See axolotl config

axolotl version: 0.4.1

adapter: lora
base_model: echarlaix/tiny-random-mistral
bf16: true
chat_template: llama3
dataset_prepared_path: null
datasets:
- data_files:
  - 6b60a00d2240e919_train_data.json
  ds_type: json
  format: custom
  path: /workspace/input_data/6b60a00d2240e919_train_data.json
  type:
    field_instruction: thinking
    field_output: raw_emails
    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/cdf1e981-a8b4-4ebe-a46c-07087cead18c
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: 128
lora_dropout: 0.1
lora_fan_in_fan_out: null
lora_model_dir: null
lora_r: 64
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: 4140
micro_batch_size: 4
mlflow_experiment_name: /tmp/6b60a00d2240e919_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: 0.044419569485532544
wandb_entity: null
wandb_mode: online
wandb_name: ecf87c37-5377-4e51-a311-a1485a15b7d2
wandb_project: Gradients-On-Demand
wandb_run: your_name
wandb_runid: ecf87c37-5377-4e51-a311-a1485a15b7d2
warmup_steps: 10
weight_decay: 0.0
xformers_attention: null

cdf1e981-a8b4-4ebe-a46c-07087cead18c

This model is a fine-tuned version of echarlaix/tiny-random-mistral on the None dataset. It achieves the following results on the evaluation set:

• Loss: 10.1454

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: 4140

Training results

Training Loss	Epoch	Step	Validation Loss
83.033	0.0003	1	10.3781
82.0369	0.0297	100	10.2480
81.7615	0.0595	200	10.2112
81.6229	0.0892	300	10.1929
81.5353	0.1190	400	10.1844
81.5785	0.1487	500	10.1795
81.4766	0.1785	600	10.1732
81.4573	0.2082	700	10.1687
81.3826	0.2380	800	10.1644
81.4176	0.2677	900	10.1614
81.457	0.2975	1000	10.1591
81.4761	0.3272	1100	10.1571
81.3986	0.3570	1200	10.1555
81.3097	0.3867	1300	10.1546
81.356	0.4165	1400	10.1536
81.339	0.4462	1500	10.1528
81.3342	0.4760	1600	10.1516
81.3775	0.5057	1700	10.1512
81.3441	0.5355	1800	10.1505
81.402	0.5652	1900	10.1501
81.3148	0.5950	2000	10.1494
81.4048	0.6247	2100	10.1489
81.2971	0.6545	2200	10.1485
81.2684	0.6842	2300	10.1479
81.3308	0.7140	2400	10.1475
81.3618	0.7437	2500	10.1469
81.2991	0.7735	2600	10.1467
81.3194	0.8032	2700	10.1465
81.3059	0.8330	2800	10.1463
81.3422	0.8627	2900	10.1460
81.3315	0.8925	3000	10.1459
81.3474	0.9222	3100	10.1458
81.3465	0.9520	3200	10.1456
81.359	0.9817	3300	10.1456
81.2519	1.0115	3400	10.1455
81.3384	1.0412	3500	10.1455
81.2254	1.0710	3600	10.1454
81.2984	1.1007	3700	10.1454
81.2588	1.1305	3800	10.1454
81.3178	1.1602	3900	10.1454
81.35	1.1900	4000	10.1454
81.2935	1.2197	4100	10.1454

Framework versions

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