This model is a fine-tuned version of fpadovani/arb-arab-100mb-ppt-shuff-dyck-100mbseed10. It has been trained using TRL. This model was trained with SFT.
Open-weight model · Text generation
swa-latn-100mb-after-ppt-Dp-100mb-packed-bfdiso-ckpt500_seed3407
by Francesca Padovani francesca9805/swa-latn-100mb-after-ppt-Dp-100mb-packed-bfdiso-ckpt500_seed3407
swa-latn-100mb-after-ppt-Dp-100mb-packed-bfdiso-ckpt500_seed3407 is an open-weight model for text generation from Francesca Padovani. It has 125M parameters. At 16-bit it needs about 0.3 GB of GPU memory, which fits on 1x MI300X from $1.85 an hour, at the lowest prices in the SAVRN Index.
This model is a fine-tuned version of francesca9805/swa-latn-100mb-ppt-Dp-100mb-packed-bfdisoseed3407. It has been trained using TRL. This model was trained with SFT.
Parameters125M
Context—
Weights1.5 GB
License—
AccessOpen weights
Monthly Downloads—
Runs On
What it takes to serve swa-latn-100mb-after-ppt-Dp-100mb-packed-bfdiso-ckpt500_seed3407 (125M 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 | 0.2 GB | 0.3 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
| 8-bit | 0.1 GB | 0.1 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
| 4-bit | 0.1 GB | 0.1 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 7, 2026.
Model Card
This model is a fine-tuned version of francesca9805/swa-latn-100mb-ppt-Dp-100mb-packed-bfdisoseed3407. It has been trained using TRL. This model was trained with SFT.
Excerpt from the card by Francesca Padovani.
Configuration
- Architecture
- GPT2LMHeadModel
- Vocabulary size
- 51,200
- Model type
- gpt2
Identity and Version
- Repository
- francesca9805/swa-latn-100mb-after-ppt-Dp-100mb-packed-bfdiso-ckpt500_seed3407
- Publisher
- Francesca Padovani
- Task
- Text generation
- Modality
- Text
- Library
- transformers
- Parameters
- 125M parameters
- Languages
- sft, trl
- Revision
- 21b951bcc5b95206227c24080208bf9c3a70aea7
- First published
- 2026-10-03
- Last updated
- 2026-10-03
Files and Weights
59 files, 1.5 GB in total. The weights are 17 files totalling 1.5 GB in bin, pth, safetensors.
Weights17 files · 1.5 GB
Configuration28 files · 2.8 MB
Tokenizer12 files · 8.1 MB
Documentation1 file · 2.0 KB
Repository1 file · 1.5 KB
Every file
| File | Type | Size | SHA-256 |
|---|---|---|---|
| checkpoint-12000/model.safetensors | Weights | 249.6 MB | d6993066d409 |
| checkpoint-12000/rng_state.pth | Weights | 14.6 KB | 8d4d21193594 |
| checkpoint-12000/training_args.bin | Weights | 6.4 KB | cf098a444220 |
| checkpoint-15000/model.safetensors | Weights | 249.6 MB | 7df172bb7a8e |
| checkpoint-15000/rng_state.pth | Weights | 14.6 KB | 418c386345f9 |
| checkpoint-15000/training_args.bin | Weights | 6.4 KB | cf098a444220 |
| checkpoint-3000/model.safetensors | Weights | 249.6 MB | 26d0dadadc0c |
| checkpoint-3000/rng_state.pth | Weights | 14.6 KB | 58596a2266b8 |
| checkpoint-3000/training_args.bin | Weights | 6.4 KB | cf098a444220 |
| checkpoint-6000/model.safetensors | Weights | 249.6 MB | 47a3b05d1a51 |
| checkpoint-6000/rng_state.pth | Weights | 14.6 KB | e81aa7b252ec |
| checkpoint-6000/training_args.bin | Weights | 6.4 KB | cf098a444220 |
| checkpoint-9000/model.safetensors | Weights | 249.6 MB | 74780cdf0409 |
| checkpoint-9000/rng_state.pth | Weights | 14.6 KB | 408f9db67ad3 |
| checkpoint-9000/training_args.bin | Weights | 6.4 KB | cf098a444220 |
| model.safetensors | Weights | 249.6 MB | 7df172bb7a8e |
| training_args.bin | Weights | 6.4 KB | cf098a444220 |
| added_tokens.json | Configuration | 27.7 KB | — |
| checkpoint-12000/added_tokens.json | Configuration | 27.7 KB | — |
| checkpoint-12000/config.json | Configuration | 803 B | — |
| checkpoint-12000/generation_config.json | Configuration | 154 B | — |
| checkpoint-12000/special_tokens_map.json | Configuration | 22.6 KB | — |
| checkpoint-12000/trainer_state.json | Configuration | 675.3 KB | — |
| checkpoint-15000/added_tokens.json | Configuration | 27.7 KB | — |
| checkpoint-15000/config.json | Configuration | 803 B | — |
| checkpoint-15000/generation_config.json | Configuration | 154 B | — |
| checkpoint-15000/special_tokens_map.json | Configuration | 22.6 KB | — |
| checkpoint-15000/trainer_state.json | Configuration | 845.2 KB | — |
| checkpoint-3000/added_tokens.json | Configuration | 27.7 KB | — |
| checkpoint-3000/config.json | Configuration | 803 B | — |
| checkpoint-3000/generation_config.json | Configuration | 154 B | — |
| checkpoint-3000/special_tokens_map.json | Configuration | 22.6 KB | — |
| checkpoint-3000/trainer_state.json | Configuration | 167.9 KB | — |
| checkpoint-6000/added_tokens.json | Configuration | 27.7 KB | — |
| checkpoint-6000/config.json | Configuration | 803 B | — |
| checkpoint-6000/generation_config.json | Configuration | 154 B | — |
| checkpoint-6000/special_tokens_map.json | Configuration | 22.6 KB | — |
| checkpoint-6000/trainer_state.json | Configuration | 337.0 KB | — |
| checkpoint-9000/added_tokens.json | Configuration | 27.7 KB | — |
| checkpoint-9000/config.json | Configuration | 803 B | — |
| checkpoint-9000/generation_config.json | Configuration | 154 B | — |
| checkpoint-9000/special_tokens_map.json | Configuration | 22.6 KB | — |
| checkpoint-9000/trainer_state.json | Configuration | 506.0 KB | — |
| config.json | Configuration | 803 B | — |
| special_tokens_map.json | Configuration | 22.6 KB | — |
| README.md | Documentation | 2.0 KB | — |
| .gitattributes | Repository | 1.5 KB | — |
| checkpoint-12000/spiece.model | Tokenizer | 1.1 MB | 0391325a5092 |
| checkpoint-12000/tokenizer_config.json | Tokenizer | 233.6 KB | — |
| checkpoint-15000/spiece.model | Tokenizer | 1.1 MB | 0391325a5092 |
| checkpoint-15000/tokenizer_config.json | Tokenizer | 233.6 KB | — |
| checkpoint-3000/spiece.model | Tokenizer | 1.1 MB | 0391325a5092 |
| checkpoint-3000/tokenizer_config.json | Tokenizer | 233.6 KB | — |
| checkpoint-6000/spiece.model | Tokenizer | 1.1 MB | 0391325a5092 |
| checkpoint-6000/tokenizer_config.json | Tokenizer | 233.6 KB | — |
| checkpoint-9000/spiece.model | Tokenizer | 1.1 MB | 0391325a5092 |
| checkpoint-9000/tokenizer_config.json | Tokenizer | 233.6 KB | — |
| spiece.model | Tokenizer | 1.1 MB | 0391325a5092 |
| tokenizer_config.json | Tokenizer | 233.6 KB | — |
License and Download
- License
- Not stated by the source
- Access
- Open weights, no gate
- Download size
- 1.5 GB
Download from Francesca Padovani
Released by Francesca Padovani through its official repository on Hugging Face.
Built From
- Derived from francesca9805/swa-latn-100mb-ppt-Dp-100mb-packed-bfdiso_seed3407
Memory Requirements
| Precision | Weights in memory |
|---|---|
| As published | 1.5 GB |
| 16-bit | 0.2 GB |
| 8-bit | 0.1 GB |
| 4-bit | 0.1 GB |
Weights only, from the published parameter count; the key-value cache and runtime add to this.
Questions About swa-latn-100mb-after-ppt-Dp-100mb-packed-bfdiso-ckpt500_seed3407
How much GPU memory does swa-latn-100mb-after-ppt-Dp-100mb-packed-bfdiso-ckpt500_seed3407 need?
About 0.3 GB at 16-bit and 0.1 GB at 4-bit: the weights (125M parameters) plus a working margin. A long context needs more.
What is the cheapest GPU to run swa-latn-100mb-after-ppt-Dp-100mb-packed-bfdiso-ckpt500_seed3407 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.
Similar Models
This model is a fine-tuned version of fpadovani/arb-arab-100mb-ppt-Dp-100mbseed10. It has been trained using TRL. This model was trained with SFT.
This model is a fine-tuned version of fpadovani/arb-arab-100mb-ppt-shuff-dyck-10mbseed10. It has been trained using TRL. This model was trained with SFT.
This model is a fine-tuned version of goldfish-models/swalatn100mb. It has been trained using TRL. This model was trained with SFT.
This model is a fine-tuned version of goldfish-models/swalatn100mb. It has been trained using TRL. This model was trained with SFT.
This model is a fine-tuned version of goldfish-models/swalatn100mb. It has been trained using TRL. This model was trained with SFT.