You can deploy this model and start earning money today!

CheapsetZero/
f27ce3b1-76bb-4e49-b8e6-b83dc38d6fba

Name: f27ce3b1-76bb-4e49-b8e6-b83dc38d6fba
Availability: InStock

generated_from_trainer

UCLA-AGI/Gemma-2-9B-It-SPPO-Iter2

adapter

gemma

Views

No views yet

See axolotl config

axolotl version: 0.4.1

yaml

adam_epsilon: 1e-6
adapter: lora
base_model: UCLA-AGI/Gemma-2-9B-It-SPPO-Iter2
bf16: true
chat_template: llama3
dataloader_num_workers: 0
dataloader_pin_memory: false
dataset_prepared_path: null
datasets:
- data_files:
  - 11a367bcb19b301a_train_data.json
  ds_type: json
  field: prompt
  path: /workspace/input_data/
  split: train
  type: completion
debug: null
eval_max_new_tokens: 128
eval_table_size: null
evals_per_epoch: 2
flash_attention: false
fp16: false
gradient_accumulation_steps: 4
gradient_checkpointing: true
gradient_clipping: 1.0
group_by_length: false
hub_model_id: CheapsetZero/f27ce3b1-76bb-4e49-b8e6-b83dc38d6fba
hub_strategy: checkpoint
learning_rate: 0.0002
load_in_4bit: false
load_in_8bit: false
local_rank: null
logging_steps: 10
lora_alpha: 64
lora_dropout: 0.05
lora_r: 32
lora_target_linear: true
lr_scheduler: constant_with_warmup
max_grad_norm: 1.0
max_steps: 4000
micro_batch_size: 16
mlflow_experiment_name: /tmp/11a367bcb19b301a_train_data.json
model_type: AutoModelForCausalLM
optimizer: adamw_torch
output_dir: miner_id_24
pad_to_sequence_len: true
sample_packing: false
save_safetensors: true
save_total_limit: 3
saves_per_epoch: 2
sequence_len: 512
strict: false
tf32: true
tokenizer_type: AutoTokenizer
torch_compile: false
train_on_inputs: false
trl:
  advantage_clip: 10.0
  beta: 0.1
  clip_rewards: true
  kl_penalty: kl
  max_completion_length: 256
  normalize_rewards: true
  num_generations: 8
  reward_clip_value: 10.0
  reward_funcs:
  - rewards_b0e2f786-9849-4865-a727-066ae6471139.reward_short_sentences
  - rewards_b0e2f786-9849-4865-a727-066ae6471139.reward_short_completions
  - rewards_b0e2f786-9849-4865-a727-066ae6471139.reward_low_difficult_words_percentage
  reward_weights:
  - 6.039041931213902
  - 1.4833111806701427
  - 0.4380501802438408
  temperature: 1.0
  use_vllm: false
trust_remote_code: true
val_set_size: 0.05
wandb_mode: online
wandb_name: b0e2f786-9849-4865-a727-066ae6471139
wandb_project: Gradients-On-Demand
wandb_run: your_name
wandb_runid: b0e2f786-9849-4865-a727-066ae6471139
warmup_steps: 100
weight_decay: 0.01
xformers_attention: false

f27ce3b1-76bb-4e49-b8e6-b83dc38d6fba

This model is a fine-tuned version of UCLA-AGI/Gemma-2-9B-It-SPPO-Iter2 on the None dataset. It achieves the following results on the evaluation set:

Loss: 2.2655

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: 16
eval_batch_size: 16
seed: 42
gradient_accumulation_steps: 4
total_train_batch_size: 64
optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-06 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: constant_with_warmup
lr_scheduler_warmup_steps: 100
training_steps: 110

Training results

Training Loss	Epoch	Step	Validation Loss
No log	0.0091	1	6.2043
2.6035	0.5011	55	2.5207
2.3603	1.0023	110	2.2655

Framework versions

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