base_model: huihui-ai/Llama-3.3-70B-Instruct-abliterated-finetuned
load_in_8bit: false
load_in_4bit: true
adapter: qlora
wandb_name: Llama-3.3-70B-Instruct-abliterated-finetuned-chem-claude-5-comp3-sort-pate-2500
output_dir: ./outputs/out/Llama-3.3-70B-Instruct-abliterated-finetuned-chem-claude-5-comp3-sort-pate-2500
hub_model_id: cgifbribcgfbi/Llama-3.3-70B-Instruct-abliterated-finetuned-chem-claude-5-comp3-sort-pate-2500

tokenizer_type: AutoTokenizer
push_dataset_to_hub:
strict: false

datasets:
  - path: dset_comp3.0_sortpatent_count_pat200_in5_num10134_2500.jsonl
    type: chat_template
    field_messages: messages

dataset_prepared_path: last_run_prepared
val_set_size: 0.04
save_safetensors: true

sequence_len: 2110
sample_packing: true
pad_to_sequence_len: true

lora_r: 64
lora_alpha: 32
lora_dropout: 0.05
lora_target_modules:
lora_target_linear: true

wandb_mode:
wandb_project: finetune-sweep
wandb_entity: gpoisjgqetpadsfke
wandb_watch:
wandb_run_id:
wandb_log_model:

gradient_accumulation_steps: 1
micro_batch_size: 2  # This will be automatically adjusted based on available GPU memory
num_epochs: 4
optimizer: adamw_torch_fused
lr_scheduler: cosine
learning_rate: 0.00002

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

gradient_checkpointing: true
gradient_checkpointing_kwargs:
  use_reentrant: true
logging_steps: 1
flash_attention: true

warmup_steps: 10
evals_per_epoch: 3
saves_per_epoch: 1
weight_decay: 0.01
fsdp:
  - full_shard
  - auto_wrap
fsdp_config:
  fsdp_limit_all_gathers: true
  fsdp_sync_module_states: true
  fsdp_offload_params: false
  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
special_tokens:
  pad_token: <|finetune_right_pad_id|>

Llama-3.3-70B-Instruct-abliterated-finetuned-chem-claude-5-comp3-sort-pate-2500

This model is a fine-tuned version of huihui-ai/Llama-3.3-70B-Instruct-abliterated-finetuned on the dset_comp3.0_sortpatent_count_pat200_in5_num10134_2500.jsonl dataset. It achieves the following results on the evaluation set:

Loss: 0.3638

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: 2e-05
train_batch_size: 2
eval_batch_size: 2
seed: 42
distributed_type: multi-GPU
num_devices: 8
total_train_batch_size: 16
total_eval_batch_size: 16
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: 10
num_epochs: 4.0

Training results

Training Loss	Epoch	Step	Validation Loss
0.6879	0.0088	1	0.6581
0.4988	0.3333	38	0.4895
0.4239	0.6667	76	0.4282
0.4554	1.0	114	0.4046
0.3848	1.3333	152	0.3902
0.3652	1.6667	190	0.3814
0.4143	2.0	228	0.3748
0.351	2.3333	266	0.3701
0.3392	2.6667	304	0.3672
0.3936	3.0	342	0.3654
0.3354	3.3333	380	0.3642
0.328	3.6667	418	0.3638
0.3875	4.0	456	0.3638

Framework versions

PEFT 0.15.2
Transformers 4.51.3
Pytorch 2.6.0+cu124
Datasets 3.5.1
Tokenizers 0.21.1