Qwen3 is the latest generation of large language models in Qwen series, offering a comprehensive suite of dense and mixture-of-experts (MoE) models. Built upon extensive training, Qwen3 delivers groundbreaking advancements in reasoning, instruction-following, agent capabilities, and multilingual support, with the following key features: - Uniquely support of seamless switching between thinking mode (for complex logical reasoning, math, and coding) and non-thinking mode (for efficient, general-purpose dialogue) within single model, ensuring optimal performance across various scenarios. - Significantly enhancement in its reasoning capabilities, surpassing previous QwQ (in thinking mode) and…
![[HYU_NLP] EVA Team logo](/media/mh-logo-hyu-nlp-eval-558e075b88.webp){kind=logo}
Open-weight model · Text generation
qwen3-4b-rar-medicine-static-r0-matched-seed11-step-000
by [HYU_NLP] EVA Team HYU-NLP-EVAL/qwen3-4b-rar-medicine-static-r0-matched-seed11-step-000
qwen3-4b-rar-medicine-static-r0-matched-seed11-step-000 is an open-weight model for text generation from [HYU_NLP] EVA Team, released under Apache License 2.0. It has 4B parameters and a 262,144-token context. At 16-bit it needs about 9.7 GB of GPU memory, which fits on 1x MI300X from $1.85 an hour, at the lowest prices in the SAVRN Index.
This is the policy after 0 global optimizer updates of the matched separate from the OnlineRubrics/dynamic-rubric checkpoints. The root files are a BF16 Transformers export for inference.
Parameters4B
Context262,144
Weights25.7 GB
Licenseapache-2.0
AccessOpen weights
Monthly Downloads—
Runs On
What it takes to serve qwen3-4b-rar-medicine-static-r0-matched-seed11-step-000 (4B parameters): the memory its weights need at each precision, and the cheapest way to rent enough data-center GPUs to hold them.
| Precision | Weights | Memory needed | Cheapest setup | Per hour | Also fits |
|---|---|---|---|---|---|
| 16-bit | 8.0 GB | 9.7 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
| 8-bit | 4.0 GB | 4.8 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
| 4-bit | 2.0 GB | 2.4 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
Memory is the weights at that precision plus 20% for the runtime and a short context; a long context needs more. Prices are the lowest on-demand hourly rates in the SAVRN Index, read Oct 7, 2026.
Model Card
By [HYU_NLP] EVA Team, published under apache-2.0, revision 8d05ed0a432b.
This is the policy after 0 global optimizer updates of the matched separate from the OnlineRubrics/dynamic-rubric checkpoints. The root files are a BF16 Transformers export for inference. The originalcheckpoint/ directory contains the exact original veRL/FSDP policy parameter checkpoint and its tokenizer/configuration files. Optimizer, trainer, and data-loader state are intentionally not published; the complete resume checkpoint remains on Daisy. This is an intermediate research checkpoint, not a clinical model. No medical capability or safety claim is made. Original actor parameter SHA256: f81409edc253a52ee9b3e6807bf280cf1ff242c77c2f6645b087c74e03a4e3d4
Read [HYU_NLP] EVA Team's full model card
Static-R0 Matched GRPO on RaR-Medicine — step 0
This is the policy after 0 global optimizer updates of the matched static-rubric GRPO run (planned total: 48). It is intentionally separate from the OnlineRubrics/dynamic-rubric checkpoints.
Experiment identity
- Method:
static_r0_matched - Reward source:
rar_static_r0_only - Domain: Medicine
- Training data: RaR-Medicine, 1,500 prompts
- Seed: 11
- Policy:
Qwen/Qwen3-4B-Instruct-2507 - Base revision:
cdbee75f17c01a7cc42f958dc650907174af0554 - Thinking: disabled
- GRPO global prompt batch: 96
- Rollouts per prompt: 16
- Learning rate: 5e-06
The root files are a BF16 Transformers export for inference. The
original_checkpoint/ directory contains the exact original veRL/FSDP policy
parameter checkpoint and its tokenizer/configuration files. Optimizer,
trainer, and data-loader state are intentionally not published; the complete
resume checkpoint remains on Daisy.
from transformers import AutoModelForCausalLM, AutoTokenizer
repo_id = "HYU-NLP-EVAL/qwen3-4b-rar-medicine-static-r0-matched-seed11-step-000"
tokenizer = AutoTokenizer.from_pretrained(repo_id)
model = AutoModelForCausalLM.from_pretrained(
repo_id, torch_dtype="bfloat16", device_map="auto"
)
This is an intermediate research checkpoint, not a clinical model. No medical capability or safety claim is made.
Original actor parameter SHA256: f81409edc253a52ee9b3e6807bf280cf1ff242c77c2f6645b087c74e03a4e3d4
Configuration
- Architecture
- Qwen3ForCausalLM
- Context length (tokens)
- 262,144
- Layers
- 36
- Hidden size
- 2,560
- Feed-forward size
- 9,728
- Attention heads
- 32
- Key/value heads
- 8
- Head dimension
- 128
- Vocabulary size
- 151,936
- Model type
- qwen3
Identity and Version
- Repository
- HYU-NLP-EVAL/qwen3-4b-rar-medicine-static-r0-matched-seed11-step-000
- Publisher
- [HYU_NLP] EVA Team
- Task
- Text generation
- Modality
- Text
- Library
- transformers
- Parameters
- 4B parameters
- Languages
- rar-medicine
- Revision
- 8d05ed0a432bc8e5ca47ed4fbd3cd483d7628ab1
- First published
- 2026-09-30
- Last updated
- 2026-09-30
Files and Weights
20 files, 25.7 GB in total. The weights are 4 files totalling 25.7 GB in pt, safetensors.
Weights4 files · 25.7 GB
Configuration7 files · 40.9 KB
Tokenizer4 files · 22.8 MB
Documentation2 files · 13.0 KB
Other2 files · 5.3 KB
Repository1 file · 1.7 KB
Every file
| File | Type | Size | SHA-256 |
|---|---|---|---|
| model-00001-of-00003.safetensors | Weights | 4.0 GB | 75311d91bb08 |
| model-00002-of-00003.safetensors | Weights | 4.0 GB | 0b48adbb1f60 |
| model-00003-of-00003.safetensors | Weights | 99.6 MB | 7dd39ccca5e4 |
| original_checkpoint/actor/model_world_size_1_rank_0.pt | Weights | 17.6 GB | f81409edc253 |
| archive_manifest.json | Configuration | 4.2 KB | — |
| config.json | Configuration | 1.6 KB | — |
| generation_config.json | Configuration | 213 B | — |
| model.safetensors.index.json | Configuration | 32.9 KB | — |
| original_checkpoint/actor/fsdp_config.json | Configuration | 46 B | — |
| original_checkpoint/actor/huggingface/config.json | Configuration | 1.6 KB | — |
| original_checkpoint/actor/huggingface/generation_config.json | Configuration | 213 B | — |
| LICENSE | Documentation | 11.3 KB | — |
| README.md | Documentation | 1.7 KB | — |
| chat_template.jinja | Other | 2.6 KB | — |
| original_checkpoint/actor/huggingface/chat_template.jinja | Other | 2.6 KB | — |
| .gitattributes | Repository | 1.7 KB | — |
| original_checkpoint/actor/huggingface/tokenizer.json | Tokenizer | 11.4 MB | be75606093db |
| original_checkpoint/actor/huggingface/tokenizer_config.json | Tokenizer | 694 B | — |
| tokenizer.json | Tokenizer | 11.4 MB | be75606093db |
| tokenizer_config.json | Tokenizer | 694 B | — |
License and Download
- License
- apache-2.0
- Access
- Open weights, no gate
- Download size
- 25.7 GB
Download from [HYU_NLP] EVA Team
Released by [HYU_NLP] EVA Team through its official repository on Hugging Face. Read the license.
Built From
- Derived from Qwen/Qwen3-4B-Instruct-2507
Memory Requirements
| Precision | Weights in memory |
|---|---|
| As published | 25.7 GB |
| 16-bit | 8.0 GB |
| 8-bit | 4.0 GB |
| 4-bit | 2.0 GB |
Weights only, from the published parameter count; the key-value cache and runtime add to this.
Questions About qwen3-4b-rar-medicine-static-r0-matched-seed11-step-000
How much GPU memory does qwen3-4b-rar-medicine-static-r0-matched-seed11-step-000 need?
About 9.7 GB at 16-bit and 2.4 GB at 4-bit: the weights (4B parameters) plus a working margin. A long context needs more.
What is the cheapest GPU to run qwen3-4b-rar-medicine-static-r0-matched-seed11-step-000 on?
At 16-bit, 1x MI300X from $1.85 an hour; at 4-bit, 1x MI300X from $1.85 an hour, at the lowest on-demand prices the SAVRN Index lists.
Can I use qwen3-4b-rar-medicine-static-r0-matched-seed11-step-000 commercially?
Yes. qwen3-4b-rar-medicine-static-r0-matched-seed11-step-000 is released under Apache License 2.0. The Apache License 2.0 is a permissive open-source license. It permits commercial use, modification and redistribution. It requires keeping the license and copyright notices and any NOTICE file, stating significant changes, and it includes an express patent grant from contributors.
What is qwen3-4b-rar-medicine-static-r0-matched-seed11-step-000's context length?
262,144 tokens, from the maximum position embeddings in its published configuration.
Similar Models
We introduce the updated version of the Qwen3-4B non-thinking mode, named Qwen3-4B-Instruct-2507, featuring the following key enhancements: - Significant improvements in general capabilities, including instruction following, logical reasoning, text comprehension, mathematics, science, coding and tool usage. - Substantial gains in long-tail knowledge coverage across multiple languages. - Markedly better alignment with user preferences in subjective and open-ended tasks, enabling more helpful responses and higher-quality text generation. - Enhanced capabilities in 256K long-context understanding. Qwen3-4B-Instruct-2507 has the following features: NOTE: This model supports only non-thinking…
Qwen3 is the latest generation of large language models in Qwen series, offering a comprehensive suite of dense and mixture-of-experts (MoE) models. Building upon extensive advancements in training data, model architecture, and optimization techniques, Qwen3 delivers the following key improvements over the previously released Qwen2.5: Qwen3-4B-Base has the following features: For more details, including benchmark evaluation, hardware requirements, and inference performance, please refer to our blog, GitHub, and Documentation. The code of Qwen3 has been in the latest Hugging Face transformers and we advise you to use the latest version of transformers. With transformers<4.51.0, you will…
Qwen3 is the latest generation of large language models in Qwen series, offering a comprehensive suite of dense and mixture-of-experts (MoE) models. Building upon extensive advancements in training data, model architecture, and optimization techniques, Qwen3 delivers the following key improvements over the previously released Qwen2.5: Qwen3-4B-Base has the following features: For more details, including benchmark evaluation, hardware requirements, and inference performance, please refer to our blog, GitHub, and Documentation. The code of Qwen3 has been in the latest Hugging Face transformers and we advise you to use the latest version of transformers. With transformers<4.51.0, you will…
SRA-RiskGate-4B is an autonomous risk scoring, compliance verification, and dispute adjudication model fine-tuned on top of Qwen/Qwen3-4B-Instruct-2507. It is engineered for both ends of a stablecoin payment's operational lifecycle: Autonomous on-chain agents can hallucinate payment transfers, sign malformed calldata, or trigger catastrophic transactions during market depegs. The deterministic pre-filter and policy firewall is available directly as a verified Action Provider for the Coinbase AgentKit framework: Register RiskGateActionProvider as the payment provider on your AgentKit instance. It checks chain support, policies, and peg deviations, executing safe ERC-20 transfers only after…
Intermediate policy from dynamic OnlineRubrics-Every GRPO training. Distinct from static-rubric GRPO. Base model: Qwen/Qwen3-4B-Instruct-2507; thinking disabled. This checkpoint is a policy state used by the Phase-1 audit. No downstream medical capability or safety claim is made. Research use only; not validated for clinical decision-making. Root files are the veRL-exported Hugging Face inference model (BF16). originalcheckpoint/ preserves the exact original FSDP parameter checkpoint and tokenizer/configuration files. Optimizer state, training data, responses, rubrics, infrastructure configuration, and credentials are not included. The original is retained because export…