Hugging Face's logo

Crystalcareai
/
PhiAlpaca2

Text Generation
Transformers
PyTorch
phi-msft
Generated from Trainer
custom_code
Inference Endpoints
Model card Files Community
PhiAlpaca2 / README.md
Crystalcareai's picture
Crystalcareai
Upload 10 files
e4b1638
preview code
|
raw
history blame contribute delete
3.15 kB
metadata
license: mit
base_model: microsoft/phi-2
tags:
  - generated_from_trainer
model-index:
  - name: phi-sft-out
    results: []

Built with Axolotl

See axolotl config

axolotl version: 0.3.0

base_model: microsoft/phi-2
model_type: AutoModelForCausalLM
tokenizer_type: AutoTokenizer
trust_remote_code: true

load_in_8bit: false
load_in_4bit: false
strict: false

datasets:
  - path: tatsu-lab/alpaca
    type: alpaca

dataset_prepared_path:
val_set_size: 0.05
output_dir: ./phi-sft-out

sequence_len: 2048
sample_packing: false  # currently unsupported
pad_to_sequence_len:

adapter:
lora_model_dir:
lora_r: 16
lora_alpha: 32
lora_dropout: 0.1
lora_target_linear: true
lora_fan_in_fan_out:
lora_modules_to_save:
  - embd
  - lm_head

wandb_project: Deepseek Wa
wandb_entity: lucasatkins81
wandb_watch:
wandb_name: Phi2 a6000 FT
wandb_log_model:

gradient_accumulation_steps: 1
micro_batch_size: 1
num_epochs: 1.5
optimizer: paged_adamw_8bit
adam_beta2: 0.95
adam_epsilon: 0.00001
max_grad_norm: 1.0
lr_scheduler: cosine
learning_rate: 1e-5

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

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

warmup_steps: 100
evals_per_epoch: 4
saves_per_epoch: 1
debug:
deepspeed:
weight_decay: 0.1
fsdp:
fsdp_config:
resize_token_embeddings_to_32x: true
special_tokens:
  pad_token: "<|endoftext|>"

phi-sft-out

This model is a fine-tuned version of microsoft/phi-2 on the None dataset. It achieves the following results on the evaluation set:

  • Loss: nan

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: 1e-05
  • train_batch_size: 1
  • eval_batch_size: 1
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.95) and epsilon=1e-05
  • lr_scheduler_type: cosine
  • lr_scheduler_warmup_steps: 100
  • num_epochs: 1.5

Training results

Training Loss Epoch Step Validation Loss
1.4382 0.0 1 nan
0.9139 0.25 12351 nan
0.016 0.5 24702 nan
0.0538 0.75 37053 nan
0.6701 1.0 49404 nan
0.0018 1.25 61755 nan

Framework versions

  • Transformers 4.37.0.dev0
  • Pytorch 2.1.1+cu121
  • Datasets 2.16.1
  • Tokenizers 0.15.0