kastan

Kastan's group workspace

Group: gpt2_3d_manual_8B

1-1

of 1

Tags

Notes

Author

kastan

State

Finished

Start time

July 15th, 2022 6:40:49 PM

Runtime

13m 22s

Tracked hours

13m

Run path

kastan/col_ai/sfir46kw

Linux-4.18.0-305.49.1.el8_4.x86_64-x86_64-with-glibc2.28

Python version

3.9.12

Git repository

git clone https://github.com/hpcaitech/ColossalAI-Examples.git

Git state

git checkout -b "gpt2_3d_manual_8B" f743872c2089d6bb5e593db6a8a48d427e6b2b1e

Command

/u/kastanday/new_colossal_ai/ColossalAI/examples/language/gpt/train_gpt.py --config gpt2_configs/gpt2_3d_manual_8B.py --host gpub020 --port 29500 --world_size 8 --rank 5

System Hardware

CPU count	64
GPU count	4
GPU type	NVIDIA A40

W&B CLI Version

0.12.21

Group

gpt2_3d_manual_8B

Config parameters are your model's inputs. Learn more

▶
Config parameters:{} 22 keys
- backend:
  "nccl"
- BATCH_SIZE:
  4
- colossal_config_file:
  "gpt2_configs/gpt2_3d_manual_8B.py"
- config:
  "gpt2_configs/gpt2_3d_manual_8B.py"
- data_dir:
  "/u/kastanday/colossal_ai/raw_json_backup/train_data_FINAL.json"
- ▶
  fp16:{} 1 key
  - mode:
    "AMP_TYPE.NAIVE"
- from_col:
  true
- from_torch:
  false
- gpt2_8B:
  "titans.model.gpt.gpt.gpt2_8B"
- gpt2_small:
  "titans.model.gpt.gpt.gpt2_small"
- host:
  "gpub020"
- local_rank:
  null
- ▶
  loss:{} 1 key
  - type:
    "titans.loss.lm_loss.gpt_lmloss.GPTLMLoss"
- ▶
  model:{} 2 keys
  - checkpoint:
    true
  - type:
    "titans.model.gpt.gpt.gpt2_8B"
- NUM_EPOCHS:
  60
- ▶
  optimizer:{} 3 keys
  - lr:
    0.00015
  - type:
    "torch.optim.adam.Adam"
  - weight_decay:
    0.01
- ▶
  parallel:{} 2 keys
  - pipeline:
    1
  - ▶
    tensor:{} 2 keys
    - mode:
      "3d"
    - size:
      8
- port:
  29,500
- rank:
  5
- SEQ_LEN:
  1,024
- TENSOR_PARALLEL:
  8
- world_size:
  8

Summary metrics are your model's outputs. Learn more

▶
Summary metrics:{} 5 keys
- hello_message:
  "Hi, we got started"
- loss:
  8.965822219848633
- per_step_metrics.loss:
  8.965822219848633
- per_step_metrics.samples_per_sec:
  0.77647
- per_step_metrics.Tflops:
  48.204

This run produced these artifacts as outputs. Total: 1. Learn more

wandb-history

run-sfir46kw-history:v0