Hugging Face's logo

Undi95
/
Llama-3-Unholy-8B-e4

Text Generation
Transformers
PyTorch
llama
Not-For-All-Audiences
nsfw
conversational
text-generation-inference
Inference Endpoints
Model card Files Community
2
Llama-3-Unholy-8B-e4 / README.md
Undi95's picture
Undi95
Upload folder using huggingface_hub
598d422 verified
preview code
|
raw
history blame
No virus
2.67 kB
metadata
license: other
base_model: meta-llama/Meta-Llama-3-8B-Instruct
tags:
  - generated_from_trainer
model-index:
  - name: out
    results: []

Built with Axolotl

See axolotl config

axolotl version: 0.4.0

base_model: meta-llama/Meta-Llama-3-8B-Instruct
model_type: AutoModelForCausalLM
tokenizer_type: AutoTokenizer

load_in_8bit: false
load_in_4bit: false
strict: false

datasets:
  - path: ./datasets/ToxicQAFinal.parquet
    type: sharegpt
    conversation: llama-3
  - path: ./datasets/toxicsharegpt-NoWarning.jsonl
    type: sharegpt
    conversation: llama-3

dataset_prepared_path: last_run_prepared
val_set_size: 0.0
output_dir: ./out

sequence_len: 8192
sample_packing: true
pad_to_sequence_len: true

wandb_project: Uncensored-8B-ChatML
wandb_entity:
wandb_watch:
wandb_name:
wandb_log_model:

gradient_accumulation_steps: 2
micro_batch_size: 1
num_epochs: 4
optimizer: adamw_bnb_8bit
lr_scheduler: cosine
learning_rate: 3e-6

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

gradient_checkpointing: true
early_stopping_patience:
resume_from_checkpoint:
local_rank:
logging_steps: 1
xformers_attention:
flash_attention: true

warmup_steps: 10
saves_per_epoch: 2
debug:
deepspeed:
weight_decay: 0.1
fsdp:
fsdp_config:
special_tokens:
  pad_token: "<|end_of_text|>"

out

This model is a fine-tuned version of meta-llama/Meta-Llama-3-8B-Instruct on the None dataset.

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: 3e-06
  • train_batch_size: 1
  • eval_batch_size: 1
  • seed: 42
  • distributed_type: multi-GPU
  • num_devices: 2
  • gradient_accumulation_steps: 2
  • total_train_batch_size: 4
  • total_eval_batch_size: 2
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: cosine
  • lr_scheduler_warmup_steps: 10
  • num_epochs: 4

Training results

Framework versions

  • Transformers 4.40.0.dev0
  • Pytorch 2.0.1+cu118
  • Datasets 2.15.0
  • Tokenizers 0.15.0