server : crash when -b > -ub with embeddings

[ggerganov]

@ggerganov Ok, I did few tests and apparently there's an issue that is subject to a separate issue.

Using the following command:

llama-server ... -ub 4096 -b 4096 -c 4096 -np 4

Everything works pretty much as expected. Amount of tokens that a task slot can handle appears to be ub / np. So in this example, each slot gets a 1024 tokens window. This does seem to give a nice boost depending on the embeddings chunking strategy (my current embeddings are up to 1024 tokens), but I haven't measured precisely yet.

However, using the following command:

llama-server ... -ub 1024 -b 4096 -c 4096 -np 4

The server crashes with GGML_ASSERT((cparams.causal_attn || cparams.n_ubatch >= n_tokens_all) && "non-causal attention requires n_ubatch >= n_tokens") failed as soon as it receives the next batch of tasks:

ggml_vulkan: Found 1 Vulkan devices:
ggml_vulkan: 0 = AMD Radeon RX 6600M (AMD proprietary driver) | uma: 0 | fp16: 1 | warp size: 32 | shared memory: 32768 | int dot: 1 | matrix cores: none
build: 5080 (997b1b42) with MSVC 19.38.33134.0 for x64
system info: n_threads = 8, n_threads_batch = 8, total_threads = 16

system_info: n_threads = 8 (n_threads_batch = 8) / 16 | CPU : SSE3 = 1 | SSSE3 = 1 | AVX = 1 | AVX2 = 1 | F16C = 1 | FMA = 1 | LLAMAFILE = 1 | OPENMP = 1 | AARCH64_REPACK = 1 |

main: binding port with default address family
main: HTTP server is listening, hostname: 192.168.0.2, port: 8080, http threads: 15
main: loading model
srv    load_model: loading model 'C:\Temp\snowflake-arctic-embed-l-v2.0-q8_0.gguf'
llama_model_load_from_file_impl: using device Vulkan0 (AMD Radeon RX 6600M) - 8176 MiB free
llama_model_loader: loaded meta data with 36 key-value pairs and 389 tensors from C:\Temp\snowflake-arctic-embed-l-v2.0-q8_0.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              = bert
llama_model_loader: - kv   1:                               general.type str              = model
llama_model_loader: - kv   2:                               general.name str              = Snowflake Arctic Embed L v2.0
llama_model_loader: - kv   3:                            general.version str              = v2.0
llama_model_loader: - kv   4:                           general.basename str              = snowflake-arctic-embed-l
llama_model_loader: - kv   5:                         general.size_label str              = 567M
llama_model_loader: - kv   6:                            general.license str              = apache-2.0
llama_model_loader: - kv   7:                               general.tags arr[str,8]       = ["sentence-transformers", "feature-ex...
llama_model_loader: - kv   8:                          general.languages arr[str,74]      = ["af", "ar", "az", "be", "bg", "bn", ...
llama_model_loader: - kv   9:                           bert.block_count u32              = 24
llama_model_loader: - kv  10:                        bert.context_length u32              = 8192
llama_model_loader: - kv  11:                      bert.embedding_length u32              = 1024
llama_model_loader: - kv  12:                   bert.feed_forward_length u32              = 4096
llama_model_loader: - kv  13:                  bert.attention.head_count u32              = 16
llama_model_loader: - kv  14:          bert.attention.layer_norm_epsilon f32              = 0.000010
llama_model_loader: - kv  15:                          general.file_type u32              = 7
llama_model_loader: - kv  16:                      bert.attention.causal bool             = false
llama_model_loader: - kv  17:                          bert.pooling_type u32              = 2
llama_model_loader: - kv  18:                       tokenizer.ggml.model str              = t5
llama_model_loader: - kv  19:                         tokenizer.ggml.pre str              = default
llama_model_loader: - kv  20:                      tokenizer.ggml.tokens arr[str,250002]  = ["<s>", "<pad>", "</s>", "<unk>", ","...
llama_model_loader: - kv  21:                      tokenizer.ggml.scores arr[f32,250002]  = [0.000000, 0.000000, 0.000000, 0.0000...
llama_model_loader: - kv  22:                  tokenizer.ggml.token_type arr[i32,250002]  = [3, 3, 3, 2, 1, 1, 1, 1, 1, 1, 1, 1, ...
llama_model_loader: - kv  23:            tokenizer.ggml.add_space_prefix bool             = true
llama_model_loader: - kv  24:            tokenizer.ggml.token_type_count u32              = 1
llama_model_loader: - kv  25:    tokenizer.ggml.remove_extra_whitespaces bool             = true
llama_model_loader: - kv  26:        tokenizer.ggml.precompiled_charsmap arr[u8,237539]   = [0, 180, 2, 0, 0, 132, 0, 0, 0, 0, 0,...
llama_model_loader: - kv  27:                tokenizer.ggml.bos_token_id u32              = 0
llama_model_loader: - kv  28:                tokenizer.ggml.eos_token_id u32              = 2
llama_model_loader: - kv  29:            tokenizer.ggml.unknown_token_id u32              = 3
llama_model_loader: - kv  30:          tokenizer.ggml.seperator_token_id u32              = 2
llama_model_loader: - kv  31:            tokenizer.ggml.padding_token_id u32              = 1
llama_model_loader: - kv  32:               tokenizer.ggml.mask_token_id u32              = 250001
llama_model_loader: - kv  33:               tokenizer.ggml.add_bos_token bool             = true
llama_model_loader: - kv  34:               tokenizer.ggml.add_eos_token bool             = true
llama_model_loader: - kv  35:               general.quantization_version u32              = 2
llama_model_loader: - type  f32:  244 tensors
llama_model_loader: - type q8_0:  145 tensors
print_info: file format = GGUF V3 (latest)
print_info: file type   = Q8_0
print_info: file size   = 598.63 MiB (8.86 BPW)
load: model vocab missing newline token, using special_pad_id instead
load: special_eos_id is not in special_eog_ids - the tokenizer config may be incorrect
load: special tokens cache size = 4
load: token to piece cache size = 2.1668 MB
print_info: arch             = bert
print_info: vocab_only       = 0
print_info: n_ctx_train      = 8192
print_info: n_embd           = 1024
print_info: n_layer          = 24
print_info: n_head           = 16
print_info: n_head_kv        = 16
print_info: n_rot            = 64
print_info: n_swa            = 0
print_info: n_swa_pattern    = 1
print_info: n_embd_head_k    = 64
print_info: n_embd_head_v    = 64
print_info: n_gqa            = 1
print_info: n_embd_k_gqa     = 1024
print_info: n_embd_v_gqa     = 1024
print_info: f_norm_eps       = 1.0e-05
print_info: f_norm_rms_eps   = 0.0e+00
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             = 4096
print_info: n_expert         = 0
print_info: n_expert_used    = 0
print_info: causal attn      = 0
print_info: pooling type     = 2
print_info: rope type        = 2
print_info: rope scaling     = linear
print_info: freq_base_train  = 10000.0
print_info: freq_scale_train = 1
print_info: n_ctx_orig_yarn  = 8192
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       = 335M
print_info: model params     = 566.70 M
print_info: general.name     = Snowflake Arctic Embed L v2.0
print_info: vocab type       = UGM
print_info: n_vocab          = 250002
print_info: n_merges         = 0
print_info: BOS token        = 0 '<s>'
print_info: EOS token        = 2 '</s>'
print_info: UNK token        = 3 '<unk>'
print_info: SEP token        = 2 '</s>'
print_info: PAD token        = 1 '<pad>'
print_info: MASK token       = 250001 '[PAD250000]'
print_info: LF token         = 0 '<s>'
print_info: EOG token        = 2 '</s>'
print_info: max token length = 48
load_tensors: loading model tensors, this can take a while... (mmap = true)
load_tensors: offloading 24 repeating layers to GPU
load_tensors: offloading output layer to GPU
load_tensors: offloaded 25/25 layers to GPU
load_tensors:      Vulkan0 model buffer size =   307.22 MiB
load_tensors:   CPU_Mapped model buffer size =   291.41 MiB
......................................................
llama_context: constructing llama_context
llama_context: n_seq_max     = 3
llama_context: n_ctx         = 4096
llama_context: n_ctx_per_seq = 1365
llama_context: n_batch       = 4096
llama_context: n_ubatch      = 1024
llama_context: causal_attn   = 0
llama_context: flash_attn    = 0
llama_context: freq_base     = 10000.0
llama_context: freq_scale    = 1
llama_context: n_ctx_per_seq (1365) < n_ctx_train (8192) -- the full capacity of the model will not be utilized
llama_context: Vulkan_Host  output buffer size =     0.00 MiB
init: kv_size = 4096, offload = 1, type_k = 'f16', type_v = 'f16', n_layer = 24, can_shift = 1
init:    Vulkan0 KV buffer size =   384.00 MiB
llama_context: KV self size  =  384.00 MiB, K (f16):  192.00 MiB, V (f16):  192.00 MiB
llama_context:    Vulkan0 compute buffer size =    88.01 MiB
llama_context: Vulkan_Host compute buffer size =    12.01 MiB
llama_context: graph nodes  = 825
llama_context: graph splits = 4 (with bs=1024), 2 (with bs=1)
common_init_from_params: setting dry_penalty_last_n to ctx_size = 4096
common_init_from_params: warming up the model with an empty run - please wait ... (--no-warmup to disable)
srv          init: initializing slots, n_slots = 3
slot         init: id  0 | task -1 | new slot n_ctx_slot = 1365
slot         init: id  1 | task -1 | new slot n_ctx_slot = 1365
slot         init: id  2 | task -1 | new slot n_ctx_slot = 1365
main: model loaded
main: chat template, chat_template: {%- for message in messages -%}
  {{- '<|im_start|>' + message.role + '
' + message.content + '<|im_end|>
' -}}
{%- endfor -%}
{%- if add_generation_prompt -%}
  {{- '<|im_start|>assistant
' -}}
{%- endif -%}, example_format: '<|im_start|>system
You are a helpful assistant<|im_end|>
<|im_start|>user
Hello<|im_end|>
<|im_start|>assistant
Hi there<|im_end|>
<|im_start|>user
How are you?<|im_end|>
<|im_start|>assistant
'
main: server is listening on http://192.168.0.2:8080 - starting the main loop
srv  update_slots: all slots are idle
slot launch_slot_: id  0 | task 0 | processing task
slot update_slots: id  0 | task 0 | new prompt, n_ctx_slot = 1365, n_keep = 0, n_prompt_tokens = 830
slot update_slots: id  0 | task 0 | kv cache rm [0, end)
slot update_slots: id  0 | task 0 | prompt processing progress, n_past = 830, n_tokens = 830, progress = 1.000000
slot update_slots: id  0 | task 0 | prompt done, n_past = 830, n_tokens = 830
slot      release: id  0 | task 0 | stop processing: n_past = 830, truncated = 0
slot launch_slot_: id  1 | task 2 | processing task
slot launch_slot_: id  2 | task 3 | processing task
slot launch_slot_: id  0 | task 4 | processing task
slot update_slots: id  0 | task 4 | new prompt, n_ctx_slot = 1365, n_keep = 0, n_prompt_tokens = 255
srv  log_server_r: request: POST /v1/embeddings 192.168.0.7 200
slot update_slots: id  0 | task 4 | kv cache rm [0, end)
slot update_slots: id  0 | task 4 | prompt processing progress, n_past = 255, n_tokens = 255, progress = 1.000000
slot update_slots: id  0 | task 4 | prompt done, n_past = 255, n_tokens = 255
slot update_slots: id  1 | task 2 | new prompt, n_ctx_slot = 1365, n_keep = 0, n_prompt_tokens = 852
slot update_slots: id  1 | task 2 | kv cache rm [0, end)
slot update_slots: id  1 | task 2 | prompt processing progress, n_past = 852, n_tokens = 1107, progress = 1.000000
slot update_slots: id  1 | task 2 | prompt done, n_past = 852, n_tokens = 1107
slot update_slots: id  2 | task 3 | new prompt, n_ctx_slot = 1365, n_keep = 0, n_prompt_tokens = 246
slot update_slots: id  2 | task 3 | kv cache rm [0, end)
slot update_slots: id  2 | task 3 | prompt processing progress, n_past = 246, n_tokens = 1353, progress = 1.000000
slot update_slots: id  2 | task 3 | prompt done, n_past = 246, n_tokens = 1353
C:\Sources\llama.cpp\src\llama-context.cpp:1220: GGML_ASSERT((cparams.causal_attn || cparams.n_ubatch >= n_tokens_all) && "non-causal attention requires n_ubatch >= n_tokens") failed

Originally posted by @deiteris in #12817

[ahmedshakill]

I would like to work on this.

[ahmedshakill]

My setup Ryzen 5700g - Ubuntu 24.04 - Cuda (not vulkan) on RTX 3050 with 6GB VRAM.

I spent some time and reproduced errors discussed in the parent issue. Should I consider crashing upon -b > -ub with embeddings? How should I proceed?

Reproducing error "input is too large to process. increase the physical batch size"

./llama-server --embeddings -ub 512 -b 1024 -c 4096 --host 127.0.0.1 --port 8080 -hf nomic-ai/nomic-embed-text-v1.5-GGUF:Q8_0 -ngl 99

--url http://localhost:8080/embedding \
--header "Content-Type: application/json" \
--data '{"content": "Building a website can be done in 10 simple steps:  27325,-0.03392419591546059,0.01808857172727585,0.032047469168901443,0.013507853262126446,-0.013308085501194,-0.0025945587549358606,2176203728,-0.0019427496008574963,-0.029237613081932068,-0.02679949626326561,0.005496430676430464,-0.020501896739006042,-0.01972820982336998,-0.047096554189920425,-0.012842271476984024,0.049323391169309616,0.02657255530357361,-0.0887509137392044,-0.03228742629289627,0.002886326750740409,0.01705707423388958,0.025091011077165604,0.04601024091243744,0.0203191377222538,-0.020455339923501015,0.009682172909379005,-0.02244047075510025,-0.0029785167425870895,-0.06640050560235977,-0.040286388248205185,0.0608222559094429,0.013841903768479824,-0.02146606333553791,0.049076247960329056,0.016104362905025482,0.04573232680559158,-0.010439070872962475,0.01650935970246792,-0.034356456249952316,-0.07073406130075455,0.047809623181819916,-0.0419883094727993,-0.0789145901799202,-0.08123131841421127,-0.002287890762090683,0.037756677716970444,0.09504104405641556,-0.043209508061409,0.03941306844353676,-0.008498760871589184,0.01557685062289238,-0.021454235538840294,-0.07289224863052368,0.05087573453783989,-0.027329236268997192,0.014062173664569855,-0.0479254275560379,-0.02572731301188469,-0.03264648839831352,-0.04332427307963371,0.014104863628745079,0.021585993468761444,-3.491506140562706e,-0.025166962295770645,0.006997351534664631,-0.030485186725854874,-0.02484430931508541,0.05264708399772644,0.007357002701610327,0.015226254239678383,0.019251862540841103,-0.01562185585498809 } 

it returns this

{"error":{"code":500,"message":"input is too large to process. increase the physical batch size","type":"server_error"}}% 

It also fails in 'completion' mode when ran without '--embedding'.

Reproducing error "non-causal attention requires n_ubatch >= n_tokens"

./llama-server -ub 1024 -b 4096 -c 4096 --host 127.0.0.1 --port 8080 -hf nomic-ai/nomic-embed-text-v1.5-GGUF:Q8_0 -ngl 99

...

main: server is listening on http://127.0.0.1:8080 - starting the main loop
srv  update_slots: all slots are idle
srv  log_server_r: request: GET / 127.0.0.1 200
srv  params_from_: Chat format: Content-only
slot launch_slot_: id  0 | task 0 | processing task
slot update_slots: id  0 | task 0 | new prompt, n_ctx_slot = 4096, n_keep = 0, n_prompt_tokens = 53
slot update_slots: id  0 | task 0 | kv cache rm [0, end)
slot update_slots: id  0 | task 0 | prompt processing progress, n_past = 53, n_tokens = 53, progress = 1.000000
slot update_slots: id  0 | task 0 | prompt done, n_past = 53, n_tokens = 53
slot process_toke: id  0 | task 0 | n_predict (-1) is set for infinite generation. Limiting generated tokens to n_ctx_train (2048) to avoid EOS-less generation infinite loop
slot      release: id  0 | task 0 | stop processing: n_past = 2047, truncated = 1
slot print_timing: id  0 | task 0 | 
prompt eval time =      28.88 ms /    53 tokens (    0.54 ms per token,  1834.93 tokens per second)
       eval time =    4746.05 ms /  1995 tokens (    2.38 ms per token,   420.35 tokens per second)
      total time =    4774.93 ms /  2048 tokens
srv  update_slots: all slots are idle
srv  log_server_r: request: POST /v1/chat/completions 127.0.0.1 200
srv  params_from_: Chat format: Content-only
slot launch_slot_: id  0 | task 1996 | processing task
slot update_slots: id  0 | task 1996 | new prompt, n_ctx_slot = 4096, n_keep = 0, n_prompt_tokens = 7850
slot update_slots: id  0 | task 1996 | input truncated, n_ctx = 4096, n_keep = 0, n_left = 4096, n_prompt_tokens = 3754
slot update_slots: id  0 | task 1996 | kv cache rm [0, end)
slot update_slots: id  0 | task 1996 | prompt processing progress, n_past = 3754, n_tokens = 3754, progress = 1.000000
slot update_slots: id  0 | task 1996 | prompt done, n_past = 3754, n_tokens = 3754
/home/shakil/ext/repos/llama.cpp/src/llama-context.cpp:1220: GGML_ASSERT((cparams.causal_attn || cparams.n_ubatch >= n_tokens_all) && "non-causal attention requires n_ubatch >= n_tokens") failed
Could not attach to process.  If your uid matches the uid of the target
process, check the setting of /proc/sys/kernel/yama/ptrace_scope, or try
again as the root user.  For more details, see /etc/sysctl.d/10-ptrace.conf
ptrace: Operation not permitted.
No stack.
The program is not being run.
[1]    414900 IOT instruction (core dumped)  ./llama-server -ub 1024 -b 4096 -c 4096 --host 127.0.0.1 --port 8080 -hf  -ng 

[ggerganov]

I haven't thought through deeply about the logic, but probably the most simple solution is to add a check in server.cpp:

// sanity-check parameters
if embeddings are enabled
  if batch > ubatch
    warn("setting batch = ubatch")
    batch = ubatch