#Mistral-7b-Instruct-v0.2
base_model: mistralai/Mistral-7B-Instruct-v0.2
model_type: MistralForCausalLM
tokenizer_type: LlamaTokenizer
is_mistral_derived_model: true

load_in_8bit: true
load_in_4bit: false
strict: false

datasets:
  - path: menna/tutor-synthetic
    type: completion
    field: text
dataset_prepared_path: last_run_prepared
hub_model_id: menna/mistral-synthetic-tutor
val_set_size: 0.05
output_dir: ./out

adapter: lora
lora_r: 8
lora_alpha: 16
lora_dropout: 0.05
lora_target_modules:
  - q_proj
  - v_proj

sequence_len: 512
sample_packing: false
pad_to_sequence_len: true

wandb_project: mistral-synthetic-tutor
wandb_entity:
wandb_watch:
wandb_name:
wandb_log_model: checkpoint

#only 2 epochs because of small dataset
gradient_accumulation_steps: 3
micro_batch_size: 2
num_epochs: 3
optimizer: adamw_bnb_8bit
lr_scheduler: cosine
learning_rate: 0.0002

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

chat_template: inst

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_table_max_new_tokens: 128
saves_per_epoch: 1
debug:
deepspeed:
weight_decay: 0.0
fsdp:
fsdp_config:
special_tokens:
  bos_token: "<s>"
  eos_token: "</s>"
  unk_token: "<unk:x
  :x