smg is a model from Belztjti (access requested at publisher). It has 9.6B parameters. At 16-bit it needs about 23 GB of GPU memory, which fits on 1x MI300X from $1.85 an hour, at the lowest prices in the SAVRN Index. It draws 86 downloads a month.
Runs On
What it takes to serve smg (9.6B 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 | 19.1 GB | 23.0 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
| 8-bit | 9.6 GB | 11.5 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
| 4-bit | 4.8 GB | 5.7 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.
smg on every accelerator the SAVRN Index prices, at every precision
Model Card
The publisher has not written a card for this model.
Identity and Version
- Repository
- belztjti/smg
- Publisher
- Belztjti
- Task
- Not stated by the source
- Modality
- Other
- Library
- Not stated by the source
- Parameters
- 9.6B parameters
- Languages
- glm
- Revision
- 12502298bc5cd3c5e8ddb896656f5cd6da77cc10
- First published
- 2026-08-15
- Last updated
- 2026-09-26
Files and Weights
10 files, 19.2 GB in total. The weights are 1 file totalling 19.1 GB in safetensors.
Every file
| File | Type | Size | SHA-256 |
|---|---|---|---|
| model.safetensors | Weights | 19.1 GB | — |
| .ipynb_checkpoints/config-checkpoint.json | Configuration | 774 B | — |
| config.json | Configuration | 774 B | — |
| generation_config.json | Configuration | 121 B | — |
| mlp_keep_indices.json | Configuration | 3.4 MB | — |
| .ipynb_checkpoints/prune-checkpoint.ipynb | Other | 15.7 KB | — |
| prune.ipynb | Other | 15.7 KB | — |
| .gitattributes | Repository | 1.6 KB | — |
| tokenizer.json | Tokenizer | 17.9 MB | — |
| tokenizer_config.json | Tokenizer | 6.2 MB | — |
License and Download
- License
- Not stated by the source
- Access
- Access requested at publisher
- Download size
- 19.1 GB
Belztjti grants access through its official repository on Hugging Face.
Memory Requirements
| Precision | Weights in memory |
|---|---|
| As published | 19.1 GB |
| 16-bit | 19.1 GB |
| 8-bit | 9.6 GB |
| 4-bit | 4.8 GB |
Weights only, from the published parameter count; the key-value cache and runtime add to this.
Questions About smg
How much GPU memory does smg need?
About 23 GB at 16-bit and 5.7 GB at 4-bit: the weights (9.6B parameters) plus a working margin. A long context needs more.
What is the cheapest GPU to run smg 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.