VERSIL91/279d57a4-c523-440d-a91d-9912c2369465

See axolotl config

axolotl version: 0.4.1

adapter: lora
base_model: EleutherAI/pythia-70m-deduped
bf16: auto
chat_template: llama3
cosine_min_lr_ratio: 0.1
data_processes: 16
dataset_prepared_path: null
datasets:
- data_files:
  - 62fb9d7cd79067b3_train_data.json
  ds_type: json
  format: custom
  path: /workspace/input_data/62fb9d7cd79067b3_train_data.json
  type:
    field_input: gpt4_explanations
    field_instruction: context
    field_output: outcome
    format: '{instruction} {input}'
    no_input_format: '{instruction}'
    system_format: '{system}'
    system_prompt: ''
debug: null
deepspeed: null
device_map: '{'''':torch.cuda.current_device()}'
do_eval: true
early_stopping_patience: 1
eval_batch_size: 6
eval_sample_packing: false
eval_steps: 25
evaluation_strategy: steps
flash_attention: false
fp16: null
fsdp: null
fsdp_config: null
gradient_accumulation_steps: 5
gradient_checkpointing: true
group_by_length: true
hub_model_id: null
hub_repo: stevemonite
hub_strategy: checkpoint
hub_token: null
learning_rate: 0.0003
load_in_4bit: false
load_in_8bit: false
local_rank: null
logging_steps: 1
lora_alpha: 32
lora_dropout: 0.05
lora_fan_in_fan_out: null
lora_model_dir: null
lora_r: 16
lora_target_linear: true
lora_target_modules:
- q_proj
- v_proj
lr_scheduler: cosine
max_grad_norm: 1.0
max_memory:
  0: 70GiB
max_steps: 375
micro_batch_size: 6
mlflow_experiment_name: /tmp/62fb9d7cd79067b3_train_data.json
model_type: AutoModelForCausalLM
num_epochs: 2
optim_args:
  adam_beta1: 0.9
  adam_beta2: 0.95
  adam_epsilon: 1e-5
optimizer: adamw_torch
output_dir: miner_id_24
pad_to_sequence_len: true
resume_from_checkpoint: null
s2_attention: null
sample_packing: false
save_steps: 50
save_strategy: steps
sequence_len: 2048
special_tokens:
  pad_token: <|endoftext|>
strict: false
tf32: false
tokenizer_type: AutoTokenizer
torch_compile: false
train_on_inputs: false
trust_remote_code: true
val_set_size: 50
wandb_entity: sn56-miner
wandb_mode: disabled
wandb_name: null
wandb_project: god
wandb_run: c5ga
wandb_runid: null
warmup_raio: 0.03
warmup_ratio: 0.03
weight_decay: 0.01
xformers_attention: null

dfe18005-ad68-4425-a804-4ece094088cb

This model is a fine-tuned version of EleutherAI/pythia-70m-deduped on the None dataset. It achieves the following results on the evaluation set:

• Loss: 3.2653

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: 0.0003
• train_batch_size: 6
• eval_batch_size: 6
• seed: 42
• distributed_type: multi-GPU
• num_devices: 4
• gradient_accumulation_steps: 5
• total_train_batch_size: 120
• total_eval_batch_size: 24
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=adam_beta1=0.9,adam_beta2=0.95,adam_epsilon=1e-5
• lr_scheduler_type: cosine
• lr_scheduler_warmup_steps: 9
• training_steps: 304

Training results

Training Loss	Epoch	Step	Validation Loss
606.47	0.0066	1	128.0950
107.0225	0.1649	25	26.2796
61.9235	0.3298	50	10.9448
44.5495	0.4947	75	6.8106
24.4823	0.6596	100	5.5373
20.8706	0.8245	125	4.7062
18.9357	0.9894	150	4.2778
16.7825	1.1544	175	3.4199
15.083	1.3193	200	3.2062
16.1497	1.4842	225	3.2653

Framework versions

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