Models
GitHub
Discord
Docs
Cloud
Sign in
Download
Models
Download
GitHub
Discord
Docs
Cloud
Sign in
mirage335
/
NVIDIA-Nemotron-Nano-9B-v2-virtuoso
:latest
59
Downloads
Updated
2 weeks ago
Cancel
NVIDIA-Nemotron-Nano-9B-v2-virtuoso:latest
...
/
model
95ee1b5339df · 9.1GB
Metadata
general.architecture
nemotron_h
nemotron_h
general.file_type
Q6_K
Q6_K
nemotron_h.attention.head_count
40
40
nemotron_h.attention.head_count_kv
[0, 0, 0, 0, 0, ...]
[0, 0, 0, 0, 0, ...]
nemotron_h.attention.key_length
128
128
nemotron_h.attention.layer_norm_epsilon
1e-05
1e-05
nemotron_h.attention.layer_norm_rms_epsilon
1e-05
1e-05
nemotron_h.attention.value_length
128
128
nemotron_h.block_count
56
56
nemotron_h.context_length
1048576
1048576
nemotron_h.embedding_length
4480
4480
nemotron_h.feed_forward_length
[0, 15680, 0, 15680, 0, ...]
[0, 15680, 0, 15680, 0, ...]
nemotron_h.rope.dimension_count
112
112
nemotron_h.rope.scaling.finetuned
false
false
nemotron_h.ssm.conv_kernel
4
4
nemotron_h.ssm.group_count
8
8
nemotron_h.ssm.inner_size
10240
10240
nemotron_h.ssm.state_size
128
128
nemotron_h.ssm.time_step_rank
128
128
nemotron_h.vocab_size
131072
131072
tokenizer.ggml.add_bos_token
true
true
tokenizer.ggml.add_eos_token
false
false
tokenizer.ggml.add_sep_token
false
false
tokenizer.ggml.bos_token_id
1
1
tokenizer.ggml.eos_token_id
12
12
tokenizer.ggml.merges
[Ġ Ġ, Ġ t, e r, i n, Ġ ĠĠĠ, ...]
[Ġ Ġ, Ġ t, e r, i n, Ġ ĠĠĠ, ...]
tokenizer.ggml.model
gpt2
gpt2
tokenizer.ggml.padding_token_id
0
0
tokenizer.ggml.pre
tekken
tekken
tokenizer.ggml.token_type
[3, 3, 3, 3, 3, ...]
[3, 3, 3, 3, 3, ...]
tokenizer.ggml.tokens
[<unk>, <s>, </s>, [INST], [/INST], ...]
[<unk>, <s>, </s>, [INST], [/INST], ...]
tokenizer.ggml.unknown_token_id
0
0
quantize.imatrix.chunks_count
822
822
quantize.imatrix.dataset
/training_dir/calibration_datav5.txt
/training_dir/calibration_datav5.txt
quantize.imatrix.entries_count
120
120
quantize.imatrix.file
/models_out/NVIDIA-Nemotron-Nano-9B-v2-GGUF/nvidia_NVIDIA-Nemotron-Nano-9B-v2-imatrix.gguf
/models_out/NVIDIA-Nemotron-Nano-9B-v2-GGUF/nvidia_NVIDIA-Nemotron-Nano-9B-v2-imatrix.gguf
Tensor
Name
Type
Shape
token_embd.weight
Q8_0
Q8_0
[4480, 131072]
blk.0
blk.0.attn_norm.weight
F32
F32
[4480]
blk.0.ssm_a
F32
F32
[1, 128]
blk.0.ssm_conv1d.bias
F32
F32
[12288]
blk.0.ssm_conv1d.weight
F32
F32
[4, 12288]
blk.0.ssm_d
F32
F32
[1, 128]
blk.0.ssm_dt.bias
F32
F32
[128]
blk.0.ssm_in.weight
Q8_0
Q8_0
[4480, 22656]
blk.0.ssm_norm.weight
F32
F32
[1280, 8]
blk.0.ssm_out.weight
Q6_K
Q6_K
[10240, 4480]
blk.1
blk.1.attn_norm.weight
F32
F32
[4480]
blk.1.ffn_down.weight
Q8_0
Q8_0
[15680, 4480]
blk.1.ffn_up.weight
Q8_0
Q8_0
[4480, 15680]
blk.2
blk.2.attn_norm.weight
F32
F32
[4480]
blk.2.ssm_a
F32
F32
[1, 128]
blk.2.ssm_conv1d.bias
F32
F32
[12288]
blk.2.ssm_conv1d.weight
F32
F32
[4, 12288]
blk.2.ssm_d
F32
F32
[1, 128]
blk.2.ssm_dt.bias
F32
F32
[128]
blk.2.ssm_in.weight
Q8_0
Q8_0
[4480, 22656]
blk.2.ssm_norm.weight
F32
F32
[1280, 8]
blk.2.ssm_out.weight
Q6_K
Q6_K
[10240, 4480]
blk.3
blk.3.attn_norm.weight
F32
F32
[4480]
blk.3.ffn_down.weight
Q8_0
Q8_0
[15680, 4480]
blk.3.ffn_up.weight
Q8_0
Q8_0
[4480, 15680]
blk.4
blk.4.attn_norm.weight
F32
F32
[4480]
blk.4.ssm_a
F32
F32
[1, 128]
blk.4.ssm_conv1d.bias
F32
F32
[12288]
blk.4.ssm_conv1d.weight
F32
F32
[4, 12288]
blk.4.ssm_d
F32
F32
[1, 128]
blk.4.ssm_dt.bias
F32
F32
[128]
blk.4.ssm_in.weight
Q8_0
Q8_0
[4480, 22656]
blk.4.ssm_norm.weight
F32
F32
[1280, 8]
blk.4.ssm_out.weight
Q6_K
Q6_K
[10240, 4480]
blk.5
blk.5.attn_norm.weight
F32
F32
[4480]
blk.5.ffn_down.weight
Q8_0
Q8_0
[15680, 4480]
blk.5.ffn_up.weight
Q8_0
Q8_0
[4480, 15680]
blk.6
blk.6.attn_norm.weight
F32
F32
[4480]
blk.6.ssm_a
F32
F32
[1, 128]
blk.6.ssm_conv1d.bias
F32
F32
[12288]
blk.6.ssm_conv1d.weight
F32
F32
[4, 12288]
blk.6.ssm_d
F32
F32
[1, 128]
blk.6.ssm_dt.bias
F32
F32
[128]
blk.6.ssm_in.weight
Q8_0
Q8_0
[4480, 22656]
blk.6.ssm_norm.weight
F32
F32
[1280, 8]
blk.6.ssm_out.weight
Q6_K
Q6_K
[10240, 4480]
blk.7
blk.7.attn_norm.weight
F32
F32
[4480]
blk.7.ssm_a
F32
F32
[1, 128]
blk.7.ssm_conv1d.bias
F32
F32
[12288]
blk.7.ssm_conv1d.weight
F32
F32
[4, 12288]
blk.7.ssm_d
F32
F32
[1, 128]
blk.7.ssm_dt.bias
F32
F32
[128]
blk.7.ssm_in.weight
Q8_0
Q8_0
[4480, 22656]
blk.7.ssm_norm.weight
F32
F32
[1280, 8]
blk.7.ssm_out.weight
Q6_K
Q6_K
[10240, 4480]
blk.8
blk.8.attn_norm.weight
F32
F32
[4480]
blk.8.ffn_down.weight
Q8_0
Q8_0
[15680, 4480]
blk.8.ffn_up.weight
Q8_0
Q8_0
[4480, 15680]
blk.9
blk.9.attn_norm.weight
F32
F32
[4480]
blk.9.ssm_a
F32
F32
[1, 128]
blk.9.ssm_conv1d.bias
F32
F32
[12288]
blk.9.ssm_conv1d.weight
F32
F32
[4, 12288]
blk.9.ssm_d
F32
F32
[1, 128]
blk.9.ssm_dt.bias
F32
F32
[128]
blk.9.ssm_in.weight
Q8_0
Q8_0
[4480, 22656]
blk.9.ssm_norm.weight
F32
F32
[1280, 8]
blk.9.ssm_out.weight
Q6_K
Q6_K
[10240, 4480]
blk.10
blk.10.attn_norm.weight
F32
F32
[4480]
blk.10.ffn_down.weight
Q8_0
Q8_0
[15680, 4480]
blk.10.ffn_up.weight
Q8_0
Q8_0
[4480, 15680]
blk.11
blk.11.attn_norm.weight
F32
F32
[4480]
blk.11.ssm_a
F32
F32
[1, 128]
blk.11.ssm_conv1d.bias
F32
F32
[12288]
blk.11.ssm_conv1d.weight
F32
F32
[4, 12288]
blk.11.ssm_d
F32
F32
[1, 128]
blk.11.ssm_dt.bias
F32
F32
[128]
blk.11.ssm_in.weight
Q8_0
Q8_0
[4480, 22656]
blk.11.ssm_norm.weight
F32
F32
[1280, 8]
blk.11.ssm_out.weight
Q6_K
Q6_K
[10240, 4480]
blk.12
blk.12.attn_norm.weight
F32
F32
[4480]
blk.12.ffn_down.weight
Q8_0
Q8_0
[15680, 4480]
blk.12.ffn_up.weight
Q8_0
Q8_0
[4480, 15680]
blk.13
blk.13.attn_norm.weight
F32
F32
[4480]
blk.13.ssm_a
F32
F32
[1, 128]
blk.13.ssm_conv1d.bias
F32
F32
[12288]
blk.13.ssm_conv1d.weight
F32
F32
[4, 12288]
blk.13.ssm_d
F32
F32
[1, 128]
blk.13.ssm_dt.bias
F32
F32
[128]
blk.13.ssm_in.weight
Q8_0
Q8_0
[4480, 22656]
blk.13.ssm_norm.weight
F32
F32
[1280, 8]
blk.13.ssm_out.weight
Q6_K
Q6_K
[10240, 4480]
blk.14
blk.14.attn_k.weight
Q8_0
Q8_0
[4480, 1024]
blk.14.attn_norm.weight
F32
F32
[4480]
blk.14.attn_output.weight
Q6_K
Q6_K
[5120, 4480]
blk.14.attn_q.weight
Q8_0
Q8_0
[4480, 5120]
blk.14.attn_v.weight
Q8_0
Q8_0
[4480, 1024]
blk.15
blk.15.attn_norm.weight
F32
F32
[4480]
blk.15.ffn_down.weight
Q8_0
Q8_0
[15680, 4480]
blk.15.ffn_up.weight
Q8_0
Q8_0
[4480, 15680]
blk.16
blk.16.attn_norm.weight
F32
F32
[4480]
blk.16.ssm_a
F32
F32
[1, 128]
blk.16.ssm_conv1d.bias
F32
F32
[12288]
blk.16.ssm_conv1d.weight
F32
F32
[4, 12288]
blk.16.ssm_d
F32
F32
[1, 128]
blk.16.ssm_dt.bias
F32
F32
[128]
blk.16.ssm_in.weight
Q8_0
Q8_0
[4480, 22656]
blk.16.ssm_norm.weight
F32
F32
[1280, 8]
blk.16.ssm_out.weight
Q6_K
Q6_K
[10240, 4480]
blk.17
blk.17.attn_norm.weight
F32
F32
[4480]
blk.17.ffn_down.weight
Q8_0
Q8_0
[15680, 4480]
blk.17.ffn_up.weight
Q8_0
Q8_0
[4480, 15680]
blk.18
blk.18.attn_norm.weight
F32
F32
[4480]
blk.18.ssm_a
F32
F32
[1, 128]
blk.18.ssm_conv1d.bias
F32
F32
[12288]
blk.18.ssm_conv1d.weight
F32
F32
[4, 12288]
blk.18.ssm_d
F32
F32
[1, 128]
blk.18.ssm_dt.bias
F32
F32
[128]
blk.18.ssm_in.weight
Q8_0
Q8_0
[4480, 22656]
blk.18.ssm_norm.weight
F32
F32
[1280, 8]
blk.18.ssm_out.weight
Q6_K
Q6_K
[10240, 4480]
blk.19
blk.19.attn_norm.weight
F32
F32
[4480]
blk.19.ffn_down.weight
Q8_0
Q8_0
[15680, 4480]
blk.19.ffn_up.weight
Q8_0
Q8_0
[4480, 15680]
blk.20
blk.20.attn_norm.weight
F32
F32
[4480]
blk.20.ssm_a
F32
F32
[1, 128]
blk.20.ssm_conv1d.bias
F32
F32
[12288]
blk.20.ssm_conv1d.weight
F32
F32
[4, 12288]
blk.20.ssm_d
F32
F32
[1, 128]
blk.20.ssm_dt.bias
F32
F32
[128]
blk.20.ssm_in.weight
Q8_0
Q8_0
[4480, 22656]
blk.20.ssm_norm.weight
F32
F32
[1280, 8]
blk.20.ssm_out.weight
Q6_K
Q6_K
[10240, 4480]
blk.21
blk.21.attn_k.weight
Q8_0
Q8_0
[4480, 1024]
blk.21.attn_norm.weight
F32
F32
[4480]
blk.21.attn_output.weight
Q6_K
Q6_K
[5120, 4480]
blk.21.attn_q.weight
Q8_0
Q8_0
[4480, 5120]
blk.21.attn_v.weight
Q8_0
Q8_0
[4480, 1024]
blk.22
blk.22.attn_norm.weight
F32
F32
[4480]
blk.22.ffn_down.weight
Q8_0
Q8_0
[15680, 4480]
blk.22.ffn_up.weight
Q8_0
Q8_0
[4480, 15680]
blk.23
blk.23.attn_norm.weight
F32
F32
[4480]
blk.23.ssm_a
F32
F32
[1, 128]
blk.23.ssm_conv1d.bias
F32
F32
[12288]
blk.23.ssm_conv1d.weight
F32
F32
[4, 12288]
blk.23.ssm_d
F32
F32
[1, 128]
blk.23.ssm_dt.bias
F32
F32
[128]
blk.23.ssm_in.weight
Q8_0
Q8_0
[4480, 22656]
blk.23.ssm_norm.weight
F32
F32
[1280, 8]
blk.23.ssm_out.weight
Q6_K
Q6_K
[10240, 4480]
blk.24
blk.24.attn_norm.weight
F32
F32
[4480]
blk.24.ffn_down.weight
Q8_0
Q8_0
[15680, 4480]
blk.24.ffn_up.weight
Q8_0
Q8_0
[4480, 15680]
blk.25
blk.25.attn_norm.weight
F32
F32
[4480]
blk.25.ssm_a
F32
F32
[1, 128]
blk.25.ssm_conv1d.bias
F32
F32
[12288]
blk.25.ssm_conv1d.weight
F32
F32
[4, 12288]
blk.25.ssm_d
F32
F32
[1, 128]
blk.25.ssm_dt.bias
F32
F32
[128]
blk.25.ssm_in.weight
Q8_0
Q8_0
[4480, 22656]
blk.25.ssm_norm.weight
F32
F32
[1280, 8]
blk.25.ssm_out.weight
Q6_K
Q6_K
[10240, 4480]
blk.26
blk.26.attn_norm.weight
F32
F32
[4480]
blk.26.ffn_down.weight
Q8_0
Q8_0
[15680, 4480]
blk.26.ffn_up.weight
Q8_0
Q8_0
[4480, 15680]
blk.27
blk.27.attn_norm.weight
F32
F32
[4480]
blk.27.ssm_a
F32
F32
[1, 128]
blk.27.ssm_conv1d.bias
F32
F32
[12288]
blk.27.ssm_conv1d.weight
F32
F32
[4, 12288]
blk.27.ssm_d
F32
F32
[1, 128]
blk.27.ssm_dt.bias
F32
F32
[128]
blk.27.ssm_in.weight
Q8_0
Q8_0
[4480, 22656]
blk.27.ssm_norm.weight
F32
F32
[1280, 8]
blk.27.ssm_out.weight
Q6_K
Q6_K
[10240, 4480]
blk.28
blk.28.attn_norm.weight
F32
F32
[4480]
blk.28.ffn_down.weight
Q8_0
Q8_0
[15680, 4480]
blk.28.ffn_up.weight
Q8_0
Q8_0
[4480, 15680]
blk.29
blk.29.attn_norm.weight
F32
F32
[4480]
blk.29.ssm_a
F32
F32
[1, 128]
blk.29.ssm_conv1d.bias
F32
F32
[12288]
blk.29.ssm_conv1d.weight
F32
F32
[4, 12288]
blk.29.ssm_d
F32
F32
[1, 128]
blk.29.ssm_dt.bias
F32
F32
[128]
blk.29.ssm_in.weight
Q8_0
Q8_0
[4480, 22656]
blk.29.ssm_norm.weight
F32
F32
[1280, 8]
blk.29.ssm_out.weight
Q6_K
Q6_K
[10240, 4480]
blk.30
blk.30.attn_k.weight
Q8_0
Q8_0
[4480, 1024]
blk.30.attn_norm.weight
F32
F32
[4480]
blk.30.attn_output.weight
Q6_K
Q6_K
[5120, 4480]
blk.30.attn_q.weight
Q8_0
Q8_0
[4480, 5120]
blk.30.attn_v.weight
Q8_0
Q8_0
[4480, 1024]
blk.31
blk.31.attn_norm.weight
F32
F32
[4480]
blk.31.ffn_down.weight
Q8_0
Q8_0
[15680, 4480]
blk.31.ffn_up.weight
Q8_0
Q8_0
[4480, 15680]
blk.32
blk.32.attn_norm.weight
F32
F32
[4480]
blk.32.ssm_a
F32
F32
[1, 128]
blk.32.ssm_conv1d.bias
F32
F32
[12288]
blk.32.ssm_conv1d.weight
F32
F32
[4, 12288]
blk.32.ssm_d
F32
F32
[1, 128]
blk.32.ssm_dt.bias
F32
F32
[128]
blk.32.ssm_in.weight
Q8_0
Q8_0
[4480, 22656]
blk.32.ssm_norm.weight
F32
F32
[1280, 8]
blk.32.ssm_out.weight
Q6_K
Q6_K
[10240, 4480]
blk.33
blk.33.attn_norm.weight
F32
F32
[4480]
blk.33.ffn_down.weight
Q8_0
Q8_0
[15680, 4480]
blk.33.ffn_up.weight
Q8_0
Q8_0
[4480, 15680]
blk.34
blk.34.attn_norm.weight
F32
F32
[4480]
blk.34.ssm_a
F32
F32
[1, 128]
blk.34.ssm_conv1d.bias
F32
F32
[12288]
blk.34.ssm_conv1d.weight
F32
F32
[4, 12288]
blk.34.ssm_d
F32
F32
[1, 128]
blk.34.ssm_dt.bias
F32
F32
[128]
blk.34.ssm_in.weight
Q8_0
Q8_0
[4480, 22656]
blk.34.ssm_norm.weight
F32
F32
[1280, 8]
blk.34.ssm_out.weight
Q6_K
Q6_K
[10240, 4480]
blk.35
blk.35.attn_norm.weight
F32
F32
[4480]
blk.35.ffn_down.weight
Q8_0
Q8_0
[15680, 4480]
blk.35.ffn_up.weight
Q8_0
Q8_0
[4480, 15680]
blk.36
blk.36.attn_norm.weight
F32
F32
[4480]
blk.36.ssm_a
F32
F32
[1, 128]
blk.36.ssm_conv1d.bias
F32
F32
[12288]
blk.36.ssm_conv1d.weight
F32
F32
[4, 12288]
blk.36.ssm_d
F32
F32
[1, 128]
blk.36.ssm_dt.bias
F32
F32
[128]
blk.36.ssm_in.weight
Q8_0
Q8_0
[4480, 22656]
blk.36.ssm_norm.weight
F32
F32
[1280, 8]
blk.36.ssm_out.weight
Q6_K
Q6_K
[10240, 4480]
blk.37
blk.37.attn_norm.weight
F32
F32
[4480]
blk.37.ffn_down.weight
Q8_0
Q8_0
[15680, 4480]
blk.37.ffn_up.weight
Q8_0
Q8_0
[4480, 15680]
blk.38
blk.38.attn_norm.weight
F32
F32
[4480]
blk.38.ssm_a
F32
F32
[1, 128]
blk.38.ssm_conv1d.bias
F32
F32
[12288]
blk.38.ssm_conv1d.weight
F32
F32
[4, 12288]
blk.38.ssm_d
F32
F32
[1, 128]
blk.38.ssm_dt.bias
F32
F32
[128]
blk.38.ssm_in.weight
Q8_0
Q8_0
[4480, 22656]
blk.38.ssm_norm.weight
F32
F32
[1280, 8]
blk.38.ssm_out.weight
Q6_K
Q6_K
[10240, 4480]
blk.39
blk.39.attn_k.weight
Q8_0
Q8_0
[4480, 1024]
blk.39.attn_norm.weight
F32
F32
[4480]
blk.39.attn_output.weight
Q6_K
Q6_K
[5120, 4480]
blk.39.attn_q.weight
Q8_0
Q8_0
[4480, 5120]
blk.39.attn_v.weight
Q8_0
Q8_0
[4480, 1024]
blk.40
blk.40.attn_norm.weight
F32
F32
[4480]
blk.40.ffn_down.weight
Q8_0
Q8_0
[15680, 4480]
blk.40.ffn_up.weight
Q8_0
Q8_0
[4480, 15680]
blk.41
blk.41.attn_norm.weight
F32
F32
[4480]
blk.41.ssm_a
F32
F32
[1, 128]
blk.41.ssm_conv1d.bias
F32
F32
[12288]
blk.41.ssm_conv1d.weight
F32
F32
[4, 12288]
blk.41.ssm_d
F32
F32
[1, 128]
blk.41.ssm_dt.bias
F32
F32
[128]
blk.41.ssm_in.weight
Q8_0
Q8_0
[4480, 22656]
blk.41.ssm_norm.weight
F32
F32
[1280, 8]
blk.41.ssm_out.weight
Q6_K
Q6_K
[10240, 4480]
blk.42
blk.42.attn_norm.weight
F32
F32
[4480]
blk.42.ffn_down.weight
Q8_0
Q8_0
[15680, 4480]
blk.42.ffn_up.weight
Q8_0
Q8_0
[4480, 15680]
blk.43
blk.43.attn_norm.weight
F32
F32
[4480]
blk.43.ssm_a
F32
F32
[1, 128]
blk.43.ssm_conv1d.bias
F32
F32
[12288]
blk.43.ssm_conv1d.weight
F32
F32
[4, 12288]
blk.43.ssm_d
F32
F32
[1, 128]
blk.43.ssm_dt.bias
F32
F32
[128]
blk.43.ssm_in.weight
Q8_0
Q8_0
[4480, 22656]
blk.43.ssm_norm.weight
F32
F32
[1280, 8]
blk.43.ssm_out.weight
Q6_K
Q6_K
[10240, 4480]
blk.44
blk.44.attn_norm.weight
F32
F32
[4480]
blk.44.ssm_a
F32
F32
[1, 128]
blk.44.ssm_conv1d.bias
F32
F32
[12288]
blk.44.ssm_conv1d.weight
F32
F32
[4, 12288]
blk.44.ssm_d
F32
F32
[1, 128]
blk.44.ssm_dt.bias
F32
F32
[128]
blk.44.ssm_in.weight
Q8_0
Q8_0
[4480, 22656]
blk.44.ssm_norm.weight
F32
F32
[1280, 8]
blk.44.ssm_out.weight
Q6_K
Q6_K
[10240, 4480]
blk.45
blk.45.attn_norm.weight
F32
F32
[4480]
blk.45.ffn_down.weight
Q8_0
Q8_0
[15680, 4480]
blk.45.ffn_up.weight
Q8_0
Q8_0
[4480, 15680]
blk.46
blk.46.attn_norm.weight
F32
F32
[4480]
blk.46.ssm_a
F32
F32
[1, 128]
blk.46.ssm_conv1d.bias
F32
F32
[12288]
blk.46.ssm_conv1d.weight
F32
F32
[4, 12288]
blk.46.ssm_d
F32
F32
[1, 128]
blk.46.ssm_dt.bias
F32
F32
[128]
blk.46.ssm_in.weight
Q8_0
Q8_0
[4480, 22656]
blk.46.ssm_norm.weight
F32
F32
[1280, 8]
blk.46.ssm_out.weight
Q6_K
Q6_K
[10240, 4480]
blk.47
blk.47.attn_norm.weight
F32
F32
[4480]
blk.47.ffn_down.weight
Q8_0
Q8_0
[15680, 4480]
blk.47.ffn_up.weight
Q8_0
Q8_0
[4480, 15680]
blk.48
blk.48.attn_norm.weight
F32
F32
[4480]
blk.48.ssm_a
F32
F32
[1, 128]
blk.48.ssm_conv1d.bias
F32
F32
[12288]
blk.48.ssm_conv1d.weight
F32
F32
[4, 12288]
blk.48.ssm_d
F32
F32
[1, 128]
blk.48.ssm_dt.bias
F32
F32
[128]
blk.48.ssm_in.weight
Q8_0
Q8_0
[4480, 22656]
blk.48.ssm_norm.weight
F32
F32
[1280, 8]
blk.48.ssm_out.weight
Q6_K
Q6_K
[10240, 4480]
blk.49
blk.49.attn_norm.weight
F32
F32
[4480]
blk.49.ffn_down.weight
Q8_0
Q8_0
[15680, 4480]
blk.49.ffn_up.weight
Q8_0
Q8_0
[4480, 15680]
blk.50
blk.50.attn_norm.weight
F32
F32
[4480]
blk.50.ssm_a
F32
F32
[1, 128]
blk.50.ssm_conv1d.bias
F32
F32
[12288]
blk.50.ssm_conv1d.weight
F32
F32
[4, 12288]
blk.50.ssm_d
F32
F32
[1, 128]
blk.50.ssm_dt.bias
F32
F32
[128]
blk.50.ssm_in.weight
Q8_0
Q8_0
[4480, 22656]
blk.50.ssm_norm.weight
F32
F32
[1280, 8]
blk.50.ssm_out.weight
Q6_K
Q6_K
[10240, 4480]
blk.51
blk.51.attn_norm.weight
F32
F32
[4480]
blk.51.ffn_down.weight
Q8_0
Q8_0
[15680, 4480]
blk.51.ffn_up.weight
Q8_0
Q8_0
[4480, 15680]
blk.52
blk.52.attn_norm.weight
F32
F32
[4480]
blk.52.ssm_a
F32
F32
[1, 128]
blk.52.ssm_conv1d.bias
F32
F32
[12288]
blk.52.ssm_conv1d.weight
F32
F32
[4, 12288]
blk.52.ssm_d
F32
F32
[1, 128]
blk.52.ssm_dt.bias
F32
F32
[128]
blk.52.ssm_in.weight
Q8_0
Q8_0
[4480, 22656]
blk.52.ssm_norm.weight
F32
F32
[1280, 8]
blk.52.ssm_out.weight
Q6_K
Q6_K
[10240, 4480]
blk.53
blk.53.attn_norm.weight
F32
F32
[4480]
blk.53.ffn_down.weight
Q8_0
Q8_0
[15680, 4480]
blk.53.ffn_up.weight
Q8_0
Q8_0
[4480, 15680]
blk.54
blk.54.attn_norm.weight
F32
F32
[4480]
blk.54.ssm_a
F32
F32
[1, 128]
blk.54.ssm_conv1d.bias
F32
F32
[12288]
blk.54.ssm_conv1d.weight
F32
F32
[4, 12288]
blk.54.ssm_d
F32
F32
[1, 128]
blk.54.ssm_dt.bias
F32
F32
[128]
blk.54.ssm_in.weight
Q8_0
Q8_0
[4480, 22656]
blk.54.ssm_norm.weight
F32
F32
[1280, 8]
blk.54.ssm_out.weight
Q6_K
Q6_K
[10240, 4480]
blk.55
blk.55.attn_norm.weight
F32
F32
[4480]
blk.55.ffn_down.weight
Q8_0
Q8_0
[15680, 4480]
blk.55.ffn_up.weight
Q8_0
Q8_0
[4480, 15680]
output.weight
Q8_0
Q8_0
[4480, 131072]
output_norm.weight
F32
F32
[4480]