See axolotl config

axolotl version: 0.4.1

adapter: lora
base_model: aisingapore/llama3-8b-cpt-sea-lionv2.1-instruct
bf16: auto
chat_template: llama3
dataset_prepared_path: null
datasets:
- data_files:
  - 33f82a72ccefd5bf_train_data.json
  ds_type: json
  format: custom
  path: /workspace/input_data/33f82a72ccefd5bf_train_data.json
  type:
    field_instruction: prompt_source
    field_output: response_model
    format: '{instruction}'
    no_input_format: '{instruction}'
    system_format: '{system}'
    system_prompt: ''
debug: null
deepspeed: null
early_stopping_patience: null
eval_max_new_tokens: 128
eval_table_size: null
evals_per_epoch: 4
flash_attention: true
fp16: true
fsdp: null
fsdp_config: null
gradient_accumulation_steps: 4
gradient_checkpointing: true
group_by_length: false
hub_model_id: sn56/0cba69c7-4711-4b90-a2d9-02a22f069154
hub_repo: null
hub_strategy: checkpoint
hub_token: null
learning_rate: 0.0002
load_in_4bit: false
load_in_8bit: false
local_rank: null
logging_steps: 1
lora_alpha: 16
lora_dropout: 0.1
lora_fan_in_fan_out: null
lora_model_dir: null
lora_r: 8
lora_target_linear: true
lr_scheduler: cosine
max_memory:
  0: 70GiB
max_steps: 25
micro_batch_size: 1
mlflow_experiment_name: /tmp/33f82a72ccefd5bf_train_data.json
model_type: AutoModelForCausalLM
num_epochs: 3
optimizer: adamw_torch
output_dir: miner_id_24
pad_to_sequence_len: true
resume_from_checkpoint: null
s2_attention: null
sample_packing: false
saves_per_epoch: 3
sequence_len: 2028
strict: false
tf32: false
tokenizer_type: AutoTokenizer
train_on_inputs: false
trust_remote_code: true
val_set_size: 0.05
wandb_entity: sn56-miner
wandb_mode: disabled
wandb_name: 0cba69c7-4711-4b90-a2d9-02a22f069154
wandb_project: god
wandb_run: 8q1f
wandb_runid: 0cba69c7-4711-4b90-a2d9-02a22f069154
warmup_steps: 10
weight_decay: 0.0
xformers_attention: null

0cba69c7-4711-4b90-a2d9-02a22f069154

This model is a fine-tuned version of aisingapore/llama3-8b-cpt-sea-lionv2.1-instruct on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.9325

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: 1
• eval_batch_size: 1
• seed: 42
• distributed_type: multi-GPU
• num_devices: 4
• gradient_accumulation_steps: 4
• total_train_batch_size: 16
• total_eval_batch_size: 4
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: cosine
• lr_scheduler_warmup_steps: 10
• training_steps: 25

Training results

Training Loss	Epoch	Step	Validation Loss
7.4423	0.0000	1	9.1654
8.9115	0.0001	3	9.0982
8.2145	0.0002	6	8.0506
5.6599	0.0003	9	4.3642
2.5568	0.0004	12	2.2055
2.4366	0.0004	15	1.6779
0.8671	0.0005	18	1.1326
1.4519	0.0006	21	0.9818
1.3446	0.0007	24	0.9325

Framework versions

• PEFT 0.13.2
• Transformers 4.46.0
• Pytorch 2.5.0+cu124
• Datasets 3.0.1
• Tokenizers 0.20.1