treyleo16/ small-minded:1b

183 4 weeks ago

Like a little baby with a small brain.

0.0001m 0.00077m 0.001m 0.26m 1m 500m 1b
... /
model
85a30206389d ยท 645MB
    Metadata
  • general.architecture
    llama
  • general.file_type
    Q4_K_M
  • llama.attention.head_count
    128
  • llama.attention.head_count_kv
    128
  • llama.attention.layer_norm_rms_epsilon
    1e-05
  • llama.block_count
    4
  • llama.context_length
    512
  • llama.embedding_length
    1792
  • llama.feed_forward_length
    43520
  • llama.rope.dimension_count
    14
  • llama.rope.freq_base
    10000
  • llama.vocab_size
    32000
  • tokenizer.ggml.add_bos_token
    true
  • tokenizer.ggml.add_eos_token
    false
  • tokenizer.ggml.bos_token_id
    1
  • tokenizer.ggml.eos_token_id
    2
  • tokenizer.ggml.model
    llama
  • tokenizer.ggml.padding_token_id
    0
  • tokenizer.ggml.scores
    [0, 0, 0, 0, 0, ...]
  • tokenizer.ggml.token_type
    [2, 3, 3, 6, 6, ...]
  • tokenizer.ggml.tokens
    [<unk>, <s>, </s>, <0x00>, <0x01>, ...]
  • tokenizer.ggml.unknown_token_id
    0
    • Tensor
  • token_embd.weight
    Q6_K
    [1792, 32000]
    • blk.0
  • blk.0.attn_k.weight
    Q4_K
    [1792, 1792]
  • blk.0.attn_norm.weight
    F32
    [1792]
  • blk.0.attn_output.weight
    Q4_K
    [1792, 1792]
  • blk.0.attn_q.weight
    Q4_K
    [1792, 1792]
  • blk.0.attn_v.weight
    Q4_K
    [1792, 1792]
  • blk.0.ffn_down.weight
    Q4_K
    [43520, 1792]
  • blk.0.ffn_gate.weight
    Q4_K
    [1792, 43520]
  • blk.0.ffn_norm.weight
    F32
    [1792]
  • blk.0.ffn_up.weight
    Q4_K
    [1792, 43520]
    • blk.1
  • blk.1.attn_k.weight
    Q4_K
    [1792, 1792]
  • blk.1.attn_norm.weight
    F32
    [1792]
  • blk.1.attn_output.weight
    Q4_K
    [1792, 1792]
  • blk.1.attn_q.weight
    Q4_K
    [1792, 1792]
  • blk.1.attn_v.weight
    Q4_K
    [1792, 1792]
  • blk.1.ffn_down.weight
    Q4_K
    [43520, 1792]
  • blk.1.ffn_gate.weight
    Q4_K
    [1792, 43520]
  • blk.1.ffn_norm.weight
    F32
    [1792]
  • blk.1.ffn_up.weight
    Q4_K
    [1792, 43520]
    • blk.2
  • blk.2.attn_k.weight
    Q4_K
    [1792, 1792]
  • blk.2.attn_norm.weight
    F32
    [1792]
  • blk.2.attn_output.weight
    Q4_K
    [1792, 1792]
  • blk.2.attn_q.weight
    Q4_K
    [1792, 1792]
  • blk.2.attn_v.weight
    Q6_K
    [1792, 1792]
  • blk.2.ffn_down.weight
    Q6_K
    [43520, 1792]
  • blk.2.ffn_gate.weight
    Q4_K
    [1792, 43520]
  • blk.2.ffn_norm.weight
    F32
    [1792]
  • blk.2.ffn_up.weight
    Q4_K
    [1792, 43520]
    • blk.3
  • blk.3.attn_k.weight
    Q4_K
    [1792, 1792]
  • blk.3.attn_norm.weight
    F32
    [1792]
  • blk.3.attn_output.weight
    Q4_K
    [1792, 1792]
  • blk.3.attn_q.weight
    Q4_K
    [1792, 1792]
  • blk.3.attn_v.weight
    Q6_K
    [1792, 1792]
  • blk.3.ffn_down.weight
    Q6_K
    [43520, 1792]
  • blk.3.ffn_gate.weight
    Q4_K
    [1792, 43520]
  • blk.3.ffn_norm.weight
    F32
    [1792]
  • blk.3.ffn_up.weight
    Q4_K
    [1792, 43520]
  • output_norm.weight
    F32
    [1792]