https://huggingface.co/DMetaSoul/Dmeta-embedding-zh-small
embedding
1,561 Pulls Updated 12 months ago
7a2d429b1755 · 295MB
-
general.architecturebertbert
-
general.file_typeF32F32
-
bert.attention.causalfalsefalse
-
bert.attention.head_count1212
-
bert.attention.layer_norm_epsilon1e-121e-12
-
bert.block_count88
-
bert.context_length10241024
-
bert.embedding_length768768
-
bert.feed_forward_length30723072
-
bert.pooling_typeCLSCLS
-
tokenizer.ggml.bos_token_id101101
-
tokenizer.ggml.bos_token_id00
-
tokenizer.ggml.cls_token_id101101
-
tokenizer.ggml.eos_token_id102102
-
tokenizer.ggml.eos_token_id22
-
tokenizer.ggml.mask_token_id103103
-
tokenizer.ggml.modelbertbert
-
tokenizer.ggml.padding_token_id00
-
tokenizer.ggml.scores[-1000, -1000, -1000, -1000, -1000, ...][-1000, -1000, -1000, -1000, -1000, ...]
-
tokenizer.ggml.seperator_token_id102102
-
tokenizer.ggml.token_type[3, 1, 1, 1, 1, ...][3, 1, 1, 1, 1, ...]
-
tokenizer.ggml.token_type_count22
-
tokenizer.ggml.tokens[[PAD], [unused1], [unused2], [unused3], [unused4], ...][[PAD], [unused1], [unused2], [unused3], [unused4], ...]
-
tokenizer.ggml.unknown_token_id100100
-
token_embd.weightF32[768, 21128]
-
blk.0.attn_k.biasF32[768]
-
blk.0.attn_k.weightF32[768, 768]
-
blk.0.attn_output.biasF32[768]
-
blk.0.attn_output.weightF32[768, 768]
-
blk.0.attn_output_norm.biasF32[768]
-
blk.0.attn_output_norm.weightF32[768]
-
blk.0.attn_q.biasF32[768]
-
blk.0.attn_q.weightF32[768, 768]
-
blk.0.attn_v.biasF32[768]
-
blk.0.attn_v.weightF32[768, 768]
-
blk.0.ffn_down.biasF32[768]
-
blk.0.ffn_down.weightF32[3072, 768]
-
blk.0.ffn_up.biasF32[3072]
-
blk.0.ffn_up.weightF32[768, 3072]
-
blk.0.layer_output_norm.biasF32[768]
-
blk.0.layer_output_norm.weightF32[768]
-
blk.1.attn_k.biasF32[768]
-
blk.1.attn_k.weightF32[768, 768]
-
blk.1.attn_output.biasF32[768]
-
blk.1.attn_output.weightF32[768, 768]
-
blk.1.attn_output_norm.biasF32[768]
-
blk.1.attn_output_norm.weightF32[768]
-
blk.1.attn_q.biasF32[768]
-
blk.1.attn_q.weightF32[768, 768]
-
blk.1.attn_v.biasF32[768]
-
blk.1.attn_v.weightF32[768, 768]
-
blk.1.ffn_down.biasF32[768]
-
blk.1.ffn_down.weightF32[3072, 768]
-
blk.1.ffn_up.biasF32[3072]
-
blk.1.ffn_up.weightF32[768, 3072]
-
blk.1.layer_output_norm.biasF32[768]
-
blk.1.layer_output_norm.weightF32[768]
-
blk.2.attn_k.biasF32[768]
-
blk.2.attn_k.weightF32[768, 768]
-
blk.2.attn_output.biasF32[768]
-
blk.2.attn_output.weightF32[768, 768]
-
blk.2.attn_output_norm.biasF32[768]
-
blk.2.attn_output_norm.weightF32[768]
-
blk.2.attn_q.biasF32[768]
-
blk.2.attn_q.weightF32[768, 768]
-
blk.2.attn_v.biasF32[768]
-
blk.2.attn_v.weightF32[768, 768]
-
blk.2.ffn_down.biasF32[768]
-
blk.2.ffn_down.weightF32[3072, 768]
-
blk.2.ffn_up.biasF32[3072]
-
blk.2.ffn_up.weightF32[768, 3072]
-
blk.2.layer_output_norm.biasF32[768]
-
blk.2.layer_output_norm.weightF32[768]
-
blk.3.attn_k.biasF32[768]
-
blk.3.attn_k.weightF32[768, 768]
-
blk.3.attn_output.biasF32[768]
-
blk.3.attn_output.weightF32[768, 768]
-
blk.3.attn_output_norm.biasF32[768]
-
blk.3.attn_output_norm.weightF32[768]
-
blk.3.attn_q.biasF32[768]
-
blk.3.attn_q.weightF32[768, 768]
-
blk.3.attn_v.biasF32[768]
-
blk.3.attn_v.weightF32[768, 768]
-
blk.3.ffn_down.biasF32[768]
-
blk.3.ffn_down.weightF32[3072, 768]
-
blk.3.ffn_up.biasF32[3072]
-
blk.3.ffn_up.weightF32[768, 3072]
-
blk.3.layer_output_norm.biasF32[768]
-
blk.3.layer_output_norm.weightF32[768]
-
blk.4.attn_k.biasF32[768]
-
blk.4.attn_k.weightF32[768, 768]
-
blk.4.attn_output.biasF32[768]
-
blk.4.attn_output.weightF32[768, 768]
-
blk.4.attn_output_norm.biasF32[768]
-
blk.4.attn_output_norm.weightF32[768]
-
blk.4.attn_q.biasF32[768]
-
blk.4.attn_q.weightF32[768, 768]
-
blk.4.attn_v.biasF32[768]
-
blk.4.attn_v.weightF32[768, 768]
-
blk.4.ffn_down.biasF32[768]
-
blk.4.ffn_down.weightF32[3072, 768]
-
blk.4.ffn_up.biasF32[3072]
-
blk.4.ffn_up.weightF32[768, 3072]
-
blk.4.layer_output_norm.biasF32[768]
-
blk.4.layer_output_norm.weightF32[768]
-
blk.5.attn_k.biasF32[768]
-
blk.5.attn_k.weightF32[768, 768]
-
blk.5.attn_output.biasF32[768]
-
blk.5.attn_output.weightF32[768, 768]
-
blk.5.attn_output_norm.biasF32[768]
-
blk.5.attn_output_norm.weightF32[768]
-
blk.5.attn_q.biasF32[768]
-
blk.5.attn_q.weightF32[768, 768]
-
blk.5.attn_v.biasF32[768]
-
blk.5.attn_v.weightF32[768, 768]
-
blk.5.ffn_down.biasF32[768]
-
blk.5.ffn_down.weightF32[3072, 768]
-
blk.5.ffn_up.biasF32[3072]
-
blk.5.ffn_up.weightF32[768, 3072]
-
blk.5.layer_output_norm.biasF32[768]
-
blk.5.layer_output_norm.weightF32[768]
-
blk.6.attn_k.biasF32[768]
-
blk.6.attn_k.weightF32[768, 768]
-
blk.6.attn_output.biasF32[768]
-
blk.6.attn_output.weightF32[768, 768]
-
blk.6.attn_output_norm.biasF32[768]
-
blk.6.attn_output_norm.weightF32[768]
-
blk.6.attn_q.biasF32[768]
-
blk.6.attn_q.weightF32[768, 768]
-
blk.6.attn_v.biasF32[768]
-
blk.6.attn_v.weightF32[768, 768]
-
blk.6.ffn_down.biasF32[768]
-
blk.6.ffn_down.weightF32[3072, 768]
-
blk.6.ffn_up.biasF32[3072]
-
blk.6.ffn_up.weightF32[768, 3072]
-
blk.6.layer_output_norm.biasF32[768]
-
blk.6.layer_output_norm.weightF32[768]
-
blk.7.attn_k.biasF32[768]
-
blk.7.attn_k.weightF32[768, 768]
-
blk.7.attn_output.biasF32[768]
-
blk.7.attn_output.weightF32[768, 768]
-
blk.7.attn_output_norm.biasF32[768]
-
blk.7.attn_output_norm.weightF32[768]
-
blk.7.attn_q.biasF32[768]
-
blk.7.attn_q.weightF32[768, 768]
-
blk.7.attn_v.biasF32[768]
-
blk.7.attn_v.weightF32[768, 768]
-
blk.7.ffn_down.biasF32[768]
-
blk.7.ffn_down.weightF32[3072, 768]
-
blk.7.ffn_up.biasF32[3072]
-
blk.7.ffn_up.weightF32[768, 3072]
-
blk.7.layer_output_norm.biasF32[768]
-
blk.7.layer_output_norm.weightF32[768]
-
position_embd.weightF32[768, 1024]
-
token_embd_norm.biasF32[768]
-
token_embd_norm.weightF32[768]
-
token_types.weightF32[768, 2]
Metadata
Tensor
blk.0
blk.1
blk.2
blk.3
blk.4
blk.5
blk.6
blk.7