AssistantPepe8B / mobile stacking / Click here for TL;DR This is a project that was a long time in the making because I wanted to get it right. I'm still not fully satisfied, as there are some rough corners to sand, but for now, this would do.
Runs On
What it takes to serve Assistant_Pepe_8B (266,240 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 | 0.0 GB | 0.0 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
| 8-bit | 0.0 GB | 0.0 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
| 4-bit | 0.0 GB | 0.0 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.
SAVRN's Notes on Assistant_Pepe_8B
Start with the file listing on this one, because the memory table came back empty at every precision. The weights on disk are 16.1 GB across 22 safetensors files, which fits the 8B in the name far better than the 266,240 parameters the metadata reports, so plan hardware around 16.1 GB. The publisher built it as a chat assistant with a deliberately irreverent forum voice, and it carries a 1,073,152-token context from its base, Llama-3.1-Nemotron-8B-UltraLong-1M-Instruct_Abliterated. On the cheapest Index setup, one MI300X with 192 GB at $1.85 per hour, the weights take under a tenth of the card.
Meta's Llama 3.1 Community License applies: commercial use is permitted unless your products had more than 700 million monthly active users on the release date, with attribution and Meta's Acceptable Use Policy required. Before committing, reconcile the parameter count with the file size and test the personality on your audience.
Model Card
AssistantPepe8B / mobile stacking / Click here for TL;DR This is a project that was a long time in the making because I wanted to get it right. I'm still not fully satisfied, as there are some rough corners to sand, but for now, this would do. The goal was to maximize shitpostness along with helpfulness, without glazing the user for every retarded idea. Not an easy needle to thread. This amphibious AI has learned the ways of /g/, and speaks fluent brainrot, but will also help you out with just about anything you'll need, and won't be ashamed to roast you while at it. For those who remember OniMitsubishi12B - it was so overtly toxic that it made me worry at first (only to quickly be verified…
Excerpt from the card by Sica Rius, licensed llama3.1.
Configuration
- Architecture
- LlamaForCausalLM
- Context length (tokens)
- 1,073,152
- Layers
- 32
- Hidden size
- 4,096
- Feed-forward size
- 14,336
- Attention heads
- 32
- Key/value heads
- 8
- Head dimension
- 128
- Vocabulary size
- 129,024
- RoPE base
- 500,000
- Stored precision
- bfloat16
- Model type
- llama
Identity and Version
- Repository
- SicariusSicariiStuff/Assistant_Pepe_8B
- Publisher
- Sica Rius
- Task
- Not stated by the source
- Modality
- Other
- Library
- Not stated by the source
- Parameters
- 266,240 parameters
- Languages
- en
- Revision
- ff7aa0bc217771cb51ec0701ba80d735ff946e24
- First published
- 2026-01-14
- Last updated
- 2026-05-03
Files and Weights
22 files, 16.1 GB in total. The weights are 4 files totalling 16.1 GB in safetensors.
Every file
| File | Type | Size | SHA-256 |
|---|---|---|---|
| model-00001-of-00004.safetensors | Weights | 5.0 GB | 42955296c3d0 |
| model-00002-of-00004.safetensors | Weights | 5.0 GB | 934c91042485 |
| model-00003-of-00004.safetensors | Weights | 4.9 GB | a6e45f082b18 |
| model-00004-of-00004.safetensors | Weights | 1.2 GB | d2f48f7545c1 |
| Example_Code/snake.py | Configuration | 6.3 KB | — |
| config.json | Configuration | 926 B | — |
| generation_config.json | Configuration | 142 B | — |
| model.safetensors.index.json | Configuration | 24.0 KB | — |
| special_tokens_map.json | Configuration | 444 B | — |
| README.md | Documentation | 14.9 KB | — |
| Images/Assistant_Pepe_8B.png | Other | 3.6 MB | edb13afbb4f8 |
| Images/Examples/code.png | Other | 5.5 MB | ca56164c9b34 |
| Images/Examples/log0.png | Other | 1.0 MB | ed97a418cfa9 |
| Images/Examples/log1.png | Other | 931.2 KB | c8fe903869f1 |
| Images/Examples/log2.png | Other | 2.4 MB | 78d8c400f928 |
| Images/Examples/log3.png | Other | 1.7 MB | d05ec30b0dd8 |
| Images/Examples/log4.png | Other | 1.0 MB | 84347c5846cb |
| Images/UGI.png | Other | 297.3 KB | 98004fa29ad4 |
| chat_template.jinja | Other | 485 B | — |
| .gitattributes | Repository | 2.1 KB | — |
| tokenizer.json | Tokenizer | 17.2 MB | 6b9e4e7fb171 |
| tokenizer_config.json | Tokenizer | 50.7 KB | — |
License and Download
- License
- llama3.1
- Access
- Open weights, no gate
- Download size
- 16.1 GB
Released by Sica Rius through its official repository on Hugging Face.
Built From
- Derived from SicariusSicariiStuff/Llama-3.1-Nemotron-8B-UltraLong-1M-Instruct_Abliterated
- Trained on (disclosed) SicariusSicariiStuff/UBW_Tapestries
Memory Requirements
| Precision | Weights in memory |
|---|---|
| As published | 16.1 GB |
| 16-bit | 0.0 GB |
| 8-bit | 0.0 GB |
| 4-bit | 0.0 GB |
Weights only, from the published parameter count; the key-value cache and runtime add to this.
Questions About Assistant_Pepe_8B
How much GPU memory does Assistant_Pepe_8B need?
About 0 GB at 16-bit and 0 GB at 4-bit: the weights (266,240 parameters) plus a working margin. A long context needs more.
What is the cheapest GPU to run Assistant_Pepe_8B 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 Assistant_Pepe_8B commercially?
Yes, with conditions. Assistant_Pepe_8B is released under Meta Llama 3.1 Community License. The Llama 3.1 Community License permits commercial use, except that a licensee whose products had more than 700 million monthly active users on the release date must request a license from Meta. It requires attribution as the license specifies and compliance with Meta's Acceptable Use Policy.
What is Assistant_Pepe_8B's context length?
1,073,152 tokens, from the maximum position embeddings in its published configuration.