Metabird-7B

See axolotl config

axolotl version: 0.3.0

base_model: leveldevai/TurdusBeagle-7B
model_type: MistralForCausalLM
tokenizer_type: LlamaTokenizer
is_mistral_derived_model: true

load_in_8bit: false
load_in_4bit: false
strict: false

datasets:
  - path: shuyuej/metamath_gsm8k
    type:
      system_prompt: ""
      field_instruction: question
      field_output: answer
      format: "[INST] {instruction} [/INST]"
      no_input_format: "[INST] {instruction} [/INST]"
      
dataset_prepared_path:
val_set_size: 0.05
output_dir: ./out

sequence_len: 8192
sample_packing: true
pad_to_sequence_len: true
eval_sample_packing: false

wandb_project:
wandb_entity:
wandb_watch:
wandb_name:
wandb_log_model:

gradient_accumulation_steps: 4
micro_batch_size: 2
num_epochs: 1
optimizer: adamw_bnb_8bit
lr_scheduler: cosine
learning_rate: 0.000005

train_on_inputs: false
group_by_length: false
bf16: true
fp16: false
tf32: false

gradient_checkpointing: true
early_stopping_patience:
resume_from_checkpoint:
local_rank:
logging_steps: 1
xformers_attention:
flash_attention: true

warmup_steps: 10
evals_per_epoch: 4
eval_table_size:
eval_table_max_new_tokens: 128
saves_per_epoch: 1
debug:
deepspeed:
weight_decay: 0.0
fsdp:
fsdp_config:
special_tokens:
  bos_token: "<s>"
  eos_token: "</s>"
  unk_token: "<unk>"

Metabird

This model is a fine-tuned version of leveldevai/TurdusBeagle-7B on the shuyuej/metamath_gsm8k dataset. It achieves the following results on the evaluation set:

• Loss: 0.4017

Model description

More information soon

Intended uses & limitations

More information soon

Training and evaluation data

More information soon

Training procedure

More information soon

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 5e-06
• train_batch_size: 2
• eval_batch_size: 2
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 8
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_steps: 10
• num_epochs: 1

Training results

Training Loss	Epoch	Step	Validation Loss
0.9074	0.05	1	0.9932
0.5012	0.26	5	0.4849
0.4204	0.53	10	0.4435
0.3748	0.79	15	0.4017

Framework versions

• Transformers 4.37.0.dev0
• Pytorch 2.0.1+cu117
• Datasets 2.16.1
• Tokenizers 0.15.0

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric	Value
Avg.	71.03
AI2 Reasoning Challenge (25-Shot)	69.54
HellaSwag (10-Shot)	87.54
MMLU (5-Shot)	65.27
TruthfulQA (0-shot)	57.94
Winogrande (5-shot)	83.03
GSM8k (5-shot)	62.85