Models
GitHub
Discord
Docs
Cloud
Sign in
Download
Models
Download
GitHub
Discord
Docs
Cloud
Sign in
rajdeopankaj
/
bond-embed-v1-fp16
:latest
10
Downloads
Updated
3 months ago
Cancel
embedding
bond-embed-v1-fp16:latest
...
/
model
3126f10be641 · 84MB
Metadata
general.architecture
bert
bert
general.file_type
F16
F16
bert.attention.causal
false
false
bert.attention.head_count
6
6
bert.attention.layer_norm_epsilon
1e-12
1e-12
bert.block_count
16
16
bert.context_length
1024
1024
bert.embedding_length
384
384
bert.feed_forward_length
1536
1536
bert.pooling_type
Mean
Mean
tokenizer.ggml.add_bos_token
true
true
tokenizer.ggml.add_eos_token
true
true
tokenizer.ggml.add_sep_token
false
false
tokenizer.ggml.bos_token_id
101
101
tokenizer.ggml.eos_token_id
102
102
tokenizer.ggml.mask_token_id
103
103
tokenizer.ggml.model
bert
bert
tokenizer.ggml.padding_token_id
0
0
tokenizer.ggml.pre
bert-custom
bert-custom
tokenizer.ggml.seperator_token_id
102
102
tokenizer.ggml.token_type
[3, 1, 1, 1, 1, ...]
[3, 1, 1, 1, 1, ...]
tokenizer.ggml.token_type_count
2
2
tokenizer.ggml.tokens
[[PAD], [unused1], [unused2], [unused3], [unused4], ...]
[[PAD], [unused1], [unused2], [unused3], [unused4], ...]
tokenizer.ggml.unknown_token_id
100
100
Tensor
Name
Type
Shape
token_embd.weight
F16
F16
[384, 32768]
blk.0
blk.0.attn_k.bias
F32
F32
[384]
blk.0.attn_k.weight
F16
F16
[384, 384]
blk.0.attn_output.bias
F32
F32
[384]
blk.0.attn_output.weight
F16
F16
[384, 384]
blk.0.attn_output_norm.bias
F32
F32
[384]
blk.0.attn_output_norm.weight
F32
F32
[384]
blk.0.attn_q.bias
F32
F32
[384]
blk.0.attn_q.weight
F16
F16
[384, 384]
blk.0.attn_v.bias
F32
F32
[384]
blk.0.attn_v.weight
F16
F16
[384, 384]
blk.0.ffn_down.bias
F32
F32
[384]
blk.0.ffn_down.weight
F16
F16
[1536, 384]
blk.0.ffn_up.bias
F32
F32
[1536]
blk.0.ffn_up.weight
F16
F16
[384, 1536]
blk.0.layer_output_norm.bias
F32
F32
[384]
blk.0.layer_output_norm.weight
F32
F32
[384]
blk.1
blk.1.attn_k.bias
F32
F32
[384]
blk.1.attn_k.weight
F16
F16
[384, 384]
blk.1.attn_output.bias
F32
F32
[384]
blk.1.attn_output.weight
F16
F16
[384, 384]
blk.1.attn_output_norm.bias
F32
F32
[384]
blk.1.attn_output_norm.weight
F32
F32
[384]
blk.1.attn_q.bias
F32
F32
[384]
blk.1.attn_q.weight
F16
F16
[384, 384]
blk.1.attn_v.bias
F32
F32
[384]
blk.1.attn_v.weight
F16
F16
[384, 384]
blk.1.ffn_down.bias
F32
F32
[384]
blk.1.ffn_down.weight
F16
F16
[1536, 384]
blk.1.ffn_up.bias
F32
F32
[1536]
blk.1.ffn_up.weight
F16
F16
[384, 1536]
blk.1.layer_output_norm.bias
F32
F32
[384]
blk.1.layer_output_norm.weight
F32
F32
[384]
blk.2
blk.2.attn_k.bias
F32
F32
[384]
blk.2.attn_k.weight
F16
F16
[384, 384]
blk.2.attn_output.bias
F32
F32
[384]
blk.2.attn_output.weight
F16
F16
[384, 384]
blk.2.attn_output_norm.bias
F32
F32
[384]
blk.2.attn_output_norm.weight
F32
F32
[384]
blk.2.attn_q.bias
F32
F32
[384]
blk.2.attn_q.weight
F16
F16
[384, 384]
blk.2.attn_v.bias
F32
F32
[384]
blk.2.attn_v.weight
F16
F16
[384, 384]
blk.2.ffn_down.bias
F32
F32
[384]
blk.2.ffn_down.weight
F16
F16
[1536, 384]
blk.2.ffn_up.bias
F32
F32
[1536]
blk.2.ffn_up.weight
F16
F16
[384, 1536]
blk.2.layer_output_norm.bias
F32
F32
[384]
blk.2.layer_output_norm.weight
F32
F32
[384]
blk.3
blk.3.attn_k.bias
F32
F32
[384]
blk.3.attn_k.weight
F16
F16
[384, 384]
blk.3.attn_output.bias
F32
F32
[384]
blk.3.attn_output.weight
F16
F16
[384, 384]
blk.3.attn_output_norm.bias
F32
F32
[384]
blk.3.attn_output_norm.weight
F32
F32
[384]
blk.3.attn_q.bias
F32
F32
[384]
blk.3.attn_q.weight
F16
F16
[384, 384]
blk.3.attn_v.bias
F32
F32
[384]
blk.3.attn_v.weight
F16
F16
[384, 384]
blk.3.ffn_down.bias
F32
F32
[384]
blk.3.ffn_down.weight
F16
F16
[1536, 384]
blk.3.ffn_up.bias
F32
F32
[1536]
blk.3.ffn_up.weight
F16
F16
[384, 1536]
blk.3.layer_output_norm.bias
F32
F32
[384]
blk.3.layer_output_norm.weight
F32
F32
[384]
blk.4
blk.4.attn_k.bias
F32
F32
[384]
blk.4.attn_k.weight
F16
F16
[384, 384]
blk.4.attn_output.bias
F32
F32
[384]
blk.4.attn_output.weight
F16
F16
[384, 384]
blk.4.attn_output_norm.bias
F32
F32
[384]
blk.4.attn_output_norm.weight
F32
F32
[384]
blk.4.attn_q.bias
F32
F32
[384]
blk.4.attn_q.weight
F16
F16
[384, 384]
blk.4.attn_v.bias
F32
F32
[384]
blk.4.attn_v.weight
F16
F16
[384, 384]
blk.4.ffn_down.bias
F32
F32
[384]
blk.4.ffn_down.weight
F16
F16
[1536, 384]
blk.4.ffn_up.bias
F32
F32
[1536]
blk.4.ffn_up.weight
F16
F16
[384, 1536]
blk.4.layer_output_norm.bias
F32
F32
[384]
blk.4.layer_output_norm.weight
F32
F32
[384]
blk.5
blk.5.attn_k.bias
F32
F32
[384]
blk.5.attn_k.weight
F16
F16
[384, 384]
blk.5.attn_output.bias
F32
F32
[384]
blk.5.attn_output.weight
F16
F16
[384, 384]
blk.5.attn_output_norm.bias
F32
F32
[384]
blk.5.attn_output_norm.weight
F32
F32
[384]
blk.5.attn_q.bias
F32
F32
[384]
blk.5.attn_q.weight
F16
F16
[384, 384]
blk.5.attn_v.bias
F32
F32
[384]
blk.5.attn_v.weight
F16
F16
[384, 384]
blk.5.ffn_down.bias
F32
F32
[384]
blk.5.ffn_down.weight
F16
F16
[1536, 384]
blk.5.ffn_up.bias
F32
F32
[1536]
blk.5.ffn_up.weight
F16
F16
[384, 1536]
blk.5.layer_output_norm.bias
F32
F32
[384]
blk.5.layer_output_norm.weight
F32
F32
[384]
blk.6
blk.6.attn_k.bias
F32
F32
[384]
blk.6.attn_k.weight
F16
F16
[384, 384]
blk.6.attn_output.bias
F32
F32
[384]
blk.6.attn_output.weight
F16
F16
[384, 384]
blk.6.attn_output_norm.bias
F32
F32
[384]
blk.6.attn_output_norm.weight
F32
F32
[384]
blk.6.attn_q.bias
F32
F32
[384]
blk.6.attn_q.weight
F16
F16
[384, 384]
blk.6.attn_v.bias
F32
F32
[384]
blk.6.attn_v.weight
F16
F16
[384, 384]
blk.6.ffn_down.bias
F32
F32
[384]
blk.6.ffn_down.weight
F16
F16
[1536, 384]
blk.6.ffn_up.bias
F32
F32
[1536]
blk.6.ffn_up.weight
F16
F16
[384, 1536]
blk.6.layer_output_norm.bias
F32
F32
[384]
blk.6.layer_output_norm.weight
F32
F32
[384]
blk.7
blk.7.attn_k.bias
F32
F32
[384]
blk.7.attn_k.weight
F16
F16
[384, 384]
blk.7.attn_output.bias
F32
F32
[384]
blk.7.attn_output.weight
F16
F16
[384, 384]
blk.7.attn_output_norm.bias
F32
F32
[384]
blk.7.attn_output_norm.weight
F32
F32
[384]
blk.7.attn_q.bias
F32
F32
[384]
blk.7.attn_q.weight
F16
F16
[384, 384]
blk.7.attn_v.bias
F32
F32
[384]
blk.7.attn_v.weight
F16
F16
[384, 384]
blk.7.ffn_down.bias
F32
F32
[384]
blk.7.ffn_down.weight
F16
F16
[1536, 384]
blk.7.ffn_up.bias
F32
F32
[1536]
blk.7.ffn_up.weight
F16
F16
[384, 1536]
blk.7.layer_output_norm.bias
F32
F32
[384]
blk.7.layer_output_norm.weight
F32
F32
[384]
blk.8
blk.8.attn_k.bias
F32
F32
[384]
blk.8.attn_k.weight
F16
F16
[384, 384]
blk.8.attn_output.bias
F32
F32
[384]
blk.8.attn_output.weight
F16
F16
[384, 384]
blk.8.attn_output_norm.bias
F32
F32
[384]
blk.8.attn_output_norm.weight
F32
F32
[384]
blk.8.attn_q.bias
F32
F32
[384]
blk.8.attn_q.weight
F16
F16
[384, 384]
blk.8.attn_v.bias
F32
F32
[384]
blk.8.attn_v.weight
F16
F16
[384, 384]
blk.8.ffn_down.bias
F32
F32
[384]
blk.8.ffn_down.weight
F16
F16
[1536, 384]
blk.8.ffn_up.bias
F32
F32
[1536]
blk.8.ffn_up.weight
F16
F16
[384, 1536]
blk.8.layer_output_norm.bias
F32
F32
[384]
blk.8.layer_output_norm.weight
F32
F32
[384]
blk.9
blk.9.attn_k.bias
F32
F32
[384]
blk.9.attn_k.weight
F16
F16
[384, 384]
blk.9.attn_output.bias
F32
F32
[384]
blk.9.attn_output.weight
F16
F16
[384, 384]
blk.9.attn_output_norm.bias
F32
F32
[384]
blk.9.attn_output_norm.weight
F32
F32
[384]
blk.9.attn_q.bias
F32
F32
[384]
blk.9.attn_q.weight
F16
F16
[384, 384]
blk.9.attn_v.bias
F32
F32
[384]
blk.9.attn_v.weight
F16
F16
[384, 384]
blk.9.ffn_down.bias
F32
F32
[384]
blk.9.ffn_down.weight
F16
F16
[1536, 384]
blk.9.ffn_up.bias
F32
F32
[1536]
blk.9.ffn_up.weight
F16
F16
[384, 1536]
blk.9.layer_output_norm.bias
F32
F32
[384]
blk.9.layer_output_norm.weight
F32
F32
[384]
blk.10
blk.10.attn_k.bias
F32
F32
[384]
blk.10.attn_k.weight
F16
F16
[384, 384]
blk.10.attn_output.bias
F32
F32
[384]
blk.10.attn_output.weight
F16
F16
[384, 384]
blk.10.attn_output_norm.bias
F32
F32
[384]
blk.10.attn_output_norm.weight
F32
F32
[384]
blk.10.attn_q.bias
F32
F32
[384]
blk.10.attn_q.weight
F16
F16
[384, 384]
blk.10.attn_v.bias
F32
F32
[384]
blk.10.attn_v.weight
F16
F16
[384, 384]
blk.10.ffn_down.bias
F32
F32
[384]
blk.10.ffn_down.weight
F16
F16
[1536, 384]
blk.10.ffn_up.bias
F32
F32
[1536]
blk.10.ffn_up.weight
F16
F16
[384, 1536]
blk.10.layer_output_norm.bias
F32
F32
[384]
blk.10.layer_output_norm.weight
F32
F32
[384]
blk.11
blk.11.attn_k.bias
F32
F32
[384]
blk.11.attn_k.weight
F16
F16
[384, 384]
blk.11.attn_output.bias
F32
F32
[384]
blk.11.attn_output.weight
F16
F16
[384, 384]
blk.11.attn_output_norm.bias
F32
F32
[384]
blk.11.attn_output_norm.weight
F32
F32
[384]
blk.11.attn_q.bias
F32
F32
[384]
blk.11.attn_q.weight
F16
F16
[384, 384]
blk.11.attn_v.bias
F32
F32
[384]
blk.11.attn_v.weight
F16
F16
[384, 384]
blk.11.ffn_down.bias
F32
F32
[384]
blk.11.ffn_down.weight
F16
F16
[1536, 384]
blk.11.ffn_up.bias
F32
F32
[1536]
blk.11.ffn_up.weight
F16
F16
[384, 1536]
blk.11.layer_output_norm.bias
F32
F32
[384]
blk.11.layer_output_norm.weight
F32
F32
[384]
blk.12
blk.12.attn_k.bias
F32
F32
[384]
blk.12.attn_k.weight
F16
F16
[384, 384]
blk.12.attn_output.bias
F32
F32
[384]
blk.12.attn_output.weight
F16
F16
[384, 384]
blk.12.attn_output_norm.bias
F32
F32
[384]
blk.12.attn_output_norm.weight
F32
F32
[384]
blk.12.attn_q.bias
F32
F32
[384]
blk.12.attn_q.weight
F16
F16
[384, 384]
blk.12.attn_v.bias
F32
F32
[384]
blk.12.attn_v.weight
F16
F16
[384, 384]
blk.12.ffn_down.bias
F32
F32
[384]
blk.12.ffn_down.weight
F16
F16
[1536, 384]
blk.12.ffn_up.bias
F32
F32
[1536]
blk.12.ffn_up.weight
F16
F16
[384, 1536]
blk.12.layer_output_norm.bias
F32
F32
[384]
blk.12.layer_output_norm.weight
F32
F32
[384]
blk.13
blk.13.attn_k.bias
F32
F32
[384]
blk.13.attn_k.weight
F16
F16
[384, 384]
blk.13.attn_output.bias
F32
F32
[384]
blk.13.attn_output.weight
F16
F16
[384, 384]
blk.13.attn_output_norm.bias
F32
F32
[384]
blk.13.attn_output_norm.weight
F32
F32
[384]
blk.13.attn_q.bias
F32
F32
[384]
blk.13.attn_q.weight
F16
F16
[384, 384]
blk.13.attn_v.bias
F32
F32
[384]
blk.13.attn_v.weight
F16
F16
[384, 384]
blk.13.ffn_down.bias
F32
F32
[384]
blk.13.ffn_down.weight
F16
F16
[1536, 384]
blk.13.ffn_up.bias
F32
F32
[1536]
blk.13.ffn_up.weight
F16
F16
[384, 1536]
blk.13.layer_output_norm.bias
F32
F32
[384]
blk.13.layer_output_norm.weight
F32
F32
[384]
blk.14
blk.14.attn_k.bias
F32
F32
[384]
blk.14.attn_k.weight
F16
F16
[384, 384]
blk.14.attn_output.bias
F32
F32
[384]
blk.14.attn_output.weight
F16
F16
[384, 384]
blk.14.attn_output_norm.bias
F32
F32
[384]
blk.14.attn_output_norm.weight
F32
F32
[384]
blk.14.attn_q.bias
F32
F32
[384]
blk.14.attn_q.weight
F16
F16
[384, 384]
blk.14.attn_v.bias
F32
F32
[384]
blk.14.attn_v.weight
F16
F16
[384, 384]
blk.14.ffn_down.bias
F32
F32
[384]
blk.14.ffn_down.weight
F16
F16
[1536, 384]
blk.14.ffn_up.bias
F32
F32
[1536]
blk.14.ffn_up.weight
F16
F16
[384, 1536]
blk.14.layer_output_norm.bias
F32
F32
[384]
blk.14.layer_output_norm.weight
F32
F32
[384]
blk.15
blk.15.attn_k.bias
F32
F32
[384]
blk.15.attn_k.weight
F16
F16
[384, 384]
blk.15.attn_output.bias
F32
F32
[384]
blk.15.attn_output.weight
F16
F16
[384, 384]
blk.15.attn_output_norm.bias
F32
F32
[384]
blk.15.attn_output_norm.weight
F32
F32
[384]
blk.15.attn_q.bias
F32
F32
[384]
blk.15.attn_q.weight
F16
F16
[384, 384]
blk.15.attn_v.bias
F32
F32
[384]
blk.15.attn_v.weight
F16
F16
[384, 384]
blk.15.ffn_down.bias
F32
F32
[384]
blk.15.ffn_down.weight
F16
F16
[1536, 384]
blk.15.ffn_up.bias
F32
F32
[1536]
blk.15.ffn_up.weight
F16
F16
[384, 1536]
blk.15.layer_output_norm.bias
F32
F32
[384]
blk.15.layer_output_norm.weight
F32
F32
[384]
position_embd.weight
F32
F32
[384, 1024]
token_embd_norm.bias
F32
F32
[384]
token_embd_norm.weight
F32
F32
[384]
token_types.weight
F32
F32
[384, 2]