nolma-8B is a model from TrustLLM EU, released under other (access requested at publisher). It has 7.2B parameters. At 16-bit it needs about 17.3 GB of GPU memory, which fits on 1x MI300X from $1.85 an hour, at the lowest prices in the SAVRN Index.
Pretraining checkpoints of the TrustLLM project's nolma-8B run, trained on the nolma-mix corpus. Research-only; not for redistribution or commercial use. Checkpoints are added while training runs.
Runs On
What it takes to serve nolma-8B (7.2B 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 | 14.4 GB | 17.3 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
| 8-bit | 7.2 GB | 8.6 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
| 4-bit | 3.6 GB | 4.3 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.
nolma-8B on every accelerator the SAVRN Index prices, at every precision
Model Card
Pretraining checkpoints of the TrustLLM project's nolma-8B run, trained on the nolma-mix corpus. Research-only; not for redistribution or commercial use. Checkpoints are added while training runs. Each checkpoint is a branch named step-N, and main holds the latest. Access is restricted.
Excerpt from the card by TrustLLM EU, licensed other.
Identity and Version
- Repository
- TrustLLMeu/nolma-8B
- Publisher
- TrustLLM EU
- Task
- Not stated by the source
- Modality
- Other
- Library
- Not stated by the source
- Parameters
- 7.2B parameters
- Languages
- Not stated by the source
- Revision
- d32bcc89c3802db30c0ad943bfafd659fa4361a4
- First published
- 2026-09-23
- Last updated
- 2026-09-26
Files and Weights
14 files, 14.4 GB in total. The weights are 3 files totalling 14.4 GB in safetensors.
Every file
| File | Type | Size | SHA-256 |
|---|---|---|---|
| model-00001-of-00003.safetensors | Weights | 5.0 GB | — |
| model-00002-of-00003.safetensors | Weights | 5.0 GB | — |
| model-00003-of-00003.safetensors | Weights | 4.4 GB | — |
| config.json | Configuration | 1.8 KB | — |
| configuration_opt_moe.py | Configuration | 21.5 KB | — |
| model.safetensors.index.json | Configuration | 417.0 KB | — |
| modeling_opt_moe.py | Configuration | 51.3 KB | — |
| special_tokens_map.json | Configuration | 293 B | — |
| README.md | Documentation | 948 B | — |
| SHA256SUMS | Other | 992 B | — |
| chat_template.jinja | Other | 303 B | — |
| .gitattributes | Repository | 1.6 KB | — |
| tokenizer.json | Tokenizer | 35.5 MB | — |
| tokenizer_config.json | Tokenizer | 2.4 KB | — |
License and Download
- License
- other
- Access
- Access requested at publisher
- Download size
- 14.4 GB
TrustLLM EU grants access through its official repository on Hugging Face.
Memory Requirements
| Precision | Weights in memory |
|---|---|
| As published | 14.4 GB |
| 16-bit | 14.4 GB |
| 8-bit | 7.2 GB |
| 4-bit | 3.6 GB |
Weights only, from the published parameter count; the key-value cache and runtime add to this.
Questions About nolma-8B
How much GPU memory does nolma-8B need?
About 17.3 GB at 16-bit and 4.3 GB at 4-bit: the weights (7.2B parameters) plus a working margin. A long context needs more.
What is the cheapest GPU to run nolma-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.
What license is nolma-8B released under?
other, as its publisher declares it. Read the license text before commercial use.