qwen2.5-coder

The latest series of Code-Specific Qwen models, with significant improvements in code generation, code reasoning, and code fixing.

tools 0.5b 1.5b 3b 7b 14b 32b

349K Pulls Updated 2 days ago

196 Tags

qwen2.5-coder:1.5b-instruct-q5_0 ... /

model

3305f09f78fb · 1.1GB

Metadata

general.architecture

qwen2
general.base_model.0.name

Qwen2.5 Coder 1.5B
general.base_model.0.organization

Qwen
general.base_model.0.repo_url

https://huggingface.co/Qwen/Qwen2.5-Coder-1.5B
general.base_model.count

1
general.basename

Qwen2.5-Coder
general.file_type

8
general.finetune

Instruct
general.languages

[en]
general.license

apache-2.0
general.license.link

https://huggingface.co/Qwen/Qwen2.5-Coder-1.5B-Instruct/blob/main/LICENSE
general.name

Qwen2.5 Coder 1.5B Instruct
general.quantization_version

2
general.size_label

1.5B
general.tags

[code, codeqwen, chat, qwen, qwen-coder, ...]
general.type

model
qwen2.attention.head_count

12
qwen2.attention.head_count_kv

2
qwen2.attention.layer_norm_rms_epsilon

1e-06
qwen2.block_count

28
qwen2.context_length

32768
qwen2.embedding_length

1536
qwen2.feed_forward_length

8960
qwen2.rope.freq_base

1e+06
tokenizer.ggml.add_bos_token

false
tokenizer.ggml.bos_token_id

151643
tokenizer.ggml.eos_token_id

151645
tokenizer.ggml.merges

[Ġ Ġ, ĠĠ ĠĠ, i n, Ġ t, ĠĠĠĠ ĠĠĠĠ, ...]
tokenizer.ggml.model

gpt2
tokenizer.ggml.padding_token_id

151643
tokenizer.ggml.pre

qwen2
tokenizer.ggml.token_type

[1, 1, 1, 1, 1, ...]
tokenizer.ggml.tokens

[!, ", #, $, %, ...]

Tensor

Name

Type

Shape
token_embd.weight

Q6_K

[1536, 151936]

blk.0

blk.0.attn_k.bias

F32

[256]
blk.0.attn_k.weight

Q5_0

[1536, 256]
blk.0.attn_norm.weight

F32

[1536]
blk.0.attn_output.weight

Q5_0

[1536, 1536]
blk.0.attn_q.bias

F32

[1536]
blk.0.attn_q.weight

Q5_0

[1536, 1536]
blk.0.attn_v.bias

F32

[256]
blk.0.attn_v.weight

Q5_0

[1536, 256]
blk.0.ffn_down.weight

Q5_0

[8960, 1536]
blk.0.ffn_gate.weight

Q5_0

[1536, 8960]
blk.0.ffn_norm.weight

F32

[1536]
blk.0.ffn_up.weight

Q5_0

[1536, 8960]

blk.1

blk.1.attn_k.bias

F32

[256]
blk.1.attn_k.weight

Q5_0

[1536, 256]
blk.1.attn_norm.weight

F32

[1536]
blk.1.attn_output.weight

Q5_0

[1536, 1536]
blk.1.attn_q.bias

F32

[1536]
blk.1.attn_q.weight

Q5_0

[1536, 1536]
blk.1.attn_v.bias

F32

[256]
blk.1.attn_v.weight

Q5_0

[1536, 256]
blk.1.ffn_down.weight

Q5_0

[8960, 1536]
blk.1.ffn_gate.weight

Q5_0

[1536, 8960]
blk.1.ffn_norm.weight

F32

[1536]
blk.1.ffn_up.weight

Q5_0

[1536, 8960]

blk.2

blk.2.attn_k.bias

F32

[256]
blk.2.attn_k.weight

Q5_0

[1536, 256]
blk.2.attn_norm.weight

F32

[1536]
blk.2.attn_output.weight

Q5_0

[1536, 1536]
blk.2.attn_q.bias

F32

[1536]
blk.2.attn_q.weight

Q5_0

[1536, 1536]
blk.2.attn_v.bias

F32

[256]
blk.2.attn_v.weight

Q5_0

[1536, 256]
blk.2.ffn_down.weight

Q5_0

[8960, 1536]
blk.2.ffn_gate.weight

Q5_0

[1536, 8960]
blk.2.ffn_norm.weight

F32

[1536]
blk.2.ffn_up.weight

Q5_0

[1536, 8960]

blk.3

blk.3.attn_k.bias

F32

[256]
blk.3.attn_k.weight

Q5_0

[1536, 256]
blk.3.attn_norm.weight

F32

[1536]
blk.3.attn_output.weight

Q5_0

[1536, 1536]
blk.3.attn_q.bias

F32

[1536]
blk.3.attn_q.weight

Q5_0

[1536, 1536]
blk.3.attn_v.bias

F32

[256]
blk.3.attn_v.weight

Q5_0

[1536, 256]
blk.3.ffn_down.weight

Q5_0

[8960, 1536]
blk.3.ffn_gate.weight

Q5_0

[1536, 8960]
blk.3.ffn_norm.weight

F32

[1536]
blk.3.ffn_up.weight

Q5_0

[1536, 8960]

blk.4

blk.4.attn_k.bias

F32

[256]
blk.4.attn_k.weight

Q5_0

[1536, 256]
blk.4.attn_norm.weight

F32

[1536]
blk.4.attn_output.weight

Q5_0

[1536, 1536]
blk.4.attn_q.bias

F32

[1536]
blk.4.attn_q.weight

Q5_0

[1536, 1536]
blk.4.attn_v.bias

F32

[256]
blk.4.attn_v.weight

Q5_0

[1536, 256]
blk.4.ffn_down.weight

Q5_0

[8960, 1536]
blk.4.ffn_gate.weight

Q5_0

[1536, 8960]
blk.4.ffn_norm.weight

F32

[1536]
blk.4.ffn_up.weight

Q5_0

[1536, 8960]

blk.5

blk.5.attn_k.bias

F32

[256]
blk.5.attn_k.weight

Q5_0

[1536, 256]
blk.5.attn_norm.weight

F32

[1536]
blk.5.attn_output.weight

Q5_0

[1536, 1536]
blk.5.attn_q.bias

F32

[1536]
blk.5.attn_q.weight

Q5_0

[1536, 1536]
blk.5.attn_v.bias

F32

[256]
blk.5.attn_v.weight

Q5_0

[1536, 256]
blk.5.ffn_down.weight

Q5_0

[8960, 1536]
blk.5.ffn_gate.weight

Q5_0

[1536, 8960]
blk.5.ffn_norm.weight

F32

[1536]
blk.5.ffn_up.weight

Q5_0

[1536, 8960]

blk.6

blk.6.attn_k.bias

F32

[256]
blk.6.attn_k.weight

Q5_0

[1536, 256]
blk.6.attn_norm.weight

F32

[1536]
blk.6.attn_output.weight

Q5_0

[1536, 1536]
blk.6.attn_q.bias

F32

[1536]
blk.6.attn_q.weight

Q5_0

[1536, 1536]
blk.6.attn_v.bias

F32

[256]
blk.6.attn_v.weight

Q5_0

[1536, 256]
blk.6.ffn_down.weight

Q5_0

[8960, 1536]
blk.6.ffn_gate.weight

Q5_0

[1536, 8960]
blk.6.ffn_norm.weight

F32

[1536]
blk.6.ffn_up.weight

Q5_0

[1536, 8960]

blk.7

blk.7.attn_k.bias

F32

[256]
blk.7.attn_k.weight

Q5_0

[1536, 256]
blk.7.attn_norm.weight

F32

[1536]
blk.7.attn_output.weight

Q5_0

[1536, 1536]
blk.7.attn_q.bias

F32

[1536]
blk.7.attn_q.weight

Q5_0

[1536, 1536]
blk.7.attn_v.bias

F32

[256]
blk.7.attn_v.weight

Q5_0

[1536, 256]
blk.7.ffn_down.weight

Q5_0

[8960, 1536]
blk.7.ffn_gate.weight

Q5_0

[1536, 8960]
blk.7.ffn_norm.weight

F32

[1536]
blk.7.ffn_up.weight

Q5_0

[1536, 8960]

blk.8

blk.8.attn_k.bias

F32

[256]
blk.8.attn_k.weight

Q5_0

[1536, 256]
blk.8.attn_norm.weight

F32

[1536]
blk.8.attn_output.weight

Q5_0

[1536, 1536]
blk.8.attn_q.bias

F32

[1536]
blk.8.attn_q.weight

Q5_0

[1536, 1536]
blk.8.attn_v.bias

F32

[256]
blk.8.attn_v.weight

Q5_0

[1536, 256]
blk.8.ffn_down.weight

Q5_0

[8960, 1536]
blk.8.ffn_gate.weight

Q5_0

[1536, 8960]
blk.8.ffn_norm.weight

F32

[1536]
blk.8.ffn_up.weight

Q5_0

[1536, 8960]

blk.9

blk.9.attn_k.bias

F32

[256]
blk.9.attn_k.weight

Q5_0

[1536, 256]
blk.9.attn_norm.weight

F32

[1536]
blk.9.attn_output.weight

Q5_0

[1536, 1536]
blk.9.attn_q.bias

F32

[1536]
blk.9.attn_q.weight

Q5_0

[1536, 1536]
blk.9.attn_v.bias

F32

[256]
blk.9.attn_v.weight

Q5_0

[1536, 256]
blk.9.ffn_down.weight

Q5_0

[8960, 1536]
blk.9.ffn_gate.weight

Q5_0

[1536, 8960]
blk.9.ffn_norm.weight

F32

[1536]
blk.9.ffn_up.weight

Q5_0

[1536, 8960]

blk.10

blk.10.attn_k.bias

F32

[256]
blk.10.attn_k.weight

Q5_0

[1536, 256]
blk.10.attn_norm.weight

F32

[1536]
blk.10.attn_output.weight

Q5_0

[1536, 1536]
blk.10.attn_q.bias

F32

[1536]
blk.10.attn_q.weight

Q5_0

[1536, 1536]
blk.10.attn_v.bias

F32

[256]
blk.10.attn_v.weight

Q5_0

[1536, 256]
blk.10.ffn_down.weight

Q5_0

[8960, 1536]
blk.10.ffn_gate.weight

Q5_0

[1536, 8960]
blk.10.ffn_norm.weight

F32

[1536]
blk.10.ffn_up.weight

Q5_0

[1536, 8960]

blk.11

blk.11.attn_k.bias

F32

[256]
blk.11.attn_k.weight

Q5_0

[1536, 256]
blk.11.attn_norm.weight

F32

[1536]
blk.11.attn_output.weight

Q5_0

[1536, 1536]
blk.11.attn_q.bias

F32

[1536]
blk.11.attn_q.weight

Q5_0

[1536, 1536]
blk.11.attn_v.bias

F32

[256]
blk.11.attn_v.weight

Q5_0

[1536, 256]
blk.11.ffn_down.weight

Q5_0

[8960, 1536]
blk.11.ffn_gate.weight

Q5_0

[1536, 8960]
blk.11.ffn_norm.weight

F32

[1536]
blk.11.ffn_up.weight

Q5_0

[1536, 8960]

blk.12

blk.12.attn_k.bias

F32

[256]
blk.12.attn_k.weight

Q5_0

[1536, 256]
blk.12.attn_norm.weight

F32

[1536]
blk.12.attn_output.weight

Q5_0

[1536, 1536]
blk.12.attn_q.bias

F32

[1536]
blk.12.attn_q.weight

Q5_0

[1536, 1536]
blk.12.attn_v.bias

F32

[256]
blk.12.attn_v.weight

Q5_0

[1536, 256]
blk.12.ffn_down.weight

Q5_0

[8960, 1536]
blk.12.ffn_gate.weight

Q5_0

[1536, 8960]
blk.12.ffn_norm.weight

F32

[1536]
blk.12.ffn_up.weight

Q5_0

[1536, 8960]

blk.13

blk.13.attn_k.bias

F32

[256]
blk.13.attn_k.weight

Q5_0

[1536, 256]
blk.13.attn_norm.weight

F32

[1536]
blk.13.attn_output.weight

Q5_0

[1536, 1536]
blk.13.attn_q.bias

F32

[1536]
blk.13.attn_q.weight

Q5_0

[1536, 1536]
blk.13.attn_v.bias

F32

[256]
blk.13.attn_v.weight

Q5_0

[1536, 256]
blk.13.ffn_down.weight

Q5_0

[8960, 1536]
blk.13.ffn_gate.weight

Q5_0

[1536, 8960]
blk.13.ffn_norm.weight

F32

[1536]
blk.13.ffn_up.weight

Q5_0

[1536, 8960]

blk.14

blk.14.attn_k.bias

F32

[256]
blk.14.attn_k.weight

Q5_0

[1536, 256]
blk.14.attn_norm.weight

F32

[1536]
blk.14.attn_output.weight

Q5_0

[1536, 1536]
blk.14.attn_q.bias

F32

[1536]
blk.14.attn_q.weight

Q5_0

[1536, 1536]
blk.14.attn_v.bias

F32

[256]
blk.14.attn_v.weight

Q5_0

[1536, 256]
blk.14.ffn_down.weight

Q5_0

[8960, 1536]
blk.14.ffn_gate.weight

Q5_0

[1536, 8960]
blk.14.ffn_norm.weight

F32

[1536]
blk.14.ffn_up.weight

Q5_0

[1536, 8960]

blk.15

blk.15.attn_k.bias

F32

[256]
blk.15.attn_k.weight

Q5_0

[1536, 256]
blk.15.attn_norm.weight

F32

[1536]
blk.15.attn_output.weight

Q5_0

[1536, 1536]
blk.15.attn_q.bias

F32

[1536]
blk.15.attn_q.weight

Q5_0

[1536, 1536]
blk.15.attn_v.bias

F32

[256]
blk.15.attn_v.weight

Q5_0

[1536, 256]
blk.15.ffn_down.weight

Q5_0

[8960, 1536]
blk.15.ffn_gate.weight

Q5_0

[1536, 8960]
blk.15.ffn_norm.weight

F32

[1536]
blk.15.ffn_up.weight

Q5_0

[1536, 8960]

blk.16

blk.16.attn_k.bias

F32

[256]
blk.16.attn_k.weight

Q5_0

[1536, 256]
blk.16.attn_norm.weight

F32

[1536]
blk.16.attn_output.weight

Q5_0

[1536, 1536]
blk.16.attn_q.bias

F32

[1536]
blk.16.attn_q.weight

Q5_0

[1536, 1536]
blk.16.attn_v.bias

F32

[256]
blk.16.attn_v.weight

Q5_0

[1536, 256]
blk.16.ffn_down.weight

Q5_0

[8960, 1536]
blk.16.ffn_gate.weight

Q5_0

[1536, 8960]
blk.16.ffn_norm.weight

F32

[1536]
blk.16.ffn_up.weight

Q5_0

[1536, 8960]

blk.17

blk.17.attn_k.bias

F32

[256]
blk.17.attn_k.weight

Q5_0

[1536, 256]
blk.17.attn_norm.weight

F32

[1536]
blk.17.attn_output.weight

Q5_0

[1536, 1536]
blk.17.attn_q.bias

F32

[1536]
blk.17.attn_q.weight

Q5_0

[1536, 1536]
blk.17.attn_v.bias

F32

[256]
blk.17.attn_v.weight

Q5_0

[1536, 256]
blk.17.ffn_down.weight

Q5_0

[8960, 1536]
blk.17.ffn_gate.weight

Q5_0

[1536, 8960]
blk.17.ffn_norm.weight

F32

[1536]
blk.17.ffn_up.weight

Q5_0

[1536, 8960]

blk.18

blk.18.attn_k.bias

F32

[256]
blk.18.attn_k.weight

Q5_0

[1536, 256]
blk.18.attn_norm.weight

F32

[1536]
blk.18.attn_output.weight

Q5_0

[1536, 1536]
blk.18.attn_q.bias

F32

[1536]
blk.18.attn_q.weight

Q5_0

[1536, 1536]
blk.18.attn_v.bias

F32

[256]
blk.18.attn_v.weight

Q5_0

[1536, 256]
blk.18.ffn_down.weight

Q5_0

[8960, 1536]
blk.18.ffn_gate.weight

Q5_0

[1536, 8960]
blk.18.ffn_norm.weight

F32

[1536]
blk.18.ffn_up.weight

Q5_0

[1536, 8960]

blk.19

blk.19.attn_k.bias

F32

[256]
blk.19.attn_k.weight

Q5_0

[1536, 256]
blk.19.attn_norm.weight

F32

[1536]
blk.19.attn_output.weight

Q5_0

[1536, 1536]
blk.19.attn_q.bias

F32

[1536]
blk.19.attn_q.weight

Q5_0

[1536, 1536]
blk.19.attn_v.bias

F32

[256]
blk.19.attn_v.weight

Q5_0

[1536, 256]
blk.19.ffn_down.weight

Q5_0

[8960, 1536]
blk.19.ffn_gate.weight

Q5_0

[1536, 8960]
blk.19.ffn_norm.weight

F32

[1536]
blk.19.ffn_up.weight

Q5_0

[1536, 8960]

blk.20

blk.20.attn_k.bias

F32

[256]
blk.20.attn_k.weight

Q5_0

[1536, 256]
blk.20.attn_norm.weight

F32

[1536]
blk.20.attn_output.weight

Q5_0

[1536, 1536]
blk.20.attn_q.bias

F32

[1536]
blk.20.attn_q.weight

Q5_0

[1536, 1536]
blk.20.attn_v.bias

F32

[256]
blk.20.attn_v.weight

Q5_0

[1536, 256]
blk.20.ffn_down.weight

Q5_0

[8960, 1536]
blk.20.ffn_gate.weight

Q5_0

[1536, 8960]
blk.20.ffn_norm.weight

F32

[1536]
blk.20.ffn_up.weight

Q5_0

[1536, 8960]

blk.21

blk.21.attn_k.bias

F32

[256]
blk.21.attn_k.weight

Q5_0

[1536, 256]
blk.21.attn_norm.weight

F32

[1536]
blk.21.attn_output.weight

Q5_0

[1536, 1536]
blk.21.attn_q.bias

F32

[1536]
blk.21.attn_q.weight

Q5_0

[1536, 1536]
blk.21.attn_v.bias

F32

[256]
blk.21.attn_v.weight

Q5_0

[1536, 256]
blk.21.ffn_down.weight

Q5_0

[8960, 1536]
blk.21.ffn_gate.weight

Q5_0

[1536, 8960]
blk.21.ffn_norm.weight

F32

[1536]
blk.21.ffn_up.weight

Q5_0

[1536, 8960]

blk.22

blk.22.attn_k.bias

F32

[256]
blk.22.attn_k.weight

Q5_0

[1536, 256]
blk.22.attn_norm.weight

F32

[1536]
blk.22.attn_output.weight

Q5_0

[1536, 1536]
blk.22.attn_q.bias

F32

[1536]
blk.22.attn_q.weight

Q5_0

[1536, 1536]
blk.22.attn_v.bias

F32

[256]
blk.22.attn_v.weight

Q5_0

[1536, 256]
blk.22.ffn_down.weight

Q5_0

[8960, 1536]
blk.22.ffn_gate.weight

Q5_0

[1536, 8960]
blk.22.ffn_norm.weight

F32

[1536]
blk.22.ffn_up.weight

Q5_0

[1536, 8960]

blk.23

blk.23.attn_k.bias

F32

[256]
blk.23.attn_k.weight

Q5_0

[1536, 256]
blk.23.attn_norm.weight

F32

[1536]
blk.23.attn_output.weight

Q5_0

[1536, 1536]
blk.23.attn_q.bias

F32

[1536]
blk.23.attn_q.weight

Q5_0

[1536, 1536]
blk.23.attn_v.bias

F32

[256]
blk.23.attn_v.weight

Q5_0

[1536, 256]
blk.23.ffn_down.weight

Q5_0

[8960, 1536]
blk.23.ffn_gate.weight

Q5_0

[1536, 8960]
blk.23.ffn_norm.weight

F32

[1536]
blk.23.ffn_up.weight

Q5_0

[1536, 8960]

blk.24

blk.24.attn_k.bias

F32

[256]
blk.24.attn_k.weight

Q5_0

[1536, 256]
blk.24.attn_norm.weight

F32

[1536]
blk.24.attn_output.weight

Q5_0

[1536, 1536]
blk.24.attn_q.bias

F32

[1536]
blk.24.attn_q.weight

Q5_0

[1536, 1536]
blk.24.attn_v.bias

F32

[256]
blk.24.attn_v.weight

Q5_0

[1536, 256]
blk.24.ffn_down.weight

Q5_0

[8960, 1536]
blk.24.ffn_gate.weight

Q5_0

[1536, 8960]
blk.24.ffn_norm.weight

F32

[1536]
blk.24.ffn_up.weight

Q5_0

[1536, 8960]

blk.25

blk.25.attn_k.bias

F32

[256]
blk.25.attn_k.weight

Q5_0

[1536, 256]
blk.25.attn_norm.weight

F32

[1536]
blk.25.attn_output.weight

Q5_0

[1536, 1536]
blk.25.attn_q.bias

F32

[1536]
blk.25.attn_q.weight

Q5_0

[1536, 1536]
blk.25.attn_v.bias

F32

[256]
blk.25.attn_v.weight

Q5_0

[1536, 256]
blk.25.ffn_down.weight

Q5_0

[8960, 1536]
blk.25.ffn_gate.weight

Q5_0

[1536, 8960]
blk.25.ffn_norm.weight

F32

[1536]
blk.25.ffn_up.weight

Q5_0

[1536, 8960]

blk.26

blk.26.attn_k.bias

F32

[256]
blk.26.attn_k.weight

Q5_0

[1536, 256]
blk.26.attn_norm.weight

F32

[1536]
blk.26.attn_output.weight

Q5_0

[1536, 1536]
blk.26.attn_q.bias

F32

[1536]
blk.26.attn_q.weight

Q5_0

[1536, 1536]
blk.26.attn_v.bias

F32

[256]
blk.26.attn_v.weight

Q5_0

[1536, 256]
blk.26.ffn_down.weight

Q5_0

[8960, 1536]
blk.26.ffn_gate.weight

Q5_0

[1536, 8960]
blk.26.ffn_norm.weight

F32

[1536]
blk.26.ffn_up.weight

Q5_0

[1536, 8960]

blk.27

blk.27.attn_k.bias

F32

[256]
blk.27.attn_k.weight

Q5_0

[1536, 256]
blk.27.attn_norm.weight

F32

[1536]
blk.27.attn_output.weight

Q5_0

[1536, 1536]
blk.27.attn_q.bias

F32

[1536]
blk.27.attn_q.weight

Q5_0

[1536, 1536]
blk.27.attn_v.bias

F32

[256]
blk.27.attn_v.weight

Q5_0

[1536, 256]
blk.27.ffn_down.weight

Q5_0

[8960, 1536]
blk.27.ffn_gate.weight

Q5_0

[1536, 8960]
blk.27.ffn_norm.weight

F32

[1536]
blk.27.ffn_up.weight

Q5_0

[1536, 8960]

output_norm.weight

F32

[1536]