Name and Version
version: 8373 (46dba9f)
built with GNU 11.4.0 for Linux x86_64
Operating systems
Linux
GGML backends
Vulkan
Hardware
Ryzen 9 7900 + rtx 4080 super + rx 7900 xtx
Models
unsloth/Qwen3.5-27B-GGUF:UD-Q4_K_XL
Problem description & steps to reproduce
When running
llama-server -hf unsloth/Qwen3.5-27B-GGUF:UD-Q4_K_XL \
-ngl 999 \
--temp 0.6 \
--top-p 0.95 \
--top-k 20 \
--min-p 0.00 \
-dev VULKAN0,VULKAN1 \
-c 16000
(both rtx 4080 super and rx 7900xtx)
when I tell the model "HI!"
the model starts outputing:
'*%%14"234+&*#1"$%1-+!4!2'3&,-1,32&0)*!&0+&2&.$',$3#,)'1),#'+'((44$+-21010)2!!$"4."%)2%%#3+&$$'2*31!...-&-$(34,-2.3".+(%2'#''((!&.")%...
Thousands of tokens like the above, and doesn't stop
However, when running
llama-server -hf unsloth/Qwen3.5-27B-GGUF:UD-Q4_K_XL \
-ngl 999 \
--temp 0.6 \
--top-p 0.95 \
--top-k 20 \
--min-p 0.00 \
-dev VULKAN1 \
-c 16000
(meaning only on the rx 7900xtx)
the model is completly functional.
when I tell it "HI!"
it outputs something like
<<<reasoning_content_start>>>Thinking Process:
1. **Analyze the Request:**
* Input: "HI!"
* Intent: Greeting, initiating conversation.
* Tone: Friendly, casual, enthusiastic (indicated by the exclamation mark).
* Expected Output: A friendly greeting in return, offering assistance.
2. **Determine the appropriate response:**
* Acknowledge the greeting.
* Match the tone (friendly, enthusiastic).
* Offer help or ask how the user is doing.
* Keep it concise but warm.
3. **Drafting options:**
* Option 1: Hello! How can I help you today?
* Option 2: Hi there! What's up?
* Option 3: HI! 👋 How's your day going? Anything I can help you with?
4. **Selecting the best option:** Option 3 is friendly, uses an emoji to match the energy, and opens the floor for assistance.
5. **Final Polish:** "HI! 👋 How's your day going? Anything I can help you with?"
6. **Safety Check:** No sensitive or harmful content. Just a greeting.
7. **Final Output Generation.** (Matches the selected draft).ot
<<<reasoning_content_end>>>HI! 👋 How's your day going? Anything I can help you with?
Is further information/logs needed?
First Bad Commit
No response
Relevant log output
log of loading the model on the 2 gpus:
llama-server -hf unsloth/Qwen3.5-27B-GGUF:UD-Q4_K_XL -ngl 999 --temp 0.6 --top-p 0.95 --top-k 20 --min-p 0.00 -dev VULKAN0,VULKAN1 -c 16000
load_backend: loaded RPC backend from /home/user/Downloads/ollama-tests/llama-cpp/bin/libggml-rpc.so
ggml_vulkan: Found 2 Vulkan devices:
ggml_vulkan: 0 = NVIDIA GeForce RTX 4080 SUPER (NVIDIA) | uma: 0 | fp16: 1 | bf16: 1 | warp size: 32 | shared memory: 49152 | int dot: 1 | matrix cores: NV_coopmat2
ggml_vulkan: 1 = AMD Radeon RX 7900 XTX (RADV NAVI31) (radv) | uma: 0 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat
load_backend: loaded Vulkan backend from /home/user/Downloads/ollama-tests/llama-cpp/bin/libggml-vulkan.so
load_backend: loaded CPU backend from /home/user/Downloads/ollama-tests/llama-cpp/bin/libggml-cpu-zen4.so
common_download_file_single_online: no previous model file found /home/user/.cache/llama.cpp/unsloth_Qwen3.5-27B-GGUF_preset.ini
common_download_file_single_online: HEAD failed, status: 404
no remote preset found, skipping
common_download_file_single_online: using cached file (same etag): /home/user/.cache/llama.cpp/unsloth_Qwen3.5-27B-GGUF_Qwen3.5-27B-UD-Q4_K_XL.gguf
common_download_file_single_online: using cached file (same etag): /home/user/.cache/llama.cpp/unsloth_Qwen3.5-27B-GGUF_mmproj-BF16.gguf
main: n_parallel is set to auto, using n_parallel = 4 and kv_unified = true
build: 8373 (46dba9fce) with GNU 11.4.0 for Linux x86_64
system info: n_threads = 12, n_threads_batch = 12, total_threads = 24
system_info: n_threads = 12 (n_threads_batch = 12) / 24 | CPU : SSE3 = 1 | SSSE3 = 1 | AVX = 1 | AVX2 = 1 | F16C = 1 | FMA = 1 | BMI2 = 1 | AVX512 = 1 | AVX512_VBMI = 1 | AVX512_VNNI = 1 | AVX512_BF16 = 1 | LLAMAFILE = 1 | OPENMP = 1 | REPACK = 1 |
Running without SSL
init: using 23 threads for HTTP server
start: binding port with default address family
main: loading model
srv load_model: loading model '/home/user/.cache/llama.cpp/unsloth_Qwen3.5-27B-GGUF_Qwen3.5-27B-UD-Q4_K_XL.gguf'
common_init_result: fitting params to device memory, for bugs during this step try to reproduce them with -fit off, or provide --verbose logs if the bug only occurs with -fit on
llama_params_fit_impl: projected memory use with initial parameters [MiB]:
llama_params_fit_impl: - Vulkan0 (NVIDIA GeForce RTX 4080 SUPER) : 16376 total, 6805 used, 8128 free vs. target of 1024
llama_params_fit_impl: - Vulkan1 (AMD Radeon RX 7900 XTX (RADV NAVI31)): 24560 total, 11870 used, 12300 free vs. target of 1024
llama_params_fit_impl: projected to use 18676 MiB of device memory vs. 39105 MiB of free device memory
llama_params_fit_impl: targets for free memory can be met on all devices, no changes needed
llama_params_fit: successfully fit params to free device memory
llama_params_fit: fitting params to free memory took 0.36 seconds
llama_model_load_from_file_impl: using device Vulkan0 (NVIDIA GeForce RTX 4080 SUPER) (0000:01:00.0) - 15179 MiB free
llama_model_load_from_file_impl: using device Vulkan1 (AMD Radeon RX 7900 XTX (RADV NAVI31)) (0000:0e:00.0) - 24532 MiB free
llama_model_loader: loaded meta data with 49 key-value pairs and 851 tensors from /home/user/.cache/llama.cpp/unsloth_Qwen3.5-27B-GGUF_Qwen3.5-27B-UD-Q4_K_XL.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 = qwen35
llama_model_loader: - kv 1: general.type str = model
llama_model_loader: - kv 2: general.sampling.top_k i32 = 20
llama_model_loader: - kv 3: general.sampling.top_p f32 = 0.950000
llama_model_loader: - kv 4: general.sampling.temp f32 = 0.600000
llama_model_loader: - kv 5: general.name str = Qwen3.5-27B
llama_model_loader: - kv 6: general.basename str = Qwen3.5-27B
llama_model_loader: - kv 7: general.quantized_by str = Unsloth
llama_model_loader: - kv 8: general.size_label str = 27B
llama_model_loader: - kv 9: general.license str = apache-2.0
llama_model_loader: - kv 10: general.license.link str = https://huggingface.co/Qwen/Qwen3.5-2...
llama_model_loader: - kv 11: general.repo_url str = https://huggingface.co/unsloth
llama_model_loader: - kv 12: general.base_model.count u32 = 1
llama_model_loader: - kv 13: general.base_model.0.name str = Qwen3.5 27B
llama_model_loader: - kv 14: general.base_model.0.organization str = Qwen
llama_model_loader: - kv 15: general.base_model.0.repo_url str = https://huggingface.co/Qwen/Qwen3.5-27B
llama_model_loader: - kv 16: general.tags arr[str,3] = ["qwen3_5_moe", "unsloth", "image-tex...
llama_model_loader: - kv 17: qwen35.block_count u32 = 64
llama_model_loader: - kv 18: qwen35.context_length u32 = 262144
llama_model_loader: - kv 19: qwen35.embedding_length u32 = 5120
llama_model_loader: - kv 20: qwen35.feed_forward_length u32 = 17408
llama_model_loader: - kv 21: qwen35.attention.head_count u32 = 24
llama_model_loader: - kv 22: qwen35.attention.head_count_kv u32 = 4
llama_model_loader: - kv 23: qwen35.rope.dimension_sections arr[i32,4] = [11, 11, 10, 0]
llama_model_loader: - kv 24: qwen35.rope.freq_base f32 = 10000000.000000
llama_model_loader: - kv 25: qwen35.attention.layer_norm_rms_epsilon f32 = 0.000001
llama_model_loader: - kv 26: qwen35.attention.key_length u32 = 256
llama_model_loader: - kv 27: qwen35.attention.value_length u32 = 256
llama_model_loader: - kv 28: qwen35.ssm.conv_kernel u32 = 4
llama_model_loader: - kv 29: qwen35.ssm.state_size u32 = 128
llama_model_loader: - kv 30: qwen35.ssm.group_count u32 = 16
llama_model_loader: - kv 31: qwen35.ssm.time_step_rank u32 = 48
llama_model_loader: - kv 32: qwen35.ssm.inner_size u32 = 6144
llama_model_loader: - kv 33: qwen35.full_attention_interval u32 = 4
llama_model_loader: - kv 34: qwen35.rope.dimension_count u32 = 64
llama_model_loader: - kv 35: tokenizer.ggml.model str = gpt2
llama_model_loader: - kv 36: tokenizer.ggml.pre str = qwen35
llama_model_loader: - kv 37: tokenizer.ggml.tokens arr[str,248320] = ["!", "\"", "#", "$", "%", "&", "'", ...
llama_model_loader: - kv 38: tokenizer.ggml.token_type arr[i32,248320] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ...
llama_model_loader: - kv 39: tokenizer.ggml.merges arr[str,247587] = ["Ġ Ġ", "ĠĠ ĠĠ", "i n", "Ġ t",...
llama_model_loader: - kv 40: tokenizer.ggml.eos_token_id u32 = 248046
llama_model_loader: - kv 41: tokenizer.ggml.padding_token_id u32 = 248055
llama_model_loader: - kv 42: tokenizer.chat_template str = {%- set image_count = namespace(value...
llama_model_loader: - kv 43: general.quantization_version u32 = 2
llama_model_loader: - kv 44: general.file_type u32 = 15
llama_model_loader: - kv 45: quantize.imatrix.file str = Qwen3.5-27B-GGUF/imatrix_unsloth.gguf
llama_model_loader: - kv 46: quantize.imatrix.dataset str = unsloth_calibration_Qwen3.5-27B.txt
llama_model_loader: - kv 47: quantize.imatrix.entries_count u32 = 496
llama_model_loader: - kv 48: quantize.imatrix.chunks_count u32 = 80
llama_model_loader: - type f32: 353 tensors
llama_model_loader: - type f16: 96 tensors
llama_model_loader: - type q8_0: 48 tensors
llama_model_loader: - type q4_K: 173 tensors
llama_model_loader: - type q5_K: 122 tensors
llama_model_loader: - type q6_K: 43 tensors
llama_model_loader: - type iq4_xs: 16 tensors
print_info: file format = GGUF V3 (latest)
print_info: file type = Q4_K - Medium
print_info: file size = 16.40 GiB (5.24 BPW)
load: 0 unused tokens
load: printing all EOG tokens:
load: - 248044 ('<|endoftext|>')
load: - 248046 ('<|im_end|>')
load: - 248063 ('<|fim_pad|>')
load: - 248064 ('<|repo_name|>')
load: - 248065 ('<|file_sep|>')
load: special tokens cache size = 33
load: token to piece cache size = 1.7581 MB
print_info: arch = qwen35
print_info: vocab_only = 0
print_info: no_alloc = 0
print_info: n_ctx_train = 262144
print_info: n_embd = 5120
print_info: n_embd_inp = 5120
print_info: n_layer = 64
print_info: n_head = 24
print_info: n_head_kv = 4
print_info: n_rot = 64
print_info: n_swa = 0
print_info: is_swa_any = 0
print_info: n_embd_head_k = 256
print_info: n_embd_head_v = 256
print_info: n_gqa = 6
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-06
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 = 17408
print_info: n_expert = 0
print_info: n_expert_used = 0
print_info: n_expert_groups = 0
print_info: n_group_used = 0
print_info: causal attn = 1
print_info: pooling type = 0
print_info: rope type = 40
print_info: rope scaling = linear
print_info: freq_base_train = 10000000.0
print_info: freq_scale_train = 1
print_info: n_ctx_orig_yarn = 262144
print_info: rope_yarn_log_mul = 0.0000
print_info: rope_finetuned = unknown
print_info: mrope sections = [11, 11, 10, 0]
print_info: ssm_d_conv = 4
print_info: ssm_d_inner = 6144
print_info: ssm_d_state = 128
print_info: ssm_dt_rank = 48
print_info: ssm_n_group = 16
print_info: ssm_dt_b_c_rms = 0
print_info: model type = 27B
print_info: model params = 26.90 B
print_info: general.name = Qwen3.5-27B
print_info: vocab type = BPE
print_info: n_vocab = 248320
print_info: n_merges = 247587
print_info: BOS token = 11 ','
print_info: EOS token = 248046 '<|im_end|>'
print_info: EOT token = 248046 '<|im_end|>'
print_info: PAD token = 248055 '<|vision_pad|>'
print_info: LF token = 198 'Ċ'
print_info: FIM PRE token = 248060 '<|fim_prefix|>'
print_info: FIM SUF token = 248062 '<|fim_suffix|>'
print_info: FIM MID token = 248061 '<|fim_middle|>'
print_info: FIM PAD token = 248063 '<|fim_pad|>'
print_info: FIM REP token = 248064 '<|repo_name|>'
print_info: FIM SEP token = 248065 '<|file_sep|>'
print_info: EOG token = 248044 '<|endoftext|>'
print_info: EOG token = 248046 '<|im_end|>'
print_info: EOG token = 248063 '<|fim_pad|>'
print_info: EOG token = 248064 '<|repo_name|>'
print_info: EOG token = 248065 '<|file_sep|>'
print_info: max token length = 256
load_tensors: loading model tensors, this can take a while... (mmap = true, direct_io = false)
load_tensors: offloading output layer to GPU
load_tensors: offloading 63 repeating layers to GPU
load_tensors: offloaded 65/65 layers to GPU
load_tensors: CPU_Mapped model buffer size = 682.03 MiB
load_tensors: Vulkan0 model buffer size = 5871.12 MiB
load_tensors: Vulkan1 model buffer size = 10241.18 MiB
.............................................................................................
common_init_result: added <|endoftext|> logit bias = -inf
common_init_result: added <|im_end|> logit bias = -inf
common_init_result: added <|fim_pad|> logit bias = -inf
common_init_result: added <|repo_name|> logit bias = -inf
common_init_result: added <|file_sep|> logit bias = -inf
llama_context: constructing llama_context
llama_context: n_seq_max = 4
llama_context: n_ctx = 16128
llama_context: n_ctx_seq = 16128
llama_context: n_batch = 2048
llama_context: n_ubatch = 512
llama_context: causal_attn = 1
llama_context: flash_attn = auto
llama_context: kv_unified = true
llama_context: freq_base = 10000000.0
llama_context: freq_scale = 1
llama_context: n_ctx_seq (16128) < n_ctx_train (262144) -- the full capacity of the model will not be utilized
llama_context: Vulkan_Host output buffer size = 3.79 MiB
llama_kv_cache: Vulkan0 KV buffer size = 378.00 MiB
llama_kv_cache: Vulkan1 KV buffer size = 630.00 MiB
llama_kv_cache: size = 1008.00 MiB ( 16128 cells, 16 layers, 4/1 seqs), K (f16): 504.00 MiB, V (f16): 504.00 MiB
llama_memory_recurrent: Vulkan0 RS buffer size = 236.91 MiB
llama_memory_recurrent: Vulkan1 RS buffer size = 361.59 MiB
llama_memory_recurrent: size = 598.50 MiB ( 4 cells, 64 layers, 4 seqs), R (f32): 22.50 MiB, S (f32): 576.00 MiB
llama_context: pipeline parallelism enabled
llama_context: graph reuse is currently not compatible with pipeline parallelism - disabling
sched_reserve: reserving ...
sched_reserve: Flash Attention was auto, set to enabled
sched_reserve: resolving fused Gated Delta Net support:
sched_reserve: fused Gated Delta Net (autoregressive) enabled
sched_reserve: fused Gated Delta Net (chunked) enabled
sched_reserve: Vulkan0 compute buffer size = 319.81 MiB
sched_reserve: Vulkan1 compute buffer size = 598.07 MiB
sched_reserve: Vulkan_Host compute buffer size = 146.08 MiB
sched_reserve: graph nodes = 3657
sched_reserve: graph splits = 3
sched_reserve: reserve took 44.66 ms, sched copies = 4
common_init_from_params: warming up the model with an empty run - please wait ... (--no-warmup to disable)
clip_model_loader: model name: Qwen3.5-27B
clip_model_loader: description:
clip_model_loader: GGUF version: 3
clip_model_loader: alignment: 32
clip_model_loader: n_tensors: 334
clip_model_loader: n_kv: 28
clip_model_loader: has vision encoder
clip_ctx: CLIP using Vulkan0 backend
load_hparams: Qwen-VL models require at minimum 1024 image tokens to function correctly on grounding tasks
load_hparams: if you encounter problems with accuracy, try adding --image-min-tokens 1024
load_hparams: more info: https://github.com/ggml-org/llama.cpp/issues/16842
load_hparams: projector: qwen3vl_merger
load_hparams: n_embd: 1152
load_hparams: n_head: 16
load_hparams: n_ff: 4304
load_hparams: n_layer: 27
load_hparams: ffn_op: gelu
load_hparams: projection_dim: 5120
--- vision hparams ---
load_hparams: image_size: 768
load_hparams: patch_size: 16
load_hparams: has_llava_proj: 0
load_hparams: minicpmv_version: 0
load_hparams: n_merge: 2
load_hparams: n_wa_pattern: 0
load_hparams: image_min_pixels: 8192
load_hparams: image_max_pixels: 4194304
load_hparams: model size: 887.99 MiB
load_hparams: metadata size: 0.12 MiB
warmup: warmup with image size = 1472 x 1472
alloc_compute_meta: Vulkan0 compute buffer size = 248.10 MiB
alloc_compute_meta: CPU compute buffer size = 24.93 MiB
alloc_compute_meta: graph splits = 1, nodes = 823
warmup: flash attention is enabled
srv load_model: loaded multimodal model, '/home/user/.cache/llama.cpp/unsloth_Qwen3.5-27B-GGUF_mmproj-BF16.gguf'
srv load_model: initializing slots, n_slots = 4
common_speculative_is_compat: the target context does not support partial sequence removal
srv load_model: speculative decoding not supported by this context
slot load_model: id 0 | task -1 | new slot, n_ctx = 16128
slot load_model: id 1 | task -1 | new slot, n_ctx = 16128
slot load_model: id 2 | task -1 | new slot, n_ctx = 16128
slot load_model: id 3 | task -1 | new slot, n_ctx = 16128
srv load_model: prompt cache is enabled, size limit: 8192 MiB
srv load_model: use `--cache-ram 0` to disable the prompt cache
srv load_model: for more info see https://github.com/ggml-org/llama.cpp/pull/16391
init: chat template, 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
<think>
'
srv init: init: chat template, thinking = 1
main: model loaded
main: server is listening on http://127.0.0.1:8080
main: starting the main loop...
srv update_slots: all slots are idle
Name and Version
version: 8373 (46dba9f)
built with GNU 11.4.0 for Linux x86_64
Operating systems
Linux
GGML backends
Vulkan
Hardware
Ryzen 9 7900 + rtx 4080 super + rx 7900 xtx
Models
unsloth/Qwen3.5-27B-GGUF:UD-Q4_K_XL
Problem description & steps to reproduce
When running
(both rtx 4080 super and rx 7900xtx)
when I tell the model "HI!"
the model starts outputing:
'*%%14"234+&*#1"$%1-+!4!2'3&,-1,32&0)*!&0+&2&.$',$3#,)'1),#'+'((44$+-21010)2!!$"4."%)2%%#3+&$$'2*31!...-&-$(34,-2.3".+(%2'#''((!&.")%...Thousands of tokens like the above, and doesn't stop
However, when running
(meaning only on the rx 7900xtx)
the model is completly functional.
when I tell it "HI!"
it outputs something like
Is further information/logs needed?
First Bad Commit
No response
Relevant log output
log of loading the model on the 2 gpus: