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
arb-arab-100mb-after-ppt-Dp-100mb-ckpt500_seed10
by Francesca Padovani fpadovani/arb-arab-100mb-after-ppt-Dp-100mb-ckpt500_seed10
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.
Parameters125M
Context—
Weights499.1 MB
License—
AccessOpen weights
Monthly Downloads—
Runs On
What it takes to serve arb-arab-100mb-after-ppt-Dp-100mb-ckpt500_seed10 (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 Sep 18, 2026.
Model Card
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.
Excerpt from the card by Francesca Padovani.
Configuration
- Architecture
- GPT2LMHeadModel
- Vocabulary size
- 51,200
- Model type
- gpt2
Identity and Version
- Repository
- fpadovani/arb-arab-100mb-after-ppt-Dp-100mb-ckpt500_seed10
- Publisher
- Francesca Padovani
- Task
- Text generation
- Modality
- Text
- Library
- transformers
- Parameters
- 125M parameters
- Languages
- trl, sft
- Revision
- 471b38e8eb4c99bed4a73e0505d09e7d19c25d5b
- First published
- 2026-09-18
- Last updated
- 2026-09-18
Files and Weights
19 files, 502.5 MB in total. The weights are 5 files totalling 499.1 MB in bin, pth, safetensors.
Weights5 files · 499.1 MB
Configuration8 files · 326.6 KB
Tokenizer4 files · 3.1 MB
Documentation1 file · 1.9 KB
Repository1 file · 1.5 KB
Every file
| File | Type | Size | SHA-256 |
|---|---|---|---|
| checkpoint-4000/model.safetensors | Weights | 249.6 MB | 6a5728d86e1d |
| checkpoint-4000/rng_state.pth | Weights | 14.6 KB | 08dfb8e5c33b |
| checkpoint-4000/training_args.bin | Weights | 6.4 KB | 9f587ccf2f2b |
| model.safetensors | Weights | 249.6 MB | 6a5728d86e1d |
| training_args.bin | Weights | 6.4 KB | 9f587ccf2f2b |
| added_tokens.json | Configuration | 27.7 KB | — |
| checkpoint-4000/added_tokens.json | Configuration | 27.7 KB | — |
| checkpoint-4000/config.json | Configuration | 803 B | — |
| checkpoint-4000/generation_config.json | Configuration | 154 B | — |
| checkpoint-4000/special_tokens_map.json | Configuration | 22.6 KB | — |
| checkpoint-4000/trainer_state.json | Configuration | 224.2 KB | — |
| config.json | Configuration | 803 B | — |
| special_tokens_map.json | Configuration | 22.6 KB | — |
| README.md | Documentation | 1.9 KB | — |
| .gitattributes | Repository | 1.5 KB | — |
| checkpoint-4000/spiece.model | Tokenizer | 1.3 MB | 9689bcf676e0 |
| checkpoint-4000/tokenizer_config.json | Tokenizer | 233.9 KB | — |
| spiece.model | Tokenizer | 1.3 MB | 9689bcf676e0 |
| tokenizer_config.json | Tokenizer | 233.9 KB | — |
License and Download
- License
- Not stated by the source
- Access
- Open weights, no gate
- Download size
- 499.1 MB
Download from Francesca Padovani
Released by Francesca Padovani through its official repository on Hugging Face.
Built From
- Derived from fpadovani/arb-arab-100mb-ppt-Dp-100mb_seed10
Memory Requirements
| Precision | Weights in memory |
|---|---|
| As published | 499.1 MB |
| 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 arb-arab-100mb-after-ppt-Dp-100mb-ckpt500_seed10
How much GPU memory does arb-arab-100mb-after-ppt-Dp-100mb-ckpt500_seed10 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 arb-arab-100mb-after-ppt-Dp-100mb-ckpt500_seed10 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-shuff-dyck-10mbseed10. It has been trained using TRL. This model was trained with SFT.
This model is a fine-tuned version of GeorgeUwaifo/iviegpt2new01cresults on an unknown dataset. It achieves the following results on the evaluation set: The following hyperparameters were used during training: - learningrate: 5e-05 - trainbatchsize: 4 - evalbatchsize: 8 - gradientaccumulationsteps: 4 - totaltrainbatchsize: 16 - lrschedulertype: linear - lrschedulerwarmupsteps: 387 - numepochs: 5 - mixedprecisiontraining: Native AMP - Transformers 5.16.1 - Pytorch 2.11.0+cu128 - Datasets 4.8.5 - Tokenizers 0.23.1
This model is a custom-code derivative of AxiomicLabs/GPT-X2-125M, adapted for experimental long-context causal language modeling and architecture research. The repository includes a Hugging Face Transformers-compatible GPT-X2 implementation with optional Symplectic Metric-RoPE Governor support and training utilities built around CIxOpt, a heterogeneous optimizer developed for efficient parameter routing across large projection matrices, sensitive normalization parameters, and optional governor modules. The model is intended as a research checkpoint for compact long-context generation, positional encoding experiments, optimizer testing, and continued fine-tuning. This implementation uses a…
SmolLM2 is a family of compact language models available in three size: 135M, 360M, and 1.7B parameters. They are capable of solving a wide range of tasks while being lightweight enough to run on-device. More details in our paper: https://arxiv.org/abs/2502.02737 SmolLM2 demonstrates significant advances over its predecessor SmolLM1, particularly in instruction following, knowledge, reasoning. The 135M model was trained on 2 trillion tokens using a diverse dataset combination: FineWeb-Edu, DCLM, The Stack, along with new filtered datasets we curated and will release soon. We developed the instruct version through supervised fine-tuning (SFT) using a combination of public datasets and our…
SmolLM2 is a family of compact language models available in three size: 135M, 360M, and 1.7B parameters. They are capable of solving a wide range of tasks while being lightweight enough to run on-device. More details in our paper https://arxiv.org/abs/2502.02737 SmolLM2 demonstrates significant advances over its predecessor SmolLM1, particularly in instruction following, knowledge, reasoning. The 135M model was trained on 2 trillion tokens using a diverse dataset combination: FineWeb-Edu, DCLM, The Stack, along with new filtered datasets we curated and will release soon. We developed the instruct version through supervised fine-tuning (SFT) using a combination of public datasets and our own…