nebel / fugaku-llm

Fugaku-LLM

1,347 Pulls Updated 6 months ago

1 Tag

fugaku-llm:13b-instruct ... /

model

b6999c1a64a5 · 11GB

Metadata

general.architecture

gpt2
general.file_type

17
general.name

fugakullm_0325best
general.quantization_version

2
gpt2.attention.head_count

36
gpt2.attention.layer_norm_epsilon

1e-05
gpt2.block_count

40
gpt2.context_length

2048
gpt2.embedding_length

5184
gpt2.feed_forward_length

20736
tokenizer.ggml.bos_token_id

1
tokenizer.ggml.cls_token_id

5
tokenizer.ggml.eos_token_id

7
tokenizer.ggml.mask_token_id

3
tokenizer.ggml.model

llama
tokenizer.ggml.padding_token_id

4
tokenizer.ggml.scores

[-1000, -1000, -1000, -1000, -1000, ...]
tokenizer.ggml.seperator_token_id

6
tokenizer.ggml.token_type

[3, 3, 3, 3, 3, ...]
tokenizer.ggml.tokens

[<unk|LLM-jp>, <s|LLM-jp>, </s|LLM-jp>, <mask|LLM-jp>, <pad|LLM-jp>, ...]
tokenizer.ggml.unknown_token_id

0

Tensor

Name

Type

Shape
token_embd.weight

Q5_1

[5184, 49152]

blk.0

blk.0.attn_norm.bias

F32

[5184]
blk.0.attn_norm.weight

F32

[5184]
blk.0.attn_output.bias

F32

[5184]
blk.0.attn_output.weight

Q5_1

[5184, 5184]
blk.0.attn_qkv.bias

F32

[15552]
blk.0.attn_qkv.weight

Q8_0

[5184, 15552]
blk.0.ffn_down.bias

F32

[5184]
blk.0.ffn_down.weight

Q6_K

[20736, 5184]
blk.0.ffn_norm.bias

F32

[5184]
blk.0.ffn_norm.weight

F32

[5184]
blk.0.ffn_up.bias

F32

[20736]
blk.0.ffn_up.weight

Q5_1

[5184, 20736]

blk.1

blk.1.attn_norm.bias

F32

[5184]
blk.1.attn_norm.weight

F32

[5184]
blk.1.attn_output.bias

F32

[5184]
blk.1.attn_output.weight

Q5_1

[5184, 5184]
blk.1.attn_qkv.bias

F32

[15552]
blk.1.attn_qkv.weight

Q8_0

[5184, 15552]
blk.1.ffn_down.bias

F32

[5184]
blk.1.ffn_down.weight

Q6_K

[20736, 5184]
blk.1.ffn_norm.bias

F32

[5184]
blk.1.ffn_norm.weight

F32

[5184]
blk.1.ffn_up.bias

F32

[20736]
blk.1.ffn_up.weight

Q5_1

[5184, 20736]

blk.2

blk.2.attn_norm.bias

F32

[5184]
blk.2.attn_norm.weight

F32

[5184]
blk.2.attn_output.bias

F32

[5184]
blk.2.attn_output.weight

Q5_1

[5184, 5184]
blk.2.attn_qkv.bias

F32

[15552]
blk.2.attn_qkv.weight

Q8_0

[5184, 15552]
blk.2.ffn_down.bias

F32

[5184]
blk.2.ffn_down.weight

Q6_K

[20736, 5184]
blk.2.ffn_norm.bias

F32

[5184]
blk.2.ffn_norm.weight

F32

[5184]
blk.2.ffn_up.bias

F32

[20736]
blk.2.ffn_up.weight

Q5_1

[5184, 20736]

blk.3

blk.3.attn_norm.bias

F32

[5184]
blk.3.attn_norm.weight

F32

[5184]
blk.3.attn_output.bias

F32

[5184]
blk.3.attn_output.weight

Q5_1

[5184, 5184]
blk.3.attn_qkv.bias

F32

[15552]
blk.3.attn_qkv.weight

Q8_0

[5184, 15552]
blk.3.ffn_down.bias

F32

[5184]
blk.3.ffn_down.weight

Q6_K

[20736, 5184]
blk.3.ffn_norm.bias

F32

[5184]
blk.3.ffn_norm.weight

F32

[5184]
blk.3.ffn_up.bias

F32

[20736]
blk.3.ffn_up.weight

Q5_1

[5184, 20736]

blk.4

blk.4.attn_norm.bias

F32

[5184]
blk.4.attn_norm.weight

F32

[5184]
blk.4.attn_output.bias

F32

[5184]
blk.4.attn_output.weight

Q5_1

[5184, 5184]
blk.4.attn_qkv.bias

F32

[15552]
blk.4.attn_qkv.weight

Q8_0

[5184, 15552]
blk.4.ffn_down.bias

F32

[5184]
blk.4.ffn_down.weight

Q6_K

[20736, 5184]
blk.4.ffn_norm.bias

F32

[5184]
blk.4.ffn_norm.weight

F32

[5184]
blk.4.ffn_up.bias

F32

[20736]
blk.4.ffn_up.weight

Q5_1

[5184, 20736]

blk.5

blk.5.attn_norm.bias

F32

[5184]
blk.5.attn_norm.weight

F32

[5184]
blk.5.attn_output.bias

F32

[5184]
blk.5.attn_output.weight

Q5_1

[5184, 5184]
blk.5.attn_qkv.bias

F32

[15552]
blk.5.attn_qkv.weight

Q8_0

[5184, 15552]
blk.5.ffn_down.bias

F32

[5184]
blk.5.ffn_down.weight

Q5_K

[20736, 5184]
blk.5.ffn_norm.bias

F32

[5184]
blk.5.ffn_norm.weight

F32

[5184]
blk.5.ffn_up.bias

F32

[20736]
blk.5.ffn_up.weight

Q5_1

[5184, 20736]

blk.6

blk.6.attn_norm.bias

F32

[5184]
blk.6.attn_norm.weight

F32

[5184]
blk.6.attn_output.bias

F32

[5184]
blk.6.attn_output.weight

Q5_1

[5184, 5184]
blk.6.attn_qkv.bias

F32

[15552]
blk.6.attn_qkv.weight

Q8_0

[5184, 15552]
blk.6.ffn_down.bias

F32

[5184]
blk.6.ffn_down.weight

Q6_K

[20736, 5184]
blk.6.ffn_norm.bias

F32

[5184]
blk.6.ffn_norm.weight

F32

[5184]
blk.6.ffn_up.bias

F32

[20736]
blk.6.ffn_up.weight

Q5_1

[5184, 20736]

blk.7

blk.7.attn_norm.bias

F32

[5184]
blk.7.attn_norm.weight

F32

[5184]
blk.7.attn_output.bias

F32

[5184]
blk.7.attn_output.weight

Q5_1

[5184, 5184]
blk.7.attn_qkv.bias

F32

[15552]
blk.7.attn_qkv.weight

Q8_0

[5184, 15552]
blk.7.ffn_down.bias

F32

[5184]
blk.7.ffn_down.weight

Q5_K

[20736, 5184]
blk.7.ffn_norm.bias

F32

[5184]
blk.7.ffn_norm.weight

F32

[5184]
blk.7.ffn_up.bias

F32

[20736]
blk.7.ffn_up.weight

Q5_1

[5184, 20736]

blk.8

blk.8.attn_norm.bias

F32

[5184]
blk.8.attn_norm.weight

F32

[5184]
blk.8.attn_output.bias

F32

[5184]
blk.8.attn_output.weight

Q5_1

[5184, 5184]
blk.8.attn_qkv.bias

F32

[15552]
blk.8.attn_qkv.weight

Q8_0

[5184, 15552]
blk.8.ffn_down.bias

F32

[5184]
blk.8.ffn_down.weight

Q5_K

[20736, 5184]
blk.8.ffn_norm.bias

F32

[5184]
blk.8.ffn_norm.weight

F32

[5184]
blk.8.ffn_up.bias

F32

[20736]
blk.8.ffn_up.weight

Q5_1

[5184, 20736]

blk.9

blk.9.attn_norm.bias

F32

[5184]
blk.9.attn_norm.weight

F32

[5184]
blk.9.attn_output.bias

F32

[5184]
blk.9.attn_output.weight

Q5_1

[5184, 5184]
blk.9.attn_qkv.bias

F32

[15552]
blk.9.attn_qkv.weight

Q8_0

[5184, 15552]
blk.9.ffn_down.bias

F32

[5184]
blk.9.ffn_down.weight

Q6_K

[20736, 5184]
blk.9.ffn_norm.bias

F32

[5184]
blk.9.ffn_norm.weight

F32

[5184]
blk.9.ffn_up.bias

F32

[20736]
blk.9.ffn_up.weight

Q5_1

[5184, 20736]

blk.10

blk.10.attn_norm.bias

F32

[5184]
blk.10.attn_norm.weight

F32

[5184]
blk.10.attn_output.bias

F32

[5184]
blk.10.attn_output.weight

Q5_1

[5184, 5184]
blk.10.attn_qkv.bias

F32

[15552]
blk.10.attn_qkv.weight

Q8_0

[5184, 15552]
blk.10.ffn_down.bias

F32

[5184]
blk.10.ffn_down.weight

Q5_K

[20736, 5184]
blk.10.ffn_norm.bias

F32

[5184]
blk.10.ffn_norm.weight

F32

[5184]
blk.10.ffn_up.bias

F32

[20736]
blk.10.ffn_up.weight

Q5_1

[5184, 20736]

blk.11

blk.11.attn_norm.bias

F32

[5184]
blk.11.attn_norm.weight

F32

[5184]
blk.11.attn_output.bias

F32

[5184]
blk.11.attn_output.weight

Q5_1

[5184, 5184]
blk.11.attn_qkv.bias

F32

[15552]
blk.11.attn_qkv.weight

Q8_0

[5184, 15552]
blk.11.ffn_down.bias

F32

[5184]
blk.11.ffn_down.weight

Q6_K

[20736, 5184]
blk.11.ffn_norm.bias

F32

[5184]
blk.11.ffn_norm.weight

F32

[5184]
blk.11.ffn_up.bias

F32

[20736]
blk.11.ffn_up.weight

Q5_1

[5184, 20736]

blk.12

blk.12.attn_norm.bias

F32

[5184]
blk.12.attn_norm.weight

F32

[5184]
blk.12.attn_output.bias

F32

[5184]
blk.12.attn_output.weight

Q5_1

[5184, 5184]
blk.12.attn_qkv.bias

F32

[15552]
blk.12.attn_qkv.weight

Q8_0

[5184, 15552]
blk.12.ffn_down.bias

F32

[5184]
blk.12.ffn_down.weight

Q5_K

[20736, 5184]
blk.12.ffn_norm.bias

F32

[5184]
blk.12.ffn_norm.weight

F32

[5184]
blk.12.ffn_up.bias

F32

[20736]
blk.12.ffn_up.weight

Q5_1

[5184, 20736]

blk.13

blk.13.attn_norm.bias

F32

[5184]
blk.13.attn_norm.weight

F32

[5184]
blk.13.attn_output.bias

F32

[5184]
blk.13.attn_output.weight

Q5_1

[5184, 5184]
blk.13.attn_qkv.bias

F32

[15552]
blk.13.attn_qkv.weight

Q8_0

[5184, 15552]
blk.13.ffn_down.bias

F32

[5184]
blk.13.ffn_down.weight

Q5_K

[20736, 5184]
blk.13.ffn_norm.bias

F32

[5184]
blk.13.ffn_norm.weight

F32

[5184]
blk.13.ffn_up.bias

F32

[20736]
blk.13.ffn_up.weight

Q5_1

[5184, 20736]

blk.14

blk.14.attn_norm.bias

F32

[5184]
blk.14.attn_norm.weight

F32

[5184]
blk.14.attn_output.bias

F32

[5184]
blk.14.attn_output.weight

Q5_1

[5184, 5184]
blk.14.attn_qkv.bias

F32

[15552]
blk.14.attn_qkv.weight

Q8_0

[5184, 15552]
blk.14.ffn_down.bias

F32

[5184]
blk.14.ffn_down.weight

Q5_K

[20736, 5184]
blk.14.ffn_norm.bias

F32

[5184]
blk.14.ffn_norm.weight

F32

[5184]
blk.14.ffn_up.bias

F32

[20736]
blk.14.ffn_up.weight

Q5_1

[5184, 20736]

blk.15

blk.15.attn_norm.bias

F32

[5184]
blk.15.attn_norm.weight

F32

[5184]
blk.15.attn_output.bias

F32

[5184]
blk.15.attn_output.weight

Q5_1

[5184, 5184]
blk.15.attn_qkv.bias

F32

[15552]
blk.15.attn_qkv.weight

Q8_0

[5184, 15552]
blk.15.ffn_down.bias

F32

[5184]
blk.15.ffn_down.weight

Q5_K

[20736, 5184]
blk.15.ffn_norm.bias

F32

[5184]
blk.15.ffn_norm.weight

F32

[5184]
blk.15.ffn_up.bias

F32

[20736]
blk.15.ffn_up.weight

Q5_1

[5184, 20736]

blk.16

blk.16.attn_norm.bias

F32

[5184]
blk.16.attn_norm.weight

F32

[5184]
blk.16.attn_output.bias

F32

[5184]
blk.16.attn_output.weight

Q5_1

[5184, 5184]
blk.16.attn_qkv.bias

F32

[15552]
blk.16.attn_qkv.weight

Q8_0

[5184, 15552]
blk.16.ffn_down.bias

F32

[5184]
blk.16.ffn_down.weight

Q6_K

[20736, 5184]
blk.16.ffn_norm.bias

F32

[5184]
blk.16.ffn_norm.weight

F32

[5184]
blk.16.ffn_up.bias

F32

[20736]
blk.16.ffn_up.weight

Q5_1

[5184, 20736]

blk.17

blk.17.attn_norm.bias

F32

[5184]
blk.17.attn_norm.weight

F32

[5184]
blk.17.attn_output.bias

F32

[5184]
blk.17.attn_output.weight

Q5_1

[5184, 5184]
blk.17.attn_qkv.bias

F32

[15552]
blk.17.attn_qkv.weight

Q8_0

[5184, 15552]
blk.17.ffn_down.bias

F32

[5184]
blk.17.ffn_down.weight

Q5_K

[20736, 5184]
blk.17.ffn_norm.bias

F32

[5184]
blk.17.ffn_norm.weight

F32

[5184]
blk.17.ffn_up.bias

F32

[20736]
blk.17.ffn_up.weight

Q5_1

[5184, 20736]

blk.18

blk.18.attn_norm.bias

F32

[5184]
blk.18.attn_norm.weight

F32

[5184]
blk.18.attn_output.bias

F32

[5184]
blk.18.attn_output.weight

Q5_1

[5184, 5184]
blk.18.attn_qkv.bias

F32

[15552]
blk.18.attn_qkv.weight

Q8_0

[5184, 15552]
blk.18.ffn_down.bias

F32

[5184]
blk.18.ffn_down.weight

Q5_K

[20736, 5184]
blk.18.ffn_norm.bias

F32

[5184]
blk.18.ffn_norm.weight

F32

[5184]
blk.18.ffn_up.bias

F32

[20736]
blk.18.ffn_up.weight

Q5_1

[5184, 20736]

blk.19

blk.19.attn_norm.bias

F32

[5184]
blk.19.attn_norm.weight

F32

[5184]
blk.19.attn_output.bias

F32

[5184]
blk.19.attn_output.weight

Q5_1

[5184, 5184]
blk.19.attn_qkv.bias

F32

[15552]
blk.19.attn_qkv.weight

Q8_0

[5184, 15552]
blk.19.ffn_down.bias

F32

[5184]
blk.19.ffn_down.weight

Q6_K

[20736, 5184]
blk.19.ffn_norm.bias

F32

[5184]
blk.19.ffn_norm.weight

F32

[5184]
blk.19.ffn_up.bias

F32

[20736]
blk.19.ffn_up.weight

Q5_1

[5184, 20736]

blk.20

blk.20.attn_norm.bias

F32

[5184]
blk.20.attn_norm.weight

F32

[5184]
blk.20.attn_output.bias

F32

[5184]
blk.20.attn_output.weight

Q5_1

[5184, 5184]
blk.20.attn_qkv.bias

F32

[15552]
blk.20.attn_qkv.weight

Q8_0

[5184, 15552]
blk.20.ffn_down.bias

F32

[5184]
blk.20.ffn_down.weight

Q5_K

[20736, 5184]
blk.20.ffn_norm.bias

F32

[5184]
blk.20.ffn_norm.weight

F32

[5184]
blk.20.ffn_up.bias

F32

[20736]
blk.20.ffn_up.weight

Q5_1

[5184, 20736]

blk.21

blk.21.attn_norm.bias

F32

[5184]
blk.21.attn_norm.weight

F32

[5184]
blk.21.attn_output.bias

F32

[5184]
blk.21.attn_output.weight

Q5_1

[5184, 5184]
blk.21.attn_qkv.bias

F32

[15552]
blk.21.attn_qkv.weight

Q8_0

[5184, 15552]
blk.21.ffn_down.bias

F32

[5184]
blk.21.ffn_down.weight

Q5_K

[20736, 5184]
blk.21.ffn_norm.bias

F32

[5184]
blk.21.ffn_norm.weight

F32

[5184]
blk.21.ffn_up.bias

F32

[20736]
blk.21.ffn_up.weight

Q5_1

[5184, 20736]

blk.22

blk.22.attn_norm.bias

F32

[5184]
blk.22.attn_norm.weight

F32

[5184]
blk.22.attn_output.bias

F32

[5184]
blk.22.attn_output.weight

Q5_1

[5184, 5184]
blk.22.attn_qkv.bias

F32

[15552]
blk.22.attn_qkv.weight

Q8_0

[5184, 15552]
blk.22.ffn_down.bias

F32

[5184]
blk.22.ffn_down.weight

Q6_K

[20736, 5184]
blk.22.ffn_norm.bias

F32

[5184]
blk.22.ffn_norm.weight

F32

[5184]
blk.22.ffn_up.bias

F32

[20736]
blk.22.ffn_up.weight

Q5_1

[5184, 20736]

blk.23

blk.23.attn_norm.bias

F32

[5184]
blk.23.attn_norm.weight

F32

[5184]
blk.23.attn_output.bias

F32

[5184]
blk.23.attn_output.weight

Q5_1

[5184, 5184]
blk.23.attn_qkv.bias

F32

[15552]
blk.23.attn_qkv.weight

Q8_0

[5184, 15552]
blk.23.ffn_down.bias

F32

[5184]
blk.23.ffn_down.weight

Q5_K

[20736, 5184]
blk.23.ffn_norm.bias

F32

[5184]
blk.23.ffn_norm.weight

F32

[5184]
blk.23.ffn_up.bias

F32

[20736]
blk.23.ffn_up.weight

Q5_1

[5184, 20736]

blk.24

blk.24.attn_norm.bias

F32

[5184]
blk.24.attn_norm.weight

F32

[5184]
blk.24.attn_output.bias

F32

[5184]
blk.24.attn_output.weight

Q5_1

[5184, 5184]
blk.24.attn_qkv.bias

F32

[15552]
blk.24.attn_qkv.weight

Q8_0

[5184, 15552]
blk.24.ffn_down.bias

F32

[5184]
blk.24.ffn_down.weight

Q5_K

[20736, 5184]
blk.24.ffn_norm.bias

F32

[5184]
blk.24.ffn_norm.weight

F32

[5184]
blk.24.ffn_up.bias

F32

[20736]
blk.24.ffn_up.weight

Q5_1

[5184, 20736]

blk.25

blk.25.attn_norm.bias

F32

[5184]
blk.25.attn_norm.weight

F32

[5184]
blk.25.attn_output.bias

F32

[5184]
blk.25.attn_output.weight

Q5_1

[5184, 5184]
blk.25.attn_qkv.bias

F32

[15552]
blk.25.attn_qkv.weight

Q8_0

[5184, 15552]
blk.25.ffn_down.bias

F32

[5184]
blk.25.ffn_down.weight

Q6_K

[20736, 5184]
blk.25.ffn_norm.bias

F32

[5184]
blk.25.ffn_norm.weight

F32

[5184]
blk.25.ffn_up.bias

F32

[20736]
blk.25.ffn_up.weight

Q5_1

[5184, 20736]

blk.26

blk.26.attn_norm.bias

F32

[5184]
blk.26.attn_norm.weight

F32

[5184]
blk.26.attn_output.bias

F32

[5184]
blk.26.attn_output.weight

Q5_1

[5184, 5184]
blk.26.attn_qkv.bias

F32

[15552]
blk.26.attn_qkv.weight

Q8_0

[5184, 15552]
blk.26.ffn_down.bias

F32

[5184]
blk.26.ffn_down.weight

Q5_K

[20736, 5184]
blk.26.ffn_norm.bias

F32

[5184]
blk.26.ffn_norm.weight

F32

[5184]
blk.26.ffn_up.bias

F32

[20736]
blk.26.ffn_up.weight

Q5_1

[5184, 20736]

blk.27

blk.27.attn_norm.bias

F32

[5184]
blk.27.attn_norm.weight

F32

[5184]
blk.27.attn_output.bias

F32

[5184]
blk.27.attn_output.weight

Q5_1

[5184, 5184]
blk.27.attn_qkv.bias

F32

[15552]
blk.27.attn_qkv.weight

Q8_0

[5184, 15552]
blk.27.ffn_down.bias

F32

[5184]
blk.27.ffn_down.weight

Q5_K

[20736, 5184]
blk.27.ffn_norm.bias

F32

[5184]
blk.27.ffn_norm.weight

F32

[5184]
blk.27.ffn_up.bias

F32

[20736]
blk.27.ffn_up.weight

Q5_1

[5184, 20736]

blk.28

blk.28.attn_norm.bias

F32

[5184]
blk.28.attn_norm.weight

F32

[5184]
blk.28.attn_output.bias

F32

[5184]
blk.28.attn_output.weight

Q5_1

[5184, 5184]
blk.28.attn_qkv.bias

F32

[15552]
blk.28.attn_qkv.weight

Q8_0

[5184, 15552]
blk.28.ffn_down.bias

F32

[5184]
blk.28.ffn_down.weight

Q6_K

[20736, 5184]
blk.28.ffn_norm.bias

F32

[5184]
blk.28.ffn_norm.weight

F32

[5184]
blk.28.ffn_up.bias

F32

[20736]
blk.28.ffn_up.weight

Q5_1

[5184, 20736]

blk.29

blk.29.attn_norm.bias

F32

[5184]
blk.29.attn_norm.weight

F32

[5184]
blk.29.attn_output.bias

F32

[5184]
blk.29.attn_output.weight

Q5_1

[5184, 5184]
blk.29.attn_qkv.bias

F32

[15552]
blk.29.attn_qkv.weight

Q8_0

[5184, 15552]
blk.29.ffn_down.bias

F32

[5184]
blk.29.ffn_down.weight

Q5_K

[20736, 5184]
blk.29.ffn_norm.bias

F32

[5184]
blk.29.ffn_norm.weight

F32

[5184]
blk.29.ffn_up.bias

F32

[20736]
blk.29.ffn_up.weight

Q5_1

[5184, 20736]

blk.30

blk.30.attn_norm.bias

F32

[5184]
blk.30.attn_norm.weight

F32

[5184]
blk.30.attn_output.bias

F32

[5184]
blk.30.attn_output.weight

Q5_1

[5184, 5184]
blk.30.attn_qkv.bias

F32

[15552]
blk.30.attn_qkv.weight

Q8_0

[5184, 15552]
blk.30.ffn_down.bias

F32

[5184]
blk.30.ffn_down.weight

Q5_K

[20736, 5184]
blk.30.ffn_norm.bias

F32

[5184]
blk.30.ffn_norm.weight

F32

[5184]
blk.30.ffn_up.bias

F32

[20736]
blk.30.ffn_up.weight

Q5_1

[5184, 20736]

blk.31

blk.31.attn_norm.bias

F32

[5184]
blk.31.attn_norm.weight

F32

[5184]
blk.31.attn_output.bias

F32

[5184]
blk.31.attn_output.weight

Q5_1

[5184, 5184]
blk.31.attn_qkv.bias

F32

[15552]
blk.31.attn_qkv.weight

Q8_0

[5184, 15552]
blk.31.ffn_down.bias

F32

[5184]
blk.31.ffn_down.weight

Q6_K

[20736, 5184]
blk.31.ffn_norm.bias

F32

[5184]
blk.31.ffn_norm.weight

F32

[5184]
blk.31.ffn_up.bias

F32

[20736]
blk.31.ffn_up.weight

Q5_1

[5184, 20736]

blk.32

blk.32.attn_norm.bias

F32

[5184]
blk.32.attn_norm.weight

F32

[5184]
blk.32.attn_output.bias

F32

[5184]
blk.32.attn_output.weight

Q5_1

[5184, 5184]
blk.32.attn_qkv.bias

F32

[15552]
blk.32.attn_qkv.weight

Q8_0

[5184, 15552]
blk.32.ffn_down.bias

F32

[5184]
blk.32.ffn_down.weight

Q5_K

[20736, 5184]
blk.32.ffn_norm.bias

F32

[5184]
blk.32.ffn_norm.weight

F32

[5184]
blk.32.ffn_up.bias

F32

[20736]
blk.32.ffn_up.weight

Q5_1

[5184, 20736]

blk.33

blk.33.attn_norm.bias

F32

[5184]
blk.33.attn_norm.weight

F32

[5184]
blk.33.attn_output.bias

F32

[5184]
blk.33.attn_output.weight

Q5_1

[5184, 5184]
blk.33.attn_qkv.bias

F32

[15552]
blk.33.attn_qkv.weight

Q8_0

[5184, 15552]
blk.33.ffn_down.bias

F32

[5184]
blk.33.ffn_down.weight

Q5_K

[20736, 5184]
blk.33.ffn_norm.bias

F32

[5184]
blk.33.ffn_norm.weight

F32

[5184]
blk.33.ffn_up.bias

F32

[20736]
blk.33.ffn_up.weight

Q5_1

[5184, 20736]

blk.34

blk.34.attn_norm.bias

F32

[5184]
blk.34.attn_norm.weight

F32

[5184]
blk.34.attn_output.bias

F32

[5184]
blk.34.attn_output.weight

Q5_1

[5184, 5184]
blk.34.attn_qkv.bias

F32

[15552]
blk.34.attn_qkv.weight

Q8_0

[5184, 15552]
blk.34.ffn_down.bias

F32

[5184]
blk.34.ffn_down.weight

Q6_K

[20736, 5184]
blk.34.ffn_norm.bias

F32

[5184]
blk.34.ffn_norm.weight

F32

[5184]
blk.34.ffn_up.bias

F32

[20736]
blk.34.ffn_up.weight

Q5_1

[5184, 20736]

blk.35

blk.35.attn_norm.bias

F32

[5184]
blk.35.attn_norm.weight

F32

[5184]
blk.35.attn_output.bias

F32

[5184]
blk.35.attn_output.weight

Q5_1

[5184, 5184]
blk.35.attn_qkv.bias

F32

[15552]
blk.35.attn_qkv.weight

Q8_0

[5184, 15552]
blk.35.ffn_down.bias

F32

[5184]
blk.35.ffn_down.weight

Q6_K

[20736, 5184]
blk.35.ffn_norm.bias

F32

[5184]
blk.35.ffn_norm.weight

F32

[5184]
blk.35.ffn_up.bias

F32

[20736]
blk.35.ffn_up.weight

Q5_1

[5184, 20736]

blk.36

blk.36.attn_norm.bias

F32

[5184]
blk.36.attn_norm.weight

F32

[5184]
blk.36.attn_output.bias

F32

[5184]
blk.36.attn_output.weight

Q5_1

[5184, 5184]
blk.36.attn_qkv.bias

F32

[15552]
blk.36.attn_qkv.weight

Q8_0

[5184, 15552]
blk.36.ffn_down.bias

F32

[5184]
blk.36.ffn_down.weight

Q6_K

[20736, 5184]
blk.36.ffn_norm.bias

F32

[5184]
blk.36.ffn_norm.weight

F32

[5184]
blk.36.ffn_up.bias

F32

[20736]
blk.36.ffn_up.weight

Q5_1

[5184, 20736]

blk.37

blk.37.attn_norm.bias

F32

[5184]
blk.37.attn_norm.weight

F32

[5184]
blk.37.attn_output.bias

F32

[5184]
blk.37.attn_output.weight

Q5_1

[5184, 5184]
blk.37.attn_qkv.bias

F32

[15552]
blk.37.attn_qkv.weight

Q8_0

[5184, 15552]
blk.37.ffn_down.bias

F32

[5184]
blk.37.ffn_down.weight

Q6_K

[20736, 5184]
blk.37.ffn_norm.bias

F32

[5184]
blk.37.ffn_norm.weight

F32

[5184]
blk.37.ffn_up.bias

F32

[20736]
blk.37.ffn_up.weight

Q5_1

[5184, 20736]

blk.38

blk.38.attn_norm.bias

F32

[5184]
blk.38.attn_norm.weight

F32

[5184]
blk.38.attn_output.bias

F32

[5184]
blk.38.attn_output.weight

Q5_1

[5184, 5184]
blk.38.attn_qkv.bias

F32

[15552]
blk.38.attn_qkv.weight

Q8_0

[5184, 15552]
blk.38.ffn_down.bias

F32

[5184]
blk.38.ffn_down.weight

Q6_K

[20736, 5184]
blk.38.ffn_norm.bias

F32

[5184]
blk.38.ffn_norm.weight

F32

[5184]
blk.38.ffn_up.bias

F32

[20736]
blk.38.ffn_up.weight

Q5_1

[5184, 20736]

blk.39

blk.39.attn_norm.bias

F32

[5184]
blk.39.attn_norm.weight

F32

[5184]
blk.39.attn_output.bias

F32

[5184]
blk.39.attn_output.weight

Q5_1

[5184, 5184]
blk.39.attn_qkv.bias

F32

[15552]
blk.39.attn_qkv.weight

Q8_0

[5184, 15552]
blk.39.ffn_down.bias

F32

[5184]
blk.39.ffn_down.weight

Q6_K

[20736, 5184]
blk.39.ffn_norm.bias

F32

[5184]
blk.39.ffn_norm.weight

F32

[5184]
blk.39.ffn_up.bias

F32

[20736]
blk.39.ffn_up.weight

Q5_1

[5184, 20736]

output.weight

Q8_0

[5184, 49152]
output_norm.bias

F32

[5184]
position_embd.weight

F16

[5184, 2048]
output_norm.weight

F32

[5184]