File size: 1,672 Bytes
3be406d
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
00fdec8
3be406d
 
 
00fdec8
3be406d
 
 
 
 
00fdec8
3be406d
 
 
 
 
 
 
 
 
 
 
 
 
 
00fdec8
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
---
license: apache-2.0
tags:
- StepLaw
- causal-lm
language:
- en
library_name: transformers
pipeline_tag: text-generation
model-index:
- name: step2v2_0618_h1024_ffnh9552_numh16_numl8_lr2.762e-03_bs1024_ti38146_mlr1.00e-05
  results: []
---

# Wandb Model Name: step2v2_0618_h1024_ffnh9552_numh16_numl8_lr2.762e-03_bs1024_ti38146_mlr1.00e-05

This model is part of the [StepLaw-N_268M-D_79.0B](https://huggingface.co/collections/StepLaw/StepLaw-N_268M-D_79.0B) collection.

## Model Specifications

### Architecture
- **Hidden size (H)**: 1024
- **Feed-forward network size (FFN)**: 9552
- **Attention heads**: 16
- **Layers**: 8
- **Parameter count**: 268M

### Training Parameters
- **Learning rate (lr)**: 2.762e-03
- **Batch size (bs)**: 2097152
- **Training iterations**: 38146
- **Training tokens (D)**: 80.0B

## Model Description

StepLaw models are trained with various hyperparameter settings to enable research on scaling laws and hyperparameter optimization. This specific model was trained with learning rate 2.762e-03 and batch size 2097152 for 38146 iterations, using a total of 80.0B training tokens.

## Usage Example

```python
from transformers import AutoModelForCausalLM, AutoTokenizer

model_name = "StepLaw/StepLaw-N_268M-D_79.0B-LR2.762e-03-BS2097152"
tokenizer = AutoTokenizer.from_pretrained(model_name, trust_remote_code=True, use_fast=False)
model = AutoModelForCausalLM.from_pretrained(model_name, trust_remote_code=True)

# Generate text
inputs = tokenizer("A long time ago in a galaxy far, far away", return_tensors="pt")
outputs = model.generate(**inputs, max_length=100)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))
```