base_model: Qwen/Qwen1.5-32B
model_type: AutoModelForCausalLM
tokenizer_type: AutoTokenizer

trust_remote_code: true

load_in_8bit: false
load_in_4bit: false
strict: false


datasets:
  - path: u-acc/mimi
    type: sharegpt
    conversation: chatml
  - path: u-acc/sonnetorcasubset
    type: sharegpt
    conversation: chatml
  - path: u-acc/claude_writing
    type: sharegpt
    conversation: chatml
  - path: kalomaze/Opus_Instruct_3k
    type: sharegpt
    conversation: chatml
  - path: kalomaze/Opus_Instruct_25k
    type: sharegpt
    conversation: chatml
chat_template: chatml
dataset_prepared_path:
val_set_size: 0.0
output_dir: ./magnum-32b-v1
default_system_message: You are an assistant that responds to the user.
sequence_len: 8192
sample_packing: true
eval_sample_packing: false
pad_to_sequence_len: false

adapter:
lora_model_dir:
lora_r:
lora_alpha:
lora_dropout:
lora_target_linear:
lora_fan_in_fan_out:

wandb_project: magnum-32b-v2
wandb_entity:
wandb_watch:
wandb_name: attempt-2
wandb_log_model:

gradient_accumulation_steps: 4
micro_batch_size: 1
num_epochs: 2
optimizer: paged_adamw_8bit
lr_scheduler: cosine
learning_rate: 0.00001

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

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

warmup_steps: 30
evals_per_epoch:
eval_table_size:
eval_max_new_tokens:
saves_per_epoch: 2
debug:
deepspeed: deepspeed_configs/zero3_bf16.json
weight_decay: 0.0
fsdp:
fsdp_config:
special_tokens: