diff --git "a/scores/Qwen3-30B-A3B-pruned-IQ3_S.md" "b/scores/Qwen3-30B-A3B-pruned-IQ3_S.md" new file mode 100644--- /dev/null +++ "b/scores/Qwen3-30B-A3B-pruned-IQ3_S.md" @@ -0,0 +1,1653 @@ +# Qwen3-30B-A3B-IQ3_S.gguf - GGUF Internal File Dump + +- Endian: LITTLE endian + +## Key Value Metadata Store + +There are 45 key-value pairs in this file + +| POS | TYPE | Count | Key | Value | +|----:|:---------|-------:|:------------------------------------------|:--------------------------------------------------------------------| +| 1 | UINT32 | 1 | GGUF.version | 3 | +| 2 | UINT64 | 1 | GGUF.tensor_count | 555 | +| 3 | UINT64 | 1 | GGUF.kv_count | 42 | +| 4 | STRING | 1 | general.architecture | `qwen3moe` | +| 5 | STRING | 1 | general.type | `model` | +| 6 | STRING | 1 | general.name | `Qwen3 30B A3B` | +| 7 | STRING | 1 | general.basename | `Qwen3` | +| 8 | STRING | 1 | general.size_label | `30B-A3B` | +| 9 | STRING | 1 | general.license | `apache-2.0` | +| 10 | STRING | 1 | general.license.link | `https://huggingface.co/Qwen/Qwen3-30B-A3B/blob/main/LICENSE` | +| 11 | UINT32 | 1 | general.base_model.count | 1 | +| 12 | STRING | 1 | general.base_model.0.name | `Qwen3 30B A3B Base` | +| 13 | STRING | 1 | general.base_model.0.organization | `Qwen` | +| 14 | STRING | 1 | general.base_model.0.repo_url | `https://huggingface.co/Qwen/Qwen3-30B-A3B-Base` | +| 15 | [STRING] | 1 | general.tags | [ `text-generation` ] | +| 16 | UINT32 | 1 | qwen3moe.context_length | 40960 | +| 17 | UINT32 | 1 | qwen3moe.embedding_length | 2048 | +| 18 | UINT32 | 1 | qwen3moe.feed_forward_length | 6144 | +| 19 | UINT32 | 1 | qwen3moe.attention.head_count | 32 | +| 20 | UINT32 | 1 | qwen3moe.attention.head_count_kv | 4 | +| 21 | FLOAT32 | 1 | qwen3moe.rope.freq_base | 1000000.0 | +| 22 | FLOAT32 | 1 | qwen3moe.attention.layer_norm_rms_epsilon | 1e-06 | +| 23 | UINT32 | 1 | qwen3moe.expert_used_count | 8 | +| 24 | UINT32 | 1 | qwen3moe.attention.key_length | 128 | +| 25 | UINT32 | 1 | qwen3moe.attention.value_length | 128 | +| 26 | UINT32 | 1 | qwen3moe.expert_count | 128 | +| 27 | UINT32 | 1 | qwen3moe.expert_feed_forward_length | 768 | +| 28 | STRING | 1 | tokenizer.ggml.model | `gpt2` | +| 29 | STRING | 1 | tokenizer.ggml.pre | `qwen2` | +| 30 | [STRING] | 151936 | tokenizer.ggml.tokens | [ `!`, `"`, `#`, `$`, `%`, ... ] | +| 31 | [INT32] | 151936 | tokenizer.ggml.token_type | [ 1, 1, 1, 1, 1, 1, 1, ... ] | +| 32 | [STRING] | 151387 | tokenizer.ggml.merges | [ `Ġ Ġ`, `ĠĠ ĠĠ`, `i n`, `Ġ t`, `ĠĠĠĠ ĠĠĠĠ`, ... ] | +| 33 | UINT32 | 1 | tokenizer.ggml.eos_token_id | 151645 | +| 34 | UINT32 | 1 | tokenizer.ggml.padding_token_id | 151643 | +| 35 | UINT32 | 1 | tokenizer.ggml.bos_token_id | 151643 | +| 36 | BOOL | 1 | tokenizer.ggml.add_bos_token | False | +| 37 | STRING | 1 | tokenizer.chat_template | `{%- if tools %}{{- '<|im_`...`{%- endif %}{%- endif %}` | +| 38 | UINT32 | 1 | general.quantization_version | 2 | +| 39 | UINT32 | 1 | general.file_type | 26 | +| 40 | BOOL | 1 | general.pruned | True | +| 41 | UINT32 | 1 | qwen3moe.block_count | 46 | +| 42 | STRING | 1 | quantize.imatrix.file | `./imatrix/imatrix-Qwen3-30B-A3B-medium.dat` | +| 43 | STRING | 1 | quantize.imatrix.dataset | `../../datasets/imatrix/combined_all_medium.txt` | +| 44 | INT32 | 1 | quantize.imatrix.entries_count | 385 | +| 45 | INT32 | 1 | quantize.imatrix.chunks_count | 6946 | + +## Tensors Overview ~29B Elements + +Total number of elements in all tensors: 29285881344 Elements + +- [Qwen3-30B-A3B-IQ3\_S.gguf - GGUF Internal File Dump](#qwen3-30b-a3b-iq3_sgguf---gguf-internal-file-dump) + - [Key Value Metadata Store](#key-value-metadata-store) + - [Tensors Overview ~29B Elements](#tensors-overview-29b-elements) + - [Tensor Data Offset](#tensor-data-offset) + - [Base Tensor Group : ~622M Elements](#base-tensor-group--622m-elements) + - [Block 0 Tensor Group : ~623M Elements](#block-0-tensor-group--623m-elements) + - [Block 1 Tensor Group : ~623M Elements](#block-1-tensor-group--623m-elements) + - [Block 2 Tensor Group : ~623M Elements](#block-2-tensor-group--623m-elements) + - [Block 3 Tensor Group : ~623M Elements](#block-3-tensor-group--623m-elements) + - [Block 4 Tensor Group : ~623M Elements](#block-4-tensor-group--623m-elements) + - [Block 5 Tensor Group : ~623M Elements](#block-5-tensor-group--623m-elements) + - [Block 6 Tensor Group : ~623M Elements](#block-6-tensor-group--623m-elements) + - [Block 7 Tensor Group : ~623M Elements](#block-7-tensor-group--623m-elements) + - [Block 8 Tensor Group : ~623M Elements](#block-8-tensor-group--623m-elements) + - [Block 9 Tensor Group : ~623M Elements](#block-9-tensor-group--623m-elements) + - [Block 10 Tensor Group : ~623M Elements](#block-10-tensor-group--623m-elements) + - [Block 11 Tensor Group : ~623M Elements](#block-11-tensor-group--623m-elements) + - [Block 12 Tensor Group : ~623M Elements](#block-12-tensor-group--623m-elements) + - [Block 13 Tensor Group : ~623M Elements](#block-13-tensor-group--623m-elements) + - [Block 14 Tensor Group : ~623M Elements](#block-14-tensor-group--623m-elements) + - [Block 15 Tensor Group : ~623M Elements](#block-15-tensor-group--623m-elements) + - [Block 16 Tensor Group : ~623M Elements](#block-16-tensor-group--623m-elements) + - [Block 17 Tensor Group : ~623M Elements](#block-17-tensor-group--623m-elements) + - [Block 18 Tensor Group : ~623M Elements](#block-18-tensor-group--623m-elements) + - [Block 19 Tensor Group : ~623M Elements](#block-19-tensor-group--623m-elements) + - [Block 20 Tensor Group : ~623M Elements](#block-20-tensor-group--623m-elements) + - [Block 21 Tensor Group : ~623M Elements](#block-21-tensor-group--623m-elements) + - [Block 22 Tensor Group : ~623M Elements](#block-22-tensor-group--623m-elements) + - [Block 23 Tensor Group : ~623M Elements](#block-23-tensor-group--623m-elements) + - [Block 24 Tensor Group : ~623M Elements](#block-24-tensor-group--623m-elements) + - [Block 25 Tensor Group : ~623M Elements](#block-25-tensor-group--623m-elements) + - [Block 26 Tensor Group : ~623M Elements](#block-26-tensor-group--623m-elements) + - [Block 27 Tensor Group : ~623M Elements](#block-27-tensor-group--623m-elements) + - [Block 28 Tensor Group : ~623M Elements](#block-28-tensor-group--623m-elements) + - [Block 29 Tensor Group : ~623M Elements](#block-29-tensor-group--623m-elements) + - [Block 30 Tensor Group : ~623M Elements](#block-30-tensor-group--623m-elements) + - [Block 31 Tensor Group : ~623M Elements](#block-31-tensor-group--623m-elements) + - [Block 32 Tensor Group : ~623M Elements](#block-32-tensor-group--623m-elements) + - [Block 33 Tensor Group : ~623M Elements](#block-33-tensor-group--623m-elements) + - [Block 34 Tensor Group : ~623M Elements](#block-34-tensor-group--623m-elements) + - [Block 35 Tensor Group : ~623M Elements](#block-35-tensor-group--623m-elements) + - [Block 36 Tensor Group : ~623M Elements](#block-36-tensor-group--623m-elements) + - [Block 37 Tensor Group : ~623M Elements](#block-37-tensor-group--623m-elements) + - [Block 38 Tensor Group : ~623M Elements](#block-38-tensor-group--623m-elements) + - [Block 39 Tensor Group : ~623M Elements](#block-39-tensor-group--623m-elements) + - [Block 40 Tensor Group : ~623M Elements](#block-40-tensor-group--623m-elements) + - [Block 41 Tensor Group : ~623M Elements](#block-41-tensor-group--623m-elements) + - [Block 42 Tensor Group : ~623M Elements](#block-42-tensor-group--623m-elements) + - [Block 43 Tensor Group : ~623M Elements](#block-43-tensor-group--623m-elements) + - [Block 44 Tensor Group : ~623M Elements](#block-44-tensor-group--623m-elements) + - [Block 45 Tensor Group : ~623M Elements](#block-45-tensor-group--623m-elements) + +### Tensor Data Offset + +This table contains the offset and data segment relative to start of file + +| T_ID | Tensor Layer Name | Data Offset (B) | Data Size (B) | +|-----:|:----------------------------|-----------------:|-----------------:| +| 0 | output.weight | 0x5b12e0 | 0x7199800 | +| 1 | output_norm.weight | 0x774aae0 | 0x2000 | +| 2 | token_embd.weight | 0x774cae0 | 0x7199800 | +| 3 | blk.0.attn_k.weight | 0xe8e62e0 | 0x62000 | +| 4 | blk.0.attn_k_norm.weight | 0xe9482e0 | 0x200 | +| 5 | blk.0.attn_norm.weight | 0xe9484e0 | 0x2000 | +| 6 | blk.0.attn_output.weight | 0xe94a4e0 | 0x370000 | +| 7 | blk.0.attn_q.weight | 0xecba4e0 | 0x310000 | +| 8 | blk.0.attn_q_norm.weight | 0xefca4e0 | 0x200 | +| 9 | blk.0.attn_v.weight | 0xefca6e0 | 0x6e000 | +| 10 | blk.0.ffn_down_exps.weight | 0xf0386e0 | 0x6c00000 | +| 11 | blk.0.ffn_gate_exps.weight | 0x15c386e0 | 0x4980000 | +| 12 | blk.0.ffn_gate_inp.weight | 0x1a5b86e0 | 0x100000 | +| 13 | blk.0.ffn_norm.weight | 0x1a6b86e0 | 0x2000 | +| 14 | blk.0.ffn_up_exps.weight | 0x1a6ba6e0 | 0x4980000 | +| 15 | blk.1.attn_k.weight | 0x1f03a6e0 | 0x62000 | +| 16 | blk.1.attn_k_norm.weight | 0x1f09c6e0 | 0x200 | +| 17 | blk.1.attn_norm.weight | 0x1f09c8e0 | 0x2000 | +| 18 | blk.1.attn_output.weight | 0x1f09e8e0 | 0x370000 | +| 19 | blk.1.attn_q.weight | 0x1f40e8e0 | 0x310000 | +| 20 | blk.1.attn_q_norm.weight | 0x1f71e8e0 | 0x200 | +| 21 | blk.1.attn_v.weight | 0x1f71eae0 | 0x6e000 | +| 22 | blk.1.ffn_down_exps.weight | 0x1f78cae0 | 0x6c00000 | +| 23 | blk.1.ffn_gate_exps.weight | 0x2638cae0 | 0x4980000 | +| 24 | blk.1.ffn_gate_inp.weight | 0x2ad0cae0 | 0x100000 | +| 25 | blk.1.ffn_norm.weight | 0x2ae0cae0 | 0x2000 | +| 26 | blk.1.ffn_up_exps.weight | 0x2ae0eae0 | 0x4980000 | +| 27 | blk.2.attn_k.weight | 0x2f78eae0 | 0x62000 | +| 28 | blk.2.attn_k_norm.weight | 0x2f7f0ae0 | 0x200 | +| 29 | blk.2.attn_norm.weight | 0x2f7f0ce0 | 0x2000 | +| 30 | blk.2.attn_output.weight | 0x2f7f2ce0 | 0x370000 | +| 31 | blk.2.attn_q.weight | 0x2fb62ce0 | 0x310000 | +| 32 | blk.2.attn_q_norm.weight | 0x2fe72ce0 | 0x200 | +| 33 | blk.2.attn_v.weight | 0x2fe72ee0 | 0x6e000 | +| 34 | blk.2.ffn_down_exps.weight | 0x2fee0ee0 | 0x6c00000 | +| 35 | blk.2.ffn_gate_exps.weight | 0x36ae0ee0 | 0x4980000 | +| 36 | blk.2.ffn_gate_inp.weight | 0x3b460ee0 | 0x100000 | +| 37 | blk.2.ffn_norm.weight | 0x3b560ee0 | 0x2000 | +| 38 | blk.2.ffn_up_exps.weight | 0x3b562ee0 | 0x4980000 | +| 39 | blk.3.attn_k.weight | 0x3fee2ee0 | 0x62000 | +| 40 | blk.3.attn_k_norm.weight | 0x3ff44ee0 | 0x200 | +| 41 | blk.3.attn_norm.weight | 0x3ff450e0 | 0x2000 | +| 42 | blk.3.attn_output.weight | 0x3ff470e0 | 0x370000 | +| 43 | blk.3.attn_q.weight | 0x402b70e0 | 0x310000 | +| 44 | blk.3.attn_q_norm.weight | 0x405c70e0 | 0x200 | +| 45 | blk.3.attn_v.weight | 0x405c72e0 | 0x6e000 | +| 46 | blk.3.ffn_down_exps.weight | 0x406352e0 | 0x6c00000 | +| 47 | blk.3.ffn_gate_exps.weight | 0x472352e0 | 0x4980000 | +| 48 | blk.3.ffn_gate_inp.weight | 0x4bbb52e0 | 0x100000 | +| 49 | blk.3.ffn_norm.weight | 0x4bcb52e0 | 0x2000 | +| 50 | blk.3.ffn_up_exps.weight | 0x4bcb72e0 | 0x4980000 | +| 51 | blk.4.attn_k.weight | 0x506372e0 | 0x62000 | +| 52 | blk.4.attn_k_norm.weight | 0x506992e0 | 0x200 | +| 53 | blk.4.attn_norm.weight | 0x506994e0 | 0x2000 | +| 54 | blk.4.attn_output.weight | 0x5069b4e0 | 0x370000 | +| 55 | blk.4.attn_q.weight | 0x50a0b4e0 | 0x310000 | +| 56 | blk.4.attn_q_norm.weight | 0x50d1b4e0 | 0x200 | +| 57 | blk.4.attn_v.weight | 0x50d1b6e0 | 0x6e000 | +| 58 | blk.4.ffn_down_exps.weight | 0x50d896e0 | 0x6c00000 | +| 59 | blk.4.ffn_gate_exps.weight | 0x579896e0 | 0x4980000 | +| 60 | blk.4.ffn_gate_inp.weight | 0x5c3096e0 | 0x100000 | +| 61 | blk.4.ffn_norm.weight | 0x5c4096e0 | 0x2000 | +| 62 | blk.4.ffn_up_exps.weight | 0x5c40b6e0 | 0x4980000 | +| 63 | blk.5.attn_k.weight | 0x60d8b6e0 | 0x62000 | +| 64 | blk.5.attn_k_norm.weight | 0x60ded6e0 | 0x200 | +| 65 | blk.5.attn_norm.weight | 0x60ded8e0 | 0x2000 | +| 66 | blk.5.attn_output.weight | 0x60def8e0 | 0x370000 | +| 67 | blk.5.attn_q.weight | 0x6115f8e0 | 0x310000 | +| 68 | blk.5.attn_q_norm.weight | 0x6146f8e0 | 0x200 | +| 69 | blk.5.attn_v.weight | 0x6146fae0 | 0x6e000 | +| 70 | blk.5.ffn_down_exps.weight | 0x614ddae0 | 0x6c00000 | +| 71 | blk.5.ffn_gate_exps.weight | 0x680ddae0 | 0x4980000 | +| 72 | blk.5.ffn_gate_inp.weight | 0x6ca5dae0 | 0x100000 | +| 73 | blk.5.ffn_norm.weight | 0x6cb5dae0 | 0x2000 | +| 74 | blk.5.ffn_up_exps.weight | 0x6cb5fae0 | 0x4980000 | +| 75 | blk.6.attn_k.weight | 0x714dfae0 | 0x62000 | +| 76 | blk.6.attn_k_norm.weight | 0x71541ae0 | 0x200 | +| 77 | blk.6.attn_norm.weight | 0x71541ce0 | 0x2000 | +| 78 | blk.6.attn_output.weight | 0x71543ce0 | 0x370000 | +| 79 | blk.6.attn_q.weight | 0x718b3ce0 | 0x310000 | +| 80 | blk.6.attn_q_norm.weight | 0x71bc3ce0 | 0x200 | +| 81 | blk.6.attn_v.weight | 0x71bc3ee0 | 0x6e000 | +| 82 | blk.6.ffn_down_exps.weight | 0x71c31ee0 | 0x6c00000 | +| 83 | blk.6.ffn_gate_exps.weight | 0x78831ee0 | 0x4980000 | +| 84 | blk.6.ffn_gate_inp.weight | 0x7d1b1ee0 | 0x100000 | +| 85 | blk.6.ffn_norm.weight | 0x7d2b1ee0 | 0x2000 | +| 86 | blk.6.ffn_up_exps.weight | 0x7d2b3ee0 | 0x4980000 | +| 87 | blk.7.attn_k.weight | 0x81c33ee0 | 0x62000 | +| 88 | blk.7.attn_k_norm.weight | 0x81c95ee0 | 0x200 | +| 89 | blk.7.attn_norm.weight | 0x81c960e0 | 0x2000 | +| 90 | blk.7.attn_output.weight | 0x81c980e0 | 0x370000 | +| 91 | blk.7.attn_q.weight | 0x820080e0 | 0x310000 | +| 92 | blk.7.attn_q_norm.weight | 0x823180e0 | 0x200 | +| 93 | blk.7.attn_v.weight | 0x823182e0 | 0x6e000 | +| 94 | blk.7.ffn_down_exps.weight | 0x823862e0 | 0x6c00000 | +| 95 | blk.7.ffn_gate_exps.weight | 0x88f862e0 | 0x4980000 | +| 96 | blk.7.ffn_gate_inp.weight | 0x8d9062e0 | 0x100000 | +| 97 | blk.7.ffn_norm.weight | 0x8da062e0 | 0x2000 | +| 98 | blk.7.ffn_up_exps.weight | 0x8da082e0 | 0x4980000 | +| 99 | blk.8.attn_k.weight | 0x923882e0 | 0x62000 | +| 100 | blk.8.attn_k_norm.weight | 0x923ea2e0 | 0x200 | +| 101 | blk.8.attn_norm.weight | 0x923ea4e0 | 0x2000 | +| 102 | blk.8.attn_output.weight | 0x923ec4e0 | 0x370000 | +| 103 | blk.8.attn_q.weight | 0x9275c4e0 | 0x310000 | +| 104 | blk.8.attn_q_norm.weight | 0x92a6c4e0 | 0x200 | +| 105 | blk.8.attn_v.weight | 0x92a6c6e0 | 0x6e000 | +| 106 | blk.8.ffn_down_exps.weight | 0x92ada6e0 | 0x6c00000 | +| 107 | blk.8.ffn_gate_exps.weight | 0x996da6e0 | 0x4980000 | +| 108 | blk.8.ffn_gate_inp.weight | 0x9e05a6e0 | 0x100000 | +| 109 | blk.8.ffn_norm.weight | 0x9e15a6e0 | 0x2000 | +| 110 | blk.8.ffn_up_exps.weight | 0x9e15c6e0 | 0x4980000 | +| 111 | blk.9.attn_k.weight | 0xa2adc6e0 | 0x62000 | +| 112 | blk.9.attn_k_norm.weight | 0xa2b3e6e0 | 0x200 | +| 113 | blk.9.attn_norm.weight | 0xa2b3e8e0 | 0x2000 | +| 114 | blk.9.attn_output.weight | 0xa2b408e0 | 0x370000 | +| 115 | blk.9.attn_q.weight | 0xa2eb08e0 | 0x310000 | +| 116 | blk.9.attn_q_norm.weight | 0xa31c08e0 | 0x200 | +| 117 | blk.9.attn_v.weight | 0xa31c0ae0 | 0x6e000 | +| 118 | blk.9.ffn_down_exps.weight | 0xa322eae0 | 0x6c00000 | +| 119 | blk.9.ffn_gate_exps.weight | 0xa9e2eae0 | 0x4980000 | +| 120 | blk.9.ffn_gate_inp.weight | 0xae7aeae0 | 0x100000 | +| 121 | blk.9.ffn_norm.weight | 0xae8aeae0 | 0x2000 | +| 122 | blk.9.ffn_up_exps.weight | 0xae8b0ae0 | 0x4980000 | +| 123 | blk.10.attn_k.weight | 0xb3230ae0 | 0x62000 | +| 124 | blk.10.attn_k_norm.weight | 0xb3292ae0 | 0x200 | +| 125 | blk.10.attn_norm.weight | 0xb3292ce0 | 0x2000 | +| 126 | blk.10.attn_output.weight | 0xb3294ce0 | 0x370000 | +| 127 | blk.10.attn_q.weight | 0xb3604ce0 | 0x310000 | +| 128 | blk.10.attn_q_norm.weight | 0xb3914ce0 | 0x200 | +| 129 | blk.10.attn_v.weight | 0xb3914ee0 | 0x6e000 | +| 130 | blk.10.ffn_down_exps.weight | 0xb3982ee0 | 0x6c00000 | +| 131 | blk.10.ffn_gate_exps.weight | 0xba582ee0 | 0x4980000 | +| 132 | blk.10.ffn_gate_inp.weight | 0xbef02ee0 | 0x100000 | +| 133 | blk.10.ffn_norm.weight | 0xbf002ee0 | 0x2000 | +| 134 | blk.10.ffn_up_exps.weight | 0xbf004ee0 | 0x4980000 | +| 135 | blk.11.attn_k.weight | 0xc3984ee0 | 0x62000 | +| 136 | blk.11.attn_k_norm.weight | 0xc39e6ee0 | 0x200 | +| 137 | blk.11.attn_norm.weight | 0xc39e70e0 | 0x2000 | +| 138 | blk.11.attn_output.weight | 0xc39e90e0 | 0x370000 | +| 139 | blk.11.attn_q.weight | 0xc3d590e0 | 0x310000 | +| 140 | blk.11.attn_q_norm.weight | 0xc40690e0 | 0x200 | +| 141 | blk.11.attn_v.weight | 0xc40692e0 | 0x6e000 | +| 142 | blk.11.ffn_down_exps.weight | 0xc40d72e0 | 0x6c00000 | +| 143 | blk.11.ffn_gate_exps.weight | 0xcacd72e0 | 0x4980000 | +| 144 | blk.11.ffn_gate_inp.weight | 0xcf6572e0 | 0x100000 | +| 145 | blk.11.ffn_norm.weight | 0xcf7572e0 | 0x2000 | +| 146 | blk.11.ffn_up_exps.weight | 0xcf7592e0 | 0x4980000 | +| 147 | blk.12.attn_k.weight | 0xd40d92e0 | 0x62000 | +| 148 | blk.12.attn_k_norm.weight | 0xd413b2e0 | 0x200 | +| 149 | blk.12.attn_norm.weight | 0xd413b4e0 | 0x2000 | +| 150 | blk.12.attn_output.weight | 0xd413d4e0 | 0x370000 | +| 151 | blk.12.attn_q.weight | 0xd44ad4e0 | 0x310000 | +| 152 | blk.12.attn_q_norm.weight | 0xd47bd4e0 | 0x200 | +| 153 | blk.12.attn_v.weight | 0xd47bd6e0 | 0x6e000 | +| 154 | blk.12.ffn_down_exps.weight | 0xd482b6e0 | 0x6c00000 | +| 155 | blk.12.ffn_gate_exps.weight | 0xdb42b6e0 | 0x4980000 | +| 156 | blk.12.ffn_gate_inp.weight | 0xdfdab6e0 | 0x100000 | +| 157 | blk.12.ffn_norm.weight | 0xdfeab6e0 | 0x2000 | +| 158 | blk.12.ffn_up_exps.weight | 0xdfead6e0 | 0x4980000 | +| 159 | blk.13.attn_k.weight | 0xe482d6e0 | 0x62000 | +| 160 | blk.13.attn_k_norm.weight | 0xe488f6e0 | 0x200 | +| 161 | blk.13.attn_norm.weight | 0xe488f8e0 | 0x2000 | +| 162 | blk.13.attn_output.weight | 0xe48918e0 | 0x370000 | +| 163 | blk.13.attn_q.weight | 0xe4c018e0 | 0x310000 | +| 164 | blk.13.attn_q_norm.weight | 0xe4f118e0 | 0x200 | +| 165 | blk.13.attn_v.weight | 0xe4f11ae0 | 0x6e000 | +| 166 | blk.13.ffn_down_exps.weight | 0xe4f7fae0 | 0x6c00000 | +| 167 | blk.13.ffn_gate_exps.weight | 0xebb7fae0 | 0x4980000 | +| 168 | blk.13.ffn_gate_inp.weight | 0xf04ffae0 | 0x100000 | +| 169 | blk.13.ffn_norm.weight | 0xf05ffae0 | 0x2000 | +| 170 | blk.13.ffn_up_exps.weight | 0xf0601ae0 | 0x4980000 | +| 171 | blk.14.attn_k.weight | 0xf4f81ae0 | 0x62000 | +| 172 | blk.14.attn_k_norm.weight | 0xf4fe3ae0 | 0x200 | +| 173 | blk.14.attn_norm.weight | 0xf4fe3ce0 | 0x2000 | +| 174 | blk.14.attn_output.weight | 0xf4fe5ce0 | 0x370000 | +| 175 | blk.14.attn_q.weight | 0xf5355ce0 | 0x310000 | +| 176 | blk.14.attn_q_norm.weight | 0xf5665ce0 | 0x200 | +| 177 | blk.14.attn_v.weight | 0xf5665ee0 | 0x6e000 | +| 178 | blk.14.ffn_down_exps.weight | 0xf56d3ee0 | 0x6c00000 | +| 179 | blk.14.ffn_gate_exps.weight | 0xfc2d3ee0 | 0x4980000 | +| 180 | blk.14.ffn_gate_inp.weight | 0x100c53ee0 | 0x100000 | +| 181 | blk.14.ffn_norm.weight | 0x100d53ee0 | 0x2000 | +| 182 | blk.14.ffn_up_exps.weight | 0x100d55ee0 | 0x4980000 | +| 183 | blk.15.attn_k.weight | 0x1056d5ee0 | 0x62000 | +| 184 | blk.15.attn_k_norm.weight | 0x105737ee0 | 0x200 | +| 185 | blk.15.attn_norm.weight | 0x1057380e0 | 0x2000 | +| 186 | blk.15.attn_output.weight | 0x10573a0e0 | 0x370000 | +| 187 | blk.15.attn_q.weight | 0x105aaa0e0 | 0x310000 | +| 188 | blk.15.attn_q_norm.weight | 0x105dba0e0 | 0x200 | +| 189 | blk.15.attn_v.weight | 0x105dba2e0 | 0x6e000 | +| 190 | blk.15.ffn_down_exps.weight | 0x105e282e0 | 0x6c00000 | +| 191 | blk.15.ffn_gate_exps.weight | 0x10ca282e0 | 0x4980000 | +| 192 | blk.15.ffn_gate_inp.weight | 0x1113a82e0 | 0x100000 | +| 193 | blk.15.ffn_norm.weight | 0x1114a82e0 | 0x2000 | +| 194 | blk.15.ffn_up_exps.weight | 0x1114aa2e0 | 0x4980000 | +| 195 | blk.16.attn_k.weight | 0x115e2a2e0 | 0x62000 | +| 196 | blk.16.attn_k_norm.weight | 0x115e8c2e0 | 0x200 | +| 197 | blk.16.attn_norm.weight | 0x115e8c4e0 | 0x2000 | +| 198 | blk.16.attn_output.weight | 0x115e8e4e0 | 0x370000 | +| 199 | blk.16.attn_q.weight | 0x1161fe4e0 | 0x310000 | +| 200 | blk.16.attn_q_norm.weight | 0x11650e4e0 | 0x200 | +| 201 | blk.16.attn_v.weight | 0x11650e6e0 | 0x6e000 | +| 202 | blk.16.ffn_down_exps.weight | 0x11657c6e0 | 0x6c00000 | +| 203 | blk.16.ffn_gate_exps.weight | 0x11d17c6e0 | 0x4980000 | +| 204 | blk.16.ffn_gate_inp.weight | 0x121afc6e0 | 0x100000 | +| 205 | blk.16.ffn_norm.weight | 0x121bfc6e0 | 0x2000 | +| 206 | blk.16.ffn_up_exps.weight | 0x121bfe6e0 | 0x4980000 | +| 207 | blk.17.attn_k.weight | 0x12657e6e0 | 0x62000 | +| 208 | blk.17.attn_k_norm.weight | 0x1265e06e0 | 0x200 | +| 209 | blk.17.attn_norm.weight | 0x1265e08e0 | 0x2000 | +| 210 | blk.17.attn_output.weight | 0x1265e28e0 | 0x370000 | +| 211 | blk.17.attn_q.weight | 0x1269528e0 | 0x310000 | +| 212 | blk.17.attn_q_norm.weight | 0x126c628e0 | 0x200 | +| 213 | blk.17.attn_v.weight | 0x126c62ae0 | 0x6e000 | +| 214 | blk.17.ffn_down_exps.weight | 0x126cd0ae0 | 0x6c00000 | +| 215 | blk.17.ffn_gate_exps.weight | 0x12d8d0ae0 | 0x4980000 | +| 216 | blk.17.ffn_gate_inp.weight | 0x132250ae0 | 0x100000 | +| 217 | blk.17.ffn_norm.weight | 0x132350ae0 | 0x2000 | +| 218 | blk.17.ffn_up_exps.weight | 0x132352ae0 | 0x4980000 | +| 219 | blk.18.attn_k.weight | 0x136cd2ae0 | 0x62000 | +| 220 | blk.18.attn_k_norm.weight | 0x136d34ae0 | 0x200 | +| 221 | blk.18.attn_norm.weight | 0x136d34ce0 | 0x2000 | +| 222 | blk.18.attn_output.weight | 0x136d36ce0 | 0x370000 | +| 223 | blk.18.attn_q.weight | 0x1370a6ce0 | 0x310000 | +| 224 | blk.18.attn_q_norm.weight | 0x1373b6ce0 | 0x200 | +| 225 | blk.18.attn_v.weight | 0x1373b6ee0 | 0x6e000 | +| 226 | blk.18.ffn_down_exps.weight | 0x137424ee0 | 0x6c00000 | +| 227 | blk.18.ffn_gate_exps.weight | 0x13e024ee0 | 0x5280000 | +| 228 | blk.18.ffn_gate_inp.weight | 0x1432a4ee0 | 0x100000 | +| 229 | blk.18.ffn_norm.weight | 0x1433a4ee0 | 0x2000 | +| 230 | blk.18.ffn_up_exps.weight | 0x1433a6ee0 | 0x5280000 | +| 231 | blk.19.attn_k.weight | 0x148626ee0 | 0x62000 | +| 232 | blk.19.attn_k_norm.weight | 0x148688ee0 | 0x200 | +| 233 | blk.19.attn_norm.weight | 0x1486890e0 | 0x2000 | +| 234 | blk.19.attn_output.weight | 0x14868b0e0 | 0x370000 | +| 235 | blk.19.attn_q.weight | 0x1489fb0e0 | 0x310000 | +| 236 | blk.19.attn_q_norm.weight | 0x148d0b0e0 | 0x200 | +| 237 | blk.19.attn_v.weight | 0x148d0b2e0 | 0x6e000 | +| 238 | blk.19.ffn_down_exps.weight | 0x148d792e0 | 0x6c00000 | +| 239 | blk.19.ffn_gate_exps.weight | 0x14f9792e0 | 0x4980000 | +| 240 | blk.19.ffn_gate_inp.weight | 0x1542f92e0 | 0x100000 | +| 241 | blk.19.ffn_norm.weight | 0x1543f92e0 | 0x2000 | +| 242 | blk.19.ffn_up_exps.weight | 0x1543fb2e0 | 0x4980000 | +| 243 | blk.20.attn_k.weight | 0x158d7b2e0 | 0x62000 | +| 244 | blk.20.attn_k_norm.weight | 0x158ddd2e0 | 0x200 | +| 245 | blk.20.attn_norm.weight | 0x158ddd4e0 | 0x2000 | +| 246 | blk.20.attn_output.weight | 0x158ddf4e0 | 0x370000 | +| 247 | blk.20.attn_q.weight | 0x15914f4e0 | 0x310000 | +| 248 | blk.20.attn_q_norm.weight | 0x15945f4e0 | 0x200 | +| 249 | blk.20.attn_v.weight | 0x15945f6e0 | 0x6e000 | +| 250 | blk.20.ffn_down_exps.weight | 0x1594cd6e0 | 0x6c00000 | +| 251 | blk.20.ffn_gate_exps.weight | 0x1600cd6e0 | 0x4980000 | +| 252 | blk.20.ffn_gate_inp.weight | 0x164a4d6e0 | 0x100000 | +| 253 | blk.20.ffn_norm.weight | 0x164b4d6e0 | 0x2000 | +| 254 | blk.20.ffn_up_exps.weight | 0x164b4f6e0 | 0x4980000 | +| 255 | blk.21.attn_k.weight | 0x1694cf6e0 | 0x62000 | +| 256 | blk.21.attn_k_norm.weight | 0x1695316e0 | 0x200 | +| 257 | blk.21.attn_norm.weight | 0x1695318e0 | 0x2000 | +| 258 | blk.21.attn_output.weight | 0x1695338e0 | 0x370000 | +| 259 | blk.21.attn_q.weight | 0x1698a38e0 | 0x310000 | +| 260 | blk.21.attn_q_norm.weight | 0x169bb38e0 | 0x200 | +| 261 | blk.21.attn_v.weight | 0x169bb3ae0 | 0x6e000 | +| 262 | blk.21.ffn_down_exps.weight | 0x169c21ae0 | 0x6c00000 | +| 263 | blk.21.ffn_gate_exps.weight | 0x170821ae0 | 0x4980000 | +| 264 | blk.21.ffn_gate_inp.weight | 0x1751a1ae0 | 0x100000 | +| 265 | blk.21.ffn_norm.weight | 0x1752a1ae0 | 0x2000 | +| 266 | blk.21.ffn_up_exps.weight | 0x1752a3ae0 | 0x4980000 | +| 267 | blk.22.attn_k.weight | 0x179c23ae0 | 0x62000 | +| 268 | blk.22.attn_k_norm.weight | 0x179c85ae0 | 0x200 | +| 269 | blk.22.attn_norm.weight | 0x179c85ce0 | 0x2000 | +| 270 | blk.22.attn_output.weight | 0x179c87ce0 | 0x370000 | +| 271 | blk.22.attn_q.weight | 0x179ff7ce0 | 0x310000 | +| 272 | blk.22.attn_q_norm.weight | 0x17a307ce0 | 0x200 | +| 273 | blk.22.attn_v.weight | 0x17a307ee0 | 0x6e000 | +| 274 | blk.22.ffn_down_exps.weight | 0x17a375ee0 | 0x6c00000 | +| 275 | blk.22.ffn_gate_exps.weight | 0x180f75ee0 | 0x4980000 | +| 276 | blk.22.ffn_gate_inp.weight | 0x1858f5ee0 | 0x100000 | +| 277 | blk.22.ffn_norm.weight | 0x1859f5ee0 | 0x2000 | +| 278 | blk.22.ffn_up_exps.weight | 0x1859f7ee0 | 0x4980000 | +| 279 | blk.23.attn_k.weight | 0x18a377ee0 | 0x62000 | +| 280 | blk.23.attn_k_norm.weight | 0x18a3d9ee0 | 0x200 | +| 281 | blk.23.attn_norm.weight | 0x18a3da0e0 | 0x2000 | +| 282 | blk.23.attn_output.weight | 0x18a3dc0e0 | 0x370000 | +| 283 | blk.23.attn_q.weight | 0x18a74c0e0 | 0x310000 | +| 284 | blk.23.attn_q_norm.weight | 0x18aa5c0e0 | 0x200 | +| 285 | blk.23.attn_v.weight | 0x18aa5c2e0 | 0x6e000 | +| 286 | blk.23.ffn_down_exps.weight | 0x18aaca2e0 | 0x6c00000 | +| 287 | blk.23.ffn_gate_exps.weight | 0x1916ca2e0 | 0x4980000 | +| 288 | blk.23.ffn_gate_inp.weight | 0x19604a2e0 | 0x100000 | +| 289 | blk.23.ffn_norm.weight | 0x19614a2e0 | 0x2000 | +| 290 | blk.23.ffn_up_exps.weight | 0x19614c2e0 | 0x4980000 | +| 291 | blk.24.attn_k.weight | 0x19aacc2e0 | 0x6e000 | +| 292 | blk.24.attn_k_norm.weight | 0x19ab3a2e0 | 0x200 | +| 293 | blk.24.attn_norm.weight | 0x19ab3a4e0 | 0x2000 | +| 294 | blk.24.attn_output.weight | 0x19ab3c4e0 | 0x370000 | +| 295 | blk.24.attn_q.weight | 0x19aeac4e0 | 0x370000 | +| 296 | blk.24.attn_q_norm.weight | 0x19b21c4e0 | 0x200 | +| 297 | blk.24.attn_v.weight | 0x19b21c6e0 | 0x6e000 | +| 298 | blk.24.ffn_down_exps.weight | 0x19b28a6e0 | 0x6c00000 | +| 299 | blk.24.ffn_gate_exps.weight | 0x1a1e8a6e0 | 0x4980000 | +| 300 | blk.24.ffn_gate_inp.weight | 0x1a680a6e0 | 0x100000 | +| 301 | blk.24.ffn_norm.weight | 0x1a690a6e0 | 0x2000 | +| 302 | blk.24.ffn_up_exps.weight | 0x1a690c6e0 | 0x4980000 | +| 303 | blk.25.attn_k.weight | 0x1ab28c6e0 | 0x6e000 | +| 304 | blk.25.attn_k_norm.weight | 0x1ab2fa6e0 | 0x200 | +| 305 | blk.25.attn_norm.weight | 0x1ab2fa8e0 | 0x2000 | +| 306 | blk.25.attn_output.weight | 0x1ab2fc8e0 | 0x370000 | +| 307 | blk.25.attn_q.weight | 0x1ab66c8e0 | 0x370000 | +| 308 | blk.25.attn_q_norm.weight | 0x1ab9dc8e0 | 0x200 | +| 309 | blk.25.attn_v.weight | 0x1ab9dcae0 | 0x6e000 | +| 310 | blk.25.ffn_down_exps.weight | 0x1aba4aae0 | 0x6c00000 | +| 311 | blk.25.ffn_gate_exps.weight | 0x1b264aae0 | 0x5280000 | +| 312 | blk.25.ffn_gate_inp.weight | 0x1b78caae0 | 0x100000 | +| 313 | blk.25.ffn_norm.weight | 0x1b79caae0 | 0x2000 | +| 314 | blk.25.ffn_up_exps.weight | 0x1b79ccae0 | 0x5280000 | +| 315 | blk.26.attn_k.weight | 0x1bcc4cae0 | 0x6e000 | +| 316 | blk.26.attn_k_norm.weight | 0x1bccbaae0 | 0x200 | +| 317 | blk.26.attn_norm.weight | 0x1bccbace0 | 0x2000 | +| 318 | blk.26.attn_output.weight | 0x1bccbcce0 | 0x370000 | +| 319 | blk.26.attn_q.weight | 0x1bd02cce0 | 0x370000 | +| 320 | blk.26.attn_q_norm.weight | 0x1bd39cce0 | 0x200 | +| 321 | blk.26.attn_v.weight | 0x1bd39cee0 | 0x6e000 | +| 322 | blk.26.ffn_down_exps.weight | 0x1bd40aee0 | 0x6c00000 | +| 323 | blk.26.ffn_gate_exps.weight | 0x1c400aee0 | 0x5280000 | +| 324 | blk.26.ffn_gate_inp.weight | 0x1c928aee0 | 0x100000 | +| 325 | blk.26.ffn_norm.weight | 0x1c938aee0 | 0x2000 | +| 326 | blk.26.ffn_up_exps.weight | 0x1c938cee0 | 0x5280000 | +| 327 | blk.27.attn_k.weight | 0x1ce60cee0 | 0x6e000 | +| 328 | blk.27.attn_k_norm.weight | 0x1ce67aee0 | 0x200 | +| 329 | blk.27.attn_norm.weight | 0x1ce67b0e0 | 0x2000 | +| 330 | blk.27.attn_output.weight | 0x1ce67d0e0 | 0x370000 | +| 331 | blk.27.attn_q.weight | 0x1ce9ed0e0 | 0x370000 | +| 332 | blk.27.attn_q_norm.weight | 0x1ced5d0e0 | 0x200 | +| 333 | blk.27.attn_v.weight | 0x1ced5d2e0 | 0x6e000 | +| 334 | blk.27.ffn_down_exps.weight | 0x1cedcb2e0 | 0x6c00000 | +| 335 | blk.27.ffn_gate_exps.weight | 0x1d59cb2e0 | 0x5280000 | +| 336 | blk.27.ffn_gate_inp.weight | 0x1dac4b2e0 | 0x100000 | +| 337 | blk.27.ffn_norm.weight | 0x1dad4b2e0 | 0x2000 | +| 338 | blk.27.ffn_up_exps.weight | 0x1dad4d2e0 | 0x5280000 | +| 339 | blk.28.attn_k.weight | 0x1dffcd2e0 | 0x6e000 | +| 340 | blk.28.attn_k_norm.weight | 0x1e003b2e0 | 0x200 | +| 341 | blk.28.attn_norm.weight | 0x1e003b4e0 | 0x2000 | +| 342 | blk.28.attn_output.weight | 0x1e003d4e0 | 0x370000 | +| 343 | blk.28.attn_q.weight | 0x1e03ad4e0 | 0x370000 | +| 344 | blk.28.attn_q_norm.weight | 0x1e071d4e0 | 0x200 | +| 345 | blk.28.attn_v.weight | 0x1e071d6e0 | 0x6e000 | +| 346 | blk.28.ffn_down_exps.weight | 0x1e078b6e0 | 0x6c00000 | +| 347 | blk.28.ffn_gate_exps.weight | 0x1e738b6e0 | 0x5280000 | +| 348 | blk.28.ffn_gate_inp.weight | 0x1ec60b6e0 | 0x100000 | +| 349 | blk.28.ffn_norm.weight | 0x1ec70b6e0 | 0x2000 | +| 350 | blk.28.ffn_up_exps.weight | 0x1ec70d6e0 | 0x5280000 | +| 351 | blk.29.attn_k.weight | 0x1f198d6e0 | 0x6e000 | +| 352 | blk.29.attn_k_norm.weight | 0x1f19fb6e0 | 0x200 | +| 353 | blk.29.attn_norm.weight | 0x1f19fb8e0 | 0x2000 | +| 354 | blk.29.attn_output.weight | 0x1f19fd8e0 | 0x370000 | +| 355 | blk.29.attn_q.weight | 0x1f1d6d8e0 | 0x370000 | +| 356 | blk.29.attn_q_norm.weight | 0x1f20dd8e0 | 0x200 | +| 357 | blk.29.attn_v.weight | 0x1f20ddae0 | 0x6e000 | +| 358 | blk.29.ffn_down_exps.weight | 0x1f214bae0 | 0x6c00000 | +| 359 | blk.29.ffn_gate_exps.weight | 0x1f8d4bae0 | 0x5280000 | +| 360 | blk.29.ffn_gate_inp.weight | 0x1fdfcbae0 | 0x100000 | +| 361 | blk.29.ffn_norm.weight | 0x1fe0cbae0 | 0x2000 | +| 362 | blk.29.ffn_up_exps.weight | 0x1fe0cdae0 | 0x5280000 | +| 363 | blk.30.attn_k.weight | 0x20334dae0 | 0x6e000 | +| 364 | blk.30.attn_k_norm.weight | 0x2033bbae0 | 0x200 | +| 365 | blk.30.attn_norm.weight | 0x2033bbce0 | 0x2000 | +| 366 | blk.30.attn_output.weight | 0x2033bdce0 | 0x370000 | +| 367 | blk.30.attn_q.weight | 0x20372dce0 | 0x370000 | +| 368 | blk.30.attn_q_norm.weight | 0x203a9dce0 | 0x200 | +| 369 | blk.30.attn_v.weight | 0x203a9dee0 | 0x6e000 | +| 370 | blk.30.ffn_down_exps.weight | 0x203b0bee0 | 0x6c00000 | +| 371 | blk.30.ffn_gate_exps.weight | 0x20a70bee0 | 0x5280000 | +| 372 | blk.30.ffn_gate_inp.weight | 0x20f98bee0 | 0x100000 | +| 373 | blk.30.ffn_norm.weight | 0x20fa8bee0 | 0x2000 | +| 374 | blk.30.ffn_up_exps.weight | 0x20fa8dee0 | 0x5280000 | +| 375 | blk.31.attn_k.weight | 0x214d0dee0 | 0x6e000 | +| 376 | blk.31.attn_k_norm.weight | 0x214d7bee0 | 0x200 | +| 377 | blk.31.attn_norm.weight | 0x214d7c0e0 | 0x2000 | +| 378 | blk.31.attn_output.weight | 0x214d7e0e0 | 0x370000 | +| 379 | blk.31.attn_q.weight | 0x2150ee0e0 | 0x370000 | +| 380 | blk.31.attn_q_norm.weight | 0x21545e0e0 | 0x200 | +| 381 | blk.31.attn_v.weight | 0x21545e2e0 | 0x6e000 | +| 382 | blk.31.ffn_down_exps.weight | 0x2154cc2e0 | 0x6c00000 | +| 383 | blk.31.ffn_gate_exps.weight | 0x21c0cc2e0 | 0x5280000 | +| 384 | blk.31.ffn_gate_inp.weight | 0x22134c2e0 | 0x100000 | +| 385 | blk.31.ffn_norm.weight | 0x22144c2e0 | 0x2000 | +| 386 | blk.31.ffn_up_exps.weight | 0x22144e2e0 | 0x5280000 | +| 387 | blk.32.attn_k.weight | 0x2266ce2e0 | 0x6e000 | +| 388 | blk.32.attn_k_norm.weight | 0x22673c2e0 | 0x200 | +| 389 | blk.32.attn_norm.weight | 0x22673c4e0 | 0x2000 | +| 390 | blk.32.attn_output.weight | 0x22673e4e0 | 0x370000 | +| 391 | blk.32.attn_q.weight | 0x226aae4e0 | 0x370000 | +| 392 | blk.32.attn_q_norm.weight | 0x226e1e4e0 | 0x200 | +| 393 | blk.32.attn_v.weight | 0x226e1e6e0 | 0x6e000 | +| 394 | blk.32.ffn_down_exps.weight | 0x226e8c6e0 | 0x6c00000 | +| 395 | blk.32.ffn_gate_exps.weight | 0x22da8c6e0 | 0x5280000 | +| 396 | blk.32.ffn_gate_inp.weight | 0x232d0c6e0 | 0x100000 | +| 397 | blk.32.ffn_norm.weight | 0x232e0c6e0 | 0x2000 | +| 398 | blk.32.ffn_up_exps.weight | 0x232e0e6e0 | 0x5280000 | +| 399 | blk.33.attn_k.weight | 0x23808e6e0 | 0x6e000 | +| 400 | blk.33.attn_k_norm.weight | 0x2380fc6e0 | 0x200 | +| 401 | blk.33.attn_norm.weight | 0x2380fc8e0 | 0x2000 | +| 402 | blk.33.attn_output.weight | 0x2380fe8e0 | 0x370000 | +| 403 | blk.33.attn_q.weight | 0x23846e8e0 | 0x370000 | +| 404 | blk.33.attn_q_norm.weight | 0x2387de8e0 | 0x200 | +| 405 | blk.33.attn_v.weight | 0x2387deae0 | 0x6e000 | +| 406 | blk.33.ffn_down_exps.weight | 0x23884cae0 | 0x6c00000 | +| 407 | blk.33.ffn_gate_exps.weight | 0x23f44cae0 | 0x5280000 | +| 408 | blk.33.ffn_gate_inp.weight | 0x2446ccae0 | 0x100000 | +| 409 | blk.33.ffn_norm.weight | 0x2447ccae0 | 0x2000 | +| 410 | blk.33.ffn_up_exps.weight | 0x2447ceae0 | 0x5280000 | +| 411 | blk.34.attn_k.weight | 0x249a4eae0 | 0x6e000 | +| 412 | blk.34.attn_k_norm.weight | 0x249abcae0 | 0x200 | +| 413 | blk.34.attn_norm.weight | 0x249abcce0 | 0x2000 | +| 414 | blk.34.attn_output.weight | 0x249abece0 | 0x370000 | +| 415 | blk.34.attn_q.weight | 0x249e2ece0 | 0x370000 | +| 416 | blk.34.attn_q_norm.weight | 0x24a19ece0 | 0x200 | +| 417 | blk.34.attn_v.weight | 0x24a19eee0 | 0x6e000 | +| 418 | blk.34.ffn_down_exps.weight | 0x24a20cee0 | 0x6c00000 | +| 419 | blk.34.ffn_gate_exps.weight | 0x250e0cee0 | 0x5280000 | +| 420 | blk.34.ffn_gate_inp.weight | 0x25608cee0 | 0x100000 | +| 421 | blk.34.ffn_norm.weight | 0x25618cee0 | 0x2000 | +| 422 | blk.34.ffn_up_exps.weight | 0x25618eee0 | 0x5280000 | +| 423 | blk.35.attn_k.weight | 0x25b40eee0 | 0x6e000 | +| 424 | blk.35.attn_k_norm.weight | 0x25b47cee0 | 0x200 | +| 425 | blk.35.attn_norm.weight | 0x25b47d0e0 | 0x2000 | +| 426 | blk.35.attn_output.weight | 0x25b47f0e0 | 0x370000 | +| 427 | blk.35.attn_q.weight | 0x25b7ef0e0 | 0x370000 | +| 428 | blk.35.attn_q_norm.weight | 0x25bb5f0e0 | 0x200 | +| 429 | blk.35.attn_v.weight | 0x25bb5f2e0 | 0x6e000 | +| 430 | blk.35.ffn_down_exps.weight | 0x25bbcd2e0 | 0x6c00000 | +| 431 | blk.35.ffn_gate_exps.weight | 0x2627cd2e0 | 0x5280000 | +| 432 | blk.35.ffn_gate_inp.weight | 0x267a4d2e0 | 0x100000 | +| 433 | blk.35.ffn_norm.weight | 0x267b4d2e0 | 0x2000 | +| 434 | blk.35.ffn_up_exps.weight | 0x267b4f2e0 | 0x5280000 | +| 435 | blk.36.attn_k.weight | 0x26cdcf2e0 | 0x6e000 | +| 436 | blk.36.attn_k_norm.weight | 0x26ce3d2e0 | 0x200 | +| 437 | blk.36.attn_norm.weight | 0x26ce3d4e0 | 0x2000 | +| 438 | blk.36.attn_output.weight | 0x26ce3f4e0 | 0x370000 | +| 439 | blk.36.attn_q.weight | 0x26d1af4e0 | 0x370000 | +| 440 | blk.36.attn_q_norm.weight | 0x26d51f4e0 | 0x200 | +| 441 | blk.36.attn_v.weight | 0x26d51f6e0 | 0x6e000 | +| 442 | blk.36.ffn_down_exps.weight | 0x26d58d6e0 | 0x6c00000 | +| 443 | blk.36.ffn_gate_exps.weight | 0x27418d6e0 | 0x5280000 | +| 444 | blk.36.ffn_gate_inp.weight | 0x27940d6e0 | 0x100000 | +| 445 | blk.36.ffn_norm.weight | 0x27950d6e0 | 0x2000 | +| 446 | blk.36.ffn_up_exps.weight | 0x27950f6e0 | 0x5280000 | +| 447 | blk.37.attn_k.weight | 0x27e78f6e0 | 0x6e000 | +| 448 | blk.37.attn_k_norm.weight | 0x27e7fd6e0 | 0x200 | +| 449 | blk.37.attn_norm.weight | 0x27e7fd8e0 | 0x2000 | +| 450 | blk.37.attn_output.weight | 0x27e7ff8e0 | 0x370000 | +| 451 | blk.37.attn_q.weight | 0x27eb6f8e0 | 0x370000 | +| 452 | blk.37.attn_q_norm.weight | 0x27eedf8e0 | 0x200 | +| 453 | blk.37.attn_v.weight | 0x27eedfae0 | 0x6e000 | +| 454 | blk.37.ffn_down_exps.weight | 0x27ef4dae0 | 0x6c00000 | +| 455 | blk.37.ffn_gate_exps.weight | 0x285b4dae0 | 0x5280000 | +| 456 | blk.37.ffn_gate_inp.weight | 0x28adcdae0 | 0x100000 | +| 457 | blk.37.ffn_norm.weight | 0x28aecdae0 | 0x2000 | +| 458 | blk.37.ffn_up_exps.weight | 0x28aecfae0 | 0x5280000 | +| 459 | blk.38.attn_k.weight | 0x29014fae0 | 0x6e000 | +| 460 | blk.38.attn_k_norm.weight | 0x2901bdae0 | 0x200 | +| 461 | blk.38.attn_norm.weight | 0x2901bdce0 | 0x2000 | +| 462 | blk.38.attn_output.weight | 0x2901bfce0 | 0x370000 | +| 463 | blk.38.attn_q.weight | 0x29052fce0 | 0x370000 | +| 464 | blk.38.attn_q_norm.weight | 0x29089fce0 | 0x200 | +| 465 | blk.38.attn_v.weight | 0x29089fee0 | 0x6e000 | +| 466 | blk.38.ffn_down_exps.weight | 0x29090dee0 | 0x6c00000 | +| 467 | blk.38.ffn_gate_exps.weight | 0x29750dee0 | 0x5280000 | +| 468 | blk.38.ffn_gate_inp.weight | 0x29c78dee0 | 0x100000 | +| 469 | blk.38.ffn_norm.weight | 0x29c88dee0 | 0x2000 | +| 470 | blk.38.ffn_up_exps.weight | 0x29c88fee0 | 0x5280000 | +| 471 | blk.39.attn_k.weight | 0x2a1b0fee0 | 0x6e000 | +| 472 | blk.39.attn_k_norm.weight | 0x2a1b7dee0 | 0x200 | +| 473 | blk.39.attn_norm.weight | 0x2a1b7e0e0 | 0x2000 | +| 474 | blk.39.attn_output.weight | 0x2a1b800e0 | 0x370000 | +| 475 | blk.39.attn_q.weight | 0x2a1ef00e0 | 0x370000 | +| 476 | blk.39.attn_q_norm.weight | 0x2a22600e0 | 0x200 | +| 477 | blk.39.attn_v.weight | 0x2a22602e0 | 0x6e000 | +| 478 | blk.39.ffn_down_exps.weight | 0x2a22ce2e0 | 0x6c00000 | +| 479 | blk.39.ffn_gate_exps.weight | 0x2a8ece2e0 | 0x5280000 | +| 480 | blk.39.ffn_gate_inp.weight | 0x2ae14e2e0 | 0x100000 | +| 481 | blk.39.ffn_norm.weight | 0x2ae24e2e0 | 0x2000 | +| 482 | blk.39.ffn_up_exps.weight | 0x2ae2502e0 | 0x5280000 | +| 483 | blk.40.attn_k.weight | 0x2b34d02e0 | 0x6e000 | +| 484 | blk.40.attn_k_norm.weight | 0x2b353e2e0 | 0x200 | +| 485 | blk.40.attn_norm.weight | 0x2b353e4e0 | 0x2000 | +| 486 | blk.40.attn_output.weight | 0x2b35404e0 | 0x370000 | +| 487 | blk.40.attn_q.weight | 0x2b38b04e0 | 0x370000 | +| 488 | blk.40.attn_q_norm.weight | 0x2b3c204e0 | 0x200 | +| 489 | blk.40.attn_v.weight | 0x2b3c206e0 | 0x6e000 | +| 490 | blk.40.ffn_down_exps.weight | 0x2b3c8e6e0 | 0x6c00000 | +| 491 | blk.40.ffn_gate_exps.weight | 0x2ba88e6e0 | 0x5280000 | +| 492 | blk.40.ffn_gate_inp.weight | 0x2bfb0e6e0 | 0x100000 | +| 493 | blk.40.ffn_norm.weight | 0x2bfc0e6e0 | 0x2000 | +| 494 | blk.40.ffn_up_exps.weight | 0x2bfc106e0 | 0x5280000 | +| 495 | blk.41.attn_k.weight | 0x2c4e906e0 | 0x6e000 | +| 496 | blk.41.attn_k_norm.weight | 0x2c4efe6e0 | 0x200 | +| 497 | blk.41.attn_norm.weight | 0x2c4efe8e0 | 0x2000 | +| 498 | blk.41.attn_output.weight | 0x2c4f008e0 | 0x370000 | +| 499 | blk.41.attn_q.weight | 0x2c52708e0 | 0x370000 | +| 500 | blk.41.attn_q_norm.weight | 0x2c55e08e0 | 0x200 | +| 501 | blk.41.attn_v.weight | 0x2c55e0ae0 | 0x6e000 | +| 502 | blk.41.ffn_down_exps.weight | 0x2c564eae0 | 0x6c00000 | +| 503 | blk.41.ffn_gate_exps.weight | 0x2cc24eae0 | 0x5280000 | +| 504 | blk.41.ffn_gate_inp.weight | 0x2d14ceae0 | 0x100000 | +| 505 | blk.41.ffn_norm.weight | 0x2d15ceae0 | 0x2000 | +| 506 | blk.41.ffn_up_exps.weight | 0x2d15d0ae0 | 0x5280000 | +| 507 | blk.42.attn_k.weight | 0x2d6850ae0 | 0x6e000 | +| 508 | blk.42.attn_k_norm.weight | 0x2d68beae0 | 0x200 | +| 509 | blk.42.attn_norm.weight | 0x2d68bece0 | 0x2000 | +| 510 | blk.42.attn_output.weight | 0x2d68c0ce0 | 0x370000 | +| 511 | blk.42.attn_q.weight | 0x2d6c30ce0 | 0x370000 | +| 512 | blk.42.attn_q_norm.weight | 0x2d6fa0ce0 | 0x200 | +| 513 | blk.42.attn_v.weight | 0x2d6fa0ee0 | 0x6e000 | +| 514 | blk.42.ffn_down_exps.weight | 0x2d700eee0 | 0x6c00000 | +| 515 | blk.42.ffn_gate_exps.weight | 0x2ddc0eee0 | 0x5280000 | +| 516 | blk.42.ffn_gate_inp.weight | 0x2e2e8eee0 | 0x100000 | +| 517 | blk.42.ffn_norm.weight | 0x2e2f8eee0 | 0x2000 | +| 518 | blk.42.ffn_up_exps.weight | 0x2e2f90ee0 | 0x5280000 | +| 519 | blk.43.attn_k.weight | 0x2e8210ee0 | 0x6e000 | +| 520 | blk.43.attn_k_norm.weight | 0x2e827eee0 | 0x200 | +| 521 | blk.43.attn_norm.weight | 0x2e827f0e0 | 0x2000 | +| 522 | blk.43.attn_output.weight | 0x2e82810e0 | 0x370000 | +| 523 | blk.43.attn_q.weight | 0x2e85f10e0 | 0x370000 | +| 524 | blk.43.attn_q_norm.weight | 0x2e89610e0 | 0x200 | +| 525 | blk.43.attn_v.weight | 0x2e89612e0 | 0x6e000 | +| 526 | blk.43.ffn_down_exps.weight | 0x2e89cf2e0 | 0x6c00000 | +| 527 | blk.43.ffn_gate_exps.weight | 0x2ef5cf2e0 | 0x5280000 | +| 528 | blk.43.ffn_gate_inp.weight | 0x2f484f2e0 | 0x100000 | +| 529 | blk.43.ffn_norm.weight | 0x2f494f2e0 | 0x2000 | +| 530 | blk.43.ffn_up_exps.weight | 0x2f49512e0 | 0x5280000 | +| 531 | blk.44.attn_k.weight | 0x2f9bd12e0 | 0x6e000 | +| 532 | blk.44.attn_k_norm.weight | 0x2f9c3f2e0 | 0x200 | +| 533 | blk.44.attn_norm.weight | 0x2f9c3f4e0 | 0x2000 | +| 534 | blk.44.attn_output.weight | 0x2f9c414e0 | 0x370000 | +| 535 | blk.44.attn_q.weight | 0x2f9fb14e0 | 0x370000 | +| 536 | blk.44.attn_q_norm.weight | 0x2fa3214e0 | 0x200 | +| 537 | blk.44.attn_v.weight | 0x2fa3216e0 | 0x6e000 | +| 538 | blk.44.ffn_down_exps.weight | 0x2fa38f6e0 | 0x6c00000 | +| 539 | blk.44.ffn_gate_exps.weight | 0x300f8f6e0 | 0x5280000 | +| 540 | blk.44.ffn_gate_inp.weight | 0x30620f6e0 | 0x100000 | +| 541 | blk.44.ffn_norm.weight | 0x30630f6e0 | 0x2000 | +| 542 | blk.44.ffn_up_exps.weight | 0x3063116e0 | 0x5280000 | +| 543 | blk.45.attn_k.weight | 0x30b5916e0 | 0x6e000 | +| 544 | blk.45.attn_k_norm.weight | 0x30b5ff6e0 | 0x200 | +| 545 | blk.45.attn_norm.weight | 0x30b5ff8e0 | 0x2000 | +| 546 | blk.45.attn_output.weight | 0x30b6018e0 | 0x370000 | +| 547 | blk.45.attn_q.weight | 0x30b9718e0 | 0x370000 | +| 548 | blk.45.attn_q_norm.weight | 0x30bce18e0 | 0x200 | +| 549 | blk.45.attn_v.weight | 0x30bce1ae0 | 0x6e000 | +| 550 | blk.45.ffn_down_exps.weight | 0x30bd4fae0 | 0x6c00000 | +| 551 | blk.45.ffn_gate_exps.weight | 0x31294fae0 | 0x5280000 | +| 552 | blk.45.ffn_gate_inp.weight | 0x317bcfae0 | 0x100000 | +| 553 | blk.45.ffn_norm.weight | 0x317ccfae0 | 0x2000 | +| 554 | blk.45.ffn_up_exps.weight | 0x317cd1ae0 | 0x5280000 | + +### Base Tensor Group : ~622M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------|:---------------------------------|:------------------|:----------------------|:--------| +| 0 | output.weight | Output (W) | (~311M) 311164928 | 2048 x 151936 x 1 x 1 | IQ3_XXS | +| 1 | output_norm.weight | Output Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 2 | token_embd.weight | Token Embedding (W) | (~311M) 311164928 | 2048 x 151936 x 1 x 1 | IQ3_XXS | + +- Total elements in base: (~622M) 622331904 +- Percentage of total elements: 2.13% + + +### Block 0 Tensor Group : ~623M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:---------------------------|:------------------------------------------------------------------------------------------|:------------------|:----------------------|:--------| +| 3 | blk.0.attn_k.weight | Block 0 Attention Key (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_XXS | +| 4 | blk.0.attn_k_norm.weight | Block 0 Attn_K_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 5 | blk.0.attn_norm.weight | Block 0 Attention Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 6 | blk.0.attn_output.weight | Block 0 Attention Output (W) | ( ~8M) 8388608 | 4096 x 2048 x 1 x 1 | IQ3_S | +| 7 | blk.0.attn_q.weight | Block 0 Attention Query (W) | ( ~8M) 8388608 | 2048 x 4096 x 1 x 1 | IQ3_XXS | +| 8 | blk.0.attn_q_norm.weight | Block 0 Attn_Q_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 9 | blk.0.attn_v.weight | Block 0 Attention Value (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 10 | blk.0.ffn_down_exps.weight | Block 0 Ffn_Down_Exps (W) | (~201M) 201326592 | 768 x 2048 x 128 x 1 | IQ4_NL | +| 11 | blk.0.ffn_gate_exps.weight | Block 0 Ffn_Gate_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_XXS | +| 12 | blk.0.ffn_gate_inp.weight | Block 0 Expert-Routing Layer For The Feed-Forward Network In Mixture Of Expert Models (W) | (~262K) 262144 | 2048 x 128 x 1 x 1 | F32 | +| 13 | blk.0.ffn_norm.weight | Block 0 Feed-Forward Network Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 14 | blk.0.ffn_up_exps.weight | Block 0 Ffn_Up_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_XXS | + +- Total elements in blk.0: (~623M) 623120640 +- Percentage of total elements: 2.13% + + +### Block 1 Tensor Group : ~623M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:---------------------------|:------------------------------------------------------------------------------------------|:------------------|:----------------------|:--------| +| 15 | blk.1.attn_k.weight | Block 1 Attention Key (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_XXS | +| 16 | blk.1.attn_k_norm.weight | Block 1 Attn_K_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 17 | blk.1.attn_norm.weight | Block 1 Attention Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 18 | blk.1.attn_output.weight | Block 1 Attention Output (W) | ( ~8M) 8388608 | 4096 x 2048 x 1 x 1 | IQ3_S | +| 19 | blk.1.attn_q.weight | Block 1 Attention Query (W) | ( ~8M) 8388608 | 2048 x 4096 x 1 x 1 | IQ3_XXS | +| 20 | blk.1.attn_q_norm.weight | Block 1 Attn_Q_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 21 | blk.1.attn_v.weight | Block 1 Attention Value (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 22 | blk.1.ffn_down_exps.weight | Block 1 Ffn_Down_Exps (W) | (~201M) 201326592 | 768 x 2048 x 128 x 1 | IQ4_NL | +| 23 | blk.1.ffn_gate_exps.weight | Block 1 Ffn_Gate_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_XXS | +| 24 | blk.1.ffn_gate_inp.weight | Block 1 Expert-Routing Layer For The Feed-Forward Network In Mixture Of Expert Models (W) | (~262K) 262144 | 2048 x 128 x 1 x 1 | F32 | +| 25 | blk.1.ffn_norm.weight | Block 1 Feed-Forward Network Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 26 | blk.1.ffn_up_exps.weight | Block 1 Ffn_Up_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_XXS | + +- Total elements in blk.1: (~623M) 623120640 +- Percentage of total elements: 2.13% + + +### Block 2 Tensor Group : ~623M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:---------------------------|:------------------------------------------------------------------------------------------|:------------------|:----------------------|:--------| +| 27 | blk.2.attn_k.weight | Block 2 Attention Key (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_XXS | +| 28 | blk.2.attn_k_norm.weight | Block 2 Attn_K_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 29 | blk.2.attn_norm.weight | Block 2 Attention Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 30 | blk.2.attn_output.weight | Block 2 Attention Output (W) | ( ~8M) 8388608 | 4096 x 2048 x 1 x 1 | IQ3_S | +| 31 | blk.2.attn_q.weight | Block 2 Attention Query (W) | ( ~8M) 8388608 | 2048 x 4096 x 1 x 1 | IQ3_XXS | +| 32 | blk.2.attn_q_norm.weight | Block 2 Attn_Q_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 33 | blk.2.attn_v.weight | Block 2 Attention Value (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 34 | blk.2.ffn_down_exps.weight | Block 2 Ffn_Down_Exps (W) | (~201M) 201326592 | 768 x 2048 x 128 x 1 | IQ4_NL | +| 35 | blk.2.ffn_gate_exps.weight | Block 2 Ffn_Gate_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_XXS | +| 36 | blk.2.ffn_gate_inp.weight | Block 2 Expert-Routing Layer For The Feed-Forward Network In Mixture Of Expert Models (W) | (~262K) 262144 | 2048 x 128 x 1 x 1 | F32 | +| 37 | blk.2.ffn_norm.weight | Block 2 Feed-Forward Network Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 38 | blk.2.ffn_up_exps.weight | Block 2 Ffn_Up_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_XXS | + +- Total elements in blk.2: (~623M) 623120640 +- Percentage of total elements: 2.13% + + +### Block 3 Tensor Group : ~623M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:---------------------------|:------------------------------------------------------------------------------------------|:------------------|:----------------------|:--------| +| 39 | blk.3.attn_k.weight | Block 3 Attention Key (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_XXS | +| 40 | blk.3.attn_k_norm.weight | Block 3 Attn_K_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 41 | blk.3.attn_norm.weight | Block 3 Attention Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 42 | blk.3.attn_output.weight | Block 3 Attention Output (W) | ( ~8M) 8388608 | 4096 x 2048 x 1 x 1 | IQ3_S | +| 43 | blk.3.attn_q.weight | Block 3 Attention Query (W) | ( ~8M) 8388608 | 2048 x 4096 x 1 x 1 | IQ3_XXS | +| 44 | blk.3.attn_q_norm.weight | Block 3 Attn_Q_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 45 | blk.3.attn_v.weight | Block 3 Attention Value (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 46 | blk.3.ffn_down_exps.weight | Block 3 Ffn_Down_Exps (W) | (~201M) 201326592 | 768 x 2048 x 128 x 1 | IQ4_NL | +| 47 | blk.3.ffn_gate_exps.weight | Block 3 Ffn_Gate_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_XXS | +| 48 | blk.3.ffn_gate_inp.weight | Block 3 Expert-Routing Layer For The Feed-Forward Network In Mixture Of Expert Models (W) | (~262K) 262144 | 2048 x 128 x 1 x 1 | F32 | +| 49 | blk.3.ffn_norm.weight | Block 3 Feed-Forward Network Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 50 | blk.3.ffn_up_exps.weight | Block 3 Ffn_Up_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_XXS | + +- Total elements in blk.3: (~623M) 623120640 +- Percentage of total elements: 2.13% + + +### Block 4 Tensor Group : ~623M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:---------------------------|:------------------------------------------------------------------------------------------|:------------------|:----------------------|:--------| +| 51 | blk.4.attn_k.weight | Block 4 Attention Key (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_XXS | +| 52 | blk.4.attn_k_norm.weight | Block 4 Attn_K_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 53 | blk.4.attn_norm.weight | Block 4 Attention Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 54 | blk.4.attn_output.weight | Block 4 Attention Output (W) | ( ~8M) 8388608 | 4096 x 2048 x 1 x 1 | IQ3_S | +| 55 | blk.4.attn_q.weight | Block 4 Attention Query (W) | ( ~8M) 8388608 | 2048 x 4096 x 1 x 1 | IQ3_XXS | +| 56 | blk.4.attn_q_norm.weight | Block 4 Attn_Q_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 57 | blk.4.attn_v.weight | Block 4 Attention Value (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 58 | blk.4.ffn_down_exps.weight | Block 4 Ffn_Down_Exps (W) | (~201M) 201326592 | 768 x 2048 x 128 x 1 | IQ4_NL | +| 59 | blk.4.ffn_gate_exps.weight | Block 4 Ffn_Gate_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_XXS | +| 60 | blk.4.ffn_gate_inp.weight | Block 4 Expert-Routing Layer For The Feed-Forward Network In Mixture Of Expert Models (W) | (~262K) 262144 | 2048 x 128 x 1 x 1 | F32 | +| 61 | blk.4.ffn_norm.weight | Block 4 Feed-Forward Network Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 62 | blk.4.ffn_up_exps.weight | Block 4 Ffn_Up_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_XXS | + +- Total elements in blk.4: (~623M) 623120640 +- Percentage of total elements: 2.13% + + +### Block 5 Tensor Group : ~623M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:---------------------------|:------------------------------------------------------------------------------------------|:------------------|:----------------------|:--------| +| 63 | blk.5.attn_k.weight | Block 5 Attention Key (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_XXS | +| 64 | blk.5.attn_k_norm.weight | Block 5 Attn_K_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 65 | blk.5.attn_norm.weight | Block 5 Attention Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 66 | blk.5.attn_output.weight | Block 5 Attention Output (W) | ( ~8M) 8388608 | 4096 x 2048 x 1 x 1 | IQ3_S | +| 67 | blk.5.attn_q.weight | Block 5 Attention Query (W) | ( ~8M) 8388608 | 2048 x 4096 x 1 x 1 | IQ3_XXS | +| 68 | blk.5.attn_q_norm.weight | Block 5 Attn_Q_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 69 | blk.5.attn_v.weight | Block 5 Attention Value (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 70 | blk.5.ffn_down_exps.weight | Block 5 Ffn_Down_Exps (W) | (~201M) 201326592 | 768 x 2048 x 128 x 1 | IQ4_NL | +| 71 | blk.5.ffn_gate_exps.weight | Block 5 Ffn_Gate_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_XXS | +| 72 | blk.5.ffn_gate_inp.weight | Block 5 Expert-Routing Layer For The Feed-Forward Network In Mixture Of Expert Models (W) | (~262K) 262144 | 2048 x 128 x 1 x 1 | F32 | +| 73 | blk.5.ffn_norm.weight | Block 5 Feed-Forward Network Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 74 | blk.5.ffn_up_exps.weight | Block 5 Ffn_Up_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_XXS | + +- Total elements in blk.5: (~623M) 623120640 +- Percentage of total elements: 2.13% + + +### Block 6 Tensor Group : ~623M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:---------------------------|:------------------------------------------------------------------------------------------|:------------------|:----------------------|:--------| +| 75 | blk.6.attn_k.weight | Block 6 Attention Key (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_XXS | +| 76 | blk.6.attn_k_norm.weight | Block 6 Attn_K_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 77 | blk.6.attn_norm.weight | Block 6 Attention Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 78 | blk.6.attn_output.weight | Block 6 Attention Output (W) | ( ~8M) 8388608 | 4096 x 2048 x 1 x 1 | IQ3_S | +| 79 | blk.6.attn_q.weight | Block 6 Attention Query (W) | ( ~8M) 8388608 | 2048 x 4096 x 1 x 1 | IQ3_XXS | +| 80 | blk.6.attn_q_norm.weight | Block 6 Attn_Q_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 81 | blk.6.attn_v.weight | Block 6 Attention Value (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 82 | blk.6.ffn_down_exps.weight | Block 6 Ffn_Down_Exps (W) | (~201M) 201326592 | 768 x 2048 x 128 x 1 | IQ4_NL | +| 83 | blk.6.ffn_gate_exps.weight | Block 6 Ffn_Gate_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_XXS | +| 84 | blk.6.ffn_gate_inp.weight | Block 6 Expert-Routing Layer For The Feed-Forward Network In Mixture Of Expert Models (W) | (~262K) 262144 | 2048 x 128 x 1 x 1 | F32 | +| 85 | blk.6.ffn_norm.weight | Block 6 Feed-Forward Network Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 86 | blk.6.ffn_up_exps.weight | Block 6 Ffn_Up_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_XXS | + +- Total elements in blk.6: (~623M) 623120640 +- Percentage of total elements: 2.13% + + +### Block 7 Tensor Group : ~623M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:---------------------------|:------------------------------------------------------------------------------------------|:------------------|:----------------------|:--------| +| 87 | blk.7.attn_k.weight | Block 7 Attention Key (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_XXS | +| 88 | blk.7.attn_k_norm.weight | Block 7 Attn_K_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 89 | blk.7.attn_norm.weight | Block 7 Attention Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 90 | blk.7.attn_output.weight | Block 7 Attention Output (W) | ( ~8M) 8388608 | 4096 x 2048 x 1 x 1 | IQ3_S | +| 91 | blk.7.attn_q.weight | Block 7 Attention Query (W) | ( ~8M) 8388608 | 2048 x 4096 x 1 x 1 | IQ3_XXS | +| 92 | blk.7.attn_q_norm.weight | Block 7 Attn_Q_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 93 | blk.7.attn_v.weight | Block 7 Attention Value (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 94 | blk.7.ffn_down_exps.weight | Block 7 Ffn_Down_Exps (W) | (~201M) 201326592 | 768 x 2048 x 128 x 1 | IQ4_NL | +| 95 | blk.7.ffn_gate_exps.weight | Block 7 Ffn_Gate_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_XXS | +| 96 | blk.7.ffn_gate_inp.weight | Block 7 Expert-Routing Layer For The Feed-Forward Network In Mixture Of Expert Models (W) | (~262K) 262144 | 2048 x 128 x 1 x 1 | F32 | +| 97 | blk.7.ffn_norm.weight | Block 7 Feed-Forward Network Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 98 | blk.7.ffn_up_exps.weight | Block 7 Ffn_Up_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_XXS | + +- Total elements in blk.7: (~623M) 623120640 +- Percentage of total elements: 2.13% + + +### Block 8 Tensor Group : ~623M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:---------------------------|:------------------------------------------------------------------------------------------|:------------------|:----------------------|:--------| +| 99 | blk.8.attn_k.weight | Block 8 Attention Key (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_XXS | +| 100 | blk.8.attn_k_norm.weight | Block 8 Attn_K_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 101 | blk.8.attn_norm.weight | Block 8 Attention Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 102 | blk.8.attn_output.weight | Block 8 Attention Output (W) | ( ~8M) 8388608 | 4096 x 2048 x 1 x 1 | IQ3_S | +| 103 | blk.8.attn_q.weight | Block 8 Attention Query (W) | ( ~8M) 8388608 | 2048 x 4096 x 1 x 1 | IQ3_XXS | +| 104 | blk.8.attn_q_norm.weight | Block 8 Attn_Q_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 105 | blk.8.attn_v.weight | Block 8 Attention Value (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 106 | blk.8.ffn_down_exps.weight | Block 8 Ffn_Down_Exps (W) | (~201M) 201326592 | 768 x 2048 x 128 x 1 | IQ4_NL | +| 107 | blk.8.ffn_gate_exps.weight | Block 8 Ffn_Gate_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_XXS | +| 108 | blk.8.ffn_gate_inp.weight | Block 8 Expert-Routing Layer For The Feed-Forward Network In Mixture Of Expert Models (W) | (~262K) 262144 | 2048 x 128 x 1 x 1 | F32 | +| 109 | blk.8.ffn_norm.weight | Block 8 Feed-Forward Network Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 110 | blk.8.ffn_up_exps.weight | Block 8 Ffn_Up_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_XXS | + +- Total elements in blk.8: (~623M) 623120640 +- Percentage of total elements: 2.13% + + +### Block 9 Tensor Group : ~623M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:---------------------------|:------------------------------------------------------------------------------------------|:------------------|:----------------------|:--------| +| 111 | blk.9.attn_k.weight | Block 9 Attention Key (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_XXS | +| 112 | blk.9.attn_k_norm.weight | Block 9 Attn_K_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 113 | blk.9.attn_norm.weight | Block 9 Attention Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 114 | blk.9.attn_output.weight | Block 9 Attention Output (W) | ( ~8M) 8388608 | 4096 x 2048 x 1 x 1 | IQ3_S | +| 115 | blk.9.attn_q.weight | Block 9 Attention Query (W) | ( ~8M) 8388608 | 2048 x 4096 x 1 x 1 | IQ3_XXS | +| 116 | blk.9.attn_q_norm.weight | Block 9 Attn_Q_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 117 | blk.9.attn_v.weight | Block 9 Attention Value (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 118 | blk.9.ffn_down_exps.weight | Block 9 Ffn_Down_Exps (W) | (~201M) 201326592 | 768 x 2048 x 128 x 1 | IQ4_NL | +| 119 | blk.9.ffn_gate_exps.weight | Block 9 Ffn_Gate_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_XXS | +| 120 | blk.9.ffn_gate_inp.weight | Block 9 Expert-Routing Layer For The Feed-Forward Network In Mixture Of Expert Models (W) | (~262K) 262144 | 2048 x 128 x 1 x 1 | F32 | +| 121 | blk.9.ffn_norm.weight | Block 9 Feed-Forward Network Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 122 | blk.9.ffn_up_exps.weight | Block 9 Ffn_Up_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_XXS | + +- Total elements in blk.9: (~623M) 623120640 +- Percentage of total elements: 2.13% + + +### Block 10 Tensor Group : ~623M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:----------------------------|:-------------------------------------------------------------------------------------------|:------------------|:----------------------|:--------| +| 123 | blk.10.attn_k.weight | Block 10 Attention Key (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_XXS | +| 124 | blk.10.attn_k_norm.weight | Block 10 Attn_K_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 125 | blk.10.attn_norm.weight | Block 10 Attention Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 126 | blk.10.attn_output.weight | Block 10 Attention Output (W) | ( ~8M) 8388608 | 4096 x 2048 x 1 x 1 | IQ3_S | +| 127 | blk.10.attn_q.weight | Block 10 Attention Query (W) | ( ~8M) 8388608 | 2048 x 4096 x 1 x 1 | IQ3_XXS | +| 128 | blk.10.attn_q_norm.weight | Block 10 Attn_Q_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 129 | blk.10.attn_v.weight | Block 10 Attention Value (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 130 | blk.10.ffn_down_exps.weight | Block 10 Ffn_Down_Exps (W) | (~201M) 201326592 | 768 x 2048 x 128 x 1 | IQ4_NL | +| 131 | blk.10.ffn_gate_exps.weight | Block 10 Ffn_Gate_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_XXS | +| 132 | blk.10.ffn_gate_inp.weight | Block 10 Expert-Routing Layer For The Feed-Forward Network In Mixture Of Expert Models (W) | (~262K) 262144 | 2048 x 128 x 1 x 1 | F32 | +| 133 | blk.10.ffn_norm.weight | Block 10 Feed-Forward Network Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 134 | blk.10.ffn_up_exps.weight | Block 10 Ffn_Up_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_XXS | + +- Total elements in blk.10: (~623M) 623120640 +- Percentage of total elements: 2.13% + + +### Block 11 Tensor Group : ~623M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:----------------------------|:-------------------------------------------------------------------------------------------|:------------------|:----------------------|:--------| +| 135 | blk.11.attn_k.weight | Block 11 Attention Key (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_XXS | +| 136 | blk.11.attn_k_norm.weight | Block 11 Attn_K_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 137 | blk.11.attn_norm.weight | Block 11 Attention Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 138 | blk.11.attn_output.weight | Block 11 Attention Output (W) | ( ~8M) 8388608 | 4096 x 2048 x 1 x 1 | IQ3_S | +| 139 | blk.11.attn_q.weight | Block 11 Attention Query (W) | ( ~8M) 8388608 | 2048 x 4096 x 1 x 1 | IQ3_XXS | +| 140 | blk.11.attn_q_norm.weight | Block 11 Attn_Q_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 141 | blk.11.attn_v.weight | Block 11 Attention Value (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 142 | blk.11.ffn_down_exps.weight | Block 11 Ffn_Down_Exps (W) | (~201M) 201326592 | 768 x 2048 x 128 x 1 | IQ4_NL | +| 143 | blk.11.ffn_gate_exps.weight | Block 11 Ffn_Gate_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_XXS | +| 144 | blk.11.ffn_gate_inp.weight | Block 11 Expert-Routing Layer For The Feed-Forward Network In Mixture Of Expert Models (W) | (~262K) 262144 | 2048 x 128 x 1 x 1 | F32 | +| 145 | blk.11.ffn_norm.weight | Block 11 Feed-Forward Network Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 146 | blk.11.ffn_up_exps.weight | Block 11 Ffn_Up_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_XXS | + +- Total elements in blk.11: (~623M) 623120640 +- Percentage of total elements: 2.13% + + +### Block 12 Tensor Group : ~623M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:----------------------------|:-------------------------------------------------------------------------------------------|:------------------|:----------------------|:--------| +| 147 | blk.12.attn_k.weight | Block 12 Attention Key (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_XXS | +| 148 | blk.12.attn_k_norm.weight | Block 12 Attn_K_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 149 | blk.12.attn_norm.weight | Block 12 Attention Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 150 | blk.12.attn_output.weight | Block 12 Attention Output (W) | ( ~8M) 8388608 | 4096 x 2048 x 1 x 1 | IQ3_S | +| 151 | blk.12.attn_q.weight | Block 12 Attention Query (W) | ( ~8M) 8388608 | 2048 x 4096 x 1 x 1 | IQ3_XXS | +| 152 | blk.12.attn_q_norm.weight | Block 12 Attn_Q_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 153 | blk.12.attn_v.weight | Block 12 Attention Value (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 154 | blk.12.ffn_down_exps.weight | Block 12 Ffn_Down_Exps (W) | (~201M) 201326592 | 768 x 2048 x 128 x 1 | IQ4_NL | +| 155 | blk.12.ffn_gate_exps.weight | Block 12 Ffn_Gate_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_XXS | +| 156 | blk.12.ffn_gate_inp.weight | Block 12 Expert-Routing Layer For The Feed-Forward Network In Mixture Of Expert Models (W) | (~262K) 262144 | 2048 x 128 x 1 x 1 | F32 | +| 157 | blk.12.ffn_norm.weight | Block 12 Feed-Forward Network Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 158 | blk.12.ffn_up_exps.weight | Block 12 Ffn_Up_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_XXS | + +- Total elements in blk.12: (~623M) 623120640 +- Percentage of total elements: 2.13% + + +### Block 13 Tensor Group : ~623M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:----------------------------|:-------------------------------------------------------------------------------------------|:------------------|:----------------------|:--------| +| 159 | blk.13.attn_k.weight | Block 13 Attention Key (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_XXS | +| 160 | blk.13.attn_k_norm.weight | Block 13 Attn_K_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 161 | blk.13.attn_norm.weight | Block 13 Attention Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 162 | blk.13.attn_output.weight | Block 13 Attention Output (W) | ( ~8M) 8388608 | 4096 x 2048 x 1 x 1 | IQ3_S | +| 163 | blk.13.attn_q.weight | Block 13 Attention Query (W) | ( ~8M) 8388608 | 2048 x 4096 x 1 x 1 | IQ3_XXS | +| 164 | blk.13.attn_q_norm.weight | Block 13 Attn_Q_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 165 | blk.13.attn_v.weight | Block 13 Attention Value (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 166 | blk.13.ffn_down_exps.weight | Block 13 Ffn_Down_Exps (W) | (~201M) 201326592 | 768 x 2048 x 128 x 1 | IQ4_NL | +| 167 | blk.13.ffn_gate_exps.weight | Block 13 Ffn_Gate_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_XXS | +| 168 | blk.13.ffn_gate_inp.weight | Block 13 Expert-Routing Layer For The Feed-Forward Network In Mixture Of Expert Models (W) | (~262K) 262144 | 2048 x 128 x 1 x 1 | F32 | +| 169 | blk.13.ffn_norm.weight | Block 13 Feed-Forward Network Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 170 | blk.13.ffn_up_exps.weight | Block 13 Ffn_Up_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_XXS | + +- Total elements in blk.13: (~623M) 623120640 +- Percentage of total elements: 2.13% + + +### Block 14 Tensor Group : ~623M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:----------------------------|:-------------------------------------------------------------------------------------------|:------------------|:----------------------|:--------| +| 171 | blk.14.attn_k.weight | Block 14 Attention Key (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_XXS | +| 172 | blk.14.attn_k_norm.weight | Block 14 Attn_K_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 173 | blk.14.attn_norm.weight | Block 14 Attention Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 174 | blk.14.attn_output.weight | Block 14 Attention Output (W) | ( ~8M) 8388608 | 4096 x 2048 x 1 x 1 | IQ3_S | +| 175 | blk.14.attn_q.weight | Block 14 Attention Query (W) | ( ~8M) 8388608 | 2048 x 4096 x 1 x 1 | IQ3_XXS | +| 176 | blk.14.attn_q_norm.weight | Block 14 Attn_Q_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 177 | blk.14.attn_v.weight | Block 14 Attention Value (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 178 | blk.14.ffn_down_exps.weight | Block 14 Ffn_Down_Exps (W) | (~201M) 201326592 | 768 x 2048 x 128 x 1 | IQ4_NL | +| 179 | blk.14.ffn_gate_exps.weight | Block 14 Ffn_Gate_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_XXS | +| 180 | blk.14.ffn_gate_inp.weight | Block 14 Expert-Routing Layer For The Feed-Forward Network In Mixture Of Expert Models (W) | (~262K) 262144 | 2048 x 128 x 1 x 1 | F32 | +| 181 | blk.14.ffn_norm.weight | Block 14 Feed-Forward Network Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 182 | blk.14.ffn_up_exps.weight | Block 14 Ffn_Up_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_XXS | + +- Total elements in blk.14: (~623M) 623120640 +- Percentage of total elements: 2.13% + + +### Block 15 Tensor Group : ~623M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:----------------------------|:-------------------------------------------------------------------------------------------|:------------------|:----------------------|:--------| +| 183 | blk.15.attn_k.weight | Block 15 Attention Key (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_XXS | +| 184 | blk.15.attn_k_norm.weight | Block 15 Attn_K_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 185 | blk.15.attn_norm.weight | Block 15 Attention Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 186 | blk.15.attn_output.weight | Block 15 Attention Output (W) | ( ~8M) 8388608 | 4096 x 2048 x 1 x 1 | IQ3_S | +| 187 | blk.15.attn_q.weight | Block 15 Attention Query (W) | ( ~8M) 8388608 | 2048 x 4096 x 1 x 1 | IQ3_XXS | +| 188 | blk.15.attn_q_norm.weight | Block 15 Attn_Q_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 189 | blk.15.attn_v.weight | Block 15 Attention Value (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 190 | blk.15.ffn_down_exps.weight | Block 15 Ffn_Down_Exps (W) | (~201M) 201326592 | 768 x 2048 x 128 x 1 | IQ4_NL | +| 191 | blk.15.ffn_gate_exps.weight | Block 15 Ffn_Gate_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_XXS | +| 192 | blk.15.ffn_gate_inp.weight | Block 15 Expert-Routing Layer For The Feed-Forward Network In Mixture Of Expert Models (W) | (~262K) 262144 | 2048 x 128 x 1 x 1 | F32 | +| 193 | blk.15.ffn_norm.weight | Block 15 Feed-Forward Network Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 194 | blk.15.ffn_up_exps.weight | Block 15 Ffn_Up_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_XXS | + +- Total elements in blk.15: (~623M) 623120640 +- Percentage of total elements: 2.13% + + +### Block 16 Tensor Group : ~623M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:----------------------------|:-------------------------------------------------------------------------------------------|:------------------|:----------------------|:--------| +| 195 | blk.16.attn_k.weight | Block 16 Attention Key (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_XXS | +| 196 | blk.16.attn_k_norm.weight | Block 16 Attn_K_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 197 | blk.16.attn_norm.weight | Block 16 Attention Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 198 | blk.16.attn_output.weight | Block 16 Attention Output (W) | ( ~8M) 8388608 | 4096 x 2048 x 1 x 1 | IQ3_S | +| 199 | blk.16.attn_q.weight | Block 16 Attention Query (W) | ( ~8M) 8388608 | 2048 x 4096 x 1 x 1 | IQ3_XXS | +| 200 | blk.16.attn_q_norm.weight | Block 16 Attn_Q_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 201 | blk.16.attn_v.weight | Block 16 Attention Value (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 202 | blk.16.ffn_down_exps.weight | Block 16 Ffn_Down_Exps (W) | (~201M) 201326592 | 768 x 2048 x 128 x 1 | IQ4_NL | +| 203 | blk.16.ffn_gate_exps.weight | Block 16 Ffn_Gate_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_XXS | +| 204 | blk.16.ffn_gate_inp.weight | Block 16 Expert-Routing Layer For The Feed-Forward Network In Mixture Of Expert Models (W) | (~262K) 262144 | 2048 x 128 x 1 x 1 | F32 | +| 205 | blk.16.ffn_norm.weight | Block 16 Feed-Forward Network Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 206 | blk.16.ffn_up_exps.weight | Block 16 Ffn_Up_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_XXS | + +- Total elements in blk.16: (~623M) 623120640 +- Percentage of total elements: 2.13% + + +### Block 17 Tensor Group : ~623M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:----------------------------|:-------------------------------------------------------------------------------------------|:------------------|:----------------------|:--------| +| 207 | blk.17.attn_k.weight | Block 17 Attention Key (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_XXS | +| 208 | blk.17.attn_k_norm.weight | Block 17 Attn_K_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 209 | blk.17.attn_norm.weight | Block 17 Attention Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 210 | blk.17.attn_output.weight | Block 17 Attention Output (W) | ( ~8M) 8388608 | 4096 x 2048 x 1 x 1 | IQ3_S | +| 211 | blk.17.attn_q.weight | Block 17 Attention Query (W) | ( ~8M) 8388608 | 2048 x 4096 x 1 x 1 | IQ3_XXS | +| 212 | blk.17.attn_q_norm.weight | Block 17 Attn_Q_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 213 | blk.17.attn_v.weight | Block 17 Attention Value (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 214 | blk.17.ffn_down_exps.weight | Block 17 Ffn_Down_Exps (W) | (~201M) 201326592 | 768 x 2048 x 128 x 1 | IQ4_NL | +| 215 | blk.17.ffn_gate_exps.weight | Block 17 Ffn_Gate_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_XXS | +| 216 | blk.17.ffn_gate_inp.weight | Block 17 Expert-Routing Layer For The Feed-Forward Network In Mixture Of Expert Models (W) | (~262K) 262144 | 2048 x 128 x 1 x 1 | F32 | +| 217 | blk.17.ffn_norm.weight | Block 17 Feed-Forward Network Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 218 | blk.17.ffn_up_exps.weight | Block 17 Ffn_Up_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_XXS | + +- Total elements in blk.17: (~623M) 623120640 +- Percentage of total elements: 2.13% + + +### Block 18 Tensor Group : ~623M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:----------------------------|:-------------------------------------------------------------------------------------------|:------------------|:----------------------|:--------| +| 219 | blk.18.attn_k.weight | Block 18 Attention Key (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_XXS | +| 220 | blk.18.attn_k_norm.weight | Block 18 Attn_K_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 221 | blk.18.attn_norm.weight | Block 18 Attention Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 222 | blk.18.attn_output.weight | Block 18 Attention Output (W) | ( ~8M) 8388608 | 4096 x 2048 x 1 x 1 | IQ3_S | +| 223 | blk.18.attn_q.weight | Block 18 Attention Query (W) | ( ~8M) 8388608 | 2048 x 4096 x 1 x 1 | IQ3_XXS | +| 224 | blk.18.attn_q_norm.weight | Block 18 Attn_Q_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 225 | blk.18.attn_v.weight | Block 18 Attention Value (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 226 | blk.18.ffn_down_exps.weight | Block 18 Ffn_Down_Exps (W) | (~201M) 201326592 | 768 x 2048 x 128 x 1 | IQ4_NL | +| 227 | blk.18.ffn_gate_exps.weight | Block 18 Ffn_Gate_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_S | +| 228 | blk.18.ffn_gate_inp.weight | Block 18 Expert-Routing Layer For The Feed-Forward Network In Mixture Of Expert Models (W) | (~262K) 262144 | 2048 x 128 x 1 x 1 | F32 | +| 229 | blk.18.ffn_norm.weight | Block 18 Feed-Forward Network Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 230 | blk.18.ffn_up_exps.weight | Block 18 Ffn_Up_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_S | + +- Total elements in blk.18: (~623M) 623120640 +- Percentage of total elements: 2.13% + + +### Block 19 Tensor Group : ~623M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:----------------------------|:-------------------------------------------------------------------------------------------|:------------------|:----------------------|:--------| +| 231 | blk.19.attn_k.weight | Block 19 Attention Key (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_XXS | +| 232 | blk.19.attn_k_norm.weight | Block 19 Attn_K_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 233 | blk.19.attn_norm.weight | Block 19 Attention Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 234 | blk.19.attn_output.weight | Block 19 Attention Output (W) | ( ~8M) 8388608 | 4096 x 2048 x 1 x 1 | IQ3_S | +| 235 | blk.19.attn_q.weight | Block 19 Attention Query (W) | ( ~8M) 8388608 | 2048 x 4096 x 1 x 1 | IQ3_XXS | +| 236 | blk.19.attn_q_norm.weight | Block 19 Attn_Q_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 237 | blk.19.attn_v.weight | Block 19 Attention Value (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 238 | blk.19.ffn_down_exps.weight | Block 19 Ffn_Down_Exps (W) | (~201M) 201326592 | 768 x 2048 x 128 x 1 | IQ4_NL | +| 239 | blk.19.ffn_gate_exps.weight | Block 19 Ffn_Gate_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_XXS | +| 240 | blk.19.ffn_gate_inp.weight | Block 19 Expert-Routing Layer For The Feed-Forward Network In Mixture Of Expert Models (W) | (~262K) 262144 | 2048 x 128 x 1 x 1 | F32 | +| 241 | blk.19.ffn_norm.weight | Block 19 Feed-Forward Network Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 242 | blk.19.ffn_up_exps.weight | Block 19 Ffn_Up_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_XXS | + +- Total elements in blk.19: (~623M) 623120640 +- Percentage of total elements: 2.13% + + +### Block 20 Tensor Group : ~623M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:----------------------------|:-------------------------------------------------------------------------------------------|:------------------|:----------------------|:--------| +| 243 | blk.20.attn_k.weight | Block 20 Attention Key (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_XXS | +| 244 | blk.20.attn_k_norm.weight | Block 20 Attn_K_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 245 | blk.20.attn_norm.weight | Block 20 Attention Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 246 | blk.20.attn_output.weight | Block 20 Attention Output (W) | ( ~8M) 8388608 | 4096 x 2048 x 1 x 1 | IQ3_S | +| 247 | blk.20.attn_q.weight | Block 20 Attention Query (W) | ( ~8M) 8388608 | 2048 x 4096 x 1 x 1 | IQ3_XXS | +| 248 | blk.20.attn_q_norm.weight | Block 20 Attn_Q_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 249 | blk.20.attn_v.weight | Block 20 Attention Value (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 250 | blk.20.ffn_down_exps.weight | Block 20 Ffn_Down_Exps (W) | (~201M) 201326592 | 768 x 2048 x 128 x 1 | IQ4_NL | +| 251 | blk.20.ffn_gate_exps.weight | Block 20 Ffn_Gate_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_XXS | +| 252 | blk.20.ffn_gate_inp.weight | Block 20 Expert-Routing Layer For The Feed-Forward Network In Mixture Of Expert Models (W) | (~262K) 262144 | 2048 x 128 x 1 x 1 | F32 | +| 253 | blk.20.ffn_norm.weight | Block 20 Feed-Forward Network Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 254 | blk.20.ffn_up_exps.weight | Block 20 Ffn_Up_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_XXS | + +- Total elements in blk.20: (~623M) 623120640 +- Percentage of total elements: 2.13% + + +### Block 21 Tensor Group : ~623M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:----------------------------|:-------------------------------------------------------------------------------------------|:------------------|:----------------------|:--------| +| 255 | blk.21.attn_k.weight | Block 21 Attention Key (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_XXS | +| 256 | blk.21.attn_k_norm.weight | Block 21 Attn_K_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 257 | blk.21.attn_norm.weight | Block 21 Attention Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 258 | blk.21.attn_output.weight | Block 21 Attention Output (W) | ( ~8M) 8388608 | 4096 x 2048 x 1 x 1 | IQ3_S | +| 259 | blk.21.attn_q.weight | Block 21 Attention Query (W) | ( ~8M) 8388608 | 2048 x 4096 x 1 x 1 | IQ3_XXS | +| 260 | blk.21.attn_q_norm.weight | Block 21 Attn_Q_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 261 | blk.21.attn_v.weight | Block 21 Attention Value (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 262 | blk.21.ffn_down_exps.weight | Block 21 Ffn_Down_Exps (W) | (~201M) 201326592 | 768 x 2048 x 128 x 1 | IQ4_NL | +| 263 | blk.21.ffn_gate_exps.weight | Block 21 Ffn_Gate_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_XXS | +| 264 | blk.21.ffn_gate_inp.weight | Block 21 Expert-Routing Layer For The Feed-Forward Network In Mixture Of Expert Models (W) | (~262K) 262144 | 2048 x 128 x 1 x 1 | F32 | +| 265 | blk.21.ffn_norm.weight | Block 21 Feed-Forward Network Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 266 | blk.21.ffn_up_exps.weight | Block 21 Ffn_Up_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_XXS | + +- Total elements in blk.21: (~623M) 623120640 +- Percentage of total elements: 2.13% + + +### Block 22 Tensor Group : ~623M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:----------------------------|:-------------------------------------------------------------------------------------------|:------------------|:----------------------|:--------| +| 267 | blk.22.attn_k.weight | Block 22 Attention Key (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_XXS | +| 268 | blk.22.attn_k_norm.weight | Block 22 Attn_K_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 269 | blk.22.attn_norm.weight | Block 22 Attention Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 270 | blk.22.attn_output.weight | Block 22 Attention Output (W) | ( ~8M) 8388608 | 4096 x 2048 x 1 x 1 | IQ3_S | +| 271 | blk.22.attn_q.weight | Block 22 Attention Query (W) | ( ~8M) 8388608 | 2048 x 4096 x 1 x 1 | IQ3_XXS | +| 272 | blk.22.attn_q_norm.weight | Block 22 Attn_Q_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 273 | blk.22.attn_v.weight | Block 22 Attention Value (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 274 | blk.22.ffn_down_exps.weight | Block 22 Ffn_Down_Exps (W) | (~201M) 201326592 | 768 x 2048 x 128 x 1 | IQ4_NL | +| 275 | blk.22.ffn_gate_exps.weight | Block 22 Ffn_Gate_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_XXS | +| 276 | blk.22.ffn_gate_inp.weight | Block 22 Expert-Routing Layer For The Feed-Forward Network In Mixture Of Expert Models (W) | (~262K) 262144 | 2048 x 128 x 1 x 1 | F32 | +| 277 | blk.22.ffn_norm.weight | Block 22 Feed-Forward Network Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 278 | blk.22.ffn_up_exps.weight | Block 22 Ffn_Up_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_XXS | + +- Total elements in blk.22: (~623M) 623120640 +- Percentage of total elements: 2.13% + + +### Block 23 Tensor Group : ~623M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:----------------------------|:-------------------------------------------------------------------------------------------|:------------------|:----------------------|:--------| +| 279 | blk.23.attn_k.weight | Block 23 Attention Key (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_XXS | +| 280 | blk.23.attn_k_norm.weight | Block 23 Attn_K_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 281 | blk.23.attn_norm.weight | Block 23 Attention Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 282 | blk.23.attn_output.weight | Block 23 Attention Output (W) | ( ~8M) 8388608 | 4096 x 2048 x 1 x 1 | IQ3_S | +| 283 | blk.23.attn_q.weight | Block 23 Attention Query (W) | ( ~8M) 8388608 | 2048 x 4096 x 1 x 1 | IQ3_XXS | +| 284 | blk.23.attn_q_norm.weight | Block 23 Attn_Q_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 285 | blk.23.attn_v.weight | Block 23 Attention Value (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 286 | blk.23.ffn_down_exps.weight | Block 23 Ffn_Down_Exps (W) | (~201M) 201326592 | 768 x 2048 x 128 x 1 | IQ4_NL | +| 287 | blk.23.ffn_gate_exps.weight | Block 23 Ffn_Gate_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_XXS | +| 288 | blk.23.ffn_gate_inp.weight | Block 23 Expert-Routing Layer For The Feed-Forward Network In Mixture Of Expert Models (W) | (~262K) 262144 | 2048 x 128 x 1 x 1 | F32 | +| 289 | blk.23.ffn_norm.weight | Block 23 Feed-Forward Network Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 290 | blk.23.ffn_up_exps.weight | Block 23 Ffn_Up_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_XXS | + +- Total elements in blk.23: (~623M) 623120640 +- Percentage of total elements: 2.13% + + +### Block 24 Tensor Group : ~623M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:----------------------------|:-------------------------------------------------------------------------------------------|:------------------|:----------------------|:--------| +| 291 | blk.24.attn_k.weight | Block 24 Attention Key (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 292 | blk.24.attn_k_norm.weight | Block 24 Attn_K_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 293 | blk.24.attn_norm.weight | Block 24 Attention Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 294 | blk.24.attn_output.weight | Block 24 Attention Output (W) | ( ~8M) 8388608 | 4096 x 2048 x 1 x 1 | IQ3_S | +| 295 | blk.24.attn_q.weight | Block 24 Attention Query (W) | ( ~8M) 8388608 | 2048 x 4096 x 1 x 1 | IQ3_S | +| 296 | blk.24.attn_q_norm.weight | Block 24 Attn_Q_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 297 | blk.24.attn_v.weight | Block 24 Attention Value (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 298 | blk.24.ffn_down_exps.weight | Block 24 Ffn_Down_Exps (W) | (~201M) 201326592 | 768 x 2048 x 128 x 1 | IQ4_NL | +| 299 | blk.24.ffn_gate_exps.weight | Block 24 Ffn_Gate_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_XXS | +| 300 | blk.24.ffn_gate_inp.weight | Block 24 Expert-Routing Layer For The Feed-Forward Network In Mixture Of Expert Models (W) | (~262K) 262144 | 2048 x 128 x 1 x 1 | F32 | +| 301 | blk.24.ffn_norm.weight | Block 24 Feed-Forward Network Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 302 | blk.24.ffn_up_exps.weight | Block 24 Ffn_Up_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_XXS | + +- Total elements in blk.24: (~623M) 623120640 +- Percentage of total elements: 2.13% + + +### Block 25 Tensor Group : ~623M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:----------------------------|:-------------------------------------------------------------------------------------------|:------------------|:----------------------|:-------| +| 303 | blk.25.attn_k.weight | Block 25 Attention Key (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 304 | blk.25.attn_k_norm.weight | Block 25 Attn_K_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 305 | blk.25.attn_norm.weight | Block 25 Attention Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 306 | blk.25.attn_output.weight | Block 25 Attention Output (W) | ( ~8M) 8388608 | 4096 x 2048 x 1 x 1 | IQ3_S | +| 307 | blk.25.attn_q.weight | Block 25 Attention Query (W) | ( ~8M) 8388608 | 2048 x 4096 x 1 x 1 | IQ3_S | +| 308 | blk.25.attn_q_norm.weight | Block 25 Attn_Q_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 309 | blk.25.attn_v.weight | Block 25 Attention Value (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 310 | blk.25.ffn_down_exps.weight | Block 25 Ffn_Down_Exps (W) | (~201M) 201326592 | 768 x 2048 x 128 x 1 | IQ4_NL | +| 311 | blk.25.ffn_gate_exps.weight | Block 25 Ffn_Gate_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_S | +| 312 | blk.25.ffn_gate_inp.weight | Block 25 Expert-Routing Layer For The Feed-Forward Network In Mixture Of Expert Models (W) | (~262K) 262144 | 2048 x 128 x 1 x 1 | F32 | +| 313 | blk.25.ffn_norm.weight | Block 25 Feed-Forward Network Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 314 | blk.25.ffn_up_exps.weight | Block 25 Ffn_Up_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_S | + +- Total elements in blk.25: (~623M) 623120640 +- Percentage of total elements: 2.13% + + +### Block 26 Tensor Group : ~623M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:----------------------------|:-------------------------------------------------------------------------------------------|:------------------|:----------------------|:-------| +| 315 | blk.26.attn_k.weight | Block 26 Attention Key (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 316 | blk.26.attn_k_norm.weight | Block 26 Attn_K_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 317 | blk.26.attn_norm.weight | Block 26 Attention Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 318 | blk.26.attn_output.weight | Block 26 Attention Output (W) | ( ~8M) 8388608 | 4096 x 2048 x 1 x 1 | IQ3_S | +| 319 | blk.26.attn_q.weight | Block 26 Attention Query (W) | ( ~8M) 8388608 | 2048 x 4096 x 1 x 1 | IQ3_S | +| 320 | blk.26.attn_q_norm.weight | Block 26 Attn_Q_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 321 | blk.26.attn_v.weight | Block 26 Attention Value (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 322 | blk.26.ffn_down_exps.weight | Block 26 Ffn_Down_Exps (W) | (~201M) 201326592 | 768 x 2048 x 128 x 1 | IQ4_NL | +| 323 | blk.26.ffn_gate_exps.weight | Block 26 Ffn_Gate_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_S | +| 324 | blk.26.ffn_gate_inp.weight | Block 26 Expert-Routing Layer For The Feed-Forward Network In Mixture Of Expert Models (W) | (~262K) 262144 | 2048 x 128 x 1 x 1 | F32 | +| 325 | blk.26.ffn_norm.weight | Block 26 Feed-Forward Network Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 326 | blk.26.ffn_up_exps.weight | Block 26 Ffn_Up_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_S | + +- Total elements in blk.26: (~623M) 623120640 +- Percentage of total elements: 2.13% + + +### Block 27 Tensor Group : ~623M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:----------------------------|:-------------------------------------------------------------------------------------------|:------------------|:----------------------|:-------| +| 327 | blk.27.attn_k.weight | Block 27 Attention Key (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 328 | blk.27.attn_k_norm.weight | Block 27 Attn_K_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 329 | blk.27.attn_norm.weight | Block 27 Attention Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 330 | blk.27.attn_output.weight | Block 27 Attention Output (W) | ( ~8M) 8388608 | 4096 x 2048 x 1 x 1 | IQ3_S | +| 331 | blk.27.attn_q.weight | Block 27 Attention Query (W) | ( ~8M) 8388608 | 2048 x 4096 x 1 x 1 | IQ3_S | +| 332 | blk.27.attn_q_norm.weight | Block 27 Attn_Q_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 333 | blk.27.attn_v.weight | Block 27 Attention Value (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 334 | blk.27.ffn_down_exps.weight | Block 27 Ffn_Down_Exps (W) | (~201M) 201326592 | 768 x 2048 x 128 x 1 | IQ4_NL | +| 335 | blk.27.ffn_gate_exps.weight | Block 27 Ffn_Gate_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_S | +| 336 | blk.27.ffn_gate_inp.weight | Block 27 Expert-Routing Layer For The Feed-Forward Network In Mixture Of Expert Models (W) | (~262K) 262144 | 2048 x 128 x 1 x 1 | F32 | +| 337 | blk.27.ffn_norm.weight | Block 27 Feed-Forward Network Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 338 | blk.27.ffn_up_exps.weight | Block 27 Ffn_Up_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_S | + +- Total elements in blk.27: (~623M) 623120640 +- Percentage of total elements: 2.13% + + +### Block 28 Tensor Group : ~623M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:----------------------------|:-------------------------------------------------------------------------------------------|:------------------|:----------------------|:-------| +| 339 | blk.28.attn_k.weight | Block 28 Attention Key (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 340 | blk.28.attn_k_norm.weight | Block 28 Attn_K_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 341 | blk.28.attn_norm.weight | Block 28 Attention Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 342 | blk.28.attn_output.weight | Block 28 Attention Output (W) | ( ~8M) 8388608 | 4096 x 2048 x 1 x 1 | IQ3_S | +| 343 | blk.28.attn_q.weight | Block 28 Attention Query (W) | ( ~8M) 8388608 | 2048 x 4096 x 1 x 1 | IQ3_S | +| 344 | blk.28.attn_q_norm.weight | Block 28 Attn_Q_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 345 | blk.28.attn_v.weight | Block 28 Attention Value (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 346 | blk.28.ffn_down_exps.weight | Block 28 Ffn_Down_Exps (W) | (~201M) 201326592 | 768 x 2048 x 128 x 1 | IQ4_NL | +| 347 | blk.28.ffn_gate_exps.weight | Block 28 Ffn_Gate_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_S | +| 348 | blk.28.ffn_gate_inp.weight | Block 28 Expert-Routing Layer For The Feed-Forward Network In Mixture Of Expert Models (W) | (~262K) 262144 | 2048 x 128 x 1 x 1 | F32 | +| 349 | blk.28.ffn_norm.weight | Block 28 Feed-Forward Network Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 350 | blk.28.ffn_up_exps.weight | Block 28 Ffn_Up_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_S | + +- Total elements in blk.28: (~623M) 623120640 +- Percentage of total elements: 2.13% + + +### Block 29 Tensor Group : ~623M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:----------------------------|:-------------------------------------------------------------------------------------------|:------------------|:----------------------|:-------| +| 351 | blk.29.attn_k.weight | Block 29 Attention Key (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 352 | blk.29.attn_k_norm.weight | Block 29 Attn_K_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 353 | blk.29.attn_norm.weight | Block 29 Attention Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 354 | blk.29.attn_output.weight | Block 29 Attention Output (W) | ( ~8M) 8388608 | 4096 x 2048 x 1 x 1 | IQ3_S | +| 355 | blk.29.attn_q.weight | Block 29 Attention Query (W) | ( ~8M) 8388608 | 2048 x 4096 x 1 x 1 | IQ3_S | +| 356 | blk.29.attn_q_norm.weight | Block 29 Attn_Q_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 357 | blk.29.attn_v.weight | Block 29 Attention Value (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 358 | blk.29.ffn_down_exps.weight | Block 29 Ffn_Down_Exps (W) | (~201M) 201326592 | 768 x 2048 x 128 x 1 | IQ4_NL | +| 359 | blk.29.ffn_gate_exps.weight | Block 29 Ffn_Gate_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_S | +| 360 | blk.29.ffn_gate_inp.weight | Block 29 Expert-Routing Layer For The Feed-Forward Network In Mixture Of Expert Models (W) | (~262K) 262144 | 2048 x 128 x 1 x 1 | F32 | +| 361 | blk.29.ffn_norm.weight | Block 29 Feed-Forward Network Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 362 | blk.29.ffn_up_exps.weight | Block 29 Ffn_Up_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_S | + +- Total elements in blk.29: (~623M) 623120640 +- Percentage of total elements: 2.13% + + +### Block 30 Tensor Group : ~623M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:----------------------------|:-------------------------------------------------------------------------------------------|:------------------|:----------------------|:-------| +| 363 | blk.30.attn_k.weight | Block 30 Attention Key (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 364 | blk.30.attn_k_norm.weight | Block 30 Attn_K_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 365 | blk.30.attn_norm.weight | Block 30 Attention Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 366 | blk.30.attn_output.weight | Block 30 Attention Output (W) | ( ~8M) 8388608 | 4096 x 2048 x 1 x 1 | IQ3_S | +| 367 | blk.30.attn_q.weight | Block 30 Attention Query (W) | ( ~8M) 8388608 | 2048 x 4096 x 1 x 1 | IQ3_S | +| 368 | blk.30.attn_q_norm.weight | Block 30 Attn_Q_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 369 | blk.30.attn_v.weight | Block 30 Attention Value (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 370 | blk.30.ffn_down_exps.weight | Block 30 Ffn_Down_Exps (W) | (~201M) 201326592 | 768 x 2048 x 128 x 1 | IQ4_NL | +| 371 | blk.30.ffn_gate_exps.weight | Block 30 Ffn_Gate_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_S | +| 372 | blk.30.ffn_gate_inp.weight | Block 30 Expert-Routing Layer For The Feed-Forward Network In Mixture Of Expert Models (W) | (~262K) 262144 | 2048 x 128 x 1 x 1 | F32 | +| 373 | blk.30.ffn_norm.weight | Block 30 Feed-Forward Network Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 374 | blk.30.ffn_up_exps.weight | Block 30 Ffn_Up_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_S | + +- Total elements in blk.30: (~623M) 623120640 +- Percentage of total elements: 2.13% + + +### Block 31 Tensor Group : ~623M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:----------------------------|:-------------------------------------------------------------------------------------------|:------------------|:----------------------|:-------| +| 375 | blk.31.attn_k.weight | Block 31 Attention Key (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 376 | blk.31.attn_k_norm.weight | Block 31 Attn_K_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 377 | blk.31.attn_norm.weight | Block 31 Attention Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 378 | blk.31.attn_output.weight | Block 31 Attention Output (W) | ( ~8M) 8388608 | 4096 x 2048 x 1 x 1 | IQ3_S | +| 379 | blk.31.attn_q.weight | Block 31 Attention Query (W) | ( ~8M) 8388608 | 2048 x 4096 x 1 x 1 | IQ3_S | +| 380 | blk.31.attn_q_norm.weight | Block 31 Attn_Q_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 381 | blk.31.attn_v.weight | Block 31 Attention Value (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 382 | blk.31.ffn_down_exps.weight | Block 31 Ffn_Down_Exps (W) | (~201M) 201326592 | 768 x 2048 x 128 x 1 | IQ4_NL | +| 383 | blk.31.ffn_gate_exps.weight | Block 31 Ffn_Gate_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_S | +| 384 | blk.31.ffn_gate_inp.weight | Block 31 Expert-Routing Layer For The Feed-Forward Network In Mixture Of Expert Models (W) | (~262K) 262144 | 2048 x 128 x 1 x 1 | F32 | +| 385 | blk.31.ffn_norm.weight | Block 31 Feed-Forward Network Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 386 | blk.31.ffn_up_exps.weight | Block 31 Ffn_Up_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_S | + +- Total elements in blk.31: (~623M) 623120640 +- Percentage of total elements: 2.13% + + +### Block 32 Tensor Group : ~623M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:----------------------------|:-------------------------------------------------------------------------------------------|:------------------|:----------------------|:-------| +| 387 | blk.32.attn_k.weight | Block 32 Attention Key (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 388 | blk.32.attn_k_norm.weight | Block 32 Attn_K_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 389 | blk.32.attn_norm.weight | Block 32 Attention Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 390 | blk.32.attn_output.weight | Block 32 Attention Output (W) | ( ~8M) 8388608 | 4096 x 2048 x 1 x 1 | IQ3_S | +| 391 | blk.32.attn_q.weight | Block 32 Attention Query (W) | ( ~8M) 8388608 | 2048 x 4096 x 1 x 1 | IQ3_S | +| 392 | blk.32.attn_q_norm.weight | Block 32 Attn_Q_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 393 | blk.32.attn_v.weight | Block 32 Attention Value (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 394 | blk.32.ffn_down_exps.weight | Block 32 Ffn_Down_Exps (W) | (~201M) 201326592 | 768 x 2048 x 128 x 1 | IQ4_NL | +| 395 | blk.32.ffn_gate_exps.weight | Block 32 Ffn_Gate_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_S | +| 396 | blk.32.ffn_gate_inp.weight | Block 32 Expert-Routing Layer For The Feed-Forward Network In Mixture Of Expert Models (W) | (~262K) 262144 | 2048 x 128 x 1 x 1 | F32 | +| 397 | blk.32.ffn_norm.weight | Block 32 Feed-Forward Network Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 398 | blk.32.ffn_up_exps.weight | Block 32 Ffn_Up_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_S | + +- Total elements in blk.32: (~623M) 623120640 +- Percentage of total elements: 2.13% + + +### Block 33 Tensor Group : ~623M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:----------------------------|:-------------------------------------------------------------------------------------------|:------------------|:----------------------|:-------| +| 399 | blk.33.attn_k.weight | Block 33 Attention Key (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 400 | blk.33.attn_k_norm.weight | Block 33 Attn_K_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 401 | blk.33.attn_norm.weight | Block 33 Attention Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 402 | blk.33.attn_output.weight | Block 33 Attention Output (W) | ( ~8M) 8388608 | 4096 x 2048 x 1 x 1 | IQ3_S | +| 403 | blk.33.attn_q.weight | Block 33 Attention Query (W) | ( ~8M) 8388608 | 2048 x 4096 x 1 x 1 | IQ3_S | +| 404 | blk.33.attn_q_norm.weight | Block 33 Attn_Q_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 405 | blk.33.attn_v.weight | Block 33 Attention Value (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 406 | blk.33.ffn_down_exps.weight | Block 33 Ffn_Down_Exps (W) | (~201M) 201326592 | 768 x 2048 x 128 x 1 | IQ4_NL | +| 407 | blk.33.ffn_gate_exps.weight | Block 33 Ffn_Gate_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_S | +| 408 | blk.33.ffn_gate_inp.weight | Block 33 Expert-Routing Layer For The Feed-Forward Network In Mixture Of Expert Models (W) | (~262K) 262144 | 2048 x 128 x 1 x 1 | F32 | +| 409 | blk.33.ffn_norm.weight | Block 33 Feed-Forward Network Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 410 | blk.33.ffn_up_exps.weight | Block 33 Ffn_Up_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_S | + +- Total elements in blk.33: (~623M) 623120640 +- Percentage of total elements: 2.13% + + +### Block 34 Tensor Group : ~623M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:----------------------------|:-------------------------------------------------------------------------------------------|:------------------|:----------------------|:-------| +| 411 | blk.34.attn_k.weight | Block 34 Attention Key (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 412 | blk.34.attn_k_norm.weight | Block 34 Attn_K_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 413 | blk.34.attn_norm.weight | Block 34 Attention Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 414 | blk.34.attn_output.weight | Block 34 Attention Output (W) | ( ~8M) 8388608 | 4096 x 2048 x 1 x 1 | IQ3_S | +| 415 | blk.34.attn_q.weight | Block 34 Attention Query (W) | ( ~8M) 8388608 | 2048 x 4096 x 1 x 1 | IQ3_S | +| 416 | blk.34.attn_q_norm.weight | Block 34 Attn_Q_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 417 | blk.34.attn_v.weight | Block 34 Attention Value (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 418 | blk.34.ffn_down_exps.weight | Block 34 Ffn_Down_Exps (W) | (~201M) 201326592 | 768 x 2048 x 128 x 1 | IQ4_NL | +| 419 | blk.34.ffn_gate_exps.weight | Block 34 Ffn_Gate_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_S | +| 420 | blk.34.ffn_gate_inp.weight | Block 34 Expert-Routing Layer For The Feed-Forward Network In Mixture Of Expert Models (W) | (~262K) 262144 | 2048 x 128 x 1 x 1 | F32 | +| 421 | blk.34.ffn_norm.weight | Block 34 Feed-Forward Network Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 422 | blk.34.ffn_up_exps.weight | Block 34 Ffn_Up_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_S | + +- Total elements in blk.34: (~623M) 623120640 +- Percentage of total elements: 2.13% + + +### Block 35 Tensor Group : ~623M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:----------------------------|:-------------------------------------------------------------------------------------------|:------------------|:----------------------|:-------| +| 423 | blk.35.attn_k.weight | Block 35 Attention Key (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 424 | blk.35.attn_k_norm.weight | Block 35 Attn_K_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 425 | blk.35.attn_norm.weight | Block 35 Attention Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 426 | blk.35.attn_output.weight | Block 35 Attention Output (W) | ( ~8M) 8388608 | 4096 x 2048 x 1 x 1 | IQ3_S | +| 427 | blk.35.attn_q.weight | Block 35 Attention Query (W) | ( ~8M) 8388608 | 2048 x 4096 x 1 x 1 | IQ3_S | +| 428 | blk.35.attn_q_norm.weight | Block 35 Attn_Q_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 429 | blk.35.attn_v.weight | Block 35 Attention Value (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 430 | blk.35.ffn_down_exps.weight | Block 35 Ffn_Down_Exps (W) | (~201M) 201326592 | 768 x 2048 x 128 x 1 | IQ4_NL | +| 431 | blk.35.ffn_gate_exps.weight | Block 35 Ffn_Gate_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_S | +| 432 | blk.35.ffn_gate_inp.weight | Block 35 Expert-Routing Layer For The Feed-Forward Network In Mixture Of Expert Models (W) | (~262K) 262144 | 2048 x 128 x 1 x 1 | F32 | +| 433 | blk.35.ffn_norm.weight | Block 35 Feed-Forward Network Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 434 | blk.35.ffn_up_exps.weight | Block 35 Ffn_Up_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_S | + +- Total elements in blk.35: (~623M) 623120640 +- Percentage of total elements: 2.13% + + +### Block 36 Tensor Group : ~623M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:----------------------------|:-------------------------------------------------------------------------------------------|:------------------|:----------------------|:-------| +| 435 | blk.36.attn_k.weight | Block 36 Attention Key (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 436 | blk.36.attn_k_norm.weight | Block 36 Attn_K_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 437 | blk.36.attn_norm.weight | Block 36 Attention Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 438 | blk.36.attn_output.weight | Block 36 Attention Output (W) | ( ~8M) 8388608 | 4096 x 2048 x 1 x 1 | IQ3_S | +| 439 | blk.36.attn_q.weight | Block 36 Attention Query (W) | ( ~8M) 8388608 | 2048 x 4096 x 1 x 1 | IQ3_S | +| 440 | blk.36.attn_q_norm.weight | Block 36 Attn_Q_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 441 | blk.36.attn_v.weight | Block 36 Attention Value (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 442 | blk.36.ffn_down_exps.weight | Block 36 Ffn_Down_Exps (W) | (~201M) 201326592 | 768 x 2048 x 128 x 1 | IQ4_NL | +| 443 | blk.36.ffn_gate_exps.weight | Block 36 Ffn_Gate_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_S | +| 444 | blk.36.ffn_gate_inp.weight | Block 36 Expert-Routing Layer For The Feed-Forward Network In Mixture Of Expert Models (W) | (~262K) 262144 | 2048 x 128 x 1 x 1 | F32 | +| 445 | blk.36.ffn_norm.weight | Block 36 Feed-Forward Network Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 446 | blk.36.ffn_up_exps.weight | Block 36 Ffn_Up_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_S | + +- Total elements in blk.36: (~623M) 623120640 +- Percentage of total elements: 2.13% + + +### Block 37 Tensor Group : ~623M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:----------------------------|:-------------------------------------------------------------------------------------------|:------------------|:----------------------|:-------| +| 447 | blk.37.attn_k.weight | Block 37 Attention Key (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 448 | blk.37.attn_k_norm.weight | Block 37 Attn_K_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 449 | blk.37.attn_norm.weight | Block 37 Attention Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 450 | blk.37.attn_output.weight | Block 37 Attention Output (W) | ( ~8M) 8388608 | 4096 x 2048 x 1 x 1 | IQ3_S | +| 451 | blk.37.attn_q.weight | Block 37 Attention Query (W) | ( ~8M) 8388608 | 2048 x 4096 x 1 x 1 | IQ3_S | +| 452 | blk.37.attn_q_norm.weight | Block 37 Attn_Q_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 453 | blk.37.attn_v.weight | Block 37 Attention Value (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 454 | blk.37.ffn_down_exps.weight | Block 37 Ffn_Down_Exps (W) | (~201M) 201326592 | 768 x 2048 x 128 x 1 | IQ4_NL | +| 455 | blk.37.ffn_gate_exps.weight | Block 37 Ffn_Gate_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_S | +| 456 | blk.37.ffn_gate_inp.weight | Block 37 Expert-Routing Layer For The Feed-Forward Network In Mixture Of Expert Models (W) | (~262K) 262144 | 2048 x 128 x 1 x 1 | F32 | +| 457 | blk.37.ffn_norm.weight | Block 37 Feed-Forward Network Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 458 | blk.37.ffn_up_exps.weight | Block 37 Ffn_Up_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_S | + +- Total elements in blk.37: (~623M) 623120640 +- Percentage of total elements: 2.13% + + +### Block 38 Tensor Group : ~623M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:----------------------------|:-------------------------------------------------------------------------------------------|:------------------|:----------------------|:-------| +| 459 | blk.38.attn_k.weight | Block 38 Attention Key (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 460 | blk.38.attn_k_norm.weight | Block 38 Attn_K_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 461 | blk.38.attn_norm.weight | Block 38 Attention Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 462 | blk.38.attn_output.weight | Block 38 Attention Output (W) | ( ~8M) 8388608 | 4096 x 2048 x 1 x 1 | IQ3_S | +| 463 | blk.38.attn_q.weight | Block 38 Attention Query (W) | ( ~8M) 8388608 | 2048 x 4096 x 1 x 1 | IQ3_S | +| 464 | blk.38.attn_q_norm.weight | Block 38 Attn_Q_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 465 | blk.38.attn_v.weight | Block 38 Attention Value (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 466 | blk.38.ffn_down_exps.weight | Block 38 Ffn_Down_Exps (W) | (~201M) 201326592 | 768 x 2048 x 128 x 1 | IQ4_NL | +| 467 | blk.38.ffn_gate_exps.weight | Block 38 Ffn_Gate_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_S | +| 468 | blk.38.ffn_gate_inp.weight | Block 38 Expert-Routing Layer For The Feed-Forward Network In Mixture Of Expert Models (W) | (~262K) 262144 | 2048 x 128 x 1 x 1 | F32 | +| 469 | blk.38.ffn_norm.weight | Block 38 Feed-Forward Network Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 470 | blk.38.ffn_up_exps.weight | Block 38 Ffn_Up_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_S | + +- Total elements in blk.38: (~623M) 623120640 +- Percentage of total elements: 2.13% + + +### Block 39 Tensor Group : ~623M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:----------------------------|:-------------------------------------------------------------------------------------------|:------------------|:----------------------|:-------| +| 471 | blk.39.attn_k.weight | Block 39 Attention Key (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 472 | blk.39.attn_k_norm.weight | Block 39 Attn_K_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 473 | blk.39.attn_norm.weight | Block 39 Attention Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 474 | blk.39.attn_output.weight | Block 39 Attention Output (W) | ( ~8M) 8388608 | 4096 x 2048 x 1 x 1 | IQ3_S | +| 475 | blk.39.attn_q.weight | Block 39 Attention Query (W) | ( ~8M) 8388608 | 2048 x 4096 x 1 x 1 | IQ3_S | +| 476 | blk.39.attn_q_norm.weight | Block 39 Attn_Q_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 477 | blk.39.attn_v.weight | Block 39 Attention Value (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 478 | blk.39.ffn_down_exps.weight | Block 39 Ffn_Down_Exps (W) | (~201M) 201326592 | 768 x 2048 x 128 x 1 | IQ4_NL | +| 479 | blk.39.ffn_gate_exps.weight | Block 39 Ffn_Gate_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_S | +| 480 | blk.39.ffn_gate_inp.weight | Block 39 Expert-Routing Layer For The Feed-Forward Network In Mixture Of Expert Models (W) | (~262K) 262144 | 2048 x 128 x 1 x 1 | F32 | +| 481 | blk.39.ffn_norm.weight | Block 39 Feed-Forward Network Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 482 | blk.39.ffn_up_exps.weight | Block 39 Ffn_Up_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_S | + +- Total elements in blk.39: (~623M) 623120640 +- Percentage of total elements: 2.13% + + +### Block 40 Tensor Group : ~623M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:----------------------------|:-------------------------------------------------------------------------------------------|:------------------|:----------------------|:-------| +| 483 | blk.40.attn_k.weight | Block 40 Attention Key (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 484 | blk.40.attn_k_norm.weight | Block 40 Attn_K_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 485 | blk.40.attn_norm.weight | Block 40 Attention Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 486 | blk.40.attn_output.weight | Block 40 Attention Output (W) | ( ~8M) 8388608 | 4096 x 2048 x 1 x 1 | IQ3_S | +| 487 | blk.40.attn_q.weight | Block 40 Attention Query (W) | ( ~8M) 8388608 | 2048 x 4096 x 1 x 1 | IQ3_S | +| 488 | blk.40.attn_q_norm.weight | Block 40 Attn_Q_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 489 | blk.40.attn_v.weight | Block 40 Attention Value (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 490 | blk.40.ffn_down_exps.weight | Block 40 Ffn_Down_Exps (W) | (~201M) 201326592 | 768 x 2048 x 128 x 1 | IQ4_NL | +| 491 | blk.40.ffn_gate_exps.weight | Block 40 Ffn_Gate_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_S | +| 492 | blk.40.ffn_gate_inp.weight | Block 40 Expert-Routing Layer For The Feed-Forward Network In Mixture Of Expert Models (W) | (~262K) 262144 | 2048 x 128 x 1 x 1 | F32 | +| 493 | blk.40.ffn_norm.weight | Block 40 Feed-Forward Network Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 494 | blk.40.ffn_up_exps.weight | Block 40 Ffn_Up_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_S | + +- Total elements in blk.40: (~623M) 623120640 +- Percentage of total elements: 2.13% + + +### Block 41 Tensor Group : ~623M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:----------------------------|:-------------------------------------------------------------------------------------------|:------------------|:----------------------|:-------| +| 495 | blk.41.attn_k.weight | Block 41 Attention Key (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 496 | blk.41.attn_k_norm.weight | Block 41 Attn_K_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 497 | blk.41.attn_norm.weight | Block 41 Attention Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 498 | blk.41.attn_output.weight | Block 41 Attention Output (W) | ( ~8M) 8388608 | 4096 x 2048 x 1 x 1 | IQ3_S | +| 499 | blk.41.attn_q.weight | Block 41 Attention Query (W) | ( ~8M) 8388608 | 2048 x 4096 x 1 x 1 | IQ3_S | +| 500 | blk.41.attn_q_norm.weight | Block 41 Attn_Q_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 501 | blk.41.attn_v.weight | Block 41 Attention Value (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 502 | blk.41.ffn_down_exps.weight | Block 41 Ffn_Down_Exps (W) | (~201M) 201326592 | 768 x 2048 x 128 x 1 | IQ4_NL | +| 503 | blk.41.ffn_gate_exps.weight | Block 41 Ffn_Gate_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_S | +| 504 | blk.41.ffn_gate_inp.weight | Block 41 Expert-Routing Layer For The Feed-Forward Network In Mixture Of Expert Models (W) | (~262K) 262144 | 2048 x 128 x 1 x 1 | F32 | +| 505 | blk.41.ffn_norm.weight | Block 41 Feed-Forward Network Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 506 | blk.41.ffn_up_exps.weight | Block 41 Ffn_Up_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_S | + +- Total elements in blk.41: (~623M) 623120640 +- Percentage of total elements: 2.13% + + +### Block 42 Tensor Group : ~623M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:----------------------------|:-------------------------------------------------------------------------------------------|:------------------|:----------------------|:-------| +| 507 | blk.42.attn_k.weight | Block 42 Attention Key (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 508 | blk.42.attn_k_norm.weight | Block 42 Attn_K_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 509 | blk.42.attn_norm.weight | Block 42 Attention Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 510 | blk.42.attn_output.weight | Block 42 Attention Output (W) | ( ~8M) 8388608 | 4096 x 2048 x 1 x 1 | IQ3_S | +| 511 | blk.42.attn_q.weight | Block 42 Attention Query (W) | ( ~8M) 8388608 | 2048 x 4096 x 1 x 1 | IQ3_S | +| 512 | blk.42.attn_q_norm.weight | Block 42 Attn_Q_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 513 | blk.42.attn_v.weight | Block 42 Attention Value (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 514 | blk.42.ffn_down_exps.weight | Block 42 Ffn_Down_Exps (W) | (~201M) 201326592 | 768 x 2048 x 128 x 1 | IQ4_NL | +| 515 | blk.42.ffn_gate_exps.weight | Block 42 Ffn_Gate_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_S | +| 516 | blk.42.ffn_gate_inp.weight | Block 42 Expert-Routing Layer For The Feed-Forward Network In Mixture Of Expert Models (W) | (~262K) 262144 | 2048 x 128 x 1 x 1 | F32 | +| 517 | blk.42.ffn_norm.weight | Block 42 Feed-Forward Network Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 518 | blk.42.ffn_up_exps.weight | Block 42 Ffn_Up_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_S | + +- Total elements in blk.42: (~623M) 623120640 +- Percentage of total elements: 2.13% + + +### Block 43 Tensor Group : ~623M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:----------------------------|:-------------------------------------------------------------------------------------------|:------------------|:----------------------|:-------| +| 519 | blk.43.attn_k.weight | Block 43 Attention Key (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 520 | blk.43.attn_k_norm.weight | Block 43 Attn_K_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 521 | blk.43.attn_norm.weight | Block 43 Attention Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 522 | blk.43.attn_output.weight | Block 43 Attention Output (W) | ( ~8M) 8388608 | 4096 x 2048 x 1 x 1 | IQ3_S | +| 523 | blk.43.attn_q.weight | Block 43 Attention Query (W) | ( ~8M) 8388608 | 2048 x 4096 x 1 x 1 | IQ3_S | +| 524 | blk.43.attn_q_norm.weight | Block 43 Attn_Q_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 525 | blk.43.attn_v.weight | Block 43 Attention Value (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 526 | blk.43.ffn_down_exps.weight | Block 43 Ffn_Down_Exps (W) | (~201M) 201326592 | 768 x 2048 x 128 x 1 | IQ4_NL | +| 527 | blk.43.ffn_gate_exps.weight | Block 43 Ffn_Gate_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_S | +| 528 | blk.43.ffn_gate_inp.weight | Block 43 Expert-Routing Layer For The Feed-Forward Network In Mixture Of Expert Models (W) | (~262K) 262144 | 2048 x 128 x 1 x 1 | F32 | +| 529 | blk.43.ffn_norm.weight | Block 43 Feed-Forward Network Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 530 | blk.43.ffn_up_exps.weight | Block 43 Ffn_Up_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_S | + +- Total elements in blk.43: (~623M) 623120640 +- Percentage of total elements: 2.13% + + +### Block 44 Tensor Group : ~623M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:----------------------------|:-------------------------------------------------------------------------------------------|:------------------|:----------------------|:-------| +| 531 | blk.44.attn_k.weight | Block 44 Attention Key (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 532 | blk.44.attn_k_norm.weight | Block 44 Attn_K_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 533 | blk.44.attn_norm.weight | Block 44 Attention Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 534 | blk.44.attn_output.weight | Block 44 Attention Output (W) | ( ~8M) 8388608 | 4096 x 2048 x 1 x 1 | IQ3_S | +| 535 | blk.44.attn_q.weight | Block 44 Attention Query (W) | ( ~8M) 8388608 | 2048 x 4096 x 1 x 1 | IQ3_S | +| 536 | blk.44.attn_q_norm.weight | Block 44 Attn_Q_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 537 | blk.44.attn_v.weight | Block 44 Attention Value (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 538 | blk.44.ffn_down_exps.weight | Block 44 Ffn_Down_Exps (W) | (~201M) 201326592 | 768 x 2048 x 128 x 1 | IQ4_NL | +| 539 | blk.44.ffn_gate_exps.weight | Block 44 Ffn_Gate_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_S | +| 540 | blk.44.ffn_gate_inp.weight | Block 44 Expert-Routing Layer For The Feed-Forward Network In Mixture Of Expert Models (W) | (~262K) 262144 | 2048 x 128 x 1 x 1 | F32 | +| 541 | blk.44.ffn_norm.weight | Block 44 Feed-Forward Network Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 542 | blk.44.ffn_up_exps.weight | Block 44 Ffn_Up_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_S | + +- Total elements in blk.44: (~623M) 623120640 +- Percentage of total elements: 2.13% + + +### Block 45 Tensor Group : ~623M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:----------------------------|:-------------------------------------------------------------------------------------------|:------------------|:----------------------|:-------| +| 543 | blk.45.attn_k.weight | Block 45 Attention Key (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 544 | blk.45.attn_k_norm.weight | Block 45 Attn_K_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 545 | blk.45.attn_norm.weight | Block 45 Attention Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 546 | blk.45.attn_output.weight | Block 45 Attention Output (W) | ( ~8M) 8388608 | 4096 x 2048 x 1 x 1 | IQ3_S | +| 547 | blk.45.attn_q.weight | Block 45 Attention Query (W) | ( ~8M) 8388608 | 2048 x 4096 x 1 x 1 | IQ3_S | +| 548 | blk.45.attn_q_norm.weight | Block 45 Attn_Q_Norm (W) | ( 128) 128 | 128 x 1 x 1 x 1 | F32 | +| 549 | blk.45.attn_v.weight | Block 45 Attention Value (W) | ( ~1M) 1048576 | 2048 x 512 x 1 x 1 | IQ3_S | +| 550 | blk.45.ffn_down_exps.weight | Block 45 Ffn_Down_Exps (W) | (~201M) 201326592 | 768 x 2048 x 128 x 1 | IQ4_NL | +| 551 | blk.45.ffn_gate_exps.weight | Block 45 Ffn_Gate_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_S | +| 552 | blk.45.ffn_gate_inp.weight | Block 45 Expert-Routing Layer For The Feed-Forward Network In Mixture Of Expert Models (W) | (~262K) 262144 | 2048 x 128 x 1 x 1 | F32 | +| 553 | blk.45.ffn_norm.weight | Block 45 Feed-Forward Network Normalization (W) | ( ~2K) 2048 | 2048 x 1 x 1 x 1 | F32 | +| 554 | blk.45.ffn_up_exps.weight | Block 45 Ffn_Up_Exps (W) | (~201M) 201326592 | 2048 x 768 x 128 x 1 | IQ3_S | + +- Total elements in blk.45: (~623M) 623120640 +- Percentage of total elements: 2.13% + + +