malayalam-moe-grpo-sft is a model from Siyahul Haque (access requested at publisher). It has 371M parameters. At 16-bit it needs about 0.9 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 malayalam-moe-grpo-sft (371M 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 | 0.7 GB | 0.9 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
| 8-bit | 0.4 GB | 0.4 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
| 4-bit | 0.2 GB | 0.2 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 19, 2026.
malayalam-moe-grpo-sft 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
- siyah1/malayalam-moe-grpo-sft
- Publisher
- Siyahul Haque
- Task
- Not stated by the source
- Modality
- Other
- Library
- Not stated by the source
- Parameters
- 371M parameters
- Languages
- Not stated by the source
- Revision
- 35a690f5e5b9605a8277ac31ddd830bdb0698a97
- First published
- 2026-09-25
- Last updated
- 2026-09-26
Files and Weights
14 files, 5.9 GB in total. The weights are 8 files totalling 5.9 GB in bin, pt, pth, safetensors.
Every file
| File | Type | Size | SHA-256 |
|---|---|---|---|
| last_checkpoint/model.safetensors | Weights | 1.5 GB | — |
| last_checkpoint/optimizer.pt | Weights | 3.0 GB | — |
| last_checkpoint/rng_state.pth | Weights | 14.6 KB | — |
| last_checkpoint/scaler.pt | Weights | 1.4 KB | — |
| last_checkpoint/scheduler.pt | Weights | 1.5 KB | — |
| last_checkpoint/training_args.bin | Weights | 5.2 KB | — |
| model.safetensors | Weights | 1.5 GB | — |
| training_args.bin | Weights | 5.2 KB | — |
| config.json | Configuration | 611 B | — |
| last_checkpoint/config.json | Configuration | 611 B | — |
| last_checkpoint/trainer_state.json | Configuration | 49.6 KB | — |
| .gitattributes | Repository | 1.5 KB | — |
| tokenizer.json | Tokenizer | 2.5 MB | — |
| tokenizer_config.json | Tokenizer | 338 B | — |
License and Download
- License
- Not stated by the source
- Access
- Access requested at publisher
- Download size
- 5.9 GB
Siyahul Haque grants access through its official repository on Hugging Face.
Memory Requirements
| Precision | Weights in memory |
|---|---|
| As published | 5.9 GB |
| 16-bit | 0.7 GB |
| 8-bit | 0.4 GB |
| 4-bit | 0.2 GB |
Weights only, from the published parameter count; the key-value cache and runtime add to this.
Questions About malayalam-moe-grpo-sft
How much GPU memory does malayalam-moe-grpo-sft need?
About 0.9 GB at 16-bit and 0.2 GB at 4-bit: the weights (371M parameters) plus a working margin. A long context needs more.
What is the cheapest GPU to run malayalam-moe-grpo-sft 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.