kzmszk/kzllm1_wiki

Model Card for Model ID

• フルスクラッチで事前学習する練習のために作ったモデル。実用的な意味はいっさいありません。
• dataset: mc4-ja の一部 + wikipedia の一部。filter/dedupe ずみで1.5GBほど。https://note.com/kan_hatakeyama/n/nbea55ed4498d
• train: https://github.com/hatakeyama-llm-team/llm/tree/gcp_1_gpu/codes/2_pretrain
• train framework: megatron-deepspeed
• os: windows11 + wsl2 + docker (cuda1.18)
• hardware: CPU AMD Ryzen 9 7900X, mem 64G, rtx4090 VRAM 24G
• wandb: https://wandb.ai/kzms/kzllm_0/runs/a6427nal?nw=nwuserkzms
• 学習時間300分ぐらい
• GPUは100%利用と出るが電力消費的には６割程度。VRAM使用量は主にglobal_batch_sizeに依存し、4だとぎりぎり。
• train_samples: 3000000　最初デカすぎて cuda OOM error が出たので小さくした。何が変わるか今ひとつわかってない。
• CPUは１割程度。処理が分散されてるかんじはしない。2、3コアぐらい使うけど
• deepspeed の設定で bf16 を true にすると学習失敗する。何か他の学習設定の関係で使えない。fp16 は ok。
• tokenizer は vocab 数65k。かなりでかい。
• wandb summary

wandb: Run summary:
wandb:                      Train/Samples/loss_scale 262144.0
wandb:                              Train/Samples/lr 3e-05
wandb:                      Train/Samples/train_loss 3.8399
wandb:                                     iteration 56470
wandb:                                iteration_time 0.3093
wandb:                      iteration_time_vs_tokens 0.0
wandb:                                loss/iteration 56470
wandb:                                  loss/lm loss 3.8399
wandb:          throughput/approx_params_in_billions 0.37099
wandb:           throughput/elapsed_ms_per_iteration 0.3093
wandb:                          throughput/iteration 56470
wandb:                     throughput/iteration-time 0.3093
wandb:                    throughput/samples_per_sec 9.69947
wandb:        throughput/samples_per_sec_per_replica 9.69947
wandb:                             throughput/tflops 53.94309
wandb:             throughput/tokens_per_gpu_per_sec 19864.51923
wandb: throughput/tokens_per_gpu_per_sec_per_replica 19864.51923
wandb:                     throughput/tokens_per_sec 19864.51923
wandb:         throughput/tokens_per_sec_per_replica 19864.51923

Model Details

Model Description

This is the model card of a 🤗 transformers model that has been pushed on the Hub. This model card has been automatically generated.

• Developed by: kzmszk
• Model type: gpt
• Language(s) (NLP): Ja, En
• License: MIT
• Finetuned from model [optional]: None

Model Sources [optional]

• Repository: https://github.com/hatakeyama-llm-team/llm/tree/gcp_1_gpu/codes/2_pretrain
• Paper [optional]: [More Information Needed]
• Demo [optional]: [More Information Needed]

Uses

Direct Use

[More Information Needed]

Downstream Use [optional]

[More Information Needed]

Out-of-Scope Use

[More Information Needed]

Bias, Risks, and Limitations

[More Information Needed]

Recommendations

Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.

How to Get Started with the Model

Use the code below to get started with the model.

[More Information Needed]

Training Details

Training Data

[More Information Needed]

Training Procedure

Preprocessing [optional]

[More Information Needed]

Training Hyperparameters

• Training regime: [More Information Needed]

Speeds, Sizes, Times [optional]

[More Information Needed]

Evaluation

Testing Data, Factors & Metrics

Testing Data

[More Information Needed]

Factors

[More Information Needed]

Metrics

[More Information Needed]

Results

[More Information Needed]

Summary

Model Examination [optional]

[More Information Needed]

Environmental Impact

Carbon emissions can be estimated using the Machine Learning Impact calculator presented in Lacoste et al. (2019).

• Hardware Type: [More Information Needed]
• Hours used: [More Information Needed]
• Cloud Provider: [More Information Needed]
• Compute Region: [More Information Needed]
• Carbon Emitted: [More Information Needed]

Technical Specifications [optional]

Model Architecture and Objective

[More Information Needed]

Compute Infrastructure

[More Information Needed]

Hardware

[More Information Needed]

Software

[More Information Needed]

Citation [optional]

BibTeX:

[More Information Needed]

APA:

[More Information Needed]

Glossary [optional]

[More Information Needed]

More Information [optional]

[More Information Needed]

Model Card Authors [optional]

[More Information Needed]

Model Card Contact

[More Information Needed]