ONNX 模型库
返回模型

说明文档

GPT-J 6B

模型描述

GPT-J 6B 是一个使用 Ben Wang 的 Mesh Transformer JAX 训练的 Transformer 模型。"GPT-J" 指的是模型类别,而 "6B" 表示可训练参数的数量。

原始实现

点击此链接查看原始实现。

如何使用

通过 git clone https://huggingface.co/OWG/gpt-j-6B 克隆仓库来下载模型。

然后你可以使用以下代码来使用该模型:

from onnxruntime import InferenceSession, SessionOptions, GraphOptimizationLevel
from transformers import AutoTokenizer


tokenizer = AutoTokenizer.from_pretrained("EleutherAI/gpt-j-6B")

options = SessionOptions()
options.graph_optimization_level = GraphOptimizationLevel.ORT_ENABLE_ALL

session = InferenceSession("path/to/model.onnx", sess_options=options)
session.disable_fallback()

TODO

OWG/gpt-j-6B

作者 OWG

↓ 0 ♥ 1

创建时间: 2022-04-14 09:20:10+00:00

更新时间: 2022-04-19 08:42:15+00:00

在 Hugging Face 上查看

文件 (316)

.gitattributes
README.md
model/lm_head.bias
model/model.onnx ONNX
model/onnx__MatMul_11436
model/onnx__MatMul_11437
model/onnx__MatMul_11438
model/onnx__MatMul_11468
model/onnx__MatMul_11469
model/onnx__MatMul_11470
model/onnx__MatMul_11471
model/onnx__MatMul_11472
model/onnx__MatMul_11473
model/onnx__MatMul_11503
model/onnx__MatMul_11504
model/onnx__MatMul_11505
model/onnx__MatMul_11506
model/onnx__MatMul_11507
model/onnx__MatMul_11508
model/onnx__MatMul_11538
model/onnx__MatMul_11539
model/onnx__MatMul_11540
model/onnx__MatMul_11541
model/onnx__MatMul_11542
model/onnx__MatMul_11543
model/onnx__MatMul_11573
model/onnx__MatMul_11574
model/onnx__MatMul_11575
model/onnx__MatMul_11576
model/onnx__MatMul_11577
model/onnx__MatMul_11578
model/onnx__MatMul_11608
model/onnx__MatMul_11609
model/onnx__MatMul_11610
model/onnx__MatMul_11611
model/onnx__MatMul_11612
model/onnx__MatMul_11613
model/onnx__MatMul_11643
model/onnx__MatMul_11644
model/onnx__MatMul_11645
model/onnx__MatMul_11646
model/onnx__MatMul_11647
model/onnx__MatMul_11648
model/onnx__MatMul_11678
model/onnx__MatMul_11679
model/onnx__MatMul_11680
model/onnx__MatMul_11681
model/onnx__MatMul_11682
model/onnx__MatMul_11683
model/onnx__MatMul_11713
model/onnx__MatMul_11714
model/onnx__MatMul_11715
model/onnx__MatMul_11716
model/onnx__MatMul_11717
model/onnx__MatMul_11718
model/onnx__MatMul_11748
model/onnx__MatMul_11749
model/onnx__MatMul_11750
model/onnx__MatMul_11751
model/onnx__MatMul_11752
model/onnx__MatMul_11753
model/onnx__MatMul_11783
model/onnx__MatMul_11784
model/onnx__MatMul_11785
model/onnx__MatMul_11786
model/onnx__MatMul_11787
model/onnx__MatMul_11788
model/onnx__MatMul_11818
model/onnx__MatMul_11819
model/onnx__MatMul_11820
model/onnx__MatMul_11821
model/onnx__MatMul_11822
model/onnx__MatMul_11823
model/onnx__MatMul_11853
model/onnx__MatMul_11854
model/onnx__MatMul_11855
model/onnx__MatMul_11856
model/onnx__MatMul_11857
model/onnx__MatMul_11858
model/onnx__MatMul_11888
model/onnx__MatMul_11889
model/onnx__MatMul_11890
model/onnx__MatMul_11891
model/onnx__MatMul_11892
model/onnx__MatMul_11893
model/onnx__MatMul_11923
model/onnx__MatMul_11924
model/onnx__MatMul_11925
model/onnx__MatMul_11926
model/onnx__MatMul_11927
model/onnx__MatMul_11928
model/onnx__MatMul_11958
model/onnx__MatMul_11959
model/onnx__MatMul_11960
model/onnx__MatMul_11961
model/onnx__MatMul_11962
model/onnx__MatMul_11963
model/onnx__MatMul_11993
model/onnx__MatMul_11994
model/onnx__MatMul_11995
model/onnx__MatMul_11996
model/onnx__MatMul_11997
model/onnx__MatMul_11998
model/onnx__MatMul_12028
model/onnx__MatMul_12029
model/onnx__MatMul_12030
model/onnx__MatMul_12031
model/onnx__MatMul_12032
model/onnx__MatMul_12033
model/onnx__MatMul_12063
model/onnx__MatMul_12064
model/onnx__MatMul_12065
model/onnx__MatMul_12066
model/onnx__MatMul_12067
model/onnx__MatMul_12068
model/onnx__MatMul_12098
model/onnx__MatMul_12099
model/onnx__MatMul_12100
model/onnx__MatMul_12101
model/onnx__MatMul_12102
model/onnx__MatMul_12103
model/onnx__MatMul_12133
model/onnx__MatMul_12134
model/onnx__MatMul_12135
model/onnx__MatMul_12136
model/onnx__MatMul_12137
model/onnx__MatMul_12138
model/onnx__MatMul_12168
model/onnx__MatMul_12169
model/onnx__MatMul_12170
model/onnx__MatMul_12171
model/onnx__MatMul_12172
model/onnx__MatMul_12173
model/onnx__MatMul_12203
model/onnx__MatMul_12204
model/onnx__MatMul_12205
model/onnx__MatMul_12206
model/onnx__MatMul_12207
model/onnx__MatMul_12208
model/onnx__MatMul_12238
model/onnx__MatMul_12239
model/onnx__MatMul_12240
model/onnx__MatMul_12241
model/onnx__MatMul_12242
model/onnx__MatMul_12243
model/onnx__MatMul_12273
model/onnx__MatMul_12274
model/onnx__MatMul_12275
model/onnx__MatMul_12276
model/onnx__MatMul_12277
model/onnx__MatMul_12278
model/onnx__MatMul_12308
model/onnx__MatMul_12309
model/onnx__MatMul_12310
model/onnx__MatMul_12311
model/onnx__MatMul_12312
model/onnx__MatMul_12313
model/onnx__MatMul_12343
model/onnx__MatMul_12344
model/onnx__MatMul_12345
model/onnx__MatMul_12346
model/onnx__MatMul_12347
model/onnx__MatMul_12348
model/onnx__MatMul_12378
model/onnx__MatMul_12379
model/onnx__MatMul_12380
model/onnx__MatMul_12381
model/onnx__MatMul_12382
model/onnx__MatMul_12383
model/onnx__MatMul_12413
model/onnx__MatMul_12414
model/onnx__MatMul_12415
model/onnx__MatMul_12416
model/transformer.h.0.attn.bias
model/transformer.h.0.ln_1.bias
model/transformer.h.0.ln_1.weight
model/transformer.h.0.mlp.fc_in.bias
model/transformer.h.0.mlp.fc_out.bias
model/transformer.h.1.attn.bias
model/transformer.h.1.ln_1.bias
model/transformer.h.1.ln_1.weight
model/transformer.h.1.mlp.fc_in.bias
model/transformer.h.1.mlp.fc_out.bias
model/transformer.h.10.attn.bias
model/transformer.h.10.ln_1.bias
model/transformer.h.10.ln_1.weight
model/transformer.h.10.mlp.fc_in.bias
model/transformer.h.10.mlp.fc_out.bias
model/transformer.h.11.attn.bias
model/transformer.h.11.ln_1.bias
model/transformer.h.11.ln_1.weight
model/transformer.h.11.mlp.fc_in.bias
model/transformer.h.11.mlp.fc_out.bias
model/transformer.h.12.attn.bias
model/transformer.h.12.ln_1.bias
model/transformer.h.12.ln_1.weight
model/transformer.h.12.mlp.fc_in.bias
model/transformer.h.12.mlp.fc_out.bias
model/transformer.h.13.attn.bias
model/transformer.h.13.ln_1.bias
model/transformer.h.13.ln_1.weight
model/transformer.h.13.mlp.fc_in.bias
model/transformer.h.13.mlp.fc_out.bias
model/transformer.h.14.attn.bias
model/transformer.h.14.ln_1.bias
model/transformer.h.14.ln_1.weight
model/transformer.h.14.mlp.fc_in.bias
model/transformer.h.14.mlp.fc_out.bias
model/transformer.h.15.attn.bias
model/transformer.h.15.ln_1.bias
model/transformer.h.15.ln_1.weight
model/transformer.h.15.mlp.fc_in.bias
model/transformer.h.15.mlp.fc_out.bias
model/transformer.h.16.attn.bias
model/transformer.h.16.ln_1.bias
model/transformer.h.16.ln_1.weight
model/transformer.h.16.mlp.fc_in.bias
model/transformer.h.16.mlp.fc_out.bias
model/transformer.h.17.attn.bias
model/transformer.h.17.ln_1.bias
model/transformer.h.17.ln_1.weight
model/transformer.h.17.mlp.fc_in.bias
model/transformer.h.17.mlp.fc_out.bias
model/transformer.h.18.attn.bias
model/transformer.h.18.ln_1.bias
model/transformer.h.18.ln_1.weight
model/transformer.h.18.mlp.fc_in.bias
model/transformer.h.18.mlp.fc_out.bias
model/transformer.h.19.attn.bias
model/transformer.h.19.ln_1.bias
model/transformer.h.19.ln_1.weight
model/transformer.h.19.mlp.fc_in.bias
model/transformer.h.19.mlp.fc_out.bias
model/transformer.h.2.attn.bias
model/transformer.h.2.ln_1.bias
model/transformer.h.2.ln_1.weight
model/transformer.h.2.mlp.fc_in.bias
model/transformer.h.2.mlp.fc_out.bias
model/transformer.h.20.attn.bias
model/transformer.h.20.ln_1.bias
model/transformer.h.20.ln_1.weight
model/transformer.h.20.mlp.fc_in.bias
model/transformer.h.20.mlp.fc_out.bias
model/transformer.h.21.attn.bias
model/transformer.h.21.ln_1.bias
model/transformer.h.21.ln_1.weight
model/transformer.h.21.mlp.fc_in.bias
model/transformer.h.21.mlp.fc_out.bias
model/transformer.h.22.attn.bias
model/transformer.h.22.ln_1.bias
model/transformer.h.22.ln_1.weight
model/transformer.h.22.mlp.fc_in.bias
model/transformer.h.22.mlp.fc_out.bias
model/transformer.h.23.attn.bias
model/transformer.h.23.ln_1.bias
model/transformer.h.23.ln_1.weight
model/transformer.h.23.mlp.fc_in.bias
model/transformer.h.23.mlp.fc_out.bias
model/transformer.h.24.attn.bias
model/transformer.h.24.ln_1.bias
model/transformer.h.24.ln_1.weight
model/transformer.h.24.mlp.fc_in.bias
model/transformer.h.24.mlp.fc_out.bias
model/transformer.h.25.attn.bias
model/transformer.h.25.ln_1.bias
model/transformer.h.25.ln_1.weight
model/transformer.h.25.mlp.fc_in.bias
model/transformer.h.25.mlp.fc_out.bias
model/transformer.h.26.attn.bias
model/transformer.h.26.ln_1.bias
model/transformer.h.26.ln_1.weight
model/transformer.h.26.mlp.fc_in.bias
model/transformer.h.26.mlp.fc_out.bias
model/transformer.h.27.attn.bias
model/transformer.h.27.ln_1.bias
model/transformer.h.27.ln_1.weight
model/transformer.h.27.mlp.fc_in.bias
model/transformer.h.27.mlp.fc_out.bias
model/transformer.h.3.attn.bias
model/transformer.h.3.ln_1.bias
model/transformer.h.3.ln_1.weight
model/transformer.h.3.mlp.fc_in.bias
model/transformer.h.3.mlp.fc_out.bias
model/transformer.h.4.attn.bias
model/transformer.h.4.ln_1.bias
model/transformer.h.4.ln_1.weight
model/transformer.h.4.mlp.fc_in.bias
model/transformer.h.4.mlp.fc_out.bias
model/transformer.h.5.attn.bias
model/transformer.h.5.ln_1.bias
model/transformer.h.5.ln_1.weight
model/transformer.h.5.mlp.fc_in.bias
model/transformer.h.5.mlp.fc_out.bias
model/transformer.h.6.attn.bias
model/transformer.h.6.ln_1.bias
model/transformer.h.6.ln_1.weight
model/transformer.h.6.mlp.fc_in.bias
model/transformer.h.6.mlp.fc_out.bias
model/transformer.h.7.attn.bias
model/transformer.h.7.ln_1.bias
model/transformer.h.7.ln_1.weight
model/transformer.h.7.mlp.fc_in.bias
model/transformer.h.7.mlp.fc_out.bias
model/transformer.h.8.attn.bias
model/transformer.h.8.ln_1.bias
model/transformer.h.8.ln_1.weight
model/transformer.h.8.mlp.fc_in.bias
model/transformer.h.8.mlp.fc_out.bias
model/transformer.h.9.attn.bias
model/transformer.h.9.ln_1.bias
model/transformer.h.9.ln_1.weight
model/transformer.h.9.mlp.fc_in.bias
model/transformer.h.9.mlp.fc_out.bias
model/transformer.ln_f.bias
model/transformer.ln_f.weight
model/transformer.wte.weight