CoolFace
Modelpublic

jplhughes2/1a_meta-llama-Llama-3.1-405B-Instruct-fsdp

sourceHugging Facellama3.1updated 2y agoView on Hugging Face
1likes6downloads
Model Card

<!-- 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: 0.6.0

yaml
# This works!

base_model: meta-llama/Llama-3.1-405B-Instruct
hub_model_id: jplhughes2/1a_meta-llama-Llama-3.1-405B-Instruct-fsdp
load_in_8bit: false
load_in_4bit: true
adapter: qlora
wandb_name: 1a_meta-llama-Llama-3.1-405B-Instruct-fsdp
output_dir: ./outputs/out/1a_meta-llama-Llama-3.1-405B-Instruct-fsdp
# base_model:
# hub_model_id:
# load_in_8bit:
# load_in_4bit:
# adapter:
# wandb_name:
# output_dir:

tokenizer_type: AutoTokenizer
push_dataset_to_hub:
strict: false

datasets:
  - path: jplhughes2/docs_only_30k_filtered
    type: completion
    field: text
    split: train
dataset_prepared_path: last_run_prepared
# val_set_size: 0.05
test_datasets:
  - path: jplhughes2/docs_only_val_5k_filtered
    type: completion
    field: text
    split: train
save_safetensors: true

sequence_len: 1024
sample_packing: true
pad_to_sequence_len: true

lora_r: 64
lora_alpha: 128
lora_dropout: 0.05
lora_target_modules:
lora_target_linear: true

wandb_mode:
wandb_project: alignment-faking
wandb_entity: academicsnyuperez
wandb_watch:
wandb_run_id:
wandb_log_model:

gradient_accumulation_steps: 1
micro_batch_size: 4
num_epochs: 1
optimizer: adamw_torch_fused
lr_scheduler: cosine
learning_rate: 0.00002

train_on_inputs: false
group_by_length: false
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|>

</details><br>

1a_meta-llama-Llama-3.1-405B-Instruct-fsdp

This model is a fine-tuned version of meta-llama/Llama-3.1-405B-Instruct on the jplhughes2/docsonly30k_filtered dataset. It achieves the following results on the evaluation set:

  • —Loss: 0.5761

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
  • —trainbatchsize: 4
  • —evalbatchsize: 4
  • —seed: 42
  • —distributed_type: multi-GPU
  • —num_devices: 8
  • —totaltrainbatch_size: 32
  • —totalevalbatch_size: 32
  • —optimizer: Use OptimizerNames.ADAMWTORCHFUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
  • —lrschedulertype: cosine
  • —lrschedulerwarmup_steps: 10
  • —num_epochs: 1.0

Training results

Training LossEpochStepValidation Loss
1.3230.001611.3262
0.61470.33442040.6174
0.58360.66894080.5761

Framework versions

  • —PEFT 0.14.0
  • —Transformers 4.48.3
  • —Pytorch 2.4.1+cu124
  • —Datasets 3.2.0
  • —Tokenizers 0.21.0