bvenkatesulu909/venki_llm_1/model

bvenkatesulu909/ venki_llm_1:latest

2 Downloads Updated yesterday

Field Real value Size 13 MB (2.57M params, F32 GGUF) Context 128 tokens (architecture limit) Input Text Parity 21/21 ✅

venki_llm_1:latest ... /

model

e9fb59aef60f · 13MB

Metadata

general.architecture

gpt2

gpt2
general.file_type

F32

F32
gpt2.attention.causal

true

true
gpt2.attention.head_count

4

4
gpt2.attention.head_count_kv

4

4
gpt2.attention.layer_norm_epsilon

1e-05

1e-05
gpt2.block_count

4

4
gpt2.context_length

128

128
gpt2.embedding_length

192

192
gpt2.feed_forward_length

768

768
tokenizer.ggml.merges

[Ġ Ġ, ĠĠ ĠĠ, Ġ t, h e, Ġ a, ...]

[Ġ Ġ, ĠĠ ĠĠ, Ġ t, h e, Ġ a, ...]
tokenizer.ggml.model

gpt2

gpt2
tokenizer.ggml.scores

[0, 0, 0, 0, 0, ...]

[0, 0, 0, 0, 0, ...]
tokenizer.ggml.tokens

[!, ", #, $, %, ...]

[!, ", #, $, %, ...]

Tensor

Name

Type

Shape
token_embd.weight

F32

F32

[192, 4000]

blk.0

blk.0.attn_norm.bias

F32

F32

[192]
blk.0.attn_norm.weight

F32

F32

[192]
blk.0.attn_output.bias

F32

F32

[192]
blk.0.attn_output.weight

F32

F32

[192, 192]
blk.0.attn_qkv.bias

F32

F32

[576]
blk.0.attn_qkv.weight

F32

F32

[192, 576]
blk.0.ffn_down.bias

F32

F32

[192]
blk.0.ffn_down.weight

F32

F32

[768, 192]
blk.0.ffn_norm.bias

F32

F32

[192]
blk.0.ffn_norm.weight

F32

F32

[192]
blk.0.ffn_up.bias

F32

F32

[768]
blk.0.ffn_up.weight

F32

F32

[192, 768]

blk.1

blk.1.attn_norm.bias

F32

F32

[192]
blk.1.attn_norm.weight

F32

F32

[192]
blk.1.attn_output.bias

F32

F32

[192]
blk.1.attn_output.weight

F32

F32

[192, 192]
blk.1.attn_qkv.bias

F32

F32

[576]
blk.1.attn_qkv.weight

F32

F32

[192, 576]
blk.1.ffn_down.bias

F32

F32

[192]
blk.1.ffn_down.weight

F32

F32

[768, 192]
blk.1.ffn_norm.bias

F32

F32

[192]
blk.1.ffn_norm.weight

F32

F32

[192]
blk.1.ffn_up.bias

F32

F32

[768]
blk.1.ffn_up.weight

F32

F32

[192, 768]

blk.2

blk.2.attn_norm.bias

F32

F32

[192]
blk.2.attn_norm.weight

F32

F32

[192]
blk.2.attn_output.bias

F32

F32

[192]
blk.2.attn_output.weight

F32

F32

[192, 192]
blk.2.attn_qkv.bias

F32

F32

[576]
blk.2.attn_qkv.weight

F32

F32

[192, 576]
blk.2.ffn_down.bias

F32

F32

[192]
blk.2.ffn_down.weight

F32

F32

[768, 192]
blk.2.ffn_norm.bias

F32

F32

[192]
blk.2.ffn_norm.weight

F32

F32

[192]
blk.2.ffn_up.bias

F32

F32

[768]
blk.2.ffn_up.weight

F32

F32

[192, 768]

blk.3

blk.3.attn_norm.bias

F32

F32

[192]
blk.3.attn_norm.weight

F32

F32

[192]
blk.3.attn_output.bias

F32

F32

[192]
blk.3.attn_output.weight

F32

F32

[192, 192]
blk.3.attn_qkv.bias

F32

F32

[576]
blk.3.attn_qkv.weight

F32

F32

[192, 576]
blk.3.ffn_down.bias

F32

F32

[192]
blk.3.ffn_down.weight

F32

F32

[768, 192]
blk.3.ffn_norm.bias

F32

F32

[192]
blk.3.ffn_norm.weight

F32

F32

[192]
blk.3.ffn_up.bias

F32

F32

[768]
blk.3.ffn_up.weight

F32

F32

[192, 768]

output.weight

F32

F32

[192, 4000]
output_norm.bias

F32

F32

[192]
position_embd.weight

F32

F32

[192, 128]
output_norm.weight

F32

F32

[192]