zagreus-test-202000-sft-12

<!-- 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"/>

<details><summary>See axolotl config</summary>

axolotl version: 0.9.2

   # ============= 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



# --- Cosine knobs (Axolotl) ---
# 1) Mantieni LR costante al massimo per l'80% degli step
cosine_constant_lr_ratio: 0.8
# 2) Floor al 30% del LR max (min_lr = 4.5e-6)
cosine_min_lr_ratio: 0.3
# Non impostare lr_div_factor quando usi cosine_min_lr_ratio


optimizer: adamw_torch_fused
lr_scheduler: cosine   # <-- 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: 3000                      # ≈ 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: 500
save_total_limit: 3
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 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 3.5482

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: cosine
• lr_scheduler_warmup_steps: 24
• training_steps: 3000

Training results

Framework versions

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