Models
GitHub
Discord
Docs
Cloud
Sign in
Download
Models
Download
GitHub
Discord
Docs
Cloud
Sign in
internlm
/
interns1
:q8_0
34.4K
Downloads
Updated
2 months ago
Intern-S1 is the most advanced open-source multimodal reasoning model to date. Intern-S1 combines strong general-task capabilities with state-of-the-art performance on a wide range of scientific tasks.
Intern-S1 is the most advanced open-source multimodal reasoning model to date. Intern-S1 combines strong general-task capabilities with state-of-the-art performance on a wide range of scientific tasks.
Cancel
vision
interns1:q8_0
...
/
projector
c0f0519f15b1 · 6.0GB
Metadata
general.architecture
clip
clip
general.file_type
Q8_0
Q8_0
clip.has_vision_encoder
true
true
clip.projector_type
internvl
internvl
clip.use_gelu
true
true
clip.vision.attention.head_count
25
25
clip.vision.attention.layer_norm_epsilon
1e-06
1e-06
clip.vision.block_count
45
45
clip.vision.embedding_length
3200
3200
clip.vision.feed_forward_length
12800
12800
clip.vision.image_mean
[0.485, 0.456, 0.406]
[0.485, 0.456, 0.406]
clip.vision.image_size
448
448
clip.vision.image_std
[0.229, 0.224, 0.225]
[0.229, 0.224, 0.225]
clip.vision.patch_size
14
14
clip.vision.projection_dim
4096
4096
clip.vision.projector.scale_factor
2
2
Tensor
Name
Type
Shape
mm.model.mlp.0.bias
F32
F32
[12800]
mm.model.mlp.0.weight
F32
F32
[12800]
mm.model.mlp.1.bias
F32
F32
[4096]
mm.model.mlp.1.weight
Q8_0
Q8_0
[12800, 4096]
mm.model.mlp.3.bias
F32
F32
[4096]
mm.model.mlp.3.weight
Q8_0
Q8_0
[4096, 4096]
v.blk.0
v.blk.0.attn_k.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.0.attn_k_norm.weight
F32
F32
[3200]
v.blk.0.attn_out.bias
F32
F32
[3200]
v.blk.0.attn_out.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.0.attn_q.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.0.attn_q_norm.weight
F32
F32
[3200]
v.blk.0.attn_v.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.0.ffn_down.bias
F32
F32
[3200]
v.blk.0.ffn_down.weight
Q8_0
Q8_0
[12800, 3200]
v.blk.0.ffn_up.bias
F32
F32
[12800]
v.blk.0.ffn_up.weight
Q8_0
Q8_0
[3200, 12800]
v.blk.0.ln1.weight
F32
F32
[3200]
v.blk.0.ln2.weight
F32
F32
[3200]
v.blk.0.ls1.weight
F32
F32
[3200]
v.blk.0.ls2.weight
F32
F32
[3200]
v.blk.1
v.blk.1.attn_k.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.1.attn_k_norm.weight
F32
F32
[3200]
v.blk.1.attn_out.bias
F32
F32
[3200]
v.blk.1.attn_out.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.1.attn_q.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.1.attn_q_norm.weight
F32
F32
[3200]
v.blk.1.attn_v.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.1.ffn_down.bias
F32
F32
[3200]
v.blk.1.ffn_down.weight
Q8_0
Q8_0
[12800, 3200]
v.blk.1.ffn_up.bias
F32
F32
[12800]
v.blk.1.ffn_up.weight
Q8_0
Q8_0
[3200, 12800]
v.blk.1.ln1.weight
F32
F32
[3200]
v.blk.1.ln2.weight
F32
F32
[3200]
v.blk.1.ls1.weight
F32
F32
[3200]
v.blk.1.ls2.weight
F32
F32
[3200]
v.blk.2
v.blk.2.attn_k.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.2.attn_k_norm.weight
F32
F32
[3200]
v.blk.2.attn_out.bias
F32
F32
[3200]
v.blk.2.attn_out.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.2.attn_q.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.2.attn_q_norm.weight
F32
F32
[3200]
v.blk.2.attn_v.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.2.ffn_down.bias
F32
F32
[3200]
v.blk.2.ffn_down.weight
Q8_0
Q8_0
[12800, 3200]
v.blk.2.ffn_up.bias
F32
F32
[12800]
v.blk.2.ffn_up.weight
Q8_0
Q8_0
[3200, 12800]
v.blk.2.ln1.weight
F32
F32
[3200]
v.blk.2.ln2.weight
F32
F32
[3200]
v.blk.2.ls1.weight
F32
F32
[3200]
v.blk.2.ls2.weight
F32
F32
[3200]
v.blk.3
v.blk.3.attn_k.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.3.attn_k_norm.weight
F32
F32
[3200]
v.blk.3.attn_out.bias
F32
F32
[3200]
v.blk.3.attn_out.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.3.attn_q.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.3.attn_q_norm.weight
F32
F32
[3200]
v.blk.3.attn_v.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.3.ffn_down.bias
F32
F32
[3200]
v.blk.3.ffn_down.weight
Q8_0
Q8_0
[12800, 3200]
v.blk.3.ffn_up.bias
F32
F32
[12800]
v.blk.3.ffn_up.weight
Q8_0
Q8_0
[3200, 12800]
v.blk.3.ln1.weight
F32
F32
[3200]
v.blk.3.ln2.weight
F32
F32
[3200]
v.blk.3.ls1.weight
F32
F32
[3200]
v.blk.3.ls2.weight
F32
F32
[3200]
v.blk.4
v.blk.4.attn_k.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.4.attn_k_norm.weight
F32
F32
[3200]
v.blk.4.attn_out.bias
F32
F32
[3200]
v.blk.4.attn_out.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.4.attn_q.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.4.attn_q_norm.weight
F32
F32
[3200]
v.blk.4.attn_v.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.4.ffn_down.bias
F32
F32
[3200]
v.blk.4.ffn_down.weight
Q8_0
Q8_0
[12800, 3200]
v.blk.4.ffn_up.bias
F32
F32
[12800]
v.blk.4.ffn_up.weight
Q8_0
Q8_0
[3200, 12800]
v.blk.4.ln1.weight
F32
F32
[3200]
v.blk.4.ln2.weight
F32
F32
[3200]
v.blk.4.ls1.weight
F32
F32
[3200]
v.blk.4.ls2.weight
F32
F32
[3200]
v.blk.5
v.blk.5.attn_k.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.5.attn_k_norm.weight
F32
F32
[3200]
v.blk.5.attn_out.bias
F32
F32
[3200]
v.blk.5.attn_out.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.5.attn_q.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.5.attn_q_norm.weight
F32
F32
[3200]
v.blk.5.attn_v.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.5.ffn_down.bias
F32
F32
[3200]
v.blk.5.ffn_down.weight
Q8_0
Q8_0
[12800, 3200]
v.blk.5.ffn_up.bias
F32
F32
[12800]
v.blk.5.ffn_up.weight
Q8_0
Q8_0
[3200, 12800]
v.blk.5.ln1.weight
F32
F32
[3200]
v.blk.5.ln2.weight
F32
F32
[3200]
v.blk.5.ls1.weight
F32
F32
[3200]
v.blk.5.ls2.weight
F32
F32
[3200]
v.blk.6
v.blk.6.attn_k.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.6.attn_k_norm.weight
F32
F32
[3200]
v.blk.6.attn_out.bias
F32
F32
[3200]
v.blk.6.attn_out.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.6.attn_q.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.6.attn_q_norm.weight
F32
F32
[3200]
v.blk.6.attn_v.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.6.ffn_down.bias
F32
F32
[3200]
v.blk.6.ffn_down.weight
Q8_0
Q8_0
[12800, 3200]
v.blk.6.ffn_up.bias
F32
F32
[12800]
v.blk.6.ffn_up.weight
Q8_0
Q8_0
[3200, 12800]
v.blk.6.ln1.weight
F32
F32
[3200]
v.blk.6.ln2.weight
F32
F32
[3200]
v.blk.6.ls1.weight
F32
F32
[3200]
v.blk.6.ls2.weight
F32
F32
[3200]
v.blk.7
v.blk.7.attn_k.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.7.attn_k_norm.weight
F32
F32
[3200]
v.blk.7.attn_out.bias
F32
F32
[3200]
v.blk.7.attn_out.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.7.attn_q.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.7.attn_q_norm.weight
F32
F32
[3200]
v.blk.7.attn_v.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.7.ffn_down.bias
F32
F32
[3200]
v.blk.7.ffn_down.weight
Q8_0
Q8_0
[12800, 3200]
v.blk.7.ffn_up.bias
F32
F32
[12800]
v.blk.7.ffn_up.weight
Q8_0
Q8_0
[3200, 12800]
v.blk.7.ln1.weight
F32
F32
[3200]
v.blk.7.ln2.weight
F32
F32
[3200]
v.blk.7.ls1.weight
F32
F32
[3200]
v.blk.7.ls2.weight
F32
F32
[3200]
v.blk.8
v.blk.8.attn_k.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.8.attn_k_norm.weight
F32
F32
[3200]
v.blk.8.attn_out.bias
F32
F32
[3200]
v.blk.8.attn_out.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.8.attn_q.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.8.attn_q_norm.weight
F32
F32
[3200]
v.blk.8.attn_v.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.8.ffn_down.bias
F32
F32
[3200]
v.blk.8.ffn_down.weight
Q8_0
Q8_0
[12800, 3200]
v.blk.8.ffn_up.bias
F32
F32
[12800]
v.blk.8.ffn_up.weight
Q8_0
Q8_0
[3200, 12800]
v.blk.8.ln1.weight
F32
F32
[3200]
v.blk.8.ln2.weight
F32
F32
[3200]
v.blk.8.ls1.weight
F32
F32
[3200]
v.blk.8.ls2.weight
F32
F32
[3200]
v.blk.9
v.blk.9.attn_k.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.9.attn_k_norm.weight
F32
F32
[3200]
v.blk.9.attn_out.bias
F32
F32
[3200]
v.blk.9.attn_out.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.9.attn_q.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.9.attn_q_norm.weight
F32
F32
[3200]
v.blk.9.attn_v.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.9.ffn_down.bias
F32
F32
[3200]
v.blk.9.ffn_down.weight
Q8_0
Q8_0
[12800, 3200]
v.blk.9.ffn_up.bias
F32
F32
[12800]
v.blk.9.ffn_up.weight
Q8_0
Q8_0
[3200, 12800]
v.blk.9.ln1.weight
F32
F32
[3200]
v.blk.9.ln2.weight
F32
F32
[3200]
v.blk.9.ls1.weight
F32
F32
[3200]
v.blk.9.ls2.weight
F32
F32
[3200]
v.blk.10
v.blk.10.attn_k.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.10.attn_k_norm.weight
F32
F32
[3200]
v.blk.10.attn_out.bias
F32
F32
[3200]
v.blk.10.attn_out.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.10.attn_q.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.10.attn_q_norm.weight
F32
F32
[3200]
v.blk.10.attn_v.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.10.ffn_down.bias
F32
F32
[3200]
v.blk.10.ffn_down.weight
Q8_0
Q8_0
[12800, 3200]
v.blk.10.ffn_up.bias
F32
F32
[12800]
v.blk.10.ffn_up.weight
Q8_0
Q8_0
[3200, 12800]
v.blk.10.ln1.weight
F32
F32
[3200]
v.blk.10.ln2.weight
F32
F32
[3200]
v.blk.10.ls1.weight
F32
F32
[3200]
v.blk.10.ls2.weight
F32
F32
[3200]
v.blk.11
v.blk.11.attn_k.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.11.attn_k_norm.weight
F32
F32
[3200]
v.blk.11.attn_out.bias
F32
F32
[3200]
v.blk.11.attn_out.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.11.attn_q.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.11.attn_q_norm.weight
F32
F32
[3200]
v.blk.11.attn_v.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.11.ffn_down.bias
F32
F32
[3200]
v.blk.11.ffn_down.weight
Q8_0
Q8_0
[12800, 3200]
v.blk.11.ffn_up.bias
F32
F32
[12800]
v.blk.11.ffn_up.weight
Q8_0
Q8_0
[3200, 12800]
v.blk.11.ln1.weight
F32
F32
[3200]
v.blk.11.ln2.weight
F32
F32
[3200]
v.blk.11.ls1.weight
F32
F32
[3200]
v.blk.11.ls2.weight
F32
F32
[3200]
v.blk.12
v.blk.12.attn_k.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.12.attn_k_norm.weight
F32
F32
[3200]
v.blk.12.attn_out.bias
F32
F32
[3200]
v.blk.12.attn_out.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.12.attn_q.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.12.attn_q_norm.weight
F32
F32
[3200]
v.blk.12.attn_v.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.12.ffn_down.bias
F32
F32
[3200]
v.blk.12.ffn_down.weight
Q8_0
Q8_0
[12800, 3200]
v.blk.12.ffn_up.bias
F32
F32
[12800]
v.blk.12.ffn_up.weight
Q8_0
Q8_0
[3200, 12800]
v.blk.12.ln1.weight
F32
F32
[3200]
v.blk.12.ln2.weight
F32
F32
[3200]
v.blk.12.ls1.weight
F32
F32
[3200]
v.blk.12.ls2.weight
F32
F32
[3200]
v.blk.13
v.blk.13.attn_k.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.13.attn_k_norm.weight
F32
F32
[3200]
v.blk.13.attn_out.bias
F32
F32
[3200]
v.blk.13.attn_out.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.13.attn_q.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.13.attn_q_norm.weight
F32
F32
[3200]
v.blk.13.attn_v.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.13.ffn_down.bias
F32
F32
[3200]
v.blk.13.ffn_down.weight
Q8_0
Q8_0
[12800, 3200]
v.blk.13.ffn_up.bias
F32
F32
[12800]
v.blk.13.ffn_up.weight
Q8_0
Q8_0
[3200, 12800]
v.blk.13.ln1.weight
F32
F32
[3200]
v.blk.13.ln2.weight
F32
F32
[3200]
v.blk.13.ls1.weight
F32
F32
[3200]
v.blk.13.ls2.weight
F32
F32
[3200]
v.blk.14
v.blk.14.attn_k.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.14.attn_k_norm.weight
F32
F32
[3200]
v.blk.14.attn_out.bias
F32
F32
[3200]
v.blk.14.attn_out.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.14.attn_q.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.14.attn_q_norm.weight
F32
F32
[3200]
v.blk.14.attn_v.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.14.ffn_down.bias
F32
F32
[3200]
v.blk.14.ffn_down.weight
Q8_0
Q8_0
[12800, 3200]
v.blk.14.ffn_up.bias
F32
F32
[12800]
v.blk.14.ffn_up.weight
Q8_0
Q8_0
[3200, 12800]
v.blk.14.ln1.weight
F32
F32
[3200]
v.blk.14.ln2.weight
F32
F32
[3200]
v.blk.14.ls1.weight
F32
F32
[3200]
v.blk.14.ls2.weight
F32
F32
[3200]
v.blk.15
v.blk.15.attn_k.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.15.attn_k_norm.weight
F32
F32
[3200]
v.blk.15.attn_out.bias
F32
F32
[3200]
v.blk.15.attn_out.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.15.attn_q.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.15.attn_q_norm.weight
F32
F32
[3200]
v.blk.15.attn_v.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.15.ffn_down.bias
F32
F32
[3200]
v.blk.15.ffn_down.weight
Q8_0
Q8_0
[12800, 3200]
v.blk.15.ffn_up.bias
F32
F32
[12800]
v.blk.15.ffn_up.weight
Q8_0
Q8_0
[3200, 12800]
v.blk.15.ln1.weight
F32
F32
[3200]
v.blk.15.ln2.weight
F32
F32
[3200]
v.blk.15.ls1.weight
F32
F32
[3200]
v.blk.15.ls2.weight
F32
F32
[3200]
v.blk.16
v.blk.16.attn_k.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.16.attn_k_norm.weight
F32
F32
[3200]
v.blk.16.attn_out.bias
F32
F32
[3200]
v.blk.16.attn_out.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.16.attn_q.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.16.attn_q_norm.weight
F32
F32
[3200]
v.blk.16.attn_v.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.16.ffn_down.bias
F32
F32
[3200]
v.blk.16.ffn_down.weight
Q8_0
Q8_0
[12800, 3200]
v.blk.16.ffn_up.bias
F32
F32
[12800]
v.blk.16.ffn_up.weight
Q8_0
Q8_0
[3200, 12800]
v.blk.16.ln1.weight
F32
F32
[3200]
v.blk.16.ln2.weight
F32
F32
[3200]
v.blk.16.ls1.weight
F32
F32
[3200]
v.blk.16.ls2.weight
F32
F32
[3200]
v.blk.17
v.blk.17.attn_k.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.17.attn_k_norm.weight
F32
F32
[3200]
v.blk.17.attn_out.bias
F32
F32
[3200]
v.blk.17.attn_out.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.17.attn_q.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.17.attn_q_norm.weight
F32
F32
[3200]
v.blk.17.attn_v.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.17.ffn_down.bias
F32
F32
[3200]
v.blk.17.ffn_down.weight
Q8_0
Q8_0
[12800, 3200]
v.blk.17.ffn_up.bias
F32
F32
[12800]
v.blk.17.ffn_up.weight
Q8_0
Q8_0
[3200, 12800]
v.blk.17.ln1.weight
F32
F32
[3200]
v.blk.17.ln2.weight
F32
F32
[3200]
v.blk.17.ls1.weight
F32
F32
[3200]
v.blk.17.ls2.weight
F32
F32
[3200]
v.blk.18
v.blk.18.attn_k.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.18.attn_k_norm.weight
F32
F32
[3200]
v.blk.18.attn_out.bias
F32
F32
[3200]
v.blk.18.attn_out.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.18.attn_q.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.18.attn_q_norm.weight
F32
F32
[3200]
v.blk.18.attn_v.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.18.ffn_down.bias
F32
F32
[3200]
v.blk.18.ffn_down.weight
Q8_0
Q8_0
[12800, 3200]
v.blk.18.ffn_up.bias
F32
F32
[12800]
v.blk.18.ffn_up.weight
Q8_0
Q8_0
[3200, 12800]
v.blk.18.ln1.weight
F32
F32
[3200]
v.blk.18.ln2.weight
F32
F32
[3200]
v.blk.18.ls1.weight
F32
F32
[3200]
v.blk.18.ls2.weight
F32
F32
[3200]
v.blk.19
v.blk.19.attn_k.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.19.attn_k_norm.weight
F32
F32
[3200]
v.blk.19.attn_out.bias
F32
F32
[3200]
v.blk.19.attn_out.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.19.attn_q.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.19.attn_q_norm.weight
F32
F32
[3200]
v.blk.19.attn_v.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.19.ffn_down.bias
F32
F32
[3200]
v.blk.19.ffn_down.weight
Q8_0
Q8_0
[12800, 3200]
v.blk.19.ffn_up.bias
F32
F32
[12800]
v.blk.19.ffn_up.weight
Q8_0
Q8_0
[3200, 12800]
v.blk.19.ln1.weight
F32
F32
[3200]
v.blk.19.ln2.weight
F32
F32
[3200]
v.blk.19.ls1.weight
F32
F32
[3200]
v.blk.19.ls2.weight
F32
F32
[3200]
v.blk.20
v.blk.20.attn_k.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.20.attn_k_norm.weight
F32
F32
[3200]
v.blk.20.attn_out.bias
F32
F32
[3200]
v.blk.20.attn_out.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.20.attn_q.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.20.attn_q_norm.weight
F32
F32
[3200]
v.blk.20.attn_v.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.20.ffn_down.bias
F32
F32
[3200]
v.blk.20.ffn_down.weight
Q8_0
Q8_0
[12800, 3200]
v.blk.20.ffn_up.bias
F32
F32
[12800]
v.blk.20.ffn_up.weight
Q8_0
Q8_0
[3200, 12800]
v.blk.20.ln1.weight
F32
F32
[3200]
v.blk.20.ln2.weight
F32
F32
[3200]
v.blk.20.ls1.weight
F32
F32
[3200]
v.blk.20.ls2.weight
F32
F32
[3200]
v.blk.21
v.blk.21.attn_k.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.21.attn_k_norm.weight
F32
F32
[3200]
v.blk.21.attn_out.bias
F32
F32
[3200]
v.blk.21.attn_out.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.21.attn_q.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.21.attn_q_norm.weight
F32
F32
[3200]
v.blk.21.attn_v.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.21.ffn_down.bias
F32
F32
[3200]
v.blk.21.ffn_down.weight
Q8_0
Q8_0
[12800, 3200]
v.blk.21.ffn_up.bias
F32
F32
[12800]
v.blk.21.ffn_up.weight
Q8_0
Q8_0
[3200, 12800]
v.blk.21.ln1.weight
F32
F32
[3200]
v.blk.21.ln2.weight
F32
F32
[3200]
v.blk.21.ls1.weight
F32
F32
[3200]
v.blk.21.ls2.weight
F32
F32
[3200]
v.blk.22
v.blk.22.attn_k.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.22.attn_k_norm.weight
F32
F32
[3200]
v.blk.22.attn_out.bias
F32
F32
[3200]
v.blk.22.attn_out.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.22.attn_q.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.22.attn_q_norm.weight
F32
F32
[3200]
v.blk.22.attn_v.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.22.ffn_down.bias
F32
F32
[3200]
v.blk.22.ffn_down.weight
Q8_0
Q8_0
[12800, 3200]
v.blk.22.ffn_up.bias
F32
F32
[12800]
v.blk.22.ffn_up.weight
Q8_0
Q8_0
[3200, 12800]
v.blk.22.ln1.weight
F32
F32
[3200]
v.blk.22.ln2.weight
F32
F32
[3200]
v.blk.22.ls1.weight
F32
F32
[3200]
v.blk.22.ls2.weight
F32
F32
[3200]
v.blk.23
v.blk.23.attn_k.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.23.attn_k_norm.weight
F32
F32
[3200]
v.blk.23.attn_out.bias
F32
F32
[3200]
v.blk.23.attn_out.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.23.attn_q.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.23.attn_q_norm.weight
F32
F32
[3200]
v.blk.23.attn_v.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.23.ffn_down.bias
F32
F32
[3200]
v.blk.23.ffn_down.weight
Q8_0
Q8_0
[12800, 3200]
v.blk.23.ffn_up.bias
F32
F32
[12800]
v.blk.23.ffn_up.weight
Q8_0
Q8_0
[3200, 12800]
v.blk.23.ln1.weight
F32
F32
[3200]
v.blk.23.ln2.weight
F32
F32
[3200]
v.blk.23.ls1.weight
F32
F32
[3200]
v.blk.23.ls2.weight
F32
F32
[3200]
v.blk.24
v.blk.24.attn_k.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.24.attn_k_norm.weight
F32
F32
[3200]
v.blk.24.attn_out.bias
F32
F32
[3200]
v.blk.24.attn_out.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.24.attn_q.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.24.attn_q_norm.weight
F32
F32
[3200]
v.blk.24.attn_v.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.24.ffn_down.bias
F32
F32
[3200]
v.blk.24.ffn_down.weight
Q8_0
Q8_0
[12800, 3200]
v.blk.24.ffn_up.bias
F32
F32
[12800]
v.blk.24.ffn_up.weight
Q8_0
Q8_0
[3200, 12800]
v.blk.24.ln1.weight
F32
F32
[3200]
v.blk.24.ln2.weight
F32
F32
[3200]
v.blk.24.ls1.weight
F32
F32
[3200]
v.blk.24.ls2.weight
F32
F32
[3200]
v.blk.25
v.blk.25.attn_k.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.25.attn_k_norm.weight
F32
F32
[3200]
v.blk.25.attn_out.bias
F32
F32
[3200]
v.blk.25.attn_out.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.25.attn_q.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.25.attn_q_norm.weight
F32
F32
[3200]
v.blk.25.attn_v.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.25.ffn_down.bias
F32
F32
[3200]
v.blk.25.ffn_down.weight
Q8_0
Q8_0
[12800, 3200]
v.blk.25.ffn_up.bias
F32
F32
[12800]
v.blk.25.ffn_up.weight
Q8_0
Q8_0
[3200, 12800]
v.blk.25.ln1.weight
F32
F32
[3200]
v.blk.25.ln2.weight
F32
F32
[3200]
v.blk.25.ls1.weight
F32
F32
[3200]
v.blk.25.ls2.weight
F32
F32
[3200]
v.blk.26
v.blk.26.attn_k.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.26.attn_k_norm.weight
F32
F32
[3200]
v.blk.26.attn_out.bias
F32
F32
[3200]
v.blk.26.attn_out.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.26.attn_q.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.26.attn_q_norm.weight
F32
F32
[3200]
v.blk.26.attn_v.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.26.ffn_down.bias
F32
F32
[3200]
v.blk.26.ffn_down.weight
Q8_0
Q8_0
[12800, 3200]
v.blk.26.ffn_up.bias
F32
F32
[12800]
v.blk.26.ffn_up.weight
Q8_0
Q8_0
[3200, 12800]
v.blk.26.ln1.weight
F32
F32
[3200]
v.blk.26.ln2.weight
F32
F32
[3200]
v.blk.26.ls1.weight
F32
F32
[3200]
v.blk.26.ls2.weight
F32
F32
[3200]
v.blk.27
v.blk.27.attn_k.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.27.attn_k_norm.weight
F32
F32
[3200]
v.blk.27.attn_out.bias
F32
F32
[3200]
v.blk.27.attn_out.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.27.attn_q.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.27.attn_q_norm.weight
F32
F32
[3200]
v.blk.27.attn_v.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.27.ffn_down.bias
F32
F32
[3200]
v.blk.27.ffn_down.weight
Q8_0
Q8_0
[12800, 3200]
v.blk.27.ffn_up.bias
F32
F32
[12800]
v.blk.27.ffn_up.weight
Q8_0
Q8_0
[3200, 12800]
v.blk.27.ln1.weight
F32
F32
[3200]
v.blk.27.ln2.weight
F32
F32
[3200]
v.blk.27.ls1.weight
F32
F32
[3200]
v.blk.27.ls2.weight
F32
F32
[3200]
v.blk.28
v.blk.28.attn_k.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.28.attn_k_norm.weight
F32
F32
[3200]
v.blk.28.attn_out.bias
F32
F32
[3200]
v.blk.28.attn_out.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.28.attn_q.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.28.attn_q_norm.weight
F32
F32
[3200]
v.blk.28.attn_v.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.28.ffn_down.bias
F32
F32
[3200]
v.blk.28.ffn_down.weight
Q8_0
Q8_0
[12800, 3200]
v.blk.28.ffn_up.bias
F32
F32
[12800]
v.blk.28.ffn_up.weight
Q8_0
Q8_0
[3200, 12800]
v.blk.28.ln1.weight
F32
F32
[3200]
v.blk.28.ln2.weight
F32
F32
[3200]
v.blk.28.ls1.weight
F32
F32
[3200]
v.blk.28.ls2.weight
F32
F32
[3200]
v.blk.29
v.blk.29.attn_k.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.29.attn_k_norm.weight
F32
F32
[3200]
v.blk.29.attn_out.bias
F32
F32
[3200]
v.blk.29.attn_out.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.29.attn_q.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.29.attn_q_norm.weight
F32
F32
[3200]
v.blk.29.attn_v.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.29.ffn_down.bias
F32
F32
[3200]
v.blk.29.ffn_down.weight
Q8_0
Q8_0
[12800, 3200]
v.blk.29.ffn_up.bias
F32
F32
[12800]
v.blk.29.ffn_up.weight
Q8_0
Q8_0
[3200, 12800]
v.blk.29.ln1.weight
F32
F32
[3200]
v.blk.29.ln2.weight
F32
F32
[3200]
v.blk.29.ls1.weight
F32
F32
[3200]
v.blk.29.ls2.weight
F32
F32
[3200]
v.blk.30
v.blk.30.attn_k.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.30.attn_k_norm.weight
F32
F32
[3200]
v.blk.30.attn_out.bias
F32
F32
[3200]
v.blk.30.attn_out.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.30.attn_q.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.30.attn_q_norm.weight
F32
F32
[3200]
v.blk.30.attn_v.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.30.ffn_down.bias
F32
F32
[3200]
v.blk.30.ffn_down.weight
Q8_0
Q8_0
[12800, 3200]
v.blk.30.ffn_up.bias
F32
F32
[12800]
v.blk.30.ffn_up.weight
Q8_0
Q8_0
[3200, 12800]
v.blk.30.ln1.weight
F32
F32
[3200]
v.blk.30.ln2.weight
F32
F32
[3200]
v.blk.30.ls1.weight
F32
F32
[3200]
v.blk.30.ls2.weight
F32
F32
[3200]
v.blk.31
v.blk.31.attn_k.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.31.attn_k_norm.weight
F32
F32
[3200]
v.blk.31.attn_out.bias
F32
F32
[3200]
v.blk.31.attn_out.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.31.attn_q.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.31.attn_q_norm.weight
F32
F32
[3200]
v.blk.31.attn_v.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.31.ffn_down.bias
F32
F32
[3200]
v.blk.31.ffn_down.weight
Q8_0
Q8_0
[12800, 3200]
v.blk.31.ffn_up.bias
F32
F32
[12800]
v.blk.31.ffn_up.weight
Q8_0
Q8_0
[3200, 12800]
v.blk.31.ln1.weight
F32
F32
[3200]
v.blk.31.ln2.weight
F32
F32
[3200]
v.blk.31.ls1.weight
F32
F32
[3200]
v.blk.31.ls2.weight
F32
F32
[3200]
v.blk.32
v.blk.32.attn_k.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.32.attn_k_norm.weight
F32
F32
[3200]
v.blk.32.attn_out.bias
F32
F32
[3200]
v.blk.32.attn_out.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.32.attn_q.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.32.attn_q_norm.weight
F32
F32
[3200]
v.blk.32.attn_v.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.32.ffn_down.bias
F32
F32
[3200]
v.blk.32.ffn_down.weight
Q8_0
Q8_0
[12800, 3200]
v.blk.32.ffn_up.bias
F32
F32
[12800]
v.blk.32.ffn_up.weight
Q8_0
Q8_0
[3200, 12800]
v.blk.32.ln1.weight
F32
F32
[3200]
v.blk.32.ln2.weight
F32
F32
[3200]
v.blk.32.ls1.weight
F32
F32
[3200]
v.blk.32.ls2.weight
F32
F32
[3200]
v.blk.33
v.blk.33.attn_k.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.33.attn_k_norm.weight
F32
F32
[3200]
v.blk.33.attn_out.bias
F32
F32
[3200]
v.blk.33.attn_out.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.33.attn_q.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.33.attn_q_norm.weight
F32
F32
[3200]
v.blk.33.attn_v.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.33.ffn_down.bias
F32
F32
[3200]
v.blk.33.ffn_down.weight
Q8_0
Q8_0
[12800, 3200]
v.blk.33.ffn_up.bias
F32
F32
[12800]
v.blk.33.ffn_up.weight
Q8_0
Q8_0
[3200, 12800]
v.blk.33.ln1.weight
F32
F32
[3200]
v.blk.33.ln2.weight
F32
F32
[3200]
v.blk.33.ls1.weight
F32
F32
[3200]
v.blk.33.ls2.weight
F32
F32
[3200]
v.blk.34
v.blk.34.attn_k.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.34.attn_k_norm.weight
F32
F32
[3200]
v.blk.34.attn_out.bias
F32
F32
[3200]
v.blk.34.attn_out.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.34.attn_q.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.34.attn_q_norm.weight
F32
F32
[3200]
v.blk.34.attn_v.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.34.ffn_down.bias
F32
F32
[3200]
v.blk.34.ffn_down.weight
Q8_0
Q8_0
[12800, 3200]
v.blk.34.ffn_up.bias
F32
F32
[12800]
v.blk.34.ffn_up.weight
Q8_0
Q8_0
[3200, 12800]
v.blk.34.ln1.weight
F32
F32
[3200]
v.blk.34.ln2.weight
F32
F32
[3200]
v.blk.34.ls1.weight
F32
F32
[3200]
v.blk.34.ls2.weight
F32
F32
[3200]
v.blk.35
v.blk.35.attn_k.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.35.attn_k_norm.weight
F32
F32
[3200]
v.blk.35.attn_out.bias
F32
F32
[3200]
v.blk.35.attn_out.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.35.attn_q.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.35.attn_q_norm.weight
F32
F32
[3200]
v.blk.35.attn_v.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.35.ffn_down.bias
F32
F32
[3200]
v.blk.35.ffn_down.weight
Q8_0
Q8_0
[12800, 3200]
v.blk.35.ffn_up.bias
F32
F32
[12800]
v.blk.35.ffn_up.weight
Q8_0
Q8_0
[3200, 12800]
v.blk.35.ln1.weight
F32
F32
[3200]
v.blk.35.ln2.weight
F32
F32
[3200]
v.blk.35.ls1.weight
F32
F32
[3200]
v.blk.35.ls2.weight
F32
F32
[3200]
v.blk.36
v.blk.36.attn_k.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.36.attn_k_norm.weight
F32
F32
[3200]
v.blk.36.attn_out.bias
F32
F32
[3200]
v.blk.36.attn_out.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.36.attn_q.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.36.attn_q_norm.weight
F32
F32
[3200]
v.blk.36.attn_v.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.36.ffn_down.bias
F32
F32
[3200]
v.blk.36.ffn_down.weight
Q8_0
Q8_0
[12800, 3200]
v.blk.36.ffn_up.bias
F32
F32
[12800]
v.blk.36.ffn_up.weight
Q8_0
Q8_0
[3200, 12800]
v.blk.36.ln1.weight
F32
F32
[3200]
v.blk.36.ln2.weight
F32
F32
[3200]
v.blk.36.ls1.weight
F32
F32
[3200]
v.blk.36.ls2.weight
F32
F32
[3200]
v.blk.37
v.blk.37.attn_k.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.37.attn_k_norm.weight
F32
F32
[3200]
v.blk.37.attn_out.bias
F32
F32
[3200]
v.blk.37.attn_out.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.37.attn_q.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.37.attn_q_norm.weight
F32
F32
[3200]
v.blk.37.attn_v.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.37.ffn_down.bias
F32
F32
[3200]
v.blk.37.ffn_down.weight
Q8_0
Q8_0
[12800, 3200]
v.blk.37.ffn_up.bias
F32
F32
[12800]
v.blk.37.ffn_up.weight
Q8_0
Q8_0
[3200, 12800]
v.blk.37.ln1.weight
F32
F32
[3200]
v.blk.37.ln2.weight
F32
F32
[3200]
v.blk.37.ls1.weight
F32
F32
[3200]
v.blk.37.ls2.weight
F32
F32
[3200]
v.blk.38
v.blk.38.attn_k.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.38.attn_k_norm.weight
F32
F32
[3200]
v.blk.38.attn_out.bias
F32
F32
[3200]
v.blk.38.attn_out.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.38.attn_q.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.38.attn_q_norm.weight
F32
F32
[3200]
v.blk.38.attn_v.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.38.ffn_down.bias
F32
F32
[3200]
v.blk.38.ffn_down.weight
Q8_0
Q8_0
[12800, 3200]
v.blk.38.ffn_up.bias
F32
F32
[12800]
v.blk.38.ffn_up.weight
Q8_0
Q8_0
[3200, 12800]
v.blk.38.ln1.weight
F32
F32
[3200]
v.blk.38.ln2.weight
F32
F32
[3200]
v.blk.38.ls1.weight
F32
F32
[3200]
v.blk.38.ls2.weight
F32
F32
[3200]
v.blk.39
v.blk.39.attn_k.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.39.attn_k_norm.weight
F32
F32
[3200]
v.blk.39.attn_out.bias
F32
F32
[3200]
v.blk.39.attn_out.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.39.attn_q.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.39.attn_q_norm.weight
F32
F32
[3200]
v.blk.39.attn_v.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.39.ffn_down.bias
F32
F32
[3200]
v.blk.39.ffn_down.weight
Q8_0
Q8_0
[12800, 3200]
v.blk.39.ffn_up.bias
F32
F32
[12800]
v.blk.39.ffn_up.weight
Q8_0
Q8_0
[3200, 12800]
v.blk.39.ln1.weight
F32
F32
[3200]
v.blk.39.ln2.weight
F32
F32
[3200]
v.blk.39.ls1.weight
F32
F32
[3200]
v.blk.39.ls2.weight
F32
F32
[3200]
v.blk.40
v.blk.40.attn_k.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.40.attn_k_norm.weight
F32
F32
[3200]
v.blk.40.attn_out.bias
F32
F32
[3200]
v.blk.40.attn_out.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.40.attn_q.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.40.attn_q_norm.weight
F32
F32
[3200]
v.blk.40.attn_v.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.40.ffn_down.bias
F32
F32
[3200]
v.blk.40.ffn_down.weight
Q8_0
Q8_0
[12800, 3200]
v.blk.40.ffn_up.bias
F32
F32
[12800]
v.blk.40.ffn_up.weight
Q8_0
Q8_0
[3200, 12800]
v.blk.40.ln1.weight
F32
F32
[3200]
v.blk.40.ln2.weight
F32
F32
[3200]
v.blk.40.ls1.weight
F32
F32
[3200]
v.blk.40.ls2.weight
F32
F32
[3200]
v.blk.41
v.blk.41.attn_k.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.41.attn_k_norm.weight
F32
F32
[3200]
v.blk.41.attn_out.bias
F32
F32
[3200]
v.blk.41.attn_out.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.41.attn_q.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.41.attn_q_norm.weight
F32
F32
[3200]
v.blk.41.attn_v.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.41.ffn_down.bias
F32
F32
[3200]
v.blk.41.ffn_down.weight
Q8_0
Q8_0
[12800, 3200]
v.blk.41.ffn_up.bias
F32
F32
[12800]
v.blk.41.ffn_up.weight
Q8_0
Q8_0
[3200, 12800]
v.blk.41.ln1.weight
F32
F32
[3200]
v.blk.41.ln2.weight
F32
F32
[3200]
v.blk.41.ls1.weight
F32
F32
[3200]
v.blk.41.ls2.weight
F32
F32
[3200]
v.blk.42
v.blk.42.attn_k.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.42.attn_k_norm.weight
F32
F32
[3200]
v.blk.42.attn_out.bias
F32
F32
[3200]
v.blk.42.attn_out.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.42.attn_q.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.42.attn_q_norm.weight
F32
F32
[3200]
v.blk.42.attn_v.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.42.ffn_down.bias
F32
F32
[3200]
v.blk.42.ffn_down.weight
Q8_0
Q8_0
[12800, 3200]
v.blk.42.ffn_up.bias
F32
F32
[12800]
v.blk.42.ffn_up.weight
Q8_0
Q8_0
[3200, 12800]
v.blk.42.ln1.weight
F32
F32
[3200]
v.blk.42.ln2.weight
F32
F32
[3200]
v.blk.42.ls1.weight
F32
F32
[3200]
v.blk.42.ls2.weight
F32
F32
[3200]
v.blk.43
v.blk.43.attn_k.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.43.attn_k_norm.weight
F32
F32
[3200]
v.blk.43.attn_out.bias
F32
F32
[3200]
v.blk.43.attn_out.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.43.attn_q.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.43.attn_q_norm.weight
F32
F32
[3200]
v.blk.43.attn_v.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.43.ffn_down.bias
F32
F32
[3200]
v.blk.43.ffn_down.weight
Q8_0
Q8_0
[12800, 3200]
v.blk.43.ffn_up.bias
F32
F32
[12800]
v.blk.43.ffn_up.weight
Q8_0
Q8_0
[3200, 12800]
v.blk.43.ln1.weight
F32
F32
[3200]
v.blk.43.ln2.weight
F32
F32
[3200]
v.blk.43.ls1.weight
F32
F32
[3200]
v.blk.43.ls2.weight
F32
F32
[3200]
v.blk.44
v.blk.44.attn_k.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.44.attn_k_norm.weight
F32
F32
[3200]
v.blk.44.attn_out.bias
F32
F32
[3200]
v.blk.44.attn_out.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.44.attn_q.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.44.attn_q_norm.weight
F32
F32
[3200]
v.blk.44.attn_v.weight
Q8_0
Q8_0
[3200, 3200]
v.blk.44.ffn_down.bias
F32
F32
[3200]
v.blk.44.ffn_down.weight
Q8_0
Q8_0
[12800, 3200]
v.blk.44.ffn_up.bias
F32
F32
[12800]
v.blk.44.ffn_up.weight
Q8_0
Q8_0
[3200, 12800]
v.blk.44.ln1.weight
F32
F32
[3200]
v.blk.44.ln2.weight
F32
F32
[3200]
v.blk.44.ls1.weight
F32
F32
[3200]
v.blk.44.ls2.weight
F32
F32
[3200]
v.class_embd
F32
F32
[3200, 1, 1]
v.patch_embd.bias
F32
F32
[3200]
v.patch_embd.weight
F16
F16
[14, 14, 3, 3200]
v.position_embd.weight
F32
F32
[3200, 1025, 1]