See axolotl config

axolotl version: 0.4.1

adapter: lora
base_model: katuni4ka/tiny-random-qwen1.5-moe
bf16: auto
chat_template: llama3
dataset_prepared_path: null
datasets:
- data_files:
  - 2b8375abdf26554a_train_data.json
  ds_type: json
  format: custom
  path: /workspace/input_data/2b8375abdf26554a_train_data.json
  type:
    field_instruction: is_hate
    field_output: tweet
    format: '{instruction}'
    no_input_format: '{instruction}'
    system_format: '{system}'
    system_prompt: ''
debug: null
deepspeed: null
early_stopping_patience: null
eval_max_new_tokens: 128
eval_table_size: null
evals_per_epoch: 1
flash_attention: true
fp16: null
fsdp: null
fsdp_config: null
gradient_accumulation_steps: 8
gradient_checkpointing: true
group_by_length: false
hub_model_id: error577/f16880e2-5a7f-4ac8-aab4-78ff4c1d642b
hub_repo: null
hub_strategy: end
hub_token: null
learning_rate: 0.0001
load_in_4bit: false
load_in_8bit: false
local_rank: null
logging_steps: 1
lora_alpha: 16
lora_dropout: 0.05
lora_fan_in_fan_out: null
lora_model_dir: null
lora_r: 32
lora_target_linear: true
lr_scheduler: cosine
max_steps: 1000
micro_batch_size: 2
mlflow_experiment_name: /tmp/2b8375abdf26554a_train_data.json
model_type: AutoModelForCausalLM
num_epochs: 8
optimizer: adamw_bnb_8bit
output_dir: miner_id_24
pad_to_sequence_len: true
resume_from_checkpoint: null
s2_attention: null
sample_packing: false
saves_per_epoch: 1
sequence_len: 2048
strict: false
tf32: false
tokenizer_type: AutoTokenizer
train_on_inputs: false
trust_remote_code: true
val_set_size: 0.02
wandb_entity: null
wandb_mode: online
wandb_name: c71231f6-9c0b-4206-b56b-cc809f1f3448
wandb_project: Gradients-On-Demand
wandb_run: your_name
wandb_runid: c71231f6-9c0b-4206-b56b-cc809f1f3448
warmup_steps: 10
weight_decay: 0.0
xformers_attention: null

f16880e2-5a7f-4ac8-aab4-78ff4c1d642b

This model is a fine-tuned version of katuni4ka/tiny-random-qwen1.5-moe on the None dataset. It achieves the following results on the evaluation set:

• Loss: 11.8567

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: 0.0001
• train_batch_size: 2
• eval_batch_size: 2
• seed: 42
• gradient_accumulation_steps: 8
• total_train_batch_size: 16
• 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: 1000

Training results

Training Loss	Epoch	Step	Validation Loss
11.9364	0.0017	1	11.9389
11.8897	0.2121	125	11.8998
11.8959	0.4242	250	11.8887
11.8681	0.6363	375	11.8712
11.8515	0.8484	500	11.8612
10.7436	1.0604	625	11.8583
9.9079	1.2725	750	11.8572
13.3687	1.4846	875	11.8568
13.3599	1.6967	1000	11.8567

Framework versions

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