base_model: manishiitg/open-aditi-hi-v1
model_type: MistralForCausalLM
tokenizer_type: LlamaTokenizer
is_mistral_derived_model: true

load_in_8bit: false
load_in_4bit: true
strict: false

rl: true
datasets:
  - path: manishiitg/argilla-ultrafeedback-binarized-preferences-cleaned
    split: train
    type: ultra_apply_chatml
  - path: manishiitg/unalignment-toxic-dpo-v0.1
    split: train
    type: apply_chatml

dataset_prepared_path: last_run_prepared
val_set_size: 0.0
output_dir: /sky-notebook/manishiitg/open-aditi-hi-v1-dpo

hub_model_id: manishiitg/open-aditi-hi-v1-dpo
hf_use_auth_token: true

wandb_project: open-aditi-hi-v1-dpo

save_safetensors: true


adapter: qlora
lora_model_dir:

sequence_len: 4096
sample_packing: true
pad_to_sequence_len: false

lora_r: 16
lora_alpha: 32
lora_dropout: 0.05
lora_target_linear: true
lora_fan_in_fan_out:
lora_target_modules:
  - gate_proj
  - down_proj
  - up_proj
  - q_proj
  - v_proj
  - k_proj
  - o_proj

lora_modules_to_save:
 - embed_tokens
 - lm_head

wandb_entity:
wandb_watch:
wandb_run_id:
wandb_log_model:

gradient_accumulation_steps: 4
micro_batch_size: 3
num_epochs: 4
optimizer: adamw_bnb_8bit
lr_scheduler: cosine
learning_rate: 0.0002

adam_beta2: 0.95
adam_epsilon: 0.00001
max_grad_norm: 1.0

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


gradient_checkpointing: true
early_stopping_patience:
resume_from_checkpoint:
auto_resume_from_checkpoints: true ## manage check point resume from here
local_rank:
logging_steps: 1
xformers_attention:
flash_attention: true

loss_watchdog_threshold: 5.0
loss_watchdog_patience: 3

warmup_steps: 10
eval_steps: 0
evals_per_epoch: 0
eval_table_size:
eval_table_max_new_tokens: 128
save_steps: 100 ## increase based on your dataset
save_strategy: steps
debug:
deepspeed:
weight_decay: 0.0
fsdp:
fsdp_config:
special_tokens:
  bos_token: "<s>"
  eos_token: "</s>"
  unk_token: "<unk>"
tokens: # these are delimiters
  - "<|im_start|>"
  - "<|im_end|>"

open-aditi-hi-v1-dpo

This model is a fine-tuned version of manishiitg/open-aditi-hi-v1 on the None dataset.

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

The following bitsandbytes quantization config was used during training:

quant_method: bitsandbytes
load_in_8bit: False
load_in_4bit: True
llm_int8_threshold: 6.0
llm_int8_skip_modules: None
llm_int8_enable_fp32_cpu_offload: False
llm_int8_has_fp16_weight: False
bnb_4bit_quant_type: nf4
bnb_4bit_use_double_quant: True
bnb_4bit_compute_dtype: bfloat16

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 0.0002
train_batch_size: 3
eval_batch_size: 8
seed: 42
gradient_accumulation_steps: 4
total_train_batch_size: 12
optimizer: Adam with betas=(0.9,0.95) and epsilon=1e-05
lr_scheduler_type: linear
lr_scheduler_warmup_steps: 10
training_steps: 6964

Training results

Framework versions

PEFT 0.7.0
Transformers 4.37.0.dev0
Pytorch 2.1.1+cu121
Datasets 2.16.1
Tokenizers 0.15.0