraining for Large Language Models},
author={Zeming Chen and Alejandro Hernández-Cano and Angelika Romanou and Antoine Bonnet and Kyle Matoba and Francesco Salvi and Matteo Pagliardini and Simin Fan and Andreas Köpf and Amirkeivan Mohtashami and Alexandre Sallinen and Alireza Sakhaeirad and Vinitra Swamy and Igor Krawczuk and Deniz Bayazit and Axel Marmet and Syrielle Montariol and Mary-Anne Hartley and Martin Jaggi and Antoine Bosselut},
year={2023},
eprint={2311.16079},
archivePrefix={arXiv},
primaryClass={cs.CL}
}
@software{epfmedtrn,
author = {Zeming Chen and Alejandro Hernández-Cano and Angelika Romanou and Antoine Bonnet and Kyle Matoba and Francesco Salvi and Matteo Pagliardini and Simin Fan and Andreas Köpf and Amirkeivan Mohtashami and Alexandre Sallinen and Alireza Sakhaeirad and Vinitra Swamy and Igor Krawczuk and Deniz Bayazit and Axel Marmet and Syrielle Montariol and Mary-Anne Hartley and Martin Jaggi and Antoine Bosselut},
title = {MediTron-70B: Scaling Medical Pretraining for Large Language Models},
month = November,
year = 2023,
url = {https://github.com/epfLLM/meditron}
}
```
| {"base_model": "meta-llama/Llama-2-7b", "datasets": ["epfl-llm/guidelines"], "language": ["en"], "license": "llama2", "metrics": ["accuracy", "perplexity"]} |
m42-health/Llama3-Med42-8B | m42-health | text-generation | [
"transformers",
"safetensors",
"llama",
"text-generation",
"m42",
"health",
"healthcare",
"clinical-llm",
"conversational",
"en",
"arxiv:2408.06142",
"license:llama3",
"autotrain_compatible",
"text-generation-inference",
"region:us"
] | 2024-07-02T10:14:40 | 2024-08-20T05:12:05 | 1,966 | 62 | ---
language:
- en
license: llama3
license_name: llama3
pipeline_tag: text-generation
tags:
- m42
- health
- healthcare
- clinical-llm
inference: false
---
# **Med42-v2 - A Suite of Clinically-aligned Large Language Models**
Med42-v2 is a suite of open-access clinical large language models (LLM) instruct and preference-tuned by M42 to expand access to medical knowledge. Built off LLaMA-3 and comprising either 8 or 70 billion parameters, these generative AI systems provide high-quality answers to medical questions.
## Key performance metrics:
- Med42-v2-70B outperforms GPT-4.0 in most of the MCQA tasks.
- Med42-v2-70B achieves a MedQA zero-shot performance of 79.10, surpassing the prior state-of-the-art among all openly available medical LLMs.
- Med42-v2-70B sits at the top of the Clinical Elo Rating Leaderboard.
|Models|Elo Score|
|:---:|:---:|
|**Med42-v2-70B**| 1764 |
|Llama3-70B-Instruct| 1643 |
|GPT4-o| 1426 |
|Llama3-8B-Instruct| 1352 |
|Mixtral-8x7b-Instruct| 970 |
|**Med42-v2-8B**| 924 |
|OpenBioLLM-70B| 657 |
|JSL-MedLlama-3-8B-v2.0| 447 |
## Limitations & Safe Use
- The Med42-v2 suite of models is not ready for real clinical use. Extensive human evaluation is undergoing as it is required to ensure safety.
- Potential for generating incorrect or harmful information.
- Risk of perpetuating biases in training data.
Use this suite of models responsibly! Do not rely on them for medical usage without rigorous safety testing.
## Model Details
*Disclaimer: This large language model is not yet ready for clinical use without further testing and validation. It should not be relied upon for making medical decisions or providing patient care.*
Beginning with Llama3 models, Med42-v2 were instruction-tuned using a dataset of ~1B tokens compiled from different open-access and high-quality sources, including medical flashcards, exam questions, and open-domain dialogues.
**Model Developers:** M42 Health AI Team
**Finetuned from model:** Llama3 - 8B & 70B Instruct
**Context length:** 8k tokens
**Input:** Text only data
**Output:** Model generates text only
**Status:** This is a static model trained on an offline dataset. Future versions of the tuned models will be released as we enhance the model's performance.
**License:** Llama 3 Community License Agreement
**Research Paper:** [Med42-v2: A Suite of Clinical LLMs](https://huggingface.co/papers/2408.06142)
## Intended Use
The Med42-v2 suite of models is being made available for further testing and assessment as AI assistants to enhance clinical decision-making and access to LLMs for healthcare use. Potential use cases include:
- Medical question answering
- Patient record summarization
- Aiding medical diagnosis
- General health Q&A
**Run the model**
You can use the 🤗 Transformers library `text-generation` pipeline to do inference.
```python
import transformers
import torch
model_name_or_path = "m42-health/Llama3-Med42-8B"
pipeline = transformers.pipeline(
"text-generation",
model=model_name_or_path,
torch_dtype=torch.bfloat16,
device_map="auto",
)
messages = [
{
"role": "system",
"content": (
"You are a helpful, respectful and honest medical assistant. You are a second version of Med42 developed by the AI team at M42, UAE. "
"Always answer as helpfully as possible, while being safe. "
"Your answers should not include any harmful, unethical, racist, sexist, toxic, dangerous, or illegal content. "
"Please ensure that your responses are socially unbiased and positive in nature. If a question does not make any sense, or is not factually coherent, explain why instead of answering something not correct. "
"If you don’t know the answer to a question, please don’t share false information."
),
},
{"role": "user", "content": "What are the symptoms of diabetes?"},
]
prompt = pipeline.tokenizer.apply_chat_template(
messages, tokenize=False, add_generation_prompt=False
)
stop_tokens = [
pipeline.tokenizer.eos_token_id,
pipeline.tokenizer.convert_tokens_to_ids("<|eot_id|>"),
]
outputs = pipeline(
prompt,
max_new_tokens=512,
eos_token_id=stop_tokens,
do_sample=True,
temperature=0.4,
top_k=150,
top_p=0.75,
)
print(outputs[0]["generated_text"][len(prompt) :])
```
## Hardware and Software
The training was conducted on the NVIDIA DGX cluster with H100 GPUs, utilizing PyTorch's Fully Sharded Data Parallel (FSDP) framework.
## Evaluation Results
### Open-ended question generation
To ensure a robust evaluation of our model's output quality, we employ the LLM-as-a-Judge approach using Prometheus-8x7b-v2.0. Our assessment uses 4,000 carefully curated publicly accessible healthcare-related questions, generating responses from various models. We then use Prometheus to conduct pairwise comparisons of the answers. Drawing inspiration from the LMSYS Chatbot-Arena methodology, we present the results as Elo ratings for each model.
To maintain fairness and eliminate potential bias from prompt engineering, we used the same simple system prompt for every model throughout the evaluation process.
Below is the scoring rubric we used to prompt Prometheus to select the best answer:
```
### Score Rubric:
Which response is of higher overall quality in a medical context? Consider:
* Relevance: Does it directly address the question?
* Completeness: Does it cover all important aspects, details and subpoints?
* Safety: Does it avoid unsafe practices and address potential risks?
* Ethics: Does it maintain confidentiality and avoid biases?
* Clarity: Is it professional, clear and easy to understand?
```
#### Elo Ratings
|Models|Elo Score|
|:---:|:---:|
|**Med42-v2-70B**| 1764 |
|Llama3-70B-Instruct| 1643 |
|GPT4-o| 1426 |
|Llama3-8B-Instruct| 1352 |
|Mixtral-8x7b-Instruct| 970 |
|**Med42-v2-8B**| 924 |
|OpenBioLLM-70B| 657 |
|JSL-MedLlama-3-8B-v2.0| 447 |
#### Win-rate
### MCQA Evaluation
Med42-v2 improves performance on every clinical benchmark compared to our previous version, including MedQA, MedMCQA, USMLE, MMLU clinical topics and MMLU Pro clinical subset. For all evaluations reported so far, we use [EleutherAI's evaluation harness library](https://github.com/EleutherAI/lm-evaluation-harness) and report zero-shot accuracies (except otherwise stated). We integrated chat templates into harness and computed the likelihood for the full answer instead of only the tokens "a.", "b.", "c." or "d.".
|Model|MMLU Pro|MMLU|MedMCQA|MedQA|USMLE|
|---:|:---:|:---:|:---:|:---:|:---:|
|**Med42v2-70B**|64.36|87.12|73.20|79.10|83.80|
|**Med42v2-8B**|54.30|75.76|61.34|62.84|67.04|
|OpenBioLLM-70B|64.24|90.40|73.18|76.90|79.01|
|GPT-4.0<sup>†</sup>|-|87.00|69.50|78.90|84.05|
|MedGemini*|-|-|-|84.00|-|
|Med-PaLM-2 (5-shot)*|-|87.77|71.30|79.70|-|
|Med42|-|76.72|60.90|61.50|71.85|
|ClinicalCamel-70B|-|69.75|47.00|53.40|54.30|
|GPT-3.5<sup>†</sup>|-|66.63|50.10|50.80|53.00|
|Llama3-8B-Instruct|48.24|72.89|59.65|61.64|60.38|
|Llama3-70B-Instruct|64.24|85.99|72.03|78.88|83.57|
**For MedGemini, results are reported for MedQA without self-training and without search. We note that 0-shot performance is not reported for Med-PaLM 2. Further details can be found at [https://github.com/m42health/med42](https://github.com/m42health/med42)*.
<sup>†</sup> *Results as reported in the paper [Capabilities of GPT-4 on Medical Challenge Problems](https://www.microsoft.com/en-us/research/uploads/prod/2023/03/GPT-4_medical_benchmarks.pdf)*.
## Accessing Med42 and Reporting Issues
Please report any software "bug" or other problems through one of the following means:
- Reporting issues with the model: [https://github.com/m42health/med42](https://github.com/m42health/med42)
- Reporting risky content generated by the model, bugs and/or any security concerns: [https://forms.office.com/r/fPY4Ksecgf](https://forms.office.com/r/fPY4Ksecgf)
- M42’s privacy policy available at [https://m42.ae/privacy-policy/](https://m42.ae/privacy-policy/)
- Reporting violations of the Acceptable Use Policy or unlicensed uses of Med42: <[email protected]>
## Acknowledgements
We thank the Torch FSDP team for their robust distributed training framework, the EleutherAI harness team for their valuable evaluation tools, and the Hugging Face Alignment team for their contributions to responsible AI development.
## Citation
```
@misc{med42v2,
Author = {Cl{\'e}ment Christophe and Praveen K Kanithi and Tathagata Raha and Shadab Khan and Marco AF Pimentel},
Title = {Med42-v2: A Suite of Clinical LLMs},
Year = {2024},
Eprint = {arXiv:2408.06142},
url={https://arxiv.org/abs/2408.06142},
}
```
| [
"QUESTION_ANSWERING",
"SUMMARIZATION"
] | [
"MEDQA"
] | BioNLP | # **Med42-v2 - A Suite of Clinically-aligned Large Language Models**
Med42-v2 is a suite of open-access clinical large language models (LLM) instruct and preference-tuned by M42 to expand access to medical knowledge. Built off LLaMA-3 and comprising either 8 or 70 billion parameters, these generative AI systems provide high-quality answers to medical questions.
## Key performance metrics:
- Med42-v2-70B outperforms GPT-4.0 in most of the MCQA tasks.
- Med42-v2-70B achieves a MedQA zero-shot performance of 79.10, surpassing the prior state-of-the-art among all openly available medical LLMs.
- Med42-v2-70B sits at the top of the Clinical Elo Rating Leaderboard.
|Models|Elo Score|
|:---:|:---:|
|**Med42-v2-70B**| 1764 |
|Llama3-70B-Instruct| 1643 |
|GPT4-o| 1426 |
|Llama3-8B-Instruct| 1352 |
|Mixtral-8x7b-Instruct| 970 |
|**Med42-v2-8B**| 924 |
|OpenBioLLM-70B| 657 |
|JSL-MedLlama-3-8B-v2.0| 447 |
## Limitations & Safe Use
- The Med42-v2 suite of models is not ready for real clinical use. Extensive human evaluation is undergoing as it is required to ensure safety.
- Potential for generating incorrect or harmful information.
- Risk of perpetuating biases in training data.
Use this suite of models responsibly! Do not rely on them for medical usage without rigorous safety testing.
## Model Details
*Disclaimer: This large language model is not yet ready for clinical use without further testing and validation. It should not be relied upon for making medical decisions or providing patient care.*
Beginning with Llama3 models, Med42-v2 were instruction-tuned using a dataset of ~1B tokens compiled from different open-access and high-quality sources, including medical flashcards, exam questions, and open-domain dialogues.
**Model Developers:** M42 Health AI Team
**Finetuned from model:** Llama3 - 8B & 70B Instruct
**Context length:** 8k tokens
**Input:** Text only data
**Output:** Model generates text only
**Status:** This is a static model trained on an offline dataset. Future versions of the tuned models will be released as we enhance the model's performance.
**License:** Llama 3 Community License Agreement
**Research Paper:** [Med42-v2: A Suite of Clinical LLMs](https://huggingface.co/papers/2408.06142)
## Intended Use
The Med42-v2 suite of models is being made available for further testing and assessment as AI assistants to enhance clinical decision-making and access to LLMs for healthcare use. Potential use cases include:
- Medical question answering
- Patient record summarization
- Aiding medical diagnosis
- General health Q&A
**Run the model**
You can use the 🤗 Transformers library `text-generation` pipeline to do inference.
```python
import transformers
import torch
model_name_or_path = "m42-health/Llama3-Med42-8B"
pipeline = transformers.pipeline(
"text-generation",
model=model_name_or_path,
torch_dtype=torch.bfloat16,
device_map="auto",
)
messages = [
{
"role": "system",
"content": (
"You are a helpful, respectful and honest medical assistant. You are a second version of Med42 developed by the AI team at M42, UAE. "
"Always answer as helpfully as possible, while being safe. "
"Your answers should not include any harmful, unethical, racist, sexist, toxic, dangerous, or illegal content. "
"Please ensure that your responses are socially unbiased and positive in nature. If a question does not make any sense, or is not factually coherent, explain why instead of answering something not correct. "
"If you don’t know the answer to a question, please don’t share false information."
),
},
{"role": "user", "content": "What are the symptoms of diabetes?"},
]
prompt = pipeline.tokenizer.apply_chat_template(
messages, tokenize=False, add_generation_prompt=False
)
stop_tokens = [
pipeline.tokenizer.eos_token_id,
pipeline.tokenizer.convert_tokens_to_ids("<|eot_id|>"),
]
outputs = pipeline(
prompt,
max_new_tokens=512,
eos_token_id=stop_tokens,
do_sample=True,
temperature=0.4,
top_k=150,
top_p=0.75,
)
print(outputs[0]["generated_text"][len(prompt) :])
```
## Hardware and Software
The training was conducted on the NVIDIA DGX cluster with H100 GPUs, utilizing PyTorch's Fully Sharded Data Parallel (FSDP) framework.
## Evaluation Results
### Open-ended question generation
To ensure a robust evaluation of our model's output quality, we employ the LLM-as-a-Judge approach using Prometheus-8x7b-v2.0. Our assessment uses 4,000 carefully curated publicly accessible healthcare-related questions, generating responses from various models. We then use Prometheus to conduct pairwise comparisons of the answers. Drawing inspiration from the LMSYS Chatbot-Arena methodology, we present the results as Elo ratings for each model.
To maintain fairness and eliminate potential bias from prompt engineering, we used the same simple system prompt for every model throughout the evaluation process.
Below is the scoring rubric we used to prompt Prometheus to select the best answer:
```
### Score Rubric:
Which response is of higher overall quality in a medical context? Consider:
* Relevance: Does it directly address the question?
* Completeness: Does it cover all important aspects, details and subpoints?
* Safety: Does it avoid unsafe practices and address potential risks?
* Ethics: Does it maintain confidentiality and avoid biases?
* Clarity: Is it professional, clear and easy to understand?
```
#### Elo Ratings
|Models|Elo Score|
|:---:|:---:|
|**Med42-v2-70B**| 1764 |
|Llama3-70B-Instruct| 1643 |
|GPT4-o| 1426 |
|Llama3-8B-Instruct| 1352 |
|Mixtral-8x7b-Instruct| 970 |
|**Med42-v2-8B**| 924 |
|OpenBioLLM-70B| 657 |
|JSL-MedLlama-3-8B-v2.0| 447 |
#### Win-rate
### MCQA Evaluation
Med42-v2 improves performance on every clinical benchmark compared to our previous version, including MedQA, MedMCQA, USMLE, MMLU clinical topics and MMLU Pro clinical subset. For all evaluations reported so far, we use [EleutherAI's evaluation harness library](https://github.com/EleutherAI/lm-evaluation-harness) and report zero-shot accuracies (except otherwise stated). We integrated chat templates into harness and computed the likelihood for the full answer instead of only the tokens "a.", "b.", "c." or "d.".
|Model|MMLU Pro|MMLU|MedMCQA|MedQA|USMLE|
|---:|:---:|:---:|:---:|:---:|:---:|
|**Med42v2-70B**|64.36|87.12|73.20|79.10|83.80|
|**Med42v2-8B**|54.30|75.76|61.34|62.84|67.04|
|OpenBioLLM-70B|64.24|90.40|73.18|76.90|79.01|
|GPT-4.0<sup>†</sup>|-|87.00|69.50|78.90|84.05|
|MedGemini*|-|-|-|84.00|-|
|Med-PaLM-2 (5-shot)*|-|87.77|71.30|79.70|-|
|Med42|-|76.72|60.90|61.50|71.85|
|ClinicalCamel-70B|-|69.75|47.00|53.40|54.30|
|GPT-3.5<sup>†</sup>|-|66.63|50.10|50.80|53.00|
|Llama3-8B-Instruct|48.24|72.89|59.65|61.64|60.38|
|Llama3-70B-Instruct|64.24|85.99|72.03|78.88|83.57|
**For MedGemini, results are reported for MedQA without self-training and without search. We note that 0-shot performance is not reported for Med-PaLM 2. Further details can be found at [https://github.com/m42health/med42](https://github.com/m42health/med42)*.
<sup>†</sup> *Results as reported in the paper [Capabilities of GPT-4 on Medical Challenge Problems](https://www.microsoft.com/en-us/research/uploads/prod/2023/03/GPT-4_medical_benchmarks.pdf)*.
## Accessing Med42 and Reporting Issues
Please report any software "bug" or other problems through one of the following means:
- Reporting issues with the model: [https://github.com/m42health/med42](https://github.com/m42health/med42)
- Reporting risky content generated by the model, bugs and/or any security concerns: [https://forms.office.com/r/fPY4Ksecgf](https://forms.office.com/r/fPY4Ksecgf)
- M42’s privacy policy available at [https://m42.ae/privacy-policy/](https://m42.ae/privacy-policy/)
- Reporting violations of the Acceptable Use Policy or unlicensed uses of Med42: <[email protected]>
## Acknowledgements
We thank the Torch FSDP team for their robust distributed training framework, the EleutherAI harness team for their valuable evaluation tools, and the Hugging Face Alignment team for their contributions to responsible AI development.
## Citation
```
@misc{med42v2,
Author = {Cl{\'e}ment Christophe and Praveen K Kanithi and Tathagata Raha and Shadab Khan and Marco AF Pimentel},
Title = {Med42-v2: A Suite of Clinical LLMs},
Year = {2024},
Eprint = {arXiv:2408.06142},
url={https://arxiv.org/abs/2408.06142},
}
```
| {"language": ["en"], "license": "llama3", "license_name": "llama3", "pipeline_tag": "text-generation", "tags": ["m42", "health", "healthcare", "clinical-llm"], "inference": false} |
seongil-dn/bge-m3-756 | seongil-dn | sentence-similarity | ["sentence-transformers","safetensors","xlm-roberta","sentence-similarity","feature-extraction","gen(...TRUNCATED) | 2025-03-07T10:43:53 | 2025-03-07T10:48:06 | 12 | 0 | "---\nbase_model: seongil-dn/unsupervised_20m_3800\nlibrary_name: sentence-transformers\npipeline_ta(...TRUNCATED) | [
"TEXT_CLASSIFICATION",
"TRANSLATION"
] | [
"CRAFT"
] | Non_BioNLP | "\n# SentenceTransformer based on seongil-dn/unsupervised_20m_3800\n\nThis is a [sentence-transforme(...TRUNCATED) | "{\"base_model\": \"seongil-dn/unsupervised_20m_3800\", \"library_name\": \"sentence-transformers\",(...TRUNCATED) |
LoneStriker/OpenBioLLM-Llama3-8B-GGUF | LoneStriker | null | ["gguf","llama-3","llama","Mixtral","instruct","finetune","chatml","DPO","RLHF","gpt4","distillation(...TRUNCATED) | 2024-04-26T19:11:19 | 2024-04-26T19:23:42 | 30 | 1 | "---\nbase_model: meta-llama/Meta-Llama-3-8B\nlanguage:\n- en\nlicense: llama3\ntags:\n- llama-3\n- (...TRUNCATED) | [
"QUESTION_ANSWERING"
] | [
"MEDQA",
"PUBMEDQA"
] | BioNLP | "\n\n<div align=\"center\">\n<img width=\"260px\" src=\"https://cdn-uploads.huggingface.co/productio(...TRUNCATED) | "{\"base_model\": \"meta-llama/Meta-Llama-3-8B\", \"language\": [\"en\"], \"license\": \"llama3\", \(...TRUNCATED) |
medspaner/mdeberta-v3-base-es-trials-misc-ents | medspaner | token-classification | ["transformers","pytorch","deberta-v2","token-classification","generated_from_trainer","arxiv:2111.0(...TRUNCATED) | 2024-01-13T12:07:27 | 2024-10-01T06:30:33 | 12 | 0 | "---\nlicense: cc-by-nc-4.0\nmetrics:\n- precision\n- recall\n- f1\n- accuracy\ntags:\n- generated_f(...TRUNCATED) | [
"NAMED_ENTITY_RECOGNITION"
] | [
"SCIELO"
] | BioNLP | "\n<!-- This model card has been generated automatically according to the information the Trainer ha(...TRUNCATED) | "{\"license\": \"cc-by-nc-4.0\", \"metrics\": [\"precision\", \"recall\", \"f1\", \"accuracy\"], \"t(...TRUNCATED) |
carsondial/slinger20241231-3 | carsondial | sentence-similarity | ["sentence-transformers","safetensors","bert","sentence-similarity","feature-extraction","generated_(...TRUNCATED) | 2025-01-01T15:30:35 | 2025-01-01T15:31:07 | 6 | 0 | "---\nbase_model: BAAI/bge-base-en-v1.5\nlanguage:\n- en\nlibrary_name: sentence-transformers\nlicen(...TRUNCATED) | [
"TEXT_CLASSIFICATION"
] | [
"CRAFT"
] | Non_BioNLP | "\n# slinger-base\n\nThis is a [sentence-transformers](https://www.SBERT.net) model finetuned from [(...TRUNCATED) | "{\"base_model\": \"BAAI/bge-base-en-v1.5\", \"language\": [\"en\"], \"library_name\": \"sentence-tr(...TRUNCATED) |
StivenLancheros/Roberta-base-biomedical-clinical-es-finetuned-ner-CRAFT_en_es | StivenLancheros | token-classification | ["transformers","pytorch","tensorboard","roberta","token-classification","generated_from_trainer","l(...TRUNCATED) | 2022-03-11T19:08:07 | 2022-03-12T11:39:55 | 115 | 0 | "---\nlicense: apache-2.0\nmetrics:\n- precision\n- recall\n- f1\n- accuracy\ntags:\n- generated_fro(...TRUNCATED) | [
"NAMED_ENTITY_RECOGNITION"
] | [
"CRAFT"
] | BioNLP | "\n<!-- This model card has been generated automatically according to the information the Trainer ha(...TRUNCATED) | "{\"license\": \"apache-2.0\", \"metrics\": [\"precision\", \"recall\", \"f1\", \"accuracy\"], \"tag(...TRUNCATED) |
bobox/DeBERTa-small-ST-v1-test-step2 | bobox | sentence-similarity | ["sentence-transformers","pytorch","deberta-v2","sentence-similarity","feature-extraction","generate(...TRUNCATED) | 2024-08-21T19:22:52 | 2024-08-21T19:23:13 | 7 | 0 | "---\nbase_model: bobox/DeBERTa-small-ST-v1-test\ndatasets:\n- jinaai/negation-dataset-v2\n- tals/vi(...TRUNCATED) | [
"TEXT_CLASSIFICATION",
"SEMANTIC_SIMILARITY"
] | [
"MEDAL",
"SCIQ",
"SCITAIL"
] | Non_BioNLP | "\n# SentenceTransformer based on bobox/DeBERTa-small-ST-v1-test\n\nThis is a [sentence-transformers(...TRUNCATED) | "{\"base_model\": \"bobox/DeBERTa-small-ST-v1-test\", \"datasets\": [\"jinaai/negation-dataset-v2\",(...TRUNCATED) |
End of preview. Expand
in Data Studio
README.md exists but content is empty.
- Downloads last month
- 69