Models
Turbo
Sign in
Download
Models
Download
Sign in
stable-code
:latest
146.5K
Downloads
Updated
1 year ago
Stable Code 3B is a coding model with instruct and code completion variants on par with models such as Code Llama 7B that are 2.5x larger.
Stable Code 3B is a coding model with instruct and code completion variants on par with models such as Code Llama 7B that are 2.5x larger.
Cancel
3b
stable-code:latest
...
/
model
2c0d857fd309 · 1.6GB
Metadata
general.architecture
stablelm
stablelm
general.file_type
Q4_0
Q4_0
stablelm.attention.head_count
32
32
stablelm.attention.layer_norm_epsilon
1e-05
1e-05
stablelm.block_count
32
32
stablelm.context_length
16384
16384
stablelm.embedding_length
2560
2560
stablelm.feed_forward_length
6912
6912
stablelm.rope.dimension_count
20
20
stablelm.use_parallel_residual
true
true
tokenizer.ggml.bos_token_id
0
0
tokenizer.ggml.eos_token_id
0
0
tokenizer.ggml.merges
[Ġ Ġ, Ġ t, Ġ a, h e, i n, ...]
[Ġ Ġ, Ġ t, Ġ a, h e, i n, ...]
tokenizer.ggml.model
gpt2
gpt2
tokenizer.ggml.padding_token_id
0
0
tokenizer.ggml.token_type
[3, 3, 1, 1, 1, ...]
[3, 3, 1, 1, 1, ...]
tokenizer.ggml.tokens
[<|endoftext|>, <|padding|>, !, ", #, ...]
[<|endoftext|>, <|padding|>, !, ", #, ...]
tokenizer.ggml.unknown_token_id
0
0
Tensor
Name
Type
Shape
token_embd.weight
Q4_0
Q4_0
[2560, 50304]
blk.0
blk.0.attn_k.weight
Q4_0
Q4_0
[2560, 2560]
blk.0.attn_norm.bias
F32
F32
[2560]
blk.0.attn_norm.weight
F32
F32
[2560]
blk.0.attn_output.weight
Q4_0
Q4_0
[2560, 2560]
blk.0.attn_q.weight
Q4_0
Q4_0
[2560, 2560]
blk.0.attn_v.weight
Q4_0
Q4_0
[2560, 2560]
blk.0.ffn_down.weight
Q4_0
Q4_0
[6912, 2560]
blk.0.ffn_gate.weight
Q4_0
Q4_0
[2560, 6912]
blk.0.ffn_norm.bias
F32
F32
[2560]
blk.0.ffn_norm.weight
F32
F32
[2560]
blk.0.ffn_up.weight
Q4_0
Q4_0
[2560, 6912]
blk.1
blk.1.attn_k.weight
Q4_0
Q4_0
[2560, 2560]
blk.1.attn_norm.bias
F32
F32
[2560]
blk.1.attn_norm.weight
F32
F32
[2560]
blk.1.attn_output.weight
Q4_0
Q4_0
[2560, 2560]
blk.1.attn_q.weight
Q4_0
Q4_0
[2560, 2560]
blk.1.attn_v.weight
Q4_0
Q4_0
[2560, 2560]
blk.1.ffn_down.weight
Q4_0
Q4_0
[6912, 2560]
blk.1.ffn_gate.weight
Q4_0
Q4_0
[2560, 6912]
blk.1.ffn_norm.bias
F32
F32
[2560]
blk.1.ffn_norm.weight
F32
F32
[2560]
blk.1.ffn_up.weight
Q4_0
Q4_0
[2560, 6912]
blk.2
blk.2.attn_k.weight
Q4_0
Q4_0
[2560, 2560]
blk.2.attn_norm.bias
F32
F32
[2560]
blk.2.attn_norm.weight
F32
F32
[2560]
blk.2.attn_output.weight
Q4_0
Q4_0
[2560, 2560]
blk.2.attn_q.weight
Q4_0
Q4_0
[2560, 2560]
blk.2.attn_v.weight
Q4_0
Q4_0
[2560, 2560]
blk.2.ffn_down.weight
Q4_0
Q4_0
[6912, 2560]
blk.2.ffn_gate.weight
Q4_0
Q4_0
[2560, 6912]
blk.2.ffn_norm.bias
F32
F32
[2560]
blk.2.ffn_norm.weight
F32
F32
[2560]
blk.2.ffn_up.weight
Q4_0
Q4_0
[2560, 6912]
blk.3
blk.3.attn_k.weight
Q4_0
Q4_0
[2560, 2560]
blk.3.attn_norm.bias
F32
F32
[2560]
blk.3.attn_norm.weight
F32
F32
[2560]
blk.3.attn_output.weight
Q4_0
Q4_0
[2560, 2560]
blk.3.attn_q.weight
Q4_0
Q4_0
[2560, 2560]
blk.3.attn_v.weight
Q4_0
Q4_0
[2560, 2560]
blk.3.ffn_down.weight
Q4_0
Q4_0
[6912, 2560]
blk.3.ffn_gate.weight
Q4_0
Q4_0
[2560, 6912]
blk.3.ffn_norm.bias
F32
F32
[2560]
blk.3.ffn_norm.weight
F32
F32
[2560]
blk.3.ffn_up.weight
Q4_0
Q4_0
[2560, 6912]
blk.4
blk.4.attn_k.weight
Q4_0
Q4_0
[2560, 2560]
blk.4.attn_norm.bias
F32
F32
[2560]
blk.4.attn_norm.weight
F32
F32
[2560]
blk.4.attn_output.weight
Q4_0
Q4_0
[2560, 2560]
blk.4.attn_q.weight
Q4_0
Q4_0
[2560, 2560]
blk.4.attn_v.weight
Q4_0
Q4_0
[2560, 2560]
blk.4.ffn_down.weight
Q4_0
Q4_0
[6912, 2560]
blk.4.ffn_gate.weight
Q4_0
Q4_0
[2560, 6912]
blk.4.ffn_norm.bias
F32
F32
[2560]
blk.4.ffn_norm.weight
F32
F32
[2560]
blk.4.ffn_up.weight
Q4_0
Q4_0
[2560, 6912]
blk.5
blk.5.attn_k.weight
Q4_0
Q4_0
[2560, 2560]
blk.5.attn_norm.bias
F32
F32
[2560]
blk.5.attn_norm.weight
F32
F32
[2560]
blk.5.attn_output.weight
Q4_0
Q4_0
[2560, 2560]
blk.5.attn_q.weight
Q4_0
Q4_0
[2560, 2560]
blk.5.attn_v.weight
Q4_0
Q4_0
[2560, 2560]
blk.5.ffn_down.weight
Q4_0
Q4_0
[6912, 2560]
blk.5.ffn_gate.weight
Q4_0
Q4_0
[2560, 6912]
blk.5.ffn_norm.bias
F32
F32
[2560]
blk.5.ffn_norm.weight
F32
F32
[2560]
blk.5.ffn_up.weight
Q4_0
Q4_0
[2560, 6912]
blk.6
blk.6.attn_k.weight
Q4_0
Q4_0
[2560, 2560]
blk.6.attn_norm.bias
F32
F32
[2560]
blk.6.attn_norm.weight
F32
F32
[2560]
blk.6.attn_output.weight
Q4_0
Q4_0
[2560, 2560]
blk.6.attn_q.weight
Q4_0
Q4_0
[2560, 2560]
blk.6.attn_v.weight
Q4_0
Q4_0
[2560, 2560]
blk.6.ffn_down.weight
Q4_0
Q4_0
[6912, 2560]
blk.6.ffn_gate.weight
Q4_0
Q4_0
[2560, 6912]
blk.6.ffn_norm.bias
F32
F32
[2560]
blk.6.ffn_norm.weight
F32
F32
[2560]
blk.6.ffn_up.weight
Q4_0
Q4_0
[2560, 6912]
blk.7
blk.7.attn_k.weight
Q4_0
Q4_0
[2560, 2560]
blk.7.attn_norm.bias
F32
F32
[2560]
blk.7.attn_norm.weight
F32
F32
[2560]
blk.7.attn_output.weight
Q4_0
Q4_0
[2560, 2560]
blk.7.attn_q.weight
Q4_0
Q4_0
[2560, 2560]
blk.7.attn_v.weight
Q4_0
Q4_0
[2560, 2560]
blk.7.ffn_down.weight
Q4_0
Q4_0
[6912, 2560]
blk.7.ffn_gate.weight
Q4_0
Q4_0
[2560, 6912]
blk.7.ffn_norm.bias
F32
F32
[2560]
blk.7.ffn_norm.weight
F32
F32
[2560]
blk.7.ffn_up.weight
Q4_0
Q4_0
[2560, 6912]
blk.8
blk.8.attn_k.weight
Q4_0
Q4_0
[2560, 2560]
blk.8.attn_norm.bias
F32
F32
[2560]
blk.8.attn_norm.weight
F32
F32
[2560]
blk.8.attn_output.weight
Q4_0
Q4_0
[2560, 2560]
blk.8.attn_q.weight
Q4_0
Q4_0
[2560, 2560]
blk.8.attn_v.weight
Q4_0
Q4_0
[2560, 2560]
blk.8.ffn_down.weight
Q4_0
Q4_0
[6912, 2560]
blk.8.ffn_gate.weight
Q4_0
Q4_0
[2560, 6912]
blk.8.ffn_norm.bias
F32
F32
[2560]
blk.8.ffn_norm.weight
F32
F32
[2560]
blk.8.ffn_up.weight
Q4_0
Q4_0
[2560, 6912]
blk.9
blk.9.attn_k.weight
Q4_0
Q4_0
[2560, 2560]
blk.9.attn_norm.bias
F32
F32
[2560]
blk.9.attn_norm.weight
F32
F32
[2560]
blk.9.attn_output.weight
Q4_0
Q4_0
[2560, 2560]
blk.9.attn_q.weight
Q4_0
Q4_0
[2560, 2560]
blk.9.attn_v.weight
Q4_0
Q4_0
[2560, 2560]
blk.9.ffn_down.weight
Q4_0
Q4_0
[6912, 2560]
blk.9.ffn_gate.weight
Q4_0
Q4_0
[2560, 6912]
blk.9.ffn_norm.bias
F32
F32
[2560]
blk.9.ffn_norm.weight
F32
F32
[2560]
blk.9.ffn_up.weight
Q4_0
Q4_0
[2560, 6912]
blk.10
blk.10.attn_k.weight
Q4_0
Q4_0
[2560, 2560]
blk.10.attn_norm.bias
F32
F32
[2560]
blk.10.attn_norm.weight
F32
F32
[2560]
blk.10.attn_output.weight
Q4_0
Q4_0
[2560, 2560]
blk.10.attn_q.weight
Q4_0
Q4_0
[2560, 2560]
blk.10.attn_v.weight
Q4_0
Q4_0
[2560, 2560]
blk.10.ffn_down.weight
Q4_0
Q4_0
[6912, 2560]
blk.10.ffn_gate.weight
Q4_0
Q4_0
[2560, 6912]
blk.10.ffn_norm.bias
F32
F32
[2560]
blk.10.ffn_norm.weight
F32
F32
[2560]
blk.10.ffn_up.weight
Q4_0
Q4_0
[2560, 6912]
blk.11
blk.11.attn_k.weight
Q4_0
Q4_0
[2560, 2560]
blk.11.attn_norm.bias
F32
F32
[2560]
blk.11.attn_norm.weight
F32
F32
[2560]
blk.11.attn_output.weight
Q4_0
Q4_0
[2560, 2560]
blk.11.attn_q.weight
Q4_0
Q4_0
[2560, 2560]
blk.11.attn_v.weight
Q4_0
Q4_0
[2560, 2560]
blk.11.ffn_down.weight
Q4_0
Q4_0
[6912, 2560]
blk.11.ffn_gate.weight
Q4_0
Q4_0
[2560, 6912]
blk.11.ffn_norm.bias
F32
F32
[2560]
blk.11.ffn_norm.weight
F32
F32
[2560]
blk.11.ffn_up.weight
Q4_0
Q4_0
[2560, 6912]
blk.12
blk.12.attn_k.weight
Q4_0
Q4_0
[2560, 2560]
blk.12.attn_norm.bias
F32
F32
[2560]
blk.12.attn_norm.weight
F32
F32
[2560]
blk.12.attn_output.weight
Q4_0
Q4_0
[2560, 2560]
blk.12.attn_q.weight
Q4_0
Q4_0
[2560, 2560]
blk.12.attn_v.weight
Q4_0
Q4_0
[2560, 2560]
blk.12.ffn_down.weight
Q4_0
Q4_0
[6912, 2560]
blk.12.ffn_gate.weight
Q4_0
Q4_0
[2560, 6912]
blk.12.ffn_norm.bias
F32
F32
[2560]
blk.12.ffn_norm.weight
F32
F32
[2560]
blk.12.ffn_up.weight
Q4_0
Q4_0
[2560, 6912]
blk.13
blk.13.attn_k.weight
Q4_0
Q4_0
[2560, 2560]
blk.13.attn_norm.bias
F32
F32
[2560]
blk.13.attn_norm.weight
F32
F32
[2560]
blk.13.attn_output.weight
Q4_0
Q4_0
[2560, 2560]
blk.13.attn_q.weight
Q4_0
Q4_0
[2560, 2560]
blk.13.attn_v.weight
Q4_0
Q4_0
[2560, 2560]
blk.13.ffn_down.weight
Q4_0
Q4_0
[6912, 2560]
blk.13.ffn_gate.weight
Q4_0
Q4_0
[2560, 6912]
blk.13.ffn_norm.bias
F32
F32
[2560]
blk.13.ffn_norm.weight
F32
F32
[2560]
blk.13.ffn_up.weight
Q4_0
Q4_0
[2560, 6912]
blk.14
blk.14.attn_k.weight
Q4_0
Q4_0
[2560, 2560]
blk.14.attn_norm.bias
F32
F32
[2560]
blk.14.attn_norm.weight
F32
F32
[2560]
blk.14.attn_output.weight
Q4_0
Q4_0
[2560, 2560]
blk.14.attn_q.weight
Q4_0
Q4_0
[2560, 2560]
blk.14.attn_v.weight
Q4_0
Q4_0
[2560, 2560]
blk.14.ffn_down.weight
Q4_0
Q4_0
[6912, 2560]
blk.14.ffn_gate.weight
Q4_0
Q4_0
[2560, 6912]
blk.14.ffn_norm.bias
F32
F32
[2560]
blk.14.ffn_norm.weight
F32
F32
[2560]
blk.14.ffn_up.weight
Q4_0
Q4_0
[2560, 6912]
blk.15
blk.15.attn_k.weight
Q4_0
Q4_0
[2560, 2560]
blk.15.attn_norm.bias
F32
F32
[2560]
blk.15.attn_norm.weight
F32
F32
[2560]
blk.15.attn_output.weight
Q4_0
Q4_0
[2560, 2560]
blk.15.attn_q.weight
Q4_0
Q4_0
[2560, 2560]
blk.15.attn_v.weight
Q4_0
Q4_0
[2560, 2560]
blk.15.ffn_down.weight
Q4_0
Q4_0
[6912, 2560]
blk.15.ffn_gate.weight
Q4_0
Q4_0
[2560, 6912]
blk.15.ffn_norm.bias
F32
F32
[2560]
blk.15.ffn_norm.weight
F32
F32
[2560]
blk.15.ffn_up.weight
Q4_0
Q4_0
[2560, 6912]
blk.16
blk.16.attn_k.weight
Q4_0
Q4_0
[2560, 2560]
blk.16.attn_norm.bias
F32
F32
[2560]
blk.16.attn_norm.weight
F32
F32
[2560]
blk.16.attn_output.weight
Q4_0
Q4_0
[2560, 2560]
blk.16.attn_q.weight
Q4_0
Q4_0
[2560, 2560]
blk.16.attn_v.weight
Q4_0
Q4_0
[2560, 2560]
blk.16.ffn_down.weight
Q4_0
Q4_0
[6912, 2560]
blk.16.ffn_gate.weight
Q4_0
Q4_0
[2560, 6912]
blk.16.ffn_norm.bias
F32
F32
[2560]
blk.16.ffn_norm.weight
F32
F32
[2560]
blk.16.ffn_up.weight
Q4_0
Q4_0
[2560, 6912]
blk.17
blk.17.attn_k.weight
Q4_0
Q4_0
[2560, 2560]
blk.17.attn_norm.bias
F32
F32
[2560]
blk.17.attn_norm.weight
F32
F32
[2560]
blk.17.attn_output.weight
Q4_0
Q4_0
[2560, 2560]
blk.17.attn_q.weight
Q4_0
Q4_0
[2560, 2560]
blk.17.attn_v.weight
Q4_0
Q4_0
[2560, 2560]
blk.17.ffn_down.weight
Q4_0
Q4_0
[6912, 2560]
blk.17.ffn_gate.weight
Q4_0
Q4_0
[2560, 6912]
blk.17.ffn_norm.bias
F32
F32
[2560]
blk.17.ffn_norm.weight
F32
F32
[2560]
blk.17.ffn_up.weight
Q4_0
Q4_0
[2560, 6912]
blk.18
blk.18.attn_k.weight
Q4_0
Q4_0
[2560, 2560]
blk.18.attn_norm.bias
F32
F32
[2560]
blk.18.attn_norm.weight
F32
F32
[2560]
blk.18.attn_output.weight
Q4_0
Q4_0
[2560, 2560]
blk.18.attn_q.weight
Q4_0
Q4_0
[2560, 2560]
blk.18.attn_v.weight
Q4_0
Q4_0
[2560, 2560]
blk.18.ffn_down.weight
Q4_0
Q4_0
[6912, 2560]
blk.18.ffn_gate.weight
Q4_0
Q4_0
[2560, 6912]
blk.18.ffn_norm.bias
F32
F32
[2560]
blk.18.ffn_norm.weight
F32
F32
[2560]
blk.18.ffn_up.weight
Q4_0
Q4_0
[2560, 6912]
blk.19
blk.19.attn_k.weight
Q4_0
Q4_0
[2560, 2560]
blk.19.attn_norm.bias
F32
F32
[2560]
blk.19.attn_norm.weight
F32
F32
[2560]
blk.19.attn_output.weight
Q4_0
Q4_0
[2560, 2560]
blk.19.attn_q.weight
Q4_0
Q4_0
[2560, 2560]
blk.19.attn_v.weight
Q4_0
Q4_0
[2560, 2560]
blk.19.ffn_down.weight
Q4_0
Q4_0
[6912, 2560]
blk.19.ffn_gate.weight
Q4_0
Q4_0
[2560, 6912]
blk.19.ffn_norm.bias
F32
F32
[2560]
blk.19.ffn_norm.weight
F32
F32
[2560]
blk.19.ffn_up.weight
Q4_0
Q4_0
[2560, 6912]
blk.20
blk.20.attn_k.weight
Q4_0
Q4_0
[2560, 2560]
blk.20.attn_norm.bias
F32
F32
[2560]
blk.20.attn_norm.weight
F32
F32
[2560]
blk.20.attn_output.weight
Q4_0
Q4_0
[2560, 2560]
blk.20.attn_q.weight
Q4_0
Q4_0
[2560, 2560]
blk.20.attn_v.weight
Q4_0
Q4_0
[2560, 2560]
blk.20.ffn_down.weight
Q4_0
Q4_0
[6912, 2560]
blk.20.ffn_gate.weight
Q4_0
Q4_0
[2560, 6912]
blk.20.ffn_norm.bias
F32
F32
[2560]
blk.20.ffn_norm.weight
F32
F32
[2560]
blk.20.ffn_up.weight
Q4_0
Q4_0
[2560, 6912]
blk.21
blk.21.attn_k.weight
Q4_0
Q4_0
[2560, 2560]
blk.21.attn_norm.bias
F32
F32
[2560]
blk.21.attn_norm.weight
F32
F32
[2560]
blk.21.attn_output.weight
Q4_0
Q4_0
[2560, 2560]
blk.21.attn_q.weight
Q4_0
Q4_0
[2560, 2560]
blk.21.attn_v.weight
Q4_0
Q4_0
[2560, 2560]
blk.21.ffn_down.weight
Q4_0
Q4_0
[6912, 2560]
blk.21.ffn_gate.weight
Q4_0
Q4_0
[2560, 6912]
blk.21.ffn_norm.bias
F32
F32
[2560]
blk.21.ffn_norm.weight
F32
F32
[2560]
blk.21.ffn_up.weight
Q4_0
Q4_0
[2560, 6912]
blk.22
blk.22.attn_k.weight
Q4_0
Q4_0
[2560, 2560]
blk.22.attn_norm.bias
F32
F32
[2560]
blk.22.attn_norm.weight
F32
F32
[2560]
blk.22.attn_output.weight
Q4_0
Q4_0
[2560, 2560]
blk.22.attn_q.weight
Q4_0
Q4_0
[2560, 2560]
blk.22.attn_v.weight
Q4_0
Q4_0
[2560, 2560]
blk.22.ffn_down.weight
Q4_0
Q4_0
[6912, 2560]
blk.22.ffn_gate.weight
Q4_0
Q4_0
[2560, 6912]
blk.22.ffn_norm.bias
F32
F32
[2560]
blk.22.ffn_norm.weight
F32
F32
[2560]
blk.22.ffn_up.weight
Q4_0
Q4_0
[2560, 6912]
blk.23
blk.23.attn_k.weight
Q4_0
Q4_0
[2560, 2560]
blk.23.attn_norm.bias
F32
F32
[2560]
blk.23.attn_norm.weight
F32
F32
[2560]
blk.23.attn_output.weight
Q4_0
Q4_0
[2560, 2560]
blk.23.attn_q.weight
Q4_0
Q4_0
[2560, 2560]
blk.23.attn_v.weight
Q4_0
Q4_0
[2560, 2560]
blk.23.ffn_down.weight
Q4_0
Q4_0
[6912, 2560]
blk.23.ffn_gate.weight
Q4_0
Q4_0
[2560, 6912]
blk.23.ffn_norm.bias
F32
F32
[2560]
blk.23.ffn_norm.weight
F32
F32
[2560]
blk.23.ffn_up.weight
Q4_0
Q4_0
[2560, 6912]
blk.24
blk.24.attn_k.weight
Q4_0
Q4_0
[2560, 2560]
blk.24.attn_norm.bias
F32
F32
[2560]
blk.24.attn_norm.weight
F32
F32
[2560]
blk.24.attn_output.weight
Q4_0
Q4_0
[2560, 2560]
blk.24.attn_q.weight
Q4_0
Q4_0
[2560, 2560]
blk.24.attn_v.weight
Q4_0
Q4_0
[2560, 2560]
blk.24.ffn_down.weight
Q4_0
Q4_0
[6912, 2560]
blk.24.ffn_gate.weight
Q4_0
Q4_0
[2560, 6912]
blk.24.ffn_norm.bias
F32
F32
[2560]
blk.24.ffn_norm.weight
F32
F32
[2560]
blk.24.ffn_up.weight
Q4_0
Q4_0
[2560, 6912]
blk.25
blk.25.attn_k.weight
Q4_0
Q4_0
[2560, 2560]
blk.25.attn_norm.bias
F32
F32
[2560]
blk.25.attn_norm.weight
F32
F32
[2560]
blk.25.attn_output.weight
Q4_0
Q4_0
[2560, 2560]
blk.25.attn_q.weight
Q4_0
Q4_0
[2560, 2560]
blk.25.attn_v.weight
Q4_0
Q4_0
[2560, 2560]
blk.25.ffn_down.weight
Q4_0
Q4_0
[6912, 2560]
blk.25.ffn_gate.weight
Q4_0
Q4_0
[2560, 6912]
blk.25.ffn_norm.bias
F32
F32
[2560]
blk.25.ffn_norm.weight
F32
F32
[2560]
blk.25.ffn_up.weight
Q4_0
Q4_0
[2560, 6912]
blk.26
blk.26.attn_k.weight
Q4_0
Q4_0
[2560, 2560]
blk.26.attn_norm.bias
F32
F32
[2560]
blk.26.attn_norm.weight
F32
F32
[2560]
blk.26.attn_output.weight
Q4_0
Q4_0
[2560, 2560]
blk.26.attn_q.weight
Q4_0
Q4_0
[2560, 2560]
blk.26.attn_v.weight
Q4_0
Q4_0
[2560, 2560]
blk.26.ffn_down.weight
Q4_0
Q4_0
[6912, 2560]
blk.26.ffn_gate.weight
Q4_0
Q4_0
[2560, 6912]
blk.26.ffn_norm.bias
F32
F32
[2560]
blk.26.ffn_norm.weight
F32
F32
[2560]
blk.26.ffn_up.weight
Q4_0
Q4_0
[2560, 6912]
blk.27
blk.27.attn_k.weight
Q4_0
Q4_0
[2560, 2560]
blk.27.attn_norm.bias
F32
F32
[2560]
blk.27.attn_norm.weight
F32
F32
[2560]
blk.27.attn_output.weight
Q4_0
Q4_0
[2560, 2560]
blk.27.attn_q.weight
Q4_0
Q4_0
[2560, 2560]
blk.27.attn_v.weight
Q4_0
Q4_0
[2560, 2560]
blk.27.ffn_down.weight
Q4_0
Q4_0
[6912, 2560]
blk.27.ffn_gate.weight
Q4_0
Q4_0
[2560, 6912]
blk.27.ffn_norm.bias
F32
F32
[2560]
blk.27.ffn_norm.weight
F32
F32
[2560]
blk.27.ffn_up.weight
Q4_0
Q4_0
[2560, 6912]
blk.28
blk.28.attn_k.weight
Q4_0
Q4_0
[2560, 2560]
blk.28.attn_norm.bias
F32
F32
[2560]
blk.28.attn_norm.weight
F32
F32
[2560]
blk.28.attn_output.weight
Q4_0
Q4_0
[2560, 2560]
blk.28.attn_q.weight
Q4_0
Q4_0
[2560, 2560]
blk.28.attn_v.weight
Q4_0
Q4_0
[2560, 2560]
blk.28.ffn_down.weight
Q4_0
Q4_0
[6912, 2560]
blk.28.ffn_gate.weight
Q4_0
Q4_0
[2560, 6912]
blk.28.ffn_norm.bias
F32
F32
[2560]
blk.28.ffn_norm.weight
F32
F32
[2560]
blk.28.ffn_up.weight
Q4_0
Q4_0
[2560, 6912]
blk.29
blk.29.attn_k.weight
Q4_0
Q4_0
[2560, 2560]
blk.29.attn_norm.bias
F32
F32
[2560]
blk.29.attn_norm.weight
F32
F32
[2560]
blk.29.attn_output.weight
Q4_0
Q4_0
[2560, 2560]
blk.29.attn_q.weight
Q4_0
Q4_0
[2560, 2560]
blk.29.attn_v.weight
Q4_0
Q4_0
[2560, 2560]
blk.29.ffn_down.weight
Q4_0
Q4_0
[6912, 2560]
blk.29.ffn_gate.weight
Q4_0
Q4_0
[2560, 6912]
blk.29.ffn_norm.bias
F32
F32
[2560]
blk.29.ffn_norm.weight
F32
F32
[2560]
blk.29.ffn_up.weight
Q4_0
Q4_0
[2560, 6912]
blk.30
blk.30.attn_k.weight
Q4_0
Q4_0
[2560, 2560]
blk.30.attn_norm.bias
F32
F32
[2560]
blk.30.attn_norm.weight
F32
F32
[2560]
blk.30.attn_output.weight
Q4_0
Q4_0
[2560, 2560]
blk.30.attn_q.weight
Q4_0
Q4_0
[2560, 2560]
blk.30.attn_v.weight
Q4_0
Q4_0
[2560, 2560]
blk.30.ffn_down.weight
Q4_0
Q4_0
[6912, 2560]
blk.30.ffn_gate.weight
Q4_0
Q4_0
[2560, 6912]
blk.30.ffn_norm.bias
F32
F32
[2560]
blk.30.ffn_norm.weight
F32
F32
[2560]
blk.30.ffn_up.weight
Q4_0
Q4_0
[2560, 6912]
blk.31
blk.31.attn_k.weight
Q4_0
Q4_0
[2560, 2560]
blk.31.attn_norm.bias
F32
F32
[2560]
blk.31.attn_norm.weight
F32
F32
[2560]
blk.31.attn_output.weight
Q4_0
Q4_0
[2560, 2560]
blk.31.attn_q.weight
Q4_0
Q4_0
[2560, 2560]
blk.31.attn_v.weight
Q4_0
Q4_0
[2560, 2560]
blk.31.ffn_down.weight
Q4_0
Q4_0
[6912, 2560]
blk.31.ffn_gate.weight
Q4_0
Q4_0
[2560, 6912]
blk.31.ffn_norm.bias
F32
F32
[2560]
blk.31.ffn_norm.weight
F32
F32
[2560]
blk.31.ffn_up.weight
Q4_0
Q4_0
[2560, 6912]
output.weight
Q6_K
Q6_K
[2560, 50304]
output_norm.bias
F32
F32
[2560]
output_norm.weight
F32
F32
[2560]