base_model: meta-llama/Llama-3.1-8B-Instruct

plugins:
  # - axolotl.integrations.liger.LigerPlugin
liger_rope: true
liger_rms_norm: true
liger_swiglu: true
liger_fused_linear_cross_entropy: true

strict: false

chat_template: llama3
datasets:
  - path: AlekseyKorshuk/rewriter-v0.3-axolotl
    type: chat_template
    field_messages: messages
    message_field_role: role
    message_field_content: content
    roles:
      user:
        - user
      assistant:
        - assistant

val_set_size: 0.05
output_dir: ./outputs/out

eval_table_size: 0
eval_max_new_tokens: 256

sequence_len: 2048
sample_packing: false
eval_sample_packing: false
pad_to_sequence_len: false

wandb_project: ai-seo-rewriter
wandb_entity:
wandb_watch:
wandb_name: rewriter-v0.3-8b
wandb_log_model:

gradient_accumulation_steps: 4
micro_batch_size: 8
eval_batch_size: 8
num_epochs: 1
optimizer: adamw_torch
lr_scheduler: cosine
learning_rate: 7.5e-6

train_on_inputs: false
group_by_length: false
bf16: true
fp16:
tf32: false

gradient_checkpointing: true
gradient_checkpointing_kwargs:
  use_reentrant: false
early_stopping_patience:
resume_from_checkpoint:
logging_steps: 1
xformers_attention:
flash_attention: true

warmup_ratio: 0.1
evals_per_epoch: 2
saves_per_epoch: 1
debug:
deepspeed:
weight_decay: 0.0
fsdp:
  - full_shard
  - auto_wrap
fsdp_config:
  fsdp_limit_all_gathers: true
  fsdp_sync_module_states: true
  fsdp_offload_params: true
  fsdp_use_orig_params: false
  fsdp_cpu_ram_efficient_loading: true
  fsdp_auto_wrap_policy: TRANSFORMER_BASED_WRAP
  fsdp_transformer_layer_cls_to_wrap: LlamaDecoderLayer
  fsdp_state_dict_type: FULL_STATE_DICT
  fsdp_sharding_strategy: FULL_SHARD
  fsdp_backward_prefetch: BACKWARD_PRE
special_tokens:
   pad_token: <|end_of_text|>


hub_model_id: AlekseyKorshuk/rewriter-v0.3-axolotl-8b

rewriter-v0.3-axolotl-8b

This model is a fine-tuned version of meta-llama/Llama-3.1-8B-Instruct on the None dataset. It achieves the following results on the evaluation set:

Loss: 1.3373

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: 7.5e-06
train_batch_size: 8
eval_batch_size: 8
seed: 42
distributed_type: multi-GPU
num_devices: 8
gradient_accumulation_steps: 4
total_train_batch_size: 256
total_eval_batch_size: 64
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: cosine
lr_scheduler_warmup_steps: 8
num_epochs: 1

Training results

Training Loss	Epoch	Step	Validation Loss
2.345	0.0112	1	2.4606
1.2621	0.5056	45	1.3373

Framework versions

Transformers 4.45.2
Pytorch 2.4.0+cu121
Datasets 3.0.1
Tokenizers 0.20.1