You can deploy this model and start earning money today!

Alphatao/
826671fb-74bf-4adf-893c-6be73c4cb670

Name: 826671fb-74bf-4adf-893c-6be73c4cb670
Availability: InStock

generated_from_trainer

JackFram/llama-160m

adapter

apache-2.0

Views

No views yet

See axolotl config

axolotl version: 0.4.1

yaml

adapter: lora
base_model: JackFram/llama-160m
bf16: true
chat_template: llama3
dataset_prepared_path: null
datasets:
- data_files:
  - 74abca5b35871829_train_data.json
  ds_type: json
  format: custom
  path: /workspace/input_data/74abca5b35871829_train_data.json
  type:
    field_instruction: chat
    field_output: system
    format: '{instruction}'
    no_input_format: '{instruction}'
    system_format: '{system}'
    system_prompt: ''
debug: null
device_map:
  ? ''
  : 0,1,2,3,4,5,6,7
early_stopping_patience: 2
eval_max_new_tokens: 128
eval_steps: 100
eval_table_size: null
flash_attention: true
gradient_accumulation_steps: 8
gradient_checkpointing: true
group_by_length: false
hub_model_id: Alphatao/826671fb-74bf-4adf-893c-6be73c4cb670
hub_repo: null
hub_strategy: null
hub_token: null
learning_rate: 0.0002
load_best_model_at_end: true
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
- k_proj
- v_proj
- o_proj
- down_proj
- up_proj
lr_scheduler: cosine
max_grad_norm: 1.0
max_steps: 12977
micro_batch_size: 4
mlflow_experiment_name: /tmp/74abca5b35871829_train_data.json
model_type: AutoModelForCausalLM
num_epochs: 2
optimizer: adamw_bnb_8bit
output_dir: miner_id_24
pad_to_sequence_len: true
resume_from_checkpoint: null
s2_attention: null
sample_packing: false
save_steps: 100
sequence_len: 2048
special_tokens:
  pad_token: </s>
strict: false
tf32: true
tokenizer_type: AutoTokenizer
train_on_inputs: false
trust_remote_code: true
val_set_size: 0.04457917261055635
wandb_entity: null
wandb_mode: online
wandb_name: 029def98-2bf0-4673-93ae-3a488c6daa7a
wandb_project: Gradients-On-Demand
wandb_run: your_name
wandb_runid: 029def98-2bf0-4673-93ae-3a488c6daa7a
warmup_steps: 10
weight_decay: 0.0
xformers_attention: null

826671fb-74bf-4adf-893c-6be73c4cb670

This model is a fine-tuned version of JackFram/llama-160m on the None dataset. It achieves the following results on the evaluation set:

Loss: 0.0007

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: 4
eval_batch_size: 4
seed: 42
gradient_accumulation_steps: 8
total_train_batch_size: 32
optimizer: Use OptimizerNames.ADAMW_BNB 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: 6698

Training results

Training Loss	Epoch	Step	Validation Loss
2.5646	0.0003	1	2.5999
0.0154	0.0299	100	0.0136
0.0037	0.0597	200	0.0048
0.0026	0.0896	300	0.0036
0.0022	0.1194	400	0.0021
0.0017	0.1493	500	0.0023
0.0022	0.1792	600	0.0010
0.0008	0.2090	700	0.0019
0.0003	0.2389	800	0.0008
0.0006	0.2688	900	0.0008
0.0002	0.2986	1000	0.0008
0.0002	0.3285	1100	0.0007
0.0004	0.3583	1200	0.0008
0.0005	0.3882	1300	0.0007

Framework versions

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