YoungPanda/SFT_Test1

See axolotl config

axolotl version: 0.4.0

base_model: 152334H/miqu-1-70b-sf
model_type: AutoModelForCausalLM
tokenizer_type: LlamaTokenizer
trust_remote_code: true

load_in_8bit: false
load_in_4bit: false
strict: false

datasets:
  - path: Drewskidang/rag
    type: alpaca
    prompt_style: chatml
  - path: Drewskidang/output
    type: alpaca
    prompt_style: chatml
  - path: Drewskidang/Instruct
    type: alpaca
    prompt_style: chatml
  - path: Drewskidang/Preference
    type: alpaca
    prompt_style: chatml
  - path: lighteval/legal_summarization
    name: BillSum
    type: summarizetldr

  
dataset_prepared_path: last_run_prepared
val_set_size: 0.0
output_dir: ./qlora-out

## You can optionally freeze the entire model and unfreeze a subset of parameters
unfrozen_parameters:
  - lm_head.*
  #- model.embed_tokens.*
#- model.layers.2[0-9]+.block_sparse_moe.gate.*
#  - model.layers.2[0-9]+.block_sparse_moe.experts.*
#  - model.layers.3[0-9]+.block_sparse_moe.gate.*
#  - model.layers.3[0-9]+.block_sparse_moe.experts.*

model_config:
  output_router_logits: true



sequence_len: 4096
sample_packing: true
pad_to_sequence_len: true

wandb_project: Leak_Mistral
wandb_entity:
wandb_watch:
wandb_name:
wandb_log_model:

gradient_accumulation_steps: 2
micro_batch_size: 10
num_epochs: 2
optimizer: adamw_bnb_8bit
lr_scheduler: cosine
learning_rate: 0.0002

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_table_max_new_tokens: 128
saves_per_epoch: 1
debug:
deepspeed: 
weight_decay: 0.0
fsdp:
fsdp_config:
special_tokens:
  eos_token: "<|im_end|>"
tokens:
  - "<|im_start|>"
trust_remote_code: true

qlora-out

This model is a fine-tuned version of 152334H/miqu-1-70b-sf 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: 0.0002
• train_batch_size: 10
• eval_batch_size: 10
• seed: 42
• distributed_type: multi-GPU
• num_devices: 8
• gradient_accumulation_steps: 2
• total_train_batch_size: 160
• total_eval_batch_size: 80
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_steps: 10
• num_epochs: 2

Training results

Framework versions

• Transformers 4.38.0.dev0
• Pytorch 2.1.1+cu121
• Datasets 2.16.1
• Tokenizers 0.15.0