github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-05 23:53:43 -05:00
Code Issues 14.5k Packages Projects Releases 120 Wiki Activity

[GH-ISSUE #5095] add support Alibaba-NLP/gte-Qwen2-7B-instruct #3216

New Issue
Closed
opened 2026-04-12 13:43:12 -05:00 by GiteaMirror · 7 comments
GiteaMirror commented 2026-04-12 13:43:12 -05:00
Owner
Copy Link

Originally created by @louyongjiu on GitHub (Jun 17, 2024).
Original GitHub issue: https://github.com/ollama/ollama/issues/5095

https://huggingface.co/Alibaba-NLP/gte-Qwen2-7B-instruct

PixPin_2024-06-17_17-35-17

Originally created by @louyongjiu on GitHub (Jun 17, 2024). Original GitHub issue: https://github.com/ollama/ollama/issues/5095 https://huggingface.co/Alibaba-NLP/gte-Qwen2-7B-instruct ![PixPin_2024-06-17_17-35-17](https://github.com/ollama/ollama/assets/16408477/67883d1c-1746-4065-98be-5464009793f1)
GiteaMirror added the model label 2026-04-12 13:43:12 -05:00
GiteaMirror commented 2026-04-12 13:43:15 -05:00
Author
Owner
Copy Link

@palindsay commented on GitHub (Jun 27, 2024):

Looks like llama.cpp needs to be updated:
./main -m models/gte-Qwen2-7B-instruct.gguf
Log start
main: build = 3041 (eb57fee5)
main: built with cc (Ubuntu 13.2.0-23ubuntu4) 13.2.0 for x86_64-linux-gnu
main: seed = 1719467592
llama_model_loader: loaded meta data with 22 key-value pairs and 339 tensors from models/gte-Qwen2-7B-instruct.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.name str = gte-Qwen2-7B-instruct
llama_model_loader: - kv 2: qwen2.block_count u32 = 28
llama_model_loader: - kv 3: qwen2.context_length u32 = 131072
llama_model_loader: - kv 4: qwen2.embedding_length u32 = 3584
llama_model_loader: - kv 5: qwen2.feed_forward_length u32 = 18944
llama_model_loader: - kv 6: qwen2.attention.head_count u32 = 28
llama_model_loader: - kv 7: qwen2.attention.head_count_kv u32 = 4
llama_model_loader: - kv 8: qwen2.rope.freq_base f32 = 1000000.000000
llama_model_loader: - kv 9: qwen2.attention.layer_norm_rms_epsilon f32 = 0.000001
llama_model_loader: - kv 10: general.file_type u32 = 32
llama_model_loader: - kv 11: tokenizer.ggml.model str = gpt2
llama_model_loader: - kv 12: tokenizer.ggml.pre str = qwen2
llama_model_loader: - kv 13: tokenizer.ggml.tokens arr[str,151646] = ["!", """, "#", "$", "%", "&", "'", ...
llama_model_loader: - kv 14: tokenizer.ggml.token_type arr[i32,151646] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ...
llama_model_loader: - kv 15: tokenizer.ggml.merges arr[str,151387] = ["Ġ Ġ", "ĠĠ ĠĠ", "i n", "Ġ t",...
llama_model_loader: - kv 16: tokenizer.ggml.eos_token_id u32 = 151643
llama_model_loader: - kv 17: tokenizer.ggml.padding_token_id u32 = 151643
llama_model_loader: - kv 18: tokenizer.ggml.bos_token_id u32 = 151643
llama_model_loader: - kv 19: tokenizer.ggml.add_eos_token bool = true
llama_model_loader: - kv 20: tokenizer.chat_template str = {% for message in messages %}{{'<|im_...
llama_model_loader: - kv 21: general.quantization_version u32 = 2
llama_model_loader: - type f32: 141 tensors
llama_model_loader: - type bf16: 198 tensors
llm_load_vocab: special tokens cache size = 3.
llm_load_print_meta: format = GGUF V3 (latest)
llm_load_print_meta: arch = qwen2
llm_load_print_meta: vocab type = BPE
llm_load_print_meta: n_vocab = 151646
llm_load_print_meta: n_merges = 151387
llm_load_print_meta: n_ctx_train = 131072
llm_load_print_meta: n_embd = 3584
llm_load_print_meta: n_head = 28
llm_load_print_meta: n_head_kv = 4
llm_load_print_meta: n_layer = 28
llm_load_print_meta: n_rot = 128
llm_load_print_meta: n_embd_head_k = 128
llm_load_print_meta: n_embd_head_v = 128
llm_load_print_meta: n_gqa = 7
llm_load_print_meta: n_embd_k_gqa = 512
llm_load_print_meta: n_embd_v_gqa = 512
llm_load_print_meta: f_norm_eps = 0.0e+00
llm_load_print_meta: f_norm_rms_eps = 1.0e-06
llm_load_print_meta: f_clamp_kqv = 0.0e+00
llm_load_print_meta: f_max_alibi_bias = 0.0e+00
llm_load_print_meta: f_logit_scale = 0.0e+00
llm_load_print_meta: n_ff = 18944
llm_load_print_meta: n_expert = 0
llm_load_print_meta: n_expert_used = 0
llm_load_print_meta: causal attn = 1
llm_load_print_meta: pooling type = 0
llm_load_print_meta: rope type = 2
llm_load_print_meta: rope scaling = linear
llm_load_print_meta: freq_base_train = 1000000.0
llm_load_print_meta: freq_scale_train = 1
llm_load_print_meta: n_yarn_orig_ctx = 131072
llm_load_print_meta: rope_finetuned = unknown
llm_load_print_meta: ssm_d_conv = 0
llm_load_print_meta: ssm_d_inner = 0
llm_load_print_meta: ssm_d_state = 0
llm_load_print_meta: ssm_dt_rank = 0
llm_load_print_meta: model type = ?B
llm_load_print_meta: model ftype = BF16
llm_load_print_meta: model params = 7.61 B
llm_load_print_meta: model size = 14.18 GiB (16.00 BPW)
llm_load_print_meta: general.name = gte-Qwen2-7B-instruct
llm_load_print_meta: BOS token = 151643 '<|endoftext|>'
llm_load_print_meta: EOS token = 151643 '<|endoftext|>'
llm_load_print_meta: PAD token = 151643 '<|endoftext|>'
llm_load_print_meta: LF token = 148848 'ÄĬ'
llm_load_print_meta: EOT token = 151645 '<|im_end|>'
llm_load_tensors: ggml ctx size = 0.16 MiB
llm_load_tensors: CPU buffer size = 14520.56 MiB
........................................................................................
llama_new_context_with_model: n_ctx = 512
llama_new_context_with_model: n_batch = 512
llama_new_context_with_model: n_ubatch = 512
llama_new_context_with_model: flash_attn = 0
llama_new_context_with_model: freq_base = 1000000.0
llama_new_context_with_model: freq_scale = 1
llama_kv_cache_init: CPU KV buffer size = 28.00 MiB
llama_new_context_with_model: KV self size = 28.00 MiB, K (f16): 14.00 MiB, V (f16): 14.00 MiB
llama_new_context_with_model: CPU output buffer size = 0.58 MiB
llama_new_context_with_model: CPU compute buffer size = 303.18 MiB
llama_new_context_with_model: graph nodes = 986
llama_new_context_with_model: graph splits = 1

system_info: n_threads = 16 / 32 | AVX = 1 | AVX_VNNI = 0 | AVX2 = 1 | AVX512 = 1 | AVX512_VBMI = 1 | AVX512_VNNI = 1 | AVX512_BF16 = 1 | FMA = 1 | NEON = 0 | SVE = 0 | ARM_FMA = 0 | F16C = 1 | FP16_VA = 0 | WASM_SIMD = 0 | BLAS = 0 | SSE3 = 1 | SSSE3 = 1 | VSX = 0 | MATMUL_INT8 = 0 | LLAMAFILE = 1 |
GGML_ASSERT: examples/main/main.cpp:248: llama_add_eos_token(model) != 1
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.
zsh: IOT instruction (core dumped) ./main -m models/gte-Qwen2-7B-instruct.gguf

<!-- gh-comment-id:2193889508 --> @palindsay commented on GitHub (Jun 27, 2024): Looks like llama.cpp needs to be updated: ./main -m models/gte-Qwen2-7B-instruct.gguf Log start main: build = 3041 (eb57fee5) main: built with cc (Ubuntu 13.2.0-23ubuntu4) 13.2.0 for x86_64-linux-gnu main: seed = 1719467592 llama_model_loader: loaded meta data with 22 key-value pairs and 339 tensors from models/gte-Qwen2-7B-instruct.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.name str = gte-Qwen2-7B-instruct llama_model_loader: - kv 2: qwen2.block_count u32 = 28 llama_model_loader: - kv 3: qwen2.context_length u32 = 131072 llama_model_loader: - kv 4: qwen2.embedding_length u32 = 3584 llama_model_loader: - kv 5: qwen2.feed_forward_length u32 = 18944 llama_model_loader: - kv 6: qwen2.attention.head_count u32 = 28 llama_model_loader: - kv 7: qwen2.attention.head_count_kv u32 = 4 llama_model_loader: - kv 8: qwen2.rope.freq_base f32 = 1000000.000000 llama_model_loader: - kv 9: qwen2.attention.layer_norm_rms_epsilon f32 = 0.000001 llama_model_loader: - kv 10: general.file_type u32 = 32 llama_model_loader: - kv 11: tokenizer.ggml.model str = gpt2 llama_model_loader: - kv 12: tokenizer.ggml.pre str = qwen2 llama_model_loader: - kv 13: tokenizer.ggml.tokens arr[str,151646] = ["!", "\"", "#", "$", "%", "&", "'", ... llama_model_loader: - kv 14: tokenizer.ggml.token_type arr[i32,151646] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ... llama_model_loader: - kv 15: tokenizer.ggml.merges arr[str,151387] = ["Ġ Ġ", "ĠĠ ĠĠ", "i n", "Ġ t",... llama_model_loader: - kv 16: tokenizer.ggml.eos_token_id u32 = 151643 llama_model_loader: - kv 17: tokenizer.ggml.padding_token_id u32 = 151643 llama_model_loader: - kv 18: tokenizer.ggml.bos_token_id u32 = 151643 llama_model_loader: - kv 19: tokenizer.ggml.add_eos_token bool = true llama_model_loader: - kv 20: tokenizer.chat_template str = {% for message in messages %}{{'<|im_... llama_model_loader: - kv 21: general.quantization_version u32 = 2 llama_model_loader: - type f32: 141 tensors llama_model_loader: - type bf16: 198 tensors llm_load_vocab: special tokens cache size = 3. llm_load_print_meta: format = GGUF V3 (latest) llm_load_print_meta: arch = qwen2 llm_load_print_meta: vocab type = BPE llm_load_print_meta: n_vocab = 151646 llm_load_print_meta: n_merges = 151387 llm_load_print_meta: n_ctx_train = 131072 llm_load_print_meta: n_embd = 3584 llm_load_print_meta: n_head = 28 llm_load_print_meta: n_head_kv = 4 llm_load_print_meta: n_layer = 28 llm_load_print_meta: n_rot = 128 llm_load_print_meta: n_embd_head_k = 128 llm_load_print_meta: n_embd_head_v = 128 llm_load_print_meta: n_gqa = 7 llm_load_print_meta: n_embd_k_gqa = 512 llm_load_print_meta: n_embd_v_gqa = 512 llm_load_print_meta: f_norm_eps = 0.0e+00 llm_load_print_meta: f_norm_rms_eps = 1.0e-06 llm_load_print_meta: f_clamp_kqv = 0.0e+00 llm_load_print_meta: f_max_alibi_bias = 0.0e+00 llm_load_print_meta: f_logit_scale = 0.0e+00 llm_load_print_meta: n_ff = 18944 llm_load_print_meta: n_expert = 0 llm_load_print_meta: n_expert_used = 0 llm_load_print_meta: causal attn = 1 llm_load_print_meta: pooling type = 0 llm_load_print_meta: rope type = 2 llm_load_print_meta: rope scaling = linear llm_load_print_meta: freq_base_train = 1000000.0 llm_load_print_meta: freq_scale_train = 1 llm_load_print_meta: n_yarn_orig_ctx = 131072 llm_load_print_meta: rope_finetuned = unknown llm_load_print_meta: ssm_d_conv = 0 llm_load_print_meta: ssm_d_inner = 0 llm_load_print_meta: ssm_d_state = 0 llm_load_print_meta: ssm_dt_rank = 0 llm_load_print_meta: model type = ?B llm_load_print_meta: model ftype = BF16 llm_load_print_meta: model params = 7.61 B llm_load_print_meta: model size = 14.18 GiB (16.00 BPW) llm_load_print_meta: general.name = gte-Qwen2-7B-instruct llm_load_print_meta: BOS token = 151643 '<|endoftext|>' llm_load_print_meta: EOS token = 151643 '<|endoftext|>' llm_load_print_meta: PAD token = 151643 '<|endoftext|>' llm_load_print_meta: LF token = 148848 'ÄĬ' llm_load_print_meta: EOT token = 151645 '<|im_end|>' llm_load_tensors: ggml ctx size = 0.16 MiB llm_load_tensors: CPU buffer size = 14520.56 MiB ........................................................................................ llama_new_context_with_model: n_ctx = 512 llama_new_context_with_model: n_batch = 512 llama_new_context_with_model: n_ubatch = 512 llama_new_context_with_model: flash_attn = 0 llama_new_context_with_model: freq_base = 1000000.0 llama_new_context_with_model: freq_scale = 1 llama_kv_cache_init: CPU KV buffer size = 28.00 MiB llama_new_context_with_model: KV self size = 28.00 MiB, K (f16): 14.00 MiB, V (f16): 14.00 MiB llama_new_context_with_model: CPU output buffer size = 0.58 MiB llama_new_context_with_model: CPU compute buffer size = 303.18 MiB llama_new_context_with_model: graph nodes = 986 llama_new_context_with_model: graph splits = 1 system_info: n_threads = 16 / 32 | AVX = 1 | AVX_VNNI = 0 | AVX2 = 1 | AVX512 = 1 | AVX512_VBMI = 1 | AVX512_VNNI = 1 | AVX512_BF16 = 1 | FMA = 1 | NEON = 0 | SVE = 0 | ARM_FMA = 0 | F16C = 1 | FP16_VA = 0 | WASM_SIMD = 0 | BLAS = 0 | SSE3 = 1 | SSSE3 = 1 | VSX = 0 | MATMUL_INT8 = 0 | LLAMAFILE = 1 | GGML_ASSERT: examples/main/main.cpp:248: llama_add_eos_token(model) != 1 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. zsh: IOT instruction (core dumped) ./main -m models/gte-Qwen2-7B-instruct.gguf
GiteaMirror commented 2026-04-12 13:43:16 -05:00
Author
Owner
Copy Link

@palindsay commented on GitHub (Jun 27, 2024):

Opened llama.cpp issue here: https://github.com/ggerganov/llama.cpp/issues/8152

<!-- gh-comment-id:2193901317 --> @palindsay commented on GitHub (Jun 27, 2024): Opened llama.cpp issue here: https://github.com/ggerganov/llama.cpp/issues/8152
GiteaMirror commented 2026-04-12 13:43:17 -05:00
Author
Owner
Copy Link

@louyongjiu commented on GitHub (Jun 27, 2024):

Opened llama.cpp issue here: ggerganov/llama.cpp#8152
already added
https://ollama.com/rjmalagon/gte-qwen2-7b-instruct-embed-f16

<!-- gh-comment-id:2194191659 --> @louyongjiu commented on GitHub (Jun 27, 2024): > Opened llama.cpp issue here: [ggerganov/llama.cpp#8152](https://github.com/ggerganov/llama.cpp/issues/8152) already added https://ollama.com/rjmalagon/gte-qwen2-7b-instruct-embed-f16
GiteaMirror commented 2026-04-12 13:43:17 -05:00
Author
Owner
Copy Link

@rjmalagon commented on GitHub (Jun 28, 2024):

Hi, just to humble clarify. I converted this model from the Hugging Face repo, it loads and runs on Ollama as an embed model, but I can't affirm it is working correctly or at its full capacity.

Early testing on some diverse materials returns enough good results, but I have this “feeling” that Qwen2 models support, including this GTE for embedding, need some extra work or refining on llama.cpp. I say this just as a user and not an expert.

Take note that the current Ollama build and repo use llama.cpp from 2 weeks ago 7c26775adb and may include (and/or lack) some additional fixes or patches.

<!-- gh-comment-id:2197147071 --> @rjmalagon commented on GitHub (Jun 28, 2024): Hi, just to humble clarify. I converted this model from the Hugging Face repo, it loads and runs on Ollama as an embed model, but I can't affirm it is working correctly or at its full capacity. Early testing on some diverse materials returns enough good results, but I have this “feeling” that Qwen2 models support, including this GTE for embedding, need some extra work or refining on llama.cpp. I say this just as a user and not an expert. Take note that the current Ollama build and repo use llama.cpp from 2 weeks ago https://github.com/ggerganov/llama.cpp/tree/7c26775adb579e92b59c82e8084c07a1d0f75e9c and may include (and/or lack) some additional fixes or patches.
GiteaMirror commented 2026-04-12 13:43:18 -05:00
Author
Owner
Copy Link

@rjmalagon commented on GitHub (Jun 28, 2024):

Looks like llama.cpp needs to be updated: ./main -m models/gte-Qwen2-7B-instruct.gguf Log start main: build = 3041 (eb57fee) main: built with cc (Ubuntu 13.2.0-23ubuntu4) 13.2.0 for x86_64-linux-gnu main: seed = 1719467592 llama_model_loader: loaded meta data with 22 key-value pairs and 339 tensors from models/gte-Qwen2-7B-instruct.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.name str = gte-Qwen2-7B-instruct llama_model_loader: - kv 2: qwen2.block_count u32 = 28 llama_model_loader: - kv 3: qwen2.context_length u32 = 131072 llama_model_loader: - kv 4: qwen2.embedding_length u32 = 3584 llama_model_loader: - kv 5: qwen2.feed_forward_length u32 = 18944 llama_model_loader: - kv 6: qwen2.attention.head_count u32 = 28 llama_model_loader: - kv 7: qwen2.attention.head_count_kv u32 = 4 llama_model_loader: - kv 8: qwen2.rope.freq_base f32 = 1000000.000000 llama_model_loader: - kv 9: qwen2.attention.layer_norm_rms_epsilon f32 = 0.000001 llama_model_loader: - kv 10: general.file_type u32 = 32 llama_model_loader: - kv 11: tokenizer.ggml.model str = gpt2 llama_model_loader: - kv 12: tokenizer.ggml.pre str = qwen2 llama_model_loader: - kv 13: tokenizer.ggml.tokens arr[str,151646] = ["!", """, "#", "$", "%", "&", "'", ... llama_model_loader: - kv 14: tokenizer.ggml.token_type arr[i32,151646] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ... llama_model_loader: - kv 15: tokenizer.ggml.merges arr[str,151387] = ["Ġ Ġ", "ĠĠ ĠĠ", "i n", "Ġ t",... llama_model_loader: - kv 16: tokenizer.ggml.eos_token_id u32 = 151643 llama_model_loader: - kv 17: tokenizer.ggml.padding_token_id u32 = 151643 llama_model_loader: - kv 18: tokenizer.ggml.bos_token_id u32 = 151643 llama_model_loader: - kv 19: tokenizer.ggml.add_eos_token bool = true llama_model_loader: - kv 20: tokenizer.chat_template str = {% for message in messages %}{{'<|im_... llama_model_loader: - kv 21: general.quantization_version u32 = 2 llama_model_loader: - type f32: 141 tensors llama_model_loader: - type bf16: 198 tensors llm_load_vocab: special tokens cache size = 3. llm_load_print_meta: format = GGUF V3 (latest) llm_load_print_meta: arch = qwen2 llm_load_print_meta: vocab type = BPE llm_load_print_meta: n_vocab = 151646 llm_load_print_meta: n_merges = 151387 llm_load_print_meta: n_ctx_train = 131072 llm_load_print_meta: n_embd = 3584 llm_load_print_meta: n_head = 28 llm_load_print_meta: n_head_kv = 4 llm_load_print_meta: n_layer = 28 llm_load_print_meta: n_rot = 128 llm_load_print_meta: n_embd_head_k = 128 llm_load_print_meta: n_embd_head_v = 128 llm_load_print_meta: n_gqa = 7 llm_load_print_meta: n_embd_k_gqa = 512 llm_load_print_meta: n_embd_v_gqa = 512 llm_load_print_meta: f_norm_eps = 0.0e+00 llm_load_print_meta: f_norm_rms_eps = 1.0e-06 llm_load_print_meta: f_clamp_kqv = 0.0e+00 llm_load_print_meta: f_max_alibi_bias = 0.0e+00 llm_load_print_meta: f_logit_scale = 0.0e+00 llm_load_print_meta: n_ff = 18944 llm_load_print_meta: n_expert = 0 llm_load_print_meta: n_expert_used = 0 llm_load_print_meta: causal attn = 1 llm_load_print_meta: pooling type = 0 llm_load_print_meta: rope type = 2 llm_load_print_meta: rope scaling = linear llm_load_print_meta: freq_base_train = 1000000.0 llm_load_print_meta: freq_scale_train = 1 llm_load_print_meta: n_yarn_orig_ctx = 131072 llm_load_print_meta: rope_finetuned = unknown llm_load_print_meta: ssm_d_conv = 0 llm_load_print_meta: ssm_d_inner = 0 llm_load_print_meta: ssm_d_state = 0 llm_load_print_meta: ssm_dt_rank = 0 llm_load_print_meta: model type = ?B llm_load_print_meta: model ftype = BF16 llm_load_print_meta: model params = 7.61 B llm_load_print_meta: model size = 14.18 GiB (16.00 BPW) llm_load_print_meta: general.name = gte-Qwen2-7B-instruct llm_load_print_meta: BOS token = 151643 '<|endoftext|>' llm_load_print_meta: EOS token = 151643 '<|endoftext|>' llm_load_print_meta: PAD token = 151643 '<|endoftext|>' llm_load_print_meta: LF token = 148848 'ÄĬ' llm_load_print_meta: EOT token = 151645 '<|im_end|>' llm_load_tensors: ggml ctx size = 0.16 MiB llm_load_tensors: CPU buffer size = 14520.56 MiB ........................................................................................ llama_new_context_with_model: n_ctx = 512 llama_new_context_with_model: n_batch = 512 llama_new_context_with_model: n_ubatch = 512 llama_new_context_with_model: flash_attn = 0 llama_new_context_with_model: freq_base = 1000000.0 llama_new_context_with_model: freq_scale = 1 llama_kv_cache_init: CPU KV buffer size = 28.00 MiB llama_new_context_with_model: KV self size = 28.00 MiB, K (f16): 14.00 MiB, V (f16): 14.00 MiB llama_new_context_with_model: CPU output buffer size = 0.58 MiB llama_new_context_with_model: CPU compute buffer size = 303.18 MiB llama_new_context_with_model: graph nodes = 986 llama_new_context_with_model: graph splits = 1

system_info: n_threads = 16 / 32 | AVX = 1 | AVX_VNNI = 0 | AVX2 = 1 | AVX512 = 1 | AVX512_VBMI = 1 | AVX512_VNNI = 1 | AVX512_BF16 = 1 | FMA = 1 | NEON = 0 | SVE = 0 | ARM_FMA = 0 | F16C = 1 | FP16_VA = 0 | WASM_SIMD = 0 | BLAS = 0 | SSE3 = 1 | SSSE3 = 1 | VSX = 0 | MATMUL_INT8 = 0 | LLAMAFILE = 1 | GGML_ASSERT: examples/main/main.cpp:248: llama_add_eos_token(model) != 1 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. zsh: IOT instruction (core dumped) ./main -m models/gte-Qwen2-7B-instruct.gguf

Looks like llama.cpp needs to be updated: ./main -m models/gte-Qwen2-7B-instruct.gguf Log start main: build = 3041 (eb57fee) main: built with cc (Ubuntu 13.2.0-23ubuntu4) 13.2.0 for x86_64-linux-gnu main: seed = 1719467592 llama_model_loader: loaded meta data with 22 key-value pairs and 339 tensors from models/gte-Qwen2-7B-instruct.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.name str = gte-Qwen2-7B-instruct llama_model_loader: - kv 2: qwen2.block_count u32 = 28 llama_model_loader: - kv 3: qwen2.context_length u32 = 131072 llama_model_loader: - kv 4: qwen2.embedding_length u32 = 3584 llama_model_loader: - kv 5: qwen2.feed_forward_length u32 = 18944 llama_model_loader: - kv 6: qwen2.attention.head_count u32 = 28 llama_model_loader: - kv 7: qwen2.attention.head_count_kv u32 = 4 llama_model_loader: - kv 8: qwen2.rope.freq_base f32 = 1000000.000000 llama_model_loader: - kv 9: qwen2.attention.layer_norm_rms_epsilon f32 = 0.000001 llama_model_loader: - kv 10: general.file_type u32 = 32 llama_model_loader: - kv 11: tokenizer.ggml.model str = gpt2 llama_model_loader: - kv 12: tokenizer.ggml.pre str = qwen2 llama_model_loader: - kv 13: tokenizer.ggml.tokens arr[str,151646] = ["!", """, "#", "$", "%", "&", "'", ... llama_model_loader: - kv 14: tokenizer.ggml.token_type arr[i32,151646] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ... llama_model_loader: - kv 15: tokenizer.ggml.merges arr[str,151387] = ["Ġ Ġ", "ĠĠ ĠĠ", "i n", "Ġ t",... llama_model_loader: - kv 16: tokenizer.ggml.eos_token_id u32 = 151643 llama_model_loader: - kv 17: tokenizer.ggml.padding_token_id u32 = 151643 llama_model_loader: - kv 18: tokenizer.ggml.bos_token_id u32 = 151643 llama_model_loader: - kv 19: tokenizer.ggml.add_eos_token bool = true llama_model_loader: - kv 20: tokenizer.chat_template str = {% for message in messages %}{{'<|im_... llama_model_loader: - kv 21: general.quantization_version u32 = 2 llama_model_loader: - type f32: 141 tensors llama_model_loader: - type bf16: 198 tensors llm_load_vocab: special tokens cache size = 3. llm_load_print_meta: format = GGUF V3 (latest) llm_load_print_meta: arch = qwen2 llm_load_print_meta: vocab type = BPE llm_load_print_meta: n_vocab = 151646 llm_load_print_meta: n_merges = 151387 llm_load_print_meta: n_ctx_train = 131072 llm_load_print_meta: n_embd = 3584 llm_load_print_meta: n_head = 28 llm_load_print_meta: n_head_kv = 4 llm_load_print_meta: n_layer = 28 llm_load_print_meta: n_rot = 128 llm_load_print_meta: n_embd_head_k = 128 llm_load_print_meta: n_embd_head_v = 128 llm_load_print_meta: n_gqa = 7 llm_load_print_meta: n_embd_k_gqa = 512 llm_load_print_meta: n_embd_v_gqa = 512 llm_load_print_meta: f_norm_eps = 0.0e+00 llm_load_print_meta: f_norm_rms_eps = 1.0e-06 llm_load_print_meta: f_clamp_kqv = 0.0e+00 llm_load_print_meta: f_max_alibi_bias = 0.0e+00 llm_load_print_meta: f_logit_scale = 0.0e+00 llm_load_print_meta: n_ff = 18944 llm_load_print_meta: n_expert = 0 llm_load_print_meta: n_expert_used = 0 llm_load_print_meta: causal attn = 1 llm_load_print_meta: pooling type = 0 llm_load_print_meta: rope type = 2 llm_load_print_meta: rope scaling = linear llm_load_print_meta: freq_base_train = 1000000.0 llm_load_print_meta: freq_scale_train = 1 llm_load_print_meta: n_yarn_orig_ctx = 131072 llm_load_print_meta: rope_finetuned = unknown llm_load_print_meta: ssm_d_conv = 0 llm_load_print_meta: ssm_d_inner = 0 llm_load_print_meta: ssm_d_state = 0 llm_load_print_meta: ssm_dt_rank = 0 llm_load_print_meta: model type = ?B llm_load_print_meta: model ftype = BF16 llm_load_print_meta: model params = 7.61 B llm_load_print_meta: model size = 14.18 GiB (16.00 BPW) llm_load_print_meta: general.name = gte-Qwen2-7B-instruct llm_load_print_meta: BOS token = 151643 '<|endoftext|>' llm_load_print_meta: EOS token = 151643 '<|endoftext|>' llm_load_print_meta: PAD token = 151643 '<|endoftext|>' llm_load_print_meta: LF token = 148848 'ÄĬ' llm_load_print_meta: EOT token = 151645 '<|im_end|>' llm_load_tensors: ggml ctx size = 0.16 MiB llm_load_tensors: CPU buffer size = 14520.56 MiB ........................................................................................ llama_new_context_with_model: n_ctx = 512 llama_new_context_with_model: n_batch = 512 llama_new_context_with_model: n_ubatch = 512 llama_new_context_with_model: flash_attn = 0 llama_new_context_with_model: freq_base = 1000000.0 llama_new_context_with_model: freq_scale = 1 llama_kv_cache_init: CPU KV buffer size = 28.00 MiB llama_new_context_with_model: KV self size = 28.00 MiB, K (f16): 14.00 MiB, V (f16): 14.00 MiB llama_new_context_with_model: CPU output buffer size = 0.58 MiB llama_new_context_with_model: CPU compute buffer size = 303.18 MiB llama_new_context_with_model: graph nodes = 986 llama_new_context_with_model: graph splits = 1

system_info: n_threads = 16 / 32 | AVX = 1 | AVX_VNNI = 0 | AVX2 = 1 | AVX512 = 1 | AVX512_VBMI = 1 | AVX512_VNNI = 1 | AVX512_BF16 = 1 | FMA = 1 | NEON = 0 | SVE = 0 | ARM_FMA = 0 | F16C = 1 | FP16_VA = 0 | WASM_SIMD = 0 | BLAS = 0 | SSE3 = 1 | SSSE3 = 1 | VSX = 0 | MATMUL_INT8 = 0 | LLAMAFILE = 1 | GGML_ASSERT: examples/main/main.cpp:248: llama_add_eos_token(model) != 1 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. zsh: IOT instruction (core dumped) ./main -m models/gte-Qwen2-7B-instruct.gguf

Looks like llama.cpp needs to be updated: ./main -m models/gte-Qwen2-7B-instruct.gguf Log start main: build = 3041 (eb57fee) main: built with cc (Ubuntu 13.2.0-23ubuntu4) 13.2.0 for x86_64-linux-gnu main: seed = 1719467592 llama_model_loader: loaded meta data with 22 key-value pairs and 339 tensors from models/gte-Qwen2-7B-instruct.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.name str = gte-Qwen2-7B-instruct llama_model_loader: - kv 2: qwen2.block_count u32 = 28 llama_model_loader: - kv 3: qwen2.context_length u32 = 131072 llama_model_loader: - kv 4: qwen2.embedding_length u32 = 3584 llama_model_loader: - kv 5: qwen2.feed_forward_length u32 = 18944 llama_model_loader: - kv 6: qwen2.attention.head_count u32 = 28 llama_model_loader: - kv 7: qwen2.attention.head_count_kv u32 = 4 llama_model_loader: - kv 8: qwen2.rope.freq_base f32 = 1000000.000000 llama_model_loader: - kv 9: qwen2.attention.layer_norm_rms_epsilon f32 = 0.000001 llama_model_loader: - kv 10: general.file_type u32 = 32 llama_model_loader: - kv 11: tokenizer.ggml.model str = gpt2 llama_model_loader: - kv 12: tokenizer.ggml.pre str = qwen2 llama_model_loader: - kv 13: tokenizer.ggml.tokens arr[str,151646] = ["!", """, "#", "$", "%", "&", "'", ... llama_model_loader: - kv 14: tokenizer.ggml.token_type arr[i32,151646] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ... llama_model_loader: - kv 15: tokenizer.ggml.merges arr[str,151387] = ["Ġ Ġ", "ĠĠ ĠĠ", "i n", "Ġ t",... llama_model_loader: - kv 16: tokenizer.ggml.eos_token_id u32 = 151643 llama_model_loader: - kv 17: tokenizer.ggml.padding_token_id u32 = 151643 llama_model_loader: - kv 18: tokenizer.ggml.bos_token_id u32 = 151643 llama_model_loader: - kv 19: tokenizer.ggml.add_eos_token bool = true llama_model_loader: - kv 20: tokenizer.chat_template str = {% for message in messages %}{{'<|im_... llama_model_loader: - kv 21: general.quantization_version u32 = 2 llama_model_loader: - type f32: 141 tensors llama_model_loader: - type bf16: 198 tensors llm_load_vocab: special tokens cache size = 3. llm_load_print_meta: format = GGUF V3 (latest) llm_load_print_meta: arch = qwen2 llm_load_print_meta: vocab type = BPE llm_load_print_meta: n_vocab = 151646 llm_load_print_meta: n_merges = 151387 llm_load_print_meta: n_ctx_train = 131072 llm_load_print_meta: n_embd = 3584 llm_load_print_meta: n_head = 28 llm_load_print_meta: n_head_kv = 4 llm_load_print_meta: n_layer = 28 llm_load_print_meta: n_rot = 128 llm_load_print_meta: n_embd_head_k = 128 llm_load_print_meta: n_embd_head_v = 128 llm_load_print_meta: n_gqa = 7 llm_load_print_meta: n_embd_k_gqa = 512 llm_load_print_meta: n_embd_v_gqa = 512 llm_load_print_meta: f_norm_eps = 0.0e+00 llm_load_print_meta: f_norm_rms_eps = 1.0e-06 llm_load_print_meta: f_clamp_kqv = 0.0e+00 llm_load_print_meta: f_max_alibi_bias = 0.0e+00 llm_load_print_meta: f_logit_scale = 0.0e+00 llm_load_print_meta: n_ff = 18944 llm_load_print_meta: n_expert = 0 llm_load_print_meta: n_expert_used = 0 llm_load_print_meta: causal attn = 1 llm_load_print_meta: pooling type = 0 llm_load_print_meta: rope type = 2 llm_load_print_meta: rope scaling = linear llm_load_print_meta: freq_base_train = 1000000.0 llm_load_print_meta: freq_scale_train = 1 llm_load_print_meta: n_yarn_orig_ctx = 131072 llm_load_print_meta: rope_finetuned = unknown llm_load_print_meta: ssm_d_conv = 0 llm_load_print_meta: ssm_d_inner = 0 llm_load_print_meta: ssm_d_state = 0 llm_load_print_meta: ssm_dt_rank = 0 llm_load_print_meta: model type = ?B llm_load_print_meta: model ftype = BF16 llm_load_print_meta: model params = 7.61 B llm_load_print_meta: model size = 14.18 GiB (16.00 BPW) llm_load_print_meta: general.name = gte-Qwen2-7B-instruct llm_load_print_meta: BOS token = 151643 '<|endoftext|>' llm_load_print_meta: EOS token = 151643 '<|endoftext|>' llm_load_print_meta: PAD token = 151643 '<|endoftext|>' llm_load_print_meta: LF token = 148848 'ÄĬ' llm_load_print_meta: EOT token = 151645 '<|im_end|>' llm_load_tensors: ggml ctx size = 0.16 MiB llm_load_tensors: CPU buffer size = 14520.56 MiB ........................................................................................ llama_new_context_with_model: n_ctx = 512 llama_new_context_with_model: n_batch = 512 llama_new_context_with_model: n_ubatch = 512 llama_new_context_with_model: flash_attn = 0 llama_new_context_with_model: freq_base = 1000000.0 llama_new_context_with_model: freq_scale = 1 llama_kv_cache_init: CPU KV buffer size = 28.00 MiB llama_new_context_with_model: KV self size = 28.00 MiB, K (f16): 14.00 MiB, V (f16): 14.00 MiB llama_new_context_with_model: CPU output buffer size = 0.58 MiB llama_new_context_with_model: CPU compute buffer size = 303.18 MiB llama_new_context_with_model: graph nodes = 986 llama_new_context_with_model: graph splits = 1

system_info: n_threads = 16 / 32 | AVX = 1 | AVX_VNNI = 0 | AVX2 = 1 | AVX512 = 1 | AVX512_VBMI = 1 | AVX512_VNNI = 1 | AVX512_BF16 = 1 | FMA = 1 | NEON = 0 | SVE = 0 | ARM_FMA = 0 | F16C = 1 | FP16_VA = 0 | WASM_SIMD = 0 | BLAS = 0 | SSE3 = 1 | SSSE3 = 1 | VSX = 0 | MATMUL_INT8 = 0 | LLAMAFILE = 1 | GGML_ASSERT: examples/main/main.cpp:248: llama_add_eos_token(model) != 1 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. zsh: IOT instruction (core dumped) ./main -m models/gte-Qwen2-7B-instruct.gguf

And early observation: you are using a Bfloat16/FP32 mix, Ollama can't load Bfloat16 models, and I converted this model to a FP16/FP32 mix (and, internally, I use pure FP32 too for comparison and testing). If you use the convert-from-hf python script, try using "--outtype" option with FP16 or FP32 , the former will give you a decent mix of FP16/FP32 instead of a strictly FP16 mix.

I don't know whether the Bfloat16 issue is an Ollama limitation or a llama.cpp thing, but you can try FP16/FP32 conversion with llama.cpp for your intended use too.

<!-- gh-comment-id:2197166960 --> @rjmalagon commented on GitHub (Jun 28, 2024): > Looks like llama.cpp needs to be updated: ./main -m models/gte-Qwen2-7B-instruct.gguf Log start main: build = 3041 ([eb57fee](https://github.com/ollama/ollama/commit/eb57fee51f7b4d78039f003249873c2eb46f12f6)) main: built with cc (Ubuntu 13.2.0-23ubuntu4) 13.2.0 for x86_64-linux-gnu main: seed = 1719467592 llama_model_loader: loaded meta data with 22 key-value pairs and 339 tensors from models/gte-Qwen2-7B-instruct.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.name str = gte-Qwen2-7B-instruct llama_model_loader: - kv 2: qwen2.block_count u32 = 28 llama_model_loader: - kv 3: qwen2.context_length u32 = 131072 llama_model_loader: - kv 4: qwen2.embedding_length u32 = 3584 llama_model_loader: - kv 5: qwen2.feed_forward_length u32 = 18944 llama_model_loader: - kv 6: qwen2.attention.head_count u32 = 28 llama_model_loader: - kv 7: qwen2.attention.head_count_kv u32 = 4 llama_model_loader: - kv 8: qwen2.rope.freq_base f32 = 1000000.000000 llama_model_loader: - kv 9: qwen2.attention.layer_norm_rms_epsilon f32 = 0.000001 llama_model_loader: - kv 10: general.file_type u32 = 32 llama_model_loader: - kv 11: tokenizer.ggml.model str = gpt2 llama_model_loader: - kv 12: tokenizer.ggml.pre str = qwen2 llama_model_loader: - kv 13: tokenizer.ggml.tokens arr[str,151646] = ["!", """, "#", "$", "%", "&", "'", ... llama_model_loader: - kv 14: tokenizer.ggml.token_type arr[i32,151646] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ... llama_model_loader: - kv 15: tokenizer.ggml.merges arr[str,151387] = ["Ġ Ġ", "ĠĠ ĠĠ", "i n", "Ġ t",... llama_model_loader: - kv 16: tokenizer.ggml.eos_token_id u32 = 151643 llama_model_loader: - kv 17: tokenizer.ggml.padding_token_id u32 = 151643 llama_model_loader: - kv 18: tokenizer.ggml.bos_token_id u32 = 151643 llama_model_loader: - kv 19: tokenizer.ggml.add_eos_token bool = true llama_model_loader: - kv 20: tokenizer.chat_template str = {% for message in messages %}{{'<|im_... llama_model_loader: - kv 21: general.quantization_version u32 = 2 llama_model_loader: - type f32: 141 tensors llama_model_loader: - type bf16: 198 tensors llm_load_vocab: special tokens cache size = 3. llm_load_print_meta: format = GGUF V3 (latest) llm_load_print_meta: arch = qwen2 llm_load_print_meta: vocab type = BPE llm_load_print_meta: n_vocab = 151646 llm_load_print_meta: n_merges = 151387 llm_load_print_meta: n_ctx_train = 131072 llm_load_print_meta: n_embd = 3584 llm_load_print_meta: n_head = 28 llm_load_print_meta: n_head_kv = 4 llm_load_print_meta: n_layer = 28 llm_load_print_meta: n_rot = 128 llm_load_print_meta: n_embd_head_k = 128 llm_load_print_meta: n_embd_head_v = 128 llm_load_print_meta: n_gqa = 7 llm_load_print_meta: n_embd_k_gqa = 512 llm_load_print_meta: n_embd_v_gqa = 512 llm_load_print_meta: f_norm_eps = 0.0e+00 llm_load_print_meta: f_norm_rms_eps = 1.0e-06 llm_load_print_meta: f_clamp_kqv = 0.0e+00 llm_load_print_meta: f_max_alibi_bias = 0.0e+00 llm_load_print_meta: f_logit_scale = 0.0e+00 llm_load_print_meta: n_ff = 18944 llm_load_print_meta: n_expert = 0 llm_load_print_meta: n_expert_used = 0 llm_load_print_meta: causal attn = 1 llm_load_print_meta: pooling type = 0 llm_load_print_meta: rope type = 2 llm_load_print_meta: rope scaling = linear llm_load_print_meta: freq_base_train = 1000000.0 llm_load_print_meta: freq_scale_train = 1 llm_load_print_meta: n_yarn_orig_ctx = 131072 llm_load_print_meta: rope_finetuned = unknown llm_load_print_meta: ssm_d_conv = 0 llm_load_print_meta: ssm_d_inner = 0 llm_load_print_meta: ssm_d_state = 0 llm_load_print_meta: ssm_dt_rank = 0 llm_load_print_meta: model type = ?B llm_load_print_meta: model ftype = BF16 llm_load_print_meta: model params = 7.61 B llm_load_print_meta: model size = 14.18 GiB (16.00 BPW) llm_load_print_meta: general.name = gte-Qwen2-7B-instruct llm_load_print_meta: BOS token = 151643 '<|endoftext|>' llm_load_print_meta: EOS token = 151643 '<|endoftext|>' llm_load_print_meta: PAD token = 151643 '<|endoftext|>' llm_load_print_meta: LF token = 148848 'ÄĬ' llm_load_print_meta: EOT token = 151645 '<|im_end|>' llm_load_tensors: ggml ctx size = 0.16 MiB llm_load_tensors: CPU buffer size = 14520.56 MiB ........................................................................................ llama_new_context_with_model: n_ctx = 512 llama_new_context_with_model: n_batch = 512 llama_new_context_with_model: n_ubatch = 512 llama_new_context_with_model: flash_attn = 0 llama_new_context_with_model: freq_base = 1000000.0 llama_new_context_with_model: freq_scale = 1 llama_kv_cache_init: CPU KV buffer size = 28.00 MiB llama_new_context_with_model: KV self size = 28.00 MiB, K (f16): 14.00 MiB, V (f16): 14.00 MiB llama_new_context_with_model: CPU output buffer size = 0.58 MiB llama_new_context_with_model: CPU compute buffer size = 303.18 MiB llama_new_context_with_model: graph nodes = 986 llama_new_context_with_model: graph splits = 1 > > system_info: n_threads = 16 / 32 | AVX = 1 | AVX_VNNI = 0 | AVX2 = 1 | AVX512 = 1 | AVX512_VBMI = 1 | AVX512_VNNI = 1 | AVX512_BF16 = 1 | FMA = 1 | NEON = 0 | SVE = 0 | ARM_FMA = 0 | F16C = 1 | FP16_VA = 0 | WASM_SIMD = 0 | BLAS = 0 | SSE3 = 1 | SSSE3 = 1 | VSX = 0 | MATMUL_INT8 = 0 | LLAMAFILE = 1 | GGML_ASSERT: examples/main/main.cpp:248: llama_add_eos_token(model) != 1 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. zsh: IOT instruction (core dumped) ./main -m models/gte-Qwen2-7B-instruct.gguf > Looks like llama.cpp needs to be updated: ./main -m models/gte-Qwen2-7B-instruct.gguf Log start main: build = 3041 ([eb57fee](https://github.com/ollama/ollama/commit/eb57fee51f7b4d78039f003249873c2eb46f12f6)) main: built with cc (Ubuntu 13.2.0-23ubuntu4) 13.2.0 for x86_64-linux-gnu main: seed = 1719467592 llama_model_loader: loaded meta data with 22 key-value pairs and 339 tensors from models/gte-Qwen2-7B-instruct.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.name str = gte-Qwen2-7B-instruct llama_model_loader: - kv 2: qwen2.block_count u32 = 28 llama_model_loader: - kv 3: qwen2.context_length u32 = 131072 llama_model_loader: - kv 4: qwen2.embedding_length u32 = 3584 llama_model_loader: - kv 5: qwen2.feed_forward_length u32 = 18944 llama_model_loader: - kv 6: qwen2.attention.head_count u32 = 28 llama_model_loader: - kv 7: qwen2.attention.head_count_kv u32 = 4 llama_model_loader: - kv 8: qwen2.rope.freq_base f32 = 1000000.000000 llama_model_loader: - kv 9: qwen2.attention.layer_norm_rms_epsilon f32 = 0.000001 llama_model_loader: - kv 10: general.file_type u32 = 32 llama_model_loader: - kv 11: tokenizer.ggml.model str = gpt2 llama_model_loader: - kv 12: tokenizer.ggml.pre str = qwen2 llama_model_loader: - kv 13: tokenizer.ggml.tokens arr[str,151646] = ["!", """, "#", "$", "%", "&", "'", ... llama_model_loader: - kv 14: tokenizer.ggml.token_type arr[i32,151646] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ... llama_model_loader: - kv 15: tokenizer.ggml.merges arr[str,151387] = ["Ġ Ġ", "ĠĠ ĠĠ", "i n", "Ġ t",... llama_model_loader: - kv 16: tokenizer.ggml.eos_token_id u32 = 151643 llama_model_loader: - kv 17: tokenizer.ggml.padding_token_id u32 = 151643 llama_model_loader: - kv 18: tokenizer.ggml.bos_token_id u32 = 151643 llama_model_loader: - kv 19: tokenizer.ggml.add_eos_token bool = true llama_model_loader: - kv 20: tokenizer.chat_template str = {% for message in messages %}{{'<|im_... llama_model_loader: - kv 21: general.quantization_version u32 = 2 llama_model_loader: - type f32: 141 tensors llama_model_loader: - type bf16: 198 tensors llm_load_vocab: special tokens cache size = 3. llm_load_print_meta: format = GGUF V3 (latest) llm_load_print_meta: arch = qwen2 llm_load_print_meta: vocab type = BPE llm_load_print_meta: n_vocab = 151646 llm_load_print_meta: n_merges = 151387 llm_load_print_meta: n_ctx_train = 131072 llm_load_print_meta: n_embd = 3584 llm_load_print_meta: n_head = 28 llm_load_print_meta: n_head_kv = 4 llm_load_print_meta: n_layer = 28 llm_load_print_meta: n_rot = 128 llm_load_print_meta: n_embd_head_k = 128 llm_load_print_meta: n_embd_head_v = 128 llm_load_print_meta: n_gqa = 7 llm_load_print_meta: n_embd_k_gqa = 512 llm_load_print_meta: n_embd_v_gqa = 512 llm_load_print_meta: f_norm_eps = 0.0e+00 llm_load_print_meta: f_norm_rms_eps = 1.0e-06 llm_load_print_meta: f_clamp_kqv = 0.0e+00 llm_load_print_meta: f_max_alibi_bias = 0.0e+00 llm_load_print_meta: f_logit_scale = 0.0e+00 llm_load_print_meta: n_ff = 18944 llm_load_print_meta: n_expert = 0 llm_load_print_meta: n_expert_used = 0 llm_load_print_meta: causal attn = 1 llm_load_print_meta: pooling type = 0 llm_load_print_meta: rope type = 2 llm_load_print_meta: rope scaling = linear llm_load_print_meta: freq_base_train = 1000000.0 llm_load_print_meta: freq_scale_train = 1 llm_load_print_meta: n_yarn_orig_ctx = 131072 llm_load_print_meta: rope_finetuned = unknown llm_load_print_meta: ssm_d_conv = 0 llm_load_print_meta: ssm_d_inner = 0 llm_load_print_meta: ssm_d_state = 0 llm_load_print_meta: ssm_dt_rank = 0 llm_load_print_meta: model type = ?B llm_load_print_meta: model ftype = BF16 llm_load_print_meta: model params = 7.61 B llm_load_print_meta: model size = 14.18 GiB (16.00 BPW) llm_load_print_meta: general.name = gte-Qwen2-7B-instruct llm_load_print_meta: BOS token = 151643 '<|endoftext|>' llm_load_print_meta: EOS token = 151643 '<|endoftext|>' llm_load_print_meta: PAD token = 151643 '<|endoftext|>' llm_load_print_meta: LF token = 148848 'ÄĬ' llm_load_print_meta: EOT token = 151645 '<|im_end|>' llm_load_tensors: ggml ctx size = 0.16 MiB llm_load_tensors: CPU buffer size = 14520.56 MiB ........................................................................................ llama_new_context_with_model: n_ctx = 512 llama_new_context_with_model: n_batch = 512 llama_new_context_with_model: n_ubatch = 512 llama_new_context_with_model: flash_attn = 0 llama_new_context_with_model: freq_base = 1000000.0 llama_new_context_with_model: freq_scale = 1 llama_kv_cache_init: CPU KV buffer size = 28.00 MiB llama_new_context_with_model: KV self size = 28.00 MiB, K (f16): 14.00 MiB, V (f16): 14.00 MiB llama_new_context_with_model: CPU output buffer size = 0.58 MiB llama_new_context_with_model: CPU compute buffer size = 303.18 MiB llama_new_context_with_model: graph nodes = 986 llama_new_context_with_model: graph splits = 1 > > system_info: n_threads = 16 / 32 | AVX = 1 | AVX_VNNI = 0 | AVX2 = 1 | AVX512 = 1 | AVX512_VBMI = 1 | AVX512_VNNI = 1 | AVX512_BF16 = 1 | FMA = 1 | NEON = 0 | SVE = 0 | ARM_FMA = 0 | F16C = 1 | FP16_VA = 0 | WASM_SIMD = 0 | BLAS = 0 | SSE3 = 1 | SSSE3 = 1 | VSX = 0 | MATMUL_INT8 = 0 | LLAMAFILE = 1 | GGML_ASSERT: examples/main/main.cpp:248: llama_add_eos_token(model) != 1 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. zsh: IOT instruction (core dumped) ./main -m models/gte-Qwen2-7B-instruct.gguf > Looks like llama.cpp needs to be updated: ./main -m models/gte-Qwen2-7B-instruct.gguf Log start main: build = 3041 ([eb57fee](https://github.com/ollama/ollama/commit/eb57fee51f7b4d78039f003249873c2eb46f12f6)) main: built with cc (Ubuntu 13.2.0-23ubuntu4) 13.2.0 for x86_64-linux-gnu main: seed = 1719467592 llama_model_loader: loaded meta data with 22 key-value pairs and 339 tensors from models/gte-Qwen2-7B-instruct.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.name str = gte-Qwen2-7B-instruct llama_model_loader: - kv 2: qwen2.block_count u32 = 28 llama_model_loader: - kv 3: qwen2.context_length u32 = 131072 llama_model_loader: - kv 4: qwen2.embedding_length u32 = 3584 llama_model_loader: - kv 5: qwen2.feed_forward_length u32 = 18944 llama_model_loader: - kv 6: qwen2.attention.head_count u32 = 28 llama_model_loader: - kv 7: qwen2.attention.head_count_kv u32 = 4 llama_model_loader: - kv 8: qwen2.rope.freq_base f32 = 1000000.000000 llama_model_loader: - kv 9: qwen2.attention.layer_norm_rms_epsilon f32 = 0.000001 llama_model_loader: - kv 10: general.file_type u32 = 32 llama_model_loader: - kv 11: tokenizer.ggml.model str = gpt2 llama_model_loader: - kv 12: tokenizer.ggml.pre str = qwen2 llama_model_loader: - kv 13: tokenizer.ggml.tokens arr[str,151646] = ["!", """, "#", "$", "%", "&", "'", ... llama_model_loader: - kv 14: tokenizer.ggml.token_type arr[i32,151646] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ... llama_model_loader: - kv 15: tokenizer.ggml.merges arr[str,151387] = ["Ġ Ġ", "ĠĠ ĠĠ", "i n", "Ġ t",... llama_model_loader: - kv 16: tokenizer.ggml.eos_token_id u32 = 151643 llama_model_loader: - kv 17: tokenizer.ggml.padding_token_id u32 = 151643 llama_model_loader: - kv 18: tokenizer.ggml.bos_token_id u32 = 151643 llama_model_loader: - kv 19: tokenizer.ggml.add_eos_token bool = true llama_model_loader: - kv 20: tokenizer.chat_template str = {% for message in messages %}{{'<|im_... llama_model_loader: - kv 21: general.quantization_version u32 = 2 llama_model_loader: - type f32: 141 tensors llama_model_loader: - type bf16: 198 tensors llm_load_vocab: special tokens cache size = 3. llm_load_print_meta: format = GGUF V3 (latest) llm_load_print_meta: arch = qwen2 llm_load_print_meta: vocab type = BPE llm_load_print_meta: n_vocab = 151646 llm_load_print_meta: n_merges = 151387 llm_load_print_meta: n_ctx_train = 131072 llm_load_print_meta: n_embd = 3584 llm_load_print_meta: n_head = 28 llm_load_print_meta: n_head_kv = 4 llm_load_print_meta: n_layer = 28 llm_load_print_meta: n_rot = 128 llm_load_print_meta: n_embd_head_k = 128 llm_load_print_meta: n_embd_head_v = 128 llm_load_print_meta: n_gqa = 7 llm_load_print_meta: n_embd_k_gqa = 512 llm_load_print_meta: n_embd_v_gqa = 512 llm_load_print_meta: f_norm_eps = 0.0e+00 llm_load_print_meta: f_norm_rms_eps = 1.0e-06 llm_load_print_meta: f_clamp_kqv = 0.0e+00 llm_load_print_meta: f_max_alibi_bias = 0.0e+00 llm_load_print_meta: f_logit_scale = 0.0e+00 llm_load_print_meta: n_ff = 18944 llm_load_print_meta: n_expert = 0 llm_load_print_meta: n_expert_used = 0 llm_load_print_meta: causal attn = 1 llm_load_print_meta: pooling type = 0 llm_load_print_meta: rope type = 2 llm_load_print_meta: rope scaling = linear llm_load_print_meta: freq_base_train = 1000000.0 llm_load_print_meta: freq_scale_train = 1 llm_load_print_meta: n_yarn_orig_ctx = 131072 llm_load_print_meta: rope_finetuned = unknown llm_load_print_meta: ssm_d_conv = 0 llm_load_print_meta: ssm_d_inner = 0 llm_load_print_meta: ssm_d_state = 0 llm_load_print_meta: ssm_dt_rank = 0 llm_load_print_meta: model type = ?B llm_load_print_meta: model ftype = BF16 llm_load_print_meta: model params = 7.61 B llm_load_print_meta: model size = 14.18 GiB (16.00 BPW) llm_load_print_meta: general.name = gte-Qwen2-7B-instruct llm_load_print_meta: BOS token = 151643 '<|endoftext|>' llm_load_print_meta: EOS token = 151643 '<|endoftext|>' llm_load_print_meta: PAD token = 151643 '<|endoftext|>' llm_load_print_meta: LF token = 148848 'ÄĬ' llm_load_print_meta: EOT token = 151645 '<|im_end|>' llm_load_tensors: ggml ctx size = 0.16 MiB llm_load_tensors: CPU buffer size = 14520.56 MiB ........................................................................................ llama_new_context_with_model: n_ctx = 512 llama_new_context_with_model: n_batch = 512 llama_new_context_with_model: n_ubatch = 512 llama_new_context_with_model: flash_attn = 0 llama_new_context_with_model: freq_base = 1000000.0 llama_new_context_with_model: freq_scale = 1 llama_kv_cache_init: CPU KV buffer size = 28.00 MiB llama_new_context_with_model: KV self size = 28.00 MiB, K (f16): 14.00 MiB, V (f16): 14.00 MiB llama_new_context_with_model: CPU output buffer size = 0.58 MiB llama_new_context_with_model: CPU compute buffer size = 303.18 MiB llama_new_context_with_model: graph nodes = 986 llama_new_context_with_model: graph splits = 1 > > system_info: n_threads = 16 / 32 | AVX = 1 | AVX_VNNI = 0 | AVX2 = 1 | AVX512 = 1 | AVX512_VBMI = 1 | AVX512_VNNI = 1 | AVX512_BF16 = 1 | FMA = 1 | NEON = 0 | SVE = 0 | ARM_FMA = 0 | F16C = 1 | FP16_VA = 0 | WASM_SIMD = 0 | BLAS = 0 | SSE3 = 1 | SSSE3 = 1 | VSX = 0 | MATMUL_INT8 = 0 | LLAMAFILE = 1 | GGML_ASSERT: examples/main/main.cpp:248: llama_add_eos_token(model) != 1 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. zsh: IOT instruction (core dumped) ./main -m models/gte-Qwen2-7B-instruct.gguf And early observation: you are using a Bfloat16/FP32 mix, Ollama can't load Bfloat16 models, and I converted this model to a FP16/FP32 mix (and, internally, I use pure FP32 too for comparison and testing). If you use the convert-from-hf python script, try using "--outtype" option with FP16 or FP32 , the former will give you a decent mix of FP16/FP32 instead of a strictly FP16 mix. I don't know whether the Bfloat16 issue is an Ollama limitation or a llama.cpp thing, but you can try FP16/FP32 conversion with llama.cpp for your intended use too.
GiteaMirror commented 2026-04-12 13:43:18 -05:00
Author
Owner
Copy Link

@waters222 commented on GitHub (Jul 9, 2024):

@rjmalagon I tried your model output. its not same as HF output at the moment.
I think there is something wrong about either the implementation from qwen2 side or certain areas for mix precision calcualtion.

<!-- gh-comment-id:2218460461 --> @waters222 commented on GitHub (Jul 9, 2024): @rjmalagon I tried your model output. its not same as HF output at the moment. I think there is something wrong about either the implementation from qwen2 side or certain areas for mix precision calcualtion.
GiteaMirror commented 2026-04-12 13:43:19 -05:00
Author
Owner
Copy Link

@whisper-bye commented on GitHub (Mar 15, 2025):

+1

<!-- gh-comment-id:2726318388 --> @whisper-bye commented on GitHub (Mar 15, 2025): +1
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 model
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#3216
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?