GitHub
185.5K
Download
Models
Docs
Pricing
Sign in
Download
Models
Docs
Pricing
Sign in
postpostmodern
/
talkie-1930
:latest
83
Downloads
Updated
1 month ago
A 13B language model trained only on pre-1931 English, packaged to run on stock Ollama. Ask it about the future and it will guess 1960.
A 13B language model trained only on pre-1931 English, packaged to run on stock Ollama. Ask it about the future and it will guess 1960.
Cancel
talkie-1930:latest
...
/
model
7bcc763f1bdc · 8.6GB
Metadata
general.architecture
talkie
talkie
general.file_type
Q4_K_M
Q4_K_M
talkie.attention.head_count
40
40
talkie.attention.key_length
128
128
talkie.attention.layer_norm_rms_epsilon
1.1920929e-07
1.1920929e-07
talkie.attention.value_length
128
128
talkie.block_count
40
40
talkie.context_length
8192
8192
talkie.embedding_length
5120
5120
talkie.feed_forward_length
13696
13696
talkie.logit_scale
3.890625
3.890625
talkie.rope.freq_base
1e+06
1e+06
talkie.rope.scaling.factor
4
4
talkie.rope.scaling.original_context_length
2048
2048
talkie.rope.scaling.type
yarn
yarn
tokenizer.ggml.eos_token_id
65535
65535
tokenizer.ggml.merges
[Ġ t, h e, Ġ a, i n, Ġ the, ...]
[Ġ t, h e, Ġ a, i n, Ġ the, ...]
tokenizer.ggml.model
gpt2
gpt2
tokenizer.ggml.padding_token_id
65535
65535
tokenizer.ggml.pre
talkie
talkie
tokenizer.ggml.token_type
[1, 1, 1, 1, 1, ...]
[1, 1, 1, 1, 1, ...]
tokenizer.ggml.tokens
[Ā, ā, Ă, ă, Ą, ...]
[Ā, ā, Ă, ă, Ą, ...]
Tensor
Name
Type
Shape
token_embd.weight
Q4_K
Q4_K
[5120, 65540]
blk.0
blk.0.attn_k.weight
Q4_K
Q4_K
[5120, 5120]
blk.0.attn_output.scale
F32
F32
[1]
blk.0.attn_output.weight
Q4_K
Q4_K
[5120, 5120]
blk.0.attn_q.weight
Q4_K
Q4_K
[5120, 5120]
blk.0.attn_q_norm.weight
F32
F32
[1, 40]
blk.0.attn_v.weight
Q6_K
Q6_K
[5120, 5120]
blk.0.ffn_down.scale
F32
F32
[1]
blk.0.ffn_down.weight
Q8_0
Q8_0
[13696, 5120]
blk.0.ffn_gate.weight
Q4_K
Q4_K
[5120, 13696]
blk.0.ffn_up.weight
Q4_K
Q4_K
[5120, 13696]
blk.0.layer_output_scale.weight
F32
F32
[1]
blk.1
blk.1.attn_k.weight
Q4_K
Q4_K
[5120, 5120]
blk.1.attn_output.scale
F32
F32
[1]
blk.1.attn_output.weight
Q4_K
Q4_K
[5120, 5120]
blk.1.attn_q.weight
Q4_K
Q4_K
[5120, 5120]
blk.1.attn_q_norm.weight
F32
F32
[1, 40]
blk.1.attn_v.weight
Q6_K
Q6_K
[5120, 5120]
blk.1.ffn_down.scale
F32
F32
[1]
blk.1.ffn_down.weight
Q8_0
Q8_0
[13696, 5120]
blk.1.ffn_gate.weight
Q4_K
Q4_K
[5120, 13696]
blk.1.ffn_up.weight
Q4_K
Q4_K
[5120, 13696]
blk.1.layer_output_scale.weight
F32
F32
[1]
blk.2
blk.2.attn_k.weight
Q4_K
Q4_K
[5120, 5120]
blk.2.attn_output.scale
F32
F32
[1]
blk.2.attn_output.weight
Q4_K
Q4_K
[5120, 5120]
blk.2.attn_q.weight
Q4_K
Q4_K
[5120, 5120]
blk.2.attn_q_norm.weight
F32
F32
[1, 40]
blk.2.attn_v.weight
Q6_K
Q6_K
[5120, 5120]
blk.2.ffn_down.scale
F32
F32
[1]
blk.2.ffn_down.weight
Q8_0
Q8_0
[13696, 5120]
blk.2.ffn_gate.weight
Q4_K
Q4_K
[5120, 13696]
blk.2.ffn_up.weight
Q4_K
Q4_K
[5120, 13696]
blk.2.layer_output_scale.weight
F32
F32
[1]
blk.3
blk.3.attn_k.weight
Q4_K
Q4_K
[5120, 5120]
blk.3.attn_output.scale
F32
F32
[1]
blk.3.attn_output.weight
Q4_K
Q4_K
[5120, 5120]
blk.3.attn_q.weight
Q4_K
Q4_K
[5120, 5120]
blk.3.attn_q_norm.weight
F32
F32
[1, 40]
blk.3.attn_v.weight
Q6_K
Q6_K
[5120, 5120]
blk.3.ffn_down.scale
F32
F32
[1]
blk.3.ffn_down.weight
Q8_0
Q8_0
[13696, 5120]
blk.3.ffn_gate.weight
Q4_K
Q4_K
[5120, 13696]
blk.3.ffn_up.weight
Q4_K
Q4_K
[5120, 13696]
blk.3.layer_output_scale.weight
F32
F32
[1]
blk.4
blk.4.attn_k.weight
Q4_K
Q4_K
[5120, 5120]
blk.4.attn_output.scale
F32
F32
[1]
blk.4.attn_output.weight
Q4_K
Q4_K
[5120, 5120]
blk.4.attn_q.weight
Q4_K
Q4_K
[5120, 5120]
blk.4.attn_q_norm.weight
F32
F32
[1, 40]
blk.4.attn_v.weight
Q6_K
Q6_K
[5120, 5120]
blk.4.ffn_down.scale
F32
F32
[1]
blk.4.ffn_down.weight
Q8_0
Q8_0
[13696, 5120]
blk.4.ffn_gate.weight
Q4_K
Q4_K
[5120, 13696]
blk.4.ffn_up.weight
Q4_K
Q4_K
[5120, 13696]
blk.4.layer_output_scale.weight
F32
F32
[1]
blk.5
blk.5.attn_k.weight
Q4_K
Q4_K
[5120, 5120]
blk.5.attn_output.scale
F32
F32
[1]
blk.5.attn_output.weight
Q4_K
Q4_K
[5120, 5120]
blk.5.attn_q.weight
Q4_K
Q4_K
[5120, 5120]
blk.5.attn_q_norm.weight
F32
F32
[1, 40]
blk.5.attn_v.weight
Q4_K
Q4_K
[5120, 5120]
blk.5.ffn_down.scale
F32
F32
[1]
blk.5.ffn_down.weight
Q5_0
Q5_0
[13696, 5120]
blk.5.ffn_gate.weight
Q4_K
Q4_K
[5120, 13696]
blk.5.ffn_up.weight
Q4_K
Q4_K
[5120, 13696]
blk.5.layer_output_scale.weight
F32
F32
[1]
blk.6
blk.6.attn_k.weight
Q4_K
Q4_K
[5120, 5120]
blk.6.attn_output.scale
F32
F32
[1]
blk.6.attn_output.weight
Q4_K
Q4_K
[5120, 5120]
blk.6.attn_q.weight
Q4_K
Q4_K
[5120, 5120]
blk.6.attn_q_norm.weight
F32
F32
[1, 40]
blk.6.attn_v.weight
Q4_K
Q4_K
[5120, 5120]
blk.6.ffn_down.scale
F32
F32
[1]
blk.6.ffn_down.weight
Q5_0
Q5_0
[13696, 5120]
blk.6.ffn_gate.weight
Q4_K
Q4_K
[5120, 13696]
blk.6.ffn_up.weight
Q4_K
Q4_K
[5120, 13696]
blk.6.layer_output_scale.weight
F32
F32
[1]
blk.7
blk.7.attn_k.weight
Q4_K
Q4_K
[5120, 5120]
blk.7.attn_output.scale
F32
F32
[1]
blk.7.attn_output.weight
Q4_K
Q4_K
[5120, 5120]
blk.7.attn_q.weight
Q4_K
Q4_K
[5120, 5120]
blk.7.attn_q_norm.weight
F32
F32
[1, 40]
blk.7.attn_v.weight
Q6_K
Q6_K
[5120, 5120]
blk.7.ffn_down.scale
F32
F32
[1]
blk.7.ffn_down.weight
Q8_0
Q8_0
[13696, 5120]
blk.7.ffn_gate.weight
Q4_K
Q4_K
[5120, 13696]
blk.7.ffn_up.weight
Q4_K
Q4_K
[5120, 13696]
blk.7.layer_output_scale.weight
F32
F32
[1]
blk.8
blk.8.attn_k.weight
Q4_K
Q4_K
[5120, 5120]
blk.8.attn_output.scale
F32
F32
[1]
blk.8.attn_output.weight
Q4_K
Q4_K
[5120, 5120]
blk.8.attn_q.weight
Q4_K
Q4_K
[5120, 5120]
blk.8.attn_q_norm.weight
F32
F32
[1, 40]
blk.8.attn_v.weight
Q4_K
Q4_K
[5120, 5120]
blk.8.ffn_down.scale
F32
F32
[1]
blk.8.ffn_down.weight
Q5_0
Q5_0
[13696, 5120]
blk.8.ffn_gate.weight
Q4_K
Q4_K
[5120, 13696]
blk.8.ffn_up.weight
Q4_K
Q4_K
[5120, 13696]
blk.8.layer_output_scale.weight
F32
F32
[1]
blk.9
blk.9.attn_k.weight
Q4_K
Q4_K
[5120, 5120]
blk.9.attn_output.scale
F32
F32
[1]
blk.9.attn_output.weight
Q4_K
Q4_K
[5120, 5120]
blk.9.attn_q.weight
Q4_K
Q4_K
[5120, 5120]
blk.9.attn_q_norm.weight
F32
F32
[1, 40]
blk.9.attn_v.weight
Q4_K
Q4_K
[5120, 5120]
blk.9.ffn_down.scale
F32
F32
[1]
blk.9.ffn_down.weight
Q5_0
Q5_0
[13696, 5120]
blk.9.ffn_gate.weight
Q4_K
Q4_K
[5120, 13696]
blk.9.ffn_up.weight
Q4_K
Q4_K
[5120, 13696]
blk.9.layer_output_scale.weight
F32
F32
[1]
blk.10
blk.10.attn_k.weight
Q4_K
Q4_K
[5120, 5120]
blk.10.attn_output.scale
F32
F32
[1]
blk.10.attn_output.weight
Q4_K
Q4_K
[5120, 5120]
blk.10.attn_q.weight
Q4_K
Q4_K
[5120, 5120]
blk.10.attn_q_norm.weight
F32
F32
[1, 40]
blk.10.attn_v.weight
Q6_K
Q6_K
[5120, 5120]
blk.10.ffn_down.scale
F32
F32
[1]
blk.10.ffn_down.weight
Q8_0
Q8_0
[13696, 5120]
blk.10.ffn_gate.weight
Q4_K
Q4_K
[5120, 13696]
blk.10.ffn_up.weight
Q4_K
Q4_K
[5120, 13696]
blk.10.layer_output_scale.weight
F32
F32
[1]
blk.11
blk.11.attn_k.weight
Q4_K
Q4_K
[5120, 5120]
blk.11.attn_output.scale
F32
F32
[1]
blk.11.attn_output.weight
Q4_K
Q4_K
[5120, 5120]
blk.11.attn_q.weight
Q4_K
Q4_K
[5120, 5120]
blk.11.attn_q_norm.weight
F32
F32
[1, 40]
blk.11.attn_v.weight
Q4_K
Q4_K
[5120, 5120]
blk.11.ffn_down.scale
F32
F32
[1]
blk.11.ffn_down.weight
Q5_0
Q5_0
[13696, 5120]
blk.11.ffn_gate.weight
Q4_K
Q4_K
[5120, 13696]
blk.11.ffn_up.weight
Q4_K
Q4_K
[5120, 13696]
blk.11.layer_output_scale.weight
F32
F32
[1]
blk.12
blk.12.attn_k.weight
Q4_K
Q4_K
[5120, 5120]
blk.12.attn_output.scale
F32
F32
[1]
blk.12.attn_output.weight
Q4_K
Q4_K
[5120, 5120]
blk.12.attn_q.weight
Q4_K
Q4_K
[5120, 5120]
blk.12.attn_q_norm.weight
F32
F32
[1, 40]
blk.12.attn_v.weight
Q4_K
Q4_K
[5120, 5120]
blk.12.ffn_down.scale
F32
F32
[1]
blk.12.ffn_down.weight
Q5_0
Q5_0
[13696, 5120]
blk.12.ffn_gate.weight
Q4_K
Q4_K
[5120, 13696]
blk.12.ffn_up.weight
Q4_K
Q4_K
[5120, 13696]
blk.12.layer_output_scale.weight
F32
F32
[1]
blk.13
blk.13.attn_k.weight
Q4_K
Q4_K
[5120, 5120]
blk.13.attn_output.scale
F32
F32
[1]
blk.13.attn_output.weight
Q4_K
Q4_K
[5120, 5120]
blk.13.attn_q.weight
Q4_K
Q4_K
[5120, 5120]
blk.13.attn_q_norm.weight
F32
F32
[1, 40]
blk.13.attn_v.weight
Q6_K
Q6_K
[5120, 5120]
blk.13.ffn_down.scale
F32
F32
[1]
blk.13.ffn_down.weight
Q8_0
Q8_0
[13696, 5120]
blk.13.ffn_gate.weight
Q4_K
Q4_K
[5120, 13696]
blk.13.ffn_up.weight
Q4_K
Q4_K
[5120, 13696]
blk.13.layer_output_scale.weight
F32
F32
[1]
blk.14
blk.14.attn_k.weight
Q4_K
Q4_K
[5120, 5120]
blk.14.attn_output.scale
F32
F32
[1]
blk.14.attn_output.weight
Q4_K
Q4_K
[5120, 5120]
blk.14.attn_q.weight
Q4_K
Q4_K
[5120, 5120]
blk.14.attn_q_norm.weight
F32
F32
[1, 40]
blk.14.attn_v.weight
Q4_K
Q4_K
[5120, 5120]
blk.14.ffn_down.scale
F32
F32
[1]
blk.14.ffn_down.weight
Q5_0
Q5_0
[13696, 5120]
blk.14.ffn_gate.weight
Q4_K
Q4_K
[5120, 13696]
blk.14.ffn_up.weight
Q4_K
Q4_K
[5120, 13696]
blk.14.layer_output_scale.weight
F32
F32
[1]
blk.15
blk.15.attn_k.weight
Q4_K
Q4_K
[5120, 5120]
blk.15.attn_output.scale
F32
F32
[1]
blk.15.attn_output.weight
Q4_K
Q4_K
[5120, 5120]
blk.15.attn_q.weight
Q4_K
Q4_K
[5120, 5120]
blk.15.attn_q_norm.weight
F32
F32
[1, 40]
blk.15.attn_v.weight
Q4_K
Q4_K
[5120, 5120]
blk.15.ffn_down.scale
F32
F32
[1]
blk.15.ffn_down.weight
Q5_0
Q5_0
[13696, 5120]
blk.15.ffn_gate.weight
Q4_K
Q4_K
[5120, 13696]
blk.15.ffn_up.weight
Q4_K
Q4_K
[5120, 13696]
blk.15.layer_output_scale.weight
F32
F32
[1]
blk.16
blk.16.attn_k.weight
Q4_K
Q4_K
[5120, 5120]
blk.16.attn_output.scale
F32
F32
[1]
blk.16.attn_output.weight
Q4_K
Q4_K
[5120, 5120]
blk.16.attn_q.weight
Q4_K
Q4_K
[5120, 5120]
blk.16.attn_q_norm.weight
F32
F32
[1, 40]
blk.16.attn_v.weight
Q6_K
Q6_K
[5120, 5120]
blk.16.ffn_down.scale
F32
F32
[1]
blk.16.ffn_down.weight
Q8_0
Q8_0
[13696, 5120]
blk.16.ffn_gate.weight
Q4_K
Q4_K
[5120, 13696]
blk.16.ffn_up.weight
Q4_K
Q4_K
[5120, 13696]
blk.16.layer_output_scale.weight
F32
F32
[1]
blk.17
blk.17.attn_k.weight
Q4_K
Q4_K
[5120, 5120]
blk.17.attn_output.scale
F32
F32
[1]
blk.17.attn_output.weight
Q4_K
Q4_K
[5120, 5120]
blk.17.attn_q.weight
Q4_K
Q4_K
[5120, 5120]
blk.17.attn_q_norm.weight
F32
F32
[1, 40]
blk.17.attn_v.weight
Q4_K
Q4_K
[5120, 5120]
blk.17.ffn_down.scale
F32
F32
[1]
blk.17.ffn_down.weight
Q5_0
Q5_0
[13696, 5120]
blk.17.ffn_gate.weight
Q4_K
Q4_K
[5120, 13696]
blk.17.ffn_up.weight
Q4_K
Q4_K
[5120, 13696]
blk.17.layer_output_scale.weight
F32
F32
[1]
blk.18
blk.18.attn_k.weight
Q4_K
Q4_K
[5120, 5120]
blk.18.attn_output.scale
F32
F32
[1]
blk.18.attn_output.weight
Q4_K
Q4_K
[5120, 5120]
blk.18.attn_q.weight
Q4_K
Q4_K
[5120, 5120]
blk.18.attn_q_norm.weight
F32
F32
[1, 40]
blk.18.attn_v.weight
Q4_K
Q4_K
[5120, 5120]
blk.18.ffn_down.scale
F32
F32
[1]
blk.18.ffn_down.weight
Q5_0
Q5_0
[13696, 5120]
blk.18.ffn_gate.weight
Q4_K
Q4_K
[5120, 13696]
blk.18.ffn_up.weight
Q4_K
Q4_K
[5120, 13696]
blk.18.layer_output_scale.weight
F32
F32
[1]
blk.19
blk.19.attn_k.weight
Q4_K
Q4_K
[5120, 5120]
blk.19.attn_output.scale
F32
F32
[1]
blk.19.attn_output.weight
Q4_K
Q4_K
[5120, 5120]
blk.19.attn_q.weight
Q4_K
Q4_K
[5120, 5120]
blk.19.attn_q_norm.weight
F32
F32
[1, 40]
blk.19.attn_v.weight
Q6_K
Q6_K
[5120, 5120]
blk.19.ffn_down.scale
F32
F32
[1]
blk.19.ffn_down.weight
Q8_0
Q8_0
[13696, 5120]
blk.19.ffn_gate.weight
Q4_K
Q4_K
[5120, 13696]
blk.19.ffn_up.weight
Q4_K
Q4_K
[5120, 13696]
blk.19.layer_output_scale.weight
F32
F32
[1]
blk.20
blk.20.attn_k.weight
Q4_K
Q4_K
[5120, 5120]
blk.20.attn_output.scale
F32
F32
[1]
blk.20.attn_output.weight
Q4_K
Q4_K
[5120, 5120]
blk.20.attn_q.weight
Q4_K
Q4_K
[5120, 5120]
blk.20.attn_q_norm.weight
F32
F32
[1, 40]
blk.20.attn_v.weight
Q4_K
Q4_K
[5120, 5120]
blk.20.ffn_down.scale
F32
F32
[1]
blk.20.ffn_down.weight
Q5_0
Q5_0
[13696, 5120]
blk.20.ffn_gate.weight
Q4_K
Q4_K
[5120, 13696]
blk.20.ffn_up.weight
Q4_K
Q4_K
[5120, 13696]
blk.20.layer_output_scale.weight
F32
F32
[1]
blk.21
blk.21.attn_k.weight
Q4_K
Q4_K
[5120, 5120]
blk.21.attn_output.scale
F32
F32
[1]
blk.21.attn_output.weight
Q4_K
Q4_K
[5120, 5120]
blk.21.attn_q.weight
Q4_K
Q4_K
[5120, 5120]
blk.21.attn_q_norm.weight
F32
F32
[1, 40]
blk.21.attn_v.weight
Q4_K
Q4_K
[5120, 5120]
blk.21.ffn_down.scale
F32
F32
[1]
blk.21.ffn_down.weight
Q5_0
Q5_0
[13696, 5120]
blk.21.ffn_gate.weight
Q4_K
Q4_K
[5120, 13696]
blk.21.ffn_up.weight
Q4_K
Q4_K
[5120, 13696]
blk.21.layer_output_scale.weight
F32
F32
[1]
blk.22
blk.22.attn_k.weight
Q4_K
Q4_K
[5120, 5120]
blk.22.attn_output.scale
F32
F32
[1]
blk.22.attn_output.weight
Q4_K
Q4_K
[5120, 5120]
blk.22.attn_q.weight
Q4_K
Q4_K
[5120, 5120]
blk.22.attn_q_norm.weight
F32
F32
[1, 40]
blk.22.attn_v.weight
Q6_K
Q6_K
[5120, 5120]
blk.22.ffn_down.scale
F32
F32
[1]
blk.22.ffn_down.weight
Q8_0
Q8_0
[13696, 5120]
blk.22.ffn_gate.weight
Q4_K
Q4_K
[5120, 13696]
blk.22.ffn_up.weight
Q4_K
Q4_K
[5120, 13696]
blk.22.layer_output_scale.weight
F32
F32
[1]
blk.23
blk.23.attn_k.weight
Q4_K
Q4_K
[5120, 5120]
blk.23.attn_output.scale
F32
F32
[1]
blk.23.attn_output.weight
Q4_K
Q4_K
[5120, 5120]
blk.23.attn_q.weight
Q4_K
Q4_K
[5120, 5120]
blk.23.attn_q_norm.weight
F32
F32
[1, 40]
blk.23.attn_v.weight
Q4_K
Q4_K
[5120, 5120]
blk.23.ffn_down.scale
F32
F32
[1]
blk.23.ffn_down.weight
Q5_0
Q5_0
[13696, 5120]
blk.23.ffn_gate.weight
Q4_K
Q4_K
[5120, 13696]
blk.23.ffn_up.weight
Q4_K
Q4_K
[5120, 13696]
blk.23.layer_output_scale.weight
F32
F32
[1]
blk.24
blk.24.attn_k.weight
Q4_K
Q4_K
[5120, 5120]
blk.24.attn_output.scale
F32
F32
[1]
blk.24.attn_output.weight
Q4_K
Q4_K
[5120, 5120]
blk.24.attn_q.weight
Q4_K
Q4_K
[5120, 5120]
blk.24.attn_q_norm.weight
F32
F32
[1, 40]
blk.24.attn_v.weight
Q4_K
Q4_K
[5120, 5120]
blk.24.ffn_down.scale
F32
F32
[1]
blk.24.ffn_down.weight
Q5_0
Q5_0
[13696, 5120]
blk.24.ffn_gate.weight
Q4_K
Q4_K
[5120, 13696]
blk.24.ffn_up.weight
Q4_K
Q4_K
[5120, 13696]
blk.24.layer_output_scale.weight
F32
F32
[1]
blk.25
blk.25.attn_k.weight
Q4_K
Q4_K
[5120, 5120]
blk.25.attn_output.scale
F32
F32
[1]
blk.25.attn_output.weight
Q4_K
Q4_K
[5120, 5120]
blk.25.attn_q.weight
Q4_K
Q4_K
[5120, 5120]
blk.25.attn_q_norm.weight
F32
F32
[1, 40]
blk.25.attn_v.weight
Q6_K
Q6_K
[5120, 5120]
blk.25.ffn_down.scale
F32
F32
[1]
blk.25.ffn_down.weight
Q8_0
Q8_0
[13696, 5120]
blk.25.ffn_gate.weight
Q4_K
Q4_K
[5120, 13696]
blk.25.ffn_up.weight
Q4_K
Q4_K
[5120, 13696]
blk.25.layer_output_scale.weight
F32
F32
[1]
blk.26
blk.26.attn_k.weight
Q4_K
Q4_K
[5120, 5120]
blk.26.attn_output.scale
F32
F32
[1]
blk.26.attn_output.weight
Q4_K
Q4_K
[5120, 5120]
blk.26.attn_q.weight
Q4_K
Q4_K
[5120, 5120]
blk.26.attn_q_norm.weight
F32
F32
[1, 40]
blk.26.attn_v.weight
Q4_K
Q4_K
[5120, 5120]
blk.26.ffn_down.scale
F32
F32
[1]
blk.26.ffn_down.weight
Q5_0
Q5_0
[13696, 5120]
blk.26.ffn_gate.weight
Q4_K
Q4_K
[5120, 13696]
blk.26.ffn_up.weight
Q4_K
Q4_K
[5120, 13696]
blk.26.layer_output_scale.weight
F32
F32
[1]
blk.27
blk.27.attn_k.weight
Q4_K
Q4_K
[5120, 5120]
blk.27.attn_output.scale
F32
F32
[1]
blk.27.attn_output.weight
Q4_K
Q4_K
[5120, 5120]
blk.27.attn_q.weight
Q4_K
Q4_K
[5120, 5120]
blk.27.attn_q_norm.weight
F32
F32
[1, 40]
blk.27.attn_v.weight
Q4_K
Q4_K
[5120, 5120]
blk.27.ffn_down.scale
F32
F32
[1]
blk.27.ffn_down.weight
Q5_0
Q5_0
[13696, 5120]
blk.27.ffn_gate.weight
Q4_K
Q4_K
[5120, 13696]
blk.27.ffn_up.weight
Q4_K
Q4_K
[5120, 13696]
blk.27.layer_output_scale.weight
F32
F32
[1]
blk.28
blk.28.attn_k.weight
Q4_K
Q4_K
[5120, 5120]
blk.28.attn_output.scale
F32
F32
[1]
blk.28.attn_output.weight
Q4_K
Q4_K
[5120, 5120]
blk.28.attn_q.weight
Q4_K
Q4_K
[5120, 5120]
blk.28.attn_q_norm.weight
F32
F32
[1, 40]
blk.28.attn_v.weight
Q6_K
Q6_K
[5120, 5120]
blk.28.ffn_down.scale
F32
F32
[1]
blk.28.ffn_down.weight
Q8_0
Q8_0
[13696, 5120]
blk.28.ffn_gate.weight
Q4_K
Q4_K
[5120, 13696]
blk.28.ffn_up.weight
Q4_K
Q4_K
[5120, 13696]
blk.28.layer_output_scale.weight
F32
F32
[1]
blk.29
blk.29.attn_k.weight
Q4_K
Q4_K
[5120, 5120]
blk.29.attn_output.scale
F32
F32
[1]
blk.29.attn_output.weight
Q4_K
Q4_K
[5120, 5120]
blk.29.attn_q.weight
Q4_K
Q4_K
[5120, 5120]
blk.29.attn_q_norm.weight
F32
F32
[1, 40]
blk.29.attn_v.weight
Q4_K
Q4_K
[5120, 5120]
blk.29.ffn_down.scale
F32
F32
[1]
blk.29.ffn_down.weight
Q5_0
Q5_0
[13696, 5120]
blk.29.ffn_gate.weight
Q4_K
Q4_K
[5120, 13696]
blk.29.ffn_up.weight
Q4_K
Q4_K
[5120, 13696]
blk.29.layer_output_scale.weight
F32
F32
[1]
blk.30
blk.30.attn_k.weight
Q4_K
Q4_K
[5120, 5120]
blk.30.attn_output.scale
F32
F32
[1]
blk.30.attn_output.weight
Q4_K
Q4_K
[5120, 5120]
blk.30.attn_q.weight
Q4_K
Q4_K
[5120, 5120]
blk.30.attn_q_norm.weight
F32
F32
[1, 40]
blk.30.attn_v.weight
Q4_K
Q4_K
[5120, 5120]
blk.30.ffn_down.scale
F32
F32
[1]
blk.30.ffn_down.weight
Q5_0
Q5_0
[13696, 5120]
blk.30.ffn_gate.weight
Q4_K
Q4_K
[5120, 13696]
blk.30.ffn_up.weight
Q4_K
Q4_K
[5120, 13696]
blk.30.layer_output_scale.weight
F32
F32
[1]
blk.31
blk.31.attn_k.weight
Q4_K
Q4_K
[5120, 5120]
blk.31.attn_output.scale
F32
F32
[1]
blk.31.attn_output.weight
Q4_K
Q4_K
[5120, 5120]
blk.31.attn_q.weight
Q4_K
Q4_K
[5120, 5120]
blk.31.attn_q_norm.weight
F32
F32
[1, 40]
blk.31.attn_v.weight
Q6_K
Q6_K
[5120, 5120]
blk.31.ffn_down.scale
F32
F32
[1]
blk.31.ffn_down.weight
Q8_0
Q8_0
[13696, 5120]
blk.31.ffn_gate.weight
Q4_K
Q4_K
[5120, 13696]
blk.31.ffn_up.weight
Q4_K
Q4_K
[5120, 13696]
blk.31.layer_output_scale.weight
F32
F32
[1]
blk.32
blk.32.attn_k.weight
Q4_K
Q4_K
[5120, 5120]
blk.32.attn_output.scale
F32
F32
[1]
blk.32.attn_output.weight
Q4_K
Q4_K
[5120, 5120]
blk.32.attn_q.weight
Q4_K
Q4_K
[5120, 5120]
blk.32.attn_q_norm.weight
F32
F32
[1, 40]
blk.32.attn_v.weight
Q4_K
Q4_K
[5120, 5120]
blk.32.ffn_down.scale
F32
F32
[1]
blk.32.ffn_down.weight
Q5_0
Q5_0
[13696, 5120]
blk.32.ffn_gate.weight
Q4_K
Q4_K
[5120, 13696]
blk.32.ffn_up.weight
Q4_K
Q4_K
[5120, 13696]
blk.32.layer_output_scale.weight
F32
F32
[1]
blk.33
blk.33.attn_k.weight
Q4_K
Q4_K
[5120, 5120]
blk.33.attn_output.scale
F32
F32
[1]
blk.33.attn_output.weight
Q4_K
Q4_K
[5120, 5120]
blk.33.attn_q.weight
Q4_K
Q4_K
[5120, 5120]
blk.33.attn_q_norm.weight
F32
F32
[1, 40]
blk.33.attn_v.weight
Q4_K
Q4_K
[5120, 5120]
blk.33.ffn_down.scale
F32
F32
[1]
blk.33.ffn_down.weight
Q5_0
Q5_0
[13696, 5120]
blk.33.ffn_gate.weight
Q4_K
Q4_K
[5120, 13696]
blk.33.ffn_up.weight
Q4_K
Q4_K
[5120, 13696]
blk.33.layer_output_scale.weight
F32
F32
[1]
blk.34
blk.34.attn_k.weight
Q4_K
Q4_K
[5120, 5120]
blk.34.attn_output.scale
F32
F32
[1]
blk.34.attn_output.weight
Q4_K
Q4_K
[5120, 5120]
blk.34.attn_q.weight
Q4_K
Q4_K
[5120, 5120]
blk.34.attn_q_norm.weight
F32
F32
[1, 40]
blk.34.attn_v.weight
Q6_K
Q6_K
[5120, 5120]
blk.34.ffn_down.scale
F32
F32
[1]
blk.34.ffn_down.weight
Q8_0
Q8_0
[13696, 5120]
blk.34.ffn_gate.weight
Q4_K
Q4_K
[5120, 13696]
blk.34.ffn_up.weight
Q4_K
Q4_K
[5120, 13696]
blk.34.layer_output_scale.weight
F32
F32
[1]
blk.35
blk.35.attn_k.weight
Q4_K
Q4_K
[5120, 5120]
blk.35.attn_output.scale
F32
F32
[1]
blk.35.attn_output.weight
Q4_K
Q4_K
[5120, 5120]
blk.35.attn_q.weight
Q4_K
Q4_K
[5120, 5120]
blk.35.attn_q_norm.weight
F32
F32
[1, 40]
blk.35.attn_v.weight
Q6_K
Q6_K
[5120, 5120]
blk.35.ffn_down.scale
F32
F32
[1]
blk.35.ffn_down.weight
Q8_0
Q8_0
[13696, 5120]
blk.35.ffn_gate.weight
Q4_K
Q4_K
[5120, 13696]
blk.35.ffn_up.weight
Q4_K
Q4_K
[5120, 13696]
blk.35.layer_output_scale.weight
F32
F32
[1]
blk.36
blk.36.attn_k.weight
Q4_K
Q4_K
[5120, 5120]
blk.36.attn_output.scale
F32
F32
[1]
blk.36.attn_output.weight
Q4_K
Q4_K
[5120, 5120]
blk.36.attn_q.weight
Q4_K
Q4_K
[5120, 5120]
blk.36.attn_q_norm.weight
F32
F32
[1, 40]
blk.36.attn_v.weight
Q6_K
Q6_K
[5120, 5120]
blk.36.ffn_down.scale
F32
F32
[1]
blk.36.ffn_down.weight
Q8_0
Q8_0
[13696, 5120]
blk.36.ffn_gate.weight
Q4_K
Q4_K
[5120, 13696]
blk.36.ffn_up.weight
Q4_K
Q4_K
[5120, 13696]
blk.36.layer_output_scale.weight
F32
F32
[1]
blk.37
blk.37.attn_k.weight
Q4_K
Q4_K
[5120, 5120]
blk.37.attn_output.scale
F32
F32
[1]
blk.37.attn_output.weight
Q4_K
Q4_K
[5120, 5120]
blk.37.attn_q.weight
Q4_K
Q4_K
[5120, 5120]
blk.37.attn_q_norm.weight
F32
F32
[1, 40]
blk.37.attn_v.weight
Q6_K
Q6_K
[5120, 5120]
blk.37.ffn_down.scale
F32
F32
[1]
blk.37.ffn_down.weight
Q8_0
Q8_0
[13696, 5120]
blk.37.ffn_gate.weight
Q4_K
Q4_K
[5120, 13696]
blk.37.ffn_up.weight
Q4_K
Q4_K
[5120, 13696]
blk.37.layer_output_scale.weight
F32
F32
[1]
blk.38
blk.38.attn_k.weight
Q4_K
Q4_K
[5120, 5120]
blk.38.attn_output.scale
F32
F32
[1]
blk.38.attn_output.weight
Q4_K
Q4_K
[5120, 5120]
blk.38.attn_q.weight
Q4_K
Q4_K
[5120, 5120]
blk.38.attn_q_norm.weight
F32
F32
[1, 40]
blk.38.attn_v.weight
Q6_K
Q6_K
[5120, 5120]
blk.38.ffn_down.scale
F32
F32
[1]
blk.38.ffn_down.weight
Q8_0
Q8_0
[13696, 5120]
blk.38.ffn_gate.weight
Q4_K
Q4_K
[5120, 13696]
blk.38.ffn_up.weight
Q4_K
Q4_K
[5120, 13696]
blk.38.layer_output_scale.weight
F32
F32
[1]
blk.39
blk.39.attn_k.weight
Q4_K
Q4_K
[5120, 5120]
blk.39.attn_output.scale
F32
F32
[1]
blk.39.attn_output.weight
Q4_K
Q4_K
[5120, 5120]
blk.39.attn_q.weight
Q4_K
Q4_K
[5120, 5120]
blk.39.attn_q_norm.weight
F32
F32
[1, 40]
blk.39.attn_v.weight
Q6_K
Q6_K
[5120, 5120]
blk.39.ffn_down.scale
F32
F32
[1]
blk.39.ffn_down.weight
Q8_0
Q8_0
[13696, 5120]
blk.39.ffn_gate.weight
Q4_K
Q4_K
[5120, 13696]
blk.39.ffn_up.weight
Q4_K
Q4_K
[5120, 13696]
blk.39.layer_output_scale.weight
F32
F32
[1]
output.weight
Q6_K
Q6_K
[5120, 65540]