| 70B Instruct | |
| perplexity: calculating perplexity over 564 chunks, n_ctx=512, batch_size=2048, n_seq=4 | |
| perplexity: 451.13 seconds per pass - ETA 17 hours 40.13 minutes | |
| [1]4.9700,[2]5.3351,[3]4.6836,[4]4.3679, | |
| 120B Instruct IQ4_XS | |
| ``` | |
| ./perplexity -f /home/user/Downloads/wikitext-2-raw/wiki.test.raw -m ../../IQ4_XS_120B | |
| main: build = 2709 (40f74e4d) | |
| main: built with cc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0 for x86_64-linux-gnu | |
| main: seed = 1713741077 | |
| llama_model_loader: loaded meta data with 22 key-value pairs and 1263 tensors from ../../IQ4_XS_120B (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 = New Volume | |
| llama_model_loader: - kv 2: llama.vocab_size u32 = 128256 | |
| llama_model_loader: - kv 3: llama.context_length u32 = 8192 | |
| llama_model_loader: - kv 4: llama.embedding_length u32 = 8192 | |
| llama_model_loader: - kv 5: llama.block_count u32 = 140 | |
| llama_model_loader: - kv 6: llama.feed_forward_length u32 = 28672 | |
| llama_model_loader: - kv 7: llama.rope.dimension_count u32 = 128 | |
| llama_model_loader: - kv 8: llama.attention.head_count u32 = 64 | |
| llama_model_loader: - kv 9: llama.attention.head_count_kv u32 = 8 | |
| llama_model_loader: - kv 10: llama.attention.layer_norm_rms_epsilon f32 = 0.000010 | |
| llama_model_loader: - kv 11: llama.rope.freq_base f32 = 500000.000000 | |
| llama_model_loader: - kv 12: general.file_type u32 = 30 | |
| llama_model_loader: - kv 13: tokenizer.ggml.model str = gpt2 | |
| llama_model_loader: - kv 14: tokenizer.ggml.tokens arr[str,128256] = ["!", "\"", "#", "$", "%", "&", "'", ... | |
| llama_model_loader: - kv 15: tokenizer.ggml.scores arr[f32,128256] = [0.000000, 0.000000, 0.000000, 0.0000... | |
| llama_model_loader: - kv 16: tokenizer.ggml.token_type arr[i32,128256] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ... | |
| llama_model_loader: - kv 17: tokenizer.ggml.merges arr[str,280147] = ["Ġ Ġ", "Ġ ĠĠĠ", "ĠĠ ĠĠ", "... | |
| llama_model_loader: - kv 18: tokenizer.ggml.bos_token_id u32 = 128000 | |
| llama_model_loader: - kv 19: tokenizer.ggml.eos_token_id u32 = 128001 | |
| llama_model_loader: - kv 20: tokenizer.chat_template str = {% set loop_messages = messages %}{% ... | |
| llama_model_loader: - kv 21: general.quantization_version u32 = 2 | |
| llama_model_loader: - type f32: 281 tensors | |
| llama_model_loader: - type q5_K: 157 tensors | |
| llama_model_loader: - type q6_K: 1 tensors | |
| llama_model_loader: - type iq4_xs: 824 tensors | |
| llm_load_vocab: special tokens definition check successful ( 256/128256 ). | |
| 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 = 128256 | |
| llm_load_print_meta: n_merges = 280147 | |
| llm_load_print_meta: n_ctx_train = 8192 | |
| llm_load_print_meta: n_embd = 8192 | |
| llm_load_print_meta: n_head = 64 | |
| llm_load_print_meta: n_head_kv = 8 | |
| llm_load_print_meta: n_layer = 140 | |
| 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 = 8 | |
| 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-05 | |
| 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 = 28672 | |
| 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 = 500000.0 | |
| llm_load_print_meta: freq_scale_train = 1 | |
| llm_load_print_meta: n_yarn_orig_ctx = 8192 | |
| 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 = IQ4_XS - 4.25 bpw | |
| llm_load_print_meta: model params = 121.89 B | |
| llm_load_print_meta: model size = 61.35 GiB (4.32 BPW) | |
| llm_load_print_meta: general.name = New Volume | |
| llm_load_print_meta: BOS token = 128000 '<|begin_of_text|>' | |
| llm_load_print_meta: EOS token = 128001 '<|end_of_text|>' | |
| llm_load_print_meta: LF token = 128 'Ä' | |
| llm_load_print_meta: EOT token = 128009 '<|eot_id|>' | |
| llm_load_tensors: ggml ctx size = 0.64 MiB | |
| llm_load_tensors: CPU buffer size = 62823.05 MiB | |
| ................................................................................................... | |
| llama_new_context_with_model: n_ctx = 2048 | |
| llama_new_context_with_model: n_batch = 2048 | |
| llama_new_context_with_model: n_ubatch = 512 | |
| llama_new_context_with_model: freq_base = 500000.0 | |
| llama_new_context_with_model: freq_scale = 1 | |
| llama_kv_cache_init: CPU KV buffer size = 1120.00 MiB | |
| llama_new_context_with_model: KV self size = 1120.00 MiB, K (f16): 560.00 MiB, V (f16): 560.00 MiB | |
| llama_new_context_with_model: CPU output buffer size = 1.96 MiB | |
| llama_new_context_with_model: CPU compute buffer size = 324.01 MiB | |
| llama_new_context_with_model: graph nodes = 4486 | |
| llama_new_context_with_model: graph splits = 1 | |
| system_info: n_threads = 16 / 32 | AVX = 1 | AVX_VNNI = 0 | AVX2 = 1 | AVX512 = 0 | AVX512_VBMI = 0 | AVX512_VNNI = 0 | FMA = 1 | NEON = 0 | ARM_FMA = 0 | F16C = 1 | FP16_VA = 0 | WASM_SIMD = 0 | BLAS = 0 | SSE3 = 1 | SSSE3 = 1 | VSX = 0 | MATMUL_INT8 = 0 | | |
| perplexity: tokenizing the input .. | |
| perplexity: tokenization took 1254.33 ms | |
| perplexity: calculating perplexity over 564 chunks, n_ctx=512, batch_size=2048, n_seq=4 | |
| perplexity: 789.77 seconds per pass - ETA 30 hours 55.97 minutes | |
| [1]6.6257,[2]7.0895,[3]7.4198,[4]7.8431,[5]8.1498,[6]8.2868,[7]8.5416,[8]16.7631,[9]16.9070,[10]17.0177,[11]17.3416,[12]17.1153,[13]17.1495,[14]15.9925,[15]15.6819,[16]14.8677,[17]14.6194,[18]14.7188,[19]14.3299,[20]13.8694,[21]13.7140,[22]13.0065,[23]13.4134,[24]12.9553,[25]12.3752,[26]15.8985,[27]15.3589,[28]14.8173,[29]17.5358,[30]17.0630,[31]16.7671,[32]16.4692,[33]16.2825,[34]16.0765,[35]15.9255,[36]15.9421,[37]15.6712,[38]15.5823,[39]15.4130,[40]15.2759,[41]15.1863,[42]14.8463,[43]14.8449,[44]15.5276,[45]15.5749,[46]15.3805,[47]15.2628,[48]14.9983,[49]14.7855,[50]14.7951,[51]14.8063,[52]15.4632,[53]15.5870,[54]15.4298,[55]15.2675,[56]15.2484,[57]15.0987,[58]16.4450,[59]16.1559,[60]16.0842,[61]16.0550,[62]16.0266,[63]16.0539,[64]16.0921,[65]15.9508,[66]15.8068,[67]15.6202,[68]15.6122,[69]15.6199,[70]15.5773,[71]15.4804,[72]15.2780,[73]15.1460,[74]15.0568,[75]14.8374,[76]14.6901,[77]14.5159,[78]14.4655,[79]14.4193,[80]14.3641,[81]14.2662,[82]14.2472,[83]14.1861,[84]14.1679,[85]14.0669,[86]13.9520,[87]14.0137,[88]13.9075,[89]13.8810,[90]13.8321,[91]13.7511,[92]13.6789, | |
| ``` |