esunn/suhyup-test-openllama-lora

See axolotl config

axolotl version: 0.4.0

base_model: openlm-research/open_llama_3b_v2
model_type: LlamaForCausalLM
tokenizer_type: LlamaTokenizer
load_in_8bit: false
load_in_4bit: true
strict: false
push_dataset_to_hub:
datasets:
  - path: "./raft_oracle_context_alpaca.json"
    type: alpaca
dataset_prepared_path: ./dataset-pre
val_set_size: 0.02
adapter: lora
lora_model_dir:
sequence_len: 1024
sample_packing: true
lora_r: 8
lora_alpha: 16
lora_dropout: 0.0
lora_target_modules:
  - gate_proj
  - down_proj
  - up_proj
  - q_proj
  - v_proj
  - k_proj
  - o_proj
lora_fan_in_fan_out:
wandb_project:
wandb_entity:
wandb_watch:
wandb_name:
wandb_log_model:
output_dir: ./outputs/lora-out
gradient_accumulation_steps: 1
micro_batch_size: 2
num_epochs: 4
optimizer: adamw_bnb_8bit
torchdistx_path:
lr_scheduler: cosine
learning_rate: 0.0002
train_on_inputs: false
group_by_length: false
bf16: false
fp16: true
tf32: false
gradient_checkpointing: true
early_stopping_patience:
resume_from_checkpoint:
local_rank:
logging_steps: 1
xformers_attention:
flash_attention: true
gptq_groupsize:
s2_attention:
gptq_model_v1:
warmup_steps: 20
evals_per_epoch: 4
saves_per_epoch: 1
debug:
deepspeed:
weight_decay: 0.1
fsdp:
fsdp_config:
special_tokens:
  bos_token: "<s>"
  eos_token: "</s>"
  unk_token: "<unk>"

outputs/lora-out

This model is a fine-tuned version of openlm-research/open_llama_3b_v2 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.4426

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: 2
• eval_batch_size: 2
• seed: 42
• distributed_type: multi-GPU
• num_devices: 7
• total_train_batch_size: 14
• total_eval_batch_size: 14
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_steps: 20
• num_epochs: 4
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
1.9628	0.0022	1	1.9150
0.5816	0.2505	115	0.6157
0.3604	0.5011	230	0.4307
0.2598	0.7516	345	0.3558
0.2227	1.0022	460	0.3434
0.1381	1.2266	575	0.3376
0.0718	1.4771	690	0.3372
0.0684	1.7277	805	0.3608
0.0817	1.9782	920	0.3663
0.0315	2.2004	1035	0.3888
0.0331	2.4510	1150	0.4003
0.0222	2.7015	1265	0.4145
0.0222	2.9521	1380	0.4216
0.0166	3.1743	1495	0.4330
0.017	3.4248	1610	0.4391
0.0142	3.6754	1725	0.4426

Framework versions

• PEFT 0.10.0
• Transformers 4.40.2
• Pytorch 2.1.2+cu118
• Datasets 2.19.1
• Tokenizers 0.19.1