File size: 13,878 Bytes
1e414c4 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 |
llama_model_loader: loaded meta data with 51 key-value pairs and 959 tensors from DeepSeek-Coder-V2-Instruct-0724-IMat-GGUF/DeepSeek-Coder-V2-Instruct-0724.Q8_0.gguf.hardlink.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 = deepseek2
llama_model_loader: - kv 1: general.type str = model
llama_model_loader: - kv 2: general.name str = DeepSeek Coder V2 Instruct 0724
llama_model_loader: - kv 3: general.version str = 0724
llama_model_loader: - kv 4: general.finetune str = Instruct
llama_model_loader: - kv 5: general.basename str = DeepSeek-Coder-V2
llama_model_loader: - kv 6: general.size_label str = 160x14B
llama_model_loader: - kv 7: general.license str = other
llama_model_loader: - kv 8: general.license.name str = deepseek-license
llama_model_loader: - kv 9: general.license.link str = LICENSE
llama_model_loader: - kv 10: general.base_model.count u32 = 1
llama_model_loader: - kv 11: general.base_model.0.name str = DeepSeek Coder V2 Base
llama_model_loader: - kv 12: general.base_model.0.organization str = Deepseek Ai
llama_model_loader: - kv 13: general.base_model.0.repo_url str = https://huggingface.co./deepseek-ai/De...
llama_model_loader: - kv 14: deepseek2.block_count u32 = 60
llama_model_loader: - kv 15: deepseek2.context_length u32 = 163840
llama_model_loader: - kv 16: deepseek2.embedding_length u32 = 5120
llama_model_loader: - kv 17: deepseek2.feed_forward_length u32 = 12288
llama_model_loader: - kv 18: deepseek2.attention.head_count u32 = 128
llama_model_loader: - kv 19: deepseek2.attention.head_count_kv u32 = 128
llama_model_loader: - kv 20: deepseek2.rope.freq_base f32 = 10000.000000
llama_model_loader: - kv 21: deepseek2.attention.layer_norm_rms_epsilon f32 = 0.000001
llama_model_loader: - kv 22: deepseek2.expert_used_count u32 = 6
llama_model_loader: - kv 23: general.file_type u32 = 7
llama_model_loader: - kv 24: deepseek2.leading_dense_block_count u32 = 1
llama_model_loader: - kv 25: deepseek2.vocab_size u32 = 102400
llama_model_loader: - kv 26: deepseek2.attention.q_lora_rank u32 = 1536
llama_model_loader: - kv 27: deepseek2.attention.kv_lora_rank u32 = 512
llama_model_loader: - kv 28: deepseek2.attention.key_length u32 = 192
llama_model_loader: - kv 29: deepseek2.attention.value_length u32 = 128
llama_model_loader: - kv 30: deepseek2.expert_feed_forward_length u32 = 1536
llama_model_loader: - kv 31: deepseek2.expert_count u32 = 160
llama_model_loader: - kv 32: deepseek2.expert_shared_count u32 = 2
llama_model_loader: - kv 33: deepseek2.expert_weights_scale f32 = 16.000000
llama_model_loader: - kv 34: deepseek2.rope.dimension_count u32 = 64
llama_model_loader: - kv 35: deepseek2.rope.scaling.type str = yarn
llama_model_loader: - kv 36: deepseek2.rope.scaling.factor f32 = 40.000000
llama_model_loader: - kv 37: deepseek2.rope.scaling.original_context_length u32 = 4096
llama_model_loader: - kv 38: deepseek2.rope.scaling.yarn_log_multiplier f32 = 0.100000
llama_model_loader: - kv 39: tokenizer.ggml.model str = gpt2
llama_model_loader: - kv 40: tokenizer.ggml.pre str = deepseek-llm
llama_model_loader: - kv 41: tokenizer.ggml.tokens arr[str,102400] = ["!", "\"", "#", "$", "%", "&", "'", ...
llama_model_loader: - kv 42: tokenizer.ggml.token_type arr[i32,102400] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ...
llama_model_loader: - kv 43: tokenizer.ggml.merges arr[str,99757] = ["Ġ Ġ", "Ġ t", "Ġ a", "i n", "h e...
llama_model_loader: - kv 44: tokenizer.ggml.bos_token_id u32 = 100000
llama_model_loader: - kv 45: tokenizer.ggml.eos_token_id u32 = 100001
llama_model_loader: - kv 46: tokenizer.ggml.padding_token_id u32 = 100001
llama_model_loader: - kv 47: tokenizer.ggml.add_bos_token bool = true
llama_model_loader: - kv 48: tokenizer.ggml.add_eos_token bool = false
llama_model_loader: - kv 49: tokenizer.chat_template str = {% if not add_generation_prompt is de...
llama_model_loader: - kv 50: general.quantization_version u32 = 2
llama_model_loader: - type f32: 300 tensors
llama_model_loader: - type q8_0: 659 tensors
llm_load_vocab: special tokens cache size = 18
llm_load_vocab: token to piece cache size = 0.6411 MB
llm_load_print_meta: format = GGUF V3 (latest)
llm_load_print_meta: arch = deepseek2
llm_load_print_meta: vocab type = BPE
llm_load_print_meta: n_vocab = 102400
llm_load_print_meta: n_merges = 99757
llm_load_print_meta: vocab_only = 0
llm_load_print_meta: n_ctx_train = 163840
llm_load_print_meta: n_embd = 5120
llm_load_print_meta: n_layer = 60
llm_load_print_meta: n_head = 128
llm_load_print_meta: n_head_kv = 128
llm_load_print_meta: n_rot = 64
llm_load_print_meta: n_swa = 0
llm_load_print_meta: n_embd_head_k = 192
llm_load_print_meta: n_embd_head_v = 128
llm_load_print_meta: n_gqa = 1
llm_load_print_meta: n_embd_k_gqa = 24576
llm_load_print_meta: n_embd_v_gqa = 16384
llm_load_print_meta: f_norm_eps = 0.0e+00
llm_load_print_meta: f_norm_rms_eps = 1.0e-06
llm_load_print_meta: f_clamp_kqv = 0.0e+00
llm_load_print_meta: f_max_alibi_bias = 0.0e+00
llm_load_print_meta: f_logit_scale = 0.0e+00
llm_load_print_meta: n_ff = 12288
llm_load_print_meta: n_expert = 160
llm_load_print_meta: n_expert_used = 6
llm_load_print_meta: causal attn = 1
llm_load_print_meta: pooling type = 0
llm_load_print_meta: rope type = 0
llm_load_print_meta: rope scaling = yarn
llm_load_print_meta: freq_base_train = 10000.0
llm_load_print_meta: freq_scale_train = 0.025
llm_load_print_meta: n_ctx_orig_yarn = 4096
llm_load_print_meta: rope_finetuned = unknown
llm_load_print_meta: ssm_d_conv = 0
llm_load_print_meta: ssm_d_inner = 0
llm_load_print_meta: ssm_d_state = 0
llm_load_print_meta: ssm_dt_rank = 0
llm_load_print_meta: ssm_dt_b_c_rms = 0
llm_load_print_meta: model type = 236B
llm_load_print_meta: model ftype = Q8_0
llm_load_print_meta: model params = 235.74 B
llm_load_print_meta: model size = 233.41 GiB (8.50 BPW)
llm_load_print_meta: general.name = DeepSeek Coder V2 Instruct 0724
llm_load_print_meta: BOS token = 100000 '<｜begin▁of▁sentence｜>'
llm_load_print_meta: EOS token = 100001 '<｜end▁of▁sentence｜>'
llm_load_print_meta: PAD token = 100001 '<｜end▁of▁sentence｜>'
llm_load_print_meta: LF token = 126 'Ä'
llm_load_print_meta: max token length = 256
llm_load_print_meta: n_layer_dense_lead = 1
llm_load_print_meta: n_lora_q = 1536
llm_load_print_meta: n_lora_kv = 512
llm_load_print_meta: n_ff_exp = 1536
llm_load_print_meta: n_expert_shared = 2
llm_load_print_meta: expert_weights_scale = 16.0
llm_load_print_meta: rope_yarn_log_mul = 0.1000
ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no
ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
ggml_cuda_init: found 1 CUDA devices:
Device 0: NVIDIA GeForce RTX 4090, compute capability 8.9, VMM: yes
llm_load_tensors: ggml ctx size = 0.80 MiB
llm_load_tensors: offloading 5 repeating layers to GPU
llm_load_tensors: offloaded 5/61 layers to GPU
llm_load_tensors: CPU buffer size = 218873.36 MiB
llm_load_tensors: CUDA0 buffer size = 20135.96 MiB
....................................................................................................
llama_new_context_with_model: n_ctx = 512
llama_new_context_with_model: n_batch = 512
llama_new_context_with_model: n_ubatch = 512
llama_new_context_with_model: flash_attn = 0
llama_new_context_with_model: freq_base = 10000.0
llama_new_context_with_model: freq_scale = 0.025
llama_kv_cache_init: CUDA_Host KV buffer size = 2200.00 MiB
llama_kv_cache_init: CUDA0 KV buffer size = 200.00 MiB
llama_new_context_with_model: KV self size = 2400.00 MiB, K (f16): 1440.00 MiB, V (f16): 960.00 MiB
llama_new_context_with_model: CUDA_Host output buffer size = 0.39 MiB
llama_new_context_with_model: CUDA0 compute buffer size = 1422.00 MiB
llama_new_context_with_model: CUDA_Host compute buffer size = 81.01 MiB
llama_new_context_with_model: graph nodes = 4480
llama_new_context_with_model: graph splits = 990
system_info: n_threads = 25 (n_threads_batch = 25) / 32 | AVX = 1 | AVX_VNNI = 0 | AVX2 = 1 | AVX512 = 1 | AVX512_VBMI = 1 | AVX512_VNNI = 1 | AVX512_BF16 = 1 | FMA = 1 | NEON = 0 | SVE = 0 | ARM_FMA = 0 | F16C = 1 | FP16_VA = 0 | WASM_SIMD = 0 | BLAS = 1 | SSE3 = 1 | SSSE3 = 1 | VSX = 0 | MATMUL_INT8 = 0 | LLAMAFILE = 1 |
compute_imatrix: tokenizing the input ..
compute_imatrix: tokenization took 526.799 ms
compute_imatrix: computing over 139 chunks with batch_size 512
compute_imatrix: 90.65 seconds per pass - ETA 3 hours 30.00 minutes
[1]5.7772,[2]3.8835,[3]3.8059,[4]4.1672,[5]4.1086,[6]3.9026,[7]4.1099,[8]3.9733,[9]4.4323,
save_imatrix: stored collected data after 10 chunks in DeepSeek-Coder-V2-Instruct-0724-IMat-GGUF/imatrix.dat
[10]4.6201,[11]4.3030,[12]4.5590,[13]4.8871,[14]5.1335,[15]5.2601,[16]5.4875,[17]5.6677,[18]5.7927,[19]5.8962,
save_imatrix: stored collected data after 20 chunks in DeepSeek-Coder-V2-Instruct-0724-IMat-GGUF/imatrix.dat
[20]5.6232,[21]5.6381,[22]5.5802,[23]5.6274,[24]5.5363,[25]5.6976,[26]5.6137,[27]5.7519,[28]5.6031,[29]5.3284,
save_imatrix: stored collected data after 30 chunks in DeepSeek-Coder-V2-Instruct-0724-IMat-GGUF/imatrix.dat
[30]5.1049,[31]5.0415,[32]4.9882,[33]4.8558,[34]4.6419,[35]4.4657,[36]4.4172,[37]4.3647,[38]4.3780,[39]4.3155,
save_imatrix: stored collected data after 40 chunks in DeepSeek-Coder-V2-Instruct-0724-IMat-GGUF/imatrix.dat
[40]4.2715,[41]4.2049,[42]4.1509,[43]4.1936,[44]4.2619,[45]4.3487,[46]4.3656,[47]4.2404,[48]4.1211,[49]4.0191,
save_imatrix: stored collected data after 50 chunks in DeepSeek-Coder-V2-Instruct-0724-IMat-GGUF/imatrix.dat
[50]3.9192,[51]3.9238,[52]3.8966,[53]3.9838,[54]4.0627,[55]4.1495,[56]4.1101,[57]4.1283,[58]4.1489,[59]4.2108,
save_imatrix: stored collected data after 60 chunks in DeepSeek-Coder-V2-Instruct-0724-IMat-GGUF/imatrix.dat
[60]4.2794,[61]4.3451,[62]4.3897,[63]4.4198,[64]4.4604,[65]4.4913,[66]4.4892,[67]4.4981,[68]4.5029,[69]4.5437,
save_imatrix: stored collected data after 70 chunks in DeepSeek-Coder-V2-Instruct-0724-IMat-GGUF/imatrix.dat
[70]4.5644,[71]4.5709,[72]4.6022,[73]4.6110,[74]4.6151,[75]4.6210,[76]4.6322,[77]4.6445,[78]4.6694,[79]4.6579,
save_imatrix: stored collected data after 80 chunks in DeepSeek-Coder-V2-Instruct-0724-IMat-GGUF/imatrix.dat
[80]4.6792,[81]4.6795,[82]4.6854,[83]4.6823,[84]4.6838,[85]4.6807,[86]4.6792,[87]4.6723,[88]4.6972,[89]4.7218,
save_imatrix: stored collected data after 90 chunks in DeepSeek-Coder-V2-Instruct-0724-IMat-GGUF/imatrix.dat
[90]4.7272,[91]4.7600,[92]4.7942,[93]4.7672,[94]4.7706,[95]4.7531,[96]4.7771,[97]4.7976,[98]4.7951,[99]4.7560,
save_imatrix: stored collected data after 100 chunks in DeepSeek-Coder-V2-Instruct-0724-IMat-GGUF/imatrix.dat
[100]4.7169,[101]4.6779,[102]4.6376,[103]4.5994,[104]4.5669,[105]4.5309,[106]4.4965,[107]4.4640,[108]4.4398,[109]4.4545,
save_imatrix: stored collected data after 110 chunks in DeepSeek-Coder-V2-Instruct-0724-IMat-GGUF/imatrix.dat
[110]4.4866,[111]4.5306,[112]4.5749,[113]4.6178,[114]4.6884,[115]4.7267,[116]4.7498,[117]4.7678,[118]4.7937,[119]4.7927,
save_imatrix: stored collected data after 120 chunks in DeepSeek-Coder-V2-Instruct-0724-IMat-GGUF/imatrix.dat
[120]4.7703,[121]4.7169,[122]4.6675,[123]4.6978,[124]4.7296,[125]4.7339,[126]4.7399,[127]4.7519,[128]4.7785,[129]4.7832,
save_imatrix: stored collected data after 130 chunks in DeepSeek-Coder-V2-Instruct-0724-IMat-GGUF/imatrix.dat
[130]4.7964,[131]4.8182,[132]4.8148,[133]4.8068,[134]4.7571,[135]4.7070,[136]4.7102,[137]4.6635,[138]4.6222,[139]4.5770,
save_imatrix: stored collected data after 139 chunks in DeepSeek-Coder-V2-Instruct-0724-IMat-GGUF/imatrix.dat
llama_perf_print: load time = 194329.99 ms
llama_perf_print: prompt eval time = 15123291.91 ms / 71168 tokens ( 212.50 ms per token, 4.71 tokens per second)
llama_perf_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second)
llama_perf_print: total time = 15252664.04 ms / 71169 tokens
Final estimate: PPL = 4.5770 +/- 0.05779
|