Misc. bug: b4907 breaks RPC #12465
Description
Name and Version
b4907 for rpc server and client both (double checked)
Operating systems
Linux
Which llama.cpp modules do you know to be affected?
llama-server
Command line
rpc-server is started on remote
server is launched with RPC specifiying the remoteProblem description & steps to reproduce
load any model over RPC SEGVs the remote and host
First Bad Commit
b4907
Relevant log output
LOCAL:
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 4070, compute capability 8.9, VMM: yes
build: 4907 (7dfad387) with cc (GCC) 11.2.0 for x86_64-slackware-linux
system info: n_threads = 8, n_threads_batch = 8, total_threads = 16
system_info: n_threads = 8 (n_threads_batch = 8) / 16 | CUDA : ARCHS = 600,610,700,750 | F16 = 1 | USE_GRAPHS = 1 | PEER_MAX_BATCH_SIZE = 128 | CPU : SSE3 = 1 | SSSE3 = 1 | AVX = 1 | AVX2 = 1 | F16C = 1 | FMA = 1 | BMI2 = 1 | LLAMAFILE = 1 | OPENMP = 1 | AARCH64_REPACK = 1 |
main: HTTP server is listening, hostname: 127.0.0.1, port: 8080, http threads: 15
main: loading model
srv load_model: loading model '/data3hd/models/QwQ-32B.IQ4_XS.gguf'
llama_model_load_from_file_impl: using device RPC[192.9.200.5:50052] (RPC[192.9.200.5:50052]) - 11810 MiB free
llama_model_load_from_file_impl: using device CUDA0 (NVIDIA GeForce RTX 4070) - 11593 MiB free
llama_model_loader: loaded meta data with 33 key-value pairs and 771 tensors from /data3hd/models/QwQ-32B.IQ4_XS.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 = qwen2
llama_model_loader: - kv 1: general.type str = model
llama_model_loader: - kv 2: general.name str = QwQ 32B
llama_model_loader: - kv 3: general.basename str = QwQ
llama_model_loader: - kv 4: general.size_label str = 32B
llama_model_loader: - kv 5: general.license str = apache-2.0
llama_model_loader: - kv 6: general.license.link str = https://huggingface.co/Qwen/QWQ-32B/b...
llama_model_loader: - kv 7: general.base_model.count u32 = 1
llama_model_loader: - kv 8: general.base_model.0.name str = Qwen2.5 32B
llama_model_loader: - kv 9: general.base_model.0.organization str = Qwen
llama_model_loader: - kv 10: general.base_model.0.repo_url str = https://huggingface.co/Qwen/Qwen2.5-32B
llama_model_loader: - kv 11: general.tags arr[str,2] = ["chat", "text-generation"]
llama_model_loader: - kv 12: general.languages arr[str,1] = ["en"]
llama_model_loader: - kv 13: qwen2.block_count u32 = 64
llama_model_loader: - kv 14: qwen2.context_length u32 = 131072
llama_model_loader: - kv 15: qwen2.embedding_length u32 = 5120
llama_model_loader: - kv 16: qwen2.feed_forward_length u32 = 27648
llama_model_loader: - kv 17: qwen2.attention.head_count u32 = 40
llama_model_loader: - kv 18: qwen2.attention.head_count_kv u32 = 8
llama_model_loader: - kv 19: qwen2.rope.freq_base f32 = 1000000.000000
llama_model_loader: - kv 20: qwen2.attention.layer_norm_rms_epsilon f32 = 0.000010
llama_model_loader: - kv 21: tokenizer.ggml.model str = gpt2
llama_model_loader: - kv 22: tokenizer.ggml.pre str = qwen2
llama_model_loader: - kv 23: tokenizer.ggml.tokens arr[str,152064] = ["!", "\"", "#", "$", "%", "&", "'", ...
llama_model_loader: - kv 24: tokenizer.ggml.token_type arr[i32,152064] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ...
llama_model_loader: - kv 25: tokenizer.ggml.merges arr[str,151387] = ["Ġ Ġ", "ĠĠ ĠĠ", "i n", "Ġ t",...
llama_model_loader: - kv 26: tokenizer.ggml.eos_token_id u32 = 151645
llama_model_loader: - kv 27: tokenizer.ggml.padding_token_id u32 = 151643
llama_model_loader: - kv 28: tokenizer.ggml.bos_token_id u32 = 151643
llama_model_loader: - kv 29: tokenizer.ggml.add_bos_token bool = false
llama_model_loader: - kv 30: tokenizer.chat_template str = {%- if tools %}\n {{- '<|im_start|>...
llama_model_loader: - kv 31: general.quantization_version u32 = 2
llama_model_loader: - kv 32: general.file_type u32 = 30
llama_model_loader: - type f32: 321 tensors
llama_model_loader: - type q5_K: 72 tensors
llama_model_loader: - type q6_K: 1 tensors
llama_model_loader: - type iq4_xs: 377 tensors
print_info: file format = GGUF V3 (latest)
print_info: file type = IQ4_XS - 4.25 bpw
print_info: file size = 16.64 GiB (4.36 BPW)
load: special tokens cache size = 26
load: token to piece cache size = 0.9311 MB
print_info: arch = qwen2
print_info: vocab_only = 0
print_info: n_ctx_train = 131072
print_info: n_embd = 5120
print_info: n_layer = 64
print_info: n_head = 40
print_info: n_head_kv = 8
print_info: n_rot = 128
print_info: n_swa = 0
print_info: n_swa_pattern = 1
print_info: n_embd_head_k = 128
print_info: n_embd_head_v = 128
print_info: n_gqa = 5
print_info: n_embd_k_gqa = 1024
print_info: n_embd_v_gqa = 1024
print_info: f_norm_eps = 0.0e+00
print_info: f_norm_rms_eps = 1.0e-05
print_info: f_clamp_kqv = 0.0e+00
print_info: f_max_alibi_bias = 0.0e+00
print_info: f_logit_scale = 0.0e+00
print_info: f_attn_scale = 0.0e+00
print_info: n_ff = 27648
print_info: n_expert = 0
print_info: n_expert_used = 0
print_info: causal attn = 1
print_info: pooling type = 0
print_info: rope type = 2
print_info: rope scaling = linear
print_info: freq_base_train = 1000000.0
print_info: freq_scale_train = 1
print_info: n_ctx_orig_yarn = 131072
print_info: rope_finetuned = unknown
print_info: ssm_d_conv = 0
print_info: ssm_d_inner = 0
print_info: ssm_d_state = 0
print_info: ssm_dt_rank = 0
print_info: ssm_dt_b_c_rms = 0
print_info: model type = 32B
print_info: model params = 32.76 B
print_info: general.name = QwQ 32B
print_info: rope_yarn_log_mul = 0.0000
print_info: vocab type = BPE
print_info: n_vocab = 152064
print_info: n_merges = 151387
print_info: BOS token = 151643 '<|endoftext|>'
print_info: EOS token = 151645 '<|im_end|>'
print_info: EOT token = 151645 '<|im_end|>'
print_info: PAD token = 151643 '<|endoftext|>'
print_info: LF token = 198 'Ċ'
print_info: FIM PRE token = 151659 '<|fim_prefix|>'
print_info: FIM SUF token = 151661 '<|fim_suffix|>'
print_info: FIM MID token = 151660 '<|fim_middle|>'
print_info: FIM PAD token = 151662 '<|fim_pad|>'
print_info: FIM REP token = 151663 '<|repo_name|>'
print_info: FIM SEP token = 151664 '<|file_sep|>'
print_info: EOG token = 151643 '<|endoftext|>'
print_info: EOG token = 151645 '<|im_end|>'
print_info: EOG token = 151662 '<|fim_pad|>'
print_info: EOG token = 151663 '<|repo_name|>'
print_info: EOG token = 151664 '<|file_sep|>'
print_info: max token length = 256
load_tensors: loading model tensors, this can take a while... (mmap = true)
load_tensors: offloading 64 repeating layers to GPU
load_tensors: offloading output layer to GPU
load_tensors: offloaded 65/65 layers to GPU
load_tensors: RPC[192.9.200.5:50052] model buffer size = 9092.39 MiB
load_tensors: CPU_Mapped model buffer size = 394.45 MiB
load_tensors: CUDA0 model buffer size = 7549.71 MiB
................................................................................................
llama_context: constructing llama_context
llama_context: n_seq_max = 1
llama_context: n_ctx = 15104
llama_context: n_ctx_per_seq = 15104
llama_context: n_batch = 128
llama_context: n_ubatch = 128
llama_context: causal_attn = 1
llama_context: flash_attn = 1
llama_context: freq_base = 1000000.0
llama_context: freq_scale = 1
llama_context: yarn_log_mul = 0
llama_context: n_ctx_per_seq (15104) < n_ctx_train (131072) -- the full capacity of the model will not be utilized
llama_context: CUDA_Host output buffer size = 0.58 MiB
init: kv_size = 15104, offload = 1, type_k = 'f16', type_v = 'f16', n_layer = 64, can_shift = 1
init: RPC[192.9.200.5:50052] KV buffer size = 2124.00 MiB
init: CUDA0 KV buffer size = 1652.00 MiB
llama_context: KV self size = 3776.00 MiB, K (f16): 1888.00 MiB, V (f16): 1888.00 MiB
llama_context: RPC[192.9.200.5:50052] compute buffer size = 40.56 MiB
llama_context: CUDA0 compute buffer size = 81.13 MiB
llama_context: CPU compute buffer size = 9.88 MiB
llama_context: graph nodes = 1991
llama_context: graph splits = 3
common_init_from_params: setting dry_penalty_last_n to ctx_size = 15104
common_init_from_params: warming up the model with an empty run - please wait ... (--no-warmup to disable)
/usr/local/src/ai/llamacpp/llama.cpp/ggml/src/ggml-rpc/ggml-rpc.cpp:682: GGML_ASSERT(status) failed
[New LWP 25218]
[New LWP 25222]
[New LWP 25223]
[New LWP 25224]
[New LWP 25225]
[New LWP 25226]
[New LWP 25227]
[New LWP 25228]
[New LWP 25229]
[New LWP 25230]
[New LWP 25231]
[New LWP 25232]
[New LWP 25233]
[New LWP 25234]
[New LWP 25235]
[New LWP 25236]
[New LWP 25237]
[New LWP 25238]
[New LWP 25239]
[New LWP 25240]
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib64/libthread_db.so.1".
0x00007f93a2d523c7 in wait4 () from /lib64/libc.so.6
#0 0x00007f93a2d523c7 in wait4 () from /lib64/libc.so.6
#1 0x00007f93a320c1c1 in ggml_abort () from /usr/lib64/libggml-base.so
#2 0x00007f93a32d369b in ggml_backend_rpc_graph_compute(ggml_backend*, ggml_cgraph*) () from /usr/lib64/libggml-rpc.so
#3 0x00007f93a32212a5 in ggml_backend_sched_graph_compute_async () from /usr/lib64/libggml-base.so
#4 0x00007f93b13c9901 in llama_context::graph_compute(ggml_cgraph*, bool) () from /usr/lib64/libllama.so
#5 0x00007f93b13cba8a in llama_context::decode(llama_batch&) () from /usr/lib64/libllama.so
#6 0x00007f93b13ccdbc in llama_decode () from /usr/lib64/libllama.so
#7 0x00000000005976d2 in common_init_from_params(common_params&) ()
#8 0x000000000049e4c1 in server_context::load_model(common_params const&) ()
#9 0x0000000000434665 in main ()
[Inferior 1 (process 25217) detached]
REMOTE:
create_backend: using CUDA backend
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 4070, compute capability 8.9, VMM: yes
Starting RPC server on 0.0.0.0:50052, backend memory: 11810 MB
Accepted client connection, free_mem=12384534528, total_mem=12592087040
Client connection closed
Accepted client connection, free_mem=12384534528, total_mem=12592087040
Client connection closed
Accepted client connection, free_mem=12384534528, total_mem=12592087040
Client connection closed
Accepted client connection, free_mem=12384534528, total_mem=12592087040
check_node_graph_compatibility_and_refresh_copy_ops: disabling CUDA graphs due to mul_mat_id
/usr/local/bin/ll_startrpc: line 14: 25876 Segmentation fault rpc-server -H 0.0.0.0 -p 50052