Models
GitHub
Discord
Docs
Cloud
Sign in
Download
Models
Download
GitHub
Discord
Docs
Cloud
Sign in
Uynaity
/
rui
:3b
8
Downloads
Updated
1 year ago
Based on the provided weather and events, automatically generate the copy for RIC Rui.
Based on the provided weather and events, automatically generate the copy for RIC Rui.
Cancel
3b
rui:3b
...
/
model
8381a63bb271 · 6.2GB
Metadata
general.architecture
qwen2
qwen2
general.file_type
F16
F16
qwen2.attention.head_count
16
16
qwen2.attention.head_count_kv
2
2
qwen2.attention.layer_norm_rms_epsilon
1e-06
1e-06
qwen2.block_count
36
36
qwen2.context_length
32768
32768
qwen2.embedding_length
2048
2048
qwen2.feed_forward_length
11008
11008
qwen2.rope.freq_base
1e+06
1e+06
tokenizer.ggml.add_bos_token
false
false
tokenizer.ggml.bos_token_id
128245
128245
tokenizer.ggml.eos_token_id
151645
151645
tokenizer.ggml.merges
[Ġ Ġ, ĠĠ ĠĠ, i n, Ġ t, ĠĠĠĠ ĠĠĠĠ, ...]
[Ġ Ġ, ĠĠ ĠĠ, i n, Ġ t, ĠĠĠĠ ĠĠĠĠ, ...]
tokenizer.ggml.model
gpt2
gpt2
tokenizer.ggml.padding_token_id
151665
151665
tokenizer.ggml.pre
qwen2
qwen2
tokenizer.ggml.token_type
[1, 1, 1, 1, 1, ...]
[1, 1, 1, 1, 1, ...]
tokenizer.ggml.tokens
[!, ", #, $, %, ...]
[!, ", #, $, %, ...]
Tensor
Name
Type
Shape
token_embd.weight
F16
F16
[2048, 151669]
blk.0
blk.0.attn_k.bias
F32
F32
[256]
blk.0.attn_k.weight
F16
F16
[2048, 256]
blk.0.attn_norm.weight
F32
F32
[2048]
blk.0.attn_output.weight
F16
F16
[2048, 2048]
blk.0.attn_q.bias
F32
F32
[2048]
blk.0.attn_q.weight
F16
F16
[2048, 2048]
blk.0.attn_v.bias
F32
F32
[256]
blk.0.attn_v.weight
F16
F16
[2048, 256]
blk.0.ffn_down.weight
F16
F16
[11008, 2048]
blk.0.ffn_gate.weight
F16
F16
[2048, 11008]
blk.0.ffn_norm.weight
F32
F32
[2048]
blk.0.ffn_up.weight
F16
F16
[2048, 11008]
blk.1
blk.1.attn_k.bias
F32
F32
[256]
blk.1.attn_k.weight
F16
F16
[2048, 256]
blk.1.attn_norm.weight
F32
F32
[2048]
blk.1.attn_output.weight
F16
F16
[2048, 2048]
blk.1.attn_q.bias
F32
F32
[2048]
blk.1.attn_q.weight
F16
F16
[2048, 2048]
blk.1.attn_v.bias
F32
F32
[256]
blk.1.attn_v.weight
F16
F16
[2048, 256]
blk.1.ffn_down.weight
F16
F16
[11008, 2048]
blk.1.ffn_gate.weight
F16
F16
[2048, 11008]
blk.1.ffn_norm.weight
F32
F32
[2048]
blk.1.ffn_up.weight
F16
F16
[2048, 11008]
blk.2
blk.2.attn_k.bias
F32
F32
[256]
blk.2.attn_k.weight
F16
F16
[2048, 256]
blk.2.attn_norm.weight
F32
F32
[2048]
blk.2.attn_output.weight
F16
F16
[2048, 2048]
blk.2.attn_q.bias
F32
F32
[2048]
blk.2.attn_q.weight
F16
F16
[2048, 2048]
blk.2.attn_v.bias
F32
F32
[256]
blk.2.attn_v.weight
F16
F16
[2048, 256]
blk.2.ffn_down.weight
F16
F16
[11008, 2048]
blk.2.ffn_gate.weight
F16
F16
[2048, 11008]
blk.2.ffn_norm.weight
F32
F32
[2048]
blk.2.ffn_up.weight
F16
F16
[2048, 11008]
blk.3
blk.3.attn_k.bias
F32
F32
[256]
blk.3.attn_k.weight
F16
F16
[2048, 256]
blk.3.attn_norm.weight
F32
F32
[2048]
blk.3.attn_output.weight
F16
F16
[2048, 2048]
blk.3.attn_q.bias
F32
F32
[2048]
blk.3.attn_q.weight
F16
F16
[2048, 2048]
blk.3.attn_v.bias
F32
F32
[256]
blk.3.attn_v.weight
F16
F16
[2048, 256]
blk.3.ffn_down.weight
F16
F16
[11008, 2048]
blk.3.ffn_gate.weight
F16
F16
[2048, 11008]
blk.3.ffn_norm.weight
F32
F32
[2048]
blk.3.ffn_up.weight
F16
F16
[2048, 11008]
blk.4
blk.4.attn_k.bias
F32
F32
[256]
blk.4.attn_k.weight
F16
F16
[2048, 256]
blk.4.attn_norm.weight
F32
F32
[2048]
blk.4.attn_output.weight
F16
F16
[2048, 2048]
blk.4.attn_q.bias
F32
F32
[2048]
blk.4.attn_q.weight
F16
F16
[2048, 2048]
blk.4.attn_v.bias
F32
F32
[256]
blk.4.attn_v.weight
F16
F16
[2048, 256]
blk.4.ffn_down.weight
F16
F16
[11008, 2048]
blk.4.ffn_gate.weight
F16
F16
[2048, 11008]
blk.4.ffn_norm.weight
F32
F32
[2048]
blk.4.ffn_up.weight
F16
F16
[2048, 11008]
blk.5
blk.5.attn_k.bias
F32
F32
[256]
blk.5.attn_k.weight
F16
F16
[2048, 256]
blk.5.attn_norm.weight
F32
F32
[2048]
blk.5.attn_output.weight
F16
F16
[2048, 2048]
blk.5.attn_q.bias
F32
F32
[2048]
blk.5.attn_q.weight
F16
F16
[2048, 2048]
blk.5.attn_v.bias
F32
F32
[256]
blk.5.attn_v.weight
F16
F16
[2048, 256]
blk.5.ffn_down.weight
F16
F16
[11008, 2048]
blk.5.ffn_gate.weight
F16
F16
[2048, 11008]
blk.5.ffn_norm.weight
F32
F32
[2048]
blk.5.ffn_up.weight
F16
F16
[2048, 11008]
blk.6
blk.6.attn_k.bias
F32
F32
[256]
blk.6.attn_k.weight
F16
F16
[2048, 256]
blk.6.attn_norm.weight
F32
F32
[2048]
blk.6.attn_output.weight
F16
F16
[2048, 2048]
blk.6.attn_q.bias
F32
F32
[2048]
blk.6.attn_q.weight
F16
F16
[2048, 2048]
blk.6.attn_v.bias
F32
F32
[256]
blk.6.attn_v.weight
F16
F16
[2048, 256]
blk.6.ffn_down.weight
F16
F16
[11008, 2048]
blk.6.ffn_gate.weight
F16
F16
[2048, 11008]
blk.6.ffn_norm.weight
F32
F32
[2048]
blk.6.ffn_up.weight
F16
F16
[2048, 11008]
blk.7
blk.7.attn_k.bias
F32
F32
[256]
blk.7.attn_k.weight
F16
F16
[2048, 256]
blk.7.attn_norm.weight
F32
F32
[2048]
blk.7.attn_output.weight
F16
F16
[2048, 2048]
blk.7.attn_q.bias
F32
F32
[2048]
blk.7.attn_q.weight
F16
F16
[2048, 2048]
blk.7.attn_v.bias
F32
F32
[256]
blk.7.attn_v.weight
F16
F16
[2048, 256]
blk.7.ffn_down.weight
F16
F16
[11008, 2048]
blk.7.ffn_gate.weight
F16
F16
[2048, 11008]
blk.7.ffn_norm.weight
F32
F32
[2048]
blk.7.ffn_up.weight
F16
F16
[2048, 11008]
blk.8
blk.8.attn_k.bias
F32
F32
[256]
blk.8.attn_k.weight
F16
F16
[2048, 256]
blk.8.attn_norm.weight
F32
F32
[2048]
blk.8.attn_output.weight
F16
F16
[2048, 2048]
blk.8.attn_q.bias
F32
F32
[2048]
blk.8.attn_q.weight
F16
F16
[2048, 2048]
blk.8.attn_v.bias
F32
F32
[256]
blk.8.attn_v.weight
F16
F16
[2048, 256]
blk.8.ffn_down.weight
F16
F16
[11008, 2048]
blk.8.ffn_gate.weight
F16
F16
[2048, 11008]
blk.8.ffn_norm.weight
F32
F32
[2048]
blk.8.ffn_up.weight
F16
F16
[2048, 11008]
blk.9
blk.9.attn_k.bias
F32
F32
[256]
blk.9.attn_k.weight
F16
F16
[2048, 256]
blk.9.attn_norm.weight
F32
F32
[2048]
blk.9.attn_output.weight
F16
F16
[2048, 2048]
blk.9.attn_q.bias
F32
F32
[2048]
blk.9.attn_q.weight
F16
F16
[2048, 2048]
blk.9.attn_v.bias
F32
F32
[256]
blk.9.attn_v.weight
F16
F16
[2048, 256]
blk.9.ffn_down.weight
F16
F16
[11008, 2048]
blk.9.ffn_gate.weight
F16
F16
[2048, 11008]
blk.9.ffn_norm.weight
F32
F32
[2048]
blk.9.ffn_up.weight
F16
F16
[2048, 11008]
blk.10
blk.10.attn_k.bias
F32
F32
[256]
blk.10.attn_k.weight
F16
F16
[2048, 256]
blk.10.attn_norm.weight
F32
F32
[2048]
blk.10.attn_output.weight
F16
F16
[2048, 2048]
blk.10.attn_q.bias
F32
F32
[2048]
blk.10.attn_q.weight
F16
F16
[2048, 2048]
blk.10.attn_v.bias
F32
F32
[256]
blk.10.attn_v.weight
F16
F16
[2048, 256]
blk.10.ffn_down.weight
F16
F16
[11008, 2048]
blk.10.ffn_gate.weight
F16
F16
[2048, 11008]
blk.10.ffn_norm.weight
F32
F32
[2048]
blk.10.ffn_up.weight
F16
F16
[2048, 11008]
blk.11
blk.11.attn_k.bias
F32
F32
[256]
blk.11.attn_k.weight
F16
F16
[2048, 256]
blk.11.attn_norm.weight
F32
F32
[2048]
blk.11.attn_output.weight
F16
F16
[2048, 2048]
blk.11.attn_q.bias
F32
F32
[2048]
blk.11.attn_q.weight
F16
F16
[2048, 2048]
blk.11.attn_v.bias
F32
F32
[256]
blk.11.attn_v.weight
F16
F16
[2048, 256]
blk.11.ffn_down.weight
F16
F16
[11008, 2048]
blk.11.ffn_gate.weight
F16
F16
[2048, 11008]
blk.11.ffn_norm.weight
F32
F32
[2048]
blk.11.ffn_up.weight
F16
F16
[2048, 11008]
blk.12
blk.12.attn_k.bias
F32
F32
[256]
blk.12.attn_k.weight
F16
F16
[2048, 256]
blk.12.attn_norm.weight
F32
F32
[2048]
blk.12.attn_output.weight
F16
F16
[2048, 2048]
blk.12.attn_q.bias
F32
F32
[2048]
blk.12.attn_q.weight
F16
F16
[2048, 2048]
blk.12.attn_v.bias
F32
F32
[256]
blk.12.attn_v.weight
F16
F16
[2048, 256]
blk.12.ffn_down.weight
F16
F16
[11008, 2048]
blk.12.ffn_gate.weight
F16
F16
[2048, 11008]
blk.12.ffn_norm.weight
F32
F32
[2048]
blk.12.ffn_up.weight
F16
F16
[2048, 11008]
blk.13
blk.13.attn_k.bias
F32
F32
[256]
blk.13.attn_k.weight
F16
F16
[2048, 256]
blk.13.attn_norm.weight
F32
F32
[2048]
blk.13.attn_output.weight
F16
F16
[2048, 2048]
blk.13.attn_q.bias
F32
F32
[2048]
blk.13.attn_q.weight
F16
F16
[2048, 2048]
blk.13.attn_v.bias
F32
F32
[256]
blk.13.attn_v.weight
F16
F16
[2048, 256]
blk.13.ffn_down.weight
F16
F16
[11008, 2048]
blk.13.ffn_gate.weight
F16
F16
[2048, 11008]
blk.13.ffn_norm.weight
F32
F32
[2048]
blk.13.ffn_up.weight
F16
F16
[2048, 11008]
blk.14
blk.14.attn_k.bias
F32
F32
[256]
blk.14.attn_k.weight
F16
F16
[2048, 256]
blk.14.attn_norm.weight
F32
F32
[2048]
blk.14.attn_output.weight
F16
F16
[2048, 2048]
blk.14.attn_q.bias
F32
F32
[2048]
blk.14.attn_q.weight
F16
F16
[2048, 2048]
blk.14.attn_v.bias
F32
F32
[256]
blk.14.attn_v.weight
F16
F16
[2048, 256]
blk.14.ffn_down.weight
F16
F16
[11008, 2048]
blk.14.ffn_gate.weight
F16
F16
[2048, 11008]
blk.14.ffn_norm.weight
F32
F32
[2048]
blk.14.ffn_up.weight
F16
F16
[2048, 11008]
blk.15
blk.15.attn_k.bias
F32
F32
[256]
blk.15.attn_k.weight
F16
F16
[2048, 256]
blk.15.attn_norm.weight
F32
F32
[2048]
blk.15.attn_output.weight
F16
F16
[2048, 2048]
blk.15.attn_q.bias
F32
F32
[2048]
blk.15.attn_q.weight
F16
F16
[2048, 2048]
blk.15.attn_v.bias
F32
F32
[256]
blk.15.attn_v.weight
F16
F16
[2048, 256]
blk.15.ffn_down.weight
F16
F16
[11008, 2048]
blk.15.ffn_gate.weight
F16
F16
[2048, 11008]
blk.15.ffn_norm.weight
F32
F32
[2048]
blk.15.ffn_up.weight
F16
F16
[2048, 11008]
blk.16
blk.16.attn_k.bias
F32
F32
[256]
blk.16.attn_k.weight
F16
F16
[2048, 256]
blk.16.attn_norm.weight
F32
F32
[2048]
blk.16.attn_output.weight
F16
F16
[2048, 2048]
blk.16.attn_q.bias
F32
F32
[2048]
blk.16.attn_q.weight
F16
F16
[2048, 2048]
blk.16.attn_v.bias
F32
F32
[256]
blk.16.attn_v.weight
F16
F16
[2048, 256]
blk.16.ffn_down.weight
F16
F16
[11008, 2048]
blk.16.ffn_gate.weight
F16
F16
[2048, 11008]
blk.16.ffn_norm.weight
F32
F32
[2048]
blk.16.ffn_up.weight
F16
F16
[2048, 11008]
blk.17
blk.17.attn_k.bias
F32
F32
[256]
blk.17.attn_k.weight
F16
F16
[2048, 256]
blk.17.attn_norm.weight
F32
F32
[2048]
blk.17.attn_output.weight
F16
F16
[2048, 2048]
blk.17.attn_q.bias
F32
F32
[2048]
blk.17.attn_q.weight
F16
F16
[2048, 2048]
blk.17.attn_v.bias
F32
F32
[256]
blk.17.attn_v.weight
F16
F16
[2048, 256]
blk.17.ffn_down.weight
F16
F16
[11008, 2048]
blk.17.ffn_gate.weight
F16
F16
[2048, 11008]
blk.17.ffn_norm.weight
F32
F32
[2048]
blk.17.ffn_up.weight
F16
F16
[2048, 11008]
blk.18
blk.18.attn_k.bias
F32
F32
[256]
blk.18.attn_k.weight
F16
F16
[2048, 256]
blk.18.attn_norm.weight
F32
F32
[2048]
blk.18.attn_output.weight
F16
F16
[2048, 2048]
blk.18.attn_q.bias
F32
F32
[2048]
blk.18.attn_q.weight
F16
F16
[2048, 2048]
blk.18.attn_v.bias
F32
F32
[256]
blk.18.attn_v.weight
F16
F16
[2048, 256]
blk.18.ffn_down.weight
F16
F16
[11008, 2048]
blk.18.ffn_gate.weight
F16
F16
[2048, 11008]
blk.18.ffn_norm.weight
F32
F32
[2048]
blk.18.ffn_up.weight
F16
F16
[2048, 11008]
blk.19
blk.19.attn_k.bias
F32
F32
[256]
blk.19.attn_k.weight
F16
F16
[2048, 256]
blk.19.attn_norm.weight
F32
F32
[2048]
blk.19.attn_output.weight
F16
F16
[2048, 2048]
blk.19.attn_q.bias
F32
F32
[2048]
blk.19.attn_q.weight
F16
F16
[2048, 2048]
blk.19.attn_v.bias
F32
F32
[256]
blk.19.attn_v.weight
F16
F16
[2048, 256]
blk.19.ffn_down.weight
F16
F16
[11008, 2048]
blk.19.ffn_gate.weight
F16
F16
[2048, 11008]
blk.19.ffn_norm.weight
F32
F32
[2048]
blk.19.ffn_up.weight
F16
F16
[2048, 11008]
blk.20
blk.20.attn_k.bias
F32
F32
[256]
blk.20.attn_k.weight
F16
F16
[2048, 256]
blk.20.attn_norm.weight
F32
F32
[2048]
blk.20.attn_output.weight
F16
F16
[2048, 2048]
blk.20.attn_q.bias
F32
F32
[2048]
blk.20.attn_q.weight
F16
F16
[2048, 2048]
blk.20.attn_v.bias
F32
F32
[256]
blk.20.attn_v.weight
F16
F16
[2048, 256]
blk.20.ffn_down.weight
F16
F16
[11008, 2048]
blk.20.ffn_gate.weight
F16
F16
[2048, 11008]
blk.20.ffn_norm.weight
F32
F32
[2048]
blk.20.ffn_up.weight
F16
F16
[2048, 11008]
blk.21
blk.21.attn_k.bias
F32
F32
[256]
blk.21.attn_k.weight
F16
F16
[2048, 256]
blk.21.attn_norm.weight
F32
F32
[2048]
blk.21.attn_output.weight
F16
F16
[2048, 2048]
blk.21.attn_q.bias
F32
F32
[2048]
blk.21.attn_q.weight
F16
F16
[2048, 2048]
blk.21.attn_v.bias
F32
F32
[256]
blk.21.attn_v.weight
F16
F16
[2048, 256]
blk.21.ffn_down.weight
F16
F16
[11008, 2048]
blk.21.ffn_gate.weight
F16
F16
[2048, 11008]
blk.21.ffn_norm.weight
F32
F32
[2048]
blk.21.ffn_up.weight
F16
F16
[2048, 11008]
blk.22
blk.22.attn_k.bias
F32
F32
[256]
blk.22.attn_k.weight
F16
F16
[2048, 256]
blk.22.attn_norm.weight
F32
F32
[2048]
blk.22.attn_output.weight
F16
F16
[2048, 2048]
blk.22.attn_q.bias
F32
F32
[2048]
blk.22.attn_q.weight
F16
F16
[2048, 2048]
blk.22.attn_v.bias
F32
F32
[256]
blk.22.attn_v.weight
F16
F16
[2048, 256]
blk.22.ffn_down.weight
F16
F16
[11008, 2048]
blk.22.ffn_gate.weight
F16
F16
[2048, 11008]
blk.22.ffn_norm.weight
F32
F32
[2048]
blk.22.ffn_up.weight
F16
F16
[2048, 11008]
blk.23
blk.23.attn_k.bias
F32
F32
[256]
blk.23.attn_k.weight
F16
F16
[2048, 256]
blk.23.attn_norm.weight
F32
F32
[2048]
blk.23.attn_output.weight
F16
F16
[2048, 2048]
blk.23.attn_q.bias
F32
F32
[2048]
blk.23.attn_q.weight
F16
F16
[2048, 2048]
blk.23.attn_v.bias
F32
F32
[256]
blk.23.attn_v.weight
F16
F16
[2048, 256]
blk.23.ffn_down.weight
F16
F16
[11008, 2048]
blk.23.ffn_gate.weight
F16
F16
[2048, 11008]
blk.23.ffn_norm.weight
F32
F32
[2048]
blk.23.ffn_up.weight
F16
F16
[2048, 11008]
blk.24
blk.24.attn_k.bias
F32
F32
[256]
blk.24.attn_k.weight
F16
F16
[2048, 256]
blk.24.attn_norm.weight
F32
F32
[2048]
blk.24.attn_output.weight
F16
F16
[2048, 2048]
blk.24.attn_q.bias
F32
F32
[2048]
blk.24.attn_q.weight
F16
F16
[2048, 2048]
blk.24.attn_v.bias
F32
F32
[256]
blk.24.attn_v.weight
F16
F16
[2048, 256]
blk.24.ffn_down.weight
F16
F16
[11008, 2048]
blk.24.ffn_gate.weight
F16
F16
[2048, 11008]
blk.24.ffn_norm.weight
F32
F32
[2048]
blk.24.ffn_up.weight
F16
F16
[2048, 11008]
blk.25
blk.25.attn_k.bias
F32
F32
[256]
blk.25.attn_k.weight
F16
F16
[2048, 256]
blk.25.attn_norm.weight
F32
F32
[2048]
blk.25.attn_output.weight
F16
F16
[2048, 2048]
blk.25.attn_q.bias
F32
F32
[2048]
blk.25.attn_q.weight
F16
F16
[2048, 2048]
blk.25.attn_v.bias
F32
F32
[256]
blk.25.attn_v.weight
F16
F16
[2048, 256]
blk.25.ffn_down.weight
F16
F16
[11008, 2048]
blk.25.ffn_gate.weight
F16
F16
[2048, 11008]
blk.25.ffn_norm.weight
F32
F32
[2048]
blk.25.ffn_up.weight
F16
F16
[2048, 11008]
blk.26
blk.26.attn_k.bias
F32
F32
[256]
blk.26.attn_k.weight
F16
F16
[2048, 256]
blk.26.attn_norm.weight
F32
F32
[2048]
blk.26.attn_output.weight
F16
F16
[2048, 2048]
blk.26.attn_q.bias
F32
F32
[2048]
blk.26.attn_q.weight
F16
F16
[2048, 2048]
blk.26.attn_v.bias
F32
F32
[256]
blk.26.attn_v.weight
F16
F16
[2048, 256]
blk.26.ffn_down.weight
F16
F16
[11008, 2048]
blk.26.ffn_gate.weight
F16
F16
[2048, 11008]
blk.26.ffn_norm.weight
F32
F32
[2048]
blk.26.ffn_up.weight
F16
F16
[2048, 11008]
blk.27
blk.27.attn_k.bias
F32
F32
[256]
blk.27.attn_k.weight
F16
F16
[2048, 256]
blk.27.attn_norm.weight
F32
F32
[2048]
blk.27.attn_output.weight
F16
F16
[2048, 2048]
blk.27.attn_q.bias
F32
F32
[2048]
blk.27.attn_q.weight
F16
F16
[2048, 2048]
blk.27.attn_v.bias
F32
F32
[256]
blk.27.attn_v.weight
F16
F16
[2048, 256]
blk.27.ffn_down.weight
F16
F16
[11008, 2048]
blk.27.ffn_gate.weight
F16
F16
[2048, 11008]
blk.27.ffn_norm.weight
F32
F32
[2048]
blk.27.ffn_up.weight
F16
F16
[2048, 11008]
blk.28
blk.28.attn_k.bias
F32
F32
[256]
blk.28.attn_k.weight
F16
F16
[2048, 256]
blk.28.attn_norm.weight
F32
F32
[2048]
blk.28.attn_output.weight
F16
F16
[2048, 2048]
blk.28.attn_q.bias
F32
F32
[2048]
blk.28.attn_q.weight
F16
F16
[2048, 2048]
blk.28.attn_v.bias
F32
F32
[256]
blk.28.attn_v.weight
F16
F16
[2048, 256]
blk.28.ffn_down.weight
F16
F16
[11008, 2048]
blk.28.ffn_gate.weight
F16
F16
[2048, 11008]
blk.28.ffn_norm.weight
F32
F32
[2048]
blk.28.ffn_up.weight
F16
F16
[2048, 11008]
blk.29
blk.29.attn_k.bias
F32
F32
[256]
blk.29.attn_k.weight
F16
F16
[2048, 256]
blk.29.attn_norm.weight
F32
F32
[2048]
blk.29.attn_output.weight
F16
F16
[2048, 2048]
blk.29.attn_q.bias
F32
F32
[2048]
blk.29.attn_q.weight
F16
F16
[2048, 2048]
blk.29.attn_v.bias
F32
F32
[256]
blk.29.attn_v.weight
F16
F16
[2048, 256]
blk.29.ffn_down.weight
F16
F16
[11008, 2048]
blk.29.ffn_gate.weight
F16
F16
[2048, 11008]
blk.29.ffn_norm.weight
F32
F32
[2048]
blk.29.ffn_up.weight
F16
F16
[2048, 11008]
blk.30
blk.30.attn_k.bias
F32
F32
[256]
blk.30.attn_k.weight
F16
F16
[2048, 256]
blk.30.attn_norm.weight
F32
F32
[2048]
blk.30.attn_output.weight
F16
F16
[2048, 2048]
blk.30.attn_q.bias
F32
F32
[2048]
blk.30.attn_q.weight
F16
F16
[2048, 2048]
blk.30.attn_v.bias
F32
F32
[256]
blk.30.attn_v.weight
F16
F16
[2048, 256]
blk.30.ffn_down.weight
F16
F16
[11008, 2048]
blk.30.ffn_gate.weight
F16
F16
[2048, 11008]
blk.30.ffn_norm.weight
F32
F32
[2048]
blk.30.ffn_up.weight
F16
F16
[2048, 11008]
blk.31
blk.31.attn_k.bias
F32
F32
[256]
blk.31.attn_k.weight
F16
F16
[2048, 256]
blk.31.attn_norm.weight
F32
F32
[2048]
blk.31.attn_output.weight
F16
F16
[2048, 2048]
blk.31.attn_q.bias
F32
F32
[2048]
blk.31.attn_q.weight
F16
F16
[2048, 2048]
blk.31.attn_v.bias
F32
F32
[256]
blk.31.attn_v.weight
F16
F16
[2048, 256]
blk.31.ffn_down.weight
F16
F16
[11008, 2048]
blk.31.ffn_gate.weight
F16
F16
[2048, 11008]
blk.31.ffn_norm.weight
F32
F32
[2048]
blk.31.ffn_up.weight
F16
F16
[2048, 11008]
blk.32
blk.32.attn_k.bias
F32
F32
[256]
blk.32.attn_k.weight
F16
F16
[2048, 256]
blk.32.attn_norm.weight
F32
F32
[2048]
blk.32.attn_output.weight
F16
F16
[2048, 2048]
blk.32.attn_q.bias
F32
F32
[2048]
blk.32.attn_q.weight
F16
F16
[2048, 2048]
blk.32.attn_v.bias
F32
F32
[256]
blk.32.attn_v.weight
F16
F16
[2048, 256]
blk.32.ffn_down.weight
F16
F16
[11008, 2048]
blk.32.ffn_gate.weight
F16
F16
[2048, 11008]
blk.32.ffn_norm.weight
F32
F32
[2048]
blk.32.ffn_up.weight
F16
F16
[2048, 11008]
blk.33
blk.33.attn_k.bias
F32
F32
[256]
blk.33.attn_k.weight
F16
F16
[2048, 256]
blk.33.attn_norm.weight
F32
F32
[2048]
blk.33.attn_output.weight
F16
F16
[2048, 2048]
blk.33.attn_q.bias
F32
F32
[2048]
blk.33.attn_q.weight
F16
F16
[2048, 2048]
blk.33.attn_v.bias
F32
F32
[256]
blk.33.attn_v.weight
F16
F16
[2048, 256]
blk.33.ffn_down.weight
F16
F16
[11008, 2048]
blk.33.ffn_gate.weight
F16
F16
[2048, 11008]
blk.33.ffn_norm.weight
F32
F32
[2048]
blk.33.ffn_up.weight
F16
F16
[2048, 11008]
blk.34
blk.34.attn_k.bias
F32
F32
[256]
blk.34.attn_k.weight
F16
F16
[2048, 256]
blk.34.attn_norm.weight
F32
F32
[2048]
blk.34.attn_output.weight
F16
F16
[2048, 2048]
blk.34.attn_q.bias
F32
F32
[2048]
blk.34.attn_q.weight
F16
F16
[2048, 2048]
blk.34.attn_v.bias
F32
F32
[256]
blk.34.attn_v.weight
F16
F16
[2048, 256]
blk.34.ffn_down.weight
F16
F16
[11008, 2048]
blk.34.ffn_gate.weight
F16
F16
[2048, 11008]
blk.34.ffn_norm.weight
F32
F32
[2048]
blk.34.ffn_up.weight
F16
F16
[2048, 11008]
blk.35
blk.35.attn_k.bias
F32
F32
[256]
blk.35.attn_k.weight
F16
F16
[2048, 256]
blk.35.attn_norm.weight
F32
F32
[2048]
blk.35.attn_output.weight
F16
F16
[2048, 2048]
blk.35.attn_q.bias
F32
F32
[2048]
blk.35.attn_q.weight
F16
F16
[2048, 2048]
blk.35.attn_v.bias
F32
F32
[256]
blk.35.attn_v.weight
F16
F16
[2048, 256]
blk.35.ffn_down.weight
F16
F16
[11008, 2048]
blk.35.ffn_gate.weight
F16
F16
[2048, 11008]
blk.35.ffn_norm.weight
F32
F32
[2048]
blk.35.ffn_up.weight
F16
F16
[2048, 11008]
output_norm.weight
F32
F32
[2048]