Open-weight model
tournament-tourn_c8068e356b7c22c8_20260928-e656b77d-f220-4bdf-8a37-969d6c803b50-5E2fNjMW
by Grads gradients-io-tournaments/tournament-tourn_c8068e356b7c22c8_20260928-e656b77d-f220-4bdf-8a37-969d6c803b50-5E2fNjMW
tournament-tourn_c8068e356b7c22c8_20260928-e656b77d-f220-4bdf-8a37-969d6c803b50-5E2fNjMW is an open-weight model from Grads. It has 8.5B parameters and a 8,192-token context. At 16-bit it needs about 20.5 GB of GPU memory, which fits on 1x MI300X from $1.85 an hour, at the lowest prices in the SAVRN Index.
Runs On
What it takes to serve tournament-tourn_c8068e356b7c22c8_20260928-e656b77d-f220-4bdf-8a37-969d6c803b50-5E2fNjMW (8.5B 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 | 17.1 GB | 20.5 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
| 8-bit | 8.5 GB | 10.2 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
| 4-bit | 4.3 GB | 5.1 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
The publisher has not written a card for this model.
Configuration
- Architecture
- GemmaForCausalLM
- Context length (tokens)
- 8,192
- Layers
- 28
- Hidden size
- 3,072
- Feed-forward size
- 24,576
- Attention heads
- 16
- Key/value heads
- 16
- Head dimension
- 256
- Vocabulary size
- 256,000
- RoPE base
- 10000
- Stored precision
- bfloat16
- Model type
- gemma
Identity and Version
- Repository
- gradients-io-tournaments/tournament-tourn_c8068e356b7c22c8_20260928-e656b77d-f220-4bdf-8a37-969d6c803b50-5E2fNjMW
- Publisher
- Grads
- Task
- Not stated by the source
- Modality
- Other
- Library
- Not stated by the source
- Parameters
- 8.5B parameters
- Languages
- Not stated by the source
- Revision
- 45b68c40c4a9d16fa772f5fc80851b669c1a6ac1
- First published
- 2026-09-29
- Last updated
- 2026-09-29
Files and Weights
9 files, 17.1 GB in total. The weights are 1 file totalling 17.1 GB in safetensors.
Every file
| File | Type | Size | SHA-256 |
|---|---|---|---|
| model.safetensors | Weights | 17.1 GB | dcf5dab84d03 |
| config.json | Configuration | 832 B | — |
| generation_config.json | Configuration | 154 B | — |
| model.safetensors.index.json | Configuration | 20.9 KB | — |
| special_tokens_map.json | Configuration | 695 B | — |
| loss.txt | Other | 27 B | — |
| .gitattributes | Repository | 1.6 KB | — |
| tokenizer.json | Tokenizer | 34.4 MB | 83d1a3e50bab |
| tokenizer_config.json | Tokenizer | 46.5 KB | — |
License and Download
- License
- Not stated by the source
- Access
- Open weights, no gate
- Download size
- 17.1 GB
Released by Grads through its official repository on Hugging Face.
Memory Requirements
| Precision | Weights in memory |
|---|---|
| As published | 17.1 GB |
| 16-bit | 17.1 GB |
| 8-bit | 8.5 GB |
| 4-bit | 4.3 GB |
Weights only, from the published parameter count; the key-value cache and runtime add to this.
Questions About tournament-tourn_c8068e356b7c22c8_20260928-e656b77d-f220-4bdf-8a37-969d6c803b50-5E2fNjMW
How much GPU memory does tournament-tourn_c8068e356b7c22c8_20260928-e656b77d-f220-4bdf-8a37-969d6c803b50-5E2fNjMW need?
About 20.5 GB at 16-bit and 5.1 GB at 4-bit: the weights (8.5B parameters) plus a working margin. A long context needs more.
What is the cheapest GPU to run tournament-tourn_c8068e356b7c22c8_20260928-e656b77d-f220-4bdf-8a37-969d6c803b50-5E2fNjMW 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 is tournament-tourn_c8068e356b7c22c8_20260928-e656b77d-f220-4bdf-8a37-969d6c803b50-5E2fNjMW's context length?
8,192 tokens, from the maximum position embeddings in its published configuration.