Built with Axolotl

See axolotl config

axolotl version: 0.5.2

base_model: meta-llama/Llama-3.1-8B
model_type: AutoModelForCausalLM
tokenizer_type: AutoTokenizer
tokenizer_use_fast: false
resize_token_embeddings_to_32x: false

flash_attention: true
xformers_attention:

load_in_8bit: false
load_in_4bit: false
strict: false

datasets:
  - path: skymizer/Llama3.1-8B-base-tokenized-fineweb-edu-45B-4096
    train_on_split: train
    type: completion

test_datasets:
  - path: skymizer/Llama3.1-8B-base-tokenized-fineweb-edu-test-4K
    split: test
    type: completion

is_preprocess: true
skip_prepare_dataset: true

dataset_prepared_path: /mnt/home/model-team/datasets/pretokenized/Llama3.1-8B-base-tokenized-fineweb-edu-45B-4096 

hf_use_auth_token: true
output_dir: /mnt/home/model-team/models/Llama3.1-8B-v0.1-relu-stage-1-fineweb-edu-45B-4096
resume_from_checkpoint:
auto_resume_from_checkpoints: true

sequence_len: 4096
sample_packing: true
sample_packing_group_size: 100000
sample_packing_bin_size: 200
pad_to_sequence_len: true

eval_sample_packing: false
# eval_causal_lm_metrics: ["perplexity"]

wandb_project: "sparse-tuning-cpt"
wandb_entity:
wandb_watch:
wandb_name: "Llama3.1-8B-relu-stage-1-fineweb-edu-45B-4096"
wandb_log_model:

# global batch size = 2 * 8 * 8 GPUs * 8 Nodes * 4096 = 4M
gradient_accumulation_steps: 8
micro_batch_size: 2
  # eval_batch_size: 2
max_steps: 10000
optimizer: adamw_torch
learning_rate: 0.000015
lr_scheduler: cosine
cosine_min_lr_ratio: 1.0 
weight_decay: 0.0
adam_beta1: 0.9
adam_beta2: 0.95
adam_eps: 0.000001
max_grad_norm: 1.0

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

hub_model_id: "skymizer/Llama3.1-8B-relu-stage-1-fineweb-edu-45B-4096"

save_strategy: "steps"
save_steps: 500

gradient_checkpointing: true
gradient_checkpointing_kwargs:
  use_reentrant: false
early_stopping_patience:
resume_from_checkpoint:
local_rank:
logging_steps: 1

warmup_steps: 1
eval_steps: 500
eval_table_size:
debug:
deepspeed: /root/train/axolotl/deepspeed_configs/zero3_bf16.json
fsdp:
fsdp_config:
seed: 42

special_tokens:
  pad_token: "<|end_of_text|>"

Llama3.1-8B-relu-stage-1-fineweb-edu-45B-4096

This model is a fine-tuned version of meta-llama/Llama-3.1-8B on an unknown dataset. It achieves the following results on the evaluation set:

  • Loss: 1.9682

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: 2
  • eval_batch_size: 2
  • seed: 42
  • distributed_type: multi-GPU
  • num_devices: 64
  • gradient_accumulation_steps: 8
  • total_train_batch_size: 1024
  • total_eval_batch_size: 128
  • optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.95) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
  • lr_scheduler_type: cosine
  • lr_scheduler_warmup_steps: 2
  • training_steps: 10000

Training results

Training Loss Epoch Step Validation Loss
12.2232 0.0001 1 12.1487
2.2025 0.0424 500 2.2272
2.1454 0.0848 1000 2.1515
2.0991 0.1273 1500 2.1142
2.0604 0.1697 2000 2.0894
2.058 0.2121 2500 2.0711
2.0582 0.2545 3000 2.0561
2.0474 0.2969 3500 2.0442
2.0268 0.3394 4000 2.0347
2.0173 0.3818 4500 2.0256
1.9941 0.4242 5000 2.0178
2.0113 0.4666 5500 2.0106
1.9949 0.5091 6000 2.0040
2.0077 0.5515 6500 1.9984
1.986 0.5939 7000 1.9935
1.9902 0.6363 7500 1.9888
1.9899 0.6787 8000 1.9841
1.9729 0.7212 8500 1.9800
1.971 0.7636 9000 1.9759
1.9784 0.8060 9500 1.9718
1.9553 0.8484 10000 1.9682

Framework versions

  • Transformers 4.46.3
  • Pytorch 2.5.1+cu124
  • Datasets 3.1.0
  • Tokenizers 0.20.3
Downloads last month
168
Safetensors
Model size
8.03B params
Tensor type
BF16
·
Inference Examples
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.

Model tree for skymizer/Llama3.1-8B-relu-stage-1-fineweb-edu-45B-4096

Finetuned
(763)
this model
Quantizations
1 model