main: build = 3010 (95f84d5c) main: built with cc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0 for x86_64-linux-gnu main: seed = 1716911218 llama_model_loader: loaded meta data with 27 key-value pairs and 561 tensors from AutoCoder-IMat-GGUF/AutoCoder.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 = llama llama_model_loader: - kv 1: general.name str = AutoCoder llama_model_loader: - kv 2: llama.block_count u32 = 62 llama_model_loader: - kv 3: llama.context_length u32 = 16384 llama_model_loader: - kv 4: llama.embedding_length u32 = 7168 llama_model_loader: - kv 5: llama.feed_forward_length u32 = 19200 llama_model_loader: - kv 6: llama.attention.head_count u32 = 56 llama_model_loader: - kv 7: llama.attention.head_count_kv u32 = 8 llama_model_loader: - kv 8: llama.rope.freq_base f32 = 100000.000000 llama_model_loader: - kv 9: llama.attention.layer_norm_rms_epsilon f32 = 0.000001 llama_model_loader: - kv 10: general.file_type u32 = 0 llama_model_loader: - kv 11: llama.vocab_size u32 = 32256 llama_model_loader: - kv 12: llama.rope.dimension_count u32 = 128 llama_model_loader: - kv 13: llama.rope.scaling.type str = linear llama_model_loader: - kv 14: llama.rope.scaling.factor f32 = 4.000000 llama_model_loader: - kv 15: tokenizer.ggml.model str = gpt2 llama_model_loader: - kv 16: tokenizer.ggml.pre str = deepseek-coder llama_model_loader: - kv 17: tokenizer.ggml.tokens arr[str,32256] = ["!", "\"", "#", "$", "%", "&", "'", ... llama_model_loader: - kv 18: tokenizer.ggml.token_type arr[i32,32256] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ... llama_model_loader: - kv 19: tokenizer.ggml.merges arr[str,31757] = ["Ġ Ġ", "Ġ t", "Ġ a", "i n", "h e... llama_model_loader: - kv 20: tokenizer.ggml.bos_token_id u32 = 32013 llama_model_loader: - kv 21: tokenizer.ggml.eos_token_id u32 = 32021 llama_model_loader: - kv 22: tokenizer.ggml.padding_token_id u32 = 32014 llama_model_loader: - kv 23: tokenizer.ggml.add_bos_token bool = true llama_model_loader: - kv 24: tokenizer.ggml.add_eos_token bool = false llama_model_loader: - kv 25: tokenizer.chat_template str = {% if messages[0]['role'] == 'system'... llama_model_loader: - kv 26: general.quantization_version u32 = 2 llama_model_loader: - type f32: 561 tensors llm_load_vocab: mismatch in special tokens definition ( 243/32256 vs 256/32256 ). llm_load_print_meta: format = GGUF V3 (latest) llm_load_print_meta: arch = llama llm_load_print_meta: vocab type = BPE llm_load_print_meta: n_vocab = 32256 llm_load_print_meta: n_merges = 31757 llm_load_print_meta: n_ctx_train = 16384 llm_load_print_meta: n_embd = 7168 llm_load_print_meta: n_head = 56 llm_load_print_meta: n_head_kv = 8 llm_load_print_meta: n_layer = 62 llm_load_print_meta: n_rot = 128 llm_load_print_meta: n_embd_head_k = 128 llm_load_print_meta: n_embd_head_v = 128 llm_load_print_meta: n_gqa = 7 llm_load_print_meta: n_embd_k_gqa = 1024 llm_load_print_meta: n_embd_v_gqa = 1024 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 = 19200 llm_load_print_meta: n_expert = 0 llm_load_print_meta: n_expert_used = 0 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 = linear llm_load_print_meta: freq_base_train = 100000.0 llm_load_print_meta: freq_scale_train = 0.25 llm_load_print_meta: n_yarn_orig_ctx = 16384 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: model type = ?B llm_load_print_meta: model ftype = all F32 llm_load_print_meta: model params = 33.34 B llm_load_print_meta: model size = 124.21 GiB (32.00 BPW) llm_load_print_meta: general.name = AutoCoder llm_load_print_meta: BOS token = 32013 '<｜begin▁of▁sentence｜>' llm_load_print_meta: EOS token = 32021 '<|EOT|>' llm_load_print_meta: PAD token = 32014 '<｜end▁of▁sentence｜>' llm_load_print_meta: LF token = 126 'Ä' ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: CUDA_USE_TENSOR_CORES: yes 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 = 0.57 MiB llm_load_tensors: offloading 10 repeating layers to GPU llm_load_tensors: offloaded 10/63 layers to GPU llm_load_tensors: CPU buffer size = 127193.42 MiB llm_load_tensors: CUDA0 buffer size = 20230.55 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 = 100000.0 llama_new_context_with_model: freq_scale = 0.25 llama_kv_cache_init: CUDA_Host KV buffer size = 104.00 MiB llama_kv_cache_init: CUDA0 KV buffer size = 20.00 MiB llama_new_context_with_model: KV self size = 124.00 MiB, K (f16): 62.00 MiB, V (f16): 62.00 MiB llama_new_context_with_model: CUDA_Host output buffer size = 0.12 MiB llama_new_context_with_model: CUDA0 compute buffer size = 959.00 MiB llama_new_context_with_model: CUDA_Host compute buffer size = 15.01 MiB llama_new_context_with_model: graph nodes = 1990 llama_new_context_with_model: graph splits = 576 system_info: n_threads = 32 / 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 235.987 ms compute_imatrix: computing over 236 chunks with batch_size 512 compute_imatrix: 85.02 seconds per pass - ETA 5 hours 34.40 minutes [1]6.2367,[2]5.0981,[3]5.3674,[4]6.3093,[5]6.6962,[6]6.5139,[7]5.6034,[8]6.3346,[9]6.2327, save_imatrix: stored collected data after 10 chunks in AutoCoder-IMat-GGUF/imatrix.dat [10]7.0182,[11]7.3080,[12]7.1931,[13]7.8130,[14]7.1667,[15]7.8930,[16]8.0319,[17]8.4303,[18]8.5621,[19]8.8975, save_imatrix: stored collected data after 20 chunks in AutoCoder-IMat-GGUF/imatrix.dat [20]8.7195,[21]9.0143,[22]8.8119,[23]8.3915,[24]8.5450,[25]7.9526,[26]7.5086,[27]7.1777,[28]7.0598,[29]7.1306, save_imatrix: stored collected data after 30 chunks in AutoCoder-IMat-GGUF/imatrix.dat [30]7.2243,[31]7.3420,[32]7.5052,[33]7.7138,[34]7.5603,[35]7.1611,[36]6.8229,[37]6.7698,[38]6.7691,[39]6.7560, save_imatrix: stored collected data after 40 chunks in AutoCoder-IMat-GGUF/imatrix.dat [40]6.7272,[41]6.8435,[42]7.0138,[43]7.1658,[44]7.3762,[45]7.3547,[46]7.4967,[47]7.7038,[48]7.9130,[49]8.1219, save_imatrix: stored collected data after 50 chunks in AutoCoder-IMat-GGUF/imatrix.dat [50]8.2618,[51]8.1807,[52]8.0399,[53]7.8965,[54]7.7444,[55]7.9113,[56]8.0362,[57]8.1050,[58]8.2624,[59]8.3243, save_imatrix: stored collected data after 60 chunks in AutoCoder-IMat-GGUF/imatrix.dat [60]8.4889,[61]8.6123,[62]8.7770,[63]8.8948,[64]8.9956,[65]9.0936,[66]9.1913,[67]9.3509,[68]9.4569,[69]9.5059, save_imatrix: stored collected data after 70 chunks in AutoCoder-IMat-GGUF/imatrix.dat [70]9.5613,[71]9.4594,[72]9.3993,[73]9.3739,[74]9.3351,[75]9.3279,[76]9.2996,[77]9.2548,[78]9.1472,[79]9.1006, save_imatrix: stored collected data after 80 chunks in AutoCoder-IMat-GGUF/imatrix.dat [80]9.1075,[81]9.0669,[82]9.1354,[83]9.1885,[84]9.2184,[85]9.0658,[86]9.0755,[87]8.9919,[88]8.8167,[89]8.8220, save_imatrix: stored collected data after 90 chunks in AutoCoder-IMat-GGUF/imatrix.dat [90]8.8249,[91]8.8873,[92]8.9070,[93]8.9475,[94]8.9910,[95]8.9723,[96]8.9361,[97]8.9400,[98]8.9640,[99]9.0133, save_imatrix: stored collected data after 100 chunks in AutoCoder-IMat-GGUF/imatrix.dat [100]9.0484,[101]9.0443,[102]9.0419,[103]9.0161,[104]9.0064,[105]9.0092,[106]8.9743,[107]8.9700,[108]8.9735,[109]8.9437, save_imatrix: stored collected data after 110 chunks in AutoCoder-IMat-GGUF/imatrix.dat [110]8.9297,[111]8.9006,[112]8.8987,[113]8.8891,[114]8.8714,[115]8.8467,[116]8.8307,[117]8.8270,[118]8.8133,[119]8.7494, save_imatrix: stored collected data after 120 chunks in AutoCoder-IMat-GGUF/imatrix.dat [120]8.7974,[121]8.8334,[122]8.8401,[123]8.8185,[124]8.8434,[125]8.8608,[126]8.8472,[127]8.7689,[128]8.7780,[129]8.7923, save_imatrix: stored collected data after 130 chunks in AutoCoder-IMat-GGUF/imatrix.dat [130]8.7391,[131]8.7488,[132]8.6890,[133]8.6261,[134]8.5618,[135]8.4958,[136]8.4314,[137]8.3654,[138]8.3062,[139]8.2436, save_imatrix: stored collected data after 140 chunks in AutoCoder-IMat-GGUF/imatrix.dat [140]8.1961,[141]8.1307,[142]8.0784,[143]8.0174,[144]7.9439,[145]7.8924,[146]7.8435,[147]7.7859,[148]7.7273,[149]7.6772, save_imatrix: stored collected data after 150 chunks in AutoCoder-IMat-GGUF/imatrix.dat [150]7.6288,[151]7.5735,[152]7.5221,[153]7.4689,[154]7.4131,[155]7.3718,[156]7.3188,[157]7.2882,[158]7.2251,[159]7.1737, save_imatrix: stored collected data after 160 chunks in AutoCoder-IMat-GGUF/imatrix.dat [160]7.1659,[161]7.2120,[162]7.2339,[163]7.2855,[164]7.3378,[165]7.3149,[166]7.3304,[167]7.3299,[168]7.3212,[169]7.3309, save_imatrix: stored collected data after 170 chunks in AutoCoder-IMat-GGUF/imatrix.dat [170]7.3366,[171]7.3420,[172]7.3335,[173]7.3637,[174]7.3576,[175]7.3754,[176]7.3666,[177]7.3789,[178]7.3875,[179]7.3976, save_imatrix: stored collected data after 180 chunks in AutoCoder-IMat-GGUF/imatrix.dat [180]7.3970,[181]7.4175,[182]7.4364,[183]7.4421,[184]7.4629,[185]7.4892,[186]7.5225,[187]7.5317,[188]7.5587,[189]7.5713, save_imatrix: stored collected data after 190 chunks in AutoCoder-IMat-GGUF/imatrix.dat [190]7.5875,[191]7.6106,[192]7.6421,[193]7.6671,[194]7.6765,[195]7.7298,[196]7.7414,[197]7.7303,[198]7.7852,[199]7.8435, save_imatrix: stored collected data after 200 chunks in AutoCoder-IMat-GGUF/imatrix.dat [200]7.8993,[201]7.9646,[202]8.0040,[203]8.0264,[204]8.0384,[205]8.0071,[206]8.0052,[207]8.0328,[208]8.0716,[209]8.0787, save_imatrix: stored collected data after 210 chunks in AutoCoder-IMat-GGUF/imatrix.dat [210]8.0894,[211]8.1033,[212]8.1210,[213]8.1402,[214]8.1424,[215]8.1519,[216]8.1691,[217]8.1997,[218]8.2578,[219]8.2313, save_imatrix: stored collected data after 220 chunks in AutoCoder-IMat-GGUF/imatrix.dat [220]8.2453,[221]8.2315,[222]8.2474,[223]8.2451,[224]8.2428,[225]8.2667,[226]8.2477,[227]8.2620,[228]8.2658,[229]8.3247, save_imatrix: stored collected data after 230 chunks in AutoCoder-IMat-GGUF/imatrix.dat [230]8.3919,[231]8.4580,[232]8.5224,[233]8.5653,[234]8.5411,[235]8.5097,[236]8.4690, save_imatrix: stored collected data after 236 chunks in AutoCoder-IMat-GGUF/imatrix.dat llama_print_timings: load time = 202649.20 ms llama_print_timings: sample time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) llama_print_timings: prompt eval time = 2051921.29 ms / 120832 tokens ( 16.98 ms per token, 58.89 tokens per second) llama_print_timings: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) llama_print_timings: total time = 2170559.57 ms / 120833 tokens Final estimate: PPL = 8.4690 +/- 0.09498