| llama_print_build_info: build = 1 (62061f910) |
| llama_print_build_info: built with GNU 13.3.0 for Linux x86_64 |
| llama_quantize: quantizing '/root/qwen38_bf16/BF16/Qwen3.8-27B-BF16-00001-of-00002.gguf' to '/mnt/geth-vol1/qwen38_out/Qwen3.8-27B-IQ1_M.gguf.part' as IQ1_M using 16 threads |
| llama_model_loader: additional 1 GGUFs metadata loaded. |
| llama_model_loader: loaded meta data with 50 key-value pairs and 866 tensors from /root/qwen38_bf16/BF16/Qwen3.8-27B-BF16-00001-of-00002.gguf (version GGUF V3 (latest)) |
| llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output. |
| llama_model_loader: - kv 0: general.architecture str = qwen35 |
| llama_model_loader: - kv 1: general.type str = model |
| llama_model_loader: - kv 2: general.sampling.top_k i32 = 20 |
| llama_model_loader: - kv 3: general.sampling.top_p f32 = 0.950000 |
| llama_model_loader: - kv 4: general.sampling.temp f32 = 1.000000 |
| llama_model_loader: - kv 5: general.name str = Qwen3.8-27B |
| llama_model_loader: - kv 6: general.basename str = Qwen3.8-27B |
| llama_model_loader: - kv 7: general.description str = Renewal of the beloved Qwen model, de... |
| llama_model_loader: - kv 8: general.quantized_by str = Unsloth |
| llama_model_loader: - kv 9: general.size_label str = 27B |
| llama_model_loader: - kv 10: general.license str = apache-2.0 |
| llama_model_loader: - kv 11: general.repo_url str = https://huggingface.co/unsloth |
| llama_model_loader: - kv 12: general.base_model.count u32 = 1 |
| llama_model_loader: - kv 13: general.base_model.0.name str = Qwen3.8 27B |
| llama_model_loader: - kv 14: general.base_model.0.organization str = Qwen |
| llama_model_loader: - kv 15: general.base_model.0.repo_url str = https://huggingface.co/Qwen/Qwen3.8-27B |
| llama_model_loader: - kv 16: general.tags arr[str,1] = ["unsloth"] |
| llama_model_loader: - kv 17: qwen35.block_count u32 = 65 |
| llama_model_loader: - kv 18: qwen35.context_length u32 = 262144 |
| llama_model_loader: - kv 19: qwen35.embedding_length u32 = 5120 |
| llama_model_loader: - kv 20: qwen35.feed_forward_length u32 = 17408 |
| llama_model_loader: - kv 21: qwen35.attention.head_count u32 = 24 |
| llama_model_loader: - kv 22: qwen35.attention.head_count_kv u32 = 4 |
| llama_model_loader: - kv 23: qwen35.rope.dimension_sections arr[i32,4] = [11, 11, 10, 0] |
| llama_model_loader: - kv 24: qwen35.rope.freq_base f32 = 10000000.000000 |
| llama_model_loader: - kv 25: qwen35.attention.layer_norm_rms_epsilon f32 = 0.000001 |
| llama_model_loader: - kv 26: qwen35.attention.key_length u32 = 256 |
| llama_model_loader: - kv 27: qwen35.attention.value_length u32 = 256 |
| llama_model_loader: - kv 28: general.file_type u32 = 32 |
| llama_model_loader: - kv 29: qwen35.nextn_predict_layers u32 = 1 |
| llama_model_loader: - kv 30: qwen35.ssm.conv_kernel u32 = 4 |
| llama_model_loader: - kv 31: qwen35.ssm.state_size u32 = 128 |
| llama_model_loader: - kv 32: qwen35.ssm.group_count u32 = 16 |
| llama_model_loader: - kv 33: qwen35.ssm.time_step_rank u32 = 48 |
| llama_model_loader: - kv 34: qwen35.ssm.inner_size u32 = 6144 |
| llama_model_loader: - kv 35: qwen35.full_attention_interval u32 = 4 |
| llama_model_loader: - kv 36: qwen35.rope.dimension_count u32 = 64 |
| llama_model_loader: - kv 37: general.quantization_version u32 = 2 |
| llama_model_loader: - kv 38: tokenizer.ggml.model str = gpt2 |
| llama_model_loader: - kv 39: tokenizer.ggml.pre str = qwen35 |
| llama_model_loader: - kv 40: tokenizer.ggml.tokens arr[str,248320] = ["!", "\"", "#", "$", "%", "&", "'", ... |
| llama_model_loader: - kv 41: tokenizer.ggml.token_type arr[i32,248320] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ... |
| llama_model_loader: - kv 42: tokenizer.ggml.merges arr[str,247587] = ["Ġ Ġ", "ĠĠ ĠĠ", "i n", "Ġ t",... |
| llama_model_loader: - kv 43: tokenizer.ggml.eos_token_id u32 = 248046 |
| llama_model_loader: - kv 44: tokenizer.ggml.padding_token_id u32 = 248055 |
| llama_model_loader: - kv 45: tokenizer.ggml.bos_token_id u32 = 248044 |
| llama_model_loader: - kv 46: split.no u16 = 0 |
| llama_model_loader: - kv 47: split.count u16 = 2 |
| llama_model_loader: - kv 48: split.tensors.count i32 = 866 |
| llama_model_loader: - kv 49: tokenizer.chat_template str = {%- set image_count = namespace(value... |
| llama_model_loader: - type f32: 360 tensors |
| llama_model_loader: - type bf16: 506 tensors |
|
|
| llama_model_quantize_impl: have importance matrix data with 496 entries |
| llama_tensor_get_type: blk.64.attn_k.weight - applying manual override: iq1_m -> q4_K |
| llama_tensor_get_type: blk.64.attn_output.weight - applying manual override: iq1_m -> q4_K |
| llama_tensor_get_type: blk.64.attn_q.weight - applying manual override: iq1_m -> q4_K |
| llama_tensor_get_type: blk.64.attn_v.weight - applying manual override: iq1_m -> q4_K |
| llama_tensor_get_type: blk.64.ffn_down.weight - applying manual override: iq1_m -> q4_K |
| llama_tensor_get_type: blk.64.ffn_gate.weight - applying manual override: iq1_m -> q4_K |
| llama_tensor_get_type: blk.64.ffn_up.weight - applying manual override: iq1_m -> q4_K |
| llama_tensor_get_type: blk.64.nextn.eh_proj.weight - applying manual override: iq1_m -> q4_K |
| [ 1/ 866] output.weight - [ 5120, 248320, 1, 1], type = bf16, |
| ====== llama_model_quantize_impl: did not find weights for output.weight |
| converting to q5_K .. load_imatrix: imatrix datasets=['/Volumes/Lexar/aq_ds4flash/imatrix_agentic.txt'] |
| load_imatrix: loaded 496 importance matrix entries from /mnt/geth-vol1/qwen38_assets/Qwen3.8-27B-agentic.imatrix computed on 1580 chunks |
| prepare_imatrix: have 496 importance matrix entries |
| size = 2425.00 MiB -> 833.59 MiB |
| [ 2/ 866] output_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 3/ 866] token_embd.weight - [ 5120, 248320, 1, 1], type = bf16, |
| ====== llama_model_quantize_impl: did not find weights for token_embd.weight |
| converting to q2_K .. size = 2425.00 MiB -> 397.85 MiB |
| [ 4/ 866] blk.0.attn_gate.weight - [ 5120, 6144, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 5/ 866] blk.0.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 6/ 866] blk.0.attn_qkv.weight - [ 5120, 10240, 1, 1], type = bf16, converting to q4_K .. size = 100.00 MiB -> 28.12 MiB |
| [ 7/ 866] blk.0.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to q2_K .. size = 170.00 MiB -> 27.89 MiB |
| [ 8/ 866] blk.0.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 9/ 866] blk.0.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 10/ 866] blk.0.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 11/ 866] blk.0.ssm_a - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 12/ 866] blk.0.ssm_alpha.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 13/ 866] blk.0.ssm_beta.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 14/ 866] blk.0.ssm_conv1d.weight - [ 4, 10240, 1, 1], type = f32, size = 0.156 MiB |
| [ 15/ 866] blk.0.ssm_dt.bias - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 16/ 866] blk.0.ssm_norm.weight - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 17/ 866] blk.0.ssm_out.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 18/ 866] blk.1.attn_gate.weight - [ 5120, 6144, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 19/ 866] blk.1.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 20/ 866] blk.1.attn_qkv.weight - [ 5120, 10240, 1, 1], type = bf16, converting to q4_K .. size = 100.00 MiB -> 28.12 MiB |
| [ 21/ 866] blk.1.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to q2_K .. size = 170.00 MiB -> 27.89 MiB |
| [ 22/ 866] blk.1.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 23/ 866] blk.1.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 24/ 866] blk.1.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 25/ 866] blk.1.ssm_a - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 26/ 866] blk.1.ssm_alpha.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 27/ 866] blk.1.ssm_beta.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 28/ 866] blk.1.ssm_conv1d.weight - [ 4, 10240, 1, 1], type = f32, size = 0.156 MiB |
| [ 29/ 866] blk.1.ssm_dt.bias - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 30/ 866] blk.1.ssm_norm.weight - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 31/ 866] blk.1.ssm_out.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 32/ 866] blk.2.attn_gate.weight - [ 5120, 6144, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 33/ 866] blk.2.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 34/ 866] blk.2.attn_qkv.weight - [ 5120, 10240, 1, 1], type = bf16, converting to q4_K .. size = 100.00 MiB -> 28.12 MiB |
| [ 35/ 866] blk.2.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to q2_K .. size = 170.00 MiB -> 27.89 MiB |
| [ 36/ 866] blk.2.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 37/ 866] blk.2.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 38/ 866] blk.2.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 39/ 866] blk.2.ssm_a - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 40/ 866] blk.2.ssm_alpha.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 41/ 866] blk.2.ssm_beta.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 42/ 866] blk.2.ssm_conv1d.weight - [ 4, 10240, 1, 1], type = f32, size = 0.156 MiB |
| [ 43/ 866] blk.2.ssm_dt.bias - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 44/ 866] blk.2.ssm_norm.weight - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 45/ 866] blk.2.ssm_out.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 46/ 866] blk.3.attn_k.weight - [ 5120, 1024, 1, 1], type = bf16, converting to iq1_m .. size = 10.00 MiB -> 1.09 MiB |
| [ 47/ 866] blk.3.attn_k_norm.weight - [ 256, 1, 1, 1], type = f32, size = 0.001 MiB |
| [ 48/ 866] blk.3.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 49/ 866] blk.3.attn_output.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq2_xxs .. size = 60.00 MiB -> 7.73 MiB |
| [ 50/ 866] blk.3.attn_q.weight - [ 5120, 12288, 1, 1], type = bf16, converting to iq1_m .. size = 120.00 MiB -> 13.12 MiB |
| [ 51/ 866] blk.3.attn_q_norm.weight - [ 256, 1, 1, 1], type = f32, size = 0.001 MiB |
| [ 52/ 866] blk.3.attn_v.weight - [ 5120, 1024, 1, 1], type = bf16, converting to q4_K .. size = 10.00 MiB -> 2.81 MiB |
| [ 53/ 866] blk.3.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to q2_K .. size = 170.00 MiB -> 27.89 MiB |
| [ 54/ 866] blk.3.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 55/ 866] blk.3.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 56/ 866] blk.3.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 57/ 866] blk.4.attn_gate.weight - [ 5120, 6144, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 58/ 866] blk.4.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 59/ 866] blk.4.attn_qkv.weight - [ 5120, 10240, 1, 1], type = bf16, converting to q4_K .. size = 100.00 MiB -> 28.12 MiB |
| [ 60/ 866] blk.4.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to q2_K .. size = 170.00 MiB -> 27.89 MiB |
| [ 61/ 866] blk.4.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 62/ 866] blk.4.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 63/ 866] blk.4.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 64/ 866] blk.4.ssm_a - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 65/ 866] blk.4.ssm_alpha.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 66/ 866] blk.4.ssm_beta.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 67/ 866] blk.4.ssm_conv1d.weight - [ 4, 10240, 1, 1], type = f32, size = 0.156 MiB |
| [ 68/ 866] blk.4.ssm_dt.bias - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 69/ 866] blk.4.ssm_norm.weight - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 70/ 866] blk.4.ssm_out.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 71/ 866] blk.5.attn_gate.weight - [ 5120, 6144, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 72/ 866] blk.5.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 73/ 866] blk.5.attn_qkv.weight - [ 5120, 10240, 1, 1], type = bf16, converting to q4_K .. size = 100.00 MiB -> 28.12 MiB |
| [ 74/ 866] blk.5.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to q2_K .. size = 170.00 MiB -> 27.89 MiB |
| [ 75/ 866] blk.5.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 76/ 866] blk.5.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 77/ 866] blk.5.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 78/ 866] blk.5.ssm_a - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 79/ 866] blk.5.ssm_alpha.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 80/ 866] blk.5.ssm_beta.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 81/ 866] blk.5.ssm_conv1d.weight - [ 4, 10240, 1, 1], type = f32, size = 0.156 MiB |
| [ 82/ 866] blk.5.ssm_dt.bias - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 83/ 866] blk.5.ssm_norm.weight - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 84/ 866] blk.5.ssm_out.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 85/ 866] blk.6.attn_gate.weight - [ 5120, 6144, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 86/ 866] blk.6.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 87/ 866] blk.6.attn_qkv.weight - [ 5120, 10240, 1, 1], type = bf16, converting to q4_K .. size = 100.00 MiB -> 28.12 MiB |
| [ 88/ 866] blk.6.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to q2_K .. size = 170.00 MiB -> 27.89 MiB |
| [ 89/ 866] blk.6.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 90/ 866] blk.6.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 91/ 866] blk.6.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 92/ 866] blk.6.ssm_a - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 93/ 866] blk.6.ssm_alpha.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 94/ 866] blk.6.ssm_beta.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 95/ 866] blk.6.ssm_conv1d.weight - [ 4, 10240, 1, 1], type = f32, size = 0.156 MiB |
| [ 96/ 866] blk.6.ssm_dt.bias - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 97/ 866] blk.6.ssm_norm.weight - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 98/ 866] blk.6.ssm_out.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 99/ 866] blk.7.attn_k.weight - [ 5120, 1024, 1, 1], type = bf16, converting to iq1_m .. size = 10.00 MiB -> 1.09 MiB |
| [ 100/ 866] blk.7.attn_k_norm.weight - [ 256, 1, 1, 1], type = f32, size = 0.001 MiB |
| [ 101/ 866] blk.7.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 102/ 866] blk.7.attn_output.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq2_xxs .. size = 60.00 MiB -> 7.73 MiB |
| [ 103/ 866] blk.7.attn_q.weight - [ 5120, 12288, 1, 1], type = bf16, converting to iq1_m .. size = 120.00 MiB -> 13.12 MiB |
| [ 104/ 866] blk.7.attn_q_norm.weight - [ 256, 1, 1, 1], type = f32, size = 0.001 MiB |
| [ 105/ 866] blk.7.attn_v.weight - [ 5120, 1024, 1, 1], type = bf16, converting to q4_K .. size = 10.00 MiB -> 2.81 MiB |
| [ 106/ 866] blk.7.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to q2_K .. size = 170.00 MiB -> 27.89 MiB |
| [ 107/ 866] blk.7.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 108/ 866] blk.7.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 109/ 866] blk.7.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 110/ 866] blk.8.attn_gate.weight - [ 5120, 6144, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 111/ 866] blk.8.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 112/ 866] blk.8.attn_qkv.weight - [ 5120, 10240, 1, 1], type = bf16, converting to q4_K .. size = 100.00 MiB -> 28.12 MiB |
| [ 113/ 866] blk.8.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 114/ 866] blk.8.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 115/ 866] blk.8.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 116/ 866] blk.8.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 117/ 866] blk.8.ssm_a - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 118/ 866] blk.8.ssm_alpha.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 119/ 866] blk.8.ssm_beta.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 120/ 866] blk.8.ssm_conv1d.weight - [ 4, 10240, 1, 1], type = f32, size = 0.156 MiB |
| [ 121/ 866] blk.8.ssm_dt.bias - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 122/ 866] blk.8.ssm_norm.weight - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 123/ 866] blk.8.ssm_out.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 124/ 866] blk.9.attn_gate.weight - [ 5120, 6144, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 125/ 866] blk.9.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 126/ 866] blk.9.attn_qkv.weight - [ 5120, 10240, 1, 1], type = bf16, converting to q4_K .. size = 100.00 MiB -> 28.12 MiB |
| [ 127/ 866] blk.9.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 128/ 866] blk.9.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 129/ 866] blk.9.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 130/ 866] blk.9.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 131/ 866] blk.9.ssm_a - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 132/ 866] blk.9.ssm_alpha.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 133/ 866] blk.9.ssm_beta.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 134/ 866] blk.9.ssm_conv1d.weight - [ 4, 10240, 1, 1], type = f32, size = 0.156 MiB |
| [ 135/ 866] blk.9.ssm_dt.bias - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 136/ 866] blk.9.ssm_norm.weight - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 137/ 866] blk.9.ssm_out.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 138/ 866] blk.10.attn_gate.weight - [ 5120, 6144, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 139/ 866] blk.10.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 140/ 866] blk.10.attn_qkv.weight - [ 5120, 10240, 1, 1], type = bf16, converting to q4_K .. size = 100.00 MiB -> 28.12 MiB |
| [ 141/ 866] blk.10.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 142/ 866] blk.10.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 143/ 866] blk.10.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 144/ 866] blk.10.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 145/ 866] blk.10.ssm_a - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 146/ 866] blk.10.ssm_alpha.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 147/ 866] blk.10.ssm_beta.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 148/ 866] blk.10.ssm_conv1d.weight - [ 4, 10240, 1, 1], type = f32, size = 0.156 MiB |
| [ 149/ 866] blk.10.ssm_dt.bias - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 150/ 866] blk.10.ssm_norm.weight - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 151/ 866] blk.10.ssm_out.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 152/ 866] blk.11.attn_k.weight - [ 5120, 1024, 1, 1], type = bf16, converting to iq1_m .. size = 10.00 MiB -> 1.09 MiB |
| [ 153/ 866] blk.11.attn_k_norm.weight - [ 256, 1, 1, 1], type = f32, size = 0.001 MiB |
| [ 154/ 866] blk.11.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 155/ 866] blk.11.attn_output.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq2_xxs .. size = 60.00 MiB -> 7.73 MiB |
| [ 156/ 866] blk.11.attn_q.weight - [ 5120, 12288, 1, 1], type = bf16, converting to iq1_m .. size = 120.00 MiB -> 13.12 MiB |
| [ 157/ 866] blk.11.attn_q_norm.weight - [ 256, 1, 1, 1], type = f32, size = 0.001 MiB |
| [ 158/ 866] blk.11.attn_v.weight - [ 5120, 1024, 1, 1], type = bf16, converting to q4_K .. size = 10.00 MiB -> 2.81 MiB |
| [ 159/ 866] blk.11.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 160/ 866] blk.11.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 161/ 866] blk.11.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 162/ 866] blk.11.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 163/ 866] blk.12.attn_gate.weight - [ 5120, 6144, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 164/ 866] blk.12.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 165/ 866] blk.12.attn_qkv.weight - [ 5120, 10240, 1, 1], type = bf16, converting to q4_K .. size = 100.00 MiB -> 28.12 MiB |
| [ 166/ 866] blk.12.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 167/ 866] blk.12.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 168/ 866] blk.12.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 169/ 866] blk.12.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 170/ 866] blk.12.ssm_a - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 171/ 866] blk.12.ssm_alpha.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 172/ 866] blk.12.ssm_beta.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 173/ 866] blk.12.ssm_conv1d.weight - [ 4, 10240, 1, 1], type = f32, size = 0.156 MiB |
| [ 174/ 866] blk.12.ssm_dt.bias - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 175/ 866] blk.12.ssm_norm.weight - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 176/ 866] blk.12.ssm_out.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 177/ 866] blk.13.attn_gate.weight - [ 5120, 6144, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 178/ 866] blk.13.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 179/ 866] blk.13.attn_qkv.weight - [ 5120, 10240, 1, 1], type = bf16, converting to q4_K .. size = 100.00 MiB -> 28.12 MiB |
| [ 180/ 866] blk.13.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 181/ 866] blk.13.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 182/ 866] blk.13.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 183/ 866] blk.13.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 184/ 866] blk.13.ssm_a - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 185/ 866] blk.13.ssm_alpha.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 186/ 866] blk.13.ssm_beta.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 187/ 866] blk.13.ssm_conv1d.weight - [ 4, 10240, 1, 1], type = f32, size = 0.156 MiB |
| [ 188/ 866] blk.13.ssm_dt.bias - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 189/ 866] blk.13.ssm_norm.weight - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 190/ 866] blk.13.ssm_out.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 191/ 866] blk.14.attn_gate.weight - [ 5120, 6144, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 192/ 866] blk.14.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 193/ 866] blk.14.attn_qkv.weight - [ 5120, 10240, 1, 1], type = bf16, converting to q4_K .. size = 100.00 MiB -> 28.12 MiB |
| [ 194/ 866] blk.14.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 195/ 866] blk.14.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 196/ 866] blk.14.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 197/ 866] blk.14.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 198/ 866] blk.14.ssm_a - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 199/ 866] blk.14.ssm_alpha.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 200/ 866] blk.14.ssm_beta.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 201/ 866] blk.14.ssm_conv1d.weight - [ 4, 10240, 1, 1], type = f32, size = 0.156 MiB |
| [ 202/ 866] blk.14.ssm_dt.bias - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 203/ 866] blk.14.ssm_norm.weight - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 204/ 866] blk.14.ssm_out.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 205/ 866] blk.15.attn_k.weight - [ 5120, 1024, 1, 1], type = bf16, converting to iq1_m .. size = 10.00 MiB -> 1.09 MiB |
| [ 206/ 866] blk.15.attn_k_norm.weight - [ 256, 1, 1, 1], type = f32, size = 0.001 MiB |
| [ 207/ 866] blk.15.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 208/ 866] blk.15.attn_output.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq2_xxs .. size = 60.00 MiB -> 7.73 MiB |
| [ 209/ 866] blk.15.attn_q.weight - [ 5120, 12288, 1, 1], type = bf16, converting to iq1_m .. size = 120.00 MiB -> 13.12 MiB |
| [ 210/ 866] blk.15.attn_q_norm.weight - [ 256, 1, 1, 1], type = f32, size = 0.001 MiB |
| [ 211/ 866] blk.15.attn_v.weight - [ 5120, 1024, 1, 1], type = bf16, converting to q4_K .. size = 10.00 MiB -> 2.81 MiB |
| [ 212/ 866] blk.15.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 213/ 866] blk.15.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 214/ 866] blk.15.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 215/ 866] blk.15.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 216/ 866] blk.16.attn_gate.weight - [ 5120, 6144, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 217/ 866] blk.16.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 218/ 866] blk.16.attn_qkv.weight - [ 5120, 10240, 1, 1], type = bf16, converting to q4_K .. size = 100.00 MiB -> 28.12 MiB |
| [ 219/ 866] blk.16.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 220/ 866] blk.16.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 221/ 866] blk.16.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 222/ 866] blk.16.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 223/ 866] blk.16.ssm_a - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 224/ 866] blk.16.ssm_alpha.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 225/ 866] blk.16.ssm_beta.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 226/ 866] blk.16.ssm_conv1d.weight - [ 4, 10240, 1, 1], type = f32, size = 0.156 MiB |
| [ 227/ 866] blk.16.ssm_dt.bias - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 228/ 866] blk.16.ssm_norm.weight - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 229/ 866] blk.16.ssm_out.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 230/ 866] blk.17.attn_gate.weight - [ 5120, 6144, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 231/ 866] blk.17.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 232/ 866] blk.17.attn_qkv.weight - [ 5120, 10240, 1, 1], type = bf16, converting to q4_K .. size = 100.00 MiB -> 28.12 MiB |
| [ 233/ 866] blk.17.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 234/ 866] blk.17.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 235/ 866] blk.17.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 236/ 866] blk.17.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 237/ 866] blk.17.ssm_a - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 238/ 866] blk.17.ssm_alpha.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 239/ 866] blk.17.ssm_beta.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 240/ 866] blk.17.ssm_conv1d.weight - [ 4, 10240, 1, 1], type = f32, size = 0.156 MiB |
| [ 241/ 866] blk.17.ssm_dt.bias - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 242/ 866] blk.17.ssm_norm.weight - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 243/ 866] blk.17.ssm_out.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 244/ 866] blk.18.attn_gate.weight - [ 5120, 6144, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 245/ 866] blk.18.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 246/ 866] blk.18.attn_qkv.weight - [ 5120, 10240, 1, 1], type = bf16, converting to q4_K .. size = 100.00 MiB -> 28.12 MiB |
| [ 247/ 866] blk.18.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 248/ 866] blk.18.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 249/ 866] blk.18.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 250/ 866] blk.18.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 251/ 866] blk.18.ssm_a - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 252/ 866] blk.18.ssm_alpha.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 253/ 866] blk.18.ssm_beta.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 254/ 866] blk.18.ssm_conv1d.weight - [ 4, 10240, 1, 1], type = f32, size = 0.156 MiB |
| [ 255/ 866] blk.18.ssm_dt.bias - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 256/ 866] blk.18.ssm_norm.weight - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 257/ 866] blk.18.ssm_out.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 258/ 866] blk.19.attn_k.weight - [ 5120, 1024, 1, 1], type = bf16, converting to iq1_m .. size = 10.00 MiB -> 1.09 MiB |
| [ 259/ 866] blk.19.attn_k_norm.weight - [ 256, 1, 1, 1], type = f32, size = 0.001 MiB |
| [ 260/ 866] blk.19.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 261/ 866] blk.19.attn_output.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq2_xxs .. size = 60.00 MiB -> 7.73 MiB |
| [ 262/ 866] blk.19.attn_q.weight - [ 5120, 12288, 1, 1], type = bf16, converting to iq1_m .. size = 120.00 MiB -> 13.12 MiB |
| [ 263/ 866] blk.19.attn_q_norm.weight - [ 256, 1, 1, 1], type = f32, size = 0.001 MiB |
| [ 264/ 866] blk.19.attn_v.weight - [ 5120, 1024, 1, 1], type = bf16, converting to q4_K .. size = 10.00 MiB -> 2.81 MiB |
| [ 265/ 866] blk.19.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 266/ 866] blk.19.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 267/ 866] blk.19.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 268/ 866] blk.19.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 269/ 866] blk.20.attn_gate.weight - [ 5120, 6144, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 270/ 866] blk.20.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 271/ 866] blk.20.attn_qkv.weight - [ 5120, 10240, 1, 1], type = bf16, converting to q4_K .. size = 100.00 MiB -> 28.12 MiB |
| [ 272/ 866] blk.20.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 273/ 866] blk.20.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 274/ 866] blk.20.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 275/ 866] blk.20.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 276/ 866] blk.20.ssm_a - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 277/ 866] blk.20.ssm_alpha.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 278/ 866] blk.20.ssm_beta.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 279/ 866] blk.20.ssm_conv1d.weight - [ 4, 10240, 1, 1], type = f32, size = 0.156 MiB |
| [ 280/ 866] blk.20.ssm_dt.bias - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 281/ 866] blk.20.ssm_norm.weight - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 282/ 866] blk.20.ssm_out.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 283/ 866] blk.21.attn_gate.weight - [ 5120, 6144, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 284/ 866] blk.21.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 285/ 866] blk.21.attn_qkv.weight - [ 5120, 10240, 1, 1], type = bf16, converting to q4_K .. size = 100.00 MiB -> 28.12 MiB |
| [ 286/ 866] blk.21.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 287/ 866] blk.21.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 288/ 866] blk.21.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 289/ 866] blk.21.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 290/ 866] blk.21.ssm_a - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 291/ 866] blk.21.ssm_alpha.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 292/ 866] blk.21.ssm_beta.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 293/ 866] blk.21.ssm_conv1d.weight - [ 4, 10240, 1, 1], type = f32, size = 0.156 MiB |
| [ 294/ 866] blk.21.ssm_dt.bias - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 295/ 866] blk.21.ssm_norm.weight - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 296/ 866] blk.21.ssm_out.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 297/ 866] blk.22.attn_gate.weight - [ 5120, 6144, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 298/ 866] blk.22.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 299/ 866] blk.22.attn_qkv.weight - [ 5120, 10240, 1, 1], type = bf16, converting to q4_K .. size = 100.00 MiB -> 28.12 MiB |
| [ 300/ 866] blk.22.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 301/ 866] blk.22.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 302/ 866] blk.22.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 303/ 866] blk.22.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 304/ 866] blk.22.ssm_a - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 305/ 866] blk.22.ssm_alpha.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 306/ 866] blk.22.ssm_beta.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 307/ 866] blk.22.ssm_conv1d.weight - [ 4, 10240, 1, 1], type = f32, size = 0.156 MiB |
| [ 308/ 866] blk.22.ssm_dt.bias - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 309/ 866] blk.22.ssm_norm.weight - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 310/ 866] blk.22.ssm_out.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 311/ 866] blk.23.attn_k.weight - [ 5120, 1024, 1, 1], type = bf16, converting to iq1_m .. size = 10.00 MiB -> 1.09 MiB |
| [ 312/ 866] blk.23.attn_k_norm.weight - [ 256, 1, 1, 1], type = f32, size = 0.001 MiB |
| [ 313/ 866] blk.23.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 314/ 866] blk.23.attn_output.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq2_xxs .. size = 60.00 MiB -> 7.73 MiB |
| [ 315/ 866] blk.23.attn_q.weight - [ 5120, 12288, 1, 1], type = bf16, converting to iq1_m .. size = 120.00 MiB -> 13.12 MiB |
| [ 316/ 866] blk.23.attn_q_norm.weight - [ 256, 1, 1, 1], type = f32, size = 0.001 MiB |
| [ 317/ 866] blk.23.attn_v.weight - [ 5120, 1024, 1, 1], type = bf16, converting to q4_K .. size = 10.00 MiB -> 2.81 MiB |
| [ 318/ 866] blk.23.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 319/ 866] blk.23.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 320/ 866] blk.23.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 321/ 866] blk.23.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 322/ 866] blk.24.attn_gate.weight - [ 5120, 6144, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 323/ 866] blk.24.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 324/ 866] blk.24.attn_qkv.weight - [ 5120, 10240, 1, 1], type = bf16, converting to q4_K .. size = 100.00 MiB -> 28.12 MiB |
| [ 325/ 866] blk.24.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 326/ 866] blk.24.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 327/ 866] blk.24.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 328/ 866] blk.24.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 329/ 866] blk.24.ssm_a - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 330/ 866] blk.24.ssm_alpha.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 331/ 866] blk.24.ssm_beta.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 332/ 866] blk.24.ssm_conv1d.weight - [ 4, 10240, 1, 1], type = f32, size = 0.156 MiB |
| [ 333/ 866] blk.24.ssm_dt.bias - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 334/ 866] blk.24.ssm_norm.weight - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 335/ 866] blk.24.ssm_out.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 336/ 866] blk.25.attn_gate.weight - [ 5120, 6144, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 337/ 866] blk.25.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 338/ 866] blk.25.attn_qkv.weight - [ 5120, 10240, 1, 1], type = bf16, converting to q4_K .. size = 100.00 MiB -> 28.12 MiB |
| [ 339/ 866] blk.25.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 340/ 866] blk.25.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 341/ 866] blk.25.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 342/ 866] blk.25.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 343/ 866] blk.25.ssm_a - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 344/ 866] blk.25.ssm_alpha.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 345/ 866] blk.25.ssm_beta.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 346/ 866] blk.25.ssm_conv1d.weight - [ 4, 10240, 1, 1], type = f32, size = 0.156 MiB |
| [ 347/ 866] blk.25.ssm_dt.bias - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 348/ 866] blk.25.ssm_norm.weight - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 349/ 866] blk.25.ssm_out.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 350/ 866] blk.26.attn_gate.weight - [ 5120, 6144, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 351/ 866] blk.26.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 352/ 866] blk.26.attn_qkv.weight - [ 5120, 10240, 1, 1], type = bf16, converting to q4_K .. size = 100.00 MiB -> 28.12 MiB |
| [ 353/ 866] blk.26.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 354/ 866] blk.26.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 355/ 866] blk.26.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 356/ 866] blk.26.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 357/ 866] blk.26.ssm_a - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 358/ 866] blk.26.ssm_alpha.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 359/ 866] blk.26.ssm_beta.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 360/ 866] blk.26.ssm_conv1d.weight - [ 4, 10240, 1, 1], type = f32, size = 0.156 MiB |
| [ 361/ 866] blk.26.ssm_dt.bias - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 362/ 866] blk.26.ssm_norm.weight - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 363/ 866] blk.26.ssm_out.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 364/ 866] blk.27.attn_k.weight - [ 5120, 1024, 1, 1], type = bf16, converting to iq1_m .. size = 10.00 MiB -> 1.09 MiB |
| [ 365/ 866] blk.27.attn_k_norm.weight - [ 256, 1, 1, 1], type = f32, size = 0.001 MiB |
| [ 366/ 866] blk.27.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 367/ 866] blk.27.attn_output.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq2_xxs .. size = 60.00 MiB -> 7.73 MiB |
| [ 368/ 866] blk.27.attn_q.weight - [ 5120, 12288, 1, 1], type = bf16, converting to iq1_m .. size = 120.00 MiB -> 13.12 MiB |
| [ 369/ 866] blk.27.attn_q_norm.weight - [ 256, 1, 1, 1], type = f32, size = 0.001 MiB |
| [ 370/ 866] blk.27.attn_v.weight - [ 5120, 1024, 1, 1], type = bf16, converting to q4_K .. size = 10.00 MiB -> 2.81 MiB |
| [ 371/ 866] blk.27.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 372/ 866] blk.27.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 373/ 866] blk.27.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 374/ 866] blk.27.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 375/ 866] blk.28.attn_gate.weight - [ 5120, 6144, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 376/ 866] blk.28.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 377/ 866] blk.28.attn_qkv.weight - [ 5120, 10240, 1, 1], type = bf16, converting to q4_K .. size = 100.00 MiB -> 28.12 MiB |
| [ 378/ 866] blk.28.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 379/ 866] blk.28.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 380/ 866] blk.28.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 381/ 866] blk.28.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 382/ 866] blk.28.ssm_a - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 383/ 866] blk.28.ssm_alpha.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 384/ 866] blk.28.ssm_beta.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 385/ 866] blk.28.ssm_conv1d.weight - [ 4, 10240, 1, 1], type = f32, size = 0.156 MiB |
| [ 386/ 866] blk.28.ssm_dt.bias - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 387/ 866] blk.28.ssm_norm.weight - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 388/ 866] blk.28.ssm_out.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 389/ 866] blk.29.attn_gate.weight - [ 5120, 6144, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 390/ 866] blk.29.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 391/ 866] blk.29.attn_qkv.weight - [ 5120, 10240, 1, 1], type = bf16, converting to q4_K .. size = 100.00 MiB -> 28.12 MiB |
| [ 392/ 866] blk.29.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 393/ 866] blk.29.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 394/ 866] blk.29.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 395/ 866] blk.29.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 396/ 866] blk.29.ssm_a - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 397/ 866] blk.29.ssm_alpha.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 398/ 866] blk.29.ssm_beta.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 399/ 866] blk.29.ssm_conv1d.weight - [ 4, 10240, 1, 1], type = f32, size = 0.156 MiB |
| [ 400/ 866] blk.29.ssm_dt.bias - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 401/ 866] blk.29.ssm_norm.weight - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 402/ 866] blk.29.ssm_out.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 403/ 866] blk.30.attn_gate.weight - [ 5120, 6144, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 404/ 866] blk.30.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 405/ 866] blk.30.attn_qkv.weight - [ 5120, 10240, 1, 1], type = bf16, converting to q4_K .. size = 100.00 MiB -> 28.12 MiB |
| [ 406/ 866] blk.30.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 407/ 866] blk.30.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 408/ 866] blk.30.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 409/ 866] blk.30.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 410/ 866] blk.30.ssm_a - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 411/ 866] blk.30.ssm_alpha.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 412/ 866] blk.30.ssm_beta.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 413/ 866] blk.30.ssm_conv1d.weight - [ 4, 10240, 1, 1], type = f32, size = 0.156 MiB |
| [ 414/ 866] blk.30.ssm_dt.bias - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 415/ 866] blk.30.ssm_norm.weight - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 416/ 866] blk.30.ssm_out.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 417/ 866] blk.31.attn_k.weight - [ 5120, 1024, 1, 1], type = bf16, converting to iq1_m .. size = 10.00 MiB -> 1.09 MiB |
| [ 418/ 866] blk.31.attn_k_norm.weight - [ 256, 1, 1, 1], type = f32, size = 0.001 MiB |
| [ 419/ 866] blk.31.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 420/ 866] blk.31.attn_output.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq2_xxs .. size = 60.00 MiB -> 7.73 MiB |
| [ 421/ 866] blk.31.attn_q.weight - [ 5120, 12288, 1, 1], type = bf16, converting to iq1_m .. size = 120.00 MiB -> 13.12 MiB |
| [ 422/ 866] blk.31.attn_q_norm.weight - [ 256, 1, 1, 1], type = f32, size = 0.001 MiB |
| [ 423/ 866] blk.31.attn_v.weight - [ 5120, 1024, 1, 1], type = bf16, converting to q4_K .. size = 10.00 MiB -> 2.81 MiB |
| [ 424/ 866] blk.31.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 425/ 866] blk.31.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 426/ 866] blk.31.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 427/ 866] blk.31.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 428/ 866] blk.32.attn_gate.weight - [ 5120, 6144, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 429/ 866] blk.32.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 430/ 866] blk.32.attn_qkv.weight - [ 5120, 10240, 1, 1], type = bf16, converting to q4_K .. size = 100.00 MiB -> 28.12 MiB |
| [ 431/ 866] blk.32.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 432/ 866] blk.32.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 433/ 866] blk.32.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 434/ 866] blk.32.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 435/ 866] blk.32.ssm_a - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 436/ 866] blk.32.ssm_alpha.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 437/ 866] blk.32.ssm_beta.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 438/ 866] blk.32.ssm_conv1d.weight - [ 4, 10240, 1, 1], type = f32, size = 0.156 MiB |
| [ 439/ 866] blk.32.ssm_dt.bias - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 440/ 866] blk.32.ssm_norm.weight - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 441/ 866] blk.32.ssm_out.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 442/ 866] blk.33.attn_gate.weight - [ 5120, 6144, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 443/ 866] blk.33.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 444/ 866] blk.33.attn_qkv.weight - [ 5120, 10240, 1, 1], type = bf16, converting to q4_K .. size = 100.00 MiB -> 28.12 MiB |
| [ 445/ 866] blk.33.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 446/ 866] blk.33.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 447/ 866] blk.33.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 448/ 866] blk.33.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 449/ 866] blk.33.ssm_a - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 450/ 866] blk.33.ssm_alpha.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 451/ 866] blk.33.ssm_beta.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 452/ 866] blk.33.ssm_conv1d.weight - [ 4, 10240, 1, 1], type = f32, size = 0.156 MiB |
| [ 453/ 866] blk.33.ssm_dt.bias - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 454/ 866] blk.33.ssm_norm.weight - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 455/ 866] blk.33.ssm_out.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 456/ 866] blk.34.attn_gate.weight - [ 5120, 6144, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 457/ 866] blk.34.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 458/ 866] blk.34.attn_qkv.weight - [ 5120, 10240, 1, 1], type = bf16, converting to q4_K .. size = 100.00 MiB -> 28.12 MiB |
| [ 459/ 866] blk.34.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 460/ 866] blk.34.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 461/ 866] blk.34.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 462/ 866] blk.34.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 463/ 866] blk.34.ssm_a - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 464/ 866] blk.34.ssm_alpha.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 465/ 866] blk.34.ssm_beta.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 466/ 866] blk.34.ssm_conv1d.weight - [ 4, 10240, 1, 1], type = f32, size = 0.156 MiB |
| [ 467/ 866] blk.34.ssm_dt.bias - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 468/ 866] blk.34.ssm_norm.weight - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 469/ 866] blk.34.ssm_out.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 470/ 866] blk.35.attn_k.weight - [ 5120, 1024, 1, 1], type = bf16, converting to iq1_m .. size = 10.00 MiB -> 1.09 MiB |
| [ 471/ 866] blk.35.attn_k_norm.weight - [ 256, 1, 1, 1], type = f32, size = 0.001 MiB |
| [ 472/ 866] blk.35.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 473/ 866] blk.35.attn_output.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq2_xxs .. size = 60.00 MiB -> 7.73 MiB |
| [ 474/ 866] blk.35.attn_q.weight - [ 5120, 12288, 1, 1], type = bf16, converting to iq1_m .. size = 120.00 MiB -> 13.12 MiB |
| [ 475/ 866] blk.35.attn_q_norm.weight - [ 256, 1, 1, 1], type = f32, size = 0.001 MiB |
| [ 476/ 866] blk.35.attn_v.weight - [ 5120, 1024, 1, 1], type = bf16, converting to q4_K .. size = 10.00 MiB -> 2.81 MiB |
| [ 477/ 866] blk.35.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 478/ 866] blk.35.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 479/ 866] blk.35.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 480/ 866] blk.35.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 481/ 866] blk.36.attn_gate.weight - [ 5120, 6144, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 482/ 866] blk.36.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 483/ 866] blk.36.attn_qkv.weight - [ 5120, 10240, 1, 1], type = bf16, converting to q4_K .. size = 100.00 MiB -> 28.12 MiB |
| [ 484/ 866] blk.36.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 485/ 866] blk.36.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 486/ 866] blk.36.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 487/ 866] blk.36.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 488/ 866] blk.36.ssm_a - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 489/ 866] blk.36.ssm_alpha.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 490/ 866] blk.36.ssm_beta.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 491/ 866] blk.36.ssm_conv1d.weight - [ 4, 10240, 1, 1], type = f32, size = 0.156 MiB |
| [ 492/ 866] blk.36.ssm_dt.bias - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 493/ 866] blk.36.ssm_norm.weight - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 494/ 866] blk.36.ssm_out.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 495/ 866] blk.37.attn_gate.weight - [ 5120, 6144, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 496/ 866] blk.37.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 497/ 866] blk.37.attn_qkv.weight - [ 5120, 10240, 1, 1], type = bf16, converting to q4_K .. size = 100.00 MiB -> 28.12 MiB |
| [ 498/ 866] blk.37.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 499/ 866] blk.37.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 500/ 866] blk.37.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 501/ 866] blk.37.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 502/ 866] blk.37.ssm_a - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 503/ 866] blk.37.ssm_alpha.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 504/ 866] blk.37.ssm_beta.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 505/ 866] blk.37.ssm_conv1d.weight - [ 4, 10240, 1, 1], type = f32, size = 0.156 MiB |
| [ 506/ 866] blk.37.ssm_dt.bias - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 507/ 866] blk.37.ssm_norm.weight - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 508/ 866] blk.37.ssm_out.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 509/ 866] blk.38.attn_gate.weight - [ 5120, 6144, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 510/ 866] blk.38.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 511/ 866] blk.38.attn_qkv.weight - [ 5120, 10240, 1, 1], type = bf16, converting to q4_K .. size = 100.00 MiB -> 28.12 MiB |
| [ 512/ 866] blk.38.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 513/ 866] blk.38.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 514/ 866] blk.38.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 515/ 866] blk.38.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 516/ 866] blk.38.ssm_a - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 517/ 866] blk.38.ssm_alpha.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 518/ 866] blk.38.ssm_beta.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 519/ 866] blk.38.ssm_conv1d.weight - [ 4, 10240, 1, 1], type = f32, size = 0.156 MiB |
| [ 520/ 866] blk.38.ssm_dt.bias - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 521/ 866] blk.38.ssm_norm.weight - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 522/ 866] blk.38.ssm_out.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 523/ 866] blk.39.attn_k.weight - [ 5120, 1024, 1, 1], type = bf16, converting to iq1_m .. size = 10.00 MiB -> 1.09 MiB |
| [ 524/ 866] blk.39.attn_k_norm.weight - [ 256, 1, 1, 1], type = f32, size = 0.001 MiB |
| [ 525/ 866] blk.39.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 526/ 866] blk.39.attn_output.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq2_xxs .. size = 60.00 MiB -> 7.73 MiB |
| [ 527/ 866] blk.39.attn_q.weight - [ 5120, 12288, 1, 1], type = bf16, converting to iq1_m .. size = 120.00 MiB -> 13.12 MiB |
| [ 528/ 866] blk.39.attn_q_norm.weight - [ 256, 1, 1, 1], type = f32, size = 0.001 MiB |
| [ 529/ 866] blk.39.attn_v.weight - [ 5120, 1024, 1, 1], type = bf16, converting to q4_K .. size = 10.00 MiB -> 2.81 MiB |
| [ 530/ 866] blk.39.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 531/ 866] blk.39.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 532/ 866] blk.39.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 533/ 866] blk.39.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 534/ 866] blk.40.attn_gate.weight - [ 5120, 6144, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 535/ 866] blk.40.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 536/ 866] blk.40.attn_qkv.weight - [ 5120, 10240, 1, 1], type = bf16, converting to q4_K .. size = 100.00 MiB -> 28.12 MiB |
| [ 537/ 866] blk.40.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 538/ 866] blk.40.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 539/ 866] blk.40.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 540/ 866] blk.40.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 541/ 866] blk.40.ssm_a - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 542/ 866] blk.40.ssm_alpha.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 543/ 866] blk.40.ssm_beta.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 544/ 866] blk.40.ssm_conv1d.weight - [ 4, 10240, 1, 1], type = f32, size = 0.156 MiB |
| [ 545/ 866] blk.40.ssm_dt.bias - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 546/ 866] blk.40.ssm_norm.weight - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 547/ 866] blk.40.ssm_out.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 548/ 866] blk.41.attn_gate.weight - [ 5120, 6144, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 549/ 866] blk.41.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 550/ 866] blk.41.attn_qkv.weight - [ 5120, 10240, 1, 1], type = bf16, converting to q4_K .. size = 100.00 MiB -> 28.12 MiB |
| [ 551/ 866] blk.41.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 552/ 866] blk.41.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 553/ 866] blk.41.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 554/ 866] blk.41.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 555/ 866] blk.41.ssm_a - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 556/ 866] blk.41.ssm_alpha.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 557/ 866] blk.41.ssm_beta.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 558/ 866] blk.41.ssm_conv1d.weight - [ 4, 10240, 1, 1], type = f32, size = 0.156 MiB |
| [ 559/ 866] blk.41.ssm_dt.bias - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 560/ 866] blk.41.ssm_norm.weight - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 561/ 866] blk.41.ssm_out.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 562/ 866] blk.42.attn_gate.weight - [ 5120, 6144, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 563/ 866] blk.42.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 564/ 866] blk.42.attn_qkv.weight - [ 5120, 10240, 1, 1], type = bf16, converting to q4_K .. size = 100.00 MiB -> 28.12 MiB |
| [ 565/ 866] blk.42.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 566/ 866] blk.42.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 567/ 866] blk.42.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 568/ 866] blk.42.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 569/ 866] blk.42.ssm_a - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 570/ 866] blk.42.ssm_alpha.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 571/ 866] blk.42.ssm_beta.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 572/ 866] blk.42.ssm_conv1d.weight - [ 4, 10240, 1, 1], type = f32, size = 0.156 MiB |
| [ 573/ 866] blk.42.ssm_dt.bias - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 574/ 866] blk.42.ssm_norm.weight - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 575/ 866] blk.42.ssm_out.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 576/ 866] blk.43.attn_k.weight - [ 5120, 1024, 1, 1], type = bf16, converting to iq1_m .. size = 10.00 MiB -> 1.09 MiB |
| [ 577/ 866] blk.43.attn_k_norm.weight - [ 256, 1, 1, 1], type = f32, size = 0.001 MiB |
| [ 578/ 866] blk.43.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 579/ 866] blk.43.attn_output.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq2_xxs .. size = 60.00 MiB -> 7.73 MiB |
| [ 580/ 866] blk.43.attn_q.weight - [ 5120, 12288, 1, 1], type = bf16, converting to iq1_m .. size = 120.00 MiB -> 13.12 MiB |
| [ 581/ 866] blk.43.attn_q_norm.weight - [ 256, 1, 1, 1], type = f32, size = 0.001 MiB |
| [ 582/ 866] blk.43.attn_v.weight - [ 5120, 1024, 1, 1], type = bf16, converting to q4_K .. size = 10.00 MiB -> 2.81 MiB |
| [ 583/ 866] blk.43.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 584/ 866] blk.43.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 585/ 866] blk.43.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 586/ 866] blk.43.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 587/ 866] blk.44.attn_gate.weight - [ 5120, 6144, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 588/ 866] blk.44.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 589/ 866] blk.44.attn_qkv.weight - [ 5120, 10240, 1, 1], type = bf16, converting to q4_K .. size = 100.00 MiB -> 28.12 MiB |
| [ 590/ 866] blk.44.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 591/ 866] blk.44.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 592/ 866] blk.44.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 593/ 866] blk.44.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 594/ 866] blk.44.ssm_a - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 595/ 866] blk.44.ssm_alpha.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 596/ 866] blk.44.ssm_beta.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 597/ 866] blk.44.ssm_conv1d.weight - [ 4, 10240, 1, 1], type = f32, size = 0.156 MiB |
| [ 598/ 866] blk.44.ssm_dt.bias - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 599/ 866] blk.44.ssm_norm.weight - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 600/ 866] blk.44.ssm_out.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 601/ 866] blk.45.attn_gate.weight - [ 5120, 6144, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 602/ 866] blk.45.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 603/ 866] blk.45.attn_qkv.weight - [ 5120, 10240, 1, 1], type = bf16, converting to q4_K .. size = 100.00 MiB -> 28.12 MiB |
| [ 604/ 866] blk.45.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 605/ 866] blk.45.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 606/ 866] blk.45.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 607/ 866] blk.45.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 608/ 866] blk.45.ssm_a - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 609/ 866] blk.45.ssm_alpha.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 610/ 866] blk.45.ssm_beta.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 611/ 866] blk.45.ssm_conv1d.weight - [ 4, 10240, 1, 1], type = f32, size = 0.156 MiB |
| [ 612/ 866] blk.45.ssm_dt.bias - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 613/ 866] blk.45.ssm_norm.weight - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 614/ 866] blk.45.ssm_out.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 615/ 866] blk.46.attn_gate.weight - [ 5120, 6144, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 616/ 866] blk.46.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 617/ 866] blk.46.attn_qkv.weight - [ 5120, 10240, 1, 1], type = bf16, converting to q4_K .. size = 100.00 MiB -> 28.12 MiB |
| [ 618/ 866] blk.46.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 619/ 866] blk.46.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 620/ 866] blk.46.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 621/ 866] blk.46.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 622/ 866] blk.46.ssm_a - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 623/ 866] blk.46.ssm_alpha.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 624/ 866] blk.46.ssm_beta.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 625/ 866] blk.46.ssm_conv1d.weight - [ 4, 10240, 1, 1], type = f32, size = 0.156 MiB |
| [ 626/ 866] blk.46.ssm_dt.bias - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 627/ 866] blk.46.ssm_norm.weight - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 628/ 866] blk.46.ssm_out.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 629/ 866] blk.47.attn_k.weight - [ 5120, 1024, 1, 1], type = bf16, converting to iq1_m .. size = 10.00 MiB -> 1.09 MiB |
| [ 630/ 866] blk.47.attn_k_norm.weight - [ 256, 1, 1, 1], type = f32, size = 0.001 MiB |
| [ 631/ 866] blk.47.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 632/ 866] blk.47.attn_output.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq2_xxs .. size = 60.00 MiB -> 7.73 MiB |
| [ 633/ 866] blk.47.attn_q.weight - [ 5120, 12288, 1, 1], type = bf16, converting to iq1_m .. size = 120.00 MiB -> 13.12 MiB |
| [ 634/ 866] blk.47.attn_q_norm.weight - [ 256, 1, 1, 1], type = f32, size = 0.001 MiB |
| [ 635/ 866] blk.47.attn_v.weight - [ 5120, 1024, 1, 1], type = bf16, converting to q4_K .. size = 10.00 MiB -> 2.81 MiB |
| [ 636/ 866] blk.47.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 637/ 866] blk.47.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 638/ 866] blk.47.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 639/ 866] blk.47.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 640/ 866] blk.48.attn_gate.weight - [ 5120, 6144, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 641/ 866] blk.48.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 642/ 866] blk.48.attn_qkv.weight - [ 5120, 10240, 1, 1], type = bf16, converting to q4_K .. size = 100.00 MiB -> 28.12 MiB |
| [ 643/ 866] blk.48.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 644/ 866] blk.48.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 645/ 866] blk.48.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 646/ 866] blk.48.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 647/ 866] blk.48.ssm_a - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 648/ 866] blk.48.ssm_alpha.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 649/ 866] blk.48.ssm_beta.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 650/ 866] blk.48.ssm_conv1d.weight - [ 4, 10240, 1, 1], type = f32, size = 0.156 MiB |
| [ 651/ 866] blk.48.ssm_dt.bias - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 652/ 866] blk.48.ssm_norm.weight - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 653/ 866] blk.48.ssm_out.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 654/ 866] blk.49.attn_gate.weight - [ 5120, 6144, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 655/ 866] blk.49.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 656/ 866] blk.49.attn_qkv.weight - [ 5120, 10240, 1, 1], type = bf16, converting to q4_K .. size = 100.00 MiB -> 28.12 MiB |
| [ 657/ 866] blk.49.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 658/ 866] blk.49.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 659/ 866] blk.49.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 660/ 866] blk.49.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 661/ 866] blk.49.ssm_a - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 662/ 866] blk.49.ssm_alpha.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 663/ 866] blk.49.ssm_beta.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 664/ 866] blk.49.ssm_conv1d.weight - [ 4, 10240, 1, 1], type = f32, size = 0.156 MiB |
| [ 665/ 866] blk.49.ssm_dt.bias - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 666/ 866] blk.49.ssm_norm.weight - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 667/ 866] blk.49.ssm_out.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 668/ 866] blk.50.attn_gate.weight - [ 5120, 6144, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 669/ 866] blk.50.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 670/ 866] blk.50.attn_qkv.weight - [ 5120, 10240, 1, 1], type = bf16, converting to q4_K .. size = 100.00 MiB -> 28.12 MiB |
| [ 671/ 866] blk.50.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 672/ 866] blk.50.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 673/ 866] blk.50.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 674/ 866] blk.50.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 675/ 866] blk.50.ssm_a - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 676/ 866] blk.50.ssm_alpha.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 677/ 866] blk.50.ssm_beta.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 678/ 866] blk.50.ssm_conv1d.weight - [ 4, 10240, 1, 1], type = f32, size = 0.156 MiB |
| [ 679/ 866] blk.50.ssm_dt.bias - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 680/ 866] blk.50.ssm_norm.weight - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 681/ 866] blk.50.ssm_out.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 682/ 866] blk.51.attn_k.weight - [ 5120, 1024, 1, 1], type = bf16, converting to iq1_m .. size = 10.00 MiB -> 1.09 MiB |
| [ 683/ 866] blk.51.attn_k_norm.weight - [ 256, 1, 1, 1], type = f32, size = 0.001 MiB |
| [ 684/ 866] blk.51.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 685/ 866] blk.51.attn_output.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq2_xxs .. size = 60.00 MiB -> 7.73 MiB |
| [ 686/ 866] blk.51.attn_q.weight - [ 5120, 12288, 1, 1], type = bf16, converting to iq1_m .. size = 120.00 MiB -> 13.12 MiB |
| [ 687/ 866] blk.51.attn_q_norm.weight - [ 256, 1, 1, 1], type = f32, size = 0.001 MiB |
| [ 688/ 866] blk.51.attn_v.weight - [ 5120, 1024, 1, 1], type = bf16, converting to q4_K .. size = 10.00 MiB -> 2.81 MiB |
| [ 689/ 866] blk.51.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 690/ 866] blk.51.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 691/ 866] blk.51.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 692/ 866] blk.51.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 693/ 866] blk.52.attn_gate.weight - [ 5120, 6144, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 694/ 866] blk.52.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 695/ 866] blk.52.attn_qkv.weight - [ 5120, 10240, 1, 1], type = bf16, converting to q4_K .. size = 100.00 MiB -> 28.12 MiB |
| [ 696/ 866] blk.52.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 697/ 866] blk.52.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 698/ 866] blk.52.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 699/ 866] blk.52.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 700/ 866] blk.52.ssm_a - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 701/ 866] blk.52.ssm_alpha.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 702/ 866] blk.52.ssm_beta.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 703/ 866] blk.52.ssm_conv1d.weight - [ 4, 10240, 1, 1], type = f32, size = 0.156 MiB |
| [ 704/ 866] blk.52.ssm_dt.bias - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 705/ 866] blk.52.ssm_norm.weight - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 706/ 866] blk.52.ssm_out.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 707/ 866] blk.53.attn_gate.weight - [ 5120, 6144, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 708/ 866] blk.53.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 709/ 866] blk.53.attn_qkv.weight - [ 5120, 10240, 1, 1], type = bf16, converting to q4_K .. size = 100.00 MiB -> 28.12 MiB |
| [ 710/ 866] blk.53.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 711/ 866] blk.53.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 712/ 866] blk.53.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 713/ 866] blk.53.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 714/ 866] blk.53.ssm_a - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 715/ 866] blk.53.ssm_alpha.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 716/ 866] blk.53.ssm_beta.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 717/ 866] blk.53.ssm_conv1d.weight - [ 4, 10240, 1, 1], type = f32, size = 0.156 MiB |
| [ 718/ 866] blk.53.ssm_dt.bias - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 719/ 866] blk.53.ssm_norm.weight - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 720/ 866] blk.53.ssm_out.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 721/ 866] blk.54.attn_gate.weight - [ 5120, 6144, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 722/ 866] blk.54.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 723/ 866] blk.54.attn_qkv.weight - [ 5120, 10240, 1, 1], type = bf16, converting to q4_K .. size = 100.00 MiB -> 28.12 MiB |
| [ 724/ 866] blk.54.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 725/ 866] blk.54.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 726/ 866] blk.54.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 727/ 866] blk.54.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 728/ 866] blk.54.ssm_a - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 729/ 866] blk.54.ssm_alpha.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 730/ 866] blk.54.ssm_beta.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 731/ 866] blk.54.ssm_conv1d.weight - [ 4, 10240, 1, 1], type = f32, size = 0.156 MiB |
| [ 732/ 866] blk.54.ssm_dt.bias - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 733/ 866] blk.54.ssm_norm.weight - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 734/ 866] blk.54.ssm_out.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 735/ 866] blk.55.attn_k.weight - [ 5120, 1024, 1, 1], type = bf16, converting to iq1_m .. size = 10.00 MiB -> 1.09 MiB |
| [ 736/ 866] blk.55.attn_k_norm.weight - [ 256, 1, 1, 1], type = f32, size = 0.001 MiB |
| [ 737/ 866] blk.55.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 738/ 866] blk.55.attn_output.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq2_xxs .. size = 60.00 MiB -> 7.73 MiB |
| [ 739/ 866] blk.55.attn_q.weight - [ 5120, 12288, 1, 1], type = bf16, converting to iq1_m .. size = 120.00 MiB -> 13.12 MiB |
| [ 740/ 866] blk.55.attn_q_norm.weight - [ 256, 1, 1, 1], type = f32, size = 0.001 MiB |
| [ 741/ 866] blk.55.attn_v.weight - [ 5120, 1024, 1, 1], type = bf16, converting to q4_K .. size = 10.00 MiB -> 2.81 MiB |
| [ 742/ 866] blk.55.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 743/ 866] blk.55.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 744/ 866] blk.55.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 745/ 866] blk.55.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 746/ 866] blk.56.attn_gate.weight - [ 5120, 6144, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 747/ 866] blk.56.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 748/ 866] blk.56.attn_qkv.weight - [ 5120, 10240, 1, 1], type = bf16, converting to q4_K .. size = 100.00 MiB -> 28.12 MiB |
| [ 749/ 866] blk.56.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 750/ 866] blk.56.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 751/ 866] blk.56.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 752/ 866] blk.56.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 753/ 866] blk.56.ssm_a - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 754/ 866] blk.56.ssm_alpha.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 755/ 866] blk.56.ssm_beta.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 756/ 866] blk.56.ssm_conv1d.weight - [ 4, 10240, 1, 1], type = f32, size = 0.156 MiB |
| [ 757/ 866] blk.56.ssm_dt.bias - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 758/ 866] blk.56.ssm_norm.weight - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 759/ 866] blk.56.ssm_out.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 760/ 866] blk.57.attn_gate.weight - [ 5120, 6144, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 761/ 866] blk.57.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 762/ 866] blk.57.attn_qkv.weight - [ 5120, 10240, 1, 1], type = bf16, converting to q4_K .. size = 100.00 MiB -> 28.12 MiB |
| [ 763/ 866] blk.57.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 764/ 866] blk.57.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 765/ 866] blk.57.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 766/ 866] blk.57.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 767/ 866] blk.57.ssm_a - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 768/ 866] blk.57.ssm_alpha.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 769/ 866] blk.57.ssm_beta.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 770/ 866] blk.57.ssm_conv1d.weight - [ 4, 10240, 1, 1], type = f32, size = 0.156 MiB |
| [ 771/ 866] blk.57.ssm_dt.bias - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 772/ 866] blk.57.ssm_norm.weight - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 773/ 866] blk.57.ssm_out.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 774/ 866] blk.58.attn_gate.weight - [ 5120, 6144, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 775/ 866] blk.58.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 776/ 866] blk.58.attn_qkv.weight - [ 5120, 10240, 1, 1], type = bf16, converting to q4_K .. size = 100.00 MiB -> 28.12 MiB |
| [ 777/ 866] blk.58.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 778/ 866] blk.58.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 779/ 866] blk.58.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 780/ 866] blk.58.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 781/ 866] blk.58.ssm_a - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 782/ 866] blk.58.ssm_alpha.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 783/ 866] blk.58.ssm_beta.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 784/ 866] blk.58.ssm_conv1d.weight - [ 4, 10240, 1, 1], type = f32, size = 0.156 MiB |
| [ 785/ 866] blk.58.ssm_dt.bias - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 786/ 866] blk.58.ssm_norm.weight - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 787/ 866] blk.58.ssm_out.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 788/ 866] blk.59.attn_k.weight - [ 5120, 1024, 1, 1], type = bf16, converting to iq1_m .. size = 10.00 MiB -> 1.09 MiB |
| [ 789/ 866] blk.59.attn_k_norm.weight - [ 256, 1, 1, 1], type = f32, size = 0.001 MiB |
| [ 790/ 866] blk.59.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 791/ 866] blk.59.attn_output.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq2_xxs .. size = 60.00 MiB -> 7.73 MiB |
| [ 792/ 866] blk.59.attn_q.weight - [ 5120, 12288, 1, 1], type = bf16, converting to iq1_m .. size = 120.00 MiB -> 13.12 MiB |
| [ 793/ 866] blk.59.attn_q_norm.weight - [ 256, 1, 1, 1], type = f32, size = 0.001 MiB |
| [ 794/ 866] blk.59.attn_v.weight - [ 5120, 1024, 1, 1], type = bf16, converting to q4_K .. size = 10.00 MiB -> 2.81 MiB |
| [ 795/ 866] blk.59.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 796/ 866] blk.59.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 797/ 866] blk.59.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 798/ 866] blk.59.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 799/ 866] blk.60.attn_gate.weight - [ 5120, 6144, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 800/ 866] blk.60.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 801/ 866] blk.60.attn_qkv.weight - [ 5120, 10240, 1, 1], type = bf16, converting to q4_K .. size = 100.00 MiB -> 28.12 MiB |
| [ 802/ 866] blk.60.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 803/ 866] blk.60.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 804/ 866] blk.60.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 805/ 866] blk.60.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 806/ 866] blk.60.ssm_a - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 807/ 866] blk.60.ssm_alpha.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 808/ 866] blk.60.ssm_beta.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 809/ 866] blk.60.ssm_conv1d.weight - [ 4, 10240, 1, 1], type = f32, size = 0.156 MiB |
| [ 810/ 866] blk.60.ssm_dt.bias - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 811/ 866] blk.60.ssm_norm.weight - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 812/ 866] blk.60.ssm_out.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 813/ 866] blk.61.attn_gate.weight - [ 5120, 6144, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 814/ 866] blk.61.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 815/ 866] blk.61.attn_qkv.weight - [ 5120, 10240, 1, 1], type = bf16, converting to q4_K .. size = 100.00 MiB -> 28.12 MiB |
| [ 816/ 866] blk.61.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 817/ 866] blk.61.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 818/ 866] blk.61.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 819/ 866] blk.61.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 820/ 866] blk.61.ssm_a - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 821/ 866] blk.61.ssm_alpha.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 822/ 866] blk.61.ssm_beta.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 823/ 866] blk.61.ssm_conv1d.weight - [ 4, 10240, 1, 1], type = f32, size = 0.156 MiB |
| [ 824/ 866] blk.61.ssm_dt.bias - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 825/ 866] blk.61.ssm_norm.weight - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 826/ 866] blk.61.ssm_out.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 827/ 866] blk.62.attn_gate.weight - [ 5120, 6144, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 828/ 866] blk.62.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 829/ 866] blk.62.attn_qkv.weight - [ 5120, 10240, 1, 1], type = bf16, converting to q4_K .. size = 100.00 MiB -> 28.12 MiB |
| [ 830/ 866] blk.62.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 831/ 866] blk.62.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 832/ 866] blk.62.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 833/ 866] blk.62.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 834/ 866] blk.62.ssm_a - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 835/ 866] blk.62.ssm_alpha.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 836/ 866] blk.62.ssm_beta.weight - [ 5120, 48, 1, 1], type = bf16, converting to iq1_m .. size = 0.47 MiB -> 0.05 MiB |
| [ 837/ 866] blk.62.ssm_conv1d.weight - [ 4, 10240, 1, 1], type = f32, size = 0.156 MiB |
| [ 838/ 866] blk.62.ssm_dt.bias - [ 48, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 839/ 866] blk.62.ssm_norm.weight - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB |
| [ 840/ 866] blk.62.ssm_out.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 60.00 MiB -> 6.56 MiB |
| [ 841/ 866] blk.63.attn_k.weight - [ 5120, 1024, 1, 1], type = bf16, converting to iq1_m .. size = 10.00 MiB -> 1.09 MiB |
| [ 842/ 866] blk.63.attn_k_norm.weight - [ 256, 1, 1, 1], type = f32, size = 0.001 MiB |
| [ 843/ 866] blk.63.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 844/ 866] blk.63.attn_output.weight - [ 6144, 5120, 1, 1], type = bf16, converting to iq2_xxs .. size = 60.00 MiB -> 7.73 MiB |
| [ 845/ 866] blk.63.attn_q.weight - [ 5120, 12288, 1, 1], type = bf16, converting to iq1_m .. size = 120.00 MiB -> 13.12 MiB |
| [ 846/ 866] blk.63.attn_q_norm.weight - [ 256, 1, 1, 1], type = f32, size = 0.001 MiB |
| [ 847/ 866] blk.63.attn_v.weight - [ 5120, 1024, 1, 1], type = bf16, converting to q4_K .. size = 10.00 MiB -> 2.81 MiB |
| [ 848/ 866] blk.63.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 849/ 866] blk.63.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 850/ 866] blk.63.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, converting to iq1_m .. size = 170.00 MiB -> 18.59 MiB |
| [ 851/ 866] blk.63.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 852/ 866] blk.64.attn_k.weight - [ 5120, 1024, 1, 1], type = bf16, |
| ====== llama_model_quantize_impl: did not find weights for blk.64.attn_k.weight |
| converting to q4_K .. size = 10.00 MiB -> 2.81 MiB |
| [ 853/ 866] blk.64.attn_k_norm.weight - [ 256, 1, 1, 1], type = f32, size = 0.001 MiB |
| [ 854/ 866] blk.64.attn_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 855/ 866] blk.64.attn_output.weight - [ 6144, 5120, 1, 1], type = bf16, |
| ====== llama_model_quantize_impl: did not find weights for blk.64.attn_output.weight |
| converting to q4_K .. size = 60.00 MiB -> 16.88 MiB |
| [ 856/ 866] blk.64.attn_q.weight - [ 5120, 12288, 1, 1], type = bf16, |
| ====== llama_model_quantize_impl: did not find weights for blk.64.attn_q.weight |
| converting to q4_K .. size = 120.00 MiB -> 33.75 MiB |
| [ 857/ 866] blk.64.attn_q_norm.weight - [ 256, 1, 1, 1], type = f32, size = 0.001 MiB |
| [ 858/ 866] blk.64.attn_v.weight - [ 5120, 1024, 1, 1], type = bf16, |
| ====== llama_model_quantize_impl: did not find weights for blk.64.attn_v.weight |
| converting to q4_K .. size = 10.00 MiB -> 2.81 MiB |
| [ 859/ 866] blk.64.ffn_down.weight - [ 17408, 5120, 1, 1], type = bf16, |
| ====== llama_model_quantize_impl: did not find weights for blk.64.ffn_down.weight |
| converting to q4_K .. size = 170.00 MiB -> 47.81 MiB |
| [ 860/ 866] blk.64.ffn_gate.weight - [ 5120, 17408, 1, 1], type = bf16, |
| ====== llama_model_quantize_impl: did not find weights for blk.64.ffn_gate.weight |
| converting to q4_K .. size = 170.00 MiB -> 47.81 MiB |
| [ 861/ 866] blk.64.ffn_up.weight - [ 5120, 17408, 1, 1], type = bf16, |
| ====== llama_model_quantize_impl: did not find weights for blk.64.ffn_up.weight |
| converting to q4_K .. size = 170.00 MiB -> 47.81 MiB |
| [ 862/ 866] blk.64.nextn.eh_proj.weight - [ 10240, 5120, 1, 1], type = bf16, |
| ====== llama_model_quantize_impl: did not find weights for blk.64.nextn.eh_proj.weight |
| converting to q4_K .. size = 100.00 MiB -> 28.12 MiB |
| [ 863/ 866] blk.64.nextn.enorm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 864/ 866] blk.64.nextn.hnorm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 865/ 866] blk.64.nextn.shared_head_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| [ 866/ 866] blk.64.post_attention_norm.weight - [ 5120, 1, 1, 1], type = f32, size = 0.020 MiB |
| llama_model_quantize_impl: model size = 52115.19 MiB (16.00 BPW) |
| llama_model_quantize_impl: quant size = 7495.00 MiB (2.30 BPW) |
|
|
| llama_quantize: quantize time = 2766880.41 ms |
| llama_quantize: total time = 2766880.41 ms |
|
|
