github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-06 16:11:34 -05:00
Code Issues 14.5k Packages Projects Releases 120 Wiki Activity

[GH-ISSUE #11661] ollama run alibayram/hunyuan:0.5b error #7709

New Issue
Closed
opened 2026-04-12 19:48:52 -05:00 by GiteaMirror · 8 comments
GiteaMirror commented 2026-04-12 19:48:52 -05:00
Owner
Copy Link

Originally created by @workmengxue on GitHub (Aug 5, 2025).
Original GitHub issue: https://github.com/ollama/ollama/issues/11661

What is the issue?

C:\Users\mx>ollama run alibayram/hunyuan:0.5b
pulling manifest
pulling 65fe0e5b8f67: 100% ▕██████████████████████████████████████████████████████████▏ 354 MB
pulling eb9280c126b3: 100% ▕██████████████████████████████████████████████████████████▏ 407 B
pulling d63f3f323c98: 100% ▕██████████████████████████████████████████████████████████▏ 132 B
pulling 7caa25a98b52: 100% ▕██████████████████████████████████████████████████████████▏ 432 B
verifying sha256 digest
writing manifest
success
Error: 500 Internal Server Error: unable to load model: C:\Users\mx.ollama\models\blobs\sha256-65fe0e5b8f677752bf7176ea42c01b466d7ff654d8577d5f62d3028c75ba4add

Relevant log output

OS

No response

GPU

No response

CPU

No response

Ollama version

No response

Originally created by @workmengxue on GitHub (Aug 5, 2025). Original GitHub issue: https://github.com/ollama/ollama/issues/11661 ### What is the issue? C:\Users\mx>ollama run alibayram/hunyuan:0.5b pulling manifest pulling 65fe0e5b8f67: 100% ▕██████████████████████████████████████████████████████████▏ 354 MB pulling eb9280c126b3: 100% ▕██████████████████████████████████████████████████████████▏ 407 B pulling d63f3f323c98: 100% ▕██████████████████████████████████████████████████████████▏ 132 B pulling 7caa25a98b52: 100% ▕██████████████████████████████████████████████████████████▏ 432 B verifying sha256 digest writing manifest success Error: 500 Internal Server Error: unable to load model: C:\Users\mx\.ollama\models\blobs\sha256-65fe0e5b8f677752bf7176ea42c01b466d7ff654d8577d5f62d3028c75ba4add ### Relevant log output ```shell ``` ### OS _No response_ ### GPU _No response_ ### CPU _No response_ ### Ollama version _No response_
GiteaMirror added the bug label 2026-04-12 19:48:52 -05:00
GiteaMirror commented 2026-04-12 19:48:53 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Aug 5, 2025):

Server logs will help in debugging. Most likely problem is unsupported architecture.

https://github.com/ollama/ollama/issues/11361
https://github.com/ollama/ollama/issues/11239
https://github.com/ollama/ollama/issues/7503

<!-- gh-comment-id:3153782699 --> @rick-github commented on GitHub (Aug 5, 2025): [Server logs](https://github.com/ollama/ollama/blob/main/docs/troubleshooting.md#how-to-troubleshoot-issues) will help in debugging. Most likely problem is unsupported architecture. https://github.com/ollama/ollama/issues/11361 https://github.com/ollama/ollama/issues/11239 https://github.com/ollama/ollama/issues/7503
GiteaMirror commented 2026-04-12 19:48:53 -05:00
Author
Owner
Copy Link

@svengong commented on GitHub (Aug 7, 2025):

ime=2025-08-07T13:32:00.959+08:00 level=INFO source=sched.go:788 msg="new model will fit in available VRAM in single GPU, loading" model=/Users/sven/.ollama/models/blobs/sha256-e59a4dd7d94142c65e0aef5e20b0a44637467b472628dac00dfe9be23dc4d18d gpu=0 parallel=2 available=22906503168 required="18.5 GiB"
time=2025-08-07T13:32:00.960+08:00 level=INFO source=server.go:135 msg="system memory" total="32.0 GiB" free="9.1 GiB" free_swap="0 B"
time=2025-08-07T13:32:00.960+08:00 level=INFO source=server.go:175 msg=offload library=metal layers.requested=-1 layers.model=33 layers.offload=33 layers.split="" memory.available="[21.3 GiB]" memory.gpu_overhead="0 B" memory.required.full="18.5 GiB" memory.required.partial="18.5 GiB" memory.required.kv="8.0 GiB" memory.required.allocations="[18.5 GiB]" memory.weights.total="4.3 GiB" memory.weights.repeating="3.9 GiB" memory.weights.nonrepeating="410.7 MiB" memory.graph.full="5.3 GiB" memory.graph.partial="5.3 GiB"
llama_model_load_from_file_impl: using device Metal (Apple M1 Pro) - 21845 MiB free
llama_model_loader: loaded meta data with 39 key-value pairs and 354 tensors from /Users/sven/.ollama/models/blobs/sha256-e59a4dd7d94142c65e0aef5e20b0a44637467b472628dac00dfe9be23dc4d18d (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 = hunyuan-dense
llama_model_loader: - kv 1: general.type str = model
llama_model_loader: - kv 2: general.name str = Hunyuan 7B Instruct
llama_model_loader: - kv 3: general.finetune str = Instruct
llama_model_loader: - kv 4: general.basename str = Hunyuan
llama_model_loader: - kv 5: general.size_label str = 7B
llama_model_loader: - kv 6: general.base_model.count u32 = 1
llama_model_loader: - kv 7: general.base_model.0.name str = Hunyuan 7B Pretrain
llama_model_loader: - kv 8: general.base_model.0.organization str = Tencent
llama_model_loader: - kv 9: general.base_model.0.repo_url str = https://huggingface.co/tencent/Hunyua...
llama_model_loader: - kv 10: hunyuan-dense.block_count u32 = 32
llama_model_loader: - kv 11: hunyuan-dense.context_length u32 = 262144
llama_model_loader: - kv 12: hunyuan-dense.embedding_length u32 = 4096
llama_model_loader: - kv 13: hunyuan-dense.feed_forward_length u32 = 14336
llama_model_loader: - kv 14: hunyuan-dense.attention.head_count u32 = 32
llama_model_loader: - kv 15: hunyuan-dense.attention.head_count_kv u32 = 8
llama_model_loader: - kv 16: hunyuan-dense.rope.freq_base f32 = 11158840.000000
llama_model_loader: - kv 17: hunyuan-dense.attention.layer_norm_rms_epsilon f32 = 0.000010
llama_model_loader: - kv 18: hunyuan-dense.attention.key_length u32 = 128
llama_model_loader: - kv 19: hunyuan-dense.attention.value_length u32 = 128
llama_model_loader: - kv 20: hunyuan-dense.rope.scaling.type str = none
llama_model_loader: - kv 21: hunyuan-dense.rope.scaling.factor f32 = 1.000000
llama_model_loader: - kv 22: hunyuan-dense.rope.scaling.original_context_length u32 = 262144
llama_model_loader: - kv 23: tokenizer.ggml.model str = gpt2
llama_model_loader: - kv 24: tokenizer.ggml.pre str = hunyuan
llama_model_loader: - kv 25: tokenizer.ggml.tokens arr[str,128167] = ["!", """, "#", "$", "%", "&", "'", ...
llama_model_loader: - kv 26: tokenizer.ggml.token_type arr[i32,128167] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ...
llama_model_loader: - kv 27: tokenizer.ggml.merges arr[str,127698] = ["Ġ Ġ", "ĠĠ ĠĠ", "i n", "Ġ t",...
llama_model_loader: - kv 28: tokenizer.ggml.bos_token_id u32 = 127958
llama_model_loader: - kv 29: tokenizer.ggml.eos_token_id u32 = 127960
llama_model_loader: - kv 30: tokenizer.ggml.seperator_token_id u32 = 127962
llama_model_loader: - kv 31: tokenizer.ggml.padding_token_id u32 = 127961
llama_model_loader: - kv 32: tokenizer.chat_template str = {%- if not add_generation_prompt is d...
llama_model_loader: - kv 33: general.quantization_version u32 = 2
llama_model_loader: - kv 34: general.file_type u32 = 15
llama_model_loader: - kv 35: quantize.imatrix.file str = Hunyuan-7B-Instruct/Hunyuan-7B-Instru...
llama_model_loader: - kv 36: quantize.imatrix.dataset str = calibration_datav3.txt
llama_model_loader: - kv 37: quantize.imatrix.entries_count u32 = 224
llama_model_loader: - kv 38: quantize.imatrix.chunks_count u32 = 130
llama_model_loader: - type f32: 129 tensors
llama_model_loader: - type q4_K: 192 tensors
llama_model_loader: - type q6_K: 33 tensors
print_info: file format = GGUF V3 (latest)
print_info: file type = Q4_K - Medium
print_info: file size = 4.30 GiB (4.92 BPW)
llama_model_load: error loading model: error loading model architecture: unknown model architecture: 'hunyuan-dense'
llama_model_load_from_file_impl: failed to load model
time=2025-08-07T13:32:01.020+08:00 level=INFO source=sched.go:455 msg="NewLlamaServer failed" model=/Users/sven/.ollama/models/blobs/sha256-e59a4dd7d94142c65e0aef5e20b0a44637467b472628dac00dfe9be23dc4d18d error="unable to load model: /Users/sven/.ollama/models/blobs/sha256-e59a4dd7d94142c65e0aef5e20b0a44637467b472628dac00dfe9be23dc4d18d"

<!-- gh-comment-id:3162561395 --> @svengong commented on GitHub (Aug 7, 2025): ime=2025-08-07T13:32:00.959+08:00 level=INFO source=sched.go:788 msg="new model will fit in available VRAM in single GPU, loading" model=/Users/sven/.ollama/models/blobs/sha256-e59a4dd7d94142c65e0aef5e20b0a44637467b472628dac00dfe9be23dc4d18d gpu=0 parallel=2 available=22906503168 required="18.5 GiB" time=2025-08-07T13:32:00.960+08:00 level=INFO source=server.go:135 msg="system memory" total="32.0 GiB" free="9.1 GiB" free_swap="0 B" time=2025-08-07T13:32:00.960+08:00 level=INFO source=server.go:175 msg=offload library=metal layers.requested=-1 layers.model=33 layers.offload=33 layers.split="" memory.available="[21.3 GiB]" memory.gpu_overhead="0 B" memory.required.full="18.5 GiB" memory.required.partial="18.5 GiB" memory.required.kv="8.0 GiB" memory.required.allocations="[18.5 GiB]" memory.weights.total="4.3 GiB" memory.weights.repeating="3.9 GiB" memory.weights.nonrepeating="410.7 MiB" memory.graph.full="5.3 GiB" memory.graph.partial="5.3 GiB" llama_model_load_from_file_impl: using device Metal (Apple M1 Pro) - 21845 MiB free llama_model_loader: loaded meta data with 39 key-value pairs and 354 tensors from /Users/sven/.ollama/models/blobs/sha256-e59a4dd7d94142c65e0aef5e20b0a44637467b472628dac00dfe9be23dc4d18d (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 = hunyuan-dense llama_model_loader: - kv 1: general.type str = model llama_model_loader: - kv 2: general.name str = Hunyuan 7B Instruct llama_model_loader: - kv 3: general.finetune str = Instruct llama_model_loader: - kv 4: general.basename str = Hunyuan llama_model_loader: - kv 5: general.size_label str = 7B llama_model_loader: - kv 6: general.base_model.count u32 = 1 llama_model_loader: - kv 7: general.base_model.0.name str = Hunyuan 7B Pretrain llama_model_loader: - kv 8: general.base_model.0.organization str = Tencent llama_model_loader: - kv 9: general.base_model.0.repo_url str = https://huggingface.co/tencent/Hunyua... llama_model_loader: - kv 10: hunyuan-dense.block_count u32 = 32 llama_model_loader: - kv 11: hunyuan-dense.context_length u32 = 262144 llama_model_loader: - kv 12: hunyuan-dense.embedding_length u32 = 4096 llama_model_loader: - kv 13: hunyuan-dense.feed_forward_length u32 = 14336 llama_model_loader: - kv 14: hunyuan-dense.attention.head_count u32 = 32 llama_model_loader: - kv 15: hunyuan-dense.attention.head_count_kv u32 = 8 llama_model_loader: - kv 16: hunyuan-dense.rope.freq_base f32 = 11158840.000000 llama_model_loader: - kv 17: hunyuan-dense.attention.layer_norm_rms_epsilon f32 = 0.000010 llama_model_loader: - kv 18: hunyuan-dense.attention.key_length u32 = 128 llama_model_loader: - kv 19: hunyuan-dense.attention.value_length u32 = 128 llama_model_loader: - kv 20: hunyuan-dense.rope.scaling.type str = none llama_model_loader: - kv 21: hunyuan-dense.rope.scaling.factor f32 = 1.000000 llama_model_loader: - kv 22: hunyuan-dense.rope.scaling.original_context_length u32 = 262144 llama_model_loader: - kv 23: tokenizer.ggml.model str = gpt2 llama_model_loader: - kv 24: tokenizer.ggml.pre str = hunyuan llama_model_loader: - kv 25: tokenizer.ggml.tokens arr[str,128167] = ["!", "\"", "#", "$", "%", "&", "'", ... llama_model_loader: - kv 26: tokenizer.ggml.token_type arr[i32,128167] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ... llama_model_loader: - kv 27: tokenizer.ggml.merges arr[str,127698] = ["Ġ Ġ", "ĠĠ ĠĠ", "i n", "Ġ t",... llama_model_loader: - kv 28: tokenizer.ggml.bos_token_id u32 = 127958 llama_model_loader: - kv 29: tokenizer.ggml.eos_token_id u32 = 127960 llama_model_loader: - kv 30: tokenizer.ggml.seperator_token_id u32 = 127962 llama_model_loader: - kv 31: tokenizer.ggml.padding_token_id u32 = 127961 llama_model_loader: - kv 32: tokenizer.chat_template str = {%- if not add_generation_prompt is d... llama_model_loader: - kv 33: general.quantization_version u32 = 2 llama_model_loader: - kv 34: general.file_type u32 = 15 llama_model_loader: - kv 35: quantize.imatrix.file str = Hunyuan-7B-Instruct/Hunyuan-7B-Instru... llama_model_loader: - kv 36: quantize.imatrix.dataset str = calibration_datav3.txt llama_model_loader: - kv 37: quantize.imatrix.entries_count u32 = 224 llama_model_loader: - kv 38: quantize.imatrix.chunks_count u32 = 130 llama_model_loader: - type f32: 129 tensors llama_model_loader: - type q4_K: 192 tensors llama_model_loader: - type q6_K: 33 tensors print_info: file format = GGUF V3 (latest) print_info: file type = Q4_K - Medium print_info: file size = 4.30 GiB (4.92 BPW) llama_model_load: error loading model: error loading model architecture: unknown model architecture: 'hunyuan-dense' llama_model_load_from_file_impl: failed to load model time=2025-08-07T13:32:01.020+08:00 level=INFO source=sched.go:455 msg="NewLlamaServer failed" model=/Users/sven/.ollama/models/blobs/sha256-e59a4dd7d94142c65e0aef5e20b0a44637467b472628dac00dfe9be23dc4d18d error="unable to load model: /Users/sven/.ollama/models/blobs/sha256-e59a4dd7d94142c65e0aef5e20b0a44637467b472628dac00dfe9be23dc4d18d"
GiteaMirror commented 2026-04-12 19:48:54 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Aug 7, 2025):

llama_model_load: error loading model: error loading model architecture: unknown model architecture: 'hunyuan-dense'

The model is currently unsupported.

<!-- gh-comment-id:3163045477 --> @rick-github commented on GitHub (Aug 7, 2025): ``` llama_model_load: error loading model: error loading model architecture: unknown model architecture: 'hunyuan-dense' ``` The model is currently unsupported.
GiteaMirror commented 2026-04-12 19:48:54 -05:00
Author
Owner
Copy Link

@frenchmustard commented on GitHub (Aug 15, 2025):

llama_model_load: error loading model: error loading model architecture: unknown model architecture: 'hunyuan-dense'

The model is currently unsupported.

Doesn’t llama.cpp support it? I thought the necessary commits were merged by them last month?

<!-- gh-comment-id:3190395162 --> @frenchmustard commented on GitHub (Aug 15, 2025): > ``` > llama_model_load: error loading model: error loading model architecture: unknown model architecture: 'hunyuan-dense' > ``` > > The model is currently unsupported. Doesn’t llama.cpp support it? I thought the necessary commits were merged by them last month?
GiteaMirror commented 2026-04-12 19:48:55 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Aug 15, 2025):

llama.cpp supports it. ollama merged #11823 four hours ago, which enables ollama support.

<!-- gh-comment-id:3190399914 --> @rick-github commented on GitHub (Aug 15, 2025): llama.cpp supports it. ollama merged #11823 four hours ago, which enables ollama support.
GiteaMirror commented 2026-04-12 19:48:55 -05:00
Author
Owner
Copy Link

@frenchmustard commented on GitHub (Aug 15, 2025):

llama.cpp supports it. ollama merged #11823 four hours ago, which enables ollama support.

Any idea on when this update is pushed? If it will take a while ill try and see if I can get the main branch running locally right now.

<!-- gh-comment-id:3191385775 --> @frenchmustard commented on GitHub (Aug 15, 2025): > llama.cpp supports it. ollama merged [#11823](https://github.com/ollama/ollama/pull/11823) four hours ago, which enables ollama support. Any idea on when this update is pushed? If it will take a while ill try and see if I can get the main branch running locally right now.
GiteaMirror commented 2026-04-12 19:48:55 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Aug 15, 2025):

https://github.com/ollama/ollama/releases/tag/v0.11.5-rc2

However, the model doesn't have the correct template or stop parameters so while it now loads, it generates gibberish.

<!-- gh-comment-id:3191389883 --> @rick-github commented on GitHub (Aug 15, 2025): https://github.com/ollama/ollama/releases/tag/v0.11.5-rc2 However, the model doesn't have the correct template or stop parameters so while it now loads, it generates gibberish.
GiteaMirror commented 2026-04-12 19:48:56 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Aug 15, 2025):

FROM alibayram/hunyuan:0.5b
TEMPLATE """
{{- $lastUserIdx := -1 }}
{{- range $i, $_ := .Messages }}
{{- if eq .Role "user" }}{{- $lastUserIdx = $i }}{{ end }}
{{- end -}}
<｜hy_begin▁of▁sentence｜>
{{- if .System }}
{{- .System }}<｜hy_place▁holder▁no▁3｜>
{{- end }}
{{- range $i, $_ := .Messages }}
{{- $last := eq (len (slice $.Messages $i)) 1 }}
{{- if eq .Role "user" }}<｜hy_User｜>
{{- .Content }}
{{- else if eq .Role "assistant" }}<｜hy_Assistant｜>
{{- if (and $.IsThinkSet (and .Thinking (or $last (gt $i $lastUserIdx)))) -}}
<think>{{ .Thinking }}</think>
{{- end }}
{{ .Content }}<｜hy_place▁holder▁no▁2｜>
{{- end }}
{{- if and (ne .Role "assistant") $last }}<｜hy_Assistant｜>
{{- if and $.IsThinkSet (not $.Think) -}}
<think>

</think>
{{ end -}}
{{- end }}
{{- end }}"""
PARAMETER stop <｜hy_begin▁of▁sentence｜>
PARAMETER stop <｜hy_place▁holder▁no▁2｜>
PARAMETER stop <｜hy_place▁holder▁no▁3｜>
PARAMETER stop <｜hy_User｜>
PARAMETER stop <｜hy_Assistant｜>

$ ollama run hunyuan:0.5b hello 
Thinking...
Okay, the user said "hello". I need to respond in a friendly and welcoming way. Let me start with a greeting. 
Maybe something like "Hello there!" That's standard for any chat. Then, maybe offer further assistance. I should 
keep it simple and open-ended so they can ask more questions or have anything else. Let me check if there's 
anything else needed. Oh, right, sometimes people get off to sleep, so maybe a light touch about that? Or 
perhaps a reminder of how much time is left. But the main thing is to be welcoming. So putting it together: 
"Hello! How can I help you today?" That should work.
...done thinking.

<answer>
Hello there! How can I help you today?
</answer>

$ ollama run hunyuan:0.5b hello --think=false
<answer>
Hello! How can I assist you today?
</answer>

There doesn't seem to be a simple way to strip out the <answer> tags.

<!-- gh-comment-id:3191593997 --> @rick-github commented on GitHub (Aug 15, 2025): ``` FROM alibayram/hunyuan:0.5b TEMPLATE """ {{- $lastUserIdx := -1 }} {{- range $i, $_ := .Messages }} {{- if eq .Role "user" }}{{- $lastUserIdx = $i }}{{ end }} {{- end -}} <｜hy_begin▁of▁sentence｜> {{- if .System }} {{- .System }}<｜hy_place▁holder▁no▁3｜> {{- end }} {{- range $i, $_ := .Messages }} {{- $last := eq (len (slice $.Messages $i)) 1 }} {{- if eq .Role "user" }}<｜hy_User｜> {{- .Content }} {{- else if eq .Role "assistant" }}<｜hy_Assistant｜> {{- if (and $.IsThinkSet (and .Thinking (or $last (gt $i $lastUserIdx)))) -}} <think>{{ .Thinking }}</think> {{- end }} {{ .Content }}<｜hy_place▁holder▁no▁2｜> {{- end }} {{- if and (ne .Role "assistant") $last }}<｜hy_Assistant｜> {{- if and $.IsThinkSet (not $.Think) -}} <think> </think> {{ end -}} {{- end }} {{- end }}""" PARAMETER stop <｜hy_begin▁of▁sentence｜> PARAMETER stop <｜hy_place▁holder▁no▁2｜> PARAMETER stop <｜hy_place▁holder▁no▁3｜> PARAMETER stop <｜hy_User｜> PARAMETER stop <｜hy_Assistant｜> ``` ```console $ ollama run hunyuan:0.5b hello Thinking... Okay, the user said "hello". I need to respond in a friendly and welcoming way. Let me start with a greeting. Maybe something like "Hello there!" That's standard for any chat. Then, maybe offer further assistance. I should keep it simple and open-ended so they can ask more questions or have anything else. Let me check if there's anything else needed. Oh, right, sometimes people get off to sleep, so maybe a light touch about that? Or perhaps a reminder of how much time is left. But the main thing is to be welcoming. So putting it together: "Hello! How can I help you today?" That should work. ...done thinking. <answer> Hello there! How can I help you today? </answer> ``` ```console $ ollama run hunyuan:0.5b hello --think=false <answer> Hello! How can I assist you today? </answer> ``` There doesn't seem to be a simple way to strip out the `<answer>` tags.
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 bug
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#7709
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?