See axolotl config

axolotl version: 0.4.1

adapter: lora
base_model: trl-internal-testing/tiny-random-LlamaForCausalLM
bf16: true
chat_template: llama3
cosine_min_lr_ratio: 0.3
dataset_prepared_path: null
datasets:
- data_files:
  - 1370a6dd48e2ecba_train_data.json
  ds_type: json
  format: custom
  path: /workspace/input_data/1370a6dd48e2ecba_train_data.json
  type:
    field_input: document_description
    field_instruction: document_type
    field_output: generated_text
    format: '{instruction} {input}'
    no_input_format: '{instruction}'
    system_format: '{system}'
    system_prompt: ''
debug: null
deepspeed: null
early_stopping_patience: 4
eval_max_new_tokens: 128
eval_steps: 200
eval_table_size: null
flash_attention: true
fp16: null
fsdp: null
fsdp_config: null
gradient_accumulation_steps: 4
gradient_checkpointing: true
group_by_length: false
hub_model_id: Romain-XV/fb7673bb-bf79-44ba-abf2-6450fe8a1f5e
hub_repo: null
hub_strategy: checkpoint
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
lr_scheduler: cosine
max_grad_norm: 1.0
max_steps: 44502
micro_batch_size: 4
mlflow_experiment_name: /tmp/1370a6dd48e2ecba_train_data.json
model_type: AutoModelForCausalLM
num_epochs: 3
optimizer: adamw_torch
output_dir: miner_id_24
pad_to_sequence_len: true
resume_from_checkpoint: null
s2_attention: null
sample_packing: false
save_steps: 200
sequence_len: 2048
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: 1ea730f0-95e7-41bf-88e5-468b7a961401
wandb_project: Gradients-On-Demand
wandb_run: your_name
wandb_runid: 1ea730f0-95e7-41bf-88e5-468b7a961401
warmup_steps: 10
weight_decay: 0.0
xformers_attention: null

fb7673bb-bf79-44ba-abf2-6450fe8a1f5e

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

• Loss: 10.3119

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: 4
• total_train_batch_size: 16
• optimizer: Use OptimizerNames.ADAMW_TORCH 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: 9854

Training results

Training Loss	Epoch	Step	Validation Loss
10.3819	0.0003	1	10.3802
10.3395	0.0609	200	10.3350
10.3307	0.1218	400	10.3291
10.3213	0.1827	600	10.3262
10.323	0.2436	800	10.3239
10.3377	0.3045	1000	10.3221
10.333	0.3654	1200	10.3209
10.3214	0.4262	1400	10.3199
10.3305	0.4871	1600	10.3189
10.3167	0.5480	1800	10.3184
10.3112	0.6089	2000	10.3176
10.3294	0.6698	2200	10.3172
10.3199	0.7307	2400	10.3167
10.3257	0.7916	2600	10.3167
10.3301	0.8525	2800	10.3158
10.3258	0.9134	3000	10.3155
10.3223	0.9743	3200	10.3153
10.0883	1.0352	3400	10.3154
10.0741	1.0961	3600	10.3150
9.6731	1.1569	3800	10.3145
9.4612	1.2178	4000	10.3143
10.1269	1.2787	4200	10.3144
10.4693	1.3396	4400	10.3140
10.3859	1.4005	4600	10.3140
10.4245	1.4614	4800	10.3138
10.3534	1.5223	5000	10.3137
9.4533	1.5832	5200	10.3134
10.9282	1.6441	5400	10.3133
9.9388	1.7050	5600	10.3133
10.3262	1.7659	5800	10.3131
10.2825	1.8268	6000	10.3130
10.2611	1.8877	6200	10.3130
10.4574	1.9485	6400	10.3129
10.3213	2.0094	6600	10.3129
10.3267	2.0703	6800	10.3127
10.3126	2.1312	7000	10.3125
10.3346	2.1921	7200	10.3126
10.3218	2.2530	7400	10.3125
10.3207	2.3139	7600	10.3124
10.3098	2.3748	7800	10.3125
10.3191	2.4357	8000	10.3123
10.3244	2.4966	8200	10.3122
10.3187	2.5575	8400	10.3123
10.3148	2.6184	8600	10.3121
10.317	2.6793	8800	10.3122
10.3115	2.7401	9000	10.3121
10.309	2.8010	9200	10.3121
10.3158	2.8619	9400	10.3120
10.3174	2.9228	9600	10.3120
10.3196	2.9837	9800	10.3119

Framework versions

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