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…
Open-weight model · Text generation
greenlight-extractor-sft-v5-ed827924-74462f5d-33d
by Machinelearningmo machinelearningmo/greenlight-extractor-sft-v5-ed827924-74462f5d-33d
greenlight-extractor-sft-v5-ed827924-74462f5d-33d is an open-weight model for text generation from Machinelearningmo, 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 a merged QLoRA derivative of Qwen/Qwen3-4B-Instruct-2507, published for Headroom, developed under the working name Greenlight. The original model identifier is retained for reproducibility.
Parameters4B
Context262,144
Weights3.4 GB
Licenseapache-2.0
AccessOpen weights
Monthly Downloads—
Runs On
What it takes to serve greenlight-extractor-sft-v5-ed827924-74462f5d-33d (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 18, 2026.
Model Card
By Machinelearningmo, published under apache-2.0, revision f774455e2c98.
This is a merged QLoRA derivative of Qwen/Qwen3-4B-Instruct-2507, published for Headroom, developed under the working name Greenlight. The original model identifier is retained for reproducibility. Training used 336 synthetic examples and 300 configured optimizer steps through Veri. Dataset SHA-256: ed827924d7ba4118fa15c5ef71fb211c557c3390d87384f21af4eb86ae59d051. Training job: 42d582a35d2f. Rank 8, 4-bit loading, learning rate 0.0001, sequence limit 2048, micro-batch 1, gradient accumulation 2. Prompt profile: reference-contract-v5. This model proposes bounded structured payment terms. Consult the repository’s matched frozen base/tuned evaluation before making any quality claim. The corpus…
Read Machinelearningmo's full model card
Headroom payment-clause extractor
This is a merged QLoRA derivative of Qwen/Qwen3-4B-Instruct-2507, published for Headroom, developed under the working name Greenlight. The original model identifier is retained for reproducibility.
Training used 336 synthetic examples and 300 configured optimizer steps through Veri. Dataset SHA-256: ed827924d7ba4118fa15c5ef71fb211c557c3390d87384f21af4eb86ae59d051. Training job: 42d582a35d2f. Rank 8, 4-bit loading, learning rate 0.0001, sequence limit 2048, micro-batch 1, gradient accumulation 2. Prompt profile: reference-contract-v5.
This model proposes bounded structured payment terms. Consult the repository’s matched frozen base/tuned evaluation before making any quality claim.
The corpus contains generated clause text and developer-reviewed labels. No credentials or private customer records were used. Final labels are not uploaded to the trainer. Local tokenization validates the intended completion-only objective; remote effective label tensors were not directly inspected.
Outputs need source validation and owner review. Invoice eligibility, invoice issuance, contractual due date and actual bank settlement are distinct. Financial calculations remain in Headroom’s deterministic cents engine. This model does not establish cash availability, authenticate documents or authorize payments.
The base model is licensed Apache 2.0. Headroom’s modifications are the QLoRA fine-tune described above. Quantized merge can change generations; merged weights require their own serving and quality verification. See the source repository for corpus hashes, evaluation boundaries and reproducible owner tooling.
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
- RoPE base
- 5,000,000
- Model type
- qwen3
- Quantization
- bitsandbytes
Identity and Version
- Repository
- machinelearningmo/greenlight-extractor-sft-v5-ed827924-74462f5d-33d
- Publisher
- Machinelearningmo
- Task
- Text generation
- Modality
- Text
- Library
- Not stated by the source
- Parameters
- 4B parameters
- Languages
- sft
- Revision
- f774455e2c98b46ccfe261bd71718587bf17653e
- First published
- 2026-10-18
- Last updated
- 2026-10-18
Files and Weights
13 files, 3.4 GB in total. The weights are 1 file totalling 3.4 GB in safetensors.
Weights1 file · 3.4 GB
Configuration4 files · 3.4 KB
Tokenizer4 files · 15.9 MB
Documentation2 files · 13.3 KB
Other1 file · 2.6 KB
Repository1 file · 1.6 KB
Every file
| File | Type | Size | SHA-256 |
|---|---|---|---|
| model.safetensors | Weights | 3.4 GB | c2cac2393fc5 |
| added_tokens.json | Configuration | 707 B | — |
| config.json | Configuration | 2.0 KB | — |
| generation_config.json | Configuration | 187 B | — |
| special_tokens_map.json | Configuration | 499 B | — |
| LICENSE | Documentation | 11.4 KB | — |
| README.md | Documentation | 1.9 KB | — |
| chat_template.jinja | Other | 2.6 KB | — |
| .gitattributes | Repository | 1.6 KB | — |
| merges.txt | Tokenizer | 1.7 MB | — |
| tokenizer.json | Tokenizer | 11.4 MB | aeb13307a71a |
| tokenizer_config.json | Tokenizer | 5.4 KB | — |
| vocab.json | Tokenizer | 2.8 MB | — |
License and Download
- License
- apache-2.0
- Access
- Open weights, no gate
- Download size
- 3.4 GB
Download from Machinelearningmo
Released by Machinelearningmo through its official repository on Hugging Face. Read the license.
Built From
- Derived from Qwen/Qwen3-4B-Instruct-2507
- Quantized from Qwen/Qwen3-4B-Instruct-2507
Memory Requirements
| Precision | Weights in memory |
|---|---|
| As published | 3.4 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 greenlight-extractor-sft-v5-ed827924-74462f5d-33d
How much GPU memory does greenlight-extractor-sft-v5-ed827924-74462f5d-33d 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 greenlight-extractor-sft-v5-ed827924-74462f5d-33d 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 greenlight-extractor-sft-v5-ed827924-74462f5d-33d commercially?
Yes. greenlight-extractor-sft-v5-ed827924-74462f5d-33d 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 greenlight-extractor-sft-v5-ed827924-74462f5d-33d'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…
![[HYU_NLP] EVA Team logo](/media/mh-logo-hyu-nlp-eval-558e075b88.webp){kind=logo}
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…