github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-07 00:22:43 -05:00
Code Issues 14.5k Packages Projects Releases 120 Wiki Activity

[GH-ISSUE #9989] how to accelerate the inference speed of the model #32305

New Issue
Closed
opened 2026-04-22 13:26:10 -05:00 by GiteaMirror · 2 comments
GiteaMirror commented 2026-04-22 13:26:10 -05:00
Owner
Copy Link

Originally created by @Tu1231 on GitHub (Mar 26, 2025).
Original GitHub issue: https://github.com/ollama/ollama/issues/9989

What is the issue?

When I entered my ollama/ollama container terminal and ran deepseek-r1:32b, its inference speed was slow, and executing ollama displayed ollama ps
NAME ID SIZE PROCESSOR UNTIL
deepseek-r1:32b 63a233c0c27b 126 GB 100% GPU 2 minutes from now
bge-m3:latest 790764642607 1.7 GB 100% GPU 12 seconds from now，
My total video memory is 160GB, and in actual testing, it is about 4-6 words per second, which doesn't feel smooth. The background Nvidia SMI shows that each GPU only has a 20% usage rate.

Relevant log output

load_tensors: offloaded 25/25 layers to GPU
load_tensors:        CUDA0 model buffer size =   577.22 MiB
load_tensors:   CPU_Mapped model buffer size =   520.30 MiB
llama_init_from_model: n_seq_max     = 1
llama_init_from_model: n_ctx         = 4096
llama_init_from_model: n_ctx_per_seq = 4096
llama_init_from_model: n_batch       = 512
llama_init_from_model: n_ubatch      = 512
llama_init_from_model: flash_attn    = 0
llama_init_from_model: freq_base     = 10000.0
llama_init_from_model: freq_scale    = 1
llama_init_from_model: n_ctx_per_seq (4096) < n_ctx_train (8192) -- the full capacity of the model will not be utilized
llama_kv_cache_init: kv_size = 4096, offload = 1, type_k = 'f16', type_v = 'f16', n_layer = 24, can_shift = 1
llama_kv_cache_init:      CUDA0 KV buffer size =   384.00 MiB
llama_init_from_model: KV self size  =  384.00 MiB, K (f16):  192.00 MiB, V (f16):  192.00 MiB
llama_init_from_model:  CUDA_Host  output buffer size =     0.00 MiB
llama_init_from_model:      CUDA0 compute buffer size =    25.01 MiB
llama_init_from_model:  CUDA_Host compute buffer size =     5.01 MiB
llama_init_from_model: graph nodes  = 849
llama_init_from_model: graph splits = 4 (with bs=512), 2 (with bs=1)
time=2025-03-26T03:04:30.907Z level=INFO source=server.go:596 msg="llama runner started in 1.51 seconds"
llama_model_loader: loaded meta data with 33 key-value pairs and 389 tensors from /root/.ollama/models/blobs/sha256-daec91ffb5dd0c27411bd71f29932917c49cf529a641d0168496c3a501e3062c (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.size_label str              = 567M
llama_model_loader: - kv   3:                            general.license str              = mit
llama_model_loader: - kv   4:                               general.tags arr[str,4]       = ["sentence-transformers", "feature-ex...
llama_model_loader: - kv   5:                           bert.block_count u32              = 24
llama_model_loader: - kv   6:                        bert.context_length u32              = 8192
llama_model_loader: - kv   7:                      bert.embedding_length u32              = 1024
llama_model_loader: - kv   8:                   bert.feed_forward_length u32              = 4096
llama_model_loader: - kv   9:                  bert.attention.head_count u32              = 16
llama_model_loader: - kv  10:          bert.attention.layer_norm_epsilon f32              = 0.000010
llama_model_loader: - kv  11:                          general.file_type u32              = 1
llama_model_loader: - kv  12:                      bert.attention.causal bool             = false
llama_model_loader: - kv  13:                          bert.pooling_type u32              = 2
llama_model_loader: - kv  14:                       tokenizer.ggml.model str              = t5
llama_model_loader: - kv  15:                         tokenizer.ggml.pre str              = default
llama_model_loader: - kv  16:                      tokenizer.ggml.tokens arr[str,250002]  = ["<s>", "<pad>", "</s>", "<unk>", ","...
llama_model_loader: - kv  17:                      tokenizer.ggml.scores arr[f32,250002]  = [0.000000, 0.000000, 0.000000, 0.0000...
llama_model_loader: - kv  18:                  tokenizer.ggml.token_type arr[i32,250002]  = [3, 3, 3, 2, 1, 1, 1, 1, 1, 1, 1, 1, ...
llama_model_loader: - kv  19:            tokenizer.ggml.add_space_prefix bool             = true
llama_model_loader: - kv  20:            tokenizer.ggml.token_type_count u32              = 1
llama_model_loader: - kv  21:    tokenizer.ggml.remove_extra_whitespaces bool             = true
llama_model_loader: - kv  22:        tokenizer.ggml.precompiled_charsmap arr[u8,237539]   = [0, 180, 2, 0, 0, 132, 0, 0, 0, 0, 0,...
llama_model_loader: - kv  23:                tokenizer.ggml.bos_token_id u32              = 0
llama_model_loader: - kv  24:                tokenizer.ggml.eos_token_id u32              = 2
llama_model_loader: - kv  25:            tokenizer.ggml.unknown_token_id u32              = 3
llama_model_loader: - kv  26:          tokenizer.ggml.seperator_token_id u32              = 2
llama_model_loader: - kv  27:            tokenizer.ggml.padding_token_id u32              = 1
llama_model_loader: - kv  28:                tokenizer.ggml.cls_token_id u32              = 0
llama_model_loader: - kv  29:               tokenizer.ggml.mask_token_id u32              = 250001
llama_model_loader: - kv  30:               tokenizer.ggml.add_bos_token bool             = true
llama_model_loader: - kv  31:               tokenizer.ggml.add_eos_token bool             = true
llama_model_loader: - kv  32:               general.quantization_version u32              = 2
llama_model_loader: - type  f32:  244 tensors
llama_model_loader: - type  f16:  145 tensors
print_info: file format = GGUF V3 (latest)
print_info: file type   = F16
print_info: file size   = 1.07 GiB (16.25 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       = 1
print_info: model type       = ?B
print_info: model params     = 566.70 M
print_info: general.name     = n/a
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
llama_model_load: vocab only - skipping tensors

OS

Docker

GPU

Nvidia

CPU

Intel

Ollama version

0.5.13

Originally created by @Tu1231 on GitHub (Mar 26, 2025). Original GitHub issue: https://github.com/ollama/ollama/issues/9989 ### What is the issue? When I entered my ollama/ollama container terminal and ran deepseek-r1:32b, its inference speed was slow, and executing ollama displayed ollama ps NAME ID SIZE PROCESSOR UNTIL deepseek-r1:32b 63a233c0c27b 126 GB 100% GPU 2 minutes from now bge-m3:latest 790764642607 1.7 GB 100% GPU 12 seconds from now， My total video memory is 160GB, and in actual testing, it is about 4-6 words per second, which doesn't feel smooth. The background Nvidia SMI shows that each GPU only has a 20% usage rate. ### Relevant log output ```shell load_tensors: offloaded 25/25 layers to GPU load_tensors: CUDA0 model buffer size = 577.22 MiB load_tensors: CPU_Mapped model buffer size = 520.30 MiB llama_init_from_model: n_seq_max = 1 llama_init_from_model: n_ctx = 4096 llama_init_from_model: n_ctx_per_seq = 4096 llama_init_from_model: n_batch = 512 llama_init_from_model: n_ubatch = 512 llama_init_from_model: flash_attn = 0 llama_init_from_model: freq_base = 10000.0 llama_init_from_model: freq_scale = 1 llama_init_from_model: n_ctx_per_seq (4096) < n_ctx_train (8192) -- the full capacity of the model will not be utilized llama_kv_cache_init: kv_size = 4096, offload = 1, type_k = 'f16', type_v = 'f16', n_layer = 24, can_shift = 1 llama_kv_cache_init: CUDA0 KV buffer size = 384.00 MiB llama_init_from_model: KV self size = 384.00 MiB, K (f16): 192.00 MiB, V (f16): 192.00 MiB llama_init_from_model: CUDA_Host output buffer size = 0.00 MiB llama_init_from_model: CUDA0 compute buffer size = 25.01 MiB llama_init_from_model: CUDA_Host compute buffer size = 5.01 MiB llama_init_from_model: graph nodes = 849 llama_init_from_model: graph splits = 4 (with bs=512), 2 (with bs=1) time=2025-03-26T03:04:30.907Z level=INFO source=server.go:596 msg="llama runner started in 1.51 seconds" llama_model_loader: loaded meta data with 33 key-value pairs and 389 tensors from /root/.ollama/models/blobs/sha256-daec91ffb5dd0c27411bd71f29932917c49cf529a641d0168496c3a501e3062c (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.size_label str = 567M llama_model_loader: - kv 3: general.license str = mit llama_model_loader: - kv 4: general.tags arr[str,4] = ["sentence-transformers", "feature-ex... llama_model_loader: - kv 5: bert.block_count u32 = 24 llama_model_loader: - kv 6: bert.context_length u32 = 8192 llama_model_loader: - kv 7: bert.embedding_length u32 = 1024 llama_model_loader: - kv 8: bert.feed_forward_length u32 = 4096 llama_model_loader: - kv 9: bert.attention.head_count u32 = 16 llama_model_loader: - kv 10: bert.attention.layer_norm_epsilon f32 = 0.000010 llama_model_loader: - kv 11: general.file_type u32 = 1 llama_model_loader: - kv 12: bert.attention.causal bool = false llama_model_loader: - kv 13: bert.pooling_type u32 = 2 llama_model_loader: - kv 14: tokenizer.ggml.model str = t5 llama_model_loader: - kv 15: tokenizer.ggml.pre str = default llama_model_loader: - kv 16: tokenizer.ggml.tokens arr[str,250002] = ["<s>", "<pad>", "</s>", "<unk>", ","... llama_model_loader: - kv 17: tokenizer.ggml.scores arr[f32,250002] = [0.000000, 0.000000, 0.000000, 0.0000... llama_model_loader: - kv 18: tokenizer.ggml.token_type arr[i32,250002] = [3, 3, 3, 2, 1, 1, 1, 1, 1, 1, 1, 1, ... llama_model_loader: - kv 19: tokenizer.ggml.add_space_prefix bool = true llama_model_loader: - kv 20: tokenizer.ggml.token_type_count u32 = 1 llama_model_loader: - kv 21: tokenizer.ggml.remove_extra_whitespaces bool = true llama_model_loader: - kv 22: tokenizer.ggml.precompiled_charsmap arr[u8,237539] = [0, 180, 2, 0, 0, 132, 0, 0, 0, 0, 0,... llama_model_loader: - kv 23: tokenizer.ggml.bos_token_id u32 = 0 llama_model_loader: - kv 24: tokenizer.ggml.eos_token_id u32 = 2 llama_model_loader: - kv 25: tokenizer.ggml.unknown_token_id u32 = 3 llama_model_loader: - kv 26: tokenizer.ggml.seperator_token_id u32 = 2 llama_model_loader: - kv 27: tokenizer.ggml.padding_token_id u32 = 1 llama_model_loader: - kv 28: tokenizer.ggml.cls_token_id u32 = 0 llama_model_loader: - kv 29: tokenizer.ggml.mask_token_id u32 = 250001 llama_model_loader: - kv 30: tokenizer.ggml.add_bos_token bool = true llama_model_loader: - kv 31: tokenizer.ggml.add_eos_token bool = true llama_model_loader: - kv 32: general.quantization_version u32 = 2 llama_model_loader: - type f32: 244 tensors llama_model_loader: - type f16: 145 tensors print_info: file format = GGUF V3 (latest) print_info: file type = F16 print_info: file size = 1.07 GiB (16.25 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 = 1 print_info: model type = ?B print_info: model params = 566.70 M print_info: general.name = n/a 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 llama_model_load: vocab only - skipping tensors ``` ### OS Docker ### GPU Nvidia ### CPU Intel ### Ollama version 0.5.13
GiteaMirror added the question label 2026-04-22 13:26:10 -05:00
GiteaMirror commented 2026-04-22 13:26:13 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Mar 26, 2025):

https://github.com/ollama/ollama/issues/7648#issuecomment-2473561990

<!-- gh-comment-id:2753661814 --> @rick-github commented on GitHub (Mar 26, 2025): https://github.com/ollama/ollama/issues/7648#issuecomment-2473561990
GiteaMirror commented 2026-04-22 13:26:13 -05:00
Author
Owner
Copy Link

@Tu1231 commented on GitHub (Mar 27, 2025):

thanks

<!-- gh-comment-id:2757189222 --> @Tu1231 commented on GitHub (Mar 27, 2025): thanks
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
amd
api
app
bug
build
cli
cloud
compatibility
context-length
create
docker
documentation
embeddings
feature request
feedback wanted
good first issue
gpt-oss
gpu
harmony
help wanted
image
install
intel
js
launch
linux
macos
memory
mlx
model
needs more info
networking
nvidia
ollama.com
performance
pull-request
Mirrored from GitHub Pull Request
 python
question
registry
rendering
thinking
tools
top
vulkan
windows
wsl
No Label question
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
GiteaMirror ninjasurge
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/ollama#32305
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?