A 2B-parameter Qwen3.5 fine-tune, part of the Opus-Distil line in the reaperdoesntknow open-weight portfolio. Text-generation / reasoning model, trained with Unsloth + Hugging Face TRL.
Open-weight model · Text generation
Qwen3.5-2B-CyberSec
by Convergent Intelligence reaperdoesntknow/Qwen3.5-2B-CyberSec
An English Qwen3.5 2B checkpoint associated with the Trendyol Cybersecurity Instruction Tuning Dataset and exported in Transformers / Safetensors format. This release is intended for research and local experimentation.
Parameters2.3B
Context262,144
Weights4.5 GB
Licenseapache-2.0
AccessOpen weights
Monthly Downloads4.1k
Runs On
What it takes to serve Qwen3.5-2B-CyberSec (2.3B 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 | 4.5 GB | 5.5 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
| 8-bit | 2.3 GB | 2.7 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
| 4-bit | 1.1 GB | 1.4 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 Sep 18, 2026.
Model Card
By Convergent Intelligence, published under apache-2.0, revision 2f34c62a6656.
An English Qwen3.5 2B checkpoint associated with the Trendyol Cybersecurity Instruction Tuning Dataset and exported in Transformers / Safetensors format. This release is intended for research and local experimentation. The repository does not currently publish benchmark or safety-evaluation results, so the model should not be treated as a validated cybersecurity authority. The configuration identifies a Qwen3.5 conditional-generation architecture with text and vision components. Use a recent Transformers release that supports this architecture. Dependency and device behavior can vary across Transformers versions. Pin a tested environment for reproducible use. - Research on small-model…
Read Convergent Intelligence's full model card
An English Qwen3.5 2B checkpoint associated with the Trendyol Cybersecurity Instruction Tuning Dataset and exported in Transformers / Safetensors format.
This release is intended for research and local experimentation. The repository does not currently publish benchmark or safety-evaluation results, so the model should not be treated as a validated cybersecurity authority.
Lineage
- Immediate base: unsloth/Qwen3.5-2B
- Upstream family: Qwen3.5 2B
- Dataset recorded in repository metadata: Trendyol/Trendyol-Cybersecurity-Instruction-Tuning-Dataset
- Format: Transformers / Safetensors
- License: Apache-2.0
Load with Transformers
The configuration identifies a Qwen3.5 conditional-generation architecture with text and vision components. Use a recent Transformers release that supports this architecture.
from transformers import AutoModelForMultimodalLM, AutoProcessor
model_id = "reaperdoesntknow/Qwen3.5-2B-CyberSec"
processor = AutoProcessor.from_pretrained(model_id)
model = AutoModelForMultimodalLM.from_pretrained(
model_id,
device_map="auto",
)
messages = [
{"role": "user", "content": [{"type": "text", "text": "Explain least privilege."}]}
]
inputs = processor.apply_chat_template(
messages,
add_generation_prompt=True,
tokenize=True,
return_dict=True,
return_tensors="pt",
).to(model.device)
outputs = model.generate(**inputs, max_new_tokens=256)
answer = outputs[0][inputs["input_ids"].shape[-1]:]
print(processor.decode(answer, skip_special_tokens=True))
Dependency and device behavior can vary across Transformers versions. Pin a tested environment for reproducible use.
Intended use
- Research on small-model responses to cybersecurity instruction prompts.
- Local prototyping and qualitative evaluation.
- Comparison with the upstream Qwen3.5 2B checkpoint.
- Conversion and quantization experiments.
Evaluation status
No versioned benchmark report, baseline comparison, held-out test result, or safety evaluation is published in the repository reviewed for this card. Dataset association and a successful model export do not establish improved cybersecurity performance.
Useful next evidence would include:
- Results against the unchanged upstream base model.
- Held-out cybersecurity QA and scenario tests.
- General-capability regression checks.
- Hallucination, harmful-output, and misuse evaluations.
- A reproducible evaluation harness and exact revision hashes.
Limitations and safety
- The model can generate incorrect or unsafe technical guidance.
- Training data may contain errors, outdated practices, or sensitive dual-use material.
- The public card does not document dataset preprocessing, contamination checks, training hyperparameters, or checkpoint-selection criteria.
- Do not execute generated commands without review and isolation.
- Do not use the model as the sole basis for incident response, vulnerability disclosure, access-control, or other consequential security decisions.
GGUF build
For local GGUF variants, see reaperdoesntknow/Qwen3.5-2B-CyberSec-GGUF.
Part of the CIx cybersecurity model collection.
Configuration
- Architecture
- Qwen3_5ForConditionalGeneration
- Context length (tokens)
- 262,144
- Layers
- 24
- Hidden size
- 2,048
- Feed-forward size
- 6,144
- Attention heads
- 8
- Key/value heads
- 2
- Head dimension
- 256
- Vocabulary size
- 248,320
- Stored precision
- bfloat16
- Model type
- qwen3_5
Identity and Version
- Repository
- reaperdoesntknow/Qwen3.5-2B-CyberSec
- Publisher
- Convergent Intelligence
- Task
- Text generation
- Modality
- Text
- Library
- transformers
- Parameters
- 2.3B parameters
- Languages
- en
- Revision
- 2f34c62a6656c58405bebea1a7e47c60f05a9409
- First published
- 2026-08-09
- Last updated
- 2026-09-18
Files and Weights
10 files, 4.6 GB in total. The weights are 1 file totalling 4.5 GB in safetensors.
Weights1 file · 4.5 GB
Configuration4 files · 69.1 KB
Tokenizer2 files · 20.0 MB
Documentation1 file · 3.8 KB
Other1 file · 8.0 KB
Repository1 file · 1.6 KB
Every file
| File | Type | Size | SHA-256 |
|---|---|---|---|
| model.safetensors-00001-of-00001.safetensors | Weights | 4.5 GB | f2331b441530 |
| config.json | Configuration | 3.2 KB | — |
| generation_config.json | Configuration | 141 B | — |
| model.safetensors.index.json | Configuration | 64.5 KB | — |
| processor_config.json | Configuration | 1.3 KB | — |
| README.md | Documentation | 3.8 KB | — |
| chat_template.jinja | Other | 8.0 KB | — |
| .gitattributes | Repository | 1.6 KB | — |
| tokenizer.json | Tokenizer | 20.0 MB | 87a7830d63fc |
| tokenizer_config.json | Tokenizer | 15.4 KB | — |
License and Download
- License
- apache-2.0
- Access
- Open weights, no gate
- Download size
- 4.5 GB
Download from Convergent Intelligence
Released by Convergent Intelligence through its official repository on Hugging Face. Read the license.
Built From
- Derived from unsloth/Qwen3.5-2B
- Trained on (disclosed) Trendyol/Trendyol-Cybersecurity-Instruction-Tuning-Dataset
Memory Requirements
| Precision | Weights in memory |
|---|---|
| As published | 4.5 GB |
| 16-bit | 4.5 GB |
| 8-bit | 2.3 GB |
| 4-bit | 1.1 GB |
Weights only, from the published parameter count; the key-value cache and runtime add to this.
Built on This Model
- Quantized fromQwen3.5-2B-CyberSec-GGUF
- Derived fromQwen3.5-2B-CyberSec-GGUF
Questions About Qwen3.5-2B-CyberSec
How much GPU memory does Qwen3.5-2B-CyberSec need?
About 5.5 GB at 16-bit and 1.4 GB at 4-bit: the weights (2.3B parameters) plus a working margin. A long context needs more.
What is the cheapest GPU to run Qwen3.5-2B-CyberSec 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 Qwen3.5-2B-CyberSec commercially?
Yes. Qwen3.5-2B-CyberSec 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 Qwen3.5-2B-CyberSec's context length?
262,144 tokens, from the maximum position embeddings in its published configuration.
Similar Models
local, typed decisions from qwen3.5-2b. one shared state is prefetched once, each question is isolated, every allowed answer is scored as a complete token sequence, and python returns validated choice, score, and noul results. this is an inference system, not rlcd training and not a clone of typesafe jev. it never calls typesafe. the published qwen checkpoint is unchanged; fp8 changes execution precision only. probabilities are normalized support over the candidates you provide, not calibrated correctness probabilities. unlike ordinary autoregressive json generation, the model does not write a response token by token. it scores only the values supplied by the caller. complete-sequence…
A long-context DSpark speculator for Kimi K3. It supports context lengths of up to 1 million tokens. A DSpark speculator for the Kimi K3 target, enabling faster inference through speculative decoding. DSpark extends the DFlash parallel-draft backbone with a Markov logit-bias head and a per-position confidence head. This checkpoint was trained with SpecForge using hidden states from a live SGLang target engine. 64 query heads / 16 KV heads, and blocksize=7 acclen is SGLang's histogram-native request acceptance length, averaged within each question and then equally across questions. RULER V2 uses the 1M input configuration. Actual prompts span 1,000,432–1,047,925 tokens; partition acclen is…
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…
Topology-Aware Knowledge Distillation from Qwen3-30B-A3B → 1.7B TopologicalQwen is a 1.7B parameter model distilled from Qwen3-30B-A3B using Topological Knowledge Distillation (TKD) — a methodology that treats the teacher's output distribution over a concatenated token stream as a bounded variation (BV) function and decomposes knowledge transfer into three channels via the Mesh Fundamental Identity: 1. Smooth distillation (AC component) — Standard KL divergence over regions where the teacher's distribution varies continuously. This is what every other KD method does and stops at. 2. Jump corrections (D^j f) — Explicit correction terms at conceptual boundaries where the teacher's…
Extended Reasoning Distillation from Qwen3-30B-A3B-Thinking → 1.7B The most downloaded model in the Convergent Intelligence portfolio. Qwen3-1.7B-Thinking-Distil captures extended deliberation patterns from the Qwen3-30B-A3B Thinking teacher — the variant that generates long-form reasoning chains before committing to an answer — and compresses them into a 1.7B student via supervised fine-tuning on the longwriter-6k dataset. The Thinking teacher produces the richest signal of the three teacher variants in the DistilQwen family (Instruct, Thinking, Coder). Where Instruct distillation captures clean instruction-following and Coder captures hierarchical decomposition, Thinking distillation…