Q4 potentially broken
#2
by
terribleplan
- opened
When attempting to load the Q4_K_M quant of the model I get the following error:
ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no
ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
ggml_cuda_init: found 3 CUDA devices:
Device 0: NVIDIA GeForce RTX 3090 Ti, compute capability 8.6, VMM: yes
Device 1: NVIDIA GeForce RTX 3090 Ti, compute capability 8.6, VMM: yes
Device 2: NVIDIA GeForce RTX 3090, compute capability 8.6, VMM: yes
build: 4412 (dfffe676) with cc (GCC) 11.5.0 20240719 (Red Hat 11.5.0-2) for x86_64-redhat-linux
system info: n_threads = 64, n_threads_batch = 64, total_threads = 128
system_info: n_threads = 64 (n_threads_batch = 64) / 128 | CUDA : ARCHS = 860 | USE_GRAPHS = 1 | PEER_MAX_BATCH_SIZE = 128 | CPU : SSE3 = 1 | SSSE3 = 1 | AVX = 1 | AVX2 = 1 | F16C = 1 | FMA = 1 | LLAMAFILE = 1 | OPENMP = 1 | AARCH64_REPACK = 1 |
main: HTTP server is listening, hostname: 0.0.0.0, port: 8080, http threads: 127
main: loading model
srv load_model: loading model '/home/ml/models/gguf/bullerwins_DeepSeek-V3-GGUF/DeepSeek-V3-Q4_K_M/DeepSeek-V3-Q4_K_M-00001-of-00010.gguf'
llama_load_model_from_file: using device CUDA0 (NVIDIA GeForce RTX 3090 Ti) - 23869 MiB free
llama_load_model_from_file: using device CUDA1 (NVIDIA GeForce RTX 3090 Ti) - 23869 MiB free
llama_load_model_from_file: using device CUDA2 (NVIDIA GeForce RTX 3090) - 23887 MiB free
llama_model_loader: additional 9 GGUFs metadata loaded.
llama_model_loader: loaded meta data with 51 key-value pairs and 1025 tensors from /home/ml/models/gguf/bullerwins_DeepSeek-V3-GGUF/DeepSeek-V3-Q4_K_M/DeepSeek-V3-Q4_K_M-00001-of-00010.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 = Models4Tb
llama_model_loader: - kv 3: general.size_label str = 256x20B
llama_model_loader: - kv 4: general.base_model.count u32 = 1
llama_model_loader: - kv 5: general.base_model.0.name str = DeepSeek V3
llama_model_loader: - kv 6: general.base_model.0.version str = V3
llama_model_loader: - kv 7: general.base_model.0.organization str = Deepseek Ai
llama_model_loader: - kv 8: general.base_model.0.repo_url str = https://huggingface.co/deepseek-ai/De...
llama_model_loader: - kv 9: deepseek2.block_count u32 = 61
llama_model_loader: - kv 10: deepseek2.context_length u32 = 163840
llama_model_loader: - kv 11: deepseek2.embedding_length u32 = 7168
llama_model_loader: - kv 12: deepseek2.feed_forward_length u32 = 18432
llama_model_loader: - kv 13: deepseek2.attention.head_count u32 = 128
llama_model_loader: - kv 14: deepseek2.attention.head_count_kv u32 = 128
llama_model_loader: - kv 15: deepseek2.rope.freq_base f32 = 10000.000000
llama_model_loader: - kv 16: deepseek2.attention.layer_norm_rms_epsilon f32 = 0.000001
llama_model_loader: - kv 17: deepseek2.expert_used_count u32 = 8
llama_model_loader: - kv 18: general.file_type u32 = 15
llama_model_loader: - kv 19: deepseek2.leading_dense_block_count u32 = 3
llama_model_loader: - kv 20: deepseek2.vocab_size u32 = 129280
llama_model_loader: - kv 21: deepseek2.attention.q_lora_rank u32 = 1536
llama_model_loader: - kv 22: deepseek2.attention.kv_lora_rank u32 = 512
llama_model_loader: - kv 23: deepseek2.attention.key_length u32 = 192
llama_model_loader: - kv 24: deepseek2.attention.value_length u32 = 128
llama_model_loader: - kv 25: deepseek2.expert_feed_forward_length u32 = 2048
llama_model_loader: - kv 26: deepseek2.expert_count u32 = 256
llama_model_loader: - kv 27: deepseek2.expert_shared_count u32 = 1
llama_model_loader: - kv 28: deepseek2.expert_weights_scale f32 = 2.500000
llama_model_loader: - kv 29: deepseek2.expert_weights_norm bool = true
llama_model_loader: - kv 30: deepseek2.expert_gating_func u32 = 2
llama_model_loader: - kv 31: deepseek2.rope.dimension_count u32 = 64
llama_model_loader: - kv 32: deepseek2.rope.scaling.type str = yarn
llama_model_loader: - kv 33: deepseek2.rope.scaling.factor f32 = 40.000000
llama_model_loader: - kv 34: deepseek2.rope.scaling.original_context_length u32 = 4096
llama_model_loader: - kv 35: deepseek2.rope.scaling.yarn_log_multiplier f32 = 0.100000
llama_model_loader: - kv 36: tokenizer.ggml.model str = gpt2
llama_model_loader: - kv 37: tokenizer.ggml.pre str = deepseek-v3
llama_model_loader: - kv 38: tokenizer.ggml.tokens arr[str,129280] = ["<｜begin▁of▁sentence｜>", "<�...
llama_model_loader: - kv 39: tokenizer.ggml.token_type arr[i32,129280] = [3, 3, 3, 1, 1, 1, 1, 1, 1, 1, 1, 1, ...
llama_model_loader: - kv 40: tokenizer.ggml.merges arr[str,127741] = ["Ġ t", "Ġ a", "i n", "Ġ Ġ", "h e...
llama_model_loader: - kv 41: tokenizer.ggml.bos_token_id u32 = 0
llama_model_loader: - kv 42: tokenizer.ggml.eos_token_id u32 = 1
llama_model_loader: - kv 43: tokenizer.ggml.padding_token_id u32 = 1
llama_model_loader: - kv 44: tokenizer.ggml.add_bos_token bool = true
llama_model_loader: - kv 45: tokenizer.ggml.add_eos_token bool = false
llama_model_loader: - kv 46: tokenizer.chat_template str = {% if not add_generation_prompt is de...
llama_model_loader: - kv 47: general.quantization_version u32 = 2
llama_model_loader: - kv 48: split.no u16 = 0
llama_model_loader: - kv 49: split.count u16 = 10
llama_model_loader: - kv 50: split.tensors.count i32 = 1025
llama_model_loader: - type f32: 361 tensors
llama_model_loader: - type q4_K: 606 tensors
llama_model_loader: - type q6_K: 58 tensors
llm_load_vocab: special_eos_id is not in special_eog_ids - the tokenizer config may be incorrect
llm_load_vocab: special tokens cache size = 818
llm_load_vocab: token to piece cache size = 0.8223 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 = 129280
llm_load_print_meta: n_merges = 127741
llm_load_print_meta: vocab_only = 0
llm_load_print_meta: n_ctx_train = 163840
llm_load_print_meta: n_embd = 7168
llm_load_print_meta: n_layer = 61
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 = 18432
llm_load_print_meta: n_expert = 256
llm_load_print_meta: n_expert_used = 8
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 = 671B
llm_load_print_meta: model ftype = Q4_K - Medium
llm_load_print_meta: model params = 671.03 B
llm_load_print_meta: model size = 376.65 GiB (4.82 BPW)
llm_load_print_meta: general.name = Models4Tb
llm_load_print_meta: BOS token = 0 '<｜begin▁of▁sentence｜>'
llm_load_print_meta: EOS token = 1 '<｜end▁of▁sentence｜>'
llm_load_print_meta: EOT token = 1 '<｜end▁of▁sentence｜>'
llm_load_print_meta: PAD token = 1 '<｜end▁of▁sentence｜>'
llm_load_print_meta: LF token = 131 'Ä'
llm_load_print_meta: FIM PRE token = 128801 '<｜fim▁begin｜>'
llm_load_print_meta: FIM SUF token = 128800 '<｜fim▁hole｜>'
llm_load_print_meta: FIM MID token = 128802 '<｜fim▁end｜>'
llm_load_print_meta: EOG token = 1 '<｜end▁of▁sentence｜>'
llm_load_print_meta: max token length = 256
llm_load_print_meta: n_layer_dense_lead = 3
llm_load_print_meta: n_lora_q = 1536
llm_load_print_meta: n_lora_kv = 512
llm_load_print_meta: n_ff_exp = 2048
llm_load_print_meta: n_expert_shared = 1
llm_load_print_meta: expert_weights_scale = 2.5
llm_load_print_meta: expert_weights_norm = 1
llm_load_print_meta: expert_gating_func = sigmoid
llm_load_print_meta: rope_yarn_log_mul = 0.1000
llama_model_load: error loading model: done_getting_tensors: wrong number of tensors; expected 1025, got 967
llama_load_model_from_file: failed to load model
common_init_from_params: failed to load model '/home/ml/models/gguf/bullerwins_DeepSeek-V3-GGUF/DeepSeek-V3-Q4_K_M/DeepSeek-V3-Q4_K_M-00001-of-00010.gguf'
srv load_model: failed to load model, '/home/ml/models/gguf/bullerwins_DeepSeek-V3-GGUF/DeepSeek-V3-Q4_K_M/DeepSeek-V3-Q4_K_M-00001-of-00010.gguf'
main: exiting due to model loading error
I am sadly a little short of RAM to run the Q8_0, so can't double check the issue isn't my build of the PR branch into lcpp by using that instead.
I'm getting the same llama_model_load: error loading model: done_getting_tensors: wrong number of tensors; expected 1025, got 967 error, for all versions - DeepSeek-V3-GGUF-bf16, DeepSeek-V3-Q4_K_M, and DeepSeek-V3-Q8_0.
EDIT: Have tried with the fairydreaming fork as well as the ggerganov master branch + PR #11049, compiled from source on both counts, CPU only.
Hi!
Yes, it needs this commit https://github.com/ggerganov/llama.cpp/pull/11049/commits/d2f784d50d3b64ce247a29f7c449bd255fe6e18a, on newer commits there has been newer changes that break the gguf's
I'm currently requanting anyways to support the newer version.
It works! Checked out fairydreaming:deepseek-v3 and reset the head to d2f7. Getting about 2 tok/sec on CPU only 2x8175M with 512GB 2400 DDR4.