CYLI310/CodeGPT/adapter

CYLI310/

CodeGPT:latest

113 Downloads Updated 1 month ago

Code assistant trained by GPT-OSS 20B

tools

CodeGPT:latest ... /

adapter

28e2c98c5d20 · 16MB

Metadata

general.architecture

llama

llama
general.file_type

F16

F16
llama.attention.head_count

32

32
llama.attention.head_count_kv

8

8
adapter.lora.alpha

16

16
adapter.type

lora

lora

Tensor

Name

Type

Shape

blk.0

blk.0.attn_k.weight.lora_a

F16

F16

[2880, 16]
blk.0.attn_k.weight.lora_b

F16

F16

[16, 512]
blk.0.attn_output.weight.lora_a

F16

F16

[4096, 16]
blk.0.attn_output.weight.lora_b

F16

F16

[16, 2880]
blk.0.attn_q.weight.lora_a

F16

F16

[2880, 16]
blk.0.attn_q.weight.lora_b

F16

F16

[16, 4096]
blk.0.attn_v.weight.lora_a

F16

F16

[2880, 16]
blk.0.attn_v.weight.lora_b

F16

F16

[16, 512]

blk.1

blk.1.attn_k.weight.lora_a

F16

F16

[2880, 16]
blk.1.attn_k.weight.lora_b

F16

F16

[16, 512]
blk.1.attn_output.weight.lora_a

F16

F16

[4096, 16]
blk.1.attn_output.weight.lora_b

F16

F16

[16, 2880]
blk.1.attn_q.weight.lora_a

F16

F16

[2880, 16]
blk.1.attn_q.weight.lora_b

F16

F16

[16, 4096]
blk.1.attn_v.weight.lora_a

F16

F16

[2880, 16]
blk.1.attn_v.weight.lora_b

F16

F16

[16, 512]

blk.2

blk.2.attn_k.weight.lora_a

F16

F16

[2880, 16]
blk.2.attn_k.weight.lora_b

F16

F16

[16, 512]
blk.2.attn_output.weight.lora_a

F16

F16

[4096, 16]
blk.2.attn_output.weight.lora_b

F16

F16

[16, 2880]
blk.2.attn_q.weight.lora_a

F16

F16

[2880, 16]
blk.2.attn_q.weight.lora_b

F16

F16

[16, 4096]
blk.2.attn_v.weight.lora_a

F16

F16

[2880, 16]
blk.2.attn_v.weight.lora_b

F16

F16

[16, 512]

blk.3

blk.3.attn_k.weight.lora_a

F16

F16

[2880, 16]
blk.3.attn_k.weight.lora_b

F16

F16

[16, 512]
blk.3.attn_output.weight.lora_a

F16

F16

[4096, 16]
blk.3.attn_output.weight.lora_b

F16

F16

[16, 2880]
blk.3.attn_q.weight.lora_a

F16

F16

[2880, 16]
blk.3.attn_q.weight.lora_b

F16

F16

[16, 4096]
blk.3.attn_v.weight.lora_a

F16

F16

[2880, 16]
blk.3.attn_v.weight.lora_b

F16

F16

[16, 512]

blk.4

blk.4.attn_k.weight.lora_a

F16

F16

[2880, 16]
blk.4.attn_k.weight.lora_b

F16

F16

[16, 512]
blk.4.attn_output.weight.lora_a

F16

F16

[4096, 16]
blk.4.attn_output.weight.lora_b

F16

F16

[16, 2880]
blk.4.attn_q.weight.lora_a

F16

F16

[2880, 16]
blk.4.attn_q.weight.lora_b

F16

F16

[16, 4096]
blk.4.attn_v.weight.lora_a

F16

F16

[2880, 16]
blk.4.attn_v.weight.lora_b

F16

F16

[16, 512]

blk.5

blk.5.attn_k.weight.lora_a

F16

F16

[2880, 16]
blk.5.attn_k.weight.lora_b

F16

F16

[16, 512]
blk.5.attn_output.weight.lora_a

F16

F16

[4096, 16]
blk.5.attn_output.weight.lora_b

F16

F16

[16, 2880]
blk.5.attn_q.weight.lora_a

F16

F16

[2880, 16]
blk.5.attn_q.weight.lora_b

F16

F16

[16, 4096]
blk.5.attn_v.weight.lora_a

F16

F16

[2880, 16]
blk.5.attn_v.weight.lora_b

F16

F16

[16, 512]

blk.6

blk.6.attn_k.weight.lora_a

F16

F16

[2880, 16]
blk.6.attn_k.weight.lora_b

F16

F16

[16, 512]
blk.6.attn_output.weight.lora_a

F16

F16

[4096, 16]
blk.6.attn_output.weight.lora_b

F16

F16

[16, 2880]
blk.6.attn_q.weight.lora_a

F16

F16

[2880, 16]
blk.6.attn_q.weight.lora_b

F16

F16

[16, 4096]
blk.6.attn_v.weight.lora_a

F16

F16

[2880, 16]
blk.6.attn_v.weight.lora_b

F16

F16

[16, 512]

blk.7

blk.7.attn_k.weight.lora_a

F16

F16

[2880, 16]
blk.7.attn_k.weight.lora_b

F16

F16

[16, 512]
blk.7.attn_output.weight.lora_a

F16

F16

[4096, 16]
blk.7.attn_output.weight.lora_b

F16

F16

[16, 2880]
blk.7.attn_q.weight.lora_a

F16

F16

[2880, 16]
blk.7.attn_q.weight.lora_b

F16

F16

[16, 4096]
blk.7.attn_v.weight.lora_a

F16

F16

[2880, 16]
blk.7.attn_v.weight.lora_b

F16

F16

[16, 512]

blk.8

blk.8.attn_k.weight.lora_a

F16

F16

[2880, 16]
blk.8.attn_k.weight.lora_b

F16

F16

[16, 512]
blk.8.attn_output.weight.lora_a

F16

F16

[4096, 16]
blk.8.attn_output.weight.lora_b

F16

F16

[16, 2880]
blk.8.attn_q.weight.lora_a

F16

F16

[2880, 16]
blk.8.attn_q.weight.lora_b

F16

F16

[16, 4096]
blk.8.attn_v.weight.lora_a

F16

F16

[2880, 16]
blk.8.attn_v.weight.lora_b

F16

F16

[16, 512]

blk.9

blk.9.attn_k.weight.lora_a

F16

F16

[2880, 16]
blk.9.attn_k.weight.lora_b

F16

F16

[16, 512]
blk.9.attn_output.weight.lora_a

F16

F16

[4096, 16]
blk.9.attn_output.weight.lora_b

F16

F16

[16, 2880]
blk.9.attn_q.weight.lora_a

F16

F16

[2880, 16]
blk.9.attn_q.weight.lora_b

F16

F16

[16, 4096]
blk.9.attn_v.weight.lora_a

F16

F16

[2880, 16]
blk.9.attn_v.weight.lora_b

F16

F16

[16, 512]

blk.10

blk.10.attn_k.weight.lora_a

F16

F16

[2880, 16]
blk.10.attn_k.weight.lora_b

F16

F16

[16, 512]
blk.10.attn_output.weight.lora_a

F16

F16

[4096, 16]
blk.10.attn_output.weight.lora_b

F16

F16

[16, 2880]
blk.10.attn_q.weight.lora_a

F16

F16

[2880, 16]
blk.10.attn_q.weight.lora_b

F16

F16

[16, 4096]
blk.10.attn_v.weight.lora_a

F16

F16

[2880, 16]
blk.10.attn_v.weight.lora_b

F16

F16

[16, 512]

blk.11

blk.11.attn_k.weight.lora_a

F16

F16

[2880, 16]
blk.11.attn_k.weight.lora_b

F16

F16

[16, 512]
blk.11.attn_output.weight.lora_a

F16

F16

[4096, 16]
blk.11.attn_output.weight.lora_b

F16

F16

[16, 2880]
blk.11.attn_q.weight.lora_a

F16

F16

[2880, 16]
blk.11.attn_q.weight.lora_b

F16

F16

[16, 4096]
blk.11.attn_v.weight.lora_a

F16

F16

[2880, 16]
blk.11.attn_v.weight.lora_b

F16

F16

[16, 512]

blk.12

blk.12.attn_k.weight.lora_a

F16

F16

[2880, 16]
blk.12.attn_k.weight.lora_b

F16

F16

[16, 512]
blk.12.attn_output.weight.lora_a

F16

F16

[4096, 16]
blk.12.attn_output.weight.lora_b

F16

F16

[16, 2880]
blk.12.attn_q.weight.lora_a

F16

F16

[2880, 16]
blk.12.attn_q.weight.lora_b

F16

F16

[16, 4096]
blk.12.attn_v.weight.lora_a

F16

F16

[2880, 16]
blk.12.attn_v.weight.lora_b

F16

F16

[16, 512]

blk.13

blk.13.attn_k.weight.lora_a

F16

F16

[2880, 16]
blk.13.attn_k.weight.lora_b

F16

F16

[16, 512]
blk.13.attn_output.weight.lora_a

F16

F16

[4096, 16]
blk.13.attn_output.weight.lora_b

F16

F16

[16, 2880]
blk.13.attn_q.weight.lora_a

F16

F16

[2880, 16]
blk.13.attn_q.weight.lora_b

F16

F16

[16, 4096]
blk.13.attn_v.weight.lora_a

F16

F16

[2880, 16]
blk.13.attn_v.weight.lora_b

F16

F16

[16, 512]

blk.14

blk.14.attn_k.weight.lora_a

F16

F16

[2880, 16]
blk.14.attn_k.weight.lora_b

F16

F16

[16, 512]
blk.14.attn_output.weight.lora_a

F16

F16

[4096, 16]
blk.14.attn_output.weight.lora_b

F16

F16

[16, 2880]
blk.14.attn_q.weight.lora_a

F16

F16

[2880, 16]
blk.14.attn_q.weight.lora_b

F16

F16

[16, 4096]
blk.14.attn_v.weight.lora_a

F16

F16

[2880, 16]
blk.14.attn_v.weight.lora_b

F16

F16

[16, 512]

blk.15

blk.15.attn_k.weight.lora_a

F16

F16

[2880, 16]
blk.15.attn_k.weight.lora_b

F16

F16

[16, 512]
blk.15.attn_output.weight.lora_a

F16

F16

[4096, 16]
blk.15.attn_output.weight.lora_b

F16

F16

[16, 2880]
blk.15.attn_q.weight.lora_a

F16

F16

[2880, 16]
blk.15.attn_q.weight.lora_b

F16

F16

[16, 4096]
blk.15.attn_v.weight.lora_a

F16

F16

[2880, 16]
blk.15.attn_v.weight.lora_b

F16

F16

[16, 512]

blk.16

blk.16.attn_k.weight.lora_a

F16

F16

[2880, 16]
blk.16.attn_k.weight.lora_b

F16

F16

[16, 512]
blk.16.attn_output.weight.lora_a

F16

F16

[4096, 16]
blk.16.attn_output.weight.lora_b

F16

F16

[16, 2880]
blk.16.attn_q.weight.lora_a

F16

F16

[2880, 16]
blk.16.attn_q.weight.lora_b

F16

F16

[16, 4096]
blk.16.attn_v.weight.lora_a

F16

F16

[2880, 16]
blk.16.attn_v.weight.lora_b

F16

F16

[16, 512]

blk.17

blk.17.attn_k.weight.lora_a

F16

F16

[2880, 16]
blk.17.attn_k.weight.lora_b

F16

F16

[16, 512]
blk.17.attn_output.weight.lora_a

F16

F16

[4096, 16]
blk.17.attn_output.weight.lora_b

F16

F16

[16, 2880]
blk.17.attn_q.weight.lora_a

F16

F16

[2880, 16]
blk.17.attn_q.weight.lora_b

F16

F16

[16, 4096]
blk.17.attn_v.weight.lora_a

F16

F16

[2880, 16]
blk.17.attn_v.weight.lora_b

F16

F16

[16, 512]

blk.18

blk.18.attn_k.weight.lora_a

F16

F16

[2880, 16]
blk.18.attn_k.weight.lora_b

F16

F16

[16, 512]
blk.18.attn_output.weight.lora_a

F16

F16

[4096, 16]
blk.18.attn_output.weight.lora_b

F16

F16

[16, 2880]
blk.18.attn_q.weight.lora_a

F16

F16

[2880, 16]
blk.18.attn_q.weight.lora_b

F16

F16

[16, 4096]
blk.18.attn_v.weight.lora_a

F16

F16

[2880, 16]
blk.18.attn_v.weight.lora_b

F16

F16

[16, 512]

blk.19

blk.19.attn_k.weight.lora_a

F16

F16

[2880, 16]
blk.19.attn_k.weight.lora_b

F16

F16

[16, 512]
blk.19.attn_output.weight.lora_a

F16

F16

[4096, 16]
blk.19.attn_output.weight.lora_b

F16

F16

[16, 2880]
blk.19.attn_q.weight.lora_a

F16

F16

[2880, 16]
blk.19.attn_q.weight.lora_b

F16

F16

[16, 4096]
blk.19.attn_v.weight.lora_a

F16

F16

[2880, 16]
blk.19.attn_v.weight.lora_b

F16

F16

[16, 512]

blk.20

blk.20.attn_k.weight.lora_a

F16

F16

[2880, 16]
blk.20.attn_k.weight.lora_b

F16

F16

[16, 512]
blk.20.attn_output.weight.lora_a

F16

F16

[4096, 16]
blk.20.attn_output.weight.lora_b

F16

F16

[16, 2880]
blk.20.attn_q.weight.lora_a

F16

F16

[2880, 16]
blk.20.attn_q.weight.lora_b

F16

F16

[16, 4096]
blk.20.attn_v.weight.lora_a

F16

F16

[2880, 16]
blk.20.attn_v.weight.lora_b

F16

F16

[16, 512]

blk.21

blk.21.attn_k.weight.lora_a

F16

F16

[2880, 16]
blk.21.attn_k.weight.lora_b

F16

F16

[16, 512]
blk.21.attn_output.weight.lora_a

F16

F16

[4096, 16]
blk.21.attn_output.weight.lora_b

F16

F16

[16, 2880]
blk.21.attn_q.weight.lora_a

F16

F16

[2880, 16]
blk.21.attn_q.weight.lora_b

F16

F16

[16, 4096]
blk.21.attn_v.weight.lora_a

F16

F16

[2880, 16]
blk.21.attn_v.weight.lora_b

F16

F16

[16, 512]

blk.22

blk.22.attn_k.weight.lora_a

F16

F16

[2880, 16]
blk.22.attn_k.weight.lora_b

F16

F16

[16, 512]
blk.22.attn_output.weight.lora_a

F16

F16

[4096, 16]
blk.22.attn_output.weight.lora_b

F16

F16

[16, 2880]
blk.22.attn_q.weight.lora_a

F16

F16

[2880, 16]
blk.22.attn_q.weight.lora_b

F16

F16

[16, 4096]
blk.22.attn_v.weight.lora_a

F16

F16

[2880, 16]
blk.22.attn_v.weight.lora_b

F16

F16

[16, 512]

blk.23

blk.23.attn_k.weight.lora_a

F16

F16

[2880, 16]
blk.23.attn_k.weight.lora_b

F16

F16

[16, 512]
blk.23.attn_output.weight.lora_a

F16

F16

[4096, 16]
blk.23.attn_output.weight.lora_b

F16

F16

[16, 2880]
blk.23.attn_q.weight.lora_a

F16

F16

[2880, 16]
blk.23.attn_q.weight.lora_b

F16

F16

[16, 4096]
blk.23.attn_v.weight.lora_a

F16

F16

[2880, 16]
blk.23.attn_v.weight.lora_b

F16

F16

[16, 512]