TinyLlama-1.1B-Instruct-3T

TinyLlama 1.1B Instruct 3T

<img src="https://huggingface.co/gardner/TinyLlama-1.1B-Instruct-3T/resolve/main/tinyllama-1.1b-instruct.webp?download=true" alt="TinyLlama Instruct" />

This is the 3T base model trained on openhermes instruct dataset for 4 epochs. It is intended to be used for further finetuning.

<img src="https://raw.githubusercontent.com/OpenAccess-AI-Collective/axolotl/main/image/axolotl-badge-web.png" alt="Built with Axolotl" width="200" height="32"/>

Loss

<img src="https://huggingface.co/gardner/TinyLlama-1.1B-Instruct-3T/resolve/main/loss.webp?download=true" alt="Loss chart" />

axolotl config file: lora.yml

base_model: TinyLlama/TinyLlama-1.1B-intermediate-step-1431k-3T
model_type: LlamaForCausalLM
tokenizer_type: LlamaTokenizer
is_llama_derived_model: true

load_in_8bit: true
load_in_4bit: false
strict: false

datasets:
  - path: teknium/openhermes
    type: alpaca
dataset_prepared_path:
val_set_size: 0.05
output_dir: ./tiny-llama-instruct-lora

sequence_len: 4096
sample_packing: true
pad_to_sequence_len: true

adapter: lora
lora_model_dir:
lora_r: 32
lora_alpha: 16
lora_dropout: 0.05
lora_target_linear: true
lora_fan_in_fan_out:

gradient_accumulation_steps: 4
micro_batch_size: 2
num_epochs: 4
optimizer: adamw_bnb_8bit
lr_scheduler: cosine
learning_rate: 0.0002

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
saves_per_epoch: 1
debug:
deepspeed:
weight_decay: 0.0
fsdp:
fsdp_config:
special_tokens: