---
library_name: transformers
base_model: giux78/zagreus-test-202000
tags:
- generated_from_trainer
model-index:
- name: ale_outputs/opendata-sft-debug
  results: []
---

<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->

[<img src="https://raw.githubusercontent.com/axolotl-ai-cloud/axolotl/main/image/axolotl-badge-web.png" alt="Built with Axolotl" width="200" height="32"/>](https://github.com/axolotl-ai-cloud/axolotl)
<details><summary>See axolotl config</summary>

axolotl version: `0.9.2`
```yaml
   # ============= SFT DEBUG (~1M conv) =============
base_model: giux78/zagreus-test-202000
strict: false
output_dir: ./ale_outputs/opendata-sft-debug
seed: 42

datasets:
  - path: /leonardo_work/EUHPC_A04_045/training/opendata-1000000
    type: chat_template
    field_messages: conversation
    roles_to_train: ["assistant"]
    train_on_eos: turn

dataset_prepared_path: ./ale_outputs/dataset_cache/opendata-sft
default_system_message: "Sei un assistente utile."
chat_template: llama3

sequence_len: 4096
sample_packing: true
eval_sample_packing: true
pad_to_sequence_len: true

optimizer: adamw_torch_fused
lr_scheduler: constant_with_warmup   # <-- per isolare il comportamento
learning_rate: 1.5e-5
warmup_ratio: 0.05                   # un po’ più lungo in debug
weight_decay: 0.01
max_grad_norm: 1.0

micro_batch_size: 1
gradient_accumulation_steps: 8
# Usa max_steps per “più step” indipendentemente dalla lunghezza effettiva del dataset
max_steps: 2000                      # ≈ 4x gli step attuali
#num_epochs: null                     # ignora epoche quando max_steps è settato

bf16: auto
flash_attention: true
gradient_checkpointing: true

logging_steps: 10
eval_strategy: steps
eval_steps: 100
save_strategy: steps
save_steps: 200
save_total_limit: 2
val_set_size: 10000

fsdp_config:
  fsdp_sharding_strategy: FULL_SHARD
  fsdp_auto_wrap_policy: TRANSFORMER_BASED_WRAP
  fsdp_transformer_layer_cls_to_wrap: LlamaDecoderLayer
  fsdp_backward_prefetch_policy: BACKWARD_PRE
  fsdp_state_dict_type: FULL_STATE_DICT

special_tokens:
  pad_token: <|end_of_text|>
  eos_token: <|end_of_text|>


```

</details><br>

# ale_outputs/opendata-sft-debug

This model is a fine-tuned version of [giux78/zagreus-test-202000](https://huggingface.co/giux78/zagreus-test-202000) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 3.4896

## Model description

More information needed

## Intended uses & limitations

More information needed

## Training and evaluation data

More information needed

## Training procedure

### Training hyperparameters

The following hyperparameters were used during training:
- learning_rate: 1.5e-05
- train_batch_size: 1
- eval_batch_size: 1
- seed: 42
- distributed_type: multi-GPU
- num_devices: 32
- gradient_accumulation_steps: 8
- total_train_batch_size: 256
- total_eval_batch_size: 32
- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: constant_with_warmup
- lr_scheduler_warmup_steps: 24
- training_steps: 2000

### Training results

| Training Loss | Epoch  | Step | Validation Loss |
|:-------------:|:------:|:----:|:---------------:|
| No log        | 0.0020 | 1    | 3.7876          |
| 3.7672        | 0.2011 | 100  | 3.7539          |
| 3.7566        | 0.4022 | 200  | 3.7122          |
| 3.6864        | 0.6033 | 300  | 3.6783          |
| 3.6557        | 0.8044 | 400  | 3.6385          |
| 3.6056        | 1.0040 | 500  | 3.6002          |
| 3.5705        | 1.2051 | 600  | 3.5707          |
| 3.599         | 1.4062 | 700  | 3.5494          |
| 3.561         | 1.6073 | 800  | 3.5330          |
| 3.5462        | 1.8084 | 900  | 3.5205          |
| 3.4593        | 2.0080 | 1000 | 3.5115          |
| 3.4559        | 2.2092 | 1100 | 3.5051          |
| 3.4954        | 2.4103 | 1200 | 3.5013          |
| 3.5144        | 2.6114 | 1300 | 3.4983          |
| 3.5199        | 2.8125 | 1400 | 3.4964          |
| 3.3811        | 3.0121 | 1500 | 3.4949          |
| 3.3811        | 3.2132 | 1600 | 3.4937          |
| 3.4178        | 3.4143 | 1700 | 3.4923          |
| 3.4667        | 3.6154 | 1800 | 3.4913          |
| 3.4475        | 3.8165 | 1900 | 3.4907          |
| 3.3144        | 4.0161 | 2000 | 3.4896          |


### Framework versions

- Transformers 4.56.2
- Pytorch 2.5.1+cu121
- Datasets 3.5.1
- Tokenizers 0.22.1