See axolotl config
axolotl version: 0.4.1
adapter: lora
base_model: Qwen/Qwen1.5-0.5B
bf16: true
chat_template: llama3
dataset_prepared_path: null
datasets:
- data_files:
- 08def8d176b3ee4d_train_data.json
ds_type: json
format: custom
path: /workspace/input_data/08def8d176b3ee4d_train_data.json
type:
field_input: tools
field_instruction: query
field_output: answers
format: '{instruction} {input}'
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/a30c98f1-5377-4fe4-bf1b-66b3d754b9b2
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
lr_scheduler: cosine
max_grad_norm: 1.0
max_steps: 4224
micro_batch_size: 4
mlflow_experiment_name: /tmp/08def8d176b3ee4d_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: 1024
strict: false
tf32: true
tokenizer_type: AutoTokenizer
train_on_inputs: false
trust_remote_code: true
val_set_size: 0.04
wandb_entity: null
wandb_mode: online
wandb_name: 7a7bc575-869c-40c2-9816-fca44c877c0b
wandb_project: Gradients-On-Demand
wandb_run: your_name
wandb_runid: 7a7bc575-869c-40c2-9816-fca44c877c0b
warmup_steps: 10
weight_decay: 0.0
xformers_attention: null
a30c98f1-5377-4fe4-bf1b-66b3d754b9b2
This model is a fine-tuned version of Qwen/Qwen1.5-0.5B on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.0594
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: 2917
Training results
Training Loss | Epoch | Step | Validation Loss |
---|---|---|---|
0.9008 | 0.0007 | 1 | 0.9272 |
0.1566 | 0.0686 | 100 | 0.0968 |
0.0716 | 0.1372 | 200 | 0.0869 |
0.081 | 0.2057 | 300 | 0.0793 |
0.1297 | 0.2743 | 400 | 0.0786 |
0.0308 | 0.3429 | 500 | 0.0740 |
0.075 | 0.4115 | 600 | 0.0740 |
0.0821 | 0.4801 | 700 | 0.0723 |
0.0985 | 0.5486 | 800 | 0.0694 |
0.051 | 0.6172 | 900 | 0.0675 |
0.0472 | 0.6858 | 1000 | 0.0666 |
0.0509 | 0.7544 | 1100 | 0.0658 |
0.1134 | 0.8230 | 1200 | 0.0651 |
0.073 | 0.8916 | 1300 | 0.0636 |
0.1127 | 0.9601 | 1400 | 0.0636 |
0.0228 | 1.0287 | 1500 | 0.0637 |
0.0262 | 1.0973 | 1600 | 0.0634 |
0.0448 | 1.1659 | 1700 | 0.0629 |
0.037 | 1.2345 | 1800 | 0.0625 |
0.1208 | 1.3030 | 1900 | 0.0620 |
0.0368 | 1.3716 | 2000 | 0.0619 |
0.0951 | 1.4402 | 2100 | 0.0612 |
0.0582 | 1.5088 | 2200 | 0.0604 |
0.0438 | 1.5774 | 2300 | 0.0602 |
0.033 | 1.6459 | 2400 | 0.0598 |
0.0626 | 1.7145 | 2500 | 0.0597 |
0.0106 | 1.7831 | 2600 | 0.0597 |
0.0312 | 1.8517 | 2700 | 0.0594 |
0.0382 | 1.9203 | 2800 | 0.0595 |
0.0622 | 1.9889 | 2900 | 0.0594 |
Framework versions
- PEFT 0.13.2
- Transformers 4.46.0
- Pytorch 2.5.0+cu124
- Datasets 3.0.1
- Tokenizers 0.20.1
- Downloads last month
- 8
Inference Providers
NEW
This model is not currently available via any of the supported Inference Providers.
The model cannot be deployed to the HF Inference API:
The model has no pipeline_tag.
Model tree for Alphatao/a30c98f1-5377-4fe4-bf1b-66b3d754b9b2
Base model
Qwen/Qwen1.5-0.5B