README.md

metadata

license: apache-2.0
base_model: kaist-ai/mpa-Mistral-7b-v0.2-hf-sft-66k
tags:
  - axolotl
  - dpo
  - trl
  - dpo
  - generated_from_trainer
model-index:
  - name: mpa-Mistral-7b-v0.2-hf-66k-dpo-5e-7
    results: []

See axolotl config

axolotl version: 0.4.0

base_model: kaist-ai/mpa-Mistral-7b-v0.2-hf-sft-66k
model_type: MistralForCausalLM
tokenizer_type: LlamaTokenizer

load_in_8bit: false
load_in_4bit: false
strict: false

rl: dpo
datasets:
  - path: kaist-ai/mpa-train-dpo-66k
    type: chatml.argilla
    # conversation: mistral
      
dataset_prepared_path:
hub_model_id: kaist-ai/mpa-Mistral-7b-v0.2-hf-66k-dpo-5e-7
hub_strategy: checkpoint
# val_set_size: 0
output_dir: /mnt/nas/seongyun/axolotl/outputs/mpa_66k_dpo-5e-7

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

wandb_project: mpa
wandb_entity: seongyun
wandb_watch:
wandb_name: mpa_mistral-7b-v0.2-hf-66k-dpo-5e-7
wandb_log_model:

gradient_accumulation_steps: 4
micro_batch_size: 1
num_epochs: 2
optimizer: adamw_bnb_8bit
lr_scheduler: cosine
learning_rate: 0.0000005

train_on_inputs: false
group_by_length: false
bf16: auto
fp16:
tf32: false

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

warmup_steps: 10
# evals_per_epoch: 4
eval_table_size:
# eval_max_new_tokens: 128
saves_per_epoch: 1
debug:
deepspeed:
weight_decay: 0.0
fsdp:
fsdp_config:
special_tokens:

mpa-Mistral-7b-v0.2-hf-66k-dpo-5e-7

This model is a fine-tuned version of kaist-ai/mpa-Mistral-7b-v0.2-hf-sft-66k 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: 5e-07
• train_batch_size: 1
• eval_batch_size: 8
• seed: 42
• distributed_type: multi-GPU
• num_devices: 4
• gradient_accumulation_steps: 4
• total_train_batch_size: 16
• total_eval_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_steps: 10
• training_steps: 8143

Training results

Framework versions

• Transformers 4.40.0.dev0
• Pytorch 2.1.1
• Datasets 2.15.0
• Tokenizers 0.15.0