Open-weight model
Llama-3.1-8B-Instruct-cpt-tesi_all_tasks
by Pietro Ferrazzi ferrazzipietro/Llama-3.1-8B-Instruct-cpt-tesi_all_tasks
Llama-3.1-8B-Instruct-cpt-tesi_all_tasks is an open-weight model from Pietro Ferrazzi. Its published files total 185.1 MB.
This model is a fine-tuned version of ferrazzipietro/Llama-3.1-8B-Instruct-cpt-tesiall on an unknown dataset.
Model Card
This model is a fine-tuned version of ferrazzipietro/Llama-3.1-8B-Instruct-cpt-tesiall on an unknown dataset. It achieves the following results on the evaluation set: The following hyperparameters were used during training: - learningrate: 0.0003 - trainbatchsize: 64 - evalbatchsize: 32 - distributedtype: multi-GPU - gradientaccumulationsteps: 2 - totaltrainbatchsize: 128 - lrschedulertype: cosine - lrschedulerwarmupratio: 0.1 - numepochs: 10 - PEFT 0.14.0 - Transformers 4.51.0 - Pytorch 2.8.0+cu128 - Datasets 3.6.0 - Tokenizers 0.21.0
Excerpt from the card by Pietro Ferrazzi.
Identity and Version
- Repository
- ferrazzipietro/Llama-3.1-8B-Instruct-cpt-tesi_all_tasks
- Publisher
- Pietro Ferrazzi
- Task
- Not stated by the source
- Modality
- Other
- Library
- peft
- Parameters
- Not stated by the source
- Languages
- Not stated by the source
- Revision
- 28f8cc4f0e1abe3d068bc6eacbd91a34d39fc076
- First published
- 2026-09-22
- Last updated
- 2026-09-23
Files and Weights
8 files, 185.1 MB in total. The weights are 2 files totalling 167.8 MB in bin, safetensors.
Every file
| File | Type | Size | SHA-256 |
|---|---|---|---|
| adapter_model.safetensors | Weights | 167.8 MB | c650fd1bb9ba |
| training_args.bin | Weights | 7.7 KB | e6473c53d9ae |
| adapter_config.json | Configuration | 823 B | — |
| special_tokens_map.json | Configuration | 434 B | — |
| README.md | Documentation | 2.1 KB | — |
| .gitattributes | Repository | 1.6 KB | — |
| tokenizer.json | Tokenizer | 17.2 MB | ddd72091ebf5 |
| tokenizer_config.json | Tokenizer | 55.6 KB | — |
License and Download
- License
- Not stated by the source
- Access
- Open weights, no gate
- Download size
- 167.8 MB
Released by Pietro Ferrazzi through its official repository on Hugging Face.
Built From
- Adapter of ferrazzipietro/Llama-3.1-8B-Instruct-cpt-tesi_all
- Derived from ferrazzipietro/Llama-3.1-8B-Instruct-cpt-tesi_all
Memory Requirements
| Precision | Weights in memory |
|---|---|
| As published | 167.8 MB |
Weights only, from the published parameter count; the key-value cache and runtime add to this.