See axolotl config

axolotl version: 0.4.1

adapter: qlora
auto_resume_from_checkpoints: true
base_model: bigscience/bloom-560m
bf16: auto
chat_template: llama3
dataloader_num_workers: 12
dataset_prepared_path: null
datasets:
- data_files:
  - 084d4537100cb186_train_data.json
  ds_type: json
  format: custom
  path: /workspace/input_data/084d4537100cb186_train_data.json
  type:
    field_input: langpair
    field_instruction: source
    field_output: good-translation
    format: '{instruction} {input}'
    no_input_format: '{instruction}'
    system_format: '{system}'
    system_prompt: ''
debug: null
deepspeed: null
early_stopping_patience: 3
eval_max_new_tokens: 128
eval_steps: 100
eval_table_size: null
evals_per_epoch: null
flash_attention: false
fp16: false
fsdp: null
fsdp_config: null
gradient_accumulation_steps: 4
gradient_checkpointing: false
group_by_length: true
hub_model_id: error577/c9014b2e-4cf8-4380-a7b9-319dcd8bb12a
hub_repo: null
hub_strategy: checkpoint
hub_token: null
learning_rate: 0.0002
load_in_4bit: true
load_in_8bit: false
local_rank: null
logging_steps: 10
lora_alpha: 128
lora_dropout: 0.05
lora_fan_in_fan_out: null
lora_model_dir: null
lora_r: 64
lora_target_linear: true
lr_scheduler: cosine
max_grad_norm: 1.0
max_steps: null
micro_batch_size: 2
mlflow_experiment_name: /tmp/084d4537100cb186_train_data.json
model_type: AutoModelForCausalLM
num_epochs: 3
optimizer: adamw_torch_4bit
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: 512
strict: false
tf32: false
tokenizer_type: AutoTokenizer
train_on_inputs: false
trust_remote_code: true
val_set_size: 0.005
wandb_entity: null
wandb_mode: online
wandb_name: 15f4c23f-d52e-4354-806b-0c5da0374350
wandb_project: Gradients-On-Demand
wandb_run: your_name
wandb_runid: 15f4c23f-d52e-4354-806b-0c5da0374350
warmup_steps: 30
weight_decay: 0.0
xformers_attention: null

c9014b2e-4cf8-4380-a7b9-319dcd8bb12a

This model is a fine-tuned version of bigscience/bloom-560m on the None dataset. It achieves the following results on the evaluation set:

• Loss: 1.5994

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
• gradient_accumulation_steps: 4
• total_train_batch_size: 8
• optimizer: Use OptimizerNames.ADAMW_TORCH_4BIT 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: 30
• num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss
No log	0.0002	1	4.4030
6.5189	0.0222	100	2.6937
7.8292	0.0444	200	2.4313
5.3937	0.0666	300	2.3118
7.8199	0.0888	400	2.2125
5.4877	0.1110	500	2.1706
4.9892	0.1332	600	2.0680
5.6608	0.1554	700	2.0517
4.671	0.1776	800	2.0242
6.3505	0.1998	900	1.8601
4.1893	0.2220	1000	1.8662
4.8748	0.2441	1100	1.8771
5.0172	0.2663	1200	1.8052
5.2344	0.2885	1300	1.8127
4.4286	0.3107	1400	1.7529
4.4394	0.3329	1500	1.7735
4.4868	0.3551	1600	1.6954
4.6618	0.3773	1700	1.7275
4.6153	0.3995	1800	1.6793
5.0591	0.4217	1900	1.7122
3.6119	0.4439	2000	1.6413
3.9039	0.4661	2100	1.6450
4.1736	0.4883	2200	1.6268
4.7935	0.5105	2300	1.6143
4.0114	0.5327	2400	1.5969
3.9156	0.5549	2500	1.5649
4.5978	0.5771	2600	1.5713
3.7863	0.5993	2700	1.5953
3.469	0.6215	2800	1.5994

Framework versions

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